Heteroaromatic side-chain analogs of pregabalin

Article abstract of DOI:10.1016/j.bmcl.2005.06.090

A series of heteroaromatic analogs of pregabalin has been identified that possess anticonvulsant activity in the DBA/2 mouse model. The methods of synthesis and preliminary pharmacology are discussed herein.

Full text of DOI:10.1016/j.bmcl.2005.06.090

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2329–2332
Heteroaromatic side-chain analogs of pregabalin
Robert M. Schelkun, Po-wai Yuen,* David J. Wustrow, Jack Kinsora,
Ti-Zhi Su and Mark G. Vartanian
PGRD Michigan Laboratories, Ann Arbor, MI 48105, USA
Received 16 February 2005; revised 27 June 2005; accepted 29 June 2005
Available online 15 August 2005
Dedicated to the memory of David W. Robertson in recognition of his outstanding accomplishments in drug discovery and research.
Abstract—A series of heteroaromatic analogs of pregabalin has been identified that possess anticonvulsant activity in the DBA/2
mouse model. The methods of synthesis and preliminary pharmacology are discussed herein.
Ó 2005 Elsevier Ltd. All rights reserved.
The c-amino acid pregabalin (1) (Fig. 1) is known to
bind to the gabapentin binding site on the a2-d subunit
of voltage sensitive calcium channels in the CNS.1 Bind-
ing to the a2-d protein by pregabalin is proposed to
modulate the functional effects these proteins have on
calcium currents in activated neurons2 and on stimu-
lated neurotransmitter release.3 Reduced release of
excitatory neurotransmitters and peptide neuromodula-
tors (especially under conditions of hyperexcitability) is
thought to account for the anticonvulsant, anxiolytic,
and analgesic effects of these compounds. Pregabalin
has been shown to reduce the frequency of partial sei-
zures, reduce pain from post-herpetic neuralgia4 and
diabetic peripheral neuropathy,5 and reduce symptoms
of generalized anxiety disorder6 in placebo-controlled
studies. Sustained efficacy in these indications was
associated with a favorable safety profile, with the most
common side effects including dizziness and
somnolence.7
CO2H
NH2
CO2H
NH2
Ar = furan,
thiophene
Ar
1
2
Figure 1.
phosphono acetate or triethylphosphono acetate
(1.1 equiv) and NaH in THF at 0 °C, gave the acrylic
acid ester as only the (E) isomer (5a, b, and d). The dou-
ble bond was reduced with H2 (50 psi) using Wilkinsonꢀs
catalyst in THF at 45 °C in the case of the furans (5a
and b), and 20%Pd/C in methanol for the 3-thienyl com-
pound (5d), giving 6a, 6b, and 6d. The resulting saturat-
ed ester (6a, 6b, 6d) was alkylated with tert-butyl
bromoacetate and LDA in THF at À78 °C. The methyl
or ethyl ester in 7a, 7b, and 7d could be selectively
hydrolyzed using 1 N LiOH in THF/i-PrOH. The car-
boxylic acid in 8a, 8b, and 8d was reduced to the alcohol
using borane dimethyl sulfide complex in THF at 0 °C.
The tosylate was obtained from the alcohol by treatment
with para-toluenesulfonyl chloride and triethyl amine in
CH2Cl2 at 0 °C. Conversion to the azide (9a, 9b, and 9d)
was carried out using sodium azide in DMSO at 60 °C.
For the furan analogs (9a and 9b), the azide was selec-
tively hydrogenated to the amine using Pd/CaCO3 in
EtOAc8 prior to hydrolyzing the tert-butyl ester using
TFA/CH2Cl2. The free base was isolated using ion
exchange chromatography to give compounds 11a and
11b. For the 3-thienyl analog (9d), the tert-butyl ester
was hydrolyzed first, followed by hydrogenation of the
azide to the amine using 10%Pd/C in THF. The amino
acid (10d) was recrystallized from methanol/EtOAc.
As part of a program to identify the scope of substitu-
ents recognized by the a2-d protein, we were interested
in replacing the isobutyl side chain of pregabalin with
other groups such as aromatic or heteroaromatic rings.
The synthesis of a series of heteroaromatic-substituted
analogs of pregabalin is depicted in Scheme 1. Starting
with either 2 or 3-furaldehyde (4a and 4b), or 3-thiophene
carboxaldehyde (4d), and treating with either trimethyl-
*
Corresponding author. Tel.: +1 734 622 1979; fax: +1 734 622
5165; e-mail: po-wai.yuen@pfizer.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.06.090

