Stimulation of GLUT4 (glucose transporter isoform 4) storage vesicle formation by sphingolipid depletion

Abstract

Insulin stimulates glucose transport in fat and skeletal muscle cells primarily by inducing the translocation of GLUT4 (glucose transporter isoform 4) to the PM (plasma membrane) from specialized GSVs (GLUT4 storage vesicles). Glycosphingolipids are components of membrane microdomains and are involved in insulin-regulated glucose transport. Cellular glycosphingolipids decrease during adipocyte differentiation and have been suggested to be involved in adipocyte function. In the present study, we investigated the role of glycosphingolipids in regulating GLUT4 translocation. We decreased glycosphingolipids in 3T3-L1 adipocytes using glycosphingolipid synthesis inhibitors and investigated the effects on GLUT4 translocation using immunocytochemistry, preparation of PM sheets, isolation of GSVs and FRAP (fluorescence recovery after photobleaching) of GLUT4-GFP (green fluorescent protein) in intracellular structures. Glycosphingolipids were located in endosomal vesicles in pre-adipocytes and redistributed to the PM with decreased expression at day 2 after initiation of differentiation. In fully differentiated adipocytes, depletion of glycosphingolipids dramatically accelerated insulin-stimulated GLUT4 translocation. Although insulin-induced phosphorylation of IRS (insulin receptor substrate) and Akt remained intact in glycosphingolipid-depleted cells, both in vitro budding of GLUT4 vesicles and FRAP of GLUT4-GFP on GSVs were stimulated. Glycosphingolipid depletion also enhanced the insulin-induced translocation of VAMP2 (vesicle-associated membrane protein 2), but not the transferrin receptor or cellubrevin, indicating that the effect of glycosphingolipids was specific to VAMP2-positive GSVs. Our results strongly suggest that decreasing glycosphingolipid levels promotes the formation of GSVs and, thus, GLUT4 translocation. These studies provide a mechanistic basis for recent studies showing that inhibition of glycosphingolipid synthesis improves glycaemic control and enhances insulin sensitivity in animal models of Type 2 diabetes.

Figure 1. Changes in the level and distribution of gangliosides during differentiation of 3T3-L1 cells

(a) Cells were induced to differentiate into adipocytes (see Experimental), and at the indicated days were lysed and the lipids were extracted and analyzed by TLC (upper panel), or total proteins were analyzed by western blotting to detect GLUT4 levels (lower panel). (b) At the indicated time points of differentiation, cells were fixed, permeabilized, labeled with AF594-labeled CtxB (to detect GM₁ ganglioside), and observed by confocal microscopy. Bar, 10 µm.

Figure 2. SL depletion accelerates insulin-induced GLUT4 translocation in 3T3-L1 adipocytes

(a,b) Following growth in differentiation medium for 6 days, the cells were further treated with inhibitors of SL synthesis (FB1, NB-DGJ or P4) for 4 days, serum starved, and then stimulated with 100 nM insulin for 5 min. The insulin-stimulated GLUT4 translocation to PM sheets was observed and quantified by image analysis (see Experimental). Bar, 10 µm. Results are expressed as fold over the level of GLUT4 on PM sheets from unstimulated control cells without inhibitor pretreatment. Values are the mean ± SD (n ≥ 100 PM sheets from 3 independent experiments). (c,d) Cells were pretreated with indicated inhibitors and insulin-stimulated as above. Insulin-stimulated GLUT4 translocation to the PM fraction was monitored by western blotting (c) and quantified by image analysis (d). Results are expressed as fold over the level of GLUT4 in PM fractions from unstimulated control cells without inhibitor pretreatment. Values are the mean ± SD from 3 independent experiments. (e,f) Cells were pretreated with NB-DGJ for 4 days, serum starved, and then stimulated with insulin for various times. The insulin-stimulated GLUT4 translocation to PM sheets was visualized over time by immune-fluorescence (e) and quantified by image analysis (f). Bar, 10 µm. Results are expressed as fold over the level of GLUT4 on the PM sheets from unstimulated control cells without inhibitor pretreatment. Values are the mean ± SD (n ≥ 80 PM sheets from 2 independent experiments).

Figure 3. Insulin signaling is unaffected by SL-depletion of 3T3-L1 cells

Cells were treated with FB1 or NB-DGJ for 4 days as in Fig. 2. The phosphorylation of IRS-1 and Akt, and total IRS-1 and Akt were measured by western blotting (see Experimental). Shown are representative blots from 3 independent experiments.

Figure 4. SL depletion accelerates GSV formation in 3T3-L1 cells

Cells were pretreated with FB1 or P4 for 4 days. (a) Donor membranes from cells treated ± inhibitors were incubated with cytosol from untreated cells for 20 min at 37°C. De novo formed vesicles and the donor membranes were analyzed by western blotting using antibodies against GLUT4 and the transferrin receptor. (b) Cells were treated with P4 for 4 days and then transfected with HA-GLUT4-GFP for 24 hrs. Half of the post-Golgi area was photobleached (bracketed region) and the fluorescence recovery after photobleaching was monitored by confocal microscopy (see Experimental). Bar, 10 µm. (c) Quantitation of fluorescence recovery after photobleaching shown in (b). Results are means ± SE for ≥ 10 cells from 2 independent experiments.

Figure 5. Insulin-stimulated translocation of VAMP2 and Cellubrevin in SL-depleted 3T3-L1 adipocytes

Cells were untreated (Control) or pretreated with P4 for 4 days, serum-starved, and then stimulated with insulin for 5 min. Insulin stimulated translocation of VAMP2 (a) and Cellubrevin (not shown) to the PM sheets (see Experimental). Bar, 10 µM (b) Quantitation of levels of GLUT4, VAMP2, TfnR and cellubrevin on PM sheets after 5 min stimulation with insulin. Results are expressed as fold over the level of each protein on PM sheets from unstimulated cells without inhibitor pretreatment. Values are the mean ± SD (n ≥ 100 PM sheets from 3 independent experiments).