INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The influx of calcium through voltage-gated Ca²⁺ channels regulates a wide range of cellular processes, including neurotransmitter release, activation of Ca²⁺-dependent enzymes and second messenger cascades, gene regulation, and proliferation. To this date, the primary structures for 10 distinct Ca²⁺ channel ₁ subunits have been identified by molecular cloning and were found to fall into three main classes: Ca_V1 with the L-type high-voltage activated Ca²⁺ channels, Ca_V2 with the non-L-type high-voltage activated Ca²⁺ channels, and Ca_V3 with the T-type low-voltage activated Ca²⁺ channels. Ca_V2.3 encodes a component of the native R-type current identified in neurons (Randall and Tsien, 1997; Piedras-Renteria and Tsien, 1998; Saegusa et al., 2000) that contributes to the synaptic transmission at hippocampal synapses (Gasparini et al., 2001) and neurohypophysial terminals (Wang et al., 1999). The Ca_V2.3 gene is expressed in islets of Langerhans where it could be involved in insulin secretion (Vajna et al., 2001). Knockout mice for Ca_V3.1 displayed a decrease in firing at the thalamocortical relay as well as a resistance to absence seizures (Kim et al., 2001).

Although a minimum voltage-gated Ca²⁺ channel can be formed by a single 1 subunit, co-expression of the full complement of subunits is required for the cardiac L-type Ca_V1.2 (Biel et al., 1991; Parent et al., 1997), brain N-type Ca_V2.1 (Williams et al., 1992; Soong et al., 1993), brain L-type Ca_V3.1 (Williams et al., 1992; Bell et al., 2001; Beguin et al., 2001), and R-type Ca_V2.3 (Parent et al., 1997) to generate Ca²⁺ currents with time course and voltage dependence similar to native currents (Catterall, 1991). -Subunits increase current density (Brice et al., 1997; Tareilus et al., 1997; Bichet et al., 2000a) by antagonizing an endoplasmic reticulum retention signal that is contained within the I-II linker (Bichet et al., 2000a). -Subunits hyperpolarize the voltage dependence of activation and inactivation, except for 2a, which decreases the kinetics and the voltage dependence of inactivation in Ca_V2.1-2.3 (Parent et al., 1997; Mangoni et al., 1997; DeWaard and Campbell, 1995; Stea et al., 1994; Jones et al., 1998; Cens et al., 1999). The mechanism underlying this effect is probably related to the palmytoylation of the cysteines 3 and 4 in the N-terminal of 2a (Chien et al., 1996; Restituito et al., 2000; Qin et al., 1998; Chien and Hosey, 1998; Stephens et al., 2000).

The auxiliary  subunits can potentially be associated with any of the six 1 pore-forming subunits of high-voltage Ca²⁺ channels (Ca_V1.1- 1.3; Ca_V2.1- 2.3) via conserved interaction domains: -interaction domain (AID) located on the I-II linker of the 1 subunit and the -interaction domain (BID) located at the beginning of the second conserved region of the  subunit (DeWaard et al., 1995, 1996; Pragnell et al., 1994; Walker and DeWaard, 1998; Walker et al., 1998, 1999). The AID binding site is composed of QQxExxLxGYxxWIxxxE where x can be any residue (see Fig. 1 A). Point mutations of conserved (Q374, Q375, E377, L380, G382, E391) or nonconserved (R378) residues within the AID motif failed to alter the binding of 3 to Ca_V2.1 (Bichet et al., 2000a). In contrast, the YWI residues appear to be critical for 1b and/or 3 binding in Ca_V2.1 (DeWaard et al., 1996). Mutations of the conserved tyrosine (Y383) residue to a serine (S) disrupted 3 binding to Ca_V2.1, although the substitution by a phenylalanine (F) or a tryptophan (W) preserved in part 3 and 1b binding (DeWaard et al., 1996; Witcher et al., 1995). The Y to S substitution in the AID motif of Ca_V1.2 and Ca_V1.1 disrupted the plasma membrane localization of the 1 subunit while preserving in part the -subunit-induced modulation of whole-cell and single-channel currents (Neuhuber et al., 1998a,b; Gerster et al., 1999).

View larger version (53K): [in this window] [in a new window]   FIGURE 1   3-Subunit binding to the I-II linker of CaV2.3. (A) Predicted secondary structure for the human brain CaV2.3 channel with the four homologous repeats and the N- and the C-termini facing the cytoplasm. The -subunit binding site on the 1 subunit (AID) is located within 20 residues of the IS6 transmembrane segment. The consensus sequence for the AID motif (QQxExxLxGYxxWIxxxE) is shown for CaV2.3 and CaV1.2 with conserved residues in bold letters. The residues mutated in the decatuple mutant RAIRENRADK-GDLEDKLDTQ (365-389) are underlined. (B) Coomassie blue-stained SDS-PAGE gel showing the wild-type and the mutant AID fusion proteins from CaV2.3. Molecular weight standards are shown to the left. The GST-proteins have a molecular mass of 25 kDa (no insert) or 33 kDa (87 AA from the I-II linker). (C) Autoradiogram of in vitro translated [35S]3 overlays on AIDE GST-I-II mutants immobilized on nitrocellulose. Whereas a strong signal was recorded for the wild-type channel, R378E, and RAIRENRADK-GDLEDKLDTQ, [35S]3 binding could not be detected for W386A. Lane 1, pGEX-4T1 vector; lane 2, wild-type CaV2.3; lane 3, W386A; lane 4, R378E; lane 5, W386A; lane 6, the decatuple mutant RAIRENRADK-GDLEDKLDTQ.

