Substituted 3-amino biaryl propionic acids as potent VLA-4 antagonists.

Article abstract of DOI:10.1016/S0960-894X(02)00460-2

A series of substituted N-(3,5-dichlorobenzenesulfonyl)-(L)-prolyl- and (L)-azetidyl-beta-biaryl beta-alanine derivatives was prepared as selective and potent VLA-4 antagonists. The 2,6-dioxygenated biaryl substitution pattern is important for optimizing potency. Oral bioavailability was variable and may be a result of binding to circulating plasma proteins.

Full text of DOI:10.1016/S0960-894X(02)00460-2

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2415–2418
Substituted 3-Amino Biaryl Propionic Acids as Potent VLA-4
Antagonists
Ihor E. Kopka,a,* Linus S. Lin,a Richard A. Mumford,b Thomas Lanza, Jr.,a
Plato A. Magriotis,a David Young,a Stephen E. DeLaszlo,a Malcolm MacCoss,a
Sander G. Mills,a Gail Van Riper,b Ermengilda McCauley,b Kathryn Lyons,c
Stella Vincent,c Linda A. Egger,b Usha Kidambi,b Ralph Stearns,c Adria Colletti,c
Yohannes Teffera,c Sharon Tong,a Karen Owens,a Dorothy Levorse,a
John A. Schmidtb and William K. Hagmanna
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000 RY 123-136, Rahway, NJ 07065, USA
bDepartment of Inflammation and Rheumatology Research, Merck Research Laboratories, Rahway, NJ 07065, USA
cDepartment of Drug Metabolism, Merck Research Laboratories, Rahway, NJ 07065, USA
Received 14 March 2002; accepted 20 May 2002
Abstract—A series of substituted N-(3,5-dichlorobenzenesulfonyl)-(l)-prolyl- and (l)-azetidyl-b-biaryl b-alanine derivatives was
prepared as selective and potent VLA-4 antagonists. The 2,6-dioxygenated biaryl substitution pattern is important for optimizing
potency. Oral bioavailability was variable and may be a result of binding to circulating plasma proteins. # 2002 Elsevier Science
Ltd. All rights reserved.
VLA-4 (a4b1; CD49d/CD29; very late antigen-4) is a
member of the integrin family that is expressed on lym-
phocytes.1 Its ligands include vascular cell adhesion
molecule-1 (VCAM-1), which is expressed on activated
endothelial cells at sites of inflammation and peptides
derived from the type III connecting segment (CS-1)
domain of fibronectin. Antibodies to a4 are effective
inhibitors of leukocyte infiltration and prevent tissue
damage in several animal models of inflammation.2
Inhibition of VLA-4 may reduce the migration and/or
activation of cell types important to sustaining a pro-
longed inflammatory response. VLA-4 inhibitors may
be considered for treatment of asthma,3 multiple
sclerosis,4 and rheumatoid arthritis.5
(rat), 7% (rhesus monkey).8 Based on these results and
the potency enhancement found with substituted
biphenylalanines, N-sulfonylated b-biphenylalanine-b-
amino dipeptides were pursued. The hope was that the
b-biaryl alanine scaffold would imparhtte same
improved potency and pharmacokinetic properties seen
in the b-phenyl-b-alanine series.8
We have previously reported the structure–activity rela-
tionship (SAR) of several series of sulfonylated dipep-
tide derivatives.6À8 The pharmacokinetic (PK) profiles
of 1 and 2 (Fig. 1) suggesthtateach has reasonably
good oral bioavailability in several species (1: F%: 49%
(rat), 27% (dog) 8% (rhesus monkey); 2: F%: 43%
Figure 1. Inhibition of VLA-4 by b-phenyl-b-alanine derivatives.
Two routes were employed to synthesize the b-biaryl
b-alanines. Boc-(S)-4-hydroxyphenyl glycine methyl
ester was treated with triflic anhydride and homo-
logated with an aryl boronic acid. (Scheme 1). The ester
was then hydrolyzed, the free acid treated with dis-
obutylchloroformate and reacted with diazomethane to
*Corresponding author. Fax: +1-732-594-5966; e-mail: ihor_kopka@
merck.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00460-2

