Rimonabant reduces obesity-associated hepatic steatosis and features of metabolic syndrome in obese Zucker fa/fa rats

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

This study investigated the effects of rimonabant (SR141716), an antagonist of the cannabinoid receptor type 1 (CB1), on obesity-associated hepatic steatosis and related features of metabolic syndrome: inflammation (elevated plasma levels of tumor necrosis factor alpha [TNFα]), dyslipidemia, and reduced plasma levels of adiponectin. We report that oral treatment of obese (fa/fa) rats with rimonabant (30 mg/kg) daily for 8 weeks abolished hepatic steatosis. This treatment reduced hepatomegaly, reduced elevation of plasma levels of enzyme markers of hepatic damage (alanine aminotransferase, gamma glutamyltransferase, and alkaline phosphatase) and decreased the high level of local hepatic TNFα currently associated with steatohepatitis. In parallel, treatment of obese (fa/fa) rats with rimonabant reduced the high plasma level of the proinflammatory cytokine TNFα and increased the reduced plasma level of the anti-inflammatory hormone adiponectin. Finally, rimonabant treatment also improved dyslipidemia by both decreasing plasma levels of triglycerides, free fatty acids, and total cholesterol and increasing the HDLc/LDLc ratio. All the effects of rimonabant found in this study were not or only slightly observed in pair-fed obese animals, highlighting the additional beneficial effects of treatment with rimonabant compared to diet. These results demonstrate that rimonabant plays a hepatoprotective role and suggest that this CB1 receptor antagonist potentially has clinical applications in the treatment of obesity-associated liver diseases and related features of metabolic syndrome. (HEPATOLOGY 2007.)

Obesity is a complex and multifactorial metabolic disorder resulting from an imbalance between energy intake and expenditure that may have genetic and/or behavioral origins involving the quantity and quality of food intake as well as lifestyle.1 It is characterized by increased body weight and abnormal development of adipose tissue with excessive fat storage, accompanied by a dramatic deregulation of the endocrine function of adipose tissue.2–8 Obesity is a condition currently associated with a cluster of chronic and progressive diseases that have several features of metabolic syndrome, including diabetes, hyperinsulinemia and insulin resistance, dyslipidemia, cardiovascular disorders, hepatic and renal pathologies, inflammation, and cancer.3–14 Studies aimed to determine the mediators of metabolic syndrome have found the involvement of deregulation of several cytokines, hormones, and adipocytokines.5–8, 12–17 Among these biologically active factors, proinflammatory cytokines such as tumor necrosis factor alpha (TNFα)5–8, 12–14, 18–20 and anti-inflammatory and protective hormones such as adiponectin (an adipocytokine exclusively produced by adipose tissue)7, 15–28 have been shown to play a pivotal role in diverse pathologies related to metabolic syndrome and its complications.5–8, 15–17

Hepatic steatosis and its related inflammatory state (steatohepatitis) are the principal hepatic complications of obesity and metabolic diseases. Hepatic steatosis, or fatty liver, is a disorder characterized by fat infiltration and excessive accumulation of lipids such as triglycerides in the liver, accompanied by an increased liver/body weight ratio and higher plasma levels of enzyme markers of liver damage (alanine aminotransferase [ALT], gamma glutamyltransferase [GGT], and alkaline phosphatase [ALP]). This pathology, which is often associated with obesity, hyperinsulinemia, and insulin resistance shows an inflammatory state called steatohepatitis, characterized by increased hepatic and plasma levels of several proinflammatory cytokines, particularly TNFα, which may play a crucial role in the progress of steatohepatitis into hepatic necrosis, fibrosis, and cirrhosis.18, 29–32 At present, no pharmacotherapy is available that can fully reverse and prevent steatohepatitis. Even though therapy focused on diet and exercise partly improves this disease in its early stages, the need for medical therapy is important. Rimonabant (SR141716), an antagonist of the cannabinoid receptor subtype 1 (CB1), has been shown to have potent anti-obesity effects and to improve several pathological features of obesity-associated chronic and progressive diseases: it reduced food intake, body weight, fat mass, and hyperinsulinemia and ameliorated insulin sensitivity and plasma lipid parameters in obese rodents33–39 and in humans.40–42

The aim of the present study was to investigate the effects of rimonabant on obesity-associated hepatic steatosis, on the high hepatic TNFα level that characterizes steatohepatitis, and on features of its related metabolic syndrome (dyslipidemia, inflammation, elevated plasma TNFα, and low plasma level of adiponectin) in an animal model of obesity and diabetes: the obese Zucker (fa/fa) rat.

