Paradoxical Role of AT-rich Interactive Domain 1A in Restraining Pancreatic Carcinogenesis

SIMPLE SUMMARY: Genes with deleterious mutations in tumors are widely accepted as tumor suppressors, since, loss of their normal expression often promotes tumor development. However, most tumors develop over a long period, with gradual accumulation of tumor-promoting events. Thus, to comprehend the role of individual genes in this evolving process of tumorigenesis, it is critical to investigate their role in both early precursors and established tumor cells. Despite recurrent mutations in ARID1A in genomes of human cancer, including pancreatic cancer, its role in tumorigenesis is not clear. We aim to understand the role of Arid1a in pancreatic cancer development and maintenance by investigating its role in both early pancreatic precursor cells and established pancreatic cancer cells. Besides novel understanding of context-dependent role of Arid1a in pancreatic cancer, this study will also enable development of therapeutic strategies for pancreatic cancers patients with ARID1A mutations, which is currently a critical unmet need in clinic. ABSTRACT: Background & Aims: ARID1A is postulated to be a tumor suppressor gene owing to loss-of-function mutations in human pancreatic ductal adenocarcinomas (PDAC). However, its role in pancreatic pathogenesis is not clear despite recent studies using genetically engineered mouse (GEM) models. We aimed at further understanding of its direct functional role in PDAC, using a combination of GEM model and PDAC cell lines. Methods: Pancreas-specific mutant Arid1a-driven GEM model (Ptf1a-Cre; Kras(G12D); Arid1a(f/f) or “KAC”) was generated by crossing Ptf1a-Cre; Kras(G12D) (“KC”) mice with Arid1a(f/f) mice and characterized histologically with timed necropsies. Arid1a was also deleted using CRISPR-Cas9 system in established human and murine PDAC cell lines to study the immediate effects of Arid1a loss in isogenic models. Cell lines with or without Arid1a expression were developed from respective autochthonous PDAC GEM models, compared functionally using various culture assays, and subjected to RNA-sequencing for comparative gene expression analysis. DNA damage repair was analyzed in cultured cells using immunofluorescence and COMET assay. Results: Retention of Arid1a is critical for early progression of mutant Kras-driven pre-malignant lesions into PDAC, as evident by lower Ki-67 and higher apoptosis staining in “KAC” as compared to “KC” mice. Enforced deletion of Arid1a in established PDAC cell lines caused suppression of cellular growth and migration, accompanied by compromised DNA damage repair. Despite early development of relatively indolent cystic precursor lesions called intraductal papillary mucinous neoplasms (IPMNs), a subset of “KAC” mice developed aggressive PDAC in later ages. PDAC cells obtained from older autochthonous “KAC” mice revealed various compensatory (“escaper”) mechanisms to overcome the growth suppressive effects of Arid1a loss. Conclusions: Arid1a is an essential survival gene whose loss impairs cellular growth, and thus, its expression is critical during early stages of pancreatic tumorigenesis in mouse models. In tumors that arise in the setting of ARID1A loss, a multitude of “escaper” mechanisms drive progression.