Residual force enhancement is reduced in permeabilized fiber bundles from mdm muscles

When skeletal muscles are stretched during activation, they exhibit a large increase in force (Abbott and Aubert, 1952; Flitney and Hirst, 1978; Katz, 1939), whereas consumption of energy decreases (Fenn, 1924; Ortega et al., 2015). When held at the stretched length, the force decreases and eventually reaches a steady state that is higher than the force produced during an isometric contraction at the same final length. The difference between steady-state force after active stretch and force during isometric contraction at the same final length is termed residual force enhancement (RFE; Edman et al., 1982). There is no widely accepted theory that provides a mechanism for RFE in skeletal muscles (Rassier and Herzog, 2004). Some alternative theories explain RFE based on contributions of cross-bridges and/or non-cross-bridge structures such as titin (Minozzo and de Lira, 2013).

Titin contributes up to 98% of passive force in myofibrils and sarcomeres (Bartoo et al., 1997; Granzier and Labeit, 2004). The titin protein has an I-band region that is composed of two spring-like elements, the proximal tandem Ig domains and the PEVK region (Gautel and Goulding, 1996). The N2A segment links the tandem Ig domains to the PEVK region (Linke et al., 1998). Earlier studies assumed that titin does not contribute to active muscle force production (e.g. Linke et al., 2002). However, numerous studies have suggested a significant role for titin in active muscle contraction (Horowits et al., 1986; Herzog, 2014; Leonard and Herzog, 2010; Linke, 2018; Monroy et al., 2012; Nishikawa et al., 2012). A role for titin in active muscle contraction was demonstrated in skeletal myofibers after exposure to low dose ionizing radiation (Horowits et al., 1986; Horowits and Podolsky, 1987), which showed that titin degradation reduced active force production and suggested that titin helps maintain active tension (Linke, 2018; Nishikawa et al., 2020).

More recently, several studies have demonstrated interactions between titin and calcium (Labeit et al., 2003; Tatsumi et al., 2001), as well as calcium-dependent interactions between titin and actin (Dutta et al., 2018; Kellermayer and Granzier, 1996), which have been suggested to play an important role in residual force enhancement (Leonard and Herzog, 2010; Tahir et al., 2020). Leonard and Herzog (2010) and Powers et al. (2014) demonstrated that titin force increases upon activation of skeletal muscles. They found that force increased faster with sarcomere length in active compared with passive myofibrils when stretched beyond overlap of the thick and thin filaments to eliminate production of force by cross-bridges. In contrast to wild-type (WT) muscles (Powers et al., 2014), Powers et al. (2016) observed that titin force fails to increase during active stretch of single muscular dystrophy with myositis (mdm) myofibrils stretched beyond overlap of thick and thin filaments. The mdm mutation (Lane, 1985) results in a small deletion in titin that includes 83 amino acids in the N2A region (Garvey et al. 2002). The observation by Powers et al. (2016) supports the idea that the mdm mutation reduces titin-based force in active muscles (Nishikawa et al., 2019, 2020) and supports the idea that the mdm mutation in mice is a good tool for assessing titin's role in active muscle.

Several recent studies have suggested that titin modulates RFE in skeletal muscles (Joumaa and Herzog, 2014; Shalabi et al., 2017; Tahir et al., 2020). In a study on skinned fibers from rabbit psoas muscles, Joumaa and Herzog (2014) found an increase in calcium sensitivity in the RFE state compared with purely isometric contraction. Furthermore, partial degradation of titin via trypsin treatment and subsequent osmotic compression of trypsin-treated fibers via dextran (T-500) eliminated the increase in calcium sensitivity (Joumaa and Herzog, 2014). These results suggest that titin regulates calcium sensitivity in RFE and that titin-based forces can modify calcium sensitivity via a mechanism that is independent of myofilament lattice spacing. Another study demonstrated the presence of RFE in rabbit skeletal (psoas) but not in cardiac myofibrils (Shalabi et al., 2017), suggesting that the presence of RFE is related to differences in titin isoforms between skeletal (N2A) and cardiac muscles (N2B). Although consideration of RFE in cardiac muscle is beyond the scope of the present study, we note here that previous studies differ in reporting the presence (Boldt et al., 2020) or absence (Shalabi et al., 2017; Tomalka et al., 2019) of RFE in cardiac muscles. A recent study (Tahir et al., 2020) also showed that RFE was reduced in intact mdm soleus (SOL) muscles compared with WT.

