Evaluation of difluoromethyl ketones as agonists of the γ-aminobutyric acid type B (GABAB) receptor

Article abstract of DOI:10.1021/jm301805e

The design, synthesis, biological evaluation, and in vivo studies of difluoromethyl ketones as GABA_B agonists that are not structurally analogous to known GABA_B agonists, such as baclofen or 3-aminopropyl phosphinic acid, are present

Full text of DOI:10.1021/jm301805e

Article
pubs.acs.org/jmc
Evaluation of Difluoromethyl Ketones as Agonists of the
γ‑Aminobutyric Acid Type B (GABAB) Receptor
Changho Han,† Amy E. Salyer,† Eun Hoo Kim,† Xinyi Jiang,† Rachel E. Jarrard,† Matthew S. Powers,‡
Aaron M. Kirchhoff,‡ Tolani K. Salvador,§ Julia A. Chester,‡ Gregory H. Hockerman,*,†
and David A. Colby*,†,§
†Department of Medicinal Chemistry and Molecular Pharmacology, Purdue University, West Lafayette, Indiana 47907, United States
‡Department of Psychological Sciences, Purdue University, West Lafayette, Indiana 47907, United States
§Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, United States
S
* Supporting Information
ABSTRACT: The design, synthesis, biological evaluation, and in vivo studies of
difluoromethyl ketones as GABAB agonists that are not structurally analogous to
known GABAB agonists, such as baclofen or 3-aminopropyl phosphinic acid, are
presented. The difluoromethyl ketones were assembled in three synthetic steps
using a trifluoroacetate-release aldol reaction. Following evaluation at clinically
relevant GABA receptors, we have identified a difluoromethyl ketone that is a
potent GABAB agonist, obtained its X-ray structure, and presented preliminary in
vivo data in alcohol-preferring mice. The behavioral studies in mice demonstrated
that this compound tended to reduce the acoustic startle response, which is consistent with an anxiolytic profile. Structure−
activity investigations determined that replacing the fluorines of the difluoromethyl ketone with hydrogens resulted in an inactive
analogue. Resolution of the individual enantiomers of the difluoromethyl ketone provided a compound with full biological
activity at concentrations less than an order of magnitude greater than the pharmaceutical, baclofen.
cocaine, heroin, and nicotine addiction in addition to its uses in
muscle relaxation (e.g., spinal and lower esophageal muscu-
lature).9−11
INTRODUCTION
■
γ-Aminobutyric acid (GABA) is the major inhibitory neuro-
transmitter in the central nervous system, and its primary roles
are to regulate the excitation of neurons and the release of other
neurotransmitters. The actions of GABA are guided by two
major types of GABA receptors, GABAA and GABAB. GABAA
receptors are ion channels that increase the threshold for
excitation in neurons and other cells, and they are
physiologically distinct, pentameric assemblies of 19 different
subunits.1 The GABAA receptor is modulated by the
benzodiazepine sedative/hypnotics, such as diazepam, and
newer nonbenzodiazepine sedatives, such as zolpidem,
zopiclone, and zaleplon.2 GABAB receptors are G-protein
coupled receptors that can inhibit the release of neuro-
transmitters such as acetylcholine, noradrenalin, serotonin,
glutamic acid, and dopamine.3,4 They are functional hetero-
dimers of GABAB1 and GABAB2, and the former contains the
GABA-binding site and the latter is the direct activator of G-
protein (Gi and Go types). The GABAB receptors mediate the
action of GABA through modulation of potassium channels,5
calcium channels,6,7 and adenylyl cyclase.8 Unlike targeting the
GABAA receptor, few pharmaceuticals are selective ligands for
the GABAB receptor, except for the muscle relaxant, baclofen.
Moreover, all of the other known GABAB agonists are structural
analogues of GABA or baclofen; therefore, structure−activity
data at this receptor is quite restricted.9 However, targeting the
GABAB receptor presents a potential strategy for pharmaco-
logical intervention for anxiety and mood disorders and alcohol,
Herein, we report the synthesis, structure−activity data and
preliminary behavioral studies in mice of difluoromethyl
ketone-based GABAB agonists. These structures are a
divergence from existing GABAB agonists, such as GABA and
baclofen, yet retain substantial potency and selectivity for the
GABAB receptor over GABAA receptors. Structure−activity
investigations determined that replacing the fluorines of the
difluoromethyl ketone with hydrogens resulted in an inactive
analogue. In the preliminary studies in mice, an active
compound tended to reduce the acoustic startle response,
which is consistent with an anxiolytic drug profile. The acoustic
startle response is commonly used in both humans and animals
to assess anxiety-related behavior.12
Baclofen and 3-aminopropyl phosphinic acid represent the
prototypical structures of selective GABAB agonists (Figure
1).9,13 Baclofen is a racemic mixture of 3-(para-chlorophenyl)-
substituted GABA, the majority of its biological activity being
displayed by the (R)-enantiomer.13 Baclofen is approved as a
muscle relaxant and has demonstrated substantial preclinical
and clinical potential in alcohol, cocaine, heroin, and nicotine
addiction, panic and post-traumatic stress disorders, and
gastroesophageal reflux (GERD).9 Other baclofen derivatives
Received: December 7, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
and simultaneously, an increasing number of pharmaceutical
agents contain fluorine, typically because this atom can enhance
drug-like qualities and limit metabolism.27 Difluoromethyl
ketones are prominent examples of such fluorinated com-
pounds because the presence of the electron-withdrawing
fluorine atom causes the ketone to revert to a hydrate, or gem-
diol, in the presence of water (eq 1).25,28,29 This change
Figure 1. Structures of GABA and GABAB agonists.
with structural modifications at the 3-position are also GABAB
agonists,14−16 notably, substituted thiophene and benzofuran
are potential surrogates for the para-chlorophenyl group.14 3-
Aminopropyl phosphinic acid is a GABAB agonist10 and also
served as a scaffold for the discovery of additional GABAB
agonists, partial agonists, and antagonists, usually through the
placement of substituents at the 2-position.9,17,18 For example,
the analogue with a (R)-2-fluoro substituent on 3-aminopropyl
phosphinic acid is a candidate molecule for inhibition of
GERD.18 3-Aminopropyl phosphinic acid and its derivatives
have demonstrated a potential role as muscle relaxants,
anxiolytics, and treatments for addiction.9 Because of the
substantial therapeutic possibilities for the GABAB ligands,
other classes of agents such as allosteric modulators have been
developed.9,19,20 Nevertheless, the published structure−activity
data for GABAB agonism is primarily limited to baclofen and 3-
aminopropyl phosphinic acid; therefore, we aimed to address
this deficiency and concomitantly demonstrate a novel role for
difluoromethyl ketones. This class of fluorinated compounds
has established their fundamental importance in modulating
unique biological activity or serving as a transition-state mimic.
increases aqueous solubility and imparts a tetrahedral shape,
which can further enable activity as a transition state mimic.
Thusly, difluoromethyl ketones have displayed biological
activity as inhibitors of matrix metalloprotease,25 esterase,28
malarial aspartic protease,29 γ-secretase,30 and renin.31 How-
ever, a potential role for difluoromethyl ketones as agonists of
GABAB receptors has not been demonstrated.
Recently, we have disclosed a novel method to assemble
difluoromethyl ketones by aldol reactions with difluoroenolates
generated following the release of trifluoroacetate (Figure 3).26
RESULTS AND DISCUSSION
■
Chemistry. Structural data for the GABAB receptor is quite
scarce, yet site-directed mutagenesis and ligand binding studies
have revealed that two key active-site residues, Ser246 and
Asp471, have critical interactions with GABA and baclofen
(Figure 2).21 Also, Tyr366 plays a pivotal role in the binding of
Figure 3. Difluoromethyl ketones from trifluoroacetate-release aldol
reaction.
