Linking Microbial Community Structure and Function During the Acidified Anaerobic Digestion of Grass

Abstract

Harvesting valuable bioproducts from various renewable feedstocks is necessary for the critical development of a sustainable bioeconomy. Anaerobic digestion is a well-established technology for the conversion of wastewater and solid feedstocks to energy with the additional potential for production of process intermediates of high market values (e.g., carboxylates). In recent years, first-generation biofuels typically derived from food crops have been widely utilized as a renewable source of energy. The environmental and socioeconomic limitations of such strategy, however, have led to the development of second-generation biofuels utilizing, amongst other feedstocks, lignocellulosic biomass. In this context, the anaerobic digestion of perennial grass holds great promise for the conversion of sustainable renewable feedstock to energy and other process intermediates. The advancement of this technology however, and its implementation for industrial applications, relies on a greater understanding of the microbiome underpinning the process. To this end, microbial communities recovered from replicated anaerobic bioreactors digesting grass were analyzed. The bioreactors leachates were not buffered and acidic pH (between 5.5 and 6.3) prevailed at the time of sampling as a result of microbial activities. Community composition and transcriptionally active taxa were examined using 16S rRNA sequencing and microbial functions were investigated using metaproteomics. Bioreactor fraction, i.e., grass or leachate, was found to be the main discriminator of community analysis across the three molecular level of investigation (DNA, RNA, and proteins). Six taxa, namely Bacteroidia, Betaproteobacteria, Clostridia, Gammaproteobacteria, Methanomicrobia, and Negativicutes accounted for the large majority of the three datasets. The initial stages of grass hydrolysis were carried out by Bacteroidia, Gammaproteobacteria, and Negativicutes in the grass biofilms, in addition to Clostridia in the bioreactor leachates. Numerous glycolytic enzymes and carbohydrate transporters were detected throughout the bioreactors in addition to proteins involved in butanol and lactate production. Finally, evidence of the prevalence of stressful conditions within the bioreactors and particularly impacting Clostridia was observed in the metaproteomes. Taken together, this study highlights the functional importance of Clostridia during the anaerobic digestion of grass and thus research avenues allowing members of this taxon to thrive should be explored.

Keywords: 16S rRNA profiling; anaerobic digestion; biomolecule co-extraction; cellulosic substrate; metaproteomics.

FIGURE 1

Non-metric multidimensional scaling (NMDS) plot based on Bray–Curtis distances of the 16S rRNA sequences from grass biofilm (DNA represented in green and cDNA in red) and leachate (DNA represented in purple and cDNA in blue) fractions from bioreactors R1 (squares), R2 (circles) and R3 (triangles). Analysis of similarity (ANOSIM) was carried out to assess the statistical significance of sample groupings and the corresponding R-values and corrected P_adj-values are displayed.

FIGURE 2

Taxonomic composition, transcriptionally, and translationally active taxa of the grass biofilm and leachate communities from the triplicate reactors R1, R2, and R3. G stands for grass and L for leachate. Community composition analysis was based on taxonomic assignment of (A) 16S rRNA gene sequences from DNA samples. Transcriptionally and translationally active taxa analyses were based on taxonomic assignment of (B) 16S rRNA gene sequences from cDNA samples and (C) proteins. Percentage relative abundances are displayed. The single percentage number displayed in each panel corresponds to the total contribution of the seven taxonomic categories represented (six microbial taxa in addition to unknown) to the datasets. Numbers in brackets represent the percentage of proteins for which the corresponding assigned lowest common ancestor was of higher taxonomic level than class.

FIGURE 3

Ratios of phylogenetic assignments in grass biofilm and leachate datasets. Ratios were calculated using (n_c/n)/(N_c/N), where n_c is the number of hits (OTUs or proteins) to a given phylogenetic assignment in the grass datasets, n is the total number of hits in the corresponding grass datasets (DNA, cDNA, and proteins), N_c is the number of hits to that phylogenetic assignment in the leachate datasets, and N is the total number of hits in the corresponding leachate datasets. Statistically significant over- and under-representation of a given phylogenetic assignment between two datasets is represented by asterisks and was determined by pairwise comparisons using two-tailed Fishers’ exact test with confidence intervals at 99% significance (P_adj < 0.05).

FIGURE 4

Non-metric multidimensional scaling (NMDS) plot based on Bray–Curtis distances of the proteins extracted from grass biofilm (represented in red) and leachate (represented in blue) fractions from bioreactors R1 (squares), R2 (circles), and R3 (triangles). Analysis of similarity (ANOSIM) was carried out to assess the statistical significance of sample groupings and the corresponding R-values and corrected P_adj-values are displayed.

FIGURE 5

Distribution of functional COG categories in the grass biofilm and leachate metaproteomes. R1, R2, and R3 stands for reactor 1, reactor 2, and reactor 3.

FIGURE 6

Krona plots displaying phylogenetic classification of proteins detected at the class level on the outer ring and functional key words on the inner ring. aa, amino acid; carb, carbohydrate; met, metabolism; ox, oxidative; prod, production. For an interactive version of the Krona plot see Section “Data Accessibility.”

FIGURE 7

Circos plots displaying broad functional activities and taxonomic assignments on the outer ring and proteins numbers on the inner ring. aa, amino acid; carbohyd, carbohydrate; fun, function; hydro, hydrolysis; met, metabolism; ox, oxidative.