Inhibition and substrate activity of conformationally rigid vigabatrin analogues with γ-aminobutyric acid aminotransferase

Article abstract of DOI:10.1021/jm990271o

Several cyclopentene GABA analogues were synthesized as conformationally rigid analogues of the epilepsy drug vigabatrin and tested as inhibitors and substrates of γ-aminobutyric acid aminotransferase (GABA-AT). None of these compounds produced time-dependent inhibition. (1R,4S)-(+)-4-Amino-2- cyclopentene-1-carboxylic acid ((+)-3), (4R)-(-)-4-amino-1-cyclopentene-1- carboxylic acid ((-)-4), and d,l-3-amino-1-cyclopentene-1-carboxylic acid (6) are good substrates. The K(m) and k(cat) values for the latter two compounds are very similar to those of GABA, suggesting that they bind in an orientation similar to that of GABA. The K(m) value for (+)-3 is 24 times lower than that for GABA, although its k(cat) value is only one-fourth that for GABA; nonetheless, it is a better substrate for GABA-AT than is GABA. All of these compounds, as well as the enantiomers of 3 and 4 and d,l-trans-4- amino-2-cyclopentene-1-carboxylic acid (5), are competitive inhibitors of GABA-AT. These results demonstrate the effects of the carboxylate group orientation and the stereochemistry of the amino and carboxylate groups on the substrate activity and inhibitor activity, and this should be important to the future design of inhibitors of GABA-AT.

Full text of DOI:10.1021/jm990271o

J . Med. Chem. 1999, 42, 4725-4728
4725
In h ibition a n d Su bstr a te Activity of Con for m a tion a lly Rigid Viga ba tr in
An a logu es w ith γ-Am in obu tyr ic Acid Am in otr a n sfer a se
J ian Qiu, J oyce M. Pingsterhaus, and Richard B. Silverman*
Departments of Chemistry and of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University,
Evanston, Illinois 60208-3113
Received J une 1, 1999
Several cyclopentene GABA analogues were synthesized as conformationally rigid analogues
of the epilepsy drug vigabatrin and tested as inhibitors and substrates of γ-aminobutyric acid
aminotransferase (GABA-AT). None of these compounds produced time-dependent inhibition.
(1R,4S)-(+)-4-Amino-2-cyclopentene-1-carboxylic acid ((+)-3), (4R)-(-)-4-amino-1-cyclopentene-
1-carboxylic acid ((-)-4), and d,l-3-amino-1-cyclopentene-1-carboxylic acid (6) are good sub-
strates. The Km and kcat values for the latter two compounds are very similar to those of GABA,
suggesting that they bind in an orientation similar to that of GABA. The Km value for (+)-3 is
24 times lower than that for GABA, although its kcat value is only one-fourth that for GABA;
nonetheless, it is a better substrate for GABA-AT than is GABA. All of these compounds, as
well as the enantiomers of 3 and 4 and d,l-trans-4-amino-2-cyclopentene-1-carboxylic acid (5),
are competitive inhibitors of GABA-AT. These results demonstrate the effects of the carboxylate
group orientation and the stereochemistry of the amino and carboxylate groups on the substrate
activity and inhibitor activity, and this should be important to the future design of inhibitors
of GABA-AT.
In tr od u ction
electrons in the γ-C-H bond of vigabatrin allow for
delocalization of those electrons in either direction. As
a possible approach to prohibiting one of the possible
tautomerizations and enhancing the other, a series of
conformationally rigid vigabatrin and related analogues
(3-6), which were previously shown to be excellent
mimetics of GABA during studies of the GABAA receptor
and GABA uptake,7 were synthesized.8,9 Furthermore,
the binding affinities of the different geometric and
enantiomeric compounds may reveal information re-
garding the binding of the various groups at the active
site and provide a basis for further inhibitor design.
Surprisingly, unlike vigabatrin, none of these analogues
exhibit time-dependent inhibition of GABA-AT; how-
ever, all of the analogues are competitive inhibitors, and
analogues (+)-3, (-)-4, and 6 are very good substrates
for GABA-AT.
