We further found that GEP100 down-regulation with shRNA significantly inhibited the Matrigel invasion ability of PaTu8988 cells, but had no effect on the cell viability, migration and adhesion, indicating that GEP100 is particularly responsible for the invasive ability of cells. In vivo experiment also showed that GEP100 knock-down significantly inhibited the liver metastasis of pancreatic cancer cells in Balb/c nude mice. This result was in line with previous data demonstrating that in breast cancer cells GEP100 was highly expressed in the invasive ones and that GEP100 down-regulation inhibited cell invasion and lung metastasis. These results indicated that GEP100 is not only a CUDC-907 target for breast cancer, it might be a general mechanism used by other cancer types. Several other biological functions for GEP100 have been reported. In HeLa cells, GEP100 regulated cell adhesion through controlling endocytosis and recycling of integrin b. In a liver carcinoma cell line HepG2, GEP100 directly interacted with acatenin and played a role in actin cytoskeleton remodeling. In myoblasts and macrophages, GEP100 was involved in cell-cell fusion. It has also been reported to take part in the regulation of nucleolar architecture and phagocytosis. In this study, we did not observe any significant effect of GEP100 on cell adhesion to collagen matrix or migration activity. Since there is not any report on the role of GEP100 in pancreatic cancer cells, this may be due to the difference in cell type. Unlike the previous report on a close relation between Arf6 expression and breast cancer cell invasive ability, we failed to detect that in the panel of pancreatic cancer cell lines we used. The above mentioned biological functions of GEP100 have been found to be dependent on the activation of Arf6, but it is believed that GEP100 is a multifunctional protein that regulates cellular functions in both an Arf-dependent and -independent manner. For example, GEP100 was involved in apoptotic cell death independent of Arf6 activity. Further studies including determination of the activation status of Arf6 will be necessary to reveal the role of Arf6 in the process of pancreatic cancer cell invasion. We found GEP100 knock-down caused a morphological change of cells from mesenchymal to epithelial phenotype. In pancreatic cancer, shifting of their epithelial features toward a mesenchymal phenotype enhances cell motility and is considered to be a prerequisite for tumor invasion. E-cadherin is the best characterized molecular marker in epithelial cells and localized at the adherens junctions. Loss of E-cadherin expression and/or function is a well-recognized marker of EMT and promotes invasion. Therefore we examined the expression of Ecadherin protein. GEP100 shRNAs treatment increased Ecadherin expression level 2 to 3 -fold. This is consistent with Hiroi’s report showing that in HepG2 cell, down-regulation of GEP100 increased the expression of E-cadherin. In addition, they also demonstrated that GEP100 interacted with a-catenin.