N-Cinnamoylanthranilates as human TRPA1 modulators: Structure-activity relationships and channel binding sites

Article abstract of DOI:10.1016/j.ejmech.2019.02.074

The transient receptor potential ankyrin 1 (TRPA1) channel is a non-selective cation channel, which detects noxious stimuli leading to pain, itch and cough. However, the mechanism(s) of channel modulation by many of the known, non-reactive modulators has not been fully elucidated. N-Cinnamoylanthranilic acid derivatives (CADs) contain structural elements from the TRPA1 modulators cinnamaldehyde and flufenamic acid, so it was hypothesized that specific modulators could be found amongst them and more could be learnt about modulation of TRPA1 with these compounds. A series of CADs was therefore screened for agonism and antagonism in HEK293 cells stably transfected with WT-human (h)TRPA1, or C621A, F909A or F944A mutant hTRPA1. Derivatives with electron-withdrawing and/or electron-donating substituents were found to possess different activities. CADs with inductive electron-withdrawing groups were agonists with desensitising effects, and CADs with electron-donating groups were either partial agonists or antagonists. Site-directed mutagenesis revealed that the CADs do not undergo conjugate addition reaction with TRPA1, and that F944 is a key residue involved in the non-covalent modulation of TRPA1 by CADs, as well as many other structurally distinct non-reactive TRPA1 ligands already reported.

Full text of DOI:10.1016/j.ejmech.2019.02.074

Accepted Manuscript
N-cinnamoylanthranilates as human TRPA1 modulators: Structure-activity
relationships and channel binding sites
Arundhasa Chandrabalan, Martin J. McPhillie, Alyn H. Morice, Andrew N. Boa, Laura
R. Sadofsky
PII:
S0223-5234(19)30206-5
DOI:
https://doi.org/10.1016/j.ejmech.2019.02.074
EJMECH 11166
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 9 January 2019
Revised Date: 27 February 2019
Accepted Date: 27 February 2019
Please cite this article as: A. Chandrabalan, M.J. McPhillie, A.H. Morice, A.N. Boa, L.R. Sadofsky,
N-cinnamoylanthranilates as human TRPA1 modulators: Structure-activity relationships and
channel binding sites, European Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/
j.ejmech.2019.02.074.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
1
2
N-Cinnamoylanthranilates as human TRPA1 Modulators: Structure-
Activity Relationships and Channel Binding Sites.
3
4
Arundhasa Chandrabalan a,b, Martin J. McPhillie c, Alyn H. Morice d, Andrew N. Boa a,*,
Laura R. Sadofsky b,*
5
a
6
Department of Chemistry & Biochemistry, University of Hull, Cottingham Road, Hull HU6
7
7RX, UK.
8
b Centre for Atherothrombosis and Metabolic Disease, Hull York Medical School, University
of Hull, Cottingham Road, Hull HU6 7RX, UK.
9
10
11
12
13
c School of Chemistry, University of Leeds, Leeds LS2 9JT, UK
d Respiratory Medicine, Hull York Medical School, Castle Hill Hospital, Cottingham HU16
5JQ, UK.
14
15
16
* Corresponding authors
Email addresses: a.n.boa@hull.ac.uk (ANB), laura.sadofsky@hyms.ac.uk (LRS)
1

ACCEPTED MANUSCRIPT
17
18
Graphical Abstract
19
20
21
22
23
24
25
26
27
28
Highlights
N-Cinnamoyl anthranilate derivatives (CADs) are TRPA1 modulators
CADs with electron-withdrawing groups are TRPA1 agonists with desensitizing effects
CADs with strongly electron-donating groups are TRPA1 antagonists
CADs modulate TRPA1 through non-covalent interactions
F944A mutants show reduced sensitivity towards CADs and many other TRPA1 modulators
2

ACCEPTED MANUSCRIPT
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Abstract
The transient receptor potential ankyrin 1 (TRPA1) channel is a non-selective cation channel,
which detects noxious stimuli leading to pain, itch and cough. However, the mechanism(s) of
channel modulation by many of the known, non-reactive modulators has not been fully
elucidated. N-cinnamoylanthranilic acid derivatives (CADs) contain structural elements from
the TRPA1 modulators cinnamaldehyde and flufenamic acid, so it was hypothesized that
specific modulators could be found amongst them and more could be learnt about modulation
of TRPA1 with these compounds. A series of CADs was therefore screened for agonism and
antagonism in HEK293 cells stably transfected with WT-human (h)TRPA1, or C621A,
F909A or F944A mutant hTRPA1. Derivatives with electron-withdrawing and/or electron-
donating substituents were found to possess different activities. CADs with inductive
electron-withdrawing groups were agonists with desensitizing effects, and CADs with
electron-donating groups were either partial agonists or antagonists. Site-directed
mutagenesis revealed the CADs do not undergo conjugate addition reaction with TRPA1 and
reveal that F944 is a key residue involved in the non-covalent modulation of TRPA1 by
CADs, as well as many other structurally distinct non-reactive TRPA1 ligands already
reported.
46
47
48
49
50
51
52
53
Keywords: transient receptor potential ankyrin 1; TRPA1; TRP; N-cinnamoylanthranilic
acid; tranilast; calcium signaling; binding site; non-covalent
Abbreviations Used
[Ca2+]i, intracellular calcium ion; CAD, N-cinnamoylanthranilic acid derivative; hTRPA1,
human transient receptor potential ankyrin 1; hTRPM8, human transient receptor potential
melastatin 8.
3

ACCEPTED MANUSCRIPT
54
4

ACCEPTED MANUSCRIPT
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
1. Introduction
The human transient receptor potential ankyrin 1 (hTRPA1) channel [1] is predominantly
expressed in the primary afferent (sensory) neurons [2] including trigeminal [3], dorsal root
[4] and nodose ganglia, [5] and acts as a detector of chemical, mechanical and thermal stimuli
[6]. TRPA1 is known to be involved in pain, itch, [7, 8] inflammatory diseases like arthritis
[9], chronic obstructive pulmonary disease (COPD), and other respiratory diseases including
chronic cough and asthma [6, 10]. There are many compounds known to modulate TRPA1,
among them compounds containing reactive electrophilic groups that activate the channel
through covalent modification of cysteine or lysine residues at the N-terminus of the protein
[11, 12]. However, TRPA1 is also modulated by compounds through non-covalent
interactions only. Mutagenesis and chimeric approaches have been utilized to elucidate the
interactions of individual non-reactive chemicals with the channel [12-17]. However, the
non-covalent mechanism(s) of modulation (activation or blocking) by compounds with
distinct structures are currently much less well determined, hindering our understanding of
the role of TRPA1 in diseases, and development of drugs targeting the channel.
Analogues of N-cinnamoylanthranilic acid (1) have shown a range of biological activities,
including anti-allergic, anti-histaminic, anti-inflammatory, anti-asthmatic, [18, 19] anti-
oxidant, [20] anti-fibrotic [21], anti-proliferative, [22] anti-cancer, [23] anti-platelet, [24]
anti-coagulant [25] and as modulators of B-cells and T-cells, [26, 27], and are under
investigations
for
several
medicinal
applications.
The
derivative
N-(3,4-
dimethoxycinnamoyl)anthranilic acid (11, Tranilast™ or Rizaben™) is an anti-inflammatory
drug that has been used in South Korea and Japan for over twenty years to treat allergic
diseases such as bronchial asthma, allergic rhinitis, hypertrophic scars and scleroderma [18,
21, 28]. N-cinnamoylanthranilate derivatives (CADs) also contain structural elements from
two classes of known TRPA1 modulators, namely the α,β-unsaturated carbonyl moiety found
5

ACCEPTED MANUSCRIPT
80
81
in the agonist cinnamaldehyde (CA), [29] and the anthranilate moiety found in the agonist
flufenamic acid (FFA) and related fenamates [30]. It was therefore anticipated that discovery
of CAD agonists or antagonists of TRPA1 could be used to learn about the modulation of the
channel by non-reactive compounds.
82
83
84
CADs have previously been shown to inhibit other TRP channels. For example N-(p-
amylcinnamoyl)anthranilic acid (ACA), a PLA2 inhibitor, [31] was characterized as a novel
TRPM2, TRPM8, TRPC3, TRPC6 and TRPV1 channel blocker [32]. Similarly, N-(3-
methoxyphenyl)-4-chlorocinnamide (SB366791) selectively blocks TRPV1 [33], yet is
inactive against TRPM8 [34]. The anthranilates 5-nitro-2-(phenethylamino)benzoic acid
85
86
87
88
89
(NPEB),
5-nitro-2-(phenethylamino)benzamide
(NPBA)
and
5-nitro-2-(3-
90
phenylpropylamino)benzoic acid (NPPB), the latter which bears some overall similarity to
CAD 1), selectively activates TRPA1 [35]. Therefore, to investigate both the structure-
activity relationship (SAR) and pharmacology of CADs as TRPA1 modulators, we set out to
synthesize and screen a series of compounds against TRPA1-transfected HEK293 cells.
91
92
93
94
95
2. Materials and Methods
96
2.1 Materials.
97
2.1.1 Commercial TRPA1 modulators. Cinnamaldehyde (CA, ≥ 95 %, natural), acrolein
(ACR, analytical standard), allyl isothiocyanate (AITC, analytical standard), menthol (99 %),
(˗)-menthol and 5-nitro-2-(3-phenylpropylamino)benzoic (NPPB) were purchased from
Sigma-Aldrich. Calcimycin (calcium ionophore, A23187), thymol (> 99.5 %), carvacrol (98
%), eugenol (99 %), cyclohexylcarbamic acid 3ʹ-carbamoyl-biphenyl-3-yl ester (URB597, ≥
98 %), (-)-nicotine (≥ 99 %), N-(p-amylcinnamoyl)anthranilic acid (ACA, ≥ 98 %),
probenecid, 4-(4-chlorophenyl)-3-methylbut-3-en-2-oxime (AP18) and 2-(1,3-dimethyl-2,6-
dioxo-1,2,3,6-tetrahydro-7H-purin-7-yl)-N-(4-isopropylphenyl)acetamide (HC030031) were
98
99
100
101
102
103
104
6

ACCEPTED MANUSCRIPT
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
purchased from Sigma-Aldrich. (1E,3E)-1-(4-fluorophenyl)-2-methylpent-1-en-3-one oxime
(A967079), N-(3-aminopropyl)-2-((3-methylbenzyl)oxy)-N-(thiophen-2-ylmethyl)benzamide
hydrochloride (AMTB.HCl) and SB366791 were purchased from Tocris Bioscience. Ethyl
((1R,2S,5R)-2-isopropyl-5-methylcyclohexane-1-carbonyl)glycinate (WS5) was from
Millennium Specialty Chemicals Inc., Procter & Gamble, flufenamic acid (analytical
standard) was from Fluka Analytical, mefenamic acid (98 %) was from Johnson Matthey
Company and diclofenac (> 98 %) was from Tokyo Chemical Industry. Nordihydroguaiaretic
acid (NDGA) and farnesylthiosalicylic acid (FTS) were purchased from Santa Cruz
Biotechnology. The structures of these various TRPA1 modulators are shown in Figure 1.
The stock solutions of the compounds were made and serially diluted to lower half-log scale
concentrations in dimethyl sulfoxide (DMSO, 100 %, analytical reagent grade, Fisher
Scientific), and thus the concentration of DMSO was maintained constant in a given total
volume of sample.
2.1.2 Chemistry.
The Orion Corporation-disclosed aryl sulfonamide derivative (S)-N-(4-chlorobenzyl)-1-((4-
fluorophenyl)sulfonyl)pyrrolidine-2-carboxamide (ASD) [36], the series of CADs and related
derivatives were synthesized in-house using standard methods (see Supporting Information
for details). Briefly, three synthetic routes were followed to synthesize the CADs. In one
approach, the cinnamic acid derivative was converted to its corresponding acid chloride,
which was coupled to methyl anthranilate, and the resulting methyl N-cinnamoylanthranilate
was then hydrolyzed to yield the corresponding N-cinnamoylanthranilic acid derivative
(Scheme 1). In the second approach, Meldrum's acid was reacted with anthranilic acid to
produce 2-[(carboxyacetyl)amino]benzoic acid, which was then condensed with a
benzaldehyde derivative via a piperidine-catalysed Knoevenagel condensation, producing N-
7

ACCEPTED MANUSCRIPT
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
cinnamoylanthranilate as the piperidinium salt which was acidified to yield the final product
(Scheme 2). In the third approach, a secondary amino acid was reacted with cinnamoyl
chloride under basic conditions, and the resulting salt was acidified to yield the product
(Scheme 3). Compound purity was assessed by satisfactory CHN combustion elemental
1
13
microanalysis, H and C NMR spectroscopy, and a constant melting point. Chemical and
spectroscopic data obtained for compounds synthesized (see Supporting Information) agreed
closely with that reported, where available, in the literature. Stock solutions of these
compounds were prepared as described above (Section 2.1.1).
2.2 Cell culture. HEK293 cells stably transfected with either pcDNA3.1(+) constructs
containing cDNA for hTRPA1 [37], hTRPM8 [38], hTRPA1 mutants or pcDNA3 (mock)
were grown in Dulbecco's Modified Eagle's Medium (DMEM) containing 25 mM 4-(2-
hydroxyethyl)-1-piperazine ethanesulfonic acid (HEPES) and 4.5 g l-1 glucose, supplemented
with 2 mM L
-glutamine, 100 U ml-1 penicillin, 100 U ml-1 streptomycin (all from Lonza), 10
% v/v heat inactivated Fetal Bovine Serum (Life Technologies) and 0.25 mg ml-1 geneticin
(G418 Sulfate, Corning). The cells were grown in T75 tissue culture flask (Greiner Bio-One
CELLSTAR) in a humidified cell culture incubator with 5 % CO2 at 37 °C. When the cells
reached ~90 % confluence, they were harvested in phosphate buffered saline (PBS, Sigma-
Aldrich) solution and centrifuged at 205 × g for 4 min for calcium signaling experiments.
2.3 Calcium signaling. The modulatory effects of the CADs were evaluated on hTRPA1,
hTRPM8 and pcDNA3 (mock) transfected-HEK293 cells using a fluorescence-based calcium
signaling assays at room temperature, either in the cuvette-based system or a micro-well plate
system as described below.
8

