Small molecule antagonists of the CCR2b receptor. Part 2: Discovery process and initial structure-activity relationships of diamine derivatives

Article abstract of DOI:10.1016/j.bmcl.2004.08.009

A lead evolution process is described that utilized the SAR of a homopiperazine series for the discovery of alternative cyclic diamine derivatives as nanomolar inhibitors of the interaction between MCP-1 and the CCR2b receptor. Structure-activity relationships (SAR) of a weakly active class of CCR2b inhibitors were utilized to initiate a lead evolution program employing the Drug Discovery Engine. Several alternative structural series have been discovered that display nanomolar activity in the CCR2b binding and CCR2b-mediated chemotaxis assays.

Full text of DOI:10.1016/j.bmcl.2004.08.009

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5413–5416
Small molecule antagonists of the CCR2b receptor. Part 2:
Discovery process and initial structure–activity relationships
of diamine derivatives
b
a,
Wilna J. Moree,a, Ken-ichiro Kataoka, Michele M. Ramirez-Weinhouse,
*
Tatsuki Shiota,b Minoru Imai,b Masaki Sudo,b,à Takaharu Tsutsumi,b Noriaki Endo,b
Yumiko Muroga,b Takahiko Hada,b,§ Hiroko Tanaka,b Takuya Morita,b
Jonathan Greene,a Doug Barnum,a John Saunders,a,– Yoshinori Kato,b Peter L. Myersa,k
and Christine M. Tarbya,
**
aDeltagen Research Laboratories, 4570 Executive Drive, Suite 400, San Diego, CA 92121, USA
bTeijin Institute for Bio-medical Research, 4-3-2 Asahigaoka, Hino, Tokyo 191-8512, Japan
Received 6 October 2003; accepted 5 August 2004
Available online 16 September 2004
Abstract—Structure–activity relationships (SAR) of a weakly active class of CCR2b inhibitors were utilized to initiate a lead evo-
lution program employing the Drug Discovery EngineTM. Several alternative structural series have been discovered that display
nanomolar activity in the CCR2b binding and CCR2b-mediated chemotaxis assays.
Ó 2004 Elsevier Ltd. All rights reserved.
Chemokines (chemotactic cytokines) are a family of
small molecular weight proteins involved in a variety
of inflammatory responses via the chemoattraction and
activation of leukocytes.1 Monocyte chemoattractant
protein-1 (MCP-1) is a 8–10kDa protein belonging to
the b- or CC chemokine subfamily and is postulated
to be primarily responsible for the selective recruitment
of leukocytes from the circulation to the site of inflam-
mation2 by binding to its 7-transmembrane G protein-
coupled receptor (CCR2b) on the surface of monocytes
and macrophages. Both CCR2b3 and MCP-14 knockout
mice as well as animal model studies with anti-MCP-1
have indicated that therapeutic intervention at the
CCR2b receptor may have beneficial effects in a variety
of inflammatory diseases including rheumatoid arthri-
tis,5 atherosclerosis,6 glomerulonephritis,7 and multiple
sclerosis.8 For this reason, we and others9 have initiated
programs to develop small molecule CCR2b antago-
nists. Screening of the Teijin sample collection, followed
by hit to lead efforts, led to the identification of lead ser-
ies 1 and 2 (Fig. 1).10
*
Corresponding author at present address: Neurocrine Biosciences,
Inc., 12790 El Camino Real, San Diego, CA 92130, USA. Tel.: +1
858 617 7506; fax: +1 858 617 7998; e-mail: wmoree@neurocrine.
com
Present address: Pfizer Global Research and Development, La Jolla
Laboratories, 3550 General Atomics Court, San Diego, CA 92121,
USA.
Despite optimization efforts, using both traditional
medicinal chemistry and library-based approaches to
identify potent CCR2b inhibitors in this series, binding
à Present address: Pfizer Global Research and Development, Nagoya
Laboratories, 5-2 Taketoyo, Aichi 470-2393, Japan.
§ Present address: Bizen Chemical Co., Ltd, Research and Develop-
ment Department, 363-Tokutomi, Kumayama-Cho, Akaiwa-Gun,
Okayama 709-0716, Japan.
R1
R2
– Present address: see *.
