Inhibition of human alcohol dehydrogenases by formamides

Article abstract of DOI:10.1021/jm9707380

Human alcohol dehydrogenase (HsADH) comprises class I (α, β, and γ), class II (π), and class IV (σ) enzymes. Selective inhibitors of the enzymes could be used to prevent the metabolism of alcohols that form toxic products. Formamides are unreactive analog

Full text of DOI:10.1021/jm9707380

1696
J . Med. Chem. 1998, 41, 1696-1701
In h ibition of Hu m a n Alcoh ol Deh yd r ogen a ses by F or m a m id es
J ohn F. Schindler, Kristine B. Berst, and Bryce V. Plapp*
Department of Biochemistry, The University of Iowa, Iowa City, Iowa 52242
Received October 31, 1997
Human alcohol dehydrogenase (HsADH) comprises class I (R, â, and γ), class II (π), and class
IV (σ) enzymes. Selective inhibitors of the enzymes could be used to prevent the metabolism
of alcohols that form toxic products. Formamides are unreactive analogues of aldehydes and
bind to the enzyme-NADH complex [Ramaswamy, S.; Scholze, M.; Plapp, B. V. Biochemistry
1997, 36, 3522-3527]. They are uncompetitive inhibitors against varied concentrations of
alcohol, and this makes them effective even with saturating concentrations of alcohols.
Molecular modeling led to the design and synthesis of a series of cyclic, linear, and disubstituted
formamides. Evaluation of 23 compounds provided structure-function information and
selective inhibitors for the enzymes, which have overlapping but differing substrate specificities.
Monosubstituted formamides are good inhibitors of class I and II enzymes, and disubstituted
formamides are selective for the R enzyme. Selective inhibitors, with Ki values at pH 7 and 25
°C of 0.33-0.74 µM, include N-cyclopentyl-N-cyclobutylformamide for HsADH R, N-benzylform-
amide for HsADH â1, N-1-methylheptylformamide for HsADH γ2, and N-heptylformamide for
HsADH σ and HsADH â1.
Ta ble 1. Active-Site Residues of Alcohol Dehydrogenases11,12
In tr od u ction
Human alcohol dehydrogenases (HsADH1) are targets
for inhibitor design since they oxidize alcohols to toxic
products and they can participate in the altered me-
tabolism of various compounds during ethanol consump-
tion.2,3 There are five enzymes with significant activity
on a variety of alcohols: class I (R, â, and γ), class II
(π), and class IV (σ). Selective, uncompetitive inhibitors
could be especially useful for controlling the metabolism
of alcohols since they are effective in the presence of
saturating concentrations of alcohols.4 Formamides and
amides bind preferentially to the enzyme-NADH com-
plex and are potent uncompetitive inhibitors against
varied concentrations of ethanol.5-8
The design of inhibitors for the human enzymes can
be based on inhibitors of horse liver alcohol dehydro-
genase (EqADH) due to the structural homology among
these enzymes.9,10 We used the three-dimensional
structure of EqADH complexed with NADH and N-
cyclohexylformamide to provide a starting model.8 Three-
dimensional structures of â and σ enzymes are known,9,10
and models of the R, γ2, and π enzymes were created by
making the appropriate amino acid substitutions in the
structure of EqADH (Table 1). Since the structures of
the active sites are different, we expected to be able to
produce selective inhibitors.
amino acid
Eq
Hs R
Hs â1
Hs γ2
Hs π
Hs σ
48
57
93
S
L
F
F
L
S
F
L
T
V
M
L
I
T
M
A
Y
V
S
F
L
I
V
M
L
I
T
L
F
Y
L
G
F
L
T
V
M
L
V
S
L
F
T
F
Y
L
L
S
F
F
T
V
E
I
T
M
F
L
I
110
116
117
140
141
143
294
306b
309b
318
Y
L
G
F
V
V
V
M
L
I
a
∆
F
M
T
V
M
F
F
V
a
Amino acid residue 117 in the human σ enzyme is deleted.
b
Residues provided by the second subunit.
bind to the enzyme-NADH complex. The inhibition
constants determined for the forward or reverse reaction
were about (within 2-fold) the same, which is expected
if coenzyme dissociation is rate-limiting in the ordered
mechanism. Most values were determined at pH 7, but
similar values were also obtained at pH 8. Inhibition
of EqADH by N-cyclohexylformamide is independent of
pH in the range 5.5-9.0 and decreases above a pK of
10.5.
Str u ctu r e-Fu n ction Relation sh ips. Table 2 shows
that the binding affinities of EqADH, R, and â1 enzymes
for the cyclic formamides are independent of ring size
(excluding cyclopropyl). Each of these enzymes has
exceptionally high affinities for the N-cyclobutyl, N-
cyclopentyl, and N-cyclohexyl derivatives. In contrast,
the γ2 and π enzymes show an increase in affinity as
the ring size is increased. The class I enzymes have
similar affinity for N-cyclohexylformamide, whereas the
σ and π enzymes have much lower affinity for this
inhibitor. The binding affinity of HsADH σ also shows
little dependence on ring size, but there was decreased
affinity for N-cyclohexylformamide, which implicates
Resu lts
En zym ology. The inhibition constants were deter-
mined by initial velocity studies with varied concentra-
tions of the inhibitor and a substrate appropriate for
each human enzyme. The Ki values were not dependent
on the substrate used. The formamides are uncompeti-
tive inhibitors against alcohols and competitive against
aldehydes or ketones, indicating that the formamides
* Author to whom correspondence should be addressed. Address:
4-370 Bowen Science Building. E-mail: bv-plapp@uiowa.edu.