In contrast to these previous studies that suggest a role for titin in RFE, Powers et al. (2017) reported similar average ‘titin force enhancement’ (TFE) in mdm and WT single psoas fibers. They defined TFE as the residual stress remaining in actively stretched individual fibers after subtracting the passive stress and contractile stress at a final sarcomere length of 3.2 µm (Powers et al., 2017). These results contradicted observations from a previous study in mdm myofibrils showing no increase in titin-based force upon activation (Powers et al., 2016), as well as a study on intact mdm SOL muscles (Tahir et al., 2020), which showed no RFE after subtracting passive stress after stretch from RFE.

The aim of our study was to revisit the question of whether RFE is reduced in fiber bundles from mdm SOL and extensor digitorum longus (EDL) muscles compared with WT. We included the relatively fast EDL and slow SOL muscles (Kushmerick et al., 1992) because previous observations suggested that the mdm mutation affects twitch and tetanic stress (Hessel et al., 2019) as well as eccentric work (Hessel and Nishikawa, 2017) to a greater extent in SOL than in EDL muscles, but RFE has not been quantified previously in mdm EDL muscles. We performed active stretch, passive stretch and isometric contraction from 2.6 to 3.0 µm on the same individual fiber bundles. We hypothesized that mdm fiber bundles from both SOL and EDL muscles will show reduced RFE compared with WT fibers, in contrast to the results of Powers et al. (2017). This result would be consistent with previous observations in single myofibrils (Powers et al., 2016) and intact muscles (Tahir et al., 2020). We also expected that RFE would be greater in the fast EDL than in the slower SOL on the basis of previous studies (Rassier, 2017; Tomalka et al., 2017).

Heterozygous mice of the strain B6C3Fe a/a-Ttn mdm/J were obtained from the Jackson Laboratory (Bar Harbor, ME, USA) and a breeding colony was maintained to produce WT and homozygous recessive (mdm) mice. For experiments, WT mice (n=5) were 29–42 days of age (mean±s.e.m. age=35.8±2.7 days; mean body mass=19.5±1.8 g) and mdm mice were 16–30 days of age (mean age=25.0±2.5 days; mean body mass=4.1±0.7 g). Mice were housed in groups of 2–4 mice per cage, maintained on 12 h:12 h light:dark cycle at 22°C and were fed ad libitum. Mice of both sexes were killed via isoflurane overdose confirmed by cervical dislocation. For each mouse, 3–4 fiber bundles were harvested from the EDL and SOL (n=5 muscles per genotype), for a total of 16–20 fibers per muscle×genotype). We used fiber bundles from younger mice to reduce the possibility of fibrosis (Lopez et al., 2008) and degeneration (Heimann et al., 1996), which might affect the active and passive tension after stretch. EDL and SOL muscles were extracted from euthanized mice following standard procedures (Hakim and Nishikawa, 2013). The Institutional Animal Care and Use Committee at Northern Arizona University approved all husbandry and experimental protocols (nos. 18-002 and 21-001).

Fiber bundles were chosen for study because, like single fibers, sarcomere lengths can be imaged in real time during testing and force can be controlled by varying the calcium concentration, yet they are stronger and less likely to be damaged during isolation or active stretch. Skinned fiber bundles were prepared using standard techniques to minimize the contributions of endomysium to passive tension (Joumaa and Herzog, 2014). Extracted muscles were placed in a collection solution (in mmol l⁻¹: Tris 50, KCl 100, MgCl₂ 2, EGTA 1, pH 7.0) for 6 h at 4°C, then transferred to an overnight solution (in mmol l⁻¹: Tris 50, KCl 2, NaCl 100, MgCl₂ 2, EGTA 1, pH 7.0) with 1:1 glycerol for 12 h at 4°C, and finally placed in a long-term storage solution (collection solution with 1:1 glycerol) for 4–6 weeks at −20°C (Joumaa et al., 2007). To protect against protein degradation, protease inhibitor tablets (Complete^®, Roche Diagnostics, Montreal, Quebec, Canada) were added to all solutions (1 tablet per 50 ml solution). Muscles were used 4–6 weeks after surgery. On the day of experiments, muscles were washed in relaxing solution (in mmol l⁻¹: potassium propionate 170, magnesium acetate 2.5, MOPS 20, K₂EGTA 5, ATP 2.5, pH 7.0) to remove the rigor–glycerol solution. Bundles of 4–5 fibers were then dissected from each muscle in a Petri dish filled with ice-cold relaxing solution and used immediately for experimentation. The solutions used during the experiments included relaxing solution, washing solution (in mmol l⁻¹: potassium propionate 185, magnesium acetate 2.5, MOPS 20, ATP 2.5, pH 7.0) and activating solution [in mmol l⁻¹: potassium propionate 170, magnesium acetate 2.5, MOPS 10, ATP 2.5, and CaEGTA and K₂EGTA mixed at different proportions to obtain a pCa (–log [Ca²⁺]) of 4.2, pH 7.0]. All experiments were conducted at room temperature (22°C). At this temperature, previous studies of rat and mouse SOL and EDL have found that these muscles retain ∼80–90% of their force at 35°C (James et al., 2015; Stephenson and Williams, 1985), although another study found a larger drop in the force of rat EDL and SOL at this temperature (Galler and Hilber, 1998).

