An efficient method for the assembly of sulfated oligosaccharides using reductive amination

Article abstract of DOI:10.1016/S0040-4039(00)02277-2

A generally applicable efficient method was developed for assembly of structurally defined sulfated oligosaccharides. Using a reductive amination reaction at low pH with newly designed linkers possessing multiple aromatic amino groups, facile preparations of oligosaccharide assemblies were achieved without any alteration in the sulfated oligosaccharide parts.

Full text of DOI:10.1016/S0040-4039(00)02277-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 1293–1296
An efficient method for the assembly of sulfated oligosaccharides
using reductive amination
Shuhei Koshida,a Yasuo Suda,a,* Akio Arano,a Michael Sobelb and Shoichi Kusumotoa
aDepartment of Chemistry, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan
bDepartment of Surgery, Syracuse Veterans Administration and Upstate Medical University, State University of New York,
Syracuse, NY 13210, USA
Received 12 October 2000; revised 24 November 2000; accepted 8 December 2000
Abstract—A generally applicable efficient method was developed for assembly of structurally defined sulfated oligosaccharides.
Using a reductive amination reaction at low pH with newly designed linkers possessing multiple aromatic amino groups, facile
preparations of oligosaccharide assemblies were achieved without any alteration in the sulfated oligosaccharide parts. © 2001
Elsevier Science Ltd. All rights reserved.
Glycosaminoglycans (GAGs) have a number of biologi-
cal functions that can be attributed to specific partial
structures in the GAG chain. The weak interactions of
individual subunit structures are often enhanced by
their specific repetition, which is called a clustering or
multi-valency effect.1–3 For an in-depth study of the
role of specific subunits, the preparation of clustered
compounds is useful to obtain high-affinity ligands.
Reductive amination is one efficient method for the
coupling of oligosaccharides. In the past, proteins were
used as carriers for oligosaccharides. But coupling
occurs at unidentified amino groups in the protein, and
that method gives a mixture of heterogeneous products
having different numbers of carbohydrate units.4–6 For
our studies of the structure-function relations between
heparin and vascular cells and proteins,7,8 it is crucial to
use structurally defined ligands possessing specific sub-
unit structures in heparin.
We designed structurally defined carriers with aromatic
amino groups as the point of linkage to carbohydrates.
To establish the reaction conditions for the reductive
amination, a synthetic sulfated disaccharide 19 was
employed as
a
model compound and sodium
cyanoborohydride (NaBH3CN) as a reducing reagent.
The reagent has been widely used in the reductive
amination with remarkable efficiency.10,11 In the first
attempt using a large excess (10 equiv.) of ethanolamine
(Scheme 1, R=CH2CH2OH), the reaction did not pro-
ceed at neutral or acidic conditions. At higher pH
around 9–10, the desired reductive amination did occur
to give the coupling product 2, but the yield was not
Scheme 1.
* Corresponding author.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)02277-2

1294
S. Koshida et al. / Tetrahedron Letters 42 (2001) 1293–1296
satisfactory due to the competitive reduction of the
anomeric position, giving the glucitol derivative 3 as a
major by-product. To favor the reductive amination, an
amine derivative capable of forming a Schiff base at
neutral or acidic pH range was desirable. The reaction
of 1 (Scheme 1, R=Ph) with aniline (pKa=4.65)
instead of ethanolamine (pKa=9.52) was performed
using NaBH3CN. This coupling proceeded smoothly at
pH 7 to yield the desired product 4 without the forma-
tion of the by-product 3.
connection of m-phenylenediamine units using an
amido spacer was next attempted. Based on this idea,
linkers (7 and 8) having free amino groups derived from
phenylenediamine moieties at the ends were prepared in
one or a few steps without any protection, as illustrated
in Scheme 3. A variety of more complex linkers can
also be prepared using the appropriate multivalent car-
boxylic acids.
The coupling reaction of 1 with the divalent linker 7
(0.4 equiv. to 1) was performed at pH 3 in water–
AcOH at 37°C, and the reaction mixture was then
purified by gel permeation chromatography (Sephadex
G-50) to give the dimeric assembly 9 with a 59% yield
Then, the preparation of a multimeric oligosaccharides
assembly by reductive amination was attempted using a
divalent aromatic amine (Scheme 2). Since NaBH3CN
can be used even at a pH range as low as 3,10,11 the
coupling reaction was tested at pH 3 in water–AcOH at
37°C.12 In this case, the mono-coupled 5 was obtained
in much higher yield than that at pH 7 as judged from
ESI–MS, but the desired dimeric-product 6 was still not
encountered at all. Reaction at a higher temperature
(80°C) did not improve the results.
1
(Scheme 4)13 as confirmed by H NMR and ESI–MS.14
With the trivalent linker 8, the reaction at all the three
amino groups proceeded equally well to give the
trimeric assembly 10.14 No loss of sulfate groups was
observed in either case. In this manner, an easy and
direct assembly of sulfated saccharides by reductive
amination was made possible by using these newly
designed linker compounds. This was archived under
acidic conditions, without undesirable direct reduction
of the reducing ends of the carbohydrate. Using the
same linkers and reaction conditions, other sulfated
disaccahrides, which were derived from a natural sul-
fated polysaccharide heparin, were also assembled.15
Although one unit of 1 was coupled with m-phenylene-
diamine almost quantitatively at pH 3, the dimeric
assembly of the sulfated saccharide failed. This seems to
be caused by the steric hindrance and/or the electric
repulsion of the sulfated saccharide units. Therefore, a
Scheme 2.
Scheme 3.

