Identification of potent and novel α4β1 antagonists using in silico screening

Article abstract of DOI:10.1021/jm020054e

The antigen α4β1 (very late antigen-4, VLA-4) plays an important role in the migration of white blood cells to sites of inflammation. It has been implicated in the pathology of a variety of diseases including asthma, multiple sclerosis, and rheumatoid art

Full text of DOI:10.1021/jm020054e

2988
J . Med. Chem. 2002, 45, 2988-2993
Id en tifica tion of P oten t a n d Novel r4â1 An ta gon ists Usin g in Silico Scr een in g
J uswinder Singh,*,† Herman van Vlijmen,† Yusheng Liao,† Wen-Cherng Lee,† Mark Cornebise,† Mary Harris,†
I-hsiang Shu,† Alan Gill,† J ulio H. Cuervo,† William M. Abraham,§ and Steven P. Adams†
Department of Drug Design and Evaluation, Biogen Inc., 12 Cambridge Center, Cambridge, MA 02142, and
Mount Sinai Medical Center, Miami Beach, FL 33140
Received February 5, 2002
The antigen R4â1 (very late antigen-4, VLA-4) plays an important role in the migration of
white blood cells to sites of inflammation. It has been implicated in the pathology of a variety
of diseases including asthma, multiple sclerosis, and rheumatoid arthritis. We describe a series
of potent inhibitors of R4â1 that were discovered using computational screening for replacements
of the peptide region of an existing tetrapeptide-based R4â1 inhibitor (1; 4-[N′-(2-methylphenyl)-
ureido]phenylacetyl-Leu-Asp-Val) derived from fibronectin. The search query was constructed
using a model of 1 that was based upon the X-ray conformation of the related integrin-binding
region of vascular cell adhesion molecule-1 (VCAM-1). The 3D search query consisted of the
N-terminal cap and the carboxyl side chain of 1 because, upon the basis of existing structure-
activity data on this series, these were known to be critical for high-affinity binding to R4â1.
The computational screen identified 12 reagents from a virtual library of 8624 molecules as
satisfying the model and our synthetic filters. All of the synthesized compounds tested inhibit
R4â1 association with VCAM-1, with the most potent compound having an IC50 of 1 nM,
comparable to the starting compound. Using CATALYST, a 3D QSAR was generated that
rationalizes the variation in activities of these R4â1 antagonists. The most potent compound
was evaluated in a sheep model of asthma, and a 30 mg nebulized dose was able to inhibit
early and late airway responses in allergic sheep following antigen challenge and prevented
the development of nonspecific airway hyperresponsiveness to carbachol. Our results demon-
strate that it is possible to rapidly identify nonpeptidic replacements of integrin peptide
antagonists. This approach should be useful in identification of nonpeptidic R4â1 inhibitors
with improved pharmacokinetic properties relative to their peptidic counterparts.
In tr od u ction
therefore be useful to identify alternative nonpeptidic
structures to replace the Leu-Asp-Val portion of the
inhibitors to yield compounds with better pharmacoki-
netic properties while maintaining potency.
The antigen R4â1 (VLA-4, very late antigen-4) is a
member of the integrin family and is involved in the
migration of mononuclear leukocytes to sites of inflam-
mation. Due to the role of R4â1 in the inflammatory
response, it is an attractive target for therapeutic
intervention. The R4-integrin-dependent adhesion path-
ways are critical intervention points in several inflam-
matory and autoimmune pathologies. These include
asthma,1 multiple sclerosis,2,3 and rheumatoid arthritis.4
The antigen R4â1 is known to bind fibronectin in the
extracellular matrix and vascular cell adhesion mole-
cule-1 (VCAM-1) on the endothelium. Komoriya et al.,5
using peptide mapping studies, were able to identify the
minimum fibronectin sequence for cell adhesion activity
as being Leu-Asp-Val. This integrin-binding motif was
used as the starting point in the design of small
molecule antagonists of R4â1. Lin et al.6 used a novel
capping strategy to identify 4-[N′-(2-methylphen-
yl)ureido]phenylacetyl-Leu-Asp-Val (1) as a potent,
highly selective inhibitor of R4â1. This class of com-
pounds exhibits efficacy in inflammatory models such
as the sheep model of allergic bronchoconstriction;6
however, their peptidic nature results in rapid clearance
in animals (S.P.A., unpublished results). It would
Database mining has emerged as a productive way
to identify more favorable replacements for peptide
fragments in lead molecules. One approach involves
defining a three-dimensional pharmacophore model for
the ligand and searching a chemical database of existing
compounds for those structures that satisfy the spatial
and chemical constraints of the model.7 Computational
tools such as CATALYST8 and UNITY3D9 have been
utilized successfully to identify novel inhibitors of HIV-
integrase10, as well as the angiotensin II11 and the
muscarinic M3 receptor.12 In this paper, the application
of this approach to the discovery of novel nonpeptidic
VLA-4 antagonists is described.
