Metabonomics Study on the Effect of Traditional Chinese Medicines Feed Addition on Growth Performance and Serum Metabolic Prole of Juvenile Chinese Softshell Turtle (Pelodiscus Sinensis Wiegmann) using UPLC–Triple /TOF–MS Analysis

Background (cid:0) Traditional Chinese medicines (TCMs) had been proven to stimulate digestion, promote growth, boost immune response, and reduce inammatory potential in aquaculture, but its action mechanism was still lack of holistic interpretation. Results: The UPLC-Triple-TOF-MS/MS based platform was developed to investigate the serum metabolic prole associated with the growth performance and immunity in Pelodiscus sinensis fed with the diet added with TCMs feeding additive. The nal weight and specic growth rate of TG2 groups was signicantly the highest (p (cid:0) 0.05), and the feed conversion ratio of TG2 groups was signicantly the lowest (p (cid:0) 0.05) compared to the CG group. In the metabolomics assay, a total of 795 (VIP > 1) out of 3089 variable ions were signicantly different between the TG2 group and CG group. 43 potential biomarkers between TG2 and CG groups were nally screened out on the basis of VIP > 1 and p < 0.05. Based on the analysis of MetPA, eleven potential biomarkers and six involved metabolic pathways were signicantly regulated with treatment of TCMs. Phenylalanine, tyrosine and tryptophan biosynthesis and Arginine and proline metabolism were the two most heavily affected pathways (Impact >0.2), while Tyrosine metabolism, Cysteine and methionine metabolism and Arginine biosynthesis were relatively lightly affected pathways. Conclusions (cid:0) The dietary supplementation nineteen TCMs as feed additives could inuence the growth performance of the healthy juvenile P. sinensis, and the optimal dose of TCMs for juvenile turtles was 2% TCMs. The UPLC-Triple-TOF-MS based metabonomics may be a useful tool for elucidating the ecacy and mechanism of complex TCM prescriptions.


Background
The Chinese softshell turtle (Pelodiscus sinensis Wiegmann) is an ancient, subaquatic reptile, which has high edible and pharmaceutical values in China and other Asian countries, such as Korea, Japan and Vietnam [1][2]. Due to its high economic value, the turtle has become a commercially important aquaculture species in China [3], and more than 340,000 tons of the turtles per year were produced in Chinese turtle farms [4]. Similar to other aquaculture species, the P. sinensis has suffered from parasitic, bacterial and viral diseases attacks due to the increase in stocking density that threaten the sustainable development of aquaculture of this species [5]. To control disease outbreaks, Traditional chemical and antibiotic drugs were used indiscriminately, which are easy to cause residual problems in the surrounding environment affecting higher animals [6]. Traditional Chinese medicines (TCM) are cheaper source for therapeutics, have greater accuracy than chemotherapeutic agents, and offer a viable solution for all problems which aquaculture faces today [7].
TCM is gaining more attention all over the world, due to the reliable therapeutic e cacy [8]. Compared with inorganic chemicals or synthetic antibiotics, TCM had been proven to be less toxic, residue free, natural, antibacterial, immune modulators, antioxidant, etc., and is thought to be ideal growth promoters in animal feeding [9][10]. A lot of TCMs have already been reported to stimulate digestion, promote growth, boost immune response, provide antioxidant and antimicrobial properties, and reduce in ammatory potential in pigs, chickens, sh and other animals [11][12][13][14]. As a natural "green" feed additive, TCMs are increasingly used in aquaculture [15][16]. Herbal medicine is considered one of the best ways to boost the sh's growth, immunity and could be suitable alternatives to antibiotics [17][18][19]. The healthy Trionyx sinensis were randomly divided into ve groups, and were fed with basal diet added Astragalus polysaccharides (APS) at the levels of 0.00, 0.25, 050, 0.75 and 1.00% for 40 days. The results showed that the speci c growth rates, the weight gain rate and the survival rate of Trionyx sinensis fed with the diet added with APS were signi cantly increased (P<0.05) [20]. Internal environment of metabolism of TCM was a dynamic process, which accorded with the dynamics, integrity and systematic characteristic of metabonomic [21].
