The industrial solvent 1,4-dioxane causes hyperalgesia by targeting capsaicin receptor TRPV1

The synthetic chemical 1,4-dioxane is used as industrial solvent, food, and care product additive. 1,4-Dioxane has been noted to influence the nervous system in long-term animal experiments and in humans, but the molecular mechanisms underlying its effects on animals were not previously known. Here, we report that 1,4-dioxane potentiates the capsaicin-sensitive transient receptor potential (TRP) channel TRPV1, thereby causing hyperalgesia in mouse model. This effect was abolished by CRISPR/Cas9-mediated genetic deletion of TRPV1 in sensory neurons, but enhanced under inflammatory conditions. 1,4-Dioxane lowered the temperature threshold for TRPV1 thermal activation and potentiated the channel sensitivity to agonistic stimuli. 1,3-dioxane and tetrahydrofuran which are structurally related to 1,4-dioxane also potentiated TRPV1 activation. The residue M572 in the S4-S5 linker region of TRPV1 was found to be crucial for direct activation of the channel by 1,4-dioxane and its analogs. A single residue mutation M572V abrogated the 1,4-dioxane-evoked currents while largely preserving the capsaicin responses. Our results further demonstrate that this residue exerts a gating effect through hydrophobic interactions and support the existence of discrete domains for multimodal gating of TRPV1 channel. Our results suggest TRPV1 is a co-receptor for 1,4-dioxane and that this accounts for its ability to dysregulate body nociceptive sensation.

exposure thus harms the upper respiratory passages, symptomized by coughing and stomach pain. However, the in vivo molecular target of 1,4-dioxane and the action mechanism remain to be understood.
Primary sensory neurons in dorsal root, trigeminal, and nodose ganglia initiate pain in response to noxious chemical, mechanical, or thermal stimuli [11]. Transient receptor potential (TRP) channels are calciumpermeable and non-selective cation channels expressed by somatosensory neurons [12]. As molecular sensors for nociceptive stimulation, TRP channels prominently regulate the pathogenesis of both inflammatory and neuropathic pain [13]. The vanilloid receptor 1 (TRPV1) is predominantly expressed by sensory neurons [14]. It is a polymodal nociceptive receptor activated by capsaicin, heat above 42°C and other irritants such as protons and inflammatory mediators [15][16][17], whereby regulating pain perception [18,19]. In addition, Kunkler and colleagues have demonstrated that daily exposures to environmental irritants like acrolein would induce functional sensitization of peripheral neural elements including TRPA1 and TRPV1 channels [20].
Here, using CRISPR-Cas9-engineered mouse model, we show that the nociceptive receptor TRPV1 is targeted by 1,4-dioxane in vivo, resulting in hyperalgesia behaviors. Combining electrophysiology, Ca 2+ imaging, and molecular genetics, we demonstrate that 1,4-dioxane activates TRPV1 to cause Ca 2+ influx in the dorsal root ganglia (DRG) and trigeminal ganglia (TG) sensory neurons. We further show that 1,4-dioxane potentiates the sensitivity of TRPV1 to inflammatory stimuli, while lowers its temperature threshold to thermal activation, hence providing a mechanistic basis for the observed hyperalgesia behavior. Our data suggest TRPV1 channel as a co-receptor for 1,4-dioxane, whereby it may dysregulate the nociceptive sensation.

