This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kobrinsky, E. Articles by Soldatov, N. M. Search for Related Content PubMed PubMed Citation Articles by Kobrinsky, E. Articles by Soldatov, N. M.

Voltage-Gated Rearrangements Associated with Differential ß-Subunit Modulation of the L-Type Ca²⁺ Channel Inactivation

Evgeny Kobrinsky ^*, Klaus J. F. Kepplinger , Alexander Yu ^*, Jo Beth Harry ^*, Heike Kahr , Christoph Romanin , Darrell R. Abernethy ^* and Nikolai M. Soldatov ^*

^* National Institute on Aging, National Institutes of Health, Baltimore, Maryland 21224 USA; and Institute for Biophysics, University of Linz, A-4040 Linz, Austria

Correspondence: Address reprints requests to N. M. Soldatov, National Institute on Aging, National Institutes of Health, 5600 Nathan Shock Dr., Baltimore, MD 21224. Tel.: 410-558-8343; Fax: 410-558-8318; E-mail: soldatovN{at}grc.nia.nih.gov.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Auxiliary ß-subunits bound to the cytoplasmic ₁-interaction domain of the pore-forming _1C-subunit are important modulators of voltage-gated Ca²⁺ channels. The underlying mechanisms are not yet well understood. We investigated correlations between differential modulation of inactivation by ß_1a- and ß₂- subunits and structural responses of the channel to transition into distinct functional states. The NH₂-termini of the _1C- and ß-subunits were fused with cyan or yellow fluorescent proteins, and functionally coexpressed in COS1 cells. Fluorescence resonance energy transfer (FRET) between them or with membrane-trapped probes was measured in live cells under voltage clamp. It was found that in the resting state, the tagged NH₂-termini of the _1C- and ß-subunit fluorophores are separated. Voltage-dependent inactivation generates strong FRET between _1C and ß_1a suggesting mutual reorientation of the NH₂-termini, but their distance vis-à-vis the plasma membrane is not appreciably changed. These voltage-gated rearrangements were substantially reduced when the ß_1a-subunit was replaced by ß₂. Differential ß-subunit modulation of inactivation and of FRET between _1C and ß were eliminated by inhibition of the slow inactivation. Thus, differential ß-subunit modulation of inactivation correlates with the voltage-gated motion between the NH₂-termini of _1C- and ß-subunits and targets the mechanism of slow voltage-dependent inactivation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Calcium channels respond to membrane depolarization with an increase in membrane permeability that develops within a few milliseconds and then spontaneously decreases (inactivates) with characteristic kinetics. In the case of the L-type class C voltage-sensitive Ca²⁺ channel (Ca_v1.2), the replacement of Ca²⁺ with Ba²⁺ ions as the charge carrier eliminates Ca²⁺-induced inactivation, and the Ba²⁺ current inactivates essentially in a voltage-dependent manner. Given the critical role of Ca_v1.2 channels in cell signaling (Soldatov, 2003), investigation of molecular correlates of inactivation has been an objective of a number of laboratories, including those cited in this article.

The Ca_v1.2 channel is composed of the pore-forming _1C-subunit and the auxiliary ß- and ₂-subunits. All three subunits can be colocalized along the plasma membrane in transfected cells expressing the functional recombinant channel (Gao et al., 1999). The subunits remain associated with each other after solubilization of the channel protein complex in mild nonionic detergents (Chang and Hosey, 1988) suggesting close association between the subunits in vivo. In fact, binding motifs involved in interactions between the _1C- and ß-subunits (Pragnell et al., 1994) as well as between the _1C- and ₂-subunits (Gurnett and Campbell, 1996) have been identified.

The ß-subunits are important modulators of the Ca²⁺ channel activity (Birnbaumer et al., 1998; Yamaguchi et al., 1998; Colecraft et al., 2002; Hullin et al., 2003; Takahashi et al., 2003). They are located on the cytoplasmic side of plasma membrane and generate the molecular signal necessary for correct plasma membrane targeting of the functional Ca_v1.2 complexes. Each of the four identified different types of ß-subunit (ß₁–ß₄) binds, via the NH₂-terminal part of the second common conserved region (amino acids 262–290 in ß_1a and 215–243 in ß_2a), to the ₁-interaction domain (AID), a conserved motif in the cytoplasmic linker between repeats I and II of the ₁-subunit (Pragnell et al., 1994; De Waard et al., 1994). This binding was shown to be a molecular correlate for the well-established effect of ß-subunits on the kinetics of Ca_v1.2 channel gating (Singer et al., 1991). Inactivation of the Ca_v1.2 channel is accelerated by the replacement of ß₂ with the ß_1a-subunit. The modulatory effect of the ß-subunit was observed in other types of Ca²⁺ channels as well (Hummer et al., 2003). For example, in Ca_v2.1 (_1A) P/Q-type channel, De Waard and Campbell (1995) identified ß_2a as the least potent among four other types of ß-subunit in acceleration of inactivation. Thus, ß-subunits modulate both the functional expression and the gating of ion conductance.

Recently we have developed a new method for measuring dynamic structural responses to changes in the cell membrane potential (Kobrinsky et al., 2003). Using measurements of fluorescence resonance energy transfer (FRET) between enhanced cyan (ECFP) and yellow (EYFP) fluorescent proteins under voltage-clamp conditions in live cells, we provided evidence that ECFP/EYFP-tagged cytoplasmic tails of the Ca_v1.2 channel are moving parts, responsive to voltage gating. The voltage-dependent mobility of the COOH-terminal tail was found to be essential for Ca²⁺-dependent inactivation and Ca²⁺ signal transduction.

Here we have applied FRET imaging in voltage-clamped living COS1 cells to directly compare the interaction between _1C and different ß-subunits in different functional states of the channel, and to determine molecular motions of the ß-subunits vis-à-vis _1C and the plasma membrane. Two pore-forming _1C-subunits showing different inactivation properties were used. The _1C,77-subunit has the structure of the most common _1C splice variant and thus can be referred to as the "wild-type" isoform. The _1C,IS-IV-subunit gives rise to a channel that does not show the slow voltage-dependent inactivation (Shi and Soldatov, 2002). This isoform was generated by the mutation of the following four amino acids in the transmembrane segments S6 of the _1C,77-subunit: S405I in IS6, A752T in IIS6, V1165T in IIIS6, and I1475T in IVS6. These mutations are located in positions –1 (IS6) and –2 (IIS6–IVS6) and thus form a ring of altered amino acid residues at the inner mouth of the pore. The corresponding structure in the _1C,77 channel was named the "annular determinant of slow inactivation" (ADSI) reflecting the specific arrangement of crucial amino acids with regards to the ion-conducting pore.

