The efficient synthesis of (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid, a chiral β-amino acid with potent affinity for the α2δ protein

Article abstract of DOI:10.1016/j.tetlet.2008.12.111

A chiral β-amino acid containing three contiguous chiral centers was synthesized efficiently in 11 steps, employing enantio-enriched β-ketoester as a key intermediate, via stereoselective catalytic hydrogenation of the corresponding enamide. Stereoselecti

Full text of DOI:10.1016/j.tetlet.2008.12.111

Tetrahedron Letters 50 (2009) 1167–1170
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
The efficient synthesis of (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid,
a chiral b-amino acid with potent affinity for the
a2d protein
a
a
a
a
a
Ji Zhang a, , Peter G. Blazecka , Derek A. Pflum , Joseph Bozelak , Derek Vrieze , Norman L. Colbry ,
Garrett Hoge a, David C. Boyles a, Brian Samas a, Timothy T. Curran a, Augustine T. Osuma b, Paul Johnson b,
Suzanne Kesten b, Jacob B. Schwarz b, Annise Goodman b, Mark Plummer b, Anne Akin c, Yun Huang c,
Michael Lovdahl c, Andrew J. Thorpe b
*
a Research API, Michigan Laboratories, Pfizer, Inc., 2800 Plymouth Road, Ann Arbor, MI 48105, USA
b Medicinal Chemistry, Michigan Laboratories, Pfizer, Inc., 2800 Plymouth Road, Ann Arbor, MI 48105, USA
c Analytical API, Pfizer Global Research & Development, Michigan Laboratories, Pfizer, Inc., 2800 Plymouth Road, Ann Arbor, MI 48105, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A chiral b-amino acid containing three contiguous chiral centers was synthesized efficiently in 11 steps,
employing enantio-enriched b-ketoester as a key intermediate, via stereoselective catalytic hydrogena-
tion of the corresponding enamide. Stereoselective 1,4-addition of a methyl group and protonation were
key to the preparation of the desired acid 12. Mild and efficient reaction conditions were applied to the
enamine formation and protection to avoid epimerization at C-4 of compounds 13 and 14. The final com-
Received 19 September 2008
Revised 15 December 2008
Accepted 22 December 2008
Available online 8 January 2009
pound was found to display potent affinity for the
treatment of a variety of diseases.
a
2d-protein that is a recognized drug target for the
Keywords:
Chiral b-amino acid
b-Ketoester
Ó 2009 Elsevier Ltd. All rights reserved.
Oxazolidinone
Asymmetric conjugate addition
Organocuprate
Stereoselective catalytic hydrogenation
The synthesis of chiral amino acids consistently attracts strong
interest from across the synthetic organic chemistry community.1
In addition to being attractive synthetic targets, amino acids often
have profound biological activity.2
cient material to support pre-clinical studies. The original route
to the targeted amino acid (Scheme 1) was based on the key inter-
mediate chiral aldehyde 7, which was prepared from crotonic acid
via oxazolidinone chemistry.6 Asymmetric conjugate addition of
organocopper reagent, prepared from n-propylmagnesium chlo-
ride and CuBr.SMe, to oxazolidinone 4,7 followed by treatment
with excess MeI (5 equiv) resulted in a diastereomeric mixture
which was extremely difficult to purify. Reduction of oxazolidi-
none 5 with LiAlH4, followed by PCC oxidation gave chiral aldehyde
7. The subsequent chemistry involved the use of Davis’ (S)-p-tolu-
enesulfinamide8 to transform 7 into chiral imine 8, followed by
stereoselective Ti (IV)-mediated Reformatsky-type reaction to ob-
tain 9. There were several issues with this approach: (1) The reac-
tion of 4 to 5 resulted in a 3:1 to 5:1 diastereomeric mixture and
Pregabalin (LyricaÒ),3 a chiral -amino acid,4 is approved in the
c
US for the treatment of post-herpetic neuralgia and as an add-on
therapy for epilepsy. With additional approval in the EU for gener-
alized anxiety, pregabalin is emerging as a key medicine available
to physicians, and has attracted much interest due to its therapeu-
tic properties, pharmacology, and structure–activity relationships.
