γ-Amino butyric acid analogs as novel potent GABA-AT inhibitors: Molecular docking, synthesis, and biological evaluation

Article abstract of DOI:10.1007/s00044-012-0023-0

A new series, of γ-amino butyric acid analogs were designed and synthesized as novel potent GABA-AT inhibitors. A structure-activity relationship study was performed by correlating the effect of different substituents with GABA-AT inhibitory activity of the title compounds. The preliminary bioassays showed that acid hydrazones exhibited excellent inhibitory activities in micromolar (0.07-0.56 μM) range, while Schiff's bases showed variable results. The most potent compound, 4-amino-N'-[(1Z)-1-(2-bromophenyl) ethylidene]butanehydrazide (AHG177) showed inhibitory potency (IC₅₀) of 0.073 μM. Aminobutyrate transaminase is a pyridoxal-P enzyme which follows a bi-bi ping pong mechanism and in pyridoxamine form can readily transaminate only with succinic semialdehyde and 2-oxoglutarate. The results strongly suggest that only the pyridoxal form of the enzyme is capable of reacting with the ligands. Our findings open up the possibility to extend this protocol to different databases in order to find new potential inhibitor for promising targets based on a rational drug design process.

Full text of DOI:10.1007/s00044-012-0023-0

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res (2013) 22:134–146
DOI 10.1007/s00044-012-0023-0
ORIGINAL RESEARCH
c-Amino butyric acid analogs as novel potent GABA-AT
inhibitors: molecular docking, synthesis, and biological evaluation
•
S. K. Bansal B. N. Sinha R. L. Khosa
•
Received: 8 December 2011 / Accepted: 7 March 2012 / Published online: 20 March 2012
Ó Springer Science+Business Media, LLC 2012
Abstract A new series, of c-amino butyric acid analogs
were designed and synthesized as novel potent GABA-AT
inhibitors. A structure–activity relationship study was
performed by correlating the effect of different substituents
with GABA-AT inhibitory activity of the title compounds.
The preliminary bioassays showed that acid hydrazones
exhibited excellent inhibitory activities in micromolar
(0.07–0.56 lM) range, while Schiff’s bases showed vari-
able results. The most potent compound, 4-amino-N0-[(1Z)-
1-(2-bromophenyl)ethylidene]butanehydrazide (AHG177)
showed inhibitory potency (IC50) of 0.073 lM. Aminobu-
tyrate transaminase is a pyridoxal-P enzyme which follows
a bi–bi ping pong mechanism and in pyridoxamine form
can readily transaminate only with succinic semialdehyde
and 2-oxoglutarate. The results strongly suggest that only
the pyridoxal form of the enzyme is capable of reacting
with the ligands. Our findings open up the possibility to
extend this protocol to different databases in order to find
new potential inhibitor for promising targets based on a
rational drug design process.
Keywords AutoDock 4.0.1 ꢀ Docking simulation ꢀ
GABA ꢀ GABA-AT inhibitor ꢀ IC50
ꢀ
Lamarckian genetic algorithm
Introduction
The search for antiepileptic compounds with a more
selective activity and lower toxicity continues to be an area
of investigation in medicinal chemistry. A rational drug
design process of a new anticonvulsant could be achieved
by the identification of new targets through better under-
standing of molecular mechanisms of epilepsy. Novel
anticonvulsant agents are discovered through conventional
screening and/or structure modification rather than a
mechanism-driven design (Barbara, 2005). Docking-based
drug design by the use of structural biology remains one of
the most logical and aesthetically pleasing approaches in
drug discovery paradigms. The structured knowledge of the
binding capabilities of the active site residues to specific
groups on the agonist or antagonist leads to proposals for
synthesis of very specific agents, with a high probability of
biological action (Hardy et al., 2003).
Virtual screening of compound libraries has become a
standard technology in modern drug discovery pipelines. If
a suitable structure of the target is available, molecular
docking can be used to discriminate between putative
binders and non-binders in large databases of chemicals
and to reduce the number of compounds to be subjected to
experimental testing substantially. Molecular docking is
one of the key computational chemistry techniques that are
routinely applied to drug discovery. The holy grail of
molecular docking is to replace experimental studies of
protein ligand complexes by modelling their structures and
binding affinities in silico (Novikov and Chilov, 2009)
Electronic supplementary material The online version of this
article (doi:10.1007/s00044-012-0023-0) contains supplementary
material, which is available to authorized users.
S. K. Bansal (&) ꢀ R. L. Khosa
Department of Medicinal Chemistry, School of Pharmacy,
Bharat Institute of Technology, Meerut 250103, UP, India
e-mail: skbansal@bitmeerut.edu.in; skbansal2003@gmail.com
B. N. Sinha
Division of Molecular Modelling, Department of Pharmaceutical
Sciences, Birla Institute of Technology, Mesra, Ranchi 835215,
Jharkhand, India
123

Med Chem Res (2013) 22:134–146
135
c-Amino butyric acid (GABA) is a predominant inhib-
itory neurotransmitter in mammalian CNS modulating
central inhibitory tone via activation of ionotropic GABAA
and GABAC receptor and G-Protein-coupled GABAB
receptor (Osolodkin et al., 2009; Smith and Simpson,
2003). c-Amino butyrate aminotransferase (GABA-AT)
catalyzes the degradation of GABA to succinic semialde-
hyde (SSA). Depleted levels of GABA have been shown to
cause convulsions (Karlsson et al., 1974). Raising GABA
levels in brain have an anticonvulsive effect (Krogsgaard–
Larsen, 1981). GABA-AT is a validated target for antiep-
ileptic drugs because its selective inhibition raises GABA
concentration in brain (Storici et al., 1999). Numerous
strategies exist to elevate GABA levels in the brain. The
strategy, which we have taken, involves the inhibition or
the inactivation of GABA-AT (Bansal et al., 2010, 2011a,
b, c; Nogardy and Weaver, 2005; Silverman and Clift,
2008; Sowa et al., 2005). GABA itself is not an effective
anticonvulsive agent since it does not cross the blood brain
barrier, a protective membrane that prevents xenobiotics
from entering the brain (Silverman et al., 1986). Conse-
quently, a real need exists to develop new anticonvulsant
compounds to cover seizures which are so far resistant to
presently available drugs. Current marketed antiepileptic
drugs consist of a variety of structural classes (lamotrigine,
oxcarbazepine, topiramate, gabapentin, and levetiracetam)
with different mechanisms of action. These agents typically
have non-overlapping efficacy and side effect profiles
presenting multiple treatment options for the patient pop-
ulation. However, approximately 30 % of seizure sufferers
fail to respond to current therapies. Currently, there is no
single drug of choice for treating all types of seizures. One
should focus on mechanism-driven discovery of novel
compounds followed by their evaluation by in vitro and in
vivo models to discover novel antiepileptic drugs. Several
recent successes (pregabalin, brivaracetam) have shown
that knowledge of the mechanism of action gives the
developer a significant advantage in improving efficacy
through increased target potency and selectivity, thereby
lowering the potential for dose-related side effects. It is the
hope that new generation AEDs with novel mechanisms
will increase the likelihood for success in treating a het-
erogeneous patient population (Gerlach and Krajewski,
2010).
Materials and methods
GABA-AT receptor modeling
The receptor model was prepared using AutoDock ToolsÒ
1.4.6 and MGL ToolsÒ 1.5.4 packages (The Scripps
Research Institute, Molecular Graphics Laboratory, 10550
North Torrey Pines Road, CA, 92037) running on Red Hat
Enterprise Linux 5.0.
It consists of several steps. First, the 3D crystal structure
of GABA-AT; PDB code 1OHV (Kwon et al., 1992;
Storici et al., 2005; Toney et al., 1995) was downloaded
from Brookhaven protein data bank (PDB; http://www.rcsb.
org/pdb) and loaded to python molecular viewer. The non-
bonded oxygen atoms of waters, present in the crystal
structure were removed. After assigning the bond orders,
missing hydrogen atoms were added, then the partial
atomic charges was calculated using Gasteiger–Marsili
(1980) method. Kollman and co-workers (1984) united
atom charges were assigned, non-polar hydrogens merged
and rotatable bonds were assigned, considering all the
amide bonds as non-rotatable. The receptor file was con-
verted to pdbqt format, which is pdb plus ‘‘q’’ charges and
‘‘t’’ AutoDock type. (To confirm to the AutoDock types,
polar hydrogens should be present where as non-polar
hydrogens and lone pair should be merged, each atom
should be assigned Gasteiger partial charges).
Since vigabatrin (Fig. 1) forms a covalent ternary
adduct with the active site LYS 329 (22) of GABA-AT,
therefore LYS 329 was included as flexible residue for
introducing conformational search of flexible side chain.
For the same macromolecule was saved in two files: one
containing the formatted, flexible LYS 329 residue, and the
other all the rest of the residues in the macromolecule.
Ligand modeling
ChemDraw Ultra 6.0.1 (Cambridge Soft.Com, 100 Cam-
bridge park drive, Cambridge, MA 02140, USA) was used
to draw the 3D structures of different ligands (Schiff’s
bases of GABA). Ligands were further refined and cleaned
in 3D by addition of explicit hydrogens and gradient
optimization function of MarvinSketch 5.0.6.1 (Chemaxon
Ltd; http://www.Chemaxon.com). All the structures were
written in Tripos mol2 file format.
A strategy along this line is to search for compounds
with new modes of action, a series of GABA with an imine
link to a wide variety of alkyl and aryl aldehydes, ketones
has been designed, synthesized, and screened
In this study, our goal was to apply computational tech-
niques in the pursuit of potential inhibitors of GABA-AT
enzyme activity. Molecular docking simulations were
employed to both find hit compounds and to rank the best fit
of the ligands.
Fig. 1 Chemical structure of vigabatrin
123

