The experimental subjects were a full-sib family of O. bidens preserved in the cooperative base of the Freshwater Fisheries Research Institute of Fujian. The research group used wild adult horse mouth fish parents for artificial oxytocin and one-to-one pair breeding. Wild mature parents of O. bidens were artificially induced to ovulate using oxytocin (DOM 3 mg, LRH-A2 5 μg, HCG 800 IU), and paired insemination was performed according to a previous study (Chen et al., 2023). Fertilized eggs from the same parent were cultured under the same environmental conditions. After the larvae hatched, the entire process was isolated in a factory breeding tank. The culture conditions were as follows: the temperature of the aquaculture water body was 24 ± 0.5°C; feed was given twice per day, and about 50% of the water was replaced twice per week.

After continuous cultivation for three months, a total of 100 healthy fish (i.e., complete coverlet scales, a smooth surface, apparent vitality, and no parasites) were randomly selected from the full-sib family. The fish were ranked by body weight, and those in the top 30 were classified as the overweight (OW) group, while those in the bottom 30 were classified as the underweight (UW) group for detection and analysis.

The biological indexes in the OW and UW groups were determined. According to the method of Ichthyology Experimental Guidance (Meng et al., 1995), the whole body length, body length, head length, trunk length, body height, tail length, tail stalk height, and tail fin length of test fish were measured with Vernier calipers (accurate to 0.01 mm), and weight was measured with an electronic scale (accurate to 0.01 g).

After anesthesia with MS222, the intestinal tissues of individuals from the OW and UW groups were dissected; the intestinal attachments were removed, and the midgut was excised and stored in 4% paraformaldehyde for paraffin sectioning. The test samples were transferred to 75% anhydrous ethanol for gradient dehydration, and xylene was applied twice for transparency, followed by three rounds of wax penetration. This was followed by embedding, wax block dressing, slicing (thickness = 5 μm), unfolding, and baking. The sections were stained with hematoxylin-eosin (HE) and sealed with neutral gum. An optical microscope (Olympus BX53) was used for microscopic observation, and the image acquisition software CCD (Olympus DP80) was used for photography. The villus height, villus width, and muscle thickness were measured using the NDP.view 2 software micrometer.

The midgut tissue of the test fish was excised, and the gut contents were cleaned with normal saline and pre-cooled to 4°C. The fat, mesentery, and other intestinal impurities were removed, and excess water was absorbed with filter paper. The intestinal tissue was accurately weighed, and a corresponding volume of normal saline was added at a ratio of 1:9. Mechanical homogenization was carried out in a low-temperature water bath. The homogenate was centrifuged at 2500 rpm for 10 min, and the supernatant was taken and stored at 4°C. The activities of digestive enzymes, including total protein (TP), lipase (LPS), trypsin (TRY), and amylase (AMS), were measured using a kit produced by the Nanjing Jiancheng Bioengineering Institute.

The v3–v4 region of a fragment of the 16S rRNA gene in the gut tract of O. bidens was amplified using the primers F: ACTCCTACGGGAGGCAGCAG and R: GGACTACHVGGGTATCTAAT. The total volume of the amplification system was 25 μL, with 12.5 μL of 2× Mix (TaKaRa), 9.5 μL of dd H₂O, 0.5 μL of 5′ and 3′ primers, and 2 μL of cDNA template. The PCR amplification procedure comprised predenaturation at 95°C for 30 s, denaturation at 95°C for 10 s, annealing at 58.6°C for 30 s, extension at 72°C for 2 min for 35 cycles, and a final elongation at 72°C for 10 min. The amplified products were extracted using a gel extraction kit, then sequenced and analyzed on an Illumina Nextseq 2000 (Illumina, USA) sequencing platform according to the specifications of Guangzhou Gidio Biotechnology Co., Ltd. The sequencing read length was 300 bp; the criterion was 97% ASVs similarity, and the rarefaction depth was 30, 171.