S0022-2623(97)00738-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/09/1998

Inhibition of Alcohol Dehydrogenases by Formamides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1697
Ta ble 2. Inhibition of Alcohol Dehydrogenases by Formamides
R1–N–CHO
R2
Ki (µM)
EqEa
Hs Rb
Hs â1
Hs γ2
Hs πe
Hs σf
c
d
R1
R2
cyclopropyl
cyclobutyl
cyclopentyl
cyclohexyl
cyclohexylmethyl
benzyl
H
H
H
H
H
H
H
H
H
H
H
H
100
1.7
7.6
8.7
12
0.74
10g
2.6
16
13
3.0
5.4
1100
7200
110
8600
2600
1900
600
1400
5200
470
33
6.4
2.9
7.3
3.4
1.0
0.33
5.5
19
11
26
700
45
13
5.2
5.8
4.9
5200
450
150
84
300
54
28
380
47
11
21
210
30
180
0.74
100
960
1800
930
1100
ND
1700
4900
ND
3.8
3.9
2.3
3.6
31
7.1
110
300
1900
110
340
11
40
31
280
96
360
ND
191
120
ND
ND
150
ND
n-propyl
isopropyl
n-butyl
isobutyl
5.6h
22
120
15
100
54
7.2
5.5
3.6
7.0
1.9
1.8
0.88
0.36
7.4
4.5
5.5
4.5
14
n-heptyl
0.33
1.7
1900
4900
4600
10000
1600
2800
920
12
1-methylheptyl
cyclopropyl
cyclobutyl
cyclopentyl
cyclopentyl
cyclopentyl
cyclopentyl
cyclohexyl
cyclohexyl
cyclohexyl
cyclohexyl
cyclohexyl
0.41
2.1
19
4.6
47
1000
21
40
heptyl
cyclobutyl
cyclopropyl
cyclobutyl
cyclopentyl
propyl
methyl
ethyl
propyl
isopropyl
cyclopropyl
1500
4000
46
33
160
83
ND
1300
640
24
5.9
440
250
120
a
b
Inhibition constants determined with ethanol (0.4-2.0 mM) as the varied substrate, pH 8.0. Ethanol (2.7-15 mM), pH 8.0.
c Acetaldehyde (0.02-0.50 mM), pH 7.0. Cyclohexanone (1.5-12 mM), pH 7.0. e Acetaldehyde (5-50 mM), pH 7.0. f Ethanol (30-150
mM), pH 7.0. Rest of column with cyclohexanone (10-50 mM), pH 7.0. Rest of column with cyclohexanone (0.069-0.96 mM), pH 7.0.
ND ) not determined. The inhibition constants had errors e15%.
d
g
h
steric effects. The cyclohexylmethyl derivative was a
good inhibitor for all enzymes except for HsADH σ.
The substitution of the R1 position with a phenyl
group resulted in a very high Ki value for EqADH (2.5
mM). In contrast, N-benzylformamide binds tightly to
EqADH7 and was the most potent and selective inhibitor
for HsADH â1 (Ki ) 0.33 µM). The π enzyme, which
has good affinity for N-cyclohexylmethylformamide, was
much less sensitive to the N-benzyl derivative.
for these compounds than they do for the monosubsti-
tuted derivatives. Surprisingly, the γ2 enzyme binds the
disubstituted derivatives about as well as the mono-
substituted compounds. This result indicates again that
the space near the catalytic zinc of the γ2 enzyme is
larger than predicted from the modeling studies.
Discu ssion
The three-dimensional structure of the EqADH-
NADH-N-cyclohexylformamide complex suggests that
the amide resembles the ground-state structure for
aldehyde substrates, as the oxygen is ligated to the
catalytic zinc and the carbonyl carbon is suitably
positioned for direct hydrogen transfer.8 The inhibitor
binds with a cis conformation, which produces a cat-
ion-π interaction between the amide N-H and the
benzene ring of Phe-93 of the enzyme. We expected that
the human enzymes would have different specificities
for the various formamides, reflecting the different
constellation of amino acid residues in the active sites
and the flexibility of binding modes. This expectation
was realized, and particularly selective inhibitors were
identified. Since three-dimensional structures of the
human enzymes complexed with the new formamides
are not available, we must explain the structure-
function relationships in terms of models based on
known structures. We presume that tight binding is
associated with good van der Waals contacts, hydro-
phobic interactions (and lack of voids that would ac-
commodate water molecules), and the absence of steric
conflicts. The binding modes will vary due to flexible
side chains that adapt to the ligand structure and
optimize interactions.
