Genetic Screening Approaches to Cancer Driver Characterization and Synthetic Lethal Target Discovery
Mengwasser, Kristen Elizabeth
MetadataShow full item record
CitationMengwasser, Kristen Elizabeth. 2018. Genetic Screening Approaches to Cancer Driver Characterization and Synthetic Lethal Target Discovery. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.
AbstractAdvances in genetic screening technology have expanded the toolkit for systematic perturbation of gene function. While the CRISPR-Cas9 system robustly probes genetic loss-of-function in mammalian cells, a barcoded ORFeome library offers the opportunity to systematically study genetic gain-of-function. We employed these two screening tools for three purposes.
First, we performed shRNA and CRISPR-based screens for synthetic lethality with BRCA2 deficiency, in two pairs of BRCA2 isogenic cell lines. BRCA2 mutation commonly drives hereditary breast and ovarian cancer, but also creates vulnerability to PARP inhibitor treatment. Among other genes, we found the AP endonuclease APEX2 and the flap endonuclease FEN1 to be synthetic lethal with BRCA2 deficiency; the base excision repair (BER) pathway was synthetic lethal overall. We demonstrated that FEN1 plays a role in microhomology-mediated end joining (MMEJ) and that a FEN1 inhibitor selectively targets BRCA-deficient cells, offering therapeutic potential in the setting of PARP inhibitor resistance.
Second, we screened a barcoded ORFeome library for genes that either upregulate PD-L1 or interfere with IFNγ signaling. Tumor-expressed PD-L1 engages PD-1 on T cells, dampening T-cell based immune responses to tumors, while defects in IFNγ signaling underlie resistance to immunotherapy. We found 12 GPCRs that significantly increase cell surface PD-L1 display, including three LPARs (LPAR1/2/5), and a family of RNA-binding proteins (CELF3/4/5) that upregulates PD-L1. Conversely, we showed that CLK2 overexpression antagonizes IFNγ signaling; CLK2 is known to phosphorylate PTPN1 and is amplified in tumor types that respond poorly to immunotherapy. Thus, segregating patients based on CLK2 amplification status or augmenting immunotherapy with CLK2 inhibitors may improve clinical responses.
Finally, we utilized CRISPR and ORF-based libraries targeting known and predicted tumor suppressor genes (TSGs) or oncogenes (OGs) to systematically screen cancer drivers for a variety of phenotypes. We found at least 20 genes not commonly annotated as drivers that enhanced proliferation, including PAWR, AMBRA1, and USP28. We screened OGs bearing tumor-associated point mutations and functionally validated the MYCN P44L mutation. We mapped drivers across multiple hallmark cancer phenotypes and found that IRF6 knockout promotes anoikis bypass. Thus, functional studies can augment computational approaches in identifying cancer driver genes.
Citable link to this pagehttp://nrs.harvard.edu/urn-3:HUL.InstRepos:41121232
- FAS Theses and Dissertations