Synthetic analogs of the lewisb-tetrasaccharide and their binding to Griffonia simplicifolia lectin-part II

Article abstract of DOI:10.1055/s-2003-40330

The lewis b tetrasaccharide is known to bind to lectin IV of Griffonia simplicifolia. In an effort to prepare superior ligands, we have synthesized analogs of this tetrasaccharide and evaluated their binding to the lectin. This paper (part II in a series of two) describes the synthesis of and binding studies with analogs where the GlcNAc unit (a) has been replaced by simpler mimicking residues that are both flexible and rigid. A simple 1,2-trans cyclohexanediol residue provided a reasonable simplified inhibitor whereas a flexible 1,2-ethanediol replacement led to a completely inactive ligand.

Full text of DOI:10.1055/s-2003-40330

LETTER
1331
Synthetic Analogs of the Lewisb-Tetrasaccharide and their Binding to
Griffonia Simplicifolia Lectin-Part II
Synthetic
Analogsiof the
v
ewisb-Tetraesaccharidekanand Kamath, Ole Hindsgaul*
Department of Chemistry, University of Alberta, Edmonton, Alberta, T6G 2G2, Canada
E-mail: ole.hindsgaul@ualberta.ca
Received 12 March 2003
Dedicated to the memory of Professor Raymond U. Lemieux.
the tetrasaccharide with the protein occurs via the terminal
galactosyl residue, with further participation of both fuc-
osyl residues. The GlcNAc residue, however, is exposed
to the bulk solvent and should mainly serve as a scaffold
Abstract: The lewis b tetrasaccharide is known to bind to lectin IV
of Griffonia simplicifolia. In an effort to prepare superior ligands,
we have synthesized analogs of this tetrasaccharide and evaluated
their binding to the lectin. This paper (part II in a series of two) de-
scribes the synthesis of and binding studies with analogs where the for the correct orientation of the three other sugar rings.
GlcNAc unit (a) has been replaced by simpler mimicking residues
that are both flexible and rigid. A simple 1,2-trans cyclohexanediol
This paper deals with synthesis of and binding studies
with analogs where the GlcNAc residue has been simpli-
residue provided a reasonable simplified inhibitor whereas a flexi-
fied. Extensive studies by Lemieux et.al have revealed
ble 1,2-ethanediol replacement led to a completely inactive ligand.
very little about the role of the GlcNAc residue.9–12 We
therefore decided to synthesize analogs (1–3) of Leb-OMe
Key words: lewis b tetrasaccharide, lectins, galactose unit
by replacing the GlcNAc residue with simpler ‘mimick-
ing’ structures.
The recognition and binding of complex carbohydrates by
The synthesis of 1 (Scheme 1) began with condensation of
proteins forms a major basis for their biological activi-
the glycosyl bromide 4 (obtained from 3,4,6-tri-O-benzyl-
ties.1–5 The binding of the Lewis-b (Leb) tetrasaccharide,
1,2-O-(1-methoxyethylidine)-a-D-galactopyranose
by
a-L-Fuc-(1→2)-b-D-Gal-(1→3)-[a-L-Fuc-(1→4)]-b-D-
GlcNAc-OMe (Leb-OMe), by the lectin IV of Griffonia
simplicifolia (GS-IV has been extensively studies by Le-
mieux and co-workers.6–13 GS-IV-Leb recognition is
therefore an ideal system to test approaches to the design
of superior inhibitors of carbohydrate-protein binding.
