IKKβ links vascular inflammation to obesity and atherosclerosis

Abstract

IκB kinase β (IKKβ), a central coordinator of inflammatory responses through activation of NF-κB, has been implicated in vascular pathologies, but its role in atherogenesis remains elusive. Here, we demonstrate that IKKβ functions in smooth muscle cells (SMCs) to regulate vascular inflammatory responses and atherosclerosis development. IKKβ deficiency in SMCs driven by a SM22Cre-IKKβ-flox system rendered low density lipoprotein receptor-null mice resistant to vascular inflammation and atherosclerosis induced by high-fat feeding. Unexpectedly, IKKβ-deficient mice were also resistant to diet-induced obesity and metabolic disorders. Cell lineage analysis revealed that SM22Cre is active in primary adipose stromal vascular cells and deficiency of IKKβ diminished the ability of these cells to differentiate, leading to accumulation of adipocyte precursor cells in adipose tissue. Mechanistically, reduction of IKKβ expression or pharmacological inhibition of IKKβ inhibited proteasome-mediated β-catenin ubiquitination and degradation in murine preadipocytes, resulting in elevated β-catenin levels and impaired adipocyte differentiation. Further, chronic treatment of mice with a potent IKKβ inhibitor decreased adipogenesis and ameliorated diet-induced obesity. Our findings demonstrate a pivotal role of IKKβ in linking vascular inflammation to atherosclerosis and adipose tissue development, and provide evidence for using appropriate IKKβ inhibitors in the treatment of obesity and metabolic disorders.

Figure 1.

Generation of LDLR^−/− mice with SMC-specific IKKβ deficiency. (A) Western blot analysis of IKKβ and IKKα expression in aorta, liver, skeletal muscle (s. muscle), and intestine of IKKβ^F/FLDLR^−/− (F/F) and SM22Cre⁺IKKβ^F/FLDLR^−/− (KO) mice. (B) SMCs isolated from aortas of IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice were stimulated with TNF (20 ng/ml) or vehicle for 30 min. Cells stained with anti-p65 primary antibodies, followed by fluorescein-labeled secondary antibodies (green). The nuclei were visualized with DAPI (blue). Bars, 20 µm. (C) SMCs were stimulated with TNF (20 ng/ml) or vehicle for 30 min. Nuclear proteins were extracted and NF-κB binding activity was determined by electrophoretic mobility shift assay. (D) SMCs were treated with LPS (5 µg/ml) or vehicle control for 3 h. Expression of proinflammatory cytokines was analyzed by QPCR (n = 5). Similar results were obtained from at least three independent experiments. All data are means ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 2.

Deficiency of SMC IKKβ inhibits diet-induced vascular inflammation and atherosclerosis development in LDLR^−/− mice. (A and B) Representative hematoxylin and eosin sections in aorta (A) and BCA (B) of 16-wk-old male IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice fed a WD for 12 wk. Bottom panels in A represent the magnification of the boxed areas in aorta shown in top panels. (C and D) Quantitative analysis of the lesion area in the aortic root (C) and BCA (D) of WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice (n = 11–23 mice). Representative oil red O–stained sections were displayed below the quantification data. (E) Expression of proinflammatory cytokines in aortas of mice fed chow or WD for 12 wk was analyzed by QPCR (n = 5 mice). (F) Sections of aortic root atherosclerotic lesions were stained with antibodies against mouse TNF, MCP-1, or IL-1β, followed by fluorescein-labeled secondary antibodies (red). The nuclei were stained with DAPI (blue). A representative figure from three mice per group and the similar result is shown. All data are means ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bars: (A, top) 500 µm; (A, bottom) 50 µm; (B) 200 µm; (C) 500 µm; (D) 100 µm; (F) 100 µm.

Figure 3.

IKKβ-deficient mice are resistant to diet-induced obesity. (A–D) Body weight (A), lean mass (B), fat mass (C), and fat percentage (D) of 16-wk-old male IKKβ^F/FLDLR^−/− (F/F) and SM22Cre⁺IKKβ^F/FLDLR^−/− (KO) mice fed a normal chow diet (Chow) or WD for 12 wk (n = 15–20 mice). (E–G) Food intake (E), oxygen consumption (F), and mean oxygen consumption (G) were monitored in WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice (n = 5 mice). Oxygen consumption data were normalized by lean body mass and mean from 4-d measurements. (H) WAT (epididymal, subcutaneous, perirenal, and omental WAT) and BAT weight measured in WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice (n = 7–8 mice). (I and J) The expression of indicated genes in BAT (I) and WAT (J) was analyzed by QPCR (n = 4–8 mice). (K) Plasma cytokine levels in chow or WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice (n = 5–7 mice). Results are representative of three independent experiments. All data are means ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 4.

