Identification of novel azaindazole CCR1 antagonist clinical candidates

Article abstract of DOI:10.1016/j.bmcl.2018.12.024

Exploring various cyclization strategies, using a submicromolar pyrazole HTS screening hit 6 as a starting point, a novel indazole based CCR1 antagonist core was discovered. This report presents the design and SAR of CCR1 indazole and azaindazole antagonists leading to the identification of three development compounds, including 19e that was advanced to early clinical trials.

Full text of DOI:10.1016/j.bmcl.2018.12.024

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of novel azaindazole CCR1 antagonist clinical candidates
⁎
Christian Harckena, , Daniel Kuzmicha,f, Brain Cooka,f, Can Maoa,f, Darren Disalvoa,f,
Hossein Razavia,f, Alan Swinamera, Pingrong Liua, Qiang Zhanga,f, Alison Kukulkab,f,
Donna Skowb,f, Mita Patelc,f, Monica Patelc,f, Kimberly Fletcherc,f, Tara Sherryc,f, David Josephc,
Dustin Smithc,f, Melissa Canfieldd, Donald Souzad, Matthew Bogdanffye, Karen Bergd,f,
Maryanne Brownd,f
a Medicinal Chemistry Department, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368, USA
b Compound Profiling Department, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368, USA
c Drug Discovery Support Department, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368, USA
d Immunology & Respiratory Disease Research Department, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368,
USA
e Non-Clinical Drug Safety Department, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877-0368, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Exploring various cyclization strategies, using a submicromolar pyrazole HTS screening hit 6 as a starting point,
a novel indazole based CCR1 antagonist core was discovered. This report presents the design and SAR of CCR1
indazole and azaindazole antagonists leading to the identification of three development compounds, including
19e that was advanced to early clinical trials.
Chemokine receptors
CCR1
Indazole
Scaffold hopping
Clinical compound
Introduction
the treatment of inflammatory disease. Antagonists that block the in-
teractions between CCR1 and its chemokine ligands could block che-
Chemotactic cytokine receptor 1 (CCR1) is a G protein-coupled re-
ceptor that belongs to a family of > 20 chemokine receptors that have
emerged as attractive targets for drug discovery.1,2 Various chemokines
interact with these chemokine receptors that are well known to mediate
basal and inflammatory leukocyte trafficking. CCR1 is expressed on
immune cells including monocytes, macrophages, T-lymphocytes,
neutrophils, basophils, eosinophils, NK cells, mast cells and dendritic
cells. The binding of the chemokines MIP-1α (CCL3), MCP3 (CCL7) and
RANTES (CCL5) to CCR1 is reported to play a role in the trafficking of
monocytes, macrophages and Th1 cells to the site of inflammation in
rheumatoid arthritis (RA) and multiple sclerosis (MS). Additionally, the
chemokines are all found in the Central Nervous System (CNS) of MS
patients, and MIP-1α and RANTES are found in the CNS of rodents in
the experimental autoimmune encephalomyelitis (EAE) disease re-
levant model. Macrophages and Th1 cells in the synovia of RA patients
are also major producers of MIP-1α and RANTES, and they recruit
leukocytes to the synovial tissues of RA patients resulting in chronic
inflammation. Thus, CCR1 has been regarded as a potential target for
motaxis of monocytes, macrophages and Th1 cells to the site of in-
flammation, ameliorating the chronic inflammation associated with
autoimmune diseases such as RA and MS.3,4 Despite set-backs in the use
of CCR1 antagonists in autoimmune disease and inflammation, phar-
macological intervention to disease via CCR1 remains of strong interest,
for example for the treatment of systemic fungal infection or in on-
cology.5,6
A number of publications have disclosed ligands as antagonists of
CCR1 and several have shown promising results in animal models of
disease.7–11 However, several early clinical candidates have failed to
show efficacy (Fig. 1):12 BX-471 (1) failed to show efficacy in a Phase II
clinical trial in patients with relapsing remitting MS.13–15 CP-481715
(2) failed to show efficacy in RA patients after 6 weeks of treatment.
Compound 1 and 2 may not have achieved sufficient inhibition of the
receptor since their PK profile led to high human doses. Millennium
compound MLN-3897 (3) was also stopped in the clinic after lack of
efficacy in a Phase II clinical trial for RA.16–18 It is presumed that
compound 3 could not be dosed sufficiently high due to safety
⁎ Corresponding author.
E-mail address: christian.harcken@boehringer-ingelheim.com (C. Harcken).
f Former employee.
https://doi.org/10.1016/j.bmcl.2018.12.024
Received 27 September 2018; Received in revised form 8 December 2018; Accepted 12 December 2018
0960-894X/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Harcken,C.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2018.12.024

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 1. Clinical CCR1 antagonists.
Fig. 2. Cyclization strategy leading to indazole scaffold.
Scheme 1. Reagents and conditions: (a) PyBop, Et3N, 3-trifluoromethylbenzyl
amine (b) R1-I, CuI, K2HPO4, trans-N,N’-dimethylcyclohexane-1,2-diamine, to-
luene or DMF.
Scheme 2. Reagents and conditions: (a) TMSCH2N2, MeOH, toluene or; me-
thanolic HCl, Δ (b) 4-fluoroiodobenzene, CuI, K2HPO4, trans-N,N’-dimethylcy-
clohexane-1,2-diamine, toluene or DMF (c) KOH, MeOH/H2O (d) PyBop, Et3N,
R2NH2.
limitations. In 2012, ChemoCentryx reported positive proof-of-clinical-
concept for an oral CCR1 antagonist CCX354 (4; dose 200 mg, once
daily) in a 12-week Phase II trial in patients (n = 160) with RA (ACR20
response at 12 weeks = 56%).19–22 In contrast, in 2014, results from a
12-week Phase II trial in patients (n = 123) with RA with BMS-817399
(5) did not show statistically significant differences in disease scores
(ACR20, DAS28) or MRI synovitis compared to placebo despite
achieving excellent coverage of the receptor.23 Additional compound
classes have been described and have advanced into clinical trials.24,25
In the immediately preceding publication in this issue, a novel
2

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 2
Benzylamide SAR.
Scheme 3. Reagents and conditions: (a) 4-fluorophenylhydrazine·HCl, NaOAc,
ethanol/water (b) CuI, trans-N,N’-dimethylcyclohexane-1,2-diamine, K2CO3,
NMP (c) CO, Pd(PhCN)2Cl2, dppf, Et3N, EtOH (d) KOH, MeOH/H2O (e) PyBop,
Et3N, R2NH2.
Compound
R2
CCR1
High-Throughput
Solubility pH 7
(μg/mL)
Ca2+ flux
IC50 (nM)
9e
3-(trifluoromethyl)-benzyl
2-(trifluoromethyl)-benzyl
3-chlorobenzyl
22
< 0.1
< 0.1
< 0.1
0.2
13a
13b
13c
13d
13e
13f
1200
450
Table 1
3-methoxybenzyl
> 3000
1700
N-1 substituted indazoles.
3-(methoxycarbonyl)-benzyl
3-carboxybenzyl
< 0.1
> 100
NT
> 3000
> 3000
3-(N-methylaminocarbonyl-
benzyl
13g
13h
3-(methylsulfonyl)-benzyl
3-(N-methylaminosulfonyl)-
benzyl
61
55
32
NT
13i
13j
3-(N-ethylaminosulfonyl)-benzyl
3-(N-isopropylaminosulfonyl)-
benzyl
170
180
0.9
0.7
13k
3-(N,N-dimethylaminosulfonyl)-
benzyl
44
< 0.1
13l
4-(methylsulfonyl)-benzyl
4-(N-methylaminosulfonyl)-
benzyl
68
< 0.1
2.6
13m
4.5
R1
CCR1 Ca2+ flux IC50 (nM)
13n
13o
13p
4-(N-isopropylaminosulfonyl)-
benzyl
5.0
6.0
12
0.7
0.2
5.5
Compound
4-(N,N-dimethylaminosulfonyl)-
benzyl
9a
9b
9c
9d
9e
9f
3,4-dichlorophenyl
benzyl
260
> 3000
750
phenyl
4-chlorophenyl
4-fluorophenyl
4-cyanophenyl
4-methoxyphenyl
2-fluorophenyl
3-fluorophenyl
methyl
210
22
> 3000
> 3000
2500
1100
> 3000
17
9g
9h
9i
13q
13r
13s
13t
(2-pyridyl)-methyl
(3-pyridyl)-methyl
(4-pyridyl)-methyl
> 3000
> 3000
1200
NT
NT
NT
21
9j
9k
4-fluoro-3-pyridyl
120
pyrazole scaffold 6, identified via a HTS screen, was disclosed as an
antagonist of CCR1 (Fig. 2).26,27 During the lead identification phase of
the program, various approaches were considered to identify additional
chemical matter, prior to transitioning to lead optimization. For various
reasons, which included achieving balance of the overall drug-like
properties and potency, the pyrazole scaffold compounds were de-
prioritized as development candidates. A number of cyclization strate-
gies were considered, using the pyrazole scaffold as an attractive tem-
plate for scaffold hopping. Several of these strategies yielded a 5,6-
fused aromatic ring system, providing an additional ring in the core to
afford new opportunities for SAR and the optimization of drug-like
properties. A cyclization via mode “a” would provide indole compounds
like 7. The Intellectual Property assessment was unfavorable for these
compounds. However, cyclization mode “b” via the methyl group and
the amide carbonyl oxygen, where the carbonyl double bond is in-
corporated in the ring and the carbonyl oxygen is mimicked
by X (where X = CH or N), afforded indazole compounds like 8. Com-
pound 8a (X = CH) could generate a potentially genotoxic aromatic
amine moiety through hydrolysis as a potential degradation product or
metabolite. This structural alert could be lessened for structure 8b
(X = N), and the SAR from the pyrazole series indicated that the amide
motif was critical for potency. Thus, compound 9 (X = CH) was pre-
pared as the initial analogue and found to have modest CCR1 antagonist
activity (IC50 = 260 nM in a MIP-1α-induced calcium flux assay in
CCR1-expressing cells).28
Scheme 1 shows how indazoles 9 with various substituents on the
indazole nitrogen were prepared.29 Commercially available indazole-4-
carboxylic acid (10), was coupled with 3-trifluoromethylbenzyl amine
to afford amide 11. A copper mediated coupling of various aryl or
heteroaryl iodides afforded analogs 9a–k.
To prepare compounds with variation of the benzylamine moiety,
the carboxylic acid 10 was esterified (Scheme 2). A copper mediated
coupling of 4-fluoroiodobenzene followed by ester hydrolysis gave the
3

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 3
Azaindazole core SAR.
Compound
W, X, Y
CCR1 Ca2+ flux IC50 (nM)
High-Throughput Solubility pH 7 (μg/mL)
9e
W, X, Y = CH
22
< 0.1
2.1
18a
18b
18c
W = N; X, Y = CH
W = CH; X = N; Y = CH
W = N; X = CH; Y = N
450
3.0
0.1
> 3000
NT
carboxylic acid 12. Activation of the carboxylic acid in the presence of
the benzyl amine provided amide analogues 13a–t. Certain substituents
(such as the alkylsulfones and sulfonamides) had to be introduced or
elaborated post coupling.29 Chiral amines were prepared using Ellman’s
sulfinamide chemistry from the corresponding bromopyridine.29–31
The 6-azaindazoles 18b, 19a–o were synthesized through copper
catalyzed cyclization of an arylhydrazone 15 derived from 3,5-di-
bromopyridyl-4-carboxaldehyde (14) to give the 4-bromo-6-azainda-
zole 16 (Scheme 3). Compound 16 was subjected to Pd-catalyzed car-
bonylation and hydrolysis of the ester to give the carboxylic acid 17.
