Journal of Neuroscience Research 93:309–320 (2015)

Phenylarsine Oxide as a Redox Modulator of Transient Receptor Potential Vanilloid Type 1 Channel Function Kevin P. Carlin,* Gang Wu, Aniket Patel, Gregg Crumley, and Victor I. Ilyin Discovery Research, Purdue Pharma LP, Cranbury, New Jersey

Transient receptor potential vanilloid type 1 (TRPV1) channels are capable of detecting and integrating noxious stimuli and play an important role in nociceptor activation and sensitization. It has been demonstrated that oxidizing agents are capable of positively modulating (sensitizing) the TRPV1 channel. The present study investigates the ability of the thiol-oxidizing agent phenylarsine oxide (PAO) to modulate TRPV1 currents under voltageclamp conditions. We assessed the ability of PAO to modulate both proton- and capsaicin-activated currents mediated by recombinant human TRPV1 channels as well as native rat and human TRPV1 channels in dorsal root ganglion (DRG) neurons. Experiments with other oxidizing and reducing agents having various membranepermeating properties supported the intracellular oxidizing mechanism of PAO modulation. The PAO modulation of proton-activated currents was consistent across the cell types studied, with an increase in current across the proton concentrations studied. PAO modulation of the capsaicin-activated current in hTRPV1/Chinese hamster ovary cells consisted of potentiation of the current elicited with low capsaicin concentrations and inhibition of the current at higher concentrations. This same effect was seen with these recombinant cells in calcium imaging experiments and with native TRPV1 channels in rat DRG neurons. Contrary to this, currents in human DRG neurons were potentiated at all capsaicin concentrations tested after PAO treatment. These results could indicate important differences in the reduction–oxidation modulation of human TRPV1 channels in a native cellular environment. VC 2014 Wiley Periodicals, Inc. Key words: sensitization; capsaicin; human DRG; rat DRG; pain

The transient receptor potential vanilloid type 1 (TRPV1) channel is a nonselective cation channel expressed in nociceptive neurons. The TRPV1 channel can be activated by capsaicin, protons, or heat and thus serves as a molecular integrator of noxious stimuli in the terminals of polymodal nociceptors (Caterina et al., 1997). The gating mechanisms of these channels are highly complex and are not fully understood. Reduction–oxidation (redox) modulation of ion channel activity has been demonstrated for many channel C 2014 Wiley Periodicals, Inc. V

classes. In nociceptors, redox modulation has been established for T-type calcium channels (Todorovic et al., 2004), acid-sensing ion channels (ASICs; Andrey et al., 2005), and transient receptor potential ankyrin type 1 channels (Macpherson et al., 2007). Like these latter channels, TRPV1 channels are sensitive to the redox environment and are positively modulated by covalent modification of intracellular cysteine residues (Wang and Chuang, 2011). The sensitized channels in nociceptors are thought to contribute to the ongoing pain associated with tissue injury. Phenylarsine oxide (PAO) is a membranepermeable cysteine-reactive agent that covalently binds the vicinal thiol groups of proteins. It has been used to mimic the effects of oxidative stress in tissues and has been used to sensitize TRPV1 channels (Wu et al., 2008; Chuang and Lin, 2009; Wang and Chuang, 2011; Li et al., 2011; Wang et al., 2011). The present study examines the effects of PAO modulation of recombinant TRPV1 channels and of native TRPV1 channels in both rat and human dorsal root ganglion (DRG) neurons. Furthermore, we assess the effects of PAO modulation when channels are gated by both protons and capsaicin. The present data support the ability of PAO to modulate the TRPV1 channel by acting as an oxidizing agent on the cytoplasmic side of the channel. We demonstrate the potentiating effect of PAO on proton activation of both human and rat TRPV1 channels and provide evidence of a potentiating effect of PAO on capsaicin-gated currents in human DRG neurons that was not seen either with the human recombinant channels or with rat DRG neurons.

K.P. Carlin and G. Wu contributed equally to this work. *Correspondence to: Kevin P. Carlin, PhD, Discovery Research, Purdue Pharma LP, 6 Cedarbrook Drive, Cranbury, NJ 08512. E-mail: [email protected] Received 6 June 2014; Revised 24 July 2014; Accepted 26 July 2014 Published online 24 September 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jnr.23479

310

Carlin et al.

