Identification of a Partial and Selective TRPV1 Agonist CPIPC for Alleviation of Inflammatory Pain

Transient receptor potential vanilloid 1 (TRPV1) is a non-selective cation channel, predominantly expressed in a subset of peripheral sensory neurons for pain signaling. Topical application of agonist capsaicin for desensitizing TRPV1 currents has been approved for relief of chronic pain. However, the potent TRPV1 capsaicin is not ingestible and even topical capsaicin causes common side effects such as skin irritation, swelling, erythema and pruritus, suggesting that a mild TRPV1 agonist might be helpful for reducing side effects while reliving pain. In this study, we reported on a partial and selective TRPV1 agonist 4-(5-chloropyridin-2-yl)-N-(1H-indazol-6-yl)piperazine-1-carboxamide named CPIPC that was modified based on targeting the residue Arg557, important for conversion between the channel antagonism and agonism. Whole-cell patch clamp recordings indicated a concentration-dependent activation of TRPV1 currents by CPIPC with an EC50 of 1.56 ± 0.13 μM. The maximum efficacy of CPIPC (30 μM) was about 60% of saturated capsaicin (10 μM). Repetitive additions of CPIPC caused TRPV1 current desensitization in both TRPV1-expressing HEK293 cells and dorsal root ganglion (DRG) sensory neurons. Oral administration of CPIPC dose-dependently alleviated inflammatory pain in mice. Further site-directed mutagenesis combined with molecular docking revealed that residue Arg557 is critical for TRPV1 activation by CPIPC. Taken together, we identified a novel partial and selective TRPV1 agonist CPIPC that exhibits antinociceptive activity in mice.

Preclinical studies or clinical trials indicate that genetic or pharmacological inhibition of TRPV1 can cause hyperthermia through the mechanism elevating heat sensing threshold [15][16][17]. In contrast, a potent TRPV1 agonist capsaicin ( Figure 1) by topical application The cryo-EM structural investigations illustrate that both TRPV1 agonists (capsaicin and RTX) and antagonists (capsazepine) can fit into the same binding pocket consisting of residues such as Tyr511, Met514, Leu518, Leu547, Thr550, Arg557, Glu570, and Leu670 [22]. The structural complexes of TRPV1 with vanilloid ligands reveal that the hydrogen bonds with residue Arg557 are essential for the channel activation by agonist capsaicin or RTX, but not antagonists. This suggests that the Arg557 might be critical for a switch between agonism and antagonism [23]. Therefore, we hypothesized that modification of TRPV1 antagonists for forming a hydrogen bond with Arg557 might lead to identification of partial TRPV1 agonists.
To test this hypothesis, we synthesized and identified a lead compound 4-(5-chloropyridin-2-yl)-N-(1H-indazol-6-yl)piperazine-1-carboxamide (CPIPC) based on hybridization of BCTC and ABT-102 ( Figure 1). BCTC, a representative TRPV1 antagonist with strong affinity, is only a tool molecule used for pharmacological studies because of its poor aqueous solubility, short half-life and limited metabolic stability. ABT-102, 1,3-disubstituted urea-derivative with drug-like properties is another potent and selective TRPV1 antagonist that displays analgesic efficacy in pain models of inflammation, postoperative, osteoarthritic and bone cancer pain [24]. Electrophysiological recordings revealed that CPIPC was a partial and selective TRPV1 agonist, with an EC50 value of 1.56 ± 0.13 μM and an efficacy about 61.82 ± 0.49% of capsaicin-mediated channel current saturation. CPIPC was also able to desensitize TRPV1 currents in DRG sensory neurons. Oral CPIPC showed antinociceptive effects in the mouse model of inflammatory pain. Mechanistically, site-directed mutagenesis combined with molecular docking further confirmed that Arg557 is critical for TRPV1 activation by CPIPC. The cryo-EM structural investigations illustrate that both TRPV1 agonists (capsaicin and RTX) and antagonists (capsazepine) can fit into the same binding pocket consisting of residues such as Tyr511, Met514, Leu518, Leu547, Thr550, Arg557, Glu570, and Leu670 [22]. The structural complexes of TRPV1 with vanilloid ligands reveal that the hydrogen bonds with residue Arg557 are essential for the channel activation by agonist capsaicin or RTX, but not antagonists. This suggests that the Arg557 might be critical for a switch between agonism and antagonism [23]. Therefore, we hypothesized that modification of TRPV1 antagonists for forming a hydrogen bond with Arg557 might lead to identification of partial TRPV1 agonists.
To test this hypothesis, we synthesized and identified a lead compound 4-(5-chloropyridin-2-yl)-N-(1H-indazol-6-yl)piperazine-1-carboxamide (CPIPC) based on hybridization of BCTC and ABT-102 ( Figure 1). BCTC, a representative TRPV1 antagonist with strong affinity, is only a tool molecule used for pharmacological studies because of its poor aqueous solubility, short half-life and limited metabolic stability. ABT-102, 1,3-disubstituted urea-derivative with drug-like properties is another potent and selective TRPV1 antagonist that displays analgesic efficacy in pain models of inflammation, post-operative, osteoarthritic and bone cancer pain [24]. Electrophysiological recordings revealed that CPIPC was a partial and selective TRPV1 agonist, with an EC 50 value of 1.56 ± 0.13 µM and an efficacy about 61.82 ± 0.49% of capsaicin-mediated channel current saturation. CPIPC was also able to desensitize TRPV1 currents in DRG sensory neurons. Oral CPIPC showed antinociceptive effects in the mouse model of inflammatory pain. Mechanistically, site-directed mutagenesis combined with molecular docking further confirmed that Arg557 is critical for TRPV1 activation by CPIPC.

