9360 J . Org. Chem., Vol. 61, No. 26, 1996
Schwartz et al.
suggests that there is no room in the active site for the
extra hydroxyl group so that binding is impeded.
tion and observation of an enzyme-hemithioacetal ad-
duct may be complicated by enzyme binding of the CoA
and the free aldehyde present in equilibrium with the
hemithioacetal. The results of this work should help to
define the proper conditions for observation of an enzyme-
hemithioacetal adduct. While the inhibition constants
are modest, the CoA hemithioacetals may be useful tools
for studies of enzyme-inhibitor complexes and will be
readily accessible in labs in which previously studied
acyl-CoA analogs have not been available.
The hemithioacetal 2d was expected to mimic the
tetrahedral intermediate or transition state 7 in the
reaction of succinic thiokinase. Studies of succinic thioki-
nase were complicated by much more potent inhibition
by the aldehyde than that observed in the studies of
chloramphenicol acetyltransferase. This inhibition ap-
pears to be nonspecific and probably not indicative of
binding of succinic semialdehyde in the succinic acid
binding site. This is supported by the noncompetitive
nature of the inhibition and the fact that other aldehydes
were also found to inhibit this enzyme. Inhibition by CoA
was modest but significant. Inhibition by the combina-
tion of CoA and succinic semialdehyde was significantly
greater than inhibition by either alone. However, since
inhibition by both was significant, it was difficult to
discount the possibility of the observed inhibition being
due primarily to an additive effect of CoA and aldehyde
rather than inhibition by the hemithioacetal. In order
to address this issue, inhibition by dethia-CoA was
studied. This compound has all of the functionality of
CoA except for the thiol group and thus cannot form a
hemithioacetal with an aldehyde. Inhibition of succinic
thiokinase by dethia-CoA was indistinguishable from
inhibition by CoA. However, the combined effect of
dethia-CoA and aldehyde was almost identical to the
effect of aldehyde alone and much less than the effect of
CoA plus aldehyde. This verifies that the additional
inhibition over that exhibited by aldehyde alone is due
primarily to the hemithioacetal and not to free CoA.
Several acyl-CoA analogs have been prepared previ-
ously which mimic the enol or enolate intermediate in
reactions of enzymes that catalyze deprotonation of the
methyl group of acetyl-CoA as part of catalysis, most
notably citrate synthase.8-13 However, few acyl-CoA
analogs have been prepared that posess a tetrahedral
center in place of the thioester carbonyl carbon, which
may mimic the tetrahedral intermediate or transition
state in the reactions of enzymes that catalyze acyltrans-
fer. Exceptions are â-hydroxyalkyl-CoA thioethers, dif-
fering from the hemithioacetals described here in that
they contain an additional methylene group between the
sulfur atom and the hydroxymethine carbon.30,31
Also recently reported have been sulfoxide and sulfone
analogs7 and secondary alcohol analogs, differing from
the hemithioacetals by replacement of the sulfur atom
with a methylene group.27 Among these, only the sec-
ondary alcohol analogs have shown enhanced binding
relative to the natural substrate for an enzyme catalyzing
acyl transfer from a CoA ester.27 The hemithioacetal
analogs are unique both in their ready accessibility and
in their potential to mimic the tetrahedral intermediate
or transition state in enzyme-catalyzed acyl transfer.
Indeed, the hemithioacetals are the best acyl-CoA analog
inhibitors of these enzymes described thus far.
Acyl-CoA analogs have been very valuable for studies
of the structures of enzyme-inhibitor complexes in part
because the actual enzyme-CoA ester complexes are often
unstable due to hydrolysis of the thioester. While the
hemithioacetals are kinetically unstable, they are suf-
ficiently stable thermodynamically that a high ratio of
hemithioacetal to free CoA can be obtained. Still, forma-
Exp er im en ta l Section
Deter m in a tion of Equ ilibr iu m a n d Dissocia tion Con -
sta n ts for Hem ith ioa ceta ls. Equilibria for hemithioacetal
formation of CoA with 1a and 1b were measured by 400 MHz
1H NMR. NMR samples were prepared by lyophilization of
an aqueous solution of potassium phosphate (pH 6.9) and
dissolving the residue in D2O to a final concentration of 5 mM.
