ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 228, No. 1, January, pp. 159-169, 1984

Binding of Ligands to the Catalytic Zinc Ion

in Horse Liver Alcohol Dehydrogenase
CHRISTIAN
Biochemical

SYVERTSEN

Institute,

University

AND

of

JOHN S. MCKINLEY-MCKEE

Oslo, P.O. Bm

1041, Blindem,

Oslo 3, Norway

Received June 22, 1983, and in revised form August 8, 1983

The affinity of nitrogen and sulfur ligands for the catalytic zinc ion in horse liver
alcohol dehydrogenase has been investigated by their influence on the affinity labeling
reaction with iodoacetate. All the nitrogen compounds including ammonia, a primary
and a secondary amine, and heterocycles containing a pyridine-type
nitrogen with the
exception of 2,2-dipyridyl were found to activate the affinity labeling reaction. Activation
results from inner-sphere ligand coordination to the catalytic zinc ion. Closely related
pyridine compounds gave a regular increase in affinity for the enzyme with increasing
basicity, as expected for coordination to a metal ion. The sulfur compounds penicillamine
and mercaptoethanol also activated the affinity labeling reaction, but dimercaptopropanol
bound very tightly as a bidentate inhibited the reaction. The anions hydrosulfide, diethyldithiocarbamate,
and cyanide coordinated to the catalytic zinc ion, whereas azide,
thiocyanate, tetrazole, and iodide complexed the anion-binding
site. The anionic metal
ligands increased the rate of inactivation
of the enzyme with iodoacetamide by binding
to the catalytic zinc ion, while the binding of iodoacetate to the anion-binding
site was
prevented.
The catalytic zinc ion in horse liver alcohol dehydrogenase is liganded by three
protein ligands, while at least one position
in the coordination sphere is occupied by
a solvent-exchangeable
water molecule (1).
The enzyme is inactivated when the protein
ligand Cys-46 is alkylated by iodoacetate,
iodoacetamide,
or a-bromo+imidazolyl
propionic acid. Iodoacetate and cy-bromofl-imidazolyl
propionic acid react via an
affinity labeling mechanism, whereas iodoacetamide reacts without forming any
detectable reversible complex with the enzyme and is less specific for Cys-46(2-4).
In the reversible complex with the enzyme,
iodoacetate is bound to the anion-binding
site of which Arg-47 is a part; cY-bromo+
imidazolyl propionic acid also binds to the
catalytic zinc ion.
Some neutral ligands like imidazole,

1 To whom correspondence

l,lO-orthophenanthroline,
and 2,2-dipyridyl coordinate the catalytic zinc ion (5).
Imidazole has been shown to activate the
affinity labeling reaction with iodoacetate
and the inactivation
with iodoacetamide.
This resulted from the electronic properties of imidazole as a metal ligand (6). In
the present work a number of nitrogen and
sulfur ligands have been investigated
to
compare their preference and affinity as
ligands for the catalytic zinc ion.
The anion-binding
site and the catalytic
zinc ion are essential for coenzyme and
substrate binding. The two binding sites
are close in space and there are positive
or negative cooperative
interactions
in
binding to these sites (7,8). Cooperativity
is seen in the decrease in affinity for anions
to the anion-binding
site associated with
the ionization of the zinc-bound water. This
cooperativity has been studied in this work
by using anions such as cyanide, hydrosulfide, and diethyldithiocarbamate,
which

should be addressed.
159

0003-9861/84 $3.00
Copyright
All rights

Q 1984 by Academic Press, Inc.

of reproduction
in any form reserved.

160

SYVERTSEN

AND

are known to be good metal ligands. These

anions, together with azide, cyanate, thiocyanate, tetrazole, iodide, and decanoate,
have been used to determine the relative
importance of the anion-binding site and
the catalytic zinc ion in complexing anions.
MATERIALS

AND

METHODS

Reagents. The enzyme horse liver alcohol dehydrogenase (EC 1.1.1.1.) was from Boehringer-Mannheim.
The purity and assay were as previously described
(2). Before use the enzyme was dialyzed against three
changes of a lOO-fold vol of 30 mM Hepps/NaOH
buffer, pH 8.2. The buffers Mes, Tes, and Hepps were
obtained as free acids from Sigma. Borax, 2,2-dipyridyl, and t-butanol were from Merck. Acridine orange
(3,6-dimethylaminoacridinium
chloride)
was from
Serva and CoClz. 6HaO from Riedel-de Haen. Iodoacetic acid and iodoacetamide were from Sigma. Iodoacetic acid was recrystallized
twice from petroleum
ether and iodoacetamide
from water. All other reagents were the purest commercially
obtainable.
Double quartz-distilled
water was used for all solutions.
Inactivations.
Enzyme inactivations
were carried
out at 23.5C using an enzyme concentration
of about
5 pM. For inactivations
with iodoacetamide 10 mM of
reagent was used. Buffer concentrations
were 20 mM
Mes at pH 6.1, 20 mM Hepps at pH 8.2, and 20 mM
borax at pH 10.0.
Data processing. Linear first-order
plots of log(activity) against time were treated by linear regression
analysis. The data obtained for the observed firstorder rate constant, /cobs, as a function of the concentration
of iodoacetate and one or two effecters
were treated by the derivative-free
nonlinear regression program BMDPAR (copyright 1981, University
of California,
Los Angeles). Computing was carried
out on the DEC-10 machine of the University of Oslo.
Cobalt(II)-substituted
enzyme. The catalytic zinc ion
was selectively removed by dialysis (9). Insertion of
Co(I1) ions was performed with dissolved enzyme apohybride in 25 mM Tes, pH 8.2. A 1.5-fold excess of
CoCl* * 6HaO in relation to the concentration
of subunits was used. Absorption spectra were recorded with
a Hewlett-Packard
8450A spectrophotometer.
The
concentration
of apo-hybride was determined using
the molar absorptivity
of e280= 36000 M-l cm- (9),
based on M, = 80,000 (10).
2,2-dimridyl
as an optical probe. The binding of 2,2-