ACCEPTED MANUSCRIPT
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
Cuvette-based system: A previously reported protocol was followed [38]. Briefly, the cell
pellet obtained from ~90 % confluence in T75 flask was resuspended in 0.5 ml complete
DMEM and incubated with 2.5 µl of 2.5 µg µl-1 Fluo-3AM fluorescent dye (λEx/Em 506/526
nm, Life Technologies) for 30 min at room temperature with gentle rotary mixing (50 rpm).
The cells were washed with PBS by centrifuging as before, and the pellet was resuspended in
isotonic assay buffer [38] (10 mM HEPES, 145 mM NaCl, 5 mM KCl, 1 mM MgCl2.6H2O, 1
mM CaCl2 and 10 mM D-(+)-glucose made in double distilled water, and pH was adjusted
with NaOH) at a density of 5 × 106 cells ml-1. The cell suspension (100 µl) was added to
cuvettes containing the assay buffer (1.9 ml). The assays were carried out using a PTI
fluorometer with FelixGX version 4.2.2 software. The amount of [Ca2+]i released upon
activation of TRP channels were measured as a real-time-based fluorescence spectrum. The
baseline of a spectrum was recorded for 20 s before the addition of a test compound (agonist)
and the response was recorded for another 90 s, followed by the addition of calcimycin (2
µM) and recording for further 30 s. In antagonist assays, the standard antagonist, the vehicle
control or the test compounds were pre-incubated with cells for 10 min prior to the addition
of a standard agonist, and the spectrum was recorded as before.
Micro-well plate system: 96-Well cell culture microplates (black polystyrene, flat µClear
bottom, Greiner Bio-One) were coated with poly-D-lysine (PDL, MW 70-150 kDa, Sigma-
Aldrich, 50 µg ml-1 prepared in sterile PBS) as follows: PDL was added to the wells (200 µl
cm-2) and incubated for an hour at room temperature, excess PDL solution was aspirated and
the wells were washed with PBS (2 × 100 µl per well), and air-dried at room temperature for
2 h in a safety cabinet. Cell suspension (200 µl of 12500 cells per well in complete DMEM)
was added to the wells and incubated in the cell culture incubator (95% air and 5% CO2 at 37
°C) for 48 h to yield a final concentration of 5 × 104 cells per well. The culture medium was
9

ACCEPTED MANUSCRIPT
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
replaced with 100 µl of 2 µM Fluo-4AM [λEx/Em 494/506 nm (Life Technologies), diluted in
phenol red-free DMEM (Life Technologies) from DMSO stock] and incubated in the dark at
room temperature for 45 min. The cells were rinsed with PBS (2 × 100 µl per well), and the
isotonic assay buffer (100 µl per well for agonist assays or 50 µl per well for antagonist
assays) was added to the wells. In antagonist assays, the cells in 50 µl buffer were incubated
with 50 µl antagonists for 10 min before assaying. The assays were carried out using a
FlexStation3 Molecular Devices with the SoftMax Pro Software version 5. The protocol
utilized was adapted from the literature [39, 40]. Briefly, the real-time-based fluorescence
spectra were recorded using the read-mode/type: Flex fluorescence (RFUs) bottom-read. An
excitation and emission wavelengths of 485 and 538 nm were used for measurements with
530 nm auto cut-off. Each spectrum was recorded for a total run time of 240 s, where the
baseline was recorded for 20 s, and at the end of which the first addition (compound) was
made followed by the second addition (calcimycin) at 180 s.
To determine the reversibility of an antagonist, after pre-incubation, the cells were washed
with the assay buffer and resuspended in fresh assay buffer, and the response for the TRPA1
standard agonist CA (30 µM) was examined. Competitiveness of an antagonist was
determined by pre-incubating cells with the test antagonist compound of a known
concentration for 10 min and by considering the shift in the dose-response curve of CA with
and without the inhibitor.
2.4 Graphical and statistical analyses. The agonism and antagonism responses were
calculated either as a percentage of calcimycin or a standard agonist respectively, using the
differences in the maximum and minimum relative fluorescence unit (∆RFU) values. Initial
screening results correspond to three independent experiments (N = 3), and the errors
10

ACCEPTED MANUSCRIPT
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
reported are the standard error of the mean (SEM). The dose-response curves were plotted
and analyzed using GraphPad Prism version 5.03, and the values of logEC50 and logIC50 were
obtained with their SEM. Each data point on a dose-response curve corresponds to the mean
of three independent experiments (N = 3), performed either in duplicates (n = 6) or triplicates
(n = 9), with their SEM. The responses were normalized by subtracting the noise/response
obtained for the vehicle control (DMSO). In antagonist assays, the maximum response was
obtained by normalising the standard agonist response to 100 %. Thus, in antagonist assays, a
higher percentage value indicates a lower inhibitory effect by the test compound. To
determine statistical significance between groups, the one-way analysis of variance
(ANOVA) at p < 0.05 was carried out.
The standard agonists and antagonists for positive controls were chosen, based on the ion
channel specificity and potency of the ligands, and a concentration near or above EC50 or IC50
were used in the assays. The standard agonists used for the hTRPA1 channels were CA (30
µM), ACR (30 µM) or AITC (10 µM), and for hTRPM8 was WS5 (1 µM). The standard
antagonists used were the potent TRPA1 selective A967079 (100 or 300 nM) and TRPM8
selective AMTB.HCl (1 or 3 µM).
2.5 Indirect measure of inertness towards covalent modification. As CADs could
conceivably react with the TRPA1 via a conjugate addition processes, we wanted first to
disprove that they reacted with a model nucleophile in the absence of the channel (see similar
studies in [41]). Thus compound 1a (4 equiv., 226 mM) and N-acetyl-L-cysteine methyl ester
(1 equiv., 56 mM) or N-acetyl-
L-cysteine (1 equiv., 61 mM) in d6-DMSO were mixed
1
together, and the progress of any reaction was monitored using H NMR spectra recorded at
25 °C and at known time intervals.
11

ACCEPTED MANUSCRIPT
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
2.6 Site-directed mutagenesis. hTRPA1 mutants C621A (site of covalent modification) [13],
F909A or F944A (sites of non-covalent interaction identified by [16, 17]) cDNAs were
produced using a QuikChange Lightening Site-Directed Mutagenesis kit (Agilent
Technologies) as instructed in their manual, using pcDNA3.1(+) as the template and hTRPA1
primers containing the mutants (synthetic oligonucleotides purchased from Eurofins
Genomics MWG Operon). Briefly, hTRPA1 mutants strands were synthesized by performing
18 cycles of PCR, followed by digestion of the parental- and hemi-methylated supercoiled
dsDNA in the amplified products with Dpn I restriction enzyme. The amplified DNA was
then transformed into XL10-Gold ultracompetent cells (E. coli) treated with β-
mercaptoethanol. Plasmid DNA was isolated from the E. coli and purified using a
NucleoSpin® Plasmid DNA purification miniprep kit (Macherey-Nagel) according to the
manufacturer's protocol. The dsDNA samples were quantified (NanoDrop Lite
Spectrophotometer, Thermo Scientific), and sequenced (Eurofins Genomics DNA sequencing
service) and compared against the parent hTRPA1 sequence (sequence ID: Y10601) to
confirm the mutations.
HEK293 cells were transfected with the hTRPA1 mutants dsDNA (0.3 µg µl-1) separately
using FuGENE® 6 Transfection Reagent (Promega) as per the manufacturer's instructions.
The transfection reagent FuGENE and DNA were added in 3:1 ratio. G418 (0.5 mg ml-1)
resistance cells were then expanded in T25 and T75 flasks sequentially. The transfected cells
were single cell cloned by serial dilution in a 96-well cell culture plate following Corning's
procedure. Monoclonal cells were transferred to a 6-well cell culture plate and then into T75
flasks for expansion. The cells were characterized with known hTRPA1 specific agonists and
antagonists using the calcium signaling technique and compared against the WT-hTRPA1
12

