Integrated Metabolome and Transcriptome Analyses Reveal the Efficacy of Steroidal Saponins for Glucose and Lipid Metabolism in Hybrid Grouper (♀Epinephelus fuscoguttatus × ♂Epinephelus lanceolatu) Fed Higher-Lipid Diets

Simple Summary The ingestion of higher-lipid diets by fish could lead to excessive lipid deposition and metabolic disorders in vivo. This experiment analyzed the effects of steroidal saponins on hepatic glucose and lipid metabolism in hybrid grouper fed higher-lipid diets using a multi-omics approach. Data from grouper fed diets with 0% and 0.1% steroidal saponins show that 0.1% steroidal saponins inhibited glycogen synthesis, gluconeogenesis, and the arachidonic acid metabolism pathway and activated glycogenolysis, glycolysis, and the fatty acid β-oxidation pathway. This study provides a deeper insight into the glycolipid metabolic processes in the liver of grouper fed higher crude-lipid diets. Abstract An analysis of the extent of the effect of steroidal saponin addition on glucose and lipid metabolism in hybrid grouper liver was performed at the transcriptomic and metabolomic levels. Feeds (52% crude protein, 14% crude lipid) were prepared containing 0% (S0), 0.1% (S0.1), and 0.2% (S0.2) steroidal saponins. After eight weeks of feeding trial, compared to the S0 group, the activities of serum albumin, alanine aminotransferase, and aspartate transaminase were significantly lower and the activities of lysozyme, acid phosphatase, and alkaline phosphatase were significantly higher in the S0.1 group (p < 0.05). The superoxide dismutase, catalase, and glutathione peroxidase activities in the livers of the S0.1 group were significantly higher than those of the S0 group, while the malondialdehyde content was significantly lower than that of the S0 group (p < 0.05). There were forty-two differentially expressed genes and thirty-two differential metabolites associated with glucose and lipid metabolism enriched using KEGG and GO. In the S0 group, the expression of prostaglandin-endoperoxide synthase 1, prostaglandin E synthase 1, and thromboxane-2 synthase mRNA was significantly higher than in the S0.1 group (p < 0.05). The expression levels of genes in the S0 group were significantly higher than those in the S0.1 group (p < 0.05), including for glycogen synthase kinase, glucose-6-phosphatase catalytic subunit 2, fructose-1,6-bisphosphatase, phosphoenolpyruvate carboxykinase, glucose transporter 4, and malate dehydrogenase. The expression of mRNA such as fatty acid synthase, acetyl-CoA carboxylase, and sterol regulatory element-binding protein 1 was significantly lower in the S0.1 group than in the S0 group, while the expression of carnitine acyltransferase 1, acyl-CoA synthetase, and acyl-CoA dehydrogenase genes was significantly higher in the S0 group (p < 0.05). In summary, glycogen synthesis, gluconeogenesis, and the arachidonic acid metabolism pathway were inhibited by 0.1% steroidal saponins, and glycogenolysis, glycolysis, the tricarboxylic acid cycle, and the fatty acid β-oxidation pathway were activated. This study aims to provide a reference for the formulation of grouper feeds with a higher crude-lipid level.


Introduction
In the past five years, China's cultivated grouper production grew from 131,500 tons to 204,100 tons, an increase of 55.21% [1,2].The development of compound diet technology is a key factor in ensuring grouper production.Further, fishmeal, which is an optimal protein source for grouper diets, is rich and balanced in nutrients, especially in protein and fatty acids [3].Between 2017 and 2021, Chinese fishmeal production rose from 639,200 tons to 659,000 tons.However, in fact, the annual demand for fishmeal in China was about 1.7 million tons, which meant that the Chinese fishmeal imports had increased from 1,571,600 tons to 1,823,400 tons with an external dependence of over 70% [1,2,4,5].In view of this, fishmeal influenced the formula composition and price of feed and limited the efficient development of aquaculture.
Appropriately increasing lipid content in diets is an effective measure to improve protein efficiency and feed utilization [6][7][8][9], but long-term intake of high-lipid diets can result in abnormal accumulation of hepatic lipids and induce metabolic diseases such as fatty liver, which leads to metabolic disorders and immune suppression in the organism, affecting fish growth and reducing the economic efficiency of culturing [10,11].
The Panax notoginseng saponin can enhance the antioxidant capacity of the organism by increasing superoxide dismutase and catalase enzyme activities and decreasing malondialdehyde content in the liver of grouper [12].Yucca schidigera saponins and soya saponins were found to reduce malondialdehyde, total cholesterol, and triglyceride levels in the serum and liver of white shrimp (Litopenaeus vannamei) and turbot (Scophthalmus maximus L.) [13,14].The steroid saponins belong to a steroidal class of saponins, including spirostanes, furanosteroids, and furanospiralosteroids, which have the ability to scavenge free radicals, such as -OH and O 2− [15], to lower serum and hepatic levels of cholesterol and triglycerides [16] as well as the ability to promote the breakdown of hepatic glycogen [17].The bitter melon saponins (Momordica charantia, steroids) have been shown to enhance glucose metabolism by preventing the absorption of glucose in the small intestine of mice and promoting the breakdown of glycogen in the liver [18,19].Saikosaponin d has been confirmed to alleviate high-fat-diet-induced hepatic steatosis in hybrid grouper by targeting the AMPK/PPARα pathway [20].Our previous study found that steroidal saponins improve the utilization of lipids and glucose in grouper liver when the fish ingested a high-lipid diet [21].What is the exact mechanism?
The transcriptome is the pivot that links the genetic information of genes to functional proteins [22].For example, it can explain that taurine can promote hepatic endogenous taurine synthesis, bile secretion, insulin secretion, and fatty acid β-oxidation efficiency to reduce fat accumulation in grouper ingesting high-lipid diets (15% crude lipid) [23].As a bridge between genotype and phenotype, the metabolome ensures the maintenance of aerobic metabolism and activates gluconeogenesis as key metabolic strategies of turbot kidney against heat stress [24].Association analysis of transcriptomic and metabolomic data provides a more detailed analysis of the relationship between critical genes and metabolites to explore the potential mechanisms of the organism [25].The down-regulated key genes (lysophosphatidylcholine acyltransferase and glycerol-3-phosphate dehydrogenase) and increased metabolites (choline and taurine) were found to contribute to the alleviation of lipid metabolism disorders caused by elevated temperature stress in turbot [26].In this study, we combine transcriptome and metabolome association analysis to investigate how steroidal saponins alter hepatic glycolipid metabolism in grouper.

