A Dipolar Cycloaddition Reaction to Access 6-Methyl-4,5,6,7-tetrahydro-1H-[1,2,3]triazolo[4,5-c]pyridines Enables the Discovery Synthesis and Preclinical Profiling of a P2X7 Antagonist Clinical Candidate

Article abstract of DOI:10.1021/acs.jmedchem.7b01279

A single pot dipolar cycloaddition reaction/Cope elimination sequence was developed to access novel 1,4,6,7-tetrahydro-5H-[1,2,3]triazolo[4,5-c]pyridine P2X7 antagonists that contain a synthetically challenging chiral center. The structure-activity relati

Full text of DOI:10.1021/acs.jmedchem.7b01279

Subscriber access provided by University of Florida | Smathers Libraries
Article
A dipolar cycloaddition reaction to access 6-methyl-4,5,6,7-tetrahydro-1H-
1,2,3]triazolo[4,5-c]pyridines enables the discovery synthesis
[
and preclinical profiling of a P2X7 antagonist clinical candidate
Christa C Chrovian, Akinola Soyode-Johnson, Alexander A Peterson, Christine F.
Gelin, Xiaohu Deng, Curt A. Dvorak, Nicholas I. Carruthers, Brian Lord, Ian Fraser,
Leah Aluisio, Kevin J Coe, Brian Scott, Tatiana Koudriakova, Freddy Schoetens,
Kia Sepassi, David J. Gallacher, Anindya Bhattacharya, and Michael A Letavic
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01279 • Publication Date (Web): 06 Dec 2017
Downloaded from http://pubs.acs.org on December 7, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A dipolar cycloaddition reaction to access 6ꢀmethylꢀ
4
,5,6,7ꢀtetrahydroꢀ1Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridines
enables the discovery synthesis and preclinical
profiling of a P2X7 antagonist clinical candidate
*,1
1
1
1
Christa C. Chrovian, Akinola Soyode-Johnson, Alexander A. Peterson, Christine F. Gelin,
1
1
1
1
1
Xiaohu Deng, Curt A. Dvorak, Nicholas I. Carruthers, Brian Lord, Ian Fraser,
1
1
1
1
1
Leah Aluisio, Kevin J. Coe, Brian Scott, Tatiana Koudriakova, Freddy Schoetens,
1
2
1
1
Kia Sepassi, David J. Gallacher, Anindya Bhattacharya and Michael A. Letavic
1
Janssen Research & Development, LLC, 3210 Merryfield Row, San Diego, California 92121
Janssen Research & Development, Janssen Pharmaceutica NV, Turnhoutseweg 30, 2340
2
Beerse, Belgium
ABSTRACT: A single pot dipolar cycloaddition reaction / Cope elimination sequence was
developed to access novel 1,4,6,7ꢀtetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridine P2X7 antagonists
that contain a synthetically challenging chiral center. The structureꢀactivity relationships of the
new compounds are described. Two of these compounds, (S)ꢀ(2ꢀfluoroꢀ3ꢀ
(trifluoromethyl)phenyl)(1ꢀ(5ꢀfluoropyrimidinꢀ2ꢀyl)ꢀ6ꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ
[
1,2,3]triazolo[4,5ꢀc]pyridinꢀ5ꢀyl)methanone
(compound
29)
and
(S)ꢀ(3ꢀfluoroꢀ2ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
trifluoromethyl)pyridinꢀ4ꢀyl)(1ꢀ(5ꢀfluoropyrimidinꢀ2ꢀyl)ꢀ6ꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ
1,2,3]triazolo[4,5ꢀc]pyridinꢀ5ꢀyl)methanone (compound 35), were found to have robust P2X7
[
receptor occupancy at low doses in rat with ED values of 0.06 and 0.07 mg/kg, respectively.
50
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound 35 had notable solubility compared to 29 and showed good tolerability in preclinical
species. Compound 35 was chosen as a clinical candidate for advancement into Phase I clinical
trials to assess safety and tolerability in healthy human subjects prior to the initiation of proof of
concept studies for the treatment of mood disorders.
INTRODUCTION:
Proꢀinflammatory cytokines may be a significant contributor to the pathology of psychiatric
malfunctioning. Indeed, elevated levels of inflammatory biomarkers have been identified in
1
individuals diagnosed with bipolar and major depressive disorders. The P2X7 receptor is an
ATPꢀgated ion channel involved in a prominent inflammatory cascade in central nervous system
(CNS) immune cells (microglia) that results in the release of the “master” inflammatory cytokine
2
ILꢀ1β, so called because of its role in triggering an acute phase response. P2X7 enables the
release of ILꢀ1β by converting proꢀILꢀ1β into its mature form. In the CNS this occurs primarily
in microglia and triggers a neuroinflammatory cascade. As part of our ongoing interest in the
chronic effects of prolonged neuroinflammation, and in particular the role of neuroinflammation
in psychiatric conditions, we have been working to advance brainꢀpenetrant selective P2X7
antagonists into clinical proof of concept (PoC) studies in order to study the effect of inhibiting
overall ILꢀ1β release (CNS + peripheral) on individuals suffering from mood disorders. Toward
this goal, we have found that P2X7 antagonists are efficacious in preꢀclinical chronic models of
ACS Paragon Plus Environment