Resu lts
Mod el Bu ild in g a n d Da ta ba se Sea r ch in g. To
discover novel replacements of the Leu-Asp-Val portion
of our peptidic inhibitors, 1 was used as a template to
build a 3D query, which consisted of the PUPA (4-[N′-
(2-methylphenyl)ureido]phenylacetyl) and the carboxyl
group (Figure 1). These groups were selected upon the
basis of structure-activity data that showed that they
were critical for high-affinity binding. For example, we
have shown previously that the starting peptide has
weak binding to R4â1 (Ile-Leu-Asp-Val ≈ 66 µM)6
* To whom correspondence should be addressed. Phone: (617)679-
2027. Fax: (617)679-2616. E-mail: J uswinder_Singh@biogen.com.
† Biogen Inc.
§ Mount Sinai Medical Center.
10.1021/jm020054e CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/11/2002

Potent and Novel R4â1 Antagonists
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 14 2989
sequences, which accounted for a large fraction of the
hits, leaving 170 compounds for consideration (2% of the
initial database of 8624 reagents). From this set of
approximately 170 reagents, 12 were selected on the
basis of synthetic feasibility and commercial availability.
A straightforward, three-step process of reduction,
coupling to PUPA, and cleavage of the ester (Scheme
1) was used to produce the set of 12 compounds
submitted to the in vitro assay (0.14% of the initial
database of 8624 reagents).
Biologica l Resu lts. A direct binding assay was used
to evaluate the ability of the synthesized compounds to
inhibit the interaction between VLA-4 and a VCAM-
Ig fusion protein. All of the compounds synthesized from
the commercially available templates derived from
computational screening demonstrated inhibition in our
direct binding assay, with the most potent compound
(2, 1.3 nM) comparable in potency to the starting
compound upon which the computational search query
was based (1, 0.6 nM). The twelve compounds built from
computational screening showed a wide range in inhibi-
tory activity from 1.3 to 20 000 nM (Figure 2). The
compound 2 was derived from a phenylpropionic acid
template with a â-methyl substitution. The R-amide
substituent of the phenylpropionic acid template also
showed potent inhibition (3, 58 nM), although decreased
relative to the â-methyl substituent. Interestingly, para
substitution of the phenyl ring in the phenylpropionic
template provided significantly more potent inhibition
than meta substitution as reflected in the 2000-fold
difference in activity exemplified by analogues 2 and 9.
Another potent template involved a urea group (4) with
an IC50 of 67 nM. The remaining compounds were much
less active (5-13, 800-20 000 nM).
F igu r e 1. In silico screening for Leu-Asp-Val replacements:
(a) the 3-D conformation of 1 used to build the 3D CATALYST
search query; (b) the CATALYST search query (hypothesis)
consisting of PUPA (cyan and yellow) and carboxyl group
(green) from 1; (c) overlay of a compound from the in silico
screen with the search query.
In the sheep model of allergic airways disease, a
single 30 mg dose of compound 2 administered 30 min
before antigen challenge inhibited the immediate antigen-
induced bronchial response by 48%. Late-phase bron-
choconstriction and the antigen-induced nonspecific
airway hyperresponsiveness to inhaled carbachol were
almost completely eliminated (Figure 3).
and that addition of the N-terminal cap PUPA results
in a 100 000-fold enhancement in binding affinity. The
PUPA alone does not inhibit the integrin. The aspartic
acid side chain appears to be crucial for inhibition
because deletion of the Asp-Val side chains to form
PUPA-Leu results in complete loss of binding while
truncation of Val, providing PUPA-Leu-Asp, results in
only a 30-fold decrease in the inhibition of R4â1.6
The conformation of 1 was based on the X-ray crystal
structure of the integrin binding region of VCAM-1, Ile-
Asp-Ser, which is homologous to the Leu-Asp-Val por-
tion of fibronectin and is known to bind to R4â1. The
conformation of the N-terminal cap (PUPA) was deter-
mined from a systematic conformational energy search.
The resulting 3D query was used to search a database
of virtual compounds that consisted of commercially
available reagents coupled to the N-terminal cap, PUPA.
A search of the Available Chemicals Directory13 for all
compounds with a free amine or nitro group and a
carboxyl group identified 8624 reagents. A virtual
chemical library was created by coupling in silico the
carboxyl of the N-terminal cap (PUPA) to the free amine
or nitro of the commercially available reagent (Scheme
1). A multiconformational database of these virtual
compounds was built and screened using our 3D query,
resulting in 416 hits. At this stage, we removed peptide
To explain the structural basis for the variation in
the activities of the compounds, a 3D QSAR model was
built using CATALYST. The CATALYST program pro-
vides a feature-based description of chemical structures
(e.g., H-bond acceptors, negative ionizable groups, etc.)
