DNA methylation correlates with lifespan and predicts remaining lifespan in the zebra finch

Abstract

Biomarkers that correlate with age are essential tools for understanding aging and lifespan variation. DNA methylation (DNAm) changes predictably with age in parts of the genome. In humans, epigenetically 'old' individuals relative to their chronological age also have a reduced life expectancy but whether the link between epigenetic age and lifespan is a general feature remains an open question. We explored age-related changes in DNAm in the zebra finch (Taeniopygia castanotis), a key avian model species, using 100 longitudinal whole genome methylomes from 50 captive adults monitored until their natural death. We found genome-wide hypomethylation with age, with DNAm decreasing faster in individuals with shorter lifespans and identified 29 CpG sites where DNAm changed significantly with age. We developed an epigenetic aging clock based on 119 CpG sites, that predicted chronological age with high accuracy (median absolute deviation=0.68 years, 8.2% of maximum lifespan in our dataset). Females raised in large broods, which have shorter lifespans, showed increased epigenetic age acceleration, consistent with faster biological aging. However, epigenetic age acceleration did not predict lifespan or remaining lifespan. In contrast, the within-individual rate of change in epigenetic age significantly predicted lifespan: faster epigenetic aging was associated with a shorter lifespan. Moreover, a 'doom' clock, trained to predict post-sampling lifespan, successfully predicted remaining lifespan. Our findings provide the first evidence that DNAm provides a window into biological aging in birds, showing how early-life environments shape the aging trajectory via the epigenome and underscore the value of longitudinal data in aging studies.