Transverse tubule remodeling enhances Orai1-dependent Ca2+ entry in skeletal muscle

Antonio Michelucci; Simona Boncompagni; Laura Pietrangelo; Maricela García-Castañeda; Takahiro Takano; Sundeep Malik; Robert T Dirksen; Feliciano Protasi

doi:10.7554/eLife.47576

. 2019 Oct 28;8:e47576. doi: 10.7554/eLife.47576

Transverse tubule remodeling enhances Orai1-dependent Ca²⁺ entry in skeletal muscle

Antonio Michelucci ^1,², Simona Boncompagni ^2,³, Laura Pietrangelo ^2,³, Maricela García-Castañeda ¹, Takahiro Takano ¹, Sundeep Malik ¹, Robert T Dirksen ^1,^†,^✉, Feliciano Protasi ^2,^4,^†

Editors: Richard S Lewis⁵, Ronald L Calabrese⁶

PMCID: PMC6837846 PMID: 31657717

Abstract

Exercise promotes the formation of intracellular junctions in skeletal muscle between stacks of sarcoplasmic reticulum (SR) cisternae and extensions of transverse-tubules (TT) that increase co-localization of proteins required for store-operated Ca²⁺ entry (SOCE). Here, we report that SOCE, peak Ca²⁺ transient amplitude and muscle force production during repetitive stimulation are increased after exercise in parallel with the time course of TT association with SR-stacks. Unexpectedly, exercise also activated constitutive Ca²⁺ entry coincident with a modest decrease in total releasable Ca²⁺ store content. Importantly, this decrease in releasable Ca²⁺ store content observed after exercise was reversed by repetitive high-frequency stimulation, consistent with enhanced SOCE. The functional benefits of exercise on SOCE, constitutive Ca²⁺ entry and muscle force production were lost in mice with muscle-specific loss of Orai1 function. These results indicate that TT association with SR-stacks enhances Orai1-dependent SOCE to optimize Ca²⁺ dynamics and muscle contractile function during acute exercise.

Research organism: Mouse

Introduction

Store-operated Ca²⁺ entry (SOCE), a Ca²⁺ entry mechanism activated by depletion of intracellular stores, is mediated by the coupling between stromal interacting molecule-1 (STIM1), a luminal Ca²⁺ sensor protein in the endoplasmic reticulum (ER) membrane (Roos et al., 2005), and Ca²⁺-permeable Orai1 channels in the plasma membrane (Feske et al., 2006). In skeletal muscle, SOCE is similarly coordinated by a functional interaction between STIM1 in the sarcoplasmic reticulum (SR) and Orai1 in the transverse tubule (TT) (Kurebayashi and Ogawa, 2001; Lyfenko and Dirksen, 2008; Stiber et al., 2008), specialized invaginations of the surface membrane that propagate action potentials to trigger SR Ca²⁺ release during excitation-contraction (EC) coupling. While the precise physiological role(s) of SOCE in muscle remains incompletely defined, SOCE has been proposed to promote myoblast fusion/differentiation (Darbellay et al., 2009) and muscle development (Kurebayashi and Ogawa, 2001; Wei-Lapierre et al., 2013; Carrell et al., 2016), reduce fatigue during periods of prolonged stimulation (Wei-Lapierre et al., 2013; Carrell et al., 2016; Pan et al., 2002; Boncompagni et al., 2017), as well as serve as a counter-flux to Ca²⁺ loss across the TT system during EC coupling (Koenig et al., 2018). Importantly, mutations in the genes that encode STIM1 and Orai1 underlie an array of disorders with clinical myopathy as central deﬁning component. SOCE dysfunction also contributes to the pathogenesis of several other muscle disorders (Michelucci et al., 2018). Thus, SOCE plays an important role in both normal muscle development/function and muscle disease.

In resting mammalian skeletal muscle, STIM1 proteins are found throughout the I band of the sarcomere (Stiber et al., 2008; Wei-Lapierre et al., 2013; Boncompagni et al., 2017). On the other hand, Orai1 channels at rest are found almost exclusively at the A-I band junction, in accordance with their localization within the TT membrane of the triad (Wei-Lapierre et al., 2013; Boncompagni et al., 2017). A modest degree of STIM1-Orai1 co-localization at the A-I band junction at rest has been suggested to explain in part the uniquely rapid activation of SOCE (<1 s) observed in skeletal muscle (Koenig et al., 2018; Launikonis and Ríos, 2007; Edwards et al., 2011). However, STIM1 and Orai1 localization in muscle is not static, as acute treadmill exercise drives a significant reorganization of the sarcotubular membranes in the I band of extensor digitorum longus (EDL) fibers of wild type (WT) mice that involves: i) remodeling of the SR into stacks of flat cisternae; ii) elongation of the TT into the I band; and iii) increased co-localization of Orai1 with STIM1 in the I band (Boncompagni et al., 2017). This exercise-induced remodeling leads to the formation of new SR-TT contacts that are structurally distinct from triad junctions in that they are oriented longitudinally at the I band (while triads are transversally oriented at the A-I band junction) and display a junctional gap of only 7–8 nm (while triads exhibit a 12 nm junctional gap) (Boncompagni et al., 2017). Importantly, the formation of these junctions following acute exercise correlates with an increased resistance to EDL muscle contractile decline during repetitive stimulation in presence of extracellular Ca²⁺, but not under conditions that reduce SOCE (0 Ca², BTP-2, 2-APB) (Boncompagni et al., 2017). As the formation of the SR-TT junctions within the I band following acute exercise are coincident with enhanced STIM1-Orai1 co-localization and force production during sustained activity, these structural elements were suggested to function as Calcium Entry Units (CEUs).

In spite of these findings, the stability of CEUs formed following exercise, process for disassembly of the components, and impact on Orai1-dependent SOCE, Ca²⁺ dynamics, and muscle force production remain unknown. In fact, the precise degree to which CEUs formed after exercise promote SOCE, optimize Ca²⁺ dynamics and contribute to the adaptation of muscle function in response to acute exercise is unknown. Here, we addressed these issues using a combination of structural (electron microscopy, EM) and functional (Mn²⁺ quench of fura-2 fluorescence, intracellular Ca²⁺ measurements, and muscle contractility) approaches in EDL muscle and flexor digitorum brevis (FDB) fibers from WT and Orai1-deficient mice in the absence of exercise (control) or <1, 6, and 24 hr after acute treadmill exercise. Together, the results support the idea that TT association with SR stacks in the I band promotes Orai1-dependent SOCE needed to replenish releasable Ca²⁺ stores, maintain SR Ca²⁺ release, and maximize contractile force during repetitive stimulation of fast twitch skeletal muscle.

Results

SR-stacks increase for up to 6 hr after exercise

We previously reported that EDL muscle fibers from mice subjected to a single bout of incremental treadmill running exhibit a remodeling of SR membranes into flat, parallel stacks of flat cisternae (SR-stacks) (Boncompagni et al., 2017). To determine their stability, we quantified SR-stacks in EM cross sections of EDL muscles fixed at different time points (<1, 6 and 24 hr) following acute treadmill exercise (Figure 1A–D). Shortly after exercise (<1 hr), the percentage of muscle fibers with SR-stacks in the I band (Figure 1E), the number of stacks/100 µm² (Figure 1F), and SR-stack length (Figure 1G) were all significantly increased compared to that of non-exercised control mice. The percentage of fibers with SR-stacks at the I band and the number of SR stacks/100 µm² further increased 6 hr after exercise. All three parameters decreased to values closer to that of fibers from non-exercised control mice 24 hr after exercise. A similar time course was also observed in FDB muscle fibers (Figure 1—figure supplement 1).

Figure 1. — (**A–D**) Representative EM images of cross sections (in proximity of I band) of EDL muscle fibers from control mice (A) and from mice < 1 hr (B), 6 hr (C), and 24 hr (D) after acute treadmill exercise (empty arrows point to stacks of SR membranes). (E) Percentage of muscle fibers exhibiting SR-stacks. (F) Number of SR stacks per 100 μm² of cross sectional area. (G) SR-stack length. Numbers in bars (n) indicate the number of fibers analyzed. Number of mice used: Control, n = 4;<1 hr after exercise, n = 4; 6 hr after exercise, n = 4; 24 hr after exercise, n = 3; *p<0.05. Data are shown as mean ± SEM.

Figure 1—figure supplement 1. — (**A–D**) Representative EM images of cross sections (in proximity of I band) of EDL muscle fibers from control mice (A) and from mice < 1 hr (B), 6 hr (C), and 24 hr (D) after acute treadmill exercise (empty arrows point to stacks of SR membranes). (E) Percentage of muscle fibers exhibiting SR-stacks. (F) Number of SR stacks per 100 μm² of cross sectional area. (G) SR-stack length. Numbers in bars (n) indicate the number of fibers analyzed. Number of mice used: Control, n = 4;<1 hr after exercise, n = 4; 6 hr after exercise, n = 4; 24 hr after exercise, n = 3; *p<0.05. Data are shown as mean ± SEM.

TT length is increased <1 hr after exercise

Acute exercise also promotes an elongation of the TT from the triad into the I band to form new contacts with stacks of SR cisternae (Boncompagni et al., 2017). Thus, we quantified the length of contacts between TT extensions and flat-parallel stacks of SR within the I band at rest, as well as <1, 6, and 24 hr following acute exercise, in samples post-fixed in the presence of ferrocyanide (Figure 2A–D). The results indicate that the TT/SR contact length within the I band region of the sarcomere increased 5-fold < 1 hr after exercise and then returned to levels not significantly different from control both 6 and 24 hr after exercise (Figure 2E). The increase in TT/SR contact length after exercise is consistent with our prior demonstration (using immunofluorescence and immuno-gold electron microscopy) of increased Orai1-STIM1 co-localization within the I band <1 hr after exercise (Boncompagni et al., 2017).

In addition, total TT length/100 µm² was also significantly increased in EDL muscle fibers from mice < 1 hr after exercise (Figure 2B and F) compared to that of control samples (Figure 2A and F). Importantly, in contrast to that observed for SR-stacks (Figure 1), increased extensions of the TT network (and TT/SR contact length) within the I band, observed <1 hr after exercise, were not different from control 6 and 24 hr after exercise (Figure 2E and F). A transient increase in total TT length/100 µm² and with a more prolonged increase in SR stacks at the I band after exercise was also observed in FDB muscle fibers (Figure 1—figure supplement 1).

Mn²⁺Quench of fura-2 fluorescence is increased <1 hr after exercise

Immunofluorescence and immunogold studies found that exercise-induced formation of CEUs promoted an increase of STIM1 and Orai1 co-localization (Boncompagni et al., 2017). To determine the degree to which these structural changes correlate with STIM1-Orai1 function, the maximum rate of Mn²⁺ quench of fura-2 fluorescence was used as an index of Ca²⁺ entry as described previously (Wei-Lapierre et al., 2013). Mn²⁺ quench experiments were performed in single FDB fibers both following SR Ca²⁺ store depletion (+depletion; 1 hr incubation with thapsigargin plus cyclopiazonic acid) and in the absence of pharmacological store depletion (-depletion). Although fura-2 emission was unaltered upon application of extracellular Mn²⁺ in non-depleted ﬁbers (Figure 4—figure supplement 1), a significant increase in quench was observed in fibers from control mice following store depletion (Figure 3A). Importantly, compared to that of control mice, the maximum rate of Mn²⁺ quench was significantly increased (~70%) in FDB fibers isolated from mice < 1 hr after treadmill exercise (Figure 3B). The maximum rate of Mn²⁺ quench was not different from that of control mice following 6 and 24 hr of recovery. An analysis of the frequency distribution of maximal rate of Mn²⁺ quench in SR depleted fibers showed a shift toward larger rates of quench in fibers from mice < 1 hr after exercise, but not in fibers isolated at 6 and 24 hr of recovery (Figure 3C).

Figure 3. — (A) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺recorded in FDB fibers following store depletion with 1 μM thapsigargin + 15 μM cyclopiazonic acid (+depletion) isolated from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (B) Quantitative analysis of the maximum rate of Mn²⁺ quench following store depletion. (C) Frequency histogram of percentage of fibers exhibiting different levels of maximal rate of Mn²⁺ quench. Histogram data were fit with a single Gaussian distribution. Numbers in bars (n) reflect the number of FDB fibers analyzed. Number of mice used: Control, n = 5;<1 hr after exercise, n = 5; 6 hr after exercise, n = 3; 24 hr after exercise, n = 3; *p<0.05. Data are shown as mean ± SEM.

Since acute exercise increases STIM1 and Orai1 co-localization within the I band (Boncompagni et al., 2017), we also quantified Mn²⁺ quench in FDB fibers in the absence of pharmacological store depletion (-depletion). While fura-2 quench did not occur upon addition of Mn²⁺ to fibers from non-exercised mice, a marked increase in quench rate was observed in fibers from mice < 1 hr after exercise (Figure 4A). Consistent with the morphological observations in TT length (Figure 2), the maximum rate of Mn²⁺ quench in the absence of pharmacological store depletion was significantly increased at <1 hr after exercise, reduced at 6 hr, but essentially undetectable at 24 hr after exercise (Figure 4B). As a result, a shift toward larger Mn²⁺ quench rates was observed in fibers isolated from mice < 1 hr after exercise (Figure 4C). Consistent with the Mn²⁺ quench observed in SR Ca²⁺ replete fibers from mice < 1 hr after exercise being mediated by Orai1 channels, Mn²⁺ quench in these fibers was rapidly blocked by an inhibitor of Orai1-dependent SOCE (10 μM BTP-2; Figure 4—figure supplement 1).

Figure 4. — (A) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺ recorded in FDB fibers in the absence of store depletion (-depletion) isolated from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (B) Quantitative analysis of the maximum rate of Mn²⁺ quench in the absence of store depletion. (C) Frequency histogram of percentage of fibers exhibiting different levels of maximal rate of Mn²⁺ quench. Histogram data were fit with a single Gaussian distribution. Numbers in bars (n) reflect the number of FDB fibers analyzed. Numbers in bars (n) reflect the number of FDB fibers analyzed. Number of mice used: Control, n = 5;<1 hr after exercise, n = 5; 6 hr after exercise, n = 3; 24 hr after exercise, n = 3; *p<0.05. Data are shown as mean ± SEM.