Experimental Design
Three isoproteic (52% crude protein) and isolipidic (14% crude lipid) diets were formed by adding 0% (S 0 ), 0.1% (S 0.1 ), and 0.2% (S 0.2 ) steroidal saponins (Guangzhou Fishtech Biotechnology Co., Ltd., Guangzhou, China) to diets with fishmeal, concentrated cottonseed protein, poultry by-product meal, and soybean meal as the main protein source and fish oil and soybean oil as the main lipid sources.Please refer to Supplementary Material (Tables S1 and S2).Experimental fish were fed the experimental diets to apparent satiation at 8:00 and 16:00 daily for 8 weeks.The hybrid grouper were separated into nine buckets (1 m 3 Fiberglass farming buckets) with twenty-five fish in each bucket.During the experiment, the fish were continuously oxygenated every day, dissolved oxygen 5-6 mg/L, ammonia nitrogen <0.2 mg/L, temperature 30.5 ± 0.8 • C, salinity 28-32.Please refer to our previous study for the detailed feeding process [21].

Sample Collection and Chemical Analysis
Blood samples were obtained from the tail vein of hybrid grouper.The samples were centrifuged at 4 • C and 3500 rpm for 15 min to extract the serum and then immediately stored at −80 • C.
Meanwhile, the soybean-sized liver (approximately 600 mg) was obtained from each replicate on ice and then washed with saline.At once, the liver samples were stored at −80 • C for analysis of the gene expression of glucose lipid metabolism enzyme.
The grouper were anesthetized with MS-222 (1:10,000), and then the fresh soybeansized liver was cut from living fish fed diets S 0 and S 0.1 and immediately washed with saline to clean the surface of blood remaining on the tissue [21].The liver samples were numbered, stored in 2 mL freeze tubes, quickly put in liquid nitrogen, and subsequently transferred to −80 • C for metabolome and transcriptome analyses.The detailed methodology was described in Yang and Li et al. [9,13].

Metabolome Analysis
The 100 mg liver samples (six replicates per group) were ground separately in liquid nitrogen, and 500 µL of precooled 80% methanol was resuspended using well vortex to homogenize.The liver samples were ice-bathed for 5 min, followed by centrifugation at 15,000× g for 20 min at 4 • C. The supernatants were diluted with mass-spectrometry-grade water to a methanol content of 53% and then centrifuged (15,000× g, 20 min, 4 • C).Thus, the supernatants were injected into the UPLC-MS/MS system for analysis.UPLC-MS/MS analyses were conducted at Gene Denovo Ltd. (Guangzhou, China), and a Vanquish UPLC system (ThermoFisher, Berlin, Germany) and an Orbitrap Q ExactiveTM HF-X mass spectrometer (Thermo Fisher, Berlin, Germany) were used.
The raw data files were processed using UPLC-MS/MS with Compound Discoverer 3.1 (CD3.1,Thermo Fisher) performing peak alignment, peak extraction, and quantification for each metabolite.Peak was matched to mzCloud (https://www.mzcloud.org/,accessed on 16 February 2022), mz Vault, and Mass List database to obtain accurate qualitative and relative quantitative results.We used the statistical software R (R-3.4.3 version), Python (Python version 2.7.6), and CentOS (CentOS version 6.6) for statistical analysis.A normal transformation was attempted using the area normalization method when the data were not normally distributed.