Mutations and deletions within the AID motif (see Fig. 1 A) were shown to disrupt the voltage dependence of inactivation in high-voltage activated (HVA) Ca²⁺ channels (Page et al., 1997; Geib et al., 2002; Berrou et al., 2001; Herlitze et al., 1997). We have recently shown that mutations of the nonconserved R378 in the AID motif specifically decreased the kinetics and voltage dependence of inactivation in Ca_V2.3, whereas the E462R mutation in Ca_V1.2 accelerated inactivation kinetics (Berrou et al., 2001; Bernatchez et al., 2001b). Despite being enclosed within the AID motif, Ca_V2.3 R378E, Ca_V1.2 E462R, and Ca_V1.2/Ca_V2.3 chimeras were found to be typically modulated by  subunits in terms of the kinetics and the voltage dependence of inactivation (Berrou et al., 2001; Bernatchez et al., 2001a). Although the -subunit binding properties of the AID locus have been well established in Ca_V2.1, its role in regard to -subunit binding and modulation remains to be investigated in Ca_V2.3.

We show here that point mutations of W386 disrupted -subunit binding to the I-II linker. Furthermore, the tryptophan residue appears to be an essential determinant at position 386 as substitutions with hydrophobic (A, G), hydrophilic (Q, R,E), and aromatic (F, Y) residues alike led to the same results. In contrast, the nonconserved mutations R378E and the multiple mutant R365G + A366D + I376L + R378E + E379D + N381K + R384L + A385D + D388T + K389Q preserved -subunit binding (see Fig. 1). When expressed in Xenopus oocytes, W386 mutants (A, G, Q, R, E, F, Y) and Y383 (A, S) yielded large whole-cell Ba²⁺ currents with typical current-voltage properties. 3-Subunit modulation of inactivation was eliminated in W386 mutants but not in Y383 mutants, suggesting that W386 is critical for -subunit binding and modulation of Ca_V2.3 channels.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Recombinant DNA materials

Standard methods of plasmid DNA preparation were used (Sambrook et al., 1989). cDNAs coding for the auxiliary 3 (Genbank M88751) and 2a (Genbank M80545) were kindly donated by Dr. E. Perez-Reyes (Castellano et al., 1993; Perez-Reyes et al., 1992). The wild-type human Ca_V2.3 (1E) (GenBank L27745) was a gift from Dr. T. Schneider (Schneider et al., 1994). The rat brain 2b subunit was provided by Dr. T. P. Snutch.

Point mutations and RNA transcription

Point mutations were performed with 35-40-mer synthetic oligos into the wild-type Ca_V2.3 using the Quick-Change XL-mutagenesis kit (Stratagene, La Jolla, CA). The nucleotide sequence of the mutant channel was bi-directionally analyzed using automatic sequencing by BioST (Lachine, Québec, Canada). cDNA constructs for wild-type and mutated 1 subunits were linearized at the 3' end by HindIII digestion whereas the 3 and 2a subunits were digested by NotI. Run-off transcripts were prepared using methylated cap analog m⁷G(5')ppp(5')G and T7 RNA polymerase with the mMessage mMachine transcription kit (Ambion, Austin, TX). The final cRNA products were resuspended in DEPC-treated H₂O and stored at 80°C. The integrity of the final product and the absence of degraded RNA were determined by a denaturing agarose gel stained with ethidium bromide.

3 overlay assays onto glutathione S-transferase (GST) fusion proteins

A fragment encoding the amino acids 338-425 in Ca_V2.3 was generated by polymerase chain reaction, cloned in-frame into the BamHI-XhoI sites of pGEX-4T1 vector (Amersham Pharmacia Biotech, Baie-D'urfée, Québec, Canada) and expressed in the Escherichia coli strain BL21-De3. The synthesis of the fusion proteins was induced using 0.5 mM isopropyl -D-thiogalactoside in a liquid culture grown to A₆₀₀ of ~1.0. After 2.5 h at 37°C, bacteria were collected by centrifugation. For overlay assays, crude BL21 bacterial extracts were boiled for 2 min in 2X Laemmli's loading buffer and separated on a denaturing SDS-polyacrylamide gel (12% acrylamide). Samples were loaded in duplicate so the proteins could be visualized by Coomassie staining in addition to the autoradiogram. This half of the gel was transferred onto a PVDF membrane (Millipore GmbH, Eschborn, Germany) using Towbin buffer (25 mM Tris, 192 mM glycine, 20% methanol, and 0.05% SDS). The membrane was blocked with 1% bovine serum albumin (BSA) in HBS-Tween (137 mM NaCl, 3 mM KCl, 10 mM HEPES, and 0.05% Tween-20, pH 7.4), washed once with HBS-Tween and incubated for 1 h at room temperature in 5 ml of HBS-Tween with 20 µl of [³⁵S]methionine-labeled 3 subunit. The blots were washed twice for 10 min in HBS-Tween and air dried, and radioactive signals were detected by autoradiography. ³⁵[S]Methionine-labeled 3 in pBluescript (0.5 µg) was synthesized by coupled in vitro transcription and translation (TNT Promega, Madison, WI) in a 50-µl reaction volume for 1 h, and the reaction mixture was applied without further treatment to the overlay membrane. To ensure equivalent protein loading, gels were stained with Coomassie blue to visualize the major protein band in each lane before autoradiography.

