The history and future of targeting cyclin-dependent kinases in cancer therapy

Abstract

Cancer represents a pathological manifestation of uncontrolled cell division; therefore, it has long been anticipated that our understanding of the basic principles of cell cycle control would result in effective cancer therapies. In particular, cyclin-dependent kinases (CDKs) that promote transition through the cell cycle were expected to be key therapeutic targets because many tumorigenic events ultimately drive proliferation by impinging on CDK4 or CDK6 complexes in the G1 phase of the cell cycle. Moreover, perturbations in chromosomal stability and aspects of S phase and G2/M control mediated by CDK2 and CDK1 are pivotal tumorigenic events. Translating this knowledge into successful clinical development of CDK inhibitors has historically been challenging, and numerous CDK inhibitors have demonstrated disappointing results in clinical trials. Here, we review the biology of CDKs, the rationale for therapeutically targeting discrete kinase complexes and historical clinical results of CDK inhibitors. We also discuss how CDK inhibitors with high selectivity (particularly for both CDK4 and CDK6), in combination with patient stratification, have resulted in more substantial clinical activity.

Figure 1. Progression of the cell cycle driven by CDKs

Mitogenic signals stimulate cyclin-dependent kinase 4 (CDK4) and CDK6 and promote entry into the cell cycle, whereas antiproliferative checkpoints inhibit CDK4 and CDK6 activity or induce the expression of the CDK4 and CDK6 inhibitor p16^INK4A. Active CDK4 and CDK6 complexes initiate the phosphorylation (P) of key substrates, including the tumour suppressor retinoblastoma protein (RB), thereby unleashing a gene expression programme that is coordinated by the E2F family of transcription factors. In this context, CDK4 and CDK6 initiate transcription and stability of E-type and A-type cyclins (CycE and CycA, respectively) and the subsequent activation of CDK2 that contributes to the further phosphorylation of RB and the initiation of DNA replication. Further checkpoints can directly inhibit CDK2 activity or induce the CDK-interacting protein/kinase inhibitory protein (CIP/KIP) class of inhibitors (p21^CIP1 and p27^KIP1) that bind to and inhibit CDK2–cyclin complexes. With the completion of DNA replication, CDK1–Cyc A and CDK1–Cyc B complexes form to phosphorylate targets in G2 phase. In the absence of DNA damage and following appropriate preparation for chromosomal segregation, the cellular default is to activate CDK1–CycB complexes and progress into mitosis. However, there are potent checkpoints that limit CDK1 activity and halt mitotic progression. Subsequent degradation of CycB is required for anaphase progression and the production of two daughter cells in G1 phase of the cell cycle. During this transition from M phase back into G1 phase, RB is dephosphorylated, and the cycle is once more responsive to mitogenic and antiproliferative signalling. CCN, cyclin; CDC, cell division cycle; CDT1, chromatin licensing and DNA replication factor 1; MAD2L1, MAD2 mitotic arrest deficient-like 1; MCM, minichromosome maintenance complex component; PLK1, polo-like kinase 1.

Figure 2. G1–S regulatory modules and relevance to cancer

Control over the G1–S transition is coordinated by distinct regulatory modules that are dysregulated in cancer. a | Initially, mitogenic signals impinge on cyclin-dependent kinase 4 (CDK4) or CDK6 activity through multiple mechanisms, including the induction of cyclin D1 (CycD1) gene (CCND1) expression, protein stability and assembly of the CDK–Cyc complex. These steps can be individually antagonized, or the induction of CDK4 and CDK6 inhibitors (that is, the inhibitor of CDK4 (INK4) family of proteins) can function to prevent complex assembly and to inhibit assembled complexes b | The net activation state of CDK4 and CDK6 initiates the phosphorylation of the tumour suppressor retinoblastoma protein (RB) that contributes to activation and release of the E2F family of transcription factors. E2F proteins control the expression of a multitude of positive-acting factors that are critical for progression through the S phase and the G2–M transition. Multiple mechanisms lead to RB inactivation in cancer, such as mutations, aberrant phosphorylation or protein sequestration. c | E2Fs and additional signals drive the expression and activation of CDK2–CycE and CDK2–CycA complexes, which contribute to DNA replication and further phosphorylation of RB. Deregulation of this activity is found in cancer through amplification of E-type cyclins or loss of CDK inhibitors. CCN, cyclin; CDC, cell division cycle; CDT1, chromatin licensing and DNA replication factor 1; CIP, CDK-interacting protein; KIP, kinase inhibitory protein; MAD2L1, MAD2 mitotic arrest deficient-like 1; MCM, minichromosome maintenance complex component; PLK1, polo-like kinase 1.

Figure 3. Summary of the biological functions of CDK complexes

A summary of the different classes of cyclin-dependent kinase (CDK)–cyclin (Cyc) complexes involved in the cell cycle or in diverse biological processes is shown. CDK–Cyc complexes shown in green promote cell cycle progression, whereas those depicted in blue are generally involved in transcriptional processes. The CDK5 complexes shown in red are involved both in the control of neuronal viability and in the promotion of the cell cycle. FOXM1, forkhead box protein M1; RB, retinoblastoma protein.

Figure 4. Deregulation of CDK regulatory genes in cancer

The frequencies of genetic amplification of cyclin-dependent kinase 4 (CDK4) and CDK6; cyclin D1 (CCND1); retinoblastoma 1 (RB1); cyclin-dependent kinase inhibitor 2A (CDKN2A); and cyclin E1 (CCNE1) and CCNE2 are summarized across multiple disease sites. For each of the indicated cancer types, the frequencies of mutation (green), amplification (red) and homozygous deletion (dark blue) were determined using genetic data from >2,000 cancer cases obtained through cBioPortal for Cancer Genomics. As shown, different types of cancer exhibit distinct predominant mechanisms of genetic alterations in cell cycle control. In many cases, the same cancer type has been evaluated in multiple independent studies. Detailed information about each case and disease is accessible through the cBioPortal for Cancer Genomics. ACC, adrenocortical carcinoma; ACYC, adenoid cystic carcinoma; AML, acute myeloid leukaemia; CCLE, Cancer Cell Line Encyclopedia; CS, carcinosarcoma; MBL, medulloblastoma; MM, multiple myeloma; NCI60, US National Cancer Institute (NCI) 60 human tumour cell line anticancer drug screen; SC, serous cystadenocarcinoma.

Figure 5. Selected CDK inhibitors

The chemical structures of several pan-cyclin-dependent kinase (CDK) and CDK4- and CDK6-selective inhibitors are shown. The published half-maximal inhibitory concentration (IC₅₀) values against selected CDK complexes are shown. ND, not determined.