Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The VR1 receptor is a ligand-gated, nonspecific cation channel activated by vanilloid compounds, low pH, and noxious heat (Tominaga et al., 1998). The best-known natural vanilloids are capsaicin, which is the pungent ingredient in hot chili peppers, and resiniferatoxin, synthesized by a cactus-like plant called Euphorbia resinifera (Szallasi and Blumberg, 1999). High temperatures, >42°C, protons, and high concentrations of anandamide gate the VR1 receptor directly (Tominaga et al., 1998). Because the VR1 receptor is permeable to numerous cations, including Ca²⁺, Na⁺, and K⁺, its activation leads to an influx of Na⁺ and Ca²⁺, and an efflux of K⁺ (Szallasi et al., 1999).

A number of ionic mechanisms have been implicated in cell death, including Ca²⁺ (Choi, 1987; Munir et al., 1995; Grant et al., 1997), Na⁺ (Raymond et al., 1996; Itoh et al., 1998), and K⁺ (Yu et al., 1999). In the dorsal root ganglion, capsaicin treatment results in selective degeneration of approximately 50% of the sensory neurons of neonatal rats, and approximately 18% of these neurons in adult rats (Holzer, 1991; Jancso, 1992). When capsaicin is administered to adult rats peripherally, it results in marked peripheral nerve fiber degeneration in organs such as skin, ureter, and duodenum (Hoyes and Barber, 1981; Chung et al., 1985, 1990). Nociceptor terminals are destroyed reliably, but complete death of the cell bodies does not always occur. In both adult and newborn rats, the capsaicin-induced neurodegeneration is restricted to the subpopulation of small C-type primary sensory neurons (Jancso et al., 1977, 1985), which are the predominant VR1-expressing neurons (Caterina et al., 1997). Evidence suggests that capsaicin-induced damage of C-fiber neurons may lead to transganglionic degeneration (Mannion et al., 1996). Toxic capsaicin treatment also results in increased calcium influx over a sustained period of time, consistent with a role for intracellular calcium in the mechanism of capsaicin-induced sensory neuron degeneration (Jancso et al., 1978, 1984).

Both the rat and human VR1 receptors have been cloned (Caterina et al., 1997; Dubin et al., 2000; Hayes et al., 2000). hVR1 consists of 839 amino acids with six putative transmembrane domains (Dubin et al., 2000; Hayes et al., 2000), 92% homology to rat VR1, and a single nucleotide polymorphism resulting in a valine to isoleucine substitution at residue 585 located in transmembrane domain 5 (Hayes et al., 2000). When expressed in Xenopus oocytes or HEK293 cells, the hVR1 receptor exhibits pharmacological and physiological properties similar to those observed for the native vanilloid receptor, originally identified and characterized in sensory neurons (Dubin et al., 2000; Hayes et al., 2000). Much research in heterologous expression systems to date has focused on the pharmacology of VR1-mediated currents and intracellular calcium responses (Caterina et al., 1997; Tominaga et al., 1998; Dubin et al., 2000; Hayes et al., 2000). However, very little research has focused on characterization of VR1-mediated intracellular sodium responses, or the downstream cellular signaling consequences of VR1-mediated ion flux.

VR1 appears to mediate capsaicin-induced pain responses in vivo because this effect is eliminated in VR1 knockout mice (Caterina et al., 2000). Prolonged capsaicin exposure to VR1-transfected non-neuronal cells induces intracellular calcium responses and cell death (Caterina et al., 1997). Therefore, VR1 is a good candidate molecule for explaining the mechanism by which a distinct subpopulation of sensory neurons dies in response to capsaicin in vivo. Because the VR1 channel is permeable to both sodium and calcium, and mediates the influx of both of these ions into the cell, it is important to determine the relative calcium and sodium ion contributions to VR1-mediated cell death. The role these ions play in cell death is important for understanding the mechanism by which VR1 activation can kill cells and may begin to explain the mechanistic basis for capsaicin-induced neurodegeneration in vivo. Therefore, we have evaluated the ionic dependence of VR1-mediated cell death in VR1-expressing HEK293 cells. These data suggest that sustained, increased intracellular calcium, but not sodium concentration, primarily mediates VR1-dependent toxicity in these cells.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Capsaicin, resiniferatoxin, capsazepine, ruthenium red, and propidium iodide (PI) were all purchased from Sigma Chemical (St. Louis, MO). Fura-2/AM, SBFI/AM, fluo-4/AM, and CM-H₂DCFDA were all purchased from Molecular Probes (Eugene, OR).