Page 3 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
depression. Data have also been presented suggesting that P2X7 antagonists that can cross the
4
,5
bloodꢀbrain barrier may be the most effective at treating psychiatric conditions.
At least two selective P2X7 antagonists have advanced into PoC studies for the treatment of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
7,8
Crohn’s disease (CD) and/or Rheumatoid Arthritis (RA). These two compounds ꢀꢀ one from
8
6,7
Pfizer and one from AstraZeneca ꢀꢀ were both effective at blocking peripheral ILꢀ1β release in
ex vivo ATPꢀactivated patient plasma samples. Neither compound ultimately met its desired
7,8
6
clinical endpoint in the studies, and their development for either RA and/or CD was
discontinued.
We recently disclosed the structure of JNJꢀ54175446 (1), which has successfully completed
9
initial Phase I clinical studies. Compound 1 is a potent P2X7 antagonist with significant brain
penetration and exhibits doseꢀdependent P2X7 receptor occupancy in the hippocampus of rats
and dogs. It is considered to be a Biopharmaceutical Classification System (BCS) Class II
compound based on its high permeability and low solubility; in aqueous buffer its solubility was
3
9 µM at pH 2 and 21 µM at pH 7, but was found to be readily absorbed in preclinical species,
9
with a bioavailability of ~100% in dog. The predicted human halfꢀlife of 1 was 29 h.
While compound 1 progressed into preclinical development, the project team remained
interested in a backꢀup compound with high solubility as well as differentiated absorption,
distribution, metabolism and elimination (ADME) properties, specifically, one with more rapid
elimination.
1
0
Throughout our work on P2X7 antagonists containing 5,6ꢀfused cores, we have found that a
9
11
12
4
Rꢀmethyl (as in 1 and 2 ) or –phenyl (as in 3) substituent provides improved potency
10,13
compared to an unsubstituted 4ꢀposition,
particularly in rodent species. The added 4Rꢀ
9
,11
9
substituent also provided improved microsomal stability and decreased CYP2C19 inhibition.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The recent report describing compound 1 establishes that a methyl 4Rꢀorientation is the preferred
9
epimer for P2X7 potency of 4ꢀmethylꢀ4,5,6,7ꢀtetrahydroꢀ1Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridines.
1
4
Separately, we have shown that 6ꢀmethyl groups also provide benefit in fused ring cores (4 and
1
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
), and the stereochemistry was determined to be S in the examples for which an assignment
14
was conclusively made, including 4 and 1ꢀ(5,6ꢀdihydroꢀ[1,2,4]triazoloꢀ[4,3ꢀa]pyrazinꢀ7(8H)ꢀ
1
0
yl)methanones . It should be noted that while the absolute configuration of 5 was not
15
determined, the specific rotation value was positive. The 6Sꢀmethyl group had been
underexplored with 4,5,6,7ꢀtetrahydroꢀ1Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridine cores. We thus became
interested in investigating (S)ꢀ6ꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridines.
Chart 1. Fused triazole and pyrimidine core P2X7 antagonists containing a C-methyl or -phenyl
substituent in the piperidine or piperazine ring.
RESULTS AND DISCUSSION:
Synthetic Methodology Development. One challenge in executing this proposal was devising
synthetic methods for the rapid assembly of final compounds. Compelling 1,2,3ꢀtriazole
1
6
syntheses includes click chemistry of an alkyne and an azide. In our case of a fused triazole,
this would be followed by elaborating the 5ꢀposition of the ring with functionality that could,
ACS Paragon Plus Environment

Page 5 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
after a ring closing reaction with a 4ꢀmethylamino group, form the fused triazole system
1
7
(
Scheme 1, eq A). However, this strategy is rarely used with Hetꢀ or arylꢀ azides, and becomes
17ꢀ19
more challenging given the 6ꢀmethyl group we needed to install.
In an alternative method,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the fused triazole can be formed from a diaminopyridine, which is followed by a reduction of the
13,20
pyridine ring (eq B).
Both of these routes require several challenging steps. With the required
absolute configuration, the synthesis becomes additionally arduous. We were interested in
developing a concise route to enable rapid access to the 1ꢀHetꢀ6Sꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ
[
1,2,3]triazolo[4,5ꢀc]pyridines, and thought that a 1,3ꢀdipolar cycloaddition of an azide and a
ketone was worth investigating.
Scheme 1. Synthetic methods for the preparation of 1-Het-1,4,6,7-tetrahydro-5H-[1,2,3]triazolo[4,5-
c]pyridines.
As early as 1967, Huisgen et al. reported 1,3ꢀdipolar cycloadditions with azides and an
2
1
enamide that could form triazoles in a fused ring system. This reaction has been more recently
elaborated using ketones and azides in the presence of an amine organocatalyst. In one example,
Ramachary and coꢀworkers reported the use of activated ketones in an enamine amination /
elimination followed by [3+2]ꢀcycloaddition / hydrolysis in an organocatalytic cascade as a
22
synthetic approach to substitutedꢀ1,2,3ꢀtriazoles (Scheme 2, eq A). Perhaps more relevant to
2
3
our work is the cascade reaction using unactivated ketones reported in 2011 by Belkheira et al.
In this work, carbocyclic ketones were reacted with aryl azides, presumably via enamine
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
formation, in a dipolar cycloaddition reaction (Scheme 2, eq B). We hoped we could extend this
methodology to the formation of 6ꢀmethylꢀ4,5,6,7ꢀtetrahydroꢀ1Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridines
from unsymmetrical ketone heterocycles containing a chiral center (Scheme 2, eq C). We kept in
mind however, that the altered electronics and added steric bulk of our starting ketone (versus the
literature), as well as the need for a specific regiochemical outcome, could make the optimization
challenging.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of fused-1,2,3-triazoles by an enamine amination / elimination followed by [3+2]-
cycloaddition / hydrolysis organocatalytic cascade.
Initial investigations were conducted using phenyl azide and ketone racꢀ6 following the
2
3
protocol reported by Belkheira et al. We found that the reaction did not proceed with 6 using
their optimized conditions of catalytic proline and CH Cl solvent (Table 1, entry 1). We found
2
2
that when pyrrolidine was used instead of proline, the reaction proceeded, albeit in low yields
entry 2). However, the products that were isolated contained a pyrrolidine adduct (7a and 8a),
(
which we supposed was an intermediate that could be readily converted to desired product. It
was not surprising that the yield improved when a full equivalent of pyrrolidine was used (entry
3
). Comparable yields were achieved with ethanol solvent (entry 4), and were higher using
toluene (entry 5).
Table 1. Initial results of a dipolar cycloaddition with 2-methylpiperidone and phenyl azide.
ACS Paragon Plus Environment

