Discovery of potent and use-dependent sodium channel blockers for treatment of chronic pain

Article abstract of DOI:10.1016/j.bmcl.2005.02.093

A new series of voltage-gated sodium channel blockers with potential for treatment of chronic pain is reported. Systematic structure-activity relationship studies, starting with compound 1, led to identification of potent analogs that displayed use-dependent block of sodium channels, were efficacious in pain models in vivo, and most importantly, were devoid of activity against the cardiac potassium channel hERG.

Full text of DOI:10.1016/j.bmcl.2005.02.093

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2943–2947
Discovery of potent and use-dependent sodium channel blockers
for treatment of chronic pain
Jun Liang,a,* Richard M. Brochu,b Charles J. Cohen,b Ivy E. Dick,b John P. Felix,b
Michael H. Fisher,a Maria L. Garcia,b Gregory J. Kaczorowski,b Kathryn A. Lyons,a
Peter T. Meinke,a Birgit T. Priest,b William A. Schmalhofer,b McHardy M. Smith,b
Jason W. Tarpley,c Brande S. Williams,b William J. Martinc and William H. Parsonsa
aDepartment of Medicinal Chemistry, Merck Research Laboratories, Rahway, NJ 07065, USA
bDepartment of Ion Channels, Merck Research Laboratories, Rahway, NJ 07065, USA
cDepartment of Pharmacology, Merck Research Laboratories, Rahway, NJ 07065, USA
Received 20 December 2004; revised 18 February 2005; accepted 22 February 2005
Available online 4 May 2005
Abstract—A new series of voltage-gated sodium channel blockers with potential for treatment of chronic pain is reported. System-
atic structure–activity relationship studies, starting with compound 1, led to identification of potent analogs that displayed use-
dependent block of sodium channels, were efficacious in pain models in vivo, and most importantly, were devoid of activity against
the cardiac potassium channel hERG.
Ó 2005 Published by Elsevier Ltd.
1. Introduction
nantly expressed in the peripheral nervous system and
dorsal root ganglion (DRG),6 and have been implicated
in pain transmission pathways.7 Sodium channel block-
ers used to treat chronic pain do not discriminate be-
tween Nav1 subtypes and derive their therapeutic
index from their use-dependent properties. The objective
of this study was to identify potent, use-dependent, and
orally bioavailable sodium channel blockers with good
therapeutic windows, that could be developed for treat-
ment of chronic pain in patients.
It is widely recognized that chronic pain encompasses a
number of heterogeneous symptoms, which can gener-
ally be characterized as inflammatory or neuropathic,
the latter representing a category for which few thera-
pies currently exist. Injuries or diseases of the peripheral
nervous system can lead to neuropathic pain.1 Experi-
mentally and clinically, such pain states are associated
with hyper-excitability and spontaneous action potential
firing in peripheral sensory neurons.
Voltage-gated sodium channels (Nav1) underlie the initi-
ation and propagation of action potentials in peripheral
neurons, and are therefore appealing molecular targets
for treating chronic pain.2 Proof of concept for Nav1
channels in the clinic has been suggested from the use
of blockers such as carbamazepine, lamotrigine, and
lidocaine.3 A shared feature among these treatments
for chronic pain is the ability to block sodium channels,
and, at therapeutic concentrations, sodium channel
block is the only known effect of lidocaine. Several sub-
types of sodium channels have been identified, including
Nav1.7 (PN1)4 and Nav1.8 (PN3),5 which are predomi-
2. Results and discussion
Compound 1 (Fig. 1) was first disclosed by scientists
at Ciba-Geigy as a highly potent blocker of cardiac
sodium channels.8 This compound was prepared and
found to be active against Nav1.7 in a functional, Volt-
age/Ion Probe Reader (VIPR) assay that measures
veratridine-induced depolarization in HEK-293 cells
stably transfected with Nav1.7 channels9 (IC50
4.0 lM). Compound 1 was also tested in a rat formalin
paw model10 as previously described.9b In this model,
injection of formalin into the hind paw induces two
phases of spontaneous pain behavior. Compound 1 re-
duced phase II (10–60 min) formalin-induced flinching
in a dose-dependent manner when injected locally into
*
Corresponding author. E-mail: jun_liang@merck.com
0960-894X/$ - see front matter Ó 2005 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2005.02.093

2944
J. Liang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2943–2947
With compound 1 as a lead, an effort was directed to de-
velop potent, use-dependent Nav1.7 sodium channel
blockers, with minimal hERG activity. Herein is re-
ported the synthesis as well as in vitro and in vivo activi-
ties of novel voltage-gated sodium channel blockers.
