Reagent switchable stereoselective β(1,2) mannoside mannosylation: OH-2 of mannose is a privileged acceptor

Article abstract of DOI:10.1039/b803999m

The discovery of novel conditions for highly β-stereoselective (>9: 1) mannosylation of OH-2 of mannosides using a straightforward perbenzylthioglycoside donor has allowed ready assembly of β-mannosyl oligosaccharides including the repeating trisaccharide motif of the O5 antigen of pathogen Klebsiella pneumoniae. The Royal Society of Chemistry.

Full text of DOI:10.1039/b803999m

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Reagent switchable stereoselective b(1,2) mannoside mannosylation:
OH-2 of mannose is a privileged acceptor†‡
Katie J. Doores and Benjamin G. Davis*
Received 7th March 2008, Accepted 22nd April 2008
First published as an Advance Article on the web 2nd June 2008
DOI: 10.1039/b803999m
The discovery of novel conditions for highly b-stereoselective (>9 : 1) mannosylation of OH-2 of
mannosides using a straightforward perbenzylthioglycoside donor has allowed ready assembly of
b-mannosyl oligosaccharides including the repeating trisaccharide motif of the O5 antigen of pathogen
Klebsiella pneumoniae.
The stereoselective formation of cis-glycosidic linkages is still
Crich et al. have used kinetic isotope effect experiments to explore
a
major synthetic challenge in the synthesis of complex
this mechanism and suggest that the 4,6-O-benzylidene acts to
prevent rehybridisation at the anomeric carbon thus favouring
the triflate intermediate.17 This methodology has been widely
used in the synthesis of biologically significant oligosaccharides
containing cis-glycosidic linkages,2,18,19 although some linkages still
remain as difficult targets.3,20
Our laboratory has been particularly interested in the synthesis
of fragments of D-mannosyl-containing sections of pathogen
coats such as the high-mannose oligosaccharide 1 found on
the surface of the HIV-1 envelope glycoprotein gp120,21 which
contains predominantly alpha mannosides, and the O5 antigen of
the opportunistic gram-negative pathogen Klebsiella pneumoniae
serotype O5 2 (Fig. 2). This paper describes the development of
mannosylation methodology and its application to the synthesis
of key di-, tri- and tetra-saccharide fragments of both of these
structures.
oligosaccharides.1–3 Although significant advances have been made
in methodology, these methods are not fully understood and
generalities are often hard to find.2
Common methods for the synthesis of b-mannosides (Fig. 1)
involve the synthesis of a b-glucoside, followed by inversion of
C-2 configuration either by a nucleophilic displacement4,5 or an
oxidation–reduction procedure.6,7 Intramolecular aglycon delivery
has also been developed by several groups as a method for the
synthesis of 1,2-cis glycosides and has been applied to the synthesis
of b-mannosides.8–14
Fig. 1 Selected, illustrative methods for the synthesis of b-mannosides.
The use of sulfoxide glycosyl donors, first activated with
triflic anhydride before addition of the acceptor, has also proved
successful.15,16 In such systems, the use of 4,6-O-benzylidene-type
protection with non-participating groups on O-2 and O-3 of the
donor is important in gaining high b-selectivity. The mechanism
proposed involves the formation of an a-triflate, which then
undergoes an SN2-like displacement giving b-mannoside product.
Fig. 2 D-Mannosyl targets: high mannose oligosaccharide of gp120 1 and
O5 antigen from gram-negative pathogen Klebsiella pneumoniae serotype
O5. Target fragments highlighted in grey.
Department of Chemistry, University of Oxford, Chemistry Research
Laboratory, Mansfield Road, Oxford, UK OX1 3TA. E-mail: ben.davis@
chem.ox.ac.uk; Fax: +44 (0) 1865 275674; Tel: +44 (0) 1865 275654
† This paper is dedicated to Prof. Andrew B. Holmes and, in particular, his
contributions to natural product and polyol syntheses that have inspired
our group and others.