The affinities of the horse, R, â1, and σ enzymes for
the noncyclic, linear formamides showed little depen-
dence on the length of the alkyl chain. This is a
somewhat surprising result, as the substrate binding
pockets are completely hydrophobic and each methylene
unit can be expected to improve affinity by about 2-fold.5
However, HsADH â1 and HsADH σ showed particularly
strong affinities for N-heptylformamide, which indicates
that a certain size of aliphatic chain can optimize
interactions. It is interesting that the HsADH γ2 prefers
the branched aliphatic chain over the corresponding
linear chain, whereas the π and σ enzymes prefer the
linear chains. The affinity of the γ2 enzyme for the
branched 1-methylheptyl derivative is particularly strong,
which must reflect the space near the catalytic zinc. The
â1 enzyme also had a high affinity for the 1-methylhep-
tyl derivative, whereas HsADH σ was poorly inhibited.
We examined a series of disubstituted formamides as
inhibitors since modeling suggested that the R enzyme
should have increased space near the catalytic zinc due
to the replacement of Phe-93 with Ala. Indeed, the
whole series of compounds with substituents at the R2
position showed high affinities for the R enzyme. In
particular, N-cyclopentyl-N-cyclobutylformamide was an
especially potent inhibitor of HsADH R. In contrast, the
horse, σ, â1, and π enzymes have much lower affinities
The disubstituted formamide N-cyclopentyl-N-cyclobu-
tylformamide is a potent and selective inhibitor for

1698 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Schindler et al.
F igu r e 1. Models of the proposed binding modes of the formamides with the human enzymes. (A) The Z isomer of N-cyclopentyl-
N-cyclobutylformamide binding to HsADH R. The model of the R enzyme was created by making the appropriate amino acid
substitutions in the structure of EqADH. (B) The cis isomer of N-benzylformamide binding to HsADH â1.9 (C) The cis isomer of
N-heptylformamide binding to HsADH σ.10
HsADH R. Molecular modeling suggests that this
inhibitor binds to the R enzyme in the Z configuration
(Figure 1A). This orientation produces favorable con-
tacts between the cyclobutyl group and Ala-93 and Ile-
318, whereas the E configuration produces unfavorable
contacts between the cyclopentyl ring and these active-
site residues. In the case of the N-methyl-N-cyclohexyl-
formamide and N-ethyl-N-cyclohexylformamide, the
methyl and ethyl groups are not capable of providing
these interactions. Binding of N-cyclopentyl-N-cyclobu-
tylformamide might be improved further by adding a
propyl or butyl group at carbon 3 of the cyclopentyl ring.
As compared to the other enzymes, the â1 enzyme
displays the highest affinity for N-benzylformamide. The
modeling of the cis conformation into the active site
provides favorable contacts between the phenyl ring and
residues Val318 and Leu309 (Figure 1B). This confor-
mation also provides a cation-π interaction between the

Inhibition of Alcohol Dehydrogenases by Formamides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1699
amide N-H and Phe-93 of the enzyme. The â1 enzyme
also shows very strong binding of N-heptylformamide.
Modeling studies show that the N-heptyl chain provides
good contacts with Val-318 and Leu-309 of the enzyme.
for substrates are steady-state kinetic constants. Thus,
the active-site topology is better defined with inhibitors
than with substrates. Although we can model the
binding, observed ligand positions may be different from
those expected from modeling because the differences
in energetics among binding modes are small. X-ray
crystallography of selected complexes would better
define the interactions.
Modeling suggests that the narrow active site in
HsADH σ could have several hydrophobic contacts with
the N-heptylformamide (Figure 1C). Comparison of the
three-dimensional structures of the â and σ enzymes
shows that the presence of Met-57, Met-141, and Phe-
309 in the σ enzyme narrows the binding pocket as
compared with the â enzyme, which has leucine residues
at these positions.10 The effect of these substitutions
for binding of N-heptylformamide is not apparent,
however, as the σ and â1 enzymes have similar binding
affinity. On the other hand, steric hindrance due to
Met-57 and Met-141 in the σ enzyme could be respon-
sible for the 10-fold or more decreases in affinity for the
branched chain derivatives as compared to the linear
alkyl chains. The 100-fold decrease in the affinity of
the σ enzyme for N-1-methylheptylformamide as com-
pared to that for N-heptylformamide is somewhat
surprising as the σ enzyme has good catalytic activity
with all-trans-retinol, and we would expect the N-1(S)-
methylheptylformamide (trans conformation) could bind
in a manner similar to all-trans-retinol.13
The inhibitors developed in this study have not been
tested in animals, but other formamides and amides
have been found to be effective in rats.6,15,16 For broad
spectrum inhibition of most alcohol dehydrogenases,
N-cyclohexylmethylformamide and N-heptylformamide
would appear to be the best inhibitors, but the N-
cyclopentyl- and N-isopropylformamides might also
have suitable pharmaceutical properties. After the
substrate specificities of the alcohol dehydrogenases are
further defined, selective inhibitors may be chosen to
prevent oxidation of specific alcohols. For treatment of
methanol poisoning, the â and γ enzymes should be
inhibited, as these are abundant in the liver and have
the highest catalytic activity.17,18 Although we have
found good inhibitors that exploit the specificities in the
active sites, more potent inhibitors might be prepared.
