Extended Data Fig. 4: Glucose tolerance testing in mice in response to extracellular vesicles derived from various treatments, tissue insulin sensitivity in isolated tissue, and blood glucose and insulin levels during a hyperinsulinemic-euglycemic clamp.
(A-D) Gating strategy for imaging flow cytometry in NAFL mice (n = 3 fasted, n = 4 glucose gavage) that were fasted and subjected to an oral gavage of water (vehicle) or glucose. The unfocused bright field signals were excluded using Gradient RMS feature (A). Particles with relative area smaller than the speed beads were gated by the discriminating aspect ratio and area of particles (previously optimised) (B). CD63-PE-Cy7 and CD9-APC positive particles were gated (C). ASGPR positive particles were plotted and gated based on the isotype control (D). 1 independent experiment. (E) Glucose AUC during an oral GTT in Control (n = 6/group) and NAFL mice (n = 5/group) in response to sonicated liver EVs across 1 cohort. P < 0.001 (main effect). (F-G) Glucose AUC during an oral GTT in Control (F, n = 9 saline, n = 10 NAFL EV, n = 9 ‘shaved EV) or NAFL (G, n = 5 saline, n = 3 Control EV, n = 6 ‘shaved’ EV) mice that received saline or ‘shaved’ liver EVs across 2 cohorts (control) and 1 cohort (NAFL). P < 0.001. *, significant from Saline; #, significant from Control EV. (H) Representative cryo-electron microscopy of ‘shaved’ liver EV preparation. 1 experiment. (I) Nanoparticle tracking analysis of ‘shaved’ liver EV. 1 experiment, n = 4 biological replicates graphed as an average. (J-K) Blood glucose levels and AUC (inset) during an oral GTT (J), and serum insulin (K) of mice that received ICV of saline/liver EVs during 1 experiment (n = 9/group). P < 0.001 (main effect: time). (L-N) Glucose uptake in isolated soleus muscle (L, n = 3 saline, n = 4 Control EV, P = 0.61, basal and insulin-stimulated samples are paired from the same animal within each treatment), white adipose tissue (M, n = 7 saline/basal, n = 6 saline/insulin, n = 7 Control EV/basal, n = 6 Control EV/insulin, P = 0.73), and brown adipose tissue (N, n = 7/group, P = 0.932) collected from NAFL mice and treated ex vivo with saline or liver EVs (10 μg/ml) obtained from healthy Control mice 1 hr prior to and during the experiment. 1 cohort across tissues. Basal and insulin-stimulated samples from adipose are paired from the same animal across both treatments, with exception of 1 mouse for the EWAT comparison (that is, basal saline vs. basal EV). (O) Blood glucose levels at baseline (that is, fasted state) and during a hyperinsulinemic-euglycemic clamp in Control mice (n = 7/group) following intraperitoneal injection of NAFL EV 1 hr prior to the experiment. P < 0.001 (main effect). 1 independent cohort. (P) Serum insulin at baseline (that is, fasted state) and during a hyperinsulinemic-euglycemic clamp in Control mice (n = 6 saline, 9 NAFL EV) following intraperitoneal injection of NAFL EV 1 hr prior to the experiment. P < 0.001 (main effect). 1 independent cohort. Data are expressed as mean ± SEM. Each datapoint represents an independent biological sample. Data was analysed using a one-way ANOVA (F, G), two-way ANOVA (E, L, M, N) or repeated measures two-way ANOVA (J, K, O, P) with Student Newman Keuls post-hoc analyses, or unpaired two-tailed t-test (J (inset)). *, significant main effect compared to basal.