Abbreviations

ALP, alkaline phosphatase; ALT, alanine aminotransferase; CB1 receptor, cannabinoid receptor type 1; GGT, gamma glutamyltransferase; HDLc, high-density lipoprotein-cholesterol; LDLc, low-density lipoprotein-cholesterol; TNFα: tumor necrosis factor alpha.

Materials and Methods

Animals.

Ten-week-old to 11-week-old male obese Zucker (fa/fa) rats and their lean littermates were purchased from Iffa Crédo (France). All animals were housed individually with food and water freely available and were maintained at room temperature under a 12-hour light/12-hour dark cycle. Once a day animals were orally administrated either vehicle (0.05% Tween 80 in water) or vehicle supplemented with rimonabant at a dose of 30 mg/kg. Body weight and food intake of individual rats were monitored daily. Pair-fed animals were housed individually with water freely available and received a daily quantity of food equal to the mean quantity of food consumed by the animals treated with rimonabant. After 8 weeks of treatments and 24 hours after the last administration of treatment, animals were anesthetized with pentobarbital (1 mg/kg) and were killed. Trunk blood from individual rats was collected by cardiac withdrawal in order to determine plasma levels of different parameters. The weights of individual collected livers were determined, and the organs were divided into 2 parts: one part was frozen immediately at −80°C for biochemical analysis, and the other part was kept in 10% formaldehyde for histological analysis.

All procedures were approved by the Comité d'Expérimentation Animale (Animal Care and Use Committee of Sanofi-Synthélabo Research).

Histological Analysis.

Livers were fixed overnight in buffered formaldehyde (10%), embedded in paraffin, cut with a microtome into 5-μm-thick paraffin sections, placed on glass slides and stained with hematoxylin and eosin (H&E). For the analysis of general hepatic morphology and fat content, 10 randomly selected sections from each liver were examined and photographed.

Measurement of Plasma Levels of TNFα, ALT, GGT, ALP, Triglycerides, Free Fatty Acids, Total Cholesterol, LDLc, and HDLc.

Plasma TNFα levels were quantified using commercial ELISA kits as described in the manufacturer's protocol (Biosource). Plasma levels of ALT, GGT, and ALP were measured using photometric assay kits and a HITACHI 912 biochemical analyzer as described in the manufacturer's protocol (Roche Diagnostics). Plasma levels of triglycerides, free fatty acids, total cholesterol, high-density lipoprotein-cholesterol (HDLc), and low-density lipoprotein-cholesterol (LDLc) were measured using commercial kits (ABX Diagnostics) and a HITACHI 912 biochemical analyzer (Roche Diagnostics) as previously described.39

RNA Preparation and RT-PCR Analysis of TNFα Expression.

Total RNA was prepared from rat livers using TRIZOL reagent (Invitrogen). For RT-PCR, 2 μg of total RNA was reverse-transcribed using oligo (dT) primer, and synthesized cDNA was quantified. PCR amplification of the first-strand cDNA was carried out according to the manufacturer's protocol (Invitrogen) in the presence of either two TNFα cDNA–specific primers (sense primer 5′-CCACGTCGTAGCAAACCAC-3′; antisense primer 5′-TGACTCCAAAGTAGACCTGC-3′) or two β actin cDNA–specific primers (sense primer 5′-GGGTCACCCACACTGTGC-3′; antisense primer 5′-TGCTTGCTGATCCACATCTG-3′). PCR amplification was realized in its log phase using starting cDNA quantities of 2 and 0.06 ng for TNFα and β actin, respectively. The following temperature profile was used for PCR amplification: an initial denaturing step at 93°C for 1 minute and 35 cycles of 93°C for 30 seconds, 56°C for 30 seconds, and 72°C for 1 minute. For Southern blot analysis, PCR products were electrophoresed and transferred to nylon Hybond N+ membranes (GE Healthcare). The membranes were hybridized with an internal and [32 P]-labeled oligonucleotide probe specific for TNFα or β actin cDNA. Membranes were scanned on a STORM phosphoimager. Messenger RNA expression was quantified with the IMAGE-QuaNT program (GE Healthcare). Results were normalized against β actin mRNA expression and are presented as percentages of the control values.