With respect to the latter, we undertook the synthesis and biolog-
ical evaluation of b-amino acid analogues of pregabalin.5 Herein,
we report an efficient preparation of the synthetically challenging
b-amino acid (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid
using a chiral b-ketoester as the key intermediate (Fig. 1).
The target amino acid 2 is notably more complex and poses a
NH2
greater synthetic challenge than c-amino acid 1. It was important
NH2
to develop a robust synthetic route to allow preparation of suffi-
COOH
COOH
* Corresponding author at present address: Process Research & Development,
Bristol-Myers Squibb, Princeton, NJ 08543, Unites States. Tel.: +1 609 252 5469; fax:
+1 609 252 7825.
1, Pregabalin
2
Figure 1. Structure of pregabalin 1 and (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic
acid 2.
E-mail address: ji.zhang@bms.com (J. Zhang).
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.12.111

1168
J. Zhang et al. / Tetrahedron Letters 50 (2009) 1167–1170
O
O
O
O
MgCl
OH
O
O
HN
Ph
N
O
CuBr.SMe2, THF
then MeI -35°C
62%
Ph
H3CC Cl
3
4
LiCl, Et3N,
THF, -44°C
72%
(chromatography)
O
PCC, CH2Cl2
Al2O3, 25°C
O
OH
O
N
79%
LAH, THF
-78°C
(Plug chromatography)
Ph
5
6
87% (crude)
O
S
O
H2N
O
OtBu
S
N
O
TiClOiPr3,LDA,
THF, -78°C
Ti(IV)OEt4, THF 25°C
53%
(chromatography)
7
8
55%
(chromatography)
CO2tBu
CO2H
NH2
TFA, MeOH, CH2Cl2
NH
S
O
76%
(Ion Exchange)
9
2
Scheme 1. Original route to enriched (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid 2 from chiral aldehyde 7.
the crude material was an oil, requiring a tedious isolation and
purification to enrich 5; (2) oxidation of 6 to 7 utilized scale-un-
friendly PCC, and provided a product prone to racemization; (3)
conversion of 7 to 8 and 8 to 9 consistently afforded low yields
(53% and 55%); (4) ion exchange chromatography used to isolate
and purify final product required large volumes of NH4OH and
aqueous HCl. Poor product recovery was compounded by the diffi-
culty of removing water as scale increased; (5) isomeric impurities
in the final isolated product were persistent, and upgrading was
extremely difficult.
chiral purity >99%). But despite the improvements in the orginal
route, chiral aldehyde 7 was found to be an unstable intermediate,
highly volatile, and prone to epimerization. In addition, intermedi-
ates 7, 8, 9, and 2 are oils, offering no opportunity for facile purifi-
cation on scale. With the realization that aldehyde 7 was not a
viable intermediate for further scaling, an alternative, chiral acid
12 was identified as a key intermediate in the preparation of
b-ketoester 13, itself a precursor for a stereoselective enamide
hydrogenation, and a new approach was designed as illustrated
(see Supplementary data).
