All procedures in the current study including animal experiments and sample collection were approved by the Experimental Animal Welfare and Ethical Committee of the Institute of Animal Science, Chinese Academy of Agricultural Sciences (No. IAS2020-104).

A total of 12 healthy sows (3–4 years old, 5th−6th parity) obtained from a commercial farm (Henan, China) along with their litters were assigned to two groups: the control group (CON) and the early-life nutrition intervention group (ENI). Each group comprised a total of 6 sows, and 11 piglets were selected from each litter. The piglets among the two groups had similar initial body weight (initial BW = 1.73 ± 0.03 kg). All sows with suckling piglets were separately housed in closed farrowing pens and provided with similar commercial diets and water ad libitum. The environment of the pigpens was kept clean, ventilated, and regularly disinfected.

All piglets were fed with the same standard creep diet (Table 1) from 7 to 30 days of age. Newborn piglets in the CON group were only fed with the standard creep diet. Newborn piglets in the ENI group were provided with additional synthetic milk, a supplementary for breast milk from Day 3 to Day 30 of age. All piglets were weaned on Day 21. Piglets in the ENI group were allowed free access to synthetic milk before 22 days of age, whereas the amount of synthetic milk was reduced from 22 to 30 days. Piglets were fed with synthetic milk through feeders. The synthetic milk and feeders were purchased from Libaowei Nutrition Technology Co., Ltd (Guangdong, China). Synthetic milk was treated with ultrahigh temperature methods. Details on the feeding mode and the nutritional constituents of the synthetic milk are presented in Figure 1A and Table 1, respectively. The living weight of all piglets was recorded on Day 3, 21 and Day 30 of age. Incidence of diarrhea of piglets was recorded every day during the experimental period and the diarrhea rate was calculated as follows: Diarrhea rate (%) = (the number of piglets with diarrhea × diarrhea days) / (total number of piglets × total observational days) × 100 (16).

Four piglets randomly selected from each group were weighed at the end of the experiment (Day 30 of the pigs' age). Blood samples of 5 ml were collected from the anterior vena cava of the piglets. Blood samples were coagulated for 30 min, then serum was obtained by centrifugation at 3,000 g for 10 min at ambient temperature. Furthermore, plasma was collected using heparin as an anticoagulant, and the plasma sample was centrifuged for 10 min at 3,000 g. The piglets were euthanized under anesthesia after collection of blood samples. The length of the colon and the whole intestine was determined. Further, the weight of livers and spleen was determined. Samples of colonic chyme and mucosa were collected, immediately transferred to liquid nitrogen, and stored at −80°C for subsequent analysis. Colon tissue samples were fixed in Carnoy's solution.

Total RNA was extracted using TRIzol reagent (Ambion, UT) from colonic mucosa samples. RNA integrity was by electrophoresis using 1.0% agarose gel. Nano Drop^® ND-1000 spectrophotometer (Nano-Drop Technologies, Wilmington, DE) was used to determine quality of RNA. cDNA was synthesized from RNA using a reverse transcription kit. cDNA samples were stored at −20°C for quantitative real-time PCR (qRT-PCR). All primers were designed by the National Center for Biotechnology Information (NCBI). Primer sequences are presented in Table 2. Primers were purchased from Sangon Biotech (Shanghai, China). qRT-PCR assays were performed using TB Green Premix Ex Taq (TaKaRa, Kusatsu, Japan) at a final volume of 20 μL containing 10 μl of 2 × Top Green qPCR SuperMix (TransStart^®, Beijing, China), and 7.8 μl of ddH₂O, 0.4 μl of each primer, 1 μl of diluted cDNA, and 0.4 μl of Passive Reference Dye (50 ×). xZO-1, Claudin4, Occludin, TNF-α, IL-1β, MCP-1, IL-8, GALNT1, B3GNT6, MUC2 mRNA expression levels were explored in colonic mucosa. Expression levels of the target genes were normalized to the housekeeping genes β-actin and GAPDH as endogenous controls. qRT-PCR reaction was programmed as follows: 94°C for 30 s, and followed by 40 cycles for 5 s at 94°C, 30 s at 60°C and 10 s at 72°C. Relative expression levels of all genes were determined using the 2^−ΔΔCT method (17).

AKP activity in serum was determined using commercial reagent kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). LPS level in plasma was determined using the commercially available Tachypleus amebocyte lysate kit (Chinese Horseshoe Crab Reagent Manufactory Co., Ltd, Xiamen, China) following the quantitative Chromogenic Limulus Amebocyte Lysate assay method.

Protein expression levels of IL-8, IL-6, and TNF-α proteins in the colonic mucosa were determined using ELISA assay kits (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. A microplate reader (BioTek-ELx808, BioTek Instruments, Inc., Texas) was used to read the protein bands.

