Microbiome-based conversion of glucose to butyrate and hexanoate: roles and limitations revealed through genome-centered metatranscriptomics

doi:10.21203/rs.3.rs-4881179/v1

Download PDF

Research Article

Microbiome-based conversion of glucose to butyrate and hexanoate: roles and limitations revealed through genome-centered metatranscriptomics

https://doi.org/10.21203/rs.3.rs-4881179/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Chain elongation is emerging as a biotechnological tool to convert waste organics to beneficial fermentation products with wide utility across industrial and agricultural sectors. Glucose, the most abundant sugar on earth, is present in organic feedstocks as both a monomer and as a constituent of complex polymers. In this work, glucose was used as the sole carbon and electron source to enrich a chain elongating microbial community using duplicate bioreactor systems and batch experiments were performed to assess the role of hydrogen supplementation on fermentation.

Results

Hexanoic acid was the most abundant MCCA produced in both bioreactors with steady-state concentrations of about 4 g COD L^− 1 while butyric acid was the most abundant carboxylic acid with concentrations of about 5 g COD L^− 1. Metagenomic and metatranscriptomic analyses showed that the microbial communities in the duplicate reactors were very similar, with Caproicibacter sp. 002316805 and Caproicibacterium sp. 002399445 emerging as the most abundant chain elongating bacteria in both bioreactors. The microbiomes also contained a high abundance of lactic acid bacteria, primarily the Coriobacteriaceae Tractidigestivibacter and Olegusella. H₂ supplementation with batch experiments resulted in marginal changes in the products compared to controls with N₂ supplementation, with butyric acid production being higher than but not hexanoate production. One low-abundance Clostridium_B increased expression of genes for the Wood Ljungdahl pathway during H₂ supplementation.

Conclusions

The duplicate bioreactors behaved similarly, with the same abundant species. Combined, Caproicibacter sp. 002316805, Caproicibacterium sp. 002399445, and Tractidigestivibacter sp902834555 accounted for > 65% of the transcriptome and analysis of transcribed genes suggests butyrate and hexanoate originate from both direct sugar conversion by Caproicibacter sp. 002316805 and Caproicibacterium sp. 002399445 as well as initial conversion to lactate by Tractidigestivibacter sp902834555. Further, it was found that Caproicibacter and Caproicibacterium differ in synteny of chain elongation genes. H₂ supplementation during batch experiments did not reliably increase hexanoate production.

Caproicbacterium

Caproicbacter

Tractidigestivibacter

Chain elongation

Butyrate

Hexanoate

Cellulose is the most abundant organic compound in nature and is primarily produced by plants consuming CO₂ while harnessing solar energy [1]. Cellulase enzymes hydrolyze cellulose to produce glucose, a fermentable carbohydrate that can be used as a feedstock for downstream anaerobic processes [2]. Despite the onset of the negative impacts of climate change, fossil fuel demands continue to increase. At the same time, societies are generating increasing quantities of organic wastes, and cellulose is often a major component of these waste streams [3]. Therefore, even if plants are not grown directly for chemical and energy production, the sugars they generate are abundant “wastes”.

Chain elongation is a fermentation process where pure- or open-cultures produce medium chain carboxylic acids (MCCAs) from a variety of feedstocks [4, 5]. MCCAs are monocarboxylic acids consisting of six to twelve carbon atoms and have gained great attention as a precursor for biofuels, biochemicals, dietary supplements, and pharmaceuticals [4, 6, 7]. So called “chain elongators” rely on reverse β-oxidation to recycle intracellular reducing equivalents and conserve energy as part of their central carbon metabolism. Acetyltransferase (ACAT) is the enzyme that initiates the process, followed by a hydroxyacyl-CoA dehydrogenase (HAD), an enoyl-CoA dehydratase (ECH), and acyl-CoA dehydrogenase (ACD). The ACD enzyme is part of a complex with two electron transfer flavoproteins (EtfA and EtfB) which couples the reduction of an enoyl-CoA (e.g., crotonyl-CoA) to an acyl-CoA (e.g., butyryl-CoA) with ferredoxin reduction [8, 9]. The final product of reverse β-oxidation is a carboxylic acid “elongated” by two atoms of carbon saturated with hydrogen atoms. Chain elongating bacteria rely on a sodium translocating ferredoxin—NAD⁺ oxidoreductase (Rnf) complex, a protein complex that couples ferredoxin oxidation with generation of an ion motive force for energy conservation [10].

The most well-studied chain elongation processes rely on lactate or ethanol as exogenous electron donors, but sugar-based chain elongation has received increasing attention, mostly with newly isolated strains of hexanoate-producing bacteria [11–13]. (Meta)transcriptomic analyses have been performed with Caproicibacterium lactatifermentans [14] and microbial communities producing MCCAs from xylose-rich feedstocks [15] that suggest multiple routes of sugar degradation, including the Embden-Meyerhoff (EM) pathway (aka, glycolysis) and the Bifid shunt pathway. Still, the conversion of glucose as sole substrate to MCCAs by a self-assembled microbial community has not been thoroughly investigated.

During chain elongation, excess electrons that are not directed to reverse β-oxidation are used to generate H₂. The initial sugar degradation pathway can impact the amount of H₂ generated; when acetyl-CoA is produced via the Bifid shunt, it is decoupled from H₂ generation [16]. H₂ has also been suggested as an electron donor for chain elongation, and previous thermodynamic analyses suggest that its accumulation is unlikely to create thermodynamic bottlenecks during sugar-based chain elongation [15]. Past studies on H₂ supplementation have demonstrated increased butyrate production by microbial communities co-consuming H₂, xylan, and complex feedstocks (e.g., corn silage and manure) [17], but the impacts of H₂ supplementation on glucose fermentation with microbial communities has not been studied previously. Further, because H₂ can be produced through renewable resources, it can potentially offset the negative environmental impacts of supplementing electron donors in chain elongation systems [18].

To further understand chain elongation from glucose as the sole carbonaceous substrate and the impacts of H₂ supplementation, duplicate bioreactors were seeded with municipal wastewater sludge and operated for 184 days under conditions known to favor chain elongation. 16S rRNA gene amplicon sequencing and shotgun metagenomics were used to identify abundant microbial taxa and to assess their genomic potential. Batch experiments were then used to assess the impacts of H₂ supplementation on fermentation products, and metatranscriptomic analyses were used to ascertain the activity of individual species within the bioreactors under differing headspace conditions. In total, the work herein demonstrates that a relatively simple microbial community emerged from a complex inocula that converted glucose to butyrate and hexanoate, but H₂ supplementation was not successful for improving chain elongation from glucose. Further, this work provides fundamental insights on the role of sugar fermenters and genomic differences between two recently classified genera of chain elongators – Caproicibacterium and Caproicibacter.

Bioreactor configuration and operation

Two bioreactors from Solaris Biotech Solutions, named Jupiter 1 and Jupiter 2 (referred to as J1 and J2), were operated. Each bioreactor consisted of a reactor vessel supported by a control cabinet with four peristaltic pumps and a human-machine interface (HMI) for the automated programming and operation of both J1 and J2. Both bioreactors were furnished with a water jacket, pH probe, impeller and overhead mixer, and two thermometers for measuring internal vessel temperature and water jacket temperature. Additionally, the bioreactors were equipped with a gas condenser, a pressure release valve, a gas sparger, and four ports for influent supply, as well as ports for buffer and acid to control pH. Both bioreactors had a total volume of 2 L, with a working volume maintained at 1 L. The hydraulic retention time (HRT) was controlled at 6 days and the water jacket maintained a constant temperature of 37°C. The pH control system maintained a pH of 5.5 by introducing 5M NaOH or 1M HCl. During the initial week of operation, the stirring speed was established at 100 rpm then reduced to 50 rpm. The system operated at atmospheric pressure throughout the study.

Batch Experiments

On Day 155, aliquots from the J1 bioreactor were used to seed small batch experiments in which multiple head space conditions were tested to assess impacts on MCCA production and microbial community gene expression. Biomass from the bioreactor was transferred to 125mL serum bottles and synthetic medium containing 4 g COD L^− 1 of glucose was added to each bottle at a ratio of 1:1, with a total working volume of 50 ml. After this, the headspace in the serum bottle was adjusted by sparging the bottles continuously with either N₂ or H₂. Biological triplicates were used for each condition. The bottles were incubated in an incubator shaker for 84 hours with a mixing speed of 100 rpm and temperature of 37°C. Samples were collected for metatranscriptomic analyses after 84 hours.

