Long-term studies using an extensive human mammary epithelial cell (HMEC) culture system and molecularly defined nomenclature have generated new models of the barriers that prevent unlimited replication of cells derived from normal human tissues. Our data support a model whereby HMEC encounter two mechanistically distinct proliferation barriers. The first barrier, stasis, is a consequence of accumulated stresses, is mediated by the retinoblastoma (RB) pathway, and is largely or fully telomere length independent. The onset of stasis in HMEC is correlated with increased levels of the CKI p16INK4a, but not p21Cip1or p14ARF, and may be associated with chromatin modifications and altered gene expression. Depending upon culture conditions, HMEC can proliferate for 10-60 population doublings prior to elevation of p16 and proliferative arrest. In HMEC, stasis can be overcome by alterations in pathways governing RB and does not require loss of p53. However, different cells may vary in perception and responses to specific stresses, with some cell types (e.g., keratinocytes) using p53-dependent p21 to prevent RB inactivation and to enforce stasis. We propose that stasis most resembles what is usually called senescence or M1, as well as “culture shock”. A second extremely stringent barrier is imposed by critically shortened telomeres producing telomere dysfunction. Where wild-type p53 is present, this barrier has been termed agonescence, and produces a mostly viable growth arrest. If the second barrier is approached with non-functional p53, then crisis, rather than agonescence, occurs. Reactivation of telomerase activity is needed to overcome the telomere dysfunction barrier. Low levels of hTERT expression and telomerase activity can be detected in some pre-stasis HMEC populations, but not in post-stasis HMEC that encountered stasis and then overcame it, associated with silencing of p16; telomerase reactivation in these cells may require multiple errors, contributing to the stringency of this barrier. Non-proliferative cells at both stasis and agonescence express senescence-associated β-galactosidase activity and a senescent morphology, but can be distinguished by other molecular markers (e.g., labeling index, presence or absence of karyotypic abnormalities, 2N to 4N ratio). Our model may offer new insights about the mechanisms involved in cancer etiology, and potential therapeutic interventions. It is consistent with what is observed in vivo, e.g., overcoming stasis may correlate with early clonal expansion/atypical hyperplasia, whereas the phenotype of cells approaching agonescence is similar to what is seen in DCIS (very short telomeres, genomic instability), where telomerase reactivation can be first detected. Our ability to generate this model was facilitated by the use of one cell type, thereby avoiding potentially confounding variables due to species and cell type diversity.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]