The Adaptive Characteristics of Cholesterol and Bile Acid Metabolism in Nile Tilapia Fed a High-Fat Diet

Since high-fat diet (HFD) intake elevates liver cholesterol and enhanced cholesterol-bile acid flux alleviates its lipid deposition, we assumed that the promoted cholesterol-bile acid flux is an adaptive metabolism in fish when fed an HFD. The present study investigated the characteristic of cholesterol and fatty acid metabolism in Nile tilapia (Oreochromis niloticus) after feeding an HFD (13% lipid level) for four and eight weeks. Visually healthy Nile tilapia fingerlings (average weight 3.50 ± 0.05 g) were randomly distributed into four treatments (4-week control diet or HFD and 8-week control diet or HFD). The liver lipid deposition and health statue, cholesterol/bile acid, and fatty acid metabolism were analyzed in fish after short-term and long-term HFD intake. The results showed that 4-week HFD feeding did not change serum alanine transaminase (ALT) and aspartate transferase (AST) enzyme activities, along with comparable liver malondialdehyde (MDA) content. But higher serum ALT and AST enzyme activities and liver MDA content were observed in fish fed 8-week HFD. Intriguingly, remarkably accumulated total cholesterol (mainly cholesterol ester, CE) was observed in the liver of fish fed 4-week HFD, along with slightly elevated free fatty acids (FFAs) and comparable TG contents. Further molecular analysis in the liver showed that obvious accumulation of CE and total bile acids (TBAs) in fish fed 4-week HFD was mainly attributed to the enhancement of cholesterol synthesis, esterification, and bile acid synthesis. Furthermore, the increased protein expressions of acyl-CoA oxidase 1/2 (Acox1 and Acox2), which serve as peroxisomal fatty acid β-oxidation (FAO) rate-limiting enzymes and play key roles in the transformation of cholesterol into bile acids, were found in fish after 4-week HFD intake. Notably, 8-week HFD intake remarkably elevated FFA content (about 1.7-fold increase), and unaltered TBAs were found in fish liver, accompanied by suppressed Acox2 protein level and cholesterol/bile acid synthesis. Therefore, the robust cholesterol-bile acid flux serves as an adaptive metabolism in Nile tilapia when fed a short-term HFD and is possibly via stimulating peroxisomal FAO. This finding enlightens our understanding on the adaptive characteristics of cholesterol metabolism in fish fed an HFD and provides a new possible treatment strategy against metabolic disease induced by HFD in aquatic animals.


