Novel and potent 3-(2,3-dichlorophenyl)-4-(benzyl)-4H-1,2,4-triazole P2X7 antagonists

Article abstract of DOI:10.1016/j.bmcl.2007.04.075

Structure-activity relationship (SAR) studies were conducted around early tetrazole-based leads 3 and 4. Replacements for the tetrazole core were investigated and the pendant benzyl substitution was reoptimized with a triazole isostere. Triazole-based P2X

Full text of DOI:10.1016/j.bmcl.2007.04.075

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4044–4048
Novel and potent 3-(2,3-dichlorophenyl)-4-(benzyl)-4H-
1,2,4-triazole P2X7 antagonists
William A. Carroll,a,* Douglas M. Kalvin,b Arturo Perez Medrano,a Alan S. Florjancic,a
Ying Wang,a Diana L. Donnelly-Roberts,a Marian T. Namovic,a George Grayson,a
a
a
´
Prisca Honore and Michael F. Jarvis
aAbbott Laboratories, Neuroscience Research, 100 Abbott Park Road, Abbott Park, IL 60064, USA
bAbbott Laboratories, Medicinal Chemistry Technologies, USA
Received 10 April 2007; revised 23 April 2007; accepted 23 April 2007
Available online 27 April 2007
Abstract—Structure–activity relationship (SAR) studies were conducted around early tetrazole-based leads 3 and 4. Replacements
for the tetrazole core were investigated and the pendant benzyl substitution was reoptimized with a triazole isostere. Triazole-based
P2X7 antagonists were identified with similar potency to the lead compound 4 but with improved physiochemical properties. Com-
pound 12 was active in a rat model of neuropathic pain.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The P2X receptors (P2X1–P2X7) comprise a family of li-
gand-gated ion channels that are activated by ATP.1 A
functional ion channel forms through the hetero- or
homomeric assembly of P2X subunits. Each subunit, in
turn, consists of a single polypeptide chain with two mem-
brane spanning regions and both amino and carboxy ter-
imini being intracellular. Of the various P2X receptors,
the P2X7 receptor is unique in that it does not form heter-
omeric complexes with other P2X receptor subunits.2
Brief (millisecond) activation of the P2X7 receptor with
ATP results in a non-selectivecation (Na+, Ca2+, K+) flux,
whereas longer exposure (>30 min) to ATP gives rise to
cytolytic pore formation.3
knockout (KO) mice support a prominent role for P2X7
in inflammation and pain. These KO mice showed a re-
duced incidence and severity of arthritic symptoms com-
pared to wild-type mice.12 Additionally, P2X7 KO mice
were protected from the development of both adjuvant-
induced inflammatory pain and partial nerve ligation in-
duced neuropathic pain.13 Blockade of P2X7 with a small
molecule antagonist, thus, could provide a novel thera-
peutic approach to the treatment of pain as well as neuro-
degenerative and inflammatory disorders.
O
H
N
N
S
N
Cl
O O
N
CH2
O
N
O
Cl
The P2X7 receptor is located primarily on cells of the im-
mune system4 such as macrophages, monocytes, mast
cells, and lymphocytes as well as on glial cells (astrocytes
and microglia) in the central nervous system. The pres-
ence of the P2X7 receptor on neurons, however, has yet
to be definitively established.5 Activation of the P2X7
receptor on cells of immunological origin causes a release
of pro-inflammatory cytokines,4 including IL-1b, a pro-
cess that may play an important role in the development
and progression of inflammatory, neurodegenerative6–9
or chronic pain10,11 conditions in either the peripheral
or central compartments. Recent studies with P2X7
O
S
2
O
N
N
Cl
N
Cl
N
N
N
Cl
Cl
N
N
N
1, KN-62
N
3
4
Structure–activity relationship (SAR) studies of P2X7
selective ligands have appeared in the literature for sev-
eral chemical series.14–19 KN-62 (1), a tyrosine-based
non-competitive P2X7 antagonist, was the first small
molecule inhibitor to be identified.20 Recently, a more
Keywords: P2X7 antagonist; Neuropathic pain.
*
Corresponding author. E-mail: william.a.carroll@abbott.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.04.075

W. A. Carroll et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4044–4048
4045
Table 2.
‘druglike’ series of adamantylbenzamides, exemplified
by 2, have been described that possess potent activity
to competitively block P2X7 receptors.18 As described
in a recent report, we have found that 1-benzyl-5-(2,3-
dichlorophenyl)-tetrazoles are potent P2X7 antagonists
with compound 3 showing activity in a rat model of neu-
ropathic pain.21a The studies described herein extend the
SAR investigations around 3 by examining heterocyclic
replacements for the tetrazole core (see Table 1) and
subsequent reoptimization of the pendant nitrogen sub-
stituent (Tables 2 and 3).