Many tetrasaccharide analogs have been prepared and
evaluated, and the crystal structure of a tetrasaccharide
bound to the protein has been solved.7,13 High-resolution
crystallographic studies reveal that the main interaction of
treatment with acetyl bromide in presence of tetraethy-
lammonium bromide10) with the alcohol 514 under Helfer-
ich conditions. The glycoside 6 was obtained in 89%
yield. Deprotection of the primary alcohol using 2,3-
dichloro-5,6-dicyano-1,4-benzoquinone (DDQ), fol-
lowed by DMTST-promoted reaction with the fucosyl do-
nor 8 gave compound 9 as an a/b mixture where the a
isomer was purified by chromatography after de-O-
acetylation. Tetraethylammonium bromide-promoted10
reaction with the fucosyl donor 11 gave compound 12 in
O
SEt
OBn
OBn
OBn
O
BnO
BnO
BnO
BnO
OBn
O
BnO
BnO
OBn
8
BnO
HO
O
PMBO
O
a
O
b
+
OPMB
c
OH
OAc
OAc
AcO
Br
5
6
7
4
HO
BnO
OH
OBn
BnO
Br
BnO
OBn
OBn
O
OH
O
OBn
O
O
OBn
OBn
OBn
11
HO
O
OBn
OH
BnO
f
BnO
OBn
O
BnO
BnO
O
BnO
BnO
OBn
O
OBn
d
O
O
e
O
HO
O
O
O
O
BnO
O
O
O
O
12
OH
10
OAc
O
OH
O
9
OBn
1
OH
HO
BnO OBn
Scheme 1 Reagents: a) Hg(CN)2–nitromethane–toluene (89%); b) DDQ (88%); c) 8, DTBMP–DMTST, CH2Cl2 (81%); d) NaOMe–MeOH
(82%); e) 11, Bu4NBr (68%); f. H2–Pd(OH)2 (92%).
Synlett 2003, No. 9, Print: 11 07 2003.
Art Id.1437-2096,E;2003,0,09,1331,1333,ftx,en;S02503ST.pdf.
© Georg Thieme Verlag Stuttgart · New York

1332
V. Kamath, O. Hindsgaul
LETTER
BnO
BnO
OBn
BnO
BnO
BnO
BnO
OBn
O
OBn
O
O
HO
O
b
HO
O
HO
HO
a
+
AcO
OAc
OH
Br
4
13
14
15
BnO
HO
OBn
HO
OH
OBn
O
O
BnO
HO
OH
OBn
O
O
d
O
O
O
O
c
BnO
HO
11
O
O
O
O
OH
OBn
16
2
OBn
HO
BnO
HO
Scheme 2 Reagents: a) Hg(CN)2–nitromethane–toluene (74%); b) NaOMe–MeOH (94%); c) 11–Bu4NBr (70%); d. H2–Pd(OH)2 (85%).
68% yield (attempts to introduce the two fucosyl residues analog 2 were to bind to GS-IV, it would leave a hydro-
in one step gave a complex mixture of compounds which phobic cyclohexane ring projecting into solvent. This
could not be separated by chromatography). Finally, de- might well lead to very unfavorable hydration properties
benzylation by catalytic hydrogenation gave the desired so we also decided to prepare the control derivative 3
compound 115 in 92% yield.
where hydrophilic OH groups were added to the more hy-
drophobic parent structure 2.
Leb-OMe analogs 2 were synthesized in a similar fashion
(Scheme 2). (1R,2R)-trans-1,2-Cyclohexane diol 13 was For the synthesis of analog 3, (1R, 2R)-trans-cyclohex-4-
prepared as described in the literature.16 The diol 4 was enediol (17) was used (Scheme 3). Donor 4 was reacted
glycosylated with 13 using mercuric cyanide to furnish 14 with 17 using mercuric cyanide to furnish 18 in 70%
in 74% yield. De-O-acetylation, followed by introduction yield. Initial attempts to introduce the two fucosyl resi-
of the two fucosyl residues in one step using tetrabutylam- dues in one step (on de-O-acetylated 18) followed by se-
monium bromide gave 16 in 70% yield. Finally, debenzyl- lective hydroxylation in the last step were unsuccessful.
ation gave the desired tetrasaccharide 215 in 82% yield.
Therefore, compound 18 was O-acetylated and subse-
quently hydroxylated (using osmium tetroxide under stan-
dard conditions) to give a mixture of two compounds (20).