Total RNA was extracted from fish muscle tissue using a TRIzol kit (AG, USA). The values of OD260 nm and OD280 nm were determined using a Cary50 ultraviolet spectrophotometer (Varian, USA), and the extraction purity and concentration of RNA were preliminarily determined according to standard curves. The quality of RNA was subsequently detected by 1% agarose gel electrophoresis. cDNA was synthesized by reverse transcription of the extracted RNA using a PrimeScript RT reagent kit (TaKaRa, Japan) and diluted 10 times with ddH2O for subsequent PCR validation and qRT-PCR tests.

The total RNA of the sample was extracted and sent to Beijing NOhe Technology Co., Ltd., for cDNA synthesis, cDNA end repair, splicing, and PCR library enrichment and purification. Sequencing, data splicing assembly, and analysis were performed after quality control. The genes were mapped to each term of the GO database (http://www.geneontology.org/), and the number of genes in each term was counted. GO entries that were significantly enriched in differentially expressed genes compared with the whole genome background were identified using hypergeometric tests. All differentially expressed genes were mapped to pathways in the KEGG database (http://www.kegg.jp/), and the number of genes in each pathway was counted.

A set of expressed genes was randomly selected from the transcriptome database, and forward and reverse primers for real-time quantitative fluorescent PCR (qRT-PCR) were designed according to the gene sequences ( Supplementary Table 1 ). The qualified primers were judged by detecting whether the amplification efficiency was within the range of 95–105%. β-actin was used as an internal reference gene to verify the expression levels of genes between different groups. qRT-PCR amplification was carried out using 10× gradient diluted cDNA as a template to construct standard curves. The total volume of the reaction system was 20 μL and contained 10 μL of 2× SYBR green master mix (TaKaRa), 6.8 μL of ddH₂O, 0.4 μL of Rox Dye, 0.4 μL each of the 5′ and 3′ primers, and 2 μL of cDNA template. The amplification procedure comprised predenaturation at 95°C for 30 s, denaturation at 95°C for 5 s, and annealing and extension at 58.6°C for 30 s, for a total of 40 cycles. The qRT-PCR amplification was performed on a Step One Plus fluorescence quantitative PCR instrument.

The stats package for R (version 3.3.1) and the scipy package for Python (v1.0.0) were used in the statistical analysis. The Tukey–Kramer post-hoc test was employed for multiple comparisons. The criteria |log2 FC| ≥ 1 and P adjust < 0.05 were used to screen for differentially expressed genes. The “ggcor” package for R Studio software version 2022.07.0-548 (2009–2022 RStudio, PBC) was used for Mantel tests to analyze the correlations between morphological data and relevant indicators. A structural equation model (SEM) was constructed using the software IBM SPSS Amos (version 24). Each variable underwent Z-score standardization prior to modelling. The goodness of fit of the model was determined by the χ² test (χ2/df ratio around 0 to –2 with P > 0.05), the root mean square error of approximation (RMSEA, 0.05 < RMSEA ≤ 0.08 with P > 0.05 defined as an adequate fit), and the root mean square residual (RMR, RMSEA < 0.01 with P > 0.05 as a good fit). SPSS23.0 (IBM, USA) and Excel were used for statistical analysis, and one-way ANOVA was used to compare the differences among groups. P < 0.05 indicated a significant difference, and P < 0.01 indicated a particularly significant difference. GraphPad Prism 8.0 software was used to map and present the results.

The growth indexes of body weight, total length, and other measures listed above were significantly higher in the OW group than in the UW group (P < 0.01). The mean body weight of fish in the OW group was more than two times higher than in the UW group; the total length of the OW group was more than 30% higher than that of the UW group. The average growth rate of the OW group was 158.55 mg/d, which was 122.28% greater than that of the UW group ( Figure 1 ).

