In situ formation of β-glycosyl imidinium triflate from participating thioglycosyl donors: Elaboration to disarmed-armed iterative glycosylation

Article abstract of DOI:10.1039/c2cc35032g

β-Glycosyl imidinium triflate is generated from participating thioglycoside donors for disarmed-armed iterative glycosylations and one-pot oligosaccharide synthesis.

Full text of DOI:10.1039/c2cc35032g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 10910–10912
www.rsc.org/chemcomm
COMMUNICATION
In situ formation of b-glycosyl imidinium triflate from participating
thioglycosyl donors: elaboration to disarmed–armed iterative
glycosylationw
Yu Hsien Lin, Bhaswati Ghosh and Kwok-Kong Tony Mong*
Received 13th July 2012, Accepted 17th September 2012
DOI: 10.1039/c2cc35032g
b-Glycosyl imidinium triflate is generated from participating
thioglycoside donors for disarmed–armed iterative glycosylations
and one-pot oligosaccharide synthesis.
Oligosaccharide synthesis is a challenging and tedious endeavour.1
To streamline the process, sequential glycosylation strategies for
oligosaccharide synthesis have emerged.2,3 Recently, we developed
a DMF-modulated glycosylation method for non-participating
glycosyl donors, and this method was elaborated to iterative
1,2-cis a-glycosylations.4 Because oligosaccharides contain
Scheme 1 Development of a DMF-modulated glycosylation procedure
b- and a-glycosidic bonds, it is reasonable and necessary
to extend the modulation concept to the construction of
b-glycosidic linkages.5
with participating thioglycoside donors 1 and 2.
Table 1 DMF-modulated glycosylation with participating thioglycoside
donors 1 and 2
In glycosylation, activation of participating glycosyl donors
produces the glycosyl oxacarbenium ion,6 which under the
influence of neighbouring group participation (NGP) would be
converted to the glycosyl dioxalenium ion.7–9 As the participating
group blocks the a-stereotopic face of the dioxalenium ion,
subsequent reaction with DMF produces b-glycosyl imidinium
triflate. Given that the dioxalenium ion and b-imidinium triflate
are in equilibrium, both may react with an acceptor to yield a
mixture of a/b anomers. In previous studies, the amount of
DMF used in modulated glycosylation was related to the
degree of a-selectivity.4 Extrapolating from this relationship,
we hypothesized that the amount of DMF might be modified so
as to subdue the a-directing effect and yet retain the pre-activation
property of DMF.
Producta
Entry Donor DMF (equiv.) Time (h) No. Yield% a/b
1
2
3
4
5
6
7
8
1
1
1
1
1
1
2
2
7.2
2.4
1.2
0.6
0.0
1.2d
0.6
1.2
6.0
3.0
4.0
4.0
4.0
3
4
4
4
4
4
4
5
5
72
70
75
1 : 5
1/22
b only
NDc
ND
ND
NDc
a only
Traceb
Traceb
Traceb
—
3
4
71
a
b
a/b-Anomeric ratios were determined by HPLC. CH3CN was used
c
d
as solvent. The reactions gave a complex reaction mixture. ND
stands for not determined.
Reducing the idea to practice, participating 2-O-benzoyl
thioglucoside 110 and thiomannoside 2 were employed as
model donors for development of a new DMF-modulated
glycosylation procedure. At the start, thioglucoside 1 (1.2 equiv.)
was activated with N-iodosuccinimide (1.2 equiv., NIS) and
trimethylsilyltrifluoromethane sulfonate (1.2 equiv., TMSOTf) in
the absence or the presence of 0.6 to 7.2 equiv. of DMF.11 Upon
completion of the activation, galactose acceptor 3 (1.0 equiv.)
was added (Scheme 1 and Table 1). In the presence of 7.2 equiv.
of DMF, the glycosylation afforded the expected disaccharide
4 in 72% yield but with a 1 : 5 a/b-anomeric ratio (entry 1).
Reducing the amount of DMF to 2.4 equiv. suppressed the
a-anomer formation (entry 2), but complete elimination
occurred with 1.2 equiv. of DMF (entry 3). Further reduction
or absence of DMF was unproductive resulting in a complex
reaction mixture (entries 4 and 5). The modulated glycosylation
did not work when acetonitrile was used as solvent (entry 6). Some
optimization of reaction temperature is required for the present
glycosylation (see Table S1 in ESIw). Besides the thioglucoside 1,
the modulated glycosylation procedure was effective for glycosyla-
tion with thiomannoside 2 (entries 7 and 8).
