Design and synthesis of a library of chemokine antagonists

Article abstract of DOI:10.1016/j.bmcl.2011.09.013

A library of chemokine antagonists has been synthesized using a combination of solid and solution-phase chemistry. Structures of known chemokine antagonists were used to produce a pharmacophore which served to guide monomer selection. Several combinations of monomers have resulted in providing novel chemokine antagonists which in some cases display dual chemokine receptor antagonism.

Full text of DOI:10.1016/j.bmcl.2011.09.013

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6249–6252
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and synthesis of a library of chemokine antagonists
Gurdip Bhalay a, , Birgit Albrecht a, Mohammed Akhlaq b, Urs Baettig a, David Beer b, Zarin Brown b,
⇑
Steven Charlton b, Andrew Dunstan a, Michelle Bradley b, Peter Gedeck a, Angela Glen a, Trevor Howe a,
Thomas Keller a, Juliet Leighton-Davies b, Alice Li b, Clive McCarthy a, Cecile Mocquet a, Charles Owen b,
Paul Nicklin b, Elizabeth Rosethorne b
a Global Discovery Chemistry, Novartis Institutes of Biomedical Research, Horsham Research Centre, Wimblehurst Road, Horsham, West Sussex RH12 5AB, UK
b Respiratory Diseases Area, Novartis Institutes of Biomedical Research, Horsham Research Centre, Wimblehurst Road, Horsham, West Sussex RH12 5AB, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 July 2011
Revised 2 September 2011
Accepted 3 September 2011
Available online 10 September 2011
A library of chemokine antagonists has been synthesized using a combination of solid and solution-phase
chemistry. Structures of known chemokine antagonists were used to produce a pharmacophore which
served to guide monomer selection. Several combinations of monomers have resulted in providing novel
chemokine antagonists which in some cases display dual chemokine receptor antagonism.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Antagonist
Chemokine
Library
Solid-phase
The trafficking and homing of leukocytes in normal homeostasis
and in disease is under the control of a variety of cytokine and lipid
mediators. One family of small cytokines particularly involved in
inflammation are the chemokines.1 Chemokines belong to a family
of highly basic structurally related proteins comprising of 70–120
amino acids with amino acid sequence identities between 20–
70%. They are divided into four groups (CC, C, CXC, and CX3C)2
dependent on the spacing and number of the first two cysteine
residues. They mainly act upon neutrophils, monocytes, lympho-
cytes and eosinophils via G-protein coupled receptors, called che-
mokine receptors on the surface of their target cells. Chemokines
A survey of representative patent literature for ligands targeting
the CC class of chemokine receptor antagonists, namely CCR1,
CCR2, CCR3 and CCR5 led us to observe a structural resemblance
between the disclosed compounds (Fig. 1, for example compounds
1–4)6 allowing us to envisage synthesis of a focused library. We
postulated a simple pharmacophore (Fig. 2) with common features
including: a hydrophobic region, a basic nitrogen atom and a link-
ing carbon chain to a second hydrophobic region. With this guiding
pharmacophore and a synthetic route in-hand, we utilised mono-
mers from the Novartis Corporate Archive as well as commercially
available sources to generate a virtual library. The primary goal of
this project was to discover novel, single and multiple11 chemokine
receptor antagonists. We also challenged ourselves to utilize auto-
mation for the dispensing of reagents for synthesis and for purifi-
cation using solid-phase extraction and polymer supported
scavenger resins.12
A three dimensional library was enumerated using the initially
selected monomers. In an effort to focus the compound properties
towards the generally accepted range for optimal oral absorption
and pharmacokinetics,13 physico-chemical filters were applied
based upon molecular weight, clogP, number of H-bond donors
and acceptors prior to synthesis resulting in a 4320 compound
virtual library. The distribution against lipophilicity, polar surface
area and molecular weight is represented graphically in Figure 3.
To illustrate the synthetic methodology, and for simplicity,
this Letter will describe the syntheses of compounds 11, 12,
have been shown to be involved in
a variety of biological
processes3 including: growth regulation, hematopoiesis, angiogen-
esis, rheumatoid arthritis, osteoarthritis, multiple sclerosis, athero-
sclerosis, Alzheimer’s disease, asthma, chronic obstructive
pulmonary disorder, allergic disease and human immunodefi-
ciency virus (HIV) infection. At present more than 45 chemokines
and almost 20 chemokine receptors have been identified in
humans. As expected this discovery generated intense activity by
pharmaceutical companies, however though many antagonists
have progressed from discovery to the clinic only two: the CCR5
inhibitor maraviroc4 and CXCR4 antagonist plerixafor5 have so
far become registered drugs.
⇑
Corresponding author.
E-mail address: gurdip.bhalay@novartis.com (G. Bhalay).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.013

