Demonstration of differential quantitative requirements for NSF among multiple vesicle fusion pathways of GLUT4 using a dominant-negative ATPase-deficient NSF.

In this study, we investigated the relative participation of N-ethylmaleimide-sensitive factor (NSF) in vivo in a complex multistep vesicle trafficking system, the translocation response of GLUT4 to insulin in rat adipose cells. Transfections of rat adipose cells demonstrate that over-expression of wild-type NSF has no effect on total, or basal and insulin-stimulated cell-surface expression of HA-tagged GLUT4. In contrast, a dominant-negative NSF (NSF-D1EQ) can be expressed at a low enough level that it has little effect on total HA-GLUT4, but does reduce both basal and insulin-stimulated cell-surface HA-GLUT4 by approximately 50% without affecting the GLUT4 fold-translocation response to insulin. However, high expression levels of NSF-D1EQ decrease total HA-GLUT4. The inhibitory effect of NSF-D1EQ on cell-surface HA-GLUT4 is reversed when endocytosis is inhibited by co-expression of a dominant-negative dynamin (dynamin-K44A). Moreover, NSF-D1EQ does not affect cell-surface levels of constitutively recycling GLUT1 and TfR, suggesting a predominant effect of low-level NSF-D1EQ on the trafficking of GLUT4 from the endocytic recycling compared to the intracellular GLUT4-specific compartment. Thus, our data demonstrate that the multiple fusion steps in GLUT4 trafficking have differential quantitative requirements for NSF activity. This indicates that the rates of plasma and intracellular membrane fusion reactions vary, leading to differential needs for the turnover of the SNARE proteins.