Identification of second-generation P2X3 antagonists for treatment of pain

Article abstract of DOI:10.1016/j.bmcl.2018.02.039

A second-generation small molecule P2X3 receptor antagonist has been developed. The lead optimization strategy to address shortcomings of the first-generation preclinical lead compound is described herein. These studies were directed towards the identific

Full text of DOI:10.1016/j.bmcl.2018.02.039

Accepted Manuscript
Identification of second-generation P2X3 antagonists for treatment of pain
Anthony Ginnetti, Daniel Paone, Shaun Stauffer, Craig Potteiger, Anthony
Shaw, James Deng, James Mulhearn, Diem Nguyen, Carolyn Segerdell, Juliana
Anquandah, Amy Calamari, Gong Cheng, Michael Leitl, Annie Liang, Eric
Moore, Jacqueline Panigel, Mark Urban, Jixin Wang, Kerry Fillgrove, Cuyue
Tang, Sean Cook, Stefanie Kane, Christopher Salvatore, Samuel Graham,
Christopher Burgey
PII:
S0960-894X(18)30144-6
https://doi.org/10.1016/j.bmcl.2018.02.039
BMCL 25635
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
5 January 2018
15 February 2018
19 February 2018
Please cite this article as: Ginnetti, A., Paone, D., Stauffer, S., Potteiger, C., Shaw, A., Deng, J., Mulhearn, J.,
Nguyen, D., Segerdell, C., Anquandah, J., Calamari, A., Cheng, G., Leitl, M., Liang, A., Moore, E., Panigel, J.,
Urban, M., Wang, J., Fillgrove, K., Tang, C., Cook, S., Kane, S., Salvatore, C., Graham, S., Burgey, C., Identification
of second-generation P2X3 antagonists for treatment of pain, Bioorganic & Medicinal Chemistry Letters (2018),
doi: https://doi.org/10.1016/j.bmcl.2018.02.039
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.

Identification of second-generation P2X3 antagonists for
treatment of pain
Anthony Ginnettia, Daniel Paonea, Shaun Stauffera, Craig Potteigera, Anthony Shawa,
James Denga, James Mulhearna, Diem Nguyena, Carolyn Segerdella, Juliana
Anquandahb, Amy Calamarib, Gong Chengb, Michael Leitlb, Annie Liangb, Eric Mooreb,
Jacqueline Panigelb, Mark Urbanb, Jixin Wangb, Kerry Fillgrovec, Cuyue Tangc, Sean
Cookb, Stefanie Kaneb, Christopher Salvatoreb, Samuel Grahama, Christopher Burgeya
a Department of Medicinal Chemistry, MRL, Merck & Co., Inc., PO Box 4, West Point, PA 19486, USA
b Department of Pain Research, MRL, Merck & Co., Inc., PO Box 4, West Point, PA 19486, USA
c Department of Drug Metabolism, MRL, Merck & Co., Inc., PO Box 4, West Point, PA 19486, USA
Abstract—A second-generation small molecule P2X3 receptor antagonist has been developed. The lead optimization strategy to address
shortcomings of the first-generation preclinical lead compound is described herein. These studies were directed towards the identification
and amelioration of preclinical hepatobiliary findings, reducing potential for drug-drug interactions, and decreasing the projected human
dose of the first-generation lead.

1
2
Despite the presence of numerous treatment options, pain due to osteoarthritis is still a
primary complaint of older populations; furthermore, only a small percentage of pain
sufferers receive a prescription treatment. New drugs with minimal side effects,
improved efficacy, and ease of administration are sought for the treatment of pain. The
P2X3 receptor is an ATP-gated ion channel expressed on nociceptors (primary afferent
neurons that sense painful stimuli) with limited receptor expression in other tissues.1 Its
role in the transmission of nociceptive stimuli along with its selective expression pattern
makes the P2X3 receptor an attractive target with the potential to provide treatment of
inflammatory, visceral, and neuropathic pain states.2 Indeed, P2X3 receptor antagonists
identified in our laboratories have already demonstrated dose-dependent reversal of pain
in the rat Complete Freund’s Adjuvant (CFA) assay.3
3
4
5
6
7
8
9
10
11
12
13
14
15
Figure 1. Profile of MK-3901.
16
17
MK-3901, disclosed previously, was the first P2X3 receptor antagonist preclinical
candidate identified by our laboratories. MK-3901 is a potent antagonist of the P2X3
2

