N-aryl-prolyl-dipeptides as potent antagonists of VLA-4.

Article abstract of DOI:10.1016/S0960-894X(02)00356-6

The design, synthesis, and biological evaluation of N-arylprolyl-dipeptide derivatives as small molecule VLA-4 antagonists is described. Potency against VLA-4 and alpha(4)beta(7) and rat pharmacokinetic evaluation revealed some advantages over the related N-(arylsulfonyl)-prolyl-dipeptide analogues.

Full text of DOI:10.1016/S0960-894X(02)00356-6

Bioorganic & Medicinal ChemistryLetters 12 (2002) 2205–2208
N-Aryl-prolyl-dipeptides as Potent Antagonists of VLA-4
Theodore M. Kamenecka,a,* Thomas Lanza, Jr.,a Stephen E. de Laszlo,a Bing Li,a
Ermengilda D. McCauley,b Gail Van Riper,b Linda A. Egger,c Usha Kidambi,c
Richard A. Mumford,b Sharon Tong,a Malcolm MacCoss,a John A. Schmidtb
and William K. Hagmanna
aDepartment of Medicinal Chemistry, Merck Research Laboratories, Rahway, NJ 07065, USA
bDepartment of Inflammation and Rheumatology Research, Merck Research Laboratories, Rahway, NJ 07065, USA
cDepartment of Pharmacology, Merck Research Laboratories, Rahway, NJ 07065, USA
Received 8 February2002; accepted 23 April 2002
Abstract—The design, synthesis, and biological evaluation of N-arylprolyl-dipeptide derivatives as small molecule VLA-4 antagonists
is described. Potencyagainst VLA-4 and a4b7 and rat pharmacokinetic evaluation revealed some advantages over the related
N-(arylsulfonyl)-prolyl-dipeptide analogues. # 2002 Elsevier Science Ltd. All rights reserved.
The cell adhesion molecule VLA-4 (verylate antigen,
a4b1, CD49d/CD29) is an integrin constitutively
expressed on circulating lymphocytes.1 VLA-4 interacts
with the alternativelyspliced CS-1 domain of fibronectin
in the extracellular matrix and with VCAM-1 (vascular
cell adhesion molecule-1) expressed on endothelial cells.
These ligands are upregulated in response to inflamma-
torycytokines and are expressed at sites of inflammation.
The binding of these ligands with VLA-4 is thought to
be required for cell trafficking, activation, and development
during normal and/or pathophysiological processes. As
such, VLA-4 antagonists maybe useful in the treatment
of diseases which are characterized bya prolonged
inflammatoryresponses where cell trafficking and acti-
vation are important such as asthma, atherosclerosis,
rheumatoid arthritis, inflammatorybowel disease, and
multiple sclerosis.2À4
roughly50- to 100-fold specificityfor VLA-4 over the
related a4b7 integrin.
In the search for new leads, we found that migration of
the sulfonyl residue from the proline nitrogen in 1 to the
a-carbon (as in A to B, Fig. 2) led to potent compounds
with better pharmacokinetic profiles.6 Furthermore, the
sulfonyl group was deemed not necessary, as direct
attachment of the aryl group to the a-carbon (as in B to
C) led to compounds that retained their potencyagainst
VLA-4 and also led to further improvements in PK.7 In
searching for the structural elements that contributed to
the poor PK characteristics of compounds like 1, the
N-sulfonyl group was proposed to be unnecessary for
potency(as in C to D) since a-aryl-cycloalkanoyl deriva-
tives (C) could be made which were potent. This sulfonyl
elimination would lead to a new class of N-aryl-prolyl
derivatives that might have better PK characteristics
and is the subject of this report.
Previously, N-arylsulfonylated dipeptide derivatives
have been reported to be potent VLA-4 antagonists.5
Typical of this class of compounds is derivative 1 (Fig.
1). Compounds of this structure class generallyhave
poor pharmacokinetic (PK) properties, characterized by
low oral bioavailability(F <5%) and high plasma
clearance rates (Clp>100 mL/min/kg. Theyalso show
The desired N-arylated prolines were constructed by
two different strategies outlined in Schemes 1 and 2.
