Three pregnant SB sows and three LA sows with the same test period and the third pregnancy with with 10–12 number of fetuses, respectively, were selected. They were raised in a single pen in the same environment and fed an antibiotic-free diet obtained from the National Research Council (18) (Table 1). After delivery, the sows and piglets were raised in the same column and the piglets suckled until they were 35 days old. For each sow, two piglets (six piglets per breed, twelve piglets in total) were randomly selected for slaughter. After the slaughter of 35-day-old piglets, quickly take about 2 cm of the middle segment of the jejunum, after preliminary cleaning, clean the intestinal cavity with sterile physiological saline, and collect the contents of the intestinal cavity by pushing them into the freezing tube. Then collect the jejunum sample (2–3 cm in the middle section) and transfer it to 4% polyformaldehyde solution for fixation. Twelve samples from the two pig breeds were analyzed using untargeted metabolomics. All animal procedures were approved by the Institutional Animal Care and Use Committee of Yunnan Agricultural University (No. YNAU20200022).

After sacrificing 35-day-old piglets, we quickly removed a 2 cm middle jejunum section. Hematoxylin and eosin (H&E) staining was performed according to the method described by Feldman et al. (19). The goblet cells of the jejunum were stained with Periodic Acid-Schiff (PAS) staining methods based on the method reported by Sorokin et al. (20). For determination of immune indices, we following antibodies from Servicebio Technology, Wuhan, CN: rabbit anti-ZO-1 (#GB11195, 1:200), rabbit anti-1L-1β, rabbit anti-1L-6 (#GB11113 and #GB11117, 1:800), rabbit anti-1L-10 (#GB11108, 1:300), mouse anti-TLR4 (#GB11269, 1:100), rabbit anti-TLR2 (#GB11554, 1:500), and rabbit anti-MUC-2 (#GB11344, 1:1,000) primary antibodies and an HRP-conjugated goat anti-rabbit IgG (#GB23303, 1:200) as a secondary antibody. The methods were based on those reported by Yang et al. (21).

The sample (50 mg) and extraction solvent (acetonitrile-methanol-water, 2:2:1, including internal standard; 1,000 μL) were added to an Eppendorf tube, vortexed for 30 s, homogenized at 45 Hz for 4 min, and sonicated for 5 min in ice water (three times). The samples were then incubated at −20 °C for 1 h, centrifuged at 12,000 rpm at 4 °C for 15 min, and then transferred and stored in liquid chromatography-mass spectrometry (LC-MS) vials at −80 °C for ultra-high-performance LC-Q Exactive Orbitrap MS (UHPLC-QE Orbitrap/MS) analysis. Quality control (QC) samples were prepared by mixing equal aliquots of supernatants from all samples.

LC-MS/MS analysis was performed using a UHPLC system (1290, Agilent Technologies) and a UPLC HSS T3 column (2.1 mm × 100 mm, 1.8 μm) coupled to a Q Exactive (Orbitrap MS, Thermo). The mobile phase A was 0.1% formic acid in water for positive, and 5 mmol/L ammonium acetate in water for negative, and the mobile phase B was acetonitrile. The elution gradient was set as follows: 0 min, 1% B; 1 min, 1% B; 8 min, 99% B; 10 min, 99% B; 10.1 min, 1% B; 12 min, 1% B. The flow rate was 0.5 mL/min. The injection volume was 2 μL. The QE mass spectrometer was used for its ability to acquire MS/MS spectra on an information-dependent basis (IDA) during an LC/MS experiment. In this mode, the acquisition software (Xcalibur 4.0.27, Thermo) continuously evaluates the full scan survey MS data as it collects and triggers the acquisition of MS/MS spectra depending on preselected criteria. ESI source conditions were set as following: Sheath gas flow rate as 45 Arb, Aux gas flow rate as 15Arb, capillary temperature 320 °C, full MS resolution as 70,000, MS/MS resolution as 17,500, collision energy as 20/40/60 eV in NCE model, spray voltage as 3.8 kV (positive) or −3.1 kV (negative), respectively.

Raw MS data (raw) files were converted to the mzML format using ProteoWizard and processed using the R package XCMS (v3.2). The data were then filtered according to the following criteria: the number of samples containing metabolites was < 50%, all sample numbers were in one group (QC was also taken as a group). Internal standardization was performed for each sample (22). Subsequently, using default values, missing values were replaced by half of the minimum values found in the dataset (23). The preprocessing generated a data matrix consisting of retention time (RT), mass-to-charge ratio (m/z), and peak intensity values. An internal MS/MS database was used for data processing, and OSI-SMMS (v1.0, Dalian Chemical Data Solutions Information Technology Co., Ltd., China) was used for peak annotation after data processing with in-house MS/MS database.

