Non-alcoholic fatty liver disease (NAFLD) is a progressive disease associated with insulin resistance and obesity. The spectrum of NAFLD ranges from simple steatosis to non-alcoholic steatohepatitis (NASH) characterized by steatosis, inflammation, hepatocyte injuries and progressive fibrosis. No treatment has been proven efficacious except for lifestyle modifications coupling physical exercise with weight reduction. Better understanding NASH pathogenesis should light on new mechanisms involved in NAFLD progression and thereby on new targets to counteract NAFLD development. High-fat diet (HFD)-fed foz/foz mice, a strain of mice carrying a mutation in a ciliary protein, represent a unique mouse model of NASH as they develop both the metabolic context and the histological features of human NASH. Their unique metabolic phenotype has been linked to hyperphagia resulting from abnormal ciliary function in the central nervous system but the links between obesity, metabolic syndrome and fibrosing NASH are not elucidated. My aim was thus to unveil mechanisms contributing to the phenotype in this model. In this doctoral thesis, we performed a pair-feeding experiment to determine whether the metabolic and liver alterations in foz/foz mice were only related to an increase in food intake. Caloric restriction failed to protect foz/foz mice against obesity, adipose inflammation and glucose intolerance. Obese foz/foz mice had similar physical activity level but reduced energy expenditure compared to their wild-type (WT) littermates. We identified an altered thermogenic adaptation to HFD or a cold exposure in foz/foz mice compared to HFD-fed WT littermates hence decreasing energy substrate utilization. Defective adaptive thermogenesis was due to a lower sympathetic tone in brown adipose tissue (BAT). Sympathetic stimulation by intermittent cold exposure restored BAT function in foz/foz mice and thereby improved glucose tolerance, adiposity and hepatic steatosis. Next, in interventional experiments, we evaluated whether increasing non-shivering thermogenesis could prevent and/or improve pre-existing NASH. In foz/foz mice with metabolic syndrome and liver steatosis, β3 adrenergic receptor (β3AR) agonist improved BAT function and induced browning of white adipose tissue. Increased thermogenic capacity was associated with a better glucose tolerance, a decreased NAFLD activity score and decreased transaminases levels with no change in body weight. When initiated after the onset of NASH in foz/foz mice, β3AR agonist treatment restored BAT function and increased glucose tolerance but had no impact on liver pathology compared to untreated mice. Similarly, β3AR agonist had no therapeutic effect when administrated for 4 weeks on methione and choline deficient diet-induced NASH neither in C57Bl6 nor in obese and diabetic db/db mice. As boosting BAT activity had no impact of body weight loss, we reasoned that increased energy substrate mobilization through weight loss therapy could divert lipids towards burning in the BAT and improve NASH. When coupled with caloric restriction, β3AR agonist enhanced weight loss and glucose tolerance compared to mice with caloric restriction only. In addition, the combined therapy, but not food restriction alone, decreased hepatic fat content, hepatocyte ballooning and NAFLD activity score, thus improving pre-established NASH. In conclusion, our findings indicate that failure of BAT adaptation drives the metabolic complications of obesity in foz/foz mice, including development of NAFLD. Increasing thermogenic capacities prevents NAFLD progression but is not effective to cure pre-existing NASH. Nonetheless, when coupled with weight loss therapy, BAT stimulation provides additional therapeutic advantage in NASH management and opens new possibilities to improve liver histology in NAFLD patients.