De novo synthesis of deoxy sugar via a Wharton rearrangement

Article abstract of DOI:10.1039/c1cc13837e

A highly divergent synthesis of α-fuco-, α-6-deoxy-allo-, α-6-deoxy-altro-pyranosides has been achieved. This route utilizes a Wharton rearrangement as part of a new post-glycosylation transformation strategy.

Full text of DOI:10.1039/c1cc13837e

View Article Online / Journal Homepage / Table of Contents for this issue
This article is part of the
Glycochemistry & glycobiology
web themed issue
Guest Editors are Professor Peter H. Seeberger and Dr Daniel B. Werz
Please visit the website to access the other articles in this issue:-
http://www.rsc.org/chemcomm/glycochemistry

Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 10251–10253
www.rsc.org/chemcomm
COMMUNICATION
De novo synthesis of deoxy sugar via a Wharton rearrangementwz
Hua-Yu Leo Wang and George A. O’Doherty*
Received 27th June 2011, Accepted 18th July 2011
DOI: 10.1039/c1cc13837e
A highly divergent synthesis of a-fuco-, a-6-deoxy-allo-, a-6-
deoxy-altro-pyranosides has been achieved. This route utilizes
a Wharton rearrangement as part of a new post-glycosylation
transformation strategy.
The importance of carbohydrate containing natural products in
medicine is indubitable.1 A survey of the structural motif used in
bioactive natural products leads one to notice the reliance on
rare sugars with uncommon stereochemistry and patterns of
deoxygenation.2–6 In an effort to mimic nature’s use of rare
Scheme 1 De novo approach toward carbohydrates.
deoxy sugars with unusual stereochemistry, medicinal chemists
of seven deoxy sugars, vide infra.14 Retrosynthetically, the
have long desired new synthetic strategies to provide practical
access to a wide range of rare sugar structural motifs. This need
has inspired various synthetic approaches.7
desired pyran 7 could be derived from the epoxy-ketone 8 via
a Wharton rearrangement (Scheme 2), which in turn can be
prepared via nucleophilic epoxidation of a-aculo-pyranoside 9.
Similarly, we have developed two de novo asymmetric
approaches to carbohydrates, where all the carbon atoms of
sugars are derived from achiral furans8 or dienoates.9 Of these
two approaches, the furan approach has proven to be the
most powerful. The route links a highly enantioselective
Noyori reduction of acylfurans, Achmatowicz oxidative
dearomatization, diastereoselective carbonate formation, and
Pd(0)-catalyzed glycosylation to stereoselectively transform
acylfurans to an aculose sugar with the desired stereochemistry
(e.g., 10 to a-/b- and D-/L-11, see Scheme 3).10 Subsequent
post-glycosylation transformations provide access to the other
pyran sugars.11 Of particular utility, is the osmium-catalyzed
dihydroxylation of the C-2/C-3 pyran double bond R/S-1,
which (depending on the C-4 stereochemistry) can be used to
install manno-, gulo-, and talo-stereochemistry (Scheme 1).11
Similarly, regio- and stereoselective anti-addition across the
same position can be used to access other sugars, such as
a-6-deoxy-altrose, a-ascarylose and a-digitoxose (Scheme 1).12
We envisioned that the use of the osmium-catalyzed
dihydroxylation on the regioisomeric allylic alcohol R/S-2 could
yield pyranoses with altro-, galacto- and allo-stereochemistry,
depending on the C-2 stereochemistry. Herein, we report the
successful application of the Wharton rearrangement13 as a new
post-glycosylation transformation, which leads to the synthesis
Scheme 2 Retrosynthetic analysis.
As previously reported, Boc-pyranone 11 can be readily
prepared in three steps from acylfuran 10 (Scheme 3).10 Using
our Pd(0)-glycosylation, Boc-pyranone 11 was converted into
a-benzyl-aculoside 9 with complete retention of anomeric
stereochemistry. Epoxidation of 9 with H2O2 and a catalytic
amount of base (10% NaOH) gave epoxide ketone 8 with
high diastereoselectivity (420 to 1).5 Next, we explored the
potential of Wharton transposition to rearrange 8 into allylic
alcohol 7, which under our optimal condition (Table 1, entry 6)
cleanly gave the desired product in 67% yield.
102 Hurtig Hall, 306 Huntington Avenue, Boston, Massachusetts,
USA. E-mail: g.odoherty@neu.edu; Fax: +1-617-373-8795;
Tel: +1-617-373-4817
w This article is part of the ChemComm ‘Glycochemistry and glyco-
biology’ web themed issue.
Scheme 3 De novo synthesis of pyran alcohol 7.
z Electronic supplementary information (ESI) available: Experimental
procedures, characterization data, and NMR spectra are available. See
DOI: 10.1039/c1cc13837e
Our initial attempts to perform the Wharton rearrangement
began with the typical condition (3 : 2 of N2H4ꢀH2O/AcOH,
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 10251–10253 10251