Figure 4—figure supplement 1. — (A) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺ recorded in FDB fibers in the absence of store depletion (-depletion) isolated from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (B) Quantitative analysis of the maximum rate of Mn²⁺ quench in the absence of store depletion. (C) Frequency histogram of percentage of fibers exhibiting different levels of maximal rate of Mn²⁺ quench. Histogram data were fit with a single Gaussian distribution. Numbers in bars (n) reflect the number of FDB fibers analyzed. Numbers in bars (n) reflect the number of FDB fibers analyzed. Number of mice used: Control, n = 5;<1 hr after exercise, n = 5; 6 hr after exercise, n = 3; 24 hr after exercise, n = 3; *p<0.05. Data are shown as mean ± SEM.

Peak Ca²⁺transient amplitude during repetitive stimulation is enhanced <1 hr after exercise

To determine the impact of increased SOCE <1 hr after exercise (Figures 3 and 4) on intracellular Ca²⁺ dynamics, myoplasmic and total Ca²⁺ store content were monitored in single FDB fibers isolated from control mice and in mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise (Figure 5). We first assessed the ability of FDB ﬁbers to maintain myoplasmic Ca²⁺ transient amplitude during repetitive high-frequency stimulation (500 ms, 50 Hz, every 2.5 s), using mag-ﬂuo-4, a rapid, low-afﬁnity Ca²⁺ dye to maximize resolution of Ca²⁺ transient magnitude and kinetics (Capote et al., 2005). Representative traces showed no significant difference between the four groups in peak Ca²⁺ transient amplitude elicited during the 1^st stimulus train (Figure 5A). In FDB ﬁbers from control mice, both absolute (Figure 5B) and relative (Figure 5C) peak Ca²⁺ transient amplitude exhibited a steady modest decline from the ﬁrst to last stimulation train, with an average decrease of ~30% over the 40 stimuli. On the other hand, fibers isolated from mice < 1 hr after exercise displayed a significant increase in both absolute and relative peak Ca²⁺ transient amplitude for all trains between stimuli 7–21 (Figure 5B and C). Finally, peak Ca²⁺ transient amplitude elicited in fibers isolated from mice either 6 or 24 hr after exercise, were not significantly different from control.

Figure 5. — (A) Representative superimposed relative (ΔF/F₀) mag-fluo-4 traces during the 1^st, 10^th, and 40^th stimulation train (500 ms, 50 Hz, every 2.5 s, duty cycle 0.2) in FDB fibers isolated from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (**B and C**) Quantitative analysis of the time course of relative (B) and normalized (C) change in peak mag-fluo-4 fluorescence. (**D and E**) Effect of high-frequency stimulation on total releasable Ca²⁺ store content in fura-FF-loaded fibers (D) and resting myoplasmicCa²⁺ concentration in fura-2-loaded fibers (E) from control mice and mice < 1 hr after exercise in the absence (solid bars) and following (shaded bars) delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 s). n = number of FDB fibers analyzed. Number of mice used in A-C: Control, n = 5;<1 hr after exercise, n = 5; 6 hr after exercise, n = 4; 24 hr after exercise, n = 3. Number of mice used in D: Control, n = 4 (before stimulation) and n = 4 (after stimulation);<1 hr after exercise, n = 3 (before stimulation) and n = 4 (after stimulation). Number of mice used in E: Control, n = 5 (before stimulation) and n = 4 (after stimulation);<1 hr after exercise, n = 5 (before stimulation) and n = 4 (after stimulation); *p<0.05. Data are shown as mean ± SEM.

Figure 5—figure supplement 1. — (A) Representative superimposed relative (ΔF/F₀) mag-fluo-4 traces during the 1^st, 10^th, and 40^th stimulation train (500 ms, 50 Hz, every 2.5 s, duty cycle 0.2) in FDB fibers isolated from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (**B and C**) Quantitative analysis of the time course of relative (B) and normalized (C) change in peak mag-fluo-4 fluorescence. (**D and E**) Effect of high-frequency stimulation on total releasable Ca²⁺ store content in fura-FF-loaded fibers (D) and resting myoplasmicCa²⁺ concentration in fura-2-loaded fibers (E) from control mice and mice < 1 hr after exercise in the absence (solid bars) and following (shaded bars) delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 s). n = number of FDB fibers analyzed. Number of mice used in A-C: Control, n = 5;<1 hr after exercise, n = 5; 6 hr after exercise, n = 4; 24 hr after exercise, n = 3. Number of mice used in D: Control, n = 4 (before stimulation) and n = 4 (after stimulation);<1 hr after exercise, n = 3 (before stimulation) and n = 4 (after stimulation). Number of mice used in E: Control, n = 5 (before stimulation) and n = 4 (after stimulation);<1 hr after exercise, n = 5 (before stimulation) and n = 4 (after stimulation); *p<0.05. Data are shown as mean ± SEM.

Total releasable Ca²⁺ Store Content is reduced <1 hr after Exercise, but Increased Following Repetitive, High-frequency Stimulation

The observed increase in peak Ca²⁺ transient amplitude (Figure 5B) during repetitive, high-frequency stimulation in fibers from mice < 1 hr after exercise could result from increased SOCE during each stimulus train progressively enhancing total Ca²⁺ store content such that the total releasable Ca²⁺ load is increased. To test this idea, we measured total Ca²⁺ store content in FDB fibers under resting conditions and after delivery of 10 consecutive, high-frequency stimulus trains. Total Ca²⁺ store content was assessed in fura-FF-loaded fibers by application of a Ca²⁺ store release cocktail consisting of 10 µM ionomycin, 30 µM cyclopiazonic acid, and 100 µM EGTA in a Ca²⁺-free Ringer’s solution (ICE). The results indicate that compared to fibers from control mice, resting total Ca²⁺ store content was significantly reduced (~25%) in fibers obtained from mice < 1 hr after exercise (Figure 5D), but not in fibers isolated from mice 6 and 24 hr after exercise (Figure 5—figure supplement 1). Consistent with the increase in constitutive SOCE observed <1 hr after exercise (Figure 4B), a significant increase in resting myoplasmic Ca²⁺ concentration was also observed only in fibers < 1 hr after exercise (Figure 5E and Figure 5—figure supplement 1). Importantly, both total releasable Ca²⁺ store content and myoplasmic resting Ca²⁺ concentration were increased significantly after 10 stimulus trains in fibers from mice < 1 hr after exercise compared to that observed for fibers from both control mice and mice < 1 hr after exercise in the absence of stimulation (Figure 5D and E). In contrast, resting free Ca²⁺ concentration in the SR lumen (Figure 5—figure supplement 2) was not altered in FDB fibers from mice < 1 hr after exercise.

In addition to an increase in total releasable Ca²⁺ store content, the increase in Ca²⁺ transient amplitude following repetitive, high-frequency stimulation could also result from a reduction in myoplasmic Ca²⁺ buffering and/or SERCA-mediated Ca²⁺ reuptake. To address these possibilities, we compared the kinetics of electrically-evoked twitch Ca²⁺ transient decay in mag-fluo-4-loaded fibers from sedentary mice and mice < 1 hr after exercise before and after delivery of 10 consecutive stimulus trains (500 ms, 50 Hz, every 2.5 s). Using this approach, the decay phase of the electrically-evoked transient is well-described by a double exponential fit in which the fast component primarily reflects Ca²⁺ binding to fast myoplasmic Ca²⁺ buffers while the slow component is dominated by SERCA-mediated SR Ca²⁺ reuptake (Capote et al., 2005; Carroll et al., 1999; Baylor and Hollingworth, 2003) (Figure 5—figure supplement 3A). No difference in either the amplitudes or time constants of twitch Ca²⁺ transient decay were observed after delivery of 10 consecutive, high-frequency stimulus trains (Figure 5—figure supplement 3B–D) in FDB fibers isolated from either sedentary mice or <1 hr after exercise.

Maintenance of specific force during repetitive stimulation is increased <1 hr after exercise

We previously reported that sustained EDL contractility during repetitive, high frequency stimulation is increased <1 hr after acute exercise (Boncompagni et al., 2017). To determine if the same time course of both enhanced SOCE and sustained Ca²⁺ release correlate with this increment of contractility (Figures 3–5), we compared the ability of EDL muscles to maintain force during repetitive stimulation (Figure 6A). Average force-frequency curves and peak tetanic-specific force were not significantly different across all four conditions (Figure 6—figure supplement 1). All experimental conditions exhibited a similar modest reduction in peak force during the 2^nd stimulus train, followed by a rebound increase (referred to as the bump-phase) that generally lasted ~10–12 stimulus trains (Figure 6B and C). While the peak-specific force generated during the first stimulus train was not different between the four conditions (Figure 6A, left), maximal specific force increased significantly during stimulus trains 7–30 in EDL muscles from mice < 1 hr after exercise (Figure 6A, middle), a potentiation that was lost by the end of the protocol (Figure 6A, right). As a result, both peak-specific force (Figure 6D) and the maximum magnitude of the bump-phase (Figure 6E) were increased 20–25%.

Figure 6. — (A) Representative specific force traces elicited during the 1^st, 10^th, and 40^th stimulus train (500 ms, 50 Hz, every 2.5 s, duty cycle 0.2) in EDL muscles excised from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (**B and C**) Quantitative analysis of the time course of specific (B) and normalized (C) peak force during 40 consecutive, high-frequency stimulus trains. (D) Quantitative analysis of the fold change of peak-specific force calculated at the 10^thstimulus train. (E) Quantitative analysis of the ratio between the maximum peak-specific force produced during the 40 stimulus trains and that of the 2^nd stimulus train. Numbers in bars (n) reflect the number of EDL muscles analyzed. Number of mice used: Control, n = 7;<1 hr after exercise, n = 5; 6 hr after exercise, n = 6; 24 hr after exercise, n = 5; *p<0.05. Data are shown as mean ± SEM.

Figure 6—figure supplement 1. — (A) Representative specific force traces elicited during the 1^st, 10^th, and 40^th stimulus train (500 ms, 50 Hz, every 2.5 s, duty cycle 0.2) in EDL muscles excised from control mice and from mice < 1 hr, 6 hr, and 24 hr after acute treadmill exercise. (**B and C**) Quantitative analysis of the time course of specific (B) and normalized (C) peak force during 40 consecutive, high-frequency stimulus trains. (D) Quantitative analysis of the fold change of peak-specific force calculated at the 10^thstimulus train. (E) Quantitative analysis of the ratio between the maximum peak-specific force produced during the 40 stimulus trains and that of the 2^nd stimulus train. Numbers in bars (n) reflect the number of EDL muscles analyzed. Number of mice used: Control, n = 7;<1 hr after exercise, n = 5; 6 hr after exercise, n = 6; 24 hr after exercise, n = 5; *p<0.05. Data are shown as mean ± SEM.

Orai1 is required for increased SOCE activity and EDL force decay following acute exercise

To address the role of Orai1 channels in the muscle adaptations to exercise reported in Figures 3–6, we determined the effect of exercise on SOCE in single FDB fibers and in intact EDL muscles from inducible, muscle-specific Orai1 knockout (iOrai1 KO) and muscle-specific (Carrell et al., 2016), dominant-negative Orai1 (dnOrai1) transgenic mice (Wei-Lapierre et al., 2013). Consistent with that reported previously (Carrell et al., 2016), SOCE induced by pharmacological SR Ca²⁺ depletion was absent in FDB fibers isolated from iOrai1 KO mice (Figure 7A and B). Similarly, FDB fibers from iOrai1 KO mice < 1 hr after exercise also lacked detectable Mn²⁺ quench following pharmacological store depletion (Figure 7A and B). In addition, constitutive SOCE in the absence of store depletion (-depletion), was also absent in fibers isolated from both control iOrai1 KO mice and iOrai1 KO mice < 1 hr after exercise (Figure 7C and D). Finally, the bump-phase of muscle contraction during repetitive stimulation was suppressed in EDL muscles from control iOrai1 KO mice and was not further enhanced <1 hr after exercise (Figure 7E and F). Similar results were obtained in dnOrai1 mice (Figure 7—figure supplement 1), which also specifically lack SOCE in skeletal muscle (Wei-Lapierre et al., 2013).

Figure 7. — Data were obtained in EDL muscles and FDB fibers from non-exercised iOrai1 KO mice and iOrai1 KO mice < 1 hr after acute treadmill. (A) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺ recorded in FDB fibers following store depletion with 1 μM thapsigargin + 15 μM cyclopiazonic (+depletion). (B) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺ recorded in FDB the absence of store depletion (-depletion). (**C–D**) Quantitative analysis of the maximum rate of Mn²⁺ quench under store-depleted (C) and non-depleted (D) conditions in FDB fibers. (E) Quantitative analysis of the time course of specific force during 40 consecutive, high-frequency stimulus trains in EDL muscles. (F) Quantitative analysis of the ratio between the maximum peak-specific force produced during the 40 stimulus trains and that of the 2^nd stimulus train in EDL muscles. Numbers (n) reflect the number of FDB fibers or muscles analyzed. Number of mice used: iOrai1 KO control, n = 5; iOrai1 KO exercised, n = 5; *p<0.05. Data are shown as mean ± SEM.

Figure 7—figure supplement 1. — Data were obtained in EDL muscles and FDB fibers from non-exercised iOrai1 KO mice and iOrai1 KO mice < 1 hr after acute treadmill. (A) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺ recorded in FDB fibers following store depletion with 1 μM thapsigargin + 15 μM cyclopiazonic (+depletion). (B) Representative traces of fura-2 fluorescence during application of 0.5 mM Mn²⁺ recorded in FDB the absence of store depletion (-depletion). (**C–D**) Quantitative analysis of the maximum rate of Mn²⁺ quench under store-depleted (C) and non-depleted (D) conditions in FDB fibers. (E) Quantitative analysis of the time course of specific force during 40 consecutive, high-frequency stimulus trains in EDL muscles. (F) Quantitative analysis of the ratio between the maximum peak-specific force produced during the 40 stimulus trains and that of the 2^nd stimulus train in EDL muscles. Numbers (n) reflect the number of FDB fibers or muscles analyzed. Number of mice used: iOrai1 KO control, n = 5; iOrai1 KO exercised, n = 5; *p<0.05. Data are shown as mean ± SEM.