* Abbreviations
used: Mes, I-morpholineethanesulfonic
acid; Tes, 2-{[2-hydroxy-l.l-bis(hydroxymethyl)ethyl]amino}ethanesulfonic
acid; Hepps, 4-(2hydroxyethyl)-1-piperazinepropanesulfonic
acid.

MC KINLEY-MC

KEE

dipyridyl to the catalytic zinc ion in the presence of

effector ligands was studied spectrophotometrically
at 307 nm (11). The observed absorbance, Aoh, as a
function of the concentration
of 2,2-dipyridyl
was fit
to
A ohs= l/2 [Et + dP, + Kapp
- \IEf + dP: + K:,,

derived

2E&t + 2.WGpp
+ ~PJG,,I,
PI

from

MW

A = e.l.[EdP],
Et = [E] + [EdP],

Kspp = and

[EdP]

dP, = [dP] + [EdP],

where c is the molar absorptivity

of the enzyme -2,2dipyridyl complex (EdP); Et is the total concentration
of enzyme; dP, is the total concentration
of 2,2-dipyridyl; and Kapp is the apparent dissociation constant
of the EdP complex.
Nonlinear regression analysis of Eq. [l] was performed with the BMDPAR program. Et was based on
active site titrations
with NADH in the presence of
isobutyramide
(12).
The dissociation constants of effector ligands, K,,
were calculated from

KL = [Ll/((&/s) - 1).

PI

In the presence and absence of effector ligand (L), S,

and S refer to the slope of the straight lines in a
double-inverse
plot of Aobe against the concentration
of free 2,2-dipyridyl,
as calculated from the values of
t and Kapp.
Theory. In affinity labeling a reversible enzymeinactivator
complex (EI) is formed prior to the formation of the inactivated
enzyme.
KI
In
E + I + Ei - E(inactive)

Scheme1

K, is the dissociation constant of the enzyme-inactivator complex, and b is the maximum first-order
rate constant for inactivation.
With excess inactivator(1) a pseudo-first-order
reaction is observed with
respect to active enzyme, for which the observed firstorder rate constant is given by
kbs

MWVG + [II).

[31

The parameters K, and b may be calculated from a

double-inverse plot of k+ against [I], or from nonlinear
regression analysis.
With one effector present, which in this work are
various nitrogen and sulfur ligands, the general
Scheme II has been used to describe the influence of
effector upon the affinity labeling reaction (6).

ZINC

BINDING

IN HORSE

LIVER

ALCOHOL

DEHYDROGENASE

k2
EI - E(inactive)

161

mixed-type inhibitor
with respect to iodoacetate. All the other nitrogen ligands
+
+
were activators of the affinity labeling reaction. Typical examples of double-inverse
plots in the presence of an activator are
K i
q
shown for thiazole at pH 8.2 in Fig. 2 and
EL + I-& EIL az E(inactive)
for cyclohexylamine
at pH 10.0 in Fig. 3.
Scheme II
Activation is demonstrated by the ordinate
intercept in the presence of activator being
Here KL is the dissociation constant of the enzymebelow the intercept without activator. This
effector complex, a is an interaction or cooperativity
is reflected by the parameter @> 1 (Scheme
constant, and fl is a constant affecting the maximum
II). Figures 2 and 3 differ regarding the
rate constant. For scheme II kob is expressed by
pattern of the intercepts on the abcissa.
k,[I](l + flL]/aKd
This is a consequence of an altered interkobs=
t41
KIU + IL]/&) + Ul(1 + [Ll/aK,)
action constant cy. In Fig. 2 (Y > 1 means
Loading the BMDPAR program with this function
there is negative cooperativity in the bindalong with the data for the dependent variable kobs ing of iodoacetate and thiazole to the enand the appropriate variables [I] and [L], allows dezyme at pH 8.2, while in Fig. 3 CY< 1 demtermination
of the parameters (Y, 8, and KL; KI and
onstrates positive cooperativity
in the
b were determined previously (13).
binding of iodoacetate and cyclohexylamine to the enzyme at pH 10.0. Table I lists
RESULTS
the dissociation constants KL and the paThe e#ect of sme neutral nitrogen and rameters (Y and p according to Scheme II
sulfur ligands. The influence of effector li- for the nitrogen and sulfur ligands exgands upon affinity labeling with iodoacamined.
For the activators cyclohexylamine,
pietate is illustrated by double-inverse plots
of kobsversus iodoacetate concentration at peridine, and 4-dimethyl-amino
pyridine,
different effector concentrations.
Figure 1 the experiments were performed at pH 10.0
to ensure some of the effector being in the
is such a plot at pH 8.2 with the effector
ligand 2,2-dipyridyl,
which is a partial
base form.
E+I2