1,4-Dioxane causes hyperalgesia via TRPV1 activation
1,4-Dioxane penetrates the skin and irritates multiple tissues in humans [9,10]. To explore its effect on pain transduction, we used the Radiant Heat and Von Frey assay to respectively evaluate thermal and mechanical sensation in mouse model. Varying concentrations of 1,4-dioxane as indicated or normal saline (10 μl) were injected into the mouse left hind paw while using the right hind paw as control (10 μl saline only). After 30 min, we counted the withdrawal latency of each hind paw in response to radiant heat as nocifensive reaction (i.e., licking foot). The paw withdrawal latency was significantly reduced along with increased concentrations of 1,4-dioxane (e.g., 7.6 ± 0.7 s for administration of 10% 1,4-dioxane vs. 14.1 ± 0.8 s of the control paw, p < 0.01) (Fig. 1a), reflecting an enhanced sensitivity to high temperature stimulus. In Von Frey assessment, 1,4-dioxane paw injection also decreased the threshold for mechanical pain sensing (e.g., painful weight = 2.1 ± 0.2 g for administration of 10% 1,4-dioxane vs. 5.4 ± 0.4 g of the control paw, p < 0.001) (Fig. 1b). Therefore, 1,4dioxane causes both thermal and mechanical hyperalgesia.
Sensory transduction in mammalian animals is importantly mediated by ion channels, among which the TRP channels expressed by the peripheral sensory neurons play an essential role. We then investigated the effects of 1,4-dioxane on nociceptive TRP ion channels known to be expressed in DRG neurons. TRPV1, TRPA1, TRPV2, TRPM8, TRPV3, and TRPM3 channels were individually expressed in HEK 293 cells and consecutively stimulated by their own agonist and 1,4dioxane. As illustrated in Fig. S1, among the tested TRP channels, only TRPV1 showed response currents in response to 3% or 5% 1,4-dioxane, indicating its direct effect on the TRPV1 channel.
Previous studies have shown that TRPV1 is closely related to mechanical and thermal hypersensitivity [21,22], thus suggesting that 1,4-dioxane may cause thermal and mechanical hyperalgesia by activating TRPV1. To explore the involvement of TRPV1 in 1,4-dioxanecaused hyperalgesia, we used a mouse model where the Trpv1 gene was deleted by CRISPR/Cas9-mediated genome editing (Fig. S2a). Consistent with previous studies, knockout of TRPV1 had no obvious effect on the development of mice [18], nor on their general appearance, gross anatomy, body weight, locomotion, or overt behavior. However, the 1,4-dioxane-caused hyperalgesia was completely lost in Trpv1 -/mice ( Fig. 1a, b), indicating that TRPV1 channel mediates the effect of 1,4-dioxane.
Previous studies have shown that the direct activation of TRPV1 resulted in the paw edema in mice [23][24][25]. We next asked whether 1,4-dioxane can also elicit paw edema in vivo. As illustrated in Fig. 1c, we found that intraplantar administration of 1,4-dioxane (5%, 10 μl) produced a significant increase in paw volume from 4 ± 1 to 32 ± 4% in the hind paws of wild-type mice. This scenario was abolished in Trpv1 -/mice, confirming that the effect of 1,4-dioxane is TRPV1-dependent.
As TRPV1 is highly permeable to Ca 2+ , we examined its activation by 1,4-dioxane with calcium imaging. We observed that 1,4-dioxane evoked robust Ca 2+ increases in acutely isolated TG and DRG neurons of wild-type (WT) mice, but not in neurons prepared from Trpv1 KO mice (Fig. 1d-h). High KCl (60 mM) was subsequently applied to ascertain neuronal viability in the end of each experiment. Patch-clamp recordings further revealed that 1,4-dioxane (2%) potentiated the capsaicin-activated TRPV1 current in DRG and TG sensory neurons and was also able to directly trigger TRPV1 currents in a concentrationdependent manner (Fig. 1i-m). As expected, the effect of 1,4-dioxane was fully suppressed in neurons of Trpv1 KO mice. Together, 1,4-dioxane-caused hyperalgesia is mediated by the enhanced activity of TRPV1 channel.