Using acetone and cellulose acetate (Merck, Darmstadt, Germany), each fiber bundle was glued lengthwise to a high-speed length controller on one end (Aurora Scientific, model 322C, Ontario, Canada) and a force transducer on the other end (Aurora Scientific, model 400A). Length control and force measurements were accomplished using an ASI802D data acquisition and control system (Aurora Scientific). The rig was connected to an inverted microscope (Leica, DMIRE2) with an ocular micrometer for measurement of fiber diameter (graticule 10 mm=100 divisions). The microscope was also fitted with a sarcomere length (SL) tracking camera (Aurora Scientific HVSL600A). Changes in SL were measured via SL tracking and acquisition software. The fiber rig was controlled using Linux software, which switches wells containing different solutions. In contrast to a previous study (Powers et al., 2017) in which laser diffraction was used to measure average SL, we used the SL tracking camera to measure initial and final SL. For measurement of SL, the Aurora 901B camera has a spatial resolution of 5 pixels µm⁻¹ at high magnification (50× optical lens of the inverted microscope) and 2 pixels µm⁻¹ at low magnification (20× optical lens of inverted microscope) at 300 frames s⁻¹ full frame (Aurora Scientific Inc., 2012).

The mechanical test protocols were designed to maximize RFE and minimize fiber damage by implementing a long stretch at a slow velocity, as the magnitude of RFE increases with stretch amplitude but is largely independent of stretch rate (Abbott and Aubert, 1952; Herzog and Leonard, 2000; Sugi and Tsuchiya, 1988). To prevent slippage or breakage of fibers during the experiments, two pre-conditioning stretches in relaxing solution were performed before the start of each experiment. The pre-conditioning stretches from 2.6 to 3.0 µm were performed at a rate of 0.04 µm s⁻¹. After the pre-conditioning stretches, fiber bundles were passively stretched from 2.6 to 3.0 µm at 0.04 µm s⁻¹ (Fig. 1), held isometrically for 60 s until steady-state force was attained, then moved to relaxing solution and returned to the original SL (2.6 µm). After a rest of 5 min, the fiber bundles were activated at 2.6 µm and were held isometrically for 118 s to reach steady-state force (Fig. 2). Thereafter, fiber bundles were actively stretched from 2.6 to 3.0 µm at 0.04 µm s⁻¹ and held for 60 s, at which time the steady-state force at 3.0 µm was measured. Fiber bundles were then deactivated and returned to the starting SL. Finally, after a rest of 5 min, fiber bundles were passively stretched from 2.6 to 3.0 µm at a rate of 0.04 µm s⁻¹, then activated (139 s) to reach steady-state isometric stress at 3.0 µm (Fig. 2).

Stress (mN mm⁻²) was calculated by dividing force by cross-sectional area, assuming a cylindrical shape of the fiber bundles (Powers et al., 2017). RFE was calculated by subtracting the steady-state force during isometric contraction at 3.0 µm from the steady-state force after active stretch to the same final SL (Fig. 2). We also calculated TFE as the stress after active stretch that remains after subtracting the total stress (passive+active) at the final SL of 3.0 µm (Powers et al., 2017). Finally, percentage of TFE (TFE%) as defined by Powers et al. (2017) was calculated as percentage of total stress following active stretch to final length (3.0 µm). Fibers in which no measurable force was observed after activation were eliminated from the trials. Based on this criterion, 5% of WT trials and 16% of mdm trials were removed from the final data set.

Statistical analysis was performed using JMP Pro14 and GraphPad Prism software. The Shapiro–Wilk test showed that the data were normally distributed (P>0.05). Levene's test showed that the variances differed significantly between genotypes (P<0.05). We corrected the data with non-homogeneous variances using Box–Cox transformation. To evaluate differences between WT and mdm fiber bundles, we used one-way ANOVA with genotype as the main effect. Because more than one fiber bundle from a single muscle was included in the data, we accounted for the within-muscle variation by including a random effect of muscle nested within genotype. Alpha values were set at 0.05 and the sample of 3–4 fiber bundles from n=5 mice per genotype was adequate to detect observed differences between means with power >0.8. The dependent variables were isometric stress at 2.6 µm, isometric stress at 3.0 µm, steady-state stress after passive stretch from 2.6 to 3.0 µm, steady-state stress after active stretch from 2.6 to 3.0 µm, RFE and TFE%. Data are presented as means±s.d. One-way ANOVA showed that the random effect of variation among muscles was significant for all dependent variables in both genotypes (WT and mdm). There were no significant correlations between mouse age and any dependent variables for either genotype (WT versus mdm) or muscle (EDL versus SOL; all P>0.05).

Total isometric stress at 2.6 µm (Fig. 3) was higher in WT EDL fiber bundles (n=20, mean±s.d.=114.5±0.27 mN mm⁻²) than in mdm EDL fiber bundles (n=19, 78.5±5.77 mN mm⁻²; ANOVA, F=166.8, P<0.0001). Similarly, total isometric stress was higher in WT SOL fiber bundles (n=19, 113.6±0.33 mm²) at 2.6 µm (Fig. 3) compared with mdm SOL fiber bundles (n=16, 78.0±5.80 mN mm⁻²; ANOVA, F=166.18, P<0.0001).

Steady-state stress following passive stretch from 2.6 to 3.0 µm (Fig. 4A) was significantly higher in mdm EDL fiber bundles (n=19, 24.1±5.03 mN mm⁻²) than in WT EDL fiber bundles (n=20, 9.8±2.82 mN mm⁻²; ANOVA, F=35.0743, P<0.0004). Similarly, mdm SOL fiber bundles (n=16, 29.4±6.39 mN mm⁻²) had higher passive stress following stretch from 2.6 to 3.0 µm than WT (n=19, 10.0±1.47 mN mm⁻²; ANOVA, F=56.31, P<0.0001; Fig. 4A).

Steady-state stress after active stretch (Fig. 4B) was not significantly different between mdm EDL fiber bundles (132.5±16.70 mN mm⁻²) and WT EDL fiber bundles (133.4±0.94 mN mm⁻², ANOVA, F=2.0159, P=0.1934). However, mdm SOL fiber bundles (Fig. 4B) had a significantly higher active stress after stretch (139.4±1.61 mN mm⁻²) than WT fiber bundles (135.6±1.76 mN mm⁻²; ANOVA, F=12.786, P=0.0072).

Total isometric stress at 3.0 µm (Fig. 5) was significantly higher in mdm EDL (122.1±15.21 mN mm⁻²) compared with WT EDL (90.4±0.88 mN mm⁻²; ANOVA, F=49.09, P<0.0001). Similarly, mdm SOL (131.7±2.15 mN mm⁻²) had significantly higher total isometric stress at 3.0 µm than WT SOL fiber bundles (91.6±2.83 mN mm⁻²; ANOVA, F=605.09, P<0.0001).

Significantly lower RFE was observed in mdm EDL (10.4±3.60 mN mm⁻²) and SOL (7.7±1.05 mN mm⁻²) fiber bundles compared with WT EDL (43.0±1.75 mN mm⁻²; ANOVA, F=199.36, P<0.0001) and SOL fiber bundles (43.9±1.44 mN mm⁻²; ANOVA, F=1815.34, P<0.0001), respectively (Fig. 6). Mdm fiber bundles from both EDL (−10±2.96%; Fig. 7A) and SOL (−16±4.7%; Fig. 7B) showed no TFE%. After subtracting the passive stress component from RFE, TFE% was significantly lower in mdm fiber bundles compared with WT EDL (26±2.14%; ANOVA, F=316.9, P<0.0001). Similarly, TFE% was lower in mdm SOL fiber bundles compared with WT SOL fiber bundles (25±0.89%; ANOVA, F=219.6, P<0.0001).

