Understanding host-microbiota interactions in the commercial piglet around weaning

Abstract

Weaning is a critical period in the life of pigs with repercussions on their health and welfare and on the economy of the swine industry. This study aimed to assess the effect of the commercial early weaning on gut microbiota, intestinal gene expression and serum metabolomic response via an integrated-omic approach combining 16S rRNA gene sequencing, the OpenArray gene expression technology and ¹H-NMR spectroscopy. Fourteen piglets from different litters were sampled for blood, jejunum tissue and caecal content two days before (- 2d), and three days after (+ 3d) weaning. A clearly differential ordination of caecal microbiota was observed. Higher abundances of Roseburia, Ruminococcus, Coprococcus, Dorea and Lachnospira genera in weaned piglets compared to prior to weaning showed the quick microbial changes of the piglets' gut microbiota. Downregulation of OCLN, CLDN4, MUC2, MUC13, SLC15A1 and SLC13A1 genes, also evidenced the negative impact of weaning on gut barrier and digestive functions. Metabolomic approach pinpointed significant decreases in choline, LDL, triglycerides, fatty acids, alanine and isoleucine and increases in 3-hydroxybutyrate after weaning. Moreover, the correlation between microbiota and metabolome datasets revealed the existence of metabolic clusters interrelated to different bacterial clusters. Our results demonstrate the impact of weaning stress on the piglet and give insights regarding the associations between gut microbiota and the animal gene activity and metabolic response.

Figure 1

Box plot of the alpha (a) and beta (b) diversity during lactation (LACT) and after weaning (WEAN) based on the calculation of different indices: Chao1, Shannon and Simpson for alpha diversity, and Whittaker for beta-diversity.

Figure 2

Bar plot of the relative abundances (RA) expressed in percentage of the phyla (a) and main genera (b) observed in the analysis of the microbiota of piglets by massive sequencing of the 16S rRNA gene. Bar plot LACT represents the relative abundances observed during lactation, while bar plot WEAN represents the values observed in weaned piglets. Only taxa with RA greater than 2.0% were annotated with RA percentage ± SD. Figure created with the online open-source tool Datawrapper (http://datawrapper.de).

Figure 3

Differentially abundant taxa from caecal content (ln change and adjusted P-value < 0.05) between suckling and weaned piglets at genus level. All significant genera are presented; positive values and negative values indicate greater and lower abundance, respectively, in weaned animals (WEAN group) compared to suckling piglets (LACT); taxa are sorted according to the general mean of relative abundance (the average of LACT and WEAN, indicated between brackets in %) and in decreasing order. Figure created with the online open-source tool Datawrapper (http://datawrapper.de).

Figure 4

Significant differing caecal microbiota pathways between suckling and weaned piglets (KEGG level 2). All sequence reads were used to predict functions against the KEGG database (http://www.genome.jp/kegg/) by means of PICRUSt v.1.1.3. (http://picrust.github.io/picrust/) bioinformatics software package. Difference values are expressed as difference from pre- to post-weaning. Figure created with the software package STAMP v2.1.3. (https://github.com/dparks1134/STAMP).

Figure 5

Weaning effect on serum metabolic profile of the piglets. (a) Principal components analysis (PCA) score plot of serum metabolites set from suckling piglets (orange) and weaned piglets (blue). (b) Orthogonal partial least squares discrimination analysis (OPLS-DA) score between suckling piglets (orange) and weaned piglets (blue).

Figure 6

Correlation plot between gene expression (Genes), caecal microbiota (16 S) and metabolomic (Metabolites) datasets. The correlation between caecal microbiota and serum metabolome is much higher than with jejunal gene expression (Multivariate correlation values (RV): Genes & 16S = 0.37; Genes & Metabolites = 0.22; 16S & Metabolites = 0.53). Figure created by using open-source software R v3.5.3. (https://www.r-project.org/foundation/) and the LinkHD package (https://github.com/lauzingaretti/LinkHD).

Figure 7

Heatmap showing the correlation analysis between gut microbiota and serum metabolome (¹H-NMR bucket area regions) in piglets. Red or green spots indicate positive or negative Pearson correlations between variables, respectively, and the colour intensity is directly proportional to the correlation coefficient. Four clusters (designated as I, II, III or IV) are identified for metabolite buckets and 3 major clusters (named A, B and C) for bacterial families. Figure created by using open-source software R21 v3.5.3. (https://www.r-project.org/foundation/).