Gating and modulation of TRPV1 by 1,4-dioxane
To characterize 1,4-dioxane potentiation of TRPV1 channel, we expressed it in HEK293 cells and used whole-cell voltage-clamp recording to analyze the electrophysiological responses. As shown in Fig. 2a-c, 1, 4dioxane dose-dependently activated TRPV1 channel with an EC 50 = 2.1 ± 0.03% concentration. Single-channel a The radiant heat test shows 1,4-dioxane-caused thermal hyperalgesia in WT mice but not in Trpv1 -/mice. b Von Frey filament assay evaluating the effect of 1,4-dioxane on mechanical pain sensing. c Intraplantar injections of 10 μl of 5% 1,4-dioxane significantly increased paw volume compared with that injected with saline controls. The paw edema ratio is the percentage increase of paw volume induced by 1,4-dioxane in Trpv1 +/+ mice. The effects of 1,4-dioxane were abolished in Trpv1 -/mice. The representative images of the dioxane-injected paw edema are shown on the left. d Responses of dorsal root ganglia (DRG) neurons acutely isolated from Trpv1 +/+ or Trpv1 -/mice were consecutively challenged with 1 μM capsaicin (Cap), 5% 1,4-dioxane, and 60 mM KCl, as indicated. Channel activation was assessed by calcium imaging in cells loaded with the fluorescent Ca 2+ indicator Fluo-4, AM. The colored bar indicates relative calcium levels. e Averaged responses of DRG neurons dissociated from Trpv1 +/+ (gray, n = 61) or Trpv1 -/-(red, n = 45) mice to capsaicin, 1,4-dioxane, and KCl. Fluo-4 epifluorescence changes were computed as (Fi-F0)/F0, where Fi represented fluorescence intensity at any frame and F0 was the baseline fluorescence calculated from the averaged fluorescence of the first 10 frames. f Percentage of DRG neurons responding to capsaicin (1 μM), 1,4-dioxane (5%), or high KCl (60 mM) in neurons isolated from Trpv1 +/+ or Trpv1 -/mice. g Trigeminal ganglia (TG) neurons from Trpv1 +/+ or Trpv1 -/mice were challenged with 1,4-dioxane (5%) followed by capsaicin (1 μM), then high KCl (60 mM). Quantification of responses were assessed by calcium imaging (n ≥ 30 cells per trace). h Percentage of TG neurons responding to capsaicin, 1,4-dioxane, or high KCl in neurons isolated from Trpv1 +/+ or Trpv1 -/mice. i Typical response of DRG neurons cultured from wild-type (i1) and Trpv1-deficient (i2) mice to 1,4-dioxane and capsaicin. j Quantification of peak currents for recordings in (i1). k-l Similar recordings and statistical results in TG neurons. m Quantification of responses to 1,4-dioxane (5%) and capsaicin (5 μM) in both DRG and TG neurons cultured from wild-type or Trpv1 -/mice (n ≥ 8). Responses to 1,4-dioxane were observed only in the capsaicin sensitive neurons recordings were acquired from the outside-out membrane patches of TRPV1-expressing HEK293 cells (Fig. 2d), perfused with 1,4-dioxane under various voltage clamps ranging from −100 to +100 mV. Comparing the i-V curve between 1,4-dioxane and capsaicin activation, we observed a similar rectified shape with a relatively linear increase at depolarizing voltages and a sublinear dependence at hyperpolarizing potentials (Fig. 2e).
TRPV1 is a multimodal channel and can be activated by various stimuli such as voltage, heat, protons, capsaicin, and the pharmacological compound 2aminoethoxydiphenyl borate (2-APB) as well as a variety of endogenous factors. Notably, one type of activation can be potentiated by another. We assessed the impact of 1,4-dioxane on TRPV1 activation by other stimulations. As shown in Fig. 2f, g, 1, 4-dioxane indeed potentiated TRPV1 currents evoked by capsaicin, acidification (pH 6.0), and 2-APB. A scenario was further confirmed by single-channel recordings showing that the presence of 1,4-dioxane increased TRPV1 channel open probability (Fig. 2h, i). There was no significant difference for the amplitude of single-channel currents elicited by different conditions (Fig. S3). The higher opening of the channel was associated with stronger noise, which tended to be more profound at hyperpolarized potentials. Notably, the single-channel currents evoked by 1,4dioxane were fully blocked by TRPV1 blocker ruthenium red (RR, 10 μM, −60 mV, Fig. 2h) but not capsazepine (Czp, Fig. 2i), the selective inhibitor for capsaicin stimulation. Notably, RR did not block TRPV1 currents evoked by either 1,4-dioxane or capsaicin at +60 mV a Representative whole-cell recordings from TRPV1-expressing HEK293 cell. The cell was exposed to varied concentrations of 1,4-dioxane or capsaicin (Cap, 1 μM), as indicated. Holding potential was −60 mV. b Summary of relative currents elicited by 1,4-dioxane or capsaicin. Currents were normalized to that evoked by 1 μM capsaicin. Numbers of cells are indicated in parentheses. c Dose-response curve of 1,4-dioxane. Fitting by Hill's equation gave an EC 50 = 2.1 ± 0.03% and n H = 3.1 ± 0.22 (n = 9). d Representative single-channel currents recorded from an outside-out patch of TRPV1-expressing HEK293 cells. Currents were evoked by 1,4dioxane at different voltages ranging from −100 to +100 mV and were low-pass filtered at 2 kHz. e Unitary current-voltage relationships activated by 1,4-dioxane (red) or capsaicin (black), showing no significant difference. f Potentiation effects of 1,4-dioxane. Currents were evoked by capsaicin, 2-APB, acidification (pH 6.0) or 1,4-dioxane as indicated. The presence of 1,4-dioxane potentiated the responses of each stimulation. Holding potential V h = −60 mV. g Comparison of the averaged response evoked by different agonists with or without 1% 1,4-dioxane. The labels over the bars indicate the number of recordings. h Potentiation by 1,4-dioxane of single-channel activity in outside-out patches excised from TRPV1-expressing HEK293 cells. Capsaicin, 1,4-dioxane, capsazepine, or RR were applied consecutively to the same patches as indicated. The holding potentials were +60 mV (upper) and −60 mV (lower), respectively. Co-application of 1,4-dioxane (0.5%) with 0.02 μM capsaicin resulted in more single-channel openings. i Average changes in TRPV1 open probability. The addition of 1,4-dioxane remarkably increased single-channel open probability. Inclusion of RR but not capsazepine reduced the channel openings (V h = −60 mV, n = 7-11 cells per condition) (Fig. S4). These data indicate that 1,4-dioxane potentiates TRPV1 activity at the single-channel level.