N
N
R1
R N
R
N
N
R
k Present e-mail address: myerspl@mindspring.com
** Present address: Bristol-Myers Squibb Company, PO Box 4000,
Princeton, NJ 08543, USA.
N
R2
R1
1
2
Figure 1. Homopiperazine lead series and alternative structural series.
Formerly CombiChem, Inc.
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.009

5414
W. J. Moree et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5413–5416
R1
O
S
R1
a,b
c
c
N
HN
N
Cl
O
R
HN
NHBoc
H
N
NH2
O
OH
S
3
7
11
R1
N
H
N
N
N
N
O
R1
HN
N
N
a,d
a
HO
1a
2a
O
Figure 2. Most potent compounds in homopiperazine series 1 and 2.
NH2
NH2
HN
R
4
5
8
9
12
R1
c
HN
R1
N
N
affinity remained in the micromolar range. These results
required us to evolve away from this homopiperazine
core and to develop alternative structural series. Our
Drug Discovery EngineTM was utilized to assist in this
lead evolution process,11 leading to the discovery of sev-
eral alternative diamine derived series with nanomolar
activity in the CCR2b binding and CCR2b chemotaxis
assays.
NH2
NH2
HN
R
13
R1
HN
R1
N
N
a,e,f
c
NH2
NH2
HN
R
O
6
10
14
Scheme 1. Reagents and conditions: (a) R1Cl or R1Br, K2CO3,
MeCN, 16h, D; (b) TFA; (c) R2CO2H, EDC, HOBt, DIPEA, or
R2SO2Cl, DIPEA, or R2NCO, DIPEA, or R2NCS, DIPEA, or RCl, or
RBr, DIPEA; (d) BH3; (e) Br2, NaOH; (f) 20% HCl, D.
As described in our previous paper,10 optimization of
the homopiperazine series, using either traditional
medicinal chemistry techniques or through library-based
approaches, led to the identification of CCR2b antago-
nists with lM binding affinities. The most potent com-
pounds identified in series 1 and 2 were 1a with an
IC50 value of 0.71lM and 2a with an IC50 value of
7.4lM, respectively (Fig. 2). Although approximately
800 compounds were prepared in the homopiperazine
series, further improvement of activity could not be
established and alternative structural series were investi-
gated using a process called the Drug Discovery
EngineTM.
mon substituent. The second substituent was varied
using hypothesis matching entities as well as entities,
which would probe a developing SAR. From the virtual
library containing 83,000 members, a set of 820 com-
pounds was synthesized around 11 different cores. Only
the diamine derivatives 11–14 will be described herein,
since the majority of active compounds were identified
in those series (vide infra).
For this purpose, screening data generated from the
homopiperazine series together with screening data from
ꢀ200 known GPCR ligands were used to generate a
computational model, composed of an ensemble of
pharmacophore ÔhypothesesÕ.11b Each of these pharma-
cophores contained two–four important features and
was selected since they distinguished active from inactive
compounds. The hypotheses were employed to filter a
virtual library, which was computationally constructed
by subjecting 20 different cores to a variety of in-silico
reactions that would be applicable to high-throughput
parallel synthesis. The cores and reactions were chosen
such that the resulting compounds would contain a
basic nitrogen, which was found to be a prerequisite
for activity in the homopiperazine series as well as for
most series described in the literature.9a–c The side
chains were derived from commercially available
reagents. In addition, side chains determined in the
optimization of the homopiperazine series, such as the
p-chlorobenzyl, benzhydryl, and 4-methyl-thiophene-2-
carboxylic acid hydrazide, were included to allow for a
direct comparison between templates. Examples of some
of the reactions utilized were alkylation, acylation, sulf-
onylation, epoxide opening, Grignard addition, and
(thio)urea formation. Generic examples of virtual
library members 11–14 are shown in Scheme 1. Temp-
lates and chemistries were prioritized based on a relative
score against the pharmacophore hypotheses11b as well
as chemical feasibility. To facilitate high-throughput
chemistry, high scoring compounds around each core
were grouped and divided in subgroups based on a com-
For the high-throughput parallel library synthesis of
diamine derivatives 11–14, intermediates 7–10 were
prepared by alkylation of starting materials 3–6. For
4-aminomethylpiperidine 5, selective alkylation on the
secondary nitrogen was achieved by the use of a large
excess of diamine to generate intermediate 9. Alkylated
prolinamide required an additional reduction step with
borane to give intermediate 8, whereas alkylated
3-(Boc-amino)pyrrolidine was treated with TFA to
remove the Boc group to give 7. Finally, alkylated nipe-
cotamide was subjected to a Hofmann rearrangement,
followed by hydrolysis of the formed isocyanate to give
intermediate 10.