Reaction of 20 with 2,2-dimethoxypropane gave 21,
Based on the crystal complex of the native Leb-OMe tet-
rasaccharide,8 we anticipated that if the cyclohexanediol
BnO
BnO
BnO
BnO
BnO
BnO
OBn
OBn
O
OBn
AcO
O
O
a
O
b
HO
O
HO
HO
+
OAc
19
OAc
18
AcO
Br
BnO
BnO
4
17
BnO
BnO
OBn
O
OBn
AcO
O
c
d
O
f
RO
O
OH
O
Br
OAc
OR
O
OH
O
OBn
11
20
21. R = Ac
OBn
e
BnO
22. R = H
BnO
BnO
OBn
OBn
OH
OH
OBn
OBn
O
O
BnO
BnO
OH
O
g
OBn
OBn
h
O
O
O
O
O
O
O
OH
HO
OH
BnO
BnO
OH
O
O
O
O
OH
O
O
OH
O
24
O
OBn
23
O
OH
OBn
OBn
3
O
OBn
OH
BnO
BnO
OH
OH
Scheme 3 Reagents: a) Hg(CN)2–nitromethane–toluene (70%); b) acetic anhydride–pyridine (93%); c) OsO4–NMO (81%); d) 2,2-dime-
thoxypropane–TsOH (87%); e) NaOMe–MeOH (94%); f) 11, Bu4NBr (72%); g) HOAc/H2O; h) H2–Pd(OH)2 (64%, two steps).
Synlett 2003, No. 9, 1331–1333 ISSN 1234-567-89 © Thieme Stuttgart · New York

LETTER
Synthetic Analogs of the Lewisb-Tetrasaccharide
1333
which was de-O-acetylated to give the diol 22. Fucosyla-
tion of 22 using 11 and tetraethylammonium bromide
gave 23 in 72% yield. Hydrolysis of the ketal followed by
debenzylation generated the desired compound 315 in 64%
yield (mixture of two isomers 3a and 3b).
Acknowledgement
We thank Dr. A. Otter for carrying out the high-field NMR spectral
analysis, Dr. S. Nilar for the molecular modeling studies, Dr. J.
Sadowska for the ELISA assays and Dr. A. Morales for mass spec-
trometric characterization. This work was supported by grants from
the Natural Sciences and Engineering Research Council of Canada
(NSERC). We are indebted to Professor Thomas Norberg of the
Swedish Agricultural University for his critical review of this
manuscript.
The binding of GS IV to compounds 1–3 and to the parent
compound Leb-OMe was studied by ELISA inhibition as
described in the preceding communication.8 The results
are summarized in Scheme 4. The IC50 of the reference
compound Leb-OMe was 36 mM. All of the three synthe-
sized analogs 1–3 poorly inhibited lectin binding com-
pared to the parent compound, Leb-OMe. The poorest
inhibitor was analog 1, for which no inhibition could be
observed at the concentrations used. Since 1 has the most
flexible linkage between the galactosyl and fucosyl resi-
dues (b and c), this shows the importance of a more con-
formationally rigid linkage between these residues.
However, even though analogs 2 and 3 both have this
more rigid linkage, they were still poor inhibitors com-
pared to Leb-OMe. However, even the analog 2 where Fuc
and Gal residues can adopt similar if not identical relative
orientations on a more rigid template results in a 6-fold
decrease in potency. Hydration of the complex is not like-
ly the reason, since the more hydrophilic 3 was 3-fold
worse again.
References
(1) Varki, A. Glycobiology 1993, 3, 97.
(2) Weis, W. I.; Drickamer, K.; Hendrickson, W. A. Nature
(London) 1992, 360, 127.
(3) Dennis, J. W. In Cell Surface Carbohydrates and Cell
Development; Fukuda, M., Ed.; CRC Press: Boca Raton,
1992, 161–194.
(4) Giannis, A. Angew. Chem., Int. Ed. Engl. 1994, 33, 178.
(5) Karlsson, K. A. Trends in Pharmacol. Sci. 1991, 12, 265.
(6) Lemieux, R. U. Am. Chem. Soc. Sym. Ser. 1993, 519, 5.
(7) Delbaere, L. T. J.; Vadonselaar, M. P. L.; Quail, J. W.;
Wilson, K. S.; Dauter, Z. J. Mol. Chem. 1993, 230, 950.
