INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Despite their typical sensitivity to the dihydropyridines (DHPs), the predominant voltage-gated Ca²⁺ current in hair cells displays atypical biophysical and functional properties. The current activates and deactivates with fast time constants ( = approximately 0.5 ms), shows weak inactivation, and activates at potentials approximately 20 mV more negative than the activation voltages of other DHP-sensitive currents (Hudspeth and Lewis 1988; Rodriguez-Contreras and Yamoah 2001; Zidanic and Fuchs 1995). These properties allow hair cells to respond to tonic and phasic changes in vestibular and auditory stimuli through Ca²⁺-mediated neurotransmitter release.

The DHP-sensitive Ca²⁺ channel in hair cells has been cloned from the chick basilar papilla and identified as _1D (Kollmar et al. 1997a) or Ca_V1.3 ₁ channel (Ertel et al. 2000). In contrast to other Ca²⁺ channels, wherein the heterologous expression of the pore-forming  subunit closely mirrors the native channel current, the biophysical properties of the expressed Ca_V1.3 are quite distinct from typical hair cell Ca²⁺ currents (Bell et al. 2001; Koschak et al. 2001; Mikami et al. 1989; Seino et al. 1992; Williams et al. 1992; Xu and Lipscombe 2001). Although it has been suggested that splice variants of Ca_V1.3₁ channel in the chick basilar papilla may suffice to confer the unique properties of the DHP-sensitive Ca²⁺ currents in hair cells (Kollmar et al. 1997b), in other systems, there is ample evidence to suggest that co-assembly of auxiliary  and ₂ subunits with the ₁ subunit may contribute to the properties of the native Ca²⁺ current. For example, co-expression of the ₂ and  subunits approximates the voltage dependence and kinetics of gating of ₁-subunit in the native cardiac Ca²⁺ current properties, as well as enhances the level of expression of the channels (Lacerda et al. 1991; Perez-Garcia et al. 1995; Wei et al. 2000).

The rapid kinetics of synaptic transmission at the hair cell afferent nerve terminals require the influx of Ca²⁺, which is supplied mainly by the Ca_V1.3 channels (Zidanic and Fuchs 1995). The strategic placement of the channels with the exocytotic apparatus is crucial for fast neurotransmitter release, as observed in the synaptic terminals of the squid giant axon (Adler et al. 1991). Previous studies from the presynaptic terminals of the squid giant axons, the rabbit retina, and chick ciliary neurons have identified ~10-nm particles that may represent Ca²⁺ channels assembled approximately <100 nm from synaptic vesicle docking sites (Heuser and Reese 1973; Mennerick and Matthews 1996; Stanley 1993). Indeed, freeze-fracture images of the active zones of hair cells have yielded findings consistent with the morphological association of the channels with synaptic proteins (Issa and Hudspeth 1994).

The goal of this study was to determine whether the known  and ₂ subunits and synaptic proteins can modulate the Ca_V1.3₁ channel to confer the kinetic and voltage-dependent properties that may approximate the DHP-sensitive current phenotype in hair cells. Our results are as follows: first, co-expression of Ca_V1.3₁ with _1A and ₂ subunits, as well as with syntaxin, enhanced the functional expression of the channel. Second,  and ₂ subunits produced at least a 9-mV hyperpolarization shift in the steady-state activation curves. Third, co-expression with syntaxin sufficed to confer the activation range of the native current and reduced the inactivation profile of the Ca_V1.3₁ Ca²⁺ current. Finally, to further confirm these findings in native systems, we were able to demonstrate that botulinum toxin C, an inhibitor of syntaxin, enhanced the inactivation of the Ca²⁺ currents in hair cells. Thus the interaction of Ca_V1.3₁ with auxiliary subunits and synaptic proteins modifies the functional expression of the channels and contributes to their physiological properties.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Hair cell preparation

Hair cells were isolated from the sacculus of the bullfrog (Rana catesbeiana, 3-4 in) using solutions and procedures described previously (Rodriguez-Contreras et al. 2002; Yamoah et al. 1998).

