Aberrations in the subsequent generations of HPV16-E6E7-expressing cells after release from APH-induced replication stress. The mechanism for the preference of pericentromeric aberrations is unclear at this stage. The acute effect of APH is known to cause chromatid breaks on newly synthesized chromatids. These chromatid breaks are often interlinked by AB1010 ultra-fine DNA bridge which may facilitate efficient end-joining of the breaks. This is in line with the idea that most of the chromatid breaks in fragile sites are rapidly end-joined. On the other hand, it has been recently discovered that hyper-condensation of chromatin during mitosis enhances DNA breakage in some fragile sites. During mitosis, pericentromeric chromatin is known to be highly condensed. It is possible that this specific feature of pericentromeric chromatin may lead to preferential DNA rupture in pericentromeric regions during mitosis. The broken chromatids in pericentromeric regions may be more difficult to repair through end-joining than non-pericentromeric ends, particularly in cells with defect in DNA damage repair. The duplicated chromatids with pericentromeric rearrangement or breaks were revealed as chromosomal type pericentromeric rearrangements or breaks in the subsequent metaphases. HPV16 E6 is known to inactivate p53, which plays important roles in DNA damage repair. In addition, it was shown that HPV16 E6-expressing cells had lower S-phase recovery rates after DNA damage. Our data in this study also confirmed that HPV16 E6E7-hTERT-expressing cells were deficient in recovering from replication stress-induced S-phase arrest when compared with hTERT-expressing counterparts. HPV16 E6 has been also shown to impair G2 checkpoint. The above information together may, at least in part, explain our finding that pericentromeric rearrangements became the predominant type of chromosome aberrations in the subsequent generations of HPV16 E6E7-expressing cells. In addition to inefficient DNA replication, over-activation of oncogenes or growth signaling pathways, which induces hyperDNA replication, can also cause replication stress and induce fragile site instability. In our study, the expression of HPV16 E6E7 is a typical example of activation of growth signaling pathways. This is because HPV16 E6 and E7 inactivate p53 and Rb, respectively, both of which play essential roles in inhibiting cell proliferation. Intriguingly, our data showed that epithelial cell lines derived from different organ sites consistently exhibited preferential pericentromeric instability upon expression of HPV16 E6E7. It appears that pericentromeric instability plays a more prominent role than nonpericentromeric instability in contributing to gross chromosome aberration formation in HPV16 E6E7-expressing cells. It is relevant to note that pericentromeric or centromeric aberrations have been reported to be a common form of chromosome aberrations in cervical cancers, as well as in many other types of cancer. Since cancer cells commonly face replication stress from the earliest stages of cancer development in vivo, and the inactivation of p53 and/or Rb pathway occurs in most cancers, we infer that our findings in this study may have important implications for genomic instability, particularly pericentromeric instability, in cancer cells. In summary, pericentromeric instability was found to be a general phenomenon in human cells expressing HPV16 E6 and E7, and was enhanced by aphidicolin-induced replication stress in successive cell generations. Since cancer development is associated with replications stress, and inactivation of p53 and Rb pathway is common in cancer cells, our finding that pericentromeric regions are refractory to prompt repair after replication.