Synthesis and SAR development of quinoline analogs as novel P2X7 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2019.04.033

The P2X7 receptor (P2X7R) plays an important role in diverse conditions associated with tissue damage and inflammation, suggesting that the human P2X7R (hP2X7R) is an attractive therapeutic target. In the present study, the synthesis and structure-activity relationship (SAR) of a novel series of quinoline derivatives as P2X7R antagonists are described herein. These compounds exhibited mechanistic activity (YO PRO) in an engineered HEK293 expressing hP2X7R as well as a functional response (IL-1β) in human THP-1 (hTHP-1) cellular assays. Compound 19 was identified as the most promising compound in this series with excellent cellular potency, low liver microsomal clearance, good permeability and low efflux ratio. In addition, this compound also displayed good pharmacokinetic properties and acceptable brain permeability (K_p,uu of 0.37).

Full text of DOI:10.1016/j.bmcl.2019.04.033

Accepted Manuscript
Synthesis and SAR development of quinoline analogs as novel P2X7 receptor
antagonists
Yufang Xiao, Srinivasa Karra, Andreas Goutopoulos, Noune Tahmassian
Morse, Susan Zhang, Mohanraj Dhanabal, Hui Tian, Jeyaprakashnarayanan
Seenisamy, Jayashankaran Jayaadevan, Richard Caldwell, Justin Potnick,
Matthew Bleich, Eugene Chekler, Brian Sherer
PII:
S0960-894X(19)30253-7
https://doi.org/10.1016/j.bmcl.2019.04.033
BMCL 26400
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
28 March 2019
18 April 2019
23 April 2019
Please cite this article as: Xiao, Y., Karra, S., Goutopoulos, A., Tahmassian Morse, N., Zhang, S., Dhanabal, M.,
Tian, H., Seenisamy, J., Jayaadevan, J., Caldwell, R., Potnick, J., Bleich, M., Chekler, E., Sherer, B., Synthesis and
SAR development of quinoline analogs as novel P2X7 receptor antagonists, Bioorganic & Medicinal Chemistry
Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.04.033
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and SAR development of quinoline analogs as novel P2X7 receptor antagonists
Yufang Xiaoa,*, Srinivasa Karraa,c, Andreas Goutopoulosa,*, Noune Tahmassian Morsea, Susan Zhanga, Mohanraj
Dhanabala, Hui Tiana, Jeyaprakashnarayanan Seenisamyb, Jayashankaran Jayaadevanb , Richard Caldwella, Justin
Potnicka, Matthew Bleicha, Eugene Cheklera, Brian Sherera
a EMD Serono Research and Development Institute, 45A Middlesex Turnpike, Billerica MA 01821
b Syngene International Ltd, Biocon Park, SEZ, Bommasandra Industrial Area - Phase-IV
Bommasandra-Jigani Link Road,Bangalore 560 099, India
c Current address: Alkermes, Inc, 852 Winter Street, Waltham, MA 02451
ABSTRACT
The P2X7 receptor (P2X7R) plays an important role in diverse conditions associated with tissue damage and
inflammation, suggesting that the human P2X7R (hP2X7R) is an attractive therapeutic target. In the present study,
the synthesis and structure-activity relationship (SAR) of a novel series of quinoline derivatives as P2X7R
antagonists are described herein. These compounds exhibited mechanistic activity (YO PRO) in an engineered
HEK293 expressing hP2X7R as well as a functional response (IL-1) in human THP-1 (hTHP-1) cellular assays.
Compound 19 was identified as the most promising compound in this series with excellent cellular potency, low
liver microsomal clearance, good permeability and low efflux ratio. In addition, this compound also displayed good
pharmacokinetic properties and acceptable brain permeability (Kp,uu of 0.37).
Key words: P2X7, P2X7R antagonist, quinoline, multiple sclerosis
*Corresponding author
Email: yufang.xiao@emdserono.com, Andreas.goutopoulos@emdserono.com