Cloning and Transfection of hVR1 into HEK293 Cells. The hVR1 cDNA was isolated by screening of a human thalamus library and cloned as previously described (Dubin et al., 2000). HEK293 cells were transfected with hVR1 in pcDNA3.1zeo(+) by using the Effectene nonliposomal lipid-based transfection kit (QIAGEN, Valencia, CA). hVR1/HEK293 cells were routinely grown as monolayers under selection (zeocin; 200 µg/ml; Invitrogen, Carlsbad, CA) in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum, glutamine (2 mM), and penicillin/streptomycin (50 units/ml) in 5% CO₂ at 37°C. Cells were passaged frequently, every 3 to 5 days, to avoid acidic medium exposure. Cells were passaged without enzymes or Ca²⁺ chelators. Stably expressing hVR1 cell lines were screened for expression by using fluo-3/AM in conjunction with a FLIPR (FL-101; Molecular Devices, Sunnyvale, CA) at room temperature. The clones exhibiting the most robust influx of Ca²⁺ upon application of capsaicin were studied further but clone 3 was used for all subsequent studies because the response to capsaicin was robust for >30 passages (hVR1/HEK). Zeocin-resistant clones were shown to express the intact exogenous human VR1 receptor gene by using reverse transcription-polymerase chain reaction of primers designed to produce a product corresponding to coding and vector 3'-untranslated sequences (data not shown).

Ca²⁺ Influx Measurements with FLIPR. Transfected cells were seeded onto poly-D-lysine-coated, black-walled 96-well plates (Biocoat, 354640; BD Biosciences, San Jose, CA) at about 40,000 cells/well and grown for at least 1 day in culture medium to confluence. On the day of the experiment, media were removed using a 12-prong aspirator, incubated in 100 µl of fluo-3 (2 µM; Molecular Probes) with pluronic acid (0.04%; Molecular Probes) for 1 h at room temperature. After loading the cells, the fluo-3 was aspirated, 160 µl of buffer was added to all wells, and intracellular Ca²⁺ levels were subsequently assayed using the FLIPR (Molecular Devices) to simultaneously monitor fluo-3 fluorescence in all wells (_excitation = 488 nm, _emission = 540 nm). Cells were challenged with agonists (at 10-fold concentration in 20 µl added to 180 µl at a velocity of 20 µl/s) and the fluorescence intensity was captured every 3 s for the first 3 min after agonist addition. The contents of the wells were mixed three times (40 µl) after the additions. Antagonists were added on line (9-fold concentration in 20 µl added to 160 µl at a velocity of 20 µl/s) and fluorescence intensity was captured every 3 s for 3 min before agonist addition. The saline buffer used for these experiments contained 130 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 20 mM HEPES, pH 7.4).