Page 7 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
amine,
solvent,
temperature,
time
N
N
N
N
NBoc
NBoc
O
N3
N
N
N
O
N
N
+
O
rac-6
7a
8a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
entry
1
amine
solvent
conditions
isolated yield
no reaction
proline,
catalytic
rt or 80 °C
microwave
DCM
DCM
DCM
EtOH
toluene
pyrrolidine,
catalytic
80 °C
a
2
3
4
5
<5% 7a+8a
microwave
pyrrolidine,
1 equiv
34% 7a+8a
80 °C
microwave
(1:1)
pyrrolidine,
120 °C
29% 7a+8a
1
equiv
pyrrolidine,
equiv
microwave
(1:1)
63% 7a+8a
100 °C
1
(1:1)
a
In these examples, the yields and relative ratios of 7a and 8a were not rigorously determined.
Unfortunately, attempts to facilitate an acid catalyzed elimination of 7a + 8a to form 9a + 10a
were unsuccessful using either catalytic or stoichiometric amounts of pꢀTSA. One thought we
had was to use oxidative conditions in order to set up a Cope elimination. Gratifyingly, after
treating an isolated mixture of 7a + 8a (1:1) with mꢀCPBA, a rapid transformation occurred to
form 9a + 10a in a 1:1 ratio and 83% isolated yield (Scheme 3). In fact, this elimination could
even be conducted in a single pot format, following the [3+2] cycloaddition step. The best
overall yields for the sequence were obtained when we combined the ketone, azide and
pyrrolidine at once with heating. Then, after formation of the intermediates 7 and 8 in roughly 2
hours, m-CPBA was added. Due to this development, our experimental procedure for forming
fused triazoles was now quite simple in that three steps, i) enamine formation, ii) dipolar
cycloaddition and iii) Cope elimination, are conducted in a single pot.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
However, an unresolved problem remained for the formation of phenylꢀsubstituted 6ꢀmethylꢀ
4
,5,6,7ꢀtetrahydroꢀ1Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridines. There was no regioselectivity for the
formation of 7a versus 8a. Since the reaction appeared to have many variables that could
potentially be optimized, we undertook a more detailed investigation of nitrogen protecting
groups, amine bases, solvents and temperatures.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3. Pyrrolidine adduct elimination via oxidation and Cope elimination.
The next reaction optimizations were conducted using 2ꢀazidoꢀ5ꢀfluoropyrimidine (the same
pyrimidine featured in 1) and Nꢀprotected 2ꢀmethylꢀpiperidones (Table 2). A solvent and
temperature screen revealed that the best yields and regioselectivities were obtained at higher
temperatures with toluene or 1,4ꢀdioxane (entries 1ꢀ6). DeanꢀStark conditions had minimal
impact on the regioselectivity and yield (entry 7). It is interesting to note that using phenyl azide
under the best reaction conditions resulted in a significantly lower regioselectivity than when the
fluoropyrimidine azide was used (compare entry 6, Table 2 (88:12) with entry 5, Table 1 (1:1)).
This is a curious and unexpected finding. One possible reason for this could be that the electron
deficiency of the fluoropyrimidine azide favors a thermodynamic product by enabling
reversibility, although this hypothesis has not been tested. The nitrogen protecting groups were
also screened, and carbamate protecting groups worked well but Nꢀmethyl and ꢀbenzyl groups
resulted in no reaction (entries 8ꢀ11). A screen of secondary amines appeared to favor the use of
unsubstituted pyrrolidine (a subset of amines is shown in entries 12ꢀ16). Proline once again did
not promote the reaction (entry 14).
ACS Paragon Plus Environment

Page 9 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. One pot dipolar cycloaddition
fluoropyrimidine.
/ Cope elimination reaction optimization using 2-azido-5-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
amine
Nꢀprotected
piperidinone
a
Entry
solvent
temp
yield, 9b:10b
(1 equiv)
b
1
2
3
4
6
6
6
6
pyrrolidine
pyrrolidine
pyrrolidine
pyrrolidine
toluene
EtOH
100 °C
80 °C
65%, 88:12
47%, 85:15
37%, 82:18
62%, 89:11
MeCN
80 °C
1,4ꢀdioxane
100 °C
0
% of 9b or
5
6
7
6
6
6
pyrrolidine
pyrrolidine
pyrrolidine
TFE
80 °C
c
1
0b
toluene
toluene
rt
19%, 60:40
reflux 4h
DeanꢀStark
60%, 88:12
8
9
pyrrolidine
pyrrolidine
toluene
toluene
100 °C
100 °C
59%
48%, 7.5:1
1
0
1
pyrrolidine
pyrrolidine
toluene
toluene
toluene
toluene
100 °C
100 °C
100 °C
100 °C
no reaction
no reaction
no reaction
no reaction
1
1
1
2
3
6
6
1
1
4
5
6
6
6
toluene
toluene
toluene
100 °C
100 °C
100 °C
no reaction
no reaction
no reaction
16
a
b
c
Reported as isolated yield of the mixture of regioisomers.
The result is the average yield of n=5 reactions.
Byproducts were observed but not characterized.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The scope of the reaction was next investigated using a variety of carboꢀ and heteroꢀ cyclic
ketones, as well as aryl and alkyl azides. The reaction’s utility is enhanced by the fact that both
aromatic and alkyl azides were reactive. All of the reactions in Table 3 were conducted using
commercially available azides. The highest yields were with phenyl azide and carbocyclic or Nꢀ
heterocyclic ketones (11ꢀ13), an exception being 5ꢀmembered ring ketones which did not react to
form product (14 and 17). tertꢀButyl 4ꢀoxoazepaneꢀ1ꢀcarboxylate reacted to form the two
possible regioisomeric products with roughly equal preference (1:1 for 13a:13b and 1:2 for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3a:23b). The oxygenꢀ and sulfurꢀcontaining ketones reacted in low to moderate yields (15ꢀ16
and 24ꢀ25). Products 18ꢀ20 are examples from reactions using different aromatic azides, and low
to moderate yields were obtained in these cases as well. It should be noted however, that despite
low to moderate yields obtained, the ability to form a fused triazole ring system using a variety
of functionalized azides is an excellent option for medicinal chemists at the hitꢀtoꢀlead or lead
optimization stage of a program, where facile access to relevant substrates is of high value in
order to quickly evaluate structureꢀactivity relationships (SAR).
Table 3. Screen of cyclic ketones in the dipolar cycloaddition / Cope elimination using aryl or alkyl azides.
a
a
pdt
ketone
azide pdt structure Yield
pdt
19
ketone
azide pdt structure Yield
b
11
N ꢀPh
66%
59%
41%
3
12
N ꢀPh
3
20
13%
1
3a
63%,
1:1
N ꢀPh
21
22
N -iPr
55%
38%
3
3
1
3b
1
4
N ꢀPh no reaction
3
ꢀꢀ
N -iPr
3
ACS Paragon Plus Environment

