Calcium plays a central role in the sensitization of TRPV3 channel to repetitive stimulations

Abstract

Transient receptor potential channels are involved in sensing chemical and physical changes inside and outside of cells. TRPV3 is highly expressed in skin keratinocytes, where it forms a nonselective cation channel activated by hot temperatures in the innocuous and noxious range. The channel has also been implicated in flavor sensation in oral and nasal cavities as well as being a molecular target of some allergens and skin sensitizers. TRPV3 is unique in that its activity is sensitized upon repetitive stimulations. Here we investigated the role of calcium ions in the sensitization of TRPV3 to repetitive stimulations. We show that the sensitization is accompanied by a decrease of Ca(2+)-dependent channel inhibition mediated by calmodulin acting at an N-terminal site (amino acids 108-130) and by an acidic residue (Asp(641)) at the pore loop of TRPV3. These sites also contribute to the voltage dependence of TRPV3. During sensitization, the channel displayed a gradual shift of the voltage dependence to more negative potentials as well as uncoupling from voltage sensing. The initial response to ligand stimulation was increased and sensitization to repetitive stimulations was decreased by increasing the intracellular Ca(2+)-buffering strength, inhibiting calmodulin, or disrupting the calmodulin-binding site. Mutation of Asp(641) to Asn abolished the high affinity extracellular Ca(2+)-mediated inhibition and greatly facilitated the activation of TRPV3. We conclude that Ca(2+) inhibits TRPV3 from both the extracellular and intracellular sides. The inhibition is sequentially reduced, appearing as sensitization to repetitive stimulations.

Figure 1. Effect of intracellular Ca²⁺ buffering strength on the activation of TRPV3 by 2APB

TRPV3-transfected HEK293 cells were held in whole cell mode at 0 mV with pipette solution containing either 10 mM BAPTA (A) or 1 mM EGTA (B). Voltage ramps were applied from -100 to +100 mV in 100 ms after a 20-ms step pulse to -100 mV every 0.5 s. 2APB (100 μM) was applied continuously as indicated by the bars above the current traces, which represent current development at -100 mV (below dashed lines) and +100 mV (above dashed lines). Dashed lines show zero current. Insets show current-voltage (I-V) relationships obtained from the voltage ramp at the time points indicated in the current traces. C and D show summary data (n = 5 for EGTA, n=16 for BAPTA) of current densities at 100 sec after 2APB application (C) and time constants (τ_act) obtained from exponential fits (D) of current developed at -100 and +100 mV using pipettes that contained either 1 mM EGTA (open bars) or 10 mM BAPTA (filled bars). Note, logarithmic scale is used in D to accommodate the large difference between data values. * p < 0.05, ^† p < 0.001, different from EGTA.

Figure 2. Effect of intracellular Ca²⁺ buffering strength on sensitization of TRPV3 to repetitive 2APB stimulations

A and B, similar to Fig. 1 but 2APB was applied repetitively with short intervals. The pipette contained 10 mM BAPTA (A) or 1 mM EGTA (B). C, two consecutive 2APB stimulations were spaced with a 10-min interval when the cell was held continuously at 0 mV. The pipette contained 1 mM EGTA. D, peak current densities developed at -100 (open symbols) and +100 mV (filled symbols) for repetitive 2APB stimulations with pipette solutions that contained 10 mM BAPTA (circles), 10 mM EGTA (squares) and 1 mM EGTA (triangles). n = 6-9. E, simulation of Ca²⁺ diffusion from the source in solutions that contain 1 and 10 mM EGTA or 10 mM BAPTA. The diffusion coefficient for Ca²⁺ (0.22 μm²/msec), dissociation constants for Ca²⁺/EGTA (1.8 × 10^-7 fmole/μm³) and Ca²⁺/BAPTA (2.2 × 10^-7 fmole/μm³), on rates for Ca²⁺ binding to EGTA (2.5 × 10³ μm³/fmole.msec) and to BAPTA (4 × 10⁵ μm³/fmole.msec) were adopted from Naraghi and Neher (26). The initial [Ca²⁺] at the source of generation, e.g. inner mouth of the pore, is arbitrarily assigned to 1 mM. F, fast deactivation detected during the 20-ms step pulse from 0 to -100 mV in the EGTA (gray) and BAPTA (black) buffered internal solutions. Representative traces at the peak of 2APB-evoked response from two fully sensitized cells were scaled to show the difference in the rate of deactivation.

