Marriett G - Methods in Enzimology v291 Caged Compounds MIE-AP (1998)

Preface
Caged compounds are biomolecules whose activity or function is

inactivated by chemical modification with a photolabile group. Excitation
of a caged compound with (near-ultraviolet) light results in specific
cleavage of a bond within the molecule that rapidly liberates the functional
biomolecule of interest. One main advantage of this light-directed pertur-
bation technique centers around its ability to overcome the mixing and
diffusion time limits of fast-flow methods. This is achieved by precharging
the biological sample with a caged substrate, ligand, or protein conjugate,
followed by light-directed activation of the caged compound which triggers
the specific reaction under investigation within the excitation volume.
In addition, concentration jumps mediated by light-directed activation
of caged compounds can be conducted without physically disturbing the
preparation. Caged groups described in this volume generate active
biomolecules within nanoseconds of absorbing light and, furthermore,
by using two-photon excitation of the caged compound, the concentration
jump can be confined to a volume having a diameter of approximately
250 nm with little or no irradiation-based damage to the preparation.
Light-directed activation of caged compounds and rapid monitoring of
the ensuing reaction using photomultiplier or imaging-based techniques
have been used to understand the molecular mechanism of several
biochemical reactions and processes in vitro and in vivo. This volume
also describes the application of caged compounds to studies that aim
to understand the molecular basis of neurotransmission, muscle contrac-
tion, and ion channel function.
Topics covered in this book include (a) an up-to-date account of the
synthetic chemistry and photochemistry of traditional and newer genera-
tions of caged groups, (b) a description of new reagents and methods to
prepare monofunctional and bifunctional complexes of a diverse array of
caged substrates, ligands, and polypeptides, (c) a state-of-the-art practical
and theoretical description of lamp- and laser-based excitation sources for
light-directed activation of caged compounds in macro- and microscopic
samples, (d) the measurement and interpretation of spectroscopic signals
derived from light-directed activation of caged compounds to monitor the
kinetics of the reaction being studied, and (e) the application of light-
directed activation of caged compounds and caged polypeptides to chemical
relaxation studies that aim to understand the molecular mechanism of
protein-catalyzed reactions in a diverse array of complex molecular environ-
ments.
xiii
xiv PREFACE
The primary objective of this volume is to provide newcomers to the

field with a central source of valuable practical and theoretical information
on caged compounds that will allow them to use light-directed activation
of caged compounds to address molecular mechanisms of biochemical reac-
tions or processes within biological samples. It provides the investigator
with the most up-to-date and comprehensive account of caged compounds
that should aid in the design of experiments to compare and interpret data
obtained from kinetic measurements of specific reactions triggered by light-
directed activation of caged biomolecules.
This volume is dedicated to the memory of Professor Fredric S. Fay.
GERARD MARRIOTT
Fredric Stewart Fay
1943-1997
The scientific community lost one of its most gifted and productive
members on March 18, 1997 when Fred Fay suffered a fatal heart attack.
Fred was born in New York City and was educated in its public school
system. On graduation from the Bronx High School of Science in 1961, he
enrolled at Cornell University where he majored in chemistry and was
elected to Phi Beta Kappa. His love and talent for science were evident
early, not only in the schoolboy experiments conducted in the family apart-
ment, but also in the summers of his undergraduate years when he worked
as a lab assistant at the Sloane-Kettering Institute. Fred began his formal
scientific career in his junior year at Cornell when he joined the lab of Dr.
R. Blake Reeves. With Dr. Reeves Fred first studied the effects of low
barometric pressure on the synthesis of a myoglobin-like oxygen-binding
pigment in a marine arthropod. This was followed by a study of B12 uptake
in a protozoan, which led to Fred's first scientific publication [Reeves,
xxxiii
xxxiv FREDRIC STEWART FAY
R.B. and Fay, F.S. Cyanocobalamin (vitamin B12) uptake by Ochromonas

malhamensis. Am. J. Physiol. 210:1273-1278, 1966]. Reeves had recently
arrived at Cornell after completing his doctoral and postdoctoral training
in the Physiology Department at Harvard Medical School. As is so often
the case, Reeves referred his star pupil to his mentor, Dr. John R. Pappen-
heimer, and on graduation from Cornell, Fred enrolled in the Medical
Sciences Program at Harvard to begin his doctoral studies in physiology.
Pappenheimer had long-standing interests in the regulation of respira-
tion and in oxygen-sensing mechanisms, and at the time was studying the
contribution of the ionic composition of cerebrospinal fluid to the regulation
of respiration in goats. It was quite natural in this environment for Fred's
interest to be tweaked by the question of how tissues sense and respond
to changes in oxygen tension. He examined the anomalously high rate of
oxygen consumption in the carotid body for clues to its oxygen-sensing
function. These studies provided the first accurate data on oxygen consump-
tion by this tissue, which, in fact, consumed oxygen at less than 20% of the
previously accepted rate. To complete his thesis project he went on to study
contractile responses of the guinea pig ductus arteriosis to increased oxygen
availability. This latter research kindled his interest in the physiology of
smooth muscle, which was to play a central role in his research career.
These early studies also foreshadowed other themes that became prominent
in all of his subsequent work. Fred had a need to see what was happening
in his experiments. His thesis and the resulting papers beautifully portrayed
the morphology of his preparations at the macroscopic and microscopic
levels. Studies of both the carotid body and the ductus arteriosus required
the custom design and fabrication of suitable apparatus and carried the
imprimatur of the inventiveness and skill that characterized all of his
later work.
After receiving his PhD in 1969, Fred stayed on at Harvard to continue
work on the guinea pig ductus arteriosus. These studies showed all of the
elegance and clarity of thought that were to become the hallmark of Fred's
science. He found that tension in the ductal wall increased in response to
increased partial pressure of oxygen just as well when oxygen was present
only in the perfusate within the vessel or when present only in the perifusate
bathing the outer adventitial surface that contained nerve endings and mast
cells. Because equal but oppositely directed gradients of Po2 produced
equivalent contractile responses, he concluded that the smooth muscle itself
acted as the oxygen sensor and contracted in a stepwise fashion (Fay, F.S.
Guinea pig ductus arteriosus I. The cellular and metabolic basis for oxygen
sensitivity. Am. J. Physiol. 221:470-479, 1971). He later showed that the
trigger for this response is a unique cytochrome and demonstrated that the
ductus could be made to contract by light of the appropriate wavelength
if the cytochrome had first been inactivated by carbon monoxide in the
FREDRIC STEWART FAY XXXV
dark. Virtually all of the themes that were to characterize his later work were
already germinating in these studies: the morphologic studies, particularly
associated with the irreversibility of ductal closure and the questions raised
with regard to coupling between metabolism and electrical activity and
contraction, and the stepwise contractile behavior of "a population of
smooth muscle cells of the single-unit type."
Meanwhile, 40 miles to the west, final plans were taking shape to open
the University of Massachusetts Medical School. Fred was recruited as an
assistant professor of physiology, and was in Worcester to greet the first
class of medical students when they arrived in September 1970. He was
intimately involved in all of the planning, organizing, and just plain hard
work that goes into starting a new department. Much of the credit for
whatever success the department now enjoys must go to him. With charac-
teristic energy and drive Fred immediately set up shop in a small lab in
the converted tobacco warehouse that was to be the temporary home of
the medical school for the next 3½years. Here he continued his studies on
the ductus arteriosus and on the morphologic basis for contractility of
smooth muscle in collaboration with Peter Cooke. He also began to lay
the groundwork for in-depth studies of smooth muscle physiology.
At that time understanding of how smooth muscle contracts was meager
at best. All that was known was derived from studies of multicellular
preparations with all of the complexities attributable to heterogeneity of
cells and extracellular matrices. Fred began to rectify that situation with
characteristic insight and boundless energy. He set about trying to develop
a homogeneous population of isolated muscle cells to define the contractile
properties of smooth muscle cells themselves. His first attempts were with
the intestinal muscles of a large salamander, Amphiuma sp., which had
unusually large smooth muscle cells. Unfortunately, obtaining a dispersed
cell preparation that maintained normal contractility was no simple matter.
As he struggled with Amphiuma, Bagby and colleagues (Bagby, R., Young,
A., Dotson, R., Fisher, B., and McKinnon, K. Contraction of single smooth
muscle cells from Bufo marinus stomach. Nature 234:351-352, 1971) pub-
lished their successful preparation of isolated smooth muscle cells from the
stomach of Bufo marinus. Fred was quick to exploit this preparation and
undertook what are now his classic studies of smooth muscle physiology.
If there is any one aspect of Fred's career that stands out, it was his
use of the single smooth muscle cell preparation. With it, Fred explored
virtually every aspect of smooth muscle physiology and molecular anatomy
at the cellular and subcellular level. Fundamental contributions were made
concerning the organization of the contractile apparatus and how it is
turned on and off in response to various physiologic signals. This unique
preparation was not only a model system for smooth muscle, but also a
proving ground for Fred's views of cell biology. Working with a medical
xxxvi FREDRIC STEWART FAY
student, Claudio Delise, Fred showed with scanning electron microscopic

images that when a smooth muscle contracts, large "blebs" appear on the
plasma membrane, and proposed that the contractile machinery attaches
at specialized regions of the plasma membrane. The now classic article
describing these findings (Fay, F.S. and Delise, C.M. Contraction of isolated
smooth muscle cells--structural changes. Proc. Natl. Acad. Sci. U.S.A.
70=641-645, 1973) was his third most cited paper.
With another medical student, Peter Canaday, he devised an ultrasensi-
tive force transducer. He then developed the techniques for attaching a
single smooth muscle cell to the transducer and recorded tensions of a
few micrograms when the cell contracted (Fay, F.S. Isometric contractile
properties of single isolated smooth muscle cells. Nature 265:553-556, 1977).
With this new technology Fred showed that, in contrast to striated muscle
or the multifiber smooth muscle preparations used by all previous investiga-
tors, the tension of a relaxed smooth muscle cell is virtually independent
of its length. These studies highlighted the major contribution of the connec-
tive tissue elements to the mechanical properties of smooth muscle tissues
and thus confirmed the critical importance of studying isolated smooth
muscle cells for the understanding of smooth muscle physiology. In these
studies, Fred also made the unique observation that there was a delay
between the excitatory stimulus and the onset of force generation. He
suggested that the delay was due either (1) to a delay in coupling of the
changes in membrane potential to the change in intracellular Ca2+; or, (2)
that Ca 2+ is not rate limiting and that CaZ+-dependent smooth muscle
myosin regulatory processes are slow.
The importance of being able to measure intracellular calcium concen-
trations ([Ca2+]i) was evident, and when he became eligible for his first
sabbatical leave, Fred decided to work with Dr. Stuart Taylor at the
Mayo Medical School to learn how to measure intracellular Ca 2+ using
aequorin as a calcium indicator (Fay, F.S., Shlevin, H.H., Granger, W.C.,
and Taylor, S.R. Aequorin luminescence during activation of single
isolated smooth muscle cells. Nature 280:506-508, 1979). He microinjected
aequorin into individual smooth muscle cells and discovered that on
stimulation [Ca2+]i rises rapidly and remains above basal levels. Therefore,
Ca 2+ delivery to the regulatory proteins is not rate limiting. Rather, it
is the myosin light-chain kinase regulatory pathway that is relatively slow,
thus generating the observed delay between stimulus and contraction. This
finding inevitably generated the next question: What Ca 2+ handling
processes might govern the decay of [Ca2+]i? Its answer would have to
wait for the development of new techniques, and he would revisit this
concept twenty years later.
Fred loved to hypothesize about molecular mechanisms. He loved to
look at a data set and turn it around and around in his mind to uncover
FREDRIC STEWART FAY xxxvii
every facet. He would challenge his colleagues or fellows to come up with

every possibility that might explain the data and then rigorously weigh every
potential explanation until only the one or two most plausible survived. He
was not shy in going out on a limb to propose unique mechanisms. A
good example of this is seen in his second most cited paper (Scheid, C.R.,
Honeyman, T.W., and Fay, F.S. Mechanism of/3-adrenergic relaxation of
smooth muscle. Nature 277:32-36, 1979). Fred and his co-workers proposed
that/3-adrenergic relaxation of smooth muscle cells was a result of cAMP-
dependent stimulation of the Na+,K+-ATPase pump, which, in lowering
intracellular sodium would, in turn, stimulate the Na+/Ca 2+ exchanger, with
the final outcome being a reduction in [Ca2+]i. It was not until 14 years
later, and after his lab had developed the digital imaging microscope (DIM),
that Fred was able to expand on this idea. Fred reasoned that if/3-adrenergic
relaxation involved both Na+,K+-ATPase activity coupled to the Na+/Ca 2+
exchanger, these two entities might be physically colocalized within the
smooth muscle cell. By pushing fluorescence detection coupled to the digital
imaging microscope to its technologic maximum, Fred and his colleagues
also pushed back the limits of spatial resolution. Using immunohistochemi-
cal methods, they demonstrated that the Na +,K+-ATPase and the Na +/Ca 2+
exchanger were indeed colocalized in the plasma membrane and were
restricted to regions that were distinct from those membrane regions that
were specialized for mechanical force transmission (Moore, E.D.W., Etter,
E.F., Phillipson, K.D., Carrington, W.A., Fogarty, K.E., Lifshitz, L. M.,
and Fay, F.S. Coupling of the Na+/Ca 2+ exchanger, Na+/K ÷ pump, and
sarcoplasmic reticulum in smooth muscle. Nature 365:657-660, 1993). With
McCarron and Walsh, Fred then went on to show that Na+/Ca 2+ exchange
regulates cytoplasmic calcium in smooth muscle.
Efforts to understand how smooth muscle contracts had been hampered
by the lack of any technology that would allow Fred to look inside the
cells and see what was happening molecule by molecule. He therefore set
out to devise optical instruments to do just that. The development of the
DIM was critical to the next steps in Fred's scientific career. He assembled
an extraordinary team of physicists, mathematicians, and computer scien-
tists who together created powerful and innovative optical technology that
placed the Imaging Group at the University of Massachusetts at the fore-
front of biomedical imaging at the cellular and subcellular levels. Since the
late 1980s, the digital imaging microscope has evolved to keep pace with
the complexity of the scientific problems to which it has been applied. For
example, the studies of the distribution of the Na+,K+-ATPase and the
Na+/Ca 2+ exchanger required three-dimensional images of immunolabeled
fixed cells. A single data set of multiple focal planes within a cell was taken
over a period of several minutes. Such a leisurely time course clearly was
inadequate to tackle the rapid dynamic changes that occur in living cells.
XXXVlll F R E D R I C S T E W A R T FAY
In the early 1990s a high-speed instrument was developed that used laser
illumination, high-speed piezo electric focus, and a conventional scientific
CCD camera partially masked so that different portions of its imaging area
could be used to take images a few milliseconds apart. This instrument
was used to measure changes in the membrane potential of individual
mitochondria that were in motion at speeds of up to 1 micron per second
in neuroblastoma cells (Loew, L.M., Tuft, R.A., Carrington, W.A., and
Fay, F.S. Imaging in five dimensions: time-dependent membrane potentials
in individual mitochondria. Biophys. J. 65:2396-2407, 1993) and was fast
enough to permit the study of Ca 2÷ gradients in the first 15 milliseconds
of systole in cardiac myocytes (Isenberg, G., Etter, E.F., Wendt-Galliteri,
M., Schiefer, A., Carrington, W.A., Tuft, R.A., and Fay, F.S. Intrasarcomere
[Ca2+] gradients in ventricular myocytes revealed by high-speed digital
imaging microscopy. Proc. Natl. Acad. Sci. U.S.A. 93:5413-5418, 1996).
Even greater speed has been achieved with a high-speed CCD camera
developed for the Strategic Defense Department, and made possible Fred's
last major work in which he studied the relationship between calcium
"sparks," which are localized quantal releases of Ca 2+, and spontaneous
transient outward potassium currents [Kirber, M.T., Etter, E.F., Singer,
J.J., Walsh, J.V., Jr., and Fay, F.S. Simultaneous 3D imaging of Ca 2÷ sparks
and ionic currents in single smooth muscle cells. Biophys. J. 72:(2)part 2
A295, 1997].
Fred's involvement in his science was far reaching. Although a skilled
machinist was available, if a custom part for the microscope had to be made
quickly, Fred would use the milling machine himself to build the part. He,
as the biologist, did not simply delegate the application of complex physical
and mathematical concepts to his team of experts, but remained an inter-
active contributor to the incorporation of these ideas into workable instru-
ments. Inherent in the workings of the digital imaging microscope was the
ability to account for fluorescence from out-of-focus optical planes. This
was no small feat. Fred thoroughly understood the physics and mathematics
behind these sophisticated deblurring techniques. His most trusted refer-
ence in this regard was the "Feynmann Lectures."
Fred was among the earliest investigators to avail themselves of Fura-2
as a Ca 2+ indicator. He quickly marshaled the resources to obtain the
necessary equipment to make ratiometric fluorescent measurements in his
isolated smooth muscle cells. Before long he had the requisite technology
incorporated into the DIM. These technical advances led to Fred's most
cited paper, now an official citation classic (Williams, D.A., Fogart, K.E.,
Tsien, R.Y., and Fay, F.S. Calcium gradients in single smooth muscle cells
revealed by the digital imaging microscope using Fura-2. Nature 318:558-
561, 1985). However, Fred was not satisfied with global measurements of
Ca 2+. He developed the ability to monitor intracellular calcium with both
FREDRIC STEWART FAY xxxix
high spatial and temporal resolution. With this technique, Fred and his
fellows demonstrated that Ca 2÷ gradients existed within the cell with higher
[Ca2÷]i within the nucleus and at the sarcolemma, presumably due to Ca 2÷
stores within the sarcoplasmic reticulum. Fred demonstrated that cyto-
plasmic Ca 2÷ was about 140 nM under resting conditions and about 700-800
nM under activated conditions. These data indicated the exquisite sensitivity
of the contractile apparatus for Ca 2÷. The specialization of the membrane
for calcium handling suggested that there may be significant differences
between measurements of global cellular calcium compared to those close
to sites of calcium release and storage. This is most evident in Fred's
most recent work in which he utilized membrane-bound calcium indicators
(Etter, E.F., Minta, A., Poenie, M., and Fay, F.S. Near-membrane [Ca 2÷]
transients resolved using the Ca 2÷ indicator FFP18. Proc. Natl. Acad. Sci.
U.S.A. 93:5368-5373, 1996), which were capable of reporting local calcium
changes adjacent to the plasma membrane.
Fred contributed not only to our understanding of the contractile ma-
chinery, but also of the regulatory processes that govern smooth muscle
contraction. A major question in this field was whether activation in smooth
muscle contained a calcium-dependent actin filament-linked regulatory
system analogous to that in striated muscle. In a novel experiment, Fred
and co-workers microinjected the catalytic subunit of myosin light-chain
kinase into single smooth muscle cells and observed that, in the absence
of changes in intracellular calcium, phosphorylation of the regulatory light
chain on smooth muscle myosin was sufficient to elicit a contraction. These
data refuted the necessity for an actin-linked regulatory system but did not
rule out a possible modulatory role for such a system (Itoh, T., lkebe, M.,
Kargacin, G.J., Hartshorne, D.J., Kemp, B.E., and Fay, F.S. Effects of
modulators of myosin light-chain kinase activity in single smooth muscle
cells. Nature 338:164-167, 1989).
Fred was a renaissance scientist whose interests were not limited to
smooth muscle. He had an ongoing interest in chemotaxis and the relation-
ship between changes in intracellular calcium and cellular motility (Brun-
dage, R.A., Fogart, K.E., Tuft, R.A., and Fay, F.S. Calcium gradients under-
lying polarization and chemotaxis of eosinophils. Science 254:703-706,
1991). His lab also examined a variety of topics that explored the relation
of molecular distribution to cell function. Studies either initiated in his
lab or conducted collaboratively with other scientists at the University of
Massachusetts or around the world included explorations of the association
of hexokinase with mitochondria, the distribution of precursor messenger
R N A in the nucleus, the codistribution of poly(A) R N A with microfila-
ments, the localization of a yeast splicing factor in subnuclear domains, the
migrations of MAP kinase, and intracellular trafficking of growth factor re-
ceptors.
xl FREDRIC STEWART FAY
Fred was a superb scholar and teacher. He was not satisfied with proving
to himself that he understood the physical principles behind the optical
techniques that he incorporated into development of optical instruments.
Fred believed that the mark of knowing whether he really understood a
concept was to then try and teach it to someone else. Therefore, he designed
a successful graduate course in "Optical Methods," which served as a great
forum for presenting established ideas and developing new ones. Fred also
was an active participant in the Medical Physiology course given each year
to first-year medical students. Over the years he taught a wide range of
physiological topics, inevitably coming up with new insights and approaches
to solidly established findings. Fred always had the patience to help his
fellows learn the art of public presentations. He would videotape his fellow's
job interview seminars and spend hours going over them point by point
until he was satisfied. His standards were high and he brought out the best
in all who worked with him.
Fred will be remembered by his students and colleagues in Worcester
and on every continent for more than his science, for he was a very special
human being. His home, his heart, and his mind were open for all to share,
and his wit and energy brought joy to all who were privileged to know
him. He worked hard and played hard. He touched many lives, and left
them all better for the contact. He took great pleasure and pride in his
wife Madeleine, his children, Andrew, Nicholas, and Isabel, and his grand-
children, Sarah, David, and Julia. He also derived great joy from the scien-
tific offspring who worked in his lab. The list includes seven medical stu-
dents, five graduate students, twenty-three postdoctoral fellows, and six
senior scholars who spent their sabbatical leaves with him. Many of his
former students and fellows have gone on to build impressive careers for
themselves, and their science will always bear Fred's indelible imprint. Fred
died far too soon, but we can take some comfort in the knowledge that he
packed a great deal into his all too short lifetime. In all, he published more
than 125 papers and reviews in the most highly critical and respected
journals. His creativity resulted in the award of eight patents, and his
technological innovations in optical methods paved the way for him and
future generations to "boldly go where no one has gone before."
H. MAURICE GOODMAN DAVID WARSHAW

DEPARTMENT OF PHYSIOLOGY DEPARTMENT OF MOLECULAR
UNIVERSITY OF MASSACHUSETTS PHYSIOLOGY AND BIOPHYSICS
MEDICAL SCHOOL UNIVERSITY OF VERMONT MEDICAL
WORCESTER, MASSACHUSETTS SCHOOL
BURLINGTON, VERMONT
[ 1] NEW PHOTOPROTECTING GROUPS 1
[1] N e w P h o t o p r o t e c t i n g Groups: Desyl and

p-Hydroxyphenacyl Phosphate and Carboxylate Esters
By F. W. W E B E R ,
R I C H A R D S. G I V E N S , J O R G ANDREAS H. JUNG,
and CHAN-Ho PARK
The well-established photochemical release of bioorganic substrates

represents one of the most ingenious methods for inducing an in vitro,
controlled biochemical or physiological response yet devised. The major
advantages of employing "photoprotecting" or "caged ''1 substrates over
coeventional methods for substrate release are threefold: (1) the photopro-
tected or caged substrate can be dispersed throughout a biological prepara-
tion without eliciting the substrate's normal stimulus; (2) the release of the
substrate can be rigorously controlled, i.e., both the concentration and the
spatial distribution can be carefully managed; and (3) the temporal release
can be varied over a range of first-order rate constants for release from
second to nanosecond domains. First-order rate constants for the release
of this magnitude are fast enough to pursue kinetic analyses of practically
any subsequent biochemical response.
The necessary properties 2 of a successful caged or phototrigger group
for common biological substrates such as ATP, L-glutamate, or y-aminobu-
tyric acid ( G A B A ) have been suggested and include the following: (1) The
photoprotected substrate must be soluble in aqueous buffered media. (2)
The cage or phototrigger must be stable to hydrolysis, especially at high
ionic strengths. (3) The photochemistry should be efficient, preferably with
a quantum efficiency >0.1, and the activating radiation should be >300 nm
to avoid competition for the light by biological media. (4) The photoprod-
uct(s) of the protecting group must be biologically benign. In addition to
these four necessary requirements, desirable properties for the phototrigger
include the following: (5) The photoproducts should not absorb at the same
wavelength as the activating radiation for the phototrigger to avoid a light
filtering effect by the product at high conversions. (6) The photorelease
reaction should be a primary photochemical step, i.e., direct bond cleavage
i "Caged compound" is a term coined in 1978for photolabile derivativesof natural substrates

such as ATP. The first such derivativeswere reported by Kaplan, Forbush, and Hoffman.4
More recently, the term "phototrigger" has become increasinglymore acceptable as a better
description of the photoactive protecting group.
2General rules for protecting groups in chemical synthesis can also be found in V. N. R.
Pillai, Synthesis 1-26 (1980) and Org. Photochem. 9, 225-343 (1988).
Copyright © 1998 by Academic Press

All rights of reproduction in any form reserved.
METHODS IN ENZYMOLOGY, VOL. 291 0076-6879/98 $25.00
2 CAGED COMPOUNDS [ 1]
NH2
NH2
NO2 347 nrn

R
I '~ 0 O O ~'N "/'~N> irradiation O O O ~-N~N
F'~~'~o~ - o- ~- o- ~-o--~ I
+ -6 -6 -6 L....-o-._J 3 mM MgCI; "O-P-O-¢-O-P-O---~t
I I I I
^ I * . H 4 +NH4+NH4
OH OH krpH
= 867.1s-I +NH4 +NH4+NH4 O H ~ H
21° 4a,b
"Caged"-ATP,s ATP ~ = 0.63
/
la,R=H.X =H (2NB)
b.R = OH3, X = H (2NPE)
(also, X = OCH3)
OPO3R' C?PO3R'-
k hydml X~ R
× ÷ ~LI~OoI:OH
X
OH
-d
3, nitroso hemiketal
2, aci-nitro intermediate
SCHEME 1
to the substrate from the reactive excited state, thus avoiding any long-
lived intermediates prior to release of the substrate.
These six criteria seldom are met in their entirety, but they do serve as
excellent guidelines for the design and development of new phototriggers.
Over the past twenty years, 3 the emphasis in the field of cage or photo-
trigger ligands has centered on derivatives of the o-nitrobenzyl group. The
two basic archetypical chromophores are the 2-nitrobenzyl (2NB, la) and
2-nitrophenethyl (2NPE, lb) ligands and their derivatives. This robust chro-
mophore undergoes an initial intramolecular redox reaction on excitation
at >300 nm (Scheme 1). An aci-nitro intermediate 2 is formed, which
subsequently undergoes further molecular reorganization to form a hemi-
ketal or acetal 3. It is the hydrolysis of 2 that leads to the release of the
substrate. The original o-nitrobenzyl cage group has thus been ejected as
an o-nitrosoacetophenone or benzaldehyde (4a,b; Scheme 1). For the
o-nitrobenzyl cages, the photorelease occurs at a modest rate, i.e., k r = 1-
1 0 3 s e c -1, and is not very sensitive to minor changes in temperature or
the media. Good quantum efficiencies (-0.1-0.6) and absorptivity at
wavelengths longer than 300 nm are the distinctive advantages that fa-
vor this ligand.
An impressive number of applications of "caged" reactions have ap-
peared that almost exclusively employ a variant of the 2-nitrobenzyl photo-
3 R. S. Givens and L. W. Kueper III, Chem. Rev. 93, 55-66 (1993).

[ 1] NEWPHOTOPROTECTINGGROUPS 3
chemistry outlined in Scheme 1. 4-7 Several of these applications have been

directed toward understanding fundamental biochemical processes includ-
ing, for example, the mechanism of release of inorganic phosphate in skele-
tal muscle,8 the action of actomyosin in muscle contraction induced by
ATP, 9 the role of cAMP in the relaxation of distal muscle,1° and the action
of Ca2+-ATPase in the sarcoplasmic reticulum in active calcium transport
during ATP hydrolysis.11 Other studies have dealt with refining, modifying,
and improving the photochemistry of the o-nitrobenzyl cage or with opti-
mizing or modifying the reaction conditions.7'12 Very little has appeared,
however, that reveals discovery of new photoactive caging ligands.
Nevertheless, there are several disadvantages that limit the application
and usefulness of 2NB/2NPE photochemistry.12Among the most problem-
atic are (1) the highly absorbing reactive nitroso ketone or aldehyde 4 that
is formed as the side product, (2) the difficulty in synthesis of some of the
caged derivatives, (3) the instability of the caged derivative to premature
decomposition or to biochemically induced consumption prior to photolysis,
and (4) competitive energy migration or electron transfer away from the
excited 2-nitroaryl group prior to C - O bond lysis. For 2NB/2NPE, addi-
tional disadvantages include the reaction of the o-nitrosobenzaldehyde or
ketone 4a,b with peptides and proteins, possibly interfering with the biologi-
cal processes under examination or with normal spectroscopic monitoring
protocols. The relatively modest microscopic rate for release of the nucleo-
tide (i.e., kr - 1-10 3 sec -1) is limited by the rate-controlling hydrolysis of
the aci-nitro intermediate 2 (Scheme 1). 12 The slow release of nucleotides,
for example, prevented the complete kinetic analysis of the mechanism of
ATP on the contraction of myocardial muscle. Niggli and Lederer 13 found
4 j. H. Kaplan, G. Forbush III, and J. F. Hoffman, Biochemistry 17, 1920-1935 (1978). The
recently adopted term "phototrigger" imparts a concept of a very rapid release of the
substrate that is now attainable with the newer photoprotecting groups.
5 See reviews: (a) H. A. Lester and J. M. Nerbonne, Annu. Rev. Biophys. Bioeng. 11, 151
(1982); (b) J. M. Nerbonne, Optical Methods Cell Physiol. 40, Ch. 24, 418-445 (1986); and
(c) A. M. Gurney and H. A. Lester, Physiol. Rev. 67, 583-617(1987).
6 j. A. McCray and D. R. Trentham, Annu. Rev. Biophys. Chem. 18, 239-270 (1989).
7 j. E. T. Corrie and D. R. Trentham, in "Bioorganic Photochemistry" (H. Morrison, ed.),
Wiley, New York (1993).
8 j. W. Walker, Z. Lu, and R. L. Moss, J. Biol. Chem 267, 2459-2466 (1992).
9 E. M. Ostap and D. T. Thomas, Biophys. J. 59, 1235-1241 (1991).
10 R. F. Willenbucher, Y. N. Xie, V. E. Eysselein, and W. J. Snape Jr., Am. J. Physiol.
(Gastrointest. Liver Physiol. 14) 262, G159-G164 (1992).
11 S. M. Lewis and D. D. Thomas, Biochemistry 30, 8331-8339 (1991).
12 (a) J. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. 110,
7170-7177 (1988); and (b) J. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham,
Methods in Enzymol. 172, 288-301 (1989).
13 E. Niggli and W. J. Lederer, Biophys. J. 59, 1123-1135 (1991).
4 CAGEDCOMPOUNDS [ 1]
that ADP formation by enzyme-induced hydrolysis of ATP was more rapid

than the rate of ATP release by photolysis of the caged nucleotide. The
presence of ADP, both from enzymatic formation and as an impurity in
the caged ATP sample, severely compromised the study.
To date, however, the many advantageous properties of 2-nitrobenzyl
cages have far outweighed the disadvantages, resulting in a plethora of
major applications of 2NB/2NPE chemistry to the study of biochemical
mechanisms. Nevertheless, future improvements are necessary to meet the
increasing demand for faster release rates and a wider selection of excita-
tion wavelengths.
New Phototriggers for Phosphates and Nucleotides:

Desyl and p-Hydroxyphenacyl "Cages"
Our studies of desyl and p-hydroxyphenacyl phototriggers provide the
first general alternative to the 2-nitrobenzyl cage approach. In our initial
investigations, we synthesized a series of desyl phosphates 14q7 (e.g., 5a,
Scheme 2) and demonstrated that the release of the phosphate ligands
occurs efficiently (~ = 0.3) with rate constants for release estimated at
>10 7 sec -1 [Eq. (1)]. 16Both the quantum efficiencies and the product yields
x' Lph
x;x O-PO3R2
5a, X = H
b, X = OCH3
hv = (RO)2PO2" +
were independent of the solvent (benzene, CH3CN, CH3OH, or H20).

X
6a,b
(1)
Furthermore, the ligands on the phosphate group had no effect on the

photochemistry.
To demonstrate the application of this phototrigger for the release of
a nucleotide, desyl-cAMP 7a was synthesized (Scheme 2) and irradiated
14 R. S. Givens and B. Matuszewski, J. Am. Chem. Soc. 106, 6860-6861 (1984).

15 R. S. Givens and L. W. Kueper III, Chem. Rev. 93, 55-66 (1993).
16 R. S. Givens, P. S. Athey, L. W. Kueper III, B. Matuszewski, and J.-y. Xue, J. Am. Chem.
Soc. 114, 8708-8709 (1992).
17 R. S. Givens, P. S. Athey, L. W. Kueper III, B. Matuszewski, J.-y. Xue, and T. Fister, J.
Am. Chem. Soc. 115, 6001-6012 (1993).
a. Synthesis of Desyl Phosphate (5a)
O 1. (MeO)3P,CH2CI2
,,~ 2. AcBr, CH3CN Ph O O 1. pyridine,Phil
Ph Ph " ~P~o 2. (COCl),, Phil
O Ph Me
benzil 2-methoxy-4,5-diphenyl-2-oxo-1,2,3-dioxaphosphate
o/PO3"2
Ph
"~O'~P~CIPh"
O O H20, R3N,THF =. p h ~h.O
p
2-chloro-4,5-diphenyl-2-oxo-1,2,3-dioxaphosphate 5a
b. Synthesis of Desyl cAMP (7a)
N,N-dimethylacetamide, ~P,~ I
+ (Bu)3N, 100", 20% Dk/ | ~0
H, H
\\
0
O OH
p Br
;H
0
Desyl bromide (Bu)3NH+ 7a, Desyl caged cAMP
cAMP
SCHEME 2
at either 300 or 350 r i m . 16'17 The results, shown in Eq. (2), illustrate the
highly efficient, rapid release by desyl as the phototrigger. The release of
cAMP occurs with a first-order rate constant of 108 sec -1 and is accompanied
by the formation of 2-phenylbenzo[b]furan (6a). This and all other desyl
photoreactions that we have examined are triplet mediated, quenchable
NH2 NH2
.o (2)
Ph O/ kr = 108 s-1 HO /
O
7a, Desyl caged cAMP ~dis = dp=pp= 0.3 6a
cAMP
2x xSx
OPO2R2
O.PO3R2 ntersystem O.PO3R2 X ~ + ph
crossing
5a, X = H
b, X = OCH 3
X PO2R2 -(RO)2PO~ -H+ X

ID
[ I[i u+ Ph h
X X Ph
X
8 9 6a,b
SCHEME 3
by low triplet energy quenchers such as naphthalene (ET = 62 kcal/mol).

Direct observation of the triplet by laser flash photolysis 18 provides defini-
tive evidence for the short-lived triplet (hma~ = 460 nm, ~'trip = 12 nsec)
that decays essentially instantaneously to the 2-phenylbenzofuran 6a
(hmax = 300 nm, kr = 108 sec-1). Based on these preliminary observations,
a reasonable sequence of events leading to the formation of the benzofuran
begins with the electrophilic n,zr* triplet attacking the benzyl group to
generate 8 and then the intermediate protonated dihydrobenzofuran 9, i.e.,
simultaneous or sequential loss of the phosphate and the proton generate
the furan and the released substrate (Scheme 3). Such an attack on a
m,m'-dimethoxybenzyl derivative should be especially favorable due to the
electron-donating contributions of the two ortho, para-directing methoxy
groups (see Scheme 3).
Although this mechanism involves the generation of a reactive interme-
diate, thus violating the requirement of primary photochemical processes,
the lifetime of the intermediate is short enough that the release process is
still five orders of magnitude faster than the corresponding o-nitrobenzyl
derivative. Therefore, benzoin phototriggers may have important applica-
tions when absorption at longer wavelengths (>350 nm) and rapid release
(>108 sec -1) are required.
18j. Wirz, R. Givens, and B. Hellrung, unpublished results, 1997.

Subsequently, Corrie and Trentham 19 explored the photochemistry of

these and the closely related m,m'-dimethoxybenzoin phosphates as a cage
ligand, z° They have reported the development of caged analog 5e [Eq. (3)]
Ph .0
x'~ 000
_ II II_
J--/-.~O II
MeO~e" (~.0~.0~)_
00 0 ~ o:d hv,, ATP+ M e O ~ (3)
IOMe~ Ph
5c, m,m'-dimethoxydesyI-ATP 6b
to replace the 2NB/2NPE groups. Their results parallel our findings 16,17
and thus indicate that the desyl group deserves further attention as a cage
ligand for nucleotides. In both cases, the photoproducts of the cage ligand
appear to be biochemically benign as biological substrates.
Several limitations on the use of benzoin derivatives as phototriggers
are apparent from these studies. Among the most severe are (1) the intro-
duction of a new chiral center, which results in diastereomeric mixtures
whenever the caged substrate contains one or more chiral centers; (2) the
generation of a highly absorbing, fluorescent photoproduct 6a,b; (3) the
instability of the phototrigger for selected derivatives of nucleotides2°; and
(4) the marginal aqueous solubility of the reactant phototrigger and its
photoproduct. Each of these limitations can, in principle, be addressed for
specific applications. For example, a single enantiomeric benzoin 21 can be
synthesized or the phototrigger mixture of diastereomers can be resolved.
Analogs containing sulfonic or carboxylic acid functions could be developed
to enhance the aqueous solubility of the phototrigger and its photoproduct.
These modifications would have a negligible effect on the photochemistry.
A more direct approach to overcoming all four limitations, however,
can be anticipated from an examination of the structural similarity of the
desyl group to another well-known o~-keto derivative, the p-methoxyphen-
acyl chromophore, first introduced by Sheehan, Wilson, and co-workers. 22
By simply removing the aryl substituent on 5 and substituting a p-hydroxy
moiety for Sheehan's p-methoxy group on the phenacyl ring, the aqueous
solubility of the resulting phototrigger 10e should improve dramatically,
and the new phototrigger is absent the complications of a chiral center.
I9 j. E. T. Corrie, G. P. Reid, D. R. Trentham, M. A. Mazid, and M. B. Hursthouse, J. Chem.

Soc. Perkin Trans. I, 1015-1019 (1992).
20 j. E. T. Corrie and D. R. Trentham, J. Chem. Soc. Perkin Trans. I, 2409-2417 (1992).
21 M. C. PilTung and S. W. Shuey, J. Org. Chem. 59, 3890-3897 (1994).
22 (a) J. C. Sheehan, R. M. Wilson, and A. W. Oxford, J. Am. Chem. Soc. 93, 7222-7228
(1971); and (b) J. C. Sheehan and J. Umezawa, J. Org. Chem. 38, 3771-3774 (1973).
8 CAGED COMPOUNDS [l ]
Furthermore, based on the earlier studies by Anderson and Reese, 23 intro-

duction of a p-hydroxy and p-methoxy substituent (i.e., 10a or 10d) pro-
motes a new, competing photorearrangement process yielding a substituted
phenylacetate 11 in addition to the normally encountered acetophenone
reduction product 12 [Eq. (4)].
O O
+x
10lg Y = OCH3; X = CI 1la, Y = OCH 3 12a, Y - OCH3

(4)
b, Y - OCH3, X = O2CR b, y = O H b, Y - O H
e, Y = OCH3, X = C~P(OR)2
O, Y - O H ; X = CI
e, Y - OH, X = O2P(OR)2
Parallel investigations of the structurally analogous p-methoxyphenacyl

phosphates have been reported by Epstein and Garrossian, 24 Baldwin et
al., 25 Pirrung, 21 Futura et al., 26 and by us, 14-17'27 that have extended the
applications of a-keto derivatives as phototriggers. However, it was not
until recently that the first applications ofp-hydroxyphenacyl phototriggers
were reported for phosphate or for a nucleotide, i.e., p-hydroxyphenacyl-
ATP. 27 Our synthesis of these two examples is outlined in Scheme 4 (struc-
tures 13-17), and the experimental details are provided in the Methods
section. Additional information on the synthesis and photochemistry, as
well as information on several analogs, appeared in our original articles? 7
In addition to the expected improved solubility and the lack of a diastereo-
meric mixture of caged derivatives for p-hydroxyphenacyl-ATP (17), we
were pleased to discover that the combined effects on the photochemistry
by the p-hydroxy substituent and the solvent change to H20 or buffers
were to promote the rearrangement reaction to p-hydroxyphenylacetic acid
[lib, Eq. (5)] to the exclusion of all other possible photoproducts from
the phototrigger.
A particularly attractive feature is the fact that the phenylacetic acid
does not compete for incident radiation >300 nm, thus permitting higher,
in fact complete, conversion of the caged phosphate to ATP [see Eq. (5)].
While this may not be a particular advantage in most physiological stud-
23j. C. Anderson and C. B. Reese, Tetrahedron Lett., 1-4 (1962).

24 W. W. Epstein and M. Garrossian, J. Chem. Soc., Chem. Commun., 532-533 (1987).
z5 j. E. Baldwin, A. W. McConnaughie, M. G. Moloney, A. J. Pratt, and S. B. Shim, Tetrahedron
46, 6879-6884 (1990).
26 T. Futura, H. Torigai, M. Sugimoto, and M. Iwamura, J. Org. Chem. 60, 3953-3956 (1995).
27 (a) R. S. Givens and C.-H. Park, Tetrahedron Lett. 37, 6259-6262 (1996); and (b) C.-H.
Park and R. S. Givens, J. Am. Chem. Soc. 119, 2453-2463 (1997). (Methods adapted in
part with permission from the American Chemical Society.)
[ 11 NEW PHOTOPROTECTING GROUPS 9
o o
O
(BnO)2PO2-÷N(Me)t, OP(OBn)2
Benzene,85%
HO.~ ~'~Br HO" "~/
13 14
o o
I I o
OP(OBn) 2 o _o II
Ethyleneglycol . . ~ . . . . . . . . ~ OP(O B n)2
HO" p-TsOH
(82%) HO"
14
15
I I o 1,P°,C o o
o o it
~
HO"
OP(OBn) 2
2) 1% HCI
3) DEAE,
Sephadex, HO ~
O- +NH4
NH4OAc +NH4
15 (96%) 16
NH2
o o o o 'I~N I ~ N ~
ADP 1) Im2CO DEAE-Cellulose O-P-O-P-O-P-O-q
,
r~ T v - o I .oI -oI ~_o_~
2) 16, HMPA NH4*HCO3-
(overallyield:68%)
HO~ ' ~ +NH4 +NH4+NH4 ~
OH OH
17
SCHEME4
ies, it may be important for applicationslwhen caged derivatives are intro-

duced as a medicinal agent. Such applications remain to be developed,
however.
The rearrangement is proposed to pats through a dienedione intermedi-
ate 18 by direct displacement of the phosphate group from the electron-rich,
7r* ~ o-* electronic excited state, releasir~g ATP in a primary photochemical
process (Scheme 5). Although many of~the mechanistic details remain to
10 CAGED COMPOUNDS [11
NH2
I ~ l ~ J 3 300 nrn
,rradiation 3
N NH~N~2
O O O O '~ N""~"N I
II II II I~ N
~L. OP-O-P-O-P-O--I 0fl 0II 0II
J~ J~ V - o I "0I -0I ~.-0~,.~
O Tris
pHbuffer
7.3 " O-P--O--P-O-P-O~
I I J
HOf'~..~ +NH4 +NH4 +NH4 OH]~OH kr=6.gx108s-1 "O -O .O ~......~O--~J

+NH4 +NH4 +NH4 ~
OH OH (5)
ATP ~app =0.30
17. "Caged"-ATP ~dis = 0.37
+ ~ 0 OH
HO
I I b,~app =0.31
be determined, the similarity of the phototrigger chemistry to recent
electrochemical studies on the radical anion-induced fragmentations of
substituted a-phenoxy- and carboxyacetophenones 28 and the correlation of
the rate constants for these fragmentations in laser flash photolysis studies 29
lend credence to the process depicted in Scheme 5. To date, however, the
dienedione 18 has not been detected. Additional support is the close analogy
of the photorearrangement to the anchimerically assisted 1,2-aryl migration
during the solvolytic reactions of/3-phenethyl halides and tosylates. 3°
Applications of Desyl and p-Hydroxyphenacyl Phototriggers to

Excitatory Amino Acids, Peptides, and Carboxylate Esters
Extension of the a-ketophosphate phototriggers to carboxylate release
has also been accomplished with both the desy131 and the p-hydroxyphen-
acy132 amino acids. For simple amino acids, the applications of these photo-
triggers has been limited to two examples, the excitatory y-aminobutyrate
GABA and y-O-substituted L-glutamate derivatives (e.g., 22a,b and 24a,b).
All of the a-amino-substituted carboxylic acid phototriggers proved to be
unstable to aqueous environments, hydrolyzing to the free amino acid and
the a-hydroxy phototrigger. The general synthesis for the amino acid deriva-
28 (a) C. P. Andrieux, J.-M. Sav6ant, A. Tallec, R. Tardivel, and C. Tardy, J. Am. Chem. Soc.
118, 9788-9789 (1996); and (b) M. L. Andersen, W. Long, and D. D. M. Wayner, J. Am.
Chem. Soc. 119, 6590-6595 (1997).
29 N. Mathivanan, L. J. Johnston, and D. D. Wayner, J. Chem. Phys. 99, 8190-8195 (1995).
30 D. J. Cram, J. Am. Chem. Soc. 71, 3863 (1952).
31 K. R. Gee, L. W. Kueper III, J. Barnes, G. Dudley, and R. S. Givens, J. Org. Chem. 61,
1228-1233 (1996).
32 R. S. Givens, A. Jung, C-H. Park, J. Weber, and W. Bartlett, J. Am. Chem. Soc. 119,
8369-8370 (1997).
0
0 3
x hv ST
Buffer~
250 C
• HO~ x
10e, X = OPO3R2 (n,z*)3

f, X = OCOR
3neighboring
o a~ group assisted
SET H20.¢TM
-HX
~,~OH
HO~ 0
(~,~,)3 11b
18
SCHEME 5
tives of the two phototriggers is outlined in Scheme 6 (structures 19-26)

and provided in detail for 22a,b and 24a,b in the Methods section. Yields
of the photoprotected carboxylates are generally quite good.
The photochemistry of the carboxylate analogs is likewise uncompli-
cated by side reactions (Schemes 7 and 8). Quantum efficiencies are high,
ranging from -0.1 to 0.3, and the photoreaction is quenchable by Na + 2-
naphthalenesulfonate, thus indicating triplet state intermediacy. The fate
of the phototriggers for the carboxylate series is identical to that observed
for the phosphate series. This observation provides further evidence that
the photochemistry of the phototriggers is independent of the leaving group,
at least within the series of phosphates and carboxylate leaving groups.
Preliminary evidence shows that the range of leaving groups can be ex-
panded to include sulfonate esters and several other conventional leav-
ing g r o u p s . 33
In an attempt to include peptides and higher homologs of amino acids
for applications of the p-hydroxyphenacyl phototriggers, the dipeptide Ala-
Ala 28 was derivatized by the same method employed for the amino acids
and its photochemistry examined in detail (Scheme 8). In contrast to the
a-amino acid derivatives noted earlier, this caged dipeptide was stable to
aqueous and buffered media, 34 probably a result of the lowered reactivity
33 L. Chimilio, A. H. Jung, J. F. W. Weber, and R. S. Givens, Unpublished results, 1997-1998.

34p-Hydroxyphenacyl Ala-Ala 27 does slowly hydrolyze in Tris buffer with a half-life of
213 min. 32
12 CAGED COMPOUNDS [ 11
O-t-Bu and N-t-BOC protectedamino acids

o
HO~NH-t..BOC
O
Arl C- -N
19, Gly Ar2 O CF3CO~
20a,b
HO~NH4"-BOC
o
O
O
21, GABA
Art ; DBU; Benzene
Ar 2 O CF3CO~
Ar2
H NH-t-BOC 22a,b GABA phototriggers
a) Aq, Ar 2 =CeH 5 (desyl bromide)
HO~co,~-Bu O H NHa
~
b) Aq = p-hydroxyphenyi, Ar2 = H
o (p-hydroxyphenacyl bromide)
23, Glu Arl C
At20
24a,b Glu phototngger

H NH-t-BOC
. ~ 2) TFA (62 - 84% overall)
t Bu.-O~co2 H
0 H3P~]~H
o
Arl--~"O~co2 ~
25, Glu Ar 2 O CF3CO~2
26a,b
SCHEME6
of the ester to hydrolysis. The lower basicity of the a-amino group as a

peptide linkage renders it less susceptible to protonation at pH - 7 . For
the simpler a-amino acid derivatives, the protonated amino group enhances
the nucleophilic attack by H20, thus leading to hydrolysis even in buf-
fered media.
O H NH3® hv, 350 nm H NI-1~3
O CO20 1:1 CH3CN:H20 CO2G +
O Ph
Glu 6a
24a Desyl Glu
O
.~....~O~.v/.-'~NH~ hv, 350nm
1:1 CH3CN:H20 H O ~ NH3~
22a Desyl GABA GABA 6a
SCHEME 7
O H NH3(~ hv, 300 nm H NH3

~
O CO2~ D20/H20;~100% ~r ~ v "CO~

HO/ ~'~'/ O No hydrolysis detected II
O + OH
24bp-hydroxyphenacyl Glu Glu

11b
O
,v3O0nm HO. A N
+ l i ~ ' F ' ~ ( °H
HO" ~ O CF3CO~ D20/H20 O CF3CO? HO,~J 0
GABA 11b
22bp-hydroxphenacyl GABA
H CH3 H O
H,, CH3 H O i f ' ~ " ~ OH ® ...... N
~ N H ~ "~OH
,, oonm • °" CF3C(~2 " ~ 1\
;F3C02e eNH3 O + U H ~H3
~]H3 D20; 18 rain HO
28 [c~]D -14.77~(c=0.13,H20)
27 p-hydroxyphe nacyI-Ala-AIa 11b (100% retention of optical ac'dvi~y
starting dipepUde had a rotation
[a]O -39.73"[0=0.59.H20] of -14.75 = under the same
20
m.p. 143"C cond~ons)
SCHEME 8
Hydrolysis thus avoided, the photochemistry of the caged dipeptide 27

was examined and shown to be in accord with the y-amino acid derivatives
as illustrated in Scheme 7. This reaction was also shown to be a triplet
process, occurring at high efficiency, and capable of 100% conversion to
photoproducts. Figure 1 illustrates the complete conversion to product for
27. Determination of the other physical properties of the dipeptide, e.g.,
melting point (mp), infrared data (IR), and [O~]D,further established that the
original stereochemistry was not compromised during the photochemical
release or the chemical manipulations? z
Applications of Desyl and p-Hydroxyphenacyl Phototriggers

Because these new phototriggers have only been known for a short
time, there are only a few reported applications. Most notable are the
examples of release of the excitatory amino acid L-glutamate, work that
has been carried out in the laboratories of Katz and co-workers at the
Howard Hughes Medical Institute at Duke University. 31'32
Katz has found that irradiation of hippocampal pyramidal neurons
bathed in either 1 mM desyl glutamate (24a) or 50 ~M p-hydroxyphenacyl
14 CAGED COMPOUNDS [li
1080SEC.
720SEC.
~
k~L ~ 240SEC.
)~ 120SEC.
, 0SEC.
FIG. 1. 1H NMR spectra of the photorelease of Ala-Ala (28) from p-hydroxyphenacyl-

Ala-Ala (27) in D20 at 300 nm. The sample of 27 in D20 was irradiated in a 5-ram-diameter
NMR tube at 300 nm in a Rayonet reactor (Southern New England Ultraviolet Company,
Hamden, CT). The NMR spectra were obtained before irradiation (lowest spectrum) and
after every 2 min of irradiation until the reaction was completed (~12 min). The change in
chemical shifts in the aromatic region (8 6.5-8.0 ppm), the appearance of the benzyl methylene
at 8 3.6 ppm, and the disappearance of the phenacyl methylene at t~5.45 ppm clearly illustrate
the conversion of the p-hydroxyphenacyl group to a p-hydroxybenzyl moiety.
glutamate (24b) in patch-clamp experiments will generate an action poten-

tial when the duration of the light flash releases sufficient L-glutamate to
directly stimulate the glutamate receptors (Figs. 2 and 3). The release of
glutamate was faster than the duration of the 8- to 10-msec flashes.
In voltage-clamp experiments (Fig. 2), the release of glutamate results
in a large inward current causing the cell to escape clamp and fire the
action potential. For desyl glutamate (24a), the excitation wavelength was
351-364 nm, whereas p-hydroxyphenacyl glutamate (24b) responded better
to 270-nm irradiation. Finally, in these experiments the desyl glutamate
24a showed a background intrinsic activity, causing as much as a 300%
increase in the holding current. In contrast, the p-hydroxyphenacyl gluta-
mate (24b) showed no detectable background activity.
As shown in Fig. 3, cells bathed in 50/zMp-hydroxyphenacyl glutamate
(24b) and exposed to flashes of different duration will evoke responses in
current-clamp recordings in proportion to the length of the exposure until
20 pA L
20 msec
FI6. 2. In a voltage-clamp experiment, desyl L-glutamate (24a; 1 mM in artificial cerebral

spinal fluid (ASCF) bathing solution) is irradiated near a recording neuron. The resulting
large inward current causes the cell to escape clamp and fire an action potential. The response
to uncaged glutamate begins as soon as the laser flash (8 msec) begins. [Reprinted in part
with permission from K. R. Gee, L W. Kueper III, J. Barnes, G. Dudley, and R. S. Givens,
Z Org. Chem. 61, 1228-1233 (1996). Copyright © 1996 American Chemical Society.]
- - 30 ms
- - 20 ms
- - 10 ms
- - 3 ms
t
shutter lOmV
100 ms
F~o. 3. In a current-clamp recording, p-hydroxyphenacyl L-glutamate (24b; 50 /zM) is

irradiated with 270-nm flashes of varying duration. Even in the shortest flash (3 msec), a large
voltage change is observed. These recordings were done in the presence of TTX to block
regenerative sodium action potentials. (Reprinted in part with permission from R. S. Givens,
A. Jung, C-h. Park, J. Weber, and W. Bartlett J. Am. Chem. Soc. 119, 8369-8380 (1997).
Copyright © 1997 American Chemical Society.)
the threshold is reached. Exposures of longer duration than 10 msec gener-
ate sufficient glutamate to cause a regenerative C a 2+ spike, as indicated by
the large jump in the response. Figure 4 shows similar photorelease studies
but in voltage-clamp recordings. Currents as large as 1500 pA are observed.
These studies were all conducted in the presence of tetrodotoxin (TTX)
to block regenerative sodium action potentials.
Summary
This chapter introduces two new phototriggers for caging nucleotides,
amino acids, peptides, and related functional group derivatives, potentially
including higher-order homologs. These new phototriggers have several
advantages that favor them over the conventional o-nitrobenzyl derivatives,
shutter: 5, 10, 20, 5 0 , 7 0 ms
200 pA
100 ms
FIG. 4. Voltage-clamp experiments from flash excitation of 50/aM of p-hydroxyphenacyl

L-glutamate (24b) at 270 nm gave increasingly larger currents, peaking at 1500 pA. Even at
very short shutter openings of ~ 5 msec, a 100-pA inward current was obtained. These experi-
ments were also carried out in the presence of tetrodotoxin (TTX) to prevent regenerative
sodium action potentials produced. (Reprinted in part with permission from R. S. Givens,
A. Jung, C-h. Park, J. Weber, and W. Bartlett 3. A m . Chem. Soc. 119, 8369-8380 (1997).
Copyright © 1997 American Chemical Society.)
[1] NEWPHOTOPROTECTINGGROUPS 17
including a more rapid release of the substrate by a primary photochemical
fragmentation process; adequate to excellent aqueous solubility; and stable,
benign photoproducts. Added attractive features for the p-hydroxyphenacyl
phototrigger include its rearrangement to a phenylacetic acid, thus shifting
the chromophore absorption to a shorter wavelength, eliminating its inter-
ference with the incident radiation. Finally, the absence of a chiral center
eliminates diastereomeric mixtures during synthesis.
It is anticipated that these features will result in added demand for use
of a-keto phototriggers for many new applications in biochemistry and
physiological studies.
Methods
General Experimental M e t h o d s
All nuclear magnetic resonance (NMR) spectra are reported in ppm

(~) from either tetramethylsilane (1H and 13C) or 85% phosphoric acid
(31p). Melting points are uncorrected. Triethylamine is dried by refluxing
over potassium hydroxide and distilling under argon, collecting the middle
fraction, and stored in the dark. Pyridine and N,N-dimethylformamide are
dried over phosphorus pentoxide and distilled under reduced pressure,
collecting the middle fraction, and stored over sodium hydroxide and molec-
ular sieves, respectively. Benzene is distilled from sodium/benzophenone
ketyl prior to use. Technical-grade diethyl ether (Et20), methanol, dichloro-
methane (CH2C12), and tetrahydrofuran are distilled from calcium hydride.
Synthesis o f D e s y l D i h y d r o g e n Phosphate, S o d i u m Salt (Sa) 16'17
2, 2, 2 - Trimethoxy - 4,5 - diphenyl-2,2 - dihydro - 1,3,22t 5 - dioxaphosphole.
Benzil (35 g, 0.17 mol) is added to a stirred solution of freshly distilled
trimethyl phosphite (24 g, 0.19 mol) in 20 ml of dichloromethane at 0° over
a 2-hr period. After an additional hour of stirring at 0°, the solvent is
removed by a Rotavapor (BUchi Labortechnik AG, Basel, Switzerland),
leaving a viscous yellow oil. The oil is used without further purification.
The crude yield is nearly quantitative as determined by aH NMR. 1H NMR
(CDCI~) ~ 3.73 (d, J = 13.2 Hz, 9H), 7.26 (m, 6H), 7.53 (d, J = 1.8 Hz,
4H); 13C NMR (CDC13) 55.6 (d, J = 10.5 Hz, CH3), 126.5, 127.8, 128.3,
130.8 [d, J 12.5 Hz, COP(OCH3)3], 133.9; 3lp NMR (CDC13, 85% H3PO4)
=
33.52 (m, J = 13.2); IR (neat) 3200-2800, 1600, 1450, 1175 (POCH3),

1070, 750 cm -1.
2_Methoxy_4,5_diphenyl_2_oxo_l,3,2,~5_dioxaphosphate. 16,17The trimeth-
oxydioxaphosphole is dissolved in 80 ml of dry acetonitrile. In one portion,
3.0 ml (3.9 g, 0.032 mol) of freshly distilled acetyl bromide is added to the
stirred solution. Because the reaction is exothermic, the remaining 11.0 ml
18 CAGEDCOMPOUNDS [ 11
(16.7 g, 0.137 mol) of acetyl bromide is added slowly over a 1-hr period at
a rate that keeps the solution at 50°. The reaction is followed by the chemical
shift and the growth of the methoxy doublet in the 1H NMR. On completing
the reaction, the solvent is evaporated to yield a thick yellow oil that is
not purified further. The crude yield is >95% based on 1H NMR. 1H NMR
(CDCI3) 8 3.94 (& J = 12.3 Hz, 3H), 7.3 (m, 6H), 7.4 (m, 4H); 31p NMR
(CDC13, 85% H3PO4) 8 11.65 (m, J = 11.3 Hz); IR (neat) 3100, 1600, 1450,
1300, 1250 cm -1.
4,5-Diphenyl-2-oxido-2-oxo-l,3,2ltS-dioxaphosphole. 16'17 A solution of
the cyclophosphate synthesized above (2-methoxy-4,5-diphenyl-2-oxo-
1,3,2-dioxaphosphate, 48 g, 0.16 mol) and pyridine (36 g, 0.53 mol) in 90 ml
of dry benzene is heated at gentle reflux with stirring for 7 hr, cooled to
20°, the solvent decanted and the precipitate filtered, and thoroughly washed
with benzene (2 × 50 ml) under argon. The fine crystalline salt is dried at
25 ° at 0.3 torr for 12 hr and stored under dry argon. The isolated yield is
95% (57 g, 0.15 tool), mp 149-151 °. 1H NMR (DMSO) 6 4.4 (s, 3H), 7.25
(m, 6H), 7.4 (m, 2H), 8.0 (t, J = 6.8 Hz, 2H), 8.5 (t, 1H), 9.0 (d, J = 5.8
Hz, 2H); 13C NMR (DMSO) 8 47.8, 126.22, 127.6, 127.73, 128.32, 131.32
(d, J = 9.25 Hz), 134.54 144.92, 145.52; 31p NMR (DMSO) ~ 18.8 (s); IR
(KBr) 3500, 3100, 1650, 1505, 1455, 1270, 1140, 1080, 960, 840, 790, 770,
710 cm -1.
Desyl Dihydrogen Phosphate, Aniline Salt (Sa, Aniline Salt). 16,17A solu-
tion of 5.0 g (13.6 mmol) of N-methylpyridinium salt of 4,5-diphenyl-2-
oxido-2-oxo-l,3,2-dioxaphosphole is treated with phosgene as previously
described. After workup, the oil is taken up in 30 ml of tetrahydrofuran and
cooled to 0°, to which 0.50 g (28 mmol) of water in 10 ml of tetrahydrofuran
(THF) is added dropwise with stirring. During the addition, the color
changes from orange to yellow. After 30 min, the solvent is removed to
yield a yellow oil that is dissolved in 30-40 ml of ethanol, to which 1.9 g
(20 mmol) of aniline is added. The resulting anilinium salt precipitates
within a few minutes. The isolated yield of the aniline salt of desyl dihydro-
gen phosphate is 4.80 g (13.0 mmol, 92% yield). The crude salt is recrystal-
lized from ethanol; mp(decomposition) 131-140 °. IR (KBr) 3400-2200,
1700, 1600, 1580, 1500, 1450, 1220, 1190, 1080, 980cm-1; analysis for
Ce0H20NOsP calculated: C, 62.34; H, 5.23; N, 3.63. Found: C, 61.58; H, 5.59;
N, 3.01.
Desyl Dihydrogen Phosphate, Sodium Salt (Sa, Sodium Salt). 16"17 The
desyl phosphoric aniline salt (0.52 g, 1.38 mmol) is dissolved by heating in
30 ml of deionized water. The solution is allowed to cool and is then passed
through a sodium ion-exchange column (1.5 × 22 cm). Approximately 125
ml of the eluant is collected in a tared flask and the water removed by
lyophilization. The sodium salt of 5a, a white powder, is isolated in 87%
yield (0.42 g, 1.2 mmol), mp/dec 180-183 °. 1H NMR (D20) ~ 6.46 (d, J =
8.5 Hz, 1H), 7.2-7.5 (m, 8H), 7.84 (d, J = 7.8 Hz, 2H); 13C NMR (D20)
81.32 (d, J = 4.22 Hz), 130.61, 131.67, 131.72, 131.96, 136.82, 137.10, 138.43
(d, J = 4.6 Hz), 202.01; IR (KBr) 3480, 3020, 1680, 1580, 1480, 1430, 1240,
1200, 1180, 1060 cm-1; UV [1-I20, Am~,(~)],251 (11,000), 323 (400). Analysis
calculated for C14H11OsPN: C, 50.02; H, 3.30. Found: C, 51.08; H, 4.00.
Synthesis of Desyl (-)-Adenosine Cyclic Y,5'-Phosphate (Ta) 16'17
Tri-n-butylammonium Salt of (-)-Adenosine Cyclic3',5'-Phosphate. To
1.0 g (2.9 mol) of (-)-adenosine cyclic 3',5'-phosphate (cAMP) in 250 ml
of deionized water is added 1.1 equivalents of tri-n-butylamine. The water
is removed by lyophilization, yielding a white salt (1.5 g, 2.9 mmol, 100%
conversion); dec ~ 120°. 1H NMR (D20) 8 0.79 (t, J = 7.3 Hz, 9H), 1.24
(m, J = 7.3 Hz, 6H), 1.52 (m, J = 3.4 Hz, 6H), 2.98 (m, 6H), 4.18 (m, 2H),
4.35 (m, 1H), 4.43 (m, 1H), 4.61 (m, 1H), 6.03 (s, 1H), 8.08 (s, 1H), 8.09
(s, 1H); 31p NMR (D20) t3 -0.91.
Desyl Adenosine Cyclic 3',5'-Phosphate (7a).16,17 The procedure of
Rameriz 35 is adapted to the synthesis of 7a. The tri-n-butylammonium salt
of cAMP (0.5 g, 0.97 mmol) is dissolved in 11 ml of freshly distilled N,N-
dimethylacetamide and 1.7 ml of tri-n-butylamine in a 100 ml round-bottom
flask. To this mixture is added dropwise, at 100° over a 10- to 15-min
period, a solution of desyl bromide (1.3 g, 4.75 mmol) in 3 ml of N,N-
dimethylacetamide. Monitoring of the reaction by thin layer chromatogra-
phy (TLC) in a 90% CHC13and 10% methanol solvent system reveals the
appearance of the phosphate triester with an RI of 0.4. After 45-60 min at
100°, the reaction mixture is cooled to room temperature. The solvent is
removed by distillation with a Kugelrohr distillation apparatus (Aldrich
Chemical Company, Milwaukee, WI). A 31p NMR (CDC13)of the resulting
dark orange oil indicates the reaction is 20-25% complete.
The oil is washed three times with hexane, then dissolved in a minimal
amount of chloroform and separated by flash chromatography on a silica
column (20 x 3 cm). Elution using 100% chloroform removes the unreacted
desyl bromide and nonpolar derivatives of desyl bromide (this is easily
noted by the yellow band that readily elutes from the column). The solvent
system is then adjusted to 5% methanol and 95% chloroform and the cyclic
triester eluted in fractions 20-25 (fraction size ~ 25 ml). Appropriate frac-
tions are combined and the solvent is removed by evaporation under re-
duced pressure (Rotavapor). The major impurity found in the triester is
tri-n-butylamine, which is detected by 1H NMR. Final purification is
achieved by dissolving the residue in chloroform and eluting it off a flash
35 F. Rameriz and J. F. Marecek, Synthesis 449 (1985).

silica column as previously described, using 100% chloroform to first elute
the amine, followed by elution of the triester with 5% methanol in chloro-
form. Detecting the eluted amine is difficult; therefore, removal of the
residual tri-n-butylamine was achieved by repeating the above chromatog-
raphy. Typically, a 100-mg sample is isolated (20% yield): 1H NMR (di-
methyl sulfoxide, DMSO) 8 4.0 (m, 1H), 4.25 (m, 2H), 4.4-4.8 (m, 5H),
5.35 (m, 1H), 5.55 (m, 1H), 6.05 (s, 1H), 6.07 (s, 1H), 6.35 (broad doublets,
2 OH), 7.05 (overlapping doublets, 2H), 7.25-7.7 (m, 16H), 7.8-7.96 (m,
6H), 8.2 (s, 1H), 8.27 (s, 1H); 31p NMR (DMSO) 8 -5.5 and -6.3; IR
(KBr) 1700, 1650, 1600, 1300, 1000, 925, 840, 700 cm-l; UV [H20, ,~max(8)]
253 (20,000), 320 (500); mass spectrum (FAB +) calculated for C24H22NsOTP:
524 (M+), 330, 195, 115.
Synthesis of p-Hydroxyphenacyl Phototriggers 27

Synthesis of 2-Bromo-4'-hydroxyacetophenone (13). 278 2-Bromo-4'-hy-
droxyacetophenone (13) was prepared by a modification of the method of
Durden and Juorio. 36 To 650 ml of a solution of chloroform and ethyl
acetate (1 : 1, v/v) is added 69.8 g (313 mmol) of cupric bromide and 17.2 g
(125 mmol) of 4-hydroxyacetophenone. The mixture was heated at reflux
under a stream of N2 to remove HBr for 24 hr and additional CuBr2 (5.58 g;
25 mmol) is added if the TLC indicates unreacted 4-hydroxyacetophenone.
The mixture is cooled to room temperature and the solution is decanted,
filtered through a pad of silica gel, the pad is washed with 9 : 1 CH2C12 : ethyl
acetate, and the combined solvent mixture evaporated. The crude product
is reerystallized from benzene to give 16.8 g (78.1 mmol, 62%) of 2-bromo-
p-hydroxyacetophenone 27 (13): mp 129-131°; 1H NMR (acetone-d6) 8 4.64
(s, 2H), 6.97 (d, J = 8.7 Hz, 2H), 7.97 (d, J = 8.7 Hz, 2H); 13C NMR
(acetone-d6) ~ 122.22, 132.83, 138.29, 169.31, 196.35.
Sequence for Conversion of 13 to p-Hydroxyphenacyl Phosphate (16)

Synthesis of Tetramethylammonium Dibenzyl Phosphate. 37 A solution
of dibenzyl monophosphoric acid (4.56 g, 16.44 mmol) in 20 ml of acetone
is placed in a 100-ml two-necked flask equipped with a thermometer and
a dropping funnel. A 25% aqueous solution of tetramethylammonium hy-
droxide (5.9 ml, 1.50 g, 16.44 mmol) is added dropwise through the dropping
funnel with efficient stirring in an ice bath at 0°. After the addition is
36(a) D. S. Durden, A. V. Juorio, and B. A. Davis, Anal Chem. 52, 1815-1820 (1980); and
(b) N. P. Buu-Hoi and D. Lavit, J. Chem. Soc., 18-20 (1955).
37(a) M. Kulbaand A. Zwierzak,Ann. Soc. Chim. Polonorum48,1603(1974);(b) A. Zwierzak
and M. Kulba, Tetrahedron22, 3163-3170 (1971);and (c) E. P. Serebryakov,L. M. Suslova,
and V. F. Kucherov, Tetrahedron34, 345-351 (1978).
complete the resulting solution [neutral to slightly alkaline (pH 7.0-7.5)]
is stirred for 2 hr. Evaporation of acetone and H20 on a Rotavapor gives
a concentrated aqueous solution of the salt. The residual H20 is removed
by azeotropic distillation with benzene to yield 5.66 g (98%) of tetramethyl-
ammonium dibenzyl phosphate: mp 62-70°; 1H NMR (D20) 83.15 (s, 12H),
4.84 (d, J = 7.4 Hz, 4H), 7.38 (m, 10H); 13C NMR (D20) 8 57.83, 70.31 (d,
J = 20.5 Hz), 130.50, 130.96, 131.37, 139.86 (d, J = 27.8 Hz); 31p NMR
( D 2 0 ) 8 0.66.
p-Hydroxyphenacyl Dibenzyl Phosphate (14).27 A mixture of tetrameth-
ylammonium dibenzyl phosphate (160 g, 4.70 mmol), p-hydroxyphenacyl
bromide (13; 1.00 g, 4.70 mmol), and 20 ml of benzene is placed in a 50-ml
round-bottom flask and refluxed with efficient stirring for 10 hr. The solu-
tion is cooled to room temperature and extracted with the HzO/ethyl acetate
solution. After removal of the solvent, the product is recrystallized from
CH2C12 : ethyl acetate (1 : 1) or is purified by a silica gel flash column chroma-
tography (CH2C12 : ethyl acetate; 60 : 40) to yield 1.63 g (85%) ofp-hydroxy-
phenacyl dibenzyl phosphate (14). Mp 116.5-118°; 1H NMR (DMSO-d6)
8 5.13 (d, J = 7.8 Hz, 4H), 5.36 (d, J = 10.7 Hz, 2H), 6.87 (d, J = 7.6 Hz,
2H), 7.38 (m, 10H), 7.82 (d, J = 7.6 Hz, 2H), 10.51 (s, 1H); ~3C NMR
(DMSO-d6) 868.65 (d, J = 22.3 Hz), 68.73 (d, J = 25.8 Hz), 115.39, 125.14,
127.81, 128.30, 128.41, 130.34, 136.0l (d, J = 29.8 Hz), 162.65, 190.80
(d, J = 16.5 Hz); 31p NMR (DMSO-d6) 8 0.56; IR (CHC13) 3212, 1684,
1600, 1456, 1373, 1281, 1115, 984 cm -1. Exact mass calculated for C22H2106P:
412.1154. Found: 412.1163. Analysis calculated for C22H2106P: C, 64.08; H,
5.13. Found: C, 64.46; H, 5.18.
p-Hydroxyphenacyl Dibenzyl Phosphate Ethylene Ketal (15). 27 The
ketal is prepared by modification of the method of Amos and Ziegler. 38
To the solution of previously dried p-hydroxyphenacyl dibenzyl phosphate
(14; 3.45 g, 8.4 mmol) in 100 ml of dry benzene, a catalytic amount of p-
toluenesulfonic acid (p-TsOH) (79.9 mg, 0.42 mmol) is added, followed by
the addition of excess amount of ethylene glycol (1.93mi, 2.15g,
34.6 mmol). The solution is refluxed for 5 hr with a Dean-Stark apparatus
to remove the water that is generated (CaSO4 is placed in the receiving
arm to further trap H20). When the reaction is complete, NaHCO3 (0.7l g,
8.4 mmol) is added to neutralize the mixture. The benzene is removed in
vacuo, ethylene glycol is removed by extraction with H20/ethyl acetate to
yield a white solid. Further purification is done by recrystallization from
Et20 to give 3.19 g (82.1%) of p-hydroxyphenacyl dibenzyl phosphate ethyl-
ene ketal (15): mp 97-98°; 1H NMR (DMSO-d6) 8 3.80 (m, 2H), 4.00 (m,
38 (a) A. A. Amos and P. Ziegler, Chem. Ber. 37, 345 (1959); and (b) R. E. Ireland and
D. M. Walba, Org. Syn. Coll. 6, 567 (1988).
2H), 4.03 (d, J = 6.7 Hz, 2H), 4.95 (d, J = 7.7 Hz, 4H), 6.74 (d, J = 8.6 Hz,
2H), 7.27 (d, J = 8.6 Hz, 2H), 7.33-7.38 (m, 10H), 9.53 (s, 1H); 13C NMR
(DMSO-d6) ~ 64.91, 68.34 (d, J = 22.7 Hz), 69.03 (d, J =- 24.9 Hz), 107.06
(d, J = 33.4 Hz), 114.71, 127.21, 127.71, 128.28, 128.40, 129.11, 135.96 (d,
J = 28.3 Hz), 157.60; 31p NMR (DMSO-d6) ~ -0.02; IR (CHCI3) 3280, 1610,
1512, 1455, 1273, 1168, 1002 cm -1. Exact mass calculated for C24H25OTP:
456.1416. Found: 456.1400. Analysis Calculated for C24H2507P: C, 63.16;
H, 5.52. Found: C, 63.00; H, 5.49.
p-Hydroxyphenacyl Diammonium Phosphate (16). 27 The hydrogeno-
lysis is performed according to the procedure of Heathcock. 39 To the ethyl-
ene ketal of p-hydroxyphenacyl dibenzyl phosphate (15; 1.86 g, 4.1 mmol)
dissolved in 30 ml of methanol is added 186 mg of 10% Pd/C. The solution
is hydrogenated under 10 psi pressure of H2 with stirring for 2 hr followed
by the addition of 1% HC1 (1 ml). After filtration, the filtrate is evaporated
in vacuo. The resulting viscous liquid is loaded on 5 g of DEAE-Sephadex
(A-50-120; Sigma, St. Louis, MO) pretreated with 10 mM of ammonium
acetate. The column is eluted with stepping gradient ammonium acetate
solution (200 ml each of 0.0 to 0.15 M solution with 0.05 M increments,
then 800 ml of 0.20 M solution to obtain the product). The diammonium
salt (16) is eluted at 0.2 M ammonium acetate. Excess ammonium acetate
and water are lyophilized off, and the resulting solid is recrystallized from
H20/methanol (1:1) to give 1.05 g (96.2%) of p-hydroxyphenacyl diam-
monium phosphate (16). Mp 177-192 ° (dec); aH NMR (D20) 6 5.08
(d, J = 5.6 Hz, 2H), 6.92 (d, J = 8.8 Hz, 2H), 7.89 (d, J = 8.8 Hz, 2H);
13C NMR (D20) 6 69.37 (d, J = 14.2 Hz), 118.46, 128.59, 133.42, 165.03,
200.50 (d, J = 33.3 Hz); 31p NMR (D20) ~ 1.18, IR (KBr) 3350, 3088, 2950,
1679, 1594, 1570, 1440, 1268, 1090 cm-1; UV-vis (CH3CN/H20) Ama~(e)220
(8400), 282 (14,000). Analysis Calculated for CsHasN206P: C, 36.10; H, 5.68;
N, 10,52. Found: C, 35.82; H, 5.38; N, 10.33.
Coupling of p-Hydroxyphenacyl Phosphate with ADP

p-Hydroxyphenacyl Adenosine 5'-Triphosphate, Triammonium Salt
(17). 27 The coupling reaction is modeled after the method of Hoard and
Ott. 4° Dowex 50W-X8 resin (Baker Chemical Co., Phillipsburg, NJ) (10 g)
in a sintered glass funnel is treated with 5% hydrochloric acid (20 ml x 2)
and washed with water until the filtrate is neutral. The resin is then treated
with 20% aqueous pyridine solution and washed with water until the filtrate
is neutral. A solution of adenosine 5'-diphosphate (ADP, potassium salt,
39 C. H. Heathcock and R. Ratcliffe, J. Am. Chem. Soc. 93, 1746-1757 (1971).

40 D. E. H o a r d and D. G. Ott, J. Am. Chem. Soc. 87, 1785-1788 (1965).
0.385 g, 0.797 mmol) in 7 ml of water is stirred with the resin for 10 rain;
the resin is filtered and washed with water (7 ml x 4). Into the filtrate is
added tributylamine (0.38 ml, 0.295 g, 1.59 mmol) followed by stirring for
45 min. The solution is evaporated to dryness and then lyophilized to
yield a yellow oily residue, which is further dried by sequential cycles of
dissolution and evaporation of dry pyridine (3.5 ml x 2) and dry dimethyl-
formamide (DMF, 3.5 ml x 2). The dried residue is redissolved in 3.5 ml
of dry DMF, carbonyl diimidazole (0.555 g, 3.42 mmol) is added, and the
mixture is stirred under argon for 24 hr at ambient temperature. The reac-
tion is quenched by adding methanol (0.11 ml, 2.8 mmol), the resulting
solution is stirred for 3 hr and pumped to dryness. Meanwhile, a solution
of p-hydroxyphenacyl diammonium phosphate (16; 400 mg, 1.5 mmol) in
13 ml of water is stirred for 45 min with the activated Dowex 50W-X8 resin
(10 g, pyridinium form) as described earlier. The resin is filtered off and
washed with water (7 ml X 3). Tri-n-octylamine (0.535 g, 0.66 ml, 1.50
mmol) is added to the filtrate, most of the water removed in vacuo, and
the residue lyophilized to yield a yellow residue, which is subjected to
dissolution and evaporation of dry pyridine (5 ml x 2) and dry DMF (5
ml x 2). The gummy residue is redissolved in 3.5 ml of dry DMF and added
to the above ADP-imidazole solution. The DMF is pumped off at room
temperature. Hexamethylphosphoramide (HMPA, 10 ml) is added to the
pale yellow residue, and the solution is stirred under argon for 3 days, for
which it is sonicated for 20 min periods every 12 hr. The reaction is quenched
by adding 40 ml of water, and is washed with chloroform (35 ml x 4)
and hexane (35 ml). The aqueous layer is lyophilized and purified by a
diethylaminoethyl (DEAE)-cellulose (170 ml dry volume of Whatman DE
52 conditioned with 10 mM NH4HCO3 solution) column eluted with a
stepping gradient of 100 ml of ammonium bicarbonate solution with the
concentration from 0.0 to 0.25 M with 0.05 M increments followed by 700
ml of 0.3 M to elute the product. The fractions are monitored by TLC on
silica gel (iPrOH:0.30M aq NHaI-ICO3 (65:35)): Rf (ATP ester) --- 0.84;
Rf (p-hydroxyphenacyl diammonium phosphate) = 0.58. The fractions con-
taining the desired phototrigger are lyophilized to yield 0.528 g (71.4%
purity by mass: yield 68.4%) of p-hydroxyphenacyl adenosine 5'-triphos-
phate (HATP), triammonium salt (17): mp 137° (dec); 1H NMR (D20)
4.21 (m, 1H), 4.27 (s, 2H), 4.37 (t, J = 4.3 Hz, 1H), 4.44 (t, J = 21.5 Hz,
1H), 5.10 (m, 2H), 5.77 (d, J = 5.1 Hz, 1H), 6.52 (d, J = 8.5 Hz, 2H), 7.46
(d, J = 8.8 Hz, 2H), 8.02 (s, 1H), 8.29 (s, 1H); 13C NMR (D20) 8 68.12 (d,
J = 20.7 Hz), 70.87 (d, J = 19.4 Hz), 72.92, 77.77, 86.51 (d, J = 37.1 Hz),
90.20, 117.93, 121.14, 128.20, 133.00, 142.91, 151.02, 152.90, 156.33, 164.12,
197.95 (d, J = 33.2 Hz); 31p NMR (D/O) 8 -25.42 (t, J = 48.3 Hz), -14.09
24 CAGEDCOr~POUNDS [11
(d, J = 46.1 Hz), -13.74 (d, J = 49.4 Hz); IR (KBr) 3320, 3150, 2991, 2821,
1710, 1679, 1594, 1570, 1453, 1372, 1248, 1112, 1041 cm 1; UV-vis (CH3CN/
H20) Amax(e) 260 (30,500), 286 (14,600). Exact mass calculated for
C18H22OxsP3 (free acid): 641.0482. Found: 641.0460. The p-hydroxyphenacyl
diammonium phosphate was also recovered.
Synthesis of 2, 4'-Dihydroxyacetophenone 41
A solution of 500 mg (2.33 mmol) of 13 is dissolved in 30 ml benzene:
CH2C12 (1 : 1), and 0.1 ml (2.65 mmol) of formic acid is added. The resulting
mixture is cooled to 6°. At this temperature a solution of 0.45 ml (3.02
mmol) of 1,8-diazabicyclo[5.4.0]-7-undecene (DBU) in 5 ml benzene is
added dropwise over a period of 10 min. The reaction mixture turns light
yellow and is stirred for another 45 min at 6°. The mixture is allowed to
warm to room temperature and then is stirred until completion (monitoring
by TLC, 48 h). The volatiles are removed in vacuo, and the resulting 2-
formyloxy-4'-hydroxyacetophenone is used without further purification. It
is dissolved in 20 ml methanol containing 0.2 g (5 mmol) NaOH and stirred
for 4 h. After neutralization with 1N HCI, the solution is extracted four
times with CH2C12 (10 ml each). The organic layers are collected, dried,
and evaporated in vacuo. The product is purified by silica gel column
chromatography (CH2C12 :ethyl acetate, 9:1) to yield colorless crystals of
2,4'-dihydroxyacetophenone (171 mg, 1.13 mmol, 49%): mp 165-167°; 1H
NMR (D20) fi 7.82 (d, 2H, J = 8.7 Hz), 6.91 (d, 2H, J = 8.7 Hz), 4.89 (s,
2H). 13C NMR (CD3CN) fi 196.81,161.72, 129.84, 115.09, 64.55. IR (KBr)
1676 cm-1; mass spectrum m/z 152 (M+), 136, 121, 93, 65.
Synthesis of Desyl Caged Amino Acids 31,42

Synthesis of y-O-Desyl L-Glutamate (24a). 31,42 A solution of desyl bro-
mide (605 mg, 2.2 mmol), N-tert-BOC-L-glutamic acid, ot-tert-butyl ester 23
(672 mg, 2.2 mmol), and D B U (0.35 ml, 2.3 mmol) is heated at reflux in
benzene (25 ml) and stirred overnight to give y-O-desyl tert-butyl N-tert-
BOC L-glutamate in a yield of 1.06 g (96%) as a colorless oil: 1H NMR
(CDCI3) ~ 7.92 (d, J = 8.1 Hz, 2H), 7.51 (t, J = 7.4 Hz, 1H), 7.47-7.32 (m,
7H), 6.86 (s, 1H), 5.12 (dd, J = 13.6, 7.3 Hz, 1H), 4.5 (br s, 1H), 2.6 (m,
2H), 2.22 (m, 1H), 1.97 (m, 1H), 1.45 (s, 9H), 1.42 (s, 9H).
y-O-Desyl tert-butyl N-tert-BOC L-glutamate (1.06 g, 2.13 mmol) and
trifluoroacetic acid (TFA, 2.0 ml) are stirred at room temperature in dichlo-
41(a) G. Fodor, O. Kovacs,and T. Mercer, Acta Chim. Sci. Hung. 1, 395-402 (1951);and (b)
M. Patzlaff and W. Z. Barz, Z. Naturforsch. 33e, 675-684 (1978).
4zA. H. Jung, R. S, Givens, and C-H. Park, unpublished results, 1997-1998.
romethane (5 ml) and give, after Sephadex LH-20 purification, y-O-desyl

L-glutamate (24a) in a yield of 0.51 g (52%) as a hygroscopic colorless
powder; TFA is removed from the product salt during lyophilization: mp
118-120 ° (dec.); aH NMR (CDC13) 8 2.18 (m, 2H), 2.71 (m, 2H), 3.76 (q,
1H), 7.13 (s, 1H) 7.39-7.49 (m, 7H), 7.62 (t, 1H, J = 7.33 Hz), 7.98 (d, 2H,
J = 7.35 Hz); 13C NMR (D20) 828.39, 32.654, 56.786, 79.04, 130.95, 131.61,
131.71,131.81,131.89, 132.02, 132.43, 132.92, 135.45, 136.35, 137.82, 176.52,
176.68; mass spectrum m/z (relative intensity) 342 (M + + 1, 95), 230 (45),
212 (95), 195 (64); exact mass calculated = 341.1341; found = 341.1346.
UV (CH3CN/H20) /~max (8) 252 (9000), 300 (1000), 350 (200). Analysis
Calculated for C19H19NO5 • ½H20: C, 65.13; H, 5.75; N, 4.00. Found: C, 64.97;
H, 5.69; N, 3.88.
Synthesis of O-Desyl y-Aminobutyric Acid (22a). 31"42m solution of N-
tert-BOC y-aminobutyric acid (21; 500 mg, 2.46 mmol), desyl bromide (677
mg, 2.46 mmol), and DBU (396 mg, 2.60 mmol) were heated at reflux in
benzene (25 ml) to give N-tert-BOC O-desyl y-aminobutyrate in a yield of
0.94 g (96%) as a clear colorless oil: tH NMR (CDCI3) 8 7.92 (d, J = 7.8
Hz, 2H), 7.4 (m, 8H), 6.86 (s, IH), 4.7 (br s, 1H), 3.20 (d, J = 5.9 Hz, 2H),
2.53 (m, 2H), 1.87 (dt, J = 6.9, 2.8 Hz, 2H), 1.4 (br s, 9H).
A solution (0.93 g, 2.3 mmol) of the N-tert-BOC O-desyl y-aminobuty-
rate and TFA (8.0 ml, 100 mmol) in chloroform (25 ml) was stirred overnight
at room temperature. The volatiles are removed in vacuo, and benzene
(1 × 15 ml) is evaporated from the residue, leaving a colorless oil. This oil
is purified by chromatography on Sephadex LH-20, using water as the
eluant. The pure product fractions are combined and lyophilized to give
O-desyl y-aminobutyrate (22a, O-desyl GABA, 0.78 g; 81%) as a colorless
hygroscopic powder: mp 82-84°; 1H NMR (D20) 8 1.82 (m, 2H), 2.38 (m,
2H), 2.84 (t, 2H, J = 7.92 Hz), 6.88 (s, 1H), 6.99 (m, 5H), 7.12 (t, 1H), 7.25
(d, 2H, J = 7.23 Hz), 7.69 (d, 2H, J = 7.23 Hz); 13C NMR (D20) 8 24.6,
32.9, 41.1, 80.6, 131.1, 131.2, 131.3, 131.7, 132.0, 135.3, 136.1, 136.7, 175.9,
198.8. Analysis calculated for Cz0H20NOsF3: C, 58.39; H, 4.90; N, 3.40.
Found: C, 58.13, H, 4.75; N, 3.39.
Synthesis of y-O-(p-Hydroxyphenacyl) Glutamate (24b) 32"43

Synthesis of Protected y-O-(p-Hydroxyphenacyl) Glutamate. A suspen-
sion of the bromoketone 13 (1.24 g, 5.77 mmol) is stirred under reflux in
20 ml of benzene in order to dissolve the bromoketone, cooled to 6° and
a solution of 1.70 g (5.60 mmol) of N-tert-BOC-L-glutamic acid, ot-tert-butyl
43 R. S. Givens, C-H. Park, A. H. Jung, J. F. W. Weber, and W. Bartlet, Full paper in prepa-
ration.
26 CAGED COMPOUNDS [ II
ester 23 in 15 ml benzene is added. At this temperature a solution of 0.90

ml (6.02 mmol) of D B U in 10 ml benzene is added dropwise over a period
of 15 min. During the addition of the base, the reaction mixture turns
light yellow. After the addition, the mixture is allowed to warm to room
temperature and is stirred for 17 hr and the solvent removed in vacuo. The
product is purified by silica gel column chromatography (CH2Cl2:ethyl
acetate, 9:1) to yield a yellow solid (1.48 g, 3.38 mmol, 60%). 1H NMR
(CDC13) 8 1.42 (s, 9H), 1.44 (s, 9H), 2.00 (m, 1H), 2.21 (m, 1H), 4.25 (m,
1H), 5.23 (s, 2H), 6.86 (d, 2H, J = 8.4 Hz), 7.76 (d, 2H, J = 8.3 Hz).
Deprotection of Protected T-O-(p-Hydroxyphenacyl) Glutamate. A solu-
tion of the fully protected glutamate (1.23 g; 3.02 mmol) is treated with 10
ml TFA at 4° and stirred for 3 hr at this temperature. Evaporation of the
T F A in vacuo is followed by the addition of 30 ml ethyl acetate and 30 ml
H20. The layers are separated and the aqueous layer is evaporated in
vacuo. Treating the oily residue with 25 ml acetonitrile : H20 immediately
causes a precipitation of colorless cyrstals of 24b (447 mg, 1.59 mmol, 53%),
mp 187-188 ° (decomposition). aH NMR (D20) 8 2.19 (m, 2H), 2.27 (m,
2H), 3.81 (m, 1H), 5.47 (s, 2H), 6.95 (d, 1H, J = 8.8 Hz), 7.89 (d, 2H, J =
8.8 Hz).
General Procedure for Photolysis

Acetonitrile (high-performance liquid chromatography, HPLC grade)
is used without further purification. All water is distilled and passed through
a Nanopure deionizing system (Sybron/Barnstead, Inc., Boston, MA). Phos-
phate buffer for analytical HPLC is prepared using 85% phosphoric acid,
tetrabutylammonium phosphate (TBAP, 0.5 mM), and potassium hydrox-
ide to afford a solution of 0.04 M at pH 6. The HPLC system employed
consists of two pumps, a controller, an autoinjector fitted with a 10-/zL loop,
a UV-VIS spectrophotometric detector set at 254 or 280 nm, a recorder, and
a Ca8 5/zm 250 x 4.6-nm C18 reversed phase column. All analytical HPLC
analyses employ either a solvent gradient or an isocratic elution with a
flow rate of 1.0-1.5 ml/min. Photolysis is performed on a merry-go-round
apparatus equipped with 16 x 300-nm or 4 x 350-nm lamps. The light
output for the determination of quantum yields is measured using the
potassium ferrioxalate method, an
Into a 20 x 180-mm Pyrex tube is placed 5 or 10 ml of an aqueous or
alcoholic solution containing the phosphate ester and an appropriate inter-
nal standard. The concentration of the phosphate is adjusted to assure
complete absorption of the incident radiation, i.e., greater than 3 absorbance
44 C. G. Hatchard and C. A. Parker, Proc. R. Soc. London A-235, 518-522 (1956).

[1] NEWPHOTOPROTECTINGGROUPS 27
units at the excitation wavelength. The tube is sealed with a septum, deaer-
ated with argon for at least 20 min at 0°, and photolyzed in the apparatus
previously described. A 100-/zl aliquot is removed periodically, stored in
the dark, and analyzed by HPLC.
Photolysis of Desyl Adenosine Cyclic 3',5'-phosphate (7a) atpH 7.2. Into
a 5-ml volumetric flask is placed 42.8 mg (0.032 mmol) of a diastereomeric
mixture of desyl adenosine cyclic 3',5'-phosphate (Ta). The cyclic triester
is diluted to 5 ml with 1 : 1 Tris buffer/dioxane solution. To each of five
10 × 75-mm Pyrex tubes was added 1 ml-aliquot of the stock solution. The
solutions are deaerated, irradiated, and prepared for HPLC and 31p NMR
analysis as previously described. The apparent pH of the solution before
and after photolysis is measured to be 7.2. The quantum efficiencies
for the disappearance of desyl adenosine cyclic 3',5'-phosphate (Ta), (I)di s =
0.39 _ 0.04, the appearance of adenosine cyclic 3',5'-phosphate, qbapp =
0.34 - 0.03, and the appearance of 2-phenylbenzo[b]furan (6a), d~app =
0.19 _ 0.03.
In a separate experiment, 85.6 mg (0.16 mM) of desyl adenosine cyclic
3',5'-phosphate (7a) is dissolved in 10 ml of a 1 : 1 mixture of dioxane; 0.05
mM tris (in D20, pH adjusted to 7.0 with NaOH/D20). Into a series of
NMR tubes is placed 1.0 ml of the desyl cAMP solution, deareated with
argon, and photolyzed in the merry-go-round apparatus with 4× R P R
3500-~, lamps. Sample tubes are removed periodically and examined by
31p NMR (see Fig. 1). Chemical shifts are reported relative to 85% H3PO4
as an external standard.
Photolysis of Desyl cAMP (7a) at Various pH Values. The cyclic
triester 9 (50.3 mg, 0.10 mmol) is dissolved in 3.0 ml of dioxane. Into
each of six 10 × 75-mm Pyrex tubes is placed 0.5 ml of the desyl cAMP
(Ta) stock solution, and the volume is brought to 1.0 ml by the addition
of 0.5 ml of one of the following buffer solutions: 0.05 M phosphate
buffer (pH 7, H20), 0.05 M phosphate buffer (pH 7, D20), 0.05 M Tris
buffer (pH 7.2, H20), 0.05 M Tris (pH 7.2, D20), 0.05 M perchloric acid
(pH 2.0, DzO). The pH values of each final solution are determined. The
solutions are deaerated with argon for 20 min and irradiated for 10 min
with 4× RPR 3500-,~ lamps. Prior to HPLC analysis, 0.5 ml from each
photolysis mixture is added to 0.25 ml of a solution containing 1,4-
diphenyl-l,3-butadiene as an internal standard. Analysis is carried out as
described in the General Methods section. The quantum efficiencies for
appearance of 6a are constant at 0.17 ___ 0.02 at all pH values. The
disappearance of 7a and appearance of cAMP are also constant at 0.38
_ 0.02 and 0.34 ___ 0.02, respectively. These two cannot be measured in
the two phosphate buffers, however, due to interferences with the 31p
signals from the buffer.
Photolysis of p-Hydroxyphenacyl Diammonium Phosphate (16). 27,35
Into five Pyrex tubes containing 10 ml of 10% CH3CN in buffer (pH 7.3)
are introduced 100 mg (0.38 mmol) of p-hydroxyphenacyl diammonium
phosphate (16) and 0.2 g (1.88 mmol) of anisole as an internal standard.
The solution is deaerated and irradiated at 300 nm for 30 min. Aliquots
are taken at 5-min intervals. Each aliquot is analyzed by HPLC. The only
photoproduct, p-hydroxyphenylacetic acid (16), is identified by injection
on HPLC with an authentic sample. The quantum efficiency for the disap-
pearance of p-hydroxyphenacyl diammonium phosphate (16) is ~dis =
0.38 and for the appearance of p-hydroxyphenylacetic acid (lib, R' -- H),
~rearr = 0.12.
Photolysis of p-Hydroxyphenacyl Adenosine 5'-Triphosphate, Triam-
monium Salt (17). 27 Into three Pyrex tubes containing 5 ml of Tris buffer
(0.05 M, pH 7.3) are introduced 20 mg (28.9 ~mol) of p-hydroxyphenacyl
adenosine 5'-triphosphate, triammonium salt (17) and 15.3 mg (125 ~mol)
of benzoic acid as an internal standard. The solution is deaerated and
irradiated at 300 nm for 12 min. Aliquots are taken at 3-min intervals
and analyzed by HPLC. p-Hydroxyphenylacetic acid (17) and ATP are
identified as photoproducts by injection on HPLC with authentic samples.
The quantum efficiency for the disappearance of triammonium salt (17) is
~ i s (17) -- 0.37, and for the appearance of p-hydroxyphenylacetic acid
( l l b ) is (I)rearr ( l l b ) = 0.31, and for ATP, (I)AT P = 0.30.
Photolysis of p-Hydroxyphenacyl Adenosine 5'- Triphosphate, Triam-
monium Salt (17), in Lactated Ringer's Solution. Into a Pyrex tube con-
taining 5 ml of Ringer's solution (each 100 ml contains 5 g of dextrose
hydrous, 600 mg of sodium chloride, 310 mg of sodium lactate, 30 mg of
potassium chloride, and 20 mg of calcium chloride, pH 6.5) is introduced
20 mg (28.9/zmol) of p-hydroxyphenacyl adenosine 5'-triphosphate, triam-
monium salt (17). The solution is deaerated and irradiated at 300 nm for
12 min. The resulting solution is taken and prepared for HPLC analysis.
p-Hydroxyphenylacetic acid ( l l b ) and ATP are identified as photoproducts
by injection on HPLC with authentic samples.
Competitive Quenching Study of p-Hydroxyphenacyl Adenosine 5'-Tri-
phosphate (17) with Sodium 2-Naphthalene Sulfonate. To a 10-ml volumetric
flask is added 15 mg (21.7/xmol) of p-hydroxyphenacyl adenosine 5'-tri-
phosphate, triammonium salt (17), 61 mg (0.5 mmol) of benzoic acid, and
Tris buffer (1 M, pH 7.3) to the fill line. To four Pyrex tubes is added
1 ml of the above solution. To three of these tubes are added, respectively,
11.5 (50 /zmol), 23.0 (100 /zmol), and 34.5 mg (150/zmol) of sodium 2-
naphthalene sulfonate. The tubes are deaerated for 20 min with argon at
room temperature and photolyzed using 4x RPR 350-nm lamps; the reac-
tions are monitored by HPLC. A least-squares analysis is performed on
the data obtained as shown in the tabulation.
Parameter a dPdis(17) t~rearr(llb) (I)ATP
Sodium
naphthalene
sulfonate [mM]
0 0.37 (0.04) 0.31 (0.03) 0.30 (0.03)
5O 0.31 0.25 0.25
100 0.27 0.20 0.23
150 0.24 0.16 0.19
Slope (K~v)(M-1) 3.56 6.38 3.61

r (nsec) 0.54 (0.09) 0.97 (0.10) 0.55 (0.06)
k, 108 sec -1 6.90 (1.15) 3.20 (0.33) 5.50 (1.00)
a Error shown in parentheses.
Stability Test of p-Hydroxyphenacyl Adenosine 5'-Triphosphate, Triam-

monium Salt (17), in Various Media. Into six 10-ml test tubes, respectively,
containing 5 ml of H 2 0 , D 2 0 , Tris buffer (50 mM, pH 7.3), Ringer's solution
(I) (each 100 ml contains 600 mg of sodium chloride, 30 mg of potassium
chloride, and 20 mg of calcium chloride, pH 6.5), Ringer's solution (II)
[each 100 ml contains (I) and 310 mg of sodium lactate, pH 6.5], Ringer's
solution (III) [each 100 ml contained (II) and 5 g of dextrose hydrous, pH
6.5] is introduced 5 mg (7.2 tzmol) of p-hydroxyphenacyl adenosine 5'-
triphosphate, triammonium salt (17). The stability of 17 in six different
media was checked by HPLC analysis. In all cases 17 was stable at least
24 hr or longer.
Acknowledgments
We thank the National Science Foundation (Grant NSF/OSR 9255223), the Petroleum
Research Fund (Grant 17541-AC4) of the American Chemical Society, and the University
of Kansas for financial support. In addition to the co-workers cited in references to our work,
we thank the following for assistance in improving the synthetic and photochemical procedures
reported herein: Sabine Amslinger, Christiane Burzik, and Russell Herpel.
30 CAGEDCOMPOUNDS [21
[21 S y n t h e s i s , P h o t o c h e m i s t r y , a n d B i o l o g i c a l
Characterization of Photolabile Protecting Groups for
Carboxylic Acids and Neurotransmitters
By KYLE R. GEE, BARRY K. CARPENTER,and GEORGE P. HESS
Introduction
The utility of biologically active compounds whose activity is tempo-
rarily masked by strategic attachment of a photolabile, or "caging," group
continues to growJ -6 The main advantages of reagent application in a
biological system by UV photolysis of caged precursors, as opposed to
conventional mixing methods, include much better temporal (-1000-fold)
and spatial resolution of reagent introduction. 4-1° This becomes important
when the processes being studied are very rapid, such as in receptor activa-
tion u and inhibition. 12'13 The spatial resolution of concentration jumps
afforded by the photolysis of caged compounds is also being exploited in
complex systems, such as receptor mapping in cells,7 neuronal mapping in
tissue slices, 8'9and in known circuits of cells controlling pharyngeal pumping
in the nematode Caenorhabditis elegans.10
Attachment of a caging group requires the presence of an appropriate
functionality in the compound of interest. Phosphates, carboxylates, phe-
nols, thiols, amines, and amides have all been shown to be suitable sites
of attachment of a caging group whose subsequent photolysis liberates the
original functionality and thus the native bioeffector molecule. 1 Carboxyl-
ares are functional groups common to a large number of natural and unnatu-
ral bioeffector molecules, including amino acid neurotransmitters, and thus
1 j. E. T. Corrie and D. R. Trentham, in "Bioorganic Photochemistry" (H. J. Morrison, ed.),

Vol. 2, p. 243, Wiley, New York, 1993.
2 S. R. Adams and R. Y. Tsien, Annu. Rev. Physiol. 55, 755 (1993).
3 j. M. Nerbonne, Curr. Opin. Neurobiol. 6, 379 (1996).
4 G. P. Hess and C. Grewer, Methods Enzymol. 291, [25] (1998) (this volume).
5 G. P. Hess, Biochemistry 32, 989 (1993).
6 G. P. Hess, L. Niu, and R. Wieboldt, Ann. N.Y. Acad. Sci. 757, 23 (1995).
7 W. Denk, Proc. Natl. Acad. Sci. U.S.A. 91, 6629 (1994).
8 M. B. Dalva and L. C. Katz, Science 265, 255 (1994).
9 A. Sawatari and E. M. Callaway, Nature 380, 442 (1996).
10H. Li, L. Avery, W. Denk, and G. P. Hess, Proc. Natl. Acad. Sci. U.S.A. 94, 5912 (1996).
11 N. Matsubara, A. P. BiUington, and G. P. Hess, Biochemistry 31, 5507 (1992).
12L. Niu and G. P. Hess, Biochemistry 32, 3831 (1993).
13 L. Niu, L. G. Abood, and G. P. Hess, Proc. Natl. Acad. Sci. U.S.A. 92, 12008 (1995).

[2] PHOTOLABILE PROTECTING GROUPS 31
general methods for caging carboxylates will prove useful for preparing a
wide range of caged bioeffectors.
Caged carboxylates were among the first described in the literature 14-16
on photosensitive protecting groups, although it is only recently that their
utility has been exploited for studies of rapid processes in biological systems.
The first report of caged versions of biologically relevant carboxylates came
from Wilcox et al., 17 who described making 4,5-dimethoxy-2-nitrobenzyl
(DMNB) and 1-(4,5-dimethoxy-2-nitrophenyl)ethyl (DMNPE) esters of the
amino acid neurotransmitters aspartate, glutamate, glycine, and ~/-aminobu-
tyric acid (GABA). These photosensitive esters were designed to be used
in the study of rapid neuronal receptor activation processes, but slow photol-
ysis rates (time constant - 1 sec for the production of free amino acid)
limited their use for such studies. The synthesis of photolabile o-nitrobenzyl
esters of the amino acids serine, TM alanine, leucine, and lysine 19 was subse-
quently reported, but again the esters were of limited utility because of
suboptimal photolysis parameters.
The use of the ot-carboxy-o-nitrobenzyl (CNB) protecting group 2° has
resulted in the preparation of caged compounds that are (i) water soluble
at neutral pH, (ii) biologically inactive prior to photolysis, (iii) photolyzed
in a wavelength region not deleterious to cells or the reaction being investi-
gated (A > 335 nm), (iv) photolyzed to the biologically active compound
in the useful microsecond time region with sufficient quantum yield to allow
kinetic investigations over a wide concentration range of the liberated
biologically active compounds, and (v) give rise to photolysis products that
are not deleterious to the cells or the reaction being investigated. The
compounds caged with the CNB group that meet all these criteria are
carbamoylcholine 2° and the neurotransmitters that were caged as CNB
esters: the inhibitory neurotransmitter G A B A 21 and the excitatory neuro-
transmitters glutamate 22 and kainic acid. 23 The photolysis of these caged
neurotransmitters generated the free amino acids in 20-80/xsec with quan-
14j. A. Barltrop, P. J. Plant, and P. Shofield, Chem. Commun. p. 822 (1966).

15 A. Patchornik, B. Amit, and R. B. Woodward, J. Am. Chem. Soc. 92, 6333 (1970).
16j. C. Sheehan, R. M. Wilson, and A. W. Oxford, J. Am. Chem. Soc. 93, 7222 (1971).
17M. Wilcox, R. W. Viola, K. W. Johnson, A. P. Billington, B. K. Carpenter, J. A. McCray,
A. P. Guzikowski, and G. P. Hess, J. Org. Chem. 55, 1585 (1990).
18 M. C. Pirrung and D. S. Nunn, Bio. Med. Chem. Lett. 2, 1489 (1992).
19 H. Aoshima, D. Tanaka, and A. Kamimura, Biosci. Biotech. Biochem. 56, 1086 (1992).
20 T. Milburn, N. Matsubara, A. P. Billington, J. B. Udgaonkar, J. W. Walker, B. K. Carpenter,
W. W. Webb, J. Marque, W. Denk, J. A. McCray, and G. P. Hess, Biochemistry 28, 49 (1989).
21 K. R. Gee, R. Wieboldt, and G. P. Hess, J. Am. Chem. Soc. 116, 8366 (1994).
22 R. Wieboldt, K. R. Gee, L. Niu, D. Ramesh, B. K. Carpenter, and G. P. Hess, Proc. Natl.
Acad. Sci. U.S.A. 91, 8752 (1994).
23 L. Niu, K. R. Gee, K. Schaper, and G. P. Hess, Biochemistry 35, 2030 (1996).
32 CAGED COMPOUNDS [2l
tum yields of 0.15-0.43. Such rapid activation allows these compounds to
be used in the study of fast receptor activation, u'2° inhibition, 12"13'24and
desensitization steps. 2°'25'26 However, caging the /3-carboxylate group of
N-methyl-D-aspartic acid (NMDA) resulted in a compound that was not
biologically inert, z7
Another variant of the nitrobenzyl ester, the 2,2'-dinitrobenzhydryl
(DNB) ester of NMDA, photolyzes on the order of 5/zsec with a quantum
yield of 0.18. 28 Following up on the favorable photolysis properties of
2-methoxy-5-nitrophenyl (MNP) acetate, 29 the MNP ester of the neuro-
transmitter glycine was prepared, but it exhibited a limited stability at
neutral pH. 3° The MNP ester of/3-alanine is rapidly photolyzed to /3-
alanine, which activates the glycine receptor, and has properties suitable
for investigation of the glycine receptor. 31 The MNP esters of both glycine
and/3-alanine photolyze in less than 2/zsec with quantum yields of 0.2.
Benzoinyl, or "desyl," esters of G A B A and glutamate were reported with
photolysis rate constants on the order of 107 s e c 1 and quantum yields of
0.14-0.3. 32 The suitability of these two compounds for investigations of the
G A B A and glutamate receptors is not yet known.
This chapter is concerned with the synthesis and purification of examples
of carboxylates caged as CNB, 2,2'-dinitrobenzhydryl (DNB), desyl, MNP,
and DMNPE esters, as well as the photochemical and biological character-
ization of representative examples of these caged carboxylates.
Synthesis of Caged Carboxylates

The general strategy for making photosensitive esters involves (A) car-
bodiimide-mediated coupling of carboxylic acids with photosensitive alco-
hols, (B) nucleophilic displacement of a halogen anion from an o-nitroben-
zyl or desyl halide by a carboxylate anion, or (C) coupling of a carboxylic
acid with an a-diazo-o-nitrophenylethyl compound. This last strategy was
24 L. Niu, C. Grewer, and G. P. Hess, Techn. Protein Chem. VII, 139 (1996).
25 L. Niu, R. W. Vazques, G. Nagel, T. Friedrich, E. Bamberg, R. E. Oswald, and G. P. Hess,
Proc. Natl. Acad. Sci. U.S.A. 93, 12964 (1996).
26 T. Otis, S. Zhang, and L. O. Trussell, J. Neurosci. 16~ 7496 (1996).
z7 K. R. Gee, L~ Niu, K. Schaper, and G. P. Hess, J. Org. Chem. 60, 4260 (1995).
28 K. R. Gee, L. Niu, K. Schaper, and G. P. Hess, submitted (1998).
29 p. Kuzmic, L. Pavlickova, and M. Soucek, Coll. Czech. Chem. Commun. 5L 1293 (1986).
3o D. Ramesh, R. Wieboldt, L. Niu, B. K. Carpenter, and G. P. Hess, Proc. Natl. Acad. Sci.
U.S.A. 90, 11074 (1993).
31 L. Niu, R. Wieboldt, D. Ramesh, B. K. Carpenter, and G. P. Hess, Biochemistry 35,
8136 (1996).
32 K. R. Gee, L. W. Kueper, III, J. Barnes, G. Dudley, and R. S. Givens, J. Org. Chem. 61,
1228 (1996).
initially described by Walker et aL 33 for preparing caged versions of biologi-
cally active phosphates such as ATP.
Reagents
All starting materials are available from Aldrich (Milwaukee, WI) as
are anhydrous solvents, which are handled under nitrogen or argon using
standard syringe techniques. Calcium ionophore A23187 and protected
amino acids are available from Sigma (St. Louis, MO); NMDA is from
Research Biochemicals, Inc. (Natick, MA).
Chromatography
Flash chromatography is performed essentially according to the method
of Still et aL 34 For reversed-phase chromatography, Sephadex LH-20 is
employed. The resin is slurried in the solvent noted (usually water) and
allowed to swell, followed by elution of the sample under gravity. Product
fractions are pooled and the organic solvent (if any) is removed by rotary
evaporation, and the remaining aqueous solution is frozen and lyophilized.
Analytical thin-layer chromatography (TLC) is performed on aluminum-
backed silica gel plates in the indicated solvent systems.
fl-O-2,2'-Dinitrobenzhydryl-N-methyl-D-aspartic Acid
The preparation of fl-O-DNB-NMDA (5) (Scheme 1) requires the pro-
tected carbamate N-t-BOC NMDA (1). Initially this carbamate was pre-
pared by acylation of NMDA with tert-butyl pyrocarbonate in water/diox-
ane mediated by sodium hydrogen carbonate. 27 However, this method gives
variable yields and is somewhat time-consuming. Pretreatment of NMDA
with bistrimethylsilyltrifluoroacetamide (BTTFA) in acetonitrile, followed
by tert-butyl pyrocarbonate, affords the carbamate reproducibly in high
yield in much less time. To a mixture of NMDA (1.0 g, 6.8 mmol) in
acetonitrile (anhydrous, 32 ml) is added BTTFA (5.5 ml, 6.8 mmol). The
resulting mixture is stirred for 30 min, then di-tert-butyl pyrocarbonate
(2.2 g, 10 mmol) is added. The resulting slightly hazy mixture is stirred
overnight, then quenched by adding 15 ml methanol. The resulting mixture
is concentrated in vacuo and then dissolved in 2.5% sodium bicarbonate
(60 ml). The resulting solution is extracted with ether (3 x 30 ml), then its
pH (9) carefully lowered to 1.8 by dropwise addition of 5% HC1. The
resulting solution is extracted with ethyl acetate (3x 35 ml). The extract
33 j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. U0,

7170 (1988).
34W. C. Still, M. Kahn, and A. Mitra,J. Org. Chem. 43, 2923 (1978).
O OH NO2 O
OH ~ DNBO
HO +2 CH2C12
N O N O
H3C / ~'BOC H3C / ~'BOC
1 2 3
NO2
1 EquivalentNaOH
DNB = ~ dioxane/water
0 O2N'x]~ 0
TFA HO~-(~]~ ODNB
( CH2Cl2
NH o / jNO N 0
H3C [ ~ H3C / ~ B O C
5
SCHEME 1
is dried over sodium sulfate, then concentrated to N-t-BOC NMDA as 1.37

g (82%) of a colorless solid. The carbamate thus obtained is identical to
that obtained using the previously described method. 27
To make the title compound, follow Scheme 1. A mixture of N-t- BOC
NMDA (1, 0.77 g, 3.1 mmol), 2,2'-dinitrobenzhydro135 (2,1.70 g, 2.20 mmol),
1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride (EDC)
(1.21 g, 6.3 mmol), and catalytic amounts of 4-dimethylaminopyridine
(DMAP) and 1-hydroxybenzotriazole (HOBt) in anhydrous dichlorometh-
ane (35 ml) are prepared in the cold and allowed to warm to room tempera-
ture with stirring. A small amount (10-15%) of monoester is formed, as
indicated by TLC analysis of the reaction mixture (Rf0.35), but the predomi-
nant product is the bisester 3 (Rr0.77), easily distinguished from the starting
alcohol (Rf 0.63) on TLC (chloroform/methanol/acetic acid, 50:5 : 1). After
aqueous workup, flash chromatography (using ethyl acetate/hexanes as
eluant) gives the starting alcohol 2 as 0.77 g (45%) and the bisester 3 as
0.82 g (35%) of a colorless oil: 1H NMR (CDCI3) ~ 8.1 (m, 4H), 7.91 (m,
35R. B. Johns and K. R. Markham,J. Chem. Soc. p. 3712 (1962).

[9.] PHOTOLABILE PROTECTING GROUPS 35
2H), 7.55 (m, 8H), 7.36 (m, 3H), 7.14 (m, 1H), 4.92 (t, 1H), 3.16 (m, 1H),
2.97 (m, 1H), 2.80 (s, 3H), 1.4 (s, 9H). Analysis calculated for C36H33N5014:
C, 56.92; H, 4.38; N, 9.22. Found: C, 56.84; H, 4.67; N, 8.87. The a-carboxylate
4 is selectively generated by treatment of a solution of 2 (0.80 g, 1.0 mmol)
in dioxane with 1.0 equivalent of aqueous sodium hydroxide. The reaction
is slow and never reaches completion; if the reaction is pushed by addition
of more NaOH, some of the/3-carboxylate is generated as well as some
nonchromophoric by-products. After 2 days at room temperature, the reac-
tion solution is neutralized with aqueous citric acid and concentrated in
vacuo, followed by flash chromatography (15% methanol/chloroform) to
yield the monoester 4 (0.17 g, 31%) as well as 2,2'-dinitrobenzhydrol (2)
and bisester 3 (0.29 g, 31%), which is subjected to a second round of alkaline
treatment to make more of 4 (80 mg, 30%): 1H NMR ( C D C I 3 ) 8.07 (m,
2H), 7.90 (br s, 1H), 7.52 (m, 4H), 7.31 (m, 2H), 7.16 (m, 2H), 4.48 (t, J =
6.5 Hz, 1H), 3.15 (m, 1H), 2.99 (dd, J = 16.4, 7.8 Hz, 1H) 2.82 (s, 3H), 1.34
(s, 9H). Analysis calculated for C23HzsN3010: C, 54.87; H, 5.01; N, 8.35.
Found: C, 54.99; H, 5.19; N, 7.79. High-performance liquid chromatography
(HPLC) analysis indicates <2% of the/3-carboxylate. The amino group in
4 (0.23 g, 0.46 mmol) is deprotected by brief treatment (10 min) with
trifluoroacetic acid (TFA) (2 ml) in dichloromethane (4 ml). After concen-
tration the residue is purified by reversed-phase chromatography using
water: dioxane (1 : 1, v/v) as eluant, giving the title compound (5) as 192
mg (81%) of a colorless powder: aH NMR [deuterated dimethyl sulfoxide
(DMSO-d6)] a 8.12 (td, J -- 7.9, 2.7 Hz, 2H), 7.79 (m, 2H), 7.68 (m, 4H),
7.49 (d, J = 8.0 Hz, 1H), 3.46 (t, J = 6.4 Hz, 1H), 2.97 (dd, J = 16.5, 6.4
Hz, 1H), 2.76 (dd, J = 16.5, 6.2 Hz, 1H), 2.45 (s, 3H); 19F (DMSO-d6) 68.9.
Analysis calculated for C18HavN308 • 0.5CF3COzH: C, 50.02; H, 3.88; N,
9.27. Found: C, 50.26; H, 3.74; N, 9.37.
ot-Carboxy-2-nitrobenzyl-y-aminobutyrate
The simplest way to make CNB esters (8) is to react carboxylic acids
with tert-butyl o~-bromo-2-nitrophenyl acetate (BBNA) (Scheme 2) in the
presence of DBU, followed by TFA treatment to deprotect the caging
group carboxylate and any tert-butyl carbamate(s) present. The method
initially described 22 for making BBNA (radical bromination of tert-butyl
2-nitrophenyl acetate) is somewhat time-consuming and involves careful
flash chromatographic separation of BBNA from the starting ester, as they
elute very close to each other on silica gel. A better route to BBNA involves
simple esterification of ot-bromo-2-nitrophenylacetic acid (BNPA) 36 with
36 C. Chang, B. Niblack, B. Walker, and H. Bayley, Chem. Biol. 2, 391 (1995).

36 CAGEDCOMPOUNDS 121
Br
o
HO/~/.x.~NHBOC + CO2-t-Bu D B U ) ~ N H B O C
C6H6 O
v "NO2
BBNA v "NO2
6 7
HOyo 0
TFA_F~)cH2C~I2 O ~ NH 2 . CF3CO2H
v "NO2
8
SCHEME2
N,N'-diisopropyl-tert-butylisourea. 37 Although tert-butyl ester formation

via the isourea route generally affords tert-butyl esters in only 40-50%
yield, the method is still a significant improvement over that originally
described for BBNA when overall ease and speed of synthesis are consid-
ered. Thus, to a solution of BNPA (0.50 g, 1.9 mmol) in dichloromethane
(15 ml) is added isourea (1.5 g, 7.5 mmol). The resulting mixture is stirred
overnight at room temperature, diluted with ethyl acetate (60 ml), and
filtered. The filtrate is washed with 5% citric acid (1x 50 ml), 5% sodium
bicarbonate (1 × 50 ml), and brine (1 x 50 ml), dried (sodium sulfate), and
concentrated. Flash chromatographic purification of the residue using ethyl
acetate/hexanes as eluants gives BBNA as 0.26 g (43%) of a colorless oil
that slowly solidifies on standing.
A solution of BBNA (0.26 g, 0.82 mmol), N-tert-BOC-T-aminobutyric
acid (6, 166 mg, 0.82 mmol) and DBU (0.13 ml, 0.87 mmol) in benzene (10
ml) is refluxed overnight. The DBU hydrobromide is filtered off, and the
filtrate concentrated to a pale brown oil and purified by flash chromatogra-
phy (5% ethyl acetate/chloroform) to give ot-tert-butylcarboxyl-2-nitroben-
zyl N-tert-BOC-T-aminobutyrate (7) as 0.36 g (100%) of a clear, pale brown
oil: 1H NMR (CDC13) 8.00 (d, J = 7.6 Hz, 1H), 7.64 (m, 2H), 7.53 (m, 1H),
6.74 (s, 1H), 4.7 (br s, 1H), 3.20 (q, J = 6.3 Hz, 2H), 2.52 (m, 2H), 1.90 (m,
J = 6.9 Hz, 2H), 1.43 (s, 9H), 1.40 (s, 9H).
Neat a-tert-butylcarboxyl-2-nitrobenzyl N-tert-BOC-T-aminobutyrate
(7, 0.30 g, 0.68 mmol) is cooled in an ice bath under argon and treated with
37L. J. Mathias, Synthesis p. 561 (1979).

trifluoroacetic acid (2.2 ml, 29 mmol). The resulting solution is incubated

overnight at room temperature, and the volatiles are removed to give a
pale brown oil. This oil is purified by chromatography on Sephadex LH-
20, using water as an eluant. The combined product fractions are lyophilized
to give the title compound 8 as its trifluoroacetate salt as 210 mg (77%) of
a compact white powder: mp 139-142 ° (dec); 1H NMR (D20) 8.13 (d, J =
8.1 Hz, 1H), 7.81 (t, J = 7.4 Hz, 1H), 7.70 (m, 2H), 6.69 (s, 1H), 3.07 (t,
J = 6.4 Hz, 2H), 2.63 (m, 2H), 2.01 (t, J = 7.0 Hz, 2H). Analysis calculated
for C14H15HzO8F3: C, 42.43; H, 3.82; N, 7.07. Found: C, 42.45; H, 3.97;
N, 6.92.
y-O-Desyl Glutamate
The synthesis of this compound (12) has been previously described. 32
Briefly, following Scheme 3, reaction of desyl bromide (10, 605 mg, 2.2
mmol), N-t-BOC-glutamic acid, a-tert-butyl ester (9, 672 mg, 2.2 mmol),
and DBU (0.35 ml, 2.3 mmol) in refluxing benzene (25 ml) for 1 hr gives
11 as 1.06 g (96%) of a colorless immobile oil after aqueous workup and
flash chromatographic purification (ethyl acetate/chloroform). Reaction of
11 (1.06 g, 2.13 mmol) with TFA (2.0 ml) in dichloromethane (5 ml) for 3
hr gives, after Sephadex LH-20 purification using water, 12 as 0.51 g (52%)
of a hygroscopic colorless powder; during lyophilization, the trifluoroacetic
acid in the product salt is volatilized: mp 129-132 ° (dec); 1H NMR (D20)
8.07 (d, J = 7.5 Hz, 2H), 7.72 (t, J = 7.4 Hz, 1H), 7.5 (m, 7H), 7.22 (s, 1H),
O.r
o o
+o oH+ CoH6 o
NHBOC
9 10 11
'CF3CO2H
CH2C12 HO i [
12
SCHEME 3
3.85 (q, J = 6.6 Hz, 1H), 2.8 (m, 2H), 2.2 (m 2H). Analysis calculated for
Ca9H19NO5 " 1/2 H20: C, 65.13; H, 5.75; N, 4.00. Found: C, 64.97; H, 5.69;
N, 3.88.
2-Methoxy-5-nitrophenyl-fl-alanine
The synthesis of compound 16 via Scheme 4 has been previously de-
scribed. 31 Briefly, a solution of N-t-BOC/3-alanine (13, 220 rag, 1.2 mmol),
EDC (230 mg, 1.2 mmol), and catalytic amounts of DMAP (10 mg) and
HOBt (10 mg) in dichloromethane (7 ml) is cooled to - 7 8 ° and treated
with 2-methoxy-5-nitrophenol (14, 170 mg, 1 mmol). The reaction mixture
is allowed to warm to room temperature overnight, then washed with water
(3x 5 ml), 5% NaHCO3 (1x 5 ml), water (2x 5 ml), 10% citric acid
(3x 5 ml), and brine ( l x 3 ml), dried (NazSO4), and purified via flash
chromatography using hexane/ether as an eluant to give the intermediate
carbamate 15 as a white powder in 80% yield: 1H NMR (CDC13) 8.16 (dd,
J = 9.2, 2.7, '1 H), 7.98 (d, J = 2.7, 1H), (d, J = 9.1, 1H), 5.12 (br s, 1H),
3.94 (s, 3H), 3.52 (dt, J = 6.0, 2H), 2.81 (t, J = 6.0, 2H), 1.64 (s, 9H). A
solution of 15 (65 rag, 0.19 mmol) is prepared in HCl-saturated ethyl acetate
at -78 ° and allowed to warm to room temperature. The precipitate is
filtered, washed with ethyl acetate, and dried in vacuo to give the title
compound (16) as 50 mg (95%) of a white powder: mp 168-175 ° (dec);
NO 2 NO 2
O
+ EDC )
HO~-/'~NHBOC
OH O NHBOC
OMe OMe
13 14 15
NO 2
HCI
ethyl acetate
)
+oA l
OMe
16
SCI-~ME 4
[2] PHOTOLABILEPROTECTINGGROUPS 39
N2
O "'"
CH3 + CH30 CHC13,
O//""OH \ CH30" "">/ "NO2
17 18
CH3NH N O O "'"
C H 3 0 ~ C H 3 /
CH30 '/ "NO2

19
SCHEME 5
1H NMR (D20) 8.08 (d, J = 8.8 1H), 7.9 (s, 1H), 7.1 (d, 1H), 4.4 (s, 3H),
3.2 (t, J = 5.8, 2H), 2.97 (t, J -- 5.8, 2H); HRMS 241.0827 [241.0824 calculated
for C10H,3N205 (MH+)].
1-(4,5-Dimethoxy-2-nitrophenyl)ethyl A23187
A23187 (17) is a calcium ionophore, i.e., a naturally occurring substance
that increases intracellular calcium ion concentrations by acting as a calcium
carrier across plasma cell membranes. It is often used to increase intracellu-
lar calcium concentrations in cell populations suspended in high-calcium
media. Attachment of a caging group to its carboxylate moiety abolishes
its calcium-channel activity.3s Ultraviolet photolysis then regenerates free
A23187 as judged by intracellular calcium concentration jumps after cells
are bathed in solutions of the caged precursor. 39
Following Scheme 5, to a pale yellow solution of 4,5-dimethoxy-2-nitro-
acetophenone hydrazone 17 (18 mg, 0.075 mmol) in chloroform (1.5 ml)
is added manganese dioxide (70 mg, 0.81 mmol). The resulting mixture is
stirred for 15 min in darkness and is then filtered through Celite with a
38K. R. Gee and P. J. Millard, unpublishedresults (1995).

39A. Ishihara,K. R. Gee, S. Schwartz,K. Jacobsen, and J. Lee, Biotechniques 23, 268 (1997).
chloroform rinse (2x 1 ml). To the combined red-orange diazoethane (18)
filtrate is added a solution of the ionophore A23187 [17, 26.9 mg, 0.0514
mmol, Sigma (St. Louis, MO)] in 1.5 ml chloroform. After the mixture is
stirred overnight, the color of the reaction solution is pale yellow and TLC
analysis shows no remaining A23187. The reaction solution is diluted with
chloroform (5 ml), and glacial acetic acid (5 drops) is added to quench any
remaining diazoalkane. Concentration in vacuo gives a pale yellow oil,
which is purified by flash chromatography on silica gel using 5% methanol/
chloroform as an eluant. The title compound (19) is obtained as 37 mg
(97%) of a pale yellow microcrystalline solid. On TLC the two product
diastereomers are resolved using chloroform/methanol/acetic acid
(100 : 5 : 1) as an eluant, giving two spots at Rr0.36 and 0.43; the fluorescence
of these spots is initially quenched, but the spots fluoresce blue on illumina-
tion with a hand-held UV lamp (unmodified A23187 has an RI of 0.21).
Two-dimensional TLC experiments show that after 10 min irradiation at 366
nm with a hand-held UV lamp, both diastereomers of the title compound are
photolyzed cleanly into native A23187. For the title compound: mp 108-
116 ° (dec); aH NMR (CDC13) ~9.6, 9.3 (two br s, 1H), 8.2, 8.0 (two s, 1H),
7.6 (m, 3H), 6.9 (m, 2H), 6.6 (t, 1H), 6.25 (dq, 1H), 4.2-2.8 (m, 13H),
1.8-0.7 (m, 26 H). HPLC analysis shows the presence of both product
diastereomers, in 93% purity.
Photochemistry of Caged Carboxylates

The mechanism of the photolytic release of carboxylates from o-nitro-
benzyl esters has been studied in detail, 4°,41 as has the photolytic release
of carboxylates from desyl estersJ 6 We describe general methodology used
to determine photochemical properties of caged carboxylates in vitro.
Instrumentation
The apparatus used for investigations of the photolysis reaction of caged

derivatives of carboxylates is illustrated in Fig. 1.
Kinetic Measurements
Figure 2A illustrates the fast conversion to products observed in the

photolysis of MNP-/3-alanine {16). 31 The absorbance of the compound re-
leased has the wavelength and pH dependence characteristic of 2-methoxy-
5-nitrophenol. The photolysis reaction apparently consists of a single expo-
40H. Schupp, W. K. Wong, and W. Shnabel, J. Photochem. ~ 85 (1987).

41Q. Q. Zhu, W. Schnabel and H. Schupp, J. Photochem. 39~317 (1987).
Flash
i[ Laser ~ ~
[ _ _| ~Optical
e r
Monoch romator
Storage ~ ~ H i g~h voltage

loscopeCurrent
osci voltotage
converter
FIG. 1. Laser-flash photolysis apparatus for studying photolysis reactions of caged deriva-
tives of neurotransmitters. Light from a XeCI excimer laser (Lumonics TE-816M, Kanata,
Ottawa, Ontario, Canada) is focused to irradiate 40 /zl solution in a 2 × 10-mm quartz
cuvette. 56 The laser provides 100-nsec pulses of 308 nm light with a pulse energy of 10-50
mJ; flux densities of 108 mJ per mm 2 are produced at the front face of the cuvette. The pulse
energy is measured with a Gentec ED200 Joulemeter (Dalton, Canada). A collimated monitor
beam is produced by an Oriel (Stratford, CT) flash lamp with a 100-/~sec pulse and passed
through a Corning WGS360 cutoff filter and then through the quartz photolysis cuvette
(Spectrocell Inc., 160/zl, Oreland, PA) oriented with faces perpendicular to the irradiation
and monitoring beams, respectively. The monitor beam is at right angles to the laser beam;
the path length for the laser and the monitoring beam is 2 and 10 mm, respectively. The monitor
beam projects onto the detection apparatus, a photomultiplier-monochromator (single pass)
combination (Thorn EM/9635QB, McPherson 275, Fairfield, NJ), for observing the transient
absorbance of intermediates produced during photolysis. The detection apparatus is arranged
at 90° to the laser. To avoid noise from the discharge of the laser power supply during
measurements in the nanosecond time region, the signal emerging from the cuvette is carried
by a 250-ktm-diameter optical fiber (Fiberguide Industries, Stirling, N J) from one room housing
the laser to the detection system in a second room. This optimizes the detection of the
photolysis intermediates and products. A current-to-voltage converter (Thorn EMI Gencom
A1, Fairfield, NJ) is used to treat the voltage output, which is filtered at four times the data
acquisition rate (Krohn-Hite Model 3396, Avon, MA); signal transients are observed on a
storage oscilloscope (Tektronix 549, Beaverton, OR), where the results are displayed and
recorded, or are digitized directly into an AST 286 personal computer. Signal capture is
optimized by using a variable time-delay trigger (World Precision Instruments A310, Sarasota,
FL). The white-light monitoring beam itself produces negligible photolysis of any of the caged
compounds we have tested so far. It also does not heat the sample, as determined by monitoring
the temperature at which the starting material gives a maximum absorbance with an inserted
thermocouple with a 0.1 ° sensitivity.56
nential process with a Ill 2 value of --1/zsec. Proof that the photolysis reaction
releases/3-alanine is ascertained by use of HPLC, 3° cell-flow, 42 and laser-
pulse photolysis 4'a1'2° experiments with hippocampal neurons containing
glycine receptors. 3a
42 j. B. Udgaonkar and G. P. Hess, Proc. Natl. Acad. ScL U.S.A. 84, 8758 (1987).
A o . ~" ^
0.1
Q)
0
c 0.1
0
0
{n 0.0
.£/
<
-0.0
-20 0 20 40 60 80 100
T i m e (/~sec)
B 0.5 . . . . i . . . . f . . . . i . . . . ~ . . . .
0.4
~ 0.3
IJ
~ .2-
,~ 0.1
0 ~
-0.1 , , , , I , ' ' ' i . . . . i . . . . , . . . .
-50 0 50 1 O0 150 200
Time (#,see)
FIG. 2. Rapid absorbance changes produced by photolysis of caged compounds at room
temperature by a single pulse of 308-rim light from an XeCI excimer laser (approximately
30-50 mJ). (A) A 2 mM solution of caged MNP-/3-alanine (16) 31 in 100 mM phosphate buffer,
pH 7.1, is photolyzed. The absorbance change is monitored at 405 nm with a single-beam
transient spectrophotometer, the signal transfer parameters of which effectively produced 500
kHz filtering of the rapid absorbance jump. (B) A 5 mM solution of CNB-GABA (8) in 100
mM phosphate buffer at pH 7.4 is photolyzed. 21 The absorbance change is monitored at 430
nm. (C) Determination of the product quantum yield for photolysis of caged kainate in 100
mM phosphate at pH 6.8 and 22 °. A sample cuvette containing 0.5 mM CNB-caged kainate 23
is irradiated, with mixing of the solution between laser pulses of 308 nm with energy output
of 50 mJ, and the absorbance, A, is monitored at 430 nm as a function of the number of laser
shots, n. The ratio of absorbed photons to targeted molecules (Ke) is 0.87, the path length
(/) was 10 mm, and the fraction of the volume irradiated (F) was 0.5. The product quantum
yield and molar absorption coefficient are determined, by using Eq. (1), to be 0.34 _+ 0.07
and 920 M -1 cm -1, respectively. The measurements are made in duplicate, and the standard
deviation from the mean is shown? 3
C .1.6
t . , - , • i •
-1.8
-2.0
-2.2
-2.4
-2.6
-2.8
-3.0
2", 0 2 4 6
(n-l) Laser Pulses
FIG. 2. (continued)
When CNB is used to protect the carboxyl function of neurotransmitters,

the photolysis reaction is followed by observing transient absorption spectra
in the 350- to 500-nm wavelength region characteristic of the aci-nitro
intermediate in the photolysis reaction. 4°,41,43 Experimental evidence
indicates that the decay of this intermediate reflects product for-
mation.33,40,41,43,44
Figure 2B illustrates the formation and decay of the aci-nitro intermedi-
ate in the photolysis of CNB-7-aminobutyric acid (8).21 The decay phase
of the absorbance (Fig. 2B) is consistent with a single exponential process
with a tl/z value of 19 tzsec. Proof that photolysis results in the release of
the inhibitory neurotransmitter of GABA in the photolysis of CNB-GABA
is shown by the use of HPLC, 21 cell-flow,45 and laser-pulse photolysis 4
experiments with cortical neurons containing GABA receptors31
Quantum Yield Determination

The quantum yield of MNP-/3-alanine (16) 31 is determined with 20-/xl
aliquots of a solution of 2 mM MNP-/3-alanine in 100 mM phosphate buffer,
pH 7.1. A quartz cuvette with a 0.2-mm path length is used. The solution
is photolyzed by a single 30-mJ light pulse at 308 nm from an excimer laser,
and the concentration of the released 2-methoxy-5-nitrophenol, which is
assumed to be produced stoichiometrically with/3-alanine, is determined
spectrophotometrically. A Gentec ED200 Joulemeter (Dalton, Canada) is
used to determine the energy absorbed by the sample and by the buffer
43 j. A. McCray and D. R. Trentham, Proc. Natl. Acad. Sci. U.S.A. 77, 7237 (1980).
44j. F. Wootton and D. R. Trentham, N A T O Adv. Study Inst. Ser. C 272, 277 (1989).
45 G. P. Hess, J. B. Udgaonkar, and W. L. Olbricht, Annu. Rev. Biophys. Biophys. Chem. 16,
507 (1987).
44 CAGED COMPOUNDS [2]
B 2.5
~- 2.o
0
uJ
0
0
I
0.0
LL_____
-250 lliSO 3550 5450 7350
TIME (msec)
FIG. 3. (A) Schematic drawing (not to scale) of the device used for rapid equilibration of
receptors on a cell with neurotransmitters, caged neurotransmitters, and inhibitors.42'45 The
device is made with stainless-steel tubing and is U shaped. Solutions can flow in from one
arm of the U and out from a central opening ( - 1 5 0 / z m diameter) at the base of the U, as
indicated by the arrows. A BC3H1 muscle cell of - 1 5 / x m diameter and containing nicotinic
aeetylcholine receptors is attached to a whole-cell current-recording electrode.52When a valve
is opened, the solution containing the biological effector emerges from the central hole of
the U tube at a rate of - 1 cm/sec. At the end of the current measurement, which is used to
characterize the opening and closing of receptor channels, a valve turns off the flow of solution
through the lower arm of the U tube. Simultaneously, another valve opens the connection
between the upper arm of the U tube and a suction device. The solution bathing the cell now
flows into the U tube at a rate of ~4 cm/sec. This removes the solution bathing the cell and
prevents leakage of solutions containing the biological effector from the lower arm of the U
tube into the cell compartment. The solution removed by the U tube is automatically replaced
by a new solution to bathe the cell. About 3 min of this procedure is sufficient to remove
the biological effector from the solution bathing the cell. Typically - 1 5 measurements can
be made with one BC3H1 cell. Leaking of solution from the central tube or one arm of the
U tube is prevented by suction applied to the other arm of the U tube. The resulting linear
flow rate of the solution is 1-2 cm/sec. The cell suspended from the current-recording electrode
is approximately 100/zm away from the hole.24 (B) An example of an acetylcholine-induced
ionic curent in a BC3H1 cell. The cell was placed under whole-cefl voltage clamp at a membrane
potential of - 6 0 mV. A solution of 200/xM acetylcholine flowed around the cell. The thin
line parallel to the abscissa was calculated from Eq. (2) and represents the current that would
be obtained in the absence of receptor desensitizationY (C) Effect of earbamoylcholine
concentration on IA. Carbamoylcholine-indueed whole-cell currents through the acetyleholine
C
5
-<2
0 """
10-2 10-1 100 101
[CARBAMOYLCHOLINE] (raM)
receptors of BC3H1 cells were measured at 23° and - 6 0 mV. The results represent measure-
ments on 10 ceils, normalized to each other by standardization of the carbamoylcholine
concentration. Triangles represent the normalized 1A values. Circles represent IA values calcu-
lated from Po values at three carbamoylcholine concentrations. The solid line was drawn using
Eq. (2) and values for K1, ~b, and IMRM of 235/zM, 0.18, and 5.1 nA, respectively, which were
obtained from the linear fit of the data using Eq. (2).42'57
alone. The quantum product yield estimated to be 0.2 in this experiment

is in good agreement with that found for MNP-glycine. 3°
Determination of the product quantum yield can be accomplished with-
out using analytical techniques to isolate the product. The assumption is
that the measured optical signal in the photolysis reaction is proportional
to the product released. The absorbance of the product An measured at
the nth laser pulse with mixing of the solution after each pulse is given by
Eq. (1): 20
A n = Em LcoCbKEexp[-qbKEF(n - 1)] (1)
where Em is the molar extinction coefficient of the absorbance change, L
is the length of the light path, and Co is the initial concentration of the
compound that is photolyzed, th is the quantum yield, KE is the ratio of
the number of absorbed photons to the number of target molecules in the
laser beam, and F is the fraction of the solution containing the compound
photolyzed through which the laser beam passes. It is important to ascertain
that the absorbance of the solution at the wavelength used for photolysis
does not change during these experiments. The quantum yield and the
molar extinction coefficient of the product can be obtained from a semiloga-
rithmic plot of An versus (n - 1) (Fig. 3C). Such a plot is shown for the
photolysis of CNB-caged kainate 23 in which the absorbance measured is
due to the aci-nitro intermediate observed in the photolysis reaction. A
product quantum yield of 0.34 ___ 0.07 is obtained from the slope of the
line. It is assumed that the concentration of transient intermediate is directly
proportional to the concentration of the compound liberated. 33'4°'41'43'44
Biological C h a r a c t e r i z a t i o n of Caged Carboxylates

Crucial in the development of caged biological effectors is the demon-
stration that the caged compound before photolysis and its photolytic side
products are biologically inert in the system under study. To determine
whether caged biological effectors are biologically inert, the following ques-
tion must be asked: Does the caged biological effector initiate, inhibit, or
potentiate the reaction to be investigated? Neurotransmitter receptors
are reversibly inactivated by their specific neurotransmitter, not in the
second time region originally assumed, 46 but in the millisecond time re-
gion. 4,26,45'47-5° Techniques were, therefore, developed to investigate recep-
tor-mediated reactions in the millisecond time region. 42,45,47'51The technique
we use to characterize caged neurotransmitters is described next.
Cell-Flow T e c h n i q u e
Instrumentation. The concentration of open receptor channels can be

determined by measuring the current passing through the channels at con-
stant transmembrane voltage using the whole-cell current recording tech-
nique. 52 A simple device was developed by Krishtal and Pidoplichko 51 to
flow neurotransmitter solutions over a single cell suspended in the center
of the stream by the recording electrode. A schematic drawing of the cell-
flow device is given in Fig. 3A.
The flow rate of the solutions emerging from the flow device is restricted
by the stability of the seal between the recording electrode and the cell
membrane. At the permissible flow rates, receptors on the cell surface
facing the flow device are in contact with the neurotransmitter before the
solution reaches the opposite side of the cell and, therefore, desensitize first.
46B. Katz and S. Thesleff, J. Physiol. (Lond.) 138, 63 (1957).

47G. P. Hess, D. J. Cash, and H. Aoshima, Nature 282, 329 (1979).
48B. Sakmann, J. Patlak, and E. Neher, Nature 286, 71 (1980).
49L. O. Trussel, G. D. Fischbach, L. I. Thio, and C. Z. Zoramski, Proc. Natl. Acad. Sci. U.S.A.
85, 2834 (1988).
50E. L. M. Ochoa, A. Chattopadhyay,and M. G. McNamee,Cell. Mol. Neurobiol. 9,141 (1989).
51O. A. Krishtal and V. I. Pidoplichko, Neuroscience 5, 2325 (1980).
52O. P. Hamill, E. Marty, E. Neher, B. Sakmann, and F. J. Sigworth, Pflugers Arch. $91,
85 (1981).
The available hydrodynamic theory53'54 allows one to correct the observed

current for receptor desensitization (inactivation) that occurs while some re-
ceptors on the cell surface farther away from the flow device are still equili-
brating with the neurotransmitter.42'45
A typical cell-flow experiment4z'45is shown in Fig. 3B. In this experiment
a 200/zM solution of acetylcholine flows over a BC3H1 muscle cell containing
nicotinic acetylcholine receptors. The current due to the opening of acetyl-
choline receptor channels is recorded with a whole-cell current-recording
electrode at constant voltage. The current (thick curve) first rises due to the
opening of receptor channels, reaches an amplitude of about 1.1 nA, and then
falls due to receptor desensitization.
Theory. The thin line in Fig. 3B rising to a value of 2.2 nA is a calculated
line and represents the current that would be observed if the receptors did
not desensitize. Equation (2) was derived to calculate this current Ig .42 We
define Ia as the amplitude of the current arising from the receptors on the
cell surface in the absence of receptor desensitization, at a definite ligand
concentration. To take into account the uneven rate at which the ligand solu-
tion flows over the cell s u r f a c e , 42'45 w e divide the current time course into
constant time intervals (we have used 5 msec) and then correct the current
for the desensitization that occurs during each time interval At. After the
current (Iobs)ati is measured for each of n constant time intervals (nAt = tn),
the correct current is given by Eq. (2):
(IA)t,, ---- ( eaat 1) ~ (Iobs)hti -]- (Iobs)5t

- - n (2)
i=1
where (Iobs)A/i is the observed current during the ith time interval. (IA)t,
becomes equal to IA when the value of tn is equal to or greater than the
current rise time, tr. The current and the current maximum corrected for
desensitization are given by the thin line in Fig. 3B. With rise times of
50-220 msec, in measurements made with the same cell and the same ligand
concentration, IA was found to vary by no more than 10%.
Comparison of Cell-Flow Method with Single-Channel Recording Tech-
nique and Laser-Pulse Photolysis Technique. The validity of the experimen-
tal approach and Eq. (2) (Fig. 3C) used to calculate the concentration of
open receptor channels has been determined by two independent tech-
53 V. G. Landau and E. M. Lifshitz, "Fluid Mechanics," Pergamon, Oxford, UK, 1958.

54 A. G. Levich, "Physicochemical Hydrodynamics," Prentice-Hall, Englewood Cliffs, N J,
1962.
niques, the single-channel current recording techique, 55 which allows one

to determine the fraction of receptor channels that are open 48 at low neuro-
transmitter concentration, and the laser-pulse photolysis technique using
caged neurotransmitters, which has a microsecond time resolution and in
which the observed current does not have to be corrected for receptor
desensitization.4'6,l 1
Determining Suitability of Caged Neurotransmitters for Biological

Experiments Using Cell-Flow Technique
The importance of determining the effects of caged compounds on the
activity of the target biological molecule is illustrated with two examples.
Both CNB-carbamoylcholine (20) and o~-methyl-o-nitrobenzyl (MNB) car-
bamoylcholine (21) are water soluble and stable in aqueous solutions at
neutral pH. On irradiation with light in the 300- to 350-nm region, both
compounds release carbamoylcholine with rates and quantum yield that
make them suitable for rapid chemical kinetic investigations of nicotinic
acetylcholine receptor-mediated reactions. However, the CNB-caged car-
bamoylcholine is biologically inert and can be used in the investigations of
receptor function, whereas the MNB-caged derivative modifies receptor
function and cannot be used.
N" O N(CH3) 21R--CH

R O +
Another example of the importance of the biological characterization

of caged compounds is illustrated by the cell-flow experiments in Fig. 4A
and 4B. The experiment in Fig. 4A shows that in the presence of both 300
/~M NMDA and 300/zM CNB-NMDA, the observed current amplitude is
lower than when CNB-NMDA is not present. This means that the N M D A
receptor is inhibited by the caged form CNB-NMDA. It is of interest to
note, therefore, that CNB-carbamoylcholine (20), 2° CNB-y-aminobutyric
acid (8), zl and CNB-glutamate 22 are biologically inert in similar experi-
ments. Figure 4B shows that 10 mM CNB-GABA does not affect the
response of a mouse cerebral cortical cell to 250/zM GABA. It is, of course,
55 E. Neher and B. Sakmann, Nature 260, 779 (1976).

56 A. P. Billington, N. Matsubara, W. W. Webb, and G. P. Hess, Adv. Protein Chem. llI,
417 (1992).
57 G. P. Hess, Arch. Phys. Biochem. 104~ 752 (1996).
1200
800,
o,
o~
400
Time (msee) Time (msec)
~ 1500
E
~ 1000
0
o,
500
c-
~... 2 4
'' ".- i,,
6 8 10
Time (sec)
FIG. 4. (A) Whole-cell current induced by NMDA and recorded from individual neonatal
rat hippocampal neurons at - 6 0 mV, pH 7.4, and 22 °. The transmembrane ion current was
generated by (a) flowing 300 p~M NMDA over the cell or (b) simultaneously flowing 300/zM
NMDA and 300/xM CNB-NMDA in which the/3-carboxyl group is caged.27 (B) Whole-cell
current responses of embryonic mouse cerebral cortical cells, pH 7.3, 21-23 °, at a transmem-
brane voltage Vm of - 70 mV. The two curves represent experiments in which the cell was
equilibrated with 250/zM GABA in the absence and presence of 10 mM CNB-GABA (8). 21
also important to ascertain that the side product that is photolytically

released together with the desired biological effector is biologically inert.
In the development of any new caged compounds it is important that
the reagents meet certain stringent criteria before they are used in the
studies for which they have been designed. We have described the synthesis,
photochemistry, and biological characterization of photolabile precursors

(caged compounds) of carboxylic acids and neurotransmitters. Some of the
compounds described can provide new levels of temporal resolution in
transient kinetic measurements of function and/or spatial resolution in
mapping studies.
Acknowledgments
Work reviewed here was supported by a grant (GM04842) awarded to G.P.H. by the
National Institutes of Health Institute of General Medical Sciences.
[31 N e w C a g e d G r o u p s : 7 - S u b s t i t u t e d Coumarinylmethyl
Phosphate Esters
B y TOSHIAKI FURUTA a n d MICHIKO IWAMURA
Introduction
The design and synthesis of caged probe molecules that function inside
of living cells are challenging problems for organic chemists. As part of a
research program directed toward these objectives we have reported the
synthesis and utilization of new types of photochemically removable pro-
tecting groups that additionally exhibit a strong fluorescence emission. 1-3 Of
the four fluorescent, photosensitive protection groups prepared for diethyl
phosphate, namely, 1-pyrenylmethyl,4 4-(7-methoxycoumarinyl)methyl
(MCM),5 2_(9,10_dioxo)anthrarylmethyl,6 and 2-naphthylmethyl, 1-pyrenyl-
methyl was found to be the best caged group in terms of its photochemical
properties. However, its extremely low stability in physiological salt solu-
tions limited its application in studies on living cells, and so we focused
our attention on the 4-(7-methoxycoumarinyl)methylgroup as a photolabile
"cage" group for phosphates. 7 Whether naturally occurring or synthetic,
coumarin and many of its related aromatic lactones show interesting photo-
1 M. Iwamura, T. Ishikawa, Y. Koyama, K. Sakuma, and H. Iwamura, Tetrahedron Lett. 28,

679 (1987).
2 M. Iwamura, K. Tokuda, N. Koga, and H. Iwamura, Chem. Lett. 1729 (1987).
3 M. Iwamura, C. Hodota, and M. Ishibashi, Synth. Lett. 35 (1991).
4 T. Furuta, H. Torigai, T. Osawa, and M. Iwamura, Chem. Lett. 1179 (1993).
5 R. S. Givens and B. Matuszewski, J. Am. Chem. Soc. 106, 6860 (1984).
6 D. S. Kemp and J. Reczek, Tetrahedron Lett. 18, 1031 (1977).
7 T. Furuta, H. Torigai, M. Sugimoto, and M. Iwamura, J. Org. Chem. 60, 3953 (1995).
Copyright© 1998by AcademicPress

All rightsof reproductionin any formreserved.
METHODS IN ENZYMOLOGY,VOL 291 0076-6879198 $25.00
photochemistry, and biological characterization of photolabile precursors

(caged compounds) of carboxylic acids and neurotransmitters. Some of the
compounds described can provide new levels of temporal resolution in
transient kinetic measurements of function and/or spatial resolution in
mapping studies.
Acknowledgments
Work reviewed here was supported by a grant (GM04842) awarded to G.P.H. by the
National Institutes of Health Institute of General Medical Sciences.
[31 N e w C a g e d G r o u p s : 7 - S u b s t i t u t e d Coumarinylmethyl
Phosphate Esters
B y TOSHIAKI FURUTA a n d MICHIKO IWAMURA
Introduction
The design and synthesis of caged probe molecules that function inside
of living cells are challenging problems for organic chemists. As part of a
research program directed toward these objectives we have reported the
synthesis and utilization of new types of photochemically removable pro-
tecting groups that additionally exhibit a strong fluorescence emission. 1-3 Of
the four fluorescent, photosensitive protection groups prepared for diethyl
phosphate, namely, 1-pyrenylmethyl,4 4-(7-methoxycoumarinyl)methyl
(MCM),5 2_(9,10_dioxo)anthrarylmethyl,6 and 2-naphthylmethyl, 1-pyrenyl-
methyl was found to be the best caged group in terms of its photochemical
properties. However, its extremely low stability in physiological salt solu-
tions limited its application in studies on living cells, and so we focused
our attention on the 4-(7-methoxycoumarinyl)methylgroup as a photolabile
"cage" group for phosphates. 7 Whether naturally occurring or synthetic,
coumarin and many of its related aromatic lactones show interesting photo-
1 M. Iwamura, T. Ishikawa, Y. Koyama, K. Sakuma, and H. Iwamura, Tetrahedron Lett. 28,

679 (1987).
2 M. Iwamura, K. Tokuda, N. Koga, and H. Iwamura, Chem. Lett. 1729 (1987).
3 M. Iwamura, C. Hodota, and M. Ishibashi, Synth. Lett. 35 (1991).
4 T. Furuta, H. Torigai, T. Osawa, and M. Iwamura, Chem. Lett. 1179 (1993).
5 R. S. Givens and B. Matuszewski, J. Am. Chem. Soc. 106, 6860 (1984).
6 D. S. Kemp and J. Reczek, Tetrahedron Lett. 18, 1031 (1977).
7 T. Furuta, H. Torigai, M. Sugimoto, and M. Iwamura, J. Org. Chem. 60, 3953 (1995).

METHODS IN ENZYMOLOGY,VOL 291 0076-6879198 $25.00
13] NEW CAGED GROUPS 51
RO
0~/ NH2
o
I N J~l,,,..~
O O~ P ~ ' ~ O ~ N
OH
MCM-cAMP : R=CH3 ACM-cAMP: R=CH3CO

HCM-cAMP : R=H PCM-cAMP: R=CH3CH2CO
F~c. 1.7-Substituted 4-coumarinylmethyl-caged cAMPs.
biological activities, and a number of these compounds are commercially

available and used as fluorescent dyes. Although the 4-coumarinylmethyl
group has been used as a fluorescent tag for biological molecules,5 this
highly fluorescent group had not been used as a caging group until our
report on a new method for the synthesis of esters of diethyl phosphate
and cAMP. 8 The silver(I) oxide promoted condensation method, as we
call it, is a one-pot synthesis of phosphate esters from the corresponding
phosphate and an alkyl halide in the presence of silver oxide. It is a mild
neutral reaction that generates the product in a modest yield along with
the silver halide, an easily separable product. Although the product yield
is not necessarily high, the main attraction of this reaction is that it is a
single-step synthesis that uses commercially available phosphate groups
and alkyl halides.
The coumarin-caged cAMP compounds, MCM-cAMP, 4-(7-acetoxy-
coumarinyl)methyladenosine (ACM)-cAMP, 4-(7-propionyloxycouma-
rinyl)methyladenosine (PCM)-cAMP and 4-(7-hydroxycoumarinyl)meth-
yladenosine (HCM)-cAMP cyclic 3',5'-monophosphates (Fig. 1), which
differ from each other only by the substituent at the 7-position of the
coumarin ring, have been prepared from reactions of the corresponding 4-
coumarinylmethyl bromide with cAMP using the silver(I) oxide promoted
method. These caged cAMP compounds exhibit similar photocleavage
properties and efficiently release cAMP in Ringer's solution on irradiation
with near-ultraviolet light. Irradiation of a 0.1 mM solution of any one of the
caged cAMPs in Ringer's solution containing 1% (v/v) dimethyl sulfoxide
(DMSO) with near-ultraviolet light for 20 min results in the loss of 60-80%
of coumarin-caged cAMP and liberation of 40-60% of free cAMP. Another
8 T. Furuta, H. Torigai, T. Osawa, and M. Iwamura, J. Chem. Soc. Perkin Trans. 1, 3139 (1993).
significant feature of these caged cAMPs is their long half-life in the dark
in physiological salt solutions compared to those of o-nitrobenzyl- or o-
nitrophenethyl-caged cAMPs, which in our experience readily hydrolyze
in aqueous solution.
Higher lipophilicity is an important property of caged nucleotides if
they are to be loaded into living cells through the plasma membrane.
However, this lipophilicity might also result in significant leakage of the
caged cAMP from the cell. To overcome this potential problem, acyloxy-
coumarin cAMP derivatives were designed and synthesized with the feature
that once they enter the cell they are hydrolyzed by intracellular esterases
to the more hydrophilic 7-hydyroxycoumarin derivative, HCM-cAMP. The
de-esterified caged cAMP compounds are better retained within the cyto-
plasm, and cells loaded with these caged compounds were found to exhibit
a higher cAMP-dependent melanin granule dispersion response compared
to other more lipophilic-caged cAMP compounds after irradiation with
near-ultraviolet light. 9
This article summarizes the methods used in our laboratory for the
preparation and characterization of 4-coumarinylmethyl-caged phosphate
esters and their application in studies of the role of cAMP on the dispersion
and aggregation of pigment granules in live cells.
Methods
Materials
(-)-Adenosine cyclic 3',5'-monophosphate (Acros, Sigma, St. Louis,

MO), 4-(bromomethyl)-7-methoxycoumarin (Nakarai, Kyoto, Japan), 4-
(bromomethyl)-7-acetoxycoumarin (TCI, Tokyo, Japan), 4-methylum-
belliferone (TCI), p-toluenesulfonyl hydrazide (Wako, Osaka, Japan), tri-
ethylamine (Wako), selenium dioxide (95%, Nakarai), and silver oxide
(Wako) are used without further purification. Acetonitrile (dehydrated,
Kanto, Tokyo, Japan), dimethyl sulfoxide (Wako), ethanol (99.5%, Wako),
and m-xylene (Wako) of the highest quality available are used without
further purification. Methanol is distilled from Mg(OCH3)2 under nitrogen.
Chromatography is performed using Kieselgel 60 (E. Merck, Darmstadt,
Germany) of 0.063- to 0.200-mm-grade for column chromatography and of
0.042- to 0.063-mm-grade silica gel for flash column.
Nuclear magnetic resonance (NMR) is performed using a JNM-GSX
270 spectrometer (Jeol, Tokyo, Japan) with tetramethylsilane as an internal
9T. Furuta,A. Momotake,M. Sugimoto,M. Hatayama,H. Torigai,and M.Iwamura,Biochem.

Biophys. Res. Commun. 228, 193 (1996).
[31 NEWCA6EO~ROUeS 53
standard. Infrared spectroscopy (IR) is performed using an FF/IR 5000
spectrometer (Jasco, Tokyo, Japan). Ultraviolet (UV) spectra are deter-
mined with a U-3210 spectrometer (Hitachi, Tokyo, Japan). High-perfor-
mance liquid chromatography (HPLC) is performed with a PU980 intelli-
gent HPLC pump equipped with a UV970 intelligent UV/vis detector
(Jasco, Tokyo, Japan) using a reversed-phase column, Jasco Crestpac C185S
(150 × 4.6 mm) and monitored at 254 nm. The mobile phase is 50%
methanol-50% water (v/v). High-resolution mass spectra are obtained using
a GC-MATE spectrometer (Jeol, Tokyo, Japan) in FAB mode.
Synthesis of Coumarin-Caged cAMPs

Synthesis of 4-(7-Methoxycoumarinyl)methyladenosine
Cyclic Y,5'-Monophosphate
Cyclic AMP (33 mg, 0.10 mmol) and 4-bromomethyl-7-methoxycouma-
rin (80.7 mg, 0.300 mmol are added to an oven-dried, 30-ml, two-necked,
round-bottom flask equipped with a magnetic stirring bar and a rubber
septum. The flask is evacuated under vacuum and flushed with argon. Dry
acetonitrile (6 ml) and DMSO (1 ml) are added to the reaction flask using
a syringe. After the addition of solid silver(I) oxide (44 mg, 0.20 mmol),
the resulting black suspension is stirred at 60° for 45 hr under argon. The
reaction mixture is filtered through filter paper and the residue washed
with chloroform. The combined filtrate is evaporated under reduced pres-
sure. The oily residue is purified by column chromatography (30 g of SiO2,
4.7% (v/v) methanol-CHzCl2, then 6.2% methanol-CHzCl2) to give 75 mg
of a yellow oil. Further purification by flash chromatography (30 g of SiO2,
4.7% methanol-CHzCl2) gives MCM-cAMP (23 mg, 44%) as a mixture of
two stereoisomers [axial (ax)/equatorial (eq) = 2.3/1]; 12.5 mg (29%) of
product is obtained from the residual black solid silver salt of cAMP.
Separation of the axial and equatorial isomer is achieved by semipreparative
reversed-phase HPLC (Cica-Merck, Tokyo, Japan, Lichrosorb RP-18 (7
m) 250 × 250 mm, mobile phase: 60% methanol-40% H20, flow rate: 5
ml/min, retention time: axial 33 min, equatorial 36 rain). Melting point
260-263 ° (decomposition); 1H NMR (5% CD3OD/CDCL3, ~ ppm) (axial
isomer) 3.90 (3H, s, OCH3), 4.47 (1H, d, J = 7.2 Hz, H'5), 4.70 (1H, dd,
J = 4.6 and 22.1 Hz, H'4), 4.85 (1H, d, J = 5 Hz, H'2), 5.40 (2H, d, J =
6.5 Hz, P-O-CH2), 5.52 (1H, m, H'3), 5.97 (1H, s, H'I), 6.54 (1H, t, J -- 1
Hz, H3), 6.89-6.91 (2H, m, H5 and H8), 7.74 (1H, dd, J = 2 and 6 Hz,
H6), 7.92 (1H, s, adenine H2), 8.25 (1H, s, adenine H8); (equatorial isomer)
3.92 (3H, s, OCH3), 4.48 (1H, d, J = 6 Hz, H'5), 4.68 (1H, m, H'4), 4.83
(1 H, d, J = 5 Hz, H'2), 5.41 (2H m, P-O-CH2), 5.51 (1H, m, H'3), 6.04
(1H, s, H'I), 6.49 (1H, t, J = 1 Hz, H3), 6.89-6.95 (2H, m, H5 and H8),
7.47 (1H, dd, J = 2 and 6 Hz, H6), 8.03 (1H, s, adenine H2), 8.28 (1H, s,
adenine H8); IR (KBr, Vmaxcm -1) (mixture of two stereoisomers) 1719
( C = O ) , 1296 ( P = O ) , 1212 (C-OCH3), 1096 (C-OH), 1017 (P-O); UV
[50% H20-dioxane,/~max (/3)] 259 (14,600), 325 (13,300).
Synthesis of 4-(7-AcetoxycoumarinyOmethyladenosine
Cyclic 3',5'-Monophosphate
Silver(l) Oxide Promoted Method. This reaction is performed as de-
scribed above except it uses 100 mg (0.304 mmol) of cAMP, 267 mg (0.899
mmol) of 7-acetoxy-4-bromomethylcoumarin, 132 mg (0.570 mmol) of sil-
ver(I) oxide, 18 ml of acetonitrile, and 3 ml of DMSO and is stirred for 27
hr at 100°. The products are purified by column chromatography (20 g of
SiO2, 6.2% methanol-CH2Cl2) and then flash column chromatography (20
g of SIO2, 6.2% methanol-CH2Cl2) to give 22 mg (0.041 mmol, 13.5% yield)
of ACM-cAMP as a mixture of two stereoisomers (ax/eq = 3/1).
Diazo-Based Method. Cyclic AMP, 64.3 mg (0.195 mmol), is added to
a 20-ml, round-bottomed flask equipped with a magnetic stirring bar and
CaC12 drying tube. Dry DMSO (2 ml) and 4-diazomethyl-7-acetoxycou-
marin (52.6 mg, 0.215 mmol) are added to the flask, and the mixture is
stirred at room temperature for 4 days. Evaporation and purification by
flash column chromatography (10 g of SIO2, 6.2% methanol-CH2C12) give
17.4 mg (0.0320 mmol, 16.5% yield) of ACM-cAMP as a mixture of two
stereoisomers (ax/eq = 1/1). 1H NMR (5% CD3OD/CDC13, 8 ppm) (axial
isomer) 2.35 [3H, s, C(=O)CH3], 4.48 (1H, d, J = 7 Hz, H'5), 4.68 (1H,
m, H'4), 4.82 (1H, d, J = 7 Hz, H'2), 5.40 (2H, d, J = 6.5 Hz, P-O-CH2),
5.49 (1H, m, H'3), 5.96 (1H, s, H'I), 6.67 (1H, t, J = 1.5 Hz, H3), 7.07-7.22
(2H, m, H5 and H8), 7.56 (1H, dd, J = 2 and 9 Hz, H6), 7.89 (1H, s, adenine
H2), 8.28 (1H, s, adenine H8); (equatorial isomer) 2.36 [3H, s, C ( = O)CH3],
4.48 (1H, d, J = 7 Hz, H'5), 4.68 (1H, m, H'4), 4.84 (1H, d, J = 7 Hz, H'2),
5.40 (2H, d, J = 6.5 Hz, P-O-CH2), 5.49 (1H, m, H'3), 6.01 (1H, s, H'I),
6.64 (1H, t, J = 1.5 Hz, H3), 7.07-7.22 (2H, m, H5 and H8), 7.56 (1H, dd,
J = 2 and 9 Hz, H6), 7.93 (1H, s, adenine H2), 8.29 (1H, s, adenine H8);
IR (KBr, /"max cm -1) (mixture of two stereoisomers) 1721 (C----O), 1267
(P----O), 1203 (C-O), 1135 (C-O), 1067 (C-O), 1015 (P-O); UV [CH3OH,
)tm~x (e)] 267 (13,400), 313 (7650).
Synthesis of Bromomethyl-7-hydroxycoumarin
4-Bromomethyl-7-acetoxycoumarin, 512.7 mg (1.726 mmol), and 35 ml
of wet toluene are added to a 100-ml, round-bottomed flask equipped
with a magnetic stirring bar, reflux condenser, and CaC12 drying tube. The
mixture is stirred at 80° for 20 min, 2.8 g of p-toluenesulfonic acid on SiO21°
lo G. Blay, M. L. Cardona, M. B. Garcia, and J. P. Pedro,Synthesis438 (1989).