ACCEPTED MANUSCRIPT
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
responses to determine the level of TRPA1 protein stably expressed in the transfected
HEK293 cells. The known TRPA1 ligands characterized in the hTRPA1 mutants were CA,
AITC, ACR, menthol, thymol, carvacrol, eugenol, FFA, MFA, DCF, NPPB, URB597, FTS,
(˗)-nicotine, NDGA, probenecid, HC030031, AP18, A967079 and ASD.
2.7 Electrophysiology. Experiments involving animals were performed by IR Pharma
(London, UK). Experiments were performed in accordance with the U.K. Home Office
Guidelines for Animal Welfare based on the Animals (Scientific Procedures) Act of 1986 and
the ARRIVE guidelines [42]. Vagus nerves of eight male Dunkin-Hartley guinea pigs
(300˗400 g) were isolated, characterized and experimented as described previously [43, 44].
Compound stock solutions were prepared in DMSO and diluted 1000× in Krebs-Henseleit
(KH) buffer. The depolarization was recorded in mV. To determine if a compound caused
activation of the nerve, non-cumulative concentration responses to potential tussive stimuli
were carried out. A control response to the TRPV1 agonist capsaicin (1 µM) was carried out
to determine nerve viability, following which the nerve was stimulated with single
concentrations of a test compound for 2 min. After application of each concentration of the
compound, the nerve was washed with KH buffer until it returned to baseline, and this was
repeated with the full range of concentrations. A similar stimulation was also carried out with
the vehicle control (DMSO). A further control response to TRPA1 agonist acrolein (300 µM)
was carried out at the end of the experiment to determine nerve viability.
In order to study antagonism, the nerve was exposed to the agonist ACR (300 µM) for 2 min
and washed with KH buffer until the response returned to baseline. This was repeated to
provide two control agonist responses. The nerve was then pre-treated with a test compound
for 10 min, and then re-stimulated with ACR (300 µM) for 2 min 20 s (the additional 20 s
13

ACCEPTED MANUSCRIPT
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
was to allow for the changeover of stimuli) to assess if the compound was able to affect the
magnitude of the depolarization induced by the ACR. Following a brief washout, the nerve
was exposed to ACR (300 µM) for 2 min to provide a recovery response to ensure nerve
viability and that the compound was washed off.
3. Results
3.1 Calcium signaling. The CADs were evaluated in HEK293 cells stably transfected with
hTRPA1, hTRPM8 or pcDNA3 using a fluorescence-based calcium signaling technique.
Agonism of the compounds were measured by assessing the Ca2+ influx in TRPA1-HEK293
cells upon exposure to the test compounds relative to calcimycin (calcium ionophore,
A23187), and antagonism of the compounds were measured by the ability of the test
compounds to antagonize the agonism of a standard agonist. Hence, the agonism and
antagonism responses of the compounds are presented as the percentage of calcimycin and
standard agonist, respectively (Tables 1 and 2). The vehicle control (0.2 or 0.4 % DMSO) had
no significant activity in hTRPA1 and in mock transfected-HEK293 cells (data not shown),
but had a slight antagonising effect (inhibits ~20 % of WS5 response) in hTRPM8.
Carboxylic acid-containing compounds were evaluated as their corresponding carboxylate
salt due to the isotonic assay buffer (pH 7.4). Even so, for some derivatives insolubility in the
buffer at higher concentrations (300 µM, or in some cases >100 µM) prevented completion of
the dose-response curve (Fig. 3). Most of the CAD esters were partially soluble or insoluble
above 3 µM, and hence screening at higher concentrations was not possible. The CADs, 14
and 15 autofluoresced in the wavelength region of Fluo-3 and Fluo-4 dyes, so screening was
not possible. Due to broad absorption ranges for these two CADs other long-wavelength
calcium dyes were found not to be suitable either.
14

ACCEPTED MANUSCRIPT
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
3.2 CAD Structure-Activity Relationship in hTRPA1. In the initial screening of N-
cinnamoylanthranilic acids (Table 1), the CADs 1-10, 20 and 21 showed ≥25 % agonism at
30 µM and ≥50 % inhibition at 100 µM. The agonists 2, 3, 20 and 21 showed weak
antagonism in the initial screen, relative to other antagonists (e.g. 4, 5 and 6), so full
inhibition curves were not done and IC50 values were not calculated. The CADs 11, 12, and
16-19 showed weak agonism (<25 %) whereas CAD 13 showed no agonism. However, the
CADs 11-13 showed ≥40 % inhibition, and 16-19 showed weak inhibitory effect (<30 %). In
the initial screening of methyl N-cinnamoylanthranilates (Table 2), all the evaluated esters,
including 1a, 3a-8a, 13a-15a, and 20a-22a, were agonists (Table 2). Most of the methyl N-
cinnamoylanthranilate esters at 3 µM showed a similar level of agonism to that of their
corresponding N-cinnamoylanthranilic acid derivatives at 30 µM. However, apart from 20a
and 21a, none of the esters showed antagonism. Dose-response curves were carried out for N-
cinnamoylanthranilic acids that showed higher than 25 % agonism at 30 µM, and for those
which had high antagonism and partial agonism. As can be seen from the curves, the effects
observed were dose-dependent (Fig. 3).
The halogenated CADs 4 and 5, with a bromo or chloro substituent at the para-position,
had similar potency, whereas the fluorinated CAD 7 showed an increase in EC50 and IC50,
relative to other halogens in the series. As the halogenated CADs 4-7 showed potent agonism
and antagonism, it was thought that they might possess desensitising effects following
activation of the channel. This was confirmed by recording the real-time spectra for 10 min
(Fig. 4a,b). Since CA and FFA are known to have desensitising effects, [30, 45] they were
also evaluated along with the CADs for comparison. At the concentration eliciting a
maximum response, the elevated Ca2+ level caused by the agonism was sustained (Fig. 4b),
15

ACCEPTED MANUSCRIPT
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
and at a submaximal concentration the effect dropped over 200-300 s. On continuous
exposure to an agonist a diminished response was obtained (Fig. 4c-f), due to desensitization
of the channel (see [30]). For the CADs tested at 10 µM the agonism response dropped back
nearer to the spectral baseline within 10 min of administration and desensitized the agonism
of the CA (30 µM) standard agonist (Fig. 4c-f). Due to this bimodal activity, potent agonism
and desensitising effect, a washout experiment was carried out to examine the reversibility of
the compounds. The desensitising effect shown by the halogenated CADs on hTRPA1 was
found to be reversible up to 10 µM for 4 and 5, and up to 30 µM for 6 and 7, and were
irreversible at higher concentrations (Fig. 5).
The CADs with a methyl (8) or methoxy (10) group possessed similar agonist activity at
30 µM in the initial screening. However, on comparing the EC50 of 8 and 10, CAD 8
appeared to be less potent. Disubstituted CADs 11 and 12 with electron donating groups
(EDGs) exhibited partial agonist and antagonist activities, whilst 13 showed only an
antagonistic effect and no agonism below 100 µM. Nevertheless, the disubstituted CADs 11
and 12 were less potent antagonists compared to the monosubstituted 13.
Ortho-substituted CADs 6 and 9 were less efficacious, relative to their corresponding
para-substituted analogues 5 and 10. In addition, α,β-saturated derivatives 16 and 17 showed
poor activity at 30 µM compared to the corresponding CADs 1 and 10 with an α,β-
unsaturation. The compounds 18 and 19 with a non-planar ring replacing the anthranilate
moiety showed weak responses relative to the parent compound 1. The α-substituted CADs
either with a methyl or phenyl group were more potent agonists (20 and 21) relative to the
unsubstituted CAD (1). Among the α-substituted CADs 20 and 21, and unsubstituted CAD 1,
the derivative 21 showed pronounced agonist potency.
16

