Ryanodine receptor fragmentation and sarcoplasmic reticulum Ca²⁺ leak after one session of high-intensity interval exercise

HIIT Causes Fragmentation of RyR1 in Recreationally Active Men.

In an initial experiment to test whether a brief period of HIIT exercise can induce long-lasting changes in muscle function, recreationally active males (SI Appendix, Table S1) performed three 30-s all-out bouts of cycling (i.e., only 90 s of total exercise time) with 4-min rest between bouts. Subsequent contractions produced by electrical stimulation of knee extensors revealed a marked length-independent force decrease, especially at low (10 Hz) stimulation frequency, which was not fully recovered even 24 h after the brief HIIT exercise (SI Appendix, Fig. S1A). Thus, these initial experiments show that as little as three 30-s intervals of HIIT exercise can induce long-lasting impairments in contractile function.

In the next series of experiments, recreationally active males performed six 30-s all-out cycling bouts, and biopsies were taken from the vastus lateralis muscle before and at ∼10 min and 24 h after the cycling bouts (SI Appendix, Fig. S1B). To assess changes in RyR1 induced by this HIIT exercise, Western blot experiments were performed with a polyclonal antibody targeted against the last nine amino acids on the C-terminal end of human RyR (no. 5029; gift from Andrew Marks, Columbia University, New York). These experiments showed no obvious change in RyR1 directly after the HIIT exercise, but 24 h later, only ∼15% remained as the full-sized RyR1 monomer, and instead major fragments emerged at ∼375, 80, and 60 kDa (Fig. 1 A and B). Similarly, a commercially available mouse monoclonal anti-RyR1 antibody (ab2868; Abcam) showed a shift from the full-length RyR1 monomer to a ∼375-kDa fragment 24 h after exercise (SI Appendix, Fig. S2); note that the ab2868 antibody did not detect the smaller ∼80- and 60-kDa fragments, possibly because the cleavage sites then interfered with the binding site of this antibody. Conversely, neither the t-tubular voltage sensor (the dihydropyridine receptor; DHPR), the SR Ca²⁺ pump (SERCA2), the SR Ca²⁺ buffer (calsequestrin 1; CSQ1), nor the structural proteins dystrophin (DMD) and actin showed any change in expression or signs of fragmentation after the HIIT exercise (Fig. 1C). Moreover, Western blotting to assess the amount of ubiquitin-conjugated proteins showed no general difference between before and 10 min and 24 h after HIIT exercise (SI Appendix, Fig. S3).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

HIIT exercise induces extensive RyR1 fragmentation in recreationally active subjects. (A) Representative Western blot reveals decreased expression of full-sized RyR1 (red arrow) 24 h after exercise, which was accompanied by the appearance of fragments of ∼375, 80, and 60 kDa (indicated by black arrows). (B) Mean relative distribution of native RyR1 and its fragments from four subjects; the total intensity of all four analyzed bands was set to 100% at each time point in each subject. (C) Representative Western blots and mean data of DHPR, SERCA2, CSQ1, DMD, and actin expression ∼10 min (n = 11) and 24 h (n = 5) after exercise; relative expression before exercise was set to 100% in each subject. Data are expressed as mean ± SEM.

To investigate whether other types of exhaustive exercise also result in RyR1 fragmentation, we studied RyR1 modifications induced by a marathon foot race performed by male subjects regularly doing endurance training at a recreational level. Western blotting showed neither decreased RyR1 expression nor fragmentation at 1 and 24 h after the marathon race (SI Appendix, Fig. S4A). However, RyR1 immunoprecipitation experiments revealed a marked dissociation of the channel-stabilizing subunit calstabin1 (also known as FKBP12; SI Appendix, Fig. S4B), which is consistent with previous results obtained after strenuous endurance exercise and in muscle pathologies and which has been linked to increased RyR1 Ca²⁺ leakage (6⇓⇓⇓⇓⇓⇓–13). Thus, the extensive challenge to muscle integrity caused by marathon running resulted in destabilizing changes to RyR1, but no fragmentation.

HIIT Causes Force Depression Due to Defective SR Ca²⁺ Release in Muscle Fibers.

Tentative mechanisms underlying the decrease in contractile performance during and after the HIIT exercise were assessed both at the neuronal and muscular levels (SI Appendix, Fig. S5). Mean power output decreased as the series of cycling bouts progressed, being decreased by ∼25% in the sixth bout, and this decrease occurred despite constant neuronal activation (SI Appendix, Fig. S6). Maximum voluntary contraction (MVC) force was decreased by ∼40% immediately and 5 min after the repeated cycling bouts, and again this decrease was not accompanied by any reduction in neuronal activation (SI Appendix, Fig. S7).

We used supramaximal electrical stimulation of the femoral nerve to assess knee extensor muscle function without influence from neuronal activation. The force induced by 10- and 100-Hz doublet stimulation as well as the rate of twitch force development were substantially decreased immediately and 5 min after the six cycling bouts (SI Appendix, Fig. S8 A–C). Conversely, the membrane excitability seemed unaffected by the HIIT exercise, as judged from measurements of the muscle compound action potential (M wave) in response to a single electrical impulse (SI Appendix, Fig. S8D). Intriguingly, no statistically significant differences from prefatigue values were observed when the above MVC contractions and experiments with electrical femoral nerve stimulation were performed 24 h after exercise—i.e., at the time when RyR1 Western blots show extensive fragmentation.

Thus far, our results show a force depression induced by a single session of HIIT that is due to defective function within the muscle fibers. The close to normal action potential characteristics (i.e., virtually unaltered M-wave properties) after exercise indicate that the force depression is due to factor(s) intrinsic to the muscle fibers—i.e., decreased SR Ca²⁺ release and/or impaired myofibrillar contractile function. To distinguish between these two possibilities, we measured the force produced during direct stimulation of the contractile proteins in skinned fibers obtained from vastus lateralis muscle biopsies taken before and ∼10 min after the repeated cycling bouts. The results showed no HIIT exercise-induced change in maximum Ca²⁺-activated force or myofibrillar Ca²⁺ sensitivity (SI Appendix, Fig. S9). Note that the cycling performed during the HIIT exercise involved mainly concentric contractions. A long-lasting force depression was observed after unaccustomed eccentric contractions, but such contractions resulted in severe impairments in myofibrillar contractility and a shift of the active force–length relationship toward longer lengths (31); neither of these defects were observed after the present HIIT exercise (see also SI Appendix, Fig. S1A). Thus, contractile function of the myofibrillar proteins was not impaired after the HIIT exercise, and the mechanism behind the force depression can be narrowed down to defective SR Ca²⁺ release.

Elite Endurance Athletes Develop a Prolonged HIIT-Induced Force Depression, but No RyR1 Fragmentation.

In the next set of experiments, we tested whether the HIIT exercise-induced RyR1 fragmentation also occurs in individuals with a high aerobic capacity. Fourteen elite endurance runners or road cyclists (SI Appendix, Table S1) performed the six bouts of 30-s all-out cycling. The mean power decreased as the series of cycling bouts progressed also in these athletes (SI Appendix, Fig. S10A), but the average decrease in the sixth bout was slightly smaller (∼15%) than in the recreationally active subjects (∼25%). Moreover, there was a marked decrease in electrically stimulated force production after exercise, especially at the low (10 Hz) stimulation frequency (SI Appendix, Fig. S10B). Intriguingly, Western blots showed no signs of increased RyR1 fragmentation after the HIIT exercise in the elite athletes (Fig. 2 A and B), which is in sharp contrast to the marked fragmentation observed in the recreationally active subjects.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

