Metagenomics of the Svalbard reindeer rumen microbiome reveals abundance of polysaccharide utilization loci

Abstract

Lignocellulosic biomass remains a largely untapped source of renewable energy predominantly due to its recalcitrance and an incomplete understanding of how this is overcome in nature. We present here a compositional and comparative analysis of metagenomic data pertaining to a natural biomass-converting ecosystem adapted to austere arctic nutritional conditions, namely the rumen microbiome of Svalbard reindeer (Rangifer tarandus platyrhynchus). Community analysis showed that deeply-branched cellulolytic lineages affiliated to the Bacteroidetes and Firmicutes are dominant, whilst sequence binning methods facilitated the assemblage of metagenomic sequence for a dominant and novel Bacteroidales clade (SRM-1). Analysis of unassembled metagenomic sequence as well as metabolic reconstruction of SRM-1 revealed the presence of multiple polysaccharide utilization loci-like systems (PULs) as well as members of more than 20 glycoside hydrolase and other carbohydrate-active enzyme families targeting various polysaccharides including cellulose, xylan and pectin. Functional screening of cloned metagenome fragments revealed high cellulolytic activity and an abundance of PULs that are rich in endoglucanases (GH5) but devoid of other common enzymes thought to be involved in cellulose degradation. Combining these results with known and partly re-evaluated metagenomic data strongly indicates that much like the human distal gut, the digestive system of herbivores harbours high numbers of deeply branched and as-yet uncultured members of the Bacteroidetes that depend on PUL-like systems for plant biomass degradation.

Figure 1. Microbial community analysis of the Svalbard reindeer rumen microbiome and comparison with microbiomes from other selected gut environments.

(A) Diversity and relative abundance of the most abundant bacterial taxa identified in the rumen of the Svalbard reindeer based on phylogenetic analysis of 16S rRNA genes. (B) OTU network map showing OTU interactions between all rarefied samples from the Svalbard reindeer, Norwegian reindeer feeding on a commercial feed, Tammar wallaby, rumen and termite. OTUs are represented by dots and dot sizes reflect sequence counts within the OTU. Dot colour indicates the number of microbiomes in which the OTU was found (1 =  white, 2 =  yellow, 3 =  orange, 4 =  red). The lines radiating from each of the five grey dots link the OTUs to their source microbiomes: Svalbard reindeer, dark blue (this study); Norwegian reindeer, light blue (this study; dataset included for comparative purposes only); Termite_PL3, yellow ; bovine, green ; Tammar wallaby, orange . (C) Principal coordinate axes (PCoA) for the unweighted UniFrac analyses are coloured by host animal; Svalbard reindeer (2 samples: ▸▴) dark-blue; Norwegian reindeer (2 samples: ♦♦) light-blue; bovine rumen (•) green; Tammar wallaby (◂) red; termite (▪) orange. (D) Composition of the Svalbard reindeer rumen metagenome sequence dataset, based on sequence composition-based binning of 1394 assembled scaffolds (∼5.4 Mb) using PhyloPythiaS. For a complete rrs inventory and comparisons between the two Svalbard reindeer samples at an OTU definition of 97% ID see Table S1.

Figure 2. Comparative gene organization of presumptive cellulolytic PULs identified in the Svalbard reindeer rumen and other gut environments.

Gene clusters were recovered from the cow rumen metagenome (partial genome of an as-yet uncultured Bacteroidales phylotype AC2a) and sequenced cellulolytic fosmids constructed from environmental DNA originating from the gut microbiomes of the wallaby foregut and the reindeer and buffalo rumen (upper four clusters). All fosmid sequences originate from Bacteroidetes, according to PhyloPythiaS binning, and for all cellulolytic activity has been detected in functional screens. Green genes represent SusE/SusF-like genes predicted to encode outer-membrane proteins whose function is currently unknown. Black genes encode putative response-regulators. BACON: indicates a carbohydrate binding domain identified by . TonB: indicates members of the TonB-dependent receptor family, a group of outer membrane spanning β-barrel proteins that transport solutes and macromolecules. TW-64 and TW-33 correspond to sample IDs for the GH9 and GH5 genes (respectively) encoded within the AC2a PUL, from which expressed proteins were tested positive for hydrolytic activity on various cellulosic substrates by Hess et al., . GenBank accession numbers and/or IMG Gene Object ID numbers are provided. Gene IDs for the AC2a PUL are provided in Table S5.