Cell culture and transfection

HEK 293 cells (a human embryonic kidney cell line) were cultured at 37°C in minimum essential medium (MEM, Invitrogen, Carlsbad, CA) containing 10% horse serum, (heat-inactivated, Invitrogen) and 1% penicillin-streptomycin (P0906, Sigma), and were incubated at 5% CO₂. Cells were transiently transfected using calcium-phosphate precipitation procedure (kit from Invitrogen) as described previously (Perez-Garcia et al. 1995). Channel subunits to be studied were subcloned into pGW1H, an expression vector using a cytomegalovirus promoter (British Biotechnology, Oxford, UK). A 7.5-µg plasmid containing the gene that encodes the pore-forming subunit was included (Ca_V1.3₁, derived from pancreatic beta cells), and mixed with 5 µg of plasmid containing the gene that encodes auxiliary subunits [none, a _1A subunit (derived from skeletal muscle) and/or the ₂ subunit, brain-derived, and syntaxin, vesicle-associated membrane protein (VAMP), and synaptosome-associated protein of 25 kDa (SNAP25), brain-derived]. Our goal was to test whether known Ca²⁺ channel subunits that can be expressed functionally are sufficient to produce the properties of L-type current in hair cells. The _1A subunit is expressed abundantly in hair cells, but _2A is not (Green et al. 1996). To allow detection of transfected cells, the S65T bright green fluorescent protein (GFP) mutant in GFPpRK5 vector was used to co-transfect the cells to quantify the transfection efficiency. Electrophysiologic recordings were performed only from GFP-positive cells. The GFPpRK5 plasmids were used with all transfections.

Electrophysiology

Ca²⁺ currents were recorded as described (Perez-Garcia et al. 1995) using the whole cell patch-clamp technique from transiently transfected HEK 293 cells and hair cells from the bullfrog saccule (Rodriguez-Contreras and Yamoah 2001) at room temperature. Outward K⁺ currents were blocked with Cs⁺. The cell capacitance was calculated as the ratio of total charge (the integrated area under the current transient) to the magnitude of the pulse (20 mV). Data were stored in a computer for analysis.

Recording solutions

The same pipette and bath solutions were used to record whole cell Ca²⁺ currents from transiently transfected HEK 293 cells and hair cells to allow direct comparison of the two preparations. Patch electrodes were filled with solution containing (in mM) 120 CsCl₂, 5 EGTA, 10 HEPES, and 3 Mg-ATP (pH 7.2, CsOH). The composition of the bath solution was as follows (in mM): 40 CaCl₂, 80 CsCl, 10 HEPES, and 10 glucose (pH 7.4, CsOH). Stock solutions of nimodipine and BayK 8644 (Calbiochem, La Jolla, CA) were dissolved in 100% dimethyl sulfoxide (DMSO), stored at 20°C, and further diluted using the bath solution just before use. Botulinum toxin C was reconstituted and activated by incubating in DTT (5 mM) for 2 h at 37°C and added to the internal solution containing (in mM) 100 CsCl, 20 TEACl, 5 HEPES, 5 EGTA, 0.1 CaCl₂, and 2 Mg-ATP. The final concentration of the botulinum toxin was 200 nM. The external solution for these experiments contained (in mM) 5 CaCl₂, 98 NaCl, 5 4-AP, 20 TEACl, 3 D-glucose, and 5 HEPES.

Data analysis

Whole cell Ca²⁺ current amplitude at varying test potentials was measured at the peak and steady-state levels using a peak and steady-state detection routine; the current was divided by the cell capacitance (pF) to generate the current density-voltage relationship. The mean capacitance of HEK 293 cells was 39 ± 6 pF (n = 72) and the capacitance of hair cells was 14 ± 3 pF (n = 18). Analysis of data were performed using custom-made software and Microcal Origin (Northampton, MA) programs. Where appropriate, pooled data are presented as means ± SE.

Immunocytochemistry

Isolated hair cells were fixed and permeabilized as described by Rodriguez-Contreras and Yamoah (2001). Cells were incubated for 1 h in a blocking solution of frog saline containing 1% bovine serum albumin and 1% horse serum, followed by 3-5 h incubation with Ca²⁺ channel antibodies anti-Ca_V1.3₁ (Alomone Laboratories, Jerusalem, Israel) and anti-syntaxin (Sigma) at dilutions of 1:100 to 1:200. Cells were then incubated with secondary antibodies conjugated with fluorescein and rhodamine for 2 h, washed, and viewed with a Zeiss LSM 510 confocal microscope. Control experiments using antigenic peptides have been shown previously to block the antibody binding reaction (Rodriguez-Contreras and Yamoah 2001).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have previously demonstrated that saccular hair cells with length-to-apical diameter (LAD) ratio of 4-6 mostly expressed (>95%) the DHP-sensitive Ca²⁺ currents (Rodriguez-Contreras and Yamoah 2001). Only cells with LAD ratio of 4-6 were used in this study. Whole cell Ca²⁺ current recordings were carried by 40 mM Ca²⁺ in both hair cells and HEK 293 cells to allow for direct comparison.

