Role of parvalbumin in fatigue-induced changes in force and cytosolic calcium transients in intact single mouse myofibers

Abstract

One of the most important cytosolic Ca²⁺ buffers present in mouse fast-twitch myofibers, but not in human myofibers, is parvalbumin (PV). Previous work using conventional PV gene (PV) knockout (PV-KO) mice suggests that lifelong PV ablation increases fatigue resistance, possibly due to compensations in mitochondrial volume. In this work, PV ablation was induced only in adult mice (PV-KO), and contractile and cytosolic Ca²⁺ responses during fatigue were studied in isolated muscle and intact single myofibers. Results were compared with control littermates (PV-Ctr). We hypothesized that the reduced myofiber cytosolic Ca²⁺ buffering developed only in adult PV-KO mice leads to a larger cytosolic free Ca²⁺ concentration ([Ca²⁺]_c) during repetitive contractions, increasing myofiber fatigue resistance. Extensor digitorum longus (EDL) muscles from PV-KO mice had higher force in unfused stimulations (∼50%, P < 0.05) and slowed relaxation (∼46% higher relaxation time, P < 0.05) versus PV-Ctr, but muscle fatigue resistance or fatigue-induced changes in relaxation were not different between genotypes (P > 0.05). In intact single myofibers from flexor digitorum brevis (FDB) muscles, basal and tetanic [Ca²⁺]_c during fatiguing contractions were higher in PV-KO (P < 0.05), accompanied by a greater slowing in estimated sarcoplasmic reticulum (SR) Ca²⁺-pumping versus PV-Ctr myofibers (∼84% reduction, P < 0.05), but myofiber fatigue resistance was not different between genotypes (P > 0.05). Our results demonstrate that although the estimated SR Ca²⁺ uptake was accelerated in PV-KO, the total energy demand by the major energy consumers in myofibers, the cross-bridges, and SR Ca²⁺ ATPase were not altered enough to affect the energy supply for contractions, and therefore fatigue resistance remained unaffected.

NEW & NOTEWORTHY Parvalbumin (PV) is a cytosolic Ca²⁺ buffer that is present in mouse myofibers but not in human muscle. We show that inducible knockout of PV leads to increases in myofiber cytosolic free Ca²⁺ concentrations and slowing of Ca²⁺ pumping during fatigue versus control mice. However, PV ablation does not interfere with fatigue-induced slowing in relaxation or fatigue resistance. These data support the use of mouse muscle as a suitable model to investigate human muscle fatigue.

INTRODUCTION

During both intense exercise and repetitive contractions in skeletal myofibers, there is a progressive decrease in force development due to a variety of intracellular factors that result in peripheral muscle fatigue (1). One of the hallmarks of muscle fatigue is a progressive slowing of relaxation that directly interferes with the contractile efficiency of muscles during dynamic movements (2). In myofibers obtained from small rodents, fish, and amphibians, the relaxation rate is dependent on 1) the rate of cross-bridges detachment; 2) the off-rate of Ca²⁺ dissociation from troponin C (TnC); 3) the rate of cytosolic Ca²⁺ buffering by parvalbumin (PV); and 4) the velocity of Ca²⁺-pumping into the sarcoplasmic reticulum (SR) by the SR Ca²⁺-ATPase (SERCA) (3). Changes in any of the aforementioned factors (1–4) may directly impact force decay during relaxation. Particularly during fatigue, slowing in relaxation in mouse myofibers seems to be strongly influenced by a reduced cross-bridges detachment rate compared with a decrease in the SR Ca²⁺-pumping velocity (4). Although the contributions of cross-bridge cycling and SR Ca²⁺ uptake on myofiber relaxation during fatiguing contractions are still uncertain, it is well established that cytosolic Ca²⁺-handling in myofibers is a major factor contributing to muscle fatigue (5).

During fatiguing contractions, the changes in cytosolic free Ca²⁺ concentration ([Ca²⁺]_c; e.g., tetanic [Ca²⁺]_c, basal [Ca²⁺]_c, SR Ca²⁺ pumping) in fast-twitch myofibers have been well described (6–10). Specifically, fatigue-induced decrease in tetanic [Ca²⁺]_c has a major role in the development of fatigue in myofibers (11). It is important to note that a great majority of these studies were performed in small animals such as mice and amphibians whose fast-twitch myofibers contain abundant PV (12). However, human skeletal muscles contain very low, or undetectable, amounts of PV (13), which may affect the cytosolic Ca²⁺ kinetics responses during fatiguing contractions differently than in PV-present myofibers.

The role of PV ablation in fast-twitch myofibers from mice has been previously investigated in conventional PV knockout mice (cPV-KO) (14), in which PV is absent from the embryonic phase of development. Adult cPV-KO mice show slower muscle relaxation of fast-twitch muscles compared with wild-type mice, both in situ (14) and ex vivo (15) under nonfatiguing conditions. cPV-KO mice also show a decreased rate of [Ca²⁺]_c decay after a short period (20 ms) of electrical stimulation in cultured myofibers (14), although maximal SERCA enzyme activity measured in vitro in SR vesicles was unchanged (14). However, it is unclear whether SR Ca²⁺ pumping during relaxation within live myofibers is affected by the presence of PV when the activity of SERCA should be functioning at both maximal and submaximal rates ([Ca²⁺]_c ≥ 0.2 µM) (16). It is also not known whether the fatigue-induced slowing in relaxation (and in SR Ca²⁺ pumping) is impacted by the absence of PV.

Interestingly, fast-twitch muscles from cPV-KO mice show increased myofiber mitochondrial volume (17, 18), and greater fatigue resistance (17), but these changes were considered compensations to the lifelong adaptations of the PV absence in these muscles. However, it is not known how PV ablation in fully developed muscles would interfere with the contractile and [Ca²⁺]_c-handling changes that occur during fatiguing contractions. Furthermore, it is not known whether changes in cytosolic Ca²⁺ buffering interfere with the myofiber slowing in relaxation that occur during fatigue.

The purpose of this study was to investigate whether inducible PV ablation in adult mice alters the contractile and [Ca²⁺]_c changes that occur during fatigue, without the lifelong compensatory effects seen in cPV-KO mice (14, 15, 17, 18), which has never been tested before in intact single myofibers. We tested the hypothesis that reducing the cytosolic Ca²⁺-buffering in fast-twitch myofibers from mouse muscle, by knocking out PV expression in adult mice, increases tetanic [Ca²⁺]_c during contractions, leading to a greater myofiber fatigue resistance. The results obtained in the present investigation help elucidate the role of cytosolic Ca²⁺ handling changes during fatigue in mouse fast-twitch myofibers that have similar cytosolic Ca²⁺ buffering as occurs in human myofibers.

