Stereoselective synthesis of a fragment of mycobacterial arabinan

Article abstract of DOI:10.1021/ol062198j

Strategies for the stereoselective synthesis of mycobacterial arabinan were explored. Arabinofuranosyl donors with various protective groups were screened in terms of suitability for β-(1,2-cis)-selective glycosylation. The protective group was found to affect the stereoselectivity of arabinofuranosylation. β-Selectivity was drastically enhanced by using donors protected with 3,5-TIDPS, possibly due to conformational constraints on the furanose ring. Synthesis of heptaarabinofuranoside was then performed to demonstrate the practicality of this methodology.

Full text of DOI:10.1021/ol062198j

ORGANIC
LETTERS
2006
Vol. 8, No. 24
5525-5528
Stereoselective Synthesis of a Fragment
of Mycobacterial Arabinan
Akihiro Ishiwata,†,‡ Hiroko Akao,† and Yukishige Ito*,†,‡
RIKEN (Institute of Physical and Chemical Research), 2-1 Hirosawa, Wako-shi,
Saitama 351-0198, Japan, and CREST (JST), Kawaguchi, Saitama 332-0012, Japan
yukito@riken.jp
Received September 5, 2006
ABSTRACT
Strategies for the stereoselective synthesis of mycobacterial arabinan were explored. Arabinofuranosyl donors with various protective groups
were screened in terms of suitability for -(1,2-cis)-selective glycosylation. The protective group was found to affect the stereoselectivity of
arabinofuranosylation. -Selectivity was drastically enhanced by using donors protected with 3,5-TIDPS, possibly due to conformational constraints
on the furanose ring. Synthesis of heptaarabinofuranoside was then performed to demonstrate the practicality of this methodology.
â
â
Glycans that consist of furanosides are widespread constitu-
ents of glycoconjugates and cell-surface polysaccharides in
bacteria,1 fungi,2 and parasites.3 They play key roles in the
infectivity and pathogenicity of these microbes. Among them,
mycobacterial cell wall arabinans are attracting particular
attention. Mycobacterium tuberculosis, the causative agent
of tuberculosis, remains rampant and is a growing threat
worldwide due to the emergence of strains that have
multidrug resistance (MDR).4 Arabinan biosynthesis by
mycobacterial arabinofuranosyl transferases (ArafTs)5 is a
novel therapeutic target,6 although the precise mechanism
has yet to be elucidated. On the other hand, adjuvants derived
from the cell wall skeleton of mycobacteria, such as Bacillus
Calmette-Gue´rin (BCG-CWS) from M. boVis, are known to
be activators of innate immunity.7,8 Although the mechanism
of BCG-CWS-initiated immunity has been proposed to
involve the activation of macrophages via a Toll-like receptor
(TLR)2,9 its structural basis is difficult to define due to the
extreme complexity of BCG-CWS. The structure of the CWS
is unique, being composed of mycolic acid (MA), D-arabinan,
D-galactan, a linker disaccharide (R-Rhap-(1,3)-R-GlcNAc),
and peptidoglycan (PG) (Figure 1).10 Mycobacterial arabinans
consist of an [R-Araf(1f5)R-Araf]n repeat, which is linked
to the inner complex and is capped by a branched motif.
The latter contains terminal â-Araf, linked to the 2-position
† RIKEN (Institute of Physical and Chemical Research).
‡ CREST (JST).
(6) For designed inhibitors against mycobacterial Araf-T, see: (a) Pathak,
A. K; Pathak, V.; Maddry, J. A.; Suling, W. J.; Gurcha, S. S.; Besra, G. S.;
Reynolds, R. C. Bioorg. Med. Chem. 2001, 9, 3145-3151. (b) Pathak, A.
K.; Pathak, V.; Kulshrestha, M.; Kinnaird, D.; Suling, W. J.; Gurcha, S.
S.; Besra, G. S.; Reynolds, R. C. Tetrahedron 2003, 59, 10239-10248.
(7) Alexandroff, A. B.; Jackson, A. M.; O’Donnell, M. A.; James, K.
Lancet 1999, 353, 1689-1694.
