Epigenetic dependence of diffuse midline glioma with H3K27M mutation

Diffuse intrinsic pontine glioma (DIPG) is a deadly disease with the median survival of DIPG patients less than one year after diagnosis. DIPG tumors initiate from the pons and midline of brain, and spread to other brain regions where they mingle with normal cells such as neurons. Recently, it has been shown that neurons promote proliferation and invasion of DIPG cells. However, it is largely unknown how the intrinsic gene expression program of DIPG cells is regulated for their interactions with neurons. Furthermore, currently there are no effective treatments for DIPG patients. Therefore, it is imperative to understand the molecular basis of pathogenesis of DIPG and to identify novel drug targets for this deadly disease. About 80% DIPG tumors contain somatic mutations at genes encoding canonical histone H3 (H3.1) or its variant (H3.3), resulting in replacement of histone H3 lysine 27 with methionine (H3K27M). We found that expression of either H3.1K27M or H3.3K27M proteins leads to a global reduction of di- and tri-methylation (H3K27me2/me3) on wild type histone H3. H3K27me2/me3 marks are catalyzed by the PRC2 complex with Ezh2 as the catalytic subunit, and play important roles in gene silencing. However, H3.1K27M and H3.3K27M DIPG tumors show distinct gene expression signatures and are associated with distinct driver genetic mutations. We hypothesize that epigenome reprograming by H3.1K27M and H3.3K27M creates a dependence of DIPG tumor cells on other chromatin regulators. To test this hypothesis, we performed CRISPR/Cas9 screens and found that Brg1(brahma-related gene 1 or called SMACAR4), the catalytic subunit of mammalian SWI/SNF (mSWI/SNF) chromatin remodeling complexes, and Ezh2, are among top hits. The identification of Ezh2 is expected as we and others have shown that Ezh2 and H3K27me2/me3 are needed to silence tumor suppressor genes in DIPG cells. However, it was not known whether Brg1 has any roles in DIPG. Our results support the hypothesis that Brg1 functions as the catalytic subunit of DIPG-specific mSWI/SNF complexes to control the gene expression and fitness of DIPG cells. Furthermore, in H3.3K27M DIPG cells, transcription factor SOX10 recruits Brg1 to regulate the expression of genes involved in cell growth, extracellular matrix and neural development. Based on these exciting observations, we will 1) identify genes whose expression is regulated by Brg1 directly H3.1K27M DIPG cells; 2) evaluate genetic and epigenetic changes in H3.1K27M DIPG cells that render these cells depends on Brg1; 3) test the hypothesis that Brg1 and its target genes involved in neural development contribute to the neuron-glioma interactions of H3.3K27M DIPG cells; 4) identify other subunits of mSWI/SNF complexes that work with Brg1 to control gene expression and fitness of DIPG cells; and 5) test the hypothesis that inhibition of Brg1 alone and in combination with Ezh2 inhibition impede the growth of DIPG tumors using patient derived xenograft mouse models. Together, these studies will not only provide molecular insight into how epigenomic re-writing by H3K27M mutant proteins promotes tumorigenesis, but also identify drug targets for this deadly disease.