Table 1 Reaction conditions for Wharton rearrangement
Reagents
Entry (equiv)
Acid/Base
(equiv)
Yield
Solventa t (h) Temp. (%)b
1
2
3
4
5
6
7
8
N2H4ꢀH2O (3)
AcOH (2) MeOH
AcOH (2) CH2Cl2 18
MeOH 24
3
0 1C – rt
0 1C – rt
rt
0 1C – rt
rt
48
10
0
55
28
67
22
o10
N2H4ꢀH2O (3)
N2H4ꢀH2O (3)
N2H4ꢀH2O (4)
N2H4ꢀH2O (4)
N2H4ꢀH2O (5)
—
AcOH (1) MeOH
Et3N (1) MeOH
AcOH (3) MeOH
2
6
1.5 0 1C – rt
rt
0.3 68 1C
Scheme 5 Synthetic approaches toward a-ascarylose 4.
N2H4ꢀHSO4 (3)c Et3N (3)
CH3CN
—
3
N2H4ꢀH2O (5)
—
manno-stereochemistry was formed as a single diastereomer
(Scheme 5). In addition to the stereochemistry, we were
surprised by the regioselectivity of addition to the iodonium
intermediate, (i.e., trans-diequatorial addition). This peculiar
regioselectivity could be explained by the destabilizing effects
of the C-2 axial hydroxyl group on the partial positive charge
that is required for addition at the C-3 position.
a
b
c
Solvents are used in 0.3 M concentration. Isolated yield. N2H4ꢀ
HSO4 are prepared by N2H4ꢀH2O and conc. H2SO4 (ref. 15).
entry 1) in MeOH.13 The protic solvent, MeOH, appeared to be
important for the reaction. Switching to an aprotic solvent CH2Cl2
led to much slower reactions (entry 2). Slow reactions occurred also
in more polar solvents (i.e., acetonitrile) and in the presence of
ammonium salts (entry 7).15 Running the reaction in neat aqueous
hydrazine at elevated temperature led to a faster reaction, although
with a significant amount of decomposition (entry 8). We found
that acetic acid is critical for the reaction to occur (entry 3), whereas
the use of Et3N as a base gave low conversion to product (entry 5).
While increasing the ratio of hydrazine to AcOH (4 : 1) improved
the yield (55%, entry 4), increasing both the amount of hydrazine
and acetic acid gave the best results (67%, entry 6). We found that
it was best to initially add the hydrazine to a methanol solution of 8
to allow for hydrazone formation (B30 min) before adding the
AcOH at 0 1C. These conditions work well on the crude reaction
product 8 from the enone epoxidation. Thus enone 9 can be
converted to 7 on a gram scale using only one column chromato-
graphic purification to give 50% yield.
To test for reversibility in the iodonium formation, we decided to
perform the reaction on the t-butyl carbonate 14, which could trap
the stereoisomeric iodonium intermediate. When carbonate 14 was
exposed to the identical conditions as above only trans-diequatorial
iodide 15 was observed. Both iodo-esters 13 and 15 were easily
reduced in one-pot with LiAlH4 to give a-ascarylose 4 in good
yields. Alternatively, in a milder condition, iodide 13 was removed
under radical conditions with tris(trimethylsilyl)silane (TTMSS)
and AIBN, followed by hydrolysis to obtain a-ascarylose 4 with a
similar yield over two steps (Scheme 5).
We next decided to explore the facial selectivity of the stereo-
isomeric allylic alcohol 17 (Scheme 6). Our initial attempts to invert
the stereochemistry of allylic alcohol 7 via a Mitsunobu reaction
failed leading to only products of SN20 allylic transposition
(7 to 16).17 To avoid this problem, we turned to an oxidation/
reduction strategy. The allylic alcohol 7 was oxidized with MnO2,
and reduced under Luche conditions to provide the allylic alcohol
17 in 76% overall yield as a single diastereomer.
With allylic alcohol 7 in hand, we explored its reduction and
oxidation chemistry. Diimide reduction (o-NO2PhSO2NHNH2/
Et3N) of 7 cleanly gave a rare a-3,4-dideoxy-rhamnose sugar 12
in 85% yield (Scheme 4). As an alternative, the alkene in 7 can
be dihydroxylated under Upjohn conditions to give 6-deoxy-altrose
3 exclusively in 92% yield. These results suggest that the C-2/C-5
flanking substituents dominate the reaction facial selectivity over
the competing 1,3-diaxial interactions between C-1-OBn and
C-3-oxygen. To our surprise, this facial preference superseded
the potential hydrogen bonding interaction between the axial
allylic alcohol and the osmate-diamine complex. Thus, when 7
was subjected to the hydroxy-directed dihydroxylation conditions
of Donohoe16 the same altro-product 3 was formed.
With allylic alcohol 17 in hand, we first investigated the
stereo- and regio-selectivity of the halo-acetoxy formation.
Under the same NIS/HOAc conditions, allylic alcohol 17 was
converted to b-iodoacetate 18 as a single regio- and stereo-
isomer in 74% yield (Scheme 7). The gulo-stereochemistry of
18 was the result of a net trans-diaxial addition. Thus, chan-
ging the C-2 allylic alcohol stereochemistry of 7 to 17 did not
In marked contrast to the osmylation, pyran 7 reacted with
the opposite facial selectivity with ‘‘I+’’ type reagents (NIS/
HOAc), suggesting that the iodonium formation is more
sensitive to the developing 1,3-diaxial interaction with the
presence of anomeric oxygen. For instance, when pyran 7
was exposed to NIS in acetic acid, the b-acetoxy iodide 13 with
Scheme 6 Inversion of allylic alcohol 7.
Scheme 7 Synthesis of a-fucose, a-6-deoxy-allose, 4,6-dideoxy-
allose, and 3,4-dideoxy sugar congener.
Scheme 4 Synthesis of altro- and 3,4-dideoxy-rhamno-se.
10252 Chem. Commun., 2011, 47, 10251–10253
c
This journal is The Royal Society of Chemistry 2011