[2.2- Dipyridyl]

- 0.4

-a2

0.2

a6

I/ Dodoacetate]
FIG. 1. Double-inverse
plot of the observed first-order
concentration
for the inactivation
of horse liver alcohol
Buffer, 20 mM Hepps, pH 8.2.

(mbl)

1.0

(mMr

rate constant,
dehydrogenase.

koba, versus iodoacetate

Effect of 2,2-dipyridyl.

162

SYVERTSEN

AND

MC KINLEY-MC

[Thiazole]

-0.2

KEE

(mM)

Q6

0.2

I/ [lodoacetate]
FIG. 2. Effect of thizole.

Buffer,

Acridine orange was an activator at pH

8.2, mainly by facilitating
the reversible
binding of iodoacetate (a < 1). At pH 6.1,
7.0, 9.1, and 10.0 with 300 PM acridine orange no activation was observed. Acridine
orange only slightly affected the affinity
labeling reaction with a-bromo-&imidazolyl propionate by facilitating
the binding
of the label. Acridine was almost an uncompetitive inhibitor
with respect to iodoacetate. Against cu-bromo-fl-imidazolyl
propionate, acridine was a mixed inhibitor.

20 mM Hepps, pH 8.2.

Assuming linear inhibition,

KL was 60 PM
at pH 8.2.
1,2,4-Triazole (10 mM) or urea (0.1 M) did
not influence the affinity labeling reaction.
With mercaptoethanol,
dimercaptopropanol, and penicillamine
as effecters the
experiments were performed at pH 6.1 to
keep the concentration
of thiolate which
reacts readily with iodoacetate low. With
mercaptoethanol
a partial mixed-type inhibition occurs at low concentrations of iodoacetate while activation is observed at

[Cyclohexylamine]

l/ [lodoacetate]
FIG. 3. Effect of cyclohexylamine.

1.0

(mM)-

Buffer,

(mM)

(mM)-
20 mM borax, pH 10.0.

ZINC

BINDING

IN HORSE

LIVER

high concentrations. Penicillamine was an

activator and dimercaptopropanol an inhibitor at all iodoacetate concentrations.
The efect of some anions upon afinity
labeling with iodoacetate in the absence and
presence of the activator imidazok.
Cyanide,
cyanate, thiocyanate, diethyldithiocarbamate, azide, iodide, and tetrazole are competitive inhibitors with respect to iodoacetate (data not shown).
Additional experiments at high inhibitor
concentrations demonstrated that inhibition was linear. For cyanide the anionic
form is the binding species as no inhibition
was observed at pH 6.1. Hydrosulfide anion
on the other hand was like mercaptoethanol, a partial mixed-type effector, showing
an activation effect with /3> 1 at high concentrations of iodoacetate (Table I).
KLZ

EL2 +Lz
+

E+l

163

DEHYDROGENASE

To further characterize the binding sites

for the anions they were added as second
effecters together with imidazole. Imidazole is known to coordinate the catalytic
zinc ion and activate the affinity labeling
reaction with iodoacetate in the same
manner as thiazole in Fig. 2 (6). Figure 4
is a double-inverse plot of kDbsagainst iodoacetate concentration at a fixed concentration of 2.0 mM azide and different
imidazole concentrations. Saturating concentrations of iodoacetate abolish the inhibition with azide at all imidazole concentrations as seen from the common ordinate intercepts. The different slopes in
the presence and absence of azide with a
saturating concentration of imidazole indicate a ternary enzyme-imidazole-azide
complex. Where L1 is imidazole, Scheme
III describes the inhibition by azide as the
second effector L,:
KI
r

EI?

E(inactive)

Ll

-&.I

ALCOHOL

Ll

KL1
J

ELILB s

Lz + EL1 + I s EILi a%E(inactive)

Scheme III

This is an extension of scheme II where KL2 is the dissociation constant for the second
effector from the enzyme and y is an interaction constant. Equation [5] expresses kobs
as a function of [I], [L,], and [L,]:
k

&JI](l + ,&]/aKd

Ohs
= KAl + M/&I

+ LI/KLz + MLI~Y&JL)

Iodide, thiocyanate, and tetrazole showed

much the same inhibition pattern as azide.
In the case of cyanide, diethyldithiocarbamate, and decanoate, saturating concentrations of iodoacetate or imidazole
abolish the inhibition of inactivation with
iodoacetate. Thus, no ternary enzyme-imidazole-anion complex is indicated. The
parameters of Scheme III for the anions
result from nonlinear regression of Eq. [5],
and are listed in Table II.