The acid 17 was coupled with the appropriate benzylamines as in
Scheme 2 to generate 6-azaindazole analogues 18b, 19a–o.
potency in the CCR1 calcium flux assay (IC50 61 nM). Furthermore,
compound 13g had appreciable solubility in a high-throughput kinetic
solubility assay (32 μg/mL) compared to compound 9e (< 0.1 ug/mL).
The sulfonamides 13h–k also generally demonstrated acceptable po-
tency. The N-methylsulfonamide 13h (CCR1 IC50 55 nM) and the N,N-
dimethylsulfonamide 13k (CCR1 IC50 44 nM), were of similar potency
suggesting the NH was not acting as a hydrogen bond donor. The N-
ethylsulfonamide 13i and the N-isopropylsulfonamide 13j were 3-fold
less potent than 13h suggesting steric bulk may be undesirable for
potency in the calcium flux assay. As observed in the 3-position, the 4-
methylsulfone 13l (CCR1 IC50 68 nM) was more potent than com-
pounds with other groups in the 4-position. While the 3- and 4-me-
thylsulfones were equipotent, in contrast, the 4-sulfonamides 13m–o
were 10- to 40-fold more potent than the corresponding 3-sulfonamides
13h, 13j, and 13k, respectively. In general, the sulfonamide analogs in
Table 2 displayed lower aqueous solubility than the methyl sulfone
analogs. Interestingly, potent activity could be retained in the sterically
demanding, but also more basic N-methylpiperidinesulfonamide 13p.
It should also be noted that the amide NH is critical for CCR1 ac-
tivity. N-alkylation of the amide nitrogen of 13l with methyl, ethyl or
propyl groups gave significantly reduced potency in the calcium flux
assay.
The 5-azaindazole and 5,7-diazaindazole synthesis has been de-
scribed previously.29 Certain substituents (such as the alkylsulfones and
sulfonamides) had to be introduced or elaborated post coupling.32
Chiral amines were prepared using Ellman’s sulfinamide chemistry
from the corresponding bromopyridine.29–31
As shown in Table 1, a number of N-1 substituted indazoles were
surveyed. Benzyl substitution was not tolerated as in 9b. The un-
substituted phenyl compound 9c was less potent than the original lead
9a. The 4-chlorophenyl analog 9d was equipotent to 9a, however the 4-
fluorophenyl analog 9e (CCR1 IC50 22 nM) was 10-fold more potent
than 9a. Substitution by larger groups at the 4-position was not toler-
To further improve aqueous solubility, the SAR of the amide moiety
was expanded to include various pyridyl substitution patterns. The
unsubstituted pyridyl groups in 13q–s were not tolerated; solely the 4-
pyridyl analogue 13s retained detectable CCR1 activity (CCR1 IC50
1200 nM). However, in combination with methylsulfone substitution as
in 13t the compound maintained potent activity (CCR1 IC50 120 nM).
This finding was explored further and will be discussed in the context of
the azaindazoles (vide infra).
ated, as seen in examples 9f and 9g, each having
a CCR1
IC50 > 3000 nM. Ortho- and meta-fluoro substituted compounds 9h
and 9i displayed low micromolar activity in the calcium flux assay and
were 50 to 100-fold less potent than the para-isomer 9e. A number of
small alkyl R1 groups were prepared (e.g. methyl as in 9j) which were
not active at the highest concentrations tested in the CCR1 calcium flux
assay. Interestingly, the 4-fluoro-3-the pyrazole series where no polar
atoms were tolerated in this area of the binding pocket (see im-
mediately preceding publication in this issue).
In an attempt to further improve physical chemical properties of the
indazole CCR1 antagonists, heteroatoms were introduced into the core.
The 5-azaindazole 18a was 20-fold less active against CCR1 than the
comparator 9e (Table 3). However, the 6-azaindazole 18b showed
improved solubility and potency increased by an additional factor of
ten (CCR1 IC50 3.0 nM). Unfortunately, the solubility of the azainda-
zoles with the trifluoromethylbenzylamine moiety remained overall
low.
Next, the benzylamine moiety SAR was explored. As shown in
Table 2, several key substitution patterns emerged. In general, mono-
substitution in the 2-position had a detrimental effect on potency in the
CCR1 assay, exemplified by the ortho-CF3 substituted compound 13a.
Substitution in the 3-position was better tolerated, however compounds
containing groups such as 3-chloro as in 13b and 3-methoxy as in 13c
lost significant potency (CCR1 IC50 450 nM and > 3000 nM, respec-
tively). The ester 13d retained some micromolar activity, while the
carboxylic acid 13e and the N-methyl carboxamide 13f had CCR1 IC50
values > 3000 nM. The 3-methylsulfone 13g displayed acceptable
It should be noted that the 7-azaindazole isomer was successfully
synthesized only with additional substituents on the ring (and with
using a different benzylamine moiety). With said changes, it lost all
activity. The 5,7-diazaindazole 18c was not active (CCR1
4

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 4
Table 4 (continued)
Benzylamine SAR on the azaindazole core.
Compound
R2
CCR1 Ca2+
flux IC50
(nM)
High-Throughput
Solubility pH 7 (μg/mL)
19i
61
9
19j
1.4
43
19k
19l
0.2
0.3
58
48
Compound
R2
CCR1 Ca2+
flux IC50
(nM)
High-Throughput
Solubility pH 7 (μg/mL)
18b
3.0
0.1
18
19m
19n
19o
12
> 45
> 45
> 45
19a
19b
8.0
1.8
2.0
1.2
3.2
19c
19d
12
2
3.5
2.4
IC50 > 3000 nM).
It should also be noted that the potency increase of the 6-azainda-
zole compared to the indazole could be realized equally through the
introduction of a hydrogen bond acceptor substituent (such as CN) in
the 6-position of the indazole. However, the solubility of the 6-cya-
noindazoles was more difficult to improve.
