Synthesis of oxazolidinyl azacycles via ring-closing olefin metathesis: A practical entry to the synthesis of deoxy-azasugars and hydroxypyrrolizidines

Article abstract of DOI:10.1016/S0040-4039(01)00635-9

Starting with D- and L-serines, an expedient method for the preparation of oxazolidinyl piperidines, azepenes and azacyclooctenes was illustrated as a route to various deoxy-azasugars and hydroxypyrrolizidines. The ring-closing olefin metathesis of oxazolidinyl di-olefins was used as a key-step to construct the azacycles. Consecutive epoxidation, hydrolysis and transannulation of oxazolidinyl azacyclooctene led to hydroxypyrrolizidines.

Full text of DOI:10.1016/S0040-4039(01)00635-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 4079–4082
Synthesis of oxazolidinyl azacycles via ring-closing olefin
metathesis: a practical entry to the synthesis of deoxy-azasugars
and hydroxypyrrolizidines
Thangaiah Subramanian, Chang-Ching Lin and Chun-Cheng Lin*
Institute of Chemistry, Academia Sinica, Nankang, Taipei, Taiwan 11529, ROC
Received 4 January 2001; revised 5 April 2001; accepted 12 April 2001
Abstract—Starting with
D- and L-serines, an expedient method for the preparation of oxazolidinyl piperidines, azepenes and
azacyclooctenes was illustrated as a route to various deoxy-azasugars and hydroxypyrrolizidines. The ring-closing olefin
metathesis of oxazolidinyl di-olefins was used as a key-step to construct the azacycles. Consecutive epoxidation, hydrolysis and
transannulation of oxazolidinyl azacyclooctene led to hydroxypyrrolizidines. © 2001 Elsevier Science Ltd. All rights reserved.
Many (deoxy-)azasugars and hydroxypyrrolizidine
alkaloids have been shown to selectively inhibit
oligosaccharide processing enzymes.1 These properties
make them potentially chemotherapeutic agents for
treating diabetes, cancer, and viral (e.g. HIV and influ-
enza) infections.1,2 The therapeutic importance of these
compounds has stimulated much synthetic effort
toward their preparation.1a,3 Oxazolidinyl piperidine 1a
has recently been reported to serve as an important
intermediate for the synthesis of 1-deoxymanno-
jirimycin (2) and 1-deoxyaltromycin (3) as well as 1b
for 1-deoxygalactostatin (4).4 In addition, compounds
1a/1b can also be used in the synthesis of hydroxy-
lated 2-oxaindolizidines, which are inhibitors of glycosi-
dases5 (Fig. 1).
di-olefin oxazolidinones serves as a key-step in our
synthetic schemes. Our method provides diversified
stereoisomers that are useful in the SAR study of
glycosidase inhibition.
O
O
O
a, b
OMe
OMe
c
O
O
O
O
HO
OH
NH
N
NH2
5
6
7
R1
R2
OH
e
d
f
O
O
O
1a/1b
N
N
O
8
9a R1 = OH, R2 = H
9b R1 = H, R2 = OH
However, the literature methods for the synthesis of
1a/1b and their analogs require several steps and
resulted in low overall yields.4,6,7a As a continuation of
our efforts to develop glycosidase inhibitors, we report
herein, an efficient and practical synthesis of oxazo-
Scheme 1. (a) SOCl2, MeOH, reflux, overnight; (b) K2CO3,
triphosgene, H2O/toluene, rt (86% in two steps); (c) NaH,
allyl bromide, DMF, 65%; (d) NaBH4, MeOH, 74%; (e) (i)
DMSO, (COCl)2, CH2Cl2, EtN(i-Pr)2; (ii) vinylmagnesium
bromide, CH2Cl2 (53% in two steps); (f) Grubb’s catalyst (10
mol%), CH2Cl2, rt, 96% (1a:1b=1:1.3)
lidinylpiperidine 1a/1b and their analogs from
-serines. The ring-closing olefin metathesis (RCM) of
D
- and
L
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
O
O
HO
O
O
HO
HO
N
N
HN
HN
HN
4
1b
1a
2
3
Figure 1.
* Corresponding author. Tel.: 886-2-27898648; fax: 886-2-27831237; e-mail: cclin@chem.sinica.edu.tw
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00635-9

4080
T. Subramanian et al. / Tetrahedron Letters 42 (2001) 4079–4082
As shown in Scheme 1,
D
-serine 5 was esterified with
CH2Cl2 at room temperature for 2 h afforded quantita-
thionyl chloride in methanol and then converted into
the more stable chiral oxazolidinone 68 by the treat-
ment with triphosgene in the presence of K2CO3 (86%
in two steps). N-Allylation of 6 using allylbromide
yielded 7 (65%), in which the ester group was subse-
quently reduced by NaBH4 to give alcohol 8 (84%).