Toxicity and metabolism of the new inhibitors have
not been evaluated, but the LD50 for N-cyclohexylform-
amide in mice is 320 mg/kg,19 which is 25 times higher
than the dose expected to be required to produce 90%
inhibition of the class I alcohol dehydrogenases. Since
the formamides are uncompetitive inhibitors, the rela-
tive velocity of the enzymes, at saturating levels of
alcohol is given by the relationship, v ) V/(1 + [I]/Ki).
The γ2 enzyme was most potently inhibited by N-1-
methylheptylformamide, and the branched chain de-
rivatives were better inhibitors than the linear alkyl
compounds. The three-dimensional structure of this
enzyme is not known, but the amino acid residues at
the active site (Table 1) resemble those of the horse
enzyme, for which structures of several complexes are
available. The γ2 and horse enzymes are the only ones
in this group with Ser-48, rather than Thr-48, and there
should be more space to accommodate the branched
compounds. For the γ2 enzyme, however, there appears
to be even more space than in the horse enzyme, as the
disubstituted derivatives were good inhibitors, albeit not
as good as they are for the R enzyme. The γ enzyme is
the only one of these enzymes with good activity with
steroids (oxidizing 3â-hydroxyl groups), indicating a
relatively large active-site pocket.14
Exp er im en ta l Section
Ma ter ia ls. Crystalline horse liver alcohol dehydrogenase,
NAD+, and NADH were purchased from Boehringer Mann-
heim. Clones for the expression of HsADH R, â1, and σ were
obtained from Dr. Thomas D. Hurley (Indiana University
School of Medicine), and the vectors for the γ2 and π enzymes
were from Dr. J an-Olov Ho¨o¨g (Karolinska Institutet). N-
Cyclohexylformamide and other reagents were obtained from
Aldrich Chemical Co. N-Cyclohexylmethylformamide and
N-isobutylformamide were prepared as described previously.8
The primary and secondary amines were synthesized and
isolated according to the procedure of Borch et al.20 Com-
pounds were characterized using a Varian 500 MHz NMR
spectrometer. Multiple resonances are due to the cis-trans
isomerism of the formamides. Elemental analyses were
performed by Galbraith Laboratories.
P r otein P u r ifica tion . The human alcohol dehydrogenases
were prepared essentially as described.21,22 Protein homogene-
ity was established by polyacrylamide gel electrophoresis in
the presence of sodium dodecyl sulfate. Enzyme activity was
determined spectrophotometrically by following the change in
absorbance at 340 nm due to NADH production (ꢀ340 ) 6.22
mM-1 cm-1). The specific activities of the purified enzymes
in the standard assay with 2.4 mM NAD+ and 33 mM ethanol
in 0.1 M sodium glycine, pH 10, at 25 °C were comparable to
the literature values.
The three-dimensional structure of the π enzyme is
not known, but the inhibition studies provide some
indication of the active-site topology. The π enzyme has
relatively lower affinity for most of the formamides, and
no compound had a Ki of less than 1 µM. The modeling
with N-cyclohexylformamide does not indicate any steric
conflicts between the cyclohexyl ring and the enzyme
active site that would explain the lower affinity. The π
enzyme shows good binding affinity for N-cyclohexyl-
methylformamide, indicating that the methylene exten-
sion can position the cyclohexyl ring to make favorable
hydrophobic contacts with active-site residues. The
relatively high affinity of the π enzyme for the disub-
stituted formamides provides evidence that the active
site near the catalytic zinc is somewhat larger than
predicted from the modeling studies.
Kin etic Stu d ies. The inhibition constants were deter-
mined in 46 mM sodium phosphate and 0.25 mM EDTA, pH
7.0, at 25 °C by measuring the change in absorbance at 340
nm due to NADH. The kinetics of EqADH and HsADH R were
also studied in 33 mM sodium phosphate and 0.25 mM EDTA,
pH 8.0, at 25 °C, and it was found that the inhibition constants
were about the same as at pH 7. The substrates used for
assays differed with the various enzymes in order to obtain
optimal activity. The coenzyme concentration was held con-
stant at 0.2 mM NADH or 2.0 mM NAD+. The concentrations
These studies provide data for potential therapeutic
agents and for correlating structure and function of the
human alcohol dehydrogenases. The inhibition con-
stants determined in this study are thermodynamic
dissociation constants that reflect the energetics of the
multiple binding interactions. In contrast, Km values

1700 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10
Schindler et al.
of inhibitor were varied over at least a 3-fold range, and 20-
25 different assay conditions, with duplicates, were tested. The
inhibition data were fitted to the equations for competitive or
uncompetitive inhibition with the appropriate computer pro-
grams.23
Molecu la r Mod elin g. Molecular modeling was carried out
using the computer program O.24 Coordinates for the human
enzymes were obtained from the Brookhaven Protein Database
(entries: 1HDX, 1AGN). All inhibitors were built and mini-
mized in SYBYL (Tripos Associates).
NH), 1.58-1.44 (m, 2H, CH2), 1.37-1.16 (m, 8H, CH2), 0.93-
0.77 (t, 3H, CH3). Anal. (C8H17NO‚1/4H2O) C, H, N.