2416
I. E. Kopka et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2415–2418
Scheme 2. (a) (i) 48% HBF4, EtOH, (ii) isoamyl nitrite 5 ꢁC; (b)
methyl acrylate, 10% Pd/C, MeOH, reflux 3 h, 70%; (c) (i) 2 equiv (S)-
(À)-N-benzyl a-methylbenzyl amine, 2 equiv n-BuLi, À78 ꢁC, 30 min,
(ii) NH4Cl (aq), 66%; (d) 10% Pd/C, H2, (50 psi), MeOH, 93%; (e)
(BOC)2O, DIEA; (f) triflic anhydride/Et3N, CH2Cl2 0 ꢁC, 2 h 85%; (g)
phenylboronic acid, Pd(PPh)4, K2CO3, (toluene/EtOH, 4/1) 60–90 ꢁC,
2–3 h 75%.
Scheme 1. (a) Triflic anhydride/Et3N, CH2Cl2 0 ꢁC, 2 h 85%; (b)
phenylboronic acid, Pd(PPh)4, K2CO3, (toluene/EtOH, 4/1) 60–90 ꢁC,
2–3 h 75%; (c) LiOH, (EtOH/H2O), H+
;
(d) (i) isobutyl-
chloroformate, Et3N, (ii) diazomethane, Et2O; (e) silver benzoate,
Et3N; (MeOH, dioxane, 0 ꢁC, 2 h, 45% overall; (f) HCl (g), EtOAc,
0 ꢁC 1 h, 95%; (g) PyBop, iPr2NEt, CH2Cl2, 0 ꢁC to rt; (h) (i) LiOH,
EtOH/H2O, rt, (ii) H+, 60% overall.
form the diazoketone to effect the Arnt–Eistert reaction.
The subsequentWolff rearrangementwith silver benzo-
ate in dioxane/methanol afforded the Boc-3-(R)-amino-
3-(4-biphenyl)propionic acid, methyl ester. Unfortu-
nately, substantial quantity (ꢀ25%) of the undesired
(S)-isomer was also obtained resulting from the unex-
pected isomerization of the stereogenic center. The
unseparated isomeric mixture was carried through the
reaction sequence by coupling the deprotected biaryl
ester to (N)-(3,5-dichlorobenzenesulfonyl)-2-(S)-proline.
Scheme 3. (a) n-BuLi, THF, À78 ꢁC; (b) B(OCH3)3; (c) 2 N HCl; (d)
PhLi, ether, 0 ꢁC, 18 h.
nic acids were synthesized by directed lithiation (s-BuLi
or PhLi) of the parent arene (Scheme 3).
The directing ligand in the ring lithiation reaction is the
methoxy group. Lithiation ortho to the directing group
was generally obtained in good yield when it was effec-
ted at À78 ꢁC. With examples 11 and 12, we raised the
temperature to 0 ꢁC to accelerate the metallation pro-
cess atht e expense of a lower yield. The 2,6 bis-phenol
34 was protected as the bis-methoxymethyl ether since
ortho metallation and acid deprotection of the MOM
protecting group were easily achieved. The same sequence
was used for the monophenol derivative 33. We dis-
covered that one could directly mono- and difluorinate the
2,6-dimethoxybenzeneboronic acid with the solid phase
supported fluorinating reagents SelectFlor1 (Scheme 4).10
Diasteroisomers 3 and 4 were separated by flash column
chromatography. This approach has two limitations.
The firstis hte Suzuki coupling of our desired aryl
boronic acid occurs prior to the Arnt–Eistert homo-
logation sequence, limiting the synthetic flexibility for
variations in that ring. The second is the partial iso-
merization of the b-carbon of the biaryl amino acid. A
more flexible and efficientapproach was developed.
A chiral synthetic strategy was employed by starting
with (R)-b-4-hydroxyphenylglycine9 (Scheme 2). This
method gave high stereoselectivity for the Michael
addition (>98% d.e.) and an excellentoverall yield for
the protected amino acid triflate. The triflate was sub-
sequently used in the Suzuki coupling reaction with a
wide variety of aryl boronates to form the desired sub-
stituted b-biaryl-b-amino acid intermediates. These
b-aminoesters were used as before to obtain the test
compounds 5–10.
The inhibition of binding of a soluble 125I-VCAM-Ig
fusion protein to VLA-4 on suspended Jurkat cells6 is
The aryl boronates were obtained from commercial
sources or synthesized by metal halogen exchange of the
aryl halide with n-butyl lithium and treatment with
methyl borate. Some of the 2,6-dioxygenated arylboro-
Scheme 4. (a) 1.1 equiv SelectFlor1, 22 ꢁC, 24 h, CH3CN; (b) 3.0 equiv
SelectFlor1, 22 ꢁC, 24 h, CH3CN.

