Agaricus subrufescens fermented rye affects the development of intestinal microbiota, local intestinal and innate immunity in suckling-to-nursery pigs

Abstract

Background: Agaricus subrufescens is considered as one of the most important culinary-medicinal mushrooms around the world. It has been widely suggested to be used for the development of functional food ingredients to promote human health ascribed to the various properties (e.g., anti-inflammatory, antioxidant, and immunomodulatory activities). In this context, the interest in A. subrufescens based feed ingredients as alternatives for antibiotics has also been fuelled during an era of reduced/banned antibiotics use. This study aimed to investigate the effects of a fermented feed additive -rye overgrown with mycelium (ROM) of A. subrufescens-on pig intestinal microbiota, mucosal gene expression and local and systemic immunity during early life. Piglets received ROM or a tap water placebo (Ctrl) perorally every other day from day 2 after birth until 2 weeks post-weaning. Eight animals per treatment were euthanized and dissected on days 27, 44 and 70.

Results: The results showed ROM piglets had a lower inter-individual variation of faecal microbiota composition before weaning and a lower relative abundance of proteobacterial genera in jejunum (Undibacterium and Solobacterium) and caecum (Intestinibacter and Succinivibrionaceae_UCG_001) on day 70, as compared to Ctrl piglets. ROM supplementation also influenced gut mucosal gene expression in both ileum and caecum on day 44. In ileum, ROM pigs showed increased expression of TJP1/ZO1 but decreased expression of CLDN3, CLDN5 and MUC2 than Ctrl pigs. Genes involved in TLR signalling (e.g., TICAM2, IRAK4 and LY96) were more expressed but MYD88 and TOLLIP were less expressed in ROM pigs than Ctrl animals. NOS2 and HIF1A involved in redox signalling were either decreased or increased in ROM pigs, respectively. In caecum, differentially expressed genes between two groups were mainly shown as increased expression (e.g., MUC2, PDGFRB, TOLLIP, TNFAIP3 and MYD88) in ROM pigs. Moreover, ROM animals showed higher NK cell activation in blood and enhanced IL-10 production in ex vivo stimulated MLN cells before weaning.

Conclusions: Collectively, these results suggest that ROM supplementation in early life modulates gut microbiota and (local) immune system development. Consequently, ROM supplementation may contribute to improving health of pigs during the weaning transition period and reducing antibiotics use.

Keywords: Agaricus subrufescens; Early life; Fermentation; Gut microbiota; Immunity; Pigs.

Fig. 1

Experimental set up. Piglets orally received either ROM or tap water (Ctrl group) every other day, starting from day 2 until day 44. An oral vaccine (Salmoporc®) was given on day 21 and day 45 (booster vaccination). A randomly selected subset of animals (eight per group) were euthanised on 3 days (days 27, 44 and 70). Body weight (BW) was monitored for pig growth performance. Faeces and digesta from jejunum, ileum and caecum were collected and processed for 16S rRNA gene sequencing. Ileal and caecal mucosa were sampled for immune gene expression analysis. Blood samples were taken to assess vaccine-specific antibodies (IgM, IgA and IgG) against Salmonella typhimurium (Salmoporc®). Furthermore, blood samples were processed and evaluated for immune cells, and mesenteric lymph node (MLN) samples were re-stimulated with the mitogen LPS

Fig. 2

The effect of ROM treatment on faecal microbial inter-individual variation. Boxplots show the inter-individual variation for ROM versus Ctrl pigs in faecal microbial composition over time based on unweighted (A) and weighted (B) UniFrac distances. Significance between treatments was assessed by Wilcoxon rank sum test per time point. Asterisks are used to indicate the statistical differences between treatments (**p < 0.01, ***p < 0.001, ****p < 0.0001). Blue and yellow colours represent control (Ctrl) and treated (ROM) groups, respectively

Fig. 3

Differentially abundant genera between ROM and Ctrl pigs during pre-or post weaning. Plotted genera were identified through GAMLSS-BEZI model with random effect. The p value was corrected by FDR for multiple testing. Blue and yellow colours represent control (Ctrl) and treated (ROM) groups, respectively

