The animal trials were performed at the Huanjiang Observation and Research Station for Karst Ecosystems, Huanjiang, China. Huanjiang mini-pigs (360; half male and half female) with similar body weight (BW) and health status were selected and divided into three corresponding groups for 28-, 28-, and 26-day trials. These three trials were categorized as the 5–10 kg, 10–20 kg, and 20–30 kg BW stages groups after acclimation for 5 days. The 5–10 kg BW stage (Exp. 1) consisted of 220 pigs (28 days of age) with an average BW of 5.32 ± 0.46 kg and were randomly allocated into five CP (14, 16, 18, 20, and 22%) content groups. In the 5–10 kg BW stage, each CP content group consisted of 8–10 pens (2.0 m × 3.0 m) with five pigs in each pen. The 10–20 kg growth stage (Exp. 2) consisted of 84 pigs (60 days of age) with an average BW of 11.27 ± 1.43 kg and were randomly allocated into five CP (12, 14, 16, 18, and 20%) content groups. In the 10–20 kg BW stage, each CP content group consisted of 15–19 pens (1.5 m × 0.6 m), with one pig per pen. The 20–30 kg growth stage (Exp. 3) consisted of 56 pigs (94 days of age) with an average BW of 18.80 ± 2.21 kg and were randomly allocated into five CP (10, 12, 14, 16, and 18%) content groups. In the 20–30 kg BW stage, each CP content group consisted of 11–12 pens (1.5 m × 0.6 m), with one pig per pen.

The experimental pigs received three times (08:00, 14:00, and 20:30) meals per day and had access to feed and water ad libitum at all times. All experimental pigs were housed in a well-ventilated piggery with controlled humidity (60 ± 5%) and at 23–25°C temperature. Supplementing CP content in diets met the NRC recommended requirements (National Research Council, 2012) and the Chinese nutrient requirements (Ministry of Agriculture of the People’s Republic of China, 2004) (Supplementary Tables S1–S3). Diet premixes for individual ingredients were formulated using the recommended values by the National Research Council (2012). All pigs were in good health condition, and had no gastrointestinal diseases or any antibiotic exposure prior to the experimental trial.

Based on the average BW of each experimental pen, one pig from each pen (each CP group had eight pens; total eight pigs) was selected for the 5–10 kg BW stage; and based on the average BW of each CP content group, eight pigs from each group for the 10–20 kg and 20–30 kg BW stages were selected (12 h fasting) for sampling after euthanization by electrical stunning (120 V, 200 Hz; Electric Pig Stunner; Qingdao Jianhua Food Machinery Manufacturing Co. Ltd., Qingdao, China) and exsanguination. Approximately 2 cm of the jejunum (10 cm below the flexure of the duodenum-jejunum) and ileum (10 cm above the ileo-cecal junction) contents were sampled into 1.5 mL sterilized frozen tubes, immediately frozen in liquid nitrogen, and finally preserved at −80°C for bacterial DNA isolation and metabolome analysis.

The total bacterial genomic DNA from all intestinal samples (jejunum and ileum) was extracted with OMEGA Soil DNA Kit (M5635-02; Omega Bio-Tek, Norcross, GA, United States) following the manufacturer’s instructions and stored at −20°C prior to further analysis. The quantity and quality of the extracted DNA were confirmed using a NanoDrop NC2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States) and agarose gel electrophoresis, respectively.

The polymerase chain reaction (PCR) amplification of the bacterial 16S rRNA genes with V3–V4 region was performed using the universal forward primer 338F (5′-ACTCCTACGGGAGG CAGCA-3′) and the reverse primer 806R (5′-GGACTACHV GGGTWTCTAAT-3′). Sample-specific 7-bp barcodes were incorporated into the primers for multiplex sequencing. The PCR reaction components contained 5.00 μL of buffer (5×), 0.25 μL of Fast pfu DNA Polymerase (5.00 U/μL), 2.00 μL (2.50 mM) of dNTPs, 1.00 μL (10 μM) of each forward and reverse primer, 1.00 μL of DNA template, and 14.75 μL of ddH₂O. Thermal cycling conditions for PCR consisted of initial denaturation at 98°C for 5 min, followed by 25 cycles consisting of denaturation at 98°C for 30 s, annealing at 53°C for 30 s, and extension at 72°C for 45 s, with a final extension of 5 min at 72°C. The amplicons of PCR were purified with the Vazyme VAHTSTM DNA Clean Beads (Vazyme, Nanjing, China) and quantified using the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA, United States), following the manufacture’s protocols. After the individual quantification step, purified amplicons were pooled in an equal amount, and the pair-end (2 × 250 bp) sequencing was performed on an Illumina NovaSeq platform with NovaSeq 6,000 SP Reagent Kit (500 cycles) by the Shanghai Personal Biotechnology Co. Ltd. (Shanghai, China).

