Replicative senescence induction in single cells is not predicted by telomere length, dysfunction, or oxidation.

The accumulation of senescent cells during aging contributes to age-associated diseases. Current models posit that replicative senescence is driven by telomere dysfunction, including telomere shortening, telomere-associated DNA damage response (DDR), and telomere oxidation. Here, we first show that aging primary human fibroblasts gradually increase the time spent in a CDK2-low non-cycling state and increase senescence biomarker expression. We then evaluate telomere features as single-cell senescence biomarkers in a workflow linking high-throughput, long-term time-lapse imaging with confocal imaging to map cell-cycle dynamics to telomere features in the same cell. Our results show that telomere length and DDR do not reliably distinguish cycling from non-cycling cells at any age, and that telomere oxidation is not associated with cell-cycle withdrawal. Instead, lysosomal content, cell size, genomic architecture, and p21 more reliably mark senescence induction, depicting replicative senescence as a complex state transition with currently measurable telomere features being weakly correlated with senescence.