All intermediates 7–10 were either alkylated using alkyl
halides, acylated with acids using EDC/HOBt coupling
conditions, reacted with sulfonylchlorides, or treated
with (thio)isocyanates to give the library compounds
11–14, where R is variably alkyl, –COR2, –SO2R2,
–CONHR2, or –CSNHR2. A variety of purification
techniques was employed, which included acid–base
extractions, ion exchange chromatography, resin scav-
enging techniques and silica gel column chromatogra-
phy. Compounds that were less than 85% pure were
subsequently subjected to reverse phase HPLC/MS puri-
fication using mass triggered fraction collection.12
The biological activity of the compounds was initially
assessed in a radioligand binding assay using human
125I-MCP-1 and THP-1 cells.13 The results for selected
compounds are summarized in Tables 1–3.

W. J. Moree et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5413–5416
Table 1. Binding affinities of dialkylated diamines to CCR2b Table 3. Chemotaxis to MCP-1 compared to binding affinity
5415
R 1
Compound
Binding to CCR2b
IC50 (lM)
Chemotaxis to MCP-1
EC50 (lM)a
R1
N
N
NH
R
11i(S)
11i(R)
12i(S)
12i(R)
13i
1.20
0.18
3.10
0.57
0.70
0.66
0.360 0.09
0.024 (n = 2)
0.630 0.17
0.810 0.268
0.043 (n = 2)
0.610 0.153
11
12
N
R
H
O
H
N
N
H
S
O
Cl
14i
a Ave (n = 3) SEM unless indicated otherwise.
I
II
III
Compound
R1
R
IC50 (lM)a
Substitution of one of the alkyl side chains by a sulfon-
amide or a (thio)urea resulted in very weak to inactive
compounds for all the templates (data not shown).
Although replacement of one of the alkyl side chains
by an amide eliminated activity in the homopiperazine
series (data not shown), this was not true for some of
the other diamine templates. Both 3-aminopyrrolidine
and 2-aminomethylpyrrolidine derivatives 11e and 12e
maintained micromolar activity containing a 5-oxo-5-
phenyl pentanoyl amide sidechain (Table 2). Modifica-
tion of this side chain to a N-(carbamoylethyl) benz-
amide did not affect activity for most templates (11f,
12f, 13f, 14f). It was rationalized that a contracted side
chain such as N-(carbamoylmethyl) benzamide could
11a
11b
11c
11d
12a
II
III
I
I
33
41
I
II
III
II
17
7.8 (n = 3)b
26
I
I
a Assayed with 125I-MCP-1 (n = 1, unless indicated otherwise).
b SEM = 2.2lM.
The new diamine cores substituted with the same comb-
ination of alkyl side chains as used on the homopipera-
zine core in various orientations led in some cases to
equipotent compounds (i.e., 11c–d, 12a) indicating that
the homopiperazine core could be replaced (Table 1).
Table 2. IC50 values (lM) of diamine derivatives for CCR2b receptor
N
N
N
H
R
NH
R
Cl
11
13
Cl
N
N
N
H
R
Cl
N
H
R
Cl
12
14
R
IC50 (lM)a series 11
11c (n = 5)
IC50 (lM)a series 12
19 (n = 2)
IC50 (lM)a series 13
IC50 (lM)a series 14
O
O
e
50
87
O
O
f
22
40 (n = 2)
64
74
N
H
O
O
H
N
g
h
i
42
66 (n = 2)
14 (n = 2)
11 (n = 2)
NAb
O
O
H
N
6.9 (n = 2)
5.5 (n = 2)
0.7d (n = 12)
17
O
H
N
0.37,
2.3,
(S) 3.1,
(R) 0.57
0.66
CF
3
(S) 1.2,
(R) 0.18
(n = 2)
O
a Assayed with 125I-MCP-1 (n = 1, unless indicated otherwise).
b Not active.
c SEM = 2.7lM.
d SEM = 0.08lM.

