Since Ca2+ is a major competitor of protons for the modulation of high voltage–activated Ca2+ channels, we have studied the modulation by extracellular Ca2+ of the effects of proton on the T-type Ca2+ channel α1G (CaV3.1) expressed in HEK293 cells. At 2 mM extracellular Ca2+ concentration, extracellular acidification in the pH range from 9.1 to 6.2 induced a positive shift of the activation curve and increased its slope factor. Both effects were significantly reduced if the concentration was increased to 20 mM or enhanced in the absence of Ca2+. Extracellular protons shifted the voltage dependence of the time constant of activation and decreased its voltage sensitivity, which excludes a voltage-dependent open pore block by protons as the mechanism modifying the activation curve. Changes in the extracellular pH altered the voltage dependence of steady-state inactivation and deactivation kinetics in a Ca2+-dependent manner, but these effects were not strictly correlated with those on activation. Model simulations suggest that protons interact with intermediate closed states in the activation pathway, decreasing the gating charge and shifting the equilibrium between these states to less negative potentials, with these effects being inhibited by extracellular Ca2+. Extracellular acidification also induced an open pore block and a shift in selectivity toward monovalent cations, which were both modulated by extracellular Ca2+ and Na+. Mutation of the EEDD pore locus altered the Ca2+-dependent proton effects on channel selectivity and permeation. We conclude that Ca2+ modulates T-type channel function by competing with protons for binding to surface charges, by counteracting a proton-induced modification of channel activation and by competing with protons for binding to the selectivity filter of the channel.

At present, there are several lines of evidence that permeant ions modulate T-type Ca2+ channel gating. Shuba et al. (1991) showed that the mean open time of the T-type channel of mouse neuroblastoma cells depends on the type and concentration of the permeant ion and proposed that the channel region controlling channel closure senses a smaller fraction of the electric field due to a dynamic interaction between the ionic flux and the selectivity filter. In frog-atrial cardiomyocytes, Alvarez et al. (2000) showed that channel reopening after strong depolarizing prepulses was enhanced in low-Na+ extracellular solutions and proposed that a competition between Na+ and Ca2+ for binding sites within the channel modulates the amplitude of the voltage-facilitated T-type current. It is also known that the α1G clone (CaV3.1) shows faster macroscopic inactivation in the presence of Ba2+ than in Ca2+ (Klugbauer et al., 1999; Staes et al., 2001).

T-type Ca2+ channels are also modulated by extracellular protons. Tytgat et al. (1990) reported that, in guinea pig cardiomyocytes, low extracellular pH (pHe) induced a positive shift in the voltages for half-maximal activation (Vact) and inactivation (Vinac) together with an increase in the slope factor of channel activation (sact). In a recent paper, Delisle and Satin (2000) concluded that proton block of T-type currents (α1H, CaV3.2) under physiological conditions is due to gating modifications. These authors related the shift in the voltage for half-maximal activation by high extracellular proton concentrations to the titration of negative surface charges. To explain the increased slope factor of activation, they proposed that external acidification titrates some of the charges involved in the voltage sensing, but suggested already to test the influence of intrapore ions on the gating mechanisms.

Protons have also been shown to affect the open pore conduction of T-type channels. Tytgat et al. (1990) showed that extracellular acidification from pH 9.0 to 6 decreases the single channel conductance of the cardiac T-type channel of the guinea pig using 110 mM Ca2+ as charge carrier. For the α1H T-type clone, Delisle and Satin (2000) found that open pore block was not the main determinant for the inhibition of whole-cell currents by extracellular protons in the presence of 2.5 mM extracellular Ca2+ and 140 mM Na+. They reported a negative shift of the reversal potential after extracellular acidification and suggested that channel protonation shifts channel selectivity toward monovalent ions.

Protons have been shown to modulate the gating processes of high voltage–activated (HVA)* Ca2+ channels (Pietrobon et al., 1989; Prod'hom et al., 1989; Kwan and Kass, 1993; Tombaugh and Somjen, 1996; Zhou and Jones, 1996; Smirnov et al., 2000; Shah et al., 2001). In these channels, Ca2+ has been shown to be a major competitor of protons for the neutralization of surface charges (Kwan and Kass, 1993) and for binding to pore residues that control ion permeation and selectivity (Chen et al., 1996; Chen and Tsien, 1997). To achieve further insight in the mechanisms of ionic modulation of T-type channel function, we investigated the influence of the extracellular Ca2+ concentration ([Ca2+]e) on the proton-induced modifications of gating and permeation mechanisms of the α1G T-type subunit and the role of the selectivity filter (the EEDD pore locus) in the modulation of channel selectivity and permeation by protons. Our results suggest for three substrates of competition between Ca2+ and protons, resulting in modifications of channel gating and ionic conduction.

Solutions

Before current recordings, cells were rinsed with Krebs solution containing: 150 mM NaCl, 6 mM KCl, 1 mM MgCl2, 1.5 mM CaCl2, 10 mM glucose, 10 mM HEPES, and titrated to pH 7.4 with 1 N NaOH. All extracellular test solutions contained 5 mM CsCl, 5 mM glucose, a mixture of MES, HEPES, and TAPS (5 mM each) to extend the buffering range from pH 5.5 to 9.1, and were kept Mg2+ free to avoid extracellular Mg2+ block (Serrano et al., 2000). We used extracellular solutions of 2 or 20 mM CaCl2 or a Ca2+ free solution (5 mM EDTA) with 20 mM NaCl, osmotically compensated with 150, 120, or 130 mM NMDG, respectively. To assess the influence of Na+ on conduction properties in the 2 mM Ca2+ solution, 150 mM NMDG was substituted by an equimolar amount of Na+. Test solutions containing NMDG were titrated with HCl and those containing 150 mM NaCl with CsOH. When 2 or 20 mM Ca2+ solutions were used, the intracellular (pipette) solution contained 102 mM CsCl, 10 mM HEPES, 5 mM MgCl2, 5 mM Na2-ATP, 10 mM TEA-Cl, 10 mM EGTA, titrated to pH 7.4 with 1 N CsOH. In the experiments in zero [Ca2+]e the intracellular solution did not contain MgCl2 and Na2-ATP in order to avoid block by Mg2+ and to limit the size of outward currents, respectively. All chemicals were purchased from Sigma-Aldrich.

Electrophysiology

We used human embryonic kidney cells (HEK293) grown and transfected with the wild-type α1G (Klugbauer et al., 1999) or the EEED pore mutant as in previous works (Staes et al., 2001; Talavera et al., 2001). Currents were recorded in the whole-cell configuration of the patch-clamp technique using an EPC-7 (LIST Electronics) patch-clamp amplifier and filtered with an eight-pole Bessel-filter (Kemo). For control of voltage-clamp protocols and data acquisition, we used an IBM-compatible PC with a TL-1 DMA interface (Axon Instruments, Inc.) and the software pCLAMP (Axon Instruments, Inc.). Bath solutions were perfused by gravity via a multibarreled pipette. Patch pipettes were pulled from Vitrex capillary tubes (Modulohm) using a DMZ-Universal puller (Zeitz-instruments). An Ag-AgCl wire was used as reference electrode. Adequate voltage control was achieved by using low pipette resistances (1–2.5 MΩ) and series resistance compensation to the maximum extent possible (50–70%). Membrane-capacitive transients were electronically compensated and the linear background components were digitally subtracted before data analysis. Current traces were filtered at 2.5–5 kHz and digitized at 5–10 kHz. All experiments were performed at room temperature (22–25°C).

