DOI: 10.1126/science.

1144504
, 1076 (2007); 317 Science
et al. Martin St. Maurice,
Biotin-Dependent Multifunctional Enzyme
Domain Architecture of Pyruvate Carboxylase, a
www.sciencemag.org (this information is current as of October 22, 2007 ):
The following resources related to this article are available online at
http://www.sciencemag.org/cgi/content/full/317/5841/1076
version of this article at:
including high-resolution figures, can be found in the online Updated information and services,
http://www.sciencemag.org/cgi/content/full/317/5841/1076/DC1
can be found at: Supporting Online Material
http://www.sciencemag.org/cgi/content/full/317/5841/1076#otherarticles
, 6 of which can be accessed for free: cites 29 articles This article
http://www.sciencemag.org/cgi/collection/biochem
Biochemistry
: subject collections This article appears in the following
http://www.sciencemag.org/about/permissions.dtl
in whole or in part can be found at: this article
permission to reproduce of this article or about obtaining reprints Information about obtaining
registered trademark of AAAS.
is a Science 2007 by the American Association for the Advancement of Science; all rights reserved. The title
Copyright American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
(print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the Science

o
n

O
c
t
o
b
e
r

2
2
,

2
0
0
7

w
w
w
.
s
c
i
e
n
c
e
m
a
g
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

