Determining the composition of a model lignocellulosic conversion residue

Given that high levels of plant material deconstruction and xylose consumption are the major engineering goals for hydrolysate production and primary microbial fermentation, respectively, we aimed to characterize a LCR with minimal plant material and decreased xylose concentration. The chosen LCR was derived from an ammonia fiber expansion (AFEX)- and enzyme-treated switchgrass hydrolysate that was filtered to remove the majority of remaining insoluble plant materials [29]. This filtered hydrolysate was then subjected to fermentation by Zymomonas mobilis and distillation to remove bioethanol. Engineered Z. mobilis 2032 is known to catabolize xylose more efficiently than the similarly engineered and evolved S. cerevisiae Y128 [29, 30]; thus, we anticipated the remaining xylose in this LCR would be significantly reduced in comparison to the yeast-fermented LCR. Furthermore, glucose was not expected to be a major component of this LCR as it was predicted to be almost fully consumed during primary fermentation. Consistent with these hypotheses, Z. mobilis-based fermentation resulted in approximately 90% xylose consumption, almost complete glucose consumption, and a final LCR COD of approximately 70 g COD/L (compared to about 146 g COD/L of the hydrolysate) [29] (Additional file 1: Table S1).

To better understand the composition of this model LCR, a combination of HPLC- and GC–MS-based assays were employed to analyze several batches of this LCR. These analyses identified and quantified the amounts of sugars and other metabolites present that may be used as substrates for further fermentation to additional bioproducts (Fig. 1, Additional file 1: Table S1). This LCR was mostly liquid with about 0.7% solids by weight. The largest component of the COD of LCR was oligomeric sugars (54.2 ± 1.8%). The monomeric sugars were the next largest component at 15% of the total COD and included mostly arabinose (6.8 ± 0.2%) and xylose (5.4 ± 0.8%), with trace amounts of glucose, galactose, mannose, fucose, and rhamnose. Metabolites containing up to four carbons (C1–C4) comprised about 11% of the total COD and included primarily acetate (5.4 ± 0.1%), pyruvate (2.5 ± 0.3%), and several other organic compounds. Acetamide, the main residue produced during the nitrogen-rich AFEX pretreatment [31], was one of the largest single components of the LCR other than sugars (ca. 6%). The alcohols glycerol and xylitol made up a very small percent of the total COD of LCR, 2.6 ± 0.9%. Other compounds present in the LCR could be inhibitory to cell growth, such as monolignols from the plant matter (1.8 ± 0.2%) [11,12,13, 32]. The remaining insoluble fraction of the COD (ca. 5%) is most likely residual cell debris from the hydrolysate-fermenting microorganisms (e.g., Zymomonas). Any remaining material and any leftover COD in the form of amino acids or sugars not accounted for previously were designated as ‘other’ (ca. 0.8%).

Fig. 1
figure 1

Composition of lignocellulosic conversion residue. Lignocellulosic conversion residue (LCR) was generated from “Cave in Rock” switchgrass harvested in 2016, subjected to AFEX and enzyme pretreatments, fermented by Zymomonas mobilis, and finally distilled to remove ethanol. The total potential chemical energy of the remaining LCR was determined by COD analysis while the remaining carbon sources, Zymomonas waste products, and other compounds that might affect future microbial metabolism were quantified via a combination of different separation techniques combined with HPLC and GC–MS analysis then converted to g/L COD to calculate the percent composition of the LCR. The most abundant components of this LCR were oligomeric sugars shown in shades of green (54.2%), monomeric sugars in shades of blue (14.7%), C1–C4 metabolites in shades of purple (10.8%) and acetamide in red (6.4%). These numbers are the average of five batches of LCR

Growth of microorganisms on model lignocellulosic conversion residue

We tested 71 Streptomyces and 163 yeast species aerobically in batch cultures grown at standard growth temperatures until they reached stationary phase to identify strains in these two groups of organisms that grew well in the model LCR. For comparison, activated sludge from a wastewater treatment plant was also fed with the model LCR and samples taken were grown aerobically in batch culture for 7 days at room temperature. Growth on LCR was measured as the average dry cell weight (DCW) of several biological replicates.

