Epigenetic regulator ARID1A and stem cell transcription factor SOX9 in the maintenance of pancreatic ductal cell differentiation state and development of intraductal papillary mucinous neoplasia (IPMN) and pancreatic ductal adenocarcinoma (PDAC)

Pancreatic ductal adenocarcinoma (PDAC) is one of the highest mortality malignancies, with average 5-year survival less than 3%. The etiology of pancreatic cancer has been investigated, and the identified risk factors including such as non-O ABO blood group (1), uncommon hereditary factors of germline mutations in p16, BRCA1 and BRCA2 (2), obesity, chronic pancreatitis, long-term diabetes mellitus (3,4), Helicobacter pylori colonization (5) and tobacco smoking (6). Somatic mutations of KRAS is a driver mutation with the frequency of over 90% in invasive PDAC (7,8), and with the frequency increasing from 36% to 87% according to disease progression from PanIN-1a to 2–3 preneoplastic lesions (9,10). Besides the KRAS mutations, there exist many other genes with somatic mutations in pancreatic cancer based on The Cancer Genome Atlas (TCGA) pancreatic cancer dataset (Table 1). However, our understanding on the molecular mechanisms of these mutated genes in the development of intraductal papillary mucinous neoplasia (IPMN) and consequently invasive PDAC is still limited (11,12), and the effective therapies against PDAC are lacking (13). In a recent study of Gastroenterology, Kimura and colleagues used available mouse models and clinical samples to investigate the functions of ARID1A (AT rich interactive domain 1A) in the development of IPMN and PDAC (14). Their investigation revealed that conditional Arid1a knockout (KO) mice had dilated pancreatic duct and acceleration of PDAC from IPMN when mutant Kras oncogene was also expressed. Eventually, multilocular cystic neoplasm developed. Interestingly, cystic neoplasm developed in these transgenic mice did not show aggressive cell proliferation, but manifested excessive intraductal mucin. Moreover, cystic neoplasms developed in these mice had the characteristics of IPMN in human, but not that of mucinous cystic neoplasms (MCN). About 20% (3 out of 15) of these IPMN lesions in the transgenic mice gradually developed into PDAC at 48 weeks of age. They observed decreased expression of Sox9 in Arid1a conditional KO mouse pancreas and marked decreased expression of p21, p53 and p16INK4a (Cdkn2a) in mouse IPMN-like cystic neoplasms. They concluded that ARID1A deficiency results in ductal cell dedifferentiation and ductal dilation due to reduced SOX9 expression. This phenotype was rescued by overexpression of SOX9 both in cell culture and in vivo. Their research illuminated an important function of ARID1A which is the maintenance of differentiation of pancreatic ductal cells and the suppression of PDAC development,