Lactobacillus rhamnosus GG (LGG, ATCC 53103) and its ΔluxS mutant were shared by Dr. J.P. van Pijkeren from the University of Wisconsin-Madison. The activated wild-type (WT) and luxS-deficient strain (ΔluxS) of LGG were spread in De Man Rogosa and Sharpe (MRS) solid medium containing streptomycin (100 μg/mL) and cultured anaerobically at 37°C for 48 h. A single colony was picked with a sterile inoculation loop, inoculated into MRS liquid medium, and cultured anaerobically at 37°C for 16 h until LGG with streptomycin resistance was selected. LGG was then anaerobically grown in MRS medium at 37°C for 16 h, and centrifuged at 5000 g for 10 min to discard the supernatant. LGG was resuspended in reconstituted skimmed milk and freeze-dried under vacuum for 48 h. Plate counting detected 109 CFU/g bacteria in the powder. During administration, the bacterial powder that stored in plastic bags at 4 °C was dissolved in PBS for each piglet.

Eighteen newborn piglets with similar body weight were selected and divided randomly into 3 groups with 6 piglets in each group ( Figure 1A ). Newborn piglets were orally administered with 2 mL of WT and ΔluxS on days 1, 3, 5, 7, 10, 13, 16 and 20, respectively (equivalent to 10⁹ CFU bacteria). Piglets in the control group were administered orally with 2 mL sterile PBS. Piglets were weaned on day 21 and euthanized 3 days after weaning. Body weights were recorded at the beginning (the 1th day), the weaning (the 21th day) and the slaughter day (the 25th day) in this work. Feces were collected on days 8, 15 and 20. The schematic drawing of the experimental flow chart can be found in Figure 1A . This animal procedure was approved by the Animal Care and Use Committee of Zhejiang University (ETHICS CODE Permit no. ZJU20170529).

The number of Lactobacilli excreted in feces was determined by MRS agar plates containing streptomycin. The feces collected on days 8, 15 and 20 were resuspended in PBS and serially diluted. The dilutions of various concentrations were grown on MRS agar plates overnight at 37°C, followed by colony enumeration.

Microbial DNA was extracted from ileum contents using the DNA extraction Kit (Omega Bio-tek, USA) following the manufacturer’s protocols. The total DNA was inspected by NanoDrop 2000 (Thermo Fisher, USA) and 1% agarose gel electrophoresis. The 16S rRNA genes were amplified using the specific primer for V3-V4 region (17). And the bacterial 16S rRNA gene was amplified on the Illumina platform using a double-indexed, paired-end sequencing strategy as described previously (18). KAPA HiFi Hotstart Ready Mix PCR kit is used for PCR amplification and product purification. Subsequently, amplicons were multiplexed and sequenced on the Illumina MiSeq platform at Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). The raw sequenced reads were demultiplexed, quality-filtered by fastp version 0.20.0 (19, 20). Operational taxonomic units (OTUs) were defined by clustering 97% identity sequences, and chimeric sequences were identified and removed using UPARSE (Version 7.1) (21). Classification of representative reads for each OTU was carried out utilizing the RDP Classifier algorithm with the confidence threshold of 70%. Taxonomy was assigned to the OTU referred to the SILVA database. Raw sequence data were deposited into the NCBI Sequence Read Archive (SRA) database (PRJNA856449).

The blood was processed to obtain serum by centrifugation at 3000 g for 10 min at 4°C. The content of diamine oxidase (DAO), intestinal trefoil factor (ITF), TNF-α and IL-6 in the serum were determined using commercially available ELISA kits (Shanghai FANKEL Industrial Co., Ltd, China). In addition, antioxidant indexes in serum were measured using commercially available kits (Nanjing Jiancheng Inc, China). These indexes included activities of superoxide dismutase (SOD), glutathione peroxidase (GSH), and total antioxidant capacity (T-AOC). All steps were done according to manufacturer protocols.

Hematoxylin-eosin (H&E) staining was performed as previously described to examine morphology and structure of the gut (22). Briefly, after being taken out from the 4% paraformaldehyde solution, the intestinal tissues of the piglet were washed and embedded in wax. Paraffin blocks were cut into 3-µm thick sections and subjected to deparaffinization and a series of dehydration processes. Subsequently, sections were immersed in a gradient of alcohol and stained with H&E. Intestinal morphology of the samples were characterized using an optical microscopy (Olympus Corporation, Japan) at 100× magnification.

We obtained electron microscopy images as described previously (23). Fresh ileal tissue samples were fixed with 2.5% glutaraldehyde in 0.1 M phosphate buffer for 3 h and then washed twice in the same buffer for 10 min., and post-fixed with 1% osmium tetroxide for 1 h. Subsequent dehydration in varying grades of ethanol is required (20 min each; 30%, 50%, 70%, 90%, 95% and 100% v/v). After the samples were dried with a critical point dryer, the sections were observed with a scanning electron microscope (Hitachi SU-8010, Japan).

