TREK Channel Family Activator with a Well-Defined Structure-Activation Relationship for Pain and Neurogenic Inflammation

Article abstract of DOI:10.1021/acs.jmedchem.9b02163

TWIK-related K⁺ (TREK) channels are potential analgesic targets. However, selective activators for TREK with both defined action mechanism and analgesic ability for chronic pain have been lacking. Here, we report (1S,3R)-3-((4-(6-methylbenzo[d]

Full text of DOI:10.1021/acs.jmedchem.9b02163

Subscriber access provided by NATIONAL UNIV OF SINGAPORE
Article
A TREK Channel Family Activator with Well-Defined Structure-
Activation Relationship for Pain and Neurogenic Inflammation
Yunguang Qiu, Lu Huang, Jie Fu, Chenxia Han, Jing Fang, Ping Liao, Zhuo Chen,
Yiqing Mo, Peihua Sun, Daqing Liao, Linghui Yang, Jing Wang, qiansen zhang,
Jin Liu, Feng Liu, Tingting Liu, Wei Huang, Huaiyu Yang, and Ruotian Jiang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b02163 • Publication Date (Web): 12 Mar 2020
Downloaded from pubs.acs.org on March 13, 2020
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 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 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.
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 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A TREK Channel Family Activator with Well-
Defined Structure-Activation Relationship for Pain
and Neurogenic Inflammation
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
†,§, ¶,#
‡,#
†,#
∥
‡
⊥
Yunguang Qiu
, Lu Huang , Jie Fu , Chenxia Han , Jing Fang , Ping Liao , Zhuo
‡
†
‡
‡
‡
⊥
Chen , Yiqing Mo , Peihua Sun , Daqing Liao , Linghui Yang , Jing Wang , Qiansen
†
‡
∥
∥
†,*
⊥
Zhang , Jin Liu , Feng Liu , Tingting Liu , Wei Huang , Huaiyu Yang and Ruotian Jiang
‡,*
†Shanghai Key Laboratory of Regulatory Biology, Institute of Biomedical Sciences,
School of Life Sciences, East China Normal University, Shanghai 200241, China.
‡
Laboratory of Anesthesia and Critical Care Medicine, National-Local Joint Engineering
Research Center of Translational Medicine of Anesthesiology, West China Hospital,
Sichuan University & The Research Units of West China (2018RU012), Chinese
Academy of Medical Sciences, Chengdu 610000, China.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 61
1
2
3
4
5
6
7
8
9
§State Key Laboratory of Drug Research and CAS Key Laboratory of Receptor Research,
Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203,
China.
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
∥Department of Integrated Traditional Chinese and Western Medicine, Sichuan Provincial
Pancreatitis Centre and West China-Liverpool Biomedical Research Centre, West China
Hospital, Sichuan University, Chengdu 610000, China
⊥
Jiangsu Key Laboratory of Neuropsychiatric Diseases and College of Pharmaceutical
Sciences, Soochow University, Suzhou 215123, China
¶
University of Chinese Academy of Sciences, Beijing 100049, China
ABSTRACT. TWIK-related K+ (TREK) channels are potential analgesic targets. However,
selective activators for TREK with both defined action mechanism and analgesic ability
for chronic pain have been lacking. Here we report C3001a, a selective activator for TREK
against other two-pore domain K+ (K2P) channels. C3001a binds to the cryptic binding
site formed by P1 and TM4 in TREK-1, as suggested by computational modeling and
ACS Paragon Plus Environment

