Role of dietary Schizochytrium sp. in improving disease resistance of zebrash through metabolic and microbial analysis

a of as a promising alternative to for and fatty acid prole in a comprehensive understanding of disease resistance on microalgae supplementation of In the present study, to understand the pathogen-resistant mechanisms of diets enriched with microalgae, the e ﬀ ects of dietary Schizochytrium sp. on the intestine microbial and metabolic prole of zebrash investigated. After being infected with E. piscicida, the cumulative survival rates of zebrash were 87.5% and 70% when fed with 120 g/kg and 60 g/kg Schizochytrium sp., respectively. These results are higher than the survival rate of sh without Schizochytrium sp. supplementation (62.5%; Fig. 1). No deaths were recorded in sh from the negative control group that was injected with saline solution. These results indicate that Schizochytrium sp. diets could protect zebrash from E. piscicida infection. TCA cycle:tricarboxylic acid

oil and could enhance sustainability in aquaculture [6]. Studies have shown that marine microalgae diets that are rich in long-chain polyunsaturated fatty acids (LC-PUFA), such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), can improve the growth, nutritional performance, and fatty acid pro le of sh [3,7,8]. Moreover, Sheikhzadeh et al. have found that dietary supplementation with 2.5% Spirulina platensis could improve mucosal immune responses and disease resistance in rainbow trout [9]. Schizochytrium sp., a type of marine microalgae, has been preferred as a potential replacement for sh oil because of its ease in large-scale heterotrophic cultivation and high lipid content [10]. Our previous study found that Schizochytrium sp. is rich in LC-PUFA, especially DHA (above 40% of total lipids) [11]. To our knowledge, many researchers have focused on improving growth and the ratio of LC-PUFA with dietary containing with Schizochytrium sp. both in juvenile and adult sh, such as the Atlantic salmon (Salmo salar) [12,13], Paci c white shrimp (Litopenaeus vannamei) [14], channel cat sh (Ictalurus punctatus) [15], seabream (Sparus aurata) [16], and zebra sh (Danio rerio) [8]. Nevertheless, few studies have shown that dietary Schizochytrium sp. signi cantly improves the non-speci c immunity [17], and modulates the intestinal microbiota of both the Nile tilapia (Oreochromis niloticus) [18] and the rainbow trout (Oncorhynchus mykiss) [19]. Therefore, there is a need for a more comprehensive and more in-depth understanding of the Schizochytrium sp. supplementation's internal effects on host health.
Recently, technologies such as high-throughput sequencing and metabolomics have considered being a rapid expansion in research related to sh immunity [20]. It is well known that the intestinal microbiota plays essential roles in nutrient distribution, regulation of innate immunity, and maintenance of intestinal tissue integrity, which can be modulated by diets [21]. Because of the central role played by metabolism in immunity, metabolomics is rapidly established as a critical analytical tool in immunity studies [22].
However, few studies have been conducted in an integrated manner to evaluate the in uence of dietary microalgae on metabolism and microbiota in sh.
In the present study, zebra sh was selected as the model organism to evaluate the effect of Schizochytrium sp. in the diet on the metabolic pro le and intestinal microbiota. A 56-day feeding trial using a diet enriched with Schizochytrium sp. was administered daily to zebra sh. Metabolomic pro ling of the humoral uid and the intestinal microbial community structure were examined by gas chromatograph-mass spectrometer (GC-MS) analysis and 16S rDNA gene sequencing technology, respectively. Additionally, zebra sh were challenged with Edwardsiella piscicida (E. piscicida) to verify the disease resistance of Schizochytrium sp. enriched diets. Our results provide a novel insight into developing the fundamental understanding of metabolic responses and disease resistance in zebra sh fed with microalgae.

Materials And Methods
Microalgal strains and culture conditions Schizochytrium sp. ATCC MYA 1381 was maintained in our laboratory. The fermentation process was consistent with our previous study [11]. In brief, the seed culture (10% v/v) was transferred to a 3 L fermenter and incubated at 28 °C for 10 days. The biomass was harvested by centrifugation (12,000 × g, 10 min), and then dried in a vacuum freeze dryer. The resulting cells were stored at -20 °C until further use.