and attempts to correlate the 3D arrangement of
features across a compound set with their biological
activities. The model that best accounted for the activi-
ties of the compounds consisted of a hydrogen-bond
acceptor, a hydrogen-bond donor, a hydrophobic feature,
and a custom-built superfeature, which described the
PUPA. The model is shown in Figure 4, overlayed with
the most potent compound, 2 (IC50 ≈ 1.3 nM). All of the
features are mapped by 2, with the carboxyl satisfying
the hydrogen-bond acceptor site and the â-methyl
satisfying the hydrophobic site. The amide NH of 2 maps
to the hydrogen-bond donor site, while the PUPA maps
to our superfeature. The weaker potencies of 3 (IC50
)
58 nM) and 4 (IC50 ) 67 nM) are attributed to inad-
equate mapping of the methyl site, although the other
features are mapped well. It appears that the com-
pounds 5-10 (IC50 ) 800-3000 nM) fail to optimally

2990 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 14
Singh et al.
Sch em e 1a
a
Synthesis of 4. Key: (a) glycine methyl ester hydrochloride, Et3N, DMF; (b) H2, Pd/C, MeOH; (c) 2-methyl-PUPA, HATU, DIEA,
DMF; (d) LiOH, DMF.
F igu r e 3. Effect of compound 2 on sheep airway responses
to antigen. Naturally allergic sheep that were previously
shown to exhibit both early and late bronchoconstrictor
responses to nebulized A. suum antigens were used in the
study. A 30 mg dose (∼1 mg/kg) of 2 formulated as the Tris
salt in saline was administered by an air-jet nebulizer 0.5 h
prior to challenge with nebulized antigen. Specific lung
resistance (SRL) was determined periodically for drug-treated
animals (b) over 8 h and is expressed as the change from
prechallenge SRL. Historical control resistance data following
antigen challenge (O) is presented for the same animals that
received drug in this study. Airway responsiveness was
determined 24 h after antigen challenge by determining the
cumulative dose of carbachol required to induce a 4-fold
increase in SRL. Results are expressed as the ratio of the
carbachol dose in the presence or absence of drug treatment
divided by the carbachol dose to induce the same increase
before antigen challenge (right panel). The data are plotted
as the mean ( SEM (mean ( 0.5(range)) for two sheep.
F igu r e 2. Structures and binding assay data for the com-
pounds of this study derived from the database searching
method.
the pharmacophore model will be in improving the
potencies of the micromolar inhibitors identified from
the virtual screening. The efficacy of 2 to inhibit early
and late airway responses in allergic sheep via aerosol
administration is encouraging. We are presently at-
tempting to improve the pharmacokinetics of this series,
map the superfeature site but are able to map all of the
other features. Similarly, 11 (IC50 ) 11 400 nM), 12 (IC50
) 16 000 nM), and 13 (IC50 ) 20 000 nM) are unable to
map the superfeature and have a poorer fit to the other
which will enable us to better assess the therapeutic
features.
utility of intervention at R4â1 in a broader range of
animal models.
Con clu sion
The successful transformation of a peptide lead
compound to a nonpeptide epitope remains a significant
challenge in drug discovery. We have demonstrated how
in silico screening resulted in identifying novel potent
nanomolar replacements for the peptidic portion of an
existing R4â1 antagonist. We are now testing how useful
Exp er im en ta l Section
Mod el Bu ild in g. A 3D model of 1 was built using the
BUILDER module within InsightII.8 All energy calculations
were performed using the CVFF force field within DISCOVER,
and the potentials were set using the ForceField/Potential
command.

Potent and Novel R4â1 Antagonists
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 14 2991
r4â1 In h ibitor Effects on Sh eep Air w a y Resp on ses to
An tigen . Methods for assessing drug effects on antigen-
induced increases in specific lung resistance (SRL) are de-
scribed elsewhere.1 A 30 mg dose of 1 formulated in 5 mL of
saline containing 2 mol equiv of Tris (pH 7.0-7.4) was
administered by an air-jet nebulizer (Raindrop, Puritan,
Bennett, Lenexa, KS) 0.5 h prior to challenge with nebulized
Ascaris suum antigen. SRL was determined periodically for
drug-treated animals over 8 h. The change in SRL was
calculated for each sheep as the difference from prechallenge
baseline SRL, and values reported are the mean ( SEM (mean
( 0.5(range)) for two sheep. Airway responsiveness was
determined 24 h after antigen challenge and is expressed as
the cumulative dose of carbachol at 24 h required to induce a
4-fold increase in SRL in the presence or absence of drug
treatment divided by the carbachol dose to induce the same
increase before antigen challenge. The protocol used in the
sheep study was approved by the Mount Sinai Medical Center
Animal Research Committee, which is responsible for ensuring
the humane care and use of experimental animals.
Ch em istr y: Gen er a l. Unless noted otherwise, all starting
materials and synthesis reagents were obtained commercially
and used without further purification. Abbreviations are as
follows: DIEA, N,N-diisopropylethylamine; DMF, dimethyl-
formamide; Et3N, triethylamine; HATU, O-(7-azabenzotriazol-
1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate; EDC,
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride;
2-methylPUPA, 2-(4-(3-(o-methylphenyl)ureido)phenyl)acetic
acid; TFA, trifluoroacetic acid.
