Early-passage human diploid fibroblasts (HDFs) undergo senescence-like growth arrest in response to sublethal concentrations of H2O2 [Chen and Ames (1994) Proc. Natl. Acad. Sci. U.S.A. 95, 4130-4134]. We determine here whether H2O2 can cause apoptosis in HDFs and the molecular changes that differ between apoptosis and senescence-like growth arrest. When exponentially growing early-passage IMR-90 cells were treated for 2 h with 50-200 μM (or 0.25-1 pmol/cell) H2O2, a fraction of cells detached at 16-32 h after the treatment. The cells remaining attached were growth-arrested and developed features of senescence in 1 week. The detached cells showed caspase-3 activation and typical morphological changes associated with apoptosis. Caspase-3 activation was H2O2 dose-dependent and preceded nuclear condensation or plasma membrane leakage. Apoptotic cells were mainly distributed in the S-phase of the cell cycle, while growth-arrested cells exhibited predominantly G1- and G2/M-phase distributions. H2O2 pretreatment induced G1 arrest and prohibited induction of apoptosis by a subsequent H2O2 challenge. The p53 protein showed an average 6.1-fold elevation in apoptotic cells and a 3.5-fold elevation in growth-arrested cells. Reduction of p53 levels with human papillomavirus E6 protein prohibited the activation of caspase-3 and decreased the proportion of apoptotic cells. Growth-arrested cells had elevated p21, while p21 was absent in apoptotic cells. Bcl-2 was elevated in both growth-arrested and apoptotic cells. Finally, although the overall level of bax did not change in growth-arrested or apoptotic cells, the solubility of bax protein increased in apoptotic cells. Our data suggest that in contrast with growth-arrested cells, apoptotic cells show an S-phase cell cycle distribution, a higher degree of p53 elevation, an absence of p21 protein and increased solubility of bax protein.

Early-passage human diploid fibroblasts (HDFs) undergo senescence-like growth arrest in response to sublethal concentrations of H(2)O(2) [Chen and Ames (1994) Proc. Natl. Acad. Sci. USA. 95, 4130-4134]. We determine here whether H(2)O(2) can cause apoptosis in HDFs and the molecular changes that differ between apoptosis and senescence-like growth arrest. When exponentially growing early-passage IMR-90 cells were treated for 2 h with 50-200 microM (or 0.25-1 pmol/cell) H(2)O(2), a fraction of cells detached at 16-32 h after the treatment. The cells remaining attached were growth-arrested and developed features of senescence in 1 week. The detached cells showed caspase-3 activation and typical morphological changes associated with apoptosis. Caspase-3 activation was H(2)O(2) dose-dependent and preceded nuclear condensation or plasma membrane leakage. Apoptotic cells were mainly distributed in the S-phase of the cell cycle, while growth-arrested cells exhibited predominantly G1- and G2/M-phase distributions. H(2)O(2) pretreatment induced G1 arrest and prohibited induction of apoptosis by a subsequent H(2)O(2) challenge. The p53 protein showed an average 6.1-fold elevation in apoptotic cells and a 3.5-fold elevation in growth-arrested cells. Reduction of p53 levels with human papillomavirus E6 protein prohibited the activation of caspase-3 and decreased the proportion of apoptotic cells. Growth-arrested cells had elevated p21, while p21 was absent in apoptotic cells. Bcl-2 was elevated in both growth-arrested and apoptotic cells. Finally, although the overall level of bax did not change in growth-arrested or apoptotic cells, the solubility of bax protein increased in apoptotic cells. Our data suggest that in contrast with growth-arrested cells, apoptotic cells show an S-phase cell cycle distribution, a higher degree of p53 elevation, an absence of p21 protein and increased solubility of bax protein.

The tumor suppressor p53 protein is known to play a critical role in apoptosis. In normal human diploid fibroblasts (HDFs), expression of the human papillomaviral (HPV) E6 gene results in a reduction of p53 protein and an inhibition of oxidant induced apoptosis within 24 h. In comparison, expression of the HPV E7 gene causes down-regulation of Rb protein without inhibiting apoptosis. Here we determine whether HDFs expressing E6 undergo cell death with a delayed time course following H 2 O 2 exposure. Appearances of caspase-3 activity, cell detachment, trypan blue uptake and aberrant nuclei were all delayed in E6 cells compared to wild type (wt) or E7 cells. A mutant E6 gene that failed to reduce p53 could not delay cell death. Morphological examination revealed nuclear condensation in dying wt or E7 cells but nuclear fragmentation in E6 cells. Flow cytometry analysis indicated an S phase distribution of dying wt or E7 cells but a G2/M phase distribution of dying E6 cells. An elevation of cyclin B was observed in dying E6 cells but not in apoptotic E7 cells. Dying E6 cells also had elevated levels of cdc-2 protein and histone kinase activity, suggesting that the cells died at mitosis. Electron microscopy studies showed that E6 cells may die at prophase or prometaphase. Overexpression of bcl-2 resulted in an inhibition of both caspase-3 and death of E7 or E6 cells. Inactivating caspases with zVAD-fmk also reduced the death rate of E7 and E6 cells. Our data indicate that expression of HPV E6 causes a delay and morphological modification of cell death induced by oxidants. E6 cells die at mitosis, which can be inhibited by bcl-2 overexpression or caspase inhibition.