Table 2 Optimization of OsO4 dihydroxylation
(entries 8–9). The best allose selectivity came from the use of
(DHQD)2DPP at ꢁ78 1C (entry 12). In order to increase
fucose selectivity, the C-2 allylic alcohol had to be protected
either as a pivaloate or TBS-ether (entries 13 or 14).
In conclusion, we have found that by incorporating the
Wharton reaction into post-glycosylation transformation,
new rare deoxy sugars can be easily obtained. These sugars
include 6-deoxy-altro-/fuco-/allo-, ascarylo-, 3,4,6-trideoxy
gluco-/manno- and 4,6-dideoxy-allo-pyranosides, which can
be prepared in either D- or L-enantiomeric series. The route
allows for the divergent synthesis of a range of sugars from the
advanced stage intermediate. Future work in exploring the
potential of employing this strategy in medicinal chemistry is
currently ongoing.
Entry R
Condition Ligand
Ratio (6 : 5)a Yield (%)b
1
2
3
4
5
6
–H
A
B
D
E
C
C
—
—
—
—
2 : 1
1 : 3
1 : 4
2 : 1
1 : 1.5
95
84
80
65
94
89
–H
–H
–H
–H
–H
(DHQ)2PHAL
(DHQD)2PHAL 1 : 1.2
Notes and references
7
–H
C
(DHQD)2DPP
97
1 V. Kren and L. Martinkova, Curr. Med. Chem., 2001, 8, 1303–1328.
2 (a) C. Thibodeaux, C. E. Melancuon III and H.-w. Liu, Nature,
2007, 446, 1008–1016; (b) C. Mendez, A. Luzhetsky, A. Bechthold
and J. A. Salas, Curr. Top. Med. Chem., 2008, 8, 710–724.
3 R. P. Gorshkova, E. N. Kalmykova, V. V. Isakov and Y. S.
Ovodov, Eur. J. Biochem., 1985, 150, 527–531.
4 (a) S. C. Timmons and C. N. Jakeman, Carbohydr. Res., 2007, 342,
2695–2704; (b) M. Shan, E. U. Sharif and G. A. O’Doherty,
Angew. Chem., Int. Ed., 2010, 49, 9492–9495.
5 (a) H. Guo and G. A. O’Doherty, Org. Lett., 2005, 7, 3921–3924;
(b) T. J. Baiga, H. Guo, Y. Xing, G. A. O’Doherty, A. Parrish,
A. Dillin, M. B. Austin, J. P. Noel and J. J. La Clair, ACS Chem.
Biol., 2008, 3, 294–304.
8
–H
D
DHQ
88
9
10
11
–H
–H
–H
D
D
D
DHQD
1 : 1
1 : 4
(DHQD)2PHAL 1 : 2
82
85
88
(DHQ)2PHAL
12
–H
D
(DHQD)2DPP
92
13
14
a
–Piv
–TBS A
A
—
—
90
95
Diastereomeric ratio are based on crude NMR analysis in the
b
comparison of anomeric H1. Combined yield after flash column
purification.
6 S. Kunimoto, T. Someno, Y. Yamazaki, J. Lu, H. Esumi and
H. Naganawa, J. Antibiot., 2003, 56, 1012–1017.
affect the inherent facial preference of 17 with ‘‘I+’’ type
reagents. In contrast, switching the C-2 alcohol from axial to
equatorial did change the regio-selectivity of iodonium ring
opening. A similar removal of both iodide and acetate moiety
was achieved successfully under either hydride A or radical-
reduction condition B to give a rare 4,6-dideoxy-allose sugar 19.
Allylic alcohol 17 was further reduced with diimide to give
3,4,6-trideoxy-glucose 20.
7 For other de novo syntheses of sugar, see: (a) S. Y. Ko, A. W. M.
Lee, S. Masamune, L. A. Reed and K. B. Sharpless, Science, 1983,
220, 949–951; (b) S. J. Danishefsky and M. P. DeNinno, Angew.
Chem., Int. Ed. Engl., 1987, 26, 15–23; (c) A. B. Northrup and
D. W. C. MacMillan, Science, 2004, 305, 1752–1755.
8 (a) J. M. Harris, M. D. Keranen and G. A. O’Doherty, J. Org.