The original route was modified as follows (Scheme 1): First,
chiral aldehyde 7 was prepared from 5 by thioester formation fol-
lowed by Fukuyama reduction (Pd/Et3SiH),9 avoiding LiAlH4 reduc-
tion and PCC oxidation. Conversion of 8 to 9 was improved to >90%
yield and the need for chromatographic purification was elimi-
nated, and instead of deprotection using TFA/MeOH, followed by
ion exchange chromatography, the HCl salt of (3R,4R,5R)-3-ami-
no-4,5-dimethyl-octanoic acid was prepared directly. Recrystalli-
zation from MeOH/MeCN effectively removed isomeric
impurities. Using this route, >10 g of enriched (3R,4R,5R)-3-ami-
no-4,5-dimethyl-octanoic acid was delivered (both chemical and
To confirm the viability of a b-ketoester as an intermediate in
the synthesis of (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid 2,
and to prepare a racemic standard to help identify and assign abso-
lute stereochemistry to the impurities in the final product, we
decided to test the b-ketoester route utilizing a mixture of acid iso-
mers of 12. This route started with an indium triflate-mediated
addition of ethyl methylmalonate to 1-pentyne. Subsequent reduc-
tion, saponification, and decarboxylation gave racemic acid 12.10
(3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid, and all possible
stereoisomers (Fig. 2) were synthesized in eight steps, in good
yield, via the b-ketoester isomer mixture (see Supplementary
COOH
NH2
COOH
NH2
COOH
NH2
COOH
NH2
D4
D3
D1
D2
COOH
NH2
COOH
NH2
COOH
COOH
NH2
NH2
D8
D6
D5
D7
Figure 2. All possible stereoisomers of 3-amino-4,5-dimethyl-octanoic acid.

J. Zhang et al. / Tetrahedron Letters 50 (2009) 1167–1170
1169
LiOH.H2O
H2O2
THF:H2O
(3:1)
CuBr. SMe2
LiCl
MeMgCl
O
O
O
O
O
N
O
OH
N
O
>95%
Set two
Ph
Ph
chiral centers
Ph
10
Ph
12
11
55-62%
NH2OMe HCl (1.1 eq),
NaOAc (1.0 eq)
NOMe
O
CDI/THF
COOEt
4
CO2Et
4
CO2Et
MeOH
rt, 24h
oil
14
13
COOK
>97%
CH3COCl
Pyridine
NHAc
oil
NH2
oil
Ra-Ni
CO2Et
CO2Et
CH2Cl2
-20 to 0 oC
MeOH
H2
16
15
>98%
>98%
Scheme 2. Mild and efficient method to prepare chiral enamine.
data). Thus, this indicated that a new route based on a chiral b-
ketoester was chemically feasible, and we therefore turned our
attention to the preparation of enantiomerically pure acid 12.
Enantio-enriched acid 12 was prepared from intermediate 5 by
LiOH/H2O2 hydrolysis to remove the chiral auxiliary. Following the
approach described (see Supplementary data), the corresponding
enamide 16 was synthesized in two operations. Raney-Ni-cata-
lyzed hydrogenation of the enamide, followed by HCl hydrolysis,
afforded a mixture of desired product and its diastereomer in a ra-
tio of 2.7:1 (or 64%:24% at the b-position). This early encouraging
result from the Ra-Ni reduction suggested that a thorough screen-
ing of hydrogenation catalysts and reaction conditions was
necessary.
tion. After screening (Table 1), the optimal conditions were imple-
mented in a successful scale-up employing crude 13. In this case,
we achieved excellent conversion (>98%) with acceptable selectiv-
ity (entry 5, in small scale: dr = 97:3, in larger scale, dr = 92:8).
With the Z-enamide as the reactant, we believe that a hydrogen
bond between the NH and the carbonyl oxygen exists, thus a Cram-
chelation model is best to explain the stereochemical outcome
(Scheme 3). It was reported that hydrogenation of d-hydroxy-b-
ketoester preferentially gave syn-1,3-diols with high diasteroselec-
tivity.12 Prasad and co-workers suggested that the formation of an
intramolecular hydrogen bond explains this stereochemical out-
come. Recently, Ma reported the preparation of syn-d-hydroxy-b-
amino ester via hydrogenation of d-hydroxy-b-ketoester-derived
enamine with high diastereoselectivity.13 To rationalize the result,
intramolecular hydrogen bonds between the d-hydroxyl group and
the nitrogen atom were invoked. In our own case, without the
hydrogen bonding we believe that the neighboring group (CH3)
plays an important additive role in controlling facial selectivity of
the addition of hydrogen to provide (3R,4R,5R)-3-acetylamino-
4,5-dimethyl-octanoic acid.
