Senescence-associated DNA methylation is stochastically acquired in subpopulations of mesenchymal stem cells

Replicative senescence has a major impact on function and integrity of cell preparations. This process is reflected by continuous DNA methylation (DNAm) changes at specific CpG dinucleotides in the course of in vitro culture, and such modifications can be used to estimate the state of cellular senescence for quality control of cell preparations. Still, it is unclear how senescence-associated DNAm changes are regulated and whether they occur simultaneously across a cell population. In this study, we analyzed global DNAm profiles of human mesenchymal stem cells (MSCs) and human umbilical vein endothelial cells (HUVECs) to demonstrate that senescence-associated DNAm changes are overall similar in these different cell types. Subsequently, an Epigenetic-Senescence-Signature, based on six CpGs, was either analyzed by pyrosequencing or by bar-coded bisulfite amplicon sequencing. There was a good correlation between predicted and real passage numbers in bulk populations of MSCs (R²  = 0.67) and HUVECs (R²  = 0.97). However, when we analyzed the Epigenetic-Senescence-Signature in subclones of MSCs, the predictions revealed high variation and they were not related to the adipogenic or osteogenic differentiation potential of the subclones. Notably, in clonally derived subpopulations, the DNAm levels of neighboring CpGs differed extensively, indicating that these genomic regions are not synchronously modified during senescence. Taken together, senescence-associated DNAm changes occur in a highly reproducible manner, but they are not synchronously co-regulated. They rather appear to be acquired stochastically-potentially evoked by other epigenetic modifications.