Diet preparation
Three iso-nitrogenous, iso-energetic, and iso-lipidic experimental diets were prepared according to Sarker et al., [23] with different Schizochytrium sp. content: 0 g/kg (C), 60 g/kg (S1), and 120 g/kg (S2), respectively. Details of the feed formula are provided in Supplemental Table 1. We used the whole dried cells of Schizochytrium sp. instead of sh oil in the S1 and S2 groups. The diet preparation followed the methods by Duan et al. [20]. Crude protein, lipid, and ash of the trial diets were measured by the guidelines of the Association of O cial Analytical Chemists.
Zebra sh treatment and sample collection Healthy juvenile zebra sh (body weights of approximately 0.59 g) were purchased from Guangzhou Flower Bird Fish and Bug market (Guangzhou, China). The zebra sh were acclimatized to the trial conditions for 14 days before conducting the feeding trial. The water quality parameters were as follows: a temperature of 25 ± 0.5 °C, pH 7.8, and 6.0 ± 0.5 mg/L of dissolved oxygen. The zebra sh were fed 4% of their body weight daily. Uneaten food and feces were removed from the tanks, and one-third of the culture water was renewed with fresh water once a day.
After acclimation, sh were split into three groups with replicate tanks (48 L) containing 50 sh per tank and fed with the corresponding experimental feed. The culture conditions were identical to those of the acclimation stage. The sh were fed twice a day, namely at 08:00 and 17:00, for 56 days. On day 56, eight sh from each group were sampled for GC-MS analysis. Intestines from three individuals of each group were pooled together as one sample to reduce individual variation. Three replicates were performed for the microbiome analysis.

Challenge experiment
Eighty sh from each group were challenged with E. piscicida isolated from diseased Nile tilapia in southeastern China. These sh were infected by an intraperitoneal injection of E. piscicida at 6 × 10 3 CFU/ sh. The concentration of bacteria selected was identi ed through previous experiments that the 96h median lethal concentration (96 h LC 50 ) of adult zebra sh exposure to E. piscicida is 8 × 10 3 CFU/ sh. The symptoms were observed twice daily for 144 h, and the cumulative deaths were collated. As a negative control, 20 sh from each group were challenged with a saline solution and maintained in aquarium water. This experiment involved the same consistent feeding strategy as the other experiments. Survival curves were estimated with Kaplan-Meier's analysis, using SPSS software (SPSS v. 20.0: SPSS, Chicago, IL, USA). The experimental protocols were permitted by the Institutional Animal Care and Use Committee of Xiamen University, Xiamen, China.

Histological analysis
The foreguts from three sh per tank were sampled on the 56th day and xed for 24 h in 4% paraformaldehyde solution. They were then dehydrated, equilibrated, and embedded in para n wax, following standard histological techniques [24]. The tissue was sectioned to a 4 µm thickness with a microtome (Leica, RM2016, Wetzlar, Germany). After being stained with hematoxylin and eosin (HE), the sections were observed and photographed under a light microscope (Leica DM500, Wetzlar, Germany).

Sample preparation for GC-MS
Zebra sh were cut on ice into pieces and weighed. Pieces from each individual were placed in tubes with 1 mL/g sterile saline, then mixed by vertexing, and centrifuged for 10 min at 4 °C to collect humoral uid. The GC-MS sample preparation was performed as described in Jiang et al., [25] with moderate modi cations. Brie y, 200 µL ice-cold methanol was added to 100 µL humoral uid to quench the metabolites, and then added 3 µL internal standard (100 µg/mL ribitol) (Sigma, St. Louis, USA). After centrifugation (12,000 × g, 15 min), the transferred supernatant was dried in a vacuum freeze dryer for 24 h. The resulting pellets were applied to further analysis.

GC-MS analysis
Derivatization was performed according to previously described [26]. First, the dried pellet was dissolved in 80 µL of 20 mg/mL methoxyamine hydrochloride (Sigma, St. Louis, USA) containing pyridine and incubated for 1.5 h at 37 °C. Subsequently, 80 µL N-methyl-N-trimethyl-silyl-tri uoroacetamide (MSTFA, Sigma, St. Louis, USA) was added and incubated for 0.5 h at 37 °C. The samples were then centrifuged at 12,000 × g for 10 min at 4 °C. Chemical analysis of the samples was performed using the Agilent 7890-5975C GC-MS solution system (Agilent, Sacramento, California, USA). The injector temperature was held at 270 °C. Using the splitless model, 0.5 µL of the derivatized sample was injected into a Dodecyl Benzene Sulfonate (DBS) column of 30 m in length, 250 µm in inner diameter, and 0.25 µm in thickness. The temperature program of the GC oven is 85 °C for 5 min. The temperature is raised to 285 °C at 5 °C/min, and nally to 310 °C at a rate of 20 °C/min. Helium was used as carrier gas with a ow rate of 1 mL/min. The scanning mass range was set at 50-600 m/z, and the ionization energy of the electron impact ionization (EI) was 70 eV.