N-1-Meth ylh ep tylfor m a m id e: 1H NMR (CDCl3) δ 8.17 (s,
0.7H, CHO), 8.08 (d, 0.3H, CHO), 5.79-5.57 (s, broad, 0.3H,
NH), 5.52-5.32 (s, broad, 0.7NH), 4.16-3.99 (m, 0.7H, CHNH),
3.55-3.39 (m, 0.3H, CHNH), 1.57-1.38 (m, 2H), 1.36-1.09
(m, 11H), 0.97-0.81 (m, 3H). Anal. Calcd for C9H19NO‚
1/4H2O: C, 66.83; H, 12.15; N, 8.66. Found: C, 66.76; H, 11.48;
N, 8.63.
N-Cyclop r op yl-N-h ep tylfor m a m id e: 1H NMR (CDCl3) δ
8.30 (s, 0.8H, CHO), 8.09 (s, 0.2H, CHO), 3.33-3.24 (t, 1.6H,
CH2NH), 3.17-3.06 (t, 0.4H, CH2NH), 2.70-2.58 (m, 0.8H,
ring-CH), 2.55-2.45 (m, 0.2H, ring-CH), 1.66-1.51 (m, 2H,
n-heptyl-CH2), 1.39-1.19 (m, 8H, n-heptyl-CH2), 0.92-0.81(t,
3H, CH3), 0.82-0.73 (m, 2H, ring-CH2), 0.72-0.59 (m, 2H,
ring-CH2). Anal. (C11H21NO‚1/10H2O) C, H, N.
Syn th esis of F or m a m id es. The primary formamides were
prepared by refluxing the amine with 88% HCOOH, removing
excess HCOOH and H2O under reduced pressure, and distill-
ing the product. Attempts to synthesize the secondary form-
amides by the same procedure gave low yields and contami-
nating formate salts. Thus, an improved procedure was used.
The amine was added dropwise to an excess of 88% formic acid.
The reaction mixture was refluxed for 1.5 h, and excess formic
acid was removed under reduced pressure. The formate salt
was suspended in toluene and refluxed with stirring for 2 h
before removing the toluene/H2O azeotrope by distillation at
1
N,N-Dicyclobu tylfor m a m id e: H NMR (CDCl3) δ 8.3 (s,
1H, CHO), 4.5-4.4 (m, 1H, CH), 4.0-3.8 (m, 1H, CH), 2.4-
2.1 (m, 8H, CH2), 1.8-1.57 (m, 4H, CH2). Anal. (C9H15NO‚
1/6H2O) C, H, N.
N-Cyclopen tyl-N-cyclopr opylfor m am ide: 1H NMR (CD-
Cl3) δ 8.29 (s, 0.8H, CHO), 8.22 (s, 0.2H, CHO), 4.47-4.32 (m,
1H), 2.62-2.46 (m, 1H), 1.90-1.79 (m, 2H), 1.78-1.50 (m, 6H),
0.85-0.74 (m, 2H), 0.74-0.62 (m, 2H). Anal. (C4H15NO‚
1/6H2O) C, H, N.
1
88 °C. Reflux and distillation were repeated until the H NMR
of the reaction mixture indicated that the reaction had gone
to at least 60-70% completion. The excess toluene was
removed under reduced pressure. The residue was dissolved
in CHCl3, and the organic layer was washed with water, dried
with MgSO4, and filtered. The solvent was removed under
reduced pressure, and the product was applied to a silica gel
column and eluted with ethyl acetate/hexane (60:40). The
formamides were isolated as oils, with the exception of
N-benzylformamide, which was a solid. The oils were not dried
and contained fractional amounts of water. Final yields
ranged from 30 to 40%.
N-Cyclop r op ylfor m a m id e: 1H NMR (CDCl3) δ 8.29 (d,
0.4H, CHO), 8.16 (s, 0.6H, CHO), 5.94-5.74 (s, broad, 0.4H,
NH,), 6.17-5.96 (s, broad, 0.6H, NH), 2.81-2.67 (m, 0.6H, CH),
2.67-2.56 (m, 0.4H, CH), 0.89-0.73 (m, 2H, CH2), 0.70-0.46
(m, 2H, CH2). Anal. (C4H7NO‚1/4H2O) C, H, N.
N-Cyclobu tylfor m a m id e: 1H NMR (CDCl3) δ 8.09 (m, 1H,
CHO), 6.25-6.01 (s, broad, 0.3H, NH), 6.00-5.62 (s, broad,
0.7H, NH), 4.55-4.40 (m, 0.7H, CH), 4.07-3.92 (m, 0.3H, CH),
2.41-2.26 (m, 2H, CH2), 2.04-1.84 (m, 2H, CH2), 1.80-1.61
(m, 2H, CH2). Anal. Calcd for C5H9NO‚1/6H2O: C, 58.97; H,
9.21; N, 13.75; Found: C, 59.47; H, 9.51; N, 13.01.
N-Cyclop en tylfor m a m id e: 1H NMR (CDCl3) δ 8.17 (m,
1H, CHO), 6.13-5.71 (d, broad, 1H, NH), 4.34-4.18 (m, 0.8H,
CH), 3.90-3.77 (m, 0.2H, CH), 2.06-1.89 (m, 2H, CH2), 1.76-
1.52 (m, 4H, CH2), 1.52-1.31 (m, 2H, CH2). Anal. (C6H11NO‚
1/10H2O) C, H, N.
