3.1. Diet-Induced Weight Gain, Insulin Resistance and NAFLD
Mice received either standard chow diet (STD), chow diet enriched with 0.75% cholesterol (CHO + STD), a soybean oil-based high fat diet with 0.75% cholesterol (CHO + SOY) or a lard-based high fat diet with 0.75% cholesterol (CHO + LAR) for 20 weeks, as described in Table 1
. Animals on both high fat diets gained more weight than animals fed either chow diet or cholesterol-enriched chow diet (Figure 1
A). The high fat diet-induced weight gain could be attributed to an increase in fat mass (Figure 1
B) while the fat-free mass remained largely unaltered. Despite similar weight gain and increase in fat mass, animals fed the CHO + LAR diet were significantly more insulin resistant than animals receiving CHO + SOY diet (Figure 1
C). As expected from the body weight data, CHO + STD-fed animals showed no signs of insulin resistance.
Serum cholesterol levels increased only slightly (20%) in animals receiving the CHO + STD diet (Figure 2
A). By contrast, serum cholesterol concentrations were doubled in comparison to the control in animals receiving either one of the high fat diets with cholesterol. Notably, no difference in serum cholesterol levels was observed between CHO + SOY- and CHO + LAR-fed animals (Figure 2
Unexpectedly, but in keeping with data of many independent studies in the literature [11
], serum triglyceride levels were not elevated but instead were decreased in animals receiving either one of the cholesterol-enriched diets, irrespective of their fat content (Figure 2
B). Total cholesterol was increased in livers of all animals receiving cholesterol-enriched diets. However, whereas CHO + STD and CHO + LAR-fed animals showed a similar approximately 2 to 3-fold increase in hepatic cholesterol content, animals receiving CHO + SOY diet exhibited a 6-fold increase in hepatic total cholesterol content (Figure 2
C). Notably, free cholesterol was not significantly increased in CHO + STD-fed or CHO + LAR-fed animals in comparison to STD-fed animals, whereas free cholesterol content was doubled in CHO + SOY-fed mice (Figure 2
C). In line with this, high amounts of cholesterol crystals could be detected only in livers of CHO + SOY-fed mice whereas only few or no cholesterol crystals were visible in livers of CHO + STD or CHO + LAR-fed mice (own observation).
Although weight gain was unaltered in animals receiving CHO + STD diet, these animals had a pronounced hepatic steatosis (see below, Figure 4
). Hepatic triglyceride content increased more than twofold in comparison to chow-fed animals (Figure 2
D). Hepatic triglyceride accumulation was more pronounced in animals receiving cholesterol-enriched high fat diets. Livers of CHO + SOY-fed and CHO + LAR-fed animals contained 7-fold or 5-fold more triglycerides than STD-fed animals, respectively (Figure 2
D). The difference between the two high fat diets was, however, not significant.
In summary, mice fed a CHO + SOY diet accumulated significantly higher amounts of free and esterified cholesterol in the liver compared to mice fed one of the other cholesterol-containing diets. Since the CHO + STD, CHO + SOY and CHO + LAD diets contained equal amounts of cholesterol, the combination of dietary cholesterol and ω6-PUFA-rich soybean oil may favor hepatic cholesterol accumulation.
The more pronounced increase in hepatic cholesterol content can either be the consequence of an enhanced uptake or a diminished excretion or conversion of cholesterol. In accordance with the latter hypothesis, the expression of the cholesterol export pump Abcg5 was induced more than fourfold in animals receiving either CHO + STD or CHO + LAR diet (Figure 3
A). By contrast, the export pump was induced merely twofold, and hence significantly less, in animals receiving CHO + SOY diet than in either of the other two cholesterol containing diets (Figure 3
A). Gene expression of Abcg8, the heterodimerization partner of Abcg5, was similar, yet it did not reach significance (Figure 3
B). In comparison to the standard chow diet, Abca1, another cholesterol transporter mainly expressed in macrophages, was induced approximately 1.32-fold in livers of mice fed any of the cholesterol-containing diets. The increase was significant only in the CHO + STD diet group and no significant differences between the cholesterol-fed groups were observed.
In addition, the expression of Cyp27a1, a key enzyme for the conversion of cholesterol into bile acids, was significantly repressed in livers of CHO + SOY-fed mice (Figure 3
C). Similarly, Cyp7a1 was repressed in CHO + SOY-fed animals, whereas it was unaffected or even induced in CHO + LAR diet and CHO + STD diet-fed animals, respectively (Figure 3
D). Gene expression of the LDL receptor was repressed to a similar extent in livers of animals receiving any of the three cholesterol-containing diets (Figure 3
E). By contrast, the expression of the LDL receptor related protein 1 (Lrp1) was slightly or significantly reduced in livers of animals receiving the CHO + STD or CHO + LAR diets, whereas expression was unaltered in livers of CHO + SOY-fed animals (Figure 3
In conclusion, the enhanced cholesterol accumulation in livers of CHO + SOY-fed mice can be explained by a decreased Abcg5-mediated cholesterol export, reduced Cyp27a1-dependent conversion of cholesterol into bile acids, as well as impaired repression of Lrp1-related cholesterol uptake into hepatocytes.
Following, livers were examined histologically to determine the NASH activity score (NAS) (Table 2
and Figure 4
). No signs of NAFLD were detected in livers of STD-fed control animals. By contrast, all animals receiving cholesterol-enriched diets had a positive NAS. However, while the average NAS for CHO + STD-fed and CHO + LAR-fed animals reached a maximum of 4 and hence indicated the presence of blunt steatosis, the average NAS of CHO + SOY-fed mice was above 7, clearly indicating the presence of active NASH (Table 2
3.2. Diet-Induced Inflammation and Fibrosis
All cholesterol-containing diets apparently triggered an inflammatory response in the liver. However, in accordance with the higher NAS, animals receiving the CHO + SOY diet showed more pronounced signs of inflammation (Figure 4
right panel, quantification in Figure 5
B). The expression of the chemotactic cytokine Ccl2 (Mcp-1) was increased about two-fold in animals receiving CHO + STD or CHO + LAR diets, whereas an almost 10-fold increase was observed in CHO + SOY diet-fed animals (Figure 5
A). Consequently, the expression of the macrophage markers F4/80, Cd68, and Cd11b was increased by all cholesterol-containing diets, but was significantly higher in CHO + SOY-fed animals than in animals receiving any of the other diets (Figure 5
B–D). Similarly, the induction of the pro-inflammatory cytokine TNF-α was two-fold higher in livers of CHO + SOY-fed mice than in animals that received the CHO + STD or CHO + LAR diet (Figure 5
E). Inducible nitric oxide synthase (Nos2, iNos), a key enzyme in inflammation-dependent NO production was only induced in livers of CHO + SOY-fed animals (Figure 5
F). Furthermore, a significantly higher amount of TUNEL-positive hepatocytes were detected in livers of CHO + SOY-fed mice compared to mice fed any of the other diets, showing increased hepatic apoptosis (Figure 5
In order to assess fibrosis, liver slices were stained with Sirius Red (Figure 4
). A significant increase in fibrosis was only observed in animals receiving CHO + SOY diet, whereas all other animals only showed minor age-appropriate positive staining for Sirius Red (Figure 5
H). In accordance with these histological data, collagen 1a1 expression was slightly induced in animals fed CHO + STD or CHO + LAR diets, whereas a more than 10-fold induction was observed in livers of CHO + SOY-fed mice (Figure 5
These results show that only mice fed a CHO + SOY diet developed clear signs of hepatic inflammation with macrophage infiltration and increased expression of pro-inflammatory cytokines as well as hepatocyte apoptosis and liver fibrosis.
3.3. Diet-Induced Mitochondrial Damage and Oxidative Stress
The cholesterol-dependent induction of liver damage has previously been attributed to mitochondrial damage resulting from cholesterol accumulation in mitochondrial membranes. However, judging from the expression of complex I, II, and IV of the respiratory chain, dietary cholesterol alone apparently did cause low but no severe mitochondrial damage in our model (Figure 6
A). A mild reduction of complex I, II, and IV content was also observed in livers from animals receiving CHO + LAR diet. In stark contrast, complex I, II, and IV proteins were dramatically reduced in livers of CHO + SOY-fed animals, indicating severe mitochondrial damage in only this group (Figure 6
A). In keeping with these data, the amount of PGC-1α protein was strongly reduced in livers of animals receiving the CHO + SOY diet (Figure 6
Impairment of mitochondrial respiration causes severe oxidative stress. Accordingly, levels of malondialdehyde, a reaction product of lipid peroxidation of unsaturated fatty acids, and protein carbonyls were only increased in livers of CHO + SOY-fed mice (Figure 6
Thus, cholesterol-induced mitochondrial damage and oxidative stress was clearly enhanced by soybean oil-derived PUFA probably due to augmented lipid peroxidation.