Redistribution of ADP-ribosylation factor during stimulation of permeabilized cells with GTP analogues

Abstract

Low-molecular-mass GTP-binding proteins of the ras family were analysed by [32P]GTP binding after PAGE and transfer to nitrocellulose membranes. By this technique, several GTP-binding proteins in the 20-30 kDa range were detected in both cytosolic and microsomal fractions of RINm5F cells. One of these, displaying an apparent molecular mass of about 20 kDa and a pI of 6.7, was mainly cytosolic and was shown to be the ADP-ribosylation factor (ARF) by using specific antibodies. When permeabilized RINm5F cells were incubated with the stable GTP analogues guanosine 5'-[gamma-thio]triphosphate (GTP[S]) and guanosine 5'-[beta gamma-imido]triphosphate (p[NH]ppG) the amount of ARF increased in a fraction containing both Golgi and plasma-membrane markers, but not in the fraction containing secretory granules, mitochondria and lysosomes. GTP, GDP and its beta-thio analogue did not enhance ARF binding to membranes, smg25/rab3 and rho, as well as all the other small GTP-binding proteins detected by the [32P]GTP binding, did not redistribute under these conditions. As GTP[S] stimulates insulin secretion in these cells, we also examined the relationship between ARF translocation and insulin secretion. Both phenomena were elicited by GTP[S] with an EC50 (median effective concentration) of about 10 microM. p[NH]ppG was equipotent with GTP[S] in inducing insulin secretion (EC50 about 10 microM), but higher concentrations (about 500 microns) were required to achieve the same maximal ARF redistribution. These results suggest that: (1) ARF is subject to cycling between a membrane-associated and a free/loosely attached form, determined by the species of bound guanine nucleotide; (2) ARF alone does not seem to regulate exocytosis in insulin-secreting cells.