Novel Potent Orthosteric Antagonist of ASIC1a Prevents NMDAR-Dependent LTP Induction

Article abstract of DOI:10.1021/jm5017329

Acid sensing ion channels 1a (ASIC1a) are of crucial importance in numerous physiological and pathological processes in the brain. Here we demonstrate that novel 2-oxo-2H-chromene-3-carboxamidine derivative 5b, designed with molecular modeling approach, inhibits ASIC1a currents with an apparent IC₅₀ of 27 nM when measured at pH 6.7. Acidification to 5.0 decreases the inhibition efficacy by up to 3 orders of magnitude. The 5b molecule not only shifts pH dependence of ASIC1a activation but also inhibits its maximal evoked response. These findings suggest that compound 5b binds to pH sensor of ASIC1a acting as orthosteric noncompetitive antagonist. At 100 nM, compound 5b completely inhibits induction of long-term potentiation (LTP) in CA3-CA1 but not in MF-CA3 synapses. These findings support the knockout data indicating the crucial modulatory role of ASIC1a channels in the NMDAR-dependent LTP and introduce a novel type of ASIC1a antagonists (Chemical Equation).

Full text of DOI:10.1021/jm5017329

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Novel orthosteric antagonist of ASIC1a
prevents NMDAR-dependent LTP induction
Andriy Buta, Oleksandr Maximyuk, Dmytro B Kovalskyy, Volodymyr Sukach, Mykhaylo Vovk,
Oleksandr Ievglevskyi, Elena Isaeva, Dmytro Isaev, Alina Savotchenko, and Oleg Krishtal
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 14 May 2015
Downloaded from http://pubs.acs.org on May 15, 2015
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 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Novel potent orthosteric antagonist of ASIC1a
prevents NMDARꢀdependent LTP induction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
§
§‡
‡†
#‡
∥
∥
⊥
Andriy Buta , Oleksandr Maximyuk , Dmytro Kovalskyy , Volodymyr Sukach , Mykhailo
#
∥
∥
∥
∥
Vovk , Oleksandr Ievglevskyi , Elena Isaeva , Dmytro Isaev , Alina Savotchenko and Oleg
§
*
∥
Krishtal .
∥Bogomoletz Institute of Physiology of NAS Ukraine, 4 Bogomoletz Str., 01024 Kyiv, Ukraine
#
Institute of Organic Chemistry of NAS Ukraine, 5 Murmanska Str., 02660 Kyiv, Ukraine
§
State Key Laboratory for Molecular and Cellular Biology, 4 Bogomoletz Str., 01024 Kyiv,
Ukraine
⊥ChemBio Center, Taras Shevchenko University of Kyiv, 67 Chervonotkatska Str., 02094 Kyiv,
Ukraine
Acid sensing ion channels 1a (ASIC1a) are of crucial importance in numerous physiological and
pathological processes in the brain. Here we demonstrate that novel 2ꢀoxoꢀ2Hꢀchromeneꢀ3ꢀ
carboxamidine derivative 5b, designed with molecular modeling approach, inhibits ASIC1a
currents with an apparent IC of 27 nM when measured at pH 6.7. Acidification to 5.0 decreases
5
0
the inhibition efficacy by up to three orders of magnitude. The 5b molecule not only shifts pHꢀ
dependence of ASIC1a activation, but also inhibits its maximal evoked response. These findings
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 51
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
suggest that the compound 5b binds to pH sensor of ASIC1a acting as orthosteric nonꢀ
competitive antagonist. At 100 nM compound 5b completely inhibits induction of long term
potentiation (LTP) in CA3ꢀCA1, but not in MFꢀCA3 synapses. These findings support the
knockout data indicating at the crucial modulatory role of ASIC1a channels in the NMDARꢀ
dependent LTP and introduce novel type of ASIC1a antagonists
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
INTRODUCTION
ASIC1a is one of six acidꢀsensitive ion channels cloned so far from the mammalian nervous
1ꢀ3
system . ASIC1a is abundantly expressed in mammalian CNS. It is activated by extracellular
2+
acidification and, in contrast to the other ASICs family members, participates in Ca signaling:
4
, 5
it conducts sodium and calcium ions . In most studies, ASIC1a starts its activity at pH 6.9; the
+
2,
amplitude of response grows with the increase of H concentrations up to the saturating pH 6.0
6
.
It has been shown that in the neurons acutely dissociated from amygdala, ASICs containing
+
7
ASIC1a subunit are even more sensitive to H : they start opening at pH 7.2 .
+
Thus, ASIC1a is an exquisitely sensitive H receptor. This hypersensitivity is crucial for its
numerous putative functions in the CNS. It is suggested that higher local concentrations of
8
, 9
protons may transiently activate ASICs during normal synaptic activity in the brain . There is
circumstantial evidence connecting ASICs with certain physiological functions including
1
, 8, 9
learning, memory, fear sensing, etc.
. Genetic knockout studies show that postsynaptic
ASICs are critical for hippocampal CA1ꢀCA3 pathway LTP facilitating NMDA receptor
9
10
function , but recent data leads to the opposite conclusion . Numerous pathological states
including epilepsy, multiple sclerosis, ischemic disorders, traumatic brain injury, depression,
spinal cord injury, inflammatory pain and headache are accompanied with tissue acidosis
ACS Paragon Plus Environment
2

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1, 3, 11, 12
presumably activating ASIC1a
. It has been shown that pharmacological blockage of
ASIC1a by amiloride or specific natural blocker PcTx1, as well as its genetic deletion or down
regulation affect the above mentioned pathological states proving ASIC1a to be a promising
molecular target against such disorders. Moreover, precise role of ASICs in numerous
physiological processes such as learning, memory, fear sensing, etc. is still underestimated due to
the lack of potent and selective small molecule inhibitors. During the last decade several
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
3ꢀ17
attempts were made to solve this problem (see
for reference).
At the moment all of the described ASIC small molecule inhibitors can be divided into three
main classes: amiloride and its derivatives, (het)arylamidines and nonꢀsteroidal antiꢀ
3
, 18, 19
inflammatory drugs (NSAIDs)
. Wellꢀknown potassiumꢀsparing diuretic amiloride which
2
0
was the first small molecule found to block ASICs
is found to be nonꢀselective among
21ꢀ23
different proteins. It blocks the pore of degenerine channel family (ENaC, including ASIC)
2
4
25
as well as the pore of cyclic GMPꢀgated cation channel , sodiumꢀhydrogen antiporter 1 , Tꢀ
2
6
27
type calcium channel and urokinase plasminogen activator , etc. The most potent amiloride
16
derivative inhibits ASIC channels with apparent IC50 of 0.49 µM . (Het)arylamidines family
comprises first nonꢀamiloride inhibitor Aꢀ317567 together with analogs developed by Dubé and
1
3, 28
29
colleagues
, serine protease inhibitor nafamostate used clinically as anticoagulant
,
1
5
fluorescent stain DAPI with other indoleamidines synthesized at Merck
and finally DNAꢀ
intercalating bisarylamidines like antiꢀprotozoal drug diminazene or fluorescent dye
17, 30
hydroxystilbamidine
. Recently, a series of benzothiophenemethylamines and other
compounds derived with the help of fragment screening approach were identified as the first
1
4, 31
nonꢀamidine chemotypes of ASIC inhibitors
. NSAIDs such as salicylic acid, aspirin,
32
diclofenac and ibuprofen directly inhibit ASICs in nociceptors but with relatively low potency .
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 51
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
33, 34
Local anesthetics lidocaine and tetracaine are also able to inhibit ASIC currents
in low mM
concentrations. Aforementioned small molecule inhibitors were identified through empirical
exploratory approach and their binding sites are yet to be validated. However, for the molecules
that resemble structural features of amiloride one might suspect a similar binding mechanism. At
the moment, the pharmacology of ASICs still lacks selective and potent inhibitors. New hopes
arise with the transition to crystallographic studies of the molecular structure of ASICs. It started
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
3
5
with pioneering work carried out by Gouaux lab .
Crystal structures of chicken ASIC1 (cASIC1) in its different functional states have recently
been published revealing structural organization of the homotrimeric channel and conformational
35ꢀ39
rearrangements associated with its function
. An extracellular domain of ASIC has C3ꢀ
symmetry and controls the channel permeability (Figure S1, A). Seven distinct structural
domains are arranged in upright forearm and clenched hand manner with only two
transmembrane domains, TM1 and TM2, whereas palm, finger, knuckle, βꢀball and thumb
domains belong to the extracellular part of the ASIC (Figure S1, B and D). The acidic pocket,
also referred as “pH sensor”, is formed at the interface of the two ASIC monomers and due to its
C3ꢀsymmetry, three cavities are found per one ion channel (Figure S1, D). Eleven negatively
charged residues Glu98, Glu220, Glu236, Asp238, Glu239, Glu243, Asp260, Asp346, Asp350,
Asp408 and Glu354 confer very high electroꢀnegative potential to the pocket (Figure S1, E).
They comprise also three carboxylic dyads Glu220/Asp408, Asp238/Asp350 and
Glu239/Asp346. Most of these residues are conserved throughout the ASICs family and may be
1
, 35
crucial for the ASIC sensitivity to protons
. The pH sensor has a flaskꢀlike shape with the
narrow entry expanding inside. From structural domain organization point of view, the thumb,
ACS Paragon Plus Environment
4