The early reaction conditions11 (MeOH/NH4OAc in rt for four
days, then acetic anhydride/pyridine at 110 °C for 16 h) to form
enamide 16 caused epimerization at the C-4 position and resulted
in low yield of the desired product. We considered that firstly, the
enamine formation gave low conversion due to the poor reactivity
of NH4OAc and steric hindrance of neighboring chiral center. Sec-
ondly, in the subsequent step, the harsh condition (with base at
elevated temperature) to make enamide 16 caused epimerization.
Thus, a mild and efficient method to prepare chiral enamide 16 be-
came our main focus.
Several reagents, such as NH2NHPh, NH2OH, NH2NH2, and
BnNH2, were studied for the enamine formation, but none provided
a single product in high yield under mild conditions. Since the
above reagents have more than one reactive site, we turned our
attention to NH2OMeÁHCl (Scheme 2). Indeed the combined re-
agent system (1.1 equiv NH2OMeÁHCl with 1.0 equiv NaOAc) in
MeOH not only gave almost quantitative conversion in 24 h at rt,
but also simplified the work-up. Intermediate 14 was isolated in
98% yield by removing MeOH after filtration. Without further puri-
fication, the N–O bond 14 was hydrogenated with Ra-Ni giving 15
in excellent yield (98%). Protection with acetyl chloride/pyridine at
À20 to 0 °C also turned out to be very successful (yield >98%). The
beauty of these three steps is that while each product is a liquid,
due to the excellent yield, mild conditions, and high purity for each
step, crude material could be carried forward in the next transfor-
mation without further purification. Moreover, epimerization at C-
4 was significantly minimized due to the mild reaction conditions
Since (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid is a chiral
b-amino acid with three contiguous chiral centers, the isomeric
impurity or quality control was critical. Similar physicochemical
properties of stereoisomers (Fig. 2) made separation of one from an-
Table 1
Hydrogenation screening
NHAc
CO2Et
NHAc
CO2Et
NHAc
CO2Et
catalyst
MeOH, 20 psi H2
room temp
Undesired
17b
Desired
16
17a
Entry
Catalyst
Conversion (%)
17a:17ba
1
2
3
4
5
6
7
8
5% Pd/SrCO3
5% Pd/MgO
5% Pd/CaCO3
5% Pd/CaCO3
5% Pd/BaSO4
5% Pd/Al2O3
5% Pd/BaSO4
5% Pd/BaCO3
66
93
99
52:1
38:1
34:1
33:1
31:1
28:1
20:1
17:1
89
100
100
100
100
(ratio of desired:undesired >20:1 at the c-position).
With pure enamide 16 in hand (>95% Z), a screen was conducted
to determine the best reaction conditions for catalytic hydrogena-
a
Measured by chiral GC.

1170
J. Zhang et al. / Tetrahedron Letters 50 (2009) 1167–1170
M H
COOEt
NHAc
Me
O
H
H
CH3
COOEt
R
OEt
H
R
AcN
R
H
AcHN
Scheme 3. Cram-chelation model for hydrogenation of chiral enamide 16.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2008.12.111.
References and notes
1. Reviews, see: (a) Ordonez, M.; Cativiela, C. Tetrahedron: Asymmetry 2007, 18, 3;
(b) Ma, J.-A. Angew. Chem., Int. Ed. 2003, 42, 4290; (c) Zhou, P.; Chen, B.-C.;
Davis, F. A. Tetrahedron 2004, 60, 8003; (d) O’Donnell, M. J. Acc. Chem. Res. 2004,
37, 506.
2. See Review: (a) Field, M. J.; Li, Z.; Schwarz, J. B. J. Med. Chem. 2007, 50, 2569; (b)
Abdel-Magid, A. F.; Cohen, J. H.; Maryanoff, C. A. Curr. Med. Chem. 1999, 6, 955;
(c) Cole, D. C. Tetrahedron 1994, 50, 32.