Microbiome bioinformatics analyses were performed using the QIIME2 v.4 with slight modifications according to the official tutorials.¹ Briefly, raw sequence data were demultiplexed using the demux plugin, followed by primers cutting with the cutadapt plugin. Sequences were then quality filtered, denoised, merged, and chimera removed with DADA2 plugin. The non-singleton amplicon sequence variants (ASVs) were combined with mafft and performed to construct a phylogeny with fasttree2. The alpha-diversity metrics (including Chao1, Observed_species, Shannon, Simpson, Faith’s PD, Pielou’s evenness, and Good’s coverage) and beta-diversity metrics (including weighted UniFrac, unweighted UniFrac, Jaccard distance, and Bray–Curtis dissimilarity) were determined with the diversity plugin. Taxonomy was assigned to ASVs using the classify-sklearn naïve Bayes taxonomy classifier in a feature-classifier plugin against the Green genes Databases.

Sequence data analyses were performed using the QIIME2 and R packages (v3.2.0). The ASV-level alpha-diversity metrics, including Chao1 richness estimator, Observed_species, Shannon index, and Simpson index, were calculated using the ASV table in QIIME2. Beta-diversity analysis was used to identify the microbial community structure variations among samples using the unweighted UniFrac distance metrics and visualized through principal coordinate analysis (PCoA). Principal component analysis (PCA) was conducted based on the compositional profiles at the genus level. The linear discriminant analysis effect size (LEfSe) was performed to detect differentially abundant taxa among different groups using the default parameters. Phylogenetic investigation of communities by reconstruction of unobserved state (PICRUSt) was used to characterize the functional capacity of the small intestinal microbiota of pigs.

The metabolite contents in the jejunum and ileum of Huanjiang mini-pigs were determined using a non-targeted metabolomics approach with the UPLC-HDMS. The metabolomics procedures included sample preparation, metabolite separation and detection, data preprocessing, and statistical analysis.

For metabolite identification, approximately 25 mg of each sample was weighed into a 2-mL EP tube and then added 500 μL extract solution [acetonitrile: methanol: water = 2:2:1 (v/v), with the isotopically-labeled internal standard mixture] to the EP tube. After 30 s of vortexing, the mixed samples were homogenized at 35 Hz for 4 min and sonicated in an ice-water bath for 5 min. The homogenization and sonication cycles were repeated three times. Then the samples were incubated for 1 h at −40°C and centrifuged at 12,000 × g for 15 min at 4°C. The resulting supernatants were filtered through a 0.22-μm membrane and transferred to fresh glass vials for further analysis. The quality control (QC) sample was obtained by mixing an equal aliquot of the supernatants from all samples.

An ultra-performance liquid chromatography (UPLC) system (Vanquish, Thermo Fisher Scientific, Waltham, MA, United States) with a UPLC BEH Amide column (2.10 × 100 mm, 1.70 μm) coupled with Q Exactive HFX mass spectrometer (Orbitrap MS, Thermo Fisher Scientific, Waltham, MA, United States) was used to perform LC–MS/MS analyses. The mobile phase A contained 25 mmol/L ammonium acetate and 25 mmol/L ammonia hydroxide in water, and the mobile phase B contained acetonitrile. The injection volume was 3 μL, and the temperature of the auto-sampler was set at 4°C. To acquire MS/MS spectra on an information-dependent acquisition (IDA) mode, the QE HFX mass spectrometer was used for its ability in the control of the acquisition software (Xcalibur, Thermo Fisher Scientific, Waltham, MA, United States). In this mode, the acquisition software continuously evaluated the full scan of the MS spectrum. The conditions for ESI source were set as follows: sheath gas flow rate 30 Arb, Aux gas flow rate 25 Arb, capillary temperature 350°C, full MS resolution 60,000, MS/MS resolution 7500, collision energy 10/30/60 in NCE mode, and spray voltage 3.60 kV (positive ion mode) or −3.20 kV (negative ion mode), respectively.

