Identification and functional characterization of malignant hyperthermia mutation T1354S in the outer pore of the Ca_vα_1S-subunit

Abstract

To identify the genetic locus responsible for malignant hyperthermia susceptibility (MHS) in an Italian family, we performed linkage analysis to recognized MHS loci. All MHS individuals showed cosegregation of informative markers close to the voltage-dependent Ca²⁺ channel (Ca_V) α_1S-subunit gene (CACNA1S) with logarithm of odds (LOD)-score values that matched or approached the maximal possible value for this family. This is particularly interesting, because so far MHS was mapped to >178 different positions on the ryanodine receptor (RYR1) gene but only to two on CACNA1S. Sequence analysis of CACNA1S revealed a c.4060A>T transversion resulting in amino acid exchange T1354S in the IVS5-S6 extracellular pore-loop region of Ca_Vα_1S in all MHS subjects of the family but not in 268 control subjects. To investigate the impact of mutation T1354S on the assembly and function of the excitation-contraction coupling apparatus, we expressed GFP-tagged α_1ST1354S in dysgenic (α_1S-null) myotubes. Whole cell patch-clamp analysis revealed that α_1ST1354S produced significantly faster activation of L-type Ca²⁺ currents upon 200-ms depolarizing test pulses compared with wild-type GFP-α_1S (α_1SWT). In addition, α_1ST1354S-expressing myotubes showed a tendency to increased sensitivity for caffeine-induced Ca²⁺ release and to larger action-potential-induced intracellular Ca²⁺ transients under low (≤2 mM) caffeine concentrations compared with α_1SWT. Thus our data suggest that an additional influx of Ca²⁺ due to faster activation of the α_1ST1354S L-type Ca²⁺ current, in concert with higher caffeine sensitivity of Ca²⁺ release, leads to elevated muscle contraction under pharmacological trigger, which might be sufficient to explain the MHS phenotype.

malignant hyperthermia (MH) is a potentially lethal pharmacogenetic disorder of skeletal muscle inherited in an autosomal dominant mode. Clinical MH episodes are triggered in MH susceptible (MHS) individuals by exposure to volatile anesthetics and depolarizing skeletal muscle relaxants (e.g., succinylcholine; Refs. 8, 19). MH episodes are characterized by uncontrolled increase of muscle activity and heat production. This results in a progressive hyperthermia and muscle rigidity associated with tachycardia, metabolic acidosis, hyperkalemia, and hypoxia that can be deadly, if not treated immediately. Apart from pharmacologically induced episodes, MH patients do not present clinically relevant symptoms.

Because of its subclinical nature, MHS has to be usually diagnosed by an invasive in vitro contracture test (IVCT) to avoid life-threatening incidents during anesthesia in patients from susceptible families. According to the European standardized protocol (37, 31), the contracture responses of muscle tissue upon exposure to halothane and caffeine discriminate individuals as MHS, MH equivocal (MHE), and MH normal (MHN). All symptoms of a MH episode are relevant consequences of aberrant excitation-contraction (EC) coupling that results in abnormal and uncontrolled Ca²⁺ release from sarcoplasmic reticulum (SR) stores in skeletal muscle cells after administration of anesthetics and muscle relaxants (25).

Skeletal muscle EC coupling is known as a signaling mechanism (10) between the sarcolemmal voltage-gated L-type Ca²⁺ channel and the SR Ca²⁺ release channel or ryanodine receptor type-1 (RyR1). Membrane depolarization induces conformational changes in the voltage sensing and pore forming Ca_Vα_1S-subunit of the L-type Ca²⁺ channel, which opens the RyR1 via protein-protein interaction. Consequently, Ca²⁺ released from the SR stores activates the contractile apparatus of the muscle fibers. Genes encoding the key proteins of the EC coupling machinery are the molecular suspects for MH susceptibility. Linkage studies established the RyR1 gene (RYR1) as the main relevant gene (23, 24) that accounts for MHS in >50% of MH families. So far, a genome-wide search has identified additional MHS loci on chromosomes 1q32 (26) (MHS5, OMIM 601887) where the gene coding for the Ca_Vα_1S-subunit is located, 7q21–22 (18) (MHS3, OMIM 154276) the gene locus of the accessory Ca_Vα₂δ-1-subunit, 17q11–24 (22) (MHS2, OMIM 154275) the gene locus of the voltage-gated Na⁺ channel Na_V1.4α-subunit, chromosomes 3q13.1 (36) (MHS4, OMIM 600467) and 5p (34) (MHS6, OMIM 601888). However, missense mutations were so far only discovered in two proteins associated with these MHS loci. More than 178 missense mutations and deletions are described for RyR1, out of which 29 have been functionally characterized (33). In contrast, only two MH-causing mutations were hitherto identified in the second locus, MHS5, mapped to gene CACNA1S (5, 26) encoding the Ca_Vα_1S-subunit. The first identified mutation, R1086H in the intracellular loop connecting homologous repeats III and IV of Ca_Vα_1S, was later also physiologically characterized (41).

Here we report about a south Italian MH family pedigree where the MHS phenotype is again associated with the CACNA1S gene. Specifically, we identified a threonine to serine exchange at position 1354 in the S5-S6 extracellular pore-loop region of homologous repeat IV of Ca_Vα_1S. We introduced this T1354S mutation into the well-characterized GFP-tagged rabbit α_1S-subunit (16) and expressed construct α_1ST1354S in dysgenic (α_1S-null) mouse myotubes to study the influence of this mutation on EC coupling and caffeine-induced SR Ca²⁺ release. Our results demonstrate that mutation T1354S accelerates L-type Ca²⁺ current kinetics and also contributes to a more pronounced caffeine sensitivity of SR Ca²⁺ release. We conclude that the accelerated Ca²⁺ channel activation kinetics together with a higher caffeine-induced Ca²⁺ release in α_1ST1354S-expressing myotubes contribute to amplification of intracellular Ca²⁺ release and thus are connected with the MH phenotype.

MATERIALS AND METHODS

In Vitro Contracture Test and Linkage Analysis

A patient that showed symptoms of an MH episode and 10 other family members (family NA-6) were subjected to the standardized European IVCT (37, 31) (Supplemental Table 1; Supplemental Material for this article is available online at the Am J Physiol-Cell Physiol website). Briefly, force production of standardized muscle biopsy material at concentrations of 2% halothane or 2 mM caffeine was determined. Samples that generated a force of ≥0.2 g for both triggering agents were judged MHS. Samples with this force produced by only one of the triggering agents were classified as MHE.

Genomic DNA was extracted from blood samples with the “Nucleon” procedure (Amersham). Microsatellite repeat marker alleles for MHS loci 1–6 were PCR amplified, labeled with ³²P, and fractionated on 4% polyacrylamide sequencing gels for subsequent visualization by autoradiography or were labeled with 6-FAM, HET, HEX, and NED fluorochromes and separated by capillary electrophoresis with an ABI PRISM 3130 Genetic Analyzer (Applied Biosystems). Linkage analysis was performed using the MLINK program of the LINKAGE package (version 5.1) as implemented in FASTLINK (21, 7). Two-point logarithm of odds (LOD) scores were calculated between each marker locus and MHS, MHE, and MHN characteristics under the standard genetic model defined by the European MH Group genetics section as reported earlier (9, 35). Disease allele frequency was 1/10,000, penetrance of the index case was taken as 1.0, penetrance of the MHS heterozygote was set at 0.98, the phenocopy rate was set at 0.02, and MHE individuals were assigned unknown status and did not contribute to the LOD scores of the pedigree. The maximum two-point LOD score of family NA-6 was calculated under the same genetic model by using an anonymous marker with five alleles, heterozygosity of 0.8, and recombination fraction θ = 0.

Sequence Analysis

PCR primers for the amplification of the 44 exons of gene CACNA1S and of their intron boundaries were designed according to the human CACNA1S sequence (OMIM 114208, primer sequence on request to carsana@unina.it). Amplicons were directly sequenced with dye-terminator chemistry (Applied Biosystems) using a PRISM 3130 Genetic Analyzer (Applied Biosystems). Nucleotide substitutions in the coding sequence were numbered according to the Ca_Vα_1S cDNA sequence (GenBank Accession No. NM_000069).

Construction of Mutant Ca_Vα_1S cDNAs

Nucleotide numbers (nt) are given below in parentheses and asterisks indicate restriction enzyme (RE) sites introduced by PCR using proofreading Pfu Turbo DNA polymerase (Stratagene). All sequences generated and modified by PCR were checked for integrity by sequence analysis (Eurofins MWG, Ebersberg, Germany).

α_1ST1354S.

To substitute Thr1354 by Ser in wild-type rabbit GFP-α_1S (16), a silent AccI* RE site (nt 4049) and a “helper” SpeI* site (nt 4520), together with an A to T transversion (nt 4060) that created triplet TCC coding for Ser instead of ACC for Thr, were introduced by PCR primers. The AatII-SpeI* fragment (nt 3591–4520) produced by a fusion PCR step was coligated with the XhoI-AatII fragment of clone GFP-α_1S (nt 2654–3591) into the corresponding XhoI/SpeI polylinker sites of pBluescript (Stratagene). Subsequently, the XhoI-BglII fragment (nt 2654–4488) of this subclone was inserted into the corresponding RE sites of GFP-α_1S. For simplicity, we refer to this GFP-tagged construct as α_1ST1354S.

α_1SΔ29.

Ca_Vα_1S splice variant α_1SΔ29, lacking exon 29 (coding for 19 of the 32 residues of the extracellular loop linking transmembrane segments IVS3 and IVS4) was found to be expressed at low levels in differentiated human and mouse muscle and up to 80% in cultured myotubes (38). To generate clone α_1SΔ29, nts 3609–3666 that code for exon 29 were deleted from rabbit GFP-α_1S (16) by PCR splicing as described in Tuluc et al. (38).

Cell Culture and Transfections

Myotubes of the homozygous Ca_Vα_1S-null (dysgenic) mouse cell line GLT were cultured on 35-mm culture dishes as described previously (32). GLT myotubes were transfected with GFP-tagged α_1S cDNA constructs at the onset of myoblast fusion using FuGENE 6 transfection reagent according to the manufacturer's protocols (Roche Diagnostics). Three to four days after transfection, myotubes were either fixed for immunocytochemistry (14) or were analyzed in combined patch-clamp and intracellular Ca²⁺ measurements. Additionally, myotubes generated from primary myoblasts (referred as MDG) derived from newborn dysgenic mice (12) were cultured and transfected with FuGENE, following the same protocol as described previously (29). Cells were cultured until day 9 for complete differentiation and were used for caffeine-induced and electrically elicited Ca²⁺ release in perfusion and field stimulation experiments (see below).

Immunofluorescence Analysis

Immunofluorescence analysis was performed as previously described (14). To stain the Ca_Vα_1S-subunit protein, monoclonal antibody mAb 1A (28, 20) (1:2,000) or a rabbit affinity-purified anti-GFP antibody (1:5,000; Molecular Probes) was used together with an Alexa-488-conjugated secondary antibody so that the intrinsic GFP signal and the antibody label were both recorded in the green channel. Affinity-purified antibody 162 against RyR1 (15) (1:2,000) and monoclonal antibody mAb 20A (27) against the α₂δ-1-subunit (1:1,000) were used in combination with Alexa-594-conjugated secondary antibody. Images were recorded on a Zeiss Axiophot fluorescence microscope with a cooled charge-coupled device camera and MetaView image-processing software (Universal Imaging, West Chester, PA).

