Revealing the combined effects of lactulose and probiotic enterococci on the swine faecal microbiota using 454 pyrosequencing

Abstract

Demand for the development of non-antibiotic growth promoters in animal production has increased in recent years. This report compared the faecal microbiota of weaned piglets under the administration of a basal diet (CON) or that containing prebiotic lactulose (LAC), probiotic Enterococcus faecium NCIMB 11181 (PRO) or their synbiotic combination (SYN). At the phylum level, the Firmicutes to Bacteroidetes ratio increased in the treatment groups compared with the CON group, and the lowest proportion of Proteobacteria was observed in the LAC group. At the family level, Enterobacteriaceae decreased in all treatments; more than a 10-fold reduction was observed in the LAC (0.99%) group compared with the CON group. At the genus level, the highest Oscillibacter proportion was detected in PRO, the highest Clostridium in LAC and the highest Lactobacillus in SYN; the abundance of Escherichia was lowest in the LAC group. Clustering in the discriminant analysis of principal components revealed distinct separation of the feeding groups (CON, LAC, PRO and SYN), showing different microbial compositions according to different feed additives or their combination. These results suggest that individual materials and their combination have unique actions and independent mechanisms for changes in the distal gut microbiota.

Figure 1

α‐Diversity measurements of pig faecal microbiota according to treatment. Microbial richness estimates (Chao1 and ACE) and diversity indices (Shannon and Simpson) provide measures of diversity within each community at an OTU identity cut‐off of 97%. Each group is labelled accordingly: CON, control; LAC, prebiotic lactulose; PRO, probiotic Enterococcus faecium NCIMB 11181; SYN, synbiotic. The Kruskal–Wallis non‐parametric test for significance was performed to assess differences among pig groups; P_{C hao1} = 0.001, P _ACE = 0.001, P _Shannon = 0.002, P _Simpson = 0.03.

Figure 2

Doughnut plots of the relative abundances of sequences at the phylum and family levels. The EzTaxon database was used to classify the taxon groups. Mean relative abundances were calculated from all samples in each group; outer and inner plots depict selected taxa at the phylum and family levels respectively.

Figure 3

Differentially abundant bacterial genera among the CON, LAC, PRO and SYN groups. Piglets in the control group received a basal diet during the entire experimental period, whereas piglets receiving treatments were fed the basal diet plus the assigned feed additive. A. The heatmap shows the 33 abundant genera (> 0.1% mean relative abundance) after normalization. The normalized levels of abundance are depicted in the colour key, where white represents the lowest (min=0) and black (max=7) shows the highest level of abundance. Columns represent treatment groups, and rows indicate the bacterial genera. B. The canonical loading plot shows peaks for the bacterial genera that had strong influences on the differentiation of the control from the treatment groups. C. Clustering of the faecal microbiota according to treatment was performed by DAPC plot using the 33 differentially abundant bacterial genera as variables.

Figure 4

OTU‐based community structures and compositions in the faecal microbiota after treatment. A. Venn diagram showing the distribution of OTUs for the CON, LAC, PRO and SYN groups. Numbers indicate the number of OTUs that were unique and the number shared (core) by two or more groups, as depicted by non‐intersecting and intersecting ellipses respectively. B. The canonical loading plot shows the peaks of all bacterial OTUs that had strong influences on the separation of the control and treatment groups. C. Clustering of the faecal microbiota according to treatment was performed by DAPC plot using all bacterial OTUs at a 95% identity cut‐off as variables.