For peak detection, extraction, alignment, and integration, obtained raw data were converted into mzXML format by ProteoWizard and then processed with an in-house program, which was developed using R and based on XCMS. The metabolites were annotated using an in-house MS2 (secondary mass spectrometry) database (BiotreeDB v2.1). The value of the cutoff was 0.3. The PCA and orthogonal partial least squares discriminant analysis (OPLS-DA) were established by the SIMCA software v.16.0.2 (Sartorius Stedim Data Analytics AB, Umea, Sweden) to visualize the distinction and detect differential metabolites among different CP content groups. Moreover, the Kyoto Encyclopedia of Genes and Genomes (KEGG) and MetaboAnalyst 5.0 were used for pathway analysis.

All experimental data were statistically analyzed by a one-way analysis of variance using SPSS v.26.0 software (SPSS Inc., Chicago, IL, United States) package. All data were checked for normal distribution, and then Tukey post-hoc test was subjected to comparative analyses among different groups. The individual pigs were considered the experimental unit. Data are presented as means ± standard error of the mean (SEM). The significance and a trend toward differences were considered as p < 0.05 and 0.05 ≤ p < 0.10, respectively. The GraphPad Prism 8.0 (San Diego, CA, United States) was used for processing images.

Effects of different CP content in diets on the small intestinal microbiota community of Huanjiang mini-pigs are presented in Figure 1. Based on the high-throughput sequencing, a total of 5,541,169 raw sequences were generated from 30 jejunal and 29 ileal samples at 5–10 kg BW stage. After a series of processing, including filtering, denoising, and removing chimeras, a total of 3,923,641 high-quality sequences were obtained and aggregated into ASVs. Based on 97% sequencing similarity, a total of 4,569 bacterial ASVs were annotated in the jejunum, and 368 bacterial ASVs were common among different CP content groups, whereas 3,842 bacterial ASVs were annotated in the ileum, and 197 bacterial ASVs were common among different CP content groups.

At 10–20 kg BW stage, a total of 7,428,973 raw sequences were generated from 39 jejunal and 39 ileal samples of Huanjiang mini-pigs. After filtering, denoising, and removing chimeras, a total of 5,281,664 high-quality sequences were obtained and aggregated into ASVs (Figure 1). A total of 3,538 bacterial ASVs were annotated in the jejunum of which 341 were common bacterial ASVs, whereas 3,002 bacterial ASVs were annotated in the ileum of which 333 were common bacterial ASVs based on 97% sequencing similarity.

At 20–30 kg BW stage, a total of 7,234,867 raw sequences were generated from 35 jejunal and 35 ileal samples of Huanjiang mini-pigs, of which 5,044,147 high-quality sequences were obtained and aggregated into ASVs. Based on 97% similarity, 3,868 bacterial ASVs and 364 common ASVs were annotated in the jejunum, and 3,680 bacterial ASVs and 261 common ASVs were annotated in the ileum.

The small intestinal microbiota alpha-diversity, including Chao 1, Shannon, Simpson, and Observed_species indices of Huanjiang mini-pigs during different growth stages, are presented in Table 1. At 5–10 kg BW stage, there were no significant differences (p > 0.05) in the jejunum alpha-diversity indices among the five CP content groups. The 14, 16, and 18% CP contents in diets decreased (p < 0.05) the Shannon and Simpson indices in the ileum of pigs compared with the 20% CP content in the diet. At 10–20 kg BW stage, different CP content in diets had no impact (p > 0.05) in the jejunum and ileum alpha-diversity indices. At 20–30 kg BW stage, 12 and 14% CP contents in diets increased (p < 0.05) the Chao 1 index in the ileum compared with the 10, 16, and 18% CP contents. The 12 and 14% CP contents in diets displayed an increasing trend (p = 0.054) in the Shannon index in the ileum of pigs compared with the 18% CP content. Moreover, 14% CP content in the diet increased (p < 0.05) the Observed_species index in the ileum of pigs compared with the other CP contents in diets. However, there were no significant differences (p > 0.05) in the jejunum alpha-diversity indices among different CP content at 20–30 kg BW stage.