Page 11 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
23a
43%,
1:2
1
5
N ꢀPh
25%
N -iPr
3
3
2
3b
N
N
S
N
a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
16
N ꢀPh
3
8%
24
N -iPr
3
ND
Ph
b,c
1
1
7
8
N ꢀPh no reaction
3
ꢀꢀ
25
N -iPr
3
18%
N ꢀBn
3
38%
a
b
c
Reported as isolated yield.
The yield is the average range of several reactions.
The product mass was observed not isolated.
SAR
of
(1)ꢀ(6Sꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridinꢀ5ꢀ
yl)methanones. At this point we were eager to apply the [3+2] cycloaddition / Cope elimination
method to the synthesis of the desired 6ꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀ
c]pyridine core in order to evaluate P2X7 SAR. After formation of the core, a Bocꢀdeprotection
followed by amide bond formation gave the desired final compounds. Compounds 26ꢀ29 and 32ꢀ
4
0 in Tables 4 and 5 were synthesized using this method.
We began with 2,3ꢀsubstituted benzamide analogs in order to evaluate SAR of different Nꢀaryl
triazole substituents. This choice was made due to previous work that established 2,3ꢀhalo,ꢀalkyl,
and ꢀhaloalkyl groups to be preferred substituents on the benzamide moiety of 5,6ꢀfused core
9
ꢀ15,24
P2X7 antagonists.
We focused almost exclusively on the 6Sꢀmethyl configuration based on
earlier findings as well with related cores; all of the 6Rꢀmethyl epimers prepared and tested on
1
0,14ꢀ15
6
,6ꢀ and 5,6ꢀfused cores had poor P2X7 potency.
Shown in Tables 4 and 5 are IC values of the final compounds as measured in a calcium
5
0
mobilization assay, as well as their in vitro CYP inhibition, solubility and liver microsomal
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
turnover extraction ratios (ER). Compounds 26ꢀ34 contain a 2ꢀchloro, ꢀfluoro or ꢀmethyl
substituted 3ꢀtrifluoromethyl phenyl amide and the fused triazole is Nꢀlinked to a pyridine,
pyrazine, pyrimidine or pyrazole (substituted or unsubstituted) group. They all demonstrated
good human P2X7 potency with IC values of 1.7ꢀ19 nM. In rat, IC values were also < 20 nM,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
50
with the exception of 33 (rP2X7 IC = 404 nM). In this set of compounds, CYP2C19 and 2C9
5
0
inhibition was prevalent, although 29 showed no significant inhibition of any of the six CYP
isoforms tested.
Many of the compounds in Table 4 have low solubilities. In addition, compounds 26 and 30ꢀ34
had poor stability in vitro human and rat liver microsomes. The compounds that demonstrated
higher stability in liver microsomes have a 5ꢀfluoropyrimidineꢀsubstituted triazole (27ꢀ29). The
electron deficiency of this moiety might explain its lower propensity for oxidative metabolism as
compared to the other heteroaromatic groups.
Table 4. In vitro data for 1-(6S-methyl-1,4,6,7-tetrahydro-5H-[1,2,3]triazolo[4,5-c]pyridin-5-yl)methanone P2X7
2
antagonists with 2-R ,3-(trifluoromethyl)phenyl substituents.
1
2
hP2X7 IC50 rP2X7 IC50
CYP IC
50
solub pH
2/7 (µM)
extraction
ratio h/r
cpd
R
R
a
a
c
b
(
nM)
(nM)
(µM)
2
C19 = 0.3
C9 = 1.6
2
6
Cl
Cl
8.0±1.1
1.9±1.1
10±3.7
10 / 18
16 / 9
0.94 / 0.85
0.50 / 0.32
2
27
4.1±0.65
2C19 = 3.4
2C19 = 4.3
>10
2
8
CH3
F
14±2.4
4.8±1.1
13±5.5
5.9±1.0
15 / 19
22 / 25
0.35 / 0.54
0.47 / 0.37
29
ACS Paragon Plus Environment

Page 13 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
C19 = 0.3
C9=1.2
2C19 = 0.5
C9 = 5.5
30
Cl
Cl
1.7±0.41
7.3±2.2
1.8±0.42
2.1±0.60
74 / 29
155 / 5
0.95 / 0.90
0.85 / 0.77
2
3
1
2
2
C19 = 2.5
C9 = 2.5
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
32
33
Cl
1.8±0.37
1.9±0.54
123 / 112
>0.95 / >0.92
2
Cl
Cl
19±12
404±157
10±3.0
2C19 = 2.6
261 / 361
nd
0.83 / 0.89
2
C19 = 0.4
C9 = 3.7
3
4
2.0±0.53
>0.95 / >0.92
2
a
IC values were determined by a calcium mobilization assay in 1321N1 cells expressing the
5
0
recombinant human or rat P2X7 channel and values reported are the mean of three experiments,
unless otherwise stated. H is human, r is rat. All data are reported as the mean value of 3ꢀ4 assays
run in triplicate; IC ’s ± SD are reported.
5
0
b
CYP IC values were obtained in human liver microsomes using a “cocktail” of CYPꢀ
selective probes for the six isoforms: 1A2, 2C19, 2C8, 2C9, 2D6, 3A4.
5
0
c
Stability in human and rat liver microsomes. Data is reported as extraction ratios where
predicted hepatic clearance is divided by the species specific hepatic blood flow.
Compound 29 demonstrated good potency and ADME properties worthy of progressing. Still,
we continued our SAR exploration by introducing isonicotinamides in place of the benzamide
with the thought that we could effect physicochemical properties of the substrates, solubility in
particular, and alter their absorption, distribution, metabolism and excretion (ADME) properties
(Table 5). Compounds with chloroꢀ or fluoroꢀ substituents in the 3ꢀposition of the
isonicotinamide were found to be potent both in human and in rat (35ꢀ36,39ꢀ40), whereas the 3ꢀ
methyl analog was potent in human but not in rat (37). A trifluoromethylꢀsubstituted pyrazole
(38) had poor P2X7 potency (human IC = 4034 nM; rat IC = 2889 nM).
5
0
50
The compounds containing isonicotinamides showed excellent aqueous solubility as compared
to the benzamide analogs (35 vs 29, 37 vs 28, 36 vs 27). This improved solubility is likely aided
by a decrease in measured partition coefficient at pH 7.4 (Log D ) for 35 (1.65) versus 29,
7
.4
(
2.23). The isonicotinic nitrogen is not basic, however, due to the strongly deficient amide. In an
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ionization constant (K ) assay with 35, no dissociation was measured in the pH range of 2 to 12.
a
It also may be that a disruption to crystal packing contributes to the superior solubility of 35.
Nevertheless, compound 35 also has superior solubility as compared to lead clinical candidate 1,
which is an advantage for both formulation work as well as intestinal absorption in preclinical
and clinical studies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The isonicotinamides also helped to improve metabolic stability as well as reduce CYP2C19
inhibition. Introducing a pyrazole, pyridine or pyrazine to the 1ꢀHet position improved
compound solubility (30ꢀ33), but these compounds are unstable in liver microsomes and inhibit
CYP2C9 and CYP2C19.
Overall compounds 29 and 35 had the most promising in vitro profile and the appeal of 35 was
augmented by its excellent solubility. The two compounds were advanced to in vivo evaluation.
Table 5. In vitro data for 1-(6S-methyl-1,4,6,7-tetrahydro-5H-[1,2,3]triazolo[4,5-c]pyridin-5-yl)methanone P2X7
antagonists with nitrogen-containing aromatic amides.
solubility
CYP IC
hP2X7 IC
rP2X7 IC
50
extraction
ratio h/r
5
0
50
cpd
structure
a
a
pH 2/7
µM)
c
b
(
nM)
(nM)
(µM)
(
3
5
10±3.7
12±6.8
26±14
15±7.4
>10
>400
>400
<0.3/ <0.2
<0.3 / <0.2
<0.3 / <0.2
nd
2
C19 =
.1
36
9.4±0.98
6199±837
2889±957
5
3
3
7
8
>10
nd
51 / >400
4034±740
nd
ACS Paragon Plus Environment