The P2X7 receptor (P2X7R) is an ATP ligand-gated ion channel and a member of the P2X superfamily of
purinoreceptors. It is expressed in cells of hematopoietic lineage such as peripherally localized macrophages and
monocytes, and in the CNS in microglia and astrocytes. The P2X7R is activated by high concentrations of ATP
resulting in a number of cell-specific downstream signaling events.1 Activation of P2X7R by extracellular ATP allows
2
the passage of small cations, such as Ca2+, Na+, and K+, across the plasma membrane resulting in the release of
the proinflammatory cytokines IL-1 and IL-18.3 The P2X7R has been associated with a number of diseases
including mood disorders, chronic inflammatory pain, rheumatoid arthritis, osteoarthritis and neurodegenerative
diseases, such as multiple sclerosis, Alzheimer’s and Huntington’s diseases.4-6
The P2X7R has attracted much attention from research groups in both academia and industry since first being
cloned in 1997.2 Several P2X7R antagonists have been evaluated in clinical trials for the treatment of rheumatoid
arthritis (RA)7-8 and Crohn’s Disease (CD).9 AZD9056 (1, Fig. 1) was the first P2X7R antagonist to be evaluated in
both RA and CD. Phase II clinical trials of compounds 1 and 2 (CE-224535) did not show efficacy in the treatment of
symptoms related to RA with doses up to 400 mg and 500 mg, respectively. GSK1482160 (3, Fig. 1) did not show
efficacy in a rheumatoid arthritis trial although a dose-dependent decrease in the release of IL-1 following LPS
stimulation was observed.10 Janssen recently reported that optimization of the pharmacokinetic properties of
several lead series of P2X7R antagonists led to the identification of the brain penetrant JNJ-42253432 and JNJ-
54175446 (4 and 5, Fig. 1) which entered phase I clinical trials for the treatment of mood disorders, depression and
pain.11-13
Figure 1. Selective P2X7R antagonists in clinical trials
Despite initial failures in the clinic, efforts to discover CNS-penetrant P2X7R antagonists for the treatment of
diseases, for example mood disorders and depression, continued to be an active area of research. Our own
internal efforts were geared towards designing novel chemotypes with good hP2X7R antagonist potency, low liver
microsomal clearance and favorable brain penetration characteristics to treat neurodegenerative diseases, such as
multiple sclerosis. AstraZeneca disclosed an adamantane amide-containing quinoline series.14-15 We explored this
motif as the basis for our literature to lead efforts seeing an opportunity whether it was possible to replace the
adamantyl group with less lipophilic moieties (general structure shown in Fig. 2). The chloro group at the 6-position
of the quinoline scaffold has been reported to be important for maintaining potency.14-15 We created structural
diversity within the series with combinations of the R groups at the 2-position of the quinoline moiety and the
various R1 groups on the cyclohexane ring to generate the leading compounds with excellent cellular potency and
ADME properties.

Figure 2. P2X7R antagonist general structure
All compounds described herein were prepared by the procedures outlined in Schemes 1 and 2.16-17 The synthesis
of these antagonists started from commercially available 2,6-dichloroquinoline. Treatment of 2,6-dichloroquinoline
with aluminum trichloride and bromine provided 2,6-dichloro-5-bromo-quinoline (6), which was transformed into
carboxylic acid (7) via addition of isopropyl magnesium chloride followed by quenching with carbon dioxide.18
Amide coupling of 2,6-dichloroquinoline-5-carboxylic acid (7) with 1-(4,4-difluorocyclohexyl)methanamine in the
presence of EDCI, HOBt and DIPEA furnished 2,6-dichloro-N-[(4,4-difluorocyclohexyl)methyl]quinoline-5-
carboxamide (8). Displacement of the 2-chloro group with amines or alcohols was easily accomplished upon
treatment of 2,6-dichloro-N-[(4,4-difluorocyclohexyl)methyl]quinoline-5-carboxamide (8) with the corresponding
amines in the presence of diisopropylethylamine (DIPEA) or the corresponding alcohols in the presence of sodium
hydride to afford the final antagonists 9-21.
Scheme 1. Synthesis of quinoline analogs with Nitrogen/Oxygen atom linker at 2-position of quinoline. Reagents
and conditions: a. AlCl3, Br2, 82%; b. iPrMgCl, CO2, THF, 0 0C, 42%; c. EDCI, HOBt, DIPEA, 1-(4,4-difluorocyclohexyl)-
methanamine, DMF; d. Amines, DIPEA, ACN; e. NaH, alcohols, DMF.
Scheme 2 illustrates the synthesis of analogs with a C-C bond at the 2-position of the quinoline. Suzuki cross-
coupling of 2,6-dichloro-N-[(4,4-difluorocyclohexyl)methyl]quinoline-5-carboxamide (8) with 2-(cyclopent-1-en-1-
yl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (10) in the presence of
a
catalytic amount of
bis(triphenylphosphine)palladium chloride and cesium fluoride provided 6-chloro-2-(cyclopent-1-en-1-yl)-N-[(4,4-
difluorocyclohexyl)methyl]quinoline-5-carboxamide (11). Hydrogenation afforded the antagonist 6-chloro-2-
(cyclopentyl)-N-[(4,4-difluorocyclohexyl)methyl]quinoline-5-carboxamide (22).