This approach was developed from the literature precedent that
hexafluoroacetone hydrate undergoes a rapid fragmentation to
release trifluoroacetate after the addition of base.32 The aldol
reaction is rapid (reaction time is 3 min at rt) and promoted
under many conditions; LiBr and Et3N were the optimum
conditions for the transformation with aldehydes. Difluor-
omethyl ketones can be accessed in only three synthetic steps
from a precursor methyl ketone. This strategy is high yielding,
versatile, and fully compatible with many alkyl and aryl
aldehydes and ketones.26 Also, a complementary approach for
difluoromethyl ketones is available by halogenation of the
difluoroenolate generated by trifluoroacetate release, followed
by copper-mediated coupling.33 This difluoromethyl ketone
scaffold from the aldol process presents the requisite
difluoromethyl ketone to propel these investigations as
potential agonists as well as a β-hydroxy group that could
recapitulate the role of the amino group of GABA (and
baclofen).
Using the three-step approach, difluoromethyl ketones 1−12
were assembled for evaluation as potential GABAB ligands
(Figure 4).26 The compounds represent a focused selection of
difluoromethyl ketones that were assembled from pairing alkyl,
aryl, and α,β-unsaturated aldehydes and ketones. All of the
molecules were prepared as racemic mixtures, except for
adducts 8 and 9, which are mixtures of epimers derived from
coupling with chiral aldehydes. The difluoromethyl ketone 12
resulted from protonation of the difluoroenolate rather than
Figure 2. Baclofen-GABAB receptor binding motif determined from
site-directed mutagenesis and ligand-binding studies.
baclofen to the GABAB receptor. In this model, the interaction
between the carboxylate of baclofen and the serine residue of
the GABAB receptor presents a unique opportunity to discover
a new potential use for α-fluorinated ketones as agonists at this
receptor. This assertion builds upon two literature precedents
from investigations with α-fluorinated ketones in medicinal
chemistry: (1) their ability to form hemiketals with serine
residues in an active site22,23 and (2) their ability to serve as
surrogates for carboxylic acids24 and phosphinic acids.25 We
envisioned that applying these data would also leverage existing
structure−activity data for GABAB agonists and together enable
a manifold of unique ligands to be identified for the GABAB
receptor. Another driving force that enabled this investigation
was that we have recently discovered a synthetic method to
assemble α,α-difluorinated ketones, or difluoromethyl ke-
tones.26 Approaches to create fluorinated molecules are rising,
B
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. GABAB and GABAA Assay Data for Difluoromethyl
Ketones
a
a
compd
GABAB EC50 (μM)
GABAA EC50 (μM)
baclofen
1.7 0.10
1.9 0.90
0.53 0.33
61.8 3.01
66.9 1.19
99.3 3.78
53.5 1.76
>100
>100
>100
2.30 0.59
>100
>100
>100
>100
nd
rac-BHFF
GABA
1
2
3
4
5
6
>100
nd
7
>100
nd
8
>100
nd
b
9
nd
nd
10
11
12
24.9 1.30
40.0 3.59
>100
>100
>100
nd
a
b
Values are given with standard error. Cytotoxic in assay. nd is not
Figure 4. Structures of difluoromethyl ketones 1−12.
determined.
reaction with an aldehyde and was prepared to provide
additional insight into structure−activity relationships. An X-ray
structure of compound 10 was obtained to reveal the
conformation of the difluoromethyl ketone backbone (Figure
5). Only Percy and co-workers have reported an X-ray structure
Figure 6. Other ligands of the GABAB receptor.
4 and adamantyl analogues 10 and 11 display significant activity
in the GABAB assay, and GABA, rac-BHFF, and baclofen served
as positive controls. Analysis of these data validated our
hypothesis that difluoromethyl ketones can serve as GABAB
agonists. The EC50 values calculated from this assay of GABAB
receptor activity for the positive controls (i.e., GABA and
baclofen) are comparable to the prior reported litera-
ture.16,18,21,38 The most potent agent is the difluoromethyl
ketone 10 (EC50 = 24.9 μM), bearing an adamantyl group
adjacent to the α,α-difluoroketone and a para-acetyl phenyl
group flanking the secondary alcohol. The other adamantyl
derivative 11 is less active (EC50 = 40.0 μM) than 10, yet more
potent than all of the naphthyl-derived analogues 1−4. Analysis
of structure−activity data emphasizes the importance of a
naphthyl or an adamantyl group adjoining the α,α-difluor-
oketone, but the functionality neighboring the hydroxy group
can tolerate a much wider range of structures. However, the
simple naphthyl difluoroketone 12 is inactive, which indicates
that importance of the hydroxy group. Additional pharmaco-
logical evaluation of 10 revealed that its effects can be reversed
with (2S)-3-[[(1S)-1-(3,4-dichlorophenyl)ethyl]amino-2-
hydroxypropyl](phenylmethyl)phosphinic acid (CGP55845),
a GABAB antagonist (Figure 6).39 Thus, the inhibition of
forskolin-stimulated AC activity is, indeed, mediated by the
GABAB receptor. Compound 10 is not cytotoxic in an MTS cell
viability assay (unlike compound 9, for example). The dose−
response curves for compound 10, GABA, and baclofen depict
the relative potencies as well as the slope of the curves (Figure
Figure 5. X-ray structures of GABAB ligands: (A) compound ( )-10,
(B) baclofen.
of an analogous acyclic α,α-difluoro-β-hydroxy ketone.34 The
X-ray structure of baclofen is also presented in Figure 5B.35 The
crystalline difluoroketone 10 exists as the ketone, rather than a
hydrate, according to the X-ray structure. Difluoroketones are
known to convert partially or completely into hydrate in an
aqueous environment.25,28,29,36 Accordingly, the tendency to
favor the hydrate-form over the ketone-form in the presence of
water is dependent upon the structure of the difluoroketone as
well as the relative concentration of water.
Pharmacology. Biological evaluation of difluoromethyl
ketones 1−12, baclofen, and GABA was performed to measure
potency and selectivity at GABAB and GABAA receptors (Table
1). GABAB agonist activity was measured using a cell line
expressing both subunits of the human GABAB receptor, along
with the reporter gene, CRE-luciferase (CRE-Luc). The
introduced GABAB receptor utilizes the endogenous signaling
pathway of the host cells, and activation of GABAB receptors
leads to inhibition of cAMP production. In this assay, forskolin,
a small-molecule activator of adenylyl cyclase (AC), was used to
stimulate cAMP production, and subsequently, inhibition of
forskolin-stimulated accumulation of cAMP was measured. The
GABAB allosteric activator, rac-BHFF,20 was synthesized
according to the published protocol37 and examined as an
additional positive control (Figure 6). Dose−response curves
and EC50 values were obtained for compounds displaying
substantial activity at 100 μM. The naphthyl aldol products 1−
7). Even though GABA and baclofen are more potent (EC50
=
0.53 and 1.7 μM, respectively) than the difluoromethyl ketone
10, the dose−response curve of compound 10 is steeper than
both of the other agonists. These differences in slope merited
further investigation to deduce potential binding modes as well
C
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
well as the effects of chirality. The pharmaceutical baclofen is
available as a racemic mixture, but it is well-known that (R)-
(−)-baclofen is the major active component.13 The trifluor-
oacetate-release aldol reaction provides racemic mixtures of
products; therefore, the separation of the difluoromethyl ketone
( )-10 was conducted to resolve each of its individual
enantiomers for subsequent pharmacological evaluation.
Using the secondary alcohol of ( )-10 as a handle, chiral
esters derived from naproxen, histidine, and proline were
prepared, and fortunately, the N-Boc proline adduct 13 allowed
complete separation by preparative TLC (Figure 9). Accord-
Figure 7. Dose−response curves for baclofen, GABA, and compound
10 at the GABAB receptor.
as the relative effects of each enantiomer of the racemic
compound 10 (see below).
All of the active compounds were also screened for agonist
activity at the GABAA receptor (α1β2γ2) because GABA also
activates GABAA receptors in vivo. This assay is an electro-
physiological assay, utilizing the whole-cell voltage-clamp
method, because the GABAA receptor is a ligand-gated chloride
channel. Specifically, HEK 293 cells were transiently transfected
with α1, β2, and γ2 subunits of the rat GABAA receptor, and
compounds were applied to cells clamped at −60 mV. Of the
compounds tested, only GABA displays activity at the GABAA
receptor, with a potency similar to that previously reported
(EC50 = 2.30 μM). Thus, the pharmacological evaluation
strategy produces data for potency and efficacy in GABAB and
GABAA as well as GABAB vs GABAA selectivity. These data
validate that the difluoromethyl ketones (i.e., 1−4, 10, and 11)
are, indeed, agonists of the GABAB receptor and selective for
GABAB over GABAA receptors (see Table 1). Electro-
physiology results for compounds 1−3 and GABA in HEK
cells are presented in Figure 8. They depict the magnitude of
the response at GABAA receptors from multiple applications of
100 μM GABA as well as its absence following sequential
treatment with 100 μM of the difluoromethyl ketones.
