Towards an unprotected self-activating glycosyl donor system: Bromobutyl glycosides

Article abstract of DOI:10.1139/v02-029

Bromobutyl mannopyranosides have been successfully used as both protected and unprotected glycosyl donors both with and without the use of an external activator.

Full text of DOI:10.1139/v02-029

555
Towards an unprotected self-activating glycosyl
donor system: Bromobutyl glycosides
Benjamin G. Davis, Steven D. Wood, and Michael A.T. Maughan
Abstract: Bromobutyl mannopyranosides have been successfully used as both protected and unprotected glycosyl do-
nors both with and without the use of an external activator.
Key words: glycosylation, unprotected glycosyl donors, oligosaccharides.
Résumé : On a utilisé avec succès des mannopyranosides de bromobutyle comme donneurs, tant protégés que non pro-
tégés, de glycosyles et avec ou sans l’aide d’activateur externe.
Mots clés : glycosylation, donneurs de glycosyles non protégés, oligosaccharides.
[Traduit par la Rédaction] Davis et al. 558
Oligosaccharides and glycoconjugates are essential tools
for the investigation of the enormous variety of biological
functions that require specific carbohydrate-containing struc-
tures (1). Furthermore, their potential as therapeutic agents
is clear (2). As a result, the formation of the glycosidic link-
age continues to be a dominant theme in carbohydrate chem-
istry (3). Yet despite the development of many elegant
strategies, there is still no generally efficient and
stereoselective method available. To this end, a number of
glycosyl-donor systems have been developed, but few chem-
ical systems allow the use of unprotected glycosyl donors;
one exception has been Hanessian’s 3-methoxypyridyl
(MOP) glycoside system (4). It has been estimated that, on
average, the need for protecting groups introduces an addi-
tional six steps to each overall glycoside bond formation. If
the use of protecting groups could, therefore, be avoided or
limited, while maintaining control of reactivity, then overall
efficiencies may be improved. To this end, and as part of an
ongoing programme to develop novel glycosylation systems
and strategies (5), we have begun to investigate a new class
of glycosyl donors: 4-bromobutyl glycosides. This commu-
nication describes our first results in this area.
Reid’s powerful pentenyl glycoside class of glycosyl donors.
(6) This would not only yield THF as a volatile, non-
nucleophilic leaving group, but such a cyclization would
also be favoured over any potentially competing cyclic ether
formation with the free hydroxyls of the unprotected donor
through, for example, 8-exo cyclization of OH-2 onto the
primary bromide of the aglycon.
Clearly, the use of activation conditions that are compati-
ble with the presence of free hydroxyl groups is essential to
the use of unprotected glycosyl donors. We chose to investi-
gate two potential activation conditions: self-activation and
soft Lewis acid activation (e.g., Hal+, Ag+).
Our first target donor, mannoside 1, was prepared as
shown in Scheme 2. Penta-O-acetyl-D-mannose 6 was syn-
thesized in quantitative yield using well-established methods
from D-mannose and acetic anhydride with pyridine as a cat-
alyst. Our strategy for creating 1 relied on the different hard
and soft Lewis basicities of the anomeric O-1 and the
aglycon primary bromide of bromobutyl glycosides, respec-
tively (Scheme 1). Formation of 4-bromobutyl tetra-O-
acetyl-a-D-mannopyranoside 2 was accomplished through a
glycosidation using pentaacetate 6 and 4-bromobutanol, in
which the hard Lewis acid BF3·Et2O catalysed the loss of the
acetate group from the anomeric position of 6 without af-
fecting the soft Lewis base aglycon primary bromide or in-
deed without activating 2 as a glycosyl donor through the
mechanism envisaged in Scheme 1. In this regard, an addi-
tional aiding factor is that because of the disarmed (7) nature
of 2, this should be less reactive than the target unprotected
donor 1. 4-Bromobutanol was readily prepared through the
Our goal was to create a glycosylation system that (i) uses
sufficiently stable donors to allow preparation in unprotected
form but that can still be activated; and (ii) may self-activate
or activate under mild conditions. We reasoned that sponta-
neous 5-exo cyclization (Scheme 1) through nucleophilic at-
tack of the C-1 oxygen atom in 1, for example, would yield
an anomeric furanosyl cation, which would closely resemble
those postulated as intermediates in the activation of Fraser-
Received 31 October 2001. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 17 April 2002.
Dedicated to Professor J. Bryan Jones on the occasion of his 65th birthday.
