Voluntary wheel running with and without follistatin overexpression improves NMJ transmission but not motor unit loss in late life of C57BL/6J mice

Abstract

Sarcopenia, or pathological loss of muscle mass and strength during aging, is an important contributor to loss of physical function in older adults. Sarcopenia is a multifactorial syndrome associated with intrinsic muscle and upstream neurological dysfunction. Exercise is well-established as an effective intervention for sarcopenia, but less is known about the long-term neurobiological impact of exercise. The goals of this study were to investigate the effects of exercise, alone or in combination with follistatin (FST) overexpression (antagonist of myostatin), on neuromuscular junction transmission and motor unit numbers in mice between the age of 22 and 27 months, ages at which prior studies have demonstrated that some motor unit loss is already evident. C57BL/6J mice underwent baseline assessment and were randomized to housing with or without voluntary running wheels and injection with adeno-associated virus to overexpress FST or vehicle. Groups for comparison included sedentary and running with and without FST. Longitudinal assessments showed significantly increased muscle mass and contractility in the 'running plus FST' group, but running, with and without FST, showed no effect on motor unit degeneration. In contrast, running, with and without FST, demonstrated marked improvement of neuromuscular junction transmission stability.

Keywords: Aging; Exercise; Follistatin; Motor Unit; Neuromuscular Junction.

Figure 1:

Experimental Overview and Timeline

Figure 2:. Examples of normal (A-B) and abnormal (C-D) single fiber action potentials (SFAPs).

A-B. SFAPs with normal jitter and no blocking (superimposed on left and rastered on right). C-D. SFAPs with abnormal jitter and blocking (superimposed on left and rastered on right). Red arrow in D denotes blocked SFAP. Sensitivity: 500 microvolts per division; sweep speed: 500 microseconds per division. [Reproduced with permission (Chugh et al., 2020)].

Figure 3:. Voluntary wheel running behavior in mice and human follistatin expression following AAV9-FS344 injection.

A. A histogram of running distances for all mice housed with wheels shows a non-normal and bimodal distribution of running behavior (range: 0.00–3.63 km/day, median: 0.2 km/day, n=36 mice, 18 males, 18 females). B. Mice housed with wheels (no FST injection) with running distance ≤ 0.2 km/day (n=12, 6 males, 6 females) were classified as ‘sedentary’ and ran a median of 0.05 km/day [95% CI: 0.00, 0.15]. Mice in the ‘running’ group (>0.2 km/day) (n=8, 4 males, 4 females) ran a median distance of 0.74 km/day [95% CI: 0.21, 1.05]. Mice (n=10, 5 males, 5 females) in the ‘running plus FST’ group ran a median distance of 0.62 km/day [95% CI: 0.41, 1.68]. There was no significant difference between the running vs running plus FST groups for running distance (p=0.8968, Mann Whitney Test). C. Increased human FST protein expression was confirmed by ELISA in gastrocnemius muscle samples from Running+FST group (n=8) compared with vehicle-injected controls (n=4) (p=0.0040, Mann Whitney Test). Data shown as median ±95% confidence interval.

Figure 4:. Voluntary Wheel Running combined with FST Overexpression improves Longitudinal In Vivo Muscle Plantarflexion Muscle Contractility.