[31 NEW CAGED GROUPS 55
is then added to the solution and the mixture stirred at 80° for 5 hr. After
cooling to room temperature, the reaction mixture is applied directly to
the column and eluted with chloroform. Evaporation of the solvent gives
almost pure 4-bromomethyl-7-hydroxycoumarin (420.4 mg, 1.636 mmol,
94.8% yield).
Synthesis of 4-Bromomethyl-7-propionyloxycoumarin
4-Bromomethyl-7-hydroxycoumarin,318.7 mg (1.240 mmol), and 10 ml
of dry acetonitrile are added to a 100-ml, round-bottomed flask equipped
with a magnetic stirring bar and CaC12 drying tube. Triethylamine (0.2 ml,
1.4 mmol) and propionyl chloride (0.2 ml, 2.3 mmol) are added and the
mixture stirred at room temperature for 20 min. The mixture is diluted
with CHC13and then washed sequentially with 0.5 N K2CO3,1N HC1, and
saturated NaCI. Drying over anhydrous MgSO4 and evaporation under
vacuum give 290.6 mg (0.9715 mmol, 78.3% yield) of 4-bromomethyl-7-
propionyloxycoumarin.
Synthesis of 4-(7-PropionyloxycoumarinyOmethyladenosine
Cyclic 3',5'-Monophosphate
This reaction is performed as described above except it uses 100 mg
(0.304 mmol) of cAMP, 341 mg (1.00 mmol) of 4-bromomethyl-7-propio-
nyloxycoumarin, 132 mg (0.570 mmol) of silver(I) oxide, 18 ml of acetoni-
trile, and 3 ml of DMSO and is stirred for 16 hr at 120°. The product is
purified by column chromatography (30 g of SIO2,6.2% methanol-CH2C12)
and then flash column chromatography (20 g of SIO2, 6.2% methanol-
CH2C12) to give 38 mg (0.0678 mmol, 22.6% yield) of PCM-cAMP as a
mixture of two stereoisomers (ax/eq = 3/2). 1H NMR (5% CD3OD/CDCI3,
ppm) (axial isomer) 1.27 (3H, t, J = 8 Hz, CH3), 2.64 (2H, q, J = 8 Hz,
CH2CH3), 4.50 (1H, d, J = 6 Hz, H'5), 4.67 (1H, m, H'4), 4.86 (1H, d,
J = 6 Hz, H'2), 5.41 (2H, d, J = 7 Hz, P-O-CH2), 5.50 (1H, m, H'3), 5.99
(1H, s, H'I), 6.65 (1H, t, J = 1 Hz, H3), 7.05-7.18 (2H, m, H5 and H8),
7.55 (!H, d, J = 9 Hz, H6), 7.94 (1H, s, adenine H2), 8.26 (1H, s, adenine
H8); (equatorial isomer) 1.28 (3H, t, J = 8 Hz, CH3), 2.64 (2H, q, J = 8
Hz, CH2CH3),4.47 (1H, d, J = 7 Hz, H'5), 4.67 (1H, m, H'4), 4.84 (1H,
d, J = 6 Hz, H'2), 5.4l (2H, d, J = 7 Hz, P-O-CH2), 5.50 (1H, m, H'3),
6.03 (1H, s, H'I), 6.61 (1H, t, J = 1 Hz, H3), 7.05-7.18 (2H, m, H5 and
H8), 7.56 (1H, d, J = Hz, H6), 7.97 (1H, s, adenine H2), 8.25 (1H, s, adenine
H8); IR (KBr, /)max Cm - l ) (mixture of two stereoisomers) 1734 ( C = O ) ,
1267 ( P = O ) , 1197 (C-O), 1135 (C-O), 1067 (C-O), 1017 (P-O); UV
[CH3OH, Xmax(e)] 265 (13,000), 314 (6600); MS (FAB +) 560 (M+H+), 330,
136, 57; exact mass (FAB +) ( M + H +) (calculated, 560.1182; observed,
560.1146).
56 CAGEDCOMeOUNDS [31
Synthesis of 4-(7-HydroxycoumarinyOmethyladenosine Cyclic
3',5'-Monophosphate
Silver(I) Oxide Promoted Method. This reaction is performed as de-
scribed above except that it uses 132 mg (0.401 mmol) of cAMP, 298 mg
(1.17 mmol) of 7-hydroxy-4-bromomethylcoumarin, 176 mg (0.759 mmol)
of silver(I) oxide, 18 ml of acetonitrile, and 3 ml of DMSO, and the reaction
is stirred for 24 hr at 80°. The products are purified by column chromatogra-
phy (20 g of SiO2,9% methanol-CH2Cl2) and then by flash column chroma-
tography (20 g of SiO2, 90% methanol-CH2Cl2) to give 32 mg (0.063 mmol,
16% yield) of HCM-cAMP as a mixture of two stereoisomers (ax/eq = 2/1).
Diazo-Based Method. This reaction is performed as described above
except that it uses 46.6 mg (0.142 mmol) of cAMP and 31.5 mg (0.156
mmol) of 4-diazomethyl-7-hydroxycoumarin.The reaction mixture is stirred
for 42 hr at room temperature and purified by flash column chromatography
(20 g of 8i02, 9% methanol-CH2C12) to give 5.5 mg (0.011 mmol, 7% yield)
of HCM-cAMP as a mixture of two stereoisomers (ax/eq = 3/5). 1H NMR
(5% CD3OD/CDC13, fi ppm) (axial isomer) 4.46 (1H, d, J = 7 Hz, H'5),
4.67 (1H, m, H'4), 4.83 (1H, d, J = 6 Hz, H'2), 5.38 (2H, m, P-O-CH2),
5.50 (1H, m, H'3), 5.98 (1H, s, H'1), 6.48 (1H, t, J = Hz, H3), 6.82-6.86
.(2H, m, H5 and H8), 7.43 (1H, m, H6), 7.96 (1H, s, adenine H2), 8.22 (1H,
s, adenine H8); (equatorial isomer) 4.46 (1H, d, J = 7 Hz, H'5), 4.67 (1H,
m, H'4), 4.83 (1H, d, J = 6 Hz, H'2), 5.39 (2H, m, P-O-CH2), 5.50 (1H, m,
H'3), 6.03 (1H, s, H'I), 6.44 (1H, t, J = 1 Hz, H3), 7.05-7.18 (2H, m, H5
and H8), 7.43 (1H, m, H6), 8.01 (1H, s, adenine H2), 8.29 (1H, s, adenine
H8); IR (KBr, Vm~xcm -1) (mixture of two stereoisomers) 1715 (C----O),
1272 ( P = O ) , 1207 (C-O), 1141 (C-O), 1067 (C-O), 1006 (P-O); UV
[CH3OH, Vmax@)] 257 (12,100), 326 (16,800); MS (FAB +) 504 (M+H+),
330, 261,157, 79.
Results and Discussion
Silver(I) Oxide Promoted Condensation Method

The advantages of this synthesis (Scheme 1) are (1) preparation of the
metal or ammonium phosphates is not necessary, (2) the reaction proceeds
under neutral conditions, (3) acid or base treatment is not required for the
workup or purification step, and (4) choice of incorporated substituents
depends only on the availability of the corresponding halide. To optimize
the conditions for the reaction between cAMP and benzyl bromide, the
effects of solvent, temperature, and reaction time on the product yield were
investigated. Unlike the case with phosphoric acid, no alkylated product
o
R1°P-OH
RI1 -I-
1/2 Ag20 iol R1-p--O'R
R~1 +
AgX
5. o A
SCHEME 1. Silver(I) oxide promoted condensation reaction.
was obtained when reactions were carried out in the absence of acetonitrile.
The reaction was found to proceed only when cAMP was dissolved in a
little DMSO, with a mixture of acetonitrile and DMSO (6:1, v/v) giving
the best result. The highest product yield was obtained at a temperature
between 60 ° and 80 °. The duration of a reaction also has an effect on the
yield, as shown in Fig. 2. Long reaction times resulted in a decreased
product yield, possibly because the cAMP triesters are hydrolyzed by water
generated during the course of the reaction.
Thin-layer chromatography (TLC) or HPLC can be used to monitor
the progress of the reaction, and it should be terminated when the amount
of product has reached its plateau value. The axial isomer always formed
more readily than the equatorial isomer (Fig. 2). This result is consistent
with earlier reports that showed the equatorial isomer of cAMP benzyl
100
80~ cAMP
o~" 60 ~/ Bn-cAMP(axial)
"o ~ ~ Bn-cAMP(equatorial)
0
1'0 2'0 310 410 5'0 60
Reaction time (h)
FIG. 2. Product distribution for silver(I) oxide promoted condensation of cAMP and
benzyl bromide.
TsOH/SiO2 ..v.tl.Cl
AcO O toluene HO O Et3N,CH2CI2 PrO O
RO
NH2 ~ )
Nv'~ NH2
OH (, i I N Br
I
.- Px-O"X ,
N.2K..,~
Ixl +
. ~ Ag20
~,.
)~
0 /~"" O
~ 1
(,N N
NI.~lx.,~
o" o ~ : J . o RO 0 0 o~.P~ "
OH CH3CN-DMSO u ~,.._.Jk
OH
MCM-cAMP : R=CH3 ACM-cAMP: R=CH3CO
HCM-cAMP : R=H PCM-cAMP: R=CH3CH2CO
SCHEME 2. Synthesis of coumarin-caged cAMPs by silver(I) oxide promoted method.
ester is hydrolyzed approximately four times faster than the axial isomer, 11
and because of the ground state stereoelectronic effects of six-membered-
ring phosphate esters reported by Gorenstein. 12
Synthesis of Coumarin-Caged cAMPs via Silver(l) Oxide Method

4-(7-Methyoxycoumarinyl)methyl and A C M - c A M P were synthesized
(Scheme 2) from the corresponding 7-substituted 4-bromomethylcouma-
rins, both of which are commercially available, in the presence of silver(I)
oxide, with a yield of 44 and 14%, respectively. Both products were obtained
as a mixture of axial and equatorial isomers of the underlying dioxaphos-
phorinan ring. The predominant isomer was axial, and the ratio of the
isomers was dependent on the reaction conditions (Fig. 1). The synthesis
of P C M - c A M P proceeds with modest yield (22%) using the corresponding
bromide that was prepared from commercially available 4-bromomethyl-
7-acetoxycoumarin. Although H C M - c A M P was prepared using the same
method, with a yield of 16% for one particular run, this synthetic approach
was not found to be reproducible. The reason for this result is not yet clear,
although the free phenolic hydroxyl substituent on the coumarin ring may
affect the condensation reaction.
11j. Engels and E. J. Schlaeger, J. Med. Chem. 20, 907 (1977).

12 G. Gorenstein, Chem. Rev. 87, 1047 (1987).
[31 NEWCAGEDGROUPS 59
Synthesis of Coumarin-Caged cAMPs via Diazo-Based Method

Ito and Maruyama a3 reported the preparation and characterization of
several 7-substituted 4-diazomethylcoumarins. The reaction of cAMP with
the 7-substituted (4-diazomethyl)coumarins in DMSO, carried out under
conditions similar to those for o-nitrophenyldiazoethane, TM yield the corre-
sponding coumarin-caged cAMPs with a yield of 10-20% The preparation
of 7-hydroxy-4-diazomethylcoumarin takes five step sequences with 46%
overall yield starting from the commercially available 4-methylum-
beriferon. 13
Purification and Structural Determination

Isolation and purification of the caged cAMP from the crude product
mixture are usually carried out by column chromatography on silica gel
using 4.7% methanol-dichloromethane or 6.2% methanol-dichloro-
methane as an eluent. Further purification by flash chromatography gave
almost pure caged cAMP. Separation of diastereomeric isomers was carried
out by semipreparative reversed-phase HPLC using 50% methanol-H20
as an eluent. Determination of the structure of each isomer was performed
mainly by 1H NMR spectroscopy. Although the coumarinylmethyl-cAMPs
are not very soluble in CDC13, they are soluble in CDC13 containing 5%
C D 3 O D . The configurations of the products were assigned on the basis of
t h e 1H NMR chemical shift differences with respect to the resonance of
the H8 proton of the adenine ring. Thus the isomer showing the lower shift
was assigned as the axial stereochemistry, in analogy with previous reports) 5
A similar propensity was observed for other cAMP triesters, benzyl (Bn),
naphthylmethyl (NM), and 2-nitrobenzyl (NB) esters of cAMP. The ratio
of the axial and equatorial isomers in the mixture was determined using
the signals of axial and equatorial isomers, which appear as two discrete
singlets around 8 ppm. For example, mixtures of the axial and equatorial
isomers of MCM-cAMP show two discrete singlets at 8.25 and 8.28 ppm.
However, as the axial isomer elutes faster than the equatorial isomer on
a reversed-phase column, it was more convenient to determine the yield
of each isomer using HPLC. Selected chemical shifts and the retention
times are summarized in Table I.
13 K. Ito and J. Maruyama, Chem. Pharm. Bull 31, 3014 (1983).

14j. W. Walker, G. P. Reid, and D. R. Trentham, Methods in Enzymol. 172, 288 (1989).
15j. Engels and A. Jager, Arch. Pharm. (Weinheim) 315, 368 (1982).
60 CA6EO COMPOUNDS [31
TABLE I
SELECrED 1H NMR CHEMICALSmvrs AND HPLC RETENXIONTIMES
1H NMR (ppm)
Substrate Isomer H8 (Ad) H2 (Ad) Retention time (min)"
MCM-cAMP ax 8.25 7.92 9.9

eq 8.28 8.03 12.3
ACM-cAMP ax 8.28 7.89 8.4
eq 8.29 7.93 10.1
PCM-cAMP ax 8.26 7.94 12.2
eq 8.25 7.97 15.9
HCM-cAMP ax 8.22 7.96 34 b
eq 8.29 8.01 45 b
Bn-cAMP ax 8.18 7.82
eq 8.31 7.82
NM-cAMP ax 8.11 7.89 6.3 c
eq 8.26 7.85 12.W
NB-cAMP ax 8.20 7.88 6.7 a
eq 8.33 7.85 9.4 d
"The mobile phase was 50% methanol-50% water (v/v), and the flow rate was 0.6 ml/
min unless otherwise noted.
b The mobile phase was 35% methanol-65% water (v/v).
c The mobile phase was 60% methanol-40% water (v/v), and the flow rate was 0.8 ml/min.
d The flow rate was 0.8 ml/min.
Stability of Caged cAMP Compounds in the Dark

T h e hydrolytic stability of the v a r i o u s c A M P derivatives in 1% D M S O
in R i n g e r ' s s o l u t i o n stored in the d a r k is q u i t e variable, a n d the half-life
of each c o m p o u n d is s h o w n in T a b l e II. M C M - c A M P exhibits the highest
stability u n d e r these conditions, w h e r e a s the half-lives of the acyloxycou-
m a r i n - c a g e d c A M P s , A C M - c A M P a n d P C M - c A M P , are far shorter, pre-
s u m a b l y b e c a u s e the C7-acyloxy g r o u p is susceptible to hydrolysis in physio-
logical salt s o l u t i o n a n d A C M - c A M P is q u a n t i t a t i v e l y c o n v e r t e d into
H C M - c A M P o n p r o l o n g e d storage in R i n g e r ' s solution. T h e r e f o r e , particu-
lar care m u s t be t a k e n in the storage of the a c y l o x y c o u m a r i n - c a g e d c A M P s .
W e r e c o m m e n d that all c o m p o u n d s b e stored in a solid f o r m in a freezer
or as 10 m M D M S O stock solutions at - 2 0 °, w h e r e they are stable for
several m o n t h s .
Photochemical and Spectral Properties of Caged cAMP Compounds

I r r a d i a t i o n of M C M - c A M P in 1% D M S O in R i n g e r ' s s o l u t i o n gave the
p a r e n t c A M P a n d 4 - h y d r o x y m e t h y l - 7 - m e t h o x y c o u m a r i nas p h o t o p r o d u c t s . 7
T A B L E II
HALF-LIVES OF CAGED cAMPs"
Half-lives (hr) c
Substrate b Axial Equatorial Reference
MCM-cAMP 1000 60 7
ACM-cAMP 17.4 15.5 This work
PCM-cAMP 25.6 18.8 This work
HCM-cAMP 28.8 18.9 This work
a In 1% D M S O in Ringer's solution.
b 10 4 M solutions in Ringer's solution containing 1%

DMSO.
c The exact concentrations were determined by HPLC
and analyzed by pseudo-first-order kinetics, using
2-hydroxymethylanthraquinone as an internal stan-
dard, which was found to be stable and had no effect
on caged cAMPs.
No significant differences were observed in the photochemical reactivity

of the axial and equatorial isomers, and therefore it is not necessary to
separate the two coumarin-caged cAMP stereoisomers. The mechanism
proposed for this photolysis reaction is presented in Scheme 3.
The absorption maxima ('~max) and molar absorptivity (e) at 340 nm
for the different coumarin-caged cAMPs are summarized in Table III.
Absorption spectra were measured in methanol because A C M - c A M P and
PCM-cAMP are gradually hydrolyzed to HCM-cAMP in Ringer's solu-
tion. Introduction of the acyloxy substituent on the C-7 position of the
coumarin ring shifts the absorption maximum to a shorter wavelength by
about 10 nm and reduces the absorptivity to almost one-half compared to
o F o 7~ o
,O-P-R / O-P-R/ e H
/~ h~ ~ CH~
H3CO O H3CO O 30 0 O
OH
H20 o
II
i, + HO-P-R
H3CO O I~
SCHEME 3. Proposed mechanism for photolysis of coumarin-caged phosphate.
TABLE III
PHOTOCHEMICAL AND SPECTRAL PROPERTIES
Substrate ~ Amax(e)b ~3340b dl0disr'd
MCM-cAMP 325 (13,300), 259 (14,600) 9,600 0.12

ACM-cAMP 313 (7,650), 267 (13,400) 2,600 0.056
PCM-cAMP 313 (6,500), 265 (12,800) 1,900 0.054
HCM-cAMP 326 (16,800), 256 (12,100) 13,100 0.062
a All samples were used as a mixture of stereoisomers.

o In methanol.
c In 1% DMSO in Ringer's solution.
a Determined by potassium ferrioxalate actinometry.
those with methoxy or hydroxy substitution. The quantum efficiencies for

the disappearance (q~ois) of the following compounds in Ringer's solution
containing 1% DMSO were determined to be 0.12 for MCM-cAMP, 0.056
for ACM-cAMP, 0.054 for PCM-cAMP, and 0.062 for HCM-cAMP, re-
spectively (Table III).
Biological Activity of Caged cAMP Compounds

To evaluate the biological function of the coumarin-based caged cAMPs,
the cAMP-dependent granule dispersion response of fish melanophores
TABLE IV
EFFICIENCY OF PHOTOCHEMICAL CONVERSION
OF CAGED c A M P AND cAMP-DEPENDENT DISPERSION
OF MELANOPHORES
Magnitude of Conversion of
Substrate dispersion (%)~ substrate (%)b
MCM-cAMP 23 - 7 60
ACM-cAMP 39 --- 6 23
PCM-cAMP 54 __+4 9
HCM-cAMP 11 ± 7 64
NPE-cAMP 25 ± 10 32
a Dispersing responses of pigment granules after 10 sec

irradiation through the objective lense of the fluores-
cent microscope.
b Photlytic consumption of caged cAMPs (100/zM solu-
tion in 1% DMSO in Ringer's solution) after 10 sec
irradiation through the objective lens of the fluores-
cence microscope.
[4] CAGED FLUORESCENT PROBES 63
isolated from the scales of the dorsal trunk of the wild-type medaka fish
(Oryzias latipes) was studied. The extent of the dispersion of melanophores
is controlled by the intracellular concentration of cAMP, and it is known
that cAMP cannot pass through the membrane of live melanophores. The
assay is based on an analysis of the distribution of phase-dense granules
by phase-contrast microscopy before and after irradiation of caged cAMP
loaded cells with near ultraviolet delivered from a mercury arc lamp. 7,9
Image analysis of the extent of granule dispersion after photolysis provides
an accurate method to quantify the intracellular concentration of cAMP.
To clarify the effect of extracellularly applied caged cAMP, melanophores
were pretreated with an ot2-adrenergic agonist (norepinephrine) together
with a phosphodiesterase inhibitor, 3-isobutyl-l-methylxanthine (IBMX),
which blocks the reaggregation of the dispersing melanin pigments by the
intracellular phosphodiesterase (see experimental procedures in Refs. 7
and 9). The results of the dispersing responses for these coumarin-caged
cAMPs are shown in Table IV, which also documents the efficiency of
these photoactivation reactions.
[4] C a g e d F l u o r e s c e n t Probes
By T. J. MITCHISON,K. E. SAWIN,J. A. THERIOT,K. GEE,
and A. MALLAVARAPU
Introduction
Fluorescence microscopy is a powerful tool for observing the distribu-
tion of specific molecules in living cells. In many cases, however, it is not
possible to infer the dynamic behavior of populations of molecules when
they are uniformly labeled with a fluorescent tag. Instead it is necessary to
create a local region within the cell where the fluorescence intensity is
different from the bulk population and to follow the evolution of this region
with time. The classic method for creating a differentially labeled region
is fluorescence photobleaching. This method has been used extensively
to follow diffusion and directed movements of macromolecule in cells.
Photobleaching suffers two main problems: (1) the chemistry of bleaching
depends on generation of activated oxygen species that can cause local
damage to proteins and membranes, and (2) it can be difficult to accurately
track a region of reduced fluorescence within a pool of higher fluorescence
under conditions of limited signal.

isolated from the scales of the dorsal trunk of the wild-type medaka fish
(Oryzias latipes) was studied. The extent of the dispersion of melanophores
is controlled by the intracellular concentration of cAMP, and it is known
that cAMP cannot pass through the membrane of live melanophores. The
assay is based on an analysis of the distribution of phase-dense granules
by phase-contrast microscopy before and after irradiation of caged cAMP
loaded cells with near ultraviolet delivered from a mercury arc lamp. 7,9
Image analysis of the extent of granule dispersion after photolysis provides
an accurate method to quantify the intracellular concentration of cAMP.
To clarify the effect of extracellularly applied caged cAMP, melanophores
were pretreated with an ot2-adrenergic agonist (norepinephrine) together
with a phosphodiesterase inhibitor, 3-isobutyl-l-methylxanthine (IBMX),
which blocks the reaggregation of the dispersing melanin pigments by the
intracellular phosphodiesterase (see experimental procedures in Refs. 7
and 9). The results of the dispersing responses for these coumarin-caged
cAMPs are shown in Table IV, which also documents the efficiency of
these photoactivation reactions.
[4] C a g e d F l u o r e s c e n t Probes
By T. J. MITCHISON,K. E. SAWIN,J. A. THERIOT,K. GEE,
and A. MALLAVARAPU
Introduction
Fluorescence microscopy is a powerful tool for observing the distribu-
tion of specific molecules in living cells. In many cases, however, it is not
possible to infer the dynamic behavior of populations of molecules when
they are uniformly labeled with a fluorescent tag. Instead it is necessary to
create a local region within the cell where the fluorescence intensity is
different from the bulk population and to follow the evolution of this region
with time. The classic method for creating a differentially labeled region
is fluorescence photobleaching. This method has been used extensively
to follow diffusion and directed movements of macromolecule in cells.
Photobleaching suffers two main problems: (1) the chemistry of bleaching
depends on generation of activated oxygen species that can cause local
damage to proteins and membranes, and (2) it can be difficult to accurately
track a region of reduced fluorescence within a pool of higher fluorescence
under conditions of limited signal.

These considerations led us to develop fluorescence photoactivation

as an alternative technology for probing cytoskeleton dynamics. In this
approach the protein is tagged with a caged fluorochrome. This probe
molecule is nonfluorescent (or fluorescent at a different wavelength) until
illuminated with a brief pulse of ultraviolet light. Such illumination leads
to photolysis of the caging groups and generation of a fluorescent species.
In principle, fluorescence photoactivation can avoid the problem of genera-
tion of local oxidative damage inherent to photobleaching and can also
produce a more favorable signal-to-noise ratio for imaging. Photoactivation
can also produce toxic by-products in the form of the nitrosoaldehyde or
nitrosoketone side products from photolysis. However, each molecule of
caged fluorochrome releases only one or two molecules of side product,
whereas each molecule of fluorochrome in a bleaching experiment probably
produces hundreds of activated oxygen molecules before being bleached
by reacting with one of them. The main limitation of the photoactivation
method has been the development of suitable caged fluorescent probes,
and most of this article focuses on our progress in this area.
Requirements for an effective caged fluorescent probe have been dis-
c u s s e d . 1'2 Most important are biostability, rapid and efficient photoactiva-
tion, good brightness and photostability of the uncaged fluorochrome, prac-
ticable synthesis, and friendly protein chemistry. The latter is most difficult
to predict from molecular structure--empirically it requires probes that
are neither too hydrophobic nor too highly charged, with convenient chem-
istry for covalent attachment to proteins.
To date, caged fluorochromes have all employed variants of the
2-nitrobenzyl caging group. Two properties have been used to control
fluorescence with such groups. In caged fluorescence and rhodamines (Rds),
two caging groups act together to pin the xanthene fluorophore in a nonflu-
orescent lactone tautomer form (Figs. 1 and 3). In caged resorufin (Fig. 2)
the fluorophore is held in a nonionizable form, resulting in a blue shift of
adsorption and quenching of fluorescence.
The first caged fluorescent probe that was used for a biological experi-
ment was a fluorescein derivative, C2CF-sulfo-N-hydroxysuccinimide
(SNHS) (Fig. 1). This probe attached to tubulin led to the discovery of
poleward flux in mitotic spindles. 1,3 C2CF is highly hydrophobic, and most
proteins other than tubulin tend to aggregate if they are labeled with it.
In C2CF-SNHS itself, the sulfosuccinimide group keeps the reagent water
1 T. J. Mitchison, J. Cell Biol. 109, 637 (1989).

2 T. J. Mitchison, K. S. Sawin, and J. A. Theriot, in "Handbook of Cell Biology," Vol. 2
(J. E. Celis, ed.), pp. 65-76. Academic Press, New York, 1994.
3 K. E. Swain and T. J. Mitchison, J. Cell Biol. 112, 941 (1991).
[41 CAOED FLUORESCENT PROBES 65
~NO2 NO2
.~~o~o.~o~.~
R3 ~ R3
(RO" "~ "0
RI
R1 R2 R3 Name
O
~1 ~ ~',=/SO3Na
8~ V ~O--N~,~T H H C2CF-SNHS
(mixed isomers) g/
O
-O-CH2-COOH H CMNB2AF-NHS
(single isomer) d/
O
/~.N/N/N/~.o_.~L"
(mixed isomers)
H -COOH (~-Carboxy-
C2CF-NHS
FIG. 1. Caged fluoresceins.
O 0
"~'~°~°G;° o
o oGo
DMF
NO2 24hr O~ NNj,J
N2
CR=IA(isomer2)
FIG. 2. Structure and synthesis of caged resorufin iodoacetate.
R3~ , ~ NO2 O2N'~,~,,J R3
R4
R1
R1 I R2 R3 R4 Name
O
-H -H CMNB2QRd-NHS
O,,oi
-H COOH a-Carboxy-
C2CQRd-NHS
(mixedisomers) O//
O
-OCH3 COOH ct-Carboxy-DM-

C.2CQRd-NHS
-oc~ COOH a-Carboxy-DM-

C2CQRd-IA
(mixedisomers)
Fie. 3. Caged Q-rhodamines.
soluble. As this group hydrolyzes during a labeling reaction, aggregates

and precipitates of the reagent usually develop. These problems led us to
develop the more water-soluble caged fluorescein CMNB2-AF-N-hydroxy-
succinimide (NHS) (Fig. 1), whose applicability is more general. The search
for a caged fluorescent probe for actin labeling led to caged resorufin (Fig.
2, see Theriot and Mitchison4). Both fluorescein and resorufin suffer rapid
photobleaching after activation, which severely limits our ability to image
them in cells. This photostability consideration led us to develop the caged
rhodamines (Fig. 3). a-Carboxydimethoxy-C2CQRd-NHS and -IA are
probably the best caged fluorochromes we currently have in terms of the
parameters listed earlier. A practical drawback has been difficulty of synthe-
sis, and in general the area of caged fluorochrome design has been limited
4 j. A. Theriot and T. J. Mitchison, Nature 352, 126 (1991).

141 CAGED FLUORESCENT PROBES 67
TABLE I
PRACTICAL CONSIDERATIONS FOR DIFFERENT FLUOROPHORES
Excitation/emission Quantum
Caged maxima (nm) Photostability Biostability efficiency of Ease of
derivatives of (uncaged) (uncaged) (caged) uncaging synthesis
Fluorescein 485-495/495-520 Poor Good Good Medium

Q-Rhodamine 540-550/560-580 Good Good OK-poor Harder
Resorufin 580-590/590-610 Poor Poor Good Easy
by the chemistry. The advantages and drawbacks of the different caged

fluorochromes are summarized in Tables I and II. Data we have summarized
are largely anecdotal, although they coincide with the experience of other
laboratories and also theoretical expectations for the different chemistries.
TABLE II
PRACTICAL CONSIDERATIONSFOR DIFFERENT CAGING GROUPSa
Water Ease of
Caging group solubility synthesis Notes
Nitrobenzyl (as in C2CF- Poor Easy Simplest chemistry

SNHS) [NB]
Nitrophenethyl (as in caged Poor Medium Faster uncaging than nitroben-
resorufin) [NPE] zyl. By-product less toxic
than nitrobenzyl
2-Nitro-5- Good Medium Carboxymethoxy group imparts
carboxymethoxynitrobenzyl water solubility. Photochemis-
(as in CMNB2AF) [CMNB] try similar to nitrobenzyl
a-Carboxynitrobenzyl (as in Good Harder a-Carboxy imparts water solu-
a-carboxy-C2QRd-NHS) bility and less toxic by-prod-
[CNB] uct. Also increases efficiency
of photolysis; promotes uncag-
ing at visible wavelengths
a-Carboxy-4,5- Good Harder As simple a-carboxy. Dimeth-
dimethoxynitrobenzyl (as in oxy groups improve efficiency
a-carboxydimethoxy- of uncaging at 360 nm
C2QRd-NHS) [CDMNB]
Efficiency of photolysis depends on both the caging group itself and the chemistry that
attaches it to the fluorochrome. The phenolic ether linkage in caged fluorescein and
caged resorufln is readily cleaved by light with all caging groups. The carbamate linkage
in caged rhodamines is more resistant to cleavage, a-Carboxy caging groups improve
the efficiency of carbamate photolysis. By-product toxicity refers to the nitrosoaldehyde
or ketone formed from the caging group after photolysis. Uncaging at visible wavelengths
is an undesirable special property of a-carboxy caging groups whose origin is unclear.
For more information, see Haugland.5 The trivial names for caging groups shown in
square brackets indicate the nomenclature used in that reference.
This information is provided in the spirit of helping to guide others in
designing biological experiments and hopefully designing new chemistry.
For additional information, see Haugland. 5
Probe Availability and Synthesis

Progress on application of fluorescence photoactivation to biological
problems has largely hinged on probe availability and properties. Commer-
cial availability depends on demand and ease of synthesis. Several caged
fluorescent probes are currently available from Molecular Probes (Eugene,
OR) .5 Where these do not correspond exactly to the probes we have worked
with, the considerations listed in Tables I and II may be useful for pre-
dicting properties.
Caged fluoresceins (Fig. 1) are fairly easy to synthesize, starting from
carboxyfluorescein (usually as a mixture of isomers) or aminofluorescein
(available as single isomers). Our synthesis of C2CF was reported in Mitchi-
son 1and of CMNB2AF in Mitchison et aL2 T h e key synthetic step, formation
of the bisphenolic ether, used conditions developed by Krafft et aL 6 The
main drawback of caged fluoresceins is the lack of photostability of fluores-
cein itself. Introducing fluorine atoms increases the photostability of fluo-
rescein and decreases its pKa.5 Thus caged derivatives of 2',7'-difluorocar-
boxyfluorescein should in principle make superior probes.
Caged resorufin (Fig. 2) is the easiest probe to synthesize and is also
the probe with lowest molecular weight, which should improve its protein
chemistry. It is made simply by applying the Trentham diazo caging chemis-
try described in Walker et al. 7 to commercial resorufin iodoacetate (Boeh-
ringer, Mannheim, Germany) and using thin-layer chromatography (TLC)
to purity the product. 4 The two isomers can be readily separated, but they
seem to have similar properties. If necessary, the 4-carboxyresorufin nucleus
can be synthesized by condensing 4-nitrosoresorcinol and 2,6-dihydroxy-
benzoic acid in the presence of manganese dioxide and reducing the re-
suiting 4-carboxyresasurin with zinc. Caged resorufln has two drawbacks:
(1) like fluorescein, resorufin is readily photobleached; because there is a
strong Hg line (585 nm) centered on the resorufin adsorption maximum
(unlike fluorescein), photobleaching may appear to be very fast indeed;
and (2) because the heterocyclic nitrogen atom makes the resorufin ring
5R. P. Haugland, in "Handbookof FluorescentProbes and Research Chemicals." Molecular

Probes, Eugene, OR, 1996. Pages 447-455, for caged compounds; pp. 22-25 for fluores-
cein substitutes.
6G. A. Krafft, W. R. Sutton, and J. P. Cummings,J. Am. Chem. Soc. 110,301 (1988).
7j. W. Walker, G. P. Reid, and D. R. Trentham, Methods Enzymol. 172,288 (1989).
system electron deficient, it tends to suffer nucleophilic attack and reduction

more readily than the other fluorochromes. This tendency is exacerbated
in the caged form, where anion formation is blocked (see, e.g., Afanas'eva
et al. 8 Under physiological conditions (pH 7.4, 1-5 mM thiol groups) we
have found that caged resorufin can undergo both nucleophilic attack and
reduction, depending on the thiol used. We have also found that the half-
life of caged resorufin in cells is quite short, as little as 20 min when attached
to dextran. Attachment to actin prolongs the half-life of the probe, but
these chemical considerations make it less attractive for general use. In
principle, appropriate substitution of the resorufin ring might ameliorate
these problems, but this has not been explored.
Caged rhodamines are the best of the current probes for most applica-
tions due to the relative resistance of rhodamines to photobleaching. The
synthesis, shown in Fig. 4, is more difficult, and we encourage others to try
and improve on our methods. The key synthetic step is the acylation reaction
(F in Fig. 4). We used a strong base to deprotonate the amino groups
and then added a strong acylating agent (a chloroformate) in a rather
uncontrolled reaction with variable and poor yields. The identity of the
desired product after purification was confirmed by mass spectrometry.
This method was based on Q-rhodamine alkylation conditions described
in Arnost et al. 9 All our caged rhodamines are based on the rhodamine
derived from 7-hydroxyquinoline, to which we have given the trivial name
Q-rhodamines. Q-Rhodamines have very similar optical properties to the
popular tetramethyl-rhodamines, including an adsorption maximum near
the strong Hg line at 546 nm, high absorption coefficients (>80,000), high
quantum efficiencies of fluorescence, and good photostability. Caged rhoda-
mines based on rhodamine 110 (no substitution on the nitrogens) are also
useful for making caged derivatives, and the acylation chemistry is some-
what easier. These probes have adsorption maxima near fluorescein (495-
515 nm) and are called Rhodamine Green by Molecular Probes. 5 A milder
acylation strategy using Steglich's reagent (4,6-diphenylthieno[3,4-d]-l,3,di-
oxol-2-one 5,5-dioxide) that works well with rhodamine 110 derivatives has
been developed (Ottl and Mariott, submitted).
Because caged rhodamines are carbamates rather than phenolic ethers,
their quantum efficiency of uncaging is inherently lower than caged fluores-
ceins. This translates into needing longer pulses of UV light to uncage,
with the possibility of damaging the irradiated region of the cell. We have
found in practice that only caged rhodamines with a-carboxy caging groups
8 G. B. Afanas'eva, T. S. Viktorova, K. I. Pashkevich, and I. Y. Postovskii, Chem. Heterocyclic

Compounds 10, 302 (1974).
9 M. J. A r n o s t et al., Polaroid Corp., M A . US Patent n u m b e r 4,900,686, 1990.
}~OH /~ OCH3
OHC'~n~3~ OCH3
1)AqNaHSO3
OoN ~ [ 2)AqNaCN
A H2, Pb20 C ~ 3) Aq HC1,reflux
.oo~ ,¢--(°°"~
HOH ' ~ ' ~ . ~ OCH3
O2N / ~ OtBu
[ HOOC
B ~ C2HsCOOH,TsOH,heat
V
/OCH3
tBuOOC
H O ~ O C H a
f•,,COO
1. II
O2N
I coc12,2,6lufidine,
(mixedisom CH3
tBuOOC
1) Di,,olv~ C-Q~d in~-,-~/~v~A ,.,_~// \~_,~,.=
2) K+tBuO" 3 eqiv. . . . /~"~k /t~ .... 3
3) Add chloroformate3 equiv. ~ " ~ O H " /
/
O2N
H3CO% F I / OCH3
?o~.~u v ,~uooc
H3CO-~ ~"~O~ ~O'--~'~'~ ~OC" 3
,_~ -. ~o o~ .. )_=,
NO2 ~ O2N
--- ~ . ~ C O
HOOC
oci oci
diaminobutan G
/
I
H ~_ H Alpha-Carboxy-DM-
IA-NItS, 98% TFA C2CQRd-IA
r,mt3, OMr
H2N HN\
CO
<,
[41 CAGED FLUORESCENT PROBES 71
uncaged fast enough to be used as nonperturbing probes of the actin cy-

toskeleton. The a-carboxy group is known to facilitate the uncaging reaction
in other caged compounds. 5 The o~-carboxy caging groups also promote
water solubility and better protein chemistry, compensating for their more
difficult synthesis. The a-carboxy caging groups have one undesirable lea-
FIG. 4. Caged Q-rhodamine synthesis. Yields were in the 50-95% range except where
indicated. Identification of compounds was based on optical spectroscopy and chemical proper-
ties, except for step F, where mass spectrometry was also used to characterize the product.
(A) 0.2 M 7-hydroxyquinoline (Eastman) in ethanol was treated with 5 mol% Pt20, 2 atm
H2, 24 hr. Solvent evaporated, no purification needed. (B) Equimolar amine and anhydride
plus 0.2 equivalents of p-toluenesulfonic acid dissolved in minimal volume of propionic acid.
Refluxed 24 hr. Extra propionic acid is added and distilled off twice to remove water. Carboxy-
Q-rhodamine purified by two rounds of dissolving in hot aqueous 1.0 M HC1, filtering, cooling,
and collecting the precipitate, followed by chromatography on Sephadex LH20 with isocratic
elution in methanol. (C) Conditions described in Org. Synth. Coll. 1, pp. 336-339. Commercial
6-nitro-veratraldehyde contains impurities that do not dissolve in aqueous bisulfite and can
be removed by filtration. Product is recrystallized from CH2Cl2/hexane. (D) 2-Nitro-4,5-
dimethoxynitromandelic acid dissolved in tetrahydrofuran (THF). Two equivalents of the
isourea 16 are added. After 2 hr, product is purified by chromatography (SIO2, 70% hexane,
30% ethyl acetate). (E) Ester is dissolved in THF; 1.5 equivalents of base is added, followed
by 1.2 equivalents of phosgene (as a toluene solution). After 10 min the solution is filtered
and evaporated. Crude chloroformate is held under vacuum for several hours and used without
further purification. (F) Rhodamine is dissolved at 0.1 M in 50% THF, 50% HMPA. K+-t -
BuO (1M in THF) is added with mixing until the solution becomes blue-black (approximately
3 to 4 equivalents). The chloroformate (0.1 M in THF, 3 equivalents) is then added all at
once with mixing. The reaction generates many products. The correct one is identified on
TLC (SIO2, 84% benzene, 14% acetone, 2% acetic acid, Rf = 0.3) by its conversion from
colorless to red on irradiation with 360-nm light. Purify by chromatography (SIO2, 94-90%
dichloromethane, 4-8% ethyl acetate, 2% acetic acid). Product is further purified by taking
up in a minimal volume of hot methanol, cooling to -20 ° overnight, and collecting the
precipitate. Identity of the product is confirmed by mass spectrometry (lsims, strong molecular
ion at 1130.5) and visible spectroscopy before and after activation. Yield: Poor, variable. (G).
Caged rhodamine in DMF is treated with benzotriazol-l-yltetramethyluronium tetrafluoro-
borate (TBTU) 2 equivalents, then diaminobutane, 3 equivalents. Crude product is isolated
by dissolving in ethyl acetate and washing 3× with water. (H) Product from G in DMF is
treated with iodoacetic acid N-hydroxysuccinimide ester, 2 equivalents, and triethylamine, 2
equivalents in DMF. Purify by chromatography (SIO2, 18% dichloromethane, 80% ethyl
acetate, 2% acetic acid). (I) Product of H is deprotected in 98% trifluoroacetic acid (TFA),
2% water, 2 hr. Acid is removed in vacuo. Product is dissolved in THF and precipitated with
9 volumes of cold water. For convenience, aliquots of approximately 1/.~mol are precipitated
in individual microfuge tubes. The product is collected by centrifugation, dried in vacuo, and
stored in the dark at - 8 0 °. For protein labeling, an aliquot was resuspended in DMF. The
NHS derivatives (Fig. 3) are synthesized by treating the protected, caged-QRd-COOH with
NHS and ethyldimethylaminopropylcarbodiimide (EDC) in DMF to give the NHS ester,
followed by e-aminocaproic acid in DMSO/water/triethylamine to give the aminocaproyl
derivative, followed by NHS and EDC in DMF to give the final product.
ture we do not understand. They uncage fairly fast in visible light, for
example, at 450-520 nm for a-carboxy-4,5 dimethoxy-C2QRd-NHS.
o~-Carboxy-caged fluorescein (Fig. 1) undergoes similar visible uncaging.
This property is peculiar because neither the caging groups nor the lactone
tautomer of the fluorochrome should absorb light at these wavelengths. We
hypothesize that a-carboxy-caged fluorochromes may contain a significant
proportion of the nonlactone tautomer of the fluorochrome in their ground
state structure that can absorb light and somehow uncage. This uncaging
in visible light has the potential to cause slow activation of the whole field
during the observation phase of a photoactivation experiment. In practice,
this problem can be largely circumvented by exciting the uncaged fluoro-
chrome with light tightly centered on the adsorption maximum of the
uncaged form. For Q-Rds this means isolating the 546-nm Hg line. The
level of this light must be controlled anyway to prevent bleaching, and under
these conditions further uncaging by the observation light is minimized.
Comparing o~-carboxynitrobenzyl caging groups with and without 4,5-
dimethoxy substitution (Fig. 3, Table II), we have found that these extra
groups significantly increase the rate of photoactivation of caged Q-Rd with
366-nm light. This is presumably due to increased adsorption of photons at
366 nm with the methoxy substituents. The effect is less than the a-carboxy
group, but still useful.
Protein Labeling
Most of our experience has been with two proteins, tubulin and actin.
Tubulin must be labeled in the polymerized state to protect cysteines and
lysines essential for polymerization. We have been successful using only
NHS and SNHS esters, which presumably modify lysines. Using polymer-
ized tubulin at pH 8.0-8.6 in glycerol containing buffers and a large excess
of reagent, stoichiometries of up to 0.5 caged fluorochrome/tubulin dimer
are obtained. These labeled tubulins are somewhat compromised in their
polymerization ability, but appear to copolymerize readily with endogenous
tubulin. C2CF-SNHS has been used for most work with tubulirl, although
the other probes work as well. The basic protocol we use to label tubulin
on lysines with NHS-ester probes is described in Hyman et at lo An updated
version of this procedure is described in Desai and Mitchison (Methods
Enzymol., in press).
Rabbit muscle actin contains a single reactive cysteine (Cys-374), which
has been used as a site of attachment for a variety of fluorescent dyes. We
10A. A. Hyman,D. Drexel, D. Kellog,S. Salser, K. Sawin, P. Steffen,L. Wordeman,and

T. J. Mitchison,Methods Enzymol. 196,478 (1991).
reacted actin monomers with a-carboxydimethoxy-C2CQRd-IA using a

protocol that minimizes the amount of time required for the preparation.
We found that this improved yield and minimized hydrolytic activation of
the caged rhodamine.
For a typical labeling reaction, 0.5 ml actin (5 mg/ml) is used. Care is
taken to minimize oxidation of the cysteine prior to labeling. For this
reason, actin monomer, purified according to the method of Pardee and
Spudich, 1t is quick frozen and stored in G-buffer (5 mM Tris-Cl, 0.2 mM
CaC12, 0.2 mM ATP, pH 8.0) + 1 mM dithiothreitol (DTT). To remove
excess thiol prior to labeling, the actin is thawed and dialyzed against
G-buffer + 10 gtM DTT for 2 hr at 4°. Because ct-carboxydimethoxy-
C2CQRd-IA is sparingly soluble in water, we add it at relatively low concen-
tration [20 mM in dimethylformamide (DMF)] at a 5:1 molar ratio to
actin. Addition and mixing are performed with rapid vortexing. As with
the synthesis of caged compounds, these steps must be performed in safe-
light. All tubes should be wrapped in an opaque material to occlude light.
The solution is then brought to room temperature, and the reaction is
allowed to proceed for 30 min to 1 hr. Following this, the sample is put on ice.
Free unreacted dye is separated from protein by gel filtration [Pharmacia
(Piscataway, NJ) 5 ml Dextran desalting columns] on a column preequili-
brated with cold G-buffer + 1 mM DTT. The protein peak is identified by
spotting 1/zl of each fraction on paper (nitrocellulose paper or TLC plate)
and irradiating the spots for 2-3 min with long-range (360-nm) UV light.
Hand-held UV lamps (e.g., ENF-260C, Spectronics Corp., Westbury, NY)
work well for this. The fractions (0.5-1 ml) corresponding to the first red/
pink peak are pooled. Actin is then polymerized by adjusting the buffer
to 75 mM KC1, 50 mM Tris, pH 8.0, 1 mM MgC12, and 1 mM ATP. The
solution is left on ice for 1 hr and is then brought to room temperature for
1 hr. Filaments are pelleted (50,000 rpm, Beckman TLA 100.3 rotor, 1 hr,
4°) and washed once with G-buffer + 1 mM D'I-T. The pellet should be
glassy and translucent. Filaments are resuspended by adding 100 ~l G-
buffer + 1 mM DTT to the pellet. After a few minutes incubation at 0° to
allow swelling, the pellet is broken up and then sonicated briefly. The
supernatant is collected and the tube is reextracted with another 100/zl
G-buffer + 1 mM DTT to collect any remaining protein. Filaments are
depolymerized by dialysis against G-buffer + 1 mM DTT for 24 hr at 4 °.
For the final 2-3 hr of dialysis, the dialysis solution is replaced by G-Buffer
+ 0.5 mM glutathione. Aggregates are removed by centrifugation (44,000
Krpm, Beckman TLA100 rotor). The supernatant is collected and quick
frozen and stored in liquid N:, typically as 3-/zl aliquots.
11 j. D. Pardee and J. A. Spudich, Methods Cell Biol. 24, 271 (1982).

Using actin labeled with ot-carboxydimethoxy-C2CQRD-IAby this pro-

cedure, and imaging with a cooled CCD camera, we obtain a substantial
improvement in signal and signal/noise compared to our earlier caged
resorufin work. An example using this probe to analyze the dynamics of
actin filaments in Listeria tails in a Xenopus egg extract is shown in Fig. 5.
Instrumentation
Fluorescence photoactivation imaging at the level of single cells or
embryos requires essentially the same apparatus as conventional low-light
fluorescence imaging, except that one more light beam must be brought to
bear on the specimen for photoactivation. For collection of fluorescence
images at low light levels we have used both intensifying cameras and a
cooled CCD. In most experiments we collect pairs of images, one showing
the uncaged fluorescence signal and the second a phase-contrast image of
the cell as a reference image. In some experiments a second fluorescence
channel is also collected.3 One advantage of the cooled CCD camera is its
ability to collect high-resolution images at both high and low light levels
due to its large dynamic range (e.g., Fig. 5).
To generate the photoactivation beam we use the 366-nm line from a
100-W mercury arc lamp. Others have used the 334-364 lines from an
FIG. 5. Listeria tail in Xenopus extract marked by photoactivating caged rhodamine actin.
Rabbit muscle aetiu was conjugated with c~-carboxy-DM-C2QRd-lA as described in the text.
This probe was mixed with X e n o p u s egg extract and killed L&teria as described in Theriot
et al) 7 (A) Image in the rhodamine channel before photoactivation, and (B) an image a few
seconds after activation with a bar of light oriented vertical in the image. (C) Corresponding
phase-contrast image. Images were collected with a cooled CCD camera. Note the improve-
ment in image resolution compared to a similar experiment reported in Theriot et al. TM due
to the improved probe and use of a CCD camera in place of an ISIT.
argon ion laser. 12 In principle, small, intense areas of illumination can be

generated from either source. Laser sources may be easier to focus and
their intensity is unlimited. Mercury arc lamps are less expensive, but they
have a maximal theoretical intensity at the specimen plane, related to the
brightness and size of the arc.
If a mercury source is used to generate the photoactivation beam, a
suitable slit or pinhole must be placed in a plane conjugate with the specimen
to delimit the illuminated area in the specimen. The simplest photoactiva-
tion apparatus simply uses the field diaphragm in the epifluorescence light
path as this pinhole. This diaphragm is closed to an appropriate size, a UV
excitation filter set is put in the light path (e.g., a DAPI set), and the
epifluorescence beam is turned on briefly. Then the diaphragm is opened,
a fluorescein or rhodamine filter set is inserted, and observation of the
activated zone commences. This setup has been used successfully to follow
microtubule flux in newt spindles, with the addition of a slit that can be
slid temporarily in place of the field diaphragm during photoactivation. 13
In general, we prefer an apparatus where the photoactivation and obser-
vation beams can be controlled independently by electronic shutters, facili-
tating reliable positioning of the slit/pinhole and rapid collection of the
first images after photoactivation. The two light paths we use are shown
in Fig. 6. Either can use a laser in place of the mercury lamp for activation.
The choice of light path is a practical consideration, governed by how easy
it is to physically access the field diaphragm plane in the epi-illumination
tube for a given microscope.
Future Directions
Despite its potential for visualizing dynamic processes in living cells
and embryos, fluorescence photoactivation to date has been used in only
a handful of laboratories for a limited range of proteins. The main limitation
has been probe chemistry and probe availability, as discussed earlier. Ex-
panding the application of the method will hinge primilarly on improving
the availability of the current probes and developing better ones.
The most obvious avenue for probe improvement is to come up with
better conventional caged fluorochromes. Rhodamines are the brightest
and most photostable of the current probes. It should be possible to improve
on the synthesis methods and perhaps find caging groups that photoactivate
as efficiently as the ot-carboxy series in UV light but do not photoactivate
in visible light. Other fluorescent molecules and other types of caging
12 W. I. Rodionov, S. S. Lim, V. I. Gelfand, and G. G. Borisy, J. Cell BioL 126, 1455 (1994).
13 T. J. Mitchison and E. D. Salmon, J. Cell BioL 119, 569 (1992).
A Hg ARC MICROSCOPE BODY

--SHUTTER I ~-FIELD DIAPHRAGM . I-
F541-551nm I /SPECIMEN PLANE"~<I_
| //°"~EC'"~ L.~S - % 1 ~ 1
. . . . . ,_
ii°sF°; o
"- II -
<400rim la~l"dl SLIT

II
II
~ \/
340-380nm
SHUTTER
CONDENSER LENS
Hg ARC
I ~ REFLECT
<400nm
II . . . . - ........ _T/
am ,';,, ~
I I
f I
~FOCUSING
/ LENS
i
m
<tz~
\/
FIG. 6. Schematic diagram of photoactivation light paths. The photoactivation beam is

shown by dashed lines, and the fluorescence excitation beam by solid lines. The wavelength
values on the excitation and emission filters are appropriate for Q-rhodamine. The schematic
depicts an inverted microscope, but the same light paths can be generated on an upright
stand. For clarity, components in common to the two light paths are labeled only in (a) and
the fluorescence emission light path (which travels through the emission filter) is omitted. (a)
In this setup the photoactivation beam is brought into the epifluorescence light path on the
objective lens side of the field diaphragm using a dichroic mirror inserted into the epifluores-
cence light path. The slit is positioned in a plane conjugate to the field diaphragm and the
I4] CAGED FLUORESCENT PROBES 77
chemistry should also be considered. It might be possible to make a caged

rhodamine by quaternizing one of the nitrogens as opposed to making a
carbamate, and such a probe would have higher water solubility. Rhodols,
molecules intermediate in structure between fluorescein and rhodamines, 5
are interesting candidates for caging. In principle, a rhodol caged with a
single nitrobenzyl ether on the oxygen might be a useful probe with lower
molecular weight than the current caged fluoresceins and rhodamines. Cya-
nine dyes have fluorescence properties superior even to rhodamines as
protein-attached probes. It is not obvious how to cage them using nitroben-
zyl chemistry, but some other strategy might be feasible. In general, caged
fluorochromes are areas where creative chemistry could generate signifi-
cant payoffs.
In conventional fluorescent imaging, green fluorescent protein (GFP)
fusion proteins have risen to prominence because the fluorescent tag can
be encoded in the cDNA for the protein. Particularly in usefully mutated
forms, TM GFP appears to be an excellent fluorochrome, with good adsorption
coefficients, quantum efficiency, and photostability. Thus it is natural to
ask if GFP can be used in any kind of photoactivation experiment. Wild-
type GFP is in fact already a caged fluorochrome of sorts. The protein
can adopt two conformations with different adsorption maxima, and light
adsorption can shift the population of molecules toward the longer wave-
length absorbing f o r m . 14 By mutating GFP and selecting appropriate vari-
ants, it might be possible to turn it into a very good caged fluorochrome
that could replace the small organic probes. An alternative, and perhaps
more likely, possibility is that GFP will bring new vigor to the older ap-
proach of photobleaching. Since the organic fluorophore is buried within
the core of GFP, it is possible that it is less efficient at generating activated
oxygen species than conventional organic fluorochromes, and also that
when activated oxygens are produced they will react preferentially with
GFP itself, rather than nearby proteins. In other words, photobleaching of
GFP may be less prone to artifacts due to collateral damage to nearby
proteins than is photobleaching with conventional probes. The first GFP
14A. B. Cubitt, R. Heim, S. R. Adams, A. E. Boyd, L. A. Gross, and R. Y. Tsien, Trends

Biochem. Sci. 20, 448 (1995).
specimen, which means it is physically close to the microscope body. This light path has been
set up on Zeiss IM35, Zeiss Universal, and Olympus inverted stands. (b) In this setup the
photoactivation beam is brought in with a dichroic mirror mounted between the epifluores-
cence light source and the back of the microscope. It requires no modification of the microscope
body, but necessitates an additional lens to focus the image of the slit onto the field diaphragm
and specimen plane. This light path has been set up on a Zeiss Axiovert stand.
bleaching experiments are quite encouraging in this light. 15 However, it
will be important to test more rigorously the extent of collateral damage
in GFP bleaching experiments before adopting this technology more gen-
erally.
15 N. B. Cole, C. L. Smith, N. Sciaky, M. Terasaki, M. Edidin, and J. Lippincott-Schwartz,

Science 273, 797 (1996).
16 L. J. Mathias, Synthesis 1979, 561 (1979).
17 j. A. Theriot, J. Rosenblatt, D. A. Portnoy, P. J. Goldschmidt-Clermont, and T. J. Mitchison,
Cell 76, 505 (1994).
18 j. A. Theriot, T. J. Mitchison, L. G. Tilney, and D. A. Portnoy, Nature 357, 257 (1992).
[5] Biologically Active Peptides Caged on Tyrosine

By R. SREEKUMAR, M I T S U O IKEBE, F R E D R I C S. F A Y , *
a n d JEFFERY W , W A L K E R
In~oducUon
Synthetic peptides show considerable promise as selective inhibitors of
protein-protein interactions involved in many basic biological processes,
including signaling, secretion, and motility. 1 Light-activated caged peptides
would be powerful tools 2 for controlling the release of peptides into cells
in order to overcome diffusional delays, circumvent microinjection artifacts,
and target selected proteins in a time domain that is inaccessible by antisense
or gene knockout techniques. In addition, caged peptides would provide a
means to spatially map out selected protein activities within living cells.
Peptide probes are usually found by identifying domains involved in
critical protein-protein interactions, then narrowing the activity of the
domain to a fragment of approximately 5-30 residues by mutagenesis or
by screening overlapping peptide sequences. Alternatively, useful peptides
are identified by a combinatorial approach involving selection of active
peptide sequences out of random libraries. 3 The search process often de-
pends on having a reliable in vitro assay of peptide activity that can be
* This article is dedicated to the memory of collaborator and friend Fredric S. Fay (deceased),
who was a constant source of energy, enthusiasm, and vision.
1 D. C. Hancock, N. J. O'Reilly, and G. I. Evan, Mol. Biotech. 4, 73 (1995).
3 p. C. Andrews, D. M. Leonard, W. L. Cody, and T. K. Sawyer, Methods in Mol. Biol. 36,
305 (1994).
Copyright © 1998by AcademicPress

All rightsof reproductionin any form reserved.
METHODS IN ENZYMOLOGY,VOL. 291 0076-6879/98 $25.00
bleaching experiments are quite encouraging in this light. 15 However, it
will be important to test more rigorously the extent of collateral damage
in GFP bleaching experiments before adopting this technology more gen-
erally.
15 N. B. Cole, C. L. Smith, N. Sciaky, M. Terasaki, M. Edidin, and J. Lippincott-Schwartz,

Science 273, 797 (1996).
16 L. J. Mathias, Synthesis 1979, 561 (1979).
Cell 76, 505 (1994).
18 j. A. Theriot, T. J. Mitchison, L. G. Tilney, and D. A. Portnoy, Nature 357, 257 (1992).
[5] Biologically Active Peptides Caged on Tyrosine

By R. SREEKUMAR, M I T S U O IKEBE, F R E D R I C S. F A Y , *
a n d JEFFERY W , W A L K E R
In~oducUon
Synthetic peptides show considerable promise as selective inhibitors of
protein-protein interactions involved in many basic biological processes,
including signaling, secretion, and motility. 1 Light-activated caged peptides
would be powerful tools 2 for controlling the release of peptides into cells
in order to overcome diffusional delays, circumvent microinjection artifacts,
and target selected proteins in a time domain that is inaccessible by antisense
or gene knockout techniques. In addition, caged peptides would provide a
means to spatially map out selected protein activities within living cells.
Peptide probes are usually found by identifying domains involved in
critical protein-protein interactions, then narrowing the activity of the
domain to a fragment of approximately 5-30 residues by mutagenesis or
by screening overlapping peptide sequences. Alternatively, useful peptides
are identified by a combinatorial approach involving selection of active
peptide sequences out of random libraries. 3 The search process often de-
pends on having a reliable in vitro assay of peptide activity that can be
* This article is dedicated to the memory of collaborator and friend Fredric S. Fay (deceased),
who was a constant source of energy, enthusiasm, and vision.
1 D. C. Hancock, N. J. O'Reilly, and G. I. Evan, Mol. Biotech. 4, 73 (1995).
3 p. C. Andrews, D. M. Leonard, W. L. Cody, and T. K. Sawyer, Methods in Mol. Biol. 36,
305 (1994).

[5] ACTIVE PEPTIDES CAGED ON TYROSINE 79
used to screen large numbers of candidate peptides rapidly. Once a peptide

is identified, development of a caged peptide requires structure-activity
information. Various amino acid substitutions need to be evaluated in
model peptides to identify positions in the sequence where certain features
(e.g., charge, hydrophobicity) are disruptive for target recognition. Alterna-
tively, N or C termini can be modified with tails that interfere with recogni-
tion, or a peptide can be conformationally caged by end-to-end or side
chain cyclization to render it inactive. The modifying group(s) must be
photosensitive so that the peptide can be returned to its original active
state on illumination. Each caged peptide must be thoroughly characterized
in terms of structure, photochemistry, and biological properties before it
can be used with confidence to address specific questions in cellular systems.
An ideal peptide probe should display a good selectivity and high (submicro-
molar) affinity for its target. In the caged form, the peptide should have
little or no activity or a greatly reduced affinity and should photolyze
efficiently and rapidly to its active form.
The purpose of this article is to describe some practical aspects in the
development of caged peptides for use in cell biology. We show here that
the relatively straightforward approach of derivatizing individual amino
acid side chains on peptides with photolabile groups can lead to useful
caged peptides. A general strategy is illustrated for preparing such caged
peptides using the tyrosine side chain as an example. Tyrosine was chosen
because it is ultraviolet (UV) active and fluorescent, properties that facili-
tate characterization, and it is a critical side chain in a number of proteins.
Successful caging strategies will be described for two model peptides:
RS-20, a 20 amino acid calmodulin inhibitory peptide, and LSM1, a 13
amino acid protein kinase autoinhibitory domain peptide (see Scheme 1
for peptide sequences).
Syntheses
Peptides typically have a variety of reactive functional groups, including
a-amines, e-amines, alcohols, carboxylic acids, and sulfhydryls. Selectively
directing caging moieties to one site on an unprotected peptide is difficult
and often results in mixtures of peptides modified at multiple sites. To avoid
this problem, we first synthesize the caged amino acid, then incorporate it
into the peptide sequence at the desired position during automated peptide
synthesis. This has many advantages: (i) final peptide structures are well
defined, (ii) tens of milligrams of peptides are readily prepared, and
(iii) more flexibility is available for optimizing peptide properties because
a wider variety of amino acid side chains and caging moieties can be consid-
80 CAGED COMPOUNDS 151
A
RS-20: NH3+-ARRKWQKTGHAVRAIGRLSS-CO2
+
LSM-I: NH3-LSKDRMKKYMARR-CO~
B
NO 2 COO- NO COO-
cH~
I II
o
hv +
6 CH20
HN-CH--C,~ ~ ,.~ H N - C H - - C , . ~
cgY y
SCHEM~1. Structures of peptides containing caged tyrosine. (A) Sequence of the calmodulin
inhibitor, RS-20, and of the MLCK autoinhibitory peptide, LSM1. Arrows mark the location
of caged tyrosines in caged forms of these peptides. (B) Structure and photocleavage reaction
for caged tyrosine.
ered. The principal drawback is that chemical synthesis is limited to peptides

of up to about 50 amino acid residues in length.
Synthesis of Caged Tyrosine

Step 1: Reactive Caging Moiety: 2-Bromo-a-carboxyl-2-nitrophenyl
Methyl Ester (I). The synthetic route for caged tyrosine is illustrated in
Scheme 2. The a-carboxyl caging moiety is found to be ideal for phenols, 4
so the preparation of this version of caged tyrosine (cgY 5) is described in
detail. Coupling is carried out with a bromomethyl ester of this moiety,
which is prepared in two steps as follows.
To a stirred solution of 1 g 2'-nitrophenylacetic acid (5.5 mmol) in
10 ml benzene is added 3.5 ml anhydrous methanol followed by three drops
of concentrated sulfuric acid. The reaction mixture is refluxed for 8 hr
with an azeotropic distillation device. Thin-layer chromatography (TLC)
in dichloromethane shows quantitative conversion of 2'-nitrophenylacetic
4j. W. Walker, H. Martin, F. Schmitt, and R. J. Barsotti, Biochemistry 32, 1338 (1993).
5 cgY, O-(a-Carboxyl-2-nitrobenzyl)tyrosine incorporated into a peptide (see Scheme 1 for
structure). Amino acids are identified by standard single-letter code: A, Ala; C, Cys; D,
Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; M, Met; N, Asn, P, Pro; Q, Gin; R, Arg;
S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr.
-%
NO2 COOCH3 NO2 COOCH3 .NO2 COOCH3
~ B r t-butoxide
I + OH
CH2 0 CH2 0
I II I II
tBOC-HN--CH--C-OtBu H2N--CH--C-OH
CH2 0 m IV
I II
tBOC-HN--CH--C-OtBu
II 10%K2CO3
aqueous
NO2 COOCH3 NO2 .COOH
CH2 0 Ctt2 0
I II
Fmoc-HN--CH~C-OH H2N--CH--C-OH
vl
SCHEME2. Synthetic route for caged tyrosine.
acid (Rf = 0.01) to the corresponding methyl ester (RI = 0.28). Purification
by silica gel flash chromatography in hexane/ethyl acetate yields 0.95 g of
product as a white solid (95% yield). Of this product, 2'-nitrophenyl methyl
acetate, 6.2 g (31.8 mmol), is then dissolved in 100 ml carbon tetrachloride
in a 500-ml round-bottom flask. While stirring with a magnetic stir bar,
0.31 g benzoyl peroxide (1.28 mmol) is added, followed by 6.22 g N-bromo-
succinimide (34.9 mmol). The flask is fitted with a reflux condenser and
the reaction mixture is refluxed for 24 hr. Thin-layer chromatography in
dichloromethane shows a new spot I (R I = 0.43) and some starting material
(R I = 0.28). Extending the reaction time does not improve the extent of
formation of I, but modest improvements are afforded by another 24 hr
of reflux after filtering the reaction through Whatman (Clifton, N J) #1
paper and recharging with fresh benzoyl peroxide and N-bromosuccinimide.
In preparation for purification, the reaction mixture is filtered through
Whatman #1 paper and rotary evaporated to near dryness. Approximately
3-ml aliquots of the crude mixture are chromatographed on a 2-in. (inside
diameter) flash column containing 400 g silica gel in hexane. Compound I
is eluted in ethyl acetate/hexane (1:18, v/v) with an overall yield of
3.08 g (50%).
Step 2: Protected Tyrosine: a-N-BOC-tyrosine tert-Butyl Ester (II). To
selectively derivatize tyrosine on its phenolic side chain, the ot-carboxyl and
or-amino groups need to be protected. This is accomplished with a tert-butyl
ester and a tert-butyloxocarbonyl (BOC) moiety, respectively, tert-Butylty-
rosine is commercially available (Bachem, Torrance, CA) so it is typically
used as starting material. The amino group is protected with the commercial
reagent BOC-ON (Sigma, St. Louis, MO). This reaction proceeds in high
yield, and the product, II, is easily purified by flash chromatography.
To 1.78 g tyrosine tert-butyl ester (7.5 mmol) dissolved in 30 ml dioxane/
water (5:1, v/v) is added 1.11 ml triethylamine (TEA, 8 mmol) and then
2.0 g BOC-ON (8.2 mmol). The reaction mixture is stirred overnight at room
temperature under a nitrogen atmosphere. The progress of the reaction is
monitored by TLC on silica gel plates containing a fluorescent indicator that
localizes UV active compounds as dark spots. Thin-layer chromatography in
ethyl acetate/hexane (1 : 1, v/v) shows three spots: tyrosine-tert-butyl ester
(Rs = 0.05), II (R s = 0.5), and BOC-ON (Ry = 0.65). The reaction solvent
is removed, and the crude product is purified in one batch by flash chroma-
tography using a 2-in. (inside diameter) glass column containing 300 g silica
gel in petroleum ether. The column is developed with 400 ml petroleum
ether/ether (6: 1, v/v), then 400 ml petroleum ether/ether (4 : 1, v/v), and
finally II elutes from the column in 300 ml petroleum ether/ether
(3:2, v/v) in a yield of 1.74 g (97%). Protected tyrosine is now ready for
coupling to the caging moiety.
Step 3: Coupling. An equal molar amount of tert-butoxide is mixed
with II to enhance the reactivity of the phenolic group. Then the reactive
photolabile group is added in the form of a bromide, and the reaction
proceeds to >70% yield. To a 100-ml round-bottom flask under a nitrogen
atmosphere is added 2.5 g protected tyrosine, I! (7.4 mmol) in 10 ml tetrahy-
drofurant (THF). Then 7.6 ml 1 M potassium tert-butoxide in THF is added
dropwise while stirring with a magnetic stir bar. The reaction mixture
immediately turns dark purple, then brown. Finally, 2.48 g 2-bromo-2'-
nitrophenyl methyl acetate, I (9 mmol), is added and left to stir overnight
at room temperature. The yellow-brown suspension is then centrifuged,
and TLC of the supernatant solution in petroleum ether/ether (2:1, v/v)
shows the product, III (Rf = 0.12), and traces of starting materials, 11
(RI = 0.01) and I (RI = 0.31). Reaction solvent is removed by rotary
evaporation, and the entire reaction mixture is chromatographed using the
2-in. (inside diameter) flash column with 400 g silica gel in hexane. The
column is washed with 500 ml dichloromethane/hexane (2 : 1, v/v), and llI
is eluted with 200 ml dichloromethane with an overall yield of 1.8 g (72%).
Step 4: Deprotection. Removal of a-amino and a-carboxyl protecting

groups is accomplished by treatment in neat trifluoroacetic acid (TFA) at
room temperature. Lower levels of TFA do not remove the tert-butyl ester.
The methyl ester on the caging moiety is stable in the presence of neat
TFA, and for most applications this methyl ester is maintained until the
final step of caged peptide preparation. The progress of deprotection of
a-amines and a-carboxyls is readily monitored by analytical reversed-phase
high-performance liquid chromatography (HPLC) using a Whatman
ODS-3 column (1 x 25 cm) and a solvent composed of acetonitrile/water/
TFA (40: 60: 0.1, v/v/v). Using a flow rate of 1.5 ml/min, III elutes at 7
min and fully deprotected caged tyrosine methyl ester, IV, elutes at 4 min.
Deprotection in neat TFA at room temperature is complete in 4 hr. Caged
tyrosine methyl ester, IV, is purified in approximately 90% yield by semi-
preparative reversed-phase HPLC using a 2.5 X 25-cm ODS-3 column and
the same solvent system at 3 ml/min.
In some instances, particularly for evaluation of the properties of caged
tyrosine itself or for the use of caged tyrosine as a control compound in
biological experiments, the methyl ester, IV, is hydrolyzed to the free acid,
VI. The methyl group is removed by treatment with aqueous 10% (w/v)
K2CO3 at room temperature for 1 hr. Caged tyrosine free acid, VI, is
purified by reversed-phase HPLC, rotary evaporated to dryness, and stored
as a solid at - 2 0 °.
Step 5: F-Moc Caged Tyrosine (I0. For caged peptide synthesis, N-(9-
fluorenyl)methoxycarbonyl(F-moc) chemistry is used for selective blocking
and deblocking of o~-amino groups during peptide elongation. Therefore,
caged tyrosine methyl ester is made ready for incorporation into peptides
by derivatizing its a-amino groups with an F-moc moiety using a modifica-
tion of the method of Ten Kortenaar et aL 6
Caged tyrosine methyl ester (IV, 1.04 g, 1.95 mmol) is suspended in 7
ml acetonitrile, 4 ml water, and 0.7 ml TEA; 0.65 g F-moc succinimide
(1.93 mmol) is added portionwise with stirring. The pH of the homogeneous
reaction mixture is initially approximately 8.5 but drops as the reaction
proceeds. Small aliquots of TEA are added to maintain the pH around 8
(using standard pH paper). Progress of the reaction is monitored by analyti-
cal reversed-phase HPLC in acetonitrile/water/TFA (60: 40: 0.1, v/v/v),
and is complete within approximately 30 min. The reaction mixture is
washed with 0.1 M HCI (3 x 25 ml), and the organic phase is separated and
dried by rotary evaporation. Compound V is purified by semipreparative
reversed-phase HPLC in acetonitrile/water/TFA 45 : 55 : 0.1 (v/v/v) at
6 p. B. W. T e n Kortenaar, B. G. V a n Duk, J. M. Peeters, R. J. Raaben, P. J. H. M. A d a m s ,

and G. I. Tesser, Int. J. Peptide Protein Res. 27, 398 (1986).
84 CA~EO coMr'ouNos 151
3 ml/min. The yield of purified V is 0.58 g (56%). Compound V is stable
for at least 6 months when stored as a solid at -20 °.
Step 6: Synthesis of Peptides Containing Caged Tyrosine. Peptide synthe-
sis is carried out on an Applied Biosystems, Inc., Model 432A (Foster City,
CA) solid-phase synthesizer using F-moc chemistry and a Wang resin. 7
Activation of a-carboxyls for peptide bond formation is accomplished with
a fivefold excess of 2-(1H-benzotriazolyl)-l,l,3,3-tetramethuronium hexa-
fluorophosphate (HBTU), and selective removal of the F-moc group is
carried out in 10% piperidine in dimethylformamide (DMF). Side chain
protecting groups are tert-butyl for S, T, and Y; trityl for Q and N; BOC
for K; and 2,2,5,7,8-pentamethylchroman-6-sulfonyl (Pmc) for R. F-moc
tyrosine methyl ester, V (45 rag, 75 ~mol), is packed into a plastic amino
acid cartridge that is compatible with the synthesizer. The peptide sequence
is determined by the alignment of F-moc amino acid cartridges in the
carriage wheel, and F-moc tyrosine methyl ester is placed in its desired
position in the sequence. Double coupling is carried out during the step(s)
in which the caged amino acid is incorporated. After the peptide synthesis
is complete, the peptide column containing the derivatized resin is washed
with 20 ml dichloromethane and dried in a nitrogen stream for 15 min.
Cleavage is accomplished by transfer of the resin into a 10-ml polypropylene
tube followed by the addition of 1 ml of cleavage cocktail with stirring for
2-4 hr. The composition of the cocktail is TFA/thioanisole/ethanedithiol
(92.5 : 5 : 2.5, v/v/v). Crude peptide is precipitated from this reaction mixture
by filtration to remove the resin followed by dropwise addition of filtrate
to 10 volumes of ice-cold tert-butyl methyl ether. The precipitate is then
washed with 3 × 25 ml of the same solvent, dried in a vacuum desiccator for
15 min, suspended in 2 ml water, frozen, and lyophilized to a white powder.
Peptides are purified by analytical reversed-phase HPLC in acetonitrile/
water/TFA. To generate the free a-carboxyl moiety on the cage, purified
peptide is demethylated by treatment with 10% K 2 C O 3 , neutralized with
1 M HC1, and repurified by reversed-phase HPLC. Each peptide requires
different elution conditions for purification depending on composition and
sequence, but peptide demethylation nearly always results in a shift to an
earlier retention time (e.g., see Fig. 6). Interestingly, in the case of 9cgY-
LSM1 (LSM1 with cgY in position 9) the demethylated caged peptide
resolves into two HPLC peaks, which appear to be stereois0mers (see
below). Typically, the first indications that the caged peptide synthesis has
been successful include (i) the peptide contains a UV absorbaoce peak
near 270 nm (Fig. 1) and (ii) the peptide photolyzes smoothly to a new
HPLC peak (Fig. 2). Final confirmation of peptide structures relies on
7 S. S. Wang, J. Am, Chem. Soc. 95, 1328 (1973).

T
CJ 10
T I t '27i nrn' ' u ....
~ 8
0
~ 6
N 4
0
rj
~ 2
.~ 0
~ LSMI
i i i ~ i i I i i i i
250 300 350
Wavelength/rim
FIG. 1. U V / V I S absorbance spectra for equal concentrations of LSM1 and 9cgY-LSM1
in a q u e o u s 20 m M Tris, p H 7.5.
amino acid analysis and mass spectrometry before and after photolysis.
Peptides are double lyophilized to remove traces of TFA and are stored
frozen in powder form prior to biological experiments.
Spectral and Photochemical Properties
Extinction Coefficient
The caging moiety and the tyrosine side chain provide useful UV-active
chromophores for estimating peptide concentrations. To determine the
extinction coefficient for each chromophore, UV/visible scans are taken of
the peptides in aqueous solution, then aliquots are quantified by amino
acid analysis. Peptide tyrosines have a UV maximum at 275 nm and a ratio
of absorbance to moles of peptide (normalized to alanine) equal to 103
(Fig. 1). Thus, the extinction coefficient for a peptide containing only tyro-
sine chromophores (i.e., no F or W) is 8 2 7 5 n m = 1000 M -1 c m - 1 per tyro-
sine. By a similar analysis, peptides containing a-carboxyl caged tyrosine,
cgY, display a UV maximum at 270 nm and an extinction coefficient of
/3270 n m = 8 0 0 0 M -1 c m -1 per caged tyrosine (Fig. 1). Errors in these values
are estimated to be _+12%.
Quantum Yield
A convenient method for measuring photolytic quantum yield involves
admixture of the compound to be tested with a reference compound whose
A28o
A 1 min
B
5Y-RS-20 5cgY-RS-20
,.,,,. ,,- pg~'~'Tf
Injection
FIG. 2. Analytical reversed-phase HPLC traces of 5cgY-RS-20 before (A) and after (B)
3 min phot01ysis. (C) Synthetic 5Y-RS-20 injected as a marker.
quantum yield is known. 8 This is valid strictly only if the two compounds
have equivalent absorption properties in the wavelength range used for
irradiation. Therefore, it is important to mix the compounds at equal con-
centrations and to ensure that the compounds possess the same photolabile
8 j. W. Walker, G. P. Reid, and D. R. Trentharn, Methods Enzymol. 172, 288 (1989).

protecting group. An HPLC method is then needed that resolves the two
compounds from each other and from photolysis by-products.
Caged tyrosine, VI, is mixed with caged phenylephrine each at 30/zM
in 50 mM Tris, pH 7.5, 1 mM dithiothreitol (DTT). One milliliter of this
mixture is placed in a quartz cuvette (1 cm path length) and irradiated with
a xenon arc lamp filtered with a U G l l filter to pass near-UV light (300-
400 nm). The solution is occasionally mixed to promote uniform illumina-
tion, and at l-rain intervals 100-/.d aliquots are removed and injected onto
the analytical HPLC. The amount of material in each peak is quantified
and plotted as a function of irradiation time on a semilogarithmic plot (Fig.
3). The slope of this plot relative to the reference compound is a measure
of the quantum yield. To check for competing side reactions during photoly-
sis it is necessary to carry out a direct measure of the amount of product
formed. For caged tyrosine, VI, and the two peptides described here, a
stoichiometric amount of tyrosine or tyrosine peptides is formed, indicating
that there are no alternative routes for breakdown of the caged compounds
and that the values reported in Fig. 3 are equivalent to the product quan-
tum yield.
Photolysis Kinetics
The rate of photolysis is initially characterized for the caged amino acid
rather than the caged peptide because larger quantities are available. Two
100" , , , , . , ,
• caged
Q)
N 9 g LS 1
= r
0 2 4 6 8
Irradiation Time/min
Fro. 3. Steady-steady quantum yields (~) determined by mixing caged tyrosine, VI, or
caged peptide, 9cgY-LSM1, with an equal concentration of reference compound caged phenyl-
ephrine ( ~ = 0.11). 4 Mixtures were irradiated with a xenon arc lamp (300-400 nrn), and
aliquots were chromatographed by HPLC as in Fig. 2 to determine the amount remaining.
q~ = 0.31 for VI; q~ = 0.32 for 9cgY-LSM1.
milliliters of 1 mM caged tyrosine, VI, in 25 mM N-[2-hydroxyethyl]pipera-

zine-N'-[2-ethanesulfonic acid] (HEPES), pH 7.0, is placed in a 1 × 1-cm
quartz cuvette with all four walls polished. The sample is photolyzed with
a xenon flash lamp (Hi Tech Scientific, Salisbury, UK), and then absorbance
or fluorecence transients are monitored at 90° to the photolysis beam. The
custom-built transient spectrophotometer has been described previously. 9
A similar absorbance transient is observed for both compound ¥ I and
the caged tyrosine peptide 9cgY-LSM1. The absorbance transient decays
biexponentially with a poorly resolved process occurring with a rate contant
of about 600 sec -1 followed by a slower decay at 50 sec-L Further character-
ization involves varying the detecting wavelength with a monochromator,
which reveals a peak absorbance at 435 nm consistent with the existence
of an aci-nitro anion intermediate in the photolytic reaction. 4 This decay
process has been equated with the rate of formation of the biologically
active species in at least one case, 9 but the mechanism of photolysis is
poorly understood for most 2'-nitrobenzyl compounds. For instance, the
significance of the two distinct decay processes is unknown. Therefore, it
is useful to have additional data concerning the rate of appearance of the
photoproduct .of interest.
Caged tyrosine is of interest in this regard because tyrosine is an intrinsi-
cally fluorescent compound, and addition of the caging moiety greatly
reduces its fluorescence (Fig. 4A). Flash photolysis with fluorescence detec-
tion allows the rate of formation of tyrosine to be monitored. Fluorescence
transients following flash photolysis reveal a step increase in tyrosine fluo-
rescence (excitation at 280 nm, emission at 300 nm) for both caged tyrosine,
VI (Fig. 4B), and the caged peptide 9cgY-LSM1. The time resolution of
these measurements is approximately 10 msec, indicating that the rate of
formation of tyrosine is greater than 100 sec -1 and too fast to measure
with our current instrumentation. Such fast photolysis rates will no doubt
enhance the utility of caged tyrosine and caged tyrosine-containing pep-
tides in kinetic studies. For example, photorelease of a tyrosyl side chain
in a disordered region of a peptide or protein could be used to initiate
protein folding. The large fluorescence increase could then be used as an
environmentally sensitive signal to monitor the folding process.
Deciding on Caging Strategy: Structure-Activity Analysis

To prepare an inactive version of a peptide, one must be armed with
knowledge of which features of the peptide are important for biological
9 j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. ll0,

7170 (1988).
280 nm
ca
300 nm
~ 100
~ ~ 5o
0
250 .500 ,.550
Wavelength/nm
O)
¢/1
+
Flash
FIG. 4. (A) Fluorescence spectra of 0.1 mM tyrosine (Y, plain lines) or 0.1 mM caged
tyrosine (compound VI, bold lines). Excitation spectra are illustrated by broken lines and
emission spectra by solid lines. (B) Fluorescence increase with a sampling rate of 500 sec -1
following pulse photolysis of 1 m M compound VI in 25 mM HEPES, pH 7.0 at room temper-
ature.
recognition and activity. Two model peptides are described (Scheme 1). One
is the widely studied calmodulin-binding peptide, RS-20, whose sequence is
derived from the calmodulin-binding domain of smooth muscle myosin
light-chain kinase (MLCK). 1° A wealth of structural information exists
10T. J. Lukas, W. H. Burgess, F. G. Prendergast, W. Lau, and D. M. Watterson, Biochemistry

25, 1458 (1986).
about how this peptide interacts with its target protein, calmodulin. 11'12
Critical hydrophobic residues have been identified in positions 5 and 18 of
this peptide. Using the two assays described below, we found that substitu-
tion of W5 or L18 with glutamate effectively eliminated its interaction with
calmodulin. In contrast, substitution into either position with lysine only
modestly reduced its binding affinity toward calmodulin. Therefore, because
negative charge is more disruptive than positive charge, we synthesized
peptides containing negatively charged caged tyrosine in these positions to
create caged RS-20 peptides. As described below, only the peptide con-
taining caged tyrosine in position 5 displays large differences in calmodulin
binding before and after photolysis. Importantly, the peptide with both W5
and L18 substituted with caged tyrosine is inactive even after photolysis.
This peptide is useful as an inactive caged peptide to control for the effects
of UV exposure, generation of photolytic by-products, and nonspecific
effects of peptides during physiological experiments. Peptides with the
same composition but with a scrambled sequence can also be used for
this purpose.
The second peptide of interest is an autoinhibitory domain peptide,
which is also from smooth muscle MLCK. This 13 amino acid peptide is
thought to have a critical tyrosine in position 9 based on mutagenesis studies
of the whole enzyme.13Substitution of Y9 with glutamate eliminated MLCK
inhibitory activity and indicated that the simple replacement with the nega-
tively charged caged tyrosine would generate a caged LSM1 peptide.
Caged RS-20: Calmodulin Binding

The high-affinity calmodulin-binding peptide RS-20 is caged by replac-
ing W5 with caged tyrosine. Caged forms of this peptide chromatographed
as a single peak in reversed-phase HPLC and photolyzed to 5Y-RS-20
(RS-20 with Y in position 5) as expected (Fig. 2). A rapid test of the
calmodulin-binding activity of these peptides is performed on a calmodulin
affinity column. Five milliliters of calmodulin agarose (Sigma) is poured
into a 1 x 10-cm glass column and equilibrated with 50 mM Tris, pH 7.5,
1 mM CaEGTA, 32/zM Ca 2+ at 4°; 100 nmol 5cgY-RS-20 (RS-20 with cgY
in position 5) is dissolved in 1 ml of eluant and is passed over the column
while 14 x 1.5-ml fractions are collected. Analysis of each fraction by UV
absorbance at 270-nm reveals a peak in the column void volume (Fig. 5),
aa M. Ikura, G. M. Clore, A. M. Gronenborn G. Zhu, C. B. Klee, and A. Bax, Science 256,

632 (1992).
12 A. Crivici and M. Ikura, Annu. Rev. Biophys. Biomol. Struct. 24, 85 (1995).
13 M. Tanaka, R. Ikebe, M. Matsuura, and M. Ikebe, E M B O J. 14, 2839 (1995).
[5] ACTIVE PEPTIDES CAGED ONTYROSINE 91
80
5egY-RS-20
60 o Before p h o t o l y s i s
r photolysis
40
0
0 5 10 15 20 25 30
Fraction Number
FIG. 5. Calmodulin binding determined by affinity chromatography on calmodulin-agarose.
5cgY-RS-20 did not bind to calmodulin-agarose in 32/zM free Ca 2+, 50 mM Tris, pH 7.5, 4 °
(O). After photolysis, the peptide bound tightly under the same conditions and was eluted
with 1 mM EGTA, 50 mM Tris, pH 7.5 (Q).
which on reversed-phase HPLC analysis is found to comigrate with 5cgY-

RS-20.
An identical aliquot of 100 nmol of 5cgY-RS-20 in eluant is then irradi-
ated for 10 min with the 300-400-nm filtered output of a 75-W xenon lamp.
This sample is applied to the calmodulin-agarose column, and fractions are
collected and analyzed as before. No peaks are observed until the eluant
is changed to 50 mM Tris, pH 7.5, 1 mM ethylene glycol N,N,N',N'-
tetraacetic acid (EGTA), no Ca 2+, at which time a peak appears (Fig. 5)
that comigrates with 5Y-RS-20 on reversed-phase HPLC. Data demonstrate
that before photolysis, 5cgY-RS-20 elutes in the column void volume and
therefore does not bind appreciably to calmodulin. After photolysis, the
peptide photoproduct 5Y-RS-20 binds strongly to calmodulin in a Ca 2÷-
dependent manner (Fig. 5). Calmodulin affinity chromatography provides
a rapid means for screening large numbers of caged peptides directed
against calmodulin. The capacity of the column is approximately 500-1000
nmol of peptide, and overall recoveries of peptides are greater than 50%.
Caged RS-20: Calmodulin-Dependent Myosin Light-Chain

Kinase Activity
For a more quantitative analysis of the interaction between peptides
and calmodulin, a calmodulin-dependent MLCK assay is used. l°J4 In this
assay, the efficacy of peptide binding is assessed by examining its ability to
14M. Ikebe, M. Stepinska, B. E. Kemp, A. R. Means, and D. J. Hartshome, J. Biol. Chem.

262, 13828 (1987).
Injection
B
9cgY-LSMI
9cgY-LSMI
• r~ethltl e s t e r
Injection
FIG. 6. Reversed-phase HPLC traces showing the shift in retention time and appearance
of stereoisomers of 9cgY-LSM1 as a result of demethylation. (A) Purified 9cgY-LSM1.
(B) Purified 9cgY-LSM1 treated with 10% aqueous K2CO3 for 60 min at room temperature.
reduce calmodulin-dependent activation of MLCK over a range of peptide

concentrations. The midpoint of the peptide inhibition curve, K0.s, is taken
as a measure of relative peptide affinity for calmodulin.
The assay cocktail consists of 100 nM calmodulin, 0.5 /zg/ml smooth
muscle MLCK, 200/zg/ml myosin regulatory light chain, and a standard
buffer containing 30 mM Tris, pH 7.5, 100 mM KCI, 0.1 mM CaCI2, and

1 mM MgCI2. The phosphorylation reaction is started by adding 50/xM
[T-32p]ATP and is continued for 5 min at 25 °. Incorporation of 32p into
myosin light chain is measured by trapping the protein on a filter and then
counting the filter in a liquid scintillation counter. Caged peptides inhibit
this calmodulin-dependent activity only weakly with apparent K0.5 = 100
txM, whereas uncaged versions of the same peptide inhibit nearly 2 orders
of magnitude more potently with an apparent K0.5 near 1/xM.
Caged LSMI: Myosin Light-Chain Kinase Catalytic Domain Activity

The inhibitory activity of the MLCK autoinhibitory peptide LSM1 is
evaluated according to the inhibition of the kinase activity of the constitu-
tively active MLCK. This unregulated form of MLCK is prepared by trypsin
treatment of the whole enzyme as described} 4 The assay conditions are
the same as above except 1 mM EGTA is included and no CaZ+/calmodulin
is added. LSM1 inhibits MLCK catalytic activity with an apparent K0.5 of
about 1/xM under the conditions employed (Fig. 7).
The caged LSM1 peptide exhibited the unique property of resolving
into two components of roughly equal quantity on the HPLC (Fig. 6). None
of the caged forms of RS-20 behaved in this way, and even caged LSM1
chromatographed as a single HPLC peak prior to K2CO3 demethylation
(Fig. 6A). Each of these demethylated caged LSM1 peaks photolyzed to
LSM1 with a similar photosensitivity, and each component had the expected
mass for caged tyrosine LSM1. The two compounds appear to be stereoiso-
100 ~ ,~ • • 9cgY-LSMI
>~ V V V •
"~ 75
<
5o
o:~ ; ~'o ,oo

[PEPTIDE], ,~M
FIG. 7. In vitro enzyme assay of a caged peptide before and after photolysis. The two
stereoisomers of 9cgY-LSM1 (Fig. 6B) were resolved and tested separately for their ability
to inhibit the activity of MLCK catalytic domain. First isomer (V) and second isomer (T)
off HPLC column. Both isomers of 9cgY-LSM1 photolyzed to LSM1 (0).
94 CAGZO coMPouNos [51
mers containing the R and S configurations in the caging group. Titration

of the M L C K activity with different concentrations of peptides revealed
that both isomers of caged LSM1 were similarly weakly potent as inhibitors
of M L C K activity (Fig. 7). Thus, asymmetry in the caging moiety has no
obvious effect on the biological activity of this caged peptide. The peptide
photoproduct of each isomer of the caged peptide inhibited M L C K approxi-
mately 2 orders of magnitude more potently than did the caged forms
(Fig. 7).
Summary
We have demonstrated the feasibility of preparing caged peptides by
derivatizing a single amino acid side chain in peptides up to 20 amino acids
long. Two peptides are illustrated whose activities are reduced by nearly
2 orders of magnitude using this caging approach. The specific strategy
described here of derivatizing tyrosine side chains with a charged caging
moiety should be generally applicable in the preparation of caged peptides
that have a critical tyrosine residue (e.g., LSM1) or that have critical hy-
drophobic patches (e.g., RS-20). Other amino acid side chains are also
accessible via this caging strategy. Derivatives of threonine, serine, lysine,
cysteine, glutamate, aspartate, glutamine, and asparagine can be prepared
and site specifically inserted into peptides in an analogous manner. The
caged peptides synthesized and purified by the methods described here are
compatible with biological samples, including living cells, 15"16and have been
used to demonstrate the central importance of calmodulin, MLCK, and,
by inference, myosin II in ameboid locomotion in polarized eosinophil
cells. 16 Photoactivation of peptides within cells should provide a wealth of
new information in future investigations by allowing specific protein activi-
ties to be knocked out in an acute and spatially defined way.
Acknowledgments
The authors thank Dr. Robert Carraway for performing peptide synthesis during early
stages of this work and for providinghelpfuladviceon syntheticprocedures.We thank Dr. Gary
Case of the Universityof WisconsinBiotechnologyCenter for adviceon peptide synthesis.The
assistance of Dr. MasanoriTanaka with the MLCKassaysis gratefullyacknowledged.Research
described here was supported by NIH Grants P01 HL47053, K04 HL03119 and NSF Grant
MCB 922082 to J.W.W., and PO1 HLA7530to M.I.
15R. M. Drummond, R. Sreekumar, J. W. Walker, R. E. Carraway, M. Ikebe, and F. S. Fay,

Biophys. J. 70, A238 (1996).
16j. W. Walker, S. H. Gilbert, R. M. Drummond, M. Yamada, R. Sreekumar, R. E. Carraway,
M. Ikebe and F. S. Fay, Proc. Natl. Acad. Sci. U.S.A. 95, 1568 (1998).
[6] LIGHT-DIRECTED ACTIVATION OF PROTEIN ACTIVITY 95
[6] Light-Directed Activation of Protein Activity from

Caged Protein Conjugates
By G E R A R D M A R R I O T T , JOHANNES OTTL, M A N F R E D H E I D E C K E R ,
and DANIELA GABRIEL
Introduction
Light-directed generation of biomolecules from their caged precursors
has emerged as a powerful technique for rapidly manipulating the concen-
tration of enzyme substrates and protein ligands and for fluorescently mark-
ing specific proteins at defined locations in living cells. 1 By monitoring the
time dependence of the relaxation of the reaction under study from its light-
induced perturbed state to the new equilibrium state, kinetic parameters of
the reaction can be determined and used to gain an understanding of the
reaction mechanism. 2'3 One of the most attractive features of light-directed
activation of caged components is that it allows kinetic parameters to be
determined for specific protein-catalyzed reactions within complex biomo-
lecular environments: for example, time-resolved X-ray diffraction of crys-
tals composed of the Ras P21-caged G T P complex has been used to study
the dynamics of protein motions in the crystal during the hydrolysis of
G T P after its photoactivation from caged GTP; 4's A T P has been rapidly
generated from caged A T P in skinned muscle fibers to investigate the
kinetics of the actomyosin ATPase reaction and its coupling to force produc-
tion; 6,7 and calcium ions and inositol triphosphate have been generated
from their caged precursors in living cells to investigate the role of calcium
ion-based signaling during cell motility and after fertilization of X e n o p u s
oocytes. 8,9 Rapid mixing techniques are unsuitable in these molecular envi-
ronments because of incomplete mixing and potential damage to the prepa-
ration under study caused by the delivery and rapid flow of solutions.
1 S. Adams and R. Y. Tsien, Methods Cell Biol. 55, 755 (1993).

2 j. A. McCray and D. R. Trentham, Annu. Rev. Biophys. 18, 239 (1989).
3j. A. Theriot and T. J. Mitchison, Nature 352, 126-131 (1991).
4 I. Schlichting,S. C. Almo, G. Rapp, K. Wilson, K. Petratos, A. Lentfer, A. Wittinghofer,
W. Kabsch, E. F. Pai, G. Petsko, and R. S. Goody, Nature 345, 309 (1995).
5A. Sheidig,C. Burmester, and R. S. Goody, Methods Enzymol. 291, [14], 1998(this volume).
6y. E. Goldman, M. G. Hibberd, and D. R. Trentham, J. Physiol. 354, 605 (1984).
7j. A. Dantzig and H. Higuchi, and Y. E. Goldman, Methods Enzymol. 291, [18], 1998
(this volume).
8I. Parker and I. Ivorram, J. Physiol. 461, 133 (1993).
9F. S. Fay, S. H. Gilbert, and R. A. Brundage, Ciba Found. Symp. 188, 121 (1995).
Copyright © 1998 by AcademicPress

All rights of reproductionin any form reserved.
96 CAGEDCOMVOUNOS [61
Furthermore, most commercial fast-flow mixing instruments have mixing
dead times that exceed 1 msec, whereas light-directed activation of caged
compounds based on a-carboxy-2-nitrobenzyl derivatives can be used to
establish a concentration jump within a few microseconds.1°'11
Light-directed perturbations of protein activity are usually achieved by
activation of a caged enzyme substrate or caged protein ligand, 12,13although
a complementary approach has been developed in which near-ultraviolet
light is used to directly trigger protein activity from an inactive, caged
protein conjugate. 14-18Caged protein conjugates are prepared by chemically
modifying amino acid residues that are essential for protein activity using
photocleavable labeling reagents. These labeling groups may be removed
from the protein on irradiation with near-ultraviolet light with a concomi-
tant recovery of protein activity, often to the prelabeled level. Caged protein
conjugates should not be confused with caged fluorophore-labeled protein
conjugates, which are fully active proteins that harbor covalently bound,
nonfluorescent chromophores (caged fluorophores) that can be rendered
fluorescent after irradiation with near-ultraviolet light. 19
Our interest in developing reagents and labeling procedures to prepare
caged proteins has emerged as part of a research program aimed at under-
standing the molecular basis of cell motility. Microinjected caged proteins,
labeled with fluorescent or caged fluorescent dyes for imaging purposes,
may be used to directly manipulate the activity of G-actin and actin-binding
proteins at defined locations in motile cells, and, in combination with time-
resolved fluorescence image microscopy, this technique provides opportuni-
ties to investigate the kinetics of G-actin polymerization/sequesterization
and the role of specific actin-binding proteins in the regulation of actin
filament dynamics and cell motility.
We have developed a number of caged reagents and associated labeling
procedures that can be used to prepare caged conjugates of almost any
protein by targeting essential lysine or cysteine residuesJ 5,16,18 Amino group
directed caged reagents based on 6-nitroveratryloxycarbonyl chloride
10 K. R. Gee, R. Wieboldt, and G. P. Hess, J. Am. Chem. Soc. 116, 8366 (1994).
11 G. P. Hess and C. Grewer, Methods Enzymol. 291, [25], 1998 (this volume).
12y. E. Goldman, M. G. Hibberd, J. A. McCray, and D. R. Trentham, Nature 300, 701 (1982).
13 G. C. Ellis-Davies and J. H. Kaplan, Proc. Natl. Acad. Sci. U.S.A. 91, 187 (1994).
14p. D. Senter, M. J. Tansey, J. M. Lambert, and W. A. Bl~ittler, Photochem. Photobiol. 42,
231 (1985).
15 G. Marriott, H. Miyata, and K. Kinosita, Biochem. Int. 26, 9943 (1992).
16 G. Marriott, Biochemistry 33, 9092-9097 (1994).
17 C. Ghung-Yu, B. Niblack, B. Walker, and H. Bayley, Chem. Biol. 2, 391 (1995).
i8 G. Marriott and M. Heidecker, Biochemistry 35, 3170 (1996).
19 T. J. Mitchison, K. E. Sawin, J. A. Theriot, K. Gee, and A. Mallavarapu, Methods Enzymol.
291, [4], (1998) (this volume).
[61 LIGHT-DIRECTED
ACTIVATIONOF PROTEINACTIVITY 97
(NVOC-CI) have been used to prepare caged conjugates of G-actin 16 and

profilin,2° and these conjugates can be reactivated at will upon illumination
with near-ultraviolet light. A new thiol reactive caged reagent, 4,5-dimeth-
oxy-2-nitrobenzyl bromide (DMNBB), has been used to prepare a caged
heavy meromyosin (HMM) conjugate that is unable to couple the energy
derived from ATP hydrolysis to the production of molecular forces that
move actin filaments,TM and light-directed activation caged HMM has been
used to restore the coupling between these two processes. This property
has proven useful in studying specific aspects of the actin-activated, force-
generating ATPase reaction of myosin and its subfragments in solution, in
the in vitro motility assay, TM and within skinned muscle fibers (Koegler,
Ruegg, and Marriott, 1988, unpublished results).
This article reviews experimental procedures used by our group to
prepare and characterize caged reagents and caged protein conjugates of
cytoskeleton proteins. The labeling methods described in this article are
quite general and may form the basis of a labeling procedure to cage almost
any protein. Similarly optimized irradiation conditions are presented that
may be used for light-directed activation of caged protein conjugates.
Methods
Materials
The following materials are used:
6-Nitroveratryloxycarbonyl chloride (NVOC-C1), stored under dry sil-
ica gel in the dark (Fluka, Ronkonkoma, NY); 4,5-Dimethoxy-2-
nitrobenzyl bromide (DMNBB); acetone (Aldrich, Milwaukee, WI).
Tetramethylrhodamine (TMR) phalloidin; Na2+-ATP; polylysine
10,000 MW; dimethylaminopyridine; catalase (CAT); glucose oxi-
dase (GO) (Sigma, St. Louis, MO).
5'-Iodoacetamidotetramethylrhodamine (IA-TMR, a gift of Dr. John
Corrie, NIMR, Mill Hill, London), also available as mixed isomers
from Molecular Probes (Eugene, OR).
AB buffer: 25 mM imidazole, 25 mM KC1, 1.0 mM dithiothreitol
(DTT), 4 mM MgCI2, pH 7.0.
AB/GO/CAT/glucose/DTT buffer; glucose oxidase 100 ~g/ml; cata-
lase 18/zg/ml; glucose 3 mg/ml.
G-buffer: 2 mM Tris, 0.2 mM ATP, 0.2 mM CaCI2, 1 mM DTT, pH 8.0.
20D. Gabriel, diplomathesis, LudwigsMaximiliansUniversitiit,Mtinchen(1995).

Instrumentation
Absorption measurements are performed using the Hewlett-Packard
(Palo Alto, CA) 8452A diode array spectrophotometer. Light-scattering
measurements are made in an Aminco Bowman Luminescence Spectrome-
ter Series 2 (Sopra, Buttelborn, Germany).
Consideration for Preparation of Caged Proteins by Modification of

L ysine Residues
NVOC-C1 was widely employed in the 1970s in peptide and carbohy-
drate chemistry as a light-sensitive protection group for the amino groups
of lysine and amino sugars. 21 The carbamate bond linking the NVOC to
the amino group may be cleaved on illumination with near-ultraviolet light,
releasing the deprotected peptide or sugar with a respectable yield. 22-23
We have shown that caged proteins can be prepared using NVOC-C1 as a
reagent to modify essential lysine residues. The carbamate bond that links
NVOC to lysine residues of the protein is also easily cleaved with ultraviolet
light, generating the free amino base of the lysine residue, carbon dioxide,
and 4,5-dimethoxy-2-nitrosobenzaldehyde, as illustrated in Fig. 1.16
In general, derivatization of lysine residues of proteins has a neutral
effect on their activity, although a couple of reports in the literature describe
how the activity of certain proteins is specifically inhibited on modification
with lysine-directed reagents. For example, the polymerization activity of
G-actin is blocked when its lysine-61 residue is modified with fluorescein
i s o t h i o c y a n a t e ( F I T C ) , 24 and the phosphodiesterase-activating function of
calmodulin is reduced about fivefold after modification of its lysine-75
residue with the N-hydroxysuccinimide ester of biotin. 25 In designing a
labeling protocol to prepare caged profilin, we took advantage of the finding
that when profilin is labeled with an average of one tetramethylrhodamine
succinimide ester group, its ability to bind G-actin is reduced by 3 0 % , 26
and presumably higher labeling ratios would lead to a further attenuation
of this interaction. Interestingly, one of the seven lysine residues of profilin
21 V. N. R. Pillai, Synthesis 26, 1 (1980).

22 B. Amit, U. Zehavi, and A. Patchornik, Z Org. Chem. 39, 192 (1974).
23 S. P. A. Fodor, J. L. Read, M. C. Pirrung, L. Stryer, A. T. Lu, and D. Solas, Science 251,
767 (1991).
24 L. D. Burtnick, Biochim. Biophys. Acta 791, 57 (1984).
25 D. Mann and T. C. Vanaman, Methods Enzymol. 139, 433 (1992).
Cell 76, 505 (1994).
[6] L I G H T - D I R E C T E D A C T I V A T I O N O F P R O T E I N ACTIVITY 99
0
/
0 CO
H
H.~. / 0
N H - (CH2)4
Borate buffer . 0
NH
NO2 p H 8.5-9.5 | NO2
I
H3CO
H3CO
OCH3
OCH3
320 n m -420 nm
5 m M DTT, pH 6.9
Hyo CO
~ NO + NH2-- (CH2)4 + CO2
Nil
H3CO I
OCH3
FIG. 1. Proposed reaction mechanism of the NVOC-lysine photodeprotection reaction.
is located in a putative actin-binding domain, and we assume the caged

profilin is formed after modification of this residue with NVOC-C1.
Absorption spectrophotometric analysis of the NVOC-Cl-modified pro-
tein conjugate, after centrifugation and removal of excess NVOC-CI and
other breakdown products, can be used to determine the number of NVOC
groups attached to the protein. In general the caged protein conjugate
should be labeled with the minimum number of NVOC groups necessary to
inactivate its activity. To illustrate this point we find that the polymerization
activity of caged G-actin containing between three and five NVOC groups
is fully restored after irradiation with near-ultraviolet light, whereas very
little polymerization activity is recovered after irradiation of a caged G-actin
containing an average of eight NVOC groups (Ref. 16 and Marriott, 1997,
unpublished observations).
Procedures
Preparation of 4-Dimethylaminopyridine Salt of

6-Nitroveratryloxycarbonyl Chloride
4-Dimethylaminopyridine (DMAP), 11 mg (0.09 mmol), dissolved in
300/zl dry acetonitrile is mixed with 20 mg of NVOC-CI dissolved in 200
/~1 dry dioxane. After a 15-min reaction time the yellow-orange precipitate
is centrifuged for 10 min at 10,000 g and the supernatant discarded. The
precipitate is washed with dioxane and centrifuged again for 10 min at
10,000 g. The solid residue is dried and stored over silica in the dark and
used within a month of its preparation.
Yield: 25 mg (0.063 mmol) = 88%
Molecular Weight = 397.83 g/mol; C16H20N30C1;cation MG = 362.38
FAB-MS: m/z 362.3 (M +)
mp = 80° (decomposition)
1H NMR: (DCC13, 500 MHz): ~ [ppm] = 8.12 (d, 2H), 7.71 (s, 1H),
7.02 (s, 1H), 6.71 (d, 2H), 5.60 (S, 2H), 3.97 (S, 3H), 3.94 (s, 3H),
3.20 (s, 6H)
Preparation and Spectroscopic Characterization of

6-Nitroveratryloxycarbonyl-Labeled Polylysine
A freshly prepared solution, 100/zl, of 10 mM NVOC-CI in acetone is
added rapidly to 2 mg/ml polylysine in 0.1 M sodium borate buffer, pH
9.0, and the reaction is left for 30 min at 20° in the dark. Unreacted and
hydrolyzed NVOC-C1, which may form a precipitate during the reaction,
are removed by centrifugation at 10,000 g for 20 min. This supernatant
fraction is dialyzed overnight against two changes of 3 liters of 10 mM PIPES
(piperazine-N,N'-bis-[2-ethane-sulfonic acid]), pH 6.9. Because polylysine
does not contain any aromatic amino acids, it can be used to define the
ultraviolet spectral properties of NVOC-lysine in the conjugate. The ab-
sorption of NVOC-labeled polylysine at 350 nm is used to calculate its
concentration in the conjugate using an extinction coefficient of 5000 M -1
cm-1. 27 Using this calculated concentration and the absorption value of
NVOC-polylysine at 280 and 290 nm, it is possible to calculate, using the
Beer-Lambert equation, the extinction coefficients of NVOC at 290 nm
(3400 M -1 cm -1) and at 280 nm (3050 M -1 cm-1). Knowledge of these two
extinction coefficients and those of the protein at 280 or 290 nm may then
be used to determine the NVOC/protein labeling ratio for almost any
NVOC-labeled protein conjugate.
27 j. F. C a m e r o n and J. M. J. Frrchet, J. Am. Chem. Soc. 113, 4202 (1991).

[61 LIGHT-DIRECTED ACTIVATION OF PROTEIN ACTIVITY 101
Extinction Coefficients Used in This Study

G-Actin (290 nm) 28 = 26,000 M -1 cm -a
Profilin I (280 nm) 29 = 19,950 M -1 cm -1
Profilin II (280 n m ) 29 = 18,700 M -1 cm -1
N V O C (350 nm) 27 = 5,000 M -1 cm -a
N V O C (290 nm) = 3,400 M -1 cm -1
N V C O (280 n m ) = 3,050 M -1 cm a
T M R - I A (549 nm) 3° = 96,400 M -a cm -1
Preparation of Caged G-Actin

A freshly prepared solution, 100/zl, of 0.1 M NVOC-C1 solution in
acetone is added in rapid succession to 1.95 ml of a sodium borate based
G-buffer (2 m M borate, 0.2 m M ATP, 0.2 m M CaC12, 1 m M DTT, p H
8.5) and 6 ml of 36 /xM G-actin, prepared according to the method of
Pardee and S p u d i c h , 3°a and the reaction is left for 30 min in the dark. The
final concentrations of G-actin and NVOC-CI are 27/~M and 1.25 mM,
respectively, with an acetone content of 1.25%. The addition of NVOC-CI
in this manner reduces the amount of denatured protein that forms by
directly mixing acetone and aqueous solution of proteins. After 30 min the
sample is centrifuged at 10,000 g for 10 min at 4 ° to remove insoluble
material and then dialyzed overnight against 2 m M PIPES-based G-buffer,
p H 6.9, at 4 ° with two changes of buffer. Magnesium chloride and KCI are
added to the G-actin conjugate at 2 m M and 0.1 M, respectively, and after
90 min at 20 °, the reaction is centrifuged at 100,000 g for 60 min at 4 °.
The pellet fraction, which contains polymerization competent G-actin, is
discarded, and the supernatant fraction, which contains caged G-actin, is
analyzed by absorption spectroscopy. The absorption value at 290 nm of
G-actin in the caged G-actin conjugate is determined by subtracting the
contribution to the 290-nm absorption from the N V O C group. The G-
actin concentration in the conjugate is then calculated using an extinction
coefficient at 290 nm of 26,000 M -a cm -a.
Characterization of Caged G-actin

Using the conditions of N V O C conjugation described above, the label-
ing ratio of N V O C to G-actin normally falls between three and five N V O C
groups per G-actin monomer, with a caged G-actin yield of 30 to 60%. The
28 D. J. Gordon, D. Eisenberg, and E. D. Korn, J. BioL Chem. 251, 4778 (1976).

29 M. Haugwitz, A. Noegel, D. Rieger, and M. Schleicher, J. Cell Sci. 1 ~ , 481 (1994).
30 j. E. T. Corrie and J. S. Craik, J. Chem. Soc. Perkin Tran. 1, 2967 (1994).
30~j. D. Pardee and J. A. Spudich, Methods Cell Biol. 24, 271 (1982).
absorption spectrum of an 11.3/xM solution of caged G-actin containing

3.9 NVOC groups per actin monomer in 2 mM PIPES-based G-buffer, pH
6.9, is shown in Fig. 2A. The absorption of NVOC on G-actin is centered
at 350 nm and extends to 420 nm. Consequently, all operations on caged
G-actin must be conducted in the dark or at least under subdued light.
Caged G-actin is best used within a day or two of its preparation and
should be centrifuged at 100,000 g for 60 min before an experiment. The
polymerization competence of G-actin and NVOC-labeled G-actin in
F-buffer can be shown by sodium dodecyl sulfate-polyacrylamide gel elec-
trophoresis (SDS-PAGE). Solutions of G-actin or caged G-actin, 20/zM,
in 2 mM PIPES-based G-buffer are made into F-buffer, and after a 90-min
period to allow for polymerization, the F-actin fraction is separated from
G-actin by high-speed centrifugation (60 min at 4° in a Beckman airfuge
at 22 psi). The amount of protein in the supernatant and pellet fractions is
analyzed by SDS-PAGE and densitometry of the Coomassie-stained gel.
04\
A 05
0.3 ",
0.2-.]\'..,' ............................
300 350 400 450 500

WAVELENGTH (rim)
o 1.o \ r'...
" ~ i -
m 0 5 "~ .... ] /"" "

< " 3\ ~............. .,, ,..--, ,~
4 \ , , / .'A,,
300 350 400 450 500 550 600 650

WAVELENGTH (rim)
FIo. 2. Absorption spectra of (A) caged G-actin; (B) 25 ~ M TMR-labeled caged G-actin
(solid lines, caged; dotted lines, irradiated F-actin pellet resuspended in 500 ~1 F-buffer;
dashed lines, irradiated supernatant).
We normally find that more than 95% of the control G-actin can polymerize,
whereas less than 5% of the caged G-actin is found in the pellet fraction.
Preparation of Tetramethylrhodamine-Labeled Caged G-Actin

In principle, fluorescence image microscopy of fluorescently labeled
caged G-actin microinjected into living cells can be used to show that caged
G-actin incorporates into the actin cytoskeleton only after it has been
irradiated with near-ultraviolet light. The following method is used to pre-
pare a caged G-actin fluorescently labeled at cysteine-374 residue with
TMRIA. G-Actin at a concentration of 100/zM is dialyzed against 2 mM
Tris-based G-buffer, pH 8.0, in the absence of DTT, clarified by centrifuga-
tion, and treated at 4° in an overnight reaction with a 2-mole excess of 5'-
IATMR, added from a stock solution of 20 mM in dimethylformamide
(DMF) that is stored at - 8 0 °. Free dye is removed from the conjugate by
dialysis against three changes of 2 mM Tris-based G-buffer, pH 8.5, con-
taining 1 mM DTF, and then centrifuged at 100,000 g for 60 min at 4°. To
remove polymerization-incompetent G-actin in this preparation, MgC12 and
KCI are added to 2 and 50 mM, respectively, to make F-buffer, and after
90 rain at 20° the solution is centrifuged at 100,000 g for 60 min at 4°. The
F-actin pellet is resuspended in 2 mM borate-based G-buffer and dialyzed
overnight against 1 liter of 2 mM borate-based G-buffer. The fluorescently
labeled G-actin in the supernatant is centrifuged once more at 100,000 g
for 60 min at 4° and the concentration of G-actin established using the
Bradford assay.31 5'-IATMR-labeled G-actin at 25 /xM is treated with
NVOC-CI, added from a fresh 0.1 M acetone stock solution, to achieve a
final concentration of 1.0 mM. After a 30-min reaction time at 20° the
double-labeled conjugate is centrifuged at 10,000 g for 20 min at 4° to
remove insoluble material, and the supernatant is dialyzed overnight against
two changes of 2 liters of 2 mm PIPES-based G-buffer, pH 6.9, containing
1 mM DTT, and then centrifuged at 100,000 g for 60 min at 4°. To initiate
polymerization, MgC12 and KC1 are added to the supernatant to final con-
centrations of 2 and 50 mM, respectively, and after 90 min at room tempera-
ture the reaction is centrifuged at 100,000 g for 60 min at 4°; the TMR-
labeled caged G-actin conjugate is present in the supernatant fraction. The
yield of TMR-labeled caged G-actin is generally higher if G-actin is first
reacted with the fluorophore and then caged with NVOC-C1. The absorp-
tion spectrum of a 25/xM solution of TMR-labeled caged actin is shown
in Fig. 2B. The concentration of TMR in the conjugate of 10/zM is calculated
using an extinction coefficient of TMR of 96,400 M -1 cm- ~at 549 nm, and the
G-actin concentration of 25/xM is determined using the Bradford a s s a y . 31
31 M. M. Bradford, Anal. Biochem. 72, 248 (1976).

104 CACZDCOMPOUNDS 161
Photoactivation of Caged Protein Conjugates in Solution

Prior to photoactivation, the caged G-actin conjugate is dialyzed into
PIPES-containing F-buffer containing 5 mM DTT, pH 6.9, and centrifuged
at 100,000 g for 60 min to remove any denatured protein or any polymeriza-
tion competent G-actin that may form during storage. Samples of caged
G-actin, 500/.d, are irradiated in a quartz cuvette (4 × 10 mm) with the
near-ultraviolet light output of a 100-W high-pressure mercury arc lamp
selected with an interference filter that transmits light between 340 and
400 nm (Dr. Rapp, Opto-Elektronik, Hamburg, Germany). The intense
infra-red irradiation of the mercury arc lamp is blocked using two copper
sulfate glass filters and a 10-mm-thick distilled water filter. To ensure a
uniform irradiation of the caged protein solution in the cuvette, especially
during a prolonged irradiation treatment, the near-ultraviolet light beam
is defocused at the center of the cuvette, which is occasionally agitated.
Fluorescence and Photoactivation Microscopy

Simultaneous fluorescence microscope-based imaging of actin filaments
and photoactivation of caged HMM in an in vitro motility assay is performed
essentially as described by Marriott and Heidecker TM and Mitchison and
co-workers 19,32 and is illustrated in Fig. 3. The near-ultraviolet photolysis
beam is selected with a broad-band near-ultraviolet interference filter (340-
400 nm) at the source of a 100-W mercury arc lamp that is aligned perpendic-
ular to the body of an Axiovert 35 microscope (Zeiss, Oberkochen). A
dichroic mirror that reflects light between 340 and 420 nm, mounted onto
the normal mercury arc lamp port of the microscope, directs the photolysis
beam into the microscope. A second 100-W mercury arc lamp, aligned
parallel to the microscope body, is used to excite rhodamine/CY3-based
fluorescent dyes in the image field. A custom-made bandpass filter that
selects the 546-nm line (Chromatech, Brattleboro, VT) is placed in front
of the lamp, and this beam passes through the external dichroic mirror and
into the microscope. The fluorescence excitation filter in the microscope
filter assembly is removed in order to transmit the two coaligned excitation
beams. The dichroic mirror and emission filter are custom made for the
CY3 fluorophore (Keio filter set; Chromatech, Brattleboro, VT); the CY3
dichroic filter reflects the photolysis and fluorescence excitation beams to
the microscope objective, while the emission filter blocks the ultraviolet
light and transmits the CY3 or rhodamine emission between 565 and 590
nm. TMR-phalloidin-labeled actin filaments in the image field are recorded
with an intensified charge-coupled device camera (C2400, Hamamatsu Pho-
32T. J. Mitchison,J. CellBiol. 109,637 (1989).

iI ~!
FIG. 3. Schematic representation of the microscope used for light-directed activation of

caged proteins with simultaneous imaging of tetramethylrhodamine fluorescence. A, 100-W
mercury arc lamp for irradiation of caged proteins; B, 340- to 400-nm interference filter;
C, manually controlled shutter with adjustable iris; D, dichroic mirror mounted in an assembly
attached to the lamp port of the microscope used to reflect light between 340-420 nm into
the microscope; E, 100-W mercury arc lamp used for excitation of fluorescence dyes; F, 546-
nm interference filter; G, dichroic mirror from CY3 custom filter set used to reflect 340- to
400-nm and 546-nm light; H, emission filter from CY3 custom filter set used to collect light
between 565 and 590 nm and to block near-ultraviolet light; I, in vitro motility chamber;
J, intensified charged-coupled device camera.
tonics, Herrsching, Germany), and the images are simultaneously saved on

a video tape and in the memory of a frame grabber at 0.5-sec intervals
(Scion Inc., Frederick, PA) under the control of a Macintosh Ilfx computer.
Functional Analysis of Caged G-Actin and Decaged G-Actin

The activity of G-actin is measured by its ability to polymerize in
F-buffer. A 25/zM solution (500/xl) of caged G-actin in 2 mM PIPES-
based F-buffer, pH 6.9, containing 5 mM DTF is irradiated in a microcuvette
(4 x 10 mm) with near-ultraviolet light as described in the previous section.
Aliquots of the protein sample, 50/zl, are removed at defined exposure
times and following a 90-min period at 20° in the dark to allow for polymer-
ization; the samples are centrifuged in an airfuge at 22 psi for 60 min at 4°
to pellet F-actin. The supernatant is removed and the pellet is resuspended
in 50/~1 of PIPES-based F-buffer. The amount of actin in the supernatant
and pellet fractions is determined by S D S - P A G E as described previously.
Approximately 50% of polymerization-competent G-actin is generated
from caged G-actin following 2 min of irradiation with near-ultraviolet
light, and this yield increases to almost 90% after 12 min of irradiation (Fig.
4). Proof that the caged groups have been removed from the polymerization-
competent, decaged G-actin fraction is shown by absorption spectrophoto-
metric analysis of the pellet and supernatant fractions of the high-speed
centrifugation (Fig. 2A). The ratio of the absorbance at 290 and 350 nm,
an indicator of the amount of decaged protein, is much higher in the pellet
fraction than in the preirradiated sample, whereas the absorption of the
supernatant fraction is dominated by the soluble photoproducts, as SDS-
PAGE shows less than 5% of the protein remains in this fraction after 12
min of irradiation (Fig. 4). Similar results are obtained with the TMR-
labeled caged G-actin conjugate, although the yield of polymerization-
competent TMR-labeled G-actin obtained after 12 min of irradiation with
ultraviolet light is lower than that found for the nonfluorescent-caged
G-actin (Fig. 2B). The in vitro motility assay is used to show that F-actin
filaments generated by the photoactivation of caged G-actin are functional
and indistinguishable from normal actin filaments in terms of the precentage
FIG. 4. SDS-PAGE and Coomassie-stained gel of the 100,000-g supernatant and pellet
fractions of 20/~M caged G-actin in 2 mM PIPES-based F-buffer as a function of the exposure
time to near-ultraviolet light.
[6] LIGHT-DIRECTED
ACTIVATIONOF PROTEIN ACTIVITY 107
FI~. 5. In vitro motility assay F-actin filaments labeled with TMR-phalloidin. The actin
was prepared by light-directed activation of a 500-/~1 sample of caged G-actin as described
in the text. Scale bar is 3/xm. The filaments were observed to slide on the HMM surface for
several minutes. The three images captured by the frame grabber show the movement of
filaments over a 1-sec interval.
of filaments that m o v e on the H M M surface with a m a x i m u m sliding velocity

of 3 / z m / s e c (Fig. 5).
Preparation of Profilin
R e c o m b i n a n t profilin f r o m Dictyosteium discoideum is isolated f r o m an
Escherichia coli strain transfected with the expression v e c t o r JM83 p I M S
profilin I or profilin II. 33 I n d u c t i o n of the JM83 p l M S profilin I and profilin
II is achieved by addition of 0.5 m M I P T G (iso p r o p l y l t h i o - g a l a c t o p y r a n o -
side) for 2 hr at 37 °. T h e cells are centrifuged (45 min, 10,000 g) and w a s h e d
with T E buffer (10 m M Tris-HC1, 2 m M E D T A , p H 8.0). A f t e r a second
33L. Eichinger, A. A. Noegel, A. A., and M. Schleicher, J. Cell Biol. 112, 665 (1991).
108 cACzD COMPOUNI~S I61
centrifugation (45 min, 10,000 g) the pellet is resuspended in TEDABP
buffer [10 mM Tris-HC1, 1 mM EGTA, 1 mM DTT, 0.02% NAN3, 1 mM
benzamidine, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), pH 8.0] and
freeze-thawed twice in liquid nitrogen. After lysis with 0.4 mg/ml lysozyme,
the cells are homogenized on ice and sonicated seven to eight times. SDS-
PAGE analysis is used to show that the majority of the profilin is solubilized
in TEDABP containing 2 and 7 M urea, respectively. The two urea wash
fractions are pooled and dialyzed against TEDABP and are then applied
to a TEDABP equilibrated polyproline-Sepharose column prepared ac-
cording to the method of Rozycki et al. 34 Affinity-bound profilin is eluted
with TEDABP buffer in two steps using 2 and 7 M urea. Profilin I and II
are found primarily in the 7 M urea fraction. This fraction is dialyzed against
several changes of 50 mM Na4B204, 1 mM EDTA, pH 8.5, prior to labeling
with the DMAP salt of NVOC-C1.
Preparation of 6-Nitroveratryloxycarbonyl-Labeled Profilin

Purified profilin, 4 ml (40 ~M), in 50 mM borate buffer, 1 mM EDTA,
pH 8.5, is mixed with a fourfold molar excess of solid NVOC-CI-DMAP,
a water-soluble and activated form of NVOC-CI.
After a 1-hr reaction at 4 °, the slightly yellow reaction mixture is centri-
fuged (10,000 g, 30 rain) and the white-yellow pellet is discarded, and the
supernatant is dialyzed four times at 4° against 2 liters of 2 mM Tris-HCl,
0.5 mM DTT, 0.1 mM CaC12, pH 8.0. The solution is then clarified by
centrifugation (10,000 g, 30 min, 4°), and the protein and NVOC concentra-
tion in the caged profilin conjugate are determined by absorption spectros-
copy using the extinction coefficients listed earlier. The best preparations
of caged profilin contain two to three NVOC groups per profilin molecule.
Functional Analysis of Caged Profilin

The activity of profilin is determined by its ability to bind to G-actin in
physiological salt with a concomitant inhibition of G-actin polymerization.
The time-dependent increase in a light-scattering signal that accompanies
the transition of G-actin to F-actin may be used to monitor the kinetics of
the polymerization reaction. The scattering signal is measured in an SLM-
AMINCO fluorimeter (R~ittelborn, Germany) through crossed polarizers
setting the excitation and emission wavelengths at 350 nm with a 4-nm band-
pass. The 350-nm probing light is highly attenuated compared to the amount
of light delivered by the 100-W mercury arc lamp and is not expected to result
in the photoactivation of any significant amount of caged protein. A 4.7 txM
34M. Rozycki,C. E. Schutt, and U. Lindberg, Methods Enzymol. 196, 100 (1991).
A 14o
120-
t
iI
.~ 100- i
ii
t
~, 80- tI
I
f
6O- iI
40-
20
0 16oo ~doo 3600 4600 5000
time [sec]
B 16~
140-
120- .'7
,7
loo-
80- .[ /" .
60-
40-
_ f
20
lOOO 2600 3600 4000
time [sec]
FIG. 6. Polymerization of G-actin monitored by light scattering of F-actin filaments. The

polymerization was induced by the addition of MgC12 and KCI to G-actin in G-buffer, and
the scattered light was monitored through cross-polarizers using 350 nm for both excitation
and emission wavelengths. (A) The upper curve is the polymerization profile for 4.7 /zM
G-actin, and the lower curve is the profile for 4.7/.tM G-actin in the presence of 12/zM
profilin. (B) The curves (from top to bottom) represent the polymerization of 4.7/zM G-actin
in the presence of caged profilin that has been irradiated with near-ultraviolet light as described
in the Methods section for 0, 2.5, 5, and 10 min, respectively.
solution of G-actin from Dictyostelium discoideum, prepared according to

Prassler e t aL,TM in 2 mM PIPES-based G-buffer is treated with MgC12 to
1 mM and KCI to 50 mM, respectively, to start polymerization, and the
light-scattering signal is measured for 100 min (Fig. 6A). Normal profilin
34aj. Prassler, S. Stocker, M. Heidecker, G. Marriott, and G. Gerisch, Mol. Cell Biol. 8,
83 (1997).
at 12/xM binds to G-actin and greatly decreases the rate and extent of
G-actin polymerization compared to G-actin alone (Fig. 6A). However,
caged profilin at 14/zM has almost no effect on the kinetics or extent of
polymerization because of its inability to bind G-actin (Fig. 6B, second
curve from top). To show that profilin is activated from caged profilin after
exposure to ultraviolet light, 100-/xl samples of caged profilin subjected to
irradiation for 2.5, 5, and 10 min as described earlier are added to separate
solutions of G-actin to achieve a full profilin concentration of 12/~M and
their effect on polymerization, initiated by the addition of MgC12 and KCI
to 2 and 50 mM, respectively, is measured by light scattering, as shown in
Fig. 6B. Increased exposure of the caged profilin conjugate to intense
ultraviolet light leads to a corresponding increase in the inhibition of G-
actin polymerization, which evidently results from the binding of functional,
photoactivated profilin to G-actin.
Preparation of Caged Proteins by Modification of Cysteine Residues

A second labeling strategy to prepare caged protein conjugates is based
on the targeted modification of essential cysteine residues of the protein
using photolabile alkyl halides. For example, each of the two heads of
HMM harbor an ATPase domain with two highly reactive cysteines (Cys-
707 and Cys-697) that can be specifically derivatized with a variety of
haloacetyl-alkylating reagents. Modification of either cysteine residue re-
suits in considerable perturbations of the actin-activated ATPase of myosin
or its subfragments. 35'36 However, it is important to recognize that at a
functional level, the activity of myosin is to couple the energy derived from
the actin-activated ATPase to the production of molecular forces that move
actin filaments. In this context the activity of HMM is better characterized
using the in vitro motility a s s a y 37 rather than assays that measure only its
ATPase activity. Thus HMM labeled at Cys-707 with thiol reactive groups
exhibits normal or even elevated Mg 2+, or CaZ+-activated ATPase, but these
conjugates do not support the sliding of F-actin filaments in the in vitro
motility assay, a8,38We reasoned that if the thiol group of Cys-707, or Cys-
697, could be alkylated using a photolabile protection group for mercap-
tans, 15 then this conjugate would not be able to couple the energy derived
from hydrolysis of ATP to force production and the movement of actin
filaments, until the caged groups are removed from the conjugate with
near-ultraviolet light. This type of caged HMM can be prepared by reacting
35 S. A. M u l h e r n and E. Eisenberg, Biochemistry 17, 4419 (1978).

36 E. Reisler, Methods Enzymol. 85, 84 (1982).
37 S. J. K r o n and J. A. Spudich, Proc. Natl. Acad. Sci. U.S.A. 83, 6272 (1985).
38 D. D. Root and E. Reisler, Biophys. J. 63, 730 (1992).
161 LIGHT-DIRECTED ACTIVATION OF PROTEIN ACTIVITY 111
bromomethyl-2-nitro-4,5-dimethoxyenzene (DMNBB: Fig. 7) with HMM

under neutral to slightly basic conditions, where it specifically alkylates
cysteine-707. The 4,5-dimethoxy-2-nitrobenzyl bromide group has a broad
near-ultraviolet absorption band that extends into the blue region of the
spectrum, and its high molar absorptivity (E350nm of 5000 M -1 cm -1) is a
useful feature if the caged protein is to be operated on as a monolayer or
in optically thin samples such as cells and single muscle fibers.
Preparation of Caged Heavy Meromyosin

Heavy meromyosin is prepared from rabbit muscle according to Marri-
ott and Heidecker TM using a purification procedure that has been optimized
to preserve the functional activity of myosin. Actin is labeled with tetra-
I
NH I
NH
"c=o X,c=o
H 0
HS --CH2- C~H
ABbuffer, pH7.4 H H S__CH2_C~/
~'O"
+
I
NH
\C=O 0+"O"~ [ ~C=O
H3CO
OCH3
OCH3
340-400nm
H 0 I
NH
. N aO X~¢C=O +H + \C=O
+ HS ~CH2-C~ H S--CH2-C~
H3CO I
OCH3 NH
\C=O
I I
OCH3
aci-nitro intermediate
FIG. 7. Structure of D M N B B and proposed mechanism of the protection and light-directed
cleavage of the thioether bond linking D M N B B to the thiol of Cys-707 of HMM.
methylrhodamine phalloidin according to the method of Heidecker et aL 39

All labeling reactions that use DMNBB are performed in a darkened room
or protected from room light with aluminum foil, and because DMNBB is
a very potent alkylating reagent it must be handled with due care and
should be dispensed and weighed only in small quantities in a well-ventilated
hood. DMNBB-labeled HMM is prepared by treating HMM at 18.6/xM
in AB in the absence of DTT with DMNBB (added from a fresh 9 mM
stock solution in DMF) to a final concentration of 186/xM for 75 min at
room temperature. After an overnight dialysis at 4° against 2 x 1 liter
of AB buffer with 10 mM DTT and clarification by centrifugation at
100,000 g for 20 min at 4°, the absorption spectrum of the protein conjugate
is recorded and the labeling ratio calculated using an extinction coefficient
of DMNBB at 350 nm of 5000 M -1 cm 1, and the HMM concentration is
determined using the Bradford assay.31
These labeling conditions normally result in the modification of two
4,5-dimethoxy-2-nitrobenzyl groups per HMM molecule, or one caged
group per ATPase. Inclusion of ATP to 1 mM does not influence the
labeling ratio, and DMNBB groups are not incorporated into HMM after
a prereaction with NEM (N-ethylmaleimide) according to the method of
Reisler36 (data not shown), which would suggest the DMNBB labeling site
is restricted to the SH-1 thiol. The ATPase activity of HMM and caged
HMM is measured using the malachite green assay for the detection of
phosphate according to Kodama et aL 4° The Ca2+-activated ATPase of
caged HMM is increased more than fivefold compared to unmodified HMM
(data not shown). This activating effect is often used as an indicator for
the modification of the SH-1 thiol group of myosin and its subfragments. 36
Irradiation of caged HMM at 18/zM in AB buffer/10 mM DTT with
near-ultraviolet light (340-400 nm) leads to a dose-dependent decrease in
the intensity of the So - $1 transition. After dialysis, a comparison of the
absorption spectrum of a nonirradiated caged HMM with the caged HMM
conjugate that had been exposed to irradiation for 30 min shows that the
ultraviolet light treatment removes most, if not all, of the protection groups.
The structureless, residual near-ultraviolet absorption spectrum of the de-
protected conjugate probably results from secondary photochemical reac-
tions of the photoproduct 2-nitrosobenzaldehyde with HMM during the
long exposure time. However, in the in vitro motility assay, full activation
of HMM requires, at most, several hundred milliseconds of illumination
with near-ultraviolet light, during which time these secondary photochemi-
cal reactions should be negligible.
39 M. Heidecker, Y. Yan-Marriott, and G. Marriott, Biochemistry 34~ 11017 (1995).

4o T. Kodama, K. Fukui, and K. Kometani, J. Biochem. 99, 1465 (1986).
[61 LIGHT-DIRECTED ACTIVATION OF PROTEIN ACTIVITY 113
Transient Kinetics Investigations of Photoactivation of Caged

Heavy Meromyosin
The kinetics of the photoactivation reaction of caged HMM is deter-
mined by time-resolved absorption spectroscopy after irradiation of caged
HMM with a pulse of near-ultraviolet light, using an instrument similar to
that described by Uhl et al. 41 Transient absorption spectra of the aci-nitro
intermediate of DMNBB recorded at defined times (0.1 to 50 msec) after
irradiation with a 50-nsec laser-generated pulse of 380-nm light are shown
in Fig. 8A. Less than 1% of the DMNBB in caged HMM is photolyzed per
irradiation pulse, as measured by the change in the value of the steady-
state absorption of the caged group at 350 rim. The peak of the DMNBB
aci-nitro transient absorption spectrum occurs at 440 nm and is already
present 100 txsec after the photolysis pulse. The value of the transient
absorption at 440 nm decays in a first-order reaction with a rate constant
of 45.6 sec -a at pH 7.4 (Fig. 8B). Because this decay defines the rate-limiting
step in the photocleavage reaction, 2 it is reasonable to assume that this will
also be the rate at which the actin-activated, force-generating ATPase
activity of HMM is generated by light-directed activation of caged HMM.
Preparation and Photoactivation of Caged Heavy Meromyosin in

In Vitro Motility Assay
The in vitro assay is performed according to Marriott and Heidecker. TM
The in vitro chamber is composed of a meticulously cleaned bottom cov-
erslip (50 x 25 mm) coated with nitrocellulose and separated from an upper
acid-washed coverslip (24 x 24 mm) by two strips of paper greased with
silicon on both sides. Heavy meromyosin at 50 /~g/ml in AB buffer is
introduced from one side of the chamber for 90 sec and then again from
the opposite side for 90 sec. The chamber is washed free of HMM by three
changes of AB buffer without DTT and then incubated for 30 min in the
dark with AB buffer containing 45 t~M DMNBB, added from a fresh DMF
stock solution. After this reaction time, excess DMNBB is washed out of
the chamber with three changes of AB buffer containing 10 mM DT1L
Actin filaments labeled with tetramethylrhodamine phalloidin are then
added to a concentration of 20 nM, and after 90 sec unbound filaments
are washed out with AB/GO/CAT/DTT buffer and 1 mM ATP is added
to the chamber in AB/GO/CAT/glucose/DTT buffer to initiate sliding of
filaments by fully functional HMM.
The labeling conditions described above produce a monolayer of caged
HMM that does not support motility of F-actin filaments in the in vitro
41 R. Uhl, B. Meyer, and H. Desel, J. Biochem. Biophys. Methods 10, 35 (1984).

A 2.0 -~
1.5
,.~ 1.0
<
0.5
. . . . ~.......~. ,....~,.
0.0
''' I ' ' ' I ' ' ' I ' ' ' I ' ' ' I ' ' ' I ' ' ' I ' '
400 420 440 460 480 500 520 540
Wavelength (nm)
B
2.0
1.5
©
'~1.o
<
0.5
,°
0.0
i~...i.,~.i.,,.i...,i..,.1..,.i,,,,i....i.,,,i,,.,i...,i,.,.i ,
) 10 20 30 40 50 60
T I M E A F T E R P U L S E [ms]
FIG. 8. (A) Selected time-resolved absorption spectra of a 21.4/zM solution of DMNBB-

labeled HMM in AB buffer containing 5 mM DTT following irradiation with a 10-nsec pulse
of 380-nm light; 0.1 msec (top), 10 msec (middle), and 50 msec (bottom). (B) Time-resolved
absorption decay of sample described in A recorded at 440 nm.
motility assay in the presence of 1 m M ATP (Fig. 9), even though its
Ca2+-activated. ATPase activity is elevated compared to unmodified HMM.
Irradiation of the caged HMM-coated surface with one or two 500-msec
pulses of 340-400 nm light delivered from the mercury arc lamp leads to
the movement of more than 90% of the actin filaments with a sliding velocity
of up to 4/zm/sec (Fig. 9), which is similar to that measured for normal

HMM. After the photolysis pulse, shorter actin filaments of 2/zm or smaller
run smoothly, whereas longer filaments sometimes sever in the first few
seconds and then run smoothly. Actin filaments in the image field adjacent
to the photolyzed image field do not move, which allows hundreds of flash
experiments to be made on a single chamber. Irradiation of a surface of
caged HMM with a polarized pulse of near-ultraviolet light, to photoselect
those DMNBB molecules attached to HMM whose absorption dipole mo-
ment lies in the direction of the electric vector of the photolysis beam,
results in filaments moving in all directions and is not restricted to the
direction of the electric vector of the photolysis light. These data suggest
indirectly that the head domains of the surface-bound HMM have a consid-
erable degree of rotational freedom around their point of attachment.
The millisecond photogeneration kinetics of actin-activated ATPase
activity from caged HMM, together with the high efficiency of light-directed
cleavage of the thioether bond, makes caged myosin a unique probe to
study the mechanism of muscle contraction using in vitro model systems and
FIG. 9. Images of single actin filaments on caged HMM before and after flash photolysis
of the image field. Images of the fluorescent actin filaments on caged HMM taken from - 3
to -0.5 sec do not show any actin sliding motion. Irradiation of the image field for 0.5 sec
initiates the movement of most actin filaments in the field. The velocity of the filaments is
2-4/zm/sec. Actin filaments in the image field adjacent to the irradiated area do not move
(data not shown).
in permeabilized muscle fibers. Interestingly, generation of actin-activated

ATPase activity from caged HMM has similar activation kinetics to caged
ATP 2 but requires a fraction of the irradiation energy and liberates far less
photoproduct and, unlike caged ATP, caged myosin does not suffer the
complication of being a competitive inhibitor of myosin. 42
Concluding Comments
We have shown that caged proteins can be prepared using both lysine-
and cysteine-directed caged reagent labeling approaches, The technique
will be useful to cage the activity of a large variety of proteins, although
one limitation of the lysine-directed labeling strategy is the requirement
that wild-type proteins must have a reactive lysine residue in the active
site. The cysteine-directed labeling approach has an important advantage
in this regard in that standard site-directed mutagenesis techniques can be
used to engineer fully active proteins that harbor one or more cysteine
residue(s) in the active site. Modification of such critically located cysteine
residues with DMNBB or other thiol-directed caged reagents will no doubt
improve the ease with which the activity of proteins can be caged, iv Light-
directed activation of protein activity may be used to investigate the function
of diverse proteins within living cells. Our interests center on the role of
actin-binding proteins, motor proteins, and the small GTPase signaling
molecules in the process of cell motility. Light-directed activation of caged
conjugates of these proteins may be used to trigger the activity of the
protein of interest several hours after their microinjection, a time period
necessary for the cell to recover from injuries associated with the im-
palement. Along this line of research, we have initiated programs to develop
and apply caged conjugates of regulatory actin-binding proteins, other myo-
sin isoforms, and constitutively active forms of small GTPases, which will
be used to study spatial and kinetic aspects of protein signaling pathways
that regulate actin filament dynamics and cell motility.
42j. Sleep, C. Herrmann, T. Barman, and F. Travers, Biochemistry 33, 6038 (1993).
[7] TARGETEDCHEMICALMODIFICATION 117
[71 C a g e d P e p t i d e s a n d P r o t e i n s
by Targeted Chemical Modification
By H A G A N BAYLEY, CHUNG-YU CHANG, W . TODD MILLER,
B R E T r NIBLACK, a n d PENG PAN
Introduction
Considerable attention has been directed at the preparation of small
caged molecules, as witnessed by the material in this volume and as reviewed
elsewhere, a-4 In contrast, only recently has momentum gathered around
attempts to produce caged macromolecules for experimental biology. The
primary focus has been on caged peptides and proteins, although other
macromolecules or supramolecular assemblies have been examined, e.g.,
caged oligonucleotides5 and photomodulatable lipid bilayers.6,7 The caged
peptides and proteins described here are 2-nitrobenzyl derivatives that
are irreversibly activated on irradiation. A degree of reversible control of
enzyme activity can be provided by the introduction of photoisomerizable
groups into proteins. 8 However, perfect on-off switching is a formidable
problem and has not yet been achieved.
Previous Approaches
Caged proteins have been made by the random introduction of photo-
cleavable protecting groups through chemical modification of reactive
amino acid side chains. 9q3 However, because of its predictability and repro-
1 H. Morrison, ed., "Biorganic Photochemistry, Vol. 2: Biological Applications of Photochemi-

cal Switches," Wiley-Interscience, New York, 1993.
3 R. S. Givens and L. W. Kueper, Chem. Rev. 93, 55 (1993).
5 p. Ordoukhanian and J.-S. Taylor, J. Am. Chem. Soc. 117, 9570 (1995).
6 A. Ulman, "Ultrathin Organic Films," Academic Press, San Diego, CA, 1991.
7 j. L. Thomas and D. A. Tirrell, Acc. Chem. Res. 25, 336 (1992).
s I. Willner and R. Shai, Angew. Chem. Int. Ed. Engl. 35, 367 (1996).
10 G. Marriott, Biochemistry 33, 9092 (1994).
11 S. Thompson, J. A. Spoors, M.-C. Fawcett, and C. H. Self, Biochem. Biophys. Res. Commun.
201, 1213 (1994).
a2 R. Golan, U. Zehavi, M. Naim, A. Patchornik, and P. Smirnoff, Biochim. Biophys. Acta
1293, 238 (1996).
13 C. H. Self and S. Thompson, Nature Med. 2, 817 (1996).

ducibility, targeted modification is likely to be more useful. Indeed, targeted
modification will be essential for caging single domains in multifunctional
proteins. In one sophisticated form of targeted modification, caged proteins
have been obtained by the incorporation of unnatural amino acids at specific
sites during in vitro translation) 4 A second rational approach to the photo-
regulation of enzymes involves the use of photolabile active-site directed
reagents. For example, the protease thrombin can be caged as an o-hydroxy-
cinnamoyl acyl enzyme. 15,16 Noncovalent photosensitive reagents that
block the active site of a protein until irradiated might also be used. A
third approach is targeted chemical modification of single cysteine resi-
dues, 17A8 which can be introduced by genetic engineering, 17 and this is
described here. Proteins might also be inactivated by site-specific cross-
linking to another macromolecule with a photocleavable bifunctional re-
agent. For example, light-activated immunotoxins have been made by cross-
linking. 19 However, in this example, the sites of attachment were not de-
fined. Surprisingly, less work still has been done on caged peptides as
compared to caged proteins, although many amino acids with side chains
blocked with photolabile protecting groups are available for peptide synthe-
sis. 2° Techniques for caging unprotected peptides in solution are given here.
Caging Peptides and Proteins on Sulfur

Two related approaches for caging peptides and proteins are described
here. They are the attachment of photocleavable protecting groups to the
sulfur atoms of (1) cysteinyl residues or (2) thiophosphoryl groups. In
general, the chemistry has been evaluated with small peptides and then
extended to proteins. Nevertheless, it is stressed that caged peptides are
themselves potentially useful reagents. For example, when uncaged, pep-
tides might prevent protein-protein interactions or be released from the
blockade of the active site of an enzyme.
The highlights of the approaches are summarized as follows. Caged
14D. Mendel, J. A. Ellman, and P. G. Sehultz, J. Am. Chem. Soc. 113, 2758 (1991).
15B. L. Stoddard, J. Bruhnke, N. Porter, D. Ringe, and G. A. Petsko, Biochemistry 29,
4871 (1990).
16 B. L. Stoddard, J. Bruhnke, P. Koenigs, N. Porter, D. Ringe, and G. A. Petsko, Biochemistry
29, 8042 (1990).
17 C.-Y. Chang, B. Niblack, B. Walker, and H. Bayley, Chem. Biol. 2, 391 (1995).
18 G. Marriott and M. Heidecker, Biochemistry 35, 3170 (1996).
19 V. S. Goldmacher, P. D. Senter, J. M. Lambert, and W. A. Bl~ittler, Bioconjug. Chem. 3,
104 (1992).
2o R. P. Haugland, "Handbook of Fluorescent Probes and Research Chemicals," 6th Ed.
Molecular Probes, Eugene, OR, 1996.
[7] TARGETED CHEMICAL MODIFICATION 119
cysteinyl or thiophosphoryl peptides have been made by the chemical modi-

fication of unprotected peptides in aqueous solution. This circumvents pro-
cedures that may seem trivial to the chemist but are generally unappealing
to a cell biologist or molecular biophysicist. Mass spectrometry has been
used to analyze the caging and uncaging of peptides. Notably, characteristic
peaks are generated from the 2-nitrobenzyl group by the laser of matrix-
assisted laser desorption/ionization time-of-flight (MALDI-TOF) instru-
ments. Proteins have also been caged on cysteinyl residues. Single-cysteine
mutants are now available for many interesting proteins. Cysteine-scanning
mutagenesis can be combined with targeted chemical modification to deter-
mine which mutants will be suitable for caging. In some cases, it has been
possible to generate sufficient protein for this analysis by in vitro translation.
In our original approach, the modification reagent differed from the reagent
ultimately used for caging. However, if the caging reagent is available in
abundance, there is no reason it cannot be used directly in conjunction
with scanning mutagenesis. Caging reactions can be monitored by examin-
ing shifts in the electrophoretic mobility of proteins in extended sodium
dodecyl sulfate (SDS)-polyacrylamide gels. Natural cysteines can also be
used as targets for caging reactions. In this case, it can simplify matters to
remove irrelevant cysteines by mutagenesis. Because many proteins are
members of large families, it is likely that a caging reaction developed for
one member of the family will be applicable to others, minimizing the need
for extensive structure-function analyses. Thiophosphoryl peptides are
readily caged. Indeed, thiophosphoryl groups can be caged in the presence
of free cysteinyl residues (Fig. 1). Many cell-signaling proteins are regulated
by phosphorylation. Therefore, if this approach can be extended to proteins,
it will obviate the need for lengthy searches for suitable cysteine mutants.
SH
LRRACLGLRRASLG
S~" "NO:z - -
, ~H 9 "~-s"
LRRACLGLRRASLG LRRACLGLRRASLG
NBB / ~ NBB
s~T%°'9®-sY -"°~ ,s. ?®-s"

LRRACLGLRRASLG LRRACLGLRRASLG
FIG. 1. Selective modification of thiophosphoryl groups in the presence of cysteinyl residues.
At low pH, modification of thiophosphorylserine by 2-nitrobenzyl bromide (NBB) in a model
peptide can be achieved in the presence of a free sulfhydryl group. [From P. Pan and H.
Bayley, FEBS Lett. 405, 81 (1997).]
TABLE I
YIELDS FOR CAGING OF PEPT1DES, AND YIELDS AND QUANTUM EFF1CIENC1ES FOR
PHOTOLYSIS OF CAGED PEPTIDES a
% yield of
Peptide Reagent Caging Photolysis" q5
C-Kemptideb NBB 95 70 0.62
C-Kemptide DMNBB 95 67 0.15
C-Kemptide BNPA 95 62 0.21
Thiophosphorylkemptide NBB 95 70 0.23
Thiophosphorylkemptide DMNBB 95 55 0.06
Thiophosphorylkemptide BNPA 10 -- --
From P. Pan and H. Bayley,F E B S Lett. 4115,81 (1997).
b Kemptide is Leu-Arg-Arg-Ala-Ser-Leu-Gly.
cPhotolysiswas at pH 5.8.
S u m m a r y of Findings
Two major considerations have dictated the choice of sulfhydryl-
directed 2-nitrobenzyl reagents for derivatizing peptides and proteins. First
is the ability to obtain a high yield of caged product. The possibility of
carrying out the caging reaction in aqueous solution is especially important
for proteins. Second, the caged reagent should be efficiently photolyzed at
wavelengths that do not damage biological molecules. Therefore, a caging
group should have a high product quantum yield (q~) in the near-ultraviolet
(UV) or visible (e.g., ~b ->0.1) at neutral pH, with a useful extinction coeffi-
cient (a) (e.g., >-1000 M -I cm-1). With regard to these criteria, the
2-nitrobenzyl groups has proved satisfactory in many applications with small
molecules, and therefore it was our first choice for peptides and proteins.
Three 2-nitrobenzyl reagents have been explored for caging cysteine-
containing peptides: 2-nitrobenzyl bromide (NBB), 2-bromo-2-(2-nitrophe-
nyl)acetic acid (BNPA), and 4,5-dimethoxy-2-nitrobenzyl bromide
(DMNBB). 21 All were effective (Table I). 2-Bromo-2-(2-nitrophenyl)acetic
acid was synthesized as a highly water-soluble reagent, 17 but the model
peptide we used was released less readily when caged with BNPA than
with NBB. When DMNBB was used to produce a caged reagent with a
higher extinction in the near-UV [e.g., 2-nitrobenzyl (NB), a312 ~ 1300;
4,5-dimethoxy-2-nitrobenzyl (DMNB), ~312~ 4500], a compensating reduc-
tion in product quantum yield was observed. Therefore, the DMNB-caged
peptide was photolyzed at about the same rate as the NB-caged peptide.
21p. Pan and H. Bayley,F E B S Lett. 405, 81 (1997).

Where yet longer wavelengths (->360 nm) are required to prevent tissue
damage and NB absorption is still weaker, DMNB peptides could still be
useful. Poorly soluble reagents such as NBB and DMNBB must be added
from a concentrated stock solution in organic solvent. Nevertheless, the
final concentration of organic solvent can be as low as 2%, which is tolerated
by peptides and most proteins.
Interestingly, all three reagents have also been used to cage proteins.
In the case of a pore-forming bacterial exotoxin, a-hemolysin (aHL), BNPA
was highly effective, yielding a caged cysteine mutant [a-hemolysin mutant
Arg-104~Cys(aHL-R104C)-a-carboxynitrobenzyl(CNB)] with no residual
activity in a very sensitive assay. I7 Uncaging occurred with ~60% efficiency.
Interestingly, when NBB was reacted with aHL-R104C, only 75% inactiva-
tion was obtained, and no further decrease in activity occurred after pro-
longed incubation (unpublished data, 1994). The negative charge at residue
104 of a-hemolysin contributed by the CNB caging group must play an
important role in abolishing pore-forming activity. In contrast, NBB has
proven to be the most effective reagent for caging the catalytic (C) subunit
of cAMP-dependent protein kinase (PKA). 22 Compared with NBB, BNPA
and DMNBB were somewhat less effective at inactivating a single-cysteine
version of the C subunit, and the caged proteins obtained with these re-
agents were markedly less well activated on irradiation. = In another exam-
ple, caged heavy meromyosin was best produced by selective derivatization
of one of two natural cysteines with DMNBB. TM In this case, an alternative
reagent, 1-(bromomethyl)-2-nitro-4-benzoic acid, 9 gave a caged protein that
was not efficiently activated by light.
The efficacy of a reagent depends, then, on the nature of the target
protein. This situation might be rationalized in several ways. For example,
the rate of caging might be affected by the microenvironment of the cystei-
nyl residue (although in the cases cited, the reactions were pushed to
completion), the activity of the protein might be significant after complete
modification with one reagent but not with another, or the efficiency of
the uncaging reaction (product quantum yield: ~b) might also be affected
by the microenvironment on the protein, and in this case by quenching of
the excited state, e.g., by nearby tryptophans. In short, with a particular
protein, it is advisable to test all three 2-nitrobenzyl reagents described
here, as well as other reagents that are becoming available. Of course, the
appropriate control incubations must also be done to ensure that the protein
is not damaged by the modification or reactivation conditions. Finally, we
have reported the caging of thiophosphorylserine peptides with NBB and
DMNBB. In this case, BNPA appeared to react too slowly to be useful. 21
22 C.-Y. Chang, T. Fernandez, R. Panchal, and H. Bayley, in preparation (1998).

Go~s
Caged peptides and proteins should prove useful in several areas of

experimental biology, including cell biology and molecular biophysics. The
capacity to control the time, dose, and site of activation of caged reagents
is a tremendous advantage. For example, in cell biology, the ability to load
cells with caged signal transduction proteins by microinjection and trigger
their activity at will should be invaluable. 23 Such studies could be carried
out with cells in culture or with living tissues, such as early embryos for
studies of development. In biophysics, supramolecular assembly or enzy-
matic catalysis might be triggered with light and monitored with techniques
ranging from single-crystal X-ray diffraction to circular dichroism or infra-
red spectroscopy. The criteria for reagents for the various applications can
differ greatly. For example, in cell biology, it would be best to produce a
caged peptide or protein with no residual activity at all. Then, the photolytic
release of even 5% of the theoretical maximum activity within a cell would
be very useful. In contrast, in many biophysics experiments, 5% residual
activity might be tolerable. However, a far greater extent of uncaging would
be needed in most circumstances (e.g., 95%). The rate of release after a
short light flash might also be more important here than in a cell biology
experiment, with the important exception of fast synaptic transmission. 4,24
This rate depends on the breakdown of photogenerated intermediates and
should not be confused with the steady-state rate of production of uncaged
products during continuous photolysis, which is a function of lamp intensity.
The 2-nitrobenzyl group is generally lost fairly slowly after excitation, in
milliseconds to seconds. 1'25'26 In favorable circumstances, release in the
microsecond range occurs. 25 Protecting groups that are lost much more
quickly are under development. 27Finally, in biotherapeutics, caged peptides
and proteins might be activated in specific tissues in controlled doses, at
defined times. 28 Design criteria here must include the use of the longest
wavelengths possible to prevent tissue damage and to allow tissue penetra-
tion by light.
In our laboratory, we are especially interested in caged peptides and
proteins for signal transduction research. We have developed reagents for
23 S.-H. Wang and G. J. Augustine, Neuron 15, 755 (1995).

z4 G. P. Hess, L. Niu, and R. Wieboldt, Ann. N.Y. Acad. Sci. 7579 23 (1995).
Acad. Sci. U.S.A. 91, 8752 (1994).
26 A. Barth, K. Hauser, W. Mantele, J. E. T. Corrie, and D. R. Trentham, J. A m . Chem. Soc.
117, 10311 (1995).
27 C.-H. Park and R. S. Givens, J. A m . Chem. Soc. 119, 2453 (1997).
28 H. Bayley, F. Gasparro, and R. Edelson, Trends Pharmacol. Sci. 8, 138 (1987).
the controlled permeabilization of cells, including protease-activated29 and

metal-regulated a-hemolysins,3° as well as a caged hemolysin.I7 Caged cys-
teinyl and thiophosphoryl model peptides have been tested. 2I Therefore,
we are now in a position to make caged peptides to control protein-protein
interactions (e.g., those involving proteins that bind to serine phosphate,
such as 14-3-3 proteins 3I) and to control enzyme activity (e.g., caged kinase
pseudosubstrates32). A caged cAMP-dependent protein kinase catalytic
subunit has been developed, 22 and we hope to apply the technology we
have used to additional protein kinases, GTP (guanosine triphosphate)-
binding proteins, transcription factors, and other signaling molecules. We
are also extending the approach to thiophosphoryl proteins. The activities
of many proteins are regulated by phosphorylation, which can often be
mimicked by thiophosphorylation. Furthermore, thiophosphorylated pro-
teins are often relatively resistant to cellular phosphatases, so an uncaged
thiophosphoryl protein would persist in a cell. Because the requirements
for phosphorylation are quite specific, we can be certain that caging will
block the functional effects of thiophosphorylation and that uncaging will
restore them.
Materials
Kemptide (Leu-Arg-Arg-Ala-Ser-Leu-Gly) is purchased from Penin-
sula Laboratories (San Carlos, CA). C-Kemptide (Leu-Arg-Arg-Ala-Cys-
Leu-Gly) is synthesized at the Worcester Foundation Protein Chemistry
Facility. Kemptide CS-dimer (Leu-Arg-Arg-Ala-Cys-Leu-Gly-Leu-Arg-
Arg-Ala-Ser-Leu-Gly) was synthesized by Quality Controlled Biochemicals
(Hopkinton, MA) and further purified by high-performance liquid chroma-
tography (HPLC). Recombinant a-hemolysin R104C was purified at the
Worcester Foundation for Biomedical Research by Stephen Cheley .and
Christopher Shustak after expression in Staphylococcus aureus, a-HL-
R104C is also obtained by in vitro transcription and translation in an Esche-
richia coli $30 system. 17 A plasmid encoding the mutant C subunit of PKA,
C subunit mutants Cys-343~Ser(C343S), was kindly provided by Dr. Susan
Taylor (University of California, San Diego). Protein kinase inhibitor pep-
tide (PKIP) Affi-Gel 10 is prepared as describedY Adenosine 5'-0-(3-
thiotriphosphate) (ATP3~S) is purchased from Calbiochem-Novabiochem
29 R. G. Panchal, E. Cusack, S. Cheley, and H. Bayley, Nature Biotechnol. 14, 852 (1996).
30 M. J. Russo, H. Bayley, and M. Toner, Nature Biotechnol. 15, 278 (1997).
31 A. J. Muslin, J. W. Tanner, P. M. Allen, and A. S. Shaw, Cell 84, 889 (1996).
32 R. B. Pearson and B. E. Kemp, Methods Enzymol. 2@0, 62 (1991).
33 S. R. Olsen and M. D. Uhler, J. Biol. Chem. 264, 18662 (1989).
124 CAGEDCOMPOUNDS I71
(La Jolla, CA). 2-Bromo-2-(2-nitrophenyl)acetic acid was synthesized 17and
is now available from Molecular Probes (Eugene, OR). 2-Nitrobenzyl bro-
mide is purchased from Fluka (Ronkonkoma, NY), and DMNBB from
Aldrich (Milwaukee, WI). 4-Acetamido-4'-[(iodoacetyl)amino]stilbene
2,2'-disulfonate (IASD) and 1-(2-nitrophenyl)ethyl (NPE) phosphate (di-
ammonium salt) are from Molecular Probes. Inorganic phosphate is mea-
sured with a kit from Sigma (St. Louis, MO).
Mass Spectrometry
Mass spectrometry is done with a PerSeptive Voyager matrix-assisted
laser desorption/ionization mass spectrometer (MALDI-MS) equipped
with a 337-nm nitrogen laser (PerSeptive Biosystems, Framingham, MA).
Analyte samples in -25% (v/v) acetonitrile in 0.1% aqueous trifluoroacetic
acid (TFA) or in a low-salt buffer solution (0.6/zl) are spotted onto the
sample plate. The matrix, a-cyano-4-hydroxycinnamic acid, is then added
as a saturated solution in 25% (v/v) acetonitrile in 0.1% aqueous TFA (0.6
/zl) and the sample is dried at room temperature before analysis at a scan
rate of 5 sec -I. When 2-nitrobenzyl peptides are examined, the following
phenomena prove useful for structure confirmation21: (1) a peak at
MH ÷- 16, presumably due to loss of oxygen from the -NO2 group34; (2) a
peak at MH÷-135 corresponding to loss of the 2-nitrobenzyl group; and
(3) in the case of caged thiophosphoryl groups, a peak at MH+-151 corre-
sponding to loss of the 2-nitrobenzyl group and a further 16 mass units,
presumably due to sulfur-oxygen exchange. These additional peaks arise
through photolysis with the 337-nm laser, as they are not observed when
electrospray ionization mass spectrometry (ESI-MS) is carried out. Sulfur-
oxygen exchange does not occur when caged thiophosphoryl peptides are
photolyzed in solution. Finally, when fragmentation for mass mapping is
performed, e.g., with trypsin, the fragment carrying the 2-nitrobenzyl group
also undergoes the same photochemistry in the MALDI-MS, which aids
in the identification of the modified residue. Mass spectra that illustrate
these phenomena are included below.
Caged Cysteinyl Peptides

C-Kemptide (Leu-Arg-Arg-Ala-Cys-Leu-Gly, 0.2 mM) is reacted with
NBB, DMNBB, or BNPA in I00 mM Tris-HCl, pH 7.2, at 25 ° in the dark
for I hr. Stock solutions (50 raM) of NBB, DMNBB, and BNPA are
34H. A. Mordson, in "The Chemistry of the Nitro and Nitroso Groups" (H. Feuer, ed.)
pp. 165-213. Interscience, New York, 1969.
[7] TARGETEDCHEMICALMODIFICATION 125
prepared in, respectively, 95% ethanol, dimethylformamide, and 100 mM

Tris-HCl, pH 7.2. Final concentrations in the reaction mixture are 2 mM
NBB, 1 mM DMNBB, and 2 mM BNPA. The caged C-kemptides are
purified by HPLC with a Vydac C18 column (Vydac, Hesperia, CA) [buffer
A: 0.1% TFA in water; buffer B: 0.08% TFA in acetonitrile/water (7:3)]
(Fig. 2A). The conversion is -95% in all three cases. When examined by
MALDI-MS, the purified peptides give the expected MH ÷ peaks as well
as the laser-generated fragments described earlier (Fig. 2B). C-Kemptide
caged by BNPA appears as two peaks (both m/z 966) on HPLC, correspond-
ing to the diastereomers of ot-carboxy-2-nitrobenzyl C-kemptide.
Thiophosphorylation of Model Peptides

Kemptide (Leu-Arg-Arg-Ala-Ser-Leu-Gly, 1 rag, 1.3/xmol) or kemptide
CS-dimer (Leu-Arg-Arg-Ala-Cys-Leu-Gly-Leu-Arg-Arg-Ala-Ser-Leu-
Gly, i rag, 0.65 tzmol) is dissolved in I00 mM 3-(N-morphilino)propanesul-
fonic acid (MOPS), pH 7.0, containing 100 mM KCI and 10 mM MgCl2
(300/xl). ATP3~S in H20 (300/xl) is added in fivefold molar excess over
peptide. Thiophosphorylation is initiated by the addition of C subunit of
PKA (2/zl; final concentration of 1.0/xg/ml), followed by incubation at
room temperature for 2 hr. Thiophosphorylated peptides are purified by
HPLC, with the Vydac C18 column, and then analyzed by MALDI-MS.
MH ÷ ions are observed at m/z 869 for thiophosphorylkemptide and m/z
1638 for the kemptide CS-dimer. Purified thiophosphorylated peptides are
lyophilized and stored at -20 °.
A a B 907 923
o
c
t~
E 788
tr
1o 1'5 2'0 700 860 96o 1o'oo

Time (min) mlz
FIG. 2. HPLC purificationand MALDI-MSanalysisof C-kemptide after reaction with
NBB. (A) HPLCchromatogramof a reactionmixtureconsistingof 20/zM C-kemptideand
0.2 mM NBB after incubationfor 60 min in 100 mM Tris-HC1, pH 7.2, at 25°. Peak a is the
caged peptide. (B) MALDI-MSof purifiedNB-C-kemptidefrom (A). The MH÷ peak is at
m/z 923. Peaks at MH÷-16 and MH+-135 are also seen.
Caged Thiophosphoryl Peptides
Caged thiophosphorylkemptides are synthesized by reaction with NBB
and DMNBB under the same conditions used for caging C-kemptide. Both
derivatives are formed in -95% yield, purified by HPLC (Fig. 3A), and
analyzed by MALDI-MS, which again gives the expected MH + peaks as
well as the laser-generated fragments described earlier (Fig. 3B). No reac-
tion occurs between thiophosphorylkemptide and BNPA at neutral or basic
pH. When thiophosphorylkemptide is incubated with 2 mM BNPA at
pH 4.0, 37°, for 4 hr, a yield of - 1 0 % of the diastereomers of the caged
thiophosphorylkemptide is obtained. Because of the low yields, under rela-
tively harsh reaction conditions, these derivatives were not further investi-
gated.
Selective Reaction at Thiophosphoryl Group

Kemptide CS-dimer (0.2 mM) or thiophosphorylkemptide CS-dimer
(0.2 raM) and NBB (2 mM) are allowed to react in 100 mM Tris-HC1,
pH 7.2, at 25° in the dark for 100 min. Kemptide CS-dimer caged on the
cysteinyl residue (MALDI-MS, m/z 1677) and thiophosphorylkemptide
CS-dimer caged on both the cysteinyl residue and thiophosphoryl serine
(MALDI-MS, m/z 1909) are obtained.
The reactivities of cysteinyl and thiophosphoryl groups toward NBB
differ at low pH and permit the caging of thiophosphorylserine in the
A b IB 1004,
t=
.¢_ 869 988

._>
¢1
or
10 1'5 2'0 8()0 960 10'00 11'00

Time (min) m/z
FIG. 3. HPLC purification and MALDI-MS analysis of thiophosphorylkemptide after
reaction with NBB. (A) HPLC chromatogram of a reaction mixture containing 20 ~M thio-
phosphorylkemptide and 0.2 mM NBB after incubation for 60 min in 100 mM Tris-HC1, pH
7.2, at 25°. Peak a is unreacted peptide and peak b the caged peptide. (B) MALDI-MS of
purified NB-thiophosphorylkemptide from (A). The MH + peak is at m/z 1004. Peaks at
MH+-16, MH+-135, and MH+-151 are also seen.
A b
(-.
rr
0 ~ 1'0 l's 2'0 22 3'0

Time (min)
FIG. 4. HPLC chromatograms of reaction mixtures containing 20/zM thiophosphoryl CS-

dimer and 0.2 m M NBB incubated at pH 4.0 and pH 7.2. (A) Incubation in 100 m M sodium
acetate, pH 4.0, at 25° for 100 min. (B) Incubation in 100 m M Tris-HC1, pH 7.2, at 25° for
100 min. a, thiophosphoryl CS-dimer; b, thiophosphoryl CS-dimer caged on thiophosphate;
c, thiophosphoryl CS-dimer caged on both cysteine and thiophosphate.
presence of free cysteine in the same peptide. Presumably, the cysteinyl

residue (pKa = 9.5) is protonated under these conditions, whereas the
thiophosphoryl group is a reactive monoanion (pKa ~- 5.5). 35 Selective
caging of thiophosphorylkemptide CS-dimer (0.2 mM) with NBB (2 mM)
in 80% yield is achieved in 100 mM sodium acetate buffer, pH 4.0, at 25°
in the dark for 100 min (Fig. 4). MALDI-MS (m/z 1774) of the HPLC
purified product is consistent with only one 2-nitrobenzyl group per peptide.
The thiophosphoryl CS-dimer, caged with NBB at pH 4.0, is digested with
trypsin, and the fragments are analyzed directly by MALDI-MS (Fig. 5).
The peak at m/z 734 is consistent with the fragment RASP*LG, where Sp*
is the caged thiophosphoryl residue, and the peak at m/z 600 with its
uncaged product.
Caging at the sulfur of thiophosphoryl groups should be applicable
to native proteins provided that they are not irreversibly denatured or
precipitated at the low pH values required for selective modification. Shifts
in the pK, values of cysteinyl and thiophosphoryl groups induced by micro-
environments on a protein may facilitate or complicate selective modifica-
tion, and the reaction conditions will have to be adjusted for each case.
35 E. K. Jaffe and M. Cohn, Biochemistry 17, 652 (1978).

