A systematic strategy for preparation of uncommon sugars through enzymatic resolution and ring-closing metathesis

Article abstract of DOI:10.1016/j.tetlet.2004.12.014

A systematic synthetic strategy has been developed for producing uncommon sugars. This method involved kinetic resolution allylic alcohol followed by ring-closing-metathesis (RCM) to generate optical pure lactones as the common precursors. After further d

Full text of DOI:10.1016/j.tetlet.2004.12.014

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 811–813
A systematic strategy for preparation of uncommon sugars
through enzymatic resolution and ring-closing metathesis
Lizhi Zhu, James P. Kedenburg, Ming Xian and Peng George Wang*
Department of Biochemistry, 876 Biological Science Building, The Ohio State University, 484 West 12th Avenue,
Columbus, OH 43210, USA
Received 23 September 2004; revised 25 November 2004; accepted 2 December 2004
Available online 16 December 2004
Abstract—A systematic synthetic strategy has been developed for producing uncommon sugars. This method involved kinetic res-
olution allylic alcohol followed by ring-closing-metathesis (RCM) to generate optical pure lactones as the common precursors. After
further derivatization, four representative uncommon sugar units were successfully synthesized.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
apply to one type of uncommon sugars. In 2002, WangÕs
group developed a strategy for generating six-membered
b,c-unsaturated lactones using ring-closing-metathesis
(RCM).7 These lactones could be conveniently con-
verted to uncommon sugar derivatives. They are the
common starting point for preparation of uncommon
sugars, which means this method provides a Ôconvergent
approachÕ to sugar derivatives. The major advantage of
this synthetic strategy is that it provides a systematic
synthesis pathway for producing desired uncommon
sugar units. For instance, as shown in Scheme 1, four
different uncommon sugar units could be synthesized
starting from the same a,b-unsaturated lactone 1.
Unfortunately, the relatively high price of the starting
materials limited the usage of this method for preparing
large quantities of the desired uncommon sugars.
Deoxysugars and their oligosaccharides are frequently
found in the structure of bioactive drugs. Here, they
play very important roles in the general mechanism of
these drugsÕ action.1 There have been continuing efforts
on the synthesis of uncommon monosaccharides,2 start-
ing from either carbohydrates or noncarbohydrate pre-
cursors.3 Traditionally uncommon sugars were
synthesized through multistep transformations of rela-
tively inexpensive common sugars. For example, L-van-
cosamine can be synthesized from methyl glucoside in 15
steps.3b One major disadvantage of this approach is the
long reaction sequences for protection and deprotection
manipulations. An alternative approach is using noncar-
bohydrate precursors. In this case, many chemists have
come up with different synthetic approaches. For exam-
ple, Nicolaou synthesized L-vancosamine starting from
L-lactate in 11 steps,4 while McDonald synthesized the
same product by tungsten-catalyzed alkynol cycloiso-
merization in nine steps.5 Recently, MacMillanÕs group
developed a synthetic pathway for stereoselectively con-
structing sugar derivatives starting from b,c-oxyalalde-
hydes.6 However, all of these approaches to
uncommon sugars from either carbohydrates or noncar-
bohydrate starting materials are Ôtarget orientedÕ, which
means that the designed synthetic approach can only
In order to address this drawback of the RCM strategy,
we developed a new synthetic pathway by using a cheaper
starting material, trans-4-phenyl-3-buten-2-one. This
O
O
1
HO
O
OH
O
OH
O
OH
O
OH
HO
HO
HO
HO
Keywords: Uncommon sugars; Enzymatic resolution; Ring-closing
metathesis.
OH
3
OH
4
N3
5
2
*
Corresponding author. Tel.: +1 614 292 9884; fax: +1 614 688
3106; e-mail: wang.892@osu.edu
Scheme 1.
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.12.014

812
L. Zhu et al. / Tetrahedron Letters 46 (2005) 811–813
candidate ketone is readily available from commercial
source at a very low price. In order to generate the opti-
cally pure form of the allylic alcohol, we utilized a novel
enzymatic resolution developed by ParkÕs group in
2000.8 Reduction of this ketone, followed by an en-
zyme–metal catalytic resolution led to the desired allylic
alcohol acetate in good yields with excellent optical puri-
ties (>97% ee).
PPh3 in CH2Cl2 at 0 °C. This led to the inversion of
the chiral center to give (S)-ester 8 as a colorless oil in
89% yield (Scheme 3). For the other enantiomer (R)-8,
it was prepared by esterification of vinylacetic acid and
(R)-6 using DCC and DMAP. The attempted esterifica-
tion in acidic conditions failed, mainly due to the insta-
bility of the allylic alcohol under acidic conditions.
