QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

HOME  |   SEARCH  |   ARCHIVE  |   SUBSCRIBE  |   CONTACT  |   HELP

The Journal of Neuroscience, July 27, 2005, 25(30):6984-6996; doi:10.1523/JNEUROSCI.1137-05.2005

This Article Abstract Full Text (PDF) A correction has been published Submit an eLetter Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Leroy, J. Articles by Dolphin, A. C. Search for Related Content PubMed PubMed Citation Articles by Leroy, J. Articles by Dolphin, A. C.

 Previous Article  |  Next Article 

Cellular/Molecular
Interaction via a Key Tryptophan in the I-II Linker of N-Type Calcium Channels Is Required for 1 But Not for Palmitoylated 2, Implicating an Additional Binding Site in the Regulation of Channel Voltage-Dependent Properties

Jérôme Leroy,¹ Mark S. Richards,^1,2 * Adrian J. Butcher,^1 * Manuela Nieto-Rostro,¹ Wendy S. Pratt,¹ Anthony Davies,¹ and Annette C. Dolphin¹

¹Laboratory of Cellular and Molecular Neuroscience, Department of Pharmacology, University College London, London WC1E 6BT, United Kingdom, and ²School of Crystallography, Birkbeck College, London WC1E 7HX, United Kingdom

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The Ca_V 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 -interacting domain (AID) in the I-II linker of Ca_V2.2. We showed that the mutation W391 virtually abolishes the binding of Ca_V1b and Ca_V2a to the Ca_V2.2 I-II linker and strongly reduced current density and cell surface expression of both Ca_V2.2/2-2/1b and/2a channels. When associated with Ca_V1b, the W391A mutation also prevented the Ca_V1b-mediated hyperpolarization of Ca_V2.2 channel activation and steady-state inactivation. However, the mutated Ca_V2.2W391A/1b 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, Ca_V2a was still able to completely modulate the biophysical properties of Ca_V2.2W391A channel and allow voltage-dependent G-protein modulation of Ca_V2.2W391A. Additional data suggest that the concentration of Ca_V2a 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 Ca_V2a, 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.

Key words: calcium channel; neuron; -interaction domain; subunit; trafficking; G-protein; palmitoylation

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Voltage-gated calcium (Ca_V) channels play a major role in the physiology of excitable cells, particularly of neurons. Three families of voltage-gated calcium channels have been identified, Ca_V1-Ca_V3 (for review, see Ertel et al., 2000). The Ca_V1 class, L-type channels and the Ca_V2 class, non-L-type channels are both high-voltage-activated (HVA). These are heteromultimers composed of the pore-forming 1 subunit, associated with auxiliary Ca_V and 2 subunits (for review, see Catterall, 2000). The Ca_V2 calcium channels are inhibited by G dimers (Herlitze et al., 1996; Ikeda, 1996), which is the main mechanism of presynaptic inhibition by G-protein-coupled receptors. Ca_V subunits are crucial for normal HVA channel function (for review, see Dolphin, 2003a), because they enhance expression of functional channels at the plasma membrane, modulate their biophysical properties, and promote the voltage dependence of modulation of Ca_V2.2 calcium channels by G dimers, although the mechanism involved remains unclear (Bichet et al., 2000; Meir et al., 2000; Canti et al., 2001). G dimers and Ca_V subunits have been shown to bind to overlapping sites in the I-II loop of Ca_V2 channels, and they induce opposite effects on biophysical properties of the channels (for review, see Dolphin, 2003b), shifting channels between a willing and a reluctant state that requires large depolarizations to be opened (Bean, 1989). This observation led to the hypothesis that the binding of G dimers might dissociate the Ca_V subunit from the I-II loop of the 1 subunit (Sandoz et al., 2004) and that dissociation of Ca_V was the mechanism responsible for the inhibition observed. However, there is also evidence that G dimers do not displace Ca_V subunits but alter the orientation of the subunit with respect to the 1 subunit (Hummer et al., 2003).

Here, we investigated the role of Ca_V subunits in the plasma membrane expression and G-protein modulation of Ca_V2.2 calcium channels by mutating the tryptophan (W391) in the 1-interacting domain (AID) in the I-II loop of Ca_V2.2. This amino acid has been shown both by the study that identified the AID motif and the recent structural studies to be key to the interaction between Ca_V subunits and the AID (Pragnell et al., 1994; Chen et al., 2004; Opatowsky et al., 2004; Van Petegem et al., 2004). This mutation prevents the enhancement of functional expression of Ca_V2.2 by Ca_V1b and also prevents modulation of Ca_V2.2 by this subunit. In addition, although the G-protein modulation of Ca_V2.2W391A was present, it was not voltage dependent. In contrast, only the expression of the channel at the plasma membrane was affected when Ca_V2a was coexpressed with this mutant channel, whereas all of the biophysical properties of the expressed Ca_V2.2W391A channels were still normally modulated by Ca_V2a, including the voltage dependence of G-protein modulation. Our results further show that this was dependent on palmitoylation of Ca_V2a and suggest that the effect of Ca_V subunits on the voltage-dependent properties of the Ca_V2.2 channels occurs not only via the high-affinity I-II linker interaction but also via low-affinity interactions, presumably with other sites on the channel.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Materials. The cDNAs used in this study were Ca_V2.2 (D14157 [GenBank] ), Ca_V1b (X61394 [GenBank] ), Ca_V3 (M88751 [GenBank] ), Ca_V2a(M88751 [GenBank] ), Ca_V2a-1b chimera (Olcese et al., 1994), 2-2 (Barclay et al., 2001), and D₂ dopamine receptor (X17458 [GenBank] ). The green fluorescent protein (GFP-mut3b, U73901 [GenBank] ) was used to identify transfected cells. All cDNAs were subcloned in pMT₂.

