Discovery of 2-chloro-N-((4,4-difluoro-1-hydroxycyclohexyl)methyl)-5-(5-fluoropyrimidin-2-yl)benzamide as a potent and CNS penetrable P2X7 receptor antagonist

Article abstract of DOI:10.1016/j.bmcl.2010.03.094

Focused SAR studies were carried out around 5-heteroaryl and 1-amide portions of the 2-chlorobenzamide scaffold, resulting in the discovery of a potent, metabolically stable and centrally penetrable antagonist against P2X₇ receptor.

Full text of DOI:10.1016/j.bmcl.2010.03.094

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3107–3111
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of 2-chloro-N-((4,4-difluoro-1-hydroxycyclohexyl)methyl)-5-(5-flu-
oropyrimidin-2-yl)benzamide as a potent and CNS penetrable P2X7 receptor
antagonist
*
Xiangyang Chen , Betsy Pierce, Win Naing, Margaret L. Grapperhaus, Dennis P. Phillion
Department of Medicinal Chemistry, Pfizer St. Louis Research Laboratories, 700 Chesterfield Parkway West, Chesterfield, MO 63017, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Focused SAR studies were carried out around 5-heteroaryl and 1-amide portions of the 2-chlorobenzamide
scaffold, resulting in the discovery of a potent, metabolically stable and centrally penetrable antagonist
against P2X7 receptor.
Received 2 March 2010
Revised 23 March 2010
Accepted 26 March 2010
Available online 30 March 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
P2X7
Antagonist
SAR
CNS exposure
Pain
The P2X7 receptor (P2X7R), a ligand-gated ion channel, is ex-
pressed primarily on both peripheral and central immune cells
including macrophages, monocytes, microglia and astrocytes,1
and can be activated by elevated concentration of extracellular
demonstrated, but the significant role of activated central glia in
initiation/maintenance of chronic pain through regulating IL-1b
and perhaps glutamate suggests that P2X7R may be a central
player.9 Based on such rationale, we initiated a program to identify
P2X7R antagonists targeting central P2X7R for treatment of inflam-
mation and pain.
ATP (>100 l
M).2 Upon brief activation, P2X7R undergoes structural
changes to form a ‘typical’ cation channel to allow influx of Ca2+
and efflux of Na+/K+. If activation is repeated or prolonged, a large
non-selective membrane pore with increased cell permeability to
molecules up to 900 Da molecular mass is gradually opened, pre-
sumably as the result of activation of pannexin-1, a membrane
pore-forming protein.3 These functional changes trigger multiple
downstream effects including Caspase 1 and glial cell activations
which in turn lead to release of mature and biologically active IL-
1b and neurotransmitters (glutamate and GABA, etc.), implicating
that P2X7R plays a potential role in pain signaling.4 The therapeutic
importance of P2X7R in mediating inflammatory and neuropathic
pain is further supported by in vivo studies demonstrating that
P2X7R knockout mice were resistant to inflammatory/neuropathic
pain5 and P2X7R selective antagonists exhibited dose-dependent
antinociceptive effects in inflammatory/neuropathic pain animal
models.6,7 Therefore P2X7R as a therapeutic target for treatment
of inflammatory and neuropathic pain has attracted considerable
interest from the pharmaceutical industry.8 However, whether to
target peripheral or central P2X7R has not yet been clearly
Numerous classes of P2X7R antagonists were disclosed featur-
ing increasing drug-like properties, but no centrally penetrable
P2X7R antagonists have been specifically described so far.10
Our approach toward developing potent P2X7R antagonists with
CNS penetration was based on the potent 2-chlorobenzamide
template disclosed previously.11 Compound 1 was identified as
a suitable lead for optimization. It not only possesses desirable
physicochemical parameters under preferred CNS drug chemical
space (HBD 6 2, TPSA 6 75, etc.),12 but also has potent antago-
nism activity against P2X7R with IC50 of 32 nM and relatively
low Pgp efflux activity assessed by MDR1 of 2.0, although meta-
bolically unstable [human liver microsomal intrinsic clearance
(HLM CLint) of 80 mL/min/kg]. Here we report our SAR efforts
around 1 to improve its druggability leading to the discovery
of a potent and CNS penetrable P2X7R antagonist with an excel-
lent PK profile.
