Design and synthesis of sialyl Lewis x mimics as E-selectin inhibitors

Article abstract of DOI:10.1016/S0960-894X(00)00623-5

The design and synthesis of novel β-C-mannosides that inhibit the binding of sialyl Lewis x to E-selectin are described. Compounds that contained a phenyl substituent at the C-6 position were found to have increased potency.

Full text of DOI:10.1016/S0960-894X(00)00623-5

Bioorganic & Medicinal Chemistry Letters 11 (2001) 151±155
Design and Synthesis of Sialyl Lewis x Mimics as
E-Selectin Inhibitors
Neelu Kaila,Ã Bert E. Thomas, IV, Paresh Thakker, Juan C. Alvarez,
Raymond T. Camphausen and Deidre Crommie
Departments of Chemical Sciences, Biological Chemistry and Immunology & Hemostasis,
Wyeth Research, 85 Bolton Street, Cambridge, MA 02140, USA
Received 9 June 2000; accepted 27 October 2000
AbstractÐThe design and synthesis of novel b-C-mannosides that inhibit the binding of sialyl Lewis x to E-selectin are described.
Compounds that contained a phenyl substituent at the C-6 position were found to have increased potency. # 2001 Elsevier Science
Ltd. All rights reserved.
The selectins are a family of calcium dependent, cell
adhesion molecules that mediate the initial rolling of leu-
kocytes during an in¯ammatory response. Two members
of the selectin family, E- and P-selectin, are expressed on
the activated endothelium at sites of in¯ammation.
They recognize and bind to a minimal lactosaminogly-
can carbohydrate epitope, sLex (sialyl Lewis x, 1) on the
leukocytes. The excess recruitment of leukocytes can con-
tribute to several acute diseases, such as stroke and reper-
fusion injury, and chronic diseases, such as psoriasis and
rheumatoid arthritis. The development of a small mole-
cule inhibitor that blocks the interaction of selectins with
sLex has become an important therapeutic target.1 7
of small molecules that mimic sLex and show in vitro
activity can be found in the literature.12 21
As part of our ongoing e€ort in development of an
E-selectin antagonist, we are interested in synthetically
accessible sLex mimics.22 Some of the sLex mimetics in the
literature contain a mannose moiety13 18 that functionally
replaces the fucose of sLex and a carboxylate connected by
a linker that plays the role of sialic acid. In one such
report by Wong et al.,18 a-mannosyl glutamate 2 was
shown to be ®ve times more potent than sLex. We were
interested in using the C-mannose core due to its che-
mical stability and ease of availability. We synthesized
222 as well as the b-mannosyl glutamate 3 and tested
them in our ELISA-based assay, which measures the
binding of E-selectin to human recombinant AGP (a-1
acid glycoprotein) containing sLex. We were surprised
to ®nd that b-mannosyl glutamate 3 showed activity.
Further probing was done using molecular modeling.
The role of the individual sugars of sLex in selectin±
ligand interactions has been identi®ed.8 11 Sialic acid, l-
fucose and galactose have been shown to be critical for
the binding of sLex to selectins. The N-acetylglucos-
amine moiety permits some variations. Several reports
ÃCorresponding author. Tel.: +1-671-665-8974; fax: +1-617-665-8993; e-mail: nkaila@genetics.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00623-5

152
N. Kaila et al. / Bioorg. Med. Chem. Lett. 11 (2001) 151±155
The conformation of sLex in solution has been studied
by a number of groups using 2-D NMR and molecular
dynamics simulation.23 The crystal structure of E-selec-
tin has also been determined.24 Consequently, a number
of proposed complex structures of sLex bound to
E-selectin have appeared in the literature.23,25 28 We
used the proposed structure of Poppe et al.23 as a start-
ing point for modeling the bound structures of manno-
sides 2 and 3 to E-selectin, respectively. In the proposed
sLex±E-selectin complex structure, the 2- and 3-hydro-
xyls of the l-fucose moiety, which are both in equatorial
orientations, are bound to the essential selectin-bound
calcium. The mannose rings of 2 and 3 were overlaid
onto the l-fucose in the proposed complex structure. The
glutamate side chains of 2 and 3 were then orientated
such that they were pointed in the general direction of the
sialic acid binding site. Using the program QXP,29 these
initial structures of 2 and 3 bound to E-selectin were
energy minimized, followed by a Monte Carlo simula-
tion in which the ligand was allowed to rotate, translate
and ¯ex torsions for up to 5000 steps while the protein
was held rigid. In both the minimizations and the
Monte Carlo simulations the 3- and 4-hydroxyls of the
mannose moiety were constrained to be proximal to the
selectin-bound calcium through the use of zero-order
bonds. The resulting structures, shown in Figure 1, were
signi®cantly di€erent from the starting point. Instead of
the mannose moiety remaining overlaid on the fucose
moiety in the sLex structure, it relaxes to lie on the sur-
face of E-selectin, optimizing the van der Waals and
electrostatic interactions.
