In skeletal muscle, excitation–contraction (EC) coupling requires depolarization-induced conformational rearrangements in L-type Ca2+ channel (CaV1.1) to be communicated to the type 1 ryanodine-sensitive Ca2+ release channel (RYR1) of the sarcoplasmic reticulum (SR) via transient protein–protein interactions. Although the molecular mechanism that underlies conformational coupling between CaV1.1 and RYR1 has been investigated intensely for more than 25 years, the question of whether such signaling occurs via a direct interaction between the principal, voltage-sensing α1S subunit of CaV1.1 and RYR1 or through an intermediary protein persists. A substantial body of evidence supports the idea that the auxiliary β1a subunit of CaV1.1 is a conduit for this intermolecular communication. However, a direct role for β1a has been difficult to test because β1a serves two other functions that are prerequisite for conformational coupling between CaV1.1 and RYR1. Specifically, β1a promotes efficient membrane expression of CaV1.1 and facilitates the tetradic ultrastructural arrangement of CaV1.1 channels within plasma membrane–SR junctions. In this paper, we demonstrate that overexpression of the RGK protein Rem, an established β subunit–interacting protein, in adult mouse flexor digitorum brevis fibers markedly reduces voltage-induced myoplasmic Ca2+ transients without greatly affecting CaV1.1 targeting, intramembrane gating charge movement, or releasable SR Ca2+ store content. In contrast, a β1a-binding–deficient Rem triple mutant (R200A/L227A/H229A) has little effect on myoplasmic Ca2+ release in response to membrane depolarization. Thus, Rem effectively uncouples the voltage sensors of CaV1.1 from RYR1-mediated SR Ca2+ release via its ability to interact with β1a. Our findings reveal Rem-expressing adult muscle as an experimental system that may prove useful in the definition of the precise role of the β1a subunit in skeletal-type EC coupling.
Excitation–contraction (EC) coupling is the physiological event in which muscle converts an electrical signal (plasma membrane depolarization) into mechanical work (muscle contraction). In the case of skeletal muscle, depolarization-induced conformational rearrangements within the L-type Ca2+ channel complex (CaV1.1) are coupled to gating of the type 1 ryanodine-sensitive Ca2+ release channel (RYR1) of the SR (Schneider and Chandler, 1973; Ríos and Brum, 1987; Tanabe et al., 1988). The resultant Ca2+ efflux from the SR into the myoplasm via RYR1 activates the contractile filaments. Because SR Ca2+ release occurs rapidly in response to depolarization and independently of transient Ca2+ fluctuations, a conformational coupling mechanism appears to support communication between the two channels (see Bannister and Beam, 2013).
Although the roles of CaV1.1 and RYR1 as voltage sensor and SR Ca2+ release channel, respectively, have been established for quite some time (Tanabe et al., 1988; Nakai et al., 1996), the molecular mechanics that support conformational coupling between these two channels remain undefined. One candidate structure to mediate such coupling is the intracellular segment that links repeats II and III of the principal α1S subunit of CaV1.1 (Tanabe et al., 1990; Lu et al., 1994; Nakai et al., 1998; Wilkens et al., 2001). Another viable candidate is the auxiliary β1a subunit of the CaV1.1 heteromultimer. In this regard, β1a is firmly established as being essential for EC coupling, as genetic deletion of β1 abolishes voltage-dependent SR Ca2+ release in both mammals and bony fish (Gregg et al., 1996; Ono et al., 2004; Schredelseker et al., 2005, 2009). Moreover, purified β1a subunits and β1a peptide fragments bind RYR1 in vitro and/or activate RYR1 in planar lipid bilayers (Cheng et al., 2005; Rebbeck et al., 2011; Karunasekara et al., 2012; Hernández-Ochoa et al., 2014). Still, the key roles of β1a in trafficking CaV1.1 to the plasma membrane (Gregg et al., 1996; Strube et al., 1996) and in the ultrastructural organization of CaV1.1 into the tetradic arrays prerequisite for EC coupling (Schredelseker et al., 2005, 2009; Dayal et al., 2010, 2013; Eltit et al., 2014) have made testing a direct role for β1a in communication between the voltage-sensing components of CaV1.1 and RYR1-mediated SR Ca2+ release highly problematic.
Members of the RGK (Rad, Rem, Rem2, Gem/Kir) family of monomeric G proteins inhibit L-type Ca2+ channels in a variety of physiological systems via interactions that occur primarily with the β subunit (Béguin et al., 2001, 2007; Finlin et al., 2003, 2006; Murata et al., 2004; Bannister et al., 2008; Yang et al., 2012; Romberg et al., 2014; Xu et al., 2015; reviewed recently by Yang and Colecraft, 2013). In the present study, we have examined the impact of Rem on EC coupling in adult mouse flexor digitorum brevis (FDB) fibers overexpressing Rem via in vivo electroporation (DiFranco et al., 2007). Using this approach, we have found that Rem effectively uncouples the CaV1.1 voltage sensor from RYR1-mediated SR Ca2+ release through its interaction with β1a. Specifically, Rem markedly reduces voltage-induced myoplasmic Ca2+ transients without appreciable effects on CaV1.1 targeting, intramembrane charge movement, or SR Ca2+ store content.
MATERIALS AND METHODS
A cDNA encoding a CFP-rabbit CaV1.1 α1S-subunit (GenBank accession no. X05921) fusion construct was created by swapping out YFP for CFP in an existing YFP–α1S fusion construct (Papadopoulos et al., 2004). The cDNA segment encoding CFP was excised from the parent pECFP-C1 vector (Takara Bio Inc.) using NheI and HindIII (761 bp). Likewise, YFP was removed from the YFP–α1S fusion construct using the same restriction enzymes, linearizing the pEYFP-C1 backbone and the α1S-coding sequence (9,555 bp). The CFP-encoding segment was then re-ligated into the linearized vector carrying the α1S-coding sequence (final, 10,316 bp).
CFP-β1a and YFP-β1a.
In vivo electroporation and dissociation of FDB fibers
All procedures involving mice were approved by the University of Colorado Denver-Anschutz Medical Campus Institutional Animal Care and Use Committee. cDNA plasmids encoding YFP, CFP-α1S, CFP-β1a, V-Rem, and/or V-Rem AAA were delivered to FDB fibers of anesthetized 2–3-mo-old male C57BL/6J mice (The Jackson Laboratory) via an in vivo electroporation protocol similar to that originally described by DiFranco et al. (2007). In brief, 10 µl of 2 mg/ml hyaluronidase solution was injected into the FDB muscle with a 30-gauge hypodermic needle. After 1 h, mice were re-anesthetized and 20 µl cDNA (3–5 µg/µl) was injected into the muscle. 5 min later, two gold-plated acupuncture needle electrodes (Lhasa OMS) coupled to an isolated pulse stimulator (A-M Systems) were placed subcutaneously near the proximal and distal tendons of the muscle (∼1 cm apart). cDNAs were then electroporated into the FDB muscle with 20 100-V, 20-ms pulses delivered at 1 Hz. For assessment of SR Ca2+ stores, the transfection mixture also contained 5 µg pmCherry-C1 (Takara Bio Inc.) as a means to identify successfully transfected fibers after loading with Fluo 3-AM dye (Invitrogen; see below).
Electroporated (9–10 d after transfection) FDB muscles were dissected in cold rodent Ringer’s solution (mM: 146 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, and 10 HEPES, pH 7.4 with NaOH). Muscles were then digested in a collagenase solution (mM: 155 Cs-aspartate, 10 HEPES, and 5 MgCl2, pH 7.4 with CsOH, supplemented with 1 mg/ml BSA [Sigma-Aldrich] and 1 mg/ml collagenase type IA [Sigma-Aldrich]) with agitation at 37°C for ∼1 h. Immediately after digestion, the collagenase solution was replaced with a dissociation solution (mM: 140 Cs-aspartate, 10 Cs2EGTA, 10 HEPES, and 5 MgCl2, pH 7.4 with CsOH, supplemented with 1 mg/ml BSA), and muscles were triturated gently with a series of fire-polished glass pipettes of descending bore. Dissociated FDB fibers destined for whole-cell patch-clamp experiments were then plated onto ECL (EMD Millipore)-coated 35-mm plastic culture dishes (Falcon). For imaging, fibers were allowed to settle onto laminin (Invitrogen)-coated 35-mm culture dishes with glass coverslip bottoms (MatTek). Experiments were performed with FDB fibers 1–6 h after dissociation; successfully transfected fibers were identified by the presence of YFP or Venus fluorescence.