Scheme 2. Synthesis of quinoline analogs with Carbon atom linker at 2-position of quinoline. Reagents and
conditions: a. PdCl2(PPh3)2, CsF2, 1,4-dioxane, 90 0C; b. Pd/C, H2, MeOH.
Inhibition of hP2X7R activity was assessed by measuring the cellular uptake of YO PRO fluorescence dye in HEK293
cells expressing hP2X7R.19 The release of mature IL-1 from THP-1 cells after activation by BzATP was measured by
20
ELISA.
Substitution at the 2-position of the quinoline moiety was tolerant of a wide array of chemical
substituents allowing incorporation of a variety of functional groups linked through N, O or C atoms (Table 1). R
groups such as polar side chains (13), heterocyclic rings (11 and 12), and saturated rings (22) linked at the 2-
position were tolerated. Substitution at this position resulted in 2-8-fold increase in potency compared to the
unsubstituted compound 9.
Table 1. SAR of the 2-position on Quinoline
hTHP-1 IL-1 IC50 (nM)a
compound
R
hP2X7R YO PRO IC50
(nM)a
230
9
H
220
75
10
11
12
13
14
21
22
31
36
37
72
89
25
94
90
110
76
180
140
160
a The values are the average of at least three experiments. h denotes human.
These results led us to conduct further investigation of the substituents on the pyrrolidine ring of 12, which
showed high liver microsomal clearance (Table 2). To increase liver microsomal stability, polar amino groups, such
as 3-(2-hydroxyethylamino (15) and dimethylamino groups (16 and 17), were introduced. These analogs exhibited
relatively good cellular (IL-1release potency (<100 nM) and liver microsomal stability. However, these basic
amino groups resulted in low cell permeability and high efflux ratio. Replacement of the basic Nitrogen with a
Fluorine atom (19 and 20) or a Carbon atom (18) resulted in acceptable liver microsomal stability and reduced

efflux ratio from 114 (15) to 3.8 (18). Reduction of the hydrogen bond donor count in R2 from 1 (18) to 0 (19 and
20) further lowered the efflux ratio from 3.8 to 0.7-0.8. Compounds 19 and 20 were key analogs of interest with
good human cellular potency, low liver microsomal clearance, good permeability and low efflux ratio. However,
these compounds showed poor potency in the rodent species with IC50 values in the micromolar range. This poor
potency in the rodent species hampered the in vivo evaluation of these compounds in rodent-based disease model
studies.
Table 2. SAR of substituted pyrrolidines
Compound
R2
H
hP2X7R YO PRO
IC50 (nM)a
37
HLM/MLM/RMLb MDR-MDCK
(PA-B/ ER)c
hTHP-1 IL-1
IC50 (nM)a
108
12
15
16
17
18
19
20
87/49/57
17/10/10
80/24/25
45/50/22
32/23/46
10/49/21
12/10/31
31.8/1.2
0.18/114
6.0/6.9
10
50
27
47
8.2
19
12
75
40
6.2/7.2
100
109
104
16.5/3.8
14.1/0.85
18.4/0.69
aThe values are the average of at least three experiments. h denotes human.
bHLM: human liver microsomal clearance (L/min/mg); MLM: mouse liver microsomal clearance (L/min/mg); RML: rat liver microsomal
clearance (L/min/mg).
cMDR-MDCK: permeability as A to B (x10-6) cm/s. ER: efflux ratio.
All compounds with different substituents at the 2-position of quinoline provided the poor rodent potency. To
improve the rodent potency of 19, we turned our attention to explore the substituents (R1 group) on the
cyclohexane ring (Table 3). Difluoro substitution at the 3- or 4- position of the cyclohexane ring did not boost the
rodent potency (19 vs 23). Introduction of polar groups such as hydroxy at the 1-position of cyclohexane also did
not increase the rodent potency but improved liver microsomal stability (23 vs 24 and 25). Further exploration of
R1 group indicated that replacement of difluoro (24) at the 3-position of the cyclohexane ring with a methyl group
(26) increased rat potency by 7-fold and mouse potency by 3-fold (24 vs 26). With the (1S,3R)-hydroxymethyl on
the cyclohexane ring in place, adding polar side chains, such as dimethylamino or 2-hydroxy-2-isopropyl groups, to
the 3-position of the pyrrolidine ring generated 27 and 28, which showed 549 nM (27) and 130 nM (28) in the rat