+ [IN1 + LI/~Kd

[51

tion constant of the effector ligands, calculated from Eq. [2]. For cyanide a plot of
absorbance against the concentration of
2,2-dipyridyl deviated from the other plots
in that the initial part of the curve was
sigmoidal. This resulted in higher standard
deviations in the results. 2,2-dipyridyl is
displaced from the enzyme by cyanide, hydrosulfide, diethyldithiocarbamate,
and
dimercaptopropanol as indicated by the
The inJEuence of eflector ligands on the values of KaPP. Except for cyanate (see Disbinding of .Z,.%dipyridyl. Table III lists KaPP cussion) there is good agreement between
and c307
for the binding of 2,2-dipyridyl cal- the values of KL in Tables II and III.
culated from Eq. [l] and KL, the dissociaInactivation
of the enzyme by iodoacet-

164

SYVERTSEN

AND

MC KINLEY-MC

TABLE

KEE

BINDINGOFNITROGEN-AND SULFUR-CONTAINING
LIGANDSTOHORSE LIVER ALCOHOL
DEHYDROGENASE
ACCORDINGTO SCHEMEII
Effector ligand
2,2-Dipyridyl
4-dimethylamino-pyridine
I-Methylpyridine
Pyridine
3-Chloropyridine
I-Cyanopyridine
Pyrazine
Imidazole
Imidazole
N-Methylimidazole
Thiazole
Acridine orange
Cyclohexylamine
Piperidine
Ammonia
Hydrogen sulfide
Mercaptoethanol
Penicillamine
Dimercaptopropanol

PH

P&

KL* 6-Q

8.2

4.44

0.55 + 0.1

0.61 + 0.13

10.0

10.15

0.71 rt 0.05

8.2
8.2
8.2
8.2
8.2
8.2

6.08
5.21
3.07

0.32 + 0.02
2.5 3~ 0.3
1.5 r 0.2

0.36
2.2
1.7
1.5

1.95

9.1 k 2.6

10.0
8.2
8.2
8.2

0.6
7.00
7.00
7.06
2.53

10.0
10.0
10.0

10.1
10.6
11.0
9.26

8.2

7.04

6.1
6.1
6.1

9.43
-

67
1.4
5.6
0.24
11
0.06

rt 13

+ 0.1
+ 0.1
* 0.02
+ 0.8
+ 0.02

1.8 k 0.1
0.83 5 0.1
220 2 33
0.07 + 0.01

1.2 f 0.1
32 k 13
3 f 1 (PM)

+
+
+
+

0.07
0.2
0.3
0.2

1.7 k 0.3
4.5 + 1.5
2.4 f 0.3
0.47 + 0.02
1.2 -c 0.2
2.0 + 0.3
0.56 f 0.07
0.34 zk 0.04
0.39 + 0.04
0.93 + 0.3
60 + 6
4.2 5 0.8
4+1
4.3 f 0.8

0.10 + 0.03
4.6
2.7
3.6
2.5

k 0.3

+ 0.1
k 0.3

+ 0.1
1.7 k 0.2

5.4 f 1.7
6.8 5 0.4
5.3 + 0.1
4.8 31 0.3
5.9 * 0.5
1.2 t- 0.1
5.4 + 0.3
1.4 f 0.1
9.4 -c 3.2
6+3
1.4 + 0.1

10 (est.)
0.9 32 0.07

aK, is the proton dissociation constant. For the nitrogen bases it refers to the conjugated acid (14, 15).
* KL is the dissociation constant for the ligands from the enzyme.
cxis the interaction constant for simultaneous binding of iodoacetate and ligand to the enzyme.
d/I is a constant affecting the maximum rate of inactivation.
ePrevious determination (8).

Iodoacetamide is the neutral analog

of iodoacetate.
With pseudo-first-order
conditions the second-order rate constant
for inactivation
was calculated using the
concentration
of iodoacetamide. Figure 5
shows that the rate of inactivation
increases 5-fold in the presence of a saturating concentration
of cyanide, 11-fold
with diethyldithiocarbamate,
and l&fold
with hydrosulfide ion. While the plots of
log activity versus time were strictly linear
in the presence of cyanide, with hydrosulfide and diethyldithiocarbamate
they
were curved due to reaction with iodoacetamide. In the presence of saturating
concentrations
of tetrazole or decanoate
the inactivation
rates were 2.3 and 1.3
times faster, respectively. Cyanate (12 mM)
or thiocyanate (10 mM) did not affect the
rate, while 10 mM azide, 20 mM iodide, and
100 mM chloride caused an approximately
twofold decrease in the rate.
amide.