The bias for a-substituents is grounded in the structure
of sLex, which has an a linkage between fucose and
glucose, and the proposed complex structures of sLex
bound to E-selectin, which indicate that substituents in
the a position of the fucose (or similarly a mannose) are
directed towards the sialic acid binding site while sub-
stituents in the b position are directed towards solvent.
The structures that we obtained from our Monte Carlo
simulations show a di€erent trend. b substituents are
pointed in the general direction of the sialic acid binding
site, while substituents in the a position are directed
away from the sialic acid binding site. This occurs
because the tetrasaccharide has been clipped to a sub-
stituted monosaccharide. In the absence of the rigid
con®rmation of the tetrasaccharide the mannose relaxes
onto the surface of E-selectin in order to optimize
interactions between the mannose hydroxyls and the
protein. The relaxation process results in the anomeric
carbon being closer to the surface of E-selectin. The
interpretation of these structures does not preclude
a substitutents from interacting with the binding site,
as they must be given the number of a substituted
mannose derivatives that have been reported in the lit-
erature. It does suggest that b substituted mannose
derivatives may be potential inhibitors. Based on the
modeling results and the fact that they had not been
previously explored in the literature, we decided to use
the b-mannose sca€old.
Accordingly, several C-1 substituted b-mannosides were
prepared (Table 1). The synthetic approach used to
prepare C-mannosides 3±11 is outlined in Scheme 1.
Methyl-2-(2,3,4,6-tetra-O-benzyl-b-d-mannopyranosyl)
acetate 12 was prepared from 2,3,4,6-tetra-O-benzyl-d-
mannopyranose in two steps via a known procedure.30
Hydrolysis of ester 12 yielded 2-(2,3,4,6-tetra-O-benzyl-
b-d-mannopyranosyl) acetic acid 13, which on coupling
Figure 1. The Monte Carlo structures of 2 (yellow) and 3 (red) bound
to E-selectin are shown. The surface of E-selectin is represented by a
Connelly surface that is colored by lipophilicity. The locations of the
bound calcium and the sialic acid binding site are indicated.
Scheme 1. Synthesis of C-1 substituted mannosides 3±11.

N. Kaila et al. / Bioorg. Med. Chem. Lett. 11 (2001) 151±155
Table 1. C-1 Mannosides in the E-selectin assay
153
There are several possible explanations for their lack of
activity.34 The sLex binding site on E-selectin is relatively
shallow, which results in the sLex±E-selectin interaction
resembling a protein±protein interaction more than a
protein±small molecule interaction. If this is truly the
case, our ligands may need to be larger to occupy more
of the sLex binding site in order to out compete sLex. In
order to address these concerns, we decided to explore a
di€erent site on the E-selectin surface.
