Ca²⁺ Overload and Sarcoplasmic Reticulum Instability in tric-a Null Skeletal Muscle

Abstract

The sarcoplasmic reticulum (SR) of skeletal muscle contains K⁺, Cl⁻, and H⁺ channels may facilitate charge neutralization during Ca²⁺ release. Our recent studies have identified trimeric intracellular cation (TRIC) channels on SR as an essential counter-ion permeability pathway associated with rapid Ca²⁺ release from intracellular stores. Skeletal muscle contains TRIC-A and TRIC-B isoforms as predominant and minor components, respectively. Here we test the physiological function of TRIC-A in skeletal muscle. Biochemical assay revealed abundant expression of TRIC-A relative to the skeletal muscle ryanodine receptor with a molar ratio of TRIC-A/ryanodine receptor ∼5:1. Electron microscopy with the tric-a^−/− skeletal muscle showed Ca²⁺ overload inside the SR with frequent formation of Ca²⁺ deposits compared with the wild type muscle. This elevated SR Ca²⁺ pool in the tric-a^−/− muscle could be released by caffeine, whereas the elemental Ca²⁺ release events, e.g. osmotic stress-induced Ca²⁺ spark activities, were significantly reduced likely reflecting compromised counter-ion movement across the SR. Ex vivo physiological test identified the appearance of “alternan” behavior with isolated tric-a^−/− skeletal muscle, i.e. transient and drastic increase in contractile force appeared within the decreasing force profile during repetitive fatigue stimulation. Inhibition of SR/endoplasmic reticulum Ca^{2+ ATPase} function could lead to aggravation of the stress-induced alternans in the tric-a^−/− muscle. Our data suggests that absence of TRIC-A may lead to Ca²⁺ overload in SR, which in combination with the reduced counter-ion movement may lead to instability of Ca²⁺ movement across the SR membrane. The observed alternan behavior with the tric-a^−/− muscle may reflect a skeletal muscle version of store overload-induced Ca²⁺ release that has been reported in the cardiac muscle under stress conditions.

Introduction

In skeletal muscle, transient elevation of intracellular Ca²⁺ causes muscle contraction upon depolarization of the plasma membrane that leads to opening of the ryanodine receptor (RyR)³/Ca²⁺ release channel on the sarcoplasmic reticulum (SR) in a process termed excitation-contraction (E-C) coupling (1–3). Efficient operation of E-C coupling in skeletal muscle requires a number of cellular components: first, the junctional structure between SR and transverse tubular invagination of the plasma membrane must be maintained for efficient signal transduction from membrane depolarization to RyR channel opening (4, 5); second, an optimal environment must be preserved in the cytosol to maintain the RyR channel in a functional state; and third, ion homeostasis inside the SR lumen must be adjusted to accommodate the rapid efflux of Ca²⁺ out of the SR, as Ca²⁺ efflux will establish a transient negative potential inside the SR. This negative potential, if not balanced, will impede Ca²⁺ release and slow down the regenerative aspect of the intracellular Ca²⁺ signaling.

In the course of searching for membrane proteins participating in E-C coupling of striated muscle, we identified two isoforms of trimeric intracellular cation (TRIC) channel that were specifically expressed in the SR membrane (6). TRIC-A is preferentially expressed in striated muscles and TRIC-B is ubiquitously present in most mammalian tissues. Purified TRIC-A and TRIC-B protein formed K⁺-permeable channels in lipid bilayer membrane with distinct pharmacological characteristics (6, 7), and share properties of the native SR K⁺ channels reported by Coronado and Miller (8).

Although tric-a^−/− mice are apparently normal and could reach adulthood, the tric-a^−/−tric-b^−/− double knock-out mice die at E9.5 day due to cardiac arrest (6), whereas the tric-b^−/− neonates die shortly after birth due to respiratory failure (9). Considering the distinct tissue-distributing pattern of TRIC-A and TRIC-B and the different modulatory functions of the TRIC-A and TRIC-B channels (7), it is likely that these two isoforms have different physiological functions in different organs. Because skeletal muscle has the most abundant expression of TRIC-A, TRIC-A may play a more important role in skeletal muscle.