Fig. 4

The effect of ROM treatment on luminal microbiota colonization. Boxplots show comparisons of jejunal luminal microbial alpha diversity between ROM and Ctrl groups on day 70 based on observed richness (A), phylogenetic diversity (B), Shannon diversity (C) and inverse Simpson (InvSimpson) (D), as well as caecal luminal microbial diversity between two groups on day 44 with indices of observed richness (E) and Shannon diversity (F). Differences in alpha diversity between groups were evaluated by nonparametric Wilcoxon rank sum test. Principal coordinate analysis (PCoA) plots for jejunal luminal microbiota (G) and caecal luminal microbiota (H) at ASV level both on day 70 based on unweighted UniFrac. Significance of the difference between ROM and Ctrl groups was assessed using PERMANOVA. The percentages at the axes indicate the variation explained. Linear discriminant analysis (LDA) Effect size (LEfSe) of differentially abundant jejunal taxa between ROM and Ctrl groups on day 70 (I). Cladogram of caecal differentially abundant taxa between ROM and Ctrl pigs on day 70 according to LEfSe (J). Blue and yellow colours represent control (Ctrl) and treated (ROM) groups, respectively

Fig. 5

Redundancy analysis (RDA) triplots show the association between the variation of gene expression and environmental variables, focusing on samples from ileum on day 27 (A) and day 44 (B), as well as caecum (C) on day 44. Blue and yellow colours represent control (Ctrl) and treated (ROM) groups, respectively. Dark grey arrows indicate environmental variables and light grey arrows show genes for which variation in expression levels is best explained in the model. The percentages at the axes indicate the variation explained by the first two canonical axes

Fig. 6

Volcano plots of all expressed genes in mucosa collected from ileum of Ctrl- (A) and ROM (B) piglets, as well as from caecum of Ctrl- (C) and ROM (D) piglets by comparisons between day 44 and day 27 within each group. The x-axis is the log2 fold change (log2 FC) for the ratio day 44 versus day 27 and the y-axis is the − log10 p value that was adjusted by FDR for multiple testing. The red dots represent genes that were either significantly more (positive fold-change) or less (negative fold-change) expressed on day 44 versus day 27 for the respective location and experimental group, with adjusted p ≤ 0.05 and a log2 FC > 0.5. Blue points indicate differentially expressed genes with an adjusted p < 0.05 and a log2 FC < 0.5. Green points indicate differentially expressed genes with adjusted p > 0.05 and a log2 FC > 0.5

Fig. 7

Volcano plots of genes expressed in ileal (A) and caecal (B) mucosa on day 44, respectively. The x-axis is the log2 fold change (log2 FC) for the ratio ROM treatment versus Ctrl group and the y-axis is the − log10 p value that was adjusted by “FDR” for multiple testing. The red dots represent genes that were either significantly more (positive fold-change) or less (negative fold-change) expressed in the ROM group as compared to the Ctrl group, with adjusted p ≤ 0.05 and a log2 FC > 0.5. Green points indicate differentially expressed genes with adjusted p > 0.05 and a log2 FC > 0.5

Fig. 8

Immunological analysis. A Serum levels of Salmonella-specific IgM, IgA, and IgG prior to vaccination (day 14), 2 weeks after the initial vaccination (day 35), and 3 weeks after the booster vaccination (day 69). B IL-10 production upon ex vivo stimulation of MLN cells. MLN cells were stimulated for 24 h with 10, 1, and 0.1 μg/mL of LPS, or left unstimulated (cell culture medium only). C Activation of Natural Killer cells in PBMCs. D Gating strategy to identify Natural Killer cells. p values indicate significant differences between treatment groups over time, and asterisks are used to indicate significant differences between treatment groups after taking into account single time points (*p < 0.05). Every point represents an individual animal from a separate pen (n = 7 or 8 per treatment group), and error bars represent standard deviations. Normal distribution and equal variances of data were checked and log-transformed when necessary. No differences in IL-10 production by unstimulated MLN cells were observed post-weaning and are therefore not presented