(l)-proline could be N-arylated in the presence of
K2CO3 and CuI without loss of enantiomeric excess as
described in the literature.10 This procedure worked well
for iodobenzene as well as for 2-, 3- and 4-substituted
phenyl halides and provided the N-arylated carboxylic
acids upon workup. Standard peptide coupling with the
biphenylalanine aminoester using the HATU/HOAt
*Corresponding author. Fax: +1-858-202-5753; e-mail: theodore_
kamenecka@merck.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00356-6

2206
T. M. Kamenecka et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2205–2208
Scheme 2. (a) Aryl-halide, Pd2dba3, NaOtBu, BINAP, toluene, 80 ꢀC;
(b) TFA, CH2Cl2; (c) 2,6-dimethoxybiphenyl alanine tert-butyl ester,
HATU, HOAt, i-Pr2NEt, CH2Cl2.
Figure 1. Initial N-arylsulfonylated dipeptide lead.
method gave the dipeptides.11 Deprotection of the tert-
butyl ester with TFA provided the final products.
This method, however, failed to work for hetero-
aromatic halides such as 2-bromopyridine and 5-iodo-
pyrimidine. In the CuI method, a catalytic cycle is
proposed wherein the amino acid salt forms a chelate
with the CuI. A p-complex is formed with the aromatic
halide which facilitates intramolecular displacement of
the halide bythe amino acid nitrogen atom. The avidity
of pyridine and other aromatic amine containing hetero-
cycles for copper salts may disrupt this mechanism and
might explain whythese systems failed.
Fortunately, other methods for N-arylation are feasible
as described in Scheme 2. (l)-Proline tert-butyl ester was
coupled with pyridyl-like halides in the presence of
sodium tert-butoxide and a suitable palladium source to
afford the desired N-arylated prolines in good yield.12
Not surprisingly, under the basic reaction conditions,
complete loss of stereochemical integritywas observed
giving racemic products. Interestingly, attempted coupling
with proline methyl or benzyl ester gave much lower
yields of product if at all. Deprotection of the tert-butyl
ester with TFA, standard HATU coupling, and ester
hydrolysis afforded the desired products in good yield.
Since the proline tert-butyl ester was racemized in the
N-arylation step, the compounds made using this
method were 1:1 mixtures of diastereomers at the proline
a-center. The diastereomers were not easilyseparated
and so theywere tested as mixtures. Compounds prepared
bythe methods outlined in Schemes 1 and 2 were eval-
uated as inhibitors of VLA-4 and the related integrin,
a4b7 and are outlined in Table 1.
Figure 2. New leads based on parent molecule A.
The unsubstituted N-phenyl proline derivative (5a)
exhibited a 10-fold loss in potencyrelative to sulfonamide
1. This loss in potencyis somewhat less than what was
observed (30-fold loss) in going from B to C as described
in Figure 2. Clearly, the interaction of the sulfonyl
group with some portion of VLA-4 is important for
binding, but it is not critical for activity. Attempts to
find this interaction via substitution of the phenyl ring
at the 2- and 3-positions (5b,c) with hydrogen bond
acceptors led to small improvements in potency, while
Scheme 1. (a) Aryl-halide, CuI, K2CO3, DMAC, 90 ꢀC; (b) HATU,
HOAt, 2,6–dimethoxybiphenyl alanine tert-butyl ester, i-Pr2NEt,
CH2Cl2; (c) TFA, CH2Cl2.

T. M. Kamenecka et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2205–2208
2207
Table 1. Inhibition of VLA-4 and a4b7 by N-aryl-prolyl biphenyl-
alanine derivatives (IC50, nM)
Although one pyridyl nitrogen is good as in 5f, two is
not advantageous as pyrimidinyl analogue 5h was less
active. The 2-pyrazinyl proline (5g), which is essentially
the combination of 2- and 3-pyridine, was the most
potent analogue.14 Introduction of another heteroatom
in the ring will reduce the basicityof the system and this
maybe enough to lessen intramolecular hdyrogen
bonding if that is an issue. Also of note is that ana-
logues 5e through 5j are a mixture of diastereomers and
one enantiomer is likelyto be more potent than the
other. In summary, these results confirm our earlier
findings that the sulfonyl residue is not necessary for
potencyas evidenced bythe conversion of B to C, but
now also in the conversion of C to D (or for that matter,
A to D). The potencylost byremoval of the sulfonyl
moietyfrom 1 is almost completelyrecovered in N-aryl
proline 5g.