[³H]Resiniferatoxin Binding Assay. Cell membranes were prepared by washing cells with Hanks' balanced salt solution (HBSS). Cells were dissociated with cell dissociation buffer (Sigma Chemical) and then centrifuged at 1000g for 5 min. Cell pellets were homogenized in cold 20 mM HEPES buffer, pH 7.4, containing 5.8 mM NaCl, 320 mM sucrose, 2 mM MgCl₂, 0.75 mM CaCl₂, and 5 mM KCl and centrifuged at 1000g for 15 min. The resultant supernatant was then centrifuged at 4000g for 15 min. The pellet membranes were stored at 80°C. The binding assay procedure was modified from what has been described previously (Szallasi and Blumberg, 1993). Briefly, about 120 µg of protein/ml membranes was incubated with the indicated concentration of [³H]RTX (PerkinElmer Life Sciences, Boston, MA) in 0.5 ml of the HEPES buffer, pH 7.4, containing 0.25 mg/ml fatty acid-free bovine serum albumin at 37°C for 60 min and the reaction mixture was cooled to 4°C. Then ₁-acid glycoprotein (0.1 mg) was added to each sample and was incubated at 4°C for 15 min. The samples were centrifuged at 18,500g for 15 min. The tip of the microcentrifuge tube containing the pellets was cut off. The nonspecific binding was tested in the presence of 200 nM unlabeled RTX. Bound radioactivity was quantified by scintillation counting.

Cytoxicity Measurements. Cell viability was measured using two methods. The first method is a single-cell digital imaging-based method. Visible light digital images of the cells were collected through a 20× objective (Zeiss Fluar, numerical aperture 0.75) and imaged using an Attofluor Imager device (Atto Instruments, Rockville, MD). Afterward, the same field of cells, loaded with 10 µM PI for 10 min, at room temperature, were imaged for PI fluorescence. The dye was excited by light from a mercury lamp source, passed through a 555-nm excitation bandpass filter (10-nm bandwidth), reflected off of a mirror, and then passed through the 20× objective. Emitted light was transmitted through a pinkel filter (Omega Optical), which allows 600-nm (20-nm bandwidth) wavelength light emitted from the dye to be collected to an Attofluor-intensified CCD camera. Images were visualized and overlaid using Attofluor RatioVision software (Atto Instruments, Brattleboro, VT).

Fura-2 Imaging. Cells were loaded with 5 µM fura-2/AM for 1 h at 37°C. They were washed once with HBSS (Invitrogen) and assayed in HBSS buffer. Cells were placed onto the stage of a modified Attofluor Imager. High-speed, dual excitation of fura-2 was carried out using a RatioArc High-Speed Excitor, which rapidly switched excitation light between 334- and 380-nm wavelengths (10-nm bandwidth filters). Emitted light was transmitted through a 400-nm dichroic mirror, collected on an Attofluor-intensified CCD camera, and ratio-images were digitized, and analyzed, using Attofluor RatioVision software. Fluorescence intensity from each wavelength was analyzed from each sample by dividing the entire microscopic field into 99 regions of interest (ROI), which corresponded to approximately 200 to 400 cells. Background for each wavelength was defined by the fluorescence intensity of a region in the field with no cells, and was excluded using Attofluor RatioVision software algorithms. Some experiments were calibrated to absolute calcium values. To achieve this a standard two-point calibration method was performed using 50 µM fura-2 pentasodium salt (Molecular Probes), a high-calcium buffer of 10 mM CaCl₂, and a zero calcium buffer of 10 mM K₂EGTA (Molecular Probes). Fluorescence intensity at 334 and 380 nm was measured for both the high-calcium buffer, and the zero calcium buffer, and values were applied to the standard radiometric dye formula [Ca²⁺] = K_d[(R  R_min)/(R_max  R)]. K_d = 224 nM, R_min = ratio for zero calcium, R_max = ratio for high calcium,  = fluorescence of zero calcium at 380 nm/fluorescence of high calcium at 380 nm, and R = ratio of the experimental 340 nm fluorescence/380 nm fluorescence (Grynkiewicz et al., 1985).

