Transient receptor potential vanilloid type 1 activation down-regulates voltage-gated calcium channels through calcium-dependent calcineurin in sensory neurons.

Calcium influx through voltage-activated Ca(2+) channels (VACCs) plays a critical role in neurotransmission. Capsaicin application inhibits VACCs and desensitizes nociceptors. In this study, we determined the signaling mechanisms of the inhibitory effect of capsaicin on VACCs in primary sensory neurons. Whole-cell voltage clamp recordings were performed in acutely isolated rat dorsal root ganglion neurons. Capsaicin caused a profound decrease in the Ca(2+) current (I(Ca)) density in capsaicin-sensitive, but not -insensitive, dorsal root ganglion neurons. At 1 mum, capsaicin suppressed about 60% of N-, P/Q-, L-, and R-type I(Ca) density. Pretreatment with iodoresiniferatoxin, a specific transient receptor potential vanilloid type 1 (TRPV1) antagonist, or intracellular application of 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid blocked the inhibitory effect of capsaicin on I(ca). However, neither W-7, a calmodulin blocker, nor KN-93, a CaMKII inhibitor, attenuated the inhibitory effect of capsaicin on I(Ca). Furthermore, intracellular dialysis of deltamethrin or cyclosporin A, the specific calcineurin (protein phosphatase 2B) inhibitors, but not okadaic acid (a selective protein phosphatase 1/protein phosphatase 2A inhibitor), abolished the effect of capsaicin on I(Ca). Interestingly, 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid, deltamethrin, cyclosporin A, and okadaic acid each alone significantly increased the I(Ca) density and caused a depolarizing shift in the voltage dependence of activation. Immunofluorescence labeling revealed that capsaicin induced a rapid internalization of Ca(V)2.2 channels on the membrane. Thus, this study provides novel information that VACCs are tonically modulated by the intracellular Ca(2+) level and endogenous phosphatases in sensory neurons. Stimulation of TRPV1 by capsaicin down-regulates VACCs by dephosphorylation through Ca(2+)-dependent activation of calcineurin.

TRPV1 1 is a nonselective cation channel with high Ca 2ϩ permeability and is the molecular target of capsaicin, the main pungent ingredient in chili peppers. The TRPV1 channel is expressed in subsets of primary sensory neurons and nerve terminals and plays an essential role in detecting noxious heat and several other nociceptive stimuli (1,2). Capsaicin can excite nociceptive sensory neurons and produce transient pain in animals and humans (3,4). Paradoxically, exposure to capsaicin desensitizes nociceptive sensory neurons and results in long lasting pain relief. For example, topical or local application of capsaicin is effective in treating many acute and chronic pain syndromes in patients (5)(6)(7). Capsaicin also causes an unexplained synaptic transmission block in the spinal cord dorsal horn (8).
The voltage-activated Ca 2ϩ channels (VACCs) play a critical role in signal transduction, synaptic neurotransmitter release, and nociceptive transmission (9 -11). VACCs also are an important molecular target of many analgesic drugs such as opioids (12,13). Interestingly, capsaicin causes a profound inhibition of VACC currents in dorsal root ganglion (DRG) neurons (14). However, the cellular and signaling mechanisms of the capsaicin effect on VACCs remain poorly understood.
Protein kinases and phosphatases are key enzymes in signal transduction pathways for a wide range of cellular processes. The enzymatic addition or removal of phosphate esters on serine and threonine hydroxyls alters the activity of many proteins that are essential to the characteristic structure and function of neurons. An important mechanism regulating VACC function is through phosphorylation by protein kinases and phosphatases (15)(16)(17)(18). We now show that a Ca 2ϩ -dependent serine/threonine phosphatase, calcineurin (protein phosphatase 2B), is critically involved in down-regulation of high voltage-activated Ca 2ϩ channels (HVACCs) by capsaicin in native DRG neurons. Furthermore, the basal intracellular Ca 2ϩ level and endogenous protein phosphatases tonically modulate the HVACC current in DRG neurons. These findings are not only important to our understanding of the functional interaction between TRPV1 and HVACCs in primary nociceptors but are also significant to our understanding of the Ca 2ϩ -dependent feedback regulation of neuronal Ca 2ϩ channels in general.

MATERIALS AND METHODS
Isolation of DRG Neurons-All procedures conformed to the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of the Pennsylvania State University College of Medicine. Male Sprague-Dawley rats (5-6 weeks old; Harlan, Indianapolis, IN) were anesthetized with halothane and then rapidly decapitated. The thoracic and lumbar segments of the vertebrate column were dissected. The DRGs were quickly removed and transferred immediately into Dulbecco's modified Eagle's medium (Invitrogen). After removal of attached nerves and surrounding connective tissues, the DRGs were minced with fine spring scissors, and the ganglion fragments were placed in a flask containing 5 ml of Dulbecco's modified Eagle's medium in which trypsin (type I, 0.5 mg/ml; Sigma), collagenase (type I A , 1 mg/ml; Sigma), and DNase (type I, 0.1 mg/ml; Sigma) had been dissolved. After incubation at 34°C in a shaking water bath for 30 min, soybean trypsin inhibitor (type II-s, 1.25 mg/ml; Sigma) was then added to stop trypsin digestion. The cell suspension was centrifuged (500 rpm, 5 min) to remove the supernatant and replenished with Dulbecco's modified Eagle's medium. The cultured DRG cells were replenished with B-27 medium containing neurobasal medium (Invitrogen) with B-27 supplement (2%; Invitrogen) and penicillin/streptomycin/glutamine supplement (1%; Invitrogen). Cells were then plated onto a 35-mm culture dish containing poly-L-lysine (50 g/ml) precoated coverslips and kept for at least 30 min before electrophysiological recordings. Recordings were made within 24 h after dissociation to keep the experiment as similar to in vivo as possible and to minimize the space clamp error, since DRG neurons produce neurites that are difficult to clamp following prolonged neuronal culture.