The ß-subunit variants ß_1a, cloned from rabbit skeletal muscle (Mori et al., 1991), and ß₂, originally cloned from rabbit heart (Hullin et al., 1992), have been studied in this work as modulators of Ca²⁺ channel inactivation. The rabbit ß₂-subunit has 92.6% overall amino acid identity to the rat ß_2a protein (Perez-Reyes et al., 1992) with some variations (17 amino acids) at the NH₂-terminus, and lacks the NH₂-terminal cysteine sites of palmitoylation. The differential effect of these ß_1a- and ß₂-subunits on voltage-dependent inactivation of the _1C,77 channel expressed in Xenopus oocytes has been described (Soldatov et al., 1997). These results confirmed other reports on the acceleration of the Ba²⁺ current decay by the ß_1a-subunit as compared with ß₂ (Birnbaumer et al., 1998; Colecraft et al., 2002). This ß-subunit dependence of inactivation, defined here as differential ß-subunit modulation of inactivation, may be associated with the different ability of ß-subunits, permanently bound to AID, to freely move with the cytoplasmic linker between repeats I and II of the _1C-subunit, thus serving as a voltage-dependent inactivation particle (Restituito et al., 2000).

The aim of this work was to investigate the correlation between the effects of ß_1a- and ß₂-subunits on inactivation of the _1C,77 and _1C,IS-IV channels and the voltage-dependent rearrangements between the _1C- and ß-subunits using FRET microscopy combined with voltage clamp in live cells (Kobrinsky et al., 2003). This method of measuring the dynamic structural responses to changes in cell membrane potential was used to determine the differences in relative proximity and/or angular orientation between the NH₂-terminal regions of different _1C- and ß-subunits, as well as relative to the plasma membrane when channels are in the resting, inactivated, or conducting states.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Molecular biology
Preparation of plasmids encoding _1C,77 (Soldatov et al., 1995) and _1C,IS-IV (Shi and Soldatov, 2002) in pBluescript vector for expression in Xenopus oocytes and plasmids encoding _1C,77, ß_1a (Mori et al., 1991), and ₂ (Singer et al., 1991) proteins in pcDNA3 vector for eukaryotic expression (Soldatov et al., 2000) has been described previously. All tagged _1C channels retained characteristic electrophysiological properties of the untagged channels. The (ECFP)_N-ß_1a expressing plasmid in a pcDNA3 vector (N stands for NH₂-terminus of a subunit) was prepared by a three-piece ligation. The 6.1-kb Not I/Blp I segment of pß₁cDNA3 plasmid (Soldatov et al., 2000) was ligated with the ECFP-coding region containing the 5'-terminal Not I linker and the 3'-terminal Xho I linker attached by PCR, and 1-kb Xho I/Blp I fragment of the ß_1a-coding plasmid pCaB1 (Mori et al., 1991) with the Xho I linker attached by PCR. The (EYFP)_N-ß_1a-pcDNA3 expressing plasmid was obtained by replacing the Not I/Xho I ECFP-coding segment of (ECFP)_N-ß_1a-pcDNA3 with the EYFP-coding region containing the 5'-terminal Not I linker and the 3'-terminal Xho I linker introduced by PCR. To prepare the (ECFP)_N-ß₂ expressing plasmid, the Avr II/Acc I segment of the ß₂-subunit coding region was ligated with the Xba I/Acc I-digested vector pUC18 (Life Technologies, Carlsbad, CA). The BamH I/Acc I fragment of this construct was subcloned into Bgl II and Acc I sites of the pECFP-C1 vector (Clontech, Palo Alto, CA). Nucleotide sequences of all PCR products as well as ligation sites were confirmed at the DNA sequencing facility of the University of Maryland. Preparations of cRNA were obtained by in vitro transcription of the linearized vectors using T7/SP6 mMessage mMachine (Ambion, Austin, TX).

Cell cultures
COS1 cells used in experiments did not show appreciable endogenous expression of Ca_v1.2 channels (Meir et al., 2000). HEK293 cells were selected for immunoprecipitation and single-channel analyses (Kepplinger et al., 2000), because of the fast time course and high level of transient expression of recombinant Ca_v1.2 channels. COS1 cells are particularly useful for single-cell electrophysiology, because they divide slower than HEK293 cells and allow for tighter control over the efficiency of expression of proteins of different size. COS1 cells were grown on poly-D-lysine coated coverslips (MatTek, Ashland, MA) in DMEM supplemented with 10% fetal calf serum. HEK293 cells used for immunoprecipitation analysis and COS1 were transfected with cDNAs coding for the indicated _1C-, ß-, and ₂-subunits of the Ca²⁺ channel (1:1:1, w/w) using the Lipofectamine (Invitrogen, Carlsbad, CA) or Effectene kit (Qiagen, Venlo, The Netherlands), respectively. In some experiments with COS1 cells, (ECFP)_N- and (EYFP)_N-labeled PH domains (van der Wal et al., 2001) were coexpressed. HEK293 cells used for single-channel experiments were transfected 24–72 h before recordings with cDNAs coding for _1C,77 or (EYFP)_N–_1C,IS-IV–(ECFP)_C, ß_1a-, ß₂-, and ₂-subunits using SuperFect transfection reagent (Qiagen). Unlike the _1C,77 channel, inactivation of which was somewhat (15%) accelerated by the N-terminal EYFP labeling, the decay of the Ba²⁺ current through the _1C,IS-IV was not significantly affected by the double ECFP/EYFP labeling (Kobrinsky et al., 2003). The rationale for using the cDNA coding for (EYFP)_N–_1C,IS-IV–(ECFP)_C was to facilitate identification of the expressing cells. No functional expression of the _1C,77 or _1C,IS-IV channels was observed in COS1 cells in the absence of ß-subunits.