Patch-Clamp Analysis and Intracellular Ca²⁺ Measurements

Ca²⁺ currents were monitored simultaneously with intracellular Ca²⁺ transients in GLT myotubes on days 7 or 8 after passaging. Myotubes were cultured in 35-mm culture dishes and mounted on an Olympus IX70 inverted microscope equipped with Hoffmann modulation contrast. L-type Ca²⁺ currents were recorded using the whole cell patch-clamp technique from myotubes bathed in an external solution that contained 145 mM TEA-Cl, 10 mM CaCl₂, and 10 mM HEPES (pH 7.4 with TEA-OH). Patch pipettes were pulled from borosilicate glass (Harvard Apparatus) and fire-polished (Microforge MF-830; Narishige), and they had resistances of 1.5–2.5 MΩ when filled with a solution consisting of 145 mM Cs-aspartate, 2 mM MgCl₂, 10 mM HEPES, 0.1 mM Cs-EGTA, and 2 mM Mg-ATP (pH 7.4 with CsOH). Fluorescence signals of depolarization-induced intracellular Ca²⁺ transients were collected during whole cell recordings from a small area of the patch-clamped myotubes with a photomultiplier detection system (Photon Technology International, South Brunswick, NJ) by including 0.2 mM Fluo-4 pentapotassium salt (Molecular Probes, Eugene, OR) to the pipette solution. All recordings were performed with an Axopatch 200A amplifier controlled by pClamp 7.0 software. Currents were elicited with 200-ms depolarizing steps from a holding potential of −80 mV to test potentials between −50 and +80 mV in 10-mV increments. Each test pulse was preceded by a 1-s prepulse to −50 mV to inactivate low voltage-activated T-type currents. Linear leak and capacitative currents were digitally subtracted with a P/4 prepulse protocol (1). Recordings were low-pass Bessel filtered at 2 kHz and sampled at 5 kHz. Ca²⁺ currents were normalized by linear cell capacitance (expressed in pA/pF).

Voltage dependence of intracellular Ca²⁺ release (ΔF/F) was fitted according to a Boltzmann distribution:

$$\Delta {\text{F/F}}_0={(\Delta {\text{F/F}}_0)}_{\text{max}}/\left\{1+\text{exp}\left[-(V-V_{1/2})/k\right]\right\}$$ (1)

where V is the membrane potential, V_1/2 is the potential at which ΔF/F₀ = (ΔF/F₀)_max/2, and k is a slope factor.

The activation phase of the Ca²⁺ current (from current level 0 to 98% of peak current amplitude) was fitted with the following single (Eq. 2) and double exponential (Eq. 3) functions:

$$I(t)={\mathrm{A}}_{\text{mono}}\left[\text{exp}(-t/{\tau }_{\text{mono}})\right]+C$$ (2)

$$I(t)={\mathrm{A}}_{\text{fast}}\left[\text{exp}(-t/{\tau }_{\text{fast}})\right]+{\mathrm{A}}_{\text{slow}}\left[\text{exp}(-t/{\tau }_{\text{slow}})\right]+C$$ (3)

where I(t) is the current at time t after depolarization; A_mono, A_fast, and A_slow are the steady-state current amplitudes of each component with their respective time constants of activation (τ_mono, τ_fast, and τ_slow); and C represents the steady-state peak current.

All recordings were performed at room temperature. Only currents with a maximal voltage error <5 mV due to series resistance were analyzed with Clampfit 8.0 (Axon Instruments, Foster City, CA) and SigmaPlot 8.0 (SPSS) software. Data are given as means ± SE. Statistical significance was determined by unpaired Student's t-test.

Measurements of Caffeine-Induced and Electrically Elicited Ca²⁺ Release

Caffeine sensitivity of intracellular Ca²⁺ release was analyzed in MDG myotubes that proved more stable under field stimulation conditions as GLT myotubes. MDG cells were cultured for 9 days until complete differentiation on glass coverslips coated with carbon and gelatin. Caffeine response curves were monitored from myotubes loaded with Indo-1, a ratiometric Ca²⁺ indicator. MDG cells were incubated for 1 h at room temperature with 6 μM Indo-1-AM (Invitrogen) and 0.12% Pluronic in colorless DMEM. Loaded myotubes were bathed in a normal rodent Ringer solution (145 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES, pH 7.4) and placed in the incubation chamber with a pair of electrodes on the two opposite sides of the chamber in a 6-mm distance. The chamber was mounted on a Zeiss Axiovert epifluorescence microscope and the sample was excited at 350 nm with an Hg 50W illumination system. Fluorescence emissions at 405 (F₄₀₅) and 485 (F₄₈₅) nm were collected with a 2 photomultiplier detection system, (Photon Technology International, S. Brunswick, NJ) and analyzed as ratio (R) of F₄₀₅ to F₄₈₅. Before each recording, a cell-free region of the glass coverslip was selected in the photomultiplier detection window to measure the background fluorescence (R_background), which was afterwards automatically subtracted. First, each cell that was preselected for Ca_Vα_1S expression by green fluorescence, was electrically stimulated three times (1.7 V per mm electrode distance for 2 ms) via the two stimulation electrodes to check for intact Ca_V-RyR1 coupling. Responding myotubes were sequentially exposed to 60-s applications of different concentrations of caffeine (0.5, 1.0, 3.0, 5.0, 10, 30, and 40 mM in Ringer solution) using an automated rapid perfusion system (VC-8 Eight Channel Perfusion Valve Control Systems, Warner Instrument, Hamden, CT). Every caffeine exposure was followed by a 60-s wash step with Ringer solution. Peak values of intracellular Ca²⁺ changes in response to agonist (caffeine) application were expressed as ΔR (R_agonist − R_baseline). Data were analyzed with FeliX (Photon Technology, Princeton, NJ) and SigmaPlot 8.0 (SPSS, Chicago, IL) software packages. Caffeine concentration-response curves were fitted according to the following equation:

$${f(x)={D_{\text{max}}}^{n_{\mathrm{H}}}{/1+(\text{EC}}_{50}/[D])}^{n_{\mathrm{H}}}$$ (4)

where D_max is the maximal response, [D] is the caffeine concentration, n_H is the Hill coefficient, and EC₅₀ is the concentration of caffeine at which the response was half-maximal. Compared with similar experiments in an earlier study (41) MDG myotubes displayed a lower sensitivity to caffeine in the recent study, possibly due to higher passage numbers or slightly different culture condition.