MATERIALS AND METHODS Recombinant TRPV1 Cells Chinese hamster ovary (CHO) cells stably overexpressing the hTRPV1 channel (GenBank accession No. AJ277028; Refseq, NM_080704.3) were cultured by using standard techniques. Cells were plated on uncoated 35-mm plastic Petri dishes and maintained in Dulbecco’s modified Eagle’s media (DMEM) with 2 mM L-glutamine and 10% fetal bovine serum (FBS) in a 5% CO2 environment at 37 C for 24–48 hr before electrophysiological experiments. DRG Neonatal DRG were harvested from 2–4-day-old Sprague-Dawley rats. Ganglia were digested with trypsin (2.5%; Invitrogen, Carlsbad, CA) in Ham’s F-12 media; then, the digested tissue was rinsed three times with F-12 media containing 10% FBS. The tissue was triturated with a fire-polished pipette and plated on poly-D-lysine (PDL)/laminin-coated coverslips (BD Biosciences, San Jose, CA) in neurobasal media supplemented with B27 (Invitrogen) and 50 ng/ml 2.5S mouse nerve growth factor (NGF; Sigma, St. Louis, MO). The plated neurons were incubated at 37 C in 5% CO2 for up to 3 days. Adult DRG were harvested from >3-month-old SpragueDawley rats. Thoracic and lumbar ganglia were initially digested with 20 lg/ml papain in a calcium- and magnesium-free Hank’s balanced salt solution (HBSS; catalog No. 14185; Gibco, Gaithersville, MD) for 30 min before being transferred to a 0.27% collagenase (type 1; catalog No. 4176; Worthington Biochemicals, Lakewood, NJ) and 0.5% dispase II (catalog No. 04942078001; Roche/Boehringer-Mannheim, Basel, Switzerland) solution prepared with calcium- and magnesium-free HBSS. The tissue was rinsed twice with DMEM/F12 medium (Cellgrow; catalog No. 10-092; Corning, Corning, NY) containing 10% FBS and 1% penicillin/streptomycin. The tissue was triturated with a fire-polished pipette and plated on PDL/laminin-coated coverslips in DMEM/F12 solution supplemented with 10% FBS and 1% penicillin/streptomycin with or without 50 ng/ml mouse NGF (Sigma). The plated neurons were incubated at 37 C in 5% CO2 for up to 6 days. Adult human thoracic DRG were procured by the National Disease Research Interchange (NDRI, Philadelphia, PA) or the International Institute for the Advancement of Medicine (IIAM, Edison, NJ). Human DRG were dissected from surrounding tissue and digested in a 0.2% collagenase, 0.25% dispase II, 0.03% DNAse (catalog No. D5025; Sigma) solution prepared with 50%/50% DMEM/F12 medium. The tissue was rinsed twice with DMEM/F12 medium containing 10% FBS and 1% penicillin/streptomycin. The tissue was triturated with a firepolished pipette and centrifuged (1,000 rpm for 10 min) through a 15% Percoll (Sigma) solution. Cells were resuspended and plated on PDL/laminin-coated coverslips in DMEM/F12 solution supplemented with 10% FBS and 1% penicillin/streptomycin with or without 50 ng/ml human NGF and glial cell line-derived neurotrophic factor (GDNF; Sigma). The plated neurons were incubated at 37 C in 5% CO2 for up to 6 days. Given that both NGF and GDNF have been shown to affect TRPV1 channel expression (Amaya et al., 2004; Anand

et al., 2006), we prepared our rat and human cultures with and without growth factors. The presence or absence of these agents did not have an impact on the PAO-induced modulation of either proton or capsaicin responses. Because data from cells cultured in the absence or presence of growth factors were comparable, data for individual rat and human cell preparations were pooled. Electrophysiology Whole-cell voltage-clamp recordings were made by using conventional patch-clamp techniques at room temperature (22– 24 C). The basic equipment configuration has been described elsewhere (Ilyin et al., 2005). Currents were low-pass filtered at 2–5 kHz and digitized at 1 kHz. Borosilicate patch pipettes had resistance values between 1.0 and 3 MX when filled with intracellular recording solution. A standard holding potential of 280 mV was used, unless otherwise noted. Solution exchanges were accomplished with a horizontal translator that moved an array of glass pipettes that were connected to individual reservoirs containing control or test solutions. The aperture of the active pipette was placed 100 mm from the cell under voltage clamp. Calcium Imaging Fluorescence imaging plate reader (FLIPR; Molecular Devices, Sunnyvale, CA) measurements of Ca21 influx were performed according to modified protocols described elsewhere (Valenzano et al., 2003). Briefly, cells were seeded in 96-well plates (20,000 cells/well) in growth media and incubated at 37 C in a CO2 incubator for 24 hr. Individual wells were washed with 0.2 ml Hank’s solution (Life Technologies, Grand Island, NY) containing 1.6 mM CaCl2 and 20 mM HEPES, pH 7.4, (wash buffer), before loading cells with Fluo-4 (3 lM 3 1 hr) in the presence of 2.5 mM Probenicid. Cells were washed once in wash buffer containing 10 mM PAO or dimethylsulfoxide (DMSO) at room temperature for 10 min and then transferred to the FLIPR. In the capsaicin assay, cells were activated with 2.4 pM to 10 mM capsaicin and read for 3 min. In the pH assay, wash buffer containing 10 mM citrate was added to the wells before being transferred to the FLIPR. Then, 0.1 ml of a 0.0125 N HCl solution was added inside the FLIPR to activate the TRPV1 channels. The addition of HCl buffered the citrate-containing solution to a final pH of 5.5. Calcium responses to this stimulation were measured for 1.5 min. To isolate fluorescent changes resulting from TRPV1 activation (i.e., to remove possible ASIC channel effects), the background responses in the presence of a saturating concentration of the TRPV1 antagonist N-(4-tertiarybutylphenyl)-4(3-chloropyridin-2-yl)tetrahydropryazine-1(2H)-carbox-amide (BCTC) were subtracted from the measured signals (Valenzano et al., 2003). The final DMSO concentration in all FLIPR assays was 1.0%. Solutions and Chemicals For CHO cell electrophysiological experiments, the external solution consisted of (in mM) NaCl 130, KCl 5, BaCl2 2, MgCl2 1, glucose 30, HEPES 25, pH 7.3. In a few hTRPV1/CHO experiments, BaCl2 was replaced with 2 mM CaCl2. The internal solution was (in mM) CsCl 140, MgCl2 4, Journal of Neuroscience Research

