Context: Although pancreatic cancer is a common, highly lethal malignancy, the molecular events that enable precursor lesions to become invasive carcinoma remain unclear. We previously reported that the high-mobility group A1 (HMGA1) protein is overexpressed in >90% of primary pancreatic cancers, with absent or low levels in early precursor lesions. Methods: Here, we investigate the role of HMGA1 in reprogramming pancreatic epithelium into invasive cancer cells. We assessed oncogenic properties induced by HMGA1 in non-transformed pancreatic epithelial cells expressing activated K-RAS.We also explored theHMGA1-cyclooxygenase (COX-2) pathway inhumanpancreatic cancer cells and the therapeutic effects of COX-2 inhibitors in xenograft tumorigenesis. Results: HMGA1 cooperates with activated K-RAS to induce migration, invasion, and anchorageindependent cell growth in a cell line derived from normal human pancreatic epithelium. Moreover, HMGA1 and COX-2 expression are positively correlated in pancreatic cancer cell lines (r2 = 0.93; p < 0.001). HMGA1 binds directly to the COX-2 promoter at an AT-rich region in vivo in three pancreatic cancer cell lines. In addition, HMGA1 induces COX-2 expression in pancreatic epithelial cells, while knock-down of HMGA1 results in repression of COX-2 in pancreatic cancer cells. Strikingly, we also discovered that Sulindac (a COX-1/COX-2 inhibitor) or Celecoxib (a more specific COX-2 inhibitor) block xenograft tumorigenesis from pancreatic cancer cells expressing high levels of HMGA1. Conclusions: Our studies identify for the first time an important role for the HMGA1-COX-2 pathway in pancreatic cancer and suggest that targeting this pathway could be effective to treat, or even prevent, pancreatic cancer.
- High mobility group A1 (HMGA1)
- Pancreatic adenocarcinoma
- Tumor progression
ASJC Scopus subject areas
- Endocrinology, Diabetes and Metabolism