For multivariate analyses, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA), the OPLS variable importance in projection (VIP) score (threshold set to 1) was adopted to rank the metabolites that best distinguished the two groups. For univariate analysis, the t-test was used to screen for differential metabolites. Differential metabolites between the two groups were evaluated at P < 0.05 and VIP ≥ 1. For non-target data, when the form of adduct is considered later, the band M+H is preferred for the positive ion mode, and the band M-H is preferred for the negative ion mode.

Intestinal morphology and immune performance were analyzed using the Welch t-test in SPSS v22.0. The Wilcoxon rank test was used to calculate the diversity between groups in the R-item (v3.4.1) functional analysis. Correlations between variables were calculated using Spearman rank correlation in GraphPad Prism v7.0.

H&E staining results for the jejunum of LA and SB piglets are shown in Figure 1. Both jejunal villus height and villus height/crypt depth ratio were significantly higher in SB piglets than in LA piglets (P<0.05).

The results for jejunal mucosal immune factors of LA and SB piglets are shown in Figure 2. The results showed that the levels of IL-10, MUC2, and ZO-1 in the jejunum were markedly higher in SB piglets than in LA piglets (P < 0.01), while the levels of IL-6, IL-1β, and TLR-2 were markedly higher in LA piglets than in SB piglets (P < 0.01). Furthermore, the number of goblet cells in the jejunum was markedly higher in SB piglets than in LA piglets (P < 0.05).

The metabolomic analysis based on LC-MS/MS revealeds the difference in the metabolites of piglets' jejunal chyme in two different breeds. QC samples were used for sample quality control, as shown in Supplementary Figure S1, which depicts the overlap of BPC detected by QC sample mass spectrometry.

The PCA results for the two groups are shown in Figure 3. The jejunum chyme samples from LA and SB piglets were quite different, and samples from each group (including QC) were completely separated. As shown in Figure 4 and Supplementary Figure S2, this model was reliable and could be used to screen for differential metabolites.

As shown in Supplementary Table S1, 4,939 compounds were identified in each group of samples under NEG mode, including 4,655 and 284 identified by primary and secondary mass spectrometry, respectively. In total, 7,234 compounds were identified in each group of samples under POS mode, including 6,643 and 591 compounds identified by primary and secondary mass spectrometry, respectively.

Compared with LA piglets, the number of metabolites whose relative content was downregulated in the NEG and POS modes was much greater than that of metabolites whose relative content was upregulated in SB piglets. Compared with those in LA piglets, the relative contents of 407 and 809 metabolites were up- and downregulated in the cationic model, respectively, while the relative contents of 130 and 583 metabolites were up- and downregulated in the anionic mode, respectively (Table 2).

Among the top 20 metabolites identified in the NEG mode, five showed significant differences between groups, including isohyodeoxycholic acid, TDCA, chitin, alanyl-valine, and thymine (Supplementary Table S2). Cholic acid metabolites accounted for 25% of the total 20 metabolites. Isohyodeoxycholic acid levels were markedly higher in the jejunum of LA piglets than in that of SB piglets (P < 0.05), whereas TDCA, crustecdysone, alanyl-valine, and thymine levels were significantly lower (P < 0.05). Among the top 20 metabolites identified under the POS mode, four showed significant differences between groups, including bilirubin, pipecolic acid, hypoletin-8-gentiobioside, and 2-piperidone (Supplementary Table S3). Bilirubin, pipecolic acid, and hypoletin-8-gentiobioside levels were markedly higher in the jejunum of LA piglets than in that of SB piglets (P < 0.05), while 2-piperidone levels were substantially lower (P < 0.05).

Differences in metabolites were analyzed in the NEG mode, which showed that cholic acid metabolites accounted for 25% of the top 20 metabolites. Thus, we further compared jejunal cholic acid metabolites. As shown in Table 3, the TDCA content in SB piglets was markedly higher than that in LA piglets (P < 0.01).

Spearman rank correlations among intestinal morphology, immune performance, and differential metabolites were evaluated. As shown in Figure 5, TDCA positively correlated with ZO-1 (R = 0.94, P = 0.017), villus height (R = 1, P = 0.003), villus height/crypt depth ratio (R = 0.94, P = 0.017), and goblet cell number (R = 0.94, P = 0.017).