Ligand binding to Co(II)-substituted en-

zyme. To examine ligand coordination

to
the metal ion in the enzyme, cobalt (II)
was substituted as an optical probe for zinc
in the active site. Figure 6 shows the absorption spectra for the cobalt enzyme in
the absence and presence of cyanate, mercaptoethanol, and 4-methylpyridine.
4-Cyanopyridine
gave essentially
the same
spectrum as 4-methylpyridine.
Decanoate
did not perturb the spectrum for the cobalt
enzyme in the visible, but the difference
spectrum showed maxima around 306 nm
and a larger absorption up to 600 nm. A
similar increase in absorption between 300
and 350 nm to that with decanoate was
observed on adding 0.15 M chloride to the
cobalt enzyme.
DISCUSSION

Activation of a@@ labeling with iodoacetate. The effect of the ligands in Table

ZINC

-0.2

BINDING

0.2

l! [lodoacetate]

IN HORSE

06

LIVER

1.0

(rnM~

FIG. 4. Effect of 2.0 mM azide upon the activation

by imidazole of the affinity labeling reaction with
iodoacetate. Double-inverse plot of the observed firstorder rate constant, kabs, versus iodoacetate concentration. Buffer, 20 mM Hepps, pH 8.2. Imidazole: (0)
1.0; (0) 2.5; (A) 20 mM (- - -) is for the corresponding
line without azide.

I on the affinity labeling reaction with iodoacetate results from their coordination
to the catalytic zinc ion of liver alcohol
dehydrogenase. For imidazole the observed
activation has been considered as an innersphere ligand effect, where acting as a
metal ligand, imidazole is a better electron
donor than water (6,16). Substituting
for
water, imidazole increases the electron
density on the metal and on Cys-46, which
becomes a better nucleophile and more reactive. This is supported by the present
results as all the monodentate nitrogen and
sulfur ligands show the same activation
pattern as imidazole. It also correlates with
the better donor properties of sulfur and
nitrogen compared to oxygen.
Unlike the other nitrogen compounds,
2,2-dipyridyl does not stimulate affinity labeling but is a mixed-type inhibitor.
For
2,2-dipyridyl
at pH 8.2 (Y < 1 and fi $ 1
which is the opposite of that found for all
the nitrogen compounds other than l,lOphenanthroline
(8). The low reactivity of
the ternary enzyme-2,2-dipyridyl-iodoacetate complex (p 4 1) indicates that steric
effects of the large bidentate ligand can
make the access of iodoacetate to Cys-46
difficult. CY< 1 means that 2,2-dipyridyl
facilitates the binding of iodoacetate to the
enzyme.

ALCOHOL

165

DEHYDROGENASE

Acridine orange (3,6-dimethylamino

acridine) has been reported to stimulate the
inactivation
by iodoacetate at pH 7.0 (17).
Our results confirm stimulation
but only
at pH 8.2. As acridine orange is not
competitive against cu-bromo-P-imidazolyl
propionate, binding to the catalytic zinc
ion is excluded as was also concluded by
Kovar et al. (17, 18). The observed activation is considered due to acridine orange,
pKa 10.1 (19) being a cation, as it mainly
affects the reversible binding of iodoacetate. Neutral acridine shows similarities
with l,lO-phenanthroline
and 2,2-dipyridyl.
However, acridine, like acridine orange,
does not bind to the catalytic zinc ion as
it is not competitive against cY-bromo-Pimidazolyl propionate.
The dissociation constant KL correlates
with the proton affinity or basicity of the
neutral nitrogen ligands. This is seen in
Table I for the closely related pyridine
compounds, where, with the exception of
3-chloropyridine,
there is a regular decrease of KL with increasing basicity. For
4-dimethylamino
pyridine KL extrapolated
to pH 8.2 is 87 pM. Thus 4-dimethylamino
pyridine with a pK, of 10.15 has a 103-fold
higher affinity for the enzyme than pyrazine with a pK, of 0.6. Pyrazole with KL
= 8 mM (20) and pK, = 2.52 (21) fits in with
the other heterocycles in Table I. An inTABLE

II

ANION BINDING TO HORSE LIVER ALCOHOL DEHYDROGENASE IN THE ABSENCE AND PRESENCE OF
IMIDAZOLE ACCORDING TO SCHEME III
Anion, Lz
Azide
Iodide
Tetrazole
Thiocyanate
Cyanate
Diethyldithiocarbamate
Cyanide
Decanoate
Decanoate

PH

KI.z (mM)

8.2
8.2
8.2
8.2
8.2

3.7
3.6
3.7
1.3
0.89

f
+
f
f
f

0.5
0.5
0.5
0.1
0.07

4.2 + 0.8
2.0 rf: 0.5
6.5 f 2.1
1.9 f 0.2
24+9

8.2
8.2
8.2
10.0

1.1
0.24
0.050
0.58

f
f
f
f

0.1
0.02
0.005
0.12

a,
03
>150
co

a KU is the dissociation constant of the anion from

the enzyme.
* y is the interaction constant for simultaneous
binding of imidazole and the anion.