Inoculum and media preparation

Both bioreactors were seeded with sludge from a mesophilic anaerobic digester at the South Burlington Water Quality Facility in Burlington, Vermont, United States. The seed had a soluble chemical oxygen demand (COD) concentration of 1.67 g COD L^− 1. The Volatile Suspended Solids (VSS) to Total Suspended Solids (TSS) ratio was 0.63, with VSS and TSS concentrations of 16.77 g L^− 1 and 26.55 g L^− 1, respectively. Both bioreactors were supplied with a synthetic medium with glucose was the sole organic carbon, with 40 g COD L^− 1 glucose. The synthetic medium also contained 2 g L^− 1 ammonium chloride, 2 g L^− 1 dipotassium hydrogen phosphate, 0.4 g L^− 1 sodium phosphate monobasic, 0.1 g L^− 1 magnesium sulfate heptahydrate, and 0.05 g L^− 1 calcium chloride. After preparing the primary media, 1 mL of trace elements solution was added containing 6.2 g L^− 1 EDTA, 1 g L^− 1 zinc sulfate heptahydrate, 8.3 g L^− 1 iron (III) chloride hexahydrate, 2 g L^− 1 cobalt chloride hexahydrate, 0.1 g L^− 1 copper(II) chloride monohydrate, 0.2 g L^− 1 nickel (II) chloride hexahydrate, 1.4 g L^− 1 manganese (II) chloride tetrahydrate, 3 g L^− 1 boric acid, 1.2 g L^− 1 sodium molybdate dihydrate, and 0.1 g L^− 1 sodium selenite. Synthetic media was autoclaved after preparation and stored at 4°C until fed to the bioreactors.

Chemical analyses

Every six days, fermentation broth was collected for analysis of glucose, lactate, and volatile fatty acids (VFAs). The separation of liquid from solids was achieved through centrifugation at 9,600 rcf with the ThermoScientific Sorvall Legend Micro 31 centrifuge. Subsequently, the supernatant underwent filtration using sterile, nylon membrane, 0.45µm pore size syringe filters. Glucose and lactate were measured with a YSI 2950D Biochemistry Analyzer. For the quantification of VFAs, a Shimadzu TQ8040NX gas chromatograph–mass spectrometer (GC–MS) was used. Concentrations of acetic acid (C2), propionic acid (C3), butyric acid (C4), isobutyric acid (iC4), valeric acid (C5), isovaleric acid (iC5), hexanoic acid (C6), isohexanoic acid (iC6), heptanoic acid (C7), and octanoic acid (C8) were obtained. The GC-MS used a 2.5 ml smart syringe at 150°C to collect 1 ml of headspace after incubating the liquid samples at 95°C for 40 with an agitator speed of 250 rpm. A DB-FATWAX UI column, with 30 m length, 0.25 µm thickness and 0.23 mm diameter was used. Upon injection of the sample, the oven temperature was held at 80°C for 1 minute and increased by 20°C min^− 1 until it reached 240°C for 1.5 minutes. MS acquisition mode was Q3 scan, with ion source temperature and interface temperature at 280°C and 250°C, respectively.

DNA and RNA Extraction and Sequencing

Samples were collected throughout the experiment for DNA and RNA extraction. Samples were collected from each bioreactor, centrifuged for one minute at 9,600 rcf, and the supernatant was discarded. Samples for DNA extraction were stored at -80°C while samples for RNA extraction were flash-frozen in liquid nitrogen and then stored at -80°C. A DNeasy PowerSoil Pro Kit and RNeasy Plus Mini Kit from Qiagen were used to extract DNA and RNA, respectively. The concentration and the quality of the samples were analyzed with the Qubit Fluorometer and NanoDrop Spectrophotometer (ThermoFischer Scientific).

For a subset of DNA samples, full metagenomic analyses were performed to further assess the taxonomy and metabolic potential of organisms present within the bioreactors. Sequencing was performed using the Illumina HiSeq platform to generate 2 x 150 bp reads as well as the Oxford Nanopore Technologies MinIon system with 9.4.1 flow cells to generate long reads. Samples for short read sequencing were prepared using a Nextera XT DNA library preparation kit (Illumina FC-131-1096). Samples for long read sequencing were prepared using a ligation sequencing kit with the native barcoding expansion (Oxford Nanopore Technologies SQK-LSK109 and EXP-NBD196). After RNA extraction, these samples underwent reduction of ribosomal RNA with a QIAseq FastSelect Bacterial RNA (5S, 16S, 23S) kit and were prepared for sequencing. Samples were sequenced using the Singular Genomics G4 platform with an F3 flow cell to generate 2 x 150 bp reads. Raw read counts are summarized in Supplement 1.