Introduction
Currently, to obtain the protein-sparing effect, high-fat diets (HFD) have been extensively formulated to meet the energy demand of fish growth. However, HFD usually leads to severe accumulation of lipids, mainly triglyceride (TG), in fish [1,2]. Also, the elevation of cholesterol is a common phenomenon in fish body. In fact, cholesterol is an essential lipid involved in many biological processes for providing a precursor for the biosynthesis of bile acids, vitamin D, and steroid hormones in animals. However, excessive free cho-lesterol (FC) is the major risk factor for nonalcoholic fatty liver disease (NAFLD) [3][4][5]. One of the mechanisms by which FC causes lipotoxicity is the generation of reactive oxygen species production and mitochondrial damage [6,7]. In fish, limited studies focus on the mechanism linking HFD-induced obesity to elevated cholesterol in the liver, while the physiological role of cholesterol in HFD-fed fish is not fully understood.
There is a strong link between cholesterol and fatty acids as previous studies reported that free fatty acids (FFAs) stimulate hepatic cholesterol biosynthesis [8,9]. It is well-known that cholesterol biosynthesis is a complex enzymatic biochemical process that initializes with substrate acetyl coenzyme A (acetyl-CoA). The acetyl-CoA for cholesterol biosynthesis mainly derives from the following two sources: one is generated by citrate cleavage, and the other is from acetate produced by fatty acid β-oxidation (FAO) [10,11]. A recent study about the mechanism of diabetes-induced hypercholesterolemia in mice revealed that stimulated liver peroxisomal FAO generated considerable free acetate, a precursor for cholesterol biosynthesis [12]. In a normal physiological condition, cholesterol homeostasis is tightly regulated in animals based on the cellular cholesterol level [13,14], and high cholesterol levels always promote cholesterol esterification and elimination to prevent lipotoxicity [15,16]. Previous reports in obese mice demonstrated that the enhancement of cholesterol metabolism, including cholesterol and bile acid efflux, alleviated the deposition of lipids such as cholesterol, FFA, and TG in the liver [17][18][19]. Therefore, to some extent, enhancing the conversion of FFA to cholesterol and bile acids is an important metabolism pathway for reducing FFA accumulation in the liver.
Previous researches believed that fish could grow normally without exogenous dietary cholesterol supplementation [20,21], indicating that robust regulation of cholesterol metabolism exists in fish and is enough to meet body cholesterol requirements compared with mammals. Additionally, although HFD feeding would disturb lipid homeostasis, resulting in excessive lipid accumulation and severe lipotoxicity in fish, there is a protective mechanism against the lipotoxicity of FFA in response to temporary HFD intake [22]. However, such a protective mechanism has not been well illustrated in fish. We assumed that its robust regulation of cholesterol metabolism is an efficient process to alleviate HFD-induced lipotoxicity. Therefore, further studies are required to demonstrate the physiological role of cholesterol in HFD-fed fish. Such knowledge helps to give insight into the role of cholesterol metabolism on metabolic disorders and provides a new possible treatment strategy against metabolic disease in aquatic animals. The present study mainly investigated the changes in cholesterol and fatty acid metabolism after short-term (4 weeks) or longterm (8 weeks) HFD feeding in Nile tilapia (Oreochromis niloticus), which is a global farmed fish widely used in metabolic studies [22,23]. Previous studies in Nile tilapia showed that approximately 6% lipid level was the optimal dietary lipid level [24][25][26], whereas 13% lipid level showed a negative impact on growth rates and feed utilization [27]. After the feeding trials, organic index, biochemical indicators, and gene/protein expression involved in cholesterol and fatty acid metabolism were analyzed.

Experimental Animals and Diet and Sample Collection.
The animal study was conducted strictly conforming to the procedures approved by the Committee on Ethics of Animal Experiments of East China Normal University (approval number F20190101). Nile tilapia were acclimated into tanks (water volume at 200 liters each tank within an indoor circu-lating water system) and fed a commercial diet (including 35% protein and 5% lipid, Tongwei, Co. Ltd., Chengdu, China) twice daily (09:00 h and 17:00 h) for two weeks. After acclimatization, visually healthy Nile tilapia fingerlings (average weight 3:50 ± 0:05 g) were randomly distributed into tanks (30 fish in each tank, three replicates, and four treatments). The experimental diets were formulated to contain two different lipid levels (6% and 13%; control diet and HFD, respectively), along with approximately 39% protein and 32% carbohydrate levels according to previous studies [24][25][26][27]. All dry ingredients were finely ground through 60-mesh screen and thoroughly mixed, and then, soybean oil and distilled water were added to thoroughly blend for producing 2 mm diameter pellets with a twin-screw extruder (South China University of Technology, Guangzhou, China). All the experimental diets were air-dried and stored at −20°C until used. The fish were fed with control diet and HFD (see Table 1) twice daily (09:00 h and 17:00 h) at 4% of their body weight per day for four weeks or eight weeks. The weight of fish in each tank was recorded to adjust diet intake each week. During the whole feeding trial, water temperature, total ammonia-nitrogen, and dissolved oxygen ranged from 27.00 to 29.00°C, 0 to 0.20 mg/L, and 5.00 to 6.50 mg/ L, respectively.
At the end of the 4-week or 8-week feeding trial, all fish of control and HFD (three tanks from each treatment) were deprived of food overnight, then anesthetized with MS-222 2 Aquaculture Nutrition (0.1 g/L, Sigma, USA), counted, and bulk weighed to analyze the survival rate and growth performance. Four fish were randomly collected from each tank and individually weighed and measured to calculate the condition factor (CF). Then, blood samples were collected from the fish caudal vein using 1 mL syringes. Serum samples were obtained for biochemical analyses after the centrifugation at 3000 rpm for 10 min at 4°C. Additionally, the fish were dissected to obtain mesenteric fat and liver tissues. The weights of these tissues were noted to determine the mesenteric fat index (MFI) and hepatosomatic index (HSI). Serum, liver, and intestinal content samples were immediately frozen into liquid nitrogen and stored at −80°C until further analysis.

