The present study was carried out to determine if oxidative stress was associated with changes in the expression of LFABP and NOX in a rat model of non alcoholic steatohepatitis and whether cocoa supplementation attenuated those changes. The results indicate an association between the MCD diet and levels of LFABP in the development of NASH in a well established model of the disease. Levels of LFABP mRNA and protein were significantly lower in animals on the MCD diet in comparison to animals on the MCS diet. Suppression of LFABP may be another mechanism by which this diet causes an increased fat content in the liver in addition to impairing phosphatidylcholine synthesis . Low levels of LFABP may lead to an inability of the hepatocyte to shuttle long chain fatty acids to different intracellular destinations for metabolism , resulting in higher levels of hepatic fat content in MCD animals as evident from the histological analysis (Figure 1; Table 4). Supplementation of MCD diet with cocoa in the C1 diet regime significantly increased levels of LFABP mRNA (Figure 2A), which we postulate leads to a restoration in trafficking of fatty acids within the hepatocyte; however this did not lead to a lower degree of observed steatosis (Table 4). Increased levels of LFABP may reduce oxidative damage by binding long chain fatty acids to its methionine residues . Low levels of LFABP in MCD fed animals may therefore result in increased oxidative damage due to its ability to act as an endogenous antioxidant . The increase in LFABP mRNA in the C1 diet regime (Figure 2A) showed a similar pattern at the protein level (Figure 2B). A decrease in LFABP may be linked to the liver's inability to cope with lipotoxicity, which is thought to contribute to NASH . LFABP has been found to be upregulated in the presence of long chain fatty acids and has been directly implicated in hepatic regeneration . This may be correlated to the effects of LFABP stimulation of PPAR-α to further increase LFABP mRNA. Findings in rat models indicate an increase in LFABP during hepatic regeneration, supporting the role of this protein in maintaining the integrity of the hepatocyte . LFABP deregulation, as shown by an inverse relationship between the ratio of LFABP and fat content in the liver, has been correlated with obesity and type 2 diabetes in the Israeli sand rat . This is further supported by a silencing of LFABP in patients with hepatocellular adenoma who had a mutation in the hepatocyte nuclear factor 1α, causing impaired trafficking of fatty acids, leading to steatosis . Since LFABP is an abundant protein in hepatocytes, it may provide a major source of intracellular antioxidant activity. Purified LFABP has been tested for its antioxidant capacity  and is able to quench up to 66% of free radicals generated from superoxide. This is in agreement with our findings of lower LFABP being present at both the mRNA level (Figure 2A) and protein level (Figure 2B) in animals with MCD derived fatty liver disease in comparison to the animals fed the MCS diet. In addition, higher levels of superoxide fluorescence and 8-isoprostane were evident in the MCD fed animals as compared to the MCS fed animals (Table 3 and 5; Figure 1M and 1N), further supporting an inverse association between levels of LFABP and levels of oxidative stress. However, supplementation with cocoa in the C1 and C2 diet regimes resulted in higher superoxide and 8-OH-2dG levels when compared to MCS animals. This may be related to higher degree of observed steatosis in these groups (Table 4). Slightly lower superoxide and 8-OH-2dG levels were seen when animals were on the C3 diet regime. This C3 cocoa group had lower levels of steatosis when compared to MCD, C1 and C2 diet regimes. Further to this, lower levels of lobular inflammation and fibrosis were observed in these groups. It cannot be concluded that the higher levels of superoxide seen in the cocoa supplemented diets are as a result of the cocoa instead of the MCD, as the animals supplemented with cocoa were on the MCD diet longer than the MCD control group, dependent on the time of cocoa supplementation.
The quantification of mRNA detected differences in the levels of NOX1 mRNA expression, but no change observed in NOX2 and NOX4 mRNA expression between the different diet regimes. NOX1 mRNA expression levels were lower in all groups fed the MCD diet in comparison to those on the MCS diet (Figure 3A). The effect of the dietary regimes on NOX1 protein levels was different to that of mRNA expression levels (Figure 3B), indicating that NOX1 may be regulated at the protein level, rather than the gene level. Higher concentrations of NOX1 protein were observed in animals on the C2 diet regime. Gene knockout of gp91
, a vital regulatory component of the assembly of NOX, showed no difference in the pathology of MCD induced NASH in mice compared to wildtype . This would indicate that NOX generation of ROS is not a key factor in the development of MCD induced NASH, which is supportive of our findings in NOX mRNA expression. However, a link can be seen between NOX1 protein levels and presence of portal inflammation in animals on the C2 diet regime, with higher NOX1 levels measured and a greater proportion of portal inflammation observed in comparison to rats on the other diets. In alcoholic liver disease, mice fed ethanol via the Tsukamoto-French intragastric enteral method, NOX was found to increase ROS and activate NF-κB, which led to an increase in TNF-α in livers. This leads not only to an increase in oxidative damage but also an increase in synthesis of fatty acids causing hepatic damage .