Construction, expression, and purification of proteins. Ca_V2.2W391A was generated by site-directed mutagenesis with primers for Ca_V2.2W391A.F (5'CGGGTACCTGGAGGCGATCTTCAAGGCTGAG) and Ca_V2.2W391A.R (5'CTCAGCCTTGAAGATCGCCTCCAGGTACCCG).

Ca_V2.2 I-II loop in pGEX2T (Bell et al., 2001) [glutathione S-transferase (GST)-Ca_V2.2 I-II loop] was modified by PCR using Pfu polymerase (Stratagene, Amsterdam, The Netherlands) with the oligonucleotide primers 5'-ATGCTGGCCGAGGAGGACAGGAATGCAGAG-3' and 5'-GACTCATTCCTCCGCCTTGAAGATCCACTCCAGGTAC-3', which truncate the Ca_V2.2 I-II loop after the region encoding the AID. In this way, a construct was generated encoding the first 40 residues of the Ca_V2.2 I-II loop, incorporating the AID, fused to GST to form GST-Ca_V2.2(357-397)WT. Mutagenesis of GST-Ca_V2.2(357-397)WT was performed as described above using the primers 5'-ATCTTCAAGGCGGAGGAATGAGTCATGCTGGCCG-3' and 5'-AGCCTCCAGGTACCCGTTGAGCTCTCGCTCGATC-3' to substitute an alanine for W391. GST-Ca_V2.2(357-397)WT, GST-Ca_V2.2(357-397)W391A, and GST were expressed and purified as described for GST-Ca_V2.2 I-II loop by Bell et al. (2001). N-Terminally His-tagged Ca_v2a (H6N-Ca_v2a) and C-terminally His-tagged Ca_v1b (Ca_v1b-H6C) were expressed and purified as described by Bell et al. (2001).

Surface plasmon resonance. Assays were performed using a BIAcore 2000 (Biacore, Uppsala, Sweden) at 25°C using running buffer (20 mM Na phosphate, pH 7.5, 500 mM NaCl, and 0.005% Tween 20). Anti-GST monoclonal antibodies (Amersham Biosciences, Little Chalfont, UK) were covalently attached to the surface of CM5 dextran sensor chips according to the instructions of the manufacturer to allow the reversible immobilization of GST-fusion proteins. Approximately equivalent molar quantities 350 resonance units (RU) of GST-Ca_V2.2(357-397)WT, 350 RU of GST-Ca_V2.2(357-397)W391A, and 300 RU of GST were immobilized on successive flow cells of a sensor chip. The flow rate of running buffer was 25 µl/min. H6N-Ca_v2a and Ca_v1b-H6C were dialyzed against running buffer and diluted to a series of concentrations. These samples were applied for 6 min over all flow cells, and each injection was followed by a similar dissociation phase. The sensor chip surface was regenerated between injections by the application of 35 µl of 20 mM glycine/HCl, pH 2.2, at 10 µl/min and the immobilization of fresh ligand.

Sensorgrams were processed using the BIAevaluation 3.0 software (Biacore). Sensorgrams recorded from the flow cells containing GST-Ca_V2.2(357-397)WT and GST-Ca_V2.2(357-397)W391A were corrected for passive refractive index changes and for nonspecific interactions by subtraction of the corresponding sensorgram recorded from the flow cell containing GST only. Sensorgrams were fitted by nonlinear regression using Prism 4 (GraphPad Software San Diego, CA). For each sensorgram, the first 120 s of the association phase and the dissociation phase were fitted to a single exponential to determine the observed association rate k_on(obs) and the dissociation rate k_off. The specific association rate, k_on, was calculated as k_on = (k_on(obs) - k_off)/[Ca_V], and the dissociation constant K_D was calculated as K_D = k_off/k_on.

Cell culture and heterologous expression. The tsA-201 cells were cultured in a medium consisting of DMEM, 10% fetal bovine serum, and 1% nonessential amino acids. The cDNAs (all at 1 µg/µl) for Ca_V1 subunits, Ca_V, ₂-2, D₂ dopamine receptor, and GFP (when used as a reporter of transfected cells) were mixed in a ratio of 3:2:2:2:0.4. The cells were transfected using Fugene6 (DNA/Fugene6 ratio of 2 µg in 3 µl; Roche Diagnostics, Lewes, UK). The tsA-201 cells were replated at low density on 35 mm tissue culture dishes on the day of recording.

Biotinylation. T75 flasks of transiently transfected tsA-201 cells were washed three times with 10 ml of PBS. Cells were incubated with 2 ml of PBS, pH 8.0, containing 800 µM EZ-link Sulfo NHS-SS-Biotin (Pierce, Rockford, IL) for 15 min at room temperature. The biotinylation reaction was terminated by addition of 10 ml of 100 mM glycine in PBS, and cells were collected by centrifugation (1000 x g) and washed an additional three times with 10 ml of PBS.

Cells were lysed by addition of 750 µl of lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, and 1.5% Triton X-100) containing one Complete protease inhibitor tablet (Roche Diagnostics) per 10 ml, followed by brief sonication. After centrifugation at 20,000 x g for 20 min at 4°C, the protein concentration of the supernatant was determined using the BCA assay (Pierce). Biotinylated proteins were precipitated by incubating 500 µg of supernatant with 50 µl of streptavidin agarose beads (Pierce) for 3 h at room temperature. Precipitated proteins were washed four times with 1 ml of lysis buffer. Proteins were eluted from the beads by incubation with 100 mM dithiothreitol in lysis buffer for 30 min, followed by an equal volume of 2x SDS-PAGE sample buffer. Samples of supernatant and precipitated proteins were analyzed by SDS-PAGE on 4-12% Tris-glycine gels (Invitrogen, Paisley, UK) followed by Western blotting with anti Ca_V2.2 II-III linker antibodies (Raghib et al., 2001).