1001, Ta Hsueh Road, Hsinchu, 300, Taiwan.
E-mail: tmong@mail.nctu.edu.tw; Fax: +886-3-5131204;
Tel: +886-3-5712121
w Electronic supplementary information (ESI) available: Preparation
of building blocks 2, 6–17, 32–34, and 38 and NMR, MS data of
glycosylation products. See DOI: 10.1039/c2cc35032g
Because the DMF-modulated glycosylation includes a pre-
activation step, this opens the door to iterative glycosylations.
Such a feature is particularly beneficial for thioglycosyl donors
c
10910 Chem. Commun., 2012, 48, 10910–10912
This journal is The Royal Society of Chemistry 2012

View Article Online
Scheme 2 Disarmed–armed glycosylation approach directed by
DMF modulation.
having a C2 acyl protecting function because they are often
regarded as disarmed (less reactive) donors in conventional
glycosylations though S-benzoxazolyl 2-O-benzoyl-3,4,6-tri-
O-benzyl-D-glycosides have been found to be super-reactive
under specific activation conditions.12 In general, a disarmed
donor would not be used for glycosylation of an armed (more
reactive) glycosyl acceptor using the conventional procedure,13
though several iterative glycosylations have been established to
overcome the restriction.14 We envisioned that by the application
of the DMF modulation concept, a convenient disarmed–armed
iterative glycosylation should be at our disposal.
As a proof-of-concept study, thioglycosides 1, 2, 6, 7, and 8
were employed as disarmed donors for glycosylations of armed
(more reactive) thioglycoside acceptors 9–17 using the DMF
modulation procedure (Scheme 2 and Fig. 1). The reactivity
of the thioglycosides was unambiguously defined by relative
reactivity values (RRV).15–17
At first, the glycosylation of O-glucoside 9 with thiogluco-
side 1 produced the desired b-anomer 18 in 65% yield,
(Table 2, entry 1). Next we evaluated the glycosylations of
armed thioglycosides 10 (RRV = 4470), 11a (RRV = 4600),
16 (RRV = 50 000), and 17 (RRV = 11 000) with disarmed
thioglucoside 1 (RRV = 1855) (entries 2, 3, 5, and 6). The
glycosylations produced the expected disaccharides 19, 20, 21,
and 22. Note that such glycosylations would not have been
successful using the conventional protocol, as witnessed in
glycosylation of thioglucoside 11a with 1 (entry 4). In addition
to 1, the modulation procedure was applied to thioglucoside 6,
which contained different protecting functions (entries 7 and 8).
After preparation of b-glucosides, the modulated glycosyla-
tion method is utilized for synthesis of a-mannosides. For
example, armed thiomannosides 12 (RRV = 12 500), 13a
(RRV = 6160), 13b (RRV = 4644), and 14 (RRV = 8400)
were coupled with disarmed thiomannoside 2 (RRV = 1912)
affording expected disaccharides 25–28 exclusively in 60% to
73% yields (entries 9–12). For the glycosylation of 13b,
silylation of the C3 hydroxyl group occurred in the presence
of TMSOTf; thus triflic acid (TfOH) was used as an acid
promoter (entry 11).
Fig. 1 (a) Thioglycoside donors 1, 2, 6, 7, and 8. (b) Glycoside
acceptors 9–17. (c) Glycosylation products 18–31.
To demonstrate the utility of our method in oligosaccharide
synthesis, trisaccharides including a-(1,2)-manan trisaccharide
35, a-manan 36, and b-(1,6)-glucan 37 were prepared from
building blocks (2, 12, 32), (2, 33, 32), and (6, 11b, 34)
respectively, by iterative one-pot disarmed–armed glycosylation
(Scheme 3).