N
N
N
Cl
R
Cl
a
Compound
R
hP2X7 pIC50
5
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
H
2-Me
6.28
7.11
6.72
6.88
6.93
7.31
7.57
6.84
7.23
7.32
7.31
6.04
6.67
5.89
5.05
6.88
6.72
6.48
6.86
2-OMe
2-CF3
2-CHF2
2-SMe
2-SMeb
Scheme 1 illustrates the straightforward 3-step method
utilized to prepare the triazole 5 as well as the examples
2-SO2Meb
2,3-Benzo
2-Me, 5-F
2-CF3,5-F
2-F, 5-CF3
3-OMe
Table 1.
Cl
N
Cl
3-CF3
4-CF3
2,3-DiMeO
2,3-DiCl
2,4-DiCl
2,5-DiCl
a
Compound
hP2X7 pIC50
N
Values are means of at least three experiments.
a Compounds tested at the recombinant hP2X7 receptor as described.21
b 2-Cl,3-CF3 substitution on left hand phenyl rather than 2,3-
dichlorophenyl.
N
N
N
4
5
7.40
6.28
N
N
N
Table 3.
N
N
N
Cl
N
R
Cl
N
N
6
7
6.52
6.22
6.37
6.68
5.74
N
a
Compound
hP2X7 pIC50
N
N
30
5.98
N
N
31
32
33
4.88
7.72
7.30
N
N
8
N
N
N
9
Values are means of at least three experiments.
a Compounds tested at the recombinant hP2X7 receptor as described.21
N
N
10
in Tables 2 and 3. Conversion of methyl 2,3-dic-
hlorobenzoate to the oxadiazole intermediate in two
steps proceeded in high yield. Transformation of the
oxadiazole to the triazole by reaction with amines
proved to be a reaction that could be run in parallel,
allowing for the rapid SAR exploration of the R group
off the triazole nitrogen. Although the yields for this
N
11
2
5.22
7.19
N
Values are means of at least three experiments.
a Compounds tested at the recombinant hP2X7 receptor as described.21

4046
W. A. Carroll et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4044–4048
Cl
O
N
N
N
Cl
N
Cl
Cl
Cl
iii
Cl
OMe
i,ii
O
N
R
5, 12-32
Scheme 1. Reagents and conditions: (i) NH2NH2, EtOH; (ii) HC(OEt)3, p-TsOH, toluene, D, 90% yield for two steps; (iii) RNH2, toluene, D.
latter transformation usually did not exceed 50%,22 the
desired product was easily separated from any by-prod-
ucts by either reverse phase HPLC or normal phase col-
umn chromatography.23
with N-formylphenylacetamide in glacial acetic acid at
reflux.
Finally, the imidazole isostere 10 was synthesized by
benzylation of 4-bromoimidazole25 and subsequent
microwave-facilitated Suzuki coupling (Scheme 5). In
the case of the benzylation, the 1,4 regioisomer again
predominated (59%). The Suzuki reaction and product
purification were not optimized, accounting for the
low yield.26
The triazole and pyrazole analogs 6 and 11 were synthe-
sized using the same general method (Scheme 2) wherein
either the starting acetophenone or benzamide was reacted
first with DMF-dimethylacetal, followed by condensation
with benzylhydrazine under buffered conditions.24
The 1,2,3-triazole 7 was prepared in three steps (Scheme
3) from 1,2-dichloro-3-iodobenzene via Sonogashira
coupling with trimethylsilylacetylene followed by cleav-
age of the TMS group and 1,3-dipolar cycloaddition of
the primary acetylene with benzylazide. This latter reac-
tion provided the desired 1,5-regioisomer in low yield
(10%) due to the preferred formation of the 1,4-regio-
isomer (ꢀ20%). Reaction (not shown) of 1-azido-2,3-
dichlorobenzene with 3-phenyl-1-propyne in toluene
for 18 h at 100 ꢁC provided 8 in low yield (20%), again
due to the preferred formation of the 1,4-regioisomer
in 66% yield.