Transcriptome Analysis
The total RNA of liver (about 300 mg) was extracted with Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol.The RNA quality integrity and concentration were assessed with agarose gel electrophoresis and Nanodrop 2000 assay.The input concentration of RNA sample per replicate in S 0 group was 765.8, 576.9, and 406.3 ng/uL; in S 0.1 group, 764.7, 537.0, and 792.5 ng/uL; in S 0.2 group, 808.7, 675.4,and 855.0 ng/uL.The enriched mRNA was then fragmented with fragmentation buffer and reversed to cDNA using the NEBNext Ultra RNA Library Prep Kit for Illumina (NEB #7530, New England Biolabs, Ipswich, MA, USA).The cDNA libraries were sequenced with Gene Denovo Biotechnology (Guangzhou, China) using Illumina Novaseq6000.

Quantitative RT-PCR Analysis of Gene
Quantitative reverse transcription PCR (qRT-PCR) was performed to verify the relative expression levels of the fourteen selected DEGs so that the gene expression data obtained with RNA-Seq could be validated.Liver total RNA was extracted with Trizol reagent (Invitrogen, Carlsbad, CA, USA).The cDNA was composed with the Prime Script RT kit (Takara, Osaka, Japan), and the quantitative thermal cycler (Light Cycler 480II, Roche Diagnostics, Basel, Switzerland) was used for qRT-PCR with the SYBR Premix Ex Taq kit (Takara, Osaka, Japan) within a reaction volume of 10 µL, including 3.2 µL of sterilized double-distilled water, 1 µL of cDNA, 0.4 µL of each primer, and 5 µL of SYBR Premix Ex Taq (Takara, Osaka-shi, Japan).Requirements were a cycle of 30 s at 95 • C followed by thirty-five cycles at 95 • C for 5 s, 60 • C for 25 s, and 72 • C for 30 s.The β-actin (AY510710.2) was used as internal control.The results were calculated using the 2 −∆∆Ct method [12].A list of primers is given in Table 1.

Statistical Analysis
Mean ± standard error (SE) was used to present the results of the experiment.The statistic was analyzed using one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison tests with the software SPSS 21.0 (SPSS Inc., Michigan Avenue, Chicago, IL, USA).Significant differences were identified as p < 0.05.The Graph Pad Prism 8.0 software (8.0.2.263) was used to image experimental results.

Activities of ALT, AST, and Levels of TP, ALB in Serum
The results of the serum biochemical indexes are shown in Table 2.The enzyme activity of ALT between the S 0.1 and S 0.2 groups was not significantly different, but both groups' ALT activity was significantly lower than that in the S 0 group (p < 0.05).The AST activity in the S 0.1 group was significantly lower than that in the other two groups (p < 0.05).There was no significant difference in the TP content of serum among the three groups (p > 0.05).The ALB level in the S 0.2 group was significantly higher than that in the S 0.1 group but significantly lower than that in the S 0 group (p < 0.05).Note: Values range from minimum to maximum.There are significant differences indicated with different letters (p < 0.05).

Levels of LZM, ACP, and AKP in Serum
The serum nonspecial immune indexes of the hybrid grouper are shown in Table 3.The enzyme activities of LZM, ACP, and AKP in the S 0 group were not significantly different from those in the S 0.2 group, while they were significantly lower than those in the S 0.1 group (p < 0.05).Note: Values range from minimum to maximum.There are significant differences indicated with different letters (p < 0.05).

Liver Antioxidative Indexes
As shown in Table 4, the content of MDA in the S 0 group was significantly higher than that in the experiment groups (p < 0.05).The SOD and CAT enzyme activities in the S 0.1 group were significantly higher than those in the other groups (p < 0.05).The enzyme activity of GSH-PX in the S 0.2 group was significantly higher than that in the S 0 group but significantly lower than that in the S 0.1 group (p < 0.05).Note: Values range from minimum to maximum.There are significant differences indicated with different letters (p < 0.05).

Transcriptome Analysis in S 0 and S 0.1 Group Fish Liver
To explore the potential mechanisms of steroidal saponins on glucose lipid metabolism in hybrid grouper liver, first, the high-throughput RNA-sequencing technique was used in this experiment.The goodness of reproducibility between samples within groups and the degree of variability of samples between groups were analyzed with PCA (Figure 1A).Then, differentially expressed genes (DEGs) were separated into the S 0 group and the best-addition S 0.1 group of steroidal saponins.A total of 5667 DEGs were identified under the log 2 |Fold change| > 2.0 threshold and with a p value < 0.05 (Figure 1C), including 2687 upregulated DEGs and 2980 down-regulated DEGs (Figure 1B).
Animals 2023, 13, x FOR PEER REVIEW 7 of activity of GSH-PX in the S0.2 group was significantly higher than that in the S0 group b significantly lower than that in the S0.1 group (p < 0.05).Note: Values range from minimum to maximum.There are significant differences indicated wi different letters (p < 0.05).

Transcriptome Analysis in S0 and S0.1 Group Fish Liver
To explore the potential mechanisms of steroidal saponins on glucose lip metabolism in hybrid grouper liver, first, the high-throughput RNA-sequencin technique was used in this experiment.The goodness of reproducibility between sampl within groups and the degree of variability of samples between groups were analyze with PCA (Figure 1A).Then, differentially expressed genes (DEGs) were separated in the S0 group and the best-addition S0.1 group of steroidal saponins.A total of 5667 DEG were identified under the log2 |Fold change| > 2.0 threshold and with a p value < 0.05 (Figu 1C), including 2687 up-regulated DEGs and 2980 down-regulated DEGs (Figure 1B).

