Carcinoembryonic antigen (CEA) may promote site-specific metastasis through an intercellular adhesion mechanism. We have a model of experimental metastasis in which human colorectal carcinoma cells injected into the spleens of athymic nude mice produce liver and lung colonies in more than 40% of recipient mice (highly metastatic tumors), less than 30% of mice (weakly metastatic tumors), or in no mice (nonmetastatic tumors). CEA injected intravenously promotes colon formation by all weakly metastatic human colorectal carcinomas but not metastasis by a non-metastatic line or by 4 highly metastatic carcinomas. CEA pretreatment also increases the retention of weakly metastatic tumor cells within the liver 4 hours after cell injection while several colorectal carcinoma cell lines bind to CEA attached to a solid phase in vitro. CEA-producing colorectal carcinoma limes bind to Kupffer cells in vitro. We propose to test this hypothesis by investigating how colorectal carcinomas bind to CEA and the extent to which CEA-mediated adhesion participates in the formation of metastasis by addressing the following specific aims: 1. To determine whether CEA enhances the binding of colorectal carcinoma cella to subsets of host cells. 2. To determine whether the binding of carcinoma cells to Kupffer and other liver cells involves homophilic CEA-CEA interaction or heterophilic CEA- binding protein interaction. 3. To elucidate the function of CEA and CEA-binding proteins in the formation of metastasis in nude mice. These studies are important because showers of tumor cells may be liberated into the portal vein when a surgeon performs a potentially curative resection of primary colorectal carcinomas. Knowledge of the mechanisms of attachment and implantation may provide strategies to prevent the establishment of such metastasis.
Showing the most recent 10 out of 49 publications
