Direct chemical synthesis of the β-mannans: Linear and block syntheses of the alternating β-(1→3)-β-(1→4)-mannan common to Rhodotorula glutinis, Rhodotorula mucilaginosa, and Leptospira biflexa

Article abstract of DOI:10.1021/ja0471931

Two stereocontrolled syntheses of a methyl glycoside of an alternating β-(1→4)-β-(1→3)-mannohexaose, representative of the mannan from Rhodotorula glutinis, Rhodotorula mucilaginosa, and Leptospira biflexa, are described. Both syntheses employ a combinati

Full text of DOI:10.1021/ja0471931

Published on Web 10/22/2004
Direct Chemical Synthesis of the â-Mannans: Linear and
Block Syntheses of the Alternating â-(1f3)-â-(1f4)-Mannan
Common to Rhodotorula glutinis, Rhodotorula mucilaginosa,
and Leptospira biflexa
David Crich,* Wenju Li, and Hongmei Li
Contribution from the Department of Chemistry, UniVersity of Illinois, 845 West Taylor Street,
Chicago, Illinois 60607-7061
Received May 13, 2004; E-mail: dcrich@uic.edu
Abstract: Two stereocontrolled syntheses of a methyl glycoside of an alternating â-(1f4)-â-(1f3)-
mannohexaose, representative of the mannan from Rhodotorula glutinis, Rhodotorula mucilaginosa, and
Leptospira biflexa, are described. Both syntheses employ a combination of 4,6-O-benzylidene- and 4,6-
O-p-methoxybenzylidene acetal-protected donors to achieve stereocontrolled formation of the â-mannoside
linkage. The first synthesis is a linear one and proceeds with a high degree of stereocontrol throughout
and an overall yield of 1.9%. The second synthesis, a block synthesis, makes use of the coupling of two
trisaccharides, resulting in a shorter sequence and an overall yield of 4.4%, despite the poor selectivity in
the key step.
Introduction
4,6-O-benzylidene-type protection and nonparticipating groups
on O-2 and O-3 in our mannosylation protocols.3,4 We first
Recently, we described the linear synthesis of two â-man-
nans,1 a â-(1f2)-mannooctaose and a â-(1f4)-mannohexaose,
employing the 4,6-O-benzylidene-directed2 â-mannosylations
developed in our laboratory from either thioglycoside3 or
glycosyl sulfoxide4 donors, with emphasis on the influence of
the linkage type, â-(1f2) or â-(1f4), on stereoselectivity in
the growing chain. We describe here the extension of these
investigations to the synthesis of an alternating â-(1f3)-â-
(1f4)-mannohexaose. This hexasaccharide may be viewed as
essentially two repeats, sandwiched between two terminal
glycosides, of the basic structural unit of the antigenic mannan,
with alternating â-(1f3) and â-(1f4) linkages from Rhodot-
orula glutinis,5 Rhodotorula mucilaginosa,6 and Leptospira
biflexa.7 The particular challenge of this target, aside from the
obvious presence of multiple â-mannosidic linkages,8 is the
alternating nature of the chain, necessitating the use of two
glycosyl donors with orthogonal protection at positions 3 and
4, which are nevertheless consistent with the requirement for a
describe in full a linear synthesis in which the challenge is met
by the judicious combination of benzyl and p-methoxybenzyl
ethers with benzylidene and p-methoxybenzylidene acetals,
respectively, in two distinct thioglycosides,9 and then in a quest
for greater efficiency,10 we present a convergent or block
synthesis11 which eliminates the need for one of these protecting
groups, namely, the p-methoxybenzyl ether.
Results and Discussion
The linear synthesis of the hexasaccharide began with the
preparation of two thioglycosides and their sulfoxides. Thus,
phenyl R-D-thiomannopyranoside 1 was converted to ben-
zylidene acetal 212 and the corresponding p-methoxybenzylidene
acetal 313 in the standard manner. Reaction of 2 with dibutyltin
oxide14 and then p-methoxybenzyl chloride and tetrabutylam-
monium bromide selectively afforded 3-O-PMB ether 415 in 94%
yield. Reaction with sodium hydride and benzyl bromide then
(9) For a preliminary communication covering the linear synthesis, see: Crich,
D.; Li, H.; Yao, Q.; Wink, D. J.; Sommer, R. D.; Rheingold, A. L. J. Am.