PAO Modulation of TRPV1 Function EGTA 10, HEPES 10, pH 7.3. For DRG neuron recordings, external solution contained (in mM) NaCl 145, KCl 5, CaCl2 2, MgCl2 1, glucose 10, HEPES 10, pH 7.3. The internal solution was (in mM) KCl 140, CaCl2 1, MgCl2 2, EGTA 11, pH 7.4. For proton activation, the external solution pH was adjusted with HCl. For solutions with a pH below 6, HEPES was replaced with MES. The final concentration of DMSO did not exceed 0.3%. BCTC was synthesized at Purdue Pharma LP (Cranbury, NJ). Capsaicin, PAO, 5,50 -dithio-bis-(2-nitrobenzoic acid; DTNB), 2,3-dimercapto-1-propanol (DMP), and 2,3-dimercaptopropane-1-sulfonic acid (DMPS) were purchased from Sigma. In all CHO cell, adult rat, and human DRG experiments, PAO treatment consisted of extracellular superfusion of 10 lM for 10 min. In some postnatal DRG experiments, PAO treatment consisted of 30 lM for 5 min to reduce recording time. Data are expressed as mean 6 SEM, and significance values (P) are given in the text.

RESULTS PAO Potentiates Proton-Gated hTRPV1 Currents The application of acidic solution to TRPV1expressing CHO cells elicited an inward current that was increased in amplitude after extracellular application of PAO (10 lM for 10 min; Fig. 1A). In control experiments, when only buffer was applied during the treatment period, current responses to repetitive applications of acidic solution were reduced by approximately 30% (n 5 6 cells). The PAO-induced potentiation was proton concentration dependent, with the current elicited by the lowest pH solution tolerated by these cells (pH 4) being increased in amplitude by approximately fourfold, without a significant alteration in the EC50 value (pH 4.7 6 0.1 before vs. pH 5.0 6 0.1 after PAO; P 5 0.07, paired t-test; Fig. 1B). For proton-activated currents, PAO significantly potentiated the channel activity across all proton concentrations tested (P 5 0.036, repeated-measures ANOVA). The potentiation of proton-activated currents by PAO was observed across a large voltage range (Fig. 1C) while maintaining the normal current–voltage relationship, displaying outward rectification and a region of negative slope conductance. A small, but significant, shift in reversal potential toward more negative potentials was seen after PAO treatment (21.1 6 0.9 mV before vs. 24.8 6 0.8 mV after PAO; n 5 6; P 5 0.007, paired t-test). This PAO-potentiated current was sensitive to the TRPV1-selective antagonist BCTC when assessed with whole-cell electrophysiology (IC50 2 nM; n 5 4; Fig. 1D). The ability of PAO to potentiate proton-activated currents was also tested in calcium imaging experiments that sampled a much larger number of cells and did not alter the intracellular environment to the same degree as occurs with whole-cell patch clamp studies. In these experiments, a similar potentiation of the florescent signal was seen after PAO treatment when the cells were stimulated with acidic solution. This potentiated signal could also be blocked with BCTC (IC50 8 nM; n 5 4 experiments; Fig. 1E). These data and the fact that PAO treatment was unable to Journal of Neuroscience Research

311

potentiate currents in CHO cells that did not have functional expression of hTRPV1 (71.4 6 32.1 pA before PAO application vs. 16.3 6 27.2 pA after PAO treatment; n 5 3) indicate that heterologously expressed hTRPV1 channels can be positively modulated by PAO. Redox Modulation Mediates PAO-Induced Potentiation of Proton-Activated Currents Because PAO is a known thiol-oxidizing agent, we sought to confirm that the potentiation that we observed was mediated through redox modulation of the channel activity. Evidence in support of this hypothesis was gained through assessing the ability of another known thioloxidizing agent to mimic the potentiating effects on proton-induced TRPV1 currents. Furthermore, we assessed the ability of thiol-reducing agents to inhibit the potentiating effects of PAO. As illustrated in Figure 2A, when 200 mM of the membrane-impermeable thiol-oxidizing agent DTNB was included in the patch pipette, proton-activated (pH 4) currents were strongly potentiated. In these experiments, the proton-activated responses were seen to increase with time, consistent with time-dependent dialysis of DTNB into the cytosol during whole-cell recording. Such potentiation occurred in all cells tested (n 5 6; 1.55 6 0.22 nA vs. 4.87 6 0.92 nA, initial responses vs. later responses at 10 min; P 5 0.006, paired t-test). Similarly to the proton-activated responses potentiated by PAO, the DTNB-potentiated currents were sensitive to the TRPV1 antagonist BCTC (100 nM; Fig. 2Aa). In contrast to the effects of intracellular dialysis of DTNB, extracellular perfusion of 200 mM DTNB for 10 min was unable to potentiate proton-activated currents (n 5 6; Fig. 2Ab). We next assessed the ability of the membraneimpermeable thiol-reducing agent DMPS to inhibit the potentiating effects of PAO treatment. When applied extracellularly, DMPS was unable to reverse established PAO-induced potentiation (n 5 6; Fig. 2B). However, when cells were whole-cell patch clamped with a pipette solution containing 300 lM DMPS (in 0.1% DMSO), potentiation of proton-evoked currents did not occur after PAO treatment (6/6 cells; Fig. 2Ba), suggesting that DMPS was able to prevent the potentiation from occurring. Proton-activated currents recorded with vehicle (0.1% DMSO) included in the pipette solution demonstrated PAO-induced potentiation (n 5 6; Fig. 2Ba). Finally, the membrane-permeable thiol-reducing agent DMP was able to inhibit the PAO-induced current potentiation when applied extracellularly (Fig. 2C). Taken together, these results indicate that PAO positively modulates TRPV1 channel function via an oxidation reaction occurring on the intracellular surface of the channel. PAO Potentiation of Proton-Gated Currents in DRG Neurons To establish that the modulation observed with the heterologously expressed channels was a phenomenon

312

Carlin et al.

