Insulin controls the spatial distribution of GLUT4 on the cell surface through regulation of its postfusion dispersal.

Reagents Mouse anti-HA antibody

(HA.11) was from Berkeley Antibody Co. (Richmond, CA). Clathrin-GFP, Caveolin-GFP and Tubulin-GFP were kindly provided by Dr. J. Lippincott-Schwartz. Construction of the HA-GLUT4-mCherry has been described previously ( Lizunov et al., 2009 ). Bovine serum albumin was from Intergen (fraction V). DMEM, Insulin, Alexa-conjugated antibodies were all from Invitrogen.

Cell Culture and Transfection

Preparation of isolated rat adipose cells from male rats

(CD strain, Charles River Laboratories, MD), electroporation of rat adipose cells, and the cell-surface antibody-binding assay were performed as described previously ( Al-Hasani et al., 1998 ; Lizunov et al., 2005 ). All plasmids were used at a final concentration of 4 μg/ml. Transfected cells were kept in culture overnight and optimal protein expression level was achieved at 20-24 h after electroporation. Insulin stimulation was performed by addition of 70 nM insulin for 30 min at 37 °C.

Live Cell Imaging and Immunoflourescence Microscopy

For live cell imaging, isolated adipose cells were kept in KRBH buffer with 1% BSA, pH 7.4, maintained at 37°C using a temperature-controlled stage and Delta-T environmental chamber (Bioptechs). For immunoflourescence microscopy, the isolated cells in suspension were either fixed with 4% formaldehyde in phosphate-buffered saline (PBS) for 10 min, or incubated with 2 mM KCN to deplete ATP and inhibit GLUT4 recycling ( Satoh et al., 1993 ). The cells were then washed with PBS and transferred to KRBH with 1% BSA for incubation with antibodies in the presence of KCN. Cells were imaged using the TIRFM setup built around an Axiovert 200 microscope (Zeiss) equipped with a 100×1.45 NA objective. A TIRF slider (Till Photonics) was used to pass laser beams from an AOTF-controlled combiner system (LSM Technologies) equipped with 405/488/561 nm lasers (Coherent). Penetration depth of the evanescent field was measured to be 110 ± 20 nm by a calibration procedure with 40-nm fluorescent beads attached to the piezo-driven micropipette. Fluorescence was separated from the excitation light using a multi-band dichroic and emission filter set (GFP/DsRed-2X-A, Semrock), and passed to an electron-multiplying CCD camera (Ixon, Andor). For WF microscopy, a fiber-coupled light source (X-cite 120, EXFO) was used together with a filter-wheel (Lambda 10-B, Sutter) to sequentially excite GFP/pHluorin and mCherry/Alexa-594 using appropriate filters. Laser-switching AOTF, shutters, filter-wheels, microscope, and EMCCD camera were synchronized and controlled using μManager v1.2.38 ( http://www.micro-manager.org ).

Image analysis

A set of automated image processing macro/subroutines was developed based on existing algorithms of ImageJ (Rolling-Ball Background Subtraction, Gaussian and Granulometric filters, Z-project, Image5D, Find Maxima, and Particle Tracker). Individual diffraction-limited fluorescent structures were segmented within representative regions of interest (ROI). The following criteria were used to detect individual structures: i) fluorescence had a local maximum; ii) integrated pixel intensity was at least 2-fold above the standard deviation of pixel intensity; and iii) 70% of the peak intensity was contained within a circular ROI of five pixels. Density was measured as the number of structures per square micron; frequency of dynamic events of fission and fusion was calculated as the number of events per square micron per minute. Particle tracking was carried out using a custom-modified Particle Tracker algorithm ( Sbalzarini and Koumoutsakos, 2005 ) that utilized spatial moment analysis to detect redistribution of GLUT4 molecules during fusion. To quantify populations of stationary and mobile structures we utilized a time-projection method ( Lizunov et al., 2009 ).

Data acquired using multi-color

TIRF were processed to measure colocalization of individual structures. Simulated data were used to estimate a percent of random overlap for given densities of structures in each channel. All data are represented as means ± SEM. Statistical significance was analyzed using Student’s t-test or ANOVA. Further details of the data analysis and modeling are described in Supplemental Materials .

Supplementary Material 01 02 03 04 05 06