Identification of non-amidine inhibitors of acid-sensing ion channel-3 (ASIC3)

Article abstract of DOI:10.1016/j.bmcl.2011.05.064

A series of benzothiophene methyl amines were examined in an effort to identify non-amidine chemotypes with reduced polypharmacology from existing leads with the goal of finding potent ASIC3 channel blockers to advance the therapeutic evaluation of ASIC3

Full text of DOI:10.1016/j.bmcl.2011.05.064

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4255–4258
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of non-amidine inhibitors of acid-sensing ion channel-3 (ASIC3)
Scott D. Kuduk a, , Ronald K. Chang a, Christina N. Di Marco a, Robert M. DiPardo a, Sean P. Cook b,
⇑
Matthew J. Cato b, Aneta Jovanovska b, Mark O. Urban b, Michael Leitl b, Robert H. Spencer b,
Stefanie A. Kane b, George D. Hartman a, Mark T. Bilodeau a
a Department of Medicinal Chemistry, Merck Research Laboratories, Sumneytown Pike, PO Box 4, West Point, PA 19486, USA
b Department of Pain Research, Merck Research Laboratories, Sumneytown Pike, PO Box 4, West Point, PA 19486, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of benzothiophene methyl amines were examined in an effort to identify non-amidine chemo-
types with reduced polypharmacology from existing leads with the goal of finding potent ASIC3 channel
blockers to advance the therapeutic evaluation of ASIC3 inhibition.
Received 27 April 2011
Revised 16 May 2011
Accepted 18 May 2011
Available online 27 May 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
ASIC3
Pain
Non-amidine
Blood-to-plasma
A significant medical need remains for new analgesics1 to treat
those who suffer from chronic pain due to damage or persistent
inflammation of neuronal or somatic tissue, and the co-morbidities
associated with it. Under conditions of tissue damage, acidosis of-
ten occurs and contributes to pain.2 Highly sensitive proton-acti-
vated ion channels are present in sensory neurons and are
activated by tissue acidification.3 Acid-sensing ion channels (ASICs)
represent a proton-activated subgroup of the degenerin/epithelial
Na+ channel family.4 Evidence indicates that ASICs serve as proton
sensors and play an important role in conveying the pain from tis-
sue acidosis.5,6 The ASICs are very sensitive, and depending on the
exact composition of the subunits, can detect an extracellular pH
decrease from 7.4 to 7.0.7 A number of inflammatory mediators en-
hance ASIC activity and expression and further implicate neuronal
ASICs as key players in pain from tissue acidosis.8
Several ASIC subunits (ASIC1a/b, ASIC1b2, ASIC2a/b, ASIC3, and
ASIC4) have been identified that are encoded by four genes.9 Chan-
nels containing the ASIC3 subunit have received particular interest
as a target for blocking chronic inflammatory pain as they are
highly abundant in nociceptors where they may act as sensors
for pain from tissue acidosis.10,11 In support of this, mice lacking
the ASIC3 channel do not develop chronic muscle pain from
repeated administration of acid, while ASIC1 knock-out mice are
similar to wild-type.12
The dearth of potent and selective small molecule inhibitors of
ASIC channels has hampered elucidation of the physiological role
of ASIC3. Amiloride 1 is a potassium sparing diuretic agent that
is also a weak, non-selective inhibitor of ASIC channels,13 for which
we have previously examined the ASIC3 SAR.14 In addition, we re-
ported on the identification of amidine derived ASIC3 inhibitors
such as indole 215 and naphthalene 3, which is an analog of A-
31756716 (Fig. 1).
While ASIC3 has generated significant interest in the pain ther-
apeutic field, ASIC1a activity has been associated with anxiety, a
Cl
H2N
Cl
N
N
NH2
N
H
N
NH2
NH2
NH2
NH
O
Cl
1: Amiloride
2
N
NH2
NH
3
⇑
Corresponding author.
E-mail address: scott_d_kuduk@merck.com (S.D. Kuduk).