Figure 3. Changes in the voltage dependence of TRPV3 during repetitive stimulations

A, a TRPV3-expressing cells was voltage-clamped using the 1 mM EGTA internal solution and repetitively stimulated with 100 μM 2APB. Currents were recorded by voltage ramps as in Fig. 1. A step protocol was applied intermittently as indicated. Top traces show currents at -100 mV (open circles) and +100 mV (filled circles) obtained from voltage ramps. The up and down deflections from the dashed line (zero current) represent the response to a brief (∼12 s) stimulation of 2APB. Lower traces are currents obtained from the step protocol with those from -100, 0, 100, 200 mV highlighted in black. B, a similar experiment as in A, but the pipette solution contained 10 mM BAPTA. C and D, tail currents from the step protocol were normalized and plotted as a function of the prepulse potentials, showing conductance-voltage (G-V) relationships. In the first and 10^th stimulations, some currents did not reach steady state during step pulses (e.g. see representative traces in A). For them, steady-state currents were projected by fitting the entire current trace with an exponential function and then used to determine the conductance at each pulse potential. The conductance values were then normalized. Note, a more pronounced left-shift is evident with 1 mM EGTA (upper) than with 10 mM BAPTA (lower) in the internal solution (C). The continued lines represent fits to the Boltzmann equation, which give rise to V_1/2, z (the valence of the gating charge), and voltage-dependent fraction (VDF) of the total activity. These values are plotted as a function of the stimulation number (D). n = 3-12 cells. * p < 0.05, ** p < 0.01, *** p < 0.001 different from BAPTA; ^# p < 0.05, ^## p < 0.01, ^### p < 0.01 different from the first stimulation.

Figure 4. The intracellular Ca²⁺ effect is mediated by CaM

A, similar to Fig. 2A, TRPV3-expressing cells were repetitively stimulated with 2APB but the pipette solution contained 10 mM BAPTA with free Ca²⁺ buffered to ∼1.6 μM without (A) or with a CaM antagonist ophiobolin A (20 μM) (B). C, summary of peak current densities obtained at -100 and +100 mV for individual responses with (circles, n=11) or without ophiobolin A (Oph, triangles, n=5).

Figure 5. Identification of a CaM-binding site from the N-terminus of TRPV3

A, diagram of murine TRPV3 and maltose-binding protein (MBP)-TRPV3 fusion proteins tested. Fragments of TRPV3 were prepared as MBP-fusion proteins and tested for binding to CaM in the presence of 50 μM Ca²⁺. The fusion proteins were designated as a, b, c, etc. and their positions in the full-length TRPV3 are indicated in parentheses. Black, gray, and open bars represent positive, weakly positive, and negative binding to CaM, respectively. The borders of the identified N-terminal CaM-binding site are shown by dashed lines. Gray bars in the full-length TRPV3 indicate transmembrane segments and the pore loop. B, representative binding results showing the sizes and the amounts of ³⁵S-labeled MBP-TRPV3 fusion proteins added to the binding reactions (upper graph) and the amounts of the fusion proteins retained by CaM-agarose in the presence of 50 μM Ca²⁺ (lower graph). MBP was included alone as a negative control (lane 1). C, R¹¹³QKKKRLKKR¹²² was mutated to SQAEASDAEG as described in Materials and Methods. The mutated l fragment was fused to MBP and tested for binding to CaM under the same condition as in B. The graphs show that the mutant (RK^-) does not bind to CaM.

