Plant biomass is renewable and rich in energy, and as such, is a promising fuel alternative to fossil fuels (Goldemberg, 2009; Liao et al., 2016). However, low efficient plant cell wall deconstruction is a major barrier for economically viable biofuel production (Klein-Marcuschamer and Blanch, 2015). Plant cell walls are composed of a matrix of heteropolymers (i.e., cellulose, hemicellulose, and lignin) that are physically structured to prevent enzymatic breakdown. Commonly, a chemical and/or physical pretreatment of lignocellulosic material is applied to make cellulose accessible to hydrolytic enzymes through the solubilization of hemicelluloses and/or lignin (Mosier et al., 2005).

Natural or engineered microbial consortia are a promising approach to overcoming plant biomass recalcitrance, because of their ability to extract the majority of the energy from plant polymers in nature (Wei et al., 2009; Gessner et al., 2010; Lewin et al., 2016). The enzymatic degradation of plant material depends on multiple biomass-degrading enzymes, with complementary and synergistic activities, which are secreted by various species of micro-organisms (Allgaier et al., 2010; DeAngelis et al., 2010; Cragg et al., 2015; Lemos et al., 2017). Microbial enrichment in specific lignocellulose material is a common approach to obtain efficient and simple microbial assemblages that are able to degrade plant cell wall components (de Lima Brossi et al., 2016; Jiménez et al., 2016; Korenblum et al., 2016; Maruthamuthu et al., 2016; Mello et al., 2016). Cellulose-degrading consortia have been obtained from leaf-cutter ant refuse dumps (Lewin et al., 2016), and mixtures of plant material and chicken feces (Haruta et al., 2002). Lignin-degrading consortia have been obtained from wetland (Wang et al., 2013), rice paddy field soil (Kato et al., 2015), and redwood compost (Ceballos et al., 2017). When compared to the environmental inoculum, the enrichment derived consortia show reduced diversity, as a result of the limited number of ecological niches and selection toward more efficient degraders (Jiménez et al., 2017). The taxonomic and functional composition of lignocellulolytic microbial consortia are mainly influenced by the community in the environmental inoculum, and the substrate used for the enrichment (Cortes-Tolalpa et al., 2016; de Lima Brossi et al., 2016; Wong et al., 2016). For example, Wong et al. (2016) compared microbial enrichments derived from beaver dropping and moose rumen and found that the substrate is a major driver of the microbial consortia composition. Jiménez et al. (2016) used forest soil as an inoculum and also found distinct taxonomic composition in bacterial communities enriched in wheat straw, switchgrass and corn stover, but with similar profiles of carbohydrate-active enzyme families. Interestingly, in these studies, even after several batch cultures, the resulting microbial consortia still presented dozens of bacterial species. In plant biomass degradation, coexistence of several microbial species is possible as different members of a consortium occupy different niches, i.e., have different carbon source preferences (Jiménez et al., 2017).

In this study, we aimed to identify members of microbial consortia that have specific preferences for two major components of plant cell wall, xylan and lignin. Eight plant biomass-degrading microbial consortia were subcultured in minimal medium with xylan or alkali lignin as the sole carbon source. Deep metagenomics sequencing of these communities revealed changes in the abundance of distinct bacterial members in response to xylan and lignin-derived compounds. Furthermore, we identified metabolic pathways that were enriched after the substrate shifting.

Selective enrichment is a common approach for obtaining lignocellulose-degrading microbial consortia that can be used for biomass conversion to biofuels and other valuable biochemical. Studies have focused on identifying how inoculum sources (Cortes-Tolalpa et al., 2016) and enrichment conditions (de Lima Brossi et al., 2016; Wong et al., 2016) can affect the yield and efficiency of a lignocellulose-degrading microbial consortia. In our study, we aimed to identify community members responsible for the degradation of xylan and lignin-derived compounds. We expected that after incubation on xylan and alkali lignin, the resulting communities would be enriched in members that consume each of these substrates, independently of the history of the communities. We observed that after sub-culturing the consortia conserved the overall genetic repertoire of the original consortia, which was assessed by their k-mer composition, but the relative abundance of members carrying specific functions changed in response to the carbon source used, as observed with taxonomic and functional annotation.