c
Compd
Ar
VLA-4b
a4b7
5a
5b
5c
5d
5e
5f
5g
5h
5i
Phenyl
2-CN-phenyl
3-CH3O-phenyl
4-NO2-phenyl
2-Pyridyla
0.61
0.37
0.41
1.1
21
3.5
5.8
17.1
18
5.5
8.9
18.8
45
1.9
3-Pyridyla
0.34
0.17
0.73
1.1
2-Pyrazinyla
5-Pyrimidinyla
4-Isoquinolyla
4,6-(NO2)2–2-pyridyla
5j
0.38
17
Despite improvements in potencyon VLA-4 from lead
5a, these compounds showed little specificitywith
regard to a4b7. Most compounds were in the range of
10- to 30-fold specific for VLA-4 with the most potent
derivative 5g exhibiting ꢁ50-fold specificity. These
results are not surprising given the lack of specificity
(ꢁ10- to 100-fold) also observed for compounds in class
C (Fig. 1) and suggests that perhaps the interactions
which confer specificityare located in a different part of
the molecule.
a(R,S)-Proline mixture of diastereomers.
bVCAM-Ig ligand.
cMAdCAM-Ig ligand.
Table 2. Pharmacokinetic parametersa of selected N-aryl-prolyl
derivatives
Compd
F (%)b
CLp (mL/min/kg)
t1/2 (h)c
1
2
14
<1
3
<1
2
127
78
69
64
>100
29
43
0.6
0.6
0.3
0.7
1.4
0.3
0.2
5a
5b
5c
5d
5e
5f
The pharmacokinetic properties of selected derivatives
were determined in rats (Table 2). As hoped, removal of
the sulfonyl group (1 to 5a) leads to a compound with
improved oral bioavailability(14%) and a lower plasma
clearance rate. With the exception of 5d, all compounds
examined had plasma clearance rates lower than that of
the corresponding sulfonylated proline 1, however they
were still high.15 Despite the apparent lower clearance
rates, recoveryof potencycomes at the expense of oral
bioavailability. As the potency of pyridyl analogues 5e
and 5f approaches that of 1, so does their lack of oral
bioavailability. This common trend was not only
observed here, but also in other efforts aimed at targeting
unique leads.4,5
<1
aSprague–Dawleyrats.
bDose: 1 mg/kg iv; 2 mg/kg po.
ct1/2=plasma half-life(0À8 h)
.
4-substitution was deleterious. Efforts to find the sulfo-
nyl binding site by using the lone pair on nitrogen via
the 2-pyridyl derivative (5e) was unsuccessful giving a
compound 3-fold less active. It is not clear what is
responsible for this loss in activity, as the 2-pyridyl
nitrogen atom is the closest proximation in space to one
of the sulfonyl oxygen atoms. However, the basicity of
aminopyridine 5e might be what is responsible for its
loss in activity. The potential for an intramolecular
hydrogen bonding interaction between the pyridyl
nitrogen atom in 5e and the amide NH bond mayalter
the conformation of the molecule. This would be less
likelyto occur with the less basic pyrizine derivative ( 5g)
or with the non-basic sulfonamide 1. The other analogues
do not have a heteroatom capable of such an interaction.13
One of the most potent derivatives was 3-pyridyl proline
(5f) which showed a 2-fold improvement in potency
over 5a. The heteroatom is apparentlybetter situated to
pick up the hydrogen bonding interaction within VLA-4
and improves activity. However, there are spatial
limitations within the receptor as 4-isoquinolyl deriva-
tive (5i) is apparentlytoo large and potencyis lost
(Table 1).