Glycan 1 contains a number of key a1,2 and a1,3 mannosyl
linkages. The alpha-mannosyl linkage configuration is, as a trans
linkage, one of the most ready to form in a stereoselective manner
being favoured both by participating groups, sterics and the
anomeric effect.22,23 Access to the cis beta-mannosyl linkage there-
fore remains a key challenge in organic chemistry (vide supra).2
‡ Electronic supplementary information (ESI) available: Experimental. See
DOI: 10.1039/b803999m
2692 | Org. Biomol. Chem., 2008, 6, 2692–2696
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
We were therefore particularly surprised and pleased to find that
using the recently reported dimethyl disulfide-triflic anhydride24
conditions (conditions A) for activation of thioglycosides, we
observed the formation of Man-b1,2-Man product in very good
yield and good selectivity (b : a > 9 : 1) (Scheme 1). We reasoned
that this unexpected selectivity might be applied to the synthesis
of the repeating trisaccharide unit of 2, which contains a key b1,2-
mannosyl linkage.
more modest, consistent with the more hindered acceptor. These
results and their switchability are summarized in Table 2.
A range of representative sugar alcohol acceptors other than
OH-2 mannosides were then tested (primary hydroxyl 8, secondary
axial hydroxyl 10, secondary equatorial hydroxyl 12, 14, 17).
Together the results suggested a unique privileged reactivity of
the 2-OH of mannosides. Glycosylation of less hindered primary
acceptor 8 using donor 3 gave a poor b-selectivity (9 b : a =
1 : 3) in 53% yield (entry 4). Glycosylation of the hindered axial
4-OH in the galactose acceptor 10 gave 11 exclusively with a-
selectivity in 75% yield (entry 5). Next the GlcNAc acceptor 12
was used, which in principle would allow access to one of the key
linkages (Manb1,4GlcNAc) of the core pentasaccharide found in
N-linked glycoproteins; under glycosylating conditions A, only the
a-product 13 was observed in 78% yield (entry 6). Glycosylation
with the glucoside acceptor 14 also only gave the a anomer in 64%
yield (entry 7). Even use of disaccharide thioglycoside donor 16
gave only the a-product 18 from secondary equatorial hydroxyl
acceptor 17 (entry 8).
These striking results suggested that this strongly selective
transformation is only applicable to the synthesis of b-(1→2)-
mannosylmannoside linkages and that this reactivity is switchable
only for this linkage (Table 2).
This highly selective yet narrowly applicable methodology was
then tested in the target synthesis of the O5 antigen of the oppor-
tunistic gram-negative pathogen Klebsiella pneumoniae serotype
O5. This organism’s cell wall contains a repeating trimannoside
motif found in 22 (Fig. 2 and Scheme 2), which critically contains
the target Manb(1,2)Man linkage that is addressable by our
method.28,29 Thus, disaccharide acceptor 21 was reacted with
donor 3 using the triflic anhydride and dimethyldisulfide activating
conditions (conditions A). As expected, this successfully installed
the b-(1→2)-mannoside linkage in 84% yield and with high b-
selectivity (a : b ratio of 1 : 11); the two anomers were found
to be easily separable by column chromatography. Again, use of
conditions B did not yield target 22 but instead gave primarily
alpha linked product (Table 2).
In conclusion, we have shown that donor 3 when activated
using triflic anhydride and dimethyldisulfide is an efficient reagent
for the synthesis of b-(1→2)-mannoside linkages. The selectivity
of this reaction dramatically decreased during the synthesis of
other glycosidic linkages. This method is therefore narrow in the
linkage type to which it might be applied but was successfully
applied here to the synthesis of the naturally-occurring pathogen
trisaccharide motif 22. This reaction is of particular note as
the donor has not been conformationally constrained by the
presence of a 4,6-O-benzylidene protecting group and is reagent
switchable. We are currently exploring the mechanistic origin of
this startling selectivity but some mechanistic rationale for the
observed outcomes may be suggested with the caveat that detailed
studies are ongoing.
Scheme 1 (i) Conditions A: Tf2O, Me2S2, TTBP, 4 A mol sieves, −78 ◦C,
˚
(ii) conditions B: DMTST, TTBP, 4 A mol sieves, −78 ◦C→RT.
˚
Reaction of fully benzylated thiophenyl mannosyl donor 3
with acceptor 4 gave the b-product 5b in a >9 : 1 ratio in 95%
yield (Scheme 1).25 However, in striking contrast, when the same
glycosylation was performed using the powerful and long used
alkylsulfenylating agent dimethyl(methylthio) sulfonium triflate
(DMTST, conditions B)26 the a-product 5a only was obtained
in 67% yield. The identity and configuration of the products
were unambiguously identified (as for all mannosides) by a range
of methods including, critically, H-13C J coupling constants,27
which in all cases were in the range 150–156 Hz for b-products
and 170–172 Hz for all a-products. The reagent switchable a↔b
selectivity observed here for the b-1,2-mannosylations is to the best
of our knowledge unprecedented. The b-selectivity obtained with
conditions A was of particular additional significance as donor 3
does not possess the 4,6-O-benzylidene protection previously used
in other systems such as the Crich sulfoxides.16
These intriguing results led us to investigate the scope of
these reaction conditions and the resulting stereoselectivity in the
formation of other mannosides (Table 1). First, mannosylation
under conditions A was performed using the same acceptor 3 with
the alternative peracetylated donor 5 that bears a participating
O-2 protecting group: unsurprisingly, participation dominates
selectivity and only the a-product 6 was obtained in a yield of 55%
(entry 3, table 1). Yet, the strong b-selectivity observed for OH-2
of D-mannose was again observed, under the dimethyl disulfide-
triflic anhydride conditions (conditions A), during the installation
of a terminal mannosyl residue to the truncated D1-arm acceptor
trisaccharide 19: only the b-linked product 20b was observed in
47% yield (entry 9). Consistent with the reagent control shown in
the reaction of 3 with 4, application of conditions B to the same
reactions gave primarily a-linked product 20a(entry 10). Yields for
both of these mannosylations of 19 to yield tetrasaccahrides are
1
1
Given the simplicity of the donor system, which contrasts
with the designed use of, for example, 4,6-O-benzylidenes and
other ‘torsionally disarmed’ donors,16,30 the mechanistic origins of
the reagent-switchable selectivity in this more conformationally
flexible system are intriguing. Electronic effects in both 4,6-O-
benzylidene mannosyl donors31 and related uronic acids32 are also
suggested to influence stereoselectivity; again these are features
that per-benzylated donor 3 lacks. This suggests that other factors
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2692–2696 | 2693
©

View Article Online
Table 1 Representative mannosylations with triflic anhydride (conditions A) or dimethyl disulfide (conditions B)
Entry
1
Donor
Acceptor
Product
Conditions
A
Yield (b/a)
95% (>9 : 1)
2
3
4
4
B
67% (a only)
3
4
A
A
55% (a only)
53% (1 : 3)
5
3
A
75% (a only)
6
7
8
3
3
A
A
A
78% (a only)
64% (a only)
68% (a only)
2694 | Org. Biomol. Chem., 2008, 6, 2692–2696
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 1 (Conti.)
Entry
9
Donor
Acceptor
Product
Conditions
A
Yield (b/a)
47% (b only)
10
3
19
B
45% (1 : 2.5)
Table 2 Summary of b(1,2)-mannosylations
Entry
Donor
Acceptor
Conditions
b
a
Yield (%)
1
2
3
4
5
6
3
3
3
3
3
3
4
A
B
A
B
A
B
>9
0
1
1
11
1
1
1
0
2.5
1
95
67
47
41
84
75
4
19
19
21
21
3
Fig. 3 Putative differential activation of 2 leading to potential equili-
bration of intermediates (selected examples of alpha sulfeniums shown,
although, beta intermediates and triflates cannot be discounted).
Scheme 2 Preparation of the repeating trisaccharide motif of the O5
antigen of pathogen Klebsiella pneumoniae.
dominate. The donors used here are exclusively alpha thioman-
nosides with an axial leaving group. Were clean SN2-chemistry
to dominate, as has been suggested in a number of systems,17
then beta products would be expected; SN2 chemistry has been
invoked to explain the beta selectivity obtained from intermediate
mannosyl alpha-triflates.17 Deuterium secondary kinetic isotope
effect experiments17 are currently underway to address this issue.
In this context, the switch in activation conditions (A to B)
differs essentially only in the nature of the leaving group for the
sulfenium reagent that is generated in situ (see Fig. 3). Putative
post-activation equilibration might give rise to a number of
intermediates with varied anomeric leaving groups (or corre-
sponding glycosyl cation·anion pairs). The different conditions
(A and B) may differ in their outcome as a result of cleaner SN2-
substitution on an intermediate such a IntA. Initial experiments
intended to observe by NMR the intermediates formed from
2 upon activation by either conditions were inconclusive. Of
course, under conditions of rapid equilibration, the differing
rates of intermediate substitution will critically determine reaction
product distribution and hence observation of the equilibrium
position of these intermediates would add little to illuminating the
origin of selectivity.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 2692–2696 | 2695
©

View Article Online
The restriction of this beta selectivity to only the OH-2 manno-
side acceptors indicates a strong role for the nucleophile/acceptor
in these reactions. The axial OH-4 hydroxyl of acceptor 9 gives
rise to exclusive alpha mannosylation products, indicating that
this substrate-controlled aspect of stereoselectivity extends beyond
the relative conformational position of the OH-2 nucleophile
of the alpha-mannoside acceptors that we have shown here
are privileged acceptors. The possibility therefore remains that
the unique O1ax/O2ax/O3eq geometry plays a key role in
determining selectivity and other acceptors containing this relative
configuration are being now being explored.
14 A. Dan, Y. Ito and T. Ogawa, J. Org. Chem., 1995, 60, 4680–4681.
15 D. Crich and S. X. Sun, J. Org. Chem., 1997, 62, 1198–1199.
16 D. Crich and S. X. Sun, Tetrahedron, 1998, 54, 8321–8348.
17 D. Crich and N. S. Chandrasekera, Angew. Chem., Int. Ed., 2004, 43,
5386–5389.
18 D. Crich, W. J. Li and H. M. Li, J. Am. Chem. Soc., 2004, 126, 15081–
15086.
19 D. Crich, A. Banerjee and Q. J. Yao, J. Am. Chem. Soc., 2004, 126,
14930–14934.
20 A. V. Demchenko, Curr. Org. Chem., 2003, 7, 35–79.
21 D. A. Calarese, C. N. Scanlan, M. B. Zwick, S. Deechongkit, Y. Mimura,
R. Kunert, P. Zhu, M. R. Wormald, R. L. Stanfield, K. H. Roux, J. W.
Kelly, P. M. Rudd, R. A. Dwek, H. Katinger, D. R. Burton and I. A.
Wilson, Science, 2003, 300, 2065–2071.
22 G. J. Boons, Tetrahedron, 1996, 52, 1095–1121.
23 B. G. Davis, J. Chem. Soc., Perkin Trans. 1, 2000, 2137–2160.
24 J. Tatai and P. Fugedi, Org. Lett., 2007, 9, 4647–4650.
25 Conditions A: acceptor (1 eq), donor (1.1 eq) and 2,6-di-tert-butyl-4-
methylpyridine (5 eq) were dried in a desiccator overnight. The reagents
were dissolved in DCM (2 mL) and transferred using a cannula to a
Acknowledgements
We would like to thank the International AIDS Vaccine Initiative
(Neutralizing Antibody Consortium) for funding.