1,4-Dioxane lowers temperature threshold for TRPV1 activation
TRPV1 is a thermal sensor responding to noxious temperature for pain initiation. We tested whether 1,4dioxane affects the temperature threshold of TRPV1. With an infrared laser system, ultrafast temperature sweeps (30 to 53°C) were generated surrounding TRPV1-expressing HEK293 cells (Fig. 3a). As previously reported [26], TRPV1 appeared to be activated at a temperature threshold of~43 o C (Fig. 3b 1 ). The presence of 1,4-dioxane dose-dependently strengthened the thermal sensitivity of TRPV1, with the temperature threshold dropping down to 37°C ( Fig. 3b-d). We then asked if temperature also affects 1,4-dioxane activation of TRPV1. As shown in Fig. 3e, TRPV1 currents were evoked by various concentrations of 1,4-dioxane, during which a temperature jump from 23 to 37°C was introduced. As the temperature increased, the activation effect of 1,4-dioxane became much stronger. The EC 50 of 1,4-dioxane on TRPV1 activation was shifted to 1.04 ± 0.02% from 2.03 ± 0.05% when temperature was increased from 23 to 37°C (Fig. 3f). The interactive facilitation of 1,4-dioxane with TRPV1 thermal activation provides a mechanistic basis for 1,4-dioxane-caused hyperalgesia.
To explore the effect of 1,4-dioxane on pain transduction under inflammation condition, we injected carrageenan (20 μL of 2% (w/v) in normal saline) into the hindpaws of mice to induce a state of local inflammation [19,32]. In half an hour, both the withdrawal latency of the injected paws in response to radiant heat and the threshold for mechanical pain sensing decreased dramatically to about 50% of the baseline level in Trpv1 +/+ mice (Fig. 4g). However, carrageenan-induced thermosensation was completely lost and carrageenan-induced mechanosensation was partially lost in Trpv1 -/mice (Fig. 4g). Thereafter, varying concentrations of 1,4dioxane were injected into the same hind paws post carrageenan injection. As compared to non-inflammatory conditions ( Fig. 1a-b), lower concentration of 1,4dioxane could significantly reduce the paw withdrawal latency to heat and the mechanical pain sensitivity in Fig. 4 Activation of TRPV1 under inflammation-related conditions. a Effects of acid on the 1,4-dioxane-induced response in TRPV1-expressing HEK293 cells. Representative current traces evoked by varied concentrations of 1,4-dioxane in combination with the neutral pH (7.4) and then switched to pH 6.0. Acid strongly potentiated 1,4-dioxane responses at 0.3%, 0.5%, and 1% concentration. b The average plot compares the 1,4dioxane response in pH 7.4 and 6.0 conditions. Currents were normalized to the responses evoked by 1 μM capsaicin. c A representative wholecell recording of the TRPV1 E600Q mutant. The cell was exposed to different concentrations of 1,4-dioxane in neutral condition (pH 7.4). d Average plot of peak currents. The constitutively acidified TRPV1 E600Q mutant potentiated 1,4-dioxane-evoked current responses. e Potentiation of 1,4-dioxane (2%) responses by PDBu treatment in primary cultured DRG neurons. The phosphorylation of the channel induced by PDBu (2 μM) was verified with two capsaicin concentrations, 0.2 and 1 μM, applied before and after treatment. After phosphorylation, 1,4-dioxane (2%) or capsaicin (0.2 μM) elicited larger responses. The pipette solution also contained 2 mM MgATP. f Relative changes of the 1,4-dioxane-and capsaicin-evoked responses before and after phosphorylation. All recordings were made at a membrane potential of −60 mV. g The PWL of hind paws to radiant heat were measured for Trpv1 +/+ and Trpv1 -/mice under carrageenan-induced inflammation conditions (left). Von Frey filament assay evaluating the effect of 1,4-dioxane on mechanical pain sensing under carrageenan-induced inflammation conditions (right) mice under an inflammatory state (Fig. 4g). Again, the 1,4-dioxane-caused hyperalgesia under inflammation condition was completely lost in Trpv1 -/mice.

