Expedient synthesis of aminooxylated-carbohydrates for chemoselective access of glycoconjugates

Article abstract of DOI:10.1016/S0040-4039(01)01614-8

Herein, we describe an efficient preparation of various biologically important carbohydrate motifs bearing an aminooxy group at the anomeric position. These nucleophilic sugar analogues represent useful intermediates for the chemoselective preparation of

Full text of DOI:10.1016/S0040-4039(01)01614-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7575–7578
Expedient synthesis of aminooxylated-carbohydrates for
chemoselective access of glycoconjugates
Olivier Renaudet and Pascal Dumy*
LEDSS, UMR-CNRS 5616, Universite´ Joseph Fourier, F-38041 Grenoble Cedex 9, France
Received 29 August 2001; accepted 30 August 2001
Abstract—Herein, we describe an efficient preparation of various biologically important carbohydrate motifs bearing an aminooxy
group at the anomeric position. These nucleophilic sugar analogues represent useful intermediates for the chemoselective
preparation of glycoconjugates. The key glycosylation step involves the coupling of fluoro-activated protected sugar and
N-hydroxyphthalimide in the presence of BF3·Et2O. Final deprotection and cleavage of the phthalimide moiety with methyl-
hydrazine afforded new Glc-b-ONH2 3, GalNAc-b-ONH2 9, Glc-a-ONH2 14, Gal-a-ONH2 17 and Man-a-ONH2 20 derivatives
with good yields. Compared to the literature results, the preparation of Gal-b-ONH2 6, GalNAc-a-ONH2 11 and Lac-b-ONH2
23 proved to be more efficient. © 2001 Elsevier Science Ltd. All rights reserved.
Glycoproteins play important roles in biological pro-
cesses, such as cell adhesion, cell differentiation or cell
growth, wherein carbohydrates represent key recogni-
tion elements.1 In order to study and elucidate the
biochemical and physiopathological involvement of
such constructions, extensive investigations are in pro-
gress to access synthetic glycopeptides. However, due to
the multifunctionalities of peptides or sugars, the
difficulty in forming the O-glycosidic linkage consti-
tutes a major problem for the classical chemical
approach.2 Recently, a number of examples have
demonstrated the efficiency of chemoselective ligation
as an alternative strategy for building complex glyco-
conjugates.3 Particularly, oxime bond formation consti-
tutes an attractive route for conjugation of
glycopeptides.4 In this context, carbohydrate motifs
glycosylated with an aminooxy function are of great
interest and some of these have been described and
used successfully.4b–d Most notably, these aminooxy
analogues are also tolerated by some glycosyltrans-
ferases and can be used as substrates for the synthesis
of more complex oligosaccharides through enzymatic
reaction. Herein, we report a general and efficient
method for the preparation of these sugar analogues
based on the coupling of glycosylfluoride and N-
hydroxyphthalimide (Fig. 1) in the presence of
BF3·Et2O followed by adapted deprotection steps. This
represents a synthetic improvement to known com-
pounds (e.g. Gal-b-ONH2 6,5 GalNAc-a-ONH2 114b–d
and Lac-b-ONH2 234b–d). It also provides ready access
to new derivatives, such as Glc-b-ONH2 3, GalNAc-b-
ONH2 9, Glc-a-ONH2 14, Gal-a-ONH2 17 and Man-a-
ONH2 20. Altogether, these compounds are of great
importance for installing carbohydrate-recognition
motifs in a user-defined position within any scaffold
molecule through oxime bond formation, thus provid-
ing ideal chemical access to complex glycoconjugates
without protecting-group manipulation or coupling
reagents.