6250
G. Bhalay et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6249–6252
and 13 (Scheme 2). We developed an expedient, robust synthetic
Cl
N
route using a combination of solid-phase (Scheme 1) and solu-
tion phase chemistry (Scheme 2). Library production began with
the loading of 6 different amino alcohols on to the Wang linker14
by their reaction with Wang para-nitrophenyl carbonate 5 in
dichloromethane (Scheme 1, illustrated for aminoethanol 6).
The resulting carbamates (IR 1690 cmÀ1) were produced in good
yields (loading determined by detachment from solid-phase, and
was typically in the range 90–100%). Next the resin bound alco-
hols 7a were converted to the corresponding iodides 7b through
treatment with a mixture of triphenylphosphine, iodine and
imidazole in acetonitrile. The iodides 7b, were then reacted with
15 different amine fragments (para-fluorobenzoylpiperidine 8)
yielding the respective products (e.g., 9). Treatment with a mix-
ture of trifluoroacetic acid, DCM and water (50:40:10 v/v) de-
tached the 90 (6 amino alcohols  15 amines) required
building blocks from solid-phase as their TFA salt forms in good
yields (75–80%) and high purity (90–95%). The free base forms
10 were obtained after treatment with the basic AmberlystÒ
A21 resin. This step also had the effect of removing small
amounts of acidic artifacts thereby further increasing the purity
of the 90 desired compounds (>95% by LC-MS) with an expected
loss in yield (product yield 55–68%). Further treatment of 10
proceeded down three work-streams. Firstly reaction with 13
different sulphonyl chlorides (DCM, Et3N, rt, 5 h) was followed
N
O
F
O
HN
O
NH2
1
O
N
H
N
2
N
H
OMe
O
OMe
OMe
Cl
Cl
O
N
N
H
N
H
3
O O
S
N
N
4
Figure 1. Examples of CCR1,7 CCR2,8 CCR39 and CCR510 antagonists.
Figure 2. Schematic representation of the pharmacophoric features common to known CCR antagonists. Green spheres indicate hydrophobic regions that are linked through
spacers of different length via a basic nitrogen (blue sphere) and a donor/acceptor groups (cyan spheres). The donor/acceptor group can be an amide, a urea, a sulphonamide
or similar isosteres.
Figure 3. Physchem properties, distribution of c log P, PSA (Å2) and MW of the virtual library, outliers away from the majority are shown as closed circles.