The preparation of compounds shown in Table 1 is illustrated in
Scheme 1. Commercially available 5-bromo-2-chlorobenzoic acid
and 1-(aminomethyl)cycloheptanol were coupled to give amide 2
in the presence of HBTU and DIEA which was subsequently con-
verted into the key intermediate, boron ester 3 according to a pub-
* Corresponding author. Tel.: +1 6365360177.
E-mail address: xiangyang.chen@hotmail.com (X. Chen).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.03.094

3108
X. Chen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3107–3111
Table 1 (continued)
a
b
Entry Ar
TPSA ClogD @
pH 7.4
IC50
HLM CLint
MDR1c
MW = 373
(nM)
(mL/min/kg)
HBD = 2
TPSA = 62
ClogD = 3.5
IC50 = 32 nM
N
N
H
N
4i
4j
75
75
2.7
3.7
22
36
1.4
MDR1 = 2.0
N
N
HLM CLint = 80 mL/min/kg
OH
Cl
O
1
8
53
21
nmd
N
N
Table 1
In vitro profiles of P2X7R antagonists with six-membered heteroaryl variation
F
Ar
4k
75
3.3
16
1.2
N
H
N
a
OH
IC50 of the P2X7R antagonism was determined by ATP-induced YO-PRO-1
Cl
O
uptake inhibition assay.19
b
Entry Ar
TPSA ClogD @
pH 7.4
IC50
HLM CLint
(mL/min/kg)
MDR1c
nmd
a
b
Human liver microsomal intrinsic clearance was measured according to the
standard human liver microsomal stability assay protocol.
(nM)
MDR1 was expressed as the ratio of permeabilityB?A/permeabilityA?B using the
c
MDCK cell line transfected with the human MDR-1 gene to assess the P-glycopro-
tein (Pgp) efflux activity.
4a
62
62
3.1
3.5
40
53
N
d
Not measured.
Table 2
In vivo pharmacokinetic profiles of P2X7R antagonists
1
32
80
2.0
1.2
Compound
Rat IV/PO crossover PKa
Rat b/pc
N
b
CL
Vdss
Effective T1/2
Bioavailability
(%)
(mL/min/kg) (L/kg) (h)
OMe
4k
7f
17
3.3
0.9
1.3
0.6
4.6
35
84
0.7
1.3
4b
71
3.3
213
122
N
a
b
c
Dosed at 1 mpk in male Sprague Dawley rats.
Effective T1/2 was calculated as Vdss/CL * ln2.
b/p was measured as the ratio of compound concentration in brain over that in
CN
plasma following a steady-state IV infusion in male Sprague Dawley rats.
4c
4d
4e
86
86
82
2.7
3.5
1.9
134
232
17
37
47
20
nmd
nmd
2.8
N
lished procedure.13 Palladium-catalyzed Suzuki cross-coupling of 3
with a variety of six-membered heteroaryl halides afforded 4 in
good overall yields.
Compounds shown in Table 3 were synthesized via a slightly
different sequence (Scheme 2). Palladium-catalyzed Suzuki cross-
coupling of commercially available boronic acid 5 with 2-chloro-
5-fluoropyrimidine gave the acid intermediate 6 which was then
converted into the desired amide 7 under typical amide coupling
conditions.
The preparation of the fluorinated amines is shown in Scheme
3. Treatment of commercial cycloheptanone with Selectfluor in
methanol at room temperature overnight yielded 2-fluoro-cyclo-
heptanone 8. Ketone 8 was readily converted into a cyanohydrin
ether intermediate via addition of trimethylsilyl cyanide catalyzed
by zinc iodide in methylene chloride.14 The crude cyanohydrin
ether was subsequently reduced with borane dimethyl sulfide
complex in THF to an ethanolamine which was further converted
into the Boc-protected intermediate 9 for the purpose of easy iso-
lation. Deprotection of Boc under the acidic conditions gave the
desired amine 10 as an HCl salt. Ring expansion of commercial
4,4-difluorocyclohexanone with trimethylsilyl diazomethane
catalyzed by boron trifluoride diethyl etherate afforded 4,4-diflu-
orocycloheptanone 11 in low yield.15 Following similar procedures
as for 10, both 11 and 4,4-difluorocyclohexanone were converted
into the corresponding amine HCl salts, 13 and 15, respectively.
