Stereoselective synthesis of 4-hydroxy-2-phenylproline framework

Article abstract of DOI:10.1016/j.tetlet.2004.12.130

Highly diastereoselective (>20:1) bromo-lactonization of N-sulfonyl-2-allyl-2-phenylglycine methyl ester (11) was observed. Successive treatment of the chiral lactone with MeONa gave the desired (2S,4R)-4-hydroxy-2- phenylproline derivative in high yield without erosion of the diastereoselectivity. The starting chiral non-racemic compound (5) was prepared from the racemic 2-phenylglycine using a classical kinetic resolution (crystallization), an asymmetric phase transfer alkylation, and an enzyme-catalyzed kinetic resolution.

Full text of DOI:10.1016/j.tetlet.2004.12.130

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 1545–1549
Stereoselective synthesis of 4-hydroxy-2-phenylproline framework
Kenji Maeda,a,* Ross A. Miller,b Ronald H. Szumigala, Jr.,b Ali Shafiee,b
Sandor Karadyb and Joseph D. Armstrong, IIIb
aProcess Research, Process R&D, Banyu Pharmaceutical Co., Ltd, 9-1 Kamimutsuna 3-chome, Okazaki, Aichi 444-0858, Japan
bDepartment of Process Research, Merck & Co., Inc. PO Box 2000, Rahway, NJ 07065, USA
Received 25 November 2004; revised 22 December 2004; accepted 24 December 2004
Available online 25 January 2005
Abstract—Highly diastereoselective (>20:1) bromo-lactonization of N-sulfonyl-2-allyl-2-phenylglycine methyl ester (11) was
observed. Successive treatment of the chiral lactone with MeONa gave the desired (2S,4R)-4-hydroxy-2-phenylproline derivative
in high yield without erosion of the diastereoselectivity. The starting chiral non-racemic compound (5) was prepared from the race-
mic 2-phenylglycine using a classical kinetic resolution (crystallization), an asymmetric phase transfer alkylation, and an enzyme-
catalyzed kinetic resolution.
Ó 2005 Elsevier Ltd. All rights reserved.
RO
Proline is an important component of many bioactive
natural compounds.1 For the past few decades, highly
functionalized proline derivatives have been used as chi-
ral synthons for the development of pharmaceuticals.2
In this regard, chiral 2-substituted prolines have not
only been useful in the synthesis of new compounds with
novel structures, but they have also been extensively
used as ligands in synthetic organic chemistry.3 Despite
the well-developed stereoselective synthetic methods for
the preparation of chiral proline derivatives,4 the prepa-
ration of chiral 2-arylprolines has been limited, mainly
due to the difficulty of simple SN2 type substitution.5
Ph
N
R
CO2Me
Figure 1.
1, was envisioned. The extension of their chemistry to
the quarternary amino acid derivative 2 was expected
to maintain the chiral center and form compound 1.
Chiral compound 2 would be synthesized from racemic
2-phenylglycine methyl ester hydrochloride (3).
During the course of our efforts toward the development
of our drug candidate, a stereoselective synthesis of
(2S,4R)-4-hydroxy-2-phenylproline framework (Fig. 1)
was required as a core structure. Herein an efficient syn-
thetic method for the preparation of this compound is
described.
For the preparation of (S)-2-allyl-2-phenylglycine ester
(5, 8), three different synthetic approaches, as outlined
in Scheme 2, were initially explored. The racemic com-
pound (4) was prepared by the published procedures5b
and the screening of chiral resolution using a library
of chiral acids was carried out. D-Tartaric acid was iden-
tified to obtain the diastereomeric salt with the desired
(S)-isomer (5). Under the optimal conditions, D-tartaric
acid salt of 5 could be obtained in over 95% ee and 33%
isolated yield after a single crystallization.7 Next, the
enzyme-catalyzed hydrolysis of 4 was investigated. A
library of hydrolytic enzymes were screened and several
enzymes, including reported pig liver esterase (PLE),
were identified as candidates.5b Among the identified en-
zyme catalysts, a microbial esterase from Bacillus stearo-
morphilus (BS3)8 was discovered that hydrolyzed the
ester to the desired acid enantiomer, and left behind
the undesired ester enantiomer (97% ee with 49% yield).