Stimulation Protocols and Data Analysis

I-V curves were obtained from the peak amplitude of currents evoked by the application 200-ms lasting voltage steps from −90 to 60 mV. These curves were fitted with:

\begin{eqnarray*}&&\mathit{I}\left(\mathit{V}\right)=\mathit{G}\left(\mathit{V}\right){\cdot}\left(\mathit{V}{-}\mathit{V}_{\mathit{r}}\right)\left\{\frac{\mathit{A}}{\mathrm{1}+\mathrm{exp}\left[{-}{\left(\mathit{V}{-}\mathit{V}_{\mathit{act}}\right)}/{\mathit{s}_{\mathit{act}}}\right]}+\right.\ \\&&\left.\ \frac{\mathrm{1}{-}\mathit{A}}{\mathrm{1}+\mathrm{exp}\left[{-}{\left(\mathit{V}{-}\mathit{V}_{\mathit{act}}^{{^\prime}}\right)}/{\mathit{s}_{\mathit{act}}^{{^\prime}}}\right]}\right\},\end{eqnarray*}
(1)

where I is the measured peak current, V the step potential and G(V) the conductance, which may be voltage dependent. The activation curves in 2 and 20 mM Ca2+ solutions were best described by the sum of two Boltzmann components (Serrano et al., 1999; Lacinova et al., 2002). Vact and V′act are the potentials of half-maximal amplitude, sact and s′act are the slope parameters and A is the amplitude of the steeper component. In Ca2+ free solutions only one Boltzmann function was needed (A = 1). To account for the strong inward rectification in the 2 and 20 mM Ca2+ solutions, we have approximated G(V) from a fit of the amplitudes of the tail current after a depolarizing step to 100 mV in the voltage range from −70 to 110 mV for α1G and from −70 to 70 mV for the EEED mutant with the equation:

\[\mathit{I}_{\mathit{tail}}\left(\mathit{V}\right)=\mathit{G}\left(\mathit{V}\right){\cdot}\left(\mathit{V}{-}\mathit{V}_{\mathit{r}}\right),\]
(2)

with

\(\mathit{G}\left(\mathit{V}\right)=\mathit{G}{\cdot}\left(\mathit{V}^{\mathrm{2}}+\mathit{b}{\cdot}\mathit{V}+\mathit{c}\right)\)
⁠. The I-V relation in the Ca2+ free, 20 mM Na+ extracellular solution is fairly linear, and hence G(V) = G.

The average values of the fit parameters b and c for each experimental condition were then used to fit the I-V curves with Eq. 1. We noticed that an adequate description of the open pore conduction was essential for an accurate estimation of the slope factor sact of the activation curve, and that the use of the classical linear function for

\(\mathit{G}{\cdot}\left(\mathit{V}{-}\mathit{V}_{\mathit{r}}\right)\)
in Eq. 1 consistently overestimated the values for sact for the currents in 2 or 20 mM Ca2+. Activation curves were calculated by dividing the experimental I-V curves by the appropriate
\(\mathit{G}\left(\mathit{V}\right){\cdot}\left(\mathit{V}{-}\mathit{V}_{\mathit{r}}\right)\)
function in each experimental condition.

The time constants of activation (τact) and inactivation (τinac) were determined from a fit of current traces with the equation:

\[\mathit{I}\left(\mathit{t}\right)=\mathit{I}_{\mathit{m}}{\cdot}\left[\mathrm{1}{-}\mathrm{exp}\left({{-}\mathit{t}}/{{\tau}_{\mathit{act}}}\right)\right]^{\mathit{2}}{\cdot}\mathrm{exp}\left({{-}\mathit{t}}/{{\tau}_{\mathit{inac}}}\right)+\mathit{I}_{\mathit{b}},\]
(3)

where Im is a normalization factor and Ib is a background component. Given the sampling rate used in the activation protocol, the values of τact at positive voltages (<0.5 ms) should be considered as an upper limit of the real values. Nevertheless, our estimates are equal or smaller than previously reported values (0.6 ms, Nilius, 1992; 0.3 ms, Chen and Hess, 1990). The decaying phase of the voltage dependence of τact was fitted with an exponential function of the form:

\[{\tau}_{\mathit{act}}\left(\mathit{V}\right)=\mathrm{exp}\left[{{-}\left(\mathit{V}{-}\mathit{V}{\tau}_{\mathit{act}}\right)}/{\left(\mathit{s}{\tau}_{\mathit{act}}\right)}\right]+{\tau}_{\mathit{act}}\left({\infty}\right),\]
(4)

where sτact is the voltage sensitivity, τact(∞) is the asymptotic value at positive potentials and Vτact is the voltage at which τact is equal to 1 + τact(∞).

Steady-state inactivation (h) was determined from the peak current recorded during a 160-ms test pulse to 0 mV after a 5,120-ms lasting prepulses to potentials between −100 and −25 mV. The voltage dependence of these peak currents, normalized to that following the prepulse to −100 mV, was fitted with the equation:

\[h_{\mathrm{{\infty}}}\left(V\right)=\frac{1}{1+\mathrm{exp}\left[{\left(V{-}V_{inac}\right)}/{s_{inac}}\right]}\mathrm{,}\]
(5)

where Vinac is the potential of half-maximal inactivation and sinac the slope factor for the inactivation.

To dissect the effects of ionic conditions on conduction properties from those on gating, tail currents were recorded during voltage steps between −120 and 110 mV after a 7.5-ms lasting depolarization to 100 mV to normalize for the positive voltage shift of the steady-state activation at low pHe. Linear background components and capacitive transients were subtracted by the application of a −P/4 protocol. In the presence of 2 mM Ca2+ there was an ∼10% decrease in the estimated maximal open probability during the prepulse to 100 mV when the pHe was changed from 9.1 to 6.2. However, at 7.5 ms the estimated open probability was not significantly different between these conditions (unpublished data). This ensured that the changes in the amplitude of the tail currents with the extracellular perfusion condition were determined by open pore properties and not by modifications in gating during the prepulse to 100 mV. We determined the amplitude (Itail) and time constant of current decay (τdecay) from a single-exponential fit of the time course of the tail currents. τdecay corresponds to the time constant of deactivation (τdeac) at negative potentials and to the time constant of the macroscopic inactivation (τinac) at potentials positive to the activation threshold. The usual way to characterize the voltage dependence of τdeac is to fit τdecay(V) with a growing exponential function in a range of very negative potentials. This approximation is expected to underestimate the steepness of the voltage dependency of τdeac since it does not consider the contribution of the inactivation to current decay. Additional underestimation of the steepness may occur when fitting tail currents at very negative potentials given that the estimated values of τdecay tend to be higher than the actual ones due to voltage clamp errors that are the consequence of large tail current amplitudes. When considering the current models of T-type channel gating (see Fig. 13), in a second order of approximation, the voltage dependence of τdecay can be expressed by:

\[\mathrm{{\tau}}_{decay}\left(V\right)=\frac{1}{\mathrm{exp}\left[{{-}\left(V{-}V\mathrm{{\tau}}_{deac}\right)}/{\left(s\mathrm{{\tau}}_{deac}\right)}\right]+k_{OIo}}\mathrm{,}\]
(6)

where V is the repolarization potential, Vτdeac the voltage at which τdeac is equal to 1 ms, sτdeac the voltage sensitivity of the time constant of deactivation, and kOIo is the rate constant of the transition between the open state (O) and the closest inactivated state (IO). This approximation is proved to be valid as the channels have low probability of being in the closed state (C3) near the open state and the inactivation process is largely absorbing (kOIo >> kIoO) (see Fig. 13).

In all voltage protocols the holding potential was −100 mV and the stimulation frequency 0.5 Hz, with the exception of the inactivation protocol in which it was 0.2 Hz. Electrophysiological data were analyzed using the WinASCD software package (ftp://ftp.cc.kuleuven.ac.be/pub/droogmans/winascd.zip; G. Droogmans, Laboratory of Physiology, KU Leuven). For all measurements pooled data are given as mean ± SEM. We used ANOVA and Student's paired t test, taking P < 0.05 or P < 0.01 as the level of significance.

Data Simulation

We used MATLAB (MathWorks) to solve a Markov model for the gating of α1G in several pHe, in the presence of 2 or 20 mM Ca2+. Parameter optimization and numerical solution of the differential equations were performed with the built-in functions fmin and expm, respectively.

Effects of pHe and Extracellular Ca2+ on the Activation of α1G

We tested the modulation of the gating of α1G by extracellular Ca2+ and protons by studying the effects of pHe in the range from 9.1 to 5.5 on I-V curves in the presence of 0, 2, or 20 mM Ca2+. Fig. 1, A–C, shows current traces at various pHe values in the presence of 2 mM Ca2+ during voltage steps to −60, −40, and −20 mV. Current reduction at low pHe was most pronounced at more negative potentials. As a consequence, the I-V curve, derived from the peak current amplitudes, was shifted in the positive direction at lower pHe (Fig. 1 D). Fig. 1, E and F, shows that extracellular acidification also shifts the voltage dependence of the time constants of activation (τact) and macroscopic inactivation (τinac).

These effects of pHe were less prominent in 20 mM Ca2+ (Fig. 2, A–C). Acidification significantly inhibited the inward current, but the block was still incomplete at pHe 5.5. The shift of the peak I-V curve and of the voltage dependence of τact and τinac were also much smaller than in 2 mM Ca2+ (Fig. 2, D–F).

The effects of pHe were most pronounced in the absence of extracellular Ca2+ (5 mM EDTA). Under these conditions, T-type Ca2+ channels, like HVA Ca2+ channels, conduct large currents of monovalent cations (Carbone and Lux, 1987; Talavera et al., 1998; Alvarez et al., 2000). Fig. 3, A–C, shows current traces recorded at pHe values ranging from 9.1 to 5.5 during voltage pulses to −80, −60, and 0 mV, respectively. The current at −80 mV was completely inhibited by changing pHe from 9.1 to 8.2, but further acidification was necessary to fully block the current at less negative potentials. Because of the reduced extracellular Na+ concentration (20 mM), currents were outward at potentials positive to −15 mV. The current at 0 mV was completely inhibited only at pHe 5.5. The voltage shift of the I-V curves due to external acidification was more pronounced than in the presence of 2 or 20 mM Ca2+ (Fig. 3 D). Remarkably, a reduction of pHe from 9.1 to 7.4 induced a positive shift of the voltage of the peak inward current by ∼20 mV, and inward currents were almost completely inhibited at pHe 6.2. The effects of protons on the voltage dependences of τact and τinac were clearly more pronounced than in the presence of 2 or 20 mM Ca2+ (Figs. 3, E and F, see below).

Fig. 4 A shows the effects of protons on the voltage-dependence of α1G activation in the presence of 0, 2, and 20 mM Ca2+. The extracellular acidification to pH 6.2 did not only shift the activation curve to depolarized potentials, but also reduced its slope. It is remarkable that this effect of acidification on the slope of the activation curve, which is incompatible with a screening of surface charges by protons, was much larger in the absence of extracellular Ca2+.

Activation curves were best described by two Boltzmann components in 2 and 20 mM Ca2+. The amplitude (1–A) and the slope factor (sact) of the shallower component were around 0.45 and 8 mV, respectively, and were pHe and [Ca2+]e independent (unpublished data). The voltage for half-maximal activation of both components (Vact and Vact) showed identical pHe and [Ca2+]e dependencies. Therefore, we only show in detail the results regarding the modification of the steeper Boltzmann component. Vact depends on both pHe and [Ca2+]e, as shown in Fig. 4 B. Changing pHe from 9.1 to 6.2 shifted Vact by ∼80 mV in the absence of Ca2+, and by ∼20 and 10 mV in the presence of 2 or 20 mM Ca2+, respectively. The values of Vact in 2 and 20 mM Ca2+ seem to converge at low pHe, suggesting that protons and Ca2+ may compete for the same binding sites. A peculiar finding is that the curves representing Vact as a function of pHe in Ca2+-free solution and those in 2 or 20 mM Ca2+ intersect, the values for Vact in Ca2+-free solutions being more negative at alkaline pHe (>6.8) and more positive at acidic pHe (see also Fig. 4 A). This observation is incompatible with the neutralization of surface charges by competing Ca2+ and protons as the only mechanism underlying the voltage shift of channel activation.

The slope factor of the steeper component of the activation sact increased with extracellular acidification, an effect that is antagonized by increasing [Ca2+]e (Fig. 4 C). Reducing pHe from 9.1 to 7.4 significantly increased sact in 2 mM Ca2+, whereas acidification to pHe 6.2 and below was necessary to affect sact in 20 mM Ca2+. Also the values of sact in 2 and 20 mM Ca2+ seem to converge at low pHe. The changes in sact were much more dramatic in the absence of Ca2+, since this parameter increased about fourfold if pHe was changed from 9.1 to 6.2.

The rightward shift and increased slope factor of the activation curve by extracellular acidification might result from a voltage-dependent open pore block (Woodhull, 1973). Such a mechanism would, however, not affect the kinetics of channel activation, and is in contrast with the observed shift of the voltage dependences of the time constants of activation τact and macroscopic inactivation τinac (Figs. 13). The rightward shift of the voltage dependence of τact at pHe 9.1 by increasing [Ca2+]e is consistent with a mechanism of screening of surface charges by Ca2+. However, the observation that a decrease of pHe did not only shift the voltage dependence of τact to more positive voltages but also decreased its slope is inconsistent with simple screening by protons. To quantify these effects, we have estimated the position of the curve τact(V) along the voltage axis (Vτact), and the steepness of the voltage dependence (sτact) from the fit of the data points of the decaying phase with Eq. 4 for each experimental condition. Fig. 5, A and B, show that the patterns of modulation of the pHe dependence of Vτact and sτact by extracellular Ca2+ are similar to those of Vact and sact (compare with Fig. 4, B and C). This indicates that the Ca2+-dependent inhibition of α1G by protons is to a large extent due to an effect of protons on gating. The issue of open pore block by protons and its Ca2+ dependence is addressed below.

Extracellular Ca2+ Modulates the Effect of Protons on Steady-state Inactivation of α1G

Tytgat et al. (1990) reported that extracellular alkalinization induced a negative shift of the voltage dependence of steady-state inactivation of native T-type Ca2+ channels, and Delisle and Satin (2000) showed a positive shift with extracellular acidification in the human-cloned α1H T-type channel. In this paper, we compared the effects of pHe on the steady-state inactivation properties of α1G in the presence of 2 or 20 mM Ca2+. Because a change of pHe from 9.1 to 8.2 in the presence of 20 mM Ca2+ did not affect channel activation, we used the results at pHe 8.2 as a reference for the proton-induced changes in channel availability for this [Ca2+]e. For 2 mM Ca2+ the results at pHe 9.1 were taken as reference. Fig. 6 A compares the changes in average channel availability after extracellular acidification from 9.1 to 6.2 in the presence of 2 mM Ca2+ and from 8.2 to 6.2 in 20 mM Ca2+. The pHe dependence of the average potential for half-maximal inactivation (Vinac), as estimated from the fits of the data points with Eq. 5, is shown in Fig. 6 B. To better characterize the changes in Vinac, we have averaged the voltage shifts of the individual cells. These results, plotted in Fig. 6 C, show that a change of pHe from 9.1 to 7.4 or lower significantly shifts the availability curves at 2 mM Ca2+ to more positive potentials, but an acidification up to pHe 6.2 was necessary to induce a significant shift at 20 mM Ca2+. External acidification up to pH 6.8 in 2 mM Ca2+ did not significantly affect the slope factor of the inactivation process (sinac, Fig. 6 D), but this factor was significantly increased at pHe 6.2. In contrast, a significant effect on sinac in 20 mM Ca2+ was only observed at pHe 5.5.

pHe-induced Changes on Activation and Inactivation Are Partially Correlated

The effects of pHe on both processes of activation and inactivation of α1G are consistent with the previously reported coupling between activation and macroscopic inactivation of single T-type Ca2+ channels (Droogmans and Nilius, 1989). We have therefore studied the correlation between the parameters describing the steady-state activation and inactivation when varying pHe. Fig. 7 A shows the correlation between Vinac and Vact obtained at different pHe values and at 2 or 20 mM Ca2+. Assuming a linear relationship between Vinac and Vact, we obtained slopes of 0.49 ± 0.09 (R = 0.97) and 0.44 ± 0.08 (R = 0.99) at 2 and 20 mM Ca2+, respectively. Proton-induced changes in sact from 3.0 to 4.1 mV (in 2 mM Ca2+) and from 3.0 to 4.6 mV (in 20 mM Ca2+) were not paralleled by significant changes in sinac, indicating that the voltage sensitivity of steady-state inactivation was less pH sensitive than that of activation (Fig. 7 B). On the other hand, proton-induced changes in Vact were consistently accompanied by changes in sact (in both 2 and 20 mM Ca2+), whereas changes in Vinac up to 12.7 mV (in 2 mM Ca2+) or 8.5 mV (in 20 mM Ca2+) did not correspond to significant changes in sinac (Fig. 7, C and D).