MATERIALS AND METHODS

Ethical Approval

All procedures were approved by the University of California, San Diego Institutional Animal Care and Use Committee (UCSD-IACUC, Protocol No. S00250) and comply with American Physiological Society’s Guiding Principles in the Care and Use of Vertebrate Animals in Research and Training. B6(Cg)-Pvalb^{tm1(cre/ERT2)Zjh}/J (The Jackson Laboratory, Bar Harbor, ME; Cat. No. 010777; PV-KO) mice (n = 11 mice total) or littermate controls without the Cre-recombinase gene (PV-Ctr; n = 11 mice total) were used in the present study. Only male mice were used in this study. Mice were kept on a 12-h:12-h light/dark cycle, and regular chow and water were provided ad libitum. PV-KO and PV-Ctr (17-wk-old) mice were treated for 5 consecutive days with tamoxifen [Tam, ip, 1 mg/mouse/day in 90% (v/v) olive oil and 10% (v/v) ethanol] and euthanized 3 wk after the first injection by an overdose of sodium pentobarbital (sleep-away; 150 mg/kg, ip). Death was confirmed by absence of movement, heartbeat, and response to toe pinching followed by rapid cervical dislocation to ensure euthanasia. Mice from each genotype (PV-Ctr and PV-KO) were randomized before tamoxifen treatment and tissue collection.

Ex Vivo Intact Muscle Contractility and Fatigue Resistance

In a group of mice (n = 5 mice each genotype), the extensor digitorum longus (EDL) and soleus muscles from both hindlimbs were carefully dissected from each mouse, mounted in experimental chambers (800MS, Danish Myo Technology) and constantly perfused with Tyrode’s solution (in mM: 121 NaCl, 5 KCl, 1.8 CaCl₂, 0.5 MgCl₂, 0.4 NaH₂PO₄, 24 NaHCO₃, 5.5 glucose, 0.1 EGTA, pH 7.4) bubbled with 95% O₂ and 5% CO₂ at 32°C during the whole experimental procedure. The length that promoted the maximal force (L₀) was determined using single twitches (0.5 ms square-waved pulses, 16 V, S88X stimulator; Grass Technologies, Quincy, MA) followed by a 10-min resting under constant perfusion. Muscles were subjected to a force-frequency protocol (1–250 Hz, 350 ms trains for EDL, or 500 ms trains for soleus, once every 100 s) and then to a fatigue protocol. The fatigue protocol consisted of a bout of repetitive contractions at 120 Hz for EDL and 80 Hz for soleus at a rate of 0.15 trains per second (tps), and the train rate was increased to 0.19, 0.25, 0.33, 0.43, 0.57, 0.75, and 1 tps every minute until peak force was 30% of the initial contraction (i.e., task failure). After the experimental protocol was performed, muscles’ L₀ were measured and muscles were blotted dry and weighed. Muscle cross-sectional area (CSA) was calculated by dividing the muscle mass with the product of L₀ and the density of the muscle (1.06 mg/mm³). Force measurements of muscles obtained from both hindlimbs from each mouse were averaged, normalized by the CSA and reported as kPa (10). Force measurements obtained during the force-frequency protocol were fitted by a sigmoid nonlinear regression using the Eq. 1:

$$\text{P }=\left({\text{P}}_0•{\text{ F}}^{\text{n}\text{H}}\right)/\left(F_{50}{}^{\text{n}\text{H}}+{\text{ F}}^{\text{n}\text{H}}\right).$$ (1)

From Eq. 1, P is the force evoked by different frequencies of stimulation, P₀ is maximum tetanic force, F is the frequency of pulse stimulation, F₅₀ is the midpoint of the curve (in Hz), and nH is the Hill coefficient, which is correlated to the steepness of the curve.

Force and Free Cytosolic Ca²⁺ Transients during Contractions in Intact Single Myofibers

Myofiber isolation.

In another group of mice (n = 6 mice for each genotype), the flexor digitorum brevis (FDB) muscles from both posterior feet were quickly removed, and a single myofiber from each mouse was mechanically isolated with tendons intact. Thus, one myofiber was tested per mouse. Each myofiber was pressure injected using a micropipette filled with FURA-2 (Life technologies, Carlsbad, CA; 12 mM, diluted in 150 mM KCl and 10 mM HEPES pH 7.0) (10). Following myofiber isolation and FURA-2 loading, platinum clips were attached to the tendons, and each myofiber was transferred to an intact muscle fiber system (model 1500 A mounted with a 400 A force transducer, Aurora Scientific Inc., Aurora, ON, Canada), placed on the stage of an inverted microscope (Nikon Eclipse Ti‐S with a ×40 long distance Fluor objective), and superfused with Tyrode’s solution constantly bubbled with 21% O₂ and 5% CO₂ at 22°C during the rest of the experimental procedures.

Myofiber electrical stimulations and force measurements.

Intact single myofibers were electrically stimulated to evoke contractions using a Grass S88X stimulator (350 ms trains, 0.5 ms square-wave pulses, 8 V), and force data were acquired and analyzed as described previously (10). Each myofiber was stimulated by 100 Hz pulse stimulations at increasing lengths to determine L₀, and rested for 10 min, followed by a force-frequency protocol (1–150 Hz, once every 100 s). Immediately following the last train of the force-frequency protocol, the chamber superfusion was switched to a Tyrode’s solution containing 10 mM caffeine to evoke a single 120 Hz train, and then switched back to a caffeine-free Tyrode’s to washout the caffeine for 20 min. A fatigue protocol was initiated after caffeine was washed out. The fatigue protocol in intact single myofibers consisted of a bout of repetitive contractions at 100 Hz with an increasing train rate starting at 0.25 tps, the rate of which was increased every 2 min by 20% (0.3, 0.36, 0.43, 0.52, 0.62, 0.75, and 0.9 tps). The fatigue protocol was interrupted when peak force was 50% of the initial contraction (i.e., task failure). Cross-sectional area (CSA) of each myofiber was determined from the myofiber diameter measured using a ×40 magnification, which was not statistically different between the two genotypes (42 ± 10 µm for PV-Ctr and 37 ± 7 µm for PV-KO; P > 0.05, n = 6 mice each strain). Force (in mN) was normalized to the CSA (in mm²) and reported as kPa. Total force-time integral was obtained by the area under the curve of force development and relaxation of each contraction during the fatigue protocol and reported as kPa·s. Force measurements obtained from myofibers during the force-frequency protocol were fitted using Eq. 1.

Free cytosolic Ca²⁺assessment during contractions.

Free cytosolic calcium concentration ([Ca²⁺]_c) changes were obtained by epifluorescence spectroscopy (10). Fluorescence excitation ratio (340/380 nm; R) was converted to [Ca²⁺]_c according to Eq. 2.

$${\left[{\text{Ca}}^{\text{2}+}\right]}_{\text{c}}={\text{K}}_{\text{D}}•\beta •\left[\left(\text{R}-{\text{R}}_{\text{m}\text{i}\text{n}}\right)/\left({\text{R}}_{\text{m}\text{a}\text{x}}-\text{R}\right)\right].$$ (2)

From Eq. 2, K_D is the dissociation constant for Ca²⁺-FURA-2, which was set to 224 nM (19), β is the fluorescence ratio between high and no [Ca²⁺]_c at 380 nm, and R_min and R_max are the fluorescence ratios at no [Ca²⁺]_c and maximum tetanic [Ca²⁺]_c, respectively. R_min was set to 0.24 (10). R_max was determined for each of the contracting myofibers (i.e., from PV-KO and PV-Ctr mice) and set to the 136% of the tetanic [Ca²⁺]_c obtained when contractions were evoked at 120 Hz in the presence of 10 mM caffeine (as described above). R_max was not different between myofibers isolated from PV-KO and PV-Ctr mice (4.93 ± 0.92 vs. 4.84 ± 0.20, respectively, P > 0.05). β was determined (7.64 ± 1.76 for PV-KO mice and 5.86 ± 0.32 for PV-Ctr mice) for each of the contracting myofibers as previously described (20).