Page 3 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
experimental analysis. Further, we identify the carboxyl group of C3001a as a structural
determinant for the binding to TREK-1/2, and the key residue that defines the subtype-
selectivity of C3001a. C3001a targets TREK channels in peripheral nervous system to
reduce the excitability of nociceptive neurons. In neuropathic pain, C3001a alleviated
spontaneous pain and cold hyperalgesia. In a mouse model of acute pancreatitis, C3001a
alleviated mechanical allodynia and inflammation. Together, C3001a represents a lead
compound which could advance the rational design of peripherally-acting analgesics
targeting K2P channels without opioid-like adverse effects.
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
INTRODUCTION
Long-term opioid administration has minimal effects on chronic pain and is generally
coupled with tolerance, dependence, and addiction1. In acute pancreatitis (AP), the use
of morphine for pain control may even exacerbate the disease by increasing pancreatic
neutrophilic infiltration and necrosis2. Thus, there is a clinically unmet demand for
developing new non-opioid analgesics. In this regard, the two-pore domain K+ (K2P)
channels are considered as emerging candidates3, 4. Recently, we showed that activation
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 61
1
2
3
4
5
6
7
8
9
of the TWIK-related acid-sensitive K+ 3 (TASK-3) channel, a K2P channel, using a newly
discovered selective activator CHET3, displayed therapeutic potential in analgesia5. K2P
family channels consist of 15 gene products, and each of those contains four
transmembrane segments (TM1–TM4) and two-pore domains (P1 and P2)6. In the K2P
family, the TWIK-related K+ channel (TREK) subfamily is comprised of TREK-1 (Kcnk2),
TREK-2 (Kcnk10), and TWIK-related arachidonic acid-stimulated K+ channel (TRAAK,
Kcnk4). The TREK family channels share >78% sequence homology and some common
activation mechanisms7. They are mechanical-, thermal- and lipid-gated channels that
can be regulated by different physical and chemical stimuli, including membrane stretch8,
9, membrane depolarization10, temperature11, pH12, 13, arachidonic acid, and other
polyunsaturated fatty acids9. In peripheral sensory neurons, TREK-1 is expressed in
peptidergic substance P (SP)-positive fibers or nonpeptidergic isolectin 4-positive (IB4+)
C-fibers and co-localized with transient receptor potential cation channel subfamily V
member 1 (TRPV1) channels14, while TREK-2 is selectively expressed in IB4+ C-fibers15.
Functionally, TREK channels produce background K+ currents which contribute to the
resting membrane potential (RMP) and action potential properties of nociceptive
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
ACS Paragon Plus Environment

Page 5 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
neurons16. A recent study by Kanda et al17 found that TREK-1 and TRAAK are highly
enriched at the nodes of Ranvier of trigeminal myelinated afferent nerves, in which they
mediate rapid action potential conduction and are involved in sensory behavioral
response. TREK-1, TREK-2 or TRAAK knockout mice showed significantly increased
sensitivity to mechanical and thermal stimuli14, 18, 19. Notably, Devilliers et al20 found that
morphine may enhance the TREK-1-mediated currents by phosphorylating the channel,
and the enhancement of TREK-1 function contributes to the analgesic effect of morphine
without its adverse side effects. Thus, activation of TREK channels may provide an
opportunity to dissociate the analgesic from major adverse effects of opioids for pain
therapies.
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
A few compounds have been previously identified that selectively activate TREK
channels including ML67-33, BL-1249, ML335, ML402, and GI-53015921-24. However, to
our knowledge, except for a derivative of BL-124925, the pharmacokinetic profiles as well
as the in vivo ability of these compounds remain largely unreported. So far, the selective
TREK-1 activators with reported in vivo analgesic properties are cinnamyl 3,4-dihydroxyl-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 61
1
2
3
4
5
6
7
8
9
α-cyanocinnamate analogues derived from caffeate esters26. However, the binding site
and the molecular basis for the subtype-selectivity of these compounds in TREK channels
are unclear. It is noteworthy that under chronic pain conditions, the expression of TREK
channels was downregulated18, 27, 28. Therefore, whether selective activation of TREK
channels is sufficient to attenuate thermal hyperalgesia and mechanical allodynia under
chronic pain remains to be determined. In the search of small molecules with new
chemical scaffolds, which selectively activate TREK channels, we discovered a subtype-
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
selective
TREK-1/2
activator,
(1S,3R)-3-((4-(6-methylbenzo[d]thiazol-2-
yl)phenyl)carbamoyl)cyclopentane-1-carboxylic
acid
(named
C3001a).
Using
computational and experimental methods, we defined the binding mode and the structural
selectivity basis of C3001a in TREK-1. Having characterized the inhibitory role of C3001a
in the excitability of nociceptive neurons, we further explored the therapeutic potential of
C3001a in mouse models for inflammatory or neuropathic pain. Particularly, since the
tight coupling between TREK-1 and -opioid receptor (OR), we evaluated the
therapeutic effects of activation of TREK channels by C3001a in a mouse model of AP.
ACS Paragon Plus Environment