(1) (a) Brennan, P. J.; Nikaido, H. Annu. ReV. Biochem. 1995, 64, 29-
63. (b) Ratledge, C.; Jeremy, D., Eds. Mycobacteria, Molecular Biology
and Virulence; Blackwell Science: Oxford, 1999. (c) Cole, S. T.; Eisenach,
K. D.; McMurray, D. N.; Jacobs, W. R., Jr., Eds. Tuberculosis and the
Tubercle Bacillus; ASM Press: Washington, DC, 1999.
(2) Notermans, S.; Veeneman, G. H.; Van Zuylen, C. W. E. M.;
Hoogerhout, P.; Van Boom, J. H. Mol. Immunol. 1998, 25, 975-979.
(3) de Lederkremer, R. M.; Colli, W. Glycobiology 1995, 5, 547-552.
(4) (a) Anti-tuberculosis drug resistance in the world; World Health
Organization: Geneva, 2000. (b) WHO Report 2004, Global tuberculosis
control-surVeillance, planning, financing; World Health Organization:
Geneva, 2004.
(8) Mathe, G.; Halle-Pannenko, O.; Bourut, C. Natl. Cancer, Inst.
Monogr. 1973, 39, 107-113.
(9) (a) Tsuji, S.; Matsumoto, M.; Takeuchi, O.; Akira, S.; Azuma, I.;
Hayashi, A.; Toyoshima, K.; Seya, T. Infect. Immun. 2000, 68, 6883-
6890. (b) For a recent review, see: Krieg, A. M. Nat. Med. 2003, 9, 831-
835.
(5) Houseknecht, J. B.; Lowary, T. D. Curr. Opin. Chem. Biol. 2001, 5,
677-682.
(10) Besra, G. S.; Khoo, K.-H.; McNeil, M.; Dell, A.; Morris, R.;
Brennan, P. J. Biochemistry 1995, 34, 4257-4266.
10.1021/ol062198j CCC: $33.50
© 2006 American Chemical Society
Published on Web 11/04/2006

Figure 1. Structure of a mycobacterial arabinan terminus.
of the penultimate R-Araf. These Araf residues are O-acylated
at their C-5 positions by mycolic acid.
particularly sensitive to the nature of R (Scheme 2 and Table
1, entries 1∼5).14 Specifically, the p-methoxybenzyl (PMB)-
To achieve the synthesis of mycobacterial arabinan,
introduction of the terminal â-Araf residue is potentially
problematic. Whereas R-Araf linkage is easily accessible with
a 2-O-acylated donor, formation of the â-Araf linkage is not
straightforward. The difficulty of the latter derives from its
1,2-cis relative stereochemistry, which prohibits the use of
neighboring group participation for stereochemical control.
Because the conformation of the furanoside ring is more
flexible than that of pyranoside, factors that affect the
stereochemistry of O-furanosylation are difficult to general-
ize. For the construction of â-linked Araf glycoside, several
innovative methods have been reported.11 Of particular note
were approaches through SN2 displacement via R-trifrate (a
2,3-anhydro-type donor and a carboxybenzyl (CB) donor)
developed by Lowary11a and by Kim,11b respectively. As an
alternative to these, we explored the possibility of achieving
â-selective furanosylation by tuning the protective group
patterns. We report herein the stereoselective synthesis of
â-Araf using 3,5-O-tetraisopropyldisiloxanylidene (TIPDS)-
protected thioglycoside as a donor. This method was then
successfully applied to the synthesis of heptaarabinofurano-
side.12
Scheme 1. Preparation of Arabinofuranosyl Donors
Because terminal â-Araf residues of AG are O-5 myco-
lated, this position must be distinguishable from others to
approach mycolated arabinan. Therefore, we began by
screening the effects of O-5 protective groups on the
stereoselectivity of glycosylation using 2,3-O-benzyl-
protected donors 1 and 2 (Scheme 1), through activation by
NIS-AgOTf.13 We previously reported that selectivity was
protected donor 1c gave a favorable result, although the
selectivity was modest. On the other hand, o-nitrobenzyl-
(1d) or acyl-substituted (1b) donors gave the R-glycoside
as the major product, suggesting that the electron density of
O-5 is important. The tert-butyldiphenylsilyl (TBDPS) group
at O-5 strongly disfavored the formation of the â-glycoside,
possibly because its bulkiness hindered the nucleophilic
attack of 1212c from the upper face. Corresponding tolylthio
glycosides 2a and 2c, prepared from known compound 8,15
(11) (a) Callam, C. S.; Gadikota, R. R.; Krein, D. M.; Lowary, T. L. J.