In summary, we found that total isometric stress was higher and less variable in WT EDL and SOL fiber bundles than in mdm fiber bundles at an SL of at 2.6 µm but was higher and more variable in mdm fiber bundles than WT at 3.0 µm in both muscles. Stress following passive stretch from 2.6 to 3.0 µm was also higher and more variable in mdm fiber bundles than WT in both muscles, whereas total stress following active stretch was significantly higher in mdm SOL but not in mdm EDL. RFE was significantly lower in mdm fiber bundles than WT in both muscles, and no fibers of any muscles exhibited positive values of TFE%.

Reduced total isometric stress in mdm skeletal muscles compared with WT at shorter SLs near the plateau of the length–tension relationship (see Fig. 3) has also been found in previous studies (Huebsch et al., 2005; Monroy et al., 2012; Powers et al., 2016, 2017; Tahir et al., 2020). This reduction in contractile stress in mdm is likely due at least in part to impaired transmission of cross-bridge force from the A-band to the Z-line (Horowits et al., 1986). In the present study, the total isometric stress was higher in mdm skeletal muscles compared with WT at the longer SL (3.0 µm), presumably owing to higher passive tension (see below). Values of total isometric stress at comparable lengths have not been reported in previous studies, but Powers et al. (2017) found no difference between mdm and WT fibers at an SL of 5.2 µm.

Steady-state stress following passive stretch was significantly reduced in WT EDL and SOL fiber bundles compared with fiber bundles from mdm mice (see Fig. 4A,B). Earlier studies also found higher and more variable passive stress in mdm fibers (Powers et al., 2017) and intact muscles (Tahir et al., 2020) compared with WT. Higher passive stress could be due to alterations in exon splicing in the PEVK region of titin that could potentially lead to expression of shorter and stiffer titin isoforms in mdm muscles (Hettige, 2021). In addition, high collagen content in mdm muscles has also been demonstrated (via colorimetric hydroxyproline assay) as a potential contributor to higher passive stress (Powers et al., 2017). Higher passive stress accounts for the higher total contractile stress at 3.0 µm (sum of active stress plus steady-state passive stress) in mdm muscles, which was found in the present study and has also been found in previous studies (Powers et al., 2017; Tahir et al., 2020; Hessel et al., 2019).

Steady-state stress was larger after active stretch than after passive stretch in WT and mdm EDL and SOL fiber bundles (Fig. 4A,B). We found that steady-state stress after active stretch was larger in mdm SOL than in WT SOL (Fig. 4B). In contrast, Powers et al. (2017) found no difference between passive and active stress after stretch to very long SLs (5.2 µm) in psoas fibers from mdm mice, likely owing to much higher passive tension at the longer sarcomere length. They also found no difference in active stress after stretch between mdm and WT fibers.

RFE, the increase in steady-state stress after active stretch (Abbott and Aubert, 1952; Edman et al., 1982), is a long-known property of skeletal muscles. Our results showed reduced RFE in mdm fiber bundles from both SOL and EDL compared with WT (see Fig. 6). The small amount of RFE that remains in mdm fiber bundles could be due to the presence of glutamine-rich motifs in the PEVK region of titin, which have a high affinity for Ca²⁺ (Labeit et al., 2003). This result is consistent with previous observations from mdm myofibrils (Powers et al., 2016) and intact muscles (Tahir et al., 2020).

Similar to Hessel et al. (2017), we found that RFE was higher in EDL than in SOL. This may be due to expression of a larger titin isoform in SOL than EDL (Freiburg et al., 2000; Prado et al., 2005; Buck et al., 2014). Prado et al. (2005) demonstrated that passive tension is higher in muscles with shorter titin isoforms (i.e. EDL) compared with muscles with longer titin isoforms (e.g. SOL; see Hettige, 2021 for isoform expression in mdm muscles). Cornachione et al. (2016) also demonstrated that increasing size of titin isoforms was related to decreasing passive tension in rabbit SOL, psoas and ventricular myofibrils. Although they did not quantify RFE directly, they found that the difference in force between passive and active ramp stretches of the same myofibril also decreased with the size of the titin isoform in these muscles.

Powers et al. (2017) reported the presence of TFE in nine of 11 mdm fibers whereas TFE was absent in only two mdm fibers. In the present study, we found no TFE in any mdm fiber bundles from either SOL or EDL muscles (see Fig. 7). Several factors may account for the difference in results between the present study and those of Powers et al. (2017). Most importantly, we used mdm fiber bundles from young mice (16–30 days old) compared with the older (27–41 days old) mdm muscles used by Powers et al. (2017). Muscle degeneration starts as early as 1–2 weeks of age in SOL but not until 3–4 weeks of age in tibialis anterior (Heimann et al., 1996). Several studies have found that mdm skeletal muscles have normal sarcomere structure at ∼24–29 days of age (Hessel et al., 2019; Lopez et al., 2008; Witt et al., 2004).

Powers et al. (2017) found significant structural changes in the mdm psoas muscles they studied, including small fiber diameter, large numbers of central nuclei, and fibrosis (see their Figs 3 and 4). They were only able to measure SL using laser diffraction in less than 20% of their fibers, whereas Hessel et al. (2019) were able to measure SL using laser diffraction in all of their 24- to 29-day-old EDL fibers. In the present study, sarcomere structure appeared normal and SL could be quantified from video images in all of the fiber bundles tested. Based on the older age of the mdm muscles used in Powers et al.’s (2017) study, it appears likely that sarcomere degeneration and fibrosis led to high variability in tension after both passive and active stretching, and ultimately to the observation of TFE in a large number of mdm fibers. However, in the present study, RFE was reduced significantly and we failed to observe TFE in any mdm fiber bundles from EDL or SOL at 16–30 days of age.

Recent studies have proposed a role for titin in RFE (Leonard and Herzog, 2010; Nishikawa et al., 2012; Rode et al., 2009; Shalabi et al., 2017). Both Leonard and Herzog (2010) and Powers et al. (2014) showed that Ca²⁺-activated myofibrils stretched beyond overlap of thick and thin filaments have a higher titin-based force than passively stretched myofibrils at the same average sarcomere length; they found a steeper increase in force in active WT myofibrils than in passive WT myofibrils when stretched to very long sarcomere lengths beyond overlap of the thick and thin filaments. In contrast, Powers et al., (2016) found no difference in slope between actively and passively stretched mdm myofibrils, which suggests that titin force fails to increase upon activation in mdm myofibrils due to the small deletion in the N2A region of titin.

Leonard and Herzog (2010) and Powers et al. (2014, 2016) suggested binding of titin to actin as a mechanism for increasing titin force in active muscle. Nishikawa et al. (2012, Nishikawa, 2016) proposed that if N2A titin binds to actin in active muscle, then tandem Ig domains of titin would straighten at low force when skeletal muscles are passively stretched. When stretched actively, only the PEVK region of titin would extend at high force. In recent experiments, Dutta et al. (2018) demonstrated that N2A titin binds to actin and thin filaments, and that Ca²⁺ increases the strength and stability of N2A titin–actin interactions (Adewale and Ahn, 2021; Tsiros et al., 2021), supporting the hypothesis that titin plays a regulatory role in active muscle contraction. It has also been proposed that the mdm mutation prevents N2A titin–actin interactions, so that active mdm muscles are more compliant than WT muscles (Monroy et al., 2017; Dutta et al., 2018; Nishikawa et al., 2020). The decrease in active muscle stiffness should be associated with a reduction in RFE in mdm muscles (Powers et al., 2014, 2016; Tahir et al., 2020).

The findings of the present study that RFE was reduced and TFE was absent in 16- to 30-day-old EDL and SOL muscles from mdm mice are consistent with previous data from single myofibrils demonstrating that titin force increases upon calcium activation in WT myofibrils stretched beyond overlap for the thick and thin filaments (Powers et al., 2014), but not in mdm myofibrils under the same conditions (Powers et al., 2016). They are also consistent with observations that RFE is reduced in intact mdm SOL muscles (Tahir et al., 2020). The conflicting observations from an earlier study (Powers et al., 2017), which appeared to show that compensatory mechanisms could restore TFE in single fibers from mdm psoas, are likely explained by degeneration and fibrosis. The present study is also consistent with the idea that the mdm mutation affects N2A titin–actin interactions (Dutta et al., 2018), which reduces titin-based force in active muscles (Powers et al., 2016; Monroy et al., 2017), thereby decreasing RFE and TFE.

We thank Anthony Hessel and Ashana Puri for helpful comments on earlier versions of the manuscript.