For this research, the cultured Pacific White Shrimp (L. vannamei) were afforded by Evergreen South Ocean Tech Co., Ltd, Zhanjiang, China. 16 aquaria (300 L, 0.6 m² bottom, 4 aquaria per dietary treatment) with 40 shrimp/aquaria were used for four experimental diets. The shrimp weight 0.85 g were selected and adapted for a week with basic feed prior to the experiments. For the evaluation of the impacts of RSP (All the main raw materials (include: Rhodobacter sphaeroides and glucose, corn steep liquor, inorganic salts, etc.) were fermented, then inactivated and dried to gain the Rhodobacter sphaeroides protein product; Zhejiang NHU Company Ltd., Xinchang, Zhejiang) and CoQ10 (Zhejiang NHU Company Ltd., Xinchang, Zhejiang) on L. vannamei, four experimental diets were designed as follows: FM diet contains 20% fishmeal; RSP diet contains 20% RSP instead of 20% fishmeal; CoQ10 diet contains 20% fishmeal supplemented with CoQ10 at 0.08 g kg^–1; RSP + CoQ10 diet contains 20% RSP supplemented with CoQ10 at 0.08 g kg^–1 (Table 1). The composition of basal feed was as follows: crude protein 38%, crude fat 6% and moisture 9%. The method of diet preparation was the same as described by Niu et al. (2010). In short, all dry ingredients of four diets were weighed and mixed fully to homogeneity in a Hobart-type mixer (A-200T Mixer Bench Model unit, Resell Food Equipment Ltd., Ottawa, Canada). After that soy oil, fish oil and soybean lecithin were then added and mixed well for 5 min, and the mixed components were added and mixed to deionized water for 10 min. Post conditioning, the wet mixture were randomly extruded using monoscrew extruder (Institute of Chemical Engineering, South China University of Technology, Guangzhou, P.R. China) having barrel inner diameter of 1.2 mm. The four diets were stored at −20°C before use. All shrimp were fed to apparent satiation three cycles per day at 8:00, 12:00, and 18:00 by hand with a total amount of approximately 6% of its body weight, and maintained in fresh seawater with the temperature ranged from 27 to 29°C, pH was 8.0–8.2, salinity was approximately 30 ppt. The aquaria conditions were kept identical during feeding experiment.

The feeding experiment was carried out for 8 weeks. At the start and the end of the feeding experiment, all shrimp were weighed. The routine nutritional compositions of the body samples were then analyzed (n = 5), meanwhile, intestine samples were obtained for the analysis of intestinal microbiota (n = 5). All samples were frozen in liquid nitrogen immediately and then stored in −80°C until measurement.

During the feeding experiment (8 weeks), the number of dead individuals were counted from beginning to end for measurement of the survival rates. In this study, weigh gain (WG), feed conversion ratio (FCR), specific growth rate (SGR), condition factor (CF), and hepatosomatic index (HSI) were used to evaluate the growth performance in L. vannamei. The parameters were calculated as follows: WG = 100 × (final body weight - initial body weight)/initial body weight; SGR = 100 × (Ln final mean weight—Ln initial mean weight)/number of days; FCR = dry diet fed/wet weight gain; CF = 100 × (body weight)/(body length)³; HSI = 100 × hepatopancreas weight/whole body weight.

Genome DNA of the intestinal contents was extracted with a DNA microbiome kit (Qiagen, Germany). The integrality of total DNA was assessed with 1% agarose gel electrophoresis, DNA purity and concentration were quantified using a spectrophotometer (Thermo Fisher Scientific Inc., United States). Purified genomic DNA was then sent for sequencing at the Novogene Biological Information Technology Co. (Tianjin, China). Quality filtering and splicing on raw data were performed using QIIME procedures (Caporaso et al., 2010). The retained clean data was performed Operational Taxonomic Unit (OTU) cluster analysis. Clustering of OTUs (at 97% similarity) were performed by VSEARCH for further annotation. The further diversity difference analysis was conducted according to the results of the relative OTU abundance. Alpha diversity measurements were evaluated with ACE, Chao1 and Rank-Abundance curves (richness), Shannon and Simpson (diversity) (Schloss et al., 2011). Beta diversity measurements were manifested as principal component analysis (PCA) and principal coordinates analysis (PCoA). Finally, the microbial communities’ differences between four diet treatments were evaluated using Linear Discriminant Analysis Effect Size (LEfSe) method.