To date, only few methods using N-hydroxysuccinimide
(HOSuc) or N-hydroxyphthalimide (HOPht) as the
nucleophilic reagent in glycosylation reactions have
been described in the literature.6 Recently, direct incor-
* Corresponding author. Tel.: +4-76 63 5545; fax: +4-76 51 4382;
e-mail: pascal.dumy@ujf-grenoble.fr
Figure 1. Structures of the various a- and b-aminooxy-glyco-
sylated sugar analogues prepared.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01614-8

7576
O. Renaudet, P. Dumy / Tetrahedron Letters 42 (2001) 7575–7578
poration of N-hydroxysuccinimide has been described
by Roy5 starting from penta-O-acetylated gluco- and
galactobromides 1 and 4 using phase-transfer catalysis
(PTC). This glycosylation occurs at room temperature
using 4 equiv. of HOSuc and 1 equiv. of tetra-
butylammonium hydrogen sulphate as catalyst in a
CH2Cl2/Na2CO3 1 M biphasic system and was found to
be stereoselective for b-gluco- and b-galactosyloxysuc-
cinimide due to the presence of an anchimeric acetate
participating group on C-2. This method was adopted
by Bertozzi to prepare compound 114d bearing an a-
linkage starting from 2-azido-galactosylchloride.7
In most cases, the O-Glycosyl-Suc are converted into
glycosylhydroxamic esters by ring opening providing a
linking arm suitable for glycoconjugate preparation.5,6a
More occasionally, the highly efficient conversion of
-OSuc or -OPht moieties into the corresponding free
O-substituted hydroxylamines with hydrazine8 could be
used in the preparation of acceptors for chemoselective
oxime bond formation.4,9 In comparison with Roy’s
PTC method, we have found that the use of the more
stable glycosylfluoride with a non-participating azide
(Table 1, entry 3) or benzyl protecting groups (Table 1,
entries 4–6) on C-2 afford both a- and b-linkages very
Table 1. Reaction conditions for the preparation of a- and b-aminooxy linkages

O. Renaudet, P. Dumy / Tetrahedron Letters 42 (2001) 7575–7578
7577
The final acetate deprotection and cleavage of the
phthalimide group was realised using a large excess of
methylhydrazine in ethanol at room temperature16 for
compounds 2, 5 and 22. Pure and ninhydrine-active free
aminooxy derivatives 3, 6 and 23 were obtained as white
powders after purification and lyophilisation. It is note-
worthy that previous attempts to access mannose deriva-
tive 20 using acetate protection failed completely
probably because of an elimination reaction occurring
during the final deprotection. On the other hand, depro-
tection of benzyl groups for compounds 13, 16 and 19
by palladium hydrogenation in methanol and reductive
acetylation of the azide moiety for derivatives 8 and 10
by palladium hydrogenation in methanol/acetic anhy-
dride at room temperature is more tricky due to the
fragility of the NꢀO bond under reductive conditions.
According to NMR and X-ray structural analysis (not
shown15), we have demonstrated, in the case of com-
pound 10, that reduced and acetylated compounds at the
anomeric position are mainly obtained after several
hours under these conditions. Consequently, the evolu-
tion of the reaction of benzyl deprotection or azide
reduction should be carefully controlled with a short
reaction time to avoid any risk of NꢀO bond breaking.17
Further, phthalimide cleavage was performed following
the procedure described for 2, 5 and 22 to give com-
pounds 14, 17, 11, 9 and 20. All these compounds, and
previously unknown derivatives 3, 14, 17, 9 and 20, were
fully characterised by NMR and MS analysis.18
efficiently with BF3·Et2O as a promoter. On the other
hand, our experiments have shown that N-hydroxyph-
thalimide is much more reactive for glycosylation and
more easily cleaved than the N-hydroxysuccinimide
moiety. In addition, we demonstrated the efficiency of
our method through several examples, such as the Glc,
Gal, GalNAc and Man series, as well as for functional-
isation of more complex oligosaccharides, such as lac-
tose (Table 1, entry 7).
In entries 1 and 2 of Table 1, we report the stereoselec-
tive b-glycosylation with N-hydroxyphthalimide from
per-O-acetate bromide derivatives 1 and 4 using PTC in
very good yield (71–80%). However, for compound 10,
this method proved to be non-reproducible in our hands
and also instability observed for 3,4,6-tri-O-acetyl-2-
azido-2-deoxy-a-D-galactosylchloride during the PTC-
coupling reaction led to a much lower yield (22%, not
shown). Therefore, we were interested in fluoride activa-
tion of the anomeric position as a more direct and
general approach.
Glycosylfluorides 7, 12, 15, 18 and 21 were prepared
very easily using diethylaminosulfur trifluoride (DAST)
as a fluorine donor in THF.10 Convenient denitration
(Table 1, entry 3) of 3,4,6-tri-O-acetyl-2-azido-2-de-
oxy-a-D-galactopyranosyl nitrate was achieved with
sodium nitrite in aqueous dioxane at 80°C.11 Selective
deacetylation of per-O-acetyl lactose (Table 1, entry 7)
at the anomeric centre was realised using an ethylene-
diamine/acetic acid mixture in THF at room tempera-
ture.12
In conclusion, we have described a convenient prepara-
tion of various aminooxy sugars analogues 3, 9, 14, 17,
20 and 23 using glycosylfluoride as a donor and
BF3·Et2O as a promoter. a- or b-Analogues of the
corresponding sugar are obtained, which is important
for assessing the influence of the anomeric configuration
in biological processes. Compared to the literature,4b–d,5
compounds 6 and 11 were obtained more efficiently and
easily using N-hydroxyphthalimide. A further advan-
tage of this approach that it provides carbohydrate-
based compounds readily accessible to a broader
community for the construction of various glycoconju-
gates. These compounds represent the structural base
for most of the carbohydrate-recognition motifs found
in glycoconjugates. Therefore, their manipulation for
chemoselective preparation of the corresponding glyco-
conjugates should provide useful tools for glycobiology,
as well as new therapeutics. Work in this direction is
currently underway in our laboratory and will be
reported in the near future.