Functional expression of wild-type and mutant channels

Oocytes were obtained from female Xenopus laevis clawed frog (Nasco, Fort Atkinson, WI) as described previously (Parent et al., 1995, 1997; Berrou et al., 2001; Bernatchez et al., 1998, 2001a; Jean et al., 2002). Briefly, stage VI oocytes free of follicular cells were injected with 46 nl of a solution containing between 35 and 50 ng of cRNA coding for the wild-type or mutated 1 subunit. The 1 subunit was always co-injected with cRNA coding for the rat brain 2b (Williams et al., 1992) and either with the rat brain 3 (Castellano et al., 1993) or the rat 2a (Perez-Reyes et al., 1992) in a 3:1:1 weight ratio, respectively. Oocytes were incubated at 19°C in a Barth's solution: 100 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 5 mM HEPES, 2.5 mM pyruvic acid, 100 U/ml penicillin, 50 µg/ml gentamicin, pH 7.6.

Electrophysiological recordings in oocytes

Wild-type and mutant channels were screened at room temperature for macroscopic barium current 4-7 days after RNA injection using a two-electrode voltage-clamp amplifier (OC-725C, Warner Instruments, Hamden, CT) as described earlier (Parent et al., 1995, 1997; Berrou et al., 2001; Jean et al., 2002). Oocytes were first impaled in a modified Ringer solution (in mM: 96 NaOH, 2 KOH 1.8 CaCl₂, 1 MgCl₂, 10 HEPES) titrated to pH 7.4 with methanesulfonic acid CH₃SO₃H (MeS). The bath was then perfused with the 10 mM Ba²⁺ solution (in mM: 10 Ba(OH)₂, 110 NaOH, 1 KOH, 20 HEPES) titrated to pH 7.3 with MeS. To minimize kinetic contamination by the endogenous Ca²⁺-activated Cl current, oocytes were injected with 18.4 nl of a 50 mM EGTA (Sigma, St. Louis, MO) 0.5-2 h before the experiments. Oocytes were superfused by gravity flow at a rate of 2 ml/min, which was fast enough to allow complete chamber fluid exchange within 30 s. Experiments were performed at room temperature (20-22°C).

Data acquisition and analysis

PClamp software, Clampex 6.02 and Clampfit 6.02 (Axon Instruments, Foster City, CA) was used for online data acquisition and analysis as previously described (Bernatchez et al., 2001a,b; Berrou et al., 2001; Jean et al., 2002). Unless stated otherwise, data were sampled at 10 kHz and low pass filtered at 5 kHz using the amplifier built-in filter. For all recordings, a series of voltage pulses were applied from a holding potential of 80 mV at a frequency of 0.2 Hz from 40 to +60 mV. Isochronal inactivation data (h or h inf) were obtained from tail currents generated at the end of a 5-s prepulse (Parent et al., 1995, 1997). Tail current amplitudes were estimated using the function Analyze in Clampfit 6.0 from the peak current arising during the first 10 ms after the capacitive transient (20 data points). Each of these currents was then normalized to the maximum current obtained before the prepulse voltage (i/i_max) and was plotted against the prepulse voltage. For the isochronal inactivation figures, data points represent the mean of n  3 and were fitted to the Boltzmann Eq. 1:

(1)

Pooled data points (mean ± SEM) were fitted to Eq. 1 using user-defined functions and the fitting algorithms provided by Origin 6.0 (Microcal Software) analysis software. Equation 1 accounts for the fraction of non-inactivating current with E_0.5,inact mid-point potential; z, slope parameter; Y₀, fraction of non-inactivating current; V_m, the prepulse potential; and RT/F with their usual meanings. The fitting process generated values estimating errors on the given fit values.

Activation potentials were estimated from the normalized I-V curves obtained for each channel combination (Canti et al., 2001). Although this calculation was not exempt from gating contamination, it provided a qualitative approximation of the 3 modulation on I-V parameters. The I-V relationships were normalized to the maximum amplitude and were fitted to Eq. 2, a Boltzmann equation coupled to a linear function:

(2)

where E_0.5,act is the potential for 50% activation, G_rel is the normalized conductance, z is slope parameter, V_m is the test potential, V_rev is the apparent reversal potential, and RT/F have their usual meanings.

Inactivation kinetics were quantified using r300 values, that is the ratio of the whole-cell current remaining at the end of a 300-ms pulse (Berrou et al., 2001; Bernatchez et al., 2001a,b). As inactivation kinetics can vary with current density, comparisons between constructs and mutants were generally restricted to whole-cell currents lower than 5 µA as much as possible. Furthermore, this range of current densities made it easier to voltage clamp the oocyte uniformly, thus decreasing the possibility of series resistance artifacts contaminating the current kinetics data. Capacitive transients were erased for clarity in the final figures. Statistical analyses and Student t-test were performed using the fitting routines provided by Origin 6.1 (Microcal Software, Northampton, MA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

