Pharmacological inhibition of adipose triglyceride lipase corrects high-fat diet-induced insulin resistance and hepatosteatosis in mice

Abstract

Elevated circulating fatty acids (FAs) contribute to the development of obesity-associated metabolic complications such as insulin resistance (IR) and non-alcoholic fatty liver disease (NAFLD). Hence, reducing adipose tissue lipolysis to diminish the mobilization of FAs and lower their respective plasma concentrations represents a potential treatment strategy to counteract obesity-associated disorders. Here we show that specific inhibition of adipose triglyceride lipase (Atgl) with the chemical inhibitor, Atglistatin, effectively reduces adipose tissue lipolysis, weight gain, IR and NAFLD in mice fed a high-fat diet. Importantly, even long-term treatment does not lead to lipid accumulation in ectopic tissues such as the skeletal muscle or heart. Thus, the severe cardiac steatosis and cardiomyopathy that is observed in genetic models of Atgl deficiency does not occur in Atglistatin-treated mice. Our data validate the pharmacological inhibition of Atgl as a potentially powerful therapeutic strategy to treat obesity and associated metabolic disorders.

Figure 1. Atglistatin transiently inhibits lipolysis and protects from HFD-induced obesity.

Six weeks old male C57Bl6J mice were fed a HFD (45 kJ% fat; 22.1 kJ g⁻¹) for 50 days. Thereafter, mice were fasted for 7 h and then re-fed a HFD with or without Atglistatin (2 mmol kg⁻¹ diet) for 2 h followed by a second, subsequent fasting period of 8 h. (a) Plasma FA levels were determined in the 2 h re-fed and 8 h fasted state (n=5). (b) FA release from gonadal adipose tissue explants of re-fed and (c) 8 h-fasted mice (n=5 per group). (d,e) Atglistatin does not inhibit human adipocyte lipolysis. (d) TG hydrolase activity was assessed in COS-7 lysates of cells overexpressing human and murine ATGL and CGI-58, respectively, in the presence and absence of the indicated concentrations of Atglistatin. (e) SGBS and 3T3-L1 preadipocytes were differentiated to adipocytes. Then, cells were preincubated with the indicated concentrations of Atglistatin for 2 h. Thereafter, the medium was replaced by DMEM containing 2% BSA, 10 μM Forskolin and the indicated concentrations of Atglistatin for 1 h. The release of FA in the medium was determined and calculated per mg cell protein. (f–k) Mice were fed a HFD for 50 days, followed by HFD-feeding in the presence or absence of Atglistatin for another 50 days. (f) Body weight, and (g) fat and lean mass development (n=7 per group). Adipose tissue depots, inguinal (i)WAT, gonadal (g)WAT, and interscapular (i)BAT were analysed for their (h) weight and (i) adipocyte size. (j,k) mRNA expression of IL-6 and macrophage markers F4/80 and Cd11c was assessed in gWAT and iBAT of re-fed mice (n=5 per group). Data represent mean±s.d. Statistical significance was determined by Student's two-tailed t-test. For analysis of multiple measurements, we performed one-way analysis of variance (ANOVA) followed by Bonferroni post-hoc test; *P<0.05, **P<0.01 and ***P<0.001 for control versus ATGLi; ^#P<0.05 and ^##P<0.01 for basal versus + Atglistatin.

Figure 2. Metabolic characterization of mice fed a HFD in the presence and absence of Atglistatin.

Six weeks old male C57Bl6J mice were fed a HFD (45 kJ% fat; 22.1 kJ g⁻¹) for 50 days. Thereafter, mice were fed a HFD in the presence and absence of Atglistatin for 30 days. Metabolic phenotyping was performed using a laboratory animal monitoring system (LabMaster, TSE Systems, n=6 per group). Mice were familiarized with the metabolic cages for 3 days prior measurement. (a) EE was calculated using the formula: EE(kJ per day)=15,818*VO₂+5,176*VCO₂/1,000*24 and is plotted against body weight. Linear regression analysis was performed using GraphPad prism software. Data represent single mice. (a, insert) EE is expressed as adjusted means based on a normalized mouse weight of 35.73 g determined using ANCOVA, P=0.94. (b) Locomotor activity per day. (c) Food intake per mouse and day, and (d) per mouse and hour during the course of the day. (e) Respiratory exchange ratio (CO₂/O₂) during the course of the day and (f) per day. Mice were kept on a 14 h light and 10 h dark cycle. Grey bars indicate dark phases. (g) Water intake per mouse and day. Data represent mean±s.d. Statistical significance between control and ATGLi was determined by Student's two-tailed t-test; *P<0.05, **P<0.01 and ***P<0.001.

