Pathogenesis of SHH medulloblastoma in mice

Medulloblastoma (MB) is the most common malignant brain tumor of childhood that comprises at least four molecularly distinct subgroups. One subgroup is characterized by aberrant Sonic hedgehog (SHH) signalling. Despite the ubiquitous activation of the SHH pathway within this subgroup, there is clear evidence that tumors with a SHH profile may vary in certain molecular and clinical aspects. Targeted therapies as a novel treatment modality for MB patients are especially intriguing for primary and relapsed SHH-MB. As more drugs targeting the Hh pathway become available and enter clinical trials, it is important to know how to stratify the patients for different drugs in order to maximize response rates and to prevent unnecessary treatment failures. In a first project, we aimed to characterize tumor localization in murine Shh-associated models. Using well established mouse models we show here that oncogenic transformation of cerebellar granule cell precursors at early developmental time points may result in the formation of midline and hemispheric medulloblastoma. On the other side, oncogenic transformation at later developmental time points exclusively results in the formation of hemispheric medulloblastoma. These data, which perfectly match to the recently published observations in human patients, indicate that granule neuron precursors are biologically distinct in different cerebellar compartments and that localization of SHH medulloblastoma is dependent on the time rather than the kind of genetic alteration. In a second project, we aimed to investigate the effect of the Wnt pathway on Shh-associated medulloblastoma. Previous studies were able to show that Wnt/ß-Catenin activation might be able to inhibit Shh-associated medulloblastoma growth through downregulation of the Shh pathway. Considering a possible therapeutic approach, in vitro treatments with lithium chloride, a well-known Wnt/ß-Catenin agonist, were carried out. Lithium chloride treatment in vitro resulted in a decrease of granule neuron precursor and tumor cell viability. In the third project, we aimed to better understand the mechanisms in MB of primary and secondary resistance to drugs targeting the Hh pathway. To this end, we generated new mouse models, characterized and used these and already existing genetically engineered mouse models (GEMMs). We aimed to further expand our pre-existing repertoire of murine SHH-activated MB models in order to have a spectrum of tumors that are driven by mutations at different levels of the Hh pathway, such as PTCH1, SMO, or MYCN, and with different combinations of other mutations in additional pathways that may co-operate with HH signalling (e.g. TP53 or PIK3CA). The mouse strains Math1-creERT2::lsl-SmoM2Fl/+, Math1-creERT2::lsl-SmoM2Fl/+ lsl-Pik3caFl/+, Math1-creERT2::Ptch1Fl/Fl, Math1-creERT2::Ptch1Fl/Fl lsl-Pik3caFl/+, and Math1-creERT2::Ptch1Fl/Fl lsl-Pik3caFl/Fl developed Shh-associated medulloblastoma. Our mouse model with a MYCN mutation did not develop any tumor, neither as such nor with an additional TP53 mutation. Math1-creERT2::lsl-SmoM2Fl/+, Math1-creERT2::lsl-SmoM2Fl/+ lsl-Pik3caFl/+, and Math1-creERT2::Ptch1Fl/Fl mice were then treated with the SMO inhibitor LDE225, that is already used in clinical trials. Math1-creERT2::lsl-SmoM2Fl/+ and Math1-creERT2::Ptch1Fl/Fl mice first benefit from the treatment, but then also die due to symptoms of the tumor, whereas Math1-creERT2::lsl-SmoM2Fl/+ lsl-Pik3caFl/+ mice did not show a better prognosis for survival at all. None the less, proliferation was reduced in all tumors treated with LDE225.