Concentration-Dependent Activation of TRPV1 by the Selective and Partial Agonist CPIPC in TRPV1-Transfected HEK293 Cells
We started testing the effect of CPIPC on hTRPV1 transiently transfected HEK293 cells in calcium-imaging assay. As shown in Figure 2A, application of 1 µM CPIPC or 1 µM capsaicin as a positive control induced a robust Ca 2+ influx in TRPV1 transfected HEK293 cells compared with blank HEK293 cells. Ionomycin, as a positive control of calcium ionophore, resulted in a drastic elevation of intracellular calcium in both TRPV1-overexpressing and non-transfected HEK293 cells. In the FlexStation 3 multi-mode microplate reader assay, CPIPC at 0.1 µM or 10 µM also increased intracellular levels of calcium in TRPV1transfected HEK293 cells, which was completely reversed by a selective TRPV1 antagonist BCTC at 10 µM ( Figure 2B). These results suggest that CPIPC is an agonist of TRPV1.
We started testing the effect of CPIPC on hTRPV1 transiently transfected HEK293 cells in calcium-imaging assay. As shown in Figure 2A, application of 1 μM CPIPC or 1 μM capsaicin as a positive control induced a robust Ca 2+ influx in TRPV1 transfected HEK293 cells compared with blank HEK293 cells. Ionomycin, as a positive control of calcium ionophore, resulted in a drastic elevation of intracellular calcium in both TRPV1overexpressing and non-transfected HEK293 cells. In the FlexStation 3 multi-mode microplate reader assay, CPIPC at 0.1 μM or 10 μM also increased intracellular levels of calcium in TRPV1-transfected HEK293 cells, which was completely reversed by a selective TRPV1 antagonist BCTC at 10 μM ( Figure 2B). These results suggest that CPIPC is an agonist of TRPV1.
To further confirm the activation of TRPV1 by CPIPC, we performed the whole-cell patch clamp recordings on TRPV1 transiently transfected HEK293 cells. As shown in Figures 3A,B, CPIPC at 1 μM activated a large outward rectification TRPV1 current, and the effect could be wash-out before further activation of TRPV1 by capsaicin (1 μM). Perfusing different concentrations of CPIPC from 0.1 μM to 30 μM resulted in a concentration-dependent activation of TRPV1 currents with EC50 of 1.56 ± 0.13 μM and a Hill coefficient of 2.07 ± 0.20 ( Figures 3C,D). The maximum efficacy induced by CPIPC (30 μM) was 61.82 ± 0.49% of the current induced by saturated capsaicin (10 μM) ( Figure S1). We also observed To further confirm the activation of TRPV1 by CPIPC, we performed the whole-cell patch clamp recordings on TRPV1 transiently transfected HEK293 cells. As shown in Figure 3A,B, CPIPC at 1 µM activated a large outward rectification TRPV1 current, and the effect could be wash-out before further activation of TRPV1 by capsaicin (1 µM). Perfusing different concentrations of CPIPC from 0.1 µM to 30 µM resulted in a concentrationdependent activation of TRPV1 currents with EC 50 of 1.56 ± 0.13 µM and a Hill coefficient of 2.07 ± 0.20 ( Figure 3C,D). The maximum efficacy induced by CPIPC (30 µM) was 61.82 ± 0.49% of the current induced by saturated capsaicin (10 µM) ( Figure S1). We also observed a further increase of saturated 30 µM CPIPC-induced currents after adding a mixture of saturated capsaicin (10 µM) and CPIPC (30 µM) ( Figure 3E), indicating CPIPC was a partial TRPV1 agonist. The CPIPC-mediated TRPV1 currents were concentration-dependently inhibited by a selective TRPV1 antagonist capsazepine ranging from 0.1 to 10 µM ( Figure 3F). Repetitive applications of CPIPC (1 µM) caused a progressive reduction of TRPV1 currents by 79.33 ± 0.79% ( Figure 3G,H), indicating a TRPV1 desensitization induced by CPIPC. The CPIPC-induced desensitization also reduced the potency of capsaicin about 10-fold ( Figure S2), demonstrating the reduced TRPV1 activation by capsaicin under CPIPCinduced desensitization.   To evaluate the selectivity of CPIPC, we tested the effect of CPIPC on mTRPV2, hTRPV3, hTRPV4, hTRPA1 and hNav1.7 channels using a whole-cell patch-clamp assay. As shown in Figure 4A, 1 µM CPIPC had no detectable activation of TRPV2 channel that was activated by agonist 2-APB [25]. Similarly, CPIPC at 1 µM caused no activation of TRPV3, TRPV4 or TRPA1 currents ( Figure 4B-D), as compared with the channel activation by their agonists 2-APB for TRPV3, GSK1016790A for TRPV4 or AITC for TRPA1 [25][26][27]. We also evaluated the effect of CPIPC or capsaicin on Nav1.7 channels expressed in HEK293 cells. As shown in Figure S3, 1 µM capsaicin induced an inhibition of Nav1.7 current about 19.14 ± 0.35%, whereas 1 µM CPIPC resulted in reduction of Nav1.7 current about 4.01 ± 0.55%. These results demonstrated that CPIPC was a selective TRPV1 agonist.

