Stereoselective 5-exo-trig radical cyclization in the enantioselective synthesis of Pregabalin

Article abstract of DOI:10.1016/j.tetlet.2007.04.054

A practical stereoselective 5-exo-trig radical cyclization procedure was developed in order to prepare enantiomerically pure GABA derivative precursors (4-alkyl-pyrrolidin-2-ones). This procedure allows much more rapid access to optically pure GABA derivatives, such as the powerful antiepileptic agent (S)-(+)-3-aminomethyl-5-methylhexanoic acid (Pregabaline).

Full text of DOI:10.1016/j.tetlet.2007.04.054

Tetrahedron Letters 48 (2007) 4305–4308
Stereoselective 5-exo-trig radical cyclization in the
enantioselective synthesis of Pregabalin
*
´
´
Veronica Rodrıguez, Leticia Quintero and Fernando Sartillo-Piscil
Centro de Investigacio´n de la Facultad de Ciencias Quı´micas, BUAP, 14 Sur Esq. San Claudio, Col. San Manuel, 72570,
Puebla, Mexico
Received 23 February 2007; revised 27 March 2007; accepted 2 April 2007
Available online 18 April 2007
Abstract—A practical stereoselective 5-exo-trig radical cyclization procedure was developed in order to prepare enantiomerically
pure GABA derivative precursors (4-alkyl-pyrrolidin-2-ones). This procedure allows much more rapid access to optically pure
GABA derivatives, such as the powerful antiepileptic agent (S)-(+)-3-aminomethyl-5-methylhexanoic acid (Pregabaline).
Ó 2007 Published by Elsevier Ltd.
Pregabalin 1 ((S)-(+)-3-aminomethylhexanoic acid) is a
novel and potent anticonvulsant agent for the treatment
of epilepsy and pain.1 Its biological activity is compara-
ble to that of Gabapentin 2.2 Pregabalin was originally
obtained along with its opposite R enantiomer,3 how-
ever, it was found that this mixture of enantiomers did
not posses pharmacological activity after animal tests.1a
Thus several enantioselective routes to obtain the
desired S enantiomer have been achieved, not only at
industrial level3,4 but also at academic level1a (Fig. 1).
not stereoselective at all.5–7 However, the presence of
the a-phenylethylamine group8 as chiral auxiliary in
the framework of the pyrrolidinones 6 and 7 made it
possible to isolate each of the diastereoisomers by
column chromatography, which eventually resulted in
the isolation of the GABA derivatives 3 and 4 (Scheme
1).
We now report a stereoselective 5-exo-trig radical cycli-
zation version of the reaction showed in Scheme 1, and
Recently, we reported an accessible method for the
synthesis of GABA derivatives 3 and 4.5 The 5-exo-trig
radical cyclization reaction of optically pure allylic
amides 5 to give diastereoisomeric pyrrolidinones 6
and 7 was employed as the key reactions.6 Under the
classical free radical condition reactions (Bu3SnH/
AIBN/in refluxing benzene or lauroyl peroxide (DLP)
in refluxing benzene), the radical cyclization step was
N
Ph
X = Br, SC(S)OEt
O
X
5
5-exo-trig
radical cyclization
6/7 = 50:50
Y
Y
+
N
N
Ph
Ph
Y = H, SC(S)OEt
O
O
6
7
OH
OH
1. Separation
2. Debenzylation
3. Hydrolysis
NH2
NH2
O
O
Gabapentin
Pregabalin
O
O
NH2
2
1
NH2
OH
OH
Z = H, SH
Z
Z
Figure 1. GABA derivatives as potent anticonvulsants.
4
3
*
Scheme 1. The synthesis of GABA derivatives 3 and 4 applying the 5-
exo-trig radical cyclization as the key reaction.
Corresponding author. Tel.: +52 2222 295500x7387; fax: +52 2222
454293; e-mail: fsarpis@siu.buap.mx
0040-4039/$ - see front matter Ó 2007 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2007.04.054

