Decay in transcriptional information flow is a hallmark of cellular aging

Abstract

Aging is marked by the progressive loss of cellular function, yet the organizing principles underlying this decline remain unclear. Although molecular fingerprints of aging are diverse, many converge on disruption of the interrelated and overlapping communication networks that coordinate molecular activity. Here, we apply information theory to quantify age-related corruption in gene regulation by modeling regulatory interactions between transcription factors (TFs) and their target genes (TGs) as a multi-input multi-output communication channel. Using an analytically tractable probabilistic model and single-cell RNA-sequencing data from multiple tissues, we find that the mutual information (a measure of information transfer) between TFs and TGs declines with age across all ten tissues analyzed, establishing loss of regulatory information transmission as a hallmark of aging. Structural analysis of the regulatory network reveals that aging degrades communication primarily through input distribution mismatch, reflecting a loss of coordinated TF activity, rather than channel corruption, or the inability of TFs to reliably activate or inhibit their targets. This mismatch is caused by increased network centralization and loss of stabilizing feedback motifs, leading to reduced robustness to random perturbations. Notably, in silico upregulation of a small set of TFs restores youthful information transfer and gene expression levels, suggesting that targeted reinforcement of key regulatory nodes may rejuvenate aged networks.