˚
flame dried flask containing 4 A molecular sieves. The mixture was
stirred for 1 h and cooled to −78 ◦C. DCM (2 mL) was added to a
˚
flame dried flask containing 4 A molecular sieves and stirred for 1 h
Notes and references
then cooled to 0 ◦C. To this flask was added dimethyldisulfide (4 eq) and
trifluoromethylsulfonic anhydride (4 eq). After 2 min, the solution was
transferred to the flask containing the sugar reagents at −78 ◦C. The
mixture was stirred at −78 ◦C under an atmosphere of argon. After
1 h, the reaction mixture was quenched with triethylamine (1 mL) and
filtered through celiteꢀR . The filtrate was concentrated in vacuo and the
residue purified by flash column chromatography.
1 H. Paulsen, Angew. Chem., Int. Ed. Engl., 1990, 29, 823–839.
2 J. J. Gridley and H. M. I. Osborn, J. Chem. Soc., Perkin Trans. 1, 2000,
10, 1471–1491.
3 A. J. Fairbanks, Synlett, 2003, 1945–1958.
4 S. David, A. Malleron and C. Dini, Carbohydr. Res., 1989, 188, 193–200.
5 J. Alais and S. David, Carbohydr. Res., 1990, 201, 69–77.
6 M. A. Shaban and R. W. Jeanloz, Carbohydr. Res., 1976, 52, 103–114.
7 N. K. Kochetkov, B. A. Dmitriev, N. N. Malysheva, A. Y. Chernyak,
E. M. Klimov, N. E. Bayramova and V. I. Torgov, Carbohydr. Res.,
1975, 45, 283–290.
26 P. Fugedi and P. J. Garegg, Carbohydr. Res., 1986, 149, C9–C12.
27 K. Bock and C. Pedersen, J. Chem. Soc., Perkin Trans. 2, 1974, 293–299.
28 P.-E. Jansson, J. Lonngren, G. Widmalm, K. Leontein, K. Slettengren,
S. B. Svenson, G. Wrangsell, A. Dell and P. R. Tiller, Carbohydr. Res.,
1985, 145, 59–66.
8 F. Barresi and O. Hindsgaul, J. Am. Chem. Soc., 1991, 113, 9376–9377.
9 F. Barresi and O. Hindsgaul, Can. J. Chem., 1994, 72, 1447–1465.
10 Y. Ito, Y. Ohnishi, T. Ogawa and Y. Nakahara, Synlett, 1998, 1102–
1104.
29 J. D. C. Codee, L. H. Hossain and P. H. Seeberger, Org. Lett., 2005, 7,
3251–3254.
30 T. Nukada, A. Berces and D. M. Whitfield, Carbohydr. Res., 2002, 337,
765–774.
11 I. Cumpstey, A. J. Fairbanks and A. J. Redgrave, Org. Lett., 2001, 3,
2371–2374.
31 H. H. Jensen, L. U. Nordstrom and M. Bols, J. Am. Chem. Soc., 2004,
12 C. M. P. Seward, I. Cumpstey, M. Aloui, S. C. Ennis, A. J. Redgrave
and A. J. Fairbanks, Chem. Commun., 2000, 1409–1410.
13 Y. Ito and T. Ogawa, Angew. Chem., Int. Ed. Engl., 1994, 33, 1765–1767.
126, 9205–9213.
32 L. J. Van Den Bos, J. Dinkelaar, H. S. Overkleeft and G. A. Van Der
Marel, J. Am. Chem. Soc., 2006, 128, 13066–13067.
2696 | Org. Biomol. Chem., 2008, 6, 2692–2696
This journal is
The Royal Society of Chemistry 2008
©