Signaling Pathway Network and Key Genes of Glucose and Lipid Metabolism
The DEGs were entered into the GO and KEGG databases to search their involve signaling pathways (Figure 2).The present experiment involved 46 key different gen (Table 5), including 18 down-regulated and 28 up-regulated key genes.Of these, 15 gen were significantly down-regulated, and 24 genes were significantly up-regulated.The 30 DEGs were selected, as shown in Table 5, and clustered into a heatmap depending o their expression level and classification (Figure 2C).

Signaling Pathway Network and Key Genes of Glucose and Lipid Metabolism
The DEGs were entered into the GO and KEGG databases to search their involved signaling pathways (Figure 2).The present experiment involved 46 key different genes (Table 5), including 18 down-regulated and 28 up-regulated key genes.Of these, 15 genes were significantly down-regulated, and 24 genes were significantly up-regulated.Then, 30 DEGs were selected, as shown in Table 5, and clustered into a heatmap depending on their expression level and classification (Figure 2C).The interesting DEGs were enriched in GO, focusing on the thirty items in this study.To be exact, among the 30 enrichment items analyzed using the rich factor value, 24 items were associated with glucose metabolism, including glycogen metabolism (positive regulation of glycogen mediated process, positive regulation of glycogen biosynthetic process, UDP-glucose-4-epimerase activity, and UDP-glucose: glycoprotein glycosyltransferase activity), glycolysis (glucose catabolic process, glucose catabolic process to pyruvate, glucose catabolic process to lactate, and glucose catabolic process to lactate via pyruvate), glycogenesis (glycolytic process through glucose-6-phosphate, glucose-6-phosphate metabolic process, and DTDP-glucose-4,6-dehydratase activity), tricarboxylic acid cycle (glucose-6-phosphate isomerase activity), glucose metabolic process, regulation of glucose-mediated signaling pathway, invasive growth in response to glucose limitation, negative regulation of glucose-mediated signaling pathway, glucose-mediated signaling pathway, glucose homeostasis, regulation of invasive growth in response to glucose limitation, response to glucose, response to oxygen glucose deprivation, cellular response to oxygen glucose deprivation, cellular glucose homeostasis, and cellular response to glucose stimulus (Figure 2A).Meanwhile, six items were related to lipid metabolism, including phospholipid-hydroperoxide glutathione peroxidase activity, regulation of lipid storage, calcium-dependent phospholipid binding, positive regulation of sphingolipid biosynthetic process, fatty acid binding, and protein-lipid complex assembly (Figure 2A).
In addition, the focused DEGs were enriched into 30 interesting KEGG signaling pathways.The 30 signaling pathways showed that glucose metabolism (glycolysis/gluconeogenesis, oxidative phosphorylation, carbon metabolism, pentose phosphate pathway, fructose and mannose metabolism, glyoxylate and dicarboxylate metabolism, and thermogenesis), amino acids metabolism (biosynthesis of amino acids, proteasome, protein processing in endoplasmic reticulum, glutathione metabolism, cysteine and methionine metabolism, and protein digestion and absorption), lipid metabolism (metabolism of xenobiotics by cytochrome P450 and drug metabolism by cytochrome P450), and other signaling pathways (antigen processing and presentation, cardiac muscle contraction, apoptosis, estrogen signaling pathway, HIF-1 signaling pathway, drug metabolism by other enzymes, nucleocytoplasmic transport, phagosome, regulation of actin cytoskeleton, necroptosis, ferroptosis, mineral absorption, lysosome, RNA degradation, and toll and imd signaling pathway) were mainly involved in this experiment (Figure 2B).

Metabolism Analysis in S 0 and S 0.1 Group Fish Liver
After constructing the model with Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA) and Partial Least Squares Discriminate Analysis (PLS-DA), the sample clustering results were described using OPLS-DA and PLS-DA score plots (Figure 3A,B).It was shown that metabolites may be clustered depending on differences in the presence or absence of steroidal saponins in high-lipid diets.For example, the S 0.1 group was located on the positive side of the T-score and Partial Least Squares-1 (PLS1).The metabolite products were identified in this research, including prostaglandin G2/H2, thromboxane-2, malate, phosphoenolpyruvic acid, and fructose-6-phosphate.Moreover, the metabolites were enriched into 15 KEGG signaling pathways (Figure 3C).The results showed that lipid metabolism (bile secretion, ppar signaling pathway, adipocytokine signaling pathway, regulation of lipolysis in adipocytes, fatty acid metabolism, fatty acid elongation, fatty acid degradation, fatty acid biosynthesis, biosynthesis of unsaturated fatty acids, arachidonic acid metabolism, and steroid hormone biosynthesis) and other metabolism pathways were mainly involved in this experiment.
In addition, thirty-two metabolites related to lipid metabolism are listed in Table 6, of which eight metabolites were significantly down-regulated, nine metabolites were significantly up-regulated, nine metabolites were down-regulated, and six metabolites were up-regulated.At the same time, there are eighteen key metabolites related to glucose metabolism listed in Table 6, of which three metabolites were significantly down-regulated, nine metabolites were down-regulated, two metabolites were significantly up-regulated, and four metabolites were up-regulated.In addition, thirty-two metabolites related to lipid metabolism a of which eight metabolites were significantly down-regulated, nin significantly up-regulated, nine metabolites were down-regulated, a were up-regulated.At the same time, there are eighteen key metabolite metabolism listed in Table 6, of which three metabolites were s regulated, nine metabolites were down-regulated, two metabolites we regulated, and four metabolites were up-regulated.