Immunoprecipitation analysis
HEK293 cells were transfected in 75-cm² flasks with the mixture of _1C, ₂, and (EYFP)_N-labeled ß_1a or (ECFP)_N-labeled ß₂-subunits as described above. After 48–72 h, the cells were collected in 0.5 ml of Tris-buffered saline (150 mM NaCl, 50 mM Tris-HCl, pH 8.0) containing a mixture of protease inhibitors composed of 1 mM PMSF, 1 mM DTT, and 5 µl of protease inhibitor cocktail (P8340, Sigma, St. Louis, MO). To remove "orphan" ß-subunits, not associated with the membrane-bound _1C channels, cells were quickly freeze-thawed three times and then permeabilized with digitonin (20 µg/ml) for 5 min at 4°C. Lyzed cells were pelleted for 30 min at 12,000 x g, 4°C and solubilized with 0.5% Nonidet P-40. The cell extract was centrifuged for 30 min at 12,000 x g, 4°C. Equal amounts of protein extracts, determined by BCA protein assay (Pierce, Rockford, IL), were diluted twofold with 0.3 M KCl containing 50 mM Tris-HCl, pH 8.0, the mixture of protease inhibitors, and 10% glycerol. Extracts were incubated overnight at 4°C on a rotary shaker with 10 µl of Living Color Full-Length A.V. polyclonal antibody (BD Biosciences Clontech) prebound to 60 µl of rProtein A Sepharose Fast Flow (Amersham Pharmacia Biotech AB, Uppsala, Sweden). Protein A-sepharose beads were washed at 4°C for 15 min with the wash buffer (0.2% Nonidet P40, 50 mM Tris-HCl, pH 8.0, and protease inhibitors) containing 0.5 M KCl, then quickly washed twice with 0.3 M KCl in wash buffer, and then 30 min with 0.15 M KCl in wash buffer. The resin was finally rinsed twice briefly with 0.5 M KCl in wash buffer, and the retained proteins were eluted by incubation for 5 min at 95°C with 40 µl of SDS-sample buffer containing 2% SDS, 10 mM Tris-HCl (pH 6.8), 7.5% sucrose, and 0.5 M 2-mercaptoethanol. Proteins were separated by SDS-gel electrophoresis on a precasted 4–12% gradient polyacrylamide gel (Invitrogen) and transferred to a 0.45-µm PVDF membrane (BioRad, Hercules, CA). The (ECFP/EYFP)_N-labeled ß-subunits were identified on Western blots with Living Color monoclonal antibody JL-8 (1:4000 dilution, BD Biosciences Clontech), whereas the coprecipitated _1C-subunits were detected with the affinity purified rabbit anti-_1C calcium channel polyclonal antibody (1:10,000; Chemicon International, Temecula, CA) using an ECL Plus Western blotting detection system (Amersham Pharmacia Biotech). Expression of labeled ß-subunits was routinely monitored by the conventional epifluorescent microscopy.

FRET imaging
Images were recorded in live transfected COS1 cells with a Hamamatsu digital camera C4742-95 (Hamamatsu City, Japan) mounted on the Nikon (Melville, NY) epifluorescent microscope TE200 (60 x 1.2 n.a. objective) equipped with multiple filter sets (Chroma Technology, Rockingham, VT). Excitation light was delivered by a 75-W Xenon lamp (Woburn, MA). C-Imaging (Compix, Tualatin, OR) and MetaMorph (Universal Imaging, Downingtown, PA) software were used to obtain and analyze FRET images. FRET was quantified with three filter sets: for EYFP cube, excitation filter 500/20 nm, dichroic beam splitter 515 nm, emission filter 535/30 nm; for FRET (ECFP/EYFP) cube, excitation 436/20 nm, dichroic beam splitter 455 nm, emission filter 540/30 nm; and for ECFP cube, excitation filer 436/20 nm, dichroic beam splitter 455 nm, emission filter 480/40 nm. Regions of interest (ROI) were selected from plasma membrane sites where there was obvious signal of fluorescence and little or no interference from fluorescence from the intracellular compartment. Selection of ROIs was carried out using the C-Imaging software where intensity (I) from three filter sets was determined after background subtraction. Fluorescence values were averaged across the ROI (with one ROI per cell). Inside the ROIs some areas did not show substantial FRET. These areas were also included in the analysis. The full scale of digitized resolution ranged from 0 to 255. Because fluorescence in ROIs, containing typically 1000 pixels, was averaged, the signal/noise ratio was, respectively, increased by two to three orders and thus resolution of FRET detection was more refined. Corrected intensity of FRET (I_FRET^c) was calculated as (I_FRET – I_ECFP x 0.585 – I_EYFP x 0.115), where bleed-through coefficients were experimentally determined by calibration according to Xia and Liu (2001). This calibration is based on the measurements of fluorescence of all ECFP- and EYFP-labeled constructs expressed in COS1 cells in all three filters set to determine bleed-through in CFP, YFP, and FRET filters (for additional details, see Kobrinsky et al., 2003). The typical values of corrected FRET varied from 10 to 40. To quantify FRET, we used a generally accepted correction method (Xia and Liu, 2001). Corrected FRET values were normalized against donor and acceptor levels: N_FRET = (I_FRET^c)/(I_EYFPxI_ECFP)^1/2. Under these conditions, FRET was directly demonstrated with the positive control of the coexpressed mixture of EYFP- and ECFP-labeled pleckstrin homology (PH) domains of phospholipase C1 (Kobrinsky et al., 2003). Acceptor photobleaching (on average >90% for 15 min with the 100-W mercury lamp) gave a FRET efficiency of 0.209 ± 0.054 (n = 17), calculated as (I_ECFP* – I_ECFP) / I_ECFP*, where I_ECFP and I_ECFP* are the intensities of ECFP fluorescence before and after acceptor (EYFP) photobleaching. The relative changes in FRET signal were determined from the pairs of consecutive FRET acquisitions recorded at two different transmembrane voltages (–90 and +20 mV), stabilizing the channel in different functional states. Acquisitions of fluorescence were obtained during the application of a conditioning pulse with simultaneous recordings of the current at indicated voltages and within the time window, ranging from 50 to 300 ms. In a recent comprehensive study comparing different FRET measurement methods (Berney and Danuser, 2003), it has been shown that use of corrected FRET values to present relative changes in FRET signal is adequate if the ratio of donor (ECFP) to acceptor (EYFP) is 1. This requirement was fulfilled throughout our study. In some cases, the entire consecutive image was shifted by one or two pixels and required pixel-by-pixel adjustment using the first (reference) image with designated ROI. Such a shift could easily be detected by comparison with the borders of ROI designated for the reference image.