Bioinformatic Analyses

Multiple bioinformatic tools were used to examine the datasets on the high-performance computing clusters of the Vermont Advanced Computing Center (VACC) at the University of Vermont. The 16S rRNA gene amplicon sequencing data were analyzed using Qiime2-2023.2 [19] with trimming and denoising with dada2 [20] and taxonomy assignment based on the Silva taxonomy database [21] (release 138.1). Raw metagenomic sequences underwent adapter trimming through TrimGalore [22] (version 0.6.10) for Illumina reads and Porechop [23] (version 0.2.4) for Nanopore reads. Canu [24] (version 2.2) and metaSPAdes [25] (version 3.15.5) assembly pipelines were implemented to construct contiguous strands of DNA based on overlapping regions of Nanopore and Illumina reads, respectively. The Burrows-Wheeler Alignment [26] (BWA, version 0.7.17-r1188) package was employed to align the reads back to the assembly for both the Canu and metaSPAdes assembly methodologies. In order to refine the metagenome assemblies generated using Nanopore technology, Pilon [27] was used to automate the enhancement of the genome assemblies, which were subsequently aligned back to the Canu reads employing BWA. For binning to construct metagenome-assembled genomes (MAGs), MaxBin [28] (version 2.2.7) and MetaBat [29] (version 2.15) were employed. In dereplicated and select the best genomes for an individual organism dRep [30] (version 3.4.3) was utilized, while CheckM [31] (version 1.2.2) was employed to evaluate the quality of the MAGs. RAST [32] was used to annotate the genomes, while GTDB-Tk [33] (version 2.3.2) was employed to assign taxonomy to MAGs. The alignment file generated after executing GTDB-tk, which contained the concatenated amino acid sequences for the analyzed MAGs, was used as input for RAxML [34] (version 8.2.12) to generate a phylogenetic trees. For the RNA data, TrimGalore [22] (version 0.6.10) was used to trim sequencing adapters. Flash [35] (version 1.2.11) was used to merge the paired end reads and SortMeRNA [36] (version 4.3.6) was used to remove rRNA sequences. BWA (version 0.7.17-r1188) was used to map the RNA reads to the MAGs, and HTSeq [37] (version 2.0.3) was used to summarize read counts by genomic feature. Transcript abundance is reported based on reads per kilobase per million reads mapped (RPKM) which normalizes for sequencing depth and length of open reading frames [38]. Information on assembled MAGs and transcript mapping is provided in Supplement 1.

Glucose is converted to butyrate and hexanoate

The entire input of reducing equivalents (reported as chemical oxygen demand, COD) for the bioreactor experiments was derived from glucose which was consumed for the duration of the experiments (Fig. 1A). Lactate is known to be an electron donor and sole-substrate for chain elongation and is likely a key intermediate of the chain elongation process from sugars [15]. During the first three weeks, there was no lactate accumulation in either bioreactor (Fig. 1B). Lactate accumulated transiently with peak concentrations in both bioreactors of 5 g COD L^− 1. Transient accumulation of lactate in chain elongation bioreactors fed sugars has been noted previously [39] and indicates that rates of fermentation of sugars to lactate are greater than conversion of lactate to VFAs.

Acetate was present in both systems after the first week of bioreactor operations and exhibited wide variations in concentration (Fig. 1C). After an initial peak, acetate concentrations in both bioreactors were typically between 0.5 g COD L^− 1 and 2.0 g COD L^− 1. This is a relatively low acetic acid concentration for chain elongating bioreactors, representing only 1.3–5% of the COD fed to the bioreactors. Butyrate was consistently the most abundant carboxylic acid in both J1 and J2 (Fig. 1D). During a period from Day 50 to Day 100, butyrate concentrations were lower in both reactors and very close to hexanoic acid, near 5 g COD L^− 1. Hexanoate, the product of the butyrate elongation, had the highest peak in both bioreactors, with maximum concentration of 8.3 g COD L^− 1 for J1 and 7.3 g COD L^− 1 for J2 (Fig. 1E). However, hexanoate and butyrate production were not stable over the duration of the study, with high fluctuations across samples. Further, octanoate (Fig. 1F) was typically present at very low concentrations (less than 0.1 g COD L^− 1), with the highest peak in both J1 and J2 occurring at Day 68 of 0.42 g COD L^− 1 (approximately 1% of the COD fed to the bioreactors). Together, these results show that butyrate and hexanoate are abundant products of glucose-based chain elongation, but product profiles are variable across time and butyrate is produced at higher concentrations than the preferred hexanoate.

