Dissection of Functional Domains of the Voltage-Dependent Ca²⁺ Channel α₂δ Subunit

Abstract

Coexpression of the cloned voltage-dependent Ca²⁺ channel α₂δ subunit with the pore-forming α₁ subunit results in a significant increase in macroscopic current amplitude. To gain insight into the mechanism underlying this interaction, we have examined the regulatory effect of either the α₂δ complex or the δ subunit on the Ca²⁺ channel α₁ subunit. Transient transfection of tsA201 cells with the cardiac L-type α_1Csubunit alone resulted in the expression of inward voltage-activated currents as well as measurable [³H]-PN200-110 binding to membranes from transfected cells. Coexpression of the α₂δ subunit significantly increased the macroscopic current amplitude, altered the voltage dependence and the kinetics of the current, and enhanced [³H]-PN200-110 binding. Except for the increase in amplitude, coexpression of the δ subunit reproduced entirely the effects of the full-length α₂δ subunit on the biophysical properties of the α_1Ccurrents. However, no effect on specific [³H]-PN200-110 binding was observed on δ subunit coexpression. Likewise, profound effects on current kinetics of the neuronal α_1A subunit were observed on coexpression of the α₂δ complex in Xenopus oocytes. Furthermore, by using a chimeric strategy, we localized the region involved in this regulation to the transmembrane domain of the δ subunit. These data strongly suggest that the molecular determinants involved in α₂δ regulation are conserved across L-type and non-L type Ca²⁺ channels. Taken together, our results indicate that the region of the α₂δ subunit involved in the modulation of the gating properties of the high voltage-activated calcium channels is localized in the δ domain of the protein. In contrast, the level of membrane expression of functional channels relies on the presence of the α₂ domain of the α₂δ complex.

Voltage-gated Ca²⁺ channels are multisubunit protein complexes that control the entry of Ca²⁺ions across the membrane of excitable cells and play a major role in several physiological processes, including neurotransmission, muscle contraction, hormone secretion, and gene expression. Five classes of voltage-gated Ca²⁺channels have been described so far on the basis of their biophysical and pharmacological properties (T-, L-, N-, P/Q-, and R-types). Functional differences among Ca²⁺ channel types are attributable to several factors, including the expression of distinct α₁ pore-forming proteins and the selective association of β and α₂δ regulatory subunits (for review, seeCatterall, 1995; Dunlap et al., 1995; De Waard et al., 1996).

According to available biochemical (Chang and Hosey, 1988; Schneider and Hofmann, 1988; Kuniyasu et al., 1992; Tokumaru et al., 1992) and molecular biological data (Mikami et al., 1989; Hullin et al., 1992;Perez-Reyes et al., 1992; Collin et al., 1993), Ca²⁺channels are composed of at least three subunits: α₁, β, and α₂δ. Expression of the cloned β subunit results in an increase in current amplitude and changes the biophysical properties of the α₁ pore-forming subunit (Mori et al., 1991; Hullin et al., 1992; Perez-Reyes et al., 1992; Neely et al., 1993, 1995; Nishimura et al., 1993; Chien et al., 1995; Massa et al., 1995; Pérez-García et al., 1995; Kamp et al., 1996). Likewise, functional coexpression of the α₂δ subunit, the product of a single gene that is post-translationally processed to yield separate subunits (α₂ and δ) linked by disulfide bonds (De Jongh et al., 1990; Jay et al., 1991), also results in significantly increased macroscopic currents through α₁/β recombinant calcium channels (Singer et al., 1991; Itagaki et al., 1992; Williams et al., 1992; Shistik et al., 1995; Bangalore et al., 1996; Wiser et al., 1996). Effects on dihydropyridine (DHP) binding also have been attributed to expression of the β and α₂δ auxiliary subunits (Welling et al., 1993; Mitterdorfer et al., 1994; Wei et al., 1995).

Although these studies suggest multiple roles for the α₂δ and β subunits in the processing and function of Ca²⁺ channels, little has been reported about the mechanisms of interaction between the α₁ and α₂δ subunits without the modulatory effect of the β subunits. Even less is known about the functional significance of the δ subunit. To begin to address how these proteins participate in channel function, we have studied the regulatory effects of both the α₂δ ancillary complex and the δ subunit on the cardiac L-type and the neuronal class A α₁ pore-forming subunits.

MATERIALS AND METHODS

Cell culture and transfection. Human embryonic kidney tsA201 cells (HEK293 cells transformed with SV40 large T antigen) were grown in DMEM–high glucose supplemented with 10% equine serum, 2 mml-glutamine, 110 mg/l sodium pyruvate, and 50 μg/ml gentamycin at 37°C in a 5% CO₂ and 95% air humidified atmosphere. Transfections were performed with the calcium phosphate method (Ausubel et al., 1995) with 10 μg of plasmid cDNA encoding the rabbit L-type Ca²⁺ channel α_1C pore-forming subunit (Wei et al., 1991) alone or in combination (molar ratio 1:1) with either plasmid cDNA encoding the rat full-length Ca²⁺ channel α₂δ regulatory complex (Kim et al., 1992; Gurnett et al., 1996) or the δ subunit alone (Gurnett et al., 1996). The plasmid cDNA-encoding δ subunit was made by assembling a PCR fragment in the pcDNA3 mammalian expression plasmid (Invitrogen, San Diego, CA) after the α₂δ signal sequence. The δN-myc plasmid used in Western blot analysis was made by using two sequential PCR reactions and ligating the product into the KpnI and EcoRI sites of the pcDNA3 vector. For electrophysiology, 3 μg of a plasmid DNA encoding the CD8 surface marker (EBO-pcD-Leu2; American Type Culture Collection, Rockville, MD) also was added to the DNA transfection mixture to select cells that expressed Ca²⁺ channels.

Immunoblotting and in vitro translation. tsA201 cells were harvested 2–3 d after transfection, and cell microsomes were prepared. Cells were lifted off plates into PBS, collected by centrifugation, and resuspended in lysis buffer (50 mmTris-HCl, pH 7.4) in the presence of protease inhibitors (in μm): 0.8 aprotinin, 640 benzamidine, 1.1 leupeptin, 0.7 pepstatin A, and 230 PMSF. Cells were homogenized and centrifuged at 1300 × g for 5 min. The microsomes in the supernatant were collected by centrifugation at 130,000 × g for 37 min, resuspended in 0.3 m sucrose, 20 mm Tris, and protease inhibitors, and stored at −80°C. Subsequently, equivalent amounts (200 μg of protein) of cell microsomes were electrophoresed on 5–16% gradient SDS-PAGE and transferred to nitrocellulose membranes. Membranes were incubated overnight with a polyclonal antibody against α₂δ (rabbit 136; 1:400) or a monoclonal antibody against the myc epitope (9E10; 1:1000), subsequently incubated with either horseradish peroxidase-conjugated goat anti-rabbit IgG or goat anti-mouse IgG (Boehringer Mannheim, Indianapolis, IN) at 1:1000, and visualized with ECL. Ca²⁺ channel subunits also were synthesized by coupled in vitro transcription and translation with the TNT lysate system (Promega, Madison, WI) from the same plasmid cDNA used in transfections and analyzed by SDS-PAGE.

Electrophysiology. One day after transfection tsA201 cells were transferred to poly-l-lysine-coated coverslips and grown for 1–2 d until used for electrophysiology. To select Ca²⁺ channel-expressing cells, we incubated a coverslip immediately before recording in bathing solution (see below) containing 1–2 × 10⁶/ml paramagnetic beads precoated with a monoclonal antibody (ITI-5C2) specific for the CD8 membrane antigen (Dynal, Lake Success, NY), and the positively transfected cells were subjected to the whole-cell mode of the patch-clamp technique (Hamill et al., 1981). Briefly, patch pipettes were pulled from borosilicate glass capillaries. Typical pipette resistances were 2–5 MΩ when filled with internal solution. Currents were recorded with an Axopatch 200A amplifier (Axon Instruments, Foster City, CA) and acquired on-line, using a TL-1 interface with pClamp 6 software (Axon). After the whole-cell mode had been established, capacitive transients were canceled with the amplifier. Currents were obtained from a holding potential of −80 mV by applying test pulses every 20 sec, and data were leak-subtracted on-line by a standard P/4 protocol. Current signals were filtered at 1–2 kHz (internal four-pole Bessel filter) and digitized at 20 or 1.67 kHz, depending on the duration of the voltage steps. The bath solution contained (in mm): 40 BaCl₂, 125 TEA-Cl, 10 HEPES, and 5 glucose, pH 7.3. The internal solution consisted of (in mm): 135 CsCl, 5 MgCl₂, 10 EGTA, 10 HEPES, 4 Mg-ATP, and 0.1 GTP, pH 7.3.

Electrophysiological analysis of calcium currents expressed inXenopus oocytes was performed as described previously (De Waard and Campbell, 1995). Briefly, the follicle membranes of stage V and VI isolated Xenopus laevis oocytes were digested enzymatically with 2 mg/ml collagenase IA (Sigma, St. Louis, MO) to facilitate injection and recording. cRNAs were transcribed in vitro, using T7 RNA polymerase or SP6 RNA polymerase in the case of α_1A. Oocytes were injected with 50 nl of various subunit composition at the following cRNA concentrations: 0.7 μg/μl α_1A, 0.1 μg/μl β₄, and 0.7 μg/μl full-length α₂δ subunit or α₂Ad chimeric construct (Gurnett et al., 1996). Currents were recorded by the two-electrode voltage-clamp method with a TEV-200 amplifier (Dagan, Minneapolis, MN). Both voltage and current electrodes were filled with 3 m KCl and had resistances of ∼0.5 MΩ. Peak Ba²⁺ currents were measured for a test potential of 0 mV from a holding potential of −90 mV. Recordings were filtered at 0.5 kHz, sampled at 5 kHz, and analyzed by pClamp 6. Leak and capacitance currents were subtracted on-line by a P/6 protocol. The bath solution contained (in mm): 40 Ba(OH)₂, 50 NaOH, 2 KCl, 1 niflumic acid, 0.1 EGTA, and 5 HEPES, pH 7.4.

Radioligand binding. The effect of the Ca²⁺ channel auxiliary subunits on the dihydropyridine binding of the α_1C-transfected cells was characterized as follows. Aliquots (80 μg of protein) of cell microsomes prepared as mentioned above were resuspended in a total volume of 400 μl of buffer A (50 mm Tris, 0.1% BSA, and protease inhibitors) and incubated with increasing concentrations of (+)-[methyl-³H]-PN200-110 (Amersham, Arlington Heights, IL) in the dark at 37°C. After 60 min the receptor–ligand complexes were collected and washed with buffer A on Whatman GF/B filters with a cell harvester. Nonspecific binding was determined by the addition of 50 μm nitrendipine 10–15 min before the addition of [³H]-PN200-110. Specific binding was calculated by subtracting nonspecific from total binding.

Statistical analysis. The data are given as mean ± SE, and the number of experiments is indicated in the figure legends. Statistical differences between two means were determined by Student’st tests. Means were considered significantly different whenp < 0.05.

RESULTS

In vitro transcription–translation and cell expression of Ca²⁺ channel subunits

The plasmids used in the transfection initially were examined via a cell-free transcription–translation system. The α_1C, α₂δ, and δ cDNA clones directed the synthesis of three polypeptides of the expected molecular weight. Likewise, the expression of the α₂δ complex and the δ subunit (containing the myc epitope fused to its N terminus) was demonstrated in tsA201 cells 48 hr post-transfection by Western blot analysis with either an antibody against the α₂protein or the anti-myc antibody, respectively (data not shown). The expression of the α_1C subunit (the pore-forming and DHP-sensitive component of the channel) was detected by electrophysiology and binding experiments, as detailed below.

Effect of the regulatory subunits on current amplitude

We initiated the study of the regulatory effects of the α₂δ complex on the L-type Ca²⁺channel by comparing the fundamental biophysical properties between cells transiently transfected with constructs encoding α_1C alone or cotransfected with the α₂δ or δ subunits. Figure1A shows representative examples of mean current traces (average of three successive sweeps) obtained during 150 msec pulses in untransfected (control) and transfected cells at a test potential of +30 mV. The top trace reveals the absence of endogenous voltage-activated Ca²⁺channels in control cells. In contrast, transfection with the Ca²⁺ channel α_1C subunit resulted in the expression of inward voltage-activated currents both in the presence and absence of the auxiliary subunits. A large increase in current amplitude was observed on coexpression of the full-length α₂δ subunit: the peak current amplitude was increased approximately threefold at +30 mV (Fig. 1B, Table1). Although coexpression of the δ subunit did not modify the magnitude of the α_1C currents, evident changes in voltage dependence and waveform were observed (see below).

Fig. 1.

Whole-cell currents in tsA201 transiently transfected cells. A, Ba²⁺ currents induced by activating pulses in four different representative cells.Top trace corresponds to the current recorded in a control (untransfected) cell. Lower traces correspond to current records that were obtained in cells transfected with Ca²⁺ channel subunits in various combinations. The voltage protocol is shown above the traces, and thedotted line represents the baseline current.B, Comparison of peak current amplitudes at +30 mV in control and transfected cells. Data are expressed as mean ± SE, and the number of recorded cells is indicated inparentheses. Statistical significance of the difference between singly transfected and cotransfected cells was determined by Student’s t test (*p < 0.05).

Table 1.

Differential effects of α₂δ and δ on the biophysical and binding properties of the α_1C currents

Properties	α1C	α1C/α2δ	α1C/δ
Peak current amplitude (pA)	−72  ± 15	−221  ± 24	−65  ± 11
Apparent activation threshold (mV)	≥−10	≥−20	≥−20
Potential for half-activation (mV)	19.6	8.7	8.5
Potential for half steady-state inactivation (mV)	−6.8	−15.9	−16.4
Decay of the current at 150 msec* (%)	19.5  ± 4.2	39.2  ± 9.8	36.4  ± 6.2
Rate of inactivation at +20 mV (μsec−1)	2.5  ± 0.5	4.3  ± 0.4	4.6  ± 0.7
Bmax (fmol/mg of protein)	67  ± 29	141  ± 39	41  ± 10
KD(nm)	0.9  ± 0.5	0.1  ± 0.04	2.4  ± 0.9

^* Currents were recorded during depolarizations from −80 to +30 mV.

B_max, Maximum binding capacity;K_D, dissociation constant (mean ± SE; n = 3–5).

Effect of the auxiliary subunits on the voltage dependence of the expressed currents

Previous studies in our laboratory with neuronal class A recombinant Ca²⁺ channels expressed inXenopus oocytes demonstrated that coexpression of the δ subunit did not result in an enhancement of current amplitude (Gurnett et al., 1996). The question was raised, however, whether this protein directly interacted with the α₁ subunit although it had no effect on the one parameter studied (current amplitude). In the present study a direct comparison of the electrophysiological properties of singly transfected and cotransfected cells showed that the concomitant expression of α₂δ not only increased current amplitude but also influenced the voltage dependence of activation and inactivation. This allowed us then to study the effect of the δ subunit on these properties. As illustrated in Figure2A, in cells transfected with α_1C only, the current begins to turn on at −10 mV and reaches the peak at +30 mV. When the peak currents measured at each test potential were normalized to the maximum current observed in each cell, averaged and plotted as a function of test potential, a ∼10 mV shift in the I–V curve in the hyperpolarizing direction was observed in the cells that expressed the α₂δ subunit. Cotransfection of the δ subunit resulted in a similar shift in the voltage dependence of activation (Table1).

Fig. 2.