In 0.5 mL samples of this buffer were then dissolved 9.6 mg
of CoA (12.5 µmol) and 1-1.5 equiv of the aldehyde. Equilibria
were determined by integration of the methylene protons next
to the thiol of CoA (2.58 ppm) relative to the same protons of
2a (2.74 ppm) and 2b (2.73 ppm). NMR spectra were obtained
at 5, 15, and 30 min. The equilibrium for trifluoroacetaldehyde
and CoA was determined by addition of 5 µL of an equilibrated
(for 24 h) solution of CoA (12.5 µmol) and 1c (30 µmol) in
aqueous HEPES buffer (0.5 mL) to a solution of DTNB (0.4
mM) in 0.1 M HEPES buffer (pH 7.0) and monitoring the
initial rapid increase in absorbance at 412 nm. The equilib-
rium was also determined by 19F NMR in 20% D2O in H2O
containing 5 mM phosphate (pH 6.9), 12.5 µmol of CoA, and
10 µmol of 1c. The sample was equilibrated at room temper-
ature for 5 h before analysis. The trifluoroacetaldehyde
hydrate (3c) was observed at 75.9 ppm and the hemithioacetal
2c at 83.8 ppm, relative to C6F6 at 0.0 ppm. The equilibrium
for succinic semialdehyde and N-acetylcysteamine was studied
by 500 MHz 1H NMR with samples containing 50 mM
potassium phosphate, pH or pD 7.4. Spectra were obtained
using water suppression, and equilibria were determined by
integration of both pairs of methylene protons of N-acetylcys-
teamine (3.33 and 2.62 ppm) and the hemithioacetal (3.37 and
2.68-2.88 ppm).
Dissociation constants for 2a and 2c were determined by
addition of 5 µL of an equilibrated solution of CoA (12.5 µmol)
and 30 µmol of 1a or 1c in aqueous HEPES buffer to a solution
of DTNB (0.4 mM) in HEPES buffer (0.1 M, pH 7.0) and
monitoring the rate of increase in absorbance at 412 nm after
the initial burst.
En zym e In h ibition Stu d ies. For inhibition studies with
chloramphenicol acetyltransferase, assays were conducted as
described previously21 by measuring the decrease in thioester
absorbance at 240 nm where ꢀ was measured to be 3.27 × 103
M-1 cm-1
. Ki values were calculated from double-reciprocal
plots of 1/v vs 1/[acetyl-CoA] at three concentrations of
inhibitor. Assays were run in 0.1 M HEPES, pH 7, containing
0.1 mM chloramphenicol, 0.015-0.120 mM acetyl-CoA, 0.01
unit of chloramphenicol acetyltransferase, and inhibitor. For
2a and 2c, an aliquot of a solution containing aldehyde (31.25
mM) and CoA (12.5 mM) in phosphate buffer (pH 7.0)
equilibrated at room temperature (incubation time 30 min for
2a and 5 h for 2c) was used in the assay. For 1b, the buffer
in each assay contained 0.080 M acetaldehyde to which an
aliquot of CoA was added. Concentrations of acetaldehyde and
formaldehyde were quantified using yeast alcohol dehydroge-
nase.32 Quantitation was repeated after performing the assays
to insure that the aldehyde concentration did not change over
the course of the assays. Ki values were calculated from
double-reciprocal plots of 1/v vs 1/[acetyl-CoA] at four concen-
trations of each inhibitor.
(30) Zheng, G.-Q.; Hu, X.; Cassady, J . M.; Paige, L. A.; Geahlen, R.
L. J . Pharm. Sci. 1994, 83, 233.
(31) Rubenstein, P.; Dryer, R. J . Biol. Chem. 1980, 255, 7858.
(32) Bergmeyer, H. A. Methods of Enzymatic Analysis; Verlag
Chemie: Deerfield Beach, FL, 1984; Vol. VI, p 606.