The intestinal tissue morphology and structure of fish in the OW and UW groups were significantly different, and the total intestinal villi size of fish in the OW group was significantly larger than that in the UW group. Further microscopic measurements showed that the intestinal villus height and intestinal muscle thickness in the OW group were significantly greater than those in the UW group (P < 0.05); the villus height was about 1.3 times that of the UW group, and the muscle thickness was about 1.4 times that of the UW group. In addition, the width of the intestinal villi in the OW group was significantly greater than in the UW group (P<0.01), approximately two times that of the UW group ( Figures 2 , 3 ).

The activities of intestinal digestive enzymes were detected, and the activities of lipase and amylase in the OW group were significantly higher than in the UW group (P < 0.05). The lipase activity of the OW was about 1.9 times that of the UW, and the amylase activity was about 2.5 times that of the UW. The total protein content and trypsin activity in the OW group were significantly higher than those in the UW (P < 0.01); the trypsin activity of the OW group was about 17 times that of the UW group ( Figure 4 ).

Based on 16S rRNA sequencing, we obtained 81,299 high-quality biological sequences from different individuals. The ratio of Valid data to Raw data was above 90%, and the proportion of Q20 data was above 98%. Most of the rarefaction curves for the samples reached saturation, suggesting sufficient sequencing depth for this study. Amplicon sequence variants (ASVs) were defined based on 97% identity. A total of 632 ASVs were obtained from all fish. The total number of ASVs shared by the two groups was 130; 376 ASVs were unique to the OW group (59.49%), and 126 ASVs were unique to the UW group (19.94%). PCA analysis based on ASV composition can reflect the differences and distances between samples. The same treatment groups were closely clustered; the principal components were similar, and the amount of variation was low, indicating good repeatability between samples. The treatment groups were clearly separated, and the degree of dispersion was high, indicating a large difference between the samples. The results also indicated that the composition of intestinal flora in small fish bodies was dissimilar ( Figure 5A ). Further analysis of the bacterial community composition at the genus level was conducted, and the top 30 significantly enriched bacterial genera were obtained. The bacterial community composition was similar among samples, but the abundance and dominant bacterial groups differed. The dominant strain in the OW group was Burkholderia-Caballeronia-Paraburkholderia, and the dominant strain in the UW group was Cetobacterium ( Figure 5B ).

High-throughput sequencing technology was used to conduct transcriptome sequencing, and initial data of clean reads and GC content were obtained after removing low-quality and junction sequences. A total of 71,513 clusters were obtained from the scratch assembly of the clean reads using the Trinity software. To verify the sequencing quality of the transcriptome, 16 genes were randomly selected from the database, and the expression levels of genes in individuals of different sizes were detected by qRT-PCR. In the gene expression pattern and expression profile data, the quantitative results of 12 genes (75%) were consistent with the transcriptome expression profile, indicating high accuracy of the transcriptome data ( Figure 6A ). A total of 6648 differentially expressed genes were identified in the OW group compared with the UW group, including 2523 upregulated genes and 4125 downregulated genes ( Figure 6B ). Several growth-related genes were differentially expressed, including growth hormone-releasing hormone and fibroblast growth factor. Analysis of metabolic pathways using the KEGG database showed that the pathways with enriched genes included Biosynthesis of amino acids, 2-Oxocarboxylic acid metabolism, Hematopoietic cell lineage, Glycerolipid metabolism, and the HIF-1 signaling pathway ( Figure 6C ). To further analyze the expression patterns of growth-related regulatory genes in different individuals of O. bidens, we screened genes related to amino acid metabolism, signal transduction, and other anabolic pathways from the enriched KEGG pathways for expression profile level analysis. The actual expression levels of the related genes on the pathway were verified by qRT-PCR. The results demonstrated that the quantitative data were consistent with the transcriptome expression profiles. The analysis revealed that the expression levels of growth hormone–releasing hormone, vascular endothelial growth factor A, fibroblast growth factor, aminoacylase, serine/threonine protein kinase, mitogen-activated protein kinase, heme oxygenase, and other coding genes in the metabolic pathway were significantly higher in the OW group than in the UW group ( Figure 7 ).