Page 7 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
AP in mice represents a specific preclinical visceral inflammatory pain model in which the
6
7
8
opioids have debatable therapeutic application2, 29
.
9
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
RESULTS
C3001a selectively activates TREK channels, preferentially TREK-1 and TREK-2
channels. In one of our in-house experimental screenings performed to find bioactive
compounds for the TREK-1, we identified C3001a (Figure 1A) as a potential TREK-1
activator. In the whole-cell voltage clamp experiments on human embryonic kidney
(HEK)-293T cells that were transiently transfected with recombinant human TREK-1,
increase in current amplitude was recorded by the extracellular application of C3001a,
and the increase could be rapidly reversed by washout (Figure 1B). The effect of C3001a
in increasing the current was concentration-dependent with a half-maximum effective
concentration (EC50) of 12.81 M (Figure 1C). To further verify that C3001a is indeed an
activator of TREK-1, we tested C3001a on TREK-1 channels using inside-out and
outside-out patches. The outside-out recordings revealed that C3001a significantly
increased the open probability of TREK-1 (P = 0.0058 for -60 mV, P = 0.0054 for +60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 61
1
2
3
4
5
6
7
8
9
mV), with no change in single channel conductance (P = 0.486 for -60 mV, P = 0.055 for
+60 mV Figure 1D-G). Interestingly, C3001a did not show any significant effects on the
open probability or channel conductance of TREK-1 in the inside-out recordings (Figure
S1). Thus, the single channel recordings suggest that C3001a directly activates TREK-1
through a binding site accessible from the extracellular side of TREK-1. We further
investigated whether C3001a also activates other K2P channels. TREK-2 and TRAAK,
two other TREK subfamily members, responded to C3001a with similar EC50 (11.31 M
and 15.29 M, respectively). C3001a activated TREK-2 with relatively high efficacy (~4.5-
fold at 10 M in comparison to control) which was similar to that on TREK-1. However,
C3001a activated TRAAK with relatively low efficacy (~2.4-fold at 10 M in comparison
to control) (Figure. 1H and Figure S2). C3001a showed little effects on other human K2P
channels, including TWIK-related acid-sensitive K+ 3 (TASK-3), TASK-1, TWIK-related
spinal cord K+ channel (TRESK) and tandem pore domain halothane-inhibited K+ channel
1 (THIK-1) (Figure 1I and Figure S2). Together, these results suggest that C3001a is a
selective TREK-1/2 activator against other K2P channels.
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
ACS Paragon Plus Environment

Page 9 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
Figure 1. Identification of a TREK-1/2 activator, C3001a. (A) Chemical structure of
C3001a. (B) Example whole-cell path-clamp recording of C3001a on TREK-1 channel. (C)
C3001a concentration-response curve for TREK-1 with EC50 = 12.81 M (n = 6-23). (D)
Representative single-channel current traces from outside-out patches. (E) Amplitude
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 61
1
2
3
4
5
6
7
8
9
histograms for the patch recording shown in (D) which were fitted by Gaussian
distributions. (F, G) Analysis of NPo (channel number times open probability, F) and
channel conductance changes (G) for the single-channel recordings (n = 7; paired t test).
(H) C3001a concentration-response curves for TREK-2 and TRAAK (n = 6-13) with EC50
= 11.31 M, and 15.29 M, respectively. (I) Summary for the effects of C3001a (10 M)
on several K2P channels (n = 5-12). Data are shown as means ± SEM. **P < 0.01, n.s.,
not significant.
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
Binding mode of C3001a to TREK-1. To determine the binding modes of C3001a to
TREK subfamily members, we first analyzed pockets in reported crystal structures of
TREK subfamily members by using Fpocket 2.0 server30. In this computation, a
druggability score greater than 0.5 (the threshold) means that the pocket might be
druggable30. 32 pockets with druggability scores greater than 0.5 were identified (Figure
S3). Then we docked C3001a into each of the pockets by using Glide. The docking results
suggested that C3001a may bind to the extracellular crevice formed by P1 and TM4. The
ACS Paragon Plus Environment