2330
R. M. Schelkun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2329–2332
CO2Me
a
b
CO2Me
CHO
Ar
Ar
Ar
4a: 2-fural
b: 3-fural
d: 3-thienyl
5a,b,d
6a: 2-fural
b: 3-fural
d: 3-thienyl
CO2H
CO2Me
c, d
e
6a,b,d
Ar
Ar
CO2t-Bu
CO2t-Bu
7a,b,d
8a,b,d
f, g, h
i, j
CO2H
NH2
8a,b,d
Ar
N3
S
CO2t-Bu
9a,b,d
10d
k,l
m
NH2
CO2H
NH2
9a, b
O
O
CO2H
11a,b
12
Scheme 1. Reagents and conditions: (a) trimethylphosphono acetate, NaH, THF, 100%; (b) Wilkinsonꢀs catalyst, H2, THF, 45 °C, 97%, or 20% Pd/
C/H2, MeOH, 89%; (c) LDA, THF, À78 °C; (d) tert-butyl bromoacetate, 44–54%; (e) LiOH, H2O, THF, i-PrOH, 90%; (f) BH3ÆSMe2, THF, 0 °C, 55–
80%; (g) TsCl, TEA, CH2Cl2, 90%; (h) NaN3, DMSO, 60 °C, 85–95%; (i) TFA, CH2Cl2, 95%; (j) 10% Pd/C/H2, MeOH, 60–70%; (k) 10% Pd/CaCO3,
EtOAc, 77%; (l) TFA, CH2Cl2, 100%; (m) 20% Pd/C/H2, MeOH.
CO2H
CO2Me c, d
The synthesis of the 2-thienyl analog (10c) is depicted in
Scheme 2. Starting with commercially available 3-(2-thie-
nyl)acrylic acid, the double bound was hydrogenated
using 20%Pd/C in methanol and H2 (50 psi). The acid
was esterified using methyl chloroformate (1 equiv), tri-
ethylamine (1.1 equiv), and catalytic DMAP in CH2Cl2
at 0 °C to give the ester (6c). The saturated ester (6c)
was alkylated with tert-butyl bromoacetate and LDA in
THF at À78 °C. The methyl ester could be selectively
hydrolyzed using 1 N LiOH in THF/i-PrOH. The carbox-
ylic acid (8c) was reduced to the alcohol using borane
dimethyl sulfide complex in THF at 0 °C. The tosylate
was obtained from the alcohol by treatment with para-tol-
uenesulfonyl chloride and triethyl amine in CH2Cl2 at
0 °C. Conversion to the azide (9c) was carried out using
sodium azide in DMSO at 60 °C. The tert-butyl ester
was hydrolyzed using formic acid, followed by hydroge-
nation of the azide to the amine using 10% Pd/C in
THF, to give the amino acid (10c). This was purified using
ion exchange chromatography.
a, b
S
S
S
S
S
6c
8c
CO2H
CO2Me
f, g, h
e
CO2t-Bu
CO2t-Bu
7c
9c
i, j
CO2H
NH2
N3
S
CO2t-Bu
10c
Scheme 2. Reagents and conditions: (a) 20% Pd/C/H2, MeOH, 100%;
(b) MeCO2Cl, TEA, DMAP, 73%; (c) LDA, THF, À78 °C; (d) tert-
butylbromoacetate, 54%; (e) LiOH, H2O, THF, i-PrOH, 94%; (f)
BH3ÆSMe2, THF, 0 °C, 70%; (g) TsCl, TEA, CH2Cl2, 90%; (h) NaN3,
DMSO, 60 °C, 95%; (i) formic acid (88%), 30 °C, 97%; (j) 10% Pd/C/
H2, THF, 58%.
The (R)- and (S)-enantiomers of the racemic 2-furan ana-
log (11a) were synthesized stereospecifically. The synthe-
sis used for the (S)-enantiomer (19) is shown in Scheme 3.
The 3-furan-2-yl-propionic acid ethyl ester (13) was
hydrolyzed to the acid using 1 N LiOH in THF/i-PrOH.
The resulting acid (14) was coupled with the chiral
auxiliary (4R,5S)-(+)-4-methyl-5-phenyl-2-oxazolidinone
(1.02 equiv) using trimethylacetyl chloride (1.5 equiv)
The tetrahydrofuran analog (12) could be obtained from
the corresponding furan amine tert-butyl ester (11b) by a
subsequent hydrogenation using 20% Pd/C in methanol.
The final amino acid was isolated using ion-exchange
chromatography.

