Uncoupling the senescent phenotype from telomere shortening in hydrogen peroxide-treated fibroblasts.

Normal human cells have a limited replicative potential and inevitably reach replicative senescence in culture. Replicatively senescent cells show multiple molecular changes, some of which are related to the irreversible growth arrest in culture, whereas others resemble the changes occurring during the process of aging in vivo. Telomeres shorten as a result of cell replication and are thought to serve as a replicometer for senescence. Recent studies show that young cells can be induced to develop features of senescence prematurely by damaging agents, chromatin remodeling, and overexpression of ras or the E2F1 gene. Accelerated telomere shortening is thought to be a mechanism of premature senescence in some models. In this work, we test whether the acquisition of a senescent phenotype after mild-dose hydrogen peroxide (H(2)O(2)) exposure requires telomere shortening. Treating young HDFs with 150 microM H(2)O(2) once or 75 microM H(2)O(2) twice in 2 weeks causes long-term growth arrest, an enlarged morphology, activation of senescence-associated beta-galactosidase, and elevated expression of collagenase and clusterin mRNAs. No significant telomere shortening was observed with H(2)O(2) at doses ranging from 50 to 200 microM. Weekly treatment with 75 microM H(2)O(2) also failed to induce significant telomere shortening. Failure of telomere shortening correlated with an inability to elevate p16 protein or mRNA in H(2)O(2)-treated cells. In contrast, p21 mRNA was elevated over 40-fold and remained at this level for at least 2 weeks after a pulse treatment of H(2)O(2). The role of cell cycle checkpoints centered on p21 in premature senescence induced by H(2)O(2) is discussed here.