Solid-phase oligosaccharide synthesis with tris(alkoxy)benzyl amine (BAL) safety-catch anchoring

Article abstract of DOI:10.1039/a908446k

A tris(alkoxy)benzylamine (BAL) handle strategy was developed for safety-catch anchoring of D-glucosamine derivatives in solid-phase synthesis of oligosaccharides; the linkage between the BAL handle and the amine proved stable to conc. TFA and Lewis acids, but after N-acylation the amide could be released by treatment with TFA-H₂O (19:1).

Full text of DOI:10.1039/a908446k

Solid-phase oligosaccharide synthesis with tris(alkoxy)benzyl amine (BAL)
safety-catch anchoring
Jakob F. Tolborg and Knud J. Jensen*
Department of Organic Chemistry, Building 201, Kemitorvet, Technical University of Denmark, DK-2800 Lyngby,
Denmark. E-mail: okkj@pop.dtu.dk
Received (in Liverpool, UK) 20th October 1999, Accepted 29th November 1999
A tris(alkoxy)benzylamine (BAL) handle strategy was devel-
oped for safety-catch anchoring of -glucosamine deriva-
tives in solid-phase synthesis of oligosaccharides; the linkage
between the BAL handle and the amine proved stable to
conc. TFA and Lewis acids, but after N-acylation the amide
could be released by treatment with TFA–H2O (19+1).
In the second technique, the resin-bound monosaccharide
was treated with 4,4A-dimethoxytrityl chloride (DMTCl) to
attach a DMT moiety at the primary 6-OH group. For
quantification, the resin was exposed to TFA–MeCN (1+49)
and released DMT estimated by UV. Quantification using DMT
was performed on resin bound monosaccharides in order to
determine the efficiency of the coupling of o-PALdehyde to the
resin as well as the coupling of the amino sugar to the linker.
D
Cell-surface oligosaccharides are involved in many biological
recognition processes.1 Also, free oligosaccharides can have
biological effects such as in rhizobial lipochitin oligosacchar-
ides which function as nodulation factors.2 Amino sugars are
found in many biologically important poly- and oligo-sacchar-
ides, e.g. in glycans of O- and N-glycoproteins and -peptides,
lipochitin nodulation factors, and amino glycoside antibiotics
such as streptomycin.
Solid-phase synthesis of oligosaccharides, when performed
combinatorially, offers the prospect of easy access to very large
numbers of well-defined oligosaccharides as tools for glyco-
biology. Despite Frechet and Schuerch’s early pioneering work3
and recent significant progress,4 considerable developments are
required before solid-phase oligosaccharide synthesis can be
put to general use. Crucial to any solid-phase synthesis plan is
the efficient and reliable anchoring of the first building block,
the stability of the anchoring linkage to the synthesis conditions,
and upon completion of the synthesis the efficient release under
conditions compatible with the structural integrity of the final
product. The most common cleavage reagent in solid-phase
peptide synthesis is conc. TFA. Previously reported handles for
anchoring of carbohydrates to solid supports have included
esters,4a silyl ethers,4b sulfonyl esters,4c acetals4d and thio-
alkyls4e and ethers.4f,g
The methyl and benzyl glycosides of -glucosamine 3b and 3c
were coupled with an efficiency of at least 95% (four steps from
D
2, including DMT coupling and cleaving) independently of the
resin loading. The solid-phase bound benzyl a- -glucosaminide
D
3c was then per-acylated with Ac2O or BzCl. This yielded the
fully acetylated and benzoylated products 4b and 4c in 91–96
and 82% yield, respectively, measured by HPLC standard
curves.
Several solution-phase glycosylation methods have been
adapted for solid-phase; they all require Lewis acids for
activation.4 While studying the use of trichloroacetimidates in
For the solid-phase synthesis of amino sugar containing
oligosaccharides, we decided upon a strategy relying on release
of final products by treatment with conc. TFA, with semi-
permanent protecting groups to be removed in solution after
purification and characterization.5 The recently developed
Backbone Amide Linker (BAL) handle strategy relies on
anchoring of a peptide not through a carboxy group but through
a backbone amide nitrogen leaving the C-terminus available for
modifications.6 Applied to anchoring of amino sugars, a BAL
strategy would allow anchoring through an acetamido function-
ality leaving all hydroxy groups, including the anomeric,
available for modification.
To investigate the anchoring concept, o-PALdehyde 1 was
coupled to high-loading aminomethylated polystyrene 2 and
unprotected -glucosamine 3a anchored through a BAL linkage
D
by reductive amination in DMF–AcOH (99+1) in the presence
of NaBH3CN (Scheme 1).† Perbenzoylation with BzCl fol-
lowed by treatment with TFA–H2O (19+1) yielded after column
chromatography the a-pentabenzoate 4a in 41% for a four step
synthesis sequence. The presence of five benzoyl groups in the
final product confirmed that anchoring had occurred through the
amine rather than the putative O,O- or O,N-acetals.
For fast and efficient quantification of resin-bound carbohy-
drates we used two techniques. In the first, UV-active amino
sugar derivatives were released into solution with TFA–H2O
(19+1) followed by quantification with HPLC–UV from
standard curves.
Scheme 1 Reagents and conditions: i, aminomethylated polystyrene 2, 1
(2.0 equiv.), HBTU (1.9 equiv.), HOBt (2.0 equiv.), Pri2NEt (3.9 equiv.),
DMF, 10 h; ii, amino sugar (2.0 equiv.), NaBH3CN (10 equiv.), DMF–
AcOH (99+1), 16 h, then protecting group chemistry including N-acylation;
iii, TFA–H2O (19+1), 30 min.
Chem. Commun., 2000, 147–148
This journal is © The Royal Society of Chemistry 2000
147