3. Removal of the basic amino group reduced hERG
activity
Figure 1. Structure of compound 1.
A convergent two-step synthesis was designed that
afforded the p-chloro analogs of 1 (Scheme 1), taking
advantage of the commercially available p-chlorophenyl-
thiazole methylchloride. The C3-desmethoxy analog 4
was also synthesized in a similar manner for comparison.
Both compounds had in vitro activities similar to 1
(Table 1), suggesting that the p-chloro substitution was
well tolerated. Not unexpectedly, significant hERG
(MK-499) activity was present in both compounds.
It was suspected that the potent hERG activity of 1 (IC50
0.18 lM) could be attributed to its basic tertiary amino
group.11 This hypothesis was verified in a series of cyclic
amides that rendered the amine non-basic. The amide
analogs were prepared by oxidation of the aldehyde 2
to the carboxylic acid, followed by amide formation with
various cyclic amines. Indeed, hERG activities were
diminished in these amide analogs (Table 2), especially
as the ring size of cyclic amide decreased. The affinities
of these amides for sodium channels were largely re-
tained, especially for the pyrrolidine analogs 11 and 12.
Figure 2. In vivo activity of 1 in a rat model of pain.
the paw (Fig. 2); however it was inactive when dosed
intravenously at 3 mg/kg (13% inhibition). Further evalu-
ation of the compound indicated a number of off-target
activities. Of these, the effect on the cardiac potassium
channel hERG11 (determined as displacement of 35S-
MK-499 binding to hERG channels stably expressed
in HEK293 cells, IC50 0.18 lM) was considered the
greatest liability.
Both 11 and 12 were studied in detail in vitro and in vivo
in a rat formalin paw model (Table 3). Whole cell elec-
trophysiology9b was used to evaluate the state depen-
dence of block, and results suggested that both
analogs blocked the inactivated state much more
Compound 1 was administered directly into a single
hind paw prior to formalin injection into the same site.
Behavioral assessment consisted of counting the number
of formalin-evoked hind paw flinches for 1 h post-injec-
tion. Data are presented as mean percent inhibition
from 10–60 min post-formalin relative to vehicle control
(*p < 0.05, **p < 0.01; ANOVA with BonferroniÕs Mul-
tiple Comparison Test compared to vehicle treatment;
n = 4–10 rats/group).
Table 1. VIPR and MK-499 activities
1
3
4
VIPR (NAv1.7, IC50, lM)
MK-499 (IC50, lM)
4.0
0.18
3.8
0.04
2.9
0.03
Scheme 1.

J. Liang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2943–2947
Table 2. VIPR and hERG activities of cyclic amides
2945
Compound
R
R0
VIPR Nav1.7 (IC50, lM)
MK-499 (inh. at 10 lM) (%)
5
OCH3
19.4
88
6
7
8
H
29
6.5
40
90
85
90
OCH3
H
9
OCH3
H
3.7
7.4
1.3
1.0
63
89
50
63
10
11
12
OCH3
H
Table 3. In vitro and in vivo profiles of analogs 11 and 12
11
12
VIPR, Nav1.7, IC50 (lM):
Nav1.7, ephys. KR/KI (lM):
MK-499, IC50 (lM):
Clp (mL/min/kg):
1.3
10/0.14
7.7
1.0
6.7/0.16
9.8
627
101
t1/2 (h):
F%
Formalin paw response (10–40 min, 3 mg/kg, iv):
0.48
9.4
0.44
17
68%
65%
potently than the resting state (Table 3). Unlike 1 which
was ineffective during the first 45 min post-dose when gi-
ven intravenously (19% inhibition), 11 and 12 were com-
pounds equally efficacious in the rat formalin paw model
(68% and 65%, respectively; p < 0.01, unpaired t-test vs
vehicle). Duration of action was brief (<60 min), but
consistent with rat pharmacokinetic data for both com-
pounds showing high clearance rates and short half-
lives.