Figure 1. Known ASIC channel blockers.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.05.064

4256
S. D. Kuduk et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4255–4258
7f, 7h, 7i
m, k
RN
Br
Br
b, c
f
l, k
O
S
S
S
7e, 7g, 7j
CN
CN
R = Boc
R = H
Cl
g
Cl
Cl
Cl
9
8
10
12a-b
k
a
d, e
7c, 7d
Br
f, g
Br
NHBoc
S
NH2
S
OH
Cl
O
11
Cl
7a
f
HO
Cl
S
NHBoc
S
NHBoc
h
i, j
7k, 7l, 7m
Cl
Cl
13
14
n, e, o
p, q
N
BocN
BocN
OH
OTf
15
16
17
Scheme 1. Reagents and conditions: (a) SOCl2, pyridine, chlorobenzene, 50?115 °C; (b) NH3, MeOH; (c) TEA, TFAA, CH2Cl2; (d) BH3–THF; (e) (Boc)2O, CH2Cl2; (f) alkyne,
Pd(Ph3P)4, CuI, Et3N, DMF; (g) 4 N HCl in dioxane, CH2Cl2; (h) SOCl2, CH2Cl2; (i) NHR1R2, K2CO3, CH3CN; (j) HCl, EtOAc; (k) LAH, THF, À40 °C; (l) RCHO, AcOH, TEA, NaBH(OAc)3;
(m) RX, TEA, CH2Cl2; (n) PtO2, H2, AcOH; (o) Tf2O, CH2Cl2; (p) trimethylsilylacetylene, Pd(Ph3P)4, CuI, Et3N, DMF, 80 °C; (q) TBAF, THF.
NH2
I-
j-m
n
o
p, h
7b
N
N+
N
O
Br
Br
Br
Br
Br
24
25
26
27
S
S
NHBoc
S
a, b
c, d
CN
CN
+
Br
Cl
Cl
Cl
18
19
20
21
7n
7o
e
i, h
h
H
N
O
NHBoc
NHBoc
S
S
f, g
22
23
Cl
Cl
Scheme 2. Reagents and conditions: (a) Trimethylsilylacetylene, Pd(Ph3P)4, CuI, Et3N, DMF, 80 °C; (b) TBAF, THF; (c) LAH, THF, À40 °C; (d) (Boc)2O, CH2Cl2; (e) Pd(Ph3P)4, CuI,
Et3N, DMF; (f) Ti(OiPr)4, NH2Me, MeOH; (g) NaBH4, MeOH; (h) 4 N HCl in dioxane, CH2Cl2; (i) RCHO, AcOH, TEA, NaBH(OAc)3; (j) ethyl formate, reflux; (k) oxalyl chloride,
CH2Cl2; (l) FeCl3, À20 °C; (m) concd H2SO4, MeOH, reflux; (n) MeI, CH3CN; (o) MeMgCl, THF, À78 °C; (p) alkyne 20, Pd(Ph3P)4, CuI, Et3N, DMF.
symptom that can often accompany depression-like behaviors.17,18
Wemmie and co-workers disclosed that A-317567 also blocks ASI-
C1a and produced antidepressant effects in mice when dosed intra-
cerebroventricularly.19 We recently reported the pharmacological
characterization of 3, a close molecular analog of A-317567, and
questioned whether the observed antinociceptive effects attrib-
uted to ASIC3 may be due to off-target effects, potentially includ-
ing, but not limited to other ASIC channel subtypes.20 A common
feature of the reported ASIC inhibitors in Figure 1 is the presence
of an acylguanidine or amidine moiety, which may contribute to
potential off-target effects. In addition, both series had poor
pharmacokinetic properties. Accordingly, we wanted to identify
non-amidine ASIC3 inhibitors that could serve as better tool com-
pounds to explore the pharmacology of ASIC3.
