Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 97 (5), 1063-1077.e4

The Sixth Transmembrane Segment Is a Major Gating Component of the TMEM16A Calcium-Activated Chloride Channel

Affiliations

The Sixth Transmembrane Segment Is a Major Gating Component of the TMEM16A Calcium-Activated Chloride Channel

Christian J Peters et al. Neuron.

Abstract

Calcium-activated chloride channels (CaCCs) formed by TMEM16A or TMEM16B are broadly expressed in the nervous system, smooth muscles, exocrine glands, and other tissues. With two calcium-binding sites and a pore within each monomer, the dimeric CaCC exhibits voltage-dependent calcium sensitivity. Channel activity also depends on the identity of permeant anions. To understand how CaCC regulates neuronal signaling and how CaCC is, in turn, modulated by neuronal activity, we examined the molecular basis of CaCC gating. Here, we report that voltage modulation of TMEM16A-CaCC involves voltage-dependent occupancy of calcium- and anion-binding site(s) within the membrane electric field as well as a voltage-dependent conformational change intrinsic to the channel protein. These gating modalities all critically depend on the sixth transmembrane segment.

Conflict of interest statement

Declaration of Interests

The authors report no competing interests.

Figures

Figure 1
Figure 1. TMEM16A channel conductance is modulated by intracellular calcium, membrane voltage and extracellular anions
A. Sample traces from whole cell patch clamped HEK293 cells expressing wild-type TMEM16A and stepped from −120 mV to +150 mV from a holding potential of 0 mV, in the presence of external 140 mM NaCl or NaI (anion indicated on the left) and internal 140 mM NaCl with 0 nM, 400 nM or 1 mM free calcium (indicated above). Red dashed lines indicate 0 nA. B. Current-voltage (IV) relationships for traces in A highlight the different factors affecting TMEM16A gating. TMEM16A is outwardly rectifying at moderate internal Ca2+ but loses rectification at high internal Ca2+, n = 5, 5 and 7 for 400 nM Ca2+ with Cl, 400 nM Ca2+ with I and 1 mM Ca2+ with Cl traces, respectively. C. Steady-state whole cell conductances, derived from the data shown in B, are corrected against the reversal potential and normalized to +150 mV. D. A simplified schematic of TMEM16A gating illustrates a conceptual model of calcium, voltage and anion-dependent modulation of channel conductance. Error bars in B and C represent mean ± SEM. See also Figures S1 and S2.
Figure 2
Figure 2. Transmembrane helix 6 lines the anion pore and harbors Ca2+-binding sites, and plays a role in Ca2+-dependence of channel activation
A. A representation of mouse TMEM16A TM6 placed above an unbiased sequence alignment of the TM5-6 linker and TM6 of mouse TMEM16A with the other nine mouse TMEM16 family members. Black shading indicates that ≥70% and grey shading indicates 50–70% of TMEM16 family members have the same amino acid as in TMEM16A, which is denoted in red. The alanine-scanned region, indicated by a green bar, includes residues of particular interest marked by red arrows. B. Cryo-EM structure of TMEM16A wuth TM6 highlighted in green, bound Ca2+ ions shown in blue, and the contours of an aqueous pore shown in dark grey. C. Sample traces of inside-out patches pulled from HEK293 cells expressing the indicated construct and exposed to a series of increasing Ca2+ concentrations at +60 mV. Solutions a-f contained 150 nM, 300 nM, 400 nM, 600 nM, 5.5 μM and 1 mM Ca2+, respectively. D. EC50 values for Ca2+ concentration-response relationships from curves fit to traces as in panel C, with statistically significant right shifts highlighted in red, and significant left shifts highlighted in blue. Bars for residues that may directly coordinate Ca2+ ions are shown with diagonal hatches. “N.C.” for constructs with protein expression but no current (see also Figure S2). n for WT, 19; L627A, 6; M628A, 5; E629A, 6; L630A, 6; C631A, 5; I632A, 8; Q633A, 4; L634A, 10; S635A, 9; I636A, 10; I637A, 8; L639A, 6; G640A, 9; K641A, 8; Q642A, 6; I644A, 8; Q645A, 7; N646A, 4; N647A, 5; L648A, 6; F649A, 7; E650A, 4; I651A, 10; G652A, 10; I653A, 11. E. Hill coefficients for fits as in panel D., with significant increased highlighted in red, and significant decrease highlighted in blue. EC50 values and Hill coefficients were compared using one-way ANOVA followed by a Bonferroni post-hoc test for statistical significance. Data in panels D and E are expressed as mean ± SEM.
Figure 3
Figure 3. Alanine substitutions for I637 and Q645 in TM6 allow outward-rectifying anion conductance in the absence of internal Ca2+ and are functionally asymmetric following removal of the G640 hinge
A. Image of TMEM16A highlighting the position (in purple) of I637, G640, and Q645 in TM6 (green), with bound Ca2+ (blue). B. Current-voltage (IV) relationships for WT, I637A, and Q645A TMEM16A with 0 nM internal Ca2+ reveal that the I637A and Q645A mutations remove the requirement for Ca2+ for activation (see also Figure S3) as revealed by whole-cell patch clamp, using 10 mV voltage steps from −120 mV to +150 mV from a holding potential of −80 mV. C. Conductance-voltage (GV) relationships for I637A (red) and Q645A (blue) derived from the IV curves shown in panel B, with filled circles depicting data gathered with chloride and open circles with iodide as the external anion. D. Log10 transformation of the data in C with grey dashed line indicating a non-saturating conductance at the highest positive voltages. G-V traces corrected for external anion-dependence can then be fit with a two-state Boltzmann relationship (see also Figure S5). E. Sample traces from whole cell patch clamp recording of HEK293 cells expressing wild-type TMEM16A or channels bearing either or both Q645A and I637A mutations exposed to a series of voltage commands from a holding potential of −80 mV with external chloride and 0 nM internal Ca2+. Red dashed lines indicate 0 nA. F, G, and H. Graphs depicting normalized GV relationships after subtraction of non-saturating conductance (“Probability-activated” or PA) reveal a conductance with a Boltzmann profile. Solid lines in panel H represent PA with chloride and dashed lines with iodide as the permeant anion. I. Sample traces of inside-out patches exposed to a series of increasing Ca2+ concentrations at +60 mV. Solutions a-f contained 150 nM, 300 nM, 400 nM, 600 nM, 5.5 μM and 1 mM Ca2+, respectively. J. Graph derived from data shown in I revealing the change in conductance with increasing Ca2+. The G640A mutation increases sensitivity to Ca2+, but mutating I636 or M638 to Gly restores the normal Ca2+ sensitivity. K. Bar graph showing that only the I636G and M638G mutations can rescue the G640A mutation. n was WT, 19; G640A, 9; GA I636G, 11; GA I637G, 4; GA M638G, 5; GA L639G, 4; GA A642G, 5; GA L643G, 6; GA I644G, 4. L and M. PA vs voltage relationships comparing the effect of I637A (L) and Q645A (M) mutations in the G640A background, with solid lines depicting the fits to I637A or Q645A mutations alone. N. A comparison of fitted V1/2s for combinations of I637A, Q645A, and G640A, representing information presented in panels F, G, H, L, and M. Data are expressed as mean ± SEM and where error bars are invisible, they are contained within the points. See also Figures S3–S5.
Figure 4
Figure 4. Neutralization and charge reversal of K641 in TM6 shifts the voltage-dependence of TMEM16A and is sensitive to the presence of Ca2+
A. Image of TMEM16A highlighting the positions of K641 (cyan) in TM6, with bound Ca2+ indicated in blue. B. Sample traces from whole cell patch clamp recording of HEK293 cells expressing TMEM16A with the Q645A mutation (left) or Q645A and K641E mutations (right), stepped from −120 mV to +150 mV in 10 mV steps from a holding potential of −80 mV. Chloride was the permeant anion in each case, with 0 nM internal Ca2+ used in the top row and 1 mM Ca2+ used in the bottom row. Red dashed lines indicate 0 nA. C and D. Graphs showing the PA-V relationship for TMEM16A Q645A K641Q (C) and Q645A K641E (D), with blue lines indicating fits to the Q645A single mutant. K641Q shifts the curve to the right and K641E causes a greater shift. E. PA-V relationship of Q645A K641E mutant with 1 mM Ca2+ resembles that of the Q645A single mutant. Solid and dashed lines indicate the fit to Q645A K641E with 0 mM Ca2+ and external chloride (solid) or iodide (dashed). G. Schematic depicting possible changes in TMEM16A conformation after depolarization or Ca2+ binding. Position of residues and side chains do not reflect actual structure and are meant to be illustrative. Data are expressed as mean ± SEM and where error bars are invisible, they are contained within the points.
Figure 5
Figure 5. I637 and Q645 are differently affected by mutations in the Ca2+ binding pocket
A. A close view of the Ca2+ binding pocket of TMEM16A with TM6 in green, Q645 in purple, I637 in cyan, residues interacting with Ca2+ in orange, and Ca2+ in blue. B and C. PA-V relationships of the I637A (B) and Q645A (C) TMEM16A mutants bearing additional TM6 mutations of the Ca2+ binding residues N646 (top) and E650 (bottom). N646A or E650A does not affect the voltage-dependence of the I637A mutant but causes a left-shift in the Q645A mutant. Red and blue lines depict the fits to I637A (red) or Q645A (blue) single mutants for comparison. D and E. Graphs showing PA-V relationships of the Q645A mutant with mutations of Ca2+ binding residues in TM7 (D) and TM8 (E) reveal that mutations in TM7 and TM8 create left shifts in the Q645A mutant background, with the exception of E698. F. A bar graph comparing the V1/2s of Boltzmann fits to the PA-V curves of (B–E). Data are expressed as mean ± SEM and where error bars are invisible, they are contained within the points. See also Figure S6.
Figure 6
Figure 6. Depolarization of TMEM16A channels causes a Ca2+-dependent stabilization of open conformation and slowing of deactivation kinetics
A. Representative traces of whole-cell patch clamp recordings of WT TMEM16A with 400 nM internal Ca2+ with external chloride (top) or iodide (bottom). B. Representative traces of whole-cell patch clamp recordings of WT TMEM16A with external chloride and 200 nM (top) or 1500 nM (bottom) internal Ca2+. In both A and B cells were stepped from −120 mV to +150 mV in 10 mV increments from a holding potential of −80 mV. C. Graph of PA-V for data represented in A and B, showing a progressive left shift in the voltage-dependence upon increasing internal concentration of Ca2+. n was 200 Ca2+ ClI, 5; 400 Ca2+ ClI, 5; 400 Ca2+ II, 8; 1500 Ca2+ ClI, 7. D. Wild-type TMEM16A with 400 nM internal Ca2+ voltage-clamped from a pre-pulse potential of −120 mV (left) or +120 mV (right). Inset shows current immediately after voltage steps. E. “Instantaneous” current amplitudes from wild-type at voltages from −120 to +120 after prepulses to the indicated voltages, normalized to amplitude at +120 mV, n = 5. F. Q645A with 0 internal Ca2+ stepped under voltage clamp from a pre-pulse potential of −120 mV (left) or +120 mV (right). Inset shows current immediately after voltage steps. G. “Instantaneous” current amplitudes from Q645A at voltages from −120 mV to +120 mV after prepulses to the indicated voltages, normalized to amplitude at +120 mV, n = 5 (see also Figure S7). H. Deactivation (top) and reactivation (bottom) time courses following activating and deactivating pulses of increasing duration to +120 mV and −120 mV, respectively, for Q645A with 0 Ca2+ (left) and WT with 400 nM internal Ca2+ (right). For WT, time