Response of Intestinal Microbiota to the Variation in Diets in Grass Carp (Ctenopharyngodon idella)

The intestinal microbiota is important for the nutrient metabolism of fish and is significantly influenced by the host’s diet. The effect of ryegrass and commercial diets on the intestinal microbiota of grass carp was compared in this study. In comparison to ryegrass, artificial feed significantly reduced the microbial diversity in the intestine, which was measured by a decrease in the observed OTUs, ACE, Shannon, and the InvSimpson index. Although grass carp fed with ryegrass and artificial feed shared a dominant phyla Firmicutes and Proteobacteria, the microbial composition was clearly distinguishable between the two groups. In grass carp fed with ryegrass, Alphaproteobacteria, Gammaproteobacteria, and Actinobacteria predominated, whereas Bacilli was significantly higher in the artificial feed group due to an increase in Weissella and an unassigned Bacillales bacteria, as well as a significant increase in a potential pathogen: Aeromonas australiensis. Grass carp fed with ryegrass exhibited a more complex ecological network performed by the intestinal bacterial community, which was dominated by cooperative interactions; this was also observed in grass carp fed with artificial feed. Despite this, the increase in A. australiensis increased the competitive interaction within this ecological network, which contributed to the vulnerable perturbation of the intestinal microbiota. The alteration of the microbial composition through diet can further affect microbial function. The intestinal microbial function in grass carp fed with ryegrass was rich in amino acids and exhibited an increased energy metabolism in order to compensate for a low-nutrient diet intake, while the artificial feed elevated the microbial lipid metabolism through the promotion of its synthesis in the primary and secondary bile acids, together with a notable enhancement of fatty acid biosynthesis. These results indicated that diet can affect the homeostasis of the intestinal microbiota by altering the microbial composition and the interspecific interactions, whilst microbial function can respond to a variation in diet.


Introduction
The intestine is the most important organ of digestion and absorption in fish, and it is also a complex ecosystem as it harbors an extremely diverse and complex microbial community [1][2][3]. It is widely acknowledged that the intestinal microbiota performs critical functions for the host, such as the production of digestion-related enzymes, vitamin synthesis, pathogen protection, and immune maturation [4][5][6]. The microbiota in the intestine provides a large number of fermented metabolites for the host, particularly herbivores [7][8][9]. The intestinal bacterial community is a complex micro-ecological system that is significantly influenced by several factors, including the nutritional components of the host's food [6,10]. Since fishmeal is scarce, plant protein is frequently utilized in aquafeed; however, this may lead to intestinal inflammation and microbial dysbiosis 2 of 14 in fish [11][12][13]. The number of species, their abundance, and their intricate microbial interactions play a critical role in the homeostasis of the intestinal microbiota [14]. The microbial interactions in the intestine are widely acknowledged to be dynamic in order to connect trillions of bacteria into a sophisticated ecological network [9,15,16]. Intestinal microbiota includes more than 100 times as many genes as the host, enabling it to encode a diverse range of enzymes with a variety of different metabolic capabilities [17]. Numerous studies have shown that the intestinal microbiota can act as an additional metabolic organ, contributing significantly to the host's amino acid, glucose, energy, and lipid metabolism by producing fermentation by-products [18][19][20][21].
The herbivorous grass carp (Ctenopharyngodon idella) is China's most productive freshwater fish. Numerous studies have characterized the grass carp's intestinal microbiota [22][23][24][25] and cellulase-producing bacteria, which assists in the digestion of fish that are fed a highfiber diet [7,8,26]. However, the existing knowledge about the bacterial community in the intestine is about the composition and function of microbial communities in grass carp; however, little is known about species-species interactions within the bacterial community. Our previous research looked at the microbiota in different parts of the intestine of grass carp, and we discovered that intricate interspecific interactions could boost the efficiency of the bacterial community fermentation [16]. It is widely acknowledged that the overuse of plant protein can disturb the interspecific interactions and result in a disorder of the intestinal bacterial community in fish.
Diets not only provide nutrients to fish and the intestinal microbiota, but they also influence fish health by modulating the intestinal microbiome. The objective of this study was to investigate the response of grass carp's intestinal microbiota to ryegrass and a commercial diet that is high in plant protein by evaluating the variation in composition, interspecific interactions, and metabolic functions of the intestinal bacterial community. Thus, the findings of this study may provide an in-depth understanding of the alteration of intestinal microbiota in grass carp in response to dietary variations, providing a theoretical basis for intervening in the intestinal microbiota to maintain grass carp health.