The initial resin loading of the aminomethylated polystyrene
2 was 0.40 mmol g21, which dropped to 0.36 mmol g21 after
coupling of o-PALdehyde 1 to the resin, and 0.32 mmol g21
after attaching the partly protected monosaccharide 3d. With
high-loading resin (1.20 mmol g21) glycosylation became much
less efficient..
In conclusion, we have developed a new and efficient
strategy for anchoring amino sugars through a BAL handle to a
solid support in which the amino sugar is attached by an
efficient reductive amination. BAL anchoring was operated in a
safety-catch mode as BAL linked amines were stable to Lewis
acids and conc. TFA, whereas the corresponding amides were
released from the support with conc. TFA. This allowed the use
of Lewis acids in excess in solid-phase glycosylations. Using
this strategy, we synthesized 1-6 linked disaccharides in high
yields. The safety-catch protocol for BAL anchoring should
also be useful for other applications.
We thank Flemming Jensen and Annette Lind Nordestgaard,
Phytera, for MS, Drs Charlotte Godfredsen and Henrik
Pedersen for MAS NMR, and the Lundbeck Foundation (K. J.)
and the Danish Technical Research Council (‘Talent Project’
grant to K. J.) for financial support.
Scheme 2 Reagents and conditions: i, TMSOTf, CH2Cl2, 3 Å (Table 1); ii,
Ac2O–Py (2:1), 16 h; iii, TFA–H2O (19+1), 30 min.
Notes and references
† All solid-phase reactions were carried out in plastic syringes at room
temperature, except for glycosylations at 250 °C, which were carried out in
glass flasks. Syringes were fitted with a polypropylene filter, a Teflon valve
in the bottom, and closed in the top with the syringe plunger. For
glycosylations, the syringe was instead closed with a septum, after the solid
reactants were placed in the syringe. It was then dried in a desiccator in
vacuo with the Teflon valve open. The desiccator was opened under argon,
the Teflon valve closed, and the solvent and Lewis acid added through the
septum. After 18 h the resins were washed and dried; final products were
released from the support with TFA–H2O (19:1). Glycosylations were
typically performed on a 5 mmol scale. Quantifications were done by
HPLC–UV integration of peak areas and use of standard curves established
from benzyl and benzoyl containing compounds. Benzyl groups were
monitored at l = 215 nm, and benzoyl groups at l = 265 nm. At the 5 mmol
scale and without purification, the structure of 6b was established by MS,
1H NMR and gCOSY spectroscopy (ref. 8).
the presence of Lewis acid promoters for the glycosylation of
BAL anchored
D
-glucosaminides, we realized that BAL
anchoring could be operated in a safety-catch mode. The
amine–BAL linkage proved stable to treatment with conc. TFA
and more than stoichiometric amounts of Lewis acids such as
BF3·OEt2 and TMSOTf. After acylation to form the amide, the
linkage became acid-labile and the carbohydrate could be
released with conc. TFA.
For solid-phase glycosylations, we first synthesized the
partially protected
D-glucosamine derivative 3d. Anchoring
through a BAL handle to a polystyrene support was achieved as
above, leaving the 6-OH free. In initial studies we used
2,3,4,6-tetra-O-benzyl-b- -glucopyranosyl trichloroacetimi-
D
date 5a7 in the presence of BF3·OEt2 or TMSOTf as promoters
(Scheme 2). Solid-phase reactions are most conveniently
carried out at room temperature, but this gave only low yields of
