Endotrophin triggers adipose tissue fibrosis and metabolic dysfunction

Abstract

We recently identified endotrophin as an adipokine with potent tumour-promoting effects. However, the direct effects of local accumulation of endotrophin in adipose tissue have not yet been studied. Here we use a doxycycline-inducible adipocyte-specific endotrophin overexpression model to demonstrate that endotrophin plays a pivotal role in shaping a metabolically unfavourable microenvironment in adipose tissue during consumption of a high-fat diet (HFD). Endotrophin serves as a powerful co-stimulator of pathologically relevant pathways within the 'unhealthy' adipose tissue milieu, triggering fibrosis and inflammation and ultimately leading to enhanced insulin resistance. We further demonstrate that blocking endotrophin with a neutralizing antibody ameliorates metabolically adverse effects and effectively reverses metabolic dysfunction induced during HFD exposure. Collectively, our findings demonstrate that endotrophin exerts a major influence in adipose tissue, eventually resulting in systemic elevation of pro-inflammatory cytokines and insulin resistance, and the results establish endotrophin as a potential target in the context of metabolism and cancer.

Figure 1. Endotrophin expression in AT induces fibrosis and inflammation-related genes in WAT

(a) Q–PCR analysis for stromal–vascular versus adipocyte distribution of Col6 mRNAs. Data represents mean±s.d. (n = 5). **P<0.01 and ***P<0.001 versus stromal-vesicular by two-way analysis of variance. (b) PCR analysis of endotrophin overexpression in different fat pads (EWAT (epididymal), SWAT (subcutaneous), MWAT (mesenteric) and BAT (brown adipose tissue) and other organs in the adiponectin promoter-driven TRE-ETP (AdnP-ETP) transgenic mice and their littermate controls 5 days after Dox induction on a chow diet (n = 5 per group). (c) Immunohistochemical staining with an anti-mouse endotrophin antibody of EWAT from the transgenic mice and their littermate controls. Scale bar, 100 μm. (d) Q–PCR analysis of pro-fibrotic and pro-inflammatory genes affected by endotrophin overexpression in SWAT of the transgenic mice and their littermate controls 5 days after Dox induction (n = 5 per group, Student's t-test for all the P-values).

Figure 2. Overexpression of endotrophin locally in adipose tissue promotes metabolic dysfunction under a HFD challenge

(a) Body weight gain in endotrophin transgenic mice and their littermate controls during HFD plus 600 mg kg^–1 Dox treatment for 8 weeks (n = per group, Student's t-test for all the P-values). (b) Serum TG and (c) NEFA levels in endotrophin transgenic mice and their littermate controls after HFD plus Dox treatment for 8 weeks (n = 5 per group, Student's t-test for all the P-values). (d) Liver TG and (e) cholesterol levels in endotrophin transgenic mice and their littermate controls after HFD plus Dox treatment for 8 weeks (n = 4 for control; n = 5 for endotrophin Tg, Student's t-test for all the P-values). (f) H&E staining of liver sections from endotrophin transgenic mice and their littermate controls 8 weeks after HFD plus Dox treatment. Scale bars, 50 μm. (g) Blood glucose levels during an OGTT and an ITT in endotrophin transgenic mice and their littermate controls after HFD plus Dox treatment for 5 weeks (n = 5 for each group, Student's t-test for all the P-values). (h) Blood glucose levels during an ITT in endotrophin transgenic mice and their littermate controls after HFD plus Dox treatment for 5 weeks (n = 5 for each group, Student's t-test for all the P-values).

Figure 3. Overexpression of endotrophin locally in adipose tissue stimulates fibrosis and exacerbates inflammation induced by HFD

(a) Q–PCR analysis for collagen 3α1 and 6α1, TGFβ and its receptor (TGFβ R2) and MMP12 in SWATof endotrophin transgenic mice and their littermate controls (n = 4 for each group, Student's t-test for all the P-values) after 8 weeks of exposure to HFD plus Dox. (b) Q–PCR analysis for adipose tissue inflammation-related genes F4/80 and SAA3 in SWAT of endotrophin transgenic mice and their littermate controls (n = 4 for each group, Student's t-test for both the P-values). (c) A Masson's trichrome stain of SWAT (left panels) and BAT (right panels) from endotrophin transgenic mice and their littermate controls. Scale bars, 50 μm for SWAT and 100 μm for BAT. The right panel shows the quantitative measurements of the stain in SWAT by ImageJ software (Student's t-test, P = 0.041). (d) H&E staining of EWAT from endotrophin transgenic mice and their littermate controls. The arrows point to the CLS formed by macrophage aggregation in HFD-fed mice. Scale bars, 200 μm. (e) Immunohistochemical staining of F4/80 in EWAT from endotrophin transgenic mice and their littermate controls (left panels). Note the CLS formed by macrophage aggregation in HFD-fed mice. The right panel shows quantitative measurements of the numbers of CLS (n = 5 per group, Student's t-test, P<0.001). Scale bar, 200 μm.

Figure 4. Endotrophin neutralization improves metabolic dysfunction

(a) Two cohorts of mice (n = 8) were exposed to HFD for 45 days after weaning. At the indicated time points (arrowheads), mice were treated with anti-endotrophin monoclonal antibodies (100 μg per mouse) or equivalent amounts of non-immune antibody. Body weights were monitored. (b) Plasma triglycerides; (c) plasma free fatty acids (NEFAs); (d) plasma cholesterol; (e) hepatic triglycerides; (f) hepatic cholesterol. Significant differences as indicated by unpaired Student's t-test. (g) Hepatic histology (H&E stain) reveals clear differences in hepatic lipid accumulation. Scale bar, 25 μm. (h) Glucose infusion rate and (i) Suppression of hepatic glucose efflux was measured during a hyperinsulinemic–euglycemic clamp. Significant differences as indicated by unpaired Student's t-test. (j) Left: H&E staining of EWAT and SWAT in either isotype IgG- or 10B6-treated HFD-induced obese mice. Scale bar, 50 μm. Quantification for CLS in EWAT and SWAT represents mean±s.d. (three different fields from n = 5 per group). Significant differences as indicated by unpaired Student's t-test. Right: Mac-2 immunostaining of EWAT and SWAT, showing decreased inflammation in 10B6-given mice compared with IgG control. Adipocyte membrane and nucleus were visualized by co-staining with anti-perilipin and DAPI. Scale bar, 200 μm. (k) Masson's trichrome C staining of SWTA shows decreased fibrosis in SWAT in 10B6-given mice compared with IgG control. Quantification for the area of TC positive stains in SWAT represents mean±s.d. (three different fields from n = 5 per group). Significant differences as indicated by unpaired Student's t-test. Scale bar, 50 μm.

Figure 5. Histological analysis of AT following endotrophin neutralization after chronic HFD exposure

(a) The schematic representation of the protocol for chronic HFD experiments. (b) H&E staining of EWAT and SWAT in either isotype IgG- or 10B6-treated HFD mice (8 weeks of HFD pre-treatment before initiation of antibody treatment). Scale bar, 50 μm. Quantification for CLS in EWATand SWATrepresents mean±s.d. (three different fields from n = 5 per group). Significant differences as indicated by unpaired Student's t-test. (c) Mac-2 immunostaining of EWAT and SWAT, showing decreased inflammation in 10B6-given mice compared with IgG control. Adipocyte membrane and nucleus were visualized by co-staining with anti-perilipin and DAPI. Scale bar, 200 μm. (d) Hepatic histology (H&E stain) reveals differences in hepatic lipid accumulation. Scale bar, 200 μm. (e) C57/Bl6J mice were challenged with HFD for 8 weeks and subsequently treated with either 10B6 (100 μg per mouse, twice weekly by IP) or IgG control for another 3 weeks. OGTT was performed. Serum glucose levels during OGTT. Data represents mean±s.d. (n = 6 per group). **P<0.01 versus IgG by multiple t-tests.

Figure 6. Endotrophin is upregulated in fat tissue from obese human subjects with HOMA-IR>2.6

(a) Human subcutaneous adipocytes from healthy obese (normal HOMA-IR) and obese patients with HOMA-IR>2.6 were stained with a polyclonal antibody that specifically recognizes human endotrophin. Scale bar, 50 μm. The right panel indicates the quantitative measurement of anti-endotrophin positive stains by ImageJ software from NIH (Student's t-test for the P-value). (b) Human subcutaneous adipocytes from the same cohort as in a were stained with Masson's trichrome C. Scale bar, 100 mm. The right panel indicates the quantitative measurement of the trichrome C positive signal by ImageJ software from NIH (Student's t-test for the P-value).