B.G. Davis,1 and M.A.T. Maughan.2 Dyson Perrins Laboratory, University of Oxford, South Parks Road, Oxford OX1 3QY UK.
S.D. Wood. Department of Chemistry, University of Durham, South Road, Durham DH1 3LE U.K.
1Corresponding author (e-mail: Ben.Davis@chem.ox.ac.uk).
2Present address: Department of Chemistry, University of Durham, South Road, Durham DH1 3LE U.K.
Can. J. Chem. 80: 555–558 (2002)
DOI: 10.1139/V02-029
© 2002 NRC Canada

556
Can. J. Chem. Vol. 80, 2002
Scheme 1.
Scheme 2. Reagents and conditions: (i) Ac2O, py (3:2), 100%;
(ii) Br(CH2)3+nOH, DCM, BF3·Et2O, 43% for 2, 39% for 8;
(iii) MeONa, MeOH, 87% for 1, 98% for 7; (iv) Br(CH2)3+nOH,
BF3·Et2O, 18% for 1, 25% for 7.
OH
OH
OAc
OAc
(i)
OH
O
O
HO
HO
AcO
AcO
OAc
5
6
(ii)
(iv)
OH
OH
O
OAc
OAc
O
(iii)
HO
HO
AcO
AcO
O
O
Br
Br
( )n
( )n
1: n = 1
7: n = 0
2: n = 1
8: n = 0
yield) were also recovered, encouragingly suggesting that
through its action as a glycosyl donor, substitution of the
anomeric 4-bromobutoxy group by MeOH had occurred at
some point during the deprotection process. Since methyl
glycosides are not products of the deprotection of
mannosides (for example, the corresponding 3-bromopropyl
mannoside (vide infra)), the formation of these methanolysis
products under conditions not typically associated with
lability of the glycosidic bond gave us our first indication of
the successful action of 4-bromobutyl glycosides as glycosyl
donors. As a consequence of this methanolysis, care was re-
quired during deacetylation of 2. As might be anticipated,
removal of the disarming effect of the acetyl groups meant
that 1 would potentially prove more reactive as a glycosyl
donor than 2 and that, similarly, each of the intervening par-
tially deacetylated intermediates would prove even more re-
active than the previous. After extensive screening, the use
of a ~0.017 M methoxide solution freshly prepared from an-
hydrous MeOH and reaction times of 18 h proved optimal;
shorter reactions times led to incomplete deprotection,
whereas extended reaction times led to significant accumula-
reaction of THF with 48% HBr (aq.) (8), which was pre-
ferred for large-scale work over the use of Me2BBr (9).
More than five equivalents of BF3·Et2O were required for
effective conversion of 6 to the a-anomer 23 (the sole prod-
uct); the exclusive stereoselectivity for this trans-glycoside
may be attributed to neighbouring group participation by the
C-2 acetate group. While higher conversions were obtained
through the use of a large excess of 4-bromobutanol, the
43% yield of 2 obtained through the use of 1.1 equivalents
represents an efficient 83% yield when based on recovered
starting material and was, therefore, the preferred procedure
for scale-up. The use of higher equivalents of Lewis acid or
increased reaction times led to no significant change in the
yield of 2.
The final step in the formation of fully deprotected donor
4-bromobutyl tetra-O-acetyl-a-D-mannopyranoside 14 was
carried out in a single efficient (87% yield) Zemplén
deacetylation. Small amounts of methyl glycoside 1 (7%
3 2: BF3·Et2O (6.00 g, 42.3 mmol, 5.5 equiv.) was added dropwise to a solution of 6 (3.00g, 7.69 mmol) and 4-bromobutanol (1.30 g,
