ABSTRACT
Tumor cell invasion is a process regulated by integrins, matrix-degrading enzymes, and interactions with host tissue stromal cells. The ADAM family of proteins plays an important role in modulating various cellular responses. Here, we show that an alternatively spliced variant of ADAM9 is secreted by hepatic stellate cells and promotes carcinoma invasion. ADAM9-S induced a highly invasive phenotype in several human tumor cell lines in Matrigel assays, and the protease activity of ADAM9-S was required for invasion. ADAM9-S binds directly to alpha6beta4 and alpha2beta1 integrins on the surface of colon carcinoma cells through the disintegrin domain. ADAM9-S was also able to cleave laminin and promote invasion. Analysis of human liver metastases revealed that ADAM9 is expressed by stromal liver myofibroblasts, particularly those that are localized within the tumor stroma at the invasive front. These results emphasize the importance of tumor-stromal interactions in invasion and suggest that ADAM9-S can be an important determinant in the ability of cancer cells to invade and colonize the liver.
Subject(s)
Cell Communication/physiology , Colonic Neoplasms/pathology , Disintegrins/physiology , Liver Neoplasms, Experimental/secondary , Liver/pathology , Membrane Proteins/physiology , Metalloendopeptidases/physiology , ADAM Proteins , Amino Acid Sequence , Animals , Cell Line, Tumor , Collagen , Disintegrins/biosynthesis , Disintegrins/metabolism , Drug Combinations , Humans , Integrin alpha2beta1/metabolism , Integrin alpha6beta4/metabolism , Laminin , Liver/metabolism , Liver Neoplasms, Experimental/metabolism , Membrane Proteins/biosynthesis , Membrane Proteins/metabolism , Metalloendopeptidases/biosynthesis , Metalloendopeptidases/metabolism , Mice , Molecular Sequence Data , Neoplasm Invasiveness , Proteoglycans , Stromal Cells/metabolism , Stromal Cells/pathologyABSTRACT
BACKGROUND & AIMS: Several lines of evidence indicate that aldosterone antagonists may exert direct antifibrogenic effects. The aim of this study was to evaluate the possible direct antifibrogenic effects of canrenone, the active metabolite of spironolactone, in activated human hepatic stellate cells. METHODS: The effects of canrenone were assessed on platelet-derived growth factor-induced mitogenic and chemotactic effects and the increased de novo synthesis of different extracellular matrix components induced by transforming growth factor-beta1. RESULTS: Canrenone dose-dependently reduced platelet-derived growth factor-induced cell proliferation and motility. This effect was not associated with either changes in the phosphorylation of platelet-derived growth factor receptor and phospholipase C gamma or in the activation of the Ras/extracellular signal-regulated kinase pathway, whereas it was accompanied by a dose-dependent inhibition of platelet-derived growth factor-induced phosphatidylinositol 3-kinase activity. In addition, canrenone inhibited the activity of the Na(+)/H(+) exchanger 1 induced by platelet-derived growth factor. The effect of canrenone on Na(+)/H(+) exchanger 1 activity was reproduced by phosphatidylinositol 3-kinase inhibitors, thus supporting an inhibitory action of canrenone on phosphatidylinositol 3-kinase activity. To further address this possibility, the action of canrenone was compared with that of 2 established Na(+)/H(+) exchanger 1 inhibitors: ethylisopropylamiloride and cariporide. Whereas ethylisopropylamiloride was able to inhibit platelet-derived growth factor-induced phosphatidylinositol 3-kinase activity, cariporide was without any effect. Both compounds reproduced the effects of canrenone on platelet-derived growth factor-induced mitogenesis and chemotaxis. Finally, canrenone was able to reduce transforming growth factor-beta1-induced de novo synthesis of procollagen type I/IV and fibronectin and thrombin-induced hepatic stellate cell contraction. CONCLUSIONS: These results indicate that canrenone may be active as an antifibrogenic drug.