Role for Retinoic Acid-Related Orphan Receptor Alpha (RORα) Expressing Macrophages in Diet-Induced Obesity

Abstract

The transcription factor RORα plays an important role in regulating circadian rhythm, inflammation, metabolism, and cellular development. Herein we show a role for RORα-expressing macrophages in the adipose tissue in altering the metabolic state of mice on a high-fat diet. The expression of Rora and RORA is elevated in white adipose tissue from obese mice and humans when compared to lean counterparts. When fed a high-fat diet Rora reporter mice revealed increased expression of Rora-YFP in macrophages in white adipose tissue deposits. To further define the potential role for Rora-expressing macrophages in the generation of an aberrant metabolic state Rora^fl/flLysM^Cre/+ mice, which do not express Rora in myeloid cells, were maintained on a high-fat diet, and metabolic parameters assessed. These mice had significantly impaired weight gain and improved metabolic parameters in comparison to Rora^fl/fl control mice. Further analysis of the immune cell populations within white adipose tissue deposits demonstrates a decrease in inflammatory adipose tissue macrophages (ATM). In obese reporter mouse there was increased in Rora-YFP expressing ATM in adipose tissue. Analysis of peritoneal macrophage populations demonstrates that within the peritoneal cavity Rora-expression is limited to myeloid-derived macrophages, suggesting a novel role for RORα in macrophage development and activation, which can impact on metabolism, and inflammation.

Keywords: RORα; inflammation; macrophage; metabolism; obesity.

Figure 1

Rora expression is increased in myeloid cells in adipose tissue isolated from obese subjects. Expression of the gene encoding RORα was determined in mice (Rora) and human (RORA) (A,B, respectively). RNA was isolated from the E-WAT isolated from both lean and obese humans and mice, gene expression was quantified by qPCR and normalized based on 18S expression (n = 3 lean mice, n = 7 obese mice; n = 6 lean or obese patients). Rora^Cre and Rora^CreRosa-YFP mice were fed a high-fat diet for 12 weeks and flow cytometry performed on epididymal and inguinal white adipose tissue (E-WAT and I-WAT, respectively). Cells were gated as Live/dead⁻CD45⁺ and assessed for Rora-YFP expression (C, data is representative of 3 Rora^Cre and 7 Rora^CreRosa-YFP mice). (D) Rora^fl/fl and Rora^fl/flLysM^Cre/+ mice were fed a high-fat diet for 16 weeks and RNA isolated from E-WAT and I-WAT; expression of Rora was assessed relative to 18S (n = 6). Data is representative of mean ± SEM. Student's t-test: *P < 0.05, ***P < 0.001.

Figure 2

Rora^fl/flLysM^Cre/+ mice show decreased weight gain and improved metabolic function associated with increased expression of genes associated with thermogenesis. Groups of Rora^fl/fl and Rora^fl/flLysM^Cre/+ mice were fed a high-fat diet (HFD; 60% fat) for 16 weeks and weight monitored weekly, percentage weight gain was calculated from the starting weight of each animal and starting and final weight is quantified for each group (A, n = 11 Rora^fl/fl; n = 19 Rora^fl/flLysM^Cre/+). E-WAT, I-WAT and S-BAT (subcutaneous brown adipose tissue) weight was determined after 16 weeks on HFD (B, n = 7 Rora^fl/fl; n = 6 Rora^fl/flLysM^Cre/+). Glucose and insulin tolerance tests (GTT and ITT, respectively) were performed on mice after 14 and 15 weeks on HFD respectively (C, n = 8 Rora^fl/fl; n = 6 Rora^fl/flLysM^Cre/+). The serum AST/ALT ratio was calculated from AST and ALT activity tests on the serum from Rora^fl/fl and Rora^fl/flLysM^Cre/+ mice after 16 weeks on HFD (D, n = 8 Rora^fl/fl; n = 6 Rora^fl/flLysM^Cre/+). E-WAT, I-WAT, and S-BAT were excised after 16 weeks on HFD (n = 6 Rora^fl/fl; n = 6 Rora^fl/flLysM^Cre/+). RNA was isolated and expression of genes associated with the metabolic and thermogenic function of the adipose tissue were assessed (E). All data is representative of mean ± SEM. Student's t test: *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 3

