Exercise preserves physical fitness during aging through AMPK and mitochondrial dynamics

Abstract

Exercise is a nonpharmacological intervention that improves health during aging and a valuable tool in the diagnostics of aging-related diseases. In muscle, exercise transiently alters mitochondrial functionality and metabolism. Mitochondrial fission and fusion are critical effectors of mitochondrial plasticity, which allows a fine-tuned regulation of organelle connectiveness, size, and function. Here we have investigated the role of mitochondrial dynamics during exercise in the model organism Caenorhabditis elegans. We show that in body-wall muscle, a single exercise session induces a cycle of mitochondrial fragmentation followed by fusion after a recovery period, and that daily exercise sessions delay the mitochondrial fragmentation and physical fitness decline that occur with aging. Maintenance of proper mitochondrial dynamics is essential for physical fitness, its enhancement by exercise training, and exercise-induced remodeling of the proteome. Surprisingly, among the long-lived genotypes we analyzed (isp-1,nuo-6, daf-2, eat-2, and CA-AAK-2), constitutive activation of AMP-activated protein kinase (AMPK) uniquely preserves physical fitness during aging, a benefit that is abolished by impairment of mitochondrial fission or fusion. AMPK is also required for physical fitness to be enhanced by exercise, with our findings together suggesting that exercise may enhance muscle function through AMPK regulation of mitochondrial dynamics. Our results indicate that mitochondrial connectivity and the mitochondrial dynamics cycle are essential for maintaining physical fitness and exercise responsiveness during aging and suggest that AMPK activation may recapitulate some exercise benefits. Targeting mechanisms to optimize mitochondrial fission and fusion, as well as AMPK activation, may represent promising strategies for promoting muscle function during aging.

Keywords: C. elegans; aging; exercise; mitochondrial fission; mitochondrial fusion.

Fig. 1.

Effects of exercise and aging on mitochondrial dynamics. (A) Physical fitness decay and (B) mitochondrial morphology in body-wall muscle cells of WT worms during aging, along with representative images of mitochondrial network organization classes (29). (Scale bar, 5 μm.) (C) Acute exercise protocol: worms maintained at 20 °C were subjected to 4 h of swimming followed by 24 h of recovery on agar plates. (D) Mitochondrial morphology in body-wall muscle cells, (E) physical fitness, and (F) recovery rate of WT worms that were subjected to acute exercise on days 1, 5, and 10 of adulthood. Recovery rate refers to the difference in body bends per second between 24 h and the start of exercise (0 h). (G) Correlation between physical fitness and mitochondrial morphology of WT worms subjected to acute exercise, including data from d1, d5, and d10 of adulthood. Data are presented as mean ± SEM. *P < 0.05 and **P < 0.001 vs. WT (d1 or 0 h); #P < 0.05 and ##P < 0.001 vs. 4 h. Detailed statistical analyses, number of biological replicates, and sample size are described in SI Appendix, Table S3.

Fig. 2.

Mitochondrial fission and fusion are required for exercise-induced benefits during aging. (A) Simplified model for mitochondrial fusion and fission: mitochondrial fragmentation requires recruitment of cytosolic DRP-1 to the organelle, which oligomerizes and constricts the mitochondria into two daughters. The outer mitochondrial membrane fuses through interaction between FZO-1 of two opposing mitochondria, while EAT-3 drives inner mitochondrial membrane fusion. (B) Physical fitness decay (Left) and overall physical fitness (Right, average of d1, d5, d10, and d15) of WT compared to mitochondrial dynamics mutants drp-1(tm1108), fzo-1(tm1133), fzo-1(tm1133);drp-1(tm1108) and eat-3(ad426);drp-1(tm1108) during aging. (C) Physical fitness of WT and mitochondrial dynamics mutants drp-1(tm1108), fzo-1(tm1133), fzo-1(tm1133);drp-1(tm1108) and eat-3(ad426);drp-1(tm1108) submitted to acute exercise on day 1 of adulthood. (D) Long-term exercise: Worms maintained at 20 °C were submitted to 1 h of exercise per day for 10 d, starting at the onset of adulthood (day 1). (E) Physical fitness and (F) mitochondrial morphology in body-wall muscle cells of WT worms submitted to long-term exercise. (G) Correlation between physical fitness and mitochondrial morphology of control and long-term exercise-trained worms. (H) Physical fitness and (I) mitochondrial morphology in body-wall muscle cells of mitochondrial dynamics mutants drp-1(tm1108), fzo-1(tm1133), fzo-1(tm1133);drp-1(tm1108) and eat-3(ad426);drp-1(tm1108) submitted to long-term exercise. (J) Physical fitness of the muscle-specific RNAi strain sid-1(qt-9);myo-3p::sid-1 fed EV control, drp-1, fzo-1 and eat-3 RNAi from the L4 stage and submitted to long-term exercise. Data are presented as mean ± SEM. *P < 0.05 and **P < 0.001 vs. WT or Ctr. Detailed statistical analyses, numbers of biological replicates, and sample size are described in SI Appendix, Table S3.