Of the 71 Streptomyces strains tested, more than half were able to grow in LCR. Of these, 28 had at least moderate growth (≥ 5 mg/mL DCW) and 22 of those had high growth of ≥ 10 mg/mL DCW (Additional file 2: Table S2). A phylogenetic analysis of a subset of these Streptomyces strains showed that strains capable of growing to high DCW form distinct phylogenetic groupings with the majority of strains capable of growth on LCR falling into 3 clades, suggesting that there might be multiple factors that contribute to the ability of Streptomyces to grow on LCR (Fig. 2A). While each of these three clades contained at least one strain capable of high growth, the majority of high-growth strains occurred in clade III. A sizable fraction of yeast species, 34 out of 163, showed moderate growth on the model LCR (Fig. 2B). Some species, including S. cerevisiae, were likely unable to grow in LCR due to the low concentration of glucose, their preferred carbon source [33], confirming our hypothesis that an appetite for diverse carbon sources is vital to yeast growth in LCR. Other species, such as Lipomyces starkeyi, are known to consume xylose and other diverse substrates [34], so the lack of growth is likely due to factors beyond carbon source availability. Several species in the same family, Lipomycetaceae, grew in LCR which suggests that essential traits are variable among close relatives. The highest growing yeast species fell into two distinct groups; several of these yeasts were in the Dipodascaceae/Trichomonascaceae clade, particularly in the genus Blastobotrys, and the rest were in the CUG-Ser1 clade, particularly in the genus Debaryomyces (major clade terminology based on [35]). Both of these genera contain yeasts with traits of biotechnological interest, including lipid production capacity in some Blastobotrys yeasts, food applications of Debaryomyces species, and tolerance by both groups to an array of stressors [36].

Fig. 2
figure 2

Phylogenetic trees and growth on lignocellulosic conversion residue. Phylogenetic trees of select Streptomyces (A) and yeast species (B) show the diversity within the tested strains. The bar graphs depict growth of these microorganisms in LCR as the average dry cell weight (mg/mL) of at least 2 mLs of culture from at least two biological replicates with the average dry cell weight (mg/mL) of the microbial consortium at the bottom of each panel. Streptomyces strains capable of moderate (≥ 5 mg/mL DCW) and high (≥ 10 mg/mL DCW) growth after seven days at 28 °C with shaking formed distinct phylogenetic groupings indicated as clade 1 (blue), 2 (red), and 3 (yellow) above. Similarly, the highest growing yeast species after four days rolling at room temperature were from two distinct clades: the Dipodascaceae/Trichomonascaceae clade containing the Blastobotrys or the CUG-Ser1 clade containing the Debaryomyces. The number (n) of species in condensed yeast clades is indicated, and the reported values are the mean and standard deviation of values for all species in that clade. Full growth data are available in Additional file 2: Table S2. Clade, species, and strain designations are available in Additional file 5: Table S7

The Streptomyces species capable of high growth had a significantly higher maximum DCW than the tested yeasts, with 8069 B3-B at 81.5 ± 2.1 mg/mL and Blastobotrys capitulata at 19.6 ± 1.6 mg/mL. For comparison, growth yield from the microbial consortium was high, at approximately 14 mg/mL DCW. Although this was lower than several of the Streptomyces strains, the growth yield was comparable to the highest growing Blastobotrys yeasts.