Histological analysis of livers from rats fed the MCD diet showed greater steatosis in comparison to those on the MCS diet (Figure 1). Steatosis has been reported by others at week 2 of MCD feeding in rat livers . The severity of steatosis was not observed to be less in any of the groups in which cocoa was added to the MCD diet, however there was a statistically significant lower degree of steatosis observed in livers of animals fed the C3 diet regime. It is extrapolated from this observation that the antioxidant properties of cocoa are more likely to affect levels of reactive oxidative species rather than hepatocyte fat content. This is supported by a lower level of ROS as determined by DHE staining and 8-OH-2dG in the C3 diet regime when compared to C1 and C2 diet regimes (Table 5). Antioxidants derived from cocoa may play a role in suppressing the activation of hepatic stellate cells to form fibrotic tissue, as fibrosis was not as severe in the animals on the C3 diet regime, a group which had lower scores for steatosis and lobular inflammation compared to other MCD and MCD/cocoa regimes (Table 4).
Circulating triglyceride levels were lower in the the MCD group compared to the control. However cocoa supplementation was associated with even lower circulating triglyceride levels (Table 5). Re-esterification of fatty acids into triglycerides has been described as a mechanism protecting the liver from lipotoxicity as inflammation, oxidative damage and fibrosis decrease . Lower levels of circulating triglycerides (Table 5) found in our study are in line with increased severity of NAFLD as shown by increased steatosis scores in Table 4. The reduction in body weight on MCD possibly led to an increase in glucose being used as an energy source causing a reduction in the circulating levels of glucose (Table 5). The MCD diet has been previously reported to decrease glucose and improve insulin sensitivity whilst not having a dampening effect on the development of hepatic inflammation or fibrosis . Although the MCD diet caused weight loss, liver weight increased as a result of higher fat content as seen in the histology of these samples (Figure 1; Table 4).
RBC GSH levels were significantly higher in the C1 and C2 groups (Table 5). This suggested that cocoa could be used to increase the availability of the reduced form of GSH to act as an antioxidant within RBC's and possibly the circulation. Liver GSH on the other hand was much lower in all cocoa supplemented animals when compared to those on the MCS and MCD diets. Low levels of this endogenous antioxidant in the cocoa supplemented animals may be due to the higher bioavailability of exogenous antioxidants derived from the cocoa. The accumulation of exogenous antioxidants from cocoa may therefore be beneficial in providing sufficient antioxidants to quench ROS in NASH. Our findings on hepatic GSH are not in agreement with most other studies which show a depletion of this endogenous antioxidant .
Despite the novel data presented from the current study there are limitations associated with the findings. Due to restrictions imposed by the institutional animal welfare committee it was not possible to include additional MCD fed rats for 80 and 108 days to match cocoa supplementation groups C1 - C4. Although pilot data indicated histologically the livers of rats fed the MCD diet are similar from 42 - 112 days, it cannot be excluded that the effects associated with cocoa supplementation in the liver are not to prolonged MCD feeding. It is possible, but unlikely, that the results observed following cocoa supplementation are not due to the antioxidants present in the cocoa, but rather the trace amounts of methionine and choline present in the cocoa. However if the trace amounts of methionine and choline present in the cocoa were responsible for the results observed it would be expected that data collected from the cocoa supplemented groups would more closely resemble the MCS group and not the MCD group. Finally although the MCD diet is a commonly used model of NASH there are a number of limitations associated with comparing the model to metabolic changes in human NAFLD/NASH . These limitations include weight loss in rats fed the MCD diet, whereas NASH patients are typically overweight or obese [1, 7]. The accumulation of fat within the liver of rats fed the MCD diet is due to a disruption of the export of hepatic lipids and subsequent lipotoxicity, unlike the human situation where the excessive hepatic fat import or storage is thought to occur [1, 7].