After the one-pot synthesis of the trisaccharides 35–37, we
investigated the reaction mechanism of the modulated glyco-
sylation (Scheme 4). Selected 2-O-benzoyl thioglucoside 38
was activated in the presence of DMF in CDCl3 at À20 1C
(Scheme 4a). Upon completion of the activation, a fraction of
the mixture was taken for NMR analysis, while the rest was
treated with acceptor 39. Referring to the NMR spectra of the
pre-activation mixture, signals corresponding to glucosyl imidinium
triflates 40a, 40b, and N-b-glucosyl succinimide 41 in a ratio of
1 : 7 : 1 were detected (Scheme 4b and Fig. S2 in ESIw). The
configuration of 40b is supported by the chemical shift of the
To assess the scope of application of our method, 2-amino-
2-deoxy thioglucoside 7 (RRV = 57)15 and 2-O-benzoyl
thiogalactoside 8 (RRV = 5200)18 were investigated for
coupling with thioglycosides 11a, 15, and 16. Gratifyingly,
the glycosylations furnished the desired disaccharides 29–31
(entries 13–15). However, the glycosylation with per-O-acetyl
thiogalactoside did not bear fruit due to the acyl transfer
reaction.19 Except for disaccharide 30, the yields of the present
glycosylations fall between 55% and 73%. We reasoned that
some un-reactive glycosyl intermediates might be formed during
the pre-activation, which decreased the yield of reaction.20
3
anomeric proton at 5.79 ppm and a corresponding JH1/H1
coupling constant of 7.8 Hz.21 Additional evidence comes from
the chemical shift of the anomeric carbon at 102 ppm and a
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 10910–10912 10911

View Article Online
Table 2 Disarmed–armed glycosylations under DMF modulation
ion by NIS–TMSOTf, which under NGP is converted to the
dioxalenium ion.7–9 As the majority of dioxalenium ions react
with DMF to form b-glycosyl imidinium triflate, remaining
dioxalenium ions escape the NMR detection.22 However, their
presence would be inferred from the reaction with an acceptor.
Upon addition of an acceptor 39, the signals of the b-glucoside
42 appeared while the signals of 40 and 41 vanished (see
Fig. S5 in ESIw). An explanation is that the b-imidinium
triflate 40 and b-succinimide 41 are in equilibrium with the
dioxalenium ion; when added the acceptor 39 is selectively
coupled with the more-reactive dioxalenium ion affording the
b-glycoside 42. Such a mechanism appears to be in line with
the Curtin–Hammet principle.23
Product
Donor,
Entry RRV
Acceptor,
RRV
Ta
Timeb
Yield
(%)
(1C) (h)
No.
1
2
3
4
5
6
7
8
1, 1855
1, 1855
1, 1855
1, 1855
1, 1855
1, 1855
6, 531
9, —
À10 13.5
À10 11.0
18
19
20
NDc
21
22
23
24
25
26
27
28
29
30
31
65
67
70
—
70
55
66
63
61
60
73d
62
64
40e
57
10, 4470
11a, 4600
11a, 4600
16, 50 000
17, 11 000
11a, 4600
11b, 1200
12, 12 500
13a, 6160
13b, 4644
14, 8400
11a, 4600
15, 187
À10
4.5
À10 18
À10
À10
À10
10
2.5
2
3
6, 531
3
9
2, 1912
2, 1912
2, 1912
2, 1912
7, 57
0
5
10
11
12
13
14
15
0
0
16
4
We thank the NSC of Taiwan (NSC 100-2627-M-009-008),
and the CIT of NCTU.
0
4
5.5
3.0
3.0
À10
À10
À30
7, 57
8, 5200
Notes and references
16, 50 000
a
b
Temperature used for glycosylations. Time required for glycosyla-
c
tion coupling. The glycosylation was conducted by the conventional
1 T. J. Boltje, T. Buskas and G. J. Boons, Nat. Chem., 2009, 1, 611.
2 C.-H. Hsu, S.-C. Hung, C.-Y. Wu and C.-H. Wong, Angew.
Chem., Int. Ed., 2011, 50, 11872.
3 Y. Wang, X.-S. Ye and L.-H. Zhang, Org. Biomol. Chem., 2007,
5, 2189.
procedure without DMF modulation and ‘ND’ means not being
d
e
detected. TfOH was used as an acid promoter. Modest 40% yield
was due to the poor solubility of 30, which affected the recovery of the
product in purification.
4 S.-R. Lu, Y.-H. Lai, J.-H. Chen, C.-Y. Liu and K.-K. T. Mong,
Angew. Chem., Int. Ed., 2011, 50, 7315.
