Interaction via a key tryptophan in the I-II linker of N-type calcium channels is required for beta1 but not for palmitoylated beta2, implicating an additional binding site in the regulation of channel voltage-dependent properties

Abstract

The CaVbeta subunits of voltage-gated calcium channels regulate these channels in several ways. Here we investigate the role of these auxiliary subunits in the expression of functional N-type channels at the plasma membrane and in the modulation by G-protein-coupled receptors of this neuronal channel. To do so, we mutated tryptophan 391 to an alanine within the alpha-interacting domain (AID) in the I-II linker of CaV2.2. We showed that the mutation W391 virtually abolishes the binding of CaVbeta1b and CaVbeta2a to the CaV2.2 I-II linker and strongly reduced current density and cell surface expression of both CaV2.2/alpha2delta-2/beta1b and/beta2a channels. When associated with CaVbeta1b, the W391A mutation also prevented the CaVbeta1b-mediated hyperpolarization of CaV2.2 channel activation and steady-state inactivation. However, the mutated CaV2.2W391A/beta1b channels were still inhibited to a similar extent by activation of the D2 dopamine receptor with the agonist quinpirole. Nevertheless, key hallmarks of G-protein modulation of N-type currents, such as slowed activation kinetics and prepulse facilitation, were not observed for the mutated channel. In contrast, CaVbeta2a was still able to completely modulate the biophysical properties of CaV2.2W391A channel and allow voltage-dependent G-protein modulation of CaV2.2W391A. Additional data suggest that the concentration of CaVbeta2a in the proximity of the channel is enhanced independently of its binding to the AID by its palmitoylation. This is essentially sufficient for all of the functional effects of CaVbeta2a, which may occur via a second lower-affinity binding site, except trafficking the channel to the plasma membrane, which requires interaction with the AID region.

Figure 1.

The W391A mutation prevents the binding of Ca_Vβ1b and Ca_Vβ2a to Ca_V2.2 I-II linker. A, Representation of the Ca_V2.2 subunit that is composed of four domains of six transmembrane segments each. The I-II linker contains an 18 amino acid domain (AID) that interacts with Ca_Vβ subunits (dotted box). The sequence of the AID within Ca_V2.2 and Ca_V2.2W391A are given below, with W391 underlined and also showing the putative overlapping binding motif for Gβγ dimer binding. B, Representative Biacore sensorgrams showing interactions between Ca_Vβ1b (top) or Ca_Vβ2a (bottom) with I-II loop GST-fusion proteins from Ca_V2.2 (left) and Ca_V2.2W391A (right). The auxiliary Ca_Vβ subunits (25-200 nm as indicated) were applied during the time indicated by the bars.

Figure 2.

Role of Ca_Vβ subunits in plasma membrane expression of Ca_V2.2. A, Left, I-V relationships for Ca_V2.2/α2δ-2 coexpressed with Ca_Vβ1b (filled circles, left; n = 18) or Ca_Vβ2a (diamonds, right; n = 13) or without a Ca_Vβ subunit (filled squares, left; n = 13) compared with I-V relationships for Ca_V2.2W391A/α2δ-2 coexpressed with Ca_Vβ1b (filled stars, left; n = 11) or Ca_Vβ2a (filled triangles, right; n = 14). The mean data are fitted with a modified Boltzmann function (see Materials and Methods), the V_{50, act} and G_max values of which are given in Table 1. Typical Ba²⁺ current traces at +20 mV (identified by the symbols used) are shown above the I-V relationships. Right, Mean current density at +20 mV for each of these combinations ± SEM. B, Cell surface expression of either Ca_V2.2 or Ca_V2.2WA391A expressed with α2δ-2, either without Ca_Vβ or with Ca_Vβ1b (left) or with Ca_Vβ2a (right). Total expression of Ca_V2.2 is shown by Western blot in the top row and biotinylated Ca_V2.2 in the bottom row. Cells were transfected with empty vector (lane 1), Ca_V2.2/α2δ-2/β1b (lane 2), Ca_V2.2W391A/α2δ-2/β1b (lane 3), Ca_V2.2/α2δ-2 (lane 4), Ca_V2.2/α2δ-2/β2a (lane 5), or Ca_V2.2W391A/α2δ-2/β2a (lane 6). Note that only the cell surface expression is affected similarly by the W391A mutation or by the absence of Ca_Vβ. Right, Histogram showing quantification of the mean amount of Ca_V2.2WA391A expressed at the plasma membrane, when coexpressed with either β1b or β2a, or Ca_V2.2 without Ca_Vβ, given as a percentage of the amount of Ca_V2.2 expressed with the relevant Ca_Vβ present under the same conditions. Data are mean ± SEM of four independent experiments. MW, Molecular weight. C, Western blot illustrating the endogenous expression of Ca_Vβ3 in tsA-201 cells. Gel loaded with 2.5 μg of protein prepared from cells transfected with Ca_Vβ3 (lane 1) compared with 2.5 or 250 μg of protein from cells transfected with an empty pMT2 vector (lanes 2, 3). An anti-β3 monoclonal antibody was used for immunoblotting.