With 2,3-dichlorophenyl and benzyl substitutions held
constant (Table 1), heterocyclic replacements for the tet-
razole moiety were evaluated. There was observed a pro-
gressive loss of P2X7 potency in the order
tetrazole > triazole > imidazole > pyrazole, with the
various triazole regioisomers 5–9 demonstrating similar
P2X7 potency regardless of connectivity (C versus N) to
the phenyl and benzyl groups. In view of this latter re-
sult, it would appear that the position of the nitrogens
in the central ring does not strongly influence activity.
Rather, a stronger relationship seems to exist between
the overall electron density of the core heterocycle and
P2X7 potency. The most electron deficient ring (tetra-
zole) demonstrates the greatest potency, whereas the
more electron rich imidazole and pyrazole rings are
The triazole 9 was prepared in a straightforward fashion
(Scheme 4) by condensing 2,3-dichlorophenylhydrazine
Cl
O
Cl
O
Cl
X
N
ii or iii
Cl
Cl
Cl
i
X
N
X
N
R
6:X=N
11:X=CH
Scheme 2. Reagents and conditions: (i) DMF dimethylacetal, D. X@CH3 yield = 90%; X@NH2 yield = 44%; (ii) X@CH: BnNH2NH2ÆHCl, NaOAc,
H2O/MeOH (1:10), 90 ꢁC; yield = 46%; (iii) X@N: BnNH2NH2ÆHCl, NaOAc, 70% AcOH/dioxane (1:1), 90 ꢁC, yield = 41%.
N
Cl
N
Cl
Cl
N
Cl
Cl
I
Cl
i, ii
iii
7
Scheme 3. Reagents and conditons: (i) HC„CTMS, PdCl2(PPh3)2, CuI, Et3N, D; (ii) K2CO3, MeOH; 70% yield for two steps; (iii) BnN3, EtOH,
80 ꢁC, 10% yield.
N
N
Cl
Cl
N
NHCHO
O
Cl
NHNH2
Cl
i
+
9
Scheme 4. Reagent and condition: (i) AcOH, reflux, 28% yield.

W. A. Carroll et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4044–4048
4047
N
N
N
N
Cl
i
ii
N
Cl
N
H
Br
Br
10
Scheme 5. Reagents and conditon: (i) BnCl, TBAB, KOtBu, 21% yield; (ii) 2,3-dichlorophenylboronic acid, Pd(PPh3)2Cl2, DME/EtOH/H2O (7:3:2),
microwave, 10 min, 6% yield.
Table 4.
Compound
IL-1b release, although with less potency than that ob-
served to inhibit Ca2+ flux. In addition, the compounds
in Table 4 were tested at the rP2X7 receptor where they
also generally showed reduced potency, a potential con-
founding factor when conducting in vivo efficacy studies
in rats.
a
a
THP-1 IL-1b pIC50
rP2X7 pIC50
12
20
32
4
6.43
6.62
7.12
5.60
NT
6.68
6.26
7.11
7.12
6.09
2
Compound 12 was further evaluated in vivo for its
antinociceptive activity in the rat Chung model28 of neu-
ropathic pain. When dosed intraperitoneally, compound
12 produced a dose-dependent reversal of mechanical
allodynia at the 30 min timepoint with an ED50 of
125 lmol/kg and a maximum efficacy of 68%. Consistent
with this activity, 12 demonstrated good bioavailability
(ip) in the rat (62%) and a half-life (1.7 h) commensurate
with the duration of the behavioral assessment.29
Values are means of at least three experiments.
a Compounds tested at the recombinant rP2X7 receptor and for IL-1b
release as described.21
the weakest and the various triazoles are intermediate in
terms of both potency and electron density. From these,
the triazole 5 was selected for further study to determine
if optimization of the N-benzyl substitution could result
in analogs that recapture the potency of 4. Using paral-
lel synthesis techniques, a library of approximately 60
analogs was prepared around 5 utilizing a variety of
benzylamines (selected analogs shown in Table 2).
Briefly, ortho-substitution tended to increase P2X7 po-
tency as for analogs 12–18, whereas para-substitution
decreased potency (25) and meta-substitution produced
mixed effects (23,24). In the ortho-position the sulfone
18 was approximately fivefold less potent than the corre-
sponding sulfide 17. From the small set of ortho-substi-
tuted examples, it is clear that a variety of electron
donating and withdrawing groups are tolerated
although the best activity is observed with the electron-
ically more neutral substitutions (12,16). There was typ-
ically a significant degree of tolerance for substitution in
the meta- and para-positions as long as an ortho-substi-
tuent was present (19–21,26–29). However the reduced
activity for compound 22 is consistent with the meta-
CF3 substituted 24 also being weaker. A number of
phenethylamine analogs were also generated in this
fashion, however, none of these showed any appreciable
P2X7 activity (data not shown). Changing the substitu-
tion on the left hand phenyl from 2,3-dichloro to
2-chloro-3-trifluoromethyl provided comparable potency
(16 vs 17). Further exploration of the left-hand side
substitution was not undertaken.