-Subunit overlay assays in the I-II linker of Ca_V2.3

We have recently shown that the nonconserved residue R378 within the AID motif of Ca_V2.3 channels disrupted specifically the kinetics and voltage dependence of inactivation, whereas the reverse mutation E462R in Ca_V1.2 accelerated inactivation kinetics (Berrou et al., 2001; Bernatchez et al., 2001a). Among nonconserved residues of the AID motif, R378 in Ca_V2.3 was shown to be particularly critical as multiple mutations of other nonconserved residues failed to significantly affect the kinetics and voltage dependence of inactivation (Berrou et al., 2001). To evaluate whether such mutations could have altered -subunit binding to the I-II linker, we constructed one series of GST fusion proteins containing 87 amino acids (AA) (338-428 in AA) between IS6 and the middle of the I-II linker of Ca_V2.3 (GST-AID_E). Because the GST fusion proteins are denatured before being transferred onto a nitrocellulose blot, overlay assays suggest that 3 binding to the I-II linker does not depend too critically upon the secondary structure of the AID motif. As seen in Fig. 1, in vitro translated ³⁵[S]methionine 3 could bind to the wild-type GST-AID_E, the GST-AID(R378E), and the multiple mutant GST-AID_C (R365G + A366D + I376L + R378E + E379D + N381K + R384L + A385D + D388T + K389Q) (RAIRENRADK-GDLEDKLDTQ), which includes all the nonconserved residues within the AID motif mutated to their counterparts in Ca_V1.2. In contrast, there was no discernible binding of radiolabeled 3 to the GST-AID(W386A). Hence our results with Ca_V2.3 generally agree with data previously reported for Ca_V2.1 where -subunit binding to the I-II linker was found to depend upon the conserved WYI residues and not to be critically dependent upon the nature of the intervening sequences within the AID motif (DeWaard et al., 1995).

W386A mutation impairs -subunit binding and modulation of Ca_V2.3

Correlation between -subunit binding and modulation was investigated in the following series of experiments. As shown in Fig. 2, the W386A Ca_V2.3 channel was expressed with or without (±) 3 or 2a in Xenopus oocytes. Whole-cell current traces of the wild-type Ca_V2.3 and the R378E mutant are shown alongside for comparison. Expression of the W386A mutant yielded robust inward Ba²⁺ currents with current-voltage relationships typical of voltage-gated Ca²⁺ channels. In the absence of exogenous  subunits, the W386A, R378E, and wild-type Ca_V2.3 channels activated within the same voltage range. The complete set of activation potentials is shown in Table 1. W386A/2b activated at E_0.5 = 7 ± 1 mV (n = 7), which is similar to E_0.5 = 6 ± 2 mV (n = 10) for Ca_V2.3wt/2b currents (see also Parent et al., 1997; Berrou et al., 2001). Hence in the absence of exogenous 3, the activation parameters of W386A, R378E, and wild-type Ca_V2.3 channels were comparable. In contrast, 3 induced a significant hyperpolarizing shift in the E_0.5,act values of R378E and Ca_V2.3 wild type, whereas it did not affect W386A. Furthermore, coexpression with 3 failed to significantly increase Ba²⁺ peak currents of W386A (Table 1).

View larger version (37K): [in this window] [in a new window]   FIGURE 2   Lack of -subunit-induced modulation in W386A channels. The wild-type CaV2.3 (first row), R378E (second row), and W386A (third row) were expressed in Xenopus oocytes in the presence of 2b without  subunits (left panel), 2b + 3 (middle panel), or 2b + 2a (right panel). The average peak current densities are given in Table 1. In the absence of  subunits, the r300 ratios ranked R378E > W386A > wild type (from the slowest to the fastest). The inactivation kinetics of W386A remained insensitive to the presence of either 3 or 2a. In contrast, the inactivation kinetics of R378E were accelerated by 3 and slowed by 2a in a manner similar to the wild-type channel. Whole-cell currents were recorded using the two-electrode voltage-clamp technique in the presence of 10 mM Ba2+ after injection of EGTA. Holding potential was 80 mV. Oocytes were pulsed from 40 mV to +60 mV using 10-mV steps for 450 ms. All mutants expressed significant inward currents with typical current-voltage properties. Capacitive transients were erased for the first millisecond after the voltage step.

View this table: [in this window] [in a new window]   TABLE 1   Biophysical parameters of CaV2.3 wild-type and mutant channels expressed in Xenopus oocytes in the presence of 2b and ± 3 subunits

In Ca_V2.3 channels, subunit modulation of inactivation, kinetics, and voltage dependence is isoform specific with 3 accelerating inactivation and 2a slowing it down (Olcese et al., 1994; Parent et al., 1997). In this regard, R378E behaved like Ca_V2.3wt with r300 values smaller in the presence of 3 and significantly larger with 2a (Fig. 3). Furthermore, the R378E channel displayed slower inactivation kinetics than Ca_V2.3wt under the same subunit background, in agreement with our previous study (Berrou et al., 2001). In contrast, the inactivation kinetics of W386A were not significantly affected by 3 or 2a subunit or significantly modulated by the membrane potential. Although its inactivation kinetics was slower than Ca_V2.3/2b, W386A/2b displayed significantly faster inactivation than R378E/2b, thus confirming the key role of R378 in the voltage-dependent inactivation of Ca_V2.3. As shown in greater detail later, this conclusion is supported by the functional characterization of the seven mutants made at position W386 as well as for the two Y383 mutants.

View larger version (53K): [in this window] [in a new window]   FIGURE 3   Inactivation kinetics of W386A were not modulated by 3 or 2a. (A) The mean r300 ratios (the fraction of the whole-cell current remaining at the end of a 300-ms pulse) are shown ± SEM from 0 to +20 mV for CaV2.3 wt (black) and CaV2.3 R378E (gray) in the absence and in the presence of 2a or 3. Co-injection with either 2a or 3 led to significant changes (p < 0.001) in the r300 values for the wild-type channels at all voltages. 3 sped up inactivation kinetics of R378E (p < 0.001), but 2a had no significant effect. (B) The mean r300 ratios failed to show any significant modulation of inactivation of W386A channels by 3 or 2a at +10 or +20 mV. The apparent decrease in r300 values at 0 mV was significant only at p < 0.05. No significant acceleration of inactivation kinetics was observed with increased depolarization under any condition. Whole-cell currents were measured in 10 mM Ba2+. The numbers to the left of the mutants refer to the numbers of experiments used for statistical analysis.