Glycosylfluorides represent very reactive donors in gly-
cosylation reactions with HOPht using BF3·Et2O13 as a
promoter. Typically, stoichiometric amounts of N-
hydroxyphthalimide and triethylamine were reacted in
the presence of glycosylfluoride and the promoter (5
equiv.) in CH2Cl2 at room temperature.14 The reaction
is usually fast and complete within 15–30 min (Table 1).
After classical work-up, separation of the a and b
anomers (Table 1, entries 3–5) is easily performed by
silica-gel chromatography.
First, we have obtained the two anomers 8 and 10 (50
and 38%, respectively) from fluoride compound 7. The
a stereochemistry of compound 10 was confirmed by
crystallographic structure.15 Starting from compounds
12 and 15 (Table 1, entries 4 and 5), we were able to
prepare 90% of a/b anomer mixture (ratio 2/1) to give
gluco- and galacto-derivatives, 13 and 16, respectively.
Moreover, the glycosylation of tetrabenzylmanno-
fluoride derivative 18 (Table 1, entry 6) by the same
method stereoselectively afforded the mannosyl-a-N-
oxyphthalimide 19 with high yield (75%). Finally, we
have also demonstrated that this method is suitable for
the easy transformation of more complex oligosaccha-
rides. In Table 1, entry 7, we have shown the stereo-
selective b-incorporation of PhtOH starting from lactose
fluoride derivative 21 with good yield (70%) according
to the same procedure.
Acknowledgements
This work was supported by the Association pour la
Recherche contre le Cancer (ARC), the Centre National
pour la Recherche Scientifique (CNRS) for specific
action (AIP) and the Institut Universitaire de France
(IUF). We also acknowledge the MENRT for donating
grant no. 98-4-23548 to O.R. We also thank C. Philouze
and M. T. Averbuch-Pouchot for performing the X-ray
studies.

7578
O. Renaudet, P. Dumy / Tetrahedron Letters 42 (2001) 7575–7578
16. The use of methylhydrazine instead of hydrazine was found
References
to be more efficient and facilitated the subsequent purifica-
tion of the final compounds.
1. Dwek, R. A. Chem. Rev. 1996, 96, 683–720.
2. (a) Seitz, O. ChemBioChem 2000, 1, 214–246; (b) Taylor,
C. M. Tetrahedron 1998, 54, 11317–11362.
3. Macmillan, D.; Bertozzi, C. R. Tetrahedron 2000, 56,
9515–9525 and references cited therein.
4. (a) Cervini, S. E.; Dumy, P.; Mutter, M. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 1230–1232; (b) Rodriguez, E. C.;
Marcaurelle, L. A.; Bertozzi, C. R. J. Org. Chem. 1998, 63,
7134–7135; (c) Marcaurelle, L. A.; Rodriguez, E. C.;
Bertozzi, C. R. Tetrahedron Lett. 1998, 39, 8417–8420; (d)
Rodriguez, E. C.; Winans, K. A.; King, D. S.; Bertozzi, C.
R. J. Am. Chem. Soc. 1997, 119, 9905–9906.
5. Cao, S.; Tropper, F. D.; Roy, R. Tetrahedron 1995, 51,
6679–6686.
6. (a) Andersson, M.; Oscarson, S. Glycoconjugate J. 1992, 9,
122–125; (b) Grochowski, E.; Jurczak, J. Carbohydr. Res.
1976, 50, C15–C16.
7. Lemieux, R. U.; Ratcliffe, R. M. Can. J. Chem. 1979, 57,
1244–1251.
8. Grochowski, E.; Jurczak, J. Synthesis 1976, 682–684.
9. (a) Nicolaou, K. C.; Groneberg, R. D. J. Am. Chem. Soc.