Measurement of intramembrane charge movements and L-type Ca2+ currents from FDB fibers
Patch pipettes were fabricated from borosilicate glass and had resistances of ≤1.0 MΩ when filled with internal solution, which consisted of (mM): 140 Cs-aspartate, 10 Cs2-EGTA, 5 MgCl2, and 10 HEPES, pH 7.4 with CsOH; fibers were dialyzed in the whole-cell configuration for >20 min before recording. For recording of L-type Ca2+ currents, the external solution contained (mM): 145 TEA-methanesulfonic acid, 10 CaCl2, 10 HEPES, 2 MgSO4, 1 4-aminopyridine, 0.1 anthracene-9-carboxylic acid, and 0.002 tetrodotoxin, pH 7.4 with TEA-OH. For measurement of charge movements, the bath contained (mM): 145 TEA-methanesulfonic acid, 10 CaCl2, 10 HEPES, 2 MgSO4, 1 4-aminopyridine, 0.1 anthracene-9-carboxylic acid, 0.002 tetrodotoxin, 1 LaCl3, and 0.5 CdCl2, pH 7.4 with TEA-OH. Linear components of leak and capacitive current were corrected with −P/4 online subtraction protocols. Output filtering was at 2–5 kHz, and digitization was either at 5 kHz (currents) or 10 kHz (charge movements). Cell capacitance was determined by integration of a transient from −80 to −70 mV using Clampex 10.3 (Molecular Devices) and was used to normalize charge movement (nC/µF) and current amplitude (pA/pF). The average value of Cm was 2.3 ± 0.1 nF (n = 48 fibers). To minimize voltage error, the time constant for decay of the whole-cell capacity transient (τm) was reduced as much as possible using the analogue compensation circuit of the amplifier; the average values of τm and Ra were 1.0 ± 0.02 ms and 467 ± 26 kΩ, respectively. QON was then normalized to Cm and plotted as a function of test potential (V), and the resultant Q-V relationship was fitted according to:
where Qmax is the maximal QON, VQ is the potential causing movement of half the maximal charge, and kQ is a slope parameter. Peak currents were normalized to Cm, and the resultant I-V was fitted according to:
where I is the normalized current for the test potential V, Vrev is the reversal potential, Gmax is the maximum Ca2+ channel conductance, V1/2 is the half-maximal activation potential, and kG is the slope factor. All electrophysiological and Ca2+-imaging experiments were performed at room temperature (∼25°C).
Measurement of intracellular Ca2+ transients in the whole-cell configuration
Voltage-induced changes in myoplasmic Ca2+ were recorded from FDB fibers with Fluo 3 single-wavelength Ca2+ indicator dye (Invitrogen). The pentapotassium salt form of the dye was added to the standard internal solution (see above) for a final concentration of 200 µM. The external solution contained (mM): 145 TEA-methanesulfonic acid, 10 CaCl2, 10 HEPES, 2 MgSO4, 1 4-aminopyridine, 0.1 anthracene-9-carboxylic acid, and 0.002 tetrodotoxin, pH 7.4 with TEA-OH. After entry into the whole-cell configuration, a waiting period of no less than 20 min was used to allow the dye to diffuse into the cell interior. A 100-W mercury illuminator and a set of fluorescein filters were used to excite the dye present in the voltage-clamped fiber. A computer-controlled shutter was used to block illumination in the intervals between test pulses. Fluorescence emission was measured by means of a fluorometer (Biomedical Instrumentation Group, University of Pennsylvania). Fluorescence data are expressed as ΔF/F, where ΔF represents the change in peak fluorescence from baseline during the test pulse, and F is the fluorescence immediately before the test pulse minus the average background fluorescence. The peak value of the fluorescence change (ΔF/F) for each test potential (V) was fitted according to:
where (ΔF/F)max is the maximal fluorescence change, VF is the potential causing half the maximal change in fluorescence, and kF is a slope parameter. Only cells with transients that could be fit with Eq. 3 were used for analysis.
Live cell imaging
Dissociated FDB fibers were examined in rodent Ringer’s solution using a confocal laser-scanning microscope (LSM 510 META; Carl Zeiss). A Plan-Apochromat 63× oil-immersion objective (1.4 NA) was used to view the fiber of interest. CFP and Venus were excited with separate sweeps of the 458- and 514-nm lines, respectively, of an argon laser (30-milliwatt maximum output, operated at 50% or 6.3 A) directed to the cell via a 458/514-nm dual dichroic mirror. The emitted fluorescence was split via a 515-nm long-pass filter; CFP was directed to a photomultiplier equipped with a 465–495-nm band-pass filter, and Venus was directed to a photomultiplier equipped with a 530-nm long-pass filter. The chosen settings precluded recording of fluorescence bleed between CFP and Venus, because CFP is not excited at 514 nm and Venus emission is negligible between 465 and 495 nm (see Papadopoulos et al., 2004). Confocal fluorescence intensity data were recorded as the average of eight line scans per pixel and digitized at 8 bits, with photomultiplier gain adjusted such that maximum pixel intensities were no more than ∼70% saturated.