Fig. 1. PAO strongly potentiates proton-activated hTRPV1 currents in CHO cells. A: Current trace illustrating the potentiating effects of PAO treatment (10 mM for 10 min) on currents activated by a range of proton concentrations. B: Concentration–response curves obtained in whole-cell voltage-clamp experiments from three cells exposed to a range of proton concentrations before and after PAO treatment. C: pH 4-activated current–voltage relationships before and after PAO treatment demonstrate PAO current modulation across the voltage range tested. The current–voltage curves were generated according to a voltage ramp protocol (170 to 2175 mV at 0.14 mV/msec). Pooled data indicated a small (4 mV), but significant, shift in reversal

potential toward more negative potentials after PAO treatment (paired t-test; n 5 6). D: Current traces show responses to pH 5 solution before and after PAO treatment. Potentiated current is sensitive to TRPV1-selective antagonist BCTC. E: Calcium imaging experiments performed on a FLIPR platform demonstrate the increase in proton (pH 5.5)-stimulated fluorescence signal after PAO treatment. This was seen in the individual traces (inset) and in the low concentration range in the concentration–response curves. Both the pre- and the post-PAO signals were sensitive to BCTC. Data are averaged from three 96-well plates. PAO was applied at 10 mM for 10 min.

Journal of Neuroscience Research

PAO Modulation of TRPV1 Function

Fig. 2. Modulation of proton-activated currents by thiol-reactive agents in hTRPV1/CHO cells. A: DTNB (200 mM), a membraneimpermeable thiol-oxidizing agent, is capable of mimicking the effects of PAO by potentiating the proton-activated current when applied internally by including this compound in the patch pipette. DTNBpotentiated proton-activated current is sensitive to BCTC (0.1 lM; Aa). Pooled data demonstrate the effectiveness of internally applied DTNB and the absence of effect when applied extracellularly (Ab; **P < 0.01, paired t-test). Time 1, time of first application of pH 4 solution after establishing the whole-cell configuration; time 2, time of largest potentiated current when DTNB was applied internally or after 5 min when applied externally. B: DMPS (0.3 mM), a

Journal of Neuroscience Research

313

membrane-impermeable thiol-reducing agent, was unable to inhibit the PAO-potentiated proton-activated current when applied extracellularly but was able to prevent the PAO-potentiation when it was included in the pipette during whole-cell recording (Ba). Cells dialyzed with vehicle (0.1% DMSO) demonstrated increased current responses to pH 4 after PAO treatment (six of six cells), whereas such potentiation was absent in the cells dialyzed with 0.3 mM DMPS (six of six cells). C: The external application of 0.3 mM DMP, a membrane-permeable thiol-reducing agent, reversed the effect of PAO treatment. The holding potential was 270 mV in all the experiments. Data are mean 6 SE.

314

Carlin et al.

Fig. 3. Proton-activated currents are potentiated by PAO treatment in neonatal rat sensory neurons. A: Current trace from a capsaicinsensitive (1 lM) DRG neuron demonstrates the current response to pH 5 solution before and after PAO treatment. The current elicited by protons prior to PAO treatment consisted of a transient and a sustained component. The sustained component was enhanced after PAO treatment. B: PAO-potentiated proton-gated currents were also

sensitive to the TRPV1 antagonist BCTC (0.1 mM) in these cells. C. Bar graph summarizes the effect of PAO treatment on current responses to pH 5 solution (n 5 9) in neonatal DRG neurons. For responses to pH 5, currents were measured at the midpoint of the responses. The holding potential was 280 mV in all experiments. Data are mean 6 SE. **P < 0.01, paired t-test.

applicable to native TRPV1 channels, we examined PAO modulation of TRPV1 channels in both rat and human sensory neurons. We first assessed the effects of PAO in neonatal rat DRG neurons. Those cells that responded to 1 mM capsaicin also responded to proton application (pH 5) with a typical current waveform containing a fast transient component, followed by a sustained component (Fig. 3A, insets). The fast transient component has been shown to be mediated by ASIC channels (Ugawa et al., 2002) and was confirmed in the present study by its sensitivity to 100 mM amiloride (data not shown), whereas the sustained component is accepted to be mediated predomi-

nantly by TRPV1 channels. PAO treatment was seen to potentiate this sustained-current component profoundly in response to pH 5 in these neurons (Fig. 3A,B). On average, the sustained current was increased approximately sixfold (Fig. 3C). The PAO-potentiated sustained currents elicited by pH 5 were largely antagonized by BCTC (n 5 2; Fig. 3B), indicating that the potentiated component was mediated by TRPV1 channels. Moreover, in capsaicin-insensitive rat DRG neurons, proton-induced currents were not potentiated by PAO (data not shown). We next assessed the effect of PAO on protonactivated currents in human DRG neurons. Exposing Journal of Neuroscience Research

PAO Modulation of TRPV1 Function

315

Fig. 4. PAO modulation of proton-activated currents in human DRG neurons. A: The sustained portion of the current responses to pH 5 solution was potentiated after PAO treatment in capsaicin-responsive human neurons. Current traces from two cells are illustrated. The first cell demonstrated smaller proton-activated currents and a clear transient current component (a), whereas the second cell displayed much

larger currents, without obvious fast transient currents (b). B: Currents in cells exposed to pH 5 solution showed varying degrees of enhancement of the sustained-current component after PAO treatment (n 5 8). Currents were normalized to the pre-PAO current amplitude (the initial challenge with protons). Post-PAO is time of the largest current, after either 5 or 10 min of PAO treatment.