In this study, we examine the physiological function of TRIC-A in Ca²⁺ signaling of skeletal muscle. We show that TRIC-A is expressed in the SR membrane of skeletal muscle at a high molecular ratio to RyR. Genetic ablation of TRIC-A leads to SR vacuolization and Ca²⁺ overload in skeletal muscle. In the absence of TRIC-A-mediated K⁺ permeability, chronic SR Ca²⁺ overload creates instability for Ca²⁺ movement across the SR membrane, which leads to the appearance of “mechanical alternans” in the mutant muscle under stress conditions, a phenotype that can be aggravated by the sarcoplasmic/endoplasmic reticulum Ca²⁺-ATPase (SERCA) inhibitor. Our data suggests that TRIC-A contributes to ion movement across SR and is an important component of the E-C coupling machinery in skeletal muscle.

MATERIALS AND METHODS

Animals

Wild type rabbits (Japanese white, 3 kg) were used for microsome isolation from the hindlimb. 3–4-Month-old wild type or tric-a^−/− male mice of the same genetic background (F1:J1 embryonic stem cells × C57BL/6J) were used for all other experiments.

Microsome Protein Profiling

Total microsome were prepared from rabbit skeletal muscle as previously described (10) and solubilized in buffer containing 1% N-dodecyl-β-d-maltoside, 0.25 m sucrose, 2 mm 2-mercaptoethanol, 50 mm Tris-HCl (pH 7.4), and protease inhibitors. Microsomal fraction were separated on 10% SDS-PAGE electrophoresis along with bovine serum albumin (BSA) and stained by Coomassie Brilliant Blue solution (Nacalai, Kyoto, Japan). Protein concentrations were determined by comparing relative density of the target protein with BSA after Coomassie Brilliant Blue staining.

Western Blotting

Total cellular protein was extracted from hindlimb, extensor digitorum longus (EDL), and soleus of wild type and tric-a^−/− mice in a buffer containing 0.3 m sucrose, 20 mm Tris-HCl (pH 7.4) and protease inhibitors. 10 μg of protein were loaded on 12% SDS-PAGE and transferred onto a polyvinylidene difluoride membrane. Antibodies used were polyclonal rabbit antibody against TRIC-A, polyclonal antibody against TRIC-B, mouse monoclonal antibody against RyR (Affinity Bioreagents), rabbit polyclonal antibody against Junctophilin-1, rabbit polyclonal antibody against Na⁺-Ca²⁺ exchanger 1 (gift from Dr. Iwamoto), mouse monoclonal antibody against Na⁺-Ca²⁺ exchanger 3 (gift from Dr. Iwamoto), rabbit polyclonal antibody against calsequestrin 1 (CSQ1), mouse monoclonal antibody against sarcomeric α-actin (Sigma), and goat anti-rabbit secondary antibody.

Oxidized Iron Staining Electron Microscopy (EM)

EDL, soleus, and flexor digitorum brevis (FDB) muscles were dissected from mice and embedded overnight in pre-fixative solution (3% paraformaldehyde, 2.5% glutaraldehyde, 0.1 m cacodylate buffer, pH 7.5) containing 0.1 m potassium oxalate. After brief washing with 0.1 m cacodylate buffer, samples were placed in post-fixative solution (0.1% OsO₄, 0.1 m potassium ferricyanide, and 0.1 m cacodylate buffer, pH 7.4) and fixed for 2 h at room temperature. Then, samples were washed briefly with the same buffer, dehydrated using ethanol/acetone, and embedded in epoxy resin. Thick sections (150–200 nm) were cut and observed without uranyl acetate and lead citrate staining. Slides were examined under a JEOL-1010 transmission electron microscope. 223 muscle cells on average were examined in each sample and the percentage of vacuolization containing electron dense Ca²⁺ deposit in each sample was calculated.

Measurement of Caffeine-induced Ca²⁺ Transient and Osmotic-induced Ca²⁺ Sparks

FDB fibers were isolated by enzymatic disassociation in 2 mg/ml of type I collagenase (Sigma) and loaded with either 10 μm Fura-2 AM (Invitrogen) for measurement of caffeine-induced Ca²⁺ transient and 10 μm Fluo-4-AM for osmotic-induced Ca²⁺ spark for 1 h at room temperature. Fibers were then washed by 2.5 mm Tyrode solution (in mm, 140 NaCl, 5 KCl, 10 Hepes, 2.5 CaCl₂, 2 MgCl₂, pH 7.2). Measurement of caffeine-induced Ca²⁺ transient was performed on microscopic system of a PTI spectrofluorometer (Photon Technology International, Princeton, NJ), and changes in [Ca²⁺]_i were determined as changes in the ratio of Fura-2 fluorescence at excitation wavelengths of 350 and 380 nm, following exposure to 20 mm caffeine in Tyrode solution without Ca²⁺.