ACCEPTED MANUSCRIPT
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
The N-cinnamoylanthranilic acid methyl esters evaluated in hTRPA1 at 3 µM (Table 2)
showed a similar level of agonism to their corresponding N-cinnamoylanthranilate derivatives
at 30 µM (Table 1), but there were no considerable antagonism shown by the esters, except
for 20a and 21a with desensitising effects.
3.3 Channel selectivity. To determine the channel selectivity of CADs, they were assessed in
pcDNA3 mock-HEK293 cells for agonism, and in hTRPM8-HEK293 cells for both agonism
and antagonism. The CADs, evaluated at 30 and 100 µM, and CAD methyl esters, at 3 µM,
did not show any response in the negative control pcDNA3 mock-transfected HEK293 cells
(N = 3, data not shown).
There was no agonism seen for any CADs in hTRPM8-transfected HEK293 at the
concentrations 30 and 100 µM of N-cinnamoylanthranilates, and 3 µM of methyl N-
cinnamoylanthranilates (data not shown), except for 21 and 21a which exhibited weak
agonism (Tables 1 and 2). The majority of the compounds which were hTRPA1 agonists,
were weak hTRPM8 antagonists at high concentrations. Dose-response curves were carried
out for, 3, 4, 6, 9, 20 and 22a, that showed high antagonism (Tables 1 and 2), amongst which
22a was the most potent hTRPM8 antagonist with IC50 10 µM (Fig. 3g). In addition, it was
observed from the IC50 values that the CAD 4, which showed antagonism in both hTRPA1
and hTRPM8, was more potent in hTRPA1. However, the CADs 6 and 9 had only a 10 µM
difference in the IC50 values obtained against hTRPM8 and hTRPA1 (Table 1).
3.4 Determination of possible binding sites through mutagenesis studies.
17

ACCEPTED MANUSCRIPT
379
380
381
382
383
384
385
386
387
388
389
390
Binding site. CADs could undergo covalent modification due to the presence of an α,β-
unsaturated carbonyl group. Previously we have used NMR scale model reactions of reactive
molecules with cysteine as a quick way of gauging susceptibility of a molecule towards
covalent modification of a TRPA1 cysteine residue, and good correlation with agonism levels
has been observed [41]. Therefore, the propensity of CADs to react through conjugate
addition of the thiol group in cysteine (N-acetyl-L-cysteine methyl ester or N-acetyl-L-
cysteine) to the α,β-unsaturated double bond was evaluated for the representative CAD 1a
(Scheme S4, Supporting information) using a proton NMR time study. This CAD showed no
reactivity to the thiol group of cysteine in d6-DMSO, as monitored by changes in intensity or
appearance of new peaks in the NMR spectrum over 6 h (data not shown). However, to rule
out covalent modification of the channel completely, studies of CADs with C621A mutants
were conducted as part of a study to determine the binding site of the CADs.
391
392
393
394
395
396
397
398
399
400
401
402
403
CADs and a range of other known hTRPA1 active ligands were characterized in HEK293
cells stably transfected with hTRPA1 mutants C621A, F909A or F944A at concentrations
corresponding to the maximum response or closer to the EC50 or IC50 obtained in wild-type
(WT)-hTRPA1 HEK293 cells. The EC50 and IC50 values obtained in the WT-hTRPA1 (Table
3) were broadly consistent with the values found in the original reports, given possible
variations in the assay methods, conditions, species and/or cell lines. Significantly reduced
agonism was obtained, as expected, for the electrophilic compounds, CA, AITC and ACR in
the mutant C621A-hTRPA1. Reduced agonism was also observed for the non-reactive
compounds menthol, carvacrol, FFA, DCF, NPPB, FTS, probenecid and 22a, relative to the
responses obtained in WT-hTRPA1-HEK293 (Fig. 6a-e). With the exception of the saturated
analogue 22a, CADs produced increased responses relative to that observed in the WT-
hTRPA1 (Fig. 6e).
18

ACCEPTED MANUSCRIPT
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
A967079, AP18, CAD 13 and ASD all had reduced antagonistic responses in F909A-
hTRPA1 expressing HEK293 cells relative to WT-TRPA1 expressing cells (Fig. 6f). No
reduction in antagonism was observed for HC030031 in the F909 mutant. Menthol and
carvacrol also showed a reduced agonist effect in the absence of F909, however most of the
other compounds showed a significantly increased agonist response in this mutant compared
to the WT (Fig. 6a-e).
In the mutant F944A-hTRPA1 HEK293 cells a significant drop in activity was found for
ACR, menthol, carvacrol, eugenol and FFA, NDGA, probenecid, CADs 3 and 22a, 8 and 20,
ACA, A967079 and AP18 and ASD, relative to the responses in WT-hTRPA1 (Fig. 6). In the
case of ACA and 3, reduced responses were obtained only at a lower concentration (10 µM)
and not at a higher concentration (30 µM), as shown in Fig. 6e.
3.5 Electrophysiology. The agonist and antagonist effects of the compounds, 3 and 5 at 100
µM, and 13 at 300 µM, were tested on fully characterized isolated guinea pig vagus nerve
preparations. The compounds showed a small degree of activation of the vagus nerve (Fig.
7a-c), with 3 having the largest effect as in the hTRPA1-HEK293 cells. However, the
potency of the responses obtained in tissue (nerve) did not compare to those obtained in
HEK293 cells overexpressing hTRPA1.
423
424
4. Discussion
425
426
427
428
429
TRPA1 is activated by a wide range of stimuli. Some of the chemical modulators activate the
channel via covalent modification of specific residues on the N-terminal of the channel [6, 11,
12], however a large number do not possess the reactive groups to do likewise [12]. Here we
have studied CADs as modulators of TRPA1 activity, and investigated where these
compounds bind to the channel. Comparing trends in the SAR of the acid CADs against
19

ACCEPTED MANUSCRIPT
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
hTRPA1, we noted the parent (unsubstituted) compound 1 had a moderate agonistic effect,
and the electron withdrawing group (EWG)-substituted CADs 2 and 3 were more potent
agonists. The halogenated CADs (4 - 7), with inductive electron withdrawing but lone pair
donating properties showed potent agonism with desensitising effect, and the CADs 8 - 10,
with a weak/moderate electron donating group (EDG), showed bimodal activity, that is they
possessed partial agonism and antagonism. As the electron donating nature of the EDG
becomes stronger the agonism of the compounds decreased and became antagonists (11 - 13).
The potent agonism and antagonism of the halogenated CADs was due to a desensitising
effect following activation of the channel, as was also shown for FFA [30]. This was
reversible at low CAD concentration (<10µM) but irreversible at higher concentrations.
α,β-Saturated analogue 16 and 17 showed poor activity compared to the
corresponding α,β-unsaturated CADs. However it is interesting to note that the related,
flexible derivative NPPB, a classic chloride ion channel antagonist, is a sub-micromolar
agonist of TRPA1 [35]. Unlike the CADs studied, NPPB possesses a nitro group in the
anthranilate ring, and replacing the anthranilate moiety in our CADs with a saturated cyclic
amino acid (18 and 19) produced weak responses. This suggests that the anthranilate moiety
probably plays a key role in the activity of the CADs.
The agonism of structurally related compounds in the literature including ACA, an
inhibitor of several TRP channels, [32], and SB366791, a selective TRPV1 antagonist [33,
34], were evaluated against hTRPA1-HEK293. ACA [32], with a 4-pentyl substituent,
activated hTRPA1 potently relative to the 4-methyl substituted CAD 8. Despite being similar
to CAD 5 (but lacking COOH on the N-aryl ring), SB366791 had no agonism in hTRPA1
(data not shown). A similar trend to that observed for the EWG and EDG substituted CADs,
has also been reported for modulation of hTRPA1 by substituted benzylidenemalononitriles.
[46]
20

ACCEPTED MANUSCRIPT
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
HEK293 cells endogenously express a number of ion channels and receptors, including
purinoceptors (P2Y1 and P2Y2) [47], voltage-gated potassium channels [48, 49], sodium
channels (β1A and Nav1.7) [50, 51], sphingosine-1-phosphate receptors (Edg-1, Edg-3 and
Edg-5) [52], calcium channels [53] including TRPC1, TRPC3, TRPC4, TRPC6 and TRPC7
[54, 55], and M3 muscarinic acetylcholine receptor [56]. The CADs and methyl N-
cinnamoylanthranilate, did not activate the negative control mock-transfected HEK293 cells,
proving that in TRPA1-HEK293 cells the compounds were selective to TRPA1.
TRPM8, a cold-sensitive channel [57], is stimulated by a number of TRPA1 modulators
[58, 59]. Therefore, CADs were evaluated against stably transfected hTRPM8-HEK293 cells
for selectivity between these two channels. CADs active in hTRPM8 contained either a
halogen, an ortho- or an α-substituent. Interestingly, most of the TRPA1 agonists were
hTRPM8 antagonists at higher concentrations. However assays against a range of other ion
channels are necessary to confirm overall channel selectivity.
Looking at the physicochemical properties (Table S1) among similar analogues,
compounds possessing a higher clogP were the more potent agonists, as also seen in
examples in the literature [35, 60]. Generally, the CADs with a higher number of hydrogen
bond donors and acceptors, larger topological polar surface area and logS, and lower logP
had greater inhibitory effects with poor or no hTRPA1 agonism (Table S1).
To determine the binding site(s) of the ligands, mutant TRPA1-expressing HEK293 cell
lines were created. The residues mutated were chosen based on the functional groups on the
ligands and the residues that could participate in either covalent modification (C621 [13]) or
21

ACCEPTED MANUSCRIPT
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
through non-covalent aromatic interactions (F909 [16] and F944 [17]). C621A-hTRPA1
showed significantly reduced agonism responses for electrophilic compounds, in line with
previous reports [11, 12]. We also observed significantly diminished agonism with non-
reactive compounds. A similar type of effect had been previously described for menthol,
NPPB and FTS [35]. However the reduced effects observed with NPPB and FTS in the
mutant C621A were not statistically significant, though a significant reduction in agonism
was observed with the mutations of different cysteine residues at the N-terminus [35, 61]. The
TRPA1 antagonists could not be evaluated in this mutant, since the hTRPA1 standard
1
agonists CA, AITC and ACR bind to the mutated site. In addition to the H NMR study
carried out with CAD 1a, the findings with the C621A mutant further suggest that CADs do
not undergo covalent modification at the TRPA1 N-terminal cysteine. The reduced agonism
observed for CAD 22a is however an anomaly, but this can be explained by chemical
reactivity observed for this ester. When attempting to hydrolyse 22a, to get the free acid, the
compound eliminated the phenoxy group, presumably via an E1cb mechanism, to yield an
acrylamide derivative. With reduced conjugation and no β-substituent, this compound would
be expected to be much more likely to undergo covalent addition reaction compared to a
cinnamide. Thus, the change in modulation of 22a in the C621A mutant could be explained if
such a process also takes place inside the cell.
For F909A-hTRPA1, significantly reduced inhibition was observed as expected for
A967079 relative to the WT-hTRPA1 response [16]. Significantly reduced antagonism was
also observed for the structurally similar oxime AP18. Since the tested antagonists with
distinct structures, including CAD 13 and ASD, also showed significant differences in this
mutant, F909 residue appears to be a key residue for antagonist binding in hTRPA1. Both
CAD 13, ASD and HC030031 incorporate two aromatic rings joined by a linking group. The
22

ACCEPTED MANUSCRIPT
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
reduced antagonism was not observed for HC030031, but this could be explained by the fact
that the aromatic methyl xanthine half of the molecule is a larger bicyclic heteroaromatic ring
structure, as opposed to a single benzenoid rings in CAD 13 and ASD; it is therefore likely to
have a different binding site. There was also a drop in the agonism of menthol and carvacrol
in the absence of the F909. As menthol lacks aromaticity and cannot participate in aromatic π
interactions, it appears a more general hydrophobic interaction may be involved here,
possibly also involving residues neighbouring F909, such as T874. For most of the other
compounds the responses were significantly higher than that observed in WT-hTRPA1. This
could be due to changes in the structural conformation of hTRPA1 as a result of the
substitution of the aromatic residue F909 to alanine.
The mutant F944A-hTRPA1-HEK293 cells displayed a significantly reduced response to
many of the ligands evaluated, including the non-aromatic compounds ACR and menthol that
cannot participate in π-π interactions. The cryo-EM structure of TRPA1 shows that F944 is
buried within the channel structure [16], and apparently closed to binding without
reorganization of the channel structure. Samanta et al [62] used limited proteolysis and mass
spectrometry to evaluate the effect of a small number of non-electrophilic channel modulators
(menthol, antagonist A967079, and agonist PF-4840154) on mouse TRPA1. In that work,
several regions of TRPA1 were proposed as experiencing conformational change upon
binding of the ligands studied. However, F944 and the S5 and S6 helices were not amongst
those implicated. Nevertheless our results, using a wider range of modulators and with
hTRPA1, indicate that it is possible that the reduced activities observed for our F944A
mutant is also indicative of significant structural reorganizations of the S5 and S6 helices
upon ligand binding. The responses obtained for A967079 in both F909 and F944 mutants
were equal. However, the analogue AP18 had significant differences in response despite the
23

ACCEPTED MANUSCRIPT
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
close similarity in their structures; the AP18 inhibition was nearly abolished in the mutant
F944A, but not in F909A. ASD, however, preferentially binds F909 over F944, possibly due
to its greater size, site accessibility and location of the amino acid in the channel structure.
The residues F909 and F944 from each TRPA1 subunit are located in a ‘ring’ amongst the
S5 and S6 helices, a portion of the channel which is thought to be membrane bound [16] (Fig.
8a-c). F909 is located at the bottom of a shallow pocket, whereas F944 is buried within the
tertiary structure of the channel (Fig. 8c-d). Whilst previously EM studies have suggested the
location of the A967079 binding site [16], this area is buried behind the F909 residue in the
atomic model 3J9P. The drop-in activities of modulators in F909A and F944A demonstrated
that the corresponding compounds may possibly bind to the aromatic phenylalanine in the
hTRPA1 putative pore region non-covalently via π-π stacking and/or hydrophobic
interactions. Whilst further structural biology will be needed to elucidate detailed ligand
binding, derivatives with hydroxy groups, anthranilic acid derivatives such as non-steroidal
anti-inflammatory drugs, CADs and aryl sulfonamide derivatives, showed significant
reductions in responses in the mutant F944A, suggesting this residue might serve as a general
determinant in the modulation of hTRPA1. Complete elimination of the responses were not
observed with most of the ligands suggesting they may also interact with other nearby
residues.
The most potent CADs (3, 5 and 13) were also characterized on isolated guinea pig vagus
nerves. Unfortunately, the potency seen in the TRPA1-HEK293 cells was not observed in the
nerve tissue. ACR was used as the TRPA1 standard agonist to determine the inhibitory
effects of the compounds in guinea pig vagus nerve tissue, as desensitization of responses
were observed with the use of CA in depolarization measurements [45, 63]. As a positive
control, incubation with the TRPA1 standard antagonist HC030031 inhibited the
24