Metabolomics is an important part of systemic biology. Metabonomics is concerned with the quantitative understandings of the metabolite component of integrated living systems and its dynamic responses to the changes of both endogenous and exogenous factors [22][23]. As a powerful analytical platform, metabonomics was a useful tool for elucidating the e cacy of TCMs, and exploring its potential mechanisms and identify potential biomarkers [21][22]. Many analytical tools have currently been employed including direct GC-MS, HPLC-MS and 1H NMR spectroscopy [21][22][23][24]. To the best of our current knowledge, using UPLC-Triple-TOF-MS/MS to research the mechanism of TCMs feed additives on the growth performance and immunity of P. sinensis can offer more useful information for elucidating the e cacy and mechanism of complex TCM prescriptions. Therefore, in this study, we studied the effect of supplementation with nineteen TCMs as feed additives on the growth performance of the healthy juvenile P. sinensis and involved possible mechanism by the UPLC-Triple-TOF-MS/MS.

Materials And Methods
Investigations and protocols were conducted according to the guiding principles for the use and care of laboratory animals and in compliance with Anhui Science and Technology University Institute of Animal Care and Use Committee. The institutional review board approved this procedure. Our study had been submitted to and approved by the Academic Ethics Committee of Anhui Science and Technology University. All sample collection was undertaken in accordance with relevant Academic Ethics Committee of Anhui Science and Technology University guidelines and regulations.
The healthy juvenile P. sinensis were produced by the Anhui Huanghuai Turtle Breeding Co., Ltd. (Bengbu, China). Three hundred the turtles (with 3.92±0.08 g initial mean body weight) were acclimated to laboratory conditions for 7 days before distributed randomly into ve groups. Fifteen turtles per plastic tank (75 cm × 50 cm × 55 cm, water volume: 75 L) were randomly assigned to one of ve experimental groups, in quadruplicate. The 5 treatment groups were as follows: the control group (CG), in which the turtles were received the basal diet . The TCMs treatment groups (TG1, TG2, TG3 and TG4), the basal diet was supplemented with 1.0%, 2.0%, 3.0%, 4.0% of TCMs feeding additive, respectively. During the experiments, water temperature was controlled at 30 ± 1°C using a thermostat, water depth at 25 -30 cm, dissolved oxygen at 7.0 -9.0 mg/L (water were aerated continuously using air compressor), pH at 7.5 -8.5, ammoniacal nitrogen at 0.2 mg/L. Approximately 30% of the water volume in each tank was renewed with fresh water (30 ± 1°C) every 4 days. The surface water and bottom of the tank were cleaned.
The turtles were fed with a formulated diets manufactured by Haining Hexin Feed Co., Ltd. (Haining, China). The basic experimental diets contained approximately 48% crude protein, 3% crude lipid, 1.2% crude ber, 9 % moisture, 16 % ash, 2 % salinity, 5 % Calcium, 3 % gross phosphorus. The basic experimental diets supplemented with 0.0%, 1.0%, 2.0%, 3.0%, 4.0% of TCMs feeding additive, respectively. The basic experimental diets and TCMs feeding additive were mixed homogeneously in a helical mixer, and were stored in a -20 °C freezer. A portion of trial diets were periodically transferred to room temperature for 4 h before feeding, and then mixed with ultrapure water and make it suitable for cold-press extrusion. The 3 mm pellets were molded, and then were used to feed the turtles [3]. The turtles were fed at 3-6% of their body weight per day in two equal meals (09:00 and 17:30). The experimental diets were maintained on for 7 weeks, and nal body weight of the experimental turtles was measured to the nearest 0.01 g using a digital scale. The half of the juvenile P. sinensis of each experimental group was euthanized by decapitation and the blood was sampled as described by Zhou et al [3]. Serum was separated described by Zhou et al., and then stored at -80 °C for later analysis [3] .

