Endothelin regulates intermittent hypoxia-induced lipolytic remodelling of adipose tissue and phosphorylation of hormone-sensitive lipase

Abstract

Obstructive sleep apnoea syndrome is characterized by repetitive episodes of upper airway collapse during sleep resulting in chronic intermittent hypoxia (IH). Obstructive sleep apnoea syndrome, through IH, promotes cardiovascular and metabolic disorders. Endothelin-1 (ET-1) secretion is upregulated by IH, and is able to modulate adipocyte metabolism. Therefore, the present study aimed to characterize the role of ET-1 in the metabolic consequences of IH on adipose tissue in vivo and in vitro. Wistar rats were submitted to 14 days of IH-cycles (30 s of 21% FiO2 and 30 s of 5% FiO2 ; 8 h day(-1) ) or normoxia (air-air cycles) and were treated or not with bosentan, a dual type A and B endothelin receptor (ETA-R and ETB-R) antagonist. Bosentan treatment decreased plasma free fatty acid and triglyceride levels, and inhibited IH-induced lipolysis in adipose tissue. Moreover, IH induced a 2-fold increase in ET-1 transcription and ETA-R expression in adipose tissue that was reversed by bosentan. In 3T3-L1 adipocytes, ET-1 upregulated its own and its ETA-R transcription and this effect was abolished by bosentan. Moreover, ET-1 induced glycerol release and inhibited insulin-induced glucose uptake. Bosentan and BQ123 inhibited these effects. Bosentan also reversed the ET-1-induced phosphorylation of hormone-sensitive lipase (HSL) on Ser(660) . Finally, ET-1-induced lipolysis and HSL phosphorylation were also observed under hypoxia. Altogether, these data suggest that ET-1 is involved in IH-induced lipolysis in Wistar rats, and that upregulation of ET-1 production and ETA-R expression by ET-1 itself under IH could amplify its effects. Moreover, ET-1-induced lipolysis could be mediated through ETA-R and activation of HSL by Ser(660) phosphorylation.

Figure 1. Plasma metabolic parameters in rats

Plasma levels of (A) FFAs (μmol L^–1), (B) triglycerides (g L^–1) and (C) glucose (mmol L^–1) in rats submitted to 14 days of normoxia (N) or intermittent hypoxia (IH) with or without oral bosentan treatment (100 mg kg^–1 day^–1) (n = 8 rats per group). ^# P < 0.05 vs. control, two‐way ANOVA. There was a significant interaction between IH exposure and bosentan treatment for glycaemia (P = 0.033).

Figure 2. ET‐1 contributes to intermittent hypoxia‐induced adipose tissue lipolysis

A, ratio (%) of epididymal fat to total body weight after 14 days of exposure to normoxia (N) or intermittent hypoxia (IH) with or without oral bosentan treatment (100 mg kg^–1 day^–1). B, hematoxylin staining of representative slices of adipose tissue from untreated N and IH rats. Scale bar = 40 μm. C, quantification of mean adipose cell area on at least three epididymal fat sections per rat (n = 8 rats per group) *P < 0.05 vs. N; two‐way ANOVA.

Figure 3. Intermittent hypoxia regulates adipose tissue ET‐1 and ETA‐R expression

A, qPCR quantification of ET‐1 mRNA in epididymal fat of rats exposed to 14 days of normoxia (N) or intermittent hypoxia (IH). *P < 0.05 vs. N; Mann–Whitney rank sum test. B, ET‐1 protein level in epididymal fat extracts of rats exposed to 14 days of N or IH with or without oral bosentan treatment (100 mg kg^–1 day^–1). ^# P < 0.05 vs. control; two‐way ANOVA. C, immunohistochemistry of epididymal fat stained with anti‐ETA‐R antibody. Scale bar = 30 μm. D, quantification of ETA‐R expression normalized to the area of cell membrane in the field, from at least three epididymal fat sections per rat (n = 8 rats per group). *P < 0.05 vs. N and ^# P < 0.05 vs. control; two‐way ANOVA.