Deficiency of IKKβ protects mice from obesity-associated metabolic disorders. (A–C) Fasting plasma glucose and insulin levels (n = 10–15 mice; A), GTT (B), and the area under the curve (AUC) of GTT (C) in 16-wk-old male IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice fed a WD for 12 wk (n = 10–15 mice). (D) Macrophage infiltration in WAT determined by F4/80 staining. (E and F) Representative appearance (E) and hematoxylin and eosin (top) and oil red O–stained (bottom) sections (F) of livers. (G–I) Hepatic cholesterol and triglyceride levels (G), plasma cholesterol and triglyceride levels (H), and plasma cholesterol distribution (I) of WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice (n = 6–10 mice). (J) Hepatic gene expression was analyzed by QPCR (n = 4–8 mice). Results are representative of three independent experiments. All data are means ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Bars: (D) 100 µm; (F) 50 µm (top) and 200 µm (bottom).

Figure 5.

Deficiency of IKKβ reduces atherosclerosis in LDLR^−/− mice in the absence of obesity and severe hyperlipidemia. (A–F) Body weight (A), lean mass (B), fat mass (C), plasma cholesterol levels (D), triglyceride (E), and cholesterol distribution (F) of 16-wk-old male IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice fed a low-fat AIN76 diet for 12 wk (n = 6–10 mice). (G and H) Quantitative analysis of the lesion area in the aortic root (G) and BCA (H). n = 8–14 mice. (I) Sections of aortic root atherosclerotic lesions were stained with antibodies against mouse TNF, MCP-1, or IL-1β, followed by fluorescein-labeled secondary antibodies (red). The nuclei were stained with DAPI (blue). A representative figure from three mice per group and the similar result is shown. (J) Plasma cytokine levels of AIN76 diet-fed mice (n = 6 mice). Results are representative of three independent experiments. All data are mean ± SD. *, P < 0.05; ***, P < 0.001. Bars, 100 µm.

Figure 6.

SM22Cre-mediated IKKβ ablation blocks white adipose differentiation. (A) Hematoxylin and eosin staining of transverse sections of subcutaneous WAT from WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice. The adipocyte numbers were calculated based on per mm epidermal length (n = 5 mice). (B) Schematic of the SM22Cre⁺Rosa26^EGFP mouse model. (C) Quantitation of GFP⁺ cells in cultures of adipose SV cells isolated from Rosa26^EGFP and SM22Cre⁺Rosa26^EGFP mice by flow cytometry. The percentage of GFP⁺ cells are as indicated in the flow profiles. (D) PCR analysis of genomic DNA from adipose SV cells of IKKβ^F/FLDLR^−/− (F/F) and SM22Cre⁺IKKβ^F/FLDLR^−/− (KO) mice (top). Expression levels of IKKβ, inflammatory genes, and leptin in adipose SV cells were analyzed by QPCR (n = 3 mice) (bottom). (E) Oil red O staining of adipose SV cells of IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice induced by differentiation medium. The nuclei were stained with hematoxylin (blue) in the bottom panel. (F) SV cells and mature adipocytes were isolated from WAT of WD-fed IKKβ^F/FLDLR^−/− and SM22Cre⁺IKKβ^F/FLDLR^−/− mice. The cell population was displayed as cell number per gram of WAT (n = 5 mice). (G) Expression of adipogenic genes and adipocyte precursor cell markers in adipose SV cells was measured by QPCR (n = 5–7 mice). (H) Sections of subcutaneous WAT were stained with antibodies against adipocyte progenitor markers PDGFRβ and preadipocyte marker Pref-1, followed by Alexa Fluor 488 (green)– or Alexa Fluor 594 (red)–labeled secondary antibodies. A representative figure from three mice per group and the similar result is shown. (I) Adipose SV cells were examined for expression of adipocyte progenitor cell makers, PDGFRβ, and NG2, and endothelial cell marker CD31 with flow cytometry. CD31⁺, NG2⁺, and CD31⁻NG2⁻ cells are highlighted with red, green, and black, respectively. The percentages of CD31⁻PDGFRβ⁺NG2⁺ cells are as indicated in the flow profiles (P < 0.001, n = 3–4 mice). Similar results were obtained from at least three independent experiments. All data are means ± SD. *, P < 0.05; **, P < 0.01. Bars: (A) 200 µm; (E, bottom) 50 µm; (H) 100 µm.

Figure 7.