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the βꢀball and the finger domains from one ASIC subunit and the palm domain from adjacent
3
5
subunit contribute residues to form the acidic pocket .
Along with apo form, complexes of cASIC1 with PcTx1, a highly potent toxin (IC ~1nM) ,
4
0
5
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
37, 39
were solved at different pH levels
. It both cases PcTx1 was found to bind every one of the
three pH sensors of the cASIC1. Hydrophobic patch of PcTx1 forms extensive van der Waals
2
6
28
contacts with 5ꢀhelix of the thumb domain, while its loop II bearing RRR motif extends into
the acidic pocket binds to palm, finger and bꢀball domains forming a number of salt bridges and
3
7, 39
hydrogen bonds
. Conformational rearrangements of structural domains associated with ASIC
41
37
function or the toxin binding involves residues from the acidic pocket and its immediate
proximity but from various domains. These properties of the pH sensor let us hypothesize that
conformational freedom of the pH sensor is a crucial factor in ASIC function and adjusting
inherent flexibility of the pocket can modulate the channel activity.
Here we report on design, synthesis and electrophysiological evaluation of novel 2ꢀoxoꢀ2Hꢀ
chromeneꢀ3ꢀcarboxamidine blockers aimed to target the pH sensor site of ASIC1a. Most potent
novel compound inhibits hASIC1a currents in nanomolar concentrations with profound
dependence of blocking efficacy on the activating pH representing a novel type of ASIC
antagonists. We demonstrate that pharmacological inhibition of ASICs by this small molecule
prevents LTP induction in hippocampal CA3ꢀCA1 pathway thus supporting important
physiological role of ASICs in the plasticity of mammalian brain.
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 51
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
RESULTS
Chemistry
Compounds 1-5 were synthesized (Scheme 1) starting from N,N'ꢀdialkylꢀ3ꢀaminophenoles 7a-
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
2
f obtained in turn by alkylation of known 3ꢀaminophenoles 6a-c . Subsequent VilsmeierꢀHaack
formylation furnished corresponding aldehydes 8a-f which were converted into target 2ꢀoxoꢀ2Hꢀ
chromeneꢀ3ꢀcarboximidamides 1-5 (as monoꢀ or dihydrochloride salts) by cyclocondensation
4
3
with ethyl cyanoacetate using procedure developed by Sakurai et al .
Electrophysiological evaluation
The potency of compounds were evaluated on hASIC1a channels endogenously expressed in
4
4
HEK 293 cells using standard patchꢀclamp technique . In response to pH drop from 7.4 to 5.0
appox. 80% of the cells clamped at ꢀ100 mV gated ionic current ranging from 200 pA to 1 nA
45
which is mediated by homomeric hASIC1a channels .
Development of pH sensor binding antagonists of ASIC1a.
3
5
Crystal structure of the cASIC1 allowed us to perform a rational structure based design of
small molecule antagonists. Homology modeling was used to construct the spatial model of
human ASIC1a (hASIC1a). High overall sequence identity (> 90%) and particularly in the pH
sensor (only two amino acids differ) between human and chicken ASIC1 allowed us to obtain a
model of the hASIC1a with a virtual Xꢀray level of confidence.
We have performed unbiased docking of small molecule diversity set of 10,000 compounds
into the pH sensor saving 5 poses per ligand, as initial attempt. Whole pH sensor cavity was
treated as the binding site. Processing of the generated poses did not reveal any clear preference
neither toward binding mode nor towards a particular chemical scaffold. In the vast majority of
ACS Paragon Plus Environment
6

Page 7 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
generated poses small molecules bound alongside of the acidic pocket wall, that is similar to the
binding to a flat protein surface. Electrophysiological evaluation of the top 10 molecules did not
reveal the molecules active at concentration as high as 100 µM.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
At the next step, we tried to mimic PcTx1 in the acidic pocket with small molecules that would
37, 39
resemble its key contacts. The guanidine group of PcTx1 Arg27 in the channelꢀtoxin comlex
is found to be in tight environment of carboxyl groups of cASIC1 Glu220/Asp408 dyad and
oxygens of cASIC1 Thr215, Gly216 and Gly218. Side chains of Arg27 of toxin forms van der
Waals contacts with Phe174. Arg28 of PcTx1 binds Glu243 and Phe242 side chains of ASIC1
forming salt bridge and cationꢀπ interactions, respectively. Finally, residue Arg26 forms
3
7, 39
26
28
hydrogen bond with Asp350
. The RRR motif efficiently spans through palm, finger and
thumb domains pointing to the triꢀstar topology of putative antagonist, which would be able to
fill acidic pocket and reach its distal domains (Figure 1). To test this hypothesis we conducted
iterative structure guided design of a new blocker that would efficiently bind amino acids from
palm, finger and thumb domains within the acidic pocket.
46
We used recently published 2ꢀoxoꢀ2Hꢀchromeneꢀ3ꢀcarboxamidine scaffold as template for
the design of novel putative ASIC antagonists possessing the mentioned above triꢀstar topology.
First 2ꢀoxoꢀ2Hꢀchromeneꢀ3ꢀcarboxamidine derivative, compound 1, inhibited hASIC1a current
to 41.0±3.6 % (n=4) at 100 ꢁM (Table 1). Binding modes of the compound 1 evaluated with
docking showed little preference to a particular position within the acidic pocket, however we
identified frequently visited sites that might denote preferable interactions. Most favorably, the
amidine group bounds to the two cavities. One of them is formed by residues from the palm
domain only (Thr214ꢀGlu219, His173ꢀGly176 and Glu355), whereas the residues Glu97,
Asp259, Try191 and Asp237ꢀSer240 contribute to the other one.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 51
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
Next, we introduced flexible tertiary amino group (compound 2) to improve anchoring of the
putative antagonist to the charged groups in the acidic pocket. Surprisingly, the compound 2 had
even weaker blocking activity than the compound 1 possibly due to the large desolvatation
penalty that unfavorably contributes to the free energy of binding. In contrast, introduction of the
benzyl group at 7ꢀamino position (compound 3) had positive impact on antagonizing potency
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
(36.6±5.6%, n=4, similar to observed with the compound 1) (Table 1). To define the possible site
4
7
of benzyl group binding we estimated hydrophobic sites in the acidic pocket using SiteMap .
There are only two regions in the pH sensor which exhibit hydrophobic propensity. The
hydrophobic region 1 (HR1) is located between residues Thr214 and Gly176 of the palm domain
and is formed by backbone atoms of corresponding residues. Another region with hydrophobic
propensity (HR2) was attributed to the cavity formed by nonꢀcharged atoms of residues Glu235ꢀ
Phe241. Docking of the compound 3 revealed that its benzyl group binds to the HR2 and forms
van der Waals contacts with Cβ of the Glu235 and Cζ of Phe241. However, the inhibition
efficacy of the compound 3 was found to be virtually the same as for the compound 1(see Table
1). In order to bind both apolar cavities we synthesized the compound 4 with two benzyl groups
at the 7ꢀ position of the coumarin core. Unfortunately, this compound was insoluble suggesting
that the second substituent at nitrogen atom position 7 should preferably contribute to efficient
polar interactions. Docking of the compound 5a bearing both tertiary amine fragment and benzyl
substituent to the acidic pocket revealed preference toward a binding pose where the coumarin
moiety binds to HR1 cavity with amidine group forming extensive hydrogen bonds with Thr214,
Glu219 and Asp408, similarly to what was observed with compound 1. Carbonyl oxygen forms
an additional hydrogen bond with NH of Gly176 and van der Waals contacts with imidazole of
His173. Positively charged tertiary nitrogen of the compound 5a binds carboxylates of the
ACS Paragon Plus Environment
8