Measurements of Ca²⁺ spark were performed on a confocal microscope (Radiance 2100, Bio-Rad) (11). Fibers were first stabilized in 2.5 mm Tyrode solution, then perfused with a 170 mosm hypotonic solution containing (in mm) 70 NaCl, 5 KCl, 10 Hepes, 2.5 CaCl₂, 2 MgCl₂, pH 7.2, for 60–180 s to induce swelling before perfusion was switched back to the initial 2.5 mm Tyrode solution with an osmolarity of 290 mosm. Spark events were caught on a line scan at a sampling rate of 2 ms per line and image analysis was performed using custom routines on IDL software (Research Systems, Inc.). All experiments were conducted at room temperature (23 ± 2 °C).

Measurement of Membrane Potential

SR vesicles were prepared from 3–6-month-old mice as described previously (10). SR vesicles containing 60 g of protein were resuspended in 2 ml of Cl⁻-free SR-loading buffer (in mm: K glutamate 107.8, EGTA-KOH 2, MgSO₄ 6.6, ATP 5.4, creatine phosphate 15, Ca²⁺ gluconate 0.98, BES-KOH 20, pH 7.2, free Ca²⁺ level of pCa = 6.6), supplemented with 0.5 m di-8-ANEPPQ (Invitrogen) for 6 min at room temperature with constant stirring. Potential dependent fluorescence changes in response to 40 μm A23187 or 0.2% Triton X-100 were monitored in a DeltaRAM fluorometer (Photon Technology International) with an emission wavelength of 570 nm and a ratio of the excitation intensity of F₅₁₀/F₄₅₀.

Contractility of Isolated Muscles

As previously described (12), intact EDL and soleus muscle bundles were dissected from wild type and tric-a^−/− mice and mounted in Radnotti glass chambers with built-in platinum stimulating electrodes (Monrovia, CA), bathing in 2.5 mm Ca²⁺ Tyrode solution with 10 mm glucose, saturated with 100% O₂. One tendon of the muscle was attached to a force transducer and the other one to a stationary arm. A Powerlab computer interface program (ADInstruments) was used to control the electrical stimulation protocols. Muscles were adjusted to produce maximal isometric force and equilibrated at 80 Hz for 30 min (1-min interval, 0.83% duty cycle) using squared-wave 300-mA electrical currents of 500-ms duration. Contractile force at a range of frequencies were tested including 1, 5, 10, 20, 30, 40, 50, 60, 80, 100, 120, 140, and 160 Hz, and the maximal force (T_max) was identified. Muscles were then exposed to a fatiguing stimulation protocol (1-s interval, 50% duty cycle) using the 50% T_max stimulation for 5 min. Mechanical alternans in tric-a null muscles were induced by 3 repeated fatiguing protocols, with a 20-min recovery period at 50% T_max stimulation between, but alternans can be seen in the first, second, or the last fatiguing process.

For testing the presence of alternans with SERCA inhibition, cyclopiazonic acid (CPA) (Sigma) was added to the bathing solution in a final concentration of 20 μm for 5 min and intact EDL and soleus muscles were subjected to fatigue stimulation at 20 Hz, followed by 60 Hz fatigue stimulation after a 5-min break. Alternans in either the first or second fatiguing process were counted.

Characterization of tric-a^−/− Mice at Whole Animal Level

Body weight of the male and female mice was measured biweekly from week 2 after birth to week 20. At 12 weeks old, wild type and tric-a^−/− mice were subjected to a graded running exercise beginning at 10 m/min and increased by 2 m/min every 2 min on a rodent treadmill. The session was terminated when the mice refused to run and maximal speed was recorded. In addition, wild type and tric-a^−/− mice were placed on a rod rotating at 28 rpm and retention time on the rod was recorded. Maximum of 120 s was allowed for one rota-rod trial.

Statistics

All data were analyzed by Origin software (OriginLab). Values are mean ± S.E. Significance was determined by Student's t test. A value of p < 0.05 was used as criteria for statistical significance and other p values were as specified in the figure legends.