Figure 9. Separation of enantiomers of compound ( )-10 and
synthesis of methyl ether 14 and nonfluorinated analogue 15.
ingly, mixture ( )-10 was esterified with (S)-Boc-proline using
CDI, and the intermediates were cleanly separated by
preparative TLC. Saponification of each compound with
KOH in methanol provided (+)-enantiomer 10 ([α]23 +36.0
D
(c 0.667, CHCl3)) and (−)-enantiomer 10 ([α]23 −35.8 (c
D
0.558, CHCl3)). Two derivatives of compound ( )-10 were
also synthesized to obtain the structure−activity data at the
secondary alcohol and difluoromethyl ketone. Methylation of
10 with CH3I in the presence of Ag2O afforded the methyl
ether 14. The preparation of the nonfluorinated analogue of 10
was accomplished by the aldol reaction of 1-adamantyl methyl
ketone with 4-acetylbenzaldehyde to give the adduct 15 in 68%
yield. The structure 15 represents the nonfluorinated counter-
part 10 and will be used to understand the pharmacophoric
function of the difluoromethyl ketone because both fluorines
are replaced with hydrogen atoms.
Following the identification of the active compounds,
structure−activity investigations were performed on 10 to
elucidate the role of difluorination and the secondary alcohol as
Pharmacological evaluation of the (+)-enantiomer 10,
(−)-enantiomer 10, methyl ether 14, and nonfluorinated
analogue 15 along with ( )-baclofen and (−)-baclofen
provided additional structure−activity data (Table 2). The
(+)-enantiomer 10 was twice as potent at GABAB (EC50 = 15.9
μM) as the (−)-enantiomer 10 (EC50 = 37.8 μM) and provides
excellent agreement with the observed activity for the racemic
mixture. The methylated compound 14 and the nonfluorinated
analogue 15 are both inactive at the GABAB receptor at 100
μM. The absence of biological activity displayed by 14 and 15
demonstrates the importance of the β-hydroxy substituent as
well as the fluorines adjacent to the methyl ketone, respectively.
The latter structure−activity relationship validates the utility of
the difluoromethyl ketone and supports that this group is
pivotal for activity at the GABAB receptor, because by simply
exchanging the two fluorine atoms on 10 with hydrogen atoms,
Figure 8. Example trace from electrophysiological experiments testing
the agonist activity of compounds 1−3 and GABA at the GABAA
receptor. Solid bars represent the time during which the indicated
compounds (100 μM) were applied to GABAA receptor-expressing
cells under voltage clamp.
D
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. Biological Activity for Resolved Enantiomers of
a
Compound 10 and Derivatives 14 and 15 .
b
c
b
compd
GABA
GABAB EC50 (μM)
Hill slope
GABAA EC50 (μM)
0.53 0.33
1.9 0.90
1.7 0.10
1.0 0.40
24.9 1.30
15.9 1.84
37.8 0.78
>100
0.54 0.11
0.92 0.30
0.98 0.05
nd
2.30 0.59
>100
>100
nd
rae-BHFF
( )-baclofen
(−)-baclofen
( )-10
3.8 0.60
2.2 0.39
9.53 1.66
nd
>100
nd
(+)-10
(−)-10
( )-14
nd
nd
( )-15
>100
nd
nd
a
b
Hill slopes are calculated at GABAB. Values are given with standard
c
error. Calculated for GABAB. nd is not determined.
the activity is lost. The dose−response curves of the individual
enantiomers and racemic mixture are as expected (Figure 10).
Figure 11. Compound ( )-10 does not potentiate or antagonize the
activity of GABA at the GABAA receptor. (A) Representative trace of
an experiment showing stimulation of Cl− current by 1 and 100 μM
GABA in the absence or presence of 100 μM ( )-10 recorded from a
cell transfected with the GABAA receptor subunits as depicted in
Figure 8. Solid bars represent the time during which the indicated
compounds were applied to the cell. (B) Normalized current
amplitude (mean SE; N = 5) stimulated by 1 μM GABA in the
absence (gray bar) or presence (black bar) of 100 μM ( )-10. (C)
Normalized current amplitude (mean SE; N = 5) stimulated by 100
μM GABA in the absence (gray bar) or presence (black bar) of 100
μM ( )-10.
Figure 10. Dose−response curves for (+)-10, (−)-10, and ( )-10 at
the GABAB receptor.
The more potent enantiomer, (+)-10, presents a dose−
response with a less steep curve (Hill slope = 2.2) than its
enantiomer (Hill slope = 9.53) and the racemic mixture (Hill
slope = 3.8), accordingly. This difference in the Hill slopes for
the enantiomers may be potentially interpreted as a difference
in the binding modes for the individual enantiomers or may
reflect additional allosteric activation. Similar to the racemate,
both enantiomers of compound 10 are not cytotoxic in an MTS
cell viability assay. Overall, the (+)-enantiomer 10 is a GABAB
agonist that displays its potency within 10-fold of the
pharmaceutical agent, baclofen, and is a key structure for
additional development. Although the determination of the
absolute stereochemistry of (+)-enantiomer 10 would provide
additional structural data, the promising nature of the racemate
( )-10 as a GABAB agonist, as well as its three-step
preparation, impelled a subsequent in vivo investigation.
To determine if compound 10 has either GABAA-
potentiating or GABAA antagonizing activity, we applied 1
μM GABA alone or simultaneously with 100 μM ( )-10 to
cells under voltage clamp expressing the GABAA receptors as in
Figure 8. As shown in parts A and B of Figure 11, compound
10 did not change the current amplitude elicited by 1 μM
GABA and thus does not potentiate GABAA receptor responses
to GABA. We also applied 100 μM compound 10
simultaneously with 100 μM GABA, a near-saturating
concentration in this assay. We found that compound 10 did
not affect the amplitude of the Cl− current elicited by 100 μM
GABA to stimulate Cl−. Combined with the data in Table 1,
these data show that compound 10 has no discernible effect on
the gating of the GABAA receptor Cl− channel.
Behavioral Pharmacology. Rodent startle models, includ-
ing plain acoustic startle and fear-potentiated startle, have
predictive validity for identifying potential candidates for
treating clinical anxiety disorders.12,40 Anxiety disorders are
highly comorbid with addictions such as alcohol depend-
ence.41,42 In two experiments, we used a mouse line selectively
bred for high-alcohol preference (HAP2) to assess the potential
anxiolytic effects of compound ( )-10 and rac-BHFF in
reference to ( )-baclofen (Figure 12). HAP2 mice have a
genetic predisposition toward high-alcohol preference, are a
genetic animal model of alcohol dependence, and also display
greater anxiety-related behavior, as shown by greater plain
acoustic startle43 and fear-potentiated startle.44 In experiment 1,
mice were pretreated by intraperitoneal (ip) dosing with either
( )-baclofen (7.5 mg/kg) or vehicle 25 min before the start of
the test session. The dose for ( )-baclofen was obtained from
prior studies.45 Figure 12A shows anxiolytic effects of
( )-baclofen in HAP2 mice. ANOVA indicated a Treatment
× Trial Type (noise-alone, light + noise) interaction [F(1,10) =
26.3, p < 0.001]. Follow up analyses of Treatment within each
Trial Type showed that ( )-baclofen significantly reduced the
startle response on both noise-alone [F(1,12) = 8.1, p = 0.015]
and light + noise [F(1,12) = 26.3, p < 0.001] trials. Analyses of
Trial Type within each treatment group [( )-baclofen, vehicle]
indicated a significant difference between noise-alone and light
E
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
difluoromethyl ketone into the CNS because the pharmacoki-
netics of difluoromethyl ketones have not been widely explored.
Two reports support that fluorinated ketones can display
favorable pharmacokinetic profiles.31,46 Specifically, difluoro-
methyl ketones developed as renin inhibitors have lowered
blood pressure in normotensive monkeys following oral
administration.31 Second, fluorinated ketones developed as
phospholipase A2 inhibitors have reduced symptom onset in
the CNS disease, experimental autoimmune encephalitis,
following intravenous and intraperitoneal dosing.46 The latter
study demonstrates that a trifluoromethyl ketone, which is
likely to exist as a hydrate, is able to access the CNS.
CONCLUSIONS
■
The design, synthesis, biological evaluation, and in vivo studies
of difluoromethyl ketones as GABAB agonists that are not
structurally analogous to any known GABAB agonists are
presented. This class of molecules provides a path to extend the
structure−activity data for GABAB agonists beyond the confines
of the existing molecules designed from baclofen or 3-
aminopropyl phosphinic acid. Biological evaluation with a
unique set of GABA receptor subtypes has enabled
investigation at clinically relevant receptors. Moreover, these
results were directly translated to an animal model to assess
drug effects on the acoustic startle response. During the course
of this work, we have identified the difluoromethyl ketone
( )-10 as a potent GABAB agonist, obtained its X-ray structure,
and presented preliminary in vivo data in mice that suggest it
has biological activity. Resolution of this racemic mixture
provided individual enantiomers, one of which was determined
to be highly active and only 10-fold less potent in receptor
binding than the pharmaceutical agent, baclofen. The
implications of these studies are that difluoromethyl ketones
can serve as potent agonists of the GABAB receptor with high
selectivity over the GABAA receptor. The in vivo data suggest
this compound warrants further study as a potential anxiolytic
drug. An additional investigation of the difluoromethyl ketone
was conducted by synthesizing the nonfluorinated counterpart
of the active 10, i.e. 15, did not display any activity at the
GABAB receptor. These structure−activity data demonstrate
the pivotal nature of the difluoromethyl ketone for activity at
the GABAB receptor. Additional studies to ascertain the exact
stereochemical configuration of the chiral secondary alcohol in
the most potent analogue are underway. We have previously
reported that the assignment of absolute stereochemistry of α-
fluorinated alcohols is more complex than their nonfluorinated
counterparts because the usual Mosher’s ester analysis provides
an ambiguous assignment.26 Difluoromethyl ketones can
display good selectivity profiles, and we now demonstrate
their selectivity for GABAB over GABAA receptors. Similarly,
Diederich and co-workers have developed difluoromethyl
ketones with selectivity for malarial asparatic proteases over
the human cathepsins D and E.29 However, these data must be
placed into context that difluoromethyl ketones are known to
bind to proteases,25,29 esterase,28 secretase,30 and other
targets;31,36 therefore, these novel GABAB agonists must also
be evaluated at other targets to further determine their
potential use in vivo. We will also need to replicate the
behavioral studies in order to increase our power to detect a
significant anxiolytic effect of compound ( )-10 and rac-BHFF.
Increasing the dosages of these two test compounds may also
enhance the anxiolytic effects in mice. The studies detailed here
open a new avenue to extend the structure−activity relation-
Figure 12. Mean ( SEM) startle amplitude measured as force in
grams in male and female HAP2 mice. Effects of drug treatments on
startle response amplitude during both the noise-alone and light +
noise trials are shown (see Experimental Section). (A) Mice were
treated by ip injection with vehicle or ( )-baclofen (7.5 mg/kg). (B)
Mice treated by ip injection with vehicle, rac-BHFF (12.5 mg/kg), or
compound ( )-10 (36 mg/kg). Initial analyses indicated no
interactions with sex, so data are reported collapsed across this factor.
* p < 0.05.
+ noise trials in the vehicle-treated group [F(1,5) = 16.5, p =
0.01; light + noise > noise-alone, indicating fear-potentiated
startle] but not in the ( )-baclofen-treated group (p = 0.6). In
experiment 2, mice were pretreated with either vehicle, rac-
BHFF (12.5 mg/kg ip), or compound ( )-10 (36 mg/kg ip)
25 min before the start of the test session. The dose for rac-
BHFF was obtained from prior studies,20 and a relative dose for
the novel agent ( )-10 was selected using the pharmacological
data (see Table 1) as well as its larger relative molecular weight
to baclofen. Figure 12B shows that both rac-BHFF and
compound ( )-10 reduced the startle response on both trial
types, with a greater reduction in the compound ( )-10 group,
suggesting anxiolytic-type effects from these compounds.
However, ANOVA indicated a main effect of Trial Type
[F(1,29) = 7.8, p < 0.01; light + noise > noise-alone] but no
treatment effect (p = 0.2) or interaction. It should also be noted
that a reduction in startle magnitude could also be related to
muscle-relaxant effects of these drugs, a possibility that will
require further testing. These in vivo data provide preliminary
evidence of bioavailability and biological activity for the
difluoromethyl ketone ( )-10 in a relevant rodent model.
Additional studies are planned to validate the distribution of the
F
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
0.337 mmol). After 20 min at −78 °C, a solution of 1-adamantyl
methyl ketone (50.0 mg, 0.280 mmol) in THF (1 mL) was added
dropwise, and the reaction mixture was stirred for 40 min at −78 °C.
Next, a solution of 4-acetylbenzaldehyde (50.0 mg, 0.337 mmol) in
THF (1 mL) was added dropwise, and the reaction mixture was stirred
for 20 min at −78 °C. Then the reaction mixture was quenched with
saturated aqueous NH4Cl (10 mL) at −78 °C. The mixture was
warmed to rt and extracted with CH2Cl2 (3 × 10 mL). The combined
organics were dried over Na2SO4 and concentration under reduced
pressure. SiO2 flash chromatography (8:2 hexanes/EtOAc) afforded
ships for potent agonist activity at the GABAB receptor, and this
novel pharmacophore provides numerous opportunities for
drug design in the modulation of muscle contractility and the
treatment of anxiety and addiction-related behavior. This work
also highlights the powerful synthetic approach of using
trifluoroacetate-release to assemble fluorinated structures for
medicinal chemistry.
EXPERIMENTAL SECTION
■
1
the product 15 as a colorless oil in 68% yield (62.5 mg). H NMR
Chemistry. The synthesis of compounds 1−12 was conducted
according to the literature procedure.26 rac-BHFF was obtained using
the reported synthesis.37 Preparation and characterization data for the
new compounds, (−)-10, (+)-10, 12, and 13 are listed below. HPLC
analysis was conducted to establish a level of ≥95% purity, and
additionally, combustion analysis is provided for (+)-10 and 12.
Separation of Compounds (−)-10 and (+)-10. To a solution of
(S)-Boc-proline (231 mg, 1.08 mmol) in CH2Cl2 (1 mL), CDI (174
mg, 1.08 mmol) was added and stirred for 5 min at rt. Then, the
reaction mixture was transferred to another vial containing 3-(4-
acetylphenyl)-1-(adamantan-1-yl)-2,2-difluoro-3-hydroxypropan-1-one
1026 (78 mg, 0.22 mmol) and stirred at 70 °C overnight. Next, the
reaction mixture was cooled to rt and quenched with saturated
aqueous NH4Cl (2 mL) at the same temperature, and the resultant
mixture was extracted with CH2Cl2 (6 mL × 5). The organics were
dried over Na2SO4 and concentrated under reduced pressure.
Preparatory TLC (SiO2, 4:2:2:1.5 hexanes:pentanes:petroleum ether:-
EtOAc) afforded the two pure diastereomers that were individually
treated with a solution of 2 M KOH (1 mL) in MeOH (2 mL). After
stirring for 2 h at rt, the reaction mixtures were quenched with
saturated aqueous NH4Cl (6 mL) and extracted with CH2Cl2 (6 mL ×
5). The organics were dried over Na2SO4 and concentrated under
reduced pressure. SiO2 flash chromatography (3:7 EtOAc:hexanes)
afforded (+)-10 in 14% yield (10.9 mg) as a solid. SiO2 flash
chromatography (3:7 EtOAc:hexanes) afforded (−)-10 in 14% yield
(10.5 mg) as a solid.
(500 MHz, CDCl3) δ 7.94 (d, J = 8.4 Hz, 2H), 7.45 (d, J = 8.2 Hz,
2H), 5.16 (dt, J = 8.6, 2.7 Hz, 1H), 3.78 (d, J = 3.0 Hz, 1H), 2.86 (dd,
J = 18.0, 3.3 Hz, 1H), 2.80 (dd, J = 18.0, 8.8 Hz, 1H), 2.59 (s, 3H),
2.03 (br s, 3H), 1.78−1.70 (m, 9H), 1.67−1.64 (m, 3H). 13C NMR
(125 MHz, CDCl3) δ 216.4, 197.8, 148.4, 136.3, 128.6 (2), 125.7 (2),
69.5, 46.6, 44.7, 37.9 (3), 36.3 (3), 27.7 (3), 26.6. IR (film) νmax 3468,
2905, 2850, 1682, 1268 cm−1. HRMS (EI) m/z calcd for C21H26O3
(M)+ 326.1882, found 326.1876.
Cell Culture for GABABR/Cre-Luc Cells. GABABR/Cre-Luc cells
were used between passages 2−26 and grown in media containing
DMEM F-12 (Invitrogen, Carlsbad, CA), 5.8 mM NaHCO3 (pH 7.0−
7.2), 100 units/mL penicillin, 100 μg/mL streptomycin, 10% fetal
bovine serum (FBS; Atlanta Biologicals, Lawrenceville, GA), 2 μg/mL
puromycin (Sigma, St. Louis, MO), and 200 μg/mL G418 (Sigma)
(growth medium).
Construction of HEK293 Cells Stably Expressing Human
GABAB1 and GABAB2 Receptor cDNAs, and the Creluc Reporter
(GABABR/Cre-Luc Cells). The human GABAB1 receptor cDNA
(GenBank BC050532) and the human GABAB2 receptor cDNA
(GenBank BC035071) were subcloned into the first and second
multiple cloning site of the pIRES plasmid vector (ClonTech,
Mountain View, CA), respectively. This vector (GABA-B1/B2/
pIRES) allows both proteins to be translated from a single mRNA,
and its integrity was verified by DNA sequencing. GABA-B1/B2/
pIRES was linearized with BstBI between the NeoR and AmpR regions.
Linearized GABA-B1/B2/pIRES was then transfected into HEK293/
Cre-luc cells47 in growth medium without G418, using calcium
phosphate transfection.48 Twenty-four hours post transfection,
selection medium (growth medium with 400 μg/mL G418) was
added to plates, and the cells were incubated at 37 °C in 5% CO2.
After several days in selection medium, colonies were picked and
amplified in selection medium. G418-resistant clonal cell lines were
screened for the expression of the GABAB receptor subunits by rtPCR,
Western blot, and by an assay to detect baclofen inhibition of
forskolin-stimulated cAMP accumulation using the Cre-Luc reporter
system (see below). For the rtPCR assay, total RNA was isolated from
several clones using TRIzol (Invitrogen, Carlsbad, CA) according to
the manufacturer’s instructions, and cDNA was prepared using
Superscript II reverse transcriptase and Taq DNA polymerase
(Invitrogen, Carlsbad, CA). The primers used to amplify the
G A B A B 1 r e c e p t o r c D N A w e r e : F o r w a r d -
GGCCCGGGGCCCCTTTTGCC; Reverse- GGATCA-
CACTTGCTGTCGTGG. The primers used to amplify the GABAB2
receptor cDNA were: Forward- GGAGCAGGTGCACACG-
GAAGCC; Reverse- GGAATACAACTTGACCCGTGACC. After
amplification with Taq polymerase, DNA products of the predicted
size (268 bp for GABAB1 receptor and 338 bp for GABAB2 receptor)
were observed upon agarose gel electrophoresis and staining with
ethidium bromide. Western blotting was performed using an antibody
against the GABAB1 receptor (Santa Cruz Biochemicals, Santa Cruz,
CA) and an antibody against the GABAB2 receptor (Alomone
Laboratories, Jerusalem, Israel). Detection with the appropriate
secondary antibodies revealed specific bands of ∼80 kD, correspond-
ing to the GABAB1 receptor, and ∼130 kD, corresponding to the
GABAB2 receptor. Once the stable GABABR/Cre-Luc cell line was
confirmed, a z-score of 0.61 was determined for the inhibition of
adenylyl cyclase assay using 10 μM forskolin for the high control, and
10 μM forskolin (Tocris, Bristol, UK) + 10 μM ( )-baclofen (Sigma,
St. Louis, MO) for the low control.
(+)-3-(4-Acetylphenyl)-1-(adamantan-1-yl)-2,2-difluoro-3-
hydroxypropan-1-one (10). [α]23 +36.0 (c 0.667, CHCl3). Anal.
D
Calcd for C21H24F2O3: C, 69.60; H, 6.67. Found: C, 66.45; H, 6.51.
Other characterization data is identical to the data reported for
compound 10.26
(−)-3-(4-Acetylphenyl)-1-(adamantan-1-yl)-2,2-difluoro-3-
hydroxypropan-1-one (10). [α]23 −35.8 (c 0.558, CHCl3). Other
D
characterization data is identical to the data reported for compound
10.26
3-(4-Acetylphenyl)-1-(adamantan-1-yl)-2,2-difluoro-3-me-
thoxypropan-1-one (14). A solution of 10 (9.8 mg, 0.028 mmol)
and CH3I (35 μL, 0.56 mmol) in CH2Cl2 (0.5 mL) was treated with
Ag2O (42 mg, 0.18 mmol). After 6 h, CH3I (50 μL, 0.8 mmol) was
added to reaction. After 16 h, additional CH3I, (100 μL, 1.6 mmol)
was added. After 20 h, reaction was diluted with MeOH (20 mL) and
filtered through Celite. The residue was concentrated under reduced
pressure, suspended in CH2Cl2 (5 mL), and then washed with water
(5 mL). The aqueous layer was separated and extracted with CH2Cl2
(2 × 5 mL). Combined organics were concentrated under reduced
pressure to give 14 as a colorless solid (7.2 mg) in 71% yield: mp 85−
87 °C. 1H NMR (500 MHz, CDCl3) δ 7.98 (d, J = 8.3 Hz, 2H), 7.52
(d, J = 8.1 Hz, 2H), 4.86 (dd, J = 19.8, 5.2 Hz, 1H), 3.28 (s, 3H), 2.62
(s, 3H), 2.04 (br s, 3H), 1.94 (m, 6H), 1.72 (m, 6H). 13C NMR (125
MHz, CDCl3) δ 204.4 (dd, JCF = 29.8, 22.2 Hz, 1C), 197.8, 138.5,
137.5, 129.0 (2), 128.2 (2), 116.5 (dd, JCF = 269, 253 Hz, 1C), 81.1
(dd, JCF = 30.1, 21.2 Hz, 1C), 58.0, 46.6, 36.7 (3), 36.4 (3), 27.6 (3),
26.7. 19F NMR (282 MHz, CDCl3) δ −102.5 (dd, JFF = 278, JHF = 4.3
Hz, 1F), −123.4 (dd, JFF = 278, JHF = 19.9 Hz, 1F). HRMS (APCI) m/
z calcd for C22H26O3F2 (M+H)+ 377.1928, found 377.1930. Anal.
Calcd for C22H26O3F2·0.5CH2Cl2: C, 64.51; H, 6.50. Found C, 67.02;
H, 6.45.
3-(4-Acetylphenyl)-1-(-adamantan-1-yl)-3-hydroxypropan-
1-one (15). To a −78 °C solution of n-BuLi (0.14 mL, 2.5 M in
hexanes) in THF (3 mL) was added hexamethyldisilazane (71 μL,
G
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Adenylyl Cyclase Inhibition Assay. GABABR/Cre-Luc cells were
plated at approximately 65000 cells per well in black-walled 96-well
tissue culture plates (VWR, West Chester, PA) and incubated
overnight at 37 °C and 5% CO2. Treatments were diluted to 2× the
working concentration in Optimem I reduced-serum medium
(Invitrogen, Carlsbad, CA) supplemented with 0.02% ascorbic acid
(Sigma, St. Louis, MO). Cell culture medium was decanted from cells
and equal amounts of the Optimem buffer and 2× treatments were
added to cells. Forskolin (10 μM) was added to cells to stimulate
cAMP accumulation, and vehicle or 10 μM ( )-baclofen were added
along with forskolin to establish the maximum cAMP accumulation,
and the portion of the response that could be inhibited by activation of
the GABAB receptor, respectively. A basal control (no forskolin or
baclofen) was included with each assay. The indicated concentrations
of test compounds, ( )-baclofen, R-(−)-baclofen (Tocris, Briston,
UK), or S-(+)-baclofen (Sigma, St. Louis, MO) or γ-aminobutyric acid
(GABA) (Sigma, St. Louis, MO) were added to cells along with 10 μM
of forskolin. Reversal of GABAB receptor-mediated inhibition of
adenylyl cyclase activity was demonstrated with the GABAB receptor
antagonist CGP 55845 (10 μM) (Tocris, Bristol, UK). Each condition
was performed in triplicate. Cells were incubated with the indicated
treatments at 37 °C and 5% CO2 for 3 h. After the incubation period,
the treatments were decanted, and assay plates were allowed to cool to
room temperature for 10 min. While protecting the samples from light,
30 μL of phosphate buffered saline (Sigma, St. Louis, MO) and 30 μL
Steady Lite Plus (PerkinElmer, Waltham, MA) were added to each
well, and the plates were shaken at 500 rpm for 5 min. Luminescence
was quantified with a Victor light luminescence reader (PerkinElmer,
Waltham, MA).
was used. Whole-cell currents were recorded at a sampling rate of 1
kHz. The current response in the presence of 100 μM GABA allowed
for comparing the activity of the synthesized compounds at this
GABAA receptor type.
In Vivo Experiments. Subjects. HAP2 mice were used as subjects.
For experiment 1 [( )-baclofen], mice (7 males, 7 females) were
between 133 and 142 days old. These mice were part of a larger
experimental design and also had previously been exposed to foot
shock/acoustic startle testing (same parameters used here), testing for
prepulse inhibition of the acoustic startle response, and two injections
(separated by 5−7 days) of either vehicle, 5.0 or 7.5 mg/kg
( )-baclofen. Following a washout period of 15 days, mice were
counterbalanced into treatment groups so that previous exposure to
( )-baclofen or vehicle was distributed across treatment groups for
experiment 1 (i.e., vehicle vs 7.5 mg/kg baclofen); baclofen’s startle
reducing effect shown in Figure 12 was similar regardless of prior drug
history. For experiment 2 (rac-BHFF/compound ( )-10), exper-
̈
imentally naive mice (16 males, 16 females) were between 316 and
332 days old at the time of testing. Mice were housed in polycarbonate
cages (29.2 cm × 19.0 cm × 12.7 cm) with aspen wood shavings in
groups of 2−4 per cage. Ambient room temperature was maintained at
21
2 °C. Mice had free access to food (Rodent Lab Diet 5001,
Purina Mills Inc., St Louis, MO) and water in the home cage at all
times, except when testing procedures took place. Experimental
procedures were conducted during the light phase of a 12:12 light:dark
cycle (lights off at 19:00). All experimental procedures were approved
by the Purdue Animal Care and Use Committee and were conducted
in accordance with the Guide for the Care and Use of Laboratory
Animals.
Cytotoxicity Assay. GABABR/Cre-Luc Cells were plated at
approximately 65000 cells per well in black-walled tissue culture
plates (VWR, West Chester, PA), incubated overnight at 37 °C and
5% CO2. Treatments were diluted to 2× the working concentration in
Optimem I reduced-serum medium (Invitrogen, Carlsbad, CA). Cell
culture medium was decanted from cells, and equal amounts of the
Optimem and 2× treatments were added to cells. Each condition was
performed in triplicate. Cells were incubated at 37 °C and 5% CO2 for
3 h with the indicated treatments, and then 20 μL of CellTiter 96
Aqueous One Solution Reagent (Promega, Madison, WI) was added
to all of the wells. The plates were protected from light and incubated
for 1.5 h at 37 °C and 5% CO2, and then absorption at 490 nm was
quantified with a Synergy 4 plate reader (BioTek, Winooski, VT).
GABAA Assay. tsA201 cells, derived from human embryonic
kidney (HEK) cell line, were maintained in Dulbecco’s Modified
Eagle’s Medium (DMEM; Life Technologies, Grand Island, NY)
supplemented with 10% v/v FBS, 100 units/mL penicillin G, and 100
μg/mL streptomycin (Sigma-Aldrich, St. Louis, MO) and were
maintained at 37 °C in 95% air/5% CO2. tsA201 cells about 50−
70% confluent in a 35 mm culture dish were transfected with the rat
GABAA receptor subunits α1, β2, and γ2s (in the plasmid vector
pCDNA3.1; Invitrogen, Carlsbad, CA) using the calcium phosphate
method.4 Rat GABAA receptor subunit plasmid cDNAs were
transfected in a 1:1:5:1 ratio for α1, β2, γ2s, and GFP, respectively.
The cell medium was replaced with fresh cell medium about 5−7 h
after transient transfections. After another 20−24 h, transfected cells
were trypsinized and plated in 35 mm dishes for electrophysiology
recordings the next day.
Voltage clamp experiments were performed at rt using the standard
whole-cell configuration. Current responses were recorded with an
Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA) and
filtered at 1 kHz (six-pole Bessel filter, −3 dB). Electrodes were pulled
from borosilicate glass (VWR, West Chester, PA) and fire-polished to
resistances of 2−5 MΩ when filled with intracellular solution
containing 147 mM N-methyl-D-glucamine hydrochloride, 5 mM
CsCl, 5 mM Na2ATP, 5 mM HEPES, 1 mM MgCl2, 0.1 mM CaCl2,
and 1.1 mM EGTA at pH 7.2 and an osmolality of 310 mOsm. The
extracellular recording solution contained 145 mM NaCl, 3 mM KCl,
1.5 mM CaCl2, 1 mM MgCl2, 5.5 mM D-glucose, 10 mM HEPES at
pH 7.4, and an osmolality of 315 mOsm. The standard holding
potential for the cells was −60 mV, and a gap-free recording protocol
Drugs. All drugs were administered ip using a 10 mL/kg injection
volume. ( )-Baclofen was dissolved in distilled water and adminis-
tered at a dose of 7.5 mg/kg. rac-BHFF and compound ( )-10 were
dissolved in 10% DMSO/45% (w/v) 2-hydroxypropyl-β-cyclodextrin
in distilled water and administered at a doses of 12.5 and 36.16 mg/kg,
respectively. The dose for ( )-baclofen45 and rac-BHFF was obtained
from prior studies,20 and a relative dose for the novel agent ( )-10
was selected using the pharmacological data (see Table 1).
Apparatus. The acoustic startle response was assessed using sound-
attenuated Coulbourn Instruments (Allentown, PA, USA) Animal
Acoustic Startle System chambers, as previously described.43 Startle
stimuli consisted of 100 dB, 40 ms white noise bursts (frequency
range: 20 Hz−20 kHz). The startle response was measured as the
amount of force in grams exerted against a weight-sensitive platform
during the 200 ms after the onset of each acoustic stimulus. The force
measurement does not include the subject’s body weight. A ventilating
fan provided continuous 70−71 dB background noise.
Acoustic Startle Procedures. Mice received one conditioning and
one test session separated by 24 h. During each conditioning session,
mice received 40 trials of a 30 s, 7 W light stimulus paired with a 0.5 s,
0.8 mA foot shock (2 min intertrial interval). The foot shock occurred
during the last 0.5 s of the light stimulus presentation. The test session
consisted of a 5 min habituation period followed by 36 total trials (2
min ITI) presented on a random schedule (range: 12−108 s) to
reduce habituation to any single trial type. Twelve of the trials were
blank (no stimuli; 40 ms), 12 were noise alone (100 dB, 40 ms), and
12 were light (7 W, 30 s) + noise (100 dB, 40 ms). On light + noise
trials, the noise stimulus was presented immediately after the light
stimulus ended.
Statistical Analyses. All 12 startle responses on each trial type were
averaged for each mouse. Data were analyzed using repeated measures
analysis of variance (ANOVA) followed by posthoc one-way
ANOVAs. Probability values <0.05 were considered to be significant.
ASSOCIATED CONTENT
■
S
* Supporting Information
X-ray data of compound 10 and crystallographic data (CIF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
H
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(13) 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
Analogues of GABA. 1. New Potent and Selective GABAB Agonists. J.
Med. Chem. 1995, 38, 3297−3312.