Selective blockage of the catabolytic pathway of the
inhibitory neurotransmitter, γ-aminobutyric acid (GABA),
has proven to be an effective means to terminate
epileptic seizures resulting from a decrease in the brain
concentration of GABA.1 The enzyme that catalyzes the
degradation of GABA, GABA aminotransferase (GABA-
AT, EC 2.6.1.19), is a pyridoxal 5′-phosphate (PLP)-
dependent enzyme.2 Compounds that inactivate this
enzyme exhibit anticonvulsant activity.1,3 One of the
most effective inactivators of GABA-AT is the antiepi-
leptic drug vigabatrin ((4S)-4-amino-5-hexenoic acid (1);
γ-vinyl-GABA),4 which was designed as a mechanism-
based inactivator of this enzyme (Scheme 1).5 Vigabatrin
was shown to inactivate GABA-AT by at least two
principal mechanisms: a Michael addition mechanism
(pathway a) and an enamine mechanism (pathway b).6
These two pathways occur via deprotonation of the
γ-carbon followed by either tautomerization through the
PLP ring (2, pathway a, leading to Michael addition) or
tautomerization through the vinyl double bond (path-
way b, leading to enamine formation). In the absence
of the crystal structure of GABA-AT, the position of the
double bond of enzyme-bound 1 is unknown. (The
crystal structure of GABA-AT was recently reported;
see: Storici, P.; Capitani, G.; De Biase, D.; Moser, M.;
J ohn, R. A.; J ansonius, J . N.; Schirmer, T. Crystal
Structure of GABA-Aminotransferase, a Target for
Antiepileptic Drug Therapy. Biochemistry 1999, 38,
8628-8634.) Presumably, the reason that two different
inactivation mechanisms arise is because the orienta-
tions of the pyridoxal ring system relative to the
Ch em istr y
The enantiomers of 38 were synthesized by acid
hydrolysis of the corresponding commercially available
enantiopure lactams 7 (Scheme 2). The enantiomers of
* Address correspondence to this author at the Department of
Chemistry. Phone: (847) 491-5653. Fax: (847) 491-7713. E-mail:
Agman@chem.nwu.edu.
10.1021/jm990271o CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/01/1999

4726 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22
Brief Articles
Sch em e 1
Sch em e 2
Ta ble 1. Kinetic Constants for the Cyclopentene Analogues
48 were synthesized by Boc-protection of each enanti-
omer of 3, esterification, base treatment, and deprotec-
tion (Scheme 2). Following published procedures, ana-
logues 58 and 69 were synthesized in racemic forms.
Km
kcat
kcat/Km
Ki
compd
GABA
(mM)
(min-1
)
(mM-1 min-1
)
(mM)
2.4 ( 0.07
49
11.7
20.4
117
(1R,4S)-(+)-3 0.1 ( 0.01
(1S,4R-)-(-)-3 no turnover
0.052 ( 0.004
2.7 ( 0.6
1.2 ( 0. 2
72 ( 3
(4R)-(-)-4
(4S)-(+)-4
d,l-5
1.6 ( 0.1
37.5
40
23.4
17.4
Resu lts a n d Discu ssion
no turnover
no turnover
2.3 ( 0.4
38 ( 5
Surprisingly, incubation of GABA-AT with compounds
3-6 resulted in no time-dependent loss of enzyme
activity. Given the potency of vigabatrin as a time-
dependent inactivator6 and the similarity of the struc-
tures, particularly of (+)-3 and 5, to vigabatrin, these
compounds are expected to be inactivators. Incubation
of (+)-3 (2 mM) with GABA-AT resulted in an immedi-
ate loss of enzyme activity (30%), presumably the result
of competitive reversible inhibition. However, the en-
zyme activity gradually recovered over 40 min, presum-
ably the result of transamination of the compound.
Although (+)-3 is not an inactivator, it is an excellent
substrate (Table 1), having a Km value 24 times lower
than that of GABA; the specificity constant (kcat/Km) for
this compound as a substrate (117 min-1 mM-1) is
almost 6 times that of GABA (20.4 min-1 mM-1).
Therefore, it is effectively being converted into an
intermediate that should resemble that derived from
d,l-6
0.6 ( 0.1
vigabatrin (see 2, Scheme 1), but it is not active as an
inactivator. Apparently, constraining the double bond
of vigabatrin into the five-membered ring of these
cyclopentene analogues misorients it so it is not properly
juxtaposed for nucleophilic attack by the enzyme, nor
is it able to act as an effective enamine for addition to
the enzyme-bound PLP (such as in pathway b, Scheme
1).