RESULTS AND DISCUSSION

We identified TRIC-A through an immuno-proteomic approach using monoclonal antibodies against the skeletal muscle triad junction proteins (13). Our initial study has established TRIC-A as a K⁺-permeable channel in the SR membrane and its role in supporting RyR-mediated Ca²⁺ release in both cardiac and skeletal muscles (6). A more recent study revealed the essential role of TRIC-B in Ca²⁺ handling of alveolar epithelial cells and in perinatal lung development and maturation (9). Specifically, tric-b^−/− epithelial cells displayed compromised inositol 1,4,5-trisphosphate receptor-mediated intracellular Ca²⁺ release even though Ca²⁺ overload was present in the endoplasmic reticulum. Because TRIC-A and TRIC-B have different biophysical properties (7), these two channels likely have distinct roles in regulating intracellular Ca²⁺ signaling in both excitable and non-excitable cells.

Embryonic lethality associated with the tric-a and tric-b double knock-out mice prevented our physiological evaluations of TRIC-A and TRIC-B in adult muscle tissues. Fortunately, the tric-a^−/− mice are viable and appear normal under non-stress conditions, for example, our observation revealed that tric-a^−/− mice had normal growth rate and mobility at the whole animal level (supplemental Fig. S1). So we set off to investigate the physiological role of TRIC-A in skeletal muscle under stress conditions in the current study. Toward defining the function of TRIC-A in skeletal muscle, microsome preparations isolated from rabbit skeletal muscle was used to quantify the relative expression level of TRIC-A to other major Ca²⁺ regulatory proteins present in the SR (e.g. RyR, CSQ, and SERCA) (Fig. 1A). Molecular identities of the corresponding proteins were confirmed by Western blot for TRIC-A (Fig. 1A) and CSQ, RyR, and SERCA (not shown). As shown in Table 1, RyR is present in the rabbit skeletal microsome preparation at 16.3 ± 1.4 pmol/mg, and TRIC-A is present at 89 ± 11 pmol/mg. Assuming tetramers of RyR are required for a functional Ca²⁺ release channel (14), and trimers of TRIC-A are required for a functional K⁺ channel (6), these numbers would lead to a molecular ratio of ∼5 to 1 for TRIC-A/RyR. Our method used to quantify the protein levels for RyR, SERCA, and CSQ is comparable with previous reports by other investigators (15, 16). The abundant expression of TRIC-A in skeletal muscle is consistent with the function as a countercurrent channel for RyR Ca²⁺ release, as a redundancy may be necessary to support the repeated Ca²⁺ release events from the SR.

FIGURE 1.

Abundant expression of TRIC protein in skeletal muscles. A, representative Coomassie Brilliant Blue staining showed the relative expression levels of RyR (552 kDa) and SERCA protein (100 kDa) (left panel), as well as CSQ (60 kDa) and TRIC-A protein (33 kDa) (middle panel) in isolated SR microsome from rabbit skeletal muscle, as compared with the BSA standard. Right panel was a Western blot of the isolated microsome from rabbit hindlimb skeletal muscle using rabbit polyclonal anti-TRIC-A antibody. B, representative Western blot showing the differential expression patterns of TRIC-A and TRIC-B in hindlimb, soleus (SOL), and EDL from the wild type mice. C, summarized data of the relative expression levels of TRIC-A and TRIC-B in hindlimb (n = 8), soleus (n = 8), and EDL (n = 8) from wild type skeletal muscle. The expression levels of TRIC-A and TRIC-B in hindlimb were set as 100%, **, p < 0.01.

TABLE 1.

Relative expression level of TRIC-A to other SR resident proteins

Protein contents of RyR (tetramer), SERCA, CSQ, and TRIC-A (trimer) in isolated microsome from rabbit skeletal muscle were summarized in this table. Values were mean ± S.E. (μg/mg microsome) and n = 4. Functional unit (pmol/mg of microsome) was calculated by the formula: protein content/(molecular weight × reported subunit numbers). Reported contents of RyR and SERCA (pmol/mg) in microsome from two other groups were included as a reference.