The contraction-induced fluorescence ratio was calculated by averaging the fluorescence ratio signal in the last 100 ms of electrical stimulation and subsequently used to determine tetanic [Ca²⁺]_c. [Ca²⁺]_c before each contraction (i.e., basal [Ca²⁺]_c) was calculated by averaging the signal in the 100 ms preceding each stimulation. To determine myofilament Ca²⁺ sensitivity, force data at each tetanic [Ca²⁺]_c during the force-frequency protocols were fitted with a sigmoidal equation described in Eq. 3:

$$\text{P }=\left({\text{P}}_0•{\left[{\text{Ca}}^{\text{2}+}\right]}_{\text{c}}{}^{\text{nH}}\right)/\left(\text{C}{\text{a}}^{2+}{{}_{50}}^{\text{nH}}+{\left[{\text{Ca}}^{\text{2}+}\right]}_{\text{c}}{}^{\text{nH}}\right).$$ (3)

From Eq. 3, P is the force produced at different tetanic [Ca²⁺]_c, P₀ is the maximal tetanic force, Ca^2 +₅₀ is the midpoint of the force-[Ca²⁺]_c curve, and nH is the Hill coefficient.

To determine whether fatiguing contractions altered myofilament Ca²⁺ sensitivity, force data 40 s after the beginning of the fatigue protocol were plotted at each tetanic [Ca²⁺]_c and fitted with Eq. 3.

Determination of [Ca²⁺]_c pumping rate by the sarcoplasmic reticulum.

To estimate SR Ca²⁺ pumping during the contractile bouts, the elevated long “tail” of [Ca²⁺]_c decline following contractions (from 0.2 to 10.0 s following the stimulation period) was used to determine the rate of [Ca²⁺]_c decline (d[Ca²⁺]_c/dt) (10). Therefore, a SERCA Ca²⁺ pumping curve was obtained by plotting −d[Ca²⁺]_c/dt during the “tail” of [Ca²⁺]_c decay versus [Ca²⁺]_c (Eq. 4).

$$-\text{d}{\left[{\text{Ca}}^{\text{2}+}\right]}_{\text{c}}/\text{d}t=A•{\left[{\text{Ca}}^{\text{2}+}\right]}_{\text{c}}{}^N-L.$$ (4)

From Eq. 4, A, N, and L are three adjustable parameters representing the rate of SR Ca²⁺ pumping (in µM^N−1/s), the power function, and the SR Ca²⁺ leak (in µM/s), respectively. To directly compare SR Ca²⁺ pumping (in µM⁻³/s) in myofibers between the PV-KO and PV-Ctr mice at the first contraction and at the task failure, the curves were fitted with N and L fixed at 4 and 12, respectively. These values were based on mean values of 4.1 ± 1.3 for N and 12 ± 3 for L obtained in individual experiments (n = 10 mice), with a maximum increase in least-square error of 15% (10).

Immunoblotting of Muscle Homogenates

The EDL muscles from both hindlimbs, from mice which the FDB muscles were used to perform the intact single myofiber experiments, were dissected, frozen in liquid N₂, pul-verized and homogenized in radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl, 1.0% IGEPAL CA-630, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8.0, Sigma Aldrich, Cat. No. R0278) supplemented with protease and phosphatase inhibitor cocktails (100 µL/10 mL; Sigma Aldrich, Cat. Nos. P8340 and P0044, respectively), 1 mM K₂EGTA, 1 mM Na₂EDTA, 5 mM NaF, and 5 mM Na₃VO₄. Homogenates were centrifuged for 10 min at 10,000 g at 4°C. Protein concentration of the supernatant was determined by the Pierce BCA protein assay, and proteins were solubilized in Laemmli sample buffer [4% SDS, 20% glycerol, 0.004% bromophenol blue, 125 mM Tris HCl, pH 6.8, 10% (v/v) 2-mercaptoethanol]. Proteins (7.5 µg each lane) were loaded in 4%–20% Mini-PROTEAN TGX gels (Bio-Rad, Cat. No. 456–1093), and electrophoresis was carried out at 100 V. Proteins were transferred to an Immobilon-FL PVDF membrane (EMD Millipore, Darmstadt, Germany) for 30 min at 15 V using a Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad, Hercules, CA) and blocked for 2 h with an Odyssey Blocking Buffer in PBS (Li-COR; Lincoln, NE). Protein was detected by incubating membranes with primary antibodies for dihydropyridine receptor (DHPR, monoclonal antibody, Cat. No. MA3-920, Thermo-Fisher Scientific, RRID: AB_2069575, dilution: 1:1,000), SERCA1 (monoclonal antibody, Cat. No. MA3-912, Thermo Fisher Scientific, RRID: AB_2061281, dilution: 1:1,000), fast skeletal muscle troponin I (TNNI2, polyclonal antibody, Cat. No. LS-B7284, LSBio, RRID: AB_11180696, dilution: 1:1,000), GAPDH (polyclonal antibody, Cat. No. sc-25778, Santa Cruz Biotechnology, RRID: AB_10167668, dilution: 1:1,000), parvalbumin (polyclonal antibody, Cat. No. PA1-18055, Thermo-Fisher Scientific, RRID: AB_2173901, dilution: 1:1,000), and Mitoprofile OXPHOS rodent antibody cocktail (monoclonal antibody cocktail, Cat. No. ab110413, Abcam, RRID: AB_2629281, dilution: 1:1,000), followed by incubation with near-infrared fluorescent (680 or 800 nm) anti-mouse (IRDye 680RD Goat anti-Mouse IgG, Cat. No. 926–68070, RRID: AB_10956588) or anti-rabbit (IRDye 800CW Goat anti-Rabbit IgG, Cat. No. 926–32211, RRID: AB_621843) secondary antibodies (Li-Cor; Lincoln, NE, dilution: 1:5,000 each). Membranes were imaged using an infrared laser-based scanner (Li-COR Odyssey; Lincoln, NE). Density of the bands was analyzed with Image J software (https://imagej.nih.gov/ij/) and individual band density was divided by the density of the GAPDH band, used as a loading control for each sample. Density of bands was normalized by the average band density obtained in PV-Ctr mice. Molecular weight of each protein was estimated by comparing each band migration with the migration of bands from Precision Plus Protein Standards (Bio-Rad Laboratories, Hercules, CA) to verify antibody specificity.

Statistics

The experimental results are presented as means ± standard error (SE). For multiple comparisons, a two-way ANOVA followed by Bonferroni test was used, as indicated. For comparisons between two means, unpaired Student’s t test was used as indicated. All the analyses were conducted using GraphPad Prism version 4.00 for Windows (San Diego, California). Statistical significance was accepted when P < 0.05.