R. M. Schelkun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2329–2332
2331
a
CO2H
CO2Et
O
O
13
14
Ph
O
c, d
b
O
O
14
+
N
N
O
O
O
O
Ph
15
Ph
O
tBuO2C
O
CO2tBu
CO2H
f, g, h
e
N
O
O
16
17
CO2H
NH2
CO2tBu
N3
c, d
O
O
19
18
Scheme 3. Reagents and conditions: (a) LiOH, H2O, THF, i-PrOH, 100%; (b) trimethylacetyl chloride, TEA, LiCl, THF, 66%; (c) LDA, THF,
À78 °C; (d) tert-butyl bromoacetate, 75%; (e) LiOH, H2O2, THF, H2O, 78%; (f) BH3SMe2, THF, 0 °C, 75%; (g) TsCl, TEA, CH2Cl2, 95%; (h) NaN3,
DMSO, 60 °C, 90%; (i) 10% Pd/CaCO3, EtOAc, 84%; (j) TFA, CH2Cl2, 50–70%.
and triethylamine (3.75 equiv) in THF at 0 °C. This acyl
oxazolidinone (15) was alkylated with tert-butyl bro-
moacetate and LDA in THF at À78 °C. The desired prod-
uct (16) was isolated as a single diastereomer as evidenced
by 1H NMR.9 The chiral auxiliary was cleaved using 1 N
LiOH (2 equiv) and H2O2 (2 equiv) in THF. This carbox-
ylic acid tert-butyl ester (17) was then carried on to amino
acid (19) using the previously described chemistry for the
racemic analog (11a). The (R)-enantiomer (20) was syn-
thesized using the (4S,5R)-(À)-4-methyl-5-phenyl-2-oxa-
zolidione in place of the (+)-chiral auxiliary, and the
same synthetic steps already described.
Compounds were tested for their ability to displace
[3H]gabapentin (10 nM final concentration) from the
a2-d subunit of CNS calcium channels ([3H]GPB) (Table
1) similar to previously published methods,10 and for
their potency to inhibit the uptake of [3H]leucine into
CHO cells.11 System L affinity (SYS L) is thought to be
important for allowing these molecules to cross the blood
brain barrier. Results for these two tests are expressed as
an IC50 SEM. In vivo profiling of compounds entailed
evaluation of their ability to elicit anticonvulsant
and anxiolytic like effects in animal models (Table 1).
The DBA/2 strain of mice was used to evaluate
Table 1. In vitro and in vivo profiling of pregabalin and test compounds
HO2C
NH2
R
Compound
R
Stereochemistry [3H]GBP Binding IC50 (lM)a,d SYS L IC50 (lM)b,d WLC %c DBA/2 % protection (2 h)
1
Isobutyl
Ph-CH2–
S
0.073 ( 0.035)
>10,000
284 ( 214)
100
100
0
3
—
—
—
—
—
S
10c
10d
11a
11b
19
2-Thienyl-CH2–
3-Thienyl-CH2–
2-Furanyl-CH2–
3-Furanyl-CH2–
2-Furanyl-CH2–
2-Furanyl-CH2–
3-THF-CH2–
2.14 ( 1.38)
0.832 ( 0.291)
0.421 ( 0.001)
0.518 ( 0.039)
0.178 ( 0.031)
1.46 ( 0.359)
>10,000
2936
10.3
À3
5.3
À17
40
60
20
0
7034
10,000
1422
10,000
10,000
40
0
20
12
R
0.28
—
0
a IC50 is the concentration (lM) producing half-maximal inhibition of the specific binding of [3H]gabapentin to pig brain membranes.
b IC50 is the concentration (lM) producing half-maximal inhibition of the uptake of [3H]Leucine in CHO cells.
c Activity of pregabalin at 30 mg/kg defined as 100%.
d Values are means of three experiments; standard deviation is given in parentheses. Values without SEM are from n = 1.