Western blot analysis. Samples (2.5-250 µg of protein) from tsA-201 whole-cell lysates [prepared as described for COS-7 cells by Raghib et al. (2001)] or from biotinylation experiments (see above) were separated by SDS-PAGE on 4-12% Tris-glycine gels and then transferred to polyvinylidene fluoride membranes. Immunodetection was performed with antibodies to the Ca_v2.2 II-III linker (Raghib et al., 2001) or 3 subunit (Canti et al., 2001) as described previously (Raghib et al., 2001).

Immunocytochemistry. Two days after transfection with Ca_V2a or Ca_V2aC3,4S, tsA-201 cells were fixed with 4% paraformaldehyde in PBS for 5 min at room temperature. For permeabilization, cells were incubated twice for 7 min in a 0.02% solution of Triton X-100 in Tris-buffered saline. For detection of Ca_V2a or Ca_V2aC3,4S, the primary antibody used was a rabbit anti-Ca_V2a (462-600) at 0.2 µg/ml (Chien et al., 1995). The secondary antibody was an anti-rabbit IgG FITC conjugated (1:500; Sigma, St. Louis, MO). For nuclear staining, 4',6-diamidino-2-phenylindole (DAPI) (300 nM; Molecular Probes, Eugene, OR) was applied. Samples were mounted in VectaShield (Vector Laboratories, Burlingame, CA) to reduce photobleaching and examined on a confocal laser scanning microscope (model LSM; Ziess, Oberkochen, Germany) using a 40x objective (1.3 numerical aperture) with constant photomultiplier settings.

Whole-cell electrophysiology. Whole-cell patch-clamp recordings were performed at room temperature (22-24°C). Only fluorescent cells expressing GFP were used for recording. The single cells were voltage clamped using an Axopatch 200B patch-clamp amplifier (Axon Instruments, Foster City, CA). The electrode potential was adjusted to give zero current between pipette and external solution before the cells were attached. The cell capacitance varied from 10 to 40 pF. Patch pipettes were filled with a solution containing the following (in mM): 140 Cs-aspartate, 5 EGTA, 2 MgCl₂, 0.1 CaCl₂, 2 K₂ATP, and 10 HEPES, titrated to pH 7.2 with CsOH (310 mOsm; with a resistance of 2-3 M). The external solution contained the following (in mM): 150 tetraethylammonium bromide, 3 KCl, 1.0 NaHCO₃, 1.0 MgCl₂, 10 HEPES, 4 glucose, and 10 BaCl₂, pH adjusted to 7.4 with Tris base (320 mOsm). The pipette and cell capacitance as well as the series resistance were compensated by 80%. Leak and residual capacitance current were subtracted using a P/4 protocol. Data were filtered at 2 kHz and digitized at 5-10 kHz. The holding potential was -100 mV, and pulses were delivered every 10 s.

Data analysis and curve fitting. Current amplitude was measured 10 ms after the onset of the test pulse, and the average over a 2 ms period was calculated and used for subsequent analysis. The current density-voltage (I-V) relationships were fitted with a modified Boltzmann equation as follows: I = G_max x (V - V_rev)/(1 + exp(-(V - V_{50, act})/k)), where I is the current density (in picoamperes per picofarad), G_max is the maximum conductance (in nanosiemens per picofarad), V_rev is the reversal potential, V_{50, act} is the midpoint voltage for current activation, and k is the slope factor. Steady-state inactivation properties were measured by applying 5 s pulse from -120 to +20 mV in 10 mV increments, followed by a 11 ms repolarization to -100 mV before the 100 ms test pulse to +20 mV. Steady-state inactivation and activation data were fitted with a single Boltzmann equation of the following form: I/I_max = (A₁ - A₂)/[1 + exp((V - V_{50, inact})/k)] + A₂, where I_max is the maximal current, and V_{50, inact} is the half-maximal voltage for current inactivation. For the steady-state inactivation, A₁ and A₂ represent the proportion of inactivating and non-inactivating current, respectively. Inactivation kinetics of the currents were estimated by fitting the decaying part of the current traces with the following equation: I(t) = C + A x exp(-(t - t_o)/_inact), where t_o is zero time, C the fraction of non-inactivating current, A the relative amplitude of the exponential, and _inact is its time constant. Activation kinetics were estimated by fitting the activation phase of the current with either a single or a double exponential. Analysis was performed using pClamp6 and Origin 7. Data are expressed as mean ± SEM of the number of replicates (n). Error bars indicate SEs of multiple determinations if not otherwise mentioned. Statistical significance was analyzed using Student's paired and unpaired t test, and p < 0.05 was considered to be statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Mutation of W391A in the I-II linker of Ca_V2.2 disrupts both Ca_V1b and Ca_V2a subunit binding
The amino acid W391 in Ca_V2.2 (Fig 1A) is conserved in the AID sequence of all HVA calcium channels (for review, see Dolphin, 2003a; Richards et al., 2004) and has been described previously to be important for the binding of the Ca_V ancillary subunits to HVA calcium channels (Pragnell et al., 1994; De Waard et al., 1996; Berrou et al., 2002). The recent structural analysis of the interaction of Ca_V subunits with the Ca_V1.2 I-II linker showed that this residue is deeply embedded in the AID binding groove in Ca_V (Chen et al., 2004; Opatowsky et al., 2004; Van Petegem et al., 2004). We first showed by surface plasmon resonance analysis that mutation of W391 to A in the AID of Ca_V2.2 prevents the binding of both Ca_V1b and Ca_V2a subunit to the I-II linker of Ca_V2.2 (Fig. 1B). In our experiments, GST-fusion proteins corresponding to the proximal I-II linker, including the AID of Ca_V2.2 or Ca_V2.2W391A, or GST alone as control, were immobilized via an anti-GST antibody to an individual flow cell of a CM5 dextran sensor chip. Ca_V subunit solutions (25-200 nM) were perfused over all flow cells. No binding of the Ca_V subunits to the control GST-fusion protein was detected (data not shown). Ca_V1b and Ca_V2a exhibited specific binding to the I-II linker of Ca_V2.2. From the data shown in Figure 1B, the dissociation constants (K_D) of Ca_V1b and Ca_V2a for the I-II loop of Ca_V2.2 were calculated to be 13.8 and 10.3 nM, respectively. Negligible binding of either Ca_V subunit was detected to the mutated Ca_V2.2W391A I-II linker, and thus the K_D values could not be determined. We can only estimate that they must be at least 1000-fold less than for the I-II linker of the wild-type channel. From this assay, we show that the W391 present in the AID is crucial for the binding of Ca_V subunits to the proximal Ca_V2.2 I-II linker, and that the affinity for binding to this construct is very similar to that observed previously for the full-length I-II linker (Bell et al., 2001).