S. Koshida et al. / Tetrahedron Letters 42 (2001) 1293–1296
1295
Scheme 4.
Furthermore, neutral oligosaccahrides, such as lactose
or maltose, were easy to be assembled by a similar
reaction conditions (data not shown). Therefore, this
method may be applicable not only to sulfated saccha-
rides but also to any oligosaccharides from natural or
synthetic sources, provided that they have reducing
ends.
9. Sulfated disaccharide 1 was synthesized from the interme-
diate in the accompanying manuscript (see Ref. 15)
10. Lane, C. F. Synthesis 1975, 135–146.
11. Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am.
Chem. Soc. 1971, 93, 2897–2904.
12. No coupling reaction with m-phenylenediamine (0.4
equiv. to 1) occurred, however, the most of starting
material was recovered under the same conditions applied
for aniline (pH 7, at room temperature). The coupling of
1 with m-phenylenediamine proceeded at 37°C, but only
one of the amino groups of the diamine reacted to give a
small amount of 5: no desired dimeric assembly 6 was
obtained and most of 1 recovered.
13. Linker 7 (0.92 mg, 3.08 mmol) and 1 (5.0 mg, 7.71 mmol)
were dissolved in a mixture of water (0.9 mL) and acetic
acid (0.1 mL). To the solution was added sodium
cyanoborohydride (4.80 mg, 76.4 mmol), and the solution
was heated at 37°C in a sealed tube. After heating for 24
h, sodium cyanoborohydride (4.80 mg, 76.4 mmol) was
added, and the solution was heated again at 37°C in a
sealed tube. After heating for further 24 h, the reaction
solution was concentrated in vacuo. The residue was
dissolved in water and NaHCO3 was added, and chro-
matographed on Sephadex G-50 (1.7 x 90 cm, water) to
give 9 as a white powder (2.84 mg, 58.9% from 7, 47.1%
from 1).
Acknowledgements
This study was supported in part by ‘Research for the
Future’ Program No. 97L00502 from the Japan Society
for the Promotion of Science.
References
1. Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995, 28, 321–327.
2. Weis, W. I.; Drickamer, K. Annu. Rev. Biochem. 1996,
65, 441–473.
3. Dimick, S. M.; Powell, S. C.; McMahon, S. A.;
Moothoo, D. N.; Naismith, J. H.; Toone, E. J. J. Am.
Chem. Soc. 1999, 121, 10286–10296.
4. Schwartz, B. A.; Gray, G. R. Arch. Biochem. Biophys.
14. Spectral data for compound 9: 1H NMR (500 MHz,
D2O), l 7.17 (2H, t, J=8.0 Hz), 6.83 (2H, s), 6.78 (2H, d,
J=7.6 Hz), 6.61 (2H, d, J=7.6 Hz), 4.96 (2H, s), 4.43
(2H, dd, J=2.3 Hz, J=8.8 Hz), 4.25–4.16 (8H, m), 4.12
(2H, dd, J=8.9 Hz, J=11.8 Hz), 3.90 (2H, m), 3.84 (2H,
m), 3.74 (2H, m), 3.70 (2H, dd, J=2.4 Hz, J=7.6 Hz),
3.65–3.58 (2H, m), 3.57 (2H, dd, J=6.1 Hz, J=10.7 Hz),
3.31 (2H, b), 3.08 (2H, b), 2.68 (4H, s); ESI–MS (nega-
tive) m/z 714.10 [(M−6Na+4H)2−]. Spectral data for com-
1977, 181, 542–549.
5. Laferriere, C. A.; Roy, R. Methods Enzymol. 1994, 242,
102–108.
6. Hashimoto, Y.; Suzuki, M.; Crocker, P. R.; Suzuki, A. J.
Biochem. 1998, 123, 468–478.
7. Suda, Y.; Marques, D.; Kermode, J. C.; Kusumoto, S.;
Sobel, M. Throm. Res. 1993, 69, 501–508.
8. Koshida, S.; Suda, Y.; Fukui, Y.; Ormsby, J.; Sobel, M.;
Kusumoto, S. Tetrahedron Lett. 1999, 40, 5725–5728.

1296
S. Koshida et al. / Tetrahedron Letters 42 (2001) 1293–1296
pound 10: 1H NMR (500 MHz, D2O), l 7.15 (3H, t,
J=13.4 Hz), 3.19 (1H, dd, J=3.4 Hz, 13.4 Hz), 3.07–2.94
(7H, m, overlapped with the doublet peak at 3.04 ppm),
3.04 (d, 14.1 Hz), 2.84 (4H, d, J=14.5 Hz); ESI–MS
(negative) m/z 719.14 [(M-9Na+6H)3−].
J=8.1 Hz), 6.76–6.71 (6H, m), 6.62–6.57 (3H, m), 4.99
(1H, s), 4.98 (2H, s), 4.47–4.43 (3H, m), 4.25–4.17 (12H,
m), 4.16–4.11 (3H, m), 3.97–3.90 (3H, m), 3.90–3.80 (6H,
m), 3.74 (3H, m), 3.71–3.65 (3H, m), 3.57 (3H, dd, J=6.0
Hz, J=11.4 Hz), 3.26 (2H, ddd, J=3.4 Hz, J=4.4 Hz,
15. Koshida, S.; Suda, Y.; Sobel, M.; Kusumoto, S. Tetra-
hedron Lett. 2001, 42, 1289–1292.
.
.