I. E. Kopka et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2415–2418
2417
Table 1. Inhibition of VLA-4 by substituted N-sulfonyl-prolyl-b-
amino acids (IC50, nM)
summarized in Table 1. Several trends are noteworthy.
First, the (R)-configuration of the substituted b-biaryl
b-amino acid diastereomer is the more active derivative
in the series (3,4 and 5,6). This corresponds to the same
orientation as the substituted N-sulfonylated-prolyl-
biphenylalanine derivatives in the a-amino acid series.6
The 3, 4, and 5-monofluorinated pairs 21–22, 17–18,
and 13–14, respectively, were less potent than the parent
compounds 6 and 7 while 15–16 were slightly more
potent than the parent compound. The a-methyl-
(l)-proline derivatives of each matched pair is sub-
stantially less potent than the des-methyl analogue. The
difference between matched pairs is more pronounced in
this b-series (7- to 40-fold) than in the a-amino acid
series (2- to 5-fold).6 The difference in potency between
cognate pairs in this b-amino acid series becomes much
smaller in the case of the 2,6-dioxygen substituted biaryl
derivatives. The 2,6-dimethoxybiaryl pair 27 and 28 are
among the most potent derivatives in the series as they
are in the a-amino acid series. The monomethoxy-
hydroxy and bis-hydroxy derivatives 33, 34 are equipo-
tent to the bis-methoxy congener 28. The 2,6 dioxogen
distribution is critical in achieving potency. Ring fluori-
nated derivatives 13, 15, 17, and 21 as well as the tri-
fluoromethoxy analogue 19 show only a small shiftin
binding potency from the corresponding hydrogen ana-
logues. Binding data for two compounds described in
the literature as a potent, selective VLA-4 antagonist
No.
R
2
3
4
5
6
VLA-4
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
H (R)
H (S)
H (R) OCH3
H (S) OCH3
CH3 OCH3
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
F
—
—
—
—
—
—
—
—
—
—
—
—
—
—
F
—
—
—
—
—
—
—
—
OCH3
OCH3
F
F
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
F
F
—
—
—
—
—
—
—
OCH3
—
—
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OH
0.8
196
0.6
5.4
10.0
0.5
1.1
7.1
0.2
2.2
0.5
17.5
0.5
3.5
1.4
38.7
2.7
40.0
1.5
31.4
1.7
H
H
CN
OCF3
CH3 OCF3
OCH3
CH3 OCH3
OCH3
CH3 OCH3
OCH3
CH3 OCH3
OCH3
CH3 OCH3
OCF3
CH3 OCF3
OCH3
CH3 OCH3
H
H
H
H
F
F
F
H
—
—
—
—
H
—
—
—
OCH3
—
—
—
—
—
—
—
—
—
—
F
(BIO-1211)3 and
a
dual a4b7/VLA-4 antagonist
(TR-14035)11 are included in Table 2 for comparative
purposes. All of the derivatives in Table 1 show >500-
fold specificity for a4b1 over a4b7 (data not shown).
H
CH3
H
CH3
H
CH3 OCH3
CH3 OCF3
H
CH3 OCH3
CH3 OCH3
CH3
CH3
F
F
H
H
OCH3
OCH3
—
F
—
8.4
6.0
OCH3
OCH3
—
—
—
F
F
F
—
—
F
88.8
0.13
0.48
1.15
0.2
1.36
1.18
0.40
0.42
0.13
0.11
—
—
—
—
—
F
Based on these results, a number of N-arylsulfonyl-
l-azetidinyl derivatives were synthesized to compare
binding potencies with selected prolyl derivatives. These
results are shown in Table 2. The proline-azetidine pairs
5,35 and 27,36 were aboutequipoetntin VLA-4 bind-
ing. All of the 2,6-dioxo species in Table 2 showed
excellent binding to the receptor and at least >500-fold
specificity for a4b1 over a4b7.
29
30
31
32
33
34
OCH3
OH
OH
—
—
BIO-1211
TR-14035
The pharmacokinetic properties of selected compounds
were also determined (Table 3). Generally, the com-
pounds had low to moderate oral bioavailability and
moderate to fast plasma clearance rates. Several trends
are apparent in the proline paired sets 13–28. The des-
methyl proline derivative are cleared from circulation
much faster than the a-methyl analogues. Metabolism
studies concluded that hydrolysis of the amide bond by
circulating peptidases does not occur to any substantial
degree. The bioavailability of the a-methyl proline ana-
logues of each paired set was substantially greater than
that of the des-methyl proline derivatives.
Table 2. Inhibition of VLA-4 by substituted N-sulfonyl-azetidinyl-b-
amino acids (IC50, nM)
No.
2
5
6
VLA-4a
Most of the compounds were excreted into the bile as
intact parent or as the conjugated acyl-glucuronide.12
The low clearance, along with the very low volume of
distribution, suggests that high plasma protein binding
may be a significant contributing factor to the observed
PK profile. Many of the compounds in Table 3 were
found to be tightly bound to plasma proteins (>99.9%;
data not shown). The same PK profile is seen with the
35
36
37
38
39
40
OCH3
OCH3
OCF3
—
—
—
F
—
—
—
OCH3
OCH3
0.81
0.19
0.19
0.14
0.20
0.31
OCH3
OH
(CH2CH2)CHO
OCH3
OH
(CH2CH2)CHO
a125I-VCAM-Ig.