18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
receptor as measured by our Ca++ mobilization FLIPR and patch clamp electrophysiology
(EP) assays (FLIPR IP = 21 nM, EP IC50 = 24 nM).4 Importantly, when evaluated in the
rat CFA model of inflammatory pain, MK-3901 demonstrated efficacy similar to that of
our positive comparator, naproxen, after 60 mg/kg (p.o.) administration with a measured
EC90 of approximately 3 µM.3 However, in non-human preclinical safety studies MK-
3901 was shown to induce hyperbilirubinemia across several species, most notably a
greater than 15-fold increase in total bilirubin levels in rhesus at exposures approximately
6-fold over the projected clinical AUC. The elevated bilirubin levels associated with
MK-3901 could potentially be attributed to off-target inhibition of UGT1A1 (IC50 = 1
µM), a glucuronosyltransferase involved in bilirubin metabolism.4 MK-3901 exhibited
excellent bioavailability (%F) in rat, dog, and rhesus monkey pharmacokinetic (PK)
studies (F = 60%, 68%, and 47%, respectively), however the compound was found to be
extensively metabolized in human liver microsomes. Analysis of these metabolites
identified both the northern pyridine methyl group and the isoxazoline ring as metabolic
soft-spots. These factors lead to unacceptably high predicted dose and dose frequency
(420 mg TID). In addition to hepatobiliary findings and a high projected human dose,
MK-3901 also presented the potential for drug-drug interactions (DDI) due to
cytochrome P450 (CYP2C9) inhibition and pregnane X receptor (PXR) activation as
shown in Figure 1.
Using MK-3901 as a starting point, we sought to identify a second-generation potent and
selective P2X3 receptor antagonist with reduced risk of hyperbilirubinemia (through
attenuating off-target UGT1A1 inhibition) and reduced potential for DDI (through
minimized CYP inhibition/induction). We also targeted a lower projected clinical dose
3

42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
through improved PK, and in-vivo potency in our rat CFA study. We describe herein the
design and synthesis of potent P2X3 receptor antagonists which address these limitations.
The team took a modular approach to SAR exploration, examining the three substituent
positions of the central benzene ring. We initially set out to replace the eastern
isoxazoline side chain of MK-3901 to eliminate its metabolic liability and identify the
minimum pharmacophore necessary for activity. Through incorporation of a wide variety
of truncated eastern substituents (select examples are shown in Table 1), it was
discovered that potency could be maintained while modulating plasma protein binding.
Another advantage of investigating the minimum pharmacophore was that there appeared
to be a correlation between eastern substituent size and UGT1A1 activity, with smaller
substituents leading to reduced UGT1A1 inhibition across multiple structural series.6
Unfortunately, incorporation of smaller eastern substituents led to higher in vivo plasma
clearance in rat and dog; additionally,the risk for drug-drug interaction persisted for these
compounds as CYP inhibition and PXR activation remained for many analogs. We were
most encouraged by the tertiary-hydroxyl compound 4 which, despite lacking the entire
pyridyl isoxazoline of MK-3901, lost only a modest amount of P2X3 potency, and
displayed increased unbound fraction in rat plasma when compared with MK-3901.
60
61
Table 1. SAR of eastern substituents.
4

62
63
64
65
66
67
68
69
70
71
72
73
While SAR of the eastern portion of the molecule was ongoing, optimization of the
western amide was being concomitantly investigated with the plan to combine the
independent findings in both areas into a final optimized molecule. Utilization of the
potency-enhancing eastern fluoropyridine present in compound 2 and varying the identity
of the western amide led to a large class of compounds, representative examples of which
are shown in Table 2. The team discovered early on in investigating western substituents
that a variety of 5- and 6-membered aromatics and heterocycles were well tolerated in
place of the western fluoropyridine of MK-3901. Another interesting finding was that the
(R)-methyl substituent at the benzylic position in compounds 2-7, 10 was preferred from
5

74
75
76
77
a binding perspective when compared with larger substituents (compound 13),
disubstitution (compound 12) or the unsubstituted methylene (compounds 14, 15). One
interesting exception to this trend is the triazole-containing compound 9, where it has
been rigorously confirmed that (S)-methyl stereochemistry is preferred.
78
79
Table 2. SAR of western amides
80
81
82
83
Table 2 shows that while we had achieved potency below 10 nM in many cases, this class
of eastern fluoropyridines continued to suffer from relatively high rat plasma protein
6

84
85
86
87
88
89
90
91
92
93
94
binding with the exception of the two pyridine N-oxide compounds, 5 and 7. The team
took particular interest in these amides which also showed substantially increased FLIPR
potency.
Next, we then set forth to conjoin the independent SAR findings of the eastern and
western regions into a single molecular entity. Indeed, as Figure 2 shows, when
incorporated into the eastern tertiary hydroxyl scaffold, the pyridine N-oxide amide gives
not only enhanced FLIPR potency and reduced plasma protein binding, but also reduced
PXR activation and CYP inhibition when compared to its non-oxidized pyridine
counterpart.
95
96
Figure 2. Advantages of Pyridine N-oxides
97
98
Having addressed the CYP inhibition/PXR activation and high plasma protein binding of
MK-3901, the team focused on improving the pharmacokinetics of the potent and
selective pyridine N-oxide antagonists. The methyl group of the northern pyridine ring
had been identified as another site of oxidative metabolism by metabolite identification
studies. As such, SAR exploration was conducted to find alternative northern rings with
reduced metabolic potential.
99
100
101
102
103
7

104
105
106
Table 3. SAR of northern substituents
107
108
109
110
111
As shown in Table 3, while aryl fluorine and chlorine substituents were well tolerated
replacements for the metabolically labile methyl group of compound 17, the
8

112
113
114
115
116
117
118
119
120
121
122
123
124
125
corresponding fluoropyridine analogs were found to significantly reduce FLIPR potency
(compound 20). Additionally, the fluoropyridines suffered from low exposures and short
half-lives by nature of their high in vivo intrinsic clearance. Conversely, the northern
chloropyridines, represented by compound 21 were found to have acceptable potency in
the FLIPR assay to go along with increased exposure and lower intrinsic clearance.
Additionally, compound 21 was subsequently shown to possess an exceptional off-target
profile and increased in vivo potency in the rat CFA model of inflammatory pain (EC90 =
0.16 µM) when compared with MK-3901 (EC90 = 3 µM). Pertaining to the
hyperbilirubinemia liability of MK-3901, compound 21 was found to display an IC50 of
greater than 20 µM in the UGT1A1 inhibition assay. Lastly, compound 21 was predicted
by allometric scaling to be a relatively low clearance (1.2 – 3.6 mL/min/kg) compound
with a terminal half-life of 3 - 10 hr in humans, likely suitable for once daily dosing
schedule.
126
127
128
Figure 3. Profile of Compound 21.
9

129
130
131
132
133
134
135
136
137
138
139
140
141
The initial synthesis of compound 21 is shown in Scheme 1. Treatment of commercially
available 22 with bis(pinacolato)diboron under standard Suzuki Conditions7 results in
selective cross-coupling with the iodide to give the boronic ester 23. A second Suzuki
reaction with 2-bromo-5-chloropyridine using water soluble triphenylphosphine
conditions8 gives the biphenyl product 24. Here we observe a loss of desired product to
subsequent cross-coupling of 24 with the starting boronate 23 leading to a modest 41%
yield after 2 steps. A third cross-coupling with isopropenyl boronic ester yields the
isopropenyl product 25. After basic hydrolysis of the ester, a solvolysis reaction on the
isoprene moiety in aqueous methanesulfonic acid gave the tertiary hydroxyl intermediate
26 in 59% yield after two steps. Completion of the sequence was affected through a
standard amide coupling with amine 27 to give compound 21 in 16% overall yield in 6
steps.
142
143
144
145
Scheme 1. Initial synthesis of Compound 21
10