RESULTS

Effects of Tamoxifen Treatment in Adult PV-KO and in PV-Ctr Mice on EDL and Soleus Muscle Contractility

Adult PV-KO and PV-Ctr mice were treated for 5 consecutive days with tamoxifen (see materials and methods for details), and EDL, soleus, and FDB muscles were dissected 17 days after the last injection. Homogenates from EDL muscles showed complete ablation of the protein levels of PV in PV-KO mice, but not in PV-Ctr mice (Fig. 1A). Although PV was ablated (Bonferroni posttest, P < 0.001 vs. PV-Ctr), there were no statistical differences detected in the posttest (P > 0.05) in the amount of DHPR, SERCA1, or TNNI2 (Fig. 1B), or mitochondrial complexes (Fig. 1C) between genotypes.

Figure 1.

Tamoxifen-induced knockout of the PV gene in adult mice on myofiber Ca²⁺-handling and mitochondrial protein levels. A: immunoblotting of EDL muscle homogenates 3 wk after tamoxifen treatment in PV-Ctr and in PV-KO mice. Each lane corresponds to a separate mouse and was loaded with a total of 15 µg of protein. B: histograms of the band density of either PV or SERCA1a or DHPR or TNNI2, normalized by the density of GAPDH [n = 4 mice for each genotype; P < 0.01 (genotype), two-way ANOVA, *P < 0.001 vs. PV-Ctr, Bonferroni post-test]. C: histograms of the band density of Mitoprofile bands normalized by the density of GAPDH [n = 4 mice for each genotype; P > 0.05 (genotype), two-way ANOVA]. Data are shown as means ± SE. DHPR, dihydropyridine receptor; EDL, extensor digitorum longus; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; PV, parvalbumin; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout; SERCA1a, isoform 1a of SR Ca²⁺ATPase; TNNI2, fast skeletal muscle type troponin I.

In EDL muscles from PV-KO, trace recordings showed a higher peak force evoked by a single twitch and by a pulse-frequency that produces submaximal tetanic force (i.e., 50 Hz) compared with PV-Ctr (Fig. 2A, top). There was a clear difference in the decay of force at a high pulse frequency of stimulation (i.e., 200 Hz; Fig. 2A, top). In soleus muscles, which contain little amounts of PV compared with EDL muscle (17, 21–23), there were no differences in the trace recordings for submaximal force or relaxation velocities between the two mouse genotypes (Fig. 2A, bottom). As shown in Fig. 2B, the force-frequency curve was left-shifted in EDL muscles from PV-KO compared with PV-Ctr mice, demonstrated by a significantly lower F₅₀ in EDL muscles from PV-KO mice (Table 1, P < 0.05 vs. PV-Ctr mice). However, maximal tetanic force (P₀) was not different between the two genotypes of mice in EDL muscles (Table 1, P > 0.05). Conversely, in soleus muscles, there were no differences in the force-frequency relationship (Fig. 2C) and in the F₅₀ (Table 1) between genotypes.

Figure 2.

Effects of inducible PV ablation on isolated EDL and soleus muscles contractility in adult mice. A: representative trace recordings of force evoked by different pulse frequencies (1, 50, and 200 Hz) in EDL (top) and in soleus (bottom) muscles from PV-Ctr (black traces) and PV-KO (gray traces) mice. B and C: peak force evoked by different pulse frequencies of stimulation in EDL [B, P < 0.001 (genotype), two-way ANOVA, *P < 0.01 vs. PV-Ctr, Bonferroni posttest] and soleus (C, P > 0.05, two-way ANOVA) muscles from PV-Ctr (closed circles; n = 5) and PV-KO (open circles; n = 5) mice. Force traces (A) and peak isometric forces (B and C) were normalized by the maximum tetanic force produced by each muscle (% P₀). Data are shown as means ± SE. EDL, extensor digitorum longus; PV, parvalbumin; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout.

Table 1.

Morphometrics and contractile properties of EDL and soleus muscles

	EDL Muscle		Soleus Muscle
	PV-Ctr	PV-KO	PV-Ctr	PV-KO
Muscle dimensions
Muscle weight, mg	13.6 ± 0.9	14.4 ± 1.3	10.9 ± 0.4†	11.0 ± 1.5†
L0, mm	13.6 ± 0.5	13.0 ± 0.5	11.5 ± 0.9†	12.1 ± 0.9†
CSA, mm2	0.96 ± 0.08	1.05 ± 0.07	0.90 ± 0.10†	0.85 ± 0.08†
Single pulse, twitch
CT, ms	13.5 ± 1.4	18.0 ± 2.0*	24.5 ± 2.7†	24.0 ± 3.8†
½ RT, ms	21.6 ± 4.3	24.2 ± 1.4	30.9 ± 4.7†	33.4 ± 4.6†
Pt, kPa	71 ± 12	83 ± 6	39 ± 10†	45 ± 12†
Pt/P0, %	14 ± 2	19 ± 1*	15 ± 3	15 ± 2
Tetanic tension
P0, kPa	515 ± 45	453 ± 47	270 ± 37†	292 ± 48†
F50, Hz	69 ± 5	43 ± 4*	36 ± 5†	35 ± 4†
½ CT, ms	21 ± 4	16 ± 4	42 ± 8†	38 ± 4†
Peak +dP/dt, kPa/s	14,834 ± 748	14,070 ± 1,595	3,766 ± 578†	4,385 ± 921†
½ RT, ms	20 ± 2	31 ± 1*	38 ± 5†	37 ± 3†
80% RT, ms	28 ± 3	41 ± 1*	63 ± 7†	57 ± 8†
Peak −dP/dt, kPa/s	28,396 ± 6,206	15,642 ± 2,596*	5,275 ± 744†	6,701 ± 1,061†
Time to task failure, s	169 ± 25	162 ± 8	309 ± 25†	326 ± 11†

Data are shown as means ± SE, n = 5 mice each. Two-way ANOVA, *P < 0.05 vs. PV-Ctr mice for either EDL or soleus, †P < 0.05 vs. EDL for either PV-Ctr or PV-KO, Bonferroni posttest. CSA, muscle cross-sectional area; CT, contraction time to peak force; ½ CT, half-contraction time to peak force; EDL, extensor digitorum longus; F₅₀, midpoint of the force-frequency relationship; L₀, optimal muscle length; peak +dP/dt, peak rate of force development; peak –dP/dt, peak rate of relaxation; P_t, twitch force normalized by the CSA; P₀, maximal tetanic force normalized by the CSA; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout; 80% RT, time to relax 80% of force during relaxation; ½ RT, half-relaxation time.

The relaxation rate of EDL muscles from PV-KO mice was ∼45% lower than in EDL muscles from PV-Ctr; in addition, the ½RT and 80%RT were ∼45% and ∼50% higher, respectively, in PV-KO compared with PV-Ctr mice (Table 1, P < 0.05). However, no differences in relaxation rate or relaxation times were present in soleus muscles between the two genotypes of mice (Table 1).

Effects of Adult Parvalbumin Ablation on EDL and Soleus Muscles Fatigue Resistance

During repetitive contractions to elicit fatigue, the acute (3 wk) ablation of PV in adult PV-KO mice did not statistically alter the progressive decay in peak force during contractions up to the designated task failure (i.e., 70% decrease in initial peak force) compared with PV-Ctr mice in EDL muscle for either specific peak force (Fig. 3A) or peak force normalized by the initial contraction (Fig. 3B). In soleus muscle, although the progressive decay in peak force during contractions was statistically different between genotypes (P < 0.001, Fig. 3, D and E), the post hoc analysis was not able to find the differences in the time course of the force decay for either specific peak force (Fig. 3D) or peak force normalized by the initial contraction (Fig. 3E, P > 0.05). In addition, the time to task failure of soleus muscles (Table 1) was also not statistically different between genotypes.

