Differential role of the menthol-binding residue Y745 in the antagonism of thermally gated TRPM8 channels

Abstract

Background: TRPM8 is a non-selective cation channel that belongs to the melastatin subfamily of the transient receptor potential (TRP) ion channels. TRPM8 is activated by voltage, cold and cooling compounds such as menthol. Despite its essential role for cold temperature sensing in mammals, the pharmacology of TRPM8 is still in its infancy. Recently, tyrosine 745 (Y745) was identified as a critical residue for menthol sensitivity of the channel. In this report, we study the effect of mutating this residue on the action of several known TRPM8 antagonists: BCTC, capsazepine, SKF96365, and clotrimazole as well as two new inhibitor candidates, econazole and imidazole.

Results: We show that Y745 at the menthol binding site is critical for inhibition mediated by SKF96365 of cold- and voltage-activated TRPM8 currents. In contrast, the inhibition by other antagonists was unaffected by the mutation (BCTC) or only partially reduced (capsazepine, clotrimazole, econazole), suggesting that additional binding sites exist on the TRPM8 channel from where the inhibitors exert their negative modulation. Indeed, a molecular docking model implies that menthol and SKF96365 interact readily with Y745, while BCTC is unable to bind to this residue.

Conclusion: In summary, we identify structural elements on the TRPM8 channel that are critical for the action of channel antagonists, providing valuable information for the future design of new, specific modulator compounds.

Figure 1

The TRPM8-Y745H mutant is insensitive to menthol, but retains cold sensitivity. A, Western blot where the lanes represent lysates of HEK293 cells transfected with TRPM8-wt and TRPM8-Y745H. B, Representative traces showing calcium imaging experiments of HEK293 cells expressing TRPM8-wt or TRPM8-Y745H. Note that only the cells expressing TRPM8-wt respond to menthol. C, Summary histogram of experiments seen in B. Intracellular calcium increases were compared using repeated-measures 2-way ANOVA in combination with Bonferroni's post test with respect to the effect of the mutation on each condition: *** p < 0.001; n = 14 (TRPM8-wt) and n = 19 (TRPM8-Y745H).

Figure 2

Electrophysiological characterization of TRPM8-Y745H mutant sensitivity to menthol, cold and voltage. A, Whole-cell current-voltage relationships of TRPM8-wt and TRPM8-Y745H expressing HEK293 cells in control and menthol-containing solutions at 33°C. Note the different current scale. B, Summary histogram of experiments seen in A showing menthol-induced whole-cell currents at various potentials, normalized with the current in control conditions. Note the logarithmic current scale. The responses of TRPM8-wt vs. Y745H were compared with repeated-measures 2-way ANOVA in combination with Bonferroni's post test with respect to the effect of the mutation at each potential: *** p < 0.001; * p < 0.05; n = 7 (TRPM8-wt) and n = 13 (TRPM8-Y745H). C-E, Parameters obtained from fits of I-V data of experiments such as the ones showed in A to equation (i), n = 4-19. C, Average values of the apparent gating charge in control (33°C), menthol (100 μM), cooling (20°C) and cooling + menthol, of the TRPM8-wt vs. TRPM8-Y745H. D, Average maximum conductance (g) of cells expressing TRPM8-wt and TRPM8-Y745H during cooling and menthol application. The data are normalized with the value of g of the same cells in control solution at 33°C. E, Agonist-induced shifts of the midpoint of voltage activation (V_1/2) of TRPM8-wt vs. TRPM8-Y745H. The data are represented with respect to the value of V_1/2 in control solution at 33°C. In panels C-E, statistical significance was assessed with Student's unpaired t-test: *** p < 0.001; ** p < 0.01; * p < 0.05.

Figure 3

Structures of the various TRPM8 antagonists and menthol. Chemical structures of the various antagonists tested at the TRPM8-wt and Y745H mutant channels. The structure of menthol is shown for comparison.