Simultaneous SBFI and Fluo-4 Imaging. Cells were loaded with 5 µM SBFI/AM (Molecular Probes) and 1 µM fluo-4-AM (Molecular Probes) for 1 h at 37°C. They were washed once with HBSS and assayed in HBSS buffer. Cells were placed onto the stage of a modified Attofluor Imager. High-speed excitation of SBFI and fluo-4 was carried out using a RatioArc High-Speed Excitor, which rapidly switched excitation light among 334-, 380-, and 488-nm wavelengths (10-nm bandwidth filters, sampling rate was approximately one ratio image and one fluo-4 image per second). Emitted light was transmitted through a 490-nm longpass filter and collected to the Attofluor-intensified CCD camera. SBFI ratio-images and fluo-4 single wavelength images were digitized, and analyzed, using Attofluor RatioVision software. No cross talk between any of the dyes was observed, because no signal above background was detected when SBFI-alone-loaded cells were excited with the fluo-4 excitation wavelength, or when fluo-4-alone-loaded cells were excited with either of the two SBFI excitation wavelengths. Analysis of the data from each wavelength was carried out as described above.

Imaging of Reactive Oxygen Species (ROS). Cells were loaded with the reactive oxygen species dye CM-H₂DCFDA (5 µM; Molecular Probes) for 1 h at 37°C. They were washed once with HBSS and assayed in HBSS buffer. Cells were placed onto the stage of a modified Attofluor Imager. Excitation of the dye was carried out at 488 nm (10-nm bandwidth filter). Emitted light was transmitted through a 490-nm longpass filter, collected on the Attofluor-intensified CCD camera, images were digitized, and analyzed, using Attofluor RatioVision software.

Statistical Analysis. Data comparisons of two groups were performed by Student's t test on GraphPad Prism software (GraphPad Software, San Diego, CA). Data comparisons of three or more groups were performed by one-way ANOVA followed by Tukey's post hoc analysis on GraphPad Prism software.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Cloning of Human VR1 and Production of Stably Transfected hVR1 Cells. The human VR1 was cloned from a human thalamus cDNA library (Dubin et al., 2000). HEK293 cells stably transfected with hVR1/pCDNA3.1(zeo) produced robust increases in intracellular Ca²⁺ when challenged with 300 nM agonists capsaicin and RTX (Fig. 1A). Pretreatment with 10 µM antagonists ruthenium red and capsazepine completely blocked the agonist-induced response (Fig. 1B). The agonist-induced increase in intracellular Ca²⁺ was dependent on extracellular Ca²⁺ (data not shown). This stable clone appropriately exhibited specific binding for [³H]resiniferatoxin. Concentration-dependent inhibition of the specific binding was observed in the presence of varying concentration of VR1 receptor agonists and antagonists (Fig. 1C).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Agonist and antagonist responses in the human VR1-expressing HEK293 cell line. A fluorescence assay measuring intracellular levels of Ca2+ was used to quantitate VR1 receptor function in fluo-3-loaded human VR1/HEK cells by using the FLIPR. A, agonists capsaicin (CAP) and RTX (both at 300 nM) elicited an increase in intracellular Ca2+. B, ruthenium red (10 µM; stippled bar) and capsazepine (10 µM; stippled bar) blocked the response to 100 nM CAP (solid bar). The CAP response in the absence of antagonists is shown (buffer). C, hVR1-HEK293 cell membranes were incubated with [3H]RTX (0.4 nM) and varying concentrations of vanilloid analogs at 37°C for 60 min. The data are representative of two experiments with each point assayed in duplicate. The results showed that vanilloid analogs used in this study dose dependently inhibited [3H]RTX binding. The IC50 values of RTX, capsaicin, and capsazepine were 0.78 ± 0.15, 630 ± 202, and 206 ± 43 nM, respectively.