DHP-sensitive current in hair cells differs from Ca_V1.3₁ currents in HEK 293 cells

Figure 1 shows examples of Ca²⁺ current traces generated from hair cells. In agreement with earlier studies (Hudspeth and Lewis 1988; Martini et al. 2000; Smotherman and Narins 1999; Zidanic and Fuchs 1995), the DHP-sensitive Ca²⁺ current in hair cells has a rapid onset and inactivates slowly. Despite the fact that the channel has been identified as Ca_V1.3₁ channel (Kollmar et al. 1997a), the current profile of the expressed-channel in HEK 293 cells is in sharp contrast to that in hair cells. The Ca_V1.3₁-channel current not only activates more slowly than the hair-cell DHP-sensitive current, it also shows robust inactivation, which can be fit with two time constants of inactivation (_inact; Table 1; Fig. 1B). By contrast, the _inact of the current in hair cells is >1 s. This is exemplified by the superposition of hair cell and Ca_V1.3₁ currents in Fig. 1B. The current-density-voltage (I-V) relationship of the DHP-sensitive current in hair cells and the _1D currents yielded curves that peaked at approximately 5 and 20 mV, respectively (Fig. 1A). Because the recordings were performed using identical di/mono-valent cation conditions, the difference in voltage-dependence of activation as illustrated in Fig. 1C could not have resulted from surface charge screening effects. With 40 mM Ca²⁺ as the charge carrier, the half-activation voltage of the currents in hair cells was approximately 18 mV below that of the activation curve for the Ca_V1.3₁ currents.

View larger version (33K): [in this window] [in a new window]   Fig. 1. Averaged, peak current density-voltage plots for CaV1.31 currents in HEK 293 cells () and hair cells () using 40 mM Ca2+ as the charge carrier. The peak current occurred at approximately 20 and 5 mV for CaV1.31 currents in HEK 293 cells and hair cells, respectively. The peak for CaV1.31 currents was 92 ± 15 pA (n = 10). Representative current traces for hair cells, CaV1.31 currents in HEK 293 cells, and effects of Bay K8644 are shown in the inset. B: currents from CaV1.31 in HEK 293 cells and hair cells were normalized to compare the activation and inactivation profiles. C: steady-state activation of CaV1.31 currents in HEK 293 cells and hair cells. The conductance (g) of Ca2+ through the channels was calculated using the equation [g = I/(V  ECa)]; where I is the current at step voltage (V) and ECa is the reversal potential determined from the instantaneous current-voltage relations. The ratio of g to the maximum gmax is plotted against voltage. The activation was fitted with the Boltzmann equation, g/gmax = {1 + exp[(V1/2  V)/k]}1, where V1/2 is the half-activation and k is the slope factor. V1/2 (mV) for the means at 40 mM Ca2+ for the DHP-sensitive current in hair cells was 19.3 ± 2.4 (n = 5) and that of the CaV1.31 currents in HEK 293 cells was 1.3 ± 1.9 (n = 11). The slope factor k (mV) for DHP-sensitive current in hair cells and CaV1.3a1 currents in HEK 293 were 6.3 ± 0.4 and 6.1 ± 0.5, respectively. D: auxiliary subunits 1A and 2 increased the level of expression and produced a leftward shift in the voltage-dependent activation of the CaV1.3a1 currents (Table 1). The CaV1.3a1 channels were expressed alone (), together with the 1A subunit (), or with both the 1A and the 2 subunits (). Shown in the insets are representative current traces generated from CaV1.31 channel and co-expression with its auxiliary subunits. The peak currents for CaV1.31 + 1A, CaV1.31 + 2, and CaV1.31 + 1A + 2 were 176 ± 14 (n = 9), 97 ± 7 (n = 7), and 183 ± 9 pA (n = 9), respectively.

View this table: [in this window] [in a new window]   Table 1. Properties of hair-cell DHP-sensitive current, CaV1.31