An association analysis identified correlations between morphological indices, intestinal structure, and digestive enzyme activity, factors closely related to intestinal microbiota but having low correlations with gene transcription ( Figure 8A ). The results of the SEM were basically consistent with the abovementioned correlation analysis, and the direct and indirect effects between variables were further verified ( Figure 8B ). Intestinal microbiota positively affected gene transcription (0.477); gene transcription positively affected the intestinal morphological structure (0.416); the intestinal morphological structure positively affected intestinal digestive enzymes (2.714), and intestinal digestive enzymes positively affected the physiological phenotype (0.516). In addition, intestinal microbiota had a positive effect on the intestinal morphological structure (0.733), and intestinal morphological structure had a positive effect on physiological phenotypes (0.513).

The microbiota and the environment are interdependent; the breeding environment is regulated by the microbiota, which are supported by the environment. The health of aquatic organisms is closely tied to their environment. As a crucial component of the internal ecosystem, the intestinal microflora not only support nutritional metabolism and immune defense but also stabilize the internal environment to maintain physiological homeostasis (Blaxter et al., 2005; Wang et al., 2012).

The intestinal tract functions to absorb food nutrients, and its morphological structure determines the digestion and absorption ability of fish. The intestinal morphological structure includes intestinal villus height, intestinal villus width, intestinal muscle thickness, crypt depth, and the pillow–crypt ratio (Zhang et al., 2020). The height and width of the intestinal villi are indicators of the intestinal tract’s ability to absorb nutrients (Sang and Fotedar, 2009). Studies have shown that intestinal microbiota is essential for maintaining normal gut structure and function (Lutgendorff et al., 2008). The soluble protein p40 derived from probiotics can activate EGFR, leading to Akt activation and thereby preventing cytokine-induced apoptosis and structural damage to intestinal epithelial cells (Yan et al., 2011). The growth and health status of fish are reflected in their intestinal microbiota and the presence of probiotics. A higher abundance of probiotics produces a more stable intestinal environment, leading to improved growth status in fish (Yu et al., 2018). Zebrafish without microflora are unable to digest or absorb large protein molecules, and the artificial construction of microbial communities can help zebrafish recover and improve their digestion and absorption of proteins (Bates et al., 2006). In this study, the intestinal flora abundance of the OW group was higher than that of the UW group, and its intestinal environment was relatively more stable, indicating a stronger ability to digest and absorb nutrients. Some species of (Burkholderia-Caballeronia-Paraburkholderia), a common component of the intestinal microbiome, produce growth hormones that can improve the growth rate and health status of the host, and thus the relative abundance of this group is often closely related to the growth and health status of the host (Blandford et al., 2018; Rong et al., 2024). Our results were similar to those of previous studies, as (Burkholderia-Caballeronia-Paraburkholderia) was dominant in the intestinal tract among the experimental groups of O. bidens. The results showed that (Burkholderia-Caballeronia-Paraburkholderia) were the most abundant bacteria in the OW group while being the second most abundant taxon in the UW group, indicating that the intestinal health of the OW group was superior to that of the UW group ( Figure 5B ). Further analysis found that the height and width of intestinal villi in the OW group were significantly greater than those in the UW group, and the higher and wider intestinal villi increased the contact area between the villi and nutrients, thereby improving the digestive and absorption functions of the intestine. The thickness of the intestinal muscle can also reflect the integrity of the intestinal tissue structure. The muscle thickness of the OW group was about 1.4 times that of the UW group, and the thicker the muscle thickness, the greater the intestinal absorption efficiency ( Figures 2 and 3). Previous studies have confirmed that improving the intestinal structure can enhance the growth performance of fish (Zhu, 2021). This study suggests a correlation between the intestinal microbiota and the intestinal morphological structure of O. bidens. The correlation analysis revealed a positive association between intestinal microbiota and intestinal morphological structure, with a correlation coefficient of 0.733 ( Figure 8 ). We hypothesized that the intestinal microbiota may influence intestinal development, thereby facilitating more complete absorption of nutrients and accelerating the growth rate.

