Tin-mediated regioselective acylation of unprotected sugars on solid phase

Article abstract of DOI:10.1016/S0040-4039(00)01492-1

Methyl α-D-glucopyranoside, methyl α-D-mannopyranoside and methyl β-D-galactopyranoside bound on O-6 to a copolystyrene-DVB resin through a trityl ether linker, have been regioselectively acylated with benzoyl chloride after treatment with Bu₂S

Full text of DOI:10.1016/S0040-4039(00)01492-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 8587–8590
Tin-mediated regioselective acylation of unprotected sugars
on solid phase
Francesco Peri,* Laura Cipolla and Francesco Nicotra
Department of Biotechnology and Biosciences University of Milano-Bicocca, Piazza della Scienza, 2,
20126 Milano, Italy
Received 26 August 2000; accepted 6 September 2000
Abstract
Methyl a-
D
-glucopyranoside, methyl a-
D
-mannopyranoside and methyl b-D-galactopyranoside bound
on O-6 to a copolystyrene–DVB resin through a trityl ether linker, have been regioselectively acylated
with benzoyl chloride after treatment with Bu2SnO and stannylene formation on solid phase. The
benzoylation reactions proved to be highly regioselective affording 2-O benzoyl derivatives for glucose
and 3-O benzoyl derivatives for galactose and mannose with high yield. © 2000 Published by Elsevier
Science Ltd.
Tin-mediated esterification, alkylation,1 and other transformations of the hydroxyl groups,
including glycosylation,2,3 have been extensively exploited for the regioselective functionalisation
of sugars in solution. The stannylene derivative of methyl 4,6-O-benzylidene-a- -glucopyra-
D
noside has been characterised by X-ray analysis in the solid state as a dimer composed of two
sugars linked through the Sn2O2 parallelogram4 and the methyl 4,6-O-benzylidene-2,3-O-
dibutylstannylene-a- -mannopyranoside has been crystallised as a pentamer constituted by a
D
repeat of dioxystannacyclopentane units.5 The consequent nucleophilic enhancement of the
oxygen atom in the equatorial position in the oligomer may provide a plausible explanation of
the regioselectivity observed in the subsequent functionalisation of the OH groups.
We present here an extension of this chemistry to the regioselective acylation of solid-phase
bound monosaccharides with unprotected secondary hydroxyl groups. The emerging use of
sugars as chiral multi-dimension-diversity scaffolds for combinatorial chemistry demands con-
vergent and efficient solid-phase chemical methods for the selective functionalisation of the
resin-bound sugar.6,7 The preparation of monosaccharides protected with a set of orthogonal
groups is a laborious but necessary step in obtaining a multifunctional carbohydrate-derived
scaffold. The solid-phase universal pharmacophore-mapping libraries are then generally pre-
* Corresponding author. Tel: +39.02.6448.3453; fax: +39.02.6448.3565; e-mail: francesco.peri@unimib.it
0040-4039/00/$ - see front matter © 2000 Published by Elsevier Science Ltd.
PII: S0040-4039(00)01492-1

