Rsk2 and Lrp5 deficiency limit osteosarcoma growth in cFos-transgenic mice by different mechanisms

Abstract

The c-Fos transgenic mouse model, where the transgene is placed under the control of the H2Kb promoter, has been described to spontaneously develop osteosarcomas due to the transformation of the osteoblastic lineage. Moreover, the ribosomal S6 kinase 2 (Rsk2), which activates c-Fos by phosphorylation on serine 362, was found to be essential for c-Fos-induced osteosarcoma formation in mice. Using cell lines derived from FosTg;Rsk2-/- tumors it was further observed that Rsk2 deficiency impaired the growth advantage of FosTg cells, which was explained by aberrant nuclei number due to impaired cytokinesis suggesting the induction of “mitotic catastrophe”. Thus, Rsk2 was defined as a potential target for the design of new osteosarcoma therapies. In this thesis a pharmacological Rsk inhibitor (BI-D1870) was tested for its ability to inhibit proliferation of osteosarcoma cells. Here, the effects observed by the genetic inactivation of Rsk2 were mimicked. Moreover, since BI-D1870 administration to FosTg cell lines reduced expression of aurora kinase B, the influence of a pharmacological aurora kinase B inhibitor (Hesperadin) was also tested. Similar to BI-D1870, Hesperadin caused impaired cytokinesis, resulting in the accumulation of polynuclear cells, an effect that was not only observed in FosTg cell lines, but also in the human osteosarcoma cell line U2OS. The second part of this thesis focused on targeting a more osteoblast-specific pathway involving the transmembrane protein Lrp5, which acts as a co-receptor of Wnt ligands. Importantly, LRP5 mutations were found to be the cause of two major human bone diseases,associated with either high bone mass (Lrp5 activation) or osteoporosis (Lrp5 inactivation). To analyze the role of Lrp5 in osteosarcoma growth, FosTg mice were crossed with Lrp5-deficient mice or with mice carrying a high bone mass mutation of Lrp5 (Lrp5A213V/+). It was found that Lrp5 deficiency drastically reduced the osteosarcoma volume in FosTg mice, which was explained by a decreased number of active osteoblasts at the tumor surfaces associated with the appearance of large unmineralized areas. Cell lines were established from the three FosTg mouse models with different Lrp5 genotypes, however growth curve and mineralization analyses did not reveal significant differences, thereby suggesting a non-cell autonomous mechanism. Nevertheless, the investigation of potential downstream effectors by microarray analysis and their confirmation by qPCR permitted to identify specific genes with differential expression, which will be further analyzed for their implication in bone remodeling and cancer development. Moreover, an antibody assay performed with serum from the different mice led to the identification of Fgf21 as a potential factor causing lipodystrophy in FosTg mice. In summary, by targeting the Rsk2 or the Lrp5 pathway, this project identified two different mechanisms to impair osteosarcoma growth in c-Fos transgenic mice.