Sample Collection
The grass carp were raised in artificial ponds (area, 200 m 2 ; depth mean, 1.5 m) in the Institute of Special Aquaculture, Yichun, China. The grass carp, in two ponds, were fed with ryegrass (GF group: 12.73% protein, 1.38% lipid, and 26.52% fiber of total dry matter) and a commercial diet, provided by Tongwei Co., Ltd., Chendu, China (CF group: 30% protein (mainly plant protein from soybean meal, cottonseed meal, and rapeseed meal), 5% lipid, and 10.13% fiber). The fish were fed to apparent satiation twice a day (8:30 and 16:00) for one month without antibiotics. Grass carp fed with ryegrass (n = 8, weight mean, 316.45 ± 32.62 g) and a commercial diet (n = 8, weight mean, 357.81 ± 68.31 g) were collected after 3 h of feeding in the morning. In each group, twelve fish in one pond were chosen randomly from among more than 20 caught fish, and the other fish were returned. Before dissection, the skin surface of the grass carp was sterilized with 70% ethanol to reduce contamination, and then the digesta from the middle intestine was collected and stored in sterile freezing tubes under −80 • C. The experimental protocols of grass carp handling and sampling have been approved by the department of Laboratory Animal Science, Nanchang University (Approval Number: NCU-208-2021).

Illumina Sequencing of Bacterial 16S rRNA Gene
PowerFecal™ DNA Isolation Kit (MoBio Laboratories, Inc., Carlsbad, CA, USA) was used for DNA extraction of digesta samples. Amplification of the 16S rRNA V3-V4 region was performed as described previously with barcoded fusion primers of 341F and 805R [27]. High-throughput sequencing was performed using the Illumina HiSeq platform at Novogen Co., Ltd., Beijing, China. All the sequencing data can be found in the Sequence Read Archive (SRA) database at NCBI under accession number PRJNA880788.

Bioinformatics and Statistical Analysis
The raw sequences were sorted into different samples according to the barcodes by using the BIPES pipeline, followed by a quality-control step to remove any low-quality amplicon sequences by VSEARCH [28]. The clean sequences were then clustered into operational taxonomic units (OTUs) with a 99% sequence similarity and annotated using the Ribosomal Database (rdp_16s_v16_sp). A total of 1,369,277 effective sequences and 1201 OTUs were generated from all samples. Alpha diversity and the relative abundance of taxa analyses were calculated by R software v 4.1.3. The Wilcoxon test was used to test the α-diversity index, and the relative abundance of taxa using R software. Treemap was used to visualize the significantly abundant OTU's, the annotated taxonomy, the P-value, and the relative abundance, in which the size of the bubbles indicated the relative abundance of the raw read counts [29]. Principal coordinates analysis (PCoA), based on the Bray-Curtis dissimilarity analyses, were employed to visualize the bacterial community structure and the differences in the bacterial community was calculated by Permutational analysis of variance (PERMANOVA) based on the Bray-Curtis distance [30].
Using abundance profiles of the individual OTUs, a molecular ecological network analysis was performed to evaluate bacterial species-to-species interactions within a community (http://ieg2.ou.edu/MENA (accessed on 15 July 2022)). A Random Matrix Theory (RMT)-based approach was used for an ecological network construction and topological role identification [31]. The network was visualized using Circos and Cytoscape 3.9.0. Based on a modularity property, each network was separated into modules by the fast greedy modularity optimization. According to values of within-module connectivity (Zi) and among module connectivity (Pi), the topological roles of different nodes can be categorized into four types: peripherals (Zi ≤ 2.5, Pi ≤ 0.62), connectors (Zi ≤ 2.5, Pi > 0.62), module hubs (Zi > 2.5, Pi ≤ 0.62) and network hubs (Zi > 2.5, Pi > 0.62).
Functional gene and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were predicted using PICRUSt2 software [32] against a Greengenes reference database (Greengenes 13.5). The non-metric multidimensional scaling (NMDS) and analysis of similarity (ANOSIM) were used to evaluate the overall differences in predicted bacterial functional compositions based on the Bray-Curtis distance at KEGG orthology (KO) level [33]. A two-sided Welch's t test was used to identify significant different metabolic pathways in the two groups by software STAMP, with p < 0.05 considered significant.

Ecological Network Analysis
A circos plot displayed a classified composition and species-species interactions within the bacterial community ( Figure 5A), which consisted of different OTUs from 27 bacterial classes ( Table 2). The GF network represented 441 OTUs and 2946 edges (gray edges: 2099; red edges: 847), and the CF network displayed 252 OTUs and 856 edges (gray edges: 592; red edges: 264). The gray and red edges indicated the positive and negative interactions between two OTUs. The GF network recorded major OTUs (≥20) from Actinobacteria (69 OTUs), Alphaproteobacteria (66 OTUs), Clostridia (41 OTUs), Gammaproteobacteria (40 OTUs), Acidobacteria (29 OTUs), Deltaproteobacteria (25 OTUs), and Bacilli (23 OTUs), whereas the major OTUs from Alphaproteobacteria (47 OTUs), Bacilli from Gammaproteobacteria, and unassigned bacteria from Bacillales (Bacilli) domin the bacterial community of grass carp from the CF group (p < 0.05). Additionally PCoA analysis exhibited a clear separation in the bacterial communities between th and the CF groups at the OTU level, and a significant difference was further reve using PERMANOVA (p = 0.003, Figure 4).

Ecological Network Analysis
A circos plot displayed a classified composition and species-species interact within the bacterial community ( Figure 5A), which consisted of different OTUs from bacterial classes ( Table 2)

Functional Predictions of Intestinal Microbiota with PICRUSt2
The NMDS analysis exhibited a clear distinction in the functional composition of the intestinal microbiota between the GF and the CF groups at a KO level, and an analysis of similarity (ANOSIM) further confirmed the remarkable differences in the bacterial functional composition between the GF and the CF groups (p = 0.007, Figure 7). To study the microbial metabolic function, the KEGG functional categories were analyzed, including amino acid metabolism, carbohydrate metabolism, energy metabolism, lipid metabolism, and protein families metabolism (Figure 8). A significant difference in 47 pathways and 6 protein families, which are involved in nutrient metabolism, were observed between the GF and the CF groups (p < 0.05). In particular, the bacterial communities in the grass carp fed with ryegrass was significantly rich in 13 pathways in amino acid metabolism, 6 pathways in carbohydrate metabolism, 3 pathways in energy metabolism, 5 pathways in lipid metabolism, and 2 protein families related to metabolism (p < 0.05). The dietary commercial diet notably promoted bacterial metabolic function including 7 pathways in amino acid metabolism, 6 pathways in carbohydrate metabolism, 2 pathways in energy metabolism, 6 pathways in lipid metabolism, and 4 protein families related to metabolism in grass carp (p < 0.05).

