De novo design, synthesis, and in vitro activity of LFA-1 antagonists based on a bicyclic[5.5]hydantoin scaffold

Article abstract of DOI:10.1016/j.bmcl.2004.12.007

LFA-1 (leukocyte function-associated antigen-1), is a member of the β₂-integrin family and is expressed on all leukocytes. The LFA-1/ICAM interaction promotes tight adhesion between activated leukocytes and the endothelium, as well as between T cells and antigen-presenting cells. Evidence from both animal models and clinical trials provides support for LFA-1 as a target in several different inflammatory diseases. This paper describes the de novo design, synthesis and in vitro activity of LFA-1 antagonists based on a bicyclic[5.5]hydantoin scaffold.

Full text of DOI:10.1016/j.bmcl.2004.12.007

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1161–1164
De novo design, synthesis, and in vitro activity of LFA-1
antagonists based on a bicyclic[5.5]hydantoin scaffold
Dominique Potin,a Michele Launay,a Eric Nicolai,a Maud Fabreguette,a Patrice Malabre,a
Franc¸ois Caussade,a Dominique Besse,a Stacey Skala,b Dawn K. Stetsko,b
Gordon Todderud,b Brett R. Beno,c Daniel L. Cheney,b Chiehying J. Chang,b
Steven Sheriff,b Diane L. Hollenbaugh,b Joel C. Barrish,b Edwin J. Iwanowicz,b
Suzanne J. Suchardb and T. G. Murali Dharb,*
aCerep, 128, rue Danton, 92500 Rueil-Malmaison, France
bBristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA
cBristol-Myers Squibb Pharmaceutical Research Institute, Wallingford, CT 06492, USA
Received 30 October 2004; revised 24 November 2004; accepted 2 December 2004
Available online 18 January 2005
Abstract—LFA-1 (leukocyte function-associated antigen-1), is a member of the b2-integrin family and is expressed on all leukocytes.
The LFA-1/ICAM interaction promotes tight adhesion between activated leukocytes and the endothelium, as well as between T cells
and antigen-presenting cells. Evidence from both animal models and clinical trials provides support for LFA-1 as a target in several
different inflammatory diseases. This paper describes the de novo design, synthesis and in vitro activity of LFA-1 antagonists based
on a bicyclic[5.5]hydantoin scaffold.
Ó 2005 Elsevier Ltd. All rights reserved.
LFA-1 (leukocyte function-associated antigen-1), also
known as CD11a/CD18 and aLb2 is a member of the
b2-integrin family and is expressed on all leukocytes.1
It is a heterodimeric transmembrane glycoprotein con-
sisting of a and b subunits. The ligands, ICAM (intercel-
lular adhesion molecule)-1, -2, and -3, are differentially
expressed and regulated on both leukocytes and the
endothelium. The LFA-1/ICAM interaction promotes
tight adhesion between activated leukocytes and the
endothelium, as well as between T cells and antigen-
presenting cells. Leukocyte activation results in both a
neutrophils.4,5 Similarly, in vitro studies demonstrate
that LFA-1 plays a critical role during T-cell activation
and proliferation, and is required for efficient cytolysis
by cytotoxic T cells.6 Clinical trials using a humanized
anti-LFA-1 antibody Efalizumab (RaptivaÒ) provide
excellent proof-of-principle for this target.7 In Phase
III clinical trials, 29% of subjects demonstrated 75% or
greater psoriasis area and severity index (PASI 75) score
improvement and 56% achieved PASI 50 when efalizu-
mab was given sc weekly at 1 mg/kg.8 US approval
was granted for RaptivaÒ in October 2003 for moder-
ate-to-severe psoriasis. In addition to psoriasis, the
humanized anti-LFA-1 antibody is in development for
prevention of solid organ transplant rejection.
clustering of LFA-1 on the cell surface and
a
conformational shift in the LFA-1 molecule that corre-
lates with an increase in affinity for binding to ICAM-
1.2,3
Because of the strong evidence from both animal models
and clinical trials for LFA-1/ICAM as a target in several
different inflammatory diseases, there has been an in-
tense effort to identify orally available, small molecule
inhibitors of this interaction.9 Figure 1 lists selected ana-
logs from several recently disclosed, structurally diverse
chemical series. Literature reports demonstrate that
some of these chemical series are in fact allosteric inhibi-
tors of the LFA-1/ICAM interaction. Specifically, both
In vivo studies using anti-LFA-1 antibodies or in LFA-
1-deficient mice demonstrate that LFA-1 plays a vital
role in extravasation of lymphocytes, eosinophils, and
Keyword: LFA-1.
*
Corresponding author. Tel.: +1 609 252 4158; fax: +1 609 252
6804; e-mail: murali.dhar@bms.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.12.007