1990, 112, 4085–4086; (b) Yang, D.; Kim, S.-H.; Khane,
D. J. Am. Chem. Soc. 1991, 113, 4715–4716.
17. A mixture of compound 19 (2.23 g, 3.25 mmol) and Pd/C
10% (0.14 g, 1.3 mmol) in methanol (20 mL) was stirred
at room temperature under an atmosphere of hydrogen for
2 h. The catalyst was removed by filtration and the solvent
was evaporated in vacuo. The residue was taken up with
water, the aqueous layer was extracted several times with
ethyl acetate and it was then lyophilised. The residual white
powder was diluted in an ethanol/methylhydrazine (1/1)
mixture (10 mL) and the solution was stirred overnight at
room temperature. Chromatographic purification
(CH2Cl2/EtOH 10/1“1/1) afforded pure compound 20
(270 mg, 44%).
18. All these compounds were identified by one- and two-
dimensional NMR techniques (1H, 13C, GCOSY,
GHMQC) as well as mass spectrometry (DCI NH3/isobu-
tane). Compound 3: 1H NMR (300 MHz, D2O) l 4.59 (d,
1H, 3J1,2=8.3 Hz, H-1), 3.96 (dd, 1H, 3J5,6a=2.3 Hz,
2J6a,6b=12.2 Hz, H-6a), 3.76 (dd, 1H, 3J5,6b=5.7 Hz, H-6b),
3.57–3.30 (m, 4H, H-2, H-3, H-4, H-5); 13C NMR (75 MHz,
D2O) l 105.4 (C-1), 76.2, 76.1 (C-3, C-4), 72.0 (C-2), 69.9
(C-5), 61.1 (C-6); Mcalcd=195.2; m/z=196.0 [M+H]+, 212.9
[M+NH4]+, 180.1 [M−ONH2]+, 147.2 [M−OH−ONH2]+,
132.1 [M−2OH−ONH2]+, 115.2 [M−3OH−ONH2]+. Com-
pound 6: 1H NMR (300 MHz, D2O) l 4.54 (d, 1H, 3J1,2=8.1
Hz, H-1), 3.94 (bd, 1H, 3J3,4=3.4 Hz, H-4), 3.87–3.73 (m,
10. (a) Rosenbrook, Jr., Wm.; Riley, D. A.; Lartey, P. A.
Tetrahedron Lett. 1985, 26, 3–4; (b) Posner, G. J.; Haines,
S. R. Tetrahedron Lett. 1985, 26, 5–8.
11. Grundler, G.; Schmidt, R. R. Justus Liebigs Ann. Chem.
1984, 00, 1826–1847.
3
3H, H-5, H-6a, H-6b), 3.69 (dd, 1H, J2,3=9.8 Hz, H-3),
3.55 (dd, 1H, H-2); 13C NMR (75 MHz, D2O) l 105.9 (C-1),
12. Zhang, J.;Kovac, P. J. Carbohydr. Chem. 1999, 18, 461–469.
13. Kunz, H.; Sager, W. Helv. Chim. Acta 1985, 68, 283–287.
14. A typical procedure follows for mannose analogue 19: To
a stirring solution of compound 18 (1.27 g, 2.33 mmol),
N-hydroxyphthalimide (0.38 g, 2.33 mmol) and triethyl-
amine (325 mL, 2.33 mmol) in CH2Cl2 (20 mL) was added
BF3·Et2O (1.47 mL, 11.6 mmol) and the reaction was stirred
at room temperature for 15 min. After classical work-up,
the residue was purified by silica-gel chromatography
(AcOEt/Hexane, 3/7) to give pure compound 19 (1.20 g,
75.5 (C-3), 73.2 (C-2), 69.7 (C-4), 69.0 (C-5), 61.4 (C-6);
M
calcd=195.2;m/z=196.0[M+H]+, 212.9[M+NH4]+, 180.1
[M−ONH2]+, 147.2 [M−OH−ONH2]+, 132.1 [M−2OH−
ONH2]+, 115.2 [M−3OH−ONH2]+. Compound 9: 1H NMR
(300 MHz, D2O) l 4.57 (d, 1H, 3J1,2=8.7 Hz, H-1),
3.98–3.89 (m, 2H, H-4, H-2), 3.86–3.70 (m, 4H, H-3, H-5,
H-6a, H-6b), 2.07 (s, 3H, HNCOCH3); 13C NMR (75 MHz,
D2O) l 175.3 (HNCOCH3), 104.5 (C-1), 75.5, 71.4 (C-3,
C-5), 68.2 (C-4), 61.4 (C-6), 51.1 (C-2), 22.5 (HNCOCH3);
M
calcd=357.3; m/z=237.0 [M+H]+, 204.1 [M−ONH2]+.