VR1-Mediated Intracellular Calcium and Sodium Ion Concentration Changes in VR1-Expressing HEK293 Cells. VR1 stably transfected HEK293 cells were loaded with fura-2/AM to monitor intracellular calcium, or SBFI/AM to monitor intracellular sodium. Single-cell fluorescence imaging experiments, carried out on an Attofluor Imager device, revealed increases in both intracellular calcium and sodium concentrations in response to 3 µM capsaicin (Fig. 2, A and B). Conversion of the fura-2 ratio data to absolute calcium values revealed that the intracellular calcium concentration changed from a baseline of 48 ± 11 to 753 ± 26 nM in response to 3 µM capsaicin (p < 0.001, n = 4-14 independent samples). Pretreatment with the VR1 receptor antagonist capsazepine (10 µM) for 1 to 5 min completely blocked both the calcium and sodium response to capsaicin (Fig. 2, A and B). Both capsaicin-induced Na⁺ and Ca²⁺ responses were concentration-dependent. The capsaicin EC₅₀ values for both the calcium and sodium responses were closely matching, 81 versus 55 nM, respectively (Fig. 2, C and D).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Characterization of hVR1 cell calcium and sodium responses to capsaicin. A, intracellular calcium response to 3 µM capsaicin () or 3 µM capsaicin after 10-min pretreatment with 10 µM capsazepine (). Data for each trace is the average of 99 ROI collected from a microscopic field of hVR1-HEK293 cells, representative of 15 to 20 like experiments with similar results. B, intracellular sodium response to 3 µM capsaicin () or 3 µM capsaicin after 10-min pretreatment with 10 µM capsazepine (). Data for each trace is the average of 99 ROI collected from a microscopic field of hVR1-HEK293 cells, representative of 10 to 15 like experiments with similar results. C, concentration-response curve for intracellular calcium. EC50 = 81 nM (95% confidence interval, 72-92 nM). Data are the average of three independent samples per point. D, concentration-response curve for intracellular sodium. EC50 = 55 nM (95% confidence interval, 31-98 nM). Data are the average of three independent samples per point.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Characterization of hVR1 cell calcium and sodium responses upon antagonist addition after agonist activation. Intracellular calcium and sodium responses to 10 µM capsaicin were imaged simultaneously in the same field of cells through concurrent monitoring of fluo-4, calcium indicator, and SBFI, sodium indicator. A, 50 µM capsazepine was added 1.5 to 2 min after capsaicin. B, 10 µg/ml ruthenium red was added 1.5 to 2 min after capsaicin. Data for each trace are the average of 99 ROI collected from a microscopic field of hVR1-HEK293 cells, representative of six to seven like experiments with similar results.

Pharmacological Characterization of VR1-Mediated Toxicity in VR1-Expressing HEK293 Cells. Morphological assessment by using digital imaging at 20× magnification revealed increased numbers of cells dead or dying after 12 to 15 h of continuous exposure to capsaicin (3 µM) treatment compared with vehicle-treated control cells. Extensive cellular debris was evident, whereas control vehicle-treated cells appeared alive and healthy (Fig. 4A). Overlay of PI fluorescence showed a significantly increased number of dead or dying cells in the capsaicin-treated condition versus control or ruthenium red-protected cells (Fig. 4, A and B). Concurrently, after 12 to 15 h of continuous capsaicin treatment, cells stained more extensively with the ROS dye CM-H₂DCFDA, compared with controls, thereby indicating increased ROS in the dying cells (Fig. 4B).

View larger version (91K): [in this window] [in a new window]   Fig. 4.   Characterization of capsaicin toxicity in hVR1 cells. Cells were treated with vehicle control, 3 µM capsaicin, or 3 µM capsaicin plus 10 µg/ml ruthenium red (RR) for 12 to 15 h in media at 37°C, 5% CO2. A, visible light digital images and fluorescence images of the dead cell dye PI (excitation = 555, emission = 620) were collected from the same field of cells at the end of the treatment and overlaid. B, average ± S.E.M. number of cells exhibiting PI fluorescence above background (n = 3/condition). ***, p < 0.001 with respect to control and 10 µg/ml RR, analyzed by one-way ANOVA followed by Tukey's post hoc analysis. C, cells were treated with vehicle control or 3 µM capsaicin for 12 to 15 h in media at 37°C, 5% CO2 then loaded with the ROS dye CM-H2DCFDA for 1 h, and dye fluorescence was imaged (excitation = 488, emission = 520). Average ± S.E.M. number of cells exhibiting CM-H2DCFDA fluorescence above background (n = 3/condition). *, p < 0.05 analyzed by Student's t test.

