|Home | About | Journals | Submit | Contact Us | Français|
In aged skeletal muscle, changes to the composition and function of the contractile machinery cannot fully explain the observed decrease in the specific force produced by the contractile machinery that characterizes muscle weakness during aging. Since modification in extracellular Ca2+ entry in aged nonexcitable and excitable cells has been recently identified, we evaluated the functional status of store-operated Ca2+ entry (SOCE) in aged mouse skeletal muscle. Using Mn2+ quenching of Fura-2 fluorescence and confocal-microscopic imaging of Ca2+ movement from the transverse tubules, we determined that SOCE was severely compromised in muscle fibers isolated from aged mice (26–27 months) as compared with those from young (2–5 months) mice. While reduced SOCE in aged skeletal muscle does not appear to result from altered expression levels of STIM1 or reduced expression of mRNA for Orai, this reduction in SOCE is mirrored in fibers isolated from young mice null for mitsugumin-29, a synaptophysin-related protein that displays decreased expression in aged skeletal muscle. Our data suggest that decreased mitsugumin-29 expression and reduced SOCE may contribute to the diminished intracellular Ca2+ homeostatic capacity generally associated with muscle aging.
Aging effects on muscle function have been associated with reduced specific contractile force and sarcopenia (Faulkner et al., 1995; Gonzalez et al., 2000). The resulting muscle weakness is a significant factor contributing to falls that lead to serious injury. A decline in skeletal muscle function is also a major contributor to decreased mobility and independence in the elderly, which leads to deterioration of quality of life (Moreland et al., 2004). Although the underlying mechanisms are not fully established, compromised intracellular Ca2+ homeostasis may contribute to the progression of aging-related effects on muscle function (Delbono, 2002; Payne et al., 2004; Weisleder et al., 2006).
To initiate contraction in skeletal muscle fibers, Ca2+ is released from the sarcoplasmic reticulum (SR) into the cytosol through the ryanodine receptor (RyR). To terminate contraction, Ca2+ is sequestered into the SR through the action of SR/endoplasmic reticulum Ca2+-ATPase (SERCA). During each cycle of contraction, a fraction of Ca2+ exits the cell through the plasma membrane Ca2+ pump (Hemmings, 2001; Kurebayashi & Ogawa, 2001; Weigl et al., 2003). Thus, maintaining functional levels of Ca2+ within the SR and the cytosol-regulated Ca2+ entry across the plasma membrane is essential for muscle fibers to function at different developmental stages, and in response to various stress conditions. Store-operated Ca2+ entry (SOCE) mediates extracellular Ca2+ ([Ca2+]o) entry in response to reduction of intracellular Ca2+ ([Ca2+]i) stores (Parekh & Putney, 2005; Brotto et al., 2007).
Extensive studies link SOCE to cell development, proliferation and apoptosis in a wide variety of cells (Parekh & Putney, 2005). In skeletal muscle, SOCE has been reported to function in both physiological and pathophysiological processes (Kurebayashi & Ogawa, 2001; Pan et al., 2002; Brotto et al., 2007). Under conditions of increased Ca2+ demand, such as muscle differentiation, exercise and fatigue, SOCE acts as a gate for [Ca2+]o entry to match the increased requirements for Ca2+-dependent processes of the muscle fiber (Dangain & Neering, 1991; Louboutin et al., 1995; Brotto et al., 2004; Zhao et al., 2005). In muscular dystrophy, excessive Ca2+ leakage into the muscle fiber through SOCE and/or other Ca2+ entry pathways contributes to the progression of muscle deterioration (Fong et al., 1990; Vandebrouck et al., 2002).