Utilization of lignocellulosic conversion residue components

While COD gives a measurement of potential chemical energy contained in the tested media, we wanted to perform a more detailed analysis on which carbon and energy sources the Streptomyces and selected yeasts were using for growth. All Streptomyces that grew in LCR, the highest growing yeast species of the Blastobotrys and Debaryomyces genera, and some related species were chosen for further characterization. We quantified the amounts of sugar alcohols, glucose, xylose, cellobiose, pyruvate, succinate, lactate, formate, acetate, and ethanol in LCR before and after growth from the previous experiment. The sum of COD concentrations of these components was reported as characterized COD (Fig. 3, Additional file 2: Table S2), and all other LCR components (i.e., oligomeric carbohydrates, monolignols from the plant matter, AFEX pretreatment residues, cell debris, or other metabolic byproducts) were reported as uncharacterized COD (Fig. 3, Additional file 2: Table S2). Individual Streptomyces strains were capable of consuming up to 62.9 ± 1.1% of the characterized COD (i.e., SID14171, 14.6 ± 0.3 g COD/L) and 33.4 ± 10.0% of the uncharacterized COD (i.e., SID8358, 13.9 ± 4.2 g COD/L), and a maximum of 37.7 ± 1.5% of the total soluble COD (i.e., SID809, 25.4 ± 1.0 g COD/L) (Fig. 3, Additional file 2: Table S2). Yeast strains consumed a similar amount of the overall COD; 36.1 ± 1.7% of the total soluble COD (e.g., Blastobotrys raffinosifermentans, 21.8 ± 1.0 g COD/L), but generally consumed a greater portion of the characterized components (e.g., Debaryomyces fabryi, 82.3 ± 0.2%, 8.3 ± 0.0 g COD/L) than Streptomyces (Fig. 3, Additional file 2: Table S2). Most individual microbes consumed a larger percentage of the characterized components as compared to the uncharacterized components. Streptomyces strains SID3915 and SID8358 were exceptions to this pattern, consuming a higher percentage of the uncharacterized COD (19.4 ± 1.7%, 8.8 ± 0.7 g COD/L; 33.4 ± 10.0%, 13.9 ± 4.2 g COD/L, respectively) than the characterized COD (14.2 ± 40%, 3.2 ± 0.9 g COD/L; 27.3 ± 3.7%, 6.1 ± 0.8 g COD/L, respectively), suggesting that these strains had a preference for the components in the uncharacterized portion of the LCR or were perhaps better at accessing those components than the other strains tested (Additional file 2: Table S2). For comparison, the MMC consumed the highest percent of the soluble COD at 65.7 ± 1.9% (40.4 ± 1.2 g COD/L), nearly 90% (ca. 14 g COD/L) of the characterized substrates comprising LCR, and almost 60% (ca. 26 g COD/L) of the uncharacterized components (Fig. 3, Additional file 2: Table S2). This consumption was approximately 25% higher overall than any individual microbe, mostly through consumption of the uncharacterized material. It was surprising that although Streptomyces strains had the highest biomass accumulation as indicated by DCW, they were not the highest consumers of LCR COD. The microbial consortium likely had a higher rate of respiration than any individual species tested which would explain the comparatively low biomass for the amount of COD consumption. Although the MMC consumed a majority of the COD in LCR, a large concentration of soluble organic compounds (both characterized and uncharacterized components) was still present following a 7-day incubation, ca. 20 g COD/L. This suggests that a portion of the LCR COD may be inaccessible to this consortium and may be entirely inaccessible to biofuel- and bioproduct-producing microbes, such as Zymomonas, Streptomyces and yeasts. Reducing the residual fraction of COD is a potential target for further improvements of the upstream processing of biomass prior to primary fermentation.

Fig. 3
figure 3

Utilization of lignocellulosic conversion residue by Streptomyces, yeasts, and mixed microbial consortium. Microbes were incubated in LCR then subjected to COD assays and metabolite analyses via HPLC. Streptomyces are shown in orange, yeasts in purple, and the mixed microbial consortium (MMC) in black on each panel. Percent of soluble COD utilized after incubation of indicated microbes in LCR is calculated relative to a media control. Characterized metabolites include C1–C6 compounds formate, acetate, ethanol, succinate, pyruvate, propionate, lactate, glycerol, xylitol, xylose, and glucose, as well as the glucose dimer cellobiose. The uncharacterized fraction includes all other soluble components such as oligomeric sugars, monolignols from the plant matter, AFEX pretreatment residues, cell debris, or other metabolic byproducts. Values reported are the average of at least two biological replicates with standard deviation denoted by error bars

HPLC analysis of the characterized components of LCR after microbial growth focused on three groups of compounds: C1–C4 metabolites, monomeric sugars, and sugar alcohols (Fig. 4). One limitation of this HPLC analysis is that xylose and galactose, both known components of LCR, eluted from the column at the same time. Since galactose is only present in small amounts in LCR (Additional file 1: Table S1), we will refer to this combined value as xylose going forward.