Figure 3.

Effects of inducible PV ablation on fatigue resistance and relaxation of isolated EDL and soleus muscles. Peak force during fatigue-inducing contractions of EDL [A and B, P > 0.05 (genotype), two-way ANOVA] and soleus [D and E, P < 0.01 (genotype), two-way ANOVA] muscles from PV-Ctr (closed symbols; n = 5) and PV-KO (open symbols; n = 4) mice. Peak specific forces (A and D) during fatiguing contractions were normalized by the peak force at initial contraction by each muscle (%Initial, B and E). C and F: time to relax 80% of peak force after the last pulse stimulation (relaxation time) during fatiguing contractions in EDL [C, P < 0.0001 (fatigue), P < 0.01 (genotype), two-way ANOVA, *P < 0.05 vs. PV-Ctr, Bonferroni posttest] and soleus [F, P < 0.0001 (fatigue), P > 0.05 (genotype), two-way ANOVA] muscles from PV-Ctr (closed circles; n = 5) and PV-KO (open circles; n = 5) mice. Data are shown as means ± SE. EDL, extensor digitorum longus; PV, parvalbumin; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout.

In EDL muscle, relaxation was significantly slowed (P < 0.001) at time to task failure in both PV-Ctr mice (24 ± 5 vs. 52 ± 8 ms, for the first contraction vs. contraction at time to task failure, respectively; Fig. 3C) and in PV-KO mice (34 ± 3 ms vs. 80 ± 25 ms, for the first contraction vs. contraction at time to task failure, respectively; Fig. 3C). Post hoc analysis detected a genotype difference in relaxation time at the time to task failure (P < 0.05). However, the increase in relaxation time was proportionally similar between genotypes and not statistically different (135 ± 66% vs. 120 ± 26% for PV-KO vs. PV-Ctr mice; P > 0.05). In soleus muscles, relaxation was also slowed with fatiguing contractions for both PV-KO and PV-Ctr mice (P < 0.0001), with no difference in relaxation times at the time to task failure between genotypes of mice (124 ± 28 ms and 119 ± 24 ms, respectively; Fig. 3F).

Contractility and Cytosolic Ca²⁺ Handling in Intact Single Myofibers from Adult Mice after Parvalbumin Gene Knockout

In intact single myofibers isolated from FDB muscle, force and [Ca²⁺]_c records show that PV-KO mice developed higher tetanic force and [Ca²⁺]_c evoked by submaximal pulse frequency (i.e., 30 Hz) compared with PV-Ctr (Fig. 4A). At a near-maximal pulse-frequency (i.e., 100 Hz), [Ca²⁺]_c was clearly higher in PV-KO mice versus PV-Ctr with a faster force development during the contractions phase and a slower force decay during relaxation (Fig. 4B). The force-frequency relationship was left-shifted in PV-KO versus PV-Ctr mice, showing a higher force at submaximal frequencies of pulse stimulation [P < 0.05 (genotype), P < 0.001 Bonferroni posttest, Fig. 4C]. This was confirmed by a statistically lower F₅₀ in myofibers from PV-KO mice (26 ± 8 Hz) compared with PV-Ctr mice (39 ± 9 Hz; P < 0.05). Force evoked by 120 Hz with 10 mM caffeine was not statistically different between genotypes (319 ± 20 kPa for PV-Ctr vs. 414 ± 69 kPa for PV-KO, P = 0.217).

Figure 4.

Force responses and [Ca²⁺]_c transients during contractions of intact single FDB myofibers from PV-Ctr and PV-KO mice. A and B: representative trace recordings of force (left) and [Ca²⁺]_c (right) from PV-Ctr (black traces) and PV-KO (gray traces) myofibers evoked either by 30 Hz (A) or 100 Hz (B) pulse frequencies. C: peak force evoked by different pulse frequencies of stimulation, normalized by the maximal tetanic force (% P₀). P < 0.05 (genotype), two-way ANOVA, *P < 0.001 vs. PV-Ctr, Bonferroni posttest. D: tetanic [Ca²⁺]_c at increasing pulse-frequencies (open and closed circles), and at 120 Hz in presence of 10 mM caffeine (open and closed triangles). P < 0.001 (genotype), two-way ANOVA, *P < 0.01, Bonferroni posttest. E: peak force responses at each tetanic [Ca²⁺]_c evoked by different pulse frequencies of stimulation. F: representative trace recordings of [Ca²⁺]_c following a 100 Hz train (starting at the end of stimulation period). G: the [Ca²⁺]_c dependence of rate of [Ca²⁺]_c decay (−d[Ca²⁺]_c/dt) during the “tail” of [Ca²⁺]_c decay following 100 Hz contractions. H: rate of SR Ca²⁺ pumping obtained from data in F. PV-Ctr (closed circles; n = 6) and PV-KO (open circles; n = 5) mice. *P < 0.05 (genotype), two-way ANOVA, Bonferroni posttest. Data are shown as means ± SE. FDB, flexor digitorum brevis; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout.

Resting [Ca²⁺]_c was not statistically different between the two genotypes of mice (30 ± 8 nM vs. 44 ± 17 nM for PV-Ctr vs. PV-KO mice, respectively; P > 0.05). However, tetanic [Ca²⁺]_c during contractions was statistically higher in PV-KO mice (P < 0.0001, genotype) when contractions were evoked between 30 and 150 Hz (P < 0.01, Bonferroni posttest, Fig. 4D). When contractions were evoked by 120 Hz in the presence of 10 mM caffeine, PV-KO myofibers also showed greater tetanic [Ca²⁺]_c (3.4 ± 1.3 µM) compared with PV-Ctr myofibers (2.4 ± 0.5 µM; P < 0.01 Bonferroni posttest; Fig. 4D). Although force and [Ca²⁺]_c during submaximal tetanic contractions were higher in PV-KO compared with PV-Ctr, when force was plotted as a function of tetanic [Ca²⁺]_c, and the plots for each genotype were superimposed, the curves were clearly not different (Fig. 4E). The midpoints of the force-[Ca²⁺]_c plots (Ca^2 +₅₀) were not statistically different between the two genotypes of mice (579 ± 172 nM vs. 635 ± 156 nM for PV-Ctr vs. PV-KO mice, respectively; P > 0.05).

Trace recordings of the decay in [Ca²⁺]_c following the stimulation period showed a faster decay in [Ca²⁺]_c at the “tail” of [Ca²⁺]_c decay ([Ca²⁺]_c < 150 nM) in myofibers from PV-KO versus PV-Ctr (Fig. 4F). The velocities of [Ca²⁺]_c decay (−d[Ca²⁺]_c/dt) during the “tail” of [Ca²⁺]_c following the stimulation period (see materials and methods for details) were higher in PV-KO myofibers than in PV-Ctr myofibers, which produced a left-shift in the −d[Ca²⁺]_c/dt−[Ca²⁺]_c curve (Fig. 4G). SR Ca²⁺ pumping rate, which was estimated by fitting the −d[Ca²⁺]_c/dt−[Ca²⁺]_c plots with Eq. 4 (see materials and methods), was significantly higher in PV-KO versus PV-Ctr (4230 ± 1747 vs. 1479 ± 996 µM⁻³/s, respectively, Fig. 4H, P < 0.05).