Acceptor (EYFP) photobleaching was performed with custom YFP photobleaching cube (Chroma Technology), consisting of a D535/50x excitation filter and 100% mirror (instead of a dichroic one); this bleaching cube spared the CFP chromophore in control experiments. The 175-W Xenon lamp was used for illumination.

For illustration purposes, results of the ratio of FRET measurements, not limited just with ROI, are presented as fluorescence images. The first step in the processing of such an illustrating image is subtraction of the background, which is typically 10–12 units on the full fluorescence scale of 255 units. Background was measured in the part of the image that was not affected by cell fluorescence or in a separate dish containing nonfluorescent cells. EYFP and EYCFP images with subtracted background were fractionally subtracted from raw FRET images based on measurements of bleed-through coefficients. The fractional subtraction generated corrected FRET images, showing sensitized FRET in COS1 cells. Because C-Imaging software converted possible negative values of fluorescence to zero, after this procedure, an average "corrected background" was typically 0 or 1. Although subtraction of images recorded at –90 and +20 mV gives absolute changes of FRET, the ratio of images allows for comparison between different experiments and thus was used in our research. To obtain illustrating images of the FRET ratio at +20 mV/–90 mV, an empirical constant value, typically equal to 5 on the fluorescence scale, was added to each pixel of both images corrected by the background subtraction. This was required to avoid division by zero background fluorescence of a corrected image. With noise ranging from 0 to 12 and the full scale of the measurement at 255, this operation did not substantially affect the relationship between the processed image and the actual calculation of corrected FRET in ROI, which did not include such processing.

Electrophysiology
Three different types of electrophysiological measurements were used. Two-electrode voltage clamp recording of Ba²⁺ currents through the wild-type and mutated Ca²⁺ channels expressed in Xenopus oocytes was carried out as described previously (Shi and Soldatov, 2002). To inhibit Ba²⁺-activated Cl^– current, the oocytes were injected with 50 nl of 100 mM 1,2-bis(o-aminophenoxy)ethane-N,N,N',N'-tetraacetate (BAPTA)-Cs (pH 7.4) 30 min before recordings.

All FRET measurements were combined with whole-cell patch clamp. Patch-clamp recording of the Ba²⁺ current in COS1 cells was carried out using the Axopatch 200B amplifier (Axon Instruments, Union City, CA) 48–72 h after transfection. The extracellular bath solution contained (in mM): NaCl 100, BaCl₂ 20, MgCl₂ 1, glucose 10, HEPES 10, pH 7.4. The pipettes had resistance 3–6 M and were filled with pipette solution containing (in mM): CsCl 110, MgATP 5, BAPTA 10, tetraethylammonium 20, cAMP 0.2, HEPES 20, pH 7.4 (Soldatov et al., 2000). Currents were sampled at 2.5–5 kHz and filtered at 1 kHz. Voltage protocols were generated and data were digitized, recorded, and analyzed using pClamp 8.1 software (Axon). Calculated parameters are means ± SEM. The unpaired t-test was used to compare the _1C,77/ß and _1C,IS-IV/ß channels.

In single-channel recordings, the bath solution contained (in mM): L-aspartic acid 110, KCl 20, MgCl₂ 2, EGTA 2, HEPES 20, pH 7.4 (KOH). The pipette solution contained (in mM): BaCl₂ 96, HEPES 5, pH = 7.4 (NaOH). The pipettes were made of borosilicate glass, coated with SigmaCote (Sigma) to reduce pipette capacitance, heat polished, and showed resistances of 3–4.5 M. Voltage protocols were applied and single-channel current traces were recorded using pClamp 6 software (Axon). Interpulse potential was held at –80 mV for 5 s, after depolarization pulse to 0 mV (maximum of activity) for 1 s. Currents were sampled at 10 kHz and filtered at 1 kHz. Analysis of single-channel data (average open probability, N.p) was performed with Matlab software (MathWorks, Natick, MA) using algorithms according to Schmid et al. (1995) and Baumgartner et al. (1997). In experiments where more than one single channel was visible, the number of channels was estimated by the maximum number of overlapping openings seen within the 80–140 traces as described by Baumgartner et al. (1997). All measurements were carried out at 20–22°C.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Lack of differential ß-subunit modulation of the _1C,IS-IV channel inactivation
For comparison with our previous results (Shi and Soldatov, 2002), the electrophysiological study of inactivation of the _1C,77 and _1C,IS-IV channels were initially carried out in the Xenopus oocyte expression system. Expression in oocytes provided control over subunit composition of the Ca²⁺ channels and yielded results fully compatible with those acquired by the whole-cell patch clamp recording in COS1 cells used for the FRET microscopy (see below). Differential ß-subunit modulation of the Ba²⁺ current through the wild-type (_1C,77) channel is illustrated in Fig. 1 A as a difference in kinetics of the current inactivation conveyed to the channel by coexpression of ß_1a- and ß₂-subunits. Replacement of the ß₂-subunit with ß_1a significantly accelerated decay of the Ba²⁺ current, which supports results of other studies reviewed by Birnbaumer et al. (1998).

View larger version (14K): [in this window] [in a new window]   FIGURE 1  Differential ß-subunit modulation of the voltage-dependent inactivation in the wild-type 1C,77 channel (A) and its lack in the 1C,IS-IV channel (B). Shown are effects of coexpression of the ß1a- or ß2-subunits on inactivation of the Ba2+ currents recorded in Xenopus oocytes at the peak of current-voltage relationships (+10 mV) with holding potential of –90 mV and 40 mM Ba2+ in the perfusion solution. A two-exponential fitting of the Ba2+ current inactivation kinetics obtained from the analysis of the 1-s pulse traces showed that the fast time constant of the 1C,77 channel was not significantly changed (p > 0.09; unpaired t-test) when the ß1a-subunit (f = 94.0 ± 9.8 ms, n = 30) was replaced by ß2 (f = 132.0 ± 18.6 ms, n = 17). The slow time constant of the Ba2+ current inactivation kinetics was significantly (p < 0.005) increased when the ß1a-subunit (s 558 ± 66 ms) was replaced by the ß2-subunit (s 925 ± 112 ms) with the 1C,77 channel (A), but remained essentially unchanged in the case of the 1C,IS-IV channel (B) (see text for details).

Data summarized in Fig. 2 show that none of the tested kinetics or voltage-dependent characteristics of the Ba²⁺ current through the _1C,IS-IV channel were notably affected by the coexpression of the ß₂-subunit instead of ß_1a. The sustained current was observed over a wide range of membrane potentials (Fig. 2 A) and, on average, accounted for 50–55% of the peak current. No significant differences in the current-voltage relationships for the peak and sustained Ba²⁺ current or the time course of fast inactivation (Fig. 2 B) were found between the _1C,IS-IV channels coexpressed with ß_1a- and ß₂-subunits. The values for half-maximal activation V_0.5 = 2.1 ± 1.3 mV (slope factor k_I-V = –9.1 ± 1.3 mV; n = 10) with ß_1a and 3.9 ± 2.0 (k_I-V = –8.9 ± 0.7; n = 3) with ß₂ were not significantly different (p < 0.002; unpaired t-test). The sustained Ba²⁺ current did not show decay even with prolonged depolarization (Fig. 2 C). Similar to the _1C,IS-IV/ß_1a channel, the _1C,IS-IV/ß₂ channel was characterized by high steady-state availability at positive voltages (Fig. 2 D) and rapid recovery from inactivation within 100 ms (data not shown). Thus, it seems that inhibition of slow voltage-dependent inactivation interferes with differential ß-subunit modulation of the kinetics of inactivation. Based on this observation, we hypothesize that ß-subunits modulate the kinetics of the Ba²⁺ current decay by targeting the slow inactivation mechanism of the channel.

View larger version (34K): [in this window] [in a new window]   FIGURE 2  Electrophysiological properties of the mutated Cav1.2 channel composed of the 1C,IS-IV-, ß2-, and 2-subunits coexpressed in Xenopus oocytes. (A) Traces of the Ba2+ current evoked by 1-s depolarizations in the range from –20 to +40 mV (10-mV increments) from a holding potential of –90 mV. (B) Comparison of the current-voltage relationships for the peak (circles) and sustained (diamonds) Ba2+ currents as well as the voltage dependence of the time constant of inactivation (f, squares) for the 1C,IS-IV/ß1a (open symbols, n = 8) and 1C,IS-IV/ß2 channels (solid symbols, n = 3) The curves were fitted by the equation IBa = Gmax (V – Erev)/{1 + exp[(V – V0.5)/kI-V]}, where IBa is the amplitude of the Ba2+ current, Gmax, the maximum conductance; Erev, the reversal potential, V0.5, the voltage at 50% of IBa inactivation, and kI-V, the slope factor. (C) Lack of inactivation of the sustained current illustrated by a typical trace of the Ba2+ current through the 1C,IS-IV/ß1a (top) and 1C,IS-IV/ß2 channels (bottom) induced by 10-s depolarization to +10 mV from a holding potential of –90 mV. (D) Comparison of steady-state inactivation curves for the peak Ba2+ current through the 1C,IS-IV/ß1a (, n = 6) and 1C,IS-IV/ß2 channels (•, n = 3). The curves were fitted by a Boltzmann function IBa = a + b x {1 + exp[(V – V0.5)/k]}, where V is the conditioning pulse voltage, V0.5 is the voltage at half-maximum of inactivation, k is a slope factor, a and b are fractions of noninactivating and inactivating components of the Ba2+ current, respectively. V0.5 = –27.7 ± 1.5 mV, k = 14.7 ± 3.0, and a = 52.1 ± 4.6 (n = 7) for the 1C,IS-IV/ß1a channel. Respectively, for the 1C,IS-IV/ß2 channel, V0.5 = –24.0 ± 4.1 mV, k = 12.6 ± 0.5, and a = 52.7 ± 2.4 (n = 3).

View larger version (34K): [in this window] [in a new window]   FIGURE 3  Effect of the ß1a- and ß2-subunits on the properties of the 1C,77 and 1C,IS-IV channels. (A) Coimmunoprecipitation of the 1C- and ß-subunits in cells expressing the 1C,77 and 1C,IS-IV channels. The (1,4) 1C,77- or (2,5) 1C,IS-IV- subunits, coexpressed with 2 and (1,2) EYFP-labeled ß1a- or (4,5) ECFP-labeled ß2-subunits in HEK293 cells, were solubilized with Nonidet P-40, immunoprecipitated with polyclonal antibody to GFP in the presence of protein A-sepharose, and analyzed by Western blot for the ß-subunits using monoclonal antibody to GFP, and for the 1C-subunits by polyclonal anti-1C antibody (arrows). Controls (3,6) show nontransfected HEK293 cells. (B) Single-channel recordings of the 1C,77 or (EYFP)N–1C,IS-IV–(ECFP)C channel expressed with 2- and ß1a- or ß2-subunits in HEK293 cells. Currents were elicited by 1-s depolarizations to 0 mV from the holding potential of –80 mV. Boxed are 0–50 ms and 800–950 ms episodes used for the subsequent analysis. Average open probabilities determined as N.p (C) or p (D) at the 800–950 ms time window of a 1-s depolarization (boxed in B). For each independent experiment, between 80 and 140 traces were analyzed (number of experiments is shown in parentheses). Only the average open probability for 1C,77/ß2 channel was significantly (p<0.05) different to the other 1C/ß channels. (E) Ratio of average open probabilities (determined as N.p) between the 0–50-ms and 800–950-ms time windows (boxed in B) during a 1-s depolarization. Only the ratio calculated for the 1C,77/ß1a channel was significantly (p < 0.05) different from the other 1C/ß channels.