rAGP/ESel Assay
IC50 (mM)
Compound
NHR or NHR1
sLex
NH2
0.3
4
3
NA at 10 mM
7
6
Using the lowest energy Monte Carlo structure of 3
bound to E-selectin as a starting point, a virtual library
that corresponds to derivatization of C-6 of mannose was
generated in the presence of E-selectin using QXP. Each
initial structure was minimized, followed by a Monte
Carlo simulation of up to 1000 steps in which the ligand
was allowed to rotate, translate and ¯ex while the pro-
tein was held rigid. The 3- and 4-hydroxyls of the man-
nose moiety of each construct were constrained to be
proximal to the selectin-bound calcium through the use of
zero-order bonds. The low energy structure of each con-
struct was retained and ranked using the QXP scoring
function.29 The top 50 structures were visually screened
and the best 10 structures were retained for synthesis.
5
6
7
8
7
10
>1 0
9
>1 0
The C-6 substituted derivatives of b-mannosyl gluta-
mate 3, were prepared by the procedure shown in
Scheme 2 (compounds 18±24). Glycoside 12 on treat-
ment with trimethylsilyl tri¯ate and excess acetic anhy-
dride gave exclusively the C-6 acetyl glycoside,35 which
on hydrolysis with sodium hydroxide gave acid 15. The
latter on coupling with dibenzyl glutamate resulted in
16. The C-6 hydroxyl was then converted to the alde-
hyde by Swern oxidation. Reductive aminations of core
17 produced amines in 60±90% yields. The latter on
further hydrogenolysis yielded mannosides 18±24.31
10
NA at 10 mM
11
NA at 10 mM
with a variety of protected amino acids gave amides in 40±
60% yields. After puri®cation the amides were subjected
to hydrogenolysis to give the desired products (3±9)31,32 in
quantitative yield. Reduction of ester 12 with DIBAL at
70 ꢀC furnished aldehyde 14 in 70% yield. Reductive
aminations using sodium cyanoborohydride gave
amines 10 and 11.31,32
These compounds were tested in our in vitro E-selectin
assay (Table 2). Compounds 18, 19, 20, and 22 that
contain phenyl substituents at the C-6 position showed
activity. Glycoside 20 was the most active; being four
times more active than the earlier lead compound 3.
These results indicate that we have possibly accessed a
hydrophobic patch on the E-selectin surface with the
phenyl substituents at C-6 position. Recently several
Biological evaluation of these compounds in our ELISA
produced disappointing results. A majority of the com-
pounds showed IC50s greater than 10 mM. Glycoside 3
was found to be the most potent compound (Table 1).33
Scheme 2. Synthesis of C-6 substituted mannosides 18±24.

154
N. Kaila et al. / Bioorg. Med. Chem. Lett. 11 (2001) 151±155
Table 2. C-6 Mannosides in the E-selectin assay
11. DeFrees, S. A.; Kosch, W.; Way, W.; Paulson, J. C.;
Sabesan, S.; Halcomb, R. L.; Huang, D.-H.; Ichikawa, Y.;
Wong, C.-H. J. Am. Chem. Soc. 1995, 117, 60 and references
cited therein.
12. For recent synthesis of sLex mimetics as E-selectin inhibi-
tors see refs 12±21. Huwe, C. M.; Woltering, T. J.; Jiricek, J.;
Weitz-Schmidt, G.; Wong, C.-H. Bioorg. Med. Chem. 1999, 7,
773. Thoma, G.; Kinzy, W.; Bruns, C.; Patton, J. T.; Magnani,
J. L.; Banteli, R. J. Med. Chem. 1999, 42, 4909.
rAGP/Esel Assay
IC50 (mM)
0.3
Compounds
NR2R3
sLex
18
6.0
13. Lin, C.-C.; Moris-Vara, F.; Weitz-Schmidt, G.; Wong, C.-
H. Bioorg. Med. Chem. 1999, 7, 425.
14. Tsai, C.-Y.; Park, W. K. C.; Weitz-Schmidt, G.; Ernst, B.;
Wong, C.-H. Bioorg. Med. Chem. Lett. 1998, 8, 2333. Kogan,
T. P.; Dupre, B.; Bui, H.; McAbee, K. L.; Kassir, J. M.; Scott,
I. L.; Hu, X.; Vanderslice, P.; Beck, P. J.; Dixon, R. A. F. J.
Med. Chem. 1998, 41, 1099.
19
20
6.7
1.6
15. Toepfer, A.; Kretzschmar, G.; Schuth, S.; Sonnentag, M.
Bioorg. Med. Chem. Lett. 1997, 7, 1311.
16. Wong, C.-H.; Moris-Varas, F.; Hung, S.-C.; Marron,
T. G.; Lin, C.-C.; Gong, K. W.; Weitz-Schmidt, G. J. Am.
Chem. Soc. 1997, 119, 8152 and references cited therein.
17. Dupre, B.; Bui, H.; Scott, I. L.; Market, R. V.; Keller,
K. M.; Beck, P. J.; Kogan, T. P. Bioorg. Med. Chem. Lett.
1996, 6, 569.
21
22
23
NA at 10 mM
8
18. Marron, T. G.; Woltering, T. J.; Weitz-Schmidt, G.;
Wong, C.-H. Tetrahedron Lett. 1996, 37, 9037.
19. Murphy, P. V.; Hubbard, R. E.; Manallack, D. T.; Wills,
R. E.; Montana, J. G.; Taylor, R. J. K. Bioorg. Med. Chem.
1998, 6, 2421. Hanessian, S.; Reddy, G. V.; Huynh, H. K.;
Pan, J.; Pedatella, S. Bioorg. Med. Chem. Lett. 1997, 7, 2729.
20. Toepfer, A.; Kretzschmar, G.; Schuth, S.; Sonnentag, M.
Bioorg. Med. Chem. Lett. 1997, 7, 1317. Cappi, M. W.; Moree,
W. J.; Qiao, L.; Marron, T. G.; Weitz-Schmidt, G.; Wong, C.-
H. Bioorg. Med. Chem. 1997, 5, 283.
NA at 10 mM
24
NA at 10 mM
21. Simanek, E. E.; McGarvey, G. J.; Jablonowski, J. A.; Wong,
C.-H. Chem. Rev. 1998, 98, 833 and references cited therein.
22. Kaila, N.; Yu, H.-A.; Xiang, Y. Tetrahedron Lett. 1995,
36, 5503.
23. Poppe, L.; Brown, G. S.; Philo, J. S.; Nikrad, P. V.; Shah,
B. H. J. Am. Chem. Soc. 1997, 119, 1727 and references cited
therein.
groups have reported a new hydrophobic binding site
on the selectins.16,36 Further e€orts are in progress to
explore the new hydrophobic area in the sLex-binding
site and optimize the potency of this series of molecules.
24. Graves, B. J.; Crowther, R. L.; Chandran, C.; Rumberger,
J. M.; Li, S.; Huang, K.-S.; Presky, D. H.; Familletti, P. C.;
Wolitzky, B. A.; Burns, D. K. Nature 1994, 367, 532.
25. Sche‚er, K.; Ernst, B.; Katopodis, A.; Magnani, J. L.;
Wang, W. T.; Weisemann, R.; Peters, T. Angew. Chem., Int.
Ed. Engl. 1995, 34, 1841.
26. Kogan, T. P.; Dupre, B.; Keller, K. M.; Scott, I. L.; Bui,
H.; Market, R. V.; Beck, P. J.; Voytus, J. A.; Revelle, B. M.;
Scott, D. J. Med. Chem. 1995, 38, 4976.
27. Cooke, R. M.; Hale, R. S.; Lister, S. G.; Shah, G.; Weir,
M. P. Biochemistry 1994, 33, 10591.
28. Hensley, P.; McDevitt, P. J.; Brooks, I.; Trill, J. J.; Field,
J. A.; McNulty, D. E.; Connor, J. R.; Griswold, D. E.;
Kumar, V. N.; Kopple, K. D.; Carr, S. A.; Dalton, B. J.;
Johanson, K. J. Biol. Chem. 1994, 269, 23949.