At the end of the 8-week feeding trial, 10 shrimp from each experimental aquarium were transferred randomly into the desalinated seawater (salinity, 2 ppt). The mortality was monitored for 4.5 h post- attack, and the relative percentage survival rates were calculated. In addition, hepatopancreas samples were collected for further analysis.

The frozen hepatopancreas were weighed and homogenized by adding ice-cold PBS (1:10 dilution) until the hepatopancreas were completely broken. The homogenate tissues were centrifuged for 20 min (3,000 × g at 4°C). T-AOC, GSH-PX and MDA were detected by commercial assay kits (cat. nos. A015, A005, and A003, respectively; Nanjing Jiancheng Bioengineering Institute, China) following the manufacturers’ instructions. All assays were completed within a week after preparation.

For the purposes of investigating the effects of RSP and CoQ10 on the immunity and antioxidative capacity of L. vannamei, the hepatopancreas samples from 4 shrimp per each dietary were collected after low salinity challenge. The total RNA samples were extracted from hepatopancreas by RNeasy™ animal RNA extraction kit (Beyotime, Shanghai, China) according to the protocol, and then transcribed to complementary DNA (cDNA) with a PrimeScript RT Reagent kit (Takara Bio, Inc.). The gene expression levels of four immune- related and oxidative stress damage related genes HSP70, Caspase-3, SOD and Relish were investigated in the hepatopancreas by the Light Cycler 480 (Roche Applied Science, Basel Switzerland). The qRT-PCR reaction program was performed as the following thermocycler condition: 95°C for 10 min, followed by 40 cycles of 95°C for 5 s, 60°C for 30 s and 72°C for 30 s, finally at 4°C for 5 min. Specific shrimp primers of related genes used in this study were shown in Table 2, the expression levels of related genes were quantified using the 2^–ΔΔCt method.

All results were presented as the mean ± standard error (SE). One-way ANOVA was performed to determine statistically significant differences of the different treatment using SPSS software. Multiple comparisons were made using Duncan’s multiple tests. Other statistical analyses were performed with the R or VSEARCH software environment. A P-value < 0.05 was considered significant.

The growth performance and survival rates of L. vannamei were shown in Figure 1. In the whole feeding period, shrimp fed with RSP diet had remarkably higher WG and FCR than shrimp fed with FM diet (P < 0.05; Figures 1A,C). However, survival rates were similar among all diets (P > 0.05; Figure 1B). A higher SGR was found in shrimp fed with CoQ10 diet than that in shrimp fed those diets and no significant differences were found among other diets (P > 0.05; Figure 1D). CF of shrimp fed with FM diet was remarkably lower than those of shrimp fed with RSP diet, CoQ10 diet and RSP + CoQ10 diets, while no significant differences were found among RSP diet, CoQ10 diet and RSP + CoQ10 diets (P > 0.05; Figure 1E). HSI of shrimp fed with FM diet was significantly higher than those of shrimp fed RSP diet and CoQ10 diets (P < 0.05), while no significant differences were found in HSI between FM diet and RSP + CoQ10 diets (P > 0.05; Figure 1F).

SA The crude lipid content in muscle was significantly lower in shrimp fed FM diet than from the other three diets (P < 0.05; Table 3). Meanwhile, no considerable differences were found in crude protein and moisture contents of muscle in all diets (P > 0.05).

The alpha diversity parameters include ACE, Chao1, Shannon and Simpson index, which represent the extents of microbial communities’ richness and diversities. Clearly, the microbial community richness from RSP diet was significantly higher than that from other diets, as supported by ACE (P < 0.05) and Chao1 (P < 0.05; Figures 2A,B). The Shannon index was highest in RSP diet, following in RSP + CoQ10 diet, and lowest in FM diet (Figure 2C). What’s more, Simpson’s index was lower in RSP diet than that in FM diet and CoQ10 diets (Figure 2D), microbial diversity in FM diet was nearly identical with CoQ10 diet (Shannon index 4.5, Simpson index 0.90). All the results manifested that 100% of fishmeal replaced by RSP can increase intestinal microbial diversity.

OTU analysis-based dilution and grade abundance curves from different treatments were given in Figures 2E,F. The horizontal axis of FM diet and RSP diets were broader than that of CoQ10 diet and RSP + CoQ10 diets, manifesting that the species composition from FM diet and RSP diets were more specific and richer compared with CoQ10 diet and RSP + CoQ10 diets.

Beta diversity analysis was to show the similarities and differences of microbial communities from different treatments. According to PCA analysis, a similar microbial community structure was observed among the basic diet, CoQ10 diet and RSP + CoQ10 diet (Figure 2G), which showed that RSP substitution of fishmeal had a bigger effect on microbial community structure. A principal coordinate analysis (PCoA) showed that the four diet treatments were divided into three parts that manifested the effects of RSP and CoQ10 on the similarity of intestinal microbial composition in L. vannamei (Figure 2H), indicating that RSP substitution had impacted L. vannamei microbiota significantly (P < 0.05) but not CoQ10 addition.

An OTU was defined by sequences were clustered at the 97% similarity, each OTU represented a corresponding sequence. Observed numbers of OTUs have shown that the highest OTU number of 530 was with RSP + CoQ10 diet and the lowest of 218 with CoQ10 diet with the remaining OUT numbers were in between (222–460) (Figure 3A). A Venn map was constructed to identify dominant OTUs presented in all diet treatments, indicating that there were 50, 141, 56, and 221 unique OTUs in FM diet, RSP diet, CoQ10 diet and RSP + CoQ10 diets, respectively, which means the dietary fishmeal replaced with RSP and the dietary CoQ10 addition can influence intestinal microbial and generate corresponding unique microbiota (Figure 3B).

Compared with the OTU sequences with related database, Proteobacteria still the only predominant phylum in all diet treatments with the abundance of 75.9, 68.4, 73.2, and 66.1% in FM diet, RSP diet, CoQ10 diet and RSP + CoQ10 diets, correspondingly, the second dominant phylum was Bacteroidota with an abundance of 12.6, 16.1, 21.1, and 24.6% in FM diet, RSP diet, CoQ10 diet and RSP + CoQ10 diets (Figure 3C). Clearly, more diverse phyla were found in RSP + CoQ10 diet, including Thermotogae, Caldatribacteriota, Gemmatimonadota, and Parcubacteria. The above results revealed that the amount of Proteobacteria was changed not due to RSP replacement for fishmeal, but as a result of the addition of CoQ10. On the other hand, both dietary fishmeal replaced with RSP and dietary CoQ10 addition have promoted the proliferation of Bacteroidota, manifesting that RSP substitution and CoQ10 supplementation could improve the intestinal microecological environment.