One of the limitations of earlier tetrazole analogs lack-
ing the pyridine moiety on the right side of the pharma-
cophore was that of low solubility precluding the
routine in vivo evaluation of many analogs.21a Triazole
analogs like 12, however, possess a nitrogen basic en-
ough to permit salt formation27 and sufficient solubility
to permit pharmacokinetic and efficacy studies in vivo.
The improved ADME properties of the triazole isostere
allowed for exhaustive exploration of the benzyl group
SAR without the restriction that pyridine be present
on the right side of the molecule. These studies thus
compliment our earlier work on the left side of the phar-
macophore21a and allowed for the diversification of
P2X7 chemical space and the identification of novel,
selective30 analogs with potency comparable to available
reference compounds. These studies have also provided
additional tool compounds with which to probe the po-
tential of P2X7 antagonists in the treatment of inflam-
matory and neurodegenerative disorders.
References and notes
1. Girdler, G.; Khakh, B. S. Curr. Biol. 2004, 14, R6.
2. North, R. A. Physiol. Rev. 2002, 82, 1013.
3. Morelli, A.; Ferrari, D.; Bolognesi, G.; Rizzuto, R.;
DiVirgilio, F. Drug Dev. Res. 2001, 52, 571.
4. DiVirgilio, F.; Vishwanath, V.; Ferrari, D. Handb. Exp.
Pharmacol. 2001, 151, 355.
5. Sim, J. A.; Young, M. T.; Sung, H.-Y.; North, R. A.;
Surprenant, A. J. Neurosci. 2004, 24, 6307; Andersen, C.
M.; Nedergaard, M. Trends Neurosci. 2006, 29, 257.
6. LeFeuvre, R. A.; Brough, D.; Touzani, O.; Rothwell, N. J.
J. Cereb. Blood Flow Metab. 2003, 23, 381.
Substitution on the benzylic carbon was also investi-
gated. Interestingly, methyl (30) and dimethyl (31) pro-
gressively decreased P2X7 potency, however, tying back
the methyl to form an indane provided the most potent
analog from this series (32).27 In this case, the (R) enan-
tiomer 32 was 2–3 times more potent than the (S) enan-
tiomer 33.
7. LeFeuvre, R.; Brough, D.; Rothwell, N. Eur. J. Pharma-
col. 2002, 447, 261.
8. Rampe, D.; Wang, L.; Ringheim, G. E. J. Neuroimmunol.
2004, 147, 56.
As shown in Table 4, several compounds were evaluated
for their ability to inhibit IL-1b release from human
THP-1 cells. All three of these compounds blocked

4048
W. A. Carroll et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4044–4048
9. Wang, X.; Arcuino, G.; Takano, T.; Lin, J.; Peng, W. G.;
Wan, P.; Li, P.; Xu, Q.; Liu, Q. S.; Goldman, S. A.;
Nedergaard, M. Nat. Med. 2004, 10, 821.
10. Raghavendra, V.; DeLeo, J. A. Adv. Mol. Cell Biol. 2004,
31, 951.
2-methylbenzylamine (1.4 mL, 1.1equiv) were heated in
toluene (3 mL) for 2 days in a sealed tube at 100 ꢁC. The
reaction mixture was directly purified by silica gel flash
chromatography eluting with 3% MeOH/EtOAC to pro-
vide 1.64 g (49%) of compound 12 as a tan solid.
23. Reverse phase HPLC conditions: waters symmetry C8
column (40 mm · 100 mm, 7 lm particle size) using a
gradient of 10–100% acetonitrile: 0.1% aqueous TFA over
12 min (15 min run time) at a flow rate of 70 mL/min.
Silica gel flash column chromatography conditions for
compound 12: 3% MeOH/EtOAc.
24. The 1,5-pyrazole regioisomer 11 was the main product in
the reaction although the 1,3-regioisomer was also formed
as a minor product (6%). The structural assignment was
confirmed by NMR studies showing NOE between the
benzylic protons and the ortho-proton of the 2,3-dichlor-
ophenyl group for 11 whereas this NOE was absent for the
minor 1,3-regioisomer. The regiochemistry for the triazole
6 was assigned by analogy to Lin, Y.-i.; Lang, S. A.;
Lovell, M. F.; Perkinson, N. A. J. Org. Chem. 1979, 44,
4160.