Figure 3.

Biophysical properties of Ca_V2.2 and Ca_V2.2W391A coexpressed with Ca_Vβ1b. A, Representative current traces to illustrate current activation. Ba²⁺ tail currents were recorded after repolarizing to -50 mV after a 20 ms test pulse to between -40 and +80 mV from a holding potential of -100 mV. Top, Ca_V2.2/β1b; middle, Ca_V2.2 without any Ca_Vβ; bottom, Ca_V2.2W391A/β1b, all coexpressed with α2δ-2. B, Voltage dependence of activation of Ca_V2.2/α2δ-2 coexpressed with Ca_Vβ1b (filled circles) or without any Ca_Vβ subunit (filled squares) or Ca_V2.2W391A/α2δ-2 expressed with Ca_Vβ1b (filled stars). The normalized data, obtained from recordings such as those shown in A, are plotted against the test pulse (n = 6-19). The mean data are fitted with a Boltzmann function, the V_{50, act} values of which are given in Table 1. C, Representative current traces (labeled as in A) to illustrate steady-state inactivation protocols. Inward Ba²⁺ currents were recorded after conditioning pulses of 5 s duration, applied from a holding potential of -100 mV in 10 mV steps between -110 and +30 mV, followed by a 50 ms test pulse to +20 mV. D, Voltage dependence of steady-state inactivation of Ca_V2.2/α2δ-2 coexpressed with Ca_Vβ1b (filled circles) or without any Ca_Vβ subunit (filled squares) or Ca_V2.2W391A/α2δ-2 expressed with Ca_Vβ1b (filled stars). The normalized data obtained from recordings such as those shown in C are plotted against the conditioning potentials (n = 6-19). The mean data are fitted with a Boltzmann function, the V_50,inact values of which are given in Table 1. E, Left, Superposition of representative current traces for the subunit combinations indicated, recorded during an 800 ms depolarizing step to +20 mV, from a holding potential of -100 mV, normalized to the peak current. Right, Mean time constants of inactivation (τ_inact) obtained by fitting the decaying phase of the Ba²⁺ currents at +20 mV with a single exponential, for Ca_V2.2/α2δ-2/β1b (black bar; n = 25), Ca_V2.2/α2δ-2 (white bar; n = 25), and Ca_V2.2W391A/α2δ-2/β1b (gray bar; n = 14).

Figure 4.