Biochemical and Histological
Analyses. The total lipid in the liver was extracted by using chloroform/methanol (2 : 1, v/v) according to previously described [28]. Liver tissue and intestinal content were homogenized in phosphatebuffered saline (PBS) at a ratio of 1 : 10 (w : v) and then centrifuged (4°C, 3000 rpm, 10 min). The supernatants were collected and used for biochemical assays. The total cholesterol (TC) and free cholesterol (FC) were determined using specific commercial assay kits (Solarbio, Beijing, China) according to the method of enzymatic assays [29]. Briefly, the esterified cholesterol was hydrolyzed to free cholesterol by cholesterol esterase; then, the free cholesterol was oxidized to Δ4-cholesterone and H 2 O 2 by cholesterol oxidase. Finally, the H 2 O 2 was oxidized by peroxidase in the presence of 4aminoantipyrine and phenol to form red quinones, which could be measured at 500 nm by using spectrophotometers. TG, free fatty acids (FFAs), total bile acids (TBAs), lowdensity lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) in serum, liver, or intestinal content were measured using specific commercial assay kits    Aquaculture Nutrition (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's instructions. Briefly, TG was measured with glycerol phosphate oxidase/peroxidase and hydrogen peroxide (GPO/PAP) method. FFA was detected by acyl CoA synthetase and acyl CoA oxidase (ACS-ACOD) method. TBA was detected in the presence of Thio-NAD and 3-α-hydroxysteroid dehydrogenase (3-α-HSD), which converted bile acids to 3-keto steroids and Thio-NADH. LDL-C and HDL-C were determined with cholesterol esterase and cholesterol oxidase after being selectively solubilized by different specific surfactants (see instructions for details). Very low-density lipoprotein cholesterol (VLDL-C) was assessed by ELISA (enzyme-linked immunosorbent assay) kit (Hengyuan Biotech Co., China) according to the manufacturer's protocols. The activities of aspartate transferase (AST) and alanine transaminase (ALT) were detected following the method [30]. Briefly, AST catalyzed α-ketoglutaric acid and aspartate to produce glutamic acid and oxaloacetic acid. Oxaloacetic acid was further decarboxylated to form pyruvate, which react with 2,4dinitrophenylhydrazine to produce 2,4-dinitrophenylhydrazone. ALT facilitated the conversion of alanine and α-keto-glutarate to glutamate and pyruvate, which reacted with 2,4-dinitrophenylhydrazine to form a yellow pyruvate phenylhydrazone. Superoxide dismutase (SOD) activity and malondialdehyde (MDA) content were assessed according to the methods described previously [31,32]. SOD catalyzed the dismutation of the superoxide anion into hydrogen peroxide and O 2 , and superoxide anions acted on WST-1 to produce a water-soluble formazan dye. MDA reacted with thiobarbituric acid (TBA) to generate an MDA-TBA adduct, which could be quantified colorimetrically. Liver tissues were fixed in 4% paraformaldehyde solution and processed for histology in paraffin wax and then cut into 5 μm thick sections for neutral lipid staining following the procedure [33].

Statistical Analyses.
The obtained data were analyzed by two-way ANOVA after testing the normality and homogeneity of variances using the Shapiro-Wilk test and Levene's test, respectively. Significant differences (P ≤ 0:05) were determined by Tukey's HSD post hoc test. Asterisks indicate the differences between the C and HFD groups in the same feeding period. Hash symbols indicate the differences between the 4-week and 8-week feeding trials of the same lipid level diet. All the obtained experimental data were analyzed by using the SPSS Statistics 23.0 software (IBM, Michigan Avenue, USA). All values in the results are presented as the means ± SD (standard deviation).