human neurons to pH 5 buffer elicited either a sustained inward current with an initial fast transient component (Fig. 4Aa) similar to the current waveform seen in the rat cells (Fig. 3A) and as previously demonstrated in human DRG neurons (Baumann et al., 1996, 2004), or, more often, a sustained inward current that lacked the fast transient component (Fig. 4Ab). Regardless of the presence of the transient current component, PAO treatment of human neurons resulted in a large potentiation of the sustained component of the proton-induced current in six of the eight neurons tested (Fig. 4A,B). Because cells were challenged with protons only every 5–10 min, it is not clear whether the TRPV1 channels in the two nonpotentiated cells were more prone to desensitization compared with the other cells. In cells that did show PAO-induced potentiation, currents were increased over a wide range, from 1.6 to greater than 30-fold (Fig. 4C). In general, these data indicate that native rat and human TRPV1 channels are modulated by

PAO similarly to the heterologously expressed human TRPV1 channels when gated by protons.

Journal of Neuroscience Research

PAO Modulation of Capsaicin-Gated hTRPV1/CHO Currents We next examined the effect of PAO treatment on capsaicin-activated currents. We examined this effect across a range of capsaicin concentrations via both electrophysiology and calcium imaging. First, by using whole-cell patch-clamp techniques, individual concentration–response relationships were assessed before and after PAO treatment for capsaicin (Fig. 5A,B). In comparison with proton-activated currents, PAO had relatively minor effects on the capsaicin-gated currents across the concentrations tested. Some potentiation of the current was seen at low concentrations, but inhibition was the major effect seen with high concentrations

316

Carlin et al.

(Fig. 5B). On average, the maximal current response elicited by 10 mM capsaicin decreased by approximately 25% after PAO treatment.

Similar effects were observed in intracellular Ca21 imaging experiments. With a larger population of cells under study, the potentiating effect on the currents elicited by low concentrations of capsaicin was more obvious (Fig. 5C), but the effect of the PAO treatment on the florescent signal elicited by the high capsaicin concentrations was, again, inhibition. Overall, PAO treatment resulted in a decrease in the slope of the capsaicin concentration–response curves generated by both methods. PAO Modulation of Capsaicin-Activated TRPV1 Currents in Sensory Neurons Just as with the proton-activated currents, we also examined the effects of PAO modulation of capsaicinactivated currents in sensory neurons. For these experiments, we again used both rat and human DRG neurons. In our initial experiments with neonatal rat DRG neurons, we saw that the responses to concentrations of capsaicin greater than 1 uM were not significantly potentiated by PAO treatment. These results were consistent with the PAO modulation seen in the hTRPV1/ CHO cells in which the capsaicin-activated currents were potentiated at low capsaicin concentrations and inhibited at high concentrations. These findings in neonatal neurons were in contrast to the effect of PAO treatment in adult human DRG neurons. In these cells, PAO treatment led to a potentiation of the capsaicin-activated current (n 5 14). In four cells in which multiple concentrations of capsaicin were tested before and after PAO treatment, this potentiation was seen to occur over the full concentration range (Fig. 6A), with the maximal response to capsaicin being increased approximately twofold, without a significant shift in the EC50 value (Fig. 6B). When capsaicin was applied as a single concentration to a cell, the PAO-potentiated current showed closer to a threefold potentiation (2.8-fold with 20 uM, n 5 2; 3.1-fold with 10 uM, n 5 2; 3.3-fold with 1 uM, n 5 6). To rule out nonspecific PAO effects, the PAO-potentiated capsaicin-activated current was tested for sensitivity to BCTC (Fig. 6C). The potentiated current activated by 1 lM capsaicin was inhibited by 31% with 0.1 lM BCTC (n 5 2) and by 79% with 1 lM BCTC (n 5 2).

Fig. 5. PAO potentiates capsaicin-activated currents in hTRPV1/ CHO cells. A: Current traces show effect of capsaicin (Cap) responses by PAO. B: Concentration–response curves obtained in whole-cell voltage clamp experiments for a range of capsaicin concentrations in six cells before and after PAO treatment. Curves demonstrate a dual modulatory effect of PAO treatment on the capsaicin response, a slight potentiation at low capsaicin concentrations and an inhibitory effect at high concentrations. Currents normalized to 10 mM capsaicin response. C. FLIPR calcium imaging experiments replicated the decrease in efficacy with high capsaicin concentrations seen in the electrophysiological data and showed an obvious PAO potentiation of the capsaicin-activated responses at low capsaicin concentrations (pooled data from three 96-well experiments). Data are mean 6 SE. Journal of Neuroscience Research

PAO Modulation of TRPV1 Function

317

Fig. 6. PAO potentiates capsaicin-activated currents in human sensory neurons. A: Current trace demonstrating the response to increasing concentrations of capsaicin and the potentiation of this response after PAO treatment. B: Pooled data from four cells tested with four concentrations of capsaicin demonstrate an increase in the efficacy of cap-

saicin, without a significant shift in the potency after PAO treatment (data are mean 6 SE). C: The 1 lM capsaicin-activated PAO potentiated current in human DRG neurons was sensitive to the TPRV1selective antagonist BCTC (1 lM).