Assessment of SR Ca2+ store content
FDB fibers were loaded with 5 µM Fluo 3-AM and 0.05% pluronic acid (both from Invitrogen) dissolved in rodent Ringer’s solution for 35 min at 37°C. Fibers were then washed three times in rodent Ringer’s solution with gentle agitation. After a 10-min de-esterification period, Fluo 3-AM–loaded cells bathed in rodent Ringer’s solution were placed on the stage of the LSM 510 META microscope and viewed with a 10× 0.3-NA objective (Carl Zeiss). Fluo 3-AM was excited with the 488-nm line of an argon laser (30-milliwatt maximum output, operated at 50% or 6.3 A, attenuated to 5%). The emitted fluorescence was directed through a dual 488/543 dichroic mirror to a photomultiplier equipped with a 500–530-nm band-pass filter. SR Ca2+ release was induced by 1 mM 4-chloro-m-cresol (4-CmC; Pfaltz & Bauer) delivered via a manually operated, gravity-driven global perfusion system. Fluorescence amplitude data are expressed as ΔF/F, where F represents the baseline fluorescence before application of 4-CmC, and ΔF represents the change in peak fluorescence during the application of 4-CmC.
tsA201 cell culture and expression of cDNA
Low (<20) passage tsA201 cells were propagated in culture medium containing 90% DMEM (Thermo Fisher Scientific), 10% defined fetal bovine serum (GE Healthcare), and 100 µg/ml penicillin-streptomycin (Life Technologies). Cells were trypsinized twice weekly and replated onto 35-mm culture dishes at ∼20% confluence. Lipofectamine 2000 (Life Technologies) was used to transfect these cells within 3–5 d of plating. The transfection mixture contained expression plasmids encoding rat CaV1.3, rabbit β1a, and rat α2δ1 channel subunits at 1 µg of each cDNA per dish. The transfection mixture also contained a plasmid-encoding Venus–Rem construct (1 µg/dish; see above) or YFP (30 ng/dish; Takara Bio Inc.). The day after transfection, cells were trypsinized and replated onto 35-mm plastic for experiments the next day.
tsA201 cells expressing YFP-β1a, YFP-β1a/V-Rem, or YFP-β1a/V-Rem AAA were lysed into 300 µl of lysis buffer (mM: 50 Tris-HCl, pH 7.5, 100 NaCl, 1 MgCl2, 1 DTT, and 0.2% Tween-20) supplemented with 0.1 mM iodoacetamide and 1 mM phenylmethylsulfonyl fluoride (both from Thermo Fisher Scientific). After insoluble material was removed by centrifugation, the homogenates were incubated with a monoclonal antibody directed to Rem (1:200; Santa Cruz Biotechnology, Inc.) for 4–6 h with gentle agitation followed by an overnight incubation with protein A agarose beads (Santa Cruz Biotechnology, Inc.). The agarose beads were then washed twice with lysis buffer and collected after gentle centrifugation at 2,500 rpm. The beads were then resuspended in 30 µl of 1% SDS buffer (Bio-Rad Laboratories) and subjected to SDS-PAGE analysis. Proteins were transferred into a nitrocellulose membrane, blocked with 3% nonfat dry milk (Kroger) in PBS-Tween, and incubated overnight at 4°C with monoclonal antibodies directed to either X(G)FP (1:1,500; Antibodies Inc.) or Rem (1:500). The nitrocellulose membrane was then washed three times with PBS-Tween and incubated at room temperature for 1 h with horseradish peroxidase–conjugated goat anti–mouse IgG (1:10,000; SouthernBiotech). Protein bands were visualized with the SuperSignal West Femto kit (Thermo Fisher Scientific) and viewed on a FluorChem HD2 scanner (Alpha Innotech). Blots were stripped using Restore Western Blot Stripping Buffer (Thermo Fisher Scientific).
L-type Ca2+ current recordings from tsA201 cells
Borosilicate pipettes (2.0–3.0 MΩ) were filled with internal solution, which consisted of (mM): 140 Cs-aspartate, 10 Cs2-EGTA, 5 MgCl2, and 10 HEPES, pH 7.4 with CsOH. The bath solution contained (mM): 145 NaCl, 10 CaCl2, and 10 HEPES, pH 7.4 with NaOH. Electronic compensation was used to reduce the effective series resistance, and linear components of leak and capacitive current were corrected with −P/4 online subtraction protocol. Filtering and digitation were at 2 and 5 kHz, respectively. For tsA201 cell experiments, the average values of Cm, τm, and Ra were 20.0 ± 1.4 pF, 202.0 ± 18.9 µs, and 10.9 ± 0.9 MΩ, respectively (n = 28 cells).