Rora^fl/flLysM^Cre/+ mice show decreased Ly6C^hi ATM and fewer CD9⁺ inflammatory ATM in obese animals. Groups of Rora^fl/fl and Rora^fl/flLysM^Cre/+ mice were fed a control diet (CD; n = 5 Rora^fl/fl, n = 3 Rora^fl/flLysM^Cre/+) or a high-fat diet (HFD; n = 7 Rora^fl/fl, n = 6 Rora^fl/flLysM^Cre/+) for 16 weeks and the E-WAT and I-WAT isolated and prepared for flow cytometry. Cells isolated from the E-WAT were gated as Live-dead⁻CD45⁺SiglecF⁻ and assessed for Ly6C expression (A). Ly6Clo cells were further assessed for CD64 and CD9 expression (B). This gating strategy was repeated in cells isolated from I-WAT (C,D, n = 2 Rora^fl/fl CD, n = 1 Rora^fl/flLysM^Cre/+ CD; n = 7 Rora^fl/fl HFD, n = 6 Rora^fl/flLysM^Cre/+ HFD). Rora^Cre and Rora^CreRosa-YFP mice were fed a high-fat diet for 12 weeks and flow cytometry performed on E-WAT. Cells were gated as Live-dead⁻CD45⁺SiglecF⁻CD11b⁺ then gated as described previously based on Ly6C expression, then CD9 expression, as indicated, and assessed for Rora-YFP expression (E, data is representative of 3 Rora^Cre and 3 Rora^CreRosa-YFP mice). (F) Cells isolated from the E-WAT from HFD fed Rora^fl/fl and Rora^fl/flLysM^Cre/+ mice were stained as CAM (CD11b⁺SiglecF⁻F4/80⁺CD206^lo) or AAM (CD11b⁺SiglecF⁻F4/80^hiCD206^hi; n = 6–7). Data is representative of mean ± SEM. Student's t-test: ns, not significant, *P < 0.05, **P < 0.01.

Figure 4

Ly6C^hi monocytes are decreased in the blood in the absence of myeloid cell expression of Rora. Blood was collected from age-matched C57Bl6/J WT and C57Bl6/J Rora^sg/sg mice by submandibular bleed. (A) Ly6C expression was quantified on blood CD45⁺CD115⁺ monocytes (n = 8 WT; n = 11 Rora^sg/sg). (B) Rora-YFP expression was quantified on Ly6C^lo and Ly6C^hi blood CD45⁺CD115⁺ monocytes (data is representative of 3 Rora^CreRosa-YFP mice). (C) Ly6C expression was quantified on blood monocytes isolated from Rora^fl/fl and Rora^fl/flLysM^Cre/+ mice (n = 3 Rora^fl/fl; n = 4 Rora^fl/flLysM^Cre/+). Data is representative of mean ± SEM. Student's t-test: ns, not significant, *P < 0.05.

Figure 5

RORα acts specifically on myeloid-derived macrophages, with limited effect on tissue resident cells. Bone-marrow derived macrophages (BMDM) were isolated from age-matched WT and Rora^sg/sg mice and expression of Rora determined by qRT-PCR relative to 18S (A, n = 3 WT, n = 5 Rora^sg/sg). BMDM isolated from Rora^CreRosa-YFP mice were stimulated with media-alone, LPS (100 ng/ml) or IL-4 (20 ng/ml) for 48 h and YFP^+ve cells determined by flow cytometry (B, n = 3). Peritoneal exudate cells were collected from age-matched WT and Rora^sg/sg mice by peritoneal lavage and stained for flow cytometry as small peritoneal macrophages (SPM; Live/dead⁻CD45⁺CD11b⁺SiglecF⁻F4/80⁺MHC class II^hi) or large peritoneal macrophages (LPM; Live/dead⁻CD45⁺CD11b⁺SiglecF⁻F4/80^hiMHC class II^lo) (C; n = 10 WT; n = 5 Rora^sg/sg). Rora^CreRosa-YFP mice were injected i.p. with PBS (n = 6) or IL-4c (n = 4) on days 0 and 2 (D,E), or thioglycollate (day 0; n = 4) with or without IL-4c (days 0 and 2; n = 4) (F). Increase in CD206⁺ macrophages in the peritoneal cavity of mice following IL-4c treatment (D, gated as Live/dead⁻CD45⁺CD11b⁺SiglecF⁻F4/80⁺CD206⁺). Flow cytometry was performed on LPM (gating as above; E) or SPM (gating as above; F) assessing Rora-YFP expression on CD206^hi and CD206^lo cells. Data is representative of mean ± SEM. Student's t-test or ANOVA (E) was used for statistical analysis: ns, not significant, *P < 0.05, **P < 0.01, ***P < 0.001.