Importance of a conserved sequence motif in transmembrane segment S3 for the gating of human TRPM8 and TRPM2

Abstract

For mammalian TRPM8, the amino acid residues asparagine-799 and aspartate-802 are essential for the stimulation of the channel by the synthetic agonist icilin. Both residues belong to the short sequence motif N-x-x-D within the transmembrane segment S3 highly conserved in the entire superfamily of voltage-dependent cation channels, among them TRPM8. Moreover, they are also conserved in the closely related TRPM2 channel, which is essentially voltage-independent. To analyze the differential roles of the motif for the voltage-dependent and voltage-independent gating, we performed reciprocal replacements of the asparagine and aspartate within the S3 motif in both channels, following the proposed idea that specific electrostatic interactions with other domains take place during gating. Wild-type and mutant channels were heterologeously expressed in HEK-293 cells and channel function was analyzed by whole-cell patch-clamp analysis as well as by Ca(2+)-imaging. Additionally, the expression of the channels in the plasma membrane was tested by Western blot analysis, in part after biotinylation. For the mutations of TRPM8, responses to menthol were only compromised if also the expression of the glycosylated channel isoform was prevented. In contrast, responses to cold were consistently and significantly attenuated but not completely abolished. For TRPM2, surface expression was not significantly affected by any of the mutations but channel function was only retained in one variant. Remarkably, this was the variant of which the corresponding mutation in TRPM8 exerted the most negative effects both on channel function and expression. Furthermore, we performed an exchange of the inner pair of residues of the N-x-x-D motif between the two channels, which proved deleterious for the functional expression of TRPM8 but ineffective on TRPM2. In conclusion, the N-x-x-D motif plays specific roles in TRPM8 and TRPM2, reflecting different requirements for voltage-dependent and voltage-independent channel gating.

Figure 1. Conservation of a short sequence motif within transmembrane segment S3.

Sequence comparison within the S3 region of selected voltage-gated as well as voltage-independent cation channels, including TRPM8 and TRPM2. The amino acid sequences are shown in single letter code. The highly conserved N-x-x-D motif is highlighted with the outer pair of amino acid residues labeled in red and the inner pair in orange. Further highly conserved amino acid residues upstream of the N-x-x-D-motif are given in bold letters. The glycine residue at position 805 in the sequence of human TRPM8 which is crucial for the icilin sensitivity of the channel is marked in blue. Accession numbers are as follows human TRPM8: Q7Z2W7; human TRPM2: O94759; human TRPA1: O75762; human TRPC3: Q13507, Shaker H4 (KCNAS_DROME): P08510; human sodium channel type 2 alpha subunit (SCN2A), domain 4: Q99250; voltage-gated sodium channel from Bacillus halodurans (NaChBac): Q9KCR8; human L-type calcium channel subunit alpha 1C (CACNA1C) domain 4: Q13936.

Figure 2. Current responses of TRPM8 variants to menthol and cold.

The variants were wild-type, D802N, N799D, and N799D+D802N. (A), Current densities (representing mean ± S.E.M of 26–31 independent experiments) obtained at room temperature during voltage ramps between −150 mV and +150mV applied over 200 ms. Note that only the voltage range from 0 to +150 mV is shown because currents are sizeable exclusively in the outward direction (B), Whole-cell patch clamp measurement on HEK-293 cells expressing wild-type TRPM8 during stimulation with menthol (100 µM) or ice-cold bath solution as indicated by the horizontal bars. Between the two stimulations an intermediate wash-step with standard bath solution at room temperature was performed. The holding potential was −60 mV. (C), Similar experiment as shown in panel B on cells expressing the TRPM8 variant N799D. Note the different scaling of the ordinates in panels B and C. The inset shows the corresponding current-voltage relation of N799D in comparison to wild-type in the presence of menthol. (D), Mean inward current densities of the TRPM8 variants in response to menthol or cold obtained at a holding potential of −60 mV. Note that in double stimulation experiments only the data from the first stimulation were used for the statistical analysis. Each column represents mean ± S.E.M. of 10–16 independent experiments. Values of ***p<0.001 were considered extremely significant.