Chem., 1999, 64, 2982–2983; (b) J. M. Harris, M. D. Keranen,
H. Nguyen, V. G. Young and G. A. O’Doherty, Carbohydr. Res.,
2000, 328, 17–36; (c) H. Guo and G. A. O’Doherty, Angew. Chem.,
Int. Ed., 2007, 46, 5206–5208; (d) M. Zhou and G. A. O’Doherty,
Org. Lett., 2008, 10, 2283–2286.
Although switching the C-2 hydroxyl stereochemistry had
no effect on the stereochemistry of iodonium formation, it
greatly influenced the facial selectivity of dihydroxylation. For
instance, when 17 was exposed to typical Upjohn conditions, it
gave a 2 to 1 mixture of fuco-sugar 6 and 6-deoxy-allo-sugar 5
(Table 2, entry 1). The facial selectivity of 17 may result from
many competing factors, such as solvent effect, reaction
temperature and the presence of chiral ligands. For example,
when t-BuOH was changed to aprotic CH2Cl2, the facial
preference switched from fucose to allose (entry 2). The
preference for allo-stereochemistry in CH2Cl2 was increased
by lowering the reaction temperature (entry 3). Unfortunately,
the hydroxy-directing Donohoe conditions gave the same
selectivity as the Upjohn condition (entry 4). The addition of
the Sharpless chiral ligands to the reaction in t-BuOH switched
the selectivity from fucose to allose (entries 5–7) with
(DHQD)2DPP being the best ligand. This effect could be
enhanced by switching to CH2Cl2 and cooling to ꢁ78 1C
(entries 8–12). In general, switching between the dimeric
pseudo-enantiomeric ligands has little effect, whereas the
monomeric ligands cause a significant difference at ꢁ78 1C
9 (a) Md. M. Ahmed, B. P. Berry, T. J. Hunter, D. J. Tomcik and
G. A. O’Doherty, Org. Lett., 2005, 7, 745–748; (b) Md. M. Ahmed
and G. A. O’Doherty, J. Org. Chem., 2005, 70, 10576–10578.
10 (a) R. S. Babu, S. R. Guppi and G. A. O’Doherty, Org. Lett., 2006,
8, 1605–1608; (b) M. Shan and G. A. O’Doherty, Org. Lett., 2006,
8, 5149–5152.
11 (a) M. H. Haukaas and G. A. O’Doherty, Org. Lett., 2001, 3,
3899–3902; (b) M. H. Haukaas and G. A. O’Doherty, Org. Lett.,
2002, 4, 1771–1774; (c) Md. M. Ahmed and G. A. O’Doherty,
Tetrahedron Lett., 2005, 46, 4151–4155.
12 M. Shan, Y. Xing and G. A. O’Doherty, J. Org. Chem., 2009, 74,
5961–5966.
13 P. S. Wharton and D. H. Bohlen, J. Org. Chem., 1961, 26,
3615–3616.
14 (a) J. Liu, J. J. Swidorski, S. D. Peters and R. P. Hsung, J. Org.
Chem., 2005, 70, 3898–3902; (b) R. Takagi, K. Tojo, M. Iwata and
K. Ohkata, Org. Biomol. Chem., 2005, 3, 2031–2036; (c) A. J.
Phillips, A. C. Hart and J. A. Henderson, Tetrahedron Lett., 2006,
47, 3743–3745.
15 C. Dupuy and J. L. Luche, Tetrahedron, 1989, 45, 3437–3444.
16 T. J. Donohoe, K. Blades, P. R. Moore, M. J. Waring, J. J. G.
Winter, M. Helliwell, N. J. Newcombe and G. Stemp, J. Org.
Chem., 2002, 67, 7946–7956.
17 Hydrolysis of 16a/b gave the previously prepared diastereomeric
benzyl pyranones (16a/b, R = H), see: M. Zhou and G. A.
O’Doherty, J. Org. Chem., 2007, 72, 2485–2493.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 10251–10253 10253