View larger version (33K): [in this window] [in a new window]   Figure 1. The W391A mutation prevents the binding of CaV1b and CaV2a to CaV2.2 I-II linker. A, Representation of the CaV2.2 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 CaV subunits (dotted box). The sequence of the AID within CaV2.2 and CaV2.2W391A are given below, with W391 underlined and also showing the putative overlapping binding motif for G dimer binding. B, Representative Biacore sensorgrams showing interactions between CaV1b (top) or CaV2a (bottom) with I-II loop GST-fusion proteins from CaV2.2 (left) and CaV2.2W391A (right). The auxiliary CaV subunits (25-200 nM as indicated) were applied during the time indicated by the bars.

 
Ca_V subunit binding is essential for functional expression of N-type calcium channels at the plasma membrane
Coexpression of Ca_V2.2 or Ca_V2.2W391A with accessory Ca_V subunits allowed us to compare the biophysical properties of the wild-type and mutated channels. The 2-2 subunit was additionally present in all conditions in this study. The first striking effect of the W391A mutation was a large reduction of the current density for the mutated channel, consistent with a role of Ca_V binding to the I-II linker in trafficking Ca_V1 and Ca_V2 channels to the plasma membrane (Bichet et al., 2000) (Fig. 2A). The G_max determined from the I-V relationships for Ca_V2.2W391A with Ca_V1b was significantly decreased by 81 ± 3% compared with the Ca_V2.2/1b combination (Fig. 2A; Table 1). The reduction in G_max was similar (94 ± 0.5%) when the wild-type channel was expressed in the absence of Ca_V (Fig. 2A; Table 1). Similarly, the G_max was decreased by 72 ± 6% for Ca_V2a with Ca_V2.2W391A compared with the Ca_V2.2/2a combination (Fig. 2A; Table 1). In cells transfected with Ca_V2.2W391A in the absence of a Ca_V subunit, none of the GFP-positive cells expressed any current (n = 21), whereas at least 90% of the GFP-positive Ca_V2.2/1b-transfected cells displayed a current. This strongly suggests that Ca_V2.2 requires interaction via the I-II linker with a Ca_V to traffic to the plasma membrane.

View this table: [in this window] [in a new window]   Table 1. Biophysical properties of Cav2.2 and Cav2.2W391A

 

View larger version (33K): [in this window] [in a new window]   Figure 2. Role of CaV subunits in plasma membrane expression of CaV2.2. A, Left, I-V relationships for CaV2.2/2-2 coexpressed with CaV1b (filled circles, left; n = 18) or CaV2a (diamonds, right; n = 13) or without a CaV subunit (filled squares, left; n = 13) compared with I-V relationships for CaV2.2W391A/2-2 coexpressed with CaV1b (filled stars, left; n = 11) or CaV2a (filled triangles, right; n = 14). The mean data are fitted with a modified Boltzmann function (see Materials and Methods), the V50, act and Gmax 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 CaV2.2 or CaV2.2WA391A expressed with 2-2, either without CaV or with CaV1b (left) or with CaV2a (right). Total expression of CaV2.2 is shown by Western blot in the top row and biotinylated CaV2.2 in the bottom row. Cells were transfected with empty vector (lane 1), CaV2.2/2-2/1b (lane 2), CaV2.2W391A/2-2/1b (lane 3), CaV2.2/2-2 (lane 4), CaV2.2/2-2/2a (lane 5), or CaV2.2W391A/2-2/2a (lane 6). Note that only the cell surface expression is affected similarly by the W391A mutation or by the absence of CaV. Right, Histogram showing quantification of the mean amount of CaV2.2WA391A expressed at the plasma membrane, when coexpressed with either 1b or 2a, or CaV2.2 without CaV, given as a percentage of the amount of CaV2.2 expressed with the relevant CaV present under the same conditions. Data are mean ± SEM of four independent experiments. MW, Molecular weight. C, Western blot illustrating the endogenous expression of CaV3 in tsA-201 cells. Gel loaded with 2.5 µg of protein prepared from cells transfected with CaV3 (lane 1) compared with 2.5 or 250 µg of protein from cells transfected with an empty pMT2 vector (lanes 2, 3). An anti-3 monoclonal antibody was used for immunoblotting.

