Solid-phase synthesis of 2,3-diphenylpropionic acid library as VLA-4 antagonists

Article abstract of DOI:10.1016/j.tetlet.2005.10.087

2,3-Diphenylpropionic acid library for VLA-4 antagonist was synthesized on solid-phase. Comparison of the two synthetic routes via an orthogonal generation of two aromatic amino functional groups are discussed. From this work, several compounds were ident

Full text of DOI:10.1016/j.tetlet.2005.10.087

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 9035–9038
Solid-phase synthesis of 2,3-diphenylpropionic acid library as
VLA-4 antagonists
Yoichiro Hoshina,a,* Satoru Ikegami,a Kyoko Fujimoto,a
Akihiko Okuyama,a Tatsuhiro Harada,a Susan James,b Peter G. Griffiths,b
Zemin Wu,b Michael Lillyb and Andrew M. Brayb,ꢀ
aCentral Research Laboratories, Kaken Pharmaceutical Co., Ltd, 14 Shinomiya-minamikawara, Yamashina, Kyoto 607-8042, Japan
bMimotopes Pty Ltd, 11 Duerdin Street, Clayton, Victoria 3168, Australia
Received 19 September 2005; revised 18 October 2005; accepted 20 October 2005
Available online 10 November 2005
Abstract—2,3-Diphenylpropionic acid library for VLA-4 antagonist was synthesized on solid-phase. Comparison of the two syn-
thetic routes via an orthogonal generation of two aromatic amino functional groups are discussed. From this work, several com-
pounds were identified as potent VLA-4 antagonists.
ꢀ 2005 Elsevier Ltd. All rights reserved.
There are many well-known, drug-like (druggable)
S
O
structural templates in medicinal chemistry. The phenyl-
alanine skeleton is one of those important templates.
The angiotensin converting enzyme inhibitor alacepril1
and the insulin promoter nateglinide2 are the two key
examples based on this template (Fig. 1). Another struc-
turally related chemical class is represented by the
recently synthesized VLA-4 antagonists.3
O
O
O
N
N
CO2H
N
CO2H
H
H
Alacepril
Nateglinide
Cl
Cl
H
H
N
N
O
O
Cl
O
Cl
Br
O
The integrin, very late antigen-4 (VLA-4), is a hetero di-
meric adhesion molecule (a4b1) expressed on the surface
of leukocytes, which binds vascular cell adhesion mole-
cule-1 (VCAM-1) on endothelial cell surfaces.4 Small
molecule antagonists for VLA-4 should be useful in
the treatment of chronic inflammatory diseases such as
asthma,5 multiple sclerosis,6 and rheumatoid arthritis.7
In our research on VLA-4, we have reported the discov-
ery of a potent and orally bioavailable antagonist 1
(Fig. 1), which is based on the 2,3-diphenylpropionic
acid template.8 This template was successfully developed
as a bioisostere of phenylalanine derivatives, such as 2,9
during the effort to improve the bioavailability and
N
CO2H
N
CO2H
H
MeO
2
1
Figure 1. Chemical structure of phenylalanine derivatives as drugs
(alacepril and nateglinide), the VLA-4 antagonist having 2,3-diphenyl-
propionic acid template 1, and one of the potent VLA-4 antagonist
having phenylalanine skeleton 2.
pharmacokinetic properties of the earlier lead com-
pounds. Moreover, the 2,3-diphenylpropionic acid tem-
plate has the potential to replace other phenylalanine
derivatives that require peptide backbone modification.
In order to examine the potential ability of the 2,3-
diphenylpropionic acid template, we chose to develop
an efficient library synthesis of diverse substituted 2,3-
diphenylpropionic acids. In this letter, we report a solid-
phase synthesis of those derivatives as VLA-4 potential
antagonists.
Keywords: Solid-phase synthesis; 2,3-Diphenylpropionic acid; VLA-4
antagonist.
*
Corresponding author. Tel.: +81 75 594 0787; fax: +81 75 594
0790; e-mail: hoshina_yoichiro@kaken.co.jp
ꢀ Present address: PharmaBank, Melbourne, Vic., Australia.
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.10.087

9036
Y. Hoshina et al. / Tetrahedron Letters 46 (2005) 9035–9038
H
NO2
N
R3
R2
O2N
HO
O
O
a, b
CO2H
H2N
R1
N
CO2Et
CO2H
MeO
MeO
5
6
3
NO2
c,d
e
Reduction
FmocHN
MeO
CO2H
and acylation
Deprotection,
reductive alkylation,
and acylation
NO2
O
7
O
N
NO2
H
FmocHN
MeO
O
O
O
O
O
f, g, h
HMP linker
FmocHN
MeO
O
O
4
4
Figure 2. Synthetic strategy of 2,3-diphenylpropionic acid library on
solid support.
8
Scheme 1. Reagents and conditions: (a) EtOH, H2SO4, 99%; (b) Ref.
8; (c) NaOH aq, EtOH, 95%; (d) FmocCl, acetone, H2O, 98%; (e) allyl
(4-hydroxymethyl)phenoxyacetate, EDCI, DMAP, CH2Cl2, quant.; (f)
Pd(PPh3)4, Et3SiH, AcOH, CH2Cl2, 94%; (g) SynPhaseTM Lantern
(aminomethylated), DIC, HOBt, DMF, CH2Cl2; (h) Ac2O, DIEA,
DMF.
The synthetic strategy is shown in Figure 2. We designed
the immobilized compound 4 as an important key inter-
mediate to provide the target compounds 3 with three
diversity points (R1, R2, R3). 2,3-Diphenylpropionic
acid with orthogonally masked amino groups on each
benzene ring was coupled onto SynPhaseTM Lanterns10
through 4-hydroxymethyl phenoxyacetic acid (HMP)
linker.11 The amine protected by 9-fluorenylmethoxy-
carbonyl (Fmoc) and the nitro group can be selectively
converted to the required amino groups as required.
For the purpose of creating a versatile VLA-4 antago-
nist library, the amino group at 30-position on the 2-phen-
yl group would be converted to the tert-amide by
reductive alkylation followed by acylation, and the
other to the sec-amide by acylation. After full function-
alization, trifluoroacetic acid (TFA) mediated cleavage
would afford the target compound 3.
Route A
a, b, c
NO2
R2
O
R1
N
O
4
O
MeO
9
d,e
d,e
Route B
H
H
N
N
R3
R3
R2
O
O
N
O
a, b, c
R1
FmocHN
MeO
O
O
O
The pivotal intermediate 4 was prepared as shown in
Scheme 1.8,12 After the hydrolysis of the ethyl ester 6,
the amino group was protected by Fmoc to give 7. Car-
boxylic acid 7 was reacted with an HMP linker segment
to yield the corresponding ester 8. The allyl ester group
of 8 was selectively removed by a palladium-catalyzed
reaction. The resulting acid was attached onto Syn-
PhaseTM Lantern (aminomethylated) by amide bond for-
mation to afford 4. The loading of 4 was determined as
29.6 lmol/Lantern by quantitative Fmoc determination.
O
MeO
11
10
O
f
O
N
=
H
3
Scheme 2. Reagents and conditions: (a) 20% piperidine/DMF, 1 h; (b)
R1CHO (36 equiv, 1.0 M), NaBH3CN (18 equiv, 0.5 M), 1% AcOH/
DMF, 14 h; (c) R2COCl (28 equiv, 0.5 M), DIEA (56 equiv, 1.0 M),
20% DMF/CH2Cl2, 14 h; (d) SnCl2Æ2H2O (70 equiv, 2.0 M), 50%
DMF/CH2Cl2, 4 h; (e) R3COCl (17 equiv, 0.5 M), DIEA (34 equiv,
1.0 M), 20% DMF/CH2Cl2, 2 h; (f) TFA, 1 h.
Since the two amino groups of 4 can be generated
orthogonally, there are two possible synthetic routes,
A and B, to give 3 as outlined in Scheme 2. We assessed
both routes. First, the Fmoc group of 4 was removed
with piperidine/DMF solution to follow route A. The
amino group was reductively alkylated with isobutyr-
aldehyde under various conditions to afford the desired
mono-alkylated amine. As shown in Table 1, the target
mono-isobutylated compound 13 could be obtained in
excellent HPLC purity13 (entry 3). The di-isobutylated
amine 15 was not detected at all. In addition, the reac-
tion with n-butyraldehyde also yields mono-butylated
amine in good selectivity. This selectivity is probably
due to the steric hindrance caused by the methoxy
group at 40-position. This secondary amine was then
reacted with 2,2-dimethylpropionic acid in various con-
ditions such as DIC–HOBt–DMAP, TFFH–DIEA, or
HATU–DIEA, but, the desirable amide 9 was not
obtained. However, treatment of pivaloyl chloride in
the presence of DIEA gave amide 9 in good conver-
sion and purity. Then, to make it easy to confirm the

Y. Hoshina et al. / Tetrahedron Letters 46 (2005) 9035–9038
Table 1. Optimization of the reductive alkylation
9037
1
1)
2)
R CHO
NO2
NO2
NO2
NaBH3CN (18 eq., 0.5 M)
1% AcOH in DMF, rt
H
N
+
R1
R42N
MeO
H2N
O
CO2H
CO2H
TFA
O
MeO
MeO
R4 = H
R4 = CH2CH(CH3)2
R4 = CH2(CH2)2CH3
14:
15:
16:
12
13
Entry
R1CHO
Concentration (M)
Time (h)
HPLC area (%)a
14
22
5.6
13
1
2
3
4
5
6
7
8b
(CH3)2CHCHO
(CH3)2CHCHO
(CH3)2CHCHO
CH3CH2(CH3)CHCHO
(CH3)3CCHO
PhCHO
4-MeOC6H4CHO
CH3(CH2)2CHO
0.2
0.5
1.0
1.0
1.0
1.0
1.0
0.5
3
3
72
88
94
94
79
90
50
89
14
14
14
14
14
3
0
0
14
4.2
43
0
a HPLC peak areas % at 254 nm.
b 16 was detected, HPLC peak area = 2.4%.
structure, we synthesized our very standard compound 1
from this amide 9 in the first place. The nitro group of 9
was reduced with tin(II) chloride dihydrate followed by
an acylation with 2,6-dichlorobenzoyl chloride to give
amide 10. Cleavage with TFA (100%) afforded the crude
compound 1 in 89% HPLC purity, and 64% isolate yield
after purification with preparative TLC. The spectral
properties of 1 were identical with those of 1 prepared
via a conventional liquid phase synthetic method. On
the basis of these results, route A was confirmed prac-
tical enough for the VLA-4 library.
ized by LC–MS and selected members of the library
1
gave excellent H NMR spectra.14
Some of the selected unpurified compounds were
assayed for their ability to inhibit the binding of VLA-4
expressing human leukemia cells (HL-60) to human
VCAM-1 expressed on Chinese hamster ovary (CHO)
cells.15 As shown in Table 3, most compounds were
found to be less potent than 1 (iaL). In most cases, the
compounds that have the building block L retained
the inhibition activities. Compound iaK has a similar
potency to 1.
Route B was examined in the same manner, that is with
the same reaction conditions for each step as optimized
in route A. We chose 1 as the first compound to evaluate
route B. The nitro group of 4 was reduced to the corre-
sponding amine followed by the acylation with 2,6-di-
chlorobenzoyl chloride to give amide 11. After the
deprotection of Fmoc group, the reductive alkylation
and subsequent acylation afforded amide 10. Treatment
of 10 with TFA gave 1 in 89% HPLC purity, and 70%
isolate yield. These results indicated that there was no
major difference in purity and yield between the routes
A and B. By using either route, the lability characteris-
tics of the R groups (i.e., to Fmoc deprotection or nitro
reduction conditions) can be accommodated.
In conclusion, we have developed a general solid-phase
library synthesis of substituted 2,3-diphenylpropionic
acids. The 96-member library was synthesized in good
purity, and several library members showed potent
activity as VLA-4 antagonist. We also demonstrated
that there were two practical orthogonal routes from
the key intermediate 4. This synthetic method will be
Table 2. HPLC purity (peak area % at 254 nm) of the products
R3COCl
R1CHO/R2COCl
ia
ib
ic
id
iia
iib
iic
iid
A
B
C
D
E
77
82
76
79
75
86
86
85
81
84
82
83
74
83
77
79
74
80
81
82
80
87
83
83
74
75
76
83
74
76
87
81
82
85
83
84
74
80
85
88
81
84
79
85
81
82
78
88
We mainly selected the building blocks in consideration
of a diversity of R3, since we were interested in exploring
alternatives for 2,6-dichlorobenzoyl group (L) or 3,5-
dichloroisonicotinoyl group (K). Two aldehydes for
R1, four acid chlorides for R2 and 12 acid chlorides
for R3 were selected as shown in Figure 3. We synthe-
sized 96 compounds via route A, by using a split and
pool method with color tagged Lanterns. As a result,
the target compounds were obtained in 81% average
HPLC purity (Table 2). All compounds were character-
F
G75
H
I
J
K
L
79
72
77
70
80
80
80
77
68
63
93
89
83
87
75
91
80
80
69
61
90
90
84
80
73
94
82
84
68
71
90
83
85
82
75
92
84
85
74
73
92
84
82
80
74
92
78

9038
Y. Hoshina et al. / Tetrahedron Letters 46 (2005) 9035–9038
Ohara and Ms. Kazumi Sumita for measurement of
NMR and mass spectra.
R1CHO
CHO
R2COCl
COCl
COCl
COCl
CHO
COCl
i
ii
c
d
a
b
References and notes
R3COCl
COCl
COCl
COCl
1. Mannhold, M.; Castaner, J. Drugs Future 1986, 11, 357–360.
2. Porous, J.; Graul, A.; Castaner, J. Drugs Future 1993, 18,
503–505.
3. Yang, G. X.; Hagmann, W. K. Med. Res. Rev. 2003, 23,
369–392.
4. Elices, M. J.; Osborn, L.; Takada, Y.; Crouse, C.;
Luhowsky, S.; Hemler, M. E.; Lobb, R. R. Cell 1990,
60, 577–584.
5. (a) Milne, A. A. Y.; Piper, P. J. Eur. J. Pharmacol. 1995,
282, 243–249; (b) Abraham, W. M.; Ahmed, A.; Sielczak,
M. W.; Narita, M.; Arrhenius, T.; Elices, M. J. Am. J.
Respir. Crit. Care Med. 1997, 156, 696–703.
6. (a) Yednock, T.; Cannon, C.; Fritz, L. C.; Sanchez-
Madrid, F.; Steinman, L.; Karlin, N. Nature 1992, 356,
63–66; (b) Keszthelyi, E.; Karlik, S.; Hyduk, S.; Rice, G.
P. A.; Gordon, G.; Yednock, T.; Horner, H. Neurology
1996, 47, 1053–1059.
COCl
A
D
B
C
G
F
COCl
F3C
COCl
F
COCl
F
O
COCl
Cl
Cl
F
Cl
F
E
F
H
Cl
Cl
S
COCl
Cl
COCl
Cl
NC
COCl
COCl
N
I
J
K
L
Figure 3. Building blocks.
7. Seiffge, D. J. Rheumatol. 1996, 23, 2086–2091.
8. Hoshina, Y.; Ikegami, S.; Okuyama, A.; Fukui, H.;
Inoguchi, K.; Maruyama, T.; Fujimoto, K.; Matsumura,
Y.; Aoyama, A.; Harada, T.; Tanaka, H.; Nakamura, T.
Bioorg. Med. Chem. Lett. 2005, 15, 217–220.
Table 3. VCAM-1/VLA-4 inhibition activity of the selected com-
pounds
Building blocks
IC50 (lM)
9. Sidduri, A.; Tilley, J. W.; Lou, J. P.; Chen, L.; Kaplan, G.;
Campbell, R.; Guthrie, R.; Huang, T.-N.; Rowan, K.;
Schwinge, V.; Renzetti, L. M. Bioorg. Med. Chem. Lett.
2002, 12, 2479–2482.
10. Parsons, J. G.; Sheehan, C. S.; Wu, Z.; James, I.; Bray, A.
M. Methods Enzymol. (Comb. Chem., Part B) 2003, 369,
39–74.
iaA
iaB
iaC
iaD
iaE
iaF
iaG
iaH
iaI
5.9
>10
2.0
>10
3.1
0.27
0.12
0.15
>10
11. Sheppard, R. C.; Williams, B. J. Int. J. Peptide Protein
Res. 1982, 20, 451–454.
12. Patek, M.; Smrcina, M.; Nakanishi, E.; Izawa, H. J.
Comb. Chem. 2000, 2, 370–377.
iaJ
>10
iaK
iaL
ibL
icL
idL
iiaL
iibL
iicL
iidL
0.0017
0.002
3.0
0.0005
0.27
0.005
0.01–0.1
0.0045
0.039
13. Reversed phase HPLC was performed on Hewlett Packard
HP-1100 series system with a linear gradient of 10% of B
(0–1 min), 10–100% of B (1–10 min), 100% of B (10–
12 min) using 0.1% formic acid in water as solvent A, 0.1%
formic acid in acetonitrile as solvent B (1.0 mL/min). The
column was GL Sciences Inc. Inertsil ODS-3, 3 lm,
4.6 · 75 mm. Peak areas were integrated with 254 nm.
14. Selected data for iaK: 3-{4-[(3,5-dichloropyridine-4-car-
bonyl)amino]phenyl}-2-{3-[(2,2-dimethylpropionyl)iso-
butylamino]-4-methoxyphenyl}propionic acid: 1H NMR
(400 MHz, DMSO-d6) d: 0.69–0.94 (15H, m), 1.38–1.67
(1H, m), 2.43–2.58 (1H, m), 2.80–3.04 (1H, m), 3.18–3.40
(1H, m), 3.77 (3H, s), 3.81–3.97 (2H, m), 6.96–7.08 (1H,
m), 7.27 (1H, d, J = 8.7 Hz), 7.09–7.18 (2H, m), 7.26–7.36
(1H, m), 7.38–7.53 (2H, m), 8.78 (2H, s), 10.79 (1H, br s);
13C NMR (100 MHz, DMSO-d6) d: 20.0, 20.2, 26.2, 28.4,
38.1, 51.2, 55.4, 56.5, 112.0, 119.3, 128.1, 128.7, 129.4,
131.1, 131.2, 135.1, 136.1, 142.5, 147.7, 154.3, 159.6, 174.4,
176.5; HRMS: calcd for C31H36 35Cl2N3O5 (M+H)+
600.2032, found 600.2042.
applicable for a variety of propionic acid derivatives
that have two aromatic amino groups. This synthetic
route is useful for the discovery of new drug candidates
derived from N-acylated phenylalanine derivatives.
Acknowledgments
We appreciate the constructive contributions of Profes-
sor Takashi Takahashi (TIT), Dr. Hiroshi Tanaka
(TIT), and Dr. Masanao Shimano (Kaken) for discus-
sions relevant to this study. We thank Ms. Miyako
15. Hession, C.; Tizard, R.; Vassallo, C.; Schiffer, S. B.; Goff,
D.; Moy, P.; Chi-Rosso, G.; Luhowskyj, S.; Lobb, R.;
Osborn, L. J. Bio. Chem. 1991, 266, 6682–6685.