No abstract

Phosphorylation of human CDC25B phosphatase by CDK1/cyclin A triggers its proteasome-dependent degradation.

CDC25B is one of the three human phosphatases that activate the CDK-cyclin complexes, thereby triggering cell-cycle progression and division. Commitment to early mitotic events depends on the activation of a centrosomal pool of CDK1–cyclin-B1, and CDC25B is thought to be involved in initiating this centrosomal CDK1–cyclin-B1 activity. Centrosome-associated checkpoint kinase 1 (CHK1) has been proposed to contribute to the proper timing of a normal cell division cycle by inhibiting the activation of the centrosomal pool of CDK1. Here, we show that CDC25B is phosphorylated by CHK1 in vitro on multiple residues, including S230 and S563. We demonstrate these phosphorylations occur in vivo and that they are dependent on CHK1 activity. S230 CHK1-mediated phosphorylation is detected in cell extracts during S phase and G2 phase in the absence of DNA damage. We show that the S230-phosphorylated form of CDC25B is located at the centrosome from early S phase until mitosis. Furthermore, mutation of S230 to alanine increases the mitotic-inducing activity of CDC25B. Our results support a model in which, under normal cell cycle conditions and in the absence of DNA damage, CHK1 constitutively phosphorylates CDC25B during interphase and thus prevents the premature initiation of mitosis by negatively regulating the activity of CDC25B at the centrosome.

CDC25B is one of the three human dual-specificity phosphatases involved in the activation of cyclin-dependent kinases at key stages of the cell division cycle. CDC25B that is responsible for the activation of CDK1-cyclin B1 is regulated by phosphorylation. The STK15/Aurora-A kinase locally phosphorylates CDC25B on serine 353 at the centrosome during the G2/M transition. Here we have investigated this phosphorylation event during the cell cycle, and in response to activation of the G2 DNA damage checkpoint. We show that accumulation of the S353-phosphorylated form of CDC25B at the centrosome correlates with the relocalization of cyclin B1 to the nucleus and the activation of CDK1 at entry into mitosis. Upon activation of the G2/M checkpoint by DNA damage, we demonstrate that Aurora-A is not activated and consequently CDC25B is not phosphorylated. We show that ectopic expression of Aurora-A results in a bypass of the checkpoint that was partially overcome by a S353A mutant of CDC25B. Finally, we show that bypass of the G2/M checkpoint by the CHK1 kinase inhibitor UCN-01 results in the activation of Aurora-A and phosphorylation of CDC25B on S353. These results strongly suggest that Aurora-A-mediated phosphorylation of CDC25B at the centrosome is an important step contributing to the earliest events inducing mitosis, upstream of CDK1-cyclin B1 activation.

Using nuclear run-on assays, we showed that the tissue-specific expression of quail Pax-6 (Pax-QNR) P0-initiated mRNAs is due in part to regulation of the gene at the transcriptional level. Regulatory sequences governing neuroretina-specific expression of the P0-initiated mRNAs were investigated. By using reporter-based expression assays, we characterized a region within the Pax-QNR gene, located 7.5 kbp downstream from the P0 promoter, that functions as an enhancer in neuroretina cells but not in nonexpressing P0-initiated mRNA cells (quail embryo cells and quail retinal pigment epithelial cells). This enhancer element functioned in a position- and orientation-independent manner both on the Pax-QNR P0 promoter and the heterologous thymidine kinase promoter. Moreover, this enhancer element exhibited a developmental stage-specific activity during embryonic neuroretina development: in contrast to activity at day E7, the enhancer activity was very weak at day E5. This paralleled the level of expression of P0-initiated mRNAs observed at the same stages. Using footprinting, gel retardation, and Southwestern (DNA-protein) analysis, we demonstrated the existence of four neuroretina-specific nuclear protein-binding sites, involving multiple unknown factors. In addition we showed that the quail enhancer element is structurally and functionally conserved in mice. All of these results strongly suggest that this enhancer element may contribute to the neuroretina-specific transcriptional regulation of the Pax-6 gene in vivo.

During investigations on the regulation of the Pax-6 gene, we characterized a cDNA from quail neuroretina showing a 5' untranslated region distinct from that previously described and initiated from an internal promoter. Using RNase protection and primer extension mapping, we localized this second quail Pax-6 promoter, termed P1. As reported for the already described P0 promoter, P1 was also transactivated in vitro by the p46Pax-QNR protein. RNase protection assays performed with quail neuroretina RNA showed that P1-initiated mRNAs were detected before the P0-initiated mRNAs, remained constant up to embryonic day 8, and decreased slowly thereafter whereas, P0-initiated mRNAs accumulated up to embryonic day 8. In contrast, quail retinal pigmented epithelium expressed only the P1-initiated mRNAs. Transformation of these cells by the v-myc oncogene induced neuronal traits in the culture, which thereafter, in addition to the P1-initiated mRNAs, expressed Pax-QNR from the P0 promoter. These results suggest that expression of the quail Pax-6 gene is under the control of different regulators through alternate promoters, P0 being activated at the onset of neuronal differentiation.

Chk2 is a key player of the DNA damage signalling pathway. To identify new regulators of this kinase, we performed a yeast two-hybrid screen and found that Chk2 associated with the B′ regulatory subunit of protein phosphatase PP2A. In vitro GST-Chk2 pulldowns demonstrated that B′c isoforms bound to Chk2 with the strongest apparent affinity. This was confirmed in cellulo by co-immunoprecipitation after overexpression of the respective partners in HEK293 cells. The A and C subunits of PP2A were present in the complexes, suggesting that Chk2 was associated with a functionnal PP2A. In vitro kinase assays showed that B′c3 was a potent Chk2 substrate. This phosphorylation increased the catalytic phosphatase activity of PP2A measured on MAP kinase-phosphorylated myelin basic protein as well as on autophosphorylated Chk2. Finally, we demonstrated that overexpressing B′c3 in HEK293 suppressed the phosphorylation of Chk2 induced by a genotoxic treatment, suggesting that PP2A may counteract the action of the checkpoint kinase in living cells.

No abstract