Age-associated DNA methylation patterns in mouse blood and tail: feasibility of tissue-specific epigenetic clock development
摘要
Aging is accompanied by widespread remodeling of DNA methylation (DNAm), which can be leveraged to build epigenetic clocks. In mice, most DNAm clocks rely on internal organs or terminal tissues, limiting longitudinal studies. Here, we compared age-associated DNAm patterns in matched blood and tail—an easily accessible but underexplored tissue—and evaluated both as substrates for tissue-specific DNAm age prediction. Matched blood and tail were collected from male C57BL/6N Per2::Luc mice aged 7–103 weeks and profiled using the Infinium Mouse Methylation BeadChip. After SeSAMe preprocessing, probe-wise regression (β ~ Age, BH-FDR < 0.05) identified 78,111 age-associated CpGs in blood and 38,348 in tail. Blood showed predominantly hypomethylating changes and a later, more abrupt age shift, whereas tail exhibited a near-balanced hyper/hypomethylation pattern with smoother transitions. Segment-level analysis (BH-FDR < 0.05, ≥ 2 CpGs) identified 31,322 age-related differentially methylated regions (DMRs) in blood and 21,883 in tail. Direction-aware overlap yielded 6229 shared age-related DMRs, characterized by CpG island–enriched hypermethylation and Open Sea–enriched hypomethylation and linked to genes enriched for developmental, neuronal, immune, and signaling pathways. Elastic-net clocks achieved high accuracy in the training cohort (blood: α = 0.8, 72 CpGs, MAE = 2.30 weeks, R2 = 0.99; tail: α = 0.5, 192 CpGs, MAE = 2.75 weeks, R2 = 0.98). The blood clock generalized to an independent female blood dataset (MAE = 13.48 weeks, R2 = 0.76). Together, these findings provide a proof-of-concept foundation for tissue-specific epigenetic age estimation in mice using minimally invasive sampling and highlight tail as a promising substrate for longitudinal DNAm aging studies without terminal tissue collection.