Introduction: Senescent cells have been implicated recently as drivers of cardiac and pulmonary fibrosis. The mechanism how these cells contribute to pathologic chronic tissue remodeling is unknown.  

Methods: We generated senescent human fibroblasts (MRC5) in two different senescence modalities: DNA-damage induced senescence via bleomycin (DS) and replicative senescence (RS), within multiple time points and in vitro conditions. We assessed classical markers of senescence (p16, p21, LMNB-1, TNFRSF10C, SA-ß-galactosidase staining, γH2AX, nuclear and cell morphology), inflammation (IL-1ß, IL-6) and fibroblast activation (ACTA2, COL1A1, COL3A1, CTGF, MMP2, SMAD4, SMAD7, migration assay) on mRNA, protein and functional level. To identify the molecular machinery driving the secretion of extracellular matrix factors, we performed a LC-MS of senescent fibroblasts. To evaluate paracrine effects of the secretome, we have treated non-senescent fibroblasts with supernatants of non-senescent or DS fibroblasts.

Results: RS and DS had lower expression levels of classical fibroblast activation markers than non-senescent fibroblasts, even after TGF-ß stimulation. However, senescent fibroblasts showed significant levels of matrix production. These cells persistently secreted a plethora of collagen types and other extracellular matrix remodeling and coagulation cascade factors. Secretion of collagens could not be modified through treatment with current clinically approved anti-fibrotic drugs nintedanib and pirfenidone. In silico analysis showed abnormalities in RNA-metabolism as a basis of the senescent cell phenotype.  

Conclusion: We have identified an unusual pro-fibrotic aspect of senescent fibroblasts. This phenotype cannot be targeted with contemporary anti-fibrotic drugs. Therefore, therapeutics specifically targeting the senescence machinery are necessary to fully address fibrosis in heart failure.