1162
D. Potin et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1161–1164
Br
compound 1 would reveal a pseudo C-2 symmetric,
bicyclic[5.5]hydantoin system that would adapt a some-
what rigid Ôhalf-open bookÕ conformation (Fig. 2)
Therefore, the question was whether we could access
the p-bromophenyl pocket of BIRT-377 by optimizing
the linker in the bicyclic[5.5]hydantoin system. This re-
port describes the synthesis and in vitro SAR of novel
LFA-1 antagonists based on the aforementioned de
novo design.
Me
Me
Cl
NO2
O
S
Ac
N
N
O
N
Me
N
Cl
Me
O
A-286982
OH
BIRT-377
OH
O
OMe
The synthetic pathways utilized in the preparation of the
bicyclic[5.5]hydantoins are outlined in Schemes 1–3.
Commercially available homochiral hydroxyproline
served as the starting material for the synthesis of the
various compounds outlined in Table 1. The amino
substituted bicyclic hydantoins (Table 1, 12–17) were
synthesized by reductive amination of the ketone fol-
lowed by separation of diastereomers by silica gel col-
umn chromatography (Scheme 3).
N
O
O
Cl
Cl
Me
O
H
Br
Me
Me
Me
N
N
O
(1)
LFA-703
Figure 1. Representative LFA-1/ICAM inhibitors known from the
literature.
Table 1 displays the in vitro inhibitor potencies against
the LFA-1/ICAM interaction for a series of bicy-
clic[5.5]hydantoins.15 As the data indicates, the length
and nature of the linker, stereochemistry at the 5- and
7a-positions of the hydantoin scaffold, and the position
and substitution on the phenyl ring (R in Table 1) all
contribute significantly to the overall activity of this ser-
ies of compounds. For example, of the oxygen, sulfur,
and nitrogen atoms, the oxygen-containing linker ap-
pears to be optimal (compare compound 5 with 11
and 16). Altering the linker length leads to a significant
loss of activity. This can be seen in the comparison of 5
with 3 and 14, respectively. Table 1 also indicates that
the 4-Br substitution on the phenyl ring is optimal (com-
pare compound 5 with 7). Substitution of the 4-bromo
residue with a cyano moiety leads to a threefold drop
in potency (compare 5 and 8). Significant reduction in
potency is observed when the 4-bromo substituent on
the phenyl ring is replaced by bulky lipophillic groups
such as in compound 2. We next examined the stereo-
chemical preferences for the residues at positions 7a
and 5 of the bicyclic[5.5]hydantoin scaffold. It is clear
from Table 1 that compounds with S absolute configu-
ration at the 7a- and 5-positions are significantly more
potent than the corresponding 7aS/5R or 7aR/5R ana-
logs (compare compound 5 with 4, 6, and 9 and com-
pound 12 with 13). From the series of compounds
NMR and X-ray crystallography studies showbinding
to the IDAS site (I-domain allosteric site) of the I-do-
main on the CD11a unit, locking the conformation of
LFA-1 in a lowaffinity state thereby preventing its bind-
ing to ICAM-1.10–12 Elegant studies by Takagi and
Springer revealed the importance of the conformational
switch of the I-domain in binding to LFA-1.13
Of the compounds outlined in Figure 1, the structure of
the bicyclic compound (1)14 was intriguing with respect
to its symmetry and relationship to BIRT377. Moving
the p-bromophenyl residue away from the quaternary
center and placing it on the cyclopentyl ring system of
Br
Cl
Cl
O
Cl
Cl
O
H
N
Br
N
N
Linker
N
O
O
Figure 2. De novo design of LFA-1 antagonists based on the bicyclic-
[5.5]hydantoin scaffold.
R1
BOC
BOC
(CH2)n
O
a or b
c
N
N
OH
MeO2C
MeO2C
n = 0,1
R1
R1
O
Cl
Cl
NCO
d
(CH2)n
O
N
(CH2)n
O
N
+
HN
MeO2C
Cl
Cl
O
Scheme 1. Reagents and conditions: (a) dry KOH, DMSO, subst. BnBr, rt, 3 h, 60–70%; (b) DIAD, Ph3P, THF, subst. phenol, rt, 18 h, 60–70%; (c)
TFA, CH2Cl2, rt, 2 h, 70–90%; (d) K2CO3, DMF, rt, 18 h, 30–40%, or K2CO3,CH2Cl2, triethylamine, rt, 24 h, 70–90%.