Page 9 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Asp237/Asp351 dyad and benzyl group points towards the second hydrophobic site (Glu235ꢀ
Phe241), similar to the compound 3. In this mode, both hydrophobic patches of the acidic pocket
were addressed as well as the carboxyl dyad Asp237/Asp351. This, along with the better filling
of the pocket could lead to the improved interaction with ASIC1a.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The derivative 5a caused substantial inhibition of hASIC1a current down to 28.2±4.2% and
6
5.7±4.7 % at 100 ꢁM and 10 ꢁM correspondingly (Table 1). Lowering concentration of 5a to
1
00 nM resulted only in negligible inhibition (96.1±3.2%, n=12) of hASIC1a currents (Table 1).
Its activity is comparable with other reported small molecule ASIC1a blockers (such as
amiloride, nafamostat, Aꢀ317567, etc). It should be noted, that all the observed blocking effects
were fully reversible.
Careful evaluation of the above binding mode revealed that the benzyl moiety of the
compound 5a does not fill the hydrophobic cavity efficiently and the extra space is available.
Shortest distance between backbone heavy atoms of the Phe241 and carbon in para position of
the benzyl group of 5a is 5.31 Ǻ indicating that a halogen atom may fit this space. Introduction
of the chlorine atom in the para position of the benzyl group (compound 5b, Figure 2) further
improved the blocking potency of the molecule (44.8±4.0 %, n=9 at 10 ꢁM for 5b vs.
6
5.7±4.7 %, n=10 for 5a). The observed increase in the efficacy supports the suggested binding
mode of the compound 5b.
We also have tested the potency of compound 5b to inhibit native ASIC currents on
hippocampal neurons of rat and found that 10 ꢁM of the compound inhibited rASIC1aꢀlike down
to 48.3±5.0 % (n=4). The experiments on DRG neurons of rat revealed that compound 5b
induces virtually the same inhibition of fast rASIC3ꢀlike currents (45.2±4.3 %, n=4) but not of
the slow rASIC2ꢀlike currents (99.3±1.3 %, n=3).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 51
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
Potency of the 2-oxo-2H-chromene-3-carboxamidine derivatives depends on
the acting pH level
All the 2ꢀoxoꢀ2Hꢀchromeneꢀ3ꢀcarboxamidine derivatives have been designed to target the
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
1
, 35
pH sensor of ASIC1a, the most sensitive to protons place in the receptor
. One should expect
that the binding efficacy of the described compounds depends on pH. In order to check this
hypothesis, all compounds derived in this study were evaluated at pH 6.7 and pH 5. As expected,
a profound difference was observed between these sets of data. The ionic current elicited by a
shift to pH 6.7 had much slower activation, smaller amplitude and slower desensitization onset as
compared to the current at pH 5.0. This current was notably inhibited by 100 nM of
compound 5a (68.9±4.8, n=10, Table 1) as compared to insignificant (96.1±3.2%; n=12)
inhibition observed at pH 5.0. Compound 5b was found to be even more potent at pH 6.7
(
100 nM, 23.7±4.0% n=14), whereas at pH 5.0 this concentration was virtually inactive
98.6%±2.4, n=7). Pronounced pHꢀdependence of its inhibition in compare with classical ASIC
(
pore blocker amiloride is demonstrated on Figure 1 (AB for 5b and CD for amiloride).
Correspondingly, a profound leftward shift of the doseꢀresponse relationship for the
compound 5b was observed when ASIC1a currents were activated by pH 6.7 as compared to
pH 5.0 (Figure 1, E, IC =27.35±1.23 nM, n=7÷9 vs. 7.34±0.84ꢁM, n=8÷10 at pH 6.7 and 5.0
5
0
correspondingly). As well as in all previous cases, inhibition by the compound 5b was fully
reversed within a few minutes of washout.
ASIC1a is antagonized by 5b through an orthosteric non-competitive
mechanism
A few decades ago, it was shown that amiloride produces the voltageꢀdependent nonꢀ
2
0
competitive blockage of ASIC currents entering, in all probability the pore of the channel . We
ACS Paragon Plus Environment
10

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
have evaluated the potency of the compound 5b vs amiloride at different membrane voltages.
Amiloride blocked ASIC1a current to 41.34±2.42% (n=7) at ꢀ100 mV, while changing holding
voltage to ꢀ20 mV decreased this blockade to 60.70±3.23% (n=6, p=0.0006). However,
compound 5b produced similar inhibition at both these voltages, 42.12±2.26 (n=8) and
45.78±3.13 correspondingly (n=9) indicating the absence of voltageꢀdependence of blocking
action (Figure 1, H, p=0.92). Current/voltage relationships for amiloride and the compound 5b
show that blocking potency of amiloride is decreased with depolarization of cellular membrane,
whereas the potency of the compound 5b remains unaltered (Figure 1, G). These data suggest
that the location of the binding site for 5b is at extracellular domains of the channel, but not
within the cellular membrane.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 5b affects steadyꢀstate activation of ASIC1a current (Figure 1, F). In the
concentration of 10 nM, it shifted the doseꢀresponse curve to a lower pH range
(
EC =6.57±0.01 pH, n=7÷8 in control and EC =6.50±0.01 pH, n=8÷10 in the presence of
50 50
10 nM of compound 5b). It is worth noting that at saturating pH a small inhibition of maximal
evoked response was observed already at this small concentration of 5b (93.18±1.04% of
control). At higher concentrations of this compound the rightward shift of the doseꢀresponse
curve became more significant (EC =6.40±0.03 pH, n=6÷9 at 100 nM and EC =6.18±0.05 pH,
5
0
50
n=5÷8 at 1 ꢁM) and the maximal evoked response was markedly reduced to 83.94±4.25% at
100 nM and 72.19±5.21% at 1 ꢁM. The doseꢀresponse shift accompanied with maximal evoked
response inhibition is characteristic for orthosteric nonꢀcompetitive or negative allosteric
4
8
modulation mechanism .
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 51
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
Compound 5b affects long-term synaptic plasticity in hippocampus
Previous reports indicate that ASICs can play a substantial role in synaptic plasticity, learning
9
and memory. . We have examined the effect of rASIC1a blockade on CA3ꢀCA1 pathway. First,
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
we evaluated the potency of the compound 5b to inhibit rASIC1a channels in isolated pyramidal
hippocampal neurons of rats. At 100 nM compound 5b induced substantial inhibition of
rASIC1aꢀlike currents down to 9.39±2.9 % (n=4, p < 0.0001), whereas at 1 µM it did not block
the main ionic currents involved in synaptic transmission (Figure S2). Application of the
compound 5b (100 nM) did not affect the slope and amplitude of field excitatory postsynaptic
potential (fEPSP) at CA3ꢀCA1 synapses (103.4±3.4 %, p=0.4 for the slope and 104.6±3.0 %,
p=0.2 for the amplitude correspondingly, n=8). Moreover, tenꢀfold higher concentration of
compound 5b did not result in the appearance of any feasible inhibition of amplitude for both
inhibitory and excitatory postsynaptic currents (Figure S3). Immediately following highꢀ
frequency stimulation 5b–treated as well as control slices showed the increase in fEPSP slope
and amplitude (shortꢀterm potentiation, STP). The degree of STP in 5bꢀtreated and control slices
was not significantly different (193.2±12.3%, n=9 vs 225.9±12.4%, n=8) correspondingly,
p=0.08). As shown in Figure 2C, application of the compound 5b occluded LTP in CA3ꢀCA1
synapses (142.9±2.5%, n=9 in control vs 106.2 ± 5.2 %, n=9 in 5bꢀtreated slices, p<0.0001).
These data are in agreement with previous studies suggesting that activation of postsynaptic
ASICs promotes the membrane depolarization and thereby facilitates NMDA receptor function
9
, 49
and contributes to LTP . With this hypothesis in mind, we examined the effect of compound
b on NMDARꢀindependent LTP. The synapses formed by the mossy fibers (MF) on CA3
neurons exhibit a form of experienceꢀdependent synaptic plasticity that is induced and expressed
5
5
0
presynaptically and does not depend on the activation of NMDA receptors . To study the effect
ACS Paragon Plus Environment
12