Chem. Soc. 2001, 123, 5826-5828.
(10) (a) Douglas, N. L.; Ley, S. V.; Lucking, U.; Warriner, S. L. J. Chem. Soc.,
Perkin Trans. 1 1998, 51-65. (b) Zhang, Z.; Ollmann, I. R.; Ye, X.-S.;
Wischnat, R.; Baasov, T.; Wong, C.-H. J. Am. Chem. Soc. 1999, 121,
734-753.
(1) Crich, D.; Banerjee, A.; Yao, Q. J. Am. Chem. Soc. 2004, in press.
(2) Crich, D.; Sun, S. J. Am. Chem. Soc. 1997, 119, 11217-11223.
(3) (a) Crich, D.; Smith, M. J. Am. Chem. Soc. 2001, 123, 9015-9020. (b)
Crich, D.; Smith, M. J. Am. Chem. Soc. 2002, 124, 8867-8869.
(4) (a) Crich, D.; Sun, S. J. Org. Chem. 1997, 62, 1198-1199. (b) Crich, D.;
Sun, S. Tetrahedron 1998, 54, 8321-8348.
(5) (a) Gorin, P. A. J.; Horitsu, K.; Spencer, J. F. T. Can. J. Chem. 1965, 43,
950-954. (b) Gorin, P. A. J. Carbohydr. Res. 1975, 39, 3-10.
(6) Takita, J.; Katohda, S.; Sugiyama, H. Carbohydr. Res. 2001, 335, 133-139.
(7) Matsuo, K.; Isogai, E.; Araki, Y. Carbohydr. Res. 2000, 328, 517-524.
(8) (a) Pozsgay, V. In Carbohydrates in Chemistry and Biology; Ernst, B.,
Hart, G. W., Sinay¨, P., Eds.; Wiley-VCH: Weinheim, Germany, 2000;
Vol. 1, pp 319-343. (b) Demchenko, A. V. Synlett 2003, 1225-1240. (c)
Demchenko, A. V. Curr. Org. Chem. 2003, 7, 35-79. (d) Barresi, F.;
Hindsgaul, O. In Modern Methods in Carbohydrate Synthesis; Khan, S.
H., O’Neill, R. A., Eds.; Harwood Academic Publishers: Amsterdam, 1996;
pp 251-276. (e) Gridley, J. J.; Osborn, H. M. I. J. Chem. Soc., Perkin
Trans. 1 2000, 1471-1491.
(11) (a) Boons, G.-J. Tetrahedron 1996, 52, 1095-1121. (b) Kanie, O. In
Carbohydrates in Chemistry and Biology; Ernst, B., Hart, G. W., Sinay¨,
P., Eds.; Wiley-VCH: Weinheim, Germany, 2000; Vol. 1, pp 407-426.
(12) Oshitari, T.; Shibasaki, M.; Yoshizawa, T.; Tomita, M.; Takao, K.;
Kobayashi, S. Tetrahedron 1997, 53, 10993-11006.
(13) Johansson, R.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1 1984, 2371-
2374.
(14) (a) David, S.; Hanessian, S. Tetrahedron 1985, 41, 643-663. (b) David,
S. In PreparatiVe Carbohydrate Chemistry; Hanessian, S., Ed.; Dekker:
New York, 1997; pp 69-83.
(15) Nicolaou, K. C.; Mitchell, H. J.; Suzuki, H.; Rodriguez, R. M.; Baudoin,
O.; Fylaktakidou, K. C. Angew. Chem., Int. Ed. 1999, 38, 3334-3339.
9
10.1021/ja0471931 CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 15081-15086
15081

A R T I C L E S
Crich et al.
Table 1. Coupling and Selective Deprotection Yields in the Linear Synthesis of 26
deprotection
(% yield)
benzylation
(% yield)
acceptor
donor
linkage
â
-glycoside(% yield)
R-glycoside(% yield)
â/R ratio
MeOH
11
14
18
21
8
6
8
6
8
9 (62)
12 (81)
15 (72)
19 (62)
22 (72)
26 (63)
10 (97)a
11 (87)
18 (87)
25 (62)
1f4
1f3
1f4
1f3
1f4
13 (8)
16 (8)
20 (7.5)
23 (8)
27 (7)
10/1
9/1
7.5/1
9/1
14 (83)b
17 (70)a
21(72)b
24 (50),a (81)c
25
30
9/1
a Removal of a PMP acetal with acetic acid. b Removal of a PMB ether with DDQ. c Yield based on recovered starting material.
provided the fully protected thioglycoside 5 in 90% yield.
