Expression of core 2 beta1,6-N-acetylglucosaminyltransferase facilitates prostate cancer progression.

Cell surface carbohydrates expressed on epithelial cells are thought to play an important role in tumor progression. Previously, we have shown that expression of core 2-branched O-glycans is closely correlated with vessel invasion and depth of invasion in colon and lung carcinomas. In this study, we found that expression of core 2 beta1,6-N-acetylglucosaminyltransferase-1, Core2GnT, is positively correlated with the progression of prostate cancer in human patients. Statistical analysis demonstrated that Core2GnT is an independent predictor for progressed pathological stage (pT3) and for prostate-specific antigen (PSA) relapse. To determine directly the roles of Core2GnT in prostate cancer progression, we set up an experimental tumor model using the LNCaP prostate cancer cell line. Because this line does not express Core2GnT, we established an LNCaP line stably expressing Core2GnT, LNCap-Core2GnT, by transfecting cDNA encoding Core2GnT. When mock-transfected LNCaP cells and LNCaP-Core2GnT were inoculated in the prostate of nude mice, LNCaP-Core2GnT cells produced three times heavier prostate tumors than mock-transfected LNCaP cells. Furthermore, we found that LNCaP-Core2GnT cells adhered more strongly to prostate stromal cells, type IV collagen and laminin than did LNCaP-mock cells, but LNCaP and LNCaP-Core2GnT cells grew almost at the same rate on plates coated with type IV collagen or laminin. These results indicate that Core2GnT is an extremely useful prognostic marker for prostate cancer progression. The results also suggest that acquiring Core2GnT in prostate carcinoma cells facilitates adhesion to type IV collagen and laminin, and this increased adhesion may be a cause for aggressive tumor formation by prostate cancer cells expressing Core2GnT.