166

SYVERTSEN

AND

MC KINLEY-MC

TABLE

KEE

III

TITRATIONOFALCOHOLDEHYDROGENASEWITH~,~-DIPYRID~INTHEPRESENCE
OFEFFECTORLIGANDSAT pH 8.2

Effector

Molar absorptivity
[em7 (lo4 M-' Cm-')]

ligand

Dimercaptopropanol
(6 PM)
Hydrosuhide
(170 PM)
Cyanide (0.6 mM)
Diethyldithiocarbamate
(3.5 mM)
Cyanate (6.7 mM)
Thiocyanate (4.0 mM)
Azide (12 mM)
Tetrazole (12 mM)

Kw;

1.71 + 0.06
1.33 + 0.05

1400 + 110

2.6 f 0.03
1.3 f 0.1

1880 f 400
1723 f 250

1080 f 70

1.1 f 0.03

763 f
227 f
480 f
460 +
400+15

2.26 + 0.01
2.04 k 0.03
2.01 f 0.06

hb (PM)

(PM)

1.79 f 0.02

3.3
67
262
900
4300

43
3
20
35

a Kapp is the apparent dissociation constant of 2,2-dipyridyl

from the enzyme in the presence of effector
ligands.
* KL is the dissociation constant of the effector ligands from the enzyme.

crease in affinity with increasing basicity

is of course characteristic for binding to a
metal ion. Compared to their affinity for
free zinc ions, the pyridine derivatives have
a higher affinity while ammonia has a
lower affinity for the active-site zinc ion.
Hydrophobic interactions between the
proteins and the ligands probably accounts
for this although such comparisons require
data for relevant mixed-ligand complexes.
The relative importance of the metal ion
and the protein for the affinity of a ligand
is illustrated by the imidazole-like compounds. Imidazole is a 4.5-pH unit stronger
base than thiazole and has an eightfold

b I

1.0

[Effector]

I I
2.0

higher affinity for the enzyme, as is to be

expected for binding to a metal ion. On the
other hand, N-methylimidazole, which is
as strong a base as imidazole, has a sixfold
higher affinity for the enzyme. Here the
stabilizing effect of a methyl group due to
hydrophobic interactions with the protein
is manifest.
Positive cooperativity in the binding of
iodoacetate and the neutral nitrogen ligands to the enzyme at pH 10.0 (cu < 1,
Table I) indicates a higher affinity of the
second ligand for the binary complex between the first ligand and the enzyme than
for the free enzyme. In the case of imidazole, positive cooperativity results from

#J

(rnbl)

FIG. 5. Inactivation
of horse liver alcohol dehydrogenase by iodoacetamide. The second-order rate constant, k, for inactivation
versus the concentration
of
(0) hydrosulfide; (0) cyanide; and (Cl) diethyldithiocarbamate. Buffer, 20 mM Hepps, pH 8.2.

460

560

700 "In

FIG. 6. Absorption spectrum of 25 PM Co (II)-substituted horse liver alcohol dehydrogenase (-)

and
its binary complex with 6 mM cyanate (- - -), 6 mM
mercaptoethanol
( . . . ), and 2.5 mM 4-methylpyridine
(-a -. -). Buffer, 25 mM Tes, pH 8.0.

ZINC

BINDING

IN HORSE

LIVER

the removal of the zinc-water

ionization
when imidazole substitutes for the water
molecule bound to the catalytic zinc ion
(8). As expected, removal of the zinc-bound
OH- deletes the electrostatic repulsion between iodoacetate and OH-.
Compared to mercaptoethanol
the 400fold higher affinity for dimercaptopropanol
(Tables I and II) indicates bidentate coordination.
For 4-methylpyridine
and
mercaptoethanol
the perturbations
in the
electronic spectrum of the cobalt-substituted enzyme supports binding to the catalytic metal.

Binding sites for the anions cyanide, cyanate, thiocyanate, diethyldithiocarbamate,

hydrosul&de, a&de, tetraxole, and iodide.
Hydrosulfide
anion forms a ternary enzyme-HS--iodoacetate
complex as is evident from the activation (p > 1) of affinity
labeling with iodoacetate. The acceleration
of the rate of alkylation of Cys-46 by both
iodoacetate and iodoacetamide is explained
by hydrosulfide anion coordinating
to the
catalytic zinc ion and increasing the reactivity of Cys-46. The effects of hydrosulfide
parallel the effects of hydroxyl as both anions also greatly decrease the affinity of
iodoacetate for the anion binding site
(4, 13).
Unlike hydrosulfide,
the other anions
were competitive inhibitors
with respect
to iodoacetate. This suggests binding to the
anion-binding
site. However, for cyanide
and diethyldithiocarbamate
other evidence
shows that these ligands coordinate to the
catalytic zinc ion. They compete with both
imidazole (Table II) and 2,2-dipyridyl
(Table III) for binding to the enzyme. Also,
like hydrosulfide,
cyanide and diethyldithiocarbamate
accelerate the rate of inactivation with iodoacetamide (Fig. 5). The
observed competition between iodoacetate
and cyanide or diethyldithiocarbamate
for
binding indicates a large negative cooperativity due to electrostatic repulsion in
the simultaneous binding of the two anions.
The dissociation
constant of cyanide
from the enzyme at pH 8.2 corrected to the
base form is as low as 21 pM. This indicates
a strong metal interaction and is in agreement with the known properties of cyanide
as a good zinc ligand. For comparison the
dissociation constant of cyanide from bo-