compounds with a phenyl thiazole side chain (Nav1.7:
IC50 0.55 vs 1.3 lM for 11). However, this compound
was not examined in animal models due to concerns
regarding the well-documented metabolism issues of thio-
phenes in vivo.12
A methyl group was introduced at the benzylic position,
in an attempt to reduce the potential O-dealkylation
pathway. The in vitro activity of 20 was comparable
to that of 11 (Nav1.7: IC50 2.2 vs 1.3 lM); however, a
moderate improvement in the rat PK parameters was
noted (20: Clp 73 mL/min/kg; t1/2 0.8 h).
4. Exploring the C2-O-alkyl side chain
Replacement of the p-chlorophenylthiazole with other
side chains was studied (Table 4). In most cases, analogs
were synthesized via O-alkylation of the C2-phenol with
commercially available electrophiles. It is noteworthy
that compounds with biaryl groups tend to be more ac-
tive than those with only one phenyl ring. The bithio-
phene analog 19 was slightly more active than the
5. Probing the C3-Substituent on the phenyl ring
The role of the C3-substituent on the phenyl ring was
investigated (Table 5). A series of compounds with dif-
ferent substituents at the C3-position were prepared. A
variety of substituents were tolerated at this position;

2946
J. Liang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2943–2947
Table 4. In vitro activity of different O-alkyl analogs
R
Compd.
VIPR
13
9.0
14
>30
15
>30
16
21
Nav1.7, IC50 (lM)
R
Compd.
VIPR
17
4.0
18
0.86
19
0.55
20
2.2
Nav1.7, IC50 (lM)
however, the more polar hydroxyl group and bulky sub-
stituents (i.e., O-tert-butyl) displayed markedly dimin-
ished potency against the Nav1.7 channel.
stituent was tied to the 2-O-alkyl side chain within a
dihydropyran ring, was prepared. It was envisioned that
removing the metabolically labile C3-methoxy and C2-
O-benzyl methylene groups might improve the PK pro-
file, while preventing any unfavorable steric interactions
that could compromise potency.
6. A conformationally restricted analog of 11
To assess the effect of conformational rigidity, a con-
formationally constrained analog in which the C3-sub-
A short synthesis to 30 was developed (Scheme 2), with a
biphenyl group replacing the p-chlorophenylthiazole in
11. With the pyrrolidine amide 27 as starting material,
it was noted that a significant amount of a byproduct,
0
arising from an SN2 displacement pathway, was formed
in the Mitsunobu reaction.13 Nevertheless, the desired
SN2 product 28 was obtained in 65% yield. 28 was
heated to reflux in xylene for 24 h to effect the thermal
Claisen rearrangement.14 Subsequent Prins cyclization15
in the presence of catalytic amount of p-toluenesulfonic
acid in refluxing benzene afforded 30.16 It is noteworthy
that the competing 5-exo-trig cyclization was not ob-
served, presumably due to the favored formation of
the more stable benzylic carbenium ion. The conforma-
tionally constrained analog 30 displayed comparable in
vitro activity (Nav1.7 VIPR 3.6 lM; ephys. KR/KI
4.3 lM/0.21 lM), was reasonably clean in hERG activ-
ity (IC50 5.4 lM), while its clearance rate was reduced
(Clp 40 mL/min/kg) and half-life in vivo was doubled
(t1/2 0.88 h).
