Have you climbed Mount Everest? Recorded a critically acclaimed album?? Written the great American novel??? These life-affirming feats of body and mind pale into insignificance when compared to the evolutionary and epigenetic tour-de-force reported in two recent papers based on mammalian DNA methylation datasets of colossal size and importance.
An expeditionary force led by Steve Horvath (University of California Los Angeles/Altos Labs) now describes the analysis of this treasure trove of data in articles. First, Haghani, Li, and colleagues use “phyloepigenetic trees” to show that DNA methylation divergence parallels genetics through evolution and link specific DNA methylation patterns to traits such as maximum lifespan. Then, Lu, Fei, Raj, Horvath, and colleagues report on the evolutional conservation of mammalian aging, the intertwined nature of aging and development, and the creation of pan-mammalian epigenetic clocks.
Evolutionary Pressures and Selection on DNA Methylation Patterns Impacts Mammalian Biological Traits
“Pressure” and “selection” might remind you of waiting to be picked during gym class; however, our first study used their epigenetic bounty to look further back into evolutionary history to resolve relationships between DNA methylation and traits such as mammalian lifespan. The team employed comparative epigenomics (combining epigenetic signatures with phylogenetic relationships) using a dataset derived from ~15,000 samples from 348 mammalian species (from 25 of the 26 mammalian taxonomic orders) covering highly conserved DNA sequences at high effective sequencing depth to understand how evolutionary pressures and selection acting on epigenetics impacts biological traits in mammals.
Let’s hear from Haghani, Li, and colleagues on how evolution has molded DNA methylation in mammals:
- Tissue-specific phyloepigenetic trees based on DNA methylation show similarities to traditional phylogenetic trees, with any resemblance deriving from differences in methylation levels at intergenic regions that are not confounded by underlying sequences
- An unsupervised weighted correlation network analysis (WGCNA) identifies modules (clusters) of co-methylated CpGs and reveals to what extent DNA methylation underpins specific biological traits
- 30 of the 55 identified co-methylation modules display significant associations with species traits (taxonomic order, maximum lifespan, and average adult weight) or individual traits (chronological age, tissue, and sex)
- The DNA methylation status of specific modules responds to multiple lifespan-extending interventions
- Epigenome-wide association and eigengene-based analysis reveal DNA methylation signatures of maximum lifespan, which display independence from aging (presumably set at birth) and may represent predictors of lifespan
- Long-lived species develop DNA methylation patterns with unique peaks/troughs
- Highly methylated CpGs in long-lived species locate near the HOXL subclass of homeobox genes, morphogenesis/development-associated genes, or genes implicated in the upstream regulator analysis (see below)
- CpGs with DNA methylation levels inversely related to lifespan display enrichment in chromatin states associated with regulatory regions, with associated genes constitutively activated and enriched for nucleic acid metabolic processes
- Long-lived species may evolve mechanisms to maintain low DNA methylation at said chromatin states to favor the higher expression of genes essential for organism survival
- Upstream regulator analysis of the epigenome-wide association analysis of lifespan identifies pluripotency transcription factors (OCT4, SOX2, and NANOG), suggesting that they may control DNA methylation levels
- The transient overexpression of pluripotency transcription factors in mouse tissues affects DNA methylation near genes implicated by the epigenome-wide association analysis of maximum lifespan
- The enhanced activity of the pluripotency network in long-lived species may result in more efficient tissue repair and maintenance to ensure a longer lifespan
Overall, this DNA methylation masterpiece reveals the intertwined co-evolutionary history of the genome and epigenome, with the latter mediating the biological characteristics and traits of distinct mammalian species. Moreover, this data trove should provide a rich information resource for fields that include evolutionary biology and longevity research.
More than Just Random Damage: The Evolutionary Conservation of Epigenetic Aging in Mammals
Aching muscles and creaking bones may suggest that accumulated cell and tissue damage drive mammalian aging; however, the development of epigenetic clocks that accurately track biological age in humans and mice suggests that DNA methylation changes represent an evolutionary conserved and integral aging-associated mechanism. The second of our studies sought to undertake the back-breaking work of identifying and characterizing shared CpGs with methylation levels that change with age from a dataset encompassing 11,754 samples from 59 tissue types, originating from 185 species across 19 taxonomic orders with ages covering prenatal stages to 139 years. Their discovery of consistent age-related alterations in methylation profiles across mammals suggests that aging does not simply occur due to the random accumulation of cellular damage and instead represents an evolutionarily conserved process.
Let’s hear from Lu, Fei, Raj, Horvath, and Colleagues on the evolutionary conservation of epigenetic aging in mammals:
- The generated dataset enables the development of DNA methylation-based epigenetic clocks that provide continuous readouts of age from early development to old age in mammals and estimate tissue age with high accuracy
- Said clocks become slowed by conditions delaying growth and/or development, while human epidemiological studies and mouse interventional studies show that the pan-mammalian clocks relate to mortality risk
- A set of CpG sites in DNA sequences conserved across mammals consistently gains methylation during aging and locate to PRC2-binding sites and bivalent chromatin states that regulate the expression of development-associated genes (highly conserved biological processes)
- CpG sites that lose methylation with age associate with genes linked to circadian rhythms and mitochondria, whose functions progressively erode with age
- Age-correlated CpG sites locate to regions that lose accessibility with age in differentiated but not progenitor cells, suggesting that DNA methylation instigates chromatin compaction and hinders PRC2 access to target sites
- Overlaps between PRC2-binding sites and positively age-related DNA methylation changes occur more robustly in proliferative than in non-proliferative tissues and are more pronounced for age-related losses of methylation
Overall, the strenuous analysis of this data gold mine, culminating in the development of pan-mammalian DNA methylation-based epigenetic clocks, suggests the evolutionary conservation of a pseudo-programmed aging process intimately linked to development across all mammals. Furthermore, the central nature of PRC2 (also present in non-mammals) provides evidence that this age-related epigenetic process may occur in all vertebrates.
The View from the Authors
Jason Ernst, who contributed to both articles, said, “The technology we designed to measure DNA methylation levels across mammals along with the tissue sample contributions from a large consortium of researchers led to the production of a highly unique data set, which, when analyzed with advanced computational and statistical tools, unveiled a deeper understanding of the relationship between DNA methylation, lifespan, aging, and other biological processes across mammals.”
If you are thirsty for more details on this epigenetic tour-de-force, head over to the Mammalian Methylation Consortium website. For additional sensational information on how evolutionary pressures and selection impact methylation in mammals, see Science, August 2023; and for more on the evolutionary conservation of epigenetic aging in mammals, see Nature Aging, 2023.