(P 69) H3.3G34W ONCOHISTONE IN OSTEOSARCOMA

Objective: Osteosarcoma (OS) is the most common primary bone tumor, with 3.4 cases per million people per year worldwide. These tumors are characterized by an extreme genetic complexity that has been extensively studied, but very little is known about their epigenetics, despite the major role of epigenetics in cancer. Based on the recent discoveries of oncohistones mutations in OS, we want to better understand the role of epigenome dysregulation in OS. The most frequent oncohistone in OS is H3.3FA G34W, with approximately 4% of the cases, but other oncohistones have also been reported, such as G34R and K27M. The H3.3 G34W oncohistone is present in more than 90% of the giant cell tumor of the bone (GCTB). In this context, the methylation mark pattern on the oncohistone tail is redistributed with a reduction of H3.3K36me3 and an increase of H3.3K27me3. In addition, the knockout of the H3.3G34W in GCTB cell lines abolishes tumorigenicity in orthotopic preclinical models. We hypothesize that the epigenetic dysregulations associated with the oncohistone could confer a pathogenic advantage in OS, and could constitute a targetable vulnerability.

Methods: We generated a spontaneous murine model of OS to determine the effect of oncohistone H3.3G34W and associated epigenetic changes in OS. OSX1_Cre recombinase mice were crossed with TRP53loxP and then crossed with H3.3G34Wflex mice. In this mouse model, preosteoblasts and osteoblasts express the oncohistone H3.3G34W in a p53 depletion context (TRP53floxP). These mice are monitored for 18 months for tumor appearances and bone defects (visual and palpation monitoring, PET scans, size measurements). The tumors detected were studied by IHC and used to generate primary cell line models. The epigenetic characterization of these primary cell lines will be assessed by CUT&RUN (H3.3K36me3 and H3.3K27me3), as well as the DNA methylation status by Whole Genome Bisulfite Sequencing. In addition, a CRISPR Cas9 dropout screening will be performed to determine the genes involved in the survival and tumorigenicity of these cells.

Results: We have confirmed the generation of H3.3G34W spontaneous OS tumors by IHC. The H&E staining of these tumors reveals high-grade tumors with areas of early osteoid formation and areas of ossification perfectly compatible with OS diagnosis. In the literature, the OS penetrance, for mice over a year old, is between 5 and 10% for mice OSX1_Cre+;TRP53fl/+, but is 21% (6/29) of tumors in our cohort OSX1_Cre+;G34W;TRP53fl/+, and reaches 34% (10/29) when the spontaneous hindlimb paralysis are added. We observed hindlimbs paralysis in the group OSX1_Cre+;G34W;TRP53fl/+ but not in the G34 counterpart group. In addition, a few animals G34W;TRP53fl/+ have multiple tumor sites (and metastatic events), which is not the case in our G34;TRP53fl/+ cohort, and extremely marginal in literature for the mice with TRP53fl/+ genotype. We can also notice that where the tumors in G34 mice groups are localized principally in the extremities of long bones, there is a good proportion of tumors in the spine and hip of the mice carrying the H3.3G34W oncohistone mutation.

Conclusion: To our knowledge, these are the first murine H3.3G34W osteosarcoma tumors generated. The generation of these spontaneous OS tumors with the presence of H3.3G34W oncohistone offers an unprecedented tool to study how oncohistone and epigenetic dysregulation affects OS pathogenesis and development. This new model will allow us to better understand the role of oncohistone in OS.