Scheme 2. Synthesis of a conformationally constrained analog.
Table 5. In vitro activity of C3-substituted analogs
R
Compd.
OCH3
11
1.3
H
OH
21
CH3
22
Cl
23
OEt
24
OiPr
25
OtBu
26
22
12
1.0
VIPR
Nav1.7, IC50 (lM)
>30
0.55
0.73
6.6
7.5

J. Liang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2943–2947
2947
8. McCulough, J. R.; Trube, G.; Conder, M. L. Biophys. J.
1987, 51, 260a.
7. Conclusion
9. (a) Felix, J. P.; Williams, B. S.; Priest, B. T.; Brochu, R.
M.; Dick, I. E.; Warren, V. A.; Yan, L.; Slaughter, R. S.;
Kaczorowski, G. J.; Smith, M. M.; Garcia, M. L. Assay
Drug Dev. Tech. 2004, 2, 260; (b) Priest, B. T.; Garcia, M.
L.; Middleton, R. E.; Brochu, R. M.; Clark, S.; Dai, G.;
Dick, I. E.; Felix, J. P.; Liu, C. J.; Reiseter, B. S.;
Schmalhofer, W. A.; Shao, P.; Tang, Y. S.; Chuo, M. Z.;
Kohler, M. G.; Smith, M. M.; Warren, V. A.; Williams, B.
S.; Cohen, C. J.; Martin, W. J.; Meinke, P. T.; Parsons, W.
H.; Wafford, K. A.; Kaczorowski, G. J. Biochemistry
2004, 43, 9866.
Voltage-gated sodium channels have recently emerged
as attractive targets for treatment of chronic pain. Po-
tent and state-dependent blockers of sodium channels
have been identified with minimal hERG activity and
acceptable pharmacokinetic properties. Such com-
pounds are also found to be efficacious in animal pain
models, suggesting that they could be developed as no-
vel analgesics to relieve chronic pain in patients.
10. (a) Le Bars, D.; Gozariu, M.; Cadden, S. W. Pharmacol.
Rev. 2001, 53, 597; (b) Wheeler-Aceto, H.; Cowan, A.
Psychopharmarcology 1991, 104, 35; (c) Wheeler-Aceto,
H.; Porreca, F.; Cowan, A. Pain 1990, 40, 229.
11. (a) Cavalli, A.; Poluzzi, E.; De Ponti, F.; Recanatini, M. J.
Med. Chem. 2002, 45, 3844; (b) Mitcheson, J. S.; Chen, J.;
Lin, M.; Culberson, C.; Sanguinetti, M. C. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 12329; (c) Wang, J.; Penna, K.
D.; Wang, H.; Karczewski, J.; Connolly, T. M.; Koblan,
K. S.; Bennett, P. B.; Salata, J. J. Am. J. Physiol. Heart
Circ. Physiol. 2003, 284, H256.
12. (a) King, L. C.; Kohan, M. J.; Brooks, L.; Nelson, G. B.;
Ross, J. A.; Allison, J.; Adams, L.; Desai, D.; Amin, S.;
Padget, W.; Lambert, G. R.; Richard, A. M.; Nesnow, S.
Chem. Res. Toxicol. 2001, 14, 661; (b) Valadon, P.;
Dansette, P. M.; Girault, J. P.; Amar, C.; Mansuy, D.
Chem. Res. Toxicol. 1996, 9, 1403; (c) Mansuy, D.;
Dansette, P. M. Adv. Exp. Med. Biol. 1996, 387, 1.
13. (a) Mitsunobu, O. Synthesis 1981, 1; (b) Tsunoda, T.;
Yamamiya, Y.; Ito, S. Tetrahedron Lett. 1993, 34, 1639.
14. Wipf, P. In Comprehensive Organic Synthesis; Trost, B.
M., Fleming, I., Paquette, L. A., Eds.; Pergamon: New
York, 1991; Vol. 5, p 827.