Discussion
As a trace contaminant, 1,4-dioxane is used as an organic solvent in the industry and could be found in some daily care products and food additives. Here, we show that 1,4-dioxane causes hyperalgesia by potentiating the nociceptive receptor TRPV1. Our data reveal that TRPV1 activation is acutely facilitated by 1,4-dioxane. Moreover, our results suggest that TRPV1 potentiation may occur in conditions exposing to high levels of 1,4dioxane. As TRPV1 channel can be gradually sensitized, recurrent exposure to 1,4-dioxane may also disturb TRPV1-mediated sensory transduction. We also observed that 1,4-dioxane analogs 1,3-dioxane and tetrahydrofuran are also able to activate TRPV1. As 1,4dioxane is widely used solvents in industry [33], for instance, up to 8% of 1,4-dioxane is used in the chlorinated solvent 1,1,1-trichloroethane [TCA] and even higher concentration used in certain cutting oils [35], our data suggest appropriate measures need be taken to avoid their potential interference with sensory transduction.
1,4-Dioxane caused hyperalgesia in vivo within minutes, compatible with its skin penetration property and consistent with its irritating effect on humans [5]. Such scenario differs from the chronic effect of 1,4-dioxane that underlies organ damage and tumorigenesis [7]. We attributed 1,4-dioxane-caused hyperalgeisa to its action on TRPV1 receptor as evidenced by CRISPR/Cas9 KO genetics, which is consistent with the wide distribution of TRPV1 in nociceptive sensory neurons and its role in nociception [36,37]. While our current data demonstrate TRPV1 as a molecular target for 1,4-dioxane, it could also dysregualte other sensory pathways.
TRPV1 functions as a multimodal signal transducer of noxious stimuli in the mammalian somatosensory system [26]. Although chemical and thermal stimuli interact allosterically with TRPV1 through independent molecular mechanism, they can cross-sensitize each other to enhance channel activation. By characterizing the action mode of 1,4-dioxane at single-channel and whole-cell level, we unveiled its direct activation of TRPV1 and its potentiation effect on the activity of TRPV1 evoked by other stimulations. We also demonstrated that 1,4-dioxane significantly lowered the thermal threshold of TRPV1, which could explain the hyperalgesia behavior observed in vivo. The facilitatory effect of inflammatory factors on the activation of TRPV1 by 1,4-dioxane likely represents a positive feedback to amplify the pain sensation. These observations substantiate the theory of the synergistic sensitization among different stimuli on TRPV1 activation.
Functional and structural data have suggested the cytosolic S4-S5 linker as a gearbox in TRP channel gating [38]. Several residues within the S4-S5 linker of TRPV1 are critical to its function. For instance, numerous charged residues are important to the allosteric coupling between the voltage-, temperature-, 2-APB-, and capsaicin-dependent activation [39]. Structural analysis of TRPV1 protein has also identified the S4-S5 linker region as an important binding pocket for vanilloid and PI(4,5)P 2 [40]. Furthermore, D. H. Kwan and colleagues found that M572 in the S4-S5 linker plays a critical role in the S6 gate opening and the mutation M572A nearly abolishes heat-dependent opening but with little effect on capsaicin response [41]. In the present study, we showed the substitution of M572 by V eliminated the 1,4-dioxane-evoked currents while retaining the capsaicin responses, thus further highlighting the importance of the S4-S5 linker in the channel gating. In addition, substitutions of M572 by A, L, and I with varied hydrophobic side-chain size in the S4-S5 linker ablated the activation of TRPV1 in varying degrees by 1,4dioxane and its analogs without affecting the capsaicin response. However, the substitutions of M by F, W, R, E, or D resulted in nonfunctional channels. These results may arise from two possibilities. One is that methionine substituted into valine compromises the potential hydrophobic interaction between the sulfur atom and the aromatic ring-like compound like dioxanes [42], thus reducing the binding efficiency of 1,4-dioxane. It is remarkable that, despite the essential role of M572 in activation of TRPV1 by 1,4-dioxane, the actual binding sites of 1,4-dioxane on TRPV1 still requires further investigation, such as co-crystallization of this compound and channel protein. Another possibility is that M572 resides at a strategic position for channel gating, and its mutation might destroy the energetic transduction between 1,4-dioxane activation and gate opening. The present study identifies TRPV1 channel as a molecular target of 1,4-dioxane, which would help to understand its action on biological processes.