the disaccharide. However, when performing the glycosylation
at 250 °C the disaccharide 6a was obtained in 35% yield as an
a/b mixture (five steps from 2).
1 A. Varki, Glycobiology, 1993, 3, 97; K. J. Yarema and C. R. Bertozzi,
Curr. Opin. Chem. Biol., 1998, 49.
2
P. Lerouge, Glycobiology, 1994, 4, 127.
3 J. M. Frechet and C. Schuerch, J. Am. Chem. Soc., 1971, 93, 492.
4 (a) M. Adinolfi, G. Barone, L. D. Napoli, A. Iadonisi and G. Piccialli,
Tetrahedron Lett., 1998, 39, 1953; (b) C. Zheng, P. H. Seeberger and S. J.
Danishefsky, J. Org. Chem., 1998, 63, 1126; (c) J. A. Hunt and W. R.
Roush, J. Am. Chem. Soc., 1996, 118, 9998; (d) Y. Ito and T. Ogawa,
J. Am. Chem. Soc., 1997, 119, 5562; (e) J. Rademann and R. R. Schmidt,
J. Org. Chem., 1997, 62, 3650; (f) G. Hodosi and J. J. Krepinsky, Synlett,
1996, 159; (g) S. P. Douglas, D. M. Whitfield and J. J. Krepinsky, J. Am.
Chem. Soc., 1995, 117, 2116.
5 Conc. TFA has been used extensively for release of final products in
solid-phase synthesis of glycopeptides. Glycosidic bonds appear in
general to be fully stable to these conditions, especially while retaining
ester protecting groups. For a review see M. Meldal, in Neoglycoconju-
gates: Preparation and Applications, ed. Y. C. Lee and R. T. Lee,
Academic Press, San Diego, 1994, p. 145.
When using the less reactive benzoyl protected glycosyl
donor 2,3,4,6-tetra-O-benzoyl-a- -glucopyranosyl trichloro-
D
acetimidate 5b in the presence of TMSOTf, yields of up to 82%
of the disaccharide 6b8 were obtained (five steps from 2),
leaving as little as 2% of the unglycosylated monosaccharide
(Table1). The disaccharide was obtained with an a+b ratio of
< 1+10, the high b-selectivity being due to the neighboring
group participation of the benzoyl group. Stoichiometric and
sub-stoichiometric amounts of Lewis acids were used, relative
to the donor. We found that TMSOTf was superior to BF3·OEt2
as promoter, and MAS NMR revealed that the secondary amine
did not carry a TMS group after exposure to TMSOTf in
glycosylations. The high yield of 82% for the five step synthesis
of 6b also indicated that release of the final product with TFA–
H2O was near quantitative.
6 K. J. Jensen, J. Alsina, M. F. Songster, J. Vágner, F. Albericio and G.
Barany, J. Am. Chem. Soc., 1998, 120, 5441.
7 R. R. Schmidt, J. Michel and M. Roos, Liebigs Ann. Chem., 1984,
1343.
Table 1 Reaction conditions for synthesis of 6b
8 Selected data for 6b: dH(CDCl3, 500 MHz) 8.01 (dd, J 1.3, 8.5, 2H), 7.90
(dd, J 1.3, 8.5, 4H), 7.82 (dd, J 1.7, 8.5, 2H), 7.54–7.48 (m, 2H),
7.44–7.41 (m, 2H), 7.38–7.22 (m, 16H), 7.14 (dd, J 1.7, 7.7, 2H), 5.91
(dd, J 9.8, 9.4, 1H), 5.69 (dd, J 9.8, 9.4, 1H), 5.59 (dd, J 9.8, 8.1, 1H), 5.21
(d, J 9.4, 1H), 4.91 (d, J 7.7, 1H), 4.71 (d, J 11.5, 1H), 4.63 (dd, J 12.0,
3.4, 1H), 4.57–4.53 (m, 2H), 4.52 (d, J 11.5, 1H), 4.48 (d, J 3.4, 1H), 4.39
(d, J 11.1, Hz, 1H), 4.19–4.13 (m, 3H), 3.75 –3.68 (m, 2H), 3.57 (dd, J
10.7, 9.0, 1H), 3.42 (dd, J 9.4, 9.0, 1H), 3.07 (s, 3H), 1.81 (s, 3H); LC-MS
(C57H55NO15) calc. [M+H+], 994.4, [M 2 MeO2], 962.4; found 994.9,
962.8.
TMSOTf/
equiv.
Yield of
6b (%) (b/a)
T/°C
5b/equiv.
Room temp.
Room temp.
Room temp.
Room temp.
Room temp.
Room temp.
Room temp.
5.0
5.0
5.0
1.0
2.0
5.0
10.0
2.0
5.0
0
76 (14)
74 (20)
64 (17)
82 (11)
82 (16)
67 (17)
5.0
10.0
10.0
10.0
10.0
Communication a908446k
148
Chem. Commun., 2000, 147–148