UPLC-MS/MS analysis
Twenty serum samples ((four samples per group) were slowly thawed at room temperature for calibration curves and quality control (QC) samples. As a part of the system conditioning and quality control process, a pooled quality control sample (QC) was prepared by mixing equal volumes of all samples.
Metabolites were pro led using a UPLC-Triple-TOF-MS-based platform. Chromatographic separation of the metabolites was performed on a ExionLC TM AD system (AB Sciex, USA) equipped with an ACQUITY UPLC BEH C18 column (100 mm × 2.1 mm i.d., 1.7 µm; Waters, Milford, USA). The sample injection volume was 20 µL and the ow rate was set to 0.4 mL/min. The column temperature was maintained at 40 o C. During the period of analysis, all these samples were stored at 4 o C. The UPLC system was coupled to a quadrupole-time-of-ight mass spectrometer (Triple TOF TM 5600+, AB Sciex, USA) equipped with a electrospray ionization (ESI) source operating in positive mode and negative mode. Data acquisition was performed with the Data Dependent Acquisition (DDA) mode. The detection was carried out over a mass range of 50-1000 m/z.

Data preprocessing and annotation
After UPLC-TOF/MS analyses, the raw data were imported into the Progenesis QI 2.3 (Nonlinear Dynamics, Waters, USA) for peak detection and alignment.
The preprocessing results generated a data matrix that consisted of the retention time (RT), mass-to-charge ratio (m/z) values, and peak intensity. Mass spectra of these metabolic features were identi ed by using the accurate mass, MS/MS fragments spectra and isotope ratio difference with searching in reliable biochemical databases as Human metabolome database and Metlin database. Furthermore, metabolic features detected at least 80 % in any set of samples were retained. After ltering, following normalization procedures and imputation, statistical analysis was performed on log transformed data to identify signi cant differences in metabolite levels between comparable groups. Multivariate statistical, differential metabolites analysis and Network Construction Data were expressed as mean ± S.D, and then analyzed using single-factor (ANOVA) to assess the effect of TCMs supplementation on the growth responses and serum metabonomics of P. sinensis. The signi cance of differences between the experimental groups has been compared by Duncan's multiple range tests using the software SPSS 22.0, and P < 0.05 was considered signi cant. A multivariate statistical analysis was performed using ropls (Version1.6.2) R package from Bioconductor on Majorbio Cloud Platform (https://cloud.majorbio.com). Principle component analysis (PCA) using an unsupervised method was applied to obtain an overview of the metabolic data, general clustering, trends, or outliers were visualized. All of the metabolite variables were scaled to unit-variances prior to conducting the PCA. Orthogonal partial least squares discriminate analysis (OPLS-DA) was used for statistical analysis to determine global metabolic changes between comparable groups. All of the metabolite variables were scaled to Pareto Scaling prior to conducting the OPLS-DA. Variable importance in the projection (VIP) was calculated in OPLS-DA model.
Statistically signi cant among groups were selected with VIP value more than 1 and p value less than 0.05. Differential metabolites among two groups were summarized , and mapped into their biochemical pathways through metabolic enrichment and pathway analysis based on database search (KEGG, http://www. genome.jp/kegg/). These metabolites can be classi ed according to the pathways they involved or the functions they performed. Enrichment analysis was usually to analyze a group of metabolites in a function node whether appears or not. The principle was that the annotation analysis of a single metabolite develops into an annotation analysis of a group of metabolites. Scipy.stats (Python packages) was exploited to identify statistically signi cantly enriched pathway using Fisher's exact test. The identi ed biomarkers were subsequently con rmed by the p-values with a critical value of 0.05 from a Student's t-test. Receiver operating characteristics (ROC) curves were plotted and areas under the curve (AUC) were calculated to evaluate the accuracy of the metabolic biomarkers in distinguishing different groups. The sensitivity and speci city of the trade-offs were calculated for the selected metabolites by using the area under the ROC curve (AUC) [25]. The construction, interaction and pathway analysis of these identi ed potential biomarkers were performed with MetPA [26]. In addition, the network of potential biomarkers-metabolic pathways-targets was constructed, and the corresponding network analysis was conducted using Cytoscape 3.5.1 software (http://www.cytoscape.org/) .

