Dapagliflozin Inhibits Cell Adhesion to Collagen I and IV and Increases Ectodomain Proteolytic Cleavage of DDR1 by Increasing ADAM10 Activity

Dapagliflozin, empagliflozin, tofogliflozin, selective inhibitors of sodium-glucose cotransporter 2 (SGLT2), is used clinically to reduce circulation glucose levels in patients with type 2 diabetes mellitus by blocking the reabsorption of glucose by the kidneys. Dapagliflozin is metabolized and inactivated by UGT1A9. Empagliflozin is metabolized and inactivated by UGT1A9 and by other related isoforms UGT2B7, UGT1A3, and UGT1A8. Tofogliflozin is metabolized and inactivated by five different enzymes CYP2C18, CYP3A4, CYP3A5, CYP4A11, and CYP4F3. Dapagliflozin treatment of HCT116 cells, which express SGLT2 but not UGT1A9, results in the loss of cell adhesion, whereas HepG2 cells, which express both SGLT2 and UGT1A9, are resistant to the adhesion-related effects of dapagliflozin. PANC-1 and H1792 cells, which do not express either SGLT2 or UGT1A9, are also resistant to adhesion related effects of dapagliflozin. On the other hand, either empagliflozin or tofogliflozin treatment of HCT116, HepG2, PANC-1, and H1792 cells are resistant to the adhesion-related effects as observed in dapagliflozin treated HCT116 cells. Knockdown of UGT1A9 by shRNA in HepG2 cells increased dapagliflozin sensitivity, whereas the overexpression of UGT1A9 in HCT116 cells protected against dapagliflozin-dependent loos of cell adhesion. Dapagliflozin treatment had no effect on cellular interactions with fibronectin, vitronectin, or laminin, but it induced a loss of interaction with collagen I and IV. In parallel, dapagliflozin treatment reduced protein levels of the full-length discoidin domain receptor I (DDR1), concomitant with appearance of DDR1 cleavage products and ectodomain shedding of DDR1. In line with these observations, unmetabolized dapagliflozin increased ADAM10 activity. Dapagliflozin treatment also significantly reduced Y792 tyrosine phosphorylation of DDR1 leading to decrement of DDR1 function and detachment of cancer cells. Concomitant with these lines of results, we experienced that CEA in patients with colon cancer, which express SGLT2 but not UGT1A9, and type 2 diabetes mellitus treated by dapagliflozin in addition to chemotherapy was decreased (case 1). CEA in patients with colon cancer, which express SGLT2 but not UGT1A9, and type 2 diabetes mellitus was treated by dapagliflozin alone after radiation therapy was decreased but started to rise after cessation of dapagliflozin (case 2). CA19-9 in two of patients with pancreatic cancer and type 2 diabetes mellitus was resistant to the combination therapy of dapagliflozin and chemotherapy (case 3 and 4 respectively). PIVKAII in patients with liver cancer and type 2 diabetes mellitus, and CYFRA in patients with squamous lung cancer and type 2 diabetes mellitus was also resistant the combination therapy of dapagliflozin and chemotherapy (case 5 and 6 respectively). Taken together, these data suggest a potential role for dapagliflozin anticancer therapy against colon cancer cells that express SGLT2, but not UGT1A9.