8.47 mmol, 1.1 equiv.) in dry DCM (20 mL) at 0°C under N2. After 1.5 h, the reaction solution was warmed to room temperature and
stirred for a further 16.5 h, when TLC (EtOAc–hexane, 50:50) showed conversion of starting material (Rf = 0.5) to a major product (Rf =
0.65). The reaction mixture was poured into ice water (10 mL) and extracted with DCM (3 × 20 mL). The organic extracts were combined,
washed with water (15 mL), dried (MgSO4), filtered, and the solvent removed. The residue was purified by flash chromatography
(EtOAc–hexane, 40:60) to give recovered bromobutanol, starting material 6 and product 2 (1.61 g, 43% yield, 83% based on recovered
starting material) as a white solid; mp 61–62°C. [a]D22 = + 46.4 (c, 0.34 in CHCl3). IR (cm–1): 1747 (C=O). 1H NMR (300 MHz, CDCl3) d:
1.78 (m, 2H, OCH2CH2CH2CH2Br), 1.88 (m, 2H, OCH2CH2CH2CH2Br), 2.00, 2.05, 2.11, 2.16 (4s, 4 × 3H, 4 × Ac), 3.46 (m, 2H, CH2Br),
3.72 (m, 2H, OCH2CH2CH2CH2Br), 3.96 (ddd, J = 10, 6, 3 Hz, 1H, H-5), 4.14 (dd, J = 12, 3 Hz, 1H, H-6), 4.29 (dd, J = 12, 6 Hz, 1H, H-
6>), 4.81 (d, J = 2 Hz, 1H, H-1), 5.23–5.33 (m, 3H, H-2, H-3, H-4). 13C NMR (300 MHz, CDCl3) d: 20.7, 20.7, 20.8, 20.9 (4 × CH3CO-),
27.9, 29.3 (OCH2CH2CH2CH2Br), 33.3 (CH2Br), 62.5, 66.1, 67.4, 68.6, 69.0, 69.6 (OCH2CH2CH2CH2Br, C-2, C-3, C-4, C-5, C-6), 97.6
(C-1), 169.8, 170.2, 170.3, 170.6 (4 × CH3CO-). ES-MS m/z (MeOH): 505, 507 (M+Na+). ES-HRMS calcd. for C18H31BrNO10: 500.1131;
found: 500.1131 ([M + NH4]+).
4 1: A freshly prepared solution of NaOMe–MeOH (3 mL, 0.1 M) was added to a solution of 2 (1.00 g, 2.07 mmol) in dry methanol (15 mL)
at room temperature under nitrogen. After 18 h, TLC (ethyl acetate–hexane, 50:50) showed the conversion of 2 (Rf = 0.65) to 1 (Rf = 0.05).
The mixture was run through a Dowex 50W (H+) plug (1 × 4 cm, eluant MeOH) and the solvent removed. The residue was purified by flash
chromatography (15% MeOH–CHCl3) to yield 1 (0.57 g, 1.82 mmol, 87%); mp 104–105°C. [a]2D2 = + 36.0 (c, 0.2 in MeOH). 1H
(250 MHz, CD3OD) d: 1.78 (m, 2H, OCH2CH2CH2CH2Br), 1.98 (m, 2H, OCH2CH2CH2CH2Br), 3.50 (m, 2H, CH2Br), 3.70 (m, 2H,
OCH2CH2CH2CH2Br), 3.60–3.90 (m, 5H, H-2, H-3, H-4, H-5, H-6), 3.96 (dd, J = 12, 3 Hz, 1H, H-6>), 4.78 (d, J = 2 Hz, 1H, H-1). 13C
(250 MHz, CD3OD) d: 30.0, 31.9 (OCH2CH2CH2CH2Br), 35.1 (CH2Br), 63.8, 68.5, 69.5, 73.1, 73.5, 75.6 (OCH2CH2CH2CH2Br, C-2, C-3,
C-4, C-5, C-6), 102.4 (C-1). ES-MS m/z (MeOH): 337, 339 ([M + Na]+, 100%). ES-HRMS calcd. for C10H23BrNO6: 332.0709; found:
332.0716 ([M + NH4]+). Anal. calcd. for C10H19BrO6: C 38.11, H 6.08; found: C 37.95, H 6.04.
© 2002 NRC Canada

Davis et al.
557
Table 1. Results of glycosylation reactions using bromobutylglycosides 1 and 2 as donors.
Donor
Activatora
Solvent
Reaction time (h)
Acceptor
Product
Yield(%)b
Product a:b ratio
1
1
1
1
1
2
2
2
2
—
—
IBr
AgOTf
AgOTf
—
IBr
AgOTf
AgOTf
DMF
DMF
DMF
DMF
DMF
DCM
DCM
DCM
DCM
10
60
12
24
20
60
60
24
20
MeOHc
MeOHc
MeOHc
MeOHc
DAGd
MeOHc
MeOHc
MeOHc
DAGd
3a
3a
3a
3a
3b
—
—
4a
4b
20
53
36
66
60
—
—
43
51
9:1
9:1
8:1
9:1
5:1
—
—
a only
a only
aAll reactions at room temperature.
bAll yields are for isolated products.
c5 equivalents used.
d1 equivalent used.