The chemistry utilized to prepare requisite target molecules is
highlighted in Schemes 1 and 2. Benzothiophene 9 was produced
via thionyl chloride and subsequently can be converted in two
steps to nitrile 10. Sonogashira coupling of the bromide with acet-

S. D. Kuduk et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4255–4258
4257
N
S
NH2
NH
NH2
NH2
NH
Cl
3
NH
6
4
hASIC3 EP: IC50 ~ 0.46 μM
hASIC3 EP: IC50 3.4 μM
hASIC3 EP: IC50 6.3 μM
NH2
S
NH2
Cl
5
hASIC3 EP: IC50 4.8
7a
hASIC3 EP: IC50 2.8 μM
M
μ
Figure 2. Derivation of amine and thiophene structures.
ylene 17 afforded Boc protected intermediate 12a, which was read-
ily converted to analogs 7c–j as described. Alternatively, 10 was
first reduced with borane followed by Boc protection to produce
11, which underwent Sonogashira with the appropriate alkynes
followed by deprotection to produce 7a and 13. Alcohol 13 was
subsequently functionalized as described to afford 7k–m.
In Scheme 2, bromide 27 was obtained via a four-step sequence
with the key step involving addition of methyl Grignard to dihy-
droisoquinolinium salt 26. Sonogashira coupling of 27 with alkyne
20 was followed by deprotection to afford 7b. Finally, alkyne 20
was similarly utilized with bromide 21, and after subsequent func-
tionalization, 7n,o were obtained.
compound in the form of amine 7b. For simplicity, SAR was con-
ducted without the methyl group on the tetrahydroisoquinoline,
which led to a very modest reduction in ASIC inhibition (7e,
IC50 = 0.68
lM). Much of this could be recovered by placing a larger
ethyl (7f; IC50 = 0.23
l
M) or iPr (7g; IC50 = 0.31 M) on the nitro-
l
gen in lieu of the methyl. Having a basic amine proved critical as
analogs 7c, 7h–i were considerably less active. Of note, the unsub-
stituted analog 7d proved most potent with the additional benefit
of high solubility.
Alternative constraints were also studied and found to be
tolerated (7j–o). The tetrahydronaphthalenes provided good inhi-
bition, with the dimethyl analog (7k; IC50 = 0.23 lM) yielding
The initial SAR exploration from amidine 3 began with trun-
cated phenyl acetylene 4, which lost ꢀ14-fold potency relative to
3 (Fig. 2). The primary amine analog (5) of amidine 4 showed a
slight improvement representing the first potent non-amidine
structure. The chlorothiophene-amidine 6 represented a non-
naphthalene template with ꢀ2-fold improvement in ASIC3 inhibi-
tion. The hybrid of these two provided chlorothiophene primary
similar activity to 7d. The ring was critical as un-constrained
dimethylamine 7n lost greater than 16-fold relative to tetrahydro-
naphthalene 7k and indene 7o.
Basedupon potency and good solubility properties, amine 7d was
evaluated further in the rat CFA model of inflammatory pain. In this
study (Fig. 4), 7d showed a robust reversal of mechanical hypersen-
sitivity in the hind paw and decreased weight bearing 30 min post-
dosing (ip) in male Sprague–Dawley rats. The magnitude of reversal
was similar to that from the NSAID naproxen (dosed at 20 mg/kg po)
at the 100 mpk dose of 7d. The average plasma concentration21 at
amine 7a with an ASIC3 IC50 = 2.8
to naphthalene 5.
lM, an improvement relative
With 7a identified as a non-naphthyl, non-amidine starting
point, SAR analysis was investigated, with a pictorial summary for
the thiophene region shown in Figure 3. The primary amine was re-
quired as substitution directly on, or alpha to the nitrogen, led to a
severe loss (>20Â) in activity. The thiophene was required as indoles
lost some activity, whereas furan derivatives were inactive. Lastly,
the chlorine was required as polar or smaller groups were
substantially weaker inhibitors. Based upon this analysis the right
hand primary amine and chlorobenzothiophene of 7a was main-
tained for all subsequent SAR analysis of the alkynyl region.