Recent studies indicate reduced permeability of plasma membrane to divalent cations in aged skeletal muscle (Fraysse et al., 2006), and reduced SOCE was identified in aged neuronal cells (Vanterpool et al., 2005) and aged fibroblasts (Papazafiri & Kletsas, 2003). However, the functional status of SOCE, and any modifications of SOCE machinery, in aged skeletal muscle have not been evaluated. In this study, we used muscles isolated from aged mice as a model for sarcopenia. Mice can act as an appropriate model for sarcopenia since aged mice over 25 months display decreased skeletal muscle mass and reduced contractility in a fashion similar to aged human skeletal muscles (Barton-Davis et al., 1998; Pagala et al., 1998; Hamrick et al., 2006; Rader & Faulkner, 2006). Our experiments allow us to identify compromised SOCE as a phenotype of muscle aging. Decreased SOCE in aged skeletal muscle does not result from altered mRNA levels of stromal interaction molecule 1 (STIM1) (Zhang et al., 2005; Huang et al., 2006; Luik et al., 2006), a putative Ca2+ sensor on the SR/ER membrane, or Orai (Feske et al., 2006; Vig et al., 2006), the pore conducting unit of SOCE in several cell types (Prakriya et al., 2006). We show that SOCE is also reduced in young muscle fibers from mice lacking mitsugumin-29 (MG29) (Nishi et al., 1999), a protein that regulates Ca2+ signaling in skeletal muscle. Our data suggest that compromised SOCE in aged muscle may result from disruption of the signaling process downstream of the currently known components of the SOCE machinery pointing towards MG29 as a molecule involved with such phenomenon.
To test the functional status of SOCE in aged skeletal muscle, we employed Mn2+ quenching of Fura-2 fluorescence to quantify the unidirectional ion flux in intact flexor digitorum brevis (FDB) fibers enzymatically isolated from young (2–5 months) and aged (26–27 months) mice. In this broadly used technique to measure SOCE, Mn2+ is supplied in the extracellular solution to act as a surrogate for Ca2+. Entry of Mn2+ results in quenching of Fura-2 fluorescence measured at a wavelength of 360 nm, the isosbestic point of Fura-2 (Merritt et al., 1989). Caffeine plus ryanodine, both agonists of RyR, was perfused onto FDB fibers to activate RyR and trigger active depletion of the SR Ca2+ store, enabling activation of SOCE. As shown in Fig. 1, aged FDB muscle displays a substantially diminished Mn2+ entry rate compared to that of young muscle, indicating that SOCE induced by SR Ca2+ depletion with RyR agonists is compromised in aged skeletal muscle.
To confirm our observation, we applied skinned muscle fiber methodology to directly visualize SOCE activation and monitor its development from the beginning of SR Ca2+ depletion. Mechanical skinning of the extensor digitorum longus (EDL) muscle fiber does not appear to disrupt the integrity of the transverse tubules (TT) system, as evidenced by the doublet pattern of the TT membrane (Fig. 2A) characteristic of mammalian muscle fibers (Brown et al., 2006). Upon exposure of the skinned-muscle fiber to an intracellular-like solution containing 500 nm free Ca2+, active transport of Ca2+ into the TT compartment leads to increased Rhod-5 N fluorescence (Fig. 2B, loading). Upon addition of 20 μm thapsigargin plus 30 mm caffeine (TG/C), depletion of the SR Ca2+ store is achieved through both the passive and active pathways. This is reflected by the concurrent reduction of Rhod-5 N fluorescence due to exit of Ca2+ from the TT compartment to the cytosol (Fig. 2B, depletion).
Using this technique we were able to trap Rhod-5 N salt into the TTs of young and aged EDL muscle fibers (Fig. 3A). As shown in Fig. 3, SOCE responses were significantly different in young and aged muscles. In aged skeletal muscle, the decrease in Rhod-5 N fluorescence 4 min after application of TG/C is minimal, indicating severely compromised SOCE compared to that in young muscle fibers. Upon addition of NiCl2, a broad inhibitor of SOCE and other Ca2+ entry mechanisms, as shown in Fig. 3(B), the SOCE activity in young muscle fibers is largely inhibited, while there is only minor alteration to the inhibited SOCE in aged muscle fibers. This small effect suggests that there is some leakage of Ca2+ that occurs from the sealed TTs in skinned fibers. These effects of NiCl2 confirm that [Ca2+]o entry is altered in aged skeletal muscle, rather than [Ca2+]i release from the SR.