In addition to (+)-3, (-)-4 and d,l-6 also are good
substrates for GABA-AT, having kcat/Km values compa-
rable to that of GABA. Note that (1R,4S)-(+)-3 and (4R)-
(-)-4 have the same stereochemistry at C-4; the rules
for nomenclature change the notation. Therefore, both
of these compounds have the γ-proton in the same
stereochemical position as the proton that is removed

Brief Articles
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22 4727
from GABA during turnover (the 4-pro-S proton).10 The
corresponding enantiomers of 3 and 4 are not sub-
strates. The Km values for (-)-4 and d,l-6 are 16 and
23 times, respectively, higher than that for (+)-3,
although they are comparable to that of GABA. This
suggests that binding of GABA to the enzyme is similar
to that for (-)-4 and d,l-6 (presumably, it is (4R)-6 that
is the substrate), in which the carboxylate is planar to
the cyclopentene ring. The carboxylate group in (+)-3,
which is not planar to the ring, must be in a more
favorable position than that in bound GABA, (-)-4, or
d,l-6. Although the binding energies are lower, the rate
constants for (-)-4 and d,l-6 are about 4 times higher
than that for (+)-3. Since the rate-determining step is
γ-proton removal,11 removal of this proton in (+)-3 must
be a lower energy process than in the case of GABA,
(-)-4, or d,l-6.
It is intriguing that d,l-5 is neither an inactivator nor
a substrate, even though one of the enantiomers (1S,4S)
has the correct stereochemistry at C-4 for deprotonation.
Apparently, the carboxylate group, which is on the
opposite face of the ring as in (+)-3, interacts with the
enzyme in such a way as to misorient the C-4 proton,
preventing its removal by the active site base. The other
enantiomer has the wrong stereochemistry at C-4.
Nonetheless, 5 is a weak competitive inhibitor of the
enzyme, and in fact, all of the compounds tested are
competitive reversible inhibitors (Table 1). Compound
(+)-3 is the most potent with a Ki value of 52 µM.
Compounds (+)-4 and d,l-5 are very poor inhibitors.
In conclusion, these results demonstrate the impor-
tance of the orientation of the carboxylate group, as well
as the stereochemistry of the amino group, on the
binding of the compound to GABA-AT and suggest
future directions for the design of inhibitors.
was diluted with ethyl acetate (100 mL). The combined organic
solution was washed with 5% HCl solution (15 mL), saturated
NaHCO3 solution (15 mL), and brine (15 mL) and then dried
over anhydrous MgSO4. After removal of solvents under
reduced pressure, the residue was purified by silica gel
chromatography to afford a colorless oil (0.86 g, 98%): Rf )
1
0.26 (EtOAc:hexane, 1:4); [R]23.5 ) -51.4° (c 5.95, CHCl3); H
NMR (CDCl3) δ 5.89 (2H, m, H2, H3), 4.79 (1H, m, H4), 3.71
(3H, s, OCH3), 3.48 (1H, m, H1), 2.51 (1H, dt, J 14, 8.5 Hz,
H
5a), 1.84 (1H, dt, J 14, 4.1 Hz, H5b), 1.45 (9H, s, -O-t-Bu);
EI-MS 175, 159, 145, 129, 101, 81; HRMS calcd for C6H3F2-
NO3 175.0081, found 175.0081.
Meth yl (4S)-4-[(ter t-Bu tyloxycar bon yl)am in o]cyclopen t-
1-en e-1-ca r boxyla te, (+)-9. Compound (1S,4R)-(-)-8 (0.88
g, 3.6 mol) was dissolved in a solution of anhydrous THF (8
mL) and DBU (0.65 g, 4.2 mmol). The resultant solution was
heated at reflux for 10 h. After cooling, the resultant solution
was concentrated under reduced pressure. The residue was
purified by silica gel chromatography to afford a white solid
(0.8 g, 95%): Rf ) 0.26 (EtOAc:hexane, 1:4); [R]23.5 ) +35.8°
(c 4.63, CHCl3); 1H NMR (CDCl3) δ 6.70 (1H, t, J 2.2 Hz, H2),
4.34 (1H, s, broad, H4), 3.71 (3H, s, -OCH3), 2.90 (2H, m, H5a
and H5b), 2.40 (2H, m, H3a and H3b), 1.45 (9H, s, -O-t-Bu);
EI-MS 241, 185, 168, 153, 141, 124, 57; HRMS calcd for C12H19
NO4 241.1314, found 241.1315.