Combined Analysis of Transcriptomic and Metabolomic Data
In an attempt to explore the changes in metabolites and mRNA transcripts during glycose lipid metabolism, we performed correlation analysis using transcriptomic and metabolomic data (Figure 4).It was shown that the addition of steroidal saponins to highlipid diets contributed to activating glycogen breakdown, glycolysis, the TCA cycle, and the fatty acid β-oxidation pathway while inhibiting glycogen synthesis, glycogenesis, fatty acids biosynthesis, cyclooxygenase, and the arachidonic acid signaling pathway.

Gene Expression of ptgs-1, ptges-1, and tbxas-2
There was no significant difference in liver ptgs-1 and tbxas-2 mRNA expression between the S 0.1 group and the S 0.2 group, but both groups' expression levels were significantly lower than those in the S 0 group (p < 0.05) (Figure 5).In the S 0.2 group, the expression of ptges-1 mRNA was significantly higher than in the S 0.1 group while significantly lower than in the S 0 group (p < 0.05).

Gene Expression of Glucose Metabolism in Liver
As indicated in Figure 6, the expression of gsk, f-1,6-bp, and pepck mRNA in the S0.1 group was not significantly different from that in the S0.2 group (p > 0.05), but the three genes' expression levels were significantly lower in the S0.1 and S0.2 groups than in the S0 group (p < 0.05).The g6pca2 mRNA expression level did not show differences between the S0 and S0.2 groups (p > 0.05), but it was significantly higher in the S0.1 group (p < 0.05).The expression of glut-4 mRNA in the S0 group was significantly higher than that in the S0.2 group but significantly lower than that in the S0.1 group (p < 0.05).The md mRNA expression levels in the S0 group and S0.2 group were not found to be significantly different, while both groups' levels were significantly lower than that in the S0 group (p < 0.05).

Gene Expression of Lipid Metabolism in Liver
As can be seen in Figure 7, the fas, acc, and srebp-1 mRNA expression levels between the S0.1 and S0.2 groups were not significantly different (p > 0.05), but they were significantly lower than those in the S0 group (p < 0.05).Although the expression of cpt-1 and acsl mRNA in the S0 group was significantly higher than that in the S0.2 group, it was significantly lower than that in the S0.1 group (p < 0.05).The acadm mRNA expression in the S0 group was significantly lower than that in the S0.1 and S0.2 groups (p < 0.05).

Gene Expression of Glucose Metabolism in Liver
As indicated in Figure 6, the expression of gsk, f-1,6-bp, and pepck mRNA in the S 0.1 group was not significantly different from that in the S 0.2 group (p > 0.05), but the three genes' expression levels were significantly lower in the S 0.1 and S 0.2 groups than in the S 0 group (p < 0.05).The g6pca2 mRNA expression level did not show differences between the S 0 and S 0.2 groups (p > 0.05), but it was significantly higher in the S 0.1 group (p < 0.05).The expression of glut-4 mRNA in the S 0 group was significantly higher than that in the S 0.2 group but significantly lower than that in the S 0.1 group (p < 0.05).The md mRNA expression levels in the S 0 group and S 0.2 group were not found to be significantly different, while both groups' levels were significantly lower than that in the S 0 group (p < 0.05).

Gene Expression of Glucose Metabolism in Liver
As indicated in Figure 6, the expression of gsk, f-1,6-bp, and pepck mRNA in the S0.1 group was not significantly different from that in the S0.2 group (p > 0.05), but the three genes' expression levels were significantly lower in the S0.1 and S0.2 groups than in the S0 group (p < 0.05).The g6pca2 mRNA expression level did not show differences between the S0 and S0.2 groups (p > 0.05), but it was significantly higher in the S0.1 group (p < 0.05).The expression of glut-4 mRNA in the S0 group was significantly higher than that in the S0.2 group but significantly lower than that in the S0.1 group (p < 0.05).The md mRNA expression levels in the S0 group and S0.2 group were not found to be significantly different, while both groups' levels were significantly lower than that in the S0 group (p < 0.05).

Gene Expression of Lipid Metabolism in Liver
As can be seen in Figure 7, the fas, acc, and srebp-1 mRNA expression levels between the S0.1 and S0.2 groups were not significantly different (p > 0.05), but they were significantly lower than those in the S0 group (p < 0.05).Although the expression of cpt-1 and acsl mRNA in the S0 group was significantly higher than that in the S0.2 group, it was significantly lower than that in the S0.1 group (p < 0.05).The acadm mRNA expression in the S0 group was significantly lower than that in the S0.1 and S0.2 groups (p < 0.05).