136
Med Chem Res (2013) 22:134–146
Input molecules files for an AutoDock experiments must
was used, since preliminary experiments using other two
(simulated annealing and genetic algorithm) showed that
they are less efficient, utilizes (discredited) Lamarckian
notation that an adaptations of an individual to its envi-
ronment can be inherited by its offspring. For all dockings,
˚
100 independent runs with step sizes of 0.2 A for transla-
˚
tions and 5 A for orientations and torsions, an initial pop-
confirm to the set of atom types supported by it. AutoDock
requires that, ligands give partial atomic charges and Auto-
Dock atom types for each atom; it also requires a description
of the rotatable bond in the ligand. AutoDock uses the idea of
a tree in which the rigid core of the molecule is a ‘‘root,’’ and
the flexible parts are ‘‘branches’’ that emanate from the root.
This set consists of united atom aliphatic carbons, aromatic
carbons in cycles, polar hydrogens, hydrogen-bonded
nitrogen, and directly hydrogen-bonded oxygen among
others, each with partial charges. Therefore, pdbqt format
was used to write ligands, recognized by AutoDock.
Torsional degree of freedom (TORSDOF) is used in
calculating the change in the free energy caused by the loss
of TORSDOF upon binding. In the AutoDock 4.0.1 force
field, the TORSDOF value for a ligand is the total number
of rotatable bonds in the ligand. This number excludes
bonds in rings, bonds to leaf atoms, amide bonds, and
guanidinium bonds.
ulation of random individuals with a population size of 150
individuals, a maximum number of 2.5 9 106 energy
evaluations, maximum number of generations of 27,000, an
elitism value of 1, a number of active torsion of 9 were
used.
AutoDock ToolsÒ along with AutoDock 4.0.1 and
AutoGrid 4.0.1 was used to generate both grid and docking
parameter files (i.e., gpf and.dpf files), respectively.
Chemistry
All reagents obtained from commercial sources were ana-
lytical reagent grade. Melting points were determined in
open capillary tubes on a Buchi melting point apparatus
and are uncorrected.
Molecular docking simulations
Prior to actual docking run, AutoGrid 4.0.1, was introduced
to pre-calculate grid maps of interaction energies of various
atom types (Goodford, 1985). In all dockings, a grid map
Infrared (IR) spectra were obtained on a Perkin Elmer
RX-1 FTIR spectrophotometer in the region 4,000–400
cm-1 using KBr disc. Proton nuclear magnetic resonance
(1H NMR) spectra were recorded on Brucker ADVANCE-
300 MHz using tetramethyl silane (TMS) as an internal
standard and chemical shifts are reported in parts per million
(ppm). Mass spectra were recorded on Shimadzu GCMS-
QP1000EX. The progress of the reaction was monitored by
ascending thin layer chromatography (TLC) on silica gel-G
(Merck)-coated aluminum plates, visualized by iodine vapor
and UV light. The eluant system was benzene:methanol
(8:2). Homogeneity of the compounds was checked by TLC
followed by high performance liquid chromatography
(HPLC), WATERS-2489, using same eluant system.
˚
with 60 9 60 9 60 points, a grid spacing of 0.375 A
(roughly a quarter of the length of a carbon–carbon single
bond) were used, and the maps were centered on the ligand
binding site. In an AutoGrid procedure, the protein is
embedded in a 3D grid and a probe atom is placed at each grid
point. The energy of interaction of this single atom with the
protein is assigned to the grid point. An affinity grid is cal-
culated for each type of atoms in the substrate, typically
carbon, oxygen, nitrogen, and hydrogens as well as grid of
electrostatic potential using a point charge of ?1 as the probe
(Allison et al., 1998; Sharp et al., 1987). AutoDock 4.0.1 uses
these interaction maps to generate ensemble of low energy
conformations (Goodsell and Olson, 1990; Morris et al.,
1996). It uses a scoring function based on AMBER force
field, and estimates the free energy of binding of a ligand to
its target. For each ligand atom types, the interaction energy
between the ligand atom and the receptor is calculated for the
entire binding site which is discretized through a grid. This
has the advantage that interaction energies do not have to be
calculated at each step of the docking process but only looked
up in the respective grid maps. Since a grid map represents
the interaction energy as a function of the coordinates, their
visual inspection may reveal the potential unsaturated
hydrogen acceptors or donors or unfavorable overlaps
between the ligand and the receptor.
General procedure for the synthesis of 4-amino-N0-aryl/
alkylidenebutanehydrazide (1–10) and 3-amino-
N0- aryl/alkylidenebenzohydrazide (22–26)
A solution of the amino acid (GABA/3-aminobenzoic acid;
AA; 100 mmol) in a mixture of dioxane (20 ml), water
(10 ml), and 1 M sodium hydroxide (10 ml) was stirred
and cooled in ice water bath. Di-t-butyl pyrocarbonate
(2.4 g, 110 mmol) was added and stirring was continued at
room temperature for 30 min. The solution after acidifi-
cation with dilute aqueous potassium hydrogen sulfate to
pH 2–3 was extracted with ethyl acetate (2 9 15 ml). The
ethyl acetate extracts were pooled, washed with water
(2 9 30 ml), dried over anhydrous sodium sulfate and
evaporated in vacuo. Recrystallization of the residue with
Of the three different search algorithms offered by
AutoDock 4.0.1, the Lamarckian genetic algorithm (LGA)
based on the optimization algorithm (Solis and Wets, 1981)
123