3. (a) Andruszkiewicz, R.; Silverman, R. B. Synthesis 1989, 953; (b)
Andruszkiewicz, R.; Barrett, A. G. M.; Silverman, R. B. Synth. Commun. 1990,
20, 159; (c) Silverman, R. B. Angew. Chem., Int. Ed. 2008, 47, 3500; (d) Yuen, P.-
O.; Kanter, G. D.; Taylor, C. P.; Vartanian, M. G. Bioorg. Med. Chem. Lett. 1994, 4,
823; (e) Tassone, D. M.; Boyce, E.; Guyer, J.; Nuzum, D. Clin. Ther. 2007, 29, 26;
(f) Taylor, C. P.; Angelotti, T.; Fauman, E. Epilepsy Res. 2007, 73, 137; (g) Burk, M.
J.; De, K.; Pieter, D.; Grote, T. M.; Hoekstra, M. S.; Hoge, G.; Jennings, R. A.;
Kissel, W. S.; Le, T. V.; Lennon, I. C.; Mulhern, T. A.; Ramsden, J. A.; Wade, R. A. J.
Org. Chem. 2003, 68, 5731.
Figure 3. X-ray crystal structure of intermediate 11, and HCl salt of (3R,4R,5R)-3-
amino-4,5-dimethyl-octanoic acid 2.
other a daunting task. We had to identify the isomeric impurities in
our final product, and determine their origin and the maximum tol-
erable level. A careful study indicated that removal of isomeric
impurity D7, which arises from poor selectivity in the 1,4-Michael
addition, would be difficult. Impurity D3 was always present be-
cause the catalytic hydrogenation was not 100% selective; fortu-
nately, D3 was the most easily removed impurity by
recrystallization.
4. Yuen, P.-W. In The Art of Drug Synthesis; Johnson, D. S., Li, J. J., Eds.; Wiley-
Interscience, 2007. Chapter 16, pp 225–239.
5. (a) Schwarz, J. B.; Colbry, N. L.; Zhu, Z.; Nichelson, B.; Barta, N. S.; Lin, K.;
Hudack, R. A.; Gibbons, S. E.; Galatsis, P.; DeOrazio, R. J.; Manning, D. D.;
Vartanian, M. G.; Kinsora, J. J.; Lotarski, S. M.; Li, Z.; Dickerson, M. R.; El-Kattan,
A.; Thorpe, A. J.; Donevan, S. D.; Taylor, C. P.; Wustrow, D. J. Bioorg. Med. Chem.
Lett. 2006, 16, 3559; (b) Schelkun, R. M.; Yuen, P-W.; Wustrow, D. J.; Kinsora, J.;
Su, T-Z.; Vartanian, M. G. Bioorg. Med. Chem. Lett. 2006, 16, 2329; (c) Schwarz, J.
B.; Gibbons, S. E.; Graham, S. R.; Colbry, N. L.; Guzzo, P. R.; Le, V-D.; Vartanian,
M. G.; Kinsora, J. J.; Lotarski, S. M.; Li, Z.; Dickerson, M. R.; Su, T-Z.; Weber, M. L.;
El-Kattan, A.; Thorpe, A. J.; Donevan, S. D.; Taylor, C. P.; Wustrow, D. J. J. Med.
Chem. 2005, 48, 3026.
In the final optimization of the synthesis, (1R,2S)- diphenyl-2-
oxazolidinone was used as chiral auxiliary to provide intermediates
with better crystallinity 11 (Fig. 3, X-ray crystal structure). EEDQ14
(N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline), a convenient
and efficient reagent for peptide synthesis, was used for the coupling
of commercially available 2-methyl-2-pentenoic acid with diphe-
nyl-2-oxazolidinone to provide 10. Addition of MeMgCl and quench-
ing of the enolate with acetic acid gave predominantly 11 which,
after recrystallization, provided enriched 11 in 55–62% yield and
93% diastereomeric purity and avoided the use of MeI.15 This mate-
rial could be further purified by recrystallization (95:5 = heptane/
toluene) to give 11 having >98% chemical and chiral purity.
6. Asymmetric synthesis of
a-amino acid using chiral oxazolidinones, see: (a)
Evans, D. A.; Britton, T. C. J. Am. Chem. Soc. 1987, 109, 6881; (b) Evans, D. A.;
Weber, A. E. J. Am. Chem. Soc. 1987, 109, 7151; (c) Evans, D. A.; Weber, A. E. J.
