The pocket protein (PP) family consists of the three members RB1, p107 and p130 all possessing tumor suppressive properties. Indeed, the PPs jointly control the G1/S transition mainly by inhibiting E2F transcription factors. Notably, several viral oncoproteins are capable of binding and inhibiting PPs. Merkel cell polyomavirus (MCPyV) is considered as etiological factor for Merkel cell carcinoma (MCC) with expression of the viral Large T antigen (LT) harboring an intact PP binding domain being required for proliferation of most MCC cells. Therefore, we analyzed the interaction of MCPyV-LT with the PPs. Co-IP experiments indicate that MCPyV-LT binds potently only to RB1. Moreover, MCPyV-LT knockdown-induced growth arrest in MCC cells can be rescued by knockdown of RB1, but not by p107 or p130 knockdown. Accordingly, cell cycle arrest and E2F target gene repression mediated by the single PPs can only in the case of RB1 be significantly reverted by MCPyV-LT expression. Moreover, data from an MCC patient indicate that loss of RB1 rendered the MCPyV-positive MCC cells LT independent. Thus, our results suggest that RB1 is the dominant tumor suppressor PP in MCC, and that inactivation of RB1 by MCPyV-LT is largely sufficient for its growth supporting function in established MCPyV-positive MCC cells.

We recently introduced CDK5 as target in HCC, regulating DNA damage response. Based on this and on our previous knowledge about vascular effects of CDK5, we investigated the role of CDK5 in angiogenesis in HCC, one of the most vascularized tumors. We put a special focus on the transcription factor HIF-1α, a master regulator of tumor angiogenesis.The interaction of CDK5 with HIF-1α was tested by Western blot, PCR, reporter gene assay, immunohistochemistry, kinase assay, co-immunoprecipitation, mass spectrometry, and mutation studies. In vivo, different murine HCC models, were either induced by diethylnitrosamine or subcutaneous injection of HUH7 or HepG2 cells. The correlation of vascular density and CDK5 was assessed by immunostaining of a microarray of liver tissues from HCC patients.Inhibition of CDK5 in endothelial or HCC cells reduced HIF-1α levels in vitro and in vivo, and transcription of HIF-1α target genes (VEGFA, VEGFR1, EphrinA1). Mass spectrometry and site directed mutagenesis revealed a stabilizing phosphorylation of HIF-1α at Ser687 by CDK5. Vascular density was decreased in murine HCC models by CDK5 inhibition.In conclusion, inhibiting CDK5 is a multi-modal systemic approach to treat HCC, hitting angiogenesis, as well as the tumor cells themselves.

Severe side effects often restrict clinical application of the widely used chemotherapeutic drug doxorubicin. In order to decrease required substance concentrations, new concepts for successful combination therapy are needed. Since doxorubicin causes DNA damage, combination with compounds that modulate DNA repair could be a promising strategy. Very recently, a role of nuclear actin for DNA damage repair has been proposed, making actin a potential target for cancer therapy in combination with DNA-damaging therapeutics. This is of special interest, since actin-binding compounds have not yet found their way into clinics. We find that low-dose combination treatment of doxorubicin with the actin polymerizer chondramide B (ChB) synergistically inhibits tumor growth in vivo. On the cellular level we demonstrate that actin binders inhibit distinctive double strand break (DSB) repair pathways. Actin manipulation impairs the recruitment of replication factor A (RPA) to the site of damage, a process crucial for homologous recombination. In addition, actin binders reduce autophosphorylation of DNA-dependent protein kinase (DNA-PK) during nonhomologous end joining. Our findings substantiate a direct involvement of actin in nuclear DSB repair pathways, and propose actin as a therapeutic target for combination therapy with DNA-damaging agents such as doxorubicin.