Figure 3. Differential effects of the mutation M8-N799D+D802N on the responses to cold and menthol.

(A) Consecutive stimulation of a cell expressing M8-N799D+D802N, initially with cold and then with menthol (100 µM). Cells were exposed to room temperature prior to the exposure to menthol. The inset shows the summarized data with each column representing mean ± S.E.M. of 6 independent experiments. The difference between the values was highly significant (***p<0.001). (B) Consecutive stimulation with the two stimuli in the reversed order. The response to cold was even weaker than in panel A, due to desensitization that also weakened a further reaction to menthol. Note that for each experiment only the data from the first stimulation were used for the statistical analysis.

Figure 4. Activity of TRPM8 variants in intracellular Ca²⁺ measurements and Western blot of plasma membrane preparations.

(A), Maximum increase in the F₃₄₀/F₃₈₀ ratio of each variant in response to 300 µM menthol or ice-cold bath solution. In double stimulation experiments only the data from the first stimulation were used for statistical analysis. The miniscule increases in fluorescence observed in mock-transfected controls are subtracted. Each column represents mean ± S.E.M. of 8–17 independent experiments. Significant differences to control are indicated with asterisks. Values of ***p<0.001 were considered extremely significant. (B), Western blot of TRPM8 protein on plasma membrane fractions prepared by the differential centrifugation method (“low speed fraction” ref. to Material and Methods) of HEK-293 cells expressing various TRPM8 variants. Note the dual bands, shown in previous studies to indicate the glycosylated and non-glycosylated channel protein. The glycosylated form is absent in the weakly functional channel variant N799D. As negative control a plasma membrane fraction of mock-transfected HEK-293cells is included.

Figure 5. Current responses of the closely related TRPM2 channel and its variant N869D.

Whole-cell patch clamp measurements on HEK-293 cells expressing wild-type TRPM2 (A) or TRPM2-variant N869D (B). Stimulation was performed with ADPR (600 µM in the absence of intracellular Ca²⁺) infused into the cell through the patch pipette. The holding potential was −60 mV. Insets show the current-voltage relations obtained during voltage ramps within the indicated voltage ranges. Inward currents were repeatedly blocked with NMDG. The TRPM2 variant N869D is analogous to the virtually non-functional TRPM8 variant N799D.

Figure 6. Cell surface expression of TRPM2 variants.

(A) Western blots were performed on two different membrane fractions of HEK-293 cells transfected with a TRPM2 variant and probed with a commercially available anti-TRPM2 antibody. Mock-transfected cells were used as negative control. Membrane fractions were obtained either with high or low speed centrifugation (20,000 g vs. 100,000 g) of whole-cell lysates. (B) Alternatively, cell surface expression of TRPM2 channel variants as well as mock-control was monitored by biotinylation assays. Eluted samples were immunoblotted and detected by the same anti-TRPM2 antibody as used in A (upper right panel). Total TRPM2 expression was assessed by loading unpurified cell lysate (upper left panel). Membrane integrity and exclusive biotin-labeling of cell surface proteins was verified by re-incubation of the blot with an anti-β-actin antibody and loss of β-actin signal in the NeurAvidin-bound fraction (lower panel). Two independent experiments gave similar results.

Figure 7. Importance of the inner residues within the N-x-x-D motif for the activity of TRPM8 and TRPM2.

Whole-cell patch clamp measurements on HEK-293 cells expressing variants of TRPM2 and TRPM8 where the inner pair of residues of the N-x-x-D motif is reciprocally exchanged. (A), TRPM8 variant V800K+M801L first stimulated with ice-cold bath solution and after an intermediate wash-step with standard bath solution stimulated with 100 µM menthol (as indicated by horizontal bars). (B), TRPM2 variant K870V+L871M stimulated by infusion of 600 µM ADPR into the cell through the patch pipette and in the absence of intracellular Ca²⁺. Insets show the corresponding current-voltage relations during stimulation.