Recent breakthroughs have identified Orai and STIM1 as components of the SOCE machinery (Zhang et al., 2005; Vig et al., 2006). Three Orai genes (Orai1, Orai2 and Orai3) are found in mammalian genomes (Feske et al., 2006). To test whether the reduced SOCE in aged skeletal muscle was associated with altered expression of STIM1 and Orai, we first conducted a series of quantitative real-time polymerase chain reaction (PCR) experiments in both C2C12 myogenic cell line and gastrocnemius muscles from young and aged mice. We found that STIM1 and Orai1 expression increased significantly during myotube differentiation in C2C12 myoblast cells, whereas the expression of Orai2 and Orai3 did not change (Fig. 4A). This muscle-differentiation mediated up-regulation of Orai1 and STIM1 has likely physiological relevance, as our functional assays have shown that SOCE is absent in undifferentiated C2C12 myoblasts (data not shown), while differentiation of C2C12 myotubes is associated with robust increase in SOCE (Shin et al., 2003).
In adult mouse skeletal muscle, we found that Orai1 was the dominant isoform, suggesting that Orai1 could comprise the principal pore-forming unit for SOCE in skeletal muscle (Fig. 4B). The mRNA for Orai2 and Orai3 were detected at lower levels, and they did not change with muscle aging. Interestingly, no significant changes in mRNA for STIM1 or Orai1 were identified in aged mouse skeletal muscle (Fig. 4B). Thus, age-related decrease in SOCE activity is not likely due to changes in the relative abundance of STIM1 or Orai1 mRNA. However, STIM1 immunostaining on soleus muscle sections showed increased intracellular clustering of STIM1 protein (unpublished observations), suggesting that altered localization of STIM 1 may contribute to the reduced SOCE activity associated with muscle aging.
Our previous study has shown that MG29 protein in FDB skeletal muscles decreases with age (Weisleder et al., 2006) and this phenomenon is also observed in gastrocnemius muscle (Fig. 4C). Our studies have also indicated that phenotypic changes observed in aged muscles from wild-type mice are mirrored in young mg29−/− mice. To determine the physiological function of MG29 in SOCE regulation in adult muscle fibers, we performed SOCE measurement in EDL muscle isolated from the mg29−/− and wild-type mice at young ages (2–5 months) using our skinned muscle fiber methodology. As shown in Fig. 5, skeletal muscle fibers from young mg29−/− mice display significantly compromised SOCE, mirroring the decline in SOCE activity observed in aged wild-type muscle fibers (Figs 1 and and2).2). Our findings suggest that MG29 may act as a modulator of SOCE in skeletal muscle and that the loss of MG29 in aged skeletal muscle could contribute to the disruption of Ca2+ homeostasis in aged skeletal muscle.
In this study, we applied two complementary methodologies to reveal a significant reduction in SOCE activity in adult skeletal muscle fibers from aged mice. Further studies show that this reduced SOCE is not a result of change in the relative abundance of Orai transcripts or STIM1 protein. This reduction in SOCE activity during aging can be associated with a decrease in MG29 expression. Considering the similar results obtained in two separate model systems (i.e. intact FDB fibers and mechanically skinned muscle fibers) and recent findings of decreased SOCE in other cell types from aged animals (Papazafiri & Kletsas, 2003; Vanterpool et al., 2005; Fraysse et al., 2006), it is possible that defective SOCE is a common phenotype of age-related cellular dysfunction.
Interestingly, under our experimental conditions, both intact and skinned muscle fibers display relatively slow activation kinetics for SOCE, which is in good agreement with many previous reports in both skeletal muscle and other tissues (Vazquez et al. 1998; Kurebayashi & Ogawa, 2001; Papazafiri & Kletsas, 2003; Collet & Ma, 2004). In fact, slow SOCE activation kinetics is also in agreement with the recent identification of STIM1 and Orai as components of the SOCE machinery. STIM1 is an ER/SR Ca2+ sensor (Liou et al., 2005; Zhang et al., 2005) that translocates from the ER/SR membrane to regions close to the plasma membrane following depletion of the [Ca2+]i stores (Wu et al., 2006). This movement of STIM1 initiates activation of Orai, a pore-forming unit of SOCE at the plasma membrane (Prakriya et al., 2006; Vig et al., 2006), to gate [Ca2+]o entry (Mercer et al., 2006). It is conceivable that this activation mechanism underlies the slow activation kinetics of SOCE we observe both in skinned and intact muscle fibers. Because specific experimental conditions contribute to studies indicating rapid SOCE activation kinetics (Launikonis & Rios, 2007), to effectively evaluate SOCE function in intact and skinned fiber systems it is important to duplicate experimental conditions in the same species. The elegant studies of Launikonis and Rios were performed in rat muscle fibers. It is important to note that there are significant differences in the fiber composition and the E-C coupling properties of cardiac and skeletal muscles of mice and rats (Kolbeck & Nosek, 1994; Brooks & Conrad, 1999; Bruton et al., 2008). Thus, caution is required when comparing results from these two species as it is possible that SOCE properties may vary between these two species.