HFD Feeding Caused
Hyperlipidemia. HFD intake (4 or 8 weeks) did not affect survival rate of fish (Figure 1(b)), but after the 4-week HFD feeding trial, although no significant difference in body weight was found between the control and HFD groups (Figure 1(a)), higher CF and MFI were observed in the HFD-fed Nile tilapia (Figures 1(c) and 1(d)). However, 8 weeks of HFD intake significantly increased the body weight of fish (Figure 1(a)) and caused remarkably higher CF, MFI, and HSI than the control (Figures 1(c)-1(e)). Furthermore, 4-week HFD intake significantly elevated serum TG (Figure 1(f)), TC (Figure 1(h)), and FC (Figure 1(i)) levels, which further accumulated after 8 weeks of HFD feeding. Notably, 4-week HFD feeding neither changed serum FFA level (Figure 1(g)) nor serum ALT and AST enzyme activities (Figures 1(k) and 1(l)), two important indicators of liver function. But higher serum FFA levels (Figure 1(g)), along with elevated ALT and AST (Figures 1(k) and 1(l)) enzyme activities were observed in the 8-week HFD-fed fish. Additionally, the 4-week HFD feeding elevated serum LDL-C content, which conversely diminished in fish after 8-week HFD intake (Figure 1(j)). Generally, short-term (4 weeks) HFD intake caused hyperlipidemia that was aggravated after long-term (8 weeks) HFD feeding and is accompanied by impaired liver functions in fish.

11
Aquaculture Nutrition remarkably increased TBA content in the liver and intestinal content after 4-week feeding trial (Figures 2(m) and 2(o)). However, after 8-week feeding trial, both of these parameters were unchanged in fish between the control and HFD groups (Figures 2(m) and 2(o)). These results indicated that 4-week HFD intake caused a high accumulation of CE rather than TG in fish liver, but elevated TG and FFA contents primarily accounted for the liver lipid deposition of fish fed 8week HFD.

Short-Term HFD Feeding Shows an Adaptive Mechanism
against Lipotoxicity in the Liver. It is well-known that HFD leads to lipid accumulation in most animals. In fish, obviously increased mass of visceral adipose tissue, elevated plasma TG, FFA, cholesterol levels, and liver lipid content were observed after HFD intake [37,38]. It should be pointed out that in response to short-term HFD intake, high levels of circulating FFAs are primarily stored in the form of TG in the adipose tissue of fish. This process has been identified as an adaptive mechanism to ease the FFA-induced lipotoxicity in animals [39]. For instance, our previous study in Nile tilapia reported that the increased number of adipocytes is the primary strategy to cope with HFD intake at the early stages of the progression of fatty liver, along with relatively stable TG content in the liver [22]. Consistently, higher CF and MFI but not HSI were also observed in the fish fed with HFD for 4 weeks in the present study. Although hyperlipidemia occurred in fish after 4 weeks of HFD intake, there is only a slight elevation of FFA in the fish liver while the TG content remains stable. The liver is the central metabolic organ, where dysregulation of lipid metabolism is a critical factor in liver steatosis [40]. In the present study, 4week HFD intake significantly inhibited fatty acid synthesis (fasn, acly, and srebp1) and lipolysis (atgl) processes. These results might partially explain why short-term HFD feeding did not cause liver damage, as evident by comparable serum ALT and AST enzyme activities, liver MDA content, and liver SOD enzyme activity in fish. We proposed that liver FFA level was tightly controlled within normal fluctuation levels so that the liver would not be impaired in fish, which supported the protective mechanism against lipotoxicity in the liver of fish after short-term HFD intake.

