Establishment and characterization of preclinical mouse models for evaluation of oncogenic and tumor-suppressive properties of p53 family members

Cancer development is a multistep process which leads to tumors composed of diverse cell populations originating from one cell which underwent differential genetic changes over time. Within one tumor, such heterogeneity provides distinct subpopulations with selection advantages promoting metastasis as well as therapy resistance. In the course of targeted cancer therapies tumor composition is meanwhile monitored via circulating tumor DNA from blood samples during treatment and accordingly allows adjustment of therapy. In experimental models of solid tumor xenografts in mice, however, this method is not applicable as required blood volumes exceed blood volume of animals, hence limiting investigation of tumorrelevant genes in preclinical mouse models. Here, a method was developed which facilitates simultaneous monitoring of growth dynamics of two distinct tumor cell populations within one tumor xenograft. Therefore, cells were labelled by stable expression of either Gaussia luciferase (GLuc) or Cypridina luciferase (CLuc) prior to injection. Both luciferases are secreted into the blood stream of transplanted mice. This allows asessment of tumor composition by enzyme activity of both luciferases requiring only very low blood volumes. Moreover, to facilitate investigation of the impact of targeted genetic manipulations luciferases were linked with (non)-targeting shRNAs. To establish this method, shRNAs were used targeting the p53 familiy members p53 and p73. Whereas p53 is acknowledged as the most important tumor suppressor, p73 can occur in two N-terminally different isoforms with opposing attributes: the tumorsuppressive full length isoform TAp73 and the N-terminally truncated tumorpromoting isoform Np73. The dominant negative function of Np73 includes its ability to form hetero-oligomers with its family members p53 and TAp73, thereby interfering with their transcriptional activity. Linking the luciferases to (non)-targeting shRNAs, the differential growth properties of transplanted cells in presence and absence of p53 (or p73) can be monitored simultaneously. The here established method was successfully validated in a model of experimental metastasis as well as under therapeutic conditions. Moreover, it could be demonstrated that the growth behaviour of p73-high-expressing cells Hs 766T is highly dependent on the relative abundance of both N-terminal isoforms. The shRNA-mediated reduction of both isoforms strongly reduces tumorigenicity of these cells. In accordance with previous publications, the reintroduction of Np73 rescued this growth defect, whereas ectopic expression of TAp73 further attenuates proliferation. In order to further investigate Np73´s role in tumor development, an inducible Np73 transgenic mouse model was characterized. Whereas mere overexpression of Np73 exhibited no tumorrelevant properties, the combination with heterozygous knockout of p53 entailed earlier and accelerated tumor development particularly of lung tumors and lymphoma. The loss of the second p53 allele in lung tumors suggests that the dominant negative effect of Np73 rather impacts TAp73 than p53. Accordingly, the observed fertility and embryonic developmental defects in this transgenic model rather pointed towards a TAp73-dependent effect of Np73 as, in contrast to p53-deficient mice, severe and partially comparable defects have been described in TAp73- and complete p73-deficient mice. Finally, transcriptomewide analysis of Np73-overexpressing murine embryonic fibroblasts revealed positive regulation of metastasis-promoting factors like ITGB4, JAG1 and 2. This tumorpromoting property of Np73 goes in line with accelerated dissemination of lymphoma into lungs of Np73;p53+/- mice. Taken together, these results clearly demonstrate growth- as well as metastasis-promoting traits of Np73. However, the specific virtue of Np73 are largely cell context-dependent