Effects of different CP content in diets on the small intestinal microbiota beta-diversity of Huanjiang mini-pigs are presented in Figure 2. The PCoA analysis showed that there were significant separations of the microbial community structure in the jejunum and ileum of pigs between 14 and 16% CP contents and 20 and 22% CP contents at 5–10 kg BW stage, as well as the microbial community structure in the ileum of pigs at 20–30 kg BW stage (Figure 2A). Further non-metric multidimensional scaling (NMDS) analysis was used to assess intergroup microbial community structure distance among different CP content in diets of the different BW groups. The microbial community structure had significant separations in the jejunum and ileum at 5–10 kg BW stage, in the jejunum at 10–20 kg BW stage, as well as in the jejunum and ileum at 20–30 kg BW stage (Figure 2B).

Based on 97% sequence similarity, the top 10 bacterial phyla and top 20 bacterial genera in the jejunum and ileum of Huanjiang mini-pigs during different BW stages were identified (Figures 3, 4). At 5–10 kg BW stage, Firmicutes (74.09, 87.41, 56.99, 48.72, and 55.81%), Proteobacteria (21.70, 10.43, 37.21, 39.82, and 39.04%), Actinobacteria (2.24, 0.78, 4.73, 7.53, and 3.37%), and Bacteroidetes (0.44, 1.30, 0.96, 1.83, and 1.23%) were the top dominant phyla in the jejunum of the 14, 16, 18, 20 and 22% CP content groups, respectively, accounting more than 90% of the total bacterial phyla (Figure 3A). Additionally, 16% CP content in the diet displayed an increasing trend (p = 0.089) of Firmicutes abundance, while 14 and 16% CP contents in diets showed a decreasing trend (p = 0.056) of Actinobacteria abundance in the jejunum compared with the 20% CP (Figure 3B). In the ileum, Firmicutes (89.33, 98.69, 76.00, 72.18, and 62.47%), Proteobacteria (10.40, 1.07, 19.98, 21.90, and 32.91%), and Actinobacteria (0.15, 0.12, 3.34, 4.70, and 3.73%) were the top three dominant phyla of the 14, 16, 18, 20, and 22% CP content groups, accounting more than 90% of the total bacterial phyla (Figure 3A). However, there were no significant differences (p > 0.05) in bacterial phyla abundances in the ileum among different CP content groups at 5–10 kg BW stage.

At 10–20 kg BW stage, the dominant phyla in the jejunum of the 12, 14, 16, 18, and 20% CP content groups were Firmicutes (86.57, 87.96, 86.76, 85.64, and 88.23%) and Proteobacteria (9.88, 10.78, 10.91, 11.57, and 10.48%), whereas in the ileum were Firmicutes (99.36, 92.93, 97.86, 99.27, and 99.58%) and Proteobacteria (0.43, 2.18, 1.85, 0.51, and 0.25%) (Figure 3A). There were no significant differences (p > 0.05) in the abundances of bacterial phyla in the jejunum and ileum at 10–20 kg BW stage.

At 20–30 kg BW stage, Firmicutes (70.38, 53.87, 65.03, 72.82, and 59.72%), Proteobacteria (26.94, 41.51, 31.34, 24.69, and 35.27%), and Bacteroidetes (1.06, 3.49, 1.88, 1.77, and 3.36%) in the jejunum and Firmicutes (98.32, 93.17, 94.44, 96.18, and 82.31%) and Proteobacteria (1.51, 6.40, 4.78, 3.69, and 17.62%) in the ileum were the top abundant bacterial phyla of the 10, 12, 14, 16, and 18% CP content groups (Figure 3A). However, there were no significant differences (p > 0.05) in the abundances of bacterial phyla in the jejunum and ileum at 20–30 kg BW stage.

