Functional Genomic Landscape of Epigenetic Dependencies in Myeloma

The recent development of comprehensive functional genomic approaches based on the CRISPR/Cas9 gene editing technology and its variants has created new opportunities to assess in a quantitative manner the significance of a given gene for survival and proliferation across different tumor cell lines; or in comparison to those of other genes within the same tumor cell. These comparisons may be particularly informative in the context of genes regulating the epigenome of tumor cells, since selective chemical probes for many epigenetic regulators are not currently available. We examined the results from our analysis of genome-wide CRISPR/Cas9-based gene editing studies in a panel of n=10 MM cell lines (with AVANA pooled library of ~70,000 sgRNAs; 3-4 sgRNAs per gene) and publicly available CRISPR data on cell lines from other hematologic malignancies and solid tumors. We ranked all genes according to the depletion of their sgRNAs; and as in prior studies of our group and others, we designated as essential genes those with sgRNA depletion rank <3000 (based on the MAGECK algorithm). We focused for this study on genes with known or presumed roles in epigenetic regulation and identified several genes (including BRD4, CHD4, PRMT5, PRMT1, SETD1A, JMJD6, HDAC3, et.c.) which appear to be essential for practically all MM cell lines tested, and also broadly essential for >80% of non-MM cell lines. We also identified genes which are essential for most MM cell lines tested and for sizeable subsets of non-MM tumors, including BRD2, SETDB1, EP300, DOT1L CREBBP, and EZH2. It is notable that MM cell lines exhibited higher dependency ranks for EP300, CREBBP and DOT1L compared to a collection of 300+ non-MM cell lines, suggesting that pharmacological inhibitors that can phenocopy the genetic dependency on these genes merit further consideration for therapeutic applications in MM. We also detected genes which are essential for smaller numbers of MM cell lines tested, including ATAD2, and various histone lysine demethylases (KDMs) or methyltransferases (KMTs); while several other epigenetic genes appear to be redundant for in vitro cell survival and proliferation of MM cells. When we examined all genes with median sgRNA depletion ranks <3000 in MM cell lines and selected those sgRNAs showing the most variable effect across the 10 MM cell lines, we identified epigenetic genes such as HDAC3, HDAC8, JMJD6 and KDM2A, implying a differential role of these epigenetic dependencies across different types of MM cell lines. Of interest, the large majority of epigenetic genes identified as dependencies in these CRISPR studies are not recurrently mutated in MM; and, given their copy number status, their essentiality could not be attributed to a confounding effect of excess DNA damage from CRISPR editing of amplified genes. Furthermore, these epigenetic dependencies were not over-represented among the most overexpressed in MM cells vs. normal plasma cells. Overall, these observations indicate that the functional dependency of MM cells on these epigenetic genes cannot be readily detected based on conventional molecular profiling platforms utilized so far to identify promising therapeutic targets. Our results provide novel insights into the epigenetic landscape of MM and establish EP300, CREBBP and DOT1L as epigenetic dependencies with preferential role in MM and thus potential new therapeutic targets in this disease.