ALCOHOL

DEHYDROGENASE

167

vine carbonic anhydrase is 2.6-3.2 pM and

for hydrosulfide
2-11 pM at pH 7.5 (22).
This work demonstrates reversible binding
of diethyldithiocarbamate
to the catalytic
zinc. When dialyzing the enzyme in the
presence of excess diethyldithiocarbamate
for 72 h, selective loss of the catalytic zinc
ions has been reported (23).
For azide, thiocyanate,
iodide, and tetrazole with y > 1 (Table II) there is negative cooperativity
in the binding of imidazole and these anions to the enzyme.
However, the negative interactions are not
infinite and ternary enzyme-imidazoleanion complexes are formed. Iodide and
thiocyanate have previously been shown to
increase the affinity of 2,2-dipyridyl
to the
enzyme by a factor of 1.2-1.8 (24). That
agrees with the result in Table III for thiocyanate. Also, the dissociation constant KL
for thiocyanate in Table II derived from
inactivation
kinetics is identical with the
value obtained from the effect of thiocyanate on the NADH association rate (24).
Azide and tetrazole at concentrations three
times the KL values (Table II) induce a 1.2fold decrease in affinity of 2,2-dipyridyl
to
the enzyme (Table III). This rules out
binding to the zinc ion in the case of these
four anions and indicates binding to the
anion-binding
site. Changes in the electronic spectra of the Co(II)- and Ni(II)substituted
enzymes on adding cyanide,
hydrosulfide, and azide have suggested direct metal coordination of these ligands (9,
25). However, in the case of azide there is
either a difference between the two metals
and zinc in their preference for ligands or
the electronic spectra are sensitive to the
binding of azide to other sites in the vicinity of the metal.
Cyanate is a special case in that different
values of KL are obtained from titration
with 2,2-dipyridyl
(Table III) and from inactivation kinetics (Table II). This indicates that both zinc and the anion-binding
site are alternative binding sites with an
approximately
equal affinity for cyanate.
The spectrum of the cobalt-substituted
enzyme in the presence of cyanate is similar
to that reported with azide (9). However,
the spectrum is not conclusive regarding
whether or not there is binding to the
metal.

168

SYVERTSEN

AND

Binding of decanoate to the free enz~w~

Decanoate is generally not a good metal
ligand, but has been proposed to coordinate
the catalytic zinc ion in the free enzyme
(26). This is surprising as the catalytic zinc
ion is considered a weak positive center
and anions usually prefer to bind to the
anion-binding
site unless coenzyme, in
particular the positive pyridinium
ring of
NAD, is present.
Decanoate, like cyanide, is competitive
with iodoacetate (2) and imidazole (Table
II). Thus either the negatively
charged
carboxyl of decanoate attaches to the zinc
ion and so prevents binding of iodoacetate
to the anion-binding
site by electrostatic
repulsion, or decanoate attaches to the anion-binding
site and sterically covers the
catalytic zinc ion. The electronic spectrum
of the cobalt-substituted
enzyme does not
clarify how decanoate binds as chloride,
which does not bind to the zinc ion in the
absence of coenzyme, causes similar spectral perturbations
with the cobalt enzyme
between 300 and 350 nm. The increased
absorption above 350 nm was nevertheless
not seen with chloride.
Previous experiments indicated that decanoate binds to a different site than shortchain fatty acids. Thus with free enzyme
the fatty acid derivatives iodoacetate, bromoacetate, and bromopropionate
bind to
the anion-binding
site and form ternary
complexes with imidazole and the enzyme
(6). Inhibition
experiments using ethanol
as substrate show that iodoacetate is competitive with respect to NAD+ (27), while
decanoate is almost uncompetitive
(28).
Likewise, the use of 2,2-dipyridyl
as an optical probe for binding to the catalytic zinc
failed to give the correct affinity for acetate
to the active site while succeeding in the
case of decanoate (11). The different findings suggest that with the free enzyme decanoate is attached to the zinc ion but
short-chain fatty acids to the anion-binding site. With the enzyme and NAD+ both
types of acids form strong ternary complexes, and in this case are bound to the
catalytic zinc ion (11,28). Fatty acids such
as decanoate represent
a special case
among anions where with the free enzyme,
in the absence of coenzyme, the long hydrophobic chain orientates the molecule in

MC

KINLEY-MC

KEE

the hydrophobic substrate pocket so that

the carboxyl can coordinate to zinc. The
anionic pyrophosphate group of coenzyme
binds to Arg-47 at the anion-binding
site.
In the presence of coenzyme the anionbinding site is therefore not available and
anions can bind to zinc particularly
when
stabilized by the positive pyridinium
ring
of NAD+. Thus of the anions used here
thiocyanate (29), azide (30), and iodoacetate
(27) are all competitive with ethanol when
using excess NAD+, indicating binding to
the catalytic metal. The same applies to
chlorate and fluoride and to chloride and
bromide, which in alcohol oxidation stimulate turnover by also forming a more labile E-NADH-anion
complex (29).
The experiments
show the binding to
zinc of ligands containing N- and S-donor
atoms. Affinity is mainly determined by
the basicity of the neutral ligands. However, anions bind to the free enzyme at the
coenzyme anion-binding
site unless they
possess strong metal-binding
properties
(CN-, HS,
diethyldithiocarbamate)
or
suitable orienting hydrophobic groups (decanoate).
ACKNOWLEDGMENT
The support of the Norwegian
Research
Council for
Science and the Humanities
(NAVF)
is acknowledged.