Discussion

We previously reported that acute treadmill exercise drives the formation of contacts between remodeled SR stacks and extensions of the TT at the I band that promote co-localization of STIM1 and Orai1 proteins. The assembly of these new junctions (CEUs) correlates with reduced EDL muscle force decline during repetitive stimulation (Boncompagni et al., 2017). Here, we determined the time course of the disassembly of CEUs following exercise, as well as the impact of these junctions on Orai1-dependent Ca²⁺ entry, myoplasmic Ca²⁺ dynamics, and muscle contractile function. The results, summarized in the model shown in Figure 8, demonstrate that: a) CEUs are dynamic structures that assemble during exercise and disassemble following recovery (Figures 1 and 2); b) TT retraction from the I band following exercise occurs before SR-stack disassembly (Figures 1 and 2); c) peak SOCE activity (i.e. maximum rate of Mn²⁺ quench), reduced Ca²⁺ store content and increased resting Ca²⁺, as well as sustained Ca²⁺ transient amplitude and force production during repetitive stimulation all increase/decrease in parallel with the time course of TT extension within the I band (Figures 3, 5 and 6); d) exercise-induced CEU formation results in constitutively-activated SOCE (Figure 4); and e) increased SOCE activity and sustained force production during repetitive stimulation requires Orai1 channels (Figure 7). Together, these results indicate that the stacks of flat-parallel SR cisternae formed after exercise are functionally competent CEUs only when associated with a TT extended from the triad.

Figure 8. — Left) In non-exercised or sedentary muscle, the SR is fully replete with Ca²⁺ ions, and the myoplasmic Ca²⁺ concentration is low. STIM1 proteins are located throughout the I band, while Orai1 proteins are almost exclusively within the TT membrane. Middle) Less than 1 hr after acute exercise, total Ca²⁺ store content is reduced, and myoplasmic Ca²⁺ is modestly elevated. In addition, the free-SR undergoes a remodeling that results in the formation of SR-stacks, while the TT elongates from the triad into the I band to create junctional contacts, referred to as *Calcium Entry Units* (CEUs). Right) 6 hr of recovery after acute exercise, the SR is fully replete with Ca²⁺, the myoplasmic Ca²⁺ concentration returns to control levels, and the TT extension is largely retracted back from contacts with SR-stacks.

Two main findings of this study are: a) TTs retract from the CEU following exercise prior to SR-stack disassembly (Figures 1 and 2) and b) exercise-dependent increases in Orai1-mediated SOCE (Figures 3, 4 and 7), intracellular Ca²⁺ dynamics (Figure 5), and maintained contractile activity during repetitive stimulation (Figure 6) declines with a time course like that of TT retraction. Together, these results show that the functional effects of acute exercise are controlled by association of TTs (containing Orai1 proteins) with SR-stacks (containing STIM1 proteins) in the I band, with an increase of TT/SR contact length that provides a greater surface area for STIM1-Orai1 interaction. The role of SR-SR interfaces that lack a TT element remains unclear. However, SR-stacks containing multiple elements virtually identical to those formed in muscle after exercise, are also observed in non-muscle cells following STIM1 overexpression (Orci et al., 2009; Perni et al., 2015). Interestingly, both ER-ER contacts in non-muscle cells overexpressing STIM1 and SR-SR contacts in muscle after acute exercise exhibit small electron dense bridges within a similar junctional gap width (~8 nm). These similarities in membrane structure and gap width are consistent with the small electron dense bridges observed in EM images reflecting STIM1 proteins. Similar to the ‘diffusion trap’ model of STIM1-Orai1 coupling in non-excitable cells proposed previously (Wu et al., 2014), one possible interpretation of these findings is that STIM1-containing SR-stacks could be needed to efficiently trap Orai1 channels during TT elongation into the I band.

An unexpected finding of this study was the demonstration of significant Mn²⁺ quench in the absence of thapsigargin/cyclopiazonic acid-induced store depletion (constitutive Ca²⁺ entry) in FDB fibers isolated from mice < 1 hr after treadmill exercise (Figure 4). As the constitutive Ca²⁺ entry in these fibers was rapidly blocked by application of BTP-2 (Figure 4—figure supplement 1), the constitutive entry in these fibers is unlikely to represent increased susceptibility of fibers to damage shortly after exercise. In support of this idea, the constitutive Ca²⁺ entry observed <1 hr after exercise is absent in fibers from two separate mouse models deficient in Orai1 function (Figure 7 and Figure 7—figure supplement 1). These findings suggest that the constitutive Ca²⁺ entry is mediated by Orai1 channels (rather than TrpC or mechano-sensitive channels). D1ER measurements of ‘spatially-averaged’ free Ca²⁺ concentration in the SR were not significantly different after exercise (though areas of local depletion are possible). It is important to note, however, that D1ER localization studies indicate that the majority of the D1ER signal originated from the triad junction (Figure 5—figure supplement 2A). Thus, the D1ER measurements suggest that the free Ca²⁺ concentration within the triad junction is not different after exercise. Thus, constitutive Ca²⁺ entry observed after exercise most likely does not originate from Orai1 channels within the triad junction. Rather, constitutive entry may arise from Orai1 channels in another location, possibly CEUs assembled during exercise in the I band region of the sarcomere that are not well-reflected in the spatially-averaged D1ER signal. These observations support the possibility that CEUs represent the site of constitutive Ca²⁺ entry after exercise. An alternative explanation is that exercise could produce a signal that directly activates Orai1 independent of store depletion.

Assuming no changes in the properties of sarcolemmal Ca²⁺ efflux mechanisms, the increase in constitutive Ca²⁺ entry observed <1 hr after exercise could explain the elevation in resting Ca²⁺ concentration observed at this time point (Figure 5E and Figure 5—figure supplement 1). Again, partial Ca²⁺ depletion within SR-stacks of the CEU represents one possible mechanism for the constitutive entry observed immediately following exercise. Alternatively, as lower levels of store depletion are sufficient for Stim2 to activate Orai1-dependent Ca²⁺ entry (Brandman et al., 2007; Thiel et al., 2013), the constitutive Ca²⁺ entry observed after exercise could reflect Stim2-mediated activation of Orai1 channels within the CEUs. Thus, it will be important for future studies to more definitively assess SERCA and Stim2 expression within CEUs, determine the relative distribution of the short and long forms of STIM1 within CEUs (Michelucci et al., 2018), and quantitatively model Ca²⁺ binding and diffusion within and out of these structures.

Another observation that deserves consideration is the non-linear correspondence between exercise-induced changes in structure (TT/SR contact length increases five fold; Figure 2E) and SOCE function (~65% increase in maximal rate of Mn²⁺ quench; Figure 3B) in FDB muscle. One possibility is that, as proposed above, CEUs may reflect the subcellular location for the constitutive SOCE observed <1 hr after exercise. However, given that total releasable Ca²⁺ store content at rest was reduced <1 hr after exercise, it is possible that pre-existing STIM and Orai1 complexes distinct from newly formed CEUs could also contribute to the observed increase in SOCE function.

We investigated the mechanism by which increased SOCE following acute exercise increases peak Ca²⁺ transient amplitude during repetitive, high-frequency stimulation. Conceivably, increased total (Figure 3) and constitutive (Figure 4) SOCE activity <1 hr after exercise could potentiate peak Ca²⁺ transient amplitude either by increasing total releasable SR Ca²⁺ content or elevating myoplasmic Ca²⁺ levels enough to reduce fast buffering. Alternatively, SERCA activity could be inhibited after exercise due to glycogen depletion. Kinetic analysis of twitch Ca²⁺ transient decay before and after 10 stimulus trains in FDB fibers from control and exercised mice did not reveal a significant change in either myoplasmic Ca²⁺ buffering or SERCA-mediated SR Ca²⁺ reuptake (Figure 5—figure supplement 3). On the other hand, both total releasable Ca²⁺ store content (Figure 5D) and resting myoplasmic Ca²⁺ concentration (Figure 5E) were significantly increased after 10 stimulus trains in FDB fibers from mice < 1 hr after exercise. Together, these findings are consistent with the observed increase in peak Ca²⁺ transient amplitude (and force production) after acute exercise resulting, at least in part, from enhanced SOCE activity promoting SR refilling and total releasable Ca²⁺ content to a level sufficient to augment Ca²⁺ flux.

EDL muscles isolated from mice < 1 hr after treadmill exercise exhibit an increased ability to produce sustained force during repetitive stimulation (Figure 6) (Boncompagni et al., 2017). Muscle force production shows a rebound increment in peak-specific force (between stimulus trains 2–12), referred to as the bump-phase of the repetitive stimulation contraction curve. The following observations are consistent with this bump-phase reflecting the contribution of Orai1-dependent SOCE: 1) it is reduced by interventions (BTP-2, 2-APB, 0 Ca²⁺) that reduce SOCE (Boncompagni et al., 2017); 2) it is increased <1 hr after treadmill exercise when SOCE is also increased; 3) it is unaltered 6 and 24 hr after treadmill exercise when SOCE activity has returned to control levels; 4) it is absent in mice lacking Orai1-dependent Ca²⁺ entry both under control conditions and after acute exercise. As some mouse models have been shown to exhibit an increase in basal levels of CEUs similar to those observed immediately following exercise (Boncompagni et al., 2012a; Boncompagni et al., 2012b; Ko et al., 2011), it will be important for future studies to determine if muscles from these mice also exhibit an increase in both SOCE and the bump-phase of the repetitive stimulation contraction curve.

In conclusion, the findings reported here support the idea that CEUs formed after acute exercise represent an adaptive response of skeletal muscle to exercise that augments Orai1-dependent Ca²⁺ entry to optimize myoplasmic Ca²⁺ dynamics in a manner that limits muscle force decline during sustained activation.

Materials and methods

Animals

Male wild type (WT) C57Bl/6 mice (4–6 months old) were housed in microisolator cages at 20°C on a 12 hr light/dark cycle while being provided free access to standard chow and water. Two previously described muscle-specific mouse lines deficient in Orai1 function were also employed: a) dominant-negative Orai1 (dnOrai1) transgenic mice (Wei-Lapierre et al., 2013) and b) tamoxifen-inducible, Orai1 (Orai1^fl/fl::ACTA^Mer-Cre or iOrai1) knockout mice (Carrell et al., 2016). Four-month old male WT, dnOrai1, and iOrai1 KO mice were randomly assigned to two experimental groups: control and exercised (mice exposed to a single bout of treadmill exercise; see below). All animal studies were designed to minimize animal suffering and were approved by the Italian Ministry of Health (992/2019-PR) and the University Committee on Animal Resources at the University of Rochester (UCAR-2006-114E).

Tamoxifen treatment

Two-three-month-old Orai1^fl/fl::ACTA^Mer-Cre mice were provided ad libitum access to tamoxifen-infused mouse chow (Envigo, Huntingdon, UK) for a period of 4 weeks, and then returned to regular chow for at least one additional week before being used for experiments. Semi-quantitative PCR of Orai1 transcript level was used to verify success of tamoxifen treatment in Orai1 ablation in muscle, as previously described (Carrell et al., 2016).

Treadmill exercise protocol

A standardized acute exercise protocol was performed at room temperature using a running treadmill (Columbus Instruments, Columbus, OH, USA) on a flat surface (0° incline) as previously described (Boncompagni et al., 2017). Briefly, immediately after a warm-up period (10 min at 5 m/min), mice were subjected to a 65 min exercise protocol consisting of an initial 25 min at a speed of 10 m/min, followed by 20 min at 15/m/min, 15 min at 20 m/min and then five final 1 min intervals where the speed was increased an additional 1 m/min for each interval. The protocol was stopped when mice either reached the end of the protocol or were unable to continue as indicated by the inability of the animal to maintain running and contact with the treadmill belt (see 10 for details). Following treadmill exercise, EDL muscles (for ex vivo muscle contractility studies) and FDB muscles (for Mn²⁺ quench/Ca²⁺ measurements) were removed and prepared for experiments at one of the following three time points: 1)<1 hr from the end of the in vivo exercise protocol (referred to as <1 hr after exercise), 2) 6 hr after exercise, or 3) 24 hr after exercise.

Electron microscopy (EM)

Intact EDL and FDB muscles were fixed at room temperature with 3.5% or 6% glutaraldehyde in 0.1M sodium cacodilate (NaCaCO) buffer (pH 7.2), and processed for EM acquisition as previously described (Boncompagni et al., 2009). For TT staining, specimens were post-fixed in a mixture of 2% OsO₄ and 0.8% K₃Fe(CN)₆ for 1–2 hr followed by a rinse with 0.1M NaCaCO buffer with 75 mM CaCl₂. Ultrathin sections (~50 nm) were cut using a Leica Ultracut R microtome (Leica Microsystems, Vienna, Austria) with a Diatome diamond knife (Diatome, Biel, Switzerland) and double-stained with uranyl acetate and lead citrate. Sections were viewed in a FP 505 Morgagni Series 268D electron microscope (FEI Company, Brno, Czech Republic), equipped with Megaview III digital camera (Olympus Soft Imaging Solutions, Munster, Germany) and Soft Imaging System at 60 kV (100 kV for TT staining).

Quantitative analyses of EM images

The percentage of fibers exhibiting SR-stacks and the number of SR-stacks per 100 µm² were determined from electron micrographs of non-overlapping regions randomly collected from transversal EM sections, as described previously (Boncompagni et al., 2017). For each specimen, 10–15 representative fibers were analyzed and 5 micrographs at 28000x magnification were taken for each fiber. Operationally, ‘stacks’ were defined as two or more adjacent SR flat elements of stackable cisternae separated by a junctional gap width of ~8 nm (Boncompagni et al., 2017). The average length of SR-stacks was measured in micrographs taken at 44,000x magnification. TT/SR contact length (i.e. length of TT adjacent to SR stack membrane) and extensions of the TT network within the I band of sarcomere (total TT length) were measured in electron micrographs of non-overlapping regions randomly collected from transversal EM sections. For each specimen, 10–15 representative fibers were analyzed and 5 micrographs at 28,000x magnification in transverse sections of samples stained with ferrocyanide were taken for each fiber. Quantitative analyses of SR-stacks, TT/SR contact length and total TT length were obtained using the Analy-SIS software of the EM digital camera (Olympus Soft Imaging Solutions, Munster, Germany).

Isolation of single FDB muscle fibers

As healthy single fibers are readily isolated from FDB muscles, all Mn²⁺ quench and Ca²⁺ measurements were conducted in single, acutely dissociated FDB fibers. FDB muscles were dissected from hind paws and placed in a dish containing Ringer’s solution consisting of (in mM): 145 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, and 10 HEPES, pH 7.4. Muscles were then incubated in Ringer’s solution supplemented with 1 mg/ml collagenase A (Roche Diagnostics, Indianapolis, IN, USA) for 60 min while rocking gently at 37°C to allow enzymatic dissociation. Single FDB fibers obtained by mechanical trituration were plated on glass-bottom dishes and allowed to settle for >20 min before experimentation (see below). Only fibers with clear striations and no signs damage were used for recordings.