Electrophysiological Recordings-Patch electrodes with a resistance of ϳ2 megaohms were pulled from GC150TF-10 glass capillaries (inner diameter 1.17 mm, outer diameter 1.5 mm; Harvard Apparatus, Holliston, MA) using a micropipette puller (P-97 Sutter Instrument Co., Novato, CA) and fire-polished (DMF1000; World Precision Instruments, Sarasota, FL). Immediately before recording, neurons were treated with IB 4 -Alexa 594 (3 g/ml; Molecular Probes, Inc., Eugene, OR) in Tyrode solution for 10 min and then rinsed for at least 3 min. IB 4positive neurons were identified using a combination of epifluorescence illumination and differential interference contrast (ϫ20 -40) optics on an inverted IX70 microscope (Olympus Optical, Tokyo, Japan). Images of cells were taken with a CCD camera and displayed on a video monitor. Neurons were recorded in the whole-cell configuration using an EPC-10 amplifier (HEKA Instruments). After whole-cell configuration was established, the cell membrane capacitance and series resistance were electronically compensated. All experiments were performed at room temperature (ϳ25°C). Leak currents were subtracted using a hyperpolarizing P/4 protocol. Signals were filtered at 1 kHz, digitized at 10 kHz, and acquired using the Pulse program (HEKA).
Calcium currents (I Ca ) were recorded using extracellular solution consisting of 140 mM tetraethylammonium (TEA)-Cl, 2 mM MgCl 2 , 5 mM CaCl 2 , 10 mM glucose, 10 mM HEPES (pH 7.4 adjusted with TEA-OH, osmolarity 320 mosM). In some cells, barium was used as the charge carrier to record Ba 2ϩ currents (I Ba ) flowing through Ca 2ϩ channels. The extracellular solution consisted of 140 mM TEA, 2 mM MgCl 2 , 3 mM BaCl 2 , 10 mM glucose, 10 mM HEPES (pH 7.4 adjusted with TEA-OH, osmolarity 320 mosM). The pipette internal solution contained 120 mM CsCl, 1 mM MgCl 2 , 10 mM HEPES, 10 mM EGTA, 4 mM Mg-ATP, and 0.3 mM Na-GTP (pH 7.2 adjusted with CsOH, osmolarity 300 mosM). GTP and ATP were included in the pipette solution to minimize the "rundown" associated with the whole-cell recording. The whole-cell Ca 2ϩ current, carried by calcium (I Ca ) or barium (I Ba ), was elicited by a series of command potentials from Ϫ70 to 50 mV for 150 ms in 10-mV steps (5-s intervals) from a holding potential of Ϫ90 mV. The steady-state inactivation of VACCs was obtained by depolarizing cells to a series of prepulse potentials from Ϫ100 to 20 mV for 500 ms followed by a command potential to 0 for 150 ms.
Most drugs were dissolved in distilled water at 1,000 times the final concentration and kept frozen in aliquots. Nimodipine, thapsigargin, and KN-93 were initially dissolved in Me 2 SO to make the stock solution. The capsaicin, cyclosporine A, and deltamethrine stock solutions were prepared using ethanol. The stock solutions were further diluted in the extracellular or intracellular solution just before use and held in a series of independent syringes connected to corresponding fused silica columns. The end of the parallel columns was connected to a common silica column. The distance from the column mouth to the cell examined was about 100 m. Cells in the recording chamber were continuously bathed in extracelluar solution. Each drug solution was delivered to the recording chamber by gravity, and rapid solution exchange was achieved by controlling the corresponding valve switch (World Precision Instruments). Drugs and chemicals were purchased from Sigma, except -conotoxin GVIA, -agatoxin IVA, and -conotoxin MVIIC (Alomone Labs, Jerusalem, Israel) and W-7 and deltamethrine (Calbiochem).
Double Fluorescence Labeling of Ca V 2.2 and TRPV1/IB 4 in DRG Cells-The N-type current (Ca V 2.2 channel) is the most predominant subtype of VACC currents in DRG neurons (22). We determined the effect of capsaicin on the spatial distribution of Ca V 2.2 using immunofluorescence and confocal microscopy. DRG neurons cultured on glass coverslips were rinsed three times with phosphate-buffered saline. The cells were incubated with 1-10 M capsaicin or vehicle in phosphatebuffered saline for 20 s or 2 min. Additional cells were pretreated with either 10 M iodoresiniferatoxin, a specific TRPV1 antagonist (19), or 1 M deltamethrin, a membrane-permeable inhibitor of calcineurin (20), for 5 min before capsaicin incubation. After rinsing, the cells were immediately fixed with 4% formaldehyde for 10 min. The cells were quenched in 1% H 2 O 2 in Tris-buffered saline for 10 min and permeabilized in 0.1% Triton X-100 in Tris-buffered saline for another 10 min at room temperature. Then the cells were blocked in 4% normal goat serum for 30 min at room temperature.
