A transcriptome-based tool for assessing skin aging reveals the ameliorative effect of stem cell-derived extracellular vesicles on UVB-induced aging.

Quantifying biological aging is essential for understanding functional decline and evaluating anti-aging interventions. We present SkinAGE, a fast and resource-efficient transcriptome-based aging clock built using a deep neural network (DNN) trained on gene expression profiles of human dermal fibroblasts. SkinAGE accurately predicts cellular aging status across independent cohorts and a UVB-induced photoaging model. In 23rd-passage HFF-1 cells, the model assigned an average age score of 44, which increased by 24 units after UVB exposure, indicating enhanced transcriptomic senescence. Treatment with human embryonic stem cell-derived extracellular vesicles (hESC-EVs) reduced this score by an average of 21.2 units, suggesting a potent ameliorative effect on cellular aging. Transcriptomic analysis supported this, showing restoration of aging-related gene expression, including enrichment of cell cycle progression and p53 signaling pathways. Overall, SkinAGE provides a scalable, accurate, and cost-effective framework for quantifying cellular aging and assessing rejuvenation strategies.