Mn²⁺ quench measurements

FDB fibers were loaded with 5 μM fura-2 AM for 1 hr at 37° C in a Ca²⁺-free Ringer's solution containing (in mM): 145 NaCl, 5 KCl, 1 MgCl₂, 0.2 EGTA, pH 7.4.To measure the maximal rate of Mn²⁺ quench, during fura-2 AM loading, fibers were also incubated with two SERCA pump inhibitors (1 μM thapsigargin; 15 μM cyclopiazonic acid) to fully deplete SR Ca²⁺ stores (+depletion) and a skeletal muscle myosin inhibitor (30 μM N-benzyl-p-toluene sulfonamide [BTS]) to prevent movement artifacts (Wei-Lapierre et al., 2013). B) In a second set of studies, FDB fibers were loaded with fura-2 AM and BTS in the absence of SERCA pump inhibitors (-depletion). Both store-depleted and non-depleted FDB fibers were then bathed in Ca²⁺-free Ringer's and excited at 362 nm (isobestic point of fura-2), while emission was detected at 510 nm using a DeltaRam illumination system (Photon Technologies Inc, Birmingham, NJ, USA). After obtaining an initial basal rate of fura-2 decay (R_baseline), fibers were exposed to Ca²⁺-free Ringer's supplemented with 0.5 mM MnCl₂. The maximum rate of fura-2 quench in the presence of Mn²⁺ (R_max) was then obtained from the peak differential of the fura-2 emission trace during Mn²⁺ application. The maximum rate of SOCE (R_SOCE) was calculated as R_SOCE = R_max R_baselineand expressed as dF/dt in counts/sec (Wei-Lapierre et al., 2013).

Ca²⁺transient measurements

Myoplasmic Ca²⁺ transients were monitored in acutely isolated FDB fibers as described previously (Ainbinder et al., 2015). Briefly, FDB fibers were loaded with 4 µM mag-fluo-4-AM for 20 min at room temperature followed by washout in dye-free solution supplemented with 25 µM BTS for 20 min. While continuously perfused with a control Ringer's solution supplemented with 25 μM BTS, fibers were electrically stimulated with a repetitive stimulation protocol (40 consecutive, 500 ms duration, 50 Hz stimulus trains delivered every 2.5 s) using an extracellular electrode placed adjacent to the fiber. Mag-fluo-4 was excited at 480 ± 15 nm using an Excite epifluorescence illumination system (Nikon Instruments, Melville, NY, USA) and fluorescence emission at 535 ± 30 nm was monitored with a 40X oil objective and a photomultiplier detection system (Photon Technologies Inc, Birmingham, NJ, USA). Relative changes in mag-fluo-4 fluorescence from baseline (F/F₀) were recorded at 10 kHz and analyzed using Clampfit 10.0 (Molecular Devices, Sunnyvale, CA, USA).

The decay kinetics of single electrically-evoked twitch Ca²⁺ transients were determined in mag-fluo-4-loaded FDB fibers isolated from both sedentary mice and mice < 1 hr after exercise. For these experiments, 10 electrically-evoked twitch transients (0.5 Hz) were collected both before and after delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 s). The decay phase of each transient was fitted according to the following equation and used to generate average values for both before and after high-frequency stimulation:

F(t)=A_fast x [exp(-t/τ_fast)] + A_slow x [exp(-t/τ_slow)]

where F(t) is the fluorescence at time t, A_fast and τ_fast are the amplitude and time constants of the fast component, and A_slow and τ_slow are the amplitude and time constants of the slow component.

Total Ca²⁺store content measurements

Total releasable Ca²⁺ store content was determined in FDB fibers loaded with 4 µM fura-FF AM, a low affinity ratiometric Ca²⁺ dye, for 30 min at room temperature in control Ringer’s solution followed by 30 min washout in dye-free Ringer’s solution supplemented with 40 µM BTS as described previously (Kimura et al., 2009; Loy et al., 2011). Briefly, fura-FF-loaded fibers were perfused in Ca²⁺-free Ringer’s solution while alternately excited at 340 and 380 nm (510 nm emission) every 250 ms (30 ms exposure per wavelength and 2 × 2 binning) using a monochromator-based illumination system (TILL Photonics, Graefelfing, Germany). Fura-FF emission at 535 ± 30 nm was captured using a high speed, digital QE CCD camera (TILL Photonics, Graefelfing, Germany). Total releasable Ca²⁺ store content was assessed from the difference between basal and peak fura-FF ratio (ΔRatio_340/380) upon application of a Ca²⁺ release cocktail containing 10 µM ionomycin, 30 µM CPA, and 100 µM EGTA in a Ca²⁺-free Ringer’s solution (ICE). For a subset of experiments, total releasable Ca²⁺ store content was measured in fura-FF-loaded fibers isolated from control mice and mice < 1 hr after treadmill exercise following delivery of 10 consecutive high-frequency stimulus trains (500 ms, 50 Hz, every 2.5 s). To confirm that the peak fura-FF signal during ICE application was not saturated, maximal fura-FF responsiveness was assessed at the end of each experiment by subsequent application of Ca²⁺-containing Ringer’s solution. Analysis of peak ICE-induced change in fura-FF ratio (ΔRatio_340/380) was calculated using Clampfit 10.0 (Molecular Devices, Sunnyvale, CA, USA).

Resting Ca²⁺measurements

Resting myoplasmic Ca²⁺ concentration was determined in FDB fibers loaded with 5 µM fura-2 AM for 30 min at room temperature in control Ringer’s solution followed by a 30 min washout with dye-free Ringer’s solution. Fura-2-loaded fibers were placed on the stage of an inverted epifluorescence microscope (Nikon Instruments) and alternatively excited at 340 and 380 nm (30 ms exposure per wavelength and 2 × 2 binning) using a monochromator-based illumination system with fluorescence emission at 510 nm was captured using a high speed, digital QE CCD camera (TILL Photonics, Graefelfing, Germany). Fura-2 340/380 ratios from myoplasmic areas of interest were calculated using TILL visION software (TILL Photonics, Graefelfing, Germany), analyzed offline using NIH ImageJ, and converted to resting free Ca²⁺ concentrations using a calibration curve for fura-2 generated as described previously (Lanner et al., 2012).

SR-free Ca²⁺measurements

The free Ca²⁺ concentration in the SR lumen was assessed in FDB fibers expressing the D1ER cameleon Ca²⁺ sensor. D1ER was expressed in FDB fibers using an in vivo electroporation approach described previously (Canato et al., 2010). Briefly hind limb footpads of anesthetized mice were first injected with bovine hyaluronidase (6 μl, 0.4 U/ μl) and then one hour later with D1ER cDNA (20 μg in 71 mM NaCl) using 30-gauge needles. The footpad was then electroporated with 20 stimulations of 100 V/cm, 20 ms duration delivered at 1 Hz using subcutaneous electrodes placed perpendicular to the long axis of the muscle, close to the proximal and distal tendons. Fibers were isolated as described above and used for experiments either 7–14 days later.

D1ER-expressing fibers from control mice and mice < 1 hr after exercise were bathed in control Ringer’s solution supplemented with 25 µM BTS on the stage of an inverted Nikon Eclipse TE-2000-U microscope (Nikon Instruments, Melville, NY, USA). A relatively large rectangular region of interest of the fiber was excited at 436 nm (10 nm bandwidth). Fluorescence emission within the region of interest was split with a 515 nm dichroic mirror and collected at 480 nm (30 nm bandwidth) and 535 (30 nm bandwidth) using a photomultiplier counting system (Chroma Technology, Bellows Falls, VT). In order to directly compare measurements across multiple fibers, all experiments used identical excitation and emission gains. Following subtraction of background fluorescence, YFP and CFP emission intensities were collected and digitized at 100 Hz first in the presence of control calcium-free Ringer’s solution supplemented with 40 µM BTS and then during perfusion of ICE supplemented with 40 µM BTS. The ratio of YFP to CFP emission (R = F_YFP/F_CFP) for each fiber was calculated offline with averages calculated both immediately before application of ICE and then again after reaching a minimum value in the presence of ICE.

Immunofluorescence

FDB fibers dissociated from control mice and mice < 1 hr after exercise were plated on glass bottom dishes and fixed in 4% paraformaldehyde for 20 min at room temperature. Following blocking for 1 hr at room temperature in 10% BSA in 1X PBS-T (PBS + 0.1% Triton X-100), fibers were co-labeled overnight at 4°C using primary antibodies for GFP (Rabbit polyclonal antibody, Thermo Fisher Scientific, Cat. # A11122, 1:3000) and either α-actinin (Mouse anti α-actinin, Sigma, Cat # A7811, 1:750) or the type1 ryanodine receptor (RYR1- Mouse anti RYR1 Cat # 34C, DHSB University of Iowa, 1:30). All primary antibody incubations were in 2% BSA in 1X PBS-T. Following three 10 min washes in PBS-T, samples were incubated with a 1:500 dilution of Goat anti Rabbit Alexa fluor 488 (Molecular Probes Cat # A11034) and Goat Anti Mouse Alexa fluor 594 (Molecular Probes Cat # R37121) for 1 hr at room temperature. Following three 10 min washes in 1X PBS-T, samples were imaged using an Olympus FV1000 laser scanning confocal microscope (Olympus Scientific Solutions, Wlatham, MA) and a 100X, UPlanSAPO NA 1.4 oil immersion objective. Alexa 488 and 594 were sequentially excited at 488 and 559 nm and detected at 515 and 617 nm respectively.

Ex vivo muscle contractility measurements

Ex vivo assessment of muscle force production was made in intact EDL muscles excised from control mice and from mice < 1 hr, 6 hr or 24 hr after acute treadmill exercise. Mice were anesthetized by intra-peritoneal injection of anesthetic (Wei-Lapierre et al., 2013). EDL muscles were isolated, tied using 4–0 surgical suture, excised, and then attached to a servo motor and force transducer (Aurora Scientific, Aurora, Ontario, Canada) between two platinum electrode plates in a chamber continuously perfused with oxygenated Ringer solution containing (in mM): 137 NaCl, 5 KCl, 1.2 NaH₂PO₄, 1 MgSO₄, 2 CaCl₂,10 glucose, and 24 NaHCO₃. Before starting measurements, optimal stimulation level and muscle length (L₀) were determined using a series of 1 Hz-twitch stimulation trains while stretching the muscle to a length that generated maximal force (F₀). After establishing L₀, muscles were first equilibrated using three tetani (500 ms, 150 Hz) given at 1 min intervals and then subjected to a standard force frequency protocol (from 1 to 250 Hz). After 3 min of rest, a single sustained high frequency tetanus (150 Hz, 2 s). After an additional 5 min of rest, muscles were then subjected to a repetitive stimulation protocol (40 consecutive, 500 ms duration, 50 Hz stimulus trains delivered every 2.5 s). All muscle contractility experiments were carried out at 30°C. Muscle force was recorded using Dynamic Muscle Control software and analyzed using a combination of both Dynamic Muscle Analysis (Aurora Scientific, Aurora, Ontario, Canada) and Clampfit 10.0 (Molecular Devices, Sunnyvale, CA, USA) software. Specific force was calculated by normalizing the absolute force to the physiological cross sectional area as previously described (Boncompagni et al., 2017).

Statistical analyses

Statistical significance was determined using either Student’s t-test for comparing two groups (Figure 7 and S3) and a one-way ANOVA followed by post-hoc Tukey test when making multiple comparisons (all other figures). Amplitude histograms of maximal rates of Mn²⁺ quench were fit according to a single Gaussian distribution. In all cases, differences were considered as statistically significant at *p<0.05. All summary data were presented as mean ± SEM.

Acknowledgements

This work was supported by grants from the National Institutes of Health (AR059646 to RTD and subcontract to FP), the Italian MIUR (PRIN #2015ZZR4W3 to FP), and the Alfred and Eleanor Wedd Fund (to AM).

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

Robert T Dirksen, Email: robert_dirksen@urmc.rochester.edu.

Richard S Lewis, Stanford University School of Medicine, United States.

Ronald L Calabrese, Emory University, United States.

Funding Information

This paper was supported by the following grants:

National Institutes of Health AR059646 to Robert T Dirksen, Feliciano Protasi.
Ministry of Education, University and Research PRIN #2015ZZR4W3 to Feliciano Protasi.
Alfred and Eleanor Wedd Fund, University of Rochester to Antonio Michelucci.

Additional information

Competing interests

No competing interests declared.

Author contributions

Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing.

Data curation, Formal analysis, Validation, Investigation, Visualization.

Conceptualization, Data curation, Formal analysis, Investigation, Methodology.

Conceptualization, Data curation, Formal analysis, Methodology.

Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing.

Conceptualization, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing.

Ethics

Animal experimentation: All animal studies were designed to minimize animal suffering and were approved by the Italian Ministry of Health (992/2019-PR) and the University Committee on Animal Resources at the University of Rochester (UCAR-2006-114E).

Additional files

Transparent reporting form

elife-47576-transrepform.docx^{(247KB, docx)}

DOI: 10.7554/eLife.47576.017

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files.