Am. Chem. Soc. 1986, 108, 6757.
7. Asymmetric conjugate addition of organocuprates to chiral a,b-unsaturated N-
acyl-4-phenyl-2-oxazolidinones, see: (a) Nicolàs, E.; Russell, K. C.; Hruby, V. J. J.
Org. Chem. 1993, 58, 766; (b) Li, G.; Jarosinski, M. A.; Hruby, V. J. Tetrahedron
Lett. 1993, 34, 2561; (c) Liao, S.; Han, Y.; Qui, W.; Bruck, M.; Hruby, V. J.
Tetrahedron Lett. 1996, 37, 7917; (d) Williams, D. R.; Kissel, W. S.; Li, J. J.
Tetrahedron Lett. 1998, 39, 8593; (e) Williams, D. R.; Kissel, W. S.; Li, J. J.;
Mullins, R. J. Tetrahedron Lett. 2002, 43, 3723.
8. (a) Davis, F. A.; Deng, J. Org. Lett. 2004, 6, 2789; (b) Davis, F. A.; Lee, S. H.; Xu, H.
J. Org. Chem. 2004, 69, 3774; (c) Davis, F. A.; Reddy, R. E.; Szewczyk, J. M. J. Org.
Chem. 1995, 60, 7037.
9. (a) Fukuyama, T.; Tokuyama, H. Aldrichim. Acta 2004, 37, 87. and references
therein; (b) Kimura, M.; Seki, M. Tetrahedron 2004, 45, 3219.
10. Zhang, J.; Blazecka, P. G.; Angell, P.; Lovdahl, M.; Curran, T. T. Tetrahedron 2005,
61, 7807.
11. Zhu, G.; Chen, Z.; Zhang, X. J. Org. Chem. 1999, 64, 6907.
12. Kathawala, F. G.; Prager, B.; Prasad, K.; Repic, O.; Shapiro, M. J.; Stabler, R. S.;
Widler, L. Helv. Chim. Acta 1986, 69, 803.
In conclusion, a potent
a
2d ligand,16 chiral b-amino acid with
three contiguous chiral centers was synthesized efficiently over
11 steps using chiral b-ketoester 13 as a key intermediate. Epimer-
ization at C-4 was minimized. Early process chemistry has been
developed which is free of chromatography and produced >300 g
of (3R,4R,5R)-3-amino-4,5-dimethyl-octanoic acid 2 having excel-
lent chemical (>99%) and chiral (99%) purity.
13. Zhu, W.; Ma, D. Org. Lett. 2003, 5, 5063.
14. Zacharie, B.; Connolly, T. P.; Penny, C. L. J. Org. Chem. 1995, 60, 7072.
15. For a description of use of 1,4- addition and selective enolate protonation of
Acknowledgments
a
,b-disubstituted enoyl substrates please see: (a) Oppolzer, W.; Kinma, A. J.
We would like to thank Mark Weber for the pharmacological
evaluation of the affinity of the compound to the a2d-protein. We
are grateful for the support, encouragement, and suggestions from
Drs. Dan Belmont, Jim Davidson, Bob Sliskovics, and Professors E. J.
Corey and Scott E. Denmark.
Helv. Chim. Acta 1989, 72, 1337; (b) Oppolzer, W.; Poli, G.; Kingma, A. J.;
Starkemann, C.; Bernardinelli, G. Helv. Chim. Acta 1987, 70 2201; (c) Oppolzer,
W.; Kingma, A. J.; Poli, G. Tetrahedron 1989, 45, 479; (d) Reyes, E.; Vicario, J. L.;
Carrillo, L.; Badia, D.; Iza, A.; Uria, U. Org. Lett. 2006, 8, 2535.
16. Amino acid 2 was found to display an IC50 of 24 nM for the
123 nM for the 2d-2 protein.
a2d-1 protein and
a