146
147
148
149
150
151
152
153
In an attempt to increase the yield and efficiency of the initial route, we turned our
attention to the first two Suzuki cross-coupling steps as areas for potential improvement.
In a recent pair of publications9,10 from our labs, it was demonstrated that copper (I)
facilitates Suzuki reactions of historically challenging 2-heterocyclic boronates. Through
the use of this methodology, the team was able to directly couple the chloropyridyl group
to 22 thereby eliminating one of the cross-coupling steps from the route and increasing
the overall yield more than 2 fold as shown in Scheme 2.
154
155
156
Scheme 2. Copper-facilitated Suzuki coupling to give intermediate 24.
157
158
159
160
161
162
163
164
165
166
In summary, we have taken a modular approach to lead optimization in identifying
second-generation P2X3 receptor antagonists which address the issues observed in our
first-generation preclinical candidate, MK-3901. We have sought to improve PK and in
vivo potency in our preclinical pain model in an effort to reduce the predicted human
dose. Truncation of the isoxazoline side chain of MK-3901 eliminated a metabolic
liability present in the initial lead. Through incorporation of the pyridine N-oxide amide
side chain, we were able to enhance both potency and free fraction. Replacement of the
northern methylpyridine present in MK-3901 with chloropyridine gave compound 21
which not only displays the pharmacokinetic and physical properties to enable QD
dosing, but also drives down unwanted off-target activities, namely UGT1A1 as well as
11

167
168
169
CYP inhibition and PXR activation. In addition to identifying an improved candidate
molecule, we improved the synthetic efficiency of the first-generation synthesis of
compound 21.
170
171
172
References and notes
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
1. Bunstock, G. Pharmacol. Ther. 2006, 110, 433
2. Ford, A. P. Pain Manage. 2012, 2, 267
3. Burgey, C. Abstracts of Papers, 241st ACS National Meeting & Exposition, Anaheim, CA, Mar 27–31,
2011; American Chemical Society: Washington, DC, 1996; MEDI15.
4. Assay protocols for FLIPR, EP, CFA models as described in: Burgey, C. S.; Deng. Z. J.; Nguyen, D.
N.; Paone, D. V.; Potteiger, C. M.; Vacca, J. P. WO Patent 058298 A1, 2009
5. Zhang D.; Chando, T; J.; Everett, D. W.; Patten, C. J.; Dehal, S. S.; Humphreys, W. G. Drug. Metab.
Dispos. 2005, 33, 1729
6. Data not shown. As measured by inhibition of UGT1A1 catalyzed estradiol-3-glucuronidation in
human liver microsomes.
7. Ali, A.; Bohn, J.; Deng, Q.; Lu, Z.; Sinclair, P. J.; Taylor, G. E.; Thompson, C. F.; Quraishi, N. WO
Patent 100298 A1, 2005
8. Burgey, C. S.; Deng, Z. J.; Nguyen, D. N.; Paone, D. V.; Potteiger, C. M.; Stauffer, S. R.; Segerdell,
C.; Nomland, A.; Lim, J. J. WO Patent 111058 A1, 2010
9. Deng, J. Z.; Paone, D. V.; Ginnetti, A. T.; Kurihara, H.; Dreher, S. D.; Weissman, S. A.; Stauffer, S.
R.; Burgey, C. S. Org. Lett. 2009, 11, 345
10. Crowley, B. M.; Potteiger, C. M.; Deng, J. Z.; Prier, C. K.; Paone, D. V.; Burgey, C. S. Tetrahedron
Lett. 2011, 52, 5055
191
12

192
193
Graphical Abstract
194
195
196
197
A second-generation small molecule P2X3 receptor antagonist has been developed. The lead optimization strategy to
address shortcomings of the first-generation preclinical lead compound is described herein. These studies were
directed towards the identification and amelioration of preclinical hepatobiliary findings, reducing potential for drug-
drug interactions, and decreasing the projected human dose of the first-generation lead.
198
199
13