Growth performance
The results of the TCMs feed additives on the growth performance indices of Pelodiscus sinensis is presented in Table1. The survival rates of the TG1, TG2,  TG3, TG4 and control groups were 93.333%, 94.998%, 91.665%, 88.333% and 90.000%, respectively. The survival rate of TG2 group was the highest but no signi cant difference from other treatments. Compared with the CG group, the TG2 group supplemented with 2% TCMs feed addition showed the highest nal weight and speci c growth rate (SGR) (p 0.05), the TG1 group supplemented with 1% TCMs feed addition gave higher nal weight and SGR (P<0.05), the TG3 group supplemented with 3% TCMs feed addition increased nal weight and SGR, whereas TG4 group supplemented with 4% TCMs feed addition showed lower nal weight and SGR (P<0.05). The TG1, TG2 and TG3 groups showed signi cantly (p 0.05) lower feed conversion ratio (FCR) compared to the CG group (Table 1).

Method validation of LC-MS analysis and potential biomarkers responsible
In total, 3089 variables (1845 peaks in ESI + mode, and 1244 peaks in ESImode) were identi ed in serum samples from CG, TG1, TG2, TG3, and TG4 groups for subsequent analyses (Supplement Table S1). Due to individual differences, it was di cult to directly observe the changes of serum metabolites in the ve groups. Thus, we performed multivariate data analysis to determine the metabolic markers in P. sinensis. The PCA performed on the whole samples revealed that the QC samples were tightly clustered in PCA score plots (shown in Figure S1), which indicated that the system stability was accommodative for this metabonomic study. Figure 1 (A and B) showed PCA scores plot of serum, but the discriminations of ve groups were not very distinct. In positive ion mode of UPLC-Triple-TOF-MS/MS analysis ( Figure 1A), the metabolic pro le of the TG2 group deviated away from the control group along both PC1 and PC2, which suggested that apparent biochemical changes were induced by TG2 treatment. The TG4 group was far away from the TG2 group, and much closer to the control group (Figure 1), indicating that the metabolic disturbances induced by TG2 were signi cantly improved with TG4 treatment. Similar results were observed when TG1 and TG3 were subtracted (TG1 and TG3 groups), except that the metabolic pro le of P. sinensis in the TG4 group was much closer to the control group than that in the TG1 and TG3 groups. Thus, the OPLS-DA model was established to maximize the distinction between groups (Figure 2 A, B).
Meanwhile, variable importance in the projection (VIP) was used to identify the potential biomarkers obtained from ve groups.
The VIP values were larger than 1 and the p values of an independent test < 0.05 as the potential biomarkers. According to the results of OPLS-DA, a total of 795 (VIP > 1) out of 3089 variable ions were signi cantly different between the TG2 group and CG group. 43 potential biomarkers between TG2 and CG groups were nally screened out on the basis of VIP > 1 and p < 0.05 ( Table 2). As shown in Table 2, compared with the control group, twenty seven metabolites were up-regulated, and sixteen metabolites were down-regulated by TCMs stimulus. Further, hierarchical cluster analysis of resulting 30 identi ed metabolites revealed the existence of distinct differences between the control and TG2 group (Figure 3). In this study, we tentatively used the term''potential biomarkers'' to acknowledge their potential value and simultaneously to indicate its uncertainty.