(14) Pirard, B.; Carrupt, P.-A.; Testa, B.; Tsai, R.-S.; Berthelot, P.;
Vaccher, C.; Debaert, M.; Durant, F. Structure−Affinity Relationships
of Baclofen and 3-Heteroaromatic Analogues. Bioorg. Med. Chem.
1995, 3, 1537−1545.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: (765) 496-3962. Fax: (765) 494-1414. E-mail:
dcolby@purdue.edu. Address: Purdue University, 575 Stadium
Mall Drive, West Lafayette, IN 47907.
Notes
The authors declare no competing financial interest.
(15) Hinton, T.; Chebib, M.; Johnston, G. A. R. Enantioselective
Actions of 4-Amino-3-hydroxybutanoic Acid and (3-Amino-2-
hydroxypropyl)methylphosphinic Acid at Recombinant GABAC
Receptors. Bioorg. Med. Chem. Lett. 2008, 18, 402−404.
(16) Xu, F.; Peng, G.; Phan, T.; Dilip, U.; Chen, J. L.; Chernov-
Rogan, T.; Zhang, X.; Grindstaff, K.; Annamalai, T.; Koller, K.; Gallop,
M. A.; Wustrow, D. J. Discovery of a Novel Potent GABAB Receptor
Agonist. Bioorg. Med. Chem. Lett. 2011, 21, 6582−6585.
ACKNOWLEDGMENTS
■
We thank Purdue University and the Ralph W. and Grace M.
Showalter Research Trust for funding these studies. C.H. was
the recipient of a McKeehan Graduate Student Assistantship.
We acknowledge Philip E. Fanwick and the X-ray Crystallog-
raphy Center at Purdue University. Also, Gustavo D. Barrenha
and Katie V. Smith are thanked for assistance with the in vivo
studies and Rachel J. Bubik is thanked for technical assistance
with the pharmacology. We thank Dr. Erwin Sigel, University of
Bern, Switzerland, for the gift of cDNAs expressing the rat
GABAA receptor subunits.
(17) Froestl, W.; Mickel, S. J.; von Sprecher, G.; Diel, P. J.; Hall, R.
G.; Maier, L.; Strub, D.; Melillo, V.; Baumann, P. A. Phosphinic Acid
Analogs of GABA. 2. Selective, Orally Active GABAB Antagonists. J.
Med. Chem. 1995, 38, 3313−3331.
(18) Alstermark, C.; Amin, K.; Dinn, S. R.; Elebring, T.; Fjellstrom,
̈
O.; Fitzpatrick, K.; Geiss, W. B.; Gottfries, J.; Guzzo, P. R.; Harding, J.
P.; Holmen
́
, A.; Kothare, M.; Lehmann, A.; Mattsson, J. P.; Nilsson,
ABBREVIATIONS USED
K.; Sunden, G.; Swanson, M.; von Unge, S.; Woo, A. M.; Wyle, M. J.;
́
■
Zheng, X. Synthesis and Pharmacological Evaluation of Novel γ-
Aminobutyric Acid Type B (GABAB) Receptor Agonists as Gastro-
esophageal Reflux Inhibitors. J. Med. Chem. 2008, 51, 4315−4320.
(19) Guery, S.; Floersheim, P.; Kaupmann, K.; Froestl, W. Syntheses
and Optimization of New GS39783 Analogues as Positive Allosteric
Modulators of GABAB Receptors. Bioorg. Med. Chem. Lett. 2007, 17,
6206−6211.
(20) Malherbe, P.; Masciadri, R.; Norcross, R. D.; Knoflach, F.;
Kratzeisen, C.; Zenner, M. T.; Kolb, Y.; Marcuz, A.; Huwyler, J.;
Nakagawa, T.; Porter, R. H. P.; Thomas, A. W.; Wettstein, J. G.;
Sleight, A. J.; Spooren, W.; Prinssen, E. P. Characterization of (R,S)-
5,7-Di-tert-butyl-3-hydroxy-3-trifluoromethyl-3H-benzofuran-2-one as
a Positive Allosteric Modulator of GABAB Receptors. Br. J. Pharmacol.
2008, 154, 797−811.
(21) Galvez, T.; Prezeau, L.; Milioti, G.; Franek, M.; Joly, C.; Froestl,
W.; Bettler, B.; Bertrand, H.-O.; Blahos, J.; Pin, J.-P. Mapping the
Agonist-binding Site of GABAB Type 1 Subunit Sheds Light on the
Activation Process of GABAB Receptors. J. Biol. Chem. 2000, 275,
41166−41174.
(22) Jauhiainen, M.; Yuan, W.; Gelb, M. H.; Dolphin, P. J. Human
Plasma Lecithin-Cholesterol Acyltransferase. Inhibition of the
Phospholipase A2-Like Activity by sn-2-Difluoroketone Phosphatidyl-
choline Analogues. J. Biol. Chem. 1989, 264, 1963−1987.
(23) Gelb, M. H.; Svaren, J. P.; Abeles, R. H. Fluoro Ketone
Inhibitors of Hydrolytic Enzymes. Biochemistry 1985, 24, 1813−1817.
(24) Baskakis, C.; Magrioti, V.; Cotton, N.; Stephens, D.;
Constantinou-Kokotou, V.; Dennis, E. A.; Kokotos, G. Synthesis of
Polyfluoro Ketones for Selective Inhibition of Human Phospholipase
A2 Enzymes. J. Med. Chem. 2008, 51, 8027−8037.
CRE-Luc, CRE-luciferase; GERD, gastroesophageal reflux;
HAP, high-alcohol preferring; HEPES, 4-(2-hydroxyethyl)-
piperazine-1-ethanesulfonic acid; ip, intraperitoneal; Osm,
osmolality; rac-BHFF, (R,S)-5,7-di-tert-butyl-3-hydroxy-3-tri-
fluoromethyl-3H-benzofuran-2-one
REFERENCES
■
(1) Olsen, R. W.; Sieghart, W. International Union of Pharmacology.
LXX. Subtypes of γ-Aminobutyric AcidA Receptors: Classification on
the Basis of Subunit Composition, Pharmacology, and Function.
Update. Pharmacol. Rev. 2008, 60, 243−260.
(2) Hanson, S. M.; Morlock, E. V.; Satyshur, K. A.; Czajkowski, C.
Structural Requirements for Eszopiclone and Zolpidem Binding to the
γ-Aminobutyric Acid Type-A (GABAA) Receptor Are Different. J. Med.
Chem. 2008, 51, 7243−7252.
(3) Ulrich, D.; Bettler, B. GABAB Receptors: Synaptic Functions and
Mechanisms of Diversity. Curr. Opin. Neurobiol. 2007, 17, 298−303.
(4) Bowery, N. G. Historical Perspective and Emergence of the
GABAB Receptor. Adv. Pharmacol. 2010, 58, 1−18.
(5) Fowler, C. E.; Aryal, P.; Suen, K. F.; Slesinger, P. A. Evidence for
Association of GABAB Receptors with Kir3 Channels and Regulators
of G Protein Signalling (RGS4) Proteins. J. Physiol. 2007, 580, 51−65.
(6) Takahashi, T.; Kajikawa, Y.; Tsujimoto, T. G-Protein-Coupled
Modulation of Presynaptic Calcium Currents and Transmitter Release
by a GABAB Receptor. J. Neurosci. 1998, 18, 3138−3146.
(7) Chen, G.; van den Pol, A. N. Presynaptic GABAB Autoreceptor
Modulation of P/Q-Type Calcium Channels and GABA Release in Rat
Suprachiasmatic Nucleus Neurons. J. Neurosci. 1998, 18, 1913−1922.
(8) Nakayasu, H.; Nishikawa, M.; Mizutani, H.; Kimura, H.;
Kuriyama, K. Immunoaffinity Purification and Characterization of
Gamma-aminobutyric acid (GABA)B Receptor from Bovine Cerebral
Cortex. J. Biol. Chem. 1993, 268, 8658−8664.
(9) Froestl, W. Chemistry and Pharmacology of GABAB Receptor
Ligands. Adv. Pharmacol. 2010, 58, 19−62.
(10) Tyacke, R. J.; Lingford-Hughes, A.; Reed, L. J.; Nutt, D. J.