G. Bhalay et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6249–6252
6251
Table 1
O
O
Number of hits discovered (hits defined as compounds exhibiting >50% inhibition
assayed at 1
O
lM)
NO2
5
Receptor
Number of hits
(>50% inhib. @ 1 M)
CCR1
CCR2
CCR3
CCR5
88
27
1136
228
H2N
a
OH
6
O
O
N
H
base (simplifying its removal through evaporation) and isocya-
nate scavenger resin removed unreacted amine 10 giving the ex-
pected sulfonamide products (e.g., 12 in Scheme 2). In the
second work-stream, reaction of 10 with 11 different isocyanates
(DCM, rt, 1 h) was followed by filtration of the reaction mixture
through cartridges containing strong cation exchange (SCX)
resin.
The desired products were retained on the sorbant allowing the
non-basic impurities to pass through. Products were then released
from SCX by flushing the cartridge with aqueous ammonia solution
(1 M) providing, after evaporation the expected products in good
yield and high purities. In the third work-stream, 10 was reacted
with 24 differing carboxylic acids (HATU, Et3N, DMF, rt, 1 h) then
the reaction mixture filtered through cartridges containing basic
alumina to remove the HOAt by-product as well as excess carbox-
ylic acids thus providing after evaporation the expected amides.
Overall the library provided 2979 compounds suitable for submis-
sion to the primary assay without the need for chromatographic
purification [purities >90% (determined using UV254 nM and ELS
detection) in >20 mg quantity by weight] resulting in a 69% success
rate.15
X
7a, X=OH
7b, X=I
b
O
c
F
HN
8
O
O
N
O
9
H
N
d
O
F
N
F
NH2
10
Trends in reaction failure were apparent amongst subsets of
reactants, presumably arising due to inherent differences in reac-
tivity brought-on through steric or electronic factors of monomers
necessary to generate a diverse library. For example 2,6-diisopro-
pylbenzoic acid presented significant steric challenge towards
amide coupling under these conditions.
Scheme 1. Reagents and conditions: (a) amino ethanol, DCM, rt, 18 h; (b) Ph3P,
imidazole, I2, THF, MeCN, rt, 4 h; (c) 8, DMF, Et3N, rt, 4 h; (d) 50% TFA in DCM, rt, 1 h
then AmberlystÒ A21.
These compounds were then evaluated for their CCR1, CCR2,
CCR3 and CCR5 binding affinity. Using scintillation proximity assay
O
format, radiolabelled chemokine ligands:
[ a125I]-
125I]-MIP-1
a
O
MCP1, [125I]-eotaxin, [125I]-MIP-1
a
respectively and Chinese ham-
ster ovarian (CHO) cell lines expressing the relevant chemokine
N
F
F
N
H
receptor.
11
12
Initially the library compounds were screened16 as 1
lM solu-
tions for single point determinations. Table 1 summarises our find-
ings. With the threshold for a hit set at >50% inhibition (testing at
1 lM) this library produced a >47% hit rate. From further analysis it
O
O
b
O O
10
was evident that the library was most successful at identifying hits
at the CCR3 receptor (1136 hits). This high hit-rate (over 38% at
CCR3) is indicative of the over proportional influence of the CCR3
patent literature as well as in-house knowledge of the CCR3 target.
Nevertheless it was gratifying to also find numerous hits for CCR1
(88 hits), CCR2 (27 hits) and CCR5 (228 hits).
N
N
S
N
N
H
N
c
Next in the screening cascade, IC50 s were determined17 for se-
’
O
lected compounds; the results for some of the compounds in
Scheme 3 are shown in Table 2. Compounds 14 and 15 showed
activity at CCR1 (IC50 261 and 387 nM respectively). 16 was
N
H
N
H
F
13
approximately equipotent for CCR2 and CCR3 being 1
CCR3 18, 19 and 20 were all below 100 nM. Notably 22 displayed
dual activity at CCR3 (1.2 M) and CCR5 (2.9 M).
lM. For
Scheme 2. Reagents and conditions: (a) carboxylic acid, HATU, Et3N, DMF, rt, 1 h;
(b) sulphonyl chloride, DCM, Et3N, rt, 5 h; (c) isocyanate, DCM, rt, 1 h.
l
l
In summary, we have synthesized a library of 2979 compounds
of high purity from a virtual set of 4320 compounds, using a com-
bination of solid-phase and solution-phase chemistry. This ligand
inspired approach has led to the discovery of novel CCR1, CCR2,
by sequential filtration through cartridges containing AmberlystÒ
A21 and then isocyanate scavenger resin. AmberlystÒ A21
converted the triethylamine hydrochloride by-product to its free