ACCEPTED MANUSCRIPT
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
depolarization induced by ACR. However, contrary to the antagonistic effects observed in
hTRPA1-HEK293 cells, neither of the compounds 5 and 13 exhibited any antagonism against
depolarization of the guinea pig vagus nerve by ACR. This discrepancy in results could be
due to, but not limited to, experimental differences or a consequence of species difference.
There is only 79% identity between the amino acid sequences of hTRPA1 and guinea pig
TRPA1. The low sequence homology with 21% variation could cause differences in
responses. Caffeine, menthol and thioaminals (e.g. 4-methyl-N-[2,2,2-trichloro-1-(4-nitro-
phenylsulfanyl)-ethyl]-benzamide), are potent antagonists in hTRPA1 but showed reduced
potency, agonism or inactivity in rodent TRPA1 [64-66]. Similarly, rodent TRPA1 is
activated by cold, whereas human and rhesus monkey TRPA1 are not [67]. The menthol and
cold species difference were attributed to a single residue, V875 in primates and G878 in
rodents [67].
A possible correlation is also noted between the various observed biological activities of
CADs and TRP channels. This is plausible due to the known participation of TRP channels in
a wide range of cellular functions [68]. Key to this suggestion is a study [69] which reports
that TRPA1 is necessary for TGF-β signaling. Loss of the receptor significantly suppresses
the mRNA expression of the inflammatory cytokines, IL-6, α-smooth muscle actin involved
in fibrosis, substance P involved in inflammation, VEGF, collagen Iα1 and the
phosphorylation of kinases induced by TGF-β, and thereby results in attenuation of
fibrogenic and inflammatory reactions. Likewise, in the literature, the anti-allergic, anti-
inflammatory [18, 19], anti-oxidant [20], anti-fibrotic [70], anti-proliferative [22] and anti-
cancer [23] properties of CADs were related to inhibition of cytokines, chemokines and
growth factors. Therefore the fundamental reason behind attenuation of cytokines and growth
factors by CADs, might be the consequence of effects imposed by CADs on TRP channels.
25

ACCEPTED MANUSCRIPT
580
581
582
583
584
585
586
587
588
589
590
591
592
5. Conclusions
The series of CADs possess various activities in hTRPA1, ranging from agonism, partial
agonism, antagonism and desensitizing effects. The structurally related compound, ACA, an
inhibitor of TRPM2, TRPM8, TRPC3, TRPC6 and TRPV1 [32] activates hTRPA1 with
similar potency to those of other CADs evaluated in the study. CAD 13 (p-OH), with a strong
EDG-substituent, and CAD 21 (α-Ph), show promise as a selective hTRPA1 antagonist and
agonist, respectively. Using the key SARs found in the study, the structures of the
compounds could be optimized to make CADs more potent modulators of TRPA1. Moreover,
F944 was found to be a key residue for many TRPA1 active modulators, including menthol,
anthranilates including fenamates and CADs, and ASD. Activity for these modulators was
much reduced in the F944A mutant, indicating that this and other residues (F909) within the
TRPA1 S5-S6 putative pore region are crucial for chemosensation.
593
594
Conflicts of Interest
595
596
597
598
599
600
601
602
603
604
Theauthorsdeclarenoconflictofinterest.
Author Contributions
A.C., L.R.S., A.N.B. and A.H.M. designed the study overall. A.C. performed the chemical
and pharmacology experiments, and wrote the initial draft of the manuscript. L.R.S., A.N.B.,
A.C. and A.H.M. contributed to the interpretation of the chemical and pharmacological data,
and M.J.McP. performed the molecular modelling study and interpretation of the combined
results therefrom. The final manuscript was reviewed and approved by all authors.
Acknowledgements
26

ACCEPTED MANUSCRIPT
605
606
607
608
609
610
611
612
613
614
615
616
617
Thanks are due to Dr Kevin Morgan for his guidance with the pharmacology, Dr Kevin
Welham, Mrs Carol Kennedy, Mr Daniel Mackenzie-Wade and Mr Christopher Crow for
their technical support. We thank Prof Colin Fishwick (University Leeds) for useful
discussions. We are also grateful to Dr Sara Bonvini, Dr Mark Birrell and Prof. Maria Belvisi
of IR Pharma (http://www.irpharma.co.uk/), for designing and performing the guinea pig
vagus nerve experiments. This research was supported by the University of Hull with a
tuition fee bursary to AC.
Supporting Information
The following supplementary data to this article can be found online at: http://xxxxxxxxxxxx
1. Chemical synthesis procedures. 2. Characterization data: NMR, MS, EA and m.p. for the
compounds synthesized. 3. Table S1: Physicochemical properties of the CADs evaluated. 4.
1H and 13C NMR spectra of the compounds prepared.
27