Western Blot Analysis of Hepatic TNFα and Adiponectin Plasma Levels.

Rat liver proteins were dissolved in a lysis buffer (50 mM Tris HCl [pH 7.5], 1% SDS, 10 mM EDTA, 100 mM NaCl, and 1% β mercaptoethanol) containing protease inhibitors (Roche Diagnostics) and centrifuged at 12,000 g for 15 minutes at 4°C. The supernatants were collected, and protein concentrations were determined using a BCA protein assay kit (Pierce). Liver protein extracts (200 μg) and plasma (0.2 μL) were analyzed by western blot on Novex precast 4%-20% Tris-glycine-SDS-PAGE gels as described38 using either goat anti-rat TNFα antibody (TEBU) or rabbit anti-rat adiponectin (Affinity Bioreagents) and horseradish peroxidase (HRP)–conjugated anti-goat or anti-rabbit antiserum (Sigma). Immunoreactivity was revealed with ECL-plus chemiluminescent substrate (GE Healthcare) and scanned on a Kodak Image Station 440 CF (Eastman Kodak). Hepatic TNFα and adiponectin plasma levels were quantified with 1D image analysis software (Kodak).

Statistical Analysis.

Results are expressed as means ± SEMs. Statistical differences between groups were analyzed by one-way ANOVA followed by the Student t test, and differences were considered significant at P < 0.05.

Results

Rimonabant Reversed Hepatic Steatosis in Obese Zucker (fa/fa) Rats.

To evaluate the effect of rimonabant on hepatic steatosis, we first investigated the effect of this compound on the liver/body weight ratio of obese (fa/fa) rats. Obese (fa/fa) rats showed hepatomegaly, characterized by a higher liver/body weight ratio (4.98% ± 0.15%) than that of their lean littermates (3.50% ± 0.18%). Once-daily treatment of obese (fa/fa) rats with rimonabant (30 mg/kg) orally for 8 weeks strongly decreased this ratio to a normal level (3.85 ± 0.09 %) comparable to that observed in lean rats. In contrast, submitting obese (fa/fa) rats to a pair-fed diet produced only a slight reduction in the liver/body weight ratio, which remained higher (4.51% ± 0.17%) than that of obese rats treated with rimonabant and of lean rats (Fig. 1A). Histological analysis of livers using a hematoxylin and eosin staining process revealed severe hepatic steatosis (fat liver infiltration) in the obese (fa/fa) rats, characterized by excessive accumulation of fat in hepatic intracellular vesicles (Fig. 1B). Once-daily treatment of the obese (fa/fa) rats with rimonabant completely eliminated this hepatic steatosis (Fig. 1B). Liver slices from the obese (fa/fa) rats treated with rimonabant were found to be histologically comparable to those from lean rats. In contrast, the severe hepatic steatosis persisted in the pair-fed obese (fa/fa) rats (Fig. 1B). Furthermore, rimonabant treatment of obese (fa/fa) rats strongly reduced the elevated plasma levels of liver injury enzyme markers ALT, GGT, and ALP. In contrast, submitting obese (fa/fa) rats to pair feeding did not affect the plasma levels of these enzymes, which remained high (Fig. 1C).

Figure 1.

Effect of rimonabant on fatty liver in obese (fa/fa) rats. (A) Liver-to-body ratio of lean rats treated with vehicle (n = 12), obese (fa/fa) rats treated with vehicle (n = 12) or vehicle supplemented with rimonabant (RIMO) at a dose of 30 mg/kg orally and daily for 8 weeks (n = 12), and pair-fed obese (fa/fa) rats (n = 12) was determined. Values represent the mean ± SEM of 12 rats; **P < 0.01 compared with group of obese (fa/fa) rats treated with vehicle. (B) Analysis of hepatic histology of representative rat from each treatment group. Liver sections were fixed in formalin, embedded in paraffin, stained with hematoxylin & eosin, and photographed. (C) Plasma levels of ALT, GGT, and ALP were determined as described in Materials and Methods. Values represent the mean ± SEM of 12 rats; **P < 0.01 compared with the group of obese (fa/fa) rats treated with vehicle.

Rimonabant Reduced Steatohepatitis-Associated High Hepatic TNFα Level in Obese Zucker (fa/fa) Rats.

