Results of our study clearly showed that HFD induced lipid peroxidation and nitrosative stress in the liver that are associated with steatosis and portal inflammation (Figure 5(c)). The role of oxidative stress in the pathogenesis of NAFLD is well known and is bidirectionally linked with inflammation [5–8]. Lipid peroxidation along with cytokine production and Fas ligand induction causes hepatocyte injury and apoptosis [7]. Besides, activated Kupffer cells and other inflammatory cells produce ROS and aggravate hepatic lipid peroxidation [8]. Additional sources of ROS in NAFLD include CYP2E1 induction by free fatty acids [4, 44], mitochondrial dysfunction, and increased NADPH oxidase activity [44, 45], as well as accumulation of ceramides within the hepatocytes [46]. To eliminate higher HFD food intake as a possible reason for ROS hyperproduction within the hepatocytes, we have measured food intake daily. In our study food intake could not be contributing factor which influenced level of hepatic oxidative/nitrosative stress (Figures 1 and 2).

Our study suggests that endocannabinoids also aggravate oxidative stress in NAFLD and contribute to the development of nonalcoholic steatohepatitis (NASH) through CB1 receptors. In fact, this is the first study that deals with effects of endocannabinoids and the selective CB1 antagonist, rimonabant, on oxidative/nitrosative stress in the liver. Namely, prooxidative effects of endocannabinoids are suggested by reduction of lipid peroxidation in both standard diet- and HFD-fed animals after rimonabant treatment (Figure 2). However, the effect of rimonabant was more prominent in HFR group and was associated with reduction of inflammation (Figure 5(d)). Previous preclinical and clinical cardiomiopathy [35] and nephropathy [25] studies have shown anti-inflammatory and cytoprotective effects of CB1 antagonists, due to reduction of tumor necrosis factor-alpha (TNF-α) production and inactivation of nuclear factor-kappaB (NF-κB). Similarly, our study clearly suggests that anti-inflammatory effect of rimonabant may be additionally mediated by decline in lipid peroxidation. Besides, this finding is in accordance with higher level of ANA and upregulation of CB1 receptor in HFD-induced NAFLD [20].

In present study, HFD caused an increase in NO_x content compared to control (Figure 2(b)). This finding is in agreement with reports demonstrating a link between HFD and nitrosative liver injury [45, 47–52]. It is known that an increased dietary supply of fat to the liver may promote steatosis by increasing hepatic lipid uptake. In this pathway, oxidative/nitrosative stress appears to play a key role in the pathogenesis of NASH [52], and nitric oxide (NO) may potentiate cytotoxicity through its reaction with superoxide anion and formation of nitrotyrosine [47, 48]. Furthermore, there is a relationship between abnormal hepatic accumulation of nitrotyrosine and pathohistological severity of NASH [47]. Besides, a strong expression of inducible nitric oxide synthase (iNOS) is observed in the liver from mice fed a HFD [52]. In addition, in our study inflammation could contribute to nitrosative stress in HF group. It is well established that hepatic inflammation is a complicated condition that is caused by various factors, including oxidative/nitrosative stress [2]. In such circumstances, activation of inflammatory cells in the liver generates ROS and RNS. Thus, peroxynitrite as a potent RNS not only can react with membrane phospholipids, DNA, and proteins, but also can produce cell damage via indirect radical-mediated mechanisms [53]. Although oxidative/nitrosative stress may not initiate hepatic inflammation, ROS and RNS overproduction could cause hepatocyte injury or death and, in turn, cytokine release that provides positive feedback on inflammatory signaling and promotes NASH pathogenesis [2]. It has revealed that NF-κB is essential for inflammatory cell recruitment in NASH [54] and induction of the iNOS gene in different cell types including macrophages [52, 54]. Also, there is new evidence that HFD may lead to mitochondrial dysfunction and increased oxidative/nitrosative stress in liver and skeletal muscle [45, 50, 51].

