Differences in regulation and function of E-cyclins in human cancer cells

Abstract

Mammalian cyclin E was cloned by groups of James Roberts and Steven Reed in a screen for human cDNAs which can complement mutant G1 cyclin genes CLN1, CLN2 and CLN3 in yeast Sacaromyces cerevisiae.1 The flurry of studies which followed this discovery demonstrated that in mammalian cells cyclin E is induced in late G1 phase when it activates cyclin-dependent kinase CDK2, and also CDK1 and CDK3. During G1 phase progression cyclin E-CDK2 kinase phosphorylates and inactivates the retinoblastoma protein, pRB, leading to activation of E2F transcription factors. Since the cyclin E gene represents one of E2F transcriptional targets, this mechanism creates a positive feedback loop which leads to full activation of cyclin E-CDK2 kinase. Once induced, cyclin E-CDK2 phosphorylates proteins governing cell cycle progression (pRB, p27Kip1, E2F5), centrosome duplication (NPM, CP110), histone gene transcription (NPAT) and others. Cyclin E and cyclin E-CDK2 kinase activity is essential for assembly of DNA pre-replication complexes and for firing of DNA replication origins. As the S phase progresses, cyclin E becomes phosphorylated by cyclin E-CDK2 and by GSK3, and is then targeted for proteosomal degradation by the SCFFbw7 ubiquitin ligase.1 Subsequently, groups of Bruno Amati, Yue Xiong and Steve Coats isolated the second mammalian E-type cyclin, which was termed cyclin E2, while the protein known as “cyclin E” was renamed as cyclin E1. The two E-cyclins show substantial aminoacid similarity, associate with the same CDK partners, and appear to perform similar biological functions.1 Their regulation seems to be similar, including transcriptional activation by E2F and protein degradation through SCFFbw7 ubiquitin ligase. Also in vivo, the two E-type cyclins seem to perform highly overlapping set of functions. Thus, genetic ablation of cyclins E1 or E2 resulted in no major phenotypes, whereas combined loss of both E-cyclins led to an early embryonic lethality due to placental abnormalities.2 In adult mice, combined ablation of cyclins E1 and E2 impairs neuronal synaptic function and leads to memory deficits, due to a function of cyclin E in regulating synaptic plasticity.3 Collectively, all these observations suggested that cyclins E1 and E2 are functionally equivalent. A recent study from Elizabeth Musgrove group4 indicates that this prevailing view may need revisions. The authors focused on the function of overexpressed cyclin E in breast cancer cells. Cyclins E1 and E2 are overexpressed in a substantial number of human cancers, where they contribute to tumorigenesis likely by driving uncontrolled cell cycle progression.1 Moreover, overexpression of cyclin E1 was shown to result in chromosome instability in in vitro cultured cells, and in vivo, in mouse tumors.5,6 While the exact molecular mechanism remains to be elucidated, this role of cyclin E1 is mediated, at least in part, by binding and phosphorylating the anaphase-promoting complex (APC) regulatory subunit, Cdh1.7 This, in turn inhibits APC activity, and results in impaired mitotic progression of cyclin E1-overexpressing cells.7 Unexpectedly, Caldon et al.4 now demonstrate that cyclin E2, when overexpressed, does not interact with Cdh1, does not inhibit APC and does not impair mitotic progression. Yet, cyclin E2 overexpression still triggers genomic instability, as evidenced by increased fraction of abnormal mitoses, as well as the presence of chromosomal aberrations such as chromosome breaks and end-to-end fusions in cyclin E2-overexpressing cells. While the mechanism through which cyclin E2 causes these abnormalities remains unclear, Caldon et al.4 propose that this effect is mediated through inactivation of pRB and pRB-like p107 and p130 proteins by hyperactive cyclin E2-CDK2. Intriguingly, the same group demonstrated that the levels of cyclin E2 in cancer cells are controlled via a distinct mechanism from that operating in normal cells.8 Specifically, while in non-transformed cells the levels of cyclins E1 and E2 are regulated by SCFFbw7, in a breast cancer cells depletion of Fbw7 affects the levels of cyclin E1, but not E2.8 These finding lead to several questions. Are results of Caldon et al.4,8 generalizable across different types of human cancers? How is the stability of cyclin E2 controlled in cancer cells, and how mechanistically cyclin E2 expression shifts from Fbw7-dependent to -independent mode? How does cyclin E2 trigger chromosomal instability? Analyses of the endogenous protein complexes associated with cyclins E1 and E2 in cancer cells may help to unravel molecular differences between these two related, but apparently distinct proteins.