Nuclear magnetic resonance spectra (NMR) were recorded
as solutions in the indicated solvent on Bruker DPX-400 and
AC-300. Chemical shifts are reported in parts per million (δ,
ppm) using CHCl3 or DMSO as an internal standard. Coupling
constants are reported in hertz (Hz). Spectral splitting patterns
are designated as follows: s, singlet; d, doublet; t, triplet; q,
quartet; m, multiplet; br, broad.
Mass spectra were obtained on a Fisons VG Platform
Quattro II mass spectrometer system and processed with
MassLynx software. High-resolution mass spectra were per-
formed by M-Scan Inc. (West Chester, PA) on a VG Analytical
ZAB 2SE high-field mass spectrometer.
F igu r e 4. The alignment of 2 to the CATALYST model. The
wireframe spheres represent CATALYST features: hydrogen-
bond acceptor (green); hydrophobe (cyan); hydrogen-bond
donor (purple); superfeature (pink).
The Leu-Asp-Val section of 1 was built using the side chain
and main chain torsion angles from the Ile39-Asp40-Ser41
region from VCAM-1 (protein databank code 1VCA).14 The
N-terminal cap, termed PUPA, was attached onto the Leu-
Asp-Val fragment. The N-terminal cap was energy minimized
using 1000 steps of steepest descent, while the rest of the
structure was held fixed. By use of a systematic conformational
search, the lowest energy conformation of the N-terminal cap
was identified. The systematic conformational search was
carried out using the SEARCH/COMPARE module of IN-
SIGHTII using the default settings with a 30° step size and
the ENERGY option set to on. During the minimization, all
atoms derived from the Ile39-Asp40-Ser41 region of VCAM-1
were held fixed. After each step in the conformational search
the N-terminal cap was subjected to 500 steps of energy
minimization using the Optimize option with the default
parameters, and the lowest energy conformation was selected.
3D Qu er y a n d Da ta ba se Sea r ch in g. The Available
Chemical Directory13 was screened for all compounds with a
carboxyl group and either a free amine or nitro group. These
commercially available reagents were computationally coupled
to the free carboxyl group of our N-terminal cap (PUPA) via
the free amine or nitro group using LEGION.9
With the use of CATALYST,8 a multiconformer 3D database
of the virtual library was created. For the conformer genera-
tion, the Fast conformer process was used with the MAX-
CONFS option set to 250 and the energy threshold set to 15
kcal/mol.
Analytical HPLC was performed on a Perkin-Elmer 200
Series HPLC system with UV detection at 214 or 254 nm or
both on an Applied Biosystems 785A programmable absorption
detector. Data analysis was run using a PE Nelson 1020
integrator. A Higgins Analytical C-18 reverse-phase column
(3 µm pore size, 4.6 mm × 15 cm) was utilized and was running
with a 7 min gradient using H2O (containing 0.1% TFA v/v)
and CH3CN (containing 0.085% TFA v/v) as the mobile phase.
Preparative HPLC was performed with a Rainin HPLX solvent
delivery system with a Rainin Dynamax UV-1 detector (set
at 225 nm) using a Rainin Microsorb C-18 reverse-phase
column (5 µm pore size, 21.4 mm × 25 cm) eluted at 12 mL/
min using the same solvent system as for the analytical HPLC.
Fractions were collected and checked for purity with analytical
HPLC, and pure fractions were combined and evaporated in
vacuo to provide purified material.
The 3D query that was used for database mining was based
upon 1 and was built using CATALYST. The query was built
using ViewHypothesis workbench and consisted of the PUPA
and the carboxyl of 1. A tolerance of 2.0 Å was used at each of
the atomic positions.
CATALYST was also used to build a 3D QSAR of the
compounds synthesized from the results of our database
search. The IC50’s used for the model generation were taken
from Figure 2. The models were built using CATHYPO in
CATALYST. The conformational models were built using the
CATCONF module with the Best option set to on and the
maximum number of conformations set to 250, and the energy
threshold was set to 15 kcal/mol. All other parameters used
were at their default settings. In addition to using the standard
feature set supplied by CATALYST (hydrogen-bond donor,
hydrogen-bond acceptor, negative ionizable, hydrophobe), a
custom-built feature that described the PUPA portion of the
compounds was added. Because all of the compounds had this
feature, it was defined as a superfeature using the Exclude/
Edit tool. This was necessary because the number of features
presented by the compounds was much greater (as judged by
the CONFIG value) than CATALYST could adequately sample.
The superfeature consisted of two individual features describ-
ing the centroid of each of the phenyl rings of the PUPA
connected by a vector. The best model had a correlation
coefficient of 0.97 between the predicted and actual activities.
All new compounds show analytical and spectral (1H NMR,
MS) data consistent with the depicted structures. All reactions
were conducted at room temperature unless otherwise noted,
and were followed to completion by analytical HPLC.