To determine whether the discrepancy that we observed between rat and human DRG neurons was due to the fact we were comparing neonatal rat DRG and adult human DRG, we also examined the effect of PAO on capsaicin-activated current in adult rat DRG neurons. This allowed us to rule out a developmental change in the rat neurons being responsible for the differences that we were observing. The capsaicin concentration–response curves of the two adult species suggested that the native rat channels were somewhat less sensitive than the human channels but confirmed that the channels from both species were activated over a similar concentration range (Fig. 7A). In the adult rat DRG cells, PAO treatment resulted in a similar modulation of the capsaicin-activated current, which was also seen in the neonatal DRG and hTRPV1/CHO cells in which the maximal response was inhibited after treatment (Fig. 7B). Even though the capsaicin-activated current was not potentiated in these cells, we did observe the ability of PAO treatment to reverse desensitization, as previously described (Chuang and Lin, 2009). By using repeated capsaicin applications prior to and during the PAO treatment, it was seen that PAO was able to reverse the decrease in current amplitude; however, we never observed the current increasing to an amplitude greater than the first application (Fig. 7C).

heterologously expressed in a CHO cell background, in neonatal and adult rat DRG neurons, and in adult human DRG neurons. We examined this modulation when channels were activated by both protons and capsaicin. We began by confirming the mechanism of PAO modulation on TRPV1 channels previously proposed (Chuang and Lin, 2009). In this set of experiments, we tested the ability of a second oxidizing agent to mimic the effects of PAO and reducing agents to inhibit these effects. We also confirmed that the modulating effects of PAO that we were observing were mediated predominantly through effects on an intracellular site by using agents that were either membrane permeable or impermeable. The intracellular site of action of this thiol-active compound is consistent with previous mutagenesis studies showing that intracellularly located cysteines near the N-terminus (Salazar et al., 2008; Chuang and Lin, 2009) and C-terminus (Chuang and Lin, 2009; Wang and Chuang, 2011) are important for oxidative sensitization of the TRPV1 channel. The loss of PAO modulation in channel mutants with multiple cysteine substitutions suggests that C158, C387, C391, and C767 (human sequence) are the most relevant cysteine residues for the modulatory effects seen in the present work (Chuang and Lin, 2009). It is believed that C158 and C767 (Chuang and Lin, 2009; Wang and Chuang, 2011) are involved in intersubunit dimerization, whereas C387 and C391 form a disulfide bond within a given subunit (Chuang and Lin, 2009). PAO modulation of proton-activated TRPV1 currents was consistent across the cell types tested,

DISCUSSION The present study examined the ability of the thioloxidizing agent PAO to modulate TRPV1 channels Journal of Neuroscience Research

318

Carlin et al.

Fig. 7. Capsaicin does not show increased efficacy after PAO treatment in adult rat sensory neurons. A: Capsaicin concentration– response curves from both adult rat and human DRG neurons show that capsaicin activates currents over a similar concentration range and suggests that adult human neurons are to some extent more sensitive to capsaicin. B: Capsaicin concentration–response curves taken before and after PAO treatment of adult rat DRG show a reduced efficacy in the high concentration range. Pooled data from four cells tested with four concentrations of capsaicin. All four cells showed varying degrees of current reduction at 10 lM capsaicin after PAO treatment. C: PAO was able to reverse desensitization in these cells but was unable to potentiate the current amplitude relative to the initial current. Capsaicin test pulses were applied at approximately 2-min intervals before and during PAO perfusion. Data are mean 6 SE.

recombinant cells and both rat and human sensory neurons. This modulation resulted in an increase in the current amplitude over a range of proton concentrations

and across a large voltage range. Conversely, PAO modulation of capsaicin-activated currents was not consistent across the capsaicin concentration range studied. There was evidence of some potentiation occurring in the lower concentration range in recombinant cells and clear reduction of the amplitude in the high end of the concentration range in both the recombinant cells and the rat sensory neurons. The capsaicin-activated current in human sensory neurons responded to PAO treatment differently from both the TRPV1/CHO cells and the rat neurons, showing an increase in current amplitude across the full concentration range and resulting in an increase in efficacy without a shift in potency. The ability of PAO to potentiate a response to a saturating concentration of capsaicin was seen whether cumulative escalating concentrations of capsaicin or single saturating concentrations (10 – 20 lM) were applied. This suggests that the potentiation is not just a reversal of desensitization, as demonstrated in Figure 5C and by Chuang and Lin (2009). The ability to increase the efficacy of a saturating concentration of capsaicin has been previously demonstrated with the use of oxidizing agents (H2O2, PAO, and DTNB) on both human and rat TRPV1-expressing HEK cells (Chuang and Lin, 2009) and with the coapplication of protons to rat DRG neurons (Kress et al. 1996), rabbit trigeminal ganglion neurons (Martenson et al., 1994), and the recombinant human (Hayes et al. 2000; Wang et al. 2010), but not rat (Wang et al. 2010), channel isoform. These studies demonstrated that TRPV1 channels are capable of operating at a greater capacity than when stimulated with saturating concentrations of capsaicin alone. We saw clear differences in this capacity between the rat and the human neurons, even though we attempted to use experimental conditions as similar as possible (same experimental equipment, capsaicin application protocol, internal and external solutions, drug stock solutions, DMSO concentration, etc.) to assess these two species of neurons. The ability to increase the efficacy of the capsaicin stimulation in the human DRG neurons was different from the results seen with the adult rat sensory neurons and the CHO cells, even though these latter cells also expressed the human channel isoform. The reason why the human TRPV1 channels were modulated differently in the recombinant system, compared with the native human sensory neurons, is unclear. Given that a number of splice variants for the human TRPV1 channel (Schumacher and Eilers, 2014) have been identified in DRG, it is possible that one of these variants accounts for the altered capsaicin response after PAO treatment in the human sensory neurons, given that it is unclear how splice variants are modulated by redox agents. A second explanation for the differences that were observed among the capsaicin-activated current modulation in the human neurons and the other cells is differences in the cellular background. It has previously been shown that endogenous TRPV1 ligands can show Journal of Neuroscience Research