	Protein content	Functional unit	Reported content
	μg/mg microsome	pmol/mg microsome
RyR (tetramer)	36.6 ± 3.2	16.3 ± 1.4	8–20
SERCA	367 ± 37	3,480 ± 350	3,000–4,000
CSQ	41.0 ± 2.0	680 ± 33
TRIC-A (trimer)	8.8 ± 1.1	89 ± 11

Using immunoblot assays, we found that in addition to TRIC-A, TRIC-B is also present in skeletal muscle (Fig. 1B). Both the fast-twitch EDL muscle and the slow-twitch soleus muscle contain TRIC-A and TRIC-B proteins, but at different concentrations (Fig. 1C). Genetic ablation of TRIC-A did not appear to affect the expression level of TRIC-B, because similar levels of TRIC-B expression were detected in the wild type and tric-a^−/− muscles (supplemental Fig. S2), supporting that TRIC-A and TRIC-B may have complementary functions in Ca²⁺ signaling in muscle tissues. In addition, other major Ca²⁺ regulatory proteins in the tric-a^−/− muscle also showed a similar expression pattern as the wild type control muscle (supplemental Fig. S2), suggesting that germ-line ablation of tric-a did not lead to significant changes in the overall Ca²⁺ signaling machinery in skeletal muscle.

Systemic ablation of tric-a did not appear to affect development of the mice, significant changes in the ultrastructure of the SR network and Ca²⁺ handling properties inside the SR organelle were observed through electron microscopy (EM) (Fig. 2). Specifically, using fixative solution supplemented with potassium oxalate and potassium ferricyanide (K₃Fe(CN)₆) to visualize concentrated Ca²⁺ deposits in intracellular organelles, we detected frequent SR vacuolization and Ca²⁺ deposits in the tric-a knock-out EDL and soleus muscles by EM. Ca²⁺ deposition could be detected only in large-sized SR vacuoles from the tric-a^−/− muscles, confirmed by transmission EM oxidized iron staining images (Fig. 2, B–E). These phenomena were rarely observed in skeletal muscle derived from the wild type controls (Fig. 2A). Summary data for the presence of vacuoles in wide type and tric-a^−/− muscles were summarized in Table 2. The slightly higher basal level of vacuolization in the fast-twitch muscle may reflect the less oxidative capacity of these muscles to endure the anaerobic environment during sample preparation. These structural defects with the adult tric-a^−/− skeletal muscle were similar to what we observed previously with the tric-a^−/−tric-b^−/− neonates, which showed frequent formation of Ca²⁺ deposits and overload of Ca²⁺ inside the SR of the embryonic cardiomyocytes (6).

FIGURE 2.

Detection of large-size vacuoles containing electron-dense Ca²⁺ deposit in tric-a^−/− muscles. A, toluidine blue-stained light microscopic cross-section image of a tric-a^−/− muscle bundle at low magnification, arrows designate myofibrils containing large-sized vacuoles. B, enlarged image of a single tric-a^−/− myofibril containing large-sized vacuoles. C, transmission EM image stained by oxidized iron showed the longitudinal thick section of tric-a^−/− FDB at high magnification. Mitochondria (Mit) was labeled in the image and arrows designate electron dense Ca²⁺ deposits in the large-sized vacuoles (>0.5 μm in diameter), preferentially formed in the terminal SR region. D, enlarged image showing a Ca²⁺-overloaded swollen SR. E, transmission EM image stained by oxidized iron showed the longitudinal thick section of wild type FDB at high magnification.

TABLE 2.

Increased large-size vacuolization containing Ca²⁺ deposit in tric-a^−/− muscles

The incidence of large-sized vacuoles (%) in wild type and tric-a^−/− skeletal muscles were calculated and compared. Data shown are mean ± S.E. Fast-twitch muscles included 4 pairs of EDL and 4 pairs of FDB, whereas slow-twitch muscle included 4 pairs of soleus.

Large vacuoles (%)	Wild type	tric-a−/−
Fast-twitch muscles (n = 8)	0.64 ± 0.26	6.18 ± 2.42a
Slow-twitch muscles (n = 4)	0.25 ± 0.25	11.13 ± 4.56a

^a p < 0.05 compared with wild type.

We used caffeine to test if the total RyR-mediated Ca²⁺ releasable pool from the SR was altered with the tric-a^−/− skeletal muscle. As shown in Fig. 3A, application of 20 mm caffeine evoked a significantly larger Ca²⁺ transient in the tric-a^−/− muscle fiber as compared with the wild type fiber. This result is consistent with the increased Ca²⁺ deposit detected with EM, and similar to our previous study with the embryonic cardiac cells (6) and neonatal epithelial cells (9).

FIGURE 3.