P2X7R (rP2X7R) YO PRO assay. However, the permeability of these two compounds in the MDR-MDCK cell line was
poor (1.4-1.5 x10-6 cm/s) with high efflux ratio (>30).
Table 3. SAR of substituted cyclohexanes
Compound
R2
hP2X7R
YO PRO
IC50
hTHP-1
IL-1 IC50
(nM)a
rP2X7R
YO PRO
IC50
mP2X7R
YO PRO
IC50
HLM/
MLM/
RMLb
MDR-
MDCK
(PA-B/ ER)c
(nM)a
8.2
(nM)a
5100
(nM)a
2040
19
23
109
118
10/49/21
14.4/0.85
4.9
NT
2000
41/102/102 16.4/1.8
24
25
19
18
36.4
26.7
4300
9400
4830
6730
50/48/31
20/34/23
6.8/8.7
8.9/6.3
26
27
28
4.5
3.6
14
24.8
6.0
640
549
130
1400
2000
3690
21/65/23
44/52/32
75/68/10
8.9/4.3
1.4/39
1.5/34
7.4
aThe values are the average of at least three experiments. NT: not tested. h, r and m denote human, rat and mouse.
bHLM: human liver microsomal clearance (L/min/mg); MLM: mouse liver microsomal clearance (L/min/mg); RML: rat liver microsomal
clearance (L/min/mg).
cMDR-MDCK: permeability as A to B (x10-6) cm/s. ER: efflux ratio.