Conclusions
We show that 1,4-dioxane targets in vivo the nociceptive capsaicin receptor TRPV1, whereby causing hyperalgesia in mouse model. With a mechanistic focus, we unveil that besides direct activation of TRPV1, 1,4-dioxane potentiates its agonistic and thermal activation. The effect of 1,4-dioxane is also upregulated by inflammatory factors, suggesting a positive feedback contributing to pain hypersensitivity. By site-directed mutagenesis, we further provide molecular insights into 1,4-dioxane-TRPV1 interaction. This study hence deepens our understanding on the pathological effect of 1,4-dioxane and its potential health concern.

Mice
The Trpv1 -/mice were generated by GemPharmatech Co., Ltd. through CRISPR/Cas9-mediated genome editing. In brief, the vectors encoding Cas9 and guide RNA (5'-GAGCCTAGGGCTAGGCA-3' and 5'-GCTACAGA GTGCAATTCCGGG-3') were in vitro transcribed into messenger RNA (mRNA) and gRNA followed by injection into the fertilized eggs that were transplanted into pseudopregnant mice. The targeted genome of F0 mice was amplified with PCR and sequenced, and the chimeras were crossed with wild-type C57BL/6 mice to obtain the Trpv1 +/mice. The F1 Trpv1 +/mice were further crossed with wild-type C57BL/6 mice for at least three generations. The age-and sex-matched Trpv1 +/+ and Trpv1 -/littermates were obtained by crossing the Trpv1 +/mice and were randomly assigned to experimental groups throughout this study. The PCR primers for genotyping were as follows: Trpv1 +/+ and Trpv1 -/forward 5'-AGGGAGTCACTCAGGATGGCTTCT-3'; Trpv1 +/+ reverse 5'-AGCCAGGTCAAGATTGGAA AAGG-3'; Trpv1 -/reverse 5'-GCTGCCCTCTGACC TCAGCTACAC-3', with the expected product sizes of 375 bps and 558 bps for Trpv1 -/and wild-type mice, respectively. All mice were housed in the specific pathogen-free animal facility at Wuhan University, and all animal experiments were in accordance with protocols (No. WDSKY0201804) approved by the Institutional Animal Care and Use Committee of Wuhan University.

Hargreaves test
Behavioral studies were performed with 6-to 8-week-old wild-type or Trpv1 -/adult C57 male mice. All tests were conducted during the light phase of the light/dark cycle by a trained observer blind to the genotype. Mice were habituated to the testing room for 60 min prior to all behavioral tests unless otherwise stated. The Hargreaves test was performed as described previously [32]. All behavioral experiments were conducted in a double-blind manner. Mouse was placed in a clear plexiglass cylinder on top of a temperature-controlled Plantar Test Instrument (Ugo Basile), which produces a high-intensity infrared light aimed at the plantar surface of the hind paw. The withdraw latency of thermal hyperalgesia was determined by the onset of paw lift and/or lifting, licking, and biting. Paw-withdrawal latency in response to heating was measured by a fixed infrared stimulus. Maximum stimulus duration was set at 20 s to prevent tissue damage. The animal was habituated in the plexiglass cylinder for 30 min. The right hind paws of mice were injected intraplantarly with 10 μl vehicle (normal saline). The left hind paws of mice were injected intraplantarly with 10 μl vehicle (normal saline supplemented with varying concentrations of 1, 4-dioxane as indicated). For measure of mechanical allodynia, a dynamic plantar aesthesiometer (von Frey apparatus, Ugo Basile, Milan, Italy) were used (stimulus rate of 1 g/s; cutoff value of 10 g). Each mouse was placed individually in clear Plexiglas chambers (8 × 8 × 12 cm) and acclimated for 30 min before testing. Mechanical threshold was measured at 30 min after injection. The mechanical threshold was averaged from up to three stimuli. Numbers of each group ≥ 6.
To develop the carrageenan-induced inflammation model, the mouse hind paw was injected with 20 μl 2% (w/v) carrageenan. In a half hour, the withdraw latency and the withdraw mechanical threshold were measured. Thereafter, 10 μl vehicle (normal saline supplemented with varying concentrations of 1, 4-dioxane as indicated) was injected into the same hind paws post carrageenan injection. The withdraw latency and the withdraw mechanical threshold were determined again 30 min later.