To evaluate the action potential-elicited Ca²⁺ release from intracellular SR stores, transient elevations of intracellular Ca²⁺ were analyzed in Indo-1-AM-loaded myotubes, stimulated by three succeeding 2-ms depolarizing pulses with 1.7 V/mm electrode distance. Action-potential-induced Ca²⁺ transients were measured in single myotubes during a sequential application of 0, 1, and 2 mM caffeine dissolved in Ringer solution containing 0 and 2 mM Ca²⁺, respectively. Only for measurements in presence of 2 mM extracellular Ca²⁺ solution, myotubes were incubated between each stimulation step for 2 min with 10 mM Ca²⁺ and caffeine-free extracellular solution to ensure complete Ca²⁺ reloading of the SR stores. Fluorescence signals were detected by a photomultiplier system (PTI 814; Photon Technology International, Lawrenceville, NJ) and analyzed by FeliX and SigmaPlot 8.0 software.

RESULTS

MHS is Linked to the CACNA1S Gene in a South Italian Family

During anesthesia for tonsillectomy a patient (II:1, Fig. 1) developed tachycardia, hypercapnia, muscular rigidity, and a severe increase in body temperature, so that surgery had to be terminated. Twenty-four hours after this incident creatine kinase levels were elevated to 22.000–31.000 U/l (normal range 0–190 U/l). Linkage analysis of the family (NA-6) of this proband II:1 (Fig. 1) revealed a cosegregation of the MH phenotype with microsatellite markers for the MHS5 locus (MIM 601887), which contains the CACNA1S gene coding for the L-type Ca²⁺ channel subunit Ca_Vα_1S. In fact, LOD score values at a recombination fraction of θ = 0 were identical or close to the theoretical maximum value of +1.76 obtained only with markers in or adjacent to the MH5 locus (Table 1). Low or negative LOD score values obtained with markers for all other MHS loci indicate their irrelevance for the MHS phenotype in family NA-6 (Table 1).

Fig. 1.

Segregation of MHS5 microsatellite markers and of the c.4060A>T malignant hyperthermia susceptibility (MHS) mutation within the pedigree of family NA-6. I, II, and III indicate the 3 generations investigated; squares symbolize male and circles female family members. Black symbols denote individuals identified as MHS by invasive in vitro contracture test; white symbols marked with N denote those identified as MH normal (MHN); black/white symbols denote those identified as MH equivocal (MHE); and symbols with a question mark denote untested family members (disease status unknown). The index patient (II:1; indicated by an arrow) developed an MH crisis during anesthesia. All identical haplotypes are symbolized by bars in identical patterns. The haplotype 4–6-T-9 (black bar) is shared by all of the affected individuals (MHS and MHE) and by none of the unaffected individuals (MHN).

Table 1.

MHS loci, microsatellite markers, and two-point LOD scores in family NA-6

Loci	LOD Score at θ = 0
MHS1 19q13.1
D19S191	+0.7
D19S220	+0.7
D19S223	+0.7
MHS2 17q11.2-24
D17S1855	−1.91
D17S794	n.i.
D17S1825	−2.45
MHS3 7q21-22
D7S669	n.i.
D7S2443	−2.1
D7S440	−1.67
MHS4 3q13.1
D3S1281	n.i.
D3S3638	−1.39
D3S695	−3.21
MHS5 1q32
D1S1726	+1.76
D1S477	+1.16
CACNL1A3.PCR	+1.28
D1S1727	+1.76
MHS6 5p
D5S418	−1.61
D5S398	−1.52

Linkage analysis was performed in family NA-6 to specific microsatellite markers adjacent to each of the six known malignant hyperthermia susceptibility (MHS) loci. Logarithm of odds (LOD) score values calculated for each marker defined the genetic association (values close to the theoretical maximum value of +1.76) or the genetic exclusion (low or negative values) with the MHS phenotype. Genetic association with MHS was found only with locus MHS5 in NA-6. n.i., Not informative (the subjects of generation II are homozygous).

DNA sequencing of all CACNA1S exons and exon-intron boundaries from MHS subjects of family NA-6 ruled out the presence of a previously identified MHS5 mutation (26). Sequence analysis not only identified several already reported polymorphisms (6) but also new nucleotide substitutions (Supplemental Table 2) Only transversion c.4060A>T was found at a heterozygous status in all MHS and MHE subjects of the family (Fig. 1) but was absent in all MHN-tested family members and also in 268 chromosomal sequences derived from MHN-tested individuals unrelated to family NA-6.

Mutation c.4060A>T results in the replacement of Thr1354 by a Ser residue in an extracellular portion of the L-type Ca²⁺ channel Ca_Vα_1S-subunit. Specifically, this position is located in the linker between the pore-forming segment (pore-loop) in homologous repeat IV and transmembrane segment IVS6, 3 amino acids before IVS6 (Fig. 2). Compared with sequences retrieved from National Center for Biotechnology Information databases, this particular Thr at the position 3 amino acids before IVS6 is conserved in all so far sequenced mammalian Ca_Vα_1S-subunits as well as in those from chicken and frog. Interestingly, however, in the faster activating and inactivating neuronal/cardiac Ca_Vα_1C and the neuroendocrine Ca_Vα_1D, this position is occupied by a conserved Ser (Fig. 2B).

Fig. 2.

