Figure 1
Experimental design and metabolic traits measured to study the contribution of dendritic cells to adipose tissue metabolism. (A) Csf2+/+ and GM-CSF deficient (Csf2−/−) male mice at 8 weeks of age were placed on low-fat (LF), high-fat (HF), or high-fat + cholesterol (HFC) diets for 8 weeks. Metabolic profile was characterized through glucose, insulin, cholesterol, triglyceride, and phospholipid measurements. Cellular composition was determined through flow cytometry on stromal vascular fractions. The whole epididymal adipose tissue proteome was analyzed for 4 mice from each experimental group through liquid-chromatography mass spectrometry using isobaric peptide labels and the key driver gene network was built from the differentially expressed protein. The following comparisons were performed for all measurements: (a) Csf2+/+ LF vs. Csf2−/− LF, (b) Csf2+/+HF vs. Csf2−/− HF, (c) Csf2+/+HFC vs. Csf2−/−HFC, (d) Csf2+/+LF vs. HF, (e) Csf2−/−LF vs. HF, (f) Csf2+/+LF vs. HFC, (g) Csf2−/−LF vs. HFC, (h) Csf2+/+HF vs. HFC, (i) Csf2−/−HF vs. HFC. (B) Csf2−/− mice have decreased body weight compared to Csf2+/+mice on each diet at time of sacrifice. Comparisons a, b, c, d, e, f, and g, p < 0.001. (C) Adiposity, (epididymal + retroperitoneal adipose depots divided by body weight; details in Supplemental Fig. S1) was comparable across genotypes. For body weight and adiposity measurements; Csf2+/+ mice (low-fat (LF) n = 18, high-fat (HF) n = 16, high-fat + cholesterol (HFC) n = 22), Csf2−/− mice (LF n = 12, HF n = 18, HFC n = 13). Comparisons d, e, f, and g, p < 0.001. Cd11B+ cells (D) and Cd11B+CD11C+MHCII+ cells (E) were quantified by flow cytometry of the stromal vascular fraction from Csf2+/+ and Csf2−/− mice on LF, HF, and HFC diets. For flow measurements; mice were chosen around the median of the weight distribution for their group, Csf2+/+ mice (LF n = 3, HF n = 4, HFC n = 5), Csf2−/− mice (LF n = 4, HF n = 3, HFC n = 4). Glucose (F) and insulin tolerance (G) tests were performed after a 4 h fast. For glucose tolerance tests, Csf2+/+ mice (LF n = 9, HF n = 13, HFC n = 14) and Csf2−/− mice (LF n = 9, HF n = 9, HFC n = 8) were measured at 0, 15, 30, 60, and 120 minutes. For insulin tolerance tests, Csf2+/+ mice (LF n = 8, HF n = 13, HFC n = 14), and Csf2−/− mice (LF n = 8, HF n = 11, HFC n = 13) were measured at 0, 15, 30, 60, and 120 minutes. Area under the curve (AUC) was calculated by normalizing to time = 0 glucose levels (inset graphs). Comparisons a, b, c, d, e, and f, p < 0.05 for both glucose tolerance and insulin tolerance tests. (H) Plasma insulin levels are significantly increased in Csf2+/+ mice on high-fat (HF) and high-fat + cholesterol (HFC) diets as measured by ELISA at the time of sacrifice after a 4 h fast. For Csf2+/+ mice, the number of animals were low-fat (LF) n = 12, HF n = 13, and HFC n = 12; for Csf2−/− mice, the numbers of animals were LF n = 12, HF n = 14, and HFC n = 11. Comparisons d & f, p < 0.05. Data is presented as mean±SD for all except F and G (mean±SE). Significance was determined by ANOVA followed by Tukey’s posthoc analysis for multiple comparisons.