PAO Modulation of TRPV1 Function

differences in potency and efficacy, depending on the cell type used, to express the hTRPV1 channels (Bianchi et al., 2006). Moreover, coexpression of the TRPV1 splice variant TRPV1var in TRPV1-expressing HEK 293 cells was seen to result in a potentiation of channel activity, whereas coexpression of this variant with the wildtype channel in a COS-7 background appeared to reduce channel activity (Tian et al., 2006). A similar cellular background effect might be influencing the PAO modulation in the present study. A third factor that might be involved in the species difference seen with the PAO modulation of capsaicingated TRPV1 function is phosphatidylinositol-4,5bisphosphate (PIP2). Not only is PAO a thio-oxidixing agent but it also acts as a kinase inhibitor. PAO was initially shown to potentiate proton-gated currents in oocytes expressing rat TRPV1 (Prescott and Julius, 2003), with the belief that it was reducing PIP2 levels through inhibition of phosphatidylinositol 4-kinase. Conflicting data exist on PIP2 modulation of TRPV1 function, with some data suggesting a facilitating effect and other data indicating an inhibitory effect (Stein et al., 1989; Lukacs et al., 2007). A naturally higher level of PIP2 inhibition of TRPV1 channels in human DRG neurons, compared with the CHO cells and rat DRG neurons, could potentially explain the difference seen in PAO-modulated capsaicin response between the rat and the human neurons. More recently, protein kinase C (PKC) bII has been shown to bind directly to TRPV1 channels and to positively modulate the basal sensitivity to thermal, pH, and capsaicin stimulation (Li et al., 2014). Given that PAO has been shown to inhibit general PKC activity in human neutrophils (Kutsumi et al., 1995), a lower level of PKCbII expression in the human neurons compared with the rat neurons might be an explanation for the differences seen between neuron species. Inhibition of this kinase would have little effect after PAO treatment in the human neurons, allowing the redox modulation effect to be appreciated, whereas in the rat neurons PAO inhibition of a higher basal level of PKCbII (leading to a reduction in TRPV1 current) would be offset by the positive modulation of PAO. If in fact PAO is having additional effects on these cells through inhibition of kinases, it is not clear why PAO modulation of proton-activated currents appears similar in neurons of both rats and humans. CONCLUSIONS TRPV1 channels are known to play a major role in transducing the presence of an inflammatory environment into pain signals in both animals and humans (Caterina et al., 2000; Ugawa et al., 2002). This environment typically consists of a combination of elevated tissue temperature, acidosis, and oxidative stress that can produce lipid metabolites and endogenous thiol-oxidizing agents (Negre-Salvayre et al., 2008; Patwardhan et al., 2010) ideal for activation and modulation of TRPV1 channels. The present results suggest that recombinant systems and rat DRG Journal of Neuroscience Research

319

neurons are suitable models for redox modulation of proton-activated currents, but these models might not fully replicate the modulation of capsaicin-activated currents occurring in human sensory neurons. The design of pharmaceutical agents that antagonize TRPV1 channels is an attractive strategy for treating human pain conditions. A more complete understanding of the redox modulation of capsaicin-activated currents seen in adult human sensory neurons might allow the design of more efficacious analgesics. ACKNOWLEDGMENTS The authors thank Dr. Laykea Tafesse and Mr. Jiangchao Yao for synthesizing the BCTC used in these experiments and NDRI and IIAM for procuring the human sensory tissues. REFERENCES Amaya F, Shimosato G, Nagano M, Ueda M, Hashimoto S, Tanaka Y, Suzuki H, Tanaka M. 2004. NGF and GDNF differentially regulate TRPV1 expression that contributes to development of inflammatory thermal hyperalgesia. Eur J Neurosci 20:2303–2310. Anand U, Otto WR, Casula MA, Day NC, Davis JB, Bountra C, Birch R, Anand P. 2006. The effect of neurotrophic factors on morphology, TRPV1 expression, and capsaicin responses of cultured human DRG sensory neurons. Neurosci Lett 399:51–56. Andrey F, Tsintsadze T, Volkova T, Lozovaya N, Krishtal O. 2005. Acid sensing ionic channels: modulation by redox reagents. Biochim Biophys Acta 1745:1–6. Baumann TK, Burchiel KJ, Ingram SL, Martenson ME. 1996. Responses of adult human dorsal root ganglion neurons in culture to capsaicin and low pH. Pain 65:31–38. Baumann TK, Chaudhary P, Martenson ME. 2004. Background potassium channel block and TRPV1 activation contribute to proton depolarization of sensory neurons from humans with neuropathic pain. Eur J Neurosci 19:1343–1351. Bianchi BR, Lee CH, Jarvis MF, El Kouhen R, Moreland RB, Faltynek CR, Puttfarcken PS. 2006. Modulation of human TRPV1 receptor activity by extracellular protons and host cell expression system. Eur J Pharmacol 537:20–30. Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, Julius D. 1997. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389:816–824. Caterina MJ, Leffler A, Malmberg AB, Martin WJ, Trafton J, PetersenZeitz KR, Koltzenburg M, Basbaum AI, Julius D. 2000. Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science 288:306–313. Chuang HH, Lin S. 2009. Oxidative challenges sensitize the capsaicin receptor by covalent cysteine modification. Proc Natl Acad Sci U S A 106:20097–20102. Hayes P, Meadows HJ, Gunthorpe MJ, Harries MH, Duckworth DM, Cairns W, Harrison DC, Clarke CE, Ellington K, Prinjha RK, Barton AJ, Medhurst AD, Smith GD, Topp S, Murdock P, Sanger GJ, Terrett J, Jenkins O, Benham CD, Randall AD, Gloger IS, Davis JB. 2000. Cloning and functional expression of a human orthologue of rat vanilloid receptor-1. Pain 88:205–215. Ilyin VI, Hodges DD, Whittemore ER, Carter RB, Cai SX, Woodward RM. 2005. V102862 (Co 102862): a potent, broad-spectrum statedependent blocker of mammalian voltage-gated sodium channels. Br J Pharmacol 144:801–812.