Metabolic Pathways, biological pathway and network analysis
The 43 identi ed potential biomarkers were furthered con rmed by the univariate ROC curve analyses (AUC > 0.8) and Student's t-tests (p < 0.05). Nineteen potential biomarkers were ultimately considered to exhibit the greatest su cient utility for discrimination of the TG2 and CG groups ( Figure S2). To further investigate the recovery condition of the 19 potential biomarkers, the relative peak areas between the four TCM-dosed groups and the control group were tested based on univariate ROC curve analyses and Student's t-tests (Table S2). For TG1 group, only 4 biomarkers were increased, 3 biomarkers were decreased comparing with the TG2 group. Based on the 19 identi ed potential biomarkers between TG3 and TG2 groups, the level of Salicylaldehyde and 2-(5-Methyl-2-furanyl)-3-piperidinol were improved, while the level of Ampeloside Bf2, Benzofuran and 2-Hydroxycinnamic acid were decreased. However, for the TG4 group, only one biomarker was decreased comparing with TG2 group. Similarly, fty of the total metabolites were identi ed when compared with the metabolomes among the CG, TG1, TG2, TG3 and TG4 groups, as shown in the Heatmap visualization ( Figure 4).
Based on the 43 identi ed potential biomarkers between TG2 and CG groups, the relevant metabolic pathways were assigned using  (Table 3). Five different metabolic pathways with Impact >0.05, including Phenylalanine, tyrosine and tryptophan biosynthesis, Arginine and proline metabolism, Tyrosine metabolism, Cysteine and methionine metabolism and Arginine biosynthesis (Table 3). Phenylalanine, tyrosine and tryptophan biosynthesis with the impact-value 0.5 and Arginine and proline metabolism with the impact-value 0.2039 (Impact >0.2) were the two most heavily affected pathways, while Tyrosine metabolism, Cysteine and methionine metabolism, and Arginine biosynthesis were relatively lightly affected pathways ( Figure 5), which suggested that ve key pathways were considered the most relevant pathways involved in the growth promoting agents of TCMs. The ingenuity network analysis was employed based on MetPA to evaluate the impact on these pathways after different TCM treatments ( Figure 3S).
Integrating the analysis of KEGG, thirteen different metabolic pathways were all demonstrated to have the most impacts in the analysis of metabolic pathway, which were recognized as the important metabolic pathways in the growth performance and immunity of P. sinensis ( Figure 6). Correspondingly, ve key metabolic pathways (Phenylalanine, tyrosine and tryptophan biosynthesis, Arginine and proline metabolism, Tyrosine metabolism, Cysteine and methionine metabolism, and Arginine biosynthesis) were considered as the most relevant compounds involved in TCMs, which could be serve as potential biomarkers of growth promotion. Taking the different metabolites and metabolic pathways into consideration, we found that TG2 group supplemented with 2 % TCMs increased mainly the amino acid contents, whereas TG4 group added to 4 % TCMs reduced mainly the contents of amino acid and fatty acids in P. sinensis. Overall the results suggest that TCMs could improve protein synthesis and regulate immune function, and the most effective concentration was 2 %.

Discussion
Effects of dietary TCMs on growth of P. sinensis Feed additives were products used in animal nutrition for purposes of getting better the quality of feed and the quality of food from animal origin, or to improve the animals' performance and health [10]. In aquaculture survival and growth rate of the selected species play vital role in maintaining a commercially viable farm [27]. In recent years, TCMs were generally considered safe and natural "green" feed additive, which were increasingly used in aquaculture [15][16]. Applications of compound formulations in aquaculture were believed to be more acceptable and bene cial as different herbs with diverse mechanisms showed complementary effects between herbs in the formulation [10,12]. TCMs are added in sh feed to improve feed conversion e ciency that result in sh growth [28]. Many studies showed that inclusion of herbs in sh diet has positive effect on growth and disease free shes [29][30]. The aim of this study was to elucidate the effect of supplementation with nineteen TCMs as feed additives on the growth performance of the healthy juvenile P. sinensis. The nal weight and SGR of TG1 and TG2 groups were signi cantly higher (p 0.05), and the FRC of TG1 and TG2 groups were signi cantly lower (p 0.05) compared to the CG group. However, the nal weight and SGR of TG4 group supplemented with 4% TCMs feed addition were the lowest (P<0.05), and the FRC of TG4 group was the highest compared to the CG group (P<0.05) ( Table 1). The present results suggested that supplementation of 1% -3% TCMs feed additives could improve feed utilization rate and promote the growth of juvenile P. sinensis, and the optimal dose of TCMs for juvenile P. sinensis was 2% TCMs feed additives, which might be related to the wide variety of active components in TCMs. The naturally occurring components of TCM and/or their degradation products in the gastrointestinal tract might be metabolized by metabolic enzymes in the intestinal epithelium, and/or in uence their activities. After the TCM components were absorbed into the bloodstream through the intestinal epithelium, they were rst delivered to the liver via the portal vein [31]. Most of the herbs and spices could stimulate the function of pancreatic enzymes, some also increase the activity of digestive enzymes of gastric mucosa, and enhance gastrointestinal digestion and absorption capacity [9]. Thus, our results were consistent with the previous literature reports [20,[32][33][34].