Colon tissue was immersed and fixed with Carnoy's solution for 24 h. Colon samples were then removed from formalin and embedded in paraffin. Subsequently, the paraffin blocks were sectioned to obtain 5 μm thick sections using a semi-automatic microtome (LONGSHOU, China). Sections were stained with hematoxylin and eosin (H&E) stain (18), and viewed under an optical microscope. The crypt depth was determined following a method described by Wang et al. (19).

Further, 5 μm thick sections were obtained as described above and analyzed using the PAS-AB kit (Beijing Solarbio Technology Co., Ltd, Beijing, China) instructions. Sections were viewed under an optical microscope, and the amount of goblet cells was determined using a method reported by Cantero-Recasens et al. (20).

Approximately 0.1 g of colonic chyme and mucosa samples were separately obtained from each sample. The samples were suspended in 1 mL of ddH₂O in 1.5-mL screw capped vials for analysis of concentration of SCFAs. Concentrations of SCFAs were determined by gas chromatography as described by Wu et al. (21).

Approximately 0.5–1 g of colonic chyme and colonic mucosa samples were separately obtained from each sample, and microbial community genomic DNA was extracted using E.Z.N.A.^® soil DNA Kit (D5625-02, Omega Bio-Tek Inc., Norcross, GA) according to the manufacturer's instructions. Genomic DNA samples were stored at −80°C for subsequent analysis. Purity and DNA concentration were determined through 1% agarose gel electrophoresis and NanoDrop2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA), respectively. V3-V4 regions of bacterial 16S rRNA gene were amplified using the following primer set: 338F (5'-ACTCCTACGGGAGGCAGCAG-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3'). The reaction system was comprised of 4 μl of 5 × FastPfu Buffer, 2 μl of 2.5 mM dNTPs, 0.8 μl of each primer (5 μM), 0.4 μl of FastPfu polymerase and 10 μl of DNA template. The reactions were performed using GeneAmp^® 9700 thermal cycler (Applied Biosystems, Foster City, CA). Amplification process was as follows: denaturation for 3 min at 95°C followed by 27 cycles of 95°C for 30 s, 55°C for 30 s, 72°C for 45 s, and a final extension of 10 min at 72°C. The amplified fragments were analyzed by electrophoresis on a 2% agarose gel. The products were then purified with AxyPrep DNA Gel Extraction Kit (Axygen Bioscience, CA) according to the manufacturer's instructions. Raw microbial sequence data were analyzed and processed at the Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). Sequences were analyzed and assigned to operational taxonomic units (OTUs; 97% identity). The alpha-diversity whose coverage is based on the Chao and Shannon index within each sample was determined by QIIME tool (Version 174 1.7.0) (22). Beta diversity was evaluated by computing the unweighted Unifrac distance and visualized using principal coordinates analysis (PCoA) plots. LDA effect size (LEfSe) was used to explore biomarkers that exhibited statistical differences. The datasets presented in this study were submitted to the NCBI Sequence Read Archive (SRA) database. The name of the repository and accession number is [PRJNA779481].

Growth performance, organs weight, intestinal length and morphology, colonic permeability, mRNA expression, and concentrations of SCFAs data were subjected to analysis of variance using SPSS (23.0) software. Student's t-test was used to determine differences between two groups. The results were presented as means ± SEM. The relationships among the intestinal barrier, inflammatory cytokines, and bacterial species were explored using Pearson's correlation analysis and a correlation matrix was generated. Figures were generated using GraphPad 8.0 software. * was used to indicate a statistically significant difference (P < 0.05), and ** indicated a highly significant difference (P < 0.01). The P values ranging from 0.05 to 0.1 were also recorded.

The results showed that the body weight of piglets in both groups gradually increased from Day 3 to Day 30. The body weight of piglets in the ENI group was significantly higher relative to that of piglets in the CON group at the age of 30 (P < 0.05) (Figure 1B). Diarrhea rate of piglets in the ENI group was lower compared with the rate of piglets in the CON group, and the difference was statistically significant at the post-weaning stage (P < 0.05) (Figure 1C). Piglets in the ENI group had higher liver and spleen weight and higher relative weights of liver and spleen to living body weight compared with those of piglets in the control group (P < 0.05) (Figures 1D–G). The colon of piglets in the CON group was longer relative to that of piglets in the ENI group, however, the differences were not statistically significant (Figures 1H,I).

The results showed that the activity of AKP in serum and the level of LPS in plasma of piglets in the ENI group were significantly lower compared with those of piglets in the CON group (P < 0.05) (Figures 2A,B). The findings indicated that early-life nutrition intervention significantly upregulated expression of ZO-1, Occludin, Claudin4, GALNT1, B3GNT6, and MUC2 genes in colonic mucosa at the mRNA level (Figures 2C,D).

Analysis of intestinal morphology did not show any noticeable pathologic changes in the two groups; however, the number of goblet cells and crypt depth of piglets was significantly higher in the ENI group compared with that of the CON group (P < 0.05) (Figures 3A–C).

Relative mRNA expression levels of MCP-1, TNF-α, IL-1β, and IL-8 in the ENI group were lower relative to those in the CON group, however, only the difference in relative mRNA expression level of MCP-1 was statistically significant (P < 0.05) (Figure 4A). Protein expression levels of inflammatory cytokines (TNF-α (P = 0.066), IL-6, and IL-8) in the colonic mucosa showed a similar trend as the relative mRNA expression levels of inflammatory cytokines (Figure 4B).

Concentrations of acetic acid, propionic acid, isobutyric acid, butyric acid, and valeric acid in colonic mucosa and colonic chyme of piglets in the ENI group were higher relative to those of piglets in the CON group (Figures 5A–D,F), however, the differences were not statistically significant. Concentration of isovaleric acid in colonic chyme of piglets in the ENI group was lower compared with that of piglets in the CON group, however, the difference was not statistically significant (Figure 5E).

Fresh colonic chyme and colonic mucosa were obtained from piglets, and 16 s rRNA gene sequencing analysis was performed to explore the effects of early-life nutrition interventions on the structures and composition of the gut microbiota. The findings showed no significant difference in Chao index and Shannon index in colonic mucosa between the two group (Figures 6C,D). However, Chao index and Shannon index in colonic chyme of piglets in the CON group were higher compared with those of piglets in the ENI group (P < 0.05) (Figures 6A,B). PCoA analysis showed significant differences in phylum and genus composition between ENI and CON groups (Figures 6E–H).

Microbial community composition at the phylum and genus level of the two groups is presented in Figures 7A,B,E,F. The results showed that colonic chyme samples comprised 6 major phyla including Firmicutes, Actinobacteria, Bacteroidota, Proteobacteria, Spirochaetota, and Campilobacterota (Figure 7A). The findings showed that colonic mucosa samples mainly comprised Desulfobacterota, Deferribacterota, Firmicutes, Actinobacteria, Bacteroidota, Proteobacteria, Spirochaetota, and Campilobacterota at the phylum level (Figure 7E). The proportion of Firmicutes in colonic chyme was higher in the ENI group compared with that in the CON group, whereas the proportion of Bacteroidota in colonic chyme was higher in the CON group relative to that in the ENI group (P < 0.05) (Figure 7C). Notably, the proportion of the top 6 bacteria in colonic mucosa was not significantly different between the two groups (Figure 7G).

Composition of the gut microbiota in the colonic chyme and colonic mucosa was further analyzed at the genus level. The proportion of Catenibacterium in colonic chyme was significantly higher in the ENI group (P < 0.05), whereas the proportion of norank-f-Muribaculaceae in colonic chyme was lower in the ENI group compared with that of the CON group (P < 0.05) (Figure 7D) The proportion of the top 10 bacteria in colonic mucosa was not significantly different between the two groups (Figure 7H). LEfSe analysis was performed and presented as LDA score ≥2.0 to further explore the differences in microbiota composition between the two groups. The results showed that 61 biomarkers (blue bar) were enriched in colonic chyme in the CON group compared with the levels in the ENI group. Notably, LEfSe analysis showed that only 5 biomarkers (red bar) were enriched in colonic chyme of the ENI group, including Lachnospiraceae, Lachnospirales, Firmicutes, Lactobacillaceae, and Lactobacillus (Figure 8A). Analysis of colonic mucosa showed a total of 23 biomarkers in the CON and ENI groups. The results showed that piglets in the ENI group had a higher composition of microbiota compared with that in the CON group (Figure 8B). Notably, Lactobacillus was enriched in colonic mucosa and chyme.

The potential association among intestinal microbiota, intestinal barrier, and immune-related indexes in colonic chyme and mucosa was explored. The results showed that the proportion of Lactobacillus in colonic chyme was negatively correlated with the relative mRNA expression level of MCP-1 and was positively correlated with the relative mRNA expression levels of ZO-1, Claudin4, and GALNT1 (Figure 9A). The proportion of Escherichia-Shigella was positively correlated with the relative mRNA expression level of IL-8. Moreover, proportions of Parabacteroides and Actinobacillus were positively correlated with the relative mRNA expression level of IL-1β, whereas the proportion of Clostridium-sensu-stricto-6 was negatively correlated with the relative mRNA expression levels of ZO-1 and Occludin.

Correlation analysis of the intestinal microbiota, intestinal barrier, and immune-related indexes in colonic mucosa showed similar results to those of colonic chyme (Figure 9B). The results showed that the proportion of Lactobacillus was negatively correlated with the relative mRNA expression level of MCP-1, and was positively correlated with the relative mRNA expression levels of ZO-1, Claudin4, and GALNT1. The proportion of Escherichia-Shigella was positively correlated with the relative mRNA expression levels of IL-8. The proportion of Parabacteroides was positively correlated with the relative mRNA expression level of IL-1β (Figure 10).