of syntaxin increases the number of clusters
rather than their size, which is in agreement with
previous observations (7). In contrast, a hypo-
thetical cylindrical syntaxin arrangement (Fig.
4C, right) would not explain our findings.
The self-organization model shows that a
simple balance between weak homophilic inter-
actions and a repulsive component can explain all
available experimental data on the composition
and the dynamics of syntaxin clusters.
Syntaxin clustering via self-assembly is likely
a paradigm that applies to a variety of membrane
protein clusters. Like the syntaxins, many other
membrane proteins are known both to form clus-
ters and to homo-oligomerize in vitro via their
cytoplasmic domain. Examples include structur-
ally diverse proteins such as synaptotagmins
(13, 20) and receptors (4, 21) (containing single
transmembrane domains), anion transporters
[e.g., (22)] containing multiple transmembrane
domains, and proteins such as reggie with lipid
anchors [e.g., (23)]. The tendency to form homo-
oligomers in solution would be enhanced in
the plane of a membrane where membrane
anchoring orients the molecules ideally for
oligomerization. Hence, in the membrane self-
association is enforced, leading finally to
nanodomains containing many copies of the
self-assembled membrane proteins. Thus, just as
for the syntaxins, homo-oligomerization will
likely lead to cluster formation for any membrane
protein that can self-associate in solution.
The concept of clustering via self-assembly is
not restricted to membrane proteins oligomeriz-
ing via their cytoplasmic domain. Rather, it can
be widely applied to membrane proteins oligo-
merizing via their transmembrane or extracellular
domains, with the only prerequisite being that
in these cases there would be additional protein
domains causing steric crowding or other re-
pulsive forces, as, for example, accumulating
charges.
As outlined above, the mechanisms underly-
ing the high degree of membrane micropatterning
are not well understood for most membrane
proteins. With syntaxin 1 as an example, we pre-
sent a conceptual framework for the description
of protein domains in membranes. This biological
principle is expected to explain a considerable
part of the compositional variability of membrane
protein clusters and hence will help to advance
our understanding of membrane micropatterning.
References and Notes
1. S. J. Singer, G. L. Nicolson, Science 175, 720 (1972).
2. K. Simons, E. Ikonen, Nature 387, 569 (1997).
3. A. Kusumi et al., Annu. Rev. Biophys. Biomol. Struct. 34,
351 (2005).
4. S. Uhles, T. Moede, B. Leibiger, P. O. Berggren,
I. B. Leibiger, J. Cell Biol. 163, 1327 (2003).
5. M. Kai et al., J. Biochem. (Tokyo) 140, 677 (2006).
6. S. H. Low et al., Mol. Biol. Cell 17, 977 (2006).
7. J. J. Sieber, K. I. Willig, R. Heintzmann, S. W. Hell,
T. Lang, Biophys. J. 90, 2843 (2006).
8. T. Lang et al., EMBO J. 20, 2202 (2001).
9. M. Ohara-Imaizumi et al., J. Biol. Chem. 279, 8403
(2004).
10. S. A. Predescu, D. N. Predescu, K. Shimizu, I. K. Klein,
A. B. Malik, J. Biol. Chem. 280, 37130 (2005).
11. J. C. Lerman, J. Robblee, R. Fairman, F. M. Hughson,
Biochemistry 39, 8470 (2000).
12. S. W. Hell, J. Wichmann, Opt. Lett. 19, 780 (1994).
13. K. I. Willig, S. O. Rizzoli, V. Westphal, R. Jahn, S. W. Hell,
Nature 440, 935 (2006).
14. G. Donnert et al., Proc. Natl. Acad. Sci. U.S.A. 103,
11440 (2006).
15. Materials and methods are available on Science Online.
16. A. K. Kenworthy et al., J. Cell Biol. 165, 735 (2004).