The Effects of Extracellular Protons on Ion Permeation and Selectivity Depend on Extracellular Ca2+ and Na+. Comparison between α1G and the EEED Pore Mutant

Besides their effects on gating, extracellular protons alter ion permeation through T-type Ca2+ channels (Tytgat et al., 1990; Delisle and Satin, 2000). It has been shown that permeation and selectivity of Ca2+ channels depend on Ca2+ and Na+ (Polo-Parada and Korn, 1997) and that T-type channels are less Ca2+ selective than HVA Ca2+ channels (Lee et al., 1999; Serrano et al., 2000). We were therefore interested to find out whether the effects of protons on ion permeation through α1G were Ca2+ and/or Na+ dependent.

We analyzed the effects of pHe on the amplitude of α1G tail currents in 2 mM Ca2+ (150 mM Na+ or NMDG+) or 20 mM Ca2+ (120 mM NMDG+) (Fig. 8, A–C). Decreasing pHe from 9.1 to 6.2 in 2 mM Ca2+ and 150 mM NMDG+ significantly reduced inward tail currents, but did not affect outward currents (Fig. 8 A). The effects of pHe on the inward currents were smaller if NMDG+ was replaced by Na+ (Fig. 8 B) or if [Ca2+]e was increased to 20 mM (Fig. 8 C). To quantify the voltage dependence of this proton block, we have calculated at each potential and for each cell and Ca2+-Na+ condition the ratio of the current amplitudes at pHe 6.2 and 9.1 (I(pH6.2)/I(pH9.1), Fig. 8 D). Extracellular protons exerted a significant (≈33%) block of the current in 2 mM Ca2+ in the absence of extracellular Na+. This block was voltage independent in the range from −120 to 20 mV and declined sharply around the reversal potential. The voltage range in which the block was voltage dependent was larger in 2 mM Ca2+ and 150 mM Na+, but the maximal block was the same as in the Na+ free solution. The proton block was significantly smaller at 20 mM Ca2+ (≈10%), indicating that extracellular Ca2+ antagonizes the effects of protons on ion conduction through α1G.

We performed similar experiments using a pore mutant in which aspartate D1487 of the P-loop in domain III was substituted by glutamate, because the divalent over monovalent cation selectivity of this mutant channel (called EEED) is reduced, and because it is more sensitive to protons than the wild-type α1G (Talavera et al., 2001). The maximum proton block of the mutant EEED in 2 mM Ca2+ and in the absence of Na+ was similar to that of the wild-type (≈35%), but it was voltage-dependent over a much broader voltage range (Fig. 9, A and D). However, extracellular acidification significantly increased the amplitude of the tail currents in the presence of Na+, indicating that channel protonation strongly enhances Na+ conduction through the mutant channel (Fig. 9, B and D). Like for the wild type channel, proton block was smaller in 20 mM Ca2+ (≈20%, which was larger than in α1G) and voltage dependent in the range from −100 to 0 mV (Fig. 9, C and D).

We can assess the effects of extracellular acidification on channel selectivity and its possible dependence on [Ca2+]e from the reversal potentials (Vr) in the various experimental conditions. Fig. 10, A and B, show the average amplitudes of tail currents from cells expressing α1G and the EEED mutant, respectively, at an expanded voltage scale in the region of Vr. Extracellular acidification shifted the Vr of α1G to less positive values in 2 mM Ca2+ but not in 20 mM Ca2+ (Fig. 10 C). Vr was less positive for the EEED mutant than for α1G under all experimental conditions. Extracellular acidification significantly shifted Vr in the mutant channel (even to a larger extent than in α1G) in 2 mM Ca2+ and 150 mM NMDG+, but did not affect it either in 2 mM Ca2+ and 150 mM Na+ nor in 20 mM Ca2+ (120 mM NMDG+). Substitution of NMDG+ for Na+ did not significantly change Vr for α1G, but shifted it significantly to more positive potentials for the EEED mutant at pHe 9.1 and 6.2.

The effects of extracellular Na+ on the currents through a1G and the EEED mutant T-type channel, in the presence of 2 mM Ca2+, appeared to be pHe dependent. To quantify these effects, we have calculated the ratio of current amplitudes in Na+ and NMDG+ (I(Na+)/I(NMDG+)) at each repolarization voltage, for pHe 9.1 and 6.2 (Fig. 11). At pHe 9.1, inward currents through α1G were 25% smaller in the presence of Na+ than those in the presence of NMDG+, whereas the outward currents were ∼30% larger. The block of inward currents was, however, weaker at pHe 6.2, especially at very negative potentials. Substitution of NMDG+ by Na+ increased inward tail currents through the EEED mutant by 25–60% at pHe 9.1, but this increase was ∼150% at pHe 6.2.

Modulation of Deactivation Kinetics of α1G by Extracellular Protons Is Dependent on Extracellular Ionic Conditions. Effects of Extracellular Ca2+ and Na+

Extracellular acidification consistently accelerated channel deactivation (insets in Fig. 8, A–C, and 9, A–C). We describe here that this effect is also dependent on the [Ca2+]e. Fig. 12, A and B, shows the voltage dependence of the time constant of the current decay (τdecay), estimated from single-exponential fits of tail currents, for α1G and the EEED mutant in 2 mM (150 mM Na+ or NMDG+) and in 20 extracellular mM Ca2+ (120 mM NMDG+). The fit of these data with Eq. 6 yield the parameters that describe the voltage dependence of the time constant of deactivation (τdeac), i.e., the voltage at which τdeac was equal to 1 ms (Vτdeac, to describe the position along the voltage axis) and the steepness of the voltage dependence (sτdeac). Fig. 12 C shows that Vτdeac was more negative for both channels and pHe values in the presence of 150 mM Na+ (2 mM Ca2+), whereas extracellular acidification and 20 mM Ca2+ shifted the voltage dependence of τdeac to less negative potentials. At pHe 6.2 the voltage dependencies in 2 mM and 20 mM Ca2+ nearly overlap, suggesting that protons and Ca2+ compete to screen and/or bind to the same surface charges as shown for the activation process. As reported for α1H (Delisle and Satin, 2000), extracellular acidification did not significantly modify the voltage sensitivity of the deactivation process (sτdeac) of α1G nor that of the EEED mutant (Fig. 12 D). Interestingly, the EEED mutant showed less negative Vτdeac values and larger sτdeac than the wild-type channel in all experimental conditions and smaller values of sτdeac in the presence of 20 mM Ca2+ than in 2 mM.

Both gating and permeation mechanisms of voltage-dependent Ca2+ channels are modulated by ionic conditions and particularly by extracellular protons and Ca2+ (Prod'hom et al., 1989; Tytgat et al., 1990; Shuba et al., 1991; Kwan and Kass, 1993; Chen et al., 1996; Polo-Parada and Korn, 1997; Alvarez et al., 2000; Delisle and Satin, 2000; Shah et al., 2001). For T-type Ca2+ channels in particular it has been shown previously that extracellular protons shift the voltage dependence of channel activation due to neutralization of surface charges and decrease the voltage sensitivity of channel activation, which is not consistent with the surface charge hypothesis (Tytgat et al., 1990; Delisle and Satin, 2000; Shah et al., 2001). In addition, protons either reduce single channel conductance in 110 mM Ca2+ (Tytgat et al., 1990), or increase the whole-cell conductance and shift of channel selectivity toward monovalent ions in nearly physiological conditions (2.5 mM Ca2+ and 140 mM Na+) (Delisle and Satin, 2000). Taking into account that Ca2+ and protons interact in the modulation of channel gating and ion permeation in HVA Ca2+ channels (Kwan and Kass, 1993; Chen et al., 1996; Chen and Tsien, 1997), we were interested to investigate if Ca2+ modulates the effects of pHe on T-type channel function.