Fatigue-Induced Changes in Force and Ca²⁺-Handling in Intact Single Myofibers from Adult Mice after Parvalbumin Gene Knockout

During fatiguing contractions in FDB myofibers, peak-specific force (in kPa) was gradually decreased in myofibers from both genotypes of mice (Fig. 5, A and B). Although there was a statistical difference on specific force between genotypes (P < 0.05), the post hoc test was not able to detect statistical differences. Also, the total force-time integrals of fatigue protocols were not statistically different between genotypes (P > 0.05, Fig. 5C). When changes in peak force were normalized by the initial contraction, there were no differences between the genotypes (Fig. 5B). Time to task failure in myofibers (i.e., time to decrease tetanic force to 50% of the initial contraction) was also not statistically different between genotypes (229 ± 81 vs. 239 ± 59 s for PV-Ctr vs. PV-KO mice, respectively; P > 0.05). However, basal [Ca²⁺]_c between contractions was significantly higher in PV-KO versus PV-Ctr mice after 140 s of contractions (P < 0.001, genotype, P < 0.05, Bonferroni posttest, Fig. 5D). Furthermore, tetanic [Ca²⁺]_c was statistically higher in PV-KO versus PV-Ctr mice during fatiguing contractions (P < 0.001, genotype, P > 0.05 Bonferroni posttest, Fig. 5E). Although both basal and tetanic [Ca²⁺]_c were different between genotypes, the increase in Ca^2 +₅₀ that occurs during fatiguing contractions (i.e., difference in Ca^2 +₅₀ between before and during fatiguing contractions) was not statistically different between genotypes (126 ± 71 vs. 225 ± 185 nM of increase in Ca^2 +₅₀ for PV-Ctr vs. PV-KO myofibers; P > 0.05).

Figure 5.

Isometric force development and [Ca²⁺]_c responses during a fatigue-inducing contraction protocol in intact single FDB myofibers from PV-Ctr and PV-KO mice. A: peak-specific force during fatiguing contractions. P < 0.05 (genotype), two-way ANOVA. B: peak force normalized by the peak force at initial contraction by each myofiber during fatiguing contractions. C: total force-time integral during fatiguing contractions. P > 0.05, unpaired Student’s t test. D: basal [Ca²⁺]_c data, obtained from the 100 ms period before each contraction, up to the time of task failure. P < 0.001 (genotype), two-way ANOVA, *P < 0.05 vs. PV-Ctr, Bonferroni posttest. E: tetanic [Ca²⁺]_c responses up to the time of task failure. P < 0.001 (genotype), two-way ANOVA. PV-Ctr (closed circles; n = 6) and PV-KO (open circles; n = 6) mice. Data are shown as means ± SE. FDB, flexor digitorum brevis; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout.

Figure 6, A and B shows representative trace recordings of force and [Ca²⁺]_c, respectively, at the first contraction and at the time of task failure for each genotype. Force and [Ca²⁺]_c traces clearly show that relaxation and [Ca²⁺]_c decay were both slowed at the time of task failure in myofibers from both genotypes. Although relaxation was statistically and similarly slowed in myofibers from both genotypes at the time of task failure (P < 0.001, fatigue, P > 0.05, genotype, Fig. 6C), there was no statistical difference in the increase in relaxation time by fatiguing contractions between genotypes (85 ± 42 vs. 115 ± 86 ms for PV-Ctr and PV-KO mice, respectively, P > 0.05).

Figure 6.

Relaxation and SR Ca²⁺ pumping changes during fatiguing contractions in single FDB myofibers from PV-Ctr and PV-KO mice. A and B: representative trace recordings of force (A) and [Ca²⁺]_c (B) from PV-Ctr (black traces) and PV-KO (gray traces) myofibers at the first contraction and at the time of task failure. C: relaxation times at the initial contraction during the fatigue-inducing contractions protocol and at the time of task failure. P < 0.001 (fatigue), P > 0.05 (genotype), two-way ANOVA. D: representative trace recordings of [Ca²⁺]_c decay following the stimulation period from PV-Ctr (black traces) and PV-KO (gray traces) myofibers at the first contraction and at the time of task failure. E and F: the [Ca²⁺]_c dependence of rate of [Ca²⁺]_c decay (−d[Ca²⁺]_c/dt) during the “tail” of [Ca²⁺]_c decay in myofibers after the initial contraction (closed symbols) and after the contraction at the time of task failure (open symbols) are illustrated for PV-Ctr (E) and PV-KO (F) mice. G: rate of SR Ca²⁺ pumping after the initial contraction and the time of task failure were obtained from data in E and F and shown as PV-Ctr (n = 6) and as PV-KO (n = 6). P < 0.05 (genotype), two-way ANOVA, *P < 0.01 vs. PV-Ctr, Bonferroni posttest. H: rate of SR Ca²⁺ pumping at the time of task failure normalized by the initial contraction. P < 0.005 (genotype), two-way ANOVA, *P < 0.05 vs. PV-Ctr, Bonferroni posttest. Data are shown as means ± SE. FDB, flexor digitorum brevis; PV-Ctr, parvalbumin control littermates; PV-KO, parvalbumin knockout; SR, sarcoplasmic reticulum.

Interestingly, the “tail” of [Ca²⁺]_c at the time of task failure was visually higher in PV-KO versus PV-Ctr myofibers (Fig. 6D). Fatiguing contractions produced a right-shift in the −d[Ca²⁺]_c/dt−[Ca²⁺]_c plots in myofibers from both genotypes of mice (Fig. 6, E and F), which is an indication that the estimated SR Ca²⁺ pumping is slowed during fatiguing contractions (9). However, the −d[Ca²⁺]_c/dt−[Ca²⁺]_c plots were clearly more right-shifted in PV-KO myofibers compared with PV-Ctr (Fig. 6F). Although estimated SR Ca²⁺ pumping rate was statistically decreased at the time of task failure in both groups of myofibers (P < 0.001, fatigue, Fig. 6G), myofibers from PV-KO mice showed a significantly greater decay in the estimated SR Ca²⁺ pumping rate at the time of task failure (P < 0.005, genotype, Bonferroni posttest, Fig. 6H).

DISCUSSION

Muscle fatigue has been studied for many decades by using different animal models assuming that the mechanisms of muscle fatigue would be comparable to human muscles. However, muscle from small rodents and amphibians contain large amounts of PV, a known cytosolic Ca²⁺-buffer involved in binding Ca²⁺ during the contraction and relaxation phases (2, 24), whereas human muscle does not contain PV (13). Since the role of cytosolic Ca²⁺-handling on the decay of force production during fatigue is well known, it is important to delineate the role that PV may have in affecting fatigue development. We hypothesized that the cytosolic Ca²⁺-buffering capacity of PV in fast-twitch myofibers lowers tetanic [Ca²⁺]_c available for the contractile sites during fatiguing contractions. Therefore, absence of PV would increase tetanic [Ca²⁺]_c during repetitive contractions, thereby increasing fatigue resistance.

