Proton-selective conductance and gating of the lysosomal cation channel TMEM175

Abstract

The lysosomal cation channel TMEM175 plays a key role in luminal pH homeostasis and lysosome function, with aberrant activity linked to Parkinson's disease. Although initially described as a K⁺-selective channel, TMEM175 exhibits substantial H⁺ permeability. Here, we dissect complex changes affecting human TMEM175 conductance and ionic properties of TMEM175-mediated current in response to pH shifts on the luminal side of the protein. A drop in pH from 7.4 to 4.7 on the side equivalent to the lysosomal lumen triggers a sustained increase in TMEM175-mediated inward and outward currents, which is accompanied by a transient shift in the reversal potential (E_rev) toward the theoretical equilibrium voltage for H⁺, yet remaining ~100 mV below the expected value even in the absence of K⁺. This discrepancy, along with low sensitivity of E_rev to the concentration gradient for K⁺, supports a model in which TMEM175-mediated H⁺ flux rapidly collapses the lysosomal pH-gradient. Molecular dynamics simulations identify H57 as a key residue on the luminal side of the open channel, which forms intra- and intersubunit salt bridges with D279 and E282. Supporting the functional importance of these interactions, the TMEM175 mutant H57Y displayed reduced H⁺- and K⁺-conductance and a reduced H⁺/K⁺ selectivity in whole-cell and lysosomal electrophysiological analyses. Our findings contribute to a better understanding of TMEM175's complex electrophysiological properties, thereby expanding the possibilities of understanding the channel's function in lysosomal physiology and pathophysiology.

Keywords: MD simulations; SSME; TMEM175; patch-clamp; proton channel.

Fig. 1.

Acidification of the luminal side induces a continuous increase in conductance but only a transient shift of E_rev to positive values in TMEM175-expressing cells. (A and B) Whole-cell recordings of TMEM175-expressing (A) and GFP-expressing (empty vector transfected) control cells (Ctrl) (B). Bath and pipette solution contained 140 mM K-MS. The internal pH (pH_in) was 7.4. The external pH (pH_ex) was changed from 7.4 to 4.7. (i) Representative current responses to voltage-ramps from +120 to −120 mV shortly before (a, black), shortly after (b, pink), and 3 min after (c, purple) a pH_ex-jump from 7.4 to 4.7 (see also panels (ii) for time indices a–c). (ii) Corresponding time-courses of reversal voltage (E_rev) as well as current densities (upper graph) and chord conductance (lower graph) at ±120 mV of TMEM175-expressing and control cells. Values from voltage-ramp recordings as in (i). (iii) Individual (black) and averaged time-courses (red) of E_rev from pH_ex jump experiments as shown in (ii). (C) Current densities at ±120 mV at time points a, b, and c in (A) and (B). (D) Maximal change in E_rev and free running membrane voltage V_mem in response to pH_ex-jump from 7.4 to 4.7 of control (gray) and TMEM175-expressing (blue) cells. (E) pH_ex-jump-induced change in chord conductance at ±120 mV for control (gray) and TMEM175-expressing (blue) cells. Bars in (C–E) represent arithmetic mean ± SD; values from individual recordings are shown as closed circles. Statistical analyses in (D) and (E) were performed with unpaired two-way ANOVA with Sidák’s and Tukey’s multiple comparison test, respectively.

Fig. 2.

The TMEM175-mediated H⁺-current causes a time-dependent erosion of the pH-gradient in whole-cell patch-clamp recordings. (A) Averaged time-courses ± SEM of E_rev of TMEM175-expressing cells in symmetrical 140 mM K⁺ concentrations, in response to a pH_ex jump from 7.4 to 4.7 under different pH-buffering conditions. The internal and external solutions were buffered to pH 7.4 with 5 mM (light red) or 50 mM HEPES (dark red and blue). The external solution was buffered to pH 4.7 with 5 mM acetate (light red), 50 mM acetate (dark red), or 50 mM citrate (blue). (B) Maximal change in E_rev in response to pH_ex jumps from 7.4 to 4.7 in control (ctrl, gray) and TMEM175-expressing cells with different pH 4.7 buffer conditions as shown in panel A. Statistical comparisons were done by unpaired two-way ANOVA with Tukey’s multiple comparison test. (C) Time-courses of E_rev and corresponding chord conductance at −120 mV (G_{−120 mV}) of three TMEM175-expressing cells in response to pH_ex jumps from 7.4 (50 mM HEPES) to 4.7 (50 mM citrate). (D) Relationship between G_{−120 mV} (90 s after external acidification) and the reciprocal of the time constant (τ) of the E_rev backshift. The time constant was obtained by fitting the falling phase (backshift) of E_rev with a single exponential function. The straight line represents the best fit with a linear function. R² and the Pearson correlation coefficient (r) are displayed in the graph. (E) Relationship between time constant of the E_rev backshift and the maximal E_rev value (Peak) measured upon acidification. The straight line represents the best fit with the Hill equation. Closed black circles highlight data points with τ > 40 s. These data points are further used for calculation of P_H+/P_K+ in Fig. 4. (F) Voltage pulse protocol (upper graph) and corresponding current response of a representative TMEM175-expressing cell, 90 s after acidification of the external solution. Internal and external solutions contained 140 mM K⁺ and were buffered with 50 mM HEPES to pH 7.4 and 50 mM citrate to pH 4.7, respectively. (G) Magnification of ramp currents shown in panel (F), plotted against the corresponding membrane voltage. Colors in (F) and (G) correspond to different holding potentials of voltage pulses shown in panel (F). (H) Arithmetic mean ± SD of E_rev values from (n) independent experiments as shown in (G), plotted as a function of V_hold.