For Ca V 2.2 and IB 4 double labeling, the cells were incubated overnight at 4°C with the primary antibody (rabbit anti-Ca V 2.2, 1:200 dilution; Chemicon (Temecula, CA) or Alomone (Jerusalem, Israel)) diluted in 2% normal goat serum. Subsequently, the cells were washed and incubated with peroxidase-conjugated AffiniPure goat anti-rabbit IgG (1:100 dilution; Jackson ImmunoResearch, West Grove, PA) for 1 h at room temperature. The cells were then rinsed and incubated with fluorescein isothiocyanate-conjugated tyramide (1:100 dilution; PerkinElmer Life Sciences) for 5 min at room temperature and washed. Finally, the cells were incubated with Alexa 594 conjugated to IB 4 (1 g/ml; Molecular Probes, Inc., Eugene, OR) for 1 h at room temperature.
For Ca V 2.2 and TRPV1 double immunolabeling, the cells were incubated overnight at 4°C with a mixture of two primary antibodies (rabbit anti-Ca V 2.2 and guinea pig anti-TRPV1, 1:1000 dilution; Neuromics, Northfield, MN). The cells were first processed with the Ca V 2.2 staining as described above. The cells were then incubated with biotinconjugated goat anti-guinea pig IgG (1:200 dilution; Jackson Immu-noResearch) for 1 h at room temperature. After rinsing, cells were incubated with Alexa 594 conjugated to streptavidin for 1 h at room temperature. Subsequently, the slides were covered and mounted with the ProLong mounting medium. Negative controls were performed by omitting the primary antibodies. The labeled cells were examined on a laser-scanning confocal microscope (Leica, Germany), and areas of interest were photodocumented. Confocal laser scanning microscopy was used for accurate localization of fluorescent markers, because the thin (0.3-m) optical sectioning generated by the confocal microscope eliminates the confounding effect of out-of-focus fluorescence.
Data Analysis-Data were analyzed using the PulseFit software program (HEKA). Whole-cell current-voltage (I-V) relationships for individual neurons were constructed by calculating the mean peak inward current at each test potential and normalized for cell capacitance. Conductance-voltage (G-V) curves were calculated by dividing current at each potential by the driving force (V Ϫ V r ), where V is the test potential and V r is the reversal potential extrapolated from the I-V curve. The normalized conductance (G/G max ) for G-V relationships and inactivation curves were fitted with a Boltzmann function G/G max ϭ G min ϩ (1/(1 ϩ exp(V Ϫ V 0.5 )/k))), where G min is the minimal conductance of VACCs, G max is the maximal conductance, V 0.5 is the voltage for 50% activation or inactivation of VACCs, and k is a voltage-dependent slope factor. The percentage inhibition of total I Ca and subtypes of I Ca was calculated as the ratio of capsaicin-inhibited I Ca to the total peak I Ca or subtypes of VACC currents during control, respectively. Statistical data are presented as means Ϯ S.E. All comparisons between means were tested for significance using Student's unpaired t test or one-way analysis of variance unless otherwise indicated. p Ͻ 0.05 was considered to be statistically significant.

Capsaicin Inhibits VACC Currents through TRPV1-
The whole-cell VACC currents were elicited by a series of depolarizing pulses (from Ϫ70 to 50 mV for 150 ms in 10-mV increments) from a holding potential of Ϫ90 mV. The inward I Ca was typically activated at Ϫ30 mV and reached its maximum at 0 mV ( Fig. 1). TRPV1 is expressed predominantly in isolectin B 4 (IB 4 )-binding DRG neurons (21). Because IB 4 can label living rat DRG neurons (15-30 m in diameter) and does not affect VACC currents (22,23), we used IB 4 as a marker to identify capsaicin-sensitive DRG neurons.
In the IB 4 -positive cells included for this study, 1 M capsaicin produced inward currents ranging from 200 to 1,000 pA (406.3 Ϯ 14.7 pA, n ϭ 108). Consistent with the previous study (24), we did not observe the low voltage-activated (T-type) Ca 2ϩ currents in capsaicin-sensitive small diameter DRG neurons. Thus, only HVACC currents were studied before capsaicin ap-plication and 1-2 min after washing of the capsaicin. The application of 1 M capsaicin caused a large inhibition of I Ca density in all 15 capsaicin-sensitive neurons tested (Fig. 1A). Capsaicin inhibited I Ca density by 55.7 Ϯ 6.9 and 60.9 Ϯ 6.8% at 0 and 10 mV, respectively, in these capsaicin-sensitive neurons (Fig. 1B). The inhibitory effect of I Ca by capsaicin was partially washed out ϳ3 min after capsaicin application. Prolonged exposure to capsaicin for more than 3 min led to complete inhibition of I Ca , which did not recover after washing. Although capsaicin did not alter the I-V curve, it shifted the reversal potential from 50.5 Ϯ 1.7 to 41.7 Ϯ 3.1 mV (Fig. 1B). In contrast, in seven capsaicin-insensitive DRG neurons, 1 M capsaicin had no significant effect on I Ca (Fig. 1C). Furthermore, capsaicin caused a similar degree of I Ca inhibition in nine IB 4 -negative but capsaicin-sensitive DRG neurons (Fig. 2C).
