Unveiling the metabolic potential of two soil-derived microbial consortia selected on wheat straw

Abstract

Based on the premise that plant biomass can be efficiently degraded by mixed microbial cultures and/or enzymes, we here applied a targeted metagenomics-based approach to explore the metabolic potential of two forest soil-derived lignocellulolytic microbial consortia, denoted RWS and TWS (bred on wheat straw). Using the metagenomes of three selected batches of two experimental systems, about 1.2 Gb of sequence was generated. Comparative analyses revealed an overrepresentation of predicted carbohydrate transporters (ABC, TonB and phosphotransferases), two-component sensing systems and β-glucosidases/galactosidases in the two consortia as compared to the forest soil inoculum. Additionally, "profiling" of carbohydrate-active enzymes showed significant enrichments of several genes encoding glycosyl hydrolases of families GH2, GH43, GH92 and GH95. Sequence analyses revealed these to be most strongly affiliated to genes present on the genomes of Sphingobacterium, Bacteroides, Flavobacterium and Pedobacter spp. Assembly of the RWS and TWS metagenomes generated 16,536 and 15,902 contigs of ≥10 Kb, respectively. Thirteen contigs, containing 39 glycosyl hydrolase genes, constitute novel (hemi)cellulose utilization loci with affiliation to sequences primarily found in the Bacteroidetes. Overall, this study provides deep insight in the plant polysaccharide degrading capabilities of microbial consortia bred from forest soil, highlighting their biotechnological potential.

Figure 1. Functional profile of the metagenomes based on KEGG identifiers.

(a) PCA of the seven metagenomes using the lower hierarchical functional level; (b) Relative abundance (%) of microbial functions and pairwise comparison between RWS (1W, 3W and 10W) and TWS (1T, 3T and 10T) samples, red dots represent functions within the ABC transporters; (c) Highly enriched specific functions (compared with soil inoculum FS1) along three sequential transfers (fold-increase). Abbreviations: ABC transporters (ABCT), two-component system proteins (TCSP) and phosphotransferase systems (PTS).

Figure 2

Comparison of functional profiles (percentage of relative abundance of SEED identifiers at lower hierarchical level) between the soil inoculum FS1 and metagenomes of the three sequential transfers of (a) RWS and (b) TWS. Numbers: are the selection of the eight most overrepresented functions in the consortial metagenomes. Asterisks (*) represent functions that were most enriched in RWS and TWS samples (p < 0.005). Letters a (adenylate cyclase), b (carbon monoxide dehydrogenase) and c (and cobalt/cadmium/zinc resistance proteins) correspond to the most deselected functions (p < 0.005).

Figure 3. Carbohydrate-active functional profiles.

(a) Heat map of the relative abundance (%) of each CAZy class (AA, CBM, CE, GH, GT and PL) in each metagenome; (b) Log₁₀ X-fold increase of each CAZy class in RWS and TWS samples comparatively to the soil inoculum (FS1); (c) Pairwise comparison between RWS (1W, 3W and 10W) and TWS (1T, 3T and 10T) samples, red dots: are functions belonging to glycosyl hydrolases (GH) families; (d) “Richness” values (number of clusters at 97% nucleotide identity over the total retrieved reads in each family) of the most enriched CAZy families at transfer-10 (10W and 10T) and comparison with the soil inoculum (FS1).

Figure 4

Differentially enriched CAZy families (p < 0.005, 95% confidence intervals) between FS1 and (a) 10W and (b) 10T metagenomes; (c) Taxonomic affiliation of reads belonging to families GH92, GH2, GH95, GH43, GH20, GH31, GH1, GH3, GH29 and CBM50, in 10W, 10T and FS1, using the Lowest Common Ancestor (LCA) algorithm.

Figure 5. Graphical representation of thirteen novel Bacteroidetes HULs (hemicellulose utilization loci) recovered from the metagenome assemblages.

Numbers represents annotated proteins that are flanked by glycosyl hydrolase (GH) genes. Abbreviations: Two-component system proteins (TCSP), ABC transporters (ABCT), TonB-dependent receptors (TBR), transketolases (TKT), xylose isomerase (XI) and xylulose kinase (XKN) genes.

Figure 6. Graphical explanation of the presumed catalytic mode of action, on the major component of (hemi)cellulose, by the most enriched GH in our consortia (left).

GH2 hydrolizes β-glycosidic bonds between galactose and its organic functional group; GH95 and GH29 are enzymes that hydrolyze Fuc-alpha1-2Gal linkages attached to the non-reducing ends of oligosaccharides; GH43 can act directly on xylan and release D-xylose and L-arabinose as main products. At the top: two different lignocellulose structures and microorganisms involved in their deconstruction (flask pictures were taken by the authors). Abbreviations: Acinetobacter (Ac), Klebsiella/Kluyvera (Kl), Flavobacterium (Fl), Pseudomonas (Ps) and Sphingobacterium (Sp). Right, partial metabolic reconstruction based on a theoretical (hemi)cellulose degradation pathway and uptake of sugars (intra - and extracellularly), in Bacteroidetes, using the contig_1110 genetic information. Abbreviations: Two-component system proteins (TCSP), ABC transporters (ABCT), TonB-dependent receptors (TBR), transketolases (TKT), xylose isomerase (XI) and xylulose kinase (XKN) and pentose phosphate pathway (PPP). Plus sign represent the regulation of the GH by the TCSP.