Am. Chem. Soc. 2003, 125, 13112-13119. (b) Lee, Y. J.; Lee, K.; Jung, E.
H.; Jeon, H. B.; Kim, K. S. Org. Lett. 2005, 7, 3263-3266.
(12) For synthetic studies of oligoarabinofuranoside including â-linkage,
see: (a) Mereyala, H. B.; Hotha, S.; Gurjar, M. K. J. Chem. Soc., Chem.
Commun. 1998, 685-686. (b) D’Souza, F. W.; Lowary, T. L. Org. Lett.
2000, 2, 1493-1495. (c) D’Souza, F. W.; Ayers, J. D.; McCarren, P. R.;
Lowary, T. L. J. Am. Chem. Soc. 2000, 122, 1251-1260. (d) Sanchez, S.;
Bamhaoud, T.; Prandi, J. Tetrahedron Lett. 2000, 41, 7447-7452. (e) Yin,
H.; D’Souza, F. W.; Lowary, T. L. J. Org. Chem. 2002, 67, 892-903. (f)
Wang, H.; Ning, J. J. Org. Chem. 2003, 68, 2521-2524. (g) Marrotte, K.;
Sanchez, S.; Bamhaoud, T.; Prandi, J. Eur. J. Org. Chem. 2003, 3587-
3598.
(13) Garegg, P. J. AdV. Carbohydr. Chem. Biochem. 1997, 52, 179-
205.
(14) Ishiwata, A.; Akao, H.; Ito, Y.; Sunagawa, M.; Kusunose, N.;
Kashiwazaki, Y. Bioorg. Med. Chem. 2006, 14, 3049-3061.
(15) See Supporting Information.
5526
Org. Lett., Vol. 8, No. 24, 2006

prepared. When glycosylation with acceptor 13 was per-
formed, it was found that the use of 3,5-O-TIPDS-protected20
donor 4g or 4h resulted in the formation of the desired
â-isomer in a markedly selective manner (Table 2). That tri-
Table 1. Effect of Protection of the Glycosyl Donor with
Various Protections at the 5 Position on Arabinofuranosylation
entry donor acceptor solventb temp time/h product yield (R/â)
1a
2a
3a
4a
5a
6
1a
1b
1c
1d
1e
2a
2c
1c
1c
1c
2c
2c
12
12
12
12
12
12
12
12
12
12
13
13
Dc
Dc
Dc
Dc
Dc
Dc
Dc
Dd
Td
Xd
Dc
Dc
-78 °C
-40 °C
-78 °C
-40 °C
-78 °C
-78 °C
-78 °C
rt
1
14a 97% (9.4:1)
14b 65% (1.3:1)
14c 93% (1:2.6)
14d 88% (1.9:1)
14e 93% (1:1.5)
14a 89% (7.6:1)
14c 81% (1:2.7)
14c 82% (1.3:1)
14c 45% (2.0:1)
14c 75% (1:1.1)
15c 94% (1:4.3)
15c 85% (1:8.6)
3.5
2.5
1
Table 2. Effect of Protection of the Glycosyl Donor on
Arabinofuranosylation
1
1
7
1
entry
donor
acceptor
product
yield (R/â)
8
9
10
11
12
24
1
2
3
4
5
3
13
13
13
13
13
15f
15g
15h
15i
96% (1:2.45)
94% (1:12.5)
93% (1:20.0)
70% (1:5.36)
99% (5.26:1)
0 °C
8
1
3
3
0 °C
-40 °C
-60 °C
4g
4h
5
a From ref 13. b D: CH2Cl2. T: toluene. X: xylene. c NIS (1.2∼2.0
equiv), AgOTf (0.3∼1.0 equiv). d MeOTf (2.4 equiv), DTBMP (2.4 equiv).