Many G protein-coupled receptor agonists, including opioids, inhibit HVACC currents of DRG neurons (12,22). It has been demonstrated that G protein ␤␥, rather than the G ␣ , subunit is responsible for inhibition of VACCs by G proteins. With a high voltage prepulse, the G ␤ ␥ subunit is believed to release from the Ca 2ϩ channel ␣ 1 subunit, thereby relieving the tonic inhibition of VACCs by G proteins and producing "prepulse facilitation" (25,26). We used a double-pulse voltage protocol (27) to determine whether G proteins are involved in the effect of capsaicin on I Ca . The voltage dependence of capsaicin-induced inhibition of I Ca was examined by a depolarizing prepulse to 100 mV, preceding the second test pulse (Fig. 1A). There was no voltage-dependent relief of inhibition on I Ca by capsaicin when a strong conditioning depolarization (ϩ100 mV) prepulse was used (Fig. 1A, n ϭ 15). Using this protocol, 1 M capsaicin produced a similar degree of inhibition of I Ca (Fig. 1, A and B). Furthermore, in nine separate DRG neurons treated with pertussis toxin (500 ng/ml) overnight to inactivate inhibitory G proteins, 1 M capsaicin still significantly inhibited I Ca (57.9 Ϯ 10.1% at 0 mV; Fig. 1D).
To determine whether the effect of capsaicin on I Ca is specifically mediated by TRPV1, we used a highly specific TRPV1 antagonist, iodoresiniferatoxin (19). Iodoresiniferatoxin (10 M) alone did not produce any currents and had no effect on I Ca (Fig. 2, A and B). In the presence of 10 M iodoresiniferatoxin, 1 M capsaicin failed to produce any inward currents and had no inhibitory effect on I Ca in all 13 IB 4 -positive neurons examined (Fig. 2, A and B).
Capsaicin Inhibits All Subtypes of HVACCs-We next determined which subtypes of HVACC currents are inhibited by capsaicin. The subtype-selective Ca 2ϩ channel blockers nimodipine (5 M, L-type), -conotoxin GVIA (2 M, N-type), -agatoxin IVA (100 nM, P/Q-type), and -conotoxin MVIIC (500 nM, N-and P/Q-type) were appropriately combined to pharmacologically isolate L-, N-, P/Q-, and R-type Ca 2ϩ currents, as described previously (22,28,29). Because 100 nM -agatoxin IVA alone is not sufficient to block the Q-type Ca 2ϩ channel (29), -conotoxin MVIIC was co-applied with -agatoxin IVA to define L-and R-type VACC currents. In this protocol, the cells were depolarized from Ϫ90 to 0 mV for 150 ms. A desired subtype of I Ca was isolated before application of 1 M capsaicin for 30 s. After capsaicin current was washed out, the I Ca was reexamined in capsaicin-sensitive neurons.
To determine the effect of 1 M capsaicin on R-type I Ca , -conotoxin MVIIC and -agatoxin IVA were co-applied with FIG. 1. Capsaicin inhibits HVACC currents in IB 4 -positive DRG neurons. A, representative traces showing I Ca before and after application of 1 M capsaicin in a capsaicin-sensitive DRG neuron. B, mean I Ca density before and after application of capsaicin at different test potentials (n ϭ 15). C, original recordings of I Ca before and after capsaicin application in a capsaicin-insensitive neuron. D, treatment with 500 ng/ml pertussis toxin had no effect on capsaicin-induced inhibition of I Ca (n ϭ 9). The peak current was divided by cell capacitance to yield the current density. All of the neurons were voltage-clamped at Ϫ90 mV and depolarized from Ϫ70 to 50 mV for 150 ms with 10-mV increments. Inset, the double-pulse voltage protocol with a depolarizing prepulse to 100 mV for 50 ms preceding the second test pulse. *, p Ͻ 0.05 compared with corresponding values before capsaicin application. nimodipine and -conotoxin GVIA (Fig. 3). The R-type (drugresistant) I Ca was very small in DRG neurons, and 1 M capsaicin suppressed 58.5 Ϯ 12.6% R-type I Ca (n ϭ 6). Cd 2ϩ (300 M) completely blocked the remaining I Ca . The N-type I Ca was isolated by application of nimodipine and -agatoxin IVA in a separate group of DRG neurons. Capsaicin inhibited 61.9 Ϯ 11.6% N-type I Ca in nine DRG neurons tested (Fig. 3). Furthermore, the L-type Ca 2ϩ current, isolated with -conotoxin GVIA, -conotoxin MVIIC, and -agatoxin IVA, was inhibited 48.5 Ϯ 8.9% (n ϭ 5; Fig. 3) by 1 M capsaicin. Additionally, 1 M capsaicin inhibited 57.1 Ϯ 14.9% P/Q-type Ca 2ϩ current, isolated using -conotoxin GVIA and nimodipine (n ϭ 5; Fig. 3).
Capsaicin Increases Steady-state Inactivation of I Ca -To further characterize the inhibitory effect of capsaicin on I Ca , we examined the effect of capsaicin on steady-state activation and inactivation kinetics of I Ca , illustrated in Fig. 4. Although 1 M capsaicin caused a pronounced inhibition of I Ca , it did not significantly alter the voltage dependence of activation (Fig.  4A, n ϭ 15). The decay phase of I Ca was fitted with one exponential function, and the decay time constant of I Ca was significantly reduced by 1 M capsaicin (from 51.6 Ϯ 3.0 to 39.9 Ϯ 3.3 ms, p Ͻ 0.05; Fig. 4B; n ϭ 15). In 13 additional capsaicinsensitive neurons, the steady-state inactivation of I Ca was examined using a series of prepulse potentials (Ϫ100 to 20 mV for 500 ms) followed by depolarizing the cell to 0 mV for 150 ms. Following application of 1 M capsaicin, the voltage-dependent steady-state inactivation of HVACC currents was shifted significantly to the left (more negative potentials) (Fig. 4B).