Cyanation of commercially available 2-cyclohexen-1-one
with KCN gave 3-cyano-cyclohexanone 16,16 where the ketone
CN
N
CH2OH
N
F
4f
62
3.0
32
29
1.2
N
Cl
4g
4h
62
75
3.9
2.0
56
18
52
23
nmd
N
N
N
1.5

X. Chen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3107–3111
3109
O
O
Br
Cl
Br
Cl
Ar
Cl
B
a
b
c
H
N
H
N
H
N
OH
OH
OH
OH
O
O
O
Cl
O
2
3
4
Scheme 1. Reagents and conditions: (a) HBTU, DIEA, 1-(aminomethyl)cycloheptanol, DMF, room temperature, 1 h, 80%; (b) bis(pinacolato)diboron, PdCl2(dppf), KOAc, DMSO,
80 °C, 2 h, 85%; (c) aryl halide, Pd(PPh3)4, Cs2CO3, DME/H2O (5:1 v/v), 80 °C, overnight, 84%.
fixed (Table 1). Removal of the 5-methyl group (4a) slightly im-
Table 3
proved HLM stability which was further enhanced by re-introduc-
ing a 5-fluoro group (4f) with retained potency. Other slightly
larger 5-substitutions, whether polar or non-polar, caused some
degrees of potency loss (4b, 4c, 4g). The 5-hydroxymethyl analog,
4e, however, considerably improved both potency and HLM stabil-
ity, but raised efflux as well due to an additional HBD. Potency was
also weakened by moving substitution from 5 to 6 position (4c vs
4d). Insertion of a nitrogen into the 5-methylpyridyl ring led to
more active derivatives, 4h, 4i and 4j. Among them, 4j was the
most potent but least stable analog because of its highest ClogD va-
lue. Further replacement of the 5-methylpyrimidyl group with a 5-
fluoropyrimidyl group gave 4k with substantially enhanced HLM
stability while retaining P2X7R antagonism. Such transformation
was not suitable for both pyridazinyl and pyrazinyl analogs as
the resulting fluoro-derivatives post a structural alert.18 Overall,
good correlation between ClogD and HLM stability was observed
and low efflux was achieved if HBD 6 2 and TPSA 6 75. Therefore,
it is important to properly balance all physicochemical properties
of new synthetic targets in the multi-parameter lead optimization.
With good overall in vitro properties, 4k was advanced for in vivo
profiling.
The in vivo rat PK profile of compound 4k is summarized in Ta-
ble 2. Compound 4k had medium clearance of 17 mL/min/kg, rela-
tively short effective half-life of 0.6 h and good CNS exposure with
a brain/plasma ratio of 0.7. The in vitro assessment of HLM stability
and efflux correlated with in vivo properties reasonably well. How-
ever, the moderate PK properties of 4k were not good enough to
lead to adequately low predicted human dose. As low human dose
is desirable for a CNS drug to minimize potential adverse CNS ef-
fects, it was necessary to further improve PK properties,
particularly clearance. With an optimal 5-fluoropyrimidyl group,
the focus was then shifted to the amide portion of the molecule.
As shown in Table 3, reducing ClogD by shrinking the amide-
carbocyclic ring size from a seven- to six- and five-membered rings
continued to lower HLM intrinsic clearance. However, the potency
deteriorated accordingly. The seven-membered carbocyclic ring
appeared to be optimal for potency while having reasonable HLM
stability. Further Met ID studies of 4k using both HLM and RLM re-
vealed that metabolism took place predominantly on the carbocy-
clic ring with oxidative hydroxylation by cyps as the main
clearance mechanism, suggesting that the metabolic stability could
be improved if metabolically soft spots on the carbocyclic ring
were blocked. As well documented in the literature that fluorina-
tion can effectively block metabolism, four fluorinated analogs
were prepared and evaluated. Mono-fluorination of the seven-
membered carbocyclic ring (7c) maintained potency but only
slightly improved HLM stability. Presumably mono-fluorination
In vitro profiles of P2X7R antagonists with amide variation
F
N
N
H
N
R
Cl
O
Entry
R
ClogD @
pH 7.4
IC50
(nM)
HLM CLint (mL/
min/kg)
MDR1
1.2
4k
3.3
16
21
OH
7a
7b
7c
2.7
2.1
3.0
32
449
17
12
12
15
1.1
1.3
1.2
OH
O
F
OH
F
F
7d
7e
2.1
2.0
46
51
<8
10
nm
1.8
OH
OH
F
F
F
F
7f
1.8
27
<8.8
1.1
OH
functionality was further transformed into difluoro with Deoxoflu-
or to yield 17. Compound 17 was sequentially treated with LDA,
oxygen and acetyl chloride to afford a peracetoxy-nitrile interme-
diate 18.17 Without further purification, 18 was reduced with bor-
ane dimethyl sulfide complex in THF to an ethanolamine, followed
by protection of the amine with Boc2O to yield the intermediate 19.