HIIT exercise does not induce RyR1 fragmentation in elite endurance athletes. (A) Representative Western blots show no signs of RyR1 fragmentation after the cycling bouts in the elite athletes. Arrows indicate full-sized RyR1 (red arrow) and the location of ∼375-, 80-, and 60-kDa fragments (black arrows) observed 24 h after exercise in recreationally active subjects (Fig. 1A). (B) Mean data (± SEM) obtained from 14 elite athletes before (Pre) and ∼10 min (Post) and 24 h after exercise; total RyR1 expression was set to 100% at each time point in each subject. (C, Upper) Representative Western blots of SOD2, catalase, and DHPR from biopsies taken before the HIIT exercise in recreationally active subjects (Rec) and elite athletes (EA). DHPR did not differ between the two groups and was used as loading control. (C, Lower) Bar graphs show mean SOD2 and catalase expressions (± SEM; n = 7) relative to the mean in the Rec group, which was set to 100%. **P < 0.01 in unpaired t test. (D) Mean data (± SEM; n = 6–8) of the transcript levels of PGC-1α1 and -1α4 expressed relative to hypoxanthine guanine phosphoribosyl transferase (HPRT), which did not differ between the groups and was used as a housekeeping gene. ^#P < 0.05; ^## P < 0.01; ^### P < 0.001 vs. before exercise (one-way repeated measures ANOVA/Holm–Sidak post hoc test). PGC-1α4 was significantly higher before exercise in Rec than in EA (P < 0.05; unpaired t test).

Increased ROS production during exercise is classically linked to enhanced mitochondrial respiration, resulting in increased superoxide (O₂⁻) production in complexes I and III of the electron transport chain (32). Superoxide dismutase 2 (SOD2) and catalase have key roles in cellular ROS metabolism by converting superoxide into hydrogen peroxide (H₂O₂) and H₂O₂ into water, respectively. We measured the protein expression of SOD2 and catalase in vastus lateralis muscle before the HIIT exercise and observed at least twice as high expression in the elite athletes as in the recreationally active subjects (Fig. 2C).

Changes in cellular Ca²⁺ handling can affect gene transcription and hence the adaptive response to physical exercise (33, 34). The peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) transcriptional coactivators have key roles for muscle adaptations, with PGC-1α1 being critically important for adaptations to endurance-type exercise and PGC-1α4 more important for resistance-type exercise (4, 35). The transcript levels for both these PGC-1α isoforms were significantly increased directly after the HIIT exercise in muscle biopsies from both recreationally active subjects and elite endurance athletes (Fig. 2D). Intriguingly, 24 h after the HIIT exercise, these transcripts were decreased by ∼80% in recreationally active subjects, whereas they were back at the pre-exercise level in the elite athletes. Moreover, transcripts of PGC-1α1–targeted genes encoding for mitochondrial proteins and several transcription factors that change in response to exercise also showed markedly decreased transcript levels 24 h after exercise only in recreationally active subjects (SI Appendix, Fig. S11). Thus, the HIIT exercise triggered prolonged changes in gene transcription in the recreationally active subjects, but not in the elite endurance athletes.

HIIT-Induced Fragmentation of RyR1 Is ROS-Dependent.

The absence of RyR1 changes combined with higher SOD2 and catalase protein expressions in the elite athletes suggests an involvement of ROS in the triggering of RyR1 fragmentation. Experiments on isolated mouse flexor digitorum brevis (FDB) muscle, which is a fast-twitch toe muscle containing mainly type IIa/IIx fibers (15), were performed to specifically study tentative ROS-induced modifications of RyR1. The mitochondrial ROS production was measured with the fluorescent indicator MitoSOX Red in single FDB fibers from sedentary control mice and mice that had free access to a running wheel in the cage. The latter mice performed voluntary endurance training by running ∼20 km each night for 40 d (SI Appendix, Fig. S12A). The isolated fibers were activated with electrical current pulses and a stimulation scheme mimicking the activation pattern during the all-out cycling bouts (six 30-s periods of 250 ms tetanic 100-Hz stimulation given every 500 ms with 4 min of rest between the stimulation periods). At 5 and 10 min after the simulated HIIT exercise, the MitoSOX Red fluorescence was increased by ∼200% in the sedentary control mice, whereas the increase was significantly smaller (by ∼80%) in the endurance-trained mice (P < 0.01; SI Appendix, Fig. S12B). The ROS-induced increase in MitoSOX Red fluorescence is not reversible. The stable fluorescence between 5 and 10 min after exercise therefore indicated that ROS production returned to a low baseline level once the HIIT-mimicking stimulation was stopped. Thus, there was a marked increase in mitochondrial ROS production during the simulated HIIT exercises, and this increase was attenuated with endurance training.