Page 11 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
docking scores (Gscores) of C3001a to this crevice in many structures are the top ones
(Figure S3). Since the docking score of C3001a to the P1-M4 pocket (the cryptic binding
site) in the reported TREK-1 structure with Protein Data Bank (PDB) code of 6CQ8 is the
best compared to other reported K2P structures, this crystal structure was adopted for
further computational study. And hereafter the P1-M4 pocket was described as the cryptic
binding site named by Lolicato et al.23. Of note, identification of the extracellular cryptic
binding site as potential binding site of C3001a is consistent with the results of inside-out
and outside-out patches, which suggests that C3001a binds to the extracellular side of
TREK-1.
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
Next, given the protein flexibility and environmental effects, RosettaLigand was applied
to further characterize the binding modes of C3001a to the cryptic binding site. Six starting
points were applied for the ligand centroid to scan every site in the pocket as far as
possible. 1000 docking poses with lowest total energy score were selected and clustered
into eight possible binding models, whose scores are shown in Figure S4A. We selected
and analyzed the three top-ranked binding models (Model 1-3), and found that the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 61
1
2
3
4
5
6
7
8
9
carboxyl group of C3001a forms a hydrogen bond with Y285 in Model 2, whereas it
interacts with K286 in Model 1 and Model 3 (Figure S4A). To determine which
computational model is the correct one, we carried out mutagenesis studies where
mutation Y285A significantly reduced the activation efficacy of C3001a on TREK-1
(current enhancement of ~2-fold at 10 M), whereas mutation K286A did not have marked
effect in the activation efficacy (current enhancement of ~4.9-fold at 10 M, Figure 2A
and Figure S4B). Therefore, the mutagenesis data are so far consistent with the
interactions between Y285 and the carboxyl group of C3001a predicted in Model 2, rather
than Model 1 or Model 3. Next, we designed two derivatives in which the carboxyl group
was substituted with hydrogen or methoxycarbonyl group to disturb the hydrogen bond
interaction of C3001a with Y285 (Scheme 1 and 2). The two derivatives (30 M) showed
little activation effects on TREK-1, supporting that the carboxyl group of C3001a is
essential for activity (Figure 2B,C).
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
ACS Paragon Plus Environment

Page 13 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
Figure 2. Binding mode of C3001a in TREK-1. (A) Concentration-response curve for
C3001a on mutation Y285A with EC50 = 7.86 M (n = 5-10). (B, C) Chemical structures
and effects of two C3001a derivatives: CLD104 and CLD105 (30 M) on TREK-1 WT (n
= 5-6, one-way ANOVA with Dunnett’s post hoc test). (D) The top view and detailed view
of the predicted binding mode. The TREK-1 structure with PDB code of 6CQ8 was used
for building this model. (E) 2D diagram shows interactions between C3001a and TREK-
1. (F) Computations show the contributions of seven residues and their mutations to
C3001a binding (n = 10, top 10 docking models). EVDW and EHB represent energy
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 61
1
2
3
4
5
6
7
8
9
distributions of VDW and hydrogen bond interactions, respectively. Energy unit is Rosetta
Energy Unit (R.E.U.). (G) Concentration-response curves for C3001a on four mutations
(F149A, T156A, W290A and I293A), yielding EC50 of 22.43 M, 7.57 M, 3.91 M and
7.98 M, respectively (n = 6-13) Data are shown as means ± SEM. ***P < 0.001.
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
Scheme 1. Synthesis of CLD104a.
S
N
COOH
S
N
NH
O
NH2
i
1
2 (CLD104)
aReagents and conditions: (i) HATU, DIPEA, DCM, rt, 12 h; (CLD104 yield: 83%).
Scheme 2. Synthesis of CLD105 and C3001aa.
O
COOH
COOMe
O
HOOC
HOOC
i
ii
iii
O
3
4
5
6
S
N
S
N
S
N
NH2
NH
O
NH
1
O
OH
O
iv
v
O
O
7 (CLD105)
8 (C3001a)
ACS Paragon Plus Environment