Short-Term HFD Feeding Triggers Cholesterol-Bile Acid
Flux to Eliminate FFA-Derived Cholesterol. Cholesterol biosynthesis begins with acetyl-CoA, which could be supplied by FAO, a critical lipid catabolism process involved in the degradation of FFA [10,11]. This might explain the previous finding that there is a strong link between cholesterol and FFA [8,9]. In the present study, short-term HFD feeding caused cholesterol accumulation (mainly CE) rather than TG in fish liver, accompanied by enhanced cholesterol synthesis (hmgcr, sqle, lss, and msmo1) and esterification (Acat1 and Acat2). The obvious elevation of CE along with slightly increased FC content might be explained by the fact that FC, as a polar lipid with lipotoxicity, can be efficiently converted to CE by ACATs, and then, CE is subsequently stored in lipid droplets [16]. Also, the liver eliminates the excessive cholesterol within cells mainly via the transformation of cholesterol to bile acids, which are then being excreted into the gallbladder and intestine [41,42]. Bile acid synthesis exclusively occurs in the liver and is the only quantitatively significant cholesterol catabolic mechanism [41,42]. In the present study, short-term HFD feeding increased TBA content in the liver and intestinal content, along with promoted bile acid synthesis (such as Cyp7a1, the first rate-limiting enzyme) in fish liver. These results indicated that the enhanced bile acid synthesis and excretion processes also served as a protective mechanism for eliminating FFAderived cholesterol in the liver of fish after short-term HFD intake. Of note, the gross energy is different between the control and high-fat diets in the present study. Thus, it raises an interesting question that whether the adaptive metabolic characteristics of cholesterol and bile acid also exist in Nile tilapia fed a high-energy diet. In fact, previous study showed that the high-carbohydrate diet intake also led to high level of hepatic cholesterol and bile acids [43], indicating the robust cholesterol and bile acid metabolism in the liver of fish under the high carbohydrate nutrition. Therefore, we hypothesized that fish could alleviate the metabolic disorders caused by the high-energy diet through enhancing cholesterol and bile acid metabolism. Future studies are necessary to assess the variety and content of bile acids in the gallbladder of fish and reveal their physiological functions in fish under the high energy nutrition.
4.3. Short-Term HFD Feeding Enhanced Peroxisomal FAO-Mediated Cholesterol and Bile Acid Synthesis. In mammals, it was reported that increased FAO activity in the liver is an efficient adaptive metabolism to prevent lipid accumulation and liver disease progression [44]. The process of FAO occurs in peroxisomes and mitochondria [45][46][47], both of which cooperate intimately [48]. For example, mitochondrial FAO was stimulated to alleviate FFA accumulation as a compensatory mechanism when peroxisomal FAO is suppressed [49]. On the contrary, stimulation of peroxisomal FAO inhibited mitochondrial FAO by promoting malonyl-CoA formation in the liver of mice fed an HFD [50]. In the present study, 4-week HFD feeding increased Acox1 and Acox2 (peroxisomal FAO rate-limiting enzymes) but inhibited Cpt1a (mitochondrial FAO rate-limiting enzyme) protein expressions in fish liver. This result suggests a similar cooperative relationship between mitochondrial and peroxisomal FAO in fish. Notably, substrate specificity studies in mammals uncovered that ACOX1 is involved in the oxidation of straight-chain fatty acids, but ACOX2 is 13 Aquaculture Nutrition responsible for the degradation of the branched-chain fatty acids and uniquely plays an important role in the transformation of cholesterol into bile acid [51]. In fact, it was estimated that the proportion of peroxisomal FAO is about 30% of total FAO under normal conditions in mice. However, this proportion elevated significantly in mice under the diabetic condition as liver peroxisomal FAO was stimulated in diabetes. Therefore, it was demonstrated that peroxisomal FAO plays an important role in regulating cholesterol synthesis, and enhanced peroxisomal FAO contributed to high cholesterol level in diabetes mice [12]. Also, in the present study, the stimulated peroxisomal FAO might primarily contribute to the enhanced cholesterol and bile acid synthesis in the liver of fish fed a short-term HFD, thereby reducing FFA accumulation and alleviating liver injury caused by lipotoxicity. Since, long-chain fatty acids as endogenous ligands for PPARα, elevated hepatic FFA level in diabetic mice activated PPARα and resulted in promoting the transcription of genes related to peroxisomal FAO [52,53]. Our previous study in Nile tilapia also indicated that the activation of PPARα by fenofibrate significantly promoted mitochondrial and peroxisomal FAO in the liver [54]. In the present study, 4-week HFD feeding induced pparα gene expression, which possibly contributed to the enhanced peroxisomal FAO. These results indicated that short-term HFD feeding enhanced peroxisomal FAO-mediated cholesterol and bile acid synthesis (see Figure 5). Therefore, peroxisomal FAO may also play a critical role in regulating cholesterol and bile acid synthesis in fish.