Ca2+ Inhibits the Effects of Protons on Channel Gating

Tytgat et al. (1990) reported that changes in pHe from 9 to 6 did not shift the voltage for half-maximal activation Vact in the presence of 110 mM Ca2+, but significantly changed it in 5.4 mM Ca2+. In the present study, we show that extracellular acidification induced a shift of Vact and increased the slope factor of activation sact in the presence of either 0, 2 or 20 mM Ca2+, and that the magnitudes of these effects were more prominent at lower [Ca2+]e. Remarkably, protons shifted Vact to more positive values in Ca2+ free solutions than in 2 or 20 mM Ca2+ (Fig. 4), indicating that protons induce an extra positive shift in the voltage dependence of the activation process that cannot be explained by the neutralization of surface charges. The shift of Vact and increase of sact could be interpreted as a voltage-dependent open pore block by protons (Woodhull, 1973). However, the finding that Ca2+-dependent proton effects on the voltage dependence of τact were similar to those on steady-state activation indicate that the proton effects on the shape of the current-voltage relationship are to a large extent due to modifications in channel gating. Moreover, the open pore block by protons was proven to be voltage independent when NMDG+ was present in the bath solution. In a more physiological condition (150 mM extracellular Na+), the voltage dependence of the proton block shows an opposite slope to that expected from a Woodhull-like open pore block (see below).

Extracellular protons also affected inactivation of T-type channels by shifting the voltage for half-maximal inactivation Vinac, as it has been reported previously for cardiac T-type Ca2+ channels (Tytgat et al., 1990) and the T-type α1H subunit (Delisle and Satin, 2000). Changes in the slope factor of the inactivation curve sinac had not been reported so far. Our results show that extracellular Ca2+ antagonizes the effects of pHe on both Vinac and sinac of α1G, in the sense that significant changes in these parameters require larger changes in pHe at higher Ca2+ concentrations. These effects of protons on channel inactivation might be indirect, due to a coupling between activation and inactivation, as discussed in the next section.

Extracellular protons shift the voltage dependence of the time constant of deactivation τdeac of α1G along the voltage axis, in agreement with the results of Delisle and Satin (2000) for α1H. Furthermore, we show that this shift, as for the other gating processes, depends on [Ca2+]e. Delisle and Satin (2000) found similar depolarizing shifts of the activation and deactivation processes for a pHe change from 8.2 to 5.5 in 2.5 mM Ca2+, and suggested that these might be due to the neutralization of surface charges by extracellular protons. However, a closer examination of the Ca2+ effects reveals peculiar features of the proton modulation. With a milder acidification we observed that the proton-induced shift in Vτdeac was larger than that in Vact at high [Ca2+]e. Extracellular acidification from pH 9.1 to 6.2 shifted Vτdeac and Vact, respectively, by 19 and 20.1 mV in 2 mM Ca2+, but by 13 and 7 mV in 20 mM Ca2+ (compare Figs. 4 B and 5 C with Fig. 12 C). We believe that these results are not in contradiction with those of Delisle and Satin (2000) because they worked in proton-favoring conditions that possibly override Ca2+ effects. However, our findings cannot be explained by the standard surface potential theory (Frankenhaeuser and Hodgkin, 1957), which implies that protons shift the voltage dependence of all gating parameters by the same amount at each [Ca2+]e. To explain the different proton-induced shifts in voltage dependence of activation and deactivation by the neutralization of surface charges it has to be assumed that this effect is state dependent, in the way that due to structural rearrangements during gating, the channel structure present different substrates for protons and/or different sensitivity of the voltage sensors (or gating machinery) to proton effects (Hille, 2001).

The standard surface-potential theory also predicts that the effects of increasing extracellular proton and Ca2+ concentrations on the positive shift of the gating parameters are additive (Kwan and Kass, 1993). Our observation that protons shift the activation curve to more positive potentials in the absence of extracellular Ca2+ (Fig. 4, A and B) is therefore incompatible with neutralization of surface charges being the sole mechanism responsible for the shift of activation kinetics. Delisle and Satin (2000) proposed that the reduced voltage sensitivity of activation is due to a proton-induced slowing of voltage dependent transitions distally to channel opening. We have extended this idea to explain our results. First, we propose that protons not only decrease the voltage sensitivity of the activation (increase sact) but also shift the voltage of half-maximal activation to more positive potentials independently of the neutralization of surface charges. Second, we postulate that Ca2+ inhibits these effects of protons on activation. On the other hand, in accord with Delisle and Satin (2000), we consider that the transition determining macroscopic deactivation is only modulated by the neutralization of negative surface charges.

We have reported previously that aspartate-to-glutamate mutations in the EEED pore locus of α1G induces changes in the activation curve of the channel (Talavera et al., 2001). The present result demonstrate that the EEED mutant shows alterations in the deactivation process, with less negative Vτdeac values and smaller voltage sensitivity for τdeac than the wild-type channel. These and other gating modifications induced by pore mutations are discussed in the accompanying paper (Talavera et al., 2003, in this issue).

Activation-inactivation Coupling and Proton Effects on Channel Gating

The voltage dependence of the inactivation of T-type Ca2+ channels arises from voltage-dependent transitions occurring during channel activation (Droogmans and Nilius, 1989; Chen and Hess, 1990; Serrano et al., 1999; Burgess et al., 2002). Since protons modify the activation process it is interesting to study the possible correlation between the proton-induced changes in the parameters describing steady-state activation and inactivation. We have found that if pHe is changed in the range of 9.1 to 6.2, for each 1 mV shift in Vact there was a 0.4–0.5 mV shift in Vinac in both 2 and 20 mM Ca2+. A similar correlation factor was calculated from the data of Delisle and Satin (2000) for α1H (their Figs. 1 D and 2 C). On the other hand, there was no strict correlation between the voltage sensitivity of inactivation (sinac) and activation (sact) and between sinac and Vinac, in contrast with that observed between sact and Vact (see Fig. 7). Thus, although proton-induced changes in the inactivation and activation processes seem to be linked, their degree of correlation is not absolute. As we discuss below, a kinetic model that includes state-dependent proton effects can explain these results.

A Kinetic Model Predicts State-dependent Effects of Extracellular Protons and Ca2+

Currently, there are several kinetic models of T-type channel gating (Droogmans and Nilius, 1989; Chen and Hess, 1990; Serrano et al., 1999; Burgess et al., 2002). The model of Burgess et al. (see Fig. 13) is particularly attractive since it accounts for the properties of the α1G and α1H channels in nearly physiological conditions and includes for the first time an explicit description of the gating charges associated with channel transitions. We adopted this model and determined which parameters had to be modified in order to describe our experimental data in the different pHe in the presence of 2 or 20 mM Ca2+. First, we readjusted the parameters of the original model to describe our own experimental data at pHe 7.4 in 2 mM Ca2+ (see Table I). We then considered as working hypothesis that the modifications of gating induced by H+ and Ca2+ are due to: (a) the neutralization of surface charges, (b) state-dependent alterations of the electric field sensed by the gating charges, and (c) the modification of the gating charges. To evaluate these effects independently from each other we used the following approach. All voltage-dependent transitions, with the exception of the deactivation transitions (O → C3 and IO → I3), were rewritten as:

\begin{eqnarray*}&&k_{C1C2}=K_{C1C2}\mathrm{exp}\left[\frac{\mathrm{{\delta}}_{1}q_{1}}{T}\left(V{-}V_{Shift}\right)\right]\mathrm{{;}}\\&&\ \\&&k_{C2C1}=K_{C2C1}\mathrm{exp}\left[{-}\frac{\left(1{-}\mathrm{{\delta}}_{1}\right)q_{1}}{T}\left(V{-}V_{Shift}\right)\right]\end{eqnarray*}
\begin{eqnarray*}&&k_{I1I2}=K_{I1I2}\mathrm{exp}\left[\frac{\mathrm{{\delta}}_{1}q_{1}}{T}\left(V{-}V_{Shift}\right)\right]\mathrm{{;}}\\&&\ k_{I2I1}=K_{I2I1}\mathrm{exp}\left[{-}\frac{\left(1{-}\mathrm{{\delta}}_{1}\right)q_{1}}{T}\left(V{-}V_{Shift}\right)\right]\end{eqnarray*}
\begin{eqnarray*}&&k_{C2C3}=K_{C2C3}\mathrm{exp}\left[\frac{\mathrm{{\delta}}_{2}q_{2}}{T}\left(V{-}V_{Shift}{-}V_{O2}\right)\right]\mathrm{{;}}\\&&\ k_{C3C2}=K_{C3C2}\mathrm{exp}\left[{-}\frac{\left(1{-}\mathrm{{\delta}}_{2}\right)q_{2}}{T}\left(V{-}V_{Shift}{-}V_{O2}\right)\right]\end{eqnarray*}
\begin{eqnarray*}&&k_{I2I3}=K_{I2I3}\mathrm{exp}\left[\frac{\mathrm{{\delta}}_{2}q_{2}}{T}\left(V{-}V_{Shift}{-}V_{O2}\right)\right]\mathrm{{;}}\\&&\ k_{I3I2}=K_{I3I2}\mathrm{exp}\left[{-}\frac{\left(1{-}\mathrm{{\delta}}_{2}\right)q_{2}}{T}\left(V{-}V_{Shift}{-}V_{O2}\right)\right]\mathrm{,}\end{eqnarray*}

where T = 25.4 mV is the thermal energy in electron-volts, q1 and q2 are the gating charges associated with the first and second boxes of the kinetic scheme, and δ1 and δ2 account for the coupling between the local electric potential sensed by q1 and q2 and the membrane potential V. We introduced a voltage offset (VShift) in all these rate constants to describe the neutralization of surface charges. To account for the state-dependent modification of the electric field sensed by the gating charges, we included the voltage offset VO2 in the second box of the kinetic scheme (see Fig. 13). Starting from the set of parameters obtained at pHe 7.4 and 2 mM Ca2+ (Table I), we adapted the parameters VShift, VO2, q1, q2, and kC3O to fit the average activation, steady-state inactivation, reactivation curves and the voltage dependence of the time-to-peak obtained in each experimental condition. The variation of δ1 and δ2 did not significantly affect the goodness of the fits and thus they were kept fixed to the values obtained in the reference condition. We introduced an extra voltage offset in order to account for the neutralization of surface charges by Ca2+ to fit the data obtained in 20 mM Ca2+. This value was set to 20.5 mV and taken from the shift of the voltage of half-maximal activation by changing the [Ca2+]e from 2 to 20 mM at pHe 9.1 (see Fig. 4 B).

The deactivation transitions O → C3 and IO → I3 were described by an equation of the form:

\(k_{OC3}=k_{IOI3}=\mathrm{exp}\left[{{-}\left(V{-}V\mathrm{{\tau}}_{deac}\right)}/{s\mathrm{{\tau}}_{deac}}\right]\)
⁠, where Vτdeac (= T/q3 ln(KOC3)) and sτdeac (= T/q3) are the experimental values at each pHe and [Ca2+]e (e.g., Fig. 12). The parameter VShift is not included in the expressions of kOC3 and kIOI3 because the effect of neutralization of surface charges is implicitly considered in the values of Vτdeac.

Fig. 14 shows that the simulated curves of steady-state activation and inactivation reproduce the experimental data in the pHe range of 9.1 to 6.2–5.5 and in the presence of 2 or 20 mM Ca2+. The parameters optimized with the fitting procedure are shown as functions of pHe and Ca2+ in Fig. 15. VShift, which describes the proton-induced voltage shift due to the neutralization of surface charges, is equal to −5.5 mV at pHe 9.1 and gradually changes to positive values at acid pHe in 2 mM Ca2+ (panel A). Panel B exemplifies that this causes an identical displacement of all the gating kinetics to less negative potentials. Fig. 15 C shows that VO2 is around −4 mV at alkaline pHe but changes to positive values at acid pHe in 2 mM Ca2+. This induces an additional positive shift of the inactivation and activation curves and a slight increase in the steepness of the latter one (panel B). The relevance of the introduction of VO2 in the model becomes clear when noticing that the voltage shift attributable to the screening of surface charges when changing the pHe from 9.1 to 6.2 (VShift(6.2) − VShift(9.1) ≈ 19 mV, Fig. 15 A) equals the shift observed in the voltage dependence of the time constant of deactivation (Fig. 12 C). In other words, VO2 allows dissecting the effect of pure screening of surface charges from specific proton effects on channel activation. The specificity of this voltage offset on the transitions between the states C2 and C3 may be due to an enhanced accessibility or/and an increased affinity of superficial binding sites for protons during these transitions. Another possibility is that protonation induces an uncoupling between the voltage sensors and the activation gates of the channel. The appropriate fit of the data requires the reduction of the gating charge. However, the gating charge associated with the second step of activation (q2) shows a much stronger pHe dependence than that associated with the first step (q1). In 2 mM Ca2+ for example, q2 shows a marked reduction in the whole pHe range studied, whereas q1 does not change in the pHe range from 9.1 to 6.8 (Fig. 15 E). As shown in panel D, the predicted decrease in q1 does not have large effect on the steady-state properties in contrast to the reduction of q2, which induces a decrease in the slope of activation and inactivation curves (panel F). The decrease in q2 shifts the inactivation curve to more negative potentials in a larger extent than the activation curve. This is, therefore, based on the application of the present gating model, the reason for the uneven effects of extracellular protons on activation and inactivation properties of α1G (Fig. 7). The fit of the voltage dependence of the time-to-peak compelled a decrease of the rate of the last step of the activation (kC3O and kI3O) in the ranges from 4.6 to 1.9 ms−1 in 2 mM Ca2+ and from 2.4 to 1.2 ms−1 in 20 mM Ca2+. This produces a positive shift of the activation and inactivation curves and an increase of the steepness of the activation curve (compare dotted and short dashed lines in Fig. 15 F).

The results of the simulation of the data obtained in 20 mM Ca2+ indicate that predicted effects of protons in this condition are much weaker than in 2 mM Ca2+. We conclude that calcium ions antagonize the effects of extracellular protons on the gating properties of α1G in two ways: first, by competing for the binding to negative surface charges and, second, by controlling the proton-induced effects on an intermediate step of the activation sequence. In the first mechanism Ca2+ and protons have equivalent effects, since both ions neutralize negative surface charges and shift all voltage-dependent processes of the channel toward positive potentials. The inhibition of channel activation by the binding of protons through the second mechanism is weakened by the presence of Ca2+, which could be due to a decreased proton affinity or a modulation of the effects downstream of proton binding. This may explain why the increase of [Ca2+]e prevents the decreased voltage sensitivity, the “extra” voltage shift of the activation induced by protons that is observed in Ca2+ free conditions (Fig. 4) and the smaller proton-induced shift of the activation curves with respect to that of the voltage dependence of the deactivation kinetics. The next section argues for the selectivity filter as a target for protonation, but, as demonstrated in the accompanying paper, the binding of protons to the EEDD pore locus is not related to the proton-induced modification of channel activation.