Voltage dependence of the currents through expressed Ca²⁺ channels. A, Plot of peak normalized current–voltage relationships in transiently transfected cells. Currents were recorded in response to 150 msec depolarizations from a holding potential of −80 mV with a 10 mV increase in the pulse amplitude per step. Symbolsrepresent mean ± SE values of five to eight cells in each condition. Fits of the I–V curves were obtained assuming an activation curve of a Boltzmann type:I_Ba = [g(V_m −V_rev)]/(1 + exp[−(V_m −V_1/2)/s]), whereg is the conductance, V_mrepresents the test potential, V_rev is the apparent reversal potential, and s is the range of potential for an e-fold change aroundV_1/2. B, Measurement of the voltage dependence of inactivation at steady state. The graph shows peak currents at +30 mV as a function of the prepulse potential for three different cells (α_1C, •; α_1C/α₂δ, ▪; α_1C/δ, □). A series of 13 different 1.2 sec prepulse potentials from −80 to +40 mV was applied first, and the inactivated currents were measured with 50 msec test pulses. The obtained inactivation curves were fit with a Boltzmann function of the form: I_Ba =I_max/(1 + exp[(V_m −V_1/2)/s]), where the current amplitude I_Ba has decreased to a half-amplitude at V_1/2 with ane-fold change over s mV. Pertinent parameters of the fits are given in Table 1.

The development of steady-state inactivation was studied by holding the cells for 1.2 sec at potentials ranging successively from −80 through +40 mV before a 50 msec step depolarization to a test potential of +30 mV. Figure 2B shows that the inactivation of the α_1C currents occurs at relatively depolarized holding potentials; this has been reported also for other expressed high-voltage-activated channels (De Waard and Campbell, 1995). More importantly, coexpression of either α₂δ or δ regulatory subunits induced a ∼10 mV hyperpolarizing shift in the voltage for 50% steady-state inactivation (Table 1).

Effect of the auxiliary subunits on current kinetics

We also investigated whether other properties of the expressed channels such as the inactivation kinetics could be modified by the presence of the regulatory subunits. Hence, the effect of the α₂δ and δ subunits on inactivation was estimated from the percentage of current remaining after 150 msec activating pulses (Fig. 3A). The presence of either regulatory subunit decreased significantly the percentage of the remaining current at the end of the depolarizing pulse, indicating that both α₂δ and δ subunits were able to accelerate inactivation of the α_1C currents.

Fig. 3.

Effect of the α₂δ complex on inactivation kinetics in transiently transfected tsA201 cells.A, The percentage of current remaining 150 msec into the depolarizing pulse is plotted at various membrane potentials in α_1C (•), α_1C/α₂δ (▪), and α_1C/δ (□) expressing cells. Astraight line provided a close fit to these data, and although of no theoretical significance, it was used to emphasize differences between singly transfected and cotransfected cells (n = 10). B, Representative records of currents obtained from three different transfected cells. The currents were recorded in response to 1 sec depolarizing pulses from −80 to +20 mV and ranged from −70 to −150 pA but are shown normalized to allow for comparison of kinetics. The inactivating phase of the currents was fit (superimposed lines) with a single exponential equation of the form: I_Ba= Aexp(−t/τ) + c, whereA is the initial amplitude, t is time, τ is the time constant for inactivation, and c is a constant. C, Time constants of α_1Ccurrents are plotted at various membrane potentials. The same cells were used in B and C and represent typical values in each group (n = 3 cells). Mean ± SE values are given in Table 1.

To analyze this action in more detail, we estimated voltage-dependent inactivation from the time course of the currents during 1 sec activating pulses to +20 mV by fitting the decaying component with a single exponential equation. The use of Ba²⁺ as the charge carrier in these experiments minimized any Ca²⁺-dependent inactivation. Traces in Figure3B exemplify normalized representative records of membrane currents in singly transfected and cotransfected cells and show that variation in channel subunit composition results in different inactivation behavior. The decay of the macroscopic currents recorded from cells expressing α_1C is slow as compared with those produced by coexpression of α_1C with either the full-length α₂δ subunit or the δ subunit. Figure3C compares the time constants of the currents expressed as a function of the step voltage and clearly indicates that the currents induced in the presence of the regulatory subunits inactivate faster than the currents recorded in cells transfected with only the α_1C subunit.

