Making an early decision

The reversible switch between cell proliferation and cell quiescence is an important element of tissue homeostasis. Whereas the mechanisms of re-entry to the cell cycle after mitogen starvation — which involve cyclin-dependent kinase (CDK) activity — have been well studied, investigations of cell cycle commitment in proliferating cells have been hindered by a dependence on the bulk analysis of synchronized cell populations. Meyer and colleagues describe the use of a fluorescent sensor to monitor CDK2 activity in single cells of unperturbed cultures. They show that cells commit to proliferation or quiescence at the end of the previous cell cycle.

cytoplasmic to nuclear ratio of fluorescence intensity can thus be used as an indicator of CDK activity

The fluorescent sensor consists of a segment of human DNA helicase B that contains CDK phosphorylation sites and nuclear localization and export signals fused to the yellow fluorescent protein mVenus (DHB-Ven). In a stably transfected human mammary epithelial cell line (MCF10A), DHB-Ven translocates from the nucleus to the cytosol as the cell cycle progresses as a result of CDK-mediated phosphorylation. The cytoplasmic to nuclear ratio of fluorescence intensity can thus be used as an indicator of CDK activity. Further experiments showed that DHB-Ven is phosphorylated mainly by CDK2 and can be used to monitor changes in CDK2 activity throughout the cell cycle, as validated in individual cells emerging from mitogen starvation.

Single-cell analysis of cycling MCF10A cells showed that in ∼75% of cells (referred to by the authors as CDK2^inc cells), CDK2 activity begins to increase from an intermediate level immediately after mitosis; these cells had short intermitotic times of 16–20 hours. The remaining 25% of cells entered a reversible state of low CDK2 activity (CDK2^low cells) and longer intermitotic times but could increase their CDK2 activity to re-enter the proliferative state. A threshold cytoplasmic to nuclear ratio of DHB-Ven of 0.9–1.1 at 2 hours after anaphase could predict whether a cell would become CDK2^inc or CDK2^low. This bifurcation of cell fate at mitotic exit was supported by a differential dependence on mitogen signalling; MEK inhibition for 1–3 hours after anaphase blocked the reactivation of CDK2^low cells, whereas CDK2^inc cells no longer required MEK activity to complete the cell cycle.

So, CDK2^inc cells exit mitosis with residual CDK2 activity above a minimal threshold, which leads to increasing CDK2 activity and cell cycle progression. To examine the molecular basis of the residual CDK2 activity, the authors compared endogenous levels of cell cycle proteins in CDK2^inc and CDK2^low cells within the first 2–3 hours after anaphase. Levels of the CDK inhibitor p21 were increased in CDK2^low cells compared with CDK2^inc cells, in which p21 levels remained low throughout the cell cycle. In support of a direct role for p21 in causing cells to become CDK2^low after mitosis, p21-deficient MCF10A cells rarely became CDK2^low. Also, the rapid induction of exogenous p21 protein in p21-deficient cells immediately prior to mitosis could increase, in a dose-dependent manner, the percentage of CDK2^low cells at exit from mitosis.

That the decision between quiescence and proliferation might occur at the end of the previous cell cycle was supported by experiments showing that mitogen withdrawal between 0 and 8 hours before mitosis led to a decreased probability of entering the CDK2^inc proliferative state, whereas mitogen withdrawal after mitosis resulted in a similar percentage of CDK2^inc cells as in control populations. This result conflicts with the classic model of cell cycle commitment in late G1 and emphasizes the importance of analysing asynchronously cycling single cells.