Page 15 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
aReagents and conditions: (i) RuCl3, NaIO4, EtOAc, MeCN, water, rt, 12 h; (ii) Ac2O,
150 °C, 30 min; (iii) MeOH, 80 °C, 5 h; (iv) HATU, DIPEA, DCM, rt, 12 h; (v) LiOH, THF,
water, rt, 8 h; (CLD105 yield: 82%; C3001a yield: 63%).
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 detailed information of Model 2 is shown in Figure 2D. In this model, in addition to
Y285, residues F149, T156, I158, F160, W290, and I293 are close to C3001a and may
provide hydrophobic interactions with the benzothiazole or benzene ring of C3001a. Also,
W290 may establish a hydrogen bond with benzothiazole nitrogen of C3001a (Figure
2D,E). To further verify the binding mode, we calculated the influence of mutations F149A,
T156A, I158A, F160A, Y285A, W290A and I293A on the binding of C3001a by comparing
binding contributions of each single residue and that of their mutants (Figure 2F). The
computation of energy distributions of the average van der Waals (VDW) interactions
predicted that compared with wild type (WT) TREK-1, F149A, T156A, F160A, Y285A,
W290A and I293A should reduce the binding of C3001a whereas I158A may not have
significant effects (Figure 2F, Left). Moreover, in Y285A and W290A, the contributions of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 61
1
2
3
4
5
6
7
8
9
hydrogen bond energy of Y285 and W290 to ligand binding were significantly reduced
(Figure 2F, Right). Then we carried out electrophysiological experiments to test the
effects of these mutations on the C3001a activation of TREK-1. I158A and F160A
mutations led to nonfunctional channels (Figure S4C). For other functional mutants,
expect for I293A, F149A, T156A, Y285A and W290A displayed markedly reduced
responses to C3001a (Figure 2A,G). So far, by combining computation with functional
mutagenesis, we provide a few structural determinants for the binding of C3001a in the
cryptic binding site (Figure 2D,E).
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
G275 is a residue located at the mouth of the ligand binding pocket (Figure 3A).
Although G275 does not form direct interaction with C3001a, introducing a side chain at
this position should reduce the entrance size of the pocket and as a result may impede
the binding of C3001a to the pocket (Figure 3A). We made the cysteine mutant G275C
and we found that G275C showed reduced activation by C3001a (Figure 3B,C).
Introducing a side chain by MTSET to G275C may further seal up the entrance of the
pocket. Indeed, application of MTSET reduced the activation of C3001a on G275C.
ACS Paragon Plus Environment

Page 17 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
MTSET did not show any significant effect on the TREK-1 WT channel (Figure 3C),
whereas MTSET further right-shifted the concentration-response curve for C3001a on
G275C (Figure 3D). These results further support that C3001a binds to the cryptic binding
site.
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
Figure 3. C3001a activity on G275C. (A) Pocket changes by G275C mutation or G275C
modified with MTSET (PDB code: 6CQ8). (B) Example whole-cell path-clamp recordings
of C3001a on G275C mutation before (Left) and after (Right) MTSET (1 mM). (C)
Summary for the effects of C3001a (10 M) on TREK-1 WT and G275C without (or with)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 61
1
2
3
4
5
6
7
8
9
MTSET (1 mM) (n = 4-5 cells, one-way ANOVA with Tukey’s post hoc test). (D)
Concentration-response curves of G275C mutation and G275C modified with MTSET on
C3001a activity with EC50 = 15.19 M and 18.40 M, respectively (n = 5-19). Data are
shown as means ± SEM. *P < 0.05, **P < 0.01, n.s., not significant.
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
Figure 4. D227 contributes to C3001a selectivity between TREK-1/2 and TRAAK. (A)
Sequence alignment of TREK subfamily. (B) Structure presentation of D227Y mutation in
the proposed binding model (Models were built based on the TREK-1 structure with the
PDB code of 6CQ8). (C, D) The effects of C3001a (10 M) (C) and the concentration-
response curve of C3001a (D) on TRAAK mutation D227Y with EC50 = 9.96 M (n = 4-
10, one-way ANOVA with Tukey’s post hoc test in (C)). Data are shown as means ±
SEM. *P < 0.05, **P < 0.01, n.s., not significant.
ACS Paragon Plus Environment