4306
V. Rodrı´guez et al. / Tetrahedron Letters 48 (2007) 4305–4308
its application to the enantioselective synthesis of the
powerful anticonvulsant agent, Pregabalin 1.
catalysts with the expectation that by controlling the
amide electron delocalization between the nitrogen and
carbonyl group, the nature of the conformational energy
of the transition state of the cyclization reaction can be
altered, and thus produce one pyrrolidinone in excess.
Having this in mind, and because the chiral auxiliary
is not highly oxygenated (there are not multiple sites
for chelation), we thought that common Lewis acids
such as BF3ÆOEt2 and TiCl4 may work properly.
Since the free radical reaction conditions depicted in
Scheme 1 were carried out at high temperatures (reflux-
ing solvent temperatures, within 80–100 °C), the absence
of diastereoselectivity during the cyclization reaction
was actually expected.5–7 Furthermore, recent studies
have shown that the stereoselectivity of free radical
cyclization is greatly enhanced as a result of the use of
a combination of low temperatures and Lewis acids.9
Additionally, the use of chiral auxiliaries such as oxazo-
lidinones, sulfoxides, chiral esters and carbohydrates
have been also investigated for stereoselective radical
cyclization reactions.10
The desired chiral auxiliaries (S)-5 and (S)-8 for the 5-
exo-trig radical cyclization study were readily prepared
following the literature procedure as summarized in
Scheme 2.5 Treatment of (S)-a-methylbenzylamine 9
with allyl iodine 10 or dimethylallyl iodine 11 in the
presence of triethylamine furnished secondary amines
(S)-12 and (S)-13, which were then converted to allylic
amides (S)-5 and (S)-8 by the treatment with bromoacet-
yl bromide and DMAP (Scheme 2).
Thus, we decided to find optimal reaction conditions for
the radical cyclization reaction in order to achieve
improved stereoselective outcome. In this sense, we
choose to explore lower temperatures and Lewis acid
The radical precursors (S)-5 and (S)-8 were next sub-
jected to reductive radical reaction conditions using
Bu3SnH/BEt3/O2/Lewis acids at different temperatures
and in THF. As we anticipated, the presence of Lewis
acids at lower temperature indeed dramatically enhance
the stereoselectivity to give (4S,10S)-pyrrolidin-4-ones as
the major diastereoisomers (entries 6–9). Thus, using
BF3ÆOEt2 and TiCl4 as the Lewis acids, 91:9 and 86:14
diatereoisomeric ratios were achieved, respectively, at
ꢀ78 °C (entries 9 and 7). However, with MgBr2, EtAlCl2
and TMSOTf, the reaction was neither stereoselective
nor effective (entries 5, 10 and 11).11 This is interesting
because the MgBr2 has been successfully used in a num-
ber of stereoselective free radical cyclization reac-
tions.9d,i,10a,d,e When no Lewis acids were used, no
diastereoselectivity was observed even at ꢀ78 °C (Table
1, entries 1–4).
I
NEt3
Ph
NH
R
+
Ph
NH2
CH2Cl2/0 ºC
R
R
R
9
12: R = H
13: R = CH3
10: R = H,
11: R = CH3
R
O
Br
Br
Ph
N
R
Br
DMAP/CH2Cl2
0 ºC
O
5: R = H
8: R = CH3
Scheme 2. Synthesis of chiral allylic amides 5 and 8.
Table 1. Stereoselective tin-mediated radical cyclization
R
R
R
Et3B / O2 /
Bu3SnH
R
N
O
R
N
O
R
N
Lewis acid
THF
O
Br
R = H, (4R,1'S)-14
R = CH3, (4R,1'S)-15
R = H, (4S,1'S)-16
R = CH3, (4S,1'S)-17
R = H, CH3
(S)-5
and (S)-8
Entry
R
Temperature (°C)
Lewis acid
Yielda (%)
Ratiob
(4R,10R)-14
(4R,10R)-15
(4S,10S)-16
(4S,10S)-17
1
2
3
4
5
6
7
8
9
H
H
Me
Me
H
H
Me
H
Me
H
H
rt
ꢀ78
rt
None
None
None
None
MgBr2
TiCl4
TiCl4
BF3OEt2
BF3OEt2
EtAlCl2
TMS-triflate
78
81
78
82
77
68
74
75
53
57
55
54
42
15
14
10
9
47
43
45
46
58
85
86
90
91
—
—
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
84
N.R.
N.R.
10
11
—
—
a Isolated yields.
b Determined ratios by 1H NMR.