A. Twitch contractility showed no significant changes with age between 22–27 months (p=0.0595) or between treatment groups (p=0.0761). There was no significant interaction between age and treatment (p=0.2801). B. Tetanic contractility was significantly reduced with aging (p<0.0001) and significantly different between treatment groups (p=0.0224), but there was no interaction between age and treatment (p=0.7022). Multiple Comparisons demonstrated significantly increased tetanic torque in running plus FST versus running at month 3 (denoted as #, adjusted p=0.0366) and month 5 (denoted as ^, adjusted p=0.0348). C. Soleus wet muscle mass was significantly different between treatment groups (p=0.0300) at end point (27 m). On multiple comparisons, running plus FST showed an increase compared with running (adjusted p=0.0238) but no difference between sedentary and running (adjusted p=0.4994). D. Gastrocnemius muscle mass was not significantly different between groups (p=0.1051). E. Twitch torque normalized to triceps surae muscle mass was not significantly different between groups (p=0.1195) at end point. Multiple comparisons demonstrated no significant changes in running versus running plus FST (adjusted p=0.1171) and sedentary (adjusted p=0.0908) groups. F. Tetanic torque normalized to triceps surae muscle mass was unchanged between groups (p=0.6928) at end point. Multiple comparisons demonstrated no significant changes in the running group compared with running plus FST (adjusted p=0.6994) or sedentary groups (adjusted p=0.9983). Mixed effect model was used for longitudinal analyses of twitch and tetanic contractility torque for effects of time and treatment. Dunnett’s multiple comparison test used to compare treatment effects at each timepoint. One Way ANOVA with Dunnett’s multiple comparisons test for analyzing end point muscle mass and specific twitch and tetanic torque. *, p < 0.05. Data shown as mean ± standard deviation.

Figure 5.. Twitch muscle contractility is positively correlated with running distance.

A. Absolute twitch contractility at endpoint showed a significant positive correlation with running distance (r²=0.3558, p=0.0314, n=13). B. Absolute tetanic contractility at endpoint showed no correlation (r²=0.1474, p=0.1953, n=13). C. Twitch contractility normalized to triceps surae mass at endpoint showed a significant positive correlation with running distance (r²=0.5000, p=0.0069, n=13). D. Tetanic contractility normalized to triceps surae mass at endpoint showed no correlation (r²=0.0008, p=0.9290, n=13). Pearson correlation coefficients, r².

Figure 6:. Voluntary wheel running with or without follistatin demonstrated no effect on compound muscle action potential or motor unit number but showed transient effects on single motor unit amplitudes.

A. Compound muscle action potential (CMAP) amplitude showed significant reduction with age (p<0.0001). There were no significant differences between treatment groups (p=0.8372) or interaction between age and treatment (p=0.3666). Multiple comparisons between groups demonstrated no significant differences. B. Average single motor unit potential (SMUP) size showed no significant change with age (p=0.0920) or with treatment (p=0.4665) and no interaction was observed between age and treatment (p=0.1457). Multiple comparisons between groups showed significantly reduced SMUP size in running versus sedentary mice at month 3 (denoted as ^, adjusted p=0.0040) C. Motor unit number estimation (MUNE) showed significant reduction with age (p<0.0001). There were no significant differences with treatment (p=0.6204) or interaction between age and running (p=0.6070). Multiple comparisons between groups for MUNE showed no significant differences. Mixed effect model for comparison of time and treatment effects with Dunnett’s Multiple Comparisons tests for treatment effects at each timepoint. Data shown as mean ± standard deviation.

Figure 7:. Voluntary wheel running with or without follistatin improves NMJ transmission stability in aged C57BL/6J mice.

A. There were significant differences for jitter between groups (p=0.0049). On multiple comparisons, running (n=5 mice, n=46 synapses) demonstrated reduced jitter compared with both the sedentary (n=19 mice, n=208 synapses) (adjusted p=0.0043) and running plus FST (n=8 mice, n=89 synapses) groups (adjusted p=0.0054). B. There was significantly reduced NMJ blocking in the running group (9%, 4 of 46 synapses) compared with the sedentary group (25%, 51 of 208 synapses) (adjusted p=0.0368) but not when compared with running plus FST (11%, 10 of 89 synapses) (adjusted p=0.9176). Kruskal Wallis test was used for comparing jitter between groups with Dunn’s multiple comparison tests for comparing running versus running plus FST and sedentary groups. Chi Square test for comparing blocking frequency between running versus running plus FST and sedentary groups, p-values adjusted for multiple comparisons. Data shown as median ± 95% confidence interval. *, p<0.05; **, p<0.01.