To test the individual contribution of the α₂ and the δ domains on the inactivation kinetics of the macroscopic currents and to determine whether the molecular determinants involved in the α₂δ subunit regulation were conserved across L-type and non-L type Ca²⁺ channels, we performed additional electrophysiological recordings in Xenopus oocytes expressing the neuronal α_1A subunit. Because the α₂ domain is completely extracellular (Gurnett et al., 1996; Wiser et al., 1996) and does not bind directly to the α₁ subunit in the absence of the δ domain (Gurnett et al., 1997), the interaction between these two proteins is difficult to assess. To overcome this problem, we created an α₂δ chimeric subunit in which the transmembrane and the cytoplasmic regions of the δ subunit were substituted by equivalent sequences of adhalin, an unrelated type I transmembrane protein (α₂Ad; Gurnett et al., 1996). This chimera has been shown to coimmunoprecipitate with the α₁ subunit (Gurnett et al., 1997) and allowed us to test the participation of the α₂ and the δ domains on the acceleration of the inactivation rate of the α₁expressed currents.

In agreement with previous electrophysiological studies inXenopus oocytes (Mori et al., 1991; Williams et al., 1992;Ellinor et al., 1993), we observed that expression of the neuronal Ca²⁺ channel α_1A pore-forming subunit alone generally resulted in small current density, although high concentrations of Ba²⁺ were used to increase the resolution of the currents. However, coexpression of the β₄ subunit allowed to us to increase the expression of the α_1A subunit to levels that permitted a systematic characterization of the biophysical properties of inactivation as detailed below.

Figure 4A compares the time course of normalized representative traces during 2 sec depolarizing steps from −90 to 0 mV recorded in oocytes injected with various subunit combinations: α_1A/β₄; α_1A/β₄/α₂δ, and α_1A/β₄/α₂Ad. In the same manner as in tsA201 transfected cells, the α₁currents in oocytes decayed during sustained depolarization as a result of the voltage-dependent inactivation of the channels. However, the time course of this inactivation consisted of two kinetic components: the first component followed a fast time constant and constituted ∼35% of the total inactivating current, and the second component was represented by a slower time constant and constituted the remaining ∼65%. As illustrated in Figure 4B, the average value of the time constant for the fast component (τ_fast) was approximately the same in the three sets of oocytes investigated. Conversely, the time constant for the slow component (τ_slow) was clearly faster in the α₂δ subunit-expressing oocytes than in the other two groups of cells. In addition, the amplitude of the slow component (Amp) also was reduced greatly in the oocytes injected with the α₂δ subunit. The expression of the chimera (α₂Ad) did not affect the inactivation kinetics significantly, although there was a slight but not statistically significant reduction in both the time constant and the amplitude of the slow component of the current. These results suggest that the presence of an intact δ domain, including transmembrane sequence, is necessary for the acceleration in the inactivation rate of the α₁ currents on α₂δ subunit expression.

Fig. 4.

Time course of inactivation of the currents through recombinant calcium channels expressed inXenopus oocytes. A, Normalized representative currents for oocytes coexpressing the α_1Aand the β₄ subunits in the absence or in the presence of the full-length α₂δ or the chimeric α₂Ad subunits. In all cases the decay of the currents was poorly fit by a single exponential. The sum of two exponential functions was necessary:I_Ba = I_∞+A_f(exp(−t/τ_f) +A_s(exp(−t/τ_s) + c, where I_∞ is the steady-state inward current, A is the amplitude,t is time, and τ_f and τ_s are time constants for the fast and slow components, respectively.B, Pooled data comparing the average time constants (τ_fast and τ_slow) and relative contribution of the slow component of inactivation amplitude (Amp) at 0 mV from various oocytes in the three groups (n = 6). Asterisks denote significant differences (p < 0.05).

Modulation of the DHP binding

To characterize the DHP-binding affinity of the α_1C-transfected cells and its possible modulation by the α₂δ subunit, we performed equilibrium radioligand binding experiments that used [³H]-PN200-110. In an initial series of experiments we observed that α_1Ccurrents were sensitive to micromolar concentrations of DHPs (data not shown). Accordingly, Figure 5Ashows representative saturation isotherms of [³H]-PN200-110 binding to microsomes from α_1C, α_1C/α₂δ, and α_1C/δ-coexpressing cells. Specific binding was not observed in untransfected cells, consistent with the absence of endogenous Ca²⁺ currents and the absence of detectable levels of Ca²⁺ channel subunit protein expression. Scatchard analysis of the [³H]-PN200-110 saturation binding data revealed very similar values for the apparent dissociation constant (K_D) and the total number of binding sites (B_max) on both singly transfected cells and cells coexpressing the δ subunit. In marked contrast, when singly transfected cells and cells coexpressing the α₂δ subunit were compared, a significant increase in both the number of binding sites and in affinity for the drug was observed (Table 1).