320

Carlin et al.

Kress M, Fetzer S, Reeh PW, Vyklicky L. 1996. Low pH facilitates capsaicin responses in isolated sensory neurons of the rat. Neurosci Lett 211:5–8. Kutsumi H, Kawai K, Johnston RB, Rokutan K. 1995. Evidence for participation of vicinal dithiols in the activation sequence of the respiratory burst of human neutrophils. Blood 85:2559–2569. Li H, Wang S, Chuang AY, Cohen BE, Chuang HH. 2011. Activitydependent targeting of TRPV1 with a pore-permeating capsaicin analog. Proc Natl Acad Sci U S A 108:8497–8502. Li L, Hasan R, Zhang X. 2014. The basal thermal sensitivity of the TRPV1 ion channel is determined by PKCbII. J Neurosci 34:8246– 8258. Lukacs V, Thyagarajan B, Varnai P, Balla A, Balla T, Rohacs T. 2007. Dual regulation of TRPV1 by phosphoinositides. J Neurosci 27:7070– 7080. Macpherson LJ, Dubin AE, Evans MJ, Marr F, Schultz PG, Cravatt BF, Patapoutian A. 2007. Noxious compounds activate TRPA1 ion channels through covalent modification of cysteines. Nature 445:541– 545. Martenson ME, Ingram SL, Baumann TK. 1994. Potentiation of rabbit trigeminal responses to capsaicin in a low pH environment. Brain Res 651:143–147. Negre-Salvayre A, Coatrieux C, Ingueneau C, Salvayre R. 2008. Advanced lipid peroxidation end products in oxidative damage to proteins. Potential role in diseases and therapeutic prospects for the inhibitors. Br J Pharmacol 153:6–20. Patwardhan AM, Akopian AN, Ruparel NB, Diogenes A, Weintraub ST, Uhlson C, Murphy RC, Hargreaves KM. 2010. Heat generates oxidized linoleic acid metabolites that activate TRPV1 and produce pain in rodents. J Clin Invest 120:1617–1626. Prescott ED, Julius D. 2003. A modular PIP2 binding site as a determinant of capsaicin receptor sensitivity. Science 300:1284–1288. Salazar H, Llorente I, Jara-Oseguera A, Garcia-Villegas R, Munari M, Gordon SE, Islas LD, Rosenbaum T. 2008. A single N-terminal cyste-

ine in TRPV1 determines activation by pungent compounds from onion and garlic. Nat Neurosci 11:255–261. Schumacher MA, Eilers H. 2014. TRPV1 splice variants: structure and function. Front Biosci 15:872–882. Stein C, Millan MJ, Shippenberg TS, Peter K, Herz A. 1989. Peripheral opioid receptors mediating antinociception in inflammation. Evidence for involvement of mu, delta and kappa receptors. J Pharmacol Exp Ther 248:1269–1275. Tian W, Fu Y, Wang DH, Cohen DM. 2006. Regulation of TRPV1 by a novel renally expressed rat TRPV1 splice variant. Am J Physiol Renal Physiol 290:F117–F126. Todorovic SM, Meyenburg A, Jevtovic-Todorovic V. 2004. Redox modulation of peripheral T-type Ca21 channels in vivo: alteration of nerve injury-induced thermal hyperalgesia. Pain 109:328–339. Ugawa S, Ueda T, Ishida Y, Nishigaki M, Shibata Y, Shimada S. 2002. Amiloride-blockable acid-sensing ion channels are leading acid sensors expressed in human nociceptors. J Clin Invest 110:1185–1190. Valenzano KJ, Grant ER, Wu G, Hachicha M, Schmid L, Tafesse L, Sun Q, Rotshteyn Y, Francis J, Limberis J, Malik S, Whittemore ER, Hodges D. 2003. N-(4-tertiarybutylphenyl)24-(3-chloropyridin-2yl)tetrahydropyrazine 21(2H)-carbox-amide (BCTC), a novel, orally effective vanilloid receptor 1 antagonist with analgesic properties: I. In vitro characterization and pharmacokinetic properties. J Pharmacol Exp Ther 306:377–386. Wang EE, Li H, Wang S, Chuang AY, Chuang HH. 2011. Induction of TRPV1 desensitization by a biased receptor agonist. Channels 5:464–467. Wang S, Poon K, Oswald RE, Chuang HH. 2010. Distinct modulations of human capsaicin receptor by protons and magnesium through different domains. J Biol Chem 285:11547–11556. Wang S, Chuang HH. 2011. C-terminal dimerization activates the nociceptive transduction channel transient receptor potential vanilloid 1. J Biol Chem 286:40601–40607. Wu G, Carlin KP, Patel A, Crumley G, Ilyin VI. 2008. Redox modulation of proton-induced activation of the TRPV1 receptor. IASP Abstracts, 12th World Congress on Pain, Glasgow, Scotland. p 1435.

Journal of Neuroscience Research