One hundred and sixty healthy weaned rabbits (Hyla line rabbits, 35-day-old, initial body weight of 0.96 ± 0.06 kg, equally distributed by gender) were randomly allocated to 4 groups with eight replicates of five rabbits each. The groups followed a completely randomized 2 × 2 factorial arrangement, with dietary treatments (basal diet or 100 mg/kg of yeast peptide) and LPS treatments (saline or LPS injection) as the primary factors. The groups were defined as follows: (1) CON group: basal diet with thigh muscle injection of saline, (2) YP group: basal diet supplemented with 100 mg/kg yeast peptide, (3) LPS group: basal diet with thigh muscle injection of LPS, and (4) L-YP group: basal diet supplemented with 100 mg/kg YP and thigh muscle injection of LPS. The main components of yeast peptide are: yeast peptide ≥5,000 mg/kg, crude protein ≥3%, mannan ≥0.5%, crude ash ≤11%, water ≤11%. Purchased from Beijing Enhalor Biotechnology Co., Ltd., it has a molecular weight of 1.9 kD and consists of 19 amino acid residues (GGVGKIIEYFIGGGVGRYG). The basal diet adhered to the nutritional requirements for rabbits recommended by the National Research Council (15), and its composition and nutritional levels are detailed in Table 1.

The feeding trial spanned a total of 35 days, comprising a 7-day pre-test and a 28-day formal test period. Rabbits were individually housed in cages (L × W × H: 120 cm × 80 cm × 60 cm), provided ad libitum access to feed and water throughout the experimental duration, feeding a fixed amount of feed twice a day (once at 8:00, and once at 16:00) and immunized following established procedures. On the 28th day of the experiment, one rabbit was randomly selected from each replicate. Consistent with our laboratory’s prior reports, the LPS and L-YP groups received a thigh muscle injection of 1 mL Escherichia coli LPS (Escherichia coli O55: B5, L2880, Sigma-Aldrich) at a dose of 200 μg/kg body weight, while the CON and YP groups received an equivalent amount of physiological saline.

On day 28 of the formal test period, one rabbit per replicate was selected, and blood was collected 4 h after injection of LPS or saline. The blood, obtained by venipuncture at the ear margin, was collected in 5 mL volumes, subjected to centrifugation at 4°C and 3,000 × g for 10 min, and the resulting supernatant was collected and stored frozen. Following blood collection, rabbits were euthanized with sodium pentobarbital [100 mg/kg IV, guidelines from the American Veterinary Medical Association (AVMA)]. A 3 cm section of duodenum, jejunum, ileum, and cecum tissue was promptly isolated without any treatment.

For intestinal histology, segments of the jejunum and cecum were gently rinsed three times with ice-cold saline, sheared into 1 cm segments, and then fixed in a 4% paraformaldehyde solution. Additionally, jejunal and cecal samples were collected in sterile sampling tubes and stored at −80°C for subsequent analysis.

Samples were tested using ELISA kits (acquired from Shanghai Enzyme-linked Biotechnology Co., Ltd.). Serum samples were taken, 10 μL of the sample to be tested was added to the sample well, then 40 μL of the sample diluent was added, and 100 μL of HRP was added to the sample well. The reaction hole was sealed with sealing plate film and incubated at 37°C for 60 min. Then discard the liquid and wash it 5 times with detergent. 50 μL of substrate A and B were added to each well and incubated at 37°C for 15 min without light. Finally, 50 μL of the termination solution was added and measured the OD value of each hole at a wavelength of 450 nm in 15 min. In the Excel worksheet, the standard product concentration is taken as the horizontal coordinate and the corresponding OD value is taken as the vertical coordinate to draw the linear regression curve of the standard product. The concentration range of the DAO calibration curve: 0–24 U/mL. The concentration range of the DLA calibration curve: 0–80 nmol/L. Concentrations of D-lactic acid (DLA, Cat. No. ml974725) and Diamine oxidase (DAO, Cat. No. ml027888) in serum were calculated by curve equation.