ACCEPTED MANUSCRIPT
618
References
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
[1] D. Jaquemar, T. Schenker, B. Trueb, An ankyrin-like protein with transmembrane
domains is specifically lost after oncogenic transformation of human fibroblasts, Journal of
Biological Chemistry 274 (1999) 7325–7333.
[2] G.M. Story, A.M. Peier, A.J. Reeve, S.R. Eid, J. Mosbacher, T.R. Hricik, T.J. Earley,
A.C. Hergarden, D.A. Andersson, S.W. Hwang, P. McIntyre, T. Jegla, S. Bevan, A.
Patapoutian, ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by
cold temperatures, Cell 112 (2003) 819–829.
[3] M.M. Salas, K.M. Hargreaves, A.N. Akopian, TRPA1-mediated responses in trigeminal
sensory neurons: interaction between TRPA1 and TRPV1, European Journal of Neuroscience
29 (2009) 1568–1578.
[4] J. Hjerling-Leffler, M. Alqatari, P. Ernfors, M. Koltzenburg, Emergence of functional
sensory subtypes as defined by transient receptor potential channel expression, The Journal of
Neuroscience 27 (2007) 2435–2443.
[5] M. Brozmanova, F. Ru, L. Surdenikova, L. Mazurova, T. Taylor-Clark, M. Kollarik,
Preferential activation of the vagal nodose nociceptive subtype by TRPA1 agonists in the
guinea pig esophagus, Neurogastroenterology and Motility 23 (2011) e437–445.
[6] P.M. Zygmunt, E.D. Högestätt, Mammalian transient receptor potential (TRP) cation
channels, in: B. Nilius, V. Flockerzi (Eds.) Handbook of Experimental Pharmacology,
Springer-Verlag, Berlin Heidelberg, 2014, pp. 583–630.
[7] D.M. Bautista, S.E. Jordt, T. Nikai, P.R. Tsuruda, A.J. Read, J. Poblete, E.N. Yamoah,
A.I. Basbaum, D. Julius, TRPA1 mediates the inflammatory actions of environmental
irritants and proalgesic agents, Cell 124 (2006) 1269–1282.
[8] B. Liu, J. Escalera, S. Balakrishna, L. Fan, A.I. Caceres, E. Robinson, A. Sui, M.C.
McKay, M.A. McAlexander, C.A. Herrick, S.E. Jordt, TRPA1 controls inflammation and
pruritogen responses in allergic contact dermatitis, FASEB Journal 27 (2013) 3549–3563.
[9] A. Horvath, V. Tekus, M. Boros, G. Pozsgai, B. Botz, E. Borbely, J. Szolcsanyi, E. Pinter,
Z. Helyes, Transient receptor potential ankyrin 1 (TRPA1) receptor is involved in chronic
arthritis: in vivo study using TRPA1-deficient mice, Arthritis Research & Therapy 18 (2016)
6.
[10] B.F. Bessac, S.E. Jordt, Breathtaking TRP channels: TRPA1 and TRPV1 in airway
chemosensation and reflex control, Physiology 23 (2008) 360–370.
[11] A. Hinman, H.H. Chuang, D.M. Bautista, D. Julius, TRP channel activation by
reversible covalent modification, Proceedings of the National Academy of Sciences of the
United States of America 103 (2006) 19564–19568.
[12] L.J. Macpherson, A.E. Dubin, M.J. Evans, F. Marr, P.G. Schultz, B.F. Cravatt, A.
Patapoutian, Noxious compounds activate TRPA1 ion channels through covalent
modification of cysteines, Nature 445 (2007) 541–545.
[13] P.K. Bahia, T.A. Parks, K.R. Stanford, D.A. Mitchell, S. Varma, S.M. Stevens, T.E.
Taylor-Clark, The exceptionally high reactivity of Cys 621 is critical for electrophilic
activation of the sensory nerve ion channel TRPA1, Journal of General Physiology 147
(2016) 451–465.
[14] B. Xiao, A.E. Dubin, B. Bursulaya, V. Viswanath, T.J. Jegla, A. Patapoutian,
Identification of transmembrane domain 5 as a critical molecular determinant of menthol
sensitivity in mammalian TRPA1 channels, The Journal of Neuroscience 28 (2008) 9640–
9651.
[15] K. Nakatsuka, R. Gupta, S. Saito, N. Banzawa, K. Takahashi, M. Tominaga, T. Ohta,
Identification of molecular determinants for a potent mammalian TRPA1 antagonist by
utilizing species differences, Journal of Molecular Neuroscience 51 (2013) 754–762.
28

ACCEPTED MANUSCRIPT
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
[16] C.E. Paulsen, J.P. Armache, Y. Gao, Y. Cheng, D. Julius, Structure of the TRPA1 ion
channel suggests regulatory mechanisms, Nature 520 (2015) 511–517.
[17] G. Klement, L. Eisele, D. Malinowsky, A. Nolting, M. Svensson, G. Terp, D. Weigelt,
M. Dabrowski, Characterization of a ligand binding site in the human transient receptor
potential ankyrin 1 pore, Biophysical Journal 104 (2013) 798–806.
[18] K. Harita, Y. Ajisawa, K. Iizuka, Y. Kinoshita, T. Kamijo, M. Kobayashi, Aromatic
carboxylic amide derivatives, US3940422A (1976).
[19] M. Spiecker, I. Lorenz, N. Marx, H. Darius, Tranilast inhibits cytokine-induced nuclear
factor κB activation in vascular endothelial cells, Molecular Pharmacology 62 (2002) 856–
863.
[20] Y. Miyachi, S. Imamura, Y. Niwa, The effect of tranilast of the generation of reactive
oxygen species, Journal of Pharmacobio-Dynamics 10 (1987) 255–259.
[21] D.J. Kelly, S.J. Williams, S. Zammit, Halogenated analogues of anti-fibrotic agents,
WO2009079692A1 (2009).
[22] H. Ogita, Y. Isobe, H. Takaku, R. Sekine, Y. Goto, S. Misawa, H. Hayashi, Synthesis
and structure-activity relationship of diarylamide derivatives as selective inhibitors of the
proliferation of human endothelial cells, Bioorganic & Medicinal Chemistry 10 (2002) 3473–
3480.
[23] D. Raffa, B. Maggio, F. Plescia, S. Cascioferro, S. Plescia, M.V. Raimondi, G. Daidone,
M. Tolomeo, S. Grimaudo, A. Di Cristina, R.M. Pipitone, R. Bai, E. Hamel, Synthesis,
antiproliferative activity, and mechanism of action of a series of 2-{[(2E)-3-phenylprop-2-
enoyl]amino}benzamides, European Journal of Medicinal Chemistry 46 (2011) 2786–2796.
[24] Y. Iwasa, T. Iwasa, K. Matsui, T. Yoshimura, N. Tanaka, K. Miyazaki, Anti-platelet
action of an anti-allergic agent, N-(3',4'-dimethoxycinnamoyl)anthranilic acid (tranilast),
European Journal of Pharmacology 120 (1986) 231–234.
[25] K. Miyazawa, J. Fukuyama, K. Misawa, S. Hamano, A. Ujiie, Tranilast antagonizes
angiotensin II and inhibits its biological effects in vascular smooth muscle cells,
Atherosclerosis 121 (1996) 167–173.
[26] L. Steinman, M. Platten, P.-K. Ho, M.L. Selley, Tranilast as modulator of T cell
functioning for use in the treatment of autoimmune diseases, EP2253313A1 (2010).
[27] M.L. Selley, J.J. Inglis, R.O. Williams, A method of modulating B cell functioning,
WO2006053390 (2006).
[28] N.L. Occleston, S. O'Kane, N. Goldspink, M.W. Ferguson, New therapeutics for the
prevention and reduction of scarring, Drug Discovery Today 13 (2008) 973–981.
[29] M. Bandell, G.M. Story, S.W. Hwang, V. Viswanath, S.R. Eid, M.J. Petrus, T.J. Earley,
A. Patapoutian, Noxious cold ion channel TRPA1 is activated by pungent compounds and
bradykinin, Neuron 41 (2004) 849–857.
[30] H. Hu, J. Tian, Y. Zhu, C. Wang, R. Xiao, J.M. Herz, J.D. Wood, M.X. Zhu, Activation
of TRPA1 channels by fenamate nonsteroidal anti-inflammatory drugs, Pflügers Archiv:
European Journal of Physiology 459 (2010) 579–592.
[31] L. Liu, Regulation of lung surfactant secretion by phospholipase A2, Journal of Cellular
Biochemistry 72 (1999) 103–110.
[32] C. Harteneck, H. Frenzel, R. Kraft, N-(p-Amylcinnamoyl)anthranilic acid (ACA): a
phospholipase A2 inhibitor and TRP channel blocker, Cardiovascular Drug Reviews 25
(2007) 61–75.
[33] M.J. Gunthorpe, H.K. Rami, J.C. Jerman, D. Smart, C.H. Gill, E.M. Soffin, S. Luis
Hannan, S.C. Lappin, J. Egerton, G.D. Smith, A. Worby, L. Howett, D. Owen, S. Nasir, C.H.
Davies, M. Thompson, P.A. Wyman, A.D. Randall, J.B. Davis, Identification and
characterisation of SB-366791, a potent and selective vanilloid receptor (VR1/TRPV1)
antagonist, Neuropharmacology 46 (2004) 133–149.
29