Fig. 3.

Changes in protein abundance of major pathways involved in muscle adaptation to exercise. Percentage difference in abundance of proteins involved in: (A) mitochondrial fission and fusion, and calcium handling, (B) protein synthesis, (C) Krebs cycle, ETC and oxidative phosphorylation, (D) glycolysis, (E) lipid metabolism, (F) redox balance, and (G) mitochondrial import machinery and proteostasis, determined by comparison between WT and fzo-1(tm1133) worms that were or were not submitted to long-term exercise [starting at the onset of adulthood (day 1), according to the protocol described in Fig. 2D]. Proteomics was performed at day 10. Data are presented as mean ± SEM. *P < 0.05 and **P < 0.001 vs. WT Ctr. Detailed statistical analyses, number of biological replicates, and sample size are described in SI Appendix, Table S3.

Fig. 4.

AMPK activation preserves physical fitness with aging. (A) Physical fitness decay of WT and the long-lived worms isp-1(qm150), nuo-6(qm200), daf-2(e1370), and eat-2(ad1116) with aging. (B) Physical fitness decay and average physical fitness (average of d1, d5, d10, and d15) of WT and long-lived worms expressing constitutively active AMPK (CA-AAK-2) during aging. (C) Physical fitness of WT and CA-AAK-2 worms submitted to acute exercise on days 1, 5, and 10 of adulthood. (D) Physical fitness and (E) mitochondrial morphology in body-wall muscle cells of CA-AAK-2 worms submitted to long-term exercise. Data are presented as mean ± SEM. *P < 0.05 and **P < 0.001 vs. WT or Ctr. Detailed statistical analyses, number of biological replicates, and sample size are described in SI Appendix, Table S3.

Fig. 5.

AMPK-induced improvement in physical fitness requires mitochondrial dynamics. (A) Physical fitness decay and average physical fitness (average of d1, d5, d10, and d15) of WT and AMPK-deficient worms [aak-2(gt33)] during aging. (B) Physical fitness during 4 h of acute exercise (average of 0 h, 1 h, 2 h, and 4 h) on day 1 of adulthood of WT and aak-2(gt33). (C) Physical fitness of aak-2(gt33) worms subjected to long-term exercise. (D) Physical fitness decay and average physical fitness (average of d1, d5, d10, and d15) of animals carrying the indicated mitochondrial dynamics mutations [drp-1(tm1108), fzo-1(tm1133), and fzo-1(tm1133);drp-1(tm1108)] in either the WT or CA-AAK-2 background during aging. CA-AAK-2 in the WT background is presented in blue. (E) Physical fitness decay of the indicated strains during aging. Data are presented as mean ± SEM. *P < 0.05 and **P < 0.001 vs. WT or Ctr. #P < 0.05 and ##P < 0.001 vs. CA-AAK-2. (F) Working model for maintenance of exercise responsiveness and physical fitness during aging: The beneficial effects of exercise are mediated through AMPK and mitochondrial dynamics and proteome remodeling. Detailed statistical analyses, number of biological replicates, and sample size are described in SI Appendix, Table S3.