N-Ben zylfor m a m id e:7 1H NMR (CDCl3) δ 8.28 (s, 0.9H,
CHO), 8.20 (d, 0.1H, CHO), 7.40-7.22 (m, 5H), 6.05-5.79 (s,
broad, 1H, NH), 4.49-4.46 (d, 1.8H, CH2NH), 4.42-4.39 (d,
0.2H, CH2NH).
N-P r op ylfor m a m id e: 1H NMR (CDCl3) δ 8.14 (s, 0.8H,
CHO), 8.0-7.92 (d, 0.2H, CHO), 6.51-6.22 (s, broad, 1H, NH),
3.25-3.14 (m, 1.6H, CH2NH), 3.14-3.06 (m, 0.4H, CH2NH),
1.59-1.43 (m, 2H, CH3CH2CH2NH), 0.94-0.81 (t, 3H, CH3).
Anal. Calcd for C4H9NO‚0.4H2O: C, 50.93; H, 10.47; N, 14.85.
Found: C, 51.17; H, 10.39; N, 13.80.
N-Isop r op ylfor m a m id e: 1H NMR (CDCl3) δ 8.14-8.00 (m,
1H, CHO), 6.23-5.99 (s, broad, 0.2H, NH), 5.94-5.65 (s, broad,
0.8H, NH), 4.23-4.05 (m, 0.8H, CH3CH(NH)CH3), 3.76-3.59
(m, 0.2H, CH3CH(NH)CH3), 1.27-1.10 (m, 6H, CH3CH(NH)-
CH3). Anal. (C4H9NO‚1/2H2O) C, H, N.
N-Cyclopen tyl-N-cyclobu tylfor m am ide: 1H NMR (CDCl3)
δ 8.31 (s, 0.6H, CHO), 8.2 (s, 0.4H, CHO), 4.42-4.26 (m, 1H),
3.93-3.75 (m, 1H), 2.41-2.07 (m, 4H), 1.87-1.45 (m, 10H).
Anal. (C10H17NO‚1/6H2O) C, H, N.
N,N-Dicyclop en tylfor m a m id e: 1H NMR (CDCl3) δ 8.23
(s, 1H, CHO), 4.16-4.00 (m, 1H, CH), 3.76-3.63 (m, 1H, CH),
2.01-1.86 (m, 2H, CH2), 1.86-1.67 (m, 8H, CH2), 1.64-1.46
(m, 6H, CH2). Anal. (C11H19NO‚1/8H2O) C, H, N.
1
N-Cycloh exyl-N-m eth ylfor m a m id e: H NMR (CDCl3) δ
8.15 (s, 0.7H, CHO), 7.90 (s, 0.3H, CHO), 4.30-4.15 (m, 0.7H,
CH), 3.37-3.19 (m, 0.3H, CH), 2.89 (s, 0.9H, CH3), 2.80 (s,
2.1H, CH3), 1.91-1.78 (m, 2H, CH2), 1.76-1.63 (m, 4H, CH2),
1.61-1.27 (m, 4H, CH2). Anal. (C8H15NO‚1/3H2O) C, H, N.
N-Cycloh exyl-N-eth ylfor m a m id e: 1H NMR (CDCl3) δ
8.16 (s, 0.7H, CHO), 8.07 (s, 0.3H, CHO), 4.10-3.94 (m, 0.3H,
ring-CH), 3.33-3.22 (m, 2H, CH3CH2N), 3.22-3.13 (m, 0.7H,
ring-CH), 1.88-1.59 (m, 6H, ring-CH2), 1.55-1.41 (m, 2H, ring-
CH2), 1.39-1.22 (m, 2H, ring-CH2), 1.22-1.16 (t, 0.9H, CH3),
1.15-1.08 (t, 2.1H, CH3). Anal. (C19H17NO‚1/3H2O) C, H, N.
1
N-Cycloh exyl-N-p r op ylfor m a m id e: H NMR (CDCl3) δ
8.21 (s, 0.7H, CHO), 8.1 (s, 0.3H, CHO), 4.0-3.9 (m, 0.3H, ring-
CH), 3.22-3.18 (t, 2H, CH2N), 3.09-3.04 (m, 0.7H, ring-CH),
1.89-1.39 (m, 8H, ring-CH2), 1.39-1.22 (m, 2H, ring-CH2),
1.19-1.03 (m, 2H, CH3CH2CH2N), 0.96-0.81 (m, 3H, CH3).
Anal. (C10H19NO‚1/3H2O) C, H, N.
N-Cycloh exyl-N-isop r op ylfor m a m id e: 1H NMR (CDCl3)
δ 8.26 (s, 0.4H, CHO), 8.14 (s, 0.6H, CHO), 4.39-4.22 (m, 0.6H,
ring-CH), 3.91-3.76 (m, 0.4H, ring-CH), 3.66-3.55 (m, 0.4H,
CH3CH(N)CH3), 3.13-2.99 (m, 0.6H, CH3CH(N)CH3), 1.90-
1.41 (m, 10H, ring-CH2), 1.38-1.23 (d, 3H, CH3), 1.22-1.16
(d, 3H, CH3). Anal. (C10H19NO‚1/8H2O) C, H, N.