Fig. 5.

Dihydropyridine binding to the expressed Ca²⁺ channels. A, Equilibrium binding curves for [³H]-PN200-110 in cells transfected with α_1C alone or in combination with the α₂δ or the δ subunits. Symbolsrepresent mean values of specific DHP binding of duplicate samples in one representative experiment. Data were fit with a one-site ligand-binding equation of the form: y = [L] · Cap/(K_D + [L]), where L is the concentration of free ligand, Cap is the maximum bound ligand, andK_D is the dissociation constant.B, Comparison of specific [³H]-PN200-110 binding to microsomes of tsA201 cells transfected with α_1C alone or in combination with α₂δ or δ subunits. Binding assays were performed with 1 nm [³H]-PN200-110. Data are given as mean ± SE, and the number of separate experiments is indicated in parentheses. Asterisksdenote significant differences (p < 0.05).

To elucidate the mechanisms underlying the large α₂δ-induced increase in α_1C current amplitude, we compared current amplitude and DHP binding in both singly transfected and cotransfected cells. Because DHPs bind to the α_1C subunit, an increase in the number of binding sites could account for the increase in current magnitude. Figure5B shows the comparison of specific DHP binding to microsomes from tsA201 transfected cells, using a near-saturating concentration of [³H]-PN200-110 (1 nm). Coexpression of α₂δ subunit resulted in a ∼3.6-fold increase in the amount of [³H]-PN200-110 binding. When it is compared with the approximately threefold increase in current amplitude seen by coexpression of the α₂δ subunit, this comparison suggests that there is an important correlation between the ability to increase current amplitude and the ability to enhance DHP binding. Consistent with this, the δ subunit was incapable of increasing either the current amplitude or the DHP binding (Table 1).

DISCUSSION

We have analyzed the effects induced by interaction of the α_1C subunit with the α₂δ and δ auxiliary subunits on the fundamental biophysical properties of the recombinant channels. Our data clearly show that the α₂δ subunit significantly increased L-type calcium channel activity in transiently transfected tsA201 cells (Fig. 1). Likewise, as a result of this interaction, the α₂δ subunit induced a hyperpolarizing shift in activation and steady-state inactivation (Fig. 2) of the expressed channels. Interestingly, we found that the δ subunit reproduced entirely the effects of the full-length α₂δ subunit on the voltage dependence of the macroscopic α_1C currents. These findings therefore provide evidence that the δ subunit per se can modulate the voltage-dependent behavior of the L-type Ca²⁺channel by interacting with the α_1C subunit. Moreover, it may localize the region involved in modulation of voltage dependence to the δ domain of the α₂δ complex such that the interaction between these two proteins may affect the S4 voltage sensor.

Another manifestation of the interaction between the α_1Csubunit with the regulatory subunits is the acceleration of the inactivation rate. The traces in Figure 3B show that, during depolarization, inward currents in transfected tsA201 cells spontaneously decay in external Ba²⁺, indicating that the expressed calcium channels undergo voltage-dependent inactivation. The time course of the decaying component of these currents is well described with a single exponential equation. We found that expression of the α₂δ subunit influenced the rate at which the currents inactivated. Again, the δ domain completely reproduced the effect of the full-length α₂δ complex on inactivation kinetics (Table 1). Consistent with the above-mentioned findings, the percentage of decay of the α_1C currents was increased significantly in the presence of either regulatory subunit when inactivation was estimated from the percentage of current remaining after 150 msec activating pulses (Fig. 3A, Table1).