1
75%). H NMR (300 MHz, CDCl3): l ppm 7.74–7.63 (m,
1
Compound 14: H NMR (300 MHz, D2O) l 5.06 (d, 1H,
4H, H(ar.)Pht), 7.37–7.10 (m, 20H, H(ar.)Bn), 5.48 (d, 1H,
3J1,2=1.8 Hz, H-1), 4.81–4.33 (m, 8H, 4CH2), 4.50 (m, 1H,
H-5), 4.13–4.03 (m, 2H, H-2, H-4), 3.91 (dd, 1H, 3J2,3=3.2
Hz, 3J3,4=8.9 Hz, H-3), 3.81 (dd, 1H, 3J5,6a=3.7 Hz,
2J6a,6b=11.2 Hz, H-6a), 3.61 (dd, 1H, 3J5,6b=1.9 Hz, H-6b);
13C NMR (75 MHz, CDCl3): l ppm 163.6 (CꢁO), 138.9
(C(ar.)Bn), 138.8 (C(ar.)Bn), 138.7 (C(ar.)Bn), 138.2 (C(ar.)Bn),
134.9 (CH(ar.)Pht), 128.8 (C(ar.)Pht), 128.7 (CH(ar.)Bn), 128.7
(CH(ar.)Bn), 128.6 (CH(ar.)Bn), 128.5 (CH(ar.)Bn), 128.3
(CH(ar.)Bn), 128.2 (CH(ar.)bn), 128.0 (CH(ar.)Bn), 128.0
(CH(ar.)Bn), 127.9 (CH(ar.)Bn), 127.8 (CH(ar.)Bn), 123.9
(CH(ar.)Pht), 104.1(C-1), 79.7 (C-3), 75.3 (CH2), 74.6, 74.1,
73.6 (CH2), 73.5, 73.3 (CH2), 72.7 (CH2), 69.0 (C-6).
15. Crystallographic data (excluding structural factors) for the
structures in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary
publication numbers CCDC 167839 and 167794. Copies of
the data can be obtained, free of charge, on application to:
The Director, CCDC, 12 Union Road, Cambridge CB2
1EZ, UK [fax: +44(0)-1223-336033 or e-mail: deposit@
ccdc.cam.ac.uk]
3J1,2=3.3 Hz, H-1), 3.94 (dd, 1H, J5,6a=1.9 Hz, J6a,6b
=
3
2
12.2 Hz, H-6a), 3.85 (dd, 1H, 3J5,6b=4.7 Hz, H-6b),
3.80–3.77 (m, 1H, H-5), 3.72–3.62 (m, 2H, H-2, H-3), 3.48
(t, 1H, 3J3,4=3J4,5=9.0 Hz, H-4); 13C NMR (75 MHz, D2O)
l 102.0 (C-1), 73.2, 72.1, 71.2, 69.8 (C-2, C-3, C-4, C-5),
60.7 (C-6); Mcalcd=195.2; m/z=196.0 [M+H]+, 212.9 [M+
1
NH4]+, 180.1 [M−ONH2]+. Compound 17: H NMR (300
3
MHz, D2O) l 5.11 (d, 1H, J1,2=3.9 Hz, H-1), 4.07–4.00
3
(m, 2H, H-6a, H-6b), 3.93 (dd, 1H, J2,3=10.4 Hz, H-2),
3.89–3.78(m, 3H, H-3, H-4, H-5);13CNMR(75MHz, D2O)
l 102.0 (C-1), 71.4, 69.7, 69.6, 68.2 (C-2, C-3, C-4, C-5),
61.5 (C-6); Mcalcd=195.2; m/z=196.1 [M+H]+, 212.9 [M+
1
NH4]+. Compound 20: H NMR (300 MHz, D2O) l 5.00
3
3
(d, 1H, J1,2=1.7 Hz, H-1), 4.03 (dd, 1H, J2,3=2.8 Hz,
H-2), 3.97 (dd, 1H, 3J5,6a=1.3 Hz, 2J6a,6b=12.3 Hz, H-6a),
3.88 (dd, 3J5,6b=3.7 Hz, H-6b), 3.76–3.72 (m, 3H, H-3, H-4,
H-5); 13C NMR (75 MHz, D2O) l 103.8 (C-1), 73.4, 71.2,
69.4 (C-2), 67.2, 61.4 (C-6); Mcalcd=195.2; m/z=196.1
[M+H]+, 212.8 [M+NH4]+. For NMR data of 11 and 23,
see Ref. 4d.