Abstract A053: Defining THOC3 as a key dependency in basal-squamous pancreas cancer subtype specification and growth

Abstract

Pancreas cancer is an extremely deadly malignancy expected to become the second leading cause of cancer-related deaths by 2030. While pancreatic ductal adenocarcinoma (PDAC) has been the main focus of the field, subclones of different subtypes of disease, including basal-squamous, are admixed in tumors. The basal-squamous signature aligns with adenosquamous and is associated with worse prognosis and gemcitabine resistance, demonstrating a pressing need to understand the fate determinants of the basal-squamous subtype so that effective therapies may be developed. The Reya lab recently developed a novel mouse model Msi2-Myc, which induces the expression of stabilized MycT58A in Musashi2 (Msi2)+ stem and progenitor cells. Msi2-Myc mice can form multiple subtypes of pancreas cancer and reliably form basal-squamous tumors. This powerful model enables the ability to track subtype specification from a common precancer pool to determine potential mediators of fate. We identified potential key regulators of basal-squamous pancreas cancer that were upregulated in basal-squamous pancreas cancer compared to precancers and PDAC by RNA-sequencing, subsequently screening them to find candidates that both impaired growth in vitro and downregulated key basal-squamous marker gene expression. The RNA binding protein THO Complex 3 (THOC3) was the only candidate screened to impact growth and identity. We defined THOC3 as a key regulator of basal-squamous fate, as resultant tumors with THOC3 knockdown exhibit a loss of key basal-squamous marker expression by immunofluorescence and altered histology. Transcriptionally, knockdown of THOC3 results in a downregulation of the Moffitt basal gene signature, as well as a significant reduction in the broader transcription signature associated with the basal subtype. Conversely, overexpression of THOC3 results in a significant upregulation of basal-squamous associated genes. We also defined THOC3 as a key regulator of basal-squamous growth. Functional assays demonstrated that THOC3 is required for 3D cell growth of basal-squamous cells across mouse and human models. Additionally, we determined THOC3 is required for tumor growth in vivo in mouse, human, and patient sample models. Combination of THOC3 inhibition with gemcitabine was highly effective against chemoresistant basal-squamous cells. Mechanistically, THOC3 loss is significantly associated with downregulation of AP-1 related genes, particularly in the MAF family, defining a potential role in associated pathways. Collectively, these data identify THOC3 as a key regulator of basal-squamous pancreas cancer identity and growth and begin to define its role in oncogenesis.