N-Cyclop en tyl-N-p r op ylfor m a m id e: 1H NMR (CDCl3) δ
8.22 (s, 0.7H, CHO), 8.12 (s, 0.3H, CHO), 4.40-4.22 (m, 0.3H,
ring-CH), 3.87-3.71 (m, 0.7H, ring-CH), 3.23-3.08 (m, 2H,
CH2), 1.96-1.84 (m, 2H, CH2), 1.82-1.70 (m, 2H, CH2), 1.68-
1.53 (m, 6H, CH2), 0.96-0.87 (m, 3H, CH3). Anal. (C9H17NO‚
1/4H2O) C, H, N.
N-Cycloh exyl-N-cyclopen tylfor m am ide: 1H NMR (CDCl3)
δ 8.29 (s, 0.5H, CHO), 8.19-8.13 (s, 0.5H, CHO), 4.29-4.11
(m, 0.5H, CH), 3.91-3.75 (m, 0.5H, CH), 3.73-3.58 (m, 0.5H,
CH), 3.11-2.93 (m, 0.5H, CH), 2.00-1.41 (m, 16H, CH2), 1.39-
1.22 (m, 2H, CH2). Anal. (C12H21NO‚1/8H2O) C, H, N.
N-Bu tylfor m a m id e: 1H NMR (CDCl3), δ 8.18 (s, 0.8H,
CHO), 7.96 (d, 0.2H, CHO), 6.21 (s, broad, 0.2H, NH), 5.61 (s,
broad, 0.8H, NH), 3.30-3.23 (m, 1.6H, CH2NH), 3.21-3.10 (q,
0.4H, CH2NH), 1.6-1.2 (m, 4H, CH2), 0.80-0.99 (t, 3H, CH3).
Anal. Calcd for C5H11NO‚1/2H2O: C, 54.52; H, 10.98; N, 12.72.
Found: C, 52.73; H, 9.75; N, 11.94.
N-Cycloh exyl-N-cyclopr opylfor m am ide: 1H NMR (CDCl3)
δ 8.29 (s, 0.9H, CHO), 8.22 (s, 0.1H, CHO), 4.16-4.02 (m, 0.9H,
CH), 3.26-3.07 (m, 0.1, CH), 2.56-2.48 (m, 0.9H, CH), 2.48-
2.43 (m, 0.1H, CH), 1.93-1.77 (m, 2H, CH2), 1.76-1.59 (m,
6H, CH2), 1.46-1.28 (m, 2H, CH2), 0.86-0.76 (m, 2H, CH2),
0.75-0.63 (m, 2H, CH2). Anal. (C10H17NO‚1/8H2O) C, H, N.
1
N-Hep tylfor m a m id e: H NMR (CDCl3) δ 8.22 (s, 0.75H,
CHO), 8.06-7.90 (d, 0.25H, CHO), 6.12-5.82 (d, broad, 1H,
NH), 3.31-3.21 (m, 1.5H, CH2NH), 3.2-3.13 (m, 0.5H, CH2-

Inhibition of Alcohol Dehydrogenases by Formamides
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 10 1701
(12) Satre, M. A.; Zˆgombic´-Knight, M.; Duester, G. The Complete
Structure of Human Class IV Alcohol Dehydrogenase (Retinol
Dehydrogenase) Determined from the ADH7 Gene. J . Biol.
Chem. 1994, 269, 15606-15612.
(13) Kedishvili, N. Y.; Bosron, W. F.; Stone, C. L.; Hurley, T. D.;
Peggs, C. F.; Thomasson, H. R.; Popov, K. M.; Carr, L. G.;
Edenberg, H. J .; Li, T.-K. Expression and Kinetic Characteriza-
tion of Recombinant Human Stomach Alcohol Dehydrogenase.
Active-site Amino Acid Sequence Explains Substrate Specificity
Compared with Liver Enzymes. J . Biol. Chem. 1995, 270, 3625-
3630.
(14) McEvily, A. J .; Holmquist, B.; Auld, D. S.; Vallee, B. L.
3â-Hydroxy-5â-steroid Dehydrogenase Activity of Human Liver
Alcohol Dehydrogenase Is Specific to γ Subunits. Biochemistry
1988, 27, 4284-4288.
(15) Plapp, B. V.; Leidal, K. G.; Smith, R. K.; Murch, B. P. Kinetics
of Inhibition of Ethanol Metabolism in Rats and the Rate-
Limiting Role of Alcohol Dehydrogenase. Arch. Biochem. Bio-
phys. 1984, 230, 30-38.
Ack n ow led gm en t. This work was supported by
United States Public Health Service Grant AA00279.
We thank Dr. Thomas D. Hurley at the Indiana Uni-
versity Medical School, Indianapolis, and Dr. J an-Olov
Ho¨o¨g, Karolinska Institutet, Stockholm, for the expres-
sion vectors for the human alcohol dehydrogenases; Dr.
Heeyeong Cho in our laboratory for advice on the
modeling of the inhibitors; and The University of Iowa
College of Medicine NMR Facility.