All data are presented as mean ± SEM. Statistical comparisons were made by unpaired t test or by one-way ANOVA (where appropriate), with P < 0.05 considered significant. Figures were made using the software program SigmaPlot (version 11.0; SSPS Inc.).
Online supplemental material
Fig. S1 shows, qualitatively, the successful expression of both V-Rem and V-Rem AAA in FDB fibers by in vivo electroporation. Confocal fluorescence images of six different live, intact FDB fibers overexpressing V-Rem or V-Rem AAA are shown with average intensity profiles for the indicated regions of interest.
Rem inhibits EC coupling in FDB fibers without affecting intramembrane charge movement
Recently, we described the effects of the RGK proteins Rad and Rem on L-type Ca2+ currents and intramembrane charge movement in adult FDB muscle fibers (Beqollari et al., 2014). Although both Rad and Rem inhibited L-type currents by ∼60 and ∼45%, respectively, charge movement was only reduced in fibers transfected with Rad; charge movement for Rem-expressing fibers was virtually identical to charge movement observed in naive fibers. To confirm the latter observation, we used in vivo electroporation (DiFranco et al., 2007) to transfect FDB muscles of otherwise normal 2–3-mo-old C57BL/6J mice with either YFP or a Venus-fused wild-type mouse Rem construct (V-Rem). As expected, FDB fibers overexpressing V-Rem again displayed maximal charge movement virtually identical to YFP-expressing fibers in both amplitude and voltage dependence (Fig. 1, A–C and Table 1). Both Q-V relationships were similar to that reported by Prosser et al. (2009) when La3+ was included in the extracellular recording solution.
Because skeletal muscle EC coupling is coupled directly to translocation of CaV1.l’s voltage-sensing structures (Schneider and Chandler, 1973; Ríos and Brum, 1987; García et al., 1994; Tanabe et al., 1988), we next investigated the impact of Rem on EC coupling by recording myoplasmic Ca2+ transients in response to membrane depolarization (as in Wang et al., 1999; Wu et al., 2012). Ca2+ transients recorded from fibers transfected with V-Rem were substantially reduced compared with the transients of YFP-expressing fibers (0.6 ± 0.1 ΔF/F, n = 8 vs. 1.6 ± 0.4 ΔF/F, n = 6, respectively; P < 0.001; Fig. 1, D–F). No significant effect on the voltage dependence of SR Ca2+ release was observed between the two groups (P > 0.05; Table 1).
SR Ca2+ store content is not significantly affected by overexpression of Rem
As a means to determine whether the V-Rem–mediated reduction in voltage-induced Ca2+ release was a consequence of an altered SR Ca2+ store, we exposed intact fibers loaded with Fluo-3 AM dye to the RYR agonist 4-CmC. In these experiments, 1 mM 4-CmC elicited SR Ca2+ release that was nearly indistinguishable between FDB fibers overexpressing V-Rem and fibers expressing YFP only (5.7 ± 1.0 ΔF/F, n = 9 vs. 6.8 ± 1.8 ΔF/F, n = 5, respectively; P > 0.05; Fig. 2, A–C). The equivalent responses of YFP- and V-Rem–expressing fibers to 4-CmC suggest that depletion of SR Ca2+ store is an unlikely explanation for the ∼65% reduction in Ca2+ transient amplitude observed in V-Rem–expressing fibers.
Rem overexpression does not alter targeting of CaV1.1 α1S or β1a subunits
Because Rem has been reported to alter high voltage-activated Ca2+ channel trafficking in heterologous systems (Béguin et al., 2007; Mahalakshmi et al., 2007; Flynn and Zamponi, 2010; Yang et al., 2010) and in cardiac myocytes (Jhun et al., 2012), one possible explanation for the disruption of EC coupling by V-Rem (Fig. 1, D and F) is that the small G protein redirects CaV1.1 away from triad junctions. For this reason, we examined the subcellular distribution of CaV1.1 α1S and β1a subunits in the absence and presence of coexpressed V-Rem. When expressed in FDB fibers, CFP-tagged α1S subunits of CaV1.1 were targeted to transverse tubules as shown previously for YFP-tagged α1S subunits (DiFranco et al., 2011; Fig. 3 A). The tubular distribution of CFP-α1S was unaffected by coexpression of V-Rem (Fig. 3 B). Likewise, coexpression of V-Rem had little, if any, effect on the subcellular distribution of CFP-β1a (Fig. 3, C and D). Interestingly, the V-Rem fluorescence extended from the transverse tubules into the region of the I band. In this regard, the subcellular distribution of V-Rem overlapped, but did not completely match, the transverse tubular distributions of CaV1.1 α1S and β1a subunits. We do not consider the presence of Rem in the I band to be an artifact of overexpression, as the related RGK protein Rad clearly targets to transverse tubules when expressed in FDB fibers via electroporation (see Beqollari et al., 2014). Moreover, this observation does not affect our interpretation of the data shown in Fig. 3: coexpression of V-Rem did not alter targeting of the channel subunits to the transverse tubules.