Page 13 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of the compound 5b on LTP in MFꢀCA3 pathway we used a modified protocol for LTP
induction (4 bursts of 100 stimuli delivered at 100 Hz with a 2 min intervals). All experiments
were performed in the presence of NMDA receptor blocker MKꢀ801. Figure 2D represents group
data for fEPSPs evoked by stimulation of MF and recorded in CA3 region of hippocampus in
control and in 5bꢀtreated slices before and after delivery of tetanic stimulation. Repeatedꢀ
measures ANOVA revealed that there was no significant difference in the LTP level between
control and 5bꢀtreated slices (118.3±9.6% [n=10] vs. 117.0±11.4% [n=12], p=0.98) supporting
the idea that ASIC1a specifically contributes to the NMDARꢀdependent plasticity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DISCUSSION
+
The ion channel activated by extracellular H was found first in mammalian sensory neurons
5
1ꢀ53
2,
and initially was associated with nociception
. However, after cloning the family of ASICs
5
4ꢀ57
it became apparent that these channels play very diverse roles in much broader repertoire of
1
, 12, 58ꢀ61
2
physiological mechanisms and pathological conditions
. Homotrimeric ASIC1a channel
is a specific brain sensor for protons considered to be the key participant in learning, memory
and fear sensing as well as an important player in different pathological states, such as epilepsy,
multiple sclerosis, ischemic stroke, traumatic brain injury, depression, spinal cord injury,
1
, 3, 7ꢀ9, 11, 12, 58
inflammation and headache
. Hence, the involvement of ASICs in so wide range of
disorders urges a development of ASIC blockers as therapeutic agents.
Compound screening remains the main source of new blockers of ASICs. However, low
throughput of such screening largely limits the progress. The main efforts of medicinal chemistry
13ꢀ17, 28, 31, 46
are focused around amiloride and (het)arylamidine derivatives
. As a result, current
small molecule ASIC inhibitors are largely a combination of amidine or guanidine groups with
aromatic moieties, while the bioisosteric replacement strategy may result in new heterocyclic
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 51
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
31, 46
scaffolds
. Largest caveat of the discovered compounds is their promiscuity among targets
1
4
other than ion channel. This imposes substantial hurdle on the further development .
To our knowledge, there are still no studies utilizing the structural information on the ASIC
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
35ꢀ39
despite recent advances in solving Xꢀray of ASIC1a
. The present study has been
commenced upon disclosure of the first structure of ASIC – chicken ASIC1 in the desensitized
3
5
form . We decided to focus our efforts on the acidic pocket located in the extracellular domain
of ASIC1a. Its three features favored our choice. First, the crystal structure of ASIC1 revealed
that 4 out of 5 structural domains located above cellular membrane contribute to the acidic
pocket suggesting that its flexibility may be important for the channel function. Second, the
pocket is heavily saturated with negatively charged residues; some of them are invariant and
4
0
37, 38
crucial for pH sensitivity. Third, highly potent toxin, PcTx1 binds the pH sensor
. From
these facts we hypothesized that altering conformational perturbations of the pocket may impair
function of the whole ion channel. In silico assessment of the small molecule compound library
against the acidic pocket resulted in very diverse binding orientations of the compounds that
reflects large discrepancy in sizes of ligands and the pocket. Patch clamp validation of top scored
10 compounds returned no active molecules. Design of specific compounds directed to fill
volumetric pH sensor would be preferable strategy then. To address size we analyzed binding
2
6
28
37, 39
mode of PcTx1 to the ASIC1a, especially its pH sensor buried part – RRR motif
. We
choose the triꢀstar shape as the key feature required by a small molecule to take advantage of the
roomy acidic pocket.
The antagonist design was based on 7ꢀaminoꢀ2ꢀoxoꢀ2Hꢀchromeneꢀ3ꢀcarboxamidine core, since
its amino group could be functionalized with two substituents providing a origin for divergent
nonꢀlinear structures. Compound 1 showed a weak blocking activity, yet comparable to the
ACS Paragon Plus Environment
14

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
29
32
33, 34
nafamostate metabolite , NSAIDs
and local anesthetics
. Similarly to amiloride,
compound 1 was too small to possess a single preferred position in the acidic pocket and
modeling of the feasible poses resulted in diverse variants guided by binding of amidine group to
carboxylates. Introduction of the flexible polar (compound 2) and aromatic (compound 3) groups
was aimed at anchoring the small molecule with specific interactions. Electrophysiological
screening of derivatives 2 and 3 at pH 5.0 showed a clear preference of hydrophobic substituent
over polar. The acidic pocket of ASIC1a is depleted in lipophilic residues so that the in silico
assessment of plausible binding of the compound 3 reveals reduced number of poses comparing
compound 2. Primary binding site for benzyl moiety was identified as hydrophobic cavity
centered at phenyl ring of Phe241 from the finger domain.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
All the compounds 1ꢀ3 did not have two bulky substituents at the position 7 that would result
in the efficient filling of the acidic pocket volume. Our attempt to introduce a second benzyl
group to coumarin unfortunately led to the insoluble compound 4. In order to balance compound
potency with itssolubility we combined compounds 2 and 3 resulting in the compound 5a.
Substantial potency increase suggested that bulky groups are beneficial for the putative ASIC1a
antagonist. The difference is even more striking since the compound 2 is largely depleted in
activity. This fact can be rationalized by the predicted binding mode. The compound 5a binds
both hydrophobic regions HR1 and HR2 with its aromatic moieties, deep in the acidic pocket (
Figure ). The remaining aminoꢀarm is pointed outwards to interact with Asp237/Asp351 dyad
or Glu355. In general, charged interactions are beneficial for the free energy of binding, when
6
2
they are prepositioned to each other . Final compound 5b was designed to enhance interaction
with residue Phe241 and its improved activity supports the suggested binding mode: the
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 51
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
compound is presumed to interconnect palm, finger and thumb domains of the ASIC through
coumarin core, benzyl and amino moieties, respectively.
Conformational rearrangements of ASIC structural domains are crucial for all transitions
between open, closed and desensitized states. Mutation of charged residues located in the
proximity to the acidic pocket and involved into interꢀdomain contacts led to the substantial
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
3
37
change in pH sensitivity . Hill coefficient in mutants Asp350Asn (chicken analog of Asp351)
6
3
and Glu355Gln
was substantially lower than in wild type ASICs. Both these residues are
located at the pocket entry and are targeted by tertiary amine moiety of the compound 5b.
Phenylalanine at the position 241 is absolutely invariant among all ASIC species highlighting
functional importance of the structural motif. Residues Glu235 and Glu355 are largely involved
41
in the close ꢀ open transition and binding to these residues might substantially interfere with
this transformation. Interaction of the amidine group with the first hydrophobic cavity is
facilitated by the network of hydrogen bonds and van der Waals contacts. Four hydrogen bonds
formed by amidine group and lactam oxygen with Thr214, Glu219 and Gly176 are enhanced
with van der Waals contacts and π ꢀ π interaction of coumarin core with imidazole ring of
His173. Position of His173 is variable between ASIC subtypes and can be used to gain
specificity between them. For instance, in human ASIC3 arginine is found in equivalent position
and its flexibility may interfere with such binding mode of the compound 5b.
Recently published structure of the complex of chicken ASIC1a with PcTx1 revealed that
hydrophobic regions HR1 and HR2 and carboxyl dyad on the αꢀhelix from the thumb domain are
3
7, 39
both addressed by the toxin
. Orientation of the coumarin core is very similar to the
conformation of the side chain of Arg27 (Figure S4). Side chain of Arg27 extends into the acidic
pocket along palm domain residues with guanidine group binds to the HR1 in the position nearly
ACS Paragon Plus Environment
16

