Metabolic influence of core ciliates within the rumen microbiome

Abstract

Protozoa comprise a major fraction of the microbial biomass in the rumen microbiome, of which the entodiniomorphs (order: Entodiniomorphida) and holotrichs (order: Vestibuliferida) are consistently observed to be dominant across a diverse genetic and geographical range of ruminant hosts. Despite the apparent core role that protozoal species exert, their major biological and metabolic contributions to rumen function remain largely undescribed in vivo. Here, we have leveraged (meta)genome-centric metaproteomes from rumen fluid samples originating from both cattle and goats fed diets with varying inclusion levels of lipids and starch, to detail the specific metabolic niches that protozoa occupy in the context of their microbial co-habitants. Initial proteome estimations via total protein counts and label-free quantification highlight that entodiniomorph species Entodinium and Epidinium as well as the holotrichs Dasytricha and Isotricha comprise an extensive fraction of the total rumen metaproteome. Proteomic detection of protozoal metabolism such as hydrogenases (Dasytricha, Isotricha, Epidinium, Enoploplastron), carbohydrate-active enzymes (Epidinium, Diplodinium, Enoploplastron, Polyplastron), microbial predation (Entodinium) and volatile fatty acid production (Entodinium and Epidinium) was observed at increased levels in high methane-emitting animals. Despite certain protozoal species having well-established reputations for digesting starch, they were unexpectedly less detectable in low methane emitting-animals fed high starch diets, which were instead dominated by propionate/succinate-producing bacterial populations suspected of being resistant to predation irrespective of host. Finally, we reaffirmed our abovementioned observations in geographically independent datasets, thus illuminating the substantial metabolic influence that under-explored eukaryotic populations have in the rumen, with greater implications for both digestion and methane metabolism.

Fig. 1. Quantities of identified protozoal proteins in the rumen microbiome vary depending on host animal and dietary conditions.

Dot plots for total proteins identified (a), and average recovered metaproteomic expression (b presented as proportion of summed LFQ intensities) belonging to protozoal, bacterial, archaeal, or fungal species across the control diet (CTL) and diets supplemented with corn oil and wheat starch (COS) for dairy cattle (n = 4) and goats (n = 4). Detected protein abundances for protozoal and bacterial populations can be found in Supplementary Table S6.

Fig. 2. Detected proteins mapped to protozoal genomes/SAGs and bacterial metagenome-assembled genomes (MAGs) in the rumen microbiome of dairy cattle (n = 4) and goats (n = 4) fed either a control diet (CTL) or one supplemented with corn oil and wheat starch (COS).

The figure displays metabolically active populations (as genomes, SAGs or MAGs), with selected expressed proteins (presented as Enzyme Commission (EC) number with short enzyme descriptions) active in fiber/starch degradation, glycolysis and production of pyruvate, butyrate, acetate, and succinate in cattle (a, b) and goats (c, d) fed CTL or COS diets. a, c depict protozoal proteomes that were detected in cattle and goats respectively are presented separately to bacteria (b, d) as the scale of their protein quantification values were ~10× larger. Protein quantification values (y-axis) were calculated by considering both the number of proteins detected per MAG/SAG/genome and their LFQ intensity: we averaged LFQ intensities for each detected protein across biological replicates for each dietary condition (CTL: green or COS: orange), which were subsequently summed for all detected proteins per MAG/SAG/genome. Heatmaps show selected MAG/SAG/genome with metabolically active proteins, presented as EC numbers, recovered from cattle (RUDB-C) and goats (RUDB-G) fed either CTL or COS diets. Where average protein abundance level for a given MAG/SAG/genome was significantly different (determined by paired Wilcoxon test, p value <0.05) between the COS and CTL diet, the given MAG/SAG/genome has been marked with a red star. MAGs are presented with their MAG ID and taxonomic annotation from GTDB-tk. Genome annotations and LFQ intensities used to create heatmaps can be found in Supplementary Table S6.

Fig. 3. Reconstructed phagolysosome formation and starch metabolism of Entodinium caudatum within the rumen microbiome of goats fed the control diet (CTL) based on metaproteomic analysis.

KO identifiers for identified proteins were analysed via KEGG mapper to reconstruct expressed key features in the metabolism of En. caudatum. Dashed arrows represent proteins or pathways that were not detected in our metaproteomes but are key steps in their respective pathways. Detailed information connecting KO identifiers to their respective gene ID, LFQ and animal/diet can be found in Supplementary Table S6.

Fig. 4. Protozoal protein detection in the rumen microbiome is influenced by diet irrespective of host.

Volcano plots indicating different rumen microbiome proteins from dairy cattle (a) and goats (b) fed either the control or COS diets and which displayed both large magnitude of fold-changes in LFQ intensities (x axis) and high statistical significance (−log10 of nominal p values using an unpaired t-test, y axis). Dashed horizontal line denotes a p value <0.05 cut-off. For both animal hosts we observed an increase in protein detection for bacterial populations (greenblue) in animals fed high-starch diets compared to protozoal populations (purple), which were detected at higher LFQ intensities in animals fed the control diet.

Fig. 5. The proteomes of rumen microbiome populations from Holstein-Friesian beef cattle are affected by high starch diets.

A total of 60 beef cattle were subjected to two dietary contrasting condition: 30 animals with ad libitum feeding and 30 subjected to 125 days of feed restriction. Dietary components in both treatments consisted of 70% concentrate, and 30% grass silage, with the concentrate containing rolled barley 72.5%, soya 22.5%, molasses 3% and calf mineral 2%. Rolled barley is high in energy and starch content (~50%). Metaproteomes for a subset of 15 animals (7 restricted and 8 ad libitum) were analysed against a database that combined 19 protozoal SAGs/genome (a) and 781 MAGs and isolate genomes (b), including fungal representatives. Protozoal proteomes are presented separately as the scale of their protein quantification values were ~10× larger. Protein quantification values were calculated by considering both the number of proteins detected per MAG/SAG/genome and their abundance; we averaged LFQ intensities for each detected protein across biological replicates for each dietary condition (ad libitum: green or restricted: orange), which were subsequently summed for all detected proteins per MAG/SAG/genome. Similar to our observations in Holstein dairy heifers and Alpine goats (Fig. 2), the proteomes of protozoal species were a major fraction of the total rumen metaproteome, however they were substantially reduced in dietary conditions where starch content was higher. Further supporting our results in dairy cattle, we observed in high-starch conditions an increase in protein detection for bacterial populations affiliated to the Succinivibrionaceae. c Volcano plot indicating different rumen microbiome proteins that displayed both large magnitude of fold-changes in LFQ intensities (x axis) and high statistical significance (−log10 of nominal p values using an unpaired t-test, y axis). Dashed horizontal line denotes a p value <0.05 cut-off. Volcano plots showed an increase in protein detection for bacterial populations (greenblue) in animals fed ad libitum compared to protozoal populations (purple), which were detected at higher LFQ intensities in animals under restricted conditions. LFQ intensities used to create (a, b) can be found in Supplementary Table S8.