Swern oxidation of 8, followed by addition of vinyl-
magnesium bromide in a one-pot procedure gave the
desired di-olefin compounds 9a/9b as an inseparable
mixture (53% in two steps). RCM of 9a/9b using
Grubbs’ catalyst,9 Cl2(PCy3)2Ru=CHPh, (10 mol%) in
tive yields of oxazolidinyl piperidines 1a/1b (1:1.3),
which were easily separated on a silica gel column.
Thus, the important azasugar precursors 1a/1b were
prepared in 21% total yield by a six-step sequence.
By a procedure similar to that for 1a/1b, L-serine was
converted into di-olefin 10a/10b (Table 1). When allyl-
magnesium bromide was used instead of vinylmagne-
sium bromide, di-olefins 11a/11b were obtained.
Hydroxyl di-olefins 12a/12b and 13a/13b were prepared
by changing the N-alkylating agent to 4-bromo-1-
a
Table 1.
anti/syn
(yield)
Reaction
time
Products
Entry
1
Reactant
R1
syn
anti
OH
R2
HO Hb
Ha
1:1.3
(96)
2 h
5 h
4 h
O
O
O
N
N
N
O
O
O
10a R1=H, R2=OH
10b R1=OH, R2=H
16a
16b
Jab = 9.0 Hz
Hb
HO
R1
HO
Ha
R2
1:1
(94)
2
3
O
O
O
N
N
N
O
O
O
11a R1=H, R2=OH
11b R1=OH, R2=H
17b
17a
Jab = 7.3 Hz
Hb
HO
R1
HO
Ha
R2
1:1.3
(93)
O
O
O
N
N
N
O
O
O
12a R1=H, R2=OH
12b R1=OH, R2=H
18b
18a
Jab = 8.5 Hz
HO
Ha
Hb
HO
R1
R2
24 hb
1:1.2
(21)
4
5
O
O
O
N
N
N
O
O
O
13a R1=H, R2=OH
13b R1=OH, R2=H
19b
19a
Jab = 8.4 Hz
BnO
Ha
Hb
BnO
R1
R2
1:1.2
(64)
4 h
O
O
O
N
N
N
O
O
O
14a R1=H, R2=OBn
14b R1=OBn, R2=H
20b
20a
J
ab = 9.1 Hz
OBn
Hb
BnO
R1
R2
Ha
1:1.2
(64)
4 h
6
O
O
O
N
N
N
O
O
O
15a R1=H, R2=OBn
15b R1=OBn, R2=H
21b
21a
Jab = 9.1 Hz
aRCM was carried out in 10mol% of catalyst in CH2Cl2 at rt.
bReaction was slow at rt and heating to reflux resulting in polymerisation of the starting
material.

T. Subramanian et al. / Tetrahedron Letters 42 (2001) 4079–4082
4081
We also had access to hydroxypyrrolizidine alkaloids
by using the eight-membered ring compounds as illus-
trated in Scheme 2. The compound 20a was treated
with m-chloroperoxybenzoic acid to give epoxides 22a
and 22b (1:1). Compound 22a was shown by single
crystal diffraction to contain the epoxy group on the
exo face. Treatment of 22a with LiOH thus provided
the desired hydroxypyrrolizidine 23a11 in 73% yield as a
consequence of oxazolidinone hydrolysis and trans-
annular cyclization.12 Similarly compound 22b was con-
verted into hydroxypyrrolizidine 23b.
In summary, we have successfully converted
D- and
Figure 2. ORTEP diagram of compounds 17b and 22a.
L
-serines to a variety of azacyclic compounds by using
ring-closing olefin metathesis as a key-step. This
method provides a structural diversity for the synthesis
of many stereoisomers of deoxy-azasugars and hydroxy-
pyrrolizidines, suitable for further studies of their glyco-
sidase inhibition.
BnO
HO
H
c
c
OBn
OH
O
O
O
O
N
N
BnO
22a
23a
a
O
O
BnO
N
HO
H
Acknowledgements
OBn
OH
O
O
20a
N
N
23b
Financial support has been generously provided by
National Science Council (NSC 89-2113-M-001-062)
and Academia Sinica. We thank Professor J.-M. Fang
for useful discussions. We also thank Dr. H.-M. Lin
and Mr. Y.-S. Wen for X-ray crystallographic analysis.