Figure 6. Voltage dependence of TRPV3RK^- mutant in response to repetitive 2APB stimulations

A, a representative experiment performed as in Fig. 3A except that the cell expressed TRPV3RK^-. B and C, G-V relationships for TRPV3RK^- were obtained from tail currents as in Figs. 3C and 3D. No significant change was found between the EGTA and BAPTA internal solutions except for some small changes in VDF. * p < 0.05 different from BAPTA; ^# p < 0.05 different from the first stimulation. D and E, peak current densities (D) and τ_act (E) at -100 and +100 mV elicited by the first (left panels) and 23^rd (right panels) applications of 100 μM 2APB to HEK293 cells that expressed TRPV3RK^- (RK^-). Data obtained from wild type TRPV3 (WT) for the 23^rd stimulation are included for comparison. The internal solution contained either 1 mM EGTA (open bars) or 10 mM BAPTA (filled bars). Currents were recorded using the voltage ramp protocol. τ_act was obtained by fitting the data points from 0.5 to 12 sec (or the time at the peak current) following 2APB application. n = 5-11 cells. * p < 0.01 different from WT/EGTA, ^† p < 0.05 different from WT/BAPTA.

Figure 7. Fast inactivation of the wild-type TRPV3 and the RK^- mutant

Data were collected from the same experiments shown in Figs. 3 and 6. A, tail currents at -100 mV after pre-steps to +100 (left panels) and +200 mV (right panels) were scaled to the peak current for cells infused with 1 mM EGTA (gray lines) or 10 mM BAPTA (black lines). Shown are representative results for the wild-type (WT, upper) and the mutant (RK^-, lower). B, deactivation kinetics were fitted with the equation: I_t = y₀ + a*(1-exp(-t/τ_inact)), where y₀ is the peak current, a is the maximal deactivation and τ_inact is the deactivation time constant. Shown are summary data of τ_inact at -100 mV following pre-steps to +100 (left) and +200 mV (right) during repetitive stimulations. Open and filled symbols are data for the EGTA and BAPTA internal solutions, respectively. No significant difference is found between WT (circles) and RK^- (triangles).

Figure 8. Activation of TRPV3 in Ca²⁺-free bath solution

A, removal of extracellular Ca²⁺ strongly increased the 2APB-evoked currents. The pipette solution contained 10 mM BAPTA. A TRPV3 expressing cell was stimulated with 2APB in the normal bath (2 mM Ca²⁺) while the solution was switched to a Ca²⁺-free bath containing 0.1 mM EGTA and the same concentration of 2APB. Note the sudden increase of the current size at both the positive (+100 mV) and negative (-100 mV) potentials and linear I-V (right plot) upon Ca²⁺ removal and the quick reinhibition when extracellular Ca²⁺ was reintroduced. B and C, response of TRPV3 to repetitive 2APB applications in the Ca²⁺-free bath. The pipette solution contained either 10 mM BAPTA (B) or 1 mM EGTA (C). D and E, summary data for peak current densities at +100 (filled symbols) and -100 mV (open symbols) for repetitive 2APB stimulations with pipette solutions that contained 10 mM BAPTA (circles), 1 mM EGTA (triangles), or 10 mM BAPTA with free Ca²⁺ buffered at 1.6 μM (E) supplemented (squares) or not (diamonds) with 20 μM ophiobolin A. F, 2APB-evoked [Ca²⁺]_i changes in HEK293 cells that expressed TRPV3 (black) or the vector control (gray). Cells were loaded with fluo4 20 hrs post transfection, washed with either normal or the Ca²⁺-free bath, and then stimulated with 166 μM 2APB in the corresponding solutions. Note the significant fluorescence increase in the Ca²⁺-free bath and the lack of response in the control.