Jejunum and cecum samples were immersed in a 4% paraformaldehyde solution for 48 h to facilitate intestinal histological observation. Subsequently, these samples underwent dehydration using a JT-12S automatic dehydrator (Junjie Electronics Co., Ltd., Wuhan, China), embedding in a BMJ-A tissue embedding machine (Zhongwei Electronic Instrument Factory, Changzhou, China), and slicing into 5 μm thin sections using a rotary slicer (Leica Biosystems, Wetzla, Germany). Staining was performed with an RS36 staining machine (P.S.J. Medical Devices Co., Ltd., Changzhou, China), followed by sealing. Digital trinocular images were acquired using a BA210Digital microscope (McAuldie Industrial Group Co., Ltd.) with 40× magnification selected for observation. The height of villi and the depth of crypts were measured using the image analysis software Motic Images Advanced 3.2, and the ratio of villi height to crypt depth was calculated. Each section underwent 10 measurements, and the results were averaged.

Samples were tested using ELISA kits (acquired from Shanghai Enzyme-linked Biotechnology Co., Ltd.). The jejunal tissue was ground into tissue homogenate at 4°C in an ice bath to make the sample to be tested. Ten microliters of the sample to be tested was added to the sample well, then 40 μL of the sample diluent was added, and 100 μL of HRP was added to the sample well. The reaction hole was sealed with sealing plate film and incubated at 37°C for 60 min. Then discard the liquid and wash it 5 times with detergent. Subsequently, 50 μL of substrate A and B were added to each well and incubated at 37°C for 15 min without light. Finally, 50 μL of the termination solution was added and measured the OD value of each hole at a wavelength of 450 nm in 15 min. In the Excel worksheet, the standard product concentration is taken as the horizontal coordinate and the corresponding OD value is taken as the vertical coordinate to draw the linear regression curve of the standard product. The concentration range of the TNF-α calibration curve: 0–320 pg/mL. The concentration range of the INF-γ calibration curve: 0–800 pg/mL. The concentration range of the SIgA calibration curve: 0–20 μg/mL. The concentration range of the IL-1β calibration curve: 0–80 pg./mL. The concentration range of the IL-6 calibration curve: 0–120 pg/mL. The concentration range of the IL-10 calibration curve: 0–64 pg/mL. Concentrations of tumor necrosis factor-alpha (TNF-α, Cat. No. ml028087), interferon-gamma (INF-γ, Cat. No. ml041640), Simplified improved Gaussian approximation (SIgA, Cat. No. ml036798), interleukin-1 beta (IL-1β, Cat. No. ml027836), interleukin-6 (IL-6, Cat. No. ml027844) and interleukin-10 (IL-10, Cat. No. ml027828) in jejunum were calculated by curve equation.

A 1 g sample of cecum was weighed in a 5 mL centrifuge tube, followed by the addition of 3 mL of ultrapure water. Thorough mixing was achieved using a vortex mixer, and subsequent centrifugation at 4°C, 12,000 × g for 10 min was performed. The resulting supernatant was collected, and 1/5 volume of the supernatant was combined with 25% metaphosphoric acid. This new solution underwent an additional centrifugation at 12,000 × g for 10 min before being filtered into a chromatographic injection vial using a 0.45 mm filter membrane for the determination of short-chain fatty acids (SCFA).

Detection was carried out using a gas chromatograph (GC-2014, Shimadzu, Japan). The chromatographic column dimensions were 30 m × 0.5 μm × 0.32 mm, with nitrogen serving as the carrier gas at a flow rate of 85.1 mL/min. The temperatures of the column, detector, and injector were set at 120°C, 250°C, and 220°C, respectively.

