Longevity & AgingResearch PaperOpen Access

A Mitochondrial Lipid Controls How Aging Muscle Changes Fiber Type

Cardiolipin decline in aging muscle triggers a shift toward slow oxidative fibers — and restoring it reverses muscle wasting in mice.

Thursday, October 1, 2026 1 view
Published in Nat Aging
Cross-section microscopy image of skeletal muscle fibers stained in contrasting colors showing slow oxidative and fast glycolytic fiber types, with a researcher in lab coat examining the slide

Summary

As we age, muscle mitochondria work less well, yet muscle shifts toward fibers that are rich in mitochondria. This mouse and human study traces that paradox to cardiolipin, a lipid in the inner mitochondrial membrane. Cardiolipin and its synthesis enzyme, Crls1, decline in aged mouse and human muscle. When researchers removed Crls1 from young mouse muscle, the animals developed aging-like features: cristae damage, weaker respiration, smaller fibers, and a shift from fast glycolytic to slow oxidative fibers. A nuclear receptor called ERRγ carried the signal from mitochondria to the nucleus. Restoring Crls1 in adult mice partly rebuilt cardiolipin, began reversing muscle wasting, and fully prevented early death. Cardiolipin loss may be an upstream driver of muscle aging.

Detailed Summary

Aging muscle presents a paradox. Mitochondrial function declines, yet muscle also loses fast glycolytic fibers and shifts toward oxidative, mitochondria-rich ones. Whether mitochondrial decline causes this adaptation or merely accompanies it has been unclear. Finger and colleagues tested whether cardiolipin (CL), a phospholipid nearly exclusive to the inner mitochondrial membrane, links the two. CL shapes cristae and supports respiratory supercomplexes, and its rate-limiting synthase is CRLS1.

CL levels fell with age in both oxidative soleus (SOL) and glycolytic extensor digitorum longus (EDL) muscles of 2-year-old mice versus 9-week-old mice (n=6 per group). Crls1 was the most significantly reduced CL-related gene. Human muscle showed the same pattern: total CL was lower in older participants (53–69 years, n=5) than in young adults (about 22–35 years, n=4), with altered CL species composition and lower expression of CL synthesis genes.

To test causality, the team deleted Crls1 in skeletal muscle (MUCKO mice). CL fell across muscle types, cristae architecture was disrupted, and supercomplex formation declined. Mice failed to gain lean mass despite normal activity and adiposity, and EDL atrophied more than SOL. Citrate synthase activity and COX IV staining showed EDL mitochondrial content rising to SOL levels. Yet respiration per unit of mitochondria fell in both muscles. Overlap with aged-muscle gene changes was 16–28% across up- and downregulated genes in SOL and EDL.

EDL transcriptomes shifted toward a SOL-like profile (2,129 genes changed more than twofold in EDL versus 811 in SOL). Type IIb fibers decreased while IIa and IIx increased. The EDL proteome changed far more (1,380 proteins versus 281 in SOL). Fiber number was unchanged, so smaller fibers, not cell death, drove atrophy. In myotubes, Crls1 knockdown raised MHC I and IIa. Blocking ERRγ by siRNA or GSK5182 abolished this, and 626 of 773 Crls1-regulated genes (about 81%) depended on ERRγ. Myogenin sat downstream. Esrrg was also elevated in aged mouse EDL.

The inducible model (iMUCKO) showed the effects were not developmental. Grip strength fell 5 weeks after induction, before ex vivo force changed. CL dropped weeks after Crls1 mRNA, reflecting CL's long half-life. Stopping tamoxifen raised CL from 27% to 67% of control, began reversing weight loss and atrophy, and fully rescued premature mortality. Constitutive MUCKO mice died around 4 months; diaphragm dysfunction appeared to be the likely driver. The authors also report that the shift promotes ROS-sensitive glucose uptake and glycolytic rerouting to sustain antioxidant defenses.

Implications are mainly mechanistic. CL loss appears sufficient to reproduce several aging hallmarks and is reversible in mice. Caveats include the small human cohort, a deletion that is more severe than natural aging, and the contribution of muscle stem cells to the recovery. Myog expression was not altered in aged mice, so aging is not a perfect copy of the knockout.

Key Findings

  • Total cardiolipin declined in aged (2-year) versus young (9-week) mouse SOL and EDL muscle (n=6), and in human muscle from 53–69-year-olds (n=5) versus young adults (n=4)
  • Crls1, the CL synthase, was the most significantly reduced CL-related gene in aged mouse muscle, with CL synthesis genes also reduced in aged human muscle
  • Muscle-specific Crls1 deletion in EDL raised mitochondrial content to SOL levels but lowered respiration per mitochondrion in both muscles, with fiber number unchanged and fiber size smaller
  • CL loss changed 2,129 genes (more than twofold) in EDL versus 811 in SOL, and 1,380 EDL proteins versus 281 in SOL, with fewer type IIb and more IIa/IIx fibers
  • In myotubes, 626 of 773 Crls1-dependent genes (about 81%) required ERRγ; ERRγ inhibition or knockdown blocked the MHC I increase after Crls1 loss
  • Restoring Crls1 expression raised CL from 27% to 67% of control levels, began reversing weight loss and atrophy, and fully rescued premature mortality in iMUCKO mice
  • Grip strength fell 5 weeks after Crls1 deletion in adult mice, before ex vivo tetanic and specific force changed, showing functional decline precedes force loss

Methodology

The study combined young (9-week) versus aged (2-year) mouse muscle, a small human cohort (young n=4, aged n=5), and constitutive (MUCKO) and tamoxifen-inducible (iMUCKO) muscle-specific Crls1 knockout mice with a tamoxifen-withdrawal recovery arm. Readouts included lipidomics, TEM, citrate synthase activity, respirometry in permeabilized fibers and isolated mitochondria, RNA-seq, proteomics, histology, grip strength, ex vivo force, and whole-body metabolic phenotyping. ERRγ dependence was tested in C2C12 myotubes using siRNA, the inhibitor GSK5182, and reanalysis of published ChIP-seq. Group sizes were mostly n=3–14 per group, and data are shown as mean ± s.e.m. with significance thresholds up to P<0.0001.

Study Limitations

The human comparison was very small (4 young versus 5 aged participants), and most causal evidence comes from mouse genetic models, in which complete Crls1 deletion is more severe than natural CL decline. Cre-based deletion is irreversible, so recovery after tamoxifen withdrawal likely reflects satellite cells supplying non-recombined myonuclei rather than true re-expression. Myog was not transcriptionally altered in aged mice, so the knockout does not fully mirror aging. The supplied text was truncated and did not include the full discussion, limitations, or conflict-of-interest statement.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: