AP-1 Imprints a Reversible Transcriptional Programme of Senescent Cells

Senescent cells affect many physiological and pathophysiological processes. While select genetic and epigenetic elements for senescence induction have been identified, the dynamics, epigenetic mechanisms and regulatory networks defining senescence competence, induction and maintenance remain poorly understood, precluding the deliberate therapeutic targeting of senescence for health benefits. Here, we examined the possibility that the epigenetic state of enhancers determines senescent cell fate. We explored this by generating time-resolved transcriptomes and epigenome profiles during oncogenic RAS-induced senescence and validating central findings in different cell biology and disease models of senescence. Through integrative analysis and functional validation, we reveal links between enhancer chromatin, transcription factor recruitment and senescence competence. We demonstrate that activator protein 1 (AP-1) 'pioneers' the senescence enhancer landscape and defines the organizational principles of the transcription factor network that drives the transcriptional programme of senescent cells. Together, our findings enabled us to manipulate the senescence phenotype with potential therapeutic implications. We performed time-series experiments on ER:RASv12-transduced WI38 fibroblasts treated with 4-hydroxytamoxifen (4OHT) for 144 hours to induce OIS. Total RNA was collected at 6 time points (0, 24, 48, 72, 96 and 144 h, 2 biological replicates), which encompass the basal proliferative (0 h), hyperproliferative (24-48 h), initiation (72-96 h) and establishment (144 h) stages of OIS, and global gene expression levels were determined using Affymetrix HTA2.0 microarrays. We reasoned that inclusion of the more physiologically prevalent cell state of Quiescence in our experimental strategy would allow us to identify the OIS-specific transcriptome profile. Accordingly, WI38 fibroblasts were induced into Quiescence by serum deprivation for 96 hours. Total RNA was collected at 6 time points (0, 12, 24, 48, 72, and 96 h, 2 biological replicates) and processed in parallel with the OIS time-series.