Med Chem Res (2013) 22:134–146
137
4-Amino-N0-[3-iodophenylmethylidene]butanehydrazide
(4)
ethyl acetate–hexane offered the N-protected amino acids
(BOC-AA). Equimolar quantities of BOC-AA (100 mmol)
and hydrazine hydrate (99–100 %) (100 mmol) were
condescend in the presence of dicyclohexylcarbodiimide
(DCC) (100 mmol) in dichloromethane, by stirring at ice
cold conditions (0–3 °C) for 7–9 h. The residue was
filtered, washed, dried, recrystallized, and yielded BOC-
AA-hydrazide. Imines were prepared by the reaction of
BOC-AA-hydrazide (100 mmol) with different aldehydes
and ketones (100 mmol) with the simultaneous removal of
water (by the addition of glacial acetic acid) for 10–12 h.
The BOC was removed by treatment of the different
products with equimolar quantities of neat trifluoroacetic
acid for 60 min. The final product was filtered, washed,
dried, and recrystallized using suitable solvents and used
further. The IR spectrum of 4-amino-N0-[aryl/alkyl-yli-
dene]butanehydrazide (1–10) and 3-amino-N0-aryl/alkyli-
denebenzohydrazide (22–26) showed the presence of peaks
at 3,370 (OH of COOH), 1,700 (CO of COOH), 1,620–
1H NMR (300 MHz, D2O): d 1.455–1.537 (m, 2H, CH2 b
to NH2), 1.870 (t, 2H, J = 2.7 Hz, CH2 c to NH2), 2.538 (t,
2H, J = 3.0 Hz, CH2 a to NH2), 6.769–8.018 (5H, aryl–H
and HC=N); MS (M?1)? m/z = 332.1.
4-Amino-N0-[1, 2, 3, 4-tetrahydronaphthalene-2-ylidene]
butanehydrazide (5)
1H NMR (300 MHz, CDCl3): d 1.818–2.056 (m, 6H,
3-CH2 of naphthyl, CH2 b to NH2 and NH2, D2O
exchangeable), 2.561–2.660 (m, 8H, CH2 a, c to NH2 and
1,4-CH2 of naphthyl), 7.268–8.770 (m, 4H, 5,6,7,8-CH of
naphthyl), 9.129 (s, H of CONH, D2O exchangeable); MS
?
(M?1) m/z = 246.1.
4-Amino-N0-(10-oxo-9, 10-dihydroanthracene-
9-ylidene)butanehydrazide (6)
1,590 (C:N) cm-1
.
1H NMR (300 MHz, d ppm) spectra and m/z of titled
compounds are as follows:
1H NMR (300 MHz, CDCl3): d 1.530–1.600 (m, 2H, CH2
b to NH2), 1.803 (s, 2H, NH2, D2O exchangeable), 2.142 (t,
2H, J = 8.4 Hz, CH2 c to NH2), 2.649 (t, 2H, J = 8.4 Hz,
CH2 a to NH2), 6.965–8.101 (m, 8H, aryl–H), 9.630 (s, H
of CONH, D2O exchangeable); MS (M?1)? m/z = 308.6.
4-Amino-N0-[1-(2-bromophenyl)ethylidene]
butanehydrazide (1)
1H NMR (300 MHz, DMSO): d 0.855 (s, 3H, CH3),
1.217–1.509 (m, 2H, CH2 b to NH2), 1.993 (s, 2H, NH2,
D2O exchangeable), 2.196 (t, 2H, J = 12.3 Hz, CH2 c to
NH2), 2.767 (t, 2H, J = 9.3 Hz, CH2 a to NH2),
7.502–7.727 (m, 4H, aryl–H), 9.964 (s, H of CONH, D2O
exchangeable); MS (M?1)? m/z = 298.9.
4-Amino-N0-[2-iodophenylmethylidene]butanehydrazide
(7)
1H NMR (300 MHz, DMSO): d 1.257–1.342 (m, 2H, CH2
b to NH2), 1.600 (t, 2H, J = 5.1 Hz, CH2 c to NH2), 2.058
(s, 2H, NH2, D2O exchangeable), 2.520 (t, 2H, J = 5.1 Hz,
CH2 a to NH2), 7.522–8.038 (6H, aryl–H and HC=N),
9.003 (s, H of CONH, D2O exchangeable); MS (M?1)?
m/z = 332.0.
4-Amino-N0-[1-(3-chlorophenyl)ethylidene]
butanehydrazide (2)
1H NMR (300 MHz, DMSO): d 0.735 (s, 3H, CH3),
1.397–1.822 (m, 2H, CH2 b to NH2), 1.995 (s, 2H, NH2,
D2O exchangeable), 2.254 (t, 2H, J = 6.3 Hz, CH2 c to
NH2), 2.781 (t, 2H, J = 6.6 Hz, CH2 a to NH2),
7.453–7.698 (m, 4H, aryl–H), 9.375 (s, H of CONH, D2O
exchangeable); MS (M?1)? m/z = 254.5.
4-Amino-N0-[1,2,3,4-tetrahydronaphthalen-1-ylidene]
butanehydrazide (8)
1H NMR (300 MHz, CDCl3): d 1.132–1.391 (m, 2H,
2-CH2 of naphthyl), 1.730 (s, 2H, NH2, D2O exchange-
able), 1.894–2.006 (m, 4H, 3-CH2 of naphthyl and CH2 b
to NH2), 2.597–2.719 (m, 6H, 4-CH2 of naphthyl and CH2
a, c to NH2), 7.196–7.876 (m, 4H, 5,6,7,8-CH of naphthyl),
8.721 (s, H of CONH, D2O exchangeable); MS (M?1)?
m/z = 246.8.
4-Amino-N0-[2-methyl-10-oxo-9,10-dihydroanthracene-
9-ylidene]butanehydrazide (3)
1H NMR (300 MHz, CDCl3): d 1.621–1.726 (m, 2H, CH2
b to NH2), 1.920 (s, 2H, NH2, D2O exchangeable), 2.252 (t,
2H, J = 8.1 Hz, CH2 c to NH2), 2.569 (s, 3H, 2-CH3 of
10-oxo-9,10-dihydroanthracene), 2.768 (t, 2H, J = 6.9 Hz,
CH2 a to NH2), 7.261–7.782 (m, 7H, aryl–H), 9.822 (s, H
of CONH, D2O exchangeable); MS (M?1)? m/z = 322.5.
4-Amino-N0-[2-oxo-1, 2-diphenylethylidene]
butanehydrazide (9)
1H NMR (300 MHz, DMSO): d 1.540–1.630 (m, 2H, CH2
b to NH2), 2.091 (s, 2H, NH2, D2O exchangeable), 2.320 (t,
123

138
Med Chem Res (2013) 22:134–146
General procedure for the synthesis of 4-(aryl/
2H, J = 10.8 Hz, CH2 c to NH2), 2.752 (t, 2H, J = 8.1 Hz,
CH2 a to NH2), 7.094–8.372 (m, 10H, aryl–H), 9.081 (s, H
of CONH, D2O exchangeable); MS (M?) m/z = 309.2.
alkylideneamino) butanoic acid (11, 13–21)
and 3-(aryl/alkylideneamino) benzoic acid (27–31)
To a suspension of amino acid (GABA/3-aminobenzoic
acid; AA; 100 mmol) in absolute ethanol (80 ml) was
added concentrated sulphuric acid (15 ml). The resulting
yellow solution was heated at a reflux for 3 h, cooled to
0 °C and neutralized with concentrated aqueous ammonia
solution. The precipitated product was then collected by
filtration, washed with cold water, and recrystallized from
aqueous ethanol to give pure amino acid ester. A mixture
of amino acid ester (100 mmol), carbonyl compound
(100 mmol), and catalytic amount of glacial acetic acid
was added to absolute ethanol (60 ml). The mixture was
refluxed for 4 h and cooled. The resultant precipitate was
filtered off, washed with water (3 9 100 ml), dried over
anhydrous magnesium sulfate. The resulted product was
boiled with 80 ml of 10 % potassium hydroxide under
reflux for 1 h and the liquid was distilled off through same
condenser. Residue in the flask (potassium salt of product)
was acidified with dilute sulphuric acid, separated product
was filtered, washed water and, ethyl ether and dried over
anhydrous magnesium sulfate. Recrystallization with
alcohol and column purification led to the desired
derivatives.
4-Amino-N0-[4-bromophenyl(phenyl)methylidene]
butanehydrazide (10)
1H NMR (300 MHz, DMSO): d 1.671–1.804 (m, 2H, CH2
b to NH2), 1.908 (s, 2H, NH2, D2O exchangeable), 2.282 (t,
2H, J = 5.4 Hz, CH2 c to NH2), 2.735(t, 2H, J = 7.2 Hz,
CH2 a to NH2), 6.780–8.225 (m, 9H, aryl–H), 9.096 (s, H
of CONH, D2O exchangeable); MS (M?1)? m/z = 360.4.
3-Amino-N0-[1, 2-dihydroxy-10-oxo-9,10-
dihydroanthracene-9-ylidene]benzohydrazide (22)
1H NMR (300 MHz, CDCl3): d 4.164 (s, 2H, NH2, D2O
exchangeable), 4.948 (s, 2H, 9,10-dihydroanthracene-1,
2-OH, D2O exchangeable), 6.767–7.436 (m, 10H, aryl–H),
9.002 (s, H of CONHN, D2O exchangeable); MS (M?)
m/z = 373.5.
3-Amino-N0-[1-(3-chlorophenyl)ethylidene]benzohydrazide
(23)
1H NMR (300 MHz, CDCl3): d 0.859 (S, 3H, CH3), 4.662
(s, 2H, NH2, D2O exchangeable), 6.761–7.410 (m, 8H,
aryl–H), 8.009 (s, H of CONHN, D2O exchangeable); MS
(M?1)? m/z = 288.0.
The IR spectrum of 4-(aryl/alkylideneamino)butanoic
acid (11, 13–21) and 3-(aryl/alkylideneamino)benzoic acid
(27–31) showed the presence of peaks at 3,220
(NH stretching vibration), 1,650–1,630 (amide bond),
1,620–1,590 (C:N) cm-1
.
1H NMR (300 MHz, d ppm) spectrum and m/z of titled
compounds are as follows:
3-Amino-N0-[4-chlorophenyl(phenyl)methylidene]
benzohydrazide (24)
1H NMR (300 MHz, CDCl3): d 3.676 (s, 2H, NH2, D2O
exchangeable), 6.716–8.073 (m, 13H, aryl–H), 8.184 (s, H
of CONHN, D2O exchangeable); MS (M?) m/z = 349.0
4-{[1,5-Diphenylpenta-1,4-dien-3-ylidene]amino}butanoic
acid (11)
1H NMR (300 MHz, CDCl3): d 0.850 (t, 2H, J = 7.5 Hz,
CH2 c to COOH), 1.649–1.787 (m, 2H, CH2 b to COOH),
2.515 (t, 2H, J = 7.8 Hz, CH2 a to COOH), 5.785 (d, 2H,
J = 3.9 Hz, diene-CH), 6.557–7.766 (m, 12H, aryl–H),
11.203 (s, 1H, COOH, D2O exchangeable); MS (M?1)?
m/z = 320.8.
3-Amino-N0-[2-chloro-10-oxo-9,10-dihydroanthracene-
9-ylidene]benzohydrazide (25)
1H NMR (300 MHz, CDCl3): d 4.715 (s, 2H, NH2, D2O
exchangeable), 6.771–7.462 (m, 13H, aryl–H), 8.655 (s, H
of CONHN, D2O exchangeable); MS (M?) m/z = 375.0
4-{[1,3,3-Trimethylbicyclo[2.2.1]heptan-2-ylidene]
amino}butanoic acid (13)
3-Amino-N0-(2,3,5,6-tetrachloro-4-oxocyclohexa-2,5-dien-
1-ylidene)benzohydrazide (26)
1H NMR (300 MHz, CDCl3): d 0.797 (d, J = 5.4 Hz, 3H,
1-CH3), 0.903–0.935 (m, 8H, 1,3-CH3 and CH2 c to
COOH), 0.950–1.020 (m, 2H, 5-CH2) 1.075–1.146 (m, 2H,
6-CH2), 1.409–1.463 (m, 1H 4-CH), 1.587–1.682 (m, 2H,
CH2 b to COOH), 1.887–1.991 (m, 2H, bridged-CH2),
2.148–2.224, 2.763–2.826 (m, 2H, CH2 a to COOH),
1H NMR (300 MHz, DMSO) d 4.237 (s, 2H, NH2, D2O
exchangeable), 6.937–7.385 (m, 4H, aryl–H), 8.403 (s, H
of CONHN, D2O exchangeable); MS (M?1)? m/z =
378.2.
123

