Abstract

Biomarkers of biological age that predict the risk of disease and expected lifespan better than chronological age are key to efficient and cost-effective healthcare1-3. To advance a personalized approach to healthcare, such biomarkers must perform on the individual rather than population level, demonstrate single cell resolution, and provide scalable and cost-effective measurements. We developed a novel approach - image-based chromatin and epigenetic age (ImAge), that utilizes image texture features based on the patterns of chromatin and epigenetic marks in single nuclei. We observed the emergence of intrinsic trajectories of ImAge using dimensionality reduction without regression on chronological age. ImAge was correlated with chronological age in all tissues and organs examined and was consistent with the expected acceleration and/or deceleration of biological age in chronologically identical mice treated with chemotherapy or following a caloric restriction regimen, respectively. ImAge from chronologically identical mice inversely correlated with their locomotor activity (greater activity for younger ImAge), consistent with the essential role of locomotion as an aging biomarker. Finally, we demonstrated that ImAge is reduced upon partial reprogramming in vivo following transient expression of OSKM cassette in the liver and skeletal muscles of old mice and validated the power of ImAge to assess the heterogeneity of reprogramming. We propose that ImAge represents the first-in-class individual-level biomarker of aging and rejuvenation with single-cell resolution.