Effects of different CP content in diets on the small intestinal microbial community composition of Huanjiang mini-pigs during different BW stages at the genus level are presented in Figure 4. At 5–10 kg BW stage, Lactobacillus (61.38, 84.12, 42.24, 13.50, and 29.75%), Sphingomonas (6.92, 3.58, 17.22, 7.29, and 16.54%), and Streptococcus (3.72, 1.26, 3.68, 14.25, and 13.61%) were the most dominant genera in the jejunum of the 14, 16, 18, 20, and 22% CP content groups. In addition, bacterial genera in the jejunum with >1% relative abundances were Aquabacterium, Brevundimonas, and Acinetobacter (Figure 4A). The diet consisting of 16% CP increased (p < 0.05) the relative abundance of Lactobacillus in the jejunum of pigs compared with the 18–22% CP contents, while 16% CP content in the diet displayed a decreasing trend (p = 0.086) of Sphingomonas abundance in the jejunum of pigs compared with the 18 and 22% CP contents. The 14, 16, and 20% CP contents in diets decreased (p < 0.05) the Pelomonas abundance in the jejunum of pigs compared with the 22% CP (Figure 4B). The dominant bacterial genera in the ileum of the 14, 16, 18, 20, and 22% CP content groups were Lactobacillus (68.03, 82.49, 37.33, 10.65, and 22.04%), Turicibacter (3.85, 7.05, 26.04, 9.91, and 14.07%), and Shigella (0.19, 0.15, 0.36, 5.64, and 15.93%) at 5–10 kg BW stage (Figure 4A). The 14–16% CP contents in diets increased (p < 0.05) the Lactobacillus abundance in the ileum of pigs compared with the 20–22% CP contents, while 14–18% CP contents in diets decreased (p < 0.05) the Streptococcus abundance compared with the 20% CP content (Figure 4C). At 10–20 kg BW stage, the most dominant genera in the jejunum of the 12, 14, 16, 18, and 20% CP content groups were Lactobacillus (70.85, 85.51, 82.72, 76.92, and 84.51%), Sphingomonas (3.80, 5.05, 3.29, 3.52, and 4.39%), and Aquabacterium (2.03, 1.60, 2.75, 3.66, and 2.13%), whereas in the ileum were Lactobacillus (60.28, 83.73, 76.83, 74.92, and 80.60%), Turicibacter (16.88, 1.42, 7.81, 14.42, and 11.29%), and Clostridiaceae-Clostridium (3.26, 3.23, 2.43, 1.19, and 0.47%) (Figure 4A). At 20–30 kg BW stage, Lactobacillus (56.62, 45.48, 50.57, 67.52, and 52.07%), Streptococcus (10.27, 11.77, 12.55, 7.35, and 20.54%), Aquabacterium (4.34, 7.35, 4.35, 4.53, and 2.93%), and Brevundimonas (3.67, 3.07, 3.94, 5.07, and 6.29%) were the most abundant genera in the jejunum, and Lactobacillus (49.91, 59.44, 31.72, 47.27, and 41.05%), Turicibacter (20.90, 15.21, 15.03, 32.62, and 28.37%), and Shigella (0, 4.46, 0, 2.69, and 16.80%) were the most abundant genera in the ileum of the 12, 14, 16, 18, and 20% CP content groups (Figure 4A). There were no significant differences (p > 0.05) in the abundances of the small intestinal genera at 10–20 kg and 20–30 kg BW stages.

Different CP content in diets on the small intestinal metabolic capacity of Huanjiang mini-pigs during different BW stages are presented in Figure 5. The LEfSe analysis revealed that 20% CP content in the diet enriched the Prevotella and Tetrathiobacter, while 22% CP content in the diet enriched the Pseudomonas, Balneimonas, and Denitromonas differential marker genera in the jejunum of pigs at 5–10 kg BW stage (Figure 5A). In the ileum, the differential marker genera Sphaerotilus (14% CP), Blautia (20% CP), and Limnobacter (22% CP) were enriched at 5–10 kg BW stage (Figure 5B). At 10–20 kg BW stage, the differential marker genera Weissella (12% CP) and Sphaerotilus (16% CP) were enriched in the ileum (Figures 5A,B). At 20–30 kg BW stage, the differential marker genera Bifidobacterium (10% CP) and Ralstonia (18% CP) in the jejunum and Elstera (10% CP), Limnobacter (10% CP), Enterococcaceae (12% CP), SMB53, Actinobacillus (14% CP), Ralstonia (16% CP), and Sphaerotilus (18% CP) in the ileum were enriched in the different CP content groups (Figures 5A,B).

