GFP (GLUT4, A.U.) Cy3 (IRAP, A.U.) C A BC G H I GFP(GLUT4) Cy3(IRAP) DE F Suppl. Fig. 1 % Colocalization GLUT4/IRAP(+): 62% IRAP/GLUT4(+): 65% J threshold 2 threshold WTFAEEFA/EE threshold A threshold B % colocalization K Colocalization of GLUT4 with IRAP

Suppl. Fig. 1 Supplemental Fig. 1: Quantification of the percent colocalization of HA-GLUT4-GFP with IRAP (Cy3). Panels A-J Step 1: TIRF microscopy GFP images (example in panel A) were processed by median ranking filter in Metamorph software to produce a background image (B). The background image was subtracted from the original image to produce a background corrected image (C). The same process was performed for TIRF images collected in the Cy3 channel of the same fields of cell as the GFP channels [background image (E) was subtracted from corresponding raw image (D) to yield a background corrected Cy3 image, F]. Step 2: Regions of interest (ROI) were identified in the background-corrected GFP images with the internal threshold objects in Metamorph using an empirically determined low-intensity threshold value (G is internal threshold processed image for GFP channel of the cell shown). A single low-intensity threshold value was used to process GFP images for all cells and all conditions collected in a given experiment. The ROI mask (G) was transferred to the corresponding background-corrected GFP (H) and Cy3 images (I), and the fluorescence power within the ROI in the GFP and Cy3 images were logged for analysis. Step 3: For the same image sets, ROI were identified in the background- corrected Cy3 images using the internal threshold objects in Metamorph and an empirically determined low-intensity threshold value (not shown). A single low- intensity threshold value was used to process all Cy3 images for all cells and all conditions collected in a given experiment. The ROI mask was transferred to the corresponding background-corrected Cy3 and GFP images, and the fluorescence power within the ROI in the Cy3 and GFP images were logged for analysis. Step 4: The fluorescence power of the individual identified ROI were plotted. In panel J these values for the example cell shown are shown. The green circles are ROI identified in the GFP channel and the red circles are ROI identified in the Cy3 channel. The empirically determined low-intensity threshold values used in the analysis are noted in panel J (threshold 1 and threshold 2, respectively). ROI to the right of the green line and above the red line were counted as positive for both probes. The percent colocalization is the percent of the ROI identified in the red channel that are also positive in the green channel and vice versa. Panel K. The colocalization percentages will vary with the low-intensity threshold. To insure that the differences in colocalizations of the mutants compared to WT GLUT4 are characteristic of the mutants, we analyzed the data using two different low-intensity threshold values. As shown in Panel K, the percentage colocalizations changed however the differences between the mutants and WT GLUT4 in colocalization with IRAP were not affected by the threshold value. These data confirm that the conclusions regarding the effects of the mutations on localizations of GLUT4 with IRAP are characteristics of the mutants.

Suppl. Fig. 2 Supplemental Figure 2. Effect of Akti1/2 and Wortmannin on phosphorylation of Akt and AS160. (A) Western blots demonstrating the phosphorylation of Akt at Serine 308, Threonine 473 and AS160 at Threonine nM wortmannin and 1 µ M Akti1/2 (B) Quantification of Akt phosphorylation at Threonine 308. Shown are averages ± SEM from 3 experiments of phospho- Akt 308 /total Akt. (C) Quantification of Akt phosphorylation at Serine 473. Shown are averages ± SEM from 5 experiments of phospho-Akt 473 /total Akt. (D) Quantification of AS160 phosphorylation. Shown are averages ± SEM from 3 experiments of phospho-AS /total AS160. Immunoblotting and antibodies Rabbit anti-actin was purchased from Cytoskeleton (Denver, CO). Antibodies against Akt and phospho-Akt (Ser 473, Thr 308 ) were purchase from Cell Signaling Technologies (Beverly, MA). Rabbit anti-AS160 and phospho-AS160 (Thr 642 ) were from Millipore (Billerica, MA). Adipocytes were lysed in Laemmli buffer (Sigma, MO) or in Triton x-100-containing lysis buffer for blotting Akt Thr 473 (Cell Signaling, MA). Lysates were resolved on SDS-PAGE, transferred to nitrocellulose and blotted with the indicated antibodies. Antibody binding was detected using enhanced Chemiluminescence (Super Signal West Pico, Thermo Scientific). C pAkt 473 Insulin pAKT473/tAKT Akti 1/2 Wortmannin pAS D pAS160/tAS160 Insulin Akti 1/2 Wortmannin B pAKT308/tAKT pAkt 308 Insulin Akti 1/2 Wortmannin A Akt AS160 pAkt 308 pAS pAkt 473 Akt Insulin Akti 1/2 Wortmannin

Supplemental figure 3. Effect of insulin-resistance on phosphorylation of Akt and AS160. (A) Western blots for the phosphorylation of Akt at Ser 473 and AS160 at Thr 642. Con, control cells; Res, insulin-resistant cells. 100 nM wortmannin. (B) Quantification of Akt phosphorylation. Shown are averages ± SEM from 4 experiments of phospho-Akt/total Akt. (C) Quantification of AS160 phosphorylation. Shown are averages ± SEM from 4 experiments of phospho-AS160/total AS160. Immunoblotting and antibodies Rabbit anti-actin was purchased from Cytoskeleton (Denver, CO). Antibodies against Akt and phospho-Akt (Ser 473 ) were purchase from Cell Signaling Technologies (Beverly, MA). Rabbit anti-AS160 and phospho-AS160 (Thr 642 ) were from Millipore (Billerica, MA). Adipocytes were lysed in Laemmli buffer (Sigma, MO). Cell lysates were resolved on SDS-PAGE, transferred to nitrocellulose and blotted with the indicated antibodies. Antibody binding was detected using enhanced Chemiluminescence (Super Signal West Pico, Thermo Scientific). Suppl. Fig. 3 pAkt S Control Insulin Resistant pAkt/Akt Insulin (nM) pAS160 T pAS160/AS160 Insulin (nM) Wortmannin ABC