A B
444 444 833
¢- 5O3
379 800 789 822

g
718
71
n'- 1133 789
200 4;0 m/z 8;0 1000 200 4;0 m)z 8;0 10'00
LRRAC*LGLRRASLG LRRACLGLRRASP*LG
Trypsin ~ caged uncaged Trypsin J caged uncaged
LRR 443 LRR v 443
AC*LGLR 766 ~ 632 ACLGLR 632
RASLG 502 R A SP*L G 733 ~ 599
RAC*LGLR 923 ~ 7 8 8 RACLGLR 788
or or
A C* L G L R R 923 ~ 788 ACLGLRR 788
R RAC*LG L 923 ~ 7 8 8 RRACLGL 788
F]6.5. MALDI-MS analysisof trypsindigests. (A) NB-caged CS-dimer(nonthiophosphory-

lated) and (B) NB-caged thiophosphoryl CS-dimer (made at pH 4.0). The peptides were
digested with trypsin (1/zg/ml) in 25 mM NH4HCO3 at pH 8.9 for 10 min and then subjected
to MALDI-MS analysis. Identified tryptic fragments, including those uncaged by the laser,
are tabulatedas unprotonatedspecies. [FromP. Pan and H. Bayley, FEBS Lett. 405, 81 (1997).]
a-Hemolysin R104C Caged on Cys-104

aHL-R104C (0.36 mg/ml; stored in 10 mM sodium acetate, 150 mM
NaC1, pH 5.2) is dialyzed against 10 mM Tris-HCl, pH 8.5, containing
5mM 2-mercaptoethanol. Dialyzed R104C (100 ~1), 10 mM dithiothreitol
(DIT) in water (20/.d), and 1.0 M Tris-HC1, pH 8.5 (40/.d) are incubated
for 5 min at room temperature, prior to the addition of 100 mM BNPA in
100 mM sodium phosphate, pH 8,5 (40/.d). After 3 hr at room temperature,
1.0 M DTY (5 /~1) is added, and excess reagents are removed from the
protein by gel filtration on Bio-Gel P-2 (Bio-Rad, Richmond, CA) eluted
with 10 mM Tris-HC1, pH 8.5, containing 50 mM NaC1. About 80% conver-
sion to caged aHL-R104C is achieved as judged by sodium dodecyl sulfate-
polyacrylamide gel electrophoresis ( S D S - P A G E ) . The remaining 20% of
the protein is converted to two inactive products that migrate near unmodi-
fied aHL-R104C.
Caging o f / n Vitro Translated Protein

In many cases, sufficient protein for analysis can be obtained by caging
the product of an in vitro translation and, often, the unfractionated transla-
tion mix can be used. An Escherichia coli $30 transcription/translation
system can produce 10-50/zg protein/ml. In calculating the molar excess
[71 TARGETED CHEMICAL MODIFICATION 129
of caging reagent to be used, thiols present in the translation mix should

be taken into account. Modification of in vitro-translated protein is a partic-
ulary useful approach for scanning the properties of a collection of cysteine
mutants, before choosing one for detailed analysis or increased production.
aHL-R104C (30 /El of $30 translation mix) is reduced for 5 min at
room temperature by the addition of 10 mM DTT (15/El in water), 1.0 M
Tris-HCl, pH 8.5 (30/El), and water (45/El). BNPA (100 mM) is added in
100 mM sodium phosphate, pH 8.5 (30 tel) and the reaction is incubated
at room temperature in the dark for 9 hr. The progress of a modification
reaction can be followed by gel-shift electrophoresis (Fig. 6). Excess BNPA
and its low-mass sulfhydryl adducts are then removed by repeated cycles of
dilution with 100 mM Tris-HC1, pH 8.5, and concentration by ultrafiltration
(Amicon, Microcon 3, Millipore Corp., Bedford, MA). The final volume
(135/El) is readjusted to 150/El with 100 mM Tris-HCl containing 10 mM
DTT so that the final concentration of DTT is 1 mM. The sample is stored
at - 2 0 ° before use.
Catalytic Subunit of cAMP-Dependent Protein Kinase Caged

at Cys- 199
The Ca subunit of mouse PKA contains two cysteines: the crucial Cys-
199 and the dispensable Cys-343. We avoid possible complications with
Cys-343, such as intermolecular disulfide bond formation, by using the
mutant C343S kindly provided by the laboratory of Dr. Susan Taylor (Uni-
1 2 3 4 5 6 7 8 9 10
-.10,c-Aso
It
FIG. 6. Photolysis of aHL-R104C-CNB examined by SDS-PAGE. 35S-labeled R104C was
treated as described in the text and subjected to electrophoresis in a 12% SDS-polyacrylamide
gel. An autoradiogram of a region between the 29.5 and 45.5-kDa markers is shown. Lane
1, R104C; lane 2, R104C treated with IASD; lane 3, R104C treated with BNPA; lane 4, R104C
treated with BNPA and then further treated with IASD; lane 5, BNPA adduct of R104C
after removal of excess reagent; lane 6, as lane 5 but photolyzed at pH 6.0; lane 7, BNPA
adduct of R104C after removal of excess reagent; lane 8, as lane 7 but photolyzed at pH 8.5;
lane 9, as lane 6, but the photolysis products were further treated with IASD; lane 10, as lane
8, but the photolysis products were further treated with IASD. [From C.-Y. Chang, B. Niblack,
B. Walker, and H. Bayley, Chem. Biol. 2, 391 (1995).]
130 CAGEDCO~POUYDS 171
versity of California, San Diego). Caging the C subunit turns out to be

more difficult than the caging of aHL-R104C. Numerous sets of conditions
were tested, as documented elsewhere. 22 In this case, NBB gives the best
results. BNPA and DMNBB give similar inactivation, but poorer reactiva-
tion on photolysis.
C343S (4.3/zM in 100 mM Tris-HC1, pH 8.5, 1 mM EDTA) is incubated
with NBB (2 mM) at 25° for 45 min. NBB is added from a 50 mM stock
in ethanol, resulting in a final ethanol concentration of 4%, which is tolerated
by the enzyme. Excess NBB is removed by gel filtration. Immediately
before use, Sephadex G-25 (Sigma) is equilibrated in 20 mM Tris-HC1,
pH 8.5,200 mM KC1, degassed, and packed into a small gel filtration column
(1 ml). The reaction mix containing the modified C subunit (50 to 150/zl)
is applied and eluted with the same buffer, and fractions (50/zl) containing
C343S-NB are collected. The specific catalytic activity of the C subunit is
measured before and after reaction with NBB. 36 The residual specific activ-
ity is -8%. As judged by SDS-PAGE, the extent of modification is -95%.
The residual activity can be reduced to - 4 % by a second modification
with N-ethylmaleimide. The reaction mix is adjusted to pH 7.4, and N-
ethylmaleimide is added to a final concentration of 10 mM and allowed to
react at 25 ° for 20 min.
Unmodified C343S can be removed with PKIP coupled to Affi-Gel 10.
To treat NBB-modified C343S, the matrix is equilibrated with 2× PKIP
binding buffer [40 mM Tris-HC1, pH 7.4, 4 mM MgCI2, 2 mM ATP, 1 mM
phenylmethylsulfonylfluoride, 2 mM EDTA, 2 mM DTT, and 0.2% Nonidet
P-40 (NP-40)]. The supernatant is removed and the gel gently mixed with
an equal volume of eluant from the Sephadex G-25 column at 4° for 2 hr.
The PKIP Affi-Gel is then separated by centrifugation. The supernatant
contains the C343S-NB (86% recovery). The activity of C343S-NB is similar
before and after the PKIP Affi-Gel treatment, suggesting that the residual
activity after NBB treatment is derived from the modified C subunit and
not from traces of unmodified protein.
Photolysis of Caged Peptides and Proteins

We estimated the quantum yields for formation of uncaged product
from both peptides and proteins by comparison with the rate of photolytic
phosphate release from 1-(2-nitrophenyl)ethyl phosphate (NPE phosphate,
"caged phosphate"), for which ~b = 0.54 as determined by the ferrioxalate
actinometer method. 37,38NPE phosphate is a convenient reagent with which
36R. Roskoski,Methods Enzymol. 99, 3 (1983).

37j. H. Kaplan,B. Forbush, and J. F. Hoffman,Biochemistry 17, 1929 (1978).
38j. W. Walker, G. P. Reid, J. A. McCray,and D. R. Trentham,J. Am. Chem. Soc. 110,
7170 (1988).
to make this comparison, although a reading of the original literature 37'38

suggests that the ~b value has not been determined with certainty and is
perhaps wavelength dependent. Further, the pH at which ~bwas determined
and the pH dependence of ~b are not stated. Our experiments suggest that
the pH dependence is weak. Therefore, at the very least, the relative ~b
values obtained with NPE phosphate are of value.
For a dilute solution, the rate of photolysis of a reagent is (see, e.g.,
Bayley39):
v ~-- 2 . 3 0 3 ~ b l o e c
where v is the rate of photolysis of the reagent in M sec -I, ~bis the product
quantum yield, I0 is the incident light intensity in millimoles of photons sec-1
cm -2, e is the extinction coefficient in M -x cm -1, and c is the concentration in
M. The half-time of photolysis follows:
r -~ 0.3/4~I0e
Therefore, for low light absorption, the half-time of the reaction is inversely
proportional to the quantum yield, light intensity, and extinction coefficient,
as intuition would suggest. It is not a function of the concentration of
the reagent.
A quantum yield can therefore be obtained by using
= (~lS1T1/,S2T2
where t/~1 is a known product quantum yield such as that of NPE phosphate.
This equation applies for a single wavelength (i.e., it is fine for a laser),
so it may be necessary to correct for the number of photons absorbed over
the entire range of wavelengths emitted by a lamp and passed by a filter. It
is especially useful for caged proteins, which are generally used as relatively
dilute solutions.
If the absorbance is high, the equation does not apply. Obviously, in
this case all the incident light is absorbed by the sample and the value of
e is irrelevant. Here, the initial rate of photolysis will be a function of 4~
and independent of the concentration of reactant:
V1/V 2 = ~)l/f~2
Photolysis Conditions
Caged peptide (300 ~1, 1.6-1.8 mM) in 100 mM sodium acetate,
pH 5.8, is placed in the well of a 96-well microplate on ice. Irradiation is
39 H. Bayley, "Photogenerated Reagents in Biochemistry and Molecular Biology," Elsevier,

Amsterdam, 1983.
132 CA6ED COMVOUNOS [71
carried out with a 30-W UV lamp, which is set 1.5 cm above the plate (2200
/zW cm -2, peak emission 312 nm, Cole-Parmer, Chicago, IL). A glass filter,
providing a window at 290-380 nm (Oriel Corp., Stratford, CT), is used.
Irradiated material is removed at various time points and analyzed by
HPLC. Values of ~b are calculated by using the equation for solutions of
high absorbance (Table I). Photolysis is considerably less efficient at neutral
pH: NB-C kemptide (~bpHS.8 = 0.62; t f l p H 7 . 2 = 0.14), NB-thiophosphoryl-
kemptide (~bpH5.8 = 0.23; ~bpH72 = 0.04).
aHL-R104C-CNB in 100 mM Tris-HC1 (pH 6.0 or 8.5) containing 1 mM
DTF (60/zl) is placed in a well of a 96-well microtiter plate and irradiated
on ice through a 285-nm cutoff filter (Oriel) at 3.5 cm from a Foto UV 300
illuminator (Fotodyne, Hartland, WI). To assay for the unmasking of Cys-
104, samples are treated with IASD and analyzed by S D S - P A G E (Fig. 6).
For the sample at pH 8.5, irradiated ctHL-R104C-CNB (5/xl) is diluted
with 100 mM Tris-HC1, pH 8.5 (3/xl), and reacted with 100 mM IASD in
water (2 /xl) for 1 hr at room temperature. For the sample at pH 6.0,
irradiated otHL-R104C-CNB (5/zl) is diluted with 1.0 M Tris-HC1, pH 8.5
(2/zl), and water (1/xl) and treated in the same way.
C343S-NB collected from Sephadex G-25 is either left at pH 8.5 or
adjusted to pH 6.0 with 0.5 M sodium phosphate, pH 5.5. Solutions are
made 8.5 mM in DTT before photolysis. Samples (40/zl, 0.5/zM protein) are
placed in wells of a 96-well microtiter plate, which is put on ice. Photolysis is
performed through the 285-nm cutoff filter with a 30-W UV lamp, which
is set 2 cm above the plate (Cole-Parmer). At various time points after the
initiation of photolysis, samples are taken for kinase assay (5/xl) and IASD
modification (32/zl). The IASD is added to 20 mM, and the sample is
incubated for 20 min before analysis by S D S - P A G E (Fig. 7). The quantum
efficiency is calculated using the equation for solutions of low absorbance
((~pH6.0 = 0.84; ~bpn8.5 = 0.14).
No clear rules for estimating ~b emerge from our studies or those of
others. There is no escaping the fact that for each new caged reagent the
efficiency of photolysis and indeed the identity of the photolysis product
min 0 5 10 15
C343S-ASD
•
C343S
C343S-NB
Fro. 7. Photolysis of C343S-NB. C343S-NB was prepared as described and photolyzed at

pH 6.0 for 0, 5,10, and 15 min. The photolyzed samples were reacted with IASD to form C343S-
ASD. [From C.-Y. Chang, T. Fernandez, R. Panchal, and H. Bayley, in preparation (1998).]
I71 TARGETED CHEMICAL MODIFICATION 133
must be determined experimentally. Obviously, a high value for ~bis worth-

less if the desired product is not formed. Reagents with ¢ >-- 0.05 should
be useful for most applications, but again the photosensitivity of each
experimental system must be examined. Furthermore, ~b should be deter-
mined under conditions relevant to the application. The efficiency of photol-
ysis of 2-nitrobenzyl reagents is pH dependent and is generally worse at
neutral pH than at mildly acidic pH. In our hands, the sensitivity of ~b to
pH in the physiological range varies with the nature of the linkage (e.g.,
cysteine versus thiophosphate 21) and in proteins with the site of attach-
ment. 17 Multisubunit proteins may present an additional problem. In the
case of aHL-R104C-CNB, uncaged protein is released in -60% yield on
irradiation, but only 10-15% of the pore-forming activity is regained. The
aHL pore is a heptamer, and it is likely that improperly photolyzed subunits
have a negative effect on activity.
Future Goals
Our immediate goals include applying the reagents we have so far
developed to cellular studies. Caged peptides should present little difficulty
in this regard. However, caged proteins must be concentrated for injection
into cells and, if possible, stored in this form. For example, it is unlikely
that glycerol can be used as a cryoprotectant. With regard to caged peptides,
one of the most useful groups of reagents will be molecules caged at
thiophosphoryltyrosine. Although there are proteins that bind to phospho-
serine (e.g., 14-3-3 protein), phosphotyrosine interactions, e.g., with SH2
domains, are of paramount importance in cell signaling. In preliminary
experiments with Arg-Arg-Leu-Ile-Glu-Asp-Ala-Ile-Tyr-Ala-Ala-Arg-Gly
thiophosphorylated on Tyr with the insulin receptor kinase domain, the
thiophosphoryl peptide reacted efficiently with NBB, but photolysis of the
caged peptide was very inefficient (~b < 0.05). The production of peptides
thiophosphorylated at tyrosine has also been a problem: receptor and non-
receptor tyrosine kinases do not catalyze this reaction efficiently. In many
cases, we were unable to detect thiophosphorylation of peptides with
ATP3~S using tyrosine kinases (e.g., Abl, Src, and epidermal growth factor
(EGF) and insulin receptors) under conditions that work well when ATP
itself is used. A promising approach is to use divalent metal ions other
than Mg2÷ in the reaction (Grace et al. 4° and W. T. Miller, unpublished,
1997). Another area of attention is to test reagents other than 2-nitrobenzyl
halides for caging cysteinyl and thiophosphoryl peptides and proteins. Espe-
cially promising is the p-hydroxyphenacyl group. 27 We have tested p-hy-
40 M. R. Grace, C. T. Walsh, and P. A. Cole, Biochemistry 36, 1874 (1997).

droxyphenacyl bromide (a gift of Dr. Richard Givens, University of Kansas,
Lawrence) on cysteinyl peptides with encouraging results. This reagent
should provide caged peptides and proteins that are rapidly uncaged after
a light flash, 27 with 05 values that are less sensitive to pH (unpublished
data, 1996), and which contain no additional chiral centers (cf. BNPA).
Furthermore, the nitrosoketones and nitroaldehydes generated from 2-
nitrobenzyl caging groups are highly reactive and should be scavenged, e.g.,
by thiols. 37'38The p-hydroxyphenacyl group is converted to the harmless 4-
hydroxyphenylacetic acid after hydrolysis of a spiroketone intermediate.
Because the intermediate has the potential to react with nucleophilic groups
on biological macromolecules, further analysis of the uncaging of p-hy-
droxyphenacyl compounds is required in a biological context.
Long-term goals include the development of better methods for getting
caged peptides and proteins into cells and the synthesis of reversible photo-
activatible reagents (i.e., reagents with on-off switches). A possible alterna-
tive to microinjection is to piggyback a caged protein on another protein
capable of translocation across cell membranes. An attractive prospect is
VP22, which is a herpes virus protein with the remarkable ability to enter
living cells even when fused to quite large polypeptides such as green
fluorescent protein. 41 VP-22 could be fused with an appropriate cysteine
mutant. The molecule would be caged in vitro before incubation with target
cells. We have already developed pores that can be switched on and off
by divalent metal ions, either during assembly or as the full assembled
pore. 3°'42'43 We hope to extend this approach to photomodulated proteins
by targeted modification with photoisomerizable groups, such as azoben-
zenes or spiropyrans. 1,8 Another interesting approach to caged proteins
might be to split the protein of interest with an intervening sequence and
prevent protein splicing by caging of a conserved cysteine in the splice
junction. 44 Control of protein splicing has already been achieved by intro-
ducing a caged amino acid during in vitro translation. 45 A alternative way
of generating cysteines at defined sites is by in vitro chemical ligation of
synthetic peptides. 44'46
41 G. Elliott and P. O'Hare, Cell 88, 223 (1997).

42 B. Walker, J. Kasianowicz, M. Krishnasastry, and H. Bayley, Protein Eng. 7, 655 (1994).
43 B. Walker, O. Braha, S. Cheley, and H. Bayley, Chem. Biol. 2, 99 (1995).
44 y . Shao and S. B. H. Kent, Chem. Biol. 4~ 187 (1997).
45 S. N° Cook, W. E. Jack, X. Xiong, L. E. Danley, J. A. Ellman, P. G. Schultz, and C. J.
Noren, Angew. Chem. Int. Ed. Engl. 34, 1629 (1995).
46 p. E. Dawson, T. W. Muir, I. Clark-Lewis, and S. B. H. Kent, Science 266, 776 (1994).
[8] PHOTOCLEAVABLE AFFINITY TAGS 135
Acknowledgments
We thank John Leszyk at the W.M. Keck Protein Chemistry Facility at the Worcester
Foundation for peptide synthesis and help with HPLC and mass spectrometry. We are also
grateful to Stephen Cheley, Rekha Panchal, Christopher Shustak, and Barbara Walker for
advice and materials. This work was supported by grants from the NIH, NS26760 (to H.B.)
and CA 58530 (to W.T.M.). P.P. was supported by the NIH postdoctoral training grant T32
NS07366. The MALDI-TOF mass spectrometer was purchased through an instrumentation
grant from the NSF, BIR-9512226 (to H.B.).
[8] P h o t o c l e a v a b l e A f f i n i t y T a g s f o r I s o l a t i o n a n d
Detection of Biomolecules
By JERZY OLEJNIK, EDYTA KRZYMANSKA-OLEJNIK,
and KENNETH J. ROTHSCHILD
Introduction
The incorporation of affinity tags into biomolecules is often used in
molecular biology and medicine for isolation and detection. Isolation usu-
ally involves the attachment or incorporation of an affinity label followed
by separation through the interaction of the label with an affinity media.
Direct detection involves monitoring the tag through radioactivity, spectro-
scopically (fluorescence, mass spectrometry), or through its secondary inter-
action with another molecule (colorimetry, luminescence). However, con-
ventional methods are limited by the difficulty of removing (releasing) the
affinity label from the target biomolecule. In particular, it is often desirable
to obtain an unlabeled biomolecule after it has been isolated and/or de-
tected by a label-affinity interaction.
The use of avidin-biotin technology has proved to be an important tool
in numerous areas of biochemistry, molecular biology, and medicine, 1-3
including detection of proteins by nonradioactive immunoassays, 4'5 cyto-
chemical staining, 6 cell separation, 7 isolation of nucleic acids, detection of
1 M. Wilchek and E. A. Bayer, Methods EnzymoL 184, 1990.

2 M. Wilchek and E. A. Bayer, in "Immobilised Macromolecules" (U. B. Sleytr, ed.), pp.
51-60, Springer, London, UK, 1993.
3 M. D. Savage, G. Mattson, S. Desai, G. W. Nielander, S. Morgensen, and E. J. Conklin,
"Avidin-Biotin Chemistry: A Handbook." Pierce Chemical Company, Rockford, IL, 1992.
4 j. L. Guesdon, J. Immunol. Methods 150, 33-49 (1992).
5 T. Sano, C. L. Smith and C. R. Cantor, Science 258, 120-122 (1992).
6 A. A. Rogalski and S. J. Singer, J. Cell Biol. 101, 785-801 (1985).
7 j. Worrnmeester, F. Stiekema, and C. DeGroot, Methods Enzymol. 184, 314-319 (1990).

Acknowledgments
We thank John Leszyk at the W.M. Keck Protein Chemistry Facility at the Worcester
Foundation for peptide synthesis and help with HPLC and mass spectrometry. We are also
grateful to Stephen Cheley, Rekha Panchal, Christopher Shustak, and Barbara Walker for
advice and materials. This work was supported by grants from the NIH, NS26760 (to H.B.)
and CA 58530 (to W.T.M.). P.P. was supported by the NIH postdoctoral training grant T32
NS07366. The MALDI-TOF mass spectrometer was purchased through an instrumentation
grant from the NSF, BIR-9512226 (to H.B.).
[8] P h o t o c l e a v a b l e A f f i n i t y T a g s f o r I s o l a t i o n a n d
Detection of Biomolecules
By JERZY OLEJNIK, EDYTA KRZYMANSKA-OLEJNIK,
and KENNETH J. ROTHSCHILD
Introduction
The incorporation of affinity tags into biomolecules is often used in
molecular biology and medicine for isolation and detection. Isolation usu-
ally involves the attachment or incorporation of an affinity label followed
by separation through the interaction of the label with an affinity media.
Direct detection involves monitoring the tag through radioactivity, spectro-
scopically (fluorescence, mass spectrometry), or through its secondary inter-
action with another molecule (colorimetry, luminescence). However, con-
ventional methods are limited by the difficulty of removing (releasing) the
affinity label from the target biomolecule. In particular, it is often desirable
to obtain an unlabeled biomolecule after it has been isolated and/or de-
tected by a label-affinity interaction.
The use of avidin-biotin technology has proved to be an important tool
in numerous areas of biochemistry, molecular biology, and medicine, 1-3
including detection of proteins by nonradioactive immunoassays, 4'5 cyto-
chemical staining, 6 cell separation, 7 isolation of nucleic acids, detection of
1 M. Wilchek and E. A. Bayer, Methods EnzymoL 184, 1990.

2 M. Wilchek and E. A. Bayer, in "Immobilised Macromolecules" (U. B. Sleytr, ed.), pp.
51-60, Springer, London, UK, 1993.
3 M. D. Savage, G. Mattson, S. Desai, G. W. Nielander, S. Morgensen, and E. J. Conklin,
"Avidin-Biotin Chemistry: A Handbook." Pierce Chemical Company, Rockford, IL, 1992.
4 j. L. Guesdon, J. Immunol. Methods 150, 33-49 (1992).
5 T. Sano, C. L. Smith and C. R. Cantor, Science 258, 120-122 (1992).
6 A. A. Rogalski and S. J. Singer, J. Cell Biol. 101, 785-801 (1985).
7 j. Worrnmeester, F. Stiekema, and C. DeGroot, Methods Enzymol. 184, 314-319 (1990).

specific D N A / R N A sequences by hybridization, 8 and probing conforma-

tional changes in ion channels. 9 The technique derives its usefulness from
the extremely high affinity of the avidin-biotin interaction (K = 1015 M -1)
and the ability to biotinylate a wide range of target biomolecules such as
antibodies, nucleic acids, and lipids.
The first step in the isolation of a target molecule is its biotinylation or
the biotinylation of an agent that binds the target molecule (e.g., an antibody
or hybridization probe). This biotinylation can be achieved by reaction of an
appropriate reagent that is most often directed at either amino or sulfhydryl
groups on the target. The biotinylated molecule or target complex is then
separated from other molecules in a heterogeneous mixture using affinity
media that rely on the avidin-biotin interaction. 1 Once the target molecule
is bound through the avidin-biotin interaction, it is often important to
recover it in an unmodified and biologically functional form.
In the case of DNA, a biotin moiety can be introduced (i) during its
solid support synthesis using biotinyl phosphoramidite; 1°-14 (ii) enzy-
matically with biotinylated nucleoside triphosphate analogs, such as biotin-
ll-dUTP; 15 or (iii) through postsynthetic modifications with suitable bio-
tinylation reagents such as biotin-NHS ester, biotin hydrazide, 8'16 or
photobiotin, t7 In the case of proteins, a biotin moiety can be introduced
by direct biotinylated or through in vitro translation using, for example,
biotinylated lysyl-tRNAJ 8
Existing methods for the release of the target include (i) dissociation
of the biotin-avidin complex (6-8 M guanidinium hydrochloride), 3 (ii)
utilization of biotin analogs with decreased affinity toward avidin (e.g.,
iminobiotin), 19 and (iii) the use of biotin derivatives with chemically cleav-
8 j. L. McInnes and R. H. Symons, in "Nucleic Acid Probes" (R. H. Symons ed.), pp. 33-80,
CRC Press, Inc., Boca Raton, FL, 1989.
9 S. L. Slatin, X. Q. Qiu, K. S. Jakes, and A. Finkelstein, Nature 371, 158-161 (1994).
10 R. T. Pon, Tetrahedron Lett. 32, 1715-1718 (1990).
n U. Pieles, B. S. Sproat, and G. M. Lamm, Nucleic Acids Res. 18, 4355-4360 (1990).
12 K. Misiura, I. Durrant, M. R. Evans, and M.J. Gait, Nucleic Acids Res. 18, 4345-4354 (1990).
13 A. M. Alves, D. Holland, and M. D. Edge, Tetrahedron Lett. 30, 3089-3092 (1989).
14 a . J. COCUZZa,Tetrahedron Lett. 30, 6287-6290 (1989).
15 p. R. Langer, A. A. Waldrop, and D. C. Ward, Proc. Natl. Acad. Sci. U.S.A. 78, 6633-
6637 (1981).
16 p. Tijssen, "Laboratory Techniques in Biochemistry and Molecular Biology: Hybridization
with Nucleic Acid Probes." Elsevier, New York, 1993.
17 A. C. Forster, J. L. McInnes, D. C. Skingle, and R. H. Symons, Nucleic Acids Res. 13,
745-761 (1985).
18T. V. Kurzchalia, M. Wiedmann, H. Breter, W. Zimmermann, E. Bauschke, and T. A.
Rapoport, Eur. J. Biochem. 172, 663-668 (1988).
19 K. Hoffman, S. W. Wood, C. C. Brinton, J. A. Montibeller, and F. M. Finn, Proc. Natl.
Acad. Sci. U.S.A 77, 4666-4668 (1980).
181 PHOTOCLEAVABLE AFFINITY TAGS 137
a b
DNA purification Protein purification
~c, ~h~? ~ hv
CH--O--P--O NO2
Ilag CHf--kl ))'--NO= / ~1
C Cell isolation d Biomolecular detection
[h.llll "I'1
FIG. 1. Selected applications of photocleavable tags in biotechnology: (a) removal of
failure sequences from synthetic DNA, (b) isolation of protein from heterologous mixture,
(c) isolation of cells displaying specific surface antigen, and (d) spatially addressable photore-
lease and detection by mass spectrometry.
able spacer arms. 2°-26 However, none of these methods allows release of
the target molecule or target complex in an unaltered form because the
biotin or part of the cleavable spacer arm remains attached. In addition,
they all require harsh reaction conditions that can denature the target bio-
molecule.
As described in this article, the use of photocleavable tags (PC tags) in
the isolation/detection of biomolecules eliminates many of the difficulties
associated with conventional affinity labels by providing a rapid and effec-
tive method of removing the label in a single step. As illustrated in Fig. 1,
PC tags offer advantages in numerous applications involving the isolation
of proteins, nucleic acids, lipids, cells, and biomolecular assemblies. More
20 M. Shimkus, J. Levy, and T. Herman, Proc. Natl. Acad. Sci. U.S.A. 82, 2593-2597 (1985).
21 T. Herman, E. Lefever, and M. Shimkus, AnaL Biochem. 156, 48-55 (1986).
22 B. A. Dawson, T. Herman, and J. Lough, J. Biol. Chem. 264, 12830-12837 (1989).
23 B. A. Dawson, T. Herman, A. L. Haas, and J. Lough, J. Cell, Biochem. 46, 166-173 (1991).
24 G. m. Soukup, R. L. Cerny, and L. J. Maher III, Bioconj. Chem. 6, 135-138 (1995).
25 B. D. Gildea, J. M. Coull, and H. Koster, Tetrahedron Len. 31, 7095-7098 (1990).
26 E. Leikauf, F. Barnekow, and H. Koster, Tetrahedon 51, 3793-3802 (1995).
NO2 1
Tag Spacer Photocleavable Reactive

linker group
I~ O(CH2)2CN
X = halogen, - - O - - C - - O - - N H S , --O--P \
N(iPr)2
FIG. 2. General design of photocleavable tags based on a-substituted (R = alkyl, aryl)
2-nitrobenzyl moiety.
advanced applications include the selective release of biomolecules from

a two-dimensional array and the construction of biomolecular assemblies
at the nanometer scale.
Design of PhotocleavableAffinity Tags

The general design of photocleavable tags is shown in Fig. 2. These
reagents consist of an affinity tag (T) (biotin or specific reactive group),
which is capable of strong specific interaction or chemical reaction with an
appropriate support. The tag is linked through a variable-length spacer
arm to the photoreactive moiety (photocleavable linker), in this case
2-nitrobenzyl derivative. The photoreactive moiety can be derivatized with
a variety of reactive groups (X), which provide for selective reactivity toward
a particular chemical group on the target biomolecule. Examples of the
reactive groups include halogen (sulfhydryl reactive), N-hydroxysuccini-
mide (NHS)-carbonate (amine reactive), and N,N-diisopropyl-2-cyanoethyl
phosphoramidite (reactive toward hydroxyl groups).
We have introduced a new class of biotin derivatives, termed photo-
cleavable biotins (PC-biotins), which eliminate many of the limitations of
conventional biotins. 27,28 These reagents consist of a biotinyl moiety linked
through a spacer arm (5-aminocaproic acid) to an a-substituted 2-nitroben-
zyl nucleus that bears a reactive group. Reactivity of these compounds
depends on the nature of the reactive group and can be directed toward
either primary amino groups (PC-biotin NHS carbonate; compound 8,
27j. OleJnik, S. Sonar, E. Krzymanska-Olejnik, and K. J. Rothschild, Proc. Natl. Acad. Sci.
U.S.A. 92, 7590-7594 (1995).
28 j. Olejnik, E. Krzymanska-Olejnik, and K. J. Rothschild, Nucleic Acids Res. 24, 361-366
(1996).
Scheme 1) or hydroxyl group (PC-biotin phosphoramidite; compound 11,

Scheme 2) on the target biomolecule. The spacer is designed to provide
sufficient distance between the target molecule and the biotinyl moiety in
order to allow for effective binding to streptavidin. The a-substituted 2-
nitrobenzyl moiety was selected because its derivatives, such as esters and
amides, are known to exhibit highly efficient and fast photoreactions. 29-31
The NHS esters and carbonates react selectively under mild conditions
(pH 8) with primary amino groups to give amides and carbamates, respec-
tively. 32,33
PC-Biotin phosphoramidite (compound 11, Scheme 2) was designed for
direct use in any automated D N A / R N A synthesizer employing standard
phosphoramidite chemistry. PC-biotin phosphoramidite contains a biotinyl
moiety that bears a 4,4'-dimethoxytrityl group on an N-1 nitrogen atom.
This group enables monitoring of coupling efficiency by trityl cation assay.
In this case, the 2-nitrobenzyl group is derivatized with N,N'-diisopropyl-
2-cyanoethyl phosphoramidite--a reagent compatible with state of the art
automated nucleic acid synthesis. The selective reaction of PC-biotin phos-
phoramidite with the free 5'-OH group of a full-length oligonucleotide
chain results in the introduction of a phosphodiester group linked to a
photocleavable biotin moiety.
Materials and Methods

All chemicals are purchased from Aldrich (Milwaukee, WI), unless
otherwise indicated. 1H NMR spectra are recorded on a Varian XL-400 or
Varian Unity Plus (Palo Alto, CA) spectrometer at 400 MHz in deuterated
DMSO or CDC13 solution with chemical shifts (8, ppm) reported relative
to a tetramethylsilane internal standard. 31p nuclear magnetic resonance
(NMR) spectra are recorded in CDCI3 on a JEOL JNM-GSX270 spectrome-
ter (Peabody, MA) at 109.36 MHz with chemical shifts (8, ppm) reported
relative to an 85% H3PO4 external standard. Emission spectra are measured
on a SLM 48000 fluorimeter (Rochester, NY) using 380 nm excitation.
Infrared spectra are recorded on a Nicolet 740 FTIR spectrometer (Madi-
29 j. E. T. Corrie, J. Chem. Soc. Perkin Trans. I, 2161-2166 (1993).

W. W. Webb, J. Marque, W. Denk, J. A. MeCray, and G. P. Hess, Biochemistry 28,
49-55 (1989).
31 j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. 110,
7170-7177 (1988).
32 M. Wilchek and E. A. Bayer, Methods Enzymol. 184, 123-138 (1990).
33 A. K. Ghosh, T. T. Duong, S. P. McKee, and W. J. Thompson, Tetrahedron Lett. 33,
2781-2784 (1992).
son, WI). Oligonucleotide synthesis is performed on an Applied Biosystem

D N A / R N A synthesizer (Model 392; Foster City, CA). Samples are irradi-
ated with a Blak Ray XX-15 UV lamp (Ultraviolet Products Inc., San
Gabriel, CA) at a distance of 15 cm (emission peak 365 nm; 300-nm cutoff;
1.1 mW intensity at 31 cm). UV-VIS spectra are recorded on a Shimadzu
(Columbia, MD) 2101PC spectrophotometer. High-performance liquid
chromatography (HPLC) analysis is performed on a Waters (Milford, MA)
system consisting of a U6K injector, 600 Controller, Novapak C18 (3.9 x
150 mm) column, and a 996 photodiode array detector. For oligonucleotide
analysis, the following conditions are used: buffer A, 0.1N triethylamine
acetate, pH 6.0; buffer B, acetonitrile. Elution is performed using a linear
gradient (8-45%) of buffer B in buffer A over 45 min at a 1-ml/min flow.
Preparative purifications are achieved on a Waters Novapak C18 RCM
cartridge (8 × 100 mm) using conditions as specified above, except for flow
rate, which is increased to 2 ml/min. Fractions are then analyzed, pooled,
and freeze-dried. For Leu-Enk experiments, buffer A is 0.1% trifluoroacetic
acid (TFA) in water, buffer B is 0.1% TFA in 3 : 1 (v/v) acetonitrile/water.
Elution is performed using a linear gradient (30-45%) of B over 10 min
followed by 45% of B isocratic for 10 min. Reagents and conjugates should
be protected from bright sunshine but are found to be stable in artificial
light.
Synthesis of Photocleavable Biotin Derivatives
Photocleavable Biotin N-hydroxysuccinimide Carbonate: Scheme 1. 5-
Methyl-2-nitrobenzoic acid (1, Scheme 1) (5 g, 27.6 mmol) is added in small
portions to thionyl chloride (16.4 g, 138 mmol). After 10 hr of stirring at
room temperature, excess thionyl chloride is evaporated to give the acid
chloride (2). Magnesium turnings (1.07 g, 44.2 mmol), absolute ethanol (6
ml), chlorobenzene (8 ml), and 0.1 ml of dry CCL4 are refluxed for 2 hr.
Diethyl malonate (4.82 g, 30 mmol in 10 ml of chlorobenzene) is added,
followed by the addition of the acid chloride (2) (5.49 g, 27.5 mmol in 10
ml of chlorobenzene). The reaction mixture is stirred for 1 hr, acidified
with 20 ml of 2 N sulfuric acid, extracted with chloroform (3 x 20 ml),
dried, and evaporated to dryness. The residue is dissolved in acetic acid
(8.25 ml); 5.4 ml of H20 and 1 ml of concentrated H2SOn are added and
the mixture is refluxed for 6 hr. The reaction mixture is neutralized with
aqueous Na2CO3 and extracted with CHC13 (3 x 10 ml). Chloroform ex-
tracts are combined, dried, and after removing the solvents in vacuo, the
residue is crystallized from 95% ethanol to give 5-methyl-2-nitroacetophe-
none (3) (4.46 g, 81%). 34Compound (3) (3.51 g, 19.6 mmol), N-bromosuccin-
34 L. M. Long and H. D. Troutman, J. Am. Chem. Soc. 71, 2473-2475 (1949).

COOH COCl COCH3

O2N'~cH3 ii, •
CH3 CH
1 2 3
(~OCH3 COCH3
02N,~ 4) i, 02N,,~ 5) i,
CH2Br ""~v/'~'CH2NI-~3
Cle
4 5
O
o o / coc.
"-S~-(CH~),-C--N.--(CH~IrC--NH--CH,
~/'--NO,
HNLH 6)1 H3C

CH--OH
~.~ OII OII ,/~
(CH2)4-'C--NH--(CH2)5--C--NH--CH2~-~NO2
HIgH CH O NH CH ; NH H ~H-N O/C-O'-NI~

\s~( =),- - -( =)~- - -c ~~ - ~ _ ~ / ~ - o~
SCHEME 1. (11 SOCI2; (2) (i)Mg[CH(COOEt)2]2, (ii) H+/H20, A; (3) NBS/peroxide; (4)
(i) hexamethylenetetramine, (ii) HC1/ethanol; (5) biotinamidocaproic acid/DCC/Et3N; (6)
NaBH4; (7) N,N'-disuccinimidyl carbonate/Et3N.
imide (3.66 g, 20.6 mmol), and benzoyl peroxide (46 mg, 0.01 equivalents)
are refluxed in 20 ml of CC14 for 5 hr. The precipitate is filtered off, filtrate
evaporated, and the residue crystallized from CC14 to give 5-bromomethyl-
2-nitroacetophenone (4) (3.64 g, 72%). 35,36Compound (4) (2.0 g, 7.75 mmol)
is added to a solution of hexamethylenetetramine (1.14 g, 8.13 mmol) in
15 ml of chlorobenzene. The mixture is stirred overnight, and the resulting
precipitate of 5-aminomethyl-2-nitroacetophenone hydrochloride (5) is fil-
tered off and washed with chlorobenzene (10 ml) and diethyl ether (20
ml). The precipitate (2.93 g, 7.36 mmol) is suspended in 39 ml of chloroben-
zene and 19.5 ml of methanol. To this suspension 4.9 ml of concentrated
HC1 is added and the mixture is stirred overnight. Diethyl ether, 70 ml, is
35 p. D. Senter, M. J. Tansey, J. M. Lambert, and W. A. Bl~ittler, Photochem. Photobiol. 42,

231-237 (1988).
36T. Doppler, H. Schmid, and H. J. Hansen, Helv. Chim. Acta 62, 291-303 (1979).
then added, and the resulting precipitate is washed with 2 × 50 ml of ether

and dried under vacuum. Dimethylformamide (10 ml) is added followed
by the addition of a 5-biotinamidocaproic acid (3.29 g, 1.25 equivalents,
in 35 ml of N-methylpyrrolidone), dicyclohexylcarbodiimide (2.28 g, 1.5
equivalents), and triethylamine (1.28 ml, 1.25 equivalents). The solution is
stirred overnight at room temperature, the precipitate is filtered off, and
the filtrate is added to 700 ml of diethyl ether. The resulting precipitate is
dried and purified on a silica gel column using step gradient (5-20%) of
methanol in CHC13 to give 5-(5-biotinamidocaproamidomethyl)-2-nitro-
acetophenone (6) (2.27 g, 58%).
Intermediate 6 (1 g, 1.87 mmol) is dissolved in 15 ml of 70% ethanol,
and after cooling the solution to 0°, sodium borohydride (141 mg, 4 equiva-
lents) is added. The solution is stirred at 0° for 30 min and at room tempera-
ture for an additional 2 hr. The reaction is quenched by the addition of
1 ml acetone, neutralized with 0.1 N HC1, and concentrated to about 5 ml.
The aqueous layer is decanted, and the residue is washed with water (3 x
5 ml) and dried to give 5-(5-biotinamidocaproamidomethyl)-l-(2-nitrophe-
nyl)ethanol (7) (0.71 g, 71%).
Compound 7 (1.07 g, 2 mmol) is dissolved in 10 ml dimethylformamide.
To this solution N,N'-disuccinimidyl carbonate (Fluka, Ronkonkoma, NY)
(1 g, 1.5 equivalents) is added followed by triethylamine (0.81 ml, 3 equiva-
lents). After 5 hr of stirring at room temperature, solvents are evaporated
to dryness, and the residue is applied to a silica gel column and purified
using a step gradient of methanol in chloroform to give 1.04 g (69%) of
5-(5-biotinamidocaproamidomethyl)-l-(2-nitrophenyl)ethyl-N-hydroxy-
succinimidyl carbonate (PC-biotin-NHS) ester (8), which is characterized
as follows: mp 113-114 ° (uncorrected); CI-MS (M + = 676.5); UV-VIS (in
phosphate buffer, pH 7.4): /~1 = 2.04 nm, el = 19190 M 1 c m - 1 ; A2 = 272
nm, ~32 = 6350 M -1 cm-1; 1H NMR (3; ppm): 8.48 (t, 1H), 8.05-8.03 (d,
1H), 7.75-7.71 (t, 1H), 7.66 (s, 1H), 7.46-7.45 (d, 1H), 6.44 (s, 1H), 6.37
(s, 1H), 6.28-6.27 (m, 1H), 4.39 (m, 2H), 4.30 (m, 1H), 4.12 (m, 1H), 3.57
(d, 2H), 3.09 (m, 1H), 3.01-2.99 (m, 2H), 2.79 (m, 5H), 2.58-2.55 (d, 1H),
2.17-2.15 (m, 2H), 2.04-2.02 (m, 2H), 1.72-1.71 (m, 2H), 1.66-1.43 (m, br,
6H), 1.38-1.36 (m, br, 2H), 1.26-1.25 (m, br, 3H); IR (KBr): Uc=o 1815
and 1790 cm -1.
Photocleavable Biotin Phosphoramidite: Scheme 2. 5-(6-Biotinamido-
caproamidomethyl)-2-nitroacetophenone (6) (0.5 g, 0.94 mmol) is dried by
coevaporation with anhydrous pyridine (3 × 2 ml) and then dissolved in
5 ml of the latter. To this solution 4,4'-dimethoxytrityl chloride (DMTr-
C1) (0.634 g, 1.87 mmol) is added, followed by 4-dimethylaminopyridine
(0.006 g, 0.046 mmol). The reaction mixture is stirred at room temperature
for 5 hr and then an additional 0.317 g of DMTr-C1 is added. After 24 hr,
the reaction is quenched with methanol (1 ml), poured into 100 ml of 0.1 M
[8] PHOTOCLEAVABLEAFFINITY TAGS 143
O
U
o o coc.3
\S'/;(C H2)4-C--NH--(CH2)s--C--NH--CH2~ - - ~ N O 2
9 Ill
DMTr~I~I~/H O O COCH3
,, ,,
\S/~(CH2)4-C--NH--(CH2)s--C--NH--CH2~ _ ~ N O 2
o o
\S/~(CH2)4 C NH (CH2)s C NH CH;~ - ~ N O 2
.O, / 3) (i-Pr)2N
DMTr~H ¥ H3%H_o/~P--O(CH2)2CN
o o
\S ./~(CH2)4-C--NH--(CH2)5--C--NH--CH2~NO2
11
SCHEME 2. (1) DMTr-C1/DMAP; (2) NaBH4; (3) 2-cyanoethoxy-N,N-diisopropyl-
chlorophosphine/DIPEA. [Adapted from J. Olejnik, E. Krzymanska-Olejnik, and K. J.
Rothschild, Nucleic Acids Res. 24, 361-366 (1996).]
sodium bicarbonate, and extracted with dichloromethane (3 × 50 ml).

Evaporation of combined extracts gives yellow oil, which is further purified
on a silica gel column using a step gradient of methanol in dichloromethane/
0.2% triethylamine. Appropriate fractions are pooled and evaporated to
give compound 9 as a white foam (0.73 g, 93% yield).
1-N-(4,4'-Dimethoxytrityl)-5-(6-biotinamidocaproamidomethyl)-2-ni-
troacetophenone (compound 9, Scheme 2) (0.85 g, 1.016 mmol) is dissolved
in 7 ml of ethanol, and sodium borohydride (0.028 g, 0.74 mmol) is added
with stirring. After 1 hr the reaction is quenched with 4 ml of acetone and
evaporated under reduced pressure to give a yellow oil, which is redissolved
in 10 ml of methanol, and the solution is added to 120 ml of water. Precipitate
is isolated by centrifugation (7000 rpm, 45 min) and dried in v a c u o over
K O H to give compound 10 (0.7 g 82%).
1-N-(4,4'-Dimethoxytrityl)-5-(6-biotinamidocaproamidomethyl)-l-(2-
nitrophenyl)ethanol (compound 10) (0.186 g, 0.22 mmol) is placed in an
oven-dried flask with a magnetic stirring bar, sealed with a septum, and
dried for at least 6 hr in vacuo. Anhydrous acetonitrile (0.003% water) (1
ml) is added through septum under argon. Subsequently, N,N-diidopropyl-

ethylamine (0.15 ml, 0.88 mmol) is added followed by 2-cyanoethoxy-N,N-
diisopropylchlorophosphine (0.052 g, 0.22 mmol). After 1 hr another 0.5
equivalent of the phosphine is added. After an additional 2 hr at room
temperature the reaction mixture is treated with 0.3 ml of ethyl acetate,
followed by a saturated saline solution (10 ml), and extracted with dichloro-
methane (3 × 10 ml). The organic layer is washed with water, dried over
sodium sulfate, evaporated under reduced pressure, and purified on a silica
gel column using a step gradient (0-3%) of triethylamine in acetonitrile.
Appropriate fractions are pooled and evaporated to give compound 11
as a white foam (0.144 g, 62% yield). Thin-layer chromatography (TLC),
CH3CN:(C2H5)3N, 95:5, v/v; Rf = 0.48. 1H NMR (ppm): 7.79-7.32 (m,
1H), 7.65-7.61 (m, 1H), 7.25-7.19 (m,5H), 7.13-7.05(m, 4H), 6.91-6.85 (m,
1H), 6.75-6.73 (m, 4H), 5.75-5.66 (br s, 1H), 5.54-5.43 (m, 1H), 5.22s,
5.12d (1H), 4.38-4.26 (m, 3H), 4.23-4.11 (m, 2H), 3.88-3.77 (m, 1H), 3.73
(s, 1H), 3.66-3.54 (m, 2H), 3.46-3.37 (m, 1H), 3.29-3.21 (m, 1H), 3.10-3.02
(m, 2H), 2.65-2.60 (m, 1H), 2.54-2.44 (m, 1H), 2.40-2.32 (m, 1H), 2.26-2.20
(dd, 1H), 2.10-2.11 (app. t, 1H), 2.08-2.01 (m, 2H), 1.62-1.58 (m, 6H),
1.55-1.45 (m, 4H), 1.39-1.35 (t, 2H), 1.31-1.27 (t, 2H), 1.16-1.07 (m, 9H),
0.87-0.83 (dd, 3H). 31p NMR (ppm): 146.7, 147.9. Analysis calculated for
C52H72N709S1Pl: C 62.32%, H 7.24%, N 9.78%; found: C 62.03%, H 6.79%,
N 9.20%.
Preparation of Photocleavable-Biotin Conjugates and

Photocleavage Reaction
Reaction of Photocleavable-Biotin-N-hydroxysuccinimide with Leucine-
Enkephalin and Photocleavage. Leucine-enkephalin (Leu-Enk) (Sigma
Chemical Co., St. Louis, MO) (200/xl, 15.5 tzmol/ml in 0.1 N NaHCO3,
pH 8.0) and PC-Biotin-NHS ester (200/zl, 17/zmol/ml in dimethylform-
amide) are mixed and stirred overnight at room temperature and used
without further purification. PC-Biotin-Leu-Enk (1.93 tzmol/ml in 25 mM
phosphate buffer, pH 7.4) is irradiated for 10 min, with aliquots being
withdrawn after 1, 2, 3, 4, 5, and 10 min.
Interaction of Photocleavable-Biotin-Leu-Enk with Immobilized A vidin
or Streptavidin. PC-Biotin-Leu-Enk (10 nmol) is added to a suspension of
monomeric avidin or streptavid.in (tetrameric) agarose beads (15 nmol)
and incubated with gentle agitation for 30 min. The suspension is then spin-
filtered using the Ultrafree MC filter unit, 0.22 tzm (Millipore, Bedford,
MA), for 3 min at 5000 rpm, and filtrate containing unbound PC-biotin-
Leu Enk is subjected to photolysis as described earlier. The released free
Leu-Enk is assayed using fluorescamine (see below). The residue containing
PC-biotin-Leu-Enk complexed with avidin or streptavidin agarose beads
[81 PHOTOCLEAVABLE AFFINITY TAGS 145
is resuspended in phosphate buffer (1 ml) and irradiated as described

earlier. The released Leu-Enk is assayed using fluorescamine.
Fluorescamine Assay. Fluorescamine solution (10/zl, 10 mg/ml in dry
acetone) is added to a solution of Leu-Enk (final volume 3 ml, 10-125
nmol in borate buffer; pH 9.0) and is incubated for 10 min. Fluorescence
is measured at 488 nm after excitation at 383 nm. The assay is found to be
linear in the concentration range studied. Time dependence for photocleav-
age of PC-biotin-Leu-Enk is studied by removing small aliquots from the
photoreaction at various times of irradiation and subsequent determination
of the released Leu-Enk.
5'-Photocleavable-Biotin-Oligonucleotide Synthesis. The PC-biotin
phosphoramidite (11), 0.1 M solution in anhydrous acetonitrile, is attached
to the extra port of the Applied Biosystem 392 DNA/RNA synthesizer.
The syntheses are carried out at a 0.2-~mol scale using cyanoethyl phos-
phoramidites. For the last coupling (introduction of 11) the coupling time
is increased to 120 sec as recommended for conventional biotin phosphora-
midite) 7 Typical coupling efficiency (as determined by trityl cation conduc-
tance) is between 95 and 97%. Standard detritylation as well as cleavage and
deprotection procedures are used. Control 5'-phosphorylated sequences
are synthesized using the chemical phosphorylation reagent Phosphalink
(Applied Biosystem, Foster City, CA) according to the manufacturer's in-
structions. 38
Affinity Purification and Photocleavage. Crude 5'-PC-biotin-oligonucle-
otide (16 nmol) is added to a suspension of streptavidin-agarose beads (700
ttl, 24 nmol) (Sigma, St. Louis, MO), and the suspension is incubated at
room temperature for 1 hr. It is then spin-filtered (5 min, 5000 rpm) using
a 0.22-/xm Ultrafree MC filter (Millipore, Bedford, MA). Beads on the
filter are washed with 100/xl of phosphate buffer (pH 7.2) and spin-filtered
(three times). Finally, the beads are resuspended in 700/zl of phosphate
buffer and irradiated for 5 min. After irradiation the suspension is spin-
filtered, the beads are washed with phosphate buffer (3 × 100/zl), and the
combined filtrate volume is adjusted to 1 ml and analyzed by UV absorption
spectroscopy or HPLC.
Time Dependence of Photocleavage. In order to calculate the time de-
pendence of the photocleavage, HPLC-purified 5'-PC-biotin-(dT)7 (48
nmol) is incubated with 1.5 equivalents of streptavidin-agarose beads for
1 hr. The beads are spin-filtered, washed, resuspended in phosphate buffer
(pH 7.2), and irradiated. Aliquots (200/zl each) are withdrawn from 0,
0.25, 0.5, 1, 2, 4, 6, and 10 min of irradiation, spin-filtered, and washed as
described earlier. The filtrate volume is adjusted to 1 ml and the absorbance
37 Applied Biosystems, User Bulletin No. 70, 1992.

38 Applied Biosystems, User Bulletin No. 86, 1994.
146 CAGEDCOMPOUNDS [8l
at 260 nm measured. A sample of 700/A of streptavidin, which has not
been incubated with oligonucleotide, is spin-filtered and the UV absorption
measured, serving as background. A similar measurement is made on a
sample of oligonucleotide not incubated with streptavidin (16 nmol, 700
~1 phosphate buffer, serving as 100% control). The molar extinction coeffi-
cient at 260 nm for the PC-biotin moiety is determined separately (4700
M -1 cm a), and this value is subtracted from the estimated (assuming molar
extinction coefficient equal to 12,000 for each dT) molar extinction coeffi-
cient of 5'-PC-biotin-(dT)7 (88,700) for photorelease efficiency calculations.
In order to determine the time course for photocleavage in solution, (dT)7-
5'-PC-biotin (1 0D260)is dissolved in 1 ml of phosphate buffer and irradi-
ated at 300-350 nm. Aliquots (10 ~1) are withdrawn after 0, 0.25, 0.5, 1, 2,
4, 6, and 10 min of irradiation and injected onto an HPLC column. The
percentage conversion is calculated from the ratio of the area of the particu-
lar peak [i.e., 5'-PC-biotin-(dT)7 or 5'-p-(dT)7 ] over the sum of the areas
of the component peaks, molar extinction coefficients of the components
adjusted as described earlier.
Evaluation of Photocleavable Biotins
Evaluation of Photocleavable-Biotin Peptides

The properties of the PC-biotin-NHS ester (8) have been studied using
leucine-enkephalin (Leu-Enk), a pentapeptide (Tyr-Gly-Gly-Phe-Leu) as
a model substrate. First, PC-biotin-NHS ester is reacted with Leu-Enk to
yield PC-biotin-Leu-Enk. High-performance liquid chromatography traces
of Leu-Enk (trace a, Fig. 3), the PC-biotin-NHS ester (trace b), and the
PC-biotin-Leu-Enk (trace c) are shown in Fig. 3. It can be seen from trace
c that the formation of PC-biotin-Leu-Enk is complete because the peaks
of Leu-Enk (trace a) and PC-biotin-NHS ester (trace b) are absent. Instead,
two new peaks (peaks 1 and 2) are seen, which have identical U V - V I S
absorbance spectra (Fig. 3). These two peaks can be attributed to PC-
biotin-Leu-Enk, with the separation most likely due to the presence of
chiral center in the reagent.
Figure 3 also shows that illumination of PC-biotin-Leu-Enk (A > 300
nm) results in almost complete photolysis in less than 5 min. Along with
the decrease in intensity of peaks 1 and 2, associated with PC-biotin-Leu-
Enk, there is a parallel appearance of a photoproduct, as indicated by the
increase in peak 3 (traces d and e, Fig. 3). The identical retention time of
this peak (3) and the Leu-Enk control (peak 4) establish this photolysis
product as Leu-Enk. Thus, we conclude that Leu-Enk is photoreleased
from PC-biotin-Leu-Enk in less than 5 min of illumination in a completely
[8] PHOTOCLEAVABLEAFFINITYTAGS 147
~.._~ ~ L e u - E n k (a)
t"q
t"q
0
~PC-Biotin-NHS ester (b)
/-,
1"-~ I~2
< l Photolysisof
PC-Biotin-Leu-Enk
J3
A .... 0~ (ol
1:~ ....... 1.5rain/a)
I:~ ; ,, - - 5 rain (e)
I I I I I I I I I
0 10 20
Retention time [min]
Fro. 3. Reaction of PC-biotin-HS with leucine-enkephalin (Leu-Enk) and its p h o t o l y s i s
analyzed by HPLC. 27Trace a, Leu-Enk; trace b, PC-biotin-NHS ester; traces c, d, and e, PC-
biotin-Leu-Enk irradiated for 0, 1.5, and 5 min, r e s p e c t i v e l y .
unaltered form. Other peaks (labeled as asterisks) that appear during pho-
tolysis are most likely due to the 2-nitrosoacetophenone derivative and
other minor photolysis products that have also been observed in the photol-
ysis of a-substituted 2-nitrobenzyl compounds. 39'4°
The time dependence for photocleavage of PC-biotin-Leu-Enk in solu-
tion is also measured using fluorescamine, which reacts only with the free
N-terminal amino group of Leu-Enk to form a fluorophore ('~ex = 383 nm,
/~em = 480 m) 41 and not with PC-biotin-Leu-Enk, which lacks any free
39T. Milburn, N. Matsubara, A. P. Billington, J. B. Udgaonkar, J. W. Walker, B. K. Carpenter,

W. W. Webb, J. Marque, W. Denk, J. A McCray, and G. P. Hess, Biochemistry 28,
49-55 (1989).
40j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, Z Am. Chem. Soc. 110,
7170-7177 (1988).
41S. Udenfriend, T. Stein, P. B6hlen, W. Dairman, W. Leimbruger, and M. Weigele, Science
178, 871-872 (1972).
1.0 Standard / 5 rain •~ 1.0 ~.1~....,~ v : i

10 nmol . e:;~)
i,j \~ 4 mill .~
- O.8 ~ -
0.6 •
1
/:~
0.8 o41
iOoW 0 1 2 3 4 5
Photolysis time [rain]
'~ 0.6
o
0.4
.// ~\,\ .
I , ,\ ,,'\\
I
;'/ \,G,.
/.//O.~o..d~C-Bio~°-L0u-E~\'~',
I LI Aft~photolysis >\ ~ , ~
400 500 600 700

Wavelength [nm]
FIo. 4. Photorelease of Leu-Enk from PC-biotin-Leu-Enk assayed by fluorescamine. 27 PC-
biotin-Leu-Enk was irradiated in solution (11) or complexed with streptavidin-agarose ( 0 )
or monomeric avidin agarose ( a ) .
amino group. Figure 4 (inset) shows the time course for the release of Leu-
Enk on photocleavage of PC-biotin-Leu-Enk in solution. No fluorescence
is detected before illumination (0 min), confirming that no unreacted Leu-
Enk exists in solution. The amount of Leu-Enk released at different times
due to photocleavage is also calculated using the fluorescamine assay. Data
confirm that the Leu-Enk is almost completely released from PC-biotin-
Leu-Enk in solution within 5 min of UV illumination.*
In order to determine if the biotinyl moiety of PC-biotin-Leu-Enk re-
tains its affinity to avidin, we complexed PC-biotin-Leu-Enk with mono-
meric avidin-coated agarose beads. For this purpose, PC-biotin-Leu-Enk
was incubated for 30 min in a suspensin of these beads at a PC-biotin-Leu-
Enk/avidin ratio of 1/1.5. The suspension was then spin-filtered for 3 min.
The filtrate was analyzed for unbound PC-biotin-Leu-Enk by illuminating
it for 6 min, which is sufficient to release Leu-Enk, as shown by the HPLC
experiments (Fig. 3). As seen in Fig. 4, only a small amount of Leu-Enk
* The small drop in measured fluorescence after 5 min is most likely due to the presence of
the PC-biotin photoproduct, which may act as either a quenching agent or an inner filter.
Note in the case of PC-biotin-Leu-Enk complexed to beads such a drop is not observed
because the PC-biotin photoproduct remains bound to the avidin.
is still present in the filtrate (-6%) after spin-filtering the beads. This
indicates that the free PC-biotin-Leu-Enk binds efficiently to both the
monomeric avidin and streptavidin-coated agarose beads.
It was also established that Leu-Enk is completely released from PC-
biotin-Leu-Enk when complexed with monomeric avidin-coated agarose.
For this experiment, avidin-bound PC-biotin-Leu-Enk was photocleaved
by illumination of the resuspended agarose beads. As seen in Fig. 4, the
fluorescamine-based assay shows that approximately 9.7 nmol of Leu-Enk
is released into solution in approximately 5 rain. This is close (within the
estimated error) to the amount of PC-biotin-Leu-Enk (9.4 nmol) immobi-
lized on monomeric avidin beads. Furthermore, HPLC analysis confirmed
that the released photoproduct is Leu-Enk (data not shown). A similar
experiment (Fig. 4, inset) using streptavidin-coated agarose beads also es-
tablished that PC-biotin-Leu-Enk is efficiently bound by streptavidin and
that Leu-Enk is completely released on illumination in less than 4 min.
Evaluation of PC-Biotin Oligonucleotides

The heptamer, 5'-PC-biotin-(dT)7, was assembled using PC-biotin phos-
phoramidite (11) in an automated DNA/RNA synthesizer. The unmodified
sequence, 5'-OH-(dT)7, and a 5'-phosphorylated sequence, 5'-p-(dT)7,
were prepared using standard procedures (see Materials and Methods).
Figure 5 shows the HPLC trace of 5'-PC-biotin-(dT)7 (trace a, Fig. 5). Two
main peaks are observed in this trace with a retention time of 23.9 and
24.5 min. These two peaks can be attributed to the two diastereoisomers
o
PCB--O--P --O--d(T-T-T-T-T-T-~
o o
O--P --O--d (T-T-T-ToT-T*]3
®O
HO~d{T-T-T-T-T-T-]}
Retention time [rain]

FIG. 5. HPLC traces of (a) 5'-PC-biotin-(dT)7, (b) 5'-PC-biotin-(dT)7 irradiated with
near-UV light for 5 min, (c) 5'-p(dT)7, and (d) 5'-OH-(dT)7. [Adapted from J. Olejnik, E.
Krzymanska-Olejnik, and K. J. Rothschild, Nucleic Acids Res. 24, 361-366 (1996).]
150 CACZDCOMPOUNDS [81
generated by the introduction of the PC-biotin moiety onto the 5' end of
the oligonucleotide.3° Compared with the unmodified oligonucleotide, 5'-
OH-(dT)7 (trace d, Fig. 5, retention time 14.5 min), the PC-biotin-modified
oligonucleotide (trace a, Fig. 5) shows an increased retention time, which
is typical for biotinylated oligonucleotides. 12'25 We conclude from these
data that 5'-PC-biotin moiety is retained during cleavage and deprotection
of the oligonucleotide with ammonia [5'-phosphorylated oligonucleotide
is not present in the 5'-PC-biotin-(dT)7 sample].
The interaction of the PC-biotin-modified oligonucleotide with streptav-
idin and the photorelease of the oligonucleotide was evaluated by incubat-
ing the 5'-PC-biotin-(dT)7 with streptavidin-agarose beads; separating the
beads from the solution by spin-filtering; and irradiating resuspended beads
with 300-350 nm light. The effects of irradiating the resuspended beads
for 4 min are shown in Fig. 5. The two peaks assigned to 5'-PC-biotin-
(dT)7 (trace a, Fig. 5) disappear and a single peak appears (trace b, Fig.
5) with a retention time of - 1 3 min. The retention time of this peak is
almost identical to that of the reference 5'-phosphorylated sequence, i.e.,
5'-p-(dT)7 (trace c, Fig. 5). These data conclusively show that irradiation
causes cleavage of the PC-biotin moiety and release of 5'-phosphorylated
oligonucleotide into solution.
We also measured the time dependence of the photoconversion of 5'-
PC-biotin- ( dT)7into 5'-p- ( dT)7 in solution. For this purpose, a 5'-PC-biotin-
(dT)7 solution was subjected to irradiation with 300-350 nm light and the
reaction mixture was analyzed by reversed-phase HPLC after different
irradiation times (Fig. 6). It can be seen from the decrease of the intensity
of peaks at 23.7 and 24.3 min assigned to 5'-PC-biotin-(dT)7 and the increase
of the intensity of single peak at -13 min assigned to 5'-p-(dT)7 that the
photoreaction is complete in approximately 4 min. The appearance of the
additional small peaks with a retention time of approximately 33 min can
be attributed to formation of the biotinyl-2-nitrosoacetophenone derivative
and other minor photoproducts identified previously.39,4°
The time dependence and efficiency of the photocleavage of 5'-PC-
biotin-(dT)7 complexed with streptavidin-agarose beads was determined
by measuring the absorbance of the supernatant at 260 nm, which reflects
the amount of 5'-p-(dT)7 released into solution. The initial A260 value at 0
min (Fig. 6, inset) corresponds to less than 3% of the 5'-PC-biotin-(dT)7
prior to complexation with streptavidin-agarose beads. This result shows
that 5'-PC-biotin-(dT)7 is almost quantitatively (97%) complexed with
streptavidin-agarose. On irradiation, the photorelease is very rapid (4 min)
and reaches 92% of the estimated absorption due to 5'-PC-biotin-(dT)7
prior to complexation (see Materials and Methods). No further increase in
absorbance was observed after 6 min of irradiation. In a separate experi-
100
8O
0.6-
0.5 ~
40
~0.4- 20
oo 0.3. i ~ ~ 4 5
Irradiation[rain]
0.2-
4.0
0.1
- ~---~.~.~
o.o. 2~ j -0.25A.~
00 ~ " ~
f0 1'5 20 i5 30 35 40
Retention time [min]
FIG. 6. H P L C traces of 5'-PC-biotin-(dT)7 after increasing times of irradiation. 2s See
Materials and Methods for more details. (Inset) Time dependence of photocleavage 5'-PC-
biotin-(dT)7-->5'-p-(dT)7 reaction in solution (solid lines); (0,) concentration of 5'-PC-biotin-
(dT)7; (11) concentration of 5'-p-(dT)7. For comparison, time dependence for photorelease
of 5'-p-(dT)7 from 5'-PC-biotin-(dT)7-streptavidin-agarose beads complex is shown (dashed
line, O).
ment, synthetic 5'-p-(dT)7 was incubated with streptavidin-agarose beads.

It was found that approximately 8% of 5'-p-(dT)7 binds nonspecifically to
the streptavidin-agarose beads (data not shown). Thus, the incomplete
release of 5'-p-(dT)7 appears to be due to nonspecific binding and not to
incomplete photocleavage.
In order to evaluate the usefulness of PC-biotin phosphoramidite for
synthesis and affinity purification/phosphorylation of longer oligonucleo-
tides, two 5'-PC-biotin-labeled sequences, a 50-mer and a 60-mer, were
prepared. After deprotection, the crude 5'-PC-biotin oligonucleotides were
separately incubated with streptavidin-agarose beads. The beads were then
washed, resuspended, and finally irradiated to obtain the full-length phos-
phorylated oligonucleotides. Figure 7 shows the results of polyacrylamide
gel electrophoresis (PAGE) of the crude 50-mer (lane 1) and 60-mer (lane
4) oligonucleotides, and the affinity purified and photocleaved oligonucleo-
tides (50-mer, lane 2; 60-mer, lane 5). Polyacrylamide gel electrophoresis
of the supernatant obtained after isolation of the oligonucleotides with
streptavidin-agarose beads is also shown (50-mer, lane 3; 60-mer, lane 6).
In agreement with earlier studies, the biotinylated oligonucleotides migrate
slower, whereas 5'-phosphorylated sequences migrate faster than sequences
with 5'-OH. a° It can be further seen that affinity purification and photocleav-
age result in a compact band, indicative of high purity and homogeneity,
i/~ i !!~i!¸¸
123456
FIG. 7. PAGE (20%) of 50-mer (lanes 1-3) and 60-mer (lanes 4-6) visualized by UV
shadowing, each lane representing 0.25 OD260 units of material. 28 Lanes 1, 4: crude 5'-PC-
biotin-oligonucleotides; lanes 2, 5: PC-biotin-affinity purified and phosphorylated oligonucleo-
tides; lanes 3, 6: filtrate containing sequences not bound to streptavidin.
in contrast to crude material, which exhibits a much broader band with

additional material appearing at lower molecular weight. Because the latter
material did not bind to streptavidin, it is likely to correspond to the
failure sequences.
Concluding Remarks
As described, we have designed, synthesized, and evaluated two photo-
cleavable biotin derivatives that can be used as reagents to label biomole-
cules. These represent a much broader class of photocleavable affinity tags
that are currently under development. Several properties are important for
a PC tag to be useful in the isolation of biomolecules. They include: (1)
Reactivity: The PC tag should react selectively and efficiently with target
molecules under mild conditions. (2) Complexation: The PC tag should
retain a high affinity toward the capture molecule, such as avidin. (3)
Photocleavage: The intensity and exposure time of light required for com-
plete photocleavage should be minimal. Additional requiremens for PC
phosphoramidites include compatibility with standard cycles and proce-
dures for automated D N A / R N A synthesis. The PC-biotin-NHS ester re-
ported here satisfies all of these criteria. It reacts with any substrate having
primary amino group under mild conditions, exhibits a high affinity toward
(strept)avidin, and allows fast ( - 4 min) and efficient (-99%) photorelease
of the substrate in an unaltered form. Biotin-NHS ester and its analogs have
been widely used as biotinylation reagents for a variety of biomolecules,
including antibodies, 3 hormones, I amino-derivatized nucleic acids, 8 and lip-
ids. 4a Similar protocols should be applicable for the biotinylation of these
42 M. R. Sofia, H. Loughrey, A. Ferraretto, A. M. Cannon, G. Acerbis, F. Sudati, G. Bottiroli,

and M. Masserini, J. Liposome Res. 3, 543-549 (1993).
molecules with PC-biotin-NHS ester, thereby expanding the range of appli-
cations of biotin-avidin technology.
PC-Biotin phosphoramidite also satisfies the above requirements, in-
cluding automated synthesis. It reacts efficiently (95-97% coupling yield)
with the 5'-hydroxyl group of the growing oligonucleotide chain under
standard conditions. The PC-biotin moiety introduced onto synthetic oligo-
nucleotide is stable during deprotection and shows high affinity toward
streptavidin. The PC-biotin moiety is rapidly (<4 min) and efficiently photo-
cleaved on illumination with near-UV light, resulting in 5'-phosphorylated
DNA. Thus, the addition of a 5'-PC-biotin moiety onto oligonucleotides
through the use of PC-biotin phosphoramidite provides a photoremovable
affinity tag for fast and efficient purification and phosphorylation of syn-
thetic DNA/RNA. This is especially important for the removal of failure
sequences that are formed as a result of imperfections during the synthesis
cycle. Note, however, that this method would not eliminate deletion se-
quences that are also known to o c c u r . 43'44
The ability to 5'-phosphorylate an oligonucleotide through photocleav-
age is another important advantage of this approach. Many applications of
synthetic oligonucleotides require phosphorylation on the 5' end, including
gene construction, cloning, oligonucleotide ligation assay (OLA), 45the liga-
tion chain reaction, 46 and total cDNA sequencing. Typically, 5'-phosphory-
lation is achieved by either enzymatic or chemical methods. The use of
enzymes involves several time-consuming steps and results often in non-
quantitative phosphorylation. Chemical phosphorylation is possible during
oligonucleotide synthesis using commercially available phosphoramidites
such as Phosphalink38; however, the resulting product still requires purifi-
cation.
In addition to PC-biotin-mediated affinity purification and phosphoryla-
tion, PC-biotin oligonucleotides should be useful in a number of other
applications. For example, they could be used as primers for polymerase
chain reaction (PCR), thereby simplifying the streptavidin-mediated affinity
purification of PCR products from a reaction mixture containing template
DNA, polymerases, and other components. This procedure would yield
unmodified amplified fragments suitable for sequencing or cloning. Other
possible applications include isolation of DNA/RNA macromolecular
complexes 2°-24 and controlled photorelease of oligonucleotides for the
triggering of DNA-protein interactions and for therapeutic purposes.
43 j. Temsamani, M. Kubert, and S. Agrawal, Nucleic Acids Res. 23, 1841-1844 (1995).
44 K. L. Fearon, J. T. Stults, B. J. Bergot, L. M. Christensen, and A. M. Raible, Nucleic Acids
Res. 23, 2754-2761 (1995).
45 U. Landegren, R. Kaiser, J. Sanders, and L. Hood, Science 241, 1077-1080 (1988).
46 F. Barany, Proc. Natl. Acad. Sci. U.S.A. 88, 189-193 (1991).
Photocleavable affinity tags can also be used for spatially addressable

photorelease of biomolecules for diagnostic applications.
In general, photocleavable biotin derivatives can overcome one of the
most critical limitations of avidin-biotin technology, i.e., the irreversibility
of the avidin-biotin interaction, without sacrificing the high affinity of this
interaction. For this purpose, it will be important to synthesize other PC-
biotin compounds that exhibit selectivity for other functional groups, such
as sulfhydryls. In addition, it is desirable to be able to vary the spacer arm
length and adjust the wavelength for optimal photocleavage to suit specific
applications. Importantly, the synthetic strategy employed here for the
synthesis of PC-biotin-NHS ester and PC-biotin phosphoramidite can be
readily adapted for the synthesis of other photocleavable affinity t a g s . 47'48
For example, the specific photoreactive moiety, reactive group, spacer arm,
and affinity tag can be replaced easily to create a variety of different reagents
using the current synthetic approach.
Photocleavable biotins are expected to be useful in a variety of applica-
tions relating to the isolation of biological molecules. Suitable target biomol-
ecules include proteins, nucleic acids, carbohydrates, lipids, and macromo-
lecular assemblies. PC-Biotin should also be useful in the isolation of special
cells expressing unique cell-surface antigens, which are targets of PC-biotin-
labeled antibodies. In addition to the direct chemical reaction of PC-biotin
with biomolecules, they can also be incorporated enzymatically during
their synthesis. For example, photocleavable biotins can be introduced into
nascent proteins during in vitro synthesis by using aminoacyl-tRNAs with
suitable amino acids, such as lysine, that will selectively react with PC-
biotin-NHS ester. This will be especially useful for the rapid and efficient
isolation of nascent proteins. PC-Biotin-labeled nucleotides can also be
incorporated into nucleic acids through enzymatic or chemical synthesis,
thus presenting new opportunities in gene cloning, PCR diagnostic assays,
and in situ hybridizations.
Acknowledgments
The authors thank Dr. Stacie Williams for proofreading of the manuscript. This work was
supported by a grant from the Army Research Office (ARO) (DAAL03-92-G-0172) to K.J.R.
The work was also supported by SBIR grants to AmberGEn, Inc., from the Army Research
Office (ARO) DAAH04-96-C-0050 and the National Institutes of Health IR43 GM54920-01.
47 j. Olejnik, E. Krzymanska-Olejnik, K. J. Rothschild, Nucleic Acids Res., submitted.

48 j. Olejnik, E. Krzymanska-Olejnik, K. J. Rothschild,/n preparation.
[9] HETEROBIFUNCTIONAL PHOTOCLEAVABLE REAGENTS 155
[9] S y n t h e s i s a n d A p p l i c a t i o n s o f H e t e r o b i f u n c t i o n a l
Photocleavable Cross-Linking Reagents
B y G E R A R D MARRIOTT a n d JOHANNES OTTL
Introduction
Absorption of near-ultraviolet light by biomolecules with one or more
functional moiety protected with a nitrobenzyl-based "caged" group leads
to a photoisomerization reaction in which the bond linking the functional
group of interest to the benzylic carbon of the caged reagent is cleaved.
The light-directed photocleavage reaction generates the biomolecule of
interest and a photoproduct in the irradiated volume within a few millisec-
onds or faster. 1-5 The ability to generate light-directed concentration jumps
of protein ligands and substrates provides an opportunity to measure tran-
sient kinetic data of specific protein activity in complex molecular environ-
ments 6 that include muscle fibers and living cellsJ '2'7 Monofunctional caged
reagents have been used to prepare inactive yet photoactivatable enzyme
substrates, 2 receptor ligands, 8 and fluorescent probes. 7,9 In addition, caged
reagents capable of labeling amino acid and thiol groups have been used
to inhibit the activity of proteins (caged proteins) by modifying one or
more essential amino acid residues. 1°-14These caged groups can be removed
from the inactive protein conjugate with concomitant recovery of its activity
after irradiation with near-ultraviolet light.
In addition to these monofunctional caged reagents, heterobifunctional
photocleavable reagents have been described to protect the amino or thiol
1 j. A. McCray and D. R. Trentham, Annu. Rev. Biophys. Chem. 18, 239 (1989).
2 j. H. Kaplan, B. Forbush, III, and J. F. Hoffman, Biochemistry 17, 1929, 1978.
3 B. Amit, U. Zehavi, and A. Patchornik, J. Org. Chem. 39, 192 (1974).
4 j. F. Cameron and J. Frechet, J. Am. Chem. Soc. 113, 4303 (1991).
5 D. Gravel, S. Murray, and G. Ladouceur, J. Chem. Soc. Chem. Commun. 1828 (1985).
6 S. R. Adams and R. Y. Tsien, Annu. Rev. Physiol. 55, 755-784 (1993).
7 j. A. Theriot and T. J. Mitchison, Nature 352, 126 (1991).
8 M. Wilcox, R. W. Viola, K. W. Johnson, A. P. Billington, B. K. Carpenter, J. A. McCray, A.
P. Cuzikowski, and G. P. Hess, J. Org. Chem. 55, 1585 (1990).
9 T. J. Mitchison, J. Cell Biol. 109, 637 (1989).
tt C.-Y. Chang, B. Niblack, B. Walker, and H. Bayley, Chem. Biol. 2, 391 (1995).
12 G. Marriott, Biochemistry 33, 9092 (1994).
13 p. D. Senter, M. Tansey, J. Lambert, and W. B1/ittler, Photochem. Photobiol. 42, 231 (1985).
14j Ottl, D. Gabriel, and G. Marriott, Bioconjugate Chemistry 9, March-April (in press).

METHODS IN ENZYMOLOGY, VOL. 291 0076-6879f98 $25.00
group of small molecules15 and proteins 1°'13 as their corresponding photola-
bile carbamate or thioether, respectively.1°,13,15 The second reactive group
of the cross-linking reagent may then be used to attach the caged compound
to a second biomolecule; for example, an antibody molecule to target the
caged compound to a specific site, 13 a derivatized surface, or a fluorescent
dye for imaging purposes or to probe protein-protein interactions. 1°'15 On
irradiation of the cross-linked protein complex with near ultraviolet light,
the bond linking the caged reagent to the protected group of the molecule
of interest is cleaved and the two biomolecules dissociate at a diffusion
controlled rate. Photocleavable cross-linked complexes have been described
in which either the activity of the protein of interest is blocked and activated
on irradiation with light,a°'13 or enhanced in the cross-linked complex and
reduced to a normal level after irradiation. 12This article describes heterobi-
functional, photocleavable cross-linking reagents and associated labeling
techniques that can be used to prepare caged protein conjugates that may be
used to achieve a photoactivation or photodeactivation of protein activity.
Materials
Reagents
4-Bromomethylbenzoic acid is purchased from Tokyo Kasei Company

(Tokyo, Japan). Fuming nitric acid, N-hydroxysuccinimide, dicyclohexyl-
carbodiimide, 3,4-dimethoxy-6-nitrobenzaldehyde, di-N-succinimidyl car-
bonate (DSC), vinylmagnesium bromide, meta-chloroperbenzoic acid
(MCPBA), 4-dimethylaminopyridine (DMAP), 4,6-diphenylthieno[3,4-
d][1,3]dioxol-2-one 5,5-dioxide (TDO) are from Aldrich (Milwaukee, WI).
Bovine serum albumin (BSA), polylysine with an average molecular mass
of 40,000 Da and 2-iminothiolane are purchased from Sigma (St. Louis,
MO). Tetramethylrhodamine iodoacetamide (IATMR), acrylodan, and
aminodextran are purchased from Molecular Probes (Eugene, OR) or ob-
tained as a gift from Dr. John Corrie, NIMR, Mill Hill, London. Rhodamine
110 is from Lambda Physik (G6ttingen, Germany). All other reagents are
of the highest quality available and obtained from Sigma or Aldrich unless
stated otherwise. G-Actin is prepared according to Marriott. 12
Instrumentation
Nuclear magnetic resonance (NMR) spectra are measured on a 500-

MHz Bruker Instrument or a 400-MHz instrument (JOEL, JNM-GX 400,
15j. Olejnik,S.Sonor,E. Kryzymanska-Olejnik,and K.Rothschild,Proc. Natl. Acad. ScL U.S.A.

92, 7590 (1995).
Tokyo, Japan), mass spectrophotometry on a Finnigan MAT 900 or Finni-

gan HSQ 30 instrument. Infrared spectra are recorded on a Perkin-Elmer
1760 X FT-IR spectrophotometer or a JASCO A202 instrument of samples
suspended in Nujol. Absorption spectra are recorded on a HP 82152 diode
array spectrophotometer (Hewlett Packard) or a Shimadzu UV350A dou-
ble-beam instrument. Fluorescence spectrometry is performed on an SLM-
AB2 fluorometer (Sopra, Buttlelborn, Germany) or a Hitachi F-4010 instru-
ment. Light-directed photoactivation of caged compounds is performed
essentially as described by Marriott) z
Miscellaneous Methods
Reaction of 2-Iminothiolane with Bovine Serum Albumin

and Aminodextran
Bovine serum albumin, 4 mg, dissolved in 1 ml of 50 mM borate buffer,
pH 8.5, is treated with 40/xl of a freshly prepared solution of 3.5 mg of 2-
iminothiolane in 250/zl distilled water. The solution is left at room tempera-
ture for 1 hr and then centrifuged for 10 min at 10,000 rpm. The protein
conjugate is dialyzed overnight at 4 ° against two changes of phosphate
buffer, pH 7.5, previously saturated with nitrogen gas to prevent oxidation
of the thiol groups. After dialysis the thiolated BSA conjugate is centrifuged
at 14,000 rpm for 20 min at 4°. A similar labeling procedure is used to
prepare the thiolated dextran from aminodextran.
Determination of Free Thiol Content of Bovine Serum Albumin

Of a 10 mM stock solution of acrylodan (the thiol reactive form of the
fluorescent dye prodan) dissolved in DMF, 15/zl is added to 300/zl of a
60/zM solution of iminothiolane-BSA at pH 8.0. The reaction is left for
2 hr at room temperature, and the solution centrifuged for 10 min at 10,000
rpm and dialyzed for 2 days against phosphate buffer, pH 7.5, at 4 ° with
three changes of buffer. The protein conjugate is centrifuged for 20 min
at 14,000 rpm at 4 ° before recording its absorption spectrum. The concentra-
tion of prodan in the conjugate is calculated from the absorption value at
385 nm using an extinction coefficient of 18,500 M -1 c m - 1 ) 6
Labeling of G-Actin with Tetramethylrhodamine Iodoacetamide

Rabbit muscle G-actin, purified according to Marriott, lz in G-buffer
(2 mM Tris, 0.2 mM CaC12, 0.2 mM ATP, pH 8.0) containing 0.5 mM
16 G. Marriott, K. Zechel, and T. M. Jovin, Biochemistry 27, 6214 (1988).

dithiothreitol (DTT) is treated with 13 /zl of a 10 mM stock solution of

IATMR dissolved in DMF, and the reaction is left for 2 hr at room tempera-
ture. The protein is centrifuged for 30 min at 100,000 g at 4° and dialyzed
overnight against two changes of G-buffer without DTT at 4°. The actin
conjugate is centrifuged at 100,000 g for 1 hr, and the supernatant is made
into F-buffer (G-buffer containing 2 mM MgC12 and 0.1 M KC1) and left
for 2 hr at room temperature. Ultracentrifugation and resuspension of
the pellet in G-buffer are followed by an overnight dialysis against G-
buffer without DTT. The protein solution is clarified by centrifugation
(100,000 g for 1 hr), and absorption spectrometry is used to calculate a la-
beling ratio of 0.4 fluorophore per actin monomer using an extinction
coefficient for IA-TMR of 96,500 M 1 cm 1 at 550 nm. TM
4-Bromomethyl-3-nitrobenzoic Acid Succinimide Ester: A Simple

Heterobifunctional, Photocleavable Cross-Linking Reagent
This reagent, described by Marriott et aL1° has been used to prepare
cross-linked complexes of actin to study properties of the nucleation stage
in the polymerization of G-actin, 1° and the stabilization of actin filament
ends by cross-linked actin oligomers containing at least three monomers. 17
Syntheses
4-Bromomethyl-3-nitrobenzoic Acid. In a well-ventilated hood, 5 g of
4-bromomethyl-3-nitrobenzoic acid (BNBA) is added gradually to 50 ml
of fuming nitric acid maintained at -11 ° using a KCl-ice bath. Control of
the reaction temperature is critical for both safety and high product yield.
After 2 hr the reaction mixture is poured onto crushed ice and the product
is recrystallized from dichloromethane-heptane (1 : 1, v/v). Mass spectro-
photometric analysis of the product shows it contains the expected molecu-
lar ions of 269 and 271 m/z and a debrominated fragment of 190 m/z.
4-Bromomethyl-3-nitrobenzoic Acid Succinimide Ester. 4-Bromo-
methyl-3-nitobenzoic acid, 2.5 g, is dissolved in 20 ml of dry acetonitrile
followed by 1.05 g of N-hydroxysuccinimide and 2.15 g of dicyclohexylcar-
bodiimide, and the reaction is left overnight at room temperature. After
removal of the dicyclohexylurea by centrifugation, filtration, and evapora-
tion of the solvent, the product is recrystallized from dichloromethane-
heptane (1 : 1), yielding 2.02 g of a white solid. Mass spectrophotometric
analysis reveals molecular ions of the product of 356 and 358 m/z and
a debrominated fragment of 227 m/z. NMR: 6 2.94 (4H, s, succinimide
methylene); ~ 4.87 (2H, s, bromoalkane methylene); ~ 7.778 (1H, dd, aro-
17 H. Miyata, K. Kinosita, and G. Marriott, J. Biochem. 121, 527 (1997).

matic); ~ 8.783 (1H, d, aromatic). This reagent and several longer chain
derivatives are now available commercially from the Dojin-do Chemical
Co. (http://www.dojindo.co.jp).
4-(2-Hydroxymercaptomethyl)-3-nitrobenzoic Acid. A solution con-
taining 100 mg of BNBA and a fivefold molar excess of 2-mercaptoethanol
in 5 ml of tetrahydrofuran is heated to reflux in the dark for 2 days. The
product identified by thin-layer chromatography (TLC) on silica gel plates
is purified by silica gel chromatography using chloroform-methanol (2 : 1,
v/v). The molar absorption extinction coefficient at the 2-nitrobenzyl
group is determined to be 500 M -1 cm -~ at 350 nm.
Biochemical and Photochemical Properties of 4-Bromomethyl-3-

nitrobenzoic Acid Succinimide Ester
4-Bromomethyl-3-nitrobenzoic acid and BNBA-SE (Fig. 1) react with
the thiol group of cysteine residues of actin and BSA between pH 7 and
8 to form the corresponding thioether. Relatively harsh reaction conditions
are required to prepare the 2-mercaptoethanol derivative of BNBA. Irradi-
ation of a solution of 4-(2-hydroxyethylmercaptylmethyl)-3-nitrobenzoic
acid with near-ultraviolet light results in the photocleavage of the thioether
bond and release of mercaptoethanol and a nitrosobenzaldehyde derivative.
Infra-red spectroscopy of 4-(2-hydroxyethylmercaptylmethyl)-3-nitroben-
zoic acid recorded before and after irradiation with near-ultraviolet light
shows the disappearance of the strong nitrobenzyl absorption band at 1530
cm -1, and this result is consistent with a bond cleavage reaction mechanism
that proceeds via the photoisomerization of the nitrobenzyl group.
Preparation of Photolabile Dimer of Actin

Cross-linking of F-Actin with 4-bromomethyl-3-nitrobenzoic acid suc-
cinimide ester. The photocleavable cross-linking reagent BNBA-SE is used
to prepare an actin dimer whose nucleating activity in the polymerization
reaction of G-actin can be controlled with near-ultraviolet light. Cross-
linking of actin oligomers is achieved by incubating a solution of F-actin
at 25/~M in thiol-free F-buffer with BNBA-SE at 12.5 tzM for 1 hr at 20 °.
The F-actin fraction is centrifuged at 100,000 g at 4° for 1 hr, and the
pellet is resuspended in G-buffer containing 1 mM 2-mercaptoethanol and
dialyzed against this buffer for 24 hr at 4°. After a 5-min bath sonication
at 4°, the protein is centrifuged again at 100,000 g, the supernatant is set
aside, and the pellet is subjected to a second round of depolymerization.
The combined supernatant fractions are subjected to gel-exclusion chroma-
tography using Sephadex G-150 (Pharmacia, Freiburg, Germany). Fractions
160 CAGED COMPOUNDS
B
- Lysine - - Lysine -
F_-NH e-NH
NH NH
FIG. 1. (A) Molecular structures of BNBA and BNBA-SE. (B) Light-directed photocleav-
age reaction of cross-linked actin.
containing cross-linked actin dimer are identified by sodium dodecyl sul-

fate-polyacrylamide gel electrophoresis (SDS-PAGE) and pooled .
Cross-linking of actin monomers in F-actin with BNBA-SE proceeds
very efficiently using the reaction conditions described. Competition label-
ing experiments using acrylodan reveal that the highly reactive cysteine-
374 of one actin monomer is the target of the reactive bromoalkane of the
caged reagent (data not shown). A lysine residue from an adjacent actin
monomer in the filament most likely reacts with the activated carboxyl
group of BNBA-SE. The cross-linking reaction generates a ladder of actin
oligomers containing up to 14 or more monomer units. 17 The actin dimer
is purified from this ensemble of cross-linked actin monomers using size-
exclusion chromatography and is shown to be fully active by measuring its
ability to polymerize in F-buffer (Fig. 2B). As expected from earlier studies,
h,
lI6K>
97K >
68K >
45K >
I
31K >
mw a b c d e f g
B
200K >
116K>
97K >
T
68K >
y~
,sK, t V
31K >
mw a b c d e f g h
FIG. 2. (A) SDS-PAGE analysis of the light-directed photocleavage of the BNBA-SE-
mediated cross link between actin monomers in the purified actin dimer. Numbers of pulses
of 355-nm light delivered by the laser to a 0.85 p.M sample of the cross-linked actin dimer
in F-buffer is (a) 0, (b) 30, (c) 60, (d) 90, (e) 150, (f) 250, and (g) 400. (B) Polymerizabitity
of the BNBA-SE cross-linked actin dimer. (a) Supernatant fraction and (b) pellet fraction of
a 0.85 p.M solution of the actin dimer in G-buffer after centrifugation at 4° for 60 min at
100,000 g. (c) Supernatant fraction and (d) pellet fraction of a 0.85/.~M solution of the actin
dimer in F-buffer after centrifugation for 60 min at 100,000 g. (e) Supernatant fraction and
(f) pellet fraction of a 0.85/zM solution of the actin dimer in F-buffer containing 5 mM DTT
after irradiation with 400 pulses of 355-nm light and centrifugation for 60 min at 100,000 g.
the actin dimer acts as a nucleating species in the polymerization of G-
actin, ls'19 The nucleating activity of the dimer is quantified by measuring
the rate of elongation of the actin polymerization reaction using the change
in the fluorescence emission of the prodan conjugate of G-actin against
different concentrations of the actin dimer (Fig. 3A). This nucleation activ-
ity is lost after irradiating the dimer with near-ultraviolet light; photocleav-
age of the cross-link and the concomitant disappearance of actin dimer
and the increase in actin monomer are confirmed by S D S - P A G E and
densitometric analysis of the Coomassie-stained bands (Fig 2A). These
data are then used to show that the elongation rate of actin polymerization
exhibits a linear dependence on the amount of actin dimer in the G-actin
solution (Figs. 3A and 3B).
4-Bromomethyl-3-nitrobenzoic acid cross-linked actin containing sev-
eral actin monomers has been shown to cap the ends of actin filaments,
thereby preventing their depolymerization in low salt buffer. 17 Irradiation
of actin filaments stabilized with these actin oligomers cleaves the cross
link and generates ends with normal actin monomers, which then dissociate
from the filament. We envision that photoactivation of these photocleavable
cross-linked actin oligomers, microinjected into motile cells, will prove
useful in understanding the role of actin filament capping proteins in the
regulation of actin filament dynamics during cell motility.
A New Class of Photocleavable Cross-Linking Reagent

A new class of photocleavable cross-linking reagent has been synthe-
sized to prepare protected amino compounds with improved spectroscopic
and photocleavage properties for applications in cell biology (Scheme l ) . TM
In this class of reagent, the two reactive groups are introduced off the
benzylic carbon atom. These reagents are designed with the following oper-
ational considerations in mind: (1) the caged cross-linking reagents should
react with the functional group of interest in an aqueous solution and at a
slightly alkaline pH; (2) the photoactivation reaction should exhibit a good
quantum yield and an action spectrum in the near-ultraviolet wavelength
region (340-400 nm) to avoid interference with other biomolecules; (3) for
photoactivation reactions performed on optically thin samples, the chromo-
phore of the caged group should exhibit a high molar absorptivity; and (4)
the photoproducts of the photoactivation reaction should not be toxic or
reactive with other functional groups in the preparation. The new reagents
presented in Scheme 1 fulfill most of these requirements.
18p. Knight and G. Offer, Biochem. J. 175, 1023 (1978).

19S. C. Mokrin and E. D. Korn, Z Biol. Chem. 256, 8228 (1981).
A L J
1 . . . .
0 tO 20
Time (minutes)
Number of 355 nm Pulses (I----m)

B 0 50 100 150 200 250 300 350
4.0
3.5
.2 3.0
o
2.5
i 2.0 ""1~,
~ 1.5
I.(3 o
i i i i i i
10 8 6 4 2 0
[Actin Dimer] nM (o o)
Fxo. 3. Effect of BNBA cross-linked actin dimer on the polymerization kinetics of prodan-
labeled G-actin. Of a 3.4/zM solution of prodan-G-actin in G-buffer, 195/zl was mixed with
5 /zl of a 0.85/zM solution of BNBA cross-linked actin dimer that had been subjected to
irradiation with 355-nm light using the following number of pulses (from top to bottom trace):
0, 30, 60, 90, 150, and 250. The lowest trace (250 pulses) represents the kinetics of the
polymerization of 3.4 ~ M prodan-G-actin without the addition of the cross-linked actin dimer.
(B) Dependence of the rate of elongation in the polymerization of prodan-G-actin calculated
using data presented in A on the number of light pulses delivered to the cross-linked actin
dimer sample (filled squares) and on the concentration of actin dimer (open circles) remaining
in samples irradiated with 355-nm light. The actin dimer concentration was calculated from
a densitometric analysis of the acrylamide gel, shown in Fig. 2A.
164 CAGEDCOMPOUNDS [9]
/0~ 0~l 0~ 0
H~c__c~..c ~o_c_c, 0 0 ~c... O~ ~~H
/~PI
NO2
hc NO2 Ph ~(3
HaCO" ~ " / pyridine TDO'carb°nat~

~¢
OCH3
II1 o•iOxane
c, o.o
H3CO" ~ "
OCH3
IV
DMAP / THF
H 3 C O ~ "NO2
OOHa 0 0
0 H2C C~. CI.I~O C - N ~ - NM(
~DSC/acetonitrill~
/O oJC~o--N/ ~ X-
H2C/ "OH C/H ~ . HaCO" ~"~
H3CO~
~ ~
NO2 O
VI
DMAP / E~20 OCH3
V
OCHa
SCHEME1
Syntheses
Preparation of 1-Hydroxy-l-(3, 4-dimethoxy-6-nitrophenyl)-2-propene:
compound L 3,4-Dimethoxy-6-nitrobenzaldehyde, 1.88 g (8.95 mmol),
freshly recrystallized out of toluene, is dissolved in 50 ml dry tetrahydrofu-
ran (THF) under an argon atmosphere. The yellow solution is cooled to
- 7 0 °, and 10.7 ml of a 1 M solution of vinylmagnesium bromide in THF
is added dropwise. The deep red solution is stirred for 3 hr, slowly warmed
to 20 °, and then 40 ml of a saturated NH4CI solution added dropwise. The
red solution is extracted three times with ethyl acetate, and the organic
phase is washed five times with saturated NaC1, dried over MgSO4, and
the solvent removed. The red, oily residue (2.44 g) is taken up in a little
ethyl acetate and eluted through a silica gel column developed in hex-
ane:ethyl acetate (3:1, v/v). The solvent is removed from the product
fractions, producing an orange-red powder. Yield: 1.53 g (6.33 mmol; 72%).
Molecular weight = 239.23, C11H13NO5; mp 101-102 ° 1H NMR (CDfl3 in
ppm): 2.63 (s, broad, 1H); 3.94 (s, 3H); 3.96 (s, 3H); 5.24 (dd, J = 11 Hz,
2 Hz, 1H); 5.42 (dd, J = 18 Hz, 2 Hz, 1H); 5.92 (d, J = 6 Hz, 1H);
6.07 (ddd, J = 18 Hz, 11 Hz, 6 Hz, 1H); 7.20 (s, 1H); 7.57 (s, 1H).
MS (70 eV), m/z (%): 239 (42) M ÷, 162 (100) C9HaNO2 ÷.
[91 HETEROBIFUNCTIONAL PHOTOCLEAVABLE REAGENTS 165
Preparation of 1-(3,4-Dimethoxy-6-nitrophenyl)-2,3-epoxypropyl Hy-

droxide: compound II. meta-Chloroperoxybenzoic acid (MCPBA, technical
grade, about 80% pure) is washed three times with phosphate buffer, pH
7.5, and dried. Purified MCPBA, 1.311 g (7.60 mmol), is dissolved in 20 ml
CH2C12 and slowly dropped into a precooled solution of 1.80 g (7.52 mmol)
of I in 15 ml CH2C12. The bright yellow solution is stirred for 48 hr at room
temperature. The white precipitate is filtered and washed with CH2C12,
and then 70 ml of a saturated solution of NaHCO3 is added to the solution,
which is stirred for 30 min. The organic phase is washed seven times with
a saturated solution of NaHCO3 and then three times with a saturated
solution of NaC1. The organic phase is dried over MgSO4 and the solvent
removed. The yellow, oily residue (1.90 g) is recrystallized out of ethyl
acetate/hexane. Yield: 1.55 g (6.10 mmol, 81%, orange needles). Molecular
weight: 255.23, CnH13NO6; mp, 118-121-. IH NMR (CDC13 in ppm): isomer
a: 2.62 (s, 1H); 2.71 (dd, J = 5 Hz, 2 Hz, 1H); 2.90 (dd, J = 5 Hz, 2 Hz,
1H); 3.32 (ddd, J = 5 Hz, 5 Hz, 2 Hz, 1H); 3.96 (s, 3H); 3.98 (s, 3H); 5.57
(dd, J = 5 Hz, 2 Hz, 1H); 7.23 (s, 1H); 7.66 (s, 1H); isomer b: 2.63 (s, 1H);
2.82 (dd, J = 5 Hz, 2 Hz, 1H); 3.02 (dd, J = 5 Hz, 2 Hz, 1H); 3.62 (ddd,
J = 5 Hz, 5 Hz, 2 Hz, 1H); 3.96 (s, 3H); 4.02 (s, 3H); 5.76 (dd, J = 5 Hz,
2 Hz, 1H); 7.28 (s, 1H); 7.67 (s, 1H). (70 eV), m/z (%): 255 (35) M +, 164
(100) C9H10NO2+. FAB-MS m / z 256.1 MH+MS.
Preparation of 1-(3,4-Dimethoxy-6-nitrophenyl)-2,3-epoxypropyl Chlo-
roformate: compound I l L In a well-ventilated hood, 566 mg of II (2.22
mmol) dissolved in 10 ml of water-free dioxane in a dried 50-ml flask is
treated with 0.18 ml of water-free pyridine (2.25 mmol) together with
0.26 ml diphosgene (2.20 mmol), which is slowly dropped into the yellow
solution. The reaction mixture is stirred for 2 hr at room temperature, the
precipitate filtered off, the solvent removed, and the flask left for 1 hr
under a high vacuum to remove excess diphosgene. The oily, orange-brown
residue is taken up in a little ethyl acetate and eluted through a silica gel
column developed in hexane : ethyl acetate (3 : 1). The solvent is removed
from the product fractions, yielding a yellow powder. Yield: 620 mg (1.95
mmol, 88%). Molecular weight: 317.68, C12H12NO7CI; mp, 178-181 °. 1H
NMR (CDCI3, in ppm): isomer a: 3.57 (dd, J = 12 Hz, 4 Hz, 1H); 3.76 (dd,
J = 12 Hz, 4 Hz, 1H); 3.94 (s, 3H); 3.97 (s, 3H); 5.52-5.56 (m, 1H); 6.32
(d, J = 8 Hz, 1H); 7.20 (s, 1H); 7.71 (s, 1H); isomer b: 3.92 (s, 3H); 3.96
(s, 3H); 4.00 (dd, J = 12 Hz, 2 Hz, 1 H); 4.19 (dd, J = 12 Hz, 3 Hz, 1H);
4.66-4.68 (m, 1H); 6.13 (d, 3 Hz, 1 H); 7.00 (s, 1H); 7.73 (s, 1H). MS
(70 eV), m/z (%): 317 (82) M +, 136 (100) C8H802 +.
Preparation of 4-[1-(3,4-Dimethoxy-6-nitrophenyl)-2,3-epoxypropyl-1-
oxycarbonyloxy] - 3 - oxo - 2, 5 - diphenyl - 2, 3 - dihydrothiophene 1,1-dioxide:
compound IV. Five hundred milligrams of II (1.959 mmol) dissolved in 4
ml water-free THF is treated with 672 mg of freshly recrystallized, 4,6-
diphenylthieno[3,4-d][1,3]dioxol-2-one 5,5-dioxide (TDO, 2.062 mmol).
The orange solution is refluxed for 4 hr in the absence of base, and the
solvent is evaporated to give an orange residue that recrystallizes out of
dry toluene. Yield: 186.7 mg (0.321 mol, 82%). Molecular weight: 581.56,
C28H23NO11; mp, 176-180 °. 1H NMR (in CDC13, in ppm); isomer a: 2.35
(dd, J = 4 Hz, 3 Hz, 1H); 3.06 (dd, J = 4 Hz, 3 Hz, 1H); 3.53 (ddd, J = 4
Hz, 4 Hz, 2 Hz, 1H); 3.90 (s, 3H); 3.95 (s, 3H); 5.14 (s, 1H); 6.62 (d, J = 4
Hz, 1H); 7.01 (s, 1H); 7.27-7.32 (m, 4H); 7.43-7.48 (m, 4H); 7.60 (s, 1H),
7.92-7.97 (m, 2H); isomer b: 2.39 (dd, J = 4 Hz, 3 Hz, 1H); 3.08 (dd, J =
4 Hz, 3 Hz, 1H); 3.64 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 3H); 3.92 (s, 3H); 3.97
(s, 3H); 5.15 (s, 1H); 6.67 (d, J = 4 Hz, 1H); 7.07 (s, 1H); 7.27-7.32 (m,
4H); 7.43-7.48 (m, 4H); 7.63 (s, 1H); 7.92-7.97 (m, 2H); isomer c: 2.83
(ddd, J = 4 Hz, 4 Hz, 2 Hz, 1H); 3.48 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 1H);
3.64 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 3H); 3.96 (s, 3H); 3.98 (s, 3H); 5.16 (s, 1H);
6.78 (d, J = 3 Hz, 1 H); 7.08 (s, 1H); 7.43-7.48 (m, 4H); 7.52-7.57 (m, 8H);
7.65 (s, 1H); 7.92-7.97 (m, 2H); isomer d: 2.92 (ddd, J = 4 Hz, 4 Hz, 2 Hz,
1H), 3.48 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 1H); 3.64 (ddd, J = 4 Hz, 4 Hz,
2 Hz, 3H); 3.96 (s, 3H); 3.99 (s, 3H); 5.18 (s, 1H); 6.81 (d, J = 3 Hz 1H);
7.14 (s, 1H); 7.43-7.48 (m, 4H); 7.52-7.57 (m, 8H); 7.67 (s, 1H); 7.92-7.97
(m, 2H). FAB-MS re~z: 582.4 MH +.
Preparation of N-l-([3, 4-Dimethoxy-6-nitrophenyl)-2,3-epoxypropyl)
oxy]carbonyl)4-(N,N-dimethylamino)pyridinium 3-oxo-2,5-diphenyl-2,3-
dihydrothiophene 1,1-dioxide 4-hydroxide: compound V. One hundred fifty
milligrams of IV (0.258 mmol) dissolved in 0.4 ml of dry THF is treated
with a solution of 35 mg DMAP (0.285 mmol) in 0.1 ml of dry THF. After
a 2-min sonification treatment, an orange precipitate forms that is left to
stand for 1 hr at room temperature, centrifuged, and the orange residue
washed with THF. Yield: 166 mg (0.237 mmol, 92%). Molecular weight:
704.7 (cation, 404.4) (cation, C19H22N307);mp 85° (decomposed). 1H NMR
(CDC13 in ppm): isomer a: 2.84 (dd, J = 5 Hz, 2 Hz, 1H); 3.02 (dd, J = 5
Hz, 2 Hz, 1H); 3.25 (s, 6H); 3.58 (ddd, J -- 5 Hz, 5 Hz, 2 Hz, 1H); 3.93 (s,
3H); 4.04 (s, 3H); 6.63 (d, J = 4 Hz, 1H); 6.70 (d, J = 7 Hz, 2H); 7.05 (s,
1 H); 7.64 (s, 1H); 8.12 (d, J = 7 Hz, 2H); isomer b: 2.89 (dd, J = 5 Hz, 2
Hz, 1H); 3.11 (dd, J = 5 Hz, 2 Hz, 1H); 3.25 (s, 6H); 3.63 (ddd, J = 5 Hz,
5 Hz, 2 Hz, 1H); 3.96 (s, 3H); 4.06 (s, 3H); 6.72 (d, J = 7 Hz, 2H); 6.75 (d,
J = 3 Hz, 1H); 7.09 (s, 1H); 7.66 (s, 1H); 8.14 (d, J = 7 Hz, 2H); counterion:
4.91 (s, 1H); 7.10-7.40 (m, 8H); 7.95-8.30 (m, 2H). FAB-MS re~z: 404.5
M + (cation).
Preparation of l-(3,4-Dimethoxy-6-nitrophenyl)-2,3-epoxypropylsuccin-
imidyl Carbonate:compound VL One hundred milligrams (0.392 mmol) of
II dissolved in 2 ml of acetonitrile is treated with 110 mg (0.431 mmol)
[9] HETEROBIFUNCq'IONAL PHOTOCLEAVABLE REAGENTS 167
of di-(N-succinimidyl) carbonate. N-Ethyldiisopropylamine, 70/zl (0.431

mmol), is added, and the reaction mixture is stirred for 5 hr at room
temperature. After evaporation of the solvent, the residue is dissolved in
ethyl acetate and washed three times with 20% citric acid, sodium bicarbon-
ate solution, and saturated sodium chloride. The organic phase is dried
over magnesium sulfate, the solvent evaporated, and the yellow residue
recrystallized out of chloroform/hexane, producing yellow-orange crystals.
Yield 138 mg 89% (0.348 mml). Molecular weight: 396.32. C 1 6 H 1 6 N 2 O l 0 .
1H NMR (CDCI3 in ppm): isomer a: 2.79 (s, 4H); 2.82 (dd, J = 5 Hz, 2 Hz,
1H); 3.02 (dd, J = 5 Hz, 2 Hz, 1H); 3.56 (ddd, J = 5 Hz, 5 Hz, 2 Hz, 1H);
3.95 (s, 3H); 4.04 (s, 3H); 6.66 (d, J = 4 Hz, 1H); 7.05 (s, 1H); 7.64 (s, 1H);
isomer b: 2.79 (s, 4H); 2.90 (dd, J = 5 Hz, 2 Hz, 1H); 3.10 (dd, J = 5 Hz,
2 Hz, 1H); 3.63 (ddd, J = 5 Hz, 5 Hz, 2 Hz, 1H); 3.96 (s, 3H); 4.06 (s, 3H);
6.77 (d, J = 3 Hz, 1H); 7.09 (s, 1H); 7.67 (s, 1H). MS (70 eV), m/z (%) =
396 (51) M ÷, 238 (100) CllHxzNO5+. FAB-MS m/z 397.4 MH +.
Spectroscopic and Photochemical Properties

The Grignard reaction of vinylmagnesium bromide on 3,4-dimethoxy-
6-nitrobenzaldehyde introduces a hydroxyl and vinyl group off the chiral
benzylic carbon and provides the starting point for the synthesis of the two
reactive groups of the cross-linking reagent (Scheme 1). The first of these
groups, an activated carbamate or carbonate, is used to react with and cage
an amino group of the compound of interest to its corresponding photolabile
carbamate. The second group, an oxirane, is used to alkylate thiol groups
and serves to cross link the caged amino compound to a second thiol-
containing biomolecule. This synthetic approach also allows the physical
and chemical separation of the two nucleophilic reactive groups from the
photoisomerizations functionality. The reagents are designed for applica-
tions in living cells, and this is facilitated by the use of the 3,4-dimethoxy-
6-nitrophenyl chromophore, which has a high molar absorptivity and a
red-shifted action spectrum around 340-420 nm that helped reduce irradia-
tion-based cell damage.
The absorption spectrum of a defined concentration of an ethanolic
solution of compound II shows the lowest energy transition in the near-
ultraviolet region and has an extinction coefficient at 350 nm of 5000 M -a
cm -1. Excitation of this ethanolic solution with near-ultraviolet light (340-
420 nm) leads to an irradiation dose-dependent loss of the 350-nm band,
whereas two new absorption transitions evolve with maxima at 268 and
378 nm. These absorption bands most likely belong to the photoproduct,
a derivative of 3,4-dimethoxy-6-nitrosoacetophenone. Evidence that the
reaction occurs via photoisomerization of the nitrophenyl group is provided
by an analysis of the infrared spectra of a film of H recorded before and

after near-ultraviolet irradiation, which shows the loss of a hydroxyl absorp-
tion band at 3450 cm -1 and a gain of a ketone absorption band at 1705 cm -1.
Reactivity of Oxirane Group with Thiols

To show that the oxirane group of compound II can react selectively
with thiol groups, a qualitative analysis of the reaction of compound II
with 2-mercaptoethanol is performed under various conditions of pH and
solvent composition. To summarize these data, alkylation of mercaptoetha-
nol with compound H occurs in aqueous solutions at a pH value of 8.0-9.5
or in organic solvents in the presence of A1203 or tetrabutylammonium
fluoride. Compound II does not react with primary amines under this
condition. Investigations are also made on the reaction of the oxirane group
of compound II with the single reduced thiol group of BSA and a 2-
iminothiolane conjugate of BSA with several thiol groups. Dialysis of these
reaction mixtures removes excess compound H and allows a calculation,
based on absorption spectrophotometric analysis, of the labeling ratio of H/
BSA for both conjugates. Native BSA is found to harbor a single molecule of
II per BSA molecule, whereas the iminothiolane-BSA is labeled with six
molecules of H per BSA. An independent determination of the number of
reactive thiol groups in the iminothiolane-BSA conjugate using the thiol-
labeling reagent acrylodan16 also shows it contains six free thiol groups
(data not shown). Evidently, at this slightly basic pH the oxirane group is
able to react selectively with reduced thiol groups of proteins.
Reactivity and Properties of the New Class of Reagents

Compound II proves to be a key intermediate in the synthesis of four
different cross-linking reagents (Scheme 1). The first of these, compound
IH, reacts poorly with the amino groups of polylysine or G-actin in bicarbon-
ate buffer, pH 8.5. We suspect the substituent on the enzylic carbon atom
of compound IH is responsible for the deactivation of the chloroformate
group because the unsubstituted chloroformate (NVOC-CI) reacts effi-
ciently with amino groups in both organic and aqueous solvents.2,xz To
improve on the poor reactivity of compound H, we use Steglich's reagent
(TDO) 2° in order to make a more reactive carbonate. This reaction proceeds
smoothly in THF under reflux and produces compound IV in high yield.
The two isomers of compound IV cannot be resolved using the purification
scheme described in the Methods section. However, the photoisomerization
20 R. Kirstgen, A. Olbrich, H. Rehwinkel, and W. Steglich, Liebigs Ann. Chem. 437 (1988).
rates of the isomers of compound IV should be similar by analogy to the
similar rates found for the isomers of compound IlL
To improve the water solubility of heterobifunctional photocleavable
cross-linking reagents, and to increase their reactivity toward the e-amino
group of lysine residues of proteins, compound V, the DMAP salt of com-
pound IV, is prepared via two synthetic routes, outlined in Scheme 1. In
the first approach, the succinimide carbonate of H is prepared in a reaction
of II with di-N-succinimidyl carbonate in acetonitrile, which produces the
crystalline product VI in high yield. Treatment of compound VI in THF
with DMAP leads to the precipitation of the DMAP salt of compound V.
In the second synthetic approach, compound V is prepared in a single-step
reaction of compound IV with DMAP, as described in the Methods section.
Compound V, a highly activated, water-soluble carbamate, reacts rapidly
with the amino groups of proteins at a pH value between 8 and 9. Compound
V is found to hydrolyze in water, and therefore high concentrations of
both protein and compound V are recommended in labeling reactions.
Compound V can be added to protein labeling reactions as a solid. It
should be noted, however, that in organic solvents and in the presence of
nucleophiles, DMAP-activated salts such as V undergo esterification or
even reaction to isocyanates, 21 and therefore stock solutions of V in acetoni-
trile should be prepared just before the protein-labeling reaction.
Selected Applications
Members of the new class of cross-linking reagents are used to prepare
a caged fluorophore and a caged G-actin.
Thiol-Reactive Caged Rhodamine 110

Preparation of Di-[1-(3,4-dimethoxy-6-nitrophenyl)-2,3-epoxypropyl]-
N,N'-rhodamine Carbonate: compound VII. Compound IV, 100 mg (0.180
mmol), dissolved in 0.5 ml of water-free THF is treated with 40/zl of N-
ethyldiisopropylamine (0.350 mmol) and 32 mg of rhodamine 110 hydro-
chloride (0.086 mmol). An orange precipitate develops over a 2-min sonifi-
cation treatment. The reaction is left to stand for 4 hr at room temperature,
centrifuged, the solvent evaporated, and the orange residue taken up in 5 ml
of ethyl acetate. Successive washing of this organic solution with saturated
NaHCO3, 20% citric acid and with saturated NaC1 is used to remove the
sulfonate. The organic phase is then dried over MgSO4, the solvent re-
moved, and the product recrystallized from the residue out of toluene.
21 H. J. Kn61ker, T. Braxmeir, and G. Schlechtingen, Angew. Chem. 107, 2746 (1995).

170 CAGEDCOMPOVNDS [9]
Yield: 58 mg (0.065 mmol, 75%). Molecular weight: 892.5, C44H36N4017,
mp 188-193 °. (CDC13 in ppm): isomer a: 2.35 (dd, J = 4 Hz, 3 Hz, 1H);
3.06 (dd, J = 4 Hz, 3 Hz, 1H); 3.53 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 1H); 3.90
(s, 3H); 3.95 (s, 3H); 6.62 (d, J = 4 Hz, 1H); 7.01 (s, 1H); 6.68 (d, 8 Hz,
1H); 7.10 (s, 1H); 7.15 (d, 8Hz, 1H); 7.17 (d, 7 Hz, 1H); 7.19 (dd, 7 Hz, 7
Hz, 1H); 7.30 (dd, 7 Hz, 7 Hz, 1 H); 7.60 (s, 1H); 8.10 (d, 7 Hz, 1 H); isomer
b: 2.39 (dd, J = 4 Hz, 3 Hz, 1H); 3.08 (dd, J = 4 Hz, 3 Hz, 1H); 3.64 (ddd,
J = 4 Hz, 4 Hz, 2 Hz, 3H); 3.92 (s, 3H); 3.97 (s, 3H); 6.65 (d, J = 4 Hz,
1H); 6.68 (d, 8 Hz, 1H); 7.07 (s, 1H); 7.10 (s, 1H); 7.15 (d, 8 Hz, 1H); 7.17
(d, 7 Hz, 1H); 7.19 (dd, 7 Hz, 7 Hz, 1H); 7.30 (dd, 7 Hz, 7 Hz, 1H); 7.63
(s, 1H); 8.10 (d, 7 Hz, 1H); isomer c: 2.83 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 1H);
3.48 (ddd, J = 4 Hz, 4 Hz, 4 Hz, 1H); 3.64 (ddd, J = 4 Hz, 4 Hz, 2 Hz,
3H); 3.96 (s, 3H); 3.98 (s, 3H); 6.68 (d, 8 Hz, 1H); 6.78 (d, J = 3 Hz, 1H);
7.08 (s, 1H); 7.10 (s, 1H); 7.15 (d, 8 Hz, 1H); 7.17 (d, 7 Hz, 1H); 7.19 (dd,
7 Hz, 7 Hz, 1H); 7.30 (dd, 7 Hz, 7 Hz, 1H); 7.65 (s, 1H); 8.10 (d, 7 Hz, 1H);
isomer d: 2.92 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 1H); 3.48 (ddd, J = 4 Hz, 4 Hz,
2 Hz, 1H); 3.64 (ddd, J = 4 Hz, 4 Hz, 2 Hz, 3H); 3.96 (s, 3H); 3.99 (s, 3H);
6.18 (d, J = 3Hz, 1H); 6.68 (d, 8 Hz, 1H); 7.10 (s, 1H); 7.14 (s, 1H); 7.15
(d, 8 Hz, 1H); 7.17 (d, 7 Hz, 1H); 7.19 (dd, 7 Hz, 7 Hz, 1H); 7.30 (dd, 7
Hz, 7 Hz, 1H); 7.67 (s, 1H); 8.10 (d, 7 Hz, 1H). FAB-MS m/z: 893.3 MH +.
Amidation of the two aromatic amino groups of the highly fluorescent
rhodamine 110 fluorophore forms the nonfluorescent, lactone form of the
dye, which lacks a visible absorption band and fluorescence emissionf 2
Similarly, compound IV reacts with rhodamine 110 and produces the non-
fluorescent, bis-substituted carbamate (VII; Scheme 2). Irradiation of the
nonfluorescent (caged) rhodamine 110 with near-ultraviolet light cleaves
the two carbamate bonds and liberates free rhodamine 110 with a concomi-
tant recovery of its visible absorption band and intense fluorescence emis-
sion centered at 523 nm (Fig. 4). This reaction is quite efficient because
the photoactivation of a 500-/~1 sample of 50/xM solution of caged rhoda-
mine 110 in the presence of 5 mM DTT using the near-ultraviolet output
of a 75-W xenon arc lamp is complete within 240 sec (Fig. 4).
Labeling of G-Actin with Thiol-Reactive Caged Rhodamine 110 Reagent

The following method is used to couple one of the oxirane groups of
compound VII to the thiol group on a cysteine residue of G-actin: 2 ml of
25/xM F-actin is depolymerized by dialysis over 2 days at 4° against nitrogen-
purged G-buffer in the absence of DTT. The G-actin solution is clarified
by centrifugation for 1 hr at 100,000 g at 4°. A 20 mM stock solution of
22S. P. Leytus,L. L. Melhado,and W. F. Mangel,Biochem.J. 2119,299 (1983).

O
/C~ IJ H
H2C--CH.~/ 0 O~//~/~ Ph
~
/0.
H2N'~~
~
~\ y NH2+ CI-
N02 Ph/~S~0 +
i~ CO2H
OCHa
IV Rhodamine 110
THF / NR3
0 O
ZO\ jC-~ i H N O
O N\~ ~ ~. O CH--CH 2
H2C--CH\c/H
t/ i
HaCO+ ~ NO2
OCH3
VII
SCHEME 2
VII, 20/xl, is added to 500/zl of the G-actin solution after adjusting its pH
to a value of 9.0. After a 1-hr reaction at room temperature the actin
conjugate is clarified by centrifugation for 30 min at 10,000 rpm and dialyzed
for 2 days against G-buffer, pH 8.0, containing 1 m M D T r . The sample is
centrifuged for 1 hr at 100,000 g at 4 °, and the labeling ratio of VII to G-actin
is determined by absorption spectroscopy using an extinction coefficient for
VII of 5000 M -1 cm -1 at 350 nm and 3400 M -1 cm -1 at 290 nm. ]2
The labeling ratio of 1 : 1 determined for compound VII/G-actin suggests
that the caged fluorophore is attached to the highly reactive cysteine-374
residue. 16 The caged fluorescent G-actin conjugate polymerizes in physio-
logical salt, as seen in a comparison of the absorption spectra of the superna-
tant and pellet fractions of a high-speed centrifugation run, because the
majority of the photoactivated actin is found in the pellet fraction (Fig. 5).
Irradiation of the nonfluorescent G-actin conjugate with near-ultraviolet
light cleaves the carbamate bond that links the caged groups to rhodamine
110 and liberates the highly fluorescent rhodamine 110 and a polymeriza-
2f t
0 I , , ~ ' ~ ' ~ '-" j I- .... I
300 400 500 600 700
wavelength [nm]
B
e
t
t i
• -. xx x
I ' ".
I I I
500 550 6b0 650
wavelength Into]
FIG. 4. (A) Absorption spectra of approximately 500/zl of a 50 /zM solution of caged
rhodamine 110 in ethanol/H20 (10:1) containing 2 m M DTT. (a) Before and after near-
ultraviolet irradiation (340-400 nm) for the following times: (b) 5 sec, (c) 10 sec, (d) 180 sec,
and (e) 240 sec. (B) Fluorescence emission spectra of caged rhodamine 10 with excitation at
488 nm of approximately 50/zM of a 500-/zl solution of caged rhodamine 110 in ethanol/H20
(10:1). (a) Before, and after near-ultraviolet irradiation (340-400 rim) for the following times:
(b) 5 sec; (c) after 20 sec; (d) 60 sec; (e) 240 sec.
ik
"~ ~ iI
~i ~ "" /
300 460 560 600 700

wavelength [nm]
FIc. 5. (a) Absorption spectrum of G-actin-labeled caged rhodamine 110 before irradiation
with near-ultraviolet light in G-buffer containing 2 mM DTT. The G-actin was treated with
MgC12 and KC1 to 2 and 100 mM, respectively, and, after 2 hr to allow for polymerization,
the protein was centrifuged at 100,000 g for 60 min. (b) The absorption spectrum of the pellet
fraction of this centrifugation run after resuspension in an equal volume of F-buffer. (c) The
absorption spectrum of the supernatant fraction of this centrifugation run. (d) The absorption
spectrum of the resuspended pellet fraction [from (b)] after irradiation with near-ultraviolet
light for 10 min in the presence of 5 mM DTI?.
tion-competent G-actin that harbors a 3,4-dimethoxy-6-nitrosophenyl ke-

tone-based photoproduct (Fig. 5).
Preparation of Caged G-Actin Complex

Compound V is used as part of a new labeling procedure to prepare
caged G-actin, i.e., a G-actin conjugate that does not polymerize in physio-
logical salt but can polymerize normally after irradiation with near-ultravio-
let light. 12
A 30/zM solution, 800/zl, of IATMR-labeled G-actin in G-buffer with-
out DT-F is mixed with 20/.d of a freshly prepared 20 m M stock solution
of V in acetonitrile. After a 45-min reaction in the dark at room temperature,
the protein is centrifuged for 20 min at 14,000 rpm at 4 ° and dialyzed for
16 hr at 4 ° against nitrogen-purged G-buffer without D D T with two buffer
changes. The protein is clarified by ultracentrifugation at 100,000 g for 1
hr and the absorption spectrum recorded. The protein-labeling ratio is
calculated using an extinction coefficient for V of 5000 M -1 cm -1 at 350
nm and 3400 M -1 cm -1 at 290 nm. 12Aminodextran labeled with 2-iminothio-
lane, as described earlier, is added to a 20/.~M solution of IATMR-labeled
G-actin in a borate-based G-buffer, p H 9.5. After a 1-hr reaction the
complex is centrifuged for 20 min at 14,000 rpm at 4° and dialyzed overnight

against two changes of G-buffer containing 1 mM DTT. The fluorescent
actin-dextran complex is clarified by centrifugation for 20 min at 14,000
rpm at 4°.
Tetramethyl rhodamine iodoacetamide-labeled G-actin is labeled at an
average of 3.5 lysine residues with compound V. One or more of the oxirane
groups of the fluorescent G-actin conjugate of compound V is then cross
linked to a thiolated dextran. Sixty minutes after the addition of physiological
salt to this caged G-actin cross-linked complex, the sample is subjected to
high-speed centrifugation and the absorption spectra recorded for the super-
natant and resuspended pellet fractions (Fig. 6). These data show that the
majority of the fluorescently labeled G-actin conjugate is contained in the
supernatant fraction and is, therefore, polymerization incompetent (Fig. 6).
Exposure of a 500-t~l sample of this caged G-actin complex in F-buffer to
near-ultraviolet light for 10 min in the presence of 5 mM DTT cleaves the
cross link and released both G-actin and the dextran. After a period of 60
min at room temperature, to allow for polymerization of the photoactivated
G-actin, the sample is subjected to high-speed centrifugation (100,000 g for
60 min at 4°), and absorption spectroscopy of the supernatant and pellet frac-
tions now reveal that the majority of the fluorescent actin is present in the
resuspended pellet fraction (Fig. 6). The inability of the G-actin cross-linked
dextran complex to polymerize in physiological salt is most likely due to the
physical masking of the actin-binding site on caged G-actin by the dextran.
0~,-~1- " a
o
300 400 5;0 600 700
wavelength[nm]
FIG. 6. (a) Absorption spectra of IATMR-labeled G-actin dextran complex in G-buffer
containing 2 mM DT1Z After addition of MgCl2 and KCl to 2 and 100 mmol, respectively,
and 2 hr at room temperature the sample was centrifuged at 100,000 g at 4° for 60 min. (b) The
absorption spectrum of the supernatant fraction. No pellet fraction was found after the
centrifugation, i.e., the actin was caged. The supernatant fraction was then irradiated with
near-ultraviolet light for 10 min, and after 2 hr at room temperature the sample was centrifuged
at 100,000 g for 60 min. (c) The absorption spectrum of the pellet fraction resuspended in
an equal volume of F-buffer.
[10] LASERSFOR PHOTOLYSIS 175
This mode of inhibition, however, is relieved after light-directed dissociation

of the carbohydrate polymer from G-actin. Importantly, the photocleavage
reaction releases native, functional G-actin, whereas the photoproduct re-
mains covalently attached to the thiolated dextran. This new approach to
prepare caged G-actin requires the modification of fewer lysine residues
compared to the NVOC-CI method, le
Summary
This study designed, synthesized, and characterized a number of new
heterobifunctional photocleavable cross-linking reagents that may be used
to photomodulate the activity of proteins or to prepare caged fluorescent
dyes. Biomolecules or fluorophores caged via a thiol group with the BNBA-
SE reagent can be covalently linked to a second protein, ligand, or deriva-
tized surface through the activated carboxyl group. Members of the new
class of photocleavable cross-linking reagent can be used to cage amino
groups in the molecule of interest, which can then be covalently linked to
a second molecule through the thiol-reactive oxirane group. These cross-
linking reagents may be used for the following applications: (1) to cage the
activity of a protein by masking its active site with a second macromolecule,
e.g., aminodextran; (2) to prepare a protein conjugate exhibiting an en-
hanced or new activity that is lost on irradiation with near-ultraviolet light,
e.g., cross-linked actin dimer; (3) to target the caged compound to a specific
site by cross linking to a specific antibody; 13 (4) to attach the caged com-
pound to a thiol or amino derivatized surface; and (5) to render the caged
compound fluorescent in order to image or to quantify the yield of the
photoactivation reaction.
[ I O] T h e U s e o f L a s e r s f o r O n e - a n d T w o - P h o t o n
Photolysis of Caged Compounds
By JAMES A. MCCRAY
Introduction
The first use of a laser for single-photon photolysis of a caged compound,
p3-1-(2 nitro) phenylethyl-adenosine triphosphate (caged ATP), was in the
late 1970s by McCray, Herbette, Kihara, and Trentham. 1 The work utilized
1 j. A . M c C r a y , L. H e r b e t t e , T. K i h a r a , a n d D. R. T r e n t h a m , Proc. Natl. Acad. Sci. U.S.A.

77, 7237 (1980).

METHODS IN ENZYMOLOGY, VOL 291 0076-6879/98 $25.00
This mode of inhibition, however, is relieved after light-directed dissociation

of the carbohydrate polymer from G-actin. Importantly, the photocleavage
reaction releases native, functional G-actin, whereas the photoproduct re-
mains covalently attached to the thiolated dextran. This new approach to
prepare caged G-actin requires the modification of fewer lysine residues
compared to the NVOC-CI method, le
Summary
This study designed, synthesized, and characterized a number of new
heterobifunctional photocleavable cross-linking reagents that may be used
to photomodulate the activity of proteins or to prepare caged fluorescent
dyes. Biomolecules or fluorophores caged via a thiol group with the BNBA-
SE reagent can be covalently linked to a second protein, ligand, or deriva-
tized surface through the activated carboxyl group. Members of the new
class of photocleavable cross-linking reagent can be used to cage amino
groups in the molecule of interest, which can then be covalently linked to
a second molecule through the thiol-reactive oxirane group. These cross-
linking reagents may be used for the following applications: (1) to cage the
activity of a protein by masking its active site with a second macromolecule,
e.g., aminodextran; (2) to prepare a protein conjugate exhibiting an en-
hanced or new activity that is lost on irradiation with near-ultraviolet light,
e.g., cross-linked actin dimer; (3) to target the caged compound to a specific
site by cross linking to a specific antibody; 13 (4) to attach the caged com-
pound to a thiol or amino derivatized surface; and (5) to render the caged
compound fluorescent in order to image or to quantify the yield of the
photoactivation reaction.
[ I O] T h e U s e o f L a s e r s f o r O n e - a n d T w o - P h o t o n
By JAMES A. MCCRAY
Introduction
The first use of a laser for single-photon photolysis of a caged compound,
p3-1-(2 nitro) phenylethyl-adenosine triphosphate (caged ATP), was in the
late 1970s by McCray, Herbette, Kihara, and Trentham. 1 The work utilized
1 j. A . M c C r a y , L. H e r b e t t e , T. K i h a r a , a n d D. R. T r e n t h a m , Proc. Natl. Acad. Sci. U.S.A.

77, 7237 (1980).

a passive dye Q-switched, frequency-doubled ruby laser and was published

in 1980. This research demonstrated that biochemical kinetics inside cells
could be followed on the millisecond time scale, because the cell membrane
diffusion problem was bypassed. The motivation for this work was the 1978
paper by Kaplan, Forbush, and Hoffman, e where they had synthesized
caged ATP and used it in the study of Na+,K+-ATPase in red cell ghosts.
With the shuttered Hg arc light source that they used, they were able to
study only slower processes, on 1-sec and 100-msec time scales. Further
studies involving lasers to photolyze caged compounds have been guided
in the choice of the pulsed laser system to be used by the action spectrum
for photolysis of caged ATP, which was published in the Ph.D thesis of L.
Herbette 3 in 1980 (Fig. 1). This action spectrum should be typical for other
caged compounds in which the nitrobenzyl blocking group is used. The
main work horse for one-photon photolysis of caged compounds has been
the frequency-doubled ruby laser at 347 n m , 4 but other laser systems have
also been used, such as the frequency-doubled liquid-dye laser at 320 nm, 5
the XeF excimer laser at 351 nm, 6 and the nitrogen laser at 337 nm. 7 The
XeC1 excimer laser at 308 nm has also been used to photolyze caged ATP
in muscle studies, but sample destruction problems occurred, 8 presumably
because of absorption of light by aromatic amino acids in some of the
proteins. Caution should, therefore, be exercised with this laser system.
If it is desired to focus a laser beam through a microscope objective to
a very small volume, on the order of a few cubic microns, for one-photon
photolysis of caged compounds in or near biological cells, a serious problem
arises because photodamage of the biological material occurs. A possible
means of overcoming this problem was demonstrated in 1990 by Denk,
Strickler, and Webb. 9 They suggested that it would be better to use lower
energy (red) photons with wavelengths in the range of 640 to 700 nm and
the quantum mechancial two-photon absorption process to reach an energy
state different than that involved in one-photon absorption, which would,
however, still lead by nonradiative decay to the uncaging excited state.
They demonstrated that two-photon photolysis of caged ATP was possible.
2 j. H. Kaplan, B. Forbush, and J. F. Hoffman, Biochemistry 17, 1929 (1978).

3 t . G. Herbette, "Structure-Function Correlation Studies of Isolated and Reconstituted
Sarcoplasmic Membranes." Ph.D Thesis, University of Pennsylvania, Philadelphia, PA, 1980.
4 y. E. Goldman, M. G. Hibberd, J. A. McCray, and D. R. Trentham, Nature 3 ~ , 701 (1982).
5 j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, JACS l l 0 , 7170 (1988).
6 j. W. Karper, A. L. Zimmerman, L. Stryer, and D. A. Baylor, Proc. Natl. Acad. Sci. U.S.A.
85, 1287 (1988).
7 F. Engert, G. G. Paulus, and T. Bonheffer, J.Neurosci. Methods 66, 47 (1996).
8 R. Goody, private communication (1994).
9 W. Denk, J. H. Strickler, and W. W. Webb, Science 248, 73 (1990).
[lO] LASERS FOR PHOTOLYSIS 177
.=.
..... i
I.I.
..I c~
9
kk
O
.... ~E.... ! ~i......i~i~

"~~! 1 ° r ~ 0~
¢~. ~" 0~ ¢D ~ N O ¢~
178 CAGED COMPOUNDS [ lOI
The advantages of this method are that, in contrast to one-photon photolysis
where UV photons not involved in uncaging can photo-oxidize proteins,
leading to damage, the excess red photons not involved in uncaging can
be absorbed only into vibrational states of proteins, leading to at most a
temperature increase of the sample. This transient temperature rise prob-
lem, then, must be considered in the analysis of any experiment. The second
advantage of two-photon photolysis is the highly localized photolysis vol-
ume because the two-photon absorption probably is highest in the region
of highest peak power density. To have reasonable two-photon absorption,
it is necessary not only to focus the laser beam to a small volume in space
but also to have very narrow pulses in time so that the peak power density
is high while the average power density is low, in order to minimize damage.
This means that one should work with femtosecond lasers. The first femto-
second laser system used was a colliding-pulse, mode-locked dye laser l°'lx
that produced 100-fsec-duration pulses at a repetition rate of 80 MHz and
an average power of several milliwatts at 620 nm. This laser system, which
was constructed at Cornell University, was rather difficult to keep running,
so the next femtosecond laser system to be utilized for two-photon pho-
tolysis was a dye laser oscillator with gain jet and saturable absorber jet,
compensated for group velocity dispersion and synchronously pumped by a
mode-locked, frequency-doubled neodymium-yttrium-aluminum-garnet
laser (neodymium-YAG). This commercial Antares/Satori laser system
from Coherent, Inc., 12 produced, for example, a train of 200-fsec pulses at
76 MHz with an average power of about 200 mW and a wavelength of 640
nm. The two-photon photolysis then occurs at an energy twice that of the
original photons and hence at half the wavelength, i.e., at 320 nm. The
corresponding region of the action spectrum (see Fig. 1) can then be covered
using different dyes. This dye laser system does require some effort to
operate properly and is not "turn-key," so for general usage it was thought
that the argon ion-pumped, titanium-sapphire solid-state laser 12'13 might be
easier for nonlaser specialists. This laser system is easier to use but can
give femtosecond laser outputs only at wavelengths above 700 nm, which
limits two-photon photolysis to the region above 350 nm. Preliminary work
at Cornell University TM with caged -/-aminobutyric acid (GABA), which
released the neurotransmitter G A B A , indicates that the two-photon ab-
sorption cross section is low at 700 nm but rises substantially as the wave-
10j. A. Valdemanis and R. L. Fork, IEEE J. Quantum Electron. QE-22, 112 (1986).
al F. W. Wise, I. A. Walmsley,and C. I. Tang, Opt. Lett. 13, 129 (1988).
12Coherent, Inc., Laser Group, 5100 Patrick Henry Dr., Santa Clara, CA 95054.
13W. Koechner, "Solid-State Laser Engineering," 2nd ed., Springer-Verlag, Berlin, Ger-
many, 1988.
14W. W. Webb, private communication(1997).
[101 LASERSFOR PHOTOLYSIS 179
length is lowered. The Ti-sapphire laser is thus not the best choice for two-
photon photolysis of caged compounds. A new laser system, the chromium-
forstarite solid-state laser 15,16 with outputs in the 1230- to 1330-nm range,
which frequency doubles to the 615-665-nm range, now being tested for
two-photon photolysis of caged compounds (at 308-333 nm excitation),
may turn out to be a good choice for the nonlaser specialist to use for two-
photon photolysis. One of the most difficult problems with two-photon
laser systems is, for many researchers, that of expense. A good discussion
of two-photon laser excitation of molecular systems may be found in a 1995
article by Denk, Piston, and Webb. 17
This article gives biologists, biochemists, physiologists, and other medi-
cal scientists the necessary basic knowledge they should have so that they
can choose laser systems wisely and use them reliably and safely in experi-
ments involving photolysis of caged compounds. A reasonable understand-
ing of laser systems can be obtained by using rate equations, algebra, and
trigonometry. Some general sources on optics and lasers can be found in
Refs. 13 and 18-27.
Stimulated Emission
The basic concept involved in lasers is that of stimulated emission; in
fact, the acronym laser means "light amplification by stimulated emission
of radiation." It was Einstein 28,29in 1917 who first showed that an excited
state of a molecule can not only decay by spontaneous emission (fluores-
cence} but also must be able to decay by stimulated emission. He based
his analysis on measurements and Max Planck's theory of black-body radia-
15 V. Petricdvic, S. K. Gayen, and R. R. Alfano, Appl. Phys. Lett. 52, 1040 (1988).
a6 V. P. Yanovsky and F. W. Wise, Opt. Lett. 19, 1952 (1994).
17 W. Denk, D. W. Piston, and W. W. Webb, in "Handbook of Biological Confocal Microscopy"
(J. B. Pawley, ed.), p. 445. Plenum, New York, 1995.
18 R. S. Longhurst, "Geometrical and Physical Optics," 2nd ed., Wiley, New York, 1967.
19 M. V. Klein and T. E. Furtak, "Optics," 2nd ed., Wiley, New York, 1986.
20 M. Born and E. Wolf, "Principles of Optics," 3rd ed., Pergamon, Oxford, England, 1965.
21 A. Gerrard and J. M. Burch, "Introduction to Matrix Methods in Optics," Dover, New
York, 1994.
22 O. Svelto, "Principles of Lasers," Plenum, New York, 1976.
23 A. E. Siegmen, "Lasers," University Science, Mill Valley, CA, 1986.
24 F. P. SchSfer, ed., "Dye Lasers," 3rd ed., Springer-Verlag, Berlin, Germany, 1990.
25 Ch. K. Rhodes, ed., "Excimer Lasers," 2nd ed., Springer-Verlag, Berlin, Germany, 1984.
26 C. O. Weiss and R. Vilaseca, "Dynamics of Lasers," VCH, New York, 1991.
27 j. R. Reitz, F. J. Milford, and R. W. Christy, "Foundations of Electromagnetic Theory,"
4th ed., Addison-Wesley, New York, 1993.
28 A. Einstein, Physik. Zeitschr. XVIII, 121 (1917).
29 R. C. Tolman, Phys. Rev. 23, 693 (1924).
180 CAGED COMPOUNDS [lOl
0 4
O
II 3
I-
0
0 1 2 3 4 5 6 7
Wavelength (~m)
FIG. 2. Energy density of black-body radiation as a function of wavelength for the absolute
temperature of 1400 Kelvin.
tion. 3° If the energy density, u (i.e., energy per unit volume), of a radiation
field in equilibrium with the molecules on the inside of an enclosed cavity
is measured as a function of wavelength of the radiation, a curve like that
shown in Fig. 2 is obtained. This is the same curve that is measured for
background cosmic radiation left over from the big bang by the COsmic
Background Explorer (COBE) satellite, although in this case the tempera-
ture is very low, about 2.7 Kelvin.31
Einstein was trying to obtain this curve for equilibrium conditions by
first writing the rate equations appropriate for absorption and emission of
radiation by molecules. He considered molecules with two energy levels,
E1 and E2, where E2 > El, such that there are N1 molecules in the ground
state, El, and N2 molecules in the excited state, E2 (see Fig. 3). The total
number of molecules is N = N1 + N2. In order to be consistent with the
curve of black-body radiation, Planck found that the factor
1
e (hc)/(Ak13 -- 1 (1)
was essential in the energy density. In this expression, h is Planck's constant,

c is the velocity of light in a vacuum, A is the wavelength of the light, k is
Boltzmann's constant, and T is the absolute temperature. Einstein, there-
3o p. A. Tipler, "Modern Physics," Worth, New York, 1978.

31 E. Chaisson and S. McMillan, "Astronomy Today," 2nd ed., Prentice-Hall, Upper Saddle
River, NJ, 1997.
[ 10] LASERSFOR PHOTOLYSIS 181
fore, also had to obtain this factor from the analysis. When he considered
the rate equation for the upper state, he found that it was necessary to add
an additional rate term that depended on energy density:
dN2 _
'A21N2 + B12uN1 -- B21t/N2 (2)
dt
The first term on the right represents spontaneous emission (fluorescence),

the second term represents absorption, and the third new term represents
stimulated emission.
Setting dN2/dt equal to zero for the case of equilibrium and solving for
u gives
An
u= (3)
B12(N1/N2) - B21
The ratio NJN2 is given by the reciprocal Boltzmann factor e(E2-E1)/kT. Use
was also made of Bohr's energy-frequency relation, 3°
g2 - E1 = hc
h f = -~- (4)
where h f is the energy per photon. Thus
m21
U = B12(hc)/(~kT ) __ B21
(5)
Now, as the temperature becomes very large, approaching infinity, the

energy density also approaches infinity. This requires that B12 = B21, yield-
ing the equation
A21( 1 ) (6)
Einstein thus found that he could obtain the Planck factor only if he assumed
that an excited state of a molecule must also be able to decay by a term
that involves the number of photons already present. This is the motivation
for calling this term stimulated emission. An immediate consequence of
this conclusion is that photons can be added to those already in the field--an
amplifier of electromagnetic radiation should be possible. However, in 1917,
technology was not adequate to produce a light amplifier. In 1954, Gordon,
Zeiger, and Townes 32 produced the first stimulated-emission device, the
maser (microwave amplification by stimulated emission of radiation). By
32 j. p. Gordon, H. Z. Zeiger, and C. H. Townes, Phys. Rev. 95, 282 (1954).

N2 E2
/,/../-/~
Fro. 3. Energy level diagram for a two-level system illustrating absorption, fluorescence,
and stimulated emission.
1958 it was clear to Schawlow and Townes 33 that a light amplifier was
technologically possible, and in 1960 Maiman 34produced the first successful
optical device, the ruby laser.
Dynamic Bleaching
When using lasers it is very important to realize that one is dealing with
high-intensity light so that normal light filters used in the biochemical
laboratory should not be used to protect the eyes from laser light. The
point here is that stimulated emission must also be taken into account
when high-intensity (laser) light is being used. For low-intensity light, only
absorption and fluorescence need to be taken into account. Consider again
the two-level molecular system with energy levels, E1 and E2, as shown
in Fig.3. 35
At any given time the population of the ground state is N1 and that of
the excited state is N2. The time-dependent rate equation for the upper
level in terms of the light intensity, I (number of photons per square
centimeter per second) that impinges on the molecules is
dN2 N2
dt - o'IN1 - - - r - o'IN2 (7)
where o-is the absorption and stimulated emission cross section (with units
of area) and r is the mean fluorescence lifetime. The total number of
molecules is N = N1 + N2. At steady state the above derivative is zero,
yielding the dependence of N2 on the light intensity, I:
33 A. L. Schawlow and C. H. Townes, Phys. Rev. 112, 1940 (1958).

34 T. H. Maiman, Nature 187, 493 (1960).
35 R. W. Keyes, IBM J. October, 334 (1963).
[ 101 LASERSFOR PHOTOLYSIS 183
N2 = (-/1)
2o1 + N (8)
A plot of Eq. (8) is given in Fig. 4.

At high light intensity Eq. (8), which is based on the steady-state assump-
tion at the peak of a pulse, shows that N2 is just one-half of the total number
of molecules, N, which means also that one-half of the molecules are in
the ground state. Just as many molecules emit light by stimulated emission
as absorb light. The laser light then just passes through the material and
provides very little filtering. This feature is called dynamic bleaching. Only
filters labeled "laser filters" should be used to attenuate laser light for eye
protection! It is also very important to make absolutely sure that laser
goggles selected for eye protection are definitely for the particular wave-
length of the laser being used. Particular care should be taken when fre-
quency-doubled laser systems are being used because two different laser
beams of widely different wavelengths are present. Dynamic bleaching is
also used in passive Q-switching to hold off lasing while an inversion popula-
tion is built up in Q-switched lasers and in saturable absorbers to help
shape femtosecond laser pulses.
0.5
0.45
i i
0.4
0.35
0.3
010102
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii
0.05
0
0 ~
\
5 10
INTENSITY
15 20 25
1/2~
FI~. 4. Fractionaloccupationof the upper state with energyE2 as a functionof incident
light intensityillustratingdynamicbleaching.
An estimate of the critical intensity, Ic = 1/(2o-~-), can be made if the
size or cross section of an absorbing group on a molecule is taken as o-
2.5 × 10 -15 cm 2 and a fluorescent lifetime is taken as z ~ 10 8 sec. Then
I~ ~ 2 × 1022 photons per cm 2 per second. This corresponds to an energy
flux or intensity, for A = 320 nm, of about I~hc/A ~ 12 KW/cm 2. It should
be noted that this is a peak power density, not average power. Most pulsed
lasers used for photolysis have peak power densities above this value; for
example, the frequency-doubled, Q-switched ruby laser generally used for
one-photon photolysis of caged compounds has a pulse energy of 100 mJ
in 30 nsec and a cross-sectional area of about 12 mm 2. These numbers
result in a peak power density of I ~ 30 MW/cmQ which is 2.5 thousand
times the critical intensity. The peak power density directly out of two-
photon laser systems used is also of the order of a few megawatts per
square centimeter. After the microscope objective, the peak power density
can reach gigawatts per square centimeter. Of course, to be useful as a
photolysis source, there must also be nonradiative decay to a state in the
absorbing molecule that leads to the uncaging of the caged compound. The
combined spontaneous decay rates lead to a combined mean lifetime, %,
found from the equation
1 1 1
- ~- (9)
Tc ~'f Tn
where rf is the fluorescent lifetime (called r above) and ~'nis the nonradiative
lifetime. Because re, which now takes the place of r above, is less than the
fluorescent lifetime, rf, the critical intensity is higher.
Dynamic bleaching can also been seen if the effect of stimulated emis-
sion on Beer's law of absorption is determined. The intensity loss, dI, in a
passive material in a distance, dx, including stimulated emission is
d I = - o ' I N l d x + o'IN2 dx 0o)

Using the constraint equation, N1 + N2 = N, and Eq. (8) for N2 gives a
differential equation for I and x that can be integrated to give a modified
Beer's law.
In +A[
[/(x) ]
I0 - 1 =-trNx (11)
where A = 2o-zI0. The intensity as a function of x for different values of

A, plotted in Fig. 5, shows that for A = 0, Beer's law is obtained, but for
larger values of A the intensity is much larger than one would expect from
Beer's law. The situation for A = 1 is that for the critical intensity, Ic.
0.9
0.8
0.7
0.6
.~°
,~ 0.5
0.4
0.3
0.2
0.1
0 2 4 6 8 10
FIG. 5. Modified Beer's law absorption of laser light in a medium with stimulated emis-
sion included.
The Basic Laser

From the discussion on dynamic bleaching it can be seen that a passive
material cannot be used as a light amplifier. It is necessary to have an
inversion population where N2 > N1 ; then stimulated emission can add net
photons to the original light intensity. Laser material hosts doped with
metastable lasing ions or molecules are used as light amplifiers, but what
we usually mean by a laser is really a laser oscillator. Figure 6 shows the
layout of a basic laser oscillator. To achieve an inversion population, some
means of pumping the ions or molecules to a higher, third state with energy,
E3, must be used, such as electronic discharge for a gas laser or optical
pumping from a flash lamp, diode laser, or another laser for solid-state
lasers or liquid-dye lasers. Nonradiative decay then leads to the long-lived
second upper state of energy, E2, making an inversion population possible.
The stimulated emission may be increased by placing the lasing material
I I II I
LAMP
Total Partial
Reflecting Reflecting
FIG. 6. T h e essential components of a basic laser.
in an optical cavity so that feedback is possible. This is done by placing a

partially reflecting mirror and a totally reflecting mirror in front and behind
the lasing rod. When the mirrors are aligned parallel to each other the
optical cavity is "tuned" and feedback of the light can take place. Part of
the laser light is extracted from the cavity through the partially reflecting
mirror. The light makes two passes through the lasing rod for each round
trip, where it is enhanced by the gain of the lasing rod. There are, however,
losses of light during this round trip by absorption, scattering, diffraction,
and extraction through the output mirror. Thus the round-trip gain must
exceed the round-trip loss before laser light can appear at the output mirror.
The result is that a threshold of some kind must be exceeded in order to
get the laser to lase. For the flash-lamp pumped laser shown in Fig. 6, this
is evident in the voltage across the flash lamp, which leads to more pump
light. As will be discussed later in this article, it is necessary to operate
sufficiently above threshold so that the transverse beam structure can be
Gaussian. Some important safety points should be noted here. The voltages
used in pulsed laser systems come from high-voltage power supplies. Users
are well protected from these voltages, but if one takes off the protective
covers, then extreme care must be exercised in order to avoid electrocution.
Also, all laser beams should be covered, and any unused reflections or final
main beams should be dumped in recessed blackened beam stops. Special
care must be taken and fail-safe protection devised if laser beams are to
be used in microscopes.
Gaussian Beam Optics

It is very important to realize that the transverse profile of a laser beam
is not that of a plane wave but instead is usually Guassian, as given by
Eq. (12):
[10l LASERS FOR PHOTOLYSIS 187
__ Wo e-(x2+y2)lw 2
f(x,y,z) - w(z) (12)
where Wo is the waist radius and
w ( z ) = Wo 1 + \-~w2o] (13)
is the beam radius at a distance z from the waist. Equations (12) and (13)
are the result of an analysis called paraxial or first-order beam optics. There
are times when aberrations require the consideration of third-order beam
optics, although for most work first-order beam optics suffices if the angles
of the rays and beams from the optical axis are small.
Higher-order transverse modes are possible with a Gaussian profile but
modified by either Hermit or Laguerre polynomials. If care is not taken
to produce the fundamental Gaussian mode e.g., with a small aperture, a
multimode output is obtained. This is particularly likely to happen if the
laser is operated near threshold.
As can be seen in Fig. 7, a Gaussian beam cannot be focused to a
geometrical point but only to a small disk, called a waist. Close to the beam
axis the wavefront is spherical. The tightness of the focusing is indicated
by the Rayleigh range, which is the distance over which the beam radius
increases to V'2 times that at the waist and is given from Eq. (13) by
the formula:
ZR = r r - - (14)
A
where Wo is the original waist radius and A is the wavelength of the laser
beam.
-3 -2 -1
FIG. 7. Inherent diffraction of a Gaussian laser beam illustrating the concepts of the waist
radius and the Rayleigh range.
°i ~'°
i z z!
t
$ 8'
FIG. 8. Focusing of a Gaussian laser beam by a thin lens.
The propagation of a Gaussian laser beam through an optical system has

been studied by Kogelnik and Li36; they used a complex beam parameter,
matrices, and Kogelnik's " A B C D " law. Self, 37 however, has put the basic
results in a simple form so that it is possible to calculate the positions and
sizes of the various waists in a manner similar to that of ordinary ray tracing.
In his paper Self treated only thin lenses with the same medium on both
sides of the lens, but it can readily be shown from Kogelnik's results that
Self's forms of the equations also hold for thick lenses, provided that
waist object and image distances and focal lengths are measured from the
principal planes. Self's equations follow directly from Kogelnik's and are,
therefore, just as rigorous. The procedure is as follows.
The Gaussian beam is followed through each optical lens system, one
lens at a time. The description is in terms of a thin lens, but the same
procedure holds for a thick lens if the waist object and image distances
and focal lengths are measured from the principal planes of the thick lens.
A typical setup is shown in Fig. 8.
The position of the new waist is given by
1 1 1
+ -- - - (15)
s + Z ~ / ( s - y) s' y
where s is the distance from the thin lens to the original waist, positive to
the left and negative to the right; s' is the distance from the thin lens to
the new waist, positive to the right and negative to the left; f is the focal
length of the thin lens, positive for a converging lens and negative for a
diverging lens; and ZR is the Rayleigh range.
36 H. Kogelnik and T. Li, Appl. Opt. 5, 1550 (1966).

37 S. A. Self, Appl. Opt. 22, 658 (1983).
The waist radius magnification is found from
wo' 1
m - - (16)
X/(1
- -
Wo -- s / f ) 2 "Jr-( Z R / f ) 2
The new Rayleigh range is determined from the equation
Z ~ = m2ZR (17)
This procedure is then repeated for the second optical lens system,
carefully using the correct magnitude and sign of the distance from the
new waist to the second optical lens system. The final waist radius from
the first calculation is used as the new waist radius for the second calculation.
First-order ray (plane wave) optics equations are found from Self's
equations by letting ZR be equal to zero.
It should be stressed that the results obtained by ray optics can be
grossly different from the correct Gaussian beam optics calculation, as can
be seen from Fig. 9. For example, if the original waist is placed at the left
focal point of a thin lens, the new waist is not at infinity but is at the right
focal point.
2.S
A
u. - - - i -
2
Q
(3
c
1.5
: ! :
c~
Q
O~ 1
e:
_e
"O 0.5
m
m : Parameter ZrlFI
E
O
0
2:
-0.5
i :
-1
-3 -2 -1 0 2 3 4
Normalized Object Distance (,~F)
F I G . 9. Self's equation: Normalized image distance as a function of normalized object

distance for an optical system with the normalized Rayleigh range as a parameter. All distances
are normalized to the focal length of the optical system, where the medium on both sides of
the optical system is the same.
A typical problem is to know the size and position of an original laser

waist and to desire to find the focal length of a thin lens and where to place
it in order to reduce the size of the waist by a given amount at a given
distance from the first waist. SeWs equations can be reduced to solving a
quadratic equation whose solution is
B+ 1
f= --~ -2-~ V'B 2 - 4 A C (18)
where
A=2m 2-m 4-1; B=-4m2/; C = m Z [ / 2 + ( m 2 + 1 ) 2 Z 2] (19)
The distance of the new waist from the original waist is l.

The position of the lens from the position of the original waist, s, is
found from
s=f 1+ 2 f2J
The position of the new waist from the lens is given by s' -- l - s. For
example, suppose it is desired to reduce a HeNe (helium-neon) laser's
(633 nm) waist radius of 0.5 mm by a factor of 10 at a distance of 1 m.
According to Self's equations, a positive lens of focal length 14.3 cm
should be placed 14.5 cm to the left of the new waist position at 1 m. A
ray-tracing analysis predicts that a positive lens of focal length 8.26 cm
should be placed 9.09 cm from the new waist position at 1 m. As can be
seen, the ray tracing calculation gives answers that are quite wrong for a
Gaussian beam.
Several important concepts must be considered in order for one to have
a reasonable understanding of the laser systems used for photolysis of caged
compounds. These concepts are considered one at a time and then put
together for a description of the laser systems.
Wave Polarization 2°
Maxwell's equations predict that sinusoidal plane waves can be propa-
gated in free space such that the electric, magnetic, and propagation vectors
are all at right angles to each other. It is desired to see what curve the tip
of the electric vector traces out in a plane perpendicular to the direction
of propagation. The electric vector can be decomposed into its components
along two perpendicular directions in that plane, and both components will
oscillate sinusoidally. Maxwell's equations, however, do not predict the
relative phase of these two components. The relative phase is set by initial
[10] LASERS FOR PHOTOLYSIS 191
conditions and can change after passing through various optical compo-
nents. For example, if the two components have relative phase either 0 °
or 180 °, then the wave is linearly polarized and the tip of the electric
vector oscillates along a straight line. If the relative phase between the two
components is 90 ° and the relative magnitudes of the two components are
the same, then the wave is circularly polarized either right-handed or left-
handed, depending on the sign of the relative phase. The general case is
right- or left-handed elliptical polarization. This can be seen from the
following equations for the oscillating electric field components of a travel-
ing plane wave:
Ex = A x cos(r + q~x) (21)

Ey = A y cos('/" + ~0y)
where r = k z - cot; k is the propagation constant, equal to 2rr/A, where A

is the wavelength; and to is the angular frequency, equal to 2rrf, where f is
the frequency. A x and Ay a r e the amplitudes of the two components, and
~x and ~y are the phases of the two components. That these equations
represent traveling waves can be seen from the following discussion: A
point on the sinusoidal wave is to be followed in space as a function of
time. Now, setting the absolute phase ~x equal to zero for simplicity,
the x component of the electric field has a space and time dependence,
E x ( z , t ) = A x cos(kz - cot). If a point (phase) of the wave is followed, then
at a time t + At the wave point is at z + AZ. This requires that cos[k(z +
Az) - oJ(t + At)] = cos(kz - cot). If the cosines are equal, then the angles
must be equal, leading to Az = (oJ/k)At. However, o / k = (2rrf)/(2~r/A) =
Af = v, the speed of the wave in the material. Hence Az = v A t and the
particular point being followed on the wave has moved a distance in the
positive direction vAt. This type of wave is called a forward wave.
Recalling the trigonometric relations
sin2A + cosZA = 1
cos(A + B) = cos A cos B - sin A sin B (22)
sin(A - B) = sin A cos B - cos A sin B
and eliminating ~"between the two equations (21) leads to Eq. (23) for the
tip of the electric vector:
Ax] kAy/ Ax cos ~0 = sin2~p (23)
where ~o = q~r - ~0, is the relative phase difference between the two compo-
nents. These results can also be used for Gaussian laser beams.
1
F
0.9 -~
0.8
_~ 0.7 nl/n2 = .667
W
r,
_e :
R
,~ 0.5 R~
'~0 0.4 :
0.3
0.2
0.1
oi
0 10 20 30 40 50 60 70 80 90
Theta (degrees)
FIG. 10. Reflectivity of an air-glass interface for the component in the plane of incidence,
Rp, and the component perpendicular to the plane of incidence, R~.
Reflection and Refraction

The laws of reflection and refraction for plane waves at an interface
between two dielectrics are well known and can still be used for Gaussian
laser beams because the additional effects of the finite cross section of the
laser beam are small and need to be considered only if there are a large
number of multiple reflectionsY Thus, the law of reflection (the angle of
reflection is equal to the angle of incidence; O] = O0) and the law of
refraction (the ratio of the sines of the angle of refraction and the angle
of incident is constant and equal to the ratio of index of refraction of the
first medium to that of the second medium: Snell's law, sin O2/sin O0 =
nl/n2) can be used. The reflectivity of the component of the electric field
perpendicular to the plane of incidence, Rs, and the reflectivity of the
component of the electric field in the plane of incidence, Rp, c a n be found
from Maxwell's equations and the appropriate boundary conditions for this
problem, yielding the Fresnel relations. Figures 10 and 11 show the situation
for the air-glass case and the glass-air case, respectively.
A striking result is that for the component of the electric field in the
plane of incidence, there is no reflectivity at a certain angle, the Brewster
angle, tan O0 = nz/na This means that unpolarized light reflected at the
Brewster angle will be polarized perpendicular to the plane of incidence.
38 y. M. Antar and W. M. Boerner, Can. J. Phys. 52, 962 (1974).

[ 101 LASERSFOR PHOTOLYSIS 193
0.6
0.5
® 0.4
.~ 0.3
o
i Rs
0.2
0.1
Y:
0
0 10 20 30 40 50 60 70 80 90
Theta (degrees)
FIG. 11. Reflectivity of an glass-air interface for the component in the plane of incidence,
Rp, and the component perpendicular to the plane of incidence, Rs.
There is also a Brewster angle for the case of going from glass to air at
the complement of the former Brewster angle. An additional phenomenon
appears in going from glass to air, as shown in Fig. 11, that of total reflection
giving rise to a critical angle, sin O0 = n2/nl. Due to refraction, a beam of
light passing through a plate with parallel sides is displaced laterally from
its original direction. In order to achieve high linear polarization using the
Brewster effect, several plates must be used, leading to an unacceptable
lateral displacement. One can, however, arrange the plates so that half are
pointing in one direction and the other half in the other so that the lateral
displacement is in the opposite direction from the first, and the beam
is returned to its original direction. Such a polarizing device is called a
compensating Brewster stack.
If the incident angle is greater than the critical angle, there is total
reflection, but this is for the situation when the intensities are averaged
over one cycle of the wave. Maxwell's equations show that there actually
is a surface wave that has an exponentially varying amplitude into the air,
which can be detected and utilized. This is called the evanescent wave, and
the process is termed frustrated total reflection. This effect is used, for
example, to extract energy from a ring prism laser, as a light modulator,
and has been able to excite molecular fluorescence. A necessary condition
for the first two applications is that a second transparent plate, which is
flat to within a fraction of a wavelength, must be brought to a distance of
about a wavelength parallel to the original glass surface, which is also flat
to a fraction of a wavelength.
If a wave arrives at the air-glass surface of a glass plate in air, the

reflected wave, according to the Fresnel relations, will have a phase shift
of 180 °, whereas the transmitted wave, which reflects from the glass-air
surface, has no phase shift. This second wave is transmitted back through
the upper surface, and the two waves can then interfere coherently. What
actually happens for a wave of wavelength A in the glass depends on the
path length in the glass because an additional phase shift occurs for the
second wave because of the (2rr/A)z term in the phase ( k z - tot) of the
traveling wave. For a proper thickness there can be constructive interference
with greater reflection or destructive interference leading to little reflection.
If multiple, thin film layers are used with materials of the proper indices
of refraction and proper thicknesses, then almost totally reflecting mirrors
(multiple dielectric mirrors) can be constructed at a given wavelength, or
antireflecting (AR) coatings can be deposited on various optical surfaces
to achieve very little reflection at a given wavelength.
B i r e f r i n g e n c e 2°
If Maxwell's equations are solved for possible electromagnetic waves

in an anisotropic medium, such as a crystal, it is found that in any given
direction there are two traveling waves possible. These waves are linearly
polarized and their electric displacement vectors are at fight angles to each
other, but they differ in a much more profound way. The phase speed of
one of the waves is the same in all directions, whereas the phase speed of
the other plane wave is dependent on the direction. The first wave is called
an ordinary wave because it acts just like a wave in an isotropic medium,
and the second wave is called the extraordinary wave because it is different.
In the general case, one finds that there are two specific directions in a
crystal in which the speeds of the waves are equal. The crystal is then called
a biaxial crystal; however, some crystals have only one direction in space
where the speeds are the same. In this case they are called uniaxial crystals,
with the special direction termed the optical axis of the crystal. Uniaxial
crystals are the crystals used most in optics. If light impinges on a uniaxial
crystal surface whose optical axis is not perpendicular to the surface, then
in a plane perpendicular to the surface the expanding Huygens wavefronts
are circular for the ordinary wave but elliptical for the extraordinary wave,
leading to two physically separated waves in the crystal. On leaving the
crystal, the two waves carry two images of the same original object. This,
then, is called double refraction. A typical demonstration of double refrac-
tion is to place a calcite crystal on top of a newspaper so that two distinct
images of the newsprint are seen.
If the two perpendicular traveling waves in the crystal have different

coefficients of absorption, then unpolarized light, when decomposed into
the two components, becomes partially or wholly polarized, depending on
the length of the crystal. This effect is called dichroism. Specially treated
materials have been made that are also dichroic, such as polaroid sheets
as polarizers, or dichroic films on substrates as dichroic mirrors.
Q-switching 13
For resonant systems, a variable is used to express the quality of the
resonant system, called the quality factor, Q, which is defined as the ratio
of the average energy stored to the energy lost in one cycle. This definition
applies to lumped parameter LCR circuits, microwave resonant cavities,
and laser optical cavities. The Q of the circuit or cavity is frequency de-
pendent, and it is usually desired to have a high Q at a given frequency.
For the laser optical cavity, it is necessary to align the cavity optically so
that efficient feedback can occur, leading to high stimulated emission and
laser action. The losses in the cavity must be minimized so that the cavity
has high Q. This can be done in different ways; for example, the cavity
can be misaligned while the inversion population is maximized and then
realigned in a short time. This is called Q-switching and can be produced
by rotating either a mirror or a totally reflecting prism and synchronizing
the pumping process.
Q-switching can also be achieved by using a compensating Brewster
stack and a voltage-controlled birefringent device called a Pockels cell,
which can rotate the line of linear polariation of the beam after passing
through the cell. The basic laser of Fig. 6 is modified by placing the Pockels
cell at the totally reflecting mirror, followed by the Brewster stack (to
linearly polarize the laser beam). The Pockels cell can be operated so that
initially a high voltage causes the crystal (KD*P, potassium dideuterium
phosphate) to be birefringent such that the line of polarization is rotated
90° after two passes through the cell. The line of polarization is then at
right angles to the original line of linear polarization, and the wave energy
is dissipated in the Brewster stack by subsequent reflections. This condition
is held while the flash lamp pumps the rod to achieve a high inversion
population; at the appropriate time the high voltage on the Pockels cell is
cut off, making the crystal transparent, and the laser light comes out in
several round trips of the cavity, yielding a giant Q-switched pulse of the
order of several Joules in tens of nanoseconds, depending on the length of
the optical cavity. The peak power density of this pulse is high enough to
achieve reasonable frequency doubling.
196 CAGED COMPOUNDS [ 1O]
F r e q u e n c y Doubling a9
When an electric field is applied to a crystal, the center of the negative
charge of the electrons, of otherwise neutral atoms, is displaced from the
center of positive charge in the nucleus. This creates what is known as an
electric dipole, which has a butterfly type of electric field pattern. The
vector quantity consisting of the product of the magnitude of the charge
of the nucleus and the vector displacement of the two charge centers is
called the electric dipole moment of the atom. The net electric vector dipole
moment per unit volume is called the electric polarization, P, of the crystal
(not to be confused with wave polarization). This variable obviously de-
pends on the applied electric field, ~, but in an anisotropic crystal P is not
in the same direction as E. Because the electrical interaction is nonlinear
in an atom, the components of P can be expressed as power series in the
components of E. The electric field in a laser beam is strong enough so
that the nonlinear terms in the electric polarization can be significant. In
fact, the first nonlinear terms, the terms quadratic in the electric field
components, are used in frequency doubling. Maxwell's theory predicts that
an accelerating charge will radiate electromagnetic energy, and therefore an
oscillating electric dipole will also radiate electromagnetic energy. If the
original electric wave is oscillating at a frequency to, the electric polarization
also oscillates, not only at the frequency to but also at higher harmonics.
The way this takes place can be seen from the following simple discussion.
The original electric wave is assumed to be oscillating at frequency to, so
that at a given point in the crystal the time dependence of the electric field
is E ( t ) = E cos tot. For simplicity the relationship is given for an isotropic
crystal. Thus the first nonlinear term in the electric polarization is given by
/'2 =/3E 2 cos2tot (24)
where/3 is the second-order polarizability coefficient.