-
(4S)-4-Am in o-1-cyclop en ten e-1-ca r boxylic Acid , (+)-4.
Compound (4S)-(+)-9 (0.7 g, 2.9 mmol) was added to a
solution of acetic acid (6 mL) and 1 M HCl (6 mL). The
resultant solution was stirred and heated to 100 °C for 1 h.
After removal of solvents under vacuum, the residue was
purified with an ion-exchange column (AG 50w-8x, H+ form),
eluting with water and then 1 M pyridine solution. After
evaporation of solvents, the resultant solid was further purified
by recrystallization from ethanol and ether to afford a white
solid (0.3 g, 77%): [R]23.5 ) +32° (c 2.14, H2O);8a 1H NMR
(CDCl3) δ 6.32 (1H, m, H2), 3.98 (1H, m, H4), 2.95 (2H, m, H5a
and H5b), 2.55 (2H, m, H3a and H3b); EI-MS 127, 109, 82, 80,
56; HRMS calcd for C6H9NO2 127.0633, found 127.0635.
Met h yl (1R,4S)-4-[(ter t-Bu t yloxyca r b on yl)a m in o]cy-
clop en t-2-en e-1-ca r boxyla te, (+)-8. This compound was
prepared as described above from (1R,4S)-4-amino-2-cyclopen-
tene-1-carboxylic acid, (+)-3, [R] ) +241° (c 5.14, H2O; >99%
ee).8b All NMR data and MS data are identical to those of the
enantiomer; the optical rotation was [R]23.5 ) +51.9° (c 4.65,
CHCl3).
Met h yl (4R)-4-[(ter t-Bu t yloxyca r b on yl)a m in o]cyclo-
p en t-1-en e-1-ca r boxyla te, (-)-9. This compound was pre-
pared as described above from (1R,4S)-(+)-8. All NMR data
and MS data are identical to those of the enantiomer; the
optical rotation was [R]23.5 ) -36.1° (c 4.41, CHCl3).
(4R)-4-Am in o-1-cyclop en ten e-1-ca r boxylic Acid , (-)-4.
This compound was prepared as described above from (4R)-
(-)-9. All NMR data and MS data are identical to those of the
enantiomer; the optical rotation was [R]23.5 ) -31.9° (c 2.41,
H2O).8a
Exp er im en ta l Section
Gen er a l Meth od s. Optical spectra and GABA-AT assays
were recorded on a Perkin-Elmer Lambda 10 UV/vis spectro-
photometer. 1H NMR spectra were recorded on a Varian
Gemini 300-MHz NMR spectrometer. Chemical shifts are
reported as δ values in parts per million downfield from Me4-
Si (δ 0.0) as the internal standard in CDCl3. For samples run
in D2O, the HOD resonance was arbitrarily set at 4.80 ppm.
IR spectra were taken with a Bio-Rad FTS60 spectrophotom-
eter. An Orion Research model 701 pH meter with a general
combination electrode was used for pH measurements. Mass
spectra were obtained on a VG Instrument VG70-250SE high-
resolution spectrometer with a Maspec Data System. Flash
column chromatography was carried out with Merck silica gel
60 (230-400 mesh ASTM). TLC was run with EM Science
silica gel 60 F254 precoated glass plates.
R ea gen t s. All reagents were purchased from Aldrich
Chemical Co. without further purification except anhydrous
ether and tetrahydrofuran, which were distilled over sodium
metal under nitrogen, and anhydrous dichloromethane, which
was distilled over calcium hydride.