Next, intact, single-digit FDB muscles were activated with the HIIT-mimicking stimulation scheme, and Western blots were performed on muscles frozen 5 min after the last contraction displayed no signs of RyR1 degradation, and DHPR expression was similar to the control level. However, there was a doubling of RyR1 malondialdehyde (MDA) adducts (Fig. 3A). MDA protein adducts reflect the degree of lipid peroxidation and are frequently used as a biomarker of increased ROS production (36), and the RyR1 protein complex is known to be highly susceptible to ROS-induced modifications (37).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Simulated HIIT exercise causes a ROS-dependent RyR1 fragmentation in mouse muscle. (A) Representative Western blots of RyR1, MDA adducts on RyR1, DHPR, and actin obtained from single-digit FDB muscles 5 min after being fatigued by six bouts of 30-s simulated HIIT exercise or kept at rest (Ctrl). The contractions had no effect on RyR1, DHPR, and actin expression, but it approximately doubled the amount of MDA adducts on RyR1. (B) Western blot of RyR1 on single-digit FDB muscles snap-frozen either at rest (Ctrl) or 3 h after exercise, which was performed in the absence or presence of NAC (20 mM). Shown is relative expression of the full-size RyR1 with the average Ctrl set to 100% (n = 5–12 muscles). (C) Calpain activity in single-digit mouse FDB muscles before (Ctrl) and 30 min and 3 h after simulated HIIT exercise (n = 4–6). Positive and negative controls were obtained by adding fully active calpain and calpain inhibitor, respectively. Data are expressed as relative fluorescence units (RFU) divided by muscle wet weight (w.w.); the mean value in Ctrl was set to 1.0. All data are expressed as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001 with unpaired t test (A) or one-way ANOVA (B and C).

In the next series of experiments, FDB muscles were frozen 3 h after being exposed to the simulated HIIT exercise. Western blots from these displayed marked RyR1 fragmentation after the contractions when experiments were performed under control conditions (i.e., in standard Tyrode solution), whereas the fragmentation was completely blocked when muscles were exposed to the general antioxidant N-acetylcysteine (NAC; 20 mM) before and during the series of contractions (Fig. 3B). Thus, our results support a model where ROS induce modifications of RyR1 during the HIIT exercise, and these then trigger RyR1 fragmentation.

The distinct pattern of HIIT exercise-induced RyR1 fragmentation suggests that it involves an enzymatic cleavage process. Calpains are the likely candidates, and we measured the calpain activity in mouse FDB muscles before and after the simulated HIIT exercise. The results show a 3- to 4-fold higher calpain activity at 30 min and 3 h after the exercise (Fig. 3C).

The fragmentation of RyR1 after the simulated all-out cycling bouts might lead to dislocation of the protein. However, immunofluorescence RyR1 staining showed a similarly striated pattern before and 5 min and 3 h after the stimulation period, and this pattern was also observed when staining for the t-tubular voltage sensors DHPR (SI Appendix, Fig. S13). Noteworthy, the overall immunofluorescence staining for RyR1 was markedly decreased 3 h after the contractions, which probably reflects impaired antibody binding due to severe posttranslational modifications of RyR1 at this time point (cf Fig. 3B).

HIIT Induces a Prolonged Force Depression and an Increase in Resting [Ca²⁺]_i.