Fig. 4
figure 4

Lignocellulosic conversion residue metabolite utilization by Streptomyces, yeasts, and mixed microbial consortium. Patterns of indicated characterized metabolites present in LCR after incubation with microbes are shown relative to media controls. Analysis of at least two biological replicates was averaged. Metabolites that were present in lower levels than the media control are shown in red, metabolites that were present in higher levels than the media control are shown in blue, and white indicates no change relative to the media control. This gives a pattern of characterized metabolite consumption (red) and generation (blue) for these LCR degrading microbes

As expected, the levels of many of the assayed compounds were lower after incubation with these microorganisms as they were energy sources for cell growth. However, some of the spent LCR samples tested showed increased levels of pyruvate, succinate, xylitol, xylose, cellobiose, and/or glucose after microbial growth. (Fig. 4) The majority of Streptomyces strains produced high amounts of pyruvate, a known behavior of Streptomyces growing in media with high nitrate concentrations [37]. Three of the yeast species produced noticeable levels of succinate, a known anaerobic byproduct of some yeasts that is hypothesized to be driven by membrane energization, which may suggest that even though the cultures were grown with rolling, oxygen was limiting under these growth conditions [38].

Since xylose, cellobiose, and glucose are rare extracellular products for microorganisms, they were most likely being generated as breakdown products from the solubilized cellulose (cellobiose, glucose) and hemicellulose (xylose) dimers or small polymers remaining from hydrolysis of the original plant material and that were not metabolized during primary fermentation. This hypothesis is consistent with the utilization of uncharacterized COD depicted in Fig. 3. The abundance of these sugars after microbial growth is likely due either to their production at a rate higher than their uptake or whose uptake was prevented by catabolite repression. Furthermore, the patterns of metabolite consumption could indicate a preference for cellulose over hemicellulose by Streptomyces, as many strains show higher consumption of glucose and cellobiose than xylose and xylitol. However, the apparent accumulation of glucose in spent LCR from the MMC and several Streptomyces strains was surprising, as glucose is a preferred carbon source for many microorganisms, and we hypothesized that the consortium would utilize all available sugars. Further examination of the HPLC traces suggest that another byproduct from these samples eluted from the HPLC at approximately the same time as glucose (Additional file 1: Figure S1).

From these growth and LCR consumption assays, we can identify microorganisms that are good candidate chassis for generating valuable bioproducts from LCR. A high overall soluble COD consumption indicates that more energy would be available to be pushed towards the production of these compounds. As microbes are engineered for higher COD consumption, we hypothesize that it will be easier to increase consumption of the characterized LCR components than the uncharacterized. This would make microbes that already have high consumption of the uncharacterized portion of the LCR more desirable production chassis. For the Streptomyces, SID14171 and SID809 perform well in overall COD consumption, but SID8358 is perhaps a more attractive target due to its superior performance on the uncharacterized portion of the LCR. For the yeast species, performance on the characterized portion of COD was strong for all species analyzed. Generally, performance was also similar across species on the uncharacterized portion, but several species including Sugiyamaella smithiae, Wickerhamiella dulciola, and Blastobotrys arbuscula proved weaker in this regard. Blastobotrys raffinosifermentans has recently been highlighted due to its potential for lipid production and may prove to be a strong chassis for bioproduct formation from LCR [39].

Development of synthetic conversion residue

Given the complexity and limited availability of LCR, we wanted to develop a defined media that mimicked LCR for future experiments. This synthetic conversion residue (SynCR) would allow us to examine how individual components affect the growth of our microorganisms and if any uncharacterized components have a relevant contribution to growth phenotypes. To design SynCR, we used the analysis of the LCR as a starting point. We included the sugar alcohols, C1–C4 metabolites, cellular waste products, and the most abundant monomeric sugars. Since it was not possible to determine the chain length of the oligomeric sugars and adding variable chain-length purified hemicellulose or cellulose consistent with our observed compositions was not logistically feasible, we added the most abundant hemicellulose components as monomers and used Sigmacell50 to represent the cellulose. Cysteine, methionine, and tryptophan were not detected in our analyses, but they were added to the SynCR at 150 μM to ensure growth and optimize future production of bioproducts [40, 41]. The remaining amino acids were added at the concentrations measured in LCR. The most abundant minerals and metals (≥ 3 mg/mL), as well as any required for growth by either yeasts or Streptomyces, were also included in the final SynCR. Acetamide was included in SynCR as it is a significant component of LCR. After the addition of all these components except for the Sigmacell50, the SynCR was filtered and adjusted to pH 6.5, the same pH as LCR for microbial growth. The Sigmacell50 was sterilized by autoclaving and added to the SynCR after final filtration (recipe in Additional file 1: Table S3).

To begin investigating how microbes utilize the wide variety of carbon sources in the LCR, we used SynCR to examine the metabolite consumption patterns of a subset of Streptomyces and yeasts that had a range of growth and metabolite consumption patterns on LCR. Since SynCR is a minimal version of LCR with approximately the same calculated COD (ca. 65 g/L) and the hemicellulose components converted to monomers, we hypothesized that we would see a greater percentage of the SynCR utilized by our microorganisms than the LCR. That was true for most of the Streptomyces strains tested, where the strains utilized as much or more of the SynCR than they did the characterized components of the LCR (Fig. 5, Additional file 3: Table S5). However, 8069 B3-B had a lower overall percent of SynCR utilized as compared to LCR, suggesting that breakdown and utilization of the uncharacterized components of LCR are an important factor in metabolite utilization by this strain. The most striking difference in the metabolites characterized in SynCR after the growth of the indicated Streptomyces strains was a relative increase in the “other” component as compared to the uninoculated SynCR (Fig. 5). Streptomyces are known to produce many specialized metabolites [42] and may be converting some of the SynCR carbon into these metabolites, which would account for the increase in the “other” COD component. It also makes it unclear how much of the uncharacterized portion of LCR was new compounds produced by the microbes assayed or material initially in the LCR that was not broken down or consumed during incubation. Another notable difference in SynCR utilization as compared to LCR utilization (Fig. 3b) was the lack of cellobiose accumulation in all strains except SID8358. Since that metabolite is present due to cellulose breakdown in LCR, this result suggests that the insoluble Sigmacell50 was not as accessible to most of the Streptomyces tested as the soluble switchgrass-derived lignocellulose polymers in LCR. Further, the ability of SID8358 to degrade both the uncharacterized, oligomeric-containing portion of LCR and Sigmacell50 indicate that it could serve as a potential source for future mining of cellulose-degrading enzymes.

Fig. 5
figure 5

Utilization of synthetic conversion residue by Streptomyces and mixed microbial consortium. Microbes were incubated aerobically in synthetic conversion residue (SynCR) with crystalline cellulose and with or without lignocellulose-derived inhibitors (LDIs) for seven days then subjected to COD assays and metabolite analyses via HPLC. The bars labeled SynCR show the metabolite levels in the uninoculated media controls while the remaining bars indicate the amounts of the those metabolites present in spent SynCR after 7 days of incubation with either the mixed microbial consortium (MMC) or the indicated Streptomyces strains. Values reported are the average amounts of metabolites remaining after incubation from 3 biological replicates

Interestingly, only one of the seven tested yeast species was able to grow in the SynCR (Additional file 1: Table S4). Since these yeasts grew in LCR, this suggested that a component of the LCR essential for yeast growth was not included in the SynCR recipe. Standard yeast synthetic media [43] contains the vitamins biotin, inositol, and the B vitamins pyridoxine and niacin. Vitamin supplementation restored growth of yeasts in SynCR (Additional file 1: Table S4) and LC–MS/MS analysis confirmed that those vitamins were present in LCR at concentrations greater than those required for growth of the tested yeast species. In the future, any LCR generated will need to be evaluated for vitamins to assess viability of yeast growth.

The MMC was also grown on SynCR to show the maximum amount of SynCR available for microbial degradation. The MMC was able to metabolize almost the entirety of the soluble SynCR (95.2 ± 1.8%, 34.1 ± 1.6 g COD/L), which was not surprising as the consortium was also able to metabolize such a large portion of the characterized components of the LCR. The composition of future SynCR recipes could also be adjusted to focus on separate carbohydrate components, e.g., arabinose versus xylose, in order to identify or engineer microorganisms better able to metabolize these components.

One of the uses for SynCR is to test how individual components affect the growth of our microorganisms. Since our microbes were able to grow in LCR, they obviously are tolerant of LDIs which had been reported in the literature to be inhibitory to some microbes [32]. We wanted to determine if LDIs had any negative impact on the growth and LCR utilization of our microbes that might be reduced by future engineering efforts. Growth experiments using SynCR both with and without LDIs allowed us to test the effect these compounds had on our organisms. Interestingly, LDIs appear to have minimal effect on the Streptomyces strains tested. Only two of the Streptomyces strains tested, SID10536 and SID809, showed differential behavior in percent SynCR consumed in the presence or absence of LDIs. The higher percent of metabolites consumed by SID10536 when grown in SynCR without LDIs suggests that LDIs may be inhibitory or inducing other pathways that shift metabolism away from consumption, while conversely the LDIs may be inducing consumption in SID809. Due to the difficulty in getting yeasts to grow on SynCR, they were not evaluated for inhibition by LDIs.

Comparisons between the most abundant OTUs in the SynCR experiments with and without LDIs showed only minimal changes in microbial abundances. This is reflected in non-metric multidimensional scaling (NMDS)-space, where the LCR consortium plots more distantly from the two SynCR consortia (Additional file 1: Fig. S2). Only one OTU, Corynebacterium, had a higher relative abundance in both the LDI-containing LCR and SynCR with LDIs experiment, consistent with the observation that some Corynebacterium species have shown tolerance to LDIs [44]. The similarity of the microbial consortia grown on SynCR with and without LDIs suggests these potential toxins have a minimal effect on the microbial consortium structure. Furthermore, the similar abundance of microbial consortium members in both SynCR incubations (Fig. 5) and the data from the Streptomyces species tested both suggest that the LDIs were not universally an impediment to microbial growth. This observation is counter to previous studies demonstrating the inhibitory effect of these compounds on fermentative organisms [11,12,13].

Consortium differences in model LCR compared to SynCR

The mixed microbial consortium serves as both a representative of the maximum LCR utilization possible and as a potential source of genetic material to engineer microbes for increased LCR utilization. The SynCR can be used to assist in identification of OTUs that are responsible for utilization of specific components of the LCR. For example, to identify which OTUs in the MMC contribute to mannose utilization, we can monitor the change in consortium composition when the MMC is grown in SynCR with mannose as compared to SynCR without mannose. Similarly, as SynCR consists mostly of the characterized components of LCR, comparisons between consortium composition when grown on LCR as compared to SynCR will allow us to examine which OTUs contribute to the utilization of the uncharacterized portion of the LCR. Since we predict that it will be more difficult to engineer strains to utilize this portion of the LCR, the MMC can thereby serve as a uniquely valuable resource for that genetic material.

For these experiments, genomic DNA extracted from the initial inoculum and biomass pellets (n = 4) from the microbial consortia grown as indicated was subjected to 16S rRNA gene amplicon sequencing analysis. A total of 2931 unique operational taxonomic units (OTUs) were identified across all samples with the OTUs present in the initial wastewater consortium included for comparison (Additional file 4: Table S6). Subsequent analyses focused on the most abundant taxonomic groups which had a relative abundance of 1% or greater at the genus level. This procedure resulted in 24 highly abundant taxonomic groups across all samples representing approximately 90% of the total DNA reads (Fig. 6). The number of distinct highly abundant taxonomic groups in each sample tested were similar (inoculum, 17; LCR, 11; SynCR with LDIs, 15; SynCR without LDIs, 14), with a slightly less diverse microbial consortium present after seven days of growth in LCR than in SynCR, which is most likely due to less easily available carbon for microbial metabolism.