The impacts of different CP content in diets of Huanjiang mini-pigs during different BW stages on the small intestinal microbiota predictive function are shown in Figure 6. The level 1 PICRUST functional prediction analysis was used to compare functional enrichment of microbiota among the five CP content groups, and metabolic pathways were divided into six functional categories (including cellular processes, environmental information processing, genetic information processing, human diseases, metabolism, and organismal systems) (Figure 6A). In addition, 45 differential gene functions were observed among different CP content groups by level 2 PICRUST function prediction (Figure 6B). At 5–10 kg BW stage, there were significant changes in the majority pathways related to cell growth and death, replication and repair, infectious diseases, energy metabolism, terpenoid and polyketide metabolism, and endocrine system and immune system-related functions or pathways in the jejunum. In the ileum, the pathways belonged to cell motility, replication and repair, translation, infectious diseases, amino acid metabolism, biosynthesis of other secondary metabolites, and lipid metabolism and digestive system related functions or pathways. At 10–20 kg BW stage, there were no significant differences in the microbial functional pathways in the jejunum and ileum of the different CP content groups. At 20–30 kg BW stage, the pathways related to cell community-prokaryote and carbohydrate metabolism were enriched in the jejunum, whereas glycan biosynthesis and metabolism and cofactors and vitamins metabolism were enriched in the ileum of pigs.

In order to reveal the effects of different CP content in diets on the small intestinal metabolome profile of Huanjiang mini-pigs during different BW stages, metabolites extracted from jejunal and ileal samples were detected by UPLC-QTOF/MS. A total of 12,494 peaks were obtained from the positive ion mode (ESI⁺) and 9,940 peaks from the negative ion mode (ESI⁻) of LC–MS. After denoising, filtering, and data standardization, 8,586 (ESI⁺) and 6,979 (ESI⁻) valid peaks were obtained from the positive and negative ion modes, respectively. Finally, 845 (ESI⁺) and 334 (ESI⁻) differential metabolite signals were detected after matching with the local database (Shanghai Biotree Biomedical Technology Co. Ltd., Shanghai, China) (Supplementary Tables S4–S7).

Combined with our previous findings on immune- and antioxidant-related data (Liu et al., 2022, 2023), metabolomics analysis was considered by comparing the highest, lowest, and best dietary CP content in diets at each BW stage. The comparison groups included 14% vs. 20% and 20% vs. 22% CP contents for the 5–10 kg BW stage, 12% vs. 16% and 16% vs. 20% CP contents for the 10–20 kg BW stage, and 10% vs. 12% and 12% vs. 18% CP contents for the 20–30 kg BW stage. Considering the higher differential metabolite signals (845 vs. 334; ESI⁺ vs. ESI⁻), this study mainly analyzed the data under the ESI ⁺ mode.

Firstly, metabolite data in the ESI⁺ mode was visualized using an unsupervised multivariate data analysis method (PCA). As shown in Figure 7A, there were no visible differences between the groups, indicating that there were no significant differences in metabolome profiles between different CP content groups. To maximize the visualization of the changes in metabolome profiles between the different CP content groups, the data collected from the ESI⁺ mode were processed using the supervised multivariate data analysis method (OPLS-DA). As shown in Figure 7B and Supplementary Figure S1, there were significant visible cluster separation effects between groups with different dietary CP content groups, indicating that different CP content in diets significantly changed the small intestinal metabolome pattern of Huanjiang mini-pigs during different BW stages.

Metabolites from different CP content groups were screened and identified based on parameter criteria of VIP > 1.0 and p < 0.05 to screen differential metabolites. Fifty-seven differential metabolites (increased 16 and decreased 41) were identified in the jejunum of the 14% CP content group compared to the 20% CP content group at 5–10 kg BW stage (Figure 8). The differential metabolites mainly included lipids and lipid-like molecules, organic acids and derivatives, and organic heterocyclic compounds. Six differential metabolites (increased three and decreased three) were identified in the jejunum of the 22% CP content group compared to the 20% CP group. These metabolites mainly included amines, amino acids, and organic oxygen compounds. In addition, 111 differential metabolites (increased three and decreased 108) were identified in the ileum of the 14% CP content group compared to the 20% CP group at 5–10 kg BW stage. The differential metabolites mainly included organic acids and derivatives, organic heterocyclic compounds, and lipids and lipid-like molecules. Furthermore, 11 differential metabolites (increased two and decreased nine) were identified in the ileum of the 22% CP content group compared to the 20% CP group at 5–10 kg BW stage. The differential metabolites mainly included organic acids and derivatives, as well as lipids and lipid-like molecules.