Finally, the voltage dependence of inactivation was measured for W386A, R378E, and the wild-type channel in the presence and in the absence of 3 using 5-s prepulses. The midpoints of inactivation (E_0.5,inact) estimated from the pooled data are shown in Table 1. As seen, coexpression with 3 did not shift the inactivation curves for W386A with E_0.5,inact = 39 ± 2 mV (n = 6) for W386A/2b as compared with E_0.5,inact = 35 ± 1 mV (n = 6) for W386A/2b/3 channels. In contrast, the E_0.5,inact values for R378E and the wild-type channels were shifted to the left by 20-30 mV in the presence of 3. Hence mutation of W386 in the AID motif was found to significantly decrease if not eliminate -subunit binding and modulation in Ca_V2.3 channels.

Functional properties of the double R378E + W386A mutant

To investigate the relationships between -subunit modulation and the inactivation properties conferred by the I-II linker, the double mutant R378E + W386A was expressed and functionally characterized in Xenopus oocytes with or without exogenous 3 (Fig. 4, A and B). The double mutant retained the dominant features of both channels, namely, the slower inactivation kinetics of R378E coupled to the absence of -subunit-induced modulation of W386A (Fig. 4 C and Table 1). As seen, the activation and inactivation properties of R378E + W386A were not modulated by 3 as was seen for W386A. The voltage dependence of inactivation of R378E + W386A/2b ± 3 was similar to R378E/2b but significantly less negative than W386A ±3 or Ca_V2.3 wt/2b as it was shifted to the right by ~+20 mV (Berrou et al., 2001). Hence, mutating two neighboring sites in the I-II linker did not alter further the inactivation properties of Ca_V2.3, suggesting that inactivation and -subunit modulation are controlled by distinct loci on the 1 subunit of Ca_V2.3.

View larger version (45K): [in this window] [in a new window]   FIGURE 4   Inactivation properties of the double mutant R378E + W386A. (A and B) Whole-cell current traces were recorded in the presence of 10 mM Ba2+ for the double R378E + W386A mutant with 2b either without exogenous 3 (A) or after co-injection of 3 (B). (C) The r300 values for the wild-type CaV2.3, W386A, R378E, R378E + W386A channels ± 3 (from left to right at each voltage) show that the inactivation kinetics of W386A and R378E + W386A were not significantly accelerated by 3. Whether in the absence or in the presence of 3, the inactivation kinetics of W386A were slower than the wild-type channel but faster than R378E. The R378E + W386A mutant combined the individual properties of R378E and W386A mutants as it inactivated like R378E and did not show any -subunit-induced modulation. (D) The voltage dependence of inactivation was estimated after a series of 5-s conditioning prepulses applied between 100 and +30 mV. The fraction of the non-inactivating current was recorded at the end of the pulse, and data were fitted to the Boltzmann Eq. 1. 3 significantly shifted the voltage dependence to the left for CaV2.3wt from 36 ± 3 mV (n = 10) to 64 ± 2 mV (n = 10) and for R378E from 26 ± 2 mV (n = 7) to 44 ± 2 mV (n = 11) but failed to influence the voltage dependence of W386A or W386A + R378E. The complete set of fit values is shown in Table 1.

3-Subunit modulation of inactivation in Y383 mutants

Mutations of the conserved tyrosine (Y) were shown to preserve in part -subunit binding (Witcher et al., 1995) to AID_A- and -subunit-induced modulation of the voltage dependence of inactivation in L-type Ca_V1.1 and Ca_V1.2 (Neuhuber et al., 1998a). -Subunit modulation was characterized after functional expression of the mutants in Xenopus oocytes. Whole-cell current traces for Y383S ± 3 (not shown) and Y383A ± 3 channels were typical of HVA Ca²⁺ channels. 3 subunits did not significantly hyperpolarize the activation potentials of either Y383A or Y383S channels (Table 1). However, 3 appeared to modulate the inactivation kinetics of Y383A, although it did not influence Y383S as shown on the r300 graph (Fig. 5 C). Furthermore, coexpression with 3 induced a significant hyperpolarizing shift of 20 mV in the voltage dependence of inactivation for both Y383A and Y383S channels (Fig. 5 D), indicating that functional modulation by 3 was preserved in part in Y383 mutants. The results obtained with Ca_V2.3 differ somehow with Ca_V1.2 (Gerster et al., 1999). In that last study, the inactivation kinetics of the Y467S mutant in Ca_V1.2 was reported to be typically modulated by  subunits, whereas the voltage dependence of inactivation remained insensitive to  subunits (Gerster et al., 1999). Nonetheless, it can be concluded that mutating the conserved Y residue does not eliminate -subunit modulation in HVA Ca²⁺ channels.

View larger version (42K): [in this window] [in a new window]   FIGURE 5   Inactivation properties of Y383A and Y383S. (A and B) Whole-cell current traces were recorded in the presence of 10 mM Ba2+ for Y383A mutant with 2b without exogenous 3 (A) or after co-injection of 3 (B). (C) The corresponding r300 values are shown for Y383A ± 3 and Y383S ± 3. The inactivation kinetics of Y383S (n = 4) was not significantly influenced by 3 (p > 0.1), whereas those for Y383A ± 3 (n = 7) were significantly different at p < 0.05. In the absence or in the presence of 3, the inactivation kinetics of Y383A and Y383S were slower than the wild-type channel and similar to W386A. (D) The voltage dependence of inactivation was estimated after a series of 5-s conditioning prepulses applied between 100 and +30 mV. 3 significantly shifted the voltage dependence of inactivation for Y383A from 32 ± 1 mV (n = 4) to 55 ± 2 mV (n = 4) and for Y383S from 24 ± 1 mV (n = 4) to 42 ± 2 mV (n = 4). The complete set of values is shown in Table 1.

