A Transfected Sialyltransferase That Is Elevated in Breast Cancer and Localizes to the medial/trans-Golgi Apparatus Inhibits the Development of core-2–based O-Glycans

Whitehouse, Caroline; Burchell, Joy; Gschmeissner, Stephen; Brockhausen, Inka; Lloyd, Kenneth O.; Taylor-Papadimitriou, Joyce

There is a formal possibility that overexpression of a glycosyltransferase may alter the fine localization of the enzyme. This has been suggested from data localizing the ␤1,2 Figure 9 . Distribution of the stably expressed myc epitope-tagged ␣2,3 sialyltransferase in the Golgi apparatus of T47D (a) and MTSV1-7 (b-d) cell lines. In a and b, the sections were stained by single label immunogold microscopy using the 9E10 antibody and a second antibody coupled to 10-nm gold particles (solid arrowheads). c and d show double labeling of MTSV1-7 3STMYC 3.14 with polyclonal anti-GalT and monoclonal 9E10 antibodies followed by protein A coupled to 5-nm gold (GalT; open arrowheads) or goat antimouse coupled to 10-nm (␣2,3 SAT [O]; closed arrowheads) gold. Fig. 9 , and the distribution of the gold particles was expressed as relative percentage frequency across the Golgi stack (nm). 307 (Gal-T) and 354 (␣2,3 SAT [O] ) gold particles were counted over 18 Golgi, and the results are expressed as the mean percentage frequency. Using the Stratified Wilcoxon rank sum test, a Mann-Whitney statistic of 76820, P Ͻ 0.0001, shows that the difference in distribution between ␣2,3 SAT (O) and GalT is statistically significant. N-acetyl glucosaminyltransferase (NAGT-1), where direct immunolocalization of the endogenous enzyme showed localization to the medial compartment of the Golgi apparatus (Dunphy et al., 1985) , while the transfected tagged enzyme (fourfold increased expression) located to both the medial and trans compartments (Nilsson et al., 1993) . It must be noted, however, that the sensitivity of the method used for quantitating the distribution throughout the Golgi apparatus can affect the outcome, and with an increase in the level of expression comes an increase in the sensitivity of detection. The data of Rabouille and colleagues argue against an effect of overexpression since the localization of a tagged N-linked ␣2,6 sialyltransferase was not altered when it was expressed at widely different levels (Rabouille et al., 1995) . Even if overexpression of the enzyme does result in a broadening of the distribution in the Golgi apparatus, this would also be expected to occur when the endogenous enzyme is overexpressed, as it is in breast cancers by 8-10-fold. Thus, any overlap or competition with other glycosyltransferases such as is suggested from our data would also be likely to occur in the malignant cells.