IKKβ regulates β-catenin ubiquitination and adipocyte differentiation. (A) Expression levels of preadipocyte, mural cell, and mature adipocyte markers in 3T3-L1 cells at 0 and 48 h after addition of differentiation media (n = 3). (B) Western blot analysis of IKKβ and IKKα levels in 3T3-L1 preadipocytes expressing control shRNA or shRNA against IKKβ (shIKKβ). (C) Oil red O staining of control or shIKKβ 3T3-L1 cells induced by differentiation medium. (D) Western blot analysis of nuclear β-catenin protein levels of control or shIKKβ 3T3-L1 cells. Nuclear proteins were also probed with anti-Histone H3 antibodies as an internal control. (E) QPCR analysis of expression of IKKβ and adipogenic genes in control or shIKKβ 3T3-L1 cells (n = 3–5). (F) Control or shIKKβ 3T3-L1 cells were treated with vehicle control or 100 nM PS-341 as indicated for 4 h. β-catenin was immunoprecipitated with anti–β-catenin antibodies, and then probed with antiubiquitin monoclonal antibodies. The whole cell lysates were probed with anti–β-catenin antibodies as an internal control. (G) QPCR analysis of Smurf2 expression in control or shIKKβ 3T3-L1 cells (n = 4). (H) Western blot analysis of Smurf2 protein levels in control or shIKKβ 3T3-L1 cells. (I) Oil red O staining of 3T3-L1 cells transfected with control vector or vector expressing IκBαM induced by differentiation medium. (J) Western blot analysis of Smurf2 protein levels in 3T3-L1 cells expressing control or IκBαM vectors. (K) Control or IκBαM-expressing 3T3-L1 cells were treated with vehicle control or 100 nM PS-341 as indicated for 4 h. β-catenin was immunoprecipitated with anti–β-catenin antibodies and then probed with anti-ubiquitin monoclonal antibodies. The whole-cell lysates were probed with anti–β-catenin antibodies as an internal control. (L) Western blot analysis of nuclear β-catenin levels in control or IκBαM-expressing 3T3-L1 cells. Nuclear proteins were probed with anti-Histone H3 antibodies as an internal control. (M) Western blot analysis of Smurf2 levels in adipose SV cells isolated from WD-fed IKKβ^F/FLDLR^−/− (F/F) and SM22Cre⁺IKKβ^F/FLDLR^−/− (KO) mice. (N) Adipose SV cells were treated with vehicle control or 100 nM PS-341, as indicated, for 4 h. β-catenin was immunoprecipitated with anti–β-catenin antibodies and then probed with anti-ubiquitin monoclonal antibodies. The whole cell lysates were probed with anti-β-catenin antibodies as an internal control. (O) Western blot analysis of nuclear β-catenin levels in SV cells. Nuclear proteins were probed with anti-Histone H3 antibodies as internal control. Similar results were obtained from at least three independent experiments. All data are means ± SD. **, P < 0.01; ***, P < 0.001. Bar: (C, bottom) 100 µm; (I, bottom) 100 µm.

Figure 8.

Pharmacological inhibition of IKKβ inhibits adipocyte differentiation. (A) Oil red O staining of 3T3-L1 cells induced by differentiation medium or medium containing IKKβ inhibitor BMS-345541 at indicated concentrations. (B) Analysis of IKKβ, Smurf2, and adipogenic genes in 3T3-L1 cells treated with control or 10 µM BMS-345541 for 48 h by QPCR (n = 3–5). (C) 3T3-L1 cells were treated with vehicle control or 10 µM BMS-345541 for 48 h before incubating with vehicle control or 100 nM PS-341 for 4 h. β-catenin was immunoprecipitated with anti–β-catenin antibodies, and then probed with antiubiquitin monoclonal antibodies. The whole-cell lysates were probed with anti–β-catenin antibodies as an internal control. (D) Western blot analysis of nuclear β-catenin levels in 3T3-L1 cells treated with vehicle control or 10 µM BMS-345541 for 48 h. Nuclear proteins were probed with anti-Histone H3 antibodies as an internal control. (E) Oil red O staining of 3T3-L1 cells induced by differentiation medium or medium containing IKKβ inhibitor sodium salicylate at indicated concentrations. Similar results were obtained from at least three independent experiments. All data are mean ± SD. **, P < 0.01; ***, P < 0.001. Bars: (A, bottom) 100 µm; (E, bottom) 100 µm.

Figure 9.

Chronic treatment of mice with IKKβ inhibitor ameliorates diet-induced obesity. (A and B) 8-wk-old male C57BL/6 mice were fed a WD and treated with vehicle or 10 mg/kg body weight of BMS-345541 by daily oral gavage for 8 wk. Body weight (A) was measured weekly and fat and lean mass (B) were measured at the end of feeding study (n = 11–12 mice). (C) Representative photographs of subcutaneous (Sub) and epididymal (Epi) WAT. (D) Expression of Smurf2 and other NF-κB target genes in WAT were analyzed by QPCR (n = 4 mice). (E) Western blot analysis of nuclear β-catenin levels in WAT of control or BMS-345541-treated mice. Nuclear proteins were probed with anti-Histone H3 antibodies an internal control. (F) Oil red O staining of adipose SV cells from control or BMS-345541–treated mice induced by differentiation medium. (G) Adipose SV cells isolated from control or BMS-345541-treated mice were incubated with vehicle or 100 nM PS-341 as indicated for 4 h. β-catenin was immunoprecipitated with anti–β-catenin antibodies and then probed with antiubiquitin monoclonal antibodies. The whole cell lysates were probed with anti–β-catenin antibodies as an internal control. Similar results were obtained from at least three independent experiments. All data are mean ± SD. *, P < 0.05; **, P < 0.01. Bars: (F, bottom) 100 µm.