Frailty Leaves Distinct DNA Methylation Marks Beyond Normal Aging in Mice

Frailty is one of aging's most consequential syndromes, yet its molecular underpinnings—especially those distinct from normal aging—remain poorly understood. This longitudinal study aimed to separate the epigenetic fingerprint of frailty from that of chronological aging, and to explore whether these signatures differ between sexes.

The team studied 40 female and 49 male C57BL/6JNIA mice at up to five time points, measuring both frailty index (FI) and genome-wide DNA methylation from peripheral blood mononuclear cells (PBMCs) using a methylation array covering ~285,000 CpG sites. After quality filtering, 242,599 CpGs were retained. An untreated discovery cohort of 45 mice (153 samples) was used for primary analyses, with treated mice used for validation.

An epigenome-wide association study (EWAS) identified 61,112 CpGs associated with frailty index. These were aggregated into 1,882 differentially methylated regions (DMRs), then further filtered to 629 high-quality DMRs. Of these, 384 were classified as frailty-related (fiDMRs): 47 age-independent (aiDMRs), 197 age-dependent (adDMRs), and 140 dual (dualDMRs). Crucially, age-independent frailty DMRs were enriched in neuroactive ligand-receptor interaction and insulin resistance pathways, while age-dependent frailty DMRs mapped to lipid metabolism and cellular signaling pathways—demonstrating that frailty has an epigenetic identity beyond aging.

Sex-stratified analyses revealed striking dimorphisms. Females had 771 frailty-related DMRs enriched primarily in signaling and neuronal activity pathways, with notable involvement of imprinted genes like GNAS and Igf2r. Males had 397 fiDMRs, with age-independent frailty signals linked to aldosterone-regulated sodium reabsorption and IL-27 signaling, and age-dependent signals tied to lipid metabolism and cytochrome P450 pathways. Of 903 total sex-stratified DMRs, 265 were shared between sexes—all classified as age-dependent—while 506 were female-specific and 132 were male-specific. The researchers also identified 13 X-chromosome inactivation escape CpGs associated with frailty in females, linked to neuronal gene Slitrk4 and DNA-binding protein Bclaf3. A joint set of 925 frailty-related DMRs was assembled, and epigenetic frailty clocks built from these features were validated in mice subjected to aging interventions, with 36 DMRs showing consistent associations across validation cohorts.

These findings carry important implications: frailty is not simply accelerated aging at the epigenetic level but involves distinct biological pathways. The sex-specific signatures suggest that interventions targeting frailty may need to be tailored by sex. The epigenetic frailty clocks developed here could serve as tools for tracking frailty progression and testing interventions. However, as a mouse study using blood-derived cells, direct translation to human frailty requires further validation, and PBMC methylation may not reflect tissue-specific changes driving frailty phenotypes.