Deciphering Drivers of Functional and Phenotypic Changes through Genome-Wide Screen (DRIVE-Genome) to Identify Genes Inducing Genomic Instability in Myeloma

Introduction: Since DNA strand breaks are required for genomic rearrangements, genes involved in DNA breaks can potentially drive genomic instability and clonal evolution. Therefore, we developed a novel genome-wide functional phenotyping screen to identify genes whose modulation impacts DNA breaks in multiple myeloma (MM).

Methods: Myeloma cell lines (JIM3 and U266) were transduced (in triplicate) with a whole genome lentiviral expression library comprised of 17,255 open reading frames (ORFs) representing 12,728 genes. Cells selected in puromycin were cultured and stained using a fluorescently labeled antibody against 𝛄-H2AX, a marker of DNA breaks. Fluorescence-activated cell sorting was used to isolate cells with increased 𝛄-H2AX expression representing presence of damaged DNA; and those cells were sequenced to identify the ORFs associated with increased DNA breaks. Selected genes were further investigated in gain as well as loss of function studies.

Results: Genome wide ORF expression screen identified, in JIM3 and U266 MM cell lines, 789 and 302 genes associated with increased DNA breaks, respectively. We further analyzed 226 genes associated with increased DNA breaks that are common in both MM cell lines by functional enrichment analysis and report 6 distinct groups. These included 1) cell cycle cyclin-dependent kinases and phosphatases; 2) serine/threonine-protein kinases including calcium/calmodulin-dependent protein kinases; 3) DNA repair and recombination proteins; 4) cytoskeleton, chromosome maintenance and mitosis/cytokinesis related proteins; 5) monooxygenases involved in metabolic detoxification of xenobiotics; and 6) RNA helicases, and proteins involved in RNA processing, splicing and degradation. We next utilized 2 large datasets to evaluate clinical impact of high expression of these genes on patient outcome in MM. Elevated expression of 20 and 23 of these genes in MM patients conferred poor overall (OS) and event free (EFS) survival, respectively (GSE24080; n=559). In second MM dataset elevated expression of 14 and 8 of these genes correlated with poor OS and EFS, respectively (IFM70; n=170). Overall, the expression of three genes correlated with both the PFS and OS in both datasets. These included; 1) DTYMK; deoxythymidylate kinase; a kinase, regulating nucleotide biosynthesis and contributes to DNA replication and repair; 2) CDC25A; cell division cycle 25A, a phosphatase involved in G1/S progression and whose degradation in response to DNA damage is essential to prevent cells with chromosomal abnormalities from progressing through cell division; and 3) EXO1; Exonuclease 1, a nuclease with roles in DNA repair and replication.

We further functionally validated the role of these select genes for their impact on genomic instability. Overexpression of all three genes (DTYMK, CDC25A and EXO1) in non-cancerous bone marrow stromal (HS5) and/or cancer cells led to increase in genomic instability, as assessed by micronucleus assay, by ∼7-fold, 3.6-fold and > 2-fold respectively. For EXO1 and two other genes also identified in the screen (HDAC8 and TPX2) we have confirmed their role in causing spontaneous and chemotherapy-induced DNA breaks, acquisition of newer genomic changes overtime as well as proliferation of MM and/or other cancer cells. Validation of all identified hits in DNA damage and assays to measure genomic instability is currently in progress.

Conclusion: We report a novel genome wide high throughput screen that can uniquely utilize protein staining properties and FACS to assess impact of individual gene manipulation at functional and phenotypic level. Using this screen, we have identified genes and pathways contributing to increased DNA breaks in MM and their function is further validated in gain and loss-of function studies in spontaneous and chemotherapy-induced DNA breaks, and genomic instability in myeloma cells. These genes have potential to serve as promising new targets to make MM cells genomically static and the functional screen provides a tool to investigate impact of genome wide perturbations on functional and phenotypic changes.