Steady-state localization of a medial-Golgi glycosyltransferase involves transit through the trans-Golgi network

The steady-state localization of medial-Golgi enzymes is likely to involve retrograde transport pathways; however, the trafficking of these resident enzymes through the Golgi stack is unclear. To investigate if the medial-Golgi enzyme beta-1,2-N-acetylglucosaminyltransferase I (GlcNAc-TI) is transported to the late Golgi, a modified GlcNAc-TI bearing an N-glycan site on the C-terminus was constructed. The modified GlcNAc-TI was demonstrated to be functionally active in vivo, and was localized to the Golgi stack of transfected cells. In stable Chinese-hamster ovary (CHO) cell clones, the N-glycosylated GlcNAc-TI carried sialylated complex N-glycan chains. Pulse-chase studies showed that the majority of GlcNAc-TI was sialylated within 60 min of synthesis. Treatment of transfected CHO cells with Brefeldin A resulted in the glycosylated GlcNAc-TI bearing endo-beta-N-acetylglucosaminidase H resistant chains; however, the sialylation of glycosylated GlcNAc-TI was dramatically reduced. These data imply that, in CHO cells, newly synthesized GlcNAc-TI is transported rapidly through the Golgi stack to the trans-Golgi network, suggesting that GlcNAc-TI continuously recycles from the late Golgi. Furthermore, this data suggests that retrograde transport pathways play an important role in establishing the asymmetric distribution of GlcNAc-TI within the Golgi stack.