Benzylation of PMP acetal 3 with sodium hydride and benzyl
bromide gave dibenzyl ether 6 in 99% yield. Thioglycosides 5
and 6 were then both transformed to the corresponding sul-
foxides 7 and 8, respectively, in the form of single diastereomers.16
the next coupling reaction, which was achieved with a 9/1 ratio
of â/R selectivity by means of donor 8. Removal of the PMP
acetal from the resulting â-trisaccharide 15, affording diol 17,
required careful monitoring of the reaction progress to avoid
concomitant cleavage of the benzylidene acetal. As reported
Activation of 8 in dichloromethane at -78 °C with triflic
anhydride in the presence of the hindered base TTBP, the
standard protocol for sulfoxide â-mannosylation,4b,17 followed
by addition of methanol, gave methyl â-mannoside 9 in 62%
yield (Table 1). Glycoside 9 was subsequently obtained directly
from thioglycoside 6 by activation with 1-benzenesulfinyl
piperidine (BSP) and triflic anhydride in the presence of TTBP
in 74% yield. Although the PMP acetal in 9 could potentially
be reduced directly to the 6-O-PMB ether,13 leaving the 4-OH
free to participate in a coupling with donor 7, this would have
resulted in the need to selectively remove one of two PMB ethers
in order for the synthesis to proceed. The preferred protocol
was, therefore, the selective removal of the PMB acetal under
acidic conditions followed by regioselective 6-O-benzylation
with dibutyltin oxide and benzyl bromide. Thus, treatment of 9
with glacial acetic acid in dichloromethane gave diol 10 in 97%
yield. Reaction with dibutyltin oxide in toluene at reflux,
followed by treatment with benzyl bromide and cesium fluoride,
provided tri-O-benzyl ether 11 in 87% yield (Table 1). Standard
coupling of acceptor 11 with donor 7 then provided the â-1,4-
linked disaccharide 12 in high yield and selectivity. Treatment
of 12 with DDQ cleaved the PMB ether and set the stage for
(16) (a) Crich, D.; Mataka, J.; Sun, S.; Lam, K.-C.; Rheingold, A. R.; Wink, D.
J. J. Chem. Soc., Chem. Commun. 1998, 2763-2764. (b) Crich, D.; Mataka,
J.; Zakharov, L. N.; Rheingold, A. L.; Wink, D. J. J. Am. Chem. Soc. 2002,
124, 6028-6036.
9
15082 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

Direct Chemical Synthesis of the
â-Mannans
A R T I C L E S
Scheme 1. Key Disconnections in the Block Synthesis
by Khorana,18 this was achieved with acetic acid, but better
yields overall were obtained if the reaction was stopped before
completion and the unreacted substrate was recovered and
resubmitted to the deprotection conditions. Terminal diol 17
was then selectively benzylated on the primary alcohol by the
stannylene acetal technology, giving 18 (Table 1). At this stage,
each of the five individual steps needed in the iterative sequence
toward the hexasaccharide had been successfully demonstrated.
Careful repetition of the reaction sequence, with the yields and
selectivities indicated in Table 1, eventually led to the fully
protected all-â-hexasaccharide 26. In the course of this synthesis,
the â-stereochemistry of 9 and all subsequent â-mannosides
were assigned on the basis of the characteristic upfield chemical
shift of the H-5 resonance (δ 3.30), an apparent doublet of
triplets, which is typical and diagnostic for 4,6-O-benzylidene-
protected â-mannosides.4b
In an attempt to increase efficiency, a convergent approach to
26 was next designed. As set out in Scheme 1, the target was dis-
sected into two trisaccharides, 18 and 28, each of which could
potentially be assembled from a common disaccharide thioglyco-
side 29, thereby reducing the overall number of steps significantly.