Figure 9. Inhibitory effect of extracellular Ca²⁺ on TRPV3 channel activity

A, changes in extracellular Ca²⁺-dependent inhibition during repetitive stimulations. Upper panel shows changes in Ca²⁺ concentrations in perfusates during six successive applications of 100 μM 2APB as indicated. Traces below show currents at -100 (open circles) and +100 mV (filled circles) recorded using voltage ramps. The internal solution contained 1 mM BAPTA (see Table 1). B, I-V curves obtained at indicated external Ca²⁺ concentrations during the first, third and sixth 2APB applications for the same representative cell shown in A. C, summary of dose-dependent inhibition of TRPV3 current by external Ca²⁺ during six consecutive stimulations. Currents at +100 mV (upper) and -100 mV (lower) are normalized to that obtained in the Ca²⁺-free solution for each stimulation. n = 4-17 cells. Solid lines are fits to a two-site formula as described in Experimental Procedures. A right shift in the dose response curves in response to successive stimulations is detected at both potentials. D, the IC₅₀ value of the high affinity state is gradually increased due to successive stimulations. E, the fraction of the high affinity state is decreased. Note: the IC₅₀ values of the low affinity state have big errors because of the limited data points and large measurement errors at higher Ca²⁺ concentrations resulting from the very low current amplitudes. The mean values ranged from 0.6 to 2.1 mM, but no clear trend is detected with successive stimulations.

Figure 10. Effect of extracellular Ca²⁺ on the activity of TRPV3D641N mutant

Asp⁶⁴¹ at the pore loop of murine TRPV3 was changed to Asn and the mutant was expressed in HEK293 cells for whole-cell experiments. Similar to Fig. 9 except that D641N was used and only the first stimulation is shown. For many cells, the seal did not last for more than one stimulation and for others, the second and later responses displayed pronounced desensitization. A, representative result. The inset to the left shows I-V curves at selected Ca²⁺ concentrations. Note: low concentrations of Ca²⁺ did not inhibit and rectification of I-V only appeared at >2 mM Ca²⁺. C, summary data show that the dose-response curves only have one (the lower) affinity state.

Figure 11. Effect of intracellular Ca²⁺-CaM binding on the activation of D641N mutant

A and B, effect of intracellular Ca²⁺-buffering strength on sensitization of D641N. Internal solution contained either 10 mM BAPTA (A) or 1 mM EGTA (B). Similar to the wild type (Figs. 2A and 2B), voltage ramps revealed that currents developed more slowly and sensitization was more prominent with EGTA than with BAPTA internal solution. C, comparison of activation kinetics in response to the first 2APB (100 μM) application between BAPTA (black circles) and EGTA (gray triangles) buffered cells that expressed D641N. Currents at -100 (open symbols) and +100 mV (filled symbols) obtained from the voltage ramp protocol are scaled to highlight the kinetic difference. In the EGTA-buffered cell, 2APB was applied longer as indicated by the bar on the top to allow currents to peak. Note desensitization before 2APB washout in the BAPTA-buffered cell. D-G, agonist-evoked activation of D641N-RK^- double mutant. D, similar to C except the double mutant was used. E and F, current densities (E) and activation time constants (F) at -100 and +100 mV elicited by the first 2APB application to cells that expressed D641N (open bars) or the double mutant (filled bars). Voltage ramps were used. The internal solution contained either 1 mM EGTA or 10 mM BAPTA as indicated. The current density represents the maximum within 12 sec of the stimulation and τ_act was obtained by fitting the data points from 0.5 to 12 sec (or the time at the peak current) following 2APB application. n = 9-13 cells. * p < 0.05, ** p < 0.001. G, the step protocol was used to compare the voltage dependence of the double mutant in EGTA (open symbols) and BAPTA (filled symbols) internal solutions. Shown are representative traces (left) and the instantaneous (I₁, left panel) and steady-state (I₃₀₀, left panel) I-V, which are currents at the beginning (1 ms) and the end (300 ms) of the voltage steps, as well as the G-V relationship (right).