Recalling the trigonometric relation cos2A = (1/2)(1 + cos 2A) leads to
/3E 2 /3E 2
P2 = T + T cos 2 tot (25)
There is thus a part of the electric polarization oscillating at the frequency

2to, which leads to an induced electric wave at twice the frequency of
the original frequency. This is called frequency doubling. Because of the
relationship between frequency and wavelength, Af = v and to = 27rf, the
new electromagnetic wave is at half the wavelength of the original wave.
For example, it is said that the output of a ruby laser at 694 nm is frequency
39A. Yariv, "Quantum Electronics,"Wiley,New York, 1967.

[1 O] LASERS FOR PHOTOLYSIS 197
1.64 o
1.62
1.6
1.58
1.56
n 1.54
1.52
1.5 %
1.48 ~" ..,.
1 . 4 6 . . . . . . . . . . . . . .
200 400 600 800 1000 1200
wavelength (nm)
F~o. 12. Dependence of the index of refraction on wavelength for the ordinary wave (solid
line) and the extraordinary wave (dashed line; at 90° from the optical axis) for a KDP crystal
at a temperature of 24.8°.
doubled to 347 nm. A n anisotropic crystal, however, must be used to generate

the second harmonic. The original wave (fundamental) is introduced into the
crystal as an ordinary wave, and the generated second harmonic wave is an
extraordinary wave at right angle to the fundamental. The speed of the ordi-
nary wave is the same in all directions, but the speed of the extraordinary
wave varies with direction. For efficient transfer of energy between the funda-
mental and the second harmonic to take place, the speeds of the two waves
must be the same. This means that the indices of refraction must also be the
same because n = c/v, where c is the speed of light in a vacuum and v is the
speed of light in the m e d i u m for a given wave. The indices of refraction are
functions of frequency, and it turns out that for certain uniaxial crystals a
direction in space can be found such that the index of refraction for the ordi-
nary wave at frequency oJ is the same as the index of refraction of the extraor-
dinary wave at frequency 2 to. For a given uniaxial crystal this condition occurs
for a direction in space at a definite angle with respect to the direction of the
optical axis. In Fig. 12 it is seen that for a potassium dihydrogen phosphate
( K D P ) crystal, the index of refraction decreases with wavelength for both
the ordinary wave and the extraordinary wave. n° Because the index of refrac-
tion of the extraordinary wave is less than that of the ordinary wave, the
crystal is called a negative uniaxial crystal. For data taken in Fig. 12, the crystal
40F. Zernike, JOSA 54, 1215 (1964).

was at 24.8 ° and the light was propagating through the crystal at an angle of
90 ° with respect to the optical axis. It can be seen that for this case both the
ordinary wave at 500 nm and the extraordinary wave at 250 nm have the same
index of refraction, 1.515. If the crystal is transparent at both wavelengths,
then frequency doubling can occur. This process is called index matching,
and the crystal is said to be "angle tuned." For different angles of the beam
with respect to the optical axis, the index of refraction of the extraordinary
wave changes and the curve moves up in the figure until at 0 ° the two curves
become one. Laser light of wavelengths above 500 nm can be frequency
doubled by KDP for some angle between 0 ° and 90 ° as long as the crystal is
transparent at both wavelengths. The angle must be set to within about a
tenth of a degree from the correct angle if reasonable second harmonic
output is to be obtained. Because the indices of refraction are temperature
dependent, the crystals need to be temperature stabilized. Some of the
crystals used, such as KDP, are hydroscopic, so they must also be kept in a
sealed, dry nitrogen housing. Because matching the indices of refraction
depends on both angle and temperature, the angle can be left fixed and the
temperature changed to obtain index matching. This technique is called
temperature tuning. The index matching angle may be calculated for
an initial laser beam from data like those given in Fig. 12 and Eq. (26), 39
sinZOm [ (n~)-2 - (n~)-2 ]

= [(need)_ 2 ~ j (26)
For example, the index matching angle for ruby light (694 nm) and KDP
is 50.4 °. The description given earlier is called first index matching. Other
conditions give rise to second index matching and vector index matching.
Mode Locking 13
For an aligned optical cavity, the boundary conditions on the electro-
magnetic field determine that only standing waves of certain wavelengths
can exist in the cavity (this is similar to the situation for sound standing
waves in an organ pipe). For example, a half-wave can be fitted in as the
largest wavelength possible and hence the lowest frequency or fundamental,
o~f. A full wave corresponding to twice the frequency of the fundamental
can also fit, as well as higher harmonics. The difference in frequency be-
tween modes is Ao) = toe. These modes of oscillation are called longitudinal
modes because the fitting is along the optical axis, in contrast to the trans-
verse modes described earlier. The lasing ion or molecule of a laser does
not lase at a precise frequency but has a certain spread of frequencies over
which gain can be achieved. This spread in frequency of the gain curve
[ 10] LASERSFOR PHOTOLYSIS 199
covers many of the longitudinal mode frequencies, and in general these

longitudinal modes are excited with random phases. If some means could
be found so that the longitudinal modes could be made to oscillate with
definite phase relations to each other, then the electric fields would be
coherent, leading to constructive interference and very sharp pulses in time.
How this might work can be seen from the following simple discussion.
If the laser is oscillating at frequency too and some means can be found
to modulate this signal sinusoidally at a frequency tOm, then the output will
have a term of the form
Em(t) = mE cos tOotcos tOmt (27)
where m is the fractional modulation coefficient. Recalling the trigonomet-
ric relation cos A cos B = (1/2)[cos(A + B) + cos(A - B ) ] leads to
Em -- - ~ COS[(tO0+ tOm)t] + --~ COS[(tO0-- tOm)t] (28)
Thus there are fields oscillating at the sum and different frequencies. A
similar situation occurs in radio, where these are called side bands. If the
modulation frequncy, tOm, is set equal to the different frequency between
longitudinal modes, tOf, then the sum and difference signals excite their
adjacent longitudinal modes, which now have a definite phase relation with
the original signal of frequency tOo. These signals are now modulated so
that they excite adjacent modes in phase, and so on. Thus all of the longitudi-
nal modes are oscillating in phase and the fields of the modes are coherent.
The modes are then said to be locked. The coherence of the outputs of
the modes leads to constructive interference and a train of narrow pulses
in time. For example, the Antares n e o d y m i u m - Y A G laser is mode locked
to give 70-psec wide pulses every 13 nsec.
Group Velocity Dispersion

As discussed in the section on frequency doubling, the index of refrac-
tion of a material is a function of frequency, which means it is a function
of the wavelength of the light passing through the material. This dependence
is normally a monotonically decreasing one, as shown in Fig. 12. This effect,
called dispersion, is clearly seen when white light is passed through a glass
prism. Blue light is bent more than red light. This is explained using Snell's
law in the form
sin 02 = -no . Oo
- sm (29)
n2
Because n b > nr, the angle #2v < ~-~2r-
Two velocities (speeds) need to be considered in a homogeneous, iso-
tropic, linear medium. The first is the phase velocity, described earlier,
which gives the speed of a point on a sinusoidal wave; and the second is
the speed of propagation of any modulation of the wave, the group velocity,
which leads to a wave packet containing various frequency plane waves as
a superposition. This can be seen by combining two traveling waves of the
same amplitude but slightly different frequencies and propagation numbers:
E(z,t) = E[cos(k2z - ~o2t) + cos(klz - O~lt)] (30)

Recalling the trigonometric relation cos A + cos B = 2 cos[(A + B ) /
2]cos[(A - B)/2] leads to
where
k2 + kl oJ2 q- 601
kaw = - -2 ; °)ave= 2 ; Ak = k2 - kl; and AoJ = (.0 2 -- (.01
The signal can be interpreted as a modulated carrier traveling wave, with

the carrier having a phase velocity, Vave = O~aw/k.... and the modulation
having a group velocity, Vg = Aoj/Ak. For a wave packet with many frequen-
cies and propagation numbers, the group velocity is given by Vg = doo/dk.
The group velocity can be expressed in terms of the wavelength and the
index of refraction (see Fig. 12) using the relations oJ = kv (from v = Af);
k = 2rr/A and n = c/v. Then 12g d ( k v ) / d k = v + k dv/dk, leading to
=
Vg = v(1 - (A/n)ldn/dA]). Because the absolute value of the derivative,

]dn/dA], is positive, the slopes of the curves of n versus A are negative, and
it is seen that the group velocity is less than the phase velocity. Also if one
compares the group velocity for part of the wave packet in the red region
with that in the blue region using Fig. 12, the red group velocity is greater
than the blue group velocity. Group velocity dispersion depends on the
second derivative of the index of refraction with respect to wavelength.
This means that the part of a narrow pulse corresponding to frequencies
in the red moves faster than the part of the narrow pulse corresponding
to frequencies in the b l u e - - t h e pulse spreads, lowering its peak power
density. This effect is called chirping the pulse. This is a disadvantage for
femtosecond pulses and two-photon photolysis because high peak power
density is desired. It can be overcome by letting the faster, red part of the
pulse travel farther in space than the blue part, thereby bringing the parts
together in space. This is called group velocity dispersion compensation
and is used in femtosecond lasers.
Laser Systems
For one-photon photolysis, a Q-switched, frequency-doubled ruby laser

is the principal laser system used. This system is composed of the basic laser
with Pockels cell Q-switching and frequency-doubling in a KDP crystal, as
discussed earlier.
For two-photon photolysis, a synchronously pumped, group velocity
compensated, liquid-dye laser system with saturable dye is discussed. This
laser is usually pumped by a mode-locked, frequency-doubled neodymium-
YAG solid-state laser. The basic laser configuration of the pump laser
shown in Fig. 6 is changed to a double elliptical cavity with two of the foci
overlapping at the laser rod position and flash lamps at the other loci. A
Gaussian beam (lowest transverse electric and magnetic mode (TEM00))
is achieved by placing a small aperture in the beam, and linear wave polar-
ization is imposed with a Brewster plate. The laser is mode locked by
having the laser beam modulated by passing it through a crystal that has
a high-Q acoustical resonant mode driven by an electronic oscillator with
a piezoelectric transducer to produce acoustic standing waves. The appro-
priate frequency-matching conditions for mode locking are obtained by
first adjusting the temperature of the mode-locking crystal so that a resonant
acoustical mode matches the fixed frequency oscillator and then adjusting
the cavity length so that the longitudinal modes match the modulation
frequency. The output of the mode-locked laser consists of a train of short
pulses with a repetition rate equal to the cavity round-trip transit time.
The output of this solid-state laser is then frequency doubled, for exam-
ple, by a potassium titanyl phosphate (KTP) crystal, by focusing the beam
so that the waist is inside the doubling crystal, and the wave polarization
is oriented with respect to the crystal so that it becomes an ordinary wave
in the crystal. The beam is normal to the face of the crystal, which has
been cut for the proper index matching angle. The crystal angle must be
changed slightly for the corresponding index matching angle. The generated
second harmonic is then an extraordinary wave that is linearly polarized
at right angles to the linearly polarized original primary wave. The primary
wave is then separated from the second harmonic wave using dichroic
mirrors, leaving the second harmonic wave as a mode-locked train of pulses
with the same pulse widths and repetition rate as the primary wave. This
second harmonic wave is then amplitude and position stabilized so that it
can be used as an optical pump source for the dye laser system.
The liquid-dye laser system consists of a laser cavity with a totally
reflecting mirror and a partially reflecting mirror. The lasing medium is a
continuously flowing, flat dye jet ribbon, oriented at the Brewster angle
for the dye cavity. The length of the dye cavity is adjusted so that the
202 CAGEDCOMPOUNDS [ I 11
produced dye laser pulses arrive at the gain jet at the same time as the
pump pulses. This is called synchronous pumping. Two methods are used
to reduce the pulse widths from the mode-locked, approximately 100-psec
width to the desired pulse width of about 200 fsec. The first is to add a
second flat, continuously flowing, saturable absorber dye jet in the dye
cavity oriented also at the Brewster angle and to use dynamic bleaching
so that the front part of the pulse brings the dye to dynamic saturation,
thereby allowing the trailing part of the pulse to pass without absorption,
thus shortening the pulse. The second method of pulse shortening involves
group velocity dispersion. There are several elements in the dye cavity that
the laser beam must pass through: the gain and saturable absorber jets, a
birefringent tuning plate (set near the proper Brewster angle), and several
group velocity dispersion compensation prisms, as well as numerous mir-
rors; all of which result in group velocity dispersion (chirping) and conse-
quent spreading of the laser pulse in time. The prisms are adjusted for the
proper angles and lengths so that the group velocity dispersion is compen-
sated, leading to a dye laser output that has about 200-fsec-wide pulses
with the same repetition rate as the pump beam.
Acknowledgments
The author thanks R. Chillingworth, J. Bunkenburg, and Drs. D. Trentham, T. Bliss, D.
Ogden, N. Kiskin, M. Ferenzi, Y. Goldman, W. Webb, F. Wise, T. Asakura, T. Kihara, and
B. Chance for helpful conversations and support. He also thanks Professor L. Narducci for
critically reading the manuscript. Help with the figures by Anthony De Simone and Paul
Heipp was much appreciated. This work was supported by Drexel University and MRC Grant
E30/533-0300-574.
[1 1] Flash Lamp-Based Irradiation of Caged Compounds

By GERT RAPP
In~oduc~on
The first part of this article contains a brief review of the various light
sources that have been used for the photolysis of caged compounds, together
with a discussion of their advantages and disadvantages. Flash lamp-based
systems are then discussed in detail, with a particular emphasis on the
individual components of these systems, including flashbulbs, the electronic
driving circuit, and optical components. The final section contains a sum-

METHODSIN ENZYMOLOGY,VOL.291 0076-6879/98 $25.00
202 CAGEDCOMPOUNDS [ I 11
produced dye laser pulses arrive at the gain jet at the same time as the
pump pulses. This is called synchronous pumping. Two methods are used
to reduce the pulse widths from the mode-locked, approximately 100-psec
width to the desired pulse width of about 200 fsec. The first is to add a
second flat, continuously flowing, saturable absorber dye jet in the dye
cavity oriented also at the Brewster angle and to use dynamic bleaching
so that the front part of the pulse brings the dye to dynamic saturation,
thereby allowing the trailing part of the pulse to pass without absorption,
thus shortening the pulse. The second method of pulse shortening involves
group velocity dispersion. There are several elements in the dye cavity that
the laser beam must pass through: the gain and saturable absorber jets, a
birefringent tuning plate (set near the proper Brewster angle), and several
group velocity dispersion compensation prisms, as well as numerous mir-
rors; all of which result in group velocity dispersion (chirping) and conse-
quent spreading of the laser pulse in time. The prisms are adjusted for the
proper angles and lengths so that the group velocity dispersion is compen-
sated, leading to a dye laser output that has about 200-fsec-wide pulses
with the same repetition rate as the pump beam.
Acknowledgments
The author thanks R. Chillingworth, J. Bunkenburg, and Drs. D. Trentham, T. Bliss, D.
Ogden, N. Kiskin, M. Ferenzi, Y. Goldman, W. Webb, F. Wise, T. Asakura, T. Kihara, and
B. Chance for helpful conversations and support. He also thanks Professor L. Narducci for
critically reading the manuscript. Help with the figures by Anthony De Simone and Paul
Heipp was much appreciated. This work was supported by Drexel University and MRC Grant
E30/533-0300-574.
[1 1] Flash Lamp-Based Irradiation of Caged Compounds

By GERT RAPP
In~oduc~on
The first part of this article contains a brief review of the various light
sources that have been used for the photolysis of caged compounds, together
with a discussion of their advantages and disadvantages. Flash lamp-based
systems are then discussed in detail, with a particular emphasis on the
individual components of these systems, including flashbulbs, the electronic
driving circuit, and optical components. The final section contains a sum-

[ 111 FLASHLAMP-BASEDIRRADIATION 203
TABLE I
SUITABLELIGHTSOURCESFOR PHOTOLYSIS OF CAGED COMPOUNDS a
Wavelength Approx. pulse Typical pulse

Laser type (nm) width (nsec) energy (mJ)
A. Lasers
Ruby 347 25 100-300
Nd-YAG 355 10 100-200
Ti-Sapphire 350-370 10 90
Dye, laser pumped From 300 10 50
Dye, flash-lamp pumped From 300 1000 100
N2 337 4 0.1-1
XeC1 excimer 308 20 >200
XeF excimer 351 20 >200
B. Arc lamps
Mercury, cwb
Xenon, cwb
Hg-Xe, cwb
Xe flash lamps From 300 1 msec 100
a Compiled from the "Laser Focus Buyers Guide," where the interested reader may find
more detailed information.
bcw, continuouslyoperated lamps.
mary of studies in which flash lamps were successfully applied in the areas
of muscle contraction, protein crystallography, and electrophysiology.
Light S o u r c e s For F l a s h Photolysis of Caged C o m p o u n d s
Photolysis of most caged compounds requires irradiation of the com-

pound with light in the near-ultraviolet (UV) range, from about 300-
370 nm. A compilation of suitable light sources is given in Table I. Important
aspects to be considered are (1) the desired time resolution of the photoacti-
ration reaction, (2) the required energy of the pulse, (3) the sample size
and volume, (4) the experimental setup, and (5) the available budget.
Experiments that require a time resolution better than approximately
100/zsec must in general employ pulsed lasers; these devices are described
in more detail in Refs. i - 3 . Besides the short pulses, lasers have the advan-
tage that their collimated beam can be guided over long distances with
negligible loss of intensity and focused to a very small spot. Disadvantages
1j. A. McCray and D. R. Trentham, Annu. Rev. Biophys. Biophys. Chem. 18, 239 (1989).
2j. A. McCray, Methods Enzymol. 291 [10] (1998) (this volume).
3E. Brown and W. Webb, Methods Enzymol. 291 [20] (1998) (this volume).
204 CAGED COMPOUNDS [1 11
are that lasers are generally bulky, expensive, and require special safety
precautions. In addition, the following points should be considered when
deciding whether to use a laser or a flash lamp:
1. The efficiency of the photolysis of some caged compounds is lower
for lasers than for flash lamps. 1
2. Short pulses may lead to two-photon absorption, 2'3 which causes
ionization of the sample if the incident wavelength is in the near-UV range.
3. If the wavelength is too close to the protein absorption band, the
sample may be destroyed. Experiments on skeletal muscle using a XeCI
excimer laser at 308 nm to photolyze caged ATP showed that the irradiation
caused a physical disruption of muscle fiber bundle. 4 However, no signs
of sample deterioration were observed when the same experiment was
performed using the 351-nm-wavelength line of the XeF excimer laser at
the same energy of 120 mJ. On the other hand, in a study of the kinetics
of the phosphorylation of Na,K-ATPase, 5 an excimer laser running at
308 nm was reported to have successfully released caged phosphate without
damaging the sample.
4. The damage threshold in flash photolysis experiments, in particular
for time-resolved measurements in protein crystals, depends strongly on
the power of the light source, i.e., the time interval within which a sample
is irradiated with a given energy.
Flash L a m p S y s t e m s
The majority of kinetic studies employing caged compounds have been
conducted with a time resolution of milliseconds to seconds. In this time
regime, pulses delivered from flash lamps are generally satisfactory from
the standpoints of both energy and time resolution. Given the importance
of flash lamps in these studies, we describe the essential details of the
components of these devices and provide an account of their performance.
The principle layout of a flash lamp system is shown in Fig. 1. It basically
consists of a capacitor, to store electrical energy, which is discharged over
the flash lamp and optical components to collect the irradiated light.
The Lamp
The bewildering number of definitions and units used to define the
emission properties of lamps complicates any assessment of the perfor-
mance of light sources. Photometry is concerned with the response of the
4K. J. V. Poole, G. Rapp, and R. S. Goody, unpublished.

s H. J. Apell, M. Roudna, J. E. T. Corrie, and D. Trentham, Biochemistry 35, 10922 (1996).
[ 11] FLASHLAMP-BASEDIRRADIATION 205
[:z3
i
R
C
Uo
opticat sample
etements
L
FIG. 1. Schematic layout of a flash lamp system.
human eye to a light source, and most lamps are specified by photometric
quantities. However, for applications in photolysis experiments, radiometric
quantities are more relevant. A compilation of the most important terms
is provided in Table II. Detailed technical specifications are supplied only
for continuously operating (cw) arc lamps. In combination with the use of
a mechanical shutter, cw-operated xenon and mercury lamps may be a
suitable alternative for some applications of caged compounds, provided
a high time resolution is not required. Emission spectra of continuously
operated xenon and mercury arc lamps are shown in Fig. 2. A 150-W xenon
lamp has typical specifications of 3000 lm, 300 cd, and 20,000
cd/cm2. Unfortunately, no useful quantities for the emission in the near-
ultraviolet range can be deduced from these specifications, although the
radiant flux in the wavelength region from 300 to 380 nm may be estimated
from these spectra. Continuously operating arc lamps usually have an effi-
ciency of 60%, and xenon lamps radiate about 6% of their electrical input
power in the UV region below 380 nm. Special care has to be taken to
eliminate the strong infrared irradiation of these lamps. Mercury lamps,
which have strong emission lines between 320 and 370 nm, may be more
suitable for photolysis experiments, although any emission below 300 nm
should be filtered out using suitable bandpass filters. Using a 100-W mercury
lamp about 25% caged Ca 2+ has been photolyzed in approximately 2 sec. 6
Compared to other noble gases, xenon converts electrical energy most
efficiently into light due to its lower ionization energy. The current density
during the discharge is determined by the parameters of the driving circuit
of the lamp. These factors can significantlyinfluence the spectral distribution
of the emission, as discussed in the next section. Different materials for
flashbulbs are available. Pure silica or quartz is transparent down to the
6 R. Zucker, in "Methods in Cell Biology," Vol. 40, p. 31. Academic Press, New York, 1994.
206 CAGED C O M P O U N D S [ 111
(,=,
)=,
.£
~9
<
p,
o II 11
r~
< '7-
Z
<
.1
),=(
o
E
Z
II
m )I ~
~ =
O9 ~1
II II II
8 O
[11] FLASHLAMP-BASEDIRRADIATION 207
1.0
-Mercury
0.5
0.0 I
400 600
Wavelength(rim)
1.0
Xenon
~ 0.5
0.0 t , I , I I i
400 600 800 1000
Wavelength(nm)
FIG. 2. Emission spectra of mercury and xenon arc lamps at continuousoperation.
deep UV. In so-called ozone-flee types, the quartz envelope is doped with
titanium, which absorbs below 250 nm, as shown in Fig. 3. However, one
should keep in mind that, depending on the driving circuit, a substantial
portion of photons harmful to proteins might be irradiated.
The Driving Circuit

Basically the electrical circuit for flash lamps consists of a capacitor to
store energy, which is discharged through the lamp as shown in Fig. 1. In
the nonionized state the flash lamp has impedance of tens of megaohms.
By applying a high voltage of 10-20 kV, xenon atoms ionize and cause the
impedance to drop, thus allowing the capacitor to discharge. It is convenient
to define the pulse length as the width at one-third of its maximum ampli-
tude. Without considering the impedance of the circuit, the duration of the
current pulse is given approximately by
tu3 = ~'(LC) 1/2
100
.,........... - ;~,.-- ....... 7.==,.,,.-~, - ........
-;, i .......
f\air,.."i: "" i" ". ......
.-' ,. ~ ~ ,'
:" / -: ~ ,
!,'C i: " ,'
~ ~i ~ ,'
,~ D~: l F
i
0 , I , I , I , I ,
200 300 400 500 600

Wavelength (nm)
F~G. 3. Spectral transmission of selected materials important for caged compound applica-
tions. (A) Silica glass; (B) fused quartz; (C) quartz light guide, 1 m long, N A = 0.22;
(D) titanium-doped quartz used for "ozone-free" flash lamp envelopes; (E) liquid light guide
for U V applications, N A = 0.45; (F) BK7 optical glass, 25 m m thick; (G) cerium-doped quartz,
used for flash lamp envelopes; (H) standard glass fiber. Note: Reflections at surfaces are not
taken into account.
with an inductance of L and a capacitance of C. Using values of 20/xH

for the inductance and 3500 IxF for the capacitance, the pulse length is
@3 = 830/xsec.
The stored energy E is given by
E = 1/2 CU e
with C the capacitance and U the charging voltage. For a given energy,
reducing the capacitance and simultaneously increasing voltage shortens the
pulse length. However, the peak current increases, which has the following
consequences:
1. The lifetime of the flashbulb decreases due to a more rapid evapora-

tion of electrode material.
2. At very short pulses (< 50/xsec at 200 J electrical) the shock wave
associated with the discharge may lead to an explosion of the bulb.
3. Circuit losses increase and the conversion efficiency of electrical
energy into light is reduced due to the creation of the shock wave.
4. The emission spectrum is shifted toward a shorter wavelength.
[111 FLASHLAMP-BASEDIRRADIATION 209
13 L W (300-380 rim)
AAA
~ 33J
I i I i
400 600
Wavelength (nm)
FIG. 4. Spectral distribution of a xenon flash lamp. Stored electrical energy from 33 to
240 J. There is a shift toward shorter wavelengths on increasing the energy and, hence, current
density. The percentage of radiant flux is shown in the inset for wavelengths 300-380 nm.
The effect of different current densities on the efficiency and spectral

distribution of flash lamps was the subject of an earlier investigation. 7 An
electrical energy of 200 J was discharged in two electrical configurations.
In the first, a bank of capacitors of 500/xF charged to 894 V was used,
which resulted in a pulse of about 130/zsec full width at half maximum
(FWHM). In the second configuration, a capacitor bank of 4500/zF charged
to 298 V resulted in a pulse with a F W H M of about 750/~sec. The spectral
distribution is shifted toward the ultraviolet at these high driving voltages,
i.e., high current density, although the radiant energy over all emission
wavelengths decreased due to the effects discussed earlier. Emission spectra
obtained by operation of the xenon flash lamp at different electrical ener-
gies, and hence at different current densities, are shown in Fig. 4. At an
input of 33 J the spectrum is almost fiat, and at higher energies the center
of gravity of the spectra is shifted toward lower wavelengths with a concomi-
tant increase in the percentage of radiant energy in the UV range.
The Optical System

In many illumination systems ellipsoidal mirrors are used to collect the
maximum amount (up to 80%) of the radiant flux (Fig. 5). In practice, a
point-source lamp is placed at the first focus, and the sample, or one end
of a light guide, is located at the second focal point. A variety of reflector
geometries are available for this purpose. A ray-tracing analysis for a typical
ellipsoidal mirror with a major axis of 135 mm and minor axis of 75 mm
7 G. Rapp and K. Giith, Pfliigers Arch. 411, 200 (1988).

FIG. 5. Ellipsoidal mirror with flashbulbmounted.
is shown in Fig. 6. For an ideal point source, all reflected rays can be collected
at the second focus. However, for a real source with finite dimensions, the
image of the lamp is enlarged by about a factor of between 5 and 10,
depending on the geometry of the mirror, and consequently the irradiance
is reduced by the square of the magnification factor. The magnification
properties of ellipsoidal mirror-imaging systems are illustrated in Fig. 7. It
shows the radius of the cross section containing 50, 70, and 100%, respec-
tively, of the rays reflected from the mirror as a function of distance around
the focus, assuming the source to be a cube of 1 mm 3.
Illumination systems that use refractive optics consist of a condenser,
an objective, and a back reflector. Both the condenser and the objective
are usually assembled from an array of lenses. The ultraviolet transmission
properties of standard glass optics, shown in Fig. 3, are not suitable for
applications with caged compounds, and therefore quartz optics should be
used. With an optical array of five lenses, reflection losses at the air-quartz
interface accumulate to about 34% (4% reflection, 10 interfaces, transmis-
sion T = 0.961° = 66%). With appropriate ultraviolet antireflection coating,
[ 11] FLASHLAMP-BASEDIRRADIATION 211
FIG. 6. Ray tracing with (A) a point source and (B) a real source of 1 × 1 x 1 m m 3 in
one focal point of the ellipsoid.
these losses can be reduced to less than 5% within the excitation bandwidth
required for activation of caged compounds. The main distortions in the
illumination system arise from spherical aberration. Due to chromatic aber-
ration of the lens system, it is important to focus the beam using only UV-
filtered light. Illumination of the sample using white light (1) results in
heating of the preparation, which is certainly harmful for protein crystals
and for photoactive samples, and (2) may cause mechanical artifacts due
212 CACZD COMPOUNDS [ 1 1]
50 %
8 -
...... 70%
......... 100 % ."
.c
-...
-10 -5 0 5 10
Distancefromfocus(mm)
FIc. 7. Caustic of an ellipsoidal mirror using a real source of 1 mm3. The diagram shows
the spot size containing 50, 70, and 100% of the reflected rays around the secondary focus,
respectively, e.g., with the source of 1 mm3 100% of the reflected rays are within a spot of
about 9 mm in diameter at the back focal plane.
to m o m e n t u m transfer of light. The second effect is a well-known p r o b l e m

in patch-clamping experiments. 6 A U G l l is a suitable U V filter, and its
residual transmission in the near infrared (IR) can be blocked with a special
coating. When necessary, harmful U V light at shorter wavelengths can be
eliminated using an additional long-pass filter.
It should be kept in mind that the following relationship is an invariant
of any optical system ( H e l m h o l t z - L a g r a n g e invariant):
Asf~sne = Ai~'~ini
where A is the area of the source and its image, f~ is the solid angle
determined by entrance and exit aperture, respectively, and n is the index
of refraction; usually ne = ni. In other words, the radiance L is determined
by the radiating source and the first optical element. It cannot be increased
and, in practice, it is normally reduced. Optical elements can increase only
the irradiance, 1. One might think that the ideal system to collect m o r e
photons consists of a tapered fiber optic with a large aperture Ai at the
input and a small aperture Ae at the exit.
Because Ai/Ae = NAe/NAi = constant, light is emitted f r o m the smaller
area with a larger divergence, and so, in effect, the task of collecting m o r e
light is shifted to the next optical element (NA, numerical aperture = n
sin 0, with 0 the half-angle at the input and exit, respectively).
In an earlier investigation, 7 a comparison was made of the energy density
in the focus using refractive optics versus ellipsoidal mirrors. As expected
f r o m the ray-tracing calculations, the irradiance is higher if refractive optics
of short focal length are used compared to the ellipsoidal mirror system.
In some applications though, the shorter working distance associated with
a short focal length imposes technical problems. In these circumstances, or
if samples under investigation are larger than about 3-4 mm (e.g., muscle
fibers), an ellipsoidal mirror should be considered as an optical component.
Several reasons favor the use of light guides to deliver the light pulse to
the sample: (1) Discharge lamps often create electrical artifacts in sensitive
electronic equipment (e.g., amplifiers used in electrophysiology) if they are
located close to the flash lamp; (2) the sample stage may be very sensitive
to mechanical vibrations and must be isolated from the lamp housing; and
(3) the space around the sample is limited. For applications using caged
compounds, only quartz fibers or liquid light guides are suitable (see Fig.
3). Single quartz fibers are available up to about 1.5 mm in diameter with
a NA up to 0.25. Standard liquid light guides have diameters of 2, 3, and
5 mm, with a NA of 0.45 or larger. Here, we are again faced with the
problem of conserving brightness. The light-coupling efficiency of the light
guide is much higher at the input side for liquid light guides compared to
quartz fibers. It is, however, a challenge to collect and collimate or refocus
light emitted from the liquid light guide, especially when large diameters
are used. The optimal choice depends on the experimental conditions, and
the best solution has to be determined for each case.
Light guides are frequently used to deliver light into microscopes for
the reasons discussed earlier. For inverted microscopes, light is often cou-
pled via the epifluorescence port. Quartz objectives are transparent over
the entire wavelength range important for photolysis of most caged com-
pounds, i.e., 300-370 nm, but they are expensive. Many experiments have
been reported (see later) in which UV-transparent objectives have been
used successfully. The transmission curves for these objectives are not
available from any supplier, although suppliers claim their transmittance
declines only around 340 nm. Note, however, that below 340 nm absorption,
and hence photolysis yield, increases for most caged compounds.
Methods to Measure the Energy of a Flash Pulse

Calibrated joulemeters are easy to use, commercially available, and
have an accuracy of about 5%. They have a large dynamic range (from mJ
to several joules) and can either be directly attached to a storage oscillo-
scope or may have special readout electronics. The detector head consists
of a coating, which has a uniform absorption from the UV to the infrared.
A voltage proportional to the absorbed energy is generated at the detector,
which can easily be recorded on a storage oscilloscope. Provided the pulse
length of the flash is shorter than the response time of the joulemeter,
energy can be calculated from the p e a k voltage and the calibration factor
supplied from the manufacturer. An inexpensive detector may be built
using a UV-sensitive silicon photodiode. However, in order to obtain abso-
lute values of radiant energy, one must take into account the nonlinear
radiant sensitivity of the diode and the time structure of the flash.
Appfications
The following section summarizes applications wherein flash or continu-

ously operated arc lamps were used for photolysis of caged compounds.
This review is biased toward references that contain a m o r e detailed descrip-
tion of technical aspects of the experiments. Reviews on the application of
flash lamps for photolysis of caged compounds, include physiological and
pharmacological experiments 8'9 and studies on smooth muscle 1° and stri-
ated muscleJ 1
S t r u c t u r a l S t u d i e s on N o n c r y s t a l l i n e S y s t e m s
X-ray diffraction is an ideal tool for determining the internal structure of
macromolecular assemblies such as muscle fibers or microtubules. Whereas
laboratory X-ray generators require exposure times on the order of minutes
to hours, these times have been shifted to the millisecond region or even
less using synchrotron radiation. In the first experiments of this type with
caged compounds, 12 small strips of muscle fibers were incubated in a rigor
solution (i.e., without ATP, see also Dantzig et aL 13) and then relaxed by
photolytic release of A T P from caged ATP. Further investigations along
this line were described, 14,a5 wherein a flash lamp was employed to induce
photolysis of caged ATP, caged Ca 2+, or caged nucleotide analogs involved
in muscle contraction. Experiments on muscle mechanics have shown that
the relaxation rate from rigor induced by photolysis of caged A T P depends
markedly on the ionic strength. 16 Diazo-2, a photolabile derivative of
8 H. A. Lester and J. M. Nerbonne, Annu. Rev. Biophys. Biophys. Eng. 11, 151 (1982).
9 A. M. Gurney and H. A. Lester, Physiol. Rev. 67, 583 (1987).
10A. P. Sorely and A. V. Somlyo, Annu. Rev. Physiol. 52, 857 (1990).
11E. Homsher and N. C. Millar, Annu. Rev. Physiol. 52, 875 (1990).
12G. Rapp, K. J. V. Poole, Y. Maeda, K. Giith, J. Hendrix, and R. S. Goody, Biophys. J. 50,
993 (1986).
13j. Dantzig, H. Higuchi, and Y. E. Goldman, Methods Enzymol. 291 [18] (1998) (this volume).
14K. J. V. Poole, G. Rapp, Y. Maeda, and R. S. Goody, in "Molecular Mechanism of Muscle
Contraction" (H. Sugi and G. H. Pollack, eds.), p. 391, Plenum, New York, 1988.
is K. Horiuti, K. Kagawa, and K. Yamada, Biophys. J. 67, 1925 (1994).
a6C. Veigel, M. R. D. Maydell, R. Wiegand-Steubing, R. S. Goody, and R. H. A. Fink, Pfl~gers
Arch. 43@,994 (1995).
[11] FLASHLAMP-BASEDIRRADIATION 215
B A P T A that converts from a low to a high affinity chelator of calcium ions

after photolysis, was used to study the calcium-dependent relaxation rate
in intact striated muscle fibers 17 and in cardiac muscle. TM Caged nucleotides
were also used to measure the nucleotide-exchange kinetics of isolated
synthetic myosin filaments using fluorescence microscopy and flash pho-
tolysis. 19
Flash lamp-induced photolysis of caged A T P and caged A M P P N P was
used with time resolved cryoelectron microscopy in order to study the
dissociation of actomyosin. 2° Grids containing a suspension of sample were
irradiated at given times (10-100 msec) prior to a rapid freezing. To avoid
excess heating of the grid, it was important to limit the bandwidth of the
flashlight to a narrow region between 320 and 370 nm.
Microtubules (MTs), elements of the eukaryotic cytoskeleton, consist
of polymers of tubulin dimers. Microtubules are highly dynamic structures
that undergo catastrophic depolymerization when the fast-growing end of
the filament contains a certain number of G D P - b o u n d tubulin molecules;
the recovery of M T growth requires polymerization of dimers in the GTP-
bound form. Oscillations in the growth and collapse of MTs have been
induced by flash photolysis of caged G T P and investigated by X-ray scatter-
ing. 21 The authors of this particular study also showed that for each absorbed
X-ray photon (8 keV) about 20 caged G T P molecules were activated, of
which 7 were legitimate G T P molecules while the rest were nonfunctional
fragments of caged G T P or GTP.
Protein Crystallography
For a detailed review of the application of caged compounds in time-

resolved crystallography, see Scheidig et al. 22 The structural dynamics of a
GTPase in crystals was studied using flash photolysis of caged GTP. 23 In
X-ray diffraction studies, the measured diffraction intensities arise from
spatially and temporarily averaged electron densities in the crystal. There-
fore, to study the kinetics of a reaction in the crystal, the light pulse should
photolyze 100% of the caged nucleotide. This efficiency is difficult to
achieve, and even if a light source of sufficient energy were available to
17j. L~nnergren and A. Arner, Z Muscle Res. Cell. Motil. 13, 630 (1992).
18R. Zang, J. Zhao, A. Mandveno, and J. D. Potter, Circ. Res. 76, 1027 (1995).
19p. B. Conibear and C. R. Bagshaw, FEBS Lett. 3811,13-16(1996).
2oj. F. Menetret, W. Hoffmann, R. R. SchrOder, G. Rapp, and R. S. Goody, J. Mol. Biol.
219, 139 (1991).
21A. Marx, A. Jagla, and E. Mandelkow, Eur. Biophys. J. 19, 1 (1990).
22A. ScheidigC. Burmester, and R. S. Goody, Methods Enzymol. 291 [14] (1998)(this volume).
I. Schlichting,G. Rapp, J. John, A. Wittinghofer,E. Pal, and R. S. Goody, Pro¢. Natl. Acad.
Sci. U.S.A. ~o, 7687 (1989).
216 CAGZD COMeOUNDS [ 1 11
%.
O"~ 10 , "c21 0.3 sec,
0 , I
1 2
Time(see)
Fro. 8. Time course of heat dissipation in a protein crystal in an aqueous environment
(r = 5.9 see<) andin air 0" = 0.3 see-l).
achieve this photolysis level, the thermal load sets the limitations. Indeed,
heating of the crystal is the limiting factor in protein crystallography. The
following example derives estimates of the temperature increase and the
subsequent temperature reequilibration times in a protein crystal due to
photolysis of a caged compound.
Consider a crystal of the protein Ras p2124 of molecular mass 21 kDa,
with a dimension of 300 x 300 x 500 p~m3 and a solvent content of 40%.
At a protei n density of 1.23 g/cm 3 there are 9.2 x 1014 protein molecules
in the crystal. To release the same number of molecules of GTP from caged
GTP in this crystal would require 4.6 × 1015 photons, assuming a quantum
efficiency of 20%, which is equivalent to the energy of 2.6 mJ for a mono-
chromatic source at 347 nm (E = n h c / h ) . A t 347 nm, the wavelength of
the ruby laser, the extinction coefficient of caged ATP ~cagedATP = 5 0 0 M -1
cm -1 (at 320 nm ~cagedATP = 1000 M -1 c m 1). That is, to photolyze 100%
of the caged GTP in our example, 4 mJ or 27 mJ/mm 2 are required. In
practice, however, even more energy is required for photolysis than that
calculated here.
Taking 1.5 J/g °C for the heat capacity of a protein, 25 the Ras p21 crystal
(33 /zg protein + 18/xg water) is heated by 21 °. The amount of energy
24I. Schlichting, S. C. Almo, G. Rapp, K. Wilson, K. Petratos, A. Lentfer, A. Wittinghofer,

W. Kabsch, E. F. Pai, G. A. Petsko, and R. S. Goody, Nature 345, 309 (1990).
25G. I. Makhatadze and P. L. Privalov, J. MoL BioL 213, 375 (1990).
[ 111 FLASHLAMP-BASEDIRRADIATION 217
TABLE III
PHOTOLYSIS EFFICIENCYOF CAGED ATP IN CRYSTALSOF GLUTATHIONE SYNTHETASEVS
ENERGY DENSITY USINGA N d - Y A G LASERAT 355 nm a
Energy density Number of Photolysis

(mJ/mm 2) flashes (%) Comment
63 1 17 Crystal was burned

31 1 12 OK
31 10 33 Crystal and capillary were cracked
20 30 90 Crystal was cracked
10 60 50 OK
a Data from H. Kato, Institute for Chemical Research, Kyoto University, Japan, personal
communication (1997).
dissipated as heat is calculated following an approach described by Hol-

man. 26 For simplification we assume the crystal is instantaneously heated
by 21 ° (which is certainly valid for nanosecond laser pulses) and remains
suspended in an environment at the preflash temperature. The kinetics of
cooling of the crystal in its mother liquor and air are shown in Fig. 8. These
data are calculated assuming heat transfer coefficients of 1000 and
50 W/m2/°C for water and air, respectively. From this example it is clear
that photolysis of 100% of the caged compound with one intense flash leads
to significant heating in the crystal. This heating effect is of even more
concern in crystals that have a low solvent content, and it takes seconds
for crystals suspended in air to dissipate the absorbed heat. This heating
effect must be considered in time-resolved studies of enzyme reactions in
crystals where the rate of the reaction is temperature dependent.
In practice, photolysis of caged compounds in crystals of p2123'24 and
phosphorylase b 27 require several flashes to achieve sufficient activation of
the substrate. For glutathione synthetase, 28 the effect of the energy density
on the efficiency of photolysis and survival of the crystals was investigated
using a N d - Y A G laser at 355 nm and a xenon flash lamp. 29 The major
differences between these photolysis sources are (1) 15 nsec versus 1 msec
pulse length and (2) monochromatic versus broad wavelength band.
Data shown in Tables III and IV were kindly provided by Dr. Hiroaki
Kato. 29 In these experiments the efficiency of photolysis was determined
26 j. p. Holman, "Heat Transfer," McGraw-Hill, New York, 1986.

27 E. M. H. Duke, S. Wakatsuki, A. Hadfield, and L. N. Johnson, Protein Sci. 3, 1178 (1994).
zs H. Yamaguchi, H. Kato, Y. Hata, T. Nishioka, A. Kimura, J. Oda, and Y. Katsube, J. Mol.
BioL 229, 1083 (1993).
29 H. Kato, Institute for Chemical Research, Kyoto University, Japan, personal communica-
tion (1997).
TABLE IV
PHOTOLYSIS OF CAGED A T P IN CRYSTALS OF GLUTATHIONE
SYNTHETASE USING A FLASH LAMP WITH A UV FILTER
DEPENDING ON (A) ENERGY DENSITY AND
(B) NUMBER OF FLASHES AT 18 m J / m m 2a
A B
Energy density Photolysis Number of Photolysis

(mJ/mm 2) (%) flashes (%) (SD) b
4 25 1 68 (5.7)
9 30 2 95 (4.8)
12 37 3 94 (3.9)
18 68 5 100 (nd)
a Data from H. Kato, Institute for Chemical Research, Kyoto

University, Japan, personal communication (1997).
b SD, standard deviation; rid, not determined.
by high-performance liquid chromatography (HPLC) analysis of the crystal

after its dissolution in buffer. At a very high energy density (63 mJ/mm 2)
one shot destroys the crystal, and even with this high energy density only
17% of the caged ATP is photolyzed. A similar investigation using a flash
lamp equipped with a U G l l filter is summarized in Table IV. Here much
lower energy densities are required to produce the activated substrate: at
18 mJ/mm 2, 68% of caged ATP was photolyzed in one flash, whereas four
to five flashes at 18 mJ/mm 2 are found to release almost 100% of the
substrate. These two experiments demonstrate the dependence of the effi-
ciency of photolysis on the pulse length and bandwidth, a property already
addressed by McCray and TrenthamJ A second potential problem using
lasers with high energy density is the generation of a shock wave created
by the rapid heating, which might be particularly harmful for protein crystals
but does not appear to cause obvious damage to other samples, such as
cells.
Binding of the inhibitor pyrone to chymotrypsin is initiated by flash
photolysis of a protecting group covalently bound to the enzyme (caged
enzyme). 3° Mosaicity in the crystals increases after photolysis, as indicated
by streaks in the diffraction pattern. Laue crystallography, i.e., using white
light, is more susceptible to this effect than standard Bragg diffraction.
Increased mosaicity may arise from the temperature jump due to the flash
or from major reorganizations in the crystal associated with the start of
the reaction.
3o B. L. Stoddard, P. Koenigs, N. Porter, K. Petratos, G. Petsko, and D. Ringe, Proc. Natl.

Acad. Sci. U.S.A. 88, 5503 (1991).
In all experiments it is important to know the concentration of the

molecules of interest after photolysis. To achieve this goal the amount of
photolytically released phosphate from caged phosphate is directly mea-
sured in a crystal of phosphorylase b using a diode-array spectrophotome-
ter. 27 In these studies, 7 m M phosphate was released from 21 m M caged
phosphate after five fashes.
F u r t h e r Applications
The final part of this contribution summarizes a variety of further experi-
ments in which flash lamps or continuous arc lamps equipped with a shutter
were used. From the sound body of publications, only those articles were
selected in which details of the experimental setup are described.
1. In studies on chemotaxis, the swimming behavior of bacteria was

investigated by video microscopy following photolytic release of serine and
protons from their caged precursors. High time resolution was not required,
and light from a mercury lamp switched on for 30 msec was sufflcientJ I
2. Confocal microfluorimetry was used to measure intracellular-free
[Ca 2÷] induced in X e n o p u s oocytes by photoreleased inositol trisphosphate.
A 100-W mercury lamp with a shutter attached to the epifluorescence port
of an inverted microscope was used for photolysis. A small response was
observed with 15 msec opening; with 60-msec flashes calcium liberation in
oocytes persisted for several seconds. 32 Local increases in the calcium ion
concentration around the nucleus have been found to trigger entry into
mitosis in experiments combining flash photolysis of the calcium chelator
nitrophenyl-EGTA and confocal microscopy. 33
3. Regulation of sarcoplasmatic Ca 2+ release was investigated following
photolysis of caged cyclic ADP-ribose in intact cardiac myocytes. 34 In an
inverted microscope with quartz optics, light for photolysis from a flash
lamp was merged to the excitation light path of the microfuorimeter with
a beam splitter.
4. In whole-cell patch-clamp experiments the effect of photolytically
released A T P or InsP3 on Ca 2÷ release in Purkinje neurons 35 and in cultured
endothelial cells 36 was investigated. In both experiments, Ca 2÷ was mea-
31S. Khan, J. L. Spudich, J. A. McCray and D. R. Trentham, Proc. Natl. Acad. Sci. U.S.A.
92, 9757 (1995).
32I. Parker, Y. Yao, and V. Ilyin, Biophys. J. 70, 222 (1996).
33M. Wilding, E. M. Wright, R. Patel, G. Ellis-Davies, and M. Whitaker, Z Cell Biol. 135,
191 (1996).
34X. Q. GUO,M. A. Laflamme,and P. L. Becker, Circ. Res. 79, 147 (1996).
35K. Khodakhah and D. Ogden, J. Physiol. (London) 487, 343 (1995).
36T. D. Carter and D. Ogden, Proc. R. Soc. Lond. B Biol. Sci. 250, 235 (1992).
220 CAGED COMPOUNDS [ 1 II
sured with a microspectrofluorimeter on an inverted microscope. Samples

were exposed directly from the flash lamp housing. With 120 mJ of U G l l
filtered light, about 50% photolysis of caged A T P was achieved.
5. Photolysis of caged glutamate was used to investigate whether
L-glutamate can produce postsynaptic activation in the giant synapse of the
squid. 37 Samples were flashed either through a silica condensor, giving a
spot about 500/~m in diameter, or directly from the lamp housing, giving
a spot about 4 m m in diameter. A b o u t 9 m M L-glutamate was released in
the synapse.
6. The distribution of secretory granules with B cells was used in combi-
nation with flash photolysis of caged Ca (DM-nitrophen) to study exocytosis
in mouse pancreatic B cells. 38 These studies used a combination of tech-
niques, such as patch clamp to record whole cell or single-channel Ca 2+
channel currents, capacitance m e a s u r e m e n t s of exocytosis, and digital im-
aging to quantify cytoplasmic Ca 2÷ gradients.
7. The activation kinetics of L-type calcium current in frog cardiac
myocytes was studied using flash photolysis of caged isoproterenol, a syn-
thetic fl-agonist. 39 In a different study, caged isoproterenol and caged c A M P
were used to investigate sympathetic stimulation of rabbit cardiac cells.
The flash was applied through the rear aperture of an inverted microscope
to a selected cell using a Fluor 40× objective. 4°
8. The kinetics of potassium binding and conformational transitions in
the enzyme were investigated using photolysis of caged K ÷ together with
microspectrofluorimetry on fluorescently labeled Na,K-ATPase. 41
9. Whole-cell patch-clamp recordings were taken to investigate the ef-
fect of Ca 2÷ on the photoreceptors of D r o s o p h i l a . Light for photolysis of
D M - n i t r o p h e n or nitr-5 and for excitation of the calcium-sensitive dyes
was directed onto the sample using a bifurcated light guide and an inverted
microscope. 42 A mechanical shutter protected the photomultiplier during
the light-directed activation of calcium from caged calcium
10. In studies on flash photolysis of caged c A M P and caged GTP-'y-S
on Ca 2÷ currents in cardiac myocytes, light was coupled to the rear port of
an inverted microscope with the use of a dichroic mirror and directed to
the cell through a Fluor 40× objective. 43
37j. E. T. Corrie, A. Desantis, Y. Katayama, K. Khodakhah, J. Messenger, D. Ogden, and

D. R. Trentham, J. Physiol. 465, 1 (1993).
38K. Bokvist, L. Eliasson, C. ,~mm~lKE. Renstr6m, and P. Rorsman, EMBO J. 14, 50 (1995).
39A. M. Frace, P. F. Mery, R. Fischmeister, and H. C. Hartzell, J. Gen. Physiol. 101,337 (1993).
40H. Tanaka, R. B. Clark, and W. R. Giles, Proc. R. Soc. Lond. B Biol. Sci. 263, 241 (1996).
41E. Grell, R. Warmuth, E. Lewitzki, and H. Ruf, Acta Physiol. Scand. 146, 213 (1992).
42R. C. Hardie, J. Neurosci. 15, 889 (1995).
43R. Z. Kozlowski, V. W. Twist, A. M. Brown, and T. Powell, Am. J. Physiol. 261, 1665
(1991).
11. The kinetics of secretory response in bovine chromaffin cells follow-

ing flash photolysis of caged Ca 2+ (DM-nitrophen) was investigated via
capacitance measurements with millisecond time resolution. Flashlight was
directed through the epifluorescence port of an inverted microscope to-
gether with the excitation light for Ca 2+ measurements by means of a
sapphire w i n d o w . 44 Using the formalisms of Z u c k e r , 45 the photolysis effi-
ciency was calculated for Ca2+-free and CaZ+-bound DM-nitrophen. At a
maximum flash energy of 240 J, about 60% of Ca2+-bound and 24% of Ca 2+-
free DM-nitrophen was photolyzed. For neurophysiological studies with
millisecond time resolution, Ca 2÷ chelators have been photolyzed using
flash l a m p s , 46 whereas in other experiments in which time resolution was
not critical, continuous lamps with a shutter were used to photolyze about
12% of nitr-5 in 1 sec. 47
12. In the experimental setup used by Almers and co-workers 48 to
study Ca2+-triggered exocytosis, light from a flash lamp was coupled to the
epifluorescence port of a microscope via a liquid light guide. Additionally,
a shutter was used to protect the photomultiplier for Ca 2+ measurements.
At the maximum flash energy, 3.87 mM DM-nitrophen was photolyzed
from a 10 mM solution. In a related study, a comparison was made of the
effectiveness of photolabile calcium chelators nitrophenyl-EGTA (NPE)
and DM-nitrophen in rising cytosolic calcium to trigger e x o c y t o s i s . 49 Al-
though NPE was found to be less efficient in rising Ca 2+, it was suggested
to be a useful tool for time-resolved studies in the presence of physiological
concentrations of magnesium.
13. Derivatives of inositol trisphosphate (InsP3) and their photolabile
precursors have been characterized and applied in a study on the kinetics
of cytosolic calcium release. 5° Whole-cell recordings and fluorescence mea-
surements were combined. Photolysis light from a flash lamp equipped
with quartz condensor and objective was directed directly onto the sample
without passing through the microscope. The objective had a working dis-
tance of about 4 cm and produced a spot of 4-5 mm in diameter. Photolysis
of caged ATP was adjusted via the electrically stored energy and was
reproducible within about 10%.
14. Caged derivatives of adrenergic receptor agonists, such as epineph-
rine, phenylephrine, and isoproterenol, have been described and photo-
44 C. Heinemann, R. H. Chow, E. Neher, and R. S. Zucker, Biophys. J. 67, 2546 (1994).

45 R. S. Zucker, Cell Calcium 14, 87 (1993).
46 L. Lando and R. S. Zucker, J. Gen. Physiol. 23, 1017 (1989).
47 D. Neveu and R. S. Zucker, Neuron 16, 619 (1996).
48 p. Thomas, J. G. Wong, A. K. Lee, and W. Almers, Neuron 11, 93 (1993).
49 T. D. Parsons, G. C. R. Ellis-Davies, and W. Almers, Cell Calcium 19, 185 (1996).
5o j. F. Wootton, J. E. T. Corrie, T. Capiod, J. Feeney, D. R. Trentham, and D. C. Ogden,
Biophys. J. 68, 2601 (1995).
chemically characterized using repetitive pulse of arc lamps. 51 In studies

on calcium antagonists and agonists, light-directed activation of caged nefe-
dipine, a calcium ion antagonist, was achieved using light from a 100-W
mercury arc lamp delivered through the epifluorescence port of an inverted
microscope in cultured atrial myocytes.52 Similar studies were performed
using photolysis light from a flash lamp that had been directed onto frog
skeletal muscle. 53
15. The effect of GTP-y-S and GDP-/3-S on Ca 2÷ channel currents in
neurons was investigated applying the whole-cell patch-clamp technique. 54
The flash lamp was mounted inside the Faraday cage. Photolysis efficiency
for various caged GTP derivates at 3 mJ/mm 2 of bandpass-filtered UV light
(300-380 nm) was around 10%. Using a 1-msec flash lamp, photolysis yields
at various energy densities were determined for Ca 2+, ATP, GTP, and
cAMP from their respective photolabile precursors in a study on Ca 2+-
dependent C1- currents in cultured rat neurons following photolysis of
DM-nitrophen. 55
Conclusion
Flash lamp systems have proved useful to photolyze a large variety of
caged compounds in diverse biological and biochemical systems. In some
applications, especially those requiring very short pulses or sophisticated
optics, flash lamps are not suitable and one must resort to the use of short,
high-energy laser pulses, although these intense light pulses can damage
the sample under study. For a given energy in the UV range the number
of photoreleased molecules is higher for flash lamps compared to lasers.
The typical pulse length of a flash lamp is around 0.5-1 msec, which is
sufficiently short for many experiments in physiology. If caged compounds
become available with their action spectrum shifted from the UV toward
longer wavelengths, the efficiency of flash lamp-based photolysis will be
enhanced even further due to increased photon flux of xenon flash lamps
at longer wavelengths.
51 S. Muralidharan and J. M. Nerbonne, J. Photochem. Photobiol. 27, 123 (1995).

52 M. Bechem and H. Hoffmann, Pfltigers Arch. 424, 343 (1993).
53 D. Feldmeyer, W. Melzer, B. Pohl, and P. Zrllner, J. Physiol. 457, 639 (1992).
54 R. H. Scott, J. F. Wootton, and A. C. Dolphin, Neuroscience 38, 285 (1990).
55 K. P. Currie, J. F. Wootton, and R. H. Scott, J. Physiol. 482, 291 (1995).
[12] FOURIER TRANSFORM INFRARED PHOTOLYSIS 223
[I 2] F o u r i e r T r a n s f o r m I n f r a r e d P h o t o l y s i s S t u d i e s of
Caged Compounds
By VALENTIN CEPUS~ CAROLA ULBRICH~ CHRISTOPH ALLIN,
AGNES TROULLIER, a n d KLAUS GERWERT
Introduction
Time-resolved Fourier transform infrared (F-FIR) difference spectros-
copy has been established as a new tool to study molecular reaction mecha-
nisms of proteins at the atomic level with nanosecond time resolutionJ '2
In photobiological proteins the reaction can be started directly by a short
laser flash using the intrinsic chromophores. This approach has been success-
fully applied to the membrane proteins bacteriorhodopsin and the photo-
synthetic reaction center. 3 A much broader applicability can be achieved
by the use of caged compounds. In this case biologically active molecules
are released from inactive photolabile precursors. The use of caged com-
pounds has become widespread in the last 10 years. 4-6 They allow the
initiation of a protein reaction with a nanosecond UV laser flash.
This article presents FTIR photolysis studies of caged phosphate, caged
GTP, caged ATP, and caged calcium. This should provide a good basis for
further FTIR studies on molecular reaction mechanisms of proteins using
caged compounds. Finally, as an example, FTIR studies on the GTP-binding
protein H-ras p21 are presented.
Experimental Setup
Fourier transform infrared measurements are performed on an IFS 66v
spectrometer (Bruker, Kadsruhe, Germany). The experimental setup is
shown in Fig. 1. The MIR beam source is a water-cooled globar. The beam
splitter consists of a KBr crystal. A HgCdTe (MCT) detector that is sensitive
in the spectral region between 900 and 5000 cm -1 is used. Single-beam
1 K. Gerwert, Curr. Opin. Struct. Biol. 3, 769 (1993).

2 R. Rammelsberg, B. Hessling, H. Chorongiewski, and K. Gerwert, Appl. Spectrosc. 51,
558 (1997).
3 K. Gerwert, Biochim. Biophys. Acta 1101, 147 (1992)
4 j. A. McCray and D. R. Trentham, Annu. Rev. Biophys. Biophys. Chem. 18, 239 (1989)
5 j. E. T. Corrie and D. R. Trentham, in "Bioorganic Photochemistry: Biological Application
of Photochemical Switches" (H. Morrison, ed.), Wiley, New York, 1993.

Sample
. ~ Plasmalens Fixed Mirror
......^ I~[_~ .............-.% " "
1064nm, ,'~',
Beam, x
Splitter /
I SController
anner- I
_/
G)obar o ..................... ....................... '
/ -. Movable Mirror
Ge-Detector "'
Excimer t ............. / -t~'
mg-,- Sample
MCT-Detector
SUN/PC .... Laser

Network ........... Raman
-- Data
FIG. 1. Experimental setup for photolysis, FTIR, and Raman measurements of caged com-
pounds.
spectra are measured. The spectrometer is controlled by a personal com-

puter (PC) equipped with the program Opus (Bruker).
Photolysis of the caged compounds is performed by ultraviolet (UV)
flashes at the wavelength A of 308 nm [excimer laser, LPX 240 (Lambda
Physics, G6ttingen, Germany)]. The laser flashes are applied on the sample
at an angle of 45° to the IR beam. The duration of a UV laser pulse is
about 20 nsec, with the energy reaching up to 200 mJ per flash. Twenty to
60 flashes are used to achieve the complete photolysis of caged compounds,
except in the case of DM-nitrophen, for which approximately 100 flashes
are necessary. The home-built sample chamber of the F-FIR spectrometer
IFS 66v is purged with dried air. The metal sample holder is connected to
a thermostat. The sample solution is prepared between two CaF2 windows
of 20 mm diameter. A spacer ring made of Mylar 2.5/.~m thick is used in
order to keep a defined distance between the windows. In the absence of
other specification, spectral resolution of 4 cm -1 is used.
Fourier transform infrared measurements of H-ras p21 are performed
with the truncated form of human H-ras p21c (1-166). For the nucleotide
exchange of the protein-bound GDP for caged GTP, an excess of the
desired caged GTP isotopomer in the presence of alkaline phosphatase is
[121 FOURIER TRANSFORM INFRARED PHOTOLYSIS 225
used. 7 During the GTPase reaction of H-ras p21-GTP, spectra are collected
up to 200 min after photolysis.
Raman spectra are recorded on a IFS 88 (Bruker) spectrometer with
an FRA106 Raman module equipped with a germanium detector. For
excitation, a Nd-YAG (neodymium/yttrium-aluminum-garnet) laser
(1064 nm) is used. The sample, in a quartz capillary 0.8 mm in diameter,
is placed in the thermostatted sample holder.
Synthesis
The synthesis of 1-(2-nitrophenyl}ethyl phosphate (caged phosphate)
is performed, as described by Dantzig et aL, 8 via the reaction of 1-(2-
nitrophenyl)diazoethane and orthophosphate. 9 1804-Labeled orthophos-
phate is synthesized as described by Hackney et aI.t°
P3-[1-(2-Nitrophenyl)ethyl]guanosine 5'-triphosphate (caged GTP, see
Scheme 3) and P3-[l-(2-nitrophenyl)ethyl]guanosine 5'-O-('y-thio)triphos-
phate (caged GTP~S) are synthesized following the procedure of Walker
et aL 9by the esterification of guanosine 5'-O-triphosphate (GTP) and the ~S
analog guanosine 5'-O-(T-thio)triphosphate (GTPTS, Merck, Darmstadt,
Germany}, respectively, with 1-(2-nitrophenyl)diazoethane. The 180-
labeled isotopomers are prepared starting with the sulfur analogs guanosine
5'-O-(a-thio)triphosphate (GTPoLS), guanosine 5'-O-(/3-thio)triphosphate
(GTP/3S), and P3-[1-(2-nitrophenyl)ethyl]guanosine 5'-O-(3~-thio)triphos-
phate (caged GTP3tS) by oxidation with N-chlorosuccinimide and hydrolysis
with H2180.11'12 Guanosine 5'-O-[cz-m80]triphosphate (GTP~180, Scheme
3), guanosine 5'-O-[/3-180]triphosphate (GTP/3tso, Scheme 3), and guano-
sine 5'-O-[~/-ISO]triphosphate (GTP'ylsO, Scheme 3) are then esterified
with 1-(2-nitrophenyl)diazoethane as described earlier.
Caged ATP (P3-[1-(2-nitrophenyl)ethyl]adenosine 5'-triphosphate)
and DM-nitrophen [1-(4,5-dimethoxy-2-nitrophenyl)-l,2-diaminoethane-
N , N , N ' , N ' - t e t r a a c e t i c acid] are available commercially (Calbiochem, La
Jolla, CA).
7 j. John, I. Schlichting, E. Schiltz, P. Rosch, and A. Wittinghofer, J. Biol. Chem. 264,

13086 (1989).
8 j. A. Dantzig, Y. E. Goldman, N. C. Millar, J. Lacktis, and E. Homsher, J. Physiol. 451,
247 (1992).
7170 (1988).
10D. D. Hackney, K. E. Stempel, and P. D. Boyer, Methods Enzymol. 64, 60 (1980).
11 B. A. Connolly, F. Eckstein, and H. H. Fuldner, J. Biol. Chem. 257, 3382 (1982).
12j. Feuerstein, Dissertation, Universittit Hannover, Germany (1987).
CH~ 0 CHa O
o -
~NO2 - + hv ~ N+ O-
~, _
O
N/b o
o o
1 2
o -o
O--P-O- +
I_
O
o
4 5
SCHEME 1. Reaction scheme for the photolysis of caged phosphate?
Caged Phosphate
The 1-(2-nitrophenyl)ethyl moiety is used to protect phosphate, nucleo-
tides, and nucleotide analogs. 9'13The application of UV flashes leads to the
release of the desired phosphate compound. The mechanism of photolysis
of compounds containing the 2-nitrobenzyl group was the topic of several
investigations. 9,14,15
Scheme i shows the generally accepted reaction pathway developed by
Walker et aL 9 for the photorearrangement of caged phosphate compounds.
After photolysis, a rapid formation of intermediate 2, called the aci-nitro
anion, occurs. It decays subsequently in a dark reaction to orthophosphate
4 and the byproduct 2-nitrosoacetophenone $. The intermediate may com-
prise a number of rapidly interconvertible forms such as 3, but UV-VIS
studies 9 and single-wavelength time-resolved IR measurements 15 are able
to resolve only a single intermediate for caged ATP.
A typical FFIR difference spectrum of caged phosphate photolysis is
shown in Fig. 2. A spectrum is measured before the photolysis and is
taken as a reference. From this spectrum and the spectrum taken after the
photolysis of the caged phosphate, the difference spectrum is calculated.
The fiat baseline shows the difference between two spectra measured before
the flash as a control. Only those vibrational modes cause bands in the
13j. W. Walker, G. P. Reid, and D. R. Trentham, Methods Enzymol. 172~ 288 (1989).
14 S. Schneider, J. Photochem. Photobiol. A 55, 329 (1991).
15 A. Barth, J. E. T. Corrie, M. J. Gradwell, Y. Maeda, W. Mantele, T. Meier, and D. R.
Trentham, J. Am. Chem. Soc. 119, 4149 (1997).
[ 121 FOURIER TRANSFORM INFRARED PHOTOLYSIS 227
AA I I I I
1688
0.005
0.000
-0.005 -
-0.010 -
1525
1800 1600 1400 1200 1000

wavenumber [cm "l]
FIG. 2. Fourier transform infrared difference photolysis spectra of caged phosphate; 50

mM caged phosphate, 0.1 M MOPS, pH 7.0, baseline and a spectrum after the first, second,
fourth, and ninth flash.
difference spectrum that undergo reaction-induced absorbance changes.

Negative bands in the difference spectrum are due to the caged phosphate
1 in Scheme 1, whereas positive bands are due to the photolysis products,
orthophosphate 4 and 2-nitrosoacetophenone 5. Characteristic bands of 1-
(2-nitrophenyl)ethyl derivatives are the disappearing asymmetric and the
symmetric (NO2) stretching vibrations [Vas(NO2) = 1525 cm-1; t:sy(NO2)
= 1346 cm-1]; and the positive band is due to the carbonyl group of 2-
nitrosoacetophenone 5 [ v ( C = O ) = 1688 cm-1]. 5'16-t8
Band Assignment by Isotopic Labeling

Unequivocal assignment of IR bands is made possible by isotopic label-
ing. For identification of phosphate bands we used 1804-1abeled caged phos-
16N. B. Colthup J. W. Daly, and S. E. Wiberly, "Introduction to Infrared Raman Spectros-

copy," Academic Press, London, 1990.
~7D. Lien-Vien, N. B. Colthup, and S. E. Wiberley, "The Handbook of Infrared and Raman
Characteristic Frequencies of Organic Molecules," Academic Press, San Diego, 1991.
18H. Georg, A. Barth, W. Kreutz, F. Siebert, and W. Mantele, Biochim. Biophys. Acta 1188,
139 (1994).
228 CAGEDC O M P O U N D S [ 121
1 , ......... , ......... ......... , ......... ,
2
t •
/A [ txm|A ....... ' ......... ' ' '1068 ' ' ' ';;2 ...... I
APt
0 . 0 4 1073~ 992
0.04~
0.02 0.02
o.oo~ 5 3
10801068 977935
1117 oo, V
-0.02 / ~ 1040 1\952 1117 /~
H
93:
1073 ......... , ........... ~,o,7,2, . . . . . . . . . . . . . .

933
I l l l l l l l l l l l l l l l l l ] l l . . . . . . . I .... till
1300 1200 1100 1000 900 1300 1200 1100 1000 800
wavenumber [cm"1] wavenumber lcm"11
Fic. 3. Photolysis of 50 mM caged phosphate; (1) pH 7.0, 0.2 M MOPS; (2) pH 8.0, 0.2
M HEPES; (A) caged phosphate [1604]; (B) caged phosphate [1804]; (C) difference B - A.
phate. The higher mass of the oxygen isotope leads to a shift of all phosphate
bands to lower wavenumbers.
Figure 3 (1) shows photolysis difference spectra of the unlabeled caged
phosphate (A) and the 180-labeled caged phosphate (B) at pH 7.0. The
double difference B minus A (C) represents the spectral differences be-
tween the unlabeled and the 180-labeled phosphate moiety of the caged
compound. Figure 3 (2) shows the equivalent spectra measured at pH 8.0.
The isotopical shifts are obvious. At pH 7.0, two species of orthophosphate,
mono- and dihydrogen phosphate (pgal -- 2, pKa2 = 6.72, pga3 -- 11.96,
I = 0.1 M, 250),19 are present [Fig. 3 (1)]. To correctly assign the bands
present at pH 7.0 to one of these two species, one has to measure the
difference photolysis spectra at different pH values. Phosphate bands ap-
pear below 1300 cm -1, and therefore only this spectral region is shown in
Fig. 3.
The positive bands in spectra A and B correspond to the orthophosphate
that is liberated by the photolysis of caged phosphate. The band at 1158
cm -1 [Fig. 3 (1)] is shifted to 1124 cm -1 and can be assigned to the (PO2-)
19"Critical Stability Constants. Vol. 4. Inorganic Complexes" (R. M. Smith and A. E. Marell,
eds.), Plenum Press, New York, 1976.
[12] FOURIER TRANSFORMINFRAREDPHOTOLYSIS 229
TABLE I
ASSIGNMENTOFPHOSPHATEBANDSDUE TOISOTOPICLABELING
Wavenumber Wavenumber
Caged phosphate (cm- 1 ) Phosphate (cm 1 )
(POa2-)deg 1117 H2PO4- (PO2)as 1158

(PO32-)deg 1090 H2PO4- (PO2)sy 1073
(PO3 2 )sy 977 HPO4 2- (PO32-)deg 1068
HPO4 z- (PO3 2 )sy 992
asymmetric stretching vibration of H2PO4-.20 Indeed, this vibration is nearly

completely absent at p H 8.0 [Fig. 3 (2)] where HPO42 is the prevalent
species. The broad band at 1073 cm -1, which is shifted to 1039 cm t, is
caused by the (PO2-) symmetric stretching vibration of H2PO4- and by
(PO32-) degenerate stretching vibration of HPO42- 20,2t that appears at
1068 cm -1 at p H 8.0. The monohydrogen phosphate has an additional
characteristic vibration at 992 cm -I caused by the symmetric stretching
vibration of the (PO32-) moiety 2°'21 that is shifted to 954 and 953 cm -1, re-
spectively.
The negative bands at 1117, 1090, and 977 cm -1 (Table I) belong to the
unlabeled caged phosphate and are shifted to 1080, 1068, and 935 cm -1,
respectively, by isotopic labeling. The deprotonated monomethyl phosphate
has two degenerate (PO32-) stretching vibrations at 1115 and 1090 cm -1
and a (PO32-) symmetric stretching vibration at 983 cm-1. 22 Monoesters of
phosphoric acid are reported to have two pKa values, pKt ~ 1.5 and
pK2 ~ 6.323; thus the F T I R spectrum of caged phosphate at p H 8.0 should
be dominated by the completely deprotonated caged phosphate dianion.
Because a band at 1117 cm -~ with a shoulder at 1090 cm -1 and a band at
977 cm -~ are observed in this spectrum, the same assignment could be given
for caged phosphate.
Influence of M a g n e s i u m o n Caged P h o s p h a t e
It is important to know the influence of magnesium on the F T I R spec-
trum of the phosphate moiety for the band assignments in protein measure-
ments because magnesium is an essential cofactor for several phosphate
and nucleotide binding enzymes, e.g., ATPases, kinases, and ion pumps.
z0A. C. Chapman and L. E. Thirlwell, Spectrochim. Acta 20, 937 (1964).

21E. Steger and K. Herzog, Z. Anorg. Allg. Chem. 331, 169 (1964).
22T. Shimanouchi, M. Tsuboi, and Y. Kyogoku, Adv. Chem. Phys. 7, 435 (1964).
z3W. D. Kumler and J. J. Eiler, J. Am. Chem. Soc. 65, 2355 (1943).
I I I ' I
AA 1043
1073 t~
0.00
-0.02
1117
-0.04
1800 1600 1400 1200 1000

wavenumber [era "l]
Fic. 4. Photolysis spectra of caged phosphate with Mg2+ (dashed line) and without Mg2+
(solid line); 50 raM caged phosphate, (100 raM MgCI2), 200 raM HEPES, pH 7.0.
Figure 4 shows the photolysis spectrum of caged phosphate at pH 8.0 in

the presence and absence of magnesium.
The positive band at 1073 cm -1, which belongs to the orthophosphate,
is shifted to 1043 cm -1. This shift to lower wavenumbers is probably caused
by the decrease of the partial double bond character of the P - O bonds
involved in magnesium coordination. The band broadening in the ortho-
phosphate region could also be due to a mixture of different magnesium
orthophosphate complexes, e.g., Mg3(PO4)2 and MgHPO4.24
The small positive band at 1158 cm -1 missing in the presence of magne-
sium is due to a slightly higher concentration of H2PO4- in the absence of
magnesium, which shifts the equilibrium of the orthophosphate species
toward HPO42-.
In the absence of magnesium the negative band belonging to the caged
phosphate moiety at 1117 cm -1 has a shoulder at 1090 cm -a. In the presence
of magnesium a clearly resolved negative band is observed at 1090 cm -1.
The appearance of this band is probably caused by the shift of the positive
band from 1073 to 1043 cm -1. Therefore, it seems reasonable to assign this
band to the degenerate stretching vibration of caged phosphate, which is
masked by the positive band in the absence of magnesium. No further
24 A. W. Taylor, A. W. Frazier, and E. L. Gurney, Trans. Faraday Soc. 59, 1585 (1963).
0
O 0 O
O--~-O-~-O-~-o-i >lOP s"= o
. H•
_ ~ -~_
O 0
6 7 8
O
CH 3
0 0 0 [
O--P-O-P-O-P-O--a r~
0
OH OH
9 10
SCHEME2. Reaction scheme for the photolysisof caged GTP.9
influence of magnesium on the bands corresponding to caged phosphate

is detected.
Caged GTP
Further important compounds for the investigation of biological systems
are caged nucleotides such as caged GTP and ATP. a,5'9'13 The photolysis
reaction of caged GTP is shown in Scheme 2 (Structures 6-10).
The photolysis spectrum of caged GTP (Fig. 5, solid line) shows bands
due to the 1-(2-nitrophenyl)ethyl moiety (1688, 1525, and 1346 cm -I, as
already discussed) and phosphate bands below 1300 cm -1. Bond cleavage
between the caged moiety and the leaving group GTP gives rise to the
negative band at 1252 cm -I. It is caused by the asymmetric stretching
vibration of the ? (PO2-) group connected to the caging group. During
photolysis this group is converted to the terminal ? (PO32-) group of the
released GTP. The degenerate stretching vibration of the ? (PO32-) group
leads to the positive band at 1118 cm -~. Generally, the spectral region
between 1050 and 1200 cm -~ is determined by a superposition of the degen-
erate ? (PO32-) and the symmetric (PO2-) stretching modes. 25-27
25H. Takeuchi, H. Murata, and I. Harada, J. Am. Chem. Soc. 110, 392 (1988).
26A. Barth, W. Kreutz, and W. Mantele, Biochim. Biophys. Acta 1194, 75 (1994).
27A. Barth, W. Mantele, and W. Kreutz, Biochim. Biophys. Acta 1057, 115 (1991).
AA - ' ' ~\lll8
0.012
0.008 llj,~.' ! 1
1688
0.000
-0.004
-0.008
1525 U/-1252
I ~ I , I
1800 1600 1400 1200 1000
wavenumber[era"1]
Fie. 5. Photolysisspectra of caged GTP with Mg2+ (dashed line) and without Mg2÷ (solid
line); 20 mM caged GTP (50 mM MgClz), 100 mM HEPES, pH 7.0.
Influence of M a g n e s i u m on Caged GTP

The formation of complexes among caged GTP, GTP, or other nucleo-
tides with divalent cations, such as Ca 2+, Mg z÷, and Mn 2÷, has a strong
effect on FTIR spectra. 25'28
The comparison of caged GTP spectra with and without Mg z÷ (Fig. 5)
shows no significant effect above 1300 cm -I. The negative band at 1252
cm -I in the absence of Mg 2÷ is less intensive and is shifted upward to 1273
cm i in the presence of the cation. This can be explained by an increase
of the partial double bond character of the P - O bonds not directly involved
in magnesium coordination. The positive band around 1118 cm -I is shifted
to 1124 cm -1 in the Mg2+-bound case.
The similarity of caged GTP and caged ATP photolysis (Fig. 6) shows,
as expected, that FTIR differences spectra are dominated by changes de-
rived from the 1-(2-nitrophenyl)ethyl group and the phosphate chain.
Time-Resolved M e a s u r e m e n t s
Various FTIR techniques have been developed to collect time-resolved
absorbance data. They are the rapid scan, 29 the stroboscopic, 3° and the
28H. Brintzinger,Biochim. Biophys. Acta 77, 343 (1963).

29K. Gerwert, G. Souvignier,and B. Hess, Proc. Natl. Acad. Sci. U.S.A. 87, 9774 (1990).
30G. Souvignierand K. Gerwert, Biophys. J. 63, 1393 (1992).
[ 12] FOURIERTRANSFORMINFRAREDPHOTOLYSIS 233
AA
0.010
0.005
0.000
-0.005
I , I I ,
1800 1600 1400 1200 1000

wavenumber [cm"1]
F~G. 6. Caged GTP (solid line), caged ATP (dashed line), 40 mM caged GTP/ATP, 200
mM MgC12, 500 m M HEPES/NaOH, pH 7.5.
step-scan techniques. 2,31,32 (and citations therein). Here we use the rapid
scan technique with a 12-msec time resolution to record the IR difference
spectra of the intermediate (Scheme 2, 7, 8) and the photolysis products
of caged GTP. Conditions close to physiological values, pH 7.5 and the
presence of MgCI2, are used for the measurements. The presence of di-
thiothreitol (DTT) is essential for the biological application of several
caged compounds because the photolysis of the 1-(2-nitrophenyl)ethyl
group yields 2-nitrosoacetophenone (Scheme 2, 10). This compound can
react with the cysteines in proteins and therefore inactivate or modify the
biological system.4 These complications can be overcome by the addition
of thiols, such as DT-F, which scavenge this nitrosoketone. 33
Figure 7 compares photolysis spectra of caged GTP with and without
DTT. The main bands are the same in the two spectra, which demonstrates
the relatively small influence of DTT on the spectrum. The largest difference
is the missing band at 1688 c m -1 in the measurement with DTI?. This shows
that the carbonyl group of the 2-nitrosoketone is absent in the final products,
which can be explained by the reaction of 2-nitrosoketone with DTT. The
lower part of the spectra shows small differences in the intensities at 1250,
31 W. Uhmann, A. Becker, C. Taran, and F. Siebert, Appl. Spectrosc. 45, 390 (1991).
32 R. A. Palmer, J. L. Chao, R. M. Ditmar, V. G. Gregoriou, and S. E. Plunkett, Appl.
Spectrosc. 47, 1297 (1993).
33 j. H. Kaplan, B. Forbush III, and J. F. Hoffman, Biochemistry 17, 1929 (1978).
A A i i i i
_ 11~4
- 1688 ]i
.'l, 1600 1452 1425 1250 /g I 1046
1800 ' 1 00 ' ' 12 00' 1200

' ' ' 1000
'
wavenumber [cma]
FIo. 7. Photolysis spectra of caged GTP with DTT (solid line, 50 mM caged GTP, 250
mM MgCI2, 250 mM DTI', 500 mM HEPES/NaOH, pH 7.5, 10°, resolution 8 cm -1) and
without DTI" (dashed line, 20 mM caged GTP, 50 mM MgCI2, 100 mM HEPES/NaOH, pH
7.5, 20*, resolution 4 cm-l).
1124, 1060, and 1046 cm -1 and a few smaller bands at 1600, 1452, and
1425 cm -1.
In order to determine the IR difference spectra of the intermediate,
spectra obtained between 1 and 26 msec, after a single laser flash are
averaged (Fig. 8A). The region between 1600 and 1700 cm -1 shows a
decreased signal-to-noise ( S / N ) ratio because of the large water background
absorption. The product difference spectrum (Fig. 8B), obtained between
86 and 105 sec after the flash, shows the IR differences between caged
GTP and the final products of the caged GTP photolysis, i.e., GTP and the
reaction products of 2-nitrosoacetophenone with DTT.
As determined by a global fit analysis, 34 the intermediate decays expo-
nentially with a decay constant of 8 sec -1 to the final products. Thus the
average intermediate concentration in the spectrum (Fig. 8A) is 96%. Barth
et al. 15have reported the IR spectrum of an intermediate during the photoly-
sis of caged ATP. In general, the difference spectra shown in A agree nicely
with their published one, even though slightly different conditions (different
buffer, pH 8.5, no magnesium) were used. Several bands, e.g., at 1379, 1328,
1244, and 1181 cm -1, which they assign to the aci-nitro anion (Scheme 2,
7), can also be found in the spectrum (Fig. 8A). Barth et al. 15 could make
34B. Hessling, G. Souvignier, and K. Gerwert, Biophys. J. 65, 1929 (1993).

' ' ' f ' ' ' J ' ' ' ~ '1125' ' ~991 I
AA 1462 1328 1244 A1094 /~ I
1124
1525 A
1147
1273 1060
I 1525 I i I I i I
1800 ! 600 1400 ! 200 1000

wavenumber [m'~l
FIG. 8. Fourier transform infrared differencespectra of laser pulse photolysisof caged

GTP, 50 mM caged GTP, 250 mM MgCI2,250 mM DTr, 500 mM HEPES/NaOH, pH 7.5,
10°, resolution8 cm-1. (A) Obtained1-26 msecafter a singlelaser flash;(B) obtained86-105
sec after a singlelaser flash.
specific band assignments by the use of mac, mSN,and 1 8 0 isotopomers of

caged ATP. Deviations to the difference spectra of Barth et a i J 5 are ob-
served for the disappearing asymmetric (NO2) stretching vibration. They
reported a shift of this educt band between the intermediate and the product
difference spectrum. However, no difference at 1525 cm -1 is observed
between spectra A and B. These differences could arise from the fact that
Barth et al. used a different buffer that has a large background absorption
in this spectral region.
R a m a n Spectroscopy
Raman spectroscopy, like infrared spectroscopy, probes the vibrational
properties of a molecule. The selection rules between these two are differ-
ent, however. This difference is a useful tool for the assignment of bands.
Asymmetric vibrations give stronger bands in the IR spectrum, whereas
bands due to symmetric vibrations are stronger in the Raman spectrum.
Figure 9 shows the solution Raman spectrum of caged GTP (A) and
GTP (B). The Raman spectrum of the solvent was subtracted. The strongest
band in the Raman spectrum of caged GTP is the symmetric (NO2) stretch-
ing vibration at 1346 cm-1, TM whereas the asymmetric (NO2) stretching
vibration gives rise only to a small band in the spectrum at 1525 cm -1. By
i i i i i i i i i i ~ i i i i i
1346
1579 1487
1~ /~ 1458 I I
II IIA / I ..
A 1525
1124
1800 1600 1400 1200 1000

Raman shiR [era"1]
Fio. 9. Solution Raman spectrum. (A) Caged GTP; 100 mM caged GTP, 250 mM MgC12,
250 mM DTF, 500 mM HEPES/NaOH, pH 7.5, 4 °. (B) GTP; 20 m M GTP, 60 m M MgCI2,
pH 7.5, 4°. The Raman spectrum of the solvent was subtracted.
comparison with the Raman spectrum of GDP of Weng et al. 35 the band
at 1579 cm -1 can be assigned to the N3-C4 and C4-N9 stretching motions
of the ring of the purine base. The band at 1487 cm -1 can be assigned to
the N7----C8 and C4 = C5 stretching vibration of the purine base. Another
strong band is at 1458 cm -a, which can tentatively be assigned to the phenyl
ring of the 2-nitrophenyl group of caged GTP. The band at 1139 cm -a can
be caused by the symmetric (POE-) vibration of the or-,/3-, and y-phosphate
groups of caged GTP. The strong buffer scattering between 1000 and 1100
cm -1 causes a subtraction artifact at 1050 cm -1 and leads to a decreased S/
N ratio in this spectral region. For Mg2+-complexed GTP (Fig. 9B) the
Raman band at 1124 cm -a is assigned, as for magnesium complexed ATP, 25
to the in-phase symmetric (PO 2-) mode.
To demonstrate the different selection rules for IR and Raman spectros-
copy, the Raman difference spectrum between GTP and caged GTP (Fig.
10A) and the IR difference spectrum of photolyzed caged GTP (Fig. 10B)
are compared. In order to subtract the two Raman spectra (Figs. 9A and
9B), spectra are normalized using the band at 1579 cm -1.
35 G. Weng, C. X. Chen, V. Balogh-Nair, R. Callender, and D. Manor, Protein Sci. 3, 22 (1994).

[12] FOURIERTRANSFORMINFRAREDPHOTOLYSIS 237
' ' I , r- ...... I I •
M,/,~I 1120
1458
1124
1346 4~
11
1060
1273
1525
i ± I, , , , l , , , I , , ,
1800 1600 1400 1200 1000

wavenumber [cm-~]
FIG. 10. (A) NIR-FF-Ramandifferencespectra of GTP (Fig. 9B) minus caged GTP (Fig.
9A); (B) FTIR differencespectra of laser pulse photolysisof caged GTP; 50 mM caged GTP,
250 mM MgCI2, 250 mM DTT, 500 mM HEPES/NaOH, pH 7.5, 10°, obtained 86-105 sec
after a single laser flash, resolution 8 crn-1.
In contrast to F T I R difference spectra, Raman difference spectra be-

tween caged GTP and GTP show no large differences. Spectra show nicely
that the symmetric vibration of the (NO2) group at 1346 cm -114 gives a
much larger signal in the Raman spectrum, whereas the asymmetric (NO2)
stretching vibration at 1525 cm -1 is much stronger in the IR difference
spectrum. The negative band at 1458 cm -1 is tentatively assigned to the
phenyl ring of the 2-nitrophenyl group of caged GTP. The negative band
at 1139 crn -1 is probably caused by the in-phase mode of the three (PO2-)
groups of caged GTP. By comparison with the solution Raman spectrum
of GTP (Fig. 9A), a small band at 1120 cm -l can be visualized. This band
is assigned to the in-phase mode of the (PO2-) groups of GTP.
Balad A s s i g n m e n t s of Caged GTP

For further band assignment, we synthesized caged GTP isotopomers
that contain one 180 isotope selectively in the a, fl, and y position, respec-
O
I I
o o o
X< ~---CH-O--P--O--P--O--P--O--'] ,., 1~ N NFI:
\\ // I I I / I/v.. I
o, o, Oo ]
NO2 " " " ~ - ~
OH OH
SCHEME 3. Position of the 180 labels introduced. All oxygens the 160 isotope (1); a: 180
isotope,/3 and y: 160 isotope (2);/3:180 isotope, a and -r: 160 isotope (3); 3': 180 isotope, ot
and/3:160 isotope (4).
tively. Photolysis spectra of the four caged GTP isotopomers (Scheme 3,

unlabeled and 180 labeled) are shown in Fig. 11. To normalize spectra, the
symmetric (NO2) vibration at 1346 cm -1 is used (Fig. 11, dashed line).
The comparison of the four photolysis spectra shows downshifts for each
of the 180-labeled compounds. Shifting of caged GTP bands is indicated by
a light shading and shifting of GTP bands by a dark shading (Fig. 11, line
1). The black area symbolizes a shift that is observed only in the spectrum
of the y-180-labeled caged GTP. Spectra are termed by the numbers ac-
cording to the respective caged GTP isotopomers in Scheme 3. The wave-
' ' I ' ' ' I L q ,
jC-112,
0.00- 1 1250 // ~ kk / 1 1 0 5
J 1078
unlabdled
-0.01
_0.02 ed*O labelled
. . . .
3nO labelled
-o.o~ il i l i
4 i " i
180 labelled
1400 1300 1200 1100 1000

wavenumber [era"l]
FIG. 11. Caged GTP photolysis spectra; 20 m M caged GTP, 50 m M MgC12, 100 m M
HEPES, pH 7.5. (1) Unlabeled; (2) a-180-labeled; (3)/3-1SO-labeled; and (4) ~/-lSO-labeled.
numbers to which the bands are shifted in the spectra of the labeled com-
pounds are indicated by the same shadings referring to line 1.
The negative band at 1273 cm -1, representing the asymmetric modes
of the three (PO2-) groups of caged GTP, is sensitive to all three labels.
The intensity of this band is decreased by a- and even more b y / 3 : 8 0 -
labeling, and the band is shifted to 1250 cm -I. This is indicated by the light
shading on lines 2 and 3. y-i80-Labeling causes a downshift of the band
at 1273 to 1270 cm -1 but does not change its intensity (black area on line
4, Fig. 11). This interpretation is supported by the double differences (Fig.
12). Here the photolysis difference spectra of the labeled compounds are
subtracted from that of the unlabeled one. For or- and/3-labeled caged
GTP, a shift can be detected from 1273 to 1253 and 1251 c m -1, respectively.
In contrast, caged GTP3"180 shows a specific shift from 1277 to 1261 cm -1.
The positive band at 1250 cm -I represents the asymmetric stretching
vibration of the a and/3 (PO2-) group of the free GTP (Fig. 11). Due to
a,/3, and 3' labeling, the band at 1250 cm -1 is downshifted to around 1240
cm -1 (dark shading, Fig. 11). This becomes clearer in double difference
spectra (Fig. 12). For the a,/3, and 3' label a shift is observed from 1253 to
1238, from 1251 to 1236, and from 1261 to 1240 cm -1, respectively. For
GTP/3~80 the shift is less obvious due to the masking shift of the band at
1273 cm -1.
A A J ' ' ' I ' ' '

1 tmlabolled -
o.ooo I ~ - - ~ unhbdied
2 1273"----... 12~38 1 03 --" 1080
/ 1253 \ / v~1093
-0.005 [ 1273"-'--'A 1125/k'] ~ , ~ 1078 unlabelled
-0.010[- / \ 12~6 11031 \,-:"~
/~ unlabelled
oo" V .7.
I ,2, \/ lo94
, , , I , "---1129 ,
1400 1300 1200 1100 1000

way,number [¢m"1]
FIG. 12. Double differences of caged GTP photolysis spectra; 20 mM caged GTP, 50 mM
MgCI2, 100 mM HEPES, pH 7.5. (1) Unlabeled; (2) a-Is-O-labeled; (3) fl-lSO-labeled; and
(4) y-Is-O-labeled.
TABLE II
ASSIGNMENTS OF CHANGING PHOSPHATE VIBRATIONS DURING CAGED G T P
PHOTOLYSIS REACTION a
Caged Wavenumber Wavenumber

GTP (cm -1) Comment GTP (cm -1) Comment
ot-(PO2- ) 1273 Coupled with/3 a-(PO2 ) 1253 a,/3, and T coupled

1125 a,/3, and y coupled
1093 a,/3, and 3' coupled
/3-(P02-) 1273 Coupled with a /3-(P02 ) (1251) a,/3, and 3" coupled
1126 a,/3, and 3" coupled
1093 a,/3, and 3" coupled
3'-(P02-) 1277 T-(P032-) 1261
1129 a,/3, and Y coupled
1094 a,/3, and y coupled
~Results are obtained by the use of the isotopomers 1 to 4 (see Scheme 3).
After photolysis the y-phosphate is converted to a nonbridging (PO32-)

group. The degenerate stretching vibration of the 7 (PO32-) group should
absorb below 1150 cm -1. Furthermore, below 1150 cm -a phosphate bands
due to symmetric stretching vibrations of (PO2-) groups are superim-
posed. 15 The most intensive positive band is located at 1124 cm -1 (Fig. 11).
It is downshifted by o~-lSO-labeling. A better insight into the shifts is again
provided by double difference (Fig. 12). z5 In principle, two different shift
patterns are possible. One possibility is a downshift from 1125 to 1080 cm -a
(Fig. 12), but the maximal downshift due to a80-labeling in a diatomic
molecule is 43 cm -a. Therefore, it is more likely that the band is downshifted
from 1125 to about 1103 cm -1 and a second shift is superimposed from
1093 to 1080 cm -t. In this case, the shifted bands at 1103 and 1093 cm -1
cancel each other. Similar shifts are seen due to/3-a80-labeling from 1126
to 1103 and 1093 to 1078 cm -a (Fig. 12). y-aSO-Labeling seems to cause
shifts from 1129 to 1103 and 1094 to 1080 cm -1. The bands at 1124 and
1093 cm -a seem to represent o~ and/3 (PO2-) that are coupled to the
(PO32- ) vibrations. The observed sensitivity of the bands to the aso-labeling
can be explained by a strong vibration coupling among all three phosphate
groups (Table II).
DM-Nitrophen: A Photolabfle Chelator of Divalent Cations

Another important group of caged compounds are photolabile deriva-
tives of cation chelating reagents such as DM-nitrophen. 33,36-39 DM-Ni-
36 j. H. Kaplan and G. C. R. Ellis-Davies, Proc. Natl. Acad. Sci. U.S.A. 85, 6571 (1988).
37 G. C. Ellis-Davies, J. H. Kaplan, and R. J. Barsotti, Biophys. J. 70, 1006 (1996).
/ CO0- / COO-
CH3
CH30.-
~ v
N\
N/
-,NO2
COO "Ca2+
CO0"
COO"
UV-flash CH30 ~ O
CH30 ~ - / ' x ~ ' / ~ N O

NN------COO"
Ca2+
/ COO"
HN~ COO"
SCHEME 4. Photolysis of DM-nitrophen.
trophen contains an EDTA molecule and has a high affinity for divalent
cations in the unphotolyzed state [KD (Ca > ) ~- 5 × 10 -9 M, g D (Mg 2+) -~
2.5 X 10 -6 M]. 36 During the photolysis of DM-nitrophen, a C-N bond is
broken and iminodiacetic acid is released (Scheme 4). Thus the KD for the
cations decreases to the millimolar range. 36'4°
This chelator has been used for investigating exocytosis, 41 muscle fi-
bers, 42 and FTIR measurements on sarcoplasmic reticulum (SR) Caz+-
ATPase. m'4°Time-resolved FTIR difference spectroscopy has been applied
to measure the kinetics of calcium binding to the SR Ca>-ATPase. 4°
A typical FTIR photolysis spectrum of DM-nitrophen is shown in Fig.
13. As is the case with the other caged compounds, the band at 1525
cm -1 is assigned to the asymmetric stretching vibration of (NO2). This
band is taken as reference to rescale spectra from different measurements
when necessary. It could be expected that the bands due to the carboxylic
groups are changed when different divalent cations are used. 43'44 As shown
in Fig. 13, the wavenumber of the negative band at around 1585 cm -a
is different in the presence of Ca2+ (1585 cm-1), Mg 2+ (1588 cm-1), and
without divalent cation (1583 cm-1). This band can be assigned to the
asymmetric stretching vibration of the (COO--) groups of the EDTA
38 A. L. Escobar, F. Cifuentes, and J. L. Vergara, FEBS Lett. 364, 335 (1995).

39j. A. McCray, N. Fidler-Lim, G. C. Ellis-Davies, and J. H. Kaplan, Biochemistry 31,
8856 (1992).
40 A. Troullier, K. Gerwert, and Y. Dupont, Biophys. J. 71, 2970 (1996).
41 T. D. Parsons, G. C. Ellis-Davies, and W. Almers, Cell Calcium 19, 185 (1996).
42 j. R. Patel, G. M. Diffee, and R. L. Moss, Biophys. J. 70, 2333 (1996).
43 M. Nara, M. Tasumi, M. Tanokura, T. Hiraoki, M. Yazawa, and A. Tsutsurni, FEBS Lett.
349, 84 (1994).
44 G. B. Deacon, R. J. Phillips, Coord. Chem. Rev. 33, 227 (1980).
' ' I ' ' ' I ' ' '
AA
0.01
0.00
-0.01
15~
1525
i i i I i
1800 1600 1400 1200 1000

wavenumber [era"l]
FIC. 13. 20 mM DM-nitrophen, 20 mM DTI', 10 mM Ca 2÷ (Mg2+ or without divalent

cation), respectively, 0.3 M MOPS, pH 7.0, resolution 2 cm -~.
moiety of DM-nitrophen) 6'17'45 Similar variation is observed for the band

assigned to the symmetric carboxylate mode in the presence of Ca 2+
(1414 cm-a), Mgz+ (1420 cm-a), and without divalent cation (1402 cm-a).
The symmetric vibration of arylic (NO2) is usually found in the range
of 1357 to 1318 cm-1.17 This vibration has been assigned to a broad
negative band at 1332 cm -1 observed in the photolysis spectrum of DM-
nitrophenJ 8 We also observe a large band in this range that shows, in
fact, two resolved peaks (Fig. 13): a band at 1340 cm -1 that is not modified
in the absence or presence of divalent cations and a band at 1326 cm -a
that is sensitive to the presence of divalent cation. The last one is also
present in EDTA absorbance spectra and is assigned to the (COO-)
moiety.45 Therefore, only the band at 1340 cm -~ in the photolysis spectrum
of DM-nitrophen should be assigned to the aromatic symmetric stretching
vibration of the (NO2) moiety.
Its ability to release a variety of cations by flash photolysis makes DM-
nitrophen a useful compound for FTIR measurements on many biologi-
cal systems.
45 I~). T. Sawyer and P. J. Paulsen, J. Am. Chem. Soc. 80, 1597 (1958).
[1 2] FOURIER TRANSFORMINFRAREDPHOTOLYSIS 243
A A . . . . . . . . . J ' . . . . . . . . J . . . . . . . . .
I124 I . . . . . . . ' '
I
0.02 / / ~
A 1250
-0.02
1214 /
1341 \ 1268
I I ~ q I I [ I I [ I P I ~ I I I I [ [ ~ [ I I p ] I p [ I I , , , , , I ,
1400 1300 1200 1100 1000

wavenumber [cm"l]
FIG. 14. (A) Photolysis spectrum of 20 mM caged GTP, 50 mM MgCI2, 100 mM HEPES,
pH 7.5 (B) Photolysis spectrum of i0 mM H-ras p21-caged GTP, 25 mM MgCI2, 10 mM
DTT, 50 mM HEPES, 1% glycerol, pH 7.5.
Biological Application: F o u r i e r T r a n s f o r m I n f r a r e d S t u d i e s of G T P a s e
M e c h a n i s m of H- r a s p21
Caged G T P can be used as a substrate for H-ras p21. It is tightly bound
similarly to G T P in close proximity to magnesium, but it is not hydrolyzed. ~6
The photolysis spectrum of H-ras p 2 1 - c a g e d G T P is shown in Fig. 14.
The shape of the H-ras p 2 1 - c a g e d G T P photolysis spectrum (Fig. 14B)
is r e m a r k a b l y different from the spectrum without protein (Fig. 14A). The
difference bands of the free nucleotide (1273/1250 cm -t) are split into two
negative bands at 1268 and 1236 cm -t and two positive bands at 1260 cm -I
and 1214 cm -I in the enzyme-bound form. The b r o a d positive band pattern
originally around 1124 cm -I is shifted to 1140 cm -~. Also, below 1050 cm -I
small changes are observed. C o m p a r e d to the photolysis without H-ras p21,
the phosphate bands gain intensity and are shifted to higher wavenumbers.
Because the degrees of f r e e d o m of the nucleotide are reduced by binding
to the protein the bands a p p e a r sharper. 47'48
46I. Schlichting, G, Rapp, J. John, A. Wittinghofer, E. F. Pai, and R. S. Goody, Proc. Natl.
Acad. Sci. U.S.A. 86, 7687 (1989).
47K. Gerwert, V. Cepus, A. J. Scheidig, and R. S. Goody, in "Proceedings of Time-Resolved
Vibrational Spectroscopy VI" (A. Lau, F. Siebert, and W. Werncke, eds.), p. 185, Springer
Verlag, Berlin, 1994.
48V. Cepus, R. S. Goody, and K. Gerwert, submitted (1998).
244 CAGEDCO~VOUNOS [121
The GTPase reaction takes place after the photolytic liberation of
GTP. 46,47 Fourier transform infrared spectra were then recorded from 30
sec to 2 hr 10 rain. The difference calculation was referred to the last
recorded spectrum (Fig. 15). Therefore, the positive bands represent the
H-ras p21-GTP state and the negative ones the H-ras p21-GDP state and
inorganic phosphate. At 1260 cm -1 a strong positive band is detected that
is characteristic for the asymmetric stretching vibration of the GTP (PO2-)
groups. The negative band at 1100 cm 1 represents the strongest GDP
band. It is dominated by the degenerate stretching vibration of the terminal
/3 (PO32-) groupY Results show that kinetic analysis of the GTPase reaction
of H-ras p21 can be performed with FTIR difference spectroscopy. Thus,
a new experimental approach, time-resolved FTIR difference spectroscopy,
is established that monitors p21 GTPase reaction with high structural and
high time resolution.
Summary
Time-resolved FTIR difference spectroscopy is a powerful tool for in-
vestigating molecular reaction mechanisms of proteins. In order to detect,
beyond the large background absorbance of the protein and the water,
absorbance bands of protein groups that undergo reactions, difference
spectra have to be performed between a ground state and an activated
state of the sample. Because the absorbance changes are small, the reaction
has to be started in situ, in the apparatus, and in thin protein films. The
use of caged compounds offers an elegant approach to initiate protein
reactions with a nanosecond UV laser flash. Here, time-resolved FTIR and
2h 10 rain
I
Fic. 15. Three-dimensionalglobalfit representationof the GTPasereaction of H-ras p21;

10 mM H-ras p21-GTP, 25 mM MgC12,10 mM DTT, 50 mM HEPES, pH 7.5, 1% glycerol.
[ l 3] BIOELECTRONIC IMAGING AND PATTERN RECOGNITION 245
F F - R a m a n photolysis studies of the commonly used caged compounds,

caged Pi, caged ATP, caged GTP, and caged calcium are presented. The
use of specific isotopic labels allows us to assign the IR bands to specific
groups. Because metal ions play an important role in many biological
systems, their influence on FTIR spectra of caged compounds is discussed.
The results presented should provide a good basis for further FTIR studies
on molecular reaction mechanisms of energy or signal transducing proteins.
As an example of such investigations, the time-resolved FTIR studies on
the GTPase reaction of H-ras p21 using caged GTP is presented.
Acknowledgments
This work was supported by the Deutsche Forschungsgemeinschaft, SFB 394, Teilprojekt
B1. We thank Professor R, S. Goody for the gift of GTPflS and GTPc~S and Dr. D. Kuschrnitz
for stimulating discussion.
[13] Use of Caged Compounds in Studies of Bioelectronic

Imaging and Pattern Recognition
By C. W. WHARTON,R. A. MELDRUM,and R. S. CHITTOCK
Introduction
The main application of caged compounds has been to investigate the
fast pre-steady-state kinetics of enzyme reactions. The photoactivatable
nature of caged compounds opens up the possibility of manipulating the
steady-state activity of enzymes using light, creating the potential for a
new range of bioelectronics and molecular electronics applications. The
bioluminescent enzyme firefly luciferase has been used to study possible
methods of enzyme switching using caged compounds. The enzyme reaction
can be summarized asl:
luciferase
ATP + 02 + luciferin ~ AMP + PPi + oxyluciferin + CO2 + light
(hv = 562 nm)
The heterocyclic substrate luciferin is converted to oxyluciferin in an excited
state that emits the bioluminescent photon as it returns to the ground state.
Photoactivation of an enzyme is clearly possible using a caged substrate
molecule, and photodeactivation is possible using a caged enzyme inhibitor.
a M. DeLuca and W. D. McElroy, Methods Enzymol. 57, 3 (1987).

[ l 3] BIOELECTRONIC IMAGING AND PATTERN RECOGNITION 245
F F - R a m a n photolysis studies of the commonly used caged compounds,

caged Pi, caged ATP, caged GTP, and caged calcium are presented. The
use of specific isotopic labels allows us to assign the IR bands to specific
groups. Because metal ions play an important role in many biological
systems, their influence on FTIR spectra of caged compounds is discussed.
The results presented should provide a good basis for further FTIR studies
on molecular reaction mechanisms of energy or signal transducing proteins.
As an example of such investigations, the time-resolved FTIR studies on
the GTPase reaction of H-ras p21 using caged GTP is presented.
Acknowledgments
This work was supported by the Deutsche Forschungsgemeinschaft, SFB 394, Teilprojekt
B1. We thank Professor R, S. Goody for the gift of GTPflS and GTPc~S and Dr. D. Kuschrnitz
for stimulating discussion.
[13] Use of Caged Compounds in Studies of Bioelectronic

Imaging and Pattern Recognition
By C. W. WHARTON,R. A. MELDRUM,and R. S. CHITTOCK
Introduction
The main application of caged compounds has been to investigate the
fast pre-steady-state kinetics of enzyme reactions. The photoactivatable
nature of caged compounds opens up the possibility of manipulating the
steady-state activity of enzymes using light, creating the potential for a
new range of bioelectronics and molecular electronics applications. The
bioluminescent enzyme firefly luciferase has been used to study possible
methods of enzyme switching using caged compounds. The enzyme reaction
can be summarized asl:
luciferase
ATP + 02 + luciferin ~ AMP + PPi + oxyluciferin + CO2 + light
(hv = 562 nm)
The heterocyclic substrate luciferin is converted to oxyluciferin in an excited
state that emits the bioluminescent photon as it returns to the ground state.
Photoactivation of an enzyme is clearly possible using a caged substrate
molecule, and photodeactivation is possible using a caged enzyme inhibitor.
a M. DeLuca and W. D. McElroy, Methods Enzymol. 57, 3 (1987).

246 CACZOCOMPOUNDS [131
In the former case we describe the use of caged ATP to activate firefly
luciferase, producing a light-in/light-out system. In the latter case, luciferase
bioluminescence has been inhibited by the photolysis of caged acetic acid.
Acetate is not itself an inhibitor of luciferase, but because the optimum
pH of luciferase (7.8) is well above the pKa of acetic acid, the net result
of the uncaging process is the release of a proton (Fig. 1). If the reaction
is carried out in a weakly buffered solution, the resultant change in pH
will be sufficient to decrease enzymatic activity. Because the activity of
all enzymes displays some pH dependence, this method will be generally
applicable to the light-activated control of biological activity.
In order to move toward potential molecular electronics devices, a
further development is required to allow spatially resolved switching of
luciferase activity by immobilizing the luciferase molecules in an agarose
gel containing luciferin and caged ATP. Illumination of the gel through a
mask will activate the luciferase by the photolysis of caged ATP only in
the illuminated areas to produce a bioluminescent image of the mask.
Combination of this method with a lithographically produced active lucifer-
+H +
\ ---% \\ \ \
-0 -Z
~ c ~
N~O
I
o
o-
FIG. 1. Mechanism of caged acetic acid photolysis showing the light-induced release of a
proton followedby a molecular rearrangement leading to the release of acetate. Physiological
pH is much greater than the pKa of acetic acid, so the overall reaction produces a proton
plus acetate, allowing the modulation of enzyme activity by a fight-induced pH change.
[131 BIOELECTRONIC IMAGING AND PATTERN RECOGNITION 247
ase image within the gel has been used to demonstrate a rudimentary
pattern correlator. 2 Inclusion of caged acetate in the agarose gel has been
used to demonstrate two-wavelength switching by using chemical modifica-
tion to shift the absorbance spectrum of one of the caging groups, allowing
one wavelength to activate luciferase by photolysis of caged ATP, while a
second wavelength can produce deactivation by uncaging of caged acetic
acidJ The absorbance shift was achieved by synthesis of the 3,4-dimethoxy-
substituted analog of caged ATP (dm-ATP), which has an absorbance
maximum at 360 nm compared to 310 nm for the standard caging group.
Luciferin has also been caged and provides a third potential method for
light-mediated control of bioluminescence.4
This article describes the synthesis of caged acetate and the dimethoxy-
substituted analog of caged ATP, methods by which luciferase may be
immobilized in an agarose gel, the characterization of changes in pH induced
by photolysis of caged acetate, and the switching of luciferase biolumines-
cence in an agarose gel.
Preparative M e t h o d s
Caging of molecules with the 1-(2-nitrophenyl)ethyl group (the "stan-
dard" caging group) uses nitroacetophenone (Aldrich, Milwaukee, WI) as
the starting point for the synthesis. The corresponding dimethoxy com-
pound required to produce the 1-(3,4-dimethoxy-6-nitro)ethyl caged com-
pounds, 3,4-dimethoxy-6-nitroacetophenone, is not available commercially
and so is synthesized by the nitration of dimethoxyacetophenone (Aldrich)
by nitric acid. 5'6
Both caging precursors are converted to the corresponding acetophe-
none hydrazone by the method of Walker et al. 7 as modified by Meldrum et
al.S and subsequently to the diazoethane by manganese(IV) oxide reduction.
The molecule to be caged is added immediately to produce the caged com-
pound.
2 D. G. Lidzey, R. S, Chittock, N. Berovic, C. W. Wharton, J. B. Jackson, and T. D. Beynon,

Adv. Mater Opt. Elec. 4, 381 (1994).
3 R. S. Chittock, D. G. Lidzey, N. Berovic, C. W. Wharton, J. B. Jackson, and T. D. Beynon,
Adv. Mater Opt. Elec. 4, 349 (1994).
4 j. Yang and D. B. Thomason, Biotechniques 15, 848 (1993).
s C. A. Fetscher, Organic Syntheses Collected Volumes 4, 735 (1963).
7 j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. U0,
7170 (1988).
8 R. A. Meldrum, S. Shall, D. R. Trentham, and C. W. Wharton, Biochem. J. 266, 885 (1990).
Caged ATP molecules are purified by high-performance liquid chroma-

tography (HPLC) using a Technopak 10C18 reversed-phase column (HPLC
Technology, Macclesfield, UK). A gradient of 0-40% methanol in 10 mM
phosphate buffer gives a final product of greater than 98% purity. Unreacted
ATP is eluted from the column after approximately 3 min; caged ATP is
eluted after approximately 10 min, representing >99% of nucleotide present
in the preparation as estimated from the relative areas of the HPLC peaks.
Caged acetic acid is purified using flash chromatography, using a silica
gel column eluted with 80% hexane : 20% ethyl acetate. The product, recov-
ered by rotary evaporation, is a yellow oil that runs as a single spot on
thin-layer chromatography (TLC) with 80% hexane : 20% ethyl acetate. The
product is dissolved in 1 ml dimethyl sulfoxide (DMSO) and stored in the
dark at - 2 0 ° until required.
The luciferase used in these experiments is purchased from Sigma (St.
Louis, MO). Stock enzyme solution is stored in 25 mM glycylglycine,5 mM
MgC12, 50 mM 2-mercaptoethanol, 0.2% Triton X-100, 50% glycerol, pH
7.8, at a concentration of 1 mg/ml. Under these conditions the enzyme is
stable for up to 2 weeks.
Light-Induced pH Changes Using Caged Acetic Acid

In order to demonstrate the effect of uncaging caged acetic acid on
solution pH, a 2.4 mM solution of caged acetic acid in 1 mM glycylglycine
buffer is illuminated with a 100-W UV lamp, and the pH of the solution
is recorded as a function of time using a pH meter (Fig. 2). Prior to UV
illumination, the pH of the solution is stable at ambient light levels. On
0.5
T i m e (sec)
, .lalLt,LJ I I I
0 .i... , i
500
-0.5
-1.5
UV L i g h t O n
-2
FIc. 2. Light-inducedpH changes induced by the photolysisof caged acetic acid. The
initial conditionswere 2.4 mM caged acetic acid in 1 mM glyclyglycine,pH 7.8. The pH of
the solutionwas measuredusing a standardpH meter.
[ 13] BIOELECTRONIC IMAGING AND PA'I'TERN RECOGNITION 249
UV illumination, a change of approximately 1.5 pH units is produced within

400 sec, sufficient to reduce luciferase activity to 9% of its initial optimum
activity at pH 7.8. The slow initial rate of pH change corresponds to the
peak buffering capacity of the glycylglycine buffer. At more acidic pH the
buffering capacity of the solution becomes less, and so the rate of pH
change at constant illumination becomes greater.
Immobilization of Luciferase and Caged Molecules in Agarose Gels

Immobilization of luciferase, a necessary prerequisite for the creation
of bioluminescent patterns, is conveniently accomplished by encapsulation
of the enzymes, substrates, and caged compounds in an agarose gel. To
prevent heat denaturation of luciferase, it is necessary to use an agarose
gel with a melting point below 30°, such as Agarose Prep (LKB, Bromma,
Sweden). Gels are prepared by dissolving solid agarose in 25 mM glycylglyc-
ine, 5 mM magnesium chloride, 50 m M DTT to give a final volume of 1%
(w/v) by heating. In preparations using caged acetic acid to produce light-
induced pH changes, the concentration of glycylglycine is reduced to 1 mM.
While still liquid the solution is transferred to a water bath at 28 °, at
which temperature the gel remains liquid but the luciferase is not thermally
denatured. Luciferase, luciferin, dm-ATP, and caged acetic acid are added.
a b
1 cm 1 cm
FIG. 3. (a) A triangular pattern written onto an agarose gel by photolysis of caged ATP.
Conditions in the gel were 25 m M glycylglycine, 5 m M MgC12,50 m M DTF, 1.6/~M luciferase,
40/zM luciferin, 80/zM caged ATP, 1% agarose, pH 7.8. Photolysis was achieved by illuminating
the gel with a 100-W UV lamp for 30 sec. (b) A bioluminescent circular pattern written onto
agarose gel by photolysis of dm-ATP by illumination through a mask using light of h > 350
nm and a dark triangular pattern written onto the same gel by deactivation of luciferase by
photolysis of caged acetic acid by illumination through a mask without a filter. The agarose
gel consisted of 1 m M glycylglycine, 5 m M MgC12, 50 m M DTT, 90 n M luciferase, 100/zM
luciferin, 2.8 m M dm-ATP, 5.7 m M caged acetic acid, 1% agarose prep, pH 7.8. Illumination
time to achieve photolyse was 30 sec for dm-ATP and 120 sec for caged acetic acid.
For uncaging of caged ATP, final concentrations are 1.6/zM luciferase, 40

/xM luciferin, and 80 ~M caged ATP. For two wavelength experiments,
final concentrations are 90 nM luciferase, 100/xM luciferin, 2.8 mM dm-
ATP, and 5.7 mM caged acetic acid. This preparation is cast between glass
plates separated by 0.75-mm spaces and allowed to set for 1 h at 4 °. After
removal of the top plate the solid gel is cut into 2- to 4-cm2 samples for
use. The gels may be stored in liquid nitrogen, losing approximately 50%
of their bioluminescent activity after 250 hr.
Production of Biolurninescent Images
A collimated 100-W UV source fitted with appropriate filters for wave-
length discrimination is used to illuminate the gel-encapsulated luciferase.
Typical uncaging illumination times are 30 sec for dm-ATP and 120 sec
for caged acetic acid. Images of the bioluminescence are recorded within
2 rain of illumination of the gel using an intensified CCD camera (Photek,
St. Leonards-on-sea, Sussex, UK) attached to a DL2867-LC frame grabber
(Data Translation, Marlboro, MA).
Figure 3a shows a bioluminescent pattern of a triangular mask produced
by a 30-sec illumination of a 0.75-mm-thick agarose gel containing 1.6/xM
luciferase, 40 tzM luciferin, and 80 /zM caged ATP. Bioluminescence is
observed only in those areas of the gel illuminated through the mask. The
pattern decays with a half-life of 10 min at 18°. In Fig. 3a, the side of the
triangular pattern is approximately 10 mm in length. The best edge resolu-
tion possible under these conditions is estimated to be approximately 0.7
mm using a variable-spaced grating. 2
Figure 3b shows a bioluminescent pattern produced by illumination of
a gel containing both dm-ATP and caged acetic acid. First, a bioluminescent
pattern is produced by illumination of the gel through a circular mask with
a 350-nm cutoff filter in front of the UV lamp. Under these conditions,
dm-ATP is photolyzed, activating luciferase in the illuminated areas but
not affecting caged acetic acid. Second, the gel is illuminated through a
triangular mask with the cutoff filter removed. Under these conditions,
caged acetic acid is photolyzed, lowering the pH of the gel in the illuminated
areas and deactivating luciferase activity. The image of the triangular mask
appears as a dark area within the bright area of bioluminescence. A control
experiment demonstrates that the deactivation of luciferase by the second
illumination is not the result of luciferase or luciferin degradation by UV
light. 3
These two experiments are a sample demonstration of the spatial re-
solved switching of luciferase activity using caged compounds, including a
means of wavelength-selective uncaging of different molecules in the same
preparation by chemical modification of the caging group.
[141 UNSTABLEINTERMEDIATES 251
[I 4] Use of Caged Nucleotides to Characterize Unstable

Intermediates by X-Ray Crystallography
By A X E L J. SCHEIDIG, CHRISTOPH BURMESTER, a n d R O G E R S. G O O D Y
Introduction
Despite the enormous contributions of X-ray crystallography to our
understanding of the structure of biological macromolecules, unraveling
the mechanism of the interaction of such a molecule with a ligand or
substrate normally involves the indirect process of combining information
from static high-resolution structures with dynamic information derived
using other methods, including enzyme kinetics and spectroscopy, as well
as molecular biological techniques, such as site-directed mutagenesis. The
dramatic progress made in increasing the speed of X-ray crystallographic
data collection has led to efforts to exploit this speed to obtain structural
information on previously inaccessible states that occur as intermediates,
particularly in enzymatic reactions. The advance in X-ray diffraction meth-
ods has been so profound that the time needed for data collection is no
longer the main limitation in such experiments. The biggest remaining
hurdle is that of triggering a reaction in a crystal of a macromolecule, and
due to the nature of the difficulties involved, it is unlikely that there will
be a general solution to this problem. One method that may have a some-
what more general application is the use of caged compounds for generating
a ligand such as a substrate or other essential molecule in crystals of the
macromolecule in order to start the reaction.
Requirements for Caged Substances Used in Starting Reactions

in Crystals
Many potential applications of caged compounds require the release of
a substrate for an enzyme in the crystalline form in such a manner that the
reaction catalyzed by the enzyme can be started rapidly and efficiently. In
principle, two extreme situations may be encountered. In one of these, the
caged substance does not interact with the active site of the enzyme and
the active site will be occupied only after photolytic release of the substrate.
In principle, this situation would also allow monitoring of events associated
with the binding of the substrate to the active site. For instance, substrate-
induced conformational changes have been identified in many mechanistic
studies and are often postulated as being required for catalysis to occur,

but they have not, in most cases, been characterized structurally in detail.
The second type of situation that might be encountered is that in which
the caged ligand binds tightly to the active site of the protein or is covalently
bound to the protein or nucleic acid, and the reaction is started by stoichio-
metric conversion of this molecule or group to its active form. Both types
of situation have been encountered in attempts to perform time-resolved
protein crystallography, and the photochemical requirements for the caged
group must be considered with respect to the properties of the caged
molecule as a potential ligand.
Most of this article is devoted to an example of the second type, but
some of the potential questions arising with the first situation are discussed
briefly. In the extreme situation in which there is no interaction between
a caged substrate and a protein in a crystal, one of the main problems is
concerned with the nature of the crystal packing and, in particular, whether
there are solvent channels that allow easy access to the active site. Even
if this is the case, the question has to be considered as to whether a suffi-
ciently high concentration of the caged substance can be reached to allow,
potentially, an amount of substrate to be released by photolysis that is
sufficient to saturate the enzyme active site, both in terms of absolute
amounts and in terms of concentration relative to the dissociation constant
for binding to the active site. To appreciate that this is not a trivial point,
it must be realized that typical protein concentrations in crystals are in the
range of about 5-50 mM. A very high concentration of the caged version
of the specific ligand would be needed to ensure reaching this concentration
in the crystal, even if the photolysis reaction is highly efficient and the
available light intensity is enough to photolyze a significant fraction with,
ideally, one short flash.
Situations of this type have been encountered with glycogen phosphory-
lase using 3,5-dinitrophenyl phosphate (DNPP) as a caged phosphate
molecule 1,2 and with isocitrate dehydrogenase using caged isocitrate [1-
(2-nitrophenyl)ethyl 1-hydroxy-l,2-dicarboxy-3-propane carboxylate] as a
substrate precursor) In the first example, DNPP can be converted to inor-
ganic phosphate and 3,5-dinitrophenol at a rate of about 104 sec -1 (the
rate-limiting step in the dark reaction) by irradiation at 300-360 nm with
a quantum yield of 0.67. 4 A crystal of the phosphorylase is soaked in 45
mM DNPP, excess mother liquor is removed, photolysis is performed with
a continuous nitrogen laser at 337 nm, and the photolytic reaction is moni-
1 A. T. Hadfield and J. Hajdu, J. Appl. Crystallogr. 26, 839-842 (1993).

z A. Hadfleld and J. Hajdu, 3". MoL Biol. 236, 995-1000 (1994).
3 M. J. Brubaker, D. H. Dyer, B. Stoddard, and D. E. Koshland, Jr., Biochemistry 35, 2854-
2864 (1996).
4 j. E. T. Corrie and D. R. Trentham, in "Bioorganic Photochemistry," vol. 2 (H. Morrison,
ed.) pp. 203-205. Wiley, New York, 1993.
l 141 UNSTABLEINTERMEDIATES 253
tored spectrophotometrically. In these crystals, the protein concentration,

and therefore the concentration of binding sites for phosphate, is about
7 mM. In a large crystal, most of the photolysis occurs after 1 min in a
surface layer due to attenuation of the photolysis beam by absorption. In
this situation, the actual rate of release of phosphate is not rate controlling;
the rate of diffusion from the front of the crystal into regions that are
inadequately illuminated is rate controlling. For slow reactions in the crystal
following photolysis, the problem of nonuniform illumination and release
can be overcome, or at least reduced, by rotation of the crystal during
illumination. In addition, crystals that are significantly smaller than those
used in these studies can be used for collecting data, especially using a
synchrotron source. Structures are frequently determined using crystals
with dimensions of significantly less than 100/zm in the thinnest direction,
which considerably reduces the problem of absorption. Moving away from
the wavelength for maximum absorption and correspondingly increasing
the illuminating light intensity to excite the same number of molecules
results in more uniform photolysis.
It should be noted here that the absorption spectrum of the photolysis
product arising from the cage group is, in general, different from that of
the caging group before photolysis. Depending on the exact nature of the
light source for photolysis, this can be of significance. Using a continuous
source, there is a build up of the photolysis product during the illumination
procedure. If the source has a higher absorption at the wavelength used,
it exacerbates the problem of nonuniform photolysis throughout the crystal.
With a pulsed source, this problem is less important, although this may not
apply to sources with pulses that are long in comparison with the rate of
release. There is the additional factor that generation of a significant fraction
of the photochemically excited state may lead to similar problems, de-
pending on its spectral properties.
A potential problem arises with systems in which the affinity of the
caged compound for the active site is not low enough so that there is no
binding at all to the protein being examined, but is not high enough to
form a stoichiometric complex. An example of this type is found in the use
of caged ATP for the initiation of relaxation or contraction of muscle fibers,
followed by monitoring of rapid structural changes by low-angle X-ray
diffraction.5-7 In early experiments of this type, it was assumed that caged
ATP does not bind to actomyosin. In a typical experiment, 10 mM caged
5 G. Rapp, K. J. V. Poole, Y. Maeda, K. Giith, and R. S. Goody, Biophys. J. 50, 993-997 (1986).
6 K. J. V. Poole, G. Rapp, Y. Maeda, and R. S. Goody, in "Topics in Current Chemistry"
(E. Mandelkow, ed.), pp. 1-29, Springer-Verlag, Berlin, Heidelberg, Germany, 147 (1988).
7 G. Rapp, K. J. V. Poole, Y. Maeda, G. C. R. Ellis-Davies, J. H. Kaplan, J. McCray, and
R. S. Goody, Berichte der Bunsen Gesellschafl 93, 410-415 (1989).
ATP is diffused into muscle fibers and approximately 2 mM ATP is released
by a pulse from a flash lamp. The results are interpreted in terms of associa-
tion of ATP release at a known rate free in solution with the active site of
myosin. Caged ATP binds to actomyosin with a Kd of about 1 mM. 8 Thus,
before photolysis, about 90% of the ATP-binding sites are occupied by
caged ATP at the concentration used.
Assuming that the quantum yield for photolysis of caged ATP at the
myosin active site is the same as that for caged ATP in solution, ATP is
generated directly in 18% of the sites, whereas the remaining 82% recruit
ATP from solution. This second group of active sites is, however, not free
for direct association of ATP because this group is largely saturated by
caged ATP. Depending on the magnitude of the rate constant of dissociation
of caged ATP (which is not known), it is rate limiting for the association
of ATP, or, if it is very rapid, caged ATP effectively buffers the rate of
ATP association. In both cases, this slows the rate of ATP binding, which
presents a potential problem, depending on the aim of the experiment. It
should be noted that these problems apply not only to time-resolved X-ray
diffraction studies, but to other studies using caged nucleotides as the
starting point for investigations on muscle fibers or other actomyosin-
containing systems, including time-resolved electron microscopy, spectros-
copy, and kinetic and mechanical measurements.
The situation becomes much clearer if the caged compound binds with
high affinity to the protein in the crystal. However, the presence of a
stoichiometric amount of the cage group with respect to the protein also
introduces new problems. The general considerations concerning unifor-
mity of release of the substance of interest throughout the crystal still apply,
particularly with respect to attenuation of the illuminating light due to
absorption by the cage group, its products, or photolysis intermediates.
The main complication is that if the reaction of interest is fast enough to
require that the released ligand, substrate, or group be generated with a
single light pulse, theoretically a quantum yield of 1 and enough light to
excite all molecules of the cage group present are required to obtain com-
plete conversion to the first complex of interest. In practice, often only an
approximation to this ideal situation is possible. For reactions slow enough
to allow use of a flash lamp with a duration of about 1 msec, this might be
achievable in some cases. This arises from the fact that with relatively long
light pulses, there is the possibility of multiple excitations, depending on
the exact kinetics and mechanism of photolysis. When the time require-
ments are not so severe, and a series of pulses or continuous radiation can
8 T. Thirlwell, J. A. Sleep, and M. A. Ferenczi, J. Muscle Res. Cell Motil. 16, 131-137 (1995).
[14] UNSTABLEINTERMEDIATES 255
be used for photolysis, these problems become less significant, and this
applies to the case discussed in detail later.
A similar situation arises in the case of caged enzymes. For chymotryp-
sin, caging of the enzyme has been achieved by esterifying the serine group
that is involved in the reaction mechanism. 9,a° The trans-p-diethylamino-
o-hydroxymethyl cinnamate group was used for this purpose, because on
light-induced isomerization of this derivative, deacylation occurs. This
would then allow a reaction to start in the crystal, although there might
still be the problem of slow diffusion out of and into the active site. In the
work reported on this system, the cinnamate group was photolyzed in the
presence of a suicide substrate, which then reacts over a period of several
hours with the enzyme. 11 Because this reaction is also very slow in solution,
the experiments do not allow conclusions regarding the applicability of the
method to more rapid reactions.
Data Collection Strategies

The enormous increase in X-ray intensity available from synchrotron
sources which has been achieved in the last two decades allows diffraction
data to be collected orders of magnitude more rapidly than using conven-
tional sources. However, if monochromatic X-rays are used, there is still
the problem that a large number of data frames must be collected at
orientations of the crystal differing with respect to the beam in order to
sample a major part of the reciprocal lattice. Although individual frames
can be collected in milliseconds or less, the technical aspects of changing the
crystal orientation and data readout can decrease the achievable temporal
resolution by orders of magnitude. One potential solution to this problem
is the use of Laue diffraction conditions, i.e., the use of polychromatic
X-rays to allow collection of a large fraction of the diffraction data in one
or a few exposures. This approach has been used in several applications,
and the advantages and disadvantages are summarized briefly here. The
main advantage is, of course, speed of data collection, although depending
on the space group of the crystal, there will generally be a need to obtain
more than one exposure at varying orientations. This often leads to the
situation in which structures are calculated with significantly less than an
optimal fraction of the theoretical data. Added to this is the fact that in Laue
9 B. L. Stoddard, J. Bruhnke, N. A. Porter, D. Ringe, and G. A. Petsko, Biochemistry 29,

4871-4879 (1990).
10B. L. Stoddard, P. Koenigs, N. Porter, D. Ringe, and G. A. Petsko, Biochemistry 29,
8042-8051 (1990).
n B. L. Stoddard, P. Koenigs, N. Porter, K. Petratos, G. A. Petsko, and D. Ringe, Proc. Natl.
Acad. Sci. U.S.A. 88, 5503-5507 (1991).
data sets, low to medium resolution data are inherently underrepresented.
Crystal quality is also more critical than for monochromatic data collection.
Thus, in one example, about 100 crystals that were suitable for monochro-
matic diffraction, only a few were usable for Laue diffractionJ z Despite
these limitations, the Laue method in combination with flash photolysis
has been used to obtain millisecond 13 and even nanosecond 14 resolution
with inherent (built-in) photosensitive groups.
Recent work, some of which is described below, suggests that for rela-
tively slow reactions (which are, however, still too fast to allow triggering
by more conventional methods) it may be possible to adopt a different
approach, which is to photolyze and then freeze crystals rapidly to a temper-
ature at which the reaction of interest is so slow that monochromatic diffrac-
tion data can be collected. A variation on this theme is to perform the
photolysis at a temperature at which the light-induced reaction can occur,
but subsequent reactions are slow. An example of this is the photodissocia-
tion of carbon monoxide from myoglobin at temperatures of 2015 or 40 K. 16
Under these conditions, relatively gentle constant illumination generates
the first dissociated state of carbon monoxide and allows structural charac-
terization before rebinding or subsequent steps in dissociation can occur.
Monitoring of Reactions in Crystals

In many applications involving the use of caged compounds for the
initiation of reactions in crystals, knowledge of the degree of photolysis
and of the extent of the reaction of interest will be needed. In most cases,
this has been achieved after completion of the experiment by analysis
of the crystal, for example, by high-performance liquid chromatography
(HPLC) to detect photolysis products and where appropriate the products
of an enzymatic reaction. Spectroscopic methods, mainly absorption spec-
troscopy, have been applied to protein crystals and have been used to
monitor reactions in crystals. 2'17For many applications of caged compounds,
12L Schlichting, S. C. Almo, G. Rapp, K. Wilson, K. Petratos, A. Lentfer, A. Wittinghofer,

W. Kabsch, E. F. Pai, G. A. Petsko, and R. S. Goody, Nature 345, 309-315 (1990).
~3U. K. Genick, G. E. O. Borgstahl, K. Ng, Z. Ren, C. Pradervand, P. M. Burke, V. Srajer,
T. Y. Teng, W. Schildkamp, D. E. McRee, K. Moffat, and E. D. Getzoff, Science 275,
1471-1475 (1997).
~4V. Srajer, T. Y. Teng, T. Ursby, C. Pradervand, Z. Ren, S. Adachi, W. Schildkamp, D.
Bourgeois, M. Wulff, and K. Moffat, Science 274, 1726-1729 (1996).
15 I. Schlichting, J. Berendzen, G. N. Phillips, Jr., and R. M. Sweet, Nature 371, 808-812 (1994).
16T.-Y. Teng, V. Srajer, and K. Moffat, Nature Struct. Biol. 1, 702-705 (1994).
17y, Chen, V. Srajer, K. Ng, A. LeGrand, and K. Moffat, Rev. Sci. Instrum. 65, 1506-1511
(1994).
[ 141 UNSTABLEINTERMEDIATES 257
the ability to monitor photolysis and enzymatic reactions in real time during
an experiment will allow better decisions to be made with respect to time
windows for data collection and will also contribute to data analysis by
removing uncertainty concerning the exact composition of the crystals at
the time of diffraction data collection. Absorption (discussed briefly earlier
for the case of glycogen phosphorylase) and fluorescence spectroscopy are
two methods likely to be useful, but other approaches, such as Fourier
transform infrared (FTIR) spectroscopy, could be of great value in cer-
tain systems.
Specific Example: Time-Resolved Crystallography on H-Ras GTPase

One of the first examples of the use of caged compounds for time-
resolved crystallography invovled the small GTPase H-Ras, a protein that
plays a key role in signal transduction in the cell. Because the most impor-
tant property of this protein is a conformational change that occurs on
GTP hydrolysis, it was of interest to characterize this change in crystals of
the protein. There was also an expectation that the results could provide
some evidence on the detailed chemical mechanism of the GTPase reaction.
This is not only of importance at the fundamental level of understanding
the hydrolysis mechanism, but also for understanding the mechanism of
oncogenic transformation of this protooncogene because many trans-
forming mutations result in the loss of ability to hydrolyze GTP.
Early experiments showed that the H-Ras protein, which has a very
high affinity to GDP and GTP, also binds caged GTP [we refer to GTP
bearing the 1-(2-nitrophenyl)ethyl group on the T-phosphate as caged GTP]
with high affinity (approximately 101° M -1) so that it is relatively easy to
prepare a 1 : 1 complex between the protein and the nucleotide. 18 As can
be seen in Fig. 1, the presence of a chiral center in the cage group means
that caged GTP exists as a mixture of two diastereomers, and the first
experiments were performed with this mixture. It was shown by HPLC
that the crystals contain an approximately equimolar mixture of the two
diastereomers. Photolysis of the caged nucleotide could be achieved in the
crystal using a xenon flash lamp. 18a High-performance liquid chromatogra-
phy analysis showed that, after photolysis, hydrolysis of GTP to GDP took
place in the crystal at a rate constant that was similar to that seen in solution.
As long as sufficient amounts of reducing agents [normally dithiothreitol
(DIT)] were present to react with and inactivate the nitrosoacetophenone
18I. Schlichting, J. John, G. Rapp, A. Wittinghofer, E. F. Pai, and R. S. Goody, Proc. Natl.
Acad. Sci. U.S.A. 86, 7687-7690 (1989).
18aG. Rapp, Methods Enzymol. 291 [11] (1998) (this volume).
hv H2 0
.~ .., p21:GTP ~ p21:GDP + Pi
A--*B ....... ~-C

100
A
~ 60
20
0 10 20 30
time (min)
photolysis
diffraction
FIG. 1. Experimental scheme. Flow diagram for the generation of p21 : GTP and p21 : GDP
complexes starting with p21 : caged GTP. After photolysis the cage group 2-nitrosoacetophe-
none is released, and with a half-time of approximately 30 min (at room temperature) the
generated GTP will be hydrolyzed to GDP and inorganic phosphate. Diffraction data were
collected before photolysis,2 rain after photolysis,and after completion of the hydrolysisre-
action.
photolysis product, the crystals survived both the photolysis and the hydro-
lysis reaction without macroscopic signs of damage. It could be confirmed by
X-ray diffraction that the combined processes of photolysis and hydrolysis
occurred without loss of crystal quality as long as several precautions were
taken to prevent heating and damage due to light in the far-UV range.
One of these was the use of a filter that allowed only light between 300
and 400 nm to illuminate the crystal. Another was to keep the crystal
bathed in mother liquor to behave as a heat sink. 18a
Our experiments with H-ras p21 showed that the photolysis yield in
dilute solutions is much higher than in the crystallized system. Multiple
factors contribute to this. Some of these are differences in surface reflection
(flat cuvette versus curvature of capillary and shape of crystal); absorption
effects of the capillary and crystallization mother liquor (the quartz material
of UV/VIS cuvettes has a different absorption profile than quartz capillaries
[ 141 UNSTABLEINTERMEDIATES 259
used for mounting crystals); and attenuation of intensity throughout the

crystal arising from absorption due to the higher concentration of chromo-
phore in the crystal. It is therefore not trivial to extrapolate from the
solution results for complete photolysis to the situation in the crystal making
use of the parameters of total effective energy per pulse, and the length and
number of pulses. We have therefore optimized conditions for photolysis
working only with crystals and analyzing the efficiency of photolysis, the
activity of the crystallized protein, and the diffraction quality of the crystal.
The actual procedure for photolysis in the earlier X-ray diffraction experi-
ments was to use about 10 pulses of the xenon lamp separated by sev-
eral seconds.
Because the half-life of the GTPase reaction of H-Ras is of the order
of 30 min at room temperature, diffraction data must be collected within
a few minutes of photolysis to characterize the GTP state. Clearly, this
cannot be done by conventional methods, and data were collected at a
synchrotron source using the white X-ray beam (actual range of wavelengths
0.6 to 2.0/~). The first data set was collected at about 4 min after photolysis.
This data set was used to calculate an electron density map at about 2.8
~.12 It showed that the overall structure of the protein was very similar to
that which had previously been determined for the complex between H-Ras
and GppNHp, a nonhydrolyzable GTP analog. The binding of GTP ap-
peared to be identical at this resolution to that of GppNHp. In further
experiments, crystals were aged after photolysis in order to allow hydrolysis
to GDP to occur, and the structure of the resulting complex was determined
by conventional monochromatic diffraction methods. This led to a detailed
description of the structural changes accompanying GTP hydrolysisY
Despite the success of these experiments, not all questions of interest
could be answered. This arises mainly from the relatively low resolution
of the structures determined and the less than optimal quality of the electron
density maps. There are two main reasons for this. One is the limitation
imposed by the Laue diffraction geometry and the general low completeness
of the data sets obtained. The other is the intrinsically low resolution of
the crystals used, which was not considerably better than 2.8/~ even under
monochromatic conditions. It was suspected that the presence of the mix-
ture of diastereomers of caged GTP in crystals used in the earlier work
was a limiting factor in crystal quality, and this was strengthened by the
results of structural determination of the H-Ras-caged GTP complex with
the mixed isomers, which suggested an anomalous mode of binding. 12 In
an attempt to overcome this problem, other cage groups were examined.
The most easily available is the 2-nitrobenzyl group, which is not chiral.
Its inferior photolysis properties (low quantum yield, relatively slow dark
reaction) would probably be acceptable for slow GTPases if the crystalliza-
tion properties of its complex with the protein were better than those of
the classical cage group. Surprisingly, the best crystals that could be obtained
diffracted only to 3.6 ~,19 compared with 2.8 ~ for the mixed isomers of
caged GTP. A further group that was examined in an attempt to solve this
problem is the bis(2-nitrophenyl)methyl group. 19 Like the 2-nitrobenzyl
group, this group does not have a chiral center, but its photolysis properties
are more similar to those of the classical cage group. Again, crystals were
obtained, but the best resolution obtained was 2.6 /~, which was not a
significant improvement over the crystals with mixed isomers of caged GTP.
A solution was provided by the preparation of the pure diastereomers of
caged GTP. 2° It was found that complexes of H-Ras with the separated
diastereomers did indeed lead to better crystal quality than the mixed
isomers. ~a The complex of H-Ras p21 with the R-isomer of caged GTP
diffracted to 1.85 ~ and that with the S-isomer to 2.2/~.21 An interesting
but puzzling observation is that the S-isomer is degraded at a significant
rate to GDP and caged phosphate in the crystals, whereas the R-isomer is
much more stable. Determination of the structure at 1.85/~ showed that
the caged nucleotide was bound normally at the active site of the protein,
but that the cage group forces two loops to adopt new conformations,zl
Crystals of the complex with the pure R- or S-isomers were used for
time-resolved structural studies in a manner similar to that used earlier,
with the modification that the crystal was rotated between flashes to coun-
teract the problem of nonuniform release of G T P . 19 A further modification
to the earlier experiments was that a mutant of H-Ras (G12P) was used.
This mutant has similar GTPase properties to wild-type protein but a higher
thermal stability. The wavelength range used in these experiments was
0.45-2.6/~. Data sets were collected at 2-4, 11-13, 20-22, 30-32, and 90-92
min after photolysis. Despite the significant improvement in crystal quality
under monochromatic conditions, under Laue conditions the resolution
was limited to 2.8/~. Despite this, the earlier interpretation could be ex-
tended. In particular, a conformation of the side chain of Gin-61 was identi-
fied that had not been previously detected. This residue is known to be
directly involved in the mechanism of GTP hydrolysis. Importantly, it could
be shown that the electron density representing the y-phosphate group
disappears smoothly during the reaction, confirming that a protein-product
complex with a long-lived inorganic phosphate does not accumulate to a
19 A. Scheidig, A. Sanchez-Llorente, A. Lautwein, E. F. Pai, J. E. T. Corrie, G. Reid, A.

Wittinghofer, and R. S. Goody, Acta Crystallographica D, DS0, 512-520 (1994).
20 j. E. T, Corrie, G. P. Reid, D. R. Trentham, M. B. Hursthouse, and M. A. Mazid, J. Chem.
Soc. Perkin Trans. 1, 1015-1019 (1992).
21 A. J. Scheidig, S. M. Franken, J. E. T. Corrie, G. P. Reid, A. Wittinghofer, E. F. Pai, and
R. S. Goody, J. Mol. Biol. 253, 132-150 (1995).
[ 14] UNSTABLEINTERMEDIATES 261
significant extent. Regions undergoing structural changes between the GTP

and the GDP states displayed electron density that at this level of resolution
and data quality could not be interpreted in terms of individual conforma-
tions present as a mixture. In addition, the relatively low resolution pre-
vented an interpretation of the changes in positions and density of important
water molecules during the GTPase reaction.
Because of the shortcomings of these experiments, other approaches
have been sought to exploit the full potential of the basic principle. In most
recent work, photolysis of caged GTP at the active site of H-Ras has been
combined with rapid freezing of the crystals to boiling nitrogen temperature
in order to stop the GTPase reaction at controlled time points. This ap-
proach is possible because of the relatively slow GTPase reaction, but it
can be applied to considerably faster reactions, because the rate of cooling
in a stream of nitrogen is likely to be high enough to effectively stop
reactions that occur on the second time scale. In the case of H-Ras, as
with many other protein crystals, the freezing procedure can be performed
without the addition of cryoprotectants because the mother liquors contain
about 30% polyethylene glycol (PEG) 400 and does not form ice crystals
on freezing. Several different strategies for photolysis were examined before
arriving at an optimal procedure. As a light source, both flash lamps and
continuous irradiation from a mercury lamp were used in combination with
filters, as previously described, or with a monochromator. The following
procedure leads to the best results. Relatively small crystals (less than
150 /zm in the largest direction, already mounted in a loop containing
mother liquor) are illuminated continuously with light of wavelength 313
nm for 5 min. The procedure is performed at 2 ° to slow the rate of GTP
hydrolysis and to reduce the negative effects of heating. The crystal is
rotated continuously during illumination. After 5 min, the crystal is allowed
to settle for 2 min and then is rapidly frozen in a stream of evaporating
nitrogen and kept at this temperature (100 K) for the collection of diffrac-
tion data using monochromatic radiation from a rotating anode source
using a multiwire detector.
Each of the factors just mentioned has been tested individually for its
effect on the efficiency of GTP release and its effect on preservation of
crsytal quality. The amount of GTP released was estimated after dissolving
the crystal in water by HPLC. TM It was found that although the total time
required for photolysis was longer using monochromatic light, there was
an advantage in terms of crystal quality. The low temperature during photol-
ysis was also found to have a positive influence. Using the previously
described protocol, frozen crystals of the H-Ras GTP state that diffracted
to 1.6/~ could be obtained routinely. Surprisingly, in many cases there was
an improvement in quality on photolysis of about 0.1-0.2/~.
An important factor in the progress made is the small size of the crystals
used. This is possible because data collection at liquid nitrogen temperature
has the advantage that the crystals become very stable to radiation and
data can therefore be collected over a long period. The small crystal size
makes photolysis more efficient. We have also noted that small crystals
survive the photolysis procedure more readily than large crystals. The
combination of all the factors mentioned has led to an improvement in
data quality from 2.8/~ resolution (<60% completeness) using the flash
lamp and the Laue method at room temperature to 1.6 ,~ resolution (>90%
completeness) using gentle monochromatic photolysis and monochromatic
diffraction in combination with the cryotechnique. Data from several GTP
experiments are presently being refined and evaluated. An example of the
electron density is shown in Fig. 2.
In principle, the protocol described can be used to generate the GDP
state from the GTP state. However, it is technically difficult to mount the
crystal in a loop, photolyze, and then wait for a period of many hours for
hydrolysis to take place, mainly because of the problem of drying. We have
therefore photolyzed crystals in a batch process in a drop of mother liquor
and allowed the crystals to age for about six half-lives of the GTPase
reaction. This results in a loss of resolution to about 2/~, with the best
crystals so far diffracting to 1.8-1.9/~. Further improvement of the protocol
will probably be needed to obtain crystals of the quality of the GTP complex.
Conclusion
The few documented examples of the use of caged substances for initiat-
ing reactions in crystals allow tantalizing glimpses of the potential of the
method but also demonstrate the substantial difficulties associated with
application of the technique to answering meaningful biological questions.
It is sobering to realize that the nearest approximation to this ideal has
been achieved in a system that offers the advantage of a relatively long half-
life of the enzyme-substrate complex (H-Ras-GTP) and that the logical
extension of this work to the biologically more interesting complex between
H-Ras-GTP and a GTPase activating protein molecule will require substan-
tial technical advances.
The reasons for the relatively slow progress and limited number of
applications should be clear from the discussion presented here. Comparing
the caged substance approach to that of choosing systems that have built
in triggers, it is clear that the technical problems involved are significantly
more difficult to solve in the examples described here. In addition to the
advantage that the stoichiometry of photoactivatable groups with respect
to the protein is normally defined by the nature of the system when the
[ 1 4] UNSTABLE INTERMEDIATES 263
A ¢x
j
B
\
FIG. 2. Stereo view of a representative part of the electron density (shown at a 20, cutoff
level) of a/t-sheet core region in H-ras p21 : GTP (residues 77-81, 110-114, and 139-142)•
For this part of H-ras p21, no significant conformational changes on GTP hydrolysis are found.
Therefore, the quality of the map can be related to the quality° of the data set. (A) Data set
• . o
collected using the Laue method; wavelength range: 0.45-2.6 A; maximum resolution 2.8 A,
completeness: 60%o(2,8-50 .~) and 20% (4•5-10 ~).19 (B) Data set collected using monochro-
matic (h = 1.54 A) diffraction in combination with freeze trapping of the intermediate;
maximum resolution 1.55 A, completeness: 95% (1.55-50 .g,) and 66% (1.55-1.7 ,~).
t r i g g e r is an intrinsic c o m p o n e n t , t h e r e is also u s u a l l y t h e a d v a n t a g e o f
reversibility. This offers t h e p o s s i b i l i t y o f r e p e t i t i o n o f e x p e r i m e n t s in t h e
s a m e crystal o r g e n e r a t i o n o f p s e u d o s t a b l e states, e.g., b y c o n t i n u o u s illumi-
nation. W i t h c a g e d substances, t h e p h o t o l y s i s r e a c t i o n is g e n e r a l l y i r r e v e r s -
ible, as is t h e s u b s e q u e n t r e a c t i o n o c c u r r i n g in t h e p r o t e i n o r o t h e r m a c r o -
molecule, thus excluding repetition. For the generation of steady states in
a crystal, flow cells relying on diffusion will be simpler and preferable to
caged compounds.
Despite these difficulties, it is important to continue to develop technol-
ogy that will allow the caged approach to be applied to other systems
because this will often be the only way of tackling the problem of structural
characterization of unstable reaction intermediates. Because it will not
always be possible to release enough of the active form of the substance
uniformly throughout the crystal at a rate fast enough to allow structural
data to be obtained in an appropriate time window, compromises will often
have to be made. The combination of flash photolysis with cryocrystallo-
graphic methods is a promising development, and one example of this is
described in this article. Further developments will include the use of low
temperatures to slow the reaction of interest to the point where photolysis
can be accomplished by less drastic procedures than single high-intensity
pulses. The use of mutants with slower reaction kinetics can also bring the
time scale into a more comfortable range, as can the use of modified
substrates. However, in both cases evidence that the mechanism is not
changed compared with the reaction of wild-type protein and the natural
substrate or ligand will have to be obtained.
In summary, it seems likely that each specific application of caged
substances to time-resolved crystallography will require extensive charac-
terization of the system involved, its interaction with the caged substances
used (unless the protein itself is caged), and the events occurring in the
crystal on and after hydrolysis. It is clear that this is not achievable for
all macromolecules of interest, but when the questions are of sufficient
importance and are appropriately configured for such an approach, there
is the potential reward of information at a mechanistic level that is not
available from any other single method.
[ 15] CAGEDCHOLINERGICLIGANDS 265
[ 15] P h o t o r e g u l a t i o n o f C h o l i n e s t e r a s e A c t i v i t i e s w i t h
Caged Cholinergic Ligands
By LING PENG a n d MAURICE GOELDNER
Introduction
Acetylcholinesterase (ACHE) and butyrylcholinesterase (BuChE) both
rapidly hydrolyze the neurotransmitter acetylcholineJ The principal role
of AChE is to terminate the neurotransmission at the cholinergic synapses
by rapid hydrolysis of acetylcholine, with a turnover number approaching
20,000 sec-1. 2 Although BuChE plays an important role in breaking down
the muscle relaxant succinylcholine,3 its physiologic function remains un-
known. This enzyme has been so named because it hydrolyzes butyrylcho-
line faster than actylcholine.4 The description of the three-dimensional (3D)
structure of AChE 5 as well as of several AChE-inhibitor complexes,6-9
and of a model of BuChE, 1° based on the 3D structure of ACHE, has
permitted a better understanding of structure-function relationships in the
cholinesterases. It has, however, also raised questions concerning the traffic
of substrate and products to and from the active site in view of the high
turnover rate. Time-resolved Laue crystallography~1would present an ideal
approach to investigate this issue by analyzing, at atomic level and in real
time, the precise structural changes occurring at the enzyme active site
during catalysis. Different from conventional X-ray diffraction techniques,
Laue X-ray crystallography using high-intensity synchrotron polychromatic
1 A. Chatonnet and O. Lockridge, Biochem. J. 260, 625 (1989).

2 D. M. Quinn, Chem. Rev. 87, 955 (1987).
3 F. Hobbiger and A. W. Peck, Br. J. Pharmacol. 37, 258 (1969).
4 K.-B Augustinsson, Methods Biochem. Anal, Suppl. 217 (1971).
5 j. L. Sussman, M. Harel, F. Frolow, C. Oefner, A. Goldman, L. Toker, and I, Silman,
Science 253, 872 (1991).
6 M. Harel, I. Schalk, L. Ehret-Sabatier, F. Bouet, M. Goeldner, C. Hirth, P. H. Axelsen, I.
Silman, and J. L. Sussman, Proc. Natl. Acad. Sci. U.S.A. 90, 9031 (1993).
7 M. Harel, G. J. Kleywegt, R. B. G. Ravelfi, I. Silman, and J. L. Sussman, Structure 3,
1355 (1995).
8 M. Harel, D. M. Quinn, H. K. Nair, I. Silman, and J. L. Sussman, J. Am. Chem. Soc. 118,
2340 (1996).
9 y. Bourne, P. Taylor, and P. Marchot, Cell 83, 503 (1995).
10M, Harel, J. L. Sussman, E. Krejei, S. Bon, P. Chanal, J. Massouli6, and I. Silman, Proc.
Natl. Acad. Sci. U.S.A. 89, 10827 (1992).
11 D. W. J. Cruickshank, J. R. Helliwell, and L. N. Johnson, "Time-Resolved Macromolecular
Crystallography," Oxford University Press, Oxford, England, 1992.

METHODS IN ENZYMOLOGY,VOL. 291 0076-6879/98$25.00
266 CAGZDCO,rOUNDS [ 151
X-ray sources has reduced the exposure time to a microsecond or even
nanosecond time scale. 12 Such studies, however, require a fast and efficient
triggering of the dynamic reaction within the enzyme crystal, and, clearly,
a temporally and spatially controlled photoregulation of enzyme activity
represents the best adapted methodology.
Two photoregulation approaches have been developed. 13 The first
method implies the "caging" of the protein and involves the blockade
of the enzyme activity by modifying a crucial amino acid residue with a
photolabile group allowing the regeneration of the enzyme activity after
photolysis. The caging of the protein can be achieved either by chemical
modification of the target amino acid residue in the enzyme or by protein
biosynthesis using unnatural amino acids, namely the caged amino acid
derivatives. TM Although the synthetic pitfalls encountered during chemical
synthesis may be overcome by the design of enzyme mutants possessing
the appropriate chemical reactivity to permit an efficient and selective
modification of the mutated amino acid residue (see H. Bayley, et al., this
volume15), the biosynthesis of protein might represent a very interesting
but challenging alternative. The second photoregulation approach uses
caged ligands that can act as reversible inhibitors of the enzyme and then
photorelease the desired enzyme substrate or product at the active site,
triggering the enzymatic reaction process under temporal and spatial con-
trol. This article deals with the second caging method and addresses the
question of the photoregulation of ChE activities by means of caged cholin-
ergic ligands and of their potential use in studies on ChE catalysis by time-
resolved crystallography.
Caged Cholinergic Ligands

Scheme 1 shows the different types of caged cholinergic ligands that
have been synthesized. These are 2-nitrobenzyl derivatives a6 of choline
1-3a a7 (type A), carbamylcholine 1-3b 18a,b (type B), and noracetylcholine
lz j. Hajdu and I. Andersson, Annu. Rev. Biophys. Biomol. Struct. 22~ 467 (1993).
13N. A. Porter, J. D. Bruhnke, and P. Koenigs, in "Bioorganic Photochemistry, vol. 2, Biological
Applications of Photochemical Switches" (H. Morrison, ed.) p. 232, Wiley, New York, 1993.
14D. Mendel, J. A. Ellman, and P. G. Schultz, J. Am. Chem. Soc. 113, 2758 (1991).
15H. Bayley, C.-Y. Chang, W. T. Miller, B. Niblack, and P. Pan, Methods Enzymol. 291 [7]
(1998) (this volume).
16For photolabile protecting groups, see elsewhere in this volume.
17L. Peng and M. Goeldner, J. Org. Chem. 61, 185 (1996).
18aj. W. Walker, J. A. McCray, and G. P. Hess, Biochemistry 25, 1799 (1986).
18bT. Milburn, N. Matsubara, A. P. Billington, J. B. Udgaonkar, J. W. Walker, B. K. Carpenter,
W. W. Webb, J. Marque, W. Dank, J. A. McCray, and G. P. Hess, Biochemistry 28, 49 (1989)
[151 CAGEDCHOLINERGICLIGANDS 267
\/ Q
R.~O~N+/ R. H .O~N. /
O2N'~ I - I o m+ ,-' O2N"~ I
A B C
la: R = OH 3 lb: R = CH 3 1¢: R = CH a, R' = CH 3
2a: R = H 2b: R = H 2¢: R=H, R'=CH 3
3a: R = CH2NCS 3b: R = COOH 3¢: R = CH3, R' = CH~CH2CH3
4¢: R = H, R' = CH=CH2CH 3
SCHEME 1. 2-Nitrobenzyl derivatives of choline, carbamylcholine, noracetylcholine, and

norbutyrylcholine as caged cholinergic ligands.
1-2e 19 or norbutyrylcholine 3-4e 2° (type C). The different substitutions at

the a-benzylic position were considered to improve the photochemical
properties (i.e., rapid photofragmentation kinetics and high quantum yield)
as well as to offer chemical variability (e.g., 3a was designed as a potential
irreversible inhibitor by covalent interaction between its isothiocyanate
group and the catalytical serine residue of ChE).
Hydrolysis of acetylcholine by ChEs (Scheme 2) involves at first a rapid
acetylation of the catalytic serine with a concomitant release of choline,
then an immediate deacetylation of the enzyme with an expulsion of the
acetate. 2 As choline is the enzymatic product of ChEs, caged choline deriva-
tives 1-3a were designed to be complexed at the ChE active site and to
generate choline by subsequent photoactivation, allowing a study of the
rapid clearance of choline from the active site (Scheme 3).
Carbamylcholine serves as a slow substrate for ACHE: Rapid carbamyla-
tion with release of choline is followed by much slower decarbamylation
(Scheme 2), which can be accelerated by dilution. 21'22Thus, caged carbamyl-
choline 1-3b can offer the possibility to study the mechanism of carbamyla-
tion of AChE as well as the enzymatic release of choline (Scheme 3).
However, the caged carbamylcholines 1-3b are not suitable for studies on
BuChE because carbamylcholine has a very low affinity for BuChE, and
carbamylation of BuChE is a very slow process that requires, in addition,
a high concentration of carbamylcholine. 22
Noracetylcholine is a close analog of acetylcholine, the endogenous
substrate of ACHE. Hydrolysis of noracetylcholine is chemically identical
19L. Peng, J. Wirz, and M. Goeldner, Angew. Chem. Int. Ed. Engl. 36, 398 (1997).
20 L. Peng, J. Wirz, and M. Goeldner, Tetrahedron Lett. 38, 2961 (1997).
21 I. B. Wilson, M. A. Hatch, and S. Ginsburg, J. Biol. Chem. 235, 2312 (1960); I. B. Wilson,
M. A. Harrison, and S. Ginsburg, J. Biol. Chem. 23t7, 1498 (1961).
22 L. Peng, I. Silman, J. Sussman, and M. Goeldner, Biochemistry 35, 10854 (1996).
HO/~N~I+"- H20 HaCL O.
kl
Enz-OH+ ACh ._ ,, " Enz-OH.ACh Acetyiation AChLOEnz ~ Deacetylation Enz-OH
/cl
SCHEME 2. Mechanism of hydrolysis of acetylcholine (ACh) by cholinesterases,
Compounds of type A
Compounds of type B
__ O2N~~ I
-~.o ~
°'°"=°~- I ~?-!
I
H O'~N ~o~o T
Compounds of type C
-I o2.~ I
".'r~" J hv
•- /
/
~"1
Lo /
I ~ 1
~o.
/H ,
~ /
"I
I
.
SCHEME3. Mode of action of different caged cholinergic ligands for controlling the catalytic
reactions at different steps.
[ 15] CAGEDCHOLINERGICLIGANDS 269
o
/---1
o ,yo./..o" 3. N(OH3)s
NO2
NaBH3CN.TiCI4
95% '~ [~ NO2
94%
• NO2 I I
R = CH3 or H R = CH30r H 1==:R = CH3

2a: R=H
I Mo3SiCN
80 % Znl2
NC'y'O~OTMS 1. BHs'THF R...T..O ~ . ~ O H 21:NCHsSO2C,R " ' ~ ' O ~ N + / / I-

/
[~ NO2 2. CSCI2 ,. [ ~ N O
90%
2 3. 80%
N(OH,), ~ "/NO2 [
R = CH2NCS 3a: R = CH2NCS

SCHEME 4. Synthesis of 2-nitrobenzyl derivatives of choline 1-3a.
to that of acetylcholine.23 After formation of the binary enzyme-substrate

complex, successive acylation and deacylation steps follow, as illlustrated
in Schemes 2 and 3. The advantage of caged noracetylcholine 1-2e over
the previously mentioned compounds 1-3a and 1-3b is that they offer the
possibility of studying the mechanism of the entire catalytic process shown
in Scheme 3. Following the same strategy, caged norbutyrylcholines 3-4e
were designed for the studies on BuChE.
Syntheses
The synthetic routes to compounds 1-3a and 1-4¢ are summarized in
Schemes 4 and 5, respectively; the caged carbamylcholines 1-3b m"'b were
commercially available.
The key step in the synthesis of caged choline 1-3a (Scheme 3) was a
Lewis acid-catalyzed reductive opening of a cyclic acetal or ketal using
sodium cyanoborohydrideJv The syntheses of caged noracetylcholine 1-
21219 o r caged norbutyrylcholine 3-4e 2° (Scheme 4) started with a reductive
amination reaction on 2-nitrobenzaldehyde or 2-nitroacetophenone, and
the final products were obtained as described, in >80% overall yields.
Photochemical Properties
The application of caged compounds for the investigation of rapid ki-
netic processes depends critically on the ability of the photolysis reaction
23 H. K. Nair, J. Seravalli, T. Arbuckle, and D. M. Quinn, Biochemistry 33, 8566 (1994); I. B.

Wilson and E. Cabib, J. Am. Chem. Soc. 78, 202 (1956).
H I
H2N oH c H2o R ~ ~ O H
NO 2 NaBH3CN' ZnCI2,. NO2 NaBHsCN,ZnCI2,, NO2
92 % 99 %
R = CH3 or H R = CH 30r H R = CH 3 or H
I 1. R'COCI
90 % 2. CH31
\ / O
t . / - . o J l - R,
NO2 I
1¢: R= CH3, R' = OH 3

2¢: R= H, R' = CH3
3¢: R= CH3, R' = CH2CH2CH 3
4¢: R= H, R' = CH2CH2CH 3
SCHEME 5. Synthesis of 2-nitrobenzyl derivatives of noracetylcholine 1-2c and of norbuty-
rylcholine 3-4e.
to give the desired products rapidly and with good yield. The kinetics
and quantum yields of the photofragmentation of the caged cholinergic
compounds are summarized in Table I.
The kinetic measurements took advantage of monitoring the decay of
TABLE I
PHOTOFRAGMENTATION PARAMETERS (HALF-TIME ill2 AND QUANTUM YIELD qb) AND
INHIBITION CONSTANTSOF CAGED CHOLINERGICLIGANDSON ChEs
K1 (M) K1 (M)
Compound tv2 (/zsec) q~ AChE BuChE Ref. a
la 10 0.27 1.30 x 10 -5 1.11 × 10 4 17

2a 82,300 0.19 1.00 x 10 5 1.94 × 10 -5 17
3a 560 0.26 3.73 × 10 6 3.13 X 10 -6 17
lb 67 0.25 4.40 × 10 5 7.84 x 10 -5 18, 22
2h 1,700 0.25 NT b NT 18
3b 40 0,80 > 10 -3 > 10 -3 18, 22
le 25 0.10 7.00 x 10 -7 NT 19
2e 24 0.01 4.52 × 10 -6 NT 19
3e 24 0.10 NT 3.44 x 10 4 20
4c 23 0.01 NT 3.30 × 10 4 20
a Numbers refer to text footnotes.
b Not tested.
[ 15] CAGED CHOLINERGIC LIGANDS 271
600 550 500 450 400 350 300

Wavelength (nm)
b
A
E 0.200
C
0
,1¢
m 0.100
o
.Q
Jo
,~ 0.000
i . . . . [ . . . . ~ . . . . , . . . . n . . . . m - - - ' ~ . . . . , . . . . , . . . . ,
-10 0 10 20 30 40 50 60 70 8 0 I~sec
FIG. 1. Kinetic analysis of the laser flash photolysis of compound la. A solution of I mM
compound la in 0.1 M phosphate buffer was exposed to a single 351-nm laser pulse at 20°.
(a) The UV spectrum of a transient was observed by recording the spectral change before
and after (0.2/zsec delay) laser flash photolysis of compound la. (b) Kinetic record at 405
nm after a single laser flash photolysis of compound la at pH 6.5. Arrow indicates the beginning
of the laser flash. The transient was formed immediately and was followed by an exponential
decay. [Reprinted with permission from L. Peng and M. Goeldner, J. Org. Chem. 61, (1996).
Copyright (1996) American Chemical Society.]
the presumed aci-nitro intemediate, which has a specific absorption around

400 nm. The decay of this aci-nitro intermediate was shown to be synchro-
nous to the formation of the corresponding photoproducts. 24 Figure 1 illus-
trates the typical kinetic analysis for the photofragmentation reaction, show-
24 j. W. Walker, G. P. Ried, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. 110,

7170 (1988).
ing the absorption spectrum of the transient and its corresponding decay,a7
Except for compounds 2a, 3a, and 2b, all the other molecules show excellent
kinetic properties for photolysis (Table I), i.e., microsecond photofragmen-
ration time range, which is compatible with the rate of the enzymatic
reaction of cholinesterases (~--50 tzsec). 25
The photoproducts of primary interest are choline, carbamylcholine,
and noracetylcholine (or norbutyrylcholine), respectively. The release of
carbamylcholine from probes 1-3b was already describedlSa,b; the release
of choline and noracetylcholine (norbutyrylcholine) from compounds of
1-3a and 1-4e was assessed and quantified by a specific enzymatic assay
for choline and for noracetylcholine (norbutyrylcholine), respectivelyJ7'19'2°
The amount of the formed photoproduct increased with increasing exposure
to light and coincided with the disappearance of the parent compound, as
quantified by a high-performance liquid chromatography (HPLC) analy-
sis. 17'19'2° These results establish a stoichiometric conversion of cholinergic
ligands from their precursors. Furthermore, the quantum yields for the
photoconversion of caged choline and of caged noracetylcholine or nor-
butyrylcholine were determined by comparison with the photolysis of
1-(2-nitrophenyl)ethylcarbamylcholine (~ = 0.25). TM Although the low
quantum yields of 2e and 4e may limit their further application in the
studies of ChEs, the observed quantum yields for the other compounds
(Table I) are sufficient to ensure an efficient photorelease of the desired li-
gands.
Biochemical Properties
All three types of compounds in Scheme 1 act as reversible inhibitors
of both Torpedo AChE and human serum BuChE (Table I). Compound
3a did not react covalently with the catalytic serine as anticipated, suggesting
an unfavorable positioning of the isothiocyanato moiety within the active
site. Except for 3b, which is a very weak inhibitor of both enzymes (Table
I), inhibition constants of all other compounds are in the micromolar range.
The low affinity of 3b for ChEs is presumably due to the negative charge
of the a-carboxylic group at physiologic pH, which reduces the recognition
of the positively charged quaternary ammonium ion by ChEs. In fact, this
property has been taken advantage of by using 3b as an ideal caged agonist
for time-resolved studies of the nicotinic acetylcholine receptor, 18b,26where
no interaction between 3b and the receptor was desired before photoactiva-
tion. In our case, the ligand must be positioned at the catalytic site of the
25M. Vigny,S. Bon, J. Massouli6,and F. Leterrier,Eur. J. Biochem. 85, 317 (1978).