Molecular determinants of -subunit binding and modulation in W386 mutants

The alanine mutation at position W386 was shown to disrupt 3-subunit binding as well as 3- and 2a-subunit-induced modulation of Ca_V2.3. The structural requirements for -subunit binding and modulation were next investigated at position W386 after substitutions with hydrophobic (A, G), hydrophilic (Q, R, E), and aromatic (F, Y) residues. [³⁵S]3 binding to GST fusion proteins mutated to W386A, W386E, W386G, W386F, W386Y, or W386Q is shown in Fig. 6. Coomassie blue stained SDS-PAGE gel attests that the GST-mutants were all expressed as 33-kDa proteins and that gel loading was equivalent in each lane. None of the mutants displayed any discernible trace of 3 overlay binding although [³⁵S]3 binding on the control wild-type channel was observed under the same experimental conditions.

View larger version (28K): [in this window] [in a new window]   FIGURE 6   3 does not bind the W386 mutants. (A) Coomassie blue-stained SDS-PAGE gel showing the wild-type and the mutant W386 fusion proteins from CaV2.3. Molecular weight standards are shown to the left. The fusion proteins have a molecular mass of 25 kDa (no insert) or 33 kDa (86AA from the I-II linker). (B) Autoradiogram of in vitro translated [35S]methionine 3 overlays on AIDE GST-W386 mutants immobilized on nitrocellulose. The pGex (no insert) is shown as a control (lane 1), whereas a strong signal was recorded for the AID motif from the wild-type CaV2.3 channel (lane 2), [35S]3 binding could not be detected for W386A (lane 3), W386E (lane 4), W386G (lane 5), W386Y (lane 6), W386Q (lane 7), W386F (lane 8).

W386 mutants were expressed ±3 and characterized in Xenopus oocytes (Fig. 7). All W386 mutants, including W386R and W386F (not shown), expressed robust inward currents in the presence of 10 mM Ba²⁺. In this regard, peak current expression was found to vary widely from day to day as compared with the wild-type channel recorded under the same conditions. The mean current-voltage relationships of the W386 mutants were not significantly shifted in the hyperpolarized direction by 3. As seen for W386A, the activation potentials for the W386 mutants were generally in the same range as the Ca_V2.3/2b channels without any significant modulation by 3 with the exception of W386R, which showed a reverse sensitivity to 3 (Table 1).

View larger version (33K): [in this window] [in a new window]   FIGURE 7   W386 mutants are not modulated by 3. Whole-cell currents were recorded in the presence of 10 mM Ba2+ for the wild-type CaV2.3, W386E, W386G, and W386Y with 2b in the absence of 3 (upper panel) and when coexpressed with 3 (middle panel). The inactivation kinetics of the W386 mutants was not accelerated by 3. The peak current voltage relationships were not shifted to the left in the presence of 3 for the W386 mutants as seen on the mean normalized current-voltage relationships (lower panel). Identical results were obtained for W386A, W386F, and W386R (not shown).

The kinetics (Fig. 8 A) and voltage dependence of inactivation (Fig. 8 B) of W386 mutants were also poorly modulated by 3. As seen at 0 mV, inactivation kinetics was similar for W386A ± 3, W386E ± 3, W386G ± 3, and W386Y ± 3 with ~30% of the whole-cell currents remaining at the end of a 300-ms pulse. 3 did not increase their inactivation kinetics (p > 0.1) in contrast to the fourfold acceleration experienced by the wild-type channel (p < 10⁴). Some W386 mutants behaved distinctively. For instance, the inactivation kinetics of W386A ± 3 and W386E ± 3 were distinctively voltage independent, whereas W386G ± 3 and W386Y ± 3 appeared to inactivate significantly faster with depolarization (p < 10³).

View larger version (38K): [in this window] [in a new window]   FIGURE 8   Inactivation properties of the W386 mutants (A) The cumulative r300 values are shown for the wild-type CaV2.3, W386A, W386E, W386G, and W386Y ± 3. The inactivation kinetics of W386 mutants was not significantly influenced by 3. Nonetheless, the inactivation kinetics of W386G and W386Y remained voltage dependent and accelerated with depolarization. (B) The voltage dependence of inactivation was estimated after a series of 5-s conditioning prepulses applied between 100 and +30 mV. The fraction of the non-inactivating current was recorded at the end of the pulse, and data were fitted to the Boltzmann Eq. 1. 3 had no significant effect on the voltage dependence of inactivation for W386A, W386G, and W386E but barely shifted the E0.5 of W386Y to the left (p > 0.05). The complete set of fit values is shown in Table 1.