View larger version (10K): [in this window] [in a new window]   Fig. 5.   Concentration-response of capsaicin toxicity in hVR1 cells. A, cells were treated with the indicated concentrations of capsaicin plus vehicle (), or capsaicin after 10-min pretreatment with 10 µM capsazepine () in parallel for 12 to 15 h in media at 37°C, 5% CO2. Alamar blue fluorescence was measured and expressed as percentage of control as described under Experimental Procedures. The EC50 for capsaicin plus vehicle is 125 nM (95% confidence interval, 105-149 nM). The EC50 for capsaicin plus 10 µM capsazepine pretreatment is 649 nM (95% confidence interval, 561-749 nM). Data are the average of eight experiments. B, cells were treated with the indicated concentrations of capsaicin plus vehicle (), capsaicin after 10-min pretreatment with 2 µg/ml ruthenium red (), or capsaicin after 10-min pretreatment with 6 µg/ml ruthenium red () in parallel for 12 to 15 h in media at 37°C, 5% CO2. The EC50 for capsaicin plus vehicle is 67 nM (95% confidence interval, 59-76 nM). The EC50 for capsaicin plus 2 µg/ml ruthenium red pretreatment was 227 nM (95% confidence interval, 148-348 nM). The EC50 for capsaicin plus 6 µg/ml ruthenium red pretreatment is >30 µM. Data are the average of eight experiments.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Effect of antagonist post-treatment on capsaicin toxicity in hVR1 cells. Cells were treated with the indicated concentrations of capsaicin plus vehicle (), capsaicin followed by 10 µM capsazepine 5 min later (), or capsaicin followed by 10 µM capsazepine 30 min later () in parallel for 12 to 15 h in media at 37°C, 5% CO2. The EC50 for capsaicin plus vehicle is 125 nM (95% confidence interval, 105-149 nM). The EC50 for capsaicin plus 10 µM capsazepine at 5 min is 591 nM (95% confidence interval, 543-642 nM). The EC50 for capsaicin plus 10 µM capsazepine at 30 min is 482 nM (95% confidence interval, 398-582 nM). Data are the average of eight experiments.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Effects of decreased extracellular calcium on capsaicin-induced calcium, sodium, and toxicity responses in hVR1 cells. Intracellular calcium and sodium responses to 10 µM capsaicin were imaged simultaneously in the same field of cells through concurrent monitoring of fluo-4 and SBFI. A, calcium and sodium responses in 1.8 mM extracellular Ca2+. B, calcium and sodium responses in 0.2 mM extracellular Ca2+. C, calcium and sodium responses in 0.2 mM extracellular Ca2+ plus 1 mM EGTA. Data for each trace is the average of 99 ROI collected from a microscopic field of hVR1-HEK293 cells, representative of three to four like experiments with similar results. D, measurement of Alamar blue fluorescence in cells treated for 12 to 15 h as indicated (***, p < 0.001 versus control, *, p < 0.05 versus control, analyzed by one-way ANOVA followed by Tukey's post hoc analysis).

pH Activation of VR1 Mediates a Transient Calcium Response but No Toxicity in VR1-Expressing HEK293 Cells. The VR1 channel can be activated by acidification as well as by pharmacological agonists (Tominaga et al., 1998). Therefore, it was of interest to characterize the calcium response to lowered pH in the VR1 stable cell line and relate it to toxicity. Basal intracellular calcium was monitored in fura-2-loaded cells at pH 7.5. Upon addition of HCl, buffer pH rapidly dropped to 6.3, and the cells underwent a transient increase in intracellular calcium, which returned to basal levels with 2 min (Fig. 8A). Pretreatment with the antagonist capsazepine completely blocked this transient calcium response to HCl, indicating that it was VR1-mediated (Fig. 8A). However, the same low pH challenge did not mediate any toxicity compared with control cells monitored in pH 7.5 buffer (Fig. 8B). Therefore, these data demonstrated that a VR1-mediated transient increase in intracellular calcium was not sufficient to trigger death of VR1-expressing HEK293 cells.