V. Rodrı´guez et al. / Tetrahedron Letters 48 (2007) 4305–4308
4307
The absolute configurations for all the pyrrolidin-4-ones
were correlated with those previously reported in the
literature.5,6,11
nBu3SnH/BEt3/O2
BF3·OEt2 in THF
at -78 ºC
76%
22/23 = 88/12,
respectively
Ph
N
Ph
N
Ph
N
O
+
O
O
Br
(S)-18
Having established the effectively stereoselective
5-exo-trig radical cyclization protocol for the synthesis
of optically pure pyrrolidin-4-ones, we next proceeded
to synthesize the desired chiral auxiliary (S)-18, which
is the precursor for the enantioselective synthesis of
Pregabalin 1. A different route had to be developed to
synthesize chiral amide (S)-18 from that for the prepara-
tion of chiral amides (S)-5 and (S)-8. This new strategy
consists of the application of a Wittig olefination reac-
tion between the chiral phosphonium salt (S)-19 and
the isobutyraldehyde followed by N-acylation of amine
(S)-20.
(4S,1'S)-22
(4R,1'S)-23
Scheme 4. Synthesis of the pyrrolidinone precursor of the Pregabalin.
A model explaining the stereo outcome of the reaction is
outlined in Scheme 5. In this model, the N–C bond of
the amide is considered as a formal double bond, which
is obviously favored by the presence of the Lewis acid.
Therefore, chairlike transition states for the 5-exo-trig
radical cyclization (A and B) are proposed. Both, struc-
tures A and B can be locked by a weak Ha–O@C hydro-
gen bonding interaction to give conformationally
restrained structures for cyclization.12 Furthermore,
molecular models suggest that B may be slightly disfa-
vored by allylic 1,3-strain (A1,3) to produce the minor
pyrrolidin-4-ones (see Scheme 5).
Chiral phosphonium salt (S)-19 was prepared by the
reaction of (S)-phenylethylamine 9 and mesylate-phos-
phonium salt 21 in the presence of triethylamine
(Scheme 3). It is worth noting that only one column
chromatography purification was necessary for the
preparation of allylic amide (S)-20 from (S)-9 and 21.
With optically pure pyrrolidinone (4S,10S)-22 in hand,
we completed the enantioselective synthesis of Pregaba-
lin 1. Thus, removal of the benzyl group of (4S,10S)-22
was achieved by using Birch reduction to afford (S)-4-
isobutyl-pyrrolidin-2-one (S)-24 in good yield, which
was then submitted to basic hydrolysis to give enantio-
merically pure Pregabalin 1 (up to 98% ee); [a]D +9.6
(c 1.0, H2O) (Lit.3 [a]D 10.1, c 1.1, H2O) Scheme 6.
Chiral allylic amide (S)-18 was then subjected to stereo-
selective 5-exo-trig radical cyclization with the best reac-
tion conditions shown in Table 1 (entries 8 and 9) and
gave (4S,10S)-4-isopropylpyrrolidinone 22 in good yield
and stereoselectivity (Scheme 4).
In conclusion, the development of a practical stereo-
selective 5-exo-trig radical cyclization protocol of allylic
amides allowed us to rapidly access to optically pure
GABA derivatives. The enantioselective synthesis of
Pregabalin was successfully achieved by using this meth-
od. Additionally, with chiral phosphonium salt (S)-19 in
O
PPh3
Br
Br
HN
Ph3P
NEt3
Ph
+
Ph
NH2
(S)-9
HN
Ph
CH2Cl2
n-BuLi
THF, -78 ºC
72% from 21
(S)-20
cis/trans mixture
91/9, respectively
OMs
21
(S)-19
O
Br
Ph
N
N
O
Br
Ph
HN
OH
Birch
KOH
68%
from
(4S,1'S)-22
DMAP
CH2Cl2
92%
NH2
O
O
O
Br
(S)-18
Pregabalin, 1
(S)-24
(4S,1'S)-22
Scheme 3. Synthesis of chiral allylic amide (S)-18.
Scheme 6. Completion of the synthesis of Pregabalin.
R1
N
R2
R1
R1
R2
(4S,1'S)-16
(4S,1'S)-17
(4S,1'S)-22
Ph
N
H
Bu3Sn
LA, -78 ºC
Ph
R2
Hα
H
O
O
LA
H
N
major
5-exo-trig
radical cyclization
Br
A
hydrogen
bonding
Ph
O
allylic amides
R1
(4R,1'S)-14
(4R,1'S)-15
(4R,1'S)-23
Ph
N
N
R2
R2
Ph
H
R1
H
Hα
O
O
minor
LA
steric repulsion
A1,3
B
Scheme 5. Radical cyclization models.