References

Ainbinder A, Boncompagni S, Protasi F, Dirksen RT. Role of Mitofusin-2 in mitochondrial localization and calcium uptake in skeletal muscle. Cell Calcium. 2015;57:14–24. doi: 10.1016/j.ceca.2014.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
Baylor SM, Hollingworth S. Sarcoplasmic reticulum calcium release compared in slow-twitch and fast-twitch fibres of mouse muscle. The Journal of Physiology. 2003;551:125–138. doi: 10.1113/jphysiol.2003.041608. [DOI] [PMC free article] [PubMed] [Google Scholar]
Boncompagni S, Rossi AE, Micaroni M, Beznoussenko GV, Polishchuk RS, Dirksen RT, Protasi F. Mitochondria are linked to calcium stores in striated muscle by developmentally regulated tethering structures. Molecular Biology of the Cell. 2009;20:1058–1067. doi: 10.1091/mbc.e08-07-0783. [DOI] [PMC free article] [PubMed] [Google Scholar]
Boncompagni S, Thomas M, Lopez JR, Allen PD, Yuan Q, Kranias EG, Franzini-Armstrong C, Perez CF. Triadin/Junctin double null mouse reveals a differential role for triadin and junctin in anchoring CASQ to the jSR and regulating ca(2+) homeostasis. PLOS ONE. 2012a;7:e39962. doi: 10.1371/journal.pone.0039962. [DOI] [PMC free article] [PubMed] [Google Scholar]
Boncompagni S, Protasi F, Franzini-Armstrong C. Sequential stages in the age-dependent gradual formation and accumulation of tubular aggregates in fast twitch muscle fibers: serca and calsequestrin involvement. Age. 2012b;34:27–41. doi: 10.1007/s11357-011-9211-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
Boncompagni S, Michelucci A, Pietrangelo L, Dirksen RT, Protasi F. Exercise-dependent formation of new junctions that promote STIM1-Orai1 assembly in skeletal muscle. Scientific Reports. 2017;7:14286. doi: 10.1038/s41598-017-14134-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Brandman O, Liou J, Park WS, Meyer T. STIM2 is a feedback regulator that stabilizes basal cytosolic and endoplasmic reticulum Ca2+ levels. Cell. 2007;131:1327–1339. doi: 10.1016/j.cell.2007.11.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
Canato M, Scorzeto M, Giacomello M, Protasi F, Reggiani C, Stienen GJ. Massive alterations of sarcoplasmic reticulum free calcium in skeletal muscle fibers lacking calsequestrin revealed by a genetically encoded probe. PNAS. 2010;107:22326–22331. doi: 10.1073/pnas.1009168108. [DOI] [PMC free article] [PubMed] [Google Scholar]
Capote J, Bolaños P, Schuhmeier RP, Melzer W, Caputo C. Calcium transients in developing mouse skeletal muscle fibres. The Journal of Physiology. 2005;564:451–464. doi: 10.1113/jphysiol.2004.081034. [DOI] [PMC free article] [PubMed] [Google Scholar]
Carrell EM, Coppola AR, McBride HJ, Dirksen RT. Orai1 enhances muscle endurance by promoting fatigue-resistant type I fiber content but not through acute store-operated Ca2+ entry. The FASEB Journal. 2016;30:4109–4119. doi: 10.1096/fj.201600621R. [DOI] [PMC free article] [PubMed] [Google Scholar]
Carroll S, Nicotera P, Pette D. Calcium transients in single fibers of low-frequency stimulated fast-twitch muscle of rat. American Journal of Physiology-Cell Physiology. 1999;277:C1122–C1129. doi: 10.1152/ajpcell.1999.277.6.C1122. [DOI] [PubMed] [Google Scholar]
Darbellay B, Arnaudeau S, König S, Jousset H, Bader C, Demaurex N, Bernheim L. STIM1- and Orai1-dependent store-operated calcium entry regulates human myoblast differentiation. Journal of Biological Chemistry. 2009;284:5370–5380. doi: 10.1074/jbc.M806726200. [DOI] [PubMed] [Google Scholar]
Edwards JN, Blackmore DG, Gilbert DF, Murphy RM, Launikonis BS. Store-operated calcium entry remains fully functional in aged mouse skeletal muscle despite a decline in STIM1 protein expression. Aging Cell. 2011;10:675–685. doi: 10.1111/j.1474-9726.2011.00706.x. [DOI] [PubMed] [Google Scholar]
Feske S, Gwack Y, Prakriya M, Srikanth S, Puppel SH, Tanasa B, Hogan PG, Lewis RS, Daly M, Rao A. A mutation in Orai1 causes immune deficiency by abrogating CRAC channel function. Nature. 2006;441:179–185. doi: 10.1038/nature04702. [DOI] [PubMed] [Google Scholar]
Kimura T, Lueck JD, Harvey PJ, Pace SM, Ikemoto N, Casarotto MG, Dirksen RT, Dulhunty AF. Alternative splicing of RyR1 alters the efficacy of skeletal EC coupling. Cell Calcium. 2009;45:264–274. doi: 10.1016/j.ceca.2008.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ko JK, Choi KH, Zhao X, Komazaki S, Pan Z, Weisleder N, Ma J. A versatile single-plasmid system for tissue-specific and inducible control ofgene expression in transgenic mice. FASEB Journal : Official Publication of the Federation of American Societies for Experimental Biology. 2011;25:2638–2649. doi: 10.1096/fj.11-181560. [DOI] [PMC free article] [PubMed] [Google Scholar]
Koenig X, Choi RH, Launikonis BS. Store-operated Ca2+ entry is activated by every action potential in skeletal muscle. Communications Biology. 2018;1:31. doi: 10.1038/s42003-018-0033-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kurebayashi N, Ogawa Y. Depletion of Ca2+ in the sarcoplasmic reticulum stimulates Ca2+ entry into mouse skeletal muscle fibres. The Journal of Physiology. 2001;533:185–199. doi: 10.1111/j.1469-7793.2001.0185b.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lanner JT, Georgiou DK, Dagnino-Acosta A, Ainbinder A, Cheng Q, Joshi AD, Chen Z, Yarotskyy V, Oakes JM, Lee CS, Monroe TO, Santillan A, Dong K, Goodyear L, Ismailov II, Rodney GG, Dirksen RT, Hamilton SL. AICAR prevents heat-induced sudden death in RyR1 mutant mice independent of AMPK activation. Nature Medicine. 2012;18:244–251. doi: 10.1038/nm.2598. [DOI] [PMC free article] [PubMed] [Google Scholar]
Launikonis BS, Ríos E. Store-operated Ca2+ entry during intracellular Ca2+ release in mammalian skeletal muscle. The Journal of Physiology. 2007;583:81–97. doi: 10.1113/jphysiol.2007.135046. [DOI] [PMC free article] [PubMed] [Google Scholar]
Loy RE, Orynbayev M, Xu L, Andronache Z, Apostol S, Zvaritch E, MacLennan DH, Meissner G, Melzer W, Dirksen RT. Muscle weakness in Ryr1I4895T/WT knock-in mice as a result of reduced ryanodine receptor Ca2+ ion permeation and release from the sarcoplasmic reticulum. The Journal of General Physiology. 2011;137:43–57. doi: 10.1085/jgp.201010523. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lyfenko AD, Dirksen RT. Differential dependence of store-operated and excitation-coupled Ca2+ entry in skeletal muscle on STIM1 and Orai1. The Journal of Physiology. 2008;586:4815–4824. doi: 10.1113/jphysiol.2008.160481. [DOI] [PMC free article] [PubMed] [Google Scholar]
Michelucci A, García-Castañeda M, Boncompagni S, Dirksen RT. Role of STIM1/ORAI1-mediated store-operated Ca2+ entry in skeletal muscle physiology and disease. Cell Calcium. 2018;76:101–115. doi: 10.1016/j.ceca.2018.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
Orci L, Ravazzola M, Le Coadic M, Shen WW, Demaurex N, Cosson P. From the cover: stim1-induced precortical and cortical subdomains of the endoplasmic reticulum. PNAS. 2009;106:19358–19362. doi: 10.1073/pnas.0911280106. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pan Z, Yang D, Nagaraj RY, Nosek TA, Nishi M, Takeshima H, Cheng H, Ma J. Dysfunction of store-operated calcium channel in muscle cells lacking mg29. Nature Cell Biology. 2002;4:379–383. doi: 10.1038/ncb788. [DOI] [PubMed] [Google Scholar]
Perni S, Dynes JL, Yeromin AV, Cahalan MD, Franzini-Armstrong C. Nanoscale patterning of STIM1 and Orai1 during store-operated Ca2+ entry. PNAS. 2015;112:E5533–E5542. doi: 10.1073/pnas.1515606112. [DOI] [PMC free article] [PubMed] [Google Scholar]
Roos J, DiGregorio PJ, Yeromin AV, Ohlsen K, Lioudyno M, Zhang S, Safrina O, Kozak JA, Wagner SL, Cahalan MD, Veliçelebi G, Stauderman KA. STIM1, an essential and conserved component of store-operated Ca2+ channel function. The Journal of Cell Biology. 2005;169:435–445. doi: 10.1083/jcb.200502019. [DOI] [PMC free article] [PubMed] [Google Scholar]
Stiber J, Hawkins A, Zhang ZS, Wang S, Burch J, Graham V, Ward CC, Seth M, Finch E, Malouf N, Williams RS, Eu JP, Rosenberg P. STIM1 signalling controls store-operated calcium entry required for development and contractile function in skeletal muscle. Nature Cell Biology. 2008;10:688–697. doi: 10.1038/ncb1731. [DOI] [PMC free article] [PubMed] [Google Scholar]
Thiel M, Lis A, Penner R. STIM2 drives Ca2+ oscillations through store-operated Ca2+ entry caused by mild store depletion. The Journal of Physiology. 2013;591:1433–1445. doi: 10.1113/jphysiol.2012.245399. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wei-Lapierre L, Carrell EM, Boncompagni S, Protasi F, Dirksen RT. Orai1-dependent calcium entry promotes skeletal muscle growth and limits fatigue. Nature Communications. 2013;4:2805. doi: 10.1038/ncomms3805. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wu MM, Covington ED, Lewis RS. Single-molecule analysis of diffusion and trapping of STIM1 and Orai1 at endoplasmic reticulum-plasma membrane junctions. Molecular Biology of the Cell. 2014;25:3672–3685. doi: 10.1091/mbc.E14-06-1107. [DOI] [PMC free article] [PubMed] [Google Scholar]

eLife. doi: 10.7554/eLife.47576.019

Decision letter

Editor: Richard S Lewis¹

Reviewed by: Eduardo Rios²

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

Following exercise, muscle fibers undergo membrane remodeling that results in increased Ca²⁺ release and muscle twitch strength. Previous work has shown that after exercise the T tubule (TT) invaginations of the plasma membrane elongate and make close appositions to newly formed stacks of sarcoplasmic reticulum (SR), and that the presence of the STIM1 Ca²⁺ sensor and Orai1 calcium channels at these newly formed junctions may increase the loading of calcium into the SR to promote contractile force. Here the authors provide new evidence for this mechanism, showing that after exercise the appearance and disappearance of new SR stack-TT structures parallels the time course of augmented calcium entry through Orai, SR calcium content and release amplitude, and contractile force. The results thus support a functional link between increased muscle performance following exercise and the operation of store-operated Orai channels at newly formed SR-TT contacts. This work has important implications for understanding muscle function and morphological adaptations to exercise as well as the newly emerging roles of store-operated calcium channels in excitable cells.

Decision letter after peer review:

[Editors’ note: the authors were asked to provide a plan for revisions before the editors issued a final decision. What follows is the editors’ letter requesting such plan.]

Thank you for sending your article entitled "Transverse tubule remodeling enhances Orai1-dependent Ca²⁺ entry in skeletal muscle" for peer review at eLife. Your article is being evaluated by three peer reviewers, and the evaluation is being overseen by a Reviewing Editor and Ronald Calabrese as the Senior Editor.

Given the list of essential revisions, including new experiments, the editors and reviewers invite you to respond with an action plan and timetable for the completion of the additional work. We plan to share your responses with the reviewers and then issue a binding recommendation.

All three reviewers thought the manuscript was an important extension of your earlier work on muscle calcium entry units (CEUs) that form after exercise. The time course of changes in SR-TT contacts, SOCE, Ca²⁺ transients, and force production following exercise provides some additional support for the idea that CEUs are involved in exercise-dependent effects on muscle contractility. However, to represent a significant advance beyond the previous work the reviewers all agreed that more quantitative experimental data are needed to establish a mechanistic link between increased SOCE and altered muscle function.

The critical question is how increased SOCE leads to the enhancement of peak Ca²⁺ transients and force. Increased SOCE cannot directly account for this because the flux is orders of magnitude smaller than the SR release flux. However, several non-exclusive mechanisms could contribute:

1) An increase in SR Ca²⁺ content leading to increased Ca²⁺ release flux might be responsible but is inconsistent with the 25% reduction in content that was reported. The method used to measure total SR content (Ca²⁺ release cocktail containing ionomycin) is not sufficiently quantitative and may also release Ca²⁺ from other compartments (e.g., mitochondria). To address this concern, a ratiometric Ca²⁺ indicator should be targeted to the SR to directly measure the store content during stimulation, and the Ca²⁺ release flux can be estimated from cytosolic transients assuming generally accepted cytosolic buffering parameters.

2) SOCE may elevate cytosolic Ca²⁺ enough to reduce fast buffering and thereby potentiate the effect of Ca²⁺ release. This could be assessed by a more detailed measurement of cytosolic Ca²⁺ between transients and estimating the effect of this increased Ca²⁺ on the buffering power.

3) Finally, SERCA may be inhibited after exercise due to glycogen depletion. Reduction of SERCA activity could enhance the transient amplitude, as well as contribute to the partial depletion of the SR and tonic SOCE that was observed. SERCA activity should be measured from the Ca²⁺ transient decay rates at different times following exercise. Importantly, such measurements may reveal a time-dependent change in Ca²⁺ handling after exercise that is not directly related to the assembly and disassembly of CEUs.

[Editors’ note: formal revisions were requested, following approval of the authors’ plan of action.]

Thank you for submitting your article "Transverse tubule remodeling enhances Orai1-dependent Ca²⁺ entry in skeletal muscle" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Ronald Calabrese as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Eduardo Rios (Reviewer #2).

After the reviewers discussed the reviews with one another, the Reviewing Editor sent the list of essential revisions to you to verify that you could complete these within about two months. Your revision plan and timeline meet the requirements, so please proceed with the revisions and additional experiments. The complete review is printed below. Please note that additional comments have been added based on the revision plan you sent.

Summary:

Michelucci et al. present evidence that acute exercise promotes the association of sarcoplasmic reticulum (SR) stacks with extended transverse tubules (TT) in skeletal muscle, and that these contacts enhance store-operated calcium entry (SOCE) to optimize Ca²⁺ signaling and muscle contraction. This paper extends upon previous work by the authors describing the formation of SR stacks, TT extensions and increased STIM1-Orai1 colocalization (called calcium entry units, or CEUs) in the I band following exercise. The authors previously proposed that the CEUs promote resistance to fatigue during repetitive stimulation. Here the authors provide new functional evidence for this mechanism by showing that the time course of development and reversal of SR stack-TT structures parallels the timing of changes in SOCE, SR Ca²⁺ content, SR Ca²⁺ release transients, and contractile force following exercise. Using knockout mice, the authors also show that Orai1 is required for these changes, presumably through interaction with STIM1 and/or STIM2.