The graded activation of SOCE observed in skeletal muscle (Collet & Ma, 2004) likely has physiological relevance as entry of Ca2+ via SOCE must be tightly regulated to prevent muscle damage due to Ca2+ overload, since a rapid surge of Ca2+ into the muscle fiber could lead to muscle damage (Mikkelsen et al., 2004; Verburg et al., 2005; Yeung et al., 2005). In fact, Lamb et al. (1995) have previously demonstrated that high Ca2+ levels may cause uncoupling of the E-C coupling process. Therefore, Ca2+ entry must be a tightly regulated mechanism, suggesting that slow activation kinetics is a likely mechanism to regulate this Ca2+ entry process without the downside of Ca2+-induced damage. The characteristic mammalian doublet morphology of the T-tubular membrane observed under our experimental conditions (Fig. 2A) indicates the quality of our preparations and that Ca2+-induced damage did not occur under these conditions.
Our previous studies have shown that ablation of MG29 leads to dysfunctional SOCE in neonatal skeletal muscle (Pan et al., 2002). A separate study by Ogawa and coworkers with isolated EDL muscle bundles from adult mice suggested that SOCE is functional in the mg29−/− skeletal muscle (Kurebayashi et al., 2003). Here we confirm that SOCE is functional in mg29−/− muscle fibers (Fig. 5); however, we find the rate of SOCE in adult mg29−/− fibers was significantly reduced. It is likely that the apparent differences between our results and those obtained by Ogawa’s group are due to the specific experimental conditions used. The methodology used in the studies presented in Kurebayashi et al. (2003) could only resolve the summation of experimental manipulations on the SR Ca2+ store and, thus, could not resolve the kinetics of SOCE during the depletion process (Kurebayashi et al., 2003). Our experimental approach allows full control of the intracellular milieu, the Ca2+ uptake and release processes in individual muscle fibers, as well as providing the capability to resolve the spatial and temporal aspects of SOCE activation. In addition, our protocols favor the investigation of SOCE under steady-state conditions, which may also account for the kinetics profile observed here.
Our recent studies have revealed multiple pathways that are involved in SOCE activation in skeletal muscle, one that is dependent on conformational changes at RyR and the other that is independent of RyR (Zhao et al., 2006). To test whether both SOCE pathways are affected in aged muscle, we used a combination of thapsigargin and caffeine to activate both the RyR-dependent and RyR-independent SOCE pathways. Furthermore, muscle fibers were exposed to NiCl2 (Fig. 3B), which proved to be effective at reducing Ca2+ entry only in young muscle fibers. Together, these results suggest that both the RyR-dependent and RyR-independent components of SOCE are compromised in aged skeletal muscle.
Since both SOCE activation pathways are disrupted in aged skeletal muscle fibers, it is likely that the components upstream of the initial Ca2+ store depletion are affected in aged skeletal muscle. While we found Orai and STIM1 expression is not altered in aged skeletal muscle, it is possible that additional components of the SOCE machinery that remain to be resolved may be modified in aged muscle. One possibility is altered spatial localization of SOCE machinery in aged skeletal muscle. Our previous findings indicate a functional and physical segregation of the SR Ca2+ pool in aged muscle (Weisleder et al., 2006), suggesting that the physical interaction necessary for SOCE activation between STIM1 in the SR and Orai1 at the plasma membrane could be disrupted in aged skeletal muscle. In fact, we have recently observed that STIM1 immunostaining on soleus muscle display increased intracellular clustering of STIM1 protein (unpublished observations), suggesting that altered localization of STIM1 may play a role to the reduced SOCE activity associated with muscle aging.