Refer en ces
(1) Abbreviations: HsADH, human (Homo sapiens) alcohol dehy-
drogenase; EqADH, horse (Equus caballus) alcohol dehydroge-
nase.
(2) J acobsen, D.; McMartin, K. E. Antidotes for Methanol and
Ethylene Glycol Poisoning. Clin. Toxicol. 1997, 35, 127-143.
(3) Edenberg, H. J .; Bosron, W. F. Alcohol Dehydrogenases. In
Comprehensive Toxicology, Vol. 3 Biotransformation; Guenge-
rich, F. P., Ed.; Pergamon Press: New York, 1997; pp 119-131.
(4) Chadha, V. K.; Leidal, K. G.; Plapp, B. V. Inhibition by
Carboxamides and Sulfoxides of Liver Alcohol Dehydrogenase
and Ethanol Metabolism. J . Med. Chem. 1983, 26, 916-922.
(5) Winer, A. D.; Theorell, H. Dissociation Constants of Ternary
Complexes of Fatty Acids and Fatty Acid Amides with Horse
Liver Alcohol Dehydrogenase-Coenzyme Complexes. Acta Chem.
Scand. 1960, 14, 1729-1742.
(16) Delmas, C.; de Saint Blanquat, G.; Freudenrich, C.; Biellmann,
J .-F. New Inhibitors of Alcohol Dehydrogenase: Studies in Vivo
and in Vitro in the Rat. Alcoholism Clin. Exp. Res. 1983 7, 264-
270.
(17) Wagner, F. W.; Burger, A. R.; Vallee, B. L. Kinetic Properties of
Human Liver Alcohol Dehydrogenase: Oxidation of Alcohols by
Class I Isoenzymes. Biochemistry 1983, 22, 1857-1863.
(18) Cho, H.; Plapp, B. V. Specificity of Alcohol Dehydrogenases for
Sulfoxides. Biochemistry 1998, 37, 4482-4489.
(19) Registry of Toxic Effects of Chemical Substances, NIOSH,
registry number 766-93-8, National Library of Medicine, Be-
thesda, MD, 1997.
(6) Porter, C. C.; Titus, D. C.; DeFelice, M. J . Liver Alcohol
Dehydrogenase Inhibition by Fatty Acid Amides, N-Alkylform-
amides and Monoalkylureas. Life Sci. 1976, 18, 953-959.
(7) Freudenreich, C.; Samama, J .-P.; Biellmann, J .-F. Design of
Inhibitors from the Three-Dimensional Structure of Alcohol
Dehydrogenase. Chemical Synthesis and Enzymatic Properties.
J . Am. Chem. Soc. 1984, 106, 3344-3353.
(20) Borch R. F.; Bernstein M. D.; Durst H. D. The Cyanohydridobo-
rate Anion as a Selective Reducing Agent. J . Am. Chem. Soc.
1971, 93, 2897-2904.
(21) Stone, C. L.; Bosron, W. F.; Dunn, M. F. Amino Acid Substitu-
tions at Position 47 of Human â1â1 and â2â2 Alcohol Dehydro-
genase Affect Hydride Transfer and Coenzyme Dissociation Rate
Constants. J . Biol. Chem. 1993, 268, 892-899.
(22) Hurley, T. D.; Steinmetz, C. G.; Xie, P.; Yang, Z.-N. Three-
Dimensional Structures of Human Alcohol Dehydrogenase En-
zymes Reveal the Molecular Basis for Their Functional Diver-
sity. In Enzymology and Molecular Biology of Carbonyl
Metabolism 6; Weiner, H., Lindahl, R., Crabb, D. W., Flynn, T.
G., Eds.; Plenum Press: New York, 1997; pp 291-302.
(23) Cleland, W. W. Statistical Analysis of Enzyme Kinetic Data.
Methods Enzymol. 1979, 63, 103-138.
(24) J ones, T. A.; Zou, J . Y.; Cowan, S. W.; Kjeldgaard, M. Improved
Methods for Binding Protein Models in Electron Density Maps
and the Location of Errors in These Models. Acta Crystallogr.
1991, A47, 110-119.
(8) Ramaswamy, S.; Scholze, M.; Plapp, B. V. Binding of Forma-
mides to Liver Alcohol Dehydrogenase. Biochemistry 1997, 36,
3522-3527.
(9) Hurley, T. D.; Bosron, W. F.; Stone, C. L.; Amzel L. A. Structures
of Three Human â Alcohol Dehydrogenase Variants. Correlations
with Their Functional Differences. J . Mol. Biol. 1994, 239, 415-
429.
(10) Xie, P.; Parsons, S. H.; Speckhard, D. C.; Bosron, W. F.; Hurley,
T. D. X-ray Structure for Human Class IV σσ Alcohol Dehydro-
genase. Structural Basis for Substrate Specificity. J . Biol. Chem.
1997, 272, 18558-18563.
(11) Sun, H.-W.; Plapp, B. V. Progressive Sequence Alignment and
Molecular Evolution of the Zn-Containing Alcohol Dehydroge-
nase Family. J . Mol. Evol. 1992, 34, 522-535.
J M9707380