Development of CXCR3 antagonists. Part 2: Identification of 2-amino(4-piperidinyl)azoles as potent CXCR3 antagonists

Article abstract of DOI:10.1016/j.bmcl.2007.10.029

Development of a lead series of piperidinylurea CXCR3 antagonists has led to the identification of molecules with alternative linkages which retain good potency. A novel 5-(piperidin-4-yl)amino-1,2,4-thiadiazole derivative was found to have satisfactory in vitro metabolic stability and to be orally bioavailable in mice, giving high plasma concentrations and a half life of 5.4 h.

Full text of DOI:10.1016/j.bmcl.2007.10.029

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6806–6810
Development of CXCR3 antagonists. Part 2: Identification of
2-amino(4-piperidinyl)azoles as potent CXCR3 antagonists
Robert J. Watson,* Daniel R. Allen, Helen L. Birch, Gayle A. Chapman,
Duncan R. Hannah, Roland L. Knight, Johannes W. G. Meissner,
David A. Owen and Elizabeth J. Thomas
UCB Inflammation Discovery, Granta Park, Great Abington, Cambridge CB21 6GS, United Kingdom
Received 7 September 2007; revised 8 October 2007; accepted 10 October 2007
Available online 17 October 2007
Abstract—Development of a lead series of piperidinylurea CXCR3 antagonists has led to the identification of molecules with alter-
native linkages which retain good potency. A novel 5-(piperidin-4-yl)amino-1,2,4-thiadiazole derivative was found to have satisfac-
tory in vitro metabolic stability and to be orally bioavailable in mice, giving high plasma concentrations and a half life of 5.4 h.
Ó 2007 Elsevier Ltd. All rights reserved.
CXCR3 is a 7-transmembrane G-protein coupled recep-
tor, which is expressed by activated T cells, NK cells,
mast cells and B cells. CXCR3 binds three members of
the CXC chemokine family: MIG (CXCL9), IP-10
(CXCL10) and I-TAC (CXCL11) resulting in receptor
activation and cellular chemotaxis. A number of block-
ing strategies have demonstrated reductions of disease
severity in animal models of arthritis,1,2 inflammatory
bowel disease,3 diabetes4 and transplant rejection.5
Studies in human patients have suggested a role for
CXCR3 in rheumatoid arthritis, multiple sclerosis, dia-
betes, transplant rejection, chronic obstructive pulmon-
ary disease6 and asthma.1 CXCR3 is thus an attractive
target for the development of anti-inflammatory thera-
pies, and a number of groups have published ap-
proaches to the development of small molecule
antagonists.7
Within this series, modification of physicochemical
properties by introduction of solubilising aryl substitu-
ents could be achieved, leading to examples with im-
proved characteristics.8 Nevertheless, we had found
that the series as a whole tended to have high logD,
low solubility and poor in vivo pharmacokinetics. There
were, therefore, significant challenges in developing
drug-like molecules, which could be used as tools with
which to explore the potential of CXCR3 antagonism
in models of human disease.
Our strategy in further development was to examine the
effect of modification of the urea linkage, so as to estab-
lish whether a robust alternative chemical series could be
generated.
As reported previously, we had found that replacing the
lipophilic myrtenyl right hand side lost potency and it
was therefore retained for this phase of the series
development.
In our previous publication,8 we described the identifica-
tion of the naphthyl urea 1a by high throughput screen-
ing of a compound set selected as being ‘GPCR-
friendly’. The hit molecule formed the basis of early
hit to lead activity leading to the identification of mole-
cules such as 1b with improved physicochemical proper-
ties and excellent potency (Fig. 1).
Our first investigation was around N-substituted urea
derivatives 4 and 5, to assess the importance of the urea
in binding of the molecules. These compounds were con-
structed from the myrtenylpiperidine 28 as shown in
Scheme 1. N-Alkyl amines 3 were generated by acylation
of 2 under basic conditions, followed by reduction with
lithium aluminium hydride. The N-alkyl ureas 4 were
then obtained in good yields by reaction with 3-trifluo-
romethyl phenyl isocyanate.
Keywords: CXCR3 antagonist; (Piperidinyl)azoles; Benzazole; Opti-
misation; Pharmacokinetics.
*
Corresponding author. Tel.: +44 1223896515; fax: +44
1223896400; e-mail: bob.watson@ucb-group.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.10.029