Altered SR Ca²⁺ release properties and membrane potential in tric-a mutant muscles. A, representative trace of intracellular Ca²⁺ transient induced by 20 mm caffeine (arrow) in FDB from wild type (gray) and tric-a^−/− (black) mice (left panel). Fura-2 fluorescence was recorded at excitation wavelengths of 350 (F₃₅₀) and 380 nm (F₃₈₀). The peak caffeine response at F₃₅₀/F₃₈₀ in wild type (open bar) and tric-a^−/− muscle (filled bar) is summarized in the right panel, n = 10. B, representative three-dimensional images of Ca²⁺ spark in wild type and tric-a^−/− FDB fibers. C, normalized amplitude (ΔF/F₀), duration of single spark (ms), and frequency per min of osmotic-induced Ca²⁺ spark in FDB of the wild type (open) and tric-a^−/− mice (filled) were summarized here, n = 13–24. The value of wild type fibers were set as 1. D, representative trace of SR membrane potential measurement by di-8-ANEPPQ dye at F₅₁₀/F₄₅₀ (left panel) in SR vesicle isolated from wild type (gray) and tric-a^−/− (black) mice, and arrows designate when A23187 (Ca²⁺ ionophore) and Triton were added (left panel). Right panel shows summarization of this data, F₅₁₀/F₄₅₀ after addition of Triton was set as 0, n = 6. All values were mean ± S.E., *, p < 0.05.

Our previous studies show that transient osmotic stress applied to the adult skeletal muscle can uncover fluttered Ca²⁺ spark events from the SR (11). Using this methodology, we identified significant defects in Ca²⁺ spark signaling in the tric-a^−/− muscle (Fig. 3B). The frequency of Ca²⁺ sparks was significantly reduced in the tric-a^−/− muscle (Fig. 3C). Although the amplitude of Ca²⁺ sparks was not significantly reduced in the mutant muscle as compared with the wild type fibers, taking into account that tric-a^−/− muscle fiber had a significantly larger releasable SR Ca²⁺ pool (Fig. 3A), it suggests that the Ca²⁺ mobilization process through the RyR channel is compromised due to lack of the TRIC-A-based counter-ion current. The duration of Ca²⁺ spark events seemed to increase in the tric-a^−/− fibers, possibly reflecting a compromised mechanism of Ca²⁺-dependent RyR inactivation.

The compromised Ca²⁺ spark signaling associated with knock-out of the tric-a gene may correlate with changes in TRIC-A-mediated membrane permeability across the SR. To test this possibility, we used the voltage-sensitive fluorescent dye di-8-ANEPPQ to probe the membrane potential of SR vesicles isolated from tric-a^−/− and wild type mice. As shown in Fig. 3D, isolated tric-a^−/− SR vesicles displayed a significantly more negative potential at the cytosolic side as compared with the wild type control, which could be eliminated using Ca²⁺ ionophore A23187, suggesting correlation between the altered SR membrane potential to the increased Ca²⁺ content inside the SR. Overall, our data in Fig. 3 revealed the following aspects with the tric-a^−/− muscle: first, there is overload of Ca²⁺ inside the SR; second, there are defects in membrane permeability across the SR membrane. The absence of TRIC-A-mediated K⁺ permeability is likely the cause for the compromised Ca²⁺ spark signaling in the mutant mice.

Consequences of these ion mobilization defects on performance of the tric-a^−/− muscle was revealed in our ex vivo muscle fatigability assays. Intact muscle bundles were subjected to repeated electrical stimulations to induce fatigue. As shown in Fig. 4A, the fatigue stimulation usually causes the contractile force in wild type muscle to decline exponentially. However, in tric-a knock-out muscle, the continuous force decay was interrupted by a sudden, transient increase in the contractile force generated, similar to the alternans described in myocardium (Fig. 4B). Although the alternans seen in skeletal muscle are of a more irregular nature, we gave these spikes of contractile forces a descriptive name, i.e. mechanical alternans. Our definition of a mechanical alternan is such that the contractile force outbreak has to be 50% higher than its previous force and the outbreaks should be seen at least 10 times during the 5-min fatigue stimulation process. Accordingly, the frequency of mechanical alternans in wild type was 0% for both EDL and soleus muscles, whereas that in tric-a knock-out muscles were 31 and 39%, respectively. To test whether the mechanical alternans were due to damage to the muscle fibers, we compared the force recovery percentage after fatigue stimulation in both wild type and tric-a^−/− muscle bundles. The mutant muscle fibers showed similar force recovery rate for both fast-twitch and slow-twitch muscle bundles (0.67 ± 0.04 for EDL and 0.73 ± 0.07 for soleus) as the wild type controls (0.74 ± 0.05 for EDL and 0.85 ± 0.03 for soleus, all p > 0.05), suggesting that the stress-induced mechanical alternans are not due to changes in contractile machinery of the tric-a^−/− muscle.

FIGURE 4.

Mechanical alternans in TRIC-A absent muscles induced by fatigue stimulation. A, isolated muscle bundles from wild type mice show monotonic decrease in contractile force, when subjected to fatiguing stimulations by the 50% T_max frequency. B, isolated muscle bundles from the tric-a^−/− mice show irregular patterns contractile force, when subjected to similar fatiguing stimulations as in A. Enlarged individual contractile forces shown in the lower panels reveal the appearance of alternans in the tric-a^−/− muscle. C, EDL muscles from the wild type mice were treated with 20 μm CPA for 5 min, and tetanic contractile force measurements reveal the appearance of the characteristic slow relaxing phase (top panel). In the presence of CPA, fatiguing stimulation leads to the appearance of mild alternans in wild type muscle (bottom panel). The trace is representative of 4 of 10 muscle preparations. D, representative traces for the tric-a^−/− muscle as shown using identical experimental conditions as in C. Clearly, CPA treatment leads to aggravation of alternan events in the tric-a^−/− muscle. Trace shown at the bottom panel of D is representative of 8 of 10 muscle preparations tested.

Alternans were first discovered in cardiomyocytes (17–20), which are closely correlated to both Ca²⁺ cycling and mechanical beat of the heart (21). Altered coupling between RyR-mediated intracellular Ca²⁺ release and various mechanisms of Ca²⁺ movement across the sarcolemmal membrane, including L-type Ca²⁺ channel and Na⁺/Ca²⁺ exchanger, may underlie the alternan behavior in cardiac muscle (22). In particular, stress-induced SR Ca²⁺ overload can lead to occurrence of spontaneous extra-systole due to store overload-induced Ca²⁺ release (23), and cardiac RyR channels contain the intrinsic regulatory mechanisms for store overload-induced Ca²⁺ release. Thus, we speculate that the mechanical alternans observed in the tric-a knock-out skeletal muscle may represent instability of the SR Ca²⁺ release machinery due to overload of the SR Ca²⁺ store and reduced membrane permeability for K ions. It is worth noting that although the skeletal RyR/Ca²⁺ release channel is less sensitive to store overload-induced Ca²⁺ release (24), it may be activated under repetitive fatiguing stimulations.

To further test the contribution of SR Ca²⁺ transport to the stress-induced alternans in the tric-a^−/− muscle, we added 20 μm CPA, a potent and reversible SERCA inhibitor (25) into the bath solution. Addition of CPA leads to the appearance of a slow-relaxing component of the contractile fore (80 Hz stimulation) in the wild type muscle (Fig. 4C, top panel), as expected for the reduced SERCA activity (26). Interestingly, in the presence of CPA, fatigue stimulation of the wild type muscle led to the appearance of mild alternan behavior (Fig. 4C, bottom panel, 4 of 10 muscle preparations). More importantly, CPA inhibition of SERCA function could aggravate the alternan behavior in the tric-a^−/− muscle, because more drastic and more frequent alternan events (8 of 10) were observed under such experimental conditions (Fig. 4D, lower panel). This data provides direct evidence that altered SR Ca²⁺ transport is associated with the stress-induced alternans in the tric-a^−/− muscle, and further suggest that ablation of TRIC-A could lead to instability of Ca²⁺ transport across the SR membrane.

In summary, our study demonstrated the important physiological function of TRIC-A in Ca²⁺ signaling of skeletal muscle. Knock-out of tric-a leads to defects in SR membrane structure and membrane permeability, which may be associated with the compromised Ca²⁺ spark signaling in the tric-a^−/− muscle. The irregular mechanical alternans in the mutant muscle is likely due to instability in Ca²⁺ movement across the SR membrane. These findings suggest that TRIC-A is an important regulating moiety for the Ca²⁺ release event in skeletal muscle. Our data should provide insights for further investigating the mechanisms of skeletal muscle E-C coupling process under physiological and pathophysiological conditions.