Mathematical Models of Vulnerability and Cell-Type Specific Analysis of DNA Modifications in Aging
Project Summary Understanding the biochemical processes underlying aging that give rise to age-related pathologies is the central concept of geroscience research. A focus in geroscience is to identify interventions that slow/prevent aging. These could both alter the trajectory of aging and have beneficial effects across a number of diseases, delaying onset or slowing progression. However, evaluating anti-aging interventions requires years-long studies in mammalian model organisms and is almost impossible in clinical trials that would require of multi-year studies to evaluate lifespan and healthspan extensions. The solution to this problem lies in developing endpoints that indicate biological age and the health-related changes that go with it with sufficient sensitivity. The recent development of epigenetic clocks, chronological age-predictive machine learning models that use DNA cytosine methylation data, are leading the field in age estimation accuracy and precision. The difference between a person?s chronological age and the estimated ?methylation age? has been proposed as a measure of aging acceleration or deceleration ? a ?biological age?. Age acceleration predicted by methylation has shown robust correlations to all-cause mortality risk, but weak correlations in age-related diseases. Age-related biological outcomes such as cognitive decline have found little to no association with methylation age acceleration. Put simply, epigenetic clocks have been developed to predict chronological age but have so far demonstrated marginal utility to predict disease states. However, our preliminary results indicate that as many as 20% of genomic cytosines have age-related methylation changes which may be leveraged to understand biological aging and its outcomes simultaneously. Using the largest collection of methylation data with disease, age, tissue, and sex labels from our previous work in natural language processing, we will create the first model of epigenetic aging to identify ?healthy? aging loci and disease-predictive loci. These loci will give us insight into the molecular pathways disturbed by age-related methylation changes, providing targets for therapeutic intervention and a predictor for patient disease risk. However, it is still unknown how epigenetic clocks ?tick?, especially considering their performance across tissues. We will test how epigenetic changes with age are distributed throughout specific cell types in the central nervous system and cell types common across all tissues ? such as vascular endothelium and immune cells. Using single-cell bisulfite sequencing and cell-type specific promoter-driven labelling, we can detect if the epigenetic clock changes are truly occurring in all cells or if they are restricted to some cell types common across tissues. This information is critical for predicting the downstream effects of changes identified by epigenetic clocks and interpreting the effects of ?reversing? epigenetic aging. These experiments will further our understanding of the role epigenetics play in the aging process and in neurological diseases of age such as Alzheimer?s and Parkinson?s Disease.