tion of methyl mannoside products. The anomeric mixture of
glycosides formed (the formation of some methyl b-D-
mannopyranoside suggests that there was a lack of neigh-
bouring group participation and therefore the lack of a C-2
acetate) and the disarming effect (7) of the acyl groups in 2,
likely means that methanolysis occurred when the donor was
fully deprotected as 1. It should be noted that although 1 is
an effectively active glycosyl donor, it is nonetheless stable
even to flash column chromatography on silica: an unusual
and highly convenient stability that may be attributed to sil-
ica as a mild, hard rather than soft, Lewis acidic medium. As
an alternative direct method for the direct synthesis of 1
from D-mannose 5, Fischer glycosidation was also investi-
Lewis acid. Using AgOTf, we were extremely pleased to ob-
serve a yield of 66% of 3a. These improved conditions were
also successfully applied to the synthesis of disaccharide 3b
(60% yield) through the use of diacetone galactose (DAG)
as an acceptor, thereby demonstrating potential for
oligosaccharide synthesis and confirming the compatibility
of the method with acid-sensitive protecting groups. The
possibility exists that although none was isolated, self-
condensation of 1 may account for the only fair yield of 3b,
and we are now investigating the use of O-6 protected do-
nors.
Having demonstrated the ability of 1 to act as a glycosyl
donor, peracetylated 2 was examined next. Consistent with
our earlier hypothesis (vide supra) and with the disarmed (7)
nature of 2, glycosylation did not proceed either in the ab-
sence of an activator or in the presence of IBr. Through the
use of AgOTf, however, fair yields of methyl mannoside 4a
(12) (43%) and disaccharide 4b (51%) were obtained. The
exclusive a-stereoselectivity that was observed for 4a,b may
be attributed to neighbouring group participation by the C-2
acetate group.
Finally, to test the hypothesis that cyclization to form a
furanosyl oxonium leaving group is an essential precursor
step in the mechanism of glycosylation, two analogous
peracetylated (8) and deprotected (7) 3-bromopropyl manno-
pyranosides were prepared using analogous routes to those
used for the preparation of 1 and 2 (Scheme 2). Consistent
with the need for cyclization, neither 7 or 8 were able to act
as glycosyl donors under the conditions that had proved suc-
cessful for 1 or 2.6
In summary, this communication describes the first exam-
ples of a new class of glycosyl donor: bromobutylglycosides
(BBGs). Deprotected BBG 1 appears to possess a remark-
ably balanced reactivity that allows its ready preparation and
gated. Treatment of
a suspension of 5 in neat 4-
bromobutanol with BF3·Et2O gave, after extensive purifica-
tion, a low yield (18%) of 1; although this was only a one-
step procedure, inferior overall yield and use of large quanti-
ties of 4-bromobutanol meant preparation via 2 was selected
as a superior route.
The results of glycosylation reactions5 with 1 and 2 are
shown in Table 1. We were delighted to see that simply by
stirring 1 with MeOH as a glycosyl acceptor and without ac-
tivator, methyl mannoside 3a (10) was formed in a low 20%
yield after 10 h, but in a fair 53% yield after 60 h. The a:b
stereoselectivity (9:1) observed is consistent with the
stereoselectivities of other non-participatory mannosyl do-
nors. (3) Although these results were encouraging, we rea-
soned that use of a soft Lewis acid might increase rate and
(or) efficiency because of the differing nature of the Lewis
basicities of 1 (Scheme 1). Thus, the use of IBr (11) led af-
ter 12 h to increased rate and yield when compared with
glycosylations without an activator. Unfortunately, pro-
longed reaction times with IBr did not lead to enhanced
yield and so AgOTf was tested as an alternative halophilic
5 General procedure for glycosylation reactions: To a solution of 1 (100 mg, 0.32 mmol, 1 equiv.) in DMF (5 mL) under nitrogen, was added
acceptor MeOH (64 L, 1.6 mmol, 5 equiv.). The solution was stirred at room temperature for 60 h, any precipitate removed by
centrifugation and the solvent removed. The residue was purified by flash chromatography (15% MeOH– CHCl3) to yield 3a (33 mg, 53%,
a:b = 9:1) as a white solid. Parallel reactions were repeated under similar conditions using, IBr (66 mg, 0.32 mmol, 1 equiv.) or silver
triflate (82 mg, 0.32 mmol, 1 equiv.) to yield 22 mg (a:b, 8:1, 36%, after 12 h) or 41 mg (a:b, 9:1, 66%, after 24 h) of 3a, respectively.