The relative mRNA expression levels of genes including Toll-like receptors 4 (TLR4), Myeloid differentiation factor 88 (MyD88), Inhibitor of nuclear factor kappaB (IκBα), Nuclear factor kappa-B (NF-κB), TNF-α, IL-1β, IL-6, IL-10, ZO-1, Occludin, and Claudin-1 in the jejunum were determined using real-time fluorescence quantification. Primers for these genes are detailed in Table 2. Total RNA extraction from jejunum was carried out using Trizol reagent (Invitrogen, Carlsbad, United States), and the concentration of the extracted RNA was measured using a Nanodrop Lite RNA concentration needle (Thermo Fisher Scientific, Massachusetts, United States). Subsequent cDNA synthesis utilized the PrimeScript RT Reagent kit (TaKaRa, Dalian, China) following the manufacturer’s instructions. The real-time fluorescence quantification (qRT-PCR) was conducted on an MA-6000 instrument (Shanghai, China), with a reaction system volume of 20 μL. The qRT-PCR mixture comprised 10 μL of 2 × SYBR real-time PCR premixture, 0.8 μL of forward and reverse primer mixture, 1 μL of cDNA, and 7.2 μL of RNase Free dH₂O. Thermal cycling conditions involved an initial cycle at 95°C for 5 min, followed by 40 cycles of 95°C for 15 s and 60°C for 30 s. The melting curve analysis concluded the experiment. Repetition of the experiment three times utilized β-actin as the internal reference gene, and the 2^−ΔΔCt method was applied to calculate the relative expression of the target gene.

The total genomic DNA of the microbial community was extracted following the guidelines of the E.Z.N.A.^® soil DNA kit (Omega Biotek, Norcross, GA, United States). The extracted genomic DNA’s quality was assessed through 1% agarose gel electrophoresis, and DNA concentration and purity were determined using NanoDrop 2000 (Thermo Scientific, United States). The DNA served as a template for PCR amplification of the V3–V4 variable region of the 16S rRNA gene, employing the primers 338F (5′-ACTCCTACGGG GAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCT AAT-3′) containing the Barcode sequence. The PCR reaction mixture comprised 4 μL of 5× TransStart FastPfu buffer, 2 μL of 2.5 mM dNTPs, 0.8 μL of each 5 μM primer, 0.4 μL of TransStart FastPfu DNA polymerase, and 10 ng of template DNA, adjusted to a final volume of 20 μL. The amplification process consisted of an initial pre-denaturation at 95°C for 3 min, followed by 27 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s, concluding with a final stabilization and extension at 72°C for 10 min. The samples were stored at 4°C (PCR instrument: ABI GeneAmp^® 9700). The resulting products underwent purification using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, United States), were verified by 2% agarose gel electrophoresis, and quantified with the Quantus^™ Fluorometer (Promega, United States). The purified PCR products were then utilized for library construction with the NEXTFLEX Rapid DNA-Seq Kit and subsequent sequencing on Illumina’s Miseq PE300 platform (Shanghai Meiji Biomedical Technology Co., Ltd.).

ANOVA test was performed for statistical analysis, and general linear model procedures of SPSS 26.0 for a 2 × 2 factorial design were used (IBM SPSS Inc., Chicago, Illinois, United States). Each rabbit served as the statistical unit (n = 8). Serum and jejunum data were checked for normal distribution using the Shapiro Wilk test. Statistical models encompassed daily supplementation with yeast peptides (0 or 100 mg/kg), LPS treatments (saline or LPS injection) and their interactions. Statistical model: y_ijk = μ + α_i + β_j + α_i × β_j + e_ijk where y_ijk is an observed trait, μ is the overall mean, α_i is the fixed effect of immunological challenge (i = saline or LPS), β_j is the fixed effect of dietary YP (j = 0 or 100 mg/kg YP), α_i × β_j is the interaction between LPS and YP and e_ijk is the random error. When there was a significant trend for interaction, data were further analyzed by use of one-way ANOVA with Duncan’s multiple range tests. A treatment difference achieved significance at p < 0.05, while a trend was denoted when 0.05 < p < 0.10. Results were reported as mean ± standard error of the mean (SEM). The calculation of the Alpha diversity index followed, with the Wilcoxon rank sum test applied to analyze between-group differences in Alpha diversity. Principal Coordinate Analysis (PCoA) based on the Bray–Curtis distance algorithm tested the beta diversity of microbial communities among the samples. The non-parametric PERMANOVA test assessed the significance of differences in microbial community structure between sample groups. Additionally, Linear Discriminant Analysis Effect Size (LEfSe) analysis (LDA >2, p < 0.05) was utilized to identify bacterial taxa exhibiting significant differences in abundance at the genus level across distinct groups. Correlations between bacterial abundance (at the genera level) and environmental factors were evaluated by Spearman’s correlation test by use of the R language package “Pheatmap.”

As depicted in Figure 1, compared with the CON group, LPS challenge significantly elevated serum DAO activity (p < 0.05) and DLA content (p < 0.05) (Figures 1A,B). There were no significant differences in DAO and DLA levels in the YP group compared to the CON group (p > 0.05). Notably, a significant interaction between LPS and yeast peptide was observed for both serum DAO activity and DLA content (p < 0.05). Compared with LPS group, serum DAO activity and DLA content in L-YP group were decreased by 19.11 and 12.36%.

Figure 2 illustrates the impact of yeast peptides and LPS stress on the intestinal morphology of rabbit. LPS challenge induced pronounced intestinal congestion (Figure 2A), leading to mucosal damage characterized by atrophied small intestinal villi and mucosal detachment (Figure 2B). In comparison to the CON group, LPS challenge resulted in significant intestinal mucosal injury, marked by a reduction in cecal mucosal thickness (p < 0.05), a notable decrease in jejunal villus height and the VCR (p < 0.05). However, no significant changes were observed in the thickness of the intestinal wall or the depth of the crypts (p > 0.05). Simultaneously, the mucosal depth of the jejunum exhibited a significant reduction (p < 0.05).

There was no significant interaction between yeast peptides and LPS stress on intestinal morphology (p > 0.05). Nevertheless, yeast peptides demonstrated a beneficial effect by enhancing the height of jejunal villi in LPS-stressed jejunum (p < 0.10). Compared with the control group, yeast peptides had no significant effects on jejunum villus height (p > 0.05) and cecal mucosal depth (p > 0.05).

As depicted in Figure 3, LPS stress resulted in a notable decrease in the relative expression of Occludin, ZO-1, and Claudin-1 mRNAs compared to the control group (p < 0.05). Conversely, yeast peptides exhibited a significant augmentation in the relative expression of Occludin, ZO-1 and Claudin-1 mRNAs (p < 0.05). Notably, yeast peptides demonstrated a tendency to increase the mRNA expression of ZO-1 and Claudin-1 compared to the LPS group (p < 0.10).

As illustrated in Figure 4, LPS stress resulted in elevated levels of TNF-α, IL-1β, IL-6 and INF-γ in jejunal tissue (p < 0.05), while concurrently reducing levels of IL-10 (p < 0.05) compared to controls. There was a significant interaction between yeast peptide and LPS challenge on the contents of IL-1β, IL-10 and INF-γ in jejunal tissue (p < 0.05). Notably, yeast peptides demonstrated a positive impact on jejunal tissues by decreasing IL-1β and increasing IL-10 levels (p < 0.05), with a tendency to decrease TNF-α (p < 0.10).