In summary, a new lead was discovered based on sulfo-
namide 1, lacking the sulfonyl group. Although removal
of the sulfonyl residue led to a 10-fold drop in potency,
it was not required for binding to the receptor. Proper
selection of the aryl group on nitrogen allowed for
recoveryof potencyin the binding assays with modest
improvements in plasma clearance rates. N-aryl proline
derivatives 5f and 5g represent two such compounds
which maybe further optimized as potent VLA-4
antagonists.
Acknowledgements
The authors are grateful to Zhen Wang and Junying
Wang for formulation and mass spectral analysis and to
Marcie Donnelly, Ken Vakerich and Chris Nunes in the
performance of the in vivo phase of the rodent phar-
macokinetic studies and sample preparation.

2208
T. M. Kamenecka et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2205–2208
130, 52736s. Briefly, human Jurkat cells were suspended in
References and Notes
binding buffer (25 mM HEPES, 150 mM NaCl, 3 mM KCl, 2
mM glucose, 0.1% bovine serum albumin, pH 7.4) supple-
mented with MnCl2 (1 mM), placed in Millipore MHVB mul-
tiscreen plates Æ compounds in DMSO, incubated at room
temperature for 30 min, filtered on a vacuum box, and washed
with 100 mL of binding buffer containing 1 mM MnCl2. After
insertion of the plates into adapter plates, 100 mL of Micro-
scint-20 (Packard cat# 6013621) was added to each well. The
plates were then sealed, placed on a shaker for 30 s, and
counted on a Topcount microplate scintillation counter
(Packard). Control wells containing DMSO alone were used
to determine the level of VCAM-Ig binding corresponding to
0% inhibition. Control wells in which cells were omitted were
used to determine the level of binding corresponding to 100%
inhibition. Percent inhibition was then calculated for each test
well and the IC50 was determined from a ten point titration
using a validated four parameter fit algorithm. All titrations
were run in duplicate.
1. Hemler, M. E.; Elices, M. J.; Parker, C.; Takada, Y.
Immunol. Rev. 1990, 114, 45.
2. Elices, M. J. Curr. Opin. Anti-inflam. Immunol. 1999, 1, 15.
3. (a) For recent reviews on VLA-4 antagonists, see: Adams,
S. P.; Lobb, R. R. Annu. Rep. Med. Chem. 1999, 34, 179. (b)
Porter, J. R. Idrugs 2000, 3, 788. (c) Adams, S. P.; Lobb, R. R.
Prog. Respir. Res. 2001, 31, 302.
4. (a) For recent reported antagonists to VLA-4, see: Chen,
L.; Tilley, J. W.; Huang, T. N.; Miklowski, D.; Trilles, R.;
Guthrie, R. W.; Luk, K.; Hanglow, A.; Rowan, K.; Schwinge,
V.; Wolitzky, B. Bioorg. Med. Chem. Lett. 2000, 10, 725. (b)
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.; Rowen, K. Bioorg. Med. Chem.
Lett. 2000, 10, 729. (c) Tilley, J. W.; Kaplan, G.; Fotouhi, N.;
Wolitzky, B.; Rowan, K. Bioorg. Med. Chem. Lett. 2000, 10,
1163. (d) Fotouhi, N.; Joshi, P.; Tilley, J. W.; Rowan, K.;
Schwinge, V.; Wolitzky, B. Bioorg. Med. Chem. Lett. 2000, 10,
1167. (e) Astles, P. C.; Harris, N. V.; Morley, A. D. Bioorg.
Med. Chem. Lett. 2001, 9, 2195.