6 As usefully highlighted by a referee, the inactivity of 3-bromopropyl glycosides does not exclude the possibility of direct Fischer
glycosylation that could be catalyzed by the equivalent of HBr (in spontaneous and IBr-activated reactions) or TfOH (in AgOTf-activated
reactions) that is liberated. To test this hypothesis, 1 was used to glycosylate MeOH activated with AgOTf in the presence of the hindered
base 2,4,6-tri-tert-butylpyridine (TTBP) [D. Crich D, M. Smith, Q.J. Yao, J. Picione, Synthesis, 323 (2001).] and yielded 64% of 3a after
24h; near identical to that obtained without base. Other aspects of the mechanism, including the detection of THF, are currently in progress.
© 2002 NRC Canada

558
Can. J. Chem. Vol. 80, 2002
2. J.C. McAuliffe and O. Hindsgaul. Chem. Ind. 170 (1997);
B.G. Davis. Chem. Ind. 134 (2000); K.M. Koeller and C.-H.
Wong. Nature Biotechnol. 18, 835 (2000).
3. H. Paulsen. Angew. Chem. Int. Ed. Engl., 21, 155 (1982); K.
Toshima and K. Tatsuta. Chem. Rev. 93, 1503 (1993); G.J.
Boons. Contemp. Org. Synth. 3, 173 (1996); G.J. Boons. Tet-
rahedron, 52, 1095 (1996); B.G. Davis. J. Chem. Soc. Perkin
Trans. 1, 2137 (2000).
4. S. Hanessian and B.L. Lou. Chem. Rev. 100, 4443 (2000).
5. R.J. Tennant-Eyles, B.G. Davis, and A.J. Fairbanks. Chem.
Commun. 1037 (1999); R.J. Tennant-Eyles, B.G. Davis, and
A.J. Fairbanks. Tetrahedron: Asymmetry, 11, 231 (2000); B.G.
Davis. J. Chem. Soc. Perkin Trans. 1, 2137 (2000); B.G. Da-
vis, S.J. Ward, and P.M. Rendle. Chem. Commun. 189 (2001).
6. B. Fraser-Reid, U.E. Udodong, Z.F. Wu, H. Ottosson, J.R.
Merritt, C.S. Rao, C. Roberts, and R. Madsen. Synlett, 927
(1992).
purification and yet is sufficiently reactive to act as an un-
protected glycosyl donor even in the absence of activator. In
addition, although 1 is strongly activated by soft Lewis acid
catalysis its relative resistance to hard Lewis acid catalysis
allows, for example, its direct preparation through Fischer
glycosidation. Although the yields for glycosylations using
BBGs are thus far only fair or low,7 the potential to avoid
the use of protecting groups might offset overall glyco-
sylation efficiencies. The breadth of the utility of 1 with
other acceptors and the investigation of bromobutyl-
glycosides of other parent carbohydrates is underway, the re-
sults of which will be presented in due course.
Acknowledgments
7. For the background to the use of armed–disarmed donors (in
which the reactivity is increased or decreased by the use of
electron donating or withdrawing protecting groups, respec-
tively), see: D.R. Mootoo, P. Konradsson, U. Udodong, and B.
Fraser-Reid. J. Am. Chem. Soc. 110, 5583 (1988); G.J. Boons,
P. Grice, R. Leslie, S.V. Ley, and L.L. Yeung. Tetrahedron
Lett. 34, 8523 (1993).
We would like to thank HEFCE (SDW), the Royal Society
(BGD), and the University of Oxford for funding and the
EPSRC for access to the Mass Spectrometry Service at
Swansea and the Chemical Database Service at Daresbury.
Last but not least we would like to thank Prof. J. Bryan
Jones for his constant inspiration and encouragement.
8. E. Vedejs, M. J. Arnost, and J. P. Hagen. J. Org. Chem. 44,
3230 (1979).
9. Y. Guindon, M. Therien, Y. Girard, and C. Yoakim. J. Org.
Chem. 52, 1680 (1987).
10. E. Fischer and L. Beensch. Chem. Ber. 29, 2927 (1896).
11. K.P.R. Kartha and R.A. Field. Tetrahedron Lett. 38, 8233
(1997).
References
1. A. Varki. Glycobiology, 3, 97 (1993); R.A Dwek. Chem. Rev.,
96, 683 (1996); B.G. Davis. J. Chem. Soc. Perkin Trans. 1,
3215 (1999).
12. J.K. Dale. J. Am. Chem. Soc. 46, 1046 (1924).
7 The possibility of competing 1,6-anhydromannose formation was raised by a referee as a reason for the moderate yields. Although none was
isolated as a by-product, the synthesis of 6-O-protected donors to test this hypothesis is underway.
© 2002 NRC Canada