Whole cell Ca²⁺ currents were recorded from HEK 293 cells transfected with Ca_V1.3₁ (_1D) channels and four different combinations of subunits: Ca_V1.3₁ alone, Ca_V1.3₁/_1A, Ca_V1.3₁/₂, and Ca_V1.3₁/₂/_1A. For HEK 293 cells that were transfected with _1D alone, step depolarization > 25 mV, from a holding potential of 60 mV, activated inward Ca²⁺ currents which peaked at approximately 20 mV. Co-expression of the Ca_V1.3₁ with _1A subunit increased the current density approximately twofold and produced a significant negative shift in the steady-state activation curve by approximately 9 mV (Fig. 1D). In sharp contrast to the effects of the _1A on the voltage-dependent activation, co-expression of the ₂ subunit did not produce a shift in the V_1/2 of the steady-state activation curve (Table 1). Moreover, the increase in the level of expression and negative shift in V_1/2 induced by the _1A subunit were maintained when Ca_V1.3₁/₂/_1A was co-expressed (Fig. 1D). Next, we examined the effects of the auxiliary subunits on the extent of inactivation. None of the auxiliary subunits altered the steepness of the inactivation curves. However, the  subunit modified the extent of voltage-dependent inactivation of the Ca_V1.3₁ subunit currents by producing a rightward shift in the half-inactivation voltage by approximately 3 mV. Although the ₂ alone did not change the extent of inactivation, it enhanced the  subunit-induced attenuation of the inactivation (data not shown). The number of channels expressed in HEK 293 cells were estimated by assuming that the unitary current amplitude, using 40 mM Ca²⁺ as the charge carrier, is approximately 0.6 pA (Rodriguez-Contreras et al. 2002) at the peak voltage and that the probability of channel openings (P_o) are similar under whole cell and single-channel recording conditions (Table 1).

Effects of synaptic proteins on the voltage dependence of activation and inactivation of recombinant Ca_V1.3₁ currents

Ca²⁺ channels at presynaptic terminals may be located as close as approximately 10 nm from the docking site for a synaptic vesicle, and there is evidence of physical and functional interaction between the channels and synaptic proteins (Atlas 2001; Degtiar et al. 2000; Stanley and Mirotznik 1997). The Ca_V1.3₁ channel is the predominant presynaptic Ca²⁺ channel that mediates transmitter release in hair cells. We predicted that some of the atypical features in the Ca_V1.3₁ current in hair cells might be shaped by the channel's interaction with synaptic proteins. We sought to delineate and characterize the role of synaptic proteins in the voltage-dependent properties of the Ca_V1.3₁ channels.

Transfection of HEK 293 cells with Ca_V1.3₁ subunit in combination with various synaptic proteins led to the expression of well-resolved Ca²⁺ currents carried by recombinant Ca²⁺ channels. The relative magnitudes of the various sets of I-Vs show that addition of syntaxin caused a striking increase in the current density (approximately twofold), which was enhanced substantially by the addition of the  subunit (Fig. 2A). Figure 2B compares activation curves for the Ca_V1.3₁ subunit current and synaptic protein combinations that produced marked effects. The largest effect was the approximate 18-mV negative voltage shift in the V_1/2 of the steady-state activation curves with the co-expression of _1A, syntaxin, and VAMP (Table 1). Although the inclusion of SNAP25 and VAMP with the ₁ did not alter the magnitude of the expressed current, it produced an approximate 3-mV positive voltage shift in the V_1/2 of the ₁/₁/syntaxin combination. This is consistent with reports that demonstrate that SNAP25 and VAMP are associated only with Ca²⁺ channels that are bound to syntaxin (Martin-Moutot et al. 1996). Addition of ₂ subunit to the various subunit combinations and synaptic proteins yielded no appreciable current enhancement and no change in the voltage-dependent activation.

View larger version (32K): [in this window] [in a new window]   Fig. 2. Functional modulation of CaV1.3a1/b1A currents by synaptic proteins. A: current density-voltage relationship from a total of 11 cells are shown data from CaV1.3a1 (), CaV1.3a1/syntaxin (), CaV1.3a1/1A/syntaxin (), and CaV1.3a1/1A/syntaxin/VAMP (). B: apart from the increase in current-density (shown in A), there was a leftward shift in the voltage-dependent activation curve compared with CaV1.3a1 alone (curve in dotted line; Table 1). Representative current traces are shown in the inset. C: effects of syntaxin and syntaxin/VAMP on the inactivation profile of CaV1.3a1/1A currents are shown. The peak currents for CaV1.31 + syntaxin, CaV1.31 + 1A + syntaxin, and CaV1.31 + 1A + syntaxin + vamp were 193 ± 8 (n = 11), 307 ± 17 (n = 11), and 299 ± 21 pA (n = 11), respectively. D: botulinum toxin C, an inhibitor of syntaxin, accelerated the inactivation of DHP-sensitive current in hair cells (n = 6). The recording electrodes were filled with 200 nM botulinum toxin C. Hair cells were held at a holding potential of 60 mV and stepped to 10 mV test potential. After 150 s of sufficient dialysis of the cells, the inactivation of the current was enhanced.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The auxiliary subunit- and synaptic protein-mediated enhancement of the macroscopic Ca²⁺ currents could stem from an increase in the unitary current amplitude, the number of functional channels, and a rise in P_o, or a combination of these properties. Previous studies have demonstrated that co-expression of the ₁ and auxiliary subunits of Ca²⁺ channels produced enhanced protein expression but may not alter the unitary current amplitude (Kamp et al. 1996). The synaptic protein-induced effects may be complimentary to the actions of the auxiliary subunits; however, further studies are required to address this question.