Figure 3. Hypophagia and reduced lipid deposition in WAT contribute to the obesity resistant phenotype.

(a) Six weeks old male C57Bl6J mice were fed a HFD (45 kJ% fat; 22.1 kJ g⁻¹) for 50 days. Thereafter, mice were fed a HFD in the presence or absence of Atglistatin for 68 days. Daily food intake in single housed mice was determined by subtracting food uneaten from food given to the bottom of the cages. Pair-feeding was performed by giving control mice the same amount of food ATGLi animals have eaten the day before (±0.1 g; n=5 per group). (b) Body weight of pair-fed and ad libitum-fed animals. (c) Faeces output was determined on 3 consecutive days (n=8 per group). (d) Faeces of mice were sampled on 3 consecutive days and analysed for the excreted energy using a bomb calorimeter (n=8 per group). (e) Lipid absorption was determined on 3 consecutive days using sucrose polybehenate (5% of dietary fat, w/w) as internal standard in the HFD. Fat absorption is calculated from the ratios of behenic acid to other FA in diet and faeces, analysed by gas chromatography of FA methyl esters (n=9 per group). (f) Lipid tolerance tests were performed by oral gavage of 200 μl olive oil and subsequent determination of acylglycerol levels in plasma of control and ATGLi mice (n=5 per group). (g) Heparin releasable LPL activity was determined in WAT of re-fed animals after 50 days of diet intervention (n=5 per group). (h) mRNA expression of PPARγ-target genes was measured in gonadal WAT of re-fed mice (n=5 per group). Data represent mean±s.d. Statistical significance between control and ATGLi was determined by two-tailed Student's t-test; *P<0.05, **P<0.01 and ***P<0.001.

Figure 4. Atglistatin-mediated inhibition of lipolysis improves glucose homeostasis.

(a,b) Six weeks old male C57Bl6J mice were fed a HFD (45 kJ% fat; 22.1 kJ g⁻¹) for 50 days. Thereafter, mice were fed a HFD in the presence and absence of Atglistatin. (c–f) Twenty weeks old ob/ob mice were fed a chow diet in the presence and absence of Atglistatin. (a,f) Insulin sensitivity and (b,e) glucose tolerance was determined after 30 and 40 days diet intervention by i.p. injection of 0.5 IU kg⁻¹ insulin and 1.5 g kg⁻¹ glucose, respectively (n=6). Area under the curve (AUC) was calculated using GraphPad Prism software. (c) Body weight of ob/ob animals fed a chow diet in the presence and absence of Atglistatin (n=6). (d) Averaged daily food intake of ob/ob animals fed a chow diet in the presence and absence of Atglistatin (n=6). Data represent mean±s.d. Statistical significance was determined by Student's two-tailed t-test; *P<0.05, **P<0.01 and ***P<0.001 for control versus ATGLi.

Figure 5. Atglistatin protects from HFD-induced NAFLD.

Six weeks old male C57Bl6J mice were fed a HFD (45 kJ% fat; 22.1 kJ g⁻¹) for 50 days. Thereafter, mice were fed a HFD in the presence and absence of Atglistatin for 140 days. (a) Atglistatin abundance was determined in extracts of adipose- and non-adipose tissues using liquid chromatography–mass spectrometry analysis (n=6 per group). (b,g,h) Total lipids were extracted and acylglycerol levels were determined in (b) m. quadriceps, and heart and (g) livers of wt mice fed a HFD or (h) in livers of ob/ob mice fed a chow diet in the presence or absence of Atglistatin (n=6 per group). (c) mRNA expression of Pparα, Cpt1b and Pgc-1α was assessed in hearts of moderately fasted animals (n=6 per group). (d) mRNA expression of inflammatory and fibrosis marker genes was assessed in hearts of re-fed animals (n=9 per group). (e) Left ventricular mass and (f) ejection fraction of hearts from control and ATGLi animals was assessed by MRI after 5 weeks of diet intervention (n=5 per group). (i) Haematoxylin–eosin sections of liver tissues were categorized according to lipid droplet abundance, inflammatory foci and cell ballooning. NAFLD score is the sum of the steatosis, inflammation and ballooning score. (j) ALT enzyme activity was measured in fresh isolated plasma using a commercial kit (n=9 per group). (k) mRNA expression of inflammatory and fibrosis marker genes was assessed in livers of re-fed animals (n=7 per group). Data represent mean+s.d. Statistical significance between control and ATGLi was determined by two-tailed Student's t-test; *P<0.05, **P<0.01 and ***P<0.001.