Page 15 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
9
0
11±2.6
14±2.9
12±0.95
40±2.8
nd
nd
nd
nd
0.85 / 0.72
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
nd
a
IC values were determined by a calcium mobilization assay in 1321N1 cells expressing the
5
0
recombinant human or rat P2X7 channel and values reported are the mean of three experiments,
unless otherwise stated. Hu is human, r is rat. All data are reported as the mean value of 3ꢀ4
assays run in triplicate; IC ’s ± SD are reported.
5
0
b
CYP IC values were obtained in human liver microsomes using a “cocktail” of CYPꢀ
5
0
selective probes for the six isoforms: 1A2, 2C19, 2C8, 2C9, 2D6, 3A4.
c
Stability in human and rat liver microsomes. Data is reported as extraction ratios where
predicted hepatic clearance is divided by the species specific hepatic blood flow.
Ex Vivo Receptor Occupancy Studies in Rat. Compounds 29 and 35 were next assessed in a
receptor occupancy duration of action study measured by ex vivo autoradiography to determine
P2X7 target engagement in the rat brain. Six animals were dosed 10 mg/kg p.o., 29 as a
suspension and 35 as a solution, and receptor occupancy was measured at three time points. The
brain versus plasma ratios of 29 and 35 were found to be approximately 1:1 or greater at the
three time points, indicating high brain distribution for both compounds (Table 6). The 10 mg/kg
dose resulted in ≥90% occupancy for both compounds at the 6ꢀhour time point. The two
compounds were next evaluated in a doseꢀresponse receptor occupancy study where they
performed especially well, with ED P2X7 occupancy values of 0.06 mg/kg and 0.07 mg/kg for
5
0
2
9 and 35, respectively, in the rat brain (Figure 1). These ED s corresponded to brain/plasma
50
EC values of 19/16 and 12/15 ng/mL for 29 and 35, respectively. (The concentrations are the
50
sum of plasma protein bound and unbound fractions).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The data for compounds 29 and 35 compared favorably with previous P2X7 antagonists, as
measured by the efficacious doses in rat (Table 7). The compounds in Table 7 represent six
different structural classes of compounds, and their ED50 values were all generated using the
9
,11ꢀ14
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
same radiotracer in the receptor occupancy dose response assay.
Of this set, a 6,6ꢀfused core
compound (5) has the lowest EC values. However, its preclinical pharmacokinetic (PK) data
50
1
5
from 3 species translated to a relatively high human dose prediction. Compound 4 also had
promising EC50 receptor occupancy data, however its rat ED50 was 10ꢀfold higher than for
compounds 29 or 35. Ameriks et al. suggests that brain uptake of 4 is perhaps limited by efflux
1
4
mechanisms, based on its unbound brainꢀtoꢀplasma partition ratio (Kp u,u) of 0.51. Compounds
2 and 3 also demonstrated inferior ED s and EC s compared to 29 and 35. Compound 1 has
5
0
50
9
advanced as the lead. We continued to profile 29 and 35 as potential backꢀup compounds.
Table 6. Plasma and brain levels, and ex vivo P2X7 occupancies of 29 and 35 in rat hippocampus at three time
a
points after a 10 mg/kg oral dose.
P2X7 %
occup.
plasma levels in ng/mL
2.0 h
brain levels in ng/mL
2.0 h
0
.5 h
6.0 h
0.5 h
6.0 h
6.0 h
2
3
9
5
854.9 ± 46
3194 ± 275
1882 ± 103
3430 ± 195
1046 ± 104
2437 ± 361
1056 ± 131
3045 ± 292
2388 ± 597
2911 ± 32
1620 ± 480
1993 ± 212
95 ± 1
96 ± 1
a Mean ± SEM (n= 2)
a
b
Figure 1. Dose dependency of P2X7 occupancy of 29 and 35 as measured 2h after oral
c
administration, as measured in rat brain sections using ex vivo receptor autoradiography.
ACS Paragon Plus Environment

Page 17 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
b
c
0
.5% HPMC suspension
0% HPꢀβꢀcyclodextrin solution
Mean ± SEM (n= 3)
2
Table 7. Comparison of ED50 values at 2h of previously reported P2X7 antagonists with 29 and 35 in ex vivo
receptor occupancy studies in rat.
EC50
plasma
ED50
mg)
EC brain
50
cpd
(
(ng/mL)
(
ng/mL)
29
0.06
0.07
2.3
19
16
3
5
12
362
66
15
125
62
1
1
4
2
2
4
3
5
1
1
0.8
4.3
408
5
748
8
15
0.3
9
1
0.5
108
105
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scale Up Syntheses and Characterization. Given the excellent results in the target
engagement study, compounds 29 and 35 were scaled up for further studies. The synthesis of the
(S)ꢀ1ꢀ(5ꢀfluoropyrimidinꢀ2ꢀyl)ꢀ6ꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridine
core was conducted using our optimized dipolar cycloaddition reaction. 2ꢀAzidoꢀ5ꢀ
fluoropyrimidine was prepared in two steps from 2ꢀchloroꢀ5ꢀfluoropyrimidine and then
condensed with tertꢀbutyl (S)ꢀ2ꢀmethylꢀ4ꢀoxopiperidineꢀ1ꢀcarboxylate (commercially available
in 99.7% enantiopurity) (Scheme 4). The yields of the cycloaddition / Cope elimination sequence
on scales from 1 to 100 g ranged from 58 to 71%, and the ratio of regioisomers ranged from
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
8:12 to 91:9 in favor of the 6ꢀmethyl regioisomer. The purification of 29 was a simple wash
with hot methanol to give crystalline 29 in 100% purity. The absolute configuration of the
25
molecule was established by anomalous dispersion using the Parson’s method (Figure 2).
Scheme 4. Route used for both medicinal chemistry and scaleꢀup syntheses of 29.
Reagents and Conditions: (a) pyrrolidine, toluene, 110 °C, 4 h; m-CPBA, NaHCO , rt, 10 min
3
(
58ꢀ71%, 9b:10b = 9:1) (b) HCl in dioxane, CH Cl , rt, 16 h (41:42 = 9:1) (c) 2ꢀfluoroꢀ3ꢀ
2 2
(trifluoromethyl)benzoyl chloride, NEt , CH Cl , rt, 1 h (85% over 2 steps) (d) stir in MeOH, 60
3
2
2
°C, 24 h (55%, >99% ee).
Figure 2. Single crystal structure and stereochemical assignment of 29.
ACS Paragon Plus Environment

Page 19 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The crystallization protocol utilized to obtain pure 29 was not successful when attempted with
compound 35, because 35 is highly soluble in organic solvents and did not crystallize under a
variety of conditions. In fact, in subsequent studies, identifying a stable polymorph of 35 has
26
been unachieveable. In contrast, intermediates 26 and 27 were found to be crystalline and could
be separated, but due to impure or low recoveries, a slightly different route was used for the
largest discovery synthesis. In this sequence, a 100 g cycloaddition / Cope elimination reaction
was conducted using intermediate 44, which has the final amide in place for the cycloaddition
(
Scheme 5). The reaction occurred in 69% yield, producing a 10:1 mixture of regioisomers 35
and 30, and was followed by SFC (supercritical fluid chromatography) separation to obtain 35 in
9.5% purity. The overall yield of 35 from commercially available starting materials was 40%,
9
and this material was used in a variety of preclinical studies.
Scheme 5. Scaleꢀup route for the synthesis of 35.
O
F
S
CF3
N
N
O
F
O
F
O
F
N
N
CF3
a
b
CF3
N
CF3
c
d
HO
N
Cl
3
5
+
35
N
N
N
F
N
O
N
3
0
4
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Reagents and Conditions: (a) oxalyl chloride, DMF CH Cl rt, 16h (b) (S)ꢀ2ꢀmethylꢀ4ꢀ
,
2
2,
oxopiperidine, NEt , CH Cl , rt, 1 h (77% over 2 steps) (c) 2ꢀazidoꢀ5ꢀfluoropyrimidine,
3
2
2
pyrrolidine, p-TSA, toluene, DeanꢀStark, 110 °C, 4 h; m-CPBA, NaHCO , CH Cl rt, 2 h (69%,
3
2
2,
3
5:30 = 10:1) (d) SFC separation, 75%.
In Vitro Studies. The key in vitro data set used to advance compounds 29 and 35 is shown in
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 8. K values were measured in the 1321N1 cell line overexpressing human and rat P2X7.
i
Compounds 29 and 35 were both quite stable in human and rat liver microsomes, demonstrated
insignificant inhibition of nine CYP isoforms (IC s ≥ 19 µM), as determined by measuring the
5
0
effect of 29 and 35 on metabolite formation by probe substrates, and had no hERG channel
inhibition up to 3 µM as measured in an automated patchꢀclamp assay. Compounds 29 and 35
were not found to be substrates for efflux transporters such as Pꢀglycoprotein (Pꢀgp) in the Cacoꢀ
2
human intestinal cell line with measured efflux ratios (BA/AB) of < 2. Neither compound
exhibited CYP3A4 timeꢀdependent inhibition (TDI) in human liver microsomes, and neither
induced CYP1A1/2, although 29 showed moderate activation of human PXR (66% of rifampicin
control at 10 µM) and thus has the potential to induce CYP3A4. No significant activity of 29 or
3
5 was measured in a panel of related P2X receptors (P2X1, P2X2, P2X3, P2X2/3, and P2X4
FLIPR assays, tested up to 10 ꢁM), and no significant binding to 50 ion channels, receptors, and
transporters was detected in commercial in vitro panel (EurofinsꢀCEREP,
a
http://www.eurofins.com, tested at 10 ꢁM). Compound 35 demonstrated moderate/low plasma
protein binding in rat, dog and human (Table 8) as determined by an equilibrium dialysis
method, and is considered to be a BCS Class I compound based on its high permeability and high
solubility.
Table 8. In vitro ADME and selectivity data for 29 and 35.
2
9
35
ACS Paragon Plus Environment

Page 21 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
rP2X7 Ki
hP2X7 Ki
1.0±0.34 nM
6.5±1.6 nM
2.9±0.42 nM
7.1±1.9 nM
a
h / rat / dog /
mky
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0.47 / 0.37 /
<0.30 / <0.20 /
<0.26 / <0.19
<
0.26 / 0.66
microsomal
b
stability
CYP isoform
IC (if IC
2
C19
2
C19
50
50
c
(30 µM)
(19 µM)
<
50 µM)
3
A4 timeꢀ
dependent
inhibition
no
no
CYP induction
3A4
73/49
>3 µM
no
permeability
45/35
>3 µM
d
(
Cacoꢀ2)
hERG QPatch™
IC50
h / r plasma
proteins (%
75 / 76
47 / 56
e
bound)
a
K values are reported as the mean of 4ꢀ7 experiments in triplicate. H is human, r is rat, K ±
SEM is reported.
i
i
b
Stability in liver microsomes. Data is reported as extraction ratios where predicted hepatic
clearance is divided by the species specific hepatic blood flow.
c
Values were determined for CYPs 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1 and 3A4 by
incubation with CYP probe substrates as described in the SI.
d
ꢀ6
ꢀ6
Papp is reported as BꢀA(x10 ) cm/s / AꢀB(x10 ) cm/s.
e
Plasma protein binding 0.2 µM for 35 and 0.5 µM for 29 determined by equilibrium dialysis
method as previously described.
9
Preclinical Pharmacokinetics and Human Dose Predictions. The PK of compounds 29 and
3
5 were investigated in preclinical species. The intravenous PK in rat, dog and monkey were
characterized by low or moderate clearance (CL) and moderate volume of distribution (V ). The
ss
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
high oral bioavailability (%F) measured across species was suggestive of complete absorption
(Table 9).
The human CL for 29 was predicted using 3 species allometry of unbound CL corrected for in
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
vitro CL. The human V was predicted to be 1.9 L/kg, after correction for plasma protein
ss
binding. Based on these CL and V values, a halfꢀlife estimate in humans was determined to be 6
ss
hours. For a dose of 76 mg twice daily, plasma trough levels at steady state would be expected to
be above the EC of 98 ng/mL over 24 h (Figure 3). The plasma concentration over time profile
5
0
(Figure 3) was simulated using the predicted human PK parameters shown in Table 9.
Table 9. Pharmacokinetic data of 29 and 35 in three preclinical species.
human
rat
dog
monkey
predicted
compound and dose (i.v. mg/kg / p.o. mg/kg)
1
2
1
2
1
2
2
9
35
29
35
29
35
parameter
F (%)
2
9
35
0.5 / 2.5
1 / 5
0.125 /
0.625
0.25 /
1.25
0.125 /
0.625
0.5 / 2.5
90
81
1.7
123
2.6
3.8
8.8
305
98
1.4
30
1.7
26
97
1.3
75
1.9
4.8
6
100
1.1
0.8
16
V (L/kg)
2
.5
12
.5
ss
CL (mL/min/kg)
t1/2 (hr) i.v.
3.7
0.8
2.6
2
5.6
21.2
1123
0.9
36
6.0
Cmax (ng/mL)
AUC 24h
541
1747
1864
368
105
3190
17549
3039
23262
122
16262
5616
1786
(
ng/mLꢂh)
1
Vehicle is 50% PEG400 / 50% water
Vehicle is 20% HPꢀβꢀCD
2
Figure 3. Simulated human plasma concentration profile of 29 at a dose of 76 mg twice
daily over five days of administration.
ACS Paragon Plus Environment