GABAB Receptors in Addiction and Its Treatment. Adv. Pharmacol.
2010, 58, 373−396.
(11) Vlachou, S.; Markou, A. GABAB Receptors in Reward Processes.
Adv. Pharmacol. 2010, 58, 315−371.
(12) Hijzen, T. H.; Houtzager, S. W. J.; Joordens, R. J. E.; Olivier, B.;
Slangen, J. L. Predictive Validity of the Potentiated Startle Response as
a Behavioral Model for Anxiolytic Drugs. Psychopharmacology 1995,
118, 150−154.
(25) Reiter, L. A.; Martinelli, G. J.; Reeves, L. A.; Mitchell, P. G.
Difluoroketones as Inhibitors of Matrix Metalloprotease-13. Bioorg.
Med. Chem. Lett. 2000, 10, 1581−1584.
(26) Han, C.; Kim, E. H.; Colby, D. A. Cleavage of Carbon−Carbon
Bonds through the Mild Release of Trifluoroacetate: Generation of
α,α-Difluoroenolates for Aldol Reactions. J. Am. Chem. Soc. 2011, 133,
5802−5805.
(27) Muller, K.; Faeh, C.; Diederich, F. Fluorine in Pharmaceuticals:
̈
Looking beyond Intuition. Science 2007, 317, 1881−1886.
́
(28) Quero, C.; Rosell, G.; Jimenez, O.; Rodriguez, S.; Bosch, M. P.;
Guerrero, A. New Fluorinated Derivatives as Esterase Inhibitors.
Synthesis, Hydration and Crossed Specificity Studies. Bioorg. Med.
Chem. 2003, 11, 1047−1055.
(29) Fah, C.; Hardegger, L. A.; Baitsch, L.; Schweizer, W. B.; Meyer,
̈
S.; Bur, D.; Diederich, F. New Organofluorine Building Blocks:
I
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Inhibition of the Malarial Aspartic Proteases Plasmepsin II and IV by
Alicyclic α,α-Difluoroketone Hydrates. Org. Biomol. Chem. 2009, 7,
3947−3957.
(30) Moore, C. L.; Leatherwood, D. D.; Diehl, T. S.; Selkoe, D. J.;
Wolfe, M. S. Difluoro Ketone Peptidomimetics Suggest a Large S1
Pocket for Alzheimer’s γ-Secretase: Implications for Inhibitor Design.
J. Med. Chem. 2000, 43, 3434−3442.
Comparison of the Enantiomers of ( )-Doxanthrine, A High Efficacy
Full Dopamine D1 Receptor Agonist, and a Reversal of Enantiose-
lectivity at D1 versus alpha2C Adrenergic Receptors. Eur. Neuro-
psychopharmacol. 2009, 19, 138−146.
(48) Margolskee, R. F.; McHendry-Rinde, B.; Horn, R. Panning
Transfected Cells for Electrophysiological Studies. Biotechniques 1993,
15, 906−911.
(31) Doherty, A. M.; Sircar, I.; Kornberg, B. E.; Quin, J., III; Winters,
R. T.; Kaltenbronn, J. S.; Taylor, M. D.; Batley, B. L.; Rapundalo, S. R.;
Ryan, M. J.; Painchaud, C. A. Design and Synthesis of Potent,
Selective, and Orally Active Fluorine-Containing Renin Inhibitors. J.
Med. Chem. 1992, 35, 2−14.
(32) Prager, J. H.; Ogden, P. H. Metal Derivatives of Fluorinated
gem-Diols. J. Org. Chem. 1968, 33, 2100−2102.
(33) John, J. P.; Colby, D. A. Synthesis of α-Halo-α,α-difluoromethyl
Ketones by a Trifluoroacetate Release/Halogenation Protocol. J. Org.
Chem. 2011, 76, 9163−9168.
(34) Balnaves, A. S.; Gelbrich, T.; Hursthouse, M. B.; Light, M. E.;
Palmer, M. J.; Percy, J. M. The Generation of Difluoroenolates from
Trifluoroethanol and Reproducible Syntheses of α,α-Difluoro-β-
hydroxy Ketones. J. Chem. Soc., Perkin Trans. 1 1999, 2525−2535.
(35) Chang, C.-H.; Yang, D. S. C.; Yoo, C. S.; Wang, B.-C.; Pletcher,
J.; Sax, M.; Terrence, C. F. Structure and Absolute Configuration of
(R)-Baclofen Monohydrochloride. Acta Crystallogr., Sect. B: Struct.
Crystallogr. Cryst. Chem. 1982, 38, 2065−2067.
(36) Neder, K. M.; French, S. A.; Miller, S. P. F. Synthesis and
Inhibitory Activity of Difluoroketone Substrate Analogs of N-
Myristoyltransferase. Tetrahedron 1994, 50, 9847−9864.
(37) Alker, A. M.; Grillet, F.; Malherbe, P.; Norcross, R. D.; Thomas,
A. W.; Mascidri, R. Efficient One-Pot Synthesis of the GABAB Positive
Allosteric Modulator (R,S)-5,7-Di-tert-butyl-3-hydroxy-3-trifluoro-
methyl-3H-benzofuran-2-one. Synth. Commun. 2008, 38, 3398−3405.
(38) Kumar, R. J.; Chebib, M.; Hibbs, D. E.; Kim, H.-L.; Johnston, G.
A. R.; Salam, N. K.; Hanrahan, J. R. Novel γ-Aminobutyric Acid ρ1
Receptor Antagonists; Synthesis, Pharmacological Activity and
Structure−Activity Relationships. J. Med. Chem. 2008, 51, 3825−3840.
(39) Brugger, F.; Wicki, U.; Olpe, H.-R.; Froestl, W.; Mickel, S. The
Action of New Potent GABAB Receptor Antagonists in the
Hemisected Spinal Cord Preparation of the Rat. Eur. J. Pharmacol.
1993, 235, 153−155.
(40) Grillon, C. Models and Mechanisms of Anxiety: Evidence from
Startle Studies. Psychopharmacology 2008, 199, 421−437.
(41) Kaplan, G. B.; Heinrichs, S. C.; Carey, R. J. Treatment of
Addiction and Anxiety using Extinction Approaches: Neural
Mechanisms and Their Treatment Implications. Pharmacol., Biochem.
Behav. 2011, 97, 619−625.
(42) Kessler, R. C.; Crum, R. M.; Warner, L. A.; Nelson, C. B.;
Schulenberg, J.; Anthony, J. C. Lifetime Co-Occurrence of DSM-III-R
Alcohol Abuse and Dependence with Other Psychiatric Disorders in
the National Comorbidity Survey. Arch. Gen. Psychiatry 1997, 54,
313−321.
(43) Chester, J. A.; Barrenha, G. D. Acoustic Startle at Baseline and
During Acute Alcohol Withdrawal in Replicate Mouse Lines
Selectively Bred for High or Low Alcohol Preference. Alcohol.: Clin.
Exp. Res. 2007, 31, 1633−1644.
(44) Barrenha, G. D.; Coon, L. E.; Chester, J. A. Effects of Alcohol on
the Acquisition and Expression of Fear-Potentiated Startle in Mouse
Lines Selectively Bred for High and Low Alcohol Preference.
Psychopharmacology 2011, 218, 191−201.
(45) Chester, J. A.; Cunningham, C. L. Baclofen Alters Ethanol-
Stimulated Activity but not Conditioned Place Preference or Taste
Aversion in Mice. Pharmacol., Biochem. Behav. 1999, 63, 325−331.
(46) Kalyvas, A.; Baskakis, C.; Magrioti, V.; Constantinou-Kokotou,
V.; Stephens, D.; Lopez-Vales, R.; Lu, J.-Q.; Yong, V. W.; Dennis, E.
A.; Kokotos, G.; David, S. Differing Roles for Members of the
Phospholipase A2 Superfamily in Experimental Autoimmune
Encephalomyelitis. Brain 2009, 132, 1221−1235.
(47) Przybyla, J. A.; Cueva, J. P.; Chemel, B. R.; Hsu, K. J.; Riese, D.
J.; McCorvy, J. D.; Chester, J. A.; Nichols, D. E.; Watts, V. J.
J
dx.doi.org/10.1021/jm301805e | J. Med. Chem. XXXX, XXX, XXX−XXX