TRPV1 Desensitization Induced by Repetitive Applications of CPIPC in DRG Neurons
We next tested the effect of CPIPC on native TRPV1 currents recorded from dissociated mouse DRG neurons at +80 mV. In whole-cell patch-clamp recordings, we observed that 10 of 19 (53%) small-or medium-diameter DRG neurons were sensitive to both capsaicin (10 µM) and CPIPC (10 µM) with an amplitude of 1.37 ± 0.22 nA to CPIPC (n = 6; range of 0.97-2.18 nA). As shown in Figure 5A, CPIPC at 10 µM elicited a large capsaicin-sensitive outward currents in DRG neuron at depolarizing potential at +80 mV, and repetitive 5 times administrations of CPIPC led to a desensitization with 67.72 ± 1.08% reduction ( Figure 5B). These results demonstrate that CPIPC can activate the native TRPV1 currents and also cause the current desensitization upon repetitive applications in capsaicin-sensitive nociceptive DRG neurons.

TRPV1 Desensitization Induced by Repetitive Applications of CPIPC in DRG Neurons
We next tested the effect of CPIPC on native TRPV1 currents recorded from dissociated mouse DRG neurons at +80 mV. In whole-cell patch-clamp recordings, we observed that 10 of 19 (53%) small-or medium-diameter DRG neurons were sensitive to both capsaicin (10 μM) and CPIPC (10 μM) with an amplitude of 1.37 ± 0.22 nA to CPIPC (n = 6; range of 0.97-2.18 nA). As shown in Figure 5A, CPIPC at 10 μM elicited a large capsaicinsensitive outward currents in DRG neuron at depolarizing potential at +80 mV, and repetitive 5 times administrations of CPIPC led to a desensitization with 67.72 ± 1.08% reduction ( Figure 5B). These results demonstrate that CPIPC can activate the native TRPV1 currents and also cause the current desensitization upon repetitive applications in capsaicinsensitive nociceptive DRG neurons.

Antinociceptive Effects of CPIPC on Mouse Model of Inflammatory and Thermal Pain
We tested the antinociceptive effect of CPIPC on formalin-induced pain model with intraplantar injection of 5% formalin into mouse hind paws. As a positive control, preoral administration of a nonsteroid anti-inflammatory drug ibuprofen (2 mg/mL) for 60 min significantly reduced the time of hind paw licking induced by formalin in Phase II but not in Phase I, as compared with the vehicle control ( Figure 6B,C). In contrast, preoral administration of CPIPC at 10 mg/kg significantly relieved pain behavior induced by formalin in Phase I. 3 and 10 mg/kg CPIPC caused a dose-dependent reduction of formalin-induced licking in Phase II. These observations demonstrated an antinociceptive effect of CPIPC on inflammatory pain.