Simultaneous introduction of alanines at Rem positions R200, L227, and H229 disrupts interactions with β1a
So far, our data indicate that Rem uncouples the CaV1.1 voltage sensor from RYR1-mediated SR Ca2+ release. However, it is unclear whether this effect of Rem is dependent on the ability of the small GTP-binding protein to interact with the β1a subunit of the CaV1.1 channel complex. In this regard, three highly conserved residues of Rem (R200, L227, or H229) have been identified as being critical for interactions with the β3-subunit isoform (Béguin et al., 2007; Puhl et al., 2014); conversion of any one of these residues to alanine severely impairs binding to β3-subunit isoforms in both yeast-2-hybrid and coimmunoprecipitation assays. To specifically test whether Rem binds to the β1a-subunit isoform, we engineered a V-Rem–based construct with alanines introduced at positions R200, L227, and H229 (V-Rem AAA) and compared its ability to coimmunoprecipitate with a YFP-fused β1a construct (YFP-β1a). In these experiments, a commercially available monoclonal Rem antibody failed to immunoprecipitate YFP-β1a in lysates obtained from tsA201 cells transfected with only YFP-β1a (shown in duplicate in Fig. 4 A, lanes 2 and 6). In contrast, the antibody efficiently immunoprecipitated YFP-β1a subunits when V-Rem was coexpressed with YFP-β1a (Fig. 4 A, lanes 3 and 7). Consistent with the earlier report of Béguin et al. (2007) showing the disruption of the Rem-β3 interaction with alanine single-point mutants, an interaction between V-Rem AAA and β1a was not detectable (Fig. 4 A, lanes 4 and 8). In control experiments, the Rem antibody detected similar levels of immunoprecipitated V-Rem and V-Rem AAA (Fig. 4 B, lanes 3–4 and 7–8). Comparable expression levels for YFP-β1a, V-Rem, and V-Rem AAA mutant were confirmed in a Western blot from total lysates collected before coimmunoprecipitation (Fig. 4 C).
V-Rem AAA fails to inhibit L-type channels expressed in tsA201 cells
We next determined the functional consequences of the disruption of the Rem-β1a interaction. In these experiments, we coexpressed YFP, V-Rem, or V-Rem AAA with CaV1.3 α1D, β1a, and α2δ-1 subunits to detect interactions that occur within a functional L-type channel complex (we used CaV1.3 as a surrogate for CaV1.1 because of its highly efficient and consistent membrane expression in tsA201 cells; see Meza et al., 2013). Predictably, tsA201 cells expressing CaV1.3, β1a, α2δ-1, and V-Rem displayed virtually no L-type current (−3.8 ± 0.7 pA/pF at 0 mV; n = 4; Fig. 5 A). In contrast, cells expressing CaV1.3, β1a, α2δ-1, and V-Rem AAA had L-type currents nearly identical in amplitude (−78.3 ± 16.0 pA/pF, n = 15; Fig. 5, B and D) to control cells expressing the same channel subunits with a YFP transfection marker (−71.0 ± 20.0 pA/pF, n = 9; P > 0.05; Fig. 5, C and D). Successful expression of V-Rem and V-Rem AAA in tsA201 cells was indicated by Venus fluorescence (Fig. 5 E).