Proton Effects on Permeation and Selectivity of α1G Are Dependent on Ca2+, Na+, and the Structure of the Selectivity Filter

Tytgat et al. (1990) reported that extracellular acidification reduces the conductance of single cardiac T-type channels in the guinea pig. However, Delisle and Satin (2000) concluded that protons do not block the α1H channel, but actually increase the whole-cell conductance and decrease the Ca2+/monovalent cation selectivity. These discrepant results might be reconciled by the different extracellular Ca2+ and Na+ concentrations that were used in these reports. In fact, it has been shown that ion conduction through Ca2+ channels depends on both Ca2+ and Na+ concentrations (Polo-Parada and Korn, 1997) and that Na+ can contribute significantly to the current through T-type channels (Lee et al., 1999; Serrano et al., 1999). Our data about the effects of extracellular Ca2+ and Na+ on the open pore block and the changes in ion selectivity induced by protons in α1G and the EEED pore mutant support this contention.

We have found that in α1G, extracellular acidification from pH 9.1 to 6.2 decreased the amplitude of inward tail currents by ∼33% in the presence of 2 mM Ca2+ (150 NMDG+), whereas the block was maximal 10% in 20 mM Ca2+, indicating that open-pore proton block is larger at low [Ca2+]e. Analogue experiments with the EEED pore mutant confirmed that open-pore proton block is Ca2+ dependent and showed that proton block of Ca2+ conduction (20 mM Ca2+) was stronger than in the wild-type channel.

The reversal potentials were less positive in the EEED mutant than in the wild-type channel under all conditions, in agreement with our previous findings (Talavera et al., 2001). The fact that this was observed at pHe 9.1 may indicate that the EEED mutant has a decreased Ca2+ selectivity respect to α1G, independently from the channel protonation. This hypothesis would mean that the exchange of the aspartate residue D1487 for glutamate disrupts the coordination of Ca2+ binding to the selectivity filter in contrast to the results in L-type Ca2+ channels (Ellinor et al., 1995). A second possibility is that some protonation occurs at nanomolar extracellular proton concentrations, implying a much higher proton affinity of the EEED arrangement in the α1G template than in that of the L-type channel (Chen and Tsien, 1997). This hypothesis fits nicely with previous evidence suggesting that an EEEE selectivity filter in the α1G channel impose a narrower ionic pathway than in HVA channels (Talavera et al., 2001), which is in turn in agreement with the fact that the pore of the wild-type T-type channels is narrower than that of HVA Ca2+ channels (Cataldi et al., 2002). In both α1G and the EEED mutant, the proton-induced decrease of the Ca2+/monovalent cation selectivity (judged by the negative shift of the reversal potential) was observed in 2 but not in 20 mM Ca2+. For the EEED mutant, the negative shift of the reversal potential in 2 mM Ca2+ was not observed in the presence of Na+, probably because in this condition the channel selectivity for monovalent cations is so high that extracellular acidification does not further increase it, in contrast with the observations in the wild-type channel (Fig. 10 C).

Intriguingly, Delisle and Satin (2000) found that extracellular acidification increased the macroscopic conductance of the T-type channel α1H. Suspecting that this was due to the presence of extracellular Na+, we also addressed the effects of sodium ions on proton block and permeation through α1G. We found that the proton block in the presence of 150 mM Na+ and 2 mM Ca2+ was significantly reduced at negative potentials compared with that in 150 mM NMDG+, suggesting that channel protonation increased the Na+ permeation through the channel (Fig. 8). Interestingly, the substitution of NMDG+ by Na+ did not significantly change the reversal potential of α1G, in contrast with the significant shift to positive potentials of Vr in the EEED mutant at pHe 9.1 and 6.2 (Fig. 10). On the other hand, the comparison of the amplitude of the tail currents recorded in the presence of NMDG+ or Na+ (Fig. 11) indicates that under alkaline conditions extracellular Na+ blocks inward current through a1G, whereas Na+ appears to contribute to the inward current in acid extracellular medium. These observations are in agreement with the previous finding that variations of extracellular Na+ do not change the basal T-type current in cardiac cells at pHe 7.4 (Alvarez et al., 2000). In the EEED mutant, substitution of NMDG+ by Na+ increased inward currents, in accord with the significant enhancement of Na+ selectivity deduced from the changes in reversal potentials.

Notably, the proton block of the Ca2+ permeation through α1G did not show voltage dependence and the same was the case for the effects of the substitution of NMDG+ by Na+ for both α1G and the EEED mutant. These results suggest that the site of protonation and Na+ binding senses very little of the membrane electric field. The voltage dependence of the proton block in the wild-type channel in the presence of extracellular Na+ and in the EEED mutant are likely to be related to the proton-induced increase of the monovalent conduction at very negative potentials.

The evidence gathered from the L-type Ca2+ channel indicate that channel protonation controlling permeation properties occurs at the selectivity filter of this channel, i.e., the carboxilates of the EEEE pore locus (Chen et al., 1996; Klockner et al., 1996; Chen and Tsien, 1997; Varadi et al., 1999), rather than at a site modulating channel conductance via an allosteric mechanism (Prod'hom et al., 1989; Kuo and Hess, 1993). Taken together, we can conclude that in T-type Ca2+ channels, (a) the open pore block of the Ca2+ conduction and the decrease of the Ca2+/monovalent cations selectivity induced by protons depend on the [Ca2+]e, (b) at low pHe monovalent cations contribute significantly to the conduction, (c) protonation leading to a modification of conduction occurs at the EEDD pore locus, and (d) the EEDD pore locus is located close to the extracellular face of the channel. We propose that protonation of glutamate and/or aspartate residues neutralizes part of the negative charge of the selectivity filter and reduces the affinity for Ca2+, decreasing Ca2+/monovalent cation selectivity. It is reasonable to think that protonation of the selectivity filter of Ca2+ channels might mimic the structure of the selectivity filter of sodium channels, which is built by a far less electronegative DEKA locus.

Functional Significance of Proton–Ca2+ Interaction

We propose that protons and Ca2+ modulate T-type channel gating by two mechanisms. First, both ions shift the voltage-dependent kinetics by neutralizing negative surface charges; and second, Ca2+ prevents the proton-induced inhibition of channel activation. The potential patho-physiological implications of the proton modulation of T-type channels has been associated with the activity of the thalamocortical network (Shah et al., 2001), the pacemaker function (Tytgat et al., 1990), and the generation of arrhythmias (Delisle and Satin, 2000) in cardiac tissue. In this context the consequences of the interrelated modifications of the activation and the inactivation of T-type channels by extracellular protons are particularly interesting. Ischemic and hypoxic conditions induce both extracellular acidification and cell depolarization in cardiac and nervous tissues. Thus, the coupled shift of steady-state activation and inactivation in extracellular acidification results in a concomitant shift of the T-type window current along the voltage axis in the direction of the change of the basal membrane potential. This means that, although protons reduce channel activation, the Ca2+ entry through T-type channels could be preserved during extracellular acidification, which in turn might regulate cellular functions by constant injection of inward current and the increase in intracellular Ca2+ concentration. On the other hand, we can speculate that the uneven proton effects on activation and inactivation properties, as well as the proton-induced change in channel selectivity toward monovalent ions and the block of Ca2+ permeation could protect against harmful Ca2+ overload.

We thank Dr. F. Hofmann for the α1G clone and Dr. M. Staes for the help in the construction of the α1G pore mutant. We are also grateful to Drs. J.L. Alvarez, G. Vassort, T. Voets, and R. Vennekens for helpful discussions. The expert technical assistance of M. Crabbé, H. Van Weijenbergh, S. de Swaef, and M. Schuermans is greatly acknowledged. We also thank Professor V. Flockerzi for providing the vector pCAGGSM2.

This work was supported by the Belgian Federal Government, the Flemish Government, and the Onderzoeksraad KU Leuven (GOA 99/07, F.W.O. G.0237.95, F.W.O. G.0214.99, F.W.O. G. 0136.00; Interuniversity Poles of Attraction Program, Prime Ministers Office IUAP).

Olaf S. Andersen served as editor.

Karel Talavera's permanent address is Instituto de Cardiologia y Cirusia Cardiovascular, 10400 Lattabana, Cuba.

*

Abbreviation used in this paper: HVA, high voltage–activated.

