Imide and lactam derivatives of N-benzylpyroglutamyl-L-phenylalanine as VCAM/VLA-4 antagonists

Article abstract of DOI:10.1016/S0960-894X(00)00582-5

A series of imides and lactams derived from 4-amino-N-benzylpyroglutamyl-L-phenylalanine was prepared and evaluated for activity as VCAM/VLA-4 antagonists. Imides were more potent than the corresponding lactams; several had subnanomolar IC₅₀s in an ELISA based assay and were also highly effective at blocking VLA-4 expressing Ramos cell binding to VCAM coated plates. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00582-5

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1±4
Imide and Lactam Derivatives of N-Benzylpyroglutamyl-
L-phenylalanine as VCAM/VLA-4 Antagonists
Je€erson W. Tilley,* Gerry Kaplan, Karen Rowan, Virginia Schwinge
and Barry Wolitzky
Roche Research Center, Ho€mann-La Roche Inc., 340 Kingsland Street, Nutley, NJ 07110, USA
Received 21 August 2000; accepted 18 September 2000
AbstractÐA series of imides and lactams derived from 4-amino-N-benzylpyroglutamyl-l-phenylalanine was prepared and
evaluated for activity as VCAM/VLA-4 antagonists. Imides were more potent than the corresponding lactams; several had
subnanomolar IC50s in an ELISA based assay and were also highly e€ective at blocking VLA-4 expressing Ramos cell binding to
VCAM coated plates. # 2000 Elsevier Science Ltd. All rights reserved.
The integrin VLA-4, a4-b1, is expressed on a variety of
lympocytes including B-cells, T-cells, basophils and
eosinophils and is involved in modulating tracking,
activation and survival of these cell types.1 Because of
its potentially key role in regulating the in¯ammatory
process, there is currently a great deal of interest in dis-
covering compounds capable of interfering with the
interaction of VLA-4 with its cognate receptor, VCAM.
Such compounds are being considered for clinical
trials in asthma,2 multiple sclerosis,3 and rheumatoid
arthritis.4
carbonyl would prefer to be out of conjugation, rather
than co-planar as demanded by the proposed cyclic
structures, our limited understanding of the details of
the SAR of phenylalanine derived VCAM/VLA-4
antagonists prompted us to carry out the investigation
described below. In this work we show that appropriate
imides 2 are nearly equipotent with the amide 1 whereas
the corresponding lactam derivatives 3 are several-fold
less potent.
We have previously reported the identi®cation of a ser-
ies of N-benzylpyroglutamyl 4-substituted-l-phenylala-
nine analogues that showed good potency as VCAM/
VLA-4 antagonists in both ELISA and cell based assay
formats.5 Among the most potent compounds were
those substituted by a benzoylamino group which
incorporated both an ortho substituent and at least one
electron withdrawing group such as 1. As part of an
e€ort to improve the pharmacokinetics of this class of
molecule, we sought to investigate whether elimination
of the para-amide NH- through incorporation of the
nitrogen atom into a cyclic imide or lactam ring to give
2 or 3, respectively, would be feasible.
The preparation of the imide derivatives is shown in
Scheme 1 starting from Na-Boc-4-Fmoc-aminophenyl-
alanine 4. Benzyl ester formation, cleavage of the Boc
group, coupling with N-benzyl pyroglutamic acid and
cleavage of the Fmoc group proceeded without incident
to give the amine 5 despite the low solubility of the
Fmoc containing intermediates. Formation of the imi-
des was generally e€ected by treatment of 5 with an
excess of the appropriate cyclic anhydride in the pre-
sence of excess carbonyl diimidazole. After stirring
overnight, the reaction mixtures were concentrated, the
residues dissolved in acetic acid and applied directly to a
C-18 RP-HPLC column eluting with acetonitrile±water.
While we recognized that the requirement for ortho-
substituents in compounds related to 1 suggested that
the p-system of the aromatic ring and the amide
*Corresponding author. Tel.: +1-973-235-4660; fax: +1-973-235-
6084; e-mail: je€erson.tilley@roche.com
0960-894X/01/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00582-5

2
J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1±4
The phthalide 9 was readily prepared by reductive
amination of the amine 5 with 2-carbomethoxy-
benzaldehyde followed by heating the intermediate
amino ester in acetic acid to e€ect ring closure and cat-
alytic hydrogenolysis of the benzyl group as shown in
Scheme 3. Re¯uxing a methylene chloride solution of
the amine 5 and the ester 10 led through a nucleophilic
ring opening and subsequent acylation to the carboxylic
acid 11 (81%). Hydrogenolysis followed by thermal
decarboxylation in DMSO at 100 ꢀC then a€orded the
lactam 12 in 62% yield for the two steps (Scheme 4).
Scheme 1. (a) Benzyl-Br, KHCO3, DMF, 18 h, 84%; (b) 4 N HCl,
dioxane, 98%; (c) N-benzyl-l-pyroglutamic acid, HBTU, DIPEA,
DMF, 95%; (d) diethylamine, DMF, 64%; (e) cyclic anhydride
(3.4 equiv), carbonyl diimidazole (3 equiv), CH2Cl2; (f) carbonyl
diimidazole, CH2Cl2 (if necessary); (g) H2, Pd(C).
Scheme 3. (a) NaHBCN, HOAc, 4 A mol sieve, EtOH; (b) HOAc,
re¯ux; (c) H2, Pd(C).
In the case of 2d and 2f, we obtained the intermediate
amido acid 6 from this process and the carbonyl diimi-
dazole treatment was repeated.
Finally, the benzyl ester was cleaved by catalytic
hydrogenation. Typically, yields for the 2- to 3-step
process were around 50%. For the preparation of
the pyridinyl derivative 2b, this process resulted in the
partial hydrogenation of the pyridine ring. As an alter-
native, imide formation was carried out directly from
the amino acid 7 by treatment with an excess of pyridine
3,4-dicarboxylic acid anhydride followed by carbonyl
diimidazole in DMF. For the 4-hydroxyphenyl suc-
cinimide 2e, we started with acetoxyphenyl succinic
anhydride and the acetoxy group was lost during puri-
®cation. In the case of the isobutyl succinimide 2h, we
started with the corresponding methylpropenylsuccinnic
anhydride and saturated the double bond during ester
cleavage.
Scheme 4. (a) CH2Cl2, re¯ux, 4 h (81%); (b) H2, Pd(C) (77%);
(c) DMSO, 100 ꢀC, 12 h (81%).
Results and Discussion
Compounds were assayed for VLA-4 antagonist activity
using a solid-phase, dual antibody ELISA in which
VLA-4 derived from Ramos cells was allowed to com-
pete for bound recombinant human VCAM in the pre-
sence of serial dilutions of test compound. VLA-4
Scheme 2. (a) Pyridine 3,4-dicarboxylic acid anhydride, THF; (b) car-
bonyl diimidazole, DMF; (c) water.

J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1±4
3
bound to VCAM-1 was detected by a complex of anti-
b1 antibody and HRP-conjugated anti-mouse IgG:
chromogenic substrate (K-Blue).6 The data listed in
Table 1 indicate that several compounds were potent
inhibitors of the VCAM/VLA-4 interaction. In parti-
cular, the phthalimide 2a, and the succinimide deriva-
tives 2h and 2j each had subnanomolar activity and
were only slightly less potent than 1.
Table 1. VCAM/VLA-4 binding inhibition of N-benzylpyro-
glutamyl-l-phenylalanine derivatives
These compounds were further evaluated for their abil-
ity to block the interaction between ¯uorescently labeled
Ramos cells, which express VLA-4, with VCAM coated
microtiter plates. The rank order of potencies is the
same as obtained in the ELISA assay and is consistent with
our previous observation5 that only compounds with sub-
nanomolar potency in our ELISA assay show double digit
nanomolar potency in the more demanding cell based
assay. Again, the three most potent compounds were 2a,
2h, and 2j and were only 2- to 5-fold less potent than 1.
No
R
ELISA (IC50 nM) Ramos Cell (IC50 nM)
1
0.37
12
2a
0.52
4.3
20
650
107
92
For comparison, the lactam derivatives 3a and 3i were
prepared. As the data in Table 1 indicate, each was
approximately 5-fold less potent than the corresponding
imide in the ELISA assay. The more potent of the two
(3a) was also evaluated in the Ramos cell assay and, as
expected, found to have only weak activity.
2b
2c
0.98
1.7
The relatively potent activity observed with members of
the imide class is somewhat surprising since they are
unable to adopt the same conformation as members of
the prototype amide series such as 1. As previously
noted, activity in the later is favored by the presence of
at least one ortho-substituent on an aromatic ring bear-
ing at least one electron withdrawing group. The pre-
sence of the ortho-substituent on amides related to 1
should favor a conformation in which the aromatic p-
system and the amide carbonyl are not fully conjugated.
This is in marked contrast to the planar phthalimide 2a.
It is possible that the presence of the second carbonyl of
the imide causes the entire phthalimido group to rotate
with respect to the phenylalanine aromatic ring thus
providing a comparable positioning of the phthalimido
p-system to the aromatic p-system of 1. In this case, the
phthalide 3a lacking the buttressing e€ect of the second
carbonyl would be expected to be less active as observed.
2da
2ea
2fa
2.6
420
123
0.94
2ga
0.99
184
2ha
0.76
3.2
41
286
57
While such arguments could account for the results
obtained with the phthalimide 2a, they do not explain
the activity of the aliphatic imides, particularly 2i and
2j, which lack a comparable p-system. We previously
speculated that the amide carbonyl of 1 may be acting
as a hydrogen bond acceptor.5 It is possible that the
second carbonyl group of the imides is also involved
in a binding or favorable solvent interaction thus
compensating for the absence of a favorably oriented
aromatic p-system.
2i
2j
0.54
2.7
3a
3i
500
±
It is also possible that subtle adjustments in the protein
quaternary structure together with di€erences in the
precise orientation of the ligand result in a di€erent
binding mode for members of the imide class of VCAM-
4/VLA-4 antagonists. While further e€ort will be
required in order to more fully understand the SAR of
16
aMixture of R- and S-forms at the side chain chiral center.

4
J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1±4
the imides, the present work suggests that they o€er an
interesting new opportunity for the discovery of novel
VCAM-4/VLA-4 antagonists.
3. (a) Tubridy, N.; Behan, P. O.; Capildeo, R.; Chaudhuri, A.;
Forbes, R.; Hawkins, C. P.; Hughes, R. A. C.; Palace, J.;
Sharrack, B.; Swingler, R.; Young, C.; Moseley, I. F.; Mac-
Manus, D. G.; Donoghue, S.; Miller, D. H. Neurology 1999,
53, 466. (b) Keszthelyi, E.; Karlik, S.; Hyduk, S.; Rice, A.;
Gordon, G.; Yednock, T.; Horner, H. Neurology 1996, 47,
1053.
References
4. Sei€ge, D. J. Rheumatology 1996, 23, 2086.
1. Elices, M. J. In Cell Adhes. Mol. Matrix Proteins; Mouse,
S. A., Ed.; Springer Verlag: Berlin, 1999; pp 133±147.
2. Lin, K.-C.; Ateeq, H. S.; Hsiung, S. H.; Chong, L. T.;
Zimmerman, C. N.; Castro, A.; Lee, W.-C.; Hammond, C. E.;
Kalkunte, S.; Chen, L.-L.; Pepinsky, R. B.; Leone, D. R.;
Sprague, A. G.; Abraham, W. M.; Gill, A.; Lobb, R. A.;
Adams, S. P. J. Med. Chem. 1999, 42, 920.
5. Chen, L.; Tilley, J. W.; Guthrie, R. W.; Mennona, F.;
Huang, T.-N.; Kaplan, G.; Trilles, R.; Miklowski, D.; Huby,
N.; Schwinge, V.; Wolitzky, B.; Rowan, K. Bioorg. Med.
Chem. Lett. 2000, 10, 729.
6. Chen, L.; Guthrie, R. W.; Huang, T.-N.; Hull, K. G.; Sid-
duri, A.; Tilley, J. W. WO 9910312, 1999. Chem. Abstr. 1999,
130, 196,952.