Confluence-induced cell cycle exit involves pre-mitotic CDK inhibition by p27(Kip1) and cyclin D1 downregulation

Tissue homeostasis requires precise control of cell proliferation and arrest in response to environmental cues. In situation such as wound healing, injured cells are stimulated to divide, but as soon as confluence is reached proliferation must be blocked. Such reversible cell cycle exit occurs in G(1), requires pRb family members, and is driven by p27(Kip1)-dependent Cdk inactivation. This implies that, while dividing, cells should simultaneously prepare the exit once mitosis is accomplished. For a long time, the decision to cycle or not was presumed to occur in G(1), prior to the restriction point, beyond which the cells were bound to divide even in the absence of mitogens, before finally arresting after mitosis. However, more recent reports suggested that the commitment to cycle in response to serum occurs already in G(2) phase and requires the Ras-dependent induction of cyclin D1, which promotes following G(1)/S transition. To test whether this hypothesis applies to arrest induced by contact inhibition, we used an in vitro wounding model where quiescent human dermal fibroblasts, stimulated to proliferate by mechanical injury, synchronously exit cell cycle after mitosis due to renewed confluence. We show that this exit is preceded by p27-dependent inhibition of cyclin A-Cdk1/2, cyclin D1 downregulation and reduced pre-mitotic pRb pocket protein phosphorylation. Overexpression of cyclin D1 but not p27 depletion reversed this phenotype and compromised confluence-driven cell cycle exit. Thus, a balance between cyclin D1 and p27 may provide sensitive responses to variations in proliferative cues operating throughout the cell cycle.