Med Chem Res (2013) 22:134–146
139
10.455 (s, 1H, COOH, D2O exchangeable); MS (M?1)?
m/z = 238.0.
6.705–6.770 (m, 1H, 3-cyclohexyl-CH); MS (M?1)?
m/z = 236.3.
4-{[3,5,5-Trimethylcyclohex-2-en-1-ylidene]amino}
butanoic acid (19)
4-{[10-Oxo-9,10-dihydrophenanthrene-9-ylidene]amino}
butanoic acid (14)
1H NMR (300 MHz, DMSO): d 1.067 (s, 6H, 5,5-dicy-
clohexyl-CH3), 1.321 (s, 2H, 6-cyclohexyl-H), 1.496 (t,
2H, J = 3.0 Hz, CH2 c to COOH), 1.712–1.758 (m, 2H,
CH2 b to COOH), 1.917 (s, 3H, 3-cyclohexyl-CH3),
2.090(s, 2H, 4-cyclohexyl-H), 2.241 (t, 2H, J = 7.2 Hz,
CH2 a to COOH), 4.695(s, 1H, 2-cyclohexyl-H), 10.207 (s,
1H, COOH, D2O exchangeable); MS (M?) m/z = 223.0.
1H NMR (300 MHz, CDCl3): d 1.980–2.006 (m, 2H, CH2
b to COOH), 2.231(t, 2H, J = 3.3 Hz, CH2 c to COOH),
3.410 (t, 2H, J = 2.7 Hz, CH2 a to COOH), 6.960–7.600
(m, 8H, aryl–H), 11.204 (s, 1H, COOH, D2O exchange-
able); MS (M?) m/z = 293.7.
4-{[2-Oxo-1, 2-dihydroacenaphthylene-1-ylidene]amino}
butanoic acid (15)
4-{[2-Methyl-4-oxocyclohexa-2,5-dien-1-ylidene]amino}
butanoic acid (20)
1H NMR (300 MHz, CDCl3): d 1.920–2.005 (m, 2H, CH2
b to COOH), 2.320 (t, 2H, J = 6.9 Hz, CH2 c to COOH),
3.702 (t, 2H, J = 4.5 Hz, CH2 a to COOH), 7.025–7.748
(m, 6H, aryl–H), 10.979 (s, 1H, COOH, D2O exchange-
able); MS (M?1)? m/z = 268.8.
1H NMR (300 MHz, D2O): d 1.258 (t, 2H, J = 6.3 Hz, CH2
c to COOH), 1.730–1.820 (m, 2H, CH2 b to COOH), 2.018(s,
3H, CH3), 2.190 (t, 2H, J = 3.9 Hz, CH2 a to COOH), 6.495
(d, 1H, J = 4.2 Hz, cyclohexyl-CH), 6.600 (d, 1H,
J = 3.0 Hz, cyclohexyl-CH); MS (M?) m/z = 206.9.
4-{[2-Methylcyclohexylidene]amino}butanoic acid (16)
4-{[1-(Naphthalen-2-yl)ethylidene]amino}butanoic acid
(21)
1H NMR (300 MHz, CDCl3): d 0.757–1.040 (m, 9H,
2-cyclohexyl-CH3, CH2 c to COOH and 3,4-cyclohexyl-
CH2), 1.253–1.280 (m, 2H, 5-cyclohexyl-CH2), 1.588–
1.667 (m, 2H, CH2 b to COOH), 1.917–2.695 (m, 5H,
CH2 a to COOH, 2-cyclohexyl-CH, 6-cyclohexyl-CH2),
10.146 (s, 1H, COOH, D2O exchangeable); MS (M?1)?
m/z = 198.1.
1H NMR (300 MHz, CDCl3): d 0.995 (s, 3H, CH3),
1.606–1.689 (m, 2H, CH2 b to COOH), 2.252 (t, 2H,
J = 8.7 Hz, CH2 c to COOH), 3.659–3.690 (t, 2H, J =
7.2 Hz, CH2 atoCOOH), 6.956–7.378(m, 7H, aryl–H), 10.907
(s, 1H, COOH, D2O exchangeable); MS (M?1)? m/z = 256.2.
4-{[5-Methyl-2-(propan-2-yl)cyclohexylidene]amino}
butanoic acid (17)
3-{[2-Ethyl-10-oxo-9,10-dihydroanthracene-9-ylidene]
amino}benzoic acid (27)
1H NMR (300 MHz, DMSO):
d
0.790 (d, 9H,
1H NMR (300 MHz, CDCl3): d 0.833 (t, 3H, J = 2.1 Hz,
CH3), 2.698–2.714 (m, 2H, CH2), 7.2579–8.179 (m, 11H,
aryl–H), 11.380 (s, 1H, COOH, D2O exchangeable); MS
(M?) m/z = 355.5.
J = 32.6 Hz, 5-methyl-2-(propan-2-yl)), 1.256–1.314 (t,
2H, J = 8.7 Hz, CH2 c to COOH), 1.557 (t, 2H,
J = 14.1 Hz, 4-cyclohexyl-CH2), 1.940 (d, 2H, J =
13.2 Hz, 2,5-cyclohexyl-CH), 2.151–2.241 (m, 6H, CH2 b
to COOH, 3,5-cyclohexyl-CH2), 2.751–2.754 (m, 3H, CH2
a to COOH, propan-2-yl-CH), 9.752 (s, 1H, COOH, D2O
exchangeable); MS (M?) m/z = 239.2.
3-[(1,4,5,8-Tetrachloro-10-oxo-9,10-dihydroanthracene-
9-ylidene)amino]benzoic acid (28)
1H NMR (300 MHz, DMSO): d 5.618–7.156 (m, 8H, aryl–
H), 11.402 (s, 1H, COOH, D2O exchangeable); MS
(M?1)? m/z = 464.2.
4-{[2-Methyl-5-(prop-1-en-2-yl)cyclohex-2-en-1-
ylidene]amino}butanoic acid (18)
1H NMR (300 MHz, D2O): d 1.226 (t, 2H, J = 6.6 Hz,
6-cyclohexyl-CH2), 1.380 (t, 2H, J = 3.0 Hz, CH2 c to
COOH), 1.475 (t, 2H, J = 3.6 Hz, 4-cyclohexyl-CH2),
1.580–1.665 (m, 2H, CH2 b to COOH), 1.842 (s, 6H, CH3),
2.175 (t, 3H, J = 5.4 Hz, CH2 a to COOH, 5-cyclohexyl-
CH), 4.750 (s, 2H, 5-cyclohexyl-{1-prop-1-en-2-yl-CH2}),
3-[(10-Oxo-9,10-dihydrophenanthrene-9-ylidene)
amino]benzoic acid (29)
1H NMR (300 MHz, DMSO): d 6.909–7.566 (m, 12H,
aryl–H) 11.251 (s, 1H, COOH, D2O exchangeable); MS
(M?1)? m/z = 328.2.
123

140
Med Chem Res (2013) 22:134–146
3-[(10-Oxo-9,10-dihydroanthracene-9-ylidene)amino]
benzoic acid (30)
3-[(2,3,5,6-Tetramethyl-4-oxocyclohexa-2,5-dien-1-
ylidene)amino]benzoic acid (31)
1H NMR (300 MHz, DMSO): d 7.360–8.688 (m, 12H,
aryl–H), 12.108 (s, 1H, COOH, D2O exchangeable); MS
(M?1)? m/z = 328.2.
1H NMR (300 MHz, CDCl3): d 1.520 (s, 6H, 2,6-cyclohexyl-
CH3), 1.898 (s, 6H, 3, 5-cyclohexyl-CH3), 6.898–7.501 (m,
4H, aryl–H), 9.824 (s, 1H, COOH, D2O exchangeable); MS
(M?1)? m/z = 284.7.
Fig. 2 Designed GABA
analogues
Table 1 Predicted and experimentally determined values of inhibition constant of acid hydrazones of GABA
O
H2N
N
R
NH
S. no. Compound
code
R
Observed binding Root mean square
energy (kcal/mol) deviation (RMSD, A) constant (Ki, lM)
Predicted inhibition Experimental
˚
inhibition constant
(IC50, lM)
1.
2.
3.
AHG177
AHG174
AHG202
1-(2-Bromophenyl)ethan-1-one
1-(3-Chlorophenyl)ethan-1-one
-9.08
-8.95
25.76
26.79
26.96
0.221
0.277
0.307
0.073 0.005
0.091 0.007
0.11 0.03
2-Methyl-9,10-dihydroanthracene- -8.88
9,10-dione
4.
5.
AHG066
AHG144
3-Iodobenzaldehyde
-8.76
-8.76
25.03
24.70
0.376
0.376
0.15 0.06
0.13 0.07
1,2,3,4-Tetrahydronaphthalen-
2-one
6.
AHG173
AHG063
AHG143
AHG161
AHG182
9,10-Dihydroanthracene-9,10-dione -8.74
2-Iodobenzaldehyde -8.68
1,2,3,4-Tetrahydronaphthalen-1-one -8.47
1,2-Diphenylethane-1,2-dione -8.35
4-Bromophenyl(phenyl)methanone -8.15
26.02
24.21
25.33
27.23
23.87
0.392
0.433
0.614
0.752
1.050
0.14 0.10
0.20 0.11
0.28 0.09
0.30 0.13
0.56 0.14
7.
8.
9.
10.
123