8588
pared by covalently linking the sugar to a resin, by deprotecting one of the hydroxyl groups and
by introducing pharmacophores at this position. The regioselective functionalisation of one of
the unprotected hydroxyl groups of a sugar directly on solid phase could be an elegant and
efficient alternative to shorten this procedure.
Methyl a-
D
-gluco-, manno- and galactopyranosides (1, 2 and 4) and methyl b- -galactopyra-
D
noside (3) were anchored to a trityl chloride polystyrene–DVB resin by trityl ether formation at
O-6 (Scheme 1).†
O
Trt
OH
O
O
HO
OH
O
HO
HO
O
O
Bu
O
Sn
Bu
OCH3
HO
HO
OCH3
Ph
OH
O
OCH3
5
1
Bu
Sn
HO
OH
OH
O
O
Trt
OH
Bu
O
OH
O
HO
O
HO
HO
O
O
Ph
OCH3
OCH3
Trt
OCH3
O
1) Cl-Trt-resin, TBAI,
sym-collidine, dry DMF,
72 hrs
6
2
3) BzCl, dry CH2Cl2
OH
OH
OH
HO
OH
O
Bu
Sn
Bu
O
O
O
OCH3
OH
O
OCH3
O
2) Bu2SnO, 4 A m.s.,
dry DMF, 90 °C, 12 hrs
OCH3
4) CH2Cl2, TFA, TIS
(90:5:5)
O
OH
Ph
OH
O
3
7
OH
O
OH
HO
Mixture of 2,3 and 3,4
stannylidenes
Mixture of 2-O and 3-O benzoyl
derivatives
OH
OCH3
4
Scheme 1. Solid-phase stannylene-promoted benzoylation of monosaccharides 1, 2, 3 and 4
To determine the loading, part of the resin (50 mg) was swelled in dry CH2Cl2 and shaken at
rt in a CH2Cl2–TFA–TIS‡ (90:5:5) mixture for 1 h in a plastic syringe with a polypropylene filter
and a Teflon valve in the bottom (the cleavage was repeated twice), the effluents were collected,
concentrated and methyl glycosides were purified by FC on silica gel (AcOEt–MeOH–H2O
70:30:1). A loading of 0.81 mmol/g after purification resulted from the average of the values
obtained for the four monosaccharides.
† Experimental conditions: Trityl chloride polystyrene–DVB resin (Novabiochem, 2 g, 2.5 mmol) was swollen for
30 min in dry CH2Cl2 under argon in a glass flask with a sintered glass filter and a Teflon valve in the bottom. This
was then suspended in dry DMF (20 mL) under argon atmosphere, methyl-glycopyranosides (1.97 g, 10.2 mmol),
TBAI (4.03 g, 10.2 mmol), sym-collidine (2.03 mL, 15.2 mmol) were added and the mixture was shaken in the dark
for 72 h at rt. The resin was then washed with dry CH2Cl2 (20 mL, three times), dry THF (20 mL, three times), dry
CH2Cl2 (20 mL) and dried overnight in a desiccator in vacuo and stored under argon at −25°C.
‡ Abbreviations: FC: flash chromatography; TBAI: tetrabutylammonium iodide; TFA: trifluoroacetic acid; TIS:
tris-isopropylsilane.

8589
For the stannylene formation, the resin (200 mg, 0.16 mmol) was suspended in dry DMF§ (1
,
mL), Bu2SnO (80 mg, 0.32 mmol) and powdered 4 A m.s. (50 mg) were added and the
suspension was heated without shaking in a glass vessel sealed with a Teflon cap at 100°C for
12 h. The resin was separated from the solids by decanting in CH2Cl2, transferred to a plastic
syringe, washed twice in dry CH2Cl2, suspended in dry CH2Cl2 (1 mL), benzoyl chloride (30 mL,
0.24 mmol) was added and the suspension was shaken at rt for 1 h. The resin was then washed
with CH2Cl2, THF, CH2Cl2 and cleaved with CH2Cl2–TFA–TIS (90:5:5). The HPLC and TLC
analyses of the effluents revealed the absence of unreacted sugars and the presence of only one
monobenzylated product in the case of a-gluco-, a-manno- and b-galactopyranosides corres-
ponding, respectively, to compounds 5, 6 and 7. The regioselectivity of the benzoylation
reactions was nearly complete (more than 98% according to HPLC analysis) for the three
sugars. The reaction on methyl a- -galactopyranoside 4 was neither complete (40% of unreacted
D
sugar was recovered after cleavage) nor regioselective (a complex mixture of mono and
dibenzoylated compounds was formed). Pure monobenzoylated sugars 5, 6 and 7 were purified
1
by FC on silica gel (AcOEt–EtOH 9:1) and their regiochemistry was elucidated by H NMR.¶
The benzoylation reactions were accomplished without added base in very high yields (Table 1).
Table 1
Compound
Acylation position
Regioselectivity (%)
Yield (%)a
1
2
3
4
O-2
O-3
O-3
Mixture
\98
\98
\98
94
98
96
60b
a Yields of products purified by FC.
b Yield of the mixture of acylated products.
As a control, the benzoylation reaction was done without added base on methyl a- -glucopy-
D
ranoside bound to trityl resin omitting the stannylene formation step. Unreacted sugar was
recovered after cleavage indicating that acylation does not take place on solid phase without
stannylene activation of the hydroxyl groups. Some practical and theoretical considerations are
derived from these results. There are no examples, to our knowledge, of stannylene formation
on solid phase. The possibility of regioselectively acylating unprotected sugars bound to a resin
paves the way to more convergent solid-phase synthesis of sugar-based libraries avoiding the
extensive use of protecting groups. More generally, all synthetic transformations of sugars on
solid phase involving ester formation from diols can take advantage of this procedure. The yield
and regioselectivity for O-2 observed in the acylation of resin-bound methyl-a- -glucopyra-
D
§ Using dry benzene as solvent the stannylene formation proceeded as well giving similar results to that reported in
the paper.
¶ Mondimensional 1H NMR (300 MHz) in D-MeOH were diagnostic for the determination of the acylation
positions. The protons on the carbon atoms that were benzoylated are downshifted by more than 1 ppm: in
compound 5 (a-glucose), l=4.89 ppm, H-2 (dd, J1–2=3.7 Hz, J2–3=9.8 Hz); in compound 6 (a-mannose), l=5.22
ppm, H-3 (dd, J2–3=3.3 Hz, J3–4=9.9 Hz), in compound 7 (b-galactose), l=4.97 ppm, H-3 (dd, J2–3=10.0 Hz,
J3–4=3.4 Hz).