GC-MS data processing and statistical analysis
The original GC-MS data were conducted as previous reports [27]. Each peak was identi ed by alignment with the mass spectra from the NIST 2.2 (National Institute of Standards and Technology, USA) library. The data of the identi ed metabolites were normalized and imported into SIMCA software (version 14.1; Umetrics, Umeå, Sweden) for principal component analysis (PCA) and orthogonal partial least-squares discriminant analysis (OPLS-DA). Differential metabolites were identi ed with variable importance in the projection (VIP) > 1.0 obtained from OPLS-DA and p-values < 0.05 obtained from two-tailed Student's ttests. The pathway enrichment analysis of differential metabolites was conducted with MetaboAnalyst 4.0 [28].

Intestinal microbiome analysis
Intestinal microbial DNA was extracted and puri ed by HiPure Stool DNA Kit (Magen, Guangzhou, China), followed by the manufacturer's protocols. For microbial analysis, the primers 341F (CCTACGGGNGGCWGCAG) and 806R (GGACTACHVGGGTATCTAAT) were used to amplify the 16S rDNA V3-V4 region. The amplicons were puri ed by the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA).
Puri ed products were pooled in equimolar amounts and paired-end sequenced (PE250) on an Illumina MiSeq platform. UPARSE (version 9.2.64) was used to truncate the clustering of operational taxa units (OTUs) with 97% similarity, and the UCHIME algorithm was used to identify and delete abnormal gene sequences. The representative sequence was classi ed into organisms for each OTU according to the SILVA database (version 132), using the RDP classi er (version 2.2) with 90% con dence. The Chao and Shannon indices were determined using QIIME (version 1.9.1), following previous research protocols [29].
Non-metric multidimensional scaling (NMDS) was performed based on the unweighted UniFrac distances. The most differentially abundant taxon between the groups was identi ed by the linear discriminant analysis (LDA) of effect size (LEfSe) method.
Correlation analysis between bacterial taxonomy and differential metabolites Spearman's correlation analysis assessed the relationship between differential metabolites and bacterial taxonomy (at the genus level) with the psych package in R. p values was adjusted with a false discovery rate (FDR). Any adjusted p-value of less than 0.05 was statistically signi cant.

Results
Effect of Schizochytrium sp. supplementation on sh response to E. piscicida infection To investigate whether Schizochytrium sp. affected the host's susceptibility to bacterial infection, zebra sh were infected with E. piscicida (6 × 10 3 CFU/ sh). After being infected with E. piscicida, the cumulative survival rates of zebra sh were 87.5% and 70% when fed with 120 g/kg and 60 g/kg Schizochytrium sp., respectively. These results are higher than the survival rate of sh without Schizochytrium sp. supplementation (62.5%; Fig. 1). No deaths were recorded in sh from the negative control group that was injected with saline solution. These results indicate that Schizochytrium sp. diets could protect zebra sh from E. piscicida infection.
Intestinal histological structure in zebra sh fed with Schizochytrium sp. diets The intestinal morphology of zebra sh fed with Schizochytrium sp. diets is presented in Fig. 2. Our results demonstrated that intestinal morphology was changed by the supplement of Schizochytrium sp.
The density of goblet cells increased in zebra sh fed with Schizochytrium sp. as a diet supplement when compared with the control group.

Effects of Schizochytrium sp. supplementation on metabolite pro les in zebra sh
There was a clear separation between the three groups (the zebra sh fed two different doses of the supplement and one control without Schizochytrium sp.) in the PCA score plot (Fig. 3A). Further models of OPLS-DA between zebra sh fed with or without Schizochytrium sp. were conducted (Fig. 3B-D). Three permutation tests indicated that all OPLS-DA models contained high discrimination and predictive capability (Supplemental Figure S1).
We used the criteria of VIP > 1 and p < 0.05 for multivariate and univariate statistical analyses, respectively, to identify 17 differential metabolites between the zebra sh fed with 0 and 120 g/kg Schizochytrium sp. Similarly, 12 differential metabolites between the zebra sh fed with 0 and 60 g/kg Schizochytrium sp. Moreover, six differential metabolites existed between the zebra sh fed with 60 and 120 g/kg Schizochytrium sp. All the differentially abundant metabolites in any two groups are visually displayed in a heatmap plot (Fig. 3E). A cross-comparison Venn diagram illustrated the variation among the three collections of differential metabolites to further characterize the metabolites involved in immune resistance. All the differential metabolites formed 7 clusters (Supplemental Fig. 2). Detailed descriptions of these metabolites are included in Supplemental Table 2.
Effects of Schizochytrium sp. supplementation on the intestinal microbiota of zebra sh The nine most dominant phyla were Proteobacteria, Fusobacteria, Actinobacteria, Planctomycetes, Firmicutes, Bacteroidetes, Tenericutes, Verrucomicrobia, and Chloro exi (Fig. 5A). Bacteroidetes were less abundant in zebra sh fed with 120 g/kg Schizochytrium sp. supplemented diet than the control group (Fig. 5B). The NMDS showed that both the Schizochytrium sp. supplemented diets were separated from the control group with nmds1 (Fig. 5C). The intestinal microbial compositions of zebra sh fed with Schizochytrium sp. were completely different from that fed without Schizochytrium sp. However, the Chao index and Shannon index indicated no statistically signi cant difference among the three groups (Supplemental Fig. 3).

Correlations between differential metabolites and bacterial genera
Spearman's correlation analysis was performed to investigate the relationship between the differential metabolites and the 37 bacterial genera that were statistically changed in the group fed with 120 g/kg Schizochytrium sp. when compared to the group fed without Schizochytrium sp. The genus Sediminibacterium was negatively correlated with metabolites such as taurine, L-serine, citric acid, creatinine, D-glucose, D-galactose, phosphorylethanolamine, and myo-inositol (Fig. 6). Flavobacterium and Shewanella were negatively correlated with taurine and pantothenic acid, while the former was also negatively associated with L-serine and citric acid. In contrast, Butyricicoccus, Dorea, Bradymonas, and Actinomycetospora, were positively correlated with ve metabolites: taurine, creatinine, D-glucose, Dgalactose, and pantothenic acid. In addition, Dorea, Bradymonas, and Actinomycetospora also showed a positive correlation with L-serine and citric acid.