With nanomolar CCR1 antagonists like 18b in hand, less potent
benzylamine moieties from Table 2 were re-evaluated. The 3-methyl-
sulfonyl compound 19a was 2- to 3-fold less potent than 18b but had
100-fold increased kinetic solubility (Table 4). Interestingly, the me-
thylsulfonyl and trifluoromethyl groups appeared to bind in different
orientations at the receptor, since 19b containing both groups was one
of the most potent compounds (CCR1 IC50 1.2 nM). The para-isomer
19c showed similar CCR1 activity (IC50 12 nM) as 19a but lower aqu-
eous solubility. The corresponding N-methylsulfonamide 19d was
slightly more potent. These trends confirmed data found in the in-
dazoles (see Table 2). However, when the sulfonyl substitution was
combined with introducing a pyridyl nitrogen atom, in this series, the
potency did not diminish. Methylsulfonyl pyridine 19e was equipotent
to 19a. Analysis of crystalline thermodynamic solubility (see also ad-
vanced profiling, vide infra) revealed that the additional nitrogen atom
improved solubility further. With a calculated logP of 2.0 and 91 Å3
polar surface area, this compound was expected to show excellent drug-
like properties. Increasing the size of the alkyl group in the sulfone
substituent to ethyl and cyclopropyl as in 19f and 19g, respectively, led
to a slight reduction in potency. The two regioisomer analogues of 19e,
19h and 19i, keeping the sulfone substituent and the pyridine nitrogen
atom in the preferred para- and meta-positions, showed a reduction of
potency too. Additional possible regioisomers were less active.
19e
19f
6.8
19
41
34
20
54
6
19 g
19 h
14
5

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 5
Profile of select compounds.
Compound CCR1 Ca2+ flux IC50 (nM) CCR1 Chemotaxis IC50 (nM)a HLMb CL (%Qh) CYP3A4 IC50 (μM) HT Solubility pH 7 (μg/mL) hERG IC50 (μM)c MW clogP
19e
6.8
0.2
2.0
28
34
> 30
17
20
30
425 2.0
454 2.9
452 2.9
(S)-19k
19n
0.3
2.0
18
> 45
> 45
> 30
68
< 30
> 30
NT, not tested.
a
CCR1chemotaxis in THP-1 cells35
.
b
c
Human liver microsomes (HLM).
Manual patch clamp system.
Lastly, the discovery was made that α-branching in the benzylic
clearance data and predicted bioavailability, a human efficacious dose
of 15 mg qd was predicted. Safety profiling showed acceptable safety
margins in acute and chronic, rodent and non-rodent toxicity studies.
The compound was advance into early clinical trials. Clinical data will
be published in a future publication.
position with small alkyl groups could significantly increase the po-
tency in the related pyrazole series (see immediately preceding pub-
lication in this issue). This concept was explored with the azaindazole
core. The indazoles 13 derived from racemic 1-(3-tri-
fluoromethylphenyl)ethylamine (CCR1 IC50 350 nM) and the commer-
cially available (S) and (R) enantiomer of 1-(3-trifluoromethylphenyl)
ethylamine to afford both single enantiomers was synthesized. This
confirmed that the S-enantiomer was the optimal stereochemistry for
further SAR studies with respect to substitution in benzylic position
((S)-enantiomer IC50 350 nM and (R)-enantiomer IC50 > 3000 nM in
the calcium flux assay).29 The racemic compounds 19j-l bearing a
methyl, ethyl and n-propyl group on the benzylic position showed ap-
proximately a 5-, 30- and 20-fold further increase in potency, respec-
tively. Compounds 19k and 19l showed sub-nanomolar CCR1 IC50 va-
lues of 0.2 and 0.3 nM, respectively. Despite the added lipophilicity, the
kinetic solubility of these compounds remained high, however, as ex-
pected, crystalline material showed lower thermodynamic solubility
(see advanced compound profiles). Finally, gem-dimethyl substitution
as in 19m was not preferred, however, the cyclized cyclopropylidine
and cyclobutylidine analogues 19n and 19o did not lose potency (IC50
1.8 nM and 2.0 nM, respectively) compared to the methyl analogues 19j
– simultaneously avoiding chirality and thus the need for an asym-
metric synthesis.
Compound (S)-19k showed 30-fold further improved potency. It
inhibited CCR1 RI in human whole blood with an IC50 of 1 nM (n = 6
donors). This represents the most potent CCR1 antagonist from any
internal or published series tested in the in-house assay, including
competitor's clinical candidates. Compound (S)-19k was indeed so
potent that despite a 200-fold reduced potency in rodent (mouse CCR1
IC50 44 nM) the compound was suitable to be profiled in a mouse col-
lagen antibody induced arthritis (CAIA) mouse model: Near maximal
inhibition (94%) of disease severity score was seen at day 14 with
100 mg/kg bid (S)-19k, approximately 50% efficacy was seen with bid
30 mg/kg. At 100 mg/kg bid (S)-19k, trough exposure 5 to 10-fold
above the (mouse) whole blood cellular IC50 values was obtained.
Exposures in mice treated with 30 mg/kg indicated compound plasma
concentration at or near the whole blood IC50. Histopathology results
demonstrated significant improvement in the fore paw histology score,
which correlated with disease efficacy as measured by the arthritic
score. These results confirmed the principle of CCR1 antagonism as a
key therapeutic target in this chronic model of arthritis and also un-
derscored the need for sufficient coverage to achieve efficacy.