References and notes
1. (a) Jensen, T. S.; Gottrup, H.; Sindrup, S. H.; Bach, F. W.
Eur. J. Pharmacol. 2001, 429, 1; (b) Julius, D.; Basbaum,
A. I. Nature 2001, 413, 203; (c) Yaksh, T. Trends
Pharmacol. Sci. 1999, 20, 329.
2. (a) Black, J. A.; Cummins, T. R.; Dib-Hajj, S. D.;
Waxman, S. G. In Mechanisms and Mediators of Neuro-
pathic Pain; Malmberg, A. B., Chaplan, S. R., Eds.;
Birkhauser: Basel, 2002; pp 23–50; (b) Lai, J.; Hunter, J.
C.; Porreca, F. Curr. Opin. Neurobiol. 2003, 13, 291; (c)
Anger, T.; Madge, D. J.; Mulla, M.; Riddall, D. J. Med.
Chem. 2001, 44, 115; (d) Catterall, W. A. Neuron 2000, 26,
13.
3. (a) Mao, J.; Chen, L. L. Pain 2000, 87, 7; (b) Kalso, E.;
Tramer, M. R.; McQuay, H. J. Eur. J. Pain 1998, 2, 3; (c)
Vu, T. N. Curr. Pain Headache Rep. 2004, 8, 15.
4. Toledo-Aral, J. J.; Moss, B. L.; He, Z. J.; Koszowski, A.
G.; Whisenand, T.; Levinson, S. R.; Wolff, J. J.; Silos-
Santiago, I.; Halegoua, S.; Mandel, G. Proc. Natl. Acad.
Sci. U.S.A. 1997, 94, 1527.
5. Sangameswaran, L.; Delgado, S. G.; Fish, L. M.; Koch, B.
D.; Jakeman, L. B.; Stewart, G. R.; Sze, P.; Hunter, J. C.;
Eglen, R. M.; Herman, R. C. J. Biol. Chem. 1996, 271,
5953.
15. (a) Overman, L. E.; Pennington, L. D. Can. J. Chem. 2000,
78, 732; (b) Cloninger, M. J.; Overman, L. E. J. Am.
Chem. Soc. 1999, 121, 1092.
1
6. (a) Akopian, A. N.; Sivilotti, L.; Wood, J. N. Nature 1996,
379, 257; (b) Tate, S.; Benn, S.; Hick, C.; Trezise, D.;
Mannion, R. J.; Costigan, M.; Plumpton, C.; Grose, D.;
Gladwell, Z. Nature Neurosci. 1998, 1, 653; (c) Sleeper, A.
A.; Cummins, T. R.; Hormuzdiar, W.; Tyrrel, L.; Dib-
Hajj, S. D.; Waxman, S. G.; Black, J. A. J. Neurosci. 2000,
20, 7279.
16. Spectra data for compound 30: H NMR (CDCl3) d 7.62
(d, 4H, J = 7.2 Hz), 7.47 (d, 4H, J = 7.3 Hz), 7.38 (t, 1H,
J = 5.2 Hz), 7.18 (1H, dt, J = 5.2 Hz), 7.12 (1H, dt,
J = 5.1 Hz), 5.20 (dd, 1H, J = 7.8, 2.4 Hz), 3.55 (m, 2H),
3.41 (br s, 1H), 3.32 (m, 1H), 3.04 (m, 1H), 2.84 (m, 1H),
2.36 (m, 1H), 2.11 (m, 1H), 1.91 (m, 4H); 13C NMR
(CDCl3) 168.0, 150.8, 140.9, 140.7, 140.6, 130.5, 129.0,
127.6, 127.4, 127.3, 127.2, 126.2, 125.9, 122.7, 120.7, 47.9,
45.7, 29.9, 26.1, 25.2, 24.9.
7. Lai, J.; Gold, M. S.; Kim, C. S.; Bian, D.; Ossipov, M. H.;
Hunter, J. C.; Porreca, F. Pain 2002, 95, 143.