Functional Predictions of Intestinal Microbiota with PICRUSt2
The NMDS analysis exhibited a clear distinction in the functional composition of the intestinal microbiota between the GF and the CF groups at a KO level, and an analysis of similarity (ANOSIM) further confirmed the remarkable differences in the bacterial functional composition between the GF and the CF groups (p = 0.007, Figure 7). To study the microbial metabolic function, the KEGG functional categories were analyzed, including amino acid metabolism, carbohydrate metabolism, energy metabolism, lipid metabolism, and protein families metabolism (Figure 8). A significant difference in 47 pathways and 6 protein families, which are involved in nutrient metabolism, were observed between the GF and the CF groups (p < 0.05). In particular, the bacterial communities in the grass carp fed with ryegrass was significantly rich in 13 pathways in amino acid metabolism, 6 pathways in carbohydrate metabolism, 3 pathways in energy metabolism, 5 pathways in lipid metabolism, and 2 protein families related to metabolism (p < 0.05). The dietary commercial diet notably promoted bacterial metabolic function including 7 pathways in amino acid metabolism, 6 pathways in carbohydrate metabolism, 2 pathways in energy metabolism, 6 pathways in lipid metabolism, and 4 protein families related to metabolism in grass carp (p < 0.05).

Discussion
The importance of intestinal microbiota in host health has been highlighted in recent decades, owing to its benefits to nutrient metabolism and immune maturation [4,34,35]. It