To determine the effect of rimonabant on the high hepatic TNFα level currently associated with steatohepatitis (inflammatory state of hepatic steatosis), we analyzed the effect of rimonabant on the expression of mRNA and protein of hepatic TNFα, a proinflammatory cytokine that has been shown to be a key mediator of the progression of liver diseases to more serious forms.29–32 Expression of hepatic mRNA and TNFα protein were 3.5-fold and 2.5-fold higher, respectively, in the livers of obese (fa/fa) rats than in the livers of their lean littermates (Fig. 2). Treatment of the obese (fa/fa) rats with rimonabant strongly reduced expression of hepatic TNFα mRNA and protein. However, submitting the obese (fa/fa) rats to pair feeding did not affect the elevated levels of either hepatic mRNA or TNFα protein (Fig. 2).

Figure 2.

Effect of rimonabant on hepatic expression of TNFα. (A) Representative RT-PCR analysis of expression of TNFα mRNA. Total RNA isolated from livers of lean rats treated with vehicle, obese (fa/fa) rats treated with vehicle or vehicle supplemented with rimonabant (RIMO) at a dose of 30 mg/kg orally and daily for 8 weeks, and pair-fed obese (fa/fa) rats was analyzed by RT-PCR as indicated in the Materials and Methods section. TNFα expression, normalized against that of β actin was quantified using the IMAGE QuaNT program (Molecular Dynamics). The bar graph shows the results of the quantification analysis. Values represent the mean ± SEM of 3 independent experiments. (B) Representative Western blot analysis of hepatic TNFα level. Proteins extracted from the livers of lean rats treated with vehicle, obese fa/fa rats treated with vehicle or vehicle supplemented with rimonabant (RIMO) at a dose of 30 mg/kg orally and daily for 8 weeks, and pair-fed obese (fa/fa) rats were analyzed by Western blot using goat anti-rat TNFα antibody at a dilution of 1:250. Values represent the mean ± SEM of 3 independent experiments; **P < 0.01 compared with the group of obese (fa/fa) rats treated with vehicle. Each experimental point represents the mean ± SEM of 12 rats.

Rimonabant Reduced TNFα Plasma Level in Obese Zucker (fa/fa) Rats.

We then investigated the effect of rimonabant on the elevated plasma concentration of TNFα, that characterizes severe obesity and that has been reported to be involved in several metabolic disorders including obesity and associated diseases as well as cardiovascular complications.8, 18–20 Figure 3 shows that plasma TNFα level was 3-fold higher in obese (fa/fa) rats than in lean rats. Treatment of obese (fa/fa) rats with rimonabant induced a strong decrease in plasma TNFα level (55%) in comparison with that observed in control animals that received vehicle (Fig. 3). In contrast, submitting obese (fa/fa) rats to pair feeding did not affect the elevated levels of circulating TNFα (Fig. 3).

Figure 3.

Effect of rimonabant on TNFα plasma level. Plasma concentration of TNFα was determined using an ELISA kit (Biosource) in plasma samples collected from lean rats treated with vehicle, obese (fa/fa) rats treated with vehicle or vehicle supplemented with rimonabant (RIMO) at a dose of 30 mg/kg orally and daily for 8 weeks, and pair-fed obese (fa/fa) rats. Each experimental point represents the mean ± SEM of 12 rats; **P < 0.01 compared with the group of obese (fa/fa) rats treated with vehicle.

Rimonabant Improved Dyslipidemia in Obese Zucker (fa/fa) Rats.

Plasma lipid analysis revealed that in comparison to lean rats, obese Zucker (fa/fa) rats showed dyslipidemia, characterized by high plasma levels of triglycerides, free fatty acids, and cholesterol and by a low plasma HDLc/LDLc ratio (Fig. 4). Treatment of obese (fa/fa) rats with rimonabant clearly improved lipid disorders observed in this animal model. Indeed, rimonabant treatment of obese (fa/fa) rats strongly reduced plasma levels of triglycerides, free fatty acids, and cholesterol, which became comparable to those observed in lean rats (Fig. 4). Interestingly, rimonabant treatment of obese (fa/fa) rats increased the plasma HDLc/LDLc ratio to normal levels (Fig. 4). In contrast, submitting obese (fa/fa) rats to pair feeding had little or no effect on the plasma levels of these four lipids, demonstrating that dyslipidemia persisted under these conditions (Fig. 4).