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
identical to the amidine group of the compound 5b. Carbon atoms of the Arg27 side chain also
form van der Waals contacts with Phe174 (numbering of chicken ASIC1). In both crystals,
Arg28 binds Glu243 and Phe242 side chains forming salt bridge and cationꢀπ interactions In case
of 5b paraꢀchlorobenzyl enhanced van der Waals contacts with Phe241. Finally, residue Apr26
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
7, 39
forms hydrogen bond with Asp350
similar to the tertiary nitrogen of the compound 5b.
PcTx1 form tight interactions with residues from thumb, palm and finger domains, altering
flexibility of the channel molecule. Such overlap in the predicted binding mode for the
compound 5b and crystal structure of the channelꢀtoxin complex is not surprising since 2ꢀoxoꢀ
2
Hꢀchromeneꢀ3ꢀcarboxamidine derivatives developed in this study were designed to occupy the
same place with PcTx1 in pHꢀsensor of ASIC1a channel and therefore antagonize its activity via
3
7, 38
orthosteric mechanism. As it was shown long before the exploration of PcTx1 binding mode
,
this toxin, when bound to the channel, slightly shifted the activation and desensitization curves to
+
64
+
lower H concentrations due to increase in the apparent affinity to H . Later it was shown that
the interaction of PcTx1 with ASIC1 channel is stateꢀdependent: PcTx1 binds more tightly to the
open state of the ASIC1b channel than to the closed and desensitized ones and facilitates
openings of the channel. In the case of ASIC1a, it binds most tightly to the open and the
65
desensitized state, promoting desensitization . Therefore, a molecule bound to the pH sensor
+
may well modify ASIC1a channel sensitivity to H as well as its gating mechanism. Here we
show, that the compound affinity increases with the change in activating pH from 5.0 to
+
2+
6
.7(Figure 1E). Thus, it looks like the compound competes with H for the binding site like Ca
6
ions . At any concentration, compound 5b promotes both rightward shift of doseꢀresponse curve
and reduction of maximal evoked response, showing a behavior typical for negative allosteric
modulator or orthosteric nonꢀcompetitive antagonist (Fig. 2F, doseꢀresponse curves performed in
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 51
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 absence or presence of 5b). 2ꢀOxoꢀ2Hꢀchromeneꢀ3ꢀcarboxamidine derivatives were designed
+
to target the pH sensor of ASIC1a. We demonstrate that the affinity to H is modulated by the
compound 5b (Figure 1E) as well as the affinity of the compound 5b depends on activating pH
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
+
(
Figure 1F), suggesting that the molecule competes with H for the pH sensor. These data serve
in fact as a proof of a concept: novel compounds interact with pH sensor. However, the observed
relationships do not fit competitive inhibition mechanism. The latter requires that antagonist
bound to receptors must dissociate quickly enough to be substituted by the agonist presented at
48
the receptor compartment . In the other case the antagonist will occupy an inordinately high
percentage of the receptors and antagonism will dominate. If the percentage of occupied
4
8
receptors is high enough, the agonist does not produce a maximal response . This is just the
case of compound 5b inhibition. We speculate that paraꢀchlorobenzyl group of 5b providing rich
van der Waals contacts with Phe241 in the acidic pocket of ASIC1a greatly decelerates
dissociation of 5b from receptor resulting in observed orthosteric nonꢀcompetitive inhibition.
Thus, the compound 5b is most effective at slight acidifications. Various pathological states,
such as ischemic stroke, epilepsy, inflammation, etc. are characterized just by mild acidification
of the tissue. Such minor drop in extracellular pH from 7.4 to 6.8 is sufficient for significant
6
6
membrane depolarization accompanied by the trains of action potentials . The activity of
1
, 3, 11, 12
ASIC1a channels induced by different pathological states leads to neuronal death
.
+
49
Recently it has been suggested that H acts as neurotransmitter in amygdala ; the novel ASIC
antagonists will help to further elucidate the role of ASICs in neurotransmission.
In conclusion, we have identified a plausible structural motif for the small molecules acting as
orthosteric nonꢀcompetitive antagonists of ASICs. Mutagenesis and deeper SAR studies will be
needed to further improve the ASIC blockers, but even the preliminary modeling helps to get
ACS Paragon Plus Environment
18

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
early insight on how compound 5b prevents the ASIC1a from sensing protons. To the best of our
knowledge, there are no small molecules with reported pHꢀsensor binding.
Endogenous ASIC1a has primarily somatodendritic localization and is particularly enriched in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
67
brain synaptosomes and dendritic spines suggesting that ASIC1a is presented in synapses .
Activation of ASIC1a should modulate synaptic responses and influence synaptic plasticity. It
was shown that genetic disruption of ASIC impaired longꢀterm synaptic plasticity in CA1 region
9
49
of hippocampus and lateral amygdala , although another study showed that LTP can be
1
0
generated in ASIC1a knockout mice . These discrepancies may be due to the differences in
strains and genetic deletion approaches. In our experiments the compound 5b did not affect
baseline fEPSP, shortꢀterm plasticity and NMDARꢀindependent LTP induced in MFꢀCA3
pathway but significantly impaired NMDARꢀdependent LTP in CA1ꢀCA3 pathway. Recent
study indicates that presynaptic stimulation reduces extracellular pH and activates postsynaptic
4
9
ASICs in the lateral amygdala synapses . Activation of postsynaptic ASICs may lead to sodium
influx into the postsynaptic terminal and depolarization of the postsynaptic membrane, which
can be sensed in its turn by NMDARs in the neurons, and thus potentiate their plasticity
9
, 68
function
. The enhanced NMDAR response can facilitate the induction of longꢀterm
potentiation. The hypothesis that ASICs are required for the induction of NMDARꢀdependent
longꢀterm potentiation was supported by two independent investigations using ASIC knockout
mice in hippocampal formation and amygdala and by the fact that lowering magnesium can
9, 49
rescue LTP in the conditions of genetic ASIC ablation
. Our study further supports this
assumption indicating that ASIC blockade selectively eliminates NMDARꢀdependent LTP in
CA3ꢀCA1 synapses and does not affect NMDARꢀindependent LTP in MFꢀCA3 pathway.
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 51
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
CONCLUSION
Numerous pathological states and diseases including epilepsy, multiple sclerosis, ischemic
disorders, traumatic brain injury, depression, spinal cord injury, inflammatory pain and headache
are accompanied with tissue acidosis presumably activating ASIC1a. It has been shown that
pharmacological blockage of ASIC1a as well as its genetic deletion or down regulation affects
the above mentioned pathological states proving ASIC1a to be a promising molecular target
against such disorders. Here we present a novel 2ꢀoxoꢀ2Hꢀchromeneꢀ3ꢀcarboxamidine derivative
compound 5b which was designed by using molecular modeling approach to target pHꢀsensor of
ASIC1a channel. 5b demonstrates its maximal potency in ASIC1a inhibition (nM range) at
physiologically attainable levels of pH. Our data indicate that compound 5b acts as orthosteric
nonꢀcompetitive antagonist of ASIC1a. Recent data on the role of ASIC1a in LTP are
contradictory. The knockꢀout data show that ASIC1a has crucial role for NMDAꢀdependent LTP
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
9
induction in hippocampus and learning ability . However, other groups just by using different
1
0
knockꢀout approach came to a contradictory conclusion . We found that novel orthosteric
ASIC1a antagonist suppresses NMDAꢀdependent LTP and does not alter the NMDAꢀ
independent one. Our findings gave a new tool for determining the physiological and
pathological role of ASIC1a channel. The suggested chemical scaffold can be used for the
development of novel potent ASIC1a antagonists that can be used for the treatment of numerous
brain diseases and pathologies.
ACS Paragon Plus Environment
20

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
General methods and procedures.
All chemicals were obtained from commercially available sources and used without further
purification (SigmaꢀAldrich, “Enamine LTD”). Solvents were purified and dried using standard
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
methods. H NMR spectra were recorded on a Varian Mercuryꢀ400 (400 MHz) instrument with
1
3
TMS as internal standard. С NMR spectra were obtained on a Bruker Avance DRXꢀ500
spectrometer (125.75 MHz) with TMS as internal standard. The data are being reported as s =
singlet, br s = broad singlet, d = doublet, t = triplet, q = quartet and m = multiplet or unresolved,
chemical shifts in ppm and coupling constant(s) in Hz. LC/MS spectra were recorded using
chromatography/mass spectrometric system that consists of highꢀperformance liquid
chromatograph “Agilent 1100 Series” equipped with diodeꢀmatrix and massꢀselective detector
“
Agilent LC MSD SL”. Elemental analysis was performed in the Microanalytical Laboratory of
the Institute of Organic Chemistry, National Academy of Sciences of Ukraine. Compounds 6a,b,
7
a and 8a are commercially available. Purity of all final compounds was 95% or higher.
All biological evaluation procedures were performed in accordance with the guidelines set by
the National Institutes of Health for the humane treatment of animals and approved by the
Animal Care Committee of Bogomoletz Institute of Physiology.
Procedure a. Compounds 7c and 7d.
A mixture of 3ꢀaminophenol 6a,b (0.01 mol), benzyl bromide (0.01 mol or 0.02 mol) and
sodium bicarbonate (3.36g, 0.04 mol) in acetonitrile (150 mL) was stirred at 40 °C for 6 h. After
cooling to room temperature, the reaction mixture was filtered and the filtrate evaporated under
reduced pressure. Water (100 mL) and ethyl acetate (100 mL) were added to the residue. Organic
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 51
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
layer was separated, washed with water, dried over Na SO and the solvent was removed under
2
4
reduced pressure. The product was used in the next step without further purification.
3-[Benzyl(methyl)amino]phenol (7c). Starting from 3ꢀ(methylamino)phenol 6b and benzyl
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
1
bromide (0.01 mol). Light brown oil. Yield 1.30 g (61%). H NMR (CDCl ,): δ = 2.98 (s, 3H,
3
CH ), 4.49 (s, 2Н, CH ), 6.22–6.30 (m, 2Н ), 6.34 (d, J = 8.0, 1H ), 7.05 (t, J = 8.0, 1H), 7.18–
3
2
.
.
7.39 (m, 5H).
3-(Dibenzylamino)phenol (7d). Starting from 3ꢀaminophenol 6a and benzyl bromide (0.02
1
mol). Light yellow solid. Yield 2.25 g (78%). H NMR (DMSOꢀd ): δ = 4.60 (s, 4Н, 2CH ),
6
2
6.01–6.13 (m, 3Н), 6.82 (t, J = 8.0, 1H), 7.10–7.42 (m, 10Н), 8.86 (s, 1Н, OH).
Procedure b. Compounds 7b,e,f.
A mixture of compound 6c (2.45 g, 0.01 mol) and alkyl halide (0.01 mol) in acetonitrile (150
mL) was stirred at 81 °C for 3 h. After cooling to room temperature, the reaction mixture was
evaporated under reduced pressure. Potassium carbonate aqueous solution (0.2 M, 100 mL) and
dichloromethane (100 mL) were added to the residue. Organic layer was separated, washed with
water, dried over Na SO and the solvent was removed under reduced pressure. The product was
2
4
used in the next step without further purification.
3-{[2-(Diethylamino)ethyl](ethyl)amino}phenol (7b). Using ethyl iodide as alkyl halide.
1
Brown oil. Yield 0.94 g (40%). H NMR (CDCl ): δ = 1.08–1.19 (m, 9H, 3CH ), 2.64–2.75 (m,
3
3
6
Н, 3CH ), 3.32 (q, J = 7.2, 2Н, CH ), 3.40 (t, J = 7.0, 2Н, CH ), 6.20–6.29 (m, 3Н), 7.03 (t, J =
2 2 2
8.0, 1H).
3-{Benzyl[2-(diethylamino)ethyl]amino}phenol (7e). Using benzyl chloride as alkyl halide.
1
Brown oil. Yield 1.43 g (48%). H NMR (CDCl ): δ = 1.01 (t, J = 7.1, 6H, 2CH ), 2.57–2.64 (m,
3
3
ACS Paragon Plus Environment
22

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
6
H, 3CH ), 3.50 (t, J = 7.1, 2H, CH ), 4.46 (s, 2H, CH ), 6.21–6.27 (m, 3H), 7.00–7.05 (m, 2H,
2
2
2
H), 7.19 (d, J = 7.0, 2H), 7.26 (d, J = 7.0, 2H).
3-{(4-Chlorobenzyl)[2-(diethylamino)ethyl]amino}phenol (7f). Using 4ꢀchlorobenzyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
chloride as alkyl halide. Brown oil. Yield 1.57 g (47%). H NMR (CDCl ): δ = 1.04 (t, J = 7.0,
3
6H), 2.58 (q, J = 7.0, 4H), 2.69 (t, J = 7.2, 2H), 3.51 (t, J = 7.2, 2H), 4.48 (s, 2H), 6.13 (s, 1H),
6
.22–6.27 (m, 2H), 7.02 (t, J = 8.5, 1H), 7.14 (d, J = 8.5, 2H), 7.28 (d, J = 8.5, 2H).
Procedure c. Compounds 8b-f.
To the stirred iceꢀcooled dry dimethylformamide (2.2 mL) was slowly added freshly distilled
phosphorus oxychloride (1.1 mL) via dropping funnel and the mixture was stirred at 5°C for 10
min. To this mixture a solution of compound 7b-f (2 mmol) in dry dimethylformamide (2 mL)
was slowly added under stirring at this temperature via dropping funnel. The mixture was
allowed to warm to room temperature and then stirred additionally for 1 h. After this reaction
mixture was heated to 80 °C for 2 h. After cooling it was quenched with ice water, mixed with
charcoal (0.5 g) and filtered. To the clear solution saturated sodium bicarbonate solution was
added and the oil formed was extracted with ethylacetate (20 mL). Organic layer was washed
with brine, dried over sodium sulfate and the solvent was evaporated. Compounds 8b-f were
obtained as dark brown oils and used for the next stage without additional purification.
4
-{[2-(Diethylamino)ethyl](ethyl)amino}-2-hydroxybenzaldehyde (8b). Yield 0.40 g (75%).
1
H NMR (CDCl ): δ = 1.11–1.16 (m, 9H, 3CH ), 2.60–2.66 (m, 6Н, 3CH ), 3.39–3.47 (m, 4Н,
3
3
2
2CH ), 6.11 (s, 1Н), 6.36 (d, J = 8.0, 1H), 7.37 (d, J = 8.0, 1H), 9.56 (s, 1Н, CHO), 11.70 (s, 1Н,
2
OH).
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 51
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
1
4
-[Benzyl(methyl)amino]-2-hydroxybenzaldehyde (8c). Yield 0.38 g (79%). H NMR
(
(
CDCl ): δ = 3.12 (s, 3H, CH ), 4.64 (s, 2Н, CH ), 6.18 (s, 1Н), 6.35 (d, J = 8.0, 1H), 7.30–7.38
3
3
2
m, 6H), 9.55 (s, 1Н, CHO), 11.63 (s, 1Н, OH).
-(Dibenzylamino)-2-hydroxybenzaldehyde (8d). Yield 0.53 g (83%). H NMR (CDCl ): δ
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
1
4
3
= 4.76 (s, 4Н, 2CH ), 6.08 (s, 1Н), 6.37 (d, J = 7.8, 1H), 7.30–7.45 (m, 11Н), 9.61 (s, 1Н, CHO),
2
1
1.11 (s, 1Н, OH).
-{Benzyl[2-(diethylamino)ethyl]amino}-2-hydroxybenzaldehyde (8e). Yield 0.51 (78%).
H NMR (CDCl ): δ = 1.00 (t, J = 7.2, 6H, 2CH ), 2.52 (q, J = 7.2, 4H, 2CH ), 2.66 (t, J = 7.8,
4
1
3
3
2
2
Н, CH ), 3.53 (t, J = 7.8, 2H, CH ), 4.64 (s, 2H, CH ), 6.11 (s, 1H), 6.30 (d, J = 8.4, 1H), 7.14
2 2 2
(
d, J = 7.4, 2H), 7.24–7.31 (m, 4H), 9.45 (s, 1H, CHO), 11.54 (s, 1H, OH).
4-{(4-Chlorobenzyl)[2-(diethylamino)ethyl]amino}-2-hydroxybenzaldehyde (8f). Yield
1
0
2
9
.58 g (80%). H NMR (CDCl ): δ = 1.01 (t, J = 6.9, 6H, 2CH ), 2.56 (q, J = 6.9, 4H, 2CH ),
3
3
2
.66 (t, J = 7.1, 2H, CH ), 3.54 (t, J = 7.1, 2H, CH ), 4.63 (s, 2H, CH ), 6.11 (s, 1H), 6.29 (d, J =
2
2
2
.0, 1H), 7.10 (d, J=7.4, 2H), 7.27–7.33 (m, 3H), 9.51 (s, 1H, CHO), 11.57 (s, 1H, OH).
Procedure d. Compounds 1-4, 5a,b.
A mixture of compound 8a-f (1 mmol), ammonium acetate (0.38 g, 5 mmol) and ethyl
cyanoacetate (0.11 g, 1 mmol) in ethanol (5 mL) was refluxed for 10 min. After cooling the
mixture was diluted with diethyl ether (20 mL). Precipitate formed was collected, washed with
diethyl ether, dried on air and triturated with saturated solution of sodium bicarbonate (5 mL)
with stirring. The solid was filtered, washed with water and dissolved in 2N aqueous
hydrochloric acid (5 mL). The mixture was evaporated to dryness under rotor vacuum. The light
yellow solid residue of target compound was finally dried at 80°C.
ACS Paragon Plus Environment
24