The major findings of the present investigation were that short-term (3 wk) inducible PV ablation (PV-KO) in adult mice did not alter fatigue resistance in either isolated whole EDL muscle or intact single FDB myofibers. However, in FDB myofibers, the initial increase in tetanic [Ca²⁺]_c and late basal [Ca²⁺]_c during fatiguing contractions were higher in PV-KO, combined with a greater slowing in SR Ca²⁺ pumping at the time of task failure, when compared with PV-Ctr mice. These data suggest that PV in fast-twitch myofibers, as detected in mouse muscle but not in human muscle, minimizes the tetanic [Ca²⁺]_c changes that occur during repetitive contractions. Conversely, the presence of PV does not interfere with the decrease in force, the slowing in relaxation during fatiguing contractions, or the final tetanic [Ca²⁺]_c at fatigue, and thereby does not directly interfere with myofiber fatigue resistance.

Role of Parvalbumin on Muscle Contraction and Relaxation under Nonfatiguing Conditions

In small animals, such as rodents, amphibians, and fish, fast-twitch myofibers contain large amounts of the Ca²⁺-buffering protein PV, which is soluble in the cytosol (25, 26). PV contains two high-affinity binding sites for divalent cations, which can bind Mg²⁺ (∼10⁴/M) and with much higher affinity for Ca²⁺ (∼10⁸/M) (24, 27). During rest, cytosolic [Ca²⁺] is low (<0.1 µM) and [Mg²⁺] is high (>1 mM), so Mg²⁺ is bound to PV (28). Upon membrane depolarization, the free Ca²⁺ present within the SR rapidly diffuses into the cytosol, almost instantly raising [Ca²⁺]_c from nanomolar to micromolar levels (2). Initially, Ca²⁺ binds to troponin C (TnC), triggering cross-bridge cycling (29), and to SERCA, activating Mg^2 + ATPase-dependent SR Ca²⁺ pumping (9, 30). Ca²⁺-PV binding is slower than for TnC or SERCA since it is limited by the dissociation of Mg²⁺ from PV (24). When stimulation ends, Ca²⁺ is transported back into the SR, leading to a fast Ca²⁺ dissociation from TnC, myofiber relaxation, and decay of [Ca²⁺]_c back to its resting levels (<0.1 µM) (2). During short contractions (<2 s stimulation), when PV is not completely saturated with Ca²⁺, it still binds Ca²⁺ during relaxation, thereby favoring the Ca²⁺ dissociation from TnC, which accelerates myofiber relaxation (7, 15).

In nonfatiguing contractions, short-term inducible ablation of PV in adult mice leads to a greater force at submaximal frequencies of stimulation (Figs. 2B and 4C) with a marked slowing in relaxation (Table 1) in isolated EDL and in FDB myofibers. Similar data were also found during in situ contractions in tibialis anterior muscle and in isolated EDL and FDB muscles from conventional PV knockout (cPV-KO) mice (14, 15). Therefore, the present data using inducible PV ablation in adult mice confirms the results found in cPV-KO myofibers.

Importantly, contractility was not altered in the soleus muscles of PV-KO mice (Fig. 2C and Table 1), a muscle with low amounts of PV due to the high proportion of slow-twitch myofibers (17, 21, 22). This confirms that the treatment with tamoxifen did not change the contractile properties of already low-expressing PV muscles. This also validates that PV only affects contractility of more glycolytic myofibers, since EDL muscles contain ∼98% Type IIb and IId(x) (i.e., more glycolytic) myofibers, whereas soleus contains ∼54% Type IIa and ∼46% Type I (i.e., more oxidative) myofibers (15, 31, 32).

The role of PV on relaxation and [Ca²⁺]_c has been studied in intact single myofibers from wild-type mice by using prolonged periods of stimulation to produce full loading of Ca²⁺ on PV (i.e., >2 s train duration) (7). However, the changes in force together with changes in [Ca²⁺]_c have never been investigated in intact single myofibers after inducible PV ablation. We found that tetanic [Ca²⁺]_c in FDB myofibers from PV-KO mice were significantly higher than PV-Ctr myofibers, whereas myofilament Ca²⁺ sensitivity was unaltered (i.e., force was proportionally activated by [Ca²⁺]_c between genotypes; Fig. 4). Although increases in Ca²⁺-transients have also been detected in collagenase-digested EDL myofibers from cPV-KO mice (14), the authors were not able to correlate the changes in Ca²⁺ with force, nor determine whether myofilament Ca²⁺ sensitivity was altered, which requires dissected single myofibers with intact tendons. Therefore, our data validate that PV acts as a cytosolic Ca²⁺ buffer right after the beginning of the stimulation period without affecting the myofilament Ca²⁺ responses.

The Ca²⁺ decay rates at low [Ca²⁺]_c ([Ca²⁺]_c < 150 nM) that occur at the “tail” of Ca²⁺ decay (>200 ms after the end of the stimulation period) were faster in PV-KO myofibers compared with PV-Ctr myofibers (Fig. 4H). At the “tail” of Ca²⁺ decay, [Ca²⁺]_c is low enough to allow the analysis of Ca²⁺ kinetics using FURA-2 (33, 34), and Ca²⁺ should be already dissociated from the low-affinity binding sites of TnC in fast-twitch myofibers (35). Also, in PV-Ctr myofibers, after an initial binding of Ca²⁺ to PV during the fast phase of Ca²⁺ decay, the Ca²⁺ availability for SERCA is dependent on the slow dissociation of Ca²⁺ from PV during the exchange of Ca²⁺ with Mg²⁺ (24, 27). However, in PV-KO myofibers, as also in human myofibers, Ca²⁺ availability for SERCA is solely dependent on cytosolic Ca²⁺ diffusion (15). Several studies have reported that the rate of Ca²⁺ decay during the “tail” is proportional to the rate of SR Ca²⁺ pumping (6, 9, 36). Interestingly, in cultured myofibers from cPV-KO mice a lower [Ca²⁺]_c at the late phase of Ca²⁺ decay was observed, which was attributed to a more efficient cytosolic Ca²⁺ removal in cPV-KO mice (17). The data from the present investigation suggest that SERCA Ca²⁺ pumping function at the same [Ca²⁺]_c levels during late-phase relaxation is accelerated when PV is absent (Fig. 4H). Although we have not measured each cytosolic Ca²⁺ buffer in the PV-KO muscles, which can limit the interpretation of these results, it is well known that TnC and PV are the most abundant Ca²⁺ buffers present in the cytosol (37). Thus, although the presence of PV accelerates the Ca²⁺ dissociation from TnC during relaxation, the results from the present work suggest that PV-Ca²⁺ dissociation rate constrains, and slows, SR Ca²⁺ pumping velocity during relaxation.

Role of Parvalbumin on Cytosolic Ca²⁺ Handling and SR Ca²⁺ Uptake during Fatigue

It has been previously shown that fatigue resistance is significantly improved in cPV-KO mice (17, 18). Although the authors have detected an increase in mitochondrial biogenesis, they were unable to determine whether the increased fatigue resistance in cPV-KO mice was due to the known, or yet unknown, effects of a lifelong PV absence or by a direct consequence of PV absence on Ca²⁺ movements during repetitive contractions.