Gene Expression of Lipid Metabolism in Liver
As can be seen in Figure 7, the fas, acc, and srebp-1 mRNA expression levels between the S 0.1 and S 0.2 groups were not significantly different (p > 0.05), but they were significantly lower than those in the S 0 group (p < 0.05).Although the expression of cpt-1 and acsl mRNA in the S 0 group was significantly higher than that in the S 0.2 group, it was significantly lower than that in the S 0.1 group (p < 0.05).The acadm mRNA expression in the S 0 group was significantly lower than that in the S 0.1 and S 0.2 groups (p < 0.05).

Discussion
After ingestion of high-lipid diets (17% and 15% crude lipid), the area of lipid droplets in hybrid grouper liver increased, and the activities of the LZM, AKP, ACP, SOD, and CAT enzymes in the liver and serum decreased, while MDA, ALB, AST, and ALT levels were higher [7,27].When high-lipid diets (15% crude lipid) were supplemented with saponins, the lipid droplet area decreased in the liver of hybrid grouper, and SOD, CAT, and GSH-PX activities significantly increased to enhance the antioxidant capacity of the fish [20].In carp [17], white shrimp [28], and swimming crab (Portunus trituberculatus) [29], studies have demonstrated that saponins can decrease serum ALB, AST, and ALT enzyme activities and increase antioxidant and nonspecific immune enzyme activities in the liver to enhance the immune capacity of the body.A previous study demonstrated that 0.1% steroidal saponin significantly increased the activities and gene expression of SOD, CAT, GSH-PX, and GR in hybrid grouper liver and serum and decreased the MDA content [21].In the present experiment, compared to the control group, supplementation of 0.1% steroidal saponins to a 14% crude lipid diet significantly increased LZM, ACP, and AKP enzyme activities and decreased ALB, ALT, and AST enzyme activities in the serum of hybrid grouper.
Previous studies on Chinese mitten crab (Eriocheir sinensis) [30] and olive flounder (Paralichthys olivaceus) [31] found that arachidonic acid can enhance the antioxidant capacity of the organism by increasing SOD, CAT, AKP, ACP, and LZM activities in the liver and serum.Three metabolic pathways-cytochrome P450, cyclooxygenase, and lipoxygenase-are all engaged in the metabolism of arachidonic acid [32], in which inhibition of the expression and content of key genes and metabolites of the cyclooxygenase metabolic pathway could improve the anti-inflammatory effects of the body [33].Compared with the control group, the sod, cat, and gsh-px genes' expression in hybrid grouper liver was up-regulated with the addition of 0.1% steroidal saponins, and the levels of PGG2/PGH2, PGA2 and TXA2 metabolites, ptgs-1 and ptgs-2, ptges-1 and ptges-2, and tbxas-2 mRNA significantly decreased.Both KEGG and GO showed that these genes and metabolites were enriched in the cyclooxygenase metabolic pathway.The upstream ptgs-1 and ptgs-2 genes can activate the cyclooxygenase metabolic pathway to release the intermediate products PGG2 and PGH2, which stimulate the downstream ptges-1, ptges-2, and tbxas-2 genes to produce the downstream products PGA2 and TXA2, mediating the inflammatory response [34,35].The present study also demonstrated that the supplementation of 0.1% and 0.2% steroidal saponins in the diets contributed to a decrease in the expression of ptgs-1 and ptgs-2, ptges-1 and ptges-2, and tbxas-2 mRNA in the liver of hybrid grouper, inhibited the expression of key genes in the cyclooxygenase metabolic pathway, and played an anti-inflammatory role in the body's immune system.

