Research ArticleCELL BIOLOGY

Obesity-induced DNA released from adipocytes stimulates chronic adipose tissue inflammation and insulin resistance

See allHide authors and affiliations

Science Advances  25 Mar 2016:
Vol. 2, no. 3, e1501332
DOI: 10.1126/sciadv.1501332
  • Fig. 1 Obesity-related adipocyte degeneration and cfDNA release.

    (A to C) Plasma levels of ssDNA (A), dsDNA (B), and nucleosomes (C) (n = 9). au, arbitrary units. (D and E) Levels of ssDNA (D) and dsDNA (E) in CM obtained from an organ culture experiment using epididymal fat (n = 5). (F) Correlation between plasma level of ssDNA and blood glucose level (n = 18). (G) Representative figure of Western blot analysis of perilipin expression in epididymal fat. Expression of perilipin was quantified by densitometry and normalized to the corresponding signal for β-actin (n = 5). (H) Quantitative RT-PCR analysis of TLR9 expression in epididymal fat (n = 5). (I) Cell-type–specific expression of TLR9 in epididymal fat obtained from fat-fed mice (n = 5). (J) Representative immunogold staining against ssDNA revealing accumulation of gold particles (10 nm) in the cytoplasm of macrophages (arrows) in epididymal fat obtained from fat-fed obese mice. The accumulation of gold particles was not observed in adipose tissue macrophages in lean mice (n = 4). Scale bar, 100 nm. Inset: lower magnification (scale bar, 2 μm). Cyto, cytoplasm; Nuc, nucleus. All samples were obtained from wild-type (WT) mice fed a high-fat diet (HFD) or NC for 12 weeks. *P < 0.05, **P < 0.01, and ***P < 0.001. All values are means ± SEM.

  • Fig. 2 Role of TLR9 in macrophage activation.

    (A) CpG-ODN1826 (CpG1826) (0.1 to 1.0 μM), a TLR9-specific ligand, increased the expression of MCP-1 in peritoneal macrophages obtained from WT mice, but not in macrophages obtained from Tlr9−/− mice (n = 6). NT, non-treatment. (B) iODN2088 (0.1 μM), a specific antagonist of TLR9, inhibited the MCP-1 expression induced by CpG1826 (0.1 μM) in WT macrophages (n = 6). (C and D) Ligation of CpG1826 (0.1 μM) to TLR9 activated the NF-κB pathways determined by the phosphorylation of IκBα in WT macrophages, which was abolished by iODN2088. Neither CpG nor iODN2088 influenced the phosphorylation of IκBα in Tlr9−/− macrophages. Representative figure of Western blot analysis of IκBα phosphorylation (C) and the result of the quantification of phosphorylated IκBα normalized to the corresponding signal for total IκBα by densitometry (D) are shown (n = 4). (E) CM from control 3T3-L1 adipocytes increased MCP-1 expression in WT and Tlr9−/− macrophages. CM from degenerated adipocytes further promoted MCP-1 expression in WT macrophages, although this response was attenuated in Tlr9−/− macrophages (n = 5). After a 24-hour pretreatment with or without TNF-α, adipocytes were cultured in a starvation medium without TNF-α for another 24 hours. Culture media were then collected as CM of degenerated or control adipocytes, respectively, and used in the experiments. (F) Coculture of macrophages and 3T3-L1 adipocytes using a Transwell membrane slightly increased MCP-1 expression in WT and Tlr9−/− macrophages. Coculture with degenerated adipocytes increased MCP-1 expression in WT macrophages more efficiently, although this response was attenuated in Tlr9−/− macrophages (n = 6). (G) cfDNA extracted from degenerated adipocyte CM promoted MCP-1 expression in WT macrophages, but not in Tlr9−/− macrophages (n = 6 to 8). cfDNA extracted from 130 μl of CM was used to stimulate each well. CM from degenerated adipocytes were collected as shown in (E). *P < 0.05 and ***P < 0.001. All values are means ± SEM.

  • Fig. 3 Effects of genetic deletion of TLR9 on adipose tissue inflammation and insulin resistance.

    (A) Mac3 staining of epididymal fat obtained from WT or Tlr9−/− mice (n = 12 to 15). Scale bar, 100 μm. (B) Quantitative RT-PCR analysis of inflammatory gene expression in epididymal fat obtained from WT or Tlr9−/− mice (n = 12). (C) Western blot analysis of the phosphorylation of IκBα in epididymal fat obtained from WT or Tlr9−/− mice (n = 11). (D) Results of the insulin tolerance test (0.75 U/kg) of WT or Tlr9−/− mice (n = 13 to 16). (E) Quantitative RT-PCR analysis of epididymal fat for genes related to insulin sensitivity (n = 12). C/EBPα, CCAAT/enhancer binding protein α. All experiments in this figure were performed using samples obtained from mice fed a high-fat diet (HFD) for 12 weeks. §P = 0.05, *P < 0.05, **P < 0.01, and ***P < 0.001. All values are means ± SEM.

  • Fig. 4 Effects of hematopoietic restoration of TLR9 on adipose tissue inflammation and insulin resistance.