Page 23 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
7
The human predictions for 35 were conducted using 3 species allometry of unbound CL and
Vss. The human CL and V were predicted to be 0.8 mL/min/kg and 1.1 L/kg, respectively,
ss
resulting in a predicted human t1/2 of 16 hours. The plasma concentration over time profile
(Figure 4) was simulated using the predicted human PK parameters shown in Table 8.
Figure 4. Simulated human plasma concentration profile of 35 at a dose of 6 mg once daily
over six days of administration.
For a dose of 6 mg once daily, plasma trough levels at steady state are expected to be above the
predicted EC50 of 45 ng/mL over 24 h, when correcting for the human vs. rat K value
i
differences. The estimated Cmax at steady state and AUC in humans for a 6 mg q.d. dose are 105
ng/mL and 1786 ng/mLꢂh, respectively. Given that greater blockade of ILꢀ1β may be important
for efficacy, the human dose required to maintain steady state plasma trough concentrations
above the EC80 (225 ng/mL) was also calculated and estimated to be 30 mg q.d., with a
corresponding Cmax of 525 ng/mL and AUC of 8,930 ng/mLꢂh.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Based on the predicted human doses, 35 at this point appeared be the compound of choice to
progress further. However, it was decided to assess both 29 and 35 in preclinical toleration in the
event that 29 turned out to have a superior safety profile.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Preclinical Toleration Studies. Compounds 29 and 35 were taken into rat and dog toleration
studies in order to assess their potential toxicity after oral administration, first in single dose
escalation, followed by 4 or 5ꢀday repeated dosing toleration studies. The doses selected for 29 in
the 4ꢀ day study were 62.5, 125 and 250 mg/kg in rats and 35, 100, and 300 mg/kg in dogs. The
9
doses for 29 were chosen for comparison purposes to the lead (1), with the top doses chosen
because of apparent saturation of absorption observed at higher doses in the single dose study.
The doses selected for 35 in the repeated dose studies were 30, 125, 250, and 500 mg/kg/day in
rats and 10, 50 and 250 mg/kg/day in dogs. The doses in dog were chosen based on exposure
levels in the single dose escalation; the Cmax and AUC were approximately the same for 250
mg/kg and 1000 mg/kg doses, thus the highest dose was not tested in the repeated dosing studies.
Mortality, body weight, clinical observations, clinical pathology, gross necropsy and microscopic
examination of selected tissues, toxicokinetic parameters and gene expression analysis of liver
samples were evaluated.
Both compounds were well tolerated in rats and dogs. The only compoundꢀrelated findings for
29 in rats were a mild increase in cholesterol (125 and 250 mg/kg) and a slight increase in total
protein and albumin (250 mg/kg). In dog, there were no clinical pathology, macroscopic, or
microscopic abnormalities for 29 at any dose. In rats dosed for 4 days at 250 mg/kg/day, the
mean Day 4 Cmax was 9,260 ng/mL and AUC0ꢀ24h was 108,000 ngꢂh/mL. In dogs, day 5 Cmax at
the 300 mg/kg/day dose was 5,330 ng/mL and AUC0ꢀ24h was 78,350 ngꢂh/mL.
ACS Paragon Plus Environment

Page 25 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 35 exhibited higher exposures overall than 29, increasing with dose up to 500
mg/kg in rats and 250 mg/kg in dogs. In rats, doses up to and including 125 mg/kg/day of 35
were well tolerated. Higher doses (250 and 500 mg/kg/day) were associated with adverse effects,
mainly on body weight, weight gain, body temperature and neurobehavioral integrity. In dogs
dosed for 5 days, there were no adverse changes at 10 mg/kg/day. Doses of 50 and 250
mg/kg/day resulted in body weight loss and lower food consumption and were associated with
emesis and salivation among other clinical observations. Serum chemistry parameters (ALT,
serum triglycerides and bilirubin) were also affected. Histopathological examination identified
the lymphoid tissue (lymphoid depletion of the spleen and/or Peyer’s patches) as target of
toxicity in the high dose animals. Although histological changes were absent at the liver, this
organ was nevertheless regarded as a potential target organ of toxicity because of the changes in
serum chemistry. In rats dosed for 5 days at 125 mg/kg/day, the mean day 5 Cmax was 26,000
ng/mL and AUC0ꢀ24h was 291,000 ngꢂh/mL. In dogs, the Cmax measured on day 4 at the 10
mg/kg/day dose was 10,680 ng/mL and AUC0ꢀ24h was 167,000 ngꢂh/mL.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The maximum tolerated doses of 29 and 35 were determined in rat and dog, and used for the
calculation of the predicted therapeutic index (Table 10).
Our initial requirements for advancing a second P2X7 compound into the clinic where: 1) a
highly soluble compound and 2) a compound with predicted PK human parameters consistent
with a predicted t1/2 <30h. Compound 35 meets both of these criteria. In addition, the safety
margins for 35 were calculated to be 5ꢀ to 10ꢀfold higher than 29. For these reasons, 35 was
chosen for further detailed evaluation in cardiovascular and pharmacodynamic models.
Table 10. Tolerated exposures in animal studies and predicted human therapeutic levels for 29 and 35 used to
calculate therapeutic index.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
predicted
safety margins
Cmax/ AUC
human predicted
concentrations
c
d
rat NOAEL
dog NOAEL
mg
dose
a
b
mg
dose
daily
dose
rat/
dog/
Cmax
AUC
Cmax
AUC
Cmax AUC
368 5616
human
human
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
5x
1
52
Cmax
/
14x Cmax/
2
3
9
5
250
125
9260
108000
291000
300
10
5330
78350
mg
19x
14x AUC
AUC
2
Cmax
47x
1
00x
Cmax
3x AUC
/
26000
10680
167000 6 mg 105 1786
/
163x
AUC
9
a
Units for Cmax are ng/mL
b
c
d
Units for AUC are ng(h)/mL
Values shown are the average Cmax and AUC for the male and female dog
Human dose is predicted to translate in a 50% central P2X7 inhibition in humans
Compound 35 was taken into cardiovascular safety studies in the anesthetized guinea pig and
dog. No notable electrophysiological effects were found up to the highest doses tested
(associated with plasma levels of 15,500 and 12,150 ng/mL). Plasma concentrations of 2820 and
4
095 ng/mL of 35 and higher in dog and guinea pig, respectively, showed increases in blood
pressures. The NOAEL were 1280 and 1850 ng/mL in dog and guinea pig, respectively.
Pharmacodynamic Models. The effect of 35 on brain ILꢀ1β release was investigated in an in
vivo microdialysis assay as previously described in freely moving rats challenged with the P2X7
28
agonist BzꢀATP. Rats were dosed orally with either vehicle or 10 mg/kg of 35 four hours prior
to infusion of BzꢀATP via a microdialysis probe. Compared to vehicle, 35 significantly inhibited
BzꢀATP induced ILꢀ1β release in the rat hippocampus at the 5 and 6 h time points (Figure 5).
Figure 5. In vivo microdialysis assay to measure brain ILꢀ1β levels in freely moving rats
after oral administration of 10 mg/kg 35. ILꢀ1β release was triggered by 2'(3')ꢀOꢀ(4ꢀ
Benzoylbenzoyl)adenosineꢀ5'ꢀtriphosphate (BzꢀATP) infusion (100 mM) in rat
hippocampus. Solid line indicates time of BzꢀATP infusion (between the 4 and 6 h time
ACS Paragon Plus Environment