R. J. Watson et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6806–6810
6807
F
O
N
O
N
N
N
CF3
N
N
H
H
H
H
1b
1a
Figure 1. Hit molecules.
N
N
a, b
HN
H2N
R
3a R = Me
3b R = Et
3c R = Bn
2
c
d, e
F
F
F
O
N
O
N
N
N
N
N
H
H
R
5
4a R = Me
4b R = Et
4c R = Bn
Scheme 1. Reagents and conditions: (a) RCOCl, Et3N, DCM, 75–90%; (b) LAH, THF, reflux, 85%; (c) 3-trifluoromethylphenyl isocyanate, DCM,
57–85%; (d) triphosgene, Hunig’s base, DCM; (e) 3-ethyl N-methyl aniline, 43% over two steps.
Table 1. Urea modifications
To access the 1-methyl urea 5, the amine 2 was con-
verted into the required isocyanate by reaction with tri-
phosgene and coupled with 3-ethyl N-methyl aniline in
modest yield.
Compound
hCXCR3 Ki (lM)
1b
4a
4b
4c
5
0.016
0.18
3.6
As shown in Table 1, these modifications to the urea moi-
ety resulted in a loss of potency when compared to 1b.
Methylation of the nitrogen adjacent to the piperidine
ring was tolerated, but larger groups resulted in a dra-
matic loss of potency. Likewise, methylation adjacent
IA
3.5
Ki’s measured in ITAC stimulated GTPcS assay using CXCR3
transfected CHO membranes and are the average of two values.8
IA = <20% inhibition at 10 lM.
R
N
N
O
a, b
H2N
N
N
H
O
2
6
O
N
O
N
O
R
R
c
N
N
N
N
8
7
O
Scheme 2. Reagents and conditions: (a) methyl bromoacetate, DMF, K2CO3, 65%; (b) ArNCO, DCM, 70%; (c) NaBH4, TFA, 45%.

6808
R. J. Watson et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6806–6810
We wished to investigate whether polar substituents at-
tached to the urea linker could lead to additional binding
interactions, thus restoring potency, and we attempted to
prepare the ester 6 as shown in Scheme 2. However, the
main product after HPLC purification was in fact the
hydantoin 7a. Having previously established that urea
substitution adjacent to the aromatic ring was not toler-
ated, we were surprised to find that 7a is a potent recep-
tor antagonist with Ki of 26 nM. A set of close analogues
was prepared in which changes to the aromatic substitu-
tion were investigated. As shown in Table 2, a range of
aromatic substituents provided active compounds
although all were less potent than 7a. The ketone 7g dem-
onstrated reduced logD, but at the expense of potency.
Table 2. Hydantoins and imidazolinones
a
Compound
R
hCXCR3
Ki (lM)
logD
CLINT
(lL/min/mg)
7a
7b
7c
7d
7e
7f
3,5-(CF3)2
3-F, 5-CF3
3-Et
0.026
0.10
0.38
0.75
0.22
0.12
IA
>5.5
4.85
4.7
220
290
190
230
318
280
ND
7
3-S-Me
3-CF3
5.0
4.6
3-CF3, 4-Cl
3-COMe
3,5-(CF3)2
3-F, 5-CF3
5.3
3.4
7g
8a
8b
0.17
0.20
>5.5
4.6
88
Ki measurements are an average of two determinations. IA = <20%
inhibition at 10 lM. logD measured by octanol-water partition at pH
7.4.
a Murine microsomal clearance at 0.5 lM.
The imidazolinones 8a and 8b were also generated, by
reduction of the hydantoins with sodium borohydride
in the presence of trifluoroacetic acid.
The hydantoin compounds tested were highly lipophilic
and proved to be very highly metabolized in vitro by
murine microsomes. In addition, some hydantoins were
found to be hydrolytically labile, opening up under
aqueous conditions. While we were encouraged by the
profile of the imidazolinone 8a which gave good potency
and microsomal stability we found that the compound
had very poor aqueous solubility and that close ana-
logues such as 8b were more highly turned over in the
microsomal stability assay. We therefore felt that the
template would be difficult to adapt to provide com-
bined potency and drug-like properties, however the
work demonstrated the principle that cyclic linkers
could serve to replace the urea.
Figure 2. Overlay of aryl urea (green carbons) with benzoxazole (pink
carbons).
to the aryl group was highly detrimental to activity, sug-
gesting that the urea NH in this position plays an impor-
tant role in binding. A number of urea replacements such
as thioureas and carbamates were also investigated (not
shown) and found to result in reduced activity.
We therefore set out to incorporate heteroaromatic
replacements for the hydantoin moieties with the aim
of improving both the physicochemical properties and
stability of the products.
R'
N
a
N
N
RHN
X
N
R
9
2
R = H
3a R = Me
R'
b
X
S
N
Cl
11
R'
X
S
N
N
N
R
10
Scheme 3. Preparation of aminopiperidinyl benzazoles and aryl azoles. Reagents and conditions: (a) 2-chlorobenzazole, NMP, Et3N, 140 °C, 23–
45%; (b) Et3N, NMP, 100 °C, 40–55%.