Metabolic stability was confirmed to be low to moderate (hepatocyte CL
22%Qh). The most stable polymorph showed thermodynamic solubility
of 20 µg/mL at pH 7. The human dose for this compound was projected
at 0.5–1.25 mg qd. Unfortunately, during pre-clinical safety testing it
was discovered that the compound caused dramatic bone marrow ab-
lation in dogs with no therapeutic margin and development was ter-
minated. In the 4-week dog study, ex vivo inhibition of neutrophil CCR-
1 receptor internalization was maximal 1 mg/kg/day. Sudden and rapid
decreases in peripheral leukocytes, with no changes to T-cells, were
observed by day 8 at doses of 5 but not 1 mg/kg/day resulting in early
euthanasia of most dogs. These effects reversed in a 5-day survivor.
Histopathology included bone marrow hematopoietic cell depletion and
lymphoid depletion most severely in the thymus. Although not clasto-
genic, in vitro human lymphocyte chromosome aberration (HLA) test,
(S)-19k induced a dose dependent increase in centromeric disruption
and low incidence of endoreduplication. (S)-19k was positive in an in
vivo micronucleus assay and an in vitro micronucleus test in CHO cells
where anti-kinetochore staining confirmed aneugenicity. Although the
cause of toxicity remained undetermined, primary toxicity to the he-
matopoietic and lymphoreticular systems could be related to aneugenic
activity of (S)-19k.
More than 100 combination molecules in this small chemical space
(variation of sulfonyl substituent and position on pyridine isomers,
combined with small mono- or di-substitution on the benzylic position)
were prepared and evaluated. The most promising candidates were
resynthesized and, if required, resolved into pure enantiomers.
Table 5 shows advanced profiling data of three preferred candi-
dates.33 In general, the potency in this series in the calcium flux assay
was increased over 1000-fold compared to the lead compound 9a, with
only a modest increase in MW (414 vs 454 g/mole), while reducing the
overall lipophilicity of the series (4.0 vs 2.9 calc logP). In general, as
expected, compounds with lower clogP displayed improved aqueous
solubility. For this series, the CYP3A4 inhibition, HLM stability and
hERG inhibition parameters were typically acceptable.
Methylsulfone 19e was the first of the preferred compounds iden-
tified. It was selected for full pre-clinical candidate profiling: The
functional potency in combination with moderately low plasma protein
binding (74%) translated well into human whole blood activity (re-
ceptor internalization (RI) assay; IC50 47 nM; mean of 8 donors).34 The
compound was approximately 100-fold less potent against rodent CCR1
(mouse CCR1 IC50 670 nM) and was therefore not profiled in any in vivo
rodent models of inflammation. The metabolic stability was confirmed
as high (9%Qh in hepatocytes). This translated into an attractive pre-
clinical PK profile: The compound showed low clearance of 8%Qh, VSS
1.2 l/kg, and moderate bioavailability of 29% in a rat PK study
(Sprague-Dawley rat, po at 10 mg/kg in 0.5% methyl cellulose/0.015%
tween 80 in water; iv at 1 mg/kg in 70% PEG400 in water). Thermo-
dynamic solubility data of crystalline material confirmed acceptable
solubility of 24 µg/mL at pH 7. Based on human whole blood potency
(and the conservative assumption of a requirement of > 90% receptor
coverage at trough), the pre-clinical PK properties, in vitro human
Cyclopropane 19n was advanced into pre-clinical development as a
third azaindazole series compound.36 The compound provided a back-
up to 19e. It was about 5–10-fold more potent than 19e (Chemotaxis
IC50 2.0 nM; RI IC50 10 nM (n = 3 donors)), it did not show the tox-
icology findings of (S)-19k, and it retained excellent drug-like prop-
erties: Hepatocyte stability was further improved with a clearance
of < 8%Qh; crystalline solubility at pH 7 was slightly lower with 7 µg/
mL. In a rat PK study, the compound showed low clearance of 15%Qh,
V
SS 1.0 l/kg, and moderate bioavailability of 20% (Sprague-Dawley rat,
6

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
po at 3 mg/kg in 0.5% methyl cellulose/0.015% tween 80 in water; iv
at 1 mg/kg in 70% PEG400 in water). The human efficacious dose was
predicted to be 55 mg qd, ensuring once-a-day dosing with one dosing
unit. Clinical testing of the compound was halted before initiation of
Phase I due to strategic reasons. The compound is now being offered to
the scientific community for basic chemokine receptor research on the
Boehringer Ingelheim Open Innovation Portal OpnMe [www.opnme.
com] as one of the most selective and potent CCR1 antagonist in the
field.
CCX354-C treatment for rheumatoid arthritis: CARAT-2, a randomized, placebo
controlled clinical trial. Ann Rheum Dis. 2012;72:337–344.
23. Santalla III JB, Gardner DS, Duncia JV, Wu H, Dhar M, Cavallaro C, Tebben AJ,
Carter PH, Barrish JC, Yarde M, Briceno SW, Cvijic ME, Grafstrom RR, Liu R, Patel
SR, Watson AJ, Yang G, Rose AV, Vickery RD, Caceres-Cortes J, Caporuscio C, Camac
DM, Khan JA, An Y, Foster WR, Davies P, Hynes Jr J. Discovery of the CCR1 an-
tagonist, BMS-817399, for the treatment of Rheumatoid Arthritis. J Med Chem.
2014;57:7550–7564.
24. Hossain N, Mensonides-Harsema M, Cooper ME, Eriksson T, Ivanova S, Bergström L.
Structure activity relationships of fused bicyclic and urea derivatives of spirocyclic
compounds as potent CCR1 antagonists. Bioorg Med Chem Lett. 2014;24:108–112.
25. Trummer D, Walzer A, Groettrup-Wolfers E, Schmitz H. Efficacy, safety and toler-
ability of the CCR1 antagonist BAY 86–5047 for the treatment of endometriosis-
associated pelvic pain: a randomized controlled trial. Acta Obstet Gynecol Scand.
2017;96:694–701.
In summary, an (aza)indazole series of CCR1 antagonists has been
designed starting from a pyrazole core discovered in a HTS screen. The
(aza)indazole series presented in this report represents a novel class of
CCR1 antagonists which are structurally distinct from compounds that
have previously entered clinical trials. Medicinal chemistry optimiza-
tion in the indazole series afforded compounds with excellent human
pharmacological potency and promising drug like features. The leading
three compounds were advanced into pre-clinical development, with
19e advancing in Phase I clinical trials.