Figure 4. ET‐1 regulates its expression and that of ETA‐R in 3T3‐L1 adipocytes

qPCR quantification of ET‐1 (A) and ETA‐R (B) mRNA in mature 3T3‐L1 adipocytes exposed to ET‐1 (10 nm) with or without BQ123, BQ788 or bosentan (Bos) (10 μm each) for 24 h. *P < 0.05; one‐way ANOVA on ranks. Western blot (C) and quantification (D) of ETA‐R protein expression in mature 3T3‐L1 adipocytes submitted to ET‐1 (10 nm) with or without bosentan (10 μm) for 24 h. Quantification was performed on at least three independent experiments. *P < 0.05 vs. control and ^# P < 0.05 vs. ET‐1; two‐way ANOVA. Vertical dividing lines indicate the components parts of a single image, where intercalating bands were suppressed.

Figure 5. ET‐1 induces lipolysis and reduces glucose uptake in 3T3‐L1 adipocytes

A, lipolysis assessed by glycerol release (% of control hypoxic cells) in culture media of 3T3‐L1 adipocytes cultivated under hypoxia (3% O₂) and treated or not with ET‐1 (10 nm) and bosentan (Bos) (10 μm) for 4 or 24 h. *P < 0.05 vs. control untreated cells. ^# P < 0.05 vs. the corresponding ET‐1‐treated cells; two‐way ANOVA (n = 4 independent experiments). Lipolysis (B), assessed by glycerol release (% of control untreated cells) in culture media and glucose uptake (C) (% of control untreated cells) in 3T3‐L1 adipocytes and treated or not for 24 h with ET‐1 (10 nm), insulin (5 nm), BQ123, BQ788 or bosentan (10 μm each) under normoxia. *P < 0.05 vs. control untreated cells. ^# P < 0.05 vs. the corresponding insulin‐treated cells; one‐way ANOVA on ranks (n = 5 independent experiments).

Figure 6. ET‐1 regulates HSL phosphorylation but not HSL or ATGL expression

Western blot quantification (normalized to actin) of HSL (A) and ATGL (B) expression in 3T3‐L1 adipocytes treated or not for 5  or 10 h with ET‐1 (10 nm) and bosentan (Bos) (10 μm). C, western blots of HSL and phosphorylated (pHSL) on Ser⁶⁶⁰. Quantification of pHSL (normalized to total HSL) on (D) activatory Ser⁶⁶⁰ and (E) inhibitory Ser⁵⁶⁵. Western blots are representative of five independent experiments performed in duplicate except for (B) and (E) (n = 3 experiments performed in duplicate). *P < 0.05 and **P < 0.01 vs. untreated cells. ^# P < 0.05 vs. ET1‐treated cells; two‐way ANOVA.

Figure 7. Endothelin‐1 induces HSL phosphorylation in 3T3‐L1 cultivated in hypoxia

A, HSL phosphorylation on Ser⁶⁶⁰ in 3T3‐L1 adipocytes cultivated under hypoxia (H) (3% O₂) and treated or not with ET‐1 (10 nm) and bosentan (Bos) (10 μm) for 4 or 24 h. B, quantification of the western blots. Western blots are representative of three independent experiments performed in triplicate. *P < 0.05 vs. control H; two‐way ANOVA. P = 0.054 at 4 h for H + ET1 + Bos vs. H + ET1.

Figure 8. Proposed mechanisms explaining the dual action of endothelin‐1 on glucose and lipid metabolism in adipocytes

ET‐1 binding to its ETA receptor inhibits insulin‐induced glucose uptake and activates HSL phosphorylation on Ser⁶⁶⁰, thereby inducing diacylglycerol (DAG) lipolysis into monoacylglycerol (MAG) + fatty acid (FA). MAG conversion by monoglyceride lipase (not represented) results in glycerol and fatty acid release. Glc, glucose.