Unexpected role of CDK4 in a G2/M checkpoint

Abstract

Cellular senescence is a state of permanent cell cycle arrest in response to a diverse range of stressors. In physiological settings, senescence has an important role in normal embryonic development and it is also considered as a safeguard mechanism that prevents uncontrolled oncogenic cell proliferation.1 Features of senescence are varied but such cells are commonly metabolically active (despite cell cycle arrest), while displaying a large flattened phenotype and secreting a variety of factors, including extracellular proteases, cytokines, chemokines, and growth factors.1 The large cell phenotype is consistent with an uncoupling between cell cycle progression and cell growth pathways during the senescent response. In this issue of Cell Cycle, Brookes et al.2 continued their studies in examining the mechanisms underlying cellular senescence. In studying a DNA damage induced senescence model of human fibroblasts, they uncovered an unexpected role for cyclin dependent kinase 4 (CDK4) in contributing to a G2/M cell cycle arrest and the senescent phenotype. This function appears inconsistent with the role CDK4 plays in promoting cell proliferation and suppressing senescence.3 Entry into the G1 phase of the cell cycle occurs when CDK4 and CDK6 form active complexes with the D-type cyclins. These complexes phosphorylate the retinoblastoma tumor suppressor protein (pRB) as well as related family members p107 and p130. CDK4/6 phosphorylation of pRB and subsequent phosphorylation by CDK2 inactivates pRB function as a transcriptional repressor and allows the activation of the E2F-dependent transcriptional program leading to S-phase entry and initiation of DNA replication. In addition to the RB proteins, the forkhead box M1 (FOXM1) transcription factor is another target of CDK4/6 kinase activity that is required for maintaining the expression of the G1/S phase genes and protecting cancer cells from senescence.4 Senescent cells exhibit elevated levels of the CDK4/6 inhibitor p16INK4a and p21CIP1. p16INK4a directly interacts with CDK4 and CDK6 blocking their association with the D-type cyclins, while, the CIP/KIP (kinase inhibitor protein) family stimulates the assembly and nuclear localization of the cyclin D-CDK4/6 complexes. In contrast to the stimulatory role p21CIP1 has on CDK4/6 activity, p21CIP1 inhibits all other cyclin-CDK complexes. Furthermore the sequestration of p21CIP1 by CDK4/6 prevents binding to CDK2 and thus prevents inhibition of cyclin E-CDK2 complexes. During senescence, the binding of p16INK4a to CDK4/6 redistributes p21CIP1 and p27KIP1 from the cyclin D-CDK4/6 complexes onto the cyclin E-CDK2 complexes. Thus, the p21CIP1/p27KIP1/cyclin D-CDK4/6 complexes serve as a buffering system that controls the availability of the CIP/KIP proteins to inhibit CDK2 activity.5 It remains less clear whether these CDK4/6 complexes are catalytically active in vivo. The data in Brookes et al.,2 demonstrate that the cyclin D-CDK4-p21CIP1 complex, which assembles during the transition to senescence due to increased p21CIP1 levels, is catalytically functional and contributes to the senescent phenotype. shRNA-mediated knockdown or chemical inhibition of CDK4 prevented the increase in cell size associated with the senescent phenotype by allowing the cells to progress from the G2/M block and then arrest in G1. This suggests that CDK4 kinase activity is required for phosphorylating an as yet unknown substrate to arrest cells in G2M. This novel role for CDK4 may provide a safeguard mechanism against the proliferation of oncogenically compromised cells that acquired the capability of bypassing the G1 checkpoint. Furthermore, this data suggests that this CDK4 complex has specificity in substrate recognition because the well-know CDK4 substrate RB was not phosphorylated. In recent years, highly specific CDK4/6 drug inhibitors have been tested in clinical trials for various cancers including breast, mantle cell lymphoma and melanoma. They induce G1 arrest that is highly dependent on the presence of pRB and often associated with a senescent phenotype.3,6 A screening study for therapeutic drugs that may cooperate with CDK4/6 inhibitors in the treatment of pancreatic ductal adenocarcinoma, revealed that agents dependent on mitotic progression including the polo-like kinase inhibitors and taxanes demonstrate an antagonistic interaction with CDK4/6 inhibitors.7 The data presented in Brookes et al.,2 demonstrating bypass of the G2M checkpoint in response to CDK4/6 inhibition provides a clear mechanism to explain the antagonistic interaction with drugs targeting the mitotic machinery. These findings suggest that a better understanding of the cell cycle is required for guiding the selection of combination treatment with inhibitors of the cell cycle in general and CDKs in particular.