Discussion
In recent decades, different microalgae species have been tested for use in aquaculture applications [8,9]. Microalgae contain many bioactive components, such as omega-3 fatty acids, β-glucans, and avonoids. These nutrients may modulate sh physiology, thereby promoting the general well-being and better overall health of sh [30]. In line with previous reports, the dietary consumption of the microalga Schizochytrium sp. dramatically improved the survival ability of zebra sh after the sh were challenged with E. piscicida.
Furthermore, the density of goblet cells was increased in the treatment groups, which is similar to the observations in golden pompano (Trachinotus ovatus) fed with a Schizochytrium sp. diet [17]. Goblet cells can synthesize and secrete mucin glycoproteins, which can cover the surface of the gastrointestinal tract epithelium, protecting the intestine from infection [31]. The elevated goblet cell density observed in our study suggests that dietary supplementation with Schizochytrium sp. improves intestinal homeostasis and assists the zebra sh in resisting the pathogen.
Treating sh with varying doses of Schizochytrium sp. led to different metabolism responses and microbial composition. Both diets with Schizochytrium sp. markedly increased phosphorylethanolamine concentrations, creatinine, myo-inositol, glucose, galactose, and decreased benzoic acid. Moreover, the 120 g/kg Schizochytrium sp. diet dramatically reduced the abundance of Bacteroidetes when compared to the control diet. These ndings indicated a conspicuous in uence of Schizochytrium sp. on the metabolic pro les and intestinal microbial community in zebra sh.
Diets with Schizochytrium sp. alter the colonic metabolite pro les of zebra sh PCA and OPLS-DA analyses revealed a clear separation of metabolites in the humoral uid of sh fed different diets, suggesting differences in the metabolic pro les. The carbohydrates, such as D-glucose, Dgalactose, maltose, and myo-inositol, were found to increase in zebra sh fed with 120 g/kg Schizochytrium sp. These carbohydrates are mainly involved in starch and sucrose metabolism, galactose metabolism, and the pentose phosphate pathway. The pentose phosphate pathway and galactose metabolism are related to energy metabolism. The starch and sucrose metabolism can in uence the tricarboxylic acid cycle (TCA cycle) through aerobic glycolysis [32]. A previous study also demonstrated that boosting the TCA cycle and energy supply enhanced the survival rate of zebra sh infected with Vibrio alginolyticus [33]. These elevated concentrations of carbohydrates indicate that diets with Schizochytrium sp. could increase the disease resistance of zebra sh by enhancing the TCA cycle and energy supply. Notably, citrate, the most important intermediate of the TCA cycle, was also increased in the zebra sh fed with 120 g/kg Schizochytrium sp. compared with zebra sh fed the control diet. Higher citrate levels also indicate an intensi cation of the TCA cycle and an enhanced energy supply.
The diet with 120 g/kg Schizochytrium sp. also increased the concentration of taurine. Taurine has many essential biological functions, including the stabilization of cell membranes and antioxidation [34]. Increased taurine has been reported to protect organisms against oxidant effects [35]. Leukocytes contain high concentrations of taurine, which can increase respiratory bursts and reduce tissue damage [36]. Therefore, the increase in taurine levels observed in this study suggests that a Schizochytrium sp. diet might enhance the innate immunity against bacterial infections.
Additionally, zebra sh fed with 120 g/kg Schizochytrium sp. signi cantly increased their concentrations of amino acid relatives (such as L-serine) compared with zebra sh fed without Schizochytrium sp. supplements. It has been demonstrated that L-serine could modulate the metabolome of tilapias to improve their innate immunity and eliminate pathogens, thereby improving the survival rate in tilapias infected with Streptococcus iniae [37]. Thus, increasing L-serine concentration in zebra sh via a diet with 120 g/kg Schizochytrium sp. might have a bene cial effect on pathogen resistance.
A previous study reported that an n-3 highly unsaturated fatty acid supplementation could reduce the cholesterol content in Ctenopharyngodon idella [38]. Similarly, in our study, zebra sh fed with the DHArich marine microalga Schizochytrium sp. supplemented diets contained less cholesterol than zebra sh fed a control diet. The cholesterol content may affect immune cell function, whereas reducing cholesterol in T cells may enhance CD8 + T cell immune activity [39]. In our study, the DHA-rich marine microalga Schizochytrium sp. might have improved the immune capacity of zebra sh by reducing their cholesterol content.
Diets containing Schizochytrium sp. alter the intestinal microbiota structure in zebra sh Schizochytrium sp. added to the diet resulted in differences in the bacterial community structure in zebra sh intestines. In concordance with the previous reports on zebra sh [40], the phyla Proteobacteria, Fusobacteria, Actinobacteria, Planctomycetes, and Firmicutes were dominant regardless of diet. Additionally, a signi cant decrease in Bacteroidetes was detected in the zebra sh fed with 120 g/kg Schizochytrium sp. Schizochytrium sp. is rich in omega-3 LC-PUFAs, especially DHA, and a recent study recorded a signi cant increase in the fatty acid content in zebra sh fed Schizochytrium sp. diets [8]. Therefore, the reduced abundance of Bacteroidetes in zebra sh fed with 120 g/kg Schizochytrium sp. supplemented diet implies that the abundance of Bacteroidetes in zebra sh is related to the dietary fat levels.
Some bene cial bacterial genera, including Lactobacillus, Dorea, Butyricicoccus, and Pseudoxanthomonas, were enriched in the zebra sh fed with 120 g/kg Schizochytrium sp. supplemented diet. Lactobacillus could be used as an immunostimulant and growth promotor. For instance, it enhanced the immune response and disease resistance in tilapia [41]. Furthermore, a higher abundance of Dorea was associated with greater resistance in human feces to pathogenic bacterial infections [42], while Butyricicoccus was shown to decrease lesion sizes and in ammation in a rat colitis model [43]. Finally, Pseudoxanthomonas could in uence the host immune response through a microbiome signature in tumors [44].
Interestingly, the relative abundances of some harmful bacterial genera (including Flavobacterium, Shewanella Pseudomonas, Citrobacter, and Mycoplasmas) were depleted after the zebra sh were fed with 120 g/kg Schizochytrium sp. The presence of Flavobacterium may explain the changes in the immunometabolism in sh and affects the nutrient metabolism, immune response, and related signaling pathways in sh. Additionally, several species are known to cause diseases in freshwater sh [45]. Shewanella, Pseudomonas, and Citrobacter are pathogens or opportunistic pathogens of sh [46][47][48]. Most Mycoplasma species are speci c commensals or parasites of eukaryotes. For instance, Mycoplasma mobile is a parasitic bacterium that binds to the gills of freshwater sh and causes necrosis [49]. This study demonstrates that dietary Schizochytrium sp. could optimize the intestinal microbial composition in zebra sh, enhance their immunity, and decrease the host's pathogen invasion.
To understand the composition and function of microbial communities, we undertook a Spearman's correlation analysis of the associations between metabolite features with microorganisms [50]. Our results indicated a close association between the circulating metabolites and the altered microbiome. It was observed that two bene cial bacterial genera in the Firmicutes phylum (namely Dorea and Butyricicoccus) correlated with some functional metabolites (i.e., taurine, D-glucose, and D-galactose). The genus Dorea is associated with glucose metabolism and plays a role in the activity of the immune system [51]. Several species of the genus Butyricicoccus are butyrate producers, and butyrate has several bene cial properties that are essential for maintaining intestinal homeostasis and resistance to speci c enteric pathogens [52]. A previous study has also shown that some species of Butyricicoccus are associated with mucosa, and their presence is reduced in patients with ulcerative colitis. As mentioned previously, taurine is involved in stabilizing cell membranes and immune capacity of the body. Therefore, we speculate that the observed improvement in the immune capacity of zebra sh was related to the changes in the abundance of Dorea and Butyricicoccus. Therefore, the potential probiotic species with two genera are worth investigating further.