Long-Term HFD Intake Suppressed Conversion of FFA to
Cholesterol. Animals possess an adaptive metabolism during the short-term HFD intake. The enlargement of adipocytes, which contain many lipid droplets, is the normal physiological consequence and protective strategy against lipotoxicity from high FFA [22,55]. However, the size or numbers of adipocytes would not continuously increase with the duration of HFD intake; thereby, the maintenance of lipid homeostasis would be disrupted after long-term HFD intake. This would result in pathological consequences such as excessive lipid deposition in other tissues such as the liver, accompanied by high levels of oxidative and inflammatory factors in animals, including fish [1,56]. The increased reactive oxygen species (ROS) produced by lipid peroxidation in hepatocytes would seriously impair lipid metabolism and further increase lipid accumulation and cause liver damage [57]. In the present study, 8-week HFD intake further elevated serum TG and TC levels and caused higher serum ALT and AST enzyme activities, indicating impaired liver in fish after long-term HFD intake. Also, more severe lipid accumulation was observed in fish liver, as is evident by obviously accumulated TG (1.4-fold increase) and FFA (1.7-fold increase). Consistently, elevated MDA content and diminished SOD activity were observed in the liver of fish fed 8-week HFD in the present study. These results might be partially due to perturbed lipid metabolism, as HFD stimulated lipolysis and suppressed peroxisomal β-oxidation (Acox2) protein expression in fish liver, although it did not affect the fatty acid synthesis, autophagy, and mito-chondrial β-oxidation. Additionally, the inhibited cholesterol synthesis might link to deficient peroxisomal FAO, which provided substrate acetyl-CoA in fish liver. The disturbance of cholesterol metabolism caused by HFD intake is a critical factor for FC accumulation [2,58]. Previous study showed that rats with HFD-induced obesity presented cholesterol accumulation in the liver via inhibiting CYP7A1 gene expression, which remarkably decreased the conversion of cholesterol to bile acids [58]. In Nile tilapia, long-term HFD feeding caused hepatic cholesterol accumulation mainly due to impaired assembly of VLDL and HLD particles [2]. However, in the present study, 8-week HFD feeding significantly elevated liver FC content that might be associated with compromised esterification of cholesterol. Additionally, LDL-C content and its receptor ldlr gene expression were increased, whereas cholesterol efflux and bile acid synthesis did not change in the liver of fish fed HFD. Accordingly, declined serum LDL-C level was observed in fish after 8-week HFD intake, accompanied by unaltered TBA levels in the liver and intestinal content. These results suggested that accumulated hepatic cholesterol resulted from the increased LDL-C uptake from the blood, and long-term HFD intake also caused the disturbance of cholesterol metabolism in fish. The contradictory conclusions from various studies might be due to the discrepancy between species and their nutrition status. Of note, the remarkably elevated FFA content (about 1.7-fold increase) was found in the liver of fish, along with suppressed Acox2 protein level and downregulated cholesterol/bile acid synthesis. Hence, the deficient liver peroxisomal FAO might block the conversion of FFA to cholesterol and bile acids, thereby failing to alleviate FFA accumulation in fish after 8-week HFD intake. Therefore, the elevated FFA and FC contents primarily accounted for liver lipotoxicity and injury in fish fed a long-term HFD (see Figure 5).

Conclusion
The present study revealed that short-term HFD intake triggered protective mechanisms against FFA accumulation and liver injury. Of note, short-term HFD feeding caused cholesterol accumulation (mainly CE) rather than TG in fish liver, along with increased TBA content in liver and intestinal content. Furthermore, the stimulated hepatic peroxisomal FAO, which provides substrate acetyl-CoA for the biosynthesis of cholesterol, might primarily contribute to cholesterol and bile acid synthesis and efflux, thereby alleviating FFA accumulation in the liver of fish during the shortterm HFD intake. However, long-term HFD intake caused evident elevation of TG and FFA, along with accumulated FC content in fish liver. The elevated FC content resulted from the enhancement of LDL-C uptake indicating longterm HFD intake also caused the disturbance of cholesterol metabolism. Of note, long-term HFD intake suppressed peroxisomal FAO and cholesterol/bile acid synthesis, indicating the blocked conversion of FFA to cholesterol and bile acids, thereby failing to alleviate FFA accumulation in fish liver. Therefore, the robust cholesterol-bile acid flux serves as an adaptive metabolism in Nile tilapia fed an HFD, possibly 14 Aquaculture Nutrition via stimulating peroxisomal FAO pathway. This finding enlightens our understanding of the adaptive characteristics of cholesterol metabolism in fish fed an HFD and provides a new possible treatment strategy against metabolic disease induced by HFD in aquatic animals.

Data Availability
The datasets generated during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest
All authors declare no conflicts of interests.