In conclusion, our study provides the first evidence that SOCE is significantly compromised in aged skeletal muscle. We find that the expression of STIM1 and Orai1 mRNA, both known molecular components of SOCE, does not appear to be altered in aged skeletal muscle. However, we do find that reduced expression of MG29 may contribute to the decrease in SOCE in aged skeletal muscle. We propose that decreased SOCE would lead to chronic reduction in the SR Ca2+ stores and a decrease in the amount of releasable Ca2+ during contraction cycles in aged skeletal muscle. This dysfunction in Ca2+ homeostasis could ultimately contribute towards muscle weakness during aging, or act as an adaptive response to minimize contraction in aged muscles with limited capacity to resist mechanical force and recover from injury.
Preparation methods for FDB fibers has been described else-where (Wang et al., 2005). Briefly, individual FDB muscle fibers from C57Bl6/J wild-type male mice (Jackson Laboratories, Bar Harbor, ME, USA) were enzymatically disassociated by 2 mg mL−1 type I collagenase (Sigma, St. Louis, MO, USA) and plated onto ΔTC dish (Bioptechs Inc., Butler, PA, USA). Fibers were allowed to attach to the bottom of the dish and loaded with 10 μm Fura-2 AM (Molecular Probes, Eugene, OR, USA) at room temperature for 1 h. To prevent motion artifact in muscle fibers, 20 μm N-benzyl-p-toluene sulphonamide (Sigma), a specific myosin II inhibitor, was applied into the bathing solution 15 min prior to SR Ca2+ depletion. Muscle fibers were then examined on the inverted microscopy (×400 magnification, N.A. 1.3) of a PTI spectrofluorometer system (Photon Technology International, Monmouth Junction, NJ, USA). Fura-2 was excited at 360 nm, a wavelength insensitive to changes in [Ca2+]i, while emission at wavelength of 510 nm was recorded. To deplete SR Ca2+ store, FDB fibers were treated with 20 mm caffeine plus 5 μm ryanodine for 5 min to induce Ca2+ release. The perfusion solution was then switched to 0.5 mm Mn2+ for 5 min to observe the extent of Mn2+ entry and establish the rate of SOCE. For all measurements of SOCE by Mn2+ quenching the fluorescence signal was normalized using values determined by lysis of the cells with 1% Triton X-100 at the end of the experiment. All experiments were conducted at room temperature.
We have recently reported this methodology in mammalian skeletal muscle in detail (Zhao et al., 2005). These techniques were adapted from the methods of Launikonis et al. (2003), originally developed in frog skeletal muscles. Briefly, single intact muscle fibers are dissected from EDL muscle of young and aged wild-type C57Bl6/J mice or mg29−/− mice and cultured in Dulbecco’s modified Eagle’s medium supplied with 2% horse serum for 72 h. Muscle fibers were then mechanically skinned in the presence of 0.5 mm Rhod-5 N tripotassium salt (Molecular Probes) to trap the dye conjugated with 0.5 mm Ca2+ into transverse tubules (TT). Under these conditions, Rhod-5 N exerts the role of Ca2+ buffer (Kd ~ 90 μm) allowing the fiber to remain relaxed during the skinning (Launikonis et al., 2003; Zhao et al., 2005, 2006). Confocal microscopy using the inverted microscope (×600 magnification, N.A. 1.4) of a Bio-Radiance 2100 (Bio-Rad, Philadelphia, PA, USA) revealed the spatial and temporal distribution of Rhod-5 N inside the TTs. First, fibers are exposed to a solution that mimics intracellular conditions [90.6 mm K Glutamate, 18 mm Na Glutamate, 2 mm EGTA-KOH, 6.7 mm MgCl2, 5.4 mm ATP, 15 creatine phosphatase/CP, 0.5 CaCl2, 20 2-bromoethanesulfonate/BES-KOH, 2.5 μg/mL creatine kinase, 5 μm carbonylcya-p-(trifluoromethoxy) phenylhydrazone/FCCP, pCa 7.0, pH 7.1] that promote physiological loading of Ca2+ into the TT and the SR. Then fibers are exposed to a solution that triggers Ca2+ release from the SR, which subsequently activates SOCE [100 mm K Glutamate, 40 mm Na Glutamate, 10 mm EGTA-KOH, 10 mm 1,2-bis(o-aminophenoxy) ethane-N,N,N′,N′-tetraacetic acid/BAPTA, 0.35 mm MgCl2, 0.5 mm ATP, 1 mm CP, 20 mm BES-KOH, 5 μm FCCP, 20 μm TG and 30 mm caffeine, pH 7.1]. The composition of all solutions was calculated using a customized computer software program (Turbo-Pascal 87, version 3.0; Borland International, Scotts Valley, CA, USA) described elsewhere (Brotto et al., 2006). In the Ca2+ depletion solution, magnesium is lowered to 350 μm to facilitate Ca2+ release from the SR (Lamb & Stephenson, 1991; Launikonis & Stephenson, 2000). ATP concentration was reduced to 0.5 mm to limit the active uptake process of Ca2+ by the SERCA pump, thus favoring the Ca2+ release process, to ensure maximal depletion of the SR. Experiments were repeated multiple times and mean values of 10 regions of interest per fiber were analyzed. Rhod-5 N intensity was normalized to the maximal loading intensity, prior to the onset of SR Ca2+ depletion.
