Generating a Genetically Engineered Mouse Model to study Pediatric Glioblastoma

Sima Khazaei1, Caterina Russo1, Adam M. Fontebasso2, Nada Jabado1,2,3

1. Department of Human Genetics, McGill University, Montreal, Quebec, Canada; 2. Department of Experimental Medicine, McGill University, Montreal, Quebec, Canada; 3. Department of Pediatrics, McGill University and the McGill University Health Center Research Institute, Montreal, Quebec, Canada

Pediatric glioblastoma (GBM) is a lethal brain tumor which has a significant degree of clinical and biological heterogeneity. Two somatic point mutations have been identified in the histone 3 variant 3 (H3F3A) in a subset of high-grade brain tumors from childhood and young patients. These two point mutations occur at residues encoding amino acid substitutions of lysine K27M and glycine G34R/V. Both are involved in histone post-translational modifications. Approximately one-third of pediatric high grade astrocytomas and over 80% of diffuse intrinsic pontine glioma (DIPG) tumors have been shown to have one of these two H3.3 mutations. The clinical characteristics of patients with each of these two mutations are different, but both reduce survival and increase resistance to treatment. Furthermore, tumors containing these mutations show age and neuroanatomical specificity, potentially because of unique developmental windows in their tumorigenesis and/or unique cellular origins. H3.3 variant K27M occurs in patients within the childhood years while the G34R/V mutation occurs later in adolescent and young adult patients. We expect that these mutations play a central role in the development of GBMs and therefore, modeling these point mutations is critical to truly identifying their mechanisms of action, their cellular specificity and their developmental timeline in inducing tumorigenesis.

To further study the role of the H3F3A mutations on the genesis of GBMs in mice, we generated three genetically engineered mouse models (GEMM) by using zinc-finger nuclease and Cre/loxP recombination systems. According to our expectation one endogenous H3F3A allele was replaced by a polycistronic cassette. This cassette contains cDNA of H3F3A (either wild-type, or harboring K27M or G34R mutations), and an IRES-luciferase gene sequence. In addition, we crossed our GEMM with Cre mouse lines that expressed Cre recombinase under the control of cell specific promoters, starting with GFAP-Cre and Nestin-CreErt2 mice that allow us to express the Cre in astrocytes and neuronal stem cells, respectively. This will enable us to monitor the expression of H3.3 mutations in specific organs and different times in brain development in a spatiotemporal manner.

The constructed mice will fill a critical gap that is absolutely required for pre-clinical validation of therapeutic targeting of these mutations in this group of deadly pediatric cancers. The mouse models are also important in future studies as potential screening tools for designing subgroup- specific therapeutic clinical trials.

Key words: Pediatric glioblastoma, High grade astrocytomas, Diffuse intrinsic pontineglioma H3.3 mutations