The induction of iNOS in the liver seemed to correlate with fatty changes. Namely, the expression of iNOS is upregulated by most inducers of obesity, hyperglycemia, and insulin resistance, including free fatty acids [49, 52]. However, it is not clear how iNOS in liver with steatohepatitis interacts to expand hepatic lipid stores, causes hepatocellular injury, and recruits inflammation [52]. Our study indicates that rimonabant treatment induced a significant decrease in NO_x content in HFR group in comparison with HF group (Figure 2). Possible explanation for this result refers to anti-inflammatory and cytoprotective effects of CB1 pharmacological inhibition or genetic deletion observed in the numerous preclinical and clinical reports [9, 11–13, 15–21, 23–29, 31, 33–36]. Since increased oxidative/nitrosative stress and inflammation are known to trigger increased endocannabinoid production or impair endocannabinoid inactivation [55–57], it is likely that endocannabinoids contribute to HFD-induced liver damage in present experimental model of NAFLD by promoting expression of iNOS in hepatic parenchymal cells and/or peripheral blood mononuclear cells through the activation of CB1 receptors [58–61]. Additionally, our result may be related to published data that highlight the importance of CB1 receptor blockade in protection against the pathological consequences of a HFD in the liver [11, 16, 20, 25–27, 29].

GSH, as a multifunctional, intracellular, nonenzymatic antioxidant, is the major component of intracellular regulation of the redox state and provides the first line of defense against oxidative/nitrosative injury. Moreover, GSH is the important substrate and cofactor in drug metabolism [62]. However, abnormally large concentrations of ROS/RNS may lead to permanent changes in signaling mechanisms that respond to alterations in the thiol/disulfide redox state [63]. Our study demonstrated that HFD induced a significant depletion in GSH liver content (Figure 4). Consistent with previous reports [2, 6, 45, 48, 50–52], this finding indicates that HFD shifts the redox state towards prooxidant. Additionally, depletion in GSH liver content obtained from current study is accompanied with increased lipid oxidative damage (Figure 2) and taken together they confirmed the existence of oxidative/nitrosative stress in mice treated with HFD [52]. Besides, possible iNOS suppression may additionally decrease GSH level in the liver, having in mind that GSH can conjugate with NO to form a S-nitroso-glutathione [64, 65].

The results of our investigation found that liver GSH level was significantly higher in HFD fed mice treated with rimonabant compared to HFD fed mice, as well as in rimonabant treated mice in comparison with control (Figure 4). This rise in GSH level can be explained by an adaptive response of hepatocytes to increased ROS production. It also suggests that CB1 receptor blockade could have a beneficial effect on the redox state in hepatocytes.

SOD, as the principal enzyme in the antioxidant defense of cells, is involved in the conversion of superoxide anion into a less toxic hydrogen peroxide. It is present in hepatocytes in two isoforms, Cu/ZnSOD (cytosolic SOD) and MnSOD (mitochondrial SOD) [7]. In our study an increase in total hepatic SOD activity was observed in HFD fed mice (Figure 3). In addition, the activity of both isoforms of this enzyme was found to be increased in response to HFD treatment (Figure 3). This rise may be interpreted as an adaptive response of hepatocytes to increased ROS production induced by HFD. These findings correspond to the results of other studies indicating that liver is the organ with most prominent increase in the activity of SOD after exposure to HFD [4, 6, 8, 45, 50–52]. An increase in the activity of MnSOD obtained in our study is not surprising, since mitochondria are the most important organelles in ROS production [45]. Moreover, MnSOD has ability to limit the peroxynitrite formation [66].

In present study higher activity of total SOD and its isoenzymes was registered in liver of rimonabant-treated mice (Figure 3). Given the role of SOD, as a major defense system against superoxide, it seems that increase in its activity upon the action of rimonabant improved antioxidant defense system in the liver and makes hepatocytes less sensitive to oxidative/nitrosative damage. This result is similar to the documented data which reports that blockade of CB1 receptors may exert beneficial effects in renal and cardiac diseases associated with oxidative/nitrosative stress [25, 35, 36].

In the present study total hepatic SOD activity and its isoenzymes were significantly decreased in HFD fed mice treated with rimonabant in comparison with mice treated with HFD (Figure 3). This decrease of total hepatic SOD activity, as well as depletion in both mitochondrial and cytosolic SOD activity, indicates that blockade of CB1 receptors may modulate liver antioxidant capacity without influence on SOD activity.

In the current study, HFD caused an increase in iron liver content when compared to control values (Table 2). This result is compatible with most reports which support the hypothesis that interactions between hepatic iron and lipid metabolism play a role in NASH/NAFLD pathophysiology [67–70]. Accordingly, iron is an integral part of some enzymes and transporters involved in lipid metabolism and, as such, may exert a direct effect on hepatic lipid load, intrahepatic metabolic pathways, and hepatic lipid secretion. On the other hand, iron in its ferrous form may indirectly affect lipid metabolism. Namely, iron overload can generate oxidative stress and lipid peroxidation, which may modify the fatty acid profile of cellular membranes, leading to their disruption, damage to cell organelles, and impairment of mitochondrial oxidative metabolism [68]. It is suggested that the free radicals that form may cause a change in the ratio of saturated to unsaturated membrane phospholipids, leading to alterations in membrane fluidity [63]. This in turn may affect the activity of the embedded enzymes, so they respond to oxidative stress differently by altering their activity [71]. Besides, iron accumulation has a proinflammatory and profibrogenic role by activating Kupffer cells to release inflammatory cytokines and by activating hepatic stellate cells, which can culminate in the replacement of parenchymal tissue with connective tissue [72].

