Novel γ-aminobutyric acid ρ1 receptor antagonists; synthesis, pharmacological activity and structure-activity relationships

Article abstract of DOI:10.1021/jm7015842

γ-Aminobutyric acid (GABA) analogues based on 4-amino-cyclopent-1- enyl phosphinic acid (34-42) and 3-aminocyclobutane phosphinic acids (51, 52, 56, 57) were investigated in order to obtain selective homomeric ρ₁ GABA_C receptor antagonists. The effect of the stereochemistry and phosphinic acid substituent of these compounds on potency and selectivity within the GABA receptor subtypes was investigated. Compounds of high potency at GABA_C ρ₁ receptors (36, K_B = 0.78 μM) and selectivity greater than 100 times (41, K_B = 4.97 μM) were obtained. The data obtained was analyzed along with the known set of GABA_C ρ₁ receptor-ligands, leading to the development of a pharmacophore model for this receptor, which can be used for in silico screening.

Full text of DOI:10.1021/jm7015842

J. Med. Chem. 2008, 51, 3825–3840
3825
Novel γ-Aminobutyric Acid G1 Receptor Antagonists; Synthesis, Pharmacological Activity and
Structure-Activity Relationships
Rohan J. Kumar,† Mary Chebib,† David E. Hibbs,† Hye-Lim Kim,† Graham A. R. Johnston,‡ Noeris K. Salam,† and
Jane R. Hanrahan*,†
Faculty of Pharmacy, UniVersity of Sydney, NSW, Australia, Adrien Albert Laboratory, Department of Pharmacology,
UniVersity of Sydney, NSW, Australia
ReceiVed December 17, 2007
γ-Aminobutyric acid (GABA) analogues based on 4-amino-cyclopent-1-enyl phosphinic acid (34-42) and
3-aminocyclobutane phosphinic acids (51, 52, 56, 57) were investigated in order to obtain selective homomeric
F1 GABAC receptor antagonists. The effect of the stereochemistry and phosphinic acid substituent of these
compounds on potency and selectivity within the GABA receptor subtypes was investigated. Compounds
of high potency at GABAC F1 receptors (36, KB ) 0.78 µM) and selectivity greater than 100 times (41, KB
) 4.97 µM) were obtained. The data obtained was analyzed along with the known set of GABAC F1
receptor-ligands, leading to the development of a pharmacophore model for this receptor, which can be
used for in silico screening.
Introduction
So ubiquitous are the γ-aminobutyric acid receptors that up
to 40% of mammalian neurons use γ-aminobutyric acid (1,
GABA,a Figure 1) as a neurotransmitter.1,2 This family of receptors
is of enormous medical significance, with GABA receptors
mediating a variety of central processes, including cognition,3,4
nociception,5 vision,6 and circadian rhythms,7,8 and are involved
in the pathophysiology of epilepsy,9–12 schizophrenia,13–16 and
mood disorders.17–20 The term GABAC receptor was coined to
describe an ionotropic GABA receptor that was insensitive to
bicuculline, benzodiazepines, neurosteroids, and sedatives21–23 and
was subsequently shown to be a homomeric ligand gated ion
channel (LGIC).24 In contrast to the GABAA and GABAB receptor
Figure 1. Structures of GABA and compounds active at GABA
subunits, the GABAC receptor subunits (designated F1-3) have a
limited distribution within the central nervous system (CNS),25
reducing the risk of potential side effects. They are highly expressed
throughout the visual system, including the retina,26 superior
colliculus27 (SC), and hippocampus,28,29 leading to the recent
investigation of GABAC receptor antagonists in treating degenera-
tive eye disorders.30,31
Clinical trials involving the nonselective GABAC antagonist
3-aminopropyl(butyl)phosphinic acid (2, SGS742, Figure 1),
formerly CGP36742 (GABAB IC50 38 µM, GABAC IC50 62
µM),20 have shown it to enhance cognition and memory.19 While
the majority of this activity has been attributed to antagonism
at GABAB receptors, the magnitude of the effects observed has
led to speculation about the involvement of another mechanism
or receptor class.32 The potential role of GABAC receptors in
receptors.
memory and cognition is also supported by the dose dependent
effect observed for selective GABAC antagonist 1,2,5,6-tetrahy-
dropyridine-4-yl-methyl-phosphinic acid (3, TPMPA, Figure 1)
upon memory in a chick model of memory formation.22 These
findings highlight the need for more selective, orally active
GABAC receptor antagonists to determine the role of GABAC
receptors in cognitive processes.
The first selective GABAC receptor antagonist developed was
TPMPA (3).33 This ligand was designed based on the observed
activities of isoguvacine (4, Figure 1), which is active at GABAA
and GABAC receptors but not GABAB receptors, and 3-ami-
nopropyl(methyl)phosphinic acid (5, 3-APMPA, Figure 1),
which is active at GABAC and GABAB receptors but not
GABAA receptors. Here, a similar approach is taken, with a
substituted phosphinic acid moiety, known to reduce activity
at GABAA receptors, used as a bioisostere of the carboxylic
acid of 4-aminocyclopent-1-enecarboxylic acid (6, 4-ACPCA,
Figure1),apotentbutnonselectiveGABACreceptorantagonist.34,35
The stereochemistry at the 4-position of these compounds is an
important determinant of activity for 4-ACPCA34 and was
therefore investigated along with the effect of increasing
substituent size at the phosphinic acid, which has previously
been observed to promote increased antagonist activity at
GABAB receptors.36
* To whom correspondence should be addressed. Phone: 61293512078. Fax:
61293514391. E-mail:janeh@pharm.usyd.edu.au.
†
Faculty of Pharmacy, University of Sydney.
Adrien Albert Laboratory, Department of Pharmacology, University
‡
of Sydney.
a Abbreviations: GABA, γ-aminobutyric acid; LGIC, ligand gated ion
channel; CNS, central nervous system; TBDMSCl,tert-butyldimethylsilyl
chloride; TBAF, tetrabutylammoniumfluoride; THF, tetrahydrofuran; BOC,
tert-butyloxycarbonyl; DMF, dimethylformamide; TMS, tetramethyl silane;
EI, electron impact; CI, chemical ionization; ESI, electrospray ionization;
BSTFA, bis(trimethylsilyl)trifluoroacetamide; SAR, structure-activity re-
lationship; DABCO, 1,4-diazabicyclo[2.2.2]octane; DMAP, 4-dimethylami-
nopyridine; DIAD, diisopropylazodicarboxylate; HMDS, hexamethyldisi-
lazane; AIBN, azobisisobutyronitrile.
It has been proposed that selective activity at GABAC recep-
tors over GABAA and GABAB receptors is achieved via a more
10.1021/jm7015842 CCC: $40.75
2008 American Chemical Society
Published on Web 06/05/2008

3826 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
Scheme 1a
Scheme 2a
a Reagents and conditions: (a) SOCl2, benzene; 70 °C, 12 h; (b) (i) NaN3,
H2O, benzene; room temp, 3 h; (ii) benzene; 60 °C, 4 h; (iii) 6 M HCl,
benzene, 90 °C, 12 h; (c) BOC2O, Et3N, DMF; room temp, 4 h.
10-12 (Scheme 1), and subsequent 1,4-reduction with trapping
as the triflate as described by Paquette et al.47 provided access
to the desired intermediates (13-15, Scheme 1).
Methodology for the formation of carbon phosphorus bonds
and the application of these to the synthesis of GABA analogues
has been the subject of a number of studies.33,48–50 These have
reported the successful palladium catalyzed coupling of numer-
ous alkenyl halides to alkylphosphinates and anilinium hypo-
phosphite. Typical isolated yields for these procedures is from
30-70%. Dumond et al.50 reported the coupling of anilinium
hypophosphite to 2-trifluoromethanesulfonyloxy-1-octene with
an optimized yield of 69% using Pd(OAc)2 as the catalyst. More
recently, a report has outlined the coupling of phosphinates to
alkenyl triflates catalyzed by Pd(OAc)2 under reflux for 18 h
with the phosphinates generated in situ.51 Here, we report the
coupling of alkylphosphinate esters (RP(O)(OR′)H) to 4-tert-
butyldimethylsiloxy-1-cyclopenten-1-yl-trifluoromethane-
sulfonate (13-15, Scheme 1) with 2.5 mol% Pd(PPh3)4 as the
catalyst. The reaction proceeds smoothly for a variety of
alkylphosphinate esters (16-24, Scheme 1) with isolated yields
of 85-95% using mild conditions. While the reaction is limited
by requiring long reactions times (∼72 h), as decomposition of
the starting materials is observed upon heating, the high yield
and ease of this coupling suggests that the leaving group
properties of the triflate group make it a superior substrate to
alkenyl halides. This observation may be of synthetic utility
for C-P bond formation, as to the authors knowledge, it is the
first demonstration of direct one-step access to esterified alkyl
phosphinic acids from an alkenyl triflate and an alkylphosphinate
ester.
Following C-P bond formation, the compounds are depro-
tected using tetrabutylammoniumfluoride (TBAF) in anhydrous
tetrahydrofuran (THF), giving the alcohol (25-33, Scheme 1)
and the amine generated via a one pot Mitsunobu-Staudinger
reaction. Finally, the phosphinate ester is hydrolyzed using
aqueous HCl and the crude product purified via ion-exchange
chromatography and recrystallization to give the free amine
(34-42,Scheme 1).
Mitsunobu reactions at unhindered, primary amines such as
these have previously been observed to occur with complete
inversion of the stereocenter.52 To confirm this effect, the
Mosher amides of compounds 35 and 36 (43 and 44, respec-
tively) were prepared via the method of Smith et al.53 Chemical
shift inequivalences were observed for the P-CH3 hydrogens
via 1H NMR and from the line fitted data (see Supporting
Information) the e.e. determined to be 92-94%. As the
compounds 40 and 41 were synthesized from the same starting
materials using identical techniques, it was assumed that the
e.e.ʼs were comparable to those determined for com-
pounds 35 and 36.
a Reagents and conditions: (a) TBDMSCl, Et3N, DMAP, CHCl3; 0 °C
to room temp 24 h; 10, ((); 11, (S); 12, (R); (b) L-selectride, ET3N PhNTf2,
THF; -78 °C to room temp, 3 h; 13, ((); 14, (R); 15, (S); (c) Pd(PPh3)4,
RP(O)(OEt)H, DABCO, toluene; room temp 72 h; (d) TBAF, THF; room
temp, 12 h; (e) (i) PPh3, DIAD, HN3; 0 °C to room temp, 12 h; 40 °C, 3 h;
H2O; 40 °C, 2 h; (ii) 6 M HCl; reflux, 36 h; 16, 25, and 34, ((), R ) CH3;
17, 26, and 35, (R), R ) CH3; 18, 27, and 36, (S), R ) CH3; 19, 28, and
37, ((), R ) CH2CH3; 20, 29, and 38, ((), R ) CH(CH3)2; 21, 30, and
39, ((), R ) (CH2)3CH3; 22, 31, and 40, (R), R ) (CH2)3CH3; 23, 32, and
41, (S), R ) (CH2)3CH3; 24, 33, and 42, ((), R ) CH2(C6H5).
folded conformation of the GABA (1) backbone.37 However,
little is known about the activity of cyclobutane analogues of
GABA at GABAC receptors, with only two early reports
outlining the synthesis and GABA-ergic activity of cis- and
trans-3-aminocyclobutane carboxylic acid.38,39
Thus, a series of cyclopentene and cyclobutane compounds
were synthesized and their activity as selective F1 GABAC re-
ceptor antagonists investigated. Finally, the compounds inves-
tigated were used to develop a pharmacophore model, with the
potential to be utilized for the in silico screening of GABAC
receptor antagonists in the future.
Chemistry. 4-tert-Butyldimethylsiloxy-1-cyclopenten-1-yl-
trifluoromethanesulfonate (13, Scheme 1) was considered a key
intermediate in designing the synthetic strategy, as the excellent
electrophilic properties of the triflate moiety and its position R
to the double bond makes this compound an ideal substrate for
palladium catalyzed generation of the C-P bond. Additionally,
the protected alcohol can be transformed to the required amine
withcompleteinversionofstereochemistry,viaaMitsunobu-Stau-
dinger reaction, to yield the desired amino phosphinic acid.
To access this intermediate, initially (()-4-hydroxycyclopent-
2-en-1-one (7, Scheme 1) was synthesized from 2-methyl furan
as described by An et al.40 Several enantioselective methods
exist for the production of the individual enantiomers of
4-hydroxycyclopent-2-en-1-one (8 and 9, Scheme 1), due to its
extensive use as a synthon in prostaglandin chemistry.41 These
includeenzymatic,42chromatographic,43andsyntheticprocedures.44–46
The method of Khanapure et al.44 using L-tartaric acid as the
source of chirality was selected due to the extremely high
reported enantiomeric excess (e.e.) and the potential for mul-
tigram scale preparation. Thus (S)-4-hydroxycyclopent-2-en-1-
one (8) and (R)-4-hydroxycyclopent-2-en-1-one (9) were syn-
thesized at 99%+ e.e. as determined by optical rotation
measurement.
Routine protection of the (()-, (S)-, and (R)-4-hydroxycy-
clopent-2-ene-1-one (7, 8, and 9, respectively) using tert-
butyldimethylsilyl chloride (TBDMSCl) produced compounds
The synthesis of the cyclobutane compounds proceeded from
3-oxocyclobutanecarboxylic acid (45, Scheme 2) as the ketone
and carboxylic acid moieties could be used to generate the