2332
R. M. Schelkun et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2329–2332
anticonvulsant activity of the compounds after inducing
seizures in the mice using an audio stimulus.12 Results are
expressed as a percentage of mice protected (out of five
animals) from tonic seizures 2 h post dose at 30 mg/kg
administered orally. The anxiolytic effect of compounds
was evaluated by measuring their ability to reverse
shock-induced suppression of drinking in the Vogel
water lick conflict assay (WLC %) in rats at a dose of
30 mg/kg administered orally.13 Activity in this assay is
expressed as a percentage of pregabalinꢀs ability to re-
store punished drinking behavior at a dose of 30 mg/kg.
understanding of the SAR requirements for future
development.
Acknowledgments
The authors wish to acknowledge the late Stephen John-
son for synthesizing the phenyl analog 3, Nirmala Su-
man-Chauhan for providing the a2-d binding data, and
Don Johnson for running the hydrogenation reactions.
Compared to pregabalin, the phenyl analog14 (3) was
completely inactive in the a2-d assay, as well as the
DBA/2 model. Surprisingly, replacement of the phenyl
group with either a furan or thiophene ring resulted in
compounds with a2-d binding affinity. Activity in the
a2-d binding assay followed the trend 2-furan ꢀ 3-fu-
ran > 3-thiophene > 2-thiophene. The 3-THF analog
(12) was inactive. While a number of compounds (10d,
11a, 11b, and 19) possessed sub-micro molar binding,
none of the compounds had better binding than pregaba-
lin (1). All of the compounds tested in the System L assay
were extremely weak when compared to pregabalin (1).
The 2-fural analog (11a) was chosen for stereoselective
synthesis of its enantiomers based on its a2-d binding,
and consistent with the stereochemistry of pregabalin,
the (S)-enantiomer (19) was more potent in the a2-d
assay.9 In the DBA/2 mouse model, the 3-thienyl analog
(10d) was most active, followed by the 2-thienyl analog
(10c) and the 2-fural analog (10b).
References and notes
1. (a) Gee, N. S.; Brown, J. P.; Dissanayake, V. U. K.;
Offord, J.; Thurlow, R.; Woodruff, G. N. J. Biol. Chem.,
271 5768; (b) Marias, E.; Klugbauer, N.; Hofmann, F.
Mol. Pharmacol. 2001, 59, 1243.
2. (a) Martin, D. J.; McClelland, D.; Herd, M. B.; Sutton, K.
G.; Hall, M. D.; Lee, K.; Pinnock, R. D.; Scott, R. H.
Neuropharmacology 2002, 42, 353; (b) Sutton, K. G.;
Martin, D. J.; Pinnock, R. D.; Lee, K.; Scott, R. H. Br. J.
Pharmacol. 2002, 135, 257; (c) Stefani, A.; Hainsworth, A.
H. Curr. Neuropharmacol. 2003, 1, 237.
3. Dooley, D. J.; Donovan, C. M.; Pugsley, T. A. J.
Pharmacol. Exp. Ther. 2000, 295, 1086.
4. Hill, C. M.; Balkenohl, M.; Thomas, D. W.; Walker, R.;
Mathe, H.; Murray, G. Eur. J. Pain 2001, 5, 119.
5. Rosenstock, J.; Tuchman, M.; LaMareaux, L.; Sharma,
U. Pain 2004, 110, 628.
6. Pande, A. C.; Crockatt, J. G.; Feltner, D. E.; Janney, C.
A.; Smith, W. T.; Weisler, R.; Londborg, P. D.; Bielski, R.
J.; Zimbroff, D. L.; Davidson, J. R. Am. J. Psychiat. 2003,
160, 533.
7. Lauria-Horner, B. A.; Pohl, R. B. Expert Opin. Invest.
Drugs 2003, 12, 663.
8. The 3-furan 9b gave a mixture of desired furan-amine and
THF-amine, which could be separated by chromatogra-
phy. For the 2-furan 9a, no reduction of the furan was
observed.
9. Stereochemistry assigned based on previously published
results on the chiral synthesis of pregabalin: Yuen, P.-W.;
Kanter, G. D.; Taylor, C. P.; Vartanian, M. G. Bioorg.
Med. Chem. Lett. 1994, 4, 823.
Consistent with what was observed with a2-d binding,
the (S)-enantiomer (19) was more active in the DBA/2
model than the corresponding (R)-enantiomer (20). In
the WLC anxiety model, the compounds tested were
all devoid of any anxiolytic properties relative to pre-
gabalin. It is believed that the poor System L binding
affinity of these compounds accounts for their poor in vi-
vo activity when compared to pregabalin (1).
In summary, a series of novel analogs of pregabalin have
been synthesized where the isobutyl side chain in pre-
gabalin has been replaced with a heteroaromatic group.
These compounds showed activity in both in vitro and
animal models. Surprisingly, only the furan and thio-
phene replacements possessed biological activity, albeit
it modest when compared to pregabalin, whereas the
phenyl and THF replacements were completely inactive
both in vitro and in vivo. These results were unexpected
given the inactivity of phenyl analog (3). While none of
the compounds synthesized possessed a better biological
profile than pregabalin, they do provide an expanded
10. Suman-Chauhan, N.; Webdale, L.; Hill, D. R.; Woodruff,
G. N. Eur. J. Pharmacol., Mol. Pharmacol. 1993, 244, 293.
11. Su, T.-Z.; Lunney, E.; Campbell, G.; Oxender, D. J. Neu-
rochem. 1995, 64, 2125.
12. (a) Schlesinger, K.; Greik, B. J. The Genetics and
Biochemistry of Audiogenic Seizures. In Contributions to
Behavior-Genetic Analysis: The Mouse as a Prototype;
Lindzey, G., Theissen, D. D., Eds.; Appleton-Centruy-
Crofts: New York, 1970, pp 219–257; (b) Seyford, T. N.
Fed. Proc. 1979, 38, 2399.
13. Milan, M. J. Prog. Neurobiol. (Amsterdam, Netherlands)
2003, 70, 83.
14. Kori, S.; Tokuhisa, T. JP Patent 45,028,769.