Paw edema test
Edema was induced by intraplantar injection of 10 μl of 1,4-dixoane freshly prepared in vehicle (normal saline supplemented with 5% 1, 4-dioxane) into the left-hind paws of Trpv1 +/+ and Trpv1 -/mice, respectively. Paw volumes were measured just before and 2 h post injection of 1,4-dixoane using a plethysmometer (IITC). The increase in percentage of paw volume was calculated based on the volume difference between the normal and abnormal paws (with or without injection of 1, 4dioxane). The following equation was used: paw edema ratio (%) = (paw volume after injection of 1, 4-dioxanepaw volume after injection of saline)/paw volume after injection of saline × 100%.

Preparation of dorsal root ganglia (DRG) neurons and trigeminal ganglia (TG)
Sensory neurons were dissociated from dorsal root ganglia (DRG) or trigeminal ganglia (TG) of mice. Primary culture of DRG neurons and TG neurons were established following enzymatically and mechanically dissociation of the ganglia as described before with minor modification [43,44]. Briefly, 6-to 8-week-old wild-type or Trpv1 -/-C57 BL/6 mice were used. Mice were deeply anesthetized and then decapitated. DRGs together with the dorsal-ventral roots and attached spinal nerves were taken out from the spinal column. After removing the attached nerves and surrounding connective tissues, about 10-14 DRGs from the thoracic and lumbar segments of spinal cords were rapidly dissected and cleaned in Ca 2+ /Mg 2+ -free Hank's balanced salt solution (HBSS). Ganglia were dissociated by enzymatic treatment with collagenase type IA (1 mg/ml), trypsin (0.4 mg/ml), and DNase I (0.1 mg/ml) and incubated at 37 o C for 40 min. After dissection of the paired TG, they were washed in phosphate-buffered saline (PBS), minced and gathered in minimal essential medium (MEM, Invitrogen) containing collagenase type IA (1 mg/ml), and incubated at 37 o C for 45 min. During digestion, gentle mechanical trituration was performed every 10 min through firepolished glass pipettes until solution became cloudy. The resulting suspension of single cells was collected by centrifuge. After three washes in DMEM/F12 medium containing 10% FBS, 50 units/ml penicillin, and 50 μg/ml streptomycin, cells were dispersed by gentle titration, plated on glass coverslips coated with poly-L-lysine, and cultured at 37 o C in a humidified incubator with 5% CO 2 . Patch-clamp recordings were carried out 12-24 h after plating. TRPV1(WT), TRPV1(M572V), or empty vector together with the expression marker GFP was reintroduced into Trpv1 -/-DRGs by electroporation using Neon transfection system (Life Technologies) as described [45].

Ca 2+ imaging
Cells on coverslips were loaded with 5 μM Fluo-4 AM (Beyotime Bio-Tech Co., Ltd.) for 40 min at 37 o C in the original culture room. The cells were then washed three times with an incubation buffer containing (in mM) 140 NaCl, 5 KCl, 2 MgCl 2 , 10 HEPES, 10 glucose, and 2 CaCl 2 (pH 7.4). Cells were incubated in incubation buffer for 30 min at 37 o C to allow deesterification of intracellular AM esters. Calcium imaging was performed on an inverted epifluorescence microscope (Olympus IX 73) equipped with a complete illumination system (Lambda XL, Sutter Instruments). Fluorescent images were acquired using a cool CCD camera (CoolSNAP ES2, Teledyne Photometrics) controlled by Micro-Manager 1.4 (Vale lab, UCSF) using a public 1394 digital camera driver (Carnegie Mellon University). Images at excitation of 448-492 nm and emission of 512-630 nm were taken in a continuous time-based mode. After baseline images were taken, 1,4-dioxane was added. High KCl (60 mM) was applied in the end of each experiment to ascertain neuronal viability. More than 90% neurons were responsive to 60 mM KCl stimulation and were included in the analysis.