V-Rem AAA fails to inhibit CaV1.1 function in FDB fibers
To establish β1a as the mechanistic target of Rem in our experimental system, we overexpressed V-Rem AAA in FDB fibers and assayed its effects on L-type Ca2+ currents, intramembrane charge movement, and EC coupling. Successful expression of V-Rem AAA in FDB fibers was confirmed by Venus fluorescence, which was comparable to the fluorescence generated by V-Rem (see examples in Fig. S1). FDB fibers expressing V-Rem AAA produced sizable L-type currents that were not different than those observed in fibers expressing YFP (−9.8 ± 1.2 pA/pF, n = 6 and −9.0 ± 0.5 pA/pF, n = 5, respectively, at 20 mV; P > 0.05; Fig. 6 A). Likewise, V-Rem AAA had no obvious effect on the magnitude of gating charge movement (24.6 ± 3.4 nC/µF, n = 5; P > 0.05, ANOVA; Fig. 6 B and Table 1), although these fibers did present a steeper Q-V relationship when compared head-to-head with YFP-expressing fibers (P < 0.05, unpaired t test). Most importantly, V-Rem AAA also failed to significantly dampen SR Ca2+ release in response to membrane depolarization (1.4 ± 0.2 ΔF/F, n = 6; P > 0.05, ANOVA; Fig. 6 C and Table 1). Taken with the results in Figs. 4 and 5 showing that V-Rem AAA is unable to interact with β1a, these data indicate that the near ablation of EC coupling by V-Rem (Fig. 1) is largely dependent on structural elements that are important for contact(s) with β1a.
In this study, we found that the RGK family small G protein Rem profoundly inhibits skeletal muscle EC coupling in adult mouse FDB muscle fibers (Fig. 1, D–F). Because the observed reduction in voltage-induced SR Ca2+ release was not likely a consequence of altered CaV1.1 targeting (Fig. 3), impaired voltage sensing (Fig. 1, A–C) or a greatly depleted SR Ca2+ store (Fig. 2), a “communication breakdown” must have occurred between CaV1.1 and RYR1. An intuitive candidate locus for such EC uncoupling is the auxiliary β1a subunit of the CaV1.1 heteromultimer because RGK proteins are established β-subunit–interacting partners (Béguin et al., 2001, 2007; Finlin et al., 2003, 2006; Yang and Colecraft, 2013; Puhl et al., 2014; Xu et al., 2015). Earlier work by Colecraft and colleagues has established that Rem can inhibit L-type CaV1.2 channels expressed in HEK 293 cells without affecting intramembrane charge movement (Yang et al., 2007, 2010), and that this particular mode of Rem-mediated inhibition is dependent solely on an interaction with the β subunit (Yang et al., 2012; Yang and Colecraft, 2013). Because Rem exclusively uses this “low Po” gating mode to inhibit CaV1.1 channel function in differentiated muscle fibers (Beqollari et al., 2014), the observed impairment of EC coupling by Rem is almost certainly dependent on a Rem-β1a interaction. The inability of V-Rem AAA, a Rem construct lacking key structural elements for β binding and channel inhibition (Figs. 4 and 5, respectively), to reduce EC coupling provides additional support for this assertion (Fig. 6).
In addition to inhibiting EC coupling, V-Rem also reduced L-type current in FDB fibers (Fig. 6 A; Beqollari et al., 2014). Because L-type Ca2+ entry has been found to contribute to SR Ca2+ store refilling in myotubes (Cherednichenko et al., 2004) and in differentiated muscle fibers (Lee et al., 2015), it is not beyond possibility that SR stores may be partially depleted in V-Rem–expressing fibers. However, the nearly equivalent responses of YFP- and V-Rem–expressing fibers to 4-CmC (Fig. 2) support the idea that such a mechanism is unlikely to account for the observed effect of Rem on voltage-induced SR Ca2+ release. Likewise, an acute contribution from L-type Ca2+ flux via the channel is also improbable, as the ΔF/F-V curves for YFP- and Rem AAA–expressing fibers both displayed sigmoidal dependencies on voltage, a hallmark indication of skeletal-type EC coupling (see Fig. 6 C, right). If Ca2+ flux were making a small contribution to the transients, its loss could not likely explain the nearly 65% decrease in SR Ca2+ release resulting from coexpression of V-Rem.
Strong circumstantial, but by no means definitive, evidence exists supporting the hypothesis that β1a is directly involved in CaV1.1–RYR1 communication (see Rebbeck et al., 2014). In particular, expression of β1a is essential for EC coupling and enhances L-type current amplitude considerably (Gregg et al., 1996; Strube et al., 1996). Unfortunately, these early results obtained with myotubes cultured from β1 null mice have been difficult to interpret because membrane expression of the principal α1S subunit of CaV1.1 was severely compromised. The confounding obstacle of poor α1S trafficking in β1 null mice was overcome by elegant work with the effectively β1 null relaxed zebrafish mutant line. In the relaxed system, unpartnered α1S subunits trafficked somewhat more effectively to plasma membrane–SR junctions than in mice (Schredelseker et al., 2005). The improved membrane expression of CaV1.1 enabled meticulous ultrastructural examination of relaxed junctions, revealing that β1a is required to organize CaV1.1 into the tetrad arrays that are prerequisite for EC coupling.