Page 19 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Key determinant for C3001a selectivity on TREK subfamily. Compared with TREK-1
and TREK-2, TRAAK was less sensitive to C3001a. According to the multiple sequence
alignment in TREK channels, among the residues near the pocket, the residue Y281 in
TREK-1 was conserved in TREK-2, but not in TRAAK (equivalent residue D227, Figure
4A). We tested whether D227 controls the subtype selectivity. TRAAK was modeled
based on C3001a/TREK-1 binding model. Apparently, in C3001a/TRAAK complex, the
carboxyl group of C3001a forms strong negative electrostatic repulsion with D227 (Figure
4B). Hence, we tested the effect of C3001a on the mutant TRAAK D227Y. As a result,
D227Y showed significantly increased activation efficacy (current enhancement of ~5.4-
fold at 10 M) by C3001a, compared with TRAAK WT (current enhancement of ~2.5-fold
at 10 M). This was comparable with that in TREK-1 WT (current enhancement of ~5.7-
fold at 10 M) by C3001a (P = 0.90, Figure 4C, D), demonstrating that the residue D227
in TRAAK is an essential structural determinant for the selectivity of C3001a against
TRAAK.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 61
1
2
3
4
5
6
7
8
9
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
Figure 5. Effects of C3001a on nociceptive neurons. (A) Left, representative
electrophysiological traces show the effects of C3001a (10 M) on K+ currents in DRG
neurons; Right, bar graph summary for experiments in (Left) (n = 35 cells in 4 rats;
Wilcoxon signed rank test). (B) Left, representative trace shows C3001a-sensitive
currents at extracellular 3 mM K+ concentrations with reversal potential at -75 mV; Middle,
bar graph shows the average reversal potential around -77 mV in (Left) (n = 4 cells of 35
cells in A); Right, representative trace shows C3001a-sensitive currents at extracellular 3
mM K+ concentrations without determined reversal potential. (C) Representative trace
and bar graph show the effects of C3001a on RMP changes (n = 18 cells in 4 rats; paired
t test). (D, E) Traces and bar graph show the effects of C3001a on rheobase and firing
ACS Paragon Plus Environment

Page 21 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
frequency in nociceptive neurons (n = 19 cells in 4 rats; Wilcoxon signed rank test in (D)
6
7
8
and (E)). Data are shown as means ± SEM. **P < 0.01, ***P < 0.001.
9
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
Functional effects of C3001a on dorsal root ganglion (DRG) neurons. The functional
effects of C3001a on TREK channels were further examined in dissociated DRG neurons.
Recordings were focused on small-sized DRG neurons (diameter of ~20 m, cell
capacitance of ~20 pF), in which the TREK channels were functionally identified14. To
minimize activation of voltage-gated K+ current, we applied voltage ramps from -20 mV
to -120 mV. In 55 cells recorded, 35 cells responded to C3001a (C3001a-sensitive cells).
In the C3001a-sensitive cells, C3001a enhanced the whole-cell current density (in pA/pF)
by about 24% (Figure 5A). We subtracted the C3001a-sensitive current, and found that
this current was strongly outwardly rectifying (Figure 5B). Only in 4 cells, the reversal
potential could be reliably determined, which was around -77 mV, a value that was close
to the reversal potential of K+. In the other 31 cells, the fraction of C3001a-sensitive
current was negligible between -120 mV and -60 mV (Figure 5B), making the reversal
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 61
1
2
3
4
5
6
7
8
potential of the C3001a-sensitive current difficult to resolve. In addition, the application of
C3001a resulted in a slight hyperpolarization of the RMP by 1.12 ± 0.42 mV (Figure 5C).
9
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
Furthermore, we explored how C3001a regulated the excitability of DRG neurons. In 19
out of 48 neurons, C3001a increased the rheobase (the minimum current required to elicit
the first action potential) by ~30% and decreased the number of action potentials evoked
by suprathreshold current injections by ~60% (Figure 5D,E). In the other 29 cells, C3001a
did not modify the rheobase and slightly increased the frequency of action potentials by
6% (Figure S5). Collectively, our electrophysiological data suggest that C3001a increases
the K+ conductance, hyperpolarizes the RMP, and reduces the excitability in a subset of
nociceptive DRG neurons.
ACS Paragon Plus Environment