Syn th esis of com p ou n d 4 (see Sch em e 1): Meth yl 2-(3-
(3-n itr op h en yl)u r eid o)a ceta te (4). To a solution of glycine
methyl ester hydrochloride (0.842 g, 6.7 mmol) in DMF (10
mL) was added Et3N (2.8 mL, 20 mmol) with stirring to give
a colorless solution. To this was added 3-nitrophenylisocyanate
(1.0 g, 6.1 mmol) in portions over 5 min to give a yellow
solution. After stirring 14 h, the solution was diluted with ethyl
acetate (EtOAc), washed three times with 1 N HCl and then
with brine, dried (MgSO4), and concentrated in vacuo to a thick
yellow oil. The product was recrystallized twice from EtOAc/
hexanes and then again from ethanol/water to give 1.0 g (65%)
of compound 14 as fine, pale-yellow needles (HPLC, >99%).
MS: m/z 254 (M + H)+.
Biologica l Assa y. The compounds were tested in binding
assays, which have been reported in detail elsewhere.15

2992 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 14
Singh et al.
Gen er a l P r oced u r e b for Red u ction of a Nitr o Gr ou p
to th e Cor r esp on d in g Am in e. Meth yl 2-(3-(3-a m in op h en -
yl)u r eid o)a ceta te (15). To solution of 14 (0.5 g, 1.98 mmol)
in 15 mL of methanol (MeOH) in a 90 mL Pyrex high-pressure
vessel flushed with nitrogen was added 10% palladium on
carbon (ca. 100 mg). The vessel was capped with a hydrogena-
tion head, flushed with nitrogen, and then charged with
hydrogen (60 psi), and the mixture was allowed to stir for 14
h. The system was purged with nitrogen, and the mixture was
filtered through a plug of Celite 545. The Celite plug was
washed with additional MeOH, and the combined filtrate was
concentrated in vacuo to give 0.44 g (99%) of compound 15 as
a foamy white solid that was used without further purification.
MS: m/z 224 (M + H)+.
Gen er a l P r oced u r e c for Cou p lin g 2-Meth yl-P UP A to
Am in e. Meth yl 2-(3-(3-(2-Meth yl)P UP a ceta m id op h en yl)-
u r eid o)a ceta te (16). To a mixture of 15 (0.23 g, 1.03 mmol),
2-methyl-PUPA1 (0.284 g, 1.0 mmol), and HATU (0.456 g, 1.2
mmol) in 6 mL of DMF was added DIEA (0.7 mL, 4 mmol) to
give a dark yellow solution. The mixture was stirred for 2 h
and then diluted with EtOAc, washed three times with 1 N
HCl and then brine, dried (MgSO4), and concentrated in vacuo
to give a yellow solid. Three iterations of dissolution/precipita-
tion from acetone/hexane gave 0.352 g (72%) of compound 16
as an off-white solid (HPLC, >98%). MS: m/z 490 (M + H)+.
(EDC and Et3N were used as the coupling reagents for
compounds 2, 3, 12, 13.)
Com p ou n d 5: 1H NMR (400 MHz, DMSO-d6) δ 10.43 (s, 1
H), 8.98 (s, 1 H), 8.68 (s, 1 H), 8.30 (s, 1 H), 7.87 (s, 1 H), 7.83
(d, J ) 7.6 Hz, 1 H), 7.77 (d, J ) 8.7 Hz, 1 H), 7.53 (dd, J )
8.8, 1.8 Hz, 1 H), 7.40 (d, J ) 8.5 Hz, 2 H), 7.25 (d, J ) 8.5 Hz,
2 H), 7.14 (m, 2 H), 6.92 (t, J ) 7.4 Hz, 1 H), 3.93 (br s, 2 H),
3.61 (s, 2 H), 2.22 (s, 3 H); MS (EI) m/z 501 (M + H)+; HRMS
m/z (M + H)+ calcd 501.1886, obsd 501.1921.
Com p ou n d 6: 1H NMR (400 MHz, DMSO-d6) δ 10.28 (s, 1
H), 8.97 (s, 1 H), 8.74 (t, J ) 5.9 Hz, 1 H), 8.05 (s, 1 H), 7.87
(s, 1 H), 7.83 (d, J ) 7.6 Hz, 1 H), 7.79 (br d, J ) 8.0 Hz, 1 H),
7.51 (d, J ) 7.8 Hz, 1 H), 7.39 (m, 3 H), 7.24 (d, J ) 8.5 Hz, 2
H), 7.16-7.10 (m, 2 H), 6.92 (t, J ) 7.4 Hz, 1 H), 3.89 (d, J )
5.9 Hz, 2 H), 3.57 (s, 2 H), 2.22 (s, 3 H); MS (EI) m/z 461 (M
+ H)+; HRMS m/z (M + H)+ calcd 461.1825, obsd 461.1854.