References and Notes
1. Ebnet, K.; Vestweber, D. Histochem. Cell Biol. 1999, 112,
1. Wong, C.-H. Acc. Chem. Res. 1999, 32, 376. Sears, P.;
Wong, C.-H. Angew. Chem., Int. Ed. 1999, 38, 2301.
2. McGarvey, G. J.; Wong, C.-H. Liebigs. Ann./Recl. 1997,
1059. Cooling, L. L. W.; Zhang, D.-S.; Koerner, T. A. W.
Trends Glycosci. Glycotechnol. 1997, 9, 191.
3. Sears, P.; Wong, C.-H. Proc. Natl. Acad. Sci. U.S.A. 1996,
93, 12086.
4. Kiessling, L. L.; Pohl, N. L. Chem. Biol. 1995, 3, 71. Lasky,
L. A. Annu. Rev. Biochem. 1995, 64, 113.
5. Bertozzi, C. R. Chem. Biol. 1995, 2, 703. Giannis, A.
Angew. Che. Int. Ed. Engl. 1994, 33, 178.
6. Rosen, S. D.; Bertozzi, C. R. Curr. Opin. Cell Biol. 1994, 6,
663.
7. Springer, T. A. Cell 1994, 76, 301. Lowe, J. B. In The
Handbook of Immunopharmacology: Adhesion Molecules;
Wegner, C. D., Ed.; Academic: London, 1994; Chapter 6 and
references cited therein.
8. Ramphal, J. Y.; Zheng, Z.-L.; Perez, C.; Walker, L.;
Defrees, S. A.; Gaeta, F. C. A. J. Med. Chem. 1994, 37, 3459.
9. Stahl, W.; Sprengard, U.; Kretzschmar, G.; Kunz, H.
Angew. Chem., Int. Ed. Engl. 1994, 33, 2096.
29. McMartin, C.; Bohacek, R. S. J. Comp.-Aided Mol. Des.
1997, 11, 333.
30. Allevi, P.; Ciu€reda, P.; Colombo, D. J. Chem. Soc., Perkin.
Trans. 1 1989, 1281. The second step in the synthesis works at
0 ꢀC and not at room temperature as suggested in the reference.
31. Final compounds were puri®ed on silica gel using ethyl acet-
ate:methanol:acetonitrile:water: 6/1.5/1.5/1.5. All compounds
gave spectroscopic data consistent with the proposed struc-
tures. Satisfactory HRMS data (EI) was obtained for all the
®nal products.
32. The acid groups in amines 3, 5, 6, and 10 were protected
as benzyl esters, which were removed in the ®nal hydro-
genolysis step. Amine 8 was used as a t-butyl ester; in this case
10. Kogan, T. P.; Revelle, B. M.; Tapp, S.; Scott, D.; Beck,
P. J. J. Biol. Chem. 1995, 270, 14047.

N. Kaila et al. / Bioorg. Med. Chem. Lett. 11 (2001) 151±155
155
an additional deprotection step using TFA/CH2Cl2/water was
performed before hydrogenolysis. Amines 7, 9, and 11 were
used as methyl esters; in these cases an additional deprotection
step using NaOH/THF/MeOH was performed before hydro-
genolysis.
ni®cantly di€erent from b-mannoside 3 (IC50=7 mM) in our
assays.
35. Hung, S.-C.; Lin, C.-C.; Wong, C.-H. Tetrahedron Lett.
1997, 38, 5419. Angibeaud, P.; Utille, J.-P. Synthesis 1991,
737.
33. Mannoside 5 showed the same activity as 3 but was not
stable for long periods of time.
34. The activity of a-mannoside 2 (IC50=3 mM) was not sig-
36. Ramphal, J. Y.; Hiroshige, M.; Lou, B.; Gaudino, J. J.;
Hayashi, M.; Chen, S. M.; Chiang, L. C.; Gaeta, F. C. A.;
DeFrees, S. A. J. Med. Chem. 1996, 39, 1357.