D. Potin et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1161–1164
Br
1163
Br
Br
Cl
Cl
O
O
Cl
Cl
H
N
BOC
BOC
N
a , b
c
180o
N
N
H
OH
N
S
O
O
N
MeO2C
HO2C
O
Br
O
5a
Pseudo C-2 axis
of symmetry
Br
5b
concave nature of
scaffold
convex nature of
scaffold
O
Cl
Cl
Cl
NCO
d
N
N
+
S
HN
Figure 3. Two different orientations of the bicyclic[5.5]hydantoin, 5.
S
Cl
MeO2C
O
The pseudo C-2 symmetric bicyclic[5.5]hydantoin sys-
tem in compound 5 that would adapt a Ôhalf-open
bookÕ conformation, can have two different orientations
as shown in Figure 3.
Scheme 2. Reagents and conditions: (a) MsCl, CH2Cl2, Et3N, rt, 12 h,
90%; (b) MeONa, MeOH, (4-bromobenzyl)mercaptan, reflux, 18 h,
50–70%; (c) SOCl2, MeOH, rt, 24 h, 70%; (d) K2CO3, DMF, rt, 18 h,
20–30%.
In order to determine the preferred bound conforma-
tion, we examined a binding model of BIRT-377 bound
to LFA-1.12 In this model, the inhibitor adopts a conform-
ation orienting the 3,5-dichlorophenyl and the p-
bromophenyl aromatic rings in a favorable edge-to-face
p–p interaction. Molecular modeling suggests that of the
two symmetry-related orientations of the bicyclic hydan-
toin moiety (Fig. 4, 5a and 5b), only 5b is consistent with
the BIRT-377 model, and thus may represent the bound
conformation of compound 5 to LFA-1 (Fig. 4).
R1
Cl
Cl
O
O
Cl
Cl
(CH2)n
N
N
a, b, c
N
N
N
OH
R2
O
n = 1,2
O
Scheme 3. Reagents and conditions: (a) PDC, CH2Cl2, rt, 36 h, 60–
70%; (b) Na(OAc)3BH, mol. sieves, CH2Cl2, benzylamine, rt, 18 h, 40–
50%; (c) separation of diasteromers by SiO2 column chromatography.
As further evidence of the importance of the Ôhalf-open
bookÕ conformation of the bicyclic[5.5]hydantoin, the
urazole analog (18), in which the bicyclic ring system
outlined in Table 1, compound 5 has been identified as
the most potent LFA-1 antagonist.
Table 1. In vitro activity of bicyclic[5.5]hydantoins
R1
O
Cl
Cl
(CH2)n
A
N
N
5
7a
O
Compd
A
O
n
R1
Stereo
HeLa/HSB
(IC50 or % inhib. @ 1 lM)
26 nMa
28% (2)
BIRT-377
2
1
7aS,5S
N
3
O
0
1
1
1
1
1
1
1
1
1
1
2
2
1
1
4-Br–Phe
4-Br–Phe
4-Br–Phe
4-Br–Phe
3-Br–Phe
4-CN–Phe
4-Cl–Phe
7aS,5S
7aS,5R
7aS,5S
7aR,5R
7aS,5S
7aS,5S
7aR,5R
7aS,5S
7aS,5S
7aS,5S
7aS,5R
7aS,5S
7aS,5R
7aS,5S
7aS,5S
480 nM (3)
935 nM (1)
85 nM (3)
275 nM (3)
755 nM (3)
270 nM (3)
620 nM (1)
14% (2)
4
O
5
O
6
O
7
O
8
O
9
O
10
11
12
13
14
15
16
17
O
4-(2-CNPh)
4-Br–Phe
4-CN–Phe
4-CN–Phe
4-Br–Phe
4-Br–Phe
4-Br–Phe
4-CN–Phe
S
290 nM (2)
175 nM (2)
9% (2)
NH
NH
NH
NH
NMe
NEt
10% (2)
3% (2)
175 nM (2)
730 nM (1)
Values in parenthesis indicate the number of determinations.
a Kd value in an LFA-1/ICAM-1 binding assay as reported in the literature.12