(8) Kamath, V.; Sadowska, J.; Nilar, S.; Bundle, D. R.;
Hindsgaul, O. Synlett 2003, 1327.
(9) Spohr, U.; Hindsgaul, O.; Lemieux, R. U. Can. J. Chem.
1985, 63, 2644.
(10) Spohr, U.; Lemieux, R. U. Carbohydr. Res. 1988, 174, 211.
(11) Nikrad, P. V.; Beierbeck, H.; Lemieux, R. U. Can. J. Chem.
1992, 70, 241.
(12) Lemieux, R. U. Chem. Soc. Rev. 1989, 18, 347.
(13) Delbaere, L. T. J.; Vandonselaar, M. P. L.; Quail, J. W.;
Nikrad, P. V.; Pearlstone, J. R.; Carpenter, M. R.; Smillie, L.
B.; Spohr, U.; Lemieux, R. U. Can. J. Chem. 1990, 68, 1116.
(14) Huang, H.; Wong, C.-H. J. Org. Chem. 1995, 60, 3100.
(15) 1H NMR (D2O): 1. d: 4.09 (q, 1 H, H-5b), 4.31 (q, 1 H, H-
5c), 4.57 (d, 1 H, H-1a), 4.93 (d, 1 H, H-1b), 5.23 (d, 1 H, H-
1c); 2. d: 4.48 (q, 1 H, H-5d), 4.56 (d, 1 H, H-1b), 4.70 (q, 1
H, H-5c), 4.99 (d, 1 H, H-1c), 5.28 (d, 1 H, H-1d); 3a. d: 4.35
(d, 1 H, H-5d), 4.52 (d, 1 H, H-1b), 4.60 (q, 1 H, H-5c), 5.00
(d, 1 H, H-1c), 5.28 (d, 1 H, H-1d); 3b. d: 4.57 (d, 1 H, H-
1b), 4.73 (q, 1 H, H-5c), 4.91 (d, 1 H, H-1c), 5.27 (d, 1 H, H-
1d).
Clearly, replacement of the 3,4-disubsituted GlcNAc res-
idue in the Leb tetrasaccahride by a flexible ethane diol as
in 1 can be expected to result in a dramatic decrease in ac-
tivity. However, when this identical strategy was applied
to the development of inhibitors of E-selectin binding to
the tetrasaccharide sialyl-LeX (SLeX), replacement of a
3,4-diglycosylated GlcNAc residue by ethane diol yielded
a greatly simplified structure that was as potent as the
parent compound.14 Replacement of the same GlcNAc
residue in SLeX by cyclohexane diol (as was done for 2)
yielded a structure that was almost an order of magnitude
more active.17 In conclusion, and as expounded by
Lemieux,12 the binding of oligosaccharides by proteins re-
mains poorly understood and rules for the design of potent
inhibitors are in the early stages of development. One of
the most important and difficult factors to incorporate into
inhibitor design is the difference in hydration of the free
protein and sugar vs the complex they form on binding.
(16) Suemune, H.; Hizuka, M.; Kamashita, T.; Sakai, K. Chem.
Pharm. Bull. 1989, 37, 1379.
(17) Banteli, R.; Ernst, B. Tetrahedron Lett. 1997, 38, 4059.
HO
HO
HO
HO
OH
OH
OH
OH
OH
OH
c
OH
OH
O
O
O
O
OH
HO
HO
OH
HO
HO
HO
HO
HO
OH
OH
OH
O
OH
O
O
O
O
O
O
a
O
O
O
O
b
O
O
OMe
O
OH
NHAc
O
O
O
OH
O
O
d
O
O
OH
OH
OH
OH
OH
OH
OH
OH
HO
HO
HO
HO
3; IC50 = 730 µM
Leb-OMe; IC50 = 36 µM
2; IC50 = 222 µM
1; IC50 = inactive
Scheme 4
Synlett 2003, No. 9, 1331–1333 ISSN 1234-567-89 © Thieme Stuttgart · New York