Discussion
The importance of intestinal microbiota in host health has been highlighted in recent decades, owing to its benefits to nutrient metabolism and immune maturation [4,34,35]. It is widely accepted that diets can influence the composition of the intestinal bacterial community because microbes require nutrients and energy from the food consumed by hosts [36]. Grass carp feed on aquatic weeds in the wild, and previous research found that grass carp fed a Sudan grass diet were rich in Firmicutes, Proteobacteria, Fusobacteria, and Actinobacteria in the intestine [37]. In the current study, Proteobacteria, Firmicutes, and Actinobacteria were found to be the most prevalent phyla in the intestines of grass carp fed with ryegrass. Grass carp, the most important aquaculture species in China, are primarily fed an artificial diet rich in plant protein. Previous research has shown that commercially formulated feed can significantly alter the structure of the intestinal bacterial community in grass carp, resulting in a decrease in microbial biodiversity [37]. Indeed, the current findings show that the artificial feed reduced diversity, and the proportion of Alphaproteobacteria and Actinobacteria, while increasing Bacilli. Furthermore, in this study, grass carp fed a high plant protein diet had a significantly higher proportion of Aeromonas australiensis from Gammaproteobacteria. Aeromonas are Gram-negative microbes that live in aquatic environments and have been identified as an opportunistic pathogen for fish [38,39]. Previous research has suggested that Aeromonas sp. infection can cause intestinal inflammation and mucosal barrier function damage in grass carp [40,41], crucian carp [42], and Nile Tilapia (Oreochromis niloticus) [43]. As a result, the significant increase in A. australiensis, caused by the artificial feed, increased the potential threat to the health of grass carp in this study.
As species perform similar or complementary functions, competitive and cooperative interactions occur between different species in the same habitat, for example competition to occupy more space and to obtain more nutrients [44]. The intestinal bacterial community can form a complex ecological network based on these interspecific interactions, allowing the intestinal microbiota to maintain a dynamic homeostasis in the host [14]. The microbial interactions were clearly degraded in the grass carp fed with an artificial diet due to a decrease in the microbial diversity, which can have a negative effect on the stability of a bacterial community because the intestinal microbiota is more easily disturbed by external factors. A cooperation dominated community is thought to be more stable because cooperative interactions are more resistant to population perturbations in spatial conditions, whereas competitive interactions are more vulnerable to disruption [45,46]. Indeed, the current findings suggested that cooperation dominated the intestinal bacterial community of grass carp, whereas the increased percentage of competitive interactions in the group fed with the artificial diet can be attributed to the significant increase in the opportunistic pathogen A. australiensis. Despite the fact that the grass carp, in two groups, showed distinctive submodules in the ecological networks, the networks in this study were primarily made up of the dominating microbial flora of each group. From an ecological perspective, connections, and module hubs generalists act as structural and functional keystones and play a crucial role in sustaining the ecological network [47]. According to the latest findings, the network's generalist population was unaffected by the artificial diet. By weakening interspecific relationships and reducing cooperation in the ecological network, the artificial diet may collectively have a detrimental influence on the gut microbiota homeostasis.
Since the genes of microbes encode several enzymes involved in protein, carbohydrate, lipid, and energy metabolism, microbial fermentation plays a critical role in the host's nutritional metabolism [35,48,49]. The structure and operation of the gut bacterial community can adjust to changes in the dietary environment in a similar manner [36]. Due to differences in gene expression across the different microbes in the current study, dietary ryegrass, and artificial feed, as expected, shaped two distinctly different intestinal microbiota structures in grass carp. These differences were accompanied by a notable variation in the metabolic activities of the microbial community. Fish were frequently affected with the issue of aberrant fat deposition during breeding due to the use of a high-protein and a high-fat artificial feed. [50,51]. Evidence in both humans and animals has shown a robust correlation between obesity and a high Firmicutes ratio [52,53]. Here, compared to a ryegrass diet, the artificial feed did increase the proportion of Firmicutes in grass carp. This was followed by an increase in lipid metabolism, specifically an improvement in the biosynthesis of fatty acids, glycerols, and glycerospholipids, as well as a promotion of the production of primary and secondary bile acids. However, when grass carp were fed with ryegrass to supplement their low-nutrient diet, the microbial function related to amino acid and energy metabolism was more active. Furthermore, it has been demonstrated that microbial fermentation of complex, non-digestible dietary carbohydrates aids the host in obtaining useable forms of energy from a plant-based diet [8]. Grass carp are generally considered herbivorous, but no gene in the grass carp genome encodes an enzyme to digest cellulose, which comprises the principal component of a plant-based diet [54]. Evidence has confirmed that the microbial function of metabolizing cellulose plays an important role in the nutrient metabolism of an herbivore, and the abundant enzymes involved in the digestion of complex carbohydrates were also observed in the intestinal microbiota of grass carp [8]. Here, grass carp in the ryegrass diet group displayed a higher proportion of Actinobacteria, which proved to harbor an increased level of carbohydrate enzyme genes for the degradation of cellulose [55]. Nevertheless, it was shown that grass carp fed with an artificial feed, where starch was the predominant source of carbohydrate, had an increased microbial starch and sucrose metabolism. Additionally, a group of microorganisms that are interdependent on the effectiveness of fermentation are necessary for the microbial process of fermentation [9,56]. Therefore, the increased cooperative interaction may encourage the microbial fermentation of a ryegrass diet, assisting the grass carp to obtain more nutrients and energy. However, the microbial function in grass carp fed with the artificial diet was more susceptible to diseases in this study because of the per durability in the bacterial population.
The modification of intestinal microbiota is an effective method for keeping fish healthy because it plays a significant role in the metabolism of nutrients and the host's health. In fact, several studies have shown that by managing the homeostasis of fish's intestinal microbiota with functional feed additives improves fish health [57,58]. The current study evaluated the intestinal bacterial communities of grass carp fed with an artificial diet and ryegrass, and it would serve as a theoretical guide for the control of grass carp intestinal microbiota.

Conclusions
Intestinal bacterial community characteristics in grass carp were altered by dietary ryegrass and artificial feed. An opportunistic pathogen, A. australiensis, was found to drastically increase in the artificial feed diet. Additionally, artificial feed may impact the lipid metabolism of grass carp by raising the ratio of Firmicutes and decreasing the disturbance resistance of intestinal microbiota.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/metabo12111115/s1, Table S1: Relative abundance of different bacterial phyla in grass carp, Table S2: Relative abundance of different bacterial classes in grass carp.