5416
W. J. Moree et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5413–5416
improve activity for the more extended template 13
compared to 11e. Indeed binding affinity of compound
13g was significantly better than of 13e or 13f. Substitu-
tion of the aromatic ring in N-(carbamoylmethyl) benz-
amide side chain with a m-methyl group (13h) improved
activity further by 2-fold and with a m-trifluoromethyl
(13i) by 12-fold. This trend was observed for all the
diamine templates 11–14. The substitution effect was
most dramatic for the 3-aminopyrrolidine derivatives
(11) and 3-aminopiperidine derivatives (14). In both ser-
ies binding affinity improved approximately 100-fold by
substitution of the N-(carbamoylmethyl) benzamide in
11g, 14g with a m-CF3 group (11i, 14i). In series 12 this
effect was less pronounced and only a 29-fold difference
in binding affinity was observed between 12g containing
N-(carbamoylmethyl) benzamide and 12i with an addi-
tional m-CF3 substituent. The observation that N-(car-
pounds described in this paper and Dr. Takeshi Hara
and Dr. Seizi Kurozumi for general support.
References and notes
1. (a) Rollins, B. J. Blood 1997, 90, 909; (b) Howard, O. M.
Z.; Ben-Baruch, A.; Oppenheim, J. J. Trends Biotechnol.
1996, 14, 46; (c) Saunders, J.; Tarby, C. M. Drug Discovery
Today 1999, 4, 80; (d) Murphy, P. M.; Baggiolini, M.;
Charo, I. F.; Hebert, C. A.; Horuk, R.; Matsushima, K.;
Miller, L. H.; Oppenheim, J. J.; Power, C. A. Pharmaco-
logic. Rev. 2000, 52, 145; (e) Horuk, R. Cytokine Growth
Factor Rev. 2001, 12, 313.
2. (a) Rollins, B. J.; Morrison, E. D.; Stiles, C. D. Proc. Natl.
Acad. Sci. U.S.A. 1988, 85, 3738; (b) Charo, I. F.; Myers,
S. J.; Herman, A.; Franci, C.; Connolly, A. J.; Coughlin,
S. R. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 2752.
3. Boring, L.; Gosling, J.; Chensue, S. W.; Kunkel, S. L.;
Farese, R. V., Jr.; Broxmeyer, H. E.; Charo, I. F. J. Clin.
Invest. 1997, 100, 2552.
bamoylmethyl)-3-trifluoromethyl benzamide has
a
similar effect on binding affinity of all series is somewhat
surprising since series 11, 12, and 14 exhibit a con-
strained ethylene diamine core, but this structural motif
is conspicuously absent in series 13.
4. Lu, B.; Rutledge, B. J.; Gu, L.; Fiorillo, J.; Lukacs, N. W.;
Kunkel, S. L.; North, R.; Gerard, C.; Rollins, B. J. J. Exp.
Med. 1998, 187, 601.
5. (a) Ogata, H.; Takeya, M.; Yoshimura, T.; Takagi, K.;
Takahashi, K. J. Pathol. 1997, 182, 106; (b) Gong, J.-H.;
Ratkay, L. G.; Waterfield, J. D.; Clark-Lewis, I. J. Exp.
Med. 1997, 186, 131.
6. (a) Boring, L.; Gosling, J.; Cleary, M.; Charo, I. F. Nature
1998, 394, 894; (b) Dawson, T. C.; Kuziel, W. A.; Osahar,
T. A.; Maeda, N. Atherosclerosis 1999, 143, 205.
7. Lloyd, C. M.; Minto, A. W.; Dorf, M. E.; Proudfoot, A.;
Wells, T. N. C.; Salant, D. J.; Gutierrez-Ramos, J.-C. J.
Exp. Med. 1997, 185, 1371.
8. Kennedy, K. J.; Strieter, R. M.; Kunkel, S. L.; Lukacs,
N. W.; Karpus, W. J. J. Neuroimmunol. 1998, 92, 98.
9. (a) Mirzadegan, T.; Diehl, F.; Ebi, B.; Bhakta, S.; Polsky,
I.; McCarley, D.; Mulkins, M.; Weatherhead, G. S.;
Lapierre, J.-M.; Dankwardt, J.; Morgans, D., Jr.; Wil-
helm, R.; Jarnagin, K. J. Biol. Chem. 2000, 275, 25562; (b)
Forbes, I. T.; Cooper, D. G.; Dodds, E. K.; Hickey, D. M.
