Specific crosslinking of cell adhesive molecules by heterobifunctional glycopeptide synthesised on the basis of chemoenzymatic strategy

Article abstract of DOI:10.1039/a904362d

Heterobifunctional glycopeptide 1 composed of Neu5Acα-(2→3)Galβ(1→4)[Fucα(1→3)]GlcNAc (sialyl Lewis(x)) and Lys-Gly-Arg-Gly-Asp-Ser (KGRGDS) having specific activity to bind concurrently with two different types of cell adhesive molecules such as selectin

Full text of DOI:10.1039/a904362d

Specific crosslinking of cell adhesive molecules by heterobifunctional
glycopeptide synthesised on the basis of chemoenzymatic strategy
Shin-Ichiro Nishimura,*a Masao Matsuda,a Hidemitsu Kitamurab and Takashi Nishimurab
a Laboratory for Bio-Macromolecular Chemistry, Division of Biological Sciences, Graduate School of Science,
Hokkaido University, Sapporo 060-0810, Japan. E-mail: shin@glyco.sci.hokudai.ac.jp
b Division of Host Defense Mechanisms, Department of Immunology, Tokai University School of Medicine, Isehara
259-1193, Japan
Received (in Cambridge, UK) 1st June 1999, Accepted 22nd June 1999
Heterobifunctional glycopeptide 1 composed of Neu5Aca-
(2?3)Galb(1?4)[Fuca(1?3)]GlcNAc (sialyl Lewisx) and
Lys-Gly-Arg-Gly-Asp-Ser (KGRGDS) having specific activ-
ity to bind concurrently with two different types of cell
adhesive molecules such as selectins and integrins was
synthesised on the basis of a combined chemical and
enzymatic strategy.
It has been well documented that many kinds of cell adhesive
molecules (CAMs) participate in a variety of biological
phenomena such as cell–cell interactions and cell–extracellular
matrix (ECM) interactions.1–6 It should also be noted that
collaboration of multiple and complex bindings between
different types of cell surface receptors and their ligands greatly
contribute to the construction of stable and specific cellular
society.
In the present study, we designed a heterobifunctional
glycopeptide 1 as a novel and potential glycoligand, because
this artificial glycopeptide composed of Neu5Aca(2?3)Galb-
(1?4)[Fuca(1?3)]GlcNAc (sialyl Lewisx) and Lys-Gly-Arg-
Gly-Asp-Ser (KGRGDS) was supposed to interact with two
different types of families of cell adhesive proteins such as
selectins and integrins concurrently. This type of synthetic
heterobifunctional glycopeptide is strongly expected to become
a novel class of antiflammatory and anticancer ‘glycodrugs’.7–9
Kunz et al. have reported a significant ‘cis’ type interaction of
the synthetic RGDS-sialyl Lewisx with P-selectin as an IC50
value of 26 mM. In addition to this type of enhanced affinity of
with receptor molecules. (A) and (B) represent cooperative-type ‘cis’
Fig. 1 Chemical structure of 1 and plausible mechanisms of binding of 1
glycoligand with protein, we found an interesting ‘trans’ type
interaction of a novel glycoligand 1 involving a flexible peptide
linker with two different protein receptors. Fig. 1 summarises
some plausible mechanisms of ‘cis’ and ‘trans’ type inter-
actions of glycoligand 1 with cell adhesive proteins.
interactions, and (C) and (D) represent crosslinking-type ‘trans’ inter-
actions (crosslinking by glycopeptide), respectively.
Scheme 1 indicates a retrosynthetic route for compound 1.
Firstly, a key intermediate 2 was prepared by chemical synthesis
from a simple n-pentenyl glycoside of N-acetyllactosamine 510
and tetrapeptide derivative 4 as starting materials (Scheme 2).
The terminal double bond of compound 5 was converted into a
carboxy group by treating with potassium permanganate in 87%
yield and the active ester of 6 was employed for the coupling
reaction with the e-amino group of Z-lysine to afford 7 in 71%
yield. To achieve satisfactory flexibility in the spacer-arm
between sialyl Lewisx and RGDS and versatility in further
chemical manipulation, H-Gly-OBn was combined with 7 to
give N-acetyllactosamine-dipeptide derivative 8 in 83% yield.
Consequently, deprotection of the benzyl ester and O-acetyl
groups of 8 under alkaline conditions was allowed to proceed
smoothly and gave compound 3 in 87% yield without
racemization. Coupling reaction of the glycopeptide 3 with an
unprotected disaccharide moiety with an amino-terminus gen-
erated from RGDS derivative 4 was carried out by employing
diphenylphosphoryl azide11 as a selective promoter of the
carboxy group of the glycopeptide. A hexapeptide derivative 9
was obtained in 77% yield with no side reaction at the free
hydroxy groups of the carbohydrate moiety. Removal of all
Scheme
(CAGP 1).
1 Retrosynthetic route to cell adhesive glycopeptide 1
Chem. Commun., 1999, 1435–1436
1435