    (A) Mac3 staining of epididymal fat obtained from BM chimeric mice (n = 7). Scale bar, 100 μm. (B) Quantitative RT-PCR analysis of inflammatory gene expression in epididymal fat obtained from BM chimeric mice (n = 7). (C) Western blot analysis of the phosphorylation of IκBα in epididymal fat obtained from BM chimeric mice (n = 6 to 7). (D) Results of insulin tolerance test (0.75 U/kg) of BM chimeric mice (n = 7 to 8). (E) Quantitative RT-PCR analysis of epididymal fat obtained from BM chimeric mice for genes related to insulin sensitivity (n = 7). All experiments in this figure were performed using samples obtained from mice fed a high-fat diet (HFD) for 12 weeks. P = 0.06, *P < 0.05, **P < 0.01, and ***P < 0.001. All values are means ± SEM.

  • Fig. 5 Effects of in vivo blockade of TLR9 on adipose tissue inflammation and insulin resistance in WT mice.

    (A) Effects of intraperitoneal injection of iODN2088 on inflammatory gene expression in WT peritoneal macrophages (n = 5 to 6). (B) Mac3 staining of epididymal fat obtained from fat-fed WT mice treated with iODN2088 or its control (Ctrl-iODN) (n = 5 to 6). Scale bar, 100 μm. (C) Effects of in vivo iODN2088 treatment on inflammatory gene expression in epididymal fat (n = 5 to 6). (D) Effects of in vivo iODN2088 treatment on the phosphorylation of IκBα in epididymal fat (n = 5 to 6). (E) Response to intraperitoneal injection of insulin (1.0 U/kg) in iODN2088-treated or Ctrl-iODN–treated mice (n = 5 to 6). (F) Effects of in vivo iODN2088 treatment on the expression of genes related to insulin sensitivity in epididymal fat (n = 5 to 6). §P = 0.08, *P < 0.05, **P < 0.01, and ***P < 0.001. All values are means ± SEM.

  • Fig. 6 Relationship between cfDNA and visceral obesity or insulin resistance in humans.

    (A) Comparison of plasma ssDNA level between humans with and without computed tomography–determined visceral obesity (VFA ≥ 100 cm2). (B) Correlation between plasma ssDNA level and VFA. (C) Correlation between plasma ssDNA level and insulin resistance determined by HOMA-IR. Human plasma samples that were collected upon medical examination at regional health checkups were used (n = 131). *P < 0.05. All values are means ± SEM.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/2/3/e1501332/DC1

    Fig. S1. Correlations of cfDNA with epididymal fat weight, liver weight, and blood glucose level.

    Fig. S2. Adipocyte degeneration induced by a high-fat diet.

    Fig. S3. Effects of TLR9 agonist and antagonist on MCP-1 expression in macrophages.

    Fig. S4. Effects of TLR9 agonist and antagonist on Tlr9−/− macrophages.

    Fig. S5. Release of cfDNA from TNF-α–treated adipocytes.

    Fig. S6. Expression of Tlr2, Tlr4, and Tlr7 in Tlr9−/− macrophages.

    Fig. S7. Effects of TLR9 antagonist on macrophages treated with cfDNA.

    Fig. S8. Level of obesity between Tlr9−/− mice and wild-type mice.

    Fig. S9. Comparison of blood glucose and serum insulin levels between wild-type mice and Tlr9−/− mice.

    Fig. S10. Food consumption and metabolic studies of mice.

    Fig. S11. Insulin signaling in VAT obtained from fat-fed mice.

    Fig. S12. Comparison between Tlr9−/− mice and wild-type mice fed NC.

    Fig. S13. Level of obesity between BMT mice and non-BMT mice fed a high-fat diet.

    Fig. S14. Replacement rate of BM cells after transplantation.

    Fig. S15. Detection of bacterial DNA in plasma cfDNA.

    Table S1. Comparison between Tlr9−/− mice and wild-type mice.

    Table S2. Comparison between BM chimeric mice.

    Table S3. Comparison between iODN2088 and control for iODN-treated wild-type mice.

    Table S4. Multivariate analysis estimating HOMA-IR.

    Table S5. Primer sequences.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Correlations of cfDNA with epididymal fat weight, liver and blood glucose level.
    • Fig. S2. Adipocyte degeneration induced by a high-fat diet.
    • Fig. S3. Effects of TLR9 agonist and antagonist on MCP-1 expression in macrophages.
    • Fig. S4. Effects of TLR9 agonist and antagonist on Tlr9−/− macrophages.
    • Fig. S5. Release of cfDNA from TNF-α–treated adipocytes.
    • Fig. S6. Expression of Tlr2, Tlr4, and Tlr7 in Tlr9−/− macrophages.
    • Fig. S7. Effects of TLR9 antagonist on macrophages treated with cfDNA.
    • Fig. S8. Level of obesity between Tlr9−/− mice and wild-type mice.
    • Fig. S9. Comparison of blood glucose and serum insulin levels between wild-type mice and Tlr9−/− mice.
    • Fig. S10. Food consumption and metabolic studies of mice.
    • Fig. S11. Insulin signaling in VAT obtained from fat-fed mice.
    • Fig. S12. Comparison between Tlr9−/− mice and wild-type mice fed NC.
    • Fig. S13. Level of obesity between BMT mice and non-BMT mice fed a high-fat diet.
    • Fig. S14. Replacement rate of BM cells after transplantation.
    • Fig. S15. Detection of bacterial DNA in plasma cfDNA.
    • Table S1. Comparison between Tlr9−/− mice and wild-type mice.
    • Table S2. Comparison between BM chimeric mice.
    • Table S3. Comparison between iODN2088 and control for iODN-treated wild-type mice.
    • Table S4. Multivariate analysis estimating HOMA-IR.
    • Table S5. Primer sequences.

    Download PDF

    Files in this Data Supplement:

Navigate This Article