Fig. 3.

Histidine 57 is involved in sensing luminal acidification and gating of TMEM175. (A) Representative current responses of a TMEM175-expressing cell to voltage-ramps applied at an interval of 1 s immediately after a pH_ex jump from 7.4 (50 mM HEPES) to 4.7 (50 mM citrate), in symmetrical 140 mM K⁺ concentrations. The pH of the internal solution was 7.4. (B) Conductance at −120 mV calculated from ramps shown in (A), plotted against the corresponding change in E_rev (relative to the last voltage-ramp before the pH_ex jump). (C) Similar data as shown in (B), derived from five TMEM175-expressing cells. To enhance comparability, ΔE_rev values were plotted against normalized changes in conductance. The G_{−120 mV} value measured in response to the voltage-ramp before the pH_ex-jump was therefore set to zero and the ΔG_{−120 mV} value measured directly before the onset of the E_rev shift to one. (D–G) Molecular dynamics (MD) simulations of wildtype (wt) TMEM175. (D) Snapshots of TMEM175 in the open and closed state with H57 either protonated or deprotonated. Intramonomer salt-bridges between H57 and D279 as well as the intermonomer interactions between H57 and E282 are displayed as green lines. (E and F) Aggregated distances between H57-D279 (E) and H57-E282 (F) during simulations in these states are depicted as histograms, with reference salt-bridge distances indicated by dotted (4.0 Å) and dashed lines (3.2 Å). Each histogram is normalized such that its total area equals 1. (G) Heatmap summary of the luminal salt-bridge network as share of MD frames below a distance threshold of 3.2 Å for the corresponding channel-state group (shown in F), for all possible H57-D279/E282 distance pair combinations in both TMEM175 chains. Minimum distances between each of the two imidazole-nitrogen atoms in H57 and carboxy-oxygens in D279/E282 are drawn from four replica MD simulations for the start structure in the respective state, with a total simulation time of 3.5 µs per state. Prior to the analysis of aggregated distances, datapoints from the first 50 ns of each MD trajectory were discarded. All underlying distance timeseries for full-length trajectories are depicted in SI Appendix, Figs. S6 and S7. (H–J) SSME recordings on lysosomes isolated from TMEM175 wt- and H57Y-expressing HEK293 cells. (H) Schematic representation of an SSME experiment. (I) pH dependence of K⁺ flux assessed using 50 mM [K⁺]-jumps at varying pH. Data for TMEM175 wt were adapted from ref. . Data points represent arithmetic means ± SD from n = 5 (wt) and n = 8 (H57Y) independent sensors. (J) H⁺-permeabilities (P_H+) for TMEM175 wt and TMEM175 H57Y from pH-jump experiments in SI Appendix, Fig. S8B in the absence of K⁺. (K) K⁺-permeabilities for TMEM175 wt and H57Y from K⁺ flux experiments in SI Appendix, Fig. S8A. External and intralysosomal pH was 7.6. P_K+ values were calculated from K⁺ flux data between [K⁺] = 32 and 80 mM.

Fig. 4.

Substitution of H57 with the nonprotonatable amino acid tyrosine alters conductance and H⁺/K⁺ selectivity of TMEM175 at low pH_ex. (A and B) Time-courses of E_rev (upper graph) and chord conductance at ±120 mV (lower graph) calculated from current responses to voltage-ramps from representative whole-cell patch-clamp experiments on cells expressing TMEM175 wt (A) and TMEM175 H57Y (B). Pipette solution contained 140 mM K⁺ and was buffered with 50 mM HEPES to pH 7.4. [K⁺]_ex and pH_ex were changed during the experiments as indicated above the traces. pH_ex was buffered with 50 mM HEPES (pH_ex = 7.4) or 50 mM citrate (pH_ex = 4.7). For reducing [K⁺]_ex from 140 to 14 mM potassium was replaced by equimolar amount of NMDG⁺. (C) Chord conductance at −120 mV of GFP (ctrl), TMEM175 wt-, and TMEM175 H57Y-expressing HEK293 cells calculated from recordings as in (A) and (B) with [K⁺]_ex = 140 mM and pH_ex = 7.4 (time point a) or pH_ex = 4.7 (time point c). (D) Peak E_rev measured after pH_ex-jump from 7.4 to 4.7 in [K⁺]_ex = 140 mM. For TMEM175 H57Y peak E_rev = steady-state E_rev (time point c). (E) ∆E_rev obtained between time points c and d in (A) and (B) following reduction of [K⁺]_ex from 140 mM to 14 mM. Values were only taken from recordings in which E_rev had reached a stable ΔpH at time point c. (F) Ratios of H⁺- and K⁺-permeabilities (P_H+/P_K+) at pH_ex = 4.7 calculated from E_rev and ΔE_rev values in (D) and (E) using Eqs. 1 and 2, respectively. (G) Ratios of H⁺ and K⁺ permeabilities of TMEM175 wt at pH_ex = 7.4 [calculated from E_rev values at time point f in (A)], pH_ex = 6.1 (calculated from ΔE_rev between time points b and a in SI Appendix, Fig. S4C) and pH_ex = 4.7 [pooled values for TMEM175 wt from (F)]. Bars in (C), (F), and (G) represent geometric mean ± geometric SD. Statistical comparisons within different groups in (C) (separated by dotted lines) were performed using a ratio paired t test. Statistical comparisons between different groups in (C), (F), and (G) were performed using a lognormal ordinary one-way ANOVA with Šídák’s multiple comparisons test. Bars in (D) and (E) represent arithmetic mean ± SD. Statistical comparisons in (D) and (E) were performed using an unpaired t test with Welch’s correction.