30 min was 3.6
lM at 100 mg/kg. Of note, very high levels in brain
(14 M) were observed at this dose, which unlike amidines 2 and 3
l
distinguishes amine 7d as the only CNS penetrant ASIC3 antagonist
we have identified and examined to date. However, rats treated with
7d appeared sedated following administration, but were easily
aroused and normal when handled, a phenomenon that has been ob-
served previously in this model at an efficacious dose of amidine 3.
Amine 7d and related analogs also had exceptionally poor phar-
macokinetic properties exhibiting super-hepatic clearance, partic-
ularly in dog. Upon further interrogation, it was realized that
these compounds exhibited very high blood to plasma ratios,
which complicates their analysis unlike their amidine counter-
parts. For example, amine 7d had a blood to plasma ratio of 15.4
in dog, while amidine 3 gave a ratio of 2.6.
An SAR study was next conducted on the phenyl acetylene
region of 7a as shown in Table 1. Incorporation of the tetrahydro-
isoquinoline ring found in 3 provided a ꢀ2-fold more potent
O
S > N >
primary
amine
One of the complicating issues with amidine 3 as a tool com-
pound was off-target activity. In addition to ASIC1a activity, 3 was
evaluated against a number of receptors and enzymes (MDS Pharma
Service, Taipei, Taiwan) and found to be highly promiscuous with
required
NH2
S
α-subs.
binding affinity IC50’s <10 lM against 39 targets that could contrib-
not tolerated
Cl
ute to the behavioral effects such as muscarinic, adrenergic, dopa-
mine, norepinephrine, and serotonin receptors. Similar to amidine
3, amine 7d exhibited no selectivity for ASIC3 with an ASIC1a
IC50 = 224 nM. In addition, 7d was evaluated in the same
7a
Cl ~ OMe >>H
Figure 3. Thiophene SAR summary.

4258
S. D. Kuduk et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4255–4258
Table 1
30 min post-administration
ASIC3 titration of select amiloride derivatives
100
75
R
NH2
S
vehicle
100 mg/kg 7d
20 mg/kg naproxen
Cl
a
Compd
Structure
ASIC3 IC50
45%
50
25
0
40%
N
7b
0.27
Me
Me
6%
N
7c
7d
7e
7f
>50
0.22
0.68
0.23
0.31
10.2
>50
17.1
0.23
0.51
2.5
Boc
HN
Figure 4. Evaluation of 7d and naproxen in the rat CFA model of inflammatory pain
at 30 min post-dosing. ⁄p <0.05 (ANOVA).
by another mechanism altogether. This is particularly an issue for
7d which is brain penetrant, unlike previously investigated ASIC3
antagonists.
N
Me
Et
In summary, amine 7d was identified as a potent, non-amidine
ASIC3 inhibitor with good brain penetration. However, 7d also
caused sedative or lethargic effects in the rat CFA model. In addition,
poly-pharmacology for a number of targets including ASIC1a pre-
vents this class from being pursued further. Accordingly the identi-
fication of an ASIC3 subtype-selective inhibitor lacking other off-
target effects remains elusive as a tool to study the relationship be-
tween efficacy in pain models and the apparent peripheral sedating
effects observed with non-selective compounds.
N
N
7g
7h
7i
iPr
F2HC
F3C
N
References and notes
N
1. Childers, W. E.; Gilbert, A. M.; Kennedy, J. D.; Whiteside, G. T. Expert Opin. Ther.
Patents 2008, 18, 1027.
2. Krishtal, O. Trends Neurosci. 2003, 26, 477.
3. Reeh, P. W.; Kress, M. Curr. Opin. Pharmacol. 2001, 1, 45.
4. Wemmie, J. A.; Price, M. P.; Welsh, M. J. Trends Neurosci. 2006, 29, 578.
5. Lingueglia, E. J. Biol. Chem. 2007, 282, 17325.
6. Jones, N. G.; Slater, R.; Cadiou, H.; McNaughton, P.; McMahon, S. B. J. Neurosci.
2004, 4, 10974.