Table 4 summarizes the in vitro ADME and physicochemical data for compounds 19 and 25. Neither compound had
significant hERG activity. Plasma protein binding for both compounds was between 90-97% in all species tested.
Liver microsomal clearance varied across different species with the lowest clearance rates in human liver
microsomes. Compound 19 showed good permeability with Papp of 14.4x10-6 cm/s and low efflux ratio of 0.85 in
this assay.
Compounds 19 and 25 were dosed in female mice at 0.2 mg/kg by intravenous (iv) and 0.5 mg/kg by oral (po)
(DMSO/PEG/water as iv and po vehicles) routes of administration in order to assess their pharmacokinetic profiles.
The two compounds had a clearance rate of 4.19 L/kg/h (19) and 2.21 L/kg/h (25), moderate bioavailability of 31%
(19) and 32% (25), respectively (Table 5). Compounds 19 and 25 displayed low volume of distribution (1.84 L/kg
and 1.19 L/kg) and a short terminal half-life (0.48 h and 0.43 h), respectively. These two compounds were also
dosed at 5 mg/kg by oral administration (0.5%HPMC/K4M/0.25%/ Tween20 in pH3 buffer as vehicles) in male mice
to investigate their brain exposure. Compounds 19 and 25 had 3% and 11% mouse brain unbound fraction,
respectively (Table 5). Compound 25 also displayed a lower brain area under curve (brain AUC, 12.6 h*ng/mL) than
compound 19 (brain AUC: 30.95 h*ng/mL) because 25 had a relatively higher efflux ratio of 6.3 than 19 (efflux ratio
of 0.85). Although 25 had the higher mouse brain free fraction in comparison with 19, the calculated Kp,uu was
higher for compound 19 (0.37) than compound 25 (0.17) owing to the large difference of AUCs between plasma
and brain. The values of clogP are 4.1 and 3.2 for 19 and 25, respectively. Compound 19 is more lipophilic than
compound 25.
Table 4. in Vitro data for compounds 19 and 25
19
8.2
2040
109
25
18
6730
26.7
442
hP2X7R YO PRO IC50 (nM)a
mP2X7R YO PRO IC50 (nM)a
hTHP-1 IL-1 IC50 (nM)
MW
426
30
>30
hERG (M)
10/49/21
14.4/0.85
5.5/3.0/4.1
20/34/23
9/6.3
9.8/7.9/8.9
Clint h/m/r (L/min/mg)
MDR-MDCK (PA-B/ ER)b
PPB (unbound fraction%, human/mouse/rat)
a h and m denote human and mouse.
bMDR-MDCK: permeability as A to B (x10-6) cm/s. ER: efflux ratio
Table 5. Mouse pharmacokinetic parameters and free fraction data for compounds 19 and 25
19
25
AUC po (h*ng/mL) *a
AUC iv (h*ng/mL) *a
C max po (ng/mL) *
C max iv (ng/mL) *
72.9
238
75.1
545
0.48
4.19
1.81
31
143
451
160
738
0.43
2.21
1.19
32
T
(h)
½
Cl (L/kg/h)
Vd (L/kg)
F%
plasma AUC (h*ng/mL)
brain AUC (h*ng/mL)
mouse plasma (Fu, %)
mouse brain (Fu, %)
Kp,uu
84.7
30.95
3.0
3.0
0.37
135.9
12.6
7.9
11
0.13