Conclusions
The present study showed that Schizochytrium sp. supplementation enhanced the disease resistance in zebra sh. The mechanism involved in the improved immunology were explained via metabolomics and microbiome analysis. The altered metabolic pathways mainly involved the TCA cycle as well as the energy, taurine, L-serine, and cholesterol metabolisms. The abundance of bene cial bacteria, such as Lactobacillus, Dorea, Butyricicoccus, and Pseudoxanthomonas increased, whereas harmful pathogens (Flavobacterium, Pseudomonas, Citrobacter, and Mycoplasma) decreased in zebra sh fed with 120 g/kg Schizochytrium sp. Spearman's correlation analysis showed that some species of the genera Dorea and Butyricicoccus might be probiotics. Furthermore, the density of goblet cells in the intestine increased signi cantly in the zebra sh fed with Schizochytrium sp. when compared to the control group. Therefore, we conclude that Schizochytrium sp. supplementation has a positive effect on the immunity of zebra sh by improving their metabolic process and intestinal health.

Declarations Ethics approval
The experimental protocols were permitted by the Institutional Animal Care and Use Committee of Xiamen University, Xiamen, China.

Availability of data and materials
All data generated or analyzed during this study are available from the corresponding author upon reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was nancially supported by the National Natural Science Foundation of China (31871779, 21676221).
Authors' contributions YS conducted the animal work and most of the laboratory work and wrote the manuscript. XC and YZ helped to conduct the animal trial and part of the laboratory work and helped to revise the manuscript. YX and YW helped to revise the manuscript. NY oversaw the development of the study and reviewed the last version of the manuscript. All authors read and approved the nal manuscript. Kaplan-Meier survival curves for zebra sh fed with or without Schizochytrium sp. diets for 56 days and then challenged with E. piscicida. * indicates p < 0.05. C, S1, and S2 represent zebra sh fed with 0, 60, and 120 g/kg Schizochytrium sp., respectively. Black arrowheads indicate goblet cells. C, S1, and S2 represent zebra sh fed with 0, 60, and 120g/kg Schizochytrium sp., respectively. The different lowercase letters indicate a signi cant difference (p<0.05) between different groups.

Figure 3
Analysis of humoral fluid metabolites from the C, S1, and S2 based on GC-MS. All data were collected from eight sh per group. (A) PCA plots among the C, S1, and S2. OPLS-DA score plots obtained from (B) C vs. S1; (C) C vs. S2; (D) S1 vs. S2. (E) Heatmap and clustering of log concentrations of characteristic metabolites in dietary supplements of Schizochytrium sp. C, S1, and S2 represent zebra sh fed with 0, 60, and 120g/kg Schizochytrium sp., respectively.
Page 20/22  Comparison of intestinal microbiota compositions between groups. (A) Relative abundances of bacterial phyla among the C, S1, and S2 zebra sh, respectively. (B) Changes in the relative abundance of the phyla Bacteroidetes per group. (C) NMDS analysis of bacterial communities based on the unweighted UniFrac distances. (D) LEfSe analysis of taxonomic abundance in the intestine. C, S1, and S2 represent zebra sh fed with 0, 60, and 120g/kg Schizochytrium sp., respectively.