In the wake of the global obesity epidemic, the incidence of non-alcoholic fatty liver disease is increasing. The condition covers the spectrum from steatosis to steatohepatitis, fibrosis and cirrhosis. Classically, the pathogenesis is simplified to a two-hit model, involving development of hepatic insulin resistance leading to accumulation of triglycerides and to alteration of intracellular metabolism making the liver sensitive to oxidative stress as the first hit [2, 13]. The second hit driving the steatosis to NASH involves overload of free radicals arising from β-oxidation of free fatty acids in the steatotic hepatocytes as well as effects of adipokines secreted from white adipose tissue . Although several pharmacological agents are being evaluated, no pharmacological treatment is currently approved for the treatment of NASH, leaving weight loss and lifestyle changes as the only treatment in the routine clinical practice [5, 13].
Roux-en-Y gastric bypass is gaining acceptance as an effective treatment of morbidly obese patients, and is surprisingly effective in treating morbidly obese type 2 diabetic patients. Improved glucose tolerance is observed merely days after surgery and hence not correlated to the weight loss. In addition, the procedure also seems to reverse steatosis, steatohepatitis and fibrosis in the liver . As a result of the re-routing of the flow of chyme through the intestine, the secretion pattern of gut hormones is changed. Particular glucagon like-peptide 1 (GLP-1) and Peptide YY (PYY) are markedly elevated in response to nutrients . Glucagon like-peptide 1 is an incretin hormone released from the endocrine L-cell in the intestine in relation to meal ingestion. Under normal conditions, the hormone stimulates insulin secretion from the β-cells and inhibits glucagon secretion from the α-cells in a glucose-dependent manner . Besides effects on the endocrine pancreas, endogenous GLP-1 has been demonstrated to play a physiological role in the regulation of gastro-intestinal secretions and gastric emptying. The effects of GLP-1 are mediated by specific binding of GLP-1 to the GLP-1 receptor (GLP-1r). This receptor belongs to the secretin family (type 2) of G-protein coupled receptors, and the downstream signalling is mediated (exclusively?) by an increase in intracellular cAMP. Patients suffering from type 2 diabetes have decreased incretin effect resulting from a decreased effect of both of the incretin hormones (the other one is glucose-dependent insulinotropic polypeptide, GIP) on the β-cell . However, unlike GIP, pharmacological doses of GLP-1 are still capable of restoring glucose-induced insulin secretion in patients suffering from type 2 diabetes . The beneficial effects of GLP-1 have been exploited in the development the incretin or GLP-1 mimetics, a whole new treatment strategy for type 2 diabetes, based on substitution with GLP-1r agonists . Treatment with GLP-1r agonists induces a marked weight loss in obese type 2 diabetic patients  as well as in non-diabetic obese subjects . In addition, it is evident that treatment of type 2 diabetic patients with GLP-1r agonist improves liver status as assessed by biochemical markers .
Originally, expression of GLP-1r was not mapped to the liver [3, 11], but this classical view was recently challenged by the observation in ob/ob mice  that 60 days of treatment with GLP-1 reduced liver content of lipids in ob/ob mice. In addition, presence of functional GLP-1 receptors were demonstrated using Western blots obtained from isolated hepatocytes, and GLP-1 receptor agonist were demonstrated to induce cAMP in primary hepatocytes. Furthermore, Gupta et al.  demonstrated GLP-1r in primary culture of human hepatocytes. In this issue of Liver International, Svegliati-Baroni et al. present data of GLP-1r expression in liver biopsies from patients undergoing hepatic resection for focal nodular hyperplasia or hepatic adenoma and in liver biopsies from patients suffering from NASH . Furthermore, the expression of GLP-1r in the biopsies from patients suffering from NASH was generally lower compared with expression in biopsies from the other patient categories. To gain enough cells to conduct analysis of cellular events downstream of the GLP-1R activation, the authors turned to primary cultures of hepatocytes isolated from rats with diet-induced NASH.
Treatment of the primary hepatic cell cultures with the GLP-1r agonist exenatide induced higher expression of mRNA encoding the transcription factors, peroxisome proliferator-activated receptor alpha (Ppara) and peroxisome proliferator-activated receptor gamma (Pparg), as well as increased expression of the mRNA encoding the two key enzymes involved in both mitochondrial and peroxisomal β-oxidation of free fatty acids (FFA): carnitine palmitoyltransferase 1A (Cpt1a) and peroxisomal acyl co-enzyme A oxidase 1 (Acox1). In hepatocytes, PPARγ is generally believed to be involved in the regulation of synthesis of fatty acids and storage of lipids . It was recently shown that Pparg is up-regulated in biopsies from patients with steatosis , and in ApoB/BAT less mice, a genetically modified mouse model of obesity, insulin resistance and steatosis, exhibit increased hepatic expression of Pparg compared with wildtype mice . Furthermore, downregulation of the hepatic Pparg expression reverses the steatotic phenotype in this mouse model . However, another study applying a gene rescue approach to PPARγ deficient mice showed that rescue of Pparg expression in both liver and adipose tissue reversed the steatohepatitis induced by methionine and choline deficient diet , indicating that PPARγ should be activated in both adipocytes as well as hepatocytes to reverse steatosis. Though stimulation of hepatocytes in primary culture with exenatide results in upregulation of Pparg, it is unclear whether this results in storage of lipids and synthesis of fatty acids, because the stimulation also upregulates adenosine monophosphate kinase (Ampk). PPARα and Ampk are involved in regulation of oxidation of free fatty acids [4, 17]. On the basis of these observations, one could speculate whether GLP-1 actually induces a futile circle burning fat in the liver. The observations of up-regulation of enzymes involved in the oxidation of lipid also raise concerns whether GLP-1 treatment, actually, as a consequence of increased oxidation of FFA, might induce increased oxidative stress in the hepatocytes. This important question was not addressed by the authors, but in ob/ob mice, Ding et al.  showed that GLP-1 actually deceased markers of oxidative stress in the hepatocytes.
The classical view that the liver is not directly influenced by GLP-1 is challenged by the work from Svegliati-Baroni et al. presented in this issue. The authors present data showing expression of GLP-1R in liver biopsies from humans, and show that GLP-1 regulates expression of transcription factors and enzymes involved in the hepatic metabolism of lipids using hepatocytes isolated from rats with diet-induced NASH. Together with recent observation also showing expression of GLP-1R in the liver of mice and humans, the classical view might need to be revised, as it appears that at least a subset of hepatocytes expresses GLP-1R and are directly stimulated by GLP-1.
On the basis of the know physiological effects of GLP-1 in type 2 diabetic patients, it is possible that patients suffering from type 2 diabetes and NASH might benefit from treatment with incretin mimetics. The data presented by Svegliati-Baroni et al. indicate that incretin mimetics also could be helpful in the small group of patients only suffering from NASH, but this interesting question needs more attention in future studies.