As envisaged, the block synthesis required two further
building blocks, the known 2,3-di-O-benzyl-4,6-O-benzylidene
30,4b readily obtained by perbenzylation of diol 2, and the 2-O-
benzyl derivative 31. In principle, 31 can be obtained by
selective removal of the PMB ether from 5, but we have
preferred to employ a simple biphasic treatment with aqueous
tetrabutylammonium hydrogen sulfate and benzyl bromide in
dichloromethane,19 which lead directly to the known20 monoben-
zyl ether 31 from diol 2 in 75% yield.21 Also in the interests of
increasing efficiency, this block synthesis has been conducted
by the thioglycoside/BSP method3a rather than by the sulfoxide
method employed in the earlier linear synthesis. Pivotal disac-
charide 29 was obtained (Scheme 2) by coupling glycosyl donor
6 to thioglycoside alcohol 31 by the BSP method in 62% yield
on a multigram scale; no R-isomer was isolated from this
reaction. The success of this enterprise, in which the acceptor
itself is a thioglycoside, does not derive from any differential
reactivities of 6 and 31, as in Fraser-Reid’s armed/disarmed
concept22 and as employed successfully in the recent one-pot
thioglycoside-based oligosaccharide syntheses of the Ley10a and
Wong10b groups. Rather, it is a function of the prior activation
of the donor before addition of the acceptor, a standard feature
of our â-mannosylation protocols, and from the recognition that
1 mole of the BSP/Tf2O combination brings about the activation
of at least 2 moles of thioglycoside. This latter observation
enables the initial activation to be conducted with <50 mol %
BSP/Tf2O, thereby minimizing the amount of thiophilic species
in solution when the acceptor is added.23 The yield of this
coupling was further enhanced by the addition of triethyl
phosphite after the acceptor was added and before the reaction
mixture was allowed to come to room temperature, with the
objective, again, being to minimize the possibility of the
activation of disaccharide thioglycoside 29 by extraneous
thiophiles present in the reaction mixture. The use of phosphites
in this manner has previously been successfully demonstrated
by van Boom and co-workers in the BSP/Tf2O and diphenyl
sulfoxide/Tf2O activation of thioglycosides and hemiacetal
glycosyl donors.24-27 The stereochemistry of 29 was assigned
in the usual manner on the basis of the chemical shift of the
(22) Mootoo, D. R.; Konradsson, P.; Udodong, U.; Fraser-Reid, B. J. Am. Chem.
Soc. 1988, 110, 5583-5584.
(23) The use of substoichiometric amounts of triflic anhydride in the activation
of glycosyl sulfoxides has been similarly noted: (a) ref 2. (b) Yan, L.;
Kahne, D. Synlett 1995, 523-524. (c) Gildersleeve, J.; Pascal, R. A.; Kahne,
D. J. Am. Chem. Soc. 1998, 120, 5961-5969.
(24) (a) Code´e, J. D. C.; Litjens, R. E. J. N.; den Heeten, R.; Overkleeft, H. S.;
van Boom, J. H.; van der Marel, G. A. Org. Lett. 2003, 5, 1519-1522. (b)
Code´e, J. D. C.; van den Bos, J.; Litjens, R. E. J. N.; Overkleeft, H. S.;
van Boom, J. H.; van der Marel, G. A. Org. Lett. 2003, 5, 1947-1950. (c)
Code´e, J. D. C.; van den Bos, L. J.; Litjens, R. E. J. N.; Overkleeft, H. S.;
van Boeckel, C. A. A.; van Boom, J. H.; van der Marel, G. A. Tetrahedron
2004, 60, 1057-1064.
(25) For a recent synthesis of a small oligosaccharide library using the
thioglycoside/BSP method and involving the use of thioglycoside alcohols
as acceptors, see: Yamago, S.; Yamada, H.; Maruyama, T.; Yoshida, J.-i.
Angew. Chem., Int. Ed. 2004, 43, 2145-2148.
(17) (a) Yan, L.; Kahne, D. J. Am. Chem. Soc. 1996, 118, 9239-9248. (b)
Gildersleeve, J.; Smith, A.; Sakurai, D.; Raghavan, S.; Kahne, D. J. Am.
Chem. Soc. 1999, 121, 6176-6182.
(18) Smith, M.; Rammler, D. H.; Goldberg, I. H.; Khorana, H. G. J. Am. Chem.
Soc. 1962, 84, 430-440.