Thr1354, 3 amino acids before segment IVS6, is highly conserved in all α_1S-subunits but is replaced by Ser in the cardiac/neuronal Ca_Vα_1C- and Ca_Vα_1D-subunits. A: location of the mutation T1354S in a transmembrane domain model of GFP-α_1S. I-IV indicate the 4 homologous repeats, and pore loop domains are indicated in dark gray. Double line indicates the sarcolemmal membrane. B: sequence alignments of the C-terminal half of the pore-forming IVS5-S6 loop of Ca_Vα_1S, Ca_Vα_1C, and Ca_Vα_1D from different species and their positions in the domain model. For each isoform, residues that are identical in all species are boxed in black. The selectivity filter Glu of pore loop domain IV (dark gray cylinder) is indicated by an asterisk. Thr, 3 amino acids before transmembrane segment IVS6 (Thr1354 in human and rabbit), is indicated by a light-gray double arrowhead. As indicated by a black double arrowhead, a conserved Ser is found at the same position in Ca_Vα_1C and Ca_Vα_1D, similar to the MH mutation T1354S. Sequences for cat and horse Ca_Vα_1C and Ca_Vα_1D, for chicken Ca_Vα_1S, and for frog Ca_Vα_1S and Ca_Vα_1C were extracted from genomic assemblies at http://www.ensembl.org using a BLAST search with rabbit Ca_Vα_1S, Ca_Vα_1C, or Ca_Vα_1D cDNA sequences. GeneBank Accession Nos. for Ca_Vα_1S, Ca_Vα_1C, or Ca_Vα_1D, respectively, of the different species are as follows: human (Homo sapiens) NM_000069 (α_1S), NM_199460 (α_1C), NM_000720 (α_1D); rabbit (Oryctolagus cuniculus) NM_001101720 (α_1S), NM_001136522 (α_1C); rat (Rattus norvegicus) NM_017298 (α_1D), mouse (Mus musculus) NM_014193 (α_1S), NM_009781 (α_1C), NM_028981 (α_1D); dog (Canis lupus familiaris) XM_843592 (α_1S), XM_534932 (α_1C), XM_844338 (α_1D); cat (Felis catus) NM_001038605 (α_1S); horse (Equus caballus) XM_001916128 (α_1S); chicken (Gallus gallus) XM_416388 (α_1C), NM_205034 (α_1D); and frog (Xenopus tropicalis) NM_001079461 (α_1D).

Mutation T1354S Accelerates Skeletal Muscle Ca_V Current Kinetics

To test the functional role of this newly discovered MH mutation, we introduced the T1354S mutation into the well-characterized GFP-tagged Ca_Vα_1S of rabbit and transfected this construct (α_1ST1354S) into dysgenic (Ca_Vα_1S-null) myotubes (32, 12). Double fluorescence immunocytochemistry showed that α_1ST1354S correctly incorporates into the plasma membrane and colocalizes with RyR1 (Supplemental Fig. 1) in a punctuate expression pattern identical to that of the wild-type α_1S (α_1SWT). After immunocytochemistry demonstrated that mutation T1354S does not interfere with proper Ca_Vα_1S membrane targeting, we addressed the question whether mutation T1354S affects voltage sensing and/or Ca²⁺-conductivity of the skeletal muscle Ca_V. Whole cell patch-clamp current recordings combined with intracellular Ca²⁺-release measurements were performed in transfected dysgenic myotubes loaded with the fluorescent Ca²⁺ indicator Fluo-4. Recordings from myotubes (Fig. 3A), transfected with α_1SWT (left) and α_1ST1354S (right), showed that both α_1S constructs restored robust L-type Ca²⁺ currents (bottom traces) as well as Ca²⁺ transients (top traces). The averaged amplitudes and voltage dependences of both Ca²⁺ inward currents and intracellular Ca²⁺ transients of myotubes expressing α_1ST1354S were comparable (P > 0.05) to those of myotubes expressing α_1SWT (Fig. 3B and Table 2).

Fig. 3.

L-type Ca²⁺ currents are accelerated in myotubes expressing α_1ST1354S. A: representative traces of intracellular Ca²⁺ release (top) and L-type Ca²⁺ currents (bottom) recorded in response to 200-ms test pulses from dysgenic myotubes expressing wild-type α_1S (α_1SWT) and α_1ST1354S. B: mean voltage dependences and amplitudes of peak Ca²⁺ transients (ΔF/F; α_1SWT, n = 40, and α_1ST1354S, n = 27) and of L-type Ca²⁺ currents (α_1SWT, n = 31, and α_1ST1354S, n = 26) are indistinguishable in α_1SWT- and α_1ST1354S-expressing myotubes. I-V, current-voltage relationship. C: time to peak of Ca²⁺ currents is significantly (**P < 0.01) reduced in myotubes expressing α_1ST1354S (α_1SWT, n = 31, and α_1ST1354S, n = 26).

Table 2.

Properties of intracellular Ca²⁺ transients and L-type Ca²⁺ currents

	α1SWT	α1ST1354S
Ca2+ transient
(ΔF/F0)max	1.81 ± 0.14	1.74 ± 0.19
n	40	27
Ca2+ currents
Imax, pA/pF	2.97 ± 0.20	3.26 ± 0.44
Vact1/2, mV	34.68 ± 0.95	33.85 ± 0.82
Vrev, mV	80 ± 1.25	83 ± 1.76
ttp, ms	99.8 ± 6.6†	67.1 ± 7.3
n	31	26
Double exponential fit
τslow, ms	30.4 ± 3.38*	19.8 ± 3.7
τfast, ms	5.05 ± 0.36*	3.69 ± 0.5
n	18	13
Monoexponential fit
τmono, ms	35 ± 3.6	24.3 ± 5.0
n	13	13

Data are means ± SE for n myotubes, analyzed by fitting each recording with appropriate equations and their definitions as described in materials and methods. Properties of depolarization-induced Ca²⁺ transients and L-type Ca²⁺ currents were determined by combined whole cell patch-clamp and intracellular Fluo-4 Ca²⁺ recordings from dysgenic myotubes expressing wild-type α_1S (α_1SWT) and α_1ST1354S.

^*P < 0.05;

^†P < 0.01.

Nevertheless, α_1ST1354S-expressing myotubes showed significantly distinct Ca²⁺ current kinetics compared with α_1SWT. In Fig. 3A, a representative recording from an α_1SWT-transfected myotube shows slow-activating Ca²⁺ currents as typical for Ca_Vα_1S (bottom traces, left), which is in contrast to the more accelerated inward currents, recorded from an α_1ST1354S-transfected myotube (bottom traces, right). On average, time to peak values of maximal current traces (40 or 50 mV) were significantly (P < 0.01) reduced from 99.87 ± 6.6 ms in α_1SWT traces (n = 31) to 67.1 ± 7 ms in α_1ST1354S (n = 26; Fig. 3C).

Fast Activation Kinetics is an Intrinsic Property of α_1ST1354S

We previously demonstrated that the siRNA knockdown of the auxiliary α₂δ-1-subunit of the skeletal muscle Ca_V in myotubes leads to accelerated activation kinetics of the skeletal muscle L-type Ca²⁺ current (30). As the T1354S mutation accelerated the current activation kinetics, and considering its localization in an extracellular loop, it appears possible that this mutation might disturb the interaction between Ca_Vα_1S and α₂δ-1. To evaluate the localization and interaction of Ca_Vα_1S and α₂δ-1, double immunofluorescence labeling of α_1S and α₂δ-1 was performed in dysgenic myotubes transfected with α_1SWT and α_1ST1354S. Figure 4A demonstrates that the T1354S mutation does not notably alter the distribution pattern of α₂δ-1. The staining patterns of α_1ST1354S-expressing myotubes were indistinguishable from those of myotubes expressing α_1SWT. To elucidate whether mutation T1354S could affect the functional interaction of α_1S and α₂δ-1, the activation of the Ca²⁺ currents was further analyzed. As previously described the activation phase of the skeletal muscle Ca²⁺ current can be best described by monoexponential or double exponential functions depending on the individual recording (2, 4). In double-exponentially fitted recordings, the fast component represents a population of channels not associated with α₂δ-1, whereas the slow component represents a population of channels regulated by α₂δ-1 (30). Kinetic analysis showed that in both channels (α_1SWT and α_1ST1354S) the slow and the fast components (A_slow and A_fast) contribute with comparable fractions (60 and 40% for WT and 64 and 36% for T1354S) to the total current amplitude (Fig. 4B), indicating similar interactions with α₂δ-1. In contrast, both time constants (τ_slow and τ_fast) were significantly (P < 0.05) reduced (19.8 ± 3.7 ms and 3.7 ± 0.5 ms; n = 13) in α_1ST1354S compared with those of α_1SWT (30.4 ± 3.38 ms and 5.05 ± 0.36 ms; n = 18; Fig. 4C and Table 2). A similar trend was found with the recordings adequately fitted monoexponentially. Although not statistically significant, the time constant (τ_mono) showed a tendency to be shorter (P = 0.1) in α_1ST1354S (24.3 ± 3.6 ms; n = 13) compared with α_1SWT (35 ± 3.6 ms; n = 13; Fig. 4C and Table 2).

Fig. 4.

Interaction between Ca_Vα_1S and α₂δ-1 is not affected by the T1354S mutation. A: double immunofluorescence labeling of GFP-tagged α_1S (a and b, clusters indicated by arrows) and of α₂δ-1 (c and d, clusters indicated by arrows) in myotubes expressing α_1SWT (left) and α_1ST1354S (right). Merged images (e and f) show proper colocalization of both α_1S-constructs with the endogenous α₂δ-1. B and C: activating phases of the Ca²⁺ currents from Fig. 3A were fitted either by double or monoexponential functions and corresponding contributions to the total current amplitudes (B) and time constants of activation (C) were analyzed. B: relative contribution of fast and slow fractions to total current amplitude was similar for α_1SWT (white background, black hatched bars) and α_1ST1354S (black background, white hatched bars). C: time constants for all slow and fast components (τ_slow, τ_fast, α_1SWT, n = 18, and α_1ST1354S, n = 13) were significantly reduced (*P < 0.05) in α_1ST1354S-expressing (black background, white hatched bars) compared with α_1SWT-expressing (white background, black hatched bars) myotubes. A similar tendency was observed for τ_mono (α_1SWT, n = 13, and α_1ST1354S, n = 13) (C).

Caffeine Sensitivity of SR Ca²⁺ Release in α_1ST1354S-Expressing Myotubes

Clinical manifestations of MH episodes are associated with a nonphysiological increase of myoplasmic Ca²⁺ concentrations. Skeletal muscles from MHS individuals show a higher response to RyR1 activators like caffeine, halothane. and 4-chloro-m-cresol (42, 44). It has been shown that Ca²⁺ release induced by caffeine is downregulated by an intact interaction of RyR1 with the α_1S-subunit (41). Accordingly, we considered the possibility that the T1354S mutation accounts for an increased caffeine sensitivity of RyR1. To this aim, we generated caffeine dose-response curves of α_1SWT- and α_1ST1354S-expressing myotubes as well as of α_1S-null (dysgenic) myotubes (Fig. 5A). Before the application of varying concentrations of caffeine, three extracellular electrical pulses were applied to each cell to check for intact Ca_V-RyR1 coupling. As expected, no electrically evoked Ca²⁺ transients were observed in dysgenic myotubes. The response of myotube to 60-s sequential exposures to caffeine concentrations of 0.5, 1, 3, 5, 10, 30, and 40 mM was determined. As shown in Fig. 5B, myotubes started to respond at a 3 mM caffeine concentration. More responding cells at 3 and 5 mM caffeine were observed in cultures expressing α_1ST1354S compared with cultures expressing α_1SWT, however, not as many as in untransfected dysgenic myotubes (Fig. 5B).

Fig. 5.