constants of activation (cyan) and deactivation (orange) are overlaid with the onset of slow deactivation (top) and the decay of instantaneous current amplitude (bottom), respectively. I. Normalized amplitudes of the slow component of current deactivation compared to total for wild-type TMEM16A following 300 ms activating pulses to +120 mV, varying the concentration of intracellular Ca2+ and the species of extracellular permeant anion, as shown J. Slow and fast time constants of current activations for wild-type TMEM16A under the indicated conditions. Also shown is the time constant of onset of a slow component/total amplitude of deactivation during activations of increasing duration (τONdeacts/t) as indicated by arrows in panel E (top, right). K. Slow and fast time constants of current deactivations for wild-type TMEM16A under the indicated conditions. Also shown is the time constant of “loss of instantaneous” (L.O.I.) current, representing the recovery of the slowly activating current at +120 mV following deactivations of increasing duration at −120 mV as indicated by arrows in panel H (bottom, right), and the single exponential “fast” τ of deactivation for Q645A with 0 Ca2+ and E650A with 1 mM Ca2+ (see also Figure S7). n for data depicted in panels J and K was τONdeacts/t, 7; 200 Ca2+ ClI 200 Ca2+ II, 9; 400 Ca2+ ClI, 13; 400 Ca2+ II, 8; τOFF”L.O.I.”, 8; Q645A, 6; E650A, 6. Red dashed lines in panels A, B, D and F indicate 0 nA. All pooled data are expressed as mean ± SEM. See also Figure S7.
Figure 7
Figure 7. A dynamic interplay of Ca2+ and voltage controls TMEM16A gating
A. Left, a simplified model depicting changes to TMEM16A. Two circles within the box represent the two Ca2+ binding sites, with filled circles indicating occupancy by Ca2+. The larger circle represents the pore, with a line indicating a closed pore, half-filled circle for the outward-rectifying open state, and fully filled circle for the ‘Ohmic’ open state. Right, A sample steady-state I-V distinguishing the two open states for channels with single or double Ca2+ occupancy. B. A basic gating scheme for the WT and Q645A mutant TMEM16A under conditions of 0 Ca2+I. Left, the WT channel does not activate in response to voltage changes (up to +200 mV) without Ca2+ while the Q645A mutant (right), activates in response to membrane voltage potential changes as though with single Ca2+ occupancy, indicated by a gray filled circle. C. The WT channel, under conditions of moderate Ca2+ concentration, exists in equilibrium (gray dotted box) with the majority of channels not bound to Ca2+ and a small number bound to a single Ca2+. After depolarization, the few channels with single Ca2+ occupancy open quickly and allow anion flux. As time passes an increasing number of channels enter the single-Ca2+-bound state whereupon they can be opened by depolarization, which also favors the binding of the second Ca2+ ion, stabilizing the channel and leading to slowly activating current upon depolarization. With elevated Ca2+ levels, the channel can readily bind two Ca2+ ions and enter a non-rectifying ‘Ohmic’ open state. D. A sample trace illustrates different phases of TMEM16A conductance after depolarization under moderate Ca2+ levels. A small, rapid outward current (1) results from opening of channels with single Ca2+ occupancy while the slower increase (2 and 3) reflect the need for channels to first bind Ca2+ before entering an open state. E. A model of TM6 gating in response to Ca2+ and voltage changes. Green residues are those in TM6 critical for the Ca2+ requirement for activation, while magenta indicates K641. Red circles and red residues indicate the Ca2+-binding residues of TM6, 7, and 8. Blue circles are Ca2+ ions.

Similar articles

See all similar articles

Cited by 18 PubMed Central articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback