Mucins and mucin binding proteins in colon cancer metastasis

Abstract

AIM: To establish more directly the roles of the MUC2 mucin ge ne and the mucin binding protein galectin-3 in colon cancer metastasis.

METHODS: MUC2 levels were manipulated highly metastatic human co lon cancer cells using eukaryotic expression constructs designed to express a portion of MUC2 cDNA in the antisense orientation. Galectin-3 levels were also manipulated in human cancer cells using constructs designed to express the complete galectin-3 complementary DNA (cDNA) in either the sense or antisense orientation. Stable transfection was confirmed by a PCR-based approach and by Southern Analysis., Alterations in mRNA were determined by competitative RT-PCR and Northern Analysis. MUC2 apoprotein and galectin-3 protein levels were determined by Western Analysis. Liver colonization was assessed in athymic mice after splenic-portal inoculation or after spontaneous metastasis during cecal growth.

RESULTS: Stable integration of the MUC2 antisense construct into metastatic colon cancer cells (LS LiM6) resulted in an 80% reduction in MUC2-s pecific mRNA and a concomitant decrease in MUC2 apomucin protein. This was associated with a 50% reduction in synthesis of mature glucosamine labeled mucin, almost complete inhibition of secretion of sialyl Le-X and sialyl Tn antigens, and a 40% decrease in binding of colon cancer cells to endothelial Eselectin. Reduction in MUC2 levels was associated with a marked decrease in liver colonization.

Introduction of galectin-3 antisense into metastatic Lim6 colon cancer cells resulted in an 80% reduction in galectin-3 specific mRNA by quantitative dot blot analysis (normalized to actin). Northern Analysis confirmed a decrease in the 1 kb product in these cells. There was a 13-fold reduction in galectin-3 protein by Western Analysis compared to parental cell line or vector-transfect ed controls. Similar results were obtained for another metastatic colon cancer cell line for HM7. Both total cellular and cell surface (FACS analysis) galectin -3 were reduced. Transfection of galectin-3 (sense) into low metastatic LS174 T cells resulted in a 4.5-fold increase in mRNA and a 10-fold increase in galectin-3 protein. Down-regulation of galectin-3 by antisense transformation resulted in a significant (P < 0.001) decrease in liver colonization (liver weight, number of tumor nodules and percentage of parenchyma replaced by tumor) by Lim6 and HM7, while introduction of galectin-3 (sense) into LS174T resulted in a significant increase in liver colonization after both splenic-portal and cecal injection. Tumor tissue levels of galectin-3 in metastatic foci correlated with levels in injected cells. Manipulation of galection in these cell lines resulted in coordinately decreased (or increased) levels of MUC2 mucin, a major ligand for this lectin. Galectin-3 was also detected in the serum of 11/13 patients with colonic adenocarcinoma, with highest levels in those with distant metastatic disease.

CONCLUSION: These studies provide direct evidence that the MUC2 mucin gene and mucin binding protein galectin-3 play an important role in colon cancer metastasis.

Key Words: Colonic neoplasms, Neoplasm metastasis, Mucins, Mucin binding proteins, MUC2 mucin gene, Galectin-3