Com p ou n d 7: 1H NMR (400 MHz, DMSO-d6) δ 10.09 (s, 1
H), 8.96 (s, 1 H), 7.87 (s, 1 H), 7.83 (d, J ) 7.9 Hz, 1 H), 7.57
(s, 1 H), 7.51 (d, J ) 8.1 Hz, 1 H), 7.39 (d, J ) 8.5 Hz, 2 H),
7.22 (m, 3 H), 7.14 (m, 2 H), 6.99 (d, J ) 7.7 Hz, 1 H), 6.92 (t,
J ) 7.4 Hz, 1 H), 4.87 (t, J ) 6.8 Hz, 1 H), 3.55 (s, 2 H), 2.47
(d, J ) 6.5 Hz, 2 H), 2.22 (s, 3 H); MS (EI) m/z 448 (M + H)+;
HRMS m/z (M + H)+ calcd 448.1872, obsd 448.1857.
Com p ou n d 8: 1H NMR (400 MHz, DMSO-d6) δ 10.06 (s, 1
H), 8.96 (s, 1 H), 7.87 (s, 1 H), 7.82 (d, J ) 7.5 Hz, 1 H), 7.44
(m, 2 H), 7.39 (d, J ) 8.5 Hz, 2 H), 7.23 (d, J ) 8.5 Hz, 2 H),
7.14 (m, 3 H), 6.92 (t, J ) 7.4 Hz, 2 H), 3.54 (s, 2 H), 3.07 (m,
1 H), 2.45 (d, J ) 7.4 Hz, 2 H), 2.22 (s, 3 H). MS (EI) m/z 446
(M + H)+; HRMS m/z (M + H)+ calcd 446.2080, obsd 446.2082.
Com p ou n d 9: 1H NMR (400 MHz, DMSO-d6) δ 10.30 (s, 1
H), 8.95 (s, 1 H), 7.86 (s, 1 H), 7.83 (d, J ) 7.6 Hz, 1 H), 7.36
(d, J ) 8.5 Hz, 2 H), 7.19 (d, J ) 8.5 Hz, 2 H), 7.14 (m, 2 H),
7.05 (s, 1 H), 6.92 (t, J ) 7.5 Hz, 1 H), 4.00 (t, J ) 6.7, 2 H),
3.47 (s, 2 H), 2.63 (t, J ) 6.7, 2 H), 2.24 (s, 3 H), 2.22 (s, 3 H);
MS (EI) m/z 435 (M + H)+; HRMS m/z (M + H)+ calcd
436.1985, obsd 436.1980.
Com p ou n d 10: 1H NMR (400 MHz, DMSO-d6) δ 10.07 (s,
1 H), 9.14 (s, 1 H), 7.98 (s, 1 H), 7.83 (d, J ) 7.5 Hz, 1 H), 7.41
(m, 4 H), 7.22 (d, J ) 8.5 Hz, 2 H), 7.15 (m, 4 H), 6.90 (t, J )
7.5 Hz, 2 H), 3.54 (s, 2 H), 2.76 (t, J ) 7.5 Hz, 2 H), 2.48 (d, J
) 7.5 Hz, 2 H), 2.23 (s, 3 H); MS (EI) m/z 432 (M + H)+; HRMS
m/z (M + H)+ calcd 432.1923, obsd 432.1911.
Gen er a l p r oced u r e d for Hyd r olysis of Ester s. 2-(3-(3-
(2-Meth yl)P UP a ceta m id op h en yl)u r eid o)a cetic a cid (4).
To a solution of 16 (0.1 g, 0.2 mmol) in 4 mL of DMF was added
aqueous 2 N LiOH (0.5 mL, 1.0 mmol). The resulting solution
was stirred for 2 h and then slowly poured into 150 mL of
aqueous 1 N HCl to give a white precipitate. The solid material
was filtered off, washed with H2O, and dried in vacuo to give
90 mg (95%) of compound 4 as a fine, beige solid (HPLC,
>99%). MS: m/z 476 (M + H)+.
Compounds 2, 4, 5, 6, 7, 8, 9, 10, and 11 were synthesized
using the general procedures b, c, and d. Compounds 3, 12,
and 13 were synthesized using the general procedures c and
d. Analytical samples were obtained from preparative HPLC
purification except compound 4 which was obtained by pre-
cipitation as described in the procedure.
Sp ectr a l Da ta . Com p ou n d 2: 1H NMR (400 MHz, DMSO-
d6) δ 9.98 (s, 1 H), 8.92 (s, 1 H), 7.82 (s, 1 H), 7.75 (d, J ) 7.9
Hz, 2 H), 7.41 (d, J ) 8.5 Hz, 2 H), 7.32 (d, J ) 8.5 Hz, 2 H),
7.15 (d, J ) 8.5 Hz, 1 H), 7.05 (m, 4 H), 6.85 (m, 1 H), 3.46 (s,
2 H), 3.00 (m, 1 H), 2.37 (d, J ) 7.5 Hz, 2 H), 2.15 (s, 3 H),
1.09 (d, J ) 6.9 Hz, 3 H); 13C NMR (100 MHz, DMSO-d6) δ
173.5, 169.5, 153.0, 141.2, 138.7, 137.8, 137.7, 130.5, 129.7,
129.6, 127.7, 127.2, 126.5, 122.9, 121.3, 119.5, 118.4, 43.0, 42.6,
35.7, 22.4, 18.2; FABMS, m/z 446 (M + H)+; HRMS m/z (M +
H)+ calcd 446.2080, obsd 446.2082.