Voltage-dependent rearrangement between the _1C,77- and ß_1a-subunits
FRET is a result of the molecular interaction between the labeled partners (Lakowicz, 1999). FRET measurement with simultaneous voltage clamp gives an opportunity to eliminate possible artifacts associated with "orphan" subunits in live cells and focus exclusively on tagged _1C- and ß-subunits of fully functional channels. One of the most commonly used approaches is to measure FRET efficiency by acceptor photobleaching. Specially designed Chroma D535/50x excitation filter (Erickson et al., 2003) provided 90% photobleaching of EYFP in 2 min, a time sufficient to determine FRET efficiency under conditions when stability of a patch is not compromised. Recently Tour et al. (2003) reported on the high photosensitivity of labeled Ca_v1.2 channel to rundown. We found that EYFP photobleaching greatly accelerated rundown of the (EYFP)_N-_1C,77 channel activity (data not shown) and thus can't be applied to the measurements of voltage-gated changes of FRET, because of uncertainty in the functional state of the channel.

To overcome this problem, we used another widely accepted approach based on FRET determination by correction methods, and combined patch clamp with FRET microscopy. The directly measured FRET intensity was corrected for the cross talk from the donor and the acceptor and by the background. FRET ratio was determined independently from donor and acceptor concentrations and additionally normalized to the direct donor and acceptor intensity according to Xia and Liu (2001). Fluorescence images b and c (Fig. 4 A) indicate that coexpressed (ECFP)_N-ß_1a and (EYFP)_N-_1C,77 subunits are similarly distributed in the cell. We focused our analysis on the plasma membrane region, where the voltage-gated functional channel molecules reside. A FRET measurement of the same cell recorded at –90 mV (panel d) shows that N_FRET between these subunits was rather small. Depolarization to +20 mV (Fig. 4 A, panel e) resulted in a substantial increase in N_FRET between the (ECFP)_N-ß_1a and (EYFP)_N-_1C,77 subunits (Table 1), which was confined to the rim of the cell, consistent with plasma membrane localization (see digitally magnified regions at the bottom of the whole-cell images).

View larger version (62K): [in this window] [in a new window]   FIGURE 4  Voltage-gated rearrangements between the (ECFP)N-ß and (EYFP)N-1C subunits of the functional Ca2+ channels determined by FRET combined with patch clamp in COS1 cells. (A and B) Voltage-dependent corrected FRET between the (EYFP)N-1C,77 and (A) (ECFP)N-ß1a or (B) (ECFP)N-ß2 subunits. Schematic diagram (boxed) depicts the arrangement of fluorophores. Panel a, phase-contrast cell image with a shadow of patch pipette. Panels b and c, images of the same cell with YFP and CFP filters, respectively. Arrows point to the plasma membrane distribution of the labeled proteins. Panels d and e, representative corrected images of FRET between the (ECFP)N-ß and (EYFP)N-1C,77 subunits recorded, respectively, at the resting (–90 mV) and inactivated ((ECFP)N-ß1a) or predominantly conducting ((ECFP)N-ß2) states of the channel (+20 mV). The time window of the +20-mV acquisition is marked on the left panel above the 1-s Ba2+ current trace by a red horizontal bar. Panel f, ratio of images e/d showing the voltage-dependent change of corrected FRET. Image areas indicated by arrows were digitally magnified to demonstrate the confinement of FRET to the plasma membrane. Three traces of the Ba2+ current on the left panel in B, recorded in response to the 1-, 5- and 45-s depolarizations, demonstrate that inactivation was not complete, and FRET imaging on panel c represents channels almost equally distributed between the inactivated and open states (ratio of the Ba2+ current after 1 s to the peak current, r1s 0.56). (C) Absence of marked FRET between the (ECFP)N-ß2a and (EYFP)N-1C,77 subunits at the end of a 5-s depolarization to +20 mV that caused an additional decrease in the fraction of open channels (r5s 0.33). Panels a and b, representative corrected images of FRET obtained at –90 mV and +20 mV at the end of a 5-s pulse at the time window marked by the red bar above the current trace (panel e). Panel c, the ratio of corrected FRET images b/a sequentially recorded at +20 and –90 mV; right panels, digitally magnified areas of the image pointed by arrows. Panel d, phase-contrast image of the expressing cell with a shadow of patch pipette. (D and E) Voltage-dependent corrected FRET between the (EYFP)N-1C,IS-IV and (ECFP)N-ß1a (D) or (ECFP)N-ß2 (E) subunits shown as a ratio of the sequentially recorded +20 mV/–90 mV images (panel a). Panel b, phase-contrast images of the expressing cell. Note that in every case the Ba2+ current recordings (shown below the respective images) provide evidence that the sustained conducting state was achieved before the +20-mV FRET image acquisitions (at time windows marked by red bars). Scaling bars, 5 µm.

View this table: [in this window] [in a new window]   TABLE 1  Voltage-dependent changes of normalized FRET between the (ECFP)N-ß and (EYFP)N-1C subunits determined at –90 mV and +20 mV under whole-cell voltage clamp in COS1 cells

When coexpressed with the unlabeled _1C,77- and ₂-subunits, the (ECFP)_N-labeled ß_1a- and ß₂-subunits did not show voltage-dependent changes in ECFP fluorescence. The ratio of ECFP fluorescence intensity measured at +20 mV to those at –90 mV did not significantly differ from 1.0 and was found to be 0.991 ± 0.006 (n = 44) with (ECFP)_N-ß_1a/_1C,77; 1.018 ± 0.016 (n = 40) with (ECFP)_N-ß₂/_1C,77; 1.006 ± 0.004 (n = 22) with (ECFP)_N-ß_1a/_1C,IS-IV, and 0.997 ± 0.008 (n = 22) with (ECFP)_N-ß₂/_1C,IS-IV. Conversely, acceptor (EYFP) fluorescence measured at +20 mV was not significantly different from those at –90 mV when the (EYFP)_N-labeled _1C,77- and _1C,IS-IV-subunits were coexpressed with the unlabeled ß_1a- and ß₂-subunits. Because a voltage-induced change of the acceptor fluorescence in EYFP cube will be subtracted from the FRET cube measurement, we did not expect a decline of the EYFP fluorescence ratio at +20 and –90 mV from 1 with our three-cube filter system even if acceptor hypothetically shows voltage-sensitive change of fluorescence. Indeed, the ratio of EYFP fluorescence intensity measured at +20 mV to those at –90 mV was on average not different from 1.0: 1.003 ± 0.007 (n = 38) with (EYFP)_N-_1C,77/ß_1a; 1.002 ± 0.007 (n = 10) with (EYFP)_N-_1C,IS-IV/ß_1a; 0.995 ± 0.010 (n = 26) with (EYFP)_N-_1C,77/ß₂, and 1.011 ± 0.011 (n = 16) with (EYFP)_N-_1C,IS-IV/ß₂. This control indicates that voltage-dependent changes of fluorescence of the individual ECFP or EYFP fluorophores fused to the NH₂-termini of the _1C- and ß-subunits do not contribute to the voltage-dependent changes in FRET between them recorded in this study.