Scheme 2 Reagents and conditions: i, KMnO4, AcOH aq., rt, 87%; ii,
HONSu, DCC, DMF, 0 °C?rt, then Z-Lys, TEA, DMF–H2O (9+1), rt,
71%; iii, DPPA, Gly-OBn, TEA, DMF, 0 °C?rt, 83%; iv, NaOMe, dry
MeOH, rt, then 1 m NaOH aq., rt, ion exchange resin (Dowex 50W-X8),
87%; v, 2 m HCl–dioxane, 0 °C?rt; vi, DPPA, TEA, DMF, 0 °C?rt, 77%
(over all yield from 4); vii, Pd/C, H2 gas, MeOH–H2O, (1+1) rt, 99%.
protecting groups of the peptide moiety of 9 such as benzyl,
nitro, and benzyloxycarbonyl groups was performed by stan-
dard hydrogenation in the presence of palladium on charcoal to
give the intermediate 2 in 92% yield.
Fig. 2 Crosslinking of selectin and integrin by cell adhesive glycopeptide
(CAGP) monitored by SPR method. The first arrow ‘A’ indicates the time
point of the beginning of the injection of cell adhesive glycopeptide (20 mL,
1.3 mg mL21) to the P-selectin immobilised cuvette. The second arrow ‘B’
Subsequently, the glycopeptide 2 was employed as a
substrate for the corresponding glycosyltransferases. It is of
interest to examine the substrate specificity of the enzymatic
glycosyltransfer reactions against non-natural glycopeptide 2,
since acceptability of synthetic glycoconjugates to the glycosyl-
transferases depends strongly on the nature of aglycons or
supporting backbones.12 In fact, a synthetic hexapeptide bearing
an N-acetyllactosamine side chain through an appropriate
hydrophobic linker (2) was proved to be an excellent substrate
both for a(2”?3)sialyltransferase and for a(1?3)fucosyl-
transferase. The sialylation and fucosylation of compound 2
proceeded smoothly and glycopeptide 1 was isolated by
preparative HPLC in 45% overall yield from 2 (Scheme 3).
The effect of hybridisation of the functional carbohydrate and
peptide on the ‘cis’ type interaction between glycoligand 1 and
each protein was investigated by surface plasmon resonance
(SPR). As anticipated, compound 1 binds P- and E-selectins
more strongly than native sialyl Lewisx, and association
constants of 1 with P- and E-selectins were assumed to be 6.6 3
107 M21 and 4.5 3 105 M21, respectively. It was also found that
1 inhibited interaction of human integrin b1 with its monoclonal
antibody more effectively than RGDS (IC50 = 0.55 mM).
indicates the time point of the injection of integrin b1 (20 mL, 9.1 mg mL21
)
to this cuvette. The third arrow ‘C’ indicates the beginning of the buffer
washout (200 mL of phosphate buffer solution). Integrin b1 was found to
bind to CAGP-P-selectin complex as indicated by a significant increase in
the response unit.
These successful results of cooperative ‘cis’ interactions
prompted us to examine the artificial crosslinking of selectin
and integrin by the synthetic glycoligand. As indicated in Fig. 2,
glycopeptide 1 exhibited the specific capacity to bind P-selectin
and human integrin b1 concurrently. We also have evidence that
1 could block the integrin-mediated adhesion of activated T
cells to a collagen-coated plate at the same level as inhibition by
a monoclonal antibody to integrin. Therefore, it is strongly
suggested that synthetic glycopeptide 1 will become an
effective practical tool, ‘cell adhesive glycopeptide (CAGP)’, to
modulate immune responses through its inhibitory effect on
lymphocyte-endothelial cell interaction.
Notes and references
1 T. A. Springer, Nature, 1990, 346, 425.
2 R. O. Hynes and A. D. Lander, Cell, 1992, 68, 303.
3 L. A. Lasky, Science, 1992, 258, 964.
4 E. A. Clark and J. A. Ledbetter, Nature, 1994, 367, 425.
5 L. Steinman, Cell, 1995, 80, 7.
6 E. C. Butcher and L. J. Picker, Science, 1996, 272, 60.
7 M. Matsuda and S.-I. Nishimura, Peptide Chem., 1996, 1995, 381.
8 U. Sprengard, G. Kretzschmar, E. Bartnik, C. Huls and H. Kunz, Angew.
Chem., Int. Ed. Engl., 1995, 34, 990.
9 M. D. Pierschbacher, E. G. Hayman and E. Ruoslahti, Proc. Natl. Acad.
Sci. U.S.A., 1983, 80, 1224.
Scheme 3 Reagents and conditions: i, 2 (20 mg), CMP-Neu5Ac (25 mg), a-
2,3-sialyltransferase (0.3 unit), calf intestine alkaline phosphatase (20 unit),
BSA (10 mg), MnCl2·4H2O, Triton CF-54, 50 mM sodium cacodylate
buffer (2.0 mL, pH 7.4), 37 °C, 3 d, 80%; ii, 10 (15 mg), GDP-Fuc (10.3
mg), a-1,3-fucosyltranseferase V (10 munit), calf intestine alkaline
phosphatase (1 unit), NaN3, MnCl2·4H2O, 100 mM HEPES buffer (1.0 mL,
pH 7.5), 37 °C, 3 d, 56%.
10 S.-I. Nishimura, K. Matsuoka, T. Furuike, S. Ishii, K. Kurita and K. M.
Nishimura, Macromolecules, 1991, 24, 4236
11 T. Tsuda and S.-I. Nishimura, Chem. Commun., 1996, 2779.
12 S.-I. Nishimura and K. Yamada, J. Am. Chem. Soc., 1997, 119,
10 555.
Communication 9/04362D
1436
Chem. Commun., 1999, 1435–1436