REFERENCES
1. BRHNDCN, C. I., JBRNVALL,
H., EKLUND, H., AND
FURUGREN, B. (1975) in The Enzymes
(Boyer,
P. D., ed.), 3rd ed., Vol. 11, pp. 104-190,
Academic Press, New York/London.
2. REYNOLDS, C. H., AND MCKINLEY-MCKEE,
J. S.
(1969) Eur. J. Bioch~wz. 10,474-478.
3. DAHL, K. H., AND MCKINLEY-MCKEE,
J. S. (1977)
Eur. J. Biochem. 81,223-235.
4. EVANS, N., ANDRABIN, B. R. (1968) Eur. J. Biochem

4.548-554.
5. BOIWE, T., AND BRHND~N, C. I. (1977) Eur. J.
Biochem 77,173-179.
J. S. (1981)
6. DAHL, K. H., AND MCKINLEY-MCKEE,
Eur. J. Biochem 120,451-459.
ANDERSSON,
P., KVASSMAN,
J., LINDSTR~M,
A.,
OLDEN, B., AND PETTERSSON, G. (1980) Eur. J.
B&hem. 108,303-312.
SWERTSEN, C., AND MCKINLEY-MCKEE,
J. S. (1983)

Arch

Biochem. Biophys.

223,213-223.

MARET, W., ANDERSSON, I., DIETRICH, H., SCHNEIDER-BERNL~~HR,

H., EINARSSON, R., AND ZEP-

ZINC

10.
11.
12.
13.
14.
15.

16.

17.
18.
19.
20.

BINDING

IN HORSE

LIVER

PEZAUER, M. (1979) Eur. J. Bi~chem. 98, 501512.

J~RNVALL, H. (1970) Eur. J. Biochem. 16,25-40.
SIGMAN, D. S. (1967) J. Biol. Chem 242.3815-3824.
THEORELL, H., AND MCKINLEY-MCKEE, J. S. (1961)
Acta Chem Scone!. 15, 1797-1810.
SWERTSEN, C., AND MCKINLEY-MCKEE, J. S. (1981)
Eur. J. Biochem 117, 165-170.
SMITH, R. M., AND MARTELL, A. E. (1975) in Critical
Stability Constants, Vol. 2, Plenum, New York.
SILL~N, L. G., AND MARTELL, A. E. (1964 and 1971)
in Stability Constants of Metal-ion Complexes,
Special publications No 17 and 25, The Chemical
Society, London.
HOUGHTON, R. P. (1979) in Metal Complexes in
Organic
Chemistry,
Cambridge
University
Press, London.
KovAii, J., D~~RROVA, E., AND SKURSK?, L. (1979)
Eur. J. B&hem. 101,507-514.
Kovi\ii, J., AND D~~RROVA, E. (1981) CoU. Czech
Chem Commun. 46,2268-2277.
ALBERT, A. (1979) in Selective Toxicity, 4th ed.,
pp. 455, Chapman & Hall, London.
ANDERSSON, P., KVASSMAN, J., LINDSTR~M, A.,
OLDEN, B., AND PETTERSSON, G. (1981) Eur. J.
Biochem. 114,549-554.

ALCOHOL

DEHYDROGENASE

169

21. PERRIN, D. D. (1972) Dissociation

Constants of
Organic Bases (Supplement),
Buttersworth,
London.
22. BERTINI, I., LUCHINAT, C., AND SCOZZAFAVA, A.
(1982) in Structure and Bonding, Vol. 48, pp.
46-91, Springer-Verlag,
Berlin.
23. DRUM, D. E., LI, T.-K., AND VALLEE, B. L. (1969)
Biochemistry f&3783-3791.
24. OLD&N, B., AND PETTERSSON, G. (1982) Eur.
B&hem. 125,311-315.

J.

25. DIETRICH, H., MARET, W., KO~LOWSKI, H., AND

ZEPPEZAUER, M. (1981) J. Iwg.
Biochem. 14,
297-311.
26. ANDERSSON, P., KVASSMAN, J., OLDEN, B., AND
PE~ERSSON, G. (1981) Eur. .I. Biochem 118,
119-123.
27. MORRIS, D. L. (1972) Ph.D. thesis, University
Edinburgh, Scotland.

of

28. THEORELL, H., AND MCKINLEY-MCKEE, J. S. (1961)

Acta Chem. Stand. 15, 1834-1865.
29. PLANE, R. A., AND THEORELL, H. (1961) Acta Chem.
Scar& 15, 1866-1874.
30. YOUNG, J. M., AND WANG, J. H. (1971) J. Biol.
Chem. 146,2815-2821.