Figure 4.

Effect of rimonabant on obesity-associated dyslipidemia. Lipid parameters ((A) plasma level of triglycerides, (B) plasma level of free fatty acids, (C) plasma level of total cholesterol, and (D) HDLc/LDLc ratio) were measured as described in the Materials and Methods section in plasma samples collected from lean rats treated with vehicle, obese (fa/fa) rats treated with vehicle or vehicle supplemented with rimonabant (RIMO) at a dose of 30 mg/kg orally and daily for 8 weeks, and pair-fed obese (fa/fa) rats. Each experimental point represents the mean ± SEM of 12 rats; **P < 0.01 compared with the group of obese (fa/fa) rats treated with vehicle.

Rimonabant Increased and Normalized Adiponectin Plasma Level in Obese Zucker (fa/fa) Rats.

We also investigated the effect of rimonabant on the reduced plasma level of adiponectin in obese (fa/fa) rats, which is characteristic of several metabolic disorders including obesity and metabolic syndrome-related diseases and their complications.10, 15–17, 20–22 Figure 5 shows that adiponectin plasma level was 40% lower in obese (fa/fa) rats than in lean rats. Treatment of obese (fa/fa) rats with rimonabant increased adiponectin plasma to a level comparable to that observed in lean rats (Fig. 5). In contrast, submitting obese (fa/fa) rats to pair feeding had only a slight, nonsignificant effect on the circulating level of adiponectin, which remained lower than that observed in lean rats (Fig. 5).

Figure 5.

Effect of rimonabant on adiponectin plasma level. Plasma derived from lean rats treated with vehicle, obese (fa/fa) rats treated with vehicle or vehicle supplemented with rimonabant (RIMO) at a dose of 30 mg/kg orally and daily for 8 weeks, and pair-fed obese (fa/fa) rats treated with vehicle were analyzed by Western blot using rabbit anti-rat adiponectin antibody in a 1:250 ratio. (A) Representative Western blot analysis of adiponectin plasma level. (B) Bar graph shows results of quantification analysis of adiponectin plasma level using Image Analysis Software (Kodak). Values represent the mean ± SEM of 3 independent experiments; **P < 0.01 compared with the group of obese (fa/fa) rats treated with vehicle. Each experimental point represents the mean ± SEM of 12 rats.

Discussion

Metabolic syndrome is a cluster of metabolic disorders including insulin resistance, type 2 diabetes, dyslipidemia, inflammation, hypertension, and obesity, in particular, visceral adiposity that significantly increase the risk of cardiovascular disease.43 Obesity is an inflammatory, chronic, and progressive disease, characterized by increased body weight and development of adipose tissue with excessive fat storage.2, 13, 14, 44 It is associated with dramatic deregulation of tissue and plasma levels of pro- and anti-inflammatory cytokines such as TNFα and hormones such as adiponectin.6, 7, 13, 14 Obesity is also a major cause of the emergence of several features of metabolic syndrome and their complications. Among these, nonalcoholic hepatic steatosis and steatohepatitis, major obesity-associated complications, have been proposed as mediators of metabolic syndrome in addition to insulin resistance in animal models.10, 20 At present, there is no pharmacological agent that has both anti-obesity effects and completely reverses and prevents obesity-associated steatohepatitis and features of metabolic syndrome and their complications, in particular, the inflammatory component associated with metabolic diseases.5, 13, 14, 44