1164
D. Potin et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1161–1164
6. Shier, P.; Ngo, K.; Fung-Leung, W.-P. J. Immunol. 1999,
163, 4826.
7. Dedrick, R. L.; Walicke, P.; Garovoy, M. Transplant
Immunol. 2002, 9, 181.
8. Cather, J. C.; Cather, J. C.; Menter, A. Expert Opin. Biol.
Ther. 2003, 3, 361.
9. Liu, G. Drugs Future 2001, 26, 767.
10. Kallen, J.; Welzenbach, K.; Ramage, P.; Geyl, D.;
Kriwacki, R.; Legge, G.; Cottens, S.; Weitz-Schmidt,
W.; Hommel, U. J. Mol. Biol. 1999, 292, 1.
11. Liu, G.; Huth, J. R.; Olejniczak, E. T.; Mendoza, R.;
DeVries, P.; Leitza, S.; Reilly, E. B.; Okasinski, G. F.;
Fesik, S. W.; von Geldern, T. W. J. Med. Chem. 2001, 44,
1202.
12. Last-Barney, K.; Davidson, W.; Cardozo, M.; Frye, L. L.;
Grygon, C. A.; Hopkins, J. L.; Jeanfavre, D. D.; Pav, S.;
Qian, C.; Stevenson, J. M.; Tong, L.; Zindell, R.; Kelly, T.
A. J. Am. Chem. Soc. 2001, 123, 5643.
Figure 4. Overlay of two symmetry-related orientations with the
putative bioactive conformation of BIRT-377 proposed by Last-
Barney et al.12
13. Takagi, J.; Springer, T. A. Immunol. Rev. 2002, 186, 141.
14. Sircar I.; Furth, P.; Teegarden, B. R.; Morningstar, M.;
Smith, N.; Griffith, R. WO 0130781, 2001.
Br
Br
15. Adhesion of the HSB-2 human lymphoblast cell line to
H1-Hela cells was used to characterize LFA-1 antagonists.
H1-HeLa cells have high constitutive expression of
ICAM-1 and are competent to support LFA-1-dependent
adhesion (Suchard et al., unpublished observation). Prior
to adhesion assays, HSB-2 cells were labeled with calcein
AM and treated with PMA for 30 min. The labeled, PMA-
activated cells were added to a monolayer of H1-HeLa
cells in the presence or absence of compound or anti-
human LFA-1 antibody. Adherent HSB-2 cells were
quantified as previously described (Akeson, A. L.;
Woods, C. W. J. Immunol. Methods 1993, 163, 181).
Adhesion in the presence of vehicle alone was considered
equivalent to 0% inhibition, and adhesion in the presence
of an excess of anti-human LFA-1 antibody was consid-
ered equivalent to 100% inhibition. IC50 values were
determined using Excel fit (Microsoft Corp., Bellevue,
WA).
O
Cl
Cl
Cl
Cl
O
N
H
N
N
N
O
N
O
O
O
18, IC50 = 190 nM
5, IC50 = 85 nM
Figure 5. Relative potencies of hydantoin and urazole analogs.
is more planar, is about twofold less potent than com-
pound 5 (Fig. 5).16
In conclusion, we have identified a novel series of LFA-1
antagonists based on de novo design employing the
bicyclic[5.5]hydantoin scaffold. Optimization of the
length and nature of the linker, stereochemistry at
the 5- and 7a-positions of the hydantoin scaffold, and
the position and substitution on the phenyl ring led
to the identification of compound 5 as a potent inhibitor
of the LFA-1/ICAM interaction.
16. Compound 18 was synthesized using the synthetic proto-
col outlined below. (a) NH2–NHCO2Me, toluene, 25 °C,
Cl
Cl
O
Cl
Cl
a
NH
NH
N
C O
N
References and notes
O
A
1. Harris, E. S.; McIntyre, T. M.; Prescott, S. M.; Zimmer-
man, G. A. J. Biol. Chem. 2000, 275, 23409.
c
18
2. Hogg, N.; Henderson, R.; Leitinger, B.; McDowall, A.;
Porter, J.; Stanley, P. Immunol. Rev. 2002, 186, 164.
3. Shimaoka, M.; Xiao, T.; Liu, J.-H.; Yang, Y.; Dong, Y.;
Jun, C.-D.; McCormack, A.; Zhang, R.; Joachimiak, A.;
Takagi, J.; Wanh, J.-H.; Springer, T. A. Cell 2003, 112, 99.
4. Andrew, D. P.; Spellberg, J. P.; Takimoto, H.; Schmits,
R.; Mak, T. W.; Zukowski, M. M. Eur. J. Immunol. 1998,
28, 1959.
b
Br
Cl
Br
Br
O
Br
B
28 h (82%), 4 M KOH, reflux, 1 h (100%); (b) epichloro-
hydrin, HgCl2 (0.05 equiv), 155 °C, 11 h (56%); (c) A,
NaOMe, MeOH then B, DMF reflux, 2 h (30%).
5. Berlin-Rufenach, C.; Otto, F.; Mathies, M.; Westermann,
J.; Owen, M. J.; Hamann, A.; Hogg, N. J. Exp. Med. 1999,
189, 1467.