We further tested the anti-nociceptive effect of CPIPC on inflammatory pain induced by intraplantar injection of CFA into mouse hind paws. The paw withdrawal mechanical threshold (PWMT) was measured by von Frey hair assay after injection of CFA for 24 h. As shown in Figure 6D and Table 1, the injection of CFA induced a mechanical allodynia as compared to the control group. In contrast, oral administration of CPIPC (1, 3, 10 mg/kg) dose-dependently increased the PWMT and the effect lasted for about 2 h. As a positive control, ibuprofen (2 mg/kg, p.o.) also reduced CFA-induced inflammatory pain, similar to the anti-nociceptive activity observed in the effect of CPIPC at 3 mg/kg. . The paw withdrawal mechanical threshold (PWMT) of mice was measured after intraplantar injection of CFA for 24 h by von Frey hair. PWMT was measured at 0, 1, 2, 3, 4 h after oral administration of ibuprofen (2 mg/kg) and CPIPC (1, 3, 10 mg/kg). Statistic difference between the vehicle control and different doses of CPIPC groups is presented in Table 1.  The paw withdrawal mechanical threshold (PWMT) of mice was measured after intraplantar injection of CFA for 24 h by von Frey hair. PWMT was measured at 0, 1, 2, 3, 4 h after oral administration of ibuprofen (2 mg/kg) and CPIPC (1, 3, 10 mg/kg). Statistic difference between the vehicle control and different doses of CPIPC groups is presented in Table 1.

Identification of Residue Arg557 Critical for TRPV1 Activation by CPIPC
We adopted the structure of TRPV1 in a complex with RTX (PDB:5IRX) for molecular docking of CPIPC using the Gold 5.3.0 software (University of Cambridge, Cambridge, UK). Our docking shows that CPIPC is inserted into a binding pocket formed by residues between S4 and the S4-S5 linker and the indazole of CPIPC forms two hydrogen bonds with Arg557 and Ala566 ( Figure 7A,B), similar to capsaicin, that also forms a hydrogen bond with the Arg557 in the same binding pocket ( Figure 7C). To test whether the region of CPIPC predicted to contact Arg557 is critical for binding, we made chemical modifications to CPIPC by substituting the indazole ring with indole (CPIPC-1) or methyl (CPIPC-2) groups for disrupting the formation of the hydrogen bond with Arg557 and Ala566, respectively ( Figure 7D). When CPIPC-1 had its indazole ring replaced by indole group it resulted in a decrease of TRPV1 activation. In contrast, for CPIPC-2, the -NH-of indazole was substituted by a methyl group while remaining the hydrogen bond with Arg557 had a stronger activation on TRPV1 ( Figure 7E,F), as compared to CPIPC-1. This indicates that the region of CPIPC predicted to contact Arg557 is critical for activating TRPV1 by CPIPC.
We further constructed a mutant R557A and tested the effect of CPIPC on TRPV1 activation. Mutating Arg557 caused a dramatic attenuation of CPIPC-(or capsaicin-) mediated TRPV1 activation ( Figures 7G and S4), consistent with the observation that residue Arg557 is critical for capsaicin-evoked response [30].

Discussion
Desensitizing TRPV1 is known to induce a long-term defunctionalization of nociceptors for pain relief, serving as an attractive strategy to develop new antinociceptive agents. Over the past years, several TRPV1 agonists have been reported for their antinociceptive activities, including capsaicin, Muralatin L and MDR-652 [21,31,32]. However, potent TRPV1 agonist capsaicin causes an initial adverse effect of pungency and even an 8% capsaicin dermal patch can lead to erythema and pain at the site of application [33]. In contrast, a partial TRPV1 agonist MDR-652 and a subpotent capsaicin analogue YB-16 have been shown to be much weaker in causing pungency [21,34]. These observations suggest developing subpotent TRPV1 partial agonists might be an effective strategy to treat inflammatory and neuropathic pain with less toxicity.
Our observations show that CPIPC can dose-dependently relieve formalin-and CFAinduced nocifensive behaviors in male mice. Injection of formalin into mouse hind paws induces pain responses in two phases: phase I ( Figure 6B), an acute pain stage caused by direct stimulation of C-fiber nociceptors; and phase II ( Figure 6C), the later phase of inflammatory response [32]. Complete Freund's Adjuvant (CFA) is a widely used reagent for generation of rodent models of inflammation. The injection of CFA into the hind paws of mice or rats results in both mechanical allodynia and thermal hyperalgesia within 24 h. The inflammation condition occurs for 5 days before a slow recovery [35,36]. CPIPC (1, 3, 10 mg/kg) dose-dependently alleviated the CFA-induced mechanical allodynia and the effect lasted for about 2 h in mice. CPIPC as a partial TRPV1 agonist is different than other TRPV1 agonists, such as capsaicin which causes burning sensation and is only for topical use. CPIPC can be internally used, holding developmental potential for an antinociceptive agent.
The cryo-EM structures of TRPV1 in complex with capsaicin and capsazepine have provided a framework for predicting agonist binding sites in TRPV1 [23,37]. TRPV1 vanilloid agonists occupy and displace the resident phosphatidylinositol lipid in the binding pocket between S4 and S4-S5 linker. The vanillyl rings of these agonists, such as RTX, function to stabilize the interaction between Arg557 and Glu570 for facilitating the movement of the S4-S5 linker away from the central axis of the channel, thereby opening the lower gate through coupled movements [23]. Such a mechanism is further supported by the mutations in S4/S4-5 regions where R557A, R557K and E570R slow TRPV1 activation rate via capsaicin [30]. Interestingly, the TRPV1 antagonist capsazepine, for example, also can fit into the binding pocket of agonists without its interaction with Arg557. Therefore, Arg557 is a key residue that TRPV1 agonists act on [23,38]. Targeting the Arg557 may lead to identification of TRPV1 agonists of different potency and efficacy.
In this study, our molecular docking reveals that CPIPC and capsaicin share a similar pose in the binding pocket formed by S4 and S4-S5 linker. Either indazole of CPIPC or phenolic hydroxyl group of capsaicin can form a hydrogen bond with Arg557, whereas mutating R557 alters CPIPC-or capsaicin-induced channel sensitivity. Besides R557, residues Y511 and S512 from S3 segment, and M547 and T550 from S4 segment, are also essential for agonistic effect of capsaicin as mutations Y511A or S512Y of rTRPV1 abolish the channel sensitivity to capsaicin. This indicates that these key residues also contribute to CPIPC-mediated activation of TRPV1 [39][40][41].
In general, our identification of CPIPC not only provides a tool molecule but also a lead with therapeutic potential for pain therapy.

Calcium-Imaging
TRPV1 transiently transfected or blank HEK293 cells were seeded on 1 cm 2 glass coverslips for 24 h before being loaded with 4 µM Fluo-8 (ab142773, Abcam, Cambridge, MA, USA) in Ringer's solution (pH 7.2) containing 140 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , 10 mM D-glucose, 10 mM HEPES, and 2 mM CaCl 2 for 30 min in 5% CO 2 at 37 • C. The tested compounds were prepared in Ringer's solution. After being washed twice with Ringer's solution, cells were subsequently treated with CPIPC (1 µM), capsaicin (1 µM), and ionomycin (10 µM). Fluorescent signals were collected by a CCD camera (DS-Qi2) of Nikon powered by CellSens Dimension at 3-s intervals with a 20× objective lens. All experiments were performed at room temperature, and data were expressed as the mean ± SEM.

Measurement of Intracellular Calcium in FlexStation 3 Multi-Mode Microplate Reader Assay
The intracellular calcium level of HEK293 cells expressing TRPV1 channels was measured in the presence or absence of agonists or antagonists using the Screen Quest™ Calbryte-520 Probenecid-Free and Wash-Free Calcium Assay Kit (36318, AAT Bioquest, Sunnyvale, CA, USA) in FlexStation 3 Microplate Reader (Molecular Devices, San Francisco, CA, USA). TRPV1 transiently transfected HEK293 cells were seeded into a black 96-well plate with a density of 30,000 cells/well before cultured in 5% CO 2 at 37 • C overnight. The calbryte™ 520 AM dye-loading solution was prepared by mixing 9 mL HBSS (137 mM NaCl, 5.

Disassociation and Culture of Mouse DRG Neurons
Dorsal root ganglion (DRG) neurons were obtained from 6-8-week-old C57BL/6 mice (Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) as previously described [32,42]. The ganglia were digested with collagenase type 2 (1 mg/mL, Worthington, Lakewood, NJ, USA) and neutral protease dispase II (7.5 mg/mL, Merck, Kenilworth, NJ, USA) at 37 • C with 5% of CO 2 for 30 min. DRG cells were mechanically dissociated and washed with DMEM plus 10% fetal bovine serum. After centrifugation at 1500 rpm, the DRG cells were suspended and plated on poly-D-lysine-coated glass coverslips. Electrophysiological recordings were carried out within 48 h.

Electrophysiology
Whole-cell patch-clamp recordings were performed on HEK293 cells co-expressing TRPV1-4 and GFP under fluorescent light at room temperature (23-25 • C) using a HEKA EPC10 amplifier with PatchMaster software (HEKA Harvard, Holliston, Church Hill, TN, USA). Currents were filtered at 2.9 kHz and sampled at 20 kHz. Borosilicate glass pipettes (BF150-86-100, Sutter Instruments, Novato, CA, USA) were pulled by the horizontal micropipette puller (P-97, Sutter Instruments, Novato, CA, USA) and fire polished by the MF-830 polisher (Narishige, Tokyo, Japan). The electrode resistance was 2~4 MΩ after being filled with pipette solution and liquid junction potential was adjusted to zero after the pipette-in-bath solution. For ramp recordings of HEK293 cells, both bath solution and pipette solution components were (mM): NaCl 130; EDTA 0.2; HEPES 3 [43]. Cells were held at 0 mV after use of whole-cell recording mode before given a ramp stimulus from −100 mV to +100 mV every 1 s. Series resistances (6~15 MΩ) were compensated by 60-80%. Gravity perfusion was administrated and the perfusion rate was adjusted to 0.5 mL/min. It is necessary to divide the whole-cell current (I m ) by the cell capacitance (C m ) to calculate the current density (I d ).

Animals
Adult male C57BL/6 mice (7 weeks) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., (Beijing, China). All experimental procedures were approved by the Animal Ethics Committee of Qilu Medical University and complied with the ethical guidelines of the International Association for the Study of Pain. Animals were housed in a 12 h alternating light/dark cycle room with free access to food and water.