Fig. 6
figure 6

source and following a 7-day incubation period in LCR or SynCR with crystalline cellulose and with and without LDIs. Individual OTUs were clustered to the highest taxonomic level (c, class; o, order; g, genus), with clusters greater than 1% total relative abundance shown above, organized by phylum (Pa., Patescibacteria; Sa., Saccharibacteria; Bacter., Bacteroidota; Actino., Actinobacteriota). Taxa with distinct differences in abundance between the microbial consortia grown on different types of CR are indicated in bold

Distribution of bacterial taxa in the mixed microbial consortium on different growth media. Bacterial taxa were identified within the initial inoculum

The microbial consortium analysis revealed several key differences in consortia composition when grown on LCR as compared to SynCR. The most profound difference in microbial consortium composition between the LCR and SynCR experiments was the variable abundance of Enterococcus and Enterobacter. In the LCR-fed cultures, Enterococcus represented approximately a quarter of the 16S rRNA sequence reads, but they were present at about 5% relative abundance in the inoculum and in both SynCR experiments. Conversely, Enterobacter represented 15–24% of the microbial population in the inoculum and SynCR experiments but less than 4% of the LCR consortium (Fig. 6). Enterobacter are endophytes, which have been associated with numerous plant species, including switchgrass [45]. These taxa have also been shown to metabolize cellulose and components of hemicellulose, such as xylose [46,47,48]. Species of Enterococcus are facultative anaerobes that are both typical commensal members of the human gut microbiome and potentially pathogenic, so their presence in these experiments is unsurprising due to the origin of the inoculum. Cellulolytic activity, as well as metabolism of other carbohydrate polymers, such as pectin, have also been reported in Enterococcus species [49,50,51,52], which is consistent with a typical greater abundance of these taxa in the gut microbiomes of vegetarians [53] and in fermented plant products [54]. The prevalence of lignocellulose-degrading enzymes and cellulolytic activity in Enterobacter and Enterococcus indicates a likely role that these taxa play in the LCR and SynCR consortia; however, the abundance of free monomer carbohydrates in SynCR likely allowed the Enterobacter to outcompete the Enterococcus in these experiments. Because of the apparent opportunistic behavior of the Enterobacter, we anticipate genomic studies of the Enterococcus members of the consortium may be a more promising resource for microbial engineering.

The other relevant difference between the LCR and SynCR microbial consortia was that Dysgonomonas and Paucilactobacillus were present in both consortia, but at a significantly increased relative abundance in LCR (ca. 11% and ca. 6%, respectively, Fig. 6) as compared to SynCR. Dysgonomonas are gut symbionts of termites and wood-boring insects, and recent genomic studies of the species have revealed an abundance of glycoside hydrolase enzymes, which suggests a role in lignocellulose degradation [55,56,57]. Similar to Enterococcus, abundance of Dysgonomonas in the LCR consortium is likely a result of their putative role in the degradation of oligomeric carbohydrates, which comprise half of the potential carbon energy in LCR, and is consistent with the high observed consumption of the uncharacterized portion of COD (Fig. 3). In contrast, Paucilactobacillus is a heterofermentative lactic acid bacteria notable for the uncommon ability to metabolize pentoses [58], so would be expected to be more abundant in SynCR experiments due to the greater concentration of monomer pentoses in the SynCR as compared to LCR. However, most of the characterized species of this genus were isolated from fermented plant material and some strains have been shown to metabolize disaccharides, such as melibiose by Paucilactobacillus hokkaidonensis [58], and may therefore be able to utilize some of the less commonly metabolized plant breakdown sugars. Depending on the carbohydrate composition of future LCRs, Dysgonomonas or Paucilactobacillus related members of the consortium may be a beneficial bioengineering resource for genes related to oligomeric or less commonly metabolized sugar utilization.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.


This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (

By admin