As depicted in Figure 5, LPS stress resulted in a significant elevation of mRNA expression levels for TLR4, MyD88, NF-kB, TNF-α, IL-1β and IL-6 in jejunal tissue compared to the control group (p < 0.05), the mRNA expression level of IkBα was decreased. Compared with the control group, the yeast peptide significantly reduced the mRNA expression of IL-1β (p < 0.05). The interaction between yeast peptides and LPS stress influenced the expression of jejunal inflammatory factors, with yeast peptides partially reversing the effects induced by LPS. In acomparison to the LPS group, the L-YP group exhibited a significant down-regulation in the expression levels of TLR4, MyD88, NF-kB, TNF-α and IL-1β (p < 0.05). Additionally, yeast peptides increased the expression level of IL-10 mRNA in jejunal tissues compared to the control group (p < 0.05).

As depicted in Figure 6, LPS stress resulted in a decrease in the concentrations of acetic acid and propionic acid in the rabbit cecum (p < 0.05). There was significant interaction between yeast peptide and LPS challenge on acetic acid concentration (p < 0.05). In contrast to the LPS-induced changes, yeast peptides increased the concentrations of acetic, propionic and butyric acids (p < 0.05).

Taxonomic analysis of OTU representative sequences at a 97% similarity level (Figure 7E) yielded a total of 11,826 OTU taxonomic units across all samples. Among these, 908 OTUs were common to all samples, accounting for 7.68% of the total. The number of OTUs unique to each group (CON, YP, LPS, and L-YP) was 25, 17, 20, and 14, respectively. Furthermore, we used Sobs, Chao, Simpson, and Shannon indices to evaluate the impact of yeast peptides and LPS stress on microbial diversity in the cecum of rabbit. As depicted in Figures 7A–D, the Sobs index decreased (p < 0.05) in the LPS group, while Simpson’s index increased and Shannon’s index decreased in the YP group (p < 0.05), compared to the control group. Additionally, there was a trend of decreasing Chao index in the LPS group (p < 0.10). Combined observations indicated no interaction between yeast peptides and LPS stress on microbial α-diversity. To explore similarities or differences in community composition among samples from different subgroups, we conducted Beta diversity through principal coordinates analysis (PCoA) and non-metric multidimensional scaling (NMDS) analysis (Figure 7F). Results demonstrated significant segregation of microbial composition in CON, YP, LPS, and L-YP groups (PCoA-PC1: 16.59% vs. PC2: 15.18%). ANOSIM analysis between groups revealed a significant difference in samples among groups (p < 0.05).

Based on the results of taxonomic analysis, we enumerated the species composition and relative abundance of each group at the phylum and genus levels, creating a community histogram for visualization (Figure 8). At the phylum level, the top 10 species in relative abundance were analyzed as dominant phyla. As illustrated in Figure 8A, Firmicutes constituted the most abundant phylum in the cecum microbiota of rabbit (71.17–76.91%), followed by Bacteroidetes (13.65–20.95%) and Verrucomicrobia (3.34–4.24%). In the cecum contents of rabbits, a total of 150 genera were detected, with the top 20 genera in abundance summarized in a bar chart (Figure 8B). The relative abundance of 11 genera exceeded 3%, including UCG-014 (9.10–13.05%), Muribaculaceae (8.91–15.21%), NK4A214 (7.13–11.25%), unclassified Lachnospiraceae (3.72–8.65%), norank Eubacteriaceae (3.61–6.22%), and Clostridia vadinBB60 group (3.35–4.88%).

Figure 8C depicts the abundance distribution of cecum microorganisms at the phylum level. Notably, Proteobacteria emerged as the primary distinctive phylum. Under LPS stress, there was a non-significant increase in Proteobacteria compared to the control group (p > 0.05). Conversely, yeast peptide administration reversed this trend, resulting in a significantly lower abundance of Proteobacteria in the L-YP group compared to the LPS group (p < 0.05). Additionally, in the YP group, Firmicutes exhibited a non-significant decrease (p > 0.05), while Bacteroidetes displayed a non-significant increase (p > 0.05) compared to the control group. At the genus level, differences were observed in 12 bacterial genera (Figure 8D). The YP group exhibited a significantly higher abundance of Butyricimonas (p < 0.05) and increased abundance of Lactobacillus (p < 0.05), alongside decreased abundance of Flavobacteriaceae and Brevibacterium (p < 0.05) compared to the control group. Conversely, LPS stress led to a significant decrease in the relative abundance of Subdoligranulum, Erysipelotrichaceae, and Eubacterium (p < 0.05). In the L-YP group, Faecalibacterium showed increased abundance (p < 0.05), while the relative abundance of Rhodospirillales, Bilophila, Olsenella and UCG-002 significantly decreased (p < 0.05). To capture additional taxonomic units, LEfSe analysis with LDA (LDA score ≥ 2, p < 0.05, Figure 8E) scoring was employed. LEfSe identified 30 divergent taxa through linear analysis. The control group exhibited enrichment in Erysipelotrichaceae, UCG-001, UCG-002, Oscillospira and Brevibacterium. In the YP group, Subdoligranulum, Butyricimonas, Clostridium and Bacteroidales were significantly enriched, including the rare Chlorella spp., with Subdoligranulum having the highest LDA score. The LPS group displayed higher abundance in the Proteobacteria, Gammaproteobacteria, Burkholderiales and Tyzzerella. In contrast, the L-YP group showed differential enrichment in Blautia, Lactobacillales, Odoribacter and Alistipe.

To evaluate the correlation between microbial taxa and environmental variables, correlation heatmap plots were generated to visually represent their relationships. At the genus level (Figure 9), cecum pH exhibited significant positive correlations with the relative abundance of Flavobacteriaceae, Ruminococcaceae, and Oscillospiraceae (p < 0.05). Conversely, it showed significant negative correlations with Subdoligranulum and Family XIII AD3011 group (p < 0.05). Subdoligranulum demonstrated a highly significant negative correlation with mRNA expression of inflammatory pathway-related factors, including TNF-α, IkBα, NF-kB, TLR4, MyD88, IL-6 and IL-1β (p < 0.05). Moreover, Family XIII AD3011 group’s relative abundance was significantly negatively correlated (p < 0.05) with IL-6 and IkBα mRNA expression, while Erysipelotrichaceae’s relative abundance was significantly negatively correlated (p < 0.05) with TNF-α, IkBα, IL-1β and MyD88 mRNA expression. IkBα mRNA expression also exhibited a significant negative correlation (p < 0.05) with unclassified Lachnospira. Conversely, IL-6 and TNF-α mRNA expression displayed a significant positive correlation with the relative abundance of Flavobacteriaceae (p < 0.05). Bacteroides’ relative abundance exhibited a strong positive correlation with the mRNA expression of NF-kB and TLR4 (p < 0.05), as well as a positive correlation with IL-1β and MyD88 mRNA expression (p < 0.05). Phascolarctobacterium’s relative abundance was significantly negatively correlated (p < 0.05) with mRNA expression of IL-6, TNF-α, TLR4, IL-1β and MyD88. Conversely, Phascolarctobacterium demonstrated a significant positive correlation (p < 0.05) with intestinal barrier-related genes (Occludin, Claudins-1 and ZO-1) and mRNA expression of IL-10. Furthermore, mRNA expression of Occludin, Claudins-1, ZO-1 and IL-10 was significantly negatively correlated with the relative abundance of Flavobacteriaceae (p < 0.05). Occludin exhibited a significant negative correlation with the relative abundance of Ruminococcus (p < 0.05). All SCFA contents in the cecum were significantly negatively correlated (p < 0.05) with Flavobacteriaceae. Specifically, acetic acid content showed a significant negative correlation (p < 0.05) with the relative abundance of Peptococcaceae, propionic acid displayed a significant negative correlation (p < 0.05) with the relative abundance of Ruminococcus, and butyric acid was significantly negatively correlated (p < 0.05) with the relative abundance of Peptococcaceae.