8590
noside is comparable to that observed in solution for the 4,6-O-benzylidene analogue8,9 and can
be explained with formation of a 2,3-O-dibutylstannylene and consequent O-2 selective activa-
tion. The acylation of the resin-bound methyl a-
D
-mannopyranoside was selective for O-3, as
observed for the methyl 6-O-trityl a-
D
-mannopyranoside in solution10 with 2,3-O-stannylene
formation. The acylation on O-3 of methyl b- -galactopyranoside is comparable in yield and
D
regioselectivity to the same reaction on the 4,6-O-benzylidene analogue in solution11 with
stannylene formation in 3,4 position. The axial orientation of the anomeric methoxy group in
the methyl-a- -galactopyranoside can favour the formation of the stannylidene in the 2,3
D
positions because of additional coordination of tin to the anomeric oxygen. In this case both
3,4-O- and 2,3-O-stannylenes are probably formed, thus explaining the lack of regioselectivity of
the subsequent benzoylation.
It is noteworthy that two resin-bound sugars cannot associate in solid phase forming the
dimeric units observed for stannylenes of monosaccharides in solution. The regioselectivity of
the acylation reactions described here need to be explained with a different rationale than that
based on the nucleophilic enhancement of the apical oxygen in the Sn2O2 parallelogram of the
dimer.4,5 Further theoretical and applied studies are needed to elucidate this aspect of the
stannylene chemistry on solid phase. We are currently developing other regioselective reactions
of unprotected or partially protected sugars on solid phase via stannylene formation to
demonstrate the general utility of this technique.
References
1. The literature on the regioselective manipulation of OH groups via organotin derivatives has been extensively
reviewed by: David, S.; Hanessian, S. Tetrahedron 1985, 41, 643–663.
2. (a) Garegg, P. J.; Maloisel, J.-L.; Oscarson, S. Synthesis 1995, 409–414. (b) Cruzado, C.; Bernabe, M.;
Martin-Lomas, M. Carbohydr. Res. 1990, 203, 296–301.
3. Kaij, E.; Harita, N. Tetrahedron Lett. 2000, 41, 53–56.
4. David, S.; Pascard, C.; Cesario, M. Nouveau J. Chim. 1979, 3, 63.
5. Holzapfel, C. W.; Kokemoer, J. M.; Marais, C. F.; Kruger, G. J.; Pretorius, J. A. S. Afr. J. Chem. 1982, 35, 80.
6. Sofia, M. J.; Hunter, R.; Chan, T. Y.; Vaughan, A.; Dulina, R.; Wang, H.; Grange, D. J. Org. Chem. 1998, 63,
2802–2803.
7. Wunberg, T.; Kallus, C.; Opatz, T.; Schmidt, W.; Henke, S.; Kunz, H. Angew. Chem. 1998, 110, 2620–2622.
8. Munavu, R.; Szmant, H. H. J. Org. Chem. 1976, 41, 1832.
9. Bredenkamp, M. W.; Spies, H. S. C. Tetrahedron Lett. 2000, 41, 543–546.
10. Varma, A. J.; Schuerch, C. J. Org. Chem. 1981, 46, 799.
11. David, S.; Thieffry, A. J. Chem. Soc., Perkin Trans. 1 1979, 1568.
.