View larger version (10K): [in this window] [in a new window]   Fig. 8.   Characterization of pH change-induced intracellular calcium responses in hVR1 cells, and lack of toxicity. A, intracellular calcium responses to addition of HCl such that the extracellular buffer decreases in pH from a baseline of pH 7.5 down to pH 6.3 at the time indicated by the arrow monitored by using fura-2. Response to HCl plus vehicle () and response to HCl upon 10-min preincubation with 50 µM capsazepine (). B, measurement of Alamar blue fluorescence in cells treated for 12 to 15 h as indicated (***, p < 0.001 versus control, analyzed by one-way ANOVA with Tukey's post hoc analysis).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present study aimed to develop an experimental system to recapitulate capsaicin-induced cell death and to use that system, an HEK cell line stably expressing exogenously transfected VR1 receptor, to dissect the roles of different intracellular ionic changes in capsaicin toxicity. We demonstrated in this experimental system, using simultaneous ion imaging experiments as well as more classical methods such as alterations in extracellular buffer composition, that a sustained increase in intracellular calcium was the predominant ionic change inside the cell that lead to hVR1-mediated toxicity. In VR1-expressing HEK cells, neither VR1-mediated transient calcium increases nor sustained intracellular sodium changes contributed to VR1-mediated toxicity.

To date, both the rat and human orthologs of VR1 have been cloned (Caterina et al., 1997; Hayes et al., 2000). Both have been extensively characterized for tissue expression, electrophysiological function, pharmacology, and calcium responsivity (Caterina et al., 1997; Cesare et al., 1999; Dubin et al., 2000; Hayes et al., 2000; Jerman et al., 2000). The VR1 channel is expressed predominantly in small nociceptive neurons of the dorsal root ganglion (Tominaga et al., 1998). It is a ligand-gated, nonselective cation channel that shows pronounced outward rectification (Caterina et al., 1997). The receptor is activated by capsaicin (Szolcsanyi, 1996) and resiniferatoxin (Szolcsanyi et al., 1991). Its activation is blocked by the antagonist capsazepine (Bevan et al., 1992) and vanilloid blocker ruthenium red (Wood et al., 1988). Pharmacology of the rat VR1 receptor intracellular calcium response correlates well with what has been described in neurons (Jerman et al., 2000; Smart et al., 2000).

When overexpressed in non-neuronal cells, prolonged activation of rat VR1 leads to cell death (Caterina et al., 1997), a property that recapitulates capsaicin excitatory neurotoxicity in primary sensory neurons (Jancso et al., 1977, 1985; Wood et al., 1988). However, the VR1-mediated ionic signaling mechanisms that trigger cell death have not been determined. Because VR1 is functionally similar to other types of nonselective ligand-gated cation channels that also lead to cell death upon excessive activation, we reasoned that the ionic signaling trigger might also be comparable. In the ionotropic glutamate receptor system, it has been well established that calcium is a necessary trigger to initiate excitotoxicity in both neurons as well as in transfected non-neuronal cell lines (Choi, 1987; Munir et al., 1995; Grant et al., 1997). Sustained intracellular calcium increases can result in the detrimental overactivation of downstream signaling enzymes such as protein kinases, phospholipases, and nitric-oxide synthase. Additionally, it may lead to a loss in the mitochondrial membrane potential that collectively may contribute to cell death (Choi, 1987; Munir et al., 1995; Grant et al., 1997). However, increased intracellular sodium has also been implicated to play a significant role in ionotropic glutamate receptor-mediated cell death (Raymond et al., 1996; Itoh et al., 1998). Therefore, the present study characterized VR1-mediated calcium and sodium signaling, and determined whether either or both of these intracellular ionic changes contribute to cell death. Very much like what has been reported for glutamate receptors, our studies have shown that an increase in intracellular calcium is necessary for induction of VR1-mediated toxicity in the stably transfected HEK293 cell model system. In addition to calcium, we characterized the VR1-mediated intracellular sodium response. Although an increase in intracellular sodium coincided with the calcium response to VR1 activation, the data do not suggest that sodium significantly contributes to VR1-mediated toxicity.