γ-Aminobutyric Acid F1 Receptor Antagonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3827
Scheme 3a
The effect of stereochemistry at the 4 position was determined
in a similar manner, by determining apparent KB values for
compounds 35(R), 36(S), 40(R), and 41(S) (Figure 2 (iv)). In
both cases the (S) enantiomer was significantly more active than
the (R) enantiomer (p ) 0.0007 for 35 and 36; p ) 0.0095 for
40 and 41).
At GABAB(1A/2) receptors, compound 34 (R ) Me) was found
to be a partial agonist with an EC50 of 51.04 ( 5.6 µM and a
Imax of 56 ( 0.5%, relative to GABA (Figure 3). The effect of
stereochemistry at the 4 position was investigated by determin-
ing the dose response curves of compounds 35 (R) and 36 (S)
at GABAB(1A/2) receptors (Figure 3), the (S) enantiomer (36,
Imax ) 75.9 ( 2.0%) was found to activate a significantly greater
a Reagents and conditions: (a) HMDS, (O)PH(OEt)CH3; 90 °C, 12 h;
(b) (i) 6 M HCl; reflux, 36 h; (ii) Dowex 50 H+; (c) Et3N·3HF, CH3CN;
0 °C to room temp, 4 h; (d) (i) methyl oxalyl chloride, Et3N, CH3CN; 0 °C
to room temp, 3 h; (ii) AIBN, (Bu)3SnH, toluene; 90 °C, 5 h. 49, 51, 54,
56 ) cis-; 50, 52, 55, 57 ) trans-.
percentage of the receptor population compared to 35 (Imax
)
52.3 ( 0.1%) (p ) 0.0003) and at a lower concentration (EC50
) 52.2 ( 0.1 µM and EC50 ) 71.9 ( 0.1 µM, respectively)
than the (R) enantiomer (p < 0.0001). At these receptors, only
compounds 34-36 (R ) Me) displayed any agonist activity,
with the larger substituents on the phosphinic acid of compounds
37-42 (R ) Et, i-Pr, Bu, and Bn, respectively) abolishing all
agonist activity and only very weak antagonist activity remaining
(IC50 > 300 µM in all cases). Thus, apparent KB values were
not determined for compounds 37-42, the % inhibition of 3
µM GABA by 300 µM of these compounds are given in Table
1. Compound 37 (R ) Et) is significantly more active than
compounds 38, 39, and 42 (p ) 0.0141, 0.0106, and 0.0010,
respectively), however no significant difference in inhibition is
observed among compounds 38, 39, and 42. For compounds
40 and 41, the (S) enantiomer (41) is again observed to be
significantly more active than the (R) enantiomer (40, p )
0.0201).
At R1ꢀ2γ2L GABAA receptors, all compounds were shown
to be inactive or very weak antagonists at 600 µM, due the low
potency observed (IC50 > 300 µM in all cases), apparent KB
values were not determined and the percentage inhibition of 30
µM GABA by these compounds at 600 µM is given in Table
1. No significant difference was observed among compounds
34 and 37-39. Compound 42 (R ) Bn) was significantly more
potent than compounds 34 and 37-39 (p < 0.0042 in all cases).
The (R) enantiomers 35 and 40 (R ) Me and Bu, respectively)
were found to be inactive at 600 µM as either agonists or
antagonists.
Compounds 56 and 57 were found to be very weak
antagonists (IC50 > 100 µM) at F1 GABAC receptors, with the
cis isomer being more active than the trans isomer (Table 1).
The hydroxylated compounds 51 and 52 were found to be
inactive at F1 GABAC receptors at the concentrations tested.
None of the compounds were active at GABAB(1A/2) receptors.
Compounds 56 and 57 displayed some selectivity for GABAC
F1 receptors over GABAB receptors, where they were inactive.
Because conformationally restricted phosphinic acid analogues
of GABA are significantly weaker at GABAA receptors than at
GABAC receptors20 and given extremely low potencies observed
for the cyclobutane compounds at GABAC, the compounds were
not evaluated at GABAA.
desired amine and phosphinic acid functionalities, respectively,
by known methods. The method of Pigou and Schiesser54 was
used to generate this intermediate. After formation of the acid
chloride (46, Scheme 2), using thionyl chloride, the amine (47,
Scheme 2) was then generated via a Curtius rearrangement in
a procedure modified from Daly and Gilheany.55 The amine
was then protected as a tert-butyloxycarbonyl (BOC) amide (48,
Scheme 2) using BOC2O and triethylamine in dimethylforma-
mide (DMF).
C-P bond formation was achieved using the modified
Pudovik reaction (Scheme 3) as outlined by Hansen and
Kehler.56 The nucleophilic addition of the silyl activated ethyl
methylphosphinate57 resulted in an approximate 1:4 ratio of cis
and trans isomers (49 and 50, respectively, Scheme 3). The trans
form predominating due to the puckered nature of the cyclobu-
tane ring and resulting steric protection of one face by the BOC
group, as previously observed for nucleophilic addition to 1,3-
cyclobutanes.57 The isomers can be separated at this stage by
column chromatography and deprotected with HCl to yield the
R-hydroxy amino phosphinic acid compounds (51 and 52,
Scheme 3).
The target compounds (56 and 57, Scheme 3) were obtained
by the modified Barton deoxygenation of Hansen and Kehler,56
which proceeds with racemization of the substrate due to the
radical mechanism of the reaction. The cis and trans BOC
protected phosphinate ester isomers (54 and 55, respectively,
Scheme 3) that are formed in approximately 1:1 ratio can be
separated by column chromatography before deprotection with
HCl.
Pharmacological Results. At F1 GABAC receptors, dose-re-
sponse curves of GABA in the presence of several concentra-
tions of 34 (R ) Me) (Figure 2(i)) were determined and the
nonlinear method of Lew and Angus58 (fitted data shown in
Figure 2(ii)) used to define the interaction as competitive
antagonism. From this analysis, the KB of 34 was determined
as (pA2 ) 5.73 ( 0.09). The effect of increasing phosphinic
acid substitution was investigated at F1 GABAC receptors via
the determination of apparent KB values for compounds 37-39
and 42 (R ) Et, i-Pr, Bu, and Bn, respectively). The dose
response curves of GABA in absence and presence of a single
concentration of antagonists are shown in Figure 2(iii). A
nonsignificant decrease in activity was observed with increasing
phosphinic acid substituent size, from methyl group 34 to ethyl
group (37, p ) 0.0596) or n-butyl group (39, p ) 0.0741),
however, statistically significant reductions in activity were
observed upon substitution with an isopropyl group (38, p )
0.0365) and a benzyl group (42, p ) 0.0002).
GABAC G1 Pharmacophore. Relatively few attempts have
been made to develop computational models of the GABAC F1
receptor agonists binding pocket,59–61 and these have used
primarily homology modeling60,61 with respect to the acetyl-
choline binding protein for which a crystal structure exists.62
This approach is obviously limited by the low sequence homo-
logy between the two proteins (∼20%). A quantitative
structure-activity relationship (SAR) model of agonist binding
to the GABAC F1 receptor has proposed a scheme for the

3828 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
Figure 2. (i) Dose response curves of GABA alone (b, n ) 3) and in the presence of 3 µM (9, n ) 3), 10 µM (1, n ) 3), and 30 µM ([, n )
3) 34. (ii) Variation of Log(GABAEC50) with concentration of 34 and results of nonlinear regression. (iii) Dose response curves for GABA alone
(b, n ) 3) and in the presence of 10 µM 34 (9, n ) 3), 10 µM 37 (1, n ) 3), 10 µM 38 ([, n ) 3), 10 µM 39 (2, n ) 3), and 100 µM 42 (O,
n ) 3). (iv) Dose response curves of GABA alone (b, n ) 9) and in the presence of 30 µM 35 (0, n ) 3), 10 µM 36 (4, n ) 3), 30 µM 40 (],
n ) 3), and 10 µM 41 (3, n ) 3). Data are mean ( SEM.
by the GABAC F1 pharmacophore H-bond donor vector. This
effect is quantitatively evident through the different vector
geometry scores: 0.88 and 0.68 for 36 and 35, respectively
(Table 2), was consequently exploited during the pharmacophore
optimization. The “find match to hypothesis” criteria65 was
instructed to reject hits with vector scores less than 0.76 (note:
optimal vector score is 1.0).66 As a result, the final GABAC F1
pharmacophore identified only compound 36 and not compound
35 as a potent antagonist.
Upon screening a diverse set of GABA ligands (see Sup-
Figure 3. Dose-response curves of GABA (b, n ) 3), 34 (9, n )
porting Information for a complete list), the GABAC F1
3), 35 (0, n ) 3), and 36 (∆, n ) 3) at GABAB(1A/2) receptors. Data
pharmacophore accurately identified only potent GABAC F1
are means ( SEM.
antagonists (Table 2). Moreover, it did not identify any GABAC
agonists, such as GABA (1), or any of the potent agonists/
GABAC receptor agonist binging pocket59 but does not provide
antagonists of GABAA and GABAB. The shape constraint of
the pharmacophore was proven to be effective in excluding
various nonpotent GABAC antagonists, for example (R)-4-
ACPCA (R-6) (Figure 5), which was not identified as a hit,
despite satisfying four of the five pharmacophore sites (DNPX),
due to having excluded volume violations with spheres at the
base of the shell, indicated by the red arrow.
more than a conceptual outline with which to assess the potency
of candidate ligands. Here, as a result of the synthesis and
pharmacological evaluation of a set of new GABAC F1 receptor
antagonists, sufficient data now exists to allow the generation
of a retrospective, ligand based GABAC F1 antagonist pharma-
cophore. The generation of such a model will expediate efforts
toward the understanding of GABAC F1 receptors role in the
CNS by the identification of the primary structural features
responsible for antagonist activity and allowing in silico
screening of potential antagonists. To this end, pharmacophore
modeling was implemented using TPMPA (3), (R)-4-ACPMPA
(35), (S)-4-ACPMPA (36), (1R,3S)-3-aminocyclopentane(meth-
yl)phosphinic acid63 (60, (1R,3S)-3-ACPMPA, Table 2), (1S,3R)-
3-aminocyclopentane(methyl)phosphinic acid63 (62, (1S,3R)-3-
ACPMPA, Table 2), and imidazole-4-acetic acid64 (63, I4AA,
Table 2) as the basis set.
Discussion
It was hypothesized that the 4-aminocyclopent-1-enyl(alkyl)-
phosphinic acids (34-42) would possess the required physico-
chemical properties for optimal interaction with the GABAC
receptor agonist binding site based on the activity of (S)-4-
ACPCA (S-6).35 It has previously been observed, for confor-
mationally restricted analogues of GABA, that the use of
phosphinic acids and alkyl phosphinic acids as a bioisostere of
a carboxylic acid gives antagonists at F1 GABAC receptors,33,67–69
with activity comparable to or greater than the parent compound.
This is reflected in the pharmacological results obtained (Table
1), with compound 36 (KB ) 0.78 µM) being a competitive
antagonist with higher affinity for the GABAC agonist binding
site than (S)-4-ACPCA (S-6, KB ) 6.0 µM).35
The GABAC F1 pharmacophore (Figure 4a) represents a the-
oretical model capable of relating pharmacophore features to
antagonist activity for compound 36, discriminating it against
35. Figure 4b shows both compounds successfully matching
each of the five pharmacophore sites, however, only 36 projects
the amine H-bond donor in a similar direction to that defined

γ-Aminobutyric Acid F1 Receptor Antagonists
Table 1. Pharmacological Data
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3829
a EC50 values of agonists. Data are mean ( SEM (n ) 3-5 oocytes). b The Hill coefficient (nH). Data are mean ( SEM (n ) 3-5 oocytes). c Dissociation
d
constants of antagonists (KB) are calculated from the pA2 values in brackets. Data are mean ( SEM (n ) 3-5 oocytes). Imax is the intrinsic activity
calculated as a percentage of the maximum whole cell current produced by a maximum dose of GABA. Data are mean ( SEM (n ) 3-5 oocytes). e Percentage
inhibition by 600 µM of compound of the current produced by a submaximal dose of GABA (30 µM, EC50). Data are mean ( SEM (n ) 3-5 oocytes).
f Percentage inhibition by 300 µM of compound of the current produced by a submaximal does of GABA (3 µM, EC50). Data are mean ( SEM (n ) 3-5
oocytes).

3830 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
Table 2. GABAC Antagonists that Successfully Matched with the GABAC F1 Pharmacophorea
a Successful alignments require a minimum of four sites corresponding to exact matches of h-bond donor (D), and positive (P) and negative (N) ionizable
sites, leaving the fourth site point as either the double bond (X) or an methyl/alkyl substituent (Y). b Fitness ) 1.00 × (1.0 - align score/1.20) + 1.00 ×
vector score + 1.00 × volume score. c Indicates compound used to generate pharmacophore, 35 not shown. The fitness score is a quantitative measure of
how accurately the compound matched the pharmacophore model. Specific match criteria involved applying the excluded volume shell and rejecting hits
with vector scores < 0.76, volume scores < 0.00, or align scores >1.20.65,66
Further alkyl substitution of the phosphinic acid led to a
reduction in the activity of these compounds (37-39, 42) at F1
GABAC receptors, but interestingly the effect was not a direct
correlation of alkyl size and activity. The isopropyl substituted
compound (38, KB ) 13.09 µM) was less potent than n-butyl
substituted compound (39, KB ) 7.24 µM), possibly indicating
an unfavorable steric interaction of the branched isopropyl chain
with the binding site. The benzyl substituent (42) led to a major
reduction in activity, indicating that large planar substituents
cannot be well accommodated. Taken together, the activities

γ-Aminobutyric Acid F1 Receptor Antagonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3831
Whereas previous modeling studies have been concerned with
identifying protein structure in the area of ligand binding, this
study sought to provide a tool for assessment of potential
antagonists. Using the final pharmacophore to screen a large
set of various GABA receptor-ligands (see Supporting Infor-
mation), only potent GABAC F1 receptor antagonists (KB e 10.0
µM) matched the hypothesis, including those that were not
incorporated into the original pharmacophore design: (S)-4-
ACPCA (S-6), (piperidin-4-yl)methylphosphinic acid69 (58,
P4MPA, Table 2), (1S,3S)-3-aminocyclopentane(methyl)phos-
phinic acid63 (59, (1S,3S)-3-ACPMPA, Table 2), (1R,3R)-3-
aminocyclopentane(methyl)phosphinic acid63 (61, (1R,3R)-3-
ACPMPA, Table 2), and 3-APMPA (5). This indicates the
GABAC F1 antagonist pharmacophore has the ability to delineate
the structural features pertinent for potent GABAC F1 receptor
antagonism and thus has the potential to explain why certain
GABAC receptor congeners exhibit antagonist activity while
others do not.
It is observed for compounds 35, 36, 40, and 41 that the (S)-
enantiomer is at least an order of magnitude more potent at
GABAC F1 receptors than the corresponding (R)-enantiomer,
and this is in keeping with observations for (S)-4-ACPCA (S-
6).35 Here, the pharmacophore model effectively matches against
35 and S-6 but did not match against their (R)-enantiomers
(Table 2). This achievement is the result of two pharmacophore
attributes: the use of a vector point defining the H-bond donor
and the excluded volume shell. Initially it was found that both
(S)- and (R)-enantiomers would satisfy the necessary sites when
matched against the pharmacophore, however, as the 4 position
H-bond site was defined as a vector site, it became apparent
that the opposed stereochemistry of each enantiomer vastly
affected the directionality of the H-bond donation. Thus, their
different vector geometry scores were used as a basis for
selecting only the compounds with optimal vectors. The rationale
behind this decision equates to selecting compounds that possess
a H-bond donor with a precise directionality toward the binding
site, while those that do not are rejected.
Figure 4. (a) Compound 36 mapped against the pharmacophore
features and molecular shape constraint required for antagonism,
indicated by the activity of compounds tested against the GABAC F1
receptor. Features include, H-bond donor (D, light-blue), negative
ionizable (N, red), positive ionizable (P, dark-blue), double bond (X,
magenta), and methyl/alkyl (Y, green). The molecular shape constraint
is comprised of excluded volume spheres (yellow). For clarity, only
half of the shell is shown. (b) Pharmacophore for GABAC F1 antagonism
with overlay of 35 (red carbons) and 36 (green carbons), each satisfying
all pharmacophore sites (DNPXY). Red and green arrows follow the
projected vector path of the two amine H-bond donors for 35 and 36,
respectively.
Moreover, this distinct conformational requirement forced
some weaker GABAC receptor antagonists to adopt conforma-
tion that could easily be rejected by the excluded volume. Such
is the case for 4-ACPCA, for which the excluded shell sterically
prohibited the conformations of the (R)-enantiomer while not
affecting its more potent (S)-enantiomer. Carrying out this latter
approach is in accordance with a literature GABAC F1 receptor
agonist pharmacophore59 because both models hypothesize that
a region of the ligand binding site is surrounded by a sterically
inaccessible region.
Another critical pharmacophore feature includes the need for
molecules to have either a methyl/alkyl substituent (a proxy
for the presence of an alkyl substituted phosphinic acid) or a
double bond. This feature is observed in the known set of
competitive full GABAC antagonists, as those that do not contain
an alkyl phosphinic acid contain a double bond, for example
I4AA (63) and (S)-4-ACPCA (S-6). While the antagonist activity
of alkyl phosphinic acids at F1 GABAC receptors is well
established,68 the role of the double bond is less so. The
importance of the double bond is suggested by the fact that
TPMPA (3) is more potent than its saturated analogue P4MPA
(58)69 and also that compound 36 is more potent than its
saturated analogues (59-62). To the authors’ knowledge, no
carboxylic acid full GABAC F1 antagonists without a double
bond R or ꢀ to the carboxylic acid have been described. The
role of the double bond may be participation in a specific
Figure 5. Pharmacophore for GABAC F1 antagonism with overlay of
(S)-4-ACPCA (S-6) (blue carbons), and (R)-4-ACPCA (R-6) (orange
carbons). Red arrows indicate excluded volume violations.
may point to the existence of a narrow hydrophobic region near
theGABACbindingsite,whichhaspreviouslybeenhypothesized.67,70
Inspection of existing models of the GABAC F1 receptor agonist
binding pocket60,61 reveals basic residues (Arg-104 and Arg-
170) that are presumably the site of acid interaction. These
residues are in the vicinity of hydrophobic residues (Val-155,
Val-159, Met-156, Tyr-109) that are also proposed to form the
agonist/antagonist binding site. These residues may interact with
the phosphinic acid substituent and form its binding site.