Identi cation of potential biomarkers associated with TCMs feeding additive
Metabolomics is one of the high-throughput "omics" techniques, which beside genomics, proteomics, and transcriptomics play an important role in systems biology [35]. Metabolomics is de ned as the quantitative measurement of the dynamic multiparametric metabolic response of living systems to pathophysiological stimuli or genetic modi cation [36], which has been proposed as a powerful tool for exploring the global metabolic state of the entire organisms, and is coincident with the integrity and systemic features of TCM [22]. Therefore, in this study, the integrated metabolic approach was applied to elucidating the effect of supplementation with TCMs as feed additives on the metabonomics of the healthy juvenile P. sinensis. 43 potential biomarkers between TG2 and CG groups were nally screened out on the basis of VIP > 1 and p < 0.05 (Table 2), twenty seven metabolites were up-regulated, and sixteen metabolites were down-regulated by TCMs stimulus, which suggested TCMs could effectively regulate the metabonomics of the healthy juvenile P. sinensis supplemented with 2% TCMs feed addition, and improved the growth performance of P. sinensis. Thirteen pathways related to the metabolites of TG2 group compared with the control group were shown by the analysis of MetPA ( Figure 6 and Table 3), which were mainly involved in amino acid metabolism (Aminoacyl-tRNA biosynthesis, Phenylalanine, tyrosine and tryptophan biosynthesis, Ubiquinone and other terpenoid-quinone biosynthesis, Arginine biosynthesis, Glutathione metabolism, Cysteine and methionine metabolism, Tyrosine metabolism, Phenylalanine metabolism), nucleotide metabolism (Purine metabolism, Pyrimidine metabolism) and lipid metabolism (Sphingolipid metabolism, Glycerophospholipid metabolism). Among them, eleven potential biomarkers, including L-Methionine, L-Proline, L-Tyrosine, Ornithine, 2-Hydroxycinnamic acid, SM(d18:0/16:1(9Z)(OH)), LysoPC(22:1(13Z)), 5-Methylcytosine, Deoxycytidine, Adenosine 3'-monophosphate and Xanthosine were screened to characterize the potential biomarkers of TCMs. Among thirteen metabolic pathways, Phenylalanine, tyrosine and tryptophan biosynthesis with the impact-value 0.5, Arginine and proline metabolism with the impactvalue 0.2039, Tyrosine metabolism with the impact-value 0.1397, Cysteine and methionine metabolism with the impact-value 0.1045 were ltered out as the most heavily affected metabolic pathways, because the pathway with the impact-value threshold above 0.10 was regarded as potential target pathway involved in the growth promoting agents of TCMs [37].