The selective cannabinoid receptor type 1 (CB1 receptor) antagonist rimonabant has been reported to have potent anti-obesity effects33–42 that may partly be a result of its “peripheral” metabolic action on adipose tissue.38, 45 Therefore, we investigated the effect of rimonabant on hepatic steatosis and features of associated metabolic syndrome (inflammation, dyslipidemia, and low plasma levels of adiponectin) in obese Zucker (fa/fa) rats. Our results show that rimonabant treatment reduces hepatomegaly, completely eradicates hepatic steatosis (fatty liver), and decreases plasma levels of enzyme markers of liver damage (ALT, GGT, and ALP). Interestingly, rimonabant treatment strongly reduces elevated levels of hepatic TNFα that characterize the inflammatory state of fatty liver (the steatohepatitis).18, 19, 32 Elevated levels of hepatic proinflammatory cytokines, such as TNFα were suggested to both induce insulin resistance in liver and to be involved in the progression of steatohepatitis to hepatic fibrosis and cirrhosis.18, 19, 29–32 The reduction of hepatic TNFα levels by rimonabant may represent one mechanism by which rimonabant ameliorates the hepatic state, probably by restoring its structure and metabolic function, in particular insulin sensitivity. Complementary research is needed to elucidate this hypothesis. Additionally, rimonabant also reduces elevated plasma TNFα, which characterizes the chronic systemic low-grade inflammation associated with obesity and metabolic diseases.7, 13, 14, 44, 46 The reduction of plasma and hepatic TNFα levels by rimonabant may be one of the principal mediators of the reversion of hepatic steatosis and may also be a key effect leading to arrest of the progression of steatohepatitis to fibrosis and cirrhosis. These data reveal that rimonabant is hepatoprotective and suggest that this CB1 receptor antagonist may have a new therapeutic role in hepatic diseases. These findings are supported by a recent study that demonstrated steatohepatitis resistance in CB1 receptor knockout (KO-CB1) mice47 and confirm the role of the endocannabinoid system in liver diseases.48, 49

Furthermore, our results demonstrate that rimonabant improves dyslipidemia, a major biochemical disorder associated with steatohepatitis leading to cardiovascular diseases.50 Rimonabant treatment of obese (fa/fa) rats reduces plasma levels of cholesterol, free fatty acids, and triglycerides and, importantly, increases the HDLc/LDLc ratio. These results, obtained in an animal model, correlate with data recently reported from human clinical trials.40–42 This global improvement of the plasma lipid profile may be mediated in part by the restoration, following rimonabant treatment, of the structure and metabolic function of the principal organs and tissues involved in lipid and glucose metabolism, such as muscle, adipose tissue, and liver. Rimonabant's improvement of dyslipidemia may play an important role in its activity against obesity and metabolic syndrome and points to its cardiovascular protective role.

Furthermore, rimonabant treatment of obese (fa/fa) rats increases and normalizes plasma levels of adiponectin. This observation is in agreement with the result of our previous studies, which demonstrated that rimonabant increases expression of adiponectin mRNA in adipose tissue by directly acting on adipocytes via CB1 receptors expressed on this type of cell.38, 45 Adiponectin has been widely reported to reduce hyperinsulinemia and insulin resistance, which in turn improves insulin sensitivity, reverses steatohepatitis, and reduces concentrations of proinflammatory cytokines, particularly TNFα, so as to improve dyslipidemia, as well as to possess anti-inflammatory, cardiovascular-protective, and hepatoprotective properties.20–28 Deregulation of expression of adipose tissue and of plasma adiponectin has been shown to be directly related to the metabolic syndrome and its associated complications, in particular, steatohepatitis.7, 15–17 On the basis of these observations, regulation of adiponectin by rimonabant may be considered a key mechanism by which rimonabant exercises its anti-metabolic syndrome effects and hepatoprotective activity.

Finally, all the effects of rimonabant described in this study were not or only slightly observed in the pair-fed obese (fa/fa) rats, clearly demonstrating the beneficial and supplementary effects of treatment with rimonabant compared with just reducing food intake, in particular for the reduction of the inflammatory component associated with obesity and metabolic diseases. This inflammatory component is the principal cause of the chronic and progressive features that characterize metabolic diseases, which can lead to various fatal complications (hepatic, renal, cardiovascular, and cancer). This beneficial effect of rimonabant is probably related to its metabolic and peripheral activity, particularly to the restoration of the integrity of adipose tissue, characterized by the increased levels and normalization of plasma adiponectin. Our hypothesis is that the multiprotective effects of rimonabant may be mediated in large part by both the reduction in proinflammatory cytokines such as TNFα and the increase in protective anti-inflammatory cytokines and hormones such as adiponectin.

In summary, the data obtained from our animal model show that rimonabant reverses and prevents steatohepatitis and related features of metabolic syndrome and demonstrate that rimonabant plays a hepatoprotective role. This suggests that this CB1 receptor antagonist may have potential clinical applications in the treatment of liver diseases associated with obesity and metabolic syndrome.

Acknowledgements

We thank John Alexander for editing the manuscript and Dr. Daniel Mirouze for his helpful scientific discussions. This work is dedicated to Dr. Jean Pierre Tauber.

Ancillary