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
-(Diethylamino)-2-oxo-2H-chromene-3-carboximidamide hydrochloride (1). Using 4ꢀ
1
(
diethylamino)ꢀ2ꢀhydroxybenzaldehyde 8a. Yield 0.25 g (83%). Mp 255–257 °C. H NMR
(
DMSOꢀd ): δ = 1.17 (t, J = 7.2, 6H, 2CH ), 3.55 (q, J = 7.2, 4H, 2CH ), 6.63 (s, 1Н), 6.90 (d, J
6
3
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
=
8.0, 1H, Harom.), 7.55 (d, J = 8.0, 1H), 8.93 (s, 1Н), 9.15 (s, 4Н, 2NH ). C NMR (DMSOꢀd ):
2 6
δ = 12.23, 44.54, 95.96, 101.63, 107.04, 110.86, 131.98, 148.03, 153.72, 157.76, 159.20, 161.40.
LC/MS (M+1): 260. Elemental analysis calculated (%) for C H ClN O : C 56.85, H 6.13, Cl
1
4
18
3
2
11.99, N 14.21. Found: C 56.67, H 6.23, Cl 12.05, N 14.19.
7-{[2-(Diethylamino)ethyl](ethyl)amino}-2-oxo-2H-chromene-3-carboximidamide
dihydrochloride (2). Using 4ꢀ{[2ꢀ(diethylamino)ethyl](ethyl)amino}ꢀ2ꢀhydroxybenzaldehyde
1
8
b. Yield 0.31 g (77%). Mp 188–190 °C. H NMR (DMSOꢀd ): δ = 1.17 (t, J = 7.0, 3H, CH ),
6
3
1.27 (t, J = 7.0, 6H, 2CH ), 3.17–3.23 (m, 6H, 3CH ), 3.58 (q, J = 7.0, 2H, CH ), 4.02 (t, J = 7.0,
3
2
2
2
H, CH ), 6.84 (s, 1Н), 7.09 (d, J = 8.0, 1H, H), 7.57 (d, J = 8.0, 1H), 9.02 (s, 1Н), 9.23 (s, 4Н,
2
+
13
2
NH ), 11.50 (s, 1Н, NH ). C NMR (DMSOꢀd ): δ = 8.33, 12.03, 44.38, 45.02, 46.01, 47.36,
2
6
9
6.88, 103.45, 107.66, 111.18,132.04, 148.42, 153.65, 157.43, 159.01, 161.47. LC/MS (M+1):
3
31. Elemental analysis calculated (%) for C H Cl N O : C 53.60, H 7.00, Cl 17.58, N 13.89.
1
8
28
2
4
2
Found: C 53.65, H 7.03, Cl 17.70, N 13.81.
7
-[Benzyl(methyl)amino]-2-oxo-2H-chromene-3-carboximidamide hydrochloride (3).
Using 4ꢀ[benzyl(methyl)amino]ꢀ2ꢀhydroxybenzaldehyde 8c. Yield 0.28 g (80%). Mp 251–253
1
°C. H NMR (DMSOꢀd ): δ = 3.25 (s, 3Н, CH ), 4.82 (s, 2Н, CH ), 6.67 (s, 1Н), 6.92 (d, J = 8.0,
6
3
2
1
H), 7.23 (d, J = 7.0, 2H), 7.27 (t, J = 7.0, 1H), 7.36 (t, J = 7.0, 2H), 7.58 (d, J = 8.0, 1H), 8.89
1
3
(
s, 1Н), 9.16 (s, 4Н, 2NH2). C NMR (DMSOꢀd ): δ = 39.50, 55.05, 96.88, 103.28, 107.47,
6
1
11.20, 126.48, 127.13, 128.65, 131.76, 137.05, 148.19, 155.14, 157.16, 158.98, 161.36. LC/MS
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 51
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
(
M+1): 308. Elemental analysis calculated (%) for C H ClN O : C 62.88, H 5.28, Cl 10.31, N
1
8
18
3
2
1
2.22. Found: C 63.01, H 5.23, Cl 10.37, N 12.27.
7-(Dibenzylamino)-2-oxo-2H-chromene-3-carboximidamide hydrochloride (4). Using 4ꢀ
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
1
(dibenzylamino)phenolꢀ2ꢀhydroxybenzaldehyde 8d. Yield 0.40 g (88%). M.p.288–290 °C. H
NMR (DMSOꢀd ): δ = 4.92 (s, 4H, 2CH ), 6.61 (s, 1Н), 6.87 (d, J = 8.0, 1H), 7.33–7.48 (m,
6
2
13
1
0Н), 7.53 (d, J = 8.0, 1H), 8.99 (s, 1Н), 9.17 (s, 4Н, 2NH ). C NMR (DMSOꢀd ): δ = 55.10,
2 6
9
6.86, 103.21, 107.45, 111.24, 126.49, 127.11, 128.62, 131.77, 137.04, 148.13, 155.20, 157.18,
158.99, 161.36. LC/MS (M+1): 384. Elemental analysis calculated (%) for C H ClN O : C
2
4
22
3
2
6
8.65, H 5.28, Cl 8.44, N 10.01. Found: C 68.60, H 5.21, Cl 8.51, N 9.96.
7-{Benzyl[2-(diethylamino)ethyl]amino}-2-oxo-2H-chromene-3-carboximidamide
dihydrochloride (5a). Using 4ꢀ{benzyl[2ꢀ(diethylamino)ethyl]amino}ꢀ2ꢀhydroxybenzaldehyde
1
9
e. Yield 0.40 g (85%). Mp 192–194 °C. H NMR (DMSOꢀd ): δ = 1.24 (t, J = 7.0, 6H, 2CH ),
6
3
3
.18 (q, J = 7.0, 4H, 2CH ), 3.29 (t, J = 7.2, 2H, CH ), 4.16 (t, J = 7.2, 2H, CH ), 4.88 (s, 2H,
2 2 2
CH ), 6.87 (s, 1H), 7.10 (d, J = 8.0, 1H), 7.24–7.30 (m, 3H), 7.35–7.41 (m, 2H), 7.60 (d, J = 9.2,
2
+
13
1
H), 8.86 (s, 1H), 9.10 (s, 2H, NH ), 9.23 (s, 2H, NH ), 11.38 (s, 1H, NH ). C NMR (DMSOꢀ
2 2
d6): δ = 8.34, 45.59, 45.99, 47.23, 53.45, 97.47, 104.69, 107.98, 111.33, 126.14, 126.97, 128.46,
31.70, 136.81, 148.18, 153.94, 157.07, 158.74, 161.39. LC/MS (M+1): 393. Elemental analysis
calculated (%) for C H Cl N O : C 59.36, H 6.50, Cl 15.23, N 12.04. Found: C 59.33, H 6.61,
1
2
3
30
2
4
2
Cl 15.12, N 12.00.
-{(4-Chlorobenzyl)[2-(diethylamino)ethyl]amino}-2-oxo-2H-chromene-3-
7
carboximidamide
dihydrochloride
(5b).
Using
4ꢀ{(4ꢀchlorobenzyl)[2ꢀ
1
(diethylamino)ethyl]amino}ꢀ2ꢀhydroxybenzaldehyde 9f. Yield 0.42 g (84%). Mp 204–206°C. H
NMR (DMSOꢀd ): δ = 1.25 (t, J = 7.0, 6H, 2CH ).3.17 (q, J = 7.0, 4H, 2CH ), 3.29 (t, J = 7.4,
6
3
2
ACS Paragon Plus Environment
26