Biophysical properties of Ca_V2.2 and Ca_V2.2WA391A coexpressed with Ca_Vβ2a. A, Representative current traces to illustrate current activation using the same protocols as described in the legend to Figure 3. Top, Ca_V2.2/β2a; bottom, Ca_V2.2W391A/β2a, all coexpressed with α2δ-2. B, Voltage dependence of activation for Ca_V2.2/α2δ-2 coexpressed with Ca_Vβ2a (filled diamonds; n = 10) or Ca_V2.2W391A/α2δ-2 expressed with Ca_Vβ2a (filled triangles; n = 12). The normalized data obtained from recordings such as those shown in A are plotted against the test pulse. The mean data are fitted with a Boltzmann function, the V_{50, act} values of which are given in Table 1. C, Representative current traces to illustrate steady-state inactivation using the same protocols as described in the legend to Figure 3. Top, Ca_V2.2/β2a; bottom, Ca_V2.2W391A/β2a. D, Voltage dependence of steady-state inactivation for Ca_V2.2/α2δ-2 coexpressed with Ca_Vβ2a (filled diamonds; n = 17) or Ca_V2.2W391A/α2δ-2 expressed with Ca_Vβ2a (filled triangles; n = 18). The normalized data obtained from recordings such as those shown in C are plotted against the conditioning potentials. The mean data are fitted with a Boltzmann function, the V_{50, inact} values of which are given in Table 1. The dotted line represents the fit for Ca_V2.2 without Ca_Vβ from Figure 3D. E, Left, Superposition of representative current traces for the subunit combinations indicated, recorded during an 800 ms depolarizing step to +20 mV, from a holding potential of -100 mV and normalized to the peak current. Right, Normalized residual I_Ba at 600 ms, for Ca_V2.2/α2δ-2/β2a (black bar; n = 15) and Ca_V2.2W391A/α2δ-2/β2a (gray bar; n = 14).

Figure 5.

G-protein modulation of Ca_V2.2 and Ca_V2.2W391A expressed with Ca_Vβ1b. Top, The pulse protocol used is depicted. A 100 ms test pulse (P1) from -30 to +60 mV was applied from a holding potential of -100 mV. After 800 ms repolarization to -100 mV, a 100 ms prepulse to +100 mV was applied. The cell was repolarized for 20 ms to -100 mV, and a second pulse (P2) identical to the first one was applied. A, D, Typical current traces obtained with this protocol are represented for Ca_V2.2 (top) and Ca_V2.2W391A (bottom) coexpressed with Ca_Vβ1b. In A, the D₂ dopamine receptor is coexpressed, and the top current traces (depicted by the open symbols) are in the presence of the agonist quinpirole (100 nm). In D, Gβ1γ2 are coexpressed, and typical traces for Ca_V2.2/β1b and Ca_V2.2W391A/β1b are shown. B, I-V curves for the calcium channel combinations are shown, obtained before (filled symbols) and during (open symbols) application of 100 nm quinpirole. Top, I-V curves for Ca_V2.2/α2δ-2 coexpressed with Ca_Vβ1b (circles; n = 18) Bottom, I-V curves for Ca_V2.2W391A/α2δ-2 coexpressed with Ca_Vβ1b (stars; n = 11) are represented. I-V curves are fitted with modified Boltzmann functions, the V_{50, act} and G_max parameters of which are given in Table 1. E, Top, I-V curves for Ba²⁺ currents during P1 for Ca_V2.2/α2δ-2/β1b coexpressed without (filled circles; n = 6) or with (open triangles; n = 9) Gβ1γ2. Bottom, For Ca_V2.2W391A/α2δ-2/β1b coexpressed without (stars; n = 6) or with (open squares; n = 8) Gβ1γ2. I-V curves obtained from currents recorded during P2 when Gβ1γ2 was coexpressed with Ca_V2.2/α2δ-2/β1b (filled triangles; n = 9) or Ca_V2.2W391A/α2δ-2/β1b (filled squares; n = 8) are also represented. All data were obtained in parallel on the same experimental days. C, Voltage-dependent facilitation was calculated by dividing the peak current value obtained in P2 by that obtained in P1 at the potentials of 0, +10, and +20 mV, for Ca_V2.2/α2δ-2 with Ca_Vβ1b (black bars; n = 18), Ca_V2.2W391A/α2δ-2 with Ca_Vβ1b (white bars; n = 11) after application of quinpirole, or when Gβ1γ2 were coexpressed with Ca_V2.2/β1b (gray bars; n = 9), Ca_V2.2W391A/β1b (hatched bars; n = 8). F, Typical current traces at +20 mV for R52, 54ACa_V2.2W391A (top) coexpressed with Ca_Vβ1b before (filled circles) and after (open circles) activation of the D₂ dopamine receptor. Corresponding I-V curves are represented in the bottom.