In contrast to Ca_V2.3wt and Y383 mutants, which experienced clear hyperpolarizing shifts in the presence of exogenous 3, the voltage dependence of inactivation of W386A, W386E, W386Q, W386G, W386F, and W386Y channels was not modulated by 3 (Table 1). The inactivation curves of the W386 mutants fell roughly in three groups. W386E ± 3, W386G ± 3, and W386Q ± 3 inactivated with E_0.5,inact  26 mV, ~10 mV more positive than the wild-type channel without exogenous 3. W386R ± 3, W386F ± 3, and W386A ± 3 inactivated with E_0.5,inact  36 mV, which is very similar to the wild-type channel without exogenous 3. Of all the mutants, W386Y ± 3 inactivated at the most negative membrane potentials E_0.5,inact  46 mV, which is ~10 mV more negative than Ca_V2.3/2b (no exogenous 3). This result suggests that the higher hydrophilicity of the Tyr residue could influence the voltage-dependent inactivation of Ca_V2.3. However, no substitution of W386 could confer the typical 3-induced modulation of kinetics and voltage dependence of inactivation in Ca_V2.3.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

W386 in the AID motif is required for -subunit binding and modulation of Ca_V2.3

In this study, the molecular determinants of -subunit binding and modulation in the Ca_V2.3 Ca²⁺ channel were investigated following mutations within the high-affinity -subunit binding site (AID) of the I-II linker. The AID motif is composed of a stretch of 18 AA located about at the end of IS6 that reads QQxExxLxGYxxWIxxxE. Before our study, little was known on the determinants of -subunit binding and modulation in Ca_V2.3. Landmark studies by the groups of Campbell and deWaard (Pragnell et al., 1994; Witcher et al., 1995; DeWaard et al., 1996; Bichet et al., 2000b) have highlighted the core YWI residues as key determinants of -subunit binding in Ca_V2.1 (Pragnell et al., 1994; DeWaard et al., 1996). In particular, point mutations of conserved residues QQxExxLxGxxxxxxxxxE did not prevent 1b (Pragnell et al., 1994; DeWaard et al., 1996) or 3 binding (Bichet et al., 2000b) to the I-II linker in Ca_V2.1. We herein confirmed that mutating W386 disrupted 3-subunit overlay binding to AID_E as it was previously shown for AID_A (Bichet et al., 2000b). We further showed that mutations of W386 dramatically decreased the -subunit modulation of activation and inactivation of Ca_V2.3 channels. In contrast, mutations of the neighboring residue Y383 preserved the -subunit modulation in the voltage- dependence of activation and inactivation. The strong correlation between -subunit binding and modulation for W386 hence suggests that W386 constitutes the primary determinant of -subunit binding and modulation in Ca_V2.3.

One of the novel pieces of information arising from our data turns out to be the strict requirement for a tryptophan at position 386 for 3-subunit binding and modulation. The tryptophan residue at position 386 appears to be an essential structural determinant for -subunit binding because substitutions with hydrophobic (A, G), hydrophilic (Q, R, E), or aromatic (F, Y) residues disrupted 3 binding as well as -subunit modulation. Previous studies had clearly demonstrated the absence of 1b binding to the WA mutant in Ca_V2.1 but reported some level of interaction between 1b and WF/WY mutants (DeWaard et al., 1996), suggesting that the delocalization of  electrons in the phenyl groups could be involved in the interaction with  subunits with Ca_V2.1. At this time, we cannot rule out that some level of weak interaction remains between 3 and the W386 mutants. Our experiments were performed with 1:3 subunits co-injected at a 1:1 molar ratio. Additional experiments aimed at elucidating the affinity of 3 to the AID_E mutants require more sophisticated tools such as fast sampling kinetic analyses (Berteloot et al., 1991; Oulianova et al., 2001) or surface plasmon resonance binding (Canti et al., 2001).

Functional expression of W386 and Y383 mutants in Xenopus oocytes

All W386 (A, G, Q, E, R, F, Y) and Y383 (A, S) mutants expressed large inward Ba²⁺ currents. -Subunits are involved in the membrane trafficking of the 1 subunit in voltage-dependent Ca²⁺ channels where they are actually believed to chaperone the 1 subunit to the membrane (Chien et al., 1995; Neuhuber et al., 1998a; Tareilus et al., 1997; Yamaguchi et al., 1998; Gerster et al., 1999). For instance, the I-II linker was found to regulate the Ca_V2.1 channel expression by interacting tightly with a retention signal in the endoplasmic reticulum (ER). The inclusion of the I-II linker of Ca_V2.1 inserted at the end of the Shaker K⁺ channel was shown to prevent the membrane expression of Shaker channels (Bichet et al., 2000a). High-affinity binding of  subunits to the AID motif is required to dislodge the I-II linker from the ER, thereby relieving the trafficking clamp and allowing membrane expression of 1 subunits (Bichet et al., 2000a). Unexpectedly, disrupting -subunit binding to the I-II linker did not eliminate functional channel expression of Ca_V2.3 (our results) or Ca_V2.1 (Bichet et al., 2000a). Indeed, mutating W386 or Y383 in AID_E as well as deleting 36 AA of AID_A yielded HVA functional channels (Bichet et al., 2000a). It remains to be seen whether mutations of the AID motif lessened the interaction between the I-II linker and the retention signal in the ER, thereby decreasing the need for a chaperone auxiliary subunit.

The contribution from additional -subunit binding sites already identified in other cytoplasmic regions of the 1 subunit in Ca_V2.1 and Ca_V2.3 using in vitro binding experiments (Birnbaumer et al., 1998; Cens et al., 1998; Olcese et al., 1994; Qin et al., 1997; Walker et al., 1998) remains to be fully investigated in Ca_V2.3. Although their functional relevance in terms of -subunit modulation has yet to be established, such binding sites could partially offset the consequences of disrupting the AID motif by providing some level of interaction between 1 and  subunits.

