Rabip4 expression induces an increase in glucose uptake and Glut 4 translocation to the plasma membrane. 3T3-L1 Rabip4 Tet-off adipocytes were cultivated for 3 days with or without tetracycline (+/-tetra) to induce the expression of Rabip4. Cells were serum starved overnight, and then treated with insulin for 20 minutes. (A) Deoxyglucose uptake was then determined as described in the Materials and Methods. The results were expressed as the mean ± s.e.m. of five experiments of the fold increase in insulin stimulation over the basal condition of cells cultivated with tetracycline (no Rabip4 expression). (B) 40 μg protein from 3T3-L1 Rabip4 Tet-off adipocytes cultivated with or without tetracycline for 3 days were separated by SDS-PAGE and analyzed by western blot for the presence of Rabip4, Glut 4 and Rab4. (C) Glut 4 translocation was determined by the amount of Glut 4 present in plasma membrane sheet preparations as described in the Materials and Methods. Data are the mean ± s.e.m. of three experiments of the fold increase in insulin stimulation over the basal condition of cells cultivated with tetracycline (no Rabip4 expression). (D) 3T3-L1 Rabip4 Tet-off adipocytes cultivated with or without tetracycline for 3 days were serum starved overnight before being stimulated with 100 nM insulin. Insulin-induced tyrosine phosphorylation was detected using anti-phosphotyrosine antibodies. YP-IR, tyrosine-phosphorylated β subunits of the insulin receptor; YP-IRS, tyrosine-phosphorylated insulin receptor substrates. Activation of protein kinase B (PKB) and extracellular-regulated kinases (ERK) was estimated using anti phospho-PKB and anti phospho-ERK. *P<0.001 compared with levels in the +tetra control.
Rabip4 and Glut 4 define distinct intracellular compartments. Endogenous Rabip4 and Glut 4 were studied by immunofluorescence. (a,b) Rabip4 was detected using rabbit antibodies to Rabip4 followed by FITC-coupled donkey anti-rabbit antibodies. (c, d) Glut 4 was detected using goat anti-Glut 4 antibodies followed by Texas-Red-coupled donkey anti-goat antibodies. (e-h) Merged images of Rabip4 and Glut 4 labeling prepared using Adobe Photoshop. Panels g and h are enlarged views of the boxes in panels e and f. Left and right columns correspond to confocal sections obtained near the bottom and in the middle of the adipocyte, respectively. Bar, 1 μm.
Rabip4-positive compartments are accessible to internalized Glut 4. 3T3-L1 adipocytes expressing GFP-Rabip4 and DsRed-Glut 4-myc were treated with insulin (100 nM, 20 minutes) and incubated with anti-myc mAb for 2 hours at 4°C. Cells were then extensively washed and either fixed (A) or incubated for 1 hour at the following temperatures: 15°C (B), 20°C (C) or 37°C (D). After fixation and permeabilization, anti-myc mAb were revealed using Cy5-coupled anti-mouse antibodies. Images corresponding to GFP-Rabip4 (in green), internalized anti myc mAb (in blue), the merge of these two images and DsRed-Glut 4-myc (in red) are shown. Colocalization of GFP-Rabip4 with internalized anti-myc mAb (internalized Glut 4) results in a light blue color. Asterisks indicate GFP-Rabip4-positive structures that also contain anti-myc mAb. Bar, 1 μm.
Characterization of Rabip4 subcellular localization by immunofluorescence studies. (A) Subcellular localization of Rabip4 compared with organelle markers. 3T3-L1 adipocytes were processed for indirect immunofluorescence using antibodies to Rabip4 (a,a′,d), together with anti EEA1 mAb (b,b′) or anti-TfR mAb (e). Panels c,c′, and f show the corresponding merged images. One confocal section taken in the middle of an adipocyte is shown for each labeling. a′-c′ are enlarged views of the boxed regions in a-c. Arrowheads indicate vesicles containing juxtaposed EEA1 and Rabip4 labeling. The insets in panels d-f correspond to an enlarged view of the perinuclear region. In panels g,h,i, TGN-38 localization (shown in h) was analyzed versus GFP-Rabip4 (shown in g). Panel i shows the merge images. (B) Rabip4 is mainly colocalized with Rab4, but not Rab5, Rab11, or Rab7. 3T3-L1 adipocytes expressing Myc-Rabip4 together with GFP-Rab4 (a-c), GFP-Rab5 (d-f), or GFP-Rab7 (g-i) or GFP-Rabip4 with myc-Rab11(j-l) were fixed and permeabilized. Panels a,d,g,j show the green labeling of the different GFP-fusion proteins. Panels b,e,h show labeling of myc Rabip4 and panel k shows myc Rab11. These myc-tagged protein are detected with anti-myc mAb followed by Texas-Red-coupled anti-mouse antibodies. Panels c,f,i,l are merged images. Bars, 1 μm.
Transferrin internalized at 20°C reaches Rabip4-positive vesicles. The same experiment was performed using 3T3-L1 adipocytes (a-f) and HeLa cells (g-l). Cells were incubated for 1 hour at 20°C (a-c, g-i) or 4°C (d-f, j-l) with Texas-Red-coupled transferrin (500 μg/ml). Cells were washed in PBS, fixed with paraformaldehyde and permeabilized. Endogenous Rabip4 was detected by using specific antibodies followed by FITC-coupled anti-rabbit antibodies. Arrowheads indicate Rabip4-positive vesicles containing internalized transferrin. The inserts are enlargements of the boxed regions. Bar, 1 μm.