Our study has shown that administration of rimonabant to HFD fed mice reduced hepatic iron content in comparison with mice fed with HFD (Table 2). Since upon high fat feeding iron overload and CB1 receptors may be implicated in hepatocellular injury, our finding suggests that beneficial effects of CB1 receptor antagonism on iron-associated oxidative stress may result from improvement of iron metabolism disturbances.

The result of the present study was found a significant decline in zinc content in mice fed with HFD compared to control values (Table 2). This decline is in agreement with published data that indicate the ability of zinc to modulate hepatic gene expression and lipid homeostasis [73]. Zinc finger transcription factors, peroxisome proliferation activator receptor-α (PPAR-α) and hepatocyte nuclear factor-4α (HNF-4α), play important roles in regulation of hepatic lipid metabolism [73, 74]. However, excessive lipid accumulation in the liver is associated with the formation of ROS and release of proinflammatory cytokines [7, 75]. In such circumstance, hepatic accumulation of fat may impaire expression of zinc-finger proteins [75]. In addition, dysfunction of zinc-importing proteins by proinflammatory cytokines and oxidative stress reduces plasma zinc concentration, and zinc deficiency can thus exaggerate hepatic lipid accumulation and inflammatory response [73, 76]. Besides, dietary zinc deficiency has been shown to induce hepatic lipid peroxidation [77]. Also, reduced hepatic zinc levels have been correlated with the impaired liver function and regeneration in a great number of acute and chronic liver diseases [76], including NAFLD [74, 76]. Thus, as expected, zinc deficiency increased hepatic triglyceride, cholesterol, and levels of free fatty acids in association with downregulation of PPAR-α and lipid metabolism genes [73].

A growing body of evidence indicates that zinc offers a protection from acute and chronic liver injury, but its hepatoprotective properties have not been fully identified [73, 76, 78–80]. In current research zinc content in both rimonabant treated groups of mice was significantly increased compared to control values and mice fed on HFD (Table 2). This finding suggests that the pharmacological CB1 inactivation in hepatocytes in some manner may contribute to zinc protective effects, having in mind that this essential trace element plays a critical role in cellular integrity and biological functions [76].

Results in our study showed that copper and manganese content was significantly higher in HFD fed mice in comparison with control (Table 2). These findings may be related to published data that put into attention the development of NAFLD as a potential contributor to the disruption of homeostasis of copper, zinc, and selenium [74, 78]. Related to this, NAFLD in hepatitis C virus patients is associated with oxidative damage, likely due to imbalances in trace metals [74].

In our study significant increase in copper and manganese content was also observed in HFD fed mice after rimonabant treatment compared to mice fed on HFD, as well as in rimonabant treated mice in comparison with control group. On the other hand, in contrast to copper content, manganese content in rimonabant treated mice was decreased when compared to control values (Table 2). Judging from these findings, it seems that CB1 receptor blockade also may influence on hepatic copper and manganese status.

Rimonabant was the first drug developed to block CB1 receptors in the brain and in the periphery [81]. The effects of this drug on human body weight reduction and the metabolic profile were very encouraging [82] and so four related large clinical trials were launched to test the long-term effects of rimonabant, including its efficacy and safety [83–86]. However, rimonabant has subsequently been withdrawn due to serious side effects, such as nausea, arthralgia, diarrhea, dizziness, anxiety, depression, and suicidal ideation [87]. Although there is no future for rimonabant in human health, our study indicates that CB1 receptor antagonists may be useful in NAFLD treatment. It is still unclear which specific CB1-signaling pathways should be activated to produce the desired therapeutic effects versus which should be excluded to reduce the unwanted side effects [88].

In recent years, it becomes clear that G-protein-coupled receptor antagonists may have agonistic effects on other signaling pathways related to their receptors [88, 89]. Thus, this phenomenon, known as biased agonism, may be also responsible for beneficial effects of rimonabant in current research.