The absence of thioredoxin-interacting protein in alveolar cells exacerbates asthma during obesity

Abstract

Obesity is associated with an increased incidence of asthma. However, the mechanisms underlying this association are not fully understood. In this study, we investigated the role of thioredoxin-interacting protein (TXNIP) in obesity-induced asthma. Asthma was induced by intranasal injection of a protease from Aspergillus oryzae in normal diet (ND)- or high fat diet (HFD)-fed mice to investigate the symptoms. We measured TXNIP expression in the lungs of patients with asthma and in ND or HFD asthmatic mice. To explore the role of TXNIP in asthma pathogenesis, we induced asthma in the same manner in alveolar type 2 cell-specific TXNIP deficient (TXNIP^Cre) mice. In addition, the expression levels of pro-inflammatory cytokines were compared based on TXNIP gene expression in A549 cells stimulated with recombinant human tumor necrosis factor alpha. Compared to ND-fed mice, HFD-fed mice had elevated levels of free fatty acids and adipokines, resulting in high reactive oxygen species levels and more severe asthma symptoms. TXNIP expression was increased in both, asthmatic patients and HFD asthmatic mice. However, in experiments using TXNIP^Cre mice, despite being TXNIP deficient, TXNIP^Cre mice exhibited exacerbated asthma symptoms. Consistent with this, in vitro studies showed highest expression levels of pro-inflammatory cytokines in TXNIP-silenced cells. Overall, our findings suggest that increased TXNIP levels in obesity-induced asthma is compensatory to protect against inflammatory responses.

Keywords: Asthma; High fat diet; Inflammasome; Obesity; Thioredoxin-interacting protein.

Graphical abstract

Fig. 1

Comparison of symptoms of normal diet (ND)-fed mice and high fat diet (HFD)-fed mice induced with asthma using a protease from Aspergillus oryzae (Asp). (A) The Penh value graph of airway hyperresponsiveness measurements. Inhaled methacholine in HFD mice caused more bronchoconstriction than in ND mice. Values: mean $\pm$ standard deviation (SD) (n = 7 per group). Significance: *p$<$ 0.05 vs ND Asthma by Welch's correction. (B) Representative images and quantification graph of Diff-Quick® staining of bronchoalveolar lavage fluid (BALF) cells (Scale bar = 60 μm). Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs ND Asthma by Welch's correction, respectively. (C) Representative population dot plot and (D) quantification graph of each inflammatory cell in BALF as analyzed by flow cytometry. PerCP-CD45⁺ cells, Lymphocyte; PerCP-CD45⁺FITC-LY6G⁺ cells, Neutrophils; PerCP-CD45⁺PE-CD11c⁺APC-SiglecF⁺ cells, Alveolar macrophages; PerCP-CD45⁺APC-SiglecF⁺ cells, Eosinophils. Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs ND Asthma, respectively. Statistical analysis: Neutrophils, by Welch's correction; Alveolar macrophages and Eosinophils, by two-tailed t-test. BALF cells staining and flow cytometry analysis revealed a significant increase in the number of eosinophils in HFD mice, which is associated with asthma development. (E) Representative images of hematoxylin/eosin (H&E; Scale bar = 100 μm) and periodic acid-Schiff (PAS; Scale bar = 60 μm) staining of lung tissue and quantification graphs for histological scores. Severe airway inflammation and mucus overproduction in lung tissue due to Asp was observed in HFD mice than ND mice. Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs ND Asthma by two-tailed t-test, respectively. (F) Cytokine levels of immunoglobulin (Ig)E, interleukin (IL)-4, IL-5, IL-13, and IL-17A as analyzed by enzyme-linked immunosorbent assay. Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs ND Asthma, respectively. Statistical analysis: IgE, by Mann-Whitney test; IL-4 and IL-5, by Welch's correction; IL-13 and IL-17A, by two-tailed t-test.

Fig. 2

Effects of high fat diet (HFD) feeding duration on the levels of free fatty acids (FFAs), adipokines and reactive oxygen species (ROS). (A) Body weight of normal diet (ND)-fed mice and HFD-fed mice verse time. Significant weight gain was observed with HFD exposure compared to ND. Values: mean $\pm$ standard deviation (SD). Significance: *,**p$<$ 0.05 and 0.01 vs ND-fed mice by Welch's correction, respectively. (B) FFA concentrations and (C) leptin/adiponectin ratio in the serum as measured by enzyme-linked immunosorbent assay. With exposure to HFD, FFA levels were highest at 13 wks, and adipokines ratio was increased with increasing exposure time. Values: mean $\pm$ SD (n = 5 per group). Significance: **p$<$ 0.01. Statistical analysis: FFA assay, by Tukey's post hoc test; Adipokines assay, by Kruskal-Wallis test. (D) Representative images of intracelluar ROS levels in bronchoalveolar lavage fluid as detected by fluorescence microscopy (Scale bar = 25 μm) and quantification graph. ROS levels gradually increased with HFD exposure, peaking at 26 wks. Values: mean $\pm$ SD (n = 5 per group). Significance: *,**p$<$ 0.05 and 0.01 by Tukey's post hoc test, respectively.