26G. P. Hess, Biochemistry 32, 989 (1993).
[151 CAGEDCHOLINERGICLIGANDS 273
target protein before light activation. Thus, the reduced affinity of 3b might
limit its further application for time-resolved crystallographic studies on
the ChEs.
Whereas inhibition of AChE by the caged cholinergic ligands (except
3b) is of the mixed type, that of BuChE is purely competitive. Both mixed-
type and competitive inhibition imply that the ligands bind to the anionic
substite of the active site. Acetylcholinesterase has two binding sites for
quaternary ammonium ligands: the anionic subsite of the active site and
the peripheral site) The anionic subsite is adjacent to the catalytic triad,
near the bottom of an aromatic gorge, which is revealed by the 3D structure
of ACHE. 5 The peripheral site is located at the entrance to this gorge. 6,7,9
The binding of compound la at the active site of AChE was further assessed
by ligand displacement experiments using site-specific fluorescent probes:
N-methylacridinium, an active site-specific ligand, 27 and propidium, a
peripheral site-specific ligand. 28The fact that la displaced N-methylacridin-
ium, but not propidium, indicates that la binds specifically at the anionic
subsite of the active site. 22
Although compounds of type C were designed as inhibitors of ChEs,
both 1-2c and 3-4e can be hydrolyzed slowly by AChE and BuChE, respec-
tively. This undesired property will limit their further utilization in the
photoregulation of ChE activity.
Photoregulation of Cholinesterase Activities with Compounds

l a and l b
The ideal caged compounds for photoregulation of the ChEs activities
should meet the following criteria: (1) they should photofragment, rapidly
and with high quantum yields, to the desired molecules; and (2) they should
show sufficient affinity on the enzymes and at the same time should be
chemically stable to the enzyme hydrolysis. Thus, among all the compounds
presented in Scheme 1, la and lb are the most promising candidates for
photoregulating ChE activities. Both la and lb display rapid kinetics of
product release on pulsed illumination, with excellent photofragmentation
half-times (10/zsec for la and 24/zsec for lb, Table I), which are in the
range required for studying both AChE and BuChE, whose turnover rates,
depending on the species, vary in the range of 70-300/~sec. 25 The quantum
yields of both la and lb ensure satisfactory conversion to choline and
carbamylcholine, respectively, and the by-product of photochemical frag-
mentation, 2-nitrosoacetophenone, has no effect on the enzymic activity
27 G. Mooser, H. Schulman, and D. S. Sigman, Biochemistry 11, 1595 (1972).

28 p. Taylor, J. Lwebuga-Mukasa, S. Lappi, and J. Rademacher, Mol. Pharmacol. 10, 703 (1974).
of either AChE or BuChE. Furthermore, both la and lb show inhibitory
effects on ChEs, with inhibition constants in the micromolar range, and
are chemically stable to the enzyme hydrolysis.
We have established the photoregulation of ChE activities using either
probe la, by regenerating the enzymatic activity after photolysis of a
C h E - l a complex (Fig. 2), or probe lb, by inactivating the AChE activity
Residual activity (%)
100,
LU
r-
O
<
40
10 , , • ,
80 160 240 320
L~_~,~_r p u l s e s
Residual activity (%)
wt -
O
10 ', ' ) '-' , ,

0 1000 2000 3000
FIG. 2. Regeneration of enzymatic activity after laser flash photolysis of (a) AChE in the
presence of la and (b) BuChE in the presence of la. Experimental conditions are given in
the Appendix: Experimentation. [Reprinted with permission from L. Peng, I. Silman,
J. Sussman, and M. Goeldner, Biochemistry 35, (1996). Copyright (1996) American Chemi-
cal Society.]
[151 CAOEDCHOLINZR6ICLIGANDS 275
through photoinduced carbamylation after photolysis of an A C h E - l b com-
plex (Fig. 3a). 22
The experiments (see Appendix: Experimentation) in which we studied
the recovery of enzymatic activity after flash laser photolysis of solutions
of AChE or BuChE containing compound la (Fig. 2) suggested that not
AChE
Activity (%)
75 •
50"
&A • • • • • •
25'
0 ! i i
30 60 90 1 2 0 min
Incubation time
b
AChE
Activity (%)
100"
80.
60.
40 ¸
20 i i |
0 6 12 18 2 4 rain
Incubation time
Fie 3. (a) Time-dependent inactivation of AChE after laser flash photolysis of AChE in
the presence of lb. e , one pulse; A, 20 pulses. (b) Regeneration of AChE activity after
subsequent dilution. Experimental conditions are given in the Appendix: Experimentation.
[Reprinted with permission from L. Peng, I. Silman, J. Sussman, and M. Goeldner, Biochemistry
35, (1996). Copyright (1996) American Chemical Society.]
only did the by-product 2-nitrosoacetophenone have no inhibitory effect
on the two enzymes in the experimental conditions employed, but also
both 2-nitrosoacetophenone and choline, which would be generated con-
comitantly within the active sites of the enzymes, were cleared from their
active sites. Thus, the experimental system can indeed serve as a paradigm
for studying the clearance of choline from the active site of AChE and of
BuChE under the conditions of time-resolved crystallography.
In the experiment of photolysis of a solution of AChE containing lb,
even a single pulse caused time-dependent loss of enzymatic activity to
60% of the control value, and 20 pulses decreased the enzyme activity by
a further 15% (Fig. 3a). That this observed inactivation was indeed due to
carbamylation by carbamylcholine photogenerated from lb was demon-
strated by the fact that >90% of control activity could be regenerated in
a progressive fashion on subsequent dilution of carbamylated enzyme (Fig.
3b), just as is observed for AChE carbamylated directly by carbamyl-
choline.22
Conclusion
The results we have obtained show that la and lb possess photochemical
and specificity characteristics that render them suitable to photoregulate
ChE activities. They each release photochemicaUy a different cholinergic
modulator controlling the catalytic reactions at different steps. Compound
la photoreleases choline, the enzymatic product, whereas lb photogener-
ates carbamylcholine, an AChE substrate, allowing the carbamylation of
the enzyme with a concomitant release of choline. Because these two com-
pounds generate choline in two different ways, either by a direct photocleav-
age reaction (la) or by enzymatic hydrolysis of a substrate generated by
photocleavage (lb), they constitute complementary tools for the time-
resolved crystallographic studies envisaged. Work is presently in progess
to establish the dynamic structure of AChE during its catalysis by time-
resolved crystallography using both la and lb separately soaked into crys-
tals of Torpedo ACHE.
Perspective
The understanding of an enzymatic reaction generally involves kinetic
studies in association with chemical methods, such as site-directed labeling
and site-directed mutagenesis, designed to identify the important amino
acid residues involved in the catalytic reaction. In addition, 3D structures
of enzymes and different enzyme complexes have considerably improved
the knowledge of such chemical reactions at the atomic level. However,
[15] CAGEDCHOLINERGICLIGANDS 277
important information is generally missing in such investigations, i.e., the
description of the different interactions occurring along the reaction path-
way, in other words, the dynamic picture of an enzymatic reaction. Time-
resolved macromolecular crystallography offers the possibility of studying
such issues at the atomic level provided that the time resolution is shorter
than the lifetimes of transient key intermediates and that a fast and efficient
triggering of the enzymatic activity in crystal is available.ll Laue crystallog-
raphy using a synchrotron X-ray source allows extremely rapid data collec-
tion, on the nanosecond time scale, and offers a direct experimental ap-
proach that may permit visualization of the putative conformational changes
of the enzyme.29Photoregulation of enzyme activity with caged compounds
can provide rapid photochemical release of enzyme product or enzyme
substrate, and thus initiate the enzymatic reaction synchronously in enzyme
crystals. Alternative methodologies, artificially prolonging the key interme-
diates' lifetimes, can be envisaged either by biological methods, i.e., produc-
tion of slower mutants, 3° or by physical methods, i.e., a freeze-trapping
approach. 31 The advantage of using caged compounds in time-resolved
Laue crystallographic studies is that we could analyze directly the authentic,
short-lived intermediates along an enzymatic reaction pathway, and thus
we could observe the true dynamic picture of an enzyme while achieving
its biological function.
Appendix: Experimentation
Materials
Acetylcholinesterase was purified from electric organ tissue of Torpedo
marmorata 32 and BuChE from human plasma. 33 Compound la was synthe-
sized as described, 17 and O-[1-(2-nitrophenyl)ethyl]carbamylcholine iodide
(lb) was purchased from Molecular Probes (Junction City, OR).
Photoregulation of Cholinesterases Activities with Compounds l a and l b

Laser pulses of about 100 mJ and of about 20 nsec duration 17are applied
to a quartz cuvette (1 × 10 x 40 mm) containing 2/zg/ml ACHE, together
29 V. Srajer, T. Teng, T. Ursby, C. Pradervand, Z. Ren, S. Adachi, W. Schildkampf, D.

Bourgeois, M. Wulff, and K. Moffat, Science 274, 1726 (1996).
30j. M. Bolduc, D. H. Dyer, W. G. Scott, P. Singer, R. M. Sweet, D. E. Koshland, and B. L.
Stoddard, Science 268, 1312 (1995).
31 K. Moffat and R. Henderson, Curr. Biol. 5, 656 (1995).
32 L. Ehret-Sabatier, I. Schalk, M. Goeldner, and C. Hirth, Eur. Z Biochem. 203, 475 (1992).
33 O. Lockridge and B. N. La Du, J. BioL Chem. 253, 361 (1978).
with 5.0 x 10 -4 M probe la in a total volume of 0.4 ml of 50 mM phosphate
buffer, pH 6.5. Enzymatic activity is monitored spectrophotometrically, at
400 nm, using 1 mM p-nitrophenyl acetate as substrate, in 50 mM phosphate
buffer, pH 7.2, 200.34 Butyrylcholinesterase (3 /zg/ml) is incubated in a
quartz cuvette (5 x 10 x 40 mm) with 5.4 × 10 -5 M probe la in 50 mM
phosphate buffer, pH 7.2, in a total volume of 1.2 ml and similarly exposed
to 351-nm laser flashes. Butyrylcholinesterase activity was monitored spec-
trophotometrically at 240 nm using 5.0 x 10 -s M benzoylcholine as sub-
strate, in 50 mM phosphate buffer, pH 7.2, at 200.35
Acetylcholinesterase (294/zg/ml) is incubated with 4.2 X 10 .3 M probe
lb, in 50 mM phosphate buffer, pH 6.5, total volume 0.4 ml, in a quartz
cuvette (1 x 10 X 40 mm). After exposure to either 1 or 20 pulses of a
351-nm laser, aliquots are withdrawn for assay of enzymatic activity both
by the Ellman method 36 and by using p-nitrophenyl acetate as substrateY
Regeneration of the enzymatic activity of carbamylated AChE was achieved
by 500-fold dilution into 50 mM phosphate buffer, pH 7.2, at 20 °.
Acknowledgments
This work was supported by the Association Fran~aise contre les Myopathies, the Centre
National de la Recherche Scientifique, the Association Franco Isra61ienne pour la Recherche
Scientifique et Technique, the Soci6t6 de Secours des Amis des Sciences, and the European
Community Biotechnology Programme under Contract 960081.
34 R. K. Tripathi, J. N. Telford, and R. D. O'Brien, Biochim. Biophys. Acta 525, 103 (1978).
35 W. Kalow and H. A. Lindsay, Can. J. Biochem. Physiol. 33, 568 (1953).
s6 G. L. Ellman, K, D. Courtney, V. Andres, and M. R. Featherstone, Biochem. Pharrnacol.
7, 88 (1961).
[I 6] C a g e d S u b s t r a t e s f o r M e a s u r i n g E n z y m a t i c A c t i v i t y
in Vivo: P h o t o a c t i v a t e d C a g e d G l u c o s e 6 - P h o s p h a t e
By ROBERT R . SWEZEY a n d DAVID EPEL
This article describes the use of caged substrate molecules to study the
activity and regulation of enzymes in vivo. As stated by Van Noorden
and Jonges, 1 "Estimations of metabolic rates in cells and tissues and their
regulation on the basis of kinetic properties of enzymes in diluted solutions
1 C. J. F. Van Noorden and G. N. Jonges, Histochem. J. 27, 101 (1995).

All rightsof reproduction in any formreserved.
METHODSIN ENZYMOLOGY,VOL.291 0076-6879198 $25.00
with 5.0 x 10 -4 M probe la in a total volume of 0.4 ml of 50 mM phosphate
buffer, pH 6.5. Enzymatic activity is monitored spectrophotometrically, at
400 nm, using 1 mM p-nitrophenyl acetate as substrate, in 50 mM phosphate
buffer, pH 7.2, 200.34 Butyrylcholinesterase (3 /zg/ml) is incubated in a
quartz cuvette (5 x 10 x 40 mm) with 5.4 × 10 -5 M probe la in 50 mM
phosphate buffer, pH 7.2, in a total volume of 1.2 ml and similarly exposed
to 351-nm laser flashes. Butyrylcholinesterase activity was monitored spec-
trophotometrically at 240 nm using 5.0 x 10 -s M benzoylcholine as sub-
strate, in 50 mM phosphate buffer, pH 7.2, at 200.35
Acetylcholinesterase (294/zg/ml) is incubated with 4.2 X 10 .3 M probe
lb, in 50 mM phosphate buffer, pH 6.5, total volume 0.4 ml, in a quartz
cuvette (1 x 10 X 40 mm). After exposure to either 1 or 20 pulses of a
351-nm laser, aliquots are withdrawn for assay of enzymatic activity both
by the Ellman method 36 and by using p-nitrophenyl acetate as substrateY
Regeneration of the enzymatic activity of carbamylated AChE was achieved
by 500-fold dilution into 50 mM phosphate buffer, pH 7.2, at 20 °.
Acknowledgments
This work was supported by the Association Fran~aise contre les Myopathies, the Centre
National de la Recherche Scientifique, the Association Franco Isra61ienne pour la Recherche
Scientifique et Technique, the Soci6t6 de Secours des Amis des Sciences, and the European
Community Biotechnology Programme under Contract 960081.
34 R. K. Tripathi, J. N. Telford, and R. D. O'Brien, Biochim. Biophys. Acta 525, 103 (1978).
35 W. Kalow and H. A. Lindsay, Can. J. Biochem. Physiol. 33, 568 (1953).
s6 G. L. Ellman, K, D. Courtney, V. Andres, and M. R. Featherstone, Biochem. Pharrnacol.
7, 88 (1961).
[I 6] C a g e d S u b s t r a t e s f o r M e a s u r i n g E n z y m a t i c A c t i v i t y
in Vivo: P h o t o a c t i v a t e d C a g e d G l u c o s e 6 - P h o s p h a t e
By ROBERT R . SWEZEY a n d DAVID EPEL
This article describes the use of caged substrate molecules to study the
activity and regulation of enzymes in vivo. As stated by Van Noorden
and Jonges, 1 "Estimations of metabolic rates in cells and tissues and their
regulation on the basis of kinetic properties of enzymes in diluted solutions
1 C. J. F. Van Noorden and G. N. Jonges, Histochem. J. 27, 101 (1995).

All rightsof reproduction in any formreserved.
METHODSIN ENZYMOLOGY,VOL.291 0076-6879198 $25.00
[16] PHOTOACTIVATED CAGED GLUCOSE 6-PHOSPHATE 279
may not be applicable to intact living cells or tissues." It is therefore
preferable to study the activity of an enzyme while it resides in its physiologi-
cal address, i.e., intracellularly.
Such endeavors to study enzyme activities in vivo have predominately
used one of the following approaches: (a) Identification of enzymes at
"crossover" points in a metabolic pathway (example in Ref. 2): Here cellular
concentrations of metabolites are measured in an attempt to identify which
substrate-product pairs in the pathway are not in thermodynamic equilib-
rium. This would indicate that the enzyme(s) involved in this catalysis is
inhibited to some degree within the cell. (b) Quantitative histochemistryl: In
this approach an exogenous substance that is chromogenic via an enzymatic
activity (e.g., formazans with dehydrogenases) is administered to cells, and
the rate of product formation in individual cells is monitored microscopi-
cally (e..g, with a scanning and integrating cytophotometer). (c) Radioactive
precursors3: By administering a radiolabeled precursor for a metabolic
pathway (e.g., [3zP]Pi for glycolysis) and then measuring the changes with
time in the specific radioactivities of all of the intermediates in a metabolic
pathway, one can determine the fluxes of substrates through each enzyme
of the pathway. This approach is precise, but it is particularly labor intensive.
In our strategy, an unreactive caged and radiolabeled substrate is first
placed in a cell, and then initiation of the enzyme assay in vivo occurs by
photolytic release of the substrate. A key advantage of this approach is
that it permits a temporally defined radiolabeling of a particular intracellular
substrate pool; the rate of conversion of this substrate to product within
the cells is then followed by chemical, radiochemical, or spectroscopic
means, and knowledge of the intracellular substrate concentration permits
calculation of enzymatic activity in vivo. This approach has also been used
by Meldrum et al. 4 to follow the rate of DNA repair in cells that have been
exposed to ultraviolet (UV) irradiation damage.
This article describes the preparation and use of caged glucose 6-phos-
phate (G6P) to study changes in the glycolytic and pentose shunt pathways.
We have used this to study the activity of glucose-6-phosphate dehydroge-
nase (G6PDH) in sea urchin eggs and the changes in activity of this enzyme
that attend the fertilization of these eggs. We describe in detail the proce-
dure for sea urchin eggs, realizing that although our protocols are highly
specific for these eggs, the principles can be transferred to many other
cell types.
2 I. Yasumasu, K. Asami, R. L. Shoger, and A. Fujiwara, Exp. Cell Res. 80, 361 (1973).
3 j. G. Reich, U. Till, J. Gunther, D. Zahn, M. Tschisgale, and H. Frunder, Eur. J. Biochem.
6, 384 (1968).
4 R. A. Meldrum, S. Shall, and C. W. Wharton, Biochem. J. 266, 891 (1990),
1E+07
8E+06- "5
6E+06- 0
4E÷06-
2E+O6-
g
D.
'L
0E+00
o
,¢
Fraction N u m b e r
FIG. 1. High-performance liquid chromatography anion-exchange purification of radio-

labeled glucose 6-phosphate. Fractions 70-90 were pooled, evaporated, and reacted with
diazoalkane caging reagent (see text for details).
Preparation of Caged [3H/4C]Glucose 6-Phosphate

The chemical properties of the caging reagents and final caged products
are covered more extensively in other articles in this volume. What follows
are some practical pointers for those who do not have access to a synthetic
organic chemistry laboratory.
Preparation of [5-3H/lJ4C]Glucose 6-Phosphate. One millicurie of [5-
3H]glucose and 0.25 mCi of [1-14C]glucose are combined and the solvents
removed in vacuo. The residue is dissolved in 200 /zl of a hexokinase
reaction medium consisting of 50 mM HEPES buffer (pH 7.5), 20 mM
adenosine triphosphate, 25 mM MgClz, and 20 units of commercial hexoki-
nase from bakers yeast. This solution is incubated for 5 h at room tempera-
ture, and then the radioactive glucose 6-phosphate is purified by an anion-
exchange high-performance liquid chromatography (HPLC) procedure that
has been described elsewhereJ A representative chromatogram for one
such synthesis is shown in Fig. 1. Fractions 70-90 were pooled for subse-
quent use. Using this procedure, approximately 15% of the radioactivity in
the chromatogram is associated with G6P; the other major chromatographic
peaks are most likely unreacted glucose (fraction 5) and the hexose diphos-
phates (fractions 140 and 144).
5R. R. Swezey,J Chromatogr.B 669, 171 (1995).

Preparation of2-Nitroacetophenone Hydrazone. In a round-bottom flask

combine 2.065 g 2-nitroacetophenone (12.5 mmol), 25 m195% ethanol (v/v),
1.36 ml hydrazine hydrate, and 0.8 ml glacial acetic acid. Add boiling chips,
attach a water-jacketed condenser to the flask, and reflux the mixture for
3 h using an electric heating mantle. Cool the solution to room temperature,
and then remove the solvent under vacuum. Add 50 ml each of CHC13and
water to the residue, and then transfer to a separatory funnel. Mix immisci-
ble phases well, allow to separate, and then retain the lower (organic)
phase. Discard the aqueous phase, and wash the organic phase twice more
with water in the separatory funnel. Evaporate the washed organic phase
under vacuum to remove the CHCI3.
The hydrazone is then purified chromatographically with standard silica
gel, using ethyl acetate as the eluent as follows. Dissolve the CHCI3residue
in a minimal volume of ethyl acetate (~5 ml) and chromatograph this on
a column of silica gel (1.5 × 18 in.) equilibrated with and eluted by ethyl
acetate. Collect the deep-yellow fraction that elutes first and remove the
ethyl acetate under vacuum to obtain the final product, which is a viscous
yellow oil. We typically recover 1.2 g of product (6.7 mmol), giving an
overall yield of ~54%.
Conversion of Hydrazone to Diazo Caging Reagent and Reaction with
[52H/1J4C]Glucose 6-Phosphate. Perform the following in subdued light-
ing (darkness or red lighting is not necessary). Transfer 65/zl of the hydra-
zone oil (0.4 mmol) to a glass test tube (16 × 100 mm), and add 5 ml
diethyl ether and a 1-cm round, cross-headed stirring bar (e.g., Aldrich,
Milwaukee, WI). With vigorous stirring carefully add 0.285 g MnO2. Stop-
per the tube with a rubber septum impaled with a hypodermic needle to
permit venting, cover the reaction tube with aluminum foil, and continue
stirring 5 min. Filter the now deep orange-red ether solution through
Whatman (Clifton, N J) #1 filter paper under vacuum, releasing the vacuum
as soon as filtration is complete to prevent drying down the solution.
We keep a 2-liter beaker containing approximately 800 ml of 20% (v/v)
acetic acid in methanol nearby for immediate immersion of glassware,
pipettes, filter paper, etc., that has come into contact with the ether solution
of diazoalkane, and which is no longer needed in the synthesis. The acetic
acid effectively quenches residual diazoalkane on these surfaces, thereby
eliminating the formation of dry diazoalkane, which is potentially explosive.
Three milliliters of this diazoalkane ether solution are then transferred
to a fresh test tube containing 1 ml of an aqueous solution of [5-3H/1 -
14C]glucose 6-phosphate at pH - 4 - 5 (pH paper is adequate for this adjust-
ment) prepared as described earlier. This biphasic mixture is vigorously
mixed for at least 3 hr at room temperature using a 1-cm round cross-
headed stirring bar. The upper (ether) phase is then removed and replaced
$O00-
Ca~
~P
4000-
3000-
2000 -
G6P
lOOO- -.,m---Im--
Fraction Number
FIG.2, Reversed-phaseHPLCpurificationof cagedglucose6-phosphate.Fractions25-28

were pooled, evaporated, and dissolvedin the permeabilizationmediumfor in vivo assays
(see text for details).
with freshly prepared diazoalkane ether solution, and this is then stirred
for at least 3 hr. After one more application of flesh diazoalkane ether
solution and 3 hr stirring, the aqueous phase is removed and washed three
times with 5 ml diethyl ether and then evaporated to dryness in vacuo. The
residue is dissolved in a minimal volume of 50 mM ammonium bicarbonate
(no pH adjustment), and the caged compound is purified by reversed-phase
HPLC as has been described elsewhere. 6 As seen in Fig. 2, the unreacted
G6P comes in the flowthrough fractions, whereas the more hydrophobic
caged substrate elutes much later. Fractions 25-28 containing the caged
G6P are pooled and taken to dryness in vacuo.
Testing Caged Substrates
After the caged substrate has been synthesized and purified it must
be tested for inertness toward G6PDH and for conversion to the parent
compound with resultant competence to be oxidized on photolysis. To do
this a small portion of the caged compound is dissolved in a reaction
medium for G6PDH and 6PGDH (6-phosphogluconic dehydrogenase), and
the ability to form 14CO2 before and after irradiation is tested. This medium
contains 50 mM glycylglycine, pH 8.4, 5 mM MgC12, and 1 mM/3-nicotin-
amide adenine dinucleotide phosphate (NADP). The reaction takes place
in a small Warburg flask with 100% (w:v) trichloroacetic acid in the side
arm and 0.5 M NaOH in the center well. The reaction is initiated by adding
a G6PDH/6PGDH enzyme mixture that is prepared as follows: 100/zl of
a commercially obtained ammonium sulfate suspension of 6-phosphogluco-
6R, R. Swezeyand D. Epel, Exp. Cell Res. 2AIl~366 (1992).

[16] PHOTOACTIVATED
CAGED GLUCOSE6-PHOSPHATE 283
nate dehydrogenase is centrifuged for 2 rain at 12,000 g (at room tempera-

ture), the supernatant is removed by aspiration, and the resulting pellet is
dissolved in 100/zl of a 1-mg/ml solution of G6PDH. Ten microliters of
the G6PDH/6PGDH enzyme mixture is then transferred to each Warburg
flask; reaction with these two enzymes results in decarboxylation of the
14C at the 1 carbon of G6P to form 14CO2, which becomes trapped in the
K O H in the center well.
We find that after 3 hr incubation at room temperature only 1.2% of
the total 14C label of the material not subjected to light flashes is present
in the center cell, whereas 66.1% of the label is in the center well (i.e., is
converted to t4C02) in the sample that had received two flashes of light
from the xenon lamp (described in the Assay of Enzyme Activities in
Vivo section). Therefore, this caged substrate demonstrates the desired
properties of being inert prior to photolysis and enzymatically reactive af-
terward.
When we subjected the photolysis product to chromatographic analysis
we made a surprising discovery. As seen in Fig. 3, the photolysis of caged
G6P results in two products, G6P and a chemically unidentified side product
labeled X. We found that X was capable of reacting with 6PGDH to form
14C02; however, the retention time of X in this HPLC procedure was
different from that of 6-phosphogluconate. 5 The relevance of X and its
reactivity with 6PGDH is discussed in the results section.
Placing Caged Substrate into the Cell." Electroporation

Once the caged substrate has been purified and demonstrated to be
enzymatically inactive until photolysis, the next step is to place the caged
G6P
15000-
10000-
5000-
IA_^
i i i i i i |
@
Fraction Number
Fie. 3. High-performance liquid chromatography analysis of the photolysis product of
caged glucose 6-phosphate. Chromatographic conditions identical to those in Fig. 1.
substrate into the cytoplasm. In our work with sea urchin eggs we have
found that permeabilization of the cell membrane by mild electroporation
is an effective means to introduce caged substrates (or other impermeant
substrates) into the egg cytoplasm. The procedure for sea urchin eggs has
been described in detail elsewhere, 7 but a brief description is given here.
It should be noted that this procedure differs from most electroporation
protocols in that the holes generated by the electroporation procedure
remain open following its cessation, allowing permeation of substrates until
the pores are resealed by addition of calcium ions to the media.
The medium surrounding the eggs during electroporation is crucial for
successful electroporation of these cells. We have obtained good results
utilizing a medium with the following composition: 300 mM glycine, 225 mM
potassium gluconate, 185 mM mannitol, 20 mM NaC1, 10 mM spermidine, 5
mM MgC12,4 mM ATP, 2 mM sodium bicarbonate, and 1.6 mM reduced
glutathione; the pH of this medium is adjusted with HCI to 6.8 just prior to
use in permeabilization of eggs. The composition of this medium, developed
after much trial and error, reflects the natural osmolarity of the sea urchin
egg, i.e., seawater, and other characteristics of this cell of a marine organism
such as high K +, low Na +, low CI-, and a high concentration of an osmolyte
such as glycine. The key features of this medium that may be transferable
to other cell types include the high K ÷, low Na + and CI-, and inclusion of
MgATP at physiological levels. It is also crucial to keep the medium low
in Ca 2+, in that Ca 2+ ions impede electroporation, and as noted earlier are
used later to reseal the egg cell membrane at the end of the electroporation
loading process. (Inclusion of EGTA (ethylene glycol-bis(/3 amino ethyl
ether)N,N,N',N'-tetraacetic acid) or other chelating agents is not recom-
mended because chelators will enter the cell and could impede any Ca +-
dependent processes.)
A typical loading of sea urchin (species Strongylocentrotuspurpuratus)
eggs with caged (3H/14C) G6P proceeds as follows. To a tube containing
the dry, purified caged G6P is added 400 /xl permeabilization medium.
Dissolution is accomplished by gentle repetitive pipetting, and the solution
is then clarified by centrifugation for 3 min at approximately 12,000 g
(at room temperature). Three hundred microliters of the supernatant are
transferred to a test tube containing 200 tzl of packed sea urchin eggs
that have been equilibrated with permeabilization medium by cycles of
suspension/centrifugation, first in calcium-free seawater and then in per-
meabilization media. This suspension is transferred to the electroporation
7 R. R. Swezey and D. Epel, Cell Regul. L 65 (1989).

chamber (described in Swezey and Epel s) where the eggs are exposed five
times at 10-sec intervals to an electric field of 150 V/cm by capacitor
discharge. Using a 0.5-/zF capacitor, the resulting electric field decays by
1/e every 70/xsec in our system. The cells are then transferred to a test
tube and incubated at 16° for 2 rain with repeated gentle bubbling from a
Pasteur pipette. The pores remain open, and this time period allows diffu-
sion of the caged substrate into the permeabilized cell. After this incubation
the eggs are placed in 10 ml of an artificial seawater (MBL formula) that
has been modified to contain 1.1 mM Ca 2÷ ions (instead of the 11 mM
normally present). The cells are allowed to settle in this low-calcium seawa-
ter, during which time the electrically generated pores reseal. The eggs are
then transferred back to seawater for experimentation.
Assay of Enzyme Activities in Vivo

Once the cells are loaded with the caged substrate and placed back into
an appropriate medium, the assay of enzyme activity within the cells can
proceed. In our work, a 1.8-ml suspension of substrate-loaded sea urchin
eggs is placed into the main compartment of a small Warburg flask con-
taining 400/zl 0.5 M K O H in the center well and 200/z1100% trichloroacetic
acid (TCA) in the side arms. For photolysis, the Warburg flasks are held
in the center of a torroidal xenon flash lamp (Speedotron Corp., Chicago,
IL, 12 W-sec per flash, commercially available for photographic use). Alumi-
num foil placed over this unit maximizes light exposure to the cells and
minimizes exposure to the experimenter.
Uncaging is accomplished by igniting the flashbulb twice, with 4 sec
between flashes (the minimal recharge time). In our work with sea urchin
eggs, this flashing is performed in three cases: (a) in unfertilized eggs, (b)
in eggs that had been fertilized in the Warburg flask and allowed to develop
2 min prior to flashing, and (c) in eggs that had been fertilized in the
Warburg flask and allowed to develop 20 min prior to flashing. In separate
flasks, the reactions are stopped at various times postflashing by tipping
the flasks to mix the TCA with the sea urchin egg suspensions (final concen-
tration of TCA in the mixture is 10%). These flasks are then shaken on an
orbital shaker for 30 min to allow quantitative accumulation of the 14CO2
into the K O H solution in the center well; the amount of 14CO2formed is
then quantitated by liquid scintillation counting of the contents of the
center well.
8 R. R. Swezey and D. Epel, in "Guide to Electroporation and Electrofusion" (D. C. Chang,

B. M. Chassy, J. A. Saunders, and A. E. Sowers, eds.), p. 347. Academic Press, San
Diego, 1992.
To measure the 3H released from the [3H]glucose 6-phosphate, the
acidified egg suspensions are transferred from the flasks to centrifuge tubes,
and the eggs are removed by centrifugation at approximately 500-1000 g
for 2 min (at room temperature). The bulk of the TCA in the resulting
supernatants (volume approximately 1.2 ml) is removed by extraction three
times with 5 ml of diethyl ether, and small aliquots (10 to 25/xl) of a 1 M
solution of ammonium bicarbonate are added until the pH is approximately
neutral as evidenced with pH paper. All extracts are brought to the same
volume (2 ml) by adding water, and then 0.6 ml are distilled until at least
0.2 ml is collected in the distillate (the distillations are never taken to
dryness to avoid pyrolytic generation of 3H20 in the distillation process).
Two hundred microliters of the distillates are then counted for 3H, which
is a measure the amount of 3H20 formed in the eggs from the [5-3H]G6P.
Therefore, in our studies the relevant rates in the cells being monitored
are the rates of formation of 14CO2 and 3H20.
Results
Our primary interest was to determine the activity of the enzyme
G6PDH in eggs before and after fertilization. We already had evidence
that the in vivo activity of the decarboxylating enzyme of the pentose
shunt, 6PGDH, was so high as not to be rate limiting to the formation
of 14CO2 from the uncaged G6P by the shunt. 6 However, because glycolysis
and Krebs cycle activities also form 14CO2 from [1-14C]G6P, we needed
an independent assay of flux through these pathways. Loss of 3H from
the [5-3H]G6P to form 3HzO at the enolase and/or triose-phosphate
isomerase steps of glycolysis provides this information. Thus in these
double-label experiments we could monitor flux rates through glycolysis/
Krebs cycle enzymes and through the pentose shunt (=rate of G6PDH ac-
tivity).
Figure 4 (data replotted from Swezey and Epel 9) depicts the results for
a typical experiment; here we have plotted the amounts of 14CO2 (Fig. 4A)
and 3H20 (Fig. 4B), expressed as the percentage of the total aac or 3H in
each of the samples to normalize between flasks, formed as a function of
time postuncaging for the three developmental states of the eggs. The
relevant points obtained from experiments such as these are that 2 min
after fertilization there is a large increase in the rate of ~4CO2 generation,
with very little change in the rate of 3H20 formation, and that by 20 rain
after fertilization this rate in 14CO2 formation subsides, but there is still
little change in the rate of 3H20 formation. The changes in ~4CO2 formation
9 R. R. Swezey and D. Epel, Dev. Biol. 169, 733 (1995).

A
30
ff 25.
2 rain
20.
15- ...A .... 20 rain

K" _..--'O
lo.. 0 "'0 . . . . o
[] 0 Unf
5-
Time postflashing (sec)
B
O
15- 20 mln
tv)
Unf
I0~ . . . . . . . . . . . .---
..---A 2 min
O _&... ° - - -
0 , , ~ °
Time postflashing (sec)

FIo. 4. Assay of G6P utilization in vivo in S. purpuratus eggs. (A) Formation of 1 4 C O 2
from caged [1-14C]glucose 6-phosphate after two light flashes. (B) Formation of 3H20 from
caged [5-3H]glucose 6-phosphate after two light flashes.
are therefore indicative of changes in G6PDH activity, and not in glycolytic/

Krebs cycle activities (because 3H20 formation rates are relatively con-
stant). These findings indicate that there is very little G6PDH activity
operating in unfertilized eggs and that there is a pronounced activation of
G6PDH shortly after fertilization. At a later developmental time (20 min)
the activity of G6PDH is less.
Note that for 14CO2 data that there is an apparent burst of 14CO2
between time 0 (which represents an unflashed sample) and 5 sec after
uncaging, followed by a slower rate of formation between 5 and 30 sec
after flashing; as seen in Fig. 4B, there is no such burst in 3H20 formation.
The burst-phase 14COz formation has been attributed to the rapid oxidation
of compound X by 6PGDH, 9 and therefore the rates of [1-14C]G6P oxida-
tion are taken from data subsequent to 5 sec postflashing. This shows that
characterization of the photolysis products can be critical in the interpreta-
tion of data from in vivo experiments.
By knowing the concentration of intracellular G6P in eggs, 24 fmol/

egg, a value that is fairly constant before and after fertilization of S. purpu-
ratus eggs, l° and knowing the efficiency of uncaging in vivo (34 +- 3.4%,
Ref. 9), it is possible to calculate maximal estimates of G6PDH activity (see
Refs. 9 and 11 for details). These maximum values of G6PDH (expressed as
10-8 units per eggs) are 0.0024 in unfertilized eggs, 1.10 at 2 min after
fertilization, and 0.53 at 20 min after fertilization,u Relative to unfertilized
eggs, G6PDH activity is 458 times greater than 2 rain postfertilization and
221 times greater at 20 min.
In vitro estimates of maximal G6PDH activity in these eggs assayed at
physiological pH and substrate concentrations indicated that there are
177 × 10-8 units per egg and that this in vitro value did not change on
fertilization.12Therefore, even at the maximal rate observed in vivo (2 rain)
the enzyme exhibited only 0.62% of that seen in vitro. This fact underscores
the quotation by Van Noorden and Jonges given at the beginning of this ar-
ticle.
Conclusions
Caged radiolabeled G6P placed into unfertilized sea urchin eggs and
then uncaged by photolysis before or after fertilization showed that fertiliza-
tion results in a pronounced increase in G6PDH activity, whereas glycolytic/
Krebs cycle activity was constant during this same period.
The basic strategy described here could be adapted for use with other
enzymatic substrates and with other cell types. One limitation is that this
approach is best suited to cytoplasmic enzymes rather than those in other
intracellular compartments, as the cytoplasmic pool is accessible to the
electroporation protocol; analysis of compartmentalized enzymes could be
complicated by the rate of the movement of the uncaged substrate from
the cytoplasm to the relevant compartment.
This procedure could be useful for many other cell types, and we look
forward to similar in vivo measurements of enzyme activity in other cells.
It will be interesting to learn how well (or how poorly) the estimates from
in vitro measurements compare to in vivo rates.
10D. Epel and R. M. Iverson in "Control of Energy Metabolism" (B. Chance, R. W. Estabrook,
and J. R. Williamson, eds.), p. 267. Academic Press, New York, 1965.
11 B. B. Rees, R. R. Swezey, H. Kibak, and D. Epel, Invertebr. Reprod. Dev. 30, 123 (1996).
12 R. R. Swezey and D. Epel, Proc. Natl. Acad. Sci. U.S.A. 85, 812 (1988).
[I 71 INVESTIGATION
OF CHARGETRANSLOCATION 289
[ 17] Investigation of Charge Translocation by Ion Pumps

and Carriers Using Caged Substrates
By K. FENDLER,K. HARTUNG,G. NAGEL,and E. BAMBERG
Introduction
Ion pumps create gradients of ions across cell or cell organelle mem-
branes, while consuming light or chemical energy, e.g., by hydrolyzing ATP.
Subsequently, the electrochemical gradients are used for passive transport
via ion channels or carriers. The combined action of these membrane
proteins serves to build up membrane potential, to regulate the intracellular
pH, or to accumulate nutrients such as sugars or amino acids inside the cell.
Ion pumps and carriers pass through an enzymatic cycle during the
transport process. The study of the charge translocation by ion translocating
proteins is one of the most direct methods to study the mechanism of these
enzymes. However, only in a few cases has it been possible to follow the
ion translocation in the native membrane, as the turnover of pumps and
carriers is low (between 1-1000 see-l). Although electrical properties of
single channels can be measured very precisely because of the large turnover
of about 107 sec -I, which corresponds to a current of 10 -12 A per channel,
the study of ion pumps requires the activation of about 104-107 molecules
to produce a current equivalent to that of a single open channel.
Therefore, a high density of the protein within the membrane is a
prerequisite to studying its ion transport. This condition can be met by
using preparations that have a naturally high density of the transport system
of interest or by purifying the protein. In addition, large membrane areas
are required to produce a measurable pump current. This is possible by
protein reconstitution on planar lipid bilayers or by the application of the
so-called giant patch clamp technique. I'2 Another technical difficulty is that
many of the investigated enzymes occur in cell organeIIes or bacteria, where
electrophysiological methods are not applicable because of the small size.
In these cases ion pumps and carriers can be reconstituted on planar
lipip membranes. 3
Because of the low basic conductance of the planar lipid membranes,
a high signal-to-noise (S/N) ratio can be expected for the registration of
the charge translocation. The introduction of photolabile substrates (caged
1 D. W. Hilgemann, Pflagers Arch. 415, 247 (1989).

2 T. Friedrich, E. Bamberg, and G. Nagel, Biophys. J. 71, 2486 (1996).
3 E. Bamberg, H.-J. Butt, A. Eisenrauch, and K. Fendler, Quart. Rev. Biophys. 26, 1 (1993).

All rightsof reproduction in any form reserved.
compounds) for ion carriers and ion-motive ATPases signifies an important

advance in the study of transport properties of these enzymes. 4 Because
the photoreaction does not produce mechanical disturbances of the system,
the combination of caged compounds with electrical or electrophysiological
methods on artificial or natural membranes is ideal for the studies of trans-
port proteins with low turnover numbers. For kinetic investigations the
application of caged compounds is advantageous because in absence of an
active substrate, such as ATP, an ATPase is locked in a defined intermedi-
ate. This represents the ideal situation for the study of presteady-state
transport.
Light-induced concentration jumps of ATP using caged ATP have been
applied successfully on ion-motive ATPases to initiate the reaction and
transport cycle for both cell membranes and planar lipid membranes. Sub-
stantial information about electrogenicity and kinetics has been obtained
for the Na,K-ATPase, for the Ca-ATPase from sarcoplasmic reticulum,
and for the H,K-ATPase from pig stomach (for references, see Table I).
The latter is an electroneutrally operating 1 : 1 exchanging pump. Despite
that, the application of caged ATP showed clearly that the H ÷ translocating
step is electrogenic, 5,6 as is the K ÷ transporting step. 7 More recently, caged
Ca 2÷ and caged ATP or ADP were applied to secondary active transporting
systems such as the Na,Ca-exchanger of heart cells and the A D P / A T P -
carrier from mitochondria. The following sections give some examples for
the application of caged ATP, caged ADP, and caged Ca 2÷. Also, problems
associated with the inhibiting effect of the caged nucleotide on the Na,K-
ATPase are demonstrated. This article presents studies on the application
of caged compounds to planar lipid membranes as well as on giant patches
of heart cell membranes.
Na,K-ATPase
Na,K-ATPase plays an important role in every animal cell. Using the
free energy of ATP hydrolysis, it transports three Na + ions out of the cell
in exchange for two K +, The Na,K-ATPase thereby maintains the ion
gradients across the cell membrane that are important for various functions
of the cell. The generally accepted reaction mechanism assumes two confor-
mations of the unphosphorylated (El and E2) and the phosphorylated
4 j. H. Kaplan, B. I. I. I. Forbush, and J. F. Hoffman, Biochemistry 17, 1929 (1978).

5 H. T. W. M. v.d. Hijden, E. GreU, J. J. H. H. M. de Pont, and E. Bamberg, J. Membr. Biol.
114, 245 (1990).
s M. Stengelin, K. Fendler, and E. Bamberg, J. Membr. Biol. 132, 211 (1993).
7 T. Lorentzon, G. Sachs, and B. Wallmark, J. Biol. Chem. 263, 10705 (1988).
[ 17] INVESTIGATION
protein (E1P and E2P). Na + and K + translocation is thought to proceed via

the E1P ~ E2P and E2 ~ E1 conformational transitions, respectively.
Membrane fragments, liposomes, or whole cells containing Na,K-
ATPase were used to investigate charge transport by this ion pump. Purified
preparations, 8-12 as well as native plasma membranes 2 were studied success-
fully with different electrophysiological techniques using rapid activation
via caged ATP.
Bilayer Experiments
Caged ATP [p3-1-(2-nitro)phenylethyladenosine 5'-triphosphate] is pre-
pared as described earlier. 4'8 To photolyse the caged ATP, light pulses of
a high-pressure 100-W Hg arc lamp and a mechanical shutter (exposure
time 125 msec) equipped with UV optics are used. Alternatively, an excimer
laser flash with a duration of 10 nsec, a wavelength of 308 nm, and an
energy density of 10-200 mJ/cm 2 is used. The fraction ~7 of caged ATP
converted to ATP from photo-release measurements is determined using
the luciferin luciferase assay (Boehringer, Ingelheim) as described earlier. 1°
Membrane fragments or liposomes containing the Na,K-ATPase are
adsorbed on a planar bilayer as shown in Fig. 1 (top). The front and rear
compartments are separated by a lipid bilayer that is formed from a solution
of diphytanoyl phosphatidylcholine (1.5%) and octadecylamine (0.025%)
in decane. The currents are measured using Ag[AgCI electrodes. Details
of the technique have been presented elsewhere. 8'13
After release of ATP by irradiation with UV light a current could be
measured that rises within ca. 100 msec and decays slowly. Under these
conditions, the rise of the signal is limited by the release of ATP during
the shutter opening time (125 msec) of the UV lamp. The polarity of the
current corresponds to the transport of positive charge to the planar bilayer
(Fig. 1). This is in agreement with the 3Na+/2K + stoichiometry of cation
transport by the Na,K-ATPase. The membrane fragments or liposomes are
not integrated in the planar bilayer but rather adsorbed onto it [Fig. 1 (top)
and Fig. 7 (left)]. This explains the slow decay of the current that is brought
about by the charging of the membrane capacitance and can be rationalized
by the equivalent circuit shown in Fig. 1 (top).
8 K. Fendler, E. Grell, M. Haubs, and E. Bamberg, E M B O J. 4, 3079 (1985).

9 K. Fendler, E. Grell, and E. Bamberg, FEBS Lett. 224, 83 (1987).
10G. Nagel, K. Fendler, E. Grell, and E. Bamberg, Biochim. Biophys. Acta 901, 232 (1987).
11 R. Borlinghaus, H.-J. Apell, and P. L~iuger, J. Membr. BioL 97, 161 (1987).
a2 H.-J. Apell, R. Borlinghaus, and P. L~iuger, J. Membr. Biol. 97, 179 (1987).
13 K. Fendler, S. Jaruschewski, A. Hobbs, W. Albers, and J. P. Froehlich, J. Gen. Physiol.
102, 631 (1993).
caged ATP
light -
~ i /JJ / ~ H+ 4
H+4
Na+,K+-, ~
::::"'
NaK-ATPase
membrane
fragments (MF) ~ BLM
ATP
5cm2
n~AI 1 ~ 0 mM Na +, 20 mM K +,
V 3 mM Mg 2+
+
5c-m2
ls + ionophores
Sm2
1s + vanadate
FIG. 1. (Top) Experimental setup for the bilayer experiments. Light enters the cuvette
through a quartz window. ATPase-containing membrane fragments and caged ATP are only
in the rear compartment. The membrane fragments adsorb to the planar bilayer as shown
on the right-hand side. The capacitive coupling of the membrane fragments to the bilayer is
described by the equivalent circuit. Stationary currents were measured in the presence of
monensin and 1799. (Bottom) Pump currents of the Na,K-ATPase measured in a solution
containing (a) 25 raM imidazole hydrochloride, pH 7.5, 130 mM NaC1, 20 mM KC1, 3 mM
MgC12, 35/zM caged ATP; irradiation: 125-msec flash from a UV lamp; enzyme preparation:
Na,K-ATPase from pig kidney. (b) 1 /xM 1799 and 10/zM monensin were present. 1 mM
vanadate inhibited the signal (c).
A d d i t i o n of the H + / N a , K - e x c h a n g i n g i o n o p h o r e m o n e n s i n and the pro-

t o n o p h o r e 1799 renders the bilayer conductive for N a ÷ and K ÷ and allows
the m e a s u r e m e n t of a stationary current (Fig. l b ) , which d e m o n s t r a t e s the
continuous p u m p i n g activity of the enzyme. Inhibitors of N a , K - A T P a s e
such as o r t h o v a n a d a t e (Fig. lc) o r p r e i n c u b a t i o n with o u a b a i n s abolish the
electrical signal.
T h e ion specificity of the transport activity of the N a , K - A T P a s e is shown
in Fig. 2. I n addition to M g 2÷, which is required for A T P binding, only
N a ÷ is necessary for the generation of an electrical signal. B a s e d on these
[17] INVESTIGATION
OF CHARGETRANSLOCAT1ON 293
~ _-------~ Mg2+, Na +, K +
j~ Mg 2+, Na +, K +
-~ . . . . . . . Mg 2+, O Na +, K +
Mg2+, Na +, O K +
.~S 12"5 nA/cm2
FIG. 2. Ion specificity of the signal. Conditions as in Fig. 1 except for the omission ( ~ ) of
certain cations as indicated.
results, an electrogenic step in the early Na+-dependent steps of the reaction

cycle is proposed, s
Caged ATP binds to the Na,K-ATPase and acts as a competitive inhibi-
tor in the protein reaction, a°'14 The stability of the biological preparations
and the power output of light sources limit the fraction of ATP that can
be released from caged ATP to about 30%. Therefore, photoactivated ATP
always has to compete with free caged ATP for the binding site on the
protein. This is shown in Fig. 3 where a dissociation constant for caged ATP
of ca. 30/zM was determined compared to a value of 1-2/zM for ATP. 1°
A rapid ATP concentration jump can be obtained by irradiation of
caged ATP with a XeCI excimer laser flash (duration 10 nsec). However,
the release of ATP from caged ATP is limited by the dark reaction of the
decay of caged ATP, which depends on pH and o n M g 2+ concentration. 15'16
At a temperature of 24° and a Mg2+ concentration of 3 mM, ATP is released
with a time constant of 2.6 msec at pH 6.2 and 16 msec at pH 7.0 [for a
consideration of the temperature correction see (•7)]. Therefore, the best
time-resolved measurements require an acidic pH.
14B. Forbush III, Proc. Natl. Acad. ScL U.S.A. 84, 5310 (1984).
15j. A. McCray, L. Herbette, T. Kihara, and D. R. Trentham, Proc. Natl. Acad. Sci. U.S.A.
77, 7237 (1980).
16j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, JACS 110, 7170 (1988).
17 K. Barabas and L. Keszthelyi, Acta Biochim. Biophys. Acad. Sci. Hung. 19, 305 (1984).
0,40
x 4rtM ATP
0.32
w
0.24
0.16
0.08 - X ~
xJ o ~ ~
o1 121xMATP
30~M ATP
I
I I I
-K I 0 80 160 240
~tM caged ATP
FIG. 3. Dixon plot of the peak current/peak at different concentrations of caged ATP and
released ATP after the light flash. The fraction of caged ATP released was varied by changing
the light intensity of the UV lamp. Conditions as in Fig. 1.
Time-resolved measurements at pH 6.2 were performed with Na,K-

ATPase from pig kidney 8 and eel electric organ. 13 In the absence of K +, a
signal similar to the one shown in Fig. 2 was obtained. The reciprocal
relaxation times of the rise and decay of the signal are shown in Fig. 4. In
this experiment, the concentration of ATP was varied by adding different
amounts of caged ATP to the electrolyte while maintaining the laser energy
constant. This resulted in a constant fraction (~7 = 0.23) of released ATP.
The influence of ATP (and caged ATP) on the reciprocal relaxation time
of the decay of the signal (72-1) is immediately apparent whereas T1-1
remains unaffected. Therefore, the decaying phase of the signal (r2 -1) was
assigned to the binding and exchange of ATP and caged ATP. This effect
can be quantitatively accounted for by a simple kinetic model.13 An indepen-
dent demonstration of the competition of ATP and caged ATP comes from
the slowing down of the apparent rate of phosphorylation with [3,-32p]ATP
in the presence of caged ATP. 13
The rapidly rising phase, having a reciprocal relaxation time of ~.-1
300 sec -1, is independent of the concentration of caged ATP and is there-
fore assigned to the electrogenic step. 13 Different propositions have been
put forward for a mechanistic explanation of an electrogenic reaction in the
Na+-dependent steps of the reaction cycle: electrogenic Na + translocation
during the conformational transition E~P ~ E2P or electrogenic release of
[ ] 7] INVESTIGATION OF CHARGE TRANSLOCATION 295
400
300 )~ _-a c o o
200 • 1"1= 302 s -1

100
0
70 - 400
xi'l (s'l)
60
- 300
5O
40
200
30
2O
100
10
0
0 21)0 4 ~ 6(~) 8()0 1000
caged A T P (/JM)
FIG. 4. Reciprocal time constants of the electrical signal generated by the release of ATP
from different concentrations of caged ATP using a short (10 nsec) laser flash that released
23% ATP. Conditions: 25 mM imidazole hydrochloride, pH 6.2, 130 mM NaCI, 3 mM MgC12,
1 mM dithiothreitol; enzyme preparation: Na,K-ATPase from eel electric organ. The concen-
trations refer to the caged ATP concentration before the flash. The signals are similar to the
ones shown in Figs. 1 and 2 except for a more rapid decay. 71 corresponds to the rapid rise,
¢2 to the slower decay of the signal.
Na + via a narrow extracellular access channel (for a review, see Rakowski

et al.18).
Patch Clamp Experiments

Isolated heart cells from guinea pig (myocytes) were investigated using
the giant patch technique. 1 Gigaohm seals are obtained on "blebs" from
myocytes (Fig. 5).
After excision of the patch, the pipette is moved into the temperature-
controlled perfusion and photolysis chamber (see Fig. 5). For more details,
see Friedrich et al. 2 To photolyse the caged ATP, light pulses of an excimer
laser are used.
Figures 6a and 6b show the response of the patch current to an ATP
concentration jump using photolysis of caged ATP at pH 7.4 and pH 6.3,
18 R. F. Rakowski, D. C. Gadsby, and P. de Weer, J. Membr. Biol. 155, 105 (1997).

FIG. 5. Schematic view of the photolysis experiment on a giant excised plasma membrane
patch from a ventricular heart cell. (Insert) Cardiac myocyte with a giant patch pipette (tip
diameter = 20 tzm) approaching a plasma membrane "bleb" (arrow).
Ip 2
(pA)
1 pH 7.4
T 40 Na +
[3 0 "~ 100 NMG+
3- ~5mMK+ I
pH 6.3
40 Na +
C 100 NMG+
3
Blockade
2 of the current
by Vanedate
50 100 150
time (ms)
FIG, 6. Patch clamp current recordings. (a) Response to photolysis of 500/zM caged ATP
at pH 7.4. (b) Response in a different patch at pH 6.3. Rat cell, 24°. Caged ATP concentration
250/zM. (c) Response to photolysis of 500/zM caged ATP on the same patch as in (a), after
inhibition of the stationary pump current by I mM orthovanadate. Bath solution: B-P. Pipette
solution: P-NMG-K. Guinea pig cell, 24°.
[ 17] INVESTIGATION OF CHARGE TRANSLOCATION 297
respectively. Figure 6c demonstrates the inhibition of this signal by pre-

treatment of the same patch with orthovanadate. In contrast to experi-
ments at pH 7.4, photolytic release of ATP at pH 6.3 results in a fast
transient outward current, followed by a stationary outward current,
which best-fit using a sum of two exponential functions and a constant.
The laser flash-induced signals are measured at different caged ATP con-
centrations, but at a constant fraction 77 of released ATP. The rise of the
signal at pH 6.3 is characterized by an ATP-independent rate constant
kl of --200 sec -1 (at 24°) and the decay by an ATP-saturable rate con-
stant k2.
We found an exponentially increasing pump current on photolysis of
caged ATP at pH 7.4 (Fig. 6a). The rate of this increase is 70 s e c --1, but
this rate is dependent on the caged ATP concentration and light intensity.
From this it has been concluded that the rise time of the signal at pH 7.4
is determined by the release of ATP from caged ATP. Therefore, all studies
are performed at pH 6.3, where the release of ATP proceeds with a time
constant of 3.2 m s e c . 16J7
The current traces obtained with the patch clamp and the bilayer tech-
nique are similar in the fast time range. At t > 100 msec the patch clamp
current has decayed to a stationary current whereas the bilayer signal decays
further and becomes negative at t > 200 msec. The latter behavior is
attributable to the capacitive coupling of the membrane fragments to the
planar bilayer. 13 Capacitive coupling does not alter the rate constants but
only adds the time constant of the system and distorts the amplitudes of
the signal. 13 Therefore, an analysis of the rate constants is meaningful
whether or not the ion pump is capacitively coupled to the measuring
system.
The rate constants obtained by both methods at different concentrations
of released ATP depend in the same way on the ATP concentration and
have approximately the same values. Both techniques show a rapid rise of
the electrical signal with a rate constant of -200 sec -1, which is independent
of the ATP concentration at 24°. However, the decay of the signal is dearly
ATP-dependent with an affinity of 10 to 20/~M. Based on experiments on
planar lipid membranes, we have previously assigned the rising phase of
the signal to a fast electrogenic step with a rate of --200 sec-1. 9'13 These
conclusions are now confirmed by the results of our patch clamp study
using cardiac myocytes. This rules out tissue-specific behavior as a reason
for discrepancies found in the literature.
We conclude that phosphorylation of cardiac Na,K-ATPase is fast and
the electrogenic E1P-E2P conformational change proceeds with a rate con-
stant of --200 sec -1 at 24°. This is not contradictory to an electroneutral
E1P-E2P conformational change (with a rate constant of 200 sec -~) followed
by a fast electrogenic Na ÷ release, 19as put forward by access channel models

for the Na,K-ATPase. 2° In our experiments this mechanism would yield
the same results because it is kinetically equivalent to an electrogenic
EIP-E2P conformational change.
ADP, ATP-Carrier
The ADP, ATP-carrier (AAC) is the most abundant protein in the
inner mitochondrial membrane. Its function consists of the export of ATP
from the mitochondria into the cytosol of the cell. The transport occurs in
exchange with ADP, which is used for the production of ATP via oxidative
phosphorylation. The function of the carrier has been studied in the past
in mitochondria as well as in reconstituted proteoliposomes. It was shown
that the AAC is potential sensitive and that the electrogenic event is related
to the ATP translocation.21'2z Roughly one net charge per ATP is trans-
ported.23,24
Because the ADP/ATP exchange was assumed to be electrogenic, the
transport of the nucleotides has to be accompanied by electrical currents.
The carrier is an ideal object for the application of caged ATP or caged
ADP. To study its electrical transport, proteoliposomes reconstituted with
AAC are adsorbed to a planar lipid bilayer as described earlier for Na,K-
ATPase. The configuration of liposomes and planar bilayer is shown in
Fig. 7. Caged ATP and caged ADP [p3-1-2-nitro)phenylethyladenosine
5'-triphosphate or diphosphate] are prepared as described earlier. 4'8 To
photolyse the caged ATP, 125-msec light pulses of a high-pressure 100-W
Hg arc lamp are used.
Although ion pumps create ion gradients by the use of the chemical
energy from ATP, secondary active transporters such as the AAC need
gradients as a driving force for the transport. This implies that proteolipo-
somes preloaded with ATP or ADP in the case of the AAC have to be
impermeable to nucleotides. Because of their charged nature the nucleo-
tides can be retained in the liposomes over a period of days. Also, in the
presence of external caged nucleotides, no exchange or loss of the internal
nucleotides occurs. From this a second prerequisite for the successful appli-
19D. W. Hilgemann, Science 263, 1429 (1994).

20 D. C. Gadsby, R. F. Rakowski, and P. De Weer, Science 260, 100 (1993).
21 E. Pfaff and M. Klingenberg, Eur. J. Biochem. 6, 66 (1968).
2z M. Klingenberg and H. Rottenberg, Eur. J. Biochem. 73, 125 (1977).
23 K. LaNoue, S. M. Mizani, and M. Klingenberg, Z Biol. Chem. 253, 191 (1978).
24 R. Kr~imer and M. Klingenberg, Biochemistry 21, 1082 (1982).
[ 17] INVESTIGATION OF CHARGE TRANSLOCATION 299
l ~~ caged ATP
ADP / ATP Carrier

I
BLM
B
l ADP / ATP Carrier

BLM
FIG. 7. Capacitive currents by the ADP,ATP carrier after a light-induced concentration
jump of ATP or ADP. Proteoliposomes preloaded with ADP (A) and ATP (B). The sign of
the current corresponds to a movement of negative charges from the outside into the lumen
of the lipid vesicles. (A) The oppositely directed current in (B) corresponds to the reversed
application of the nucleotides; 100 tzM caged nucleotide was applied. One UV flash releases
30% of ATP. Flash duration was 125 msec with 2 W/cm 2 intensity. 0.1 M NaC1, pH 7.2.
cation of the caged nucleotides is fulfilled, namely that the caged nucleotides
are not transported by the AAC.
In principle, six different transport modes can be studied. In the "nor-
mal" heteroexchange with either A D P or ATP inside the vesicles (caged
ATPex/ADPin and caged ADPex/ATPin), the current is initiated by photoly-
sis either of caged ATP or caged ADP. In both cases, the obligatory counter
exchange is observed. Figure 7 shows such an experiment. As expected,
the transient capacitive currents are oppositely directed. From the sign of
the current it was derived that negative charge during ATP/ADP exchange
is translocated concomitant with ATP transport. 25In addition to the hetero-
25 N. Brustovetsky, A. Becker, M. Klingenberg, and E. Bamberg, Proc. Natl. Acad. Sci. U.S.A.
93, 664 (1996).
exchange A T P / A D P , the homoexchange ADPex/ADPin and ATPex/ATPi.

are also of interest as well as the zero-trans conditions ATPex/0~, and
ADPex/0i,. From these results, information about the electrical nature of
the transport is obtained. It was demonstrated that all transport modes are
accompanied with an electrical current and therefore with charge move-
ments. The current due to the A D P transport is connected with a movement
of positive charge.
In all systems, whether in hetero- or homoexchange modes or in un-
loaded vesicles, the currents have an opposite sign depending on whether
they started with caged A T P or with caged ADP. This agrees with the
postulate that the charge difference between A T P 4- and A D P 3- is responsi-
ble for the currents. However, it contradicts the concept derived from other
experimental evidence that A T P and not A D P carries the charge. Thus, it
has been suggested that A T P 4- combines with three positive charges of the
binding center forming A T p 4 - c 3+ with one negative excess charge, whereas
A D P 3- would neutralize the binding center forming ADp3-c3+. 26 The re-
sults obtained under homoexchange conditions lead to the conclusion that
A T P 4- binds with a negative charge excess and A D P 3- binds with a positive
charge excess to the carrier. With 2 A T P 4- one negative charge and with
2 A D P 3- one positive charge are moved across the membrane or, alterna-
tively, one carrier site works alone, giving rise to the net transfer of a
fractional charge equivalent to one-half of an elementary charge:
(ATp4-c35+) °'5- and (ADp3-c35+) °-5+.
The experiments just shown demonstrate the possibilities of measuring
currents associated with AAC-catalyzed A D P / A T P transport or with inter-
action of A A C and ligands. Because of high time resolution, this will allow
us to study the kinetics of the charge translocation in the A A C and thus
give us more information about the single steps in the reaction cycle of the
carrier similarly as obtained in the earlier work on ATP-driven ion pumps.
The approach used in the present work may also be applied to study
other carriers.
Caged Ca 2÷ a n d N a , C a - E x c h a n g e r
Photolabile Ca 2+ chelators (caged Ca 2+) have been used to study the
Na+,Ca 2÷ exchange current from mammalian heart 27-3° and lobster skeletal
26N. Brustovetsky and M. Klingenberg,J. Biol. Chem. 269, 27329 (1994).

27T. A. Powell, A. Noma, T. Shioya,and R. Z. Kozlowski,Jr. Physiol. (Lond.) 472, 45 (1993).
[ 1 7] INVESTIGATION OF CHARGE TRANSLOCATION 301
muscle. 31 In the experiments described in this article, the EDTA derivative

DM-nitrophen 32was used to study the presteady-state current of the Na,Ca-
exchanger from rat and guinea pig ventricular cells by means of the giant
patch clamp technique 1 (see also the section on the Na,K-ATPase) on the
submiUisecond time s c a l e . 33 DM-nitrophen is suitable for these studies
because it is photolyzed rapidly (r -< 30/zsec 34) and with a dramatic change
of the affinity for Ca 2÷ from 0.5 nM to 3 mM. Using a laser flash (308 nm,
10 nsec, 480 mJ/cm2), 20% of the chelator can be cleaved and the Ca 2÷
concentration jumps from 0.1 to 100/zM. Because the affinity of the unpho-
tolyzed DM-nitrophen fits to the apparent affinity of the transport sites for
Ca z÷ on the cytoplasmic side of the Na,Ca-exchanger (5/zM), it allows for
the adjustment of suitable free Ca 2÷ concentrations (-<0.5/zM) without an
excess of chelator over total Ca/÷ (see Fig. 8), which would lead to so-
called "Ca 2÷ spikes ''35 after photolysis by a flash.
In addition to transport-binding sites, a regulatory Caa+-binding site has
been identified on the cytoplasmic side of the exchanger 36,37that inactivates
the exchanger if no Ca 2÷ is bound. The regulatory action of this site can
be eliminated by a mild treatment with chymotrypsin. 37a Because the Ca 2÷
affinity of the regulatory site is higher than that of the transport site it has
only been possible to study its action in the reverse mode of the Na,Ca-
exchanger 38 or in mutants with drastically reduced Ca 2÷ affinity. 39 It could
be shown, however, that photolytic Ca 2÷ concentration jumps allow the
experimenter to distinguish between both Ca2÷-binding sites on the wild-
type exchanger under conditions of forward Na ÷, cae÷-exchange. 33
Figure 8 shows the current signal obtained from an excised patch from
a ventricular cell after the photolytic release of Ca 2÷ (0.5 ~ 100/zM) on
38 E. Niggli and W. J. Lederer, Nature 349, 621 (1991).

29 E. Niggli and W. J. Lederer, Biophys. J. 65, 882 (1993).
3o E. Niggli and P. Lipp, Biophys. J. 67, 1516 (1994).
31 A. Eisenrauch, M. Juhaszova, G. C. R. Ellis-Davies, J. H. Kaplan, E. Bamberg, and M. P.
Blaustein, Z Membr. Biol. 145, 151 (1995).
32j. Kaplan and G. C. R. Ellis-Davies, Proc. Natl. Acad. Sci. U.S.A. 85, 6571 (1988).
33 M. Kappl and K. Hartung, Biophys. J. 71, 2473 (1996).
34 G. C. R. Ellis-Davies, J. H. Kaplan, and R. J. Barsotti, Biophys. J. 70, 1006 (1996).
35 R. S. Zucker, Cell Calcium 14, 87 (1993).
36 R. DiPolo and L. Beaugr, Biochim. Biophys. Acta 854, 298 (1986).
37j. Kimura, A. Noma, and H. Irisawa, Nature 319, 596 (1986).
37a W. Stiirmer, H.-J. Apell, I. Wuddel, and P. Lauger, J. Membr. Biol. 110, 67 (1989).
38 D. W. Hilgemann, A. Collins, and S. Matsuoka, J. Gen. Physiol. 100, 993 (1992).
39 S. Matsuoka, D. A. Nicoll, L. V. Hryschko, D. O. Levitsky, J. N. Weiss, and K. D. Philipson,
J. Gen. Physiol. 105, 403 (1995).
302 CAGED COMVOUNDS [ 17]
. *° °~
o, o; ,'.:.,'-'. i4~
o
Xl = 0 . 0 1 9 m s
x2 = 0.56 ms
1 ms
FIG, 8. Current recorded from an excised membrane patch (guinea pig ventricular cell)
after a saturating Caz+ concentration jump under conditions that promote inward Na-Ca
exchange current (forward mode of Na-Ca exchange). The time course has been fit by using
a sum of two exponential functions plus a constant. The rising phase is determined by the
rise time of the recording system (15/zsec). The extracellular solution contained (in raM)
100 Na +, 20 Cs+, 10 TEA, 2 Mg2+, 119 CI , 10 EGTA, 10 HEPES, 0.02 Verapamil. The
solution on the cytoplasmic side contained (in raM): 100 Li÷, 0.44 Ca2+, 0.45 DM-nitrophen,
20 Cs+, 20 TEA, 10 HEPES. The Ca2+ concentration before the flash was 0.5/xM. After the
laser flash, Ca2+increased to > 100 brM, saturating the cytoplasmic transport sites. All solutions
were adjusted to pH 7.1. The temperature was maintained at 21°. The cytoplasmic side of
the membrane patch has been treated with a-chymotrypsin (1 mg/ml) for 1 min to eliminate
the action of the regulatory-binding site.
the cytoplasmic side u n d e r conditions that p r o m o t e f o r w a r d Na+,Ca 2+-

exchange, i.e., Naex = 100 m M , Caex = 0, Nain = 0. F o r further experimental
details, see the figure legend. Negative current flow indicates an inward
current f r o m the extracellular to the intracellular side. T h e current rises
within 50 tzsec to a p e a k and declines exponentially to a plateau with a
time constant of about 0.6 msec. T h e rising phase c a n n o t be resolved
because it is disturbed by an initial artifact due to the dissipation of laser
energy. This artifact, which varies in size f r o m experiment to experiment,
could be r e d u c e d by using petri dishes m a d e f r o m glass instead of f r o m
polystyrene. A further reduction could be achieved by short-circuiting the
input of the A x o p a t c h 2 0 0 A amplifier ( A x o n Instruments, F o s t e r City, C A )
for 50/xsec beginning 20/xsec before the laser flash is triggered. T h u s the
rising phase of the current signal does n o t provide useful information on
the kinetics of the Na,Ca-exchanger.
T h e direction of stationary current is compatible with the 3Na+/1Ca 2+
stoichiometry of the exchanger. With respect to the transient current we
have suggested that the presteady-state current is mainly d e t e r m i n e d by
the o u t w a r d translocation of Ca2+. 33 If this is true, a plausible assumption
is that the exchanger m o v e s negative net charge t o w a r d the extracellular
[17] INVESTIGATION OF CHARGE TRANSLOCATION 303
A
~ =0.05 ms
=0.33 ms
2
FIG. 9. Transient current signals recorded from an excised membrane patch under condi-
tions promoting Ca-Ca exchange before (A) and after the application of chymotrypsin (B).
The transient currents can be described by a sum of two exponential functions without a
stationary component. The free Ca 2÷ concentration before the flash was adjusted to 0.1/zM;
after the flash it was >100/zM. The pipette solution contained (in mM) 100 Li ÷, 5 Ca 2÷, 20
TEA, 20 Cs*, 10 HEPES, 0.02 verapamil. The cytoplasmic solution contained 100 Li +, 0.44
Ca 2÷, 0.455 DM-nitrophen, 20 Cs ÷, 20 TEA, 10 HEPES, 130 CI-. The bandwidth of the
recording system was 5 kHz. Other conditions as described in Fig. 8.
phase during Ca 2÷ translocation. This idea is supported by Ca 2÷ concentra-

tion jumps under conditions that promote Ca-Ca exchange, which is elec-
troneutral in the stationary state. Figure 9 shows a current signal generated
by a Ca e÷ concentration jump in the absence of extracellular Na ÷ with 5
mM Ca 2÷ in the pipette. Under the conditions of Ca-Ca exchange, a tran-
sient inward current is observed that decays with a time constant of 0.3
msec. No stationary current is observed in the absence of extracellular Na ÷
in agreement with the electroneutrality of Ca-Ca exchange. The observa-
tion of a transient inward current following a Ca 2÷ concentration jump in
both transport modes, forward Na-Ca exchange and Ca-Ca exchange, is
most easily explained by assuming that the initial inward current is related
to the outward movement of Ca 2÷. If net charge movement of the Ca-
loaded enzyme occurs in the same direction as the Ca 2÷ translocation, then
one has to assume that the enzyme carries negative charge in this state. A
different situation has been observed in the Ca-ATPase from sarcoplasmic
reticulum that carries positive charge if it is loaded with Ca2+. 4°
The application of rapid Ca 2+ concentration jumps offers a unique
chance to study the action of the regulatory Ca>-binding site under condi-
tions that allow for the binding of Ca 2÷ to both sites (i.e., forward Na-Ca
exchange and Ca-Ca exchange). Partial inactivation of the Na,Ca-
exchanger can be induced by lowering the preflash Ca a+ concentration to
about 0.1/£M. 33 Under these conditions a Ca 2÷ concentration jump elicits
40 K. Hartung, J. P. Froehlich, and K. Fendler, Biophys. J. 72, 2503 (1997).

A B
• '° A
10 pA s
x2= 0,54 ms x = 620 ms
C D
20 p A ~ '-~
l~ x2= 0.55 ms
. . . . . . . . . . . .
-t o i 2 6 2c;oo46oo'
Time [ ms ] Time [ ms ]
FIG. 10. Effect of the regulatory cytoplasmic Ca2+-binding site on the stationary forward
N a - C a exchange current. Current signals following a Ca 2+ concentration jump were recorded
at two different time resolutions before (A,B) and after (C,D) treatment of the patch with
a-chymotrypsin (1 mg/ml) for 40 sec. The preflash Ca 2+ concentration was less than 0.1/zM.
Other conditions as in Fig. 9. At high temporal resolution (A,C), the time course of the
current signals is identical, but the amplitudes of transient and stationary current are consider-
ably increased after the treatment with chymotrypsin. At a lower resolution another effect
of the treatment with chymotrypsin is visible. In the untreated membrane, an initial rise,
corresponding to the stationary amplitude shown in (A), is followed by a slowly rising phase
that can be fit using an exponential function with a time constant r = 90.6 sec. Because this
secondary rise is abolished by the treatment with chymotrypsin (C), it has been attributed to
the binding of Ca 2+ to the regulatory binding site.
a signal as shown in Fig. 10A which, on the time scale of microseconds, is

comparable to the signal shown in Fig. 8. However, on a slower time scale
(Fig. 10B) it can be seen that there is a secondary increase of the stationary
current with a time constant of about 0.6 sec. This secondary rise of outward
current is attributed to the action of the regulatory Ca2+-binding site because
it is eliminated if either the prefiash Ca 2÷ concentration is >0.5/zM or if
the membrane has been treated with chymotrypsin (Fig. 10D). Both means
are known to release the exchanger from the inactivated state. 41 These
experiments show that Ca2÷ binding to the regulatory binding sites is much
slower than Ca2÷ binding to the transport sites.
Treatment with chymotrypsin increases the amplitude of stationary and
peak inward current in both transport modes without modification of the
41 D. W. Hilgemann, Nature 344, 242 (1990).

[ 1 7] INVESTIGATION
TABLE I
CHARGE TRANSPORTOF IoNoTRANSLOCATINGPROTEINS USING CAGED SUBSTRATES
Caged
Pump Techniquea compound Ref.
Na,K-ATPase MF Caged ATP 8-13, 37a, 47-49

INT 44
PC 2, 50
H,K-ATPase MF Caged ATP 5, 6
Ca-ATPase MF Caged ATP 40, 45
INT 52
FoF1-ATPase LIP Caged ATP 51
Kdp-ATPase LIP Caged ATP 46
Na,Ca-exchanger PC Caged Ca2+ 33
MF 31
ADP, ATP-exchanger LIP Caged ATP 25
Caged ADP
a MF, Purified or native membrane fragments or vesicles adsorbed to a planar bilayer; LIP,
purified ion pumps reconstituted into liposomes adsorbed to a planar bilayer; INT, ion pumps
integrated in a planar bilayer; PC, giant patch technique.
kinetics of the transient current, indicating that a fraction of transporter

molecules is fully inactivated at low cytosolic Ca 2+ concentrations. 42
The observation of a transient inward current after a saturating Ca 2+
concentration j u m p in the C a - C a and the N a - C a exchange mode, which
is followed by a stationary inward current in the N a - C a exchange mode,
provides some insight into the reaction cycle of the Na,Ca exchanger (for
details, see Kappl and Hartung33). First, it shows that Ca 2+ binding to the
transport sites on the cytosolic side is fairly rapid (kon > 108 M -1 sec -a)
because the time to p e a k is <50/zsec. Second, it shows that Ca 2+ transloca-
tion is electrogenic and, together with previous findings that Na ÷ transloca-
tion is a major electrogenic transition, it indicates that the Na,Ca-exchanger
contains at least two electrogenic steps. Third, it demonstrates that a transi-
tion following the first electrogenic step (Ca 2+ translocation) is rate limiting,
e.g., Na ÷ translocation.
Conclusions
We have given a few examples of transport ATPases and carriers that
have been investigated using caged substrates. The advantage of this ap-
proach is the high time resolution, which allows the investigation of pre-
42D, W. Hilgemann, Biophys. I. 71, 759 (1996),

steady-state kinetics. In addition to the systems described here, a number

of other ion-translocating ATPases were investigated using caged ATP: the
Ca-ATPase from sarcoplasmatic reticulum, the H,K-ATPase from stomach,
the FoF1-ATPase from chloroplasts, and the Kdp-ATPase from the plasma
membrane of Escherichia coli. These systems are summarized together with
the respective electrophysiological technique in Table I.
In the case of caged ATP the time resolution is limited by the time
constant of ATP release from the caged compound used in those studies
[p3-1-(2-nitro) phenylethyladenosine 5'-triphosphate]. By working at a
somewhat acidic pH (6.2), ATP is released rapidly enough (within ca. 3
msec) to study most of the enzymatic reactions of interest. Nevertheless,
compounds with an improved time resolution at neutral pH are desirable
and will be available in the near future¢ 3 In the case of caged Ca 2+, the
release is extremely rapid so that the limit there is still set by the time
resolution of the electrophysiological techniques.
Acknowledgments
The results presented in this article are based on the work of numerous colleagues and
collaborators. Special thanks go to M, Klingenberg and N. Brustovetsky, University of Munich,
and to M. Kappl and T. Friedrich, Max-Planck Institut for Biophysik, Frankfurt. The work
was supported by the DFG (SFB 169 and 472) and the Max-Planck Society.
43 R. S. Givens, J. F. W. Weber, A. H. Jung, and C.-H. Park, Methods Enzymol. 291 [29]
(1998) (this volume).
44 A. Eisenrauch, E. Grell, and E. Bamberg, "The Sodium Pump: Structure, Mechanism, and
Regulation." Rockefeller Univ. Press, New York, 1991.
45 K. Hartung, E. Grell, W. Hasselbach, and E. Bamberg, Biochim. Biophys. Acta 900, 209
(1987).
46 K, Fendler, S. DrOse, K. Altendorf, and E. Bamberg, Biochemistry 35, 24, 8009 (1996).
47 S. Heyse, I. Wuddel, H.-J. Apell, and W. Stttrmer, J. Gen. Physiol. 104, 197 (1994).
48 I. Wuddel and H.-J. Apell, Biophys. J. 69, 909 (1995).
49 R. Borlinghaus and H.-J. Apell, Biochim. Biophys. Acta 939, 197 (1988).
50T, Friedrieh and G. Nagel, Biophys. J. 73, 186 (1997).
5a B. Christensen, M. Gutweiler, E. Grell, N. Wagner, R. Pabst, K. Dose, and E. Bamberg,
J. Membr. BioL 104~ 179 (1988).
52 A Eisenrauch and E. Bamberg, FEBS Lett. 268, 152 (1990).
[18] CAGED MOLECULES AND MOTOR PROTEINS 307
[ 18] Studies of M o l e c u l a r M o t o r s
Using Caged Compounds
By JODY A. DANTZIG, HIDEO HIGUCHL and YALE E. GOLDMAN
Introduction
The dynamic nature of cell motility and the macromolecular organiza-
tion of contractile structures in cells make the study of muscle contraction
and nonmuscle cell motility natural targets for the application of photolabile
compounds and flash photolysis. After the synthesis of caged ATP and its
application to Na+-K÷-ATPase pumps were reported by Kaplan et aL, I the
study of muscle contraction was one of the first fields to benefit from this
new technology.2
Cell motility includes contraction of skeletal, cardiac, and smooth mus-
cles, locomotion of nonmuscle cells, dynamic alteration of cell shape by
rearrangements of the cytoskeleton, intracellular motions and targeting of
organelles, karyokinesis (congression and segregation of the chromosomes
to the daughter cells in cell division), cytokinesis (splitting of the progenitor
cell into two daughter cells), and other motions. The motor proteins that
carry out these essential functions are members of three superfamilies: the
myosins, kinesins, and dyneins.
Myosin, in muscle or in the cytoplasm of nonmuscle cells, slides along
actin filaments (F-actin) toward their "barbed" (plus) ends using the energy
liberated from the hydrolysis of cellular ATP to ADP and orthophosphate
(Pi). More than 13 subgroups of myosins have been identified with homolo-
gous N-terminal globular regions, termed the heads or motor domains, but
with highly variable molecular mass (<100-500 kDa), assembly, enzymatic
activity, and function# In muscle, two-headed myosin molecules polymerize
into bipolar filaments that interdigitate with F-actin to form almost crystal-
line cylindrical organelles ( - 1 / z m in diameter) termed myofibrils.
Kinesins are smaller motor proteins (-120 kDa) that translocate along
microtubules (stiff cytoskeletal structures polymerized from a- and/3-tubu-
lins) toward the plus end (the fast growing end, generally peripheral in
cells). Other members of the kinesin superfamily translocate toward the
minus (slower growing, central) end of microtubules. Dozens of kinesin-like
proteins have been identified with roles in mitosis and targeted intracellular

2 y . E. Goldman, M. G. Hibberd, J. A. McCray, and D. R. T r e n t h a m , Nature 3@0,701 (1982).
3 M. S. M o o s e k e r and R. E. Cheney, Annu. Rev. Cell Dev. Biol. 11, 633 (1995)

METHODS IN ENZYMOLOGY, VOL. 291 0076-6879/98$25.00
308 CAGED COMPOUNDS [1 81
transport. 4 Dynein is a giant (1000-2000 kDa) motor protein that slides
toward the minus end of microtubules. 5'6 Flagellar dyneins drive the bending
of eukaryotic axonemes and cytoplasmic dyneins perform targeted trans-
port of intracellular vesicles and nuclear migration. Kinesins and dyneins
also transduce the energy liberated by the hydrolysis of ATP to perform
mechanical work.
The cytoskeletal structures, filamentous actin (F-actin) and microtu-
bules, are tracks that guide and direct the translocation motions. A diverse
group of proteins that bind to actin or tubulin control the remarkably
dynamic formation and remodeling of these cytoskeletal structures and
regulate the activity of molecular motors. 7'8
Solution of the crystal structures of actin 9 and the motor domains of
myosin 1° and kinesin 11'12 has extended research on the functional mecha-
nisms of these proteins to the amino acid and even the atomic level. The
catalytic ATP-binding core of kinesin and myosin have surprisingly similar
secondary structure even though there is little sequence homology. The
atomic structure of this catalytic core also maps closely onto the nucleotide-
binding regions of many otherwise unrelated enzymes, including creatine
kinase, ~3 the mitochondrial ATP synthase, 14 and heterotrimeric and small
GTP-binding proteins, such as a-transducin, a5 EF-Tu, 16 and the H-ras p21
oncogene.17 Thus understanding the operation of motor proteins may reveal
general principles in enzyme biophysics and help to unravel the molecular
mechanisms of these other signaling and energy transducing enzymes.
Some features of the chemomechanical events in the energy converting
ATPase cycle are similar among the various motor proteins (Fig. 1). Each
protein associates and dissociates with its cytoskeletal track during the
4 L. S. B. Goldstein, Annu. Rev. Genet. 27, 319 (1993),

5 E. L. F. Holzbaur and R. B. Vallee, Annu. Rev. Cell Biol. 10, 339 (1994).
6 I. R. Gibbons, Cell Struct. Funct. 21, 331 (1996).
7 T. D. Pollard, S. Almo, S. Quirk, V. Vinson, and E. E. Lattman, Annu. Rev. Cell Biol. 10,
207 (1994).
8 R. B. Vallee, Cell Motil. Cytoskel. 15, 204 (1990).
9 W. Kabsch, H. G. Mannherz, D. Suck, E. F. Pai, and K. C. Holmes, Nature 347, 37 (1990).
10 I. Rayment, W. R. Rypniewski, K. Schmidt-B~ise, R. Smith, D. R. Tomchick, M. M. Benning,
D. A. Winkelmann, G. Wesenberg, and H. M. Holden, Science 261, 50 (1993).
11 F. J. Kull, E. P. Sablin, R. Lau, R. J. Fletterick, and R. D. Vale, Nature 3811, 550 (1996).
12 E. P. Sablin, F. J. Kull, R. Cooke, R. D. Vale, and R. J. Fletteriek, Nature 3 ~ , 555 (1996).
13 K. Fritz-Wolf, T. Schnyder, T. Wallimann, and W. Kabsch, Nature 381, 341 (19%).
14j. p. Abrahams, A. G. W. Leslie, R. Lutter, and J. E. Walker, Nature 370, 621 (1994).
is j. p. Noel, H. E. Harem, and P. B. Sigler, Nature 366, 654 (1993).
16H. Berchtold, L. Reshetnikova, C. O. A. Reiser, N. K. Schirmer, M. Sprinzl, and R. Hilgen-
feld, Nature 36S, 126 (1993).
17 E. F. Pai, U. Krengel, G A, Petsko, R. S. Goody, W. Kabsch, and A. Wittinghofer, EMBO
J. 9, 2351 (1990).
[ 18] CAGED MOLECULES AND MOTOR PROTEINS 309
Myosin and
Dvnein Kinesin
strong E strong
weak E-ATP strong

2
weak E-ADP-Pi strong

3
strong E-ADP weak

4
strong E strong
FIo. 1. Biochemical scheme for the ATPase activity of motor proteins (E). "Strong" and
"weak" refer to the affinity of the motor domain of the proteins for their respective cytoskeletal
tracks, actin filaments in the case of myosin and microtubules for dynein and kinesin. In
strongly bound states, the motor proteins are mostly bound to the track and can exert sliding
force to translocate their cargo.
enzymatic cycle, and the biochemical transitions are probably associated

with structural changes that lead to the generation o f force or filament
sliding. Myosin typically is tuned for rapid filament sliding, which requires
the cooperative action of many molecules. Myosin heads are strongly bound
to actin in the nucleotide-free and ADP-bound states and associate weakly
with actin in the ATP- and ADP-Pi- bound states. 18'~9 The transition to
force generation seems to be associated with the release of Pi (Fig. 1, step
3) from actomyosin-ADP-pi.20-22
Kinesin translocates more slowly than myosin, but a single kinesin mole-
cule, with two motor domains, can slide along a microtubule without diffus-
ing away, presumably by alternate power strokes of the two heads. 23 The
18 R. W. Lymn and E. W. Taylor, Biochemistry 10, 4617 (1971).

19E. Eisenberg and L. E. Greene, Annu. Rev. PhysioL 42, 293 (1980).
20 M. G. Hibberd, J. A. Dantzig, D. R. Trentham, and Y. E. Goldman, Science 228, 1317 (1985).
21 j. A. Dantzig, Y. E. Goldman, N. C. Millar, J. Lacktis, and E. Homsher, J. Physiol. 451,
247 (1992).
22 R. Cooke and E. Pate, Biophys. J. 48, 789 (1985).
23 D. D. Hackney, Nature 377, 448 (1995).
kinesin head is strongly associated with a microtubule in the nucleotide-

free, ATP- and ADP-Pj-bound states and weakly associated in the ADP-
bound state. 24"25The pathway of the dynein ATPase is more similar to that
of myosin than kinesin. 26
In organized systems capable of transducing chemical to mechanical
energy, the kinetics and pathway of the elementary reaction steps within
the ATPase cycles can be probed by photorelease of specific ligands, such
as ATP, 27 ADP, 28 Pi, 21'29 and Ca2+, 30'31 o r by the photochelation of Ca2+.31
Sudden photoliberation of these biological activators and inhibitors initi-
ates the motor protein reactions from well-defined starting configurations
or perturbs the dynamics of their enzymatic cycles. Photolysis avoids tempo-
ral limitations due to diffusion of ligands into the macromolecular assem-
blies and mechanical disruption associated with exchange or mixing of
solutions. The mechanical and structural responses initiated by photolysis
indicate the course of events and their rates. Varying the mechanical condi-
tions, such as external load, identifies which steps are dependent on mechan-
ical strain and thereby likely to be associated with force generation and
which reactions control the overall ATPase rate. Another potential applica-
tion of photolysis in motility studies is to deliver substrate or signaling
molecules at highly localized regions in functioning intact cell systems.
This article summarizes methods developed to study cell motility using
photolysis of photolabile precursors of nucleotides, nucleotide analogs, and
Ca 2÷. Applications of caged compounds to signalling in muscle are described
elsewhere in this volume. We describe preparations of motor proteins,
apparatus, appropriate solutions, reagents, caged compounds, photolysis
light sources, special precautions required for photolysis experiments on
motor proteins, optimization of the caged molecule concentration and pho-
tolysis chamber dimensions, binding of Ca 2+ and Mg 2+ to caged ATP, and
inhibition of sliding velocity of myosin and kinesin by caged ATP. Many of
these considerations should be applicable to other macromolecular systems.
24 L. Romberg and R. D. Vale, Nature 361, 168 (1993).

25 I. M.-T. C. Crevel, A. Lockhart, and R. A. Cross, J. Mol. Biol. 257, 66 (1996).
26 D. D. Hackney, Annu. Rev. Physiol. 58, 731 (1996).
z7 y. E. Goldman, Annu. Rev. Physiol. 49, 637 (1987).
2s Z. Lu, R. L. Moss, and J. W. Walker, J. Gen. Physiol. 101, 867 (1993).
29 N. C. Millar and E. Homsher, J. Biol. Chem, 265, 20234 (1990).
3°T. D. Lenart, T. St. C. Allen, R. J. Barsotti, G. C. R. Ellis-Davies, J. H. Kaplan,
C. Franzini-Armstrong, and Y. E. Goldman, in "Mechanism of Myofilament Sliding in
Muscle Contraction" (H. Sugi and G. H. Pollack, eds.), p. 475. Plenum Press, New York,
1993.
31 C. C. Ashley, I. P. Mulligan, and T. J. Lea, Quart. Rev. Biophys. 24, 1 (1991).
Actomyosin ATPase Activity in Muscle Fibers
Preparation of Muscle Fibers

The most common sample for photolysis experiments on muscle is a
segment of a single fiber, 2-3 mm in length, dissected from a bundle of
glycerol-extracted fibers of rabbit psoas muscle. 32This preparation is conve-
nient, retains good mechanical integrity, and can be compared to many
experiments in the literature on isolated, purified proteins from rabbit
muscle. Glycerol extraction effectively removes the diffusion barrier of the
surface membrane (sarcolemma), allowing rapid (1-2 min) equilibration
of bathing constituents, such as caged molecules, with the interior of the
myofilament lattice.32 The procedure for preparing muscle is to expose the
psoas muscle anteriorly, tie small (<0.5-1 mm diameter) bundles of fibers
to glass or wooden sticks and soak them in a membrane disruption solution
containing a high concentration of ATP, low free [Ca 2+] and a cocktail of
protease inhibitors. The solution is gently stirred at 4° overnight, taking care
not to mechanically agitate the muscles. The bundles are then transferred to
a solution with Mg-ATP and 50% glycerol to further extract the membranes
and stored at - 2 0 ° for up to 6 weeks. 32 Single fiber segments are separated
from bundles under silicone oil or in glycerol relaxing solution, with low free
[Ca 2+] and high Mg-ATP concentration, maintained at 10-12 ° to minimize
proteolysis and bacterial contamination. Glycerol-extracted psoas muscle
fibers generate ca. 200 kN.m -2 of isometric (fixed length) force at 20° and
shorten with maximum filament sliding velocities33 of 2.1/~m.sec -1 at 15°.
These values are similar to the performance of intact mammalian muscle
fibers, indicating that the energy transduction mechanism is preserved
through removal of the membrane and cold storage.
Single fibers from rabbit and frog muscle are also permeabilized by
treatment with detergents 34or by removal of the sarcolemma by mechanical
dissection. 35 Insect fibrillar flight muscle fibers are prepared by glycerol
extraction. 36Bundles of cardiac37or smooth muscle38cells are also amenable
32 y . E. Goldman, M. G. Hibberd, and D. R. Trentham, Z Physiol. 354, 577 (1984).

33 R. L. Moss, J. Muscle Res. Cell Motil. 3, 295 (1982).
34 L. C. Yu and B. Brenner, Biophys. J. 55, 441 (1989).
35 y . E. Goldman and R. M. Simmons, J. Physiol. 350, 497 (1984).
36 M. Yamakawa and Y. E. Goldman, J. Gen. Physiol. 98, 657 (1991).
37 H. Martin and R. J. Barsotti, Biophys. J. 66, 1115 (1994).
38 K. Horiuti, A. V. Somlyo, Y. E. Goldman, and A. P. Somlyo, J. Gen. Physiol. 94, 769
(1989).
312 CAGZDCOMPOUNDS [ 181
750 pm
Rg~ ,ira I
1
Teflon
;arriers
I-used ~lllCa ,& ~unace

Window
i 8 rnm I I ,,."
FIG. 2. Mechanical arrangements for mounting single muscle fibers in a photolysis setup
and for exchanging solutions. (A) A T-shaped aluminum foil clip (-180/zg) used for attaching
the fiber to the setup. Tapered hooks on the force transducer and length-step motor pass
through the diamond-shaped hole in each T clip (C). (B) Single muscle fiber held in a T clip
near its end. The T clip is crimped onto the fiber away from the foil edges, causing the edges
to flare away from the fiber to avoid damage. (C) Temperature-controlled trough assembly
used for solution exchange. Solution troughs (60/zl) are milled into a stainless-steel block.
A chilled mixture of ethylene glycol and water flows through the hollow center of the block
to adjust the temperature. Slotted Teflon barriers are glued, with epoxy, to the ends of each
trough to retain the aqueous contents by surface tension. The "front trough" of the assembly
is the photolysis chamber made from an L-shaped section of a quartz cuvette attached under-
neath the block (see also Fig. 3). Laser light traversing the photolysis trough is reflected back
from the polished surface of the steel block to increase photolysis yield. Dashed lines indicate
the path of the trough assembly for solution exchange.
to photolysis experiments after permeabilization by detergents, glycerol, 39

or bacterial a-toxin. 4°
Because the functional output of a muscle is mechanical force and
shortening, attachment of its ends to the experimental setup must minimize
physical damage and extraneous mechanical compliance. A common
method is to fold T-shaped aluminum foil clips (Fig. 2A), made by photo-
lithography, directly around the fiber (Fig. 2B)Y A clip applied with the
optimum pressure holds the muscle fiber reliably during full contraction
while minimizing crushing of the muscle ends. A hole in the stem of the
39 A. V. Somlyo, Y. E. Goldman, T. Fujimori, M. Bond, D. R. Trentham, and A. P. Somlyo,

J. Gen. Physiol. 91, 165 (1988).
40 T. Kitazawa, S. Kobayashi, K. Horiuti, A. V. Somlyo, and A. P. Somlyo, J. Biol. Chem.
264, 5339 (1989).
[1 8] CAGED MOLECULES AND MOTOR PROTEINS 313
clips slips over the tapered tips of hooks attached to the mechanical trans-
ducers of the apparatus (Fig. 2C). Single fibers can also be mounted by
gluing the ends of the fiber segment to hooks or pins on the apparatus 41
or by using specialized mounting clamps. 42,43
Mechanical compliance at the fiber attachment points compromises
quantitative measurements of fiber elasticity and kinetics of mechanical
transients initiated by photolysis. 44 The series compliance can be reduced
by stiffening the ends of the fiber by fixation of a localized region with
glutaraldehyde and attaching the fixed regions to the setup. 45
Because cardiac and smooth muscle fibers are much smaller than skeletal
muscle fibers, photolysis experiments typically involve bundles of these
cells. Such preparations are attached to the experimental apparatus by -I"
clips, clamps, or glue. Substantial mechanical compliance between the cells
of these preparations cannot be eliminated by treating the ends of the
preparation. However, cardiac muscle is cross-striated, enabling monitoring
of extent of sarcomere length changes during photolysis-initiated transients.
The effects of series compliances can be estimated and taken into account
in smooth muscle experiments as well, but this is more difficult because
the direct measure of filament sliding velocity from the sarcomere length
is not available.
Experimental Apparatus
Apparatus for photolysis experiments on muscle fibers must enable
convenient, complete, and reliable exchange of media bathing the muscle,
optically couple the photolysis chamber to the ultraviolet (UV) photolysis
light source, and allow samples to be collected after photolysis for determi-
nation of the photochemical conversion of the caged compound. The vol-
ume of solution required to exchange or add the caged compound should
be minimized because caged compounds are often costly or time-consuming
to prepare. The ends of the muscle fiber must be available for attachment
to fixed supports or mechanical transducers. The requirements of the other
signals being recorded from the preparation often dictate some of the
mechanical and optical arrangements of the apparatus.
Two types of muscle fiber setups for photolysis experiments are shown
schematically in Figs. 2 and 3. The muscle fiber is typically connected to a
tension transducer at one end and a fixed support or a length driver (motor)
41 B. Brenner, Biophys. J. 41, 99 (1983).

42 y . Zhao and M. Kawai, Am. J. Physiol. 266, C437 (1994).
43 R. L. Moss, J. Physiol. 292, 177 (1979).
44 y . Luo, R. Cooke, and E. Pate, Am. J. Physiol. 265, C279 (1993).
45 p. B. Chase and M. J. Kushmerick, Biophys. J. 53, 935 (1988).
A
Force
Transducer
B C
I /~"~vacuation
>/ System
Muscle
"•e
/
J
r
Capillary
Cuvette
J
~
Muscle
Fiber
Capillary
Cuvette
Lamp Collimating X Secondary

Condensor Filter Focusing Lens ~ i o n
Motor Inlets
FIG. 3. Examples of experimental setups used in photolysis studies of single skeletal muscle
fibers or cardiac or smooth muscle fiber bundles. (A) Schematic side view of a frequency-
doubled ruby laser system, optics, and translating trough assembly. The ruby rod (1/2-inch
diameter) is optically pumped with a helical xenon flash lamp (not shown). The laser cavity
is formed by the partially reflecting front face of the ruby rod and a rear fully reflecting
mirror. A triggered Pockel cell and Brewster stack polarizer assembly (not shown) Q-switch
the laser resonator to output a 50-nsec "giant pulse" at 694 nm. Brewster stack (clean up)
polarizers and a red Wratten (Kodak) filter remove scattered pump light and unpolarized
laser fight. A temperature-tuned rubidium dihydrogen arsenate crystal or an angle-tuned
potassium dihydrogen* phosphate (KD*P) crystal serves as a harmonic generator to convert
the 694-nm laser fight to 347 nm radiation. A U350 glass filter (Hoya, Fremont, CA), a U G l l
filter (Schott, Duryea, PA), or a 1 M copper sulfate liquid filter after the harmonic generator
is used to block the 694-nm primary beam. A fused silica cylindrical lens focuses the 347-nm
beam to the height of the trough (3 mm). The working distance between the lens and the
trough is --100 mm. Masks on the lens or near the trough reduce the width of the beam to
the fiber length (2-3 mm). Four to six troughs are milled into the stainless-steel block (see
also Fig. 2). The quartz trough nearest the UV light source is illuminated from the front by
the photolysis pulse and from below by other light sources for structural signals such as
fluorescence polarization or light scattering (Table I). A microcomputer-driven stepper motor
lowers, translates, and raises the trough assembly to exchange solutions while the fiber remains
in place. (B) Schematic side view of a flash lamp and capillary flow system. Light from a
xenon flash lamp is shaped into a collimated 320- to 360-nm beam (1-inch diameter) using a
spherical mirror, fused silica condensor, and bandpass filter such as a U350 with an interference
at the other end (Figs. 2C and 3C). Solutions are exchanged either by a
flow arrangement (Fig. 3C) or by a mechanism to exchange solution troughs
while keeping the muscle fiber in place (Figs. 2C and 3A). The photolysis
trough is constructed of fused silica for transparency at the near-UV wave-
length (320-360 nm) used for photolysis. Often another optical system, such
as a fluorometer or a sarcomere length detector, uses the axis orthogonal to
the fiber and the photolysis optics.
Several techniques, listed in Table I, have been used in conjunction
with photolysis to determine the relationship among biochemical, mechani-
cal, and structural events in the actomyosin ATPase cycle. Mechanical
signals, such as force, stiffness, shortening velocity, and sarcomere length,
give direct macroscopic indications of the functional output. Biochemical
assays provide transient or steady-state information on progress of the
ATPase. The other signals are structural probes that indicate protein mo-
tions. Thus methods are available using photolysis of caged molecules to
elucidate the temporal and functional relationship among the enzymatic
turnover, mechanical steps, and structural changes within the contractile
proteins during energy transduction.
Photolysis Light Sources

Near-UV light sources used for photolysis with motor proteins are
listed in Table II. The wavelength and pulse duration are important
criteria for choosing a photolysis light source. The wavelength must be
matched to the photolysis action spectrum of the caged compound, but
as discussed in "Uniformity of Photolysis in the Sample Chamber," the
peak of the absorption spectrum is not always the best choice because
propagation of the light into the sample is limited due to absorption by
the caged molecule. Wavelengths above 310 nm are less injurious to
muscle fiber proteins than shorter wavelengths.
Pulse duration should be shorter than the intended time resolution
heat reflecting coating. The pulse is directed onto the muscle fiber cuvette by a fused silica
astigmatic secondary focusing doublet consisting of a planoconvex spherical lens and a cylindri-
cal lens. The working distance between the refocusing lens and the sample is 15-25 mm. (C)
Schematic top view of the capillary cuvette and solution exchange system used in a flash lamp
photolysis system available commercially from Scientific Instruments, Heidelberg, Germany.
The muscle fiber is wrapped around a hook on a force transducer and both ends are glued
to a pin on a length driver (motor) within a quartz capillary cuvette (1-mm inner diameter).
The cuvette is perfused through a solution inlet at the motor end. A seal to the motor pin
directs flow through the cuvette. Usually one of two inlets is dedicated to caged compounds.
Outflow is removed through an evacuated pipette.
316 CAGED CO~POUYOS [181
eo
-~.~
0
,.m
o ~o~
o ~.~.o ~ = ~ .- .-o
0"-~ ~ 0 ~ ~:~
)- .fix
o[- .=
0
~o o~'~
,< e> ~ o~
,.-a
0
t--q (-q
,..a
£
~a
.l
¢n
"~ 0
o
m
= o
,¢
~ ~ ~- o.
<
&
Z
em
..= ?.~ .gaa
.a
<
r/l
a:
o~ ~ ,~
eal)
b~
~, .~_
[18] C A G E D M O L E C U L E S A N D M O T O R PROTEINS 317
~ ~ °~
~ ' ~ -~
"~ ~ "~ ,~
['~ c* , . ~
< .~ ~-~
O ~ 0
.< •= ~ .©
m ° ° "
~ =
o ~~
~.~ o=~o= ~
~J O
. % ©
0 0 o
=o ..=
318 CAOZD COMPOU~OS [ 181
== =,
'-!.
,.a,
v
m ~ o~
i...-
~
m ~
,-~ -.~ "4
~ ~ ~' r~ ~ ~ 0~'"'~"0
;~ ~ ~ ~" ~ . . • ~ --~. . ~ .~..,
~ ~ ~ ~ 00 .~ . ~ ~ . ~.~ ~ ~ .~,
[18] C A G E D M O L E C U L E S A N D M O T O R PROTEINS 319
o~
tmN
¢,4
<
[-
0
0
Z eq
<
m
m
~' .'4 ~,~

Z
r - "4 ~ ~ =.=
M
<e
<
z
N
,-1
C x-
0m r~ t~
m
M
<
< .'~ ~ 0 ~'~ I ~

> .~z = ~,- ~ ~.,~
<
¢)
o
0
o
Z
C
<
z ~,~,~, = ~ -'~ ~-~
M
i=. •= ~ . ~ ~ ~.. ~i ~ .~ ~, ,ff ~"~ ,.~ .~ ~ o ~
;;~:=~=~o.
of the experiment, which is often limited by the dark photochemical

reaction rate (the rate of release of the active compound after the light
pulse, see Table III). If the caged compound has a low product quantum
yield (Qp, the proportion of optically excited caged molecules that pho-
tolyze), a theoretical advantage of longer light pulses is the potential for
repeated excitation of unphotolyzed molecules, thereby increasing the
extent of photolysis. 46 If the photolysis pulse duration is comparable to
the optically excited state, repeated excitation cannot occur. Lifetimes of
optically excited states of 2-nitrobenzyl compounds are subnanosec-
ond, 46'46a so the light sources in Table II have sufficient pulse durations
to allow multiple excitations.
Laser sources have the advantages of high power and simple optical
interface to the sample due to the monochromatic, collimated nature of
laser radiation. A simple fused silica lens shapes the beam and projects
it onto the sample (Fig. 3A). Many of the muscle fiber experiments have
used frequency-doubled ruby l a s e r s 27'46 o r frequency-doubled tunable dye
l a s e r s , 21'46 but these two sources are not readily available commercially.
In the trough of our muscle fiber setup (Fig. 3A), with a 3 x 8-mm
optical window and 1-mm path length, a 100-mJ pulse of 347 nm radiation
photolyzes -10% of 10 mM NPE-caged ATP to liberate - 1 mM ATP.
Shuttered continuous arc lamps or flash lamps (Fig. 3B) are also conve-
nient light s o u r c e s . 47 The choice between a laser and a flash lamp system
weighs considerations of wavelength, working distance, pulse duration,
energy, and cost. Flash lamp systems have the disadvantage that a relatively
high aperture secondary focusing lens is required to project the photolysis
light onto the fiber, reducing the working distance to typically 15-25 mm.
This constraint may limit access to the preparation by other mechanical or
optical components. However, the cost of a flash lamp system is much less
than high-powered lasers. Flash lamp systems for photolysis experiments
are described in more detail in another chapter in this v o l u m e . 48
Light sources for in vitro experiments are not as demanding as with
muscle fibers because the volume irradiated (30-tzm diameter field x 10
/xm-2 mm path length) is much smaller and focusing the photolysis light
through a microscope objective increases the power density in the small
photolysis volume. Low-power lasers are listed in Table II for in vitro experi-
ments.
46 j. A. McCray and D. R. Trentham, Annu. Rev. Biophys. Biophys. Chem. 18, 239 (1989).
46a S. R. Adams, J. P. Y. Kao, G. Grynkiewicz, A. Minta, and R. Y. Tsien, Z Am. Chem. Soc.
110, 3212 (1988).
47 G. Rapp and K. Gtith, Pfliigers Arch. 411, 200 (1988).
48 G. Rapp, Methods Enzymol. 291 [11] (1998) (this volume).
Caged Compounds
Table III lists the principal caged components used to study cell motility
and some of their properties. They are available commercially except where
indicated. 1-(2-Nitrophenyl)ethyl (NPE) esters of nucleotides and phos-
phate were the first caged compounds used in muscle fiber studies. They
have high product quantum yields (Qp) but the leaving group, 2-nitrosoace-
tophenone, is reactive and forms covalent adducts with reactive sulfhydryls,
such as protein cysteine residues. Thus, reduced glutathione (GSH), dithio-
threitol (DTT), or 2-mercaptoethanol must be included in the photolysis
media to protect the contractile apparatus, as discussed later.
Typical samples of NPE-caged nucleotides are mixtures of diastereo-
isomers due to a chiral center at the benzylic carbon of the 1-(2-nitrophe-
n y l ) e t h y l g r o u p . 46'48a The R- and S-diastereoisomers of NPE-caged
nucleotides are separable using reversed-phase chromatography, but the
separation does not confer any advantage because the diastereoisomers
have the same properties in photolysis experiments on muscle fibers. 49
Desyl (diphenyloxoethyl) 5° and dimethoxybenzoin 51 derivatives have
faster dark reactions than the NPE compounds and the benzofuran pho-
toproducts are less reactive. However, the quantum yield of some of the
derivatives is lower and they are less stable in the dark. 5°-52
NP-EGTA and DM-nitrophen are Ca 2+ chelators that photolyze to
iminodiacetic acids that have Ca 2+ affinities several orders of magnitude
lower, thus releasing bound Ca 2+. DM-nitrophen is a derivative of EDTA
and binds Mg 2+ quite tightly (Table IV), enabling it to serve as a caged
Mg 2+ or else necessitating low free Mg 2+ concentrations when used as a
caged Ca 2+. NP-EGTA is selective for Ca 2+ (Table IV). Nitr-5 and Nitr-7
are highly selective Ca 2+ compounds that photolyze at longer wavelengths
(>360 nm), but they have much lower quantum yields than NP-EGTA
or DM-nitrophen. DMNPE-caged BAPTA and Diazo-2 are caged Ca 2+
chelators used to abruptly reduce free [Ca2+].
Effects of impurities or contaminants should be considered in all samples
of caged compounds. Muscle fibers are particularly sensitive to contaminant
nucleotides because actomyosin has high affinity for ATP and ADP. For
4Sa j. E. T. Corrie, G. P. Reid, D. R. Trentham, M. B. Hursthouse, and M. A. Mazid, J. Chem.

Soc. Perkin Trans. I, 1015 (1992).
49 H. Thirlwell, J. A. Sleep, and M. A. Ferenczi, J. Muscle Res. Cell Motil. 16, 131 (1995).
50 R. S. Givens, P. S. Athey, B. Matuszewski, L. W. Kueper III, J. Xue, and T. Fister, Z Am.
Chem. Soc. 115, 6001 (1993).
5t j. E. T. Corrie and D. R. Trentham, J. Chem. Soc. Perkin Trans. I, 2409 (1992).
52 H. Thirlwell, J. E. T. Corrie, G. P. Reid, D. R. Trentham, and M. A. Ferenczi, Biophys. J.
67, 2436 (1994).
© ~ "0
~m ~ ~~ 2 o~ ~' ~- - ~
,._, .o 8~
N 0 + +
~o
o
0 "
= 0 0=---o=
"0
0 < < ~ -+ ~* .o
e.., m n=l 0
~.-~ dd
".o ~ ~, +
Z 0~ 0 0J ~ ~ ~1 m
o
oZ ',= o=Z
z Z Z N ~.
0 ~ t"-
o o . o o ~
0
0
~o~
r-4 X"~A ',~
t'q e¢~
m
< ,,n ,..., u.!u.-~t-,a~
"~''~'":::1-
'0
0
.o< ~ <
8 ~,~ ~o1~ ,~1 ,.~I ..0 ' ~ ' ~
"0
ezz z ezeee~-
f,.)
rj ..-k ,,~,
mz~. .+-.k
oo
Qo
~'~
~zag
N
~o
~'~
~uN
~+
.,.,~ Q+
,j •
t-xl
~5
,+Mm ~- +-
..4 ~,.~ ~ . ,,._+ ~.-],+.-,~ oo ,=_+
zm+~
+-,. ~2 m ~. + m m r.. ~__.
X i~ • ~ - • "-4
¢+4 ~
N~ 8 ~ M o
0 ,.~ - ,
m~ ~'m'+ B B,,.,; =_+

~~
~ , , ~ = • ~-~., .+',~"~
•r. • ~ c N ~ d ~" ~ ~ - ~"- "~I +-+

.=<
,+..) 0.) ~ . ;- ,,-,+ - ~ +"u
~,,.~ o &++m = om
~'~ ~ + . ~ - ~ u ~:.+3~ +
m "a ~ - .0 +~:~_.. + ,u , ' ~ m
~ .~ • . C~ ~ ,-,,-~ m c~ ,.~ +,,,~ +-~
m~N~ ~ ~.~.~-.~ ~
<
"~ ~ .?~,
Z
324 CAGED COr~POUNDS [ 181
T A B L E IV
APPARENT pica VALUESa FOR BINDING OF Ca 2+ AND Mg 2+
Before
photolysis
After photolysis b
Caged compound Ca 2+ Mg 2+ Ca 2+
NPE-caged A T P ' 2.52 2.45 3.65

DM-nitrophen d'e 8.30 5.65 0.21
N P - E G T A e'y 6.93 2.40 0.11
Nitr-5g 6.84 2.07 2.22
Nitr-7 g 7.30 2.27 2.07
diazo-2 h 2.66 -- 7.14
Apparent pKa values for NPE-caged ATP and N P - E G T A

were calculated for pH 7.1. All others were measured at
p H 7.2.
b Most apparent pKa values for Mg 2+ binding after photolysis
are not available.
c Apparent pK~ values for NPE-caged ATP were determined
as described in the text. Values were adjusted to p H 7.1.
d j. H, Kaplan and G. C. R. Ellis-Davies, Proc. Natl. Acad. Sci.
U.S.A. 85, 6571 (1988).
Apparent pKa values for the photolysis by-product are for
iminodiacetic acids.
f G. C. R. Ellis-Davies and J. H. Kaplan, Proc. Natl. Acad. Sci.
U.S.A. 91, 187 (1994).
g S. R. Adams, J. P. Y. Kao, G. Grynkiewicz, A. Minta, and
R. Y. Tsien, J. Am. Chem. Soc. 110, 3212 (1988).
h S. R. Adams, J. P. Y. Kao, and R. Y. Tsien, J. Am. Chem.
Soc. 111, 7957 (1989).
example, A D P runs very close to NPE-caged ATP in anion-exchange chro-

matography, so it is a common contaminant. With only 1% A D P contamina-
tion, a photolysis solution with 10 mM caged ATP contains 100/~M ADP,
in the range of the 20-200/~M dissociation constant for A D P binding to
actomyosin. 2z'53-55 Contaminant A D P at this level would markedly inhibit
the rate of binding of photoreleased ATP to actomyosin and the ATP-
induced dissociation (step 1, Fig. 1). 52 Purity of caged nucleotides is evalu-
ated using high-performance liquid chromatography (HPLC). 32,46,52A col-
orimetric assay is used to determine phosphate (Pi) contamination in
caged Pi .21
53 S. Marston, Biochim. Biophys. Acta 305, 397 (1973).

54 j. A. Dantzig, M. G. Hibberd, D. R. Trentham, and Y. E. Goldman, J. Physiol. 432, 639 (1991).
55 M. Schoenberg and E. Eisenberg, J. Gen. Physiol. 89, 905 (1987).
Due to the severe requirements for purity of caged ATP, we synthesize

and purify NPE-caged ATP in the laboratory using 1-(2-nitrophenyl)-di-
azoethane to modify ATP according to a procedure modified from that of
Walker et aL 56'56a Purification by alternating ion exchange (Whatman DE-
52, DEAE) and reversed-phase (Whatman C18, /xBondpack, bead size
50-100/xm) HPLC chromatography (totaling four column runs) reduces
nucleotide contamination to <0.02%. 54 The purified caged compound is
divided into aliquots for single experiments, rapidly frozen using liquid
nitrogen or a mixture of 2-propanol and dry ice, and then stored at - 8 0 °.
Considerations of purity for caged Ca 2+ compounds include contaminant
Ca 2+ and Mg 2+. The concentration of Ca 2+ that brings a caged Ca (e.g.,
NP-EGTA) solution to the threshold for activation of muscle fiber force
can vary considerably among batches of nominally the same chemical. This
variability is presumably an indication of Ca 2+ contamination or impurities
in the caged chelator stock.
Experimental Solutions
Experimental solutions for muscle fiber experiments are designed to
mimic the intracellular milieu in ionic strength, pH, free Ca 2+, free Mg 2+, and
nucleotides. Buffering systems should minimize changes in these parameters
that occur on photorelease of the active experimental compound. Because
these ionic and biochemical factors all influence the performance of the
fully constituted and regulated contractile system, the design of the solutions
must take account of the multiple metal-ligand interactions among the
constituents and regulate the total ionic strength, F = (Ei CiVi2)[2, where
Ci is the concentration and Vi is the net charge of each charged species in
the solution. The interaction of myosin with actin is partly ionic, so the
affinity is strongly affected by F.
Computer programs that analyze or design solutions to specified concen-
trations and ionic strength have been described. 57'58 We use a program
originally written by J. Thorson, D. C. S. White, and Y. E. Goldman and
then upgraded to a modern user interface by R. J. Barsotti and A. Fielding
(personal communication). It uses stability constants for each of the metal-
ligand complexes in the solution to solve the multiple binding equilibria.

7170 (1988).
56a j. W. Walker, G. P. Reid, and D. R. Trentham, Methods Enzymol. 172, 289 (1989).
57 A. Fabiato and F. Fabiato, J. PhysioL (Paris) 75, 463 (1979).
58 Commercially available software for determining solution constituents: Max Chelator ver-
sion 4.61 supplied by C. Patton, Hopkins Marine Station, Stanford University, Pacific
Grove, California.
TABLE V
COMPOSITIONS OF SOLUTIONS USING CAGED ATP AND CAGED CALCIUM DURING PHOTOLYSIS TRIALSa
Caged
Solution MgC12 ATP Ca 2+ EGTA HDTA Na2CP compound
High ATP relax 7.7 5.4 -- 25 -- 19.1 --

Prerigor relax 2.7 0.12 -- 30 -- 21.6 --
Rigor 3.1 -- -- 53.4 -- -- --
Ca rigor 1.2 -- 20 20 34.3 -- --
Rigor + NPE-caged ATP 4.6 -- -- 35.9 -- -- 10
Ca rigor + NPE-caged ATP 3,2 -- 36.5 36.5 -- -- 10
Pre-NP-EGTA 6,3 5.4 -- -- 48.1 -- --
NP-EGTA 6,5 5.4 1.256 -- 46.0 20 2.0
ATP, adenosine 5'-triphosphate; EGTA, ethyleneglycol-bis(/~-aminoethyl ether)-N,N,N',N'-tetra-

acetic acid; HDTA, hexamethylenedinitrilotetraacetic acid; TES, N-tris(hydroxymethyl)methyl-2-
aminoethanesulfonic acid; CP, phosphocreatine; NP-EGTA, nitrophenylethyleneglycol-bis(/3-
aminoethyl ether)-N,N,N',N'-tetraacetie acid. Solution constituents were designed using computer
software originally written by J. Thorson, D. C. S. White, and Y. E. Goldman and upgraded by
R. J. Barsotti and A. Fielding. All values are total concentrations in mM. All solutions contain
100 mM TES and 10 mM reduced glutathione. Ionic strength of the solutions is 200 mM, pH 7.1,
20°. The K + concentration is 120-140 mM and the Na ÷ concentration is 10-30 raM. For photolysis
solutions, caged compounds, concentrated stocks, and H20 are combined on the day of an experi-
ment. Free [Mg2÷] is calculated to be 1 mM in all solutions. Free [Ca 2÷] is calculated to be ~30
/zM in the Ca rigor and Ca rigor + NPE-caged ATP solutions. The NP-EGTA photolysis solution
has 1.256 mM Ca added and 250 nM calculated free Ca 2+. All other solutions contain <10 -s M
free Ca 2÷. Creatine phosphokinase (1 mg/ml, 100-150 units/ml) is added to all solutions with
creatine phosphate.
G i v e n t h e d e s i r e d f r e e m e t a l ( C a 2÷, M g 2÷) c o n c e n t r a t i o n s , f r e e , b o u n d , o r
total concentration of each anion, pH, temperature, and total ionic strength,
the program calculates the total concentration of each constituent and a
r e c i p e f o r m a k i n g t h e s o l u t i o n f r o m g i v e n s t o c k s o l u t i o n s ( T a b l e V). T h e
m e t a l - l i g a n d s t a b i l i t y c o n s t a n t s u s e d in this c a l c u l a t i o n a r e m o s t l y o b t a i n e d
f r o m p u b l i s h e d d a t a , t a b u l a t e d in M a r t e l l a n d S m i t h . 59
C a 2÷ a n d M g a÷ affinities f o r c a g e d A T P a n d t h e c a g e d c a l c i u m c o m -
p o u n d s a r e l i s t e d in T a b l e I V . M e a s u r e m e n t s o f C a 2÷ a n d M g 2+ b i n d i n g t o
N P E - c a g e d A T P a r e d e s c r i b e d l a t e r in this article. C a g e d A T P h a s a v a l e n c e
o f - 3 at n e u t r a l p H a n d a p K a o f 4.23 f o r p r o t o n b i n d i n g t o f o r m H - c a g e d
A T P 2- (J. W . W a l k e r , p e r s o n a l c o m m u n i c a t i o n ) , p K a v a l u e s f o r N P - E G T A
a r e 9.95, 7.41, 2.64, a n d 2.0 ( G . C. R. E l l i s - D a v i e s , p e r s o n a l c o m m u n i c a t i o n ) .
For the photoproducts of NP-EGTA, we use pKa values for iminodiacetic
acid, 9.45 a n d 2.4. 59 T h u s t h e p h o t o p r o d u c t s h a v e a v a l e n c e o f - 1 at n e u -
tral pH.
59 A. E. Martell and R. M. Smith, "Critical Stability Constants." Plenum Press, New York, 1974.
Examples of solutions used for single muscle fiber experiments with

caged ATP and NP-EGTA-caged Ca 2+ are listed in Table V. Design criteria
for these solutions are 1 mM free Mg 2+, 200 mM total ionic strength, pH
7.1 at 20 °, and full activation either by photolysis of caged ATP from rigor
(no nucleotide) or by photolysis of NP-EGTA from full relaxation (low
[Ca2+], + MgATP). Creatine phosphate (CP) and creatine phosphokinase
(CPK) buffer the concentration of ATP by rephosphorylating ADP pro-
duced by the contractile proteins. Ca 2+ is buffered by EGTA. pH is buffered
by TES buffer.
Caged compounds contribute significantly to ionic strength. For in-
stance, the - 3 valence of caged ATP at pH 7.1 implies that a 10 mM
concentration with its 30 mM associated cation (TEAB+; triethylamine
bicarbonate salt) contributes 60 mM to F [(10 mM × 32 + 30 mM)/2].
Therefore, to keep F constant when the caged compound is included in a
solution, other constituents must be reduced.
A fast pH buffer with a pKa near the working pH limits changes of pH
on photolysis of the caged compound. The pH buffering capacity is maximal
when pH = pKa. BES buffer 21 (2-[bis(2-hydroxyethyl)amino]ethanesul-
fonic acid with pKa of 7.12 at 20 ° and MOPS buffer [3-(N-morpholino)pro-
panesulfonic acid], pKa of 7.2 at 20 °, are potential choices. We prefer TES,
pK, of 7.5 at 20°, for the following reason: the caged molecules and other
constituents required to buffer ATP and Ca 2+ make considerable contribu-
tions to F. The amount of pH buffer that can be added to the solution is
limited by the requirement to keep total F within 200 raM. Therefore, a
figure of merit for the pH buffer is the buffering capacity per unit contribu-
tion to F. When pK, is somewhat greater than pH, more than half of the
buffer molecules are liganded with protons and do not contribute to F. At
pH 7.1, only 25% of TES molecules are charged, leading to high buffering
capacity per unit contribution to the ionic strength.
The species of anions present in the solution affect the longevity of the
muscle fiber preparation. In solutions with high concentrations of chloride,
the strength of contraction markedly diminishes after a few activations. 6°
This decline is less prominent with propionate and acetate. 6° In the solutions
shown in Table V, the main anions are EGTA, HDTA, CP, and caged
ATP, which do not seem to be deleterious. 32'6°'6x
GSH or DT-I" is added to the photolysis solutions to decrease damage
to the contractile system from reaction with the photolysis leaving group,
2-nitrosoacetophenone. This by-product of NPE-caged compounds reacts
60 M. A. W. Andrews, D. W. Maughan, T. M. Nosek, and R. E. Godt, J. Gen. Physiol. 98,

1105 (1991).
61 D. G. Moisescu, Nature 262, 610 (1976).
with free sulfhydryls, such as cysteine residues in the proteins. In early
caged ATP experiments using NPE-caged ATP in the absence of Ca 2÷,
photoreleased ATP failed to relax rigor fibers when GSH was not present.
The concentration of ATP required to relax the fiber was irreversibly
increased, indicating modification of the contractile apparatus by the pho-
tolysis by-products. When sulfhydryl reagents such as GSH or DTT are
included in the photolysis medium in excess, these compounds react with the
photoreleased by-product, thereby avoiding modification of the proteins. 61a
The second order rate constant for reaction of 2-nitrosoacetophenone with
DTT is 3.5 x 103 M -I sec -1 at pH 7.0, 21 °, and 0.18 mM ionic strength, s6
At 10 mM DTT, the time constant for removal of the reactive by-product
is thus ~- = 1/(10 mM × 3.5 × 103 M -1 sec -a) -- 29 msec.
The concentration of caged compound added to the solution is based
on several factors. The desired amount of biologically active compound
photoreleased depends on the number of photons absorbed during the flash
and Qp for photolysis, as discussed later. Other factors that bear on the
amount of caged compound to be added to the medium are spatial unifor-
mity of photorelease and effects of the caged compound on the biological
preparation before photolysis. Caged compounds bind to the contractile
proteins, generally with lower affinity than the intended photoproduct,
and either activate or inhibit the reaction. Uniformity of photolysis and
inhibition of motor proteins by NPE-caged ATP are discussed later.
In some experiments, it is necessary to reduce nucleotide concentrations
before photolysis below the levels caused by nucleotide contamination of
the caged compound. Glucose (200 mM) and hexokinase (5-10 U/ml), 54
diadenosine pentaphosphate (250 / x M ) , 54 o r apyrase (10-50 U / m l ) 37'52'62
have been used for this purpose in muscle fiber experiments. Apyrase
hydrolyzes adenine nucleotides to adenosine and Pi. Apyrase coupled to
CNBr-activated Sepharose beads retains activity and can be removed from
a treated solution by centrifugation, s2 Apyrase has also been useful in
removing tightly bound nucleotides from the contractile proteins. It diffuses
rapidly into muscle fibers, but washes out poorly. 37,62
Given the starting concentrations of the solution constituents and the
amount of photolysis, the final concentrations of ionic species and bound
complexes can be calculated. Table VI shows the changes expected for
typical NPE-caged ATP and caged Ca 2+ experiments. For the experiment
starting with 10 mM caged ATP, assuming 1 mM total ATP is photore-
leased, 0.9 mM ATP complexes Mg 2÷ and is available for binding to the
61a A. Barth, J. E. T. Corrie, M. J. Gradwell, Y. Maeda, W. M~intele, T. Meier, and D. R.

Trentham, J. A m . Chem. Soc. 119, 4149 (1997).
62 H. Martin and R. J. Barsotti, Biophys. J. 67, 1933 (1994).
TABLE VI
CHANGES IN SOLUTION CONSTITUENTS FOLLOWING THE PHOTOLYSIS OF CAGED A T P
OR NP-EGTA
Constituent a Before photolysis After photolysis
Activation solution with NPE-caged ATP

ATP (total) 0 1.0
Free Mg2+ 1.0 0.74
Free Ca 2+ 0.053 0.056
Mg-ATP 0 0.9
Caged ATP (total) 10.0 9.0
Mg-caged ATP 2.17 1.54
pH 7.10 7.05
Ionic strength 200 199
Activation solution with NP-EGTA
ATP (total) 5.39 5.39
Free Mg2+ 1.0 1.28
Free Ca 2+ 250 nM 0.023
Mg-ATP 5.0 5.005
NP-EGTA (total) 2.0 1.2
Ca-NP-EGTA 1.26 1.19
Mg-NP-EGTA 0.15 0.001
IDA (total) 0 1.6
Ca-IDA b 0 47 nM
Mg-IDA 0 0.005
pH 7.1 7.09
Ionic strength 200 199
a All concentrations are in mM except where indicated. The total concentrations of the
constituents before photolysis are given in Table V.
b IDA, iminodiacetic acid.
contractile proteins. The free [Mg 2÷] declines from 1 mM starting concentra-
tion to 0.74 mM. Free Ca 2+ changes only slightly; ionic strength decreases
by i mM. Protons (1 mM) liberated on photolysis are mostly buffered and
the pH changes from 7.1 to 7.05. Heating of the solution caused by absorp-
tion of the laser pulse is less than 2 °.
A similar calculation for a typical NP-EGTA caged Ca 2÷ experiment
gives a final flee [Ca2+] of 23/xM, corresponding to pCa of 4.6. Free [Mg 2+]
increases from 1.0 to 1.28 mM due to release from the NP-EGTA. Ionic
strength decreases by 1 mM, and pH decreases 0.01 unit. The changes in
concentrations other than the intended increases of ATP and Ca 2÷ in these
examples are minimal, suggesting that the transient phenomena induced
by the photolysis pulse are due primarily to the released ATP or Ca 2÷, not
to incidental changes or by-products of the photolysis reaction. Control
experiments liberating ineffective compounds, such as Pi in rigor, support

this conclusion.63
Probing the Actomyosin ATPase in Fibers
A TP Binding, Cross-Bridge Detachment, and Relaxation of Tension

To determine the second-order rate constant for ATP binding and
detachment of myosin from actin, the muscle fiber is started from rigor
with all of the myosin heads in the AM nucleotide-free state. ATP released
from caged ATP binds to the myosin heads, promoting detachment and then
relaxation of tension and stiffness in the absence of Ca 2+ or development of
active tension in the presence of C a 2+. The protocol to induce rigor for a
caged ATP photolysis experiment with a skinned psoas fiber is to start with
the fiber relaxed at high ATP (Table V) and incubate for >5 min in prerigor
solution (0.1 mM ATP). The fiber is then transferred to the rigor solution
for - 3 0 sec and then to a second rigor wash to reduce the ATP concentration
around the fiber rapidly. Tension and stiffness increase in the rigor solution
and stabilize after 2-3 min.
Inhibitors of force such as Pi2°'21 or butanedione monoxime ( B D M ) 64
improve the uniformity of striation spacing by reducing filament sliding as
the fiber goes into rigor. The tension is very low when rigor is induced in
the presence of Pi, BDM, low Mg2+ concentration, or low temperature.
After the ATP is removed and the fiber becomes mechanically stable, it
can be stretched slightly if necessary to increase tension. The solution is
then exchanged for a rigor solution with 2-10 mM caged ATP and >2 rain
are allowed for the caged ATP to equilibrate in the fiber interior before
the photolysis pulse is triggered.
Photoreleased ATP complexes Mg2÷ quickly and then is available for
binding to actomyosin. Following the release of ATP into the rigor fiber
in the absence of Ca 2+, tension and stiffness decline biphasically with NPE-
c a g e d A T P 32,5z (Fig. 4A) and monotonically with DMB-caged ATP in the
presence of apyrase. 52 The rate at which tension and stiffness fall depends
on the concentration of ATP released, which is adjusted by varying the
laser pulse energy or by altering the concentration of caged ATP. These
data provide estimates of the second-order rate constant for ATP binding
and ATP-induced dissociation of actomyosin.32 Methods to analyze the
tension transients in these experiments can be found in Goldman et al. 32
and Sakoda and Horiuti.65
63 j. W. Tanner, D. D. Thomas, and Y. E. Goldman, J. MoL Biol. 223, 185 (1992).

64 H. Higuchi, T. Yanagida, and Y. E. Goldman, Biophys. J. 69, 1000 (1995).
65 T. Sakoda and K. Horiuti, J. Muscle Res. Cell Motil. 13, 464 (1992).
A B
1 a~\
Ca2+k nb ~] ~
~.. RelativeStiffness ^ 2+
ATP// +Ca2+
z [---~ Tension -Ca2.
L I ! !
200 ms 200 ms
FIG. 4. Mechanical recordings from single glycerol-extracted fibers relaxed and activated
by photolysis of caged ATP or caged calcium. (A) Transient changes in tension (lower traces)
and 500-Hz in-phase stiffness (upper traces, stiffness relative to that in the rigor state where
all of the myosin heads are attached to actin). Following liberation of --1 mM ATP from
NPE-caged ATP (arrow), the fiber relaxes ( - C a 2+) or isometrically contracts (+Ca2+). The
fiat lines show the baseline tension and stiffness in the absence of Ca 2+. Solution conditions
are given in Table V. (B) Transient changes in tension during the activation of a single fiber
from the relaxed state by photolysis (arrow) of NP-EGTA [trace a in (B), see Table V
for solution constituents] and DM-nitrophen [trace b]. (B) was modified from Fig. 6 in
G. C. R. Ellis-Davies and J. H. Kaplan, Proc. Natl. Acad. Sci. U.S.A. 91, 187 (1994). Copyright
(1994) National Academy of Sciences, U,S.A. The total ATP (3 mM) and Mg (1.2 mM)
concentrations for DM-nitrophen were lower than for NP-EGTA, but the time courses and
plateau tensions were very similar.
Activation by Photolysis of Caged A TP and Caged Calcium

The fiber is put into rigor as described earlier, then rigor with - 3 0 / z M
free Ca 2+, and, finally, rigor solution with Ca 2+ and caged ATP (Table V).
Two washes of rigor solution with Ca 2+ are necessary to ensure that the
Ca 2+ regulatory system is saturated with Ca 2+. Release of ATP causes a
transient dip in tension followed by the development of full active isometric
tension (Fig. 4A). 66 This tension increase reflects the attachment and force
generation of partly synchronized cross-bridges, so it has been combined
with many other signals that indicate biochemical or structural changes in
the contractile proteins (Table I). After steady tension is reached, the fiber
is transferred to relaxing solution and the procedure can be repeated 10-20
times with active tension declining typically to 80% of the original value.
Creatine phosphate and creatine phosphokinase can be added to the
photolysis medium to buffer the ATP concentration and keep the A D P
concentration low during the steady contraction initiated by photolysis.
However, very pure caged ATP is required (---0.02% contamination) to
66 y. E. Goldman, M. G. Hibberd, and D. R. Trentham, J. Physiol. 354, 605 (1984).

332 CAGEDCOMeOUNDS [181
obtain reliable results in the presence of CP because ADP present at
submicromolar concentration is converted by CPK to ATP, which causes
slow cross-bridge cycling.
In caged ATP experiments, the Ca 2+ regulatory system is fully switched
on and the kinetics are determined by the chemomechanical events of
actomyosin. Kinetics of the Ca 2+ regulatory system become relevant when
caged Ca 2+ is used. For activation of a fiber by photolysis of NP-EGTA,
the fiber is initially in relaxing solution and is incubated in two exchanges
of pre-NP-EGTA solution (Table V) containing low Ca 2+ buffering capacity
and then the photolysis solution containing 2 mM NP-EGTA and typically
1-1.8 mM Ca 2+. The amount of Ca 2+ in the photolysis solution is adjusted
to poise the fiber just below the threshold for tension development. Photoly-
sis of the NP-EGTA or DM-nitrophen rapidly activates full tension develop-
ment (Fig. 4B).
A TPase Rate
The overall turnover rate of the actomyosin ATPase in skinned muscle

fibers has been measured in steady contractions by linked enzymatic
assays67'68or by accumulation of the product ADP using HPLC. 69 Analysis
of the rate of ATPase or product release shortly after initiation of contrac-
tion by photolysis of caged molecules identifies transient biochemical steps
of the reaction pathway that are faster, or earlier, than the rate-determining
step of the cycle. Ferenczi et al. 7° rapidly froze muscle fibers shortly after
activation by photolysis of caged [3H]ATP and determined the production
of [3H]ADP by HPLC fractionation followed by liquid scintillation count-
ing. They identified a burst of ADP production followed by a steady-state
rate of - 2 sec -1 per myosin head.
A fluorescent-labeled Pi-binding protein incorporated within the muscle
fiber has been used to monitor Pi release after photolysis of caged ATP
and caged Ca 2+ in fibers. 68 The rate of Pi release was markedly higher
during the first several turnovers of the ATPase than in other studies. The
basis of this discrepancy remains unclear since the signals in these studies
were carefully calibrated.
A third method for transient ATPase measurements, described here,
is much simpler than freeze quenching or fluorescent product-binding pro-
67 E. J. Potma and G. J. M. Stienen, J. Physiol. 496, 1 (1996).

68 Z.-H. He, R. K. Chitlingworth, M. Brune, J. E. T. Corrie, D. R. Trentham, M. R. Webb,
and M. A. Ferenczi, J. Physiol. 501, 125 (1997).
69 M. G. Hibberd, M. R. Webb, Y. E. Goldman, and D. R. Trentham, J. Biol. Chem. 260,
3496 (1985).
70 M. A. Ferenczi, E. Homsher, and D. R. Trentham, J. Physiol. 352, 575 (1984).
40
"E IATP 0 CP
[ _ ~ e n s i o - f f - - - - ts 20
"" 5 CP Out-of-phase (s) 10
~-_ ~ Stiffness (Q)
0
L ,lOs 0 5 10 15 20
[CP] (mi)
FIG. 5. Determination of the ATPase rate in a single fiber. Transient tension (upper traces)
and quadrature (out-of-phase) stiffness (Q, lower traces) were recorded during activation
from rigor by photolysis of caged ATP in the presence of 32/zM free Ca 2+, 5 mg/ml (1500
U/ml) creatine phosphokinase, and 0 or 5 m M creatine phosphate (CP), temperature 10°. At
the arrow, 1.7 m M ATP was liberated from NPE-caged ATP. As the CP was depleted, ATP
was no longer regenerated from ADP and the fiber returned to rigor (declining phase of Q).
The duration of the steady-state activation (&) was determined from the intersection of lines
fitted to the steady and declining phases of Q. The ATPase rate is given by the reciprocal of
the slope of a line fitted to ts data plotted against the initial CP concentration.
teins, but with a lower time resolution. The muscle fiber is loaded in rigor
with 30/zM Ca 2÷, 5 m M caged ATP, 1500 U/ml CPK, and various selected
CP concentrations. The fiber is transferred to a photolysis trough filled
with silicone oil so that ATP is photoreleased only within the fiber. A
photolysis pulse initiates isometric contraction of the fiber and the conse-
quent ATPase activity. The CPK reaction buffers the ATP concentration
until the CP is depleted, at which point the fiber returns to rigor. 71'72Figure
5 shows recordings of tension and quadrature stiffness (Q) 90 ° out-of-phase
with an imposed 500-Hz length change on a skinned psoas fiber. 7z The
quadrature stiffness is 0.1-0.2 of the in-phase stiffness during the initial
contraction and then declines to zero when the CP runs out. Lines are fit
to the steady-state and declining phases of Q to determine the duration of
the steady-state contraction (ts). On a plot of ts vs [CP] (Fig. 5), the reciprocal
of the slope of a line fitted to the data points gives the ATPase rate, which
is constant from 0 to 20 m M CP. The mean ATPase rate in this assay
was 0.355 _+ 0.164 m M see -I (mean _+ SD, n = 16, 71) at 10°. Taking the
concentration of myosin heads in the muscle fiber7° to be 154/zM and the
space available to CP within the fiber to be 0.84 of the fiber volume, 73 the
ATPase rate is calculated to be 1.93 sec -1, in agreement with Ferenczi et
al. 7° and Hibberd et aL 69 No initial burst of ATPase activity is detected
with this method, but that may be due to its limited sensitivity.
71 A. Ishijima, H. Kojima, H. Higuchi, Y. Harada, T. Funatsu, and T. Yanagida, Biophys. J.

70, 383 (1996).
72 C. S. Cook, H. Higuchi, and Y. E. Goldman, Biophys. J. 70, A291 (1996).
73 I. Matsubara, Y. E. Goldman, and R. M. Simmons, Z MoL Biol. 173, 15 (1984).
Shortening Velocity and Interaction Distance
To initiate steady shortening by photolysis of caged ATP, the fiber is
put into rigor, Ca 2+ rigor, and then Ca 2+ rigor with caged ATP as described
earlier. Photolysis of caged ATP causes cross-bridge detachment, reattach-
ment, and initial redevelopment of isometric tension. For a shortening
velocity experiment, when tension reaches a preset threshold below the
isometric force level, the feedback controlling the motor switches to clamp
tension constant. The fiber shortens at the preset isotonic tension until an
electronic or mechanical stop is reached that limits the amount of shorten-
ing. Tension redevelops and the fiber is then relaxed. Curves of velocity
vs force or velocity vs [ATP] are assembled from shortening velocities in
a series of such trials. For isotonic force levels below the steady rigor
tension, the fiber is released slightly or the fiber is put into rigor in the
presence of BDM to lower rigor tension below the preset isotonic level.
A variation of the steady velocity experiment is to transfer the fiber to
a trough containing silicone or paraffin oil before photolysis. TM In this case,
the caged ATP is contained within the fiber and a limited, predetermined
quantity of ATP is released by the photolysis pulse. The isotonic shortening
continues only as long as the ATP remains, after which the fiber returns
to rigor. From the amount of shortening (Ds), the ATP concentration
utilized by actomyosin ([ATPu]), and the average stiffness during the iso-
tonic phase (S,), the interaction distance (i.e., the isotonic sliding distance
per ATP molecule hydrolyzed by myosin heads while interacting with actin
filaments, Di) can be calculated according to Di = DsSa [Mo]/[ATPu], where
[Mo] is the concentration of myosin heads within the muscle fiber. 7°'74 At
high velocity (low load), Di is surprisingly high (->60 nm) relative to the
length of the myosin head (16 nm), leading to the suggestion that myosin
heads interact with actin multiple times for each ATP molecule hy-
drolyzed. TM
Application of Caged ATP to /n Vitro Motility Systems

Development of assays of motor protein activity reconstituted in vitro
from purified proteins has contributed extensively to recent progress in
understanding the function and biophysical characteristics of these impor-
tant enzymes. For actomyosin-based motility and microtubule-based dynein
and kinesin motility, assays of motor protein sliding velocity, the unit dis-
tance traversed per biochemical cycle (termed step size) and force produc-
tion by one or a few molecules have been measured. 71,75-81 Photolysis of
74H. Higuchi and Y. E. Goldman, Biophys. J. 69, 1491 (1995).

75S. J. Kron, Y. Y. Toyoshima, T. Q. P. Uyeda, and J. Spudich, Methods Enzymol. 196,
399 (1991).
[ 1 8] CAGED MOLECULES AND MOTOR PROTEINS 335
caged compounds in conjunction with these in vitro assays of motor protein

function provides an opportunity to link the mechanical events of energy
transduction with the elementary biochemical steps in the ATPase reaction
of these systems (Fig. 1) and to determine the coupling between the ATPase
reaction and motion. Only a few experiments of this kind have been pub-
lished to date, 81-83 but this is a promising area for further progress. As an
example we list methods used for studying force and sliding velocity of
kinesin on microtubules. Myosin and actomyosin have also been studied
using caged ATP in vitro with evanescant wave video microscopy8z and in
suspension with electron paramagnetic resonance spectroscopy. 83'84
Preparation of Kinesin and Tubulin

Kinesin is purified from bovine brain by microtubule affinity followed
by D E A E chromatography (Fractogel DEAE, Merck, Darmstadt, Ger-
many) and sedimentation. 85 Tubulin is prepared from bovine brain by
repeated temperature-dependent polymerization/depolymerization cycles
and phosphocellulose chromatography, and then labeled with tetrameth-
ylrhodamine 86 (Molecular Probes, Eugene, OR). Purified kinesin (-0.2
mg/ml) and tubulin ( - 1 0 mg/ml) are rapidly frozen in liquid nitrogen and
stored at - 8 0 °.
Microtubules are prepared by copolymerization of fluorescent and non-
fluorescent tubulin in a molar ratio of 1:20, total protein concentration
- 1 0 mg/ml, for 30 min at 35 ° in a solution containing 1 mM GTP, 100 mM
MES buffer, pH 6.7, 5 mM MgClz, 1 mM EGTA, and 33% glycerol. After
polymerization, microtubules are stabilized by adding 40/xM taxol, 81 pel-
leted at 250,000 g for 10 min at 25 °, resuspended at 10 mg/ml in a solution
containing 80 mM PIPES, pH 6.8, 2 mM MgCI2, 40/zM taxol, and 1 m M
EGTA, and stored at 4 ° for up to a few months.
76 y. Harada, K. Sakurada, T. Aoki, D. D. Thomas, and T. Yanagida, J. Mol. Biol. 216, 49 (1990).
77 A. Ishijima, T. Doi, K. Sakurada, and T. Yanagida, Nature 352, 301 (1991).
78 A. Ishijima, Y. Harada, H. Kojima, T. Funatsu, H. Higuchi, and T. Yanagida, Biochem.
Biophys. Res. Commun. 199, 1057 (1994).
79 K. Svoboda, C. F. Schmidt, B. J. Schnapp, and S. M. Block, Nature 365, 721 (1993).
80j. T. Finer, R. M. Simmons, and J. A. Spudich, Nature 368, 113 (1994).
81 H. Higuchi, E. Muto, Y. lnoue, and T. Yanagida, Proc. Natl. Acad. Sci. U,S.A. 94, 4395 (1997).
8~p. B. Conibear and C. R. Bagshaw, FEBS Lett. 380, 13 (1996).
83 C. L. Berger, E. C. Svensson, and D. D. Thomas, Proc. Natl. Acad. Sci. U.S.A. 86, 8753 (1989).
84 E. M. Ostap, H. D. White, and D. D. Thomas, Biochemistry 32, 6712 (1993).
85 B. J. Schnapp and T. S. Reese, Proc. Natl. Acad. Sci. U.S.A. 86, 1548 (1989).
86 A. Hyman, D. Drechsel, D. Kellogg, S. Salser, K. Sawin, P. Steffen, L. Wordeman, and T.
Mitchison, Methods Enzymol. 196, 478 (1991).
Apparatus for Combined Motility Assays and Photolysis of Caged A TP
T o detect force or m o t i o n f r o m k i n e s i n i n t e r a c t i n g with m i c r o t u b u l e s ,
either f l u o r e s c e n t - l a b e l e d m i c r o t u b u l e s or k i n e s i n - c o a t e d fluorescent b e a d s
are o b s e r v e d b y fluorescence microscopy (Fig. 6). F l u o r e s c e n c e is excited
by light f r o m a m e r c u r y arc l a m p t h r o u g h a c o m m e r c i a l r h o d a m i n e epiflu-
orescence filter b l o c k or an i n t e n s e laser such as a f r e q u e n c y - d o u b l e d
N d : Y A G [ n e o d y m i u m - ( y t t r i u m - a l u m i n u m - g a r n e t ) ] g r e e n laser (A = 532
nm). F l u o r e s c e n c e images are r e c o r d e d by a high sensitivity c a m e r a such
as a S I T camera.
M i r r ~
Quadrant
photodiode Displacement
Piezo _ ~
stage , ,J ,
",, Laser trap.,,"
P Trapping~ Kinesin Force
F°rce~/=.-,~ ^ ~ ,a --~
Green laser
S~er ?oeO-,,
i Dichroic
mirrors " " K i ~
UV laser
M i e r o t u ~
Infrared laser
Mirror~
SIT camera Fluorescent
images
FIG. 6. Schematic diagram of an experimental arrangement for laser trapping nanomet@
of displacement and force generation by single kinesin molecules initiated by flash photolysis
of caged ATP. The green laser (frequency-doubled Nd : YAG, 100 mW at ;t = 532 nm; Model
140-0534-100, Lightwave Electronics, Mountain View, CA) for fluorescence excitation, the
UV laser (He-Cd, 12 mW at )t = 354 nm; Model 72121, mfg. Liconix, Santa Clara, CA), and
the infrared laser (diode pumped, Nd:YAG, 300 roW, A = 1064 nm; Model MYLS-300-S
OEM, Santa Fe Laser Co., CA) are all focused onto the sample through an objective lens
(Plan NCF Fluor 100x, Nikon). The UV laser is equipped with a rapid mechanical shutter
(response time 0.2 msec, Cambridge Technology, Inc.). Fluorescence images are recorded by
a SIT camera (C2741, Hamamatsu Photonics). The transmitted bright-field image of the
sample, illuminated by a 50-W tungsten-halogen lamp or the green laser, is projected onto
a quadrant photodiode (s944-13, Hamamatsu Photonics) by an objective lens (Plan 100x,
Nikon). (Inset) A representation of a kinesin molecule producing active translocation force
along a microtubule against the force applied to a latex bead by the optical trap. [Modified
from H. Higuchi, E. Muto, Y. Inoue, and T. Yanagida, Proc.Natl.Acad. Sci. U.S.A.94, 4395
(1997). Copyright (1997) National Academy of Sciences, U.S.A.]
[181 CAGED MOLECULES AND MOTOR PROTEINS 337
To measure the forces produced by single motor protein molecules

bound to beads, an optical trapping system is combined with fluorescence
microscopy79'81 (Fig. 6). A diode-pumiCed, N d : Y A G infrared laser (A =
1064 nm) is focused at the sample plane by an objective lens to form the
optical trap. The transmitted bright-field image of the bead, illuminated by
a 50-W tungsten-halogen lamp or the green laser, is projected onto a
quadrant photodiode by an objective lens. Displacements of the bead in
orthogonal directions are determined from the outputs of the quadrant
photodiode boosted by differential amplifiers. Data are recorded using a
computerized data acquisition system at a sampling rate of 200 Hz-20 kHz.
For photolysis of caged ATP, the sample is illuminated through the
lower microscope objective in Fig. 6 with a pulse from a continuous wave,
near-ultraviolet laser such as a He-Cd laser (A = 354 nm) equipped with
a rapid mechanical shutter or from a N d : Y A G or Nd:YLF pulsed laser
(Table II). The laser beam passes through a 3-mm-diameter field diaphragm
and is then focused to a diameter of ~30/zm at the sample plane by the
objective lens. With the He-Cd laser propagating 3.8 mW of energy through
the objective lens, the energy density in the 707-/.~m2 laser spot at the sample
is 5400 mW mm -2.
In Vitro Assays of Motor Protein Activity
Microtubule Gliding Assay

A 3- to 10-~1 flow chamber is made of large (24 × 32 mm 2) and small
(18 × 18 mm 2) coverslips separated by 2-mm-wide slivers of 10- to
30-/~m-thick polyester film. To reduce adhesion of proteins, the surfaces
of the coverslips are coated with casein by incubating in a standard solution
(80 mM PIPES buffer, pH 6.8, 2 mM MgC12, 1 mM EGTA, and 0.7 mg/
ml casein) for 2 min. Kinesin is rapidly thawed, diluted to 10 t~g/ml ( - 25
nM) in the standard solution, and introduced into the flow chamber for 2
rain. After unbound kinesin is washed out with the standard solution,
fluorescent microtubules ( - 1 ~g/ml) are put into the flow chamber in a
gliding assay solution containing an enzymatic oxygen scavenging systemv6
[80 mM PIPES, pH 6.8, 2 mM MgClz, 1 mM EGTA, 0.5-2 mM ATP or
caged ATP, 10/.~M taxol, 10 mM glucose, 0.5% (v/v) 2-mereaptoethanol,
50 ~g/ml ( - 5 U/ml) glucose oxidase, and 9 ~g/ml (-100 U/ml) catalase
(Sigma, St. Louis, MO), ionic strength - 50 m M and temperature, 25°-27°].
Finally, the four edges of the flow chamber are sealed with nail polish. The
rate of microtubule sliding (after photolysis in caged ATP experiments) is
determined by fluorescence video microscopy.
Bead Assay for Force Measurement

Kinesin-coated beads are prepared as follows: the surface of 0.5- to
1-tzm-diameter fluorescent latex beads [blue fluorescent, Molecular Probes,
final concentration of 5% (w/v) or 140 pM] is coated with casein by incubat-
ing on ice in the standard solution for >30 min. The beads (final concentra-
tion, 70 pM) are then rapidly mixed for about 30 sec with an equal volume
of the standard solution and kinesin (final concentration, 70-1000 pM) and
incubated on ice for >30 min.
A flow chamber is made with coverslips spaced -10/~m apart. Fluores-
cent microtubules are attached to the coverslips by incubation for 2 rain
in the flow chamber at - 1 0 / x g / m l in the standard solution with 10/zM
taxol added. The remaining coverslip surfaces are then coated with casein
by incubation for 2 min with the standard solution plus 10/xM taxol. With
this treatment, the beads do not bind to the coverslips. Finally, the chamber
is filled with a bead-assay solution containing the beads and a gliding assay
solution [-0.1 pM kinesin-coated beads, 0.7 mg/ml casein, 1 mM caged
ATP, 1-25 U/ml apyrase (Sigma)]. The apyrase is added to hydrolyze any
ATP or ADP contamination in the solution. To measure force, kinesin-
coated beads are trapped with the infrared laser and brought into contact
with microtubules attached to the coverslip (inset in Fig. 6A). Then, active
movement is initiated by photoliberating ATP from caged ATP with a
5-msec pulse of UV laser light. The concentration of released ATP is varied
by changing the caged ATP concentration. After photolysis of caged ATP
and a variable time lag, the beads actively translate in a stepwise manner
and then stabilize at a high force. The rate and magnitude of these events
are measured from the quadrant photodiode records.
Quantitating Photochemical Conversion
Photochemical Conversion in Muscle Fibers

Techniques to analyze the extent of photochemical conversion depend
on the volume of sample that can be collected following the experiment.
In a typical muscle fiber experiment, the fiber is bathed in a 10- to 50-/zl
trough that is illuminated fairly evenly by the photolysis light source. For
caged ATP or caged ADP, this macroscopic sample is collected and ana-
lyzed for released nucleotide by HPLC. Two microliters of photolyzed
solution is loaded onto an anion-exchange column (Radial Pak 8PSAX,
bead size 10/xm, Waters Corp., Milford, MA) and eluted with an isocratic
running buffer of 0.3 M (NH4)H2PO4 adjusted to pH 4.3 and mixed with
acetonitrile (97 : 3, v/v). Signal amplitude and chromatographic separation
are sufficient to enable detection of 2-30% photolysis of 10 mM caged nucle-
otide.
For caged Pi, HPLC analysis or a colorimetric assayzl is used to deter-

mine the amount of photolysis. For the colorimetric assay, Pi is first sepa-
rated from phosphoesters, such as ATP and CP, that increase background
Px by hydrolysis. Pi in the photolysis solution is precipitated as calcium
phosphate using cold Ca(OH)2 (0.2 M), pelleted, and resuspended twice.
The pellet is then dissolved in 1 N HCI and assayed for Pi using a molybdate-
type phosphorus reagent (Sigma). Comparing the absorption measurements
in this assay with those on standards made by serial dilution of a 10 mM
stock of Pi, the method is linear up to 10 mM and can detect 0.05-0.1 mM
photoreleased Pi.
For photolysis of caged ATP in fibers immersed in silicone or paraffin
oil, the volume of photolysis solution within the fiber ( - 2 0 nl) is not
sufficient for these chemical methods. In this case, the ATP concentration
liberated by photolysis inside the fiber is determined using caged [3H]ATP.
The fiber is prepared for photolysis using the same solutions and procedures
as described earlier except that the Ca rigor solution contains ~2 mM
caged [3H]ATP (loading solution). ATP is released by photolysis, and the
fiber is allowed to contract isometrically in the oil for 1-2 min after photoly-
sis. The fiber is then immersed for ~4 rain in a washout solution, relaxing
solution with 0.2 mM unlabeled caged ATP, 0.2 mM ADP, and 0.1 mM
AMP added to reduce loss of radioactive counts. The fiber is immersed
for 2 min each in two more aliquots of the washout solution. The three
postphotolysis washes, the loading solution, paraffin oil after the photolysis,
and the fiber after washing are saved for counting of radioactivity. Nucleo-
tides and caged nucleotides in the solutions are separated using a SAX
analytical HPLC column as described earlier. The elution is monitored at
254 nm. One-milliliter fractions are collected, mixed with scintillation fluid,
and each sample is counted for 1 rain in a liquid scintillation counter
along with standards, the loading solution, and the oil collected from the
photolysis trough. The concentration of photolyzed caged ATP is calculated
from the initial amount of labeled caged ATP in the fiber, the proportion
photolyzed, and the bulk concentration of caged ATP in the Ca rigor
solution. It is important to measure the laser energy actually striking the
fiber in this technique because inhomogeneity of the laser beam can other-
wise markedly increase variability between experiments.
In experiments using DM-nitrophen or NP-EGTA to release C a 2+, the
muscle fiber itself provides a convenient Ca 2+ sensor. The relationship
between steady active tension and the free Ca 2+ concentration is determined
before or after the photolysis trials. Ionic conditions during this control
experiment are chosen to be identical to those in the photolysis experiment,
except a C a 2+ chelator with well-characterized affinity, such as EGTA, is
used to buffer the free [Ca 2+] at saturating and various intermediate values.
The fiber is activated at saturating [Ca 2*] at the beginning of the experiment
and several times later to guard against, or correct for, decline of the
fiber performance. Comparison of the fiber tension on photolysis with this
standard curve provides an estimate of the [Ca 2+] released. Of course, this
estimate depends on the similarity in ionic conditions between the control
and test portions of this experiment and it assumes that photolysis does
not alter the Ca 2+ response of the muscle fiber. When DM-nitrophen or
NP-EGTA is photolyzed, bound Ca is released rapidly, creating an initial
spike of free Ca 2+ in the range of 100-500/zM. Some of the released Ca 2+
binds to unphotolyzed caged Ca, reducing the free [Ca 2+] to 10-50/zM
within a few milliseconds. 87 This initial transient of high free Ca 2+ does not
affect the steady tension reached in a contraction, but it might alter the
kinetics. The effects of this initial spike in caged Ca experiments are contro-
versial in the excitation-contraction coupling field.88,89
Fraction of Caged A T P Photolyzed in Vitro

The concentration of photoreleased ATP for in vitro experiments is
measured by illuminating a sample chamber through an objective lens at
a known energy density. In Fig. 6, the photolysis pulse propagates through
the lower objective lens and a restricting aperture is removed. 81 The cham-
ber, made of fused silica (path length 2 mm), is filled with 90 txl of a standard
motility assay solution (described earlier) containing caged ATP at 0.2 m M
or less. The concentration of caged ATP should be kept low to minimize the
light gradient due to absorption by the caged ATP (see further discussion of
this point later). ~354 for caged ATP is - 6 0 0 M -1 cm -1, giving an optical
density of 0.012 (97% transmission) through the 2-mm path length at 0.1
m M caged ATP. The chamber is illuminated by the UV laser at known
low energy density, e.g., 0.13 mW mm -2, for 1, 2, and 3 sec at 25-27 °.
The photoreleased ATP is then hydrolyzed by adding 90/xl of a solution
containing hexokinase ( - 1 0 U/ml, Sigma) or 0.2 mg ml 1 skeletal muscle
myosin in 200 m M HEPES buffer, pH 7.8, 0.6 M KC1, 20 m M E D T A and
incubating at 25-27 ° for 10-60 min. The amount of Pi produced is measured
by the malachite green method. 76 The percentage photolysis, R, was calcu-
lated to be 8.7% of the starting caged ATP per mJ mm -2 of illumination. 81
This value is consistent with that obtained by photolysis of caged ATP
within an 80-tzm-diameter muscle fiber (7.4% per mJ mm -2) using a pulsed
347-nm laser. 74 At the energy density used in the motility experiments, the
rate constant kr for removal of caged ATP is calculated to be 470 sec -~
8vG. C. R. Ellis-Davies,J. H. Kaplan, and R. J. Barsotti, Biophys. J. 70, 1006 (1996).

88S. GyOrkeand M. Fill, Science260, 807 (1993).
89G. D. Lamb, M. W. Fryer, and D. G. Stephenson, Science263, 986 (1994),
[ ] 8] CAGED MOLECULES AND MOTOR PROTEINS 341
(0.087 mm 2 rrd -1 × 5400 mJ mm -2 s-l). The extent of photolysis is given

by (1 - e-~rt), where t is the duration of illumination. For a 5-msec shutter
time, the fraction of photolyzed caged ATP is calculated to be 0.90.
Uniformity of Photolysis in Sample Chamber

As the photolysis beam propagates through the sample chamber, its
intensity decreases due to absorption by the caged compound or scattering
by the biological sample. The amount of photolysis thus decays through
the sample. The consequences of nonuniform release of the active molecule
depend on the experimental preparation and the geometry. The concentra-
tion of photoreleased biological effector R0 at the entrance face of the
sample chamber is given by R0 = (IoQpACe)/(abhcNA), where I0 is the
photolysis pulse energy at wavelength A, a and b are the height and width
of the photolysis chamber, Qo is the quantum yield for photolysis, C is the
concentration of the caged molecule, e is its molar absorption coefficient
at A, h is Planck's constant, c is the velocity of light, and NA is Avogadro's
number. In the case of a skinned muscle fiber, the decay of intensity is due
largely to absorption by the caged compound in the bathing medium. The
light intensity decreases according to 1 = Io e -c~x, where x is the distance
into the sample chamber. Assuming photolysis is proportional to light
intensity (valid except for very short, focused laser pulses or high percentage
photolysis), R(x) -- (IoQpACe)/(abhcNA) e -c~x. Figure 7A shows the profile
of relative light intensity expected for the typical conditions of a muscle
fiber experiment in which NPE-caged ATP is photolyzed using a frequency-
doubled ruby laser, A = 347 nm, e --- 660 M -1 cm -1, Qp = 0.63, and d =
0.1 cm. The average concentration of photoreleased ligand over a path
ligand d is Rav= IoQpA(1 - e-C~d)/(abhcNg d), which is shown by a hori-
zontal dashed line in Fig. 7A. The position, Xav, within the trough where
the local photolysis is equal to the average value is the intersection of these
two lines 470/xm from the front surface of the trough, very close to the
center. The photolysis at the back of the l-ram trough is 0.5 of that in the
front, but this can be increased to 0.7 with a polished stainless-steel surface
that reflects one-half of the light incident on it 32 (see Fig. 2C). The decay
of intensity across a 100-/zm muscle fiber (circle in Fig. 7A) is only 6% at
10 mM caged ATP.
The optimal concentration of caged ATP is dictated by the concentration
of ATP to be released and by agonist or antagonist action of the caged
compound before photolysis. Taking the position of the fiber to be fixed
at a certain depth, df (typically 500 /zm), the released concentration of
ATP [R(d 0 at the fiber] increases with caged ATP concentration up to a
A B
3.0,
o~
E:
r'-
.~. o.8 v
2,5[
2.0
.E 0.6
1.5
~ 0.4
ol.0
~ 0.2 o 0.5
13_
0. 0.2 0.4 0.6 0.8 1.0
0.0
20 40 60 80 100
Optical Path (ram) Caged Compound (raM)
FIG. 7. Characteristics of spatial and concentration dependence of photorelease in a typical

muscle fiber experiment. (A) Light intensity distribution along a 1-mm optical path in a muscle
fiber photolysis trough. The solid line shows the exponential delay of light intensity and
photorelease in a 1-mm optical path photolysis trough containing 10 mM NPE-caged ATP,
e347 = 660 M -a cm -a. The average concentration of photoreleased A T P is shown by the dashed
line. A 100-/zm-diameter muscle fiber (circle) in the center of the trough receives almost the
same light intensity and photolysis as the average value. (B) Concentration dependence of
photolysis. The concentration of A T P released at df = 500 /zm depth in the trough (the
position of the fiber) by photolysis of NPE-caged A T P at 347 nm is plotted vs the initial
[caged A T P 1. Photolysis quantum yield (Qp) is 0.63. Assuming that photolysis is proportional
to light absorption, the maximum photolysis occurs at caged ATP concentration, Cmax =
1/(e dr) = 30 m M in this example, but at a lower concentration if e or df is higher.
maximum (Fig. 7B) and then declines due to excessive absorption of the
light in the first half of the sample chamber. The maximum release occurs
at Cm~x= 1/(/3 dr) = 30 mM. In the fiber experiment, caged ATP concentra-
tions are kept lower (<10 mM) because caged ATP competitively inhibits
ATP binding (see later).
In general, the concentration of caged molecule is limited by the sample
depth and the absorptivity at the photolysis wavelength because R(df) peaks
at Cmax = 1/(/3 dr). This phenomenon also bears on the choice of illumination
wavelength. For example, NPE-caged ATP absorbs very strongly at 260
nm, suggesting that a laser emitting at wavelengths less than 300 nm would
be more suited to the experiment. However, the absorptivity,/3260 19,600 =
M -1 cm -1, is too high for efficient photolysis at df = 500/zm. The caged

ATP concentration giving maximal photolysis [1/(/3 dr)] is only 1 mM in
this case, making it difficult to approach 1 mM ATP release. At given light
levels, increasing the caged ATP concentration above 1 mM would decrease
rather than increase the amount of ATP released at the fiber. Higher
photolysis wavelengths avoid this problem because/3 is smaller and there
is the further advantage that absorption by proteins and consequent optical
damage is lower for wavelengths above 310 nm.
Prephotolysis Effects of Caged ATP on Motor Proteins

NPE-caged ATP binds weakly to the active site of myosin 52and kinesin 8~
and thereby inhibits ATP binding and sliding velocity. The extent of this
inhibition measured by the "slack test" 90 in muscle fibers and in a kinesin
microtubule gliding assay is presented below. 8a
Inhibition of Muscle Shortening Velocity

Glycerol-extracted single psoas muscle fibers are prepared from rabbit
muscle and mounted in the photolysis setup as described earlier. Relaxed
fibers are stretched by 10% of slack length to ~2.5/~m sarcomere length.
They are incubated for ~2 min in a preactivating solution containing low
Ca-buffering capacity and then activated at 30/~M free [Ca2+], F = 200
m M (Table V), T = 10°, and 0.3-5 mM [ATP]. After active tension reaches
a steady level, a quick release, typically 5-20% of the fiber length, is applied
and the fiber briefly goes slack until active shortening causes redevelopment
of tension (Fig. 8A, inset). The fiber is then relaxed again. From pairs of
slack releases of different amplitudes, the times elapsed until the initial
redevelopment of tension are measured. 9° Shortening velocity is calculated
from the difference in amplitudes of the releases and the differences in
slack times.
Shortening velocity increases with MgATP concentration and saturates
in the mM r a n g e . 91'92 5 mM caged ATP suppresses shortening velocity (Fig.
8A). When plotted on an inverse plot (Lineweaver-Burk plot, not shown),
data fall on two straight lines that intercept the ordinate at approximately
the same maximum velocity, indicating that caged ATP is a competitive
inhibitor of ATP. Equation (1) is fit to velocity data to give a half-saturation
value for ATP concentration
V/Vmax = [ATP]/{[ATP] + KATP (1 + [cATP]/K,)} (1)
where KI is an inhibition constant for caged ATP (cATP), KATe = 280 +
50 /zM (n = 21 data points from 10 fibers), and K~ = 1.93 + 0.74 mM
(n = 16 points from 9 fibers). In a series of experiments using free [Mga+]
from 300/zM to 10 mM, the inhibitory effect of caged ATP did not depend
appreciably o n [Mg2+], suggesting that both cATP 3- and Mg-cATP- are
equally effective inhibitors. The R-isomer of caged ATP is slightly (20%)
less effective at inhibiting shortening velocity than the S-isomer.
90 K. A. P. Edman, J. Physiol. 291, 143 (1979).

