Design and synthesis of potent and selective inhibitors of integrin VLA-4

Article abstract of DOI:10.1016/S0960-894X(01)00586-8

The synthesis and identification of a novel series of inhibitors of integrin VLA-4 are described. Their in vitro activity and selectivity against closely related integrins are also presented.

Full text of DOI:10.1016/S0960-894X(01)00586-8

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2955–2958
Design and Synthesis of Potent and Selective Inhibitors of
Integrin VLA-4
Sompong Wattanasin,* Beat Weidmann, Didier Roche, Stewart Myers, Amy Xing,
Qin Guo, Michael Sabio, Peter von Matt, Ronald Hugo, Susan Maida, Philip Lake
and Marla Weetall
Novartis Institute for Biomedical Research, Novartis Pharmaceuticals Corporation, 556 Morris Avenue, Summit, NJ 07901, USA
Received 16 July 2001; accepted 28 August 2001
Abstract—The synthesis and identification of a novel series of inhibitors of integrin VLA-4 are described. Their in vitro activity and
selectivity against closely related integrins are also presented. # 2001 Elsevier Science Ltd. All rights reserved.
Antagonism of cell surface receptors belonging to the
integrin superfamily is a new paradigm for drug dis-
covery in transplantation and cancer, as well as in other
diseases.1 Integrin VLA-4 is an attractive therapeutic
target: blockade of integrin VLA-4 by monoclonal
antibodies (mAbs) has shown clear-cut efficacy in ani-
mal models of autoimmune disease, asthma, inflamma-
tion, and acute and chronic rejection of allogeneic
grafts.2,3 VLA-4 antagonists act to block leukocyte
migration as well as leukocyte activation on the cell
types thought to mediate these diseases. VLA-4 is
expressed on memory T cells, memory B cells, NK cells,
monocytes and macrophages. Although orally active
antagonists of the platelet integrin receptor are in
advanced development,1d a lack of similar progress is
evident in the discovery of orally active low molecular
weight (LMW) antagonists of the important leukocyte
integrins such as VLA-4, despite their validation as tar-
gets by mAbs. Because of the importance of this integ-
rin and its implication in various disease processes,
efforts toward the design and synthesis of orally bioa-
vailable LMW inhibitors have intensified in recent
years.4
potent inhibition in the low nanomolar range was
achieved with the progenitor members of the series,
such as 1 and 2,5 concerns with their peptidic features
led us to make various structural modifications (Scheme
1). We have successfully generated antagonists that
retain nM potency, but are structurally simpler than the
original leads.
Scheme 2 outlines the general approach to aza analo-
gues 3–6. Thus, t-butyl carbazate was reductively alky-
lated to give 15. Coupling of 15 with the isocyanate 16,
which was derived from racemic ethyl 3-amino-3-phe-
nylpropanoate, gave 17. Deprotection of the t-BOC
group, acylation with the bisarylurea acid 185c and
hydrolysis of the ethyl ester finally resulted in the aza
analogue 3. Aza analogues 4 and 5 were similarly pre-
pared from racemic ethyl 3-amino-4-hexenoate and
ethyl valine, respectively. The aza analogue 6 was pre-
pared as shown in Scheme 2. Sequential alkylation of t-
butyl carbazate with ethyl bromoacetate and allyl bro-
mide gave 19. Removal of the t-BOC protecting group
followed by coupling with N-t-BOC-l-leucine and
deprotection afforded 20. Acylation with the bisarylurea
acid 18 and hydrolysis of the ester gave 6.
Herein, we describe the identification of potent and
selective inhibitors of VLA-4 based on the known
bisarylurea series5 of LDV peptidomimetics. Although
The general route to b-aminoacid analogues (7–10) is
illustrated by the synthesis of 7 in Scheme 3. Con-
densation of 18 with 1,1-dimethylethyl (3R)-3-amino-5-
methylhexanoate followed by removal of the t-butyl
ester with trifluoroacetic acid provided 21. Standard
coupling of the acid 21 with 1,1-dimethylethyl (3S)-3-
*Corresponding author. Fax: +1-908-277-4885; e-mail: sompong.
wattanasin@pharma.novartis.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00586-8

2956
S. Wattanasin et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2955–2958
amino-3-phenylpropanoate and deprotection of the
t-butyl ester gave 7. The epimer 8 and other analogues
(9 and 10) were analogously prepared in high yields.
22, which was coupled with N-t-BOC-l-leucine
hydroxysuccinate ester to give 23, after removal of the t-
BOC group. Acylation of 23 with bromoacetyl bromide
and reaction with isoamylamine gave 24. Acylation of
23 and 24 with 18 followed by ester hydrolysis gave 12
and 13, respectively. Acylation of the b-aminoacid ester
25 with bromoacetyl bromide followed by alkylation
with an amine gave 26. Thus, compounds 14 and 27
were prepared in high yields from commercially avail-
able 1,1-dimethylethyl (3S)-3-amino-3-phenylpropano-
ate and 1,1-dimethylethyl (3S)-3-amino-hexenoate via
the intermediate 26. Analogues 28–33 were similarly
synthesized from the ester of b-aminoacid 256 in good
yields.
Peptoid analogue 11 was also prepared according to
Scheme 3 by coupling N-benzylglycine ethyl ester with
21 (DIEA, EDC, CH2Cl2, rt, 48 h) followed by hydro-
lysis (LiOH, aq THF, rt, 2 h) and acidification. Other
analogues were prepared as outlined in Scheme 4.
Michael reaction of phenethylamine with ethyl acrylate
in methanol at room temperature for 4 h cleanly gave
Scheme 4. (a) N-t-BOC-l-leucine hydroxysuccinate ester, CH2Cl2,
reflux, 8 h; (b) HCl gas, Et2O, 0oC to rt, 1 h; (c) bromoacetyl bromide,
Et3N, CH2Cl2, 0 oC to rt, 1 h, 86–97%; (d) the amine (isoamylamine
or 3-methoxypropylamine or 4-methoxybutylamine), Et3N, DMF, rt,
1 h, 70–82%; (e) 18, DIEA, EDC, CH2Cl2, rt, 12 h, 67–97% then
LiOH, THF, H2O, rt, 2–3 h and HCl acidification, 95–99% (f) 18,
DIEA, EDC, CH2Cl2, rt, 12 h, 67–97% then TFA, CH2Cl2, rt, 2–3 h,
95–99%.
Scheme 1. Structural modifications of VLA-4 inhibitors.
Compounds were evaluated for VLA-4 inhibitory
activity in a cell adhesion assay5c,7 that is based on the
binding of fluorescently labeled Ramos cells to immo-
bilized VCAM-1. IC50 values are listed in Table 1. The
aza analogues of the leucine moiety, 3 (IC50=2900 nM)
and 4 (IC50=5900 nM), were about 175 and 350 times
less active than the known inhibitor 2 (IC50=16.5 nM),
when tested side-by-side in the same assay. Replacement
of the b-aminoacid moiety with an a-aminoacid such as
Table 1. In vitro VLA-4 cell adhesion inhibitory activity (VCAM-1/
Ramos cells)7
Scheme 2. (a) Isobutyraldehyde, THF, 24 h; (b) H2, platinum/carbon,
MeOH, rt, 12 h, 74%; (c) HCl gas, 0.5 h, then phosgene, rt, 12 h; (d)
CH2Cl2, rt, 12 h, 54%; (e) HCl gas, Et2O, 0 oC to rt, 1 h, quant. (f) 18,
DIEA, EDC, CH2Cl2, rt, 12 h, 34–57%; (g) LiOH, THF, H2O, rt, 12
h, quant. (h) ethyl bromoacetate, Et3N, DMF, rt, 12 h, 24%; (i) allyl
bromide, Et3N, DMF, rt, 12 h, 90%; (j) DIEA, EDC, CH2Cl2, rt, 12
h, 39%.
Compd
IC50 (nM)a
Compd
IC50 (nM)a
1
2
3
4
5
6
7
8
3b
16.5
2900
5900
12
13
14
27
28
29
30
31
32
33
34c
37,000
98
29
45
15.4
20
11.7
4.7
7.1
1.3
>500,000
11,000
7300
262,000
103,000
6400
9
10
11
247,000
75,000
aEach IC50 value is an average of 2–5 determinations run in triplicate.
bReported IC50 data.5
cCompound 34 is the ethyl ester of 31.
Scheme 3. (a) 18, DIEA, EDC, CH2Cl2, rt, 12 h, 97%; (b) TFA,
CH2Cl2, thioanisole, rt, 20 min, 90%; (c) DIEA, EDC, CH2Cl2, rt, 12
h, quant.

S. Wattanasin et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2955–2958
2957
Table 2. Selectivity of selected inhibitors against related integrins
(IC50 values in nM)a
of the phenyl and isoamyl moieties in 14 provided a
significant (4- to 6-fold) improvement in potency as in
the case of compound 31 (IC50=4.7 nM) and 32
(IC50=7.1 nM). The methoxybutyl analogue 33
(IC50=1.3 nM) showed a slight improvement in activ-
ity. The (S)-absolute configuration of the b-aminoacid
chiral center is important for potency. For example, the
opposite (R)-enantiomer (IC50=630 nM) of 28 was
more than 40-fold less active than the (S)-isomer 28. As
expected from other LDV mimics, the corresponding
ethyl ester of 31 (34) was inactive (IC50=75,000 nM).
Integrin/ligand
28
30
31
32
VLA-4/VCAM-1
VLA-4/FN
VLA-4(mouse)/VCAM-1
VLA-4(rat)/VCAM-1
a4b7/VCAM-1
15.4
7.9
7.7
20.5
11.7
14.4
21.7
24.5
4.7
36
51.9
27.2
6,140
7.1
13.4
27.3
21.6
>100,000 >100,000
>100,000
VLA-5/FN
>500,000 >500,000 >500,000 >500,000
Mac-1 (mouse)/ICAM-1 >500,000 >500,000 >500,000 >500,000
aVb3+aVb5/VN 59,000 >100,000 >100,000 >100,000
aValues represent the average of at least four experiments with side-by-
side comparisons.
Additional studies were carried out with highly potent
compounds such as 28, 30, 31, and 32 to assess their
selectivity against other closely related integrins5c,7 and
the results are summarized in Table 2. VLA-4 binds
either VCAM-1 or the CS-1 domain of fibronectin9
(FN). Therefore, it was necessary to assess the ability of
these compounds to block VLA-4 dependent cellular
binding to FN as well as to VCAM-1. All compounds
potently blocked VLA-4 binding to FN. Cross-reactiv-
ity with murine or rat VLA-4 was assessed using cell
adhesion assays based on the binding of murine or rat
cells to immobilized human VCAM-1. The compounds
effectively block mouse and rat VLA-4-mediated cell
binding to VCAM-1 so that these compounds could be
used in rodent models of disease. Compound 31 blocked
mouse VLA-4 about 11-fold less effectively than human
VLA-4, suggesting that studies in the mouse may
underestimate the activity of the compound in humans.
valine as in 5 resulted in an inactive compound
(IC50>500,000 nM). The aza analogue of the b-ami-
noacid moiety of 2 as in 6 (IC50=10,100 nM) was less
potent in comparison with 3 and 4. The b-aminoacid
analogues such as 7, 8, and 9, in which the a-leucine
moiety of 2 was replaced with the corresponding b-leu-
cine residue, also exhibited very poor activity. Com-
pound 7 (IC50=7300 nM), which has the same absolute
configuration of the (S)-leucine side chain as in 2, how-
ever, was about 14- to 36-fold more active than the (R)-
b-leucine containing compounds 8 (IC50=262,000 nM)
and 9 (IC50=103,000 nM). An attempt to improve the
activity of 7 by introducing the a-aminoacid phenylgly-
cine to compensate for an extra carbon on the b-leucine
moiety led to the synthesis of 10 (IC50=6400 nM), but
no improvement in activity was observed. Further loss
in activity was also observed when the b-phenylglycine
unit of 7 was replaced with a peptoid mimic as in 11
(IC50=247,000 nM). A hybrid containing an a-leucine
and a peptoid as in 12 (IC50=37,000 nM) also did not
significantly improve potency. The peptoid–peptoid
hybrid 13 (IC50=98 nM), however, showed more than a
370-fold improvement in comparison with 12. A further
improvement in potency was finally obtained with 14,
the leucine peptoid and the b-aminoacid hybrid, with an
IC50 value of 29 nM. This simple inhibitor containing
only one chiral center, which is one chiral center less
than the known starting compounds such as 1 and 2,
became the lead for further optimization.
Selectivity of these compounds for VLA-4 against other
integrins was demonstrated using cell adhesion assays
for integrins a4b7, VLA-5, aVb3 and aVb5, and Mac-1.
They were found to be selective for integrin VLA-4.
Compound 31 also weakly blocked integrin a4b7
(IC50=6140 nM), but more than 1300-fold less effec-
tively than integrin VLA-4 (IC50=4.7 nM). Thus, the
selectivity of these compounds against the homologous
integrin a4b7 are far much better than the bisarylurea
series of peptide analogues.5c Compound 28 reacted
weakly with aV-integrins, but was about 3800-fold more
selective for VLA-4. Other compounds did not block
integrin a4b7 or aV-integrins at concentrations up to
100 mM. No inhibition of VLA-5- or Mac-1-dependent
adhesion was observed with these compounds at
concentrations up to 500 mM.
Earlier work on LDV peptidomimetics suggested that
an ether oxygen atom in the leucine mimic chain of 14
could have a profound effect in increasing inhibitory
activity.8 In addition, known SAR5 on analogues of 1
and 2 indicated that the phenyl chain on the b-phe-
nylglycine moiety is not optimal. In order to demon-
strate that the ether-containing chain is important for
the potency improvement of 14 and to find a better
replacement of the phenyl group, analogues 27–33 were
prepared. There was no improvement in potency with
the propenyl compound 27 (IC50=45 nM). A slight
improvement in potency was observed with the 3,4-
dimethoxyphenyl analogue of 14, 29 (IC50=20 nM).
The methoxypropyl replacement of the leucine mimic
chain of 14 and 27 as in 30 (IC50=11.7 nM) and 28
(IC50=15.4 nM), respectively, also provided a slight
improvement in potency. However, replacement of both
In conclusion, we have identified a new structural class
of non-peptidic compounds as potent and selective
inhibitors of VLA-4. The potent compounds reported
herein provide not only useful leads for further devel-
opment, but also an important tool for evaluating the
role of VLA-4 blockade by LMW compounds in
inflammation and immune responses in animal models.
The results of their pharmacokinetic and pharmacological
studies will be forthcoming.
Acknowledgements
We would like to thank Professor Robert Boeckmann,
Jr. for helpful discussions.

2958
S. Wattanasin et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2955–2958
95, 2601. (c) Sadahiro, M.; McDonald, T. O.; Allen, M. D.
References and Notes
Am. J. Pathol. 1995, 142, 675. (d) Kling, D.; Fingerle, J.;
Harlan, J. M.; Lobb, R. R.; Lang, F. Circ. Res. 1995, 77, 1121.
4. (a) Zimmerman, C. N. Exp. Opin. Ther. Patents 1999, 129.
(b) Tilley, J. W.; Kaplan, G.; Rowan, K.; Schwinge, V.;
Wolitzky, B. Bioorg. Med. Chem. Lett. 2001, 11, 1. (c) Sourers,
A.; Virgilio, A.; Shrurer, S.; Ellman, J. Bioorg. Med. Chem.
Lett. 1998, 8, 2297.
5. (a) Lin, K.; Zimmerman, C. N.; Castro, A.; Lee, W.;
Hammond, C. E.; Ateeq, H. S.; Hsiung, S. H.; Chong, L. T.;
Liao, Y.; Adams, S. P. Book of Abstract, 213th ACS National
Meeting, San Francisco, 1997; MEDI-315. (b) Lobb, R. R.;
Adams, S. P. Expert Opin. Invest. Drugs 1999, 8, 935. (c) Lin,
K.; Ateeq, H. S.; Hsiung, S. H.; Chong, L. T.; Zimmerman,
C. N.; Castro, A.; Lee, W.; Hammond, C. E.; Kalkunte, S.;
Chen, L.; Pepinsky, R. B.; Leone, D. R.; Sprague, A. G.;
Abraham, W. M.; Gill, A.; Lobb, R. R.; Adams, S. P. J. Med.
Chem. 1999, 42, 920.
6. For non-commercially available chiral b-aminoacid esters
for the synthesis of analogues 29, 31–33, they were synthesized
in high overall yield and optical purity by stereoselective con-
jugate addition of the chiral lithium amide to the cinnamic
acid ester followed by hydrogenolysis according to Davies, S.
G.; Ichihara. O. Tetrahedron: Asymmetry 1991, 2, 183.
7. Weetall, M.; Hugo, R.; Friedman, C.; Maida, S.; West, S.;
Wattanasin, S.; Bouhel, R.; Weitz-Schmidt, G.; Lake, P. Anal.
Biochem. 2001, 293, 277.
1. For recent reviews: (a) Lobb, R.; Hemler, M. J. Clin. Invest.
1994, 94, 1722. (b) Heavner, G. A. Drug Discovery Today
1996, 1, 295. (c) Mousa, S. A.; Cheresh, D. A. Drug Discovery
Today 1997, 2, 187. (d) Bochner, B. S. In Cell Adhes. Mol.
Matrix Proteins; Mousa, S. A., Ed.; Springer: Berlin, 1998; pp
113–131. (e) Elices, M. J. Curr. Opin. Anti-Inflammat. Imuno-
modulat. Invest. Drugs 1999, 1, 14. (f) Mousa, S. A. Exp. Opin.
Ther. Patents 1999, 9, 1237. (g) Adams, S. P., Lobb, R. R.
Ann. Rep. Med. Chem.; Bristol, J., Ed.; Academic: New York,
1999; Vol. 34, pp 179–188.
2. For references on the role of VLA-4 in autoimmune dis-
ease, inflammation, and asthma: (a) Barbadillo, C.; G-Arroyo,
A.; Salas, C.; Mulero, J.; Sanchez-Madrid, F.; Andreu, J. L.
Springer Semin. Immunopathol. 1995, 16, 427. (b) Abraham,
W. M.; Gill, A.; Ahmed, A.; Sielczak, M. W.; Lauredo, I. T.;
Botinnikova, Y.; Lin, K. C.; Pepinsky, B.; Leone, D. R.;
Lobb, R. R.; Adams, S. P. Am. J. Respir. Crit. Care Med.
2000, 162, 603. (c) Abraham, W. M.; Ahmed, A.; Sielczak,
M. W.; Narita, M.; Arrhenius, T.; Elices, M. J. Am. J. Respir.
Crit. Care Med. 1997, 156, 696. (d) Dutta, A. S.; Gormley,
J. J.; Coath, M.; Hassall, L.; Hayward, C. F.; Gellert, P. R.;
Kittlety, R. S.; Alcock, P. J.; Ferguson, R.; Halterman, T.;
Jamieson, A.; Moors, J. A.; Moores, J. M.; Rees, A.; Wood,
L. J.; Reilly, C. F.; Haworth, D. J. Pept. Sci. 2000, 6, 398.
3. For references on the role of VLA-4 in acute and chronic
rejection: (a) Isobe, M.; Suzuki, J.; Yagita, H.; Okumura, K.;
Yamazaki, S.; Nagai, R.; Yazaki, Y.; Sekiguchi, M. J. Immu-
nol. 1994, 153, 5810. (b) Molossi, S.; Elices, M.; Arrhenius, T.;
Diaz, R.; Coulber, C.; Rabinovitch, M. J. Clin Invest. 1995,
8. Sabio, M., Fotsch, C., von Matt, P. Novartis Pharmaceuticals
Corporation. Unpublished results.
9. (a) Guan, J.-L.; Hynes, R. Cell 1990, 60, 53. (b) Komoriya,
A.; Green, L. J.; Mervic, M.; Yamada, S.; Yamada, K. M.;
Humphries, M. J. Biol. Chem. 1991, 266, 15075.