The colons were gently separated and immediately stored in Carnoy’s fixative for Alcian blue staining (24). The sample was embedded in paraffin and then was cut into-5μm sections and deposited on the glass slides. Tissue sections were preheated at 60°C for 10 min, immersed in xylene at 60°C for 10 min, and treated with gradient ethanol (100%, 90%, 70%) for deparaffinization. After being stained with Alcian Blue, the samples under glass slide were observed with an Olympus BX63 upright fluorescence microscope equipped with Image J software to measure the number of goblet cells and mucus thickness (24). Anti-Muc2 anti-body immunofluorescence staining was used to show the internal condition of the colon directly. For immunofluorescence staining, tissue sections were covered with primary antibody at 1:200 dilution (Muc2, Biorbyt, USA) at 4°C overnight. After washed three times in PBS for 10 min, the slides were incubated with diluted secondary antibody. Finally, the slides are mounted with a mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI) (25).

Considering that the colon secretes mucus and covers the surface of intestinal epithelial cells to form a mucus barrier, the colonic mucins expression was tested by quantitative real-time PCR to (26). Total RNA was extracted from the intestinal segments using TRIzol reagent (Tiangen, China) following manufacturer’s instructions. RNA concentration and quality was determined by Nanodrop 2000. Finally, quantitative real-time PCR was performed in a Bio-Rad CFX96 real time PCR machine with the use of SYBR Green Supermix (Takara, Japan) detection protocol. Reaction program was as follows: the mixtures were incubated at 95°C for 5 min, followed by 40 cycles of 5 s at 95°C, 20 s at 60°C, and finally 30 s at 60°C. Melt curves were obtained by increasing the temperature from 65°C to 95°C in 0.5°C increment for 10 s. GAPDH was used as the housekeeping gene. Raw data were analyzed using the 2^-ΔΔCt method. The list of primers was available in Supplementary Table S1 .

Approximately 50 mg of ileum tissue was homogenized in lysis buffer supple-mented with protease and phosphatase inhibitors (Keygen Biotech, China) and centrifuged at 12,000 g for 10 min at 4°C. Total protein concentrations were detected with a bicinchoninic acid (BCA) protein assay (Keygen Biotech, China). Western blot assay was conducted as previously described using a standard protocol (27). In short, the extracted proteins were resolved on 10% or 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels and transferred to polyvinylidene fluoride (PDVF) membranes. Then incubation with primary antibodies was carried out over night at 4°C. Membranes were then washed and incubated with suitable secondary antibodies. The primary antibodies included anti-claudin-3, anti-occludin, anti-zonula occludens 1 (ZO-1) (Abcam, Cambridge, UK), anti-GAPDH, anti-p38, anti-phospho-p38 (p-p38), anti-extracellular signal-regulated kinase (anti-ERK), anti-p-ERK, anti-TLR4 and anti-IL-1β (Cell Signaling Technology, USA; 1:1000). The secondary antibody was HRP-conjugated goat anti-rabbit IgG (Abbkine, Beijing; 1:5000). Afterwards, the ultrahyper-sensitive ECL chemiluminescence kit (Beyotime Biotechnology, China) was finally used to detect the protein. The target protein bands were visualized with a gel-documentation system and analyzed with Image J software. The density values of bands were corrected by subtraction of the background values.

Data analyses were generated by one-way ANOVA using the GraphPad Prism version 8 Software. Data are presented as the mean ± SEM. Mean without sharing a common letter differ, P<0.05. P<0.05 was considered as significant and 0.05< P<0.1 was considered as a tendency.

The experimental design and workflow are shown in Figure 1A . Piglets increased their average weight from 1.36, 1.48 and 1.40 kg at initial for the Con, WT and ΔluxS groups ( Figure 1A ) to 5.34, 6.41 and 5.54 kg, respectively. The weight of piglet in the WT group was significantly higher than that in the Con and ΔluxS groups on the 21st day (weaning day) and three days after weaning (25 days of age) (P<0.05, Figure 1B ). The number of WT strain excreted in the feces was significantly higher on days 8 (P<0.05) and had a higher trend on day 15 (P=0.082) than that of the ΔluxS strain. No significant difference was observed between the experimental groups on days 20 ( Figure 1C ).