More than a decade ago, Ogasawara and co-workers re-
ported a single-step synthesis of (2S,4R)-4-hydroxypro-
line from N-homoallylamide by an iodine-mediated
consecutive cyclization.6 Based on the reported method-
ology, a retrosynthetic approach, as outlined in Scheme
Keywords: 4-Hydroxyproline; Optical resolution; Phase transfer reac-
tion; Iodo cyclization; Halo lactonization.
*
Corresponding author. Tel.: +81 564 51 5668; fax: +81 564 51
7086; e-mail: kenji_maeda@merck.com
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.12.130

1546
K. Maeda et al. / Tetrahedron Letters 46 (2005) 1545–1549
AcO
HCl·NH2
CO2Me
AcHN
CO2Me
Ph
Ph
Ph
N
H
CO2Me
1
2
3
Scheme 1.
H2N
CO2Me
Ph
H2N
CO2Me
Ph
5 ·D-Tartaric acid
D-Tartaric acid
33%, >95%ee
4
H2N
CO2H
Esterase from BS3
+
ent. 5
Ph
6
(49%, >97%ee)
Asymmetric PT reaction see ref. 9b
Cl-Ph
(S,S)-3,4,5-trifluorophenyl-NAS-bromide
H2N
H2N
CO2tBu
Ph
N
CO2tBu
ref. 9b
Br
77%, 91%ee
(literature 81%, 97%ee)
Ph
7
8
5
Cl-Ph
N
CO2tBu
Ph
CO2Me
Ph
H2SO4
MeOH, 50°C
72%
9
Scheme 2.
NOE
AcO
AcO
I2
AcHN
CO2Me
Ph
Ph
Ph
Ac2O
+
5
N
H
10a
N
H
10b
CO2Me
CO2Me
iPAC-H2O
90%
EtOAc-NaHCO3aq
>90%
2.5 : 1
2
Scheme 3.
The third approach took advantage of the previously
reported studies by Maruoka and co-workers.9b,10 tert-
Butyl ester derivative (8) was synthesized by the
asymmetric phase transfer alkylation9a with their cata-
lyst in high ee and yield. In the work-up sequence, both
the hydrolysis and trans-esterification of benzylimine
intermediate (9) occurred in one pot by the addition of
H2SO4 in MeOH at 50 °C to afford the methyl ester (5).
and selenium, mediated 5-endo cyclization of the C-
allylic secondary amine or amide has become a general
tool to construct pyrrolidine skeleton.11 Thus, this ap-
proach was adopted to construct the 4-halo-pyrrolidine
framework (Scheme 4).
After making a sulfonamide from 5,12 the 5-endo-halo-
cyclization was attempted. No reaction was observed
by the treatment with either NIS or NBS. After several
conditions and reagents were tried, we found that the
reaction proceeded rapidly by the addition of Br2 in
CH2Cl2 at À10 °C. Unexpectedly, the resultant products
appeared to be c-lactones 13a and 13b, formed in high
stereoselectivity instead of the predicted pyrrolidine
(Scheme 5). The major isomer of the lactones possessed
the desired stereochemistry. The stereochemistry and
After acetylation of the resulting amine (5) with acetic
anhydride, 2 was subjected to the iodine cyclization con-
ditions (Scheme 3). The reaction was first attempted
using in THF–H2O as reported in the literature.6a While
the reaction proceeded as expected, a number of impuri-
ties and intermediates were observed. A variety of condi-
tions were then screened and the two phase reaction
conditions (isopropyl acetate–H2O) were observed to
be the cleanest. The product was obtained as a diastereo-
meric mixture at the 4 position in a 2.5:1 ratio. The ob-
served selectivity was not sufficient for the further
manipulations. The stereochemistry and the ratio were
determined by NOE experiment and HPLC, respectively.
1
ratio were determined by NOE and H NMR analysis.
RO
X
RHN
CO2Me
Ph
Ph
Ph
N
R
N
R
CO2Me
CO2Me
Due to the above results, an alternative synthetic route
was formulated. To date, electrophile, such as halogen
Scheme 4.

K. Maeda et al. / Tetrahedron Letters 46 (2005) 1545–1549
1547
Ph
Ph
NOE
Br
TsHN
CO2R
Ph
NHTs
O
NHTs
O
TsCl
Br2, CH2Cl2
+
Br
5
or 8
O
O
NMe2
H
H
-10°C, 5min
> 80%
Me2N
3
11: R = Me
13a
13b
13 : 1 from 11
4 : 1 from 12
MeCN, rt
quant.
12: R = tBu
Ph
NHTs
OR
Br+
O
14
from Methyl ester (11)
HO
13a : 13b
Temp./ °C
Ph
NaH, MeOH
rt, 2h
40
25
-10
-70
5.5 : 1
6 : 1
13 : 1
>20 : 1
N
CO2Me
Ts
15
80% (2 steps)
Scheme 5.
Although a halo-lactonization with carboxylic acid or
amide in the presence of its secondary amide or carba-
mate has been well established,13 there are few examples
that the ester reacts with allyl cation rather than the sec-
ondary amide. The present unexpected results seem to
come from the rigid and stabilized transition state
caused by the tosyl amide and phenyl group, which
would create a suitable conformation for the ester
attack.13b,14
Next, the temperature effect of the bromo-cyclization
was investigated. Under reflux conditions (40 °C) and
25 °C, the resultant lactone showed 5.5:1 and 6:1 selecti-
vity, respectively, at c-position, compared to 13:1 at
À10 °C. At À70 °C, we observed selectivity of 20:1.15
The Br2 cyclization was applied to tert-butyl ester deriva-
tive (12) under the exact same conditions. However, the
ratio dropped to 4:1 as compared to 13:1 for the methyl
ester (Scheme 5). The steric repulsion of the phenyl and
tert-butyl groups apparently affected the transition state
of the cyclization.14
The obtained c-lactones were readily transformed into
the desired 4-hydroxyproline methyl ester by treatment
with NaH in MeOH at room temperature without any
by-products. During the reaction, no intermediates were
detected in HPLC and TLC. We speculated that once
the lactone was opened by methoxide, a successive
ring-closure reaction would occur immediately. After
silica gel purification, (2S,4R)-4-hydroly-2-phenylpro-
line derivative (15) was isolated in 80% from 11.
The above lactonization was similarly carried out with
iodine by aging for 20 h at room temperature (25 °C)
and the resultant iodo-lactone was transformed to
the desired hydroxy proline in the same manner as the
bromo-lactone (Scheme 6). However, the observed
selectivity was only 2.4:1, lower than that of
Ph
Ph
NHTs
O
NHTs
O
TsHN
CO2Me
Ph
I2, CH2Cl2
+
I
I
O
O
H
H
25°C, 20h
quant.
2.4 : 1
11
17a
17b
Ph
NHTs
O
I
OH
18
OMe
HO
Ph
CO2Me
NaH, MeOH
N
rt, 2h
71%
Ts
15
Scheme 6.

1548
K. Maeda et al. / Tetrahedron Letters 46 (2005) 1545–1549
M.; Kato, M. J. Antibiotics 1990, 43, 519–532; (f) Iso, Y.;
Ph
Irie, T.; Nishino, Y.; Motokawa, K.; Nishitani, Y. J.
Antibiotics 1996, 49, 199–209; (g) Ohtake, N.; Okamoto,
O.; Mitomo, R.; Kato, Y.; Yamamoto, K.; Haga, Y.;
Fukatsu, H.; Nakagawa, S. J. Antibiotics 1997, 50, 598–
613; (h) Kress, M. H.; Yang, C.; Yasuda, N.; Grabowski,
E. J. J. Tetrahedron Lett. 1997, 38, 2633.
NHTs
O
TsHN
a) I2, CH2Cl2
CO2Me
Ph
HO
I
or
O
I
b) TsOH·H2O
CH2Cl2
H
18
17
quant.
3. (a) Harrison, T.; Williams, B. J.; Swain, C. J. BioMed.
Chem. Lett. 1994, 4, 2733; (b) Chatani, H.; Nakajima, H.;
Kawasaki, H.; Koga, K. Heterocycles 1997, 46, 53.
4. (a) Seebach, D.; Boes, M.; Naef, R.; Schweizer, W. B. J.
Am. Chem. Soc. 1983, 105, 5390; (b) Seebach, D.; Weber,
T. Helv. Chim. Acta 1985, 68, 155; (c) Noe, C. R.;
Knollmuller, M.; Vollenkle, H.; Noe-letschnig, M.; Wei-
gand, A.; Muhl, J. Pharmazie 1996, 51, 800; (d) Nagano,
S.; Mizukami, M.; Akutsu, A.; Kawahara, N. Tetrahedron
Lett. 1999, 40, 3209; (e) Sato, T.; Kawasaki, S.; Oda, N.;
Yagi, S.; El Bialy, S. A. A.; Uenishi, J.; Yamauchi, M.;
Ikeda, M. J. Chem. Soc., Perkin Trans. 1 2001, 2623; (f)
Tamura, O.; Yanagimachi, T.; Kobayashi, T.; Ishibashi,
H. Org. Lett. 2001, 3, 2427.
Scheme 7.
bromo-lactonization under the same conditions. Based
on analysis of the reaction mixture by LC-MS, we
observed a different intermediate than what had been
predicted in the case of bromine. In this case, the reac-
tion went through the iodohydrin intermediate (18). A
trace amount of water in the solvent accelerated the
reaction.16 In order to provide more evidence for this
mechanism, the iodohydrin (18) was isolated by a silica
gel column chromatography and transformed to the
iodo-lactones (17) under the original reaction conditions
(Scheme 7). In addition, the isomer ratio of the iodohy-
drin was 2.4:1, identical to the ratio observed in the
iodo-lactones (17). The lactonization was promoted by
the treatment of TsOHÆH2O as well.
5. (a) Semmelhack, M. F.; Hall, H. T.; Yoshifuji, M.; Clarck,
G. J. Am. Chem. Soc. 1975, 97, 1247; (b) Betsbrugge, J. V.;
Tourwe, D.; Kaptein, B.; Kierkels, H.; Broxterman, R.
Tetrahedron 1997, 53, 9233.
6. (a) Takano, S.; Iwabuchi, Y.; Ogasawara, K. J. Chem.
Soc., Chem. Commun. 1988, 1527; (b) Takano, S.; Iwabu-
chi, Y.; Ogasawara, K. Heterocycles 1989, 29, 1861.
7. Enantiopurity was determined by chiral SFC analysis
using Chiralpak OJ with MeOH–CO2 as an eluent after
acetylation of 5. Its absolute configuration was determined
by conversion of known tert-butyl ester (7) into its methyl
ester as mentioned in Scheme 2 followed by comparison of
the retention time of Chiral SFC analysis after acetylation.
8. The enzyme BS3 was obtained from Julich Fine Chemical,
Julich, Germany.
9. (a) Dolling, U.-H.; Davis, P.; Grabowski, E. J. J. J. Am.
Chem. Soc. 1984, 106, 446; (b) Ooi, T.; Takeuchi, M.;
Ohara, D.; Maruoka, K. Synlett 2001, 1185.
10. Enantiopurity was determined by chiral SFC analysis
using Chiralpak AS-H with MeOH–CO2 as an eluent after
tosylation of 7, as mentioned in Scheme 3.
11. (a) Tamaru, Y.; Kawamura, S.; Tanaka, K.; Yoshida, Z.
Tetrahedron Lett. 1984, 25, 1063; (b) Tamaru, Y.;
Kawamura, S.; Bando, T.; Tanaka, K.; Hojo, M.; Yos-
hida, Z. J. Org. Chem. 1988, 53, 5491; (c) Smaele, D. D.;
Kimpe, N. D. J. Chem. Soc., Chem. Commun. 1995, 2029;
(d) Lee, W. S.; Jang, K. C.; Kim, J. H.; Park, K. H. Chem.
Commun. 1999, 251; (e) Knight, D. W.; Redfern, A. L.;
Gilmore, J. J. Chem. Soc., Perkin Trans. 1 2001, 2874; (f)
Outurquin, F.; Pannecoucke, X.; Berthe, B.; Paulmier, C.
Eur. J. Org. Chem. 2002, 1007; (g) McCormick, J. L.;
Osterman, R.; Chan, T.; Das, P. R.; Pramanik, B. N.;
Ganguly, A. K.; Girijavallabhan, V. M.; McPhail, A. T.;
Saksena, A. K. Tetrahedron Lett. 2003, 44, 7997; (h) Kim,
J. H.; Long, M. J. C.; Kim, J. Y.; Park, K. H. Org. Lett.
2004, 6, 2273.