Efficiency of FRET depends on the distance between the donor and acceptor fluorophores and their mutual angular orientation. Although (ECFP)_N-ß and (EYFP)_N-_1C subunits were expressed at 1:1 molar ratio, the uncertainties in their relative concentration can be minimized by the ratiometric analysis of FRET (Xia and Liu, 2001). Relative changes in normalized FRET values (N_FRET) were detected for all combinations of the _1C- and ß-subunits, with the exception of the _1C,77/ß₂ channel (Table 1; Figs. 4, B and C). In the resting state (–90 mV), all tested channels were characterized by low FRET between the tagged NH₂-termini of the _1C- and ß-subunits. Depolarization to +20 mV caused almost complete inactivation of the (EYFP)_N-_1C,77/(ECFP)_N-ß_1a channels within 1 s (see current trace in Fig. 4 A) and produced a fully reversible significant increase in N_FRET (Table 1; Fig. 4 A, panel e) measured in the predominantly inactivated state (marked by red bar above the current trace). This result suggests substantial rearrangement between the _1C- and ß-subunits in the inactivated state accompanied by a reduction of the apparent distance between the donor and acceptor fluorophores and/or more parallel orientation of the respective dipoles.

Voltage-dependent rearrangements between the _1C,77- and ß₂-subunits
Replacement of the ß_1a-subunit for ß₂ produced an increase in the Ba²⁺ current amplitude corresponding to higher efficiency of expression of the _1C,77 channel. Complete inactivation of the Ba²⁺ current through the _1C,77/ß₂ channel was delayed, even with a 45-s depolarization at +20 mV (Fig. 4B, bottom trace in the left panel), but not inhibited like in the case of the _1C,IS-IV channel. No measurable current was observed when _1C,77 was expressed in COS1 cells in the absence of ß-subunits (data not shown), thus precluding comparison of the current decay in the absence of the ß-subunit modulation. Fluorescence measurements corroborate the results of the immunoprecipitation/Western blot analysis and additionally confirm the expression of the ß₂-subunit and its association with _1C,77. In the case of the (EYFP)_N-_1C,77/(ECFP)_N-ß₂ channel, equally strong ECFP and EYFP fluorescence was observed (Fig. 4 B, panels b and c), similar to that found for the (EYFP)_N-_1C,77/(ECFP)_N-ß_1a channel (compare with Fig. 4 A, panels b and c). Moreover, FRET images (Fig. 4 B, panels d and e, and the respective digitally magnified membrane regions) clearly point to colocalization of the labeled _1C,77- and ß₂-subunits (notice different scale in FRET images when comparing Fig. 4 B with Fig. 4 A, panels d–f). Taken together, the large amplitude of the Ba²⁺ current, the immunoprecipitation analysis, the equally strong expression of the fluorescent _1C,77- and ß₂-subunits, and their colocalization by FRET suggest that the low state-dependent change in FRET determined for the ratio of images recorded at +20/–90 mV (Fig. 4 B, panel f) was not due to the lack of functional expression of the (EYFP)_N-_1C,77/(ECFP)_N-ß₂ channel.

Insignificant change in relative N_FRET (Table 1) recorded at the end of the 1-s depolarization pulse to +20 mV from the holding potential of –90 mV (marked by the red bar above the top current trace in Fig. 4 B) corresponded to a transition of the (EYFP)_N-_1C,77/(ECFP)_N-ß₂ channel from the resting to a predominantly noninactivated state, as seen from the slow and incomplete decay of the whole-cell Ba²⁺ current and single-channel-gating behavior in Fig. 3, B and C. An average fraction of inactivated channels in the 1-s recordings accounted for 44%. Prolongation of +20 mV depolarization pulse to 5 s (Fig. 4 C) increased the fraction of the inactivated channels to 67%. In the case of the (EYFP)_N-_1C,77/(ECFP)_N-ß_1a channel, a similar fractional inactivation would correspond to almost fivefold increase in relative N_FRET. However, differential FRET between the (EYFP)_N-_1C,77 and (ECFP)_N-ß₂ subunits recorded at –90 mV before and at the end of the 5-s pulse to +20 mV (Fig. 4 C, panel c) was not substantially increased. Thus, in sharp contrast with the ß_1a-subunit, no significant voltage-dependent changes of FRET between the NH₂-terminal fluorescent tags of the _1C,77- and ß₂-subunits were found. This result suggests that the voltage-gated rearrangements of the labeled parts in the (EYFP)_N-_1C,77/(ECFP)_N-ß₂ channel complex are either insignificantly small or occur outside the detection limits of the method (i.e., when linear distance between the fluorophores becomes >10 nm, or the dipoles approach perpendicular orientation, or both). Overall, the difference in voltage-dependent FRET between (EYFP)_N-_1C,77 and the (ECFP)_N-labeled ß_1a- or ß₂-subunits (Table 1) reflects the differential ß-subunit modulation of inactivation of the _1C,77 ("wild-type") calcium channel.