N
7j
Bn
7. Waldmann, R. Adv. Exp. Med. Biol. 2001, 502, 293.
8. Mamet, J.; Baron, A.; Lazdunski, M.; Voilley, N. J. Neurosci. 2002, 22, 10662.
9. Kellenberger, S.; Schild, L. Physiol. Rev. 2002, 82, 735.
10. Babinski, K.; Le, K. T.; Seguela, P. J. Neurochem. 1999, 72, 51.
11. Yiangou, Y.; Facer, P.; Smith, J. A.; Sangameswaran, L.; Eglen, R.; Knowles, C.;
Williams, N.; Anand, P. Eur. J. Gastroenerol. Hepatol. 2001, 13, 891.
12. Sluka, K. A.; Price, M. P.; Breese, N. M.; Stucky, C. L.; Wemmie, J. A.; Welsh, M. J.
Pain 2003, 106, 229.
N
7k
7l
N
13. Ugawa, S.; Ueda, T.; Ishida, Y.; Nishigaki, M.; Shibata, Y.; Shimada, S. J. Clin.
Invest. 2002, 110, 1185.
14. Kuduk, S. D.; Di Marco, C. N.; Chang, R. K.; DiPardo, R. M.; Cook, S. P.; Cato, M. J.;
Jovanaskova, A.; Urban, M. O.; Leitl, M.; Spencer, R. H.; Kane, S. A.; Bilodeau, M.
T.; Hartman, G. D.; Bock, M. G. Bioorg. Med. Chem. Lett. 2009, 19, 2514.
15. Kuduk, S. D.; Chang, R. K.; Wai, J.; Cofre, V.; Di Marco, C. N.; DiPardo, R. M.;
Cook, S. P.; Cato, M. J.; Jovanaskova, A.; Urban, M. O.; Leitl, M.; Spencer, R. H.;
Kane, S. A.; Hartman, G. D.; Bilodeau, M. T. Bioorg. Med. Chem. Lett. 2009, 19,
4059.
16. Dubé, G. R.; Lehto, S. G.; Breese, N. M.; Baker, S. J.; Wang, X.; Matulenko, M. A.;
Honore, P.; Stewart, A. O.; Moreland, R. B.; Brioni, J. D. Pain 2005, 117, 88.
17. Coryell, M. W.; Ziemann, A. E.; Westmoreland, P. J.; Haenfler, J. M.; Kurjakovic,
Z.; Zha, X. M.; Price, M.; Schnizler, M. K.; Wemmie, J. A. Biol. Psychiatry 2007, 62,
1140.
N
7m
7n
7o
N
4.3
0.19
N
18. Coryell, M. W.; Wunsch, A. M.; Haenfler, J. M.; Allen, J. E.; McBride, J. L.;
Davidson, B. L.; Wemmie, J. A. J. Neurosci. 2008, 28, 13738.
19. Coryell, M. W.; Wunsch, A. M.; Haenfler, J. M.; Allen, J. E.; Schnizler, M. K.;
McBride, J. L.; Ziemann, A. E.; Cook, M. N.; Dunning, J. P.; Price, M. P.; Rainier, J.
D.; Liu, A.; Light, A. R.; Langbehn, D. R.; Wemmie, J. A. J. Neurosci. 2009, 29,
5381.
20. Kuduk, S. D.; Di Marco, C. N.; Cook, S. P.; Cato, M. J.; Jovanaskova, A.; Urban, M.
O.; Leitl, M.; Spencer, R. H.; Kane, S. A.; Hartman, G. D.; Bilodeau, M. T. ACS.
Chem. Neuro. 2010, 1, 19.
a
Electrophysiology recording, inhibition (lM) of current was expressed as per-
cent inhibition of peak current versus baseline peak current. N = 3.
aforementioned MDS Pharma Service panel, and found to have bind-
ingaffinitieswith IC50’s <10 lMagainst35targets, witha substantial
overlap with amidine 3. Accordingly, the analgesic efficacy of non-
selective ASIC3 antagonists as represented by these compounds, is
further complicated by their CNS-related effects through ASIC1 or
21. Amine 7d (brain to plasma ratio ꢀ6) was found to be 100% bound to rat plasma
proteins.