Phenylalanine, tyrosine and tryptophan biosynthesis
Amino acids play important roles in many metabolic pathways as main substrates and as regulators [38]. In the current study, Figure 6 shows the network of the potential biomarkers changing for TG2 group according to the KEGG Pathway database, and peptides were found to be the main contents associated with TCMs, which mainly involved in ve important metabolism pathways ( Figure 5 and Figure 6). In TG2 group, improved levels of amino acids (L-Methionine, L-Proline, L-Tyrosine and Ornithine) suggested a high demand and rapid utilization of metabolites to protein producing pathways (Figure 6), which may be relate to TCMs formulation contained different active ingredients that may have a synergistic effect on growth, combating viral infection, appetite stimulation, and stress relief [12]. Phenylalanine and tryptophan are two essential amino acids which cannot be synthesized in vivo and can be obtained only from the daily foods, they are the precursors of tyrosine and serotonin, respectively [38]. Phenylalanine which is one of the essential aromatic amino acids, under normal circumstances, can be used as amino acids to produce various proteins in cells of the body tissue. The steady state of phenylalanine metabolism can maintain the body's normal growth, development and physiology [39]. Tyrosine is an aromatic amino acid that serves as a precursor for a variety of biologically important substances; e.g., melanin pigments, catecholamines, thyroid hormones, and protein [40]. Tyrosine production via phenylalanine hydroxylation by phenylalanine hydroxylase is a major metabolic pathway for phenylalanine. Phenylalanine and its metabolite tyrosine, as catecholamine precursors, participate in dopamine synthesis and are associated with excessive stimulation of the sympathetic nervous system [41]. A signi cant fate of tyrosine is conversion into catecholamine, e.g., dopamine, norepinephrine and epinephrine [42]. Based on the present study, L-Tyrosine, Ornithine and L-Proline were all up-regulated prominently in P. sinensis supplemented with 2% TCMs feed addition. In the role of L-tyrosine, L-phenylalanine, glutathione and coenzyme could produce fumaric acid, which was involved in the TCA, and nally promoted the metabolism of hemoglobin and heme [22]. The up-regulated trend of L-tyrosine in TG2 group could be indicative for the improvement in phenylalanine metabolism which is associated with the phenylalanine 4hydroxylase, a responsible enzyme for phenylalanine conversion into tyrosine [43]. The content of L-tyrosine was increased in the TG2 group compared with CG group, which suggested that the synthesis of glutathione was improved because of the inhibition of glucose-6-phosphate dehydrogenase activity after P. sinensis supplemented with 2% TCMs feed addition [22], therefore, TCMs could contribute to increasing the metabolism of hemoglobin.
Arginine and proline metabolism Arginine and proline are two important fatty acid amides (FAAs) that play important roles in osmotic stress [44]. It has been reported the arginine and proline metabolism was closely related to the progression of oxidative stress [45]. Arginine could exert its potential protection from the gastric mucosal damage through inhibition of oxidative stress derived via xanthine-XO [46]. Arginine is synthesized from citrulline through the sequential action of the cytosolic enzymes argininosuccinate synthetase (EC.6.3.4.5, ASS) and argininosuccinate lyase (EC.4.3.2.1, ASL) [44]. Ornithine is an amino acid that may be generated from glutamic acid or produced in the urea cycle by the hydro lysis of urea from arginine [47]. Glutamic acid is one of the 20 proteinogenic amino acids and a key molecule in cellular metabolism. Meanwhile, proline and citrulline, two metabolites of arginine, were also involved into the progression. In animals, intracellular proline levels are mainly controlled via biosynthetic and catabolic pathways [44,48]. Proline, the only proteinogenic secondary amino acid, is metabolized by its own family of enzymes responding to metabolic stress and participating in metabolic signaling. Metabolism of proline generates electrons to produce ROS and initiates a variety of downstream effects, including blockade of the cell cycle, autophagy, and apoptosis [49]. Under osmotic stress condition, proline is synthesized and degraded mainly through the glutamate metabolism pathway [48]. In this work, the increase of proline and arginine, were observed in TG2 group (Table 2), which may be explained due to the accumulation of proline, as a compatible solute, resulting in an increase in cellular osmolarity that can drive the in ux of water or reduce its e ux [44]. Arginine and proline metabolism and cysteine and methionine metabolism has previously been found to be closely associated with metabolic syndrome including dyslipidemia, obesity, hypertension and elevated plasma glucose level [49]. The analysis of MetPA proved the importance of arginine and proline metabolism involved into the formation and development of TCMs.