 
We used a biotinylation assay to assess biochemically whether there were fewer channels present at the surface of the tsA-201 cells transfected with Ca_V2.2W391A and a Ca_V subunit or when Ca_V2.2 was expressed without a Ca_V subunit, compared with wild-type Ca_V2.2/Ca_V combination. Whereas the total expression of Ca_V2.2W391A was identical to the expression of Ca_V2.2 transfected with or without a Ca_V, the amount of biotinylated channels at the plasma membrane was clearly lower (Fig. 2B). The W391A mutation decreased by 62 ± 5 and 37 ± 3% (n = 4) the number of channels that were present at the plasma membrane when coexpressed with Ca_V1b or Ca_V2a, respectively. A similar diminution of the surface expression (by 52 ± 6%) was observed for the wild-type Ca_V2.2 when expressed without Ca_V (Fig. 2B). Even when the binding of Ca_V subunits to the proximal I-II linker, including the W391A AID, was negligible in vitro, some channels were still able to traffic to the plasma membrane. It is therefore possible that Ca_V subunits can bind to other binding sites present, for example, in the distal I-II linker or on the N and C terminus of the channel as suggested previously (Cornet et al., 2002; Maltez et al., 2005).

Given the foregoing results, the reason for the presence of a low level of wild-type Ca_V2.2 current in the absence of expressed Ca_V may therefore be the presence of an endogenous subunit as described previously in Xenopus oocytes (Canti et al., 2001). This was examined by Western blotting using specific Ca_V antibodies. We clearly detected an endogenously expressed Ca_V3 in tsA-201 cells (Fig. 2C). This endogenous Ca_V may be responsible for trafficking wild-type Ca_V2.2, allowing small currents to be recorded in cells transfected without Ca_V subunits, whereas the much lower affinity of the Ca_V2.2W391A channel would preclude any interactions with the low level of endogenous Ca_V.

Biophysical properties of Ca_V2.2W391A expressed with Ca_V1b are similar to those of Ca_V2.2 expressed without Ca_V
Ca_V1b is known to hyperpolarize the activation and the steady-state inactivation of HVA calcium channels (for review, see Dolphin, 2003a). If the W391A mutation were effectively disrupting the binding of Ca_V1b to the channel, the biophysical properties of Ca_V2.2W391A should be comparable with those of Ca_V2.2 expressed without any Ca_V subunit. This was indeed the case, because tail current analysis showed the V_{50, act} to be depolarized by +9.8 and +11.2 mV, respectively, for Ca_V2.2W391A/1b and Ca_V2.2 in the absence of Ca_V compared with Ca_V2.2/1b (Fig. 3A,B; Table 1), confirming the V_{50, act} estimates obtained from the I-V relationships (Table 1). This indicated that the W391A mutation abolished the effect of Ca_V1b on the voltage dependence of activation of Ca_V2.2, such that it behaved like the wild-type channel expressed without a Ca_V subunit.

Another important feature of Ca_V1, 3, and 4 subunits is that they hyperpolarize the steady-inactivation curves of Ca_V2.2 as well as other HVA calcium channels (Bogdanov et al., 2000). The potential for half-inactivation (V_{50, inact}) was -48.6 mV for Ca_V2.2 expressed with Ca_V1b (Fig. 3C,D; Table 1). Again, in agreement with the assumption that the I-II linker of the Ca_V2.2W391A channel does not bind Ca_V1b, we found a significant depolarizing shift of the steady-state inactivation for Ca_V2.2W391A/1b [by +13 mV (Table 1)], whose steady-state inactivation curve is superimposed on that of Ca_V2.2 expressed without Ca_V1b (Fig. 3C,D; Table 1). Ca_V subunits are known to modulate not only the voltage dependence of the inactivation of calcium channels but also the kinetics of current decay (for review, see Dolphin, 2003a). The time constant of inactivation (_inact) of the Ca_V2.2W391A/1b current at +20 mV (209.2 ± 14.5 ms; n = 14) was smaller than that of the wild-type Ca_V2.2/1b combination (510.3 ± 51.6 ms; n = 25; p < 0.01), whereas the wild-type Ca_V2.2 expressed without a Ca_V subunit showed intermediate inactivation (Fig. 3E). The faster inactivation kinetics for the Ca_V2.2W391A/1b currents might be explained by the introduction of the mutation itself, because it has been shown previously that mutations in the AID are able to alter the inactivation kinetics (Dafi et al., 2004; Berrou et al., 2005).

Altogether, these results indicate that Ca_V2.2W391A is not regulated by Ca_V1b in the plasma membrane. Furthermore, if, as our evidence suggests, Ca_V2.2 expressed without Ca_V subunit is trafficked to the plasma membrane by endogenous Ca_V, this Ca_V does not regulate the channels once they have reached the membrane, as suggested previously (Canti et al., 2001).

Differential modulation of N-type channels by Ca_V2a subunits
Despite the fact that the W391A mutation effectively decreased the expression of Ca_V2.2W391A channels at the plasma membrane in the presence of Ca_V2a as well as Ca_V1b, no shift of activation was observed of the I-V curves when Ca_V2a was coexpressed with Ca_V2.2W391A compared with the wild-type Ca_V2.2/2a combination (Fig. 2A). This was confirmed by analysis of tail currents. The V_{50, act} was equivalent when either Ca_V2.2 or Ca_V2.2W391A was coexpressed with Ca_V2a (Fig. 4A,B; Table 1).