3832 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
the oral availability and effects on memory and cognition
compound 41 is currently under investigation in animal systems.
These results and its potential as an in vivo tool for determining
the physiological roles of the GABAC receptor and will be
reported in due course.
The phosphinic amino acid cyclobutane analogues 51, 52,
56, 57 of GABA showed little activity at GABAC F1 receptors
and were inactive at GABAB receptors. Their conformationally
restricted structure and relative lack of activity GABAC receptors
made them useful in generating the pharmacophore model of
the GABAC F1 receptor. In addition, these GABA analogues
are now available for testing on a variety GABA receptors,
transporters, and enzymes.
Figure 6. Comparison of the structures of SGS742 and 41.
receptor interaction, for instance a π-π interaction, or the
alteration of the pKa or orientation of the carboxylic acid.
Optimal antagonist activity is thus believed to be the result
of the correct interaction of a number of ligand features with
the receptor. These are the correct relative orientations of the
acid group, alkyl substituent, or double bond and amine. The
high affinity of compound 36 may be due to the optimal
interaction of all features with the receptor. Thus, the incorpora-
tion of the H-bond donor vector site and both the double bond
and alkyl substituents as pharmacophore features was considered
warranted.
Conclusion
The synthesis of novel phosphinic acid analogues of GABA
is reported; this has extended the utility of palladium catalysis
in the formation of C-P bonds. The activity of these compounds
has been investigated at the three major GABA receptor
subtypes, and those based on the cyclopent-1-enyl phosphinic
acid scaffold showed excellent potency at GABAC F1 receptors.
The pharmacological data collected shows that (i) the nature of
the alkyl substituent, (ii) the stereochemistry, and (iii) existence
of a double bond R or ꢀ to the acid functionality are important
in determining optimal antagonist activity and selectivity. These
features have been incorporated into a GABAC F1 pharmacoph-
ore that can be used for in silico screening of antagonist
candidates. Compound 41 is currently under investigation in
animal models of cognition and as a pharmacological tool for
elucidating the role of GABAC F1 receptors in vivo.
At the GABAB receptor, 34 is a moderately potent partial
agonist in contrast with TPMPA (3), which is a very weak
agonist, however a direct comparison of potencies is not possible
as the EC50 of TPMPA (3) at GABAB receptors was assayed
using rat hippocampal slices71 as opposed to electrophysiological
recordings on Xenopus laeVis oocytes. The GABAB receptor is
known to have quite strict conformational requirements for
agonist binding. For analogues of GABA, substitution at the R
and γ positions leads to a significant reduction in agonist
activity, whereas substitution at the ꢀ carbon is well tolerated,
as seen in the selective agonist baclofen and the antagonist
phaclofen. In comparison to TPMPA (3), 34 has the steric load
located closer to ꢀ carbon, which may explain its increased
potency relative to TPMPA, which has more steric load around
the γ position. Phosphinic acid and methyl phosphinic acid
analogues of GABA have found to be potent agonists at GABAB
receptors,72,73 whereas higher order alkyl substitution gives
compounds that are antagonists at GABAB receptors.74,75 This
is observed for compounds 37-42, which are very weak
antagonists at GABAB receptors, indicating that any phosphinic
acid alkyl substituent larger than a methyl group can effectively
abolish GABAB activity for conformationally restricted ligands.
At R1ꢀ2γ2L containing GABAA receptors, all compounds
displayed only very weak antagonist activity, which is consistent
with the reported poor affinity of the GABAA receptor for
phosphinic acids.67,72,74 Interestingly, 42 (R ) Bn) displays
greater antagonist activity at GABAA receptors than any of the
other compounds. This is in keeping with the pharmacophore
model based on alkyl and aryl substituted 5-(4-piperidyl)isox-
azol-3-ol compounds,76 which includes a large hydrophobic
pocket. Thus the affinity of 42 for the GABAA receptor may be
enhanced by interaction with this region of the receptor.
With the exception of 42, all cyclopent-1enyl phosphinic acid
compounds are potent and selective F1 GABAC receptors. While
the selectivity of 34-36 is only moderate, the higher order alkyl
substituents result in good selectivity for F1 GABAC receptors
over GABAB(1A/2) and R1ꢀ2γ2L GABAA receptors, with com-
parable potency and selectivity to TPMPA (3).
Experimental Section
All chemicals used were purchased from Aldrich Chemical Co.
Ltd. (St Luois, MO) unless otherwise stated and were of highest
commercially available purity. Solvents were distilled by standard
techniques prior to use. Where stated, reactions were performed
under inert atmosphere. 1H NMR spectra were recorded at 300 MHz
using a Varian (Palo Alto, CA) Gemini 300 spectrometer. Chemical
shifts (δ) are quoted in parts per million (ppm), referenced externally
to tetramethyl silane at 0 ppm. 13C NMR were recorded at 75 MHz
using a Varian (USA) Gemini 300 spectrometer. Chemical shifts
(δ) are quoted in ppm, referenced internally to CDCl3 at 77.0 ppm.
31P NMR were recorded at 162 MHz using a Bruker (Billerica,
MA) DPX300 spectrometer. Chemical shifts (δ) are quoted in ppm,
referenced externally to neat trimethyl phosphite at 140.85 ppm.
All coupling constants (J) are given in Hertz. Unless otherwise
noted, the duplication of peaks in the NMR spectra arise from the
presence of diastereomers associated with the chirality of the
phosphorus atom in the phosphinate ester. Low resolution electron
impact (EI) and chemical ionization (CI) MS was carried out on a
ThermoFinnigan (Waltham, MA) PolarisQ Ion Trap system using
a direct exposure probe. CI MS was performed using methane as
the reagent gas. High resolution EI MS was carried out on a Kratos
(Chestnut Ridge, NY) MS25 RFA instrument at 70 eV in magnetic
scan with perfluorokerosene as standard. Low and high resolution
electrospray ionization (ESI) MS was carried out using a Bruker
(USA) Daltronics BioApexII with a 7T superconducting magnet
and an analytical ESI source. Elemental analyses were performed
at the Research School of Chemistry, Australian National Univer-
sity, Canberra, ACT, Australia. Thin layer chromatography was
performed on Merck aluminum backed plates, precoated with silica
(0.2 mm, 60F254), which were developed using one of the following
techniques: UV fluorescence (254 nm), alkaline potassium per-
manganate solution (0.5% w/v), or ninhydrin (0.2% w/v). Flash
chromatography was performed on silica gel (Merck silica gel 60H,
particle size 5-40 µm). Melting points were determined using a
Stuart (Stone, Staffordshire, UK) SMP10 melting point apparatus.
The butyl phosphinic acid analogue of GABA, SGS742
exhibits cognitive enhancing effects as mentioned previously
and is also orally active.77 The close structural similarity between
SGS742 and 41 (Figure 6) but greater GABAC F1 selectivity of
41 suggests it offers a means of establishing the basis of the
cognition enhancement. TPMPA (3) is currently the best tool
for elucidating the role of GABAC receptors in vivo but suffers
from its lack of activity following systemic administration, thus

γ-Aminobutyric Acid F1 Receptor Antagonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3833
(()-4-Hydroxycyclopent-2-en-1-one (7). 7 was synthesized via
the method of An et al.40 1H NMR (CDCl3) δ: 7.63 (dd, 1H, J )
2.4, J ) 5.6), 6.21 (dd, 1H, J ) 1.3, J ) 5.7), 4.25 (s, 1H),
5.02-5.05 (m, 1H), 2.76 (dd, 1H, J ) 6.4, J ) 18.5), 2.27 (dd,
1H, J ) 2.1, J ) 18.7). 13C NMR (CDCl3) δ: 207.8, 164.4, 134.5,
70.0, 44.1. MS (CI) m/z (%): 99.0 ([M + H]+, 100), 81.0 (75),
71.1 (68), 53.0 (64).
(-)-(S)-4-Hydroxycyclopent-2-en-1-one (8). 8 was synthesized
via the method of Khanapure et al.44 [R]D ) -89.7 (c ) 1.2,
CH3OH, lit.44 ) -90). NMR spectra identical to 7. MS (CI) m/z
(%): 99.0 ([M + H]+, 100), 81.0 (80), 71.2 (72), 52.9 (75).
(+)-(R)-4-Hydroxycyclopent-2-en-1-one (9). 9 was synthesized
via the method of Khanapure et al.44 [R]D ) +88.31 (c ) 1.1,
CH3OH, lit.44 ) +90). NMR spectra identical to 7. MS (CI) m/z
(%): 99.0 ([M + H]+, 100).
(-)-(S)-4-tert-Butyldimethylsiloxy-1-cyclopenten-1-yl-trifluo-
romethanesulfonate (15). Method as for 13 (89%). [R]D ) -1.92
(c ) 1.3, CH3OH, lit47 ) -1.69). NMR spectra identical to 13.
MS (CI) m/z (%): 346.7 ([M + H]+, 34), 331.0 (41), 288.9 (40),
214.9 (100), 151.1 (21).
Ethyl (()-4-(tert-butyldimethylsiloxy)cyclopent-1-enyl(meth-
yl)phosphinate (16). To a solution of 1,4-diazabicyclo[2.2.2]octane
(DABCO, 200 mg, 2.0 mmol), ethyl methylphosphinate (94 mg,
1.2 mmol) and 13 (200 mg, 0.58 mmol) in toluene (10 mL) was
added Pd(PPh3)4 (17 mg, 2.5 mol%). The solution was stirred at
room temp for 18 h, at which time a second portion of Pd(PPh3)4
(10 mg, 2.5 mol%) was added. The solution was stirred for a further
24 h, the toluene removed in vacuo, and the residue purified via
flash chromatography on silica gel (ethyl acetate:ethanol, 90:10)
to give 16 as a colorless oil (143 mg, 0.43 mmol, 78%). 1H NMR
(CDCl3) δ: 6.62 (br dd, 1H, C(2)H, J ) 10.8, J ) 17.8), 4.63 (m,
1H, C(4)H), 4.10 (m, 2H, POCH2CH3), 2.72 (m, 2H, C(5)H2), 2.45
(m, 2H, C(3)H2), 1.49 (d, 3H, PCH3, JPH ) 17.5), 1.34 (t, 3H,
POCH2CH3, J ) 7.1), 0.88 (s, 9H, SiC(CH3)3), 0.07 (s, 6H,
(()-4-(tert-Butyldimethylsiloxy)cyclopenten-1-one (10). To a
stirred solution of 7 (1.7 g, 16.9 mmol), dimethylaminopyridine
(DMAP, 206 mg, 10 mol%) in anhydrous CH2Cl2 (33 mL) at 0 °C
under an atmosphere of N2, a solution of TBDMS (3.1 g, 20.7
mmol) in anhydrous CH2Cl2 (11 mL) was added dropwise over 10
min. The solution was allowed to warm to room temp and stirred
for 3 h, after which time deionized H2O (100 mL) was added. The
organic layer was separated, the aqueous phase extracted with
CH2Cl2 (3 × 50 mL), and the combined organic fractions dried
with anhydrous Mg2SO4 and concentrated in vacuo. The resulting
oil was filtered through a short column of silica gel eluted with
10% ethyl acetate in hexane. The fractions containing the product
were concentrated in vacuo and purified via vacuum distillation
(bp 81 °C, 0.05 mm, lit.78 60 °C, 0.01 mm) to give 10 as a colorless
oil that crystallized upon standing (3.05 g, 14.4 mmol, 85%, mp
2
Si(CH3)2). 13C NMR (CDCl3) δ: 145.44 and 144.83 (C(2), JPCC
2
) 11.5, JPCC ) 10.63) 134.28 (d, C(1), JPC ) 124.3), 72.99 and
72.94 (C(4), 3JPCCC ) 11.6, 3JPCCC ) 11.4), 61.77 (d, POCH2CH3,
2JPOC ) 6.3), 44.48 and 44.43 (C(5), 2JPCC ) 16.3, 2JPCC ) 15.7),
3
43.33 (d, C(3), JPCCC ) 12.1), 25.92 (SiC(CH3)3), 18.25 (SiC
CH3)3), 16.73 and 16.62 (POCH2CH3, 3JPOCC ) 6.2, 3JPOCC ) 6.7),
11.40 (d, PCH3, JPC ) 144.4), -4.55 (Si (CH3)2). 31P NMR (CDCl3)
δ: 38.5, 38.2. MS (CI) m/z (%): 305.1 ([M + H]+, 100), 289.2
(30), 261.2 (14), 247.1 (24), 173.1 (20), 145.1 (20). HRMS (EI)
m/z (%): calcd for C14H29O3PSi (M+·) 304.1624; found 304.1631.
Eth(y-l )-(R)-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(meth-
yl)phosphinate (17). Method as for 16; colorless oil (79%). [R]D
) -5.40 (c ) 1.17, CH3OH). NMR spectra identical to 16. MS
(CI) m/z (%): 305.1 ([M + H]+, 100), 289.1 (30), 261.1 (16), 247.0
(28), 195.1 (34), 173.0 (24), 145.1 (22).
1
26-28 °C lit.79 27-28 °C). H NMR (CDCl3) δ: 7.33 (dd, 1H, J
) 2.3, J ) 5.7), 6.06 (dd, 1H, J ) 1.3, J ) 5.6), 4.86 (m, 1H),
2.59 (dd, 1H, J ) 5.9, J ) 18.2), 2.12 (dd, 1H, J ) 2.3, J ) 18.2),
0.79 (s, 9H), 0.01 (s, 6H). 13C NMR (CDCl3) δ: 164.07, 135.70,
71.32, 60.65, 45.23, 25.89, 18.44, -4.43. MS (CI) m/z (%): 213.0
([M + H]+, 100), 197.1 (32).
Ethyl (+)-(S)-4-(tert-butyldimethylsiloxy)cyclopent-1-enyl(meth-
yl)phosphinate (18). Method as for 16; colorless oil (96%). [R]D
) +5.21 (c ) 0.9, CH3OH). NMR spectra identical to 16. MS
(CI) m/z (%): 305.0 ([M + H]+, 100), 289.1 (28), 261.1 (12), 247.0
(22), 173.0 (22), 145.0 (22).
(-)-(S)-4-tert-Butyldimethylsiloxy-2-cyclopenten-1-one (11).
Method as for 10 (89%, mp 26-28 °C, lit.78 27-28 °C). [R]D
)
-64.7 (c ) 1.3, CH3OH, lit.78 ) -65.3). NMR spectra identical
to 10. MS (CI) m/z (%): 212.9 ([M + H]+, 100), 197.1 (34).
(+)-(R)-4-tert-Butyldimethylsiloxy-2-cyclopenten-1-one (12).
Ethyl (()-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(eth-
yl)phosphinate (19). Method as for 16, using ethyl ethylphosphi-
nate; colorless oil (76%). 1H NMR (CDCl3) δ: 6.65 (m, 1H, C(2)H),
4.62 (m, 1H, C(4)H), 4.11 (m, 2H, POCH2CH3), 2.74 (m, 2H,
C(5)H2), 2.48 (m, 2H, C(3)H2), 1.76 (m, 2H, PCH2CH3), 1.23 (m,
6H, PCH2CH3 and POCH2CH3), 0.89 (s, 9H), 0.08 (s, 6H). 13C
Method as for 10 (87%, mp 26-28 °C, lit.79 27-28 °C). [R]D
)
+64.1 (c ) 0.5, CH3OH, lit.79 ) +65.3). NMR spectra identical
to 10. MS (CI) m/z (%): 213.0 ([M + H]+, 100), 197.1 (32).
(()-4-tert-Butyldimethylsiloxy-1-cyclopenten-1-yl-trifluo-
romethanesulfonate (13). To a solution of L-selectride (1 M in
THF, 4.7 mL, 4.7 mmol) in anhydrous THF (35 mL) at -78 °C
under an atmosphere of N2 was added dropwise a solution of 10
(1.0 g, 4.7 mmol) and Et3N (0.2 mL) in THF (15 mL) over 30
min. After an additional 30 min, N-phenyltrifluoromethanesulfo-
nimide (1.5 g, 4.1 mmol) was added in two portions. The resulting
solution was gradually warmed to room temp overnight, the solvent
removed in vacuo, and the residue partitioned between saturated
NaHCO3 solution (50 mL) and hexanes (100 mL) overnight. The
aqueous layer was extracted with hexane (2 × 25 mL), and the
combined organic fractions were washed with brine, dried with
anhydrous Mg2SO4, and concentrated in vacuo. The residue was
then purified via flash chromatography on silica gel using (hexane:
ethyl acetate:Et3N, 98:1:1) to give 13 as a colorless oil (1.59 g, 4.6
2
NMR (CDCl3) δ: 145.88 (apparent t, C(2), JPCC ) 10.7), 133.07
(d, C(1), JPC ) 123.6), 73.05 and 73.04 (C(4), 3JPCCC ) 10.9, 3JPCCC
) 10.9), 61.71 (d, POCH2CH3, 2JPOC ) 6.5), 44.55 and 44.47 (C(5),
2JPCC ) 15.3), 43.83 and 43.67 (C(5), 3JPCCC ) 9.2, 3JPCCC ) 8.9),
25.91 (SiC(CH3)3), 21.47 (d, PCH2CH3, JPC ) 102.6), 16.71 (t,
POCH2CH3, J ) 6.1), 6.79 (d, PCH2CH3, J ) 6.8). 31P NMR
(CDCl3) δ: 43.1, 42.9. MS (CI) m/z (%): 319.1 ([M + H]+, 100),
303.2 (26), 261.1 (24), 187.1 (20), 159.1 (17). HRMS (EI): calcd
for C15H31O3PSi (M+·) 318.1780; found 318.1789.
Ethyl (()-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(isopro-
pyl)phosphinate (20). Method as for 16, using ethyl isopropy-
lphosphinate; colorless needles (mp 28-31 °C, 90%). 1H NMR
(CD3OD) δ: 6.71 (bd, 1H, C(2)H, J ) 9.9), 4.69 (m, 1H, C(4)H),
4.08 (m, 2H, POCH2CH3), 2.81 (m, 2H, C(5)H2), 2.48 (m, 2H,
C(3)H2), 2.01 (m, 1H, PCH(CH3)2), 1.34 (t, 3H, POCH2CH3, J )
7.1), 1.20 (dd, 6H, PCH(CH3)2, J ) 7.2, J ) 18.7), 0.92 (d, 9H,
SiC(CH3)3, J ) 0.64), 0.12 (s, 6H, Si(CH3)2). 13C NMR (CD3OD)
δ: 147.57 (apparent t, C(2), J ) 9.6), 131.55 and 135.41 (C(1), JPC
1
mmol, 98%). H NMR (CDCl3) δ: 5.58 (br t, 1H, J ) 2.2), 4.60
(hept, 1H, J ) 3.8), 2.86 (dd, 1H J ) 16.3, J ) 7.3), 2.73 (dd, 1H,
J ) 16.3, J ) 7.3), 2.56 (br d, 1H, J ) 16.3), 2.37 (br d, 1H, J )
16.3), 0.88 (s, 9H), 0.06 (s, 6H). 13C NMR (CDCl3): 146.66, 118.78
(q, JCF ) 320.8), 115.59, 69.81, 41.46, 39.29, 25.97, 18.26, -4.63.
MS (CI) m/z (%): 346.8 ([M + H]+, 34), 330.9 (52), 288.9 (54),
214.9 (100), 151.1 (22).
3
) 121.0, JPC ) 121.1), 72.88 (d, C(4), JPCCC ) 10.8), 60.91 and
2
2
60.87 (POCH2CH3, JPOC ) 7.0, JPOC ) 7.0), 44.41 and 44.34
(C(5), 2JPCC ) 15.2, 2JPCC ) 15.3), 44.04 and 43.90 (C(3), 3JPCCC
) 11.3, 3JPCCC ) 11.1), 26.81 and 26.64 (PCH (CH3)2, JPC ) 103.2,
JPC ) 102.9), 25.11 (SiC(CH3)2), 17.67 (SiC(CH3)2), 15.68 (d,
POCH2CH3, 3JPOCC ) 6.0), 13.77 and 13.72 (d, PCH (CH3)2, 2JPCC
(+)-(R)-4-tert-Butyldimethylsiloxy-1-cyclopenten-1-yl-trifluo-
romethanesulfonate (14). Method as for 13 (78%). [R]D ) +1.97
(c ) 1.01, CH3OH, lit47 ) +1.69). NMR spectra identical to 13.
MS (CI) m/z (%): 346.7 ([M + H]+, 33), 331.0 (40), 288.9 (40),
214.9 (100), 151.1 (21).
2
) 6.1, JPCC ) 6.2), -5.82 (Si (CH3)2). 31P NMR (CD3OD) δ:
49.0, 48.9. MS (CI) m/z (%): 333.1 ([M + H]+, 100), 317.2 (20),

3834 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
275.1 (24), 201.0 (14). HRMS (EI): calcd for C16H33O3PSi ([M -
Ethyl (-)-(R)-4-Hydroxy-1-cyclopent-1-enyl(methyl)phosphi-
nate (26). Method as for 25; light-yellow oil (91%). [R]D ) -8.9
(c ) 0.53, CH3OH). NMR spectra identical to 25. MS (EI) m/z
(%): 190.1 (M+·, 100).
Ethyl (+)-(S)-4-Hydroxy-1-cyclopent-1-enyl(methyl)phosphi-
nate (27). Method as for 25; light-yellow oil (89%). [R]D ) +8.7
(c ) 0.72, MeOH). NMR spectra identical to 25. MS (EI) m/z (%):
190.1 (M+·, 100).
C(CH3)3]+·) 275.1232; found 275.1228.
Ethyl (()-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(bu-
tyl)phosphinate (21). Method as for 16, using ethyl butylphos-
phinate; colorless needles (94%, mp 31-34 °C). 1H NMR (CD3OD)
δ: 6.69 (bd, 1H, C(2)H, J ) 10.7), 4.69 (m, 1H, C(4)H), 4.06 (m,
2H, POCH2CH3), 2.82 (m, 2H, C(5)H2), 2.47 (m, 2H, C(3)H2), 1.82
(m, 2H, PCH2(CH2)2CH3), 1.54 (m, 4H, PCH2(CH2)2CH3), 1.36 (t,
3H, POCH2CH3, J ) 3.0), 1.29 (t, 3H, P(CH2)3CH3, J ) 6.7), 0.92
(d, 9H, SiC(CH3)3, J ) 1.2), 0.12 (s, 6H, Si(CH3)2). 13C NMR
Ethyl (()-4-Hydroxy-1-cyclopent-1-enyl(ethyl)phosphinate
(38). Method as for 25; light-yellow oil (quant). 1H NMR (CDCl3)
δ: 6.62 (bd, 1H, C(2)H, J ) 10.0), 4.64 (t, 1H, C(4)H, J ) 5.5),
4.02 (m, POCH2CH3, 2H), 2.86 (s, 1H, OH), 2.78 (m, C(5)H2, 2H),
2.53 (m, C(3)H2, 2H), 1.78 (m, PCH2CH3, 2H), 1.32 (dt, 3H,
POCH2CH3 J ) 2.0, J ) 7.1), 1.13 (ddd, 3H, PCH2CH3, J ) 2.2,
J ) 7.7, J ) 18.6). 13C NMR (CDCl3) δ: 145.55 and 145.46 (C(2),
2JPCC ) 10.3, 2JPCC ) 10.2), 133.32 (d, C(1), JPC ) 124.2), 72.00
2
(CD3OD) δ: 146.60 (apparent t, C(2), JPCC ) 11.8), 132.54 (d,
C(1), JPC ) 124.8), 72.82 (d, C(4), 3JPCCC ) 10.7), 60.79 and 60.74
2
2
(POCH2CH3, JPOC ) 6.5, JPOC ) 6.6), 44.46 and 44.41 (C(5),
2JPCC ) 15.8, 2JPCC ) 15.8), 43.52 and 43.49 (C(3), 3JPCCC ) 12.1,
3JPCCC ) 11.8), 27.17 and 26.92 (PCH2(CH2)2CH3, JPC ) 101.4,
JPC ) 100.8), 25.13 (SiC(CH3)2), 23.59 (d, P(CH2)2CH2CH3, 3JPCCC
3
2
2
(d, C(4), JPCCC ) 10.4), 60.64 and 60.61 (POCH2CH3, JPOC
)
) 23.9), 23.52 (d, PCH2CH2CH2CH3, JPCC ) 30.3), 17.68 (s,
2
2
SiC(CH3)2), 15.67 and 15.59 (POCH2CH3, 3JPOCC ) 9.0, 3JPOCC
)
6.4, JPOC ) 6.4), 44.39 (d, C(5), JPCC ) 15.6), 43.62 and 43.57
(C(3), 3JPCCC ) 11.6, 3JPCCC ) 11.7), 21.40 and 21.37 (PCH2CH3,
8.7), 12.76 (P(CH2)3CH3), -5.84 (Si(CH3)2). 31P NMR (CDCl3)
δ: 41.6, 41.4. MS (CI) m/z (%): 347.1 ([M + H]+, 100), 331.3
3
JPC ) 101.7, JPC ) 101.4), 16.78 (d, POCH2CH3, JPOCC ) 6.3),
5.85 (d, PCH2CH3, 2JPCC ) 4.8). 31P NMR (CDCl3) δ: 41.6, 41.8.
MS (EI) m/z (%): 204.1 (M+·, 4), 187.1 (16), 175.1(60), 147.1(100),
65.1 (38). HRMS (EI): calcd for C9H17O3P (M+·) 204.0915; found
204.0913.
(21), 289.2 (18). HRMS (EI): calcd for C17H35O3PSi (M+·
346.2093; found 346.2096.
)
Ethyl (-)-(R)-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(bu-
tyl)phosphinate (22). Method as for 21; colorless needles (99%,
mp 30-32 °C). [R]D ) -2.38 (c ) 1.0, MeOH).NMR spectra
identical to 21. MS (CI) m/z (%): 347.1 ([M + H]+, 100), 331.2
(25), 289.1 (24).
Ethyl (()-4-hydroxy-1-cyclopent-1-enyl(isopropyl)phosphi-
nate (29). Method as for 25; light-yellow oil (98%). 1H NMR
(CDCl3) δ: 6.64 (bd, 1H, C(2)H, J ) 9.2), 4.65 (m, C(4)H, 1H),
4.05 (m, POCH2CH3, 2H), 2.80 (m, C(5)H2, 2H), 2.60 (s, 1H), 2.53
(m, C(3)H2, 2H), 1.93 (m, PCH(CH3)2, 1H), 1.33 (dt, 3H,
POCH2CH3, J ) 1.9, J ) 7.0), 1.21 (d, 3H, PCH(CH3)2, J ) 7.2),
1.16 (d, 3H, PCH(CH3)2, J ) 7.1). 13C NMR (CDCl3) δ: 146.08
(apparent t, C(2), 2JPCC ) 9.4), 132.72 (d, C(1), JPC ) 119.9), 72.21
(d, C(4), 3JPCCC ) 9.9), 60.73 and 60.71 (POCH2CH3, 2JPOC ) 6.8,
Ethyl (+)-(S)-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(bu-
tyl)phosphinate (23). Method as for 21; colorless needles (92%,
mp 30-32 °C). [R]D ) +2.22 (c ) 2.5, CH2Cl2). NMR spectra
identical to 21. MS (CI) m/z (%): 347.1 ([M + H]+, 100), 331.2
(22), 289.1 (24), 215.1 (12).
2
Ethyl (()-4-(tert-Butyldimethylsiloxy)cyclopent-1-enyl(ben-
zyl)phosphinate (24). Method as for 16, using ethyl benzylphos-
phinate; colorless oil. 1H NMR (CDCl3) δ: 7.12-7.51 (m, 5H,
PCH2(C6H5)), 6.55 (bd, 1H, C(2)H, JPH ) 18.0), 4.65 (m, 1H,
C(4)H), 3.95 (m, 2H, POCH2CH3), 3.23 (dd, 2H, PCH2(C6H5), J
) 3.2, JPH ) 18.0), 2.76 (m, 2H, C(5)H2), 2.45 (m, 2H, C(3)H2),
1.29 (t, 3H, POCH2CH3, J ) 7.0), 0.90 (s, 9H, SiC(CH3)3), 0.09
(s, Si(CH3)2, 6H). 13C NMR (CDCl3) δ: 146.00 (apparent t, C(2),
J ) 6.8), 44.37 (d, C(5), JPCC ) 15.2), 44.23 and 44.14 (C(3),
3JPCCC ) 11.0, J ) 11.1), 27.46 (d, POCH(CH3)2, JPC ) 102.3),
3
16.80 (d, POCH2CH3, JPOCC ) 6.3), 15.40 and 15.37 (2JPCC
)
25.7, 2JPCC ) 28.0). 31P NMR (CDCl3) δ: 46.1, 45.8. MS (EI) m/z
(%): 218.1 (M+·, 6), 203.1 (14), 201.1 (32), 189.1 (75), 175.1 (48),
161.1 (78), 147.1 (100), 129.0 (74), 101.1 (56), 65.1 (90). HRMS
(EI): calcd for C10H19O3P (M+·) 218.1072; found 218.1076.
Ethyl (()-4-Hydroxy-1-cyclopent-1-enyl(butyl)phosphinate
(30). Method as for 25; light-yellow oil (99%). 1H NMR (CDCl3)
δ :6.62 (bd, 1H, C(2)H, J ) 10.1), 4.64 (m, C(4)H, 1H), 4.08 (m,
POCH2CH3, 2H), 2.81 (m, C(5)H2, 2H), 2.53 (m, C(5)H2, 2H), 1.75
(m, PCH2(CH2)2CH3, 2H), 1.56 (m, PCH2CH2CH2CH3, 2H), 1.41
(hept, 2H, P(CH2)2CH2CH3, J ) 7.3), 1.32 (dt, 3H, POCH2CH3, J
) 2.0, J ) 7.0), 0.92 (t, 1H, P(CH2)3CH3, J ) 7.2). 13C NMR
(CDCl3) δ: 145.31 and 145.12 (C(2), 2JPCC ) 10.7, 2JPCC ) 10.8),
133.76 (d, C(1), JPC ) 125.6), 72.01 (d, C(4), 3JPCCC ) 10.5), 60.56
(d, POCH2CH3, 2JPOC ) 6.3), 44.38 (d, C(5), 2JPCC ) 15.6), 43.61
2
2JPCC ) 9.9), 132.37 (d, benzene C(1), JPCC ) 10.9), 131.04 (d,
3
C(1), JPC ) 126.96), 130.02 (d, benzene ortho, JPCCC ) 6.6),
4
128.76 (d, benzene meta, JPCCCC ) 3.1), 127.1 (d, benzene para,
3
5JPCCCCC ) 3.7), 72.17 (d, C(4), JPCCC ) 10.5), 62.21 and 62.15
2
2
(POCH2CH3, JPOC ) 8.2, JPOC ) 8.5), 44.37 and 44.33 (C(5),
2JPCC ) 16.3, 2JPCC ) 16.2), 43.85 (d, C(3), 3JPCCC ) 11.0), 34.01
(d, PCH2(C6H5), JPC ) 138.2), 25.92 (SiC(CH3)3), -3.32 (Si(CH3)2.
31P NMR (CDCl3) δ: 37.41. MS (CI) m/z (%): 281.1 ([M + H]+,
100), 265.2 (14), 323.1 (22), 249.1 (12), 221.1 (16). HRMS (EI):
calcd for C20H33O3PSi (M+·) 380.1937; found 380.1942.
3
3
and 43.55 (C(3), JPCCC ) 11.6, JPCCC ) 11.7), 28.14 (d, PCH2
2
(CH2)2CH3, JPC ) 100.3), 23.99 (d, PCH2CH2CH2CH3, JPCC
)
Ethyl (()-4-Hydroxy-1-cyclopent-1-enyl(methyl)phosphinate
(25). To a solution of 16 (350 mg, 1.2 mmol) under an atmosphere
of N2 was added TBAF (1 M in THF, 1.4 mL). The solution was
stirred at room temp for 12 h and the solvent was removed in vacuo.
The residue was purified by flash chromatography on silica gel
(ethyl acetate:ethanol, 85:15) to give 25 as a light-yellow oil (208
3
16.3), 23.86 (d, P(CH2)2CH2CH3, JPCCC ) 3.7), 16.79 (d,
POCH2CH3, 2JPOC ) 6.3), 13.83 (P(CH2)3CH3). 31P NMR (CDCl3)
δ: 42.9, 41.9. MS (EI) m/z (%): 232.1 (M+·, 5), 217.1 (79), 203.1
(50), 175.1 (92), 146.1 (100), 65.1 (60). HRMS (EI): calcd for
C11H21O3P (M+·) 232.1228; found 232.1233.
1
Ethyl (-)-(R)-4-Hydroxy-1-cyclopent-1-enyl(butyl)phosphi-
nate (31). Method as for 25; light-yellow oil (92%). [R]D ) -7.0
(c ) 0.52, CH3OH). NMR spectra identical to 30.
Ethyl (+)-(S)-4-Hydroxy-1-cyclopent-1-enyl(butyl)phosphi-
nate (32). Method as for 25; light-yellow oil (95%). [R]D ) +6.9
(c ) 0.71, CH3OH). NMR spectra identical to 30.
mg, 1.1 mmol, 98%). H NMR (CDCl3) δ: 6.62 (bd, 1H, C(2)H,
JPH ) 10.6), 4.65 (m, C(4)H, 1H), 4.02 (m, POCH2CH3, 2H), 3.04
(m, 1H), 2.82 (d, 1H, C(5)H2, J ) 16.5), 2.55 (d, C(3)H2, 1H, J )
18.0), 1.54 (dd, 3H, PCH3, J ) 4.3, J ) 14.5), 1.33 (dt, POCH2CH3,
3H, J ) 2.0, J ) 7.1). 13C NMR (CDCl3) δ: 145.63 and 145.51
(C(2), 2JPCC ) 11.4, 2JPCC ) 11.8), 133.50 (d, C(1), JPC ) 131.6),
3
71.27 (d, C(4), JPCCC ) 11.2), 60.87 and 60.85 (dd, POCH2CH3,
Ethyl (()-4-Hydroxy-1-cyclopent-1-enyl(benzyl)phosphinate
(33). Method as for 25; light-yellow oil (79%). 1H NMR (CDCl3)
δ: 7.38-7.22 (m, 5H, PCH2(C6H5)), 6.53 (bd, 1H, C(2)H, J ) 10.3),
4.56-4.49 (m, C(4)H, 1H), 4.17-3.92 (m, POCH2CH3, 2H), 3.22
(dd, 2H, PCH2(C6H5), J ) 3.0, J ) 17.9), 2.90-2.28 (m, C(5)H2
2
2
2JPOC ) 6.3, JPOC ) 6.3), 43.65 (d, C(5), JPCC ) 16.4), 42.53
3
3
and 42.45 (C(3), JPCCC ) 12.3, JPCCC ) 12.1), 15.64 (d,
3
POCH2CH3, JPOCC ) 6.6), 12.74 and 12.57 (PCH3, JPC ) 102.5,
J ) 102.2). 31P NMR (CD3OD) δ: 42.9, 42.7. MS (EI) m/z (%):
190.1 (M+·, 9), 175.1 (14), 161.1 (96), 147.1 (41), 133.1 (100),
55.1 (37). HRMS (EI): calcd for C8H15O3P (M+·) 190.0759; found
190.0762.
and C(3)H2, 4H), 1.31 (dt, 3H, POCH2CH3, J ) 1.1, J ) 7.0). 13
C
)
NMR (CDCl3) δ: 146.40 and 146.37 (C(2), 2JPCC ) 10.7, 2JPCC
10.3), 133.01 (d, C(1), JPC ) 123.3), 131.53 (d, benzene C(1), 2JPCC

γ-Aminobutyric Acid F1 Receptor Antagonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3835
) 7.4), 130.16 and 130.13 (d, benzene ortho, 3JPCCC ) 5.8, 3JPCCC
) 5.9), 129.79 and 129.78 (d, benzene meta, 4JPCCCC ) 2.7, 4JPCCCC
J ) 2.3, J ) 27.0). 31P NMR (D2O) δ: 35.7. MS (EI) m/z (%):
190.1 (M+·, 5), 172.0 (100), 130.0 (65), 80.1 (60), 68.1 (45), 66.1
(40). HRMS (EI): calcd for C8H16NO2P (M+·) 189.0919; found
189.0920. Anal. (C8H16NO2P·0.2H2O) C, H, N.
5
) 2.7), 127.20 (apparent t, benzene para, JPCCCCC ) 2.8), 72.08
3
2
(d, C(4), JPCC ) 10.6), 61.26 and 61.25 (POCH2, JPOC ) 6.5,
2JPOC ) 6.5), 44.37 and 44.34 (C(5), 2JPCC ) 16.1, 2JPCC ) 16.0),
(()-4-Amino-1-cyclopent-1-enyl(butyl)phosphinic acid (39).
3
1
43.78 (d, C(3) JPCCC ) 11.2), 37.06 and 37.03 (PCH2(C6H5), JPC
Method as for 34; white solid (65%, mp (dec.) 192-195 °C). H
) 95.6, 2JPC ) 95.4), 16.75 (d, POCH2CH3, 2JPOC ) 6.2). 31P NMR
(CDCl3): 37.5, 37.4. MS (EI) m/z (%): 266.0 (M+·), 222.0 (59),
191.0 (22), 147.0 (97), 129.0 (66), 101.0 (47), 91.1 (70), 65.1 (100).
HRMS (EI): calcd for C14H19O3P (M+·) 266.1072; found 266.1074.
(()-4-Amino-1-cyclopent-1-enyl(methyl)phosphinic acid (34).
To a solution of 25 (150 mg, 0.85 mmol), diisopropylazodicar-
boxylate (DIAD, 0.7 mL, 1.9 mmol) and HN3 (1.9 M in benzene,
1.1 mL, 1.7 mmol) in anhydrous THF (10 mL) at 0 °C under an
atmosphere of N2 was added PPh3 (0.89 g, 3.4 mmol) in small
portions over 1 h. The reaction mixture was allowed to warm to
room temp and stirred for 12 h. The reaction mixture was then
heated to 50 °C for 3 h, at which time water (0.2 mL) was added
and heating continued for a further 2 h. The mixture was the cooled
to room temp and the solvent removed in vacuo. The residue was
partitioned between HCl (1 M, 10 mL) and CH2Cl2 (10 mL). The
organic layer was separated and extracted with aqueous HCl (2 ×
10 mL). The combined aqueous layers were washed with CH2Cl2
(2 × 10 mL) and concentrated in vacuo. The crude amino ester
hydrochloride salt was hydrolyzed by refluxing in aqueous HCl (6
M, 15 mL) for 30 h, after which time the mixture was cooled to
room temp and the solvent removed in vacuo. The residue was
purified by ion exchange chromatography (Dowex 50, H+), eluting
first with water until the eluent was neutral and colorless, followed
by pyridine (1 M). Fractions containing the product were collected
and concentrated in vacuo to give 34 as a white solid (80 mg, 0.54
mmol, 64%, mp (dec.) 202-205 °C). 1H NMR (D2O) δ: 7.51 (m,
2H, NH2), 6.11 (bd, 1H, C(2)H J ) 9.8), 3.96 (m, C(4)H 1H),
2.86 (m, 2H, C(5)H2), 2.46 (m, C(3)H2, 2H), 1.22 (d, 3H, PCH3, J
) 14.1). 13C NMR (D2O) δ: 139.85 (d, C(2), 2JPCC ) 11.5), 136.85
NMR (D2O) δ: 7.54 (m, NH2, 2H), 6.12 (bd, 1H, C(2)H, J ) 9.2),
3.95 (m, C(4)H, 1H) 2.85 (dd, 2H, C(5)H2, J ) 7.7, J ) 16.1),
2.46 (dd, 2H, C(3)H2, J ) 3.4, J ) 18.2), 1.47 (m, PCH2(CH2)2CH3,
2H), 1.27 (m, PCH2(CH2)2CH3, 4H), 0.76 (t, 1H, P(CH2)3CH3, J
) 7.1). 13C NMR (D2O) δ: 137.68 (d, C(2), 2JPCC ) 10.2), 132.23
(d, C(1), JPC ) 121.8), 50.94 (d, C(4), J ) 10.1), 38.90 (d, C(3),
2
3JPCCC ) 14.8), 38.53 (d, C(5), JPCC ) 12.9), 29.61 (d, PCH2
3
(CH2)2CH3, J ) 98.7), 24.24 (d, P(CH2)2CH2CH3, JPCCC ) 3.6),
23.74 (d, PCH2CH2CH2CH3, 2JPCC ) 16.1), 13.15 (s, P(CH2)3CH3).
31P NMR (D2O) δ: 31.4; MS (EI) m/z (%): 204.1 (M+·, 10),
186.1(47), 80.1 (100), 68.1 (52). HRMS (EI): calcd for C9H18NO2P
(M+·) 203.1075; found 203.1069. Anal. (C9H18NO2P·0.33H2O) C,
H, N.
(-)-(R)-4-Amino-1-cyclopent-1-enyl(butyl)phosphinic acid
(40). Method as for 34; white crystalline solid (75%, mp (dec.)
193-196 °C). [R]D ) -7.1 (c ) 0.5, H2O). NMR spectra identical
to 39. Anal. (C9H18NO2P·0.5H2O) C, H, N.
(+)-(S)-4-Amino-1-cyclopent-1-enyl(butyl)phosphinic acid
(41). Method as for 39; white crystalline solid (33%, mp (dec.)
192-195 °C). [R]D ) +7.3 (c ) 0.4, H2O). NMR spectra identical to
39. Anal. (C9H18NO2P·0.33H2O) C, H, N.
(()-4-Amino-1-cyclopent-1-enyl(benzyl)phosphinic acid (42).
1
Method as for 34; White solid (67%, mp (dec.) 216-219 °C). H
NMR (D2O) δ: 7.26-7.10 (m, 5H, PCH2(C6H5)), 6.05 (bd, 1H,
C(2)H J ) 9.7), 3.88 (m, C(4)H, 1H), 2.92 (d, 2H, PCH2(C5H6), J
) 17.2), 2.84-2.62 (m, C(5)H2 2H), 2.47-2.29 (m, C(3)H2, 2H).
13C NMR (D2O) δ: 139.38 (d, C(2), 2JPCC ) 10.0), 138.07 (d, C(1),
JPC ) 127.5), 135.50 (d, C(1) of benzene ring, 2JPCC ) 7.1) 129.86
(d, benzene ortho, 3JPCCC ) 5.3), 128.67 (d, benzene meta, 4JPCCCC
3
5
(d, C(1), JPC ) 127.1), 50.81 (d, C(4), JPCCC ) 10.3), 39.09 (d,
) 2.9), 126.40 (d, benzene para, JPCCCCC ) 3.3), 50.94 (d, C(4),
C(5), 2JPCC ) 15.4), 37.99 (d, C(3), 3JPCCC ) 13.5), 14.85 (d, PCH3,
JPC ) 101.5). 31P NMR (D2O) δ: 33.7. MS (EI) m/z (%): 161.1
(M+·, 5), 144.0 (100), 80.1 (62), 66.1 (32). HRMS (EI): calcd for
C6H12NO2P (M+·) 161.0606; found 161.0610. Anal. (C6H12NO2P·
0.33H2O) C, H, N.
3JPCCC ) 10.0), 38.88 (d, C(3), J ) 12.8), 38.82 (d, C(5), J ) 15.29),
38.63 (d, PCH2(C6H5), J ) 91.6). 31P NMR (D2O) δ: 26.4. MS
(CI) m/z (%): 240.1 ([M + H]+, 100). HRMS (ESI) calcd for
C12H16NNaO2P ([M + Na]+) 260.08163; found 260.07998. Anal.
(C11H14NO2P·0.5H2O) C, H, N.
(-)-(R)-4-Amino-1-cyclopent-1-enyl(methyl)phosphinic acid
(35). Method as for 34; white crystalline solid (77%, mp (dec.)
202 - 205 °C). [R]D ) -9.7 (c ) 0.2, H2O). NMR spectra identical
to 34. Anal. (C9H18NO2P·0.33H2O)C, H, N.
Methyl((S)-4-((S)-3,3,3-trifluoro-2-methoxy-2-phenylpropana-
mido)cyclopent-1-enyl)phosphinic acid (43). To a suspension of
34 (4.0 mg, 24.8 µmol) in CH2Cl2 (1 mL) was added bis(trimeth-
ylsilyl)trifluoroacetamide (BSTFA, 19.2 mg, 74.5 µmol) and the
resulting solution stirred for 18 h. Further BSTFA was added as
required until the solution was clear. (S)-(-)-R-methoxy-R-(trif-
luoromethyl)phenylacetic acid chloride (8.0 mg, 29.8 mmol) was
added to the solution and the resulting solution stirred for 40 h.
Water (2 mL) was then added and the solution stirred for 30 min.
The aqueous layer was then separated, washed with diethyl ether/
CH2Cl2 (2:1), and concentrated in vacuo. The resulting solid was
used without further purification for 1H analysis of diastereomeric
(+)-(S)-4-Amino-1-cyclopent-1-enyl(methyl)phosphinic acid
(36). Method as for 34; white crystalline solid (93%, mp (dec.)
203 - 206 °C). [R]D ) +10.1 (c ) 0.20, H2O). NMR spectra
identical to 34. Anal. (C9H18NO2P·0.33H2O) C, H, N.
(()-4-Amino-1-cyclopent-1-enyl(ethyl)phosphinic acid (37).
1
Method as for 34. White solid (58%, mp (dec.) 201-203 °C). H
NMR (D2O) δ: 7.50 (m, 2H, NH2), 6.12 (bd, 1H, C(2)H, J ) 9.2),
3.95 (m, 1H, C(4)H), 2.85 (bdd, 2H, C(5)H2, J ) 7.5, J ) 16.7),
2.46 (dm, 2H, C(3)H2, J ) 15.8), 1.46 (dq, 2H, PCH2CH3, J )
15.3, J ) 7.7), 0.88 (dt, 3H, PCH2CH3, J ) 17.9, J ) 7.7). 13C
NMR (D2O) δ: 138.12 (d, C(1), JPC ) 123.3), 137.91 (d, C(2),
2JPCC ) 10.3), 50.93 (d, C(4), 3JPCCC ) 9.8), 38.93 (d, C(3), 2JPCC
1
purity. H NMR (CD3OD) δ: 7.54-7.45 (m, 2H, benzene ortho),
7.41-7.33 (m, 3H, benzene para and meta), 6.45 (bd, J ) 10.96,
1H, C(2)H), 4.69-4.55 (m, 1H, C(4)H), 3.59 (s, 3H, OCH3),
2.94-2.77 (m, 2H, C(3)H2), 2.57-2.35 (m, 2H, C(5)H2), 1.41 (d,
JPH ) 14.63, 3H, PCH3). e.e. 93.8% from line fitting data.
Methyl((R)-4-((S)-3,3,3-trifluoro-2-methoxy-2-phenylpropan-
amido)cyclopent-1-enyl)phosphinic acid (44). Method as for 43.
1H NMR (CD3OD) δ: 7.54-7.44 (m, 2H, benzene ortho), 7.40-7.33
(m, 3H, benzene para and meta), 6.43 (bd, J ) 11.05, 1H, C(2)H),
4.67-4.55 (m, 1H, C(4)H), 3.59 (s, 3H, OCH3), 3.09-2.75 (m,
2H, C(3)H2), 2.60-2.31 (m, 2H, C(5)H2), 1.45 (d, JPH ) 14.60,
3H, PCH3). e.e. 92.0% from line fitting data.
3
) 14.8), 38.55 (d, C(5), JPCCC ) 13.0), 22.71 (d, PCH2CH3, JPC
2
) 99.67), 5.94 (d, PCH2CH3, JPCC ) 4.9). 31P NMR (D2O) δ:
32.6; MS (EI) m/z (%): 175.1 (M+·, 2), 158.0 (100), 130.1 (43),
80.1 (71), 66.1 (44). HRMS (EI): calcd for C7H14NO2P (M+·
)
175.0762; found 175.0767. Anal. (C7H14NO2P·5H2O) C, H, N.
(()-4-Amino-1-cyclopent-1-enyl(isopropyl)phosphinic acid
(38). Method as for 34; white solid (67%, mp (dec.) 190-193 °C).
1H NMR (D2O) δ: 6.17 (bd, 1H, C(2)H, J ) 9.1), 3.95 (m, C(4)H
1H), 2.86 (dd, 2H, C(5)H2, J ) 7.5, J ) 16.8), 2.47 (d, 2H, C(3)H2,
J ) 16.2), 1.63 (m, 1H, CH(CH3)2), 0.94 (t, 3H, PCHCH3, J )
7.1), 0.88 (t, 3H, PCHCH3, J ) 7.2). 13C NMR (D2O) δ: 138.87
3-Oxocyclobutanecarbonyl chloride (46). A solution of 45 (4.0
g, 35 mmol) and thionyl chloride (12.0 g, 104 mmol) in benzene
(40 mL) was heated to 70 °C for 24 h. The solvent and excess
thionyl chloride were removed in vacuo and the crude product was
purified by vacuum distillation (bp 48-50 °C, 0.05 mm) to give
2
(d, C(2), JPCC ) 9.3), 137.56 (d, C(1), JPC ) 120.0), 50.98 (d,
3
3
C(4), JPCCC ) 9.4), 38.98 (apparent t, C(3) and C(5), JPCCC
)
1
13.8), 27.84 (d, PCH (CH3)2, JPC ) 101.1), 15.19 (dd, PCH (CH3)2,
46 as a colorless oil (3.6 g, 24 mmol, 69%). H NMR (CDCl3) δ:

3836 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
3.80-3.67 (m, 1H, C(3)H), 3.57 (m, 2H, C(2)H and C(4)H), 3.45
(m, 2H, C(2)H and C(4)H). 13C NMR (CDCl3) δ: 200.39 (C(1)),
174.85 (C(O)Cl), 52.17 (C(3)), 38.90 (C(2) and C(4)).
-1.09 (Si(CH3)3). 31P NMR (CD3OD) δ: 56.33. HRMS (ESI): calcd
for C15H32NO5PSiNa ([M + Na]+) 388.1680; found 388.1633.
cis-3-Amino-1-hydroxycyclobutyl(methyl)phosphinic acid
(51). 49 (59 mg, 161 µmol) was refluxed in HCl (6 M, 6 mL) for
36 h. The HCl was removed in vacuo to give the crude hydrochlo-
ride salt, which was purified by ion exchange chromatography
(Dowex 50, H+) to give 51 a colorless crystalline solid (26 mg,
157 µmol, 98%). 1H NMR (D2O) δ: 3.84 (p, 1H, C(1)H, J ) 6.9),
2.45 (dd, 4H, C(2)H2 and C(3)H2, J ) 10.7, J ) 6.0), 1.12 (d, 3H,
PCH3, J ) 13.4). 13C NMR (D2O) δ: 70.66 (d, C(1), JPC ) 112.9),
3-Aminocyclobutanone.HCl (47). A solution of 46 (1.0 g, 6.5
mmol) in benzene (8 mL) was added dropwise to a solution of
NaN3 (0.86 g, 13.2 mmol) in water (8 mL) at 0 °C. The resulting
biphasic solution was allowed to warm to room temp and was stirred
vigorously for 2 h. The benzene layer (caution:explosive) was then
separated, washed with saturated NaHCO3, water, and dried over
CaCl2. The benzene solution was then slowly heated to 60 °C,
during which time gas evolution (N2) was observed. After cessation
of gas evolution, the solution was heated for a further 1 h before
being cooled to room temp and 20% HCl (9 mL) added. The
solution was then heated overnight at 90 °C, and gas evolution
(CO2) was again observed. The aqueous layer was then separated,
the benzene layer washed with water, and the combined aqueous
layers concentrated in vacuo to give 47 as white crystals (0.7 g,
5.8 mmol, 76%, mp (dec.) 127-130 °C), which were used without
3
2
40.45 (d, C(3), JPCCC ) 6.9), 35.59 (d, C(2) and C(4), JPCC
)
2.6), 9.80 (d, PCH3, JPC ) 92.5). 31P NMR (D2O) δ: 39.95. HRMS
(ESI) calcd for C5H12NO3PNa ([M + Na] +) 188.0447; found
188.0452.
trans-3-Amino-1-hydroxycyclobutyl(methyl)-phosphinic acid
(52). Method as for 51; white solid (43 mg, 95%). 1H NMR (D2O)
δ: 3.44 (p, 1H, C(3)H, J ) 7.9), 2.76 (ddt, 2H, C(2)H and C(4)H,
J ) 10.7, J ) 7.9, J ) 3.0), 2.13 (dddd, 1H, C(2)H and C(4)H, J
) 12.8, J ) 11.3, J ) 8.2, J ) 2.9), 1.09 (d, 3H, PCH3, J ) 13.3).
13C NMR (D2O) δ: 67.15 (d, C(1), JPC ) 114.9), 38.00 (d, C(3),
3JPCCC ) 3.4), 37.27 (d, C(2) and C(4), 2JPCC ) 2.9), 9.94 (d, PCH3,
JPC ) 92.3). 31P NMR (D2O) δ: 40.41. HRMS (ESI) calcd for
C5H13NO3P ([M + H]+) 166.0628; found 166.0630.
1
further purification in the next step. H NMR (D2O) δ: 4.05 (m,
1H, C(3)H), 3.47 (m, 2H, C(2)H and C(4)H), 3.26 (m, 2H, C(2)H
and C(4)H). 13C NMR (D2O) δ: 207.11 (C(1)), 52.34 (C(3)), 36.63
(C(2) and C(4)). MS (CI) m/z (%): 86.0 (100, [M + H]+). (ESI)
m/z (%): 100.1 (65, [M + Na]+), 86.0 (100, [M + H]+).
3-(tert-Butoxycarbonylamino)cyclobutanone (48). To a stirred
solution of 47 (1.67 g, 13.7 mmol) and BOC2O (3.1 g, 14.0 mmol)
in DMF (30 mL) at room temp was added Et3N (1.7 g, 16 mmol)
dropwise. The solution was stirred at room temp for 2 h and
partitioned between water (50 mL) and diethyl ether (50 mL). The
ether layer was washed with brine, dried with Mg2SO4, and
concentrated in vacuo. The crude product was the purified via flash
chromatography (hexane:ethyl acetate, 60:40) to give 48 as colorless
crystals (1.38 g, 68%). 1H NMR (CDCl3) δ: 4.95 (bs, 1H, C(3)H),
4.30 (m, 1H, C(1)H), 3.44 (m, 2H, C(2)H and C(4)H), 3.07 (m,
2H, C(2)H and C(4)H), 1.49 (s, 9H, C(CH3)3). 13C NMR (CDCl3)
δ: 205.19 (C(1)), 155.63 (C(O)OC(CH3)3), 80.41 (C(CH3)3), 54.96
(C(2), C(4)), 37.03 (C(3)), 28.60 (C(CH3)3). HRMS (EI): calcd for
C9H15NO3 (M·+) 185.1052; found 185.1054.
trans-tert-Butyl 3-(ethoxy(methyl)phosphoryl)-3-hydroxycy-
clobutylcarbamate (53). To a solution of 52 (0.5 g, 1.4 mmol) in
acetonitrile (5 mL) was added Et3N·3HF (0.09 g, 0.55 mmol). A
precipitate formed rapidly and the resulting solution stirred at room
temp for a further 1 h. At this time CH2Cl2 (50 mL) was added
and the solution washed with saturated NaHCO3 (50 mL) and brine
(50 mL). Each aqueous phase was back extracted with CH2Cl2 (25
mL) and the combined organic phases dried over Mg2SO4 and
concentrated in vacuo to give 53 as a white solid (0.34 g, 85%).
1H NMR (CD3OD) δ: 4.16 (m, 2H, POCH2CH3), 3.88 (m, 1H,
C(1)H), 3.93 (m, 2H, C(2)H and C(4)H), 2.13 (m, 2H, C(2)H and
C(4)H), 1.52 (d, 3H, PCH3, J ) 13.6), 1.46 (s, 9H, C(CH3)3), 1.38
(t, 3H, POCH2CH3, J ) 7.0). 13C NMR (CD3OD) δ: 157.41
(C(O)Ot-Bu), 80.13 (C(CH3)3), 67.32 (d, C(1), JCP ) 120.2), 62.82
tert-Butyl 3-(ethoxy(methyl)phosphoryl)-3-(trimethylsilyloxy)-
cyclobutylcarbamate (49 and 50). Ethyl methylphosphinate (0.90
g, 4.9 mmol), 48 (1.0 g, 10.0 mmol), and hexamethyldisilazane
(HMDS, 1.6 g, 0.7 mmol) were mixed under N2 and heated to 90
°C for 24 h. The resulting mixture was diluted with CH2Cl2 (50
mL), washed with water (2 × 50 mL) and brine (50 mL), dried
over Mg2SO4, and concentrated in vacuo. The crude product was
purified via flash chromatography on silica gel (ethyl acetate, 1%
EtOH). Separation of the cis- (minor) and trans- (major) isomers
was possible under these conditions, the cis product eluting first
and the trans later to give a combined yield was 1.4 g, 98%.
2
2
(d, POCH2CH3, JPOC ) 7.3), 41.78 (d, C(2) and C(4), JPCC
)
3
1.8), 38.43 (C(3)), 28.77 (C(CH3)3), 17.03 (d, POCH2CH3, JPOCC
) 5.6), 8.34 (d, PCH3, JCP ) 89.5). 31P NMR (CD3OD) δ: 56.43.
HRMS (ESI): calcd for C12H24NO5PNa ([M + Na]+) 316.1284;
found 316.1286.
tert-Butyl 3-(ethoxy(methyl)phosphoryl)cyclobutyl carbam-
ate (54 and 55). 53 (0.33 g, 1.1 mmol) and 4-DMAP (0.28 g, 2.2
mmol) were suspended in anhydrous CH3CN (10 mL) under N2
and cooled to 0 °C. Methyl oxalyl chloride (0.21 g, 1.7 mmol) was
added and the mixture allowed to warm rapidly to room temp and
stirred an additional 2.5 h. After addition of ethyl acetate (40 mL),
the mixture was filtered and washed with saturated NaHCO3 (25
mL) and brine (25 mL), dried over MgSO4, and concentrated in
vacuo to give the oxalate as a slightly colored oil (0.42 g, 99%).
The oil, azobisisobutyronitrile (AIBN, 46 mg, 0.28 mmol), and
Bu3SnH (0.50 g, 1.7 mmol) were dissolved in anhydrous toluene
(10 mL) under N2 and the solution heated to 90 °C for 2.5 h. The
solution was concentrated in vacuo and the individual isomers
isolated cis- and trans- by flash chromatography on silica gel
(CH2Cl2, 5% CH3OH). Combined yield 0.21 g, 68%.
cis-tert-Butyl 3-(Ethoxy(methyl)phosphoryl)-3-(trimethylsi-
lyloxy)cyclobutylcarbamate (49). 1H NMR (CD3OD) δ: 4.45-4.32
(m, 1H, NH), 4.24-4.02 (m, 3H, C(3)H and POCH2CH3),
2.66-2.40 (m, 4H, C(2)H2 and (C(4)H2), 1.43 (s, 9H, C(CH3)3),
1.43 (d, 3H, PCH3, J ) 13.6), 1.34 (t, 3H, POCH2CH3, J ) 7.0),
0.25 (s, 9H, Si(CH3)3). 13C NMR (CD3OD) δ: 157.44 (NH-
C(O)OC(CH3)3), 80.31 (C(CH3)3), 73.41 (d, C(3), JPC ) 124.8),
62.86 (d, POCH2CH3, 2JPOC ) 7.3), 41.09 (d, C(1), 3JPCCC ) 4.6),
2
39.76 (d, C(2) and C(4), JPCC ) 1.0), 28.77 (C(CH3)3), 17.04 (d,
3
POCH2CH3, JPOCC ) 5.7), 8.35 (d, PCH3, JPC ) 89.8), 2.00
(Si(CH3)3). 31P NMR (CD3OD) δ: 54.02. HRMS (ESI): calcd for
cis-tert-Butyl 3-(Ethoxy(methyl)phosphoryl)cyclobutyl car-
C15H32NO5PSiNa (MNa+) 388.1680; found 388.1635.
1
bamate (54). H NMR (CD3OD) δ: 4.18-3.93 (m, 3H, C(3) and
POCH2CH3), 2.49-2.29 (m, 3H, C(2)H, C(4)H and C(1)H),
2.20-1.94 (m, 2H, C(2)H and C(4)H), 1.39 (s, 9H, C(CH3)3), 1.40
(d, 3H, PCH3, J ) 13.6), 1.27 (t, 3H, POCH2CH3, J ) 7.0). 13C
NMR (CD3OD) δ: 157.30 (C(O)OC(CH3)3), 80.22 (C(O)OC-
trans-tert-Butyl 3-(ethoxy(methyl)phosphoryl)-3-(trimethyl-
silyloxy)cyclobutylcarbamate (50). 1H NMR (CD3OD) δ: 4.21-4.05
(m, 3H, POCH2CH3 and NH), 3.93-3.77 (m, 1H, C(1)H),
2.99-2.82 (m, 2H, C(2)H and C(4)H), 2.33-2.02 (m, 2H, C(2)H
and C(4)H), 1.47 (d, 3H, PCH3, J ) 13.6), 1.43 (s, 9H, C(CH3)3),
1.36 (t, 3H, POCH2CH3, J ) 7.1), 0.21 (s, 9H, Si(CH3)3). 13C NMR
(CD3OD) δ: 157.40 (NHC(O)OC(CH3)3), 80.14 (C(CH3)3), 67.32
2
3
(CH3)3), 62.15 (d, JPOC ) 6.66, POCH2CH3), 44.65 (d, JPCCC
)
30.1, C(3)), 32.21 (d, C(2) or C(4), 2JPCC ) 3.3), 31.63 (d, C(2) or
2
C(4), JPCC ) 5.4), 28.75 (C(O)OC(CH3)3), 26.78 (d, C(1), JPC
)
2
3
(d, C(3), JPC ) 120.2), 62.81 (d, POCH2CH3, JPOC ) 7.2), 41.78
102.5), 16.96 (d, POCH2CH3, JPOCC ) 5.9), 11.39 (d, PCH3, JPC
) 92.3). 31P NMR (CD3OD) δ: 57.47. HRMS (ESI): calcd for
C12H24NO4PNa+ ([M + Na]+) 300.1335; found 300.1334.
2
(d, C(2) and C(4), JPCC ) 1.9), 38.44 (C(1)), 28.78 (C(CH3)3),
3
17.03 (d, POCH2CH3, JPOCC ) 5.6), 8.34 (d, PCH3, JPC ) 89.6),

γ-Aminobutyric Acid F1 Receptor Antagonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3837
trans-tert-Butyl 3-(ethoxy(methyl)phosphoryl)cyclobutyl car-
bamate (55). H NMR (CD3OD) δ: 4.25-4.10 (m, 1H, C(3)H),
of drug were added to the buffer solution. For GABAA and GABAB
containing oocytes, a washout period of 7 min was allowed to
minimize desensitization.
1
4.10-3.97 (m, 2H, POCH2CH3), 2.61-2.41 (m, 3H, C(2)H, C(4)H
and C(1)H), 2.35-2.14 (m, 2H, C(2)H and C(4)H), 1.45 (d, 3H,
PCH3, J ) 13.5), 1.38 (s, 9H, C(CH3)3), 1.29 (t, 3H, POCH2CH3,
J ) 7.0). 13C NMR (CD3OD) δ: 157.35 (C(O)OC(CH3)3), 80.16
Recombinant GABA receptors where expressed in Xenopus
laeVis oocytes with similar properties to those reported earlier, these
were: homomeric F1,64 heteromeric GABAB(1A/2) (coexpressed with
Kir 3.1 and Kir 3.4 K+ channels),80 and heteromeric R1ꢀ2γ2L
receptors.81 The voltage potential was clamped at -60 mV, and
the amplitudes of the whole-cell currents varied upon receptor
opening. For all receptors, the currents ranged from 200 to 10000
nA. The maximal GABA response was attained by the application
of 100 µM GABA for F1 receptors, 600 µM GABA for both
GABAB(1A/2) and R1ꢀ2γ2L receptors (see Supporting Information for
representative electrophysiological recordings).
Data Analysis. The amplitude of the current (I) recorded in
response to each drug concentration [A] was normalized to the
maximum amplitude (Imax) of the current response to GABA (F1,
100 µM; GABAB, 600 µM; R1ꢀ2γ2L, 600 µM). Normalized
concentration-response data were analyzed using the GraphPad
“Prism” version 4.00, expressed as mean ( SEM, plotted on a
semilogarithmic scale and analyzed using the Gaussian distribution
equation. The equation used was
2
(C(O)OC(CH3)3), 62.15 (d, POCH2CH3, JPOC ) 6.8), 45.26 (d,
C(3), 3JPCCC ) 4.6), 30.99 (d, C(2) or C(4), 2JPCC ) 5.7), 30.92 (d,
C(2) or C(4), 2JPCC ) 4.7) 28.76 (C(O)OC(CH3)3), 27.79 (d, C(1),
JPC ) 100.8), 16.99 (d, POCH2CH3, 3JPOCC ) 5.9), 11.18 (d, PCH3,
JPC ) 91.0). 31P NMR (CD3OD) δ: 60.34. HRMS (ESI): calcd for
C12H24NO4PNa+ ([M + Na]+) 300.1335; found 300.1338.
cis-3-Amino-cyclobutyl(methyl)phosphinic acid (56). Method
as for 51; off-white solid (75 mg, 0.50 mmol, 93%, mp (dec.)
185-187 °C). 1H NMR (D2O) δ: 3.82 (d, C(3)H, J ) 7.6),
2.54-2.36 (m, 3H, C(2)HA, C(4)HA and C(1)H), 2.26-2.06 (m,
2H, C(2)HB and C(4)HB), 1.27 (d, 3H, PCH3, J ) 13.8). 13C NMR
3
(D2O) δ: 43.01 (d, C(3), JPCCC ) 26.7), 27.31 (d, C(2) and C(4),
2JPCC ) 4.6), 26.88 (d, C(1), JPC ) 100.6), 12.22 (d, PCH3, JPC
)
91.8). 31P NMR (CD3OD) δ: 50.52. HRMS (ESI): calcd for
C5H13NO2P ([M + H]+) 150.0678; found 150.0682.
trans-3-Amino-cyclobutyl(methyl)phosphinic acid (57). Method
as for 51; white solid (20 mg, 0.13 mmol, 93%, mp (dec.) 145-147
°C). 1H NMR (D2O) δ: 3.82 (d, 1H, C(3)H, J ) 7.6), 2.63-2.27 (m,
Y ) (area ⁄ [SD(2π)0.5] exp(-0.5(X - mean)SD)2)
(1)
5H, C(2)H2, C(4)H2 and C(1)H), 1.26 (d, 3H. PCH3, J ) 13.6). 13
C
)
)
This data was then fit by least-squares to the Hill equation:
NMR (D2O) δ: 44.06 (d, C(3), 3JPCCC ) 6.4), 27.47 (d, C(1), JPC
98.6), 26.40 (d, C(2) and C(4), 2JPCC ) 4.7), 12.20 (d, PCH3, JPC
nH
nH
[ ]
A
nH
[ ]
+ A
%I ) I
⁄ EC
(
)
50
(2)
)
(
max
91.9). 31P NMR (D2O) δ: 51.15. HRMS (ESI): calcd for C5H13NO2P
([M + H]+) 150.0678; found 150.0682.
where Imax is the maximal response, EC50 is the agonist concentra-
tion, which elicits 50% of the maximal response, and nH is the Hill
coefficient. The intrinsic activity of partial agonists was calculated
as a persentage of the maximum current produced by a maximum
dose of GABA.
pKB values were calculated by determination of the pEC50 of
GABA alone and in the presence of one or more concentrations of
the antagonist [B], tested for statistical significance via 2-way
ANOVA and fitted to the equation:
Pharmacology. Expression of Recombinant Human GABA
Receptors in Xenopus Oocytes. Xenopus laeVis were anaesthetised
with 0.17% ethyl 3-aminobenzoate and a lobe of an ovary was
removed and rinsed with oocyte releasing buffer, OR2 (82.5 mM
NaCl, 2 mM KCl, 1 mM MgCl2 ·6H2O, 5 mM HEPES, pH 7.5). It
was then treated with Collagenase A (2 mg/ml or OR2, Bohringer
Manheim) for 2 h. The released oocytes were rinsed in modified
frog Ringers solution (96 mM NaCl, 2 mM KCl, 1 mM
MgCl2 ·6H2O, 1.8 mM CaCl2, 5 mM HEPES, 2.5 mM pyruvate,
0.5 mM theophylline, 50 ng/mL gentamycin, pH 7.5). Stage V-VI
oocytes were collected and stored in this buffer.
Human GABAC F1 DNA cloned into the vector pcDNA1.1
(Invitrogen, CA) was donated by Dr. George Uhl (National Institute
for Drug Abuse, Baltimore, MD). Human R1, ꢀ2, and γ2L GABAA
DNA cloned into pCMV6-XL5 or pCMV6-XL4 plasmid vectors
were purchased from Origene Technologies Inc. (Rockville, MD).
Human GABAB1a, GABAB2, rat GIRK1, and rat GIRK4 cloned into
their corresponding vectors were donated by Dr. Fiona Marshall
(GlaxoWellcome, Hertfordshire, UK). GABAB1a was cloned in the
pcDNA3.1 (-) (Invitrogen, USA), GABAB2 and rat GIRK1 were
cloned into pcDNA3 (Invitrogen, USA), while rat GIRK4 was
cloned into pBluescript KS(-) (Stratgene, USA). Xenopus laeVis
were obtained from South Africa and housed in the Department of
Veterinary Science, University of Sydney. Plasmids containing the
F1, R1, γ2L, and girk4 were linearized with restriction enzyme Xba1,
similarly GABAB1A was linearized with BST981; GABAB2 and
girk1 were linearized with EcoR1 and not1, respectively. ꢀ2 was
expelled from the vector using a double digest of Sac1 and Sma1.
cRNA was synthesized using the “mMessage mMachine” kit from
Ambion (Austin, TX). cRNA was injected into defolliculated
oocytes at a concentration of 50 ng/50 nL. Oocytes were stored
for 1-5 days at 18 °C.
log K
( )
B
(
)
- P
[ ]
pEC50 ) -log B + 10
(3)
where P is a constant. Where several antagonist concentrations were
used, the data was tested to see if it was consistent with the criteria
for simple competitive antagonism via an F-test using eq 3 above
and:
pEC50)-log([B]slope+10log(K)) - P
(4)
If eq 3 fits the data better, the interaction is defined as competitive
antagonism.
Molecular Modeling. All molecules investigated in this study
were built with the molecular modeling and graphical user interface
package, Maestro 8.0,82 unless otherwise stated. Each structure was
subsequently minimized using the OPLS_200583 with a constant
dielectric of 1.0. All calculations were performed on an AMD64+
workstation using the SUSE10.1 Linux operating system.
Pharmacophore Perception. Pharmacophore development was
carried out using PHASE 2.5 in the Maestro modeling environment
in accordance with the PHASE 2.5 user manual.66 Primary attention
was placed on a small subset of potent F1 GABAC receptor
antagonists: TPMPA (3), (S)-4-ACPMPA (36), (1R,3S)-3-ACPMPA
(60), (1S,3R)-3-ACPMPA (63), and I4AA (63), with the intention
of highlighting common pharmacophore sites, while concomitantly
contrasting the pharmacophore differences presented by the weak
or inactive GABAC receptor antagonist, (R)-4-ACPMPA (35).
Compounds TPMPA (3), (R)-4ACPMPA (35), (S)-4ACPMPA
(36), (1R,3S)-3-ACPMPA (60), (1S,3R)-3-ACPMPA (63), and
I4AA (63) were divided into an active class, 3, 36, 60, 62, and 63,
and an inactive class, 35, and thus encompassing a broad range of
activity, potent and weak, respectively. A maximum of 100
conformations were generated for each molecule, sampling the
torsional space that may contain possible active conformers. This
was achieved using a combination of Monte Carlo Multiple
Electrophysiology. Receptor activity was measured by two
electrode voltage clamp recording using a Geneclamp 500 amplifier
(Axon Instruments, Foster City, CA), a MacLab 2e recorder (AD
Instruments, Sydney, NSW, Australia), and Chart version 5.1
program. Oocytes were voltage clamped at -60 mV and continu-
ously superfused with frog Ringers solution (96 mM NaCl, 2 mM
KCl, 1 mM MgCl2 ·6H2O, 1.8 mM CaCl2, 5 mM HEPES) or in
the case of GABAB receptors a buffer consisting of 45 mM KCl,
45 mM NaCl, 1 mM MgCL2, 1.8 mM CaCl2, and 5 mM HEPES.
For receptor activation measurements, the indicated concentrations