The data demonstrated that VR1 antagonist addition after agonist activation resulted in a diverging directional response of the intracellular calcium versus the intracellular sodium concentrations. Antagonists caused the calcium to return toward baseline, whereas the sodium remained elevated, or in some cases even slightly increased. Although the data in the present studies do not mechanistically explain these observations, some possibilities may be speculated. Antagonist binding after the channel already has been opened could result in an intermediate channel conformation, distinct from either the open conformation or the closed conformation, which may be more selectively permeable to monovalent cations. Evidence that a conformation with greater monovalent cation selectivity can occur has been reported where mutation of aspartate residue 646 on VR1 to an asparagine confers greater selectivity for sodium over divalent cations compared with wild type (Garcia-Martinez et al., 2000). Alternatively, the effect may be due to the way the HEK293 cells handle the ionic load. The cells may pump out calcium more rapidly and efficiently than sodium. Therefore, once the ion source is blocked, the calcium ions may be cleared quickly, whereas sodium clearance may be relatively delayed.

The present studies demonstrated that the intracellular calcium response to protons was transient. This response was VR1-mediated because it was completely blocked in the presence of capsazepine. The response was not sufficient to confer VR1-mediated toxicity in these cells, suggesting that a sustained intracellular calcium response, like that which is elicited by capsaicin, is necessary. The hVR1 channel has been shown to mediate sustained whole-cell current responses to protons (Hayes et al., 2000). However, reports of VR1-mediated intracellular calcium responses to protons are consistent with our data in that the calcium responses were transient (Jerman et al., 2000). In fluorescence studies, the intracellular milieu is more similar to the viability studies in which the cells are intact than in whole-cell patch-clamp recordings where intracellular second messengers are lost. The discrepancy in time course of the response to protons when comparing a whole-cell current response and an intracellular calcium response might be explained by a number of possibilities. It is possible that the Ca²⁺-free solutions used in the electrophysiological studies modified the kinetics of the low pH response. Other possibilities may include a proton-dependent, selective desensitization of channel calcium permeability, or a proton-dependent activation of cellular clearance mechanisms for calcium.

A detailed understanding of hVR1 ionic signaling is crucial for discerning the downstream cellular consequences of activation of this important molecular target in peripheral pain processing pathways. Both agonists, as well as antagonists, of VR1 have been evaluated for analgesic efficacy in the clinic. Topical application of vanilloids, such as capsaicin (0.025% Zostrix and 0.075% Zostrix HP), have been used to mitigate neuropathic pain and treat the intractable pain associated with postherpetic neuralgia, diabetic neuropathy, postmastectomy pain, complex regional pain syndromes, and rheumatoid arthritis (Rowbotham, 1994; Szallasi and Blumberg, 1996; Robbins et al., 1998). With prolonged exposure to capsaicin, nociceptor cells not only become insensitive to this agonist but also to other noxious stimuli (Szolcsanyi, 1993). The mechanism by which capsaicin produces analgesia is not known but likely includes desensitization of nociceptive sensory neurons, depletion of peptides from peripheral terminals, as well as damage to sensory nerves (Jancso et al., 1977; Rowbotham, 1994). The irritancy of capsaicin severely limits its use, and the discovery of novel compounds that block acidic and/or thermal activation of capsaicin-sensitive receptors is sought. However, the vanilloid blocker ruthenium red and the antagonist capsazepine, although exerting antinociceptive effects in a behavioral study (Santos and Calixto, 1997; Kress and Zeilhofer, 1999), have not proven to be effective analgesics in humans (Kress and Zeilhofer, 1999). Perhaps, the present studies, in conjunction with others, will advance a comprehensive understanding of the role of the VR1 channel in pain processing and neurodegeneration. Thereby, more effective treatments for pain, as well as for other nervous system disorders, may be more rationally discovered.