When 2% H2SO4 or 5% camphor-10-sulfonic acid
(CSA) was used as the catalyst, only the elimination
product was observed. Compound (S)-8 was subjected
to ring-closing-metathesis under standard conditions
using GrubbÕs second generation catalyst. However,
the phenyl group at the terminal vinyl position did give
negative effects to the RCM reaction. The reaction was
sluggish and even with the presence of 5% catalyst the
desired product was isolated in 68% yield after 3 days
in refluxing CH2Cl2. The b,c-unsaturated lactone 9 can
then be transformed into the corresponding osmunda-
lactones (1 and 10) by treatment with DMDO in
acetone at pH = 8.5. Under these conditions, the
epoxidation and the ring open of the epoxides happened
subsequently, to give two osmundalactones, 1 and 10, in
1.2:1 ratio. Compounds 1 and 10 both are desired prod-
ucts for us because they can be separated and used for
synthesis of uncommon sugars individually. In order
to clarify the synthetic pathway, only lactone 1 was used
to explore the further derivatization steps. Four different
types of deoxysugars were successfully prepared from
compound 1 (Scheme 3).
2. Results and discussion
The commercially available trans-4-phenyl-3-buten-2-
one was readily reduced by NaBH4 in MeOH at 0 °C
to give the racemic mixture of the allylic alcohol in
>95% yield (Scheme 2). The enzyme–metal catalyzed
reaction for dynamic kinetic resolution of the allylic alco-
hol was performed according to the published method.8
Previous studies by ParkÕs group indicated that the di-
chloro(p-cymene)-ruthenium(II) complex is an excellent
racemization catalyst for dynamic kinetic resolution of
allylic alcohol.9 With the presence of isopropenyl acetate
as an acyl donor, trans-4-phenyl-3-buten-2-ol (6) was
converted to the corresponding acetate (R)-7 by using
commercially
available
dichloro(p-cymene)-ruthe-
nium(II) dimmer and lipase (from Pseudomonas cepacia,
immobilized on ceramic particles) as catalysts. It was
found that triethylamine is crucial for this reaction. No
reaction happened without adding Et3N even after the
reaction mixture was stirred at rt for 24 h. Beside the de-
sired acetate, a saturated ketone was obtained as a side
product (10%). Fortunately, the side product was easily
removed by column chromatography. After reduction
with DIBAL-H (2.0 equiv), alcohol (R)-6 was obtained
quantitatively in excellent optical purities (>97% ee).
After hydrogenation in the presence of Pd/C under H2
atmosphere (25 psi), lactone 1 was transformed to c-hy-
droxy lactone 11. Subsequently, reduction with DIBAL-
H, L-amicetose (2) was obtained in excellent yields. On
the other hand, when lactone 1 was subjected to the
epoxidation conditions (NaClO/pyridine), epoxide 12
was isolated in moderate yield.12 Fortunately, these con-
ditions enabled us to produce the epoxide compound 12
The esterification was performed under Mitsunobu reac-
tion conditions by adding diethylazodicarboxylate
(DEAD) to a mixture of (R)-6, vinylacetic acid, and
O
OH
d
a
b
c
O
OH
OAc
89%
95%
78%
99%
6
(R)-7
(R)-6
O
(S)-8
O
O
O
O
O
O
e
f
+
68%
70%
HO
HO
9
1
10
Scheme 2. Reagents and conditions: (a) NaBH4/MeOH, 0 °C; (b) lipase, Ru(II) complex, isopropenyl acetate, Et3N/CH2Cl2; (c) DIBAL-H
(2.0 equiv)/CH2Cl2; (d) vinylacetic acid, Ph3P, DEAD; (e) GrubbÕs catalyst/CH2Cl2; (f) DMDO/acetone, PBSbuffer, pH = 8.5.
O
O
O
O
O
O
a
c
b
f
d, h
O
OH
HO
OH
O
HO
HO
95%
45%
85%
95%
95%
TBSO
14 trans only
HO
TBSO
O
O
O
11
2 L-amicetose
13
4 L-canarose
O
O
HO
O
O
O
O
f, c, d
c
N3
O
i
j
d, e
OH
1
OH
O
OH
HO
34%
80%
88% HO
56% HO
12 cis:trans > 20:1
TBSO
TBSO
O
3 L-digitoxose
N3
16
OH
15
5
Scheme 3. Reagents and conditions: (a) H2, Pd/C; (b) DIBAL-H/CH2Cl2; (c) NaClO, pyridine; (d) NaBH4, PhSeSePh, AcOH, iPrOH; (e)
NaBH3CN; (f) TBSCl, imidazole; (h) TBAF then NaBH3CN; (i) DPPA, PPh3, DEAD; (j) TBAF then DIBAL-H.