5. (a) Furth, P. S.; Chiang, S. L.; Sircar, I.; Griffith, R. C.;
Nowlin, D.; Gorscan, F. S.; Mah, J.; Lazarides, E. Abstracts
of Papers, Part 1, 218th National Meeting of the American
Chemical Society, New Orleans, LA, American Chemical
Society: Washington, DC, 1999, MEDI 61. (b) Semko, C. M.;
Dressen, D. B.; Grant, F. S.; Konradi, A. W.; Pleiss, M. A.;
Thorsett, E. D.; Freedman, S. B.; Holsztynska, E. J.; Quinn,
K. P.; Yednock, T. Abstracts of Papers, Part 2, 220th National
Meeting of the American Chemical Society, Washington, DC,
American Chemical Society: Washington, DC, 2000, MEDI
133. (c) Hagmann, W. K.; Durette, P. L.; Lanza, T.; Kevin,
N. J.; deLaszlo, S. E.; Kopka, I. E.; Young, D.; Magriotis,
P. A.; Li, B.; Lin, L. S.; Yang, G.; Kamenecka, T.; Chang,
L. L.; Wilson, J.; MacCoss, M.; Mills, S. G.; Van Riper, G.;
McCauley, E. D.; Egger, L. A.; Kidambi, U.; Lyons, K.; Vin-
cent, S.; Stearns, R.; Colletti, A.; Teffera, J.; Tong, S.; Fenyk-
Melody, J.; Owens, K.; Levorse, D.; Kim, P.; Schmidt, J. A.;
Mumford, R. A. Bioorg. Med. Chem. Lett. 2001, 11, 2709.
6. de Laszlo, S. E., Gutteridge, C. Unpublished results.
7. Li, B.; de Laszlo, S. E.; Kamenecka, T.; Kopka, I. E.; Lin, L.
S.; Chang, L. L.; Truong, Q.; Durette, P. L.; Lanza, T.; Mac-
Coss, M.; Tong, S.; McCauley, E.; Van Riper, G.; Schmidt, J.
A.; Hagmann W. K. Bioorg. Med. Chem Lett. In press.
8. Details of a competitive binding assaybetween human
Jurkat cells and a radiolabeled 125I-VCAM-immunoglobulin
fusion protein (125I-VCAM-Ig) have been disclosed: Durette,
P. L.; Hagmann, W. K.; MacCoss, M.; Mills, S.; Mumford, R.
PCT Application WO98/53817A1, 1998, Chem. Abstr. 1998,
9. Details of a competitive binding assaybetween human
RPMI-8866 cells (a human B-cell line a4b7 was a gift from
Professor John Wilkins, Universityof Manitoba, Canada) and
radiolabeled 125I-MAdCAM-immunoglobulin fusion protein
(
the VLA-4 binding assay.
125I-MAdCAM-Ig) have been disclosed5 and are similar to
10. (a) Ma, D.; Yao, J. Tetrahedron: Asymmetry 1996, 7,
3075. (b) Ma, D.; Zhang, Y.; Yao, J.; Wu, S.; Tao, F. J. Am.
Chem. Soc. 1998, 120, 12459.
11. (a) Carpino, L. A.; El-Faham, A.; Albericio, F. J. Org.
Chem. 1995, 60, 3561. (b) Carpino, L. A.; Ionescu, D.; El-
Faham, A. J. Org. Chem. 1996, 61, 2460.
12. (a) Guram, S.; Rennels, R. A.; Buchwald, S. L. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1348. (b) Wolfe, J. P.; Tomori,
H.; Sadighi, J. P.; Yin, J.; Buchwald, S. L. J. Org. Chem. 2000,
65, 1158. (c) Ali, M. H.; Buchwald, S. L. J. Org. Chem. 2001,
66, 2560.
13. The potential hydrogen bond in 5e would exist in a seven-
membered ring. Other analogues such as 5f or 5h would make
an eight-membered ring, which is less likely.
14. The other combination of 2- and 3-pyridines leading to a
pyridazinyl derivative was not made.
15. Although not fullyinvestigated, these compounds most
likelysuffer from similar metabolism issues as alreadydescri-
bed. For a discussion, see ref 5c and: Kopka, I. E.; Young,
D. N.; Lin, L. S.; Mumford, R. A.; Magriotis, P. A.; Mac-
Cross, M.; Mills, S. G.; Van Ripper, G.; McCauley, E.; Egger,
L. E.; Kidambi, U.; Schmidt, J. A.; Lyons, K.; Stearns, R.;
Vincent, S.; Colletti, A.; Wang, Z.; Tong, S.; Wang, J.; Owens,
K.; Levorse, D.; Hagmann, W. K. Bioorg. Med. Chem. Lett.
2002, 12, 637.