Venn diagram showed that there were 51, 166 and 52 unique taxa in the Con, WT, ΔluxS groups, respectively ( Figure 2A ). No significant differences were found in α-diversity such as Simpson index and Shannon index among the three groups ( Figures 2B, C ). The β-diversity analysis showed that no significant differences were found within and between the treatment groups, indicating that the structure of the microbial community is relatively stable ( Figure 2D ). We further compared the abundance of bacterial taxa at the phylum and genus level in piglets. The Firmicutes and Actinobacteria constituted the two predominant phyla in the ileal microbiota of the piglets ( Figure 2E ). The relative abundance of Actinobacteria and Cyanobacteria was significantly lower in the WT group than in the other two groups (P<0.05). The relative abundance of Euryarchaeota was significantly lower in the WT group than in the ΔluxS group and in the Con group (P<0.05), while there was no significant differences between the ΔluxS and the Con group ( Figures 2F-H ). At the genus level, the relative abundance of Bifidobacterium in the Con group was significantly less than the WT and ΔluxS groups (P<0.05, Figure 2J ). The relative abundances of Lachnospiraceae, Eubacterium, Frisingicoccus and Enterorhabdus in the ΔluxS group were significantly higher than the other two groups (P<0.05). The relative abundances of Catenibacterium and Collinsella were significantly increased in the ΔluxS group than in the WT group (P<0.05, Figures 2I-P ).

The impairment of intestinal barrier function was evaluated by detecting the levels of diamine oxidase (DAO) and intestinal trefoil factor (ITF) in piglet serum. There was no significant difference in DAO levels among the experimental groups, while the ITF levels were significantly increased in the WT and ΔluxS group than in the Con group (P<0.05) ( Figure 3A, B ). In the duodenum, compared with the Con group, the WT group had significantly higher villus height and VCR ratio (P<0.05), and had a decrease trend in the crypt depth (P=0.07). The VCR ratio significantly increased in the WT group than in the ΔluxS group (P<0.05) ( Figure 3C ). In the jejunum, the WT group significantly decreased the crypt depth and improved the VCR ratio compared with the ΔluxS group (P<0.05, Figure 3D ). For the ileum, the VCR ratio in the WT group were significantly higher than that in the Con and ΔluxS groups (P<0.05), the villus height in the WT group was significant higher than the ΔluxS group and had an increase trend than the Con group ( Figure 3E ).

We further visualized the inside of the ileum tissue using scanning microscopy ( Figure 3F ). Part of the ileum villi in the Con group was damaged after weaning, while the villus structure was improved in the WT and ΔluxS groups. The denser and more compact ileum microvilli were found in the WT and ΔluxS groups compared that in the Con group. At the same time, expressions of tight junction proteins were detected in ileum tissue of piglet by immunoblotting. Compared with the Con group, the WT group and ΔluxS groups had significantly increased expression of occludin protein (P<0.05) ( Figure 3G ). There was no significant difference in the expression of claudin-3 and ZO-1 proteins between the experimental groups ( Figures 3H, I ).

Colonic goblet cells and mucus layer were visualized with alcian blue staining and immunofluorescence, respectively ( Figures 4A–C ). The number of goblet cells in the in the WT and ΔluxS groups was significantly higher than that in the Con group (P<0.05). The thickness of the mucus layer in the WT group was significantly higher than that in Con and ΔluxS (P<0.05). We further determined the relative mRNA expression of mucin 1, mucin 2, mucin 13 and mucin 20 in colon tissue ( Figures 4D–G ). The relative mRNA expression of intestinal mucins proteins including mucin 1, mucin 2 and mucin 20 were significantly increased in the WT group than in the Con group (P<0.05). Meanwhile, the relative mRNA expression of mucin 1, mucin 13 and mucin 20 in the WT group were significantly higher than those in the ΔluxS group (P<0.05). The above results demonstrated that feeding of WT could promote the expression of colonic mucin in weaned piglets more than ΔluxS.

The serum inflammatory factors and antioxidant characteristics of piglets after weaning are shown in Figure 5 . The level of TNF-α in the WT group was significantly higher than the Con group (P<0.05), and there was no significant difference in TNF-α level in the ΔluxS group from the Con or WT groups ( Figure 5A ). The expression of IL-6 levels was significantly increased in the WT and ΔluxS groups compared with the Con group (P<0.05), while there was not significantly different between the WT and ΔluxS groups ( Figure 5B ). No significant differences were detected in GSH, SOD, T-AOC and MDA concentrations among the experimental groups ( Figures 5C–F ).

Our experimental results indicated that the addition of WT and ΔluxS did not alter the weight of immune organs indexes, such as spleen and liver (Not shown). We further evaluated the expression of ileum inflammatory signaling pathways by western blot. Compared with the Con group, the expression of TLR4 protein was significantly decreased in the WT and ΔluxS groups (P<0.05, Figures 6A ). The expression of p-ERK protein in the WT group was significantly lower than that in the Con (P<0.05, Figures 6C ). There was no significant difference in the expression levels of p-p38 and IL-1β among the experimental groups ( Figures 6B, D ).