L. Zhu et al. / Tetrahedron Letters 46 (2005) 811–813
813
lipids; Elsevier: Amsterdam, 1985; (c) Varki, A. Glycobiology
1993, 3, 97–130; (d) Nagarajan, R. Glycopeptide Antibiot-
ics; Dekker: New York, 1994; (e) Allen, H. J. Glycocon-
jugates: Composition, structure, and function; Dekker: New
York, 1992; (f) Weymouth-Wilson, A. C. Nat. Prod. Rep.
1997, 14, 99–110.
stereoselectively by giving the ratio of cis trans greater
than 20:1. Compound 12 was then subjected to the selec-
tively ring opening reaction with sodium phenyl-
seleno(triisopropyloxy)borate (NaBH4 and PhSeSePh
in acetic acid). Finally, after reduction with NaBH3CN,
L-digitoxose (3) was obtained in 56% yield (over all yield
of two reduction steps).10
2. (a) Kirschning, A.; Jesberger, M.; Schoning, K.-U. Syn-
thesis 2001, 507–540; (b) Marzabadi, C. H.; Franck, R. W.
Tetrahedron 2000, 56, 8385–8417; (c) Toshima, K.; Tats-
uta, K. Chem. Rev. 1993, 93, 1503–1531.
In order to produce sugars with 3,4-trans-difunctional-
ities, a different strategy has to be taken. As shown in
Scheme 3, lactone 1 was first protected as tert-butyldim-
ethylsilyl ether. This bulky functional group helped to
force the subsequent epoxidation to happen at the
anti-face of the existing hydroxyl group. As a result,
when compound 13 was treated with NaClO/pyridine,
only trans-epoxide 14 was obtained.12 Reductively open-
ing the epoxide ring in 14 with sodium phenylseleno(tri-
isopropyloxy)borate allowed the formation of 3,4-trans-
dihydroxyl lactone. After it was treated with TBAF and
NaBH3CN L-canarose (4) was obtained in very good
yields. Through this strategy, 2,6-dideoxy sugars with
3,4-trans-difunctionalities were achieved, which largely
expended the synthetic potential of the RCM methodol-
ogy. (The previous strategy, RCM followed by asym-
metric dihydroxylation, can only produce uncommon
sugars with 3,4-cis-difunctionalities.7)
3. (a) Sztaricskai, F.; Pelyvas-Ferenczik, I. Glycopeptide
Antibiotics; Dekker: New York, 1994; (b) Pelyvas-Ferenc-
zik, I.; Monneret, C.; Herczegh, P. Synthetic Aspects of
Aminodeoxy Sugars of Antibiotics; Springer: Berlin, 1988;
(c) Hauser, F. M.; Ellenberger, S. R. Chem. Rev. 1986, 86,
35–67; (d) Hudlicky, T.; Entwistle, D. A.; Pitzer, K. K.;
Thorpe, A. J. Chem. Rev. 1996, 96, 1195–1220.
4. Nicolaou, K. C.; Mitchell, H. J.; Jain, N. F.; Winssinger,
N.; Hughes, R.; Bando, T. Angew. Chem., Int. Ed. 1999,
38, 240–244.
5. Cutchins, W. W.; McDonald, F. E. Org. Lett. 2002, 4,
749–752.
6. (a) Northrup, A. B.; Mangion, I. K.; Hettche, F.;
MacMillan, D. W. C. Angew. Chem., Int. Ed. 2004, 43,
2152–2154; (b) Northrup, A. B.; MacMillan, D. W. C.
Science 2004, 305, 1752–1755.
7. Andreana, P. R.; McLellan; Jason, S.; Chen, Y.; Wang, P.
G. Org. Lett. 2002, 4, 3875–3878.
8. Lee, D.; Huh, E. A.; Kim, M.-J.; Jung, H. M.; Koh, J. H.;
Park, J. Org. Lett. 2000, 2, 2377–2379.
Meanwhile, the stereoselective approach to 3-azido-
2,3,6-trideoxy sugar was achieved by taking advantage
of Mitsunobu reaction,11 which was depicted in Scheme
3. Following the same procedure for preparation of the
L-canarose (4), the precursor 15 was obtained. It was
then treated with diphenylphosphoryl azide (DPPA)
with the presence of DEAD and triphenylphosphine to
give the azido-compound 16 in 80% yield. As a result,
the azido-group was introduced inversely to afford the
deoxysugar with 3,4-cis-difunctionalities (3,4-cis:3,4-
trans >97:3 by NMR). Finally, compound 16 was suc-
cessfully converted into the 3-azido-2,3,6-trideoxy sugar
5 after the treatment with TBAF and DIBAL-H
subsequently.12
9. (a) Koh, J. H.; Jung, H. M.; Kim, M.-J.; Park, J.