26. Cook BN, Harcken C, Lee TW-H, et al. Pyrazole Compounds as CCR1 Antagonists.
PCT Int Appl. 2009;WO2009137338.
27. Carter PH, Hynes J. N-aryl pyrazoles, indazoles and azaindazoles as antagonists of CC
chemokine receptor 1: patent cooperation treaty applications WO2010/036632,
WO2009/134666 and WWO2009/137338. Expert Opin Ther Pat.
2010;20:1609–1618.
28. Compounds are assessed for the ability to block the interaction of CCR1 and its ligand
in a functional cellular assay measuring calcium flux in response to MIP-1α in CCR1-
transfected cells. Published CCR1 reference compounds gave the following IC50 values
in this assay: BX-471 (1): 1.3 nM; CP-481715 (2): 59 nM; MLN-3897 (3): 0.8 nM;
CCX354 (4): 2.0 nM; BMS-817399 (5): 3.0 nM. Method A: Non-adherent cells pur-
chased from Chemicon Corporation (HTS005C), stably expressing recombinant CCR1
and G-alpha-16 are grown in RPMI 1640 medium (Mediatech 10-080-CM) supple-
mented with 10% heat-inactivated FBS, 0.4 mg/mL Geneticin and penicillin/strep-
tomycin. On the day of the assay, the cells are transferred to a beaker and dye-loaded
in bulk using a Fluo-4 NW Calcium Assay Kit with probenecid (Invitrogen F36205) at
0.8E6 cells/mL for 1 hour at room temperature. After 1 hour, they are seeded in a
384-well tissue culture-treated plate at a density of 20,000 cells/well. Appropriately
diluted test compound is added to the well to achieve a top concentration of 3,000 nM
(diluted 3-fold with 10 doses total). The final concentration of DMSO is 1%. The
buffer is HBSS (Invitrogen 14025) with 20 mM HEPES at pH 7.4. The cells are allowed
to incubate 1 hour in the dark at room temperature. The plates are transferred to the
FLIPR TETRA where MIP-1 alpha in 1% BSA is added at the EC80 final concentration.
Wells +/- MIP-1 alpha containing diluted DMSO instead of compound serve as the
controls. Intracellular calcium flux is recorded on the FLIPR TETRA, using excitation
at 470/495 nm and emission at 515/575 nm. Data are analyzed using Activity Base
software. Method B: Non-adherent cells purchased from Chemicon Corporation
(HTS005C), stably expressing recombinant CCR1 and G-alpha-16 are grown in RPMI
1640 medium (Mediatech 10-080-CM) supplemented with 10% FBS, 0.4 mg/mL
Geneticin and penicillin/streptomycin. On the day of the assay, the cells are loaded
with Calcium 4 dye (Molecular Devices R7448) with Probenecid (Invitrogen
P346400) at 8E5 cells/mL for 1 hour at room temperature. After 1 hour, they are
seeded in a 384-well tissue culture-treated plate at a density of 20,000 cells/well.
Appropriately diluted test compound is added to the well to achieve a top con-
centration of 3,000 nM (diluted 4-fold with 10 doses total). The final concentration of
DMSO is 1%. The buffer is HBSS (Invitrogen 14025) with 20 mM HEPES at pH 7.4.
The cells incubate 30 minutes at 37 C and then 30 minutes at room temperature. The
plates are transferred to the HAMAMATSU FDSS6000 where MIP-1alpha in 1% BSA is
added at the EC80 final concentration. All plates must be read within 4 hours of the
start of dye-loading. Wells +/- MIP-1alpha containing diluted DMSO instead of
compound serve as the controls. Data are analyzed using Activity Base software.
29. Disalvo D, Kuzmich D, Mao C, et al. Indazoles compounds as CCR1 receptor an-
tagonist. PCT Int Appl. 2009;WO2009134666.
Competing interest statement
The authors have no competing interests to declare.
References
1. Godessart N. Chemokine receptors attractive targets for drug discovery. Ann NY Acad
Sci. 2005;1051:647–657.
2. Horuk H, Ribeiro S. The clinical potential of chemokine receptor antagonists.
Pharmacol Ther. 2005;107:44–58.
3. Szczucinski A, Losy J. Chemokines and chemokine receptors in multiple sclerosis.
Potential targets for new therapies. Acta Neurol Scand. 2007;115:137–146.
4. Ubogu EE, Cossoy MB, Ransohoff RM. The expression and function of chemokines
involved in CNS inflammation. Trends Pharmacol Sci. 2006;27:48–55.
5. Lionakis MS, Albert ND, Swamydas M, Lee C-CR, Loetscher P, Kontoyiannis DP.
Pharmacological blockade of the chemokine receptor CCR1 protects mice from sys-
temic candidiasis of hematogenous origin. Antimicrob Agents Chemother.
2017;61:e02365–e2416.
6. Conroy MJ, Galvin KC, Kavanagh ME, et al. CCR1 antagonism attenuates T cell
trafficking to omentum and liver in obesity-associated cancer. Immunol Cell Biol.
2016;94:531–537.
7. Revesz L, Bollbuck B, Buhl T, et al. Novel CCR1 antagonists with oral activity in the
mouse collagen induced arthritis. Bioorg Med Chem Lett. 2005;15:5160–5164.
8. Horuk R, Shurey S, Ng HP, et al. CCR1-specific non-peptide antagonist: efficacy in a
rabbit allograft rejection model. Immunol Lett. 2001;76:193–201.
9. Gladue PR, Cole SH, Roach ML, et al. The human specific CCR1 antagonist CP-
481,715 inhibits cell infiltration and inflammatory responses in human CCR1
transgenic mice. J Immunol. 2006;176:3141–3148.