In conclusion, we report the highly stereoselective syn-
thesis of (2S,4R)-4-hydroxy-2-phenylproline derivative
achieved by developing chiral syntheses of 2-allyl-2-phen-
ylglycine derivatives followed by stereoselective bro-
mine-mediated cyclization. During the investigation,
we revealed that the halo-lactonization of compound
(5) proceeded via the different pathway depending on
the halogen (Br2 or I2). A key finding of our studies
was that the ester could react with an epi-bromonium
cation much faster than the NH-Ts group under these
particular conditions. Additionally, we observed a
strong dependence of stereoselectivity with temperature,
with >20:1 selectivity at À70 °C.
Acknowledgements
We would like to thank Mr. Robert Reamer for assis-
tance and valuable discussion with NMR experience,
Ms. Jennifer Chilenski for the chiral acid salt screening
and Ms. Mirlinda Biba for the chiral assays.
References and notes
1. (a) Cody, W. L.; Wilkes, B. C.; Muska, B. J.; Hruby, V. J.;
Castrucci, A. M. D.; Hadley, M. E. J. Med. Chem. 1984,
27, 1186; (b) Prenner, E. J.; Lewis, R.; McElhaney, R. N.
Biochim. Biophys. Acta—Biomembr. 1999, 1462, 201.
2. (a) Ondetti, M. A.; Cushman, D. W. Science 1977, 196,
441; (b) Shibata, T.; Iino, K.; Sugimura, Y. Heterocycles
1986, 24, 1331; (c) Renand, P.; Seebach, D. Synthesis 1986,
424; (d) Krapcho, J.; Turk, C.; Cushman, D. W.; Powell,
J. R.; Deforrest, J. M.; Spitzmiller, E. R.; Karanewsky, D.
S.; Duggan, M.; Rovnyak, R.; Schwartz, J.; Natarajan, S.;
Godfrey, J. D.; Ryono, D. E.; Neubeck, R.; Atwal, K. S.;
Petrillo, E. W. J. Med. Chem. 1988, 31, 1148; (e)
Sunagawa, M.; Matsumura, H.; Inoue, T.; Fukasawa,
12. Yoshida, Y.; Shimonishi, K.; Sakakura, Y.; Okada, S.;
Aso, N.; Tanabe, Y. Synthesis 1999, 1633.
13. (a) Ohfune, Y.; Kurokawa, N. Tetrahedron Lett. 1985, 26,
5307; (b) Ohfune, Y.; Hori, K.; Sakaitani, M. Tetrahedron
Lett. 1986, 27, 6079; (c) Kitagawa, O.; Sato, T.; Taguchi,
T. Chem. Lett. 1991, 177; (d) Guillerm, D.; Guillerm, G.
Tetrahedron Lett. 1992, 33, 5047; (e) Kitagawa, O.;
Hanano, T.; Kikuchi, N.; Taguchi, T. Tetrahedron Lett.
1993, 34, 2165; (f) Rousseau, G.; Robin, S. Tetrahedron
1998, 54, 13681; (g) Burger, K.; Mutze, K.; Osipov, S. N.;
Tsouker, P.; Schier, A. Monatshefte fur Chemie 2003, 134,
69.

K. Maeda et al. / Tetrahedron Letters 46 (2005) 1545–1549
1549
Ph
14. A plausible transition state. See Ref. 13b.
>20 : 1
Br
2 : 1
Br
NHTs
O
TsHN
*
TsHN
CO2Me
Ph
CO2Me
Ph
Br2, CH2Cl2
-70°C, 5min
HO
+
H
*
O
NTs
H
Br
+
11
13
16
32%
Ph
OMe
61%
O
.
.
16. The iodo-lactonization completed more quickly in the
presence of water.
15. In this case, the bromohydrin (16) formed in 32% yield at
the same time. Probably, the bromo-lactonization step
might be competing with the attack of trace amount of
water in the solvent to the bromonium cation intermediate
(14) due to the relatively low reactivity of the ester at that
temperature.
Ph
I2
NHTs
TsHN
CO2Me
Ph
CH2Cl2-H2O (10:1)
I
O
O
25ºC, 5h
H
11
17
100% conversion
.