6
15j
O-silyl protection itself was not the dominant factor for this
selectivity was evident from the results with regioisomerically
protected (3), 3,5-O-di-tert-butylsilylene21,22 masked (5), and
noncyclically protected (6) donors. Comparison of these
results suggested that the conformational restraint introduced
by the eight-membered ring 3,5-O-protection was responsible
for the enhanced â-selectivity.
gave similar results (Table 1, entries 6 and 7). A weaker
activating agent, MeOTf,16 gave low selectivity (entries 8
and 9); however, as we previously reported, substantial rate
enhancement17 was observed in frozen solvent (entry 10).
The stereochemistry of the anomeric center of glycosylated
3
products was confirmed by δ (C-1) and JH1-H2 values:18
3
â-isomer, δ (C-1) 97∼103 ppm, JH1-H2 ) 4∼5 Hz;
R-isomer, δ (C-1) 104∼111 ppm, 3JH1-H2 ) 1∼3 Hz. Further
validation for the stereochemsitry of the â-isomer was
obtained by differential NOE experiments, which revealed
the NOE between H-1 and H-2.
The marked differences in selectivity between 4h (R/â )
1:20) and 5 (R/â ) 1:5.36) are reminiscent of the work of
Woerpel et al.,23 who investigated the allylation of bicyclic
lactol acetates 16 and 17 (Figure 2). They found that eight-
five bicyclic acetate 16 gave higher selectivity in favor of
the â-isomer than the six-five counterpart 17. In light of
Woerpel’s hypothesis, we surmise that the nucleophilic attack
from the R-face is disfavored for both 5 and 4h due to the
1,2-gauche interaction between the entering acceptor and the
pseudoaxially oriented C-2 hydrogen. In the case of six-five
bicyclic 5, however, the â-attack should lead to an initial
conformer with a 3,5-silylene group possessing a distorted
nonchairlike conformation, thereby reducing the preference
of the pathway toward the â-isomer. Molecular modeling
studies15 of glycosylated products provided an alternative
interpretation of the selectivity; the total energy of the
â-linked product was 3.7 kcal/mol lower than that of the
R-isomer, suggesting that the formation of the â-isomer was
the thermodynamically favored process, which may rational-
Subsequent investigation revealed that the selectivity was
also sensitive to the structure of the acceptor. When 2c was
reacted with 3,5-O-benzyl-protected acceptor 13,15 significant
enhancement of â-selectivity (R/â ) 1:4.6) was observed.
Selectivity was improved to 1:8.6, when the reaction was
conducted at -60 °C. To achieve higher selectivity, we
turned our attention to the effects of cyclic protective groups
(Scheme 1). Our hope was that the conformational perturba-
tion associated with the formation of fused rings would cause
favorable stereoelectronic effects.19 To that end, donors
having 2,3-O- (3) or 3,5-O- (4g, 4h, 5) cyclic protection were
Scheme 2. Various Arabinofuranosyl Donors
(16) Lo¨nn, H. Carbohydr. Res. 1985, 139, 105-113.
(17) Takatani, M.; Nakano, J.; Arai, M. A.; Ishiwata, A.; Ohta, H; Ito,
Y. Tetrahedron Lett. 2004, 45, 3929-3932.
(18) Mizutani, K.; Kasai, R.; Nakamura, M.; Tanaka, O. Carbohydr. Res.
1989, 185, 27-38.
(19) For a concept in the pyranose series, see: (a) Fraser-Reid, B.; Wu,
Z.; Andrews, C. W.; Skowronski, E.; Bowen, J. P. J. Am. Chem. Soc. 1991,
113, 1434-1435. (b) Crich, D.; Sun, S. J. Am. Chem. Soc. 1997, 119,
11217-11223. (c) Ito, Y.; Ohnishi, Y.; Ogawa, T.; Nakahara, Y. Synlett
1998, 1102-1104. (d) Jensen, H. H.; Nordstrøm, L. U.; Bols, M. J. Am.
Chem. Soc. 2004, 126, 9205-9213.