10. Amat M, Benjamin CF, Williams LM, et al. Pharmacological blockade of CCR1
ameliorates murine arthritis and alters cytokine networks in vivo. Br J Pharmacol.
2006;149:666–675.
11. Horuk R. Promiscuous drugs as therapeutics for chemokine receptors. Expert Rev Mol
Med. 2009;11:1–16.
30. Chelucci G, Salvatore B, Chessa S, Pinna GA. Soccolini F. An easy route to optically
active 1-substituted-1-pyridyl-methylamines by diastereoselective reduction of en-
antiopure N-tert-butanesulfinyl ketimines. Tetrahedron Asymmetry.
12. Zhang P, Dairaghi DJ, Jaen JC, Powers JP. Recent advances in the discovery and
development of CCR1 antagonists. Annu Rep Med Chem. 2013;48:133–147.
13. Liang M, Mallari C, Rosser M, et al. Identification and characterization of a potent,
selective, and orally active antagonist of the CC chemokine Receptor-1. J Biol Chem.
2000;275:19000–19008.
2006;17:3163–3169.
31. Reeves JT, Tan Z, Reeves DC, et al. Development of an enantioselective hydro-
genation Route to (S)-1-(2-(Methylsulfonyl)pyridin-4-yl)propan-1-amine. Org Process
Res Dev. 2014;18:904–911.
14. Horuk R. BX471: a CCR1 antagonist with anti-inflammatory activity in man. Mini-Rev
Med Chem. 2005;5:791–804.
32. Cook, B.; DiSalvo, D.; Harcken, C.; Kuzmich, D.; Liu, P.; Lord, J.; Mao, C.;
Raudenbush, B.; Razavi, H.; Reeves, J.; Song, J.; Swinamer, A.; Fandrick, D.; Lee, T.;
Neu, J.; Tan, Z. Azaindazole Compounds As CCR1 Receptor Antagonists. US Patent 7,
879,873.
15. Zipp F, Hartung HP, Hillert J, et al. Blockade of chemokine signaling in patients with
multiple sclerosis. Neurology. 2006;67:1880–1883.
33. Analytical data for select compounds: 19e: 1H-NMR (400 MHz, DMSO-d6) δ 9.62 (t, J
= 6 Hz, 1H), 9.42 (s, 1H), 8.92 (s, 1H), 8.75 (d, J = 5 Hz, 1H), 8.73 (s, 1H), 8.07 (s,
1H), 7.90-7.95 (m, 2H), 7.76 (dd, J = 5 Hz, J = 1 Hz, 1H), 7.47-7.52 (m, 2H), 4.75
(d, J = 6 Hz, 2H), 3.30 (s, 3H). ES-MS m/z = 426 (M + H)+. HPLC rt = 12.77
min, > 99%. 19k: 1H-NMR (400 MHz, DMSO-d6) δ 9.41 (s, 1H), 9.39 (d, J = 8 Hz,
1H), 8.95 (s, 1H), 8.76 (d, J = 5 Hz, 1H), 8.64 (d, J = 1 Hz, 1H), 8.16 (s, 1H), 7.89-
7.94 (m, 2H), 7.82 (dd, J = 5 Hz, J = 2 Hz, 1H), 7.45-7.51 (m, 2H), 5.17 (dd, J = 14
Hz, J = 8 Hz, 1H), 3.32 (s, 3H), 2.50 (m, 1.85-1.99, 2H), 1.00 (t, J = 7 Hz, 3H). ES-
MS m/z = 454 (M + H)+. HPLC rt = 13.91 min, 98.7%. 19n: 1H-NMR (400 MHz,
DMSO-d6) δ 9.74 (s, 1H), 9.42 (s, 1H), 8.94 (s, 1H), 8.69 (d, J = 1 Hz, 1H), 8.67 (d, J
= 5 Hz, 1H), 7.90-7.95 (m, 2H), 7.81 (d, J = 1 Hz, 1H), 7.46-7.54 (m, 3H), 5.17 (dd,
16. Gladue R, Zwillich S, Clucas A, Brown M. CCR1 antagonists for the treatment of
autoimmune diseases. Curr Opin Invest Drugs. 2004;5:499–504.
17. Vergunst CE, Gerlag DM, von Moltke L, et al. MLN3897 plus methotrexate in patients
with rheumatoid arthritis: safety, efficacy, pharmacokinetics, and pharmacody-
namics of an oral CCR1 antagonist in a phase IIa, double-blind, placebo-controlled,
randomized, proof-of-concept study. Arthritis Rheum. 2009;60:3572–3581.
18. Pusalkar S, Plesescu M, Milton M, Balani SK, Chowdhury S, Prakash S. Metabolism,
excretion and pharmacokinetics of MLN3897, a CCR1 antagonist, in humans. Drug
Metab Lett. 2016;10:22–37.
19. Dairaghi DJ, Zhang P, Wang Y, et al. Pharmacokinetic and pharmacodynamic eva-
luation of the Novel CCR1 antagonist CCX354 in healthy human subjects: implica-
tions for selection of clinical dose. Clin Pharmacol Ther. 2011;89:726–734.
20. Pease JE. CCR1 antagonism for the treatment of inflammatory diseases: focus on
CCX-354. Drugs Future. 2012;37:735–739.
J = 14 Hz, J = 8 Hz, 1H), 3.28 (s, 3H), 1.60 (s, 4H). ES-MS m/z = 452 (M + H)+
.
HPLC rt = 13.55 min, 99.3%.
34. Compounds are assessed for the ability to block the internalization of CCR1 in fresh
human whole blood: Compound dilutions are prepared in DMSO and in room tem-
perature Hanks buffered saline solution, +Mg, +Ca (Gibco 14025) with 20 mM
Hepes (Gibco 15630), pH 7.4, 0.2% BSA (Sigma 2058-25G). Pipetted to mix between
21. Pennell AMK, Aggen JB, Sen S, et al. 1(4-Phenylpiperazin-1-yl)-2-(1H-pyrazol-1-yl)
ethanones as novel CCR1 antagonists. Biorg Med Chem Lett. 2013;23:1228–1231.