The mRNA expression level of Orai1, Orai2, Orai3 and STIM1 was determined by real-time PCR in the C2C12 myogenic cell line and gastrocnemius muscle from C57Bl6/J mice. First, mRNA was extracted from C2C12 myotubes at 0, 6, 12 days of differentiation and gastrocnemeuis muscles of 3- or 26-month-old wild-type mice using a RNeasy mini kit (Qiagen, Valencia, CA, USA) and transcribed into cDNA by M-MLV Reverse Transcriptase (Promega, Madison, WI, USA). Real-time PCR was performed using the following primer set: (i) Orai1 (product size = 191): forward ‘ctgctgggtcaagttctta’, reverse ‘agtgaacagcaaagacgata’; (ii) Orai2 (product size = 159): forward ‘ctgaggtggtcctgctct’, reverse ‘ggtagaagtggatggtgaag’; (iii) Orai3 (product size = 320): forward ‘catccacaatctcaactctg’, reverse ‘atagaagcagaggatggtgt’; and (iv) STIM1 (product size = 199): forward ‘aagagtctaccgaagcagag’, reverse ‘gtgctatgtttcactgttgg’. The glyseraldehyde-3-phosphate dehydrogenase (GAPDH) gene was used as the internal control. GAPDH primer set is the following (product size = 189): forward ‘tatgtcgtggagtctactgg’, reverse ‘cattgctgacaatcttgagt’. Real-time PCR was performed using SYBR Green PCR supermix (Invitrogen, Carlsbad, CA, USA) on a Bio-Rad MyIQ 96-well PCR detection system. The quality of the PCRs was confirmed by detection of uniform melting curve peaks for each primer set. One hundred nanograms of cDNA was added per reaction and the final primer concentration was 200 nm. Experiments were run for multiple times and duplicate wells were included in each repetition. Relative Ct values were calculated as 2CtGAPDH−CtTarget and the value of Orai1 was set as 1.
Gastrocnemius muscle from C57Bl6/J mice at 3 months and 26 months of age were dissected and whole muscle extracts were generated by a hand-held motorized rotary homogonizer (Kontes, Vineland, NJ, USA) using a lysis buffer containing 150 mm NaCl, 1% Trition-X, 0.1% sodium dodecyl sulfate (SDS), 50 mm Tris-HCl (pH 8.0). Protein concentrations were determined by DC protein assay (Bio-Rad) and 10 μg per sample was separated by SDS–polyacrylamide gel electrophoresis at room temperature on 4–12% Tris-glycine gradient gels for 2 h at 60 mAmps on a Mini PROTEAN II gel system (Bio-Rad). Gels were loaded in parallel and one set was stained with Novex Colloidal Blue stain (Invitrogen), per manufacturer’s instructions. Other gels were used for Western blotting using standard techniques with a concentrated mouse monoclonal antibody against MG29, 2.8 μg mL−1 (customer purified by Harlan Bioproducts for Sciences, Indianapolis, IN, USA). Equivalent loading was confirmed using monoclonal β-actin antibody (Sigma), 0.2 μg mL−1. Results were visualized with an ECL + kit (GE Healthcare, Piscataway, NJ, USA) following the manufacturer’s directions.
Values are mean ± SEM. Significance was determined by either Student’s t-test or analysis of variance followed by a Tukey’s post ad hoc test. A value of p < 0.05 was used as criterion for statistical significance.
This work was supported by National Institutes of Health grants to J.M., an American Heart Association (AHA) Scientist Development Grant to M.B. and an AHA postdoctoral fellowship to X.Z.