At 10–20 kg BW stage, 42 differential metabolites (increased four and decreased 38) were identified in the jejunum of the 12% CP content group compared to the 16% CP content group. These differential metabolites mainly included lipids and lipid-like molecules, organic acids and derivatives, organic heterocyclic compounds, and benzenes. In addition, 34 differential metabolites (all metabolites were decreased) were identified in the jejunum of the 20% CP content group compared to the 16% CP content group. These differential metabolites mainly included lipid and lipid-like molecules, phenylpropane and polyketide, and benzene. A total of 40 differential metabolites (increased two and decreased 38) were identified in the ileum of the 12% CP content group compared to the 16% CP content group. These metabolites mainly included lipid and lipid-like molecules, organic acids and derivatives, organic heterocyclic compounds, and benzene. Moreover, 14 differential metabolites (increased three and decreased 11) were identified in the ileum of the 20% CP content group compared to the 16% CP content group, and the differential metabolites mainly included lipids and lipid-like molecules (Figure 8).

At 20–30 kg BW stage, 23 differential metabolites (increased six and decreased 17) were identified in the jejunum of the 10% CP content group compared to the 12% CP content group. The differential metabolites mainly included lipids and lipid-like molecules, organic acids, and derivatives. In addition, 15 significantly increased differential metabolites were identified in the jejunum of the 18% CP content group compared to the 12% CP content group. The differential metabolites mainly included organic acids and derivatives and organic heterocyclic compounds. A total of 84 differential metabolites (increased six and decreased 78) were identified in the ileum of the 10% CP content group compared to the 12% CP content group. The differential metabolites mainly included organic heterocyclic compounds, organic acids and derivatives, lipids and lipid-like molecules, and benzenes. Furthermore, 227 differential metabolites (increased 20 and decreased 207) were identified in the ileum of the 18% CP content group compared to the 12% CP content group. The differential metabolites mainly included organic heterocyclic compounds, organic acids and derivatives, and lipids and lipid-like molecules (Figure 8).

The KEGG analysis was used to further analyze the metabolism pathways and metabolite marker of differential metabolites between the high and low CP content in diets and the optimal CP content in the diet (Figure 9). At 5–10 kg BW stage, the steroid hormone biosynthesis pathway was enriched in the jejunum of the 14% CP content group compared to the 20% CP content group, which was related to the four detected steroid differential metabolites. However, no differential metabolic pathway was detected in the jejunum of the 22% CP content group compared to the 20% CP content group (Figure 9A). Moreover, phenylalanine metabolism, aminoacyl-tRNA biosynthesis, phenylalanine/tyrosine/tryptophan biosynthesis, arginine/proline metabolism, and β-alanine metabolism pathways were enriched in the ileum of the 14% CP content group compared to the 20% CP content group, which were associated with 10 amino acids and phytohormone differential metabolites detected. However, no differential metabolite pathways were detected in the ileum of the 20% vs. 22% CP content groups (Figure 9B).

At 10–20 kg BW stage, the histidine metabolism pathway was enriched in the jejunum of the 16% CP content group compared to the 20% CP content group (Figure 9A). In addition, the primary bile acid biosynthesis pathway was enriched in the ileum of the 16% CP content group compared with the 20% CP content group, and it was associated with the detected 25-hydroxycholesterol differential metabolite (Figure 9B). However, there was no pathway enrichment in the jejunum and ileum between the 12 and 16% CP content groups.

At 20–30 kg BW stage, alanine, aspartate, and glutamate metabolic pathways were enriched in the jejunum of the 10% CP content group compared to the 12% CP content group, which were associated with differential metabolites of argininosuccinic acid (Figure 9A). Arginine and proline metabolism, and aminoacyl-tRNA biosynthesis pathways were enriched in the ileum of the 10% CP content group compared to the 12% CP content group, which were associated with six amino acids differential metabolites detected. Moreover, arginine and proline metabolism, phenylalanine metabolism, aminoacyl-tRNA biosynthesis, phenylalanine, tyrosine and tryptophan biosynthesis, d-glutamine and d-glutamate metabolism, pantothenic acid and CoA biosynthesis, β-alanine metabolism, and nitrogen metabolism pathways were enriched in the ileum of the 12% CP content group compared to the 18% CP content group. These pathways were associated with 17 amino acids, phytohormones, and carbonyl compounds (Figure 9B). However, there was no metabolic pathway enrichment in the jejunum between the 12 and 18% CP content groups.