Search

A hypothesis that accounts for most of the available literature on insulin-stimulated GLUT4 translocation is that insulin action controls the access of GLUT4 vesicles to a constitutively active plasma-membrane fusion process. However, using an in vitro fusion assay, we show here that fusion is not constitutively active. Instead, the rate of fusion activity is stimulated 8-fold by insulin. Both the magnitude and time course of stimulated in vitro fusion recapitulate the cellular insulin response. Fusion is cell cytoplasm and SNARE dependent but does not require cell cytoskeleton. Furthermore, insulin activation of the plasma-membrane fraction of the fusion reaction is the essential step in regulation. Akt from the cytoplasm fraction is required for fusion. However, the participation of Akt in the stimulation of in vitro fusion is dependent on its in vitro recruitment onto the insulin-activated plasma membrane.

The translocation of glucose transporter-4 (GLUT4) to the cell surface is a complex multistep process that involves movement of GLUT4 vesicles from a reservoir compartment, and docking and fusion of the vesicles with the plasma membrane. It has recently been proposed that a p38 mitogen-activated protein kinase (MAPK)-dependent step may lead to intrinsic activation of the transporters exposed at the cell surface. In contrast to data obtained in muscle and adipocyte cell lines, we found that no insulin activation of p38 MAPK occurred in rat adipose cells. However, the p38 MAPK inhibitor SB203580 consistently inhibited transport activity after preincubation with the adipose cells. These apparently contradictory findings led us to hypothesize that the inhibitor may have a direct effect on the transport catalytic activity of GLUT4 that was independent of inhibition of the kinase. Kinetic analysis of 3-O-methyl-d-glucose transport activity revealed that SB203580 was a noncompetitive inhibitor of zero-trans (substrate outside but not inside) transport, but was a competitive inhibitor of equilibrium-exchange (substrate inside and outside) transport. This pattern of inhibition of GLUT4 was also observed with cytochalasin B. The pattern of inhibition is consistent with interaction at the endofacial surface, but not the exofacial surface of the transporter. Occupation of the endofacial substrate site reduces maximum velocity under zero-trans conditions, because return of the substrate site to the outside is blocked, and no substrate is present inside to displace the inhibitor. Under equilibrium-exchange conditions, internal substrate competitively displaces the inhibitor, and the transport K(m) is increased.

Abnormalities in intracellular pH regulation have been proposed to be important in type 2 diabetes and the associated cardiomyopathy and hypertension. We have therefore investigated the dependence of insulin-stimulated glucose transport on cytosolic pH in cardiomyocytes. Insulin treatment of cardiomyocytes resulted in a marked alkalinization of the cytoplasm as measured using carboxy-semi-napthorhodofluor-1. The alkalinizing effect of insulin was blocked by treatment with either cariporide (which inhibits the Na+/H+ exchanger) or by bafilomycin A1 (which inhibits H+-ATPase activity). After treatments with cariporide or bafilomycin A1, insulin stimulation of insulin receptor and insulin receptor substrate-1 phosphorylation and Akt activity were normal. In contrast, glucose transport activity and the levels of functional GLUT4 at the plasma membrane (detected using an exofacial photolabel) were reduced by approximately 50%. Immunocytochemical analysis revealed that insulin treatment caused a translocation of the GLUT4 from perinuclear structures and increased its co-localization with cell surface syntaxin 4. However, neither cariporide nor bafilomycin A1 treatment reduced the translocation of immunodetectable GLUT4 to the sarcolemma region of the cell. It is therefore hypothesized that insulin-stimulated cytosol alkalinization facilitates the final stages of translocation and incorporation of fully functional GLUT4 at the surface-limiting membrane.

Nycodenz gradients have been used to examine the in vitro effects of GTP-(gamma)-S on adaptor complex association with GLUT4 vesicles. On addition of GTP-(gamma)-S, GLUT4 fractionates as a heavier population of vesicles, which we suggest is due to a budding or coating reaction. Under these conditions there is an increase in co-sedimentation of GLUT4 with AP1, but not with AP3. Western blotting of proteins associated with isolated GLUT4 vesicles shows the presence of high levels of AP1 and some AP3 but very little AP2 adaptor complexes. Cell free, in vitro association of the AP1 complex with GLUT4 vesicles is increased approximately 4-fold by the addition of GTP-(gamma)-S and an ATP regenerating system. Following GTP-(gamma)-S treatment in vitro, ARF is also recruited to GLUT4 vesicles, and the temperature dependence of ARF recruitment closely parallels that of AP1. The recruitment of both AP1 and ARF are partially blocked by brefeldin A. These data demonstrate that the coating of GLUT4 vesicles can be studied in isolated cell-free fractions. Furthermore, at least two distinct adaptor complexes can associate with the GLUT4 vesicles and it is likely that these adaptors are involved in mediating distinct intracellular sorting events at the level of TGN and endosomes.

New cell-impermeant bis-mannose photolabels have been developed with biotinyl groups attached to 4-(1-azi-2,2,2-trifluoroethyl)-benzoyl-1, 3-bis(d-mannos-4-yloxy)-2-propylamine (ATB-BMPA) by either a polyethoxy spacer (Bio-ATB-BMPA) or an additional hexanoic acid spacer (Bio-LC-ATB-BMPA). The half-maximal inhibition constants, Ki values, for inhibition of glucose transport activity in insulin-stimulated rat adipocytes were determined to be 359+/-10 and 273+/-28 microM for Bio-ATB-BMPA and Bio-LC-ATB-BMPA, respectively. These values are similar to those previously reported for the non-biotinylated compound ATB-BMPA. Following UV-irradiation-induced cross-linking of the biotinylated photolabels to rat adipocytes, the biotinylated glucose transporter isoform 4 (GLUT4) could be detected by non-radioactive and radioactive methods that utilized the interaction with streptavidin. Biotinylated GLUT4 from 1-2 microg of adipose cell membranes, precipitated onto magnetic streptavidin beads, could be sensitively and quantitatively detected using an electrochemiluminescent assay method. This utilized a ruthenium-tagged anti-GLUT4 antibody that on excitation at an electrode generated an electrochemiluminescent signal in an ORIGEN analyser. Alternatively, surface-biotinylated GLUT4 could be easily, but less sensitively, detected in streptavidin agarose precipitates which were analysed by conventional GLUT4 Western blotting. Data obtained using the non-radioactive methods compared favourably with those using tritiated versions of the biotinylated probes. Insulin treatment of adipocytes increased the levels of signals from surface biotinylated GLUT4 by approximately 10-fold or approximately 20-fold, respectively, when the electrochemiluminescent or the Western blot detection methods were used and these signals were blocked by cytochalasin B.