J . Org. Chem. 2001, 66, 4939-4940
4939
4-Chlorophenyl-S-glycidyl ether 2 was conveniently
prepared on a large scale using the Sharpless procedure6
(Scheme 1). Thus, reaction of 4-chlorophenol and (2S)-
glycidyl 3-nitrobenzenesulfonate in the presence of so-
dium hydride afforded the desired ether derivative 2 in
80% yield.
Initially, we subjected this epoxide to react with liquid
ammonia in a sealed tube. Unfortunately, this epoxide
opening reaction with ammonia also gave the bis-alky-
lated product 5, requiring the tedious chromatographic
purification of the highly basic amino alcohol intermedi-
ate 4. To address this problem, we employed various
ammonia equivalents such as dibenzylamine and ben-
zhydrylamine in the epoxide-opening reaction. While
these reactions provided the desired protected amino
alcohol derivatives in good yields, efforts to unmask the
latent amino functionality using either phase-transfer
hydrogenation conditions or Pd/C catalysts were also
accompanied, to varying degrees, by the reduction of the
chloro substituent on the phenyl ring.
Faced with this difficulty, we explored the use of
tritylamine as a masked ammonia surrogate in the
reaction with the epoxide 2. While the use of tritylamine
as an ammonia equivalent in the electrophilic reactions
with halides is precedented,7 to the best of our knowledge,
the corresponding reaction with epoxides as electrophile
has not been described. Thus, reaction of 2 with trity-
lamine gave the protected amino alcohol 3 (R ) C(C6H5)3)
in 75% yield. Treatment of this compound with either
TFA in methylene chloride or HCl in methanol furnished
intermediate 4. Alternatively, we also employed success-
fully dibenzosuberylamine (DBS-NH2)8,9 as an ammonia
equivalent in the reaction with 2. Elaboration of the
intermediate 4 to 15 was straightforward as shown in
Scheme 1.
A Con ven ien t Syn th esis of th e Novel
Hyp oglycem ic Agen t SDZ P GU 693
Ranjit C. Desai
Department of Medicinal Chemistry, Merck Research
Laboratories, P.O. Box 2000, Rahway, New J ersey 07065
ranjit_desai@merck.com
Received March 5, 2001
Noninsulin-dependent diabetes mellitus (NIDDM, type
2) is a chronic, multifactorial metabolic disease, typically
characterized by insulin resistance in the liver and
peripheral tissues.1 The resulting hyperglycemia and
hyperlipidemia in type 2 diabetes patients may also
contribute to long-term organ complications such as
neuropathy, retinopathy, nephropathy, and atheroscle-
rosis.2 Thus, in diabetic patients, maintaining tight
control of blood glucose would be desirable to lessen the
risk of the above-mentioned complications.3 Recently,
researchers at Novartis laboratories published a report
describing a novel class of tetrahydropyrrolo[2,1-b]oxazol-
5(6H)-one based hypoglycemic agents.4 In this class, SDZ
PGU 693 (1) was reported to significantly improve
glucose metabolism in diabetic and insulin-resistant
animals. Though the mechanism of this novel class of
insulin-sensitizing agents as represented by 1 is not
known at present, it provides a novel approach to
improving glucose lowering in type 2 diabetic patients.
In conclusion, we have developed a convenient synthe-
sis of SDZ PGU 693. The unprecedented use of trity-
lamine as well as dibenzosuberylamine as ammonia
equivalents in the reaction with epoxide to provide 1,2-
amino alcohols should be quite useful in the synthesis of
other analogues of this pharmacologically important class
of compounds.
In connection with an ongoing project in these labora-
tories, multigram quantities of 1 were required. We found
the published patent procedure5 for the synthesis of this
compound to be less than optimal for various reasons.
First, it involves a tedious resolution procedure for the
synthesis of the key intermediate 3, and second, unlike
the reported procedure, in our hands, the reaction of the
chiral epoxide 2 with ammonia in the sealed tube was
accompanied with varying amounts of the bis-alkylated
product 5, thus necessitating the cumbersome chromato-
graphic purification of the highly basic amino alcohol
intermediate 4. We wish to report herein a facile syn-
thesis of 1.
Exp er im en ta l Section
3-(4-Ch lor op h en oxy)-2S-h yd r oxyp r op yla m in e (4). To a
solution of 4-cholrophenyl-S-(+)-glycidyl ether 2 (50 g, 0.27 mol)
i
in PrOH (60 mL) was added tritylamine (76.6 g, 0.29 mol), and
the resulting mixture was heated under reflux for 24 h.
The solution was allowed to cool to room temperature, and
the solvent was evaporated under reduced pressure. The residue
was purified by flash chromatography on a column of silica gel,
eluting with CH2Cl2 to give 90 g of 3 (R ) C(C6H5)3) as a gum:
1H NMR (500 MHz, CDCl3) δ 7.67-7.22 (m, 18H), 6.83 (d, 1H),
4.1 (m, 1H), 4.01 (m, 2H), 2.45 (m, 2H).
(1) Bennet, P. H. Diabet. Metab. Rev. 1997, 13, 583-605.
(2) Feldt-Rasmussen, B.; Mathiesen, E. R.; Deckert, T. Lancet. 1986,
II, 1300.
(6) Klunder, J . M.; Onami, T.; Sharpless, K. B. J . Org. Chem. 1989,
54, 1295-1304.
(3) J aspan, J . B. Metabolism. 1987, 36 (Suppl. 1), 22.
(4) Aicher, T. D.; Balkan, B.; Bell, P. A.; Brand, L. J .; Cheon, S. H.;
Deems, R. O.; Fell, J . B.; Fillers, W. S.; Fraser, J . D.; Gao, J .; Knorr,
D. C.; Kahle, G. G.; Leone, C. L.; Nadelson, J .; Simpson, R.; Smith, H.
C. J . Med. Chem. 1998, 41, 4556-4566.
(7) Huszthy, P.; Bradshaw, J . S.; Krakowiak, K. E.; Wang, T.; Dalley,
N. K. J . Heterocycl. Chem. 1993, 5, 1197-1208.
(8) Pless, J . Helv. Chim. Acta 1976, 59, 499-512.
(9) To the best of our knowledge, there is only one instance of epoxide
opening reaction with DBS-NH2. However, in this case, DBS-NH2 was
not used as an ammonia surrogate. Vejdelek, Z.; Protiva, M. Collect.
Czech. Chem. Commun. 1990, 55, 1290-129.
(5) Aicher, T. D.; Cheon, S. H.; Nadelson, J .; Simpson, R.; Houlihan,
W. J . Eur. Patent 0 702 015 A1, 1996.
10.1021/jo010239d CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/31/2001
Desai