Complete Freund's Adjuvant (CFA)-Induced Inflammatory Pain Model
Adult male C57BL/6 mice were anesthetized with isoflurane and then intraplantarally injected 20 µL complete Freund's adjuvant (CFA) (Sigma, St. Louis, MO, USA) into the right hind paw. In the control group, 20 µL saline was injected into the right hind paw of mice. After injection of CFA for 24 h, CPIPC (1, 3, 10 mg/kg), ibuprofen (2 mg/kg) or saline was orally administered, while mice in the control group were administered the same volume saline. The paw withdrawal mechanical threshold (PWMT) was measured at time points 0, 1, 2, 3 and 4 h after administration of drugs with a von Frey hair by an up-and-down method [35,36]. Mice were individually placed into transparent boxes with a metal mesh for about 1 h. Then, a series of von Frey hairs with different forces (0.4 g as a starter before 0.008, 0.02, 0.04, 0.07, 0.16, 0.60, 1.0, 1.4 and 2.0 g) were applied to stimulate the plantar of hind paws until a bend in a slight S-shape was observed for 5 to 6 s. If a paw withdrawal reaction was observed, the force was marked as positive; otherwise, it was marked as negative. Once crossover was found, the test was continued four times. The PWMT was calculated by a formula: PWMT (g) = 10 [Xf+Kδ] , where Xf is the log value of the last von Frey hair force, K is the value retrieved from the standardized table based on the up-and-down pattern.

Formalin-Induced Inflammatory Pain Model
Adult male C57BL/6 mice were placed individually into transparent cages for about 1 h and then intraplantarally injected 20 µL formalin (5% formalin in vehicle saline) into the plantar surface of right hind paw. The duration of licking behaviors was recorded as a nociceptive response. The licking time was divided into two phases: phase I (0-5 min) and phase II (15-45 min). In the control group, mice were administered saline (0.1 mL/10 g). In the positive control group, ibuprofen (2 mg/kg) was orally administrated into mice for 1 h before formalin injection. In experimental groups, CPIPC (1, 3, 10 mg/kg) was intraperitoneally injected into mice for 1 h before formalin injection.

Measurement of Eye-Wiping
The pungency of capsaicin and CPIPC was evaluated in eye-wiping assay, using a protocol described previously. 10 µL of capsaicin (10 µg/mL) or CPICP (10, 30, 100 µg/mL) was dropped into mouse left eyes. In the vehicle group, the same volume of solvent was dropped into the left eyes of mice. The number of eye wipes was recorded after chemical administration for 60 s.

Randomization and Blinding
In vivo studies, animals were randomly distributed in the different experimental groups with each group having the same number of animals. Investigators for outcome assessments were blinded to group allocation. No blinding or randomization methods were used for in vitro studies.

Molecular Docking
The crystal structure of TRPV1 in a complex with RTX (PDB:5IRX) was obtained from PDB bank (http://www.rcsb.org/, accessed on 10 January 2022). The existing ligand, water and ions were removed from the PDB file. Molecular docking was performed via a Molecular Operating Environment (MOE v201608). The binding pocket was composed of residues Tyr511, Ser512, Leu515, Thr550, Asn551, Leu553, Tyr554, Arg557, Ala566, Ile569, Ile573 and Leu577 between S4 and S4-S5 linker. The 20 best docking results were exported and analyzed. The exported ligand-TRPV1 complexes were edited using UCSF Chimera 1.15 software (RBVI, University of California, San Francisco, CA, USA).

Statistical Analysis
Statistical analysis was performed by GraphPad Prism 8. Each data point is expressed as the mean ± SEM. Dose-response curves were fitted with a hill equation: y = V max x n /(k n + x n ), where V max was the maximum effect, x was the drug concentration, k was EC 50 (half maximal effective concentration) and n was the hill coefficient of sigmoidicity. Statistical analysis for differences between groups was performed by paired Student's t-tests. For multiple comparisons among the groups for antinociceptive experiments, data were analyzed using one-way or two-way ANOVA followed by Bonferroni post hoc tests. A value of p < 0.05 was considered to have statistical significance.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data used to support the findings of this study are available. Further inquiries can be directed to the corresponding authors.