Caffeine sensitivity of RyR1 in α_1ST1354S-expressing myotubes. A: representative caffeine concentration-response curves recorded from Indo-1-loaded myotubes transfected with α_1SWT (top), α_1ST1354S (middle), and untransfected dysgenic myotubes (bottom). At the beginning of each trace, a train of 3 extracellular electrical stimuli was applied to elicit voltage-gated Ca²⁺ release (black arrows). No electrically induced Ca²⁺ release was observed in untransfected dysgenic myotubes. Subsequently, increasing concentrations of caffeine were applied for 60 s to each myotube separated by 60-s washing steps. B: number of cells (%) responding to caffeine concentrations of 3 and 5 mM. C: concentration dependence of caffeine-induced Ca²⁺ release for α_1SWT (●, n = 19), α_1ST1354S (○, n = 22), and dysgenic myotubes (▾, n = 21).

Accordingly, also the dose-response curve of Ca²⁺ release in untransfected dysgenic myotubes tended to shift towards lower caffeine concentrations compared with α_1SWT-expressing myotubes. In α_1ST1354S-expressing myotubes, the sensitivity of caffeine-induced Ca²⁺ release was closer to that of dysgenic myotubes than to α_1SWT-expressing (Fig. 5C). The EC₂₅ values that best describe the differences between the dose-response curves were 4.78 ± 0.47 mM for dysgenic, 4.6 ± 0.50 mM for α_1ST1354S-expressing, and 5.34 ± 0.50 mM for α_1SWT-expressing myotubes.

The averaged values of the maximal caffeine-induced Ca²⁺ release were similar in untransfected myotubes (0.416 ± 0.018; n = 21) and in α_1ST1354S (0.426 ± 0.025; n = 22)- and α_1SWT (0.424 ± 0.025; n = 19)-expressing myotubes.

Accelerated Ca²⁺ Current Activation Through α_1ST1354S Leads to an Increase of SR Ca²⁺ Release in the Presence of Caffeine

Above we showed that L-type Ca²⁺ currents through α_1ST1354S have accelerated activation kinetics. It is to be expected that faster activation leads to higher Ca²⁺ influx during the time frame of a skeletal muscle action potential (3–5 ms). This correlation is reflected as a trend by integrating the early phase of L-type Ca²⁺ current traces through α_1SWT- and α_1ST1354S-expressing myotubes (Supplemental Fig. 2). Consequently, we were interested whether this higher initial Ca²⁺ influx leads to an increased activation of SR Ca²⁺ release under the influence of caffeine. To test this hypothesis, we analyzed action-potential-induced Ca²⁺ release induced by extracellular electrical stimuli in α_1SWT- and α_1ST1354S-expressing myotubes in presence of subthreshold (1 and 2 mM) concentrations of caffeine (Fig. 6). Three single electrical pulses were applied to the myotubes during sequential 30-s application of 0, 1, and 2 mM caffeine under exposure to 2 mM extracellular Ca²⁺ (Fig. 6A). Between each electrical stimulation phase, myotubes were incubated for 2 min with 10 mM extracellular Ca²⁺ to allow optimal loading of the SR. To test the potential effect of higher Ca_Vα_1S Ca²⁺ influx on SR Ca²⁺ release, we also included control experiments on myotubes expressing construct α_1SΔ29, a Ca_Vα_1S splice variant with an eightfold increased current density and accelerated current activation (38). Figure 6B shows that Ca²⁺ influx through the distinct Ca_V constructs indeed correlates with intracellular Ca²⁺ release under caffeine trigger. In the presence of 2 mM caffeine, RyR1 Ca²⁺ release was significantly (P < 0.05) increased in myotubes expressing α_1ST1354S (0.22 ± 0.04; n = 23) and α_1SΔ29 (0.24 ± 0.04; n = 29) compared with those expressing α_1SWT (0.11 ± 0.02; n = 19). Additionally, we measured action potential-induced Ca²⁺ release during sequential 30-s application of 0, 1, and 2 mM caffeine in Ca²⁺ free extracellular solution (Fig. 6C). Ca²⁺ release was undistinguishable in α_1SWT (n = 13)-, α_1ST1354S (n = 6)-, and α_1SΔ29 (n = 14)-expressing myotubes also in the presence of 1 and 2 mM caffeine (Fig. 6D).

Fig. 6.

Ca²⁺ release induced by extracellular electrical stimuli in α_1SWT- and α_1ST1354S-expressing myotubes. A and C: Ca²⁺ transients from myotubes were elicited by 3 extracellular electrical pulses during sequential 30-s application of 0, 1, and 2 mM caffeine (indicated by gray bars of different width) in extracellular solutions containing 2 mM Ca²⁺ (A) and 0 mM Ca²⁺ (C, bottom lines). Between the pulse trains myotubes were incubated for 2 min with extracellular caffeine-free solution containing 10 mM Ca²⁺ for SR reloading (A). B: integral of Ca²⁺ transients in 2 mM extracellular Ca²⁺ as a measure for total Ca²⁺ release in response to one action potential. A trend towards higher Ca²⁺ release with constructs α_1ST1354S (n = 23) and α_1SΔ29 (n = 29) compared with α_1SWT (n = 19) is observable under 1 mM caffeine and is significant under 2 mM caffeine (*P, #P < 0.05, respectively). D: integral of Ca²⁺ transients in 0 mM extracellular Ca²⁺ shows no difference between α_1SWT (n = 13), α_1ST1354S (n = 6), and α_1SΔ29 (n = 14) in relation to different caffeine exposures.