In the present study, PV expression was absent 3 wk after the Cre recombinase expression activation by tamoxifen, which minimized any adaptive effects of the absence of PV compared with cPV-KO mice. We hypothesized that inducible ablation of PV in myofibers in adult mice leads to an increase in fatigue resistance due to a higher tetanic [Ca²⁺]_c available to the contractile sites during repetitive contractions. The data show that fatigue resistance was clearly not different between PV-KO versus PV-Ctr mice in EDL isolated muscle (Fig. 3A) or in intact single myofibers (Fig. 5A). Furthermore, the fatigue-induced slowing in relaxation (2) was proportionally similar between the two genotypes, in both EDL muscles and FDB myofibers. These results do not support the original hypothesis of this study as well as the results obtained in cPV-KO mice (17, 18) and suggest that the presence of PV has no or minimal role on fatigue development or fatigue-induced slowing in relaxation. In addition, the data from the present investigation support the general concept that fatigue resistance is mostly dependent on the balance between the rate of energy consumption and adequate ATP supply (38). Our data show that force production, total force-time integral, and changes in peak force during fatiguing contractions were similar between in PV-KO and PV-Ctr myofibers, suggesting that total cross-bridge energy consumption during contractions was not altered by inducing PV ablation in myofibers. The energy cost of Ca²⁺ uptake in the SR accounts for ∼40% of total energy cost of contractions (39), and the total amount of SR Ca²⁺ pumped during each contraction and relaxation depends on SR Ca²⁺ release. Although PV-KO had an increased [Ca²⁺]_c during contractions (see discussion below), this could be a result of the decreased cytosolic Ca²⁺ buffering, augmenting Ca²⁺-FURA-2 binding, and not necessarily from SR Ca²⁺ release. Thus, the total energy costs of cross-bridge cycling and of SR Ca²⁺ pumping were likely not changed, which explains the similar fatigue resistance of PV-Ctr and PV-KO myofibers.

The present study investigated the [Ca²⁺]_c responses during fatiguing contractions in myofibers from inducible PV ablated adult mice. This is the first study to perform these measurements in intact single myofibers in this transgenic mouse model during fatigue. Both tetanic and basal [Ca²⁺]_c changes during contractions were significantly different in PV-KO myofibers (Fig. 5, B and C), although changes in force were not significantly different between genotypes. PV-KO myofibers showed a significant increase in tetanic [Ca²⁺]_c, most apparent in the first minute of contractions, which has been extensively demonstrated that the first minute of contractions represents Phase I of muscle fatigue (2). At this phase, there is a 10–20% decay of force associated with an increase in tetanic [Ca²⁺]_c, which was also detected in the present study in myofibers from both genotypes (Fig. 5). The decrease in force at the Phase I of fatigue has been associated as a decrease in myofilament Ca²⁺ sensitivity and myofibrillar force-generating capacity due to an increase in cytosolic inorganic phosphate concentration ([P_i]_c) (2, 40, 41). Although [P_i]_c during contractions was not measured (i.e., there is no P_i-sensitive fluorophore for the expected [P_i]_c during contractions), it is unlikely that PV-KO myofibers had produced a higher [P_i]_c during fatigue compared with PV-Ctr, since there was no difference in Ca^2 +₅₀ at the time of task failure between genotypes. These data suggest that free Ca²⁺ released by the SR during contractions at Phase I of fatigue is partially buffered by PV, which did not occur in PV ablated myofibers, resulting in an increase in tetanic [Ca²⁺]_c. Also, the higher tetanic [Ca²⁺]_c in PV-KO versus PV-Ctr myofibers with the same decay in force during Phase I suggests that myofibrillar force-generating capacity is a major mechanism for the decrease in force at the beginning of fatiguing contractions (2). However, tetanic [Ca²⁺]_c was similarly decreased between genotypes at the time of task failure. Recently, force and [Ca²⁺]_c were measured in fast-fatiguing (i.e., possibly fast-twitch) isolated single myofibers from human intercostal muscles (42), which should resemble PV-KO myofibers from the present study. In intact single human myofibers, the authors found similar changes of force and [Ca²⁺]_c (tetanic and basal) (42) as detected in PV-KO, which confirms that PV buffering does not alter the changes in force and [Ca²⁺]_c that occur during fatiguing contractions. However, SR Ca²⁺ pumping changes during fatigue were not measured and could not be compared with the present work. Therefore, PV does not seem to interfere with the decreases in tetanic [Ca²⁺]_c available for contractions during fatigue as occurs in human myofibers.

Myofibers from PV-KO mice showed a higher basal [Ca²⁺]_c between contractions compared with PV-Ctr myofibers. Although the results obtained in nonfatiguing contractions suggest that SR Ca²⁺ pumping was faster in PV-KO myofibers, the higher basal [Ca²⁺]_c could also be a result of slower SERCA function during fatigue. Our group has previously shown that there is an inverted correlation between SR Ca²⁺ pumping rate and cytosolic Ca²⁺ accumulation during fatiguing contractions in intact single myofibers (9, 10, 43). As shown in Fig. 6, the SR Ca²⁺ pumping rate was significantly more diminished compared with the first contraction PV-KO than in PV-Ctr myofibers. This may explain the increased cytosolic Ca²⁺ accumulation during fatiguing contractions in PV-KO. Interestingly, the significant slowing in SR Ca²⁺ pumping did not translate to a slowing in the relaxation rate in PV absent myofibers. This strengthens the hypothesis that, at least in healthy mice, the slowing in cross-bridge detachment, and not the slowing in SR Ca²⁺ uptake, is the major factor responsible for the slowing in relaxation that occurs during fatigue (4, 5). In addition, it is expected that the cytosolic Ca²⁺ handling changes during fatigue found in myofibers from PV absent mice should be similar to human myofibers under fatiguing contractions.

Summary

In summary, ablation of the PV in myofibers of adult mice (as a model of humanized cytosolic Ca²⁺ buffering in myofibers) revealed that the Ca²⁺-buffering capacity of PV, although significantly contributing to the acceleration of the relaxation, slows SR Ca²⁺ pumping during nonfatiguing contractions. In addition, the presence of PV during fatiguing contractions in fast-twitch myofibers minimizes the changes in [Ca²⁺]_c (basal and tetanic) that occur in myofibers not expressing PV, as are expected to occur in fast-twitch myofibers from human muscle. Intriguingly, the differences in [Ca²⁺]_c found in PV-KO versus PV-Ctr myofibers did not influence the contractile (force and relaxation) and energy requirements detected during fatiguing contractions, nor did they alter myofiber fatigue resistance. This also supports that myofibers obtained from mice are suitable models to investigate the mechanisms of human muscle fatigue resistance.

GRANTS

This research was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR-069577 (to M.C.H), by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq–Universal 424527/2016-2; to L.N.), and by the Tobacco-Related Disease Research Program New Investigator Award T29KT0397 (to L.N.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

L.N. and M.C.H. conceived and designed research; L.N. and N.K.G. performed experiments; L.N. and N.K.G. analyzed data; L.N. and M.C.H. interpreted results of experiments; L.N. prepared figures; L.N. drafted manuscript; L.N., N.K.G., and M.C.H. edited and revised manuscript; L.N., N.K.G., and M.C.H. approved final version of manuscript.