4308
V. Rodrı´guez et al. / Tetrahedron Letters 48 (2007) 4305–4308
´
5. Rodriguez, V.; Sanchez, M.; Quintero, L.; Sartillo-Piscil,
hand, a number of chiral amides with different side
chains can be prepared for the synthesis of different opti-
cally pure GABA derivatives. Thus, the preparation of a
library of GABA derivatives will be reported in due
course.
F. Tetrahedron 2004, 60, 10809–10815.
6. See previous work of Orena: Cardillo, B.; Galeazzi, R.;
Mobbili, G.; Orena, M.; Rossetti, M. Heterocycles 1994,
38, 2663–2676.
7. Similar results were observed by Ishibashi, H.; Fuke, Y.;
Yamashita, T.; Ikeda, M. Tetrahedron: Asymmetry 1996,
7, 2531–2538.
Acknowledgments
´
8. Juaristi, E.; Leon-Romo, J. L.; Reyes, A.; Escalante, J.
Tetrahedron: Asymmetry 1999, 10, 2441–2495.
9. (a) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry
of Radical Reactions; VCH: Eeinheim, 1995; (b) Sibi, M.
P.; Porter, N. A. Acc. Chem. Res. 1999, 32, 163–171;
(c) Porter, N. A.; Giese, B.; Curran, D. P. Acc. Chem. Res.
1991, 24, 296–304; (d) Renaud, P.; Gerster, M. Angew.
Chem., Int. Ed. 1998, 37, 2562–2579and references
cited therein; (e) Murakata, M.; Jono, T.; Mizuno, Y.;
Hoshino, O. J. Am. Chem. Soc. 1997, 119, 11713–11714;
(f) Nishida, M.; Hayashi, H.; Yamaura, Y.; Yanaginuma,
E.; Yonemitsu, O. Tetrahedron Lett. 1995, 36, 269–272; (g)
Nishida, M.; Ueyama, E.; Hayashi, H.; Ohtake, Y.;
Yamaura, Y.; Yanaginuma, E.; Yonemutsu, O.; Nishida,
A.; Kawahara, N. J. Am. Chem. Soc. 1994, 116, 6455–
6456; (h) Badone, D.; Bernassau, J.-M.; Cardamone, R.;
Guzzi, U. Angew. Chem., Int. Ed. 1996, 35, 535–538; (i)
Enholm, E. J.; Cottone, J. S. Org. Lett. 2001, 3, 3959–
3962.
10. (a) Nishida, M.; Hayashi, H.; Yamaura, Y.; Yanaginuma,
E.; Yonemitsu, O. Tetrahedron Lett. 1995, 36, 269–272; (b)
Nishida, M.; Ueyama, E.; Hayashi, H.; Ohtake, Y.;
Yamaura, Y.; Yanaginuma, E.; Yonemutsu, O.; Nishida,
A.; Kawahara, N. J. Am. Chem. Soc. 1994, 116, 6455–
6456; (c) Badone, D.; Bernassau, J.-M.; Cardamone, R.;
Guzzi, U. Angew. Chem., Int. Ed. 1996, 35, 535–538; (d)
Enholm, E. J.; Cottone, J. S.; Allais, F. Org. Lett. 2001, 3,
145–147; (e) Keum, G.; Kang, S. B.; Kim, Y.; Lee, E. Org.
Lett. 2004, 6, 1895–1897.
We thank CONACyT and PROMEP-BUAP for finan-
cial support. Authors also thank Dr. Xianhai Huang
(from Schering Plough) for helpful discussions.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2007.04.054.
References and notes
1. (a) Yuen, P.; Kanter, G. D.; Taylor, C.-P.; Vartanian, M.
G. Bioorg. Med. Chem. Lett. 1994, 4, 823–826; (b) Bryans,
J. S.; Davies, N.; Gee, N. S.; Dissanayake, V. U. K.;
Ratcliffe, G. S.; Horwell, D. C.; Kneen, C. O.; Morrel, A.
I.; Oles, R. J.; O’Toole, J. C.; Perkins, G. M.; Singh, L.;
Suman-Chauhan, N.; O’Neill, J. A. J. Med. Chem. 1998,
41, 1838–1845; (c) Butters, M.; Catterick, D.; Craig, A.;
Curzons, A.; Dale, D.; Gillmore, A.; Green, S. P.;
Marziano, I.; Sherlock, J.-P.; White, W. Chem. Rev.
2006, 106, 3002–3027.
2. Taylor, C. P. In New Trends in Epilepsy Management;
Chadwick, D., Ed.; Royal Society of Medicine Services
Ltd: London, 1993; pp 13–40.
3. Hoekstra, M. S.; Sobieray, D. M.; Schwindt, M. A.;
Mulhern, T. A.; Grote, T. M.; Huckabee, B. K.; Hend-
rickson, V. S.; Franklin, L. C.; Granger, E. J.; Karrick, G.
L. Org. Process Res. Dev. 1997, 1, 26–38.
4. (a) Burk, M. J.; de Koning, P. 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, 5731–5734; (b) Hoge, G.; Wu,
H.-P.; Kissel, W. S.; Pflum, D. A.; Greene, D. J.; Bao, J. J.
Am. Chem. Soc. 2004, 126, 5966–5967.
11. (a) Meyers, A. I.; Snyder, L. J. Org. Chem. 1993, 58, 36–
42; (b) Kamimura, A.; Omata, Y.; Tanaka, K.; Shirai, M.
Tetrahedron 2003, 59, 6291–6299.
12. The existence of a weak internal hydrogen bond between
Ha and the carbonyl group has been determined by
computational calculations for chiral amide 5. See Ref. 5.
Additionally see the work of Hanessian, which has been
invoked hydrogen bonding as a stereocontrol element in
radical reactions: Hanessian, S.; Yang, H.; Schaum, R. J.
Am. Chem. Soc. 1996, 118, 2507–2508.