Histone modification clocks for robust cross-species biological age prediction and elucidating senescence regulation.

Abstract

Histone modifications represent an untapped resource for biological age prediction that overcomes limitations of traditional DNA methylation-based epigenetic clocks. Here, we developed and validated histone modification-based epigenetic clocks by systematically analyzing publicly available ChIP-seq datasets spanning six tissue types and six histone marks. We identified age-associated loci and constructed 36 tissue-specific epigenetic clocks that demonstrated strong resilience to technical and biological noise, with performance comparable to established DNA methylation clocks. Our models successfully detected biological age acceleration in leukemia samples and captured age reversal following therapeutic interventions. Importantly, we found that many aging-associated loci follow nonlinear trajectories with peak modification levels at midlife, revealing previously unrecognized dynamics in epigenetic aging. We observed age-related fragmentation of super enhancer regions, suggesting progressive chromatin disorganization during aging. Functional validation of a model-selected H3K27ac peak near IGF2BP3 confirmed its causal role in cellular senescence through regulation of TRA2A expression. Extending beyond mammals, we demonstrated the applicability of histone-based clocks in