Final deprotection of Boc gave the desired amine HCl salt 20.
In order to improve the metabolic stability of 1, replacement of
the metabolically liable 5-methylpyridyl as revealed in Met ID
studies was first investigated while keeping the rest of molecule
at the
tions away from the hydroxyl group while bis-fluorination at b-
and -position indeed efficiently reduced HLM clearance although
causing some drops in potency. More than 1 unit decrease in ClogD
a-position could not effectively block cyps’ oxidation at posi-
c

3110
X. Chen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3107–3111
F
F
B(OH)2
F
N
N
N
N
a
b
+
OH
N
N
H
N
Cl
O
Cl
OH
R
5
Cl
O
Cl
O
6
7
Scheme 2. Reagents and conditions: (a) Pd(PPh3)4, Cs2CO3, DMF/H2O (5:1 v/v), 80 °C, overnight, 83%; (b) HBTU, DIEA, RNH2, DMF, room temperature, 1 h, 90%.
F
F
F
a
b
c
BocHN
HCl-H2N
O
O
OH
OH
9
10
8
F
F
F
F
F
F
F
F
b
F
c
d
b
F
F
BocHN
HCl-H2N
O
O
O
OH
12
OH
13
11
F
c
F
F
HCl-H2N
BocHN
OH
15
OH
14
h
e
f
g
c
F
BocHN
F
F
HCl-H2N
F
NC
NC
O
NC
O
F
F
F
OOAc
OH
19
F
OH
20
16
17
18
Scheme 3. Reagents and conditions: (a) Selectfluor, MeOH, overnight, 53%; (b) (i) TMS–CN, ZnI2, CH2Cl2, 0 °C to room temperature, 2 h; (ii) BH3ÁDMS, THF, room temperature,
overnight; (iii) Boc2O, NEt3, DMAP, CH2Cl2, room temperature, overnight; 22–68% for three steps; (c) 4 N HCl in dioxane, room temperature, 1 h, 85–96%; (d) TMS–CH2N2,
BF3ÁOEt2, CH2Cl2, À10 °C, 2 h, 20%; (e) KCN, Et3NÁHCl, MeOH–H2O (2:1 v/v), 60 °C, 4 h, 45%; (f) Deoxofluor, CH2Cl2, room temperature, overnight, 50%; (g) (i) LDA, O2, AcCl, THF,
À78 °C; (ii) BH3ÁDMS, THF, room temperature, overnight; (iii) Boc2O, NEt3, DMAP, CH2Cl2, room temperature, overnight; 10% for steps (g) and (h).
2. Jacobson, K. A.; Jarvis, M. F.; Williams, M. J. Med. Chem. 2002, 45, 4057.
3. Gunosewoyo, H.; Coster, M. J.; Bennett, M. R.; Kassiou, M. Bioorg. Med. Chem.
by bis-fluorination certainly added additional beneficial effect on
HLM stability.20 Good potency, low HLM clearance and efflux,
2009, 17, 4861.
along with achirality, made 7f an attractive candidate for further
evaluation. The in vivo rat PK assessment of 7f showed that it
had much improved PK properties over 4j, with low clearance of
3.3 mL/min/kg, long effective half-life of 4.6 h and high oral bio-
availability of 84% (Table 2). Coupled with its excellent CNS expo-
sure (b/p of 1.3), 7f was endorsed for further development.
In conclusion, focused SAR studies were carried out to optimize
the drug properties of 1 following CNS drug design guidelines. By
keeping the ClogD low, fluorinating the carbocyclic ring and con-
straining the number of hydrogen bond donors to two, the meta-
bolic stability was significantly improved while maintaining
potent antagonism and CNS penetration. Such effort led to the dis-
covery of 7f as a potent P2X7R antagonist with excellent PK prop-
erties and CNS exposure. Further assessment of 7f could provide
more insights into the role that central P2X7R plays in mediating
pain signaling. This could in turn lead to development of therapeu-
tic P2X7R antagonists for treatment of inflammation and pain.
4. Carroll, W. A.; Donnelly-Roberts, D.; Jarvis, M. F. Purinerg. Signal. 2009, 5, 63.
5. Chessell, I. P.; Hatcher, J. P.; Bountra, C.; Michel, A. D.; Hughes, J. P.; Green, P.;
Egerton, J.; Murfin, M.; Richardson, J.; Peck, W. L.; Grahames, C. B. A.; Casula, M.
A.; Yiangou, Y.; Birch, R.; Anand, P.; Buell, G. N. Pain 2005, 114, 386.
6. Nelson, D. W.; Gregg, R. J.; Kort, M. E.; Perez-Medrano, A.; Voigh, E. A.; Wang,
Y.; Grayson, G.; Namovic, M. T.; Donnelly-Roberts, D. L.; Niforatos, W.; Honor,
P.; Jarvis, M. F.; Faltynek, C. R.; Carroll, W. A. J. Med. Chem. 2006, 49, 3659.
7. Honore, P.; Donnelly-Roberts, D.; Namovic, M. T.; Hsieh, G.; Zhu, C. Z.; Mikusa,
J. P.; Hernandez, G.; Zhong, C.; Gauvin, D. M.; Chandran, P.; Harris, R.; Medrano,
A. P.; Carrol, W.; Marsh, K.; Sullivan, J. P.; Faltynek, C. R.; Jarvis, M. F. J.
Pharmacol. Exp. Ther. 2006, 319, 1376.
8. Romagnoli, R.; Baraldi, P. G.; Cruz-Lopez, O.; Lopet-Cara, C.; Preti, D.; Borea, P.
A.; Gessi, S. Expert Opin. Ther. Targets 2008, 12, 647.
9. Sperlagh, B.; Vizi, E. S.; Wirkner, K.; Illes, P. Prog. Neurobiol. 2006, 78, 327.
10. Guile, S. D.; Alcaraz, L.; Birkinshaw, T. N.; Bowers, K. C.; Ebden, M. R.; Furber,
M.; Stocks, M. J. J. Med. Chem. 2009, 52, 3123.
11. Dombroski, M. A.; Duplantier, A. J.; Subramanyam, C. WO 2004099146, 2004.
12. Hitchcock, S. A.; Pennington, L. D. J. Med. Chem. 2006, 49, 7560.
13. Ishiyama, T.; Murata, M.; Miyaura, N. J. Org. Chem. 1995, 60, 7508.
14. Vincek, W. C.; Aldrich, C. S.; Borchardt, R. T.; Grunewald, G. L. J. Med. Chem.
1981, 24, 7.
15. Tagat, J. R.; Guzi, T. J.; Labroli, M.; Poker, C.; Xiao, Y.; Kerekes, A. D.; Yu, T.;
Paliwal, S.; Tsui, H.-C.; Shih, N.-Y.; Mccombie, S. W.; Madison, V. S.; Lesburg, C.
A.; Duca, J. S. WO 200698961, 2006.
16. Sekiyama, Y.; Palaniappan,N.; Reynolds, K. A.; Osada, H. Tetrahedron 2003, 59, 7465.
17. Freerksen, R. W.; Pabst, W. E.; Raggio, M. L.; Sherman, S. A.; Wroble, R. R.; Watt,
D. S. J. Am. Chem. Soc. 1977, 99, 1536.
References and notes
1. Collo, G.; Neidhart, S.; Kawashima, E.; Kosco-Vilbrois, M.; North, R. A.; Buell, G.
18. Blagg, J. Annu. Rep. Med. Chem. 2006, 41, 353.
Neuropharmacology 1997, 36, 1277.

X. Chen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3107–3111
3111
19. ATP-induced YO-PRO-1 uptake assay to assess potency of P2X7R antagonists by
measuring inhibition of dye uptake. Briefly, HEK-293 cells stably expressing
human P2X7R harvested with 0.05% trypsin were suspended at 2 Â 106 cells/
mL in the assay media (15 mM HEPES, 135 mM NaCl, 5 mM KCl, 1.8 mM CaCl2,
0.8 mM MgCl2, 5 mM glucose, 1% dialyzed fetal bovine serum, pH 7.4). Ten
media at various concentrations. The plates were incubated in dark at room
temperature for 15 min, followed by addition of 10 L of 10 mM ATP in 20 mM
HEPES at pH 7.4. After incubation for additional 2 h, fluorescence derived from
dye uptake was measured using a Tecan Safire fluorescence plate reader with
excitation at 450 nm and emission at 535 nm.
l
microliters of the cell suspension and 10
384-well non-treated black plates containing 20
l
L of 10
l
M YO-PRO-1 were added to
20. Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur, V. Chem. Soc. Rev. 2008, 37,
237.
lL of antagonist in the assay