Fig. 3

Effects of lifespan extension on a mouse epigenetic clock. a, b The 148 CpG sites used in the mouse epigenetic-aging model (used for a mouse epigenetic clock) were subjected to principal component analysis. Principal component 1 is plotted for wild-type mice according to age and lifespan extension status, for wild-type Ames or dwarf mice (a) or wild-type UM-HET3, rapamycin-treated or calorie-restricted mice (b). c The mouse epigenetic-aging model applied to long-lived mice, with colors and shapes representing the different lifespan-enhancing conditions. The gray markers are the wild-type mice (identical to Fig. 2b), and the black line represents the linear fit of the epigenetic age versus chronological age of the wild-type mice. The green line represents the linear fit of the epigenetic age versus chronological age for long-lived mice. The gray dashed line represents the diagonal. d The residual (epigenetic age minus chronological age) is plotted for all mice according to their strain and treatment, and colors represent 2 or 22 months of age. p-values were calculated by comparing ages of long-lived mice to age-matched controls of the same genetic background using a t-test. *p < 0.05; **p < 0.01. e Hierarchical clustering of the top 20 most variable sites used by this epigenetic clock using average linkage with Euclidean distance. Treatment is depicted under the dendrogram, CpG sites are to the right of the heatmap (chromosome:start, 0-based) and rows are blocked according to clusters of sites that increase or decrease methylation with age. m Months, R Rapamycin treatment, C, CR Calorie restriction, D Ames Dwarf, W wild-type Ames or untreated, wild-type UM-HET3