Essential revisions:

All three reviewers thought the manuscript is an important extension of your earlier work on muscle calcium entry units (CEUs) formed after exercise. The time course of changes in SR-TT contacts, SOCE, Ca²⁺ transients, and force production following exercise provides indirect support for the idea that CEUs are involved in exercise-dependent effects on muscle contractility. However, there was also strong agreement that more quantitative experimental data are needed to show how the observed changes in SOCE lead to altered muscle function during exercise. This could occur by several non-exclusive mechanisms, which need to be tested:

1) An increase in SR Ca²⁺ content leading to increased Ca²⁺ release flux might be responsible. The free Ca²⁺ in the SR needs to be measured during stimulus trains to determine if it is elevated. The Ca²⁺ release cocktail method will indicate total stored Ca²⁺, not free Ca²⁺. Instead, a ratiometric Ca²⁺ indicator such as D1ER should be targeted to the SR to directly measure the free luminal Ca²⁺ during stimulation.

2) SOCE may elevate cytosolic Ca²⁺ enough to reduce fast buffering and thereby potentiate the effect of Ca²⁺ release. The effect on buffering could be assessed by a more detailed measurement of cytosolic Ca²⁺ between individual transients and after stimulus trains in control and exercised mice.

3) SERCA may be inhibited after exercise due to glycogen depletion. Reduction of SERCA activity could enhance the transient amplitude, as well as contribute to the partial depletion of the SR and tonic SOCE that was observed in resting muscle after exercise. Kinetic analysis of cytosolic Ca²⁺ twitch transients (fast and slow components) can be used to estimate changes in SERCA activity and Ca²⁺ buffering from control and exercised mice before and after 10 stimulus trains. Importantly, such measurements may reveal a time-dependent change in Ca²⁺ handling after exercise that is not directly related to the assembly and disassembly of CEUs.

4) The idea that CEUs are responsible for the constitutive SOCE after exercise is not well supported. In the Discussion section, the authors suggest that the new CEUs can account for the increased tonic SOCE after exercise. The supporting argument is that "non-depletion" quench after exercise (1200 cps) + TG/CPA quench before exercise (1700 cps) = TG/CPA quench rate after exercise (2800 cps). However, the quench after exercise is driven by only partial SR depletion, so TG/CPA should add more than 1200. This argument does not make quantitative sense to me. Unless it can be better substantiated, it should be deleted.

5) A related issue concerns the idea that CEUs are solely responsible for the SOCE that augments muscle function after exercise. The loss of SR Ca²⁺ content peaks at the same time (<1 hour) as "non-depletion" SOCE, Ca²⁺ release transients and force production during trains. This complicates things. Because of tonic depletion, it is possible that pre-existing STIM and Orai complexes (activated by tonic store depletion) distinct from the new CEUs are being activated and contribute to augmented function. This possibility needs to be explicitly acknowledged, and the conclusions regarding CEU functions toned down.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Transverse tubule remodeling enhances Orai1-dependent Ca²⁺ entry in skeletal muscle" for further consideration at eLife. Your revised article has been favorably evaluated by Ronald Calabrese (Senior Editor) and a Reviewing Editor.

The manuscript has been improved, and the new data have addressed several of the issues that were raised in the original review. However, there are some remaining issues that need to be addressed before acceptance, as outlined below:

1) (Subsection “Total Releasable Ca²⁺ Store Content is Reduced <1 Hour After Exercise, but Increased Following Repetitive, High-frequency Stimulation” and Discussion section). After exercise, the constitutive SOCE is greatly increased, but the new D1ER measurements show that resting free Ca²⁺ in the SR is not altered. Because STIM1 and SOCE are controlled by free Ca²⁺ in ER/SR, and not the total Ca²⁺ content, how can Orai activity be increased without a reduction of free SR Ca²⁺? The decrease in total SR Ca²⁺ that was observed cannot explain the appearance of constitutive SOCE. This conundrum needs to be clearly stated and possible explanations provided. While uncertainty in the D1ER measurement might mask a small depletion of free Ca²⁺, this is unlikely to be the whole explanation since the SOCE amplitude was so large (i.e., about the same as maximal SOCE from full SR depletion before exercise).

2) The new twitch analyses are convincing, but it is not clear to anyone but an expert muscle physiologist how these were used to draw conclusions about Ca²⁺ binding and SERCA activity. In Subsection “Total Releasable Ca²⁺ Store Content is Reduced <1 Hour After Exercise, but Increased Following Repetitive, High-frequency Stimulation”, the statement that kinetics of twitch Ca²⁺ transient decay do not change needs to be explained; why was this experiment done, and what is the interpretation? In the Discussion section, it would help to discuss the possibilities that were brought up in review, including saturation of buffers and depletion of glycogen and effects on SERCA, and how the new results argue against these alternative explanations. An example of a transient should be shown along with the analysis in Figure 5—figure supplement 3 to illustrate the fast and slow components and how they were fitted.

eLife. 2019 Oct 28;8:e47576. doi: 10.7554/eLife.47576.020

Author response

[Editors’ note: what follows is the authors’ plan to address the revisions.]

All three reviewers thought the manuscript was an important extension of your earlier work on muscle calcium entry units (CEUs) that form after exercise. The time course of changes in SR-TT contacts, SOCE, Ca²⁺ transients, and force production following exercise provides some additional support for the idea that CEUs are involved in exercise-dependent effects on muscle contractility. However, to represent a significant advance beyond the previous work the reviewers all agreed that more quantitative experimental data are needed to establish a mechanistic link between increased SOCE and altered muscle function.

The critical question is how increased SOCE leads to the enhancement of peak Ca²⁺ transients and force. Increased SOCE cannot directly account for this because the flux is orders of magnitude smaller than the SR release flux.

We thank the Reviewing Editor and three reviewers for their acknowledgment of the importance of the current work in supporting the conclusion that calcium entry units (CEUs) are involved in exercise-dependent effects on muscle contractility. We also appreciate the question raised regarding how increased SOCE after exercise enhances peak Ca²⁺transients and force during repetitive stimulation. Before describing our plan to address this issue, we would like to reiterate the backstory, motivation and objectives of the current study. Our initial report describing exercise-induced formation of CEUs in muscle (Boncompagni et al., 2017) concluded that CEUs formed following exercise enhance SOCE to optimize Ca²⁺release for muscle force production during repetitive stimulation. However, the implication of CEUs on SOCE and Ca²⁺release was based on indirect evidence (ICC and immuno-gold EM studies showing that exercise increases STIM1-Orai1 localization within CEUs and force production during repetitive stimulation). Direct evidence to support a role for CEUs in promoting Orai1- dependent SOCE and Ca²⁺release were lacking. Thus, a primary objective of the current study was to quantify Orai1-dependent SOCE and myoplasmic calcium transients during repetitive stimulation under conditions that result in CEU formation and disassembly. EM studies (in both EDL and FDB muscles) revealed that CEUs form < 1 hour after exercise and that TTs retract from the CEU earlier than SR stacks disassemble. We took advantage of this time course to assess changes in SOCE (Mn quench), calcium (mag-fluo4), and force (ex vivo EDL contractility) at different times during the assembly/disassembly of CEU component parts. Our results provide the first direct evidence that Orai1-dependent SOCE and Ca²⁺ release during repetitive stimulation are increased when TTs are associated with SR stacks. Unexpectedly, we found that acute exercise also results in substantial constitutive Orai1-dependent SOCE (with a magnitude similar to of maximal SOCE in control after full store depletion). These findings provide critical needed evidence to support the assertion that acute exercise promotes formation of CEUs to augment Orai1-dependent SOCE in order to optimize Ca²⁺ release and muscle contractility during repetitive stimulation.

We agree with the Editor and reviewers that the next important question to address is to explain how increased SOCE after exercise optimizes the myoplasmic Ca²⁺transient during repetitive stimulation. Since we did not have direct evidence for a specific mechanism, we refrained from speculating on this in the Discussion section of the manuscript. If provided an opportunity to submit a revised version of the manuscript, we would be happy to consider these potential mechanisms and their implications in the Discussion section. We have also initiated new experiments (outlined below) in hopes of being able to provide at least some evidence to support/exclude these possibilities. If successful, we are prepared to include the results on these studies in a revised version of the manuscript. However, given the importance and extent of the results/contributions of the study outlined above (none of which were questioned in the Editor’s summary of reviewer comments), we would hope that publication of our study will not be inexorably linked to our providing compelling evidence for an answer to this question.

We plan to revise the Discussion section of the manuscript to provide a comprehensive description of potential reasons for how increased SOCE function after acute exercise could result in enhanced peak Ca²⁺ transients during repetitive stimulation. Our discussion would include each of the three points below raised by the Editor and reviewers. Provided sufficient time, we also compete the series of experiments below designed to provide supportive evidence for one or more of these explanations.

However, several non-exclusive mechanisms could contribute:

1) An increase in SR Ca²⁺ content leading to increased Ca²⁺ release flux might be responsible but is inconsistent with the 25% reduction in content that was reported. The method used to measure total SR content (Ca²⁺ release cocktail containing ionomycin) is not sufficiently quantitative and may also release Ca²⁺ from other compartments (e.g., mitochondria). To address this concern, a ratiometric Ca²⁺ indicator should be targeted to the SR to directly measure the store content during stimulation, and the Ca²⁺ release flux can be estimated from cytosolic transients assuming generally accepted cytosolic buffering parameters.

To clarify, our measurements of a ~25% reduction in total Ca²⁺ store content after acute exercise were conducted on resting non-stimulated fibers, whereas the increase in Ca²⁺release during repetitive high-frequency stimulation was only observed after 8-10 consecutive stimulation trains. In addition, the increase in peak Ca²⁺transient amplitude in fibers from exercised mice developed during the time course of each 500 ms stimulus (i.e. the peak was similar at the beginning of the train but larger by the end of the train – see blue trace in Figure 5A, middle). Thus, the evolution of the increase in evoked Ca²⁺ release is dynamic, making resting steady state store content measurements irrelevant Ca²⁺ regard to SR Ca²⁺ available for release at the end of the 10th stimulus train. Thus, one possibility is that in fibers from exercised mice, increased SOCE during each stimulus train enhances store Ca²⁺such that Ca²⁺load is increased to a level sufficient to augment Ca²⁺flux during the 10^th and subsequent trains. We are planning experiments where we will measure either total C Ca²⁺ a store content (using ionomycin/CPA/EGTA or ICE as before) or just RYR1-releasable store content (using 4-chloro-m-cresol) after delivery of 10 stimulus trains. Since the longitudinal SR (with SERCA) and CEU junctions (with STIM1-Orai1) lie outside the triad and lack RYR1 release channels, we expect that the ICE assay will provide a more accurate assessment of the impact of SOCE on total Ca²⁺ store content. However, it will be interesting to see if the RYR1-releasable Ca²⁺ pool following 10 tetani is increased in fibers from exercised mice. Given the extent of SR and amount of SR Ca²⁺binding proteins in skeletal muscle, the peak ICE response in experiments is dominated by Ca²⁺within the SR compartment. Consistent with this, we previously used the Ca²⁺dependent UV absorbance spectra of BAPTA to show that the vast majority (~80%) of total Ca²⁺in EDL muscle resides within the SR (Lambolay, et al., J Gen Physiol, 2015; commented on by Manno and Rios, J Gen Physiol, 2015). In addition, the peak ICE response in fura-2ff-loaded FDB fibers is markedly reduced (~90%) in fibers from CASQ1 KO mice (see Author response image 1 below). Thus, Ca²⁺within mitochondria and other compartments contributes <10% of the ICE response in these experiments. Nevertheless, we plan to repeat the store content experiments in FDB fibers expressing a low-affinity, SR targeted Ca²⁺probe (D1ER, Reggiani et al., 2010; RCepia1-ER, Suzuki et al., 2014) introduced via electroporation.

Author response image 1. — A) Representative fura-FF ratio traces of total Ca²⁺ store content elicited during application of ICE in FDB fibers from wild-type (WT) and CASQ1-null mice. B) Peak change in fura-FF ratio during ICE application. n = number of FDB fibers analyzed; *p<0.05. Data are shown as mean ± SEM.

2) SOCE may elevate cytosolic Ca²⁺ enough to reduce fast buffering and thereby potentiate the effect of Ca²⁺ release. This could be assessed by a more detailed measurement of cytosolic Ca²⁺ between transients and estimating the effect of this increased Ca²⁺ on the buffering power.

We agree. Consistent with our observation of significant constitutive SOCE after exercise, resting Ca²⁺levels were increased in FDB fibers from exercised mice (Figure 5E). Thus, the increased SOCE during repetitive stimulation could further elevate myoplasmic Ca²⁺levels enough to reduce fast Ca²⁺ buffering (e.g. parvalbumin) during subsequent stimuli. In fact, this could in part explain the time-dependent increase observed in the peak of the Ca²⁺transient during a given stimulus train in fibers from exercised mice (see blue trace in Figure 5A, middle). To further test this idea, we plan to quantify resting Ca²⁺levels in fibers from control and exercised mice immediately after 10 consecutive high-frequency stimulus trains.

3) Finally, SERCA may be inhibited after exercise due to glycogen depletion. Reduction of SERCA activity could enhance the transient amplitude, as well as contribute to the partial depletion of the SR and tonic SOCE that was observed. SERCA activity should be measured from the Ca²⁺ transient decay rates at different times following exercise. Importantly, such measurements may reveal a time-dependent change in Ca²⁺ handling after exercise that is not directly related to the assembly and disassembly of CEUs.

This is an interesting suggestion that we had not considered previously. To obtain an estimate of SERCA activity in intact fibers from control and exercised mice before and after repetitive stimulation, we will fit the decay phase of electrically-evoked twitch Ca²⁺transients measured using the low affinity Ca²⁺dye, mag-fluo-4. The decay of the Ca²⁺transient is well-described by a double exponential fit in which the fast component primarily reflects binding of Ca²⁺to fast Ca²⁺buffers (e.g. dye, parvalbumin, troponin C; Carroll et al., 1997; Baylor and Hollingsworth, 2003; Capote et al., 2005) while the slow component reflects SERCA-mediated Ca²⁺ reuptake. The relative contribution of the slow (A_slow) and fast (A_fast) amplitudes will be calculated by dividing the absolute value of each component by the sum of the two components (A_slow/A_total and A_fast/A_total). This kinetic analysis will be completed on twitch Ca²⁺transients measured in FDB fibers from control and exercised mice both immediately before or after delivery of 10 consecutive high frequency stimulation trains. Comparison of the fast and slow amplitudes (absolute and relative) and the time constants of the two kinetic components will provide an index of the effects of exercise and repetitive stimulation on fast Ca²⁺buffering and SERCA activity.

[Editors’ notes: the authors’ response after being formally invited to submit a revised submission follows.]

Essential revisions:

All three reviewers thought the manuscript is an important extension of your earlier work on muscle calcium entry units (CEUs) formed after exercise. The time course of changes in SR-TT contacts, SOCE, Ca²⁺ transients, and force production following exercise provides indirect support for the idea that CEUs are involved in exercise-dependent effects on muscle contractility. However, there was also strong agreement that more quantitative experimental data are needed to show how the observed changes in SOCE lead to altered muscle function during exercise. This could occur by several non-exclusive mechanisms, which need to be tested:

1) An increase in SR Ca²⁺ content leading to increased Ca²⁺ release flux might be responsible. The free Ca²⁺ in the SR needs to be measured during stimulus trains to determine if it is elevated. The Ca²⁺ release cocktail method will indicate total stored Ca²⁺, not free Ca²⁺. Instead, a ratiometric Ca²⁺ indicator such as D1ER should be targeted to the SR to directly measure the free luminal Ca²⁺ during stimulation.

Thus, one possibility is that in fibers from mice <1 hour after exercise, increased SOCE during each stimulus train progressively enhances total Ca²⁺ store content such that the total releasable Ca²⁺ load is increased to a level sufficient to augment Ca²⁺ flux during the seventh and subsequent stimulus trains. To test this idea, we conducted new experiments in fura-FF-loaded fibers to assess total releasable Ca²⁺ store content (using ionomycin/CPA/EGTA or ICE as before) after delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 seconds). Results from these experiments are quite informative as both myoplasmic resting and total releasable Ca²⁺ store content were increased significantly after 10 stimulus trains in fibers from acutely exercised mice compared to that observed for fibers from both sedentary mice and <1 hour after exercise in the absence of stimulation (see Author response image 2). These results are consistent with the increase in Ca²⁺ transient amplitude following the seventh stimulation train, resulting at least in part from increased SOCE activity in fibers from mice <1 hour after exercise enhancing total SR Ca²⁺ store content such that releasable Ca²⁺ load is increased to a level sufficient to augment Ca²⁺ flux. These new data are included in revised Figure 5D and E (original Figure 5D and E panels are now presented as Figure 5—figure supplement 1).

Author response image 2. — A) Quantitative analysis of total Ca²⁺ store content elicited during application of ICE (10 µM ionomycin, 30 µM CPA, and 100 µM EGTA) in fura-FF-loaded FDB fibers in the absence (solid bars) and following (shaded bars) delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 seconds). E) Quantitative analysis of resting myoplasmic Ca²⁺ concentration recorded in fura-2-loaded FDB fibers in the absence (solid bars) and following (shaded bars) delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 seconds). n = number of FDB fibers analyzed; *p<0.05. Data are shown as mean ± SEM.

Given the extent of SR and level of SR Ca²⁺ binding proteins in skeletal muscle, the peak response to application of Ca²⁺ free solution containing ionomycin/cyclopiazonic acid/EGTA (ICE) is dominated by Ca²⁺ within the SR compartment. Consistent with this, we previously used the Ca²⁺ dependent UV absorbance spectra of BAPTA to show that the vast majority (~80%) of total Ca²⁺ in EDL muscle resides within the SR (Lambolay et al., 2015; commented on by Manno and Rios, 2015). Importantly, the peak ICE response in fura-FF-loaded fibers is markedly reduced (~90%) in FDB fibers from calsequestrin-1 knockout (CASQ1-null) mice, as expected since CASQ1 is the primary SR Ca²⁺ binding protein in muscle (see Author response image 1). These data suggest that Ca²⁺ within mitochondria and other compartments contributes <10% of the ICE response in these experiments.

Nevertheless, we conducted experiments using a SR-targeted Ca²⁺probe (D1ER). However, it is important to point out upfront that the D1ER probe reports the “free Ca²⁺concentration” within the SR/ER lumen, not the total releasable Ca²⁺content as assessed in our fura-FF studies. For these experiments, the D1ER probe was expressed in FDB fibers via electroporation (Rudolf et al., 2006; Canato et al., 2010). As reported previously, the D1ER probe localizes nicely to both triad junctions and longitudinal extensions of the SR, which are plentiful within the I band of the sarcomere (Figure 5—figure supplement 2A). Using identical excitation and emission gains, the D1ER YFP/CFP ratio was not significantly different in fibers from control mice and mice <1 hour after exercise both under resting conditions with a full SR Ca²⁺ store and after depletion of SR Ca²⁺ with ICE (Figure 5—figure supplement 2B and C). Thus, the spatially-averaged Ca²⁺ concentration within the SR lumen is not significantly altered by acute exercise (though areas of local depletion are possible). These findings are consistent with the strong buffering power in the SR in skeletal muscle and prior observations (e.g. Sztretye et al., 2011; Manno et al., 2013) that the free Ca²⁺ concentration in the SR lumen (~500 μM) remains largely unaltered in face of marked changes in total SR Ca²⁺ content. These results are included in the revised manuscript (Figure 5—figure supplement 2).

2) SOCE may elevate cytosolic Ca²⁺ enough to reduce fast buffering and thereby potentiate the effect of Ca²⁺ release. The effect on buffering could be assessed by a more detailed measurement of cytosolic Ca²⁺ between individual transients and after stimulus trains in control and exercised mice.

As shown above, resting myoplasmic Ca²⁺ levels were significantly elevated in FDB fibers isolated from mice <1 hour after acute exercise. Importantly, we now show (see panel B in Author response image 2) that this increase in resting Ca²⁺ in FDB fibers after acute exercise (where SOCE activity is increased) was even greater immediately after delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 seconds). Thus, the elevation in resting Ca²⁺ after acute exercise could conceivably reduce fast buffering enough to potentiate the effect of Ca²⁺ release. However, detailed quantitative analysis of twitch Ca²⁺ transient decay kinetics before and after 10 stimulus trains does not support a significant change in either fast Ca²⁺ buffering or SERCA-mediated Ca²⁺ reuptake (see Figure 5—figure supplement 3). We updated the Discussion section of the revised manuscript to include a consideration of these.

3) Finally, SERCA may be inhibited after exercise due to glycogen depletion. Reduction of SERCA activity could enhance the transient amplitude, as well as contribute to the partial depletion of the SR and tonic SOCE that was observed in resting muscle after exercise. Kinetic analysis of cytosolic Ca²⁺ twitch transients (fast and slow components) can be used to estimate changes in SERCA activity and Ca²⁺ buffering from control and exercised mice before and after 10 stimulus trains. Importantly, such measurements may reveal a time-dependent change in Ca²⁺ handling after exercise that is not directly related to the assembly and disassembly of CEUs.

We thank the reviewer for this excellent suggestion. We completed the suggested experiments to quantify the kinetics of electrically-evoked twitch Ca²⁺ transient decay using a low affinity Ca²⁺ dye (mag-fluo-4) before and after delivery of 10 consecutive stimulus trains (500 ms, 50 Hz, every 2.5 seconds) in FDB fibers from both sedentary mice and mice <1 hour after exercise. Using this approach, the decay phase of the electrically-evoked mag-fluo-4 twitch transient is well-described by a double exponential fit in which the fast component primarily reflects Ca²⁺ binding to fast Ca²⁺ buffers while the slow component is dominated by SERCA-mediated SR Ca²⁺ reuptake (Carroll et al., 1997; Baylor and Hollingsworth, 2003; Capote et al., 2005). For these experiments, we first recorded 10 electrically-evoked twitch transients (0.5 Hz) in FDB fibers from either sedentary mice or mice <1 hour after exercise before delivery of 10 stimulus trains (500 ms, 50 Hz, every 2.5 seconds). After the 10 stimulus trains, a second set of 10 electrically-evoked twitch transients (0.5 Hz) were recorded. Despite a significant increase in the peak Ca²⁺ transient amplitude during the 10^th stimulus train (1.51 ± 0.07) compared to the first stimulus train (1.32 ± 0.09) only in fibers from mice <1 hour after exercise, no significant changes in Ca²⁺ transient decay kinetics were observed compared to that obtained for fibers isolated from sedentary mice. These results are included in the revised manuscript (Figure 5—figure supplement 3).

4) The idea that CEUs are responsible for the constitutive SOCE after exercise is not well supported. In the Discussion section, the authors suggest that the new CEUs can account for the increased tonic SOCE after exercise. The supporting argument is that "non-depletion" quench after exercise (1200 cps) + TG/CPA quench before exercise (1700 cps) = TG/CPA quench rate after exercise (2800 cps). However, the quench after exercise is driven by only partial SR depletion, so TG/CPA should add more than 1200. This argument does not make quantitative sense to me. Unless it can be better substantiated, it should be deleted.

As suggested, we removed this speculation from the Discussion section of the revised manuscript.

5) A related issue concerns the idea that CEUs are solely responsible for the SOCE that augments muscle function after exercise. The loss of SR Ca²⁺ content peaks at the same time (<1 hour) as "non-depletion" SOCE, Ca²⁺ release transients and force production during trains. This complicates things. Because of tonic depletion, it is possible that pre-existing STIM and Orai complexes (activated by tonic store depletion) distinct from the new CEUs are being activated and contribute to augmented function. This possibility needs to be explicitly acknowledged, and the conclusions regarding CEU functions toned down.

We thank the reviewer for pointing out this possibility, which is now explicitly acknowledged in the Discussion section of the revised manuscript.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

The manuscript has been improved, and the new data have addressed several of the issues that were raised in the original review. However, there are some remaining issues that need to be addressed before acceptance, as outlined below:

1) (Subsection “Total Releasable Ca²⁺ Store Content is Reduced <1 Hour After Exercise, but Increased Following Repetitive, High-frequency Stimulation” and Discussion section). After exercise, the constitutive SOCE is greatly increased, but the new D1ER measurements show that resting free Ca²⁺ in the SR is not altered. Because STIM1 and SOCE are controlled by free Ca²⁺ in ER/SR, and not the total Ca²⁺ content, how can Orai activity be increased without a reduction of free SR Ca²⁺? The decrease in total SR Ca²⁺ that was observed cannot explain the appearance of constitutive SOCE. This conundrum needs to be clearly stated and possible explanations provided. While uncertainty in the D1ER measurement might mask a small depletion of free Ca²⁺, this is unlikely to be the whole explanation since the SOCE amplitude was so large (i.e., about the same as maximal SOCE from full SR depletion before exercise).

We thank the reviewer for raising this important point.

The D1ER measurements indeed show that the “spatially-averaged” free Ca²⁺ concentration in the SR measured with this probe is not significantly altered after exercise (though areas of local depletion are possible). It is important to note, however, that our D1ER localization results indicate that the majority of the D1ER signal is in the triad junction (Figure 5—figure supplement 1A). Thus, the D1ER data primarily reflect the free Ca²⁺ concentration within the triad junction, which is not different after exercise. These results suggest that the constitutive SOCE observed after exercise does not originate from Orai1 channels in the triad junction. Rather, the constitutive Ca²⁺ entry more likely arises from Orai1 channels localized in the elongated TT within the I band region of the sarcomere that couples with SR-stack membranes forming CEUs, that are not well-reflected in the spatially-averaged D1ER signal. These observations support the idea that CEUs represent the site of constitutive Ca²⁺ entry after exercise. An alternative explanation is that exercise could produce a signal that directly activates Orai1 independent of store depletion (even within the triad).

While admittedly too superficial, we attempted to address this issue in the following sentence from the Discussion section of the prior version of the revised manuscript:

"Thus, the constitutive Ca²⁺ entry could be due to either a fraction of the SR store being partially depleted (Figure 5—figure supplement 1) or generation of a direct activator of Orai1 channels after exercise."

By “fraction of the SR being partially depleted” we were indirectly referring to CEUs formed after exercise. We also left open the possibility that exercise could produce a direct activator of Orai1 that does not require a reduction in the SR Ca²⁺ concentration. However, we appreciate the reviewers point and realize that our attempt at brevity missed several important salient points (discussed above). Therefore, we further revised this part of the Discussion section to more explicitly consider these issues.

2) The new twitch analyses are convincing, but it is not clear to anyone but an expert muscle physiologist how these were used to draw conclusions about Ca²⁺ binding and SERCA activity. In Subsection “Total Releasable Ca²⁺ Store Content is Reduced <1 Hour After Exercise, but Increased Following Repetitive, High-frequency Stimulation”, the statement that kinetics of twitch Ca²⁺ transient decay do not change needs to be explained; why was this experiment done, and what is the interpretation? In the Discussion section, it would help to discuss the possibilities that were brought up in review, including saturation of buffers and depletion of glycogen and effects on SERCA, and how the new results argue against these alternative explanations. An example of a transient should be shown along with the analysis in Figure 5—figure supplement 3 to illustrate the fast and slow components and how they were fitted.

We apologize for not providing a clear rationale in the Results section for why the twitch Ca²⁺ transient decay kinetics were analyzed (the rationale was presented in the Methods section). We revised the manuscript to include this rationale in the Results section, just before these data are presented (subsection “Total Releasable Ca²⁺ Store Content is Reduced”). As suggested, we also now include example twitch Ca²⁺ traces with superimposed two-exponential fits in Figure 5—figure supplement 3. We now also discuss potential effects of saturation of Ca²⁺ buffers, depletion of glycogen and SERCA activity in the Discussion section of the revised manuscript.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Transparent reporting form

elife-47576-transrepform.docx^{(247KB, docx)}

DOI: 10.7554/eLife.47576.017

Data Availability Statement

All data generated or analysed during this study are included in the manuscript and supporting files.