Med Chem Res (2013) 22:134–146
141
In vitro GABA transaminase activity
fractions were ultracentrifuged in 105,0009g for 1 h, and
supernatant was used as enzymatic source in the GABA
transaminase assay.
The enzymatic activity was determined by coupling the
transaminase reaction (Reaction 1) with an excess of SSA
dehydrogenase (Reaction 2), so that the formation of
reduced pyridine nucleotide is a function of transaminase
activity (Scott and Jkoby, 1959; Jakoby, 1962). The reac-
tion was followed spectrophotometrically at 340 nm .
GABA (0.060 M, 0.30 ml), a-ketoglutarate (0.100 M,
0.15 ml), 2-mercaptoethanol (0.100 M, 0.10 ml), potassium
pyrophosphate buffer, pH 8.6 (0.100 M, 2.30 ml), tissue
homogenate (1 U/ml, 0.02 ml) were incubated in 37 °C for
30 min, followed by the addition of b-NADP (0.025 M,
0.15 ml) and excess of succinic semialdehyde dehydroge-
nase. The amount of NADPH generated in the brain tissue
for 20 min was measured spectrophotometrically at 340 nm
using potassium phosphate buffer (0.075 M) and glycerol
(25 % v/v) as enzyme dilutor for blank. Increase in A340 was
observed and recorded. Consecutively, GABA was replaced
by synthesized compounds (0.1–100 lM) and positive
control vigabatrin (5–50 lM) and IC50 was determined.
A linear curve of NADPH concentration versus absor-
bance was obtained at a wavelength of 340 nm. One unit
(1 U) of enzyme activity is equivalent to conversion of
1.0 lmol of GABA to SSA and then to succinate per
minute with a stoichiometric reduction of 1.0 lmol of
NADP. Protein concentration was determined by the
method of Bradford (1976).
c - Amino butyrate þ a - ketoglutarte
$ succinic semialdehyde þ glutamate
ð1Þ
ð2Þ
Succinic semialdehyde þ b - NADP
! succinic acid þ NADPH
Rat brain GABA-AT was partially purified by the
method described by Koo et al., (2003) and Ricci et al.,
(2006). Enzyme preparation procedures were carried out at
4 °C, unless specified otherwise. The whole brain was
isolated, and then homogenized with a glass Teflon
homogenizer in four volumes of 0.1 M potassium
phosphate buffer (pH 7.4). Homogenate were centrifuged
in 6009g for 10 min, supernatant was collected and
recentrifuged in 10,0009g for 20 min. Postmitochondrial
Table 2 Predicted and experimentally determined values of inhibition constant of Schiff’s bases of GABA
O
R
N
OH
S. no.
Compound
code
R
Observed
binding
energy
Root mean
square
deviation
Predicted
inhibition
constant
(Ki, lM)
Experimental
inhibition
constant
˚
(RMSD, A)
(kcal/mol)
(IC50, lM)
11.
12.
SBG164
SBG195
1,5-Diphenylpenta-1,4-dien-3-one
-6.12
-6.05
14.24
17.10
32.74
36.76
9.95 0.11
–
1-Methyl-9,10-
dihydroanthracene-9,10-dione
13.
SBG110
1,3,3-Trimethylbicyclo[2.2.1]heptan-
2-one
-5.91
26.03
46.59
10.02 0.08
14.
15.
16.
17.
SBG171
SBG211
SBG099
SBG116
9,10-Dihydrophenanthrene-9,10-dione
1,2-Dihydroacenaphthylene-1,2-dione
2-Methylcyclohexan-1-one
-5.83
-5.71
-5.66
-5.66
26.69
25.70
15.20
25.17
53.72
65.72
70.45
70.55
15.53 0.35
16.78 1.03
20.76 2.46
20.82 1.89
5-Methyl-2-(propan-2-yl)cyclohexan-
1-one
18.
SBG120
2-Methyl-5-(prop-1-en-2-yl)cyclohex-
2-en-1-one
-5.61
25.18
77.83
23.48 3.02
19.
20.
SBG117
SBG192
3,5,5-Trimethylcyclohex-2-en-1-one
-5.60
-5.59
24.42
14.08
78.22
79.33
25.94 3.55
30.12 3.00
2-Methylcyclohexa-2,5-diene-1,4-
dione
21.
SBG153
1-(Naphthalen-2-yl)ethan-1-one
-5.54
15.41
87.49
44.06 5.23
123

142
Med Chem Res (2013) 22:134–146
Result and discussion
runs. The central processing unit for a single docking
experiment took 75–120 min, on a 2.19-GHz Intel
(R) core2 Duo machine with 2.96 GB of RAM and Red
Hat Enterprise Linux 5.0 operating system.
The design of new and selective inhibitors of an enzyme is
one of the most important applications in contemporary
rational drug design. A total of 932 GABA derivatives
were designed retaining the original structure of GABA, as
shown in Fig. 2. To study GABA-AT inhibition, all the
designed molecules were docked and the results of LGA
docking experiments of different GABA analogs using
AutoDock 4.0.1 and AutoGrid 4.0.1 are summarized in
Tables 1, 2, and 3. For each docking experiment, the
lowest energy docked conformation was selected from 100
To evaluate accuracy of docking, lower value of binding
energy and higher value of inhibition constant (Ki) were set as
a criteria for in silico screening. 31 flexible docks were con-
sidered well docked with the binding energy values\-5.54
kcal/mol (binding energy of vigabatrin). Modeling and
docking analysis reveal the nature of the active site and
some key interactions that enable the strong binding of
acid hydrazones in comparison to Schiff’s bases. Docking
Table 3 Predicted and experimentally determined values of inhibition constant of acid hydrazones and schiff’s bases of 3-amino benzoic acid
R2
O
N
R1
R3
S. no. Compound
code
R1
R2
R3
Observed Root mean Predicted Experimental
binding
energy
(kcal/
mol)
square
deviation
(RMSD,
inhibition inhibition
constant constant
(Ki, lM) (IC50, lM)
˚
A)
22.
23.
24.
25.
26.
27.
AHABA 199
AHABA 174
AHABA 181
AHABA 215
AHABA 201
H
H
H
H
H
H
H
H
H
H
10-Hydrazinylidene-7,8-dihydroxy-
9,10-dihydroanthracene-9-one
-8.38
-8.20
-8.11
-8.03
24.46
26.89
22.10
23.54
21.25
25.26
0.72
0.98
1.13
1.29
1.81
3.86
0.24 0.03
0.32 0.02
0.37 0.01
0.46 0.02
–
[1-(3-Chlorophenyl)ethylidene]
hydrazine
4-Chlorophenyl(phenyl)methylidene]
hydrazine
7-Chloro-10-hydrazinylidene-4a,9,9a,
10-tetrahydroanthracene-9-one
2,3,5,6-Tetrachloro-4-hydrazinylidene- -7.83
cyclohexa-2,5-dien-1-one
SBABA 210 2-Ethyl-9,10-dihydro-
anthracene-
9,10-dione
OH
-7.39
-7.36
1.21 0.06
28.
SBABA 204 1,4,5,8-Tetrachloro-9,
OH
17.42
4.04
–
10-dihydroanthra-
cene-9,
10-dione
29.
30.
SBABA 171 9,10-Dihydrophe-
nanthrene-9,10-dione
OH
OH
-6.72
-6.68
24.77
25.36
11.81
12.78
2.78 0.21
4.96 1.01
SBABA 173 9,10-Dihydroan-
thracene-9,
10-dione
31.
32.
SBABA 203 2,3,5,6-Tetramethylcy- OH
clohexa-2,
5-diene-1,4-dione
-6.37
-5.54
26.05
19.98
21.46
86.55
6.54 1.69
Vigabatrin
–
–
41.21 3.38
123

Med Chem Res (2013) 22:134–146
143
interactions of 4-amino-N0-[1-(2-bromophenyl)ethylidene]
butanehydrazide (AHG177), 3-amino-N0-[1,4-dihydroxy-
10-oxo-9,10-dihydroanthracene-9-ylidene]benzohydrazide
(AHABA199), 4-{[1,5-diphenylpenta-1,4-dien-3-ylidene]
amino}butanoic acid (SBG164), and 3-{[2-ethyl-10-oxo-
9,10-dihydroanthracene-9-ylidene]amino}benzoic acid
(SBABA210) with Lys 329 appears to be in close prox-
imity and explains the high GABA-AT inhibitory activity
observed in their respective series. Docking poses and
binding interactions of vigabatrin, AHG177, AHABA199,
SBG164, and SBABA210 are shown in Figs. 3, 4, and
Table 4. The internal aldimine linkage between Lys-329
and PLP of the enzyme has been broken. Instead, a cova-
lent ternary adduct is formed among the cofactor, the
inhibitor, and the lysine residue. The presence of aromatic
ring was found to play a major role in determining inhib-
itory activity for GABA-AT. An amino acid ILE72 and
SER74 seems to be in key interactions with ligands.
Only 5–10 % of control activity is restored upon exhaus-
tive dialysis for 4 days, suggesting that the observed inhi-
bition is due to covalent bond formation between the
inhibitor and the enzyme.
All the synthesized compounds have been characterized
by physicochemical and spectral studies and the data was
within accordance of theoretical values. In general, the IR
spectra showed C:N peak at 1,620–1,590 cm-1 and
1
characteristic amide bonds at 1,650–1,630 cm-1. The H
NMR spectrum reveled that the hydrazino proton (=N–NH)
showed a singlet at d 8.55–10.00. All compounds showed a
characteristic D2O exchangeable signal due to COOH
proton of acid function group at d 11.00–11.55. The aro-
matic ring proton resonate at d is *7.25–7.6 ppm.
A discussion of structure–activity relationships of the
GABA-AT inhibitors is, at best, tentative. Aminobutyrate
transaminase is a pyridoxal-P enzyme which follows a
bi–bi ping pong mechanism and in pyridoxamine form can
readily transaminate only with SSA and 2-oxoglutarate.
The above results strongly suggest that only the pyridoxal
form of the enzyme is capable of reacting with the ligands
(Fig. 6). The Pharmacophore observed from the ongoing
studies is indicated in Fig. 3. The 4-amino-N0-[(1Z)-
1-(2-bromophenyl)ethylidene]butanehydrazide (AHG177)
has been identified as a highly potent inhibitor of the GABA-
AT. This compound has displayed significant activity at the
inhibitory concentration (IC50) of 0.073 lM.
In addition, a number of acid hydrazones have shown to
be powerful GABA-AT inhibitors. From these studies, and
from related molecular modeling investigations it became
apparent that the majority of the inhibitors are closely
related to the structure of GABA, as the values of inhibi-
tion constant of GABA derivatives was observed higher
than 3-aminobenzoic acid derivatives. Moreover, the
presence of –CONH– group and free terminal NH2 group
afforded potent inhibitors. This may be through the
hydrogen bond formation. Introduction of acetophenone
In order to rationalize the putative binding mode of
newly designed compounds, few of them were synthesized
as per Schemes 1 and 2 and screened for in vitro GABA
transaminase inhibitory activity. The biological results
were found in agreement with the docking results and are
shown in Fig. 5. Incubation of GABA analogs with rat
brain aminobutyrate results in a rapid, irreversible, and
complete loss of biological activity. The inactivation is
progressive with time and follows pseudo-first-order
kinetics. Enzymatic half life ranges from 11 to 20 min.
substituent yielded another agent (AHG174; IC50
=
0.091 lM) with potent GABA-AT enzyme affinity. In
these inhibitors, replacement of acetophenone with
anthraquinone and benzaldehyde derivatives caused the
reduction in potential. Likewise, halogen-substituted ana-
logs inhibited the enzyme more firmly, probably because of
higher electronegativity. Introduction of rigidity (decreased
number of conformation of methyl chain between amino
and carboxyl group) caused the reduction in inhibitory activity.
Replacement of COOH group by CONHN(R1R2)
showed variable effects like; (a) R1: H or CH3 and R2:
2-bromophenyl, 3-chlorophenyl, 3-iodophenyl, and
2-iodophenyl (in decreasing order of potency); (b) R1 and
R2 may be collectively replaced by polynuclear aromatic
hydrocarbon: 2-methylanthraquinone, b-tetralone, anthra-
quinone, a-tetralone, benzyl, alizarin, benzophenone,
Fig. 3 Docking predicted poses and interactions of vigabatrin with
GABA-AT (structure of vigabatrin is displayed in green) (Color
figure online)
123

144
Med Chem Res (2013) 22:134–146
Fig. 4 Docking predicted poses
and interactions of AHG177 (a),
AHABA199 (b), SBG164 (c),
and SBABA210 (d) with
GABA-AT (structure of ligands
are displayed in green)
(Color figure online)
Table 4 Binding interactions of AHG177, AHABA199, SBG164, and SBABA210
Compound
AHG177
Amino acid in vicinity
Interactions observed
ILE72, LYS329, ACT500, and
PLP600
Hydrophobic interaction between LYS329 and phenyl group
Hydrophobic interaction between ILE72 and CH3 of acyl group
Hydrogen bonding interactions between PLP600 and amide NH and terminal
NH2 group
Hydrogen bonding interactions between ACT500 and terminal NH2 group
AHABA199
GLN71, ILE72, SER74,
CYS135 LYS329, ACT500,
and PLP600
Hydrogen bonding interactions between SER74 and oxo group of
anthraquinone nucleus
Hydrogen bonding interactions between PLP600 and amide NH
Hydrogen bonding interactions between ACT500 and terminal NH2 group
Hydrophobic interaction between ILE72 and phenyl group
SBG164
ILE72, SER74, ALA 134,
CYS135, SER137, ARG192,
LYS329, and PLP600
Hydrogen bonding interactions between SER137 and hydroxyl group of
terminal carboxylic group
SBABA210
ILE72, SER74, CYS135,
LYS329, ACT500, and
PLP600
Hydrophobic interaction between ILE72 and anthraquinone group
123

Med Chem Res (2013) 22:134–146
145
Scheme 1 Scheme for the
synthesis of 1–10 and 22–26.
a Dioxane, water, IM NaOH,
BOC (di-t-butyl pyrocarbonate);
b dichloromethane, DCC,
NH2NH2.H20; c EtOH, AcOH,
R1.CO.R2, reflux; d trifluoro
acetic acid
Scheme 2 Scheme for the
synthesis of 11, 13–21 and
27–31. a EtOH, H2SO4, reflux;
b EtOH, AcOH, R1.CO.R2,
reflux; c water, KOH, reflux,
dilute H2SO4
Fig. 5 Effect of GABA analogs on rat brain GABA transaminase
2-chloroanthraquinone, and chloroanil (in decreasing order
of potency). Formation of imine link with free NH2 further
reduces the potency in the order: Dibenzylideneacetone,
fenchone, 2-methylcyclohexanone, menthone, carvone, and
isophorone.
This study contributes molecular insight into the binding
process, which is of great pivotal importance for designing
new ligands interfering with GABA-AT and shows that
new wave of flexible ligand docking program like Auto-
Dock can produce unbiased docking of GABA-AT inhib-
itors in the enzyme active site. There is still significant
Fig. 6 Proposed mechanism of inhibition of aminobutyrate transam-
inase with acid hydrazones
123