2418
I. E. Kopka et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2415–2418
Table 3. Pharmacokinetic parametersa of selected compounds
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 for
dosing of animals for pharmacokinetic evaluation.
No.
F
b (%)
Clp (mL/kg/min)
Vdss (L/kg)
t1/2c (h)
1
4
5
43.2
8.9
22.1
14.6
1.1
13.1
2.3
16.3
2.3
14.6
1.5
48.1
2.5
51.6
1.7
10.4
24.9
2.3
1.5
1.1
1.8
1.3
3.0
1.2
0.7
1.3
ND
1.7
ND
0.9
1.8
1.8
ND
1.6
3.2
2.4
ND
0.9
0.7
0.61
0.08
0.58
0.08
0.19
0.13
NDd
0.15
ND
0.12
1.4
0.1
0.26
1.1
0.08
0.12
0.25
0.48
0.37
21.6
13.9
19.4
12.2
42.0
8.2
55.4
19.6
38.7
7.3
25.7
1.0
45.1
2.6
21.9
1.0
18.0
1.0
13
14
15
16
17
18
25
26
27
28
28e
29
34
35
35e
36
36e
References and Notes
1. Hemler, M. E. Annu. Rev. Immunol. 1990, 8, 365.
2. Elices, M. J. Curr. Opin. Anti-inflam. Immun. 1999, 1, 15.
3. 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. R.;
Adams, S. P. J. Med. Chem. 1999, 42, 90.
4. Tubridy, N.; Behan, P. O.; Capildeo, R.; Chaudhuri, A.; For-
bes, R.; Hawkins, C. P.; Hughes, R. A. C.; Mosely, I. F.; Mac-
Manus, D. G.; Donoghue, S.; Miller, D. H. Neurology 1999, 53,
466.
5.0
22.1
13.9
21.7
5. Seiffge, D. J. Rheumatology 1996, 23, 2086.
6. 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.; Wil-
son, J.; MacCoss, M.; Mills, S. G.; Van Riper, G.; McCauley,
E.; Egger, L. A.; Kidambi, U.; Lyons, K.; Vincent, 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.
aSprague–Dawley rats.
bDose: 1 mg/kg iv; 2 mg/kg po.
ct1/2=plasma half-life(0À8 h)
.
dND=Notdetermined.
eRhesus monkey. Dose: 0.75 mg/kg iv; 1.5 mg/kg po.
7. Kopka, I. E.; Young, D.; Lin, L. S.; Mumford, R. A.;
MacCoss, M.; Mills, S. G.; Van Riper, G.; McCauley, E.;
Egger, L. A.; Kidambi, U.; Schmidt, J. A.; Lyons, K.; Stearns,
R.; Vincent, S.; Colletti, A.; Wang, Z.; Tong, S.; Wang, J.;
Zheng, S.; Owens, K.; Levorse, D.; Hagmann, W. K. Bioorg.
Med. Chem. Lett. 2002, 12, 637.
8. Lin, L.; Kopka, I. E.; Mumford, R.; Kamenecka, T.; Lanza,
T.; Magriotis, P.; Young, D.; DeLaszlo, S.; Van Riper, G.; Lyons,
K.; Vincent, S.; Tong, S.; Egger, L.; MacCoss, M. G.; Schmidt, J.;
Hagmann, W. Bioorg. Med. Chem. Lett. 2002, 12, 611.
9. Davies, S.; Ochihara, I. Tetrahedron: Asymmetry 1991, 2, 183.
10. SelectFlor1 is a trademark product registered to Allied-
Signal Inc.
11. Sircar, I.; Gudmundsson, K. S.; Martin, R.; Nomura, S.;
Jayakumar, H.; Nowlin, D. M.; Caradelli, P. M.; Mah, J. R.;
Castro, M. A.; Cao, Y.; Griffith, R. C.; Lazarides, E. Abstracts
of Papers, 218th National Meeting of the American Chemical
Society, New Orleans, LA, Aug, 1999; American Chemical
Society: Washington, DC, 1999; MEDI 59.
azetidine derivatives in Table 2. For compounds with
long plasma half-lives (13, 34, and 35), the plasma con-
centration in rats versus time curves (AUC) exhibit an
initial rapid drop in concentration followed by low
(ꢀ1 nM) but sustained circulating plasma drug levels in
vivo.
In summary, N-sulfonylated dipeptide b-biaryl-b-amino
acid derivatives are potent and specific antagonists of
VLA-4 and compliment the corresponding substituted
biphenylalaine analogues. The 2,6-dioxygenated biaryl
substitution pattern is important for optimizing
potency in both the azetidine and the proline series.
The variable oral bioavailability, clearance and poor
distribution may be directly attributed to the binding
of these peptide-like compounds to circulating plasma
proteins.
12. Results to be published in due course.