91 M. A. Ferenczi, Y. E. Goldman, and R. M. Simmons, J. Physiol. 350, 519 (1984).
92 E, Pate, M. Lin, K. Franks-Skiba, and R. Cooke, Am. J. Physiol. 262~ C1039 (1992).
344 CAGED COMPOUNt)S [ 181
A
1.2
B c
c- 1.2 r ] t- 0.!
•~ | B caged ATP = 0 [
~ ::" 0
,..,_o
~.01 ,-~ • "7 0,,
LT.~ 0.8 ~2 E .
-o~> 0.6
~ o.4
~o.'
o>
o.g ~>oo , ,| ~ OJ . . . . .
0 1 2 3 4 5 6 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.O 0.2 0.4 0.6 0.8 1.0 1.2
ATP concentration(mM) ATP concentration(mM) cagedATP conc. (mM)

Fro. 8. Inhibition of motor protein activity by NPE-caged ATP. (A) Shortening velocity
measured by the "slack test" (see inset) plotted against ATP concentration in the absence
(upper curve) and presence of 5 mM caged ATP, All velocities are relative to that at 5 mM
ATP, 0 caged ATP (square) at 10°, corresponding to 4.25 +__0.38/~m sec-1 per half sarcomere
(mean --_ SEM, n = 19 fibers). The curves represent Eq. (1) of the text plotted with K A T e =
280 ~M and the inhibition constant for NPE-caged ATP, KI = 1.93 mM. (Inset) Force
recordings for a single fiber activated at -32/zM free Ca2+, 0.3 mM ATP, 0 caged ATP at
10°. Quick releases of 12 and 18% were applied at the vertical line, causing the fiber to go
slack. The slack time is measured by recording the times elapsed between the release and
until the initial redevelopment of tension. (B) Velocity of microtubule gliding driven by
kinesin plotted against ATP concentration in the gliding assay solution (see text) with no
added caged ATP at 25-27°. The curve represents Eq. (1) of the text with K A T P = 30/xM
and Vm~x= 0.95/~m sec-1. (C) Velocity of microtubule gliding as in (B), but plotted against
concentration of NPE-caged ATP (mixed R- and S-isomers) in the presence of 30/xM total
ATP and 2 mM total Mg. The curve represents Eq. (1) with KATpand Vmaxas described
earlier and K1 = 126/~M.
A c c o r d i n g to these values for inhibition of muscle fiber shortening

velocity, u n p h o t o l y z e d caged A T P remaining after a flash in laser photolysis
experiments w o u l d inhibit shortening. F o r example, w h e n 1 m M caged
A T P is released f r o m 5 m M starting caged A T P , the remaining 4 m M of
caged A T P w o u l d suppress the shortening velocity to 70% of its value at
1 m M A T P in the absence of caged A T P .
C a g e d A T P binding to the active site also affects the time course of
isometric force d e v e l o p m e n t following activation of a fiber by photolysis
of caged A T P . T h e expected slowing of reaction kinetics by this effect is
illustrated in Fig. 9, which shows simulations of the simple reaction scheme
(Scheme 1).
T h e effective rate t h r o u g h the cross-bridge d e t a c h m e n t step in S c h e m e
I is given by Eq. (2):
[ A T P ] [ A M l k T = [ A T P ] [ A M t ] K I / ( [ c A T P ] + KI) (2)
w h e r e K] is the inhibition constant of caged A T P for d e t a c h m e n t and [AMt]

is the total actomyosin c o n c e n t r a t i o n ( = [AM] + [ A M . c A T P ] ) at the
[ 181 CAGED MOLECULES AND MOTOR PROTEINS 345
E 1.0
.0
tn 0.8
E
I-- 0.6
q)
> 0.4
t~
• 0.2
n"
0.0
0 25 50 75 100 125 150 175 200
Time (ms)
FIG. 9. Simulations of tension transients of a muscle fiber activated by photolysis of caged

ATP. Differential equations corresponding to Scheme I of the text were integrated with a
fourth-order R u n g e - K u t t a algorithm. Photoreleased A T P concentration, 1 m M ; total myosin
head concentration, 200 /~M; kp, 100 sec-1; kR, 50 sec 1; kc, 1 x 106 M -1 sec-1; k - c ,
2 × 103 sec -1. For trace a, [cATP], 6 m M ; kT, 1 × 106 M 1 sec 1. For trace b, [cATP], 0; kT,
2.5 × 105 m -1 sec -1. For trace c, [cATP], 0, kT, 1 X 106 M -1 sec 1.
beginning of the experimental trial. Including the dark photolysis reaction

(kp), the photolysis-induced detachment is a two-step series reaction whose
rate (kobs) is approximately the product divided by the sum of kp and
([ATPlkT[AM]/{[AM] + [AM.cATPI} ).
( cAT I +
kobs = (3)
([cATp]I + KI)kT[ATP] + kp
As an example of the expected effect of caged A T P binding on the

detachment kinetics, let kp = 100 sec -1, kT = 1 × 106 M -1 sec -1, released
[ATP] = 1 mM, KI = 2 mM, and [cATP] = 6 mM. kobs is then calculated
from Eq. (3) to be 71.4 sec -1, and a simulation of detachment and reattach-
ment is shown in Fig. 9 (trace a). If caged ATP did not inhibit detachment
(set [cATP] = 0 in Eq. (3)), then the same kobs would be obtained with
kT = 2.5 X 105 M -1 sec -1, and a kinetic simulation (Fig. 9, trace b) shows
that the time course of cross-bridge detachment and reattachment would
be the same. Trace c in Fig. 9 shows the expected faster time course at
k T = 1 )< 106 M -1 s e c -1 with no inhibition by caged ATP. kobs in this case
is 90.9 sec -1. Thus caged ATP inhibits the rate of detachment in typical
cATPk---ZT--~ATP
cATP ATP
AM.cATP k~--2-2~ A M + Detached ~ Active
SCHEME I
conditions of a photolysis experiment. This inhibition must be taken into

account in estimating kT, either by fitting a model, such as Scheme I, which
includes the AM.cATP state, to the transient data or by adjusting kT upward
by the factor ([cATP] + K~)/KI) implied by Eq. (3). This adjustment
increases kT fourfold in the case considered here. Suppression of the detach-
ment rate would be less at lower [cATP].
Inhibition of Kinesin Translocation Velocity by Caged A TP

To determine the inhibition constant of caged ATP on the velocity of
microtubule sliding driven by kinesin, concentrations of ATP and caged
ATP were varied in the incubation medium of the in vitro microtubule
gliding assay. Microtubule sliding velocity increased with [ATP], giving a
half-saturation value of K A T P = 30/.~M and Vma x = 0.95/xm sec -1 (Fig. 8B,
correlation coefficient r = 1.0). At constant total [ATP] = 30 txM and
constant total [Mg] = 2 mM, increasing [cATP] up to 1 mM inhibited
velocity (Fig. 8C). Assuming competitive inhibition, fitting Eq. (1) to these
data gave a KI = 126/zM (r = 0.99) (Y. Inoue and H. Higuchi, personal
communication).
Binding of Mg2÷ and Ca 2÷ to Caged ATP

In calculating the free [Mg2+] and free [Ca 2+] in solutions containing
caged ATP, the affinities of these ions to caged ATP must be taken into
account. This section presents measurements of these affinities. The pre-
dominant species of NPE-caged ATP near neutral pH is cATP 3- and that of
the calcium complex of caged ATP is Ca-cATP-. To determine the affinity
gca = [Ca-cATP-]/([CaZ+][cATp3-]) (4)
we made up test solutions containing known total concentrations of Ca,

Mg, and cATP (Cat, Mgt and cATPt, respectively) and measured free
[Ca 2+] using a Ca-sensitive electrode (WPI Inc., New Haven, CT, type
TIPCA). The caged ATP stock was a pure sample (<0.01% contamination)
of a mixture of both enantiomers of NPE-caged ATP.
A solution containing 5 mM caged ATP, 146 mM KCI, and 100 m M
TES buffer, pH 7.1, 20 ° was mixed with 10 mM CaC12, 5 mM caged ATP,
137 mM KC1, and 100 mM TES buffer in various ratios to result in test
solutions with 0.63-10.0 mM Cat. cATPt was determined from the absorp-
tion at 260 nm, assuming ecATP = 19.6 M -1 cm -1. The free [Ca z+] is given
by the quadratic equation
[Ca2+]2Kca + lEa2+](1 + c A T P t K c a - C a t K c a ) - C a t = 0 (5)

[ 181 CAGEDMOLECULESAND MOTORPROTEINS 347
and the expected output potential difference of the Ca-sensitive electrode

from that in a solution with 1 mM free [Ca 2÷] is given by
V - V~M = S log {[Ca2+]/(1 mM)} (6)
Before and after each calcium titration, the response of the electrode was
tested in calibrating solutions containing 0, 0.1, 1, and 10 mM free Ca 2÷
(WPI, CALBUF). The slope (S) of the electrode standard curve was 30.1
mV per decade change of [Ca2÷]. After adjusting Kc, in Eqs. (4) and (5)
for the best fit, Eqs. (5) and (6) predicted calcium titration data quite well
(correlation coefficients r = 0.99).
The affinity of Mg 2÷ for caged ATP,
KMg = [Mg-cATP-I/([Mg2+I[cATp3-I) (7)
was similarly measured by displacing Ca from caged ATP with Mg and
measuring free [Ca2+] with the calcium-sensitive electrode. Mgt was altered
from 0 to 10 mM in solutions containing 5 mM caged ATP, 100 mM TES
buffer, 137-146 mM KCI, and Cat at 0.5, 1.0, or 2.0 raM. The free [cATP 3- ]
in solutions containing Mg is given by the cubic equation [Eq. (8)]:
[cATp3-]3KcaKMg + [cATp3-]2(Kca + KMg -- cATPtKcaKMg + CatKcaKMg
+ MgtK~.gMg) + [cATp3-](1 - cATPtKca
-- cATPtKMg+ CatKca + MgtKMg) - cATPt = 0) (8)
Free [Ca 2+] is then given by
[Ca 2÷] = Cat/(1 + Kc~[cATp3-]) (9)
After adjusting/(Ca and KMg in Eqs. (7) and (8) for the best fit, Eqs. (6)
and (9) predicted magnesium titration quite well (correlation coefficients
r = 0.99).
pKca = - l o g / ( c a was found to be 2.52 _ 0.03 (mean ___SD, n = 6) at
20 °, pH 7.1, corresponding to a dissociation constant of 3 mM for Ca-cATP-.
pKMg = --log KMg was found to be 2.45 +__0.02 (n = 4), corresponding to
a dissociation constant of 3.5 m M for Mg-cATP-. The dissociation constant
for Mg for the reaction intermediate populated upon UV irradiation of
caged ATP (aci-nitro intermediate) is also 3 raM. 56 pKc~ and pKMg for
binding of Ca 2+ and Mg 2÷ to ATP at pH 7.1 were found to be 3.55 _ 0.02
(n = 6) and 3.81 +_ 0.02 (n = 3), respectively, in agreement with values
obtained by Moisescu and Thieleczek 93 under very similar conditions, thus
validating this potentiometric method.
Conclusions
Photolysis of caged molecules, particularly caged ATP and caged Ca e÷,
provides a very flexible experimental method to activate muscle fibers and
93D. G. Moisescuand R. Thieleczek, Biochim. Biophys. Acta 546, 64 (1979).

motor proteins in vitro. Hence many protocols for laser pulse photolysis
with these proteins have been described. Combination of photolysis with
other biophysical signals that probe molecular events is a fertile area for
investigation. Many of the considerations in photolysis experiments on
skeletal muscle and motor proteins are applicable to other types of muscle
and other enzymes.
Acknowledgments
This work was supported by NIH Grant HL15835 to the Pennsylvania Muscle Institute
and the Yanagida Biomotron Project E R A T O JST. We thank Drs. J. W. Walker and
G. C. R. Ellis-Davies for access to unpublished data, Drs. J. E. T. Corrie and D. R. Trentham for
helpful comments on the manuscript, and Ms. Kimberly L. Dopke for help with the manuscript.
[19] A p p l i c a t i o n o f C a g e d F l u o r e s c e i n - L a b e l e d T u b u l i n to
Studies of Microtubule Dynamics and Transport of
Tubulin Molecules in Axons
B y TAKESHI FUNAKOSHI a n d N O B U T A K A H I R O K A W A
Introduction
Microinjection of caged fluorescein-labeled cytoskeleton associated pro-
teins into cultured ceils and subsequent photoactivation of the fluorescence
in these conjugates has proved to be a very powerful tool in studying the
dynamics of the microtubule and actin cytoskeleton in living cells, t-4 Caged
fluorescein-labeled tubulin has been successfully applied to study microtu-
bule dynamics in mitotic spindles and neuronal axons. This approach has
allowed us to investigate molecular mechanisms of mitosis and the origin
of the slow transport of tubulin molecules in nerve axons. This article
describes methods and techniques using caged fluorescein-labeled tubulin
for investigations of the dynamics and transport of axonal tubulin molecules
in neurons.
In these studies, caged fluorescein-labeled tubulin is introduced into
cultured neurons by microinjection and, after a recovery period that also
allows for diffusion of the caged fluorescein tubulin throughout the cell,
l T. J. Mitchison, J. Cell Biol. 109, 637 (1989).

2 j. A. Theriot and T, J. Mitchison, Nature 352, 126 (1991).
3 S. Okabe and N. Hirokawa, Z Cell Biol. 117, 105 (1992).
4 j. Sabry, T. P. O'Connor, and M. W. Kirschner, Neuron 14, 1247 (1995).

METHODS IN ENZYMOLOGY,VOL, 291 0076-6879/98 $25.00
motor proteins in vitro. Hence many protocols for laser pulse photolysis
with these proteins have been described. Combination of photolysis with
other biophysical signals that probe molecular events is a fertile area for
investigation. Many of the considerations in photolysis experiments on
skeletal muscle and motor proteins are applicable to other types of muscle
and other enzymes.
Acknowledgments
This work was supported by NIH Grant HL15835 to the Pennsylvania Muscle Institute
and the Yanagida Biomotron Project E R A T O JST. We thank Drs. J. W. Walker and
G. C. R. Ellis-Davies for access to unpublished data, Drs. J. E. T. Corrie and D. R. Trentham for
helpful comments on the manuscript, and Ms. Kimberly L. Dopke for help with the manuscript.
[19] A p p l i c a t i o n o f C a g e d F l u o r e s c e i n - L a b e l e d T u b u l i n to
Studies of Microtubule Dynamics and Transport of
Tubulin Molecules in Axons
B y TAKESHI FUNAKOSHI a n d N O B U T A K A H I R O K A W A
Introduction
Microinjection of caged fluorescein-labeled cytoskeleton associated pro-
teins into cultured ceils and subsequent photoactivation of the fluorescence
in these conjugates has proved to be a very powerful tool in studying the
dynamics of the microtubule and actin cytoskeleton in living cells, t-4 Caged
fluorescein-labeled tubulin has been successfully applied to study microtu-
bule dynamics in mitotic spindles and neuronal axons. This approach has
allowed us to investigate molecular mechanisms of mitosis and the origin
of the slow transport of tubulin molecules in nerve axons. This article
describes methods and techniques using caged fluorescein-labeled tubulin
for investigations of the dynamics and transport of axonal tubulin molecules
in neurons.
In these studies, caged fluorescein-labeled tubulin is introduced into
cultured neurons by microinjection and, after a recovery period that also
allows for diffusion of the caged fluorescein tubulin throughout the cell,
l T. J. Mitchison, J. Cell Biol. 109, 637 (1989).

2 j. A. Theriot and T, J. Mitchison, Nature 352, 126 (1991).
3 S. Okabe and N. Hirokawa, Z Cell Biol. 117, 105 (1992).
4 j. Sabry, T. P. O'Connor, and M. W. Kirschner, Neuron 14, 1247 (1995).

METHODS IN ENZYMOLOGY,VOL, 291 0076-6879/98 $25.00
[19] PHOTOACTIVATION
OF TUBULINMOLECULESIN NEURONS 349
small segments of axons are exposed to a brief pulse of near-ultraviolet

light. The caged protection groups on the fluorescein absorb these photons,
and in a dark reaction the ether bond that links the caged groups to the
phenolic groups of fluorescein is cleaved, liberating highly fluorescent tu-
bulin conjugates only in the irradiated segment. The fate of the photoacti-
vated tubulin molecules was observed with low-light level fuorescence
microscopy or with electron microscopy by staining with antibody directed
against fluorescein.
Preparation of Caged Fluorescein-Labeled Tubulin

Caged ftuorescein-labeled tubulin is prepared according to the method
of Mitchison 1 with slight modifications. Briefly, phosphocellulose-purified
hog brain tubulin is labeled with Bis caged-fluorescein Sulfo-OSu (Dojindo
Laboratories, Japan) in solution at high pH (pH 8.5). This fluorescein
derivative is not in itself fluorescent, but generates highly fluorescent car-
boxyfluorescein after irradiation with near-ultraviolet light (photoactiva-
tion). After the labeling reaction, the tubulin preparation is subjected to
two cycles of polymerization and depolymerization to select for assembly
competent conjugates. Labeled tubulin preparations are analyzed after
photoactivation by sodium dodecyl sulfate gel electrophoresis to confirm
that the protein is labeled with the caged fluorescein reagent and is free
of unlabeled caged fluorescein. Free caged carboxyfluorescein runs at the
front of the gel and can be easily detected by irradiating the gel under an
ultraviolet light transilluminator.
Cell Culture
Cultured mouse dorsal root ganglion (DRG) neurons are used in these
experiments. These neurons have large cell bodies (25-60/xm), and the
growth rate of their axons plated on laminin-coated coverslips is very fast,
with a protrusion rate of 10-80/zm/hr. 3
1. About 20 dorsal root ganglions are isolated from adult mice 5'6 and
kept in Hanks' balanced salt solution (HBSS, GIBCO, Grand Island, NY)
at 4°. To minimize injury to these neurons, this step is completed within
45 min.
2. Dorsal root ganglions are washed twice with Ca 2÷, Mg2+-free
HBSS (Gibco).
s S. S. S. Goldenberg and U. De Boni, J. Neurobiol. 14, 195 (1983).

6S. Okabe and N. Hirokawa,J. Neurosci. 11, 1918 (1991).
3. Ganglions are treated with 0.25% (w/v) collagenase in Ca 2÷, Mg 2+-
free HBSS for 1 hr at 37°.
4. Ganglions are further incubated with 0.25% (w/v) trypsin in Ca 2÷,
Mg2+-free HBSS for 15 min.
5. After washing with culture medium (MEM, Nissui, Tokyo, Japan)
supplemented with glutamine and 15 mM HEPES, 5% horse serum, and
5% new calf serum, the ganglions are homogenized in the culture medium
by pipetting several times through a Pasteur pipette. The pipetting should
be performed gently to avoid damaging cells. If the ganglions cannot be
resuspended easily, harsher treatment with protease (with shaking) is rec-
ommended.
6. Suspended cells are plated on laminin-coated glass coverslips.
Healthy cells should be smooth and round, whereas heavily damaged cells
have a wrinkled appearance. The addition of Nerve Growth Factor (100
ng/ml) to the culture medium is optional.
Glass coverslips are secured with vacuum grease on the bottom of plastic
dishes that have been drilled with a hole smaller than the diameter of the
coverslip (Dow Coming, high vacuum silicone grease). Care should be
taken as some kinds of adhesive agents are toxic to the cells. For electron
microscopic analysis, we use CELLocate micro-grid coverslips (Eppendorf)
in order to identify the axons of cells irradiated with near-ultraviolet light.
The square meshes etched onto these coverslips are transferred to the resin
block and allow us to easily and precisely identify the photoactivated axons
after embedding.
Microinjection of Caged Fluorescein-Labeled Tubulin Into Neurons 3,7

Neurons that do not show any axonal processes or those with minor
sprouts are microinjected with bis-caged carboxyfluorescein-labeled tubulin
3-8 hr after plating and are incubated for a further 10-20 hr. This ensures
that microinjected-caged fluorescein tubulin is equally distributed through-
out the neuron. To identify microinjected neurons, cells are coinjected
with caged fluorescein-labeled tubulin (50-100/xM) and rhodamine-bovine
serum albumin (BSA) (0.5 mg/ml) in injection buffer (50 mM potassium-
glutamate, i00 mM KCI, and 1 mM MgCl2).
Photoactivation and Low Light Level Video Microscopy

Photoactivation is performed as described by Mitchison and colleagues
(see also the chapter in this volume) with slight modifications. I'3,s Essen-
7T. Funakoshi, S. Takeda, and N. Hirokawa,J. Cell BioL 133, 1347 (1996).
8T. Umeyama,S. Okabe, Y. Kanai, and N. Hirokawa,J. Cell Biol. 120, 451 (1993).
[ 191 PHOTOACTIVATION OF TUBULIN MOLECULES IN NEURONS 351
tially, the 365-nm line of a mercury lamp selected with a bandpass filter
(390-nm long pass) is introduced into the epifluorescent light path of an
inverted microscope (Axiovert; Carl Zeiss, Inc.) via a dichroic mirror posi-
tioned between the original mirror box and the field diaphragm (Fig. 1).
A handmade slit is placed at the point where the slit makes an image at
the field diaphragm. The beam of the mercury lamp is focused onto the
$
Halogen lamp Pathfinding
mirror
Mercury lamp
$
Lamps for
( - fluorescence
observation
• Shutter
< Field diaphragm
Excitation filter ~.j~ Band pass filter for UV introduction

for fluorescence l [ Slit ] e~- Shutter
observation
/1 •
Dichroic mirror for
UV introduction
Mercury lamp for

photoactivation
i
Y ~"~'~ Sample
I Dichro!c mirror for
fluorescence observation
Emission filter for

fluorescence observation
e---- Eyepiece
2 Objective lens
I
Fie. 1. A schematic of the light paths in our photoactivation/fluorescence microscope. See
text and the chapter in this volume by Mitchison et al. for further details.
slit using a lens. We can easily illuminate a small region of an axon (3-5
/zm wide) using this microscope with a 100× objective lens (Plan-Neo-
fluar, Zeiss).
To observe the fluorescence of cells we use a 100-W halogen lamp, the
excitation-emission filter set for fluorescein and/or rhodamine, and a cooled
CCD camera (C3640; Hamamatsu Photonics) with an exposure time of
1 sec. The halogen lamp is turned on only to record the fluorescence region
in order to limit photobleaching of the dye and any accompanying damage
to the cells. Mercury lamp epi-illumination of the sample is used in combina-
tion with a low power objective lens to identify injected cells.
Photomicroscopic Observation Following Photoactivation3.8

Although fluorescence microscopy does not allow us to resolve a single
microtubule in an axon, as microtubules are so densely packed, we can
observe a single cell many times after photoactivation.
Approach
1. Illuminate a small segment of the axon of the microinjected cell with
near-ultraviolet light. Fluorescein-labeled tubulin molecules are generated
in the illuminated segment.
2. Observe the fluorescence images immediately after and at appro-
priate time intervals following photoactivation.
Narrow bars of fluorescent tubulin marked in the axon by photoactiva-
tion do not move from the irradiation site even if in extending axons,
although the fluorescence intensity of the marked tubulin does decay
gradually (Fig. 2). These observations strongly imply that most of the
axonal microtubules are stationary, and we suggest that the intensity decay
reflects the polymerization-depolymerization turnover of microtubules
in the photoactivated segments. The limited spatial resolution and sensitiv-
ity of fluorescence light microscope make it difficult to detect any pop-
ulation of tubulin molecules that move out of the photoactivated seg-
ments.
Electron Microscopic Observation Following Photoactivation7

This method allows us to observe the fate of photoactivated molecules
at very high resolution with high sensitivity. To detect the population of
mobile tubulin molecules that might have been overlooked in observations
with the fluorescence microscope, we determined the distribution of fluo-
resceinated tubulin molecules in the same photoactivated axons using elec-
[ 191 PHOTOACTIVATION
OF TUBULINMOLECULESIN NEURONS 353
L
(
2~
4{3
e
60
FI~. 2. Photoactivation of an extending axon of a mouse DRG neuron. (a-e) Fluorescence

images showingthe marks (arrows) made on the microtubules by the photoactivation technique
remain stationary as the axon moves forward. (f and g) DIC images of the same axon. Arrows
indicate the position of the photoactivated region. During this experimental run, there was
a rapid extension of the axon. Elapsed time (minutes) after photoactivation is shown in the
lower left-hand corner. Reproduced from S. Okabe and N. Hirokawa, Z Cell Biol. 117, 105
(1992) by copyright permission of Rockefeller University Press. Bars: 10/zm.
tron microscope-based detection of an antiflorescein antibody. Although

this m e t h o d does not allow us to observe a single axon at m a n y time points
after the photoactivation, it provides both the resolution and the sensitivity
to detect transported tubulin molecules in a single axon. Antiserum against
fluorescein was produced using fluorescein-labeled keyhole limpet hemocy-
anin for the antigen. The antibody directed against the fluorescein hapten
was affinity purified using a CNBr-Sepharose affinity column (Pharmacia,
Piscataway, NJ) coupled with fluorescein-conjugated BSA.
We employ two different procedures to address the issue of whether
tubulin molecules are transported in the axon. To unambiguously observe
single microtubules, we permeabilize cells using taxol before fixation, as
previously described 9 with some slight modifications. This procedure ex-
tracts soluble cytosolic protein from axons (Fig. 3). To preserve the soluble
population of tubulin molecules (tubulin oligomers or heterodimers), cells
are fixed without a permeabilizing agent. U n d e r this condition, using high-
resolution electron microscopy, a transported population of tubulin mole-
cules can be seen (Fig. 4).
9 p. W. Baas and M. M. Black, J. Cell Biol. 111, 495 (1990).

354 CA6EO COMPOUNDS [191
~ I ............
'
i,ii==i,;,Lilii==i=I:i===i
i
ii
i i ii Ly,]~Q~ ............ .........
" iiiBi~],
~!i!=
FIG. 3. Behavior of photoactivated microtubules in extending axons. One hour after the
photoactivation of small regions of the axon, the cells were permeabilized, fixed, and stained
by antifluorescein antibody. (a and b) Photoactivated region. Microtubules decorated with
gold particles (arrowheads) were observed. (c and d) Examples of the areas outside the
photoactivated region: (c) an area approximately 4/zm distal to the photoactivated region
and (d) an area approximately 4/zm proximal to the photoactivated region. No microtubules
with gold labels were observed outside of the photoactivated regions. Reproduced from T.
Funakoshi, S. Takeda, and N. Hirokawa, J. Cell Biol. 133,1347 (1996) by copyright permission
of Rockefeller University Press.
FIG. 4. Transported tubulin molecules in the axon. (a-c) One minute after the photoactiva-
tion of a small section of the axon; cells were fixed without the preceding permeabilization
and visualized using an antifluorescein antibody. Many gold particles were observed outside
of the photoactivated region. (a) Photoactivated region. Many gold particles can be seen.
(b) An area about 20 tzm distal and (c) an area about 20 t~m proximal to the photoactivated
region. (a-c) Arrowheads indicate some of the gold particles. (d) An axon photoactivated
after fixation. The photoactivated region is underlined. (a-c) Many gold labels appear outside
of the photoactivated regions; the density of gold label is higher in the distal region than in
the proximal region. (a-c) Large spots are used to exaggerate the positions of the gold
particles, as they are too small to be resolved at this magnification. Reproduced from T.
Funakoshi, S. Takeda, and N. Hirokawa, J. Cell Biol. 133~ 1347 (1996) by copyright permission
of Rockefeller University Press. Bars: 1 ~m.
[ 19] P H O T O A C T I V A T I O N OF T U B U L I N M O L E C U L E S IN N E U R O N S 355
The procedure to observe single microtubules is as follows:
1. Illuminate a small segment of the axon with a beam of near-ultravio-

let light.
2. After appropriate time intervals, permeabilize and fix the cells ac-
cording to the method described previously. 9 Permeabilize the cells for 5
min with 1% Triton X-100 in PHEM buffer (60 mM PIPES, 25 mM HEPES,
10 mM EGTA, 2 mM MgCI2, pH 6.9) containing 10 txM taxol and 0.2
M NaC1.
3. Fix the cells for 30 min by adding an equal volume of 1% glutaralde-
hyde in PHEM.
4. Wash with phosphate-buffered saline (pH 7.2) (PBS) and treat the
cells with 50 mM glycine in PBS for 30 min.
5. Incubate the cells with blocking solution containing 5% skimmed
milk in PBS.
6. Incubate the cells overnight at 4 ° with the antifluorescein antibody.
Antiserum is used at a dilution of 1 : 500 in PBS with 0.5% skimmed milk.
7. Wash the cells for 1 hr with PBS.
8. Incubate the cells with a 5-nm gold-conjugated anti-rabbit antibody
(Amersham Life Science) for 5 hr at a dilution of 1 : 10 in PBS at 37°.
9. Wash the cells completely with PBS and fix again with 1% glutaralde-
hyde in PBS and process for silver enhancement (Amersham).
10. Osmify (for 1 min), block stain, dehydrate, and embed the cells.
11. Serially section the cells and examine the block.
The procedure to preserve the soluble population of tubulin molecules

is as follows:
1. Illuminate a small segment of the axon with a UV beam.

2. Fix the cells 1-20 min after photoactivation with 0.1% glutaraldehyde
and 2% paraformaldehyde in PEM.
3. Treat the neurons with 1% Triton X-100 in PBS for 30 min.
4. Wash with PBS.
5. Incubate with 50 mM glycine in PBS for 30 rain.
6. Incubate the cells with blocking solution containing 5% skimmed
milk in PBS for 30 rain.
7. Incubate the samples overnight with affinity-purified antifluorescein
antibody in BSA-PBS (0.5% (w/v) BSA and 0.1% (w/v) gelatin in PBS)
containing 0.5% (w/v) skimmed milk at 4 °. Completely wash the cells with
BSA-PBS containing 0.5% skimmed milk.
8. Incubate the cells overnight at 4° with a 1.4-nm gold-conjugated anti-
rabbit Fab' fragment (Nanoprobes, Inc., Nanogold anti-rabbit) diluted 500
times in BSA-PBS containing 1% skimmed milk, 500 mM NaC1, and 0.05%
Tween 20 (the 5 nM gold-conjugated antibody usually employed for immu-

noelectron microscopy does not penetrate D R G neurons that have not
been permeabilized before fixation).
9. Completely wash (about 1 hr) the cells with BSA-PBS containing
skimmed milk, NaC1, and Tween 20 and then with PBS.
10. Fix the cells again with 1% glutaraldehyde in PBS.
11. Silver enhancement in developer solution (30% gum arabic, 0.85%
hydroquinone, 0.11% silver lactate, citric acid monohydrate, 25.5 g/liter,
sodium citrate dihydrate, 23.5 g/liter) 1° for about 1 hr.
12. Block stain, dehydrate, and embed cells without osmification for
serial sectioning and observation.
To evaluate this procedure and to check whether labels found outside
of the photoactivated regions really represent moving tubulin molecules in
the axons, we illuminated small axonal regions of previously fixed cells
with near-ultraviolet light, stained and observed in the electron microscope
as described earlier. In these axons, no gold labels were found out of
the photoactivated regions (Fig. 4d). No translocated microtubules were
detected in cells that had been fixed after permeabilization. In photoacti-
vated axons of cells fixed without permeabilization, where fluorescently
labeled oligomers, heterodimers, and polymers are preserved, a significantly
higher amount of gold label was found in regions distal to the photoactivated
regions than in the proximal region. These data indicate that tubulin mole-
cules are not transported as polymers, but as heterodimers or oligomers
by an active mechanism rather than by simple diffusion.
10 M. A. Hayat, "Principles and Techniques of Electron Microscopy." CRC Press, Boca Raton,
FL, 1989.
[20] T W o - P h o t o n A c t i v a t i o n o f C a g e d C a l c i u m w i t h
Submicron, Submillisecond Resolution
By E D W A R D B. BROWN and WATT W. WEBB
In~oduc~on
Photolabile calcium chelators, or "cages," are a popular tool to explore
the calcium dynamics of a variety of biological systems. On absorption of
a single UV or short-wavelength visible photon a calcium cage can undergo
a conformational shift and release a calcium ion. This release can take

Tween 20 (the 5 nM gold-conjugated antibody usually employed for immu-

noelectron microscopy does not penetrate D R G neurons that have not
been permeabilized before fixation).
9. Completely wash (about 1 hr) the cells with BSA-PBS containing
skimmed milk, NaC1, and Tween 20 and then with PBS.
10. Fix the cells again with 1% glutaraldehyde in PBS.
11. Silver enhancement in developer solution (30% gum arabic, 0.85%
hydroquinone, 0.11% silver lactate, citric acid monohydrate, 25.5 g/liter,
sodium citrate dihydrate, 23.5 g/liter) 1° for about 1 hr.
12. Block stain, dehydrate, and embed cells without osmification for
serial sectioning and observation.
To evaluate this procedure and to check whether labels found outside
of the photoactivated regions really represent moving tubulin molecules in
the axons, we illuminated small axonal regions of previously fixed cells
with near-ultraviolet light, stained and observed in the electron microscope
as described earlier. In these axons, no gold labels were found out of
the photoactivated regions (Fig. 4d). No translocated microtubules were
detected in cells that had been fixed after permeabilization. In photoacti-
vated axons of cells fixed without permeabilization, where fluorescently
labeled oligomers, heterodimers, and polymers are preserved, a significantly
higher amount of gold label was found in regions distal to the photoactivated
regions than in the proximal region. These data indicate that tubulin mole-
cules are not transported as polymers, but as heterodimers or oligomers
by an active mechanism rather than by simple diffusion.
10 M. A. Hayat, "Principles and Techniques of Electron Microscopy." CRC Press, Boca Raton,
FL, 1989.
[20] T W o - P h o t o n A c t i v a t i o n o f C a g e d C a l c i u m w i t h
Submicron, Submillisecond Resolution
By E D W A R D B. BROWN and WATT W. WEBB
In~oduc~on
Photolabile calcium chelators, or "cages," are a popular tool to explore
the calcium dynamics of a variety of biological systems. On absorption of
a single UV or short-wavelength visible photon a calcium cage can undergo
a conformational shift and release a calcium ion. This release can take

[9.0] TWO-PHOTON ACTIVATION OF CAGED CALCIUM 357
place as rapidly as 10/~sec after photoexcitation, 1 allowing calcium levels

to be locally altered with high temporal resolution. Photolysis of calcium
cages has been used to study cytosolic calcium oscillations 2 and membrane
ion channels. 3 The field of muscle contractility has also benefited from the
use of calcium cages,4-6 as has the study of exocytosis.7'8
Calcium cages have commonly been photolyzed using UV flash lamps
or focused laser beams. 9 A focused laser beam generates a double cone of
illumination with the narrowest point at the focal plane of the objective
lens, so within the focal plane the spatial resolution of conventional calcium
uncaging can be quite high. However, significant excitation of cage mole-
cules can occur above and below the focal plane. If a series of slices are
envisioned running through the double cone of focused illumination, paral-
lel to the focal plane and at increasing distances from it, conservation of
energy dictates that the same number of excitation photons must pass
through each slice. This means that the same total amount of cage excitation
will also be generated in each slice, with the result that conventional excita-
tion of calcium cages (or any chromophore) has essentially no resolution
along the axis of the focused excitation light.
Two-photon excitation (TPE) is a recently developed microscopy tech-
nique that can excite chromophores, including calcium cages, with intrinsic
three-dimensional resolution) ° Two-photon excitation of a UV-excitable
cage can be accomplished using a single near-IR (~700 rim) focused laser
beam with a wavelength chosen such that each individual photon generated
by the laser has roughly half the energy required to excite the cage. Two
such photons can be absorbed by the cage in a single quantum event, II
provided that the photons interact with the cage at essentially the same
time (within ~10 -16 sec). The requirement for simultaneous absorption of
two photons gives TPE a quadratic dependence on the incident intensity,
1 G. Ellis-Davies, J. Kaplan, and R. Barsotti, Biophys. J. 70, 1006 (1996).

2 G. M. Calder, V. E. Frank-Tong, P. J. Shaw, and B. K. Drobak, Biochem. Biophys. Res.
Commun. 234, 690 (1997).
3 p. Velez, S. Gyorke, A. Escobar, J. Vergara, and M. Fill, Biophys. J. 72, 691 (1997).
4 j. R. Patel, K. S. McDonald, M. R. Wolf, and R. L. Moss, J. Biol. Chem. 272, 6018
(1997).
5 B. K. Hoskins, P. J. Griffiths, C. C. Ashley, and R. Rapp, Biophys. J. 72, A275 (1997).
6 T. D. Lenart, J. M. Murray, C. Franzini-Armstrong, and Y. Goldman, Biophys. J. 71,
2289 (1996).
7 A. F. Oberhauser, I. M. Robinson, and J. M. Fernandez, Biophys. J. 71, 1131 (1996).
8 T. D. Parsons, G. C. R. Ellis-Davies, and W. Almers, Cell Calcium 19, 185 (1996).
9 j. p. y. Kao and S. R. Adams, in "Optical Microscopy: Emerging Methods and Applications"
(B. Herman and J. L. Masters, eds.), p. 27. Academic Press, London, 1993.
10W. Denk, J. Strickler, and W. Webb, Science 248, 73 (1990).
11 M. G6ppert-Mayer, Ann. Phys. 9, 273 (1931).
with the consequence that excitation is intrinsically confined to a small

region in the immediate vicinity of the focal point of the focused laser
beam. A brief flash of focused near-IR light can therefore generate a
localized elevation of calcium over a volume of less than a femtoliter.
Chromophores generally have extremely small multiphoton absorption
cross sections, requiring pulsed ("mode-locked") lasers in order to generate
photon densities high enough to accomplish significant TPE. The extremely
localized nature of this distribution means that the elevated soluble calcium
ion concentration will rapidly dissipate due to diffusion of the liberated
ions into the surrounding volume, as well as due to uptake of the ions by
local calcium buffers.
This work explores the spatial resolution of calcium release possible
with TPE and reviews the equipment necessary to accomplish this release.
We will calculate the quantities of released calcium that can be achieved
with a TPE instrument, using previously measured two-photon uncaging
action cross sections. Finally, we will explore the temporal behavior of the
released calcium concentration distribution as it is influenced by diffusion
out of the small TPE volume, as well as by intrinsic and extrinsic calcium
buffers. These calculations apply to TPE of any caged compound with
substitution of the appropriate diffusion and cellular uptake times and
illustrate general principles of the spatial and temporal regimes that can
be accessed with TPE. A more detailed treatment of many of these experi-
ments and calculations can be found in the source referenced in footnote 13.
A "user's summary" of relevant calculations and physical data reveals
that complete photolysis of calcium cages using TPE is indeed possible.
Using 700-nm light from a commercially available Ti : sapphire mode-locked
laser (100-fsec pulse length, 80-MHz repetition rate), focused with a 1.3
numerical aperture (NA) lens into a solution containing the calcium cage
with the highest known uncaging action cross section, Azid-1,12 essentially
all the cage molecules within an ellipsoidal volume of radial dimension 0.25
/.~m and axial dimension of 0.86 ~m can be photolyzed with a 10-/zsec flash
of - 7 mW average power at the sample. 13This highly localized distribution
of free calcium ions dissipates rapidly due to diffusion of calcium ions out
of the two-photon focal volume and calcium uptake by intrinsic and extrinsic
calcium buffers. In the absence of calcium buffers, the elevated calcium
concentration at the center of the TPE focal volume will last for -32/zsec
(FWHM), whereas the introduction of buffers can significantly shorten
this duration.
12 S. R. Adams, V. Lev-Ram, and R. Tsien, Chem. and Biol. 4~ 867 (1997).

13E. Brown, J. Shear, S. Adams, R. Tsien, and W. Webb, manuscript submitted (1997).
[20] TWO-PHOTON ACTIVATION OF CAGED CALCIUM 359
fl
FIG. 1. Spatial distribution of cage activation. (a) One-photon activation. The diffraction
limited intensity distribution? 4 with the direction of laser propagation along the z axis (oriented
vertically on the page). The effective local uncaging rate per cage-calcium complex via one
photon excitation is proportional to the intensity. Note the poor confinement of the distribution
in the z dimension. (b) Two-photon activation. The square of the diffraction limited intensity
distribution. The effective local uncaging rate per cage-calcium complex via two photon
excitation is proportional to the intensity squared. Note the superior confinement of the
distribution in the z dimension.
Spatial Resolution of Two-Photon Excitation

The exact representation of the average rate of one- and two-photon
excitation (O.P.E. and T.P.E.) for a single cage molecule illuminated by a
focused laser beam is
O.P.E. = tr(I(r, z)) (1)
T.P.E. = (1/2)~(/2(r, z))
where o"is the one-photon excitation cross section in units of c m 2 and 8 is
the two-photon excitation cross section in units of cm4 sec -1 (10 -5o cm4 sec -1
equals one GSppert-Mayer, or GM). The bracket symbols, ( ) , represent a
time average, while I(r, z) is the local excitation intensity. The highest
spatial resolution possible with a given objective lens and given illumination
wavelength is accomplished by uniformly filling the back aperture of the
objective with the laser beam, producing the "diffraction limited" intensity
14D. Sandison and W. Webb, J. App. Optics 33, 603 (1994).

distribution. This diffraction-limited intensity distribution is plotted in

Fig. 1. The first power of this distribution (Fig. la) is proportional to the
one-photon excitation probability and has poor confinement along the z
axis, demonstrating the poor three-dimensional resolution of this excitation
method. The second power of this distribution (Fig. lb) is proportional to
the two-photon excitation probability and exhibits the spatial confinement
intrinsic to two-photon excitation.
The exact form of the diffraction-limited intensity distribution is quite
complicated 15 and not generally amenable to use in mathematical analysis.
The square of this distribution, however, is closely approximated by the
mathematically tractable "ellipsoidal Gaussian"
12 (r, z ) : I2e -4r2/w2r e -4zz/w~ (2)
where I0 is the intensity at the center of the focal spot, The 1/e 2 characteristic
radial and axial dimensions are Wr and wz, respectively. These characteristic
dimensions ultimately dictate the spatial resolution. Values of Wr and wz can
either be measured directly or the following approximations can be used;
2.6A
Wr -- 2 r t N A
8.8nA (3)
Wz = 27rNA2
where h is the wavelength of the excitation light in vacuum, n is the index

of refraction of the sample, and NA is the numerical aperture of the illumi-
nation optics. For 700-nm light overfilling a 1.3 NA lens, therefore, the
two-photon focal volume will have a radial dimension of 0.22/~m and an
axial dimension of 0.77/~m. These increase to 0.29 and 1.3 txm for a 1.0
NA lens, demonstrating that the axial resolution of two-photon excitation
decreases rapidly with decreasing NA.
One drawback of using an overfilled objective lens to photolyze cages
using two-photon excitation is that this method is relatively wasteful of
laser power. The transverse intensity profile of lasers operating in the
TEM00 mode is a Gaussian with a characteristic width on the order of a
millimeter. In order to overfill an objective lens, the width of this profile
is greatly expanded using a pair of lenses such that the intensity distribution
actually entering the objective lens is essentially uniform. This means that
a large fraction of the laser power is rejected by the back aperture stop of
the objective lens. This wastefulness is not a significant problem with the
common mode-locked tunable Ti : sapphire lasers in use today, with average
power outputs of a watt or more. A growing trend in two-photon micros-
25 M. Born and E. Wolf, "Principles of Optics," p. 436. Pergamon Press, Tarrytown, NY, 1980.
copy, however, is the use of single-wavelength, all solid-state lasers that

generally have extremely limited power outputs, on the order of 50 mW.
Properly overfilling an objective lens with a laser of such limited total
output reduces the laser power reaching the sample to levels so small that
it becomes problematic to accomplish any significant uncaging. Instead of
overfilling the objective, however, the lens can be completely underfilled,
allowing the entire laser power (minus the -10-30% absorption of the
objective glass) to reach the sample. The resultant intensity distribution,
known as the "Gaussian-Lorentzian" distribution, has significantly lower
axial and radial resolution than the diffraction limited intensity distribution.
The square of the Gaussian-Lorentzian distribution is
412oW4 e_4r2/w2(z)
I2(r, z ) - rrw4(z )
(4)
w2(z) = 1+ /
where w0, the minimum radial beam waist, is dictated by the size of the
laser in the back aperture of the objective and the focusing properties of
the objective.
This work is concerned with the generation of calcium distributions of
the highest spatial and temporal resolutions, and consequently will deal
exclusively with diffraction-limited optics, using the intensity profile given
by Eq. (2). Most of the results presented here are easily extrapolated to
the Gaussian-Lorentzian distribution.
Apparatus for Two-Photon Uncaging

A variety of two-photon photolysis geometries are possible, and two
variants using the commercially available two-photon microscopy system
sold by Bio-Rad (Hercules, CA) are shown in Fig. 2. In both cases, multipho-
ton excitation is generated with a mode-locked Ti:sapphire laser, rapid
modulation of the near-IR beam is accomplished with a Pockels Cell (Con-
optics, Danbury, CT), and the imaging system consists of a Bio-Rad MRC
1024 scanning box and a microscope such as the Zeiss Axiovert 135 (Zeiss,
Jena, Germany) or the Olympus BX50WI (Olympus America, Inc., Mel-
ville, NY.). The extent to which the objective lens is overfilled can be
adjusted with a separate beam expander, whereas dispersion compensation
can be accomplished using a pair of SFI0 equilateral prisms. Measurement
of the two-photon spot size can be accomplished with an objective-mounted
piezoelectric translator.
In the first variant, using a multiband dichroic available with the Bio-
Rad system, the 488-nm argon laser line is used to generate conventional
"• Pockels
Cell
~ Dichroi¢
Mirror
~ Beam
Expander
~ Objective
~ SampleChamber
FIG. 2. Cage activation apparatus configuration. (a) The confocal imaging/two-photon

photolysisexperimentalapparatus. (b) The two-photonimaging/two-photonphotolysisexperi-
mental apparatus.
one-photon confocal images of the target area, whereas photolysis is accom-
plished using the Ti:sapphire beam directed along the same excitation
path. During the imaging process, the near-IR beam is completely attenu-
ated using the Pockels cell, and a confocal image is formed of the target
area. The scan mirrors are then used to direct to the target point a high
intensity flash of Ti:sapphire light transmitted through the Pockels cell.
The subsequent effects of the photolysis are studied using confocal line
scans, full images, or other one-photon fluorescence microscopy techniques.
The advantage of this technique is that the imaging beam cannot cause
unwanted cage photolysis even if the beam dwells in one location for an
extremely long time due to slow scan speeds or high zoom factors. The
disadvantages of this technique is that the photolysis experiments are lim-
ited to the relatively thin specimens available for study with confocal micros-
copy, and the significant out-of-plane photobleaching caused by one photon
excitation during the imaging process will limit the length of cell viability.
[201 TWO-PHOTON ACTIVATION OF CAGED CALCIUM 363
~der
670/p
PMT Dichroic
Mirror
Objective
[~ Sample
Chamber
FIG. 2. (Continued)
In the second variant, using a simple 670-nm short pass dichroic, the
same two-photon beam is used both to image the target area and to pho-
tolyze the calcium cage. During imaging, the Pockels cell attenuates the
near-IR beam to low powers, sufficient to generate a fluorescent signal but
not cause significant photolysis of the calcium cage. The Pockels cell then
generates a rapid high power pulse when the scan mirrors have directed
the excitation beam to the target area, and various two-photon microscopy
techniques (line scans, full images, etc.) can then be used to study the
ensuing biological effects. The disadvantage of this technique is that the
imaging powers transmitted by the Pockels cell must be carefully chosen
to be low enough that no significant uncaging is caused during image
formation. The advantage of this technique is that it uses two-photon excita-
tion in the imaging process, with the concomitant benefits this has over
confocal microscopy; greater imaging depth and longer sample lifetimes
due to reduced photobleaching. The Pockels cell is ideally suited to power
modulation of mode-locked lasers because of its rapid response time and
wavelength insensitivity, making it superior to shutters and acousto-optical
modulators (AOMs) for this purpose. This device works by the application
364 CAGEOCOMPOUNDS [20]
of a high voltage to a set of capacitor plates enclosing a nonlinear optical

material. The laser passes through this material, and the plane of polariza-
tion of the beam is rotated through an angle dictated by the voltage applied
to the plates. An analyzing polarizer is used after the nonlinear material
to translate the variation in laser polarization angle into a variation in laser
intensity. Pockels cells can easily modulate the transmitted laser power by
over a factor of 100 with a response time of a microsecond or less. As will
be discussed later, this sort of rapid response time is necessary for TPE
generation of calcium concentration elevations with the highest spatial
resolution. This temporal requirement eliminates most common mechanical
shutters, which have response times of - 1 msec, although pairs of shutters
may be used to generate more rapid beam modulation. The pulsed nature
of the lasers required for efficient generation of two-photon excitation
makes the use of AOMs problematic, although they have sufficiently rapid
response times for highly localized generation of calcium. An AOM relies
on the diffraction of laser light through the grating generated by an acoustic
standing wave in a transparent crystal. A 100-fsec pulse of laser light has a
bandwidth of - 1 5 nm, which can produce undesirable spatial and temporal
distortion of the beam as it passes through the diffractive region of the
AOM. This leaves Pockels cells as the most viable alternative for rapid
modulation of pulsed laser beams.
Dispersion compensation using a pair of SF10 prisms allows the maxi-
mum two-photon excitation to be generated with a given average laser
power. Group delay dispersion in optical materials causes temporal broad-
ening of transmitted laser pulses, lowering the peak photon densities and
thereby reducing TPE. This dispersion arises because the index of refraction
of the glass in optical components, such as objective lenses, beam expanders,
and tube lenses, varies with wavelength. The speed of light through a
material is inversely proportional to this index of refraction, so the speed
of light through a material also scales with wavelength. A 100-fsec pulse
of light, with - 1 5 nm of bandwidth, will therefore broaden in time when
it passes through an optic because its constituent wavelength components
travel at different speeds, with longer wavelengths outrunning shorter ones
(positively dispersive) in the Ti : sapphire wavelength range. An appropri-
ately placed pair of prisms will disperse the laser pulses in the opposite,
negatively dispersive, sense, 16 effectively giving the long wavelengths a
shorter overall optical path length so that the laser pulse will be recombined
by passage through the positively dispersive optics and have a pulse width
at the sample that is close to its original, nondispersed temporal width.
An SF10 prism pair separated by approximately 60 cm provides sufficient
16 R. L. Fork, O. E. Martinez, and J. P. Gordon, Opt. Lett. 9, 150 (1984).

[201 TWO-PHOTON ACTIVATION OF CAGED CALCIUM 365
negative dispersion to counter the broadening induced in 100-fsec pulses by

a typical high NA objective, a Pockels cell, and a two-lens beam expander.
A method for the measurement of the characteristic axial and radial
distances of the two-photon focal volume generated at the sample is highly
desirable for direct determination of the spatial resolution of two-photon
calcium uncaging. A piezoelectric drive mounted on an objective lens mount
with a step size significantly smaller than the expected two-photon axial
dimension can be used to directly measure this dimension. The two-photon
focal volume can be stepped from within the coverslip to within a fluores-
cent solution of dye, and the resultant sigmoidal increase in fluorescence
can be easily fit to yield w~, the characteristic axial dimension of the two-
photon spot. Likewise, the two photon spot can be scanned transversely
across the interface between a fluorescent and nonfluorescent sample using
the scanning mirrors of the imaging system in order to determine wr,
the characteristic radial dimension of the two-photon spot. Immobilized
subresolution fluorescent beads are also suitable for these measure-
ments.
TWo-Photon Uncaging Action Cross Section of Calcium Cages

Once a two-photon uncaging system has been assembled, it is useful
to predict the amount of calcium release that is possible under different
illumination conditions. However, in order to predict the amount of calcium
that can be generated, it is necessary to know the uncaging action cross
section of the cage. A convenient method for determining the action cross
section of a calcium cage is to focus a photolysis pulse train into a solution
containing a calcium cage, calcium, and a high concentration of a rapidly
binding calcium indicator dye such as Fluo-3. If a low intensity near-IR
laser beam is focused into the sample, a steady fluorescence signal will
be generated via TPE of the equilibrium concentration of calcium-bound
indicator dye. If the near-IR beam is briefly flashed to a higher intensity,
some of the calcium-bound cage within the two-photon focal volume can
be photolyzed. This released calcium can then bind to the surrounding
indicator dye, resulting in an increase in fluorescence generated by the low-
intensity near-IR beam immediately after the photolysis pulse train. This
additional fluorescence will dissipate as the newly generated calcium-bound
indicator dye diffuses out of the two-photon focal volume. The amplitude
of the transient fluorescence increase generated by the photolysis pulse
train can be analyzed to determine the two-photon uncaging action cross
section of the calcium cageJ 3'17
17j. Shear, E. Brown S. Adams,R, Tsien, and W. Webb, Biophys. J. 70, A211 (1996).
Neglecting saturation, which will be addressed later, an uncaging pulse
train of duration At focused very briefly into a sample containing an equilib-
rium concentration of calcium-loaded cage, [CCa2+]0, will release the fol-
lowing distribution of calcium ions:
A[Ca2+(r, z)] = 1 - e (l/2)Su{I2(r'z))~'t[CCa2+]0 (5)
where 8u is the two-photon action cross section of the calcium cage and
consists of the two-photon absorption cross section multiplied by the quan-
tum efficiency of uncaging. Iu(r, z) is the spatial intensity distribution of
the focused laser excitation. Equation (5) assumes that At, the duration of
the uncaging pulse, is significantly shorter than the diffusion time of the
cage across the two-photon focal volume. For 700-nm light overfilling a 1.3
NA lens this is satisfied if Ate10/xsec. The newly liberated calcium enters
a solution with a high concentration of rapid indicator dye. If the rate of
uptake of liberated calcium by the indicator dye is significantly faster than
the rate of diffusion of the calcium across the two-photon focal volume,
then Eq. (5) also represents the concentration distribution of newly calcium-
bound indicator dye immediately after the end of the photolysis pulse. For
uncaging pulses generated with 700-nm light overfilling a 1.3 NA lens, this
requires at least - 1 mM of a rapid indicator such as Fluo-3.
The newly generated calcium-bound indicator dye is a nonequilibrium,
highly localized concentration distribution superimposed on the equilibrium
distribution of calcium-bound indicator. The fluorescence of this concentra-
tion distribution can be excited with the same laser, albeit greatly attenu-
ated, that photolyzed the cage. Assuming diffraction limited optics, the
fluorescence signal that is emitted by this newly calcium-bound indicator
as it dissipates out of the focal volume via diffusion is
Fo (1/2)8u(I2)At [CCa2+]0 1 1
AF(r) - 2 V~ [FCa2+]0(1 + r) (1 + Rr) 1/2 (6)
where F0 is the steady-state fluorescence signal generated by the attenuated

laser beam exciting the equilibrium concentration of calcium-bound indica-
tor dye. The equilibrium concentration of calcium-bound indicator dye is
[FCa2+]0 whereas the equilibrium concentration of calcium-bound cage is
[CCa2+]0. The quantity ris a dimensionless "time" equal to 8Dt/wr 2, where
D is the diffusion coefficient of the fluorescent indicator dye and R is the
square of the ratio of the two 1/e 2 beam waists; R = (wz/wr) 2. Equation
(6) assumes that only a small fraction of the available cage-calcium has
been photolyzed ((1/2)Su(/Zuu}At~ 1) in order to avoid significant depletion
of free calcium indicator dye in the focal spot. An example of a transient
fluorescent curve generated by two-photon photolysis of the calcium cage
Azid-1 in a solution of Fluo-3, along with the resultant fit using Eq. (6), is
shown in Fig. 3.
Three calcium cages have been studied with this technique: the commer-
cially available cages NPEGTA (Molecular Probes, Eugene, OR) and DM-
nitrophen (Calbiochem, La Jolla, CA), as well as a new calcium cage,
Azid-l.l: The response of these cages at selected wavelengths between 700
0.2
0.15-
0.1
0.05
J
O- • • 00 •
• • ~
• •
-0.05 t , I , , ,
-0.5 0 0.5 1 1.5 2 2.5 3 3.5
Time(ms)
FIG. 3. Dynamics of two-photon calcium cage activation detected by a fluorescent indicator.
The fluorescence signal generated by two-photon point photolysis of Azid-1 within a solution
of the calcium indicator dye Fluo-3. [Data From E. Brown, J. Shear, S. Adams, R. Tsien, and
W. Webb, manuscript submitted (1997).] The bright flash of 700-nm laser fight occurs at
t = 0 on this curve, generating newly calcium-bound Fluo-3, which produces an increase in
fluorescence signal AF(t). This fluorescence decays away to equilibrium levels as the new
Fluo-calcium diffuses out of the two-photon focal volume generated by the 1.2 NA objective
lens. The solid line is a fit to data using Eq. (6). The resultant uncaging action cross section
at this wavelength is --1.4 GM.
10 .::::2"2..2::::::::::::'22 .:=:2::!"'. _.:2::::::7".:=2:2::'2LL=:~::'""L..:22T "2...::::7"'" L..t2::2:'" L..2::27"". 2=22-"_[2:::: :'" .::7":" "
Azid-1 [iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!•!!!!!!!!!!!!!!!!!!!!!!
--'~DM-Nitrophen I .................
! !!!!!!!!!!!!!!!!!!!
o 0.1
0.01
¢3
0.001
0.0001 I I I !
700 720 740 760 780 800
Wavelength (nm)
FIG. 4. The two-photon uncaging action on cross sections of Azid-1 and DM-nitrophen
with data from 700 to 800 nm. NPEGTA did not produce enough calcium for an uncaging
signal to be detected above background noise at any of the wavelengths studied. [Data from
E. Brown, J. Shear, S. Adams, R. Tsien, and W. Webb, manuscript submitted (1997).]
and 800 nm were probed using a Ti:sapphire laser source, whereas 622-
nm light was provided using a frequency-doubled Cr: Lisgaf (unpublished
data). NPEGTA showed no significant reproducible calcium uncaging signal
at any of the wavelengths probed, DM-nitrophen displayed a peak uncaging
action cross section of -0.01 GM at 720 rim, and Azid-1 had a peak uncaging
action cross section of - 1 . 4 GM at 700 nm. Data for DM-nitrophen and
Azid-1 between 700 and 800 nm is shown in Fig. 4. The uncaging cross
section of DM-nitrophen at 622 nm is roughly a factor of six less than the
lowest value in the 700- to 800-nm wavelength range, and the cross section
of Azid-1 at 622 nm is roughly a factor of 40 lower than in the longer
wavelength range. Photolysis experiments using DM-nitrophen and
NPEGTA in living cells have qualitatively confirmed the relative efficacy
of these two cages. TM
18p. Lipp, J. Kleinle, C. Amstutz, C. Luschner, and E. Niggli, Biophys. Z 72, A44 (1997).
The fact that NPEGTA does not display any uncaging signal using this
technique does not necessarily mean that the multiphoton excitation action
cross section of this cage is vastly smaller than that of the species with
the smallest measurable action cross section at these wavelengths, DM-
nitrophen. Equation (6) describes a time-dependent fluorescent signal F(r)
superimposed on a baseline fluorescent signal F0 generated by the equilib-
rium loading of the fluorescent indicator dye. Within the range of uncaging
intensities used in these experiments (those that avoid significant dye
bleaching), the transient signal F(~') is much smaller than the baseline signal
F0. This means that the total noise in the fluorescence amplitude is going
to be dominated by the shot noise from the baseline fluorescence /70.
The desired signal is the change in fluorescence due to newly generated
Fluo-calcium AF(r), therefore the signal-to-noise (S/N) ratio of this tech-
nique, assuming that F0 is shot noise limited, is proportional to
S AF(r) [CCa2+]o
(7)
~[ ~ V~o oc t3u([FCa2+]o),/:
and is not due solely to the uncaging action cross section, but has important
contributions from the equilibrium loading of cage and indicator. Hence,
when all other factors are equal, the overall S/N is superior for cages that
have lower Kd values. DM-nitrophen has the lowest dissociation constant
of the three cages tested here, with a Kd of 5 nM, yielding the most favorable
environment for measuring the uncaging peak. The Kd of NPEGTA is 80
nM, more than 10 times greater than that of DM-nitrophen. This factor
reduces the S/N ratio for this species by more than threefold compared to
DM-nitrophen, making it that much more difficult to measure the uncaging
action cross section. The Kd of Azid-1 is -40-fold greater than that of DM-
nitrophen, 12 yielding an equilibrium distribution of calcium that reduces
the signal to noise ratio by a factor of six compared to DM-nitrophen.
The two-photon action cross section of Azid-1 is 140-fold larger, however,
providing sufficient overall signal to noise to measure calcium uncaging.
In conclusion, although NPEGTA certainly has a smaller action cross sec-
tion than DM-nitrophen, the lack of an uncaging signal using this technique
can also be contributed to the high equilibrium constant of this species,
which contributes to the difficulty in detecting uncaging signal from this
cage by decreasing the signal-to-noise ratio by a factor of three.
Total Calcium Yield via Two-Photon Excitation

When the two-photon action cross section of some calcium cages has
been measured, the total possible calcium release possible under specific
illumination conditions can be determined and the feasibility of two-photon
calcium uncaging experiments can be predicted. The total number of cal-
cium ions released by a photolysis pulse train of length At can be calculated
by integrating Eq. (5) over all space (assuming the total sample volume is
much larger than the two-photon focal volume). Assuming a diffraction
limited intensity profile, the total number of calcium ions, Nca, produced
by a photolysis pulse that is significantly shorter than the diffusion time of
the cage molecules across the two-photon focal volume is then
7T3/2. 2. ¢~
Nca= -NA[ffaa+]0 8rWz ~_1 (n!n312
-°0n (8)
where NA is Avogadro's number, [CCa2+]0again is the equilibrium concen-

tration of calcium-loaded cage, and o~ --- (1/2)ru(~)At is the "central uncag-
ing dose," representing in the low intensity limit the average number of
uncaging events per molecule at the center of the two-photon spot. Note
that in the shallow uncaging limit (a ~ 1), the total number of calcium
ions Nca does not scale with the NA of the objective lens, just as in the
limit of no saturation the total number of fluorescent photons generated
by TPE of a fluorophore does not scale with NA. This NA independence
of the "total excitation" is a general feature of TPE, whereas increasing
the NA will increase the peak intensity squared, (~) ~ NA 4 (assuming a
fixed total power), the size of the focal volume will decrease by the same
amount, and the two effects exactly cancel out. Once significant amounts
of cage-calcium complexes are photolyzed (a - 1), the total number of
calcium ions liberated does indeed scale with NA, albeit in a nonlinear
fashion.
Equation (8) contains terms proportional to the equilibrium number of
cage-calcium complexes in the two-photon focal volume, as well as terms
that describe the fractional release of these ions. This can be simplified
greatly by dividing the entire equation by the equilibrium number of cage-
calcium complexes in the two-photon volume and expressing the uncaging
production as a fractional yield. The two-photon focal volume is a poorly
defined region, however, with no specific boundary. A variety of expressions
can be chosen to represent it, such as the volume contained within the
lo/e 2 isointensity surface of the excitation beam or the volume contained
within the (1/2)Io isointensity surface. A useful definition of the two-photon
volume V arises naturally out of analysis of multiphoton fluorescence corre-
lation spectroscopy, howevera9;
V - [f WZ(r) dv]2 (9)

f W 4 ( r ) dv
19j. Mertz, C. Xu, and W. Webb, Opt. Lett. 20, 2532 (1995).
where the unitless intensity profile W2(r) is derived from Eq. (2) by setting
Io = 1. This volume is defined in the low excitation intensity limit without
ground-state depletion, or saturation, of the fluorophore undergoing TPE.
At high intensities, when ground-state depletion of fluorophores arises,
the spatial distribution of two-photon-excited fluorophores loses its three-
dimensional Gaussian shape as the fraction excited at the center of the
focal volume approaches its maximum value of one. At higher excitation
intensities, the spatial distribution of excited fluorophores begins to appear
more like a "top hat" with edges that move out from the focal point as
the excitation intensity increases. Consequently, Eq. (9) is most representa-
tive of the volume of the two-photon excitation spot in the low intensity
limit and underestimates the effective excitation volume at high excita-
tion intensities.
The equilibrium number of cage-calcium complexes contained in the
two-photon focal volume is just the equilibrium concentration of cage-
calcium complexes times the focal volume as defined for our intensity
distribution via Eq. (9). We can now define the uncaging ratio Ru as the
number of calcium ions generated by a photolysis pulse [Eq. (8)] divided
by the number of cage-calcium complexes available to be photolyzed in
this two-photon focal volume;
(10)
Ru - 4- n=l n!n3/2
For appropriate central uncaging doses o~between zero and 35, this expres-
sion can be approximated as
Ru ~ 2.54 - 1.967e -°'°543~ - 0.572e -°'4°s~ (11)
which is plotted in Fig. 5. An uncaging ratio of one corresponds to a release
of a total number of calcium ions equal to the number of cage-calcium
complexes contained in the two-photon excitation volume before the pho-
tolysis flash. The uncaging ratio does not asymptotically approach one,
however, as a - (1/2)8~(~)At grows large because the two-photon volume
[Eq. (9)] is defined in the low excitation limit. At high central uncaging
doses (large a), significant numbers of cage-calcium complexes that lie
outside this equivalent volume V, which is in the lower intensity periphery
of the illumination profile, can be photolyzed and, in fact, the uncaging
ratio and the effective volume of release grows without bound as the central
uncaging dose a approaches infinity.
Equation (10) or (11) can be used to predict the photolysis pulses
required to accomplish desired amounts of calcium release. The equilibrium
cage-calcium concentration can be set by the cage-loading conditions (the
amount of cage-calcium in the loading patch pipette, for example), and
372 CACZD COMPOUNDS [20]
2.5 i J , i J I
Ltncaging Ratio
1.5
0.5
0 5 10 15 20 25 30 35
Central Uncaging Dose ~=(1/2)~<12>~t
FIG. 5. Excitation dose dependent of calcium cage activation. The total calcium production
as a function of the central uncaging dose ~ = (1/2)Su(~)At. The total production is expressed
as a ratio Ru of the total number of calcium ions released to the total equilibrium number
of cage-calcium complexes in the two-photon probe volume and is calculated using Eq. (11).
the t w o - p h o t o n focal v o l u m e can be set with the illumination optics [see

Eqs. (2), (3), and (9)]. This dictates the total a m o u n t of c a g e - c a l c i u m
available for photolysis in the t w o - p h o t o n focal volume. All of this caged
calcium can then be released via a photolysis pulse that generates an uncag-
ing ratio o f one, which can be accomplished with a central uncaging dose
of ot -- (1/2)Su(~)At - 6. Given the r e p o r t e d p e a k action cross section for
Azid-1 of - 1 . 4 G M at 700 nm, this m e a n s that an average p o w e r of - 7
m W at the focal v o l u m e o f an overfilled 1.33 N A objective will achieve an
uncaging ratio of one in 10/zsec. D M - n i t r o p h e n requires an average p o w e r
of ~ 8 5 m W of 720-nm light to achieve the same results.
The uncaging intensities required for significant, rapid, photolysis of

DM-nitrophen are high enough to cause fluorescence saturation, or ground-
state depletion, if applied to some fluorescent molecules, so it is conceivable
that excitation saturation limits the maximum possible uncaging rate, just
as it limits the maximum possible rate of fluorescence.
The kinetics of excitation saturation using pulsed lasers 2° are different
from the kinetics using continuous illumination. 21 If commercially available
mode-locked Ti:sapphire lasers are exciting common cages, the excited
state lifetimes I of the cages are generally much longer than the duration
of individual laser pulses (nanoseconds versus femtoseconds), with the
result that a cage can be excited no more than once per laser pulse. Conse-
quently, the maximum rate of excitation of cages using pulsed lasers is
limited to the repetition rate of the laser and will not scale quadratically
with the excitation power if a significant fraction of molecules are excited
during a single laser pulse. This will limit the maximum rate of calcium
release possible with a calcium cage. If the uncaging quantum efficiency of
a calcium cage is sufficiently high, however, essentially all of the cage in a
given region of the two-photon focal volume will be photolyzed during an
uncaging flash with intensities that are lower than intensities required for
excitation saturation, and this limit will be irrelevant.
The minimum uncaging quantum efficiency required to render excita-
tion saturation irrelevant for a given experiment can be calculated by model-
ing the laser pulse train as a sequence of square pulses of intensity I,
duration r, and repetition rate ~0. The fraction Fa, of cage-calcium com-
plexes photolyzed during an uncaging flash of duration At is then
FAt = 1 -- e -(1/2)qS12°~rAt (12)
whereas the fraction of cage-calcium complexes excited during a single

laser pulse in the uncaging flash, assuming that the intensity is low enough
to avoid saturation, is
F~ = (1/2)8/2r (13)
We can therefore solve for the uncaging quantum efficiency q that allows
the fraction Fat of cage-calcium complexes photolyzed during an uncaging
flash to reach 0.9 (an arbitrarily chosen value that represents nearly com-
plete photolysis) while restricting the fraction F~ of cage-calcium complexes
excited during a single laser pulse to 0.1 (an arbitrarily chosen value that
20 C. Xu and W. Webb, in "Topics in Fluorescence Spectroscopy: Volume 5 of Multi-Photon

Excitation and Light Quenching" (J, Lackowicz, ed.). Plenum Press, New York, 1997.
21 D. Sandison, R. Williams, K. Wells, J. Strickler, and W. Webb, in "Handbook of Biological
Confocal Microscopy" (J. Pawley, ed.), p. 39. Plenum Press, New York, 1995.
is certainly below saturation). Setting Eq. (12) equal to 0.9, substituting 0.1
for ~/2r, and solving for q reveals
23
q~ (14)
o~At
If the quantum efficiency of uncaging is greater than this value, excitation
saturation will not effectively limit the total calcium release possible with
a flash of duration At because 90% of the cage-calcium complexes will
have been photolyzed in a given location at excitation intensities where
only 10% of the complexes are excited with each laser pulse, still well below
saturation. The calcium yield of DM-nitrophen, with a quantum efficiency22
of 0.18, will be saturation limited at high intensities only for photolysis
pulses that are 1.6 tzsec or shorter (assuming a laser repetition rate of 80
MHz). Azid-1, with a quantum efficiencyle of 1, is even less sensitive to
excitation saturation. This demonstrates that excitation saturation will not
be a limiting factor in most photolysis experiments, because for uncaging
pulses longer than ~2 tzsec, the quantum efficiencies of these two calcium
cages are high enough that nearly complete photolysis is accomplished at
local intensities well below saturation.
Temporal Behavior of Three-Dimensionally Localized

Calcium Distributions
The temporal behavior of the calcium concentration distribution re-
leased by a multiphoton photolysis pulse is an important factor in the design
of uncaging experiments. This behavior can be influenced by the duration
of the uncaging pulse, the overall central uncaging dose, the diffusional
mobility of the calcium ions, the intrinsic release times of the calcium cage,
and the rate of uptake of calcium ions by local buffers, both extrinsic
and intrinsic. The influence of buffering on calcium pulses generated via
conventional one-photon uncaging has been widely studied. 23,24 However,
the extremely localized nature of two-photon uncaging greatly increases
the influence of diffusion on the postphotolysis kinetics of the liberated
calcium distribution, warranting another look at the problem.
The complex nature of the interplay between buffering and diffusion
over the subfemtoliter spatial scales of multiphoton uncaging makes exact
calculation of the resultant calcium concentration distributions problematic.
Advances in the related problem of buffered calcium influx through mem-
22 j. H. Kaplan and G. Ellis-Davies, Proc. Natl. Acad. Sci. U.S.A. 85, 6571 (1988).
23 T. Xu, M. Naraghi, H. Kang, and E. Neher, Biophys. J. 73, 532 (1997).
24 A. Fleet, G. Ellis-Davies, J. Kaplan, and S. Bolsover, Soc. Neurosci. Abs. 21, 775 (1995).
[9.0] TWO-PHOTON ACTIVATION OF CAGED CALCIUM 375
brane channels have involved numerical integration of the relevant equa-

tions in order to determine the steady-state distributions of calcium ions
in the vicinity of an open calcium channel. 25 In order to gain a physical
understanding of the dynamics of the calcium concentration after a photoly-
sis pulse without attempting a numerical solution of the relevant differential
equations, we will restrict our attention to the temporal behavior of the
calcium concentration at the center of the two-photon focal volume (the
peak concentration) and generate a model of this behavior using a few
approximations and taking a few limits that greatly simplify the problem.
This model will allow us to calculate the concentrations of the three popula-
tions that calcium ions flow between during and immediately after an uncag-
ing pulse; they are the population of ground-state cage-calcium complexes,
of photoexcited cage-calcium complexes, and of free calcium ions.
The first simplifying approximation is to treat the diffusion of free
calcium ions out of the two-photon focal volume as a simple first-order
process with a characteristic escape rate ke given by the reciprocal of the
average dwell time. For NA between 0.6 and 1.3 we will approximate this as
1.2Dca
ke ~- 2 (15)
Wr
where Dca is the diffusion coefficient of free (unbuffered) calcium ions and
Wr is the 1/ee radius of the excitation volume. The second simplification is
to restrict the amount of calcium generated with an uncaging pulse train
such that the peak concentration of liberated ions is significantly lower
than the local concentration of calcium buffers. In this limit there is no
local buffer saturation and we can assign a single characteristic uptake rate
constant for intrinsic cellular buffers, ki, and extrinsic buffers, kx, where
this uptake rate constant is the usual "on rate" of the buffer, with units of
M -1 sec -1, times the local concentration of the buffer. For freely diffusing
calcium buffers with "off rates" similar to Fluo-3 (kon - 240 sec -1) based
on the equilibrium constant and the "on rate, ''13 the newly calcium-bound
buffers will have had ample time to diffuse away from the two-photon
probe volume before approaching equilibrium with the new calcium distri-
bution (i.e., for the buffer, ke > korf) and we will therefore assume they
act only as calcium sinks, not as calcium sources. We can now assign a
single rate constant, ko ~ ke + ki + kx, to account for all the processes
that remove free calcium from the two-photon focal volume.
Two other characteristic rates relevant to this problem are the rate at
which ground-state cage-calcium complexes are photoexcited, (1/2)3u(~),
and the rate at which photoexcited cages release calcium, kr. This intrinsic
25 j. Klingauf and E. Neher, Biophys. J. 72, 674 (1997).

release rate is set by the time required for the internal conformational
shifts of a photoexcited calcium-cage complex, which lead to the release
of the calcium ion. For the three cages studied here, this rate is on the
order of - 1 0 5 sec 1.1,12
For some calcium cages there is another relevant rate constant; the
rate at which photoexcited cage-calcium complexes revert to ground state
cage-calcium complexes without releasing calcium. The presence of this
extra pathway can add significant complexity to the present calculation.
We therefore explicitly consider Azid-1 in these calculations because its
quantum efficiency is close to one, 12 suggesting that this pathway can be
neglected. The temporal behavior that the following calculations predict is
also valid for cages of nonunity quantum efficiency, but the overall ampli-
tude differs by a constant factor. 13
Using the rates defined earlier, we can now define an extremely simple
dynamical relationship between the concentration of the species relevant
when a photolysis pulse acts on a solution of caged calcium
( 1/2)~'u{I2 ) kr
[CCa2+(0, 0; t)] , [C'Ca2+(0, 0; t)] ,
k (16)
A[Ca2+(0, 0; t)] ° ; Escape
where [CCa2+(0, 0; t)] is the concentration of ground-state cage-calcium

complexes at the center of the two-photon spot, [C*cae+(0, 0; t)] is the
concentration of photoexcited cage-calcium complexes at the center of the
two-photon spot, A[Ca2+(0, 0; t)] is the concentration of liberated calcium
ions at the center of the two-photon spot, and "Escape" represents escape
of liberated calcium ions from the system either through diffusion out of
the two-photon probe volume or chelation by buffers.
We can now form the relatively simple system of differential equations
that describe the behavior of these three populations:
I,cca20,0t,]I l/2u/ 0 :l cca

[C*Ca2+(0, 0; t)] = °°tJl,
C*Ca2
00t,
(1/2)fiu(IZu) -kr
A[CaZ+(0, 0; t)] 0 kr -ko L A[Ca2+(0' 0; t)] 3
(17)
In order to generate highly localized calcium distributions of high con-
centration, rapid uncaging pulses (At ~ ko 1) of high uncaging rate
(1/2)26u(~) >> kr, ko) will be used. Under these conditions, the free calcium
concentration at the center of the two-photon focal volume generated by

an uncaging pulse is given by
kr
a[Ca2+(0, 0; t)] = (kr - ko) (e-k°' - e-krt)[CCa2+]° (18)
where [CCa2+]0 is the equilibrium concentration of cage-calcium com-

plexes. In this high uncaging rate and rapid uncaging pulse limit, the peak
calcium concentration generated does not scale with the excitation rate,
and the duration of the elevated calcium distribution does not scale with
the uncaging pulse duration.
We can now use Eq. (18) to predict the temporal behavior of the peak
calcium concentration under a variety of experimental conditions. The
simplest case is one in which cage is phytolyzed via two-photon excitation
in vitro with no extrinsic buffer added. This is also the case that yields the
longest possible localized elevations of free calcium with a given illumina-
tion pattern. The release rate k r is assumed to be 105 sec -1, and the total
rate constant for free calcium loss, ko, is due only to diffusional escape
(ko = ke). The diffusional escape rate of free calcium ions from the two-
photon spot generated with a 1.33 NA lens overfilled with 700-nm light is
ke - 7 × 104 sec -1, whereas from a 0.5 NA lens it is ke - 1 × 1 0 4 s e c -1.
The predicted peak calcium concentration as a function of time after the
rapid photolysis pulse is plotted in Fig. 6. We see that the calcium elevations
generated with the superior spatial resolution of high NA optics can last
up to -32/xsec FWHM, Both the amplitude and the duration of the free
calcium pulse decrease with increasing NA due to the faster diffusional
escape time from the smaller focal volume of the higher NA objective.
Note, however, that producing similar central uncaging doses while lowering
the NA of the objective requires a significant increase in input power.
The next case to consider is two-photon uncaging within a living cell.
Prediction of the peak free calcium concentration dynamics now requires
knowledge of the intrinsic buffering rate, kj, extrinsic buffering rate, kx,
and their relation to the characteristic diffusional escape time, k~. The
intrinsic buffering dynamics are is not known for most experimental sys-
tems and are, in fact, an attractive subject of study for two-photon cal-
cium uncaging. The extrinsic buffering rate, kx, is generally known because
it is controlled by the quantity and type of calcium buffer added to the
system, usually in the form of a fluorescent calcium indicator such as
Fluo-3 but also including any calcium-free unphotolyzed cage33 In order
to explore how buffering interacts with diffusion out of the two-photon
spot, we will assume that the total buffering rate in the cellular system
(ki + kx) is equivalent to that exerted by 100/xM of Fluo-3 (a reasonable
concentration of indicator that will maintain cell viability). We can then
0.7
0.6
[ca(o,o;t)]
...... t~;b--- 0.5
0.4
0.3
0.2
0.1
0
0 20 40 60 80 100
Time (microseconds)
FIG. 6. Buffer and diffusion dependence of the kinetics of the maximum calcium concentra-
tion. Free calcium concentration at the center of the two-photon focal volume, as given by
Eq. (18). The concentration is expressed as a ratio of the local peak liberated calcium concentra-
tion to the equilibrium concentration of cage-calcium complexes before the uncaging pulse.
Compare heavy or light lines to observe the effect of calcium buffers, and compare solid or
dashed lines for the effect of varying the NA of the objective. The dotted line is calculated
using experimentally measured values of the cellular average concentration ( - 4 mM) and
on rate ( - 1 × 108 M 1 see-l) of calcium buffers in bovine chromaffin cells. 23
calculate the peak free calcium concentration in both the high NA and the
low NA case, with results shown in Fig. 6. The presence of calcium buffers
obviously greatly accelerates the rate of loss of calcium ions from the
two-photon volume, leading to lower free calcium concentrations. This
demonstrates the familiar tradeoff between fluorescent signal and biological
effect', in order to quantify a dynamical calcium process (such as calcium
release by an uncaging pulse), a large amount of fluorescent indicator is
desirable to maximize signal. However, the presence of fluorescent indicator
dye reduces the amount of free calcium available to induce biological re-
sponses.
[20] T W O - P H O T O N A C T I V A T I O N OF C A G E D C A L C I U M 379
Experimental measurements have been made of the calcium buffering

dynamics in bovine chromaffin cells, and these determined that the cells
in question had 4 mM of a calcium buffer with an on rate of 1 x 10s M -I
sec-123 for a characteristic intrinsic buffering rate of ki ~ 4 X 105 s e c -1,
provided that the total calcium release is 400/xM or less. In a cell containing
100/~M of Fluo-3 and this rapid rate of intrinsic calcium buffering, we can
estimate that the total effective calcium removal rate constant ko = ki +
k~ + kx = 5.6 X 105 sec -I for calcium liberated with a 1.33 NA lens. The
peak free calcium concentration generated in such an experiment is also
plotted in Fig. 6. With such a rapid biochemical mechanism of calcium
removal, the rate of loss of free calcium concentration is dominated by
calcium buffering over free calcium diffusion.
In conclusion, we find that the longest elevated calcium concentrations,
those due to two-photon release of calcium in a nonbuffered environment,
are on the order of 32/zsec FWHM for high spatial resolution photolysis
using 1.3 NA optics. This duration can be considerably lengthened by
reducing the NA of the excitation optics, with the unavoidable consequence
of reducing the initial spatial resolution of release. Addition of calcium
buffering to the system, whether through intrinsic or extrinsic calcium
buffers, can significantly shorten this duration.
Conclusion
This work described equipment for the two-photon excitation of caged
calcium and delineated the spatial resolution of calcium release possible
with this apparatus. The total calcium release can be calculated easily for
any illumination parameters using the equations described here, as can the
temporal behavior of the peak calcium concentration generated using rapid
uncaging pulses of high release rate.
In practical terms, complete photolysis of caged calcium with high spatial
resolution is possible with TPE. Using - 7 mW of average power over a
pulse train of 10 /zsec, 700-nm light overfilling a 1.3 NA objective can
completely photolyze the calcium cage Azid-1 contained within the TPE
focal volume. A second calcium cage, DM-nitrophen, requires considerably
more excitation power ( - 8 5 mW of 720-nm light). The resultant nonequilib-
rium highly localized distribution of calcium ions will have a radial and
axial length scale of 0.25 and 0.86/zm, respectively. The duration of the
elevated calcium concentrations produced with TPE will vary depending
on the diffusion times and buffering conditions of the released ions. In an
unbuffered environment, the aforementioned distribution generated with
a high (1.3) NA lens will last for ~32/~sec (FWHM), which is limited by
the diffusional escape time of the liberated ions out of the TPE focal
volume. This duration can be significantly extended by lowering the spatial
resolution of release and can be significantly reduced by the addition of
calcium-buffering agents, whether intrinsic cellular calcium buffers or ex-
trinsic calcium indicator dyes.
There are many possible uses for TPE of calcium cages. The competition
of buffers for locally released calcium could provide a window into the
spatial heterogeneities of the intrinsic calcium-buffering dynamics in living
cells. This could be accomplished through utilization of the relative uptake
of liberated calcium ions by a fluorescent calcium indicator and intrinsic
calcium buffers as a measure of the intrinsic buffering rate and capacity of
cells. As has been shown in Fig. 6, for at least one cell type these kinetics
should happen significantly faster than the diffusional escape of calcium
ions from the TPE focal volume, allowing these buffering properties to
be measured with submicron, three-dimensional resolution and thereby
extending previous whole cell studies. 23
These measurements and calculations demonstrate that complete two-
photon release of calcium with high three-dimensional spatial resolution
is possible and can be an extremely useful tool in the study of a variety of
cellular processes.
Acknowledgments
This work was carried out in the Developmental Resource for Biophysical Imaging and
Optoelectronics with funding provided by the NSF (Grant DIR 88002787) and NIH (Grant
RR07719 and RR04224). EB was supported as a predoctoral trainee under NIH Grant
T32GM08267. Laser excitation at 622 nm was supported by Dr. Frank Wise, Xiang Liu, and
Lie-Jia Qian of Cornell University Applied and Engineering Physics.
[2 i] C a g e d I n o s i t o l 1 , 4 , 5 - T r i s p h o s p h a t e for Studying
R e l e a s e o f C a 2÷ f r o m I n t r a c e l l u l a r Stores
By NICK CALLAMARAS a n d IAN PARKER
Introduction
Inositol 1,4,5-trisphosphate (InsP3) is a second messenger that mediates
Ca z÷ release during many physiological processes, including development,
gene regulation, secretion, contraction, synaptic transmission, and apoptotic

volume. This duration can be significantly extended by lowering the spatial
resolution of release and can be significantly reduced by the addition of
calcium-buffering agents, whether intrinsic cellular calcium buffers or ex-
trinsic calcium indicator dyes.
There are many possible uses for TPE of calcium cages. The competition
of buffers for locally released calcium could provide a window into the
spatial heterogeneities of the intrinsic calcium-buffering dynamics in living
cells. This could be accomplished through utilization of the relative uptake
of liberated calcium ions by a fluorescent calcium indicator and intrinsic
calcium buffers as a measure of the intrinsic buffering rate and capacity of
cells. As has been shown in Fig. 6, for at least one cell type these kinetics
should happen significantly faster than the diffusional escape of calcium
ions from the TPE focal volume, allowing these buffering properties to
be measured with submicron, three-dimensional resolution and thereby
extending previous whole cell studies. 23
These measurements and calculations demonstrate that complete two-
photon release of calcium with high three-dimensional spatial resolution
is possible and can be an extremely useful tool in the study of a variety of
cellular processes.
Acknowledgments
This work was carried out in the Developmental Resource for Biophysical Imaging and
Optoelectronics with funding provided by the NSF (Grant DIR 88002787) and NIH (Grant
RR07719 and RR04224). EB was supported as a predoctoral trainee under NIH Grant
T32GM08267. Laser excitation at 622 nm was supported by Dr. Frank Wise, Xiang Liu, and
Lie-Jia Qian of Cornell University Applied and Engineering Physics.
[2 i] C a g e d I n o s i t o l 1 , 4 , 5 - T r i s p h o s p h a t e for Studying
R e l e a s e o f C a 2÷ f r o m I n t r a c e l l u l a r Stores
By NICK CALLAMARAS a n d IAN PARKER
Introduction
Inositol 1,4,5-trisphosphate (InsP3) is a second messenger that mediates
Ca z÷ release during many physiological processes, including development,
gene regulation, secretion, contraction, synaptic transmission, and apoptotic

[2 1 ] CAGED InsP3 381
cell death. 1,2 The development of caged-InsP3 [c-InsP3: myo-inositol 1,4,5-

trisphosphate (1,2-nitrophenyl ester)] has provided an elegant and efficient
means by which to control levels of this active messenger in the cytoplasm
of intact cells, giving an unparalleled degree of temporal and spatial resolu-
tion for detailed studies of this ubiquitous intracelhlar pathway. 3-5 In com-
parison to other techniques, such as fast-flow perfusion of permeabilized
c e l l s 6 o r direct introduction from a point source via a pressure or iono-
phoresis microelectrode,7 it allows rapid (millisecond), spatially defined
elevations in InsP3 concentration to be achieved without compromising the
organization and biochemical composition of the cytoplasm or disrupting
the electrical properties of the cell.
The essence of the technique is first to introduce the physiologically
inert c-InsP3 into a cell (e.g., by microinjection or diffusion from a whole
cell patch pipette) and allow it to diffuse and equilibrate throughout the
cytosol. Subsequent illumination with near-UV light then provides a means
to photorelease active InsP3. Major advantages of this technique are: (1)
rapid and reproducible millisecond steps in intracelhlar concentration of
InsP3 are achieved; (2) the relative concentration of InsP3 can be precisely
regulated by the intensity and duration of illumination; and (3) manipula-
tion of the illumination field allows for homogeneous release of InsP3
throughout spatially defined regions of the cell.
Several practical points further simplify the use of c-InsP3 and extend
its utility. The caged precursor is extremely stable in the cytosol, and most
cells are exquisitely sensitive to InsP3 (nanomolar range), so that a large
"reservoir" of c-InsP3 may be loaded, allowing numerous photolysis flashes
to be delivered, each of which consume only a tiny fraction of the total
c-InsP3. Also, c-InsP3 is efficiently photolyzed by light with wavelengths
<400 nm, thus freeing up the visible spectrum for simultaneous use of
long-wavelength indicator dyes to monitor Ca 2+ liberation induced by
InsP3.
This article describes techniques developed in the authors' laboratory
for use of c-InsP3 in Xenopus laevis oocytes and illustrates some of their
applications in studying the InsP3/Ca 2÷ second messenger pathway. The
1 M. J. Berridge, Nature 361, 315 (1993).

2 T. Michikawa, A. Miyawaki, T. Furuichi, and K. Mikoshiba, Crit. Rev. Neurobiol. 10,
39 (1996).
3 D. C. Ogden, K. Khodakhah, T. D. Carter, P. T. Gray, and T. Capiod, J. Exp. Biol. 184,
105 (1993).
4 S. S. Wang and G. J. Augustine, Neuron 15, 755 (1995).
s I. Parker, Neuromethods 20, 369 (1992).
6 T. Meyer, T. Wensel, and L. Stryer, Biochemistry 29, 32 (1990).
7 M. J. Berridge, Proc. R. Soc. Lond. B 238, 235 (1989).
382 CAOED COMPOUNDS [211
Xenopus oocyte has long been a common cell type for studies of the InsP3-
mediated Ca 2÷ release system due to its enormous size (>1 mm diameter)
and simple geometry, which facilitate many experimental procedures. Addi-
tional advantages are the ability of the oocyte to express foreign proteins
encoded by microinjected mRNA s and a lack of other confounding intracel-
lular Ca 2+release channels (i.e., ryanodine receptors). The combined advan-
tages of flash-photolysis of c-InsP3 and the oocyte system have, for example,
permitted high-resolution studies of the kinetics of InsP3 receptor (InsP3R)
gating in an intact cell system, 9 allowed elucidation of the relationship
between InsP3-evoked Ca 2÷ signals and Ca2÷-dependent C1- membrane
current, 1° and facilitated the resolution of "elementary" Ca 2÷ release events
underlying global Ca 2÷ signals. 11-13Moreover, the ability to evoke reproduc-
ible InsP3 signals has aided pharmacological studies of agents affecting this
intracellular messenger pathway. 14"15
Methodology
Photochemistry of c-InsP3
In the native oocyte, InsP3 is generated from receptor-mediated break-
down of phosphatidylinositol bisphosphate. InsP3 then binds to the intracel-
lular InsP3R to release Ca 2÷ from a subset of the intracellular Ca 2+ stores
and is subsequently metabolized into a bewildering array of other inositol
phosphates, some of which retain physiological activity. The flash photolysis
of c-InsP3 avoids the time delays, nonlinearity, and modulation by other
messengers associated with ligand-activated InsP3 generation. This is
achieved due to the efficiency of the photochemistry involved in the cleav-
age of InsP3 from its protecting nitrophenyl ester group. 16 In brief, the
reaction proceeds in two steps: the rapid (nanosecond) absorption of a
high energy photon generating active intermediates followed by a slower
(millisecond) dark reaction whereby intermediates decay to release InsP3
8 R. Miledi, I. Parker, and K. Sumikawa, in "Fidia Neuroscience Award Lectures" (J. Smith,
ed.), Vol. 3, p. 57. Raven Press, New York, 1989.
9 I. Parker, Y. Yao, and V. Ilyin, Biophys. J. 70, 222 (1996).
101. Parker and Y. Yao, Cell Calcium 15, 276 (1994).
u I. Parker and Y. Yao, Proc. R. Soc. Lond. B 246, 269 (1991).
a2 y. Yao, J. Choi, and I. Parker, J. Physiol. (Lond.) 482, 533 (1995).
13 I. Parker, J. Choi, and Y. Yao, Cell Calcium 20, 105 (1996).
14 I. Parker and I. Ivorra, J. Physiol. (Lond.) 433, 207 (199l).
15 V. Ilyin and I. Parker, J. Physiol. (Lond.) 448, 339 (1992).
16 j. W. Walker, J. Feeney, and D. R. Trentham, Biochemistry 28, 3272 (1989).
[211 CAGED InsP3 383
,.
O
,
.cH3 / O--P.. !
oH
,vl
O_Ipl ..-0 -" O--P--OH
I
I OH O-
O-
Caged InsP3
O
II OH C Ha
HO--P.. I
/ O~,,..~OH "~0 H+
°
,.~ I1..O" y "O--P--OH -~J "NO2
~u--r =" I
I OH O-
O-
InsP 3 2-nitrosoacetophenone
FIG. 1. Schematic photolysis reaction for caged InsP3. In addition to free ]nsP3, photolysis
releases a proton and a by-product, 2-nitrosoacetophenone.
(Fig. 1). The by-products of this reaction, a proton and nitrosoacetophe-

none, can be toxic at high levels (1 mmol-1), altering the local pH or
reacting with -SH groups of proteins, respectively,t7 Fortunately, the con-
centrations of InsP3 necessary to elicit even maximal responses in oocytes
and many other cells are sufficiently small that by-product toxicity appears
not to be an issue.
Caged InsP3 can presently be obtained from either Calbiochem (La
Jolla, CA) or Molecular Probes (Eugene, OR). In our experience the
c-InsP3 available from Molecular Probes is preferable because of lower
contamination by active InsP3.
Introduction of c-InsP3 into Cells

Caged InsP3 is a highly charged molecule and thus does not permeate
through the cell membrane. For work with single cells, c-InsP3 may be
introduced into the cytosol by injection through a micropipette (pneumatic
presure injection or ionophoresis) or by diffusion from a whole cell patch
pipette. In the latter case, the resulting intracellular concentration should
approximate that in the pipette. Pressure injection is convenient for use
with large cells, and the final intracellular concentration of c-InsP3 can be
17M. G. Hibberd, Y. E. Goldman, and D. R. Trentham, Curr. Topics CelL ReguL 24, 357 (1984).
384 CAGZOCOMVOUNDS [211
estimated from the amount of fluid injected (measured as the diameter of

a fluid droplet expelled with the pipette tip in air), the concentration in
the injection solution, and the cytosolic volume.
Injection into defolliculated Xenopus oocytes TM is made through broken
(tip diameter ca. 5 /zm) glass micropipettes. We typically inject about
20 nl of an aqueous solution of 1-5 mM c-InsP3 (to which Ca 2+ indicator
dyes may also be added), resulting in a final intracellular concentration of
about 20-100/zM, assuming a cytosolic volume of 1 tzl. It is then necessary
to wait for 30 min or longer for diffusional equilibration of c-InsP3 through-
out the cell; this time also allows the cell to metabolize any active free
InsP3 that may be present as a contaminant. 5 Standard room lights, and
even halogen fiber-optic illuminators, cause surprisingly little photolysis of
c-InsP3, and we have not found it necessary to take any special precautions
with lighting either in handling injection solutions or after loading cells
with c-InsP3.
Clearly, mechanical introduction of c-InsP3 through pipettes would be
impractical for studies on populations of cells. An elegant approach that
has recently been developed involves loading c-InsP3 by extracellular appli-
cation at a membrane-permeable ester, 19analogous to the well-known tech-
nique for loading indicators such as Fura-2. Alternatively, the plasma mem-
brane may be permeabilized by a variety of techniques 2° to allow access
of charged c-InsP3 to the cytosol.
Linearity and Calibration of Photorelease

The amount of free InsP3 resulting from a given photolysis flash is
expected to be linearly proportional to the starting concentration of c-InsP3
and the flash intensity and duration. In practice, this linear dependence on
flash parameters is followed almost exactly; as illustrated in Fig. 2A, where
caged ATP was used to monitor photolysis in place of c-InsP3 because of
the ease of detecting release of ATP by luminescence of a luciferin/lucifer-
ase assay system. Thus, an extremely precise control of relative amounts
of InsP3 formation within a given cell can be obtained by appropriate
adjustment of flash parameters. Our usual method is to adjust the flash
intensity using neutral density filters to set an initial working range and
then vary the flash duration to control the extent of photorelease. This
has advantages in that the duration can be readily altered in very small
increments, as opposed to the relatively coarse steps available with filter
18 K. Sumikawa, I. Parker, and R. Miledi, Methods NeuroscL 1, 30 (1989).

19W. H. Li, C. Schultz, J. Lopi, and R. Y. Tsien, Tetrahedron 53, 12017 (1997).
z0 K. A. Oldershaw, D. L. Nunn, and C. W. Taylor, Biochem. J. 278, 705 (1991).
[2 ll CAGEDInsP3 385
A :
I00
./:
80
./
e °/°/'~
B
1000
60 AM
I
/" ./,/ ,°° |.
s/ /
40 / /
,~= /: / I0 ~ ~
E il | ~"
I
20 ./o t
0.01
1
0.1 11 110
I ° E Flash Duration ( s )
I"
,|
, I I I I I I 1 I I I
0 20 40 60 80 I O0
Flash Duration ( ms )
FIo. 2. Linearity of photorelease with flash strength and calibration of extent of photolysis
using a luminescence system to assay release of ATP from a caged precursor. (A) Formation
of ATP varies linearly with flash duration. Recordings were made from a 20-/~1 droplet
containing 100/~M caged ATP together with undiluted ATP assay reagent (Sigma FL-AAM).
The photolysis light was focused by the microscope objective to a small spot within the droplet,
and the graph plots the peak amplitude of the ATP-dependent luminescence (monitored by
a photomultiplier) evoked following photolysis flashes of identical intensity and varying dura-
tion. (B) Estimation of the fractional photolysis of caged ATP resulting from a given photolysis
flash. The experiment was performed as in (A), except for use of a smaller droplet that was
entirely irradiated by the photolysis spot. A series of repeated flashes of increasing duration
were applied, and the peak luminescence recorded following each flash. Double logarithmic
plot shows cumulative luminescence (i.e., sum of peak signals evoked by successive flashes)
as a function of cumulative duration of exposure to photolysis light. [Reproduced from I.
Parker and I. Ivorra, Am. Z Physiol. 263, C154 (1992), with permission.
wheels, and individual flashes within a paired-pulse protocol can be inde-

pendently controlled. Care should be taken, however, that the longest
flashes employed are brief compared to the kinetics of the response un-
der investigation.
For many purposes, control of the relative amounts of InsP3 photore-
leased in a given experiment is sufficient, but it may also be necessary to
estimate the absolute concentrations of free InsP3. This is a more difficult
and less precise task, for which we have taken two approaches. The first
386 CAGZDCOMPOUYDS I211
involves an in vitro calibration employing caged ATP. 21 The protocol is
illustrated in Fig. 2B. A small droplet containing a known amount of ATP
together with assay mixture is placed on the microscope stage and is com-
pletely irradiated by photolysis flashes of the same intensities as used for
physiological experiments. Repeated flashes are applied, and the lumines-
cence signals (linearly proportional to the amount of ATP) are recorded
following each flash. After many exposures, the luminescence response
declines to zero, indicating complete consumption of the c-ATP (luciferin
is present in excess). Thus, the total cumulative luminescence over all
responses corresponds to the photorelease of all the ATP, and the fractional
amount released by a given light flash can be calculated from the ratio of
the luminescence resulting from that flash compared with the total lumines-
cence. This fractional release should then apply equally to photolysis of
intraceUular c-InsP3 under the same conditions, with the major caveat that
no account is taken of light absorption within the specimen. A second
approach is to employ the cell as a bioassay system. In the oocyte, Ca 2+
waves are initiated at a sharply defined concentration of InsP3 so that
photolysis flashes below threshold evoke little or no Ca 2+ release or Ca 2+-
dependent C1- current response, whereas just-suprathreshold stimuli give
clear responses (cf. Fig. 4). Estimates of the absolute concentration of InsP3
corresponding to this threshold can then be obtained by injecting increasing
amounts of a poorly metabolized but equipotent InsP3 analog (3-F-InsP3)
to determine the resulting intracellular concentration at which waves are
triggered. 12
Measurements of the InsP3 concentrations by both of these methods
indicate that Ca 2+ waves are evoked in the oocyte by roughly 50 nM InsP3,
and maximal responses are evoked by concentrations about 20-fold greater. 9
Because the cells can readily be loaded to concentrations of c-InsP3 as
high as 100 ixM, flashes giving physiological responses will photolyze only
negligible proportions (0.05-1%) of the c-InsP3. Further, the turbidity of
the oocyte cytoplasm limits light penetration to a few tens of/xm, so that
the greater volume of the cell is effectively shielded from the photolysis
light and acts as a virtually inexhaustible reservoir of fresh c-InsP3. The
net result is that it is possible to evoke numerous (hundreds) reproducible
responses with repeated flashes, without significant depletion of available
c-InsP3 in the oocyte.
Photolysis Light Sources

The photolysis of caged InsP3 requires a sufficiently intense source of
illumination at wavelengths around 350 nm. Possible light sources capable
2aI. Parker and I. Ivorra,Am. J. Physiol. 263, C154 (1992).

[211 CAGEDInsP3 387
of efficient photolysis in most biological preparations include (1) continuous

arc lamps with electronically controlled shutters, (2) xenon flash lamps,
and (3) UV lasers. The preceding systems are ordered in terms of increasing
energy, complexity, and cost. Each system has relative merits and disadvan-
tages but, fortunately, bigger and costlier systems are not necessarily better.
A continuous UV arc lamp is a stable source of illumination that, when
combined with commercially available high-speed electric shutters, provides
an inexpensive and highly efficient photolysis system. The stability of the
lamp output is very good if the arc is mounted vertically and powered
from a high-quality constant current power supply, with a between-flash
reproducibility of better than 1%. For use with caged-InsP3, the limited
power of the lamp is not a major consideration because photolysis of only
a small percentage of the c-InsP3 loaded into a cell is sufficient to evoke
responses. Either mercury or xenon arc lamps (50-100 W) work well and,
indeed, it is usually necessary to attenute their output with neutral density
filters. High-speed electronically controlled shutters provide an affordable
and efficient means to produce flashes as brief as 1-2 msec. Durations
shorter than this would provide little advantage because the photogenera-
tion of InsP3 from its caged precursor takes 5-10 msec. Finally, unlike flash
lamps or pulsed lasers, it is possible to continuously irradiate the specimen,
mimicking physiological InsP3 generation over several seconds during ago-
nist stimulation. Therefore, a continuous arc lamp is the simplest and most
cost-effective choice for most experimental situations. Such systems can be
readily constructed from an existing epifluorescence microscope, merely
by substituting a UV filter cube to restrict irradiation to wavelengths <400
nm and inserting an electronic shutter into the light path.
Xenon flash lamps 22 provide a much higher (1000 times or greater)
energy output in a brief flash than a shuttered arc lamp system but, as
noted earlier, this higher energy is not likely to be required when working
with c-InsP3 provided that an efficient optical system is used to channel
light to the specimen. Further, flash lamp systems have disadvantages in
regard to their greater cost, generation of electrical artifacts, lack of inde-
pendent control of the duration of each flash in paired-pulse experiments,
and an inability to provide continuous illumination of sustained photore-
lease.
Several types of laser are available that produce high-energy pulses at
near-UV wavelengths, including the frequency-doubled ruby laser and the
frequency-tripled N d - Y A G (neodymium-ytrrium-aluminum-garnet) la-
ser. In our experience, the only application for which the high cost of these
lasers is justified involves focusing the photolysis light to a near-diffraction
limited spot in the specimen so as to achieve a virtual point-photorelease
22 G. Rapp and K. Guth, Pflugers Arch. 411, 200 (1988).

388 CAGEDCOMPOUNDS [2 1]
of InsP3. Because of the perfectly parallel beam from a laser this can be
achieved with little loss of light, whereas to achieve the same result with
a conventional (arc lamp) source involves insertion of a pinhole aperture
with an accompanying drastic loss in light throughput.
Two-photon excitation of caged compounds by femtosecond pulses from
mode-locked lasers is a recent development 23 that offers the possibility of
even more spatially restricted photorelease because the quadratic depen-
dence of two-photon photolysis on light intensity ensures that this will be
restricted only to the plane of focus of the light spot formed by an objective
lens. Although this approach has been elegantly employed to map out the
sensitivity of neurons to extracellular photorelease of caged neurotransmit-
ter, 24 the possible advantages of this expensive technology for use with c-
InsP3 are unclear and are likely to be limited by the physiology of the InsP3
pathway. In particular, the rapid intracellular diffusion of InsP3, together
with the long latencies (many milliseconds) before Ca 2+ release begins,
suggests that the spatial localization of InsP3 will be limited by those factors
and not by the minimum volume throughout which photorelease occurs.
Simultaneous Use o f c-InsP3 and Fluorescent Calcium Indicators

The primary action of InsP3 in the cell is to liberate calcium from
intracellular stores. One approach to monitor the resulting rise in cytosolic-
free calcium concentration is to measure currents through endogenous
calcium-activated membrane channels (e.g., the CaZ+-dependent C1- chan-
nels in the oocyte membraneY This is advantageous in terms of simplicity
and ensures that the measurement technique does not perturb the signals
of interest. Interpretation may, however, be complicated by uncertainties
regarding the steady-state and kinetic relationships between free [Ca 2÷]
and current amplitude, 1° and voltage-clamp recordings of current provide
no information regarding the spatial distribution of calcium within the cell.
Thus, for many purposes it is desirable to have a more direct measure of
cytosolic calcium. The availability of a wide range of fluorescent calcium
indicator dyes operating in the visible spectrum makes it possible to combine
fluorescence calcium monitoring or imaging together with simultaneous
photolysis of c-InsP3.
Currently, our favorite dyes are the Oregon green family (Molecular
Probes), which are optimized for excitation by the 488-nm line of the argon
ion laser and fluoresce in the green. The excitation spectra of these dyes
are sufficiently separated from that of c-InsP3 that we find little practical
23W. Denk and K. Svoboda,Neuron 18, 351 (1997).

24W. Denk, Proc. Natl. Acad. Sci. U.S.A. 91, 6629 (1994).
25R. Miledi and I. Parker, Z Physiol. (Lond.) 357, 173 (1984).
[211 CAGEDInsP3 389
difficulty with photolysis of c-InsP3 resulting from the blue excitation light
and, conversely, photolysis flashes at wavelengths <400 nm cause little or
no artifacts in fluorescence recordings. Furthermore, the dyes are available
as variants with different calcium affinities (e.g., Oregon green 488
BAPTA-1 has an affinity of a few hundred nanomolar, and Oregon green
BAPTA-5N has a low affinity of a few tens of micromolar), so that they
may be used selectively to examine small calcium signals close to the basal
calcium level or large calcium transients evoked by maximal InsP3 levels.
A disadvantage of all currently available long-wavelength indicators is that,
unlike the UV-excited dyes Fura-2 and Indo-1, none show shifts in either
excitation or emission spectra on binding calcium. Thus, using a single dye,
it is not possible to ratio signals at two wavelengths to obtain a calibration
of free calcium levels independent of variations in dye loading and path
length. A reasonable compromise, however, is to form a "pseudo ratio"
signal by expressing calcium-dependent fluorescence signals relative to the
resting fluorescence before stimulation. 2
The basic principle in designing an optical system for simultaneous
fluorescence monitoring and photolysis is to "stack" dichroic mirrors so as
to split off appropriate parts of the visible and UV spectra. Thus, a UV
dichroic mirror placed close to the microscope objective will reflect short
wavelengths (<400 nm) from a photolysis light source onto the specimen
but transmit longer wavelengths, which may then be further split by an
additional dichroic mirror to separate fluorescence excitation and emission
wavelengths appropriate for specific dyes. In our earlier optical system5
employing a Zeiss Universal microscope, this "stacking" was possible by
physically mounting one epifluorescence attachment on top of another.
More modern microscope designs fail to anticipate the need for such flexi-
bility, and in our present system (described in the following section) a UV
dichroic mirror is placed in the regular epifluorescence unit, whereas the
fluorescence excitation dichroic mirror is located externally to the micro-
scope with light directed through the video port. Excellent dichroic mirror
and filter sets tailored for use with particular dyes are available from Omega
Optical Inc. and Chroma Technology Corp. (both at Brattleboro, VT).
Fluorescence calcium monitoring systems may vary in their degree of
spatial and temporal resolution (with corresponding increases in cost and
complexity) from photomultiplier-based detectors to monitor calcium
throughout an entire cell or defined region of a cell,5 through CCD camera-
based wide-field imaging systems11 to laser-scan confocal microscopes pro-
viding millisecond and submicron resolution. 26
26 I. Parker, N. Callamaras, and W. Wier, Cell Calcium 21, 441 (1997).

Dump, ~2,,~ ~U lit L V-I.alserj
I
£'g'5
i ~ !
t U U J
FIG. 3. Schematic diagram of a versatile photolysis and imaging system constructed around
an Olympus IX70 inverted microscope. The photolysis system is shown at the top, and
interfaces through the epifluorescence port (EP) of the microscope. A confocal scanning
system on the left interfaces through the video side port (VP). L1, piano-concave lens; f, - 1
cm; L2, biconvex lens; f, 5 cm; $1, electronic shutter; M1, cover glass acting as beam splitter;
ND, neutral density filter wheels.
Versatile Photolysis System for Imaging and Electrophysiological Studies

This section describes details of our current photolysis system, which
operates through the epifluorescence port of an Olympus IX70 inverted
microscope (Lake Success, NY), and can be used either independently
or together with a confocal Ca 2÷ imaging system interfaced through the
microscope video port (Fig. 3). For further details of the confocal scanner,
see Parker et aL, 26 and for more general information pertaining to the
principles and applications of biological confocal microscopy, see Pawley.27
The photolysis system employs both a continuous arc lamp source (for
wide field photorelease) and a pulsed UV laser (for "point" photorelease),
with light from both systems being combined by a beam-splitting mirror
and directed into the epifluorescence port of the Olympus IX70, which
is equipped with a standard UV filter cube and fluor objective lenses.
Components are mounted on an optical breadboard, using standard post-
mounts (New Focus Inc., Santa Clara, CA). Note that with the exception
of the filter cube and holder, no Olympus epifluorescence components
are required.
The upper section of Fig. 3 shows the layout of the photolysis light
27j. B. Pawley, "Handbook of Biological and Confocal Microscopy." Plenum Press, New
York, 1990.
[21] CAGEDInsP3 391
paths. All lenses are fused silica for optimal UV transmission, and optical
components are postmounted at a height corresponding to the center line
of the Olympus epifluorescence port. For experiments where photorelease
of caged compound is required over a wide area (10- to 100-~m-diameter
spot using a 40x objective), UV light is derived from an arc lamp (75-W
xenon lamp, mounted in a Zeiss housing and operated from a stabilized
constant-current power supply). An electronic shutter (Uniblitz, Vincent
Associates, NY), triggered manually or via TTL input from a Digitimer or
computer, controls exposure duration while a set of neutral density wheels
(3.0 OD in steps of 0.1 OD, New Focus Inc.) allow control of light intensity.
Adjustment of a collector lens in the arc lamp housing provides uniform
(Koehler) illumination throughout the photolysis spot in the microscope
image plane.
The laser system employs a Mini-Lite frequency-tripled (355 nm)
Nd: YAG laser (Continuum, Santa Clara, CA), with the laser head mounted
on the optical table by nylon screws and insulating stand-offs to avoid hum
loops for electrophysiological recording. Lenses L1 and L2 form a beam
expander so that the laser beam fills the back aperture of the objective
lens, thus making use of its full numerical aperture. Because the laser beam
at full power may crack an objective lens with poor UV transmission, a
microscope cover glass is used as a mirror (M4) to reflect only a small
percentage of the laser beam into the microscope through the neutral
density filter wheels, which allow further attenuation. The invisible beam
from the Mini-Lite presents a considerable safety hazard, and appropriate
precautions, including use of laser safety goggles, should be taken when
the beam is exposed for alignment. A beam dump is placed after M4 to
avoid the possibility of the beam passing into the room, and the entire
beam path is covered while the laser is operating. The laser spot formed
by an objective lens can be viewed using a coverslip marked with a yellow
"highlighter" pen, and its position centered by small adjustments of laser
position and deflection of M1. Finally, the spot can be brought to a sharp
focus by axial adjustment of L2. A UV-blocking filter in the Olympus filter
cube, together with an additional long-pass filter (A >510 nm) inserted in
the microscope binocular head, permits safe viewing while the laser is in
use. The laser can be operated in single-shot mode (triggered by a push
switch or TTL input) or pulsed repeatedly at up to 10 Hz.
Applications of c-InsPa for Study of Ca 2÷ Signaling in Oocytes
Dose-Response Relation of Ca2+ Release

Bath application of calcium-mobilizing agonists to oocytes results in a
complex oscillatory C1- current, which varies in a highly nonlinear manner
392 CA6ZD COMPOUNDS [21]
2
o*
i"
i I
N I *
"-" I
I I
1 2
Normalized Flash Duration
FlG. 4. Dose-response relationship of membrane currents evoked by photorelease of InsP3

by different durations of irradiation. Pooled data from five oocytes (indicated by different
symbols) showing the relationship between flash duration and peak size of evoked currents.
Flash durations are normalized with respect to that which evoked just-suprathreshold responses
in each oocyte, and currents are normalized with respect to currents evoked in each oocyte
evoked by a flash of twice threshold duration. In this experiment the entire vegetal hemispheres
were exposed to UV light. ]Reproduced from I. Parker and I. Ivorra, Am. J. Physiol. 263,
C154 (1992), with permission.]
with agonist concentration.28 The intracellular location of the numerous

stages in the signaling pathway between cell surface receptors and activation
of the current, however, makes it difficult to determine at which stage the
nonlinearity arises. Photolysis of caged InsP3 offers an elegant means to
circumvent this problem bypassing earlier stages in the pathway. Indeed, the
reproducibility and linearity of photolysis allow intracellular dose-response
relationships to be determined with the same ease as bath application of
compounds to an extracellular receptor. For example, Fig. 4 illustrates
the dependence of Ca2+-activated membrane currents on the strength of
photolysis flashes and demonstrates a nonlinear relationship in that a thresh-
old amount of InsP3 is needed before any current is generated. These data
also illustrate the reproducibility of the photolysis technique, as evident in
the close overlap of measurements from five oocytes when flash strengths
are normalized relative to the threshold required to evoke a detectable
current in each cell so as to compensate for differing amounts of microin-
jected c-InsP3.
28 I. Parker, K. Sumikawa, and R. Miledi, Proc. R. Soc. Lond. B. 231, 37 (1987).

[211 CAGEDInsPa 393
Pharmacological Studies of InsP3R in Intact Oocytes
The InsP3R through which calcium liberation occurs is a potential site
for modulation by many endogenous messenger compounds and exogenous
pharmacological agents. Studies of such effects on calcium mobilization are
complicated, however, if extracellular agonists are used to activate the InsP3
pathway because different agents may act on stages between the cell surface
receptor and InsP3 formation, as well as on the InsP3R itself. Again, caged
InsP3 provides a means to circumvent these difficulties by allowing the
effects of microinjected or (for membrane-permeant substances) bath-ap-
plied agents to be monitored on signals evoked by repeated and identical
pulses of intracellular InsP3. Two examples are shown in Fig. 5.
The first concerns the possible physiological roles of Ins(1,3,4,5)P4
(InsP4), a higher-order inositol polyphosphate that is formed transiently
by phosphorylation of InsP3 during activation of the signaling pathway. As
shown in Fig. 5A, evidence that InsP4 acts as a weak agonist at the InsP3R
to potentiate calcium mobilization is obtained by microinjecting InsP3 into
oocytes while monitoring calcium-dependent CI- currents evoked by InsP3
photoreleased by successive, just-suprathreshold flashes. TMA strong potenti-
ation of the currents is observed, even with small doses of InsP4 which,
themselves, evoke almost detectable responses. This experiment further
illustrates a technical point, in that it is necessary to microinject the highly
charged InsP4 into the oocyte. The spatial spread of InsP4 in the cell
is restricted to a region around the pipette tip because of diffusion and
metabolism, and the photolysis light is therefore focused as a small spot
centered around the injection pipette. Even though voltage-clamp re-
cordings reflect current from the whole membrane area of the oocyte,
the evoked C1- currents thereby arise only from the local region exposed
to InsP4.
A second example of the use of c-InsP3 for pharmacological studies
concerns the actions of caffeine, which is widely used as a tool to discrimi-
nate between calcium mobilization mediated through InsP3R and ryanodine
receptors, by virtue of its ability to potentiate ryanodine receptor-mediated
responses. Because caffeine readily permeates the cell membrane, we were
able to study its actions on InsP3-evoked C1- current responses by bath
application during trains of repetitive photolysis flashes.29 This causes a
dramatic and reversible reduction in responses (Fig. 5B) and a rightward
shift in the dose-response curve for InsP3 (Fig. 5C), suggesting that caffeine
may act as a reversible antagonist at the InsP3 receptor. In this and other
experiments where calcium liberation is monitored by endogenous calcium-
29 I. Parker and I. Ivorra, J. PhysioL (Lond.) 433, 229 (1991).

394 CAGED COMeOUNDS I211
A
Ins(1,3,4,5)P~
Control I--~-
T- ~'f
1 "Y- l- l " '[
-! -
30 s
B
Caffeine
50 s
C
250
o//°
/
200 =/a /e
c 150 0o
o
¢,,,.
1170
o •
0 50 100 150 200 250 300 350

Flash durationIrns)
Fro. 5. Utility of c-InsP~ for pharmacological studies. (A) Responses evoked by photore-
lease of InsP3 are facilitated by injection of low doses of InsP4. Trace shows calcium-dependent
membrane current responses (inward currents at a clamp potential of - 6 0 mV) evoked by
repetitive, identical photolysis flashes, with a duration just above the threshold to evoke
[211 CAGEDInsP3 395
activated currents it is, however, important to countrol for possible effects

on the calcium-activated C1-1 channels themselves. Monitoring of intracellu-
lar calcium by means of indicator dyes is one obvious approach, but in the
experiment shown in Fig. 5B, intracellular injections of calcium through a
micropipette are interspersed with photolysis flashes to directly activate
CI- currents and serve as a control.
Temporal Control and Kinetics of the Release Process

The photorelease of InsP3 is rapid (a few milliseconds) following a
photolysis flash, so that the kinetics of calcium liberation in response to a
virtually instantaneous step of InsP3 concentration can be studied. Further-
more, photolysis by a broad light spot results in a homogeneous release of
InsP3 throughout a volume of the cell, thus obviating problems of diffusion
delays that arise in other techniques such as intracellular microinjection or
extracellular applications of InsP3 to permeabilized cells. An example is
shown in Fig. 6, where a laser confocal microscope is used to monitor
calcium fluorescence signals from a minute volume within the oocyte (ca.
1 r) in response to InsP3 photoreleased throughout a larger region of the
cell surrounding the laser spot. Calcium signals following photolysis flashes
of increasing strength begin with progressively shorter latencies and show
increasing rates of rise (Fig. 6A). Because the fluorescence signal reflects
accumulation of calcium in the cytosol, differentiation of the signal further
provides a measure of the rate of calcium efflux (Fig. 6B), allowing a more
detailed study of the dependence of the kinetics of opening of the release
channels on the concentration of InsP3.9
As is evident in Fig. 6B, calcium liberation following a photolysis flash
terminates within a few hundred milliseconds, even though the levels of
InsP3 remain elevated for several seconds. 21 Thus, the InsPaR enters a
refractory state from which its subsequent recovery can be investigated by
applying paired photolysis flashes at varying intervals, in a manner analo-
detectable signals. InsP4 (about 1 fmol) was injected into the oocyte when marked by the
arrow. [Reproduced from I. Parker and I. Ivorra, J. Physiol. (Lond.) 433, 207 (1991), with
permission of the Physiological Society.] (B) Bath-applied caffeine inhibits InsP3-evoked
membrane current responses, but not the activation of the Ca2+-dependent CI- current. Trace
shows currents evoked by alternate stimulation by photolysis flashes (filled arrowheads) and
intracellular injections of calcium (open arrowheads). Caffeine (5 raM) was bath-applied
when marked by the bar. (C) Caffeine increases the threshold amount of InsP3 required to
evoke membrane currents. Graph shows amplitudes of currents evoked by photolysis flashes
of varying durations in the absence (open symbols) and presence (filled symbols) of 2.5 mM
caffeine. [B, C reproduced from I. Parker and I. Ivorra, J. Physiol. (Lond.) 433, 229 (1991)
with permission of the Physiological Society.]
396 CAGED COMPOUNDS [21 ]
5O
A
k 8
B activatio._.ntime
/~ ~inactivation
.l~ .~/ timeconstant
latency I k.
FIG. 6. Calcium transients and kinetics of calcium liberation in response to photorelease

of varying amounts of InsP3. (A) Superimposed fluorescent signals evoked by light flashes
of varying durations (indicated in milliseconds at the right of each trace). The arrowhead
indicates the time of the flash and circles indicate the point of maximal rate of rise of the
signal. The oocyte was loaded with the low-affinity indicator calcium green-5N, together with
c-InsP3. (B) The time differential [d(hF/F)/dt] of the 50-msec trace from (A), indicating the
kinetics of Ca 2÷ flux. [Modified from I. Parker, Y. Yao, and V. Ilyin, Biophys. J. 70, 222
(1996), with permission of the Biophysical Society.]
gous to the paired-pulse experiments used to examine the refractory state

of voltage-gated channels. 3°'31 The onset and recovery from inactivation
are determined by delivering paired, identical photolysis flashes at varying
intervals and monitoring the calcium-activated currents evoked by the sec-
ond flash in each pair relative to that evoked by a single flash (Fig. 7). At
short intervals the calcium signal is potentiated by a preceding flash, but
becomes almost completely suppressed as the interval is lengthened to
about 2 sec, and subsequently recovers over several seconds.
Entry into and recovery from the refractory state undoubtedly have an
important role in generation of the repetitive calcium spikes observed
in various cells during sustained activation by extracellular agonists. 32 A
characteristic feature is that the frequency of spiking increases with agonist
30 I. Parker and I. Ivorra, Proc. Natl. Acad. Sci. U.S.A. 87, 260 (1990),
31 V. Ilyin and I. Parker, J. Physiol. (Lond.) 477, 503 (1994).
32 M. J. Berridge and A. Galione, FASEB J. 2, 3074 (1988).
[21] CAGEDInsP3 397
L25nA
5s
B 250
A
2
"E
200
O
O
N
150
100
"I: ele .............. t__
'/ -
....
%
- 7w - - . . . . : ................
c-
" /
o
50
e)
eN~• • I-*" •ef
I ~'|-t--P'" 5" I I I I I I I
0
0 2 4 6 8 10
Interval ( s )
FIo. 7. Paired-flash experiment to investigate the onset and recovery of InsPa-mediated
calcium release. (A) Examples of membrane currents evoked by paired, identical flashes
delivered at different intervals. Arrowheads mark the times of each flash. (B) Size of the
response to the second flash plotted against interval between flashes. Data are shown from
two oocytes (different symbols) and are scaled as a percentage of the response evoked by a
single flash. [Reproduced from I. Parker and I. Ivorra, Proc. Natl. Acad. Sci. U.S.A. 87, 260
(1990), with permission.]
concentration but, as noted before, interpretation of these findings is com-

plicated by the complex stages in the messenger pathway between the cell
surface receptor and the generation of InsP3. An alternative approach is to
use sustained illumination with low-intensity UV light to cause a prolonged
elevation of intracellular [InsP3] to levels proportional to the light inten-
sity.33 This results in repetitive spikes in the fluorescence calcium signal
during the period of photolysis, which increase in frequency with increasing
photorelease but become smaller and superimposed upon a more sustained
calcium elevation (Fig. 8).
Spatial Control and Heterogeneity of Calcium Release Sites

The use of light as a stimulus to evoke photolysis of c-InsP3 permits
not only control of the magnitude and kinetics of InsP3 formation, but
also control of its spatial distribution. Thus, information can be obtained
33 I. Parker and I. Ivorra, J. Physiol. (Lond.) 461, 133 (1993).

398 CAGED COMPOUNDS [2 1]
3.7
1 ~ 1 0 s
FIG. 8. Repetitive Ca2÷spikes during prolonged photorelease of InsP3. Tracaes show point-
confocal Ca2+ signals (rhod-2 fluorescence) evoked at a single recording spot by various
intensities of photolysis light. Horizontal bar indicates the duration of exposure, and numbers
next to each trace indicate the intensity of illumination as a percentage of maximum. [Repro-
duced from I. Parker and I. Ivorra, J. Physiol. (Lond.) 461, 133 (1993), with permission of
the Physiological Society.]
regarding the spatial aspects of subcellular calcium liberation, either by

restricting photolysis to defined regions of the cell or by fluorescence im-
aging of the patterns of calcium liberation evoked by spatially homogeneous
photolysis. The following examples illustrate use of these approaches to
study calcium liberation at increasingly fine levels of resolution, from global
signals involving the whole cell to elementary calcium events located to
within a few microns.
Figure 9 illustrates an experiment in which the regional sensitivity to
InsP3 is mapped across the two hemispheres (animal and vegetal) of the
polarized oocyte cell. The oocyte was loaded with caged InsP3 and stimu-
lated by identical light flashes focused as a 50 × 50-/zm square. As the
photolysis square is moved from the vegetal to the animal pole, the whole-
cell C1- currents evoked by the flashes grow progressively larger (Fig. 9A),
indicating increased sensitivity to InsP3. Thus, even though the voltage-
clamp recording summates membrane currents from the whole cell, spatial
information is provided by the localized photorelease of InsP3. One compli-
cation in such experiments, however, is that the extent of photorelease may
[211 CAGEDInsP3 399
1100nA
ls
B o
o
300 --
A
<
c 200 im
~d
C
O
\o
\ (>
100 m
° = I~--8
0 -
I '~-- •- ** I I
Animal Equator Vegetal
FIG. 9. Spatial variation in currents evoked by InsP3 photoreleased at different locations

across the animal/vegetal axis of the oocyte. (A) Traces show currents evoked by identical
light flashes, with the photolysis light focused as a small square at different locations on the
albino oocyte as denoted in the diagram. The oocyte was oriented with the germinal vesicle
(animal pole) to the left. (B) Variation in the peak size of the current at different photolysis
positions. Filled symbols are measurements from a pigmented oocyte and open symbols from
an albino oocyte. Curves were drawn by eye.
be affected by optical factors in the cell. Wild-type oocytes are strongly

pigmented in the animal hemisphere. In these cells, sensitivity to photolysis
flashes is greatly reduced in the animal hemisphere (filled symbols, Fig.
9B), presumably because the superficial pigment blocks penetration of
photolysis light into the cell. In the experiment illustrated (Fig. 9A and
400 CAGEDCOraPOVNDS [21]
open symbols Fig. 9B), this problem was avoided by use of oocytes from
albino frogs, which lack pigmentation.
To examine heterogeneity in sensitivity between different functional
calcium release sites at a finer (micrometer) scale, ~3 the UV light was
focused to a spot of about 2/~m diameter by a pinhole aperture placed
in the arc lamp photolysis system, and fluorescence calcium signals were
monitored by a confocal laser spot concentric with the photolysis spot
(Fig. 10). Calcium liberation at this localized level shows an all-or-none
40 ms 50 m s 60 ms
B 100 ~m
80 • •
.E
~ 40
~ 2o
0 , •
0 2'0 ,0 6'0 ;0
flash duration (ms)
FIG. 10. Local pinhole photorelease of InsP3 reveals httle variation in sensitivity to InsP3
at different sites. (A) Traces show Ca 2÷ fluorescence monitored from a stationary confocal
spot in response to photorelease of InsP3 evoked by flashes of UV light focused to a 2-/zm
spot concentric with the confocal spot. Each trace was obtained with the oocyte moved to a
new random position, and the columns of traces show responses to three different flash
durations as indicated. Note the all-or-none characteristics of the responses. (B) Percentage
of sites responding to flashes of varying durations. Data were obtained from traces similar
to those in (A). Points indicate the percentage of trials (n > 20) with each flash duration in
which responses were observed. All responses were from a single albino oocyte, at randomly
chosen locations, constrained to fall within a 200 × 200-/zm area of the animal hemisphere.
[Reproduced by I. Parker, I. Choi, and Y. Yao, Cell Calcium 20, 105 (1996), with permission
of Cell Calcium.]
[2 II CAGED InsP3 401
FI6. 11. Uniform photorelease of InsP3 over a wide area evokes transient, localized,
"elementary" calcium release events. The image was obtained by confocal line-scan micros-
copy, in which the laser spot was repeatedly scanned along a fixed line 50/zm long. Traces
from successive lines are stacked from left to right so that the vertical dimension of the image
represents distance along the scan line and the horizontal dimension of the image represents
distance along the scan line and the horizontal dimension represents time. Increasing Ca2÷
levels are depicted on a gray scale and are shown as a ratio relative to the resting fluorescence
at each pixel before stimulation. A photolysis flash covering a spot approximately 100/zm in
diameter around the scan line was applied when indicated by dashed line.
characteristic, 34 so that identical flashes delivered at different, r a n d o m ,

locations on the o o c y t e e v o k e either no response or responses of similar
amplitudes (Fig. 10A). T h e p r o p o r t i o n of sites responding grows progres-
sively as the flash d u r a t i o n is increased. H o w e v e r , this gradation occurs
o v e r a very n a r r o w range of flash intensities, suggesting that the microscopic
sensitivity to InsP3 varies only slightly f r o m site to site.
A different a p p r o a c h to obtain spatial i n f o r m a t i o n is to image calcium
release e v o k e d by h o m o g e n e o u s photolysis of c-InsP3 over a relatively wide
( 5 0 - 1 0 0 / x m ) area, as illustrated in Fig. 11. In this case, fluorescence calcium
images are o b t a i n e d using a rapid (8.0 msec p e r line), high-resolution
(0.2/zm/pixel) confocal line-scan system. Such studies reveal that h o m o g e -
n e o u s stimulation gives rise to localized " e l e m e n t a r y " release events, origi-
nating in a stochastic m a n n e r at particular sites that p r o b a b l y represent
clusters of I n s P a R J 3
Finally, line-scan calcium imaging can be c o m b i n e d with localized p h o t o -
release i n d u c e d by a focused spot f r o m a U V laser to visualize calcium
34I. Parker and I. Ivorra, Science 250, 977 (1990).

402 CAGED COMPOUNDS [21 ]
FIG. 12. Examples of UV laser "point" photolysis of caged compounds monitored by

confocal line-scan imaging. (A) Control experiment monitoring fluorecsence of fluorecein
photoreleased in a droplet of caged precursor. The UV laser flash was delivered at the location
and time indicated by the tip of the arrowhead. (B) Ca2+release evoked followingphotolysis
of caged InsP3 by a focused spot from a UV laser. The oocyte was loaded with Oregon green-
1 (50/xM) and caged InsP3 (5 IxM), and recordings were made at a depth of 7/zm from the
surface. The laser spot was positioned in the center of the line scan, and the tip of the
arrowhead denotes the location and time of the flash. Line-scan imaging was performed as
in Fig. 11. [Reproduced from I. Parker, N. Callamaras, and W. Wier, Cell Calcium 21, 441
(1997), with permission of Cell Calcium.]
liberation as InsP3 diffuses from a virtual point source. Figure 12A shows
a control experiment in which fluorescein is photoreleased from a droplet
of caged precursor (Molecular Probes). The image shows that fluorescein
is formed almost immediately following the flash and subsequently spreads
rapidly, consistent with diffusion in free solution. However, calcium libera-
tion evoked by "point" photorelease of InsP3 in the oocyte shows more
complex characteristics (Fig. 12B). Calcium signals do not begin until after
a latency of several tons of milliseconds and are first seen at the site
of photorelease. Liberation then occurs at increasing distances from the
photolysis spot, giving an indication of the "range of action" of InsP3,
refecting both its diffusion and the requirement for [InsP3] to exceed a
specific threshold to evoke calcium release.
Final C o m m e n t s
In summary, c-InsP3 is an invaluable tool for the study of intracellular

Ca 2÷ signaling due to the degree of control that can be exercised over the
intensity, duration, and spatial extent of stimulation. This technique has
[22] CAGED cADP-RIBOSEAND CAGEDNAADP 403
been used to extend whole cell studies of the InsP3-mediated Ca 2÷ signaling

system in Xenopus oocytes. Combination of flash photolysis together with
fluorescence Ca 2÷imaging or membrane current recording provides a means
to study the transient kinetics and modulation of spatially complex, dynamic
Ca 2÷ signaling events in an intact cell. Moreover, given the flexibility, sim-
plicity, and relative affordability of the basic apparatus necessary to conduct
these experiments, the approach should be well within the reach of
many investigators.
Acknowledgments
We thank Dr. Jennifer Kahle for editorial help. Financial support was provided by NIH
Grant GM48071. Reprint requests and requests for further information should be addressed
to Dr. Ian Parker, Laboratory of Cellular and Molecular Neurobiology, Department of Psycho-
biology, University of California Irvine, CA 92697-4550; e-mail, [email protected].
[22] C h a r a c t e r i z a t i o n a n d A p p l i c a t i o n o f P h o t o g e n e r a t i o n
of Calcium Mobilizers cADP-Ribose and Nicotinic Acid
Adenine Dinucleotide Phosphate from Caged Analogs
B y KYLE R . G E E and HON CHEUNG LEE
Introduction
Two independent mechanisms for the mobilization of internal calcium
stores have been identified in sea urchin eggs. The calcium-mobilizing
metabolites in these mechanisms are the novel nucleotides cyclic ADP-
ribose (cADPR, 1) 1 and nicotinic acid adenine dinucleotide phosphate
(NAADP, 2), 2 derived from N A D and NADP, respectively. Cyclic ADP-
ribose can function as a modulator of the calcium-induced calcium response
mechanism and it also functions as a calcium messenger itself. 3 In addition
to the calcium-releasing activity of cADPR in sea urchin eggs, a variety of
mammalian, amphibian, and plant cells have been shown to be responsive
to the molecule. 4 The calcium release mechanism activated by N A A D P is
1 H. C. Lee, T. F. Walseth, G. T. Bratt, R. N. Hayes, and D. L. Clapper, J. Biol. Ch~m. 264,

1608 (1989).
2 H. C. Lee and R. Aarhus, J. Biol. Chem. 270, 2152 (1995).
3 H. C. Lee, Recent Prog. Horm. Res. 52, 357 (1996).
4 H. C. Lee, in "CRC Series on Pharmacology and Toxicology" (V. Sorrentino, ed.), p. 31,
CRC Press, Boca Raton, FL, 1995.

been used to extend whole cell studies of the InsP3-mediated Ca 2÷ signaling

system in Xenopus oocytes. Combination of flash photolysis together with
fluorescence Ca 2÷imaging or membrane current recording provides a means
to study the transient kinetics and modulation of spatially complex, dynamic
Ca 2÷ signaling events in an intact cell. Moreover, given the flexibility, sim-
plicity, and relative affordability of the basic apparatus necessary to conduct
these experiments, the approach should be well within the reach of
many investigators.
Acknowledgments
We thank Dr. Jennifer Kahle for editorial help. Financial support was provided by NIH
Grant GM48071. Reprint requests and requests for further information should be addressed
to Dr. Ian Parker, Laboratory of Cellular and Molecular Neurobiology, Department of Psycho-
biology, University of California Irvine, CA 92697-4550; e-mail, [email protected].
[22] C h a r a c t e r i z a t i o n a n d A p p l i c a t i o n o f P h o t o g e n e r a t i o n
of Calcium Mobilizers cADP-Ribose and Nicotinic Acid
Adenine Dinucleotide Phosphate from Caged Analogs
B y KYLE R . G E E and HON CHEUNG LEE
Introduction
Two independent mechanisms for the mobilization of internal calcium
stores have been identified in sea urchin eggs. The calcium-mobilizing
metabolites in these mechanisms are the novel nucleotides cyclic ADP-
ribose (cADPR, 1) 1 and nicotinic acid adenine dinucleotide phosphate
(NAADP, 2), 2 derived from N A D and NADP, respectively. Cyclic ADP-
ribose can function as a modulator of the calcium-induced calcium response
mechanism and it also functions as a calcium messenger itself. 3 In addition
to the calcium-releasing activity of cADPR in sea urchin eggs, a variety of
mammalian, amphibian, and plant cells have been shown to be responsive
to the molecule. 4 The calcium release mechanism activated by N A A D P is
1 H. C. Lee, T. F. Walseth, G. T. Bratt, R. N. Hayes, and D. L. Clapper, J. Biol. Ch~m. 264,

1608 (1989).
2 H. C. Lee and R. Aarhus, J. Biol. Chem. 270, 2152 (1995).
3 H. C. Lee, Recent Prog. Horm. Res. 52, 357 (1996).
4 H. C. Lee, in "CRC Series on Pharmacology and Toxicology" (V. Sorrentino, ed.), p. 31,
CRC Press, Boca Raton, FL, 1995.

very sensitive to low concentrations of the molecule, i.e., nanomolar levels
provide sufficient activation. The calcium release mechanisms activated by
these molecules are distinct from the one dependent on inositol 1,4,5-
trisphosphate (IP3), e.g., heparin, an antagonist of the IP3 receptor, has no
effect on cADPR or N A A D P mechanisms. Additionally, the mechanisms
activated by cADPR and N A A D P are distinct from each other. The
NAADP-dependent calcium release mechanism is not inhibited by high
concentrations of magnesium 5 or the specific cADPR antagonist 8-amino-
cADPR, 6 distinguishing it from the cADPR-dependent pathway. Also,
N A A D P and cADPR appear to be operating through distinct receptors;
specific binding of [32p]NAADP to sea urchin egg microsomes has been
observed, and cADPR has no effect on this binding. 7 A novel property of
the N A A D P mechanism is that at a subthreshold concentration, N A A D P
can completely inactivate the release system such that a subsequent chal-
lenge with maximal concentrations of N A A D P is ineffective. 7
The cyclic structure of cADPR is formed by linking the N-1 nitrogen
of the adenine ring in NAD + to the terminal ribose, displacing the nicotin-
amide group. This cyclization can be performed by treatment of NAD +
with the ubiquitous enzyme ADP-ribosyl cyclase 1 or by heating solutions
of NAD ÷ with sodium bromide in dimethyl sulfoxide (DMSO). 8 Nicotinic
acid adenine dinucleotide phosphate is formed by replacing the nicotin-
amide group of NADP with nicotinic acid. This can be accomplished by
treatment of NADP with an aqueous base 2 or by ADP-ribosyl cyclase-
catalyzed exchange of the nicotinamide in NADP with nicotinic acid. 9
The use of photolabile, or "caged," precursors of cADPR and N A A D P
makes it possible to generate these signaling molecules with precise spatial
and temporal resolution by UV photolysis when and where desired. This
article describes the preparation of cADPR and N A A D P caged as
1-(2-nitrophenyl)ethyl (NPE) phosphoesters, and characterization of their
photolyses (see Scheme 1).
5 R. Graeff, R. J. Podein, R. Aarhus, and H. C. Lee, Biochem. Biophys. Res. Commun. 206,
786 (1995).
6 T. F. Walseth and H. C. Lee, Biochim. Biophys. Acta 1178, 235 (1993).
7 R. Aarhus, D. M. Dicky, R. M. Graeff, K. R. Gee, T. F. Walseth, and H. C. Lee, J. Biol.
Chem. 271, 8513 (1996).
8 Q.-M. Gu and C. I. Sih, J. Am. Chem. Soc. 116, 7481 (1994).
9 R. Aarhus, R. M. Graeff, D. M. Dickey, T. F. Walseth, and H. C. Lee, J. Biol. Chem. 270,
30327 (1995).
122] CAGED cADP-RIBOSEAND CAGEDNAADP 405
OH OH ~ -C02H
o [ H
HO-:o
0 2
OH OH "~-NO2 OH 0
3 O=P-OR
6H
1 R,R'=H [--2 R=H
3 O2N 3 O2N
~- 4 R o r R ' = H , H 3 C Z ~ x ~ 5 R= H3C ~ ~-;x
SCHEME 1
Synthesis
Materials
Cyclic ADP-ribose and NAADP are available commercially from Mo-
lecular Probes (Eugene, OR). Other reagents are used as received from
various suppliers. Thin-layer chromatography (TLC) is performed on alumi-
num-backed silica gel plates impregnated with a fluorescent (254 nm) indica-
tor, using the indicated solvent systems. Reversed-phase liquid chromatog-
raphy is performed on Sephadex LH-20 using nanopure water as an eluant.
Caging reactions and manipulations are performed under subdued light.
1-(2-Nitrophenyl)diazoethane (3) is prepared according to the method of
Walker et alJ °
Preparation of Caged cADPR

The structure of cADPR presents two obvious sites for attachment of
a caging group, i.e., either phosphate of the bridging diphosphate moiety.
In theory, a caging group could be attached to the exocyclic amino group
of the adenine, but this would likely entail considerable synthetic difficulty.
10j. W. Walker, G. P. Reid, J. A. McCray, and D. R. Trentham, J. Am. Chem. Soc. 110,
7170 (1988).
406 CACEO COMPOUNDS [22]
Simple and efficient caging of nucleotide phosphates such as ATP has been
previously developed by Walker et al.,l° and this was the first method used
to cage cADPR. It was not known beforehand whether or how modification
of the bridging diphosphate moiety would affect the biological activity of
the resulting caged cADPR. Fortunately, the synthesis detailed here gives
caged cADPR that is biologically inert and, on photolysis, generates
cADPR that is active in mobilizing calcium in sea urchin egg microsomes
and in live eggs. The procedure for the preparation is similar to that de-
scribed previously. 11
Purified cADPR in free acid form (48 mg, 0.092 mmol) is dissolved in
3 ml of ice-cold water. The solution should be acidic at about pH 2.3. To
this stirring solution is added the caging reagent 1-(2-nitrophenyl)diazo-
ethane (3, 0.28 retool) in 3 ml of diethyl ether. It was found that low pH
is necessary for protonation of the phosphates of cADPR, as deprotonated
phosphates are not reactive with the caging reagent. It was also found that
a different caging reagent, 1-(4,5-dimethoxy-2-nitrophenyl)diazoethane, at
this low pH is unstable and gives a slower reaction with cADPR. The
resulting biphasic mixture is vigorously stirred at 0-5 ° in the dark for 3 hr,
during which time the diazoethane solution color changes from amber to
pale yellow. The ether layer is drawn off, and the diazoethane/ether treat-
ment is repeated three more times. The aqueous portion is applied to a
Sephadex LH-20 column (2 x 20 cm) and eluted with water; 2-ml fractions
are collected. The caged product 4 (TLC RI 0.55, methanol/CHC13/H20/
acetic acid, 12.5 : 10 : 3.5 : 0.2) is isolated as a fluffy white powder after lyophi-
lization of the combined product fractions (30 mg, 49%). Unreacted
cADPR, which elutes first, is also recovered (RI 0.13, 15 mg, 31%). The
caged product is efficiently photolyzed into free cADPR using a hand-held
UV lamp as analyzed by TLC. The caged product represents a mixture of
two monocaged isomers, which can be separated by AG MP-1 chromatogra-
phy. For the unseparated product the molar extinction coefficient at
259 nm is 17,200 using 0.8 mM of the product dissolved in water. The
two isomers have similar photolysis efficiency and both are biologically
inactive until photolysis. Therefore, for biological applications, the isomers
need not be separated.11 Nuclear magnetic resonance (NMR) and elemental
analysis data for caged cADPR indicate monocaging: 1H NMR
[(CH3)2 SO-d6] ~ 9.1-8.6 (m, 2H), 8.0-7.5 (m, 4H), 6.1-5.8 (m, 4H), 5.5 (m,
3H), 4.9 (m, 1H), 4.5-3.8 (m, 10H), 1.65 (m, 2H), 1.50 (dd, J = 18.4, 6.5
Hz, 1H). Anal. calcd for C23Hz8N6015P2 • H20: C, 39.05; H, 4.27; N, 11.88.
Found: C, 38.42; H, 3.94; N, 10.89. In vitro laser (308 nm) photolysis experi-
ments have determined that the photolysis quantum yield for caged cADPR
11R. Aarhus, K. Gee, and H. C. Lee, J. Biol. Chem. 270, 7745 (1995).
[22] CAGED cADP-gmOSE AND CAGED NAADP 407
is 0.11 and that photolysis produces free cADPR with a time constant of
55 msec. 12
Preparation of Caged NAADP

The synthesis of caged N A A D P presents more challenges, as there are
several potential sites for attachment of a caging group. Also, it was not
known beforehand how modification of these various positions on the
molecule would affect biological activity. N A A D P is synthesized by incu-
bating NADP (1 mM) at pH 5.0 with the Aplysia ADP-ribosyl cyclase
(25 ng/ml) in the presence of 30 mM nicotinic acid for several hours at
20-23 ° and purified by high-performance liquid chromatography (HPLC)
using an A G MP-1 column as described previously. 9 The procedure for
preparing caged N A A D P is similar to that described previously. 13
N A A D P has three phosphate groups and one carboxylate, all of which
are potentially reactive with the caging reagent, provided they are proton-
ated during the caging reaction. When the caging reaction is performed
with the aqueous phase at pH 4.5, a relatively nonpolar product is formed
(RI 0.87, methanol: chloroform: water: acetic acid, 13 : 10 : 3.5 : 0.2) This
product is biologically inactive and is photolyzed to N A A D P very slowly.
Reducing the pH of the caging reaction to 1.3 results in a caged product
that is much more photolabile. N A A D P (24 mg, 32/zmol) is dissolved in
2.5 ml of water, and the solution is adjusted to pH 1.3 by addition of dilute
aqueous HC1. The diazoethane 3 (0.2 mmol) in 3 ml ether is added and
stirred vigorously for 8 hr in darkness. The caging reagent solution is
replaced twice and stirred each time for 3 hr. Thin-layer chromatography
analysis indicates at this point no remaining N A A D P (RI 0.13) and two
major products with RI values of 0.87 and 0.46. The least polar product
(R1 0.87) coelutes on TLC with the higher pH product described earlier.
After the aqueous layer is extracted with ether (2 × 3 ml), the more polar
product (RI 0.46) is separated on a column of Sephadex LH-20 (2 × 8 cm).
A pale yellow powder of 8.6 mg (32% yield) of the caged product 5 is
obtained, which is much more photolabile than the higher pH product
described earlier. Photolysis with a hand-held UV lamp shows significant
conversion to free NAADP, as judged by TLC analysis.
High-Performance Liquid Chromatography Analyses of Caged
NAADP. HPLC separation of the reaction products is performed with
columns packed with A G MP-1 resin (Bio-Rad, Richmond, CA) and eluted
with a nonlinear gradient of trifluoroacetic acid. The chromatogram of the
12T. F. Walseth, R. Aarhus, M. E. Gurnack, L. Wong, H.-G. A. Breitinger, K. R. Gee, and

H. C. Lee, Methods Enzymol. 2811, 294 (1997).
13 H. C. Lee, R. Aarhus, K. R. Gee, and T. Kestner, J. BioL Chem. 272, 4172 (1997).
caged N A A D P product mixture using the AG MP-1 column, as well as
the gradient of TFA, is shown in Fig. 1 (top). Of the three components,
the one indicated with an asterisk is the most effective caged form of
NAADP, i.e., most responsive to photolysis. The final purification of the
caged N A A D P is achieved using a 0.5 × 5-cm Mono Q column (Pharmacia
Biotech Inc.). The product is eluted using a gradient of water (solvent A)
and 1 M triethylamine bicarbonate (solvent B, pH 8.8): 0-12 min, 0% B;
linearly increased to 20% B from 12 to 16 min, linearly increased to 30%
B from 16 to 36 min, linearly increased to 100% from 36 to 37 min, and
held at 100% B for 3 min before returning to 0% B. A chromatogram
1st AGMP-1
150
[...... i
....... ~ [, !
30
2rid Mono-Q r-- a.0

i
i
i o.6 ~
t i
10 20 30 40
Time (rain)
FIo. 1. Purification of caged N A A D P (5) by HPLC. Sequential chromatographic steps on
anion-exchange columns are used to purify a particular form of caged N A A D P from a mixture
of products generated by the caging reaction. The first step uses an A G MP-1 column (top).
The fraction most easily photolyzed is indicated by the asterisk. Final purification is achieved
using a Mono-Q column (bottom).
[221 CAGED cADP-RIBOSEAND CAGEDNAADP 409
representing the isolation of caged N A A D P using this gradient is shown
in Fig. 1 (bottom).
Characterization of Caged NAA D P. Proton-decoupled 3ap N M R spectra
of N A A D P and caged N A A D P are obtained on samples dissolved in a
D20 solution buffered at pH 7.5 with 10 mM HEPES. A phosphoric acid
solution in methanol is used as an external (zero) chemical shift standard.
The spectrum of N A A D P shows two signals in an integration of 2 : 1, with
the upfield resonance at -12.2 ppm and the smaller, downfield signal
(2.3 ppm). This downfield signal is absent in the spectrum of NAAD,
indicating that the downfield signal in the spectrum of N A A D P is due to
the 2'-ribosyl phosphate and that the upfield signal originates from the
two phosphorus atoms in the internucleotide diphosphate bridge. In the
spectrum for caged NAADP, the chemical shift of the upfield signal is
essentially unchanged, whereas that corresponding to the 2'-ribosyl phos-
phate is shifted upfield by about 5 ppm to -2.6 ppm, indicating that this
phosphate is the one modified in caged NAADP, as opposed to the diphos-
phate bridge.
To further confirm the structural characterization, enzymatic degrada-
tion of caged and uncaged N A A D P with alkaline phosphatase (AP) and
nucleotide pyrophosphatase (NP) is performed using the malachite green
assay TM to quantify free phosphate released during reaction. Treatment of
N A A D P with AP generates ca. 1 mol of free phosphate, by cleavage of
the 2'-ribosyl phosphate. Treatment of N A A D P with NP generates no free
phosphate, but subsequent treatment with AP results in the production of
about 3 mol free phosphate (2 mol from the internucleotide diphosphate
after cleavage by NP and 1 from the 2'-ribosyl phosphate). In contrast,
treatment of caged N A A D P with AP results generates no free phosphate.
Reaction of caged N A A D P with NP, followed by AP, results in the produc-
tion of 2 mol free phosphate, indicating that the 2-ribosyl phosphate is
modified (caged) and thus not available for enzymatic cleavage by AP.
Photolysis of Caged cADPR

Photoactivation of caged cADPR in sea urchin egg (Strongylocentrotus
purpuratus) homogenates can be achieved in a spectrofluorimeter (Hitachi
S-2000) using 350-nm excitation. Every 2 see the excitation wavelength is
alternated between 490 nm for monitoring Fluo-3 fluorescence and then
back to 350 nm for further uncaging. Frozen egg homogenates are thawed
at 17° for 20 rain, then diluted to 5% with a medium containing 250 mM
N-methylglucamine, 250 mM potassium gluconate, 20 mM HEPES, 1 mM
~4p. A. Lanzetta, L. J. Alvarez,P. S. Reinach, and O. A. Candia, Anal.Biochem.100,95 (1979).

MgC12, 2 units/ml creatine kinase, 8 m M phosphocreatine, 0.5 m M ATP,

and 3 / z M Fluo-3, p H 7.2, adjusted with acetic acid. The homogenates are
diluted to 2.5% and finally to 1.25% with the medium just described and
are incubated at 17 ° for 1 hr between dilutions. Calcium release is measured
in the 1.25% homogenates using an emission wavelength of 535 nm. The
m e a s u r e m e n t s are p e r f o r m e d in a cuvette maintained at 17 °, and the homog-
enates are stirred continuously. The volume of h o m o g e n a t e is 0.2 ml. Figure
2 shows that adding caged c A D P R to egg microsomes does not produce
calcium release, but that uncaging with 350-nm light results in calcium
mobilization after a brief delay. Addition of the specific c A D P R antagonist
8-NH2-cADPR stops the calcium release immediately, indicating that the
calcium release is due to uncaged c A D P R . Addition of the antagonist
before photolysis completely abolishes calcium release during uncaging. A
control using A T P caged with the same caging group shows no calcium mo-
bilization.
23
(-)8NH2
19
8
I--500 scc----I
ce~
Caged cADPR
11 ~ ~ z C~dATP
~ 2 . 5 pM
7 ~'~.~x'~UV~
FIG. 2. Calcium release from egg homogenates induced by caged cADPR after photolysis.
Photolysis of caged cADPR is performed by alternating the excitation wavelength of the
spectrofluorimeter between UV (350 nm) and the monitoring wavelength of 490 nm. Calcium
release from egg homogenates is monitored by Fluo-3 fluorescence in the absence [(-)8NH2]
or presence [(+)8NH2] of 2/zM of 8-amino-cADPR, a specific antagonist of cADPR. Caged
cADPR and caged ATP are added to the final concentrations indicated.
19 A aM B t--5 min--q
"~ 210
~M
17
100
- 75
15
0
¢- 5o
o 11
FI6. 3. Efficiency of photolysis of caged cADPR. The efficiency is assessed by comparing

the Ca 2÷ release activity of caged cADPR under constant UV illumination with cADPR.
Caged cADPR and cADPR are added at the final concentrations indicated. The time scale
shown applies to (A) and (B).
The uncaging efficiency can be determined by calibrating the calcium

release as a function of caged cADPR concentration versus that obtained
using free cADPR. Figure 3 shows the response of egg homogenates to
increasing concentrations of caged cADPR on UV photolysis compared to
the response obtained using increasing concentrations of cADPR. In this
series, uncaging 10/zM caged cADPR induces about the same calcium
release as about 100 nm free cADPR, indicating a photolysis efficiency of
about 1%; the low number is likely a function of the weak UV source
present in the spectrofluorimeter, but it is sufficient for most applications.
Photolysis of intact sea urchin eggs containing caged cADPR also results
in calcium release, monitored by Fluo-3, which is coinjected into the eggs
with the caged reagent. Lytechinus pictus eggs are microinjected by pressure
as described previously.15 Samples of Fluo-3 and caged cADPR are dis-
solved in an injection buffer containing 0.5 M KCI, 50/zM EGTA, 10 mM
HEPES, pH 6.7. Injection volumes are about 1.8-2.5% of the egg volume.
For photolysis of caged cADPR in individual eggs, the epifluorescence
attachment of a Nikon-inverted fluorescence microscope is modified. A
second mercury lamp is attached at 90° to the epifluorescence tube. The
light of 300-400 nm for photolysis is selected with a UGI filter (Omega)
and reflected 90°, first with a 400 DCLP dichroic filter and then directed
toward the objective with a second BCECF Sp dichroic filter. For monitor-
15 H. C. Lee, R. Aarhus, and T. F. Walseth, Science 261, 352 (1993).

412 CAGED COMVOUNDS [22]
ing Fluo-3 fluorescence, 490-nm light f r o m the first m e r c u r y lamp is selected

with a 485DF22 filter. T h e excitation light is passed t h r o u g h the same 400
D C L P dichroic filter and then reflected by the B C E C F Sp dichroic filter
t o w a r d the objective. T h e Fluo-3 fluorescence is selected by a long-pass
filter with a 500-nm cutoff and m o n i t o r e d by a SIT camera. This optical
a r r a n g e m e n t allows simultaneous m e a s u r e m e n t of Fluo-3 fluorescence dur-
ing uncaging photolysis. Figure 4 shows the expected calcium increase on
photolysis of caged c A D P R in intact eggs. A f t e r a slight delay, a threefold
increase in Fluo-3 fluorescence is o b s e r v e d consistently in eggs injected
(-)uv (+)uv
03
0 i J i i i i i i
0 100 200 300 400

Time (second)
FIG. 4. Calcium release induced by photolysis of caged cADPR in live sea urchin eggs.
Calcium release in individual eggs is monitored by Fluo-3 fluorescence. Caged cADPR
(-4.9 ~m, intracellular) or caged ATP (-72 /~m, intracellular) is coinjected with Fluo-3
(-0.25 mM, intracellular) into an egg. The injection volumes are about 1.8-2.5% of the egg.
Photolysis is induced by UV light at around 360 nm. The micrographs show an egg loaded
with caged cADPR before and after UV exposure. A fertilization membrane surrounding
the egg is formed after photolysis.
[22] CAGED cADP-RIBOSE AND CAGED NAADP 413
with 2.3 ± 0 .7 M.M caged cADPR. Also shown in Fig. 4 is the control
experiment using caged ATP. In addition to inducing calcium changes,
uncaging also induces the cortical reaction, an index ofcalcium mobilization .
The micrographs in Fig. 4 show an egg injected with caged cADPR before
and after UV exposure . After about 20-30 sec of UV exposure, a fertiliza-
tion membrane is formed surrounding the whole egg. No eggs injected with
caged ATP undergo the cortical reaction.
22
A
20
12
10
12
0
1 10 100 1000 10000
Concentration (W)
FIG. 5. Efficiency of photolysis of caged NAADP using aspectrofluorimeter . (A) Photolysis
was performed by alternating the excitation wavelength of the spectrofluorimeter between
350nm (UV) andthemonitoring wavelength of 485nm . Calcium release from egghomogenates
is monitored by Fluo-3 fluorescence . The calcium release activity of 450 nM caged NAADP
during UV is similar to that of 4-5 nM NAADP, indicating a photolysis efficiency of about
1% . (B) Concentration-response curves of NAADP, photolyzed caged NAADP (caged
NAADP + UV), and caged NAADP.
Photolysis of Caged NAADP
As for caged cADPR, alternating the excitation wavelength in a spectro-
fluorimeter between 350 nm for photolysis and 485 nm for monitoring
Fluo-3 fluorescence indicates that photolysis of caged N A A D P mobilizes
calcium in sea urchin egg homogenates. Figure 5 shows that addition of
caged N A A D P to the egg homogenates with UV excitation results in
calcium release. Comparison of the calcium release activity with that in-
duced by N A A D P itself shows that about 1% of the caged N A A D P is
photolyzed. Figure 5B shows the comparison of the concentration response
of N A A D P and caged N A A D P with or without UV photolysis. Because
N A A D P is effective in releasing calcium at nanomolar concentrations, the
low efficiency of photolysis (most likely due to the low intensity in the
spectrofluorimeter) does not limit its applications in cell biology.
A novel property of the NAADP-sensitive calcium release is that the
release mechanism can be totally inactivated by pretreatment with a sub-
8 A Iq 40 nM NAADP
t
"x
6
1'
0 2'
f~ Caged NAADP
4' 900nM
2_2. z
40 1 40 NAADP (nM) 90 180 360 540 900

Caged NAADP (aM)
FtG. 6. Assay of the contamination level of NAADP in the caged NAADP preparation.
The self-inactivation property of NAADP is used as an assay to estimate its contamination
in the caged NAADP preparation. (A) Treatment of sea urchin egg homogenates with 1 nM
NAADP for 8 min (A, 8' ) effectivelyinactivates the Ca 2+ release system such that the response
to 40 nM NAADP is essentially eliminated. The extent of inactivation induced by 540 nM
caged NAADP (B) is similar to that elicited by 1 nM NAADP (A, trace 2' ), indicating that
the level of contamination is about 0.1-0.2%. Ca 2+ release is monitored by Fluo-3.
123] CAGED CYCLIC NUCLEOTIDES 415
threshold concentration of NAADP as low as 1 nM. 7 Thus it is critical that

purification of caged NAADP results in a concentration of contaminating
free NAADP below the self-inactivating levels. Figure 6B shows that addi-
tion of purified caged NAADP from 90 to 900 nM to egg homogenates
produces no calcium release; the small jumps in Fluo-3 fluorescence at high
concentrations of caged NAADP are due to addition artifacts. Figure 6
also shows the results of testing the inactivating effect of caged NAADP.
Egg homogenates are pretreated with 1 nM NAADP and subsequently
challenged with a maximal concentration (40 nM) of NAADP (Fig. 6A).
After 2 min of pretreatment, the response to 40 nM NAADP is reduced
substantially (tracing labeled 2') as compared to no pre-treatment (trace
labeled 0') and is totally eliminated after 8 min of pretreatment (trace
labeled 8'). Figure 6B shows that pretreatment of the homogenates with
90 nM caged NAADP for 2 min produces very little inactivation. At
540 nM caged NAADP, the extent of inactivation following the pretreat-
ment is similar to that affected by 1 nM of NAADP (comparing Fig. 6A,
2', with Fig. 6B, 540 nM). This extent of inactivation by the caged compound
can be accounted for if the sample is contaminated with 0.1-0.2% free
NAADP. To ensure complete freedom from contaminating NAADP, sam-
ples of caged NAADP can be treated with AP immediately before use to
remove any free NAADP present. Alkaline phosphatase effectively re-
moves the T-phosphate of contaminating NAADP and converts it to
NAAD, which is biologically inactive. 13 The caging group attached to the
2'-ribosyl phosphate in caged NAADP protects the caged form from being
inactivated by AP.
[23] A p p l i c a t i o n s o f C a g e d C o m p o u n d s o f
Hydrolysis-Resistant Analogs of cAMP and cGMP
By U. BENJAMIN KAUPP, CLAUDIA DZEJA, STEPHAN FRINGS,
Jt3RGEN BENDIG, and VOLKER HAGEN
Cyclic nucleotides, i.e., cAMP and cGMP, control a variety of important

cellular processes (for reviews, see Walter I and Lincoln and Cornwell2).
The principal targets of cyclic nucleotides are cAMP- and cGMP-dependent
kinases (cAK and cGK), 1,3 cAMP- and cGMP-regulated phosphodiester-
1 U. Walter, Rev. Physiol. Biochem. Pharmacol. 113, 41 (1989).

z T. M. Lincoln and T. L. Cornwell, FASEB J. 7, 328 (1993).
3 S. S. Taylor, J, A. Buechler, and W. Y o n e m o t o , Annu. Rev. Biochem. 59, 971 (1990).

METHODS IN ENZYMOLOGY,VOL. 291 0076-6879f9g$25.00
123] CAGED CYCLIC NUCLEOTIDES 415
threshold concentration of NAADP as low as 1 nM. 7 Thus it is critical that

purification of caged NAADP results in a concentration of contaminating
free NAADP below the self-inactivating levels. Figure 6B shows that addi-
tion of purified caged NAADP from 90 to 900 nM to egg homogenates
produces no calcium release; the small jumps in Fluo-3 fluorescence at high
concentrations of caged NAADP are due to addition artifacts. Figure 6
also shows the results of testing the inactivating effect of caged NAADP.
Egg homogenates are pretreated with 1 nM NAADP and subsequently
challenged with a maximal concentration (40 nM) of NAADP (Fig. 6A).
After 2 min of pretreatment, the response to 40 nM NAADP is reduced
substantially (tracing labeled 2') as compared to no pre-treatment (trace
labeled 0') and is totally eliminated after 8 min of pretreatment (trace
labeled 8'). Figure 6B shows that pretreatment of the homogenates with
90 nM caged NAADP for 2 min produces very little inactivation. At
540 nM caged NAADP, the extent of inactivation following the pretreat-
ment is similar to that affected by 1 nM of NAADP (comparing Fig. 6A,
2', with Fig. 6B, 540 nM). This extent of inactivation by the caged compound
can be accounted for if the sample is contaminated with 0.1-0.2% free
NAADP. To ensure complete freedom from contaminating NAADP, sam-
ples of caged NAADP can be treated with AP immediately before use to
remove any free NAADP present. Alkaline phosphatase effectively re-
moves the T-phosphate of contaminating NAADP and converts it to
NAAD, which is biologically inactive. 13 The caging group attached to the
2'-ribosyl phosphate in caged NAADP protects the caged form from being
inactivated by AP.
[23] A p p l i c a t i o n s o f C a g e d C o m p o u n d s o f
Hydrolysis-Resistant Analogs of cAMP and cGMP
By U. BENJAMIN KAUPP, CLAUDIA DZEJA, STEPHAN FRINGS,
Jt3RGEN BENDIG, and VOLKER HAGEN
Cyclic nucleotides, i.e., cAMP and cGMP, control a variety of important

cellular processes (for reviews, see Walter I and Lincoln and Cornwell2).
The principal targets of cyclic nucleotides are cAMP- and cGMP-dependent
kinases (cAK and cGK), 1,3 cAMP- and cGMP-regulated phosphodiester-
1 U. Walter, Rev. Physiol. Biochem. Pharmacol. 113, 41 (1989).

z T. M. Lincoln and T. L. Cornwell, FASEB J. 7, 328 (1993).
3 S. S. Taylor, J, A. Buechler, and W. Y o n e m o t o , Annu. Rev. Biochem. 59, 971 (1990).

METHODS IN ENZYMOLOGY,VOL. 291 0076-6879f9g$25.00
ases (PDE), 4 cyclic nucleotide-gated (CNG) ion channels, 5-7 and transcrip-
tion factors. 8 The study of cAMP- and cGMP-signaling pathways has been
greatly facilitated by the design of chemical derivatives, dubbed "caged"
compounds. 9,1°The popular term cagedrefers to molecules whose biological
recognition or activity has been disabled by chemical modification. Photoly-
sis cleaves the modifying group, uncaging and rapidly releasing active mole-
cules. Caged cyclic nucleotides were first synthesized and physiologically
tested by Korth and Engels H and then successfully employed in photolysis
experiments to study a cAMP-dependent slow Ca 2+ current in the heart.12'~3
Caged cAMP and caged cGMP have also been used to study the activation
kinetics of the cGMP-gated channel from rod photoreceptor cells, TM the
signaling pathways in olfactory sensory n e u r o n s ( O S N s ) , 15'16 the enhance-
ment of CI- currents in ventricular cells, 17 and the motile response of fish
melanophores. 18
In the just-mentioned studies, caged compounds were used that released
either cAMP or cGMP. This can limit the experimental application of these
compounds as most cells are equipped with powerful PDE whose activity
rapidly degrades cyclic nucleotides and thus prevents persistent cellular
activation. Further, the low efficiency of photolysis (quantum yield) of
many caged compounds limits the size of the concentration step that can
be produced inside a cell by pulsed conventional light sources. For many
physiological studies it would be desirable to use caged compounds that
liberate cAMP or cGMP derivatives that are hydrolysis resistant or have
higher biological efficiency. Derivatives that meet both these criteria are
8-Br-cAMP and 8-Br-cGMP. They are poorly hydrolyzable by phosphodies-
4 j. A. Beavo and M. D. Houslay, in "Cyclic Nucleotide Phosphodiesterases: Structures,

Regulation and Drug Action." Wiley, Chichester, 1990.
5 K.-W. Yau and D. A. Baylor, Annu. Rev. Neurosci. 12, 289 (1989).
6 U. B. Kaupp, Curr. Opin. Neurobiol. 5, 434 (1995).
7 j. T. Finn, M. E. Grunwald, and K.-W. Yau, Annu. Rev. Physiol. 58, 395 (1996).
8 A. Kolb, S. Busby, H. Buc, S. Garges, and S. Adhya, Annu. Rev. Biochem. 62, 749 (1993).
10j. E. T. Corrie and D. R. Trentham, in "Bioorganic Photochemistry" (H. Morrison, ed.)
Vol. 21, p. 243. Wiley, New York, 1993.
11 M. Korth and J. Engels, Naunyn-Schmiedeberger's Arch. Pharmacol. 310, 103 (1979).
12j. Nargeot, J. M. Nerbonne, J. Engels, and 14. A. Lester, Proc. Natl. Acad. Sci. U.S.A. 80,
2395 (1983).
13j. M. Nerbonne, S. Richard, J. Nargeot, and H. A. Lester, Nature 310, 74 (1984).
14j. W. Karpen, A. L. Zimmerman, L. Stryer, and D. A. Baylor, Proc. Natl. Acad. Sci, U.S.A.
85, 1287 (1988).
15 G. Lowe and G. H. Gold, Z Physiol. 462, 175 (1993).
16T. Kurahashi and A. Menini, Nature 385, 725 (1997).
17 K. Ono, Y. Nakashima and T. Shioya, Pflagers Arch. 42,4, 546 (1993).
18T. Furuta, H. Torigai, M. Sugimoto, and M. Iwamura, J. Org. Chem. 60, 3953 (1995).
[23] CAGED CYCLIC NUCLEOTIDES 417
terases, 19 and 8-Br-cGMP activates cGK 2° and CNG channels 19~1'22 more
effectively than cGMP.
We have described previously the synthesis and properties of 4,5-di-
methoxy-2-nitrobenzyl (DMNB) derivatives of 8-Br-cAMP (in the follow-
ing referred to as DMNB-caged 8-Br-cAMP) and of 8-Br-cGMP (DMNB-
caged 8-Br-cGMP), which respectively release 8-Br-cAMP and 8-Br-cGMP
after irradiation with ultraviolet light. 23 The following article discusses the
properties of these compounds and their usefulness for electrophysiological
studies of CNG channels in situ using heterologously expressed a subunits
of CNG channels from bovine OSNs and cone photoreceptor cells.24,25
Design of Caged Derivatives of Hydrolysis-Resistant

Cyclic Nucleotides
Cyclic nucleotides such as cAMP or cGMP can be rendered hydrolysis
resistant via the addition of substituents to the heterocyclic purine base
and caged via esterification at the phosphate moiety with a photolabile
group. Carbon C-8 of the purine ring system is the position by which cyclic
nucleotides can be modified without compromising their biological efficacy.
It is likely that many C-8 substituted congeners are hydrolysis resistant,
although this has been demonstrated experimentally for only a few deriva-
tives. Some C-8 substituted derivatives are several orders of magnitude
more effective in activating CNG channels than their respective natural
agonist. 19,21'22'26'27 Specifically bulky and apolar substituents at C-8 render
derivatives very potent agonist. For example, 8-fluoresceinylcarbamoyl-
methylthio-cGMP activates the rod CNG channel at almost 100-fold lower
concentrations than cGMP. 26 In addition, apolar C-8 substituents facilitate
permeation across biological membranes. One drawback of this type of
modification in combination with apolar caging groups is that it results in
19 A. L. Zimmerman, G. Yamanaka, F. Eckstein, D. A. Baylor, and L. Stryer, Proc. Natl.