22b
Scheme 2. (a) MCPBA, 81%; (b) LiOH, EtOH/H2O (1:1),
90°C, 3 h, 73%.
butene and following a similar procedure to that previ-
ously described. In addition, the hydroxyl groups of
13a/13b were protected with benzylbromide to give
References
14a/14b. Similarly, starting from
D-serine, compounds
1. (a) Stu3 tz, A. E. Iminosugars as Glycosidase Inhibitors:
15a/15b were synthesized. As shown in Table 1, the
RCM of the di-olefins gave excellent to moderate yields
of the corresponding products. However, in the pres-
ence of a free hydroxyl group, the formation of the
eight-membered ring compounds, 19a/19b from 13a/
13b, occurred in low yields (21%, entry 4). The unex-
pected low reactivity of these compounds may be due
to conformational constraint resulting in deactivation
of the catalyst by the nearby free hydroxyl group.10 In
order to eliminate the influence of the hydroxyl group,
it was protected as the benzyl ether (14a/14b) with the
results showing a significant increase in the yield (64%,
entry 5) of product, thus, demonstrating that the
hydroxyl group near the olefin plays a role in RCM. It
should be noted that this phenomenon was not
observed in the cases of formation of six- and seven-
membered rings.
Nojirimycin and Beyond; Wiley-VCH: Weinheim, 1999;
(b) Jacob, G. S. Curr. Opin. Struct. Biol. 1995, 5, 605; (c)
Winchester, B.; Fleet, G. W. J. Glycobiology 1992, 2, 199;
(d) Sinnott, M. L. Chem. Rev. 1990, 90, 1171.
2. (a) Bols, M. Acc. Chem. Res. 1998, 31, 1; (b) Davies, G.;
Sinnott, M. L.; Wither, S. G. In Comprehensive Biological
Catalysis; Sinnott, M. L., Ed.; Academic Press: London,
1998; p. 119; (c) van der Broek, L. A. G. M. In Carbohy-
drates in Drug Design; Witczak, Z. J.; Nieforth, K. A.,
Eds.; Marcel Dekker: New York, 1997; p. 471.
3. Carbohydrate Mimics: Concepts and Methods; Chapleur,
Y., Ed.; Wiley-VCH: Weinheim, 1998.
4. Asano, K.; Hakogi, T.; Iwama, S.; Katsumura, S. Chem.
Commun. 1999, 41.
5. D´ıaz Pe´rez, V. M.; Garc´ıa Moreno, M. I.; Ortiz Mellet,
C.; Fuentes, J.; D´ıaz Arribas, J. C.; Can˜ada, F. J.; Garc´ıa
Ferna´ndez, J. M. J. Org. Chem. 2000, 65, 136.
6. (a) Agami, C.; Couty, F.; Rabasso, N. Tetrahedron Lett.
2000, 41, 4113; (b) Shirai, M.; Okamoto, S.; Sato, F.
Tetrahedron Lett. 1999, 40, 5331; (c) Ciufolini, M. A.;
Hermann, C. Y. W.; Dong, Q.; Shimizu, T.; Swami-
nathan, S.; Xi, N. Synlett 1998, 105.
7. (a) Martin, R.; Moyano, A.; Pericas, M. A.; Riera, A.
Org. Lett. 2000, 2, 93; (b) White, J. D.; Hrnciar, P.;
Yokochi, A. F. T. J. Am. Chem. Soc. 1998, 120, 7359.
8. Falb, E.; Bechor, Y.; Hassner, A.; Albeck, A.; Gottlib, H.
E. J. Org. Chem. 1999, 64, 498.
The structures of 16a (anti-isomer) and 16b (syn-iso-
mer) were determined by comparing their 1H NMR
spectra with those of enantiomeric counterparts 1a/1b.
Compound 17b11 was crystallized from hexane/EtOAc,
and the syn-configuration was unambiguously deter-
mined by X-ray diffraction (Fig. 2). The stereochem-
1
istry of the other isomers were established by the H
NMR analysis. For the syn-isomers (18b, 19b, 20b and
21b), the two protons on the stereogenic carbons did
not couple with each other due to their orthogonal
disposition.
9. (a) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413;
(b) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem.
Res. 1995, 28, 446.

4082
T. Subramanian et al. / Tetrahedron Letters 42 (2001) 4079–4082
10. (a) Lin, C.-C.; Subramanian, T.; Hus, T.-S.; Fan, G.-T.;
Lin, C.-C. J. Chin. Chem. Soc. 2001, 48, 55; (b) Hyldtoft,
L.; Madsen, R. J. Am. Chem. Soc. 2000, 122, 8444; (c)
Furstner, A.; Langemann, K. J. Org. Chem. 1996, 61,
8746.
4.32–4.40 (m, 2H), 5.94 (m, 1H), 6.03 (m, 1H). Com-
pound 19a: 1H NMR (400 MHz, CDCl3): l 2.11 (br, 1H),
2.29–2.40 (m, 2H), 2.42–2.50 (m, 1H), 2.55–2.65 (m, 1H),
3.11 (ddd, J=3.5, 8.2, 13.8Hz, 1H), 3.46 (ddd, J=2.5,
8.4, 11.50 Hz, 1H), 3.83 (br, 1H), 3.92 (ddd, J=3.50, 8.4,
13.8 Hz, 1H), 4.31–4.39 (2H, m), 5.78–5.87 (m, 2H).
Compound 19b: 1H NMR (400 MHz, CDCl3): l 2.15 (br,
1H), 2.29–2.65 (m, 4H), 3.19 (ddd, J=2.4, 6.3, 13.6 Hz,
1H), 3.88–3.97 (m, 3H), 4.28–4.38 (m, 1H), 4.47 (dd,
J=3.0, 8.2 Hz, 1H), 5.66–5.73 (m, 1H), 5.78-5.85 (m,
1H). Compound 20a: 1H NMR (400 MHz, CDCl3): l
2.29 (ddd, J=3.0, 7.3, 7.3, 14.8 Hz, 1H), 2.48-2.58 (m,
3H), 3.05 (ddd, J=3.0, 9.2, 13.8 Hz, 1H), 3.50–3.60 (m,
2H), 3.96 (ddd, J=3.4, 7.3, 13.8 Hz, 1H), 4.11 (dd,
J=2.0, 8.8 Hz, 1H), 4.28 (dd, J=7.3, 8.8 Hz, 1H), 4.45
(d, J=11.5 Hz, 1H), 4.72 (d, J=11.5 Hz, 1H), 5.73–5.88
(m, 2H), 7.30-7.38 (m, 5H). Compound 23a: 1H NMR
(400 MHz, CDCl3): l 2.0–2.31 (m, 6H), 3.0 (dd, J=10.1,
18.8 Hz, 1H), 3.25 (br, 1H), 3.38 (m, 2H), 3.65 (dd,
J=4.4, 11.3 Hz, 1H), 3.89 (m, 1H), 4.02 (dd, J=2.5, 5.1
Hz, 1H), 4.19 (m, 1H), 4.56 (dd, J=11.5, 18.6 Hz, 2H),
7.30–7.37 (m, 5H)
11. Selected data for 17a: 1H NMR (400 MHz, CDCl3): l
2.45–2.60 (m, 2H), 2.86 (br, 1H), 3.53–3.62 (m, 2H), 3.78
(ddd, J=5.6, 7.3, 8.6 Hz, 1H), 4.20 (dd, J=5.6, 8.9 Hz,
1H), 4.28 (dd, J=6.8, 16.7 Hz, 1H), 4.44 (dd, J=8.6, 8.9
1
Hz, 1H), 5.75 (m, 1H), 5.94 (m, 1H). Compound 17b: H
NMR (400 MHz, CDCl3): l 1.88 (d, J=10.6 Hz, 1H),
2.41 (dddd, J=2.1, 2.4, 6.5, 15.9 Hz, 1H), 2.76 (dddd,
J=1.0, 7.0, 7.0, 15.9, 1H), 3.51–3.58 (m, 1H), 3.90 (dddd,
J=2.0, 2.0, 7.0, 10.6 Hz, 1H), 3.97 (dt, J=2.3, 8.2 Hz,
1H), 4.35 (d, J=8.5 Hz, 2H), 4.41 (dd, J=7.2, 16.8 Hz,
1H), 5.68–5.74 (m, 1H), 5.94–6.00 (m, 1H). Compound
1
18a: H NMR (400 MHz, CDCl3): l 2.25–2.35 (m, 2H),
2.48 (br, 1H), 2.92 (ddd, J=3.4, 9.6, 16.8 Hz, 1H), 3.63
(ddd, J=5.6, 8.5, 14.2, 1H), 3.86 (ddd, J=3.9, 8.5, 16.8
Hz, 1H), 4.31–4.37 (m, 1H), 4.34 (dd, J=5.6, 9.1 Hz,
1H), 4.41 (dd, J=9.1, 8.5 Hz, 1H), 5.77 (ddd, J=2.04,
1
4.2, 13.8 Hz, 1H), 5.91–5.96 (m, 1H). Compound 18b: H
NMR (400 MHz, CDCl3): l 2.24 (br, 1H), 2.38 (m, 1H),
2.46–2.55 (m, 1H), 3.30 (ddd, J=2.9, 8.1, 13.1 Hz, 1H),
3.74 (ddd, J=3.1, 8.5, 13.1 Hz, 1H), 4.06 (m, 2H),
12. Harrowven, D. C.; Pattenden, G. In Comprehensive
Organic Synthesis; Trost, B. M., Ed.; Pergamon: Oxford,
UK, 1991; Vol. 3, p. 379.
.