Com p ou n d 3: 1H NMR (400 MHz, DMSO-d6) δ 10.02 (s, 1
H), 9.01 (s, 1 H), 8.06 (d, J ) 8.1 Hz, 1 H), 7.91 (s, 1 H), 7.76
(d, J ) 7.8 Hz, 1 H), 7.42 (d, J ) 8.5 Hz, 2 H), 7.34 (d, J ) 8.5
Hz, 1 H), 7.27 (d, J ) 8.5 Hz, 2 H), 7.08 (m, 4 H), 6.86 (m, 1
H), 4.28 (m, 1 H), 3.48 (s, 2 H), 2.90 (m, 1 H), 2.2.70 (m, 1 H),
2.16 (s, 3 H), 1.70 (s, 3 H); 13C NMR (100 MHz, DMSO-d6) δ
173.5, 169.6, 169.5, 153.0, 138.8, 138.0, 137.8, 132.8, 130.5,
129.8, 129.6 (two peaks), 127.8, 126.5, 122.9, 121.4, 119.3,
118.4, 54.0, 43.0, 36.6, 22.7, 18.2; FABMS, m/z 489 (M + H)+;
HRMS m/z (M + H)+ calcd 489.2138, obsd 489.2122.
Com p ou n d 11: 1H NMR (400 MHz, DMSO-d6) δ 10.30 (s,
1 H), 9.01 (s, 1 H), 7.90 (s, 1 H), 7.82 (d, J ) 7.8 Hz, 1 H), 7.40
(d, J ) 8.5 Hz, 2 H), 7.21 (d, J ) 8.5 Hz, 2 H), 7.14 (m, 2 H),
6.92 (m, 1 H), 4.04 (s, 2 H), 3.73 (s, 2 H), 2.22 (s, 3 H); MS (EI)
m/z 458 (M + H)+; HRMS m/z (M + H)+ calcd 458.0957, obsd
458.0970.
Com p ou n d 12: 1H NMR (400 MHz, DMSO-d6) δ 8.95 (s, 1
H), 7.84 (s, 1 H), 7.77 (d, J ) 7.3 Hz, 1 H), 7.34 (d, J ) 8.6 Hz,
2 H), 7.16 (d, J ) 8.6 Hz, 2 H), 7.06 (m, 2 H), 6.86 (m, 2 H),
3.59 (s, 2 H), 3.52 (s, 2 H), 2.47 (s, 1 H), 2.16 (s, 3 H, Me); 13
C
NMR (75 MHz, DMSO-d6) δ 171.9, 169.7, 157.8, 153.0, 144.6,
139.0, 137.8, 130.5, 129.9, 128.4, 127.7, 126.4, 121.3, 118.4,
110.4, 41.4, 37.2, 18.2; FABMS, m/z 425 (M + H)+; HRMS m/z
(M + H)+ calcd 425.1284, obsd 425.1293.
Com p ou n d 13: 1H NMR (400 MHz, DMSO-d6) δ 8.94 (s, 1
H), 7.83 (s, 1 H), 7.51 (m, 2 H), 7.41-7.29 (m, 5 H), 7.18 (d, J
) 8.5 Hz, 2 H), 7.10 (d, J ) 87.9 Hz, 1 H), 7.05 (d, J ) 7.7 Hz,
1 H), 6.87 (m, 1 H), 6.45 (m, 1 H), 3.75 (s, 2 H), 3.63 (s, 2 H),
2.17 (s, 3 H, Me); FABMS, m/z 501 (M + H)+; HRMS m/z (M
+ H)+ calcd 501.1597, obsd 501.1581.
Ack n ow led gm en t. We thank Diane Leone, William
Yang, and Dan Scott for generating the in vitro biologi-
cal data. Also, we thank Paul Lyne, Parul Matani,
Prashant Singh, Laurie Castonguay and Russ Petter for
critical evaluation of the manuscript.