Acad. Sci. U.S.A. 82, 8813 (1985).
20j. D. Corbin, D. Ogreid, J. P. Miller, R. H. Suva, B. Jastorff, and S. O. DCskeland, J. Biol.
Chem. 261, 1208 (1986).
21 K.-W. Koch and U. B. Kaupp, J. Biol. Chem. 260, 6788 (1985).
22 j. C. Tanaka, J. F. Eccleston, and R. E. Furman, Biochemistry 2,8, 2776 (1989).
23 V. Hagen, C. Dzeja, S. Frings, J. Bendig, E. Krause, and U. B. Kaupp, Biochemistry 35,
7762 (19%).
24 j. Ludwig, T. Margalit, E. Eismann, D. Lancet, and U. B. Kaupp, FEBS Lett. 270, 24 (1990).
25 I. Weyand, M. Godde, S. Frings, J. Weiner, F. MUller, W. Altenhofen, H. Hatt, and U. B.
Kaupp, Nature 368, 859 (1994).
26 A. Caretta, A. Cavaggioni, and R. T. Sorbi, Eur. J. Biochem. 153, 49 (1985).
27 R. L. Brown, R. J. Bert, F. E. Evans, and J. W. Karpen, Biochemistry 32, 10089 (1993).
TABLE I
SOLUBILITYOF DMNB ESTERS OF 8-Br-cAMP AND
8-Br-cGMP a
Cyclic nucleotide ester Solubility
DMNB-caged 8-Br-cAMP
Axial isomer 120
Equatorial isomer 80
DMNB-caged 8-Br-cGMP
Axial isomer 100
Equatorial isomer 10
a In 5% acetonitrile/0.01 M HEPES/KOH buffer,

pH 7.2, containing 0.12 M KC1 at room tempera-
ture. Saturation concentrations in izM 1 hr
after dissolution.
reagents that are significantly less soluble (Table I). This disadvantage is
offset by a similarly large increase in ligand affinity at most C-8 substi-
tuted derivatives.
The rather low quantum yield of the uncaging process limits the concen-
tration step achieved by short flashes from pulsed conventional light sources,
such as xenon flash lamps or mercury lamps. Therefore, full activation by
a single flash is difficult to achieve. The much higher biological efficacy of
C-8 substituted derivatives, in principle, would be also beneficial in this
respect. For most purposes, however, it is the contamination by "free"
cyclic nucleotides that sets a practical upper limit for the concentration of
caged compound rather than its low solubility (see "Optimal Concentra-
tions and Handling of Solutions").
Synthesis
The synthesis of DMNB esters of 8-Br-cAMP and 8-Br-cGMP (Fig. 1,
see Hagen et aL 23) is based on a procedure described for DMNB-cAMP
and DMNB-cOMP, 13 but using 4,5-dimethoxy-2-nitrophenyldiazomethane
prepared from the hydrazone of 4,5-dimethoxy-2-nitrobenzaldehyde ac-
cording to Wootton and Trentham. 28
Briefly, freshly prepared 4,5-dimethoxy-2-nitrophenyldiazomethane is
coupled directly to the free acid of 8-Br-cAMP or 8-Br-cGMP by stirring
suspensions of the reactants in dimethyl sulfoxide (DMSO) at room temper-
28 j. F. Wootton and D. R. Trentham, in "Photochemical Probes in Biochemistry" (P. E.

Nielsen, ed.), Vol. 272, p. 277. Kluver Academic, The Netherlands, 1989.
[23] CAGED CYCLICNUCLEOTIDES 419
RIt
H3CO~__,
R , ~ . . ~ N ~ I ~X~-Br
NO2
DMSO o'~P~-~~ O
40h, r.t. HO
~w - N2
R"
R,L N ~ x~--Br
"3C P H3CO
?
OH ~ O
N.02
axial isomer equatorial isomer
8-Br-cAMP and DMNB--caged8-Br-cAMP: R' = H, R" = NH2

8-Br-cGMP and DMNB-caged 8-Br-cGMP: R' = NH2, R" = OH
FIG. 1. Structures of diastereoisomers of DMNl~-,.aged 8-Br-~.AMP and DMNB-caged

8-Br-cGMP.
ature in the dark for 40 hr. Due to a center of asymmetry on phosphorus,

all triesters occur as diastereomeric mixtures. The resulting pairs of diaste-
reomers of the caged compounds are isolated and separated into the axial
and equatorial forms (for structures, see Fig. 1) using preparative reversed-
phase (RP) high-performance liquid chromatography (HPLC) (for details
of the chromatographic procedure, see Hagen et aL23). For maximum stabil-
ity the pure products are lyophilized (see later). Isomeric species are as-
signed by 31p nuclear magnetic resonance (NMR) in analogy to the corre-
sponding cAMP and cGMP esters. 13,29The 31p NMR signals at higher field
(-5.19 for DMNB-caged 8-Br-cAMP and -5.17 for DMNB-caged 8-Br-
cGMP) correspond to the axial isomers, and those at lower field (-3.37
for DMNB-caged 8-Br-cAMP and -4.27 for DMNB-caged 8-Br-cGMP)
29j. Engels, Bioorg. Chem. 8, 9 (1979).

A B
r |
15 30 0 15
retention time (min) retention time (rnin)
FIr. 2. Reversed-phase HPLC profiles of DMNB-caged 8-Br-cGMP. (A) Mixture of the

axial (peak a, retention time 10.2 min) and the equatorial isomer (peak e, retention time 13.2
min) after flash chromatography.The retention time for 8-Br-cGMPwas 1.4 min. (B) HPLC
profileof the purifiedaxialisomer after preparative HPLC and lyophilisation(PLRP-Scolumn,
120 × 4.6 mm i.d., 8 m; linear gradient of 20-80% B in 40 min; eluent A, water, eluent B,
acetonitrile; flow rate: 1 ml/min; detection at 280 nm). [Reprinted with permission from V.
Hagen, C. Dzeja, S. Frings, J. Bendig, E. Krause, and U. B. Kanpp. Caged compounds
of hydrolysis-resistant analogues of cAMP and cGMP: synthesis and application to cyclic
nucleotide-gated channels. Biochemistry35, 7762-7771 (1996). Copyright (1996) American
Chemical Society.].
correspond to the equatorial forms. In R P - H P L C profiles (Fig. 2) the axial

isomers show shorter retention times compared to the equatorial isomers.
Thin-layer chromatography (silica 60-F2s4, chloroform/methanol 5 : 1, v/v)
gives R: values of 0.71 and 0.61 for the axial and the equatorial isomers of
DMNB-caged 8-Br-cAMP, respectively, and 0.54 and 0.60 for the axial and
the equatorial isomers of DMNB-caged 8-Br-cGMP. The synthesis shows
slight (approximately 60%) preferential yield of the caged axial isomer, but
all isomers are isolated in acceptable yields and high purities.
Purity
The purity of caged compounds in general, and in particular of hydroly-
sis-resistant cyclic nucleotides, is of critical importance. The following sec-
tion discusses two aspects of reagent purity, namely separation of isomer
mixtures into their respective diastereomeric components and contamina-
tion by the "free," i.e., uncaged, cyclic nucleotide.
On UV illumination the natural cyclic nucleotide is released from both

the axial and the equatorial isomer of the caged compound; therefore, in
principle, both isomers would be useful for cellular studies. Isomers differ
in their solubility and solvolytic stability by at least one order of magnitude
(see later); it is therefore advisable to employ pure diastereomers, then,
for example, progression of solvolysis inside the cell during an experiment
is reduced as well as being more predictable.
Contamination of caged compounds with free cyclic nucleotides is of
particular importance with hydrolysis-resistant derivatives because the con-
taminant is not taken care of by endogenous PDE activity. Free cyclic
nucleotide contamination results from two different sources: incomplete
caging of the primary product and solvolysis of the triester in aqueous
solution. Separation of isomers by RP-HPLC also removes free cyclic nucle-
otides. Thereby, reagents of 99.9% purity are routinely obtained. Practical
hints that help avoid contamination by solvolysis are given later (see "Opti-
mal Concentrations and Handling of Solutions").
Solubility
The successful use of caged compounds in cellular research requires
adequate solubilities of the probes in aqueous solutions of moderately high
ionic strength. Therefore, the diastereomers of DMNB-caged 8-Br-cAMP
and 8-Br-cGMP are tested for their solubilities in 5% acetonitrile/0.01 M
HEPES/KOH buffer, pH 7.2, containing 0.12 M KCI at room temperature
using RP-HPLC.
When the isomers are dissolved first in pure acetonitrile and then diluted
into the buffer, saturation concentrations are obtained as shown in Table
I. The greater solubility of axial isomers in aqueous solution, particularly
DMNB-caged 8-Br-cGMP, favors the use of pure axial forms in cellular
physiology studies.
Photochemical Properties
The DMNB esters of 8-Br-cAMP and 8-Br-cGMP have an absorption
peak around 350 nm, a wavelength, that is relatively harmless to cells.
Irradiation at 300-400 nm of the axial and the equatorial isomers of DMNB-
caged forms of 8-Br-cAMP and 8-Br-cGMP resulted in the liberation of
8-Br-cAMP or 8-Br-cGMP anions, 4,5-dimethoxy-2-nitrosobenzaldehyde,
and a proton. The photolysis mechanism is well-known. 1°'29,3°'31 The effi-

31 j. p. y. Kao and S. R. Adams, in "Optical Microscopy: Emerging Methods and Applications
(B. Herman and J. L. Lemasters, eds.), p. 27. Academic Press, San Diego, 1993.
ciency of photolysis may be evaluated by both molar extinction coefficients

(e) and quantum efficiencies (~) for the disappearance of phosphate esters.
Table II summarizes e values obtained at the long-wavelength absorption
maximum and the qb values of the axial and equatorial forms of DMNB-
caged 8-Br-cAMP and DMNB-caged 8-Br-cGMP. Within experimental
accuracy, the photochemical properties of the two diastereomers of DMNB-
caged 8-Br-cAMP and of DMNB-caged 8-Br-cGMP were identical. The
compounds have a relatively high absorbance at 300-400 nm but rather
low quantum yields. The quantum yields of the diastereomers of DMNB-
caged 8-Br-cAMP are 10-fold larger than of the diastereomers of DMNB-
caged 8-Br-cGMP. The photochemical properties are similar to those of
DMNB-caged cAMP and DMNB-caged cGMP. 28
Solvolysis
An ideal caged compound must be hydrolytically stable. With respect
to caged derivatives of 8-Br-cAMP and 8-Br-eGMP, hydrolytic stability is
crucially important because "free" 8-Br-cAMP or 8-Br-cGMP is a particu-
TABLE II
PHOTOCHEMICALPROPERTIESFOR DMNB ESTERSOF 8-Br-cAMP AND 8-Br-cGMP a
Long wavelength
absorption maximum
Cyclic nucleotide ester (nm) Quantum yield qbb
Axial isomer 345 (e 5900) 0.050
Equatorial isomer 345 (e 5900) 0.045
Axial isomer 346 (e 5800) 0.0049
Equatorial isomer 346 (e 5800) 0.0052 c
a In 5% acetonitrile/0.01 M HEPES/KOH buffer, pH 7.2, containing 0.12 M KC1.

d~ values were determined by the relative method [from H. J. Kuhn, S. E. Braslav-
sky, and R. Schmidt, Pure Appl. Chem. 61, 187 (1989)] using the standard E,E-
1,4-diphenyl-l,3-butadiene [Reprinted with permission from V. Hagen, C. Dzeja,
S. Frings, J. Bendig, E. Krause, and U. B. Kaupp, Caged compounds of hydrolysis-
resistant analogues of cAMP and cGMP: synthesis and application to cyclic nucleo-
tide-gated channels. Biochemistry 35, 7762 (1996). Copyright (1996) American
Chemical Society.].
b Aexc 333 nm.
c In dioxane/0.01 M sodium phosphate buffer, pH 7.2.
[231 CAGED CYCLIC NUCLEOTIDES 423
lady powerful activator and furthermore their degradation by endogenous

phosphodiesterases proceeds at least 500 times slower than that of natural
cyclic nucleotidesJ 9
In DMNB esters of 8-Br-cAMP and 8-Br-cGMP, as well as of cAMP
and cGMP, the electron-releasing dimethoxy groups on the benzene ring
increase the stability of the benzyl carbocation, which makes these com-
pounds particularly susceptible to hydrolysis. Half-life values of the pseudo
first-order hydrolysis of the diastereomers of DMNB esters of 8-Br-cAMP
and 8-Br-cGMP in 5% acetonitrile/0.01 M H E P E S / K O H buffer, pH 7.2,
containing 0.12 M KC1, are given in Table III. Additionally, half-life values
are listed for the DMNB-caged 8-Br-cGMP isomers in different aqueous
buffer solutions.
The axial isomers of both caged compounds are sufficiently stable with
respect to the time required for photochemical or electrophysiological ex-
periments. In contrast, the respective equatorial isomers are hydrolyzed at
pH 7.2 or 4.6 approximately 7-10 times faster than the axial forms. These
differences make the axial form much more convenient for physiological
experiments than the equatorial form or mixtures of the respective diaste-
reomers. For example, after 0.5 hr in aqueous buffer, the axial form of
DMNB-caged 8-Br-cGMP contains approximately 1% 8-Br-cGMP, whereas
the equatorial form contains 6%.
The diastereomers of DMNB-caged cAMP and cGMP show similar
stabilities in aqueous buffer solutions (V. Hagen, unpublished). In all exper-
iments with these derivatives, we recommend using pure axial isomers.
TABLE III
CALCULATED HALF-LIFE (tl/2) FOR DMNB ESTERS Of 8-Br-cAMP A~D
8-Br-cGMP a
Cyclic nucleotide ester Aqueous buffer pH tl/2 (hr)
Axial isomer HEPES/KC1 7.2 60
Equatorial isomer HEPES/KCI 7.2 8
Axial isomer HEPES/KC1 7.2 50
Axial isomer Phosphate 7.2 50
Axial isomer Phosphate 4.6 50
Equatorial isomer HEPES/KCI 7.2 5
Equatorial isomer Phosphate 7.2 6
Equatorial isomer Phosphate 4.6 5
In aqueous buffer/acetonitrile (95 : 5).

A Mg2+_free MgZ+-free
50 500ms
30s
B
2+
Idg -free Mg2+-free
50 500ms
30s
C idgZ+_free
50 lOOms
1 1
20?pA
los 50 lOOms
FIG. 3. Whole cell current responses recorded after photolysis of DMNB-caged 8-Br-cGMP
(A) and DMNB-caged 8-Br-cAMP (B) from HEK 293 cells expressing the olfactory ot subunit.
(A) Current was activated by 50- and 500-msec flashes of U V light (arrows). Black and white
bars indicate superfusion of the cell with Mg 2÷-containing and Mg 2÷-free solutions, respectively.
Recording pipette contained 6.25/zM DMNB-caged 8-Br-cGMP. The membrane voltage was
held at V,, = - 7 0 inV. The initial Mg2+-sensitive current was 106.5 pA and the maximal
current was 2154 pA. This corresponds to a 8-Br-cGMP contamination of 41 n M or 0.66%.
[231 CAGED CYCLICNUCLEOTIDES 425
"Optimal" Concentrations a n d Handling of Solutions

The following section discusses criteria to choose the optimal concentra-
tion of caged cyclic nucleotides. This concentration is determined by (1)
ligand affinity, (2) the contamination by free cyclic nucleotides, and (3) the
tolerable level of "background" activation. For example, the mean constant
of half-maximal activation KI/2 and the Hill coefficient, n (at - 5 0 mV), are
0.13 _ 0.03/zM and 2.6, respectively, for the activation of the olfactory
CNG channel a subunit by 8-Br-cGMP (see Fig. 5B). If 10% of the maximal
current is considered tolerable as a background current, and if the caged
compound is contaminated with 0.5% of free 8-Br-cGMP, then up to 11.2
~M of DMNB-caged 8-Br-cGMP can be used in the pipette. If experiments
require a lower background current, either the DMNB-caged 8-Br-cGMP
must be further purified or its concentration must be decreased accordingly.
However, in order to produce the same incremental increase in A[8-Br-
cGMP] at lower concentrations of the caged compounds a stronger (or
longer) light flash is needed. Therefore, a practical lower limit of the concen-
tration of caged cyclic nucleotides is set by the energy output of the actinic
light source and possibly also by deleterious effects of strong UV illumina-
tion on the cell. The KI/2 value of activation of the olfactory a subunit by
8-Br-cAMP is roughly 10-fold larger compared to 8-Br-cGMP; accordingly,
10-fold higher concentrations of DMNB-caged 8-Br-cAMP may be em-
ployed (Fig. 3B). In summary, an "optimal" concentration of caged com-
pound is the highest concentration that can be employed, given a tolerable
level of background activation by free cyclic nucleotides. This "optimal"
concentration may be estimated from the KI~2 value of channel activation
and the level of contamination.
The solution may also become contaminated with free cyclic nucleotide
during an experiment due to solvolysis in aqueous solution. Solvolysis can
be kept at a minimum by dissolving small aliquots of lyophilized reagent in
water-free DMSO. Final concentrations are adjusted by preparing aqueous
dilutions of this stock solution shortly before the experiment.
(B) Flash-inducedwholecell current responseswith DMNB-caged8-Br-cAMP.The recording

pipette contained 100/zM DMNB-caged 8-Br-cAMP. All other conditionswere as in (A),
except Vm= -50 mV. (C) Control experimentswith transfectedHEK 293 cells and 100/zM
caged ATP in the recording pipette (upper trace) and nontransfected HEK 293 cells with
6.25/~M DMNB-caged 8-Br-cGMP in the recording pipette (lower trace); Vm = -50 mV.
[Reprinted with permission from V. Hagen, C. Dzeja, S. Frings, J. Bendig, E. Krause, and
U. B. Kaupp. Caged compounds of hydrolysis-resistantanalogues of cAMP and cGMP:
synthesis and application to cyclicnucleotide-gated channels. Biochemistry 35, 7762-7771
(1996). Copyright (1996) American ChemicalSociety.].
Cyclic Nucleotide-Gated Channel Activation by Photolysis of Caged

Cyclic Nucleotides
DMNB-caged 8-Br-cGMP and DMNB-caged 8-Br-cAMP were tested
in H E K 293 cells expressing the ~ subunit of the CNG channel from
b o v i n e O S N s . 24'32
Figure 3A shows a whole cell recording of membrane current evoked
by photolysis of DMNB-caged 8-Br-cGMP. The cell was initially bathed
in a solution containing 10 mM Mg 2+, which effectively blocks currents
activated by 8-Br-cGMP during equilibration of the cell with the caged
compound. 32 Shortly before illumination by UV light, the perfusion was
switched to a solution that contained 10 mM EGTA. In most experiments,
a small current of several tens of picoamprres developed after removal of
Mg 2+. The small Mg2+-sensitive currents are not due to the caged compound
itself, but are completely accounted for by the activation of CNG channels
with contaminating 8-Br-cGMP. No, or negligible, Mg2+-sensitive currents
were observed in nontransfected cells or in cells that have been equilibrated
with caged ATP (Fig. 3C). This Mg2+-sensitive current progressively in-
creased during an experimental session due to solvolytic hydrolysis of
DMNB-caged 8-Br-cGMP.
On irradiation of the cell with flashes of UV light (50 or 500 msec),
large currents were recorded (Fig. 3A). The following control experiments
indicate that these currents were activated by liberation of 8-Br-cGMP
from the caged compound: (i) in the absence of the caged compound or
in nontransfected H E K 293 cells, there was no current response to a light
flash (Fig. 3C); (ii) the light-evoked current responses were saturable. A
500-msec flash activated ->90% of the maximal current; subsequent flashes
caused much smaller increments of current or no changes at all (see also
Fig. 5A). (iii) The flash-induced current was almost entirely and reversibly
suppressed by extracellular Mg 2+ (Fig. 3A) and (iv) the nitrosobenzalde-
hyde photolysis product did not change the membrane conductance because
photolysis of caged ATP (100/zM), which has the same caging group as
DMNB-caged 8-Br-cGMP, was ineffective (Fig. 3C). We conclude that
the light-evoked current was activated by 8-Br-cGMP produced by the
photolysis of DMNB-caged 8-Br-cGMP.
Figure 4 shows a photolysis experiment using either NPE-caged cGMP
or DMNB-caged 8-Br-cGMP in the pipette. Because the Kla constant of
the olfactory channel for cGMP is ~ 12-fold larger than that for 8-Br-cGMP,
the pipette solution contained 75/~M NPE-caged cGMP. The flash-evoked
currents differ from each other in that the rapid rise produced by NPE-
32S. Frings, R. Seifert, M. Godde, and U. B. Kaupp, Neuron 15, 169 (1995).
UV
0.0 r
x ~ cGMP
E 0.5
.... _ _ cGMP+IBMX
1.o 8-Br-cGMP
los
FIG. 4. Comparison of the time courses of flash-induced whole cell currents produced by
either NPE-caged cGMP or DMNB-caged 8-Br-cGMP. Concentrations in the pipette were
75 g M for NPE-caged cGMP, 6.25 tzM for DMNB-caged 8-Br-cGMP, and 0.3 mM for IBMX.
Flash-induced currents measured from three different cells were normalized to the peak of
the response shortly after the flash. Maximal whole cell currents produced by cGMP were
-1.2 nA ( - I B M X ) , -2.3 nA (+IBMX), and -2.1 nA produced by 8-Br-cGMP; V,, =
- 5 0 mV, 50-msec flashes. [Reprinted with permission from V. Hagen, C. Dzeja, S. Frings, J.
Bendig, E. Krause, and U. B. Kaupp. Caged compounds of hydrolysis-resistant analogues of
cAMP and cGMP: synthesis and application to cyclic nucleotide-gated channels. Biochemistry
35, 7762-7771 (1996). Copyright (1996) American Chemical Society.].
caged cGMP is followed by a slower decay, whereas the current produced

by DMNB-caged 8-Br-cGMP remains constant. The time of half-maximal
decay of the cGMP-induced current was ~'t~ ~- 34.1 + 15.6 sec (9 experi-
ments), range 14.9-65.6 sec. The current decline was almost completely
abolished by 0.3 mM of the PDE inhibitor 3-isobutyl-l-methyl-xanthine in
the pipette solution. We interpret these results to indicate that cGMP
released by light is hydrolyzed by endogenous PDE activity, whereas 8-
Br-cGMP is not.
Calibration
The photolysis-induced step in cyclic nucleotide concentration gener-
ated inside the cell can be accurately calibrated in situ. It requires two
different measurements. First, the dose-response relationship of the CNG
channel must be determined in excised inside-out patches. Figure 5B shows
dose-response relationships of 8-Br-cAMP and 8-Br-cGMP for the olfac-
tory a subunit. Second, the ratio Inash/Im~ between the flash-induced current
(Ina~h) and the maximal current (/max) at saturating concentrations of cyclic
nucleotides must be determined./max can be determined either by illuminat-
ing cells with repetitive light flashes until whole cell currents saturate or
by a single bright flash. From the Ina~h/Ima~ ratio and the dose-response
A
Mg2+-free
. . . . . . . . . . . .
flash
I x ..... 2 - ~
B C
0.4
O
o
E 0.5 oo 0.2 0
L o
? o
o
0.0 ........ I ........ I ~ ....... t ........ I

0.0 I I I I I
0.01 0.1 1 10 100 0 1 2 3 4 5

Llgand (/~,M) No. of UV flashes (50ms)
FIG. 5. Calibration of flash-induced changes in intracellular 8-Br-cOMP. (A) Current

responses to successive flashes of UV light. Flash duration was 50 msec. The recording pipette
contained 6.25 /zM DMNB-caged 8-Br-cGMP. The initial concentration of uncaged 8-Br-
cGMP in the pipette solution was 34 nM, corresponding to a contamination of 0.54%. The
concentration after the fifth flash was 226 nM. The average flash-induced increment of the
8-Br-cGMP concentration was 38.4 _+ 2.4 nM (5); Vm = - 5 0 mV. (B) Dose-response relation
of the heterologously expressed c~ subunit of the olfactory CNG channel for 8-Br-cGMP (11)
and 8-Br-cAMP (I~). Steady-state cyclic nucleotide-activated currents I measured from inside-
out patches at - 5 0 mV were normalized by the maximal current/max at saturating concentra-
tions of the respective nucleotide. Data were obtained from five experiments for 8-Br-cAMP
and from nine experiments for 8-Br-cGMP. The curves represent a least-squares fit to the
Hill equation. Mean values _ SD of the constant of half-maximal activation K1~2and the Hill
coefficient n were 13.3 -+ 1.7/zM and 2.2 -+ 0.1 for 8-Br-cAMP and 0.13 _+ 0.03/~M and 2.6 _+
0.1 for 8-Br-cGMP. (C) Increase of 8-Br-cGMP concentration with number of U V flashes,
calculated from the dose-response relation of (A) and the ratios of lh~/Imax determined from
the current recording of (B). [Reprinted with permission from V. Hagen, C. Dzeja, S. Frings,
J. Bendig, E. Krause, and U. B. Kaupp. Caged compounds of hydrolysis-resistant analogues
of cAMP and cGMP: synthesis and application to cyclic nucleotide-gated channels. Biochemis-
try 35, 7762-7771 (1996). Copyright (1996) American Chemical Society.].
relationship, the concentration step per flash can be calculated. Figure 5A

shows current recordings after repetitive irradiation with 50-msec flashes.
The current amplitude increased incrementally with the number of flashes
until it saturated after roughly 10 flashes. A concentration jump of 38.4 __+
2.4 nM 8-Br-cGMP/flash was calculated from the first five flashes. This
experiment also demonstrates that the amount of cyclic nucleotide released
per flash is roughly constant, i.e., the [8-Br-cGMP] increases linearly with
the flash number (Fig. 5C). This conclusion is only valid when the incremen-
tal increase in cyclic nucleotide concentration is much smaller than the
concentration of caged compound (i.e., A[8-Br-cGMP] ~ [DMNB-caged
8-Br-cGMP]).
Instead of applying successive flashes, light flashes of longer duration
may be used. Figure 6A shows such an experiment with flashes of 10, 50,
A
1 Ores 50ms 100rns
5s 2s 5s
oo 2
I
i 1
00
I I I I I I I I I I
10 50 100
f l a s h duratTon ( m s )
FIo. 6. Amplitude of flash-induced whole cell currents recorded from HEK 293 cells
expressing the a subunit of bovine cone photoreceptors. (A) Currents evoked by flashes of
10 (left), 50 (middle), and 100 (right) msec duration. The pipette contained 15/~M of DMNB-
caged 8-Br-cGMP; Vm = --50 mV. The saturating light flash was 1 sec in duration. (B)
Incremental increase of 8-Br-cGMP concentration with flash duration, calculated from the
dose-response relation and the ratios of Ihv/Imax (m).
and 100 msec duration. The incremental change in intracellular [8-Br-

cGMP] is linearly related to the duration of the flash (Fig. 6B).
Poorly Hydrolyzable vs Hydrolyzable Analogs of Cyclic Nucleotides

The principal advantage of C-8 substituted derivatives of cyclic nucleo-
tides is their hydrolysis-resistant nature. Caged derivatives of these com-
pounds therefore allow persistent elevation of cyclic nucleotides in situ.
The decay of the flash-induced concentration step on a time scale of several
minutes is probably determined by slow dialysis of the cell with the pipette
solution and diffusion of liberated C-8 substituted cyclic nucleotides across
the cell membrane rather than by PDE hydrolysis. The hydrolysis-resistant
nature along with the high biological efficacy of most C-8 substituted cyclic
nucleotides require reagents of extremely high purity. For example, a con-
tamination of DMNB-caged 8-Br-cGMP (100/xM) with 1% of the free (8-
Br-substituted) cyclic nucleotide would result in the activation of ->90% of
the olfactory CNG channel in a typical transiently expressing H E K 293 cell.
In this respect, caged derivatives of the natural cyclic nucleotides are
more convenient to use than their C-8 substituted congeners. Residual free
cyclic nucleotides either become degraded by endogenous PDE activity
while the cell is equilibrated with the caged compound or it can be removed
beforehand by treatment of solutions with trypsin-activated PDE. TMIn addi-
tion, due to hydrolysis of the flash-generated cyclic nucleotide, the experi-
ment can be repeated many times on the same cell as long as it contains
sufficient amounts of caged compound. However, considering the modest,
if not poor, purity of commercially available caged cyclic nucleotides, that
advantage may be elusive because unintentional activation of the cellular
process under study may occur before endogenous PDE had time to remove
free cyclic nucleotide. For this reason, it is mandatory to check chromato-
graphically the purity of each batch, even of hydrolyzable cyclic nucleotides.
Acknowledgments
We thank Dr. J. Bradley (Paris) for critical reading and A. Eckert for preparing the
manuscript. This work was supported by a grant from the Deutsche Forschungsgemeinschaft
(to S.F.). ClaudiaDzeja is the recipientof a fellowshipfrom the Studienstiftungdes Deutschen
Volkes. We are also grateful to the Fonds der ChemischenIndustrie for financialsupport.
[24] Nmoc-CAGED
GLUTAMATEAND Ca2+ATPase INHIBITOR 431
[24] C a g e d P r o b e s f o r S t u d y i n g C e l l u l a r P h y s i o l o g y :
A p p l i c a t i o n of o - N i t r o m a n d e l y l o x y c a r b o n y l (Nmoc) C a g i n g
Method to Glutamate and a Ca2+-ATPase Inhibitor
B y FRANCISM. RossI, MICHAELMARGULIS,ROBERT E. HOESCH,
CHA-MIN TANG, and JOSEPH P. Y. KAO
The release of bioactive molecules from inert, photosensitive precursors

by flash photolysis is an increasingly useful technique in cellular experimen-
tation. 1'2 The precursors, wherein some functional group that is essential
for bioactivity has been chemically modified by a photolabile masking group
to abolish activity temporarily, are often referred to as "caged" molecules.3
This article focuses on a newly developed caging group, two caged molecules
based thereon, as well as biological applications of these two new caged re-
agents.
o-Nitromandelyloxycarbonyl (Nmoc} Caging Group

The newly introduced photosensitive o-nitromandelyloxycarbonyl
(Nmoc) caging group offers several advantages. First, flash photolysis of
Nmoc-caged species releases bioeffector molecules on the millisecond time
scale, with quantum yields of -0.1. Second, the Nmoc cage is synthesized
with relative ease. Third, Nmoc-imidazole (Fig. 1), a chemically activated
form of the Nmoc cage, reacts smoothly with hydroxyl and amino groups
to form, respectively, carbonate and carbamate linkages that are highly
resistant to spontaneous cleavage in the absence of light. These characteris-
tics combine to make Nmoc an attractive caging group for biomolecules
bearing hydroxyl or amino groups.
The active caging reagent, Nmoc-imidazole, is prepared in two versions:
the methyl and tert-butyl ester forms [Fig. 1, R = CH3 and C(CH3)3,
respectively]. For caging hydroxyl groups in biomolecules, the methyl ester
form of Nmoc-imidazole is quite satisfactory. This form readily reacts with
hydroxyl groups to form stable carbonate ester linkages that remain intact
while the methyl group on the cage is removed by base hydrolysis to liberate
1S. R. Adams and R. Y. Tsien,Ann. Rev. Physiol. 55, 755 (1993).

2j. p. y. Kao and S. R. Adams,in "OpticalMicroscopy:EmergingMethodsand Applications"
(B. Herman and J. J. Lemasters,eds.), p. 27. AcademicPress, San Diego, 1993.
3j. H. Kaplan,B. Forbush III, and J. F. Hoffman,Biochemistry 17, 1929 (1978).

RO2C- O,~NC N
O2N~ 0
1-Nmoc-imidazole
R= methyl,or tea-butyl
FIG.1. Structureof Nmoc-imidazole.
the final caged molecule. Nmoc-DBHQ, 4 a reagent for photoreleasing

DBHQ, an inhibitor of SR/ER Ca2÷-ATPase,5-7 was prepared using methyl
Nmoc-imidazole.8 For caging amino groups, however, the tert-butyl ester
form of Nmoc-imidazole should be used. When the methyl ester form of
Nmoc-imidazole is used for caging primary amino groups in biomolecules
(e.g., the a-amino group in amino acids), the major product is not the
desired caged molecule, but rather a cyclic oxazolidinedione (Fig. 2). This
undesirable side reaction can be suppressed by replacing the methyl ester
in Nmoc-imidazole with the tert-butyl ester. The tert-butyl form of Nmoc-
imidazole readily couples with free amino groups to form stable carbamate
linkages that are unaffected by the acidic conditions required for removing
the tert-butyl group to generate the final caged molecule. Nmoc-Glu, a
caged version of the neurotransmitter glutamate, was prepared with tert-
butyl Nmoc-imidazole.
Syntheses
Synthetic procedures for methyl Nmoc-imidazole, Nmoc-DBHQ, and
the membrane-permeant acetoxymethyl (AM) ester of Nmoc-DBHQ have
been published. 8 Preparative procedures for tert-butyl-Nmoc-imidazole and
Nmoc-Glu are outlined in Fig. 3 and are described below.
4 APV, DL-2-amino-5-phosphonovaleric acid; CNB, ot-carboxy-2-nitrobenzyl; DBHQ, 2,5-

di(tert-butyl)hydroquinone; EGTA, ethylene glycol bis(2-aminoethyl)ether-N,N,N',N'-tet-
raacetic acid; GluR, glutamate receptor; HEPES, N-[2-hydroxyethyl]piperazine-N'-[2-eth-
anesulfonic acid]; NMDA, N-methyl-D-aspartate; NMR, nuclear magnetic resonance; NPE,
1-(2-nitrophenyl)ethyl; PIPES, piperazine-N,N'-bis(2-ethanesulfonic acid); SERCA,
sarcoplasmic/endoplasmic reticulum Ca2+-ATPase; TTX, tetrodotoxin.
5 G. A. Moore, D. J. McConkey, G. E. Kass, P. J. O'Brien, and S. Orrenius, FEBS Lett. 224,
331 (1987).
6 G. E. Kass, S. K. Duddy, G. A. Moore, and S. Orrenius, J. Biol. Chem. 264, 15192 (1989).
7 D. Thomas and M. R. Hanley, Methods Cell Biol. 41@9 65 (1994).
8 F. M. Rossi and J. P. Y. Kao, J. Biol. Chem. 272, 3266 (1997).
[24] Nmoc-CAGED
HI
Me02C- O,~N~z
O2N~ 0
Z-NH2 / ~ Desiredproduct:
Nmocattachedto aminogroupof Z
+
MethylNmoc-imidazolex ,
O2N
Actualproduct:
Anoxazolidinedione
FIG.2. Side-productformationin reactionof methylNmoc-imidazolewithamines.
tert-Butyl-2-acetoxy-2-(2-nitrophenyl)acetate (2)
Step a of Fig. 3:40 mmol (7.88 g) o-nitromandelic acid (1) 8 is refluxed
for 45 min in 60 ml acetic anhydride. After the reaction mixture cools to
room temperature, it is diluted with 50 ml each of tetrahydrofuran and
water. The mixture is stirred for 2 hr. The organic layer is separated from
the aqueous layer, diluted with 100 ml toluene, and extracted twice with
100 ml water. The organic phase is dried over MgSO4 and evaporated to
yield a brown oil that is directly used in step b of Fig. 3.
HO2C_ OH t-BuO2C .OH t-BuO2C- Ox~NCN

O2N.~ a,b t-BuO2C_O..~O c
O2N'~ CHa • O2N'~ P O2N~ O
I 2 3 4 tert-ButylNmoc-irnidazole
t'BuO2C ~ COzH
5 6 Nmoc-Glu
FIG.3. Schemeoutliningsynthesesof tert-butyl Nmoc-imidazoleand Nmoc-Glu.
Step b of Fig. 3:40 mmol (8.74 g) tert-butyl trichloroacetimidate is added
to a solution of the product of step a in 50 ml benzene. After 1 hr of stirring,
the reaction mixture is filtered to remove solids. The filtrate is concentrated
on a rotary evaporator and then chromatographed with ethyl acetate/hex-
ane (1 : 5) to yield 7.41 g (63%) of compound 2 as an oil. 1H NMR: 8.02
(d, J = 8.1 Hz, 1H), 7.65 (d, J = 3.9 Hz, 1H), 7.56-7.51 (m, 2H), 6.75 (s,
1H), 2.21 (s, 3H), 1.40 (s, 9H).
tert-Butyl-2-hydroxy-2-(2-nitrophenyl)acetate (3)
Step c of Fig. 3: To a stirred solution of compound 2 (6.63 g, 22.4 mmol)
in 50 ml methanol, 1.12 mmol Cs2CO3 (0.365 g) is added. After 1 hr, the
reaction mixture is diluted with 100 ml ethyl acetate and filtered through
a plug of 15 g silica gel. Concentration of the filtrate on a rotary evaporator
yields 5.36 g (94%) of compound 3 as an oil. 1H NMR: 7.96 (dd, J = 1.2,
6.8 Hz, 1H), 7.72-7.60 (m, 2H), 7.51-7.45 (m, 1H), 5.84 (d, J = 3.7 Hz,
1H), 3.67 (br d, J = 4.4 Hz, 1H), 1.38 (s, 9H).
tert-Butyl-Nmoc-imidazole [tert-butyl-2-(2-nitrophenyl)-2-
(oxycarbonylimidazole)acetate] (4)
Step d of Fig. 3: Compound 3 (8.00 mmol, 2.02 g) and carbonyldiimida-
zole (8.00 mmol, 1.30 g) are dissolved in 25 ml dichloromethane. After 1
hr of stirring, the reaction mixture is extracted three times with 25 ml water,
dried over MgSO4, concentrated by rotary evaporation, and chromato-
graphed with ethyl acetate/hexane (2 : 3) to yield 2.15 g (77%) of compound
4 as an oil. 1H NMR: 8.20 (s, 1H), 8.11 (dd, J = 1.3, 6.6 Hz, 1H), 7.73-7.63,
(m, 3H), 7.48 (s, 1H), 7.11 (s, 1H), 6.88 (s, 1H), 1.43 (s, 9H).
N-[tert-Butyl-2-[2-nitrophenyl]-2-oxycarbonyl acetate]-L-glutamic Acid,

Di-tert-butyl Ester (5)
Step e in Fig. 3: L-Glutamic acid di-tert-butyl ester hydrochloride (2.49
mmol, 0.736 g) and compound 4 (2.49 mmol, 0.866 g) are dissolved in 10
ml dichloromethane. After 0.35 ml triethylamine is added, the reaction
mixture is stirred for 40 hr. The residue obtained after the solvent evapora-
tion is chromatographed with ethyl acetate/hexane (1:5) to yield 0.858
(64%) of compound 5. 1H NMR: 8.00, (d, J = 7.82 Hz, 1H), 7.65-7.62 (m,
2H), 7.53-7.47 (m, 1H), 6.72 (s, 0.5H), 6.68 (s, 0.5H), 5.64-5.58 (m, 1H),
4.29-4.21 (m, 1H), 2.32-2.08 (m, 4H), 1.93-1.87 (m, 2H), 1.57 (s, 4.5H),
1.54 (s, 4.5H), 1.47 (s, 4.5H), 1.44 (s, 4.5H), 1.43 (s, 4.5H), 1.42 (s, 4.5H).
[24] Nmoc-CAGED
NO2 CO2R O
~ OH
Nmoe-DBHQ: R = H
Nmoc-DBHQ/AM: R = CH202CCH 3
FIG. 4. Structures of Nmoc-DBHQ and the corresponding acetoxymethyl(AM) ester.
Nmoc-Glu (N-[2-[2-nitrophenyl]-2-oxycarbonyl acetic acid]-L-glutamic

acid) (6)
Step f of Fig. 3: To a solution of 5 ml trifluoroacetic acid in 5 ml
dichloromethane is added 0.882 mmol (0.475 g) of compound 5. After 2
hr, the solution is concentrated by rotary evaporation. After azeotropic
removal of residual acid with several small volumes of benzene, the yield
of Nmoc-Glu (compound 6) is 0.32 g (98%), in the form of a colorless
glass. This material is dissolved directly in physiologic buffers for biological
experiments. High-resolution mass spectrometry (electron ionization): cal-
culated for C14H14NzO10[M +] m/z = 370.0649, observed 370.0635.
Nmoc-DBHQ: P h o t o m o d u l a t i o n of S R / E R Ca2÷-ATPase
(SERCA) Activity
The structures of N m o c - D B H Q and the corresponding A M ester are
shown in Fig. 4. Because the A M ester can be passively loaded into cells
by incubation, it is the most convenient form of the reagent for typical
cellular experiments. Loading is accomplished by incubating ceils with
1 0 - 4 0 / x M of the A M ester in aqueous medium for 30-60 min at room
temperature. In order to monitor the effects of D B H Q photorelease on
intracellular Ca 2+ dynamics, it is often desirable to load cells with a fluores-
cent indicator such as Fluo-3. For such applications, a few micromolar of
the A M ester of the indicator can also be present in the loading medium.
A small amount of the surfactant Pluronic F-127 (final concentration
<0.02% w/v) may be used to aid dispersion of the hydrophobic A M esters
into aqueous medium. 9
9j. p. y. Kao, Methods Cell Biol. 40, 155 (1994).

436 CAGZDCOMVOUNDS [24]
1.0
1 2 4
LL° 0.8
U_
® 0.6
==
0.4
~ 0.2
o.o
i i i i i i i i i
Time [1 div. = 60 s]
FIG. 5. Effect of DBHQ photorelease on [Ca2+]i. A REF52 rat embryo fibroblast was
loaded with Nmoc-DBHQ and Fluo-3 Ca 2÷ indicator by incubation with 40 and 5/zM of the
corresponding AM esters in DMEM at room temperature for 60 min. The experiment was
conducted in HBSS. Indicator fluorescence data are presented as relative changes in intensity,
AF(t)IFo, where AF(t) is the change of intensity at time t relative to the average initial intensity,
F0, which is measured with the cell at rest. Arrowheads and associated numbers indicate the
times and durations (sec), respectively, of photolysis light flashes. Corresponding dips in the
traces result from temporary interruptions of fluorescence data acquisition by an electrome-
chanical shutter to avoid potential damage to the fluorescence photomultiplier tube that might
be caused by leakage from the high-intensity photolysis flash.
Figure 5 illustrates the effects of transiently and reversibly inhibiting

SERCA pumps by photoreleasing D B H Q in rat fibroblasts loaded with
Nmoc-DBHQ. Changes in [Ca 2÷]i are monitored with the Ca 2÷ indicator
Fluo-3. Photolysis is performed as previously described. 8 A stable resting
[Ca2÷]i is the result of a dynamic balance between active pumping processes
that remove Ca 2÷ from the cytosol and passive leakage of Ca 2÷ into the
cytosol. Photoreleased DBHQ, which blocks the SERCA Ca 2÷ pumps and
thus disrupts the pump-leak balance, causes transient, dose-dependent
rises of [Ca 2÷] in the cytosol. The [Ca 2÷]i rise is transient because SERCA
inhibition by D B H Q is reversible and because D B H Q is an uncharged,
membrane-permeant molecule that easily diffuses out of the cell. As soon
as photoreleased D B H Q is cleared from the cell by diffusion into the
extracellular medium, the SERCA pumps are no longer inhibited, and the
resting [Ca 2÷]i becomes reestablished.
Because D B H Q is fairly water soluble and is membrane permeant, one
might wonder whether delivery of D B H Q by rapid-flow superfusion would
be as efficacious as by photorelease. A comparison of the two delivery
[241 Nmoc-CAOED
10%
<1
>
IIiillllt iii,t i i i i i l l l l l l [llll ~ l l l l l l l , i ] l l l l ]
0 50 100 150 200 250 300 350 400 450
Time [s]
FIG. 6. Comparison of the kinetics of changes in [Ca2÷]i in response to DBHQ delivered

by photorelease and by superfusion. (Upper trace) Response of REF52 cells to 1-see photore-
lease of DBHQ (marked by upward arrowhead). Loading and experimental conditions are
exactly as in Fig. 5. (Lower trace) Response of REF52 cell to 1-see pulse of 60/zM DBHQ
delivered by superfusion (marked by downward arrowhead). Cell was loaded with Fura-2
indicator by incubation with 1/zM of the AM ester in DMEM for 60 rain at room temperature.
Experiment was conducted in HBSS. Results are presented as changes of [Ca2÷]i (A[Ca2+]i)
relative to resting [Ca2÷]i (typically 70-90 nM, as determined by separate calibrations). Rise
and decay of the response to DBHQ photorelease are clearly much faster than for delivery
by superfusion. Time constants (ti/e) for rise and decay are 1.9 -+ 0.1 and 9.7 -4- 0.3 sec,
respectively, for photorelease and 17 _+ 5 and 82 +- 4 sec, respectively, for superfusion.
techniques is shown in Fig. 6, where cellular responses to 1-see DBHQ

application by photorelease and by superfusion are presented. Data in Fig.
6 show that both the rise and the decay of the response to photorelease
are much faster than for supeffusion delivery. More quantitatively, the rising
phases of the photorelease and superfusion responses are characterized
by time constants of 1.9 and 17 sec, respectively; the decay phases have
corresponding times constants of 9.7 and 82 sec, respectively. These results
demonstrate that even for molecules that are quite water soluble and mem-
brane permeant, photorelease still affords much better temporal and kinetic
control in cellular experimentation than conventional flow techniques.
Figure 7 shows that DBHQ photorelease can be used to probe dynamic
cellular phenomena, such as [Ca2÷]i oscillations. In REF52 cells (rat embryo
fibroblasts), activation of the inositol trisphosphate signaling pathway cou-
pled with depolarization-induced Ca 2÷ influx causes persistent and regular
o8
&
8
t,0
¢-
0~
¢-
m
o?,
O
LL
10 20 30 40 50
Time [rain]
b 1.8
1.6
"0
0
1.4
0_
"0
N
0
Z
1.0
i t i i J i i i i I i i i i I i i i i J
0 5 10 15 20
Cycle Number
FIG. 7. Effect of transient SERCA Ca 2+ pump inhibition on [Ca2+]i oscillations in rat

embryo fibroblasts. (a) [Ca 2÷]i oscillations induced in a REF52 cell by application of 500 nM
each of R8-vasopressin, which activates Ins(1,4,5)P3 production, and gramicidin D, an iono-
phore that depolarizes the cell by facilitating Na ÷ influx. Arrowheads and associated numbers
mark the times and durations (msec) of photolytic light flashes. Cell loading and experimental
conditions are exactly as in Fig. 5. (b) Data from (a) analyzed to reveal alterations in the
period between oscillation peaks caused by photoreleased pulses of DBHQ. Periods are
normalized to the unperturbed period length. Arrowheads mark the oscillation cycles during
which photorelease occurred; the corresponding numbers indicate the duration of the light
flash used.
[24] Nmoc-CAGED
oscillations in [Ca2+]i.l° The mechanism underlying these oscillations re-

quires positive cooperativity between [Ca 2+]i rises and Ins(1,4,5)P3 produc-
tion. 11'12The period between oscillation spikes is dependent on the ability
of intracellular Ca z+ stores to actively sequester Ca z+ from the c y t o s o l J 3
Nmoc-DBHQ makes it possible to study the effect of attenuating the ability
of intracellular Ca 2+ stores to sequester Ca 2÷. The behavior of [Ca2+]i oscilla-
tions in response to photomodulation of SERCA activity is shown in Fig.
7a. The results are presented more quantitatively in Fig. 7b, which shows
that photoreleased D B H Q causes a dose-dependent lengthening of the
period of that oscillation cycle during which photorelease occurred. This
type of period lengthening is just what is expected from modeling studies. 13
Such an experiment thus provides a direct test of the role of SERCA pumps
in the [Ca 2+]i oscillation mechanism.
Nmoc-Glu: Caged Glutamate for Studying Cellular Neurophysiology

Nmoc-Glu is a new caged glutamate that photoreleases free glutamate
on the one-to-few millisecond time scale, at a quantum yield of 0.11. The
principal advantage of Nmoc-Glu over y-O-CNB-GIu, 14 a commonly used
caged glutamate with rapid photorelease kinetics, is resistance to spontane-
ous hydrolysis under physiological conditions. Hydrolytic instability is detri-
mental for two reasons. First, non-NMDA glutamate receptors (GluRs)
require more than i mM glutamate to activate fully and yet are significantly
desensitized by less than I0/xM glutamate. 15,16Second, micromolar levels
of glutamate are sufficient to activate NMDA GluRs. 17Thus, achieving full
activation of non-NMDA GluRs by photolysis of caged glutamate requires
the caged reagent concentration to approach 10 mM in the medium. Any
tendency to hydrolyze spontaneously would cause a caged glutamate to
liberate free glutamate gradually, even in the absence of light. Even if
free glutamate contamination is insignificant by the standards of analytical
chemistry, it may still be significant biologically--0.1% spontaneous hydro-
10 A. T. Harootunian, J. P. Y. Kao, and R. Y. Tsien, Cold Spring Harb. Symp. Quant. Biol.
53, 935 (1988).
11 A. T. Harootunian, J. P. Y. Kao, S. Paranjape, S. R. Adams, B. V. L. Potter, and R. Y.
Tsien, Cell Calcium 12, 153 (1991).
12 A. T. Harootunian, J. P. Y. Kao, S. Paranjape, and R. Y. Tsien, Science 251, 75 (1991).
13T. Meyer and L. Stryer, Proe. Natl. Acad. Sci. U.S.A. 85, 5051 (1988).
Acad. Sci. U.S.A. 9L 8752 (1994).
15 L. O. Trussell and G. D. Fischbach, Neuron 3, 209 (1989).
16 C.-M. Tang, M. Dichter, and M. Morad, Science 243, 1474 (1989).
17 M. L. Mayer and G. L. Westbrook, J. Physiol. 354, 29 (1984).
440 CAGEDCOMVOUNDS [241
lysis in 10 mM caged glutamate yields 10/xM free glutamate, a concentration
sufficient to desensitize non-NMDA GluRs and to activate NMDA GluRs.
Glutamate photorelease from Nmoc-Glu is represented in Fig. 8. An
important feature is that photolysis generates an intermediate species, the
carbamate of glutamate, which then decarboxylates to yield free glutamate
and carbon dioxide. The decarboxylation step is rate limiting under physio-
logical conditions; therefore, photorelease of free glutamate from Nmoc-
Glu is slower than from 3~-O-CNB-GIu. The decarboxylation step is also
accelerated at lower pH. The foregoing suggests that the actual rate of
free glutamate generation by photolysis can be increased in two ways: by
increasing the starting concentration of Nmoc-Glu in the medium and by
decreasing pH.
For photolysis, an argon ion laser (Coherent 190-5, Santa Clara, CA)
with an output of -400 mW in the 351- to 364-nm region is used. The laser
beam is steered into a Nikon Diaphot inverted microscope through the
epifluorescence port and is directed through the back aperture of the objec-
tive (Fluor ×40, N.A. 1.3, Nikon, Columbia, MD) by a 400-nm long-pass
dichroic mirror. A divergent fused silica lense with a focal length of -150
mm positioned 20 cm in front of the dichroic mirror allows imaging of the
laser beam into a -50-/zm-diameter spot in the image plane. The photolysis
spot can be further delimited in diameter by closing the diaphragm of the
microscope objective. Final intensity in the image plane is estimated at - 7 0
/xW/zm -2. Light pulse duration is controlled by a laser shutter (Unblitz
LS2, Vincent Associates, Rochester, NY). Electrophysiological measure-
ments are performed in the whole cell voltage-clamp mode with a patch-
clamp amplifier (Dagan 3900, Minneapolis, MN). Borosilicate glass elec-
trodes with resistances of 3-5 Mf~ are used for whole cell recordings.
Electrode series resistance is compensated 80-90% during whole cell re-
cordings. Signals are filtered at 2 kHz, sampled at 5 kHz, and collected
and analyzed with pClamp software (Axon Instruments, Foster City, CA).
Extracellular solutions contain (in mM) 150 NaCI, 3 KCI, 2 CaC12, 1 MgCI2,
and 10 HEPES (or PIPES for pH < 7), adjusted to pH 7.3 with NaOH.
UV
kdee b CO + H3N*'~CO2-
02N i_ ii 0 ~ •
CO2- -02C . 0 CO2- H+ CO 2-
Nmoe-Glu ON~ Carbamate L-Glutamate

intermediate
Fro. 8. Scheme outlining photorelease of glutamate from Nmoc-Glu.
[24] Nmoc-CAGED
0.0
g -0.5
S
-o
-t
~.-1.0
E
<
r-
-1 .s
L-
-2.0
[Na*]o = 25 mM
I = I i i i I I ~ i = = I i i h i
-2.5
10o 150 200 250 300
Time [ms]
FIG. 9. Glutamate photorelease from Nmoc-Glu evokes robust inward current in a hippo-
campal neuron. Whole cell current response of a hippocampal neuron TM to a 5-msec photore-
lease from 10 m M Nmoc-Glu, p H 6.3, at room temperature. To attenuate the peak current
amplitude, the extracellular Na + concentration was reduced to 25 m M and photolysis was
restricted to a 30-/zm-diameter spot over the soma. The 10-90% rise time of the response is
0.98 msec. The double bars above the trace delimit the duration of the U V light pulse. Pipette
and extracellular solution compositions were as given in the text.
Where extracellular [Na +] is reduced, Na + is replaced with choline. TTX

(1 tzM) and APV (100/.~M) are added to external solutions. The internal
pipette contains (in mM) 140 CsCI, 5 EGTA, 1 MgC12, 2 CaCI2, and 10
HEPES, titrated to pH 7.3 with CsOH.
An experiment where glutamate photorelease from Nmoc-Glu is used
to stimulate a whole cell voltage-clamped rat hippocampal neuron TM is
shown in Fig. 9. Pulse photorelease of glutamate evokes a robust inward
current with a 10-90% rise time of 0.98 msec. The current declines rapidly
as the GluR channels desensitize.
Figure 10 illustrates experiments that demonstrate the high chemical
stability of Nmoc-Glu. These experiments show the effect of applying
freshly prepared I mM solutions of two different caged glutamates, without
photolysis, to hippocampal neurons. Whereas y-O-CNB-Glu a9 evokes sig-
nificant inward current even in the absence of light, Nmoc-Glu is essentially
inert. By repeating such paired tests on fresh cells, it was observed that
18 C.-M. Tang, M. Margulis, Q.-Y. Shi, and A. Fielding, Neuron 13, 1385 (1994).
19 Purchased from Molecular Probes, Inc., Eugene, OR.
v-O-CNB-Glu Nmoc-Glu
-50
<
-100
-150
-200
-250
-300 i i i i i i i i i i ~ t ~ i i I i i ~ J = i
Time [1 div.= 100 ms]
FIG. 10. Prephotolysis activity of 7-O-CNB-Glu and Nmoc-Glu. Solutions of the caged
compounds were pressure delivered, in the absence of light, to each cell tested. The upper
trace in each pair of experiments depicts the baseline current. Arrowheads mark the times
of reagent delivery. Nmoc-Glu was inert, whereas 7-O-CNB-Glu definitely evoked an inward
current. These traces were collected within 30 min of preparing 1 mM solutions of the reagents
from solid samples. The solutions were kept on ice and in the dark until use. Pipette and
external solution compositions were as given in the text.
the prephotolysis activity of the 7-O-CNB-GIu solution increases with time

(the average evoked whole cell current increases approximately threefold,
from --120 to -330 pA, over the period from 30 to 90 min after the
7-O-CNB-Glu solution was prepared). The Nmoc-Glu solution, however,
remains inert, evoking whole cell inward currents of only - 1 0 pA. These
results suggest that the intrinsic lability of 7-O-CNB-GIu causes gradual
liberation and accumulation of free glutamate under physiological condi-
tions, even in the absence of photolysis. In contrast, Nmoc-Glu appears to
be stable in physiologic solution indefinitely.
In the absence of an ideal caged glutamate, deciding which available
reagent is most appropriate depends on the demands of a particular experi-
ment. For example, if the rapid gating kinetics of non-NMDA GIuR chan-
nels are of primary interest, then 3,-O-CNB-Glu, which exhibits very fast
photorelease, would be appropriate. However, Nmoc-Glu has utility in
applications where its principal advantages, low prephotolysis activity and
high hydrolytic stability, are most useful. Such applications include (1)
experiments where caged glutamate needs to be in physiologic solution for
extended periods of time, (2) experiments where background activation of
neurons must be minimized, (3) experiments where desensitization of non-
NMDA GluRs must be minimized, and (4) experiments where the photore-
leased free glutamate concentration must be high (i.e., in the millimolar
range), which requires that the caged glutamate be present initially at high
concentration. Thus, potential applications for Nmoc-Glu include quantita-
[25] TRANSIENT KINETIC TECHNIQUES 443
tive mapping of the spatial distribution of functional non-NMDA GluRs on

dendrites and in the use of photostimulation to analyze brain circuitry3°,21
In summary, the advantage of the Nmoc caging method is threefold:
(1) ease of preparation and use of the active caging reagent, Nmoc-imidaz-
ole; (2) stable caging of hydroxyl and amino groups in bioactive molecules;
and (3) moderately fast photorelease kinetics.22-24The illustrative examples
described in this article show that the Nmoc approach can have wide use
in developing new reagents for cellular physiological studies.
Acknowledgments
This work was supported by NIH Grant GM-46956 to J.P.Y.K.
20 E. M. Callaway and L. C. Katz, Proc. Natl. Acad. Sci. U.S.A. 90, 7661 (1993).
21 L. C. Katz and M. B. DaNa, J. Neurosci. Methods 54, 205 (1994).
22 The Nmoc-caged molecules described here photorelease on the millisecond time scale,
comparable to NPE-caged ATP and Ins(1,4,5)P3, which have time constants (tVe) of release
of 4-5 and 12 msec, respectively (Refs. 23 and 24).
23 j. W. Walker, P. R. Gordon, J. A. MeCray, and D. R. Trentham, J. Am. Chem. Soc. 110,
7170 (1988).
[25] D e v e l o p m e n t a n d A p p l i c a t i o n o f C a g e d L i g a n d s
for Neurotransmitter Receptors in Transient Kinetic
and Neuronal Circuit Mapping Studies
By GEORGE P. HESS and CHRISTOF GREWER
Neurotransmitter receptor-mediated reactions on the surface of some

1012 cells in the central, sympathetic, and peripheral mammalian nervous
systems play a central role in the way in which environmental information
is received, transmitted, transduced, encoded, and stored. 1'2 Malfunctions
of these reactions are implicated in a number of diseases (e.g., Huntington's
disease, Parkinson's disease, epilepsy) and mediate the effects of many
clinically important compounds (e.g., tranquilizers, antidepressants) and of
1 F. Crick, "The Astonishing Hypothesis: The Scientific Search for the Soul," Charles
Scribner & Sons, New York, 1994.
2 E. R. Kandel, J. H. Schwartz, and T. M. Jessell, "Essentials of Neural Science and Behavior,"
Appleton & Lange, Noiwalk, CT, 1995.

METHODS IN ENZYMOLOGY, VOL, 291 0076-6879198 $25.00
tive mapping of the spatial distribution of functional non-NMDA GluRs on

dendrites and in the use of photostimulation to analyze brain circuitry3°,21
In summary, the advantage of the Nmoc caging method is threefold:
(1) ease of preparation and use of the active caging reagent, Nmoc-imidaz-
ole; (2) stable caging of hydroxyl and amino groups in bioactive molecules;
and (3) moderately fast photorelease kinetics.22-24The illustrative examples
described in this article show that the Nmoc approach can have wide use
in developing new reagents for cellular physiological studies.
Acknowledgments
This work was supported by NIH Grant GM-46956 to J.P.Y.K.
20 E. M. Callaway and L. C. Katz, Proc. Natl. Acad. Sci. U.S.A. 90, 7661 (1993).
21 L. C. Katz and M. B. DaNa, J. Neurosci. Methods 54, 205 (1994).
22 The Nmoc-caged molecules described here photorelease on the millisecond time scale,
comparable to NPE-caged ATP and Ins(1,4,5)P3, which have time constants (tVe) of release
of 4-5 and 12 msec, respectively (Refs. 23 and 24).
23 j. W. Walker, P. R. Gordon, J. A. MeCray, and D. R. Trentham, J. Am. Chem. Soc. 110,
7170 (1988).
[25] D e v e l o p m e n t a n d A p p l i c a t i o n o f C a g e d L i g a n d s
for Neurotransmitter Receptors in Transient Kinetic
and Neuronal Circuit Mapping Studies
By GEORGE P. HESS and CHRISTOF GREWER
Neurotransmitter receptor-mediated reactions on the surface of some

1012 cells in the central, sympathetic, and peripheral mammalian nervous
systems play a central role in the way in which environmental information
is received, transmitted, transduced, encoded, and stored. 1'2 Malfunctions
of these reactions are implicated in a number of diseases (e.g., Huntington's
disease, Parkinson's disease, epilepsy) and mediate the effects of many
clinically important compounds (e.g., tranquilizers, antidepressants) and of
1 F. Crick, "The Astonishing Hypothesis: The Scientific Search for the Soul," Charles
Scribner & Sons, New York, 1994.
2 E. R. Kandel, J. H. Schwartz, and T. M. Jessell, "Essentials of Neural Science and Behavior,"
Appleton & Lange, Noiwalk, CT, 1995.

METHODS IN ENZYMOLOGY, VOL, 291 0076-6879198 $25.00
abused drugs (e.g., cocaine). 2'3 Transient kinetic techniques suitable for
investigating the mechanism of neurotransmitter receptor-mediated reac-
tions on cell surfaces have been developed 4,5 and are the subject of this
article.
Here we describe the instrumentation and theory involved in transient
kinetic investigations of neurotransmitter receptor-mediated reactions on
the surface of nervous system cells and in the membranes of Xenopus
laevis oocytes in which many receptors have been expressed. 6 Photolabile,
biologically inert precursors of neurotransmitters (caged neurotransmitters)
that can be photolyzed to the neurotransmitters in the microsecond time
region form an integral part of the transient kinetic techniques. The design,
synthesis, and characterization of caged excitatory 7-9 and inhibitory 1°-12
neurotransmitters that have been used in transient kinetic investigations
are reviewed by Gee et al. 13
Six structurally related excitatory and inhibitory neurotransmitter recep-
tors, and many isoforms, have been identified using recombinant DNA
technology. 14-a6 These cell surface proteins, on binding chemical signals
(neurotransmitters) released from an adjacent cell, transiently form open
transmembrane channels (Fig. 1A). Excitatory receptors (for acetylcholine
and glutamate) form cation-specific channels and inhibitory receptors [for
y-aminobutyric acid (GABA) and glycine] anion-specific ones. If the sign
and amplitude of the resulting transmembrane voltage change reach a
critical threshold value of about - 4 0 mV, the cell will transmit an electrical
signal along its long projection, an axon. When this signal reaches the end
3 A. G. Gilman, T. W. Randal, A. S. Neis, and P. Taylor, "The Pharmacological Basis of

Therapeutics," 8th Ed. Macmillan, New York, 1990.
4 N. Matsubara, A. P. Billington, and G. P. Hess, Biochemistry 31, 5507 (1992).
5 G. P. Hess, L. Niu, and R. Wieboldt, Ann. N.Y. Acad. Sci. 757, 23 (1995).
6 D. Langosch, B. Laube, N. Rundstrom, V. Schmieden, J. Borman, and H. Betz, EMBO J.
13, 4223 (1994).
W. W. Webb, J. Marque, W. Denk, 3. A. McCray, and G. P. Hess, Biochemistry 29, 49 (1989).
Acad. Sci. U.S.A. 91, 8752 (1994).
9 L. Niu, K. R. Gee, K. Schaper, and G. P. Hess, Biochemistry 35, 2030 (1996).
10D. Ramesh, R. Wieboldt, L. Niu, B. K. Carpenter, and G. P. Hess, Proc. Natl. Acad. Sci.
U.S.A. 90, 11074 (1993).
11 K. R. Gee, R. Wieboldt, and G. P. Hess, J. Am. Chem. Soc. 116~ 8366 (1994).
12 L. Niu, R. Wieboldt, D. Ramesh, B. K. Carpenter, and G. P. Hess, Biochemistry 35,
8136 (1996).
13 K. R. Gee, B. K. Carpenter, and G. P. Hess, Methods Enzymol. 291, [2] (1998) (this volume).
14 H. Betz, Neuron 5, 383 (1990).
15 R. M. Stroud, M. P. McCarthy, and M. Shuster, Biochemistry 29, 11009 (1990).
16 G. Mandel and D. McKinnon, Annu. Rev. Neurosci. 16, 323 (1993).
of an axon, neurotransmitter is released near receptors on the surface of

an adjacent cell2 (Fig. 1A).
The transient kinetic techniques described here were developed to mea-
sure the rate and equilibrium constants that determine the concentration
of open receptor channels as a function of neurotransmitter concentration
and time in the physiologically relevant time region of microseconds to a
few milliseconds. The concentration of open receptor channels, and their
conductance and ion specificity, determine the electrical signals in the mem-
brane of the cells that trigger signal transmission to another cell. 17'18A9
The minimum reaction scheme for a receptor-mediated reaction is
shown in Fig. lB. It is based on experiments with the excitatory nicotinic
acetylcholine receptor 27a,z8 about which most information is available. It is
the only neurotransmitter receptor about which we have three-dimensional
structural information.17'18The conductance, lifetime, and ion specificity of
the open receptor channel are conveniently determined by the single-chan-
nel current-recording technique, a9aThe new techniques described here were
developed to solve another problem, To account for the integration of
excitatory and inhibitory signals arriving at one cell and the resulting trans-
membrane voltage changes, one needs to know the receptor-controlled rate
at which inorganic cations and anions cross the membrane through the
receptor channels. The rates of the cation and anion fluxes across a semiper-
meable membrane are related to the transmembrane voltage.2°,zl They
depend not only on the conductance of the open receptor channels, but
also on their concentration in the cell membrane. 2z To determine the con-
centration of open channels as a function of neurotransmitter concentration
and time, transient kinetic techniques are required. One needs to determine
(i) the reaction path starting with free receptor and neurotransmitter and
leading to the formation of neurotransmitter: receptor complexes and open
receptor channels, (ii) the reaction path leading to desensitized (transiently
inactivated) receptor forms, and (iii) the rate and equilibrium constants
associated with individual steps in the reaction. The reactions steps outlined
in Fig. 1B, like those of many enzyme-catalyzed reactions, occur in the
17 M. Planck, Ann. Physik u. Chem 40, 561 (1890).

18 D. E. Goldman, J. Gen. PhysioL 27, 37 (1943).
19 G. P. Hess, H.-A. Kolb, P. L~iuger, E. Schoffeniels, and W. Schwarze, Proc. Natl. Acad.
Sci. U.S.A. 81, 5281 (1984).
19a E. Neher and B. Sakmann, Nature 260, 799 (1976).
20 N. Unwin, J. Mol. BioL 229, 1101 (1995).
21 N. Unwin, Nature 373, 37 (1995).
2za C. L. Weill, M. G. McNamee, and A. Karlin, Biochem. Biophys. Res. Commun. 61, 997
(1974).
ZZbj. A. Reynolds and A. Karlin, Biochemistry 17, 2035 (1978).
446 CAOEDCOMeOUNDS [25]
A ASimple
r ~ Animal
Senson/Cell
Neurotransmitte - -/~ll ~ - ~ Synapse
Receptor~ ~..N~ Axon
..sc..c.
B
kop _
A + L - - , K I " AL . K I , , AL2x__.~k~IAL,'
cl (open channel)
k21 ~k12 k43~k34
.I{2 ,~
IL IL2
FIG. l. (A) Two sensory cells are located on the surface of an animal. Activation of the
sensory cell on the left leads to release of inhibitory neurotransmitters (0). Activation of the
sensory cell on the right leads to release of excitatory neurotransmitters (O). Inhibitory
neurotransmitters (e.g., glycine and GABA) on binding to their specific receptors on the cell
surface of neurons located inside the animal, transiently open anion-conducting transmem-
brane channels. Excitatory neurotransmitters (e.g., acetylcholine and glutamate) transiently
open cation-conducting transmembrane channels. If the transmembrane voltage change re-
sulting from the activation of excitatory and inhibitory neurotransmitters is of appropriate
sign and amplitude ( - - 4 0 mV), an electrical signal is elicited that travels along the axon
of the neuron and results in the release of neurotransmitters adjacent to receptors in other
cells, thus repeating the signal transmission process. (B) A minimum mechanism for the
nicotinic acetylcholine receptor from the Electrophorus electricus electroplax and BC3H1
muscle cells. The reaction scheme is consistent with the available structural ay'ls'ls~'~sb and
kinetic5~8,52 information. The reaction is initiated when a neuron secretes neurotransmitter
L, which then binds to the active, nondesensitized receptor form, A, in the membrane of an
adjacent cell. The chemical reaction contains several ligand-binding steps, characterized by
the dissociation constant K1, and leads to receptor-ligand complexes, where the subscript
represents the number of ligand molecules bound. The binding of two ligand molecules before
the acetylcholine receptor channel opens was first suggested by Katz and Thesleff. 2s Several
transitions in protein conformation result: One change in the submillisecond time region leads
to the formation of a transmembrane channel (AL2), characterized by the rate constants for
channel opening (kop) and closing (kc~). In the presence of neurotransmitter, receptor proteins
are also rapidly (15-200 msec) converted to inactive forms, L with altered ligand-binding
properties, characterized by the dissociation constant K2. Katz and Thesleff2s observed a slow
(seconds) desensitization process in the muscle acetylcholine receptor. It is now known that
a slow desensitization process is associated with a minor receptor form in BC3H1 cells that
contain muscle acetylcholine receptors. 33 It is assumed that a single Kl value pertains to
micro- to millisecond time domain and must, therefore, be investigated

with techniques that have a micro- to millisecond time resolution. Such
techniques exist for investigations of reactions mediated by soluble pro-
teins3 3-27 The function of neurotransmitter receptors, however, must be
studied in intact cells or vesicles with the receptor protein embedded in a
membrane separating two solutions of different ionic composition. Kinetic
techniques for studying cell surface receptors in the micro- to millisecond
time region and over a wide range of reactant concentrations only recently
became available. 4'5'7'53 We shall give a few examples of the application of
the new techniques using caged neurotransmitters in studies of (i) the
channel-opening process of neurotransmitter receptors, (ii) drug-receptor
interactions, (iii) integration by a cell of inorganic cation and anion trans-
membrane fluxes mediated by excitatory and inhibitory receptors, respec-
tively, and (iv) locating and elucidating the neurotransmitters and receptors
that control signal transmission in known circuits of neurons controlling
measurable responses.
Rapid Chemical Reaction Techniques for Studying Reactions

on Cell Surfaces
Cell-Flow Technique
Rapid receptor desensitization (the reversible inactivation of a receptor
by its specific neurotransmitter) provided the impetus for developing tech-
niques for rapidly equilibrating receptors with the neurotransmitter in solu-
tions flowing over cells. Desensitization of neurotransmitter receptors was
first discovered in studies with the nicotinic acetylcholine receptor and
appeared to be slow, occurring in the second time region. 28 When it was
23 M. Eigen, in "Fast Reactions and Primary Processes in Chemical Kinetics: Nobel Symp.
5" (S. Claesson, ed.), p. 333. Interscience, New York, 1967.
24 G. G. Hammes, "Enzyme Catalysis and Regulation." Academic Press, New York, 1982.
25 A. Fersht, "Enzyme Structure and Mechanism." Freeman, New York, 1985.
26 K. A. Johnson, in "The Enzymes," Vol. 20, p. I. Academic Press, New York, 1993.
27 H. Gutfreund, "Kinetics for the Life Sciences: Receptors, Transmitters and Catalysts."
Cambridge University Press, UK, 1995.
27a G. P. Hess, D. J. Cash, and H. Aoshima, Annu. Rev. Biophys. Bioeng. 12, 443 (1983).
28 B. Katz and S. Thesleff, J. Physiol, (Lond.) 138, 63 (1957).
the dissociation of L from the AL and AL2 forms. The direct conversion of A, or AL2 to the
inactive I forms is neither required nor excluded by data available so far. 27a,52The equations
used to evaluate the rate and equilibrium constants of the reaction scheme using rapid chemical
reaction techniques are listed in Table I.
TABLE I
EQUATIONS FOR REACTION SCHEME SHOWN IN FIG. 1B
Equation
Equation number
__ A-'-T2 L2
(AL2)° A + A L + AL2 + A"~2 (L + Ka)2 qb + L 2 = Po (I-A)
Ig = IMRM (ALz)o (I-B)
[/MR M (IA)-I _ 111/2 = (I)1/2 _}_ (I)1/2 K1 [L1] -1 (I-C)
Eq. (I-A): (AL2)o represents the fraction of receptor molecules in the open channel form.
1 = kop/kc~ is the channel-opening equilibrium constant, L represents the molar concentra-
tion of activating ligand, and Po is the conditional probability, determined in single-channel
current recordings, that the receptor is in the open-channel form. 3° All other symbols have
been defined (Fig. 1B).
Eq. (I-B): IA is the current due to open receptor channels in the cell membrane corrected
for receptor desensitization. IM is the current due to 1 mol of open receptor channels and
RM represents the number of moles of receptors in the cell membrane.
Eq. (I-C): A linear version of Eq. (I-B) 27a
(lobs)t- (lobs)t= = [(lob~)t=0 -- (lobs)t=]e-" (II-A)
L k43 q- 2 1(2 kz, (L k34 4- 2 K1 k~z) L ] (II-B)

a=~ -(-£7-2-~z)~ + L Z ( I + @) + 2 K 1 L * + KId~J
k34 (I) L 2 ] (II-C)
a (L+K1) 2qb+L 2 J
Eq. (II-A): a represents the rate coefficient for receptor desensitization obtained from the
falling phase of the current in cell-flow experiments. 33 lobs represents the current during
the falling phase. The subscripts t and t = 0 refer to the time of measurement, and t = o:
to the time when an equilibrium between active and desensitized forms has been reached.
Eq. (If-C): When k~4 (Fig. 1B) is the dominant rate constant, a simplified equation is obtained 33
IA/I'A = 1 + Io/K I (Ill-A)
1 FA FAC FAC2 F-A"£2 (III-A')

gi (gI)l q- ~ + ~ -I (gl)4
I,, 1o (2K1L + K~)~, (m-B)

--=1+
I'A KI (L + K1)zdP + L 2
IA = 1 + Io (AL----2)o (III-C)
t;, ~ii
IA = 1 + Io Iio (III-D)
Ik ~ q- Kn
TABLE I (Continued)
Equation
Equation number
1A= I +__ -----

Iio -Io- - Ilo
- -Io- Ilo (K~ + lo~ (III-E)
I'A KI + KtI + KI Ktt = I + KI + KII \ KI ]
Eq. (III-A): The dissociation constant of the inhibitor from the nondesensitized receptor can
be determined by both cell-flow and photolysis measurements. In order to simplify the
equations, we use a ratio method, IA/I'A, where IA and I~, represent the current maxima
corrected for receptor desensitization in the absence and presence of inhibitor, respectively.
Io represents the inhibitor concentration and KI the observed dissociation constant
Eq. (III-A'): The relationship between the observed inhibitor dissociation constant K1 and
the inhibitor dissociation constant for A, AL, AL2, and ALz receptor forms. FA, FAL,
FAL2, and F~-E2 represent the fraction of receptors in forms A, A, AL2, and AL2
Eq. (III-B): For a competitive inhibitor 1/KI is multiplied by the fraction of receptor molecules
in the A and the A L form
Eq. (III-C): An inhibitor binding only to the open-channel form
Eq. (III-D): Two inhibitors Io and IIo binding to the same receptor site
Eq. (III-E): Two inhibitors Io and Iio binding to two different receptor sites
It = Imax[1 -- exp(-kobd)] (IV-A)
( L )2 (IV-B)
kobs = kcl + kop
• K, +" { L V (fV-C)
/obs = ~Cl K ~ ' ~ o J(op ~L---"-~I )
( L )2 KI (IV-D)
k°bs = kcl + k°P ~ KI+lo
Eq. (IV-A): In the laser-pulse photolysis experiments with BC3H1 cells containing nicotinic
acetylcholine receptors, the current rise time was observed to follow a single exponential
rate law over 85% of the reaction. 4 It represents the observed current at time t and/max
represents the maximum current
Eq. (IV-B): The relationship between the observed rate constant for the current rise kob~and
kop, kcl, K1 of the reaction scheme (Fig. 1B)
Eq. (IV-C): kobs in the presence of an inhibitor that binds only to the open-channel form of
the receptor
Eq. (IV-D): kobs in the presence of an inhibitor that binds only to the closed-channel form
of the receptor. If the inhibitor binds both to the open- and closed-channel forms a combina-
tion of Eqs. (IV-C) and (IV-D) is obtained
450 CAGED COMPOUNDS [9..5]
2.5
<
I-
z1.1.1
rr 2.0 t
nr' 1.5
0
._1
1.0
o,
U.I
0.5
"1-
0.0
-250 1650 3550 5450 7350 9250
TIME (msec)
FIG. 2. Cell-flow measurement with a BC3H1 muscle cell containing nicotinic acetylcholine
receptors, pH 7.4, 23°, - 6 0 mV transmembrane potential. A solution of 200/zM acetylcholine
emerged from the cell-flow device (Fig. 3A) at a rate of 1 cm sec-L The thick solid line
represents the observed current, and the thinner line represents the calculated current cor-
rected for receptor desensitization using Eq. (1). [From G. P. Hess, J. B. Udgaonkar, and
W. L. Olbricht, Annu. Rev. Biophys. Biophys. Chem. 16, 507 (1987)].
discovered, using rapid reaction techniques z9 and subsequently single-chan-

nel recording measurements, 3° that this receptor desensitizes in the millisec-
ond time region, an elegant device for flowing solutions containing neuro-
transmitter over a single cell was developed. 31The current due to opening of
receptor channels can then be recorded by the whole cell current recording
technique. 32 Equilibration of ligands in the flowing solutions with cell sur-
face receptors, however, can be slow compared to receptor desensitization,
and the amplitude of the current recorded must be corrected to take into
account desensitization that occurs during the equilibration of the neuro-
transmitter in the flowing solutions with the cell surface receptors. 33'34 The
thick solid line in Fig. 2 is the observed whole cell current recorded when
a BC3H1 muscle cell containing acetylcholine receptors was equilibrated
with 200/zM acetylcholine in a solution flowing over the cell. The current
first rises due to cations moving through receptor channels that open in
29 G. P. Hess, D. J. Cash, and H. Aoshima, Nature 282, 329 (1979).

30 B. Sakmann, J. Patlak, and E. Neher, Nature 286, 71 (1980).
31 O. A. Krishtal and V. I. Pidoplichko, Neuroscience 5, 2325 (1980).
32 O. Hamill, A. Marty, E. Neher, B. Sakmann, and F. J. Sigworth, Pfluegers Arch. 391,
85 (1981).
33 j. B. Udgaonkar and G. P. Hess, Proc. Natl. Acad. Sci. U.S.A. 84, 8758 (1987).
34 O. P. Hess, J. B. Udgaonkar, and W. L. Olbricht, Annu. Rev. Biophys. Biophys. Chem. 16,
507 (1987).
response to the neurotransmitter. The falling phase of the current is due

to receptor desensitization in which the receptor channels close. The thin
line in Fig. 2, which reaches an amplitude about twice that reached in the
recorded current trace, is a calculated line [Eq. (1)] and is the current after
correction for receptor desensitization that occurs during the equilibration
of the receptors with neurotransmitters in the flowing solution. 33'34
Equation (1) was derived 33'34to calculate the current IA, which is associ-
ated with the active nondesensitized receptor form A (Fig. 2). To obtain
the value of IA from measurements of the observed current lobs, the current
time course is divided into constant (1-5 msec) time intervals to take into
account the equilibration time of small segments of the cell surface with
ligand. The current is then corrected for the desensitization occurring during
each time interval At. After n constant time intervals ( n A t = tn), during
each of which the current (lobs)at is measured, the corrected current is
given by 33'34
(IA)tn = ( e"at -- 1) ~ (/obs)a,i + (Iobs)A/n (1)

i=1
where (Iobs)ati is the observed current during the ith time interval and tn is
equal to or greater than the current rise time, the time it takes the current
to reach a maximum value. Under conditions of laminar flow of the neuro-
transmitter solution over the cell, the value of IA was found to be indepen-
dent of the solution velocities used and could be determined with good
p r e c i s i o n (+_10%). 33 Equation (1) is based on the theory of the flow of
solutions over spherical objects. 35,36BC3H1 cells are nearly spherical when
suspended by the current-recording electrode in solutions bathing the cell.
In experiments with central nervous system neurons, which are firmly
attached to their stratum, it has been possible to obtain spherical, tightly
sealed receptor-containing membrane vesicles of about 15 tzm diameter
attached to the recording electrode 37a (which can be suspended in the fluid
stream emerging from the flow device).
A diagram of a cell-flow device is shown in the lower left-hand
corner of Fig. 3A. Solutions flow at rates of less than 2 cm sec -1 from a
150-250-tzm-diameter porthole of the flow device over a single cell32'34 (or
vesicle of 10-15/xm diameter. The cell (or vesicle) is suspended from a
35 V. G. Landau and E. M. Lifshitz, "Fluid Mechanics," p. 219. Pergamon, Oxford, 1959.

36 A. G. Levich, "Physicochemical Hydrodynamics." Prentice-Hall, Engelwood Cliffs, N J,
1962.
37 K. M. Walstrom and G. P. Hess, Biochemistry 33, 7718 (1994).
37aW. Sather, S. Dieudonne, J. F. MacDonald, and P. Ascher, J. Physiol. (London) 450,
643 (1992).
452 CAGZD COMPOUNDS [25]
/
3006 -C
A 14-200Fm--*loptical fiber g6. ° ..~--z~ - 2500
40 • " 200°
hv ~ " '..
400 <~ .:.f '\,. . ~"
El/ o.o ............. " ,oo6

,,o%_ -- Bc3.,ce., /
] I and electrode -2.0n . . . . ~, . t
...] . . -1 0 1 2 3 4 104 204 -0.i f .o 03 o12 o.~
)application Time (ms) L21(L+KI )2
--'~ flow device
FIG. 3. (A) BC3H1 cell carrying nicotinic acetylcholine receptors, of - 2 0 / z m diameter,

attached to an electrode for whole-cell current recording was equilibrated with caged carbamo-
ylcholine. 4 The side view of the cell-flow device 31 used to equilibrate the cell surface with
ligands is shown on the left. A Candela SLL500 dye laser is used. Rhodamine 640 or sulforho-
damine 640 laser dye, together with a second harmonic generator, produces wavelengths of
320 and 328 nm, respectively. The laser beam is introduced from an optical fiber of 200/xm
diameter. The fiber is adjusted to be - 4 0 0 ~m away from the cell so that the area illuminated
around the cell has a diameter of 300-400/xm. The energy of the laser pulse emerging from
the fiber is -500/zJ, and the pulse length is 600 nsec. By projecting visible light through the
optical fiber, the cell is illuminated and the fiber properly positioned. The wavelength of 328
nm was chosen to avoid cell damage at lower wavelengths and too low a product yield at
higher wavelengths. Current is recorded in the whole-cell configuration.3z Data were low-
pass filtered (Krohn-Hite 3322) with a 1- to 10-kHz cutoff frequency ( - 3 dB point) and then
digitized at a 2- to 20-kHz sampling frequency using a PDP 11/23 minicomputer; data were
then transferred to a Convex C210 computer where the constants for the rising and decaying
phases of the whole-cell current were analyzed using the data analysis program PLOT. (B)
Whole-cell current induced by 200/.cM released carbamoylcholine at pH 7.4, 22-23 °, and - 6 0
m V : The points represent digitized current data [Table I, Eqs. (IV-A) and (IV-B)]. An
observed first-order rate constant of 2140 sec -1 was evaluated from the rise-time. (C) Determi-
nation of kop, kcl, and Ka for the opening of acetylcholine receptor-channels in BC3H1 cells
at pH 7.4, 23°, and - 6 0 mV. 4 The values of kobs determined from experiments shown in (B)
are plotted according to kob~ = kcl + kop L2/(L + K1)2 [Table I, Eq. (IV-A) and (IV-B)].
The values of kop, kd, and K1 are given in Table IV.
current-recording electrode 32 and is about 50-100/xm away from the port-

hole. Because the orientation of the porthole of the flow device with respect
to the cell is critical, we modified the device 38 first designed by Krishtal
and Pidoplichko 31 in which the porthole is stationary and mixing of solutions
occurs before leaving the porthole. A flow device that enables one to flow
solutions of different composition over the cell has been described in detail
by Niu e t al. 38 As many as 20 measurements have been made with the same
BC3H1 muscle cell. At the flow rates used the current must be corrected
for desensitization (Fig. 2). The time resolution of the cell-flow techniques
38 L. Niu, C. Grewer, and G. P. Hess, Tech. Protein Chem. VII, 139 (1996).
[25] TRANSIENTKINETICTECHNIQUES 453
is about 10 msec. 33 T o obtain a better time resolution, the flow rate of the
neurotransmitter solution could be increased. However, if one increases
the flow rate of the solutions over the cell, turbulent flow and mixing occurs
between the neurotransmitter solution and the buffer solutions surrounding
the cell. Consequently, the concentration of neurotransmitter in the solution
equilibrating with the cell surface receptors is not known.
L a s e r - P u l s e Photolysis T e c h n i q u e
In contrast to the cell-flow technique, the time resolution of a laser-

pulse photolysis technique is two to three orders of magnitude better. For
this method, biologically inert photolabile precursors of neurotransmitters
(caged neurotransmitters) suitable for investigations, in the microsecond
time region, of the excitatory acetylcholine and glutamate receptors and
the inhibitory ~/-aminobutyric acid and glycine receptors have been devel-
oped. 3'7-12'39 A cell attached to a current-recording electrode 31 is preequili-
brated with a caged neurotransmitter. At zero time a pulse of laser light
photolyzes the caged c o m p o u n d in the microsecond time region. The liber-
ated neurotransmitter binds to receptors on the cell surface and initiates
the formation of t r a n s m e m b r a n e channels; the resulting current is measured
(Fig. 3B). The apparatus used is illustrated in Fig. 3A. The time resolution
of the m e t h o d is governed by the rate with which the neurotransmitter is
liberated by a light pulse. Photolysis rates with ti/2 values between 1 and
45 tzsec have been obtained with the compounds listed in Tables II and
III. Caged neurotransmitters that are not suitable for transient kinetic
m e a s u r e m e n t s because they either photolyze too slowly, in the millisecond
time region, or are not biologically inert 39 have also been synthesized 4°-43
but are not listed on Tables II and III. Several caged phenylephrine deriva-
tives e x i s t 44-46 suitable for investigations of the oq-adrenergic receptor.
These are also not reviewed here.
39K. R. Gee, L. Niu, K. Schaper, and G. P. Hess, J. Org. Chem. 60, 4260 (1995).
40M. Wilcox, R. W. Viola, K. W. Johnson, A. P. BiUington, B. K. Carpenter, J. A. McCray,
A. P. Guzikowski, and G. P. Hess, J. Org. Chem. 55, 1585 (1990).
41A. P. Billington, K. M. Walstrom, D. Ramesh, A. P. Guzikowski, B. K. Carpenter, and
G. P. Hess, Biochemistry 31, 5500 (1992).
42j. E. T. Corrie, A. DeSantis, Y. Katayama, K. Khodakhah, J. B. Messenger, D. C. Ogden,
and D. R. Trentham, J. PhysioL (Lond.) 465, 1 (1993).
43R. Wieboldt, D. Ramesh, B. K. Carpenter, and G. P. Hess, Biochemistry 33, 1526 (1994).
44S. M. Muralidharan, G. M. Mayer, W. B. Boyle, and J. M. Nerbonne, Proc. Natl. Acad.
Sci. U.S.A. 90, 5199 (1993).
45j. W. Walker, H. Martin, F. R. Schmitt, and R. J. Barsotti, Biochemistry 32, 1338 (1993).
46S. Muralidharan and J. M. Nerbonne, J. Photochem. Photobiol. B 27, 123 (1995).
In kinetic measurements the concentration of neurotransmitter gener-

ated by photolysis of a caged precursor is measured by using the cell-flow
technique. 32,33Before photolysis, a known concentration of free neurotrans-
mitter is applied to the cell and the amplitude of the resulting current,
together with a dose-response curve (e.g., Fig. 4), is used to calibrate the
concentration of neurotransmitter generated in the photolysis experiment.
Cell-flow experiments are also used at the end of the experiment to check
if the laser pulse has damaged the cell or the receptors. The cell-flow device
(Fig. 3A) 38 (see Fig. 1 in Niu et al. 38 for details) is used both to remove the
neurotransmitter from the culture dish and to add caged neurotransmitter.
The results obtained by laser-pulse photolysis and the single-channel cur-
rent-recording technique 19a can be compared only at low neurotransmitter
concentrations at which receptor desensitization is relatively slow. As can
be seen in Fig. 4 there is good agreement among the results obtained by
laser-pulse photolysis of caged neurotransmitters, the single channel cur-
rent-recording technique, 19a and the cell-flow technique. 33
5
4
if-3
=.=..
[]
1
0 i
10 -2 10 -1 10 ° 101
[Carbamoylcholine](mM)
F]6. 4. Concentration dependence of the current amplitude corrected for receptor desensiti-
zation, IA. BC3HI muscle cells, pH 7.4, 22-23 °, and - 6 0 mV transmembrane potential. Data
are from J. B. Udgaonkar and G. P. Hess, Proc. Natl. Acad. Sci. U.S.A. 84, 8758 (1987).
(O, single-channel current recordings; ,t, cell-flow measurements) and N. Matsubara, A. P.
Billington, and G. P. Hess, Biochemistry 31, 5507 (1992) ([~, laser-pulse photolysis). The line
indicating the concentration of open receptor channels was calculated from the constants
pertaining to the channel-opening process determined in laser-pulse photolysis experiments
(Table IV).
[251 TRANSIENT KINETIC TECHNIQUES 455
Photochemical Properties of Caged Neurotransmitters

Attempts to make chemical kinetic studies of acetylcholine receptors
on cell surfaces were initiated in innovative studies by Bartels et aL,46a who
synthesized the photoisomerizable 3,3'-bis(trimethyl aminomethyl)azoben-
zene (Bis Q) and reported that the trans, but not the c/s, form was an
activating ligand for the acetylcholine receptor. Lester then used photo-
isomerization of Bis Q to study the kinetics of the opening of the acetylcho-
line receptor channel. 46b In our development of a rapid reaction method
using light activation, we foresaw the need to measure adverse effects of
caged neurotransmitters and their photolysis products; it became one of
the strategic considerations. Using rapid mixing techniques we had devel-
oped for use with vesicles, we found that "inactive" cis-Bis Q causes recep-
tor desensitization and trans-Bis Q becomes a receptor inhibitor even at
low concentrations. 47 We, therefore, began to consider the synthesis of
photolabile inert precursors of neurotransmitters.
Photocleavable protecting groups for biologically important compounds
have many uses in cell biology. This is particularly true of cases where
access of a compound to its reaction partner is slow but the induced reaction
is fast. 48-53 The most frequently used photocleavable protecting group is
the 2-nitrobenzyl group, with various substituents. The photochemically
induced transfer of oxygen in 2-nitrobenzyl derivatives by which the ortho
CH group becomes C-OH and the NO2 group NO on irradiation was first
described by De Mayo 54in 1960. Barltrop et aL55 then protected the carboxy
group of glycine and leucine using a 2-nitrobenzyl derivative. Subsequently,
Patchornik eta/. 56'56a used the 2-nitrobenzyloxycarbonyl group as a photo-
sensitive blocking reagent in peptide synthesis. The use of this protecting
group with biologically important phosphates was pioneered by Kaplan et
46a E. Bartels, N. H. Wasserman, and B. F. Erlanger, Proc. Natl. Acad. Sci. U.S.A. 68, 1820
(1971).
46b H. A. Lester and J. M. Nerbonne, Annu. Rev. Biophys. Bioeng. U , 151 (1982).
47 A. H. Delcour and G. P. Hess, Biochemistry 25, 1793 (1986).
49j. A. McCray and D. R. Trentham, Proc. Natl. Acad. Sci. U.S.A. 77, 7237 (1980).
50 R. S. Adams and R. Y. Tsien, Annu. Rev. Physiol. 55, 755 (1993).
51 j. E. T. Corrie and D. R. Trentham, "Bioorganic Photochemistry 2" (H. Morrison, ed.).
Wiley, New York, 1993.
52 G. P. Hess, Biochemistry 32, 989 (1993).
54 p. DeMayo, Adv. Org. Chem. 2, 367 (1960).
55 j. A. Barltrop, P. J. Plant, and P. Scholield, J. Chem. Soc. Chem. Commun. 822 (1966).
56 p. Patchornik, B. Amit, and R. B. Woodward, J. Am. Chem. Soc. 92, 6333 (1970).
56ap. Kuzmic, L. Pavlickova, and M. Soucek, Czech. Chem. Commun. 51, 1293 (1986).
aL 48 and McCray and Trentham 49 and led to widespread use of "caged

compounds" in which the 2-nitrobenzyl group with and without substituents
is the protecting group. When the a-amino group of glutamate was blocked
by this reagent, the photolysis rate was 17 sec -1 and the quantum yield
0.65. 51 Several other caging groups that can be used to protect the amino
group or carboxyl group of neurotransmitters have been reported, e.g., the
3',5-dimethoxybenzoin group 57-6° and the p-hydroxyphenacyl group. 61 For
transient kinetic investigations of neurotransmitter receptors on the surface
of muscle cells and neurons, caged neurotransmitters must have a number
of properties that have not yet been examined with these compounds.
Caged neurotransmitters suitable for transient kinetic experiments, in
addition to being photolyzed in the microsecond time region so that photoly-
sis is not rate limiting, must also be (i) photolyzed to give the neurotransmit-
ter with sufficient quantum yield to allow kinetic investigations to be made
over a wide range of neurotransmitter concentration, (ii) water soluble and
sufficiently stable in aqueous solution before photolysis, (iii) photolyzed at
a wavelength greater than 335 nm to avoid cell damage, and (iv) last but
not least neither the caged compound nor the photolysis products, with
the exception of the liberated neurotransmitter, may modify the receptor-
mediated reaction being studied.
All these criteria were kept in mind when we initiated the synthesis of
photolabile inert precursors of neurotransmitters. 2-Nitrobenzyl derivatives
were used to protect the amino group of carbamoylcholine,7'62 an analog
of acetylcholine, and of the inhibitory neurotransmitters glycine63 and
G A B A . 64 For the acetylcholine receptor, carbamoylcholine is used because
it is a stable and well-characterized analog of acetylcholine and, unlike
acetylcholine, it contains a functional group, an amino group, to which the
protecting group can be attached. The introduction of the a-carboxy-2-
nitrobenzyl protecting group ( a - C N B ) 7 improved the photolysis rate of
caged carbamoylcholine; the carbamate is photolyzed with a rate of 17,000
sec -t and a quantum yield of 0.8. 7 It also improved the photolysis rate of
the caged amino groups compared to the caging group used previously44
for amino groups. However, although the N-protected photolabile precur-
57 j. C. Sheehan, R. M. Wilson, and A. W. Oxford, J. Am. Chem. Soc. 93, 7222 (1971).
58 K. R. Gee, L. Niu, K. Schaper, and G. P. Hess, Biochemistry (in press).
59 G. Papageorgiou and J. E. T. Corrie, Tetrahedron 53, 3917 (1997).
60 G. G. Hammes and C. W. Wu, Annu. Rev. Biophys. Biophys. Chem. 16, 507 (1974).
61 C. H. Park and R. S. Givens, J. Am. Chem. Soc. 119, 2453 (1997).
61 D. L. Purich, Enzymes 249, 3 (1995).
62 j. W. Walker, J. A. McCray, and G. P. Hess, Biochemistry 25, 1799 (1986).
63 m. P. Billington, K. W. Johnson, D. Ramesh, A. P. Guzikowski, B. K. Carpenter, and G. P.
Hess, Biochemistry 31, 5500 (1992).
64 R. Wieboldt, D. Ramesh, B. K. Carpenter, and G. P. Hess, Biochemistry 33, 1526 (1994).
sors of glycine 63 and GABA, 64 and we presume of other amino group-

containing compounds, are photolyzed 35 times faster than the protecting
groups previously used for amino groups, 51 they are photolyzed 30 times
slower than the caged carbamoylcholine we synthesized. 7
When we used the a-CNB group to protect the carboxyl group of
neurotransmitters, we obtained compounds that meet all the criteria listed
earlier for transient kinetic investigations of the excitatory glutamate 8 and
kainate 9 receptors and the inhibitory G A B A 11 receptors. The photochemi-
cal properties of neurotransmitters caged with the a-CNB group 7 are listed
in Table II.
The 2-methoxy-5-nitrophenyl group (MNP) 56a was used to cage the
inhibitory neurotransmitter glycine. 1° Although this derivative is pho-
tolyzed in the 1-/xsec time region, it is not stable in aqueous solution.
/3-Alanine also activates the glycine receptor 65 and MNP-caged/3-alanine
has all the desirable properties for transient kinetic investigations of the
glycine receptor. 9 The photochemical properties of the MNP caged com-
pounds are listed in Table III.
Determination of Rate and Equilibrium Constants of Channel-

Opening Mechanism
The nicotinic acetylcholine receptor in BC3H1 muscle cells will be used
as an example. Figure 3B shows some results obtained in a laser-pulse
photolysis experiment. The cell was equilibrated with caged carbamoylcho-
line. At zero time the caged carbamoylcholine was photolyzed, liberating
free carbamoylcholine in the microsecond time region. The time resolution
of the technique allows one to observe three distinct phases of the reaction:
a rising phase of the current reflecting the opening of acetylcholine receptor
channels, a maximum current amplitude, a measure of the concentration
of open receptor channels, and on a different and slower time scale, the
falling phase of the current reflecting receptor desensitization. In experi-
ments with the excitatory acetylcholine, glutamate and kainate receptors
and the inhibitory G A B A and glycine receptors 4,7-12 the rise time of the
current follows a single exponential rate equation over 85% of the reaction.
In Fig. 3C the dependence of the first-order rate constant for the current
rise time is plotted versus carbamoylcholine concentration (Fig. 3C) ac-
cording to Eq. (IVB) (Table I). The ordinate intercept of the line gives
the value of the channel-closing rate constant, kcl, and the slope allows
evaluation of the channel-opening rate constant, kop. From the dependence
of the maximum current amplitude on the concentration of carbamoylcho-
65 D. Choquet and H. Korn, Neurosci. Lett. 84, 329 (1988).

65a M. A m a d o r and J. A. Dani, Synapse 7, 207 (1991).
TABLE II
PHOTOLYTIC PROPERTIES OF BIOLOGICALLY INERT, PHOTOLABILE DERIVATIVES OF
NEUROTRANSMITTERS CAGED WITH ot-CARBOXY-2-NITROBENZYL GROUP a
a-Carboxy-2-nitrobenzyl group OOH
Product
Caged Group Photolysis quantum Target
neurotransmitter caged tl/2 (/zsec) yield receptor ReL
1. Carbamoylcholine Carbamate 45 0.8 Acetylcholine Milburn et al.,

(1989)7
NO2 O
5 . , +
II. Glutamate y-Carboxyl 21 0.14 Glutamate Wieboldt et al.

NO2 O (1994)8
II
~r--CH--O--C--(CH2) 2-CH-NH 2
Q~ f
COOH
J
COOH
III. y-Aminobutyric y-Carboxyl 19 0.16 GABA Gee et al.
acid (GABA) (1994)11
~o~ ?
CH--O --C--(CH2) 2-NH 2
COOH
IV. Kainate -y-Carboxyl 45 0.37 Kainate Niu et al.
O (1996)a,9
NO~ ? )--OH
COOH H3C~
CHz
a pH 7.5, 22°.
[25] TRANSIENTKINETICTECHNIQUES 459
TABLE III
PHOTOLYTICPROPERTIESOF BIOLOGICALLYINERT,PHOTOLABILEDERIVATIVESOF
NEUROTRANSMITI~RSCAGEDWITH2-METHOXY-5-NITROPHENYLGROUPa
NO2
2-Methoxy-5-nitrophenyl group + O - -
OCH 3
Caged Photolysis Product

neurotransmitter /1/2 (/zsec) quantum yield Ref.
I. Caged glycine
Ramesh et aL
0.2
(1993)a,l°
O--C--CH2--NH 2
OCH3
II. Caged fl-alanine
Niu et al.
0.2 (1996)~,12
O--C--(CH2)2--NH 2
OCH3
pH 7.5, 22°C.
line, one can obtain the value of the channel-opening equilibrium constant,
= kdkop, and the value of K1, the dissociation constant of the receptor
site controlling channel opening (Fig. 1B), by using Eq. (I-C) (Table I).
The value of KI and • ( ~ = kjkop) obtained from the effect of ligand
concentration on the current amplitude can be compared to the values of
K1, and k~, and kop obtained from the effect of ligand concentration on
the observed rate constant, kobs, for the rise time of the current [Fig. 3C;
Eq. (IV-B) in Table I].
The falling phase of the current gives information about the rate of
receptor desensitization. Receptor desensitization is slow compared to
channel opening and occurs in a different time scale. It is investigated more
conveniently by the cell-flow method. 33
Comparison of Values Obtained for Rate and Equilibrium Constants of

Channel-Opening Process Using Independent Techniques
The values of kop, kcl, and K1 (Fig. 1B) obtained for the acetylcholine
receptor in BC3H1 muscle cells by the laser-pulse photolysis technique and
two independent techniques (cell-flow and single-channel current re-
cordings) are compared in Table IV. (i) The cell-flow technique is a chemical
kinetic method, even though its time resolution is at least two orders of
magnitude less than that of the laser-pulse photolysis technique. It does
not allow one to determine kop and kc~. It does, however, allow one to
determine the current IA corrected for desensitization. 33'34IA is a measure
of the concentration of open receptor channels. The relationship between
IA and the constants of the mechanism in Fig. 1B (K1, • = kd/kop) and
neurotransmitter concentration is given in Eqs. (I-A) and (I-B) (Table I).
(ii) In the single-channel current-recording technique 19a the receptor is
equilibrated with neurotransmitter before the measurements are made.
At low concentrations of neurotransmitter, at which the receptor is not
completely desensitized, the technique allows one to determine the ion
specificity, conductance, and lifetime of the open receptor channel 19 where
the lifetime is a measure of kc~. Under conditions where the receptor site
density is low and the recording electrode ideally records the current from a
single receptor channel it is possible to determine Po.29 po is the conditional
probability that the channel is open while the receptor is in an active,
nondesensitized state. The relationship between Po, the fraction of nonde-
sensitized receptors in the open-channel form, and the constants of the
mechanism in Fig. 1B is given by Eq. (I-A) (Table I).
The values of the constants for the acetylcholine receptor in BC3H1 cells
determined by the three techniques are listed in Table IV. The relationship
between the concentration of open acetylcholine receptor channels and
carbamoylcholine concentration, obtained by laser-pulse photolysis, the
cell-flow,33 and the single-channel current-recording 1% technique, is given
in Fig. 4. As can be seen from Table IV, when the single-channel current-
recording and cell-flow techniques are used at appropriate ligand concentra-
tions and in the appropriate time domain, there is excellent agreement
between the results obtained with these techniques and with the laser-pulse
photolysis technique.
Investigations of Mechanism of Receptor Inhibition

Effects of lnhibitors on kop and kcz. The nicotinic acetylcholine receptor
in BC3H1 muscle cells will be taken as an example. Inspection of Fig. 3B
and 3C and of Eq. (IV-A) and (IV-B) (Table I) indicates that the laser-
pulse photolysis technique allows one to determine the rate constant of
T A B L E IV
CONSTANTS FOR REACTION STEPS IN OPENING OF NICOTINIC ACETYLCHOLINE RECEPTOR
CHANNEL IN BCaH1 MUSCLE CELLSa
Constant Method Value of constant
kop Laser-pulse photolysis 12,000 sec -1

Previous techniques (22 studies) b 450-40,000 sec -1
kcl Laser-pulse photolysis 500 4- 100 sec -1
Single channel current 400 _+ 130 sec 1
Fraction of open channels Laser-pulse photolysis 0.94
Cell flow 0.84
K1 Laser-pulse photolysis 210 -4- 90/zM
Cell flow 240/xM
a p H 7.4, at 22-23 °, and - 6 0 mV. From N. Matsubara, A. P. Billington, and G. P. Hess,

Biochemistry 31, 5507 (1992).
b See Jackson, J. Physiol. (Lond.) 397, 555 (1988); reviewed Madsen and Edeson, Trends
Pharm. Sci. 9, 315 (1988); Liu and Dilger Biophys. J. 60, 424 (1991).
the current rise indicative of channel opening and, therefore, to study the
effects of inhibitors on kcl [Eq. (IV-C), Table I] and on kop [Eq. (IV-D),
Table I] independently of one another, Before the laser-pulse photolysis
technique was developed, it was possible to determine the effects of inhibi-
tors only on the channel-closing rate, using the single-channel current-
recording technique. 65'66It was noticed that in the presence of the noncom-
petitive inhibitor procaine, the acetylcholine receptor channel rapidly
opened and closed (flickered) while it was open. This was ascribed to the
inhibitor rapidly entering and leaving the open channel, thereby blocking
and unblocking it. This inhibition mechanism was generally accepted before
the introduction of transient kinetic techniques for investigations of recep-
tor-inhibitor interactions (e.g., see Lena and Changeux 67). An alternative
mechanism is not, however, excluded by the measurements. In an alterna-
tive mechanism, the inhibitor binds to a regulatory site on the receptor
before, and after, the channel opens. In this mechanism, binding of a
compound to the receptor per se may not cause receptor inhibition, but
the subsequent conformational change does. Such mechanisms are well
documented in enzymes regulating metabolism. 24-27
How can the two mechanisms be distinguished? In both mechanisms,
the rate constant for channel closing is predicted to decrease as the inhibitor
concentration is increased. However, a mechanism in which an inhibitor
binds to the receptor before the channel opens makes a clear-cut prediction
66 E. Neher and J. H. Steinbach, J. Physiol. (London) 277, 779 (1978).

67 C. Lena and J. P. Changeux, Trends Neurosci. 16, 181 (1993).
K = 4D±60 w.M
750 t El= 110±50/~.v a
200O
kcl k~p
( s - 1) 500
(s-~)
lOOO
250
l
o I 0
o 260 460 0 2~,o 460
PROCAINE (/~M) PROCAINE (/~M)
b
'S
1
A
%
e)
40 1200
[MK-801] (~Vl) [MK-801] (1-~1)
C
I +L + A K 1 ~ AL2 l ~ n ~ AL2 open channel
k'b ~ k'f
IA*L2. inhibited eh~-~l
that is not made by the channel-blocking mechanism. It predicts that the

channel-opening rate will change as the inhibitor concentration is increased.
The two mechanisms can, therefore, be distinguished, providing one can
measure the effects of inhibitors not only on the channel-closing rate, but
also on the channel-opening rate. This can now be done using the laser-
pulse photolysis method. 68
Laser-pulse photolysis experiments in which the mechanism of inhibi-
tion of the acetylcholine receptor can be studied are illustrated in Fig. 5A.
The graph on the left shows the effect of the inhibitor procaine on the
channel-closing rate, which decreased as the procaine concentration in-
creased. This decrease in the channel-closing rate is predicted by both
mechanisms and agrees with the results obtained in single-channel current
recordings. The graph on the right shows the effect of procaine on the
channel-opening rate [Eq. (IV-D), Table I]. The channel-opening rate also
decreases as the procaine concentration is increased. This observation is
consistent with a mechanism in which the inhibitor can bind and exert its
effect both before and after the channel opens.
Additional details of the inhibition mechanism have been obtained
in laser-pulse photolysis experiments with the noncompetitive inhibitors
cocaine69 and MK-801.7° MK-801 is a noncompetitive inhibitor of the excit-
atory acetylcholine65 and N-methyl-o-aspartic acid receptors 79 and has anti-
convulsant properties. 72 The results in Fig. 5B demonstrate that MK-801
increases the channel-closing rate without affecting the channel-opening
68 L. Niu and G. P. Hess, Biochemistry 32, 3831 (1993).

69 L. Niu, L. G. Abood, and G. P. Hess, Proc. Natl. Acad. Sci. U.S.A. 92, 12008 (1995).
70 C. Grewer, L. Niu, and G. P. Hess, Biochemistry submitted (1998).
71 j. E. Huettner and B. P. Bean, Proc. Natl. Acad. Sci. U.S.A. 85, 1307 (1998).
72 E. H. F. Wong, J. A. Kemp, T. Priestley, A. R. Knight, G. N. Woodruff, and L. L. Iversen,
FIG. 5. Laser-pulse photolysis measurements with the acetylcholine receptor in BC3H1 ceils,
pH 7.4, 22-24 °, and - 6 0 mV. (A) The effect of procaine on k'cl and k'op. The concentration of
carbamoylcholine released from caged carbamoylcholine is 20 and 115/zM in the experiments
on the left and right, respectively. (B) The effect of MK-801 on k'd and k % . The concentration
of photoreleased carbamoylcholine is 25 and 160/zm in the experiments on the left and right,
respectively. Solutions of caged carbamoylcholine without and with MK-801 flowed over the
cell for 200 msec before the laser pulse. TM (C) Tentative scheme for the reaction of inhibitors
with the receptor based on the experiments shown. A, AL2, and IA, IAL2 represent the closed
form of the receptor without and with inhibitor bound, respectively. AL2 and IAL2 represent
the open-channel form of the receptor without and with inhibitor bound respectively, KI and
K'~ are the receptor-inhibitor dissociation constants, and IA*L 2 represents the inhibited
closed-channel form of the receptor. TM
464 CAGEDCOMPOUNDS 125]
rate. The resulting decrease in the channel-opening equilibrium constant
accounts for receptor inhibition in the millisecond time region. A subse-
quent and slower process characterized by the observed rate constants k'b
and k'f leads to complete inhibition of the receptor (Fig. 5C). TM
Evaluation of Receptor: Inhibitor Dissociation Constants. The time reso-
lution of the kinetic methods just described allows one to determine the
maximum current amplitudes and, therefore, the concentration of open
receptor channels before desensitization occurs. This allows one to deter-
mine the effects of inhibitors on the current amplitudes at low concentra-
tions of neurotransmitter, when the receptor is mainly in the closed-channel
form, and at high concentrations, when the receptor is mainly in the open-
channel form. 68-7° These measurements allow one to differentiate between
noncompetitive and competitive inhibitors [Eqs. (III-A) and (Ill-B), Table
I] and to determine whether two noncompetitive inhibitors bind to two
different sites [Eq. (III-E), Table I] or to the same site [Eq. (III-D), Table
I]. An example of a plot of the maximum current amplitudes obtained
before receptor desensitization occurs, in the absence and presence of
inhibitors, In and I ' A respectively, is shown in Fig. 6. In these experiments,
25
20-
/
- /
15- ""
/
-_..< - ..
"~ 10- -"
0 50 100 150 200

[Cocaine] (~M)
FIG. 6. Cell-flow experiments, pH 7.4, 22-23 °, and - 6 0 mY. Effect of cocaine concentration
on the ratio of current amplitudes, IA/I'A, in the presence of 20 ~ M carbamoylcholine and
the absence or presence of MK-801. IA is the current amplitude in the absence of inhibitor,
and I'A is the current amplitude in the presence of inhibitor. @, 20/~M carbamoylcholine,
and varying cocaine concentrations, in the absence of MK-801; K1(app) = 35/zM [Eq. (III-
A), Table I]. ©, 20/zM carbamoylcholine, and varying cocaine concentrations, in the presence
of 100 p,M MK-801 (Ki(,pp) for MK-801 is 53 /zM); Kl(app) for cocaine evaluated by use of
Eq. (III-D) (Table I) is 30/~M and is similar to the value (35 ~M) obtained in the absence
of MK-801. The dotted line was calculated for the case in which two inhibitors do not compete
for one site [Eq. (III-E), Table I] using the dissociation constants for the inhibitors determined
in experiments in which only one or the other inhibitor is present.
cocaine and MK-801 were the inhibitors used. The slopes of the lines,
proportional to the apparent cocaine-receptor dissociation constant KI,
are the same when the concentration of cocaine is varied, but 100 tzM
MK-801 is absent (0) or present ((3). Equation (III-D) (Table I) indicates
that this is expected when two noncompetitive inhibitors bind to the same
site. The dashed line is calculated for the case in which cocaine and MK-
801 bind to two different noncompetitive binding sites, using Eq. (III-E)
(Table I) and using the apparent dissociation constants for cocaine and
MK-801 from experiments in which only one of the inhibitors is present.
Integration by a Single Cell of Signals Elicited by Excitatory and

Inhibitory Neurotransmitters
In classical experiments, Hubel and Wiese173'74 showed that individual
nerve cells in the visual pathway respond to specific environmental signals.
For instance, the direction from which a light dot moves across the visual
field of a monkey and the time interval between the movements can be
discerned from the electrical recordings of a single cell. TM Many different
and complicated theories involving complex circuits of cells have been
proposed to explain how electrical signals can arise in a single cell that are
indicative of both the direction and the time interval between environmental
signals. None of the theories take into account the chemical mechanism of
receptor-mediated reactions because the reaction mechanisms were not
known. The results in Fig. 7 suggest that when the reaction mechanism and
the associated rate and equilibrium constants of the receptor-mediated
reactions are considered, a relatively simple interaction of excitatory and
inhibitory receptors can explain electrical signals in single cells, which are
indicative of the direction from which a signal comes and the time interval
between the signals. Assume that the sensory cells in Fig. 1A contain
photoreceptors and that activation of the photoreceptor on the left leads
to the release of the inhibitory neurotransmitter GABA and that activation
of the photoreceptor some distance away on the right leads to the release
of the excitatory neurotransmitter glutamate. Figure 7 shows the current
oscillation expected when (i) the glutamate and GABA receptors are acti-
vated at the same time (Fig. 7B), (ii) the GABA receptor is activated first
and then the glutamate receptor (Fig. 7A), and (iii) the glutamate receptor
is activated first and then the GABA receptor (Fig. 7C). Figures 7A and
7C are based on calculations in which the currents due to the GABA and
glutamate receptors are subtracted from each other. In the central panel
73D. H. Hubel and T. N. Wiesel,Proc. Roy. Soc. (London) (Biology) 198, (1977).
74D. Hubel, "Eye, Brain, and Vision."ScientificAmericanLibrary,New York, 1988.
A 5O0
B 5OO (~ 500
ol
glUtamnte
tors
t2k
o v o
v
c E
-500
0
, I
250
time (ms)
-500
V receptors
250
time (ms)
V
time (ms)
FIG. 7. Calculated integration of currents by a single cell arising from activation of inhibitory
(anion-conducting) G A B A receptors and excitatory (cation-conducting) glutamate receptors.
The simulations are based on experiments with hippocampal neurons equilibrated with 500
/xM caged G A B A or 500/zM caged glutamate. In each case the neurotransmitter was released
by a 308-nm, 10-nsec laser pulse of ~500/.d. The concentration of the released neurotransmit-
ters was assumed to remain unchanged. The observed current versus time trace was normalized
to the same maximum current amplitude. The current due to the opening of GABA- and
glutamate-activated receptor channels was then subtracted from each other. (A) The time
interval between activation of first the inhibitory G A B A receptors and then the excitatory
glutamate receptors was taken as ~40 msec. (B) Simultaneous activation of glutamate and
G A B A receptors. (C) The time interval between activation of first the glutamate and then
the G A B A receptors was taken as ~200 msec.
(Fig. 7B) the concentrations of the excitatory glutamate and inhibitory

G A B A were chosen such that the amplitudes of the current flowing through
excitatory and inhibitory receptor channels in a single cell were the same.
The current flows through the receptors with opposite polarity (Fig. 7B).
We would, therefore, not expect to detect a signal. This is not what happens.
When glutamate and G A B A arrive at their respective receptor sites at the
same time, the current due to the excitatory glutamate receptor rises and
is not neutralized by the current due to the inhibitory G A B A receptor.
Later the current trace falls to negative values, due to the inhibitory G A B A
receptor. The currents do not neutralize each other because the channel-
opening rates, and the rates of desensitization, are quite different for the
two receptors. Both rates are much faster for the glutamate receptor than
for the G A B A receptor. The repeated simultaneous release of glutamate
and G A B A would lead to oscillation of the electrical signal. Such oscilla-
tions are frequently observed in central nervous system neurons. 68a Figure
7C shows what happens when the glutamate receptors are activated first
and then the inhibitory G A B A receptors. The signals due to the excitatory
glutamate and inhibitory G A B A receptor are now separated by a clearly
discernible time interval. Figure 7A shows what happens when the GABA
receptors are activated first and then the glutamate receptors. The electrical
signals are quite different from those shown in the other panels of Fig. 7C.
The amplitudes of the excitatory and inhibitory signals have decreased
because the slower GABA receptor is activated first and the faster gluta-
mate receptor somewhat later. As can be seen from Fig. 7, depending on
which receptor type is activated first and which second, and the time interval
between the activations, the resulting electrical signals are different. This
difference is accounted for by the underlying mechanism of the receptor-
mediated reactions and its associated constants. We have assumed here
that the concentration of the released neurotransmitter remains constant
for 125 msec. The conclusions reached, however, are valid as long as channel
opening is fast compared to receptor desensitization or removal of neuro-
transmitter from the receptors. Under these conditions it takes only two
receptor types in a single cell to indicate the direction and time interval
between stimuli that contain spatial and temporal information.
Use of Caged Neurotransmitters to Determine How Neuronal

Circuits Function
Caged neurotransmitters have been used in identifying and locating
receptors in cells75 and tissue slices. 76'77The use of caged neurotransmitters
has also made it possible to identify receptors and cells that secrete the
neurotransmitter in a known circuit of cells controlling a measurable re-
sponse in the nematode Caenorhabditis elegans. 78 Contraction of the C.
elegans pharynx is controlled by 20 neurons. 79's° Analogous to the mamma-
lian heart, the C. elegans pharynx pumps liquid in and out of the organism.
The positions of the 20 neurons controlling pharyngeal pumping have been
identified by electron microscopy.79 Among the important questions to be
answered are which cells secrete which neurotransmitters and which cells
contain the responding neurotransmitter receptors? Because a tough cuticle
surrounds the worm, and because the animal is under turgor pressure,
electrical recordings from single cells cannot be made routinely. Electrical
recordings from the intact pharynx are, however, possible (electropha-
75 W. Denk, K. R. Delaney, A. Gelperin, D. Kleinfeld, B. W. Strowbridge, D. W. Tank, and

R. Yuste, J. Neurosci. Methods 54, 151 (1994).
76 E. M. Callaway and L. C. Katz, Proc. Natl. Acad. Sci. U.S.A. 9@, 7661 (1993).
77 L. C. Katz and M. B. Dalva, J. Neurosci. Methods 54, 205 (1994).
78 H. Li, L. Avery, W. Denk, and G. P. Hess, Proc. Natl. Acad. Sci. U.S.A. 94, 5912 (1997).
79 D. G. Albertson and J. N. Thomson, Phil. Trans. R. Soc. L o n d B 275, 299 (1976).
80 D. M. Raizen and L. Avery, Neuron 12, 483 (1994).
ryngeograms, EPGs 8°-82) (Fig. 8A). Laser ablation studies in which a single
known neuron is eliminated by a laser pulse 83,84 have shown that two
of the neurons (M3s) control the time interval between contraction and
relaxation of the pharyngeal muscleTM (Fig. 8A). The M3 neurons are also
necessary for the small negative spikes in the EPGs; the spikes disappear
when M3 neurons have been destroyed8°'81 (Fig. 8A).
It was found that in nematodes without the M3 neurons, application of
the glutamate to the pharynx via photolysis of caged glutamate restored
the negative spikes in the EPG and shortened the period between contrac-
tion and relaxation of the pharynx. Photolytic release of carbamoylcholine
or GABA from the corresponding caged compound had no effect on the
EPG. Cells responding to glutamate were identifiedTM by photolyzing caged
glutamate on the pharynx over areas 7/zm in diameter. Responses were
found on muscle cells pm4, pm5, and pm6 in the metacorpus, isthmus, and
the front region of the terminal bulb, respectively (Fig. 8B). Connections
between motor neurons M3 and muscle cells pm4 and pm5 have been
identified in electron micrographs. 79 The experiments suggest that these
cells and pm6 contain receptors that respond to glutamate and that cell
M3 secretes a neurotransmitter (presumably glutamate) that activates these
glutamate receptors. The experiments also indicate that the M3 neurons
regulate the time interval between contraction and relaxation of the pharynx
by activating glutamate receptors and that the glutamate receptors in muscle
cell pm6 in the terminal bulb, innervated by a yet unidentified neuron,
determine the relaxation of the pharynx.
Neurotransmitter Receptors that Can Now Be Investigated in the

Microsecond Time Region Using Caged Neurotransmitters
Suitable photolabile precursors for transient kinetic investigations in
the microsecond time region of the excitatory acetylcholine receptor have
been available for a number of years.7 Caged neurotransmitters suitable
for transient kinetic investigations in the microsecond time region have
now become available for investigations of the excitatory glutamate,8 and
kainate 9 receptors, and the inhibitory GABA u and glycine receptors it
(Tables II and III). Some representative laser-pulse photolysis experiments
using these new caged neurotransmitters are shown in Fig. 9. They demon-
81L. Avery,D. M. Raizen, and S. Lockery,Methods Cell Biol 48, 251 (1995).
82L. Avery,J. Exp. Biol. 175,283 (1993).
83j. E. Sulston and J. G. White,Dev. Biol. 78, 577 (1980).
84C. I. Bargmannand L. Avery,Methods Cell Biol. 48, 225 (1995).
A
200
< 0
-200
-400
200 B
~. b [PC MC ] I "I'B
-200
-400 ' <o.

~O
OOo c
n=17 n=35
Q" ~ - n=27
_200 I ~ 30 - n=36 n=29
R phase
-4oo L- ' ' ' ~ is
-50 0 50 100 150 200 250 n
Time (ms) • • ~ a • A .
FIG. 8. (A) (a) A typical electropharyngeogram (EPG) 8°'81 recorded from wild-type (N2)
C. elegans.TM This portion of the EPG corresponds to a single pumping action of the pharynx,
a contraction followed by a relaxation. It contains three phases. The E or excitation phase,
indicative of muscle contraction, is the positive spike in the EPG. The P or plateau phase is
defined as the period between the E phase and R phase; the large negative spike, the R
phase, is indicative of muscle relaxation. In the P phase, there are small negative spikes
(inhibitory postsynaptic potentials, IPSPs). 8°'81 (b) An EPG from a nematode whose M3
neurons were killed by laser ablation. 8°'84 All IPSPs disappeared in this trace and the pump
duration, the time interval between the E and R phases, increased when compared to the
wild type. (c) An EPG from the same nematode as in (b). The pharynx was equilibrated with
0.5 mM caged glutamate from which free glutamate was generated by photolysis 20 msec
after the E spike. (B) 0.5 mM caged glutamate was equilibrated with the pharynx of a nematode
in which the M3 neurons were ablated and no IPSPs were generated. The caged glutamate
was photolyzed in areas 7/,~m in diameter, indicated by circles in the diagram of the pharynx;
the circles are drawn in correct proportion to the length of the pharynx. The procorpus (PC),
metacorpus (MC), and isthmus (I) of the pharynx each contain one type of muscle cell. TMThe
bar graph below the drawing of the pharynx gives the IPSP current amplitudes obtained from
various areas of the pharynx upon photolysis of 0.5 mM caged glutamate. Error bar is SEM.
[Copyright (1996) National Academy of Sciences, U.S.A.]
A B C
5O0
<,~ 200 0.8
40(
I,-
Z150
~ 0.6
t~ ~ 0.4
P'loo 20£
,,-I,
o. 1
N so
o
~ 0,0
°-10 ~ I~ 2~ 3; ,~ I°°-422;~; ; do I~0 i~0 20~ 0 0.2 20 40 60 80

Time (ms) Time (ms) Time (ms)
FIG. 9. Current recorded from mouse neurons at pH 7.4, 22-23 °, - 6 0 mV. In each experi-
ment a neuron was incubated for 200 msec with caged compound in physiological saline. The
concentrations of the caged compound were 500/xM caged glutamate in the experiment on
the left, 500/zM caged GABA in the experiment in the middle, and 620/~M caged glycine
in the experiment on the right. At time 0 the cell was exposed to a 0.5-mJ flash of 343-nm
light generated by a Candella SLL500 pulsed dye laser and delivered through a fiber optic.
The release of the neurotransmitter in the microseconds time region led to activation of the
receptors on the cell surface. The concentration of open receptor channels was determined
by a whole-cell current-recording method.32
strate that techniques are now available to investigate the mechanism of

the channel-opening process, and of receptor activation and inhibition by
clinically used compounds and abused drugs, for the excitatory glutamate
and inhibitory GABA and glycine receptors, in addition to the nicotinic
acetylcholine receptor.
Transient Kinetic Experiments with Neurotransmitter Receptors

Expressed in Xenopus Oocytes
Xenopus laevis oocytes are used extensively to express neurotransmitter
receptors from various regions of the nervous system, from different species,
and of varying subunit composition and also to express receptors altered
by site-specific mutations, nervous system .diseases, or exposure to abused
drugs. The chemical kinetic techniques suitable for use with cells of 15 to
20/zm diameter are not suitable for use with oocytes,85 the diameter of
which is 50 to 70 times larger. One reason for this is that flow techniques
are required in conjunction with laser-pulse photolysis experiments to cali-
brate the concentration of the photoreleased neurotransmitters and to
assess possible receptor or membrane damage by the laser pulse. An ap-
proach to overcome these difficulties has been developeds5 using electrical
85 L. Niu, R. W. Vazquez, G. Nagel, T. Friedrich, E. Bamberg, R. E. Oswald, and G. P. Hess,

[9,5] TRANSIENT KINETIC TECHNIQUES 471
recordings from giant (~20/zm diameter) membrane patches. 86,87The ex-

periments in Fig. 10 illustrate that receptors expressed in Xenopus oocytes
can now be investigated, by using giant oocyte membrane patches, with
the same time resolution as receptors in muscle cells and neurons. The
measurements 85 shown were made with oocytes expressing the muscle re-
ceptor from BC3H1 cells. 88 The maximum current amplitudes observed in
the experiments were normalized to each other. On the left (solid line, Fig.
10A) is an experiment with an oocyte membrane patch and the experiment
on the right (dashed line, Fig. 10A) is an experiment with a whole oocyte
using the two-electrode voltage clamp. In the experiment on the left the
time it takes the current to reach a maximum value is 2 msec. This current
rise time of 2 msec reflects the equilibration time of cell surface receptors
with ligand. The current decay, indicative of receptor desensitization, occurs
in two phases: ~70% of the current decays with a tl/2 value of - 1 4 msec
and the remaining current decays with a tl/2 value of - 6 5 msec. In contrast,
when the measurements were made with the whole oocyte using the two-
electrode voltage-clamp method (the experiment on the right, dashed line,
Fig. 10A), the current rise time is 2 sec. The two rapid desensitization
processes seen in the experiment on the left (solid line) can no longer be
seen. They are expected to have gone to completion during the 2-sec
equilibration process between cell surface receptors and the receptor-acti-
vating ligand carbamoylcholine. From the difference in surface area be-
tween the oocyte membrane patch and the whole oocyte, the current ob-
served in the whole oocyte experiment that desensitizes slowly may reflect
the property of only a small fraction of receptors present in the oocyte
membrane that desensitize relatively slowly. The properties of the major
forms of the receptors present, which desensitize rapidly and which can be
observed in the experiment on the left, are not observed in the experiment
on the right. Similar results have been obtained in experiments with the
neuronal a7-acetylcholine receptor. 85Such results are expected in all investi-
gations with receptors expressed in oocytes in which the two-electrode
voltage-clamp method is used and in which the current rise time is slow
compared to receptor desensitization. An example of a laser-pulse photoly-
sis measurement with a giant oocyte membrane patch is shown in Fig.
86 D. W. Hilgemann, Pfluegers Arch. 415, 247 (1989).

87j. Rettinger, L. A. Vasilets, S. Eisner, and W. Schwartz, in "The Sodium Pumps"
(E. Bamberg and W. Schoner, eds.), p. 553. Steikopff, Darmstadt, Germany, 1994.
87a D. Bertrand, E. Cooper, S. Valera, D. Rungger, and M. Ballivet, Methods Neurosci. 4,
174 (1991).
87b W. Sttihmer and A. B. Parekh, in "Single-Channel Recording" (B. Sakmann and E. Neher,
eds.), 2nd Ed., p. 341, Plenum, New York, 1995.
88 D. Schubert, A. J. Harris, E. E. Devine, and S. Heinemann, J. Cell Biol. 61, 398 (1974).
A
100 ...... o 100
75 i 75
=E
E
"~ 50 50
"6
o~ 25
/ 25
i ,,
!
0
10 20"~" 150 30'~ ~oo 2~oo'~ 10b00 20000

Timo (ms)
B
80
E
,~ 40
o
0 5 10 15
Time (ms)
FIG. 10. Mouse muscle acetylcholine receptors expressed in X. laevis oocytes; measurements
made at pH 7.4, 22-23 °, and -60 mV. 77 (A) The solid line (on the left) represents the current
response obtained using a giant outside-out oocyte membrane patch with a flow technique.
The dotted line on the right represents the response obtained with the whole oocyte two-
electrode voltage-clamp technique.87a,87bThe maximum current amplitude of the response in
the experiment on the left (solid line) is 4.0 nA and for the experiment on the right (dotted
line) 17.7/~A. (B) Current induced by flash photolysis of 700/~M caged carbamoylcholine
preequilibrated with a giant outside-out oocyte membrane patch. The light source for photoly-
sis was a flash lamp (Chadwick-Helmuth, E1 Monte, CA; Model 278). The wavelength was
in the 300- to 390-nm range because of the cutoff by the lens and dichroic mirror used. The
pulse energy was -700 b~Jand the pulse length 300/xsec. The spikes at time zero are instrument
artifactsY [Copyright (1996) National Academy of Sciences, U.S.A.]
lOB. It may, t h e r e f o r e , be of i n t e r e s t to r e e x a m i n e the m a n y i n t e r e s t i n g

o b s e r v a t i o n s that have b e e n r e p o r t e d in e x p e r i m e n t s with oocytes in which
i n a d e q u a t e t i m e r e s o l u t i o n t e c h n i q u e s have b e e n e m p l o y e d .
Conclusions
T e c h n i q u e s h a v e b e e n described, using caged n e u r o t r a n s m i t t e r s , that

e n a b l e t r a n s i e n t kinetic i n v e s t i g a t i o n s to b e m a d e of cell surface receptors
expressed in muscle cells, neurons, and Xenopus oocytes with the same
time resolution previously possible only with soluble proteins. The use
of these techniques was illustrated in investigations of the formation of
transmembrane channels, receptor-drug interactions, and integration of
excitatory and inhibitory signals by a single central nervous system neuron
that have temporal and spatial information. The use of caged neurotransmit-
ters in identifying cells that contain a specific receptor and cells that secrete
the neurotransmitter in a circuit of cells controlling a measurable response
has also been illustrated.
The techniques described have been developed recently so that only a
few examples could be given. Many receptors and their isoforms from
various regions of the nervous system, from different species, and from
animals that exhibit defects in receptor-mediated reactions are now avail-
able. Hundreds of therapeutically useful compounds and abused drugs have
been identified that affect receptor-mediated reactions, and new therapeutic
agents are continuously being discovered. By analogy with the use of tran-
sient kinetics in the investigations of reactions mediated by soluble proteins,
which rapidly (millisecond) interconvert between various forms with differ-
ent ligand-binding properties and biological activities, 23-27 and the few ex-
amples available in investigations of cell surface receptors, 5 it appears likely
that the new transient kinetic techniques, and caged compounds, will be
useful for studying the many interesting and unexplored areas of cell surface
receptor function that have already been identified and those that will come
to light in the future.
Acknowledgments
Work reviewed here was supported by grants (GM04842 and NS08527) awarded to G.P.H.
by the National Institutes of Health Institute of General Medical Sciences and the National
Institutes of Health Institute of Neurological Diseases and Stroke, respectively, and by a grant
(922061) awarded to G.P.H. by the National Science Foundation. C.G. was supported by a
Feodor Lynen Fellowship of the Alexander von Humboldt Foundation.
[26] C a g e d P l a n t G r o w t h R e g u l a t o r s
By A N D R E W C. ALLAN, JANEL. W A R D , M I C H A E L H. BEALE,
and ANTHONYJ. TREWAVAS
Introduction
Although plant growth regulators (plant hormones) have been the focus
of intense research since their first isolation early this century, is it only
now becoming possible to elucidate how and where these small molecules
are first perceived by the plant cell. Most natural plant growth regulators
are weak permeable acids, making it unlikely that outwardly facing plasma
membrane receptors are the exclusive site for hormone perception. Al-
though such receptors have indeed been inferred [e.g., in the perception
of gibberellin (GA) by aleurone protoplasts 1], as the hormone diffuses past
these receptors many internal sites will also experience the signal. For
example, it has been shown that part of the growth response to auxin
resides in the action of internal receptors. 2
To answer the question of the functional localization of plant hormone
perception, studies have been made of the effects of injecting natural regula-
tor molecules [e.g., abscisic acid (ABA) 3'4 and GA 4] into plant cells. How-
ever, this technique allows for no temporal separation between the injection
event (being quite traumatic for the plant cell due to piercing the cell wall)
and subsequent exposure to the signal. We have developed UV-activated
caged plant growth regulators to allow intracellular release of these com-
pounds. "Caging" provides the ideal way to release the hormone precisely,
after the cell has recovered from injection. It also allows internal release
of plant growth regulators whose permeability does not allow for easy
loading into the cell.
This article describes the synthesis of caged plant growth regulators,
and the implications of our finding that cell responses to plant growth
regulators, mediated by pH and calcium signals, can be elicited after release
within the plasma membrane.
1 R. Hooley, M. H. Beale, and S. J. Smith, Planta 183, 274 (1991).

2 M. Claussen, H. Ltithen, and M. B6ttger, Physiol. Plant 98, 861 (1996).
3 B. E. A n d e r s o n , J. M. Ward, and J. I. Schroeder, Plant Physiol. 104, 1177 (1994).
4 S. Gilroy and R. L. Jones, Plant Physiol. 104, 1185 (1994).

All rightsof reproduction in any form reserved.
METHODS IN ENZYMOLOGY.VOL. 291 0076-6879/98$25.00
[261 CAGED PLANT GROWTH REGULATORS 475
H ~ c ~,2----~
02L3~__/~-N02
C~BsX ~ /
Abscisie acid 1-(2-nitrophenyl)ethyl ester Al~.i~eadd 4-nitrolmmzylester
(ABA-2NPE) (ABA.4NBE)
o~
OH O2N
o2e
Co
Salicylicacidl-(2-nitmphenyl)ethylester
(SA-2NPE)
1-Nlq~hlt~ylae~lcadd 1-(~dl'~l~yl)t~yl ~
0 [ / ~ ::=cH2 O
&,~=/--
lasmonicacid 1.(2-nitrophenyl)ethyl ester
(JA-2NIPE)
NO2 O ~ O NO2
Flu~ bis(f~carlmxyme~hoxy-2-nitrobenzyl)
ether
FIG. 1. The structures of caged plant growth regulators and of caged fluorescein.
S y n t h e s i s of Caged Plant H o r m o n e s 5
Those plant growth regulators that are carboxylic acids are relatively
easy to cage as photocleavable esters (Fig. 1). These are formed by coupling
to appropriate derivatives of 2-nitrobenzyl alcohol. Although 2-nitrobenzyl
esters themselves are suitable, m o r e efficient photolysis is achieved when
5j. L. Ward and M. H. Beale, Phytochemistry 38, 811 (1995).

the benzyl carbon is substituted as in 1-(2-nitrophenyl)ethyl esters. 6 In our
work, we also make use of the corresponding 4-nitrobenzyl esters, which
are nonphotocleavable and, as such, serve as excellent controls in microin-
jection experiments.
Many methods are available to activate carboxylic acids for coupling
to alcohols. For GA, AGA, jasmonic acid (JA), and 1-naphthylacetic acid
(1-NAA), we have used 5 the mixed anhydride method because this requires
no protection of other functional groups on the hormone. This method is
not applicable to acids such as salicylic acid (SA), which contain a reactive
hydroxy group. In this case, we found that activation by formation of
N-hydroxysuccinimidyl esters was convenient.
Procedure for Preparation of Caged ABA

Commercial A B A (Lancaster Synthesis, Morecambe, Lancashire, UK)
(150 mg), dissolved in anhydrous tetrahydrofuran (distilled from calcium
hydride) (3 ml), is treated with isobutylchloroformate (80/zl) and triethyl-
amine (80 tzl) at room temperature until complete formation of a less polar
product (TLC monitoring) has been observed (ca. 10 min). The solution
is then diluted in ethyl acetate containing 5% (w/v) aqueous citric acid.
The organic layer is dried (magnesium sulfate) and evaporated in vacuo
to yield ABA-isobutyloxycarbonyl anhydride (189 mg, 91%), 1H NMR
(CDCI3) ~: 7.82 (1H, d, J = 16Hz, H-5), 6.32 (1H, d, J = 16Hz, H-4), 5.95
(1H, s, H-3' ), 5.76 (1H, s, H-2), 4.06 [2H, d, J = 7Hz, OCH2CH(CH3)2],
2.48 and 2.32 (2H, 2d, J = 17Hz, H2-5'), 2.07 and 1.92 (6H, 2s, H3-7' and
H3-6); 2.05 [1H, m, CH2CH(CH3)2], 1.12 and 1.04 (6H, 2s, H3-8' and
H3-9' ), 0.99 [6H, d, J = 7Hz, OCH2CH(CH3)2].
The anhydride (68 mg) in acetonitrile (2 ml) is then treated with 1-(2-
nitrophenyl)ethanol (63 mg) and 4-dimethylaminopyridine (30 mg) for 1
hr at room temperature. After addition of ethyl acetate and 5% aqueous
citric acid, and recovery as described earlier, the product is purified by flash
chromatography on silica gel (BDH, 40-63/zm) eluting with hexane : ethyl
acetate (3 : 1) to give ABA-l-(2-nitrophenyl)ethyl ester (caged ABA, ABA-
2NPE) (60 mg, 78%) ~H NMR (CDCI3) ~: 7.98 (1H, dd, J = 8 and 1Hz,
ArH), 7.80 (1H, d, J = 16Hz, H-5); 7.65 (2H, m, ArH2), 7.45 (1H, dt, J =
8 and 1Hz, ArH); 6.40 (1H, q, J = 7Hz, OCH(CH3)Ar); 6.18 (1H, d, J =
16Hz, H-4); 5.90 (1H, br s, H-3'); 5.80 (1H, br s, H-2); 2.45 and 2.29 (2H,
2d, J = 17Hz, H2-5'); 2.10 and 2.02 (6H, 2s, H3-7' and H3-6); 1.67 (3H, d,
J=7Hz, CHCH3Ar); 1.10 and 0.98 (6H, 22s, H3-8' and H3-9'). MS m/z
6j. W. Walker, in "Cellular and Molecular Biology:A Practical Approach" (J. Chad and
H. Wheal, eds.), p. 179. IRL Press, Oxford, 1991.
[261 CAGEDPLANTGROWTHREGULATORS 477
(relative intensity) 413 ([M+], 0.4), 383(1), 263(5), 190(49), 162(19),
150(100), 135(33), 120(43) and 91(65).
This procedure is also used to prepare ABA-4-nitrobenzyl ester (ABA-
4NBE; see Fig. 1) and the corresponding gibberellin A4 (GAa-2NPE and
GAa-4NBE), JA (JA-2NPE and JA-4NBE), and 1-NAA (1-NAA-2NPE
and 1-NAA-4NBE) esters. 5
Procedure for Preparation of Caged Salicylic Acid

Salicylic acid (2.5 g) in acetone (75 ml) is treated with N-hydroxysuccini-
mide (2.08 g). Dicyclohexylcarbodiimide (3.73 g) is added slowly, and the
mixture is stirred for 3 hr, giving a white precipitate. The mixture is filtered
and the filtrate evaporated in vacuo to yield salicylic acid N-hydroxysuccini-
midyl ester (SA-NHS ester; 4.41 g), which is used in subsequent coupling
reactions without further purification.
SA-NHS ester (100 mg) in dry acetonitrile (10 ml) is treated with 1-(2-
nitrophenyl)ethanol (71 mg) and 4-dimethylaminopyridine (30 mg). The
solution is stirred under nitrogen at room temperature for 1 hr. Ethyl
acetate (50 ml) and 5% citric acid (50 ml) are added, and the product,
recovered from the organic layer, is purified by flash chromatography using
hexane-ethyl acetate mixtures to give salicylic acid 1-(2-nitrophenyl)ethyl
ester (caged SA, SA-2NPE) (92 rag). 1H NMR (CDC13) 6:7.99 (1H, dd,
J 8 and 1Hz, ArH), 7.95 (1H, dd, J = 8 and 1.5Hz, ArH), 7.71 (1H, dd,
----
J = 8 and 1.5Hz, ArH), 7.64 (1H, dt, J = 8 and 1Hz, ArH), 7.46 (2H, m,
ArH2); 6.96 (1H, dd, J = 8 and 0.7Hz, ArH), 6.92 (1H, dt, J = 7 and 0.8
Hz, ArH), 6.61 [1H, q, J = 7Hz, OCH(CH3)Ar]; 1.80 (3H, d, J = 7Hz,
CHCH3Ar). MS m/z (relative intensity) 287 (M +, 3), 219(3), 151(11),
150(100), 138(31), 120(82), 103(17) and 92(64). This procedure is also used
to prepare SA-4-nitrobenzyl ester (SA-4NBE).
The caged plant growth regulators are made up as a 10 mM stock in
methanol, vacuum dried, and frozen until use.
Injection of Plant Cells and Flash Photolysis

Although direct loading of cells by these caged molecules is possible,
injection is preferred as a means to deliver the cage into the cell. Therefore,
the choice of plant tissue used for study is usually a compromise between
what is possible to inject (the optimum being a single layer of cells with
little autofluorescence) and what has a measurable response to the plant
growth regulator. Guard cells, protoplasts, pollen tubes, and epidermal cells
have been studied. These are mounted on thin coverslips and placed on
the stage of a Nikon Diaphot TMD inverted microscope (Nikon, UK Ltd.,
Shropshire, UK) to enable the micromanipulators (Narashige NT-88 4D)

of the microinjection unit to approach from above. Microelectrodes with
tip diameters of approximately 0.3/z were pulled from GC120 filamented
electrode glass (Clark Electromedical Instruments, Reading, UK) and back-
filled with 100/zM Calcium Green-1 or SNARF-1 (Molecular Probes, Eu-
gene, OR) and 200 ~M caged probe. The remainder of the electrode was
back-filled with 3 M KC1. Care was taken to keep the cage-filled electrodes
out of direct light; filling was possible under diffuse light in an otherwise
darkened room.
For data discussed in this study, guard cells of Commelina communis
were injected. These have the advantage of having visual responses to many
phytohormones (a change in pore aperture), although they are harder to
successfully inject than neighboring epidermal cells. The fluorescent probe
and the cage were coinjected by ionotophoresis using a current of 0.1-
0.2 nA for approximately 3 sec. At least 15 min was then allowed for the
cell to recover from the injection event. Cells with visible signs of damage
(e.g., collapse, condensation of the nucleus, irregularities in chloroplast
arrangement) were discarded at this stage. Calcium or pH measurements
were then made using fluorescence microscopy with an attached photometry
system (Newcastle Photonic Systems, Newcastle-upon-Tyne, UK) or a Bio-
Rad MRC-600 laser scanning confocal microscope. 7,8
We have often used Calcium Green-1 (Molecular Probes), 9 which has
high affinity and selectivity for free Ca 2+. Although Calcium Green-1 has
the disadvantage that it is a single wavelength dye, we chose to use it for
these studies, as it may be visualized using an argon laser-scanning confocal
microscope. In addition, experiments with ratiometric calcium dyes, such
as Fura-2 and Indo-1, are not possible as they are usually excited at 340-
350 nm, overlapping with the wavelengths used for photolysis of caged
probes. However, SNARF-1, used to measure intraceUular pH, has the
advantage of being both ratiometric and excited by visual wavelengths,
which avoid release of the caged probe (excitation around 510 nm, and
emissions at 580 and >610 nm).
For in vivo photolysis of the injected probe, we have used two flash
release systems with success. First, a "pulse" of up to 30 sec, focused
through the Plan Apo 40x objective, uses the epifluorescence lamp (Xenon
75-W) of the microscope via a 360 _+ 10-nm interference filter under com-
puter control (Newcastle Photonic Systems). Such exposures could be easily
7 A. C. Allan, M. D. Fricker, J. L. Ward, M. H. Beale, and A. J. Trewavas, Plant Cell 6,

685 (1994).
8 V. E. Franklin-Tong, B. K. Drcbak, A. C. Allan, P. A. C. Watkins, and A. J. Trewavas,
Plant Cell 8, 1305 (1996).
9 R. P. Haugland, "Handbook of Fluorescent Probes and Research Chemicals." Molecular
Probes, Inc., Eugene, OR, 1992.
[26] CAGED PLANT GROWTH REGULATORS 479
30 sec UV flash
800
600
400
,f
LL
200
I I I I I
0 1O0 200 300 400
Time (see)
FIG. 2. Calibrating photolysis by the use of caged fluorescein. An individual guard cell
was injected with the caged probe and then exposed to successive UV pulses. Fluorescence
intensity was monitored using microphotometry (excitation, 480 nm; emission, 525 nm).
calibrated to a relative release by using caged fluorescein (see Fig. 2). For
example, we injected individual epidermal ceils with caged fluorescein; after
10 sec UV exposure, fluorescence increased to 400,000 counts. Aqueous
solutions of fluorescein in cell-sized droplets (within an oil emulsion on the
microscope) were used to calibrate this release to approximately 1 /~M.
One million counts correlated with a fluorescein concentration of 7/zM.
Generally, 80% of the injected cage was released following a 20-sec ex-
posure.
The second system for cage release that we have used successfully1°
requires purchase of a flash photolysis "gun." This system produces a
1-msec flash of UV light (300 mJ in the 330- to 380-nm band) from a XF-
10 flash photolysis system (Hi-Tech Scientific, Salisbury, UK). Exposures
are therefore much faster, but photometry (as shown in Figs. 2 and 4) cannot
be performed with this setup because the epifluorescent lamp (needed for
excitation of Calcium Green-1 or SNARF-1) is replaced by the flash gun.
However, it is ideal for studies using laser-scanning confocal microscopy.
The viability of cells after photolysis events is always checked using
noninjected controls or cells injected with compounds such as caged IP3
10 R. Malh6 and A. J. Trewavas, Plant Cell 8, 1935 (1996).

(Molecular Probes). Another good control molecule for our studies are
the 4-nitrobenzyl esters (e.g., ABA-4NBE) of the caged plant growth regu-
lators. Changing the position of the nitro group from the ortho to the
para position makes ABA-4NBE nonphotolabile. The 4-nitrobenzyl esters
therefore control against any activity of caged compounds without release
or any nonspecific UV activation.
Caged Plant Growth Regulators: Biological Function after Photolysis

We have found that release of plant growth regulators in the cytosol
of cells, following photolysis of the cage, results in biological activity as
measured by changes in stomatal pore aperture. For example, a decrease
in stomatal aperture occurs following a 30-sec UV photolysis of ABA-
2NPE in the injected cell without affecting the uninjected companion guard
cell (Fig. 3). Closure could be observed within 2 min of A B A release
inside the cell. Measurements of closure rate can be made using either an
adjustable Nikon Filar micrometer 10× eyepiece with an accuracy of
_+0.2/zm or on the computer after image capture. We have photolyzed
ABA-2NPE in a total of over 50 cells and observed closure in every case.
The guard cell could be incubated after injection of ABA-2NPE for up to
1 hr with no deterioration in subsequent closing activity after photolysis.
Closure can also result from damage due to microinjection. However, this
possibility can be excluded using subsequent treatment with fusicoccin (10
A ~ 1-NAA-2NPE
~-NAA-4NBE
ABA-4NBE
30sUVflash
ABA-2NPE
• , . , . , . ,
10 20 30 40
Time(rain)
Fio. 3. The effect of caged ABA or caged 1-NAA on stomatal aperture after photolytic
release in the cytosol of the guard cell. (A) Apertures of individual guard cells injected with
cages or their nonphotolysible analogs. (B) Guard cell injected with caged ABA (right cell
of the pair) before photolysis. (C) Thirty minutes after UV photolysis. (D) One hour after
subsequent perfusion with 10 p.M fusicoecin. Bar: 10/zm. [Modified from A. C. Allan, M. D.
Fricker, J. L. Ward, M. H. Beale, and A. J. Trewavas, Plant Cell 6, 685 (1994).]
[26] CAGED PLANT GROWTH REGULATORS 481
TABLE I
EFFECTS OF INTRACELLULAR RELEASE OF PLANT GROWTH REGULATORS ON CYTOPLASMIC
pH, Ca e÷, AND FINAL APERTURE RESPONSE OF Cornmelina STOMATAL GUARD CELL
Effect of release on
Caged plant Effect on
growth regulator pHeyt [Ca2+]cy t stomatal aperture
1-NAA-2NPE Acidification, some- None Slight opening

times oscillatory
ABA-2NPE Slight acidification Increase dependent Closure
on plant growth
JA-2NPE Little change Slight increase None
SA-2NPE Acidification None None
/zM), which opens guard cells (by activating the plasma membrane H ÷-
ATPase u ). Opening of the guard cell resulted from the release of 1-NAA
within the guard cell (Fig. 3), after photolysis of 1-NAA-2NPE. Ultraviolet
exposure of cells injected with the nonphotolyzable analogs, ABA-4NBE
and 1-NAA-4NBE, resulted in no significant change in guard cell aperture.
We have also studied the effect of plant growth regulator release within
the cell on cytoplasmic Ca 2+ concentration and pH. Table I summarizes
these results. Figure 4 shows how photometry experiments can provide a
high temporal resolution to hormone-induced pH changes, as monitored
by SNARF-1 fluorescence. To calibrate changes in the ratio of SNARF-1
fluorescence mixtures of isobutyric acid and ammonium hydroxide were
added at the end of each experiment to verify a fixed pH reading. Although
these results are only preliminary (our major study has involved caged
ABA and calcium measurements7), they provide an indication that auxin
and ABA have effects on secondary messengers, leading to changes in
aperture when released within the cell. Other plant growth regulators,
particularly salicylic acid, also have interesting effects worthy of future
study. Gibberellin A4 was not tested in this study.
Concluding Remarks
The use of caged phytohormones has already helped in our understand-
ing of the basis of Ca2+-induced guard cell closure. 7 We have found that
the release of ABA inside cells results in guard cell closure. Controlled
release of ABA inside the cells allowed for reduction in variation (loading,
effects of differing cell populations) seen in previous experiments involving
11 E. Marr6, Annu. Rev. Plant Physiol. 30, 273 (1979).

10 30 sec UV 66
8] flash F
7 cagedATP I I 1-6.8
6-1 I I~ ~ ..... I-7.o
4 -J ~ r ~~ t,., "r .~L.;%~.~, 17.2
10-- I I / . ..|J r-66
8_lcagedNAA I ~'1 Jail=,, ,I~IB~ L618
°j ,.o
1 I 6.6 z=
caged ABA I 6.8
• 7.0
h-
I 72' 0~.,
caged JA I
6.8
~ " 7.0
7.2
caged SA I I 6.8
I
7.2
I I I I I I I
0 1O0 200 300 400 500 600
Time (sec)
FIG. 4. The effect of photolysis of caged plant growth regulators on the cytosolic pH of
Commelina guard cells, as measured by SNARF-1 ratio photometry.
exogenously applied hormones. We were therefore able to conclude that

Ca2+-mediated events are up-regulated in plants that have experienced
elevated temperatures.
The finding that release of a plant hormone within the plant cell produces
a rapid response indicates that our ideas of tightly controlled receptors
pointing out into the apoplast, perceiving extracellular ligands, should be
adjusted for the plant world. It may be that phytohormone-binding proteins
in the cytoplasm or nucleus directly perceive the primary signal and so
transduce it to a response in an entirely intracellular system. We can not
exclude the possibility that a fraction of the cytosolicaUy released hormone
leaks out after the photolysis, thereby activating outward-facing receptors.
However, the rapid perfusion of these cells, and the observation that the
release of approximately 1 ~M A B A induces closure at the same rate as
exogenously applied ABA, at the same concentration, suggests that interior
[27] DNA REPAIR 483
receptors can also transduce this signal. Future studies with caged plant
growth regulators, perhaps complexed to dextrans, may provide us with
information about the functionality of cytosolic receptors for plant re-
sponses.
[271 U s e o f C a g e d C o m p o u n d s i n S t u d i e s
of the Kinetics of DNA Repair
By R. A . MELDRUM, R. S. CHITTOCK, and C. W. WHARTON
Introduction
DNA is susceptible to a wide range of damaging influences that can
cause severe problems for a cell if it is not repaired with high fidelity.
Oxygen radicals, alkylating agents, and radiation are perhaps the three
most potent types of damaging agents. The fidelity of the genome in living
cells is maintained by a complex series of processes that repair damaged
DNA. The repair system constantly scans the genome for damage and
affects repair of a wide range of types of damage on a continuous basis.
The predominant forms of DNA damage are base mismatches, UV-induced
pyrimidine dimers, base hydroxylation/alkylation, and single- or double-
strand breaks induced by ionizing radiation. There are several pathways
for repair, but probably the most important are nucleotide excision repair
(e.g., UV-induced pyrimidine dimers) and base excision repair (e.g., ionizing
radiation-induced single-strand breaks, where base damage and/or sugar
damage occurs). These two pathways are illustrated in outline in Fig. 1,
which shows that several sequential steps are involved in each process (for
an excellent general reference to DNA repair, see Friedburg et al.1). In
each case repair is completed by polymerization using the undamaged
strand as the template tofill the gap created by the repair process, followed
by ligation to form continuous strands.
Elegant genetic experiments have delineated the sequence of events
that take place and the enzymes involved in the catalysis of each step.
Relatively little information is available concerning the details of the kinet-
ics of repair in vivo as severe perturbations (i.e., added inhibitors) have to
be applied if specific steps are to be interrogated. Almost nothing is known
1 E. C. Freidburg, G. C. Walker, and W. Siede, "DNA Repair and Mutagenesis." ASM Press,
Washington, DC, 1995.

METHODS IN ENZYMOLOGY, VOL. 291 0076-6879198 $25.00
[27] DNA REPAIR 483
receptors can also transduce this signal. Future studies with caged plant
growth regulators, perhaps complexed to dextrans, may provide us with
information about the functionality of cytosolic receptors for plant re-
sponses.
[271 U s e o f C a g e d C o m p o u n d s i n S t u d i e s
of the Kinetics of DNA Repair
By R. A . MELDRUM, R. S. CHITTOCK, and C. W. WHARTON
Introduction
DNA is susceptible to a wide range of damaging influences that can
cause severe problems for a cell if it is not repaired with high fidelity.
Oxygen radicals, alkylating agents, and radiation are perhaps the three
most potent types of damaging agents. The fidelity of the genome in living
cells is maintained by a complex series of processes that repair damaged
DNA. The repair system constantly scans the genome for damage and
affects repair of a wide range of types of damage on a continuous basis.
The predominant forms of DNA damage are base mismatches, UV-induced
pyrimidine dimers, base hydroxylation/alkylation, and single- or double-
strand breaks induced by ionizing radiation. There are several pathways
for repair, but probably the most important are nucleotide excision repair
(e.g., UV-induced pyrimidine dimers) and base excision repair (e.g., ionizing
radiation-induced single-strand breaks, where base damage and/or sugar
damage occurs). These two pathways are illustrated in outline in Fig. 1,
which shows that several sequential steps are involved in each process (for
an excellent general reference to DNA repair, see Friedburg et al.1). In
each case repair is completed by polymerization using the undamaged
strand as the template tofill the gap created by the repair process, followed
by ligation to form continuous strands.
Elegant genetic experiments have delineated the sequence of events
that take place and the enzymes involved in the catalysis of each step.
Relatively little information is available concerning the details of the kinet-
ics of repair in vivo as severe perturbations (i.e., added inhibitors) have to
be applied if specific steps are to be interrogated. Almost nothing is known
1 E. C. Freidburg, G. C. Walker, and W. Siede, "DNA Repair and Mutagenesis." ASM Press,
Washington, DC, 1995.

METHODS IN ENZYMOLOGY, VOL. 291 0076-6879198 $25.00
484 CA~ZI~ COMPOU~OS [27]
A B
9q gT:Tq q 9
DNA
exinuclease
97q J 99Y
5"AP
endonuclease
YT:T 9Yo.
DNApolymerase+ ligase
9 9 9~H79 9 [~
.os.o,e=e.se
YY TT Y
DNA polymerase ~ Ligase
979 q q 9
FIG. 1. The two predominant pathways for repair of lesions in mammalian DNA. (A)
Pathway of base excision repair, which is responsible for the repair of oxidative and ionizing
radiation (strand break) damage. Only one to five bases are replaced in this process. (B)
Pathway for nucleotide excision repair. In this process, which is responsible for the repair of
UV-induced pyrimidine dimers, some 30 nucleotides are replaced in the repair process. Solid
square symbols show bases that have been replaced in the repair process and so can be used
to monitor repair if radiolabeled nucleotides are used in repair experiments.
a b o u t events that occur within a few seconds of damage, and it is to this

aspect that the application o f caged nucleotides has b e e n addressed.
A l t h o u g h overall repair processes take minutes or hours, it is extremely
unlikely that cells will take a leisurely attitude to d a m a g e assessment and
the r e q u i r e m e n t for e m e r g e n c y repairs. F o r a multicellular organism the
d a n g e r of m u t a t i o n (and h e n c e the chance of carcinogenesis) is m u c h m o r e
serious than cell death so significant differences will be expected, and are
seen to occur, b e t w e e n repair in single cells, e.g., bacteria, and multicellular
cells, e.g., m a m m a l i a n . It is thus i m p o r t a n t that we k n o w w h a t criteria
[271 DNA REPAIR 485
operate in the selection and kinetics of repair pathways in the face of

various types of D N A damage.
Against this somewhat uncertain background the application of caged
nucleotides to the study of D N A repair in mammalian cells carries a number
of advantages as well as some disadvantages. 2-5 Perhaps the main advantage
is that a clean start can be made to studies of repair, as the reagent can be
decaged with a laser flash at a time point that is well-defined relative to
the damaging event. Use of readily permeable nucleoside bases would blur
the start point of the experiment due to the requirement for (tri)phosphor-
ylation before incorporation into DNA. Also, the start point can be ren-
dered indistinct by incorporation of nucleotides by replication synthesis
prior to the application of D N A damage.
The damaging event itself can be applied by using lasers (either UV-
or laser-induced X-rays), so both the damage and the trigger for monitoring
of the subsequent repair kinetics can be well defined. 2-5 The caged nucleo-
tide is radiolabeled so that after quenching of the repair process, nucleotide
incorporation into D N A can be followed, although this requires the tedious
procedure of D N A isolation (or, at the least, separation from unincorpo-
rated nucleotides). Because the method monitors the common polymeriza-
tion step, both the nucleotide excision repair pathway and the base excision
pathway are assayed. Differentiation of the pathways will require a more
sophisticated approach, although the type of damage applied directs the
pathway that is predominantly utilized. With the exception of the radio-
chemical assay for incorporation of nucleotides into DNA, the experiment
readily lends itself to automation and this has been achieved as will be
described later.
The method necessarily involves one procedure that can be regarded
as a potential disadvantage and this stems from the impermeability of the
mammalian cell wall to nucleotides, even when in caged form. In order to
gain access for the caged reagents, electroporation has been used to load
the cells. This process, although kinetically advantageous compared with,
for example, detergent-induced permeabilization, which requires minutes
rather than milliseconds, it does induce damage in DNA and the nuclear
envelope. Although it has proved possible to optimize the electrical parame-
ters for electroporation so as to minimize D N A damage, it seems that the
lamins of the nuclear envelope are disrupted by voltages lower than those
2 R. A. Meldrum, S. Shall, D. R. Trentham, and C. W. Wharton, Biochem. J. 266, 885 (1990).

3 R. A. MeldruIn, S. Shall, and C. W. Wharton, Biochem. J. 266, 885 (1990).
4 C. W. Wharton, R. A. Meldrum, C. Reason, J. Boone, and W. Lester, Biochem. J. 293,
825 (1993).
5 W. S. Meaking, J. Edgerton, C. W. Wharton, and R. A. Meldrurn, Biochim. Biophys. Acta
1264, 357 (1995).
required to achieve poration. It is possible to introduce a refractory period

in which cellular functions fully recover prior to application of DNA dam-
age. An important fact is that electroporation-induced DNA damage is not
repaired on the time scale of repair experiments with caged nucleotides,
although it may influence the kinetics and fidelity of the repair response.
This aspect is discussed in more detail later.
The synthesis and application of the 2-hydroxyethyl o-nitrophenyl de-
rivatives of [methyl-3H]thymidine triphosphate, [ot-32p]thymidine triphos-
phate, and [a-32p]dideoxyadenosine triphosphate (ddATP) are described.
The latter is a chain terminator and so when this is employed only a single
label can be inserted at a given repair site. Ideally, dideoxythymidine should
be used, as thymidine is exclusive to DNA; however, it was not possible
to obtain this material in 32p-labeled form. A variety of unlabeled caged
nucleotide triphosphates have also been used; these were prepared using
the standard method. 6 The only serious problem that was encountered in
the caging reaction was when [methyl-l',2'-aH]thymidine triphosphate was
used. The additional tritium atoms on the ribose ring caused double caging
to occur, one presumably placed on the ribose hydroxyl group. This is
assumed to arise from a tritium isotope effect relayed to the 3'-OH reaction
center. This material was very hydrophobic and could not be electroporated
into cells, although it decaged satisfactorily.
Synthesis of High Specific Activity [~-a2P]Dideoxyadenosine

Triphosphate and [methyl-3H]Thymidine Triphosphate
The preparative procedure 2 is based on using 1 mCi [a-32p]dideoxy-
adenosine triphosphate labeled at 5000 Cilmmol. Due to the very high
specific activity the mass of ddATP is very small and it proved necessary
to use a higher concentration of 1-(2-nitrophenyl)diazoethane than the 80
mM used in the standard preparation. 6 Because there is also almost no
buffering power available in such a dilute solution, ATP was added to
provide buffering at pH 4.0. Accordingly, 50/xl of 5 mM ATP, adjusted to
pH 4.0, was added to i mCi ddATP (in 100/xl aqueous solution as delivered
by Amersham International), To this solution, contained in a vial that could
be tightly stoppered, was added 250/zl I M 1-(2-nitrophenyl)diazoethane
in chloroform, prepared as described. 2'6'7 Stirring of the solution proved
only partly effective so the reaction vial, wrapped in aluminum foil, was
vigorously shaken for 24 hr.
6j. W. Walker, G. P. Reid, J. A. McCray,and D. R. Trentham, J. Am. Chem. Soc. U0,

7170 (1988).
7j. W. Walker, G. P. Reid, and D. R. Trentham,Methods Enzymol. 172,288 (1989).
[271 DNA REPAIR 487
Safety Note
Considerable pressure can build up during the shaking process so it is
essential to release this pressure before opening the vial. This is achieved
by inserting a hypodermic needle through the plastic lid of the vial before
removing the aluminum foil. Both the shaking and the venting should be
performed in a well-protected fume hood.
To the reaction mixture was added 1 ml water, followed by 2 ml chloro-
form. The mixture was agitated by drawing a stream of air through the
solution and then the chloroform was removed with a pipette and discarded
to a 32p store vessel. This procedure was repeated twice more before traces
of chloroform were removed by drawing a stream of air through the solution
for 15 min.
Purification of Caged [a-a2P]ddATP

This was achieved by using high-performance liquid chromatography
(HPLC) with a reversed-phase column (Waters /~Bondapak C-18). The
column was eluted isocratically with 10 mM phosphate buffer, pH 5.5,
containing 25% methanol at a flow rate of 1 ml/min. The ATP added to
act as buffer served to perform another important function in the purifica-
tion stage. It acts as a carder to prevent the severe tailing that occurred
when ATP was not added to the reaction mix. ATP eluted essentially
coincidentally with uncaged ddATP whereas caged ATP eluted well before
caged ddATP. The elution of ATP and caged ATP was detectable using
the photometric detector set at 260 nm, whereas the elution of the 32p_
labeled materials was detected by scintillation counting of small aliquots
of the 1-ml samples collected during elution. Caged ATP elution is suffi-
ciently well separated from caged ddATP elution that essentially complete
removal of caged ATP is achieved on two passages through the HPLC
column. In practice, the presence of a small quantity of caged ATP in the
caged ddATP should have no adverse effects on D N A repair experiments,
but because of the potential cytotoxic effects of the nitrosoketone photolysis
product, material to be used in experiments with cells was always passed
through the column twice.
Analysis of Purified Caged [a-32P]ddATP

Analytical HPLC, performed using the same conditions as for the pre-
parative runs, showed that the radiolabeled material was purified to 96-98%
in terms of caging after two passages through the HPLC column, whereas
80 r I!
I!
I!
t!
|1
|1
|1
"6 Ii
I|
I I
I I
I I
| |
0.1 - ~ II
•~ 8 40 -
X
Ao
',
',
], I ',I '
I I
I- tx~ .....
0- 0~
210
I I I
0 5 1=0 15
Time (min)
FIG. 2. Photolysis and HPLC analysis of purified 32p-caged ddATP. Purified unlabeled
compound (10 n M ) was added to act as a carrier. A Waters/zBondapak C-18 reversed-phase
column was eluted at 1 ml/min with 10 mM phosphate buffer, p H 5.5, containing 25% (v/v)
methanol. Open circles and line, unphotolyzed compound; triangles and dashed line, 3-min
photolysis using a 366-nm chromatography lamp. Samples were exposed in a 3.5-cm cell
culture dishes.
the total yield was >70%. The results of a typical run are shown in Fig. 2.
Note that poor results with severe tailing occurred if a carrier was not
added. In this case unlabeled ddATP was used. Similar experiments showed
that if ATP was not added as carrier and buffer in the synthesis of caged
material the yield was 30-40% and the purity in terms of caging was 70%
after two passages through the HPLC column. The presence of the two
expected diastereoisomers of the caged compounds was observed for all
compounds tested, except caged ara-CTP.