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
H, CH ), 4.15 (t, J = 7.4, 2H, CH ), 4.87 (s, 2H, CH ), 6.86 (s, 1H), 7.08 (d, J = 9.0, 1H), 7.27
2
2
2
(
d, J = 8.5, 2H), 7.42 (d, J = 8.5, 2H), 7.59 (d, J = 9.0, 1H), 8.91 (s, 1H), 9.21 (s, 4H, 2NH2),
+
13
11.40 (s, 1H, NH ). C NMR (DMSOꢀd ): δ = 8.31, 45.61, 46.03, 47.29, 52.83, 97.71, 104.99,
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
08.09, 111.42, 128.38, 128.65, 131.74, 131.96, 135.99, 148.22, 157.08, 158.72, 161.37. LC/MS
(M+1): 427. Elemental analysis calculated (%) for C H Cl N O : C 55.26, H 5.85, Cl 21.28, N
2
3
29
3
4
2
11.21. Found: C 55.17, H 5.82, Cl 21.30, N 11.25.
Homology modelling and Structure based design
The structure of the human ASIC1a was modelled based on the crystal structure of chicken
3
5
ASIC1a in the desensitized state solved at 1.9 Å resolution, 2QTS . Sequences of human and
chicken ASIC1a share high identity level (90.65%) what allow to use homology modeling with
6
9
high confidence. Automated homology modeling routine of SwissModel Server was applied.
Sequence of chicken ASIC1a was derived from uniprot database (UNIPROT ID: P78348).
Transmembrane domains TM1 and TM2 of ASICs undergo major conformational
transformations during channel function. In the crystal structure conformations of the three
subunits are not equal, so we model every chain of human ASIC using corresponding chain from
chicken ortholog (chains A, B and C). Human has two amino acids insertion comparing to
chicken Asp298 and Leu299 which are located in the solvent exposed loop. Superposition of the
Cα atoms of the ECDs (His74ꢀLys423, chicken numbering) omitting residues Asp297 and
Ser298 from chicken and Asp296ꢀLeu299 from human ASICs returned RMSD of 0.62 Å, what
denotes reliability of the model. The full trimer was reconstructed by superposition of the model
onto each monomer in the chicken model omitting transmembrane regions and the above
mentioned loop. In the generated model positions of the side chain atoms were further
energetically optimized using GROMACS v3.3, with OPLS 2001 force field in 5000 steps of
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 51
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
conjugated gradient routine. The quality of the generated model was monitored using MolProbity
7
0
server .
Small molecule modeling and docking studies were performed with Schrodinger suite. Maestro
was used for small molecule structure sketching and visualization of docking results.
For the pilot docking study we used Enamine diversity set, a library of 10 000 compounds
selected basing on chemical structure diversity. Structures of small molecules were optimized
with LigPrep module and their properties predicted with QuickProp. Aꢀ317567 was assumed to
have trans conformation of its cyclopropyl group. Human ASIC1a structure was prepared with
protein preparation wizard. Topologies of all atoms were reassigned and their positions were
verified to prevent steric clashes. For docking studies, Glide module of Schrödinger was used.
To generate a grid of the acidic pocket, amiloride molecule was manually inserted into the acidic
pocket and all residues within 25 Å from any amiloride atom were included into the grid
generation. Conformations of small molecules were flexible, while protein atoms were fixed to
their positions. For every generated pose post docking minimization was performed and 10
complexes per structure were saved. To determine hydrophobic sites SiteMap from Schrödinger
suite was applied. For the residues within 10 Å from bound Aꢀ317567 (as more extended ligand)
were included into calculation. Within isovalue cutꢀoff of ꢀ0.58 only two regions in the pH
sensor were detected (see Results).
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
HEK 293 cells culture
Human embryonic kidney 293 (HEK 293) cells (American Type Culture Collection, Manassas,
VA, USA) were cultured in Dulbecco’s modified eagle medium (DMEM) supplemented with
10% fetal bovine serum and 10 U/ml penicillin and 10 mg/ml streptomycin (all from Invitrogen,
USA). Dissociated cells were either reꢀplated for a new passage or used for patch clamp
ACS Paragon Plus Environment
28

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
o
experiments. Cells were cultured at 37 C under an atmosphere of 5% CO and 95% air with
2
approximately 95% humidity.
Hippocampal slices preparation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acute temporal lobe slices including neocortical areas (Te2 and Te3), entorhinal cortex,
subiculum and hippocampus were prepared from Wistar rats aged postnatal day 19 to 21 (P19ꢀ
71
2
1) as previously described with some modifications . On the day of experiment rat was deeply
anesthetized using sevoflurane and decapitated. Cerebellum, frontal lobe region (coronal section)
and ventralꢀlateral areas (sections at the angle 20° to 30° off the horizontal axis) were removed
from the brain. The remaining part of the brain was mounted on the stage of a Vibroslice NVSL
(World Precision Instruments Inc., Sarasota, FL, USA) and cut (400 µm) through the
hemispheres at an angle of 30–35° of their horizontal planes. All manipulations were performed
in freshly prepared iceꢀcold oxygenated (95% O ꢀ 5% CO ) artificial cerebrospinal fluid (ACSF)
2
2
of the following composition (mM): NaCl 119, KCl 2.5, CaCl 2.0, MgSO 1.3, NaHCO 26,
2
4
3
NaH PO 1.0 and glucose 11 (pH 7.35). For the experiments we took 3ꢀ4 slices from the septal
2
4
7
2
(
dorsal) part of the hippocampus . Slices were maintained in an oxygenated ACSF at a room
temperature for at least 1.5 h before use.
Electrophysiological recordings
Whole cell patch clamp recordings were made with EPCꢀ8/LIH 1600 amplifier/acquisition
system; data were collected using PatchMaster software (all from HEKA, Lambrecht/Pfalz,
Germany). Current was recorded at holding potential of ꢀ100 mV, unless otherwise indicated.
Current traces were samples at 10 kHz and filtered online at 3 kHz. Patch electrodes (2 ꢀ 3 MΩ)
were filled with a solution containing (in mM): 120 KF, 20 TrisꢀCl, (adjusted to pH 7.3 with
KOH). Extracellular solution contained (in mM): 130 NaCl, 5 KCl, 2 MgCl , 2 CaCl , 20
2
2
ACS Paragon Plus Environment
29