(19) (a) Garegg, P. J.; Iversen, T.; Oscarson, S. Carbohydr. Res. 1976, 50, C12-
C14. (b) Garegg, P. J.; Iversen, T.; Oscarson, S. Carbohydr. Res. 1977,
53, C5-C7. (c) Pozsgay, V. Carbohydr. Res. 1979, 69, 284-286. (d)
Lergenmuller, M.; Nukada, T.; Kuramochi, K.; Dan, A.; Ogawa, T.; Ito,
Y. Eur. J. Org. Chem. 1999, 1367-1376.
(20) Szurmai, Z.; Balatoni, L.; Lipta´k, A. Carbohydr. Res. 1994, 254, 301-
309.
(21) We note that this biphasic alkylation is very convenient on a multigram
scale with 31 being isolated in high yield by simple recrystallization.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15083

A R T I C L E S
Crich et al.
Scheme 2. Construction of Tri- and Disaccharide Donors 28 and 29 and of Acceptor 33
Scheme 3. Synthesis of Trisaccharides 15 and 18
trisaccharide 15 could be routinely isolated in 400 mg amounts,
sufficient for the completion of the synthesis. In preparation
for the final coupling, the PMP acetal of 15 was removed and
diol 17 converted to the terminal benzyl ether 18 by the
stannylene acetal protocol (Scheme 3).
In the final step of this convergent synthesis, the two tri-
saccharides, 18 and 28, were combined by the BSP protocol to
give â-hexasaccharide 26 and it’s R-anomer 36, in 35 and 53%
yields, respectively, with a â/R ratio of 1/1.5 (Scheme 4). Al-
though the selectivity in this final coupling was, again, disap-
pointing, it is stressed that the coupling was conducted with al-
most stoichiometric quantities of the two coupling partners (28/
18 ) 1.38/1) and, despite the poor selectivity, afforded 156 mg
of the desired hexasaccharide 26 in a single reaction. The spec-
tral data of hexasaccharide 26, obtained by the convergent block
synthesis, matched those from the sample obtained by the linear
route exactly, thereby substantiating the methods of stereochem-
ical assignment used in both syntheses. Finally, hydrogenolysis
of 26 over palladium on carbon afforded the target, alternating
mannan 37 in 93% isolated yield.30 This mannan displayed three
4,6-O-benzylidene-protected â-mannoside H-5 chemical shift
at δ 2.97. In addition, 29 displayed two anomeric carbons at δ
98.1 and 85.8 with 1JCH coupling constants of 153.0 and 166.2
Hz, respectively, consistent with the presence of a â-O-
glycoside28 and an R-S-glycoside.29 These same methods were
used to assign stereochemistry throughout the remainder of the
synthesis. Cleavage of the PMP acetal in 29 with acetic acid in
the usual manner afforded diol 32, and this was selectively
benzylated on the primary hydroxyl, giving 33, again, in the
standard manner using dibutyltin oxide and benzyl bromide
(Scheme 2). Coupling of disaccharide 33 to donor 30, under
the BSP deficient conditions, then completed the synthesis of
trisaccharide 28, which was obtained in 79% yield together with
7% of the R-anomer 34 (â/R ) 11/1).
1
distinct sets of anomeric resonances in the H and 13C NMR
spectra (D2O), consistent with the two â-(1f3)- and three
â-(1f4)-linkages and the terminal â-methyl mannoside. Impor-
(30) Needless to say, the identical substance was obtained on hydrogenolysis
of 26 derived from either the linear or the convergent routes.
(31) (a) Crich, D.; Dudkin, V. J. Am. Chem. Soc. 2002, 124, 2263-2266. (b)
Dudkin, V. Y.; Crich, D. Tetrahedron Lett. 2003, 44, 1787-1789. (c)
Dudkin, V. Y.; Miller, J. S.; Danishefsky, S. J. Tetrahedron Lett. 2003,
44, 1791-1793. (d) Kim, K. S.; Kang, S. S.; Seo, Y. S.; Kim, H. J.; Jeong,
K.-S. Synlett 2003, 1311-1314. (e) Crich, D.; de la Mora, M. A.; Cruz, R.