-Subunit binding is preserved after multiple mutations of nonconserved residues in AID

The molecular determinants of -subunit binding and functional modulation of Ca_V2.3 were investigated following mutations of conserved and nonconserved residues within the high-affinity -subunit binding site (AID) of the I-II linker. As far as -subunit binding is concerned, multiple mutations of the nonconserved residues within the AID motif, as in the Ca_V2.3 R365G + A366D + I376L + R378E + E379D + N381K + R384L + A385D + D388T + K389Q mutant, did not disrupt [³⁵S]3 overlay binding to GST fusion proteins from Ca_V2.3, suggesting that 3 binding to the I-II linker is not critically sensitive to the nature of residues interwoven in the AID motif. Our findings thus extend previous reports that the equivalent mutation R387E in Ca_V2.1 retains the ability to bind [³⁵S]3 (Bichet et al., 2000b). The binding experiments further agree with our previous report showing that -subunit modulation of inactivation was preserved in the R378E channel (Berrou et al., 2001). Furthermore, our current study confirmed that the inactivation kinetics of R378E were consistently slower than Ca_V2.3wt when expressed under the same subunit background (Figs. 2, 3, 4, 6, and 7). Hence, mutating R378 in Ca_V2.3 did not appear to modify significantly the extent to which  subunits regulate inactivation. This contrasts with the recent observation that the similar mutation in Ca_V2.1 (R387E) slowed down the inactivation kinetics of Ca_V2.1 but only when measured in a 4 background (Geib et al., 2002). Altogether, our data support the conclusion that the changes in the inactivation properties reported for R378E in Ca_V2.3 were likely to be conferred by the 1 subunit itself (Berrou et al., 2001).

Role of the I-II linker and -subunits in the voltage-dependent inactivation of Ca_V2.3

Mounting evidence suggests that the I-II linker of HVA 1 subunits behaves as a tethered inactivating blocking particle (Berrou et al., 2001; Stotz and Zamponi, 2001). Besides the control of inactivation, the I-II linker contains, however, other key regulatory sites for channel activity because it anchors -subunit interaction and provides modulation by G proteins and protein kinase C (DeWaard et al., 1997; Zamponi et al., 1997), which in turn could affect inactivation properties (kinetics and voltage dependence). We attempted to unravel the role of the I-II linker in the functional modulation of Ca_V2.3 by investigating the -subunit binding and modulation as well as the inactivation properties of the R378E, W386, and the R378E + W386A mutants.

The R378E channel displayed decreased inactivation kinetics and voltage dependence, although it preserved typical 3-subunit binding and modulation. Although W386A channels failed to be typically modulated by 3 subunits, their inactivation properties (kinetics and voltage dependence) remained weaker than the wild-type channel expressed under the same conditions. Hence, with the exception of W386G at +20 mV, all W386 mutants inactivated significantly slower than the Ca_V2.3wt channel expressed in the absence of 3 subunits. Nonetheless, the voltage dependence of inactivation properties of the W386 mutants appeared to be variable. For instance, although the voltage dependence of inactivation of W386A, W386R, and W386F mutants was indistinguishable from the wild-type Ca_V2.3/2b (without 3), the W386E, W386G, and W386Q channels inactivated distinctively at more positive voltages. Furthermore, the tryptophan-to-tyrosine substitution that preserved an aromatic residue at position 386 (W386Y) yielded channels that inactivated at more negative potentials than the wild-type channel expressed without 3 subunits. Interestingly, the slightly less polar phenylalanine residue, albeit bearing a similar phenyl group, behaved like the W386A channel. This observation suggests that mutations of conserved and nonconserved residues alike in the I-II linker could alter the inactivation kinetics of Ca_V2.3 channels (Berrou et al., 2001). We examined this proposition by investigating the properties of the double mutant R378E + W386A. The double mutant lacks 3-subunit modulation, but its inactivation properties were shown to be identical to the R378E/2b channel. Hence, mutating the neighboring residue W386 does not further decrease the kinetics and voltage dependence of inactivation of R378E. These results suggest altogether that R378 and W386, albeit localized on the same motif, control distinct functions in HVA Ca²⁺ channels and confirm that R378 in the I-II linker is the key determinant of voltage-dependent inactivation in Ca_V2.3.

In line with our observations, deWaard and collaborators have recently proposed that the altered inactivation properties of the R387E mutant in Ca_V2.1 resulted from the disruption of the intra-subunit interaction between the cytoplasmic I-II and the III-IV linkers (Geib et al., 2002). In their model, the strong interaction between the I-II and the III-IV linkers would prevent fast inactivation kinetics of the wild-type Ca_V2.1 channel. The  subunits would then promote faster inactivation kinetics by weakening the linkers' interaction. This stimulating proposition might, however, not extend to all HVA Ca²⁺ channels. In Ca_V2.1, R387E was shown to inactivate faster than the wild-type channel in the absence of 4. The situation appears to be different with the R378E mutant in Ca_V2.3 as it displayed slower inactivation kinetics and reduced voltage dependence of inactivation as compared with the wild-type channel whether in the presence or absence of coexpressed 3 or 2a subunits (Fig. 2) (Berrou et al., 2001). These contradictory results might either result from intrinsic differences in the inactivation mechanism of HVA Ca²⁺ channels or else reflect -subunit isoform binding specificity. It has been reported that 4 interacted with relatively high affinity to numerous intracellular loops (Walker et al., 1998, 1999), whereas 1b and 3 displayed a significantly higher affinity for AID in the I-II linker than for any other cytoplasmic loop of Ca_V2.1 (Walker et al., 1998, 1999). It hence remains to be seen whether this pattern of selectivity among  subunits is preserved in Ca_V2.3 and more specifically in R378E.