[bib1] Ainbinder A, Boncompagni S, Protasi F, Dirksen RT. Role of Mitofusin-2 in mitochondrial localization and calcium uptake in skeletal muscle. Cell Calcium. 2015;57:14–24. doi: 10.1016/j.ceca.2014.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] Baylor SM, Hollingworth S. Sarcoplasmic reticulum calcium release compared in slow-twitch and fast-twitch fibres of mouse muscle. The Journal of Physiology. 2003;551:125–138. doi: 10.1113/jphysiol.2003.041608. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] Boncompagni S, Rossi AE, Micaroni M, Beznoussenko GV, Polishchuk RS, Dirksen RT, Protasi F. Mitochondria are linked to calcium stores in striated muscle by developmentally regulated tethering structures. Molecular Biology of the Cell. 2009;20:1058–1067. doi: 10.1091/mbc.e08-07-0783. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] Boncompagni S, Thomas M, Lopez JR, Allen PD, Yuan Q, Kranias EG, Franzini-Armstrong C, Perez CF. Triadin/Junctin double null mouse reveals a differential role for triadin and junctin in anchoring CASQ to the jSR and regulating ca(2+) homeostasis. PLOS ONE. 2012a;7:e39962. doi: 10.1371/journal.pone.0039962. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] Boncompagni S, Protasi F, Franzini-Armstrong C. Sequential stages in the age-dependent gradual formation and accumulation of tubular aggregates in fast twitch muscle fibers: serca and calsequestrin involvement. Age. 2012b;34:27–41. doi: 10.1007/s11357-011-9211-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] Boncompagni S, Michelucci A, Pietrangelo L, Dirksen RT, Protasi F. Exercise-dependent formation of new junctions that promote STIM1-Orai1 assembly in skeletal muscle. Scientific Reports. 2017;7:14286. doi: 10.1038/s41598-017-14134-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] Brandman O, Liou J, Park WS, Meyer T. STIM2 is a feedback regulator that stabilizes basal cytosolic and endoplasmic reticulum Ca2+ levels. Cell. 2007;131:1327–1339. doi: 10.1016/j.cell.2007.11.039. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] Canato M, Scorzeto M, Giacomello M, Protasi F, Reggiani C, Stienen GJ. Massive alterations of sarcoplasmic reticulum free calcium in skeletal muscle fibers lacking calsequestrin revealed by a genetically encoded probe. PNAS. 2010;107:22326–22331. doi: 10.1073/pnas.1009168108. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] Capote J, Bolaños P, Schuhmeier RP, Melzer W, Caputo C. Calcium transients in developing mouse skeletal muscle fibres. The Journal of Physiology. 2005;564:451–464. doi: 10.1113/jphysiol.2004.081034. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] Carrell EM, Coppola AR, McBride HJ, Dirksen RT. Orai1 enhances muscle endurance by promoting fatigue-resistant type I fiber content but not through acute store-operated Ca2+ entry. The FASEB Journal. 2016;30:4109–4119. doi: 10.1096/fj.201600621R. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] Carroll S, Nicotera P, Pette D. Calcium transients in single fibers of low-frequency stimulated fast-twitch muscle of rat. American Journal of Physiology-Cell Physiology. 1999;277:C1122–C1129. doi: 10.1152/ajpcell.1999.277.6.C1122. [DOI] [PubMed] [Google Scholar]

[bib12] Darbellay B, Arnaudeau S, König S, Jousset H, Bader C, Demaurex N, Bernheim L. STIM1- and Orai1-dependent store-operated calcium entry regulates human myoblast differentiation. Journal of Biological Chemistry. 2009;284:5370–5380. doi: 10.1074/jbc.M806726200. [DOI] [PubMed] [Google Scholar]

[bib13] Edwards JN, Blackmore DG, Gilbert DF, Murphy RM, Launikonis BS. Store-operated calcium entry remains fully functional in aged mouse skeletal muscle despite a decline in STIM1 protein expression. Aging Cell. 2011;10:675–685. doi: 10.1111/j.1474-9726.2011.00706.x. [DOI] [PubMed] [Google Scholar]

[bib14] Feske S, Gwack Y, Prakriya M, Srikanth S, Puppel SH, Tanasa B, Hogan PG, Lewis RS, Daly M, Rao A. A mutation in Orai1 causes immune deficiency by abrogating CRAC channel function. Nature. 2006;441:179–185. doi: 10.1038/nature04702. [DOI] [PubMed] [Google Scholar]

[bib15] Kimura T, Lueck JD, Harvey PJ, Pace SM, Ikemoto N, Casarotto MG, Dirksen RT, Dulhunty AF. Alternative splicing of RyR1 alters the efficacy of skeletal EC coupling. Cell Calcium. 2009;45:264–274. doi: 10.1016/j.ceca.2008.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] Ko JK, Choi KH, Zhao X, Komazaki S, Pan Z, Weisleder N, Ma J. A versatile single-plasmid system for tissue-specific and inducible control ofgene expression in transgenic mice. FASEB Journal : Official Publication of the Federation of American Societies for Experimental Biology. 2011;25:2638–2649. doi: 10.1096/fj.11-181560. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] Koenig X, Choi RH, Launikonis BS. Store-operated Ca2+ entry is activated by every action potential in skeletal muscle. Communications Biology. 2018;1:31. doi: 10.1038/s42003-018-0033-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] Kurebayashi N, Ogawa Y. Depletion of Ca2+ in the sarcoplasmic reticulum stimulates Ca2+ entry into mouse skeletal muscle fibres. The Journal of Physiology. 2001;533:185–199. doi: 10.1111/j.1469-7793.2001.0185b.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] Lanner JT, Georgiou DK, Dagnino-Acosta A, Ainbinder A, Cheng Q, Joshi AD, Chen Z, Yarotskyy V, Oakes JM, Lee CS, Monroe TO, Santillan A, Dong K, Goodyear L, Ismailov II, Rodney GG, Dirksen RT, Hamilton SL. AICAR prevents heat-induced sudden death in RyR1 mutant mice independent of AMPK activation. Nature Medicine. 2012;18:244–251. doi: 10.1038/nm.2598. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] Launikonis BS, Ríos E. Store-operated Ca2+ entry during intracellular Ca2+ release in mammalian skeletal muscle. The Journal of Physiology. 2007;583:81–97. doi: 10.1113/jphysiol.2007.135046. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] Loy RE, Orynbayev M, Xu L, Andronache Z, Apostol S, Zvaritch E, MacLennan DH, Meissner G, Melzer W, Dirksen RT. Muscle weakness in Ryr1I4895T/WT knock-in mice as a result of reduced ryanodine receptor Ca2+ ion permeation and release from the sarcoplasmic reticulum. The Journal of General Physiology. 2011;137:43–57. doi: 10.1085/jgp.201010523. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] Lyfenko AD, Dirksen RT. Differential dependence of store-operated and excitation-coupled Ca2+ entry in skeletal muscle on STIM1 and Orai1. The Journal of Physiology. 2008;586:4815–4824. doi: 10.1113/jphysiol.2008.160481. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] Michelucci A, García-Castañeda M, Boncompagni S, Dirksen RT. Role of STIM1/ORAI1-mediated store-operated Ca2+ entry in skeletal muscle physiology and disease. Cell Calcium. 2018;76:101–115. doi: 10.1016/j.ceca.2018.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] Orci L, Ravazzola M, Le Coadic M, Shen WW, Demaurex N, Cosson P. From the cover: stim1-induced precortical and cortical subdomains of the endoplasmic reticulum. PNAS. 2009;106:19358–19362. doi: 10.1073/pnas.0911280106. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib25] Pan Z, Yang D, Nagaraj RY, Nosek TA, Nishi M, Takeshima H, Cheng H, Ma J. Dysfunction of store-operated calcium channel in muscle cells lacking mg29. Nature Cell Biology. 2002;4:379–383. doi: 10.1038/ncb788. [DOI] [PubMed] [Google Scholar]

[bib26] Perni S, Dynes JL, Yeromin AV, Cahalan MD, Franzini-Armstrong C. Nanoscale patterning of STIM1 and Orai1 during store-operated Ca2+ entry. PNAS. 2015;112:E5533–E5542. doi: 10.1073/pnas.1515606112. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] Roos J, DiGregorio PJ, Yeromin AV, Ohlsen K, Lioudyno M, Zhang S, Safrina O, Kozak JA, Wagner SL, Cahalan MD, Veliçelebi G, Stauderman KA. STIM1, an essential and conserved component of store-operated Ca2+ channel function. The Journal of Cell Biology. 2005;169:435–445. doi: 10.1083/jcb.200502019. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] Stiber J, Hawkins A, Zhang ZS, Wang S, Burch J, Graham V, Ward CC, Seth M, Finch E, Malouf N, Williams RS, Eu JP, Rosenberg P. STIM1 signalling controls store-operated calcium entry required for development and contractile function in skeletal muscle. Nature Cell Biology. 2008;10:688–697. doi: 10.1038/ncb1731. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] Thiel M, Lis A, Penner R. STIM2 drives Ca2+ oscillations through store-operated Ca2+ entry caused by mild store depletion. The Journal of Physiology. 2013;591:1433–1445. doi: 10.1113/jphysiol.2012.245399. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] Wei-Lapierre L, Carrell EM, Boncompagni S, Protasi F, Dirksen RT. Orai1-dependent calcium entry promotes skeletal muscle growth and limits fatigue. Nature Communications. 2013;4:2805. doi: 10.1038/ncomms3805. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] Wu MM, Covington ED, Lewis RS. Single-molecule analysis of diffusion and trapping of STIM1 and Orai1 at endoplasmic reticulum-plasma membrane junctions. Molecular Biology of the Cell. 2014;25:3672–3685. doi: 10.1091/mbc.E14-06-1107. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Transverse tubule remodeling enhances Orai1-dependent Ca2+ entry in skeletal muscle

Antonio Michelucci

Simona Boncompagni

Laura Pietrangelo

Maricela García-Castañeda

Takahiro Takano

Sundeep Malik

Robert T Dirksen

Feliciano Protasi

Roles

Abstract

Introduction

Results

SR-stacks increase for up to 6 hr after exercise

Figure 1. SR-stack incidence, number/area, and length after exercise.

Figure 1—figure supplement 1. SR-stack incidence, number/area, and T-tubule length after exercise.

TT length is increased <1 hr after exercise

Figure 2. TT extension at the I band after exercise.

Mn2+Quench of fura-2 fluorescence is increased <1 hr after exercise

Figure 3. Effect of exercise on the maximum rate of Mn2+ quench following store depletion.

Figure 4. Effect of exercise on the maximum rate of Mn2+ quench in the absence of store depletion.

Figure 4—figure supplement 1. Sensitivity of constitutive Mn2+ quench in non-depleted FDB fibers after exercise to SOCE inhibitor BTP-2.

Peak Ca2+transient amplitude during repetitive stimulation is enhanced <1 hr after exercise

Figure 5. Effect of exercise on peak Ca2+ transient amplitude during repetitive, high-frequency stimulation.

Figure 5—figure supplement 1. Effect of exercise on total Ca2+store content and resting myoplasmic Ca2+ concentration.

Figure 5—figure supplement 2. Measurements of free SR Ca levels in D1ER-expressing FDB fibers.

Figure 5—figure supplement 3. Effect of repetitive, high-frequency stimulation on Ca transient decay kinetics in FDB fibers from control and exercised mice.

Total releasable Ca2+ Store Content is reduced <1 hr after Exercise, but Increased Following Repetitive, High-frequency Stimulation

Maintenance of specific force during repetitive stimulation is increased <1 hr after exercise

Figure 6. Effect of exercise on contractile force during repetitive, high-frequency stimulation.

Figure 6—figure supplement 1. Effect of exercise on EDL muscle force-frequency curve and maximum tetanic force.

Orai1 is required for increased SOCE activity and EDL force decay following acute exercise

Figure 7. Effect of exercise on the maximum rate of Mn2+ quench and contractile force in iOrai1 KO mice.

Figure 7—figure supplement 1. Effect of exercise on the maximum rate of Mn2+ quench and contractile force in dnOrai1 mice.

Discussion

Figure 8. Model of structural events and functional effects of CEU assembly and disassembly following acute exercise and during recovery.

Materials and methods

Animals

Tamoxifen treatment

Treadmill exercise protocol

Electron microscopy (EM)

Quantitative analyses of EM images

Isolation of single FDB muscle fibers

Mn2+ quench measurements

Ca2+ transient measurements

Total Ca2+ store content measurements

Resting Ca2+ measurements

SR-free Ca2+ measurements

Immunofluorescence

Ex vivo muscle contractility measurements

Statistical analyses

Acknowledgements

Funding Statement

Contributor Information

Funding Information

Additional information

Competing interests

Author contributions

Ethics

Additional files

Data availability

References

Decision letter

Roles

Author response

Author response image 1. Effect of calsequestrin type 1 knockout (CASQ1-null) on total releasable Ca2+ store content.

Author response image 2. Effect of repetitive, high-frequency stimulation on total releasable Ca2+ store content and resting myoplasmic Ca2+ concentrationin FDB fibers from control and <1 hour after exercise mice.

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Transverse tubule remodeling enhances Orai1-dependent Ca²⁺ entry in skeletal muscle

Mn²⁺Quench of fura-2 fluorescence is increased <1 hr after exercise

Figure 3. Effect of exercise on the maximum rate of Mn²⁺ quench following store depletion.

Figure 4. Effect of exercise on the maximum rate of Mn²⁺ quench in the absence of store depletion.

Figure 4—figure supplement 1. Sensitivity of constitutive Mn²⁺ quench in non-depleted FDB fibers after exercise to SOCE inhibitor BTP-2.

Peak Ca²⁺transient amplitude during repetitive stimulation is enhanced <1 hr after exercise

Figure 5. Effect of exercise on peak Ca²⁺ transient amplitude during repetitive, high-frequency stimulation.

Figure 5—figure supplement 1. Effect of exercise on total Ca²⁺store content and resting myoplasmic Ca²⁺ concentration.

Total releasable Ca²⁺ Store Content is reduced <1 hr after Exercise, but Increased Following Repetitive, High-frequency Stimulation

Figure 7. Effect of exercise on the maximum rate of Mn²⁺ quench and contractile force in iOrai1 KO mice.

Figure 7—figure supplement 1. Effect of exercise on the maximum rate of Mn²⁺ quench and contractile force in dnOrai1 mice.

Mn²⁺ quench measurements

Ca²⁺transient measurements

Total Ca²⁺store content measurements

Resting Ca²⁺measurements

SR-free Ca²⁺measurements

Author response image 1. Effect of calsequestrin type 1 knockout (CASQ1-null) on total releasable Ca²⁺ store content.

Author response image 2. Effect of repetitive, high-frequency stimulation on total releasable Ca²⁺ store content and resting myoplasmic Ca²⁺ concentrationin FDB fibers from control and <1 hour after exercise mice.