A simple microbial community of Actinobacteria and Firmicutes was enriched

To further understand the observed bioreactor behavior, microbial communities were assessed initially using 16S rRNA gene amplicon sequencing. The microbial communities enriched in both bioreactors were relatively simple, with six genera making up > 95% of the microbial community after the first six days of operation (Fig. 2). Specifically, Caproiciproducens, which belongs to the Firmicutes/ Bacillota phylum, was identified as the most abundant genus in both J1 and J2. Another member of the Firmicutes/ Bacillota Clostridium sensu stricto 12 was the second most abundant bacteria in both bioreactors. An Unclassified Ruminococcaceae was highly abundant initially, but abundance decreased over time in both bioreactors. All three of the abundant genera of Firmicutes/ Bacillota belong to taxa known to perform chain elongation [40, 41]. In addition, Actinomycetota/ Actinobacteria Olsenella and Atopobium showed very similar abundance patterns in both bioreactors, with Atopobium decreasing in abundance throughout operations and Olsenella emerging as a dominant genera by the end of the experiments. The co-occurrence of Actinomycetota / Actinobacteria and chain-elongating bacteria has been documented previously [15, 42].

Genome-scale taxonomic classification reveals mostly uncultured species

Following 16S rRNA gene amplicon sequencing, shotgun metagenomics was performed to create MAGs. Taxonomy and abundance of MAGs (Fig. 3) align with results of 16S rRNA gene amplicon sequencing (Fig. 2). There were shifts in abundant microbial communities across the three samples from each bioreactor and bioreactors behaved similarly. In both bioreactors, Olegusella sp. 002407925 and Tractidigestivibacter sp. 902834555 showed marked changes in relative abundance, with Tractidigestivibacter sp. 902834555 becoming the most abundant species in both reactors while the relative abundance of Olegusella sp. 002407925 decreased. Both species are members of the Coriobacteriales phylum within the Actinomycetota (previously Actinobacteria) phylum, and related organisms are commonly found during microbiome-based chain elongation [43].

Caproicibacter sp. 002316805 and Caproicibacterium sp. 002399445 were abundant in the duplicate bioreactors. Both belong to newly described chain-elongating genera named in the past four years [13, 44]. The Caproicibacter sp. 002316805 MAG accounted for approximately 26% of the mapped DNA reads and the Caproicibacterium sp. 002399445 for approximately 24% of the mapped DNA reads. Both MAGs cluster within the Acutalibacteraceae family, in the order Oscillospirales, Bacillota_A phylum. Based on the GTDB, Caproicibacter sp. 002316805 contains three MAGs that have been obtained from environmental samples as part of a large-scale analysis of publicly available metagenomic data [45, 46]. The type organism is referenced as Ruminococcaceae bacterium UBA1036 in the NCBI taxonomy database (taxonomy ID: 1952270) [47]. Caproicibacterium sp. 002399445 contains eight MAGs that have been obtained from the same sources as Caproicibacter sp. 002316805 [45]. The NCBI taxonomy database classifies this organism as Ruminococcaceae bacterium UBA4871 (taxonomy ID: 1952342) [47]. The origin of the previously existing MAGs for these microorganisms is uncertain given the wide variety of sample types used to assemble the associated MAGs, but this is the first time they have been described in chain elongating bioreactors.

Hydrogen supplementation does not increase hexanoate yield

Hydrogen has the potential to be an electron donor for sugar elongation, and the overall thermodynamics of H₂ and glucose co-utilization are favorable (Table 1). Therefore, batch experiments were used to test if H₂ supplementation could increase MCCA production during glucose consumption. Batch experiments were performed with biomass obtained from J1 on Day 155 and incubated with a complete head space of N₂ or H₂ at atmospheric pressure. Prior to glucose addition, there was a COD concentration of 5.3 g COD L^− 1 and after glucose addition the COD increased to 7.5 g COD L^− 1.

Table 1

Proposed chemcial reactions and free energy released under standard conditions at neautral pH (ΔG⁰’) for H₂ and glucose co-consumption to produce hexanoate where glucose is first converted to butyrate (Rxn 1) followed by butyrate conversion to hexanoate using CO₂ and H₂ (Rxn 2). The combined reaction is also shown (Rxn 3). CO2 is balanced when Rxn 1 and Rxn 2 are combined, but a net input of four moles of H₂ per mol glucose is required.
