Truncation of the extracellular region abrogrates cell contact but retains the growth-suppressive activity of E-cadherin.

Abstract

E-cadherin has been demonstrated to induce growth suppression and decrease the invasiveness of cancer cells and thus has been proposed to be a tumor suppressor gene. The ability of E-cadherin to mediate cell-cell contact and contact inhibition presumably accounts for its antitumor effects, which are attributed to the extracellular domain of the protein. Here we report that blocking the ability of E-cadherin to mediate contact inhibition by either antagonistic antibodies or expression of a mutant form of E-cadherin with the extracellular region deleted does not abrogate growth suppression. Transfection of the E-cadherin gene into the human prostate cancer cell line TSU.Pr-1 induced cell-cell contact formation, growth suppression, and redistribution of beta-catenin to the cell membrane. Treatment of the E-cadherin transfectant (CAD) with blocking antibodies disrupted cell-cell contact formation but did not influence the growth rate, suggesting that cell-cell interaction is not required for E-cadherin-mediated growth suppression. Similarly, transfection of an E-cadherin construct in which the NH2-terminal (extracellular) region was deleted did not allow cell-cell contact formation but induced growth suppression. In contrast, transfection of an E-cadherin construct in which the COOH-terminal (cytoplasmic) region was deleted did not induce suppression but promoted cell contact formation. In cells expressing E-cadherin lacking the cytoplasmic region, beta-catenin was evenly distributed in the cytoplasm. By contrast, in cells expressing E-cadherin lacking the extracellular region, beta-catenin was cell membrane associated. Growth suppression was always associated with the localization of beta-catenin to the cell membrane. The redistribution of beta-catenin from the cytoplasm to the cell membrane initially suggested the involvement of the Wnt signaling pathway in regulating cell growth. However, only small differences in beta-catenin/T-cell factor signaling were detected in control and E-cadherin-expressing cells, suggesting that the Wnt pathway is not involved. Taken together, these findings suggest that E-cadherin-induced growth inhibition may not be solely attributed to contact inhibition but may involve the redistribution of beta-catenin from the cytoplasm to the cell membrane, and this redistribution may affect growth pathways independent of T-cell factor.