22. Tak PP, Balanescu A, Tseluyko V, et al. Chemokine receptor CCR1 antagonist
7

C. Harcken et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
each dilution. Added 10 µl of 10X compound dilution or DMSO to 80 µl blood in
bullet tubes (BioRad 598281) and pipetted to mix 4-6 times. Mixed compound and
blood again with pipette set to 80 µl to insure total incorporation. Incubated for 30
min at 37°C. Prepared 10X MIP1α/Alexa 647; Almac Sciences; 100 µg/ml (10 µM) in
PBS/0.1% BSA. Added 10 µl of MIP-1a mixtures or RI buffer to appropriate tubes at
appropriate times. Mixed with pipette set to 80 µl to insure total ligand incorporation
and through mixing. Added 10 µl of the 50 nM Alexa solution to compensation tubes.
Mixed gently and incubated at 37°C for 40 min. Stopped reaction by putting plate on
ice and addition of 0.6 mL lysing buffer per well (R&D Whole Blood Erythrocyte
Lysing Kit Cat WL1000, Lot BB018042, exp 4/10 (dilute 1:10 in sterile distilled H2O)).
Mixed by pipetting up and down 300 µl ∼8 times. Incubated at 4°C for 20 minutes to
lyse. Centrifuged (1300 rpm, 4°C, 5 min) then removed supernatant with multi-
channel pipettor (400-500 µl). Repeated lysis with 600 µl fresh lysis buffer; pipetted
6-8X. Incubated 20 min depending on first lysis efficiency. Centrifuged, aspirated and
washed with 600 µl PBS/Azide (PBS (Gibco 14040) + 3% FCS (HI) (Gibco 16140-
071) + 0.1% sodium azide (Fisher S-227)). Centrifuged, aspirated and washed with
600 µl acid treatment to remove surface bound ligand. Centrifuged, aspirated and
washed with 600 µl PBS/Azide. Resuspended in 50 µl PBS +10% human serum
buffer. Mixed 3-4 times; Incubated 30 min at 4°C. Added 50 µl of either the isotype or
antibody in staining solution. Mixed 3-4X with addition. Incubated an additional 30
min at 4°C. Added PBS/5% human serum to 600 µl. Centrifuged, aspirated and wa-
shed with 600 µl PBS/Azide. Fixed the samples in 200 µl 1% formaldehyde for 10 min
room temp. Transferred to FACS plate (BD Cat#353918) and stored over night at 4°C
in dark. Read on LSRII
10 mM HEPES pH 7.4 + 0.1% BSA (Sigma #A7030)) are added to media (Cell Gro
#10-040-CM) add L-Glutamin (Cell-Gro # 25-005-CI (final 2 mM), Pen/strep (Cell-
Gro # 30-002-CI) final 100 Units & FBS (Cell-Gro #35-015-CV) final 10%) in T175
flask and grown in 37˚C incubator with 5% CO2. Cells grown between 2.0 x 105 and
1.0 x 106 cell/mL. Resuspended cells to 1.1 x 107 cells/mL. Compounds received as
dry powder are dissolved in 100% DMSO (Dri-solve EMD #MX 1457-6) at 10 mM or
as liquid in 15 µl tubes (10 mM). Added reagents by using a multidrop with a dis-
posable head: 90 µl assay buffer to 96 well plate (Falcon 35-1190), 60 µl cell sus-
pension to a 96 well plate (Falcon 35-1190), 280 µl of assay buffer to column 12 of
chemotaxis chamber, 280 µl of 6.75 nM RANTES to columns 1-11 of chemotaxis
chamber. Transferred 10 µl compound titration (100% DMSO) into 90 µl assay buffer
mix 6 times. Transferred 6.7 µl diluted compound to 60 µl cells in Falcon 96 well
plate. Cover using lids (Falcon 35-1191), mixed by shaking. Incubated cells for 30
minutes at 37°C and 5% CO2. Changed tips and transferred 32 µl diluted compound to
lower chamber containing 280 µL 6.75 nM RANTES (R&D systems #278-RN/CF).
Placed membrane on lower chamber. After 30 min incubation, mixed cells on shaker.
On the Cybio added 50 µl cells to top of membrane. Incubated 3 hours at 37°C and 5%
CO2. In a 96 well (Falcon 35-1190) added 260 µl assay buffer to column 1 row A-D
and 100 µl to columns 2-8 in Rows A-D. Added 50 µl cell suspension to column 1,
rows A-D. Titrate 1:2 to column 7. Added lid (Falcon 35-1191) place in incubator for 3
hours. On Cybio mixed % chemotaxis plate and added 40 µl to white half area plate
containing 40 µl Promega Cell Titer-Glo reagent (#G-7572). On cybio; mixed lower
chamber and added 40 µl to white half area plate containing 40 µl Promega Cell Titer-
Glo reagent (#G-7572). Mixed on shaker. Incubated at room temperature for 15 min.
Read luminescence on LJL Analyst. % Chemotaxis generated by excel template.
36. Marsini MA, Buono FG, Lorenz JC, et al. Development of a concise, scalable synthesis
of a CCR1 antagonist utilizing a continuous flow Curtius rearrangement. Green Chem.
2017;19:1454–1461.
35. Compounds are assessed for the ability to block the chemotaxis of THP-1 cells.
Published CCR1 reference compounds gave the following IC50 values in this assay:
BX-471 (1): 1.1 nM; CP-481715 (2): 39 nM; MLN-3897 (3): 1.8 nM; CCX354 (4): 4.0
nM; BMS-817399 (5): 1.5 nM. THP-1 cells (500,000 cells/well, each plate will use 60
µl of cells per well at 1.1 x 107 cells/mL. Washed cells 3 times in assay buffer (HBSS +
8