We next sought to evaluate the effects of synaptic proteins on the extent of inactivation of the Ca_V1.3₁ current. Co-expression of syntaxin with Ca_V1.3₁ plus _1A prolonged the inactivation time constants as exemplified in the traces illustrated in Fig. 2C and summarized in Table 1. Addition of VAMP did not accentuate the effects of syntaxin on the extent of inactivation (Fig. 2C). Similar to the effects on the voltage-dependent activation, SNAP25 did not alter the extent of inactivation of the Ca_V1.3₁ current. Thus syntaxin and VAMP may modulate the Ca_V1.3₁ current functionally.

Botulinum toxin C is an endopeptidase neurotoxin that cleaves syntaxin 1A, disrupting its interaction with Ca²⁺ channels and inhibiting neurotransmitter release (Degtiar et al. 2000; Rettig et al. 1996; Sheng et al. 1994). To test whether the effects of syntaxin on Ca_V1.3₁ current in HEK 293 cells also occur in hair cells, we directly examined the role played by botulinum toxin C on hair cell Ca²⁺ currents. Consistent with the effects of syntaxin on the expressed Ca_V1.3₁ current, dialysis of hair cells with botulinum toxin C noticeably altered the inactivation of the Ca²⁺ current (Fig. 2D, n = 6). The present findings accord with other studies that have shown that co-expression of synaptic proteins, e.g., syntaxin 1A with Ca_V1.2 and Ca_V2.2, reduces the rate of inactivation (Wiser et al. 1996). Whereas VAMP potentiates the inhibitory effects of syntaxin on the Ca_V1.2 and Ca_V2.1 currents, SNAP25 reverses (Atlas 2001; Bajjalieh 1999; Zhong et al. 1999). The effects of SNAP25 on the Ca_V1.3 current, however, were minimal. Localization of Ca²⁺ channels and syntaxin at the basal pole of hair cells with anti-Ca_V1.3₁ and anti-syntaxin revealed that the two proteins are expressed closely together (Fig. 3), in accordance with previous co-immunoprecipitation and fluorescent-tagged -subunit experiments that demonstrated the association of syntaxin 1A with Ca_V2.2 and Ca_V1.3 channels (Bennett et al. 1992; Yang et al. 1999). The present findings show that in a heterologous expression system, the association of the , , and ₂ subunits and synaptic proteins approximate the voltage dependence and inactivation kinetics of the native properties of the Ca_v1.3₁ in hair cells. Although auxiliary subunits and the synaptic proteins did not restore completely the native Ca²⁺ current properties, this may result from the limitation of reconstituting multiple (5-6) different DNAs in HEK 293 cells. For example, by reducing the Ca_v1.3₁ threefold and increasing the _1A, syntaxin, and VAMP twofold, the resulting current showed little or no inactivation (_inactivation = 1.2 ± 0.5 s, n = 5). However, as expected, the magnitude of the current was small (0.5 + 0.1 pA/pF, n = 5). Furthermore, as suggested in a previous report, splice variants in the Ca_V1.3₁ channel (Kollmar et al. 1997b) may provide additional differences that would determine distinct properties of hair cell L-type channel currents. Thus reconstitution of the native hair cell Ca²⁺ current phenotype may require the hair cell-specific splice variant channel as well as their auxiliary subunits and synaptic proteins.

View larger version (21K): [in this window] [in a new window]   Fig. 3. Immunocytochemistry showing co-localization of CaV1.31 and syntaxin at the basal pole of hair cells. A-C: basal half of a hair cell that was treated with fluorescein-conjugated secondary antibodies (A, green), rhodamine-conjugated (B, red) secondary antibodies, and double-labeling (C), with A and B showing no positive staining. In contrast, hair cells labeled with anti-syntaxin antibody, which was detected with rhodamine-conjugated secondary antibody (D) and anti-1D antibody detected with fluorescein-conjugated secondary antibody (E), showed positive staining and co-localization of the two proteins as represented in F, indicating double-labeling of D and E. Scale bar = 5 µm.