17. A. T. Brunger, Q. Rev. Biophys. 38, 1 (2005).
18. R. Laage, J. Rohde, B. Brosig, D. Langosch, J. Biol. Chem.
275, 17481 (2000).
19. I. Dulubova et al., EMBO J. 18, 4372 (1999).
20. M. Fukuda, E. Kanno, Y. Ogata, K. Mikoshiba, J. Biol.
Chem. 276, 40319 (2001).
21. K. Kanazawa, A. Kudo, J. Bone Miner. Res. 20, 2053
(2005).
22. D. Navaratnam, J. P. Bai, H. Samaranayake,
J. Santos-Sacchi, Biophys. J. 89, 3345 (2005).
23. C. Neumann-Giesen et al., Biochem. J. 378, 509 (2004).
24. Single-letter abbreviations for the amino acid residues
are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe;
G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro;
Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.
25. R. B. Sutton, D. Fasshauer, R. Jahn, A. T. Brunger, Nature
395, 347 (1998).
26. The authors thank R. Jahn and S. O. Rizzoli for helpful
suggestions on the manuscript, V. Westphal and J. Keller
for help with analysis software, K.-H. Drexhage for
providing the dyes Atto532 and Atto647N, and F. E. Zilly
and M. G. Holt for gifts of a plasmid and a recombinant
protein, respectively. T.L. was supported by a grant from
the Deutsche Forschungsgemeinschaft (LA1272/2).
S.W.H. consults for Leica Microsystems CMS, Gmbh,
Mannheim, Germany.
Supporting Online Material
www.sciencemag.org/cgi/content/full/317/5841/1072/DC1
Materials and Methods
Figs. S1 to S7
References
23 February 2007; accepted 13 July 2007
10.1126/science.1141727
Domain Architecture of Pyruvate
Carboxylase, a Biotin-Dependent
Multifunctional Enzyme
Martin St. Maurice,
1
Laurie Reinhardt,
1
Kathy H. Surinya,
2
Paul V. Attwood,
3
John C. Wallace,
2
W. Wallace Cleland,
1
Ivan Rayment
1
*
Biotin-dependent multifunctional enzymes carry out metabolically important carboxyl group
transfer reactions and are potential targets for the treatment of obesity and type 2 diabetes. These
enzymes use a tethered biotin cofactor to carry an activated carboxyl group between distantly
spaced active sites. The mechanism of this transfer has remained poorly understood. Here we report
the complete structure of pyruvate carboxylase at 2.0 angstroms resolution, which shows its domain
arrangement. The structure, when combined with mutagenic analysis, shows that intermediate transfer
occurs between active sites on separate polypeptide chains. In addition, domain rearrangements
associated with activator binding decrease the distance between active-site pairs, providing a
mechanism for allosteric activation. This description provides insight into the function of biotin-
dependent enzymes and presents a new paradigm for multifunctional enzyme catalysis.
I
n biochemical pathways, metabolites must
be efficiently transferred between enzymes
to avoid the energetic penalty associated
with their loss to diffusion, degradation, or com-
peting side-reactions. Particularly efficient trans-
fer is afforded by multifunctional enzymes that
directly transfer products from one reactive site
to the next through tunnels and channels (1) or
through the use of covalently attached prosthetic
groups (2). Although the reactions catalyzed at
the individual active sites of many multifunc-
tional enzymes are well understood, few studies
have detailed their complete domain architec-
ture. Consequently, descriptions of intermediate
transfer between active sites in multifunctional
enzymes remain largely incomplete.
The multifunctional enzymes of the biotin-
dependent family use a covalently attached biotin
prosthetic group to directly transfer an activated
CO
2