Page 27 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
points). ILꢀ1β concentration are given as mean values ± SEM (n= 7); P < 0.05 Robust tꢀtest
at 5 and 6 hour time points.
350
300
250
200
150
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
100
50
0
0
2
4
6
8
10
Time Post-dose (hours)
Vehicle PO (n=7)
Compound 20 10mg/kg PO (n=7)
BZꢀATP 100mM (n=14)
5
Compound 35 was also evaluated in models of depression and pain. In a depression model of
chronic mild unpredictable stress in rats, the animals’ sucrose intake was lowered after being
subjected to a chronic stress regimen for 2 weeks. Compound 35 was found to reverse this
5
deficit. The details of this model as well as the full pharmacology of compound 35 will be
published in due course.
CONCLUSION:
Compound 35, JNJꢀ55308942, has been chosen as the second P2X7 antagonist from Janssen to
enter clinical trials, as a backꢀup to compound 1. The compound was designed to have high
solubility in aqueous media. This physicochemical property was achieved by incorporating an
isonicotinamide moiety on a 1ꢀ(6Sꢀmethylꢀ1,4,6,7ꢀtetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridinꢀ5ꢀ
yl)methanone core. Synthetically, the preparation of the 1ꢀ((5ꢀfluoropyrimidinꢀ2ꢀyl)ꢀ1,4,6,7ꢀ
tetrahydroꢀ5Hꢀ[1,2,3]triazolo[4,5ꢀc]pyridinꢀ5ꢀyl)methanone core featuring the 6S- methyl group
was foreseen to be challenging, as existing methods would have made the synthesis lengthy and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cumbersome. As such, a novel single pot dipolar cycloaddition / Cope elimination sequence was
developed, which constructed the core with a heteroarylꢀ substituted triazole in place in a single
step. This novel method is expected to have utility in the rapid construction of a broader class of
fused triazole systems.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Besides having high aqueous solubility, compound 35 showed an increased plasma free
fraction compared to closely related fused 1,2,3ꢀtriazole core compounds. This is also thought to
be due to the isonicotinamide moiety. Compound 35 showed excellent P2X7 receptor occupancy
in the hippocampus of rats, demonstrating a low ED of 0.07 mg/kg and unbound plasma EC
5
0
50
of 12 ng/mL. In addition, 35 suppressed brain ILꢀ1β release in vivo in freely moving rats
challenged with the P2X7 agonist BzꢀATP. The compound also demonstrated good tolerability
margins in preclinical species, as well as an acceptable cardiovascular safety profile in vivo. FIH
enabling safety studies are currently underway for compound 35 in anticipation of Phase I
clinical trials to determine the safety and tolerability of 35 in healthy human subjects.
EXPERIMENTAL SECTION:
Solvents and reagents were used as supplied by the manufacturer. Concentrated refers to
concentrated using a rotary evaporator under reduced pressure.
Unless specified otherwise, normalꢀphase silica gel column chromatography was performed on
silica gel (SiO ) using prepackaged cartridges and the indicated solvents. Preparative reverseꢀ
2
phase high performance liquid chromatography (HPLC) was performed on an Agilent HPLC
with a Waters XBridge C18 column or Xterra Prep RP column (5 µm, 30x100 mm or 50x150
1
8
mm) and a gradient of 5ꢀ99% acetonitrile/water (20 mM NH OH) over 12 to 18 min and a flow
4
rate of 30 or 80 ml/min, unless otherwise indicated. Preparative SFC was performed either on a
JASCO preparative SFC system, an APS 1010 system from Berger instruments, or an SFCꢀ
ACS Paragon Plus Environment

Page 29 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
PICLABꢀPREP 200 (PIC SOLUTION, Avignon, France). The separations were conducted
between 100ꢀ150 bar with a flow rate ranging from 40ꢀ60 mL/min. The columns used were
heated to 35ꢀ40 °C.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Mass spectra were obtained on an Agilent series 1100 MSD using electrospray ionization (ESI)
in positive mode. The calculated (calcd) mass corresponds to the exact mass. HRMS was
obtained on a Bruker microTOF, using ESI in positive mode with a scan range of 200ꢀ700 amu.
Sodium formate solution (Fluka 97574ꢀ100mL) was used to calibrate the TOF and as an internal
standard. High accuracy MS was run on an Agilent 5975 Series GCMS; EI source. Cerno
Biosciences Mass works software was used to process the data. PFTBA was used as an internal
calibrant. Nuclear magnetic resonance (NMR) spectra were obtained on Bruker model DRX
spectrometers. If rotamers were present, VT NMR was used to determine which peaks coalesce
1
at 120 °C in dꢀDMSO. The H spectra are reported as a mixture of all rotamers at rt in dꢀCDCl
3
1
3
or dꢀDMSO and the C are reported as the major rotamer only at rt in dꢀDMSO.
Melting points (mp) were determined using differential scanning calorimetry (DSC) on a TA
TM
Instruments Q200 series instrument equipped with a standard cell (FC). The DSC was cooled
using a TA Instruments Refrigerated Cooling System 90 (RCS 90). Nitrogen gas purge was set
to 50 mL/min. The DSC method equilibrated at 25°C and used a heating ramp of 10°C/min up
to 300°C. Samples were analyzed in pinhole hermetic pans. Data analysis was performed using
TA Universal Analysis software. Data is reported as the range of Tonset ꢀ Tpeak
.
All compounds tested were of a minimum of 95% purity as determined by HPLC. The HPLC
method used for purity determinations is as follows: Analytical LCMS was obtained on an
Agilent 1100/1200 series LCMS with ESI in positive mode and a scan range of 100ꢀ700 amu.
Samples were run on a Halo C18 column (2.7um, 50 x 3 mm), with a mobile phase of 5ꢀ99%
ACS Paragon Plus Environment