Figure 6.

The kinetics of inactivation do not contaminate the properties of G-protein modulation of Ca_V2.2 and Ca_V2.2W391A. A, Typical current traces obtained for Ca_V2.2/α2δ-2 and Ca_V2.2W391A/α2δ-2 coexpressed with Ca_Vβ3 and the D₂ dopamine receptor are represented (top). I-V curves obtained for these combinations before (filled symbols) and during (100 nm) application of quinpirole (open symbols) are also shown (bottom). B, Voltage-dependent facilitation for Ca_V2.2/α2δ-2 (black bars; n = 7) or Ca_V2.2W391A/α2δ-2 coexpressed with Ca_Vβ3 (white bars; n = 7) obtained from the P2/P1 ratio when the D₂ dopamine receptor was activated by 100 nm quinpirole. C, Facilitation rate of G-protein-modulated channels. The duration of the prepulse was increased from 0 to 200 ms. The P2/P1 facilitation ratios are given for each prepulse, for Ca_V2.2 (open circles; n = 16) and Ca_V2.2W391A (open stars; n = 17) coexpressed with Ca_Vβ1b. Data are fitted with a single exponential, the time constant (τ_facil) of which is given in Results.

Figure 7.

G-protein modulation of the kinetics and voltage dependence of activation of Ca_V2.2 and Ca_V2.2W391A coexpressed with β1b. A, The τ_act for currents recorded in P1 as described in Figure 5 before and during perfusion of quinpirole in cells transfected with Ca_V2.2/α2δ-2/Ca_Vβ1b (white and black bars, respectively; n = 18) and with Ca_V2.2W391A/α2δ-2/β1b (gray and hatched bars, respectively; n = 11). For Ca_V2.2/α2δ-2/Ca_Vβ1b in the presence of quinpirole, the data were fit by a double exponential with a fast τ_act and a slow τ_act, the percentage of each component being given above the bars. B, Activation curves derived from tail current amplitude measurements for Ca_V2.2/α2δ-2/β1b (left) and Ca_V2.2W391A/α2δ-2/β1b (right). Top, Typical tail current traces recorded after a test pulse to +80 mV before (filled symbols) and during (open symbols) application of 100 nm quinpirole. Bottom, Peak tail current density, before (filled symbols) and during (open symbols) application of quinpirole, for Ca_V2.2/α2δ-2/β1b (circles, left) or Ca_V2.2W391A/α2δ-2/β1b (stars, right). Data are the mean ± SEM of 12-19 cells, and the solid lines are Boltzmann functions fits, the parameters of which are given in Results.

Figure 8.

G-protein modulation of the kinetics and voltage dependence of activation of Ca_V2.2 and Ca_V2.2W391A coexpressed with Ca_Vβ2a. A, The effect of quinpirole (100 nm) is compared on the Ca_V2.2/α2δ-2/β2a (left) and Ca_V2.2W391A/α2δ-2/β2a (right) combinations. Representative currents traces using the P1/P2 pulse protocol shown in Figure 5, obtained before (Con) and during (Quin) application of 100 nm quinpirole. B, The P2/P1 facilitation ratios for Ca_V2.2/α2δ-2/β2a (black bars; n = 13) or Ca_V2.2W391A/α2δ-2/β2a (white bars; n = 14) at potentials between 0 and +20 mV are given. C, The τ_act values for currents activated by +20 mV steps before and during application of quinpirole. Values are given for Ca_V2.2/α2δ-2/β2a in control conditions (white bars) and in the presence of quinpirole (black bars; n = 13) or for the Ca_V2.2W391A (gray and hatched bars, respectively; n = 14). In the presence of quinpirole, the data were fit by two exponentials, with the percentage of the total represented by each component given above the bar. D, Activation derived from normalized tail current density before (filled symbols) and during (open symbols) application of quinpirole (100 nm) for the subunit combination Ca_V2.2/α2δ-2/β2a (left, diamonds; n = 10) or Ca_V2.2W391A/α2δ-2/β2a (right, triangles; n = 12). The solid lines are Boltzmann function fits to the mean data, the parameters of which are given in Results.

Figure 9.

Effect of palmitoylation of Ca_Vβ subunits on the modulation of Ca_V2.2 and Ca_V2.2W391A channels. A, Voltage dependence of activation of Ca_V2.2 (filled symbols) or Ca_V2.2W391A (open symbols) with either Ca_Vβ2aC3,4S (triangles) or Ca_Vβ2a-β1b chimera (pentagons) and α2δ-2. The solid lines are Boltzmann function fits to the mean data, the parameters of which are given in Results. B, Voltage dependence of steady-state inactivation of Ca_V2.2 (filled symbols) or Ca_V2.2W391A (open symbols) with either Ca_Vβ2aC3,4S (triangles) or Ca_Vβ2a-β1b chimera (pentagons) and α2δ-2. The solid lines are Boltzmann function fits to the mean data, the parameters of which are given in Results. C, Left, Representative Ba²⁺ current traces recorded during an 800 ms depolarizing step to +20 mV from a holding potential of -100 mV. Right, The normalized amount of residual current at 600 ms, obtained for Ca_V2.2 or Ca_V2.2W391A coexpressed with α2δ-2 and Ca_Vβ2aC3,4S (black and white bars, respectively; n = 18 for both) or the Ca_Vβ2a-β1b chimera (gray and hatched bars; n = 23 and n = 19, respectively). D, Left, The effect of quinpirole (100 nm) is compared for the subunit combinations Ca_V2.2W391A/α2δ-2/β2aC3,4S (top traces) and Ca_V2.2W391A/α2δ-2/β2a-β1b chimera (bottom traces). Representative currents traces using the P1/P2 pulse protocol shown in Figure 5, obtained before (Con) and during (Quin) application of 100 nm quinpirole. Right, The P2/P1 ratio obtained from such traces for Ca_V2.2 and Ca_V2.2W391A coexpressed with α2δ-2 and Ca_Vβ2aC3,4S (black and white bars; n = 7 and n = 8, respectively) or the Ca_Vβ2a-β1b chimera (gray and hatched bars; n = 12 and n = 11 respectively).

Figure 10.

Proposed mechanism for G-protein modulation of Ca_V2.2 calcium channels and the effect of palmitoylation on the interaction of Ca_Vβ subunits with the I-II linker. A, Confocal immunofluorescent images of β2a and β2aC3,4S subunits expressed in tsA-201 cells, using an anti-β2 antibody (green). Nuclei are visualized with DAPI (blue). Scale bars, 20 μm. Note the membrane localization of the β2a subunit (left), whereas the β2aC3,4S subunit is localized to the cytoplasm (right). B, Model for the mechanism of action of Gβγ binding to the channel to inhibit its activity. The N terminus of Ca_V2.2 containing two arginines (R52, R54) constitutes a motif implicated in the inhibition by G-proteins, whereas the voltage-dependent facilitation (loss of inhibition induced by strong depolarization) requires a bound Ca_Vβ subunit on the I-II linker. C, The GK domain of Ca_Vβ subunits interacts with the AID, allowing a low-affinity binding of its SH3 domain elsewhere on the channel (e.g., to the I-II loop of Ca_V2.2) (Maltez et al., 2005) to modulate the properties of the channel. Palmitoylation of Ca_Vβ by anchoring the subunit to the plasma membrane allows this low-affinity interaction to occur when the W391 in the AID is mutated to disrupt its interaction with the GK domain.