Alvarez, J.L., A. Artiles, K. Talavera, and G. Vassort.
2000
. Modulation of voltage-dependent facilitation of the T-type calcium current by sodium ion in isolated frog atrial cells.
Pflugers Arch.
441
:
39
–48.
Burgess, D.E., O. Crawford, B.P. Delisle, and J. Satin.
2002
. Mechanism of inactivation gating of human T-type (low-voltage activated) calcium channels.
Biophys. J.
82
:
1894
–1906.
Carbone, E., and H.D. Lux.
1987
. Single low-voltage-activated calcium channels in chick and rat sensory neurones.
J. Physiol.
386
:
571
–601.
Cataldi, M., E. Perez-Reyes, and R.W. Tsien.
2002
. Differences in apparent pore sizes of low and high voltage-activated Ca2+ channels.
J. Biol. Chem.
277
:
45969
–45976.
Chen, C.F., and P. Hess.
1990
. Mechanism of gating of T-type calcium channels.
J. Gen. Physiol.
96
:
603
–630.
Chen, X.H., I. Bezprozvanny, and R.W. Tsien.
1996
. Molecular basis of proton block of L-type Ca2+ channels.
J. Gen. Physiol.
108
:
363
–374.
Chen, X.H., and R.W. Tsien.
1997
. Aspartate substitutions establish the concerted action of P-region glutamates in repeats I and III in forming the protonation site of L-type Ca2+ channels.
J. Biol. Chem.
272
:
30002
–30008.
Delisle, B.P., and J. Satin.
2000
. pH modification of human T-type calcium channel gating.
Biophys. J.
78
:
1895
–1905.
Droogmans, G., and B. Nilius.
1989
. Kinetic properties of the cardiac T-type calcium channel in the guinea- pig.
J. Physiol.
419
:
627
–650.
Ellinor, P.T., J. Yang, W.A. Sather, J.F. Zhang, and R.W. Tsien.
1995
. Ca2+ channel selectivity at a single locus for high-affinity Ca2+ interactions.
Neuron.
15
:
1121
–1132.
Frankenhaeuser, B., and A.L. Hodgkin.
1957
. The action of calcium on the electrical properties of squid axons.
J. Physiol.
137
:
218
–244.
Hille, B. 2001. Ion Channels of Excitable Membranes. 3rd edition. Sinauer Associates, Inc., Sunderland, MA. 646–662.
Klockner, U., G. Mikala, A. Schwartz, and G. Varadi.
1996
. Molecular studies of the asymmetric pore structure of the human cardiac voltage-dependent Ca2+ channel. Conserved residue, Glu-1086, regulates proton-dependent ion permeation.
J. Biol. Chem.
271
:
22293
–22296.
Klugbauer, N., E. Marais, L. Lacinova, and F. Hofmann.
1999
. A T-type calcium channel from mouse brain.
Pflugers Arch.
437
:
710
–715.
Kuo, C.C., and P. Hess.
1993
. Characterization of the high-affinity Ca2+ binding sites in the L-type Ca2+ channel pore in rat phaeochromocytoma cells.
J. Physiol.
466
:
657
–682.
Kwan, Y.W., and R.S. Kass.
1993
. Interactions between H+ and Ca2+ near cardiac L-type calcium channels: evidence for independent channel-associated binding sites.
Biophys. J.
65
:
1188
–1195.
Lacinova, L., N. Klugbauer, and F. Hofmann.
2002
. Gating of the expressed Cav3.1 calcium channel.
FEBS Lett.
531
:
235
–240.
Lee, J.H., J.C. Gomora, L.L. Cribbs, and E. Perez-Reyes.
1999
. Nickel block of three cloned T-type calcium channels: low concentrations selectively block alpha1H.
Biophys. J.
77
:
3034
–3042.
Nilius, B. 1992. T-type calcium channels in cardiac muscle: News in kinetics and modulation. Intracellular Regulation of Ion Channels. Vol. H 60. Z. Morad, editor. Springer-Verlag, Berlin. 181–189.
Pietrobon, D., B. Prod'hom, and P. Hess.
1989
. Interactions of protons with single open L-type calcium channels. pH dependence of proton-induced current fluctuations with Cs+, K+, and Na+ as permeant ions.
J. Gen. Physiol.
94
:
1
–21.
Polo-Parada, L., and S.J. Korn.
1997
. Block of N-type calcium channels in chick sensory neurons by external sodium.
J. Gen. Physiol.
109
:
693
–702.
Prod'hom, B., D. Pietrobon, and P. Hess.
1989
. Interactions of protons with single open L-type calcium channels. Location of protonation site and dependence of proton-induced current fluctuations on concentration and species of permeant ion.
J. Gen. Physiol.
94
:
23
–42.
Serrano, J.R., S.R. Dashti, E. Perez-Reyes, and S.W. Jones.
2000
. Mg2+ block unmasks Ca2+/Ba2+ selectivity of alpha1G T-type calcium channels.
Biophys. J.
79
:
3052
–3062.
Serrano, J.R., E. Perez-Reyes, and S.W. Jones.
1999
. State-dependent inactivation of the alpha1G T-type calcium channel.
J. Gen. Physiol.
114
:
185
–201.
Shah, M.J., S. Meis, T. Munsch, and H.C. Pape.
2001
. Modulation by extracellular pH of low- and high-voltage-activated calcium currents of rat thalamic relay neurons.
J. Neurophysiol.
85
:
1051
–1058.
Shuba, Y.M., V.I. Teslenko, A.N. Savchenko, and N.H. Pogorelaya.
1991
. The effect of permeant ions on single calcium channel activation in mouse neuroblastoma cells: ion-channel interaction.
J. Physiol.
443
:
25
–44.
Smirnov, S.V., G.A. Knock, A.E. Belevych, and P.I. Aaronson.
2000
. Mechanism of effect of extracellular pH on L-type Ca2+ channel currents in human mesenteric arterial cells.
Am. J. Physiol. Heart Circ. Physiol.
279
:
H76
–H85.
Staes, M., K. Talavera, N. Klugbauer, J. Prenen, L. Lacinova, G. Droogmans, F. Hofmann, and B. Nilius.
2001
. The amino side of the C-terminus determines fast inactivation of the T- type calcium channel alpha1G.
J. Physiol.
530
:
35
–45.
Talavera, K., J. Alvarez, and G. Vassort.
1998
. A model of Ca ion permeation through the cardiac T-type Ca-channel.
Biophys. J.
74
:
A108
.
Talavera, K., M. Staes, A. Janssens, N. Klugbauer, G. Droogmans, F. Hofmann, and B. Nilius.
2001
. Aspartate residues of the Glu-Glu-Asp-Asp (EEDD) pore locus control selectivity and permeation of the T-type Ca2+ channel alpha(1G).
J. Biol. Chem.
276
:
45628
–45635.
Talavera, K., A. Janssens, N. Klugbauer. G. Droogmans, and B. Nilius.
2003
. Pore structure influences gating properties of the T-type Ca2+ channel α1G.
J. Gen Physiol.
121
:
529
–540.
Tombaugh, G.C., and G.G. Somjen.
1996
. Effects of extracellular pH on voltage-gated Na+, K+ and Ca2+ currents in isolated rat CA1 neurons.
J. Physiol.
493
:
719
–732.
Tytgat, J., B. Nilius, and E. Carmeliet.
1990
. Modulation of the T-type cardiac Ca channel by changes in proton concentration.
J. Gen. Physiol.
96
:
973
–990.
Varadi, G., M. Strobeck, S. Koch, L. Caglioti, C. Zucchi, and G. Palyi.
1999
. Molecular elements of ion permeation and selectivity within calcium channels.
Crit. Rev. Biochem. Mol. Biol.
34
:
181
–214.
Woodhull, A.M.
1973
. Ionic blockage of sodium channels in nerve.
J. Gen. Physiol.
61
:
687
–708.
Zhou, W., and S.W. Jones.
1996
. The effects of external pH on calcium channel currents in bullfrog sympathetic neurons.
Biophys. J.
70
:
1326
–1334.