As expected, Ca_V2a depolarized the V_{50, inact} of Ca_V2.2 by +42 ± 2 mV compared with Ca_V2.2 expressed without Ca_V (Fig. 4C,D; Table 1). However, contrary to all expectations, we found that Ca_V2a was also able to depolarize the steady-state inactivation properties of Ca_V2.2W391A to the same extent as for wild-type Ca_V2.2 (Fig. 4D; Table 1). Furthermore, Ca_V2a also slowed the inactivation kinetics of Ca_V2.2W391A currents (Fig. 4E), abolishing the acceleration of the inactivation observed when Ca_V2.2W391A was coexpressed with Ca_V1b (Fig. 3E), as noted previously for other mutations in the AID region (Dafi et al., 2004). The decay of the currents was <40% during the 800 ms depolarizing pulses and therefore could not be fitted with an exponential function, so we estimated the inactivation rate from the ratio of the current at 600 ms to that at the peak. The ratio at +20 mV was equivalent when Ca_V2a was coexpressed with either Ca_V2.2 (0.76 ± 0.06) or the mutated channel (0.77 ± 0.08) (Fig. 4E).

View larger version (29K): [in this window] [in a new window]   Figure 3. Biophysical properties of CaV2.2 and CaV2.2W391A coexpressed with CaV1b. 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, CaV2.2/1b; middle, CaV2.2 without any CaV; bottom, CaV2.2W391A/1b, all coexpressed with 2-2. B, Voltage dependence of activation of CaV2.2/2-2 coexpressed with CaV1b (filled circles) or without any CaV subunit (filled squares) or CaV2.2W391A/2-2 expressed with CaV1b (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 V50, 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 CaV2.2/2-2 coexpressed with CaV1b (filled circles) or without any CaV subunit (filled squares) or CaV2.2W391A/2-2 expressed with CaV1b (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 V50,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 (inact) obtained by fitting the decaying phase of the Ba2+ currents at +20 mV with a single exponential, for CaV2.2/2-2/1b (black bar; n = 25), CaV2.2/2-2 (white bar; n = 25), and CaV2.2W391A/2-2/1b (gray bar; n = 14).

 
In summary, despite the fact that the W391 mutation was able to diminish the trafficking of the Ca_V2.2W391A channels to the plasma membrane with all of the Ca_V subunits examined and in contrast to the results obtained for Ca_V1b, the Ca_V2.2W391A/2a channels expressed at the cell surface were still modulated by Ca_V2a.

View larger version (25K): [in this window] [in a new window]   Figure 4. Biophysical properties of CaV2.2 and CaV2.2WA391A coexpressed with CaV2a. A, Representative current traces to illustrate current activation using the same protocols as described in the legend to Figure 3. Top, CaV2.2/2a; bottom, CaV2.2W391A/2a, all coexpressed with 2-2. B, Voltage dependence of activation for CaV2.2/2-2 coexpressed with CaV2a (filled diamonds; n = 10) or CaV2.2W391A/2-2 expressed with CaV2a (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 V50, 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, CaV2.2/2a; bottom, CaV2.2W391A/2a. D, Voltage dependence of steady-state inactivation for CaV2.2/2-2 coexpressed with CaV2a (filled diamonds; n = 17) or CaV2.2W391A/2-2 expressed with CaV2a (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 V50, inact values of which are given in Table 1. The dotted line represents the fit for CaV2.2 without CaV 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 IBa at 600 ms, for CaV2.2/2-2/2a (black bar; n = 15) and CaV2.2W391A/2-2/2a (gray bar; n = 14).

 
Interaction with a Ca_V subunit is essential for the voltage dependence of the modulation of Ca_V2.2 calcium channels by G-protein activation
To investigate the importance of the W391 residue for the modulation by G-proteins of Ca_V2.2 calcium channels, we coexpressed a D₂ dopamine receptor with the Ca_V2.2, Ca_V1b, 2-2 combination and activated the receptor with a maximal concentration (100 nM) of the agonist quinpirole. Figure 5A shows representative currents obtained before (P1) and immediately after (P2) a 100 ms depolarizing prepulse to +100 mV, before and during application of quinpirole. The currents measured at +10 mV were inhibited by quinpirole by 64.2 ± 4% for the wild-type channel (Fig. 5A, B). The P2/P1 ratio obtained from traces such as those represented in Figure 5A reflects the voltage-dependent loss of inhibition. A value of 2.3 ± 0.1 (n = 18) was obtained for P2/P1 at +10 mV for the wild-type channel expressed with Ca_V1b (Fig. 5C).

View larger version (27K): [in this window] [in a new window]   Figure 5. G-protein modulation of CaV2.2 and CaV2.2W391A expressed with CaV1b. 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 CaV2.2 (top) and CaV2.2W391A (bottom) coexpressed with CaV1b. 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, G12 are coexpressed, and typical traces for CaV2.2/1b and CaV2.2W391A/1b 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 CaV2.2/2-2 coexpressed with CaV1b (circles; n = 18) Bottom, I-V curves for CaV2.2W391A/2-2 coexpressed with CaV1b (stars; n = 11) are represented. I-V curves are fitted with modified Boltzmann functions, the V50, act and Gmax parameters of which are given in Table 1. E, Top, I-V curves for Ba2+ currents during P1 for CaV2.2/2-2/1b coexpressed without (filled circles; n = 6) or with (open triangles; n = 9) G12. Bottom, For CaV2.2W391A/2-2/1b coexpressed without (stars; n = 6) or with (open squares; n = 8) G12. I-V curves obtained from currents recorded during P2 when G12 was coexpressed with CaV2.2/2-2/1b (filled triangles; n = 9) or CaV2.2W391A/2-2/1b (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 CaV2.2/2-2 with CaV1b (black bars; n = 18), CaV2.2W391A/2-2 with CaV1b (white bars; n = 11) after application of quinpirole, or when G12 were coexpressed with CaV2.2/1b (gray bars; n = 9), CaV2.2W391A/1b (hatched bars; n = 8). F, Typical current traces at +20 mV for R52, 54ACaV2.2W391A (top) coexpressed with CaV1b before (filled circles) and after (open circles) activation of the D2 dopamine receptor. Corresponding I-V curves are represented in the bottom.