*data is normalized to 1 mg/kg dose.
a AUC: area under curve; po: by oral; iv: intravenous.
In conclusion, our efforts have identified potent, brain-penetrant P2X7R antagonists with significant improvement
in potency on hP2X7R YO PRO and hTHP-1 IL-1 release assays. Compound 19 displayed excellent cellular potency,
low liver microsomal clearance, good permeability, low efflux ratio as well as acceptable brain penetration. But this
compound did not show the promising potency in the rodent species, which hampered the in vivo evaluation in
rodent-based disease model studies. Introduction of hydroxy group at the cyclohexane ring boosted rat and mouse
potency but had high efflux ratio. Extra efforts have been made to explore the other scaffolds which
exhibited the more potent at human, rat and mouse species with IC50 values below 100 nM. Results of
these efforts will be presented at the near future.
References and Notes
1. Surprenant, A, Rassendren, F, Kawashima, E, et al. Science. 1996;272:735.
2. Rassendren, F, Buell, GN, Virginio, C, et al. J Biol Chem. 1997;272:5482.
3. Ledeboer, A, Sloane, EM, Milligan, ED, et al. Pain 2005;115:71.
4. Romagnoli, R, Baraldi, PG, Cruz-Lopez, O, et al. Expert Opin Ther Targets. 2008;12:647.
5. Chrovian, CC, Rech, JC, Bhattacharya, A, et al. Prog Med Chem. 2014;53: 65.
6. Chen, X, Pierce, B, Naing, W, et al. Bioorg Med Chem Lett. 2010;20:3107.
7. Keystone, EC, Wang, MM, Layton, M, et al. Ann Rheum Dis. 2012;71:1630.
8. Duplantier, AJ, Dombroski, MA, Subramanyam, C, et al. Bioorg Med Chem Lett. 2011;21:3708.
9. Eser, A, Colombel, JF, Rutgeerts, P, et al. Inflammatory Bowel Dis. 2015; 21:2247.
10. Abdi, MH, Beswick, PJ, Billinton, A, et al. Bioorg Med Chem Lett. 2010;20:5080.
11. Letavic, ML, Savall, BM, Allison, BD, et al. J Med Chem. 2017;60:4559.
12. Chrovian, CC, Soyode-Johnson, A, Peterson, AA, et al. J Med Chem. 2018;61:207.
13. Lovenberg, T, Aluisio, L, Ceusters, M, et al. Proceedings of the CINP Thematic meeting on Stress,
Inflammation and Depression: Focus on Novel Antidepressant targets, Dublin, Ireland, 2015, June 4-6.
14. Evans, R, Eyssade, C, Ford, R. et al. PCT Int Appl. 2004;WO2004106305 A1.
15. Rech, JC, Bhattacharya, A, Branstetter, BJ, et al. Bioorg Med Chem Lett. 2016;26:4781.
16. The racemic mixture was separated by chiral chromatography using Chiralcel OD-H, 250x4.6 mm, 5M
column and 0.1% diethylamine in hexanes/isopropylalcohol from 80/20 to 0/100 in 10 minutes as mobile
phase and flow rate of 10 mL/mL.
17. Karra, SR, Xiao, Y, Seenisamy, J, et al. PCT Int Appl. 2015;WO2015187095 A1.
18. Ford, R, Thompson, T, Willis, P. PCT Int Appl. 2005;WO2005009968 A1.
19. IC50s were measured with hP2X7R YO PRO assay. HEK293 cells over-expressing hP2X7R were harvested
using HQTase reagent to detach them from T75. Harvested cells were centrifuged at 1200 rpm for 5
minutes at room temperature. The viability of cells was determined by Trypan blue dye and then cells
were plated at 10,000 cell/well density in 50 L volume in a 384W BD Poly-Lysine-coated plate and
0
incubated at 37 C. After overnight incubation, the culture medium was replaced with 35 L/well assay
buffer. Serial dilution and dispensing of compounds were performed using Bravo liquid handling
instrument. Compounds were introduced to assay plates starting at 2.5 M with three dilutions for 10
points. Cells were incubated with compounds for 20 minutes at room temperature. After the incubation
period, YO PRO dye (1 M) along with BzATP (10 M) were added to cells in 10 L/well. The plates were
centrifuged at 1000 rpm for 5 seconds and incubated at room temperature for 30 minutes. The uptake of
YO PRO dye inside the cells was measured using Envision Fluorescence plate reader instrument (Perkin
Elmer).
20. The effect of compounds on the release of mature IL-1 in the culture medium of THP-1 cells was
measured by sandwich ELISA. Cells were maintained in complete growth medium (RPMI 1640 + 10% HI-
FCS + 2mM L-glutamine + 1x PS). Every 3 days, the medium was renewed by diluting the cells 1/3 to 1/4

while cell density did not exceed 0.5 million cells per mL (seeding cell density at 1 x 105 /mL). THP-1 cells
were harvested from the flask in 50 mL by centrifugation for 3 minutes. Cells were re-suspended to 2x105
cells/mL in medium supplemented with 0.5 M PMA and incubated. They were then washed and
resuspended to 1.5x105 cells/mL in medium complemented with 10 ng/mL LPS, and then they were
0
primed for 4 hours at 37 C. After addition of 20 l of prediluted test compounds, blank, standard and
0
control reagents, cells were incubated for a further 30 minutes at 37 C and stimulated with 0.8 mM
BzATP for 30 minutes. The cells were centrifuged, supernatant was collected and the presence of mature
IL-1 was detected using Dual human IL-1 kit (DY201) following manufacturer's instruction.

Synthesis and SAR development of quinoline analogs as novel P2X7 receptor antagonists
Yufang Xiaoa,*, Srinivasa Karraa,c, Andreas Goutopoulosa,*, Noune Tahmassian Morsea, Susan Zhanga, Mohanraj
Dhanabala, Hui Tiana, Jeyaprakashnarayanan Seenisamyb, Jayashankaran Jayaadevanb , Richard Caldwella, Justin
Potnicka, Matthew Bleicha, Eugene Cheklera, Brian Sherera
a EMD Serono Research and Development Institute, 45A Middlesex Turnpike, Billerica MA 01821
b Syngene International Ltd, Biocon Park, SEZ, Bommasandra Industrial Area - Phase-IV
Bommasandra-Jigani Link Road,Bangalore 560 099, India
c Current address: Alkermes, Inc, 852 Winter Street, Waltham, MA 02451
Graphical abstract
21.