Rabip4 translocates to the plasma membrane in response to insulin. (A) Plasma membrane sheets were prepared from 3T3-L1 adipocytes, untreated or treated with 100 nM insulin, for 20 minutes. Rabip4 and Glut 4 were detected by immunofluorescence using rabbit anti-Rabip4 antibodies and goat anti-Glut 4 antibodies followed by incubation with FITC-coupled anti-rabbit antibodies and Texas-Red-coupled anti-goat antibodies, respectively. Bar, 10 μm. (B) Quantification of fluorescence levels from A, performed using MetaMorph software as described in the Materials and Methods. (C) 3T3-L1 adipocytes were untreated or treated with 100 nM insulin for 20 minutes. Plasma membranes (PM), intracellular microsomes (IM), and cytosol (Cyt) were then prepared and 40 μg proteins from each fraction were separated by SDS-PAGE and transferred to PVDF membrane. Antibodies against Rabip4, Glut4 and Rab4 were then used to probe the blot.
Expression of Rabip4 unable to bind Rab4 does not affect insulin-induced glucose transport, but induces the translocation of Glut 4-myc-DsRed to the plasma membrane. (A-B) 3T3-L1 Rabip4 Δ507-517 Tet-off adipocytes were cultivated for 3 days with or without tetracycline (tetra) to induce the expression of Rabip4Δ507-517. (A) Serum-deprived 3T3-L1 Rabip4 Δ507-517 Tet-off adipocytes were untreated or treated with the indicated concentrations of insulin for 20 minutes. Deoxyglucose uptake was then measured and the results of three independent experiments are shown, expressed, as the fold increase compared with basal levels in tetracycline-treated cells. The inset shows the expression of Rabip4 Δ507-517 in adipocytes cultivated with or without tetracycline by western blot using anti-Rabip4 antibodies. (B) Serum-deprived cells were untreated or treated with 100 nM insulin for 20 minutes and used to prepare plasma membrane sheets as described in the Materials and Methods. The graph represents the mean ± s.e.m. of three experiments of the fold increase in insulin stimulation over the basal condition of cells cultivated with tetracycline (no Rabip4 expression). The amounts of Glut 4 and of transferrin receptors were quantified. (C-D) 3T3-L1 adipocytes expressing Glut 4-myc-DsRed in combination with GFP or GFP-Rabip4 Δ507-517 were untreated or treated with 100 nM insulin for 20 minutes. Fixed cells were incubated with anti-myc mAb that binds the extracellular myc epitope of Glut 4-myc-DsRed incorporated into the plasma membrane. The anti-myc antibody is then detected using Cy5-coupled anti mouse antibodies to visualize the plasma membrane Glut 4 (mGlut4). (C) Typical labeling for GFP and GFP-Rabip4 Δ507-517 (in green), Glut 4-myc-DsRed (Glut 4, in red) and membrane Glut 4-myc-DsRed (mGlut 4) (in blue) are shown for basal and insulin-stimulated conditions. (D) The number of adipocytes with Glut 4-myc-DsRed detected at the plasma membrane was counted. 50-100 co-transfected cells were counted in each experimental condition and the figure shows the mean ± s.e.m. of the numbers obtained in six independent experiments. *P<0.001 compared with levels in the +tetra control.
Expression of a mutated form of Rabip4 unable to bind Rab4 affects Glut 4 trafficking between early endosomes and its storage compartment. 3T3-L1 Rabip4 Δ507-517 Tet-off adipocytes were cultivated for 2 days in the presence (a-c), or absence of tetracycline (d-f) to induce the expression of Rabip4 Δ507-517. Cells were serum deprived before a 20 minute treatment with 100 nM insulin. After extensive washes in DMEM containing 0.5% BSA, the adipocytes were incubated for 1 hour at 37°C. They were then fixed, permeabilized and used for immunofluorescence analysis. Glut 4 was detected using anti-Glut 4 goat polyclonal antibodies followed by FITC-coupled anti-goat antibodies (a,d), whereas TfR were detected with anti-TfR mAb followed by Texas-Red-coupled anti-mouse antibodies (b,e). The merged images are shown in panels c and f. The dotted square inserts show cytoplasmic vesicles with Glut 4 labeling in green and VAMP2 labeling in red. Bar, 1 μm.
Model of Rabip4 involvement in Glut 4 recycling. We propose that Rabip4, located on specialized subdomains of early endosomes and at the plasma membrane, favors the trafficking of Glut 4 towards its sequestration compartment. Rabip4-positive endosomes would constitute a post early endosome Glut 4 sorting compartment. Uncoupling between Rabip4 and Rab4 would lead to a specific recruitment of inactive Glut 4 (in blue) towards the plasma membrane. Insulin recruits Glut 4 from its sequestration compartments as well as from the endosomes and is necessary to allow plasma membrane Glut 4 to transport glucose (active Glut 4 in green). TfR, transferrin receptor; Rabip4ΔRab4BD, Rabip4 without Rab4 binding domain or Rabip4 Δ507-517.