Met h yl (1S,4R)-4-[(ter t-Bu t yloxyca r b on yl)a m in o]cy-
clop en t-2-en e-1-ca r boxyla te, (-)-8. Crude (1S,4R)-4-amino-
2-cyclopentene-1-carboxylic acid, (-)-3, [R]23.5 ) -236° (c 1.7,
H2O; 98% ee)8b (0.50 g, 3.0 mmol), obtained from acid hydroly-
sis of the corresponding lactam (1S,4R)-(+)-7 (Aldrich Chemi-
cal Co.), was added to a solution of methanol (10 mL)
containing triethylamine (1.6 mL, 12 mmol) and di-tert-butyl
dicarbonate (1.46 g, 6.6 mmol) at room temperature. After
being stirred for 3 h the reaction mixture was concentrated
under vacuum. The resultant mixture was dissolved in dichlo-
romethane (10 mL), and excess ethereal diazomethane solution
(10 mL) was added. After being stirred for 1 h, the solution
En zym es a n d Assa ys. Pig brain GABA-AT12 (specific
activity 3.9 units/mg), GABAase, and succinic semialdehyde
dehydrogenase were obtained and assayed as previously
described.13
Deter m in a tion of Kin etic Con sta n ts for Su bstr a tes.
GABA analogues of varying concentrations (e.g. 0.11, 0.21,
0.32, 0.54. 1.07, and 1.60 mM for (+)-3) were incubated at 25
°C with GABA-AT (0.004 unit) in 50 mM potassium pyrophos-
phate buffer, pH 8.5, containing 5 mM â-mercaptoethanol and
5 mM [5-14C]2-ketoglutarate (0.1 mCi/mmol) for 60 min. The
resulting [14C]glutamate was isolated and quantified as de-
scribed previously.12 The ranges of concentrations for (-)-4 and
d,l-6 are 0.65-6.5 and 0.8-8.5 mM, respectively. Controls
consisted of the entire incubation mixture with either enzyme
or substrate omitted. Nonlinear regression analysis of the
Michaelis-Menten data was carried out for the calculation of
the kinetic constants according to Cleland.14
In h ibition of GABA-AT by GABA An a logu es. The activ-
ity of GABA-AT (0.004 unit) upon the introduction of varying

4728 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 22
Brief Articles
(5) Lippert, B.; Metcalf, B. W.; J ung, M. J .; Casara, P. 4-Amino-
hex-5-enoic Acid, a Selective Catalytic Inhibitor of 4-Amino-
butyric-Acid Aminotransferase in Mammalian Brain. Eur. J .
Biochem. 1977, 74, 441-445.
(6) Nanavati, S. M.; Silverman, R. B. Mechanism of Inactivation of
γ-Aminobutyric Acid Aminotransferase by the Antieplieptic Drug
γ-Vinyl GABA (Vigabatrin). J . Am. Chem. Soc. 1991, 113, 9341-
9349.
(7) Allan, R. D.; Dickenson, H. W.; Fong, J . Structure-Activity
Studies on the Activity of a Series of Cyclopentane GABA
Analogues on GABAA Receptors and GABA Uptake. Eur. J .
Pharmacol. 1986, 122, 339-348.
(8) (a) Allan, R. D.; Fong, J . Synthesis of Analogues of GABA. XV
Preparation and Resolution of Some Potent Cyclopentene and
Cyclopentane Derivatives. Aust. J . Chem. 1986, 39, 855-864.
(b) Csuk, R.; Dorr, P. Biocatalytical Transformations. IV. Enan-
tioselective Enzymatic Hydrolyses of Building Blocks for the
Synthesis of Carbocyclic Nucleosides. Tetrahedron: Asymmetry
1994, 5, 269-276.
(9) Allan, R. D.; Twitchin, B. Synthesis of Analogues of GABA. IV
Three Unsaturated Derivatives of 3-Aminocyclopentane-1-Car-
boxylic Acid. Aust J . Chem. 1980, 33, 599-604.
(10) Burnett, G.; Walsh, C. Sterospecificity of Enzymatic Transami-
nation of γ-Aminobutyrate. J . Chem. Soc., Chem. Commun.
1979, 826-828.
(11) Yu, P. H.; Durden, D. A.; Davis, B. A.; Boulton, A. A. Deuterium
Isotope Effect in gamma-Aminobutyric Acid Transamination:
Determination of Rate-limiting Step. J . Neurochem. 1987, 48,
440-446.