We have not made kinetic measurements of caged compound photolysis
and the release of uncaged material, but we have assessed the efficiency
of the decaging process using UV difference spectroscopy. When caged
compounds are photolyzed, a difference spectrum appears at 310 nm and
has a differential extinction coefficient of 4040 M -I cm -I for all the com-
pounds analyzed. This can be used to follow the decaging process, provided
the complete experiment is completed within I0 min. A slow decay of the
difference spectrum occurs, which deafly has to be avoided in experiments
where quantitative measurements of decaging are required. During analyti-
[27] DNA REPAIR 489
cal procedures, an Hg (nom. 366 nm) chromatography lamp has been used
to induce photolysis, but photolysis has also been measured using a 351-
nm XeF excimer laser (10-nsec flashes, 200 mJ) as used in DNA repair
experiments. No significant differences were found in decaging behavior,
despite the very large difference in illumination intensity. Figure 3 shows
the results of a typical photolysis experiment.
[3H]Thymidine triphosphate (40-80 Ci/mmol) was caged using this pro-
cedure, although a carrier proved unnecessary for HPLC purification.
Stability of Caged Compounds in Mammalian Cells

Serum-supplemented medium proved to contain one or more enzymes
capable of causing rapid (16% in 2 hr) degradation of caged compounds,
and the degradative activity was not eliminated by heat treatment. 2 It is
therefore necessary to use a serum-free medium or phosphate-buffered
saline for cellular experiments with caged compounds. When [32p]ddATP
was incubated with thoroughly washed mouse (PY3T3) fibroblasts, there
was very little degradation in 30 min, but extensive degradation was ob-
served after 18 hr. Similar experiments in which permeabilized or lysed
cells were used gave very similar results, suggesting that a cell surface
enzyme may be involved. It thus seems that caged nucleotides appear to
be sufficiently stable in the cellular milieu to permit short term kinetic
experiments up to 30 min without the complication of the degradation of
caged nucleotides. It is unlikely that the nature of the nucleotide will have
7" 0.3
Q
'~ 5 0.2
= 3 0.1
1 I i I
0 0.5 1.0 1.5
Laser energy (J)
FIG. 3. Photolysis of caged ddTTP with 351-nm excimer laser irradiation, ddTTP (69 tzM)
in 30 mM HEPES buffer, pH 7.0, was irradiated in a 1-ml quartz microcuvette with pulses
of 25 mJ. The absorbance at 310 nm was measured by removal from the target position, and
the experiment was completed in less than 10 rain. The open symbols represent a semilogarith-
mic plot of the data, which shows that it is a first-order process.
a large effect on the degradative process as phosphodiesterases, presumed

to be responsible for the hydrolysis, tend to be very nonspecific.
DNA Repair Experimental Protocol

The overall scheme of the experiments involves the following steps, as
shown in Fig. 4. The caged nucleotides are loaded into cells by electropora-
tion (see later). An interval of a few seconds or longer is introduced to
allow some recovery and reequilibration of the cells following the electric
shock. Damage to the D N A is applied in the form of a 248-nm laser pulse
(or a pulse of soft X-rays generated by a laser-induced plasma). A second
variable delay allows repair to start, which is followed by photoactivation
of the caged nucleotides with a pulse of 351-nm U V laser light. A further
variable incubation period allows radiolabeled nucleotides to be incorpo-
rated at repair sites before the incubation is quenched and the cells are
lysed with sodium hydroxide. There are two possible ways to collect data.
The caged nucleotides may be photoactivated immediately after the D N A
damaging pulse and a variable interval allowed before quenching with
NaOH. This approach leads to profiles of the accumulation of radiolabeled
bases in the D N A known as a progress curve. The second approach is to
vary the interval between the D N A damaging laser pulse and the pulse
that photoactivates the caged reagent. This gives rise to profiles of the rate
of incorporation of radiolabeled bases in the D N A or rate profiles.
Caged 32p
nucleotides XeF laser2
placed in cuvettes (lkJ/m 2 351nm) Pipette 2
-manually Quench
(0.5ml of
Electrodes Electrodesout 5M NaOH)
2 x l c m in
Pipette 1 KrF laser 1 Incubation
Add cells Shock 248 nm O.ls - l O m ~
(0-100J/m 2) /
(0.9ml; 106cells) (500v/cm) /
/
/
/
/
/
/ Rate o f JEp
m~rl~r-
ationinto
0.Is 0.1s DNA
Time
FIG. 4. The experimental s c h e m e for the study of early events in the repair of U V and
soft X-ray-induced D N A d a m a g e in m a m m a l i a n cells. See text for experimental details.
[271 DNA REPAIR 491
Design of C o m p u t e r - C o n t r o l l e d A p p a r a t u s
The features of the apparatus are shown in Fig. 5. The turntable has
eight stations that accommodate 1-ml quartz microcuvettes, which have all
four faces polished. The 248-nm light is applied on the larger face of the
cuvette (I × 2 cm) as the dose (e.g., <100 J/m e) is much less than that
required from 351-nm laser and the power loss from expansion of the beam
is inconsequential. Also, because the 248-nm radiation is more strongly
absorbed than the 351-nm photoactivating light, it is advantageous to ar-
range it so that it has to penetrate only the short dimension of the cuvette
(4 mm), where attenuation is about 20% when 106 cells/ml are used. Custom-
made cuvettes are used in which the side walls are fabricated from high-
quality quartz in order to minimize the attenuation of the 248-nm light.
Control measurements ensure the nominal dose reaches the center of the
cuvette. Although similar considerations apply to the application of the
351-nm light, the absorption/scattering is much less. The 351-nm U V laser
beam traverses 1 cm, the normal path length of the cuvette. The apparatus
is interfaced with a microcomputer that allows flexibility in programming
so that a wide range of experiments can be conducted with this apparatus.
Following an experiment, the lysed cells were treated with trichloroace-
tic acid, and the precipitated material was collected on glass fiber filters
and extensively washed prior to scintillation counting. In some experiments
Laser 1- 248nm
KrF
DNADamage __ H.V. Powersupply
PC 200v
+ interface
Controlsall ", I_-rl
t____
operations
~te2 = NaOH
I
Laser2 - 351nm[
XeF 1
Photoactivates
cagedreagent
Pipettes and electrodesan
drivenby pneumaticram
Turntabledevicefor automaticdispensing,
laser exposure, permeabilizationand quenchingsamples
FIG.5. The computer-controlledapparatus used in DNA repair experimentswith caged nu-
cleotides.
the DNA was isolated and it was shown that >90% of the radioactivity
incorporated into precipitated material was in the DNA.
Electroporation
Electroporation is a widely used technique for introducing molecules
into cells to which the cell membrane is naturally impermeable. The most
common use and thus the best assessed for its efficiency is to introduce
foreign DNA into cells (i.e., transfection). It is the most suitable method
to introduce caged nucleotides into cells because it is rapid and potentially
reversible. These properties minimize leakage of cellular components in
permeabilized cells. When the process is used for transfection of DNA, it
is only important that a sufficient proportion of the cells into which the
foreign DNA has integrated remain viable and that the surviving cells
maintain a clonogenic potential. For experiments with caged compounds,
it is important that permeation of the whole cell population is optimal. It
is also important that the cellular process being studied, and for which the
caged compounds are intermediates, is not disturbed by the effects of
electroporation. Since we have studied the repair of DNA damage using
caged nucleotides, we have examined, in some depth, the possibility that
electroporation may itself induce DNA damage and how this can be mini-
mized.
The electroporation methods and apparatus used follows closely that
of Chu et al. 8 The permeabilization buffer used consists of 20 mM HEPES,
pH 7.05, 137 mM NaCI, 5 mM KCI, 0.7 mM Na2HPO4, 6 mM dextrose,
pH 7.05. Two different capacitor sizes were used: 1160 and 690/~F. The
t112 for discharge of a capacitor of 1160/zF is 10 msec and for a capacitor
of 690/zF is 5.8 msec. HL60 cells (0.9 ml, 106/ml in 1-ml microcuvettes, as
earlier) in permeabilization buffer were subjected to an electrical discharge
at varying voltages and at two different values of capacitance.
The DNA from the cells was then analyzed by the alkaline elution
method for breaks. Figure 6 shows the relationship between the electric
charge (coulombs) and the strand scission factor (SSF), which is a standard
parameter related to strand breaks? When these results were combined
with those of an experiment that estimated the sizes of molecules that
permeated the cells at different field strengths, it was fortunately apparent
8G. Chu, H. Hayakawa,and P. Berg, Nucleic Acids Res. 15, 1311 (1987).
9C. Baumstark-Kahn,S. Griensenbach,and R. R. Weichselbaum,Free Rad. Res. Commun.
16, 381 (1992).
10j. S. Cleggand S. A. Jackson,Biochem. J. 255, 335 (1988).
[27] DNA REPAIR 493
0.6- A 1160~1F A
• 690p.F
0.4
0.2 A
o I I I
0.1 0.2 0.3 0.4 0.5
Charge (coulombs)
FIG. 6. The relationship between total charge and SSF (DNA breaks) for electroporation
of HL60 cells. The total charge passed through the cell suspension is Q (coulombs) = C
(capacitance in Farads) × V (voltage discharge across the electrodes). Triangular symbols
represent points at which the capacitance was 1160/.LF, and circular symbols represent points
at which the capacitance was 690/~F; the voltage was varied at both levels of capacitance.
that caged nucleotides would permeate the cell membrane under electro-
poration conditions that would not induce D N A damage, i.e., below a
charge of 0.15 coulombs. Cells were electroporated at different voltages in
the presence of various sizes of fluorescein isothiocyanate (FITC)-labeled
dextrans, spun through silicon oil to separate the cells from the surrounding
aqueous medium, lysed, and the fluorescence measured. Figure 7 shows
profiles of the permeability of two different sized molecules into the cells
at different field strengths. The caged nucleotide is of intermediate size
between these two molecules, and optimum permeation of a molecule of
this size can take place at a field strength of 500 V/cm without inducing
D N A damage. At this field strength, it is estimated from these experiments
that between 5000 and 10,000 caged nucleotides gain access to the cytosol
in a single cell when the external concentration of caged molecules is
0.02/zM.
Notably there was no apparent repair of electroporation-induced D N A
breaks in a 2-hr period after application of the electric pulse. We assume
the breaks induced are double stranded and that these must severely com-
promise the survival of those cells in which they occur.
100 - • FITC 389Da
[] Fluorexon622 Da
6o
0
40
20
i I
0 200 400 600 800 1000 1200
Applied deetric field (volts/era)
FIG.7. FITC-dextrans were added to electroporation buffer at concentrations that standard-

ized the fluorescence per unit volume. HL60 ceils were electroporated in these solutions, and
the cells were then separated from the aqueous medium by centrifugation through oil and
lysed in 1% Triton X-100. The fluorescence of each sample is expressed as the percentage of
total possible fluorescence if 100% of the cells were permeabilized and took up the dye. The
capacitance used in these experiments was 1160/.~F.
R e s u l t s of DNA R e p a i r E x p e r i m e n t s
A n u n e x p e c t e d result was f o u n d in studies of the early stages of D N A

repair using the apparatus and m e t h o d s described earlier. 11 F o r b o t h 248-
n m U V light a n d 1-keV soft X-rays, a transient i n c o r p o r a t i o n of nucleotides
into D N A occurred. Maximal i n c o r p o r a t i o n occurs after ca. 30 sec after
which m o s t o f the radiolabel is r e m o v e d , the lowest level of i n c o r p o r a t i o n
being seen after ca. 1 min. With thymidine triphosphate there is t h e n a
steady i n c o r p o r a t i o n following the initial transient, but with d d A T P all
nucleotides are r e m o v e d and there is only m i n o r further incorporation.
T h e maximal value of i n c o r p o r a t i o n in the transient phase is a p p r o x i m a t e l y
equal for each nucleotide, which indicates that only a single base is incorpo-
rated at each repair site as d d A T P is a chain terminator. T w o m a i n deduc-
tions have b e e n m a d e f r o m these data.
~1R. A. Meldrum, W. Meaking, and C. W. Wharton, Nucleic Acids Res. 22, 1234 (1994).
[271 DNA REPAIR 495
The first is that a common form of damage, caused by both UV laser

and soft X-ray irradiation, is being "repaired" in the transient phase, which
we have named the "panic" response. Based on measurements of strand
breaks induced by each type of radiation, it seems most likely that direct
breaks in D N A are being "repaired." The high intensity of the UV laser
radiation induces ionizing radiation-like direct breaks and so this is pro-
posed to be the common feature. Pyrimidine dimers, normally the primary
photoproduct of UV irradiation, do not seem to play a role in the tran-
sient response.
That T-FP behaves very similarly compared with ddATP suggests that
the transient phase represents low fidelity base incorporation and that the
fidelity is sufficiently low that almost all bases need to be removed before
"correct" repair can start. It has been suggested that this response may be
related to the need to attain strand continuity (to maintain the reading
frame) in circumstances where severe damage may be imminent. Single-
strand breaks in close proximity on either strand may look like double-
strand breaks to the enzymes that survey the DNA for integrity.
An analogy for these surprising results has been proposed which likens
the panic response to the emergency repairs a dweller in a hurricane zone
might attempt to make to the roof of their house during the hurricane.
Exactness would not be important at this time, whereas a more leisurely
and purposeful repair could be made after the passage of the hurricane.
Whatever the explanation, it is clear that much remains to be done to
achieve a clear insight into the biological role of the panic response. Could,
for example, it be related to a switch into apotosis?
C o n t r i b u t o r s to V o l u m e 2 9 1
Article numbersare in parenthesesfollowingthe names of contributors.
Affiliationslistedare current.
ANDREW C. ALLAN (26), Department of Plant CHUNG-Yu CHANG (7), Worcester Founda-
Genetics, Weizmann Institute of Science, tion for Biomedical Research, Shrewsbury,
Rehovot, Israel 76100 Massachusetts 01545
CHRISTOPH ALLIN (12), Lehrstuhl far Bio- R. S. CHrrrocK (13, 27), School of Biochemis-
physik, Fakultiit Biologie, 44780 Bochum, try, University of Birmingham, Birmingham
Germany B15 2TT, United Kingdom
E. BAMBERG (17), Max Planck Institute for JODY A. DANTZIG (18), Pennsylvania Muscle
Biophysics, D-60596 Frankfurt, Germany, Institute, University of Pennsylvania School
and Johann Wolfgang Goethe-Universiti~t of Medicine, Philadelphia, Pennsylvania
Frankfurt am Main Biozentrum, D-60439 19104
Frankfurt am Main, Germany
CLAUDIA DZEJA (23), Institut far Biologische
HAGAN BAYLEY (7), Department of Medical
Informationsverarbeitung, Forschungszen-
Biochemistry and Genetics, Texas A&M
trum Jiilich, D-52425 Jiilich, Germany
Health Science Center, College Station,
Texas 77843 DAVID EPEL (16), Department of Biological
MICHAEL H. BEALE (26), IA CR-Long Ashton Sciences, Hopkins Marine Station, Stanford
Research Station, Department of Agricul- University, Pacific Grove, California 93950
tural Sciences, University of Bristol, Bristol FREDRIC S. FAY* (5), Department of Physiol-
BS18 9AF, United Kingdom ogy, University of Massachusetts, Worces-
JORGEN BENDIG (23), Institut far Organische ter, Massachusetts 01605
und Bioorganische Chemie, Humboldt- K. FENDLER (17), Max Planck Institute for
Universitiit Berlin, D-10115 Berlin, Biophysics, D-60596 Frankfurt, Germany
Germany
STEPHAN FRINGS (23), Institutfar Biologische
EDWARD B. BROWN (20), Department of
Informationsverarbeitung, Forschungszen-
Physics, Cornell University, Ithaca, New
trum Jiilich, D-52425 Jiilich, Germany
York 14853
CHmSTOPH BURMESTER(14), Abteilung Phy- TAKESHI FUNAKOSHI(19), Department of Cell
sikalische Biochemie, Max Planck Institute Biology and Anatomy, Graduate School of
for Molecular Physiology, 44139 Dort- Medicine, University of Tokyo, Hongo,
mund, Germany Tokyo 113, Japan
NICK CALLAMARAS(21), Laboratory of Cellu- TOSHIAKI FURUTA (3), Department of Biomo-
lar and Molecular Neurobiology, Depart- lecular Science, Faculty of Science, Toho
ment of Psychobiology, University of Cali- University, Chiba 274, Japan
fornia, Irvine, California 92697 DANIELA GABRIEL (6), Max Planck Insti-
BARRY K. CARPENTER (2), Department of tute for Biochemistry, 82152 Martinsried,
Chemistry, Section of Biochemistry, Molec- Germany
ular and Cell Biology, Cornell University, KYLE R. GEE (2, 4, 22), Molecular Probes,
Ithaca, New York 14853 Inc., Eugene, Oregon 97402
VALENTIN CEPUS(12), Lehrstuhl far Bio-
physik, Fakulti~t Biologie, 44780 Bochum,
Germany *Deceased
ix
X CONTRIBUTORS TO VOLUME 291
KLAUS GERWERT (12), Lehrstuhl far Bio- MITSUO IKEBE (5), Department of Physiology,
physik, Fakultgit Biologie, 44780 Bochum, University of Massachusetts, Worcester,
Germany Massachusetts 01605
RICHARD S. GIVENS (1), Department of Chem- MICHIKO IWAMURA (3), Department of Bio-
istry, University of Kansas, Lawrence, Kan- molecular Science, Faculty of Science, Toho
sas 66045 University, Chiba 274, Japan
MAURICE GOELDNER (15), Laboratoire de ANDREAS H. JUNG (1), Department of Chem-
chimie Bioorganique, Facultd de Pharma- istry, University of Kansas, Lawrence, Kan-
cie, Universitd Louis Pasteur Strasbourg, sas 66045
BP 24-F67401 lllkirch Cedex, France
JOSEPH P. Y. KAO (24), Department of Physi-
YALE E. GOLDMAN(18), Pennsylvania Muscle
ology, University of Maryland School of
Institute, University of Pennsylvania School
Medicine, and Medical Biotechnology Cen-
of Medicine, Philadelphia, Pennsylvania
ter, University of Maryland Biotechnology
19104
Institute, Baltimore, Maryland 21201
H. MAURICE GOODMAN, Department of Phys-
iology, University of Massachusetts Medical U. BENJAMIN KAUPP (23), Institut fur Bio-
School, Worcester, Massachusetts 01605 logische Informationsverarbeitung, For-
schungszentrum Jiilich, D-52425 Jiilich,
ROGER S. GOODY (14), Abteilung Physikal-
Germany
ische Biochemie, Max Planck Institute for
Molecular Physiology, 44139 Dortmund, EDYTA KRZYMAlqSKA-OLEJNIK (8), Am-
Germany bergen, Inc., Boston, Massachusetts 02215
CHRISTOF GREWER (25), Section of Biochem- HON CHEUNG LEE (22), Department of Physi-
istry, Molecular and Cell Biology, Cornell ology, University of Minnesota, Minneapo-
University, Ithaca, New York 14853 lis, Minnesota 55455
VOLKER HAGEN (23), Forschungsinstitut fiir A. MALLAVARAPU(4), Biochemistry Depart-
Molekulare Pharmakologie, D-10315 Ber- ment, University of California, San Fran-
lin, Germany cisco, California 94143
K. HARTUNG (17), Max Planck Institute for MICHAEL MARGULIS (24), Department of
Biophysics, D-60596 Frankfurt, Germany Neurology, University of Maryland School
MANFRED HEIDECKER (6), Biomolecular and of Medicine, Baltimore, Maryland 21201
Cellular Dynamics Group, Max Planck In- GERARD MARRIOTT (6, 9), Max Planck Insti-
stitute for Biochemistry, 82152 Martins- tute for Biochemistry, 82152 Martinsried,
ried, Germany Germany
GEORGE P. HESS (2, 25), Section of Biochem-
istry, Molecular and Cell Biology, Cornell JAMES A. MCCRAY (10), Department of Phys-
University, Ithaca, New York 14853 ics, Drexel University, Philadelphia, Penn-
sylvania 19104, and Division of Neurophys-
HIDEO HIGUCH1 (18), Department of Metal- iology, National Institute for Medical
lurgy, Faculty of Engineering, Tohoku Uni- Research, London NW7 1AA, England
versity, Sendai 980-77, Japan
NOBUTAKA HIROKAWA (19), Department of R. A. MELDRUM (27), School of Biochemistry,
Cell Biology and Anatomy, Graduate University of Birmingham, Birmingham
School of Medicine, University of Tokyo, B15 2TT, United Kingdom
Hongo, Tokyo 113, Japan W. TODD MILLER (7), Department of Physiol-
ROBERT E. HOESCH (24), Medical Biotechnol- ogy and Biophysics, SUNY Health Sciences
ogy Center, University of Maryland Bio- Center, Stony Brook, New York 11794
technology Institute, and University of T. J. MITCHlSON (4), Department of Cell Biol-
Maryland School of Medicine, Baltimore, ogy, Harvard Medical School, Boston, Mas-
Maryland 21201 sachusetts 02115
CONTRIBUTORS TO VOLUME 291 xi
G. NAGEL (17), Johann Wolfgang Goethe- lecular Physiology, 44139 Dortmund,

UniversitatFrankfurt am Main Biozentrum, Germany
D-60439 Frankfurt am Main, Germany R. SREEKUMAR (5), Department of Physiol-
BRETT NIBLACK (7), Worcester Foundation ogy, University of Wisconsin, Madison,
for Biomedical Research, Shrewsbury, Mas- Wisconsin 53706
sachusetts 01545 ROBERT R. SWEZEY (16), Department of
JERZY OLEJNIK (8), Department of Physics Pharmacokinetics and Metabolism, Sha-
and Molecular Biophysics Laboratory, man Pharmaceuticals, Inc., South San Fran-
Boston University, Boston, Massachusetts cisco, California 94080
02215, and AmberGen, Inc., Boston, Mas- CHA-MIN TANG (24), Department of Neurol-
sachusetts 02215 ogy, Universityof Maryland School of Med-
JOHANNES OTFL (6, 9), Max Planck Institute icine, Baltimore, Maryland 21201
for Biochemistry, 82152 Martinsried, J. A. THERIOT (4), Whitehead Institute, Cam-
Germany bridge, Massachusetts 02142
PENG PAN (7), Boehringer Ingelheim Phar- ANTHONY J. TREWAVAS(26), Institute of Cell
maceuticals, Inc., Ridgefield, Connecticut and Molecular Biology, University of Edin-
06877 burgh, Edinburgh EH3 9JH, United
Kingdom
CHAN-Ho PARK (1), Research and Develop-
ment Center, Hansol Institute of Science and AGNES TROULLIER (12), Lehrstuhl far Bio-
Technology, Nomyangju-Si, Kyung Ki- physik, Fakultiit Biologic, 44780 Bochum,
Do, Korea Germany
CAROLA ULBRICH (12), Lehrstuhl far Bio-
IAN PARKER (21), Laboratory of Cellular and
physik, Fakultiit Biologie, 44780 Bochum,
Molecular Neurobiology, Department of
Germany
Psychobiology, University of California,
Irvine, California 92697 JEEEERY W. WALKER (5), Department of
Physiology, University of Wisconsin School
LING PENG (15), Laboratoire de chimie Bioor-
of Medicine, Madison, Wisconsin 53706
ganique, Facult~ de Pharmacie, Universit~
Louis Pasteur Strasbourg, BP 24-F67401 JANE L. WARD (26), IACR-Long Ashton Re-
lllkirch Cedex, France search Station, Department of Agricultural
Sciences, University of Bristol, BristoI BS18
GERT RAPP (11), European Molecular Biol- 9AF, United Kingdom
ogy Laboratory, Outstation at DESY,
DAVID WARSHAW,Department of Molecular
D-22603 Hamburg, Germany
Physiology and Biophysics, University of
FRANCIS M. RossI (24), Medical Biotechnol- Vermont Medical School, Burlington, Ver-
ogy Center, University of Maryland Bio- mont 05405
technology Institute, Baltimore, Maryland WATT W. WEBB (20), Department of Applied
21201 and Engineering Physics, Cornell Univer-
KENNETH J. ROTHSCHILD (8), Department of sity, Ithaca, New York 14853
Physics, Boston University, Boston, Massa- JORG F. W. WEBER (1), Department of Chem-
chusetts 02215 istry, University of Kansas, Lawrence, Kan-
K. E. SAWIN (4), ICRF, Lincolns Inn, London sas 66045
WC2 3PX, England C. W. WHARTON (13, 27), School of Biochem-
AXLE SCHEIDIG (14), Abteilung Physikalische istry, University of Birmingham, Bir-
Biochemie, Max Planck Institute for Mo- mingham B15 2TT, United Kingdom
METHODS IN ENZYMOLOGY
VOLUME I. Preparation and Assay of Enzymes

Edited by SIDNEY P. COLOWICKAND NATHAN O. KAPLAN
VOLUME II. Preparation and Assay of Enzymes
VOLUME III. Preparation and Assay of Substrates
Edited by SIDNEY P. COLOWICKAND NATHAN O, KAPLAN
VOLUME IV. Special Techniques for the Enzymologist
VOLUME V. Preparation and Assay of Enzymes
VOLUME VI. Preparation and Assay of Enzymes (Continued)
Preparation and Assay of Substrates
Special Techniques
VOLUME VII. Cumulative Subject Index
Edited by SIDNEY P. COLOWICKAND NATHAN O. KAPEAN
VOLUME VIII. Complex Carbohydrates
Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG
VOLUME IX. Carbohydrate Metabolism
Edited by WILLIS A. WOOD
VOLUME X. Oxidation and Phosphorylation
Edited by RONALD W. ESTABROOKAND MAYNARD E. PULLMAN
VOLUME XI. Enzyme Structure
Edited by C. H. W. HIRS
VOLUME XII. Nucleic Acids (Parts A and B)
Edited by LAWRENCE GROSSMANAND KIVIE MOLDAVE
VOLUME XIII. Citric Acid Cycle
Edited by J. M. LOWENSTEIN
VOt,UME XIV. Lipids
Edited by J. M. LOWENSTEIN
VOLUME XV. Steroids and Terpenoids
Edited by RAYMOND B. CLAYTON
VOLUME XVI. Fast Reactions
Edited by KENNETH KUSTIN
XV
XVi METHODS IN ENZYMOLOGY
VOLUMEXVII. Metabolism of Amino Acids and Amines (Parts A and B)

Edited by HERBERTTABORAND CELIAWHITE TABOR
VOLUMEXVIII. Vitamins and Coenzymes (Parts A, B, and C)
Edited by DONALDB. McCoRMICKAND LEMUELD. WRIGHT
VOLUMEXIX. Proteolytic Enzymes
Edited by GERTRUDEE. PERLMANNAND LASZLOLORAND
VOLUMEXX. Nucleic Acids and Protein Synthesis (Part C)
Edited by KIVIE MOLDAVEAND LAWRENCEGROSSMAN
VOLUMEXXI. Nucleic Acids (Part D)
Edited by LAWRENCEGROSSMANAND KIVIE MOLDAVE
VOLUMEXXII. Enzyme Purification and Related Techniques
Edited by WILLIAMB. JAKOBY
VOLUMEXXIII. Photosynthesis (Part A)
Edited by ANTHONYSAN PIETRO
VOLUMEXXIV. Photosynthesis and Nitrogen Fixation (Part B)
VOLUMEXXV. Enzyme Structure (Part B)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUMEXXVI. Enzyme Structure (Part C)
VOLUMEXXVII. Enzyme Structure (Part D)
VOLUMEXXVIII. Complex Carbohydrates (Part B)
Edited by VICTOR GINSBURG
VOLUMEXXIX. Nucleic Acids and Protein Synthesis (Part E)
Edited by LAWRENCEGROSSMANAND KIVIE MOLDAVE
VOLUMEXXX. Nucleic Acids and Protein Synthesis (Part F)
Edited by KIVIE MOLDAVEAND LAWRENCEGROSSMAN
VOLUMEXXXI. Biomembranes (Part A)
Edited by SIDNEYFLEISCHERAND LESTERPACKER
VOLUMEXXXII. Biomembranes (Part B)
Edited by SIDNEYFLEISCHERAND LESTER PACKER
VOLUMEXXXIII. Cumulative Subject Index Volumes I-XXX
Edited by MARTHAG. DENNISAND EDWARDA. DENNIS
VOLUMEXXXIV. Affinity Techniques (Enzyme Purification: Part B)
Edited by WILLIAMB. JAKOBYAND MEIR WILCHEK
VOLUMEXXXV. Lipids (Part B)
Edited by JOHN M. LOWENSTEIN
METHODS IN ENZYMOLOGY xvii
VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones)

Edited by BERT W. O'MALLEY AND JOEL G. HARDMAN
VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones)
VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides)
Edited by JOEL G. HARDMANAND BERT W. O'MALLEY
VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues, and Organ
Systems)
Edited by JOEL G. HARDMANAND BERT W. O'MALLEY
VOLUME XL. Hormone Action (Part E: Nuclear Structure and Function)
VOLUME XLI. Carbohydrate Metabolism (Part B)
Edited by W. A. WOOD
VOLUME XLII. Carbohydrate Metabolism (Part C)
Edited by W. A. WOOD
VOLUME XLIII. Antibiotics
Edited by JOHN H. HASH
VOLUME XLIV. Immobilized Enzymes
Edited by KLAUS MOSBACH
VOLUME XLV. Proteolytic Enzymes (Part B)
Edited by LASZLO LORAND
VOLUME XLVI. Affinity Labeling
Edited by WILLIAM B. JAKOBYAND MEIR WILCHEK
VOLUME XLVII. Enzyme Structure (Part E)
VOLUME XLVIII. Enzyme Structure (Part F)
VOLUME XLIX. Enzyme Structure (Part G)
VOLUME L. Complex Carbohydrates (Part C)
VOLUME LI. Purine and Pyrimidine Nucleotide Metabolism
Edited by PATRICIA A. HOFFEE AND MARY ELLEN JONES
VOLUME LII. Biomembranes (Part C: Biological Oxidations)
Edited by SIDNEY FLEISCHERAND LESTER PACKER
VOLUME LIII. Biomembranes (Part D: Biological Oxidations)
Edited by SIDNEY FLEISCHERAND LESTER PACKER
VOLUME LIV. Biomembranes (Part E: Biological Oxidations)
Edited by SIDNEYFLEISCHERAND LESTER PACKER
xviii METHODS IN ENZYMOLOGY
VOLUME LV. Biomembranes (Part F: Bioenergetics)

Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME LVI. Biomembranes (Part G" Bioenergetics)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME LVII. Bioluminescence and Chemiluminescence
Edited by MARLENE A. DELuCA
VOLUME LVIII. Cell Culture
Edited by WILLIAMB. JAKOBYAND IRA PASTAN
VOLUME LIX. Nucleic Acids and Protein Synthesis (Part G)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN
VOLUME LX. Nucleic Acids and Protein Synthesis (Part H)
Edited by K~VIEMOLDAVE AND LAWRENCE GROSSMAN
VOLUME 61. Enzyme Structure (Part H)
VOLUME 62. Vitamins and Coenzymes (Part D)
Edited by DONALD B. McCORMICK AND LEMUEL D. WRIGHT
VOLUME 63. Enzyme Kinetics and Mechanism (Part A: Initial Rate and Inhibitor
Methods)
Edited by DANIEL L. PURICH
VOLUME 64. Enzyme Kinetics and Mechanism (Part B" Isotopic Probes and Com-
plex Enzyme Systems)
VOLUME 65. Nucleic Acids (Part I)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE
VOLUME 66. Vitamins and Coenzymes (Part E)
Edited by DONALD B. McCORMICK AND LEMUEL D. WRIGHT
VOLUME 67. Vitamins and Coenzymes (Part F)
Edited by DONALD B. McCoRMICK AND LEMUEL D. WRIGHT
VOLUME 68. R e c o m b i n a n t D N A
Edited by RAY Wu
VOLUME 69. Photosynthesis and Nitrogen Fixation (Part C)
VOLUME 70. Immunochemical Techniques (Part A)
Edited by HELEN VAN VUNAKIS AND JOHN J. LANGONE
VOLUME 71. Lipids (Part C)
VOLUME 72. Lipids (Part D)
METHODS IN ENZYMOLOGY xix
VOLUME 73. Immunochemical Techniques (Part B)

Edited by JOHN J. LANGONEAND HELEN VAN VUNAKIS
VOLUME 74. Immunochemical Techniques (Part C)
VOLUME75. Cumulative Subject Index Volumes XXXI, XXXII, XXXIV-LX
Edited by EDWARD A. DENNISAND MARTHA G. DENNIS
VOLUME 76. Hemoglobins
Edited by ERALDO ANTONINI,LUIGI ROSSI-BERNARDI,AND EMILIA CHIANCONE
VOLUME 77. Detoxication and Drug Metabolism
Edited by WILLIAMB. JAKOBY
VOLUME 78. Interferons (Part A)
Edited by SIDNEYPESTKA
VOLUME 79. Interferons (Part B)
Edited by SIDNEY PESTKA
VOLUME 80. Proteolytic Enzymes (Part C)
Edited by LASZLO LORAND
VOLUME 81. Biomembranes (Part H: Visual Pigments and Purple Mem-
branes, I)
Edited by LESTER PACKER
VOLUME 82. Structural and Contractile Proteins (Part A: Extracellular Matrix)
Edited by LEON W. CUNNINGHAM AND DIXIE W. FREDERIKSEN
VOLUME 83. Complex Carbohydrates (Part D)
VOLUME 84. Immunochemical Techniques (Part D: Selected Immunoassays)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 85. Structural and Contractile Proteins (Part B: The Contractile Appara-
tus and the Cytoskeleton)
Edited by DIXIE W. FREDERIKSENAND LEON W. CUNNINGHAM
VOLUME 86. Prostaglandins and Arachidonate Metabolites
Edited by WILLIAM E. M. LANDS AND WILLIAM L. SMITH
VOLUME 87. Enzyme Kinetics and Mechanism (Part C" Intermediates, Stereo-
chemistry, and Rate Studies)
VOLUME 88. Biomembranes (Part I: Visual Pigments and Purple Mem-
branes, II)
VOLUME 89. Carbohydrate Metabolism (Part D)
VOLUME 90. Carbohydrate Metabolism (Part E)
XX M E T H O D S IN E N Z Y M O L O G Y
VOLUME 91. Enzyme Structure (Part I)

VOLUME 92. Immunochemical Techniques (Part E: Monoclonal Antibodies and
General Immunoassay Methods)
Edited by JOHN J. LANGONEAND HELEN VAN VUNAKXS
VOLUME 93. Immunochemical Techniques (Part F: Conventional Antibodies, Fc
Receptors, and Cytotoxicity)
VOLUME 94. Polyamines
Edited by HERBERT TABOR AND CELIA WHITE TABOR
VOLUME 95. Cumulative Subject Index Volumes 61-74, 76-80
Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS
VOLUME 96. Biomembranes [Part J: Membrane Biogenesis: Assembly and Tar-
geting (General Methods; Eukaryotes)]
Edited by SIDNEY FLEISCHERAND BECCA FLEISCHER
VOLUME 97. Biomembranes [Part K: Membrane Biogenesis: Assembly and Tar-
geting (Prokaryotes, Mitochondria, and Chloroplasts)]
VOLUME 98. Biomembranes (Part L: Membrane Biogenesis: Processing and Re-
cycling)
VOLUME 99. Hormone Action (Part F: Protein Kinases)
Edited by JACKIE D. CORBINAND JOEL G. HARDMAN
VOLUME 100. Recombinant D N A (Part B)
Edited by RAY Wu, LAWRENCEGROSSMAN,AND KIVIE MOLDAVE
VOLUME 101. Recombinant D N A (Part C)
Edited by RAY Wu, LAWRENCEGROSSMAN,AND KIVIE MOLDAVE
VOLUME 102. Hormone Action (Part G: Calmodulin and Calcium-Binding Pro-
teins)
Edited by ANTHONY R. MEANS AND BERT W. O'MALLEY
VOLUME 103. Hormone Action (Part H: Neuroendocrine Peptides)
Edited by P. MICHAEL CONN
VOLUME 104. Enzyme Purification and Related Techniques (Part C)
Edited by WILLIAM B. JAKOBY
VOLUME 105. Oxygen Radicals in Biological Systems
VOLUME 106. Posttranslational Modifications (Part A)
Edited by FINN WOLD AND KIVIE MOLDAVE
VOLUME 107. Posttranslational Modifications (Part B)
Edited by FINN WOLD AND KIVIE MOLDAVE
METHODSIN ENZYMOLOGY xxi
VOLUME 108. Immunochemical Techniques (Part G: Separation and Characteriza-

tion of Lymphoid Cells)
Edited by GIOVANNIDI SABATO, JOHN J. LANGONE, AND
HELEN VAN VUNAKIS
VOLUME 109. Hormone Action (Part I: Peptide Hormones)

Edited by Luxz BIRNBAUMERAND BERT W. O'MALLEY
VOLUME 110. Steroids and Isoprenoids (Part A)
Edited by JOHN H. LAW AND HANS C. RILLING
VOLUME 111. Steroids and Isoprenoids (Part B)
Edited by JOHN H. LAW AND HANS C. RILLING
VOLUME 112. Drug and Enzyme Targeting (Part A)
Edited by KENNETHJ. WIDDERAND RALPH GREEN
VOLUME 113. Glutamate, Glutamine, Glutathione, and Related Compounds
Edited by ALTON MEISTER
VOLUME 114. Diffraction Methods for Biological Macromolecules (Part A)
Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. T1MASHEFF
VOLUME 115. Diffraction Methods for Biological Macromolecules (Part B)
Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF
VOLUME 116. Immunochemical Techniques (Part H: Effectors and Mediators of
Lymphoid Cell Functions)
Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN
VUNAKIS
VOLUME 117. Enzyme Structure (Part J)

VOLUME 118. Plant Molecular Biology
Edited by ARTHUR WEISSBACHAND HERBERT WEISSBACH
VOLUME 119. Interferons (Part C)
Edited by SIDNEY PESTKA
VOLUME 121. Immunochemical Techniques (Part I: Hybridoma Technology and
Monoclonal Antibodies)
VOLUME 122. Vitamins and Coenzymes (Part G)
Edited by FRANK CHYTILAND DONALD B. McCORMICK
VOLUME 123. Vitamins and Coenzymes (Part H)
Edited by FRANK CHYTIL AND DONALD B. McCORMICK
VOLUME 124. Hormone Action (Part J: Neuroendocrine Peptides)
Edited by P. MICHAELCONN
xxii METHODS IN ENZYMOLOGY
VOLUME 125. Biomembranes (Part M: Transport in Bacteria, Mitochondria, and

Chloroplasts: General Approaches and Transport Systems)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 126. Biomembranes (Part N: Transport in Bacteria, Mitochondria, and

Chloroplasts: Protonmotive Force)
VOLUME 127. Biomembranes (Part O: Protons and Water: Structure and Translo-
cation)
VOLUME 128. Plasma Lipoproteins (Part A: Preparation, Structure, and Molecu-
lar Biology)
Edited by JERE P. SEGREST AND JOHN J. ALBERS
VOLUME 129. Plasma Lipoproteins (Part B: Characterization, Cell Biology, and

Metabolism)
Edited by JOHN J. ALBERS AND JERE P. SEGREST
VOLUME 130. Enzyme Structure (Part K)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEEF
VOLUME 131. Enzyme Structure (Part L)
VOLUME 132. Immunochemical Techniques (Part J: Phagocytosis and Cell-Medi-
ated Cytotoxicity)
Edited by GIOVANNI DI SABATO AND JOHANNES EVERSE
VOLUME 133. Bioluminescence and Chemiluminescence (Part B)

Edited by MARLENE D E L u c A AND WILLIAM D. MCELRoY
VOLUME 134. Structural and Contractile Proteins (Part C: The Contractile Appa-
ratus and the Cytoskeleton)
Edited by RICHARD B. VALLEE
VOLUME 135. Immobilized Enzymes and Cells (Part B)

VOLUME 136. Immobilized Enzymes and Cells (Part C)
VOLUME 137. Immobilized Enzymes and Cells (Part D)
VOLUME 138. Complex Carbohydrates (Part E)
VOLUME 139. Cellular Regulators (Part A: Calcium- and Calmodulin-Binding
Proteins)
Edited by ANTHONY R. MEANS AND P. MICHAEL CONN

°..
METHODS IN ENZYMOLOGY XXlll
VOLUME 141. Cellular Regulators (Part B: Calcium and Lipids)

Edited by P. MICHAEL CONN AND ANTHONY R. MEANS
VOLUME 142. Metabolism of Aromatic Amino Acids and Amines
Edited by SEYMOUR KAUFMAN
VOLUME 143. Sulfur and Sulfur Amino Acids
Edited by WILLIAM B. JAKOBY AND OWEN GRIFEITH
VOLUME 144. Structural and Contractile Proteins (Part D: Extracellular Matrix)
Edited by LEON W. CUNNIN~HAM
VOLUME 145. Structural and Contractile Proteins (Part E: Extracellular Matrix)
Edited by LEON W. CUNNINGHAM
VOLUME 146. Peptide Growth Factors (Part A)
Edited by DAVID BARNESAND DAVID A. SIRBASKU
VOLUME 147. Peptide Growth Factors (Part B)
Edited by DAVID BARNESAND DAVID A. SIRBASKU
VOLVME 148. Plant Cell Membranes
Edited by LESTER PACKER AND ROLAND DOUCE
VOLUME 149. Drug and Enzyme Targeting (Part B)
Edited by RALPH GREEN AND KENNETH J. WIDDER
VOLUME 150. Immunochemical Techniques (Part K: In Vitro Models of B and T
Cell Functions and Lymphoid Cell Receptors)
Edited by GIOVANNI DI SABATO
VOLUME 151. Molecular Genetics of Mammalian Cells
Edited by MICHAEL M. GOTFESMAN
VOLUME 152. Guide to Molecular Cloning Techniques
Edited by SHELBY L. BERGER AND ALAN R. KIMMEL
VOLUME 153. Recombinant DNA (Part D)
Edited by RAY Wu AND LAWRENCE GROSSMAN
VOLUME 154. Recombinant DNA (Part E)
Edited by RAY W u AND LAWRENCE GROSSMAN
VOLUME 155. Recombinant DNA (Part F)
Edited by RAY Wu
VOLUME 156. Biomembranes (Part P: ATP-Driven Pumps and Related Trans-
port: The Na,K-Pump)
VOLUME 157. Biomembranes (Part Q: ATP-Driven Pumps and Related Trans-
port: Calcium, Proton, and Potassium Pumps)
VOLUME 158. Metalloproteins (Part A)
Edited by JAMES F. RIORDAN AND BERT L. VALLEE
xxiv METHODS IN ENZYMOLOGY
VOLUME 159. Initiation and Termination of Cyclic Nucleotide Action

Edited by JACKIE D. CORBIN AND ROGER A. JOHNSON
VOLUME 160. Biomass (Part A: Cellulose and Hemicellulose)
Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG
VOLUME 161. Biomass (Part B: Lignin, Pectin, and Chitin)
Edited by WILLISA. WOOD AND SCOTt T. KELLOGG
VOLUME 162. Immunochemical Techniques (Part L: Chemotaxis and Inflam-
mation)
Edited by GIOVANNXDI SABATO
VOLUME 163. Immunochemical Techniques (Part M: Chemotaxis and Inflam-
mation)
Edited by GIOVANNIDI SABATO
VOLUME 164. Ribosomes
Edited by HARRY F. NOLLER, JR., AND KIVIE MOLDAVE
VOLUME 165. Microbial Toxins: Tools for Enzymology
Edited by SIDNEY HARSHMAN
VOLUME 166. Branched-Chain Amino Acids
Edited by ROBERT HARRIS AND JOHN R. SOKATCH
VOLUME 167. Cyanobacteria
Edited by LESTER PACKER AND ALEXANDER N. GLAZER
VOLUME 168. Hormone Action (Part K: Neuroendocrine Peptides)
Edited by P. MICHAELCONN
VOLUME 169. Platelets: Receptors, Adhesion, Secretion (Part A)
Edited by JACEK HAWIGER
VOLUME 170. Nucleosomes
Edited by PAUL M. WASSARMAN AND ROGER D. KORNBERG
VOLUME 171. Biomembranes (Part R: Transport Theory: Cells and Model Mem-
branes)
VOLUME 172. Biomembranes (Part S: Transport: Membrane Isolation and Char-
acterization)
VOLUME 173. Biomembranes ]Part T: Cellular and Subcellular Transport: Eukary-
otic (Nonepithelial) Cells]
VOLUME 174. Biomembranes [Part U: Cellular and Subcellular Transport: Eukar-
yotic (Nonepithelial) Cells]
METHODSIN ENZYMOLOGY XXV
VOLUME176. Nuclear Magnetic Resonance (Part A: Spectral Techniques and Dy-

namics)
Edited by NORMANJ. OPPENHEIMERAND THOMASL. JAMES
VOLUME177. Nuclear Magnetic Resonance (Part B: Structure and Mechanism)
Edited by NORMANJ. OPPENHEIMERAND THOMASL. JAMES
VOLUME178. Antibodies, Antigens, and Molecular Mimicry
Edited by JOHN J. LANGONE
VOLUME179. Complex Carbohydrates (Part F)
VOLUME180. RNA Processing (Part A: General Methods)
Edited by JAMES E. DAHLBERGAND JOHN N. ABELSON
VOLUME181. RNA Processing (Part B: Specific Methods)
Edited by JAMES E. DAHLBERGAND JOHN N. ABELSON
VOLUME182. Guide to Protein Purification
Edited by MURRAYP. DEUTSCHER
VOLUME183. Molecular Evolution: Computer Analysis of Protein and Nucleic
Acid Sequences
Edited by RUSSELLF. DOOLITTLE
VOLUME184. Avidin-Biotin Technology
Edited by MEIR WILCHEKAND EDWARDA. BAYER
VOLUME185. Gene Expression Technology
Edited by DAVID V. GOEDDEL
VOLUME186. Oxygen Radicals in Biological Systems (Part B: Oxygen Radicals
and Antioxidants)
Edited by LESTER PACKERAND ALEXANDERN. GLAZER
VOLUME 187. Arachidonate Related Lipid Mediators
Edited by ROBERTC. MURPHYAND FRANKA. FITZPATRICK
VOLUME188. Hydrocarbons and Methylotrophy
Edited by MARY E. LIDSTROM
VOLUME189. Retinoids (Part A: Molecular and Metabolic Aspects)
VOLUME190. Retinoids (Part B: Cell Differentiation and Clinical Applications)
Edited by LESTERPACKER
VOLUME191. Biomembranes (Part V: Cellular and Subcellular Transport: Epithe-
lial Cells)
Edited by SIDNEYFLEISCHERAND BECCA FLEISCHER
VOLUME 192. Biomembranes (Part W: Cellular and Subcellular Transport: Epi-
thelial Cells)
Edited by SIDNEYFLEISCHERAND BECCA FLEISCHER
xxvi METHODS IN ENZYMOLOGY
VOLUME 193. Mass Spectrometry

Edited by JAMES A. McCLOSKE¥
VOLUME 194. Guide to Yeast Genetics and Molecular Biology
Edited by CHRISTINE GUTHRIE AND GERALD R. FINK
VOLUME 195. Adenylyl Cyclase, G Proteins, and Guanylyl Cyclase
Edited by ROGER A. JOHNSON AND JACKIE D. CORBIN
VOLUME 196. Molecular Motors and the Cytoskeleton
Edited by RICHARD B. VALLEE
VOLUME 197. Phospholipases
Edited by EDWARD A. DENNIS
VOLUME 198. Peptide Growth Factors (Part C)
Edited by DAVID BARNES, J. P. MATHER, AND GORDON H. SATO
VOLUME 200. Protein Phosphorylation (Part A: Protein Kinases: Assays, Purifica-
tion, Antibodies, Functional Analysis, Cloning, and Expression)
Edited by TONY HUNTERAND BARTHOLOMEWM. SEFrON
VOLUME 201. Protein Phosphorylation (Part B: Analysis of Protein Phosphoryla-
tion, Protein Kinase Inhibitors, and Protein Phosphatases)
Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON
VOLUME 202. Molecular Design and Modeling: Concepts and Applications (Part
A: Proteins, Peptides, and Enzymes)
VOLUME 203. Molecular Design and Modeling: Concepts and Applications (Part
B: Antibodies and Antigens, Nucleic Acids, Polysaccharides, and Drugs)
VOLUME 204. Bacterial Genetic Systems
Edited by JEFFREY H. MILLER
VOLUME 205. Metallobiochemistry (Part B: Metallothionein and Related Mole-
cules)
VOLUME 206. Cytochrome P450
Edited by MICHAEL R, WATERMAN AND ERIC F. JOHNSON
VOLUME 207, Ion Channels
Edited by BERNARDO RUbY AND LINDA E. IVERSON
VOLUME 208, Protein-DNA Interactions
Edited by ROBERT T. SAUER
VOLUME 209. Phospholipid Biosynthesis
Edited by EDWARD A. DENNIS AND DENNIS E. VANCE
VOLUME 210. Numerical Computer Methods
Edited by LUDWIGBRANDAND MICHAELL. JOHNSON
METHODS IN ENZYMOLOGY xxvii
VOLUME211. DNA Structures (Part A: Synthesis and Physical Analysis of

DNA)
Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG
VOLUME212. DNA Structures (Part B: Chemical and Electrophoretic Analysis
of DNA)
Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG
VOLUME213. Carotenoids (Part A: Chemistry, Separation, Quantitation, and
Antioxidation)
VOLUME214. Carotenoids (Part B" Metabolism, Genetics, and Biosynthesis)
VOLUME215. Platelets: Receptors, Adhesion, Secretion (Part B)
Edited by JACEK J. HAWIGER
VOLUME 216. Recombinant DNA (Part G)
Edited by RAY Wu
VOLUME217. Recombinant DNA (Part H)
Edited by RAy Wu
VOLUME218. Recombinant DNA (Part I)
Edited by RAY Wu
VOLUME219. Reconstitution of Intracellular Transport
Edited by JAMES E. ROTHMAN
VOLUME220. Membrane Fusion Techniques (Part A)
Edited by NEJAT Dt3ZGONE~
VOLUME221. Membrane Fusion Techniques (Part B)
Edited by NEJAT Dt3ZGONE~
VOLUME222. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement
Activation (Part A: Mammalian Blood Coagulation Factors and Inhibitors)
Edited by LASZLO LORAND AND KENNETH G. MANN
VOLUME223. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement
Activation (Part B: Complement Activation, Fibrinolysis, and Nonmammalian
Blood Coagulation Factors)
Edited by LASZLO LORAND AND KENNETH G. MANN
VOLUME224. Molecular Evolution: Producing the Biochemical Data
Edited by ELIZABETH ANNE ZIMMER, THOMAS J. WHITE, REBECCA L. CANN,
AND ALLAN C. WILSON
VOLUME225. Guide to Techniques in Mouse Development

Edited by PAUL M. WASSARMAN AND MELVIN L. DEPAMPHILIS
VOLUME226. Metallobiochemistry (Part C" Spectroscopic and Physical Methods
for Probing Metal Ion Environments in Metalloenzymes and Metalloproteins)
xxviii METHODS IN ENZYMOLOGY
VOLUME 227. Metallobiochemistry (Part D: Physical and Spectroscopic Methods

for Probing Metal Ion Environments in Metalloproteins)
VOLUME 228. Aqueous Two-Phase Systems
Edited by HARRY WALTER AND GOTE JOHANSSON
VOLUME 230. Guide to Techniques in Glycobiology
Edited by WILLIAMJ. LENNARZ AND GERALD W. HART
VOLUME 231. Hemoglobins (Part B: Biochemical and Analytical Methods)
Edited by JOHANNESEVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW
VOLUME 232. Hemoglobins (Part C: Biophysical Methods)
Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW
VOLUME 233. Oxygen Radicals in Biological Systems (Part C)
VOLUME 234. Oxygen Radicals in Biological Systems (Part D)
VOLUME 235. Bacterial Pathogenesis (Part A: Identification and Regulation of
Virulence Factors)
Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL
VOLUME 236. Bacterial Pathogenesis (Part B: Integration of Pathogenic Bacteria
with Host Cells)
Edited by VIRGINIAL. CLARKAND PATRIKM. BAVOIL
VOLUME 237. Heterotrimeric G Proteins
Edited by RAVI IYENGAR
VOLUME 238. Heterotrimeric G-Protein Effectors
Edited by RAVI IYENGAR
VOLUME 239. Nuclear Magnetic Resonance (Part C)
Edited by THOMAS L. JAMES AND NORMAN J. OPPENHEIMER
VOLUME 240. Numerical Computer Methods (Part B)
Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND
VOLUME 241. Retroviral Proteases
Edited by LAWRENCE C. KUO AND JULES A. SHAFER
VOLUME 242. Neoglycoconjugates (Part A)
Edited by Y. C. LEE AND REIKO T. LEE
VOLUME 243. Inorganic Microbial Sulfur Metabolism
Edited by HARRY D. PECK, JR., AND JEAN LEGALL
VOLUME 244. Proteolytic Enzymes: Serine and Cysteine Peptidases
Edited by ALAN J. BARRETT
METHODS IN ENZYMOLOGY xxix
VOLUME245. Extracellular Matrix Components

Edited by E. RUOSLAHTIAND E. ENGVALL
VOLUME246. Biochemical Spectroscopy
Edited by KENNETHSAUER
VOLUME247. Neoglycoconjugates (Part B: Biomedical Applications)
Edited by Y. C. LEE AND REIKO T. LEE
VOLUME248. Proteolytic Enzymes: Aspartic and Metallo Peptidases
Edited by ALAN J. BARRETT
VOLUME249. Enzyme Kinetics and Mechanism (Part D" Developments in En-
zyme Dynamics)
VOLUME250. Lipid Modifications of Proteins
Edited by PATRICKJ. CASEYAND JANICEE. Buss
VOLUME251. Biothiols (Part A: Monothiols and Dithiols, Protein Thiols, and
Thiyl Radicals)
VOLUME252. Biothiols (Part B: Glutathione and Thioredoxin; Thiols in Signal
Transduction and Gene Regulation)
VOLUME253. Adhesion of Microbial Pathogens
Edited by RON J. DOYLEAND ITZHAKOFEK
VOLUME254. Oncogene Techniques
Edited by PETER K. VOGTAND INDERM. VERMA
VOLUME255. Small GTPases and Their Regulators (Part A: RaN Family)
Edited by W. E. BALCH, CHANNINGJ. DER, AND ALAN HALL
VOLUME256. Small GTPases and Their Regulators (Part B: Rho Family)
VOLUME257. Small GTPases and Their Regulators (Part C: Proteins Involved in
Transport)
VOLUME258. Redox-Active Amino Acids in Biology
Edited by JUDITH P. KLINMAN
VOLUME259. Energetics of Biological Macromolecules
Edited by MICHAELL. JOHNSONAND GARY K. ACKERS
VOLUME260. Mitochondrial Biogenesis and Genetics (Part A)
Edited by GIUSEPPEM. AT/'ARDIAND ANNE CHOMYN
VOLUME261. Nuclear Magnetic Resonance and Nucleic Acids
Edited by THOMASL. JAMES
XXX M E T H O D S IN E N Z Y M O L O G Y
VOLUME262. DNA Replication

Edited by JUDITHL. CAMPBELL
VOLUME263. Plasma Lipoproteins (Part C: Quantitation)
Edited by WILLIAMA. BRADLEY,SANDRAH. GIANTURCO,AND JERE P. SEGREST
VOLUME 264. Mitochondrial Biogenesis and Genetics (Part B)
Edited by GIUSEPPEM. ATTARDIAND ANNE CHOMYN
VOLUME 265. Cumulative Subject Index Volumes 228, 230-262
VOLUME 266. Computer Methods for Macromolecular Sequence Analysis
Edited by RUSSELLF. DOOLrrrLE
VOLUME 267. Combinatorial Chemistry
Edited by JOHN N. ABELSON
VOLUME 268. Nitric Oxide (Part A: Sources and Detection of NO; NO Synthase)
VOLUME269. Nitric Oxide (Part B: Physiological and Pathological Processes)
VOLUME270. High Resolution Separation and Analysis of Biological Macromole-
cules (Part A: Fundamentals)
Edited by BARRY L. KARGERAND WILLIAMS. HANCOCK
VOLUME271. High Resolution Separation and Analysis of Biological Macromole-
cules (Part B: Applications)
Edited by BARRYL. KARGERAND WILLIAMS. HANCOCK
VOLUME 272. Cytochrome P450 (Part B)
Edited by ERIC F. JOHNSONAND MICHAELR. WATERMAN
VOLUME 273. RNA Polymerase and Associated Factors (Part A)
Edited by SANKARADHYA
VOLUME 274. RNA Polymerase and Associated Factors (Part B)
Edited by SANKARADHYA
VOLUME 275. Viral Polymerases and Related Proteins
Edited by LAWRENCEC. Kuo, DAVID B. OLSEN, AND STEVENS. CARROLL
VOLUME 276. Macromolecular Crystallography (Part A)
Edited by CHARLESW. CARTER,JR., AND ROBERT M. SWEET
VOLUME 277. Macromolecular Crystallography (Part B)
Edited by CHARLESW. CARTER,JR., AND ROBERT M. SWEET
VOLUME 278. Fluorescence Spectroscopy
Edited by LUDWIGBRANDANDMICHAELL. JOHNSON
VOLUME 279. Vitamins and Coenzymes, Part I
Edited by DONALDB. McCORMICK,JOHN W. SUTTIE,AND CONRADWAGNER
VOLUME 280. Vitamins and Coenzymes, Part J
Edited by DONALDB. McCORMICK,JOHN W. SUTTIE,AND CONRADWAGNER
METHODS IN ENZYMOLOGY xxxi
VOLUME 281. Vitamins and Coenzymes, Part K

Edited by DONALD B. McCoRMICK, JOHN W. SUTTIE, AND CONRAD WAGNER
VOLUME 282. Vitamins and Coenzymes, Part L
Edited by DONALD B. McCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER
VOLUME 283. Cell Cycle Control
Edited by WILLIAM G. DUNPHY
VOLUME 284. Lipases (Part A: Biotechnology)
Edited by BYRON RUmN AND EDWARD A. DENNIS
VOLUME 285. Cumulative Subject Index Volumes 263, 264, 266-289
VOLUME 286. Lipases (Part B: Enzyme Characterization and Utilization)
Edited by BYRON RUBIN AND EDWARD A. DENNIS
VOLUME 287. Chemokines
Edited by RICHARD HORUK
VOLUME 288. Chemokine Receptors
Edited by RICHARD HORUK
VOLUME 289. Solid Phase Peptide Synthesis
Edited by GREGG B. FIELDS
VOLUME 290. Molecular Chaperones
Edited by GEORGE H. LORIMERAND THOMAS BALDWIN
VOLUME 299. Oxidants and Antioxidants (Part A)
VOLUME 300. Oxidants and Antioxidants (Part B)
VOLUME 301. Nitric Oxide (Part C: Biological and Antioxidant Properties) (In
preparation)
AUTHOR INDEX 497
Author Index
Numbers in parentheses are footnote reference numbers and indicate that an author's work
is referred to although the name is not cited in the text.
A Andersson, I., 266

Andres, V., 278
Aarhus, R., 403, 404, 406, 407, 407(9), 411,
Andrews, M. A. W., 327
415(7, 13)
Andrews, P. C., 78
Abood, L. G., 30, 32(13), 463, 464(69)
Andrieux, C. P., 10
Abrahams, J. P., 308
Anson, M., 323
Acerbis, G., 152
Antar, Y. M., 192
Adachi, S., 256, 277
Aoki, T., 335, 337(76), 340(76)
Adams, P. J. H. M., 83
Aoshima, H., 31, 46, 447, 448(27a), 450,
Adams, R. S., 455
460(29)
Adams, S. R., 30, 77, 78, 95, 117, 155, 223,
Apell, H.-J., 204, 291, 305(11), 305(47),
323, 324, 357, 358, 365, 365(13), 367, 368,
305(48), 305(49), 305(50), 306
369(12), 374(12), 376(12), 416, 421,
Applied Biosystems, 145, 153(38)
431,439
Arbuckle, T., 269
Adhya, S., 416
Arner, A., 215
Afanas'eva, G. B., 69
Arnost, M. J., 69
Agrawal, S., 153
Asami, K., 279
Albers, W., 291, 294(13), 297(13), 298(13),
Ashley, C. C., 310, 323, 357
305(12) Athey, P. S., 4, 5(16, 17), 7(16, 17), 8(16, 17),
Albertson, D. G., 467, 468(79)
17(16,17), 18(16,17), 19(16, 17), 321,323
Alfano, R. R., 179 Augustine, G. J., 122, 381
Allan, A. C., 474, 478, 480
Augustinsson, K.-B., 265
Allen, P. M., 123 Avery, L., 30, 467, 468, 468(78, 80), 469(78,
Allen, T. S., 310, 318
80, 81, 83)
Allin, C., 223 Axelsen, P. H., 265, 273(6)
Almers, W., 221, 241, 357
Almo, S. C., 95,216,217(24), 256,259(12), 308
Altendorf, K., 305(46), 306
Altenhofen, W., 417 B
Alvarez, L. J., 409
Alves, A. M., 136 Baas, P. W.; 353
Amador, M., 457 Bagshaw, C. R., 215, 335
Amit, B., 31, 98, 155, 455 Baldwin, J. E., 8
Amm~il~i, C., 220 Ballivet, M., 471,472(87a)
Amos, A. A., 21 Balogh-Nair, V., 236
Amstutz, C., 368 Bamberg, E., 32, 289, 290, 291,291(2), 293(8,
Andersen, M. L., 10 10), 294(9), 298(9), 299(8), 300, 301,
Anderson, B. E., 474 305(5, 6, 8-10, 25, 44, 46, 52), 306, 470,
Anderson, J. C., 8 471(85), 472(85)
498 AUTHOR INDEX
Barabas, K., 293, 296(17) Black, M. M., 353

Barany, F., 153 Blattler, W. A., 96, 118, 141, 155, 156(13),
Bargmann, C. I., 468, 469(83) 175(13)
Barltrop, J. A., 31, 455 Blaustein, M. P., 301
Barman, T., 116 Blay, G., 54
Barnekow, F., 137 Block, S. M., 335, 337(79)
Barner, J., 10, 13(31), 24(31), 25(31) Boerner, W. M., 192
Barnes, J., 15, 32, 37(32) Bokvist, K., 220
Barsotti, R. J., 80, 88(4), 240, 301,310, 311, Boldac, J. M., 277
318, 323, 325, 328, 328(37), 340, 357, Bolsover, S., 374
376(1) Bon, S., 265, 272, 273(25)
Barth, A., 122, 226, 227,231,234(15), 235(15), Bond, M., 312
237(18), 240(15), 241(18) Bonhoeffer, T., 176
Bartlett, W., 10, 13(32), 15, 25, 25(32) Boone, J., 485
Barz, W. Z., 24 Borgstahl, G. E. O., 256
Baumstark-Kahn, C., 492 Borisy, G. G., 75
Bauschke, E., 136 Borlinghaus, R., 291,305(11, 50), 306
Bax, A., 90 Borman, J., 444
Bayer, E. A., 135, 135(1), 139, 152(1) Born, M., 179, 190(20), 194(20), 360
Bayley, H., 35, 96, 116(17), 117, 118, 119, 120, B6ttger, M., 474
121, 122, 122(21), 123, 123(17, 21, 22), Bottiroli, G., 152
128, 129, 131,132, 133(21), 134, 134(30), Bouet, F., 265, 273(6)
155, 266 Bourgeois, D., 256, 277
Baylor, D. A., 176, 416, 417, 430(14) Bourne, Y., 265, 273(9)
Beale, M. H., 474, 475, 476(5), 478, 480 Boyd, A. E., 77
Beaug6, L., 301 Boyer, P. D., 225
Beavo, J. A., 416 Bradford, M. M., 103, 112(31)
Bechem, M., 222 Braha, O., 134
Becker, A., 233, 300, 305(25) Braslavsky, S. E., 421
Becker, P. L., 219 Bratt, G. T., 403, 404(1)
Bendig, J., 415, 417, 419, 419(23), 420, 422, Braxmeir, T., 169
425, 427,428 Brenner, B., 31t, 313
Benning, M. M., 308 Breter, H., 136
Berchtold, H., 308 Brinton, C. C., 136
Berendzen, J., 256 Brintzinger, H., 232
Berg, P., 492 Brown, A. M., 220
Berger, C. L., 318, 335 Brown, E., 203,204(3), 356,358, 365, 365(13),
Bergot, B. J., 153 367, 368, 375(13), 377(13)
Berovic, N., 247, 250(2, 3) Brown, R. L., 417
Berridge, M. J., 381,396 Brubaker, M. J., 252
Bert, R. J., 417 Bruhnke, J., 118, 255, 266
Bertrand, D., 471,472(87a) Brundage, R. A., 95
Betz, H., 444 Brune, M., 318, 332
Beynon, T. D., 247, 250(2, 3) Brustovetsky, N., 300, 305(25)
B'hlen, P., 147 Buechler, J. A., 415
Billington, A. P., 30, 31, 32(11, 20), 41(11, 20, Burch, J. M., 179
56), 48, 48(11, 20), 139, 147, 150(30, 39), Burgess, W. H., 89, 91(10)
155, 266, 269(18b), 270(18b), 272(18b), Burke, P. M., 256
447(4), 449(4), 452(4), 453(7), Burmester, C., 95, 215, 251
453(7), 454, 456, 456(7), 457(7), 461, Burtnick, L. D., 98
468(7) Bus, H, 416
AUTHOR INDEX 499
Busby, S., 416 Chorongiewski, H., 223, 233(2)

Butt, H.-J., 289 Chow, R. H., 221
Buu-Hoi, N. P., 20 Christensen, B., 305(52), 306
Christensen, L. M., 153
Christy, R. W., 179
C Chu, G., 492
Chung, S., 463, 464(68), 466(68a)
Cabib, E., 269 Cifuentes, F., 241
Calder, G. M., 357 Clapper, D. L., 403, 404(1)
Callamaras, N., 380, 389, 390(26), 402 Clark, R. B., 220
Callaway, E. M., 30, 443, 467 Clark-Lewis, I., 134
Callender, R., 236 Claussen, M., 474
Cameron, J. F., 100, 155 Clegg, J. S., 492
Candia, O. A., 409 Clore, G. M., 90
Cannon, A. M., 152 Cocuzza, A. J., 136
Cantor, C. R., 135 Cody, W. L., 78
Capiod, T., 221,381 Coherent, Inc., 178
Cardona, M. L., 54 Cohn, M., 127
Caretta, A., 417 Cole, N. B., 78
Carpenter, B. K., 30, 31, 32, 32(20), 35(22), Cole, P. A., 133
41(31), 48(20, 22), 122, 139, 147, 150(30, Collins, A., 301
39), 155, 266, 269(18b), 270(18b), Colthup, N. B., 227, 242(16, 17)
272(18b), 439, 444, 453(7, 10, 12), 456, Conibear, P. B., 215, 335
456(7), 457(7, 8, 10), 468(7, 8) Conklin, E. J., 135, 135(3), 152(3)
Carraway, R. E., 94 Connolly, B. A., 225
Carter, T. D., 219, 381 Cook, C. S., 333
Cash, D. J., 46, 447, 448(27a), 450, 460(29) Cook, S. N., 134
Cavaggioni, A., 417 Cooke, R., 308, 309, 313, 324(22), 343
Cepus, V., 223, 243, 244(47) Cooper, E., 471, 472(87a)
Cerny, R. L., 137, 153(24) Corbin, J. D., 417
Chageux, J. P., 461 Cornwell, T. L., 415
Chaisson, E., 180 Corrie, J. E. T., 3, 7, 30, 101, 122, 139, 204,
Chanal, P., 265 220, 221, 223, 226, 227(5), 231(5),
Chang, C., 35 234(15), 235(25), 240(15), 252, 260,
Chang, C.-Y., 117, 118, 121, 123(17, 22), 129, 263(19), 306, 318, 321, 323, 324(52),
132, 155, 266 328(52), 330(52), 332, 343(52), 416,
Chao, J. L., 233 421(10), 455, 456, 456(52), 457(52)
Chapman, A. C., 229 Coull, J. M., 137, 150(25)
Chase, P. B., 313 Courtney, K. D., 278
Chatonnet, A., 265 Craik, J. S., 101,318
Chattopadhyay, A., 46 Cram, D. J., 10
Cheley, S., 123, 134 Crevel, I. M.-T. C., 310
Chen, C. X., 236 Crick, F., 443
Chen, Y., 256 Crivici, A., 90
Cheney, R. E., 307 Cross, R. A., 310
Chillingworth, R. K., 318, 332 Cruickshank, D. W. J., 265, 277(11)
Chimilio, L., 11 Cubitt, A. B., 77
Chittock, R. S., 245, 247, 250(2, 3), 483 Cummings, J. P., 68
Choi, I., 400, 400(13) Currie, K. P., 222
Choi, J., 382, 386(12) Cusaek, E., 123
Choquet, 457, 461(65) Cuzikowski, A. P., 155
500 AUTHOR INDEX
D Durden, D. S., 20
Durrant, I., 136
Dairman, W., 147 Dyer, D. H., 252, 277
Dalva, M. B., 30, 443, 467, 472(77) Dzeja, C., 415, 417, 419, 419(23), 420, 422,
Daly, J. W., 227, 242(16) 425, 427, 428
Dani, J. A., 457
Danley, L. E., 134
Dantzig, J. A., 95, 214, 225, 307, 309, 318, E
319, 320(21), 323, 324, 324(21), 325(54),
Eccleston, J. F., 417
328(54), 330(20), 331(21), 339(21)
Eckstein, F., 225, 417
Davis, B. A., 20
Edelson, R., 122
Dawson, B. A., 137, 153(22, 23)
Edeson, 461
Dawson, P. E., 134
Edge, M. D., 136
Deacon, G. B., 241
Edgerton, J., 485
De Boni, U., 349
Edidin, M., 78
DeGroot, C., 135
Edman, K. A. P., 343
Delaney, K. R., 467
Ehret-Sabatier, L., 265, 273(6), 277
Delcour, A. H., 455
Eichinger, L., 107
DeLuca, M., 245
Eigen, M., 447, 473(23)
DeMayo, P., 455
Eiler, J. J., 229
Denk, W., 30, 31, 32(20), 41(20), 48(20), 139,
Einstein, A., 179
147, 150(30, 39), 176, 179, 266, 269(18b),
Eisenberg, D., 101
270(18b), 272(18b), 357, 388,444, 453(7),
Eisenberg, E., 110, 309, 324
456(7), 457(7), 467, 468(7, 78), 469(78)
Eisenrauch, A., 301,305(44), 306
de Pont, J. J. H. H. M., 290, 305(5)
Eismann, E., 417, 426(24)
Desai, S., 135, 135(3), 152(3)
Eliasson, L., 220
Desantis, A., 220
Elliott, G., 134
Desel, H., 114
Ellis-Davies, G. C. R., 219, 221, 240, 241,
Devine, E. E., 471
241(36), 253, 301,310, 318, 323,324, 331,
de Weer, P., 295, 298
340, 357, 374, 376(1)
Dicky, D. M., 404, 407(9), 415(7)
Ellman, G. L., 278
Diffee, G. M., 241
Ellman, J. A., 118, 134, 266
Dilger, 461
Elsner, S., 471
DiPolo, R., 301
Engels, J., 58, 59, 416, 419, 421(29)
Ditmar, R. M., 233
Engert, F., 176
Dolphin, A. C., 222
Epel, D., 278, 282, 284, 286, 286(6), 287, 288
Doppler, T., 141
Epstein, W. W., 8
Dose, K., 305(52), 306
Escobar, A., 241, 357
D¢skeland, S. O., 417
Evan, G. I., 78
Drechsel, D., 335
Evans, F. E., 417
Drexel, D., 72
Evans, M. R., 136
DrCbak, B. K., 357, 478
Eysselein, V. E., 3
Dr6se, S., 305(46), 306
Drummond, R. M., 94
Duddy, S. K., 432 F
Dudley, G., 10, 13(31), 15, 24(31), 25(31),
32, 37(32) Fabiato, A., 325, 347(56)
Duke, E. M. H., 217, 219(27) Fabiato, F., 325, 347(56)
Duong, T. T., 139 Fawcett, M.-C., 117
Dupont, Y., 241 Fay, F. S., 78, 94, 95
AUTHOR INDEX 501
Fearon, K. L., 153 Fritz-Wolf, K., 308

Featherstone, M. R., 278 Froehlich, J. P., 291, 294(13), 297(13),
Feeney, J., 221,382, 443 298(13), 304, 305(12, 40)
Feldmeyer, D., 222 Frolow, F., 265, 273(5)
Fendler, K., 289, 290, 291, 293(8, 10), 294(9, Frunder, H., 279
13), 297(13), 298(9, 13), 299(8), 304, Fryer, M. W., 340
305(6, 8-10, 12, 40, 46), 306 Fujimori, T., 312
Ferenczi, M. A., 254, 318, 321, 323, 324(52), Fujiwara, A., 279
328(52), 330(52), 332, 334(70), 343, Fukui, K., 112
343(52) Fuldner, H. H., 225
Fernandez, J. M., 357 Funakoshi, T., 348, 350, 352(7), 354(7)
Ferraretto, A., 152 Funatsu, T., 318, 333, 335
Fersht, A., 447, 473(25) Furman, R. E., 417
Ferster, D., 463, 464(68), 466(68a) Furtak, T. E., 179
Fetscher, C. A., 247 Furuichi, T., 381,389(2)
Feuerstein, J., 225 Furuta, T., 50, 51, 52, 60(7), 61(7), 63(7, 9)
Fidler-Lim, N., 241 Futura, T., 8
Fielding, A., 325,441
Fill, M., 340, 357
Finer, J. T., 335
Fink, R. H. A., 214 G
Finkelstein, A., 136
Finn, F. M., 136 Gabriel, D., 95, 97, 155, 158(14), 162(14)
Finn, J. T., 416 Gadsby, D. C., 295, 298
Fischbach, G. D., 46, 439 Gait, M. J., 136
Fischmeister, R., 220 Galione, A., 396
Fister, T., 321,323 Garcia, M. B., 54
Fleet, A., 374 Garges, S., 416
Fletterick, R. J., 308 Garrossian, M., 8
Fodor, G., 24 Gartner, W., 291, 294(9), 298(9), 305(9)
Fodor, S. P. A., 98 Gasparro, F., 122
Foldman, A., 265 Gayen, S. K., 179
Forbush, B. III, 1(4), 3,130, 131(37), 134(37), Gee, K. R., 10, 13(31), 15, 24(31), 25(31), 30,
155, 168(2), 176, 233, 240(33), 247, 290, 31, 32, 34(27), 35(22), 37(32), 39, 42(21,
291(4), 293, 299(4), 307, 421,431,455 23), 43(21), 45(23), 48(21, 22), 49(21, 27),
Fork, R. L., 178, 364 63, 96, 104(19), 110(18), 111(18), 122,
Forster, A. C., 136 403, 404, 406, 407, 415(7, 13), 439, 444,
Frace, A. M., 220 453(8, 9,11 ), 456,457(8, 9,11), 458,468(8,
Franken, S. M., 260 9, 11)
Franklin-Tong, V. E., 357, 478 Gelfand, V. I., 75
Franks-Skiba, K., 343 Gelperin, A., 467
Franzini-Armstrong, C., 310, 318, 357 Genick, U. K., 256
Frazier, A. W., 230 Georg, H., 227, 237(18), 241(18)
Fr6chet, J., 100, 155 Gerisch, G., 109
Fricker, M. D., 478, 480 Gerrard, A., 179
Friedburg, E. C., 483 Gerwert, K., 223, 232, 233(2), 234, 241, 243,
Friedrich, T., 32, 289, 291(2), 305(51), 306, 244(47)
470, 471(85), 472(85) Ghosh, A. K., 139
Frings, S., 415, 417, 419, 419(23), 420, 422, Ghung-Yu, C., 96, 116(17)
425, 426, 427, 428 Gibbons, I. R., 308
502 AUTHORINDEX
Gilbert, S. H., 94, 95 Grewer, C., 30, 32, 41(4), 43(4), 46(4), 48(4),
Gildea, B. D., 137, 150(25) 96, 443, 446(38), 452, 454(38)
Giles, W. R., 220 Griensenbach, S., 492
Gilman, A. G., 444, 453(3) Griffiths, P. J., 357
Gilroy, S., 474 Gronenborn, A. M., 90
Ginsburg, S., 267 Gross, L. A., 77
Givens, R. S., 1, 2, 4, 5(16, 17), 6, 7(16, 17), Grunwald, M. E., 416
8, 8(14-17), 10, 11, 13(31, 32), 15, 17(16, Grynkiewicz, G., 323, 324
17), 18(16,17), 19(16, 17), 20(27), 21(27), Gu, Q.-M., 404
22(27), 24, 24(31), 25, 25(31, 32), 28(27), Guesdon, J. L., 135
32, 37(32), 50, 117, 122, 134(27), 306, 321, Gunther, J., 279
323, 456 Guo, X. Q., 219
Godde, M., 417, 426 Gurnack, M. E., 407
Godt, R. E., 327 Gurney, A. M., 3, 214
Goeldner, M., 265, 266, 267, 269(17, 19), Gurney, E. L., 230
270(17, 19, 20, 22), 271, 272(17, 19, 20), Gutfreund, H., 447, 473(27)
273(6, 22), 274, 275, 275(22), 276(22), Giath, K., 209, 212(7), 214, 253, 320, 387
277, 277(17) Gutweiler, M., 305(52), 306
Golan, R., 117 Guzikowski, A. P., 31, 456
Gold, G. H., 416 Gy6rke, S., 340, 357
Goldenberg, S. S. S., 349
Goldmacher, V. S., 118
H
Goldman, A., 265, 273(5)
Goldman, D. E., 445 Haas, A. L., 137, 153(23)
Goldman, Y. E., 95, 96, 176, 214, 225, 307, Hackney, D. D., 225, 309, 310
309, 310, 311, 312, 312(35), 318, 319, Hadfield, A. T., 217, 219(27), 252, 256(2)
320(21, 27), 323, 324, 324(21, 32), 325, Hagen, C., 423, 425, 427, 428
325(54), 327(32), 328(54), 330, 330(20, Hagen, V., 415, 417, 419, 419(23), 420
32), 331,331(21), 332, 333,334, 334(69), Hajdu, J., 252, 256(2), 266
339(21), 341(32), 343, 357, 383 Hamill, O., 450, 451(32), 452(32), 454(32),
Goldschmidt-Clermont, P. J., 74(17), 78, 98 460(32), 470(32)
Goldstein, L. S. B., 308 Hamm, H. E., 308
Goody, R. S., 95, 176, 204, 214, 215, 216, Hammes, G. G., 447, 456, 473(24)
217(23, 24), 243, 244(46, 47), 251, 253, Hammill, O. P., 44(52), 46
256, 257, 259(12), 260, 261(18), 263(19), Hancock, D. C., 78
308, 318, 319 Hanley, M. R., 432
Gtippert-Mayer, M., 357 Hansen, H. J., 141
Gordon, D. J., 101 Hao, J. P. Y., 323
Gordon, J. P., 181,364 Harada, I., 231, 232(25), 236(25), 240(25),
Gordon, P. R., 443 244(25)
Gorenstein, G., 58 Harada, Y., 333, 335, 337(76), 340(76)
Grace, M. R., 133 Hardie, R. C., 220
Graeff, R. M., 404, 407(9), 415(7) Harel, M., 265, 273(5-7)
Gravel, D., 155 Harootunian, A. T., 439
Gray, P. T., 381 Harris, A. J., 471
Greene, L. E., 309 Hartshorne, D. J., 91, 93(14)
Gregoriou, V. G., 233 Hartung, K., 289, 301,302(33), 303(33), 304,
Grell, E., 220, 290, 291, 293(8, 10), 299(8), 304(33), 305(40), 306
305(5, 8, 10, 44, 52), 306 Hartzell, H. C., 220
AUTHOR INDEX 503
Hasselbach, W., 306 Higuchi, H., 95, 214, 307, 318, 319, 330, 333,
Hata, Y., 217 334, 335, 336, 337(81), 340(81), 343(81)
Hatayama, M., 52, 63(9) Hijden, H. T. W. M. v. d., 290, 305(5)
Hatch, M. A., 267 Hilgemann, D. W., 289, 295(1), 298, 301,
Hatchard, C. G., 26 305, 471
Hatt, H., 417 Hilgenfeld, R., 308
Haubs, M., 291,293(8), 299(8), 305(8) Hiraoki, T., 241
Haugland, R. P., 67(5), 68, 69(5), 71(5), 77(5), Hirokawa, N., 348, 349, 350, 350(3), 352(3, 7,
118, 478 8), 353(3), 354(7)
Haugwitz, M., 101 Hirose, K., 318
Hauser, K., 122,226, 234(15), 235(15), 240(15) Hirth, C., 265, 273(6), 277
Hayakawa, H., 492 Hoard, D. E., 22
Hayat, M. A., 356 Hobbiger, F., 265
Hayes, R. N., 403, 404(1) Hobbs, A., 291, 294(13), 297(13), 298(13),
He, Z.-H., 318, 332 305(12)
Heathcock, C. H., 22 Hodota, C., 50
Heidecker, M., 95, 96, 97(18), 104(18), 109, Hoesch, R. E., 431
112, 114(18), 118, 121(18) Hoffman, J. F., 1(4), 3, 130, 131(37), 134(37),
Heim, R., 77 155, 168(2), 176, 233, 240(33), 247, 290,
Heinemann, C., 221 291(4), 299(4), 307, 421,431,455
Heinemann, S., 471 Hoffman, K., 136
Helliwell, J. R., 265, 277(11) Hoffmann, H., 222
Hellrung, B., 6 Hoffmann, W., 215
Henderson, R., 277 Holden, H. M., 308
Hendrix, J., 214 Holland, D., 136
Herbette, L., 175, 176, 293 Holman, J. P., 217
Herman, T., 137, 153(20-23) Holmes, K. C., 308
Herrmann, C., 116 Holzbaur, E. L. F., 308
Herzog, K., 229 Homsher, E., 214, 225, 309, 310, 318, 319,
Hess, B., 232 320(21), 323, 324(21), 331(21), 332,
Hess, G. P., 30, 31, 32, 32(11-13, 20), 34(27), 334(70), 339(21)
35(22), 41, 41(4, 11, 20, 31, 56), 42(21, Homsher, H., 318
23), 43, 43(4, 21), 44(42, 45), 45(23, 30, Hood, L., 153
42, 57), 46, 46(4, 42, 45), 47(42, 45), 48, Hooley, R., 474
48(4, 6, 11, 20-22), 49(21, 27), 96, 122, Horiuti, K., 214, 311, 312, 330
147, 150(30, 39), 155,266, 269(18a; 18b), Hoskins, B. K., 357
270(18a; 18b), 272, 272(18a; 18b), 439, Houslay, M. D., 416
443, 444, 445, 446(5, 38, 53), 447, 447(4, Hryschko, L. V., 302
5, 33, 53), 448(27a; 33), 449(4), 450, 451, Hubel, D., 465
451(33, 34), 452, 452(4, 33), 453(7-12), Hursthouse, M. B., 7, 260
454, 454(33, 38), 455, 456, 459(33), Hyman, A. A., 72, 335
460(29, 33), 467, 468(7-9, 11, 12, 78),
469(78), 470, 471(85), 472(85), 473(5)
Hessling, B., 223, 233(2), 234
Heyse, S., 305(48), 306 Ikebe, M., 78, 90, 93(14), 94
Hibberd, M. G., 95, 96,176,307, 309,311,318, Ikebe, R., 90, 91
319, 323, 324, 324(32), 325(54), 327(32), Ikura, M., 90
328(54), 330(20, 32), 331, 332, 334(69), Ilyin, V., 219, 382, 395(9), 396
341(32), 383 Inoue, Y., 335, 336, 337(81), 340(81), 343(81)
504 AUTHOR INDEX
Ireland, R. E., 21 Kaiser, R., 153

Irisawa, H., 301 Kalow, W., 278
Irving, M., 318 Kamimura, A., 31
Ishibashi, M., 50 Kanai, Y., 350, 352(8)
Ishihara, A., 39 Kandel, E. R., 443, 445(2)
Ishijima, A., 318, 333, 335 Kang, H., 374, 378, 380(22)
Ishikawa, T., 50 Kao, J. P. Y., 323, 324, 357, 421, 431, 432,
lto, K., 59 433(8), 435, 436(8), 439
lversen, L. L., 463 Kaplan, J. H., 1(4), 3, 130, 131(37), 134(37),
Iverson, R. M., 287 155, 168(2), 176, 233, 240, 240(33), 241,
Ivorra, I., 95, 382, 385,386, 392, 393,393(14), 241(36), 247, 253, 290, 291(4), 299(4),
395, 395(21), 396, 397, 398, 401 301,307, 310, 318, 323, 324, 331,340, 357,
Iwamura, H., 50 374, 376(1), 421,431,455
Iwamura, M., 8, 50, 51, 52, 60(7), 61(7), 63(7, Kappl, M., 301,302(33), 303(33), 304(33)
9), 416 Karlin, A., 445, 446(18a; 18b)
Karpen, J. W., 176, 416, 417, 430(14)
d Kass, G. E., 432
Katayama, Y., 220, 323
Jack, W. E., 134 Kato, H., 217, 218
Jackson, 461 Katsube, Y., 217
Jackson, J. B., 247, 250(2, 3) Katz, B., 46, 446(28), 447
Jackson, S. A., 492 Katz, L. C., 30, 443, 467, 472(77)
Jacobsen, K., 39 Kaupp, U. B., 415,417, 419,419(23), 420, 422,
Jaffe, E. K., 127 425, 426, 426(24), 427, 428
Jager, A., 59 Kawai, M., 313
Jagla, A., 215 Kellogg, D., 72, 335
Jakes, K. S., 136 Kemp, B. E., 91, 93(14), 123
Jaruschewski, S., 291, 294(13), 297(13), Kemp, D. S., 50
298(13), 305(12) Kemp, J. A., 463
Jastorff, B., 417 Kent, S. B. H., 134
Jessel, T. M., 443, 445(2) Kestner, T., 407, 415(13)
Johaszova, M., 301 Keszthelyi, L., 293, 296(17)
John, J., 215, 217(23), 225, 243, 244(46), Keyes, R. W., 182
257, 261(18) Khan, S., 219
Johns, R. B., 34 Khodakhah, K., 219, 220, 381
Johnson, K. A., 447, 473(26) Kibak, H., 288
Johnson, K. W., 31,155, 456 Kihara, T., 175, 293
Johnson, L. N., 217, 219(27), 265, 277(11) Kimura, A., 217
Johnston, L. J., 10 Kimura, J., 301
Jones, R. L., 474 Kinosita, K., 96, 117, 155, 156(10), 158,
Jonges, G. N., 278 158(10), 161(17), 162(17)
Jovin, T. M., 157, 168(16), 171(16) Kirschner, M. W., 348
Jung, A. H., 1,10,11,13(32), 15, 24, 25, 25(32) Kirstgen, R., 168
Juorio, A. V., 20 Kitazawa, T., 312
Klee, C. B., 90
K Klein, M. V., 179
Kleinfeld, D., 467
Kabsch, W., 95,216,217(24), 256, 259(12), 308 Kleinle, J., 368
Kagawa, K., 214 Kleywegt, G. J., 265, 273(7)
Kahn, M., 33 Klingauf, J., 375
AUTHOR INDEX 505
Klingenberg, M., 298, 300, 305(25) Kusianowicz, J., 134

Knaupp, U. B., 416 Kuzmic, P., 32, 455
Knight, A. R., 463 Kyogoku, Y., 229
Knight, P., 162
Kn61ker, H. J., 169 L
Kobayashi, S., 312
Koch, K.-W., 417 Lacktis, J., 309, 318, 319, 320(21), 323,
Kodama, T., 112 324(21), 331(21), 339(21)
Koechner, W., 178, 179(13), 195(13), 198(13) Ladouceur, G., 155
Koenigs, P., 118, 218, 255, 266 La Du, B. N., 277
Koga, N., 50 Laflamme, M. A., 219
Kogelnik, H., 188 Lamb, G. D., 340
Kojima, H., 318, 333, 335 Lambert, J. M., 96, 118, 141, 155, 156(13),
Kolb, A., 416 175(13)
Kolb, H.-A., 445 Lamm, G. M., 136
Kometani, K., 112 Lancet, D., 417, 426(24)
Kono, E., 318 Landau, V. G., 47, 451
Korn, 457, 461(65) Landegren, U., 153
Korn, E. D., 101 Lando, L., 221
Korth, M., 416 Langer, P. R., 136
Koshland, D. E., 277 Langosh, D., 444
Koshland, D. E., Jr., 252 Liinnergren, J., 215
Koster, H., 137, 150(25) LaNoue, K., 298
Kovacs, O., 24 Lanzetta, P. A., 409
Koyama, Y., 50 Lappi, S., 273
Kozlowski, R. Z., 220, 301 Lattman, E. E., 308
Krafft, G. A., 68 Lau, R., 308
Kr~imer, R., 298 Lau, W., 89, 91(10)
Krause, E., 417, 419, 419(23), 420, 422, 425, Laube, B., 444
427, 428 L~iuger, P., 291,305(11, 12, 47), 306, 445
Krejci, E., 265 Lautwein, A., 260, 263(19)
Krengel, U., 308 Lavit, D., 20
Kreutz, W., 227, 231,237(18), 241(18) Lea, T. J., 310, 323
Krishnasastry, M., 134 Lederer, W. J., 3, 301
Krishtal, O. A., 46, 450, 452(31), 453(31) Lee, A. K., 221
Kron, S. J., 110, 335 Lee, H. C., 403,404,404(1, 2), 406, 407,407(9),
Krzymanska-Olejnik, E., 135, 138, 154, 156 411,415(7, 13)
Kubert, M., 153 Lee, J., 39
Kucherov, V. F., 20 Lefever, E., 137, 153(21)
Kueper, L. W. III, 2, 4, 5(16, 17), 7(16, 17), LeGrand, A., 256
8(15-17), 10, 13(31), 15, 17(16, 17), Leikauf, E., 137
18(16, 17), 19(16, 17), 24(31), 25(31), 32, Leimbruger, W., 147
37(32), 117, 321,323 Lenart, T. D., 310, 318, 357
Kuhn, H. J., 421 Lend, C., 461
Kulba, M., 20 Lentfer, A., 95, 216, 217(24), 256, 259(12)
Kull, F. J., 308 Leonard, D. M., 78
Kumler, W. D., 229 Leslie, A. G. W., 308
Kurahashi, T., 416 Lester, H. A., 3, 214, 416, 418(13), 419(13)
Kurzchalia, T. V., 136 Lester, W., 485
Kushmerick, M. J., 313 Leterrier, F., 272, 273(251
506 AUTHOR INDEX
Levich, A. G., 47, 451 Makhatadze, G. I., 216

Levitsky, D. O., 302 Malh6, R., 478
Lev-Ram, V., 358, 369(12), 374(12), 376(12) Mallavarapu, A., 63, 96, 104(19), 110(18),
Levy, J., 137, 153(20) 111(18)
Lewis, S. M., 3 Mandel, G., 444
Lewitzki, E., 220 Mandelkow, E., 215
Leytus, S. P., 170 Mandveno, A., 215
Li, H., 30, 467, 468(78), 469(78) Mangel, W. F., 170
Li, T., 188 Mann, D., 98
Lidzey, D. G., 247, 250(2, 3) Mannherz, H. G., 308
Lien-Vien, D., 227, 242(17) Manor, D., 236
Lifshitz, E. M., 47, 451 Mantele, W., 122, 226, 227, 231, 234(15),
Lim, S. S., 75 235(15), 237(18), 240(15), 241(18)
Lin, M., 343 Marchot, P., 265, 273(9)
Lincoln, T. M., 415 Marell, A. E., 228
Lindberg, U., 108 Margalit, T., 417, 426(24)
Lindsay, H. A., 278 Margulis, M., 431,441
Ling, N., 318 Markham, K. R., 34
Lipp, P., 301, 368 Marque, J., 31, 32(20), 41(20), 48(20), 139,
Lippincott-Schwartz, J., 78 147, 150(30, 39), 266, 269(18b), 270(18b),
Liu, 461 272(18b), 444, 453(7), 456(7), 457(7),
Lockery, S., 468, 469(81) 468(7)
Lockhart, A., 310 Marr6, E., 480
Lockridge, O., 265, 277 Marriott, G., 95, 96, 97(16, 18), 98(16), 99(16),
Locktis, J., 225 104(18), 109, 110(15), 112, 114(18), 117,
Long, L. M., 140 118, 121(9, 18), 156(10), 155, 157,
Long, W., 10 157(12), 158, 158(10, 14), 161(17),
Longhurst, R. S., 179 162(14, 17), 168(12, 16), 171(12, 16),
Lorentzon, T., 290 173(12), 175(12)
Lough, J., 137, 153(22, 23) Marston, S., 324
Loughrey, H., 152 Martell, A. E., 325
Lowe, G., 416 Martin, H., 80, 88(4), 311,328, 328(37)
Lu, A. T., 98 Martinez, O. E., 364
Lu, Z., 3, 310 Marty, A., 450, 451(32), 452(32), 454(32),
Ludwig, J., 417, 426(24) 460(32), 470(32)
Lukas, T. J., 89, 91(10) Marty, E., 44(52), 46
Luo, Y., 313 Maruyama, J., 59
Luschner, C., 368 Marx, A., 215
L•then, H., 474 Masserini, M., 152
Lutter, R., 308 Massouli& J., 265, 272, 273(25)
Lwebuga-Mukasa, J., 273 Mathias, L. J., 36, 71(16), 78
Lymn, R. W., 309 Mathivanan, N., 10
Matsubara, I., 333
Matsubara, N., 30, 31, 32(11, 20), 41(11, 20,
M 56), 48, 48(11, 20), 139, 147, 150(30, 39),
266, 269(18b), 270(18b), 272(18b), 444,
Madsen, 461 447(4), 452(4), 453(7), 454, 456(7), 457(4,
Maeda, Y., 214, 253, 318, 319 7), 461,468(7)
Maher, L. J. III, 137, 153(24) Matsuoka, S., 301, 302
Maiman, T. H., 182 Matsuura, M., 90
AUTHOR ~NDEX 507
Mattson, G., 135, 135(3), 152(3) Milford, F. J., 179

Matuszewski, B., 4, 5(16, 17), 7(16, 17), 8(14, Millar, N. C., 214, 225, 309, 310, 318, 319,
16, 17), 17(16, 17), 18(16, 17), 19(16, 17), 320(21), 323, 324(21), 331(21), 339(21)
50, 321,323 Millard, P. J., 39
Maughan, D. W., 327 Miller, J. P., 417
Maydell, M. R. D., 214 Miller, W. T., 117, 266
Mayer, M. L., 439 Minta, A., 323, 324
Mazid, M. A., 7, 260 Misiura, K., 136, 150(12)
McCarthy, M. P., 444 Mitchison, T. J., 63, 64, 66, 68(4), 72, 74(17),
McConkey, D. J., 432 75, 78, 95, 96, 98, 104, 104(19), 110(18),
McConnaughie, A. W., 8 111(18), 155, 335, 348, 350(1)
McCray, J. A., 3, 31, 32(20), 33, 41(20), 43, Mitra, A., 33
43(33), 46(33, 43), 48(20), 88, 95, 96, 130, Miyata, H., 96, 117,155,156(10), 158, 158(10),
131(38), 134(38), 139, 147, 150(30, 39, 161(17), 162(17)
40), 155, 175, 176, 203, 204(1, 2), 218(1), Miyawaki, A., 381,389(2)
219, 223, 225, 226, 226(9), 231(4, 9), Mizani, S. M., 298
233(4), 241,247,253,266, 269(18a; 18b), Moffat, K., 256, 277
270(18a; 18b), 271, 272(18a; 18b), 293, Moisescu, D. G., 327, 347
296(16), 307, 320, 323, 324(46), 325, Mokrin, S. C.
347(56), 405, 406(10), 443, 444, 453(7), Moloney, M. G., 8
455, 456, 456(7), 457(7), 468(7), 486 Momotake, A., 52, 63(9)
McDonald, K. S., 357 Montibeller, J. A., 136
McElroy, W. D., 245 Moore, G. A., 432
Mclnnes, J. L., 136, 152(8) Mooseker, M. S., 307
McKee, S. P., 139 Mooser, G,, 273
McKinnon, D., 444 Morgensen, S., 135, 135(3), 152(3)
McMillan, S., 180 Morrat, K., 277
McNamee, M. G., 46, 445, 446(18a) Morrison, H., 117, 121(1), 124, 134(1)
Meaking, W. S., 485 Moss, R. L., 3, 241,310, 311, 313, 357
Means, A. R., 91, 93(14) Muir, T. W., 134
Meldrum, R. A., 245, 247, 279, 483, 485, Mulhern, S. A., 110
486(2), 489(2) Mtiller, F., 417
Melhado, L. L., 170 Mulligan, I. P., 310, 323
Melzer, W., 222 Muralidharan, S., 222
Mendel, D., 118, 266 Murata, H., 231, 232(25), 236(25), 240(25),
Menetret, J. F., 215 244(25)
Menini, A., 416 Murray, J. M., 318, 357
Mercer, T., 24 Murray, S., 155
Mertz, J., 370 Muslin, A. J., 123
Mery, P. F., 220 Muto, E., 319, 335, 336, 337(81), 340(81),
Messenger, J., 220 343(81)
Meyer, B., 114
Meyer, T., 381,439
Michikawa, T., 381, 389(2) N
Mikoshiba, K., 381,389(2)
Milburn, T., 31, 32(20), 41(20), 48(20), 139, Nagel, G., 32, 289, 291, 291(2), 293(10),
147,150(30, 39), 266, 269(18b), 270(18b), 305(10, 51), 306, 470, 471(85),472(85)
272(18b), 444, 453(7), 456(7), 457(7), Naim, M., 117
458, 468(7) Nair, H. K., 265,269
Miledi, R., 382, 384, 388, 392 Nara, M., 241
508 AUTHOR INDEX
Naraghi, M., 374, 378, 380(22) Oldershaw, K. A., 384

Nargeot, J., 416, 418(13), 419(13) Olejnik, J., 135, 138, 154, 156
Neher, E., 44(52), 46, 48, 48(48), 221, 374, Olsen, S. R., 123
375, 378, 380(22), 445, 448(30), 450, Ono, K., 416
454(19), 460(19) Ordoukhanian, P., 117
Neis, A. S., 444, 453(3) O'Reilly, N. J., 78
Nerbonne, J. M., 3, 30, 117, 121(4), 214, 222, Orrenius, S., 432
416, 418(13), 419(13), 455 Osawa, T., 50, 51
Neveu, D., 221 Ostap, E, M., 3, 318, 335
Ng, K., 256 Oswald, R. E., 32, 470, 471(85), 472(85)
Niblack, B., 35, 96, 116(17), 117, 118, 123(17), Otis, T., 32
129, 155, 266 Ott, D. G., 22
Nieoll, D. A., 302 Ottl, J., 95, 155, 158(14), 162(14)
Nielander, G. W., 135, 135(3), 152(3) Oxford, A. W., 7, 31, 40(16)
Niggli, E., 3, 301,368
Nishioka, T., 217
Niu, L., 30, 31, 32, 32(12, 13), 34(27), 35(22),
41(31), 42(23), 45(23, 30), 48(22), 49(27),
P
122, 439,444, 446(5, 36), 447(4, 5), 449(4),
452,453(8, 9,10,12), 454(38), 456, 457(8- Pabst, R., 305(52), 306
10, 12), 458, 459, 463, 464(69), 468(8, 9, Pai, E. F., 95, 215, 216, 217(23, 24), 243,
12), 470, 471(85), 472(85), 473(5) 244(46), 256, 257, 259(12), 260, 261(18),
Noegel, A. A., 101, 107 263(19), 308
Noel, J. P., 308 Palmer, R. A., 233
Noma, A., 301 Pan, P., 117, 119, 120, 122(21), 123(21), 128,
Noren, C. J., 134 133(21), 266
Nosek, T. M., 327 Panchal, R., 121, 123, 123(22), 132
Nunn, D. L., 384 Papageorgiou, G., 456
Nunn, D. S., 31 Paranjape, S., 439
Pardee, J. D., 73, 101
O Parekh, A. B., 471, 472(87b)
Park, C.-H., 1, 10, 13(32), 24, 25, 25(32), 122,
Oberhauser, A. F., 357 134(27), 456
O'Brien, P. J., 432 Park, S.-H., 8, 15, 20(27), 21(27), 22(27),
O'Brien, R. D., 278 28(27)
Ochoa, E. L. M., 46 Parker, C. A., 26
O'Connor, T. P., 348 Parker, I., 95, 219, 380, 381,382, 384, 385, 386,
Oda, J., 217 386(12), 388, 388(10), 389, 389(5, 11),
Oefner, C., 265, 273(5) 390(26), 392, 393, 393(14), 395, 395(9,
Oesterhelt, D., 291, 294(9), 298(9), 305(9) 21), 396, 397, 398, 400, 400(13), 401,402
Offer, G., 162 Parsons, T. D., 221, 241,357
Ogden, D. C., 219, 220, 221, 381 Pashkevich, K. I., 69
Ogreid, D., 417 Patchornik, A., 31, 98, 117, 155
O'Hare, P., 134 Patchornik, P., 455
Okabe, S., 348, 349, 350, 350(3), 352(3), Pate, E., 309, 313, 324(22), 343
352(8), 353(3) Patel, J. R., 241,357
Olbrich, A., 168 Patel, R., 219
Olbricht, W. L., 43, 44(45), 46(45), 47(45), Patlak, J., 46, 48(48), 448(30), 450
450, 451(34) Patton, C., 325
AUTHORINDEX 509
Patzlaff, M., 24 Q
Paulsen, P. J., 242
Paulus, G. G., 176 Oiu, X. O., 136
Pavlickova, L., 32, 455 Quinn, D. M., 265, 267(2), 269, 273(2)
Pawley, J. B., 390 Quirk, S., 308
Pearson, R. B., 123 R
Peck, A. W., 265
Peckham, M., 318 Raaben, R. J., 83
Pedro, J. P., 54 Rademacher, J., 273
Peeters, J. M., 83 Raible, A. M., 153
Peng, L., 265, 266, 267, 269(17, 19), 270(17, Raizen, D. M., 467, 468, 469(61, 80, 81)
19, 20, 22), 271,272(17, 19, 20), 273(22), Rakowski, R. F., 295, 298
274, 275, 275(22), 276(22), 277(17) Rameriz, F., 19, 28(35)
Petratos, K., 95, 216, 217(24), 218, 255, 256, Ramesh, D., 31, 32, 35(22), 41(31), 45(30),
259(12) 48(22), 122, 439, 444, 453(8, 10, 12), 456,
Petricdvic, V., 179 457(8, 10), 459, 468(8)
Petsko, G. A., 95, 118, 217(24), 218, 255, 256, Rammelsberg, R., 223, 233(2)
259(12), 308 Randal, T. W., 444, 453(3)
Pfaff, E., 298 Rapoport, T. A., 136
Philipson, K. D., 302 Rapp, G., 95, 202, 204, 209, 212(7), 214, 215,
Phillips, G. N., Jr., 256 216, 217(23, 24), 243, 244(46), 253, 256,
Phillips, R. J., 241 257, 258(18a), 259(12), 261(18), 318, 319,
Pidoplichko, V. I., 46, 450, 452(31), 453(31) 320, 387
Pieles, U., 136 Rapp, R., 357
PiUai, V. N. R., 1, 98 Ratcliffe, R., 22
Pirrung, M. C., 7, 8(21), 31, 98 Ravelli, R. B. G., 265, 273(7)
Piston, D. W., 179 Rayment, I., 308
Planck, M., 445 Read, J. L., 98
Plant, P. J., 31,455 Reason, C., 485
Plunkett, S. E., 233 Recek, J., 50
Podein, R. J., 404 Rees, B. B., 288
Pohl, B., 222 Reese, C. B., 8
Pollard, T. D., 308 Reese, T. S., 335
Pon, R. T., 136, 151(10) Regnier, M., 318
Poole, K. J. V., 204, 214, 253, 318, 319 Rehwinkel, H., 168
Porter, N. A., 118, 218, 255, 266 Reich, J. G., 279
Portnoy, D. A., 74(17), 78, 98 Reid, G. P., 3, 7, 33, 43(33), 46(33), 59, 68,
Postovskii, I. Y., 69 86, 88, 130, 131(38), 134(38), 147,
Potma, E. J., 332 150(40), 176, 225,226, 226(9), 231(9, 13),
Potter, B. V. L., 439 247, 260, 263(19), 293, 296(16), 318, 321,
Potter, J. D., 215 323, 324(52), 325, 328(52), 330(52),
Powell, T. A., 220, 301 343(52), 347(56), 405, 406(10), 486
Pradervand, C., 256, 277 Reinach, P. S., 409
Prassler, J., 109 Reiser, C. O. A., 308
Pratt, A. J., 8 Reisler, E., 110, 112(36)
Prendergast, F. G., 89, 91(10) Reitz, J. R., 179
Priestley, T., 463 Ren, Z., 256, 277
Privalov, P. L., 216 Renstr0m, E., 220
Purich, D. L., 456 Reshetnikova, L., 308
510 AUTHOR INDEX
Rettinger, J., 471 Sawyer, T. K., 78

Reynolds, J., 445, 446(18b) Schafer, F. P., 179
Rhodes, Ch. K., 179 Schalk, I., 265, 273(6), 277
Richard, S., 416, 418(13), 419(13) Schaper, K., 3l, 32, 34(27), 42(23), 45(23),
Ried, G. P., 271 49(27), 444, 453(9), 456, 457(9), 468(9)
Rieger, D , 101 Schawlow, A. L., 182
Ringe, D., 118, 218, 255 Scheidig, A., 215, 243, 244(47), 251, 260,
Robinson, I. M., 357 263(19)
Rodionov, V. I., 75 Schildkamp, W., 256, 277
Rogalski, A. A., 135 Schiltz, E., 225
Romberg, L., 310 Schlaeger, E. J., 58
Root, D. D., 110 Schlechtingen, G., 169
Rorsman, P., 220 Schleicher, M., 101, 107
Rosch, P., 225 Schlichting, I., 95, 215, 216, 217(23, 24), 225,
Rosenblatt, J,, 74(17), 78, 98 243, 244(46), 256, 257, 259(12), 261(18)
Roskoski, R., 130 Schmid, H., 141
Rossi, F. M., 431,432, 433(8), 436(8) Schmidt, C. F., 335, 337(79)
Rothschild, K. J., 135, 138, 154, 156 Schmidt, R., 421
Rottenberg, H , 298 Schmidt-B~ise, K., 308
Roudna, M., 204 Schmieden, V., 444
Rozycki, M., 108 Schmirmer, N. K., 308
Ruf, H., 220 Schmitt, F., 80, 88(4)
Rundstrom, N., 444 Schnabel, W., 40, 43(40, 41), 46(40, 41)
Rungger, D., 471,472(87a) Schnapp, B. J., 335, 337(79)
Russo, M. J., 123, 134(30) Schneider, S., 226, 237(14)
Rypniewski, W. R., 308 Schnyder, T., 308
Schoenberg, M., 324
Schoffeniels, E., 445
S
Schofield, P., 455
Sablin, E. P., 308 SchrOder, R. R., 215
Sabry, J., 348 Schroeder, J. I., 474
Sachs, G., 290 Schubert, D., 471
Sakmann, B., 44(52), 46, 48, 48(48), 445, Schulman, H., 273
448(30), 450, 451(32), 452(32), 454(19, Schultz, P. G., 118, 134, 266
32), 460(19, 32), 470(32) Schupp, H., 40, 43(40, 41), 46(40, 41)
Sakoda, T., 330 Schutt, C. E., 108
Sakurna, K., 50 Schwartz, J. H., 443, 445(2)
Sakurada, K., 335, 337(76), 340(76) Schwartz, S., 39
Salmon, E. D., 75 Schwartz, W., 471
Salser, S., 72, 335 Schwarze, W., 445
Sanchez-Llorente, A., 260, 263(19) Sciaky, N., 78
Sanders, J., 153 Scott, R. H., 222
Sandison, D., 373 Scott, W. G., 277
Sano, T., 135 Seifert, R., 426
Savage, M. D., 135, 135(3), 152(3) Self, C. H., 117
Sav6ant, J.-M., 10 Self, S. A., 188
Sawatari, A., 30 Senter, P. D., 96, 118, 141, 155, 156(13),
Sawin, K. E., 63, 64, 72, 96, 104(19), 110(18), 175(13)
111(18), 335 Seravalli, J., 269
Sawyer, D. T., 242 Serebryakov, E. P., 20
AUTHOR INDEX 511
Shai, R., 117, 134(8) Sproat, B. S., 136

Shall, S., 247, 279, 485, 486(2), 489(2) Spudich, J. A., 73, 101, 110, 335
Shao, Y., 134 Spudich, J. L., 219
Shaw, A. S., 123 Srajer, V., 256, 277
Shaw, P. J., 357 Sreekumar, R., 78, 94
Shear, J., 358, 365, 365(13), 367, 368, Steffen, P., 72, 335
375(13), 377(13) Steger, E., 229
Sheehan, J. C., 7, 31, 40(16), 456 Steglich, W., 168
Sheidig, A., 95 Stein, T., 147
Shi, Q.-Y., 44l Stempel, K. E., 225
Shim, S. B., 8 Stengelin, M., 290, 305(6)
Shimanouchi, T., 229 Stephenson, D. G., 340
Shimkus, M., 137, 153(20, 21) Stepinska, M., 91, 93(14)
Shioya, T., 30l Stiekema, F., 135
Shofield, P., 31 Stienen, G. J. M., 332
Shoger, R. L., 279 Still, W. C., 33
Shuey, S. W., 7, 8(21) Stocker, S., 109
Shuster, M., 444 Stoddard, B. U, 118, 218, 252, 255, 277
Siebert, F., 227, 233, 237(18), 241(18) Strickler, J., 176, 357, 373
Siede, W., 483 Stroud, R. M., 444
Siegmen, A. E., 179 Strowbridge, B. W., 467
Sigler, P. B., 308 Stryer, L., 98, 176, 381,416, 417, 430(14), 439
Sigman, D. S., 273 Sttihmer, W., 471,472(87b)
Sigworth, F. J., 44(52), 46, 450, 451(32), Stults, J. T., 153
452(32), 454(32), 460(32), 470(32) Stiirmer, W., 305(47, 48), 306
Sih, C. I., 404 Suck, D., 308
Silman, I., 265, 267, 270(22), 273(5-7, 22), Sudati, F., 152
274, 275, 275(22), 276(22) Sugimoto, M., 8, 50, 52, 60(7), 61(7), 63(7,
Simmons, R. M., 311,312(35), 333, 335, 343 9), 416
Singer, P., 277 Sulston, J. E., 468
Singer, S. J., 135 Sumikawa, K., 382, 384, 392
Skingle, D. C., 136 Suslova, L. M., 20
Slatin, S. L., 136 Sussman, J. L., 265, 267, 270(22), 273(5-7,
Sleep, J. A., 116, 254, 321 22), 274, 275, 275(22), 276(22)
Smirnoff, P., 117 Sutton, W. R., 68
Smith, C. L., 78, 135 Suva, R. H., 417
Smith, R., 308 Svelto, O., 179
Smith, R. M., 228, 325 Svensson, E. C., 318, 319, 335
Smith, S. J., 474 Svoboda, K., 335, 337(79), 388
Snape, W. J., Jr., 3 Swain, K. E., 64
Solas, D., 98 Sweet, R. M., 256, 277
Somlyo, A. P., 214, 311,312 Swezey, R. R., 280, 282, 283(5), 284, 286,
Somlyo, A. V., 214, 311,312 286(6), 288
Sonar, S., 138 Symons, R. H., 136, 152(8)
Soria, M. R., 152
Soucek, M., 32, 455 T
Soukup, G. A., 137, 153(24)
Souvignier, G., 232, 234 Takeda, S., 350, 352(7), 354(7)
Spoors, J. A., 117 Takeuchi, H., 231, 232(25), 236(25),
Sprinzl, M., 308 240(25), 244(25)
512 AUTHOR INDEX
Tallec, A., 10 Tipler, P.A., 180

Tanaka, D., 31 Tirrell, D.A., 117
Tanaka, H., 220 Toker, L., 265, 273(5)
Tanaka, J. C., 417 Tokuda, K.,50
Tanaka, M., 90 Tolman, R. C., 179
Tang, C. I., 178 Tomchick, D.R., 308
Tang, C.-M., 431, 439, 441 Toner, M., 123, 134(30)
Tank, D. W., 467 Torigai, H., 8, 50, 51, 52, 60(7), 61(7), 63(7,
Tanner, J. W., 123, 330 9), 416
Tanokura, M., 241 Townes, C.H., 181, 182
Tansey, M. J., 96, 141, 155, 156(13), 175(13) Toyoshima, Y. Y.,335
Taran, C., 233 Travers, F., 116
Tardivel, R., 10 Trentham, D. R.,3,7,30,33,43,43(33),46(33,
Tardy, C., 10 43, 44), 59, 68, 86, 88, 95, 96, 122, 130,
Tasumi, M., 241 131(38), 134(38), 139, 147, 150(40), 155,
Taylor, A. W., 230 175, 176, 203, 204, 204(1), 218(1), 219,
Taylor, C. W., 384 220, 221, 223, 225, 226, 226(9), 227(5),
Taylor, E. W., 309 231(4,5, 9,13),233(4),234(15),235(15),
Taylor, J.-S., 117 240(15),247, 252,260,271,293,296(16),
Taylor, P., 265, 273, 273(9), 444, 453(3) 307,309,311,312,318,319,320,321,323,
Taylor, S. S., 415 324,324(32,46,52),325,325(54),328(52,
Telford, J. N., 278 54),330(20,32,52),331,332,334(69,70),
Temsamani, J., 153 341(32), 343(52), 347(32, 56), 382, 383,
Teng, T.-Y., 256, 277 405, 406(10),416, 418,421(10), 422(28),
Ten Kortenaar, P.B.W., 83 443, 455,456(52), 457(52), 486
Terasaki, M., 78 Trewavas, A. J., 474, 478, 480
Tesser, G. I., 83 Tripathi, R. K., 278
Theriot, J. A., 63, 64, 66, 68(4), 74(17), 78, 95, Troullier, A., 223, 241
96, 98, 104(19), 110(18), 111(18), 155,348 Troutman, H. D., 140
Thesleff, S., 46, 446(28), 447 Trussell, L. O., 32, 46, 439
Thieleczek, R., 347 Tschisgale, M., 279
Thio, L. I., 46 Tsien, R. Y., 30, 77, 78, 95, 117, 155,223, 323,
Thirlwell, H., 318, 321, 324(52), 328(52), 324,358,365,365(13),367,368,369(12),
330(52), 343(52) 374(12), 375(13), 376(12), 377(13), 384,
Thirlwell, T., 254 416, 431, 439, 455
Thomas, D., 432 Tsuboi, M., 229
Thomas, D. D., 3, 318, 319, 330, 335, Tsukita, S., 318
337(76), 340(76) Tsutsumi, A., 241
Thomas, D. T., 3 Twist, V. W., 220
Thomas, J. L., 117
Thomas, P., 221
Thomason, D. B., 247
Thompson, S., 117 U
Thompson, W. J., 139
Thomson, J. N., 467, 468(79) Uden~iend, S., 147
Thorson, J., 325 Udgaonkar, J. B., 31, 32(20), 41, 41(20), 43,
Tibor, R. S., 417 44(42, 45), 45(42),46(42,45),47(42, 45),
Tijssen, P., 136 48(20), 139, 147, 150(30, 39), 266,
Till, U., 279 269(18b), 270(18b), 272(18b), 444,
Tilney, L. G., 78 447(33), 448(33), 450, 451(33, 34),
AUTHOR INDEX 513
452(33), 453(7), 454, 454(33), 456(7), 46(33), 48(20), 59, 68, 78, 80, 86, 94, 130,
457(7), 459(33), 460(33), 468(7) 131(38), 134(38), 139, 147, 150(30, 39,
Uhl, R., 114 40), 176, 225, 226, 226(9), 231(9, 13), 247,
Uhler, M. D., 123 266, 269(18a; 18b), 270(18a; 18b), 271,
Uhmann, W., 233 272(18a; 18b), 293,296(16), 310, 323,325,
Ulbrich, C., 223 347(56), 382, 405, 406(10), 443, 444,
Ulman, A., 117 453(7), 456, 456(7), 457(7), 468(7), 476,
Umeyama, T., 350, 352(8) 486
Umezawa, J., 7 Wallimann, T., 308
Unoue, Y., 319 Wallmark, B., 290
Unwin, N., 445, 446(17, 18) Walmsley, I. A., 178
Ursby, T., 256, 277 Walseth, T. F., 403, 404, 404(1), 407, 407(9),
Uyeda, T. Q. P., 335 411, 415(7)
Walsh, C. T., 133
Walstrom, K. M., 451
Walter, U., 415
V Wang, S.-H., 122
Wang, S. S., 84, 381
Valdemanis, J. A., 178
Ward, D. C., 136
Vale, R. D., 308, 310
Ward, J. L., 474, 475, 476(5), 478, 480
Valera, S., 471,472(87a)
Ward, J. M., 474
Vallee, R. B., 308
Warmuth, R., 220
Vanaman, T. C., 98
Watkins, P. A. C., 478
Van Duk, B. G., 83
Watterson, D. M., 89, 91(10)
Van Noorden, C.J.F., 278
Wayner, D. D. M., 10
Vasilets, L. A., 471
Webb, M. R., 318, 332, 334(69)
Vazquez, R. W., 32, 470, 471(85), 472(85)
Webb, W. W., 31, 32(20), 41(20, 56), 48,
Veigel, C., 214
48(20), 139, 147, 150(30, 39), 176, 178,
Velez, P., 357
179, 203,204(3), 266, 269(18b), 270(18b),
Vergara, J., 357
272(18b), 356,357,358,365,365(13), 367,
Vergara, J. L., 241
368, 370, 373, 375(13), 377(13), 444,
Vigny, M., 272, 273(25)
453(7), 456(7), 457(7), 468(7)
Viktorova, T. S., 69
Weber, J., 10, 13(32), 15, 25(32)
Vilaseca, R., 179
Weber, J. F. W., 1, 11, 25
Vinson, C,, 308
Weiboldt, R., 444, 447(4, 5)
Viola, R. W., 31, 155
Weichselbaum, R. R., 492
Weigele, M., 147
Weill, C. L., 445, 446(18a)
W Weiner, J., 417
Weiss, C. O., 179
Wagner, N., 305(52), 306 Weiss, J. N., 302
Wakatsuki, S., 217, 219(27) Wells, K., 373
Walba, D. M., 21 Weng, G., 236
Waldrop, A. A., 136 Wensel, T., 381
Walker, B., 35, 96, 116(17), 118, 123(17), 129, Wesenberg, G., 308
134, 155 Westbrook, G. L., 439
Walker, G. C., 483 Weyand, I., 417
Walker, J. E., 308 Wharton, C. W., 245, 247, 250(2, 3), 279, 483,
Walker, J. M., 3, 88, 88(4) 485, 486(2), 489(2)
Walker, J. W., 3, 31, 32(20), 33, 41(20), 43(33), Wheat, H., 463, 464(68), 466(684)
514 AUTHORINDEX
Whitaker, M., 219 Xu, C., 370, 373

White, D. C. S., 325 Xu, T., 374, 378, 380(22)
White, H. D., 318, 335 Xue, J., 321, 323
White, J. G., 468 Xue, J.-y., 4, 5(16, 17), 7(16, 17), 8(16, 17),
Wiberley, S. E., 227, 242(16, 17) 17(16, 17), 18(16, 17), 19(16, 17)
Wieboldt, R., 31, 32, 35(22), 41(31), 42(21),
43(21), 45(30), 48(21, 22), 49(21), 96, 122,
439, 444, 446(5), 453(8, 10-12), 456,
457(8, 10-12), 458, 468(8, 11, 12), 473(5) Y
Wiedmann, M., 136
Yamada, K., 214
Wiegand-Steubing, R., 214
Yamaguchi, H., 217
Wier, W., 389, 390(26), 402
Yamakawa, M., 311
Wiesel, 465
Yamanaka, G., 417
Wilchek, M., 135, 135(1), 139, 152(1)
Yanagida, T., 319, 330, 333,335, 336, 337(76,
Wilcox, M., 31, 155
81), 340(76, 81), 343(81)
Wilding, M., 219
Yang, J., 247
Willenbucher, R. F., 3
Yan-Marriott, Y., 112
Williams, R., 373
Yanovsky, V. P., 179
Willner, I., 117, 134(8)
Yao, Y., 219, 382, 386(12), 388(10), 389(11),
Wilson, I. B., 267, 269
395(9), 396, 400, 400(13)
Wilson, K., 95, 216, 217(24), 256, 259(12)
Yariv, A., 196, 198
Wilson, R. M., 7, 31, 40(16)
Yasumasu, I., 279
Winkelmann, D. A., 308
Yau, K.-W., 416
Wirz, J., 6, 267, 269(19), 270(19, 20),
Yazawa, M., 241
272(19, 20)
Yonemoto, W., 415
Wise, F. W., 178, 179
Yu, L. C., 311
Wittinghofer, A., 95, 215, 216, 217(23, 24),
Yuste, R., 467
225, 243, 244(46), 256, 257, 259(12), 260,
261(18), 263(19), 308
Wolf, E., 179, 190(20), 194(20), 360
Wolf, M. R., 357 Z
Wong, E. H. F., 463
Wong, J. G., 221 Zahn, D., 279
Wong, W. K., 40, 43(40), 46(40) Zang, R., 215
Wood, S. W., 136 Zechel, K., 157, 168(16), 171(16)
Woodruff, G. N., 463 Zehavi, U., 98, 117, 155
Woodward, R. B., 31,455 Zeiger, H. Z., 181
Wootton, J. F., 43, 46(44), 221, 222, 323, Zernike, F., 197
418, 422(28) Zhang, S., 32
Wordeman, L., 72, 335 Zhao, J., 215
Wormmeester, J., 135 Zhao, Y., 313
Wright, E. M., 219 Zhu, G., 90
Wu, C. W., 456 Zhu, Q. Q., 40, 43(41), 46(41)
Wuddel, I., 305(47-49), 306 Ziegler, P., 21
Wulff, M., 256, 277 Zimmerman, A. L., 176, 416, 417, 430(14)
Zimmermann, W., 136
X Z611ner, P., 222
Zoramski, C. Z., 46
Xie, Y. N., 3 Zucker, R. S., 205, 212(6), 221, 301
Xiong, X., 134 Zwierzak, A., 20
SUBJECT INDEX 515
Subject Index
A labeling with tetramethylrhodamine

iodoacetamide, 157-158
A23187, 1-(4,5-dimethoxy-2-nitrophenyl)
labeling with thiol-reactive caged rhoda-
ethyl carboxylate synthesis, 39-40
mine reagent, 170-171, 173
AAC, s e e ADP, ATP carrier
G-actin, 6-nitroveratryloxycarbonylchlo-
Abscisic acid, caged
ride caged conjugate
biological function after photolysis,
fluorescence photoactivation micros-
480-483
copy, 104-105
calcium response measurement with fluo-
rescent probes, 478, 481-482 functiona! analysis of caged and de-
injection into plant cells, 477-478 caged proteins, 105-107
photolysis conditions, 478-480 labeling ratio, 101-102
structure, 475 photoactivation conditions, 104
synthesis, 475-477 polymerization competence, 102-103
Acetic acid, caged compound preparation, 101
pH changes using photolysis, 246, tetramethylrhodamine labeling of
248-249 caged protein, 103
synthesis, 247-248 labeling with fluorescent caged com-
Acetylcholinesterase pounds, 72-74
caged ligands ADP, p-hydroxyphenacyl caged compound
carbamylcholine, 266-267, 269, synthesis, 22-24
273-276 ADP, ATP carrier
choline, 266-267, 269, 273-276 electrophysiological measurements with
criteria for ideal compounds, 273 caged ATP
inhibition analysis, 272-274 bilayer system, 298-299
kinetics of photolysis, 273 time-resolved measurements, 300-301
noracetylcholine, 266-267, 269 transport modes, 300
norbutyrylcholine, 267, 269 physiological function, 298
photochemical properties, 269-272 Affinity labeling, s e e Biotin affinity label-
photolysis conditions, 277-278 ing; Photocleavable affinity tags
recovery of activity following photoly- /3-Alanine
sis, 275-276, 278 2-methoxy-5 -nitrophenyl carboxylate
physiological function, 265 caged compound
ACHE, s e e Acetylcholinesterase kinetics of photolysis, 40-41
Actin, s e e a l s o Muscle fiber quantum yield determination, 43, 45
cross-linking with photocleavable re- synthesis, 38-39
agents receptor, s e e Neurotransmitter receptor
caged G-actin complex preparation, y-Aminobutyric acid
173-175 a-Carboxy-o-nitrobenzyl carboxylate
cross-linking with 4-bromomethyl-3-ni- caged compound
trobenzoic acid succinimide ester, kinetics of photolysis, 43
159-162 synthesis, 35-37
516 SUBJECT INDEX
desyl caged compound, 10-11 binding of caged compounds, 293

p-hydroxyphenacyl caged compound, patch clamp studies, 295-298
10-11 photolysis conditions, 291, 293
receptor, s e e Neurotransmitter receptor time-resolved measurements, 293-294,
Aminodextran, thiolation with 2-iminothio- 297-298
lane, 157 Azid-1, calcium uncaging efficiency by two-
ATP, caged photon photolysis, 367-369, 379
ADP, ATP carrier, electrophysiological
measurements with caged ATP
bilayer system, 298-299 B
time-resolved measurements, 300-301
transport modes, 300 Beam radius, Gaussian beam optics, 187
calcium affinity, 32l, 324, 346-347 Benzoin phosphate
dideoxy ATP in DNA repair kinetic limitations in photochemical applications,
analysis 7
analytical high-performance liquid chro- photolysis properties, 6
matography, 487-488 Biotin affinity labeling, s e e also Photocleav-
apparatus design for studies, 491-492 able affinity tags
DNA damage and analysis, 490, 494 applications, 135-136, 152
panic response, 494-495 avidin affinity, 136
photolysis conditions, 488 biotinylation of macromolecules, 136
purification, 487 release of target molecule, 136-137
stability in mammalian cells, 489-490 Birefringence, Gaussian beam optics,
synthesis of high specific activity ot-32p 194-195
compound, 486-487 Bovine serum albumin, thiolation
efficiency of photoconversion, 341-342 2-iminothiolane reaction, 157
p-hydroxyphenacyl caged compound thiol content, determination, 157
photolysis conditions, 28-29 8-Bromo-cyclic AMP
properties and applications, 8 4,5-dimethoxy-2-nitrobenzyl caged com-
stability testing, 29 pound
luciferase activation for bioelectronic im- calibration of flash-induced changes,
aging, 246-248, 250 427, 429-430
magnesium affinity, 321, 324, 346-347 concentration optimum determination,
muscle fiber photolysis studies 425
actomyosin studies design, 417-418
ATPase assays, 332-333 gated channel activation studies,
ATP binding, 330-332 426-427
caged calcium studies, 332 handling of solutions, 425
cross-bridge detachment, 330, 334, photochemical properties, 421-422
344-346 purity requirements, 420-421,430
relaxation of tension, 330, 334 solubility, 418, 421
muscle shortening velocity, inhibition solvolysis, 422-423
before photolysis, 343-346 synthesis, 418-420
photochemical conversion, quantification, phosphodiesterase resistance, 416-417
338-341 8-Bromo-cyclic GMP
sodium, potassium-ATPase, electrophysio- 4,5-dimethoxy-2-nitrobenzyl caged com-
logical measurements with caged pound
ATP calibration of flash-induced changes,
bilayer experiments, 291-295 427, 429-430
SUBJECT INDEX 517
concentration optimum determination, ATPase, see Sarcoplasmic reticulum/endo-

425 plasmic reticulum calcium ATPase
design, 417-418 caged, see Calcium, caged
gated channel activation studies, fluorescent probe measurement
426-427 apparatus for simultaneous electrophys-
handling of solutions, 425 iological measurement, 388-391
photochemical properties, 421-422 response to calcium release by caged
purity requirements, 420-421, 430 compounds
solubility, 418, 421 cyclic ADP-ribose, 410-413
solvolysis, 422-423 2 , 5 - d i ( t e r t - b u t y l ) h y d r o q u i n o n e caged
synthesis, 418-420 compound, 435-437, 439
phosphodiesterase resistance, 416-417 inositol 1,4,5-triphosphate, 388-391,
4-Bromomethyl-3-nitrobenzoic acid succini- 395, 403
mide ester nicotinic acid adenine dinucleotide
photochemical properties, 159 phosphate, 414-415
reactivity, 159 plant hormones, 478, 481-482
synthesis, 156-159 signaling, see Cyclic ADP-ribose; Inositol
BSA, see Bovine serum albumin 1,4,5-triphosphate; Nicotinic acid ade-
BuChE, see Butyrylcholinesterase nine dinueleotide phosphate; Sarco-
Butyrylcholinesterase plasmic reticulum/endoplasmic reticu-
caged ligands lum calcium ATPase
carbamylcholine, 266-267, 269, Calcium, caged
273 -276 flash lamp photolysis studies with caged
choline, 266-267, 269, 273-276 compounds, 219-222
criteria for ideal compounds, 273 sodium, calcium-exchanger, DM-ni-
inhibition analysis, 272-274 trophen caged calcium studies
kinetics of photolysis, 273 electrophysiologicalmeasurements,
noracetylcholine, 266-267, 269 302-306
norbutyrylcholine, 267, 269 photolysis kinetics, 301
photochemical properties, 269-272 rate constant determinations, 305-306
photolysis conditions, 277-278 regulatory site studies, 301-302,
recovery of activity following photoly- 304-305
sis, 275-276, 278 photochemical conversion, quantification,
physiological function, 265 339-340
two-photon photolysis
apparatus, 361-365
C applications, 356-357, 380
calcium yield determination, 358,
Caffeine, calcium release studies with caged 369-374
inositol 1,4,5-triphosphate, 393, 395 temporal behavior of calcium concen-
Caged compounds, see also specific c o m - tration distribution, 374-379
pounds and functional groups uncaging action cross section determi-
advantages in substrate release studies, 1, nation, 365-369
30, 64, 95-96 cAMP, see Cyclic AMP
criteria in design, 1-2, 64, 120 Carbamylcholine
signal transduction studies, see Calcium caged compound studies with crystallized
Calcium cholinesterases, 266-267, 269,
affinity for caged compounds, 321, 324, 273-276
346-347 receptor, see Neurotransmitter receptor
518 SUBJECT INDEX
a-Carboxy-o-nitrobenzyl carboxylates structural determination, 59

3,-aminobutyric acid fluorescent properties, 50-51
kinetics of photolysis, 43 Cross-linking, heterobifunctional photo-
synthesis, 35-37 cleavable reagents
applications, 31-32 actin
suitability for neurotransmitter study caged G-actin complex preparation,
with cell-flow, 48-50 173-175
Cell-flow, caged neurotransmitter character- cross-linking with 4-bromomethyl-3-ni-
ization trobenzoic acid succinimide ester,
channel opening, rate and equilibrium 159-162
constant determination, 460 labeling with tetramethylrhodamine
instrumentation, 46-47, 451-453 iodoacetamide, 157-158
quantification of photolysis, 454 labeling with thiol-reactive caged rhoda-
suitability of compounds, determination, mine reagent, 170-171, 173
48-49 applications, 155-156, 175
theory, 47, 447, 450-451 bovine serum albumin
validation, 47-48 thiolation with 2-iminothiolane, 157
cGMP, s e e Cyclic GMP thiol content, determination, 157
Choline, caged compound studies with crys- 4-bromomethyl-3-nitrobenzoicacid succin-
tallized cholinesterases, 266-267, 269, imide ester
photochemical properties, 159
273-276
reactivity, 159
CNB, s e e a-Carboxy-o-nitrobenzyl carbox-
synthesis, 156-159
ylates
criteria in design, 162
4-Coumarinylmethyl group, phosphate com-
1-(3,4-dimethoxy-6-nitrophenyl)-2,3-
pound caging
epoxypropyl chloroformate
cyclic AMP derivatives
reactivity, 168
4-(7-acetoxycoumarinyl) methyladeno- synthesis, 165
sine derivative 1-(3,4-dimethoxy-6-nitrophenyl)-2,3-
properties, 51-52 epoxypropyl hydroxide
synthesis, 51-54, 58 absorption spectroscopy, 167-168
biological activities, 62-63 reactivity with thiol groups, 168
4-(7-hydroxycoumarinyl)methyladeno- synthesis, 165
sine derivative N-1-([3,4-dimethoxy-6-nitrophenyl)-2,3-
properties, 51-52 epoxypropyl)oxy]carbonyl)4-(N,N-
synthesis, 51-53, 56, 58 dimethylamino)pyridinium, 3-oxo-
4-(7-methoxycoumarinyl)methyladeno- 2,5-diphenyl-2,3-dihydrothiophene
sine derivative 1,1-dioxide 4-hydroxide
properties, 51-52 caged G-actin complex preparation,
synthesis, 50-54, 58 173-175
photochemical and spectral properties, solubility, 169
60-62 synthesis, 166
4-(7-propionyloxycoumarinyl)meth- 1-(3,4-dimethoxy-6-nitrophenyl)-2,3-
yladenosine derivative epoxypropylsuccinimidylcarbonate,
properties, 51-52 synthesis, 166-167
synthesis, 51-53, 55, 58 1-hydroxyl-l-(3,4-dimethoxy-6-nitro-
purification, 59 phenyl)-2-propene, synthesis, 164
silver(l) oxide promoted condensation, 4-[1-(3,4-dimethoxy-6-nitrophenyl)-2,3-
51, 56-58 epoxypropyl-1 -oxycarbonyloxy]-3-
stability in dark, 60 oxo-2,5-diphenyl-2,3-dihydrothio-
SUBJECT INDEX 519
phene 1,1-dioxide Cyclic AMP-dependent protein kinase

reactivity, 168-169 caging at essential cysteine, 129-130
rhodamine, preparation of thiol-reac- thiophosphorylation of peptides, 125
tive caged compound, 169-170 Cyclic GMP
synthesis, 165-166 hydrolysis-resistant analogs, see 8-Bromo-
Cyclic ADP-ribose cyclic GMP
caged 1-(2-nitrophenyl)ethyl compound targets in signaling, 415-416
calcium release analysis with fluores- Cysteine, caged protein conjugates
cent probes, 410-413 caging of in vitro translated protein,
photolyis conditions, 409-410 128-129
synthesis, 405-407 cyclic AMP-dependent protein kinase,
uncaging efficiency determination, 411 129-130
calcium release in sea urchin eggs, engineering for site modification, 118,
403-404 134
synthesis, 404 heavy meromyosin
Cyclic AMP functional analysis
4-coumarinylmethy derivatives ATPase assay, 110, 114
in vitro motility assay, 110, 114-116
4-(7-acetoxycoumarinyl)methyladeno-
sine derivative photoactivation
properties, 51-52 duration of irradiation, 112-114
synthesis, 51-54, 58 transient kinetics analysis, 113-116
preparation
biological activities, 62-63
4,5-dimethoxy-2-nitrobenzyl bromide
4-(7-hydroxycoumarinyl) methyladeno-
conjugation, 111-112, 114-115
sine derivative
materials, 97
properties, 51-52
purification of protein, 111-112
synthesis, 51-53, 56, 58
a-hemolysin R104C
4-(7-methoxycoumarinyl)methyladeno-
modification with 2-nitrobenzyl group,
sine derivative 121,128-129
properties, 51-52 photolysis conditions, 130-133
synthesis, 50-54, 58 kemptide
photochemical and spectral properties, modification with 2-nitrobenzyl group,
60-62 124-125
4-(7-propionyloxyeoumarinyl)meth- photolysis conditions, 130-132
yladenosine derivative mass spectrometry analysis, 119, 124
properties, 51-52 2-nitrobenzyl reagents, efficiency in prepa-
synthesis, 51-53, 55, 58 ration, 120-121
purification, 59
silver(I) oxide promoted condensation,
51, 56-58
stability in dark, 60
structural determination, 59
D
desyl caged compound Desyl group
photolysis conditions, 27 carboxylate ester compound caging
properties and applications, 4-6 3,-aminobutyric acid, 10-11
synthesis, 19-20 applications, 13-14, 16
hydrolysis-resistant analogs, see 8-Bromo- L-glutamate
cyclic AMP properties and applications, 10-11,
targets in signaling, 415-416 13-14
520 SUBJECT INDEX
synthesis, 25-26, 37-38 2,2'-Dinitrobenzhydryl carboxylate

photochemistry, 11 applications, 32
phosphate compound caging N-methyl-n-aspartate, synthesis, 33-35
cyclic AMP DMNB, s e e 4,5-Dimethoxy-2-nitrobenzyl
photolysis conditions, 27 carboxylate
properties and applications, 4-6 DM-nitrophen
synthesis, 19-20 calcium uncaging efficiency by two-pho-
desyl dihydrogen phosphate synthesis, ton photolysis, 367-369, 379
17-19 cation chelation, 240-241
limitations, 7 Fourier transform infrared spectroscopy
photolysis, 4, 6 difference spectra of photolysis,
photolysis of caged compounds, 26-27 241-242
2,4'-Dihydroxyacetophenone, synthesis, 24 photolysis reaction, 241
4,5-Dimethoxy-2-nitrobenzyl bromide, s e e sodium, calcium-exchanger, caged cal-
Cysteine, caged protein conjugates cium studies
4,5-Dimethoxy-2-nitrobenzyl carboxylate electrophysiological measurements,
applications, 31 302-306
1-(3,4-Dimethoxy-6-nitrophenyl)-2,3- photolysis kinetics, 301
epoxypropyl chloroformate rate constant determinations, 305-306
reactivity, 168 regulatory site studies, 301-302,
synthesis, 165 304-305
1-(3,4-Dimethoxy-6-nitrophenyl)-2,3- DMNPE, s e e 1-(4,5-Dimethoxy-2-nitrophe-
epoxypropyl hydroxide nyl)ethyl carboxylate
absorption spectroscopy, 167-168 DNA repair
reactivity with thiol groups, 168 caged radioactive nucleotides in kinetic
synthesis, 165 studies
N-l-([3,4-Dimethoxy-6-nitrophenyl)-2,3- advantages and disadvantages, 485
epoxypropyl)oxy]carbonyl)4-(N,N- apparatus design for studies, 491-492
dimethylamino)pyridinium, 3-oxo-2,5- dideoxy ATP
diphenyl-2,3-dihydrothiophene 1,1-diox- analytical high-performance liquid
ide 4-hydroxide chromatography, 487-488
caged G-actin complex preparation, high specific activity a_3zp compound
173-175 synthesis, 486-487
solubility, 169 purification, 487
synthesis, 166 DNA damage and analysis, 490, 494
4-[1-(3,4-Dimethoxy-6-nitrophenyl)-2,3- electroporation of compounds, 485-
epoxypropyl-l-oxycarbonyloxy]-3-oxo- 486, 492-493
2,5-diphenyl-2,3-dihydrothiophene 1,1- panic response, 494-495
dioxide photolysis conditions, 488
reactivity, 168-169 stability in mammalian cells, 489-490
rhodamine, preparation of thiol-reactive thymidine triphosphate synthesis, 489
caged compound, 169-170 pathways, overview, 483
synthesis, 165-166 single cell versus multicellular organisms,
1-(3,4-Dimethoxy-6-nitrophenyl)-2,3- 484-485
epoxypropylsuccinimidyl carbonate, DNB, s e e 2,2'-Dinitrobenzhydryl carbox-
synthesis, 166-167 ylate
1-(4,5-Dimethoxy-2-nitrophenyl)ethyl car- Dynamic bleaching
boxylate conditions causing bleaching, 184
A23187, synthesis, 39-40 critical intensity estimation, 184
applications, 31 theory, 182-183
SUBJECTINDEX 521
Dynein tubulin, s e e Tubulin

caged compounds in photolysis studies types and structures, 64-66, 475
calcium chelators, 321, 324 Fluorescence microscopy, s e e Fluorescence
nucleotides, 321,325 photoactivation imaging
purity requirements, 321, 324-325 Fluorescence photoactivation imaging
table of properties, 322-323 cameras, 74
mechanism of motion, 308, 310 G-actin studies, 104-105
photolysis applications, 310 green fluorescent protein, 77-78
light sources, 74-75
E Fourier transform infrared spectroscopy
caged GTP photolysis
Electron microscopy, caged fluorescein-la- band assignment by isotopic labeling,
beled tubulin, 352-353, 355-356 237-240
Electroporation, caged substrates, 283-285 difference spectra, 231
Enkephalin, s e e Leucine-enkephalin dithiothreitol effects on spectra,
233-234
F intermediates in decay, 234-235
magnesium effects on spectra, 232
Flash lamp photolysis photolysis reaction, 231
applications synthesis of compounds, 225
calcium flux in response to caged com- caged phosphate photolysis
pounds, 219-222 band assignment by isotopic labeling,
chemotaxis, serine effects on bacteria, 227-229
219 difference spectra, 226-229
microtubule dynamics, 215 magnesium effects on spectra, 229-231
muscle contraction, 214-215, 320 DM-nitrophen photolysis
driving circuit, 207-209 cation chelation, 240-241
energy of pulse, measurement, 213-214 difference spectra, 241-242
lamp selection, 204-205, 207 photolysis reaction, 241
optics instrumentation for photolysis studies,
lenses, 210-212 223-224, 244
light guides, 213 Ras studies with caged GTP
mirrors, 209-210 data collection, 224-225
refractive optics versus ellipsoidal mir- difference spectra, 243-245
rors, 212-213 time-resolved measurement
pulse duration, 222 data collection techniques, 232-233
X-ray crystallography temporal resolution, 223
heat damage of crystals by flash, Frequency doubling, theory, 196-198, 201
216-218 FTIR, s e e Fourier transform infrared spec-
measuring compounds released by pho- troscopy
tolysis, 219
Ras GTPase studies in crystals, G
215-216
Flash photolysis, s e e Flash lamp photolysis; G-actin, s e e Actin
Laser photolysis GABA, s e e y-Aminobutyric acid
Fluorescein, caged compounds Gaussian beam optics
caging group selection, 67 beam radius, 187
practical considerations for use, 67 birefringence, 194-195
protein labeling, 72-74 focusing, 187
synthesis, 68 Rayleigh range, 189
522 SUBJEC'F INDEX
reflection, 192-193 Green fluorescent protein, fluorescence pho-

refraction, 194 toactivation imaging, 77-78
Self's equations, 188-190 GTP, caged
waist position, 188, 190 Fourier transform infrared spectroscopy
waist radius magnification, 189 of photolysis
wave polarization, 190-191 band assignment by isotopic labeling,
GFP, see Green fluorescent protein 237-240
Gibberellin, caged difference spectra, 231
calcium response measurement with fluo- dithiothreitol effects on spectra,
rescent probes, 478, 481-482 233-234
injection into plant cells, 477-478 intermediates in decay, 234-235
photolysis conditions, 478-480 magnesium effects on spectra, 232
structure, 475 photolysis reaction, 231
synthesis, 475-476 Ras studies, 224-225, 243-245
Glucose 6-phosphate, caged compound stud- synthesis of compounds, 225
ies in vivo synthesis of compounds, 225
electroporation of caged substrate into
sea urchin eggs, 283-285 H
glucose 6-phosphate dehydrogenase
assay, in vivo, 285-286 Heavy meromyosin, caged protein con-
fertilization of sea urchin eggs, effect jugate
on activity, 286-288 functional analysis
inertness testing of caged compound, ATPase assay, 110, 114
282-283 in vitro motility assay, 110, 114-116
photolysis conditions, 285 photoactivation
synthesis of caged compound duration of irradiation, 112-114
2-nitroacetophenone hydrazone prepa- transient kinetics analysis, 113-116
ration and reaction with glucose 6- preparation
phosphate, 281 4,5-dimethoxy-2-nitrobenzylbromide
radiolabeling, 280 conjugation, 111-112, 114-115
L-Glutamate materials, 97
desyl caged compound purification of protein, 111-112
properties and applications, 10-11, a-Hemolysin R104C
13-14 modification with 2-nitrobenzyl group,
synthesis, 25-26, 37-38 121, 128-129
p-hydroxyphenacyl caged compound photolysis of caged protein, 130-133
properties and applications, 10-11, 14, HMM, see Heavy meromyosin
16 1-Hydroxyl-l-(3,4-dimethoxy-6-nitrophe-
synthesis, 24-25 nyl)-2-propene, synthesis, 164
o-nitromandelyloxycarbonyl caged com- p-Hydroxyphenacyl group
pound carboxylate ester compound caging
electrophysiology studies of glutamate y-aminobutyric acid, 10-11
receptor effects, 440-443 applications, 13-14, 16
photochemistry, 440 L-glutamate
synthesis, 434-435 properties and applications, 10-11,
receptor, see Neurotransmitter receptor 14, 16
Glycine receptor, see Neurotransmitter re- synthesis, 24-25
ceptor peptides, 11-13
Glycogen phosphorylase, photolysis studies photochemistry, 11
in crystals, 252-253 phosphate compound caging
SUBJECT INDEX 523
ADP coupling in synthesis, 22-24 photolysis conditions, 478-480

ATP structure, 475
photolysis conditions, 28-29 synthesis, 475-476
properties and applications, 8
stability testing, 29 K
p-hydroxyphenacyl phosphate
photolysis, 28 Kainate receptor, see Neurotransmitter re-
synthesis, 20-22 ceptor
photolysis reactions, 8-10 Kemptide
photolysis of caged compounds, 26-27 modification with 2-nitrobenzyl group,
124-125
photolysis of caged peptide, 130-132
I
thiophosphorylation and modification,
Infrared spectroscopy, see Fourier trans- 125-126
form infrared spectroscopy Kinesin
Inositol 1,4,5-triphosphate ATP-binding core, 308
caged compound caged compounds in photolysis studies
applications, 381-382 calcium chelators, 321, 324
calcium fluorescent probe measure- nucleotides, 321,325
ment with electrophysiological purity requirements, 321, 324-325
measurement, 388-391, 395, 403 table of properties, 322-323
kinetics of calcium release, 395-397 mechanism of motion, 307-310
light sources for photolysis, 386-388 motility assay
linearity and calibration of photore- apparatus for combined photolysis stud-
lease, 384-386 ies, 336-337
photochemistry, 382-383 bead assay for force measurement, 338
spatial control and heterogeneity of cal- photolysis applications, 310
cium release sites, 397-402 preparation from bovine brain, 335
X e n o p u s oocyte studies translocation inhibition by caged ATP,
caffeine studies, 393, 395 346
dose-response relationship of cal-
cium release, 391-392 L
InsP4 agonist studies, 393
microinjection, 383-384 Laser photolysis
mediation of calcium release, 380-381, caged carboxylates, 40
404 characteristics of laser types, 203, 319
Ion pump, see also Calcium; Sodium, cal- comparison to flash lamp systems,
cium-exchanger; Sodium, potassium- 203-204
ATPase components of basic laser, 185-186
electrophysiological measurements with dynamic bleaching
caged substrates, 289-290 conditions causing bleaching, 184
overview of caged substrate studies, 290, critical intensity estimation, 184
305-306 theory, 182-183
frequency doubling, theory, 196-198, 201
Gaussian beam optics
d
beam radius, 187
Jasmonic acid, caged birefringence, 194-195
calcium response measurement with fluo- focusing, 187
rescent probes, 478, 481-482 Rayleigh range, 189
injection into plant cells, 477-478 reflection, 192-193
524 SUBJECT INDEX
refraction, 194 functional analysis of caged and de-

Self's equations, 188-190 caged proteins, 105-107
waist position, 188, 190 labeling ratio, 101-102
waist radius magnification, 189 photoactivation conditions, 104
wave polarization, 190-191 polymerization competence, 102-103
group velocity dispersion, 199-200 preparation of 6-nitroveratryloxycarbo-
mode locking, theory, 198-199 nyl chloride conjugate, 101
Q-switching, production, 195 tetramethylrhodamine labeling of
safety caged protein, 103
fail-safe protection, 186 inactivation of proteins, 98-99
goggles, 183 photolytic mechanism, 98
voltage supply, 186 preparation of 6-nitroveratryloxycarbonyl
single-photon photolysis chloride conjugates
development, 175-176 4-dimethylaminopyridine salt prepara-
laser systems, 176, 201-202 tion, 100
spatial resolution, 359 materials, 97
stimulated emission, theory, 179-182 quantification of conjugation, 99-101
two-photon photolysis, s e e Two-photon profilin
photolysis functional analysis of caged protein,
Laue diffraction, s e e X-ray crystallography, 108-110
photolysis studies preparation of 6-nitroveratryloxycarbo-
Leucine-enkephalin, photocleavable biotin nyl chloride conjugate, 108
labeling recombinant protein purification,
avidin affinity purification, 144-145, 107-108
148-149
conjugation, 144
fluorescamine assay, 145
photocleavage kinetics, 146-148 M
LSM1
caged tyrosine peptide Magnesium
design of inactive caged peptide, 88-90 affinity for caged ATP, 321, 324, 346-347
kinetics of photolysis, 87-88 effects on caged GTP infrared spectra,
myosin light-chain kinase inhibition, 229-231
93-94 Mass spectrometry, 2-nitrobenzyl caged pro-
spectral and photochemical properties, tein analysis, 119, 124
85-87 Meromyosin, s e e Heavy meromyosin
synthesis, 80-85 2-Methoxy-5-nitrophenyl carboxylate
structure and function, 79-80 /3-alanine
Luciferase kinetics of photolysis, 40-41
activation by caged ATP, 246, 248, 250 quantum yield determination, 43, 45
bioelectronic imaging and pattern recogni- synthesis, 38-39
tion applications, 246-247, 250 applications, 32
immobilization in agarose gels, 249-250 N-Methyl-D-aspartate, 2,2'-dinitrobenzhy-
inhibition by caged acetate pH effects, dryl carboxylate synthesis, 33-35
246, 248-249 Microtubule, s e e a l s o Tubulin
reaction catalyzed, 245 motility assay
Lysine, caged protein conjugates apparatus for combined photolysis stud-
G-actin ies, 336-337
fluorescence photoactivation micros- gliding assay, 337
copy, 104-105 preparation from tubulin, 335
SUBJECT INDEX 525
MK-801, neurotransmitter receptor inhi- ionic strength, 325-327

bition nucleotide depletion, 328
dissociation constant determinations, Myosin, see also Muscle fiber
464-465 ATP-binding core, 308
effects on opening and closing rates, caged compounds in photolysis studies
463-464 calcium chelators, 321, 324
M N P , see 2-Methoxy-5-nitrophenyl carbox- nucleotides, 321,325
ylates purity requirements, 321,324-325
Mode locking, theory, 198-199 table of properties, 322-323
Muscle fiber, see also Actin; Heavy mero- mechanism of motion, 307-309
myosin; Myosin photolysis applications, 310
apparatus for photolysis experiments
light sources, 315, 319-320
setups, 313, 315 N
biophysical signals recorded in conjunc-
tion with photolysis, 315-317 NAADP, see Nicotinic acid adenine dinucle-
caged ATP otide phosphate
actomyosin studies Neurotransmitter receptor
ATPase assays, 332-333 caged compounds
ATP binding, 330-332 apparatus for photolysis, 453
caged calcium studies, 332 cz-carboxy-2-nitrophenyl neurotransmit-
cross-bridge detachment, 330, 334, ters, photolytic properties, 453,
344-346 455-458
relaxation of tension, 330, 334 cell-flow quantification of photolysis,
muscle shortening velocity, inhibition 454
before photolysis, 343-346 channel opening, rate and equilibrium
caged compounds in photolysis studies constant determination, 457,
calcium chelators, 321, 324 459-460
nucleotides, 321,325 inhibitor studies
purity requirements, 321, 324-325 dissociation constant determinations,
table of properties, 322-323 464-465
motility assays, 334-335 effects on opening and closing rates,
photochemical conversion of caged com- 460-461,463-464
pounds, quantification 2-methoxy-5-nitrophenylneurotransmit-
ATP, 338-341 ters, photolytic properties, 453,
calcium, 339-340 457, 459
phosphate, 339 microsecond electrophysiology mea-
photolysis surements, 468, 470
applications, 214-215, 253-254, 310 neuronal circuit function analysis,
uniformity in sample chamber, 341-342 467-468
preparation for photolysis experiments X e n o p u s oocyte studies, 470-472
detergent treatment, 311 cell-flow studies
glycerol extraction, 311 apparatus, 451-453
mounting, 312-313 channel opening, rate and equilibrium
solutions for photolysis experiments constant determination, 460
buffer, 327 principle, 447, 450-451
cations, 326 classification, 444-445
changes following photolysis, 328-329 malfunctions in disease, 443-444
computer analysis, 325-326 minimum reaction scheme for mediated
free radical scavengers, 327-328 reactions
526 SUBJECT INDEX
characteristics, 445-446 6-Nitroveratryloxycarbonyl chloride conju-

equations for analysis, 448-449 gate, see Lysine, caged protein conju-
single cell excitatory and inhibitory neuro- gates
transmitter signal integration, NMDA, see N-Methyl-D-aspartate
465-467 Nmoc, see o-Nitromandelyloxycarbonyl
Nicotinic acetylcholine receptor, s e e Neuro- group
transmitter receptor Noracetylcholine, caged compound studies
Nicotinic acid adenine dinucleotide phos- with crystallized cholinesterases, 266-
phate 267, 269
caged 1-(2-nitrophenyl)ethyl compound Norbutyrylcholine, caged compound studies
calcium release analysis with fluores- with crystallized cholinesterases, 267,
cent probes, 414-415 269
high-performance liquid chromatogra- NPEGTA, calcium uncaging efficiency by
phy, 407-409 two-photon photolysis, 367-369
photolyis conditions, 414
structural characterization, 409
synthesis, 407
calcium release in sea urchin eggs, O
403-404
2-Nitrobenzyl group Oligonucleotide, photocleavable biotin la-
applications in caged compounds, 3-4, beling
64, 118-119 affinity purification, 145, 150-151
limitations in photochemical applications, high-performance liquid chromatography,
3-4 149-150
mass spectrometry analysis of modified phosphorylation at 5'-end, 153
proteins, 119, 124 photocleavage, 145-146, 150-151
photolysis reaction, 2-3, 226 polyacrylamide gel electrophoresis analy-
rate of photolysis, 122 sis, 151-152
scavenging free radicals after photolysis, synthesis, 145, 149, 153
134, 233, 328
o-Nitromandelyloxycarbonyl group
2 , 5 - d i ( t e r t - b u t y l ) h y d r o q u i n o n e caged com-
pound P
calcium studies with fluorescent indica-
tors, 435-437, 439 Peptide synthesis, caged tyrosine peptides,
cell loading, 435 83-85
synthesis, 432 Phosphate, caged
glutamate caged compound caging groups, see 4-Coumarinylmethyl
electrophysiology studies of glutamate group; Desyl group;p-Hydroxyphena-
receptor effects, 440-443 cyl group
photochemistry, 440 Fourier transform infrared spectroscopy
synthesis, 434-435 band assignment by isotopic labeling,
imidazole 227-229
reactivity, 431-432 difference spectra, 226-229
synthesis, 431-434 magnesium effects on spectra, 229-231
2-Nitrophenethyl group photochemical conversion, quantification,
limitations in photochemical applications, 339
3-4 Photobleaching, limitations compared to
photolysis reaction, 2 fluorescence photoactivation, 63
SUBJECT INDEX 527
Photocleavable affinity tags R

applications, 137-138
Raman spectroscopy
design, 138
caged versus uncaged GTP, 235-237
photocleavable biotins
instrumentation for photolysis studies,
applications, 152-154
225
criteria in design, 152
Ras
leucine-enkephalin
caged GTP studies in crystals
avidin affinity purification, 144-145,
caged GTP affinity, 257, 259-260
148-149
data collection, 259
conjugation, 144
flash lamp photolysis, 95, 215-216, 257,
fluorescamine assay, 145
261
photocleavage kinetics, 146-148
rapid freezing followed by data collec-
oligonucleotide labeling tion, 261
affinity purification, 145, 150-151 Ras mutation for thermostability, 260
high-performance liquid chromatog- resolution of data, 259-260, 262
raphy, 149-150 Fourier transform infrared spectroscopy
phosphorylation at 5'-end, 153 studies with caged GTP
photocleavage, 145-146, 150-151 data collection, 224-225
polyacrylamide gel electrophoresis difference spectra, 243-245
analysis, 151-152 Rayteigh range, Gaussian beam optics, 189
synthesis, 145, 149, 153 Reflection, Gaussian beam optics, 192-193
structures, 138-139 Refraction, Gaussian beam optics, 194
synthesis Resorufin, caged compounds
materials, 139-140 caging group selection, 67
photocleavable biotin NHS carbon- practical considerations for use, 67-69
ate, 140-142 protein labeling, 72-74
photocleavable biotin phosphorami- synthesis, 68-69
dite, 142-144, 153 types and structures, 65-66
Photolysis rate, calculation, 131 Rhodamine, caged compounds
Photoprotecting compounds, s e e Caged applications, 75, 77
compounds caging group selection, 67, 72
PKA, s e e Cyclic AMP-dependent protein practical considerations for use, 67, 69,
kinase 71-72
Profilin, 6-nitroveratryloxycarbonylchloride protein labeling, 72-74
caged conjugate synthesis, 69
functional analysis of caged protein, thiol-reactive caged compound
108-110 actin labeling and cross-linking, 170-
preparation, 108 171, 173-175
recombinant protein purification, preparation, 169-170
107-108 types and structures, 66
RS-20
caged tyrosine peptide
Q calmodulin binding analysis, 90-93
design of inactive caged peptide, 88-90
Q-switching, production, 195 kinetics of photolysis, 87-88
Quantum yield spectral and photochemical properties,
caged carboxylates, 43, 45-46 85-87
caged tyrosine peptides, 85-87 synthesis, 80-85
calculation, 131 structure and function, 79-80
528 SUBJECT INDEX
S kemptide modification with 2-nitrobenzyl

group, 126
Salicylic acid, caged modification in presence of cysteinyl resi-
calcium response measurement with fluo- dues, 119
rescent probes, 478, 481-482 selective reactivity in 2-nitrobenzyl modi-
injection into plant cells, 477-478 fication, 126-127
photolysis conditions, 478-480 thiophosphorylation of peptides, 125
structure, 475 Thymidine triphosphate, caged compound
synthesis, 475-477 synthesis, 489
Sarcoplasmic reticulurrdendoplasmic reticu- Tubulin, s e e a l s o Microtubule
lum calcium ATPase caged fluorescein-labeled tubulin
calcium oscillation role, 439 applications in transport and microtu-
2 , 5 - d i ( t e r t - b u t y l ) h y d r o q u i n o n e caged com-
bule dynamics, 348, 353, 356
pound studies electron microscopy following photoac-
calcium studies with fluorescent indica- tivation, 352-353, 355-356
tors, 435-437, 439 microinjection into neurons, 348, 350
cell loading, 435 neuron culture for microinjection,
synthesis, 432 349-350
SERCA, s e e Sarcoplasmic reticulum/endo- photoactivation, 350-351
plasmic reticulum calcium ATPase preparation, 349
Single-photon photolysis, s e e Laser pho- video microscopy, 351
tolysis microtubule preparation, 335
Sodium, calcium-exchanger preparation from bovine brain, 335
DM-nitrophen caged calcium studies Two-photon photolysis
electrophysiological measurements, caged calcium
302-306 apparatus, 361-365
photolysis kinetics, 301 applications, 356-357, 380
rate constant determinations, 305-306 calcium yield determination, 358,
regulatory site studies, 301-302, 369-374
304-305 temporal behavior of calcium concen-
stoichiometry, 303 tration distribution, 374-379
Sodium, potassium-ATPase uncaging action cross section determi-
electrophysiological measurements with nation, 365-369
caged ATP development, 176, 178
bilayer experiments, 291-295 laser systems, 178-179, 201-202, 357-
binding of caged compounds, 293 358, 361-365
patch clamp studies, 295-298 pulse duration, 178
photolysis conditions, 291, 293
spatial resolution, 359-361
time-resolved measurements, 293-294, Tyrosine, caged peptides
297-298 applications, 78-79, 94
ion specificity, 292-293 design of inactive caged peptides, 88-90,
physiological function, 290 94
reaction mechanism, 290-291 extinction coefficient determination, 85
Stimulated emission, theory, 179-182 kinetics of photolysis, 87-88
quantum yield determination, 85-87
T synthesis
2-bromo-cz-carboxyl-2-nitrophenyl
Thiophosphoryl group, caged protein conju- methyl ester, 80-82
gates coupling reaction, 82
SUBJECT INDEX 529
deprotection, 83 measuring compounds released by photol-

peptide synthesis, 83-85 ysis, 219, 253, 256-257
protecting of tyrosine, 82 muscle fibers, 253-254
thiophosphoryltyrosine modification, 133 photolysis yield considerations, 254-255,
258-259
Ras GTPase studies in crystals
W caged GTP affinity, 257, 259-260
data collection, 259
Waist position, Gaussian beam optics, 188, flash lamp photolysis, 215-216, 257,
190 261
Waist radius magnification, Gaussian beam rapid freezing followed by data collec-
optics, 189 tion, 261
Wave polarization, Gaussian beam optics, Ras mutation for thermostability, 260
190-191 resolution of data, 259-260, 262
time-resolved measurements with Laue
crystallography, 251,255-256, 259,
X
264, 277
X-ray crystallography, photolysis studies Xenopus oocyte
caged enzymes, 255, 266 caged inositol 1,4,5-triphosphate studies
concentration requirements for caged sub- of calcium release
strates, 252-254 caffeine studies, 393, 395
crystal size, 262 dose-response relationship, 391-392
glycogen phosphorylase, 252-253 InsP4 agonist studies, 393
heat damage of crystals by flash, 216- microinjection, 383-384
218, 258 neurotransmitter receptor studies with
irreversibility of photolysis, 262-264 caged compounds, 470-472

Marriett G - Methods in Enzimology v291 Caged Compounds MIE-AP (1998)

Uploaded by

Copyright:

Available Formats

Marriett G - Methods in Enzimology v291 Caged Compounds MIE-AP (1998)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Marriett G - Methods in Enzimology v291 Caged Compounds MIE-AP (1998)

Uploaded by

Copyright:

Available Formats

Preface

Caged compounds are biomolecules whose activity or function is

The primary objective of this volume is to provide newcomers to the

R.B. and Fay, F.S. Cyanocobalamin (vitamin B12) uptake by Ochromonas

student, Claudio Delise, Fred showed with scanning electron microscopic

every facet. He would challenge his colleagues or fellows to come up with

H. MAURICE GOODMAN DAVID WARSHAW

[1] N e w P h o t o p r o t e c t i n g Groups: Desyl and

The well-established photochemical release of bioorganic substrates

i "Caged compound" is a term coined in 1978for photolabile derivativesof natural substrates

Copyright © 1998 by Academic Press

NO2 347 nrn

3 R. S. Givens and L. W. Kueper III, Chem. Rev. 93, 55-66 (1993).

chemistry outlined in Scheme 1. 4-7 Several of these applications have been

that ADP formation by enzyme-induced hydrolysis of ATP was more rapid

New Phototriggers for Phosphates and Nucleotides:

were independent of the solvent (benzene, CH3CN, CH3OH, or H20).

Furthermore, the ligands on the phosphate group had no effect on the

14 R. S. Givens and B. Matuszewski, J. Am. Chem. Soc. 106, 6860-6861 (1984).

a. Synthesis of Desyl Phosphate (5a)

b. Synthesis of Desyl cAMP (7a)

X PO2R2 -(RO)2PO~ -H+ X

by low triplet energy quenchers such as naphthalene (ET = 62 kcal/mol).

18j. Wirz, R. Givens, and B. Hellrung, unpublished results, 1997.

Subsequently, Corrie and Trentham 19 explored the photochemistry of

I9 j. E. T. Corrie, G. P. Reid, D. R. Trentham, M. A. Mazid, and M. B. Hursthouse, J. Chem.

Furthermore, based on the earlier studies by Anderson and Reese, 23 intro-

10lg Y = OCH3; X = CI 1la, Y = OCH 3 12a, Y - OCH3

Parallel investigations of the structurally analogous p-methoxyphenacyl

23j. C. Anderson and C. B. Reese, Tetrahedron Lett., 1-4 (1962).

ies, it may be important for applicationslwhen caged derivatives are intro-

HOf'~..~ +NH4 +NH4 +NH4 OH]~OH kr=6.gx108s-1 "O -O .O ~......~O--~J

Applications of Desyl and p-Hydroxyphenacyl Phototriggers to

10e, X = OPO3R2 (n,z*)3

tives of the two phototriggers is outlined in Scheme 6 (structures 19-26)

33 L. Chimilio, A. H. Jung, J. F. W. Weber, and R. S. Givens, Unpublished results, 1997-1998.

O-t-Bu and N-t-BOC protectedamino acids

24a,b Glu phototngger

of the ester to hydrolysis. The lower basicity of the a-amino group as a

O H NH3® hv, 350 nm H NI-1~3

O CO20 1:1 CH3CN:H20 CO2G +

22a Desyl GABA GABA 6a

O H NH3(~ hv, 300 nm H NH3

O CO2~ D20/H20;~100% ~r ~ v "CO~

24bp-hydroxyphenacyl Glu Glu

Hydrolysis thus avoided, the photochemistry of the caged dipeptide 27

Applications of Desyl and p-Hydroxyphenacyl Phototriggers

FIG. 1. 1H NMR spectra of the photorelease of Ala-Ala (28) from p-hydroxyphenacyl-

glutamate (24b) in patch-clamp experiments will generate an action poten-

FI6. 2. In a voltage-clamp experiment, desyl L-glutamate (24a; 1 mM in artificial cerebral

F~o. 3. In a current-clamp recording, p-hydroxyphenacyl L-glutamate (24b; 50 /zM) is

shutter: 5, 10, 20, 5 0 , 7 0 ms

FIG. 4. Voltage-clamp experiments from flash excitation of 50/aM of p-hydroxyphenacyl

All nuclear magnetic resonance (NMR) spectra are reported in ppm

33.52 (m, J = 13.2); IR (neat) 3200-2800, 1600, 1450, 1175 (POCH3),

35 F. Rameriz and J. F. Marecek, Synthesis 449 (1985).

Synthesis of p-Hydroxyphenacyl Phototriggers 27

Sequence for Conversion of 13 to p-Hydroxyphenacyl Phosphate (16)

Coupling of p-Hydroxyphenacyl Phosphate with ADP

39 C. H. Heathcock and R. Ratcliffe, J. Am. Chem. Soc. 93, 1746-1757 (1971).