(20) Markiewicz, M. T. J. Chem. Res. Synop. 1979, 24-25.
(21) Corey, E. J.; Hopkins, P. B. Tetrahedron Lett. 1982, 23, 4871-
4874.
(22) Very recently, Boons et al. reported a practical approach for â-L-
arabinofuranosylation using benzyl-protected L-11 as a donor. Zhu, X.;
Kawatkar, S.; Rao, Y.; Boons, G.-J. J. Am. Chem. Soc. 2006, 128, 11948-
11957.
(23) Smith, D. M.; Tran, M. B.; Woerpel, K. A. J. Am. Chem. Soc. 2003,
125, 14149-14152.
Org. Lett., Vol. 8, No. 24, 2006
5527

With an arabinosyl donor suitable for the stereoselective
formation of the â-Araf1f2Araf linkage at our disposal, we
conducted the synthesis of heptasaccharide 32 (Scheme 3),
Scheme 3. Synthesis of Heptaarabinofuranoside
Figure 2. Plausible explanation for the â-selective addition to the
activated donor. The black arrow shows the direction for â-attack
to the anomeric carbon, and the white arrow shows the direction
for R-attack. In the case of R-attack, there seems to be a large steric
repulsion from the R-hydrogen atom at C2.
ize the selectivity based on Hammond’s postulate.24 In the
global minimum structure, the â-product had pseudoaxially
oriented glycoside linkages, which may be favorable in light
of the anomeric effect (see Supporting Information).
Inspection of other acceptors shows that, to achieve the
â-selective glycosylation, the acceptor should be moderately
bulky. For instance, a reaction of 4g with less-hindered
acceptor 2025 displayed no selectivity (Table 3). We speculate
which corresponds to the branched terminal structure of
mycobacterial arabinogalactan (Figure 3). Thus, disaccharide
2312e was first glycosylated with 2-O-Bz-protected donor
2412c to give 25a. Removal of the TIPDS group gave 25b,
which was further glycosylated with 2-O-CAc-protected
26.12b The resultant pentasaccharide 27 was converted to diol
28, which was subjected to a reaction with the â-selective
donor 4g to give 29 in high yield and selectivity (29/other
isomers ) 10.8:1). Subsequent deprotection was conducted
in a stepwise manner to give 32.
Table 3. Results of Glycosylation of 4g/4h with Other
Acceptors
entry
donor
acceptor
yield/%
R/â
1
2
3
4
5
6
7
4g
4h
4g
4g
4h
4g
4h
13
13
20
21
21
22
22
94
93
97
100
61
77
1:12.5
1:20.0
1:1.15
1:7.35
1.80:1
1:2.66
1:2.80
47
In conclusion, the stereoselective synthesis of mycobac-
terium arabinan was achieved. This â-selective arabinofura-
nosylation was applied to the synthesis of nonreducing termi-
nal heptaarabinofuranoside in the mycobacterial cell wall.
Acknowledgment. Financial support from a Grant-in-
Aid for Encouragement of Young Scientists from the
Ministry of Education, Culture, Sports, Science, and Tech-
nology (No. 187110196) is acknowledged. We thank Daini-
hon-Sumitomo Pharm Co., Ltd. for their support and Ms.
A. Takahashi (RIKEN) for her technical assistance.
that, when the acceptor is not sufficiently hindered, steric
repulsion with the C-2 hydrogen would be inconsequential.
In addition, the results with hindered acceptor 21 were
drastically different between 4g and 4h, thus suggesting that
the steric factor plays a major role in controlling the
stereochemical outcome, and there seems to be a matched/
mismatched effect in the donor-acceptor combination.
Supporting Information Available: Experimental pro-
cedures and characterization data. This material is available
free of charge via the Internet at http://pubs.acs.org.
(24) Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334-338.
(25) 20: Montgomery, J. A.; Shortnacy, A. T.; Thomas, H. J. J. Med.
Chem. 1974, 17, 1197-1207. 21: Ishiwata, A.; Ito, Y. Tetrahedron Lett.
2005, 46, 3521-3524. 22: See Supporting Information.
OL062198J
5528
Org. Lett., Vol. 8, No. 24, 2006