Discussion
After ingestion of high-lipid diets (17% and 15% crude lipid), the area of lipid droplets in hybrid grouper liver increased, and the activities of the LZM, AKP, ACP, SOD, and CAT enzymes in the liver and serum decreased, while MDA, ALB, AST, and ALT levels were higher [7,27].When high-lipid diets (15% crude lipid) were supplemented with saponins, the lipid droplet area decreased in the liver of hybrid grouper, and SOD, CAT, and GSH-PX activities significantly increased to enhance the antioxidant capacity of the fish [20].In carp [17], white shrimp [28], and swimming crab (Portunus trituberculatus) [29], studies have demonstrated that saponins can decrease serum ALB, AST, and ALT enzyme activities and increase antioxidant and nonspecific immune enzyme activities in the liver to enhance the immune capacity of the body.A previous study demonstrated that 0.1% steroidal saponin significantly increased the activities and gene expression of SOD, CAT, GSH-PX, and GR in hybrid grouper liver and serum and decreased the MDA content [21].In the present experiment, compared to the control group, supplementation of 0.1% steroidal saponins to a 14% crude lipid diet significantly increased LZM, ACP, and AKP enzyme activities and decreased ALB, ALT, and AST enzyme activities in the serum of hybrid grouper.
Previous studies on Chinese mitten crab (Eriocheir sinensis) [30] and olive flounder (Paralichthys olivaceus) [31] found that arachidonic acid can enhance the antioxidant capacity of the organism by increasing SOD, CAT, AKP, ACP, and LZM activities in the liver and serum.Three metabolic pathways-cytochrome P450, cyclooxygenase, and lipoxygenase-are all engaged in the metabolism of arachidonic acid [32], in which inhibition of the expression and content of key genes and metabolites of the cyclooxygenase metabolic pathway could improve the anti-inflammatory effects of the body [33].Compared with the control group, the sod, cat, and gsh-px genes' expression in hybrid grouper liver was up-regulated with the addition of 0.1% steroidal saponins, and the levels of PGG2/PGH2, PGA2 and TXA2 metabolites, ptgs-1 and ptgs-2, ptges-1 and ptges-2, and tbxas-2 mRNA significantly decreased.Both KEGG and GO showed that these genes and metabolites were enriched in the cyclooxygenase metabolic pathway.The upstream ptgs-1 and ptgs-2 genes can activate the cyclooxygenase metabolic pathway to release the intermediate products PGG2 and PGH2, which stimulate the downstream ptges-1, ptges-2, and tbxas-2 genes to produce the downstream products PGA2 and TXA2, mediating the inflammatory response [34,35].The present study also demonstrated that the supplementation of 0.1% and 0.2% steroidal saponins in the diets contributed to a decrease in the expression of ptgs-1 and ptgs-2, ptges-1 and ptges-2, and tbxas-2 mRNA in the liver of hybrid grouper, inhibited the expression of key genes in the cyclooxygenase metabolic pathway, and played an anti-inflammatory role in the body's immune system.Research in mice demonstrated that saponins can downregulate ptgs-1 and ptgs-2, ptges-1 and ptges-2, and tbxas-2 mRNA expression in the liver and decrease the levels of PGG2, PGH2, PGA2, and TXA2 metabolites [36,37].Consequently, steroid saponins can protect the liver by effectively removing lipid peroxidation products from the body by maintaining oxidative homeostasis through non-specific immune factors, thus improving the immunity of the fish and enabling the orderly metabolism of nutrients in the fish.However, compared to the S 0.2 group, 0.1% steroidal saponin significantly decreased serum and liver ALB and AST enzyme activities and ptges-1 gene expression and significantly increased LZM, ACP, and AKP enzyme activities.Research on Atlantic salmon (Salmo salar L.) [38] found that the dietary addition of 0.2% soya saponins increased serum SOD and CAT activities, but supplementation of 0.2% or 0.32% soy saponin decreased carnivorous field eels' (Monopterus albus) [39] and Japanese flounder's [40] antioxidant capacity.Research on omnivorous carp [41] found that antioxidant capacity was significantly inhibited when dietary Momordica charantia saponins were provided at levels above 0.64%.Thus, high doses of saponins are toxic to fish and decrease antioxidant capacity.
A previous study showed that supplementation of 0.1% steroidal saponins to highlipid diets significantly increased the activity and gene expression of hexokinase, pyruvate kinase, and phosphofructokinase in the liver and significantly decreased glucose and glycogen content in the liver and serum of grouper [21].Those phenomena indicated that steroid saponins had a potential role in stimulating carbohydrate utilization by the organism by elevating the activity and gene expression of the enzymes mentioned above.In this experiment, analysis of transcriptomic and metabolomic data showed that 0.1% steroidal saponins increased the expression of genes such as gp, pk, and md and metabolites such as G-6-P, F-6-P, and PEP in hybrid grouper liver, which were enriched in the glycolysis pathway, and decreased the expression of genes such as gsk, g6pca2, f-1,6-bp, and pepck and metabolite levels such as PYR and OAA in the gluconeogenesis pathway.It was suggested that steroidal saponins can inhibit the gluconeogenesis pathway by activating the glycolysis pathway and gradually reducing intermediate products, such as pyruvate, a product of glycolysis and a substrate of the gluconeogenesis pathway.At the same time, the ingestion of diets containing 0.1% and 0.2% steroidal saponins up-regulated glut-4 and md gene expression in hybrid grouper liver, while down-regulating gsk, g6pca2, f-1,6-bp, and pepck mRNA expression, consistent with the analysis of omics data.An investigation on grouper and carp found that hepatic expression of hk, gk, and pk genes was up-regulated by saponins and down-regulated by g-6-p, f-1,6-bp, and pepck genes in the liver [42,43].As a result, steroidal saponins can effectively promote the ability of hybrid grouper to utilize carbohydrates in the diet by activating the expression of genes and metabolite content in the glycolysis pathway while inhibiting the gluconeogenesis pathway.
The ingestion of high-lipid diets by grass carp (Ctenopharyngodon idella) (11% and 16% crude lipid) [44,45] and Nile tilapia (Oreochromis niloticus) (12% crude lipid) [46] caused an increase in serum and liver triglyceride and cholesterol levels.Studies on field eel [39], catfish (heterobranchus longifilis) [47], and largemouth bass (Micropterus salmoides) [48] revealed that saponin reduced fas, acc, ppar-γ, and srebp-1 mRNA expression in the liver while elevating the gene expression of cpt-1, cpt-2, acsl, and acadm.In this experiment, supplementation with 0.1% steroidal saponins significantly reduced triglyceride and cholesterol levels in the liver and serum of hybrid grouper [21], decreased the expression of genes such as fas and ppar-γ and metabolite levels such as eicosapentaenoic acid and palmitic acid in the liver, thereby inhibiting fatty acid biosynthesis while elevating the fatty acid β-oxidation pathway in cpt-1, cpt-2, acsl, and acadm gene expression and metabolite levels such as carnitine and citric acid.Meanwhile, the mRNA expression levels of fas, acc, and srebp-1 in the liver significantly decreased, while the gene expression levels of cpt-1, acsl, and acadm increased in the hybrid grouper who were fed a diet supplemented with 0.1% and 0.2% steroidal saponins.Steroid saponins can activate the fatty acid β-oxidation signaling pathway, increase the expression of key genes and metabolite levels, promote fatty acid transport to mitochondria, and, thus, reduce lipid deposition in grouper liver [49].However, previous studies have found that saikosaponin alleviates hepatic steatosis in hybrid grouper by activating the AMPK/PPARα signaling pathway after the fish consume a high-lipid diet (15% crude fat) [20].This was inconsistent with the findings of the present study.Saikosaponin belongs to the triterpenoid saponins, while our experimental steroid saponins belong to the steroid group of saponins.It was possible that the saponin species had different effects on hybrid grouper.