At bacterial class level, Alphaproteobacteria, Gammaproteobacteria, and Bacteroidia were confirmed to be abundant in all diet treatments. The relative abundance of Alphaproteobacteria revealed that a highest relative abundance of 49.9% in CoQ10 diet and a lowest of 32.6% in RSP + CoQ10 diet. Whereas the relative abundance of Gammaproteobacteria displayed the opposite trend that shrimp fed with RSP + CoQ10 diet has the highest abundance of 35.4% and fed with CoQ10 diet has the lowest abundance of 23.2%. The third dominant class was Bacteroidia with an abundance of 12.6, 16.1, 21.1, and 23.8% in FM diet, RSP diet, CoQ10 diet and RSP + CoQ10 diets (Figure 3D). At the family level (Figure 3E), Rhodobacteraceae was the dominant family in all diets with its relative abundance at 45.9, 35.2, 49.6, and 32.1% in FM diet, RSP diet, CoQ10 diet and RSP + CoQ10 diets, separately, indicating its predominance was weakened in RSP diet and RSP + CoQ10 diets independent of the presence of absence of CoQ10. At the genus level (Figure 3F), the number and abundance of predominant genus in different diets varied considerably. Ruegeria was the predominant genus in all diets, the relative abundance of Ruegeria manifested that the highest relative abundance of 31.2% was with FM diet and a lowest of 15.1% with RSP + CoQ10 diet with the rest in between (19.1–22.4%). At the end of the test, compared with FM diet, the relative abundance of Vibrio was decreased significantly in RSP diet, CoQ10 diet and RSP + CoQ10 diets, however, in RSP diet, CoQ10 diet and RSP + CoQ10 diets, almost an addition of Pseudoruegeria was observed. Altogether, the microbial compositions of predominant phyla were obviously similar among these four diet treatments. All these data indicated that RSP substitution and CoQ10 supplementation may maintain intestinal homeostasis and decrease the colonization by intestinal opportunistic pathogens, like Vibrio.

LefSe analysis was carried out to further investigate the changes of L. vannamei’ intestinal microbial biomarkers caused by RSP substitution and CoQ10 supplementation (Figure 4). Based on the threshold that a Linear Discriminant Analysis (LDA) score of > 4.0, the dominant phylum, order, class, family and genus were screened. The biomarkers at different levels associated with FM diet were, in descending, Cellvibrionales, Halieaceae, and Bacteroidia. In RSP diet, the specific biomarkers were, Rhodobacteraceae, Rhodobacterales, and Alphaproteobacteria. Moreover, in CoQ10 diet, the specific biomarkers were, Photobacterium_damselae, Photobacterium, and Actinobacteriota, by contrast, the specific biomarkers were, Bacteria, Protebacteria, and Oceanospirillales in RSP + CoQ10 diet.

Based on PICRUSt, the functional capacity of the intestinal microbiota was predicted. The detailed results were given in Figure 5, evidently, the most abundant function was related to environmental information processing in FM diet. In RSP + CoQ10 diet, genetic information processing functions were doubled when compared to FM diet. Correspondingly, human diseases were enriched in CoQ10 diet. In combination, these COG function classification results showed that the L. vannamei intestinal microbial taxa presented the distinct biological functions with the presence of absence of RSP substitution and CoQ10 supplementation.

The survival rates of four experimental diet treatments were shown in Table 4 after acute low salinity challenge. The survival rate in shrimp fed with RSP diet was considerably lower than those in shrimp fed with CoQ10 diet and RSP + CoQ10 diets (P < 0.05). No statistically appreciable differences were observed in shrimp fed with FM diet and RSP diets (P > 0.05).

The hepatopancreas content of MDA and activities of T-AOC and GSH-PX were illustrated in Table 5. The hepatopancreas activities of T-AOC in shrimp fed CoQ10 diet and RSP + CoQ10 diets were remarkably higher than that in shrimp fed FM diet and RSP diets (P < 0.05), while no significant differences were presented in T-AOC activity between FM diet and RSP diets (P > 0.05). In addition, no considerable differences were found in hepatopancreas MDA contents and GSH-PX activities among all diets (P > 0.05).

As showed in Figure 6, after 4.5 h acute low salinity challenge, the expression level of SOD in hepatopancreas of shrimp fed CoQ10 diet and RSP + CoQ10 diets were remarkably higher than those of shrimp fed FM diet and RSP diets (P < 0.05). Relative expression of HSP70 for shrimp fed CoQ10 diet and RSP + CoQ10 diets were higher than that for shrimp fed FM diet and RSP diets (P < 0.05). However, no significant differences were found in Caspase-3 expression among all diets (P > 0.05). The lowest Relish expression was found in shrimp fed FM diet (P < 0.05).