In contrast with the monoexponential decay of the currents in the transfected tsA201 cells, the Ca²⁺ channels expressed in Xenopus oocytes inactivated in a biexponential manner (Fig. 4). Although many factors may give rise to two components, several experimental findings suggested to us that this result may reflect an intrinsic property of the neuronal expressed channels: the endogenous currents were negligible; the ratio of the slow and fast current amplitudes remained constant from oocyte to oocyte and was independent of the total current amplitude recorded. Moreover, our results indicate that the slow inactivating component of these currents was sensitive to the regulatory effect of the α₂δ subunit. However, coexpression of the α₂Ad chimera, in which the transmembrane domain of the α₂δ complex (δ subunit) was replaced with that of adhalin, did not modify the inactivation kinetics of the α_1A/β₄channels. This suggests that the δ transmembrane domain may be the primary moiety involved in this regulation. Interestingly, because the molecular determinants of voltage-dependent inactivation in Ca²⁺ channels have been localized to the membrane-spanning segment S6 of the first repeat (IS6) of the α₁ subunit (Zhang et al., 1994), this region also may be involved in the interaction with the δ subunit.

Another regulatory action of the α₂δ complex was the drastic increase in α_1C current amplitude. The mechanisms underlying this action may be better understood by comparing current amplitude and DHP binding in both singly transfected and cotransfected cells. Our results demonstrated that the total number of DHP binding sites increased on coexpression of the α₂δ subunit. In addition, there was a large augmentation in the affinity for the DHP when α₂δ was present (Fig. 5, Table 1). These results confirm that the α₂δ subunit is crucial to the reconstitution of DHP binding (Wei et al., 1995), because the binding affinity and number of binding sites approached that of rabbit cardiac microsomes (Nishimura et al., 1993; Wei et al., 1995) only when the α₂δ subunit was coexpressed. Furthermore, these observations suggest that the α₂δ complex acts primarily by inducing important conformational changes in the pore-forming subunit. These changes in α_1C conformation then would be responsible for an increase not only in the accessibility of the drug to its site but also in the opening probability of the channel, as it has been observed in Xenopus oocytes (Shistik et al., 1995). Because the binding site for DHPs has been localized to the IIIS5–S6 and IVS5–S6 regions (Grabner et al., 1996; Peterson et al., 1996), these sites also may have been involved in the interaction with the α₂δ complex.

An alternative possibility to explain the results could be that the conformational changes induced by the α₂δ complex may play an important role in the localization of the expressed channels on the cell surface. In support of this interpretation, Bangalore et al. (1996) have demonstrated that the coexpression of the α₂δ complex increases the number of functional L-type calcium channels in the cell membrane as gauged by gating charge movement. In addition, Shistik et al. (1995) have shown that the α₂δ complex triples the amount of α_1Cprotein localized in the plasma membrane of Xenopus oocytes as detected by immunoprecipitation. As mentioned above, our results indicate that coexpression of the L-type calcium channel pore-forming subunit with α₂δ increased both the number of the total binding sites and the affinity for the radiolabeled PN200-110. In contrast, coexpression with the δ subunit did not affect specific DHP binding (Fig. 5, Table 1). Interestingly, the chimeric α₂Ad subunit has been shown to mimic the effects of the full-length α₂δ subunit on PN200-110 binding when expressed in tsA201 cells (Gurnett et al., 1997). These findings indicate not only that expression of the α₂ domain is necessary for the formation of a stable interaction capable of reconstituting normal DHP binding but also suggest that coexpression of this domain of the α₂δ protein may facilitate proper insertion of channel proteins into the cell membrane. The mechanisms of the α₂ subunit effect on the membrane trafficking of the α₁ subunit require further investigation. Gating current studies will be needed to examine this issue.

Taken together, our findings indicate that the α₂δ complex and the δ subunit interact in specific ways with the α₁ subunit and participate in the functional regulation of the L-type and non-L type calcium channels. Our previous studies indicated that the extracellular α₂ domain, which is particularly sensitive to structural modification, provides the elements required for channel stimulation (Gurnett et al., 1996). Here, we localize the region of the α₂δ subunit involved in the shift in voltage-dependent activation and steady-state inactivation as well as the acceleration of the inactivation kinetics to the δ subunit, whereas the effects on increased currents and DHP binding affinity require the presence of the α₂ domain of the α₂δ complex.