A metaproteomic analysis of the piglet fecal microbiome across the weaning transition

Abstract

Microbiome analysis has relied largely on metagenomics to characterize microbial populations and predict their functions. Here, we used a metaproteomic analysis of the fecal microbiome in piglets before and after weaning to compare protein abundances as they pertain to microbial populations specific to either a milk- or plant-based diet. Fecal samples were collected from six piglets on the day of weaning and 4 weeks after transitioning to a standard nursery diet. Using the 12,554 protein groups identified in samples, we confirmed the shift in protein composition that takes place in response to the microbial succession following weaning and demonstrated the redundancy in metabolic processes between taxa. We identified taxa with roles as primary degraders based on corresponding proteins synthesized, thereby providing evidence for cross-feeding. Proteins associated with the breakdown of milk-specific carbohydrates were common among pre-weaned pigs, whereas the proteome of post-weaned piglets contained a greater abundance of proteins involved in the breaking down plant-specific carbohydrates. Furthermore, output revealed that production of propionate takes place via the propionaldehyde pathway in pre-weaned piglets, but changes to production via the succinate pathway in post-weaned piglets. Finally, a disproportionate quantity of carbohydrate-active enzymes (CAZymes) (~8%) were produced by fungi, which typically only represent ~0.1% of the microbiome taxa. Information gathered through this characterization of the metaproteome before and after weaning revealed important differences regarding the role of members in the microbial community, thereby providing information for the optimization of diets and products for both piglet and microbiome health.

Keywords: digestion; metaproteome; microbiome; swine; weaning.

Figure 1

Protein abundance of the gut microbiome in pigs before and after weaning (n = 6). (A) Boxplot of top phyla ordered by increasing protein abundance. (B) Box plots of top genera ordered by increasing protein abundance at pre-weaning. (C) Connecting lines displaying the highest shift in protein abundance by genera during weaning. Red denotes an increase, and blue denotes a decrease in the abundance of protein groups after weaning.

Figure 2

Overview of oligosaccharide degradation, carbohydrate utilization and production of SCFAs by the pig’s microbiota. (1) Complex non-digestible oligosaccharides are broken into simple monosaccharides units by the action of specialized microbes (Primary degraders). (2) Monosaccharides (hexoses/pentoses) then enter the glycolysis pathway where these are converted into phosphoenolpyruvate and ultimately into pyruvate by microbial fermenters. (3) Pyruvate then serves as the main precursor for the generation of short chain fatty acids (acetate, propionate, and butyrate). (4) Propionate production largely takes place via succinate. (5) Other sugars (like the deoxy sugars, fucose and rhamnose) can enter the propanediol pathway to generate propionate. (6) Production of acetate can also take place via the degradation of ethanolamine by specialized microbes.

Figure 3

Taxonomic and abundance distribution of proteins involved in oligosaccharide digestion (CAZymes). Matrix of LogFC for CAZymes per taxon against their specific substrate for pre- and post-weaned pigs (n = 6). Protein groups assigned to the same genus that work on the same substrate were summed and represented by a single square. Annotations in bold denote phylum and represent the summed abundances of all lower taxonomic levels within that phylum.

Figure 4

Carbohydrate and SCFA metabolism in pre- and post-weaned pigs. Differences in pathway abundance (KEGG pathways) between pre- and post-weaned groups (n = 6), were calculated using Wilcoxon paired test followed by Bonferroni correction. Proteins groups that could not be assigned to a single pathway, were joined into a single larger pathway. Asterisks * denote significant differences between groups (p-value ≤ 0.05). Mean abundance values are reported, and error bar indicates the standard error of the mean (SEM). (A) Comparison of carbohydrate metabolism between pre- and post-weaned pigs, only six pathways with the highest trending differences are shown. (B) Comparison of SCFA metabolism between pre- and post-weaned pigs.

Figure 5

Taxonomic and abundance distribution of proteins involved in carbohydrate metabolism. Matrix of LogFC for carbohydrate associated proteins per taxon against their specific carbohydrate pathway (n = 6). Protein groups assigned to the same genus that work on the same substrate were summed and represented by a single square. Annotations in bold denote phylum and represent the summed abundances of all lower taxonomic levels within that phylum.

Figure 6

Taxonomic and abundance distribution of proteins involved in SCFA metabolism. Matrix of LogFC for SCFA proteins detected in pre- and post-weaned pigs (n = 6). Protein groups assigned to the same genus within the same pathway were summed and represented by a single square. Annotations in bold denote phylum and represent the summed abundances of all lower taxonomic levels within that phylum.

Figure 7

Propionate metabolism in pre- and post-weaned pigs. (A) Schematic of propionate production via the propanediol, succinate and acrylate pathways. (B) Protein homologous for propionaldehyde dehydrogenase (PduP), methylmalonyl-CoA decarboxylase (MmdA), lactoyl-CoA dehydrogenase (LcdA) and terminal enzymes (propionate kinase, propionate CoA transferase, propionate CoA ligase and Phosphate propionyl transferase) were compared between pre- and post-weaned pigs using Wilcoxon paired test followed multiple testing correction. Asterisks * denote significant differences between groups (p-value ≤ 0.05). Mean abundance values are reported, and error bar indicates the standard error of the mean (SEM).