 
For the Ca_V2.2W391A/1b currents, inhibition by quinpirole was similar, being 58.8 ± 5.2% at +10 mV (n = 11) (Fig. 5A, B). This inhibition was prevented by preincubating the cells for 16 h with 100 ng/ml pertussis toxin (PTX), as was inhibition of the wild-type currents (data not shown). However, the P2/P1 ratio was markedly diminished to 1.3 ± 0.1 at +10 mV (n = 11; p < 0.01) (Fig. 5C), demonstrating a lack of voltage-dependent loss of the G-protein modulation for the mutated channel. In agreement with this, when the Ca_V2.2W391A/1b combination was coexpressed with G12, it resulted in tonic inhibition compared with controls in the absence of G12, but these currents exhibited no slowed activation and no prepulse facilitation, in contrast to the wild-type Ca_V2.2/1b currents (Fig. 5C-E). Furthermore, although the V_50,act of the I-V relationship was depolarized by +12.9 mV for the wild-type channel when G12 were cotransfected, it was only depolarized by +4.5 mV for the Ca_V2.2W391A channels, in agreement with a reduced voltage dependence of the modulation by G of Ca_V2.2W391A channels. Moreover, two arginines (R52 and R54) present in the N terminus of Ca_V2.2 are essential for the modulation by G of Ca_V2.2 calcium channels (Canti et al., 1999). After mutation of these two amino acids to alanines in the N terminus of Ca_V2.2W391A, quinpirole no longer inhibited the currents. This shows that these residues, and therefore G, are involved in the voltage-independent inhibition of Ca_V2.2W391A induced by activation of the D₂ dopamine receptor by quinpirole (Fig. 5F).

To investigate whether these data were contaminated by the increased inactivation rate of the Ca_V2.2W391A/1b combination compared with wild-type Ca_V2.2/1b currents, we also examined the quinpirole-mediated inhibition of the wild-type and mutated channels expressed with Ca_V3, which produces more inactivation of wild-type Ca_V2.2 than does Ca_V1b (Fig. 6A). Unsurprisingly, when Ca_V2.2W391A was coexpressed with Ca_V3, the G_max was dramatically reduced by 83 ± 4% compared with wild-type Ca_V2.2/3 (Fig. 6A; Table 1). There was also a +4.5 mV shift of the V_{50, act} for Ca_V2.2W391A compared with wild-type Ca_V2.2/3, and the Ca_V3 subunit did not hyperpolarize the steady-state inactivation of Ca_V2.2W391A (Table 1). This suggests that, like Ca_V1b, Ca_V3 is not able to modulate the biophysical properties of Ca_V2.2W391A. However, quinpirole still inhibited Ca_V2.2W391A/3 currents by 57.8 ± 11.8% at +10 mV (p < 0.01) (Fig. 6A). Furthermore, the voltage-dependent facilitation at +10 mV was greatly diminished from 2.5 ± 0.2 when Ca_V3 was coexpressed with wild-type Ca_V2.2 to 1.3 ± 0.2 (n = 7; p < 0.01) for the Ca_V2.2W391A/3 combination (Fig. 6B).

Although the Ca_V2.2W391A currents exhibited faster inactivation kinetics than the wild-type currents, the decrease of facilitation observed for the Ca_V2.2W391A currents was not attributable to inactivation occurring during either P1 or the prepulse, because the facilitation was similar for currents formed from Ca_V2.2 with either Ca_V1b or Ca_V3, despite the fact that the latter showed accelerated inactivation kinetics (at +20 mV; _inact = 174.1 ± 17.1 ms; n = 7). In addition, at all prepulse durations from 5 to 200 ms, the P2/P1 ratio for the Ca_V2.2W391A/1b currents remained markedly reduced compared with Ca_V2.2/1b (Fig. 6C). Therefore, the inactivation occurring during the prepulse is not responsible for the apparent decrease in facilitation of Ca_V2.2W391A. Furthermore, the kinetics of facilitation could be fit to a single exponential, whose time constant (_facil) for Ca_V2.2/1b was 9.0 ± 1.4 ms (Fig. 6C), whereas the small residual facilitation was much slower for the Ca_V2.2W391A/1b combination (_facil = 16.7 ± 3.2 ms). These results indicate that Ca_V accelerates the kinetics of facilitation as described previously (Canti et al., 2000), but also that the presence of a Ca_V subunit, bound with high affinity to the I-II linker of Ca_V2.2, is essential for the voltage-dependent loss of inhibition, i.e., the ability of a +100 mV prepulse to remove G-mediated inhibition in P2.

View larger version (24K): [in this window] [in a new window]   Figure 6. The kinetics of inactivation do not contaminate the properties of G-protein modulation of CaV2.2 and CaV2.2W391A. A, Typical current traces obtained for CaV2.2/2-2 and CaV2.2W391A/2-2 coexpressed with CaV3 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 CaV2.2/2-2 (black bars; n = 7) or CaV2.2W391A/2-2 coexpressed with CaV3 (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 CaV2.2 (open circles; n = 16) and CaV2.2W391A (open stars; n = 17) coexpressed with CaV1b. Data are fitted with a single exponential, the time constant (facil) of which is given in Results.

 
A related characteristic of the G-protein modulation of Ca_V2 channels is that the kinetics of current activation are slowed. Activation of the D₂ dopamine receptor significantly slowed the activation kinetics of Ca_V2.2 channels coexpressed with Ca_V1b, increasing the time constant of activation (_act) from 2.0 ms to a combination of a similar fast _act,fast (2.2 ms) and a much slower _act,slow of 23.4 ms representing 36.5% of the current (Figs. 5A, 7A). In contrast, the activation kinetics for Ca_V2.2W391A/1b measured during a pulse to +10mV were unaffected by quinpirole. _act remained fast both before (2.1 ms) and during (2.1 ms) activation of the receptor (Figs. 5A, 7A).

Another hallmark of G-protein modulation is the depolarization of the voltage dependence of current activation. For the wild-type Ca_V2.2 coexpressed with Ca_V1b, application of quinpirole induced a +17.5 mV shift of the V_{50, act}, from +14.6 ± 1.8 to +32.1 ± 2.5 mV (n = 19; p < 0.01) (Fig. 7B). This shift was reversible on washout of quinpirole (data not shown). There was no significant inhibition of tail current amplitude at +80 mV (13.5 ± 18.0%), indicating that all of the inhibition was entirely voltage dependent. In contrast, for Ca_V2.2W391A coexpressed with Ca_V1b, the tail currents were still significantly inhibited by 55.8 ± 11.9% at +80 mV (p < 0.01) (Fig. 7B). However, the V_{50, act} for the residual Ca_V2.2W391A/1b was still depolarized by application of quinpirole by +14.9 mV, from +27.5 ± 3.7 to +42.4 ± 1.9 mV (p < 0.01) (Fig. 7B), as was also true for Ca_V2.2 expressed without a Ca_V subunit (data not shown), possibly indicating that G can still affect the voltage dependence of gating of Ca_V2.2W391A by shifting the channel to a reluctant state. Altogether, these results strongly suggest that the loss of ability of Ca_V2.2W391A to bind Ca_V1b was accompanied by the almost complete loss of voltage dependence of the G-protein modulation of N-type calcium channels.

Ca_V2.2W391A coexpressed with Ca_V2a shows voltage-dependent G-protein modulation
When Ca_V2a was coexpressed with either Ca_V2.2 or Ca_V2.2W391A, quinpirole significantly inhibited the currents, by 64.4 ± 6.2% (p < 0.01) and 77.5 ± 6.1% (p < 0.01), respectively (Fig. 8A). Furthermore, in agreement with our evidence that Ca_V2.2W391A was still associated with Ca_V2a (Fig. 4), we observed voltage-dependent facilitation of the Ca_V2.2W391A/2a combination, equivalent to that observed for the wild-type Ca_V2.2/2a (Fig. 8A,B). The P2/P1 ratio at +10 mV was 4.3 ± 0.6 for the wild-type channel and 3.7 ± 0.4 for Ca_V2.2W391A (Fig. 8B).

For the Ca_V2.2/2a combination, _act was slowed by application of quinpirole from 2.2 ± 0.1 ms to a combination of a similar _{act, fast} (2.2 ms) and a very slow _{act, slow} (>400 ms), representing 51% of the current at 100 ms (Fig. 8C). The activation kinetics of Ca_V2.2W391A were similarly slowed from a _act of 2.4 ± 0.1 ms to a combination of a similar _{act, fast} (2.4 ms) and a very slow _{act, slow} (>400 ms), representing 46% of the current at 100 ms (Fig. 8C). From tail currents recorded before and during activation of D₂ dopamine receptors, we observed an equivalent shift of V_{50, act} of +12.0 ± 1.7 and +15.6 ± 2.3 mV for Ca_V2.2 and the mutated channel, respectively (Fig. 8D). At +80 mV, the control tail current amplitude was not significantly greater that that in the presence of quinpirole for Ca_V2.2W391A/2a (Fig. 8D), in contrast to the results obtained when Ca_V1b was coexpressed with Ca_V2.2W391A. We can conclude that Ca_V2a was still able to support voltage-dependent removal of the inhibition of Ca_V2.2W391A induced by activation of the D₂ dopamine receptors.

Palmitoylation is responsible for the ability of Ca_V2a to regulate the voltage-dependent properties of Ca_V2.2W391A
A peculiarity of Ca_V2a is that it contains two cysteines in its N terminus that are palmitoylated. Palmitoylation of these residues strongly modulates the biophysical properties of calcium channels associated with Ca_V2a, particularly their inactivation (Olcese et al., 1994; Chien et al., 1996; Qin et al., 1998; Bogdanov et al., 2000; Hurley et al., 2000; Restituito et al., 2000). To investigate the role of this palmitoylation, we used a Ca_V2a with the cysteines C3 and C4 mutated to serines (Ca_V2aC3,4S) (Bogdanov et al., 2000) that is unable to incorporate palmitate (Qin et al., 1998). Coexpressing Ca_V2.2 with Ca_V2aC3,4S resulted in a significant hyperpolarization of the tail current V_{50, act} by -11.4 mV compared with Ca_V2.2/2a (Fig. 9A; Table 1). The effect of Ca_V2aC3,4S was therefore similar to Ca_V1b. In contrast, when Ca_V2aC3,4S was coexpressed with Ca_V2.2W391A, the voltage dependence of activation of the channel did not show any hyperpolarizing shift compared with Ca_V2.2W391A/2a (Fig. 9A; Table 1), suggesting that, when not palmitoylated, the Ca_V2aC3,4S subunit was not able to modulate the Ca_V2.2W391A channel.

As described previously (Qin et al., 1998; Bogdanov et al., 2000; Hurley et al., 2000; Restituito et al., 2000), coexpressing this mutated Ca_V2aC3,4S subunit with Ca_V2.2 significantly hyperpolarized the steady-state inactivation compared with wild-type Ca_V2a, by -51 mV (Fig. 9B; Table 1), and accelerated the kinetics of inactivation (