3838 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
Minimum84 and low mode conformational searching,85 in conjuga-
tion with the OPLS_2005 force field, GB/SA solvation treatment,
and a minimum atom deviation of 1.0 Å. This was conducted to
exhaustively search all conformational space, in a water solvated
environment. Minimized structures were filtered with a maximum
relative energy difference of 10.0 kcal/mol to exclude redundant
conformers.
On the basis of visual inspection and careful analysis of chemical
features in the actives, pharmacophore variants considered for the
models consisted of five features: H-bond donor, negatively charged
feature, positively charged feature, and two custom group features
representing the double bond and alkyl substituent. Thus, model
variant DNPXY was preselected, as it was visually apparent that
active compounds possessed DNP features with the option of
additionally having either X and/or Y.
The five active antagonists were analyzed on their potential to
present the DNPXY features in a common 3D spatial arrangement
of functional groups (sites) and with a minimum intersite distance
of 1.0 Å, suggesting a highly specific 3D pharmacophore model.
Several pharmacophore hypotheses were found. The chosen hy-
pothesis corresponded to the model, which gave the best overall
PHASE scoring functions66 for each active when screened for three
sites to exact matches of D, N, and P site points, leaving the fourth
site point as either the double bond (X) or methyl substituent (Y).
The resulting hypothesis was deemed the preliminary “F1 GABAC
receptor antagonist pharmacophore” and applied in subsequent
refinement procedures.
References
(1) Bloom, F. E.; Iversen, L. L. Localizing 3H-GABA in nerve terminals
of rat cerebral cortex by electron microscopic autoradiography. Nature
1971, 229, 628–630.
(2) Chebib, M.; Johnston, G. A. R. GABA-Activated Ligand Gated Ion
Channels: Medicinal Chemistry and Molecular Biology. J. Med. Chem.
2000, 43, 1427–1447.
(3) Wainwright, A.; Sirinathsinghji, D. J. S.; Oliver, K. R. Expression of
GABAA receptor R5 subunit-like immunoreactivity in human hippoc-
ampus. Mol. Brain Res. 2000, 80, 228–232.
(4) O’Connell, A. W.; Fox, G. B.; Kjoller, C.; Gallagher, H. C.; Murphy,
K. J.; Kelly, J.; Regan, C. M. Anti-ischemic and cognition-enhancing
properties of NNC-711, a γ-aminobutyric acid reuptake inhibitor. Eur.
J. Pharmacol. 2001, 424, 37–44.
(5) Zheng, W.; Xie, W.; Zhang, J.; Strong, J. A.; Wang, L.; Yu, L.; Xu,
M.; Lu, L. Function of γ-aminobutyric acid receptor/channel F1
subunits in spinal cord. J. Biol. Chem. 2003, 278, 48321–48329.
(6) Sernagor, E.; Young, C.; Eglen, S. J. Developmental modulation of
retinal wave dynamics: Shedding light on the GABA saga. J. Neurosci.
2003, 23, 7621–7629.
(7) Arnaud, C.; Gauthier, P.; Gottesmann, C. Study of a GABAC receptor
antagonist on sleep-waking behavior in rats. Psychopharmacology
(Berlin) 2001, 154, 415–419.
(8) Wisor, J. P.; DeLorey, T. M.; Homanics, G. E.; Edgar, D. M. Sleep
states and sleep electroencephalographic spectral power in mice lacking
the ꢀ3 subunit of the GABAA receptor. Brain Res. 2002, 955, 221–
228.
(9) Billinton, A.; Baird, V. H.; Thom, M.; Duncan, J. S.; Upton, N.;
Bowery, N. G. GABAB(1) mRNA expression in hippocampal sclerosis
associated with human temporal lobe epilepsy. Mol. Brain Res. 2001,
86, 84–89.
(10) Baulac, S.; Huberfeld, G.; Gourfinkel-An, I.; Mitropoulou, G.;
Beranger, A.; Prud’homme, J.-F.; Baulac, M.; Brice, A.; Bruzzone,
R.; LeGuern, E. First genetic evidence of GABAA receptor dysfunction
in epilepsy: a mutation in the γ2-subunit gene. Nat. Genet. 2001, 28,
46–48.
(11) Kang, J.; MacDonald, R. L. The GABAA receptor γ2 subunit R43Q
mutation linked to childhood absence epilepsy and febrile seizures
causes retention of R1ꢀ2γ2S receptors in the endoplasmic reticulum.
J. Neurosci. 2004, 24, 8672–8677.
(12) Prosser, H. M.; Gill, C. H.; Hirst, W. D.; Grau, E.; Robbins, M.; Calver,
A.; Soffin, E. M.; Farmer, C. E.; Lanneau, C.; Gray, J.; Schenck, E.;
Warmerdam, B. S.; Clapham, C.; Reavill, C.; Rogers, D. C.; Stean,
T.; Upton, N.; Humphreys, K.; Randall, A.; Geppert, M.; Davies, C. H.;
Pangalos, M. N. Epileptogenesis and enhanced prepulse inhibition in
GABAB1-deficient mice. Mol. Cel. Neurosci. 2001, 17, 1059–1070.
(13) Berretta, S.; Lange, N.; Bhattacharyya, S.; Sebro, R.; Garces, J.; Benes,
F. M. Long-term effects of amygdala GABA receptor blockade on
specific subpopulations of hippocampal interneurons. Hippocampus
2004, 14, 876–894.
Pharmacophore Refinement: Finding Matches to Known
GABAC Receptor-Ligands. Quality assurance and refinements
measures were conducted on the pharmacophore model as a means
of optimizing its robustness. The pharmacophore was iteratively
screened against a comprehensive list of pharmacologically char-
acterized GABA receptor-ligands (see Supporting Information)
using the Find Matches to Hypothesis Panel.65 The list included
GABAA and GABAB ligands as well as five additional potent
GABAC receptor antagonists that were not used in the original
pharmacophore development: 3-APMPA (5), (S)-4-ACPCA (S-6),
(1R,3R)-3-ACPMPA (61), (1S,3S)-3-ACPMPA (59), and P4MPA
(58). Successful alignments to the pharmacophore require a
minimum of three sites corresponding to exact matches of D, N,
and P site points, leaving the fourth site point as either the double
bond (X) or an alykyl substituent (Y). Distance matching tolerance
was set to 2.0 Å.
(14) Ishikawa, M.; Mizukami, K.; Iwakiri, M.; Hidaka, S.; Asada, T.
Immunohistochemical and immunoblot study of GABAA R1 and ꢀ2/3
subunits in the prefrontal cortex of subjects with schizophrenia and
bipolar disorder. Neurosci. Res. 2004, 50, 77–84.
(15) Mizukami, K.; Ishikawa, M.; Hidaka, S.; Iwakiri, M.; Sasaki, M.;
Iritani, S. Immunohistochemical localization of GABAB receptor in
the entorhinal cortex and inferior temporal cortex of schizophrenic
brain. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 2002, 26, 393–
396.
(16) Huntsman, M. M.; Tran, B.-V.; Potkin, S. G.; Bunney, W. E., Jr.;
Jones, E. G. Altered ratios of alternatively spliced long and short γ2
subunit mRNAs of the γ-amino butyrate type A receptor in prefrontal
cortex of schizophrenics. Proc. Natl. Acad. Sci. U.S.A. 1998, 95,
15066–15071.
The “create_xvolClash” and “create_xvolShell” Schro¨dinger
utility scripts65 were invoked to systematically locate regions where
excluded volume spheres would cause steric violations only for
inactives, in tandem with generating a shell of excluded volume
spheres to cater only for the active set. A grid spacing of 0.5 Å
and a buffer distance of 0.1 Å was used to create the shell. Hits
with align scores greater than 1.2 and vector and volume scores
less than 0.70 and 0.0, respectively, were rejected. These latter
refinement steps were not performed until the models performance
in identifying other GABAC active antagonists had been validated.
The resulting DNPXY pharmacophore, with the excluded volume
shell, was deemed the final F1 GABAC antagonist pharmacophore.
(17) Baeckstroem, T.; Andersson, A.; Andree, L.; Birzniece, V.; Bixo, M.;
Bjoern, I.; Haage, D.; Isaksson, M.; Johansson, I.-M.; Lindblad, C.;
Lundgren, P.; Nyberg, S.; Oedmark, I.-S.; Stroemberg, J.; Sundstroem-
Poromaa, I.; Turkmen, S.; Wahlstroem, G.; Wang, M.; Wihlbaeck,
A.-C.; Zhu, D.; Zingmark, E. Pathogenesis in menstrual cycle-linked
CNS disorders. Ann. N.Y. Acad. Sci. 2003, 1007, 42–53.
(18) Goddard, A. W.; Narayan, M.; Woods, S. W.; Germine, M.; Kramer,
G. L.; Davis, L. L.; Petty, F. Plasma levels of gamma-aminobutyric
acid and panic disorder. Psychiatry Res. 1996, 63, 223–225.
(19) Chang, L.; Cloak Christine, C.; Ernst, T. Magnetic resonance
spectroscopy studies of GABA in neuropsychiatric disorders. J. Clin.
Psychiatry 2003, 3 (64 Suppl.), 7–14.
(20) Prosser, J.; Hughes, C. W.; Sheikha, S.; Kowatch, R. A.; Kramer,
G. L.; Rosenbarger, N.; Trent, J.; Petty, F. Plasma GABA in children
and adolescents with mood, behavior, and comorbid mood and
behavior disorders: a preliminary study. J. Child Adolescent Psychop-
harmacol. 1997, 7, 181–99.
Acknowledgment. We thank Dr. Hue Tran, Dr. Erica
Campbell, and Dr. Suzanne Habjan for harvesting oocytes used
in receptor activation measurements and Dr. Jane Carland for
preparing human F1 mRNA. We also thank Bruce Tattam, Dr.
Ian Luck, and Dr. Keith Fischer for technical assistance with
31P NMR and mass spectrometry measurements. R.J.K. was
supported by an Australian Postgraduate Award.
Supporting Information Available: e.e. determination, repre-
sentative electrophysiological recordings, compounds used in
computational analysis and elemental analysis of key compounds.
This material is available free of charge via the Internet at http://
pubs.acs.org.

γ-Aminobutyric Acid F1 Receptor Antagonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13 3839
(21) Polenzani, L.; Woodward, R. M.; Miledi, R. Expression of mammalian
γ-aminobutyric acid receptors with distinct pharmacology in Xenopus
oocytes. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 4318–4322.
(22) Shimada, S.; Cutting, G.; Uhl, G. R. gamma-Aminobutyric acid A or
C receptor? gamma-Aminobutyric acid F1 receptor RNA induces
bicuculline-, barbiturate-, and benzodiazepine-insensitive gamma-
aminobutyric acid responses in Xenopus oocytes. Mol. Pharmacol.
1992, 41, 683–687.
(23) Drew, C. A.; Johnston, G. A. R.; Weatherby, R. P. Bicuculline-
insensitive GABA receptors: studies on the binding of (-)-baclofen
to rat cerebellar membranes. Neurosci. Lett. 1984, 52, 317–321.
(24) Amin, J.; Weiss, D. S. Homomeric rho 1 GABA channels: activation
properties and domains. Recept. Channels 1994, 2, 227–326.
(25) Rozzo, A.; Armellin, M.; Franzot, J.; Chiaruttini, C.; Nistri, A.;
Tongiorgi, E. Expression and dendritic mRNA localization of GABAC
receptor F1 and F2 subunits in developing rat brain and spinal cord.
Eur. J. Neurosci. 2002, 15, 1747–58.
tion of enzymic hydrolysis and Mitsunobu reaction. Synlett 1991, 44,
1–2.
(46) Ogura, K.; Yamashita, M.; Tsuchihashi, G. Synthesis of (R)- and (S)-
4-hydroxy-3-cyclopentenones. Tetrahedron Lett. 1976, 759, 62.
(47) Paquette, L. A.; Gao, Z.; Ni, Z.; Smith, G. F. Total synthesis of
spinosyn A. 1. Enantioselective construction of a key tricyclic
Intermediate by a multiple configurational inversion scheme. J. Am.
Chem. Soc. 1998, 120, 2543–2552.
(48) Hanrahan, J. R.; Mewett, K. N.; Chebib, M.; Burden, P. M.; Johnston,
G. A. R. An improved, versatile synthesis of the GABAC antagonists
(1,2,5,6-tetrahydropyridin-4-yl)methylphosphinic acid (TPMPA) and
(piperidin-4-yl)methylphosphinic acid (P4MPA). J. Chem. Soc., Perkin
Trans. 1 2001, 2389–2392.
(49) Johnston, G. A. R.; Burden, P. M.; Mewett, K. N.; Chebib, M.
Neurologically active phosphinic acid compound GABAC receptor
antagonists, therapeutic methods, and compositions. PCT Int. Appl.
WO 9858939, 23 June, 1998.
(26) Enz, R.; Brandstaetter, J. H.; Waessle, H.; Bormann, J. Immunocy-
tochemical localization of the GABAC receptor F subunits in the
mammalian retina. J. Neurosci. 1996, 16, 4479–4490.
(27) Boue-Grabot, E.; Roudbaraki, M.; Bascles, L.; Tramu, G.; Bloch, B.;
Garret, M. Expression of GABA receptor F subunits in rat brain.
J. Neurochem. 1998, 70, 899–907.
(28) Ge, S.; Goh, E. L. K.; Sailor, K. A.; Kitabatake, Y.; Ming, G.-l.; Song,
H. GABA regulates synaptic integration of newly generated neurons
in the adult brain. Nature 2006, 439, 589–593.
(29) Liu, B.; Hattori, N.; Jiang, B.; Nakayama, Y.; Zhang, N.-Y.; Wu, B.;
Kitagawa, K.; Taketo, M.; Matsuda, H.; Inagaki, C. Single cell RT-
PCR demonstrates differential expression of GABAC receptor F
subunits in rat hippocampal pyramidal and granule cells. Mol. Brain
Res. 2004, 123, 1–6.
(30) Leung, C. K. S.; Yeung, C. K.; Chiang, S. W. Y.; Chan, K. P.; Pang,
C. P.; Lam, D. S. C. GABAA and GABAC (GABAA0r) receptors affect
ocular growth and form-deprivation myopia. J. Toxicol., Cutaneous
Ocul. Toxicol. 2005, 24, 187–196.
(31) Stone Richard, A.; Liu, J.; Sugimoto, R.; Capehart, C.; Zhu, X.;
Pendrak, K. GABA, experimental myopia, and ocular growth in chick.
InVest. Opthalmol. Vis. Sci. 2003, 44, 3933–3946.
(50) Dumond, Y. R.; Montchamp, J.-L. Palladium-catalyzed cross-coupling
reaction of anilinium hypophosphite with alkenyl bromides and
triflates: application to the synthesis of GABA analogs. J. Organomet.
Chem. 2002, 653, 252–260.
(51) Bravo-Altamirano, K.; Huang, Z.; Montchamp, J.-L. Palladium-
catalyzed phosphorus-carbon bond formation: cross-coupling reactions
of alkyl phosphinates with aryl, heteroaryl, alkenyl, benzylic, and
allylic halides and triflates. Tetrahedron 2005, 61, 6315–6329.
(52) Mitsunobu, O.; Wada, M.; Sano, T. Stereospecific and stereoselective
reactions. I. Preparation of amines from alcohols. J. Am. Chem. Soc.
1972, 94, 697–680.
(53) Smith, E. C. R.; McQuaid, L. A.; Paschal, J. W.; DeHoniesto, J. An
enantioselective synthesis of D-(-)- and L-(+)-2-amino-3-phospho-
nopropanoic acid. J. Org. Chem. 1990, 55, 4472–4474.
(54) Pigou, P. E.; Schiesser, C. H. Convenient route to 1,3-disubstituted
cyclobutanes. An inexpensive synthesis of 3-oxocyclobutanecarboxylic
acid. J. Org. Chem. 1988, 53, 3841–3843.
(55) Daly, A. M.; Gilheany, D. G. The synthesis and use in asymmetric
epoxidation of metal salen complexes derived from enantiopure trans-
cyclopentane- and cyclobutane-1,2-diamine. Tetrahedron: Asymmetry
2003, 14, 127–137.
(32) Mondadori, C.; Moebius, H.-J.; Zingg, M. CGP 36742, an orally active
GABAB receptor antagonist, facilitates memory in a social recognition
test in rats. BehaV. Brain Res. 1996, 77, 227–229.
(33) Murata, Y.; Woodward, R. M.; Miledi, R.; Overman, L. E. The first
selective antagonist for a GABAC receptor. Bioorg. Med. Chem. Lett.
1996, 6, 2073–2076.
(34) Allan, R. D.; Dickenson, H. W.; Fong, J. Structure-activity studies
on the activity of a series of cyclopentane GABA analogs on GABAA
receptors and GABA uptake. Eur. J. Pharmacol. 1986, 122, 339–
348.
(56) Hansen, H. I.; Kehler, J. A convenient preparation of difficultly
accessible sec-alkylmethylphosphinates using sequential Pudovik/
Abramov-Barton/McCombie reactions. Synthesis 1999, 1925, 1930.
(57) Overman, L. E.; Okazaki, M. E.; Jacobsen, E. J. Synthesis applications
of cationic aza-Cope rearrangements. 16. Stereocontrolled synthesis
of substituted cis-cyclopenta[b]pyrrolidines. J. Org. Chem. 1985, 50,
2403–2405.
(58) Lew, M. J.; Angus, J. A. Analysis of competitive agonist-antagonist
interactions by nonlinear regression. Trends Pharmacol. Sci. 1995,
16, 328–337.
(35) Chebib, M.; Duke, R. K.; Allan, R. D.; Johnston, G. A. R. The effects
of cyclopentane and cyclopentene analogues of GABA at recombinant
GABAC receptors. Eur. J. Pharmacol. 2001, 430, 185–192.
(36) Froestl, W.; Mickel, S. J.; Schmutz, M.; Bittiger, H. Potent, orally
active GABAB receptor antagonists. Pharmacol. ReV. Commun. 1996,
8, 127–133.
(59) Crittenden, D. L.; Chebib, M.; Jordan, M. J. T. A quantitative
structure-activity relationship investigation into agonist binding at
GABAC receptors. THEOCHEM 2005, 755, 81–89.
(60) Harrison Neil, J.; Lummis Sarah, C. R. Molecular modeling of the
GABA(C) receptor ligand-binding domain. J. Mol. Model. 2006, 12,
317–324.
(37) Duke, R. K.; Chebib, M.; Balcar, V. J.; Allan, R. D.; Mewett, K. N.;
Johnston, G. A. R. (+)- and (-)-cis-2-aminomethylcyclopropanecar-
boxylic acids show opposite pharmacology at recombinant F1 and F2
GABAC receptors. J. Neurochem. 2000, 75, 2602–2610.
(38) Allan, R. D.; Curtis, D. R.; Headley, P. M.; Johnston, G. A. R.;
Kennedy, S. M. E.; Lodge, D.; Twitchin, B. Cyclobutane analogs of
GABA. Neurochem. Res. 1980, 5, 393–400.
(39) Safanda, J.; Sobotka, P. 3-Aminocyclobutane-1-carboxylic acid:
synthesis and some neurochemical properties. Collect. Czech. Chem.
Commun. 1982, 47, 2440–7.
(61) Sedelnikova, A.; Smith, C. D.; Zakharkin, S. O.; Davis, D.; Weiss,
D. S.; Chang, Y. Mapping the F1 GABAC Receptor Agonist Binding
Pocket: Constructing a Complete Model. J. Biol. Chem. 2005, 280,
1535–1542.
(62) Brejc, K.; van Dijk, W. J.; Klaassen, R. V.; Schuurmans, M.; van der
Oost, J.; Smit, A. B.; Sixma, L. K. Crystal structure of an ACh-binding
protein reveals the ligand-binding domain of nicotinic receptors. Nature
2001, 411, 269–276.
(63) Chebib, M.; Hanrahan, J. R.; Wooller, S.; Morriss, G.; Kumar, R. J.;
Mewett, K. N.; Johnston, G. A. R. (3-Aminocyclopentyl)methylphos-
(40) An, M.; Toochinda, T.; Bartlett, P. A. Five-Membered Ring Analogues
of Shikimic Acid. J. Org. Chem. 2001, 66, 1326–1333.
(41) Hetmanski, M.; Purcell, N.; Stoodley, R. J.; Palfreyman, M. N. Studies
related to cyclopentanoid natural products. Part 3. Synthesis of
pentenomycin and its racemate. J. Chem. Soc., Perkin Trans. 1 1984,
2089, 96.
phinic acids: Novel GABA
2007, 52, 779–787.
(64) Kusama, T.; Spivak, C. E.; Whiting, P.; Dawson, V. L.; Schaeffer,
J. C.; Uhl, G. R. Pharmacology of GABA F1 and GABA R/ꢀ receptors
expressed in Xenopus oocytes and COS cells. Br. J. Pharmacol. 1993,
109, 200–206.
C receptor antagonists. Neuropharmacology
(42) Ghorpade, S. R.; Bastawade, K. B.; Gokhale, D. V.; Shinde, P. D.;
Mahajan, V. A.; Kalkote, U. R.; Ravindranathan, T. Enzymatic kinetic
resolution studies of racemic 4-hydroxy-2-cyclopenten-1-one using
Lipozyme IM. Tetrahedron: Asymmetry 1999, 10, 4115–4122.
(43) Okamoto, Y.; Aburatani, R.; Kawashima, M.; Hatada, K.; Okamura,
N. Chromatographic resolution. XIII. Resolution of 4-hydroxy-2-
cyclopentenone derivatives by HPLC on cellulose tris(phenylcarbam-
ate) derivatives. Chem. Lett. 1986, 1767, 70.
(44) Khanapure, S. P.; Najafi, N.; Manna, S.; Yang, J.-J.; Rokach, J. An
Efficient Synthesis of 4(S)-Hydroxycyclopent-2-enone. J. Org. Chem.
1995, 60, 7548–51.
(45) Danda, H.; Furukawa, Y.; Umemura, T. Preparation of (4S)-4-hydroxy-
3-methyl-2-(2-propynyl)-2-cyclopenten-1-one by sequential combina-
(65) Phase, Version 2.5; Schro¨dinger LLC: New York, 2005.
(66) Dixon, S. L.; Smondyrev, A. M.; Knoll, E. H.; Rao, S. N.; Shaw,
D. E.; Friesner, R. A. PHASE: a new engine for pharmacophore
perception, 3D QSAR model development, and 3D database screening:
1. Methodology and preliminary results. J. Comput.-Aided Mol. Des.
2006, 20, 647–671.
(67) Chebib, M.; Vandenberg, R. J.; Johnston, G. A. R. Analogs of
γ-aminobutyric acid (GABA) and trans-4-aminocrotonic acid (TACA)
substituted in the 2 position as GABAC receptor antagonists. Br. J.
Pharmacol. 1997, 122, 1551–1560.
(68) Chebib, M.; Vandenberg, R. J.; Froestl, W.; Johnston, G. A. R.
Unsaturated phosphinic analogs of γ-aminobutyric acid as GABAC
receptor antagonists. Eur. J. Pharmacol. 1997, 329, 223–229.

3840 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 13
Kumar et al.
(69) Krehan, D.; Frolund, B.; Krogsgaard-Larsen, P.; Kehler, J.; Johnston,
G. A. R.; Chebib, M. Phosphinic, phosphonic and seleninic acid
bioisosteres of isonipecotic acid as novel and selective GABAC
receptor antagonists. Neurochem. Int. 2003, 42, 561–565.
(70) Krehan, D.; Frolund, B.; Ebert, B.; Nielsen, B.; Krogsgaard-Larsen,
P.; Johnston, G. A. R.; Chebib, M. Aza-THIP and related analogues
of THIP as GABAC antagonists. Bioorgan. Med. Chem. 2003, 11,
4891–4896.
(71) Ragozzino, D.; Woodward, R. M.; Murata, Y.; Eusebi, F.; Overman,
L. E.; Miledi, R. Design and in vitro pharmacology of a selective
γ-aminobutyric acid C receptor antagonist. Mol. Pharmacol. 1996,
50, 1024–1030.
(72) Froestl, W.; Mickel, S. J.; Hall, R. G.; von Sprecher, G.; Strub, D.;
Baumann, P. A.; Brugger, F.; Gentsch, C.; Jaekel, J.; Olpe, H.-R.;
Rihs, G.; Vassout, A.; Waldmeier, P. C.; Bittiger, H. Phosphinic Acid
Analogs of GABA. 1. New Potent and Selective GABAB Agonists.
J. Med. Chem. 1995, 38, 3297–312.
(73) Hills, J. M.; Dingsdale, R. A.; Parsons, M. E.; Dolle, R. E.; Howson,
W. 3-Aminopropylphosphinic acid: a potent, selective GABAB receptor
agonist in the guinea pig ileum and rat anococcygeus muscle. Br. J.
Pharmacol. 1989, 97, 1292–1296.
site through ligand design and pharmacophore modeling. Mol. Neu-
ropharmacol. 2004, 113–127.
(77) Gleiter, C. H.; Farger, G.; Mobius, H. J. Pharmacokinetics of CGP
36742, an orally active GABAB antagonist, in humans. J. Clin.
Pharmacol. 1996, 36, 428–438.
(78) Paquette, L. A.; Heidelbaugh, T. M. (4S)-(-)-tert-Butyldimethylsiloxy-
2-cyclopenten-1-one. Org. Synth. 1996, 73, 44–49.
(79) Paquette, L. A.; Earle, M. J.; Smith, G. F. (4R)-(+)-tert-butyldimeth-
ylsiloxy-2-cyclopenten-1-one. Org. Synth. 1996, 73, 36–43.
(80) Taniyama, K.; Takeda, K.; Ando, H.; Kuno, T.; Tanaka, C. Expression
of the GABAB receptor in Xenopus oocytes and inhibition of the
response by activation of protein kinase C. FEBS Lett. 1991, 278,
222–224.
(81) Ebert, B.; Wafford, K. A.; Whiting, P. J.; Krogsgaard-Larsen, P.;
Kemp, J. A. Molecular pharmacology of γ-aminobutyric acid type A
receptor agonists and partial agonists in oocytes injected with different
R, ꢀ, and γ receptor subunit combinations. Mol. Pharmacol. 1994,
46, 957–963.
(82) Maestro, Version 8.0; Schro¨dinger, LLC: New York, 2006.
(83) MacroModel, Version 9.1; Schro¨dinger, LLC: New York, 2005.
(84) Chang, G.; Guida, W.; Still, W. C. An internal coordinate Monte Carlo
method for searching conformational space. J. Am. Chem. Soc. 1989,
111, 4379.
(85) Kolossvary, I.; Guida, W. C. Low Mode Search. An Efficient,
Automated Computational Method for Conformational Analysis:
Application to Cyclic and Acyclic Alkanes and Cyclic Peptides. J. Am.
Chem. Soc. 1996, 118, 5011–5019.
(86) Woodward, R. M.; Polenzani, L.; Miledi, R. Characterization of
bicuculline/baclofen-insensitive (F-like) γ-aminobutyric acid receptors
expressed in Xenopus oocytes. II. Pharmacology of γ-aminobutyric
acid A and γ-aminobutyric acid B receptor agonists and antagonists.
Mol. Pharmacol. 1993, 43, 609–625.
(74) Froestl, W.; Mickel, S. J.; von Sprecher, G.; Diel, P. J.; Hall, R. G.;
Maier, L.; Strub, D.; Melillo, V.; Baumann, P. A.; Bernasconi, R.;
Gentsch, C.; Hauser, K.; Jaekel, J.; Karlsson, G.; Klebs, K.; Maitre,
L.; Marescaux, C.; Pozza, M. F.; Schmutz, M.; Steinmann, M. W.;
van Riezen, H.; Vassout, A.; Mondadori, C.; Olpe, H.-R.; Waldmeier,
P. C.; Bittiger, H. Phosphinic Acid Analogs of GABA. 2. Selective,
Orally Active GABAB Antagonists. J. Med. Chem. 1995, 38, 3313–
3331.
(75) Seabrook, G. R.; Howson, W.; Lacey, M. G. Electrophysiological
characterization of potent agonists and antagonists at pre- and
postsynaptic GABAB receptors on neurones in rat brain slices. Br. J.
Pharmacol. 1990, 101, 949–957.
(76) Frolund, B.; Jorgensen, A. T.; Liljefors, T.; Mortensen, M.; Krogs-
gaard-Larsen, P. Characterization of the GABAA receptor recognition
JM7015842