Fig. 3

Thioredoxin-interacting protein (TXNIP) expression in asthma pathogenesis. (A) Immunohistochemical staining for TXNIP in lung sections from human healthy control and asthma patient (Scale bar = 80 μm) and quantification graph of TXNIP positive area. TXNIP expression is increased in asthma patients, confirming its involvement in the pathogenesis of asthma. Values: mean $\pm$ standard deviation (SD) (n = 5 field per group). Significance: **p$<$ 0.01 vs Control by two-tailed t-test. (B) Immunohistochemical staining (Scale bar = 30 μm) and quantification graph of TXNIP positive area and (C) immunofluorescence staining (Scale bar = 30 μm) for TXNIP in lung sections from normal diet (ND)-fed mice and high fat diet (HFD)-fed mice induced with asthma using a protease from Aspergillus oryzae. Lung tissues were stained with TXNIP (green) and DAPI (blue). TXNIP expression in lung tissue was observed to be higher in HFD asthmatic mice than ND asthmatic mice. Values: mean $\pm$ SD (n = 7 per group). Significance: *p$<$ 0.05 vs ND Asthma by two-tailed t-test. (D), (E) Western blot analysis of the indicated proteins in the lung tissue and graphs representing the densitometric values of protein expression. NLRP3, NOD-like receptor family pyrin domain containing 3; ASC, apoptosis-associated speck-like protein containing a caspase-recruitment domain; IL-1β, interleukin-1β. Values: mean $\pm$ SD (n = 4 per group). Significance: *,**p$<$ 0.05 and 0.01 vs ND Asthma by two-tailed t-test, respectively.

Fig. 4

Comparison of symptoms of alveolar type 2 cell-specific thioredoxin-interacting protein (TXNIP) deficient (TXNIP^Cre) mice induced with asthma using a protease from Aspergillus oryzae (Asp). (A) The Penh value graph of airway hyperresponsiveness measurements. Increased airway resistance was observed in TXNIP^Cre compared to TXNIP^fl/fl. Values: mean $\pm$ standard deviation (SD) (n = 7 per group). Significance: **p$<$ 0.01 vs TXNIP^fl/fl by Welch's correction. (B) Representative images and quantification graph of Diff-Quick® staining of bronchoalveolar lavage fluid (BALF) cells (Scale bar = 60 μm). Values: mean $\pm$ SD (n = 7 per group). Significance: **p$<$ 0.01 vs TXNIP^fl/fl by Welch's correction. (C) Representative population dot plot and (D) quantification graph of each inflammatory cells in bronchoalveolar lavage fluid analyzed by flow cytometry. PerCP-CD45⁺ cells, Lymphocyte; PerCP-CD45⁺FITC-LY6G⁺ cells, Neutrophils; PerCP-CD45⁺PE-CD11c⁺APC-SiglecF⁺ cells, Alveolar macrophages; PerCP-CD45⁺APC-SiglecF⁺ cells, Eosinophils. Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs TXNIP^fl/fl by two-tailed t-test, respectively. BALF cells staining and flow cytometry analysis revealed a significant increase in the number of eosinophils in TXNIP^Cre, which is associated with asthma development. (E) Representative images of hematoxylin/eosin (H&E; Scale bar = 100 μm) and periodic acid-Schiff (PAS; Scale bar = 40 μm) staining of lung tissue and quantification graphs for histological scores. Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs TXNIP^fl/fl by two-tailed t-test, respectively. Higher airway inflammation and mucus overproduction in lung tissue due to Asp was observed in TXNIP^Cre than TXNIP^fl/fl.(F) Cytokine levels of immunoglobulin (Ig)E, interleukin (IL)-4, IL-5, IL-13, and IL-17A analyzed by enzyme-linked immunosorbent assay. Values: mean $\pm$ SD (n = 7 per group). Significance: *,**p$<$ 0.05 and 0.01 vs TXNIP^fl/fl, respectively. Statistical analysis: IgE and IL-13, by two-tailed t-test; IL-4, IL-5 and IL-17A, by Mann Whitney test. (G) Immunohistochemical staining for TXNIP in lung sections (Scale bar = 30 μm) and quantification graph of TXNIP positive area. Values: mean $\pm$ SD (n = 7 per group). Significance: *p$<$ 0.05 vs TXNIP^fl/fl by two-tailed t-test. (H) Western blot analysis of the indicated proteins in the lung tissue and graphs representing the densitometric values of protein expression. NLRP3, NOD-like receptor family pyrin domain containing 3; ASC, apoptosis-associated speck-like protein containing a caspase-recruitment domain; IL-1β, interleukin-1β. Values: means $\pm$ SD (n = 4 per group). Significance: *,**p$<$ 0.05 and 0.01 vs TXNIP^fl/fl by two-tailed t-test, respectively.

Fig. 5

Effects of thioredoxin-interacting protein (TXNIP) expression on pro-inflammatory cytokines and reactive oxygen species (ROS) levels in A549 cells. The mRNA expression levels of (A) interleukin (IL)-1β, (B) IL-6, and (C) tumor necrosis factor (TNF)-α as determined by quantitative reverse transcription polymerase chain reaction. Values: mean $\pm$ standard deviation (SD) (n = 4 per group). Significance: *,**p$<$ 0.05 and 0.01 by Tukey's post hoc test, respectively. (D) Representative images of intracelluar ROS levels in A549 cells as detected by fluorescence microscopy (Scale bar = 25 μm). ROS were detected in cells as indicated by distinct green fluorescence in the cytoplasm, and nuclei were stained with DAPI (blue). An increase in fluorescence intensity was observed around the nucleus in cells transfected with TXNIP siRNA, which was reduced in TNXIP-overexpressing cells. NC, non-treated cells; rTNF-α, cells stimulated with recombinant human TNF-α protein (10 ng/mL); DCFDA, 2′,7′-dichlorofluorescin diacetate.