11. Milligan, E. D.; Maier, S. F.; Watkins, L. R. Sem. Pain
Med. 2003, 1, 171.
12. Labasi, J. M.; Petrushova, N.; Donovan, C.; McCurdy, S.;
Lira, P.; Payette, M. M.; Brissette, W.; Wicks, J. R.;
Audoly, L.; Gabel, C. A. J. Immunol. 2002, 168, 6436.
13. 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.
14. Baraldi, P. G.; Romagnoli, R.; Tabrizi, M. A.; Falzoni, S.;
DiVirgilio, F. Bioorg. Med. Chem. Lett. 2000, 10, 681.
15. Ravi, R. G.; Kertesy, S. B.; Dubyak, G. R.; Jacobsen, K.
A. Drug Dev. Res. 2001, 54, 75.
16. Chen, W.; Ravi, R. G.; Kertesy, S. B.; Dubyak, G. R.;
Jacobsen, K. A. Bioconjugate Chem. 2002, 13, 1100.
17. Baraldi, P. G.; del Carmen Nunez, M.; Morelli, A.;
Falzoni, S.; DiVirgilio, F.; Romagnoli, R. J. Med. Chem.
2003, 46, 1318.
18. Baxter, A.; Bent, J.; Bowers, K.; Braddock, M.; Brough,
S.; Fagura, M.; Lawson, M.; McInally, T.; Mortimore,
M.; Robertson, M.; Weaver, R.; Webborn, P. Bioorg.
Med. Chem. Lett. 2003, 13, 4047.
25. Diez-Barra, E.; Hoz, A. de la; Sanchez-Migallon, A.;
Tejeda, J. Synth. Commun. 1993, 23, 1783.
26. Compound 10 was purified using the HPLC conditions
described above in footnote 23. The analogous Suzuki
reaction using 5-bromo-1-(2-methyl-benzyl)-1H-imidazole
rather than 5-bromo-1-benzyl-1H-imidazole provided the
product in improved yield (25%) using silica gel chroma-
tography (70:29:1 EtOAc:Hex:Et3N) to purify the
product.
27. Attempts to form the HCl salt of N-indanyl analogs like
32 resulted in some decomposition, presumably due to an
E1-type process. By contrast, compound 12 was stable to
such treatment and the HCl salt was crystallized from
Et2O/EtOH.
19. Alcaraz, L.; Baxter, A.; Bent, J.; Bowers, K.; Braddock,
M.; Cladingboel, D.; Donald, D.; Fagura, M.; Furber, M.;
Laurent, C.; Lawson, M.; Mortimore, M.; McCormick,
M.; Roberts, N.; Robertson, M. Bioorg. Med. Chem. Lett.
2003, 13, 4043.
20. Gargett, C. E.; Wiley, J. S. Br. J. Pharmacol. 1997, 120,
1483.
21. (a) Nelson, D. W.; Gregg, R. J.; Kort, M. E.; Perez-
Medrano, A.; Voight, E. A.; Wang, Y.; Grayson, G.;
Namovic, M. T.; Donnelly-Roberts, D. L.; Niforatos, W.;
Honore, P.; Jarvis, M. F.; Faltynek, C. R.; Carroll, W. A.
J. Med. Chem. 2006, 49, 3659; (b) 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.; Perez Medrano, A.; Carroll, W.;
Marsh, K.; Sullivan, J. P.; Faltynek, C. R.; Jarvis, M. F.
J. Pharmacol. Exp. Ther. 2006, 319, 1376.
28. Kim, S. H.; Chung, J. M. Pain 1992, 50, 355.
29. Compound 12 was administered at a dose of 5 lmol/kg for
the pharmacokinetics study. The following additional
pharmacokinetic parameters were measured for com-
pound 12. Oral bioavailability = 15%; clearance
(iv) = 3.1 L/h/kg; Vb = 1.9 L/kg; Tmax (ip) = 0.33 h; Cmax
(ip) = 0.21 lg/mL.
30. Compound 12 was profiled across a range of ion channels
and receptors at Cerep and was found to interact
significantly only with the peripheral benzodiazepine site
(86% @ 10 lM). Compound 12 was also inactive at P2X1,
P2X2, P2X2/3, P2X4, P2Y1, and P2Y2 receptors at 10 lM.
22. A typical experimental procedure goes as follows: 2-(2,3-
dichloro-phenyl)-[1,3,4]oxadiazole (2.28 g, 10.6 mmol) and