Figure 4

Differential effect of the Y745H mutation on the antagonism of TRPM8 by BCTC and SKF96365. A-B, Cold-evoked [Ca²⁺]_i responses in HEK293 cells expressing TRPM8-wt or TRPM8-Y745H channels, showing the inhibition by A, 3 μM BCTC, and B, 3 μM SKF96365. C-D, Summary histograms of the [Ca²⁺]_i responses of TRPM8-wt and TRPM8-Y745H channels to repeated cooling stimuli in the absence and presence of C, 3 μM BCTC (n = 9/11 wt/mut); and D, 3 μM SKF96365 (SKF; n = 33/47). Note the reversible nature of the inhibition. E, Comparison of block of TRPM8-wt vs. TRPM8-Y745H by BCTC and SKF96365. In panels C-E, intracellular calcium increases were normalized to the first cold application in control solution. In panel E, the block of each antagonist condition was compared between TRPM8-wt and -Y745H using Student's unpaired t-test: *** p < 0.001; ns. = not significant. The block by 20 μM SKF96365 over the Y745H mutant is included to show the complete lack of inhibition.

Figure 5

Variable inhibition of the Y745H mutant channel by capsazepine, clotrimazole, econazole and imidazole. A-D, Dose-inhibition curves of various antagonists at the cooling-activated TRPM8-wt and TRPM8-Y745H channels: A, capsazepine (n = 17-23/13-29 wt/mut); B, clotrimazole (n = 7-17/12-25); C, econazole (n = 14-27/9-27); and D, imidazole (n = 12-21/13-16). E, Comparison of the IC₅₀ of block of TRPM8-wt vs. TRPM8-Y745H by the above antagonists: capsazepine (CPZ); clotrimazole (CLOT); econazole (ECO). In panels A-D, the red traces represent the fits to the Hill equation. Error bars were used as weights in fitting. At each concentration, the block was compared between TRPM8-wt and -Y745H using Student's unpaired t-test, and indicated where significant by: *** p < 0.001; ** p < 0.01; * p < 0.05.

Figure 6

Electrophysiology confirms the differential effects of the Y745H mutation on BCTC and SKF96365 antagonism. Whole-cell I-V curves from voltage ramps (-100/+150 mV) of TRPM8-wt and TRPM8-Y745H expressing HEK293 cells during cooling in the presence and absence of A, 0.6 μM BCTC; and B, 3 μM SKF96365. Wash traces recorded 3 minutes after removal of the antagonist from the bath are included to show the reversible nature of the inhibition. C, Summary histogram of experiments depicted in A and B, showing the block of cold-evoked currents by 0.6 μM BCTC and 3 μM SKF96365 at a membrane potential of +80 mV. D-E, Parameters obtained from fits of I-V data to equation (i). D, Antagonist-induced change in maximum conductance during cooling in cells expressing TRPM8-wt and TRPM8-Y745H. The data are normalized to the value during cooling in control solution of the same cells, (g_cold+blocker/g_cold). E, Antagonist-induced shifts of the midpoint of voltage activation (V_1/2) of TRPM8-wt vs. TRPM8-Y745H during cooling. The data are represented with respect to the value of V_1/2 in the absence of blocker (ΔV_1/2 = V_{1/2, cold+blocker} - V_{1/2, cold}). In panels C-E, statistical significance was assessed with Student's unpaired t-test, n = 2-6.

Figure 7

Differential block of voltage-activated TRPM8-Y745H by BCTC and SKF96365. Whole-cell I-V curves from voltage ramps (-100/+150 mV) of TRPM8-wt and TRPM8-Y745H expressing HEK293 cells at 33°C in the presence and absence of A, 3 μM BCTC; and B, 3 μM SKF96365. C, Summary histogram of experiments seen in A and B, showing the block of voltage-evoked currents by BCTC and SKF96365 at a membrane potential of +120 mV. Statistical significance was assessed with the unpaired t-test: ** p < 0.01, n = 2-5.

Figure 8

Molecular modelling of ligand docking into the putative menthol binding site of the TRPM8 receptor. Parts of the S1-S4 domains of TRPM8 are depicted and residues Y745 in S2 and N799 and D802 in S3 are highlighted. A, Menthol makes a hydrogen bond with Y745 (S2), disrupting its interaction with D802 (S3). B, BCTC is located externally of Y745 preventing the interaction with this residue. C, SKF96365 interacts with Y745 and N799, stabilizing the positions of the S2 and S3 domains. The TRPM8 model was built using the coordinates of Pedretti et al. [38].