Beyond ultrastructure, the zebrafish model system poses nearly the same challenges to deciphering the function of β1a as does the β1 null mouse model. Specifically, the introduction of chimeric β1a constructs or other CaVβ isoforms has been highly useful in the identification of functionally important domains, but information regarding essential intermolecular interactions remains frustratingly difficult to glean (Beam and Bannister, 2010). In efforts to avoid such ambiguity, in vitro approaches have been used to identify interactions of potential functional significance between β1a and RYR1. Indeed, purified full-length β1a subunits do bind fragments of RYR1 in vitro (Cheng et al., 2005; Rebbeck et al., 2011), and a peptide corresponding to β1a residues V490–M524 increases RYR1 Po when applied to lipid layers (Karunasekara et al., 2012). Likewise, dialysis of FDB fibers with a slightly shorter peptide (V490–M508) peptide potentiates both EC coupling and L-type Ca2+ current by nearly 50% (Hernández-Ochoa et al., 2014). Although the use of β1a-based peptide approaches has provided support for the idea that β1a residues V490–M508 are involved in transmitting the signal between CaV1.1 and RYR1, the interpretation of these results has been somewhat limited because of uncertainty of substrate and lack of peptide specificity; one must take into account that a variety of small peptides binds to the ginormous ∼2.3-MDa RYR1 tetramer and/or modulates RYR1 Po in lipid bilayers (e.g., peptides corresponding to the A domain of the CaV1.1 II–III linker, Imperatoxin A, Maurocalcine; El-Hayek and Ikemoto, 1998; Gurrola et al., 1999; Fajloun et al., 2000; Nabhani et al., 2002; Chen et al., 2003; Cui et al., 2009).
In light of the frustrating limitations of the experimental approaches described above, new strategies are needed to further investigate the role of β1a in skeletal-type EC coupling. The use of wild-type Rem or modified Rem constructs to probe junctional architecture represents such an advance because the small G protein disrupts CaV1.1–RYR1 communication in intact, differentiated muscle fibers without deleting or altering the peptide sequences of the endogenous components of the Ca2+ release unit (CRU). Obviously, the next step in this line of investigation is to determine the precise mechanism by which Rem cuts communication between CaV1.1 and RYR1. Based on what is currently known, β1a coordinates the juxtaposition of CaV1.1 with RYR1 in tetrads (Fig. 7 A). So, it is quite possible that the Rem–β1a interaction merely impairs the ability of β1a to facilitate the ultrastructural configuration of CaV1.1 and RYR1 that is requisite for conformational coupling (Fig. 7 B). However, the preservation of tetrad arrays in fibers overexpressing Rem would indicate that the RGK protein is exerting its inhibitory influence on β1a within the intact CRU, which in turn would imply that conformational changes in β1a are involved in CaV1.1–RYR1 coupling (Fig. 7 C). A correlate of the latter interpretation would be that other structures (e.g., II–III loop of the α1S subunit) thought to be involved in transmission of the EC coupling signal are adversely impacted by Rem-induced conformational changes in β1a. Of course, these ideas remain to be tested. In this regard, our current observations provide a new means for the investigation of the β1a subunit as mediator of the communication between CaV1.1 and RYR1 that underlies EC coupling skeletal muscle.
We thank Mr. M.P. Scheele for technical assistance, Drs. S.R. Ikeda and H.L. Puhl, III, for sharing the Venus-Rem expression plasmid, and Drs. K.G. Beam and D. Oskar for continued support.
This work was supported by grants from the National Institutes of Health (AG038778 to R.A. Bannister), the Colorado Clinical and Translational Sciences Institute (J-11-122 to R.A. Bannister), the Boettcher Foundation (to R.A. Bannister), and CONACyT (169006 to U. Meza). D. Beqollari received a stipend from 2T32AG000279-11 (to R.S. Schwartz, University of Colorado Denver-AMC Department of Medicine-Geriatrics Division). Confocal images were acquired in the University of Colorado Denver-AMC Advanced Light Microscopy Core (funded in part by National Institutes of Health/NCRR Colorado CTSI grant UL1 RR025780).
The authors declare no competing financial interests.
Author contributions: D. Beqollari and R.A. Bannister designed research, performed research, analyzed data, and wrote the paper. C.F. Romberg and D. Filipova performed experiments and analyzed data. U. Meza and S. Papadopoulos analyzed data and wrote the paper.
Richard L. Moss served as editor.