Identifizierung von Kaiso Like 1 (Zbtb4) als negativem Regulator der P21CIP1-Expression, Interaktionspartner von Miz1 und Histondeacetylasen sowie Modulator der zellulären p53-Antwort

The newly identified protein Kaiso Like 1 is a member of zinc finger proteins with a BTB/POZ domain and is therefore also named Zbtb4. Due to its homology to the transcription factor and POZ protein Kaiso it is named Kaiso Like 1 (KL1). Little was known about this protein before this thesis was initiated: A downregulation of KL1 mRNA expression was observed in bad prognostic neuroblastoma tumours versus good prognostic neuroblastoma tumours. Data obtained from the oncomine-database confirmed this pattern of expression for many other solid tumour entities. In addition, data showed the downregulated expression of KL1 mRNA in solid tumours compared to normal tissue. Therefore the characterisation of KL1, by e.g. identifying interaction partners, target genes and its potential biological function was the aim of this thesis. In order to characterize KL1 overexpression and depletion assays were performed in human cell lines. Data showed that strong overexpression of KL1 leads to cellular death, identified as apoptosis by FACS-analysis. In contrast, cells overexpressing low amounts of the KL1 protein as well as cells depleted for KL1 do not behave differently with respect to control cells under standard tissue culture conditions. To test the potential effect of KL1 downregulation observed in bad prognostic tumours on the response of these tumours to chemotherapeutic agents, the neuroblastoma cell line SH-EP was depleted for KL1, then treated with vincristine or other chemotherapeutic agents commonly used in neuroblastoma therapy. Data showed that KL1-depleted SH-EP cells survive the treatment with low doses of vincristine while non-depleted cells die. Vincristine does not damage DNA but, at low doses, inhibits the dynamics of microtubles. This inhibition leads to an activation of p53 which can thereafter transactivate its pro-apoptotic genes as well as P21CIP1, through which p53 mediates a G1 arrest of the cell cycle. This resistance of KL1-depleted cells to vincristineinduced G1 arrest was found to be mediated by KL1. Therefore KL1 was identified as a repressor protein that induces a specific blockage of p53-mediated G1 arrest of the cell cycle mainly by inhibiting the transactivation of its newly identified target gene P21CIP1. When the expression of KL1 is depleted, the repression of P21CIP1 expression can no longer occur. As a consequence the p53 response shifts from apoptosis to a G1 arrest with a following low proliferation rate after treatment with low-dosed vincristine. This might explain why most cells, depleted simultaneously for P21CIP1 and KL1, lose their resistance to low doses of vincristine. To demonstrate that KL1 does not only modify the p53 response due to vincristine treatment but acts in more general manner, cells overexpressing KL1 and control cells were treated with nutlin-3, a stabilisator of p53 activation. While control cells arrest in the G1 phase under the conditions used in the experiment (and do not become apoptotic), cells overexpressing KL1 cycle normally. The KL1-mediated repression of the P21CIP1 expression is independent of p53. Instead, KL1 probably interacts with Miz1 and histone deacetylase repressor complexes at the P21CIP1 promoter: Experiments demonstrated that a histone deacetylase activity is needed for KL1-mediated repression of the P21CIP1 expression. KL1 represses the Miz1-mediated transactivation of the P21CIP1 promoter as shown by reporter assays. By performing co-immunoprecipitation analyses an interaction between the exogenous proteins KL1 and Miz1 and i.e. the histone deacetylase 2 was identified, as well as an interaction between KL1 and endogenous Miz1. Moreover the results showed that KL1 binds in vivo to the P21CIP1 promoter, in vitro to KL1 binding sites and – when Miz1 is co-expressed – to Miz1 bindings sites of the P21CIP1 promoter. Through the inhibition of P21CIP1 KL1 could be identified in neuroblastoma cells as a further factor controlling the activity of p53. Recent data showed that dependent on the cellular context, p21Cip1 acts either as a tumour suppressor or as an oncogene. KL1 as a repressor of P21CIP1 gene expression might inhibit oncogenic function of p21Cip1 in solid tumours. This could therefore be the reason for the observed downregulation of KL1 mRNA expression in bad prognostic solid tumours. KL1 is not suitable as a general therapeutic target since its downregulation leads only to specific resistance against low doses of vincristine, not against other chemotherapeutic agents used in neuroblastoma therapy. This is due to the fact that the chemotherapeutic agents inducing DNA damage promote not only a p53-mediated G1 arrest, efficiently blocked by KL1 in a p21Cip1-dependent manner, but also a G2 arrest that is triggered independently from p21Cip1.