Page 23 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
Figure 6. Analgesic and anti-inflammatory effects of C3001a in rodents. (A) Left, time
course of dose-dependent analgesia by C3001a (1.25, 2.5, 5, 10 mg/kg i.p.) in tail
immersion test at 52 °C (n = 6-9; unpaired t test or Mann-Whitney test); Right, the
analgesic effect of C3001a (5 mg/kg i.p.) in tail immersion tests at different temperatures
(n = 9-17; unpaired t test or Mann-Whitney test); (B) Left, time course of the effect of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 61
1
2
3
4
5
6
7
8
9
C3001a (10 mg/kg i.p.) on mechanical allodynia in AP mice (n = 5-6, unpaired t test);
Right, scatter plots show the effect of C3001a on MPO activity in AP mice (n = 6, unpaired
t test); (C) Representative images and scatter plots show the effect of C3001a (10 mg/kg
i.p.) on pancreatic histopathological changes and scores including edema, inflammation
and necrosis in AP mice (n = 6, unpaired t test or Mann-Whitney test. Scale bar: 100 m).
(D) Summary of C3001a (5 mg/kg i.p.) analgesia in CFA-induced chronic inflammatory
pain. Left, time course of the effects of C3001a on heat hyperalgesia (n = 9-10; unpaired
t test); Right, time course of the effects of C3001a on mechanical allodynia (n = 9-10;
Wilcoxon signed rank test). (E) Effects of C3001a (5 mg/kg i.p.) on spontaneous pain
within a 5 min duration on the day 7 after SNI (measured at 30 min post injection, n = 7;
unpaired t test). (F) Left, summary for the effect of C3001a (5 mg/kg i.p.) on cold
hyperalgesia at different stages in SNI (n = 8-10; unpaired t test or Mann-Whitney test);
Middle, time course of the effects of C3001a on cold hyperalgesia in SNI mice (n= 8;
unpaired t test); Right, comparison of the effects of C3001a and pregabalin (30 mg/kg,
i.p.) on cold plantar test in SNI mice (n = 9; unpaired t test); Data are shown as means ±
SEM. *P < 0.05, **P < 0.01, ***P < 0.001, n.s., not significant.
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
ACS Paragon Plus Environment

Page 25 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Analgesic and anti-inflammatory effects of C3001a in rodents. The TREK activators
reported by Vivier et al26, showed in vivo anti-nociceptive activity. However, the
expression of TREK members was downregulated in chronic pain15, 27, 28 and whether the
activation of TREK-1/2 under chronic pain is analgesic remains to be determined. The
pharmacokinetic profile of C3001a (Table S1) showed a negligible brain concentration of
C3001a and a relatively high concentration of C3001a in the plasma, which suggests that
C3001a may exert analgesic functions through TREK family expressed in the peripheral
nervous system (PNS). We systematically evaluated the effects of C3001a in analgesia.
The anti-nociceptive effect of C3001a was first assessed in the tail immersion test at 52°C
in mice. C3001a displayed dose-dependent analgesia, with a maximal effect at a dose of
5 mg/kg reached 30 min after intraperitoneal (i.p.) injection (Figure 6A, Left). C3001a was
effective in response to noxious cold (5 °C) or heat stimuli (46 °C and 52 °C) and to
physiological stimuli (20 °C and 40 °C) (Figure 6A, Right). To further confirm that C3001a
was targeting TREK family to be analgesic, CLD104, the derivative of C3001a with no
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 61
1
2
3
4
5
6
7
8
9
activity on TREK-1 up to 30 M (Figure 2B), was used as a negative control. CLD104
showed no analgesic function in tail immersion test (Figure S6A). AP represents a
clinically relevant visceral pain model with strong neurogenic inflammation, which could
be inhibited by TRPV1 and transient receptor potential ankyrin 1 (TRPA1) antagonists31,
32. Recent study suggested that morphine, widely used for the management of pain in AP,
may worsen its severity in AP2. Furthermore, there are few effective analgesics without
side effects or with beneficial influences on anti-inflammatory effects in AP in current
clinical managements33, 34. Thus, we assessed the analgesic effects of C3001a in a
mouse model of cerulein-induced AP35. In this model, mechanical allodynia, increased
myeloperoxidase (MPO) activity, and significant morphological damage in the form of
necrosis, inflammatory infiltration and pancreatic edema were observed (Figure S6B,C).
Administration of C3001a (10 mg/kg, i.p.) significantly alleviated mechanical allodynia,
reduced MPO activity, and attenuated pancreatic necrosis, inflammation and edema
(Figure 6B,C). We detected robust neuronal expression of TREK-1 and TREK-2 mRNA
by in situ hybridization using RNAscope in DRG in naïve mice, which were both
downregulated in cerulein treated mice (Figure S6D). In contrast, no pancreatic
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
ACS Paragon Plus Environment