Com p ou n d 4: 1H NMR (400 MHz, DMSO-d6) δ 10.04 (s, 1
H), 8.97 (s, 1 H), 8.76 (s, 1 H), 7.87 (s, 1 H), 7.83 (d, J ) 8.0
Hz, 1 H), 7.70 (s, 1 H), 7.39 (d, J ) 8.4 Hz, 2 H), 7.22 (d, J )
8.4 Hz, 2 H), 7.20-7.10 (m, 5 H), 6.92 (t, J ) 7.2 Hz, 1 H),
6.30 (br s, 1 H), 3.78 (br d, 2 H), 3.54 (s, 2 H), 2.22 (s, 3 H); 13
C
Refer en ces
NMR (100 MHz, DMSO-d6) δ 172.5, 169.7, 155.4, 153.0, 140.9,
139.9, 138.7, 137.8, 130.5, 129.7, 129.2, 127.8, 126.5, 122.9,
121.3, 118.4, 113.0, 112.6, 109.0, 43.1, 41.6, 18.2; MS (EI) m/z
476 (M + H)+; HRMS m/z (M + H)+ calcd 476.1934, obsd
476.1959.
(1) Abraham, W. M.; Sielczak, M. W.; Ahmed, A.; Cortes, A.;
Lauredo, I. T.; Kim, J .; Pepinsky, B.; Benjamin, C. D.; Leone,
D. R.; Lobb, R. R.; Weller, P. F. a4-Integrins Mediate Antigen-
induced Late Bronchial Responses and Prolonged Airway Hy-
perresponsiveness in Sheep. J . Clin. Invest. 1994, 93, 776-787.

Potent and Novel R4â1 Antagonists
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 14 2993
(2) Chabot, S.; Williams, G.; Yong, V. W. Microglial production of
TNF-alpha is induced by activated T lymphocytes. Involvement
of VLA-4 and inhibition by interferonbeta-1â. J . Clin. Invest.
1997, 100, 604-612.
(3) Leger, O. J .; Yednock, T. A.; Tanner, L.; Horner, H. C.; Hines,
D. K.; Keen, S.; Saldanha, J . Humanization of a mouse antibody
against human alpha-4 integrin: a potential therapeutic for the
treatment of multiple sclerosis. Hum. Antibodies 1997, 8, 3-16.
(4) Seiffge, D. Protective effects of monoclonal antibody to VLA-4
on leukocyte adhesion and course of disease in adjuvant arthritis
in rats. J . Rheumatol. 1996, 23, 2086-91.
(10) Nicklaus, M. C.; Neamati, N.; Hong, H.; Mazumder, A.; Sunder,
S.; Chen, J .; Milne, G. W. A.; Pommier, Y. HIV-1 Integrase
Pharmacophore: Discovery of Inhibitors through 3-Dimensional
Database Searching. J . Med. Chem. 1997, 40, 920-929.
(11) Kiyama, R.; Honma, T.; Hayashi, K.; Ogawa, M.; Hara, M.;
Fujimoto, M.; Fujishita, T. Novel Angiotensin II Receptor
Antagonists. Design, Synthesis, and in Vitro Evaluation of
Dibenzo[a,d]cycloheptene and Dibenzo[b,f]oxepin Derivatives.
Searching for Bioisosteres of Biphenylyltetrazole Using a Three-
Dimensional Search Technique. J . Med. Chem. 1995, 38, 2728-
2741.
(5) Komoriya, A.; Green, L. J .; Mervic, M.; Yamada, S. S.; Yamada,
K. M.; Humphries, M. J . The minimal essential sequence for a
major cell type-specific adhesion site(CS1) within the alterna-
tively spliced Type III connecting segment domain of fibronectin
is Leucine-Aspartic Acid-Valine. J . Biol. Chem. 1991, 266,
15075-15079.
(6) Lin, K. C.; Ateeq, H. S.; Hsiung, H. S.; 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. R.; Adams, S. P. Selective,
Tight-Binding Inhibitors of Integrin R4â1 That Inhibit Allergic
Airway Responses. J . Med. Chem. 1999, 42, 920-934.
(7) Milne, G.; Nicklaus, M.; Wang, S. S Pharmacophores in drug
Design and Discovery. QSAR Environ. Res. 1998, 9, 23-38.
(8) InsightII; Accelrys: San Diego, CA, 1998.
(12) Marriott, D. P.; Dougall, I. G.; Meghani, P.; Liu, Y.; Flower, D.
R. Lead Generation Using Pharmacaophore Mapping and 3D
Database Searching: Application to Muscarinic M3 Receptor.
J . Med. Chem. 1999, 42, 3210-3216.
(13) Available Chemical Directory; MDL: San Leandro, CA, 1998.
(14) J ones, E. Y.; Harlos, K.; Bottomley, M. J .; Robinson, R. C.;
Driscoll, P. C.; Edwards, R. M.; Clements, J . M.; Dudgeon, T.
J .; Stuart, D. I. Crystal structure of an integrin-binding fragment
of vascular cell adhesion molecule-1 at 1.8 Å resolution. Nature
1995, 373, 539-544.
(15) Lobb, R. R.; Antognetti, G.; Pepinsky, R. B.; Burkly, L. C.; Leone,
D. R. et al. A Direct binding assay for the vascular cell adhesion
molecule-1 (VCAM-1) interaction with alpha4 integrins. Cell
Adhes. Commun. 1995, 3, 385-397.
(9) Legion; Tripos: St. Louis, MO, 1998.
J M020054E