Tetrahedron Lett. 1999, 40, 6281–6284; (b) Jung, H. M.;
Koh, J. H.; Kim, M.-J.; Park, J. Org. Lett. 2000, 2, 409–
411.
10. (a) Takano, S.; Shimazaki, Y.; Sekiguchi, Y.; Ogasawara,
K. Synthesis 1989, 539–541; (b) Miyashita, M.; Suzuki, T.;
Yoshikoshi, A. Tetrahedron Lett. 1987, 28, 4293–4296.
11. Dermatakis, A.; Luk, K.-C.; DePinto, W. Bioorg. Med.
Chem. 2003, 11, 1873–1881.
12. Spectroscopic data: Compound 5 was obtained as a
mixture of a/b-anomers, selected data for b-anomer 1H
NMR (400 MHz, CD3OD) d 5.03 (dd, J = 9.0, 2.3 Hz,
1H), 4.10 (q, J = 3.5 Hz, 1H), 3.64 (dq, J = 9.2, 6.3 Hz,
1H), 3.41 (m, 1H), 2.14 (q, J = 14.0 Hz, 1H), 1.83 (m, 1H),
1.32 (d, J = 6.7 Hz, 3H); 13C NMR (100 MHz, CD3OD) d
25
90.8, 69.7, 67.0, 63.1, 32.1, 15.3; ½aŠ ¼ ꢀ38 (c = 1.0,
D
MeOH-d4); HRMS(EI) m/z calculated for C6H11 N3O3
173.0796, found 173.0801. Compound
In summary, by utilizing the chemo-enzymatic synthetic
pathway, we were able to generate the chiral allylic alco-
hols conveniently in large scale. In addition, starting
from one osmundalactone, generated through RCM
method, four different types of uncommon sugars were
successfully prepared. Additional synthetic applications
of the RCM methodology are currently undergoing in
our laboratory.
8
1H NMR
(500 MHz, CDCl3) d 7.37 (d, J = 7.5 Hz, 2H), 7.31 (t,
J = 7.0 Hz, 2H), 7.23 (t, J = 7.0 Hz, 1H), 6.59 (d,
J = 16.0 Hz, 1H), 6.17 (dd, J = 16.0, 7.0 Hz, 1H), 5.92
(m, 1H), 5.53 (quint, J = 7.0 Hz, 1H), 5.18 (m, 2H), 3.11
(dt, J = 7.8, 1.5 Hz, 2H), 1.40 (d,J = 7.0 Hz, 3H); 13C
NMR (125 MHz, CDCl3) d 170.8, 136.3, 131.7, 130.4,
128.7, 128.6, 127.9, 126.6, 118.5, 71.3, 39.5, 20.4; HRMS
(EI) m/z calculated for C14H16O2 216.1145, found
216.1146. Compound 12 1H NMR (400 MHz, CD3OD)
d 4.37–4.32 (m, 1H), 3.86 (d, J = 9.0 Hz, 1H), 3.67 (d,
J = 4.2 Hz, 1H), 3.62 (d, J = 4.2 Hz, 1H), 1.34 (d,
J = 6.3 Hz, 3H); 13C NMR (100 MHz, CD3OD) d 167.6,
73.4, 71.1, 55.9, 50.5, 17.1; HRMS(ESI) m/z calculated for
C6H8O4Na 167.0320 (M+Na+), found 167.0324. Com-
Acknowledgements
This work was supported by NSF (CH-0316806) and a
fund from Michigan Life Science Corridor Fund
(1632) to P.G.W. The author thanks the Ohio State Uni-
versity for providing research resources, analysis instru-
ments and founding support.
1
pound 14 H NMR (400 MHz, CD3OD) d 4.48–4.42 (m,
1H), 3.92 (d, J = 9.2 Hz, 1H), 3.63 (d, J = 4.4 Hz, 1H),
3.54 (d, J = 4.4 Hz, 1H), 1.33 (d, J = 6.4 Hz, 3H), 0.93 (s,
9H), 0.18 (s, 3H), 0.15 (s, 3H); 13C NMR (100 MHz,
CD3OD) d 166.7, 73.3, 71.0, 55.9, 50.7, 25.8, 18.3, ꢀ3.9,
References and notes
ꢀ4.5; HRMS(EIS)
m/z calculated for C12H22O4SiNa
1. (a) Montreuil, J.; Vleigenthart, J. F. G. Glycoproteins;
Elsevier: Amsterdam, 1995; (b) Wiegandt, H. E. Glyco-
281.1185 (M+Na+), found 281.1190.