Tetrahedron 2002, 58, 35-44. (f) Crich, D.; Dai, Z. Tetrahedron 1999,
55, 1569-1580. (g) Nicolaou, K. C.; Mitchell, H. J.; Rodriguez, R. M.;
Fylaktakidou, K. C.; Suzuki, H.; Conley, S. R. Chem.sEur. J. 2000, 6,
3149-3165. (h) Crich, D.; Li, H. J. Org. Chem. 2002, 67, 4640-4646. (i)
Dudkin, V. K.; Miller, J. S.; Danishefsky, S. J. J. Am. Chem. Soc. 2004,
126, 736-738. (j) Mandal, M.; Dudkin, V. Y.; Geng, X.; Danishefsky, S.
J. Angew. Chem., Int. Ed. 2004, 43, 2557-2561. (k) Wu, X.; Schmidt, R.
R. J. Org. Chem. 2004, 69, 1853-1857. (l) Mong, T. K.-K.; Lee, H. K.;
Duro´n, S. G.; Wong, C.-H. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 797-
802. (m) Yun, M.; Shin, Y.; Chun, K. H.; Jen, S. Bull. Korean Chem. Soc.
2000, 21, 562-566. (n) Weingart, R.; Schmidt, R. R. Tetrahedron Lett.
2000, 41, 8753-8758. (o) Tsuda, T.; Sato, S.; Nakamura, S.; Hashimoto,
S. Heterocycles 2003, 59, 509-515. (p) Kim, K. S.; Kim, J. H.; Lee, Y.
J.; Lee, Y. J.; Park, J. J. Am. Chem. Soc. 2001, 123, 8477-8481. (q) Litjens,
R. E. J. N.; Leeuwenburgh, M. A.; van der Marel, G. A.; van Boom, J. H.
Tetrahedron Lett. 2001, 42, 8693-8696. (r) Duro´n, S. G.; Polat, T.; Wong,
C.-H. Org. Lett. 2004, 6, 839-841.
Trisaccharide 15 was readily assembled by coupling donor
29 with the simple acceptor 11 under the standard BSP
conditions (Scheme 3). Unfortunately, the selectivity of this
particular coupling was poor (∼1/1) and has remained so despite
considerable effort directed toward improving it. Nevertheless,
(26) The use of phosphites to clean up extraneous thiophiles in sulfoxide
glycosylations has also been described by van Boom and others: (a)
Sliedregt, L. A. J. M.; van der Marel, G. A.; van Boom, J. H. Tetrahedron
Lett. 1994, 35, 4015-4018. (b) Alonso, I.; Khiar, N.; Martin-Lomas, M.
Tetrahedron Lett. 1996, 37, 1477-1480.
(27) Other traps for thiophilic agents previously employed in the sulfoxide
method include alkenes and alkynes: (a) ref 17b. (b) Raghavan, S.; Kahne,
D. J. Am. Chem. Soc. 1993, 115, 1580-1581.
(28) Bock, K.; Pedersen, C. J. Chem. Soc., Perkin Trans. 2 1974, 293-297.
(29) Crich, D.; Li, H. J. Org. Chem. 2000, 65, 801-805.
9
15084 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004

Direct Chemical Synthesis of the
â-Mannans
A R T I C L E S
Scheme 4. Block Synthesis of the Hexasaccharide 26 and Completion of the Synthesis
tantly, all of the anomeric carbon resonances had 1JCH coupling
constants in the region of 157-162 Hz and serve, therefore, as
a final vindication of the stereochemical assignment.
synthesis begins with acceptor 11 and the block synthesis with
acceptor 31, both of which are coupled to donor 6 (Table 1 and
Scheme 1). Thus, starting from 11, we observe that the linear
sequence proceeds in 11 steps and 1.9% overall yield to
hexasaccharide 26 (Table 1). The block synthesis, on the other
hand, progresses from 31 to 26, with a longest linear sequence
of 8 steps and an overall yield of 4.4%, despite the modest yields
associated with the lack of stereoselectivity in the final two
couplings (Schemes 2-4). In oligosaccharide synthesis, as in
the total synthesis of other classes of molecules, the arithmetic
demon is best outwitted by well-designed, convergent routes.