Page 27 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
expression of TREK-1 and TREK-2 mRNA was detected in the pancreas in naïve or
cerulein treated mice (Figure S6E), suggesting a neurogenic role of TREK channels in
the pathogenesis of analgesia and anti-inflammation in AP. Additionally, the negative
control compound CLD104 was ineffective in attenuating pancreatic MPO activity or
morphological damage of pancreas in cerulein treated mice (Figure S6F,G). In a mouse
model of chronic inflammatory pain developed by intraplantar injection with the complete
Freund’s adjuvant (CFA), C3001a alleviated heat hyperalgesia and mechanical allodynia
(Figure 6D). In the spared nerve injury (SNI)-induced neuropathic pain mouse model, the
number of hindpaw lifting, a robust behavior associated with spontaneous pain, was
reduced following C3001a administration on day 7 after SNI (Figure 6E). In the cold
plantar test, during SNI development (SNI 7 days) and maintenance (SNI 14 days and 21
days), C3001a attenuated cold hyperalgesia with similar potency compared to pregabalin
(Figure 6F). In summary, our data suggests that C3001a is effective in treating thermal
hyperalgesia and mechanical allodynia in mouse models for acute or chronic pathological
pain and pain-associated neurogenic inflammation.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 61
1
2
3
4
5
DISCUSSION
6
7
8
9
For the discovery and modification of small molecule modulators, understanding the
molecular recognition basis between the small molecules and the targeted proteins is
critical. The currently available crystal structures of TREK channels provide a platform for
revealing the binding site of small molecules by using computations. We analyzed
potential druggable pockets across TREK subfamily crystal structures followed by
molecular docking to identify the possible binding sites of C3001a. By solving the complex
crystal structures of TREK-1 with ML335 and its derivatives, Lolicato et al. revealed a
cryptic selectivity filter binding site for TREK-1 activators23. We compared the binding
modes of ML335 and its derivatives in crystal structures with that of C3001a. We found
that C3001a forms hydrophobic interactions with F149 and W290 which were similar to
the interactions between the equivalent residues F134 and W275 with ML335 and its
derivatives in crystal structures. However, C3001a forms a hydrogen bond interaction
with Y285 (Y270 in the crystal structures) through its carboxyl group whereas the sulfonyl
group of ML335 or its derivatives forms interactions with H141 and S146 (equivalent to
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
ACS Paragon Plus Environment

Page 29 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
H126 and S131 in the crystal structures). These differences are likely due to the flexibility
of both the extracellular loops and the polar groups of the compounds. The identified
binding pocket of C3001a was partially overlapped with that of ML335, which further
supports that this region is a ligand binding site which could accommodate different
chemotypes23. The C3001a binding site was confirmed using several approaches. First,
the inside-out and outside-out patch experiments indicated that C3001a binds to the
extracellular side of the channel. Second, the contributions of several key residues that
line the binding site were calculated by RosettaLigand and validated by functional
mutagenesis test. Third, we synthesized two C3001a analogues to further confirm the
critical role of the carboxyl group of C3001a in TREK-1 binding. Last, the accessibility of
C3001a to the binding pocket was examined by a real time covalent modification of the
cysteine mutant of a residue located at the entrance of the pocket. Thus, with computation,
mutagenesis and chemistry, we provide the binding mechanisms of C3001a to TREK-1.
Of note, because the carboxyl group faces to the solvent, we cannot rule out the possibility
that the carboxyl group may be structurally flexible. Future structural biology work may be
needed to further elucidate the binding of C3001a to TREK channels. Furthermore, the
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
ACS Paragon Plus Environment