Cysteine and methionine metabolism
To all organisms, L-cysteine and L-methionine are important as sulfur-containing amino acids. Animals uptake these amino acids as the sulfur sources, and catabolizes them to various cellular components [50][51]. Methionine occupies a central position in the cellular metabolism, as it is involved in the protein synthesis and synthesis of S-adenosylmethionine, the primary source of the methyl groups in the cell. Methionine biosynthesis involves three metabolic pathways: carbon backbone, sulfur utilization, formation and methylation [52]. L-Methionine is converted to S adenosylmethionine under catalysis by methionyladenosine transferase, which is produced by synthesis of S-adenosine homocysteine catalyzed by S-adenosylmethionine transylase [53]. The increase in S-adenosine homocysteine level affected the cystathionine metabolic pathway of methionine, which ultimately affected the production of its metabolite taurine [22]. Taurine (2-aminoethanesulfonic acid), a sulfur-containing amino acid, is one of the most abundant free amino acids in many tissues [54]. Taurine metabolism occurs mainly via two pathways: through synthesis from cysteine and through taurine uptake facilitated by a high-a nity transport system [44]. In bivalves, the system involved in taurine synthesis from cysteine is well-documented [55]. Synthesis of cysteine as a product of the transsulfuration pathway can be viewed as part of methionine or homocysteine degradation, with cysteine being the vehicle for sulfur conversion to end products (sulfate, taurine) that can be excreted in the urine [56]. The lack of methionine intake make the organ index decreased which would result in immunosuppression [57]. Compared to CG group, the content of methionine in P. sinensis of TG2 group was increased, which indicated TCMs supplementation to the basic experimental diets could improve methionine intake of P. sinensis, and increase immune-stimulants and the resistance to tumors. In the present study, TCMs could enhance the formation of taurine by regulating the content of L-methionine in P. sinensis. The main enzyme for taurine synthesis is mainly distributed in the liver, which is considered to be the main organ capable of taurine synthesis, suggesting that taurine may play vital roles in the liver protection [58]. Methionine, which is vital for mammalian metabolism, such as one-carbon metabolism, the transsulfuration pathway, and protein synthesis, is an essential, sulfur-containing amino acid required for human and animal health [59].

Conclusion
In summary, compared with CG group, TG1, TG2 and TG3 gave higher nal weight and SGR (P<0.05), whereas TG4 group showed lower nal weight and SGR (P<0.05), and the optimal dose of TCMs for juvenile P. sinensis was 2% TCMs feed additives. 43 potential biomarkers between TG2 and CG groups were nally screened out on the basis of VIP > 1 and p < 0.05, twenty seven metabolites were up-regulated, and sixteen metabolites were down-regulated. Eleven metabolites mainly involved in amino acid metabolism, nucleotide metabolism and lipid metabolism. These results illustrated that Phenylalanine, tyrosine and tryptophan biosynthesis, Arginine and proline metabolism, Tyrosine metabolism and Cysteine and methionine metabolism were regarded as potential target pathways involved in the growth promoting agents of TCMs (Impact >0.10). It is concluded that TCMs can be used as feeding additive for better growth of healthy juvenile P. sinensis. Weight gain (WG %) =100× ( nal body weight−initial body weight)/initial body weight.
Feed conversion ratio (FCR) = G d /G f -G i , G d : Dry feed intake, G f : Final weight, G i : Initial weight.
Survival rate (%) = 100×P. sinensis number in each group remaining at the end of the experiment/initial number of P. sinensis. a Change trend compared with control group. The "Total" is the number of compounds in the pathway; the "Hits" represents the actual matched number from the user uploaded data; the Raw p is the original p value calculated from the enrichment analysis; the Holm p is the p value adjusted by Holm-Bonferroni method; the FDR p is the p value adjusted using False Discovery Rate; the Impact is the pathway impact value calculated from pathway topology analysis.   The metabolic pathway network of the potential biomarkers for TG2 group according to the KEGG PATHWAY database. The blue boxes are the most heavily affected metabolic pathways found.