6252
G. Bhalay et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6249–6252
Table 2
IC50 data for selected compounds (values in nM)
HO
O
N
Compound
CCR1
CCR2
CCR3
CCR5
O
S
O
O
N
14
15
16
17
18
19
20
21
22
23
261
387
—
—
—
—
—
—
—
—
—
925
1878
—
—
—
—
—
—
—
N
H
—
1000
—
—
—
—
Cl
14
15
65
82
76
—
1190
—
—
—
—
1000
2850
1300
F
O
N
N
H
N
H
—
—
Br
O
N
N
H
CCR3 and CCR5 antagonists. We have also discovered compounds
which possess dual chemokine receptor activity; combinations of
CCR2–CCR3 and CCR3–CCR5 appear to be the most prevalent. Fur-
ther details will be published in due course.
O
F
16
17
18
Br
Acknowledgment
OCF3
O
We would like to thank John Hastewell who was involved in the
management of this initiative.
N
N
N
H
F
H
References and notes
F
N
OMe
O
S
1. Monika, S. Crit. Rev. Biotechnol. 2010, 30, 1.
2. Saunders, J.; Tarby, C. M. Drug Discov. Today 1999, 4, 80.
3. Horuk, R.; Ng, H. P. Med. Res. Rev. 2000, 20, 155.
N
N
N
O
4. Perry, C. M. Drugs 2010, 70, 1189.
5. Steinberg, M.; Silva, M. Clin. Therapeut. 2010, 32, 821.
6. For leading ref to patent structures see: Proudfoot, A. E. I.; Power, C. A.;
Schwarz, M. K. Exp. Opin. Invest. Drugs 2010, 19, 345.
7. Liang, M.; Mallari, C.; Rosser, M.; Ng, H.; May, K.; Monahan, S.; Bauman, J. G.;
Islam, I.; Ghannam, A.; Buckman, B.; Shaw, K.; Wei, G.-P.; Xu, W.; Zhao, Z.; Ho,
E.; Shen, J.; Oanh, H.; Subramanyam, B.; Vergona, R.; Taub, D.; Dunning, L.;
Harvey, S.; Snider, R. M.; Hesselgesser, J.; Morrissey, M. M.; Perez, H. D.; Horuk,
R. J. Biol. Chem. 2000, 275, 19000.
OMe
Cl
F
O
N
H
N
H
OMe
8. 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.
19
F
O
F
F
9. Gong, L.; Kertesz, D. J.; Smith, D. B.; Wilhelm, R. S. World Patent Application
WO 2000029377.
O
10. Baroudy, B. M.; Clader, J. W.; Josien, H. B.; McCombie, S. W.; McKittrick, B. A.;
Miller, M. W.; Neustadt, B. R.; Palani, A.; Steensma, R.; Tagat, J. R.; Vice, S. F.;
Laughlin, M. A. World Patent Application WO 2000066559.
11. (a) Pease, J. E.; Horuk, R. Curr. Top. Med. Chem. 2010, 10, 1351; (b) Pasternak, A.;
Goble, D. S.; Struthers, M.; Pasquale, P. V.; Ayala, J. M.; Di Salvo, J.; Kilburn, R.;
Wisniewski, T.; DeMartino, J. A.; Mills, S. G.; Yang, L. A. C. S. Med. Chem. Lett.
2010, 1, 14.
12. Bhattacharyya, S. Curr. Opin. Drug Disc. Dev. 2004, 6, 752–764.
13. (a) Van de Waterbeemd, H.; Smith, D. A.; Beaumont, K.; Walker, D. J. Med.
Chem. 2001, 44, 1313; (b) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney,
P. J. Adv. Drug Del. Rev. 2001, 46, 3.
N
N
N
H
H
O
20
F
CN
N
N
H
O
21
22
MeO
O
S
14. Reagents for solid-phase chemistry were purchased from: Agilent Technologies
or Merck Biosciences.
N
15. A representative set of 1H NMR were also aquired on the library collection (220
sampled, approx 8% of the collection).
N
O
OMe
16. Antagonist activity was confirmed for selected compounds using functional
assays.
17. Sources of Membranes:
F
MeO
O
CCR1 CHO-K1 cells. Used membranes at 1.3
of membrane in assay.
lg/mL final protein concentration
N
S
O
CCR2 CHO-ducX#84 cells. Used membranes at 40
concentration of membrane in assay.
lg/mL final protein
N
CCR3 RBL-2H3 cells. Used membranes at 100
of membrane in assay.
lg/mL final protein concentration
23
CCR5 CHO-K1 cells. Used membranes at 25–150
l
g/mL final protein
concentration of membrane in assay. Please note that the membrane
concentration used was dependent on the activity of individual batches of
membrane. The data (IC50) was fitted using a four parameter logistic curve
fitting in a 10 point curve. Each point on the curve was a duplicate.
F
Scheme 3. Selected compounds from the library.