Integration of transcriptomes of senescent cell models with multi-tissue patient samples reveals reduced COL6A3 as an inducer of senescence

Senescent cells are a major contributor to age-dependent cardiovascular tissue dysfunction, but knowledge of their in vivo cell markers and tissue context is lacking. To reveal tissuerelevant senescence biology, we integrate the transcriptomes of 10 experimental senescence cell models with a 224 multi-tissue gene co-expression network based on RNA-seq data of seven tissues biopsies from-600 coronary artery disease (CAD) patients. We identify 56 senescence-associated modules, many enriched in CAD GWAS genes and correlated with cardiometabolic traits-which supports universality of senescence gene programs across tissues and in CAD. Cross-tissue network analyses reveal 86 candidate senescence -associated secretory phenotype (SASP) factors, including COL6A3. Experimental knockdown of COL6A3 induces transcriptional changes that overlap the majority of the experimental senescence models, with cell-cycle arrest linked to modulation of DREAM complextargeted genes. We provide a transcriptomic resource for cellular senescence and identify candidate biomarkers, SASP factors, and potential drivers of senescence in human tissues.