Both Extracellular and Intracellular Ca 2ϩ Are Required for the Effect of Capsaicin on I Ca -The TRPV1 channel is highly permeable to Ca 2ϩ (1). Because the degree of capsaicin-produced inhibition of I Ca was proportional to the amplitude of capsaicin currents (14), we determined the role of Ca 2ϩ influx in the inhibitory effect of capsaicin on I Ca . BAPTA, a rapid Ca 2ϩ chelator (30), was included in the pipette solution. Following the membrane rupture to achieve the whole-cell configuration, the intracellular solution exchange was allowed for 5 min before examining the effect of capsaicin on I Ca . In 11 capsaicin-sensitive neurons tested, 1 M capsaicin had no significant effect on I Ca when 10 mM BAPTA was included in the pipette solution (Fig. 5A). Capsaicin also failed to alter the inactivation kinetics of I Ca (Fig. 5B). Notably, I Ca was activated at about Ϫ30 mV and reached its peak at 10 mV in the presence of 10 mM BAPTA. Compared with the peak I Ca density and I-V curve recorded in the absence of BAPTA (95.4 Ϯ 10.7 pA/pF at 0 mV, n ϭ 15), intracellular dialysis of BAPTA significantly shifted the I-V curve to the right (more positive potentials) and increased the density of I Ca (164.3 Ϯ 22.2 pA/pF at 10 mV, n ϭ 11; Fig. 5A).
TRPV1 is present on both the cell membrane and endoplasmic reticulum in the DRG (31, 32). A rapid rise in intracellular Ca 2ϩ levels may be due to either Ca 2ϩ influx or release of Ca 2ϩ

FIG. 4. Effect of capsaicin on voltage dependence activation and steady-state inactivation of I Ca .
A, the voltage dependence activation curves of I Ca before and after 1 M capsaicin application in 15 capsaicin-sensitive neurons. Before capsaicin application, the V 0.5 and slope factor were Ϫ7.1 Ϯ 0.2 and 5.0 Ϯ 0.2 mV, respectively. After capsaicin, the V 0.5 and slope factor were Ϫ7.8 Ϯ 0.2 and 5.0 Ϯ 0.1 mV, respectively. Cells were depolarized to a series of depolarizing potentials (from Ϫ60 to 20 mV with 10-mV increments) from the holding potential of Ϫ90 mV. B, representative current traces and summary data showing steady-state inactivation of I Ca before and after 1 M capsaicin in 13 capsaicin-sensitive neurons. The V 0.5 was Ϫ38.9 Ϯ 2.4 and Ϫ45.9 Ϯ 2.5 mV (p Ͻ 0.05) before and after capsaicin application, respectively. The slope factor was Ϫ22.5 Ϯ 3.4 and Ϫ17.9 Ϯ 3.0 mV (p Ͻ 0.05) before and after capsaicin application, respectively. Cells were depolarized at a series of prepulse potentials (from Ϫ90 to 10 mV, 500 ms) and then depolarized to 0 mV. The normalized conductance (G/G max ) was fit to a Boltzmann's function. Data are presented in mean Ϯ S.E. *, p Ͻ 0.05 compared with corresponding values before capsaicin application.

FIG. 5. Effect of BAPTA on I Ca and capsaicin-induced inhibition of I Ca .
A, original current traces and summary data showing lack of effect of 1 M capsaicin on I Ca at different test potentials in 11 capsaicin-sensitive neurons recorded with 10 mM BAPTA internal solution. The peak current was divided by cell capacitance to yield the current density. Note that I Ca was activated at about Ϫ30 mV and reached its peak at 10 mV in the presence of BAPTA. B, representative current traces and summary data showing steady-state inactivation of I Ca recorded with 10 mM BAPTA (from Ϫ90 to 10 mV, 500 ms) before and after capsaicin (n ϭ 11). Before capsaicin, the V 0.5 and slope factor were Ϫ35.7 Ϯ 0.7 and Ϫ11.9 Ϯ 0.7 mV, respectively. After capsaicin, the V 0.5 and slope factor were Ϫ31.3 Ϯ 0.9 and Ϫ14.4 Ϯ 0.9 mV, respectively. Data are presented as mean Ϯ S.E. from an intracellular store. In 12 capsaicin-sensitive DRG neurons, we replaced Ca 2ϩ with Ba 2ϩ in the extracellular solution to determine the role of extracellular Ca 2ϩ in the effect of capsaicin on HVACCs. I Ba was elicited at Ϫ40 mV and reached its peak at Ϫ10 mV (Fig. 6A). In the Ca 2ϩ -free external solution, 1 M capsaicin still significantly inhibited I Ba , although its inhibitory effect on I Ba was significantly attenuated (37.5 Ϯ 5.0% at Ϫ10 mV, n ϭ 6; Fig. 6A). However, the capsaicin effect can be completely washed out in 3 min. Capsaicin did not significantly alter the steady-state inactivation kinetics of I Ba (Fig. 6B).
We subsequently determined whether the remaining inhibitory effect of capsaicin on I Ba was mediated by Ca 2ϩ released from the intracellular stores, such as the endoplasmic reticulum. The intracellular Ca 2ϩ store was depleted with thapsigargin, a specific Ca 2ϩ -ATPase inhibitor (33). In a total of 15 capsaicin-sensitive neurons, we pretreated the DRG cells with 5 M thapsigargin for 10 min and washed it out for at least another 3 min using Ba 2ϩ external solution before testing the effect of 1 M capsaicin on I Ba . The inhibitory effect of capsaicin on I Ba was completely blocked in all 15 neurons treated with thapsigargin (Fig. 6C).
Furthermore, to determine whether Ca 2ϩ release from the endoplasmic reticulum independent of TRPV1 activation can inhibit HVACCs, we examined the effect of caffeine, a selective activator for endoplasmic reticulum ryannodine receptors (34,35), on I Ba in 12 additional DRG cells. Bath perfusion of 30 mM caffeine for 1-2 min caused a significant and reversible reduction of I Ba (23.7 Ϯ 3.2% inhibition at Ϫ10 mV) (Fig. 6D).
Calmodulin and Ca 2ϩ /Calmodulin-dependent Protein Kinase II Are Not Involved in the Effect of Capsaicin on I Ca -Because the increase in intracellular Ca 2ϩ appeared essential for the inhibitory effect of capsaicin on VACCs, we next determined if calmodulin and Ca 2ϩ /calmodulin-dependent protein kinase II are involved in this effect. The specific calmodulin antagonist W-7 (36) or the selective Ca 2ϩ /calmodulin-dependent protein kinase II inhibitor KN-93 (37)  sity of I Ca (193.8 Ϯ 20.7 pA/pF at 10 mV, n ϭ 6) and shifted the I-V curve to the right (Fig. 7B). However, inclusion of 0.5 M okadaic acid in the pipette solution failed to attenuate the inhibitory effect of 1 M capsaicin on I Ca in all six capsaicinsensitive neurons tested (Fig. 7B).
Calcineurin (protein phosphatase 2B) is a Ca 2ϩ -sensitive protein phosphatase and can be activated by a rise in intracellular Ca 2ϩ . To determine if capsaicin inhibits I Ca through dephosphorylation mediated by calcineurin, 1 M deltamethrin, a specific inhibitor of calcineurin (20), was included in the pipette solution. Deltamethrin completely blocked the inhibitory effect of 1 M capsaicin on I Ca in all 10 DRG neurons tested (Fig. 8, A  and B). Similarly, intracellular dialysis with cyclosporin A (50 M), another selective calcineurin inhibitor (39), also abolished the effect of capsaicin on I Ca in another 10 DRG neurons examined (Fig. 8C). In the presence of deltamethrin or cyclosporin A, 1 M capsaicin failed to alter the inactivation kinetics of I Ca (Fig. 8, B and D). In separate DRG neurons, intracellular application of a lower concentration of deltamethrin (0.5 M, n ϭ 13) or cyclosporin A (20 M, n ϭ 10) only partially blocked the inhibitory effect of 1 M capsaicin on I Ca . Rapamycin binds to FK506-binding protein but does not inhibit calcineurin, because it cannot interact with the CNB subunit for steric reasons (40). In five separate DRG neurons, we found that intracellular dialysis of 20 M rapamycin did not significantly alter the effect of capsaicin on I Ca (data not shown).
Unexpectedly, inclusion of 1 M deltamethrin alone signifi-cantly increased the density of I Ca (175.9 Ϯ 13.4 pA/pF at 10 mV, n ϭ 10) and shifted the I-V curve of I Ca to more positive potentials by ϳ10 mV (Fig. 8A), compared with the I Ca recorded with the regular pipette solution. Similar effects on the I Ca density (185.5 Ϯ 29.3 pA/pF at 10 mV, n ϭ 10) and the I-V curve were observed with intracellular dialysis of 50 M cyclosporin A (Fig. 8C).