(12) (a) Churchich, J . E.; Moses, U. J . Biol. Chem. 1981, 256, 1101.
(b) Koo, Y. K.; Nandi, D.; Silverman, R. B. Multiple Active
Enzyme Species of γ-Aminobutyric Acid Aminotransferase are
not Isozymes. Submitted.
concentrations of GABA analogues (e.g. 0.022, 0.045, 0.062,
0.090, and 0.122 mM for (+)-3) was determined spectropho-
tometrically, by measuring the formation of NADH upon the
conversion of succinic semialdehyde dehydrogenase at 25 °C
and at several GABA concentrations (e.g. 1.77, 2.47, 3.51, 5.20,
and 8.45 mM) in 100 mM potassium pyrophosphate buffer,
pH 8.5, containing 5 mM â-mercaptoethanol.15 Kinetic param-
eters were derived from Dixon16 and Cornish-Bowden17 plots
with at least five different concentrations of GABA analogues
for every concentration of GABA. The ranges of concentrations
for those analgues are as follows: 0-10 mM for (-)-3, 0-4.9
mM for (-)-4, 0-18 mM for (+)-4, 0-24 mM for d,l-5, and 0.0-
6.5 mM for d,l-6.
Tim e-Dep en d en t In a ctiva tion of GABA-AT by An a -
logu es (+)-3, (-)-3, a n d d ,l-5. Incubation mixtures were set
up consisting of GABA-AT (0.02 unit), 5 mM R-ketoglutarate,
5 mM â-mercaptoethanol, and a range of GABA analogue
concentrations from 0-10 mM in 100 mM potassium pyro-
phosphate buffer, pH 8.5, at 25 °C. Periodically, 20-µL aliquots
were withdrawn and assayed spectrophotometrically for re-
maining enzyme activity in a total volume of 600 µL.
Ack n ow led gm en t. The authors are grateful to the
National Institutes of Health (Grant NS15703) for
financial support of this research. We thank Dr. Yumee
Kim Koo for assistance in the purification of the enzyme.
Refer en ces
(1) Nanvati, S. M.; Silverman, R. B. Design of Potential Anticon-
vulsant Agents: Mechanistic Classification of GABA Ami-
notransferase Inactivators. J . Med. Chem. 1989, 32, 2413-2421.
(b) J ung, M. J . Neurotransmitter Interaction and Compartmen-
tation; Plenum: New York, 1982; pp 329-344.
(2) Cooper, A. J . Glutamate-gamma-aminobutyrate transaminase.
Methods Enzymol. 1985, 113, 80-82 and references therein.
(3) (a) Metcalf, B. W.; Lippert, B.; Casara, P. In Enzyme-Activated
Irreversible Inhibitors; Seier, N., J ung, M. J ., Eds.; Elsevier/
North-Holland Press: Amsterdam, 1978; pp 123-133. (b)
Schechter, P. J .; Tranier, Y. In Enzyme-Activated Irreversible
Inhibitors; Seier, N., J ung, M. J ., Koch-Weser, J ., Eds.; Elsevier/
North-Holland Press: Amsterdam, 1978; pp 149-162.
(4) Sabers, A.; Gram, L. Pharmacology of Vigabatrin. Pharmacol.
Toxicol. 1992, 70, 237-243.
(13) Silverman, R. B.; Levy, M. A. 3-Substituted 4-Aminobutanoic
Acids: Substrates for 4-Aminobutyric Acids-R-Ketoglutaric Acid
Aminotransferase. J . Biol. Chem. 1981, 256, 11565-11568.
(14) Cleland, W. W. Methods Enzymol. 1979, 63, 103-138.
(15) Silverman, R. B.; Nanavati, S. M. Selective Inhibition of γ-Ami-
nobutyric acid Aminotransferase by (3R,4R), (3S,4S) and (3S,4R)-
4-Amino-5-fluoro-3-phenylpentanoic Acids. J . Med. Chem. 1990,
33, 931-936.
(16) Dixon, M. The Determination of Enzyme Inhibitor Constants.
Biochem. J . 1953, 55, 170-171.
(17) Cornish-Bowden, A. Fundamentals of Enzyme Kinetics; Portland
Press: London, 1995; pp 102-108.
J M990271O