The poor selectivity in the couplings of disaccharide donor
29 with acceptor 11 (Scheme 3) and of trisaccharide donor 28
with acceptor 18 (Scheme 4) is striking and stands in marked
contrast to the several successful highly stereoselective couplings
reported in the literature using our 4,6-O-benzylidene-protected
â-mannosyl donors or subsequent derivations.9,24c,31 However,
such poor selectivity is not entirely without precedent as we
have previously noted that 2-O-benzyl-4,6-O-benzylidene man-
nosyl donors carrying a bulky silyl ether on O-3 gave poor
selectivity in coupling reactions.32,33 The problem, a clear
limitation of our method, obviously lies with donors carrying
bulky groups on O-3. As before,32 we can only suggest that
bulky groups on O-3, whether protecting groups or glycosidic
units, suffer from compression between the benzylidene ring and
the benzyl ether on O-2. This results in a situation in which the
O-2 benzyl group populates a conformation in which it is located
above the anomeric carbon, thereby hindering nucleophilic
attack on the â-face and resulting in significantly reduced selec-
tivity. This rationale is bolstered by the recent observation of
high â-selectivity in the coupling reactions of donor 38, follow-
ing activation with diphenyl sulfoxide and triflic anhydride.24a,c
Whatever the eventual reason, the obvious conclusion is that
future block syntheses should be designed so as to avoid
mannosyl donors carrying oligosaccharide chains on O-3.
1
It is of some interest that the 13C and H NMR spectra of
hexasaccharide 37 (recorded in D2O at 65 °C) closely mimic
those of the naturally occurring â-(1f3)-â-(1f4)-mannan
(recorded in D2O at 70 °C5b and 65 °C7,34). Thus, in the natural
mannan, two anomeric carbon resonances are observed at δ
98.75b (97.5,6 98.77) and δ 101.65b (100.9,6 101.77). Likewise,
the natural mannan has two anomeric protons signals at δ 4.856
(4.857) and 4.736 (4.727), consistent with a highly regular
structure of alternating (1f3) and (1f4) linkages. Synthetic
mannan 37 has a pair of superimposed anomeric carbon signals
at δ 99.9, representing the two â-(1f3)-linkages, a second
unresolved pair at δ 102.9 together with a closely allied third
resonance at δ 103.1, indicative of the two internal and the one
terminal â-(1f4) anomeric bonds, and a final signal at δ 103.9,
belonging to the â-methyl glycoside. In the proton spectrum,
37 has two overlapping anomeric resonances at δ 4.86, a group
of three unresolved signals at δ 4.73, and an isolated resonance
at δ 4.59. Again, this is consistent with the presence of two
â-(1f3)-linkages, three â-(1f4) anomeric bonds, and the
terminal methyl glycoside (δ 4.59). It is evident from the
1
grouping of the anomeric signals in both the 13C and the H
NMR spectra and the parallels with the spectra of the natural
mannan that hexasaccharide 37 already shows the beginnings
of an organized, regular structure. This is to be contrasted with
the irregular structure of the collapsed helical â-(1f2)-
mannooctaose and the highly regular structure of the â-(1f4)-
mannohexaose, whose syntheses we have also completed.1 In
agreement with these observations, hexasaccharide 37, a white
microcrystalline solid, has a melting point of 168-170 °C,
It is appropriate to compare the two syntheses of the protected
hexasaccharides presented here. As the initial steps and donor
preparations are largely common to both syntheses, the linear
(32) Crich, D.; Dudkin, V. Tetrahedron Lett. 2000, 41, 5643-5646.
(33) This is to be contrasted with the 3-O-carboxylate-protected mannosyl donors
and the 2,3-O-carbonate-protected mannosyl donors, both with 4,6-O-
benzylidene acetals, which are extremely R-selective for a different set of
reasons: (a) ref 31h. (b) Crich, D.; Cai, W.; Dai, Z. J. Org. Chem. 2000,
65, 1291-1297. (c) Crich, D.; Vinod, A. U.; Picione, J. J. Org. Chem.
2003, 68, 8453-8458.
(34) The temperature at which the NMR spectra in ref 6 were measured was
not recorded.
9
J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004 15085

A R T I C L E S
Crich et al.
which, fittingly, sits between the melting points of >300 °C
that were recorded for the â-(1f4)-mannohexaose and the
gummy, noncrystalline nature of the â-(1f2)-mannooctaose.
Acknowledgment. We thank the NIGMS (GM 57335) for
support of this work.
Supporting Information Available: Full experimental details
and characterization data. This material is available free of
charge via the Internet at http://pubs.acs.org.
JA0471931
9
15086 J. AM. CHEM. SOC. VOL. 126, NO. 46, 2004