R. J. Watson et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6806–6810
6809
Table 3. Benzazole compounds
a
Compound
X
R0
R
hCXCR3 Ki (lM)
logD
CYP 2D6 (lM)
CLINT (lL/min/mg)
9a
9b
9c
O
H
H
0.68
0.72
0.35
0.14
0.95
0.46
0.29
0.17
0.62
0.15
0.13
0.88
0.46
0.26
0.27
4.7
4.6
>5
>5
4.2
5
ND
ND
13
ND
ND
1
O
H
Me
H
S
H
9d
9e
S
H
Me
Me
Me
Me
Me
Me
H
2.2
ND
ND
3.5
7.3
4.1
1.2
0.7
IA
78
NH
S
H
ND
ND
41
9f
4-Cl
5-Cl
6-Cl
7-Cl
4-Br
6-Br
H
9g
9h
9i
S
5
S
5
48
S
>5
>5
>5
4.7
>5
>5
5
ND
140
150
ND
ND
34
9j
S
9k
10a
10b
10c
10d
S
H
CH
CH
CH
N
H
3,4-di-Cl
3,5-(CF3)2
3,5-(CF3)2
H
IA
IA
H
H
15
16
Ki measurements are a mean of two determinations. ND, not determined. IA = <20% at 50 lM.
a Murine microsomal clearance at 0.5 lM.
Similarly, a set of aryl-substituted aminoazoles 10 was
constructed by nucleophilic substitution of chloroazoles
1110 with 2 under basic conditions (Scheme 3).
Compound 10d
Mean concentration profiles in mice
F C
3
10000
1000
100
10
N
S
N
N
As shown in Table 3, moderate potency was achieved
with the benzoxazoles 9a and 9b. The benzothiazole 9c
appeared to give increased potency and low microsomal
clearance. Activity was further improved by methylation
of the linker to give 9d, although this change also led to
increased microsomal clearance and CYP inhibition.
The benzimidazole 9e was less potent than either benz-
oxazole or benzothiazole derivatives and was not inves-
tigated further.
N
H
F C
3
1100mmgg//kkgg ppoo
00..55mmgg//kkgg iiv
1
0
4
8
12
16
20
24
Time (h)
Figure 3. In vivo pharmacokinetics of compound 10d.
A series of chlorobenzothiazoles showed that the
6-substituted derivative 9h was equipotent to the unsub-
stituted analogue 9d, while the 7-chloro compound 9i
was less potent. The 4 and 6-bromobenzothiazoles gave
good activity, but were potent CYP2D6 inhibitors and
were also highly metabolized in the microsome incuba-
tion assay.
Comparison of an aryl urea with a benzoxazole using
the PyMol program9 showed a close overlay. Both mol-
ecules displayed the piperidine in a similar orientation,
the azole ring nitrogen occupying the same position as
one of the urea nitrogens and the aromatic rings in sim-
ilar orientation (Fig. 2).
The aryl azoles 10 gave moderate CXCR3 activity, the
most potent examples bearing the familiar 3,5-bis-tri-
fluoromethyl substitution pattern. Although 10d was
found to be highly lipophilic, it exhibited good in vitro
microsomal stability and showed only weak inhibition
of cytochrome P450 enzymes. This compound was taken
forward into a PK study to establish whether its in vitro
stability would result in satisfactory in vivo exposure.
We chose to construct a series of benzazole and aryl
azole derivatives, with a view to evaluating the effect
of substitution on the aromatic ring and on the nitrogen
linker.
A set of commercially available 2-chlorobenzazoles was
assembled and it was coupled in parallel by heating un-
der microwave conditions with amines 2 and 3a. The
crude products were purified directly by preparative
HPLC of the reaction mixtures with no work-up re-
quired, thus allowing rapid SAR generation.
As shown in Figure 3 and Table 4, compound 10d was
orally bioavailable with a long plasma half-life and
low clearance of 2 mL/min/kg, thus demonstrating that
the template could deliver non-urea CXCR3 antagonists
with improved PK. This compound had human plasma
Table 4. In vivo pharmacokinetics of compound 10d
Compound
cMax (ng/ml)
2012
AUC (ng h/ml)
10,800
T1/2 (h)
CLp (ml/min/kg)
2.8
logD
Sol (lg/ml)
10d
5.4
5
2.7

6810
R. J. Watson et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6806–6810
Dairaghi, D.; Lawrence, C.; Sullivan, T.; Tonn, G.;
Schall, T.; Collins, T.; Medina, J. Bioorg. Med. Chem.
Lett. 2007, 17, 3339; (b) Heise, C. E.; Pahuja, A.;
Hudson, S. C.; Mistry, M. S.; Putnam, A. L.; Gross, M.
M.; Gottlieb, P. A.; Wade, W. S.; Kiankarimi, M.;
Schwartz, D.; Crowe, P.; Zlotnik, A.; Alleva, D. G.
J. Pharmacol. Exp. Ther. 2005, 313, 1263; (c) Baba, M.;
Nishimura, O.; Kanzaki, N.; Okamoto, M.; Sawada, H.;
Iizawa, Y.; Shiraishi, M.; Aramaki, Y.; Okonogi, K.;
Ogawa, Y.; Meguro, K.; Fujino, M. PNAS 1999, 96,
5698; (d) Storelli, S.; Verdijk, P.; Verzijl, D.; Timmer-
man, H.; van de Stolpe, A. C.; Tensen, C. P.; Smit, M.
J.; De Esch, I. J. P.; Leurs, R. Bioorg. Med. Chem. Lett.
2005, 15, 2910; (e) Cole, A. G.; Stroke, I. L.; Brescia,
M.-R.; Simhadri, S.; Zhang, J. J.; Hussain, Z.; Snider,
M.; Haskell, C.; Ribeiro, S.; Appell, K. C.; Henderson,
I.; Webb, M. L. Bioorg. Med. Chem. Lett. 2006, 16, 200;
(f) Storelli, S.; Verzijl, D.; Al-Badie, J.; Elders, N.;
Bosch, L.; Timmerman, H.; Smit, M. J.; De Esch, I. J.
P.; Leurs, R. Arch. Pharm. Chem. Life Sci. 2007, 340,
281; (g) Medina, J. C.; Johnson, M. G.; Collins, T. L.
Ann. Rep. Med. Chem. 2005, 40, 215.
protein binding of 99.7% and was found to have activity
of 1 lM in a murine CXCR3 assay.8
In conclusion, we have developed a novel series of po-
tent CXCR3 antagonists based on a piperidinyl-azole
scaffold, suitable for further optimization. Further
development of this series and in vivo characterization
of resulting molecules are the subject of future
publications.
References and notes
1. Lazzeri, E.; Romagnani, P. Curr. Drug Targets Immune
Endocr. Metabol. Disord. 2005, 5, 109.
2. Salomon, I.; Netzer, N.; Wildbaum, G.; Schif-Zuck, S.;
Maor, G.; Karin, N. J. Immunol. 2002, 169, 2685.
3. Singh, U. P.; Singh, S.; Taub, D. D.; Lillard, J. W., Jr.
J. Immunol. 2003, 171, 1401.
4. Baker, M. S.; Chen, X.; Rotramel, A. R.; Nelson, J. J.; Lu,
B.; Gerard, C.; Kanwar, Y.; Kaufman, D. B. Surgery
2003, 134, 126.
5. Hancock, W. W.; Lu, B.; Gao, W.; Csizmadia, V.; Faia,
K.; King, J. A.; Smiley, S. T.; Ling, M.; Gerard, N. P.;
Gerard, C. J. Exp. Med. 2000, 192, 1515.
6. Donnelly, L. E.; Barnes, P. J. Trends Pharmacol. Sci. 2006,
27, 546.
7. (a) Johnson, M.; Li, A. R.; Liu, J.; Fu, Z.; Zhu, L.;
Miao, S.; Wang, X.; Xu, Q.; Huang, A.; Marcus, A.; Xu,
F.; Ebsworth, K.; Sablan, E.; Danao, J.; Kumer, J.;
8. Allen, D. R.; Bolt, A.; Chapman, G. A.; Knight, R. L.;
Meissner, J. G.; Owen, D. A.; Watson, R. J. Bioorg. Med.
Chem. Lett. 2007, 17, 697.
9. DeLano, W. L. The PyMOL Molecular Graphics System,
DeLano Scientific, San Carlos, CA, USA, 2002.
10. Chloroazoles 11 were commercially available, or were
prepared using the procedure of Merijanian et al.11
11. Merijanian, A.; Sharma, G.; Moushati, J.; Gunderson, K.
J. Org. Chem. 1986, 51, 543.