Capsaicin Induces Internalization of Ca V 2.2-Because capsaicin caused a profound and long lasting decrease in I Ca density, the pronounced inhibition of VACCs by capsaicin exposure cannot be fully explained with increased inactivation of HVACCs. Thus, we determined whether redistribution (e.g. internalization) of HVACCs plays a role in the effect of capsaicin on I Ca in DRG cells. Because the N-type Ca 2ϩ current (mediated by Ca V 2.2 channels) is the most prominent subtype of HVACC currents, we chose to examine the effect of 1-10 M capsaicin on the spatial distribution of Ca V 2.2 in DRG cells. The Ca V 2.2 distribution in TRPV1-and IB 4 -positive DRG neurons was studied using double immunofluorescence labeling and confocal microscopy. All negative controls (omitting primary antibodies) displayed no detectable staining. In vehicletreated DRG neurons, the cross-sectional confocal images of Ca V 2.2 immunoreactivity showed a bright staining pattern with a clear and well defined Ca V 2.2 immunoreactivity on the plasma membrane (Fig. 9). By contrast, the cells treated with 1-10 M capsaicin for 20 s and 2 min displayed a more diffuse and duller immunoreactivity of Ca V 2.2 in the cytoplasm near the cell membrane surface (Fig. 9). The inward spread of Ca V 2.2 immunoreactivity was more evident when compared with the labeling of TRPV1 and IB 4 present in the same DRG cell (Fig. 9). Furthermore, in DRG cells pretreated with either 10 M iodoresiniferatoxin or 1 M deltamethrin, the spatial distribution of Ca V 2.2 immunoreactivity on the cell membranes was not altered by capsaicin (data not shown). In DRG cells exposed to 1-10 M capsaicin for 3-5 min, Ca V 2.2 immunoreactivity completely disappeared from the plasma membrane. The cell membrane also became disintegrated following prolonged exposure to capsaicin, suggesting permanent damage to the cells (data not shown). DISCUSSION We have investigated the signaling mechanism underlying down-regulation of HVACCs by TRPV1 stimulation in native DRG neurons. Small sized DRG neurons are richly endorsed with TRPV1 (41, 42). Application of capsaicin produced a profound and sustained suppression of HVACC currents in DRG cells. We found that this inhibition was only limited to capsaicin-sensitive DRG neurons, and iodoresiniferatoxin, a highly specific TRPV1 antagonist, eliminated the effect of capsaicin. Hence, capsaicin inhibits HVACC currents specifically through TRPV1 in DRG neurons. Furthermore, we observed that capsaicin produced a similar degree of inhibition in all the subtypes (N-, P/Q-, L-, and R-type) of HVACCs. This suggests that a common mechanism is probably responsible for the effect of capsaicin on different subtypes of HVACCs. Although capsaicin had no effect on the voltage dependence activation of HVACCs, it increased the steady-state inactivation of HVACC currents. Thus, the inhibitory effect of capsaicin on HVACC currents appears to be due in part to a decrease in the open state probability of HVACCs in DRG neurons.
What then are the downstream signaling mechanisms responsible for the down-regulation of HVACCs by capsaicin in DRG neurons? Stimulation of G protein-coupled receptors inhibits HVACCs through G protein ␤␥-subunits (25,26). However, we obtained no evidence that G proteins are involved in the effect of capsaicin on HVACCs. This is because depolarizing pulses did not lead to a prepulse-induced facilitation (relief of inhibition) of Ca 2ϩ currents. Furthermore, pretreatment of DRG neurons with pertussis toxin to inactivate inhibitory G i/o proteins had no effect on capsaicin-induced inhibition on HVACC currents. Thus, the G proteins are not involved in the rapid inhibition of HVACCs by capsaicin.
Because TRPV1 is highly permeable to Ca 2ϩ (1, 2), capsaicin can induce substantial Ca 2ϩ influx into the DRG cell. In our study, intracellular dialysis with BAPTA, a rapid Ca 2ϩ chelator (30), abolished the inhibitory effect of capsaicin on Ca 2ϩ currents, suggesting that the rise in intracellular Ca 2ϩ is the key event in capsaicin-induced inhibition of HVACC currents. Although these data clearly indicate the importance of intracellular Ca 2ϩ in the effect of capsaicin on HVACCs, they do not discriminate the sources of Ca 2ϩ . Capsaicin can activate TRPV1 located on both the plasma membrane and endoplasmic reticulum (31,32). Consequently, we examined more closely the relative contribution of intracellular and extracellular Ca 2ϩ to the effect of capsaicin on HVACC currents. When the extracellular Ca 2ϩ was replaced with barium, the inhibitory effect of capsaicin was significantly attenuated, and its effect was completely washed out. Furthermore, depletion of intracellular Ca 2ϩ with thapsigargin, a highly specific inhibitor of Ca 2ϩ -ATPases (33), abolished the capsaicin-induced inhibition on HVACC currents. Thus, the inhibitory effect of capsaicin on HVACCs is triggered by both Ca 2ϩ influx and release of Ca 2ϩ from intracellular stores. We found that stimulation of endoplasmic reticulum ryannodine receptors with caffeine (34,35), significantly inhibited I Ba in DRG neurons. These data strongly suggest that Ca 2ϩ release from the endoplasmic reticulum contributes to the inhibitory effect of capsaicin on HVACCs. In addition, capsaicin increases the steady-state inactivation of HVACC currents through an increase in intracellular Ca 2ϩ , which highlights a direct Ca 2ϩ -dependent inactivation of HVACCs in primary sensory neurons.
The most salient finding of this study is that calcineurin plays a pivotal role in down-regulation of HVACCs caused by TRPV1 stimulation in DRG neurons. Calcineurin is a Ca 2ϩ -dependent protein phosphatase and is enriched in the DRG neurons (43). We found that inhibition of calcineurin with intracellular dialysis of either deltamethrin or cyclosporin A, two structurally dissimilar calcineurin inhibitors, completely blocked the effect of capsaicin on HVACC currents. On the other hand, intracellular application of okadaic acid, a specific inhibitor for protein phosphatases protein phosphatase 1 and protein phosphatase 2A, failed to attenuate the effect of capsaicin. Deltamethrin and structurally related Type II pyrethroids are potent inhibitors of calcineurin (20,44,45). Cyclosporin A specifically binds to the intracellular receptor protein cyclophilin A, a member of the immunophilin protein family (39), and the resulting complex is an effective inhibitor of calcineurin. Interestingly, rapamycin, an inhibitor of FK506-binding protein but ineffective in inhibiting calcineurin activity (40), did not alter the effect of capsaicin on Ca 2ϩ currents. Calcineurin may be activated by Ca 2ϩ , calmodulin, and inhibitory G proteins (46,47). However, it appears that Ca 2ϩ is a direct activator of calcineurin following stimulation of TRPV1 in DRG neurons, since treatment with W-7, KN-93, or pertussis toxin failed to attenuate the effect of capsaicin. These data strongly suggest that an increase in intracellular Ca 2ϩ by TRPV1 leads to down-regulation of HVACCs by protein dephosphorylation mediated by Ca 2ϩ -dependent calcineurin.
We also demonstrate for the first time that stimulation of TRPV1 causes a rapid internalization of HVACCs in DRG neurons. Using double immunofluorescence labeling and confocal microscopy, we found that capsaicin induced a rapid internalization of Ca V 2.2 immunoreactivity in IB 4 -and TRPV1positive DRG neurons. Since capsaicin caused a similar degree of down-regulation in all four subtypes of HVACCs, it is probable that internalization occurs in the other three subtypes of HVACCs. It is not clear if the substrate of calcineurin is the HVACCs or a regulatory phosphoprotein that interacts with the channel in the DRG. We found that capsaicin shifted the FIG. 8. Effect of calcineurin inhibitors on I Ca and capsaicin-induced inhibition of I Ca . A, representative current traces and summary data showing lack of effect of 1 M capsaicin on I Ca at different test potentials (from Ϫ70 to 50 mV for 150 ms) following intracellular dialysis with 1 M deltamethrin (n ϭ 10). Note to the right of the I-V curve of I Ca , the increased I Ca density in the presence of deltamethrin. B, original current traces and summary data showing that capsaicin failed to alter the steady-state inactivation curve of I Ca following intracellular application of deltamethrin (n ϭ 5). The V 0.5 and slope factor before capsaicin application were Ϫ50.4 Ϯ 1.9 and Ϫ19.3 Ϯ 2.0 mV, respectively. After capsaicin, the V 0.5 and slope factor were Ϫ46.7 Ϯ 1.8 and Ϫ15.9 Ϯ 2.0 mV, respectively. C, representative current traces and summary data showing a lack of inhibitory effect of 1 M capsaicin on I Ca at different test potentials (from Ϫ70 to 50 mV) following intracellular dialysis with 50 M cyclosporin A (n ϭ 10). Note the I-V current shift and increased I Ca density by cyclosporin A. D, current traces and summary data showing that 1 M capsaicin failed to shift the inactivation curve in the presence of cyclosporin A (n ϭ 5). The V 0.5 was Ϫ47.8 Ϯ 2.0 and Ϫ46.8 Ϯ 2.7 mV before and after capsaicin application, respectively. The slope factor was Ϫ20.5 Ϯ 2.7 and Ϫ21.6 Ϯ 3.4 mV before and capsaicin application, respectively. reversal potential of HVACCs, which is probably due to dephosphorylation of HVACCs, resulting in a structural change of channel proteins, through activation of Ca 2ϩ -dependent calcineurin. In addition to direct phosphorylation and dephosphorylation of HVACCs (16,48), protein kinases and calcineurin can phosphorylate and dephosphorylate, respectively, the cytoskeletal proteins, which in turn affect the HVACC activity (49,50). Therefore, internalization of HVACCs appears to be an important mechanism responsible for capsaicin-induced profound loss of the HVACC density in DRG neurons. Further studies are warranted to identify the exact substrates dephosphorylated by calcineurin in the DRG neurons.
Another intriguing and important finding is that the basal intracellular Ca 2ϩ and endogenous calcineurin tonically modulate the HVACC currents in DRG neurons. In this regard, intracellular dialysis with BAPTA, deltamethrin, or cyclosporin A all significantly increased the Ca 2ϩ current density and induced a depolarizing shift in the voltage dependence of activation. The observation that BAPTA and calcineurin inhibitors resulted in increased density of Ca 2ϩ currents points to a basal activity of calcineurin in DRG cells in the absence of any prior evoked increases in intracellular Ca 2ϩ . Notably, a similar effect was observed when okadaic acid was applied intracellularly, suggesting that the phosphatases protein phosphatase 1 and protein phosphatase 2A may also be involved in the basal regulation of HVACCs. Therefore, a substantial proportion of HVACCs or closely associated proteins in the DRGs are present in the dephosphorylated form in the DRG. Consistent with our findings, overexpression of calcineurin in NG108-15 cell lines causes a decreased current density of HVACCs. By contrast, the current density of HVACCs is increased in those cells transfected with the calcineurin antisense (16). HVACCs in the dynamic equilibrium of calcineurin-protein kinase activity allow DRG cells to both increase and reduce Ca 2ϩ influx and, by this mechanism, contribute to regulation of neurotransmitter release and nociceptive inputs to spinal dorsal horn neurons.
In summary, this study provides substantial new evidence that calcineurin, constitutively expressed in the cytoplasm of DRG neurons, is a key feedback regulator of intracellular Ca 2ϩ and plays an important role in down-regulation of HVACCs by TRPV1 stimulation. Increased calcineurin activity produced by TRPV1 activation could limit Ca 2ϩ influx through HVACCs in the plasma membrane by down-regulation of HVACCs through dephosphorylating the HVACC or a closely associated cytoskeletal protein. This new information is important for our understanding of the molecular mechanism of the analgesic action and diminished spinal synaptic transmission produced by capsaicin and its related analogs. This study highlights the pivotal role of intracellular Ca 2ϩ level and calcineurin in negative modulation of HVACCs in primary sensory neurons.