The functional and taxonomic structure of the communities changed considerably after growth in alkali lignin, both in their taxonomic and functional profiles. Moreover, despite of their enrichment history, most of the consortia converged to a similar composition, with prevalence of Pseudomonas spp. and aromatic degradation genes. As mentioned in the introduction, substrate-driven convergence has already been observed in other studies (Jia et al., 2016; Wong et al., 2016). Jia et al. (2016) proposed that community convergence indicates that the selective pressure imposed by the culturing conditions is strong enough to change the initial relative abundance of community members. Lignin-derived compounds are known to be toxic to a various microorganisms, and may provide a competitive advantage to more tolerant species, such as Pseudomonas spp. (Abdelaziz et al., 2016). An increase in the abundance of Pseudomonas was associated with enrichment of genes related to aromatic compound degradation. Lignin is a heterogeneous aromatic polymer, which is found in the plant cell wall covalently cross-linked with polysaccharides (Li M. et al., 2016). Members of the Pseudomonas genus are considered major degraders of aromatic compounds in various environments (Ma et al., 2013; Terrón-González et al., 2016) and several studies have demonstrated that the Pseudomonas has the capability to catabolize low molecular weight lignin (Salvachúa et al., 2015; Tian et al., 2016; Mohan and Phale, 2017; Ravi et al., 2017). The strain P. putida CSV86 was shown to utilize aromatics, such as veratryl alcohol and ferulic acid, preferentially over glucose (Basu et al., 2006) via the ortho-cleavage pathway (Mohan and Phale, 2017). In our study, Catechol 1,2-dioxygenase (C12D) genes were enriched after growth with alkali lignin as a sole carbon source and were mostly attributed to Pseudomonas (Figure 2). C12D is the key enzyme in the ortho-cleavage pathway and catalyzes the intradiol cleavage of the aromatic ring generating cis,cis-muconic acid (Fuchs et al., 2011) and it is likely involved in the degradation of kraft lignin by Cupriavidus basilensis B-8 (Shi et al., 2013).

On the other hand, after growth in xylan medium, we observed that the shift in taxonomic composition (Figure 2C) was not followed by a significant change of the functional potential of the microbial assemblages (Figure 2B). This can be explained by different substrate preferences of species with similar functional potential, such as Sphingobacterium and Cellulomonas. Sphingobacterium increased after growth in xylan medium and it has been reported in various plant biomass-degrading microbial consortia and was found to secrete endo-beta-1,4-xylanases during growth on wheat straw (Jiménez et al., 2015). Cellulomonas strains have also been found to secrete xylanases (Sánchez-Herrera et al., 2007; Lisov et al., 2017). However, in the growth conditions used in our study, Cellulomonas seems to be outcompeted by Sphingobacterium when xylan is the sole carbon source. The original consortia were enriched in genes encoding exoglucanases, endoglucanases, and beta-glucosidases. Three of the original consortia were enriched in filter paper, which is composed of crystalline cellulose, and five consortia were bred in poplar wood chips, which is mainly composed of cellulose. Therefore it suggests that the observed association of Cellulomonas with the original consortia is related to their preference to degrade cellulose (Meinke et al., 1993).

Resource partitioning among distinct bacterial phyla has been demonstrated in soil microbial communities and is a major driver of microbial diversity (Wilson and Lindow, 1994; Goldfarb et al., 2011; Kramer et al., 2016). Our findings provides strong evidence that distinct players in lignocellulolytic microbial consortia have distinct substrate preferences. These bacteria are promising candidates for future studies of engineered microbial consortia in lignocellulose degradation. Studies have observed an increase in degradation rates in mixed cultures, when compared to monocultures (Tiunov and Scheu, 2005; Evans et al., 2017). Identifying key degraders of each plant cell wall polymer is the first step toward the development of engineered microbial consortia for an “optimal” plant biomass degradation. Using metagenomics, we confirmed the major role of Pseudomonas during degradation of lignin-derived aromatic compounds. Additionally, Sphingobacterium is identified as a good candidate for xylan degradation and Cellulomonas seems to play a crucial role during aerobic degradation of cellulose in complexes substrates.

CC designed and performed the experiments and analyses and also wrote the manuscript. HF performed the bioinformatics analysis and wrote the manuscript. CC designed and wrote the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.