To assess the effect of RyR1 fragmentation on SR Ca²⁺ handling, we used mechanically dissected single mouse FDB fibers with intact tendons—i.e., a preparation that allows detailed measurements of [Ca²⁺]_i as well as the resulting force (15, 38). [Ca²⁺]_i during the initial 250 ms tetanic contraction of the simulated cycling bouts decreased with increasing number of bouts, being decreased by ∼35% at the start of the sixth bout (Fig. 4A); i.e., defects in SR Ca²⁺ release induced by the previous 30-s bouts of intense activation were not reversed during the 4-min rest periods between bouts.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Simulated HIIT exercise induces prolonged decrease in tetanic [Ca²⁺]_i and increase in resting [Ca²⁺]_i. (A, Upper) Representative records of [Ca²⁺]_i during 100-Hz tetanic stimulation trains evoked at the start of the first and sixth simulated cycling bout in a single FDB fiber. (A, Lower) Mean [Ca²⁺]_i in the initial tetanus of the six cycling bouts. (B) Force measured 5–120 min after the simulated HIIT exercise; data are expressed relative to the force before exercise, which in each fiber was set to 100% at both 40- and 120-Hz stimulation. (C) Tetanic [Ca²⁺]_i before (PRE) and 5–120 min after exercise. (D) The relation between tetanic force and [Ca²⁺]_i before (white circle; obtained by stimulating fibers at 15–150 Hz at 1-min intervals) and 120 min after exercise (black and red circles; data taken from B and C). (E) Resting [Ca²⁺]_i before and 5–120 min after exercise. (F) Representative [Ca²⁺]_i records and mean data obtained from 100-Hz tetanic stimulations in the presence of 5 mM caffeine produced before and 5 min after exercise. All data are expressed as mean ± SEM (n = 6–14 fibers). ***P < 0.001 vs. the first simulated cycling bout (A) or before exercise (B, C, and F) (one-way repeated measures ANOVA). *P < 0.05 in paired t test (F).

In agreement with the results of the above human experiments, isolated mouse FDB fibers entered a prolonged state of severely depressed force after the simulated HIIT exercise, especially at the lower (40 Hz) stimulation frequency (Fig. 4B); it should be noted that fusion occurs at higher frequencies in the mouse than in the human muscles, and 40 Hz stimulation of the mouse FDB fibers gave about the same proportion of the maximum force as 10 Hz for the human quadriceps muscle. Tetanic [Ca²⁺]_i also displayed a prolonged decrease after the contraction bouts, but in this case the decrease was larger at 120-Hz than at 40-Hz stimulation (Fig. 4C). These seemingly conflicting results are explained by the shape of the force–[Ca²⁺]_i relationship (5), and Fig. 4D shows that the force–[Ca²⁺]_i relations in 40- and 120-Hz contractions produced 120 min after the simulated HIIT exercise overlap with the force–[Ca²⁺]_i relationship under control conditions (obtained by producing 350-ms contractions at 15–150 Hz at 1-min interval in the same fibers before exercise).

Modified RyR1 can become leaky (6), which may result in an increase in resting [Ca²⁺]_i. Accordingly, the simulated HIIT exercise induced a prolonged ∼40% increase in resting [Ca²⁺]_i (Fig. 4E). Caffeine interacts with RyR1 to potentiate SR Ca²⁺ release (39). Fig. 4F shows [Ca²⁺]_i records from 100 Hz tetani produced in an FDB fiber exposed to caffeine (5 mM) before and after the simulated HIIT exercise; mean data show ∼30% lower [Ca²⁺]_i during caffeine tetani produced after the exercise (P < 0.05; Fig. 4F). These findings indicate that the simulated HIIT exercise induces RyR1 leakage, promoting Ca²⁺ fluxes from the SR toward the cytosol, which then results in increased resting [Ca²⁺]_i while tetanic [Ca²⁺]_i is reduced due to a decline in the releasable SR Ca²⁺ pool. It might also be noted that a prolonged increase in resting [Ca²⁺]_i stimulates mitochondrial biogenesis and can thereby improve muscle endurance (14⇓–16). Thus, the observed exercise-induced increase in resting [Ca²⁺]_i provides a tentative trigger for HIIT-induced mitochondrial biogenesis (21).