Voltage-dependent rearrangement between _1C,IS-IV and different ß-subunits
In full agreement with the data obtained in oocyte expression system, the replacement of the ß_1a-subunit for ß₂ did not affect the kinetics of inactivation of the Ba²⁺ current through the _1C,IS-IV channel expressed in COS1 cells (compare current traces in panels c in Fig. 4, D and E). On average, the fast time constant of inactivation was 13.0 ± 1.2 ms (n = 13) with the ß_1a-subunit and 12.9 ± 0.8 ms (n = 12) with the ß₂-subunit. The sustained fraction of the current (52 ± 8% with ß_1a and 50 ± 9% with ß₂) lasted for the duration of depolarization (up to 30 s) without decay. The relatively small current amplitudes obtained with the _1C,IS-IV channel is characteristic for other Ca_v1.2 channel mutants with impaired Ca²⁺-induced inactivation property (Soldatov et al., 1998). This was found to be due to a lower open probability and 10–15% reduction in the single-channel conductance (Kepplinger et al., 2000), and had no significant impact on the measurement of intramolecular FRET between the termini of the _1C,IS-IV channel (Kobrinsky et al., 2003).

The lack of differential ß-subunit modulation of the _1C,IS-IV channel inactivation correlated with the absence of the ß-subunit specificity of voltage-dependent FRET between the (ECFP)_N-ß and (EYFP)_N-_1C,IS-IV subunits. N_FRET between the NH₂-terminal tags of the _1C,IS-IV- and either ß-subunits was found to increase in the sustained conducting state (Fig. 4, D and E; Table 1). However, this increase in FRET essentially did not depend on the type of the coexpressed ß-subunit. The transition of the (EYFP)_N-_1C,IS-IV channel into the conducting noninactivated state, stabilized by depolarization to +20 mV (Fig. 4, D and E, panels c), showed an equally large (by unpaired t-test) increase in N_FRET with either (EYFP)_N-ß_1a (to 0.17 ± 0.04) or (EYFP)_N-ß₂ subunit (to 0.19 ± 0.03). This result may reflect the loss of the ß-subunit specificity of voltage-dependent rearrangements between the NH₂-terminal regions of the _1C,IS-IV- and ß-subunits and correlates with the lack of differential ß-subunit modulation of the Ba²⁺ current through the _1C,IS-IV channel.

Voltage-dependent rearrangements of the _1C- and ß-subunits vis-à-vis the plasma membrane
FRET microscopy combined with whole-cell patch clamp was used to estimate whether the relative proximity of the _1C- and ß-subunits to the plasma membrane is changed in different functional states of the channel (Fig. 5). As membrane probes we used the ECFP/EYFP-labeled pleckstrin homology (PH) domains of phospholipase C1 (van der Wal et al., 2001) that localize via PIP₂ to the inner leaflet of the plasma membrane. Unlike the _1C,77-(ECFP)_C channel labeled at the C-terminal tail (Kobrinsky et al., 2003), which did not show FRET with the labeled PH domains (data not shown), all NH₂-terminally labeled _1C- and ß-subunits showed substantial FRET (Fig. 5 A). This finding suggests that the NH₂-termini of the _1C- and ß-subunits are situated near the plasma membrane within the linear detection limits of the method (100 Å), i.e., close enough to determine their state-dependent rearrangements with regards to the lipid bilayer.

View larger version (73K): [in this window] [in a new window]   FIGURE 5  Voltage-dependent mobility of the NH2-labeled 1C- and ß-subunits vis-à-vis the plasma membrane. (A) Comparison of normalized FRET between (EYFP)N-PH and the indicated (ECFP)N-ß subunits in the 1C,77 or 1C,IS-IV channels at –90 mV (open bars) and +20 mV (shaded bars) (conditions of FRET acquisition were similar to those in Figure 4). n, number of cells; p, paired t-test; ns, nonsignificant differences. Schematic diagram (inset) depicts the arrangement of fluorophores. (B) Representative FRET microscopy data corresponding to the paired bars in panel A. Panels a, phase-contrast images of the expressing cells, panels b, images of corrected FRET recorded at +20 mV, and panels c, ratio of FRET images sequentially recorded at +20 and –90 mV. (C) Comparison of normalized FRET between the indicated (EYFP)N-1C subunits and (ECFP)N-PH recorded at steady state at –90 mV (open bars) and +20 mV (shaded bars). Inset diagram shows the arrangement of fluorophores. (D) Panels a, phase-contrast images of the expressing cells. Panels b and c show the corresponding representative corrected FRET at +20 mV and the ratio of FRET images sequentially recorded at +20 and –90 mV, respectively. Scaling bars, 5 µm.

Fig. 5, C and D, show that FRET between the (ECFP)_N-PH domain and the (EYFP)_N-tagged _1C,77 or _1C,IS-IV channels containing the ß_1a-subunit is essentially voltage independent. No significant changes in FRET between the NH₂-terminal fluorophores of the _1C-subunits and the plasma membrane PH domain probes were observed for the channels in the resting and the inactivated (_1C,77) or conducting (_1C,IS-IV) states.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Results of our work show that voltage-gated rearrangements between the NH₂-termini of the _1C- and ß-subunits, associated with inactivation of the Ca_v1.2 channel, depend on ADSI, the type of the ß-subunit, and differential ß-subunit modulation of inactivation.

Differential ß-subunit modulation and slow inactivation
The effect of ß-subunits on the kinetics of the Ba²⁺ current decay is mediated by their interactions with the ion-conducting _1C-subunits. However, is not clear whether inactivation of the Ca_v1.2 channel is inherent to association between the _1C- and ß-subunits. Earlier, Ferreira et al. (1997) observed in tsA201 cells an expression of the functional cardiac rabbit _1C channel in the absence of the rat brain ß₂-subunit. In their report, a coexpression of the ß₂-subunit predominantly affected the amplitude of the Ba²⁺ current but did not accelerate the kinetics of inactivation. In contrast, our studies in COS1 cells showed no functional expression of the _1C,77 and _1C,IS-IV channels in the absence of ß-subunits. Thus, we investigated the ß-subunit modulation of inactivation as a differential effect of two different ß-subunits, ß_1a and ß₂, on two variants of the Ca_v1.2 channels showing completely different inactivation properties.

Inactivation of the Ba²⁺ current through the "wild-type" _1C,77 channel is characterized by fast and slow components and is subject to differential ß-subunit modulation (Fig. 1 A). Inhibition of the voltage-dependent slow inactivation (_1C,IS-IV) abolished differential ß-subunit modulation of the Ba²⁺ current (Fig. 1 B). The replacement of the ß_1a-subunit with ß₂ did not cause appreciable changes in the inactivation kinetics of the _1C,IS-IV channel over a wide range of membrane potentials (Figs.