Conclusions
Compared with the control group, the 0.1% of steroidal saponins effectively scavenged lipid peroxidation products and enhanced the immune defense mechanism of fish by inhibiting the levels of metabolites and genes in the cyclooxygenase metabolic pathway.Further, 0.1% of steroidal saponins promoted more efficient utilization of carbohydrates and lipids in the diets by activating the expression of genes and metabolites levels in the glycolysis pathway and fatty acid β-oxidation pathway and inhibiting the gluconeogenesis pathway, facilitating the conversion and deposition of feed proteins in grouper.

Figure 1 .
Figure 1.Transcriptome analysis in hybrid grouper liver.Note: (A) principal component analysis of livers in S 0 and S 0.1 (PCA); (B) number of up-regulated and down-regulated differentially expressed genes (DEGs); (C) volcano plot of DEGs between S 0 group and S 0.1 group.

Figure 2 .
Figure 2. GO and KEGG enrichment of DEGs in fish liver.Note: (A) GO enrichment of DEGs; (B) KEGG enrichment of DEGs; (C) heatmap of DEGs.

Figure 2 .
Figure 2. GO and KEGG enrichment of DEGs in fish liver.Note: (A) GO enrichment of DEGs; (B) KEGG enrichment of DEGs; (C) heatmap of DEGs.

Figure 3 .
Figure 3. Metabolome analysis in hybrid grouper liver.Note: (A) the result of of Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-D analysis was performed with Partial Least Squares Discriminate Analysis enrichment of metabolites.

Figure 3 .
Figure 3. Metabolome analysis in hybrid grouper liver.Note: (A) the result of discriminant analysis of Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA); (B) identification analysis was performed with Partial Least Squares Discriminate Analysis (PLS-DA); (C) KEGG enrichment of metabolites.

Figure 5 .
Figure 5. Efficacy of dietary steroidal saponins on arachidonic acid signaling pathway of experimental fish livers.Note: (a-c): Values from smallest to largest.

Figure 6 .
Figure 6.Efficacy of dietary steroidal saponins on glucose metabolism in hybrid grouper liver.Note: (a-c): Values from smallest to largest.

Figure 5 .
Figure 5. Efficacy of dietary steroidal saponins on arachidonic acid signaling pathway of experimental fish livers.Note: (a-c): Values from smallest to largest.

Figure 5 .
Figure 5. Efficacy of dietary steroidal saponins on arachidonic acid signaling pathway of experimental fish livers.Note: (a-c): Values from smallest to largest.

Figure 6 .
Figure 6.Efficacy of dietary steroidal saponins on glucose metabolism in hybrid grouper liver.Note: (a-c): Values from smallest to largest.

Figure 6 .
Figure 6.Efficacy of dietary steroidal saponins on glucose metabolism in hybrid grouper liver.Note: (a-c): Values from smallest to largest.

Figure 7 .
Figure 7. Effect of dietary steroidal saponins on lipid metabolism in the liver of hybrid grouper.Note: (a-c): Values from smallest to largest.

Figure 7 .
Figure 7. Effect of dietary steroidal saponins on lipid metabolism in the liver of hybrid grouper.Note: (a-c): Values from smallest to largest.

Table 2 .
Effects of dietary steroidal saponins on biochemical indexes of the hybrid grouper serum.

Table 3 .
Effects of dietary steroidal saponins on nonspecial immune indexes of fish serum.

Table 4 .
Efficacy of dietary steroidal saponins on antioxidative indexes of hybrid grouper livers.

Table 4 .
Efficacy of dietary steroidal saponins on antioxidative indexes of hybrid grouper livers.

Table 5 .
Different expressions of DEGs related to glucose lipid metabolism in liver transcriptome of hybrid grouper.

Table 6 .
Diverse levels of metabolites related to glucose lipid metabolism in hybrid grouper.

Table 6 .
Diverse levels of metabolites related to glucose lipid metabolism in liver metabolomic of hybrid grouper.