Identifying Age-Modulating Compounds Using a Novel Computational Framework for Evaluating Transcriptional Age

The differentiation of human pluripotent stem cells (hPSCs) provides access to a wide range of cell types and tissues. However, hPSC-derived lineages typically represent a fetal stage of development, and methods to expedite the transition to an aged identity to improve modeling of late-onset disease are limited. In this study, we introduce RNAge, a transcriptome-based computational platform designed to enable the evaluation of an induced aging or a rejuvenated state. We validated this approach across independent datasets spanning different tissues and species, and show that it can be used to evaluate the effectiveness of existing age-retaining or age-modulating interventions. We also used RNAge to perform an in silico compound screen using the LINCS L1000 dataset. This approach led to the identification and experimental confirmation of several novel compounds capable of inducing aging or rejuvenation in primary fibroblasts or hPSC-derived neurons. Additionally, we observed that applying this novel induced aging strategy to an hPSC model of Alzheimer's disease (AD) accelerated neurodegeneration in a genotype-specific manner. Our study offers a robust method for quantifying age-related manipulations and unveils compounds that significantly broaden the toolkit for age-modifying strategies in hPSC-derived lineages.