intermediate between distinct active sites

in several essential metabolic reactions (3).
Some members of this enzyme family, including
acetylcoenzyme A carboxylase (ACC) and pyr-
uvate carboxylase (PC), have recently attracted
interest as potential targets in the treatment of
obesity and type 2 diabetes (4, 5). Although
several individual domain structures have been
determined for various family members (69),
the relative arrangement of these domains in a
complete multifunctional enzyme remains un-
known. PC is typically composed of three
distinct functional domains arranged on a single
120- to 130-kD polypeptide chain. The three
domains are an N-terminal biotin carboxylase
(BC) domain, a central carboxyltransferase (CT)
1
Department of Biochemistry, University of Wisconsin,
Madison, WI 53706, USA.
2
School of Molecular and
Biomedical Science, University of Adelaide, SA 5005,
Australia.
3
School of Biomedical, Biomolecular, and Chemical
Sciences, University of Western Australia, Crawley, WA 6009,
Australia.
*
To whom correspondence should be addressed. E-mail:
[email protected]
24 AUGUST 2007 VOL 317 SCIENCE www.sciencemag.org 1076
REPORTS

o
n

O
c
t
o
b
e
r

2
2
,

2
0
0
7

w
w
w
.
s
c
i
e
n
c
e
m
a
g
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

domain, and a C-terminal biotin carboxyl carrier
protein (BCCP) domain (Fig. 1A). The enzyme
uses a covalently attached biotin cofactor to
catalyze the adenosine triphosphate (ATP)
dependent carboxylation of pyruvate to oxalo-
acetate in two steps (10). Biotin is initially
carboxylated at the BC active site by ATP and
bicarbonate. The carboxyl group is subsequently
transferred by carboxybiotin to a second active
site in the CT domain, where pyruvate is carbo-
xylated to generate oxaloacetate. The BCCP
domain transfers the tethered cofactor between
the two remote active sites. To discover what
distance is traversed by the BCCP domain and
what domain motions facilitate the transfer of
biotin between active sites, we have determined
the crystal structure of the complete PC enzyme.
The crystal structure of PC from Rhizobium
etli (RePC) was determined by single-wavelength
anomalous diffraction and solvent flattening,
followed by molecular replacement (11). The
final model was refined to a resolution of 2.0
(see table S1 for refinement statistics and fig. S1
for representative electron density). The crystals
contain a dimer in the asymmetric unit. Monomer
A includes the complete polypeptide chain,
composed of the three functional domains [BC
(blue), CT (yellow), and BCCP (red)], and a
central allosteric domain (green) that has not
been previously described (Fig. 1B). Monomer B
is nearly complete except for the C-terminal
BCCP domain, which was disordered and could
not be modeled (model completeness is de-
scribed in table S2). The dimer is asymmetric,
and the notable differences in the position and
orientation of the BC domain between the two
monomers have important implications in the
mechanism of PC catalysis and regulation.
The two active sites of RePC were identified
through structural homology with related enzyme
subunits and from the position of bound ligands.
The nonhydrolyzable ATP analog, adenosine 5-
O-(3-thiotriphosphate) (ATP-g-S), is bound at
the active site of the BC domain (Fig. 2A). The
binding site consists of an ATP-grasp fold (12),
typical of the arrangement reported for ortholo-
gous enzymes (6, 13). The overall BC domain
fold is very similar [1.1 to 1.2 root mean square
deviation (RMSD)] to the BC subunits from other
family members (fig. S2A). However, the position
of ATP-g-S in the binding site of RePC reveals
differences with the position of ATP in the BC
subunit of ACC (13) (fig. S3). The position of
ATP-g-S in RePC is much more typical of other
ATP-grasp binding sites (14, 15) [supporting
online material (SOM) text]. The CT domain
of RePC also shares high structural similarity (1.1
to 1.5 RMSD) with the equivalent subunits
from orthologous enzymes (7, 9) (fig. S2B). This
domain consists of a core a
8
b
8
barrel capped by a
funnel that centers on a structurally conserved
metal ion at the active site. The metal ion in RePC
superimposes closely with the active-site metal
ions of orthologous enzymes. Metal-ion analysis
indicates that Zn
2+
is the active-site metal in
RePC, which is consistent with the observed elec-
tron density. Mutations of residues directly ligating
this Zn
2+
ion either eliminate or substantially re-
duce the catalytic activity of PC (16).
PC is subject to a variety of allosteric control
mechanisms (5), and acetylcoenzyme A(acetyl-
CoA) is an allosteric activator for PC enzymes
from many species (17). Both acetyl-CoA and
the nonhydrolyzable analog, ethyl-CoA, are
activators of RePC, with activation constant K
A
values of 30.4 and 360 mM, respectively (table
S3). RePC was cocrystallized with ethyl-CoA,
and the partial electron density allowed the
nucleotide portion of the activator to be modeled
into the structure, revealing the location of the
allosteric binding site. The binding site is located
in a previously unrecognized domain that serves
as a mediator among the BC, CT, and BCCP do-
mains. The allosteric domain consists of four
antiparallel b strands bracketing a central a helix
(Fig. 1B, green) and superficially resembles the
protein fold and acetyl-CoAbinding of the GCN-
5related acetyltransferases (18) [secondary
structure matching (19) gives a weak Q score of
0.053 with a GCN-5related aminoglycoside 3-
N-acetyltransferase, 1bo4]. The central a helix
Fig. 1. (A) Schematic
drawing of the primary
structure arrangement for
the multidomain PC from
R. etli. The allosteric do-
main, indicated with aster-
isks, includes residues 471
to 489 and 1002 to 1073.
(B) The structure of the R.
etli PC monomer A. The
BC, CT, BCCP, and alloste-
ric domains are colored
blue, yellow, red, and
green, respectively. The
chemical reactions cata-
lyzed in the individual
domains are illustrated
below the corresponding
domain structure. ADP,
adenosine diphosphate;
P
i
, inorganic phosphate.
BC CT BCCP * *
4
6
5
4
7
1

-

4
8
9
5
1
5
1
0
0
0
1
0
0
2
1
0
7
3
1
0
8
4
1
1
5
4
1
A
ATP
CoA
Zn
2+
Biotin Carrier (BCCP) Domain
Biotin Carboxylation (BC) Domain Carboxyltransferase (CT) Domain
B
BCCP-biotin-CO2

+ pyruvate
BCCP-biotin + oxaloacetate
BCCP-biotin + HCO3

+ ATP
BCCP-biotin-CO2

+ ADP + Pi
Fig. 2. (A) The ATP binding site. The revised def-
inition of the precise ATP binding site reveals
several interacting residues not previously noted in
the structure of the BC subunit from ACC. In par-
ticular, the interaction of Lys
245
with the g-phosphate
of ATP is consistent with substrate-labeling studies
that suggest that this residue directly interacts
with ATP (30). Several previously unrecognized
residues, including Glu
283
, Glu
297
, and Asn
299
, are
now seen to be important for coordinating the two
essential Mg
2+
ions (depicted as green spheres)
and ATP binding. (B) The ethyl-CoA binding site.
Interactions with the nucleotide portion of ethyl-
CoA include residues from both BC subunits and
from the allosteric domain. Most notably, Arg
472
of the allosteric domain creates a strong interac-
tion with the 5 a-phosphate of ethyl-CoA.
Lys 245
Gln 240
His 216
Glu 283
Asn 299
Glu 297
Lys 166
Glu 208
C=O
(Lys 209)
Arg 472
Arg 427
Asn 1055
backbone NH
(Asp 471)
backbone NH
(Gln 470)
C=O
(Asp 47)
A B
www.sciencemag.org SCIENCE VOL 317 24 AUGUST 2007 1077
REPORTS

o
n

O
c
t
o
b
e
r

2
2
,

2
0
0
7

w
w
w
.
s
c
i
e
n
c
e
m
a
g
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

includes Asp
471
to His
488
and connects the BC
and CT domains, whereas Asp
1018
to Asp
1065
fold around this helix in an antiparallel (b
4
) sheet
and connect the CT domain to the BCCP do-
main. The allosteric domain displays no se-
quence similarity with other members of the
biotin-dependent enzyme family, suggesting
that it is specific to PC. Ethyl-CoA is bound at
the N-terminal end of the central a helix, near
the BC-BC dimer interface (Fig. 2B). Residues
interacting with the nucleotide portion of CoA
are highly conserved among PC enzymes, ex-
plaining why PCfromCorynebacterium(20) and
subunit-type PCs (17) that have mutations or
deletions in these residues are not activated by
acetyl-CoA.
A critical feature of multifunctional enzyme
catalysis is the requisite transfer of intermediates
between remote active sites. The structure of
RePC monomer A includes a complete BCCP
domain relative to the individual catalytic domains
of the enzyme, which advances our understand-
ing of intermediate transfer. Although the electron
density for the biotin cofactor was disordered, the
global position and structure of the BCCP domain
provide a powerful restraint on the arrangement
of the prosthetic group carrier (SOM text). It is
clear from the structure that the distance between
active sites greatly exceeds the ~16 length of the
biotin arm. Thus, the BCCP domain itself must
move as it carries biotin between the active sites.
The 34 linker connecting BCCP to the allosteric
domain is highly flexible (8, 21), facilitating do-
main movement. The linker is similar to those
described for the lipoyl domains of 2-oxo acid
dehydrogenase (22). If the BCCP domain were to
transfer carboxybiotin between active sites on
the same polypeptide chain, a dramatic domain
movement would be required. Alternatively, serv-
ing between active sites on opposing polypep-
tide chains necessitates substantially less motion.
The quaternary structure and kinetic evidence de-
scribed below support the latter scenario.
The RePC crystal lattice contains a tetramer,
composed of a dimer of dimers. The tetramer is
stabilized by conserved dimerization interfaces
between neighboring BC domains and between
neighboring CT domains. PC exists predominant-
ly as a tetramer in solution, and only the tetramer
catalyzes the overall reaction (23). The tetramer is
asymmetric, with ethyl-CoA bound to each of
the two (monomer A) monomers on the top face,
whereas the two (monomer B) monomers on the
bottom face are unbound. On the top face, the BC
active site is positioned ~65 from the CTactive
site of an opposing polypeptide chain, with the
biotinylated BCCP domain positioned equidistant
between the two. Based on the distance and ori-
entation between these opposing active sites, the
BCCP domain is predicted to swing between
active-site pairs on opposing polypeptide chains
rather than between active sites on the same chain
(Fig. 3).
To confirm this prediction, we created two
RePC mutants. In one construct, the biotinylated
Lys
1119
residue was mutated to prevent biotinyl-
ation of the BCCP domain, and, in the second
construct, a carbamylated Lys
718
in the CT active
site was mutated to impair the second half-reaction.
80
C
ATP
ATP
Zn
2+
Zn
2+
180
Monomer B
65
Monomer A
B
CoA
CoA
ATP
ATP Zn
2+
Zn
2+
CoA
CoA
BC
CC BC BC BC BC BC BBB
C BC BC BC BC
ATP
nn Zn Zn ZZ
22+ 2+ 2+ 2
CT
TTTTTT CT CT CCCC
TTTTTT CT CT CCCCCCC
BC
ATP
BCCP
BCCP
Zn Zn Z
2+
CT
A
Fig. 3. (A) Model of the RePC tetramer showing the movement of the BCCP
domain between neighboring active sites on opposing polypeptide chains. (B)
Surface representation of the top face of the tetramer. For clarity, one of the two
individual monomers has been outlined in black. The distance between ATP-g-S
in the BC active site and Zn
2+
in the CT active site of the opposing polypeptide
chain is 65 . (C) Surface representation of the bottom face of the tetramer,
after a 180 rotation about the y axis. The BCCP domain is disordered in these
monomers and could not be modeled. The distance between ATP-g-S in the BC
active site and Zn
2+
in the CT active site of the opposing polypeptide chain
increases to 80 as a result of the altered orientation of the BC domain.
CT
BC
Chain A (+ ethyl CoA)
Chain B ( ethyl CoA)
Arg 472
Chain B
Chain A
domain movement
Fig. 4. Ethyl-CoA binding alters the position and orientation of the BC domain. The allosteric domain
of monomer A was aligned with the allosteric domain of monomer B. The superposition shows minor
deviations in the CT domain but substantial deviations in the BC domain. The BC domain begins to
deviate precisely at the three residues forming tight hydrogen-bonding interactions with the 5 a- and
b-phosphates of ethyl-CoA: Arg
472
and the backbone amides of Gln
470
and Asp
471
. Where ethyl-CoA is
unbound in monomer B, the side chain of Arg
472
is pointed away from the position occupied by the
activator, and the backbone of Gln
470
and Asp
471
threads from a radically different direction.
24 AUGUST 2007 VOL 317 SCIENCE www.sciencemag.org 1078
REPORTS

o
n

O
c
t
o
b
e
r

2
2
,

2
0
0
7

w
w
w
.
s
c
i
e
n
c
e
m
a
g
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m

When assayed individually, the mutant enzymes
Lys
1119
Gln
1119
(K1119Q) and K718Q exhib-
ited 0.1 and 4% wild-type activity, respectively
(table S4). Hybrid tetramers were created by mix-
ing and diluting the mutants together before as-
saying for enzyme activity. Dilution of PCpromotes
equilibration among monomers, dimers, and tet-
ramers (23) and allows a mixed heterotetramer pop-
ulation to reassociate. The mixed population of
heterotetramers exhibited 20% wild-type activity
(table S3), nearly five times as much as that ob-
served with either mutant homotetramer and near
to the maximumpredicted activity of 26%(fig. S4).
The recovery of activity on reassociation is possible
only if the hybrid tetramers recombine to restore a
functional pair of neighboring active sites, capable
of transferring the tethered carboxybiotin interme-
diate between two opposing chains. The transfer of
a carboxybiotin intermediate between active sites
on separate polypeptide chains is a previously un-
recognized feature of PC catalysis. Several multi-
functional enzymes have similarly been shown to
transfer their tethered intermediates between ac-
tive sites on opposing polypeptide chains (24, 25),
suggesting that intermolecular intermediate trans-
fer is a common and essential feature of catalysis.
Ethyl-CoA is bound to only one monomer
of the RePC asymmetric dimer, permitting a di-
rect comparison of the consequences of activator
binding on domain arrangement and orientation.
A superposition of the two monomers reveals a
40 rotation and a translocation of nearly 40
in the BC active site, centered at the ethyl-CoA
binding site of the allosteric domain (Fig. 4). In
the tetramer, ethyl-CoA is bound to both mono-
mers on the top face, and the BC active site is
positioned ~65 from its opposing CT active-
site pair (Fig. 3B). On the bottom face of the
tetramer, ethyl-CoA is unbound, and the distance
between the opposing active-site pairs increases
to ~80 (Fig. 3C). The rotation in the BC do-
main inhibits acetyl-CoA binding on the bottom
face of the tetramer. Thus, only two binding sites
are available per tetramer, which is consistent
with the Hill coefficient observed for yeast PC
(26) and with the observation that only 50% of
acetyl-CoA binding sites are occupied in yeast
PC (27). The structure suggests that acetyl-CoA
activates PC by decreasing the distance between
neighboring active sites. While the active-site pairs
on the top face of the tetramer are pushed closer
together, the active-site pairs on the bottom face
are pulled farther apart. This is a rare example of
allosteric activation paired with negative coop-
erativity and implies that half of the active-site
pairs are more active than the others. Recent ki-
netic studies and numerical simulations support
half-sites reactivity for the BC subunit of ACC
(28), suggesting that this mechanism is conserved
among enzymes of the biotin-dependent family.
Such half-sites reactivity may permit PC to affect
efficient catalysis while maintaining its in vivo
association with other metabolic enzymes (29).
The allosteric binding site in PC offers a
target for modifiers of activity that may be useful
in the treatment of obesity or type 2 diabetes, and
the mechanistic insights gained from the com-
plete structural description of RePC permit de-
tailed investigations into the individual catalytic
and regulatory sites of the enzyme. Furthermore,
as a consequence of its fully defined domain ar-
chitecture, PC represents a paradigm for under-
standing interdomain arrangement and allosteric
regulation in multifunctional enzymes.
References and Notes
1. F. M. Raushel, J. B. Thoden, H. M. Holden, Acc. Chem.
Res. 36, 539 (2003).
2. R. N. Perham, Annu. Rev. Biochem. 69, 961 (2000).
3. S. Jitrapakdee, J. C. Wallace, Curr. Protein Pept. Sci. 4,
217 (2003).
4. L. Tong, Cell. Mol. Life Sci. 62, 1784 (2005).
5. S. Jitrapakdee, A. Vidal-Puig, J. C. Wallace, Cell. Mol. Life
Sci. 63, 843 (2006).
6. S. Kondo et al., Acta Crystallogr. D 60, 486 (2004).
7. P. R. Hall et al., EMBO J. 23, 3621 (2004).
8. E. L. Roberts et al., Biochemistry 38, 5045 (1999).
9. R. Studer et al., J. Mol. Biol. 367, 547 (2007).
10. P. V. Attwood, J. C. Wallace, Acc. Chem. Res. 35, 113
(2002).
11. Materials and methods are available as supporting
material on Science Online.
12. M. Y. Galperin, E. V. Koonin, Protein Sci. 6, 2639 (1997).
13. J. B. Thoden, C. Z. Blanchard, H. M. Holden, G. L.
Waldrop, J. Biol. Chem. 275, 16183 (2000).
14. J. B. Thoden, G. Wesenberg, F. M. Raushel, H. M. Holden,
Biochemistry 38, 2347 (1999).
15. J. B. Thoden, S. Firestine, A. Nixon, S. J. Benkovic,
H. M. Holden, Biochemistry 39, 8791 (2000).
16. J. Yong-Biao, M. N. Islam, S. Sueda, H. Kondo,
Biochemistry 43, 5912 (2004).
17. S. Jitrapakdee, J. C. Wallace, Biochem. J. 340, 1
(1999).
18. F. Dyda, D. C. Klein, A. B. Hickman, Annu. Rev. Biophys.
Biomol. Struct. 29, 81 (2000).
19. E. Krissinel, K. Henrick, Acta Crystallogr. D 60, 2256
(2004).
20. P. G. Peters-Wendisch, V. F. Wendisch, S. Paul,
B. J. Eikmanns, H. Sahm, Microbiology 143, 1095 (1997).
21. D. V. Reddy, S. Rothemund, B. C. Shenoy, P. R. Carey,
F. D. Sonnichsen, Protein Sci. 7, 2156 (1998).
22. R. N. Perham, Biochemistry 30, 8501 (1991).
23. P. V. Attwood, W. Johannssen, A. Chapman-Smith,
J. C. Wallace, Biochem. J. 290, 583 (1993).
24. V. S. Rangan, A. K. Joshi, S. Smith, J. Biol. Chem. 273,
34949 (1998).
25. Y. Tang, C. Y. Kim, I. I. Mathews, D. E. Cane, C. Khosla,
Proc. Natl. Acad. Sci. U.S.A. 103, 11124 (2006).
26. J. J. Cazzulo, A. O. Stoppani, Arch. Biochem. Biophys.
127, 563 (1968).
27. A. Chapman-Smith, G. W. Booker, P. R. Clements,
J. C. Wallace, D. B. Keech, Biochem. J. 276, 759 (1991).
28. M. S. de Queiroz, G. L. Waldrop, J. Theor. Biol. 246, 167
(2007).
29. L. A. Fahien, J. W. Davis, J. Laboy, J. Biol. Chem. 268,
17935 (1993).
30. Y. Kazuta, E. Tokunaga, E. Aramaki, H. Kondo, FEBS Lett.
427, 377 (1998).
31. This work was supported by grants to W.W.C., J.C.W., and
P.V.A. from NIH (GM070455) and to I.R. from NIH
(AR35186). M.St.M. was supported in part by a fellowship
from the Natural Science and Engineering Research
Council of Canada. Use of the Structural Biology BM19
beamline Argonne National Laboratory Advanced Photon
Source was supported by the U.S. Department of Energy,
Office of Energy Research, under contract number W-31-
109-ENG-38. We thank M. Dunn for providing the R. etli
PC clone; V. Klenchin for helpful discussions and critical
reading of the manuscript; J. B. Thoden and H. M. Holden
for preliminary crystallography trials; A. Netting
(Adelaide Microscopy) for technical assistance with metal-
ion analysis; and A. Steinberg for his assistance in
generating the illustrations. Coordinates and structure
factors have been deposited in the Protein Data Bank
(www.rcsb.org) with the accession number 2QF7.
Supporting Online Material
www.sciencemag.org/cgi/content/full/317/5841/1076/DC1
Materials and Methods
SOM Text
Figs. S1 to S4
Tables S1 to S4
References
1 May 2007; accepted 5 July 2007
10.1126/science.1144504
When Fear Is Near:
Threat Imminence Elicits Prefrontal
Periaqueductal Gray Shifts in Humans
Dean Mobbs,* Predrag Petrovic, Jennifer L. Marchant, Demis Hassabis, Nikolaus Weiskopf,
Ben Seymour, Raymond J. Dolan, Christopher D. Frith
Humans, like other animals, alter their behavior depending on whether a threat is close or distant. We
investigated spatial imminence of threat by developing an active avoidance paradigm in which
volunteers were pursued through a maze by a virtual predator endowed with an ability to chase,
capture, and inflict pain. Using functional magnetic resonance imaging, we found that as the virtual
predator grew closer, brain activity shifted from the ventromedial prefrontal cortex to the
periaqueductal gray. This shift showed maximal expression when a high degree of pain was anticipated.
Moreover, imminence-driven periaqueductal gray activity correlated with increased subjective degree of
dread and decreased confidence of escape. Our findings cast light on the neural dynamics of threat
anticipation and have implications for the neurobiology of human anxiety-related disorders.
C
ritical to an organisms survival is the
ability to switch flexibly between defen-
sive states in response to threat. Within
behavioral ecology, a key component of de-
fensive switching is the predatory imminence
continuum where distinct threat states are con-
figured according to whether a predator is distal
or proximal to the prey (15). This continuum
www.sciencemag.org SCIENCE VOL 317 24 AUGUST 2007 1079
REPORTS

o
n

O
c
t
o
b
e
r

2
2
,

2
0
0
7

w
w
w
.
s
c
i
e
n
c
e
m
a
g
.
o
r
g
D
o
w
n
l
o
a
d
e
d

f
r
o
m