Electrophysiological recordings
Conventional whole-cell and excised outside-out patchclamp recording methods were used. For the recombinant expressing system, EGFP fluorescence was used as a marker for gene expression. Patch-clamp recordings were voltage clamped using an EPC10 amplifier (HEKA, Lambrecht, Germany). Voltage commands were made from the Patchmaster program. For a subset of recordings, currents were amplified using an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA) and recorded through a BNC-2090/MIO acquisition system (National Instruments, Austin, TX) using QStudio developed by Dr. Feng Qin at State University of New York at Buffalo. Recording pipettes were pulled from borosilicate glass capillaries (World Precision Instruments), and firepolished to a resistance between 2 and 4 MΩ for wholecell recordings, and 6-10 MΩ for single-channel recordings when filled with internal solution. Whole-cell recordings were typically sampled at 5 kHz and filtered at 1 kHz, and single-channel recordings were sampled at 25 kHz and filtered at 10 kHz.
Bath saline for whole-cell recording in HEK293 cells consisted of (in mM): 140 NaCl, 5 KCl, 5 EGTA, and 10 HEPES, pH 7.4 adjusted with NaOH. MES ethanesulfonic acid] was used as the pH buffer when pH<6.5 and HEPES was used for pH 7.0-7.4. Electrodes were filled with (in mM): 140 CsCl, 5 EGTA, 10 HEPES, and pH 7.4 adjusted with CsOH. The bath and pipette solutions for singlechannel recording were symmetrical and contained 140 NaCl, 5 KCl, 5 EGTA, and 10 HEPES, pH 7.4 (adjusted with NaOH). For recording DRG and TG neurons, the bath solution contained (in mM): 140 NaCl, 5 KCl, 2 MgCl 2 , 2 CaCl 2 , 10 glucose, 10 HEPE S, and pH 7.4 adjusted with NaOH and the pipette solution contained (in mM): 140 CsCl, 5 NaCl, 5 EGTA, and 10 HEPES, pH 7.3 adjusted with CsOH. Exchange of external solutions was performed using a gravity-driven local perfusion system. As determined by the conductance tests, the solution around a patch under study was fully controlled by the application of a solution with a flow rate of 100 μl/min or greater. All pharmacological experiments met this criterion. For recordings under low pH conditions, the solution also contained 50 μM amiloride to inhibit native acid sensing ion channels. Capsaicin and capsazepine were dissolved in pure ethanol to make a stock solution. All the stocks were diluted in the bath solution to the desired concentrations right before the experiment. The final concentrations of ethanol did not exceed 0.3%, which had no effect on the currents. Unless otherwise noted, all chemicals were purchased from Sigma (Millipore Sigma, St. Louis, MO). The suppliers, catalog numbers and storage methods of some major chemical reagents are shown in Table 1. All patch-clamp recordings were performed at room temperature (RT, 22-24 o C) except for temperature stimulation (see below).

Ultrafast temperature jump achievement
A single-emitter infrared laser diode (1470 nm) was designed to produce temperature jumps, as previously described [46]. A multimode fiber with a core diameter of 100 μm was used to transmit the launched laser beam. The other end of the fiber exposed the fiber core was placed close to the cell of interest where the perfusion pipette is usually located. The laser diode was driven by a pulsed quasi-CW current power supply (Stone Laser, Beijing, China), and pulsing of the controller was controlled from a computer through BNC-2090/MIO data acquisition card, which was also responsible for patchclamp recordings. A blue laser line (460 nm) was coupled into the same fiber to aid alignment. Constant temperature steps were generated by irradiating the tip of an open pipette filled with the bath solution and the current through the electrode was used as a readout for feedback control. The laser diode was first powered on for a brief duration to reach the set temperature and was then modulated to achieve a constant pipette current. The profile of the modulation pulses was stored and subsequently played back to apply the temperature jump to the cell of interest. Temperature was calibrated offline from the electrode current based on the temperature dependence of electrolyte conductivity. The threshold temperature for heat activation of TRPV1 was determined as the temperature at which the slow inward current was elicited.

Statistics and reproducibility
Data were processed with Qstudio developed by Dr. Feng Qin at State University of New York at Buffalo, Clampfit (Molecular Devices, Sunnyvale, CA), IGOR (Wavemetrics, Lake Oswego, OR, USA), SigmaPlot (SPSS Science, Chicago, IL, USA), and OriginPro (Origi-nLab Corporation, MA, USA). For concentration response analysis, the modified Hill equation was used: Y = A1 + (A2-A1)/[1 + 10^(logEC 50 -X) * n H ], in which EC 50 is the half maximal effective concentration, and n H is the Hill coefficient. Experiments were performed in two or more separate batches of DRG, TG, and HEK293 cells to confirm reproducibility. Unless stated otherwise, the summary data are presented as mean ± standard error (s.e.m.), from a population of cells (n) with statistical significance assessed by Student's t test for one or two group comparison and one-way analysis of variance (ANOVA) tests for multiple group comparisons. Significant difference is indicated by a p value less than 0.05 (*p < 0.05, **p < 0.01, *** p < 0.001).