Figure 2.

Insulin sensitivity and adipocyte size

(A) Transmission confocal micrograph of 3T3-L1 adipocytes with software segmented boundaries (overlaid colored lines) and measured diameters shown.

(B) Relative FRET signal, normalized to starting value, as a function of time of serum starved adipocytes treated with 1μM insulin (red lines) or vehicle (blue lines) as control for basal condition. Thin lines show individual cells and thick show the population mean (N = 140 cells from five independent replicates).

(C) Distribution of relative FRET change over 30 min for each adipocyte in B generated as a kernel density estimate.

(D) Relative FRET change over 30 min for each insulin stimulated adipocyte in B as a function of cell diameter. Dotted lines show free linear fits, with correlation coefficients in insert.

(E) The mean total relative FRET change comparing basal and 30 min insulin treatment of adipocytes with diameters below or above the mean cell diameter of ∼30 μm, normalized to small basal cells. Error bars showing SE and ∗ indicating significant differences to p < 0.05 using Student's t-test.

(F) Distribution of relative FRET change over 30 min for each insulin stimulated adipocyte in B generated as a kernel density estimate, separated by adipocyte diameter. ∗ indicating significant differences to p < 0.05 using Kolmogorov-Smirnov test. Compared to the initial FRET ratios (Figure S5B), only after insulin stimulation is there a difference between small and large adipocytes suggesting there is no bias between the two groups based on expression levels.