5 D. B. Werz, R. Ranzinger, S. Herget, A. Adibekian and
P. H. Seeberger, ACS Chem. Biol., 2007, 2, 685.
6 (a) L. Bohe and D. Crich, C. R. Chim., 2011, 14, 3;
´
(b) D. M. Whitfield, Adv. Carbohydr. Chem. Biochem., 2009, 62, 83.
7 B. Capon and S. P. McManus, Neighbouring Group Participation,
Plenum Press, New York, 1976.
8 D. Crich, Z. Dai and S. Gastaldi, J. Org. Chem., 1999, 64, 5224.
9 B. Fraser-Reid, S. Grimme, M. Piacenza, M. Mach and
U. Schlueter, Chem.–Eur. J., 2003, 9, 4687.
10 H. D. Premathilake, L. K. Mydock and A. V. Demchenko, J. Org.
Chem., 2010, 75, 1095.
11 G. H. Veeneman, S. H. van Leeuwen and J. H. van Boom,
Tetrahedron Lett., 1990, 31, 1331.
12 L. K. Mydock and A. V. Demchenko, Org. Lett., 2008, 10, 2103.
13 (a) D. R. Mootoo, P. Konradsson, U. E. Udodong and B. Fraser-
Reid, J. Am. Chem. Soc., 1988, 110, 5583; (b) B. Fraser-Reid and
J. C. Lopez, Top. Curr. Chem., 2011, 309, 1.
14 (a) X. Huang, L. Huang, H. Wang and X.-S. Ye, Angew. Chem.,
Int. Ed., 2004, 43, 5221; (b) S. Yamago, T. Yamada, T. Maruyama
and J.-I. Yoshida, Angew. Chem., Int. Ed., 2004, 43, 2145;
(c) S. J. Hasty, M. A. Kleine and A. V. Demchenko, Angew.
Chem., Int. Ed., 2011, 50, 4197.
Scheme 3 One-pot synthesis of trisaccharides 35, 36, and 37.
15 (a) Z. Zhang, I. R. Ollmann, X.-S. Ye, R. Wischnat, T. Baasov and
C.-H. Wong, J. Am. Chem. Soc., 1999, 121, 734; (b) K. M. Koeller
and C.-H. Wong, Chem. Rev., 2000, 100, 4488.
16 In the method, 1 equiv. of tested and reference thioglycosides were
reacted with 5 equiv. of MeOH. After 2 h of reaction, the reaction was
quenched and the crude mixture was analyzed by HPLC. Peak
integrals corresponding to thioglycosides (before and after the reac-
tion) were obtained for RRV calculation (see pp. 36 and 37 in ESIw).
17 One of the reviewers asked if a C6 OH unprotected donor would
form 1,6-anhydro product in RRV determination. Indeed, the
formation of such product was not observed in our cases.
18 B.-L. Tsai, J.-L. Han, C.-T. Ren, C.-Y. Wu and C.-H. Wong,
Tetrahedron Lett., 2011, 52, 2312.
19 Z. Li and J. C. Gildersleeve, J. Am. Chem. Soc., 2006, 128, 11612.
20 Similar explanation was given for glycosylations with bromine
addition: S. Kaeothip, J. P. Yasomanee and A. V. Demchenko,
J. Org. Chem., 2012, 77, 291.
21 K. Bock and C. Pedersen, J. Chem. Soc., Perkin Trans. 2, 1974,
293.
Scheme 4 (a) NMR studies for modulated glycosylation of 39 with
38. (b) Reaction intermediates 40a, 40b, 41 and product 42.
1JC1/H1 coupling constant of 174 Hz (Table S2 in ESIw). Note
that the 1JC1/H1 values of the anomeric centers for 40a and 40b
are ca. 10 Hz higher than that of 4C1 O-glycoside. To the best of
our knowledge, this set of NMR data represents the first of
its kind. Further confirmation of the imidinium structure is
provided by COSY, HSQC, and HMBC experiments (see
NMR spectra and Fig. S3 in ESIw).
The aforementioned NMR data enable us to propose a
mechanism for the glycosylation method (Fig. S4 in ESIw).
Thioglycoside is activated to form the glycosyl oxacarbenium
22 NMR detection of the xylopyranosyl dioxalenium ion was
reported at À78 1C (see ref. 8).
23 J. I. Seeman, Chem. Rev., 1983, 83, 83.
c
10912 Chem. Commun., 2012, 48, 10910–10912
This journal is The Royal Society of Chemistry 2012