Intestinal microbiota play a crucial role in regulating digestion and metabolism, as well as growth, development, and immune defense. Symbiotic bacteria in the intestine can directly affect the host’s digestive function. The activity of digestive enzymes in fish intestines is an important index of the ability of fish to digest and absorb nutrients (Chen et al., 2019). Lipase (LPS), amylase (AMS), and trypsin (TRY) are the main digestive enzymes in fish intestines, and their activities not only reflect the absorption and utilization efficiency of nutrients but also play a key role in growth (Qin et al., 2018; Jin et al., 2019). Probiotics can produce the aforementioned digestive enzymes in the intestine, promoting nutrient absorption and the assimilation of food by fish (Douglas, 2014). The enhancement of lipase and protein activities could improve the activity of intestinal proteases, further promoting the growth of fish (Wang et al., 2023). Trypsin, as the most important protease among intestinal digestive enzymes, is also an important indicator of the digestion and absorption ability of fish (Ding and Li, 2011). An increase in intestinal trypsin activity improved the utilization rate of carbohydrates and promoted the growth of groupers (Liu et al., 2023). This study confirmed the results of previous research, finding that the intestinal total protein content of the OW group was much higher than that of the UW group, and the activities of lipase and amylase were two to three times higher than those of the UW group. In addition, the trypsin activity of the OW group was approximately 17 times higher than that of the UW group ( Figure 4 ). The results indicated that intestinal microbiota can enhance intestinal digestive enzyme activity, thereby increasing nutrient absorption and promoting the growth of O. bidens.

The role of intestinal microbiota extends beyond regulating nutrient absorption in fish to influencing the expression of specific genes. Rawls et al. (2004) studied 79 young zebrafish and found that apolipoprotein, myeloperoxidase, serum amyloprotein, and other genes exhibited strong responses to the presence or absence of ungraded microbiota in the gut and that intestinal microbiota controlled the expression of 212 genes during aseptic basal culture. In this study, the results of gene expression profiles showed that the KEGG pathways of differentially expressed genes were concentrated in cell differentiation, the PI3K-Akt signaling pathway (promoting cell survival), the phosphatidylinositol signaling system, phospholipid metabolism, and biosynthesis of secondary metabolites ( Figure 6C ). The expression patterns of differentially expressed genes in the enriched pathways were further analyzed, and GH (growth hormone), GHRH (growth hormone-releasing hormone), FGF2 (fibroblast growth factor), pfkA (vascular endothelial growth factor A), ALT (alanine aminotransferase) and other differentially expressed genes directly or indirectly related to growth were screened. Studies have shown that the GH/IGF axis plays a crucial role in regulating the growth and development of fish, with GH and GHRH being key genes in this endocrine axis (Zheng and Ru, 2012). GHRH is an important neuroendocrine factor in fish. GHRH releases GH by stimulating the anterior pituitary, which in turn regulates the growth and development of fish by binding to the growth hormone receptors (GHRs) in liver cells (Nam et al., 2011). Therefore, GH and GHRH play positive roles in regulating the growth and development of fish. In this study, GH and GHRH gene expression in the OW group was significantly higher than in the UW group. In addition, the expression levels of most genes related to positive growth regulation, such as FGF2, pfk2, and ALT, in the OW group were significantly higher than those in the UW group ( Figure 7 ). The correlation analysis of different indicators identified a positive association between intestinal microbiota and gene transcription and expression ( Figure 8 ). We hypothesized that intestinal microbiota may further participate in the regulation of the rapid growth of the OW group by positively regulating genes.