146
Med Chem Res (2013) 22:134–146
Jakoby WB (1962) Enzymes of c-aminobutyrate metabolism. Meth-
ods Enzymol 5:771–774
room for improvement especially for the empirical binding
free energy force field and inhibition constant prediction.
Designing and synthesis followed by in vitro interactions
provide us important conclusions; decreased number of
conformation of methyl chain between amino and carboxyl
group, i.e., increase in rigidity causes the reduction in
inhibitory activity observed, introduction of halogen atoms
especially bromo and chloro significantly improved the
activity. Free terminal NH2 group and –CONH– group is
mandatory for the activity. Thus, our findings open up the
possibility to extend this protocol to different databases in
order to find new potential inhibitor for promising targets
based on a rational drug design process.
Karlsson A, Fonnum F, Malthe-Sørenssen D, Storm-Mathisen J (1974)
Effect of convulsive agent 3-mercaptopropionic acid on the levels
of GABA other amino acids and glutamate decarboxylase in
different regions of rat brain. Biochem Pharmacol 23:3053–3061
Koo BS, Park KS, Ha JH, Park JH, Lim JC, Lee DU (2003) Inhibitory
effects of the fragrance inhalation of essential oil from Acorus
gramineus on central nervous system. Biol Pharm Bull 26:978–982
Krogsgaard-Larsen P (1981) Gamma-aminobutyric acid agonists
antagonists and uptake inhibitors Design and therapeutic aspects.
J Med Chem 24:1377–1383
Kwon OS, Park J, Churchich JE (1992) Brain 4-aminobutyrate
aminotransferase: isolation and sequence of a cDNA encoding
the enzyme. J Biol Chem 267:7215–7216
Morris GM, Goodsell DS, Huey R, Olson AJ (1996) Distributed
automated docking of flexible ligands to proteins: parallel applica-
tions of Autodock 2.4. J Comput Aided Mol Des 10:293–304
Nogardy T, Weaver DF (2005) Medicinal chemistry: a molecular and
biochemical approach. Oxford University Press, New York
Novikov FN, Chilov GG (2009) Molecular docking: theoretical
background, practical applications and perspectives. Mendeleev
Commun 19:237–242
Acknowledgments Author wishes to thank Vice Chancellor, Birla
Institute of Technology, Mesra, Ranchi and Chairman, Bharat Insti-
tute of Technology, Meerut to provide necessary infrastructural and
other facilities.
Osolodkin DI, Chupakhin VI, Palyulin VA, Zefirov NS (2009)
Molecular modeling of ligand-receptor interaction in GABAC
receptor. J Mol Graph Model 27:813–821
References
Ricci L, Frosini M, Gaggelli N, Valensin G, Machetti F, Sgaragli G,
Valoti M (2006) Inhibition of rabbit brain 4-aminobutyrate
transaminase by some taurine analogues: a kinetic analysis.
Biochem Pharm 71:1510–1519
Scott EM, Jakoby WB (1959) Soluble c-aminobutyric-glutamic
transaminase from Pseudomonas fluorescens. J Biol Chem
234:932–936
Sharp K, Fine R, Honig B (1987) Computer simulations of the
diffusion of a substrate to an active site of an enzyme. Science
236:1460–1463
Silverman RB, Clift MD (2008) Synthesis and evaluation of novel
aromatic substrates and competitive inhibitors of GABA ami-
notransferase. Bioorg Med Chem Lett 18:3122–3125
Silverman RB, Durkee SC, Invergo BJ (1986) 4-Amino-2-(substituted
methyl)-2-butenoic acids: substrates and potent inhibitors of gamma-
aminobutyric acid aminotransferase. J Med Chem 29:764–770
Smith AJ, Simpson PB (2003) Methodological approaches for the
study of GABAA receptor pharmacology and functional
responses. Anal Bioanal Chem 377:843–851
Allison SA, Bacquet RJ, McCammon JA (1998) Simulation of the
diffusion-controlled reaction between superoxide and superoxide
dismutase II: detailed models. Biopolymers 27:251–269
Bansal SK, Sinha BN, Khosa RL, Olson AJ (2010) Novel GABA-AT
inhibitors: QSAR and docking based virtual screening of phenyl-
substituted b-phenyl ethylidene hydrazine analogs. Med Chem
Res 19:S114
Bansal SK, Sinha BN, Khosa RL (2011a) QSAR and docking-based
computational chemistry approach to novel GABA-AT inhibitors:
kNN-MFA based 3DQSAR model for phenyl-substituted analogs
of b-phenylethylidene hydrazine. Med Chem Res 20:549–553
Bansal SK, Sinha BN, Khosa RL (2011b) Docking based virtual
screening of schiff’s bases of GABA- a prospective to novel
GABA-AT inhibitors. Med Chem Res. doi:10.1007/s00044-011-
9843-6
Bansal SK, Sinha BN, Khosa RL, Olson AJ (2011c) Novel GABA-
AT inhibitors: QSAR and docking based virtual screening of
phenyl-substituted b-phenyl ethylidene hydrazine analogs. Med
Chem Res 20:1482–1489
Barbara M (2005) New anticonvulsant agents. Curr Top Med Chem
5:69–85
Bradford MM (1976) Rapid and sensitive method for quantitation of
microgram quantities of protein utilizing the principle of dye
binding. Anal Biochem 72:248–254
Gasteiger J, Marsili M (1980) Iterative partial equalization of orbital
electronegativity—a rapid access to atomic charges. Tetrahedron
36:3219–3228
Gerlach AC, Krajewski JL (2010) Antiepileptic drug discovery and
development: what have we learned and where are we going?
Pharmaceuticals 3:2884–2899
Goodford PJ (1985) A computational procedure for determining
energetically favorable binding sites on biologically important
macromolecules. J Med Chem 28:849–857
Goodsell DS, Olson AJ (1990) Automated docking of substrates to
proteins by simulated annealing. Proteins Struct Funct Genet
8:195–202
Hardy LW, Abraham DJ, Sapo MK (2003) Structure based drug
design. In: Abraham DJ (ed) Burger’s medicinal chemistry and
drug discovery. Wiley, New Jersey
Solis FJ, Wets RJ-B (1981) Minimization by random search
techniques. Math Oper Res 6:19–30
Sowa B, Rauw G, Davood A, Fassihi A, Knaus EE, Baker GB (2005)
Design and biological evaluation of phenyl-substituted analogs of
b-phenyl ethylidene hydrazine. Bioorg Med Chem 13:4389–4395
Storici P, Capitani G, Baise DD, Moser M, John RA, Jansonius JN,
Schirmer T (1999) Crystal structure of GABA aminotransferase a
target for antiepileptic drug therapy. Biochemistry 38:8628–8634
Storici P, Baise DD, Bossa F, Bruno S, Mozzarelli A, Peneff C,
Silverman RB, Schirmer T (2005) Structure of c-amino butyric
acid (GABA) aminotransferase a pyridoxal 50-phosphate and
[2Fe-2S] cluster-containing enzyme complexed with c-ethynyl-
GABA and with the antiepilepsy drug vigabatrin. J Biol Chem
279:363–373
Toney MD, Pascarella S, Baise DD (1995) Active site model for
c-amino butyrate aminotransferase explains substrate specificity
and inhibitor reactivities. Protein Sci 4:2366–2374
Weiner SJ, Kollman PA, Case DA, Singh UC, Ghio C, Alagona G,
Profeta S, Weiner P (1984) A new force field for molecular
mechanical simulation of nucleic acid and proteins. J Am Chem
Soc 106:765–784
123