DISCUSSION

While >178 MH-causing mutations are described for the skeletal muscle RyR1, with 29 of them functionally characterized (33), so far only two MH mutations for Ca_Vα_1S, R1086H, and R174W were identified (5, 26) with only one (R1086H) physiologically characterized (41). Here, we identified and functionally characterized a T1354S exchange in Ca_Vα_1S found in an Italian MHS family as the third MHS-related mutation in the CACNA1S gene. While the previously described mutation R174W leads to a loss of a positive charge in the IS4 voltage sensing segment and the R1086H mutation involves the intracellular linker between repeat III and IV, the newly identified T1354S mutation is located in the extracellular portion of the linker between the pore loop and the S6 transmembrane helix in homologous repeat IV. Intuitively, it might appear unlikely that a mutation in an extracellular domain could influence RyR1 Ca²⁺ release. However, in our experiments this mutation indeed influences the activation properties of the skeletal muscle Ca_V and in consequence leads to increased RyR1 Ca²⁺ release under caffeine exposure. Extracellular regions of voltage-dependent ion channels were already identified to be important for the regulation of inward currents. Episodic ataxia in humans has been associated with the G293R missense mutation in the N-terminal extracellular part of the S5-S6 linker of homologous repeat I of the neuronal Ca²⁺ channel Ca_Vα_1A (43). When heterologously expressed, mutant Ca_Vα_1AG293R leads to reduced current density without changing single channel conductance, to a shift in the current-voltage relationship to more positive potentials, and to faster inactivation (40) of Ca²⁺ currents. Similarly, the P601L mutation in the extracellular S5-S6 linker region in repeat II of mouse Ca_Vα_1A (11, 13) induces the tottering phenotype, a neurological disorder, characterized by seizures and ataxia. If mutant Ca_Vα_1AP601L was heterologously expressed, considerably reduced Ca²⁺ currents were recorded despite intact gating charges (39). This current reduction was shown to be a consequence of reduced channel open probability (11). Recently, an embryonic splice variant of Ca_Vα_1S has been identified in which the lack of exon 29 shortens the extracellular linker connecting transmembrane segments IVS3 and IVS4 (38). This splice variant is characterized by an eightfold increase in current density and a 30-mV left shift in the voltage dependence of the current. Consequently, the observed acceleration of Ca²⁺ inward currents induced by the T1354S mutation is not entirely unexpected.

Ca²⁺ currents recorded from skeletal myotubes, in which the α₂δ-1-subunit was depleted display current kinetics that are very similar to the recordings obtained with α_1ST1354S (30). Like in α_1ST1354S-expressing myotubes, Ca²⁺ inward currents in α₂δ-1-depleted myotubes do not show a change in maximal current density compared with wild-type myotubes but display significantly accelerated current activation. Because of these similarities and due to the fact that mutation T1354S is located in an extracellular region, potentially in the interaction domain with the extracellular portion of the α₂δ-1 (17), we tested whether T1354S exchange might influence the interaction with α₂δ-1. However, immunocytochemical staining revealed perfect colocalization of α_1ST1354S with α₂δ-1. Further, it was shown that the acceleration of Ca²⁺ currents in α₂δ-1-depleted myotubes results from a strong reduction or even elimination of the slow component contribution to the total current amplitude, whereas no change was observed in the time constants of the individual component (30). In contrast, the recordings with α_1ST1354S showed reduced time constants for both kinetic components without a change in the relative contribution of fast and slowly activating channel populations. Furthermore, the time constant of the currents, best fitted by a monoexponential function, was also reduced. Thus the number of channels with and without association to α₂δ-1 does not appear to change upon expression of α_1ST1354S. Instead, these data support a model that understands the acceleration of activation as an intrinsic property of Ca_Vα_1S induced by the T1354S mutation.

We propose that the acceleration of Ca_Vα_1S Ca²⁺ inward currents leads to an increased influx of Ca²⁺ during the skeletal muscle action potential. Although peak current sizes of α_1ST1354S were identical to α_1SWT after the current has reached its maximal amplitude in 200-ms test pulses, only the initial phase of these 200-ms is relevant for the in vivo situation, as action potentials in skeletal muscle last only a few milliseconds. In this initial phase, an accelerated current activation can greatly contribute to the total Ca²⁺ influx into the myotube. Indeed, integration of the first milliseconds (5, 10, or 15 ms) of the traces recorded from α_1ST1354S suggested increased net Ca²⁺ influxes compared with α_1SWT (Supplemental Fig. 2). According to our model, this elevated Ca²⁺ influx leads to an additional Ca²⁺-dependent activation of RyR1 especially after sensitization of the RyR1 under caffeine (Figs. 5 and 6) and thus induces enhanced SR Ca²⁺ release. Thus, while in cardiac muscle-type EC coupling the fast Ca²⁺ influx through Ca_Vα_1C essentially determines RyR2 Ca²⁺ release (3), we show here that also in skeletal muscle the Ca²⁺ influx through a mutated Ca_Vα_1S can amplify EC coupling, at least under the influence of RyR1 activators, to a significant extent.

We further show that RyR1 exhibits a tendency to a higher sensitivity to caffeine when it is under the control of α_1ST1354S similar to the previously described Ca_Vα_1S mutation R1086H. The combination of the higher Ca²⁺ influx due to accelerated Ca²⁺ currents together with the amplification of RyR1 sensitivity for caffeine (and supposedly also for other RyR1 activators) might appropriately explain the MHS phenotype in subjects carrying the Ca_Vα_1S T1354S mutation.

The lack of an enhancing effect of the T1354S mutation on Ca²⁺ release in the absence of extracellular Ca²⁺ (Fig. 6, C and D) and the unchanged resting Ca²⁺ levels between α_1SWT- and α_1ST1354S-expressing myotubes (Supplemental Fig. 3) strengthen the hypothesis that only the combined effect of both, caffeine and enhanced Ca²⁺ influx affect the SR Ca²⁺ release in the MH phenotype and do not support the idea of a pure activating effect of the T1354S mutation on RyR1.

GRANTS

This work was supported by the Austrian Science Fund (FWF-16098-B04 to M. G. and FWF-P20059-B05 to B. E. F.), by the Medizinische Forschungsförderung Innsbruck (MFI-6180 to J. S. and MFI-2007-417 to P. T.), by the grant of MIUR-Prin 2007, Italy (20074T4MLS_002 to A. C.), by the Ministero della Salute (Roma), Convenzione CEINGE-MIUR (2000) art 5.2 (to F. S.) and Convenzione CEINGE-Regione Campania (to F. S.), and by the progetto S.co.Pe, Centro di eccellenza riconosciuto dal MIUR ex DM 11/2000.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).