B.; Ife, R. J.; Meeson, M.; Stockley, M.; Berkhout, T. A.;
Gohil, J.; Groot, P. H. E.; Moores, K. Bioorg. Med. Chem.
Lett. 2000, 10, 1803; (c) Witherington, J.; Bordas, V.;
Cooper, D. G.; Forbes, I. T.; Gribble, A. D.; Ife, R. J.;
Berkhout, T.; Gohil, J.; Groot, P. H. E. Bioorg. Med.
Chem. Lett. 2001, 11, 2177; For short reviews, see: (d)
Trivedi, B. K.; Low, J. E.; Carson, K.; LaRosa, G. J. Ann.
Reports Med. Chem. 2000, 35, 191; (e) Howard, O. M. Z.;
Yoshimura, T. Expert Opin. Ther. Patents 2001, 11(7),
1147.
10. Imai, M.; Shiota, T.; Kataoka, K.; Tarby, C. M.; Moree,
W. J.; Tsutsumi, T.; Sudo, M.; Ramirez-Weinhouse, M.
M.; Comer, D.; Sun, C.-M.; Yamagami, S.; Tanaka, H.;
Morita, T.; Hada, T.; Greene, J.; Barnum, D.; Saunders,
J.; Myers, P. L.; Kato, Y.; Endo, N. Bioorg. Med. Chem.
Lett. 2004, 14, preceding paper. doi:10.1016/j.bmcl.
2004.08.008.
Chiral resolution indicated a preference for the R enantio-
mer over the S enantiomer in both series 11 and 12.
Compounds 11i(R) and 12i(R) were 5- to 7-fold more
active than 11i(S), and 12i(S). Compounds containing
the N-(carbamoylmethyl)-3-trifluoromethyl benzamide
side chain were not only potent CCR2b binders but also
showed potent inhibition of cell chemotaxis caused by
MCP-1 using THP-1 cells as the chemotactic cell line.14
The results are summarized in Table 3. In addition,
potent compounds in series 11 inhibited chemokine
induced Ca2+ immobilization in a calcium flux [Ca2+
]
i
assay (compounds and data not shown). The effect of
all potent compounds in series 11–14 on CCR1 was
measured using [125I]-MIP-1a and THP cells indicating
a moderate selectivity, for example, 49% inhibition at
10lM was observed for 11i(R) (IC50 > 5lM), corre-
sponding to a ꢀ30-fold selectivity on the related chemo-
kine receptor.
In conclusion, we described a lead evolution process that
led to the generation of alternative structural series to
the homopiperazines as CCR2b antagonists. A synergis-
tic approach between computational library design and
traditional medicinal chemistry in high-throughput for-
mat led to the identification of four novel lead series that
displayed submicromolar activity in CCR2b binding
and chemotaxis assays. Interestingly, the N-(carbamoyl-
methyl)-3-trifluoromethyl benzamide side chain led to a
dramatic increase in binding affinity in all series shown
in Table 3. Optimization of the lead series and detailed
SAR studies will be the subject of our next
communication.
11. (a) Myers, P. L.; Greene, J. W.; Saunders, J.; Teig, S. L.
Today’s Chemist Work 1997, 6(7), 46; (b) Bradley, E. K.;
Beroza, P.; Penzotti, J. E.; Grootenhuis, P. D. J.;
Spellmeyer, D. C.; Miller, J. L. J. Med. Chem. 2000, 43,
2770.
Acknowledgements
12. Zeng, L.; Kassel, D. B. Anal. Chem. 1998, 70, 4380.
13. See Ref. 10 in preceding paper.
14. Falk, W.; Goodwin, R. H., Jr.; Leonard, E. J. J. Immunol.
Methods 1980, 3, see Ref. 11 in preceding paper.
The authors wish to thank Dr. Dan Kassel, Dr. Lu
Zeng, and Ms. Xiaoli Wang for their technical expertise
in setting up systems for HPLC purification of com-