Adipose tissue microRNAs as regulators of CCL2 production in human obesity

Abstract

In obesity, white adipose tissue (WAT) inflammation is linked to insulin resistance. Increased adipocyte chemokine (C-C motif) ligand 2 (CCL2) secretion may initiate adipose inflammation by attracting the migration of inflammatory cells into the tissue. Using an unbiased approach, we identified adipose microRNAs (miRNAs) that are dysregulated in human obesity and assessed their possible role in controlling CCL2 production. In subcutaneous WAT obtained from 56 subjects, 11 miRNAs were present in all subjects and downregulated in obesity. Of these, 10 affected adipocyte CCL2 secretion in vitro and for 2 miRNAs (miR-126 and miR-193b), regulatory circuits were defined. While miR-126 bound directly to the 3'-untranslated region of CCL2 mRNA, miR-193b regulated CCL2 production indirectly through a network of transcription factors, many of which have been identified in other inflammatory conditions. In addition, overexpression of miR-193b and miR-126 in a human monocyte/macrophage cell line attenuated CCL2 production. The levels of the two miRNAs in subcutaneous WAT were significantly associated with CCL2 secretion (miR-193b) and expression of integrin, α-X, an inflammatory macrophage marker (miR-193b and miR-126). Taken together, our data suggest that miRNAs may be important regulators of adipose inflammation through their effects on CCL2 release from human adipocytes and macrophages.

FIG. 1.

Regulation of CCL2 by miRNAs. A: Each candidate miRNA was overexpressed in in vitro differentiated adipocytes, and secreted levels of CCL2 were assessed. B: In 3T3-L1 cells, a subset of the candidate miRNAs were individually transfected together with reporter constructs, and alterations of luciferase activity were monitored. C: CCL2 mRNA levels were determined after miR-126 overexpression in human adipocytes differentiated in vitro. Protein secretion, luciferase activity, and mRNA levels were measured 24/48 h after transfection. To rule out unspecific effects, control cells were transfected with a negative control miRNA. Results were analyzed using t test and are presented as fold change ± SEM relative to the negative control. ***P < 0.001, **P < 0.01, *P < 0.05.

FIG. 2.

Characterization of TF motif activities in obesity. A: Principal component analysis of TF motif activities in obese (○, n = 30) and nonobese (●, n = 26) subjects. B: Regression analysis comparing the second principal component with BMI (n = 56; r = 0.730; P < 0.0001).

FIG. 3.

Regulators of CCL2 in human adipocytes and obesity. A: The proposed CCL2 network depicting miRNAs (squares) as well as motifs and cognate TFs (circles) altered by obesity and potentially targeting CCL2 (diamond). T-bars represent inhibition; arrows represent stimulation. B–D: miR-193b was overexpressed (white bar) or inhibited (black bar) in in vitro differentiated adipocytes, and mRNA levels of direct (MAX and ETS1) and indirect targets (RELB, STAT6, and NFKB1) (B) and CCL2 (C and D) were determined. To rule out unspecific effects, control cells were transfected with a negative control miRNA (striped bar). mRNA levels were measured 48 h after transfection. Results were analyzed using t test and are presented as fold change ± SEM relative to the negative control. ***P < 0.001, **P < 0.01, *P < 0.05. SP1, Sp1 transcription factor; NFATC3, nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 3; BHLH, basic helix-loop-helix; MYC, v-myc myelocytomatosis viral oncogene homolog; TFDP1, transcription factor Dp-1. (A high-quality color representation of this figure is available in the online issue.)

FIG. 4.

Validated regulatory circuits controlling CCL2 levels. A: The validated transcriptional regulatory network in fat cells included two miRNAs (miR-126 and miR193b; squares) and five TFs (MAX, ETS1, RELB, NFKB1, and STAT6; circles) all directly or indirectly regulating the production of CCL2 (diamond) in human adipocytes. Bold edges represent physical interactions between nodes described previously or demonstrated in this study, and thin lines (ETS1-STAT6-CCL2) are predicted by network analyses. B–D: The effect of overexpression of miR-126 and miR-193 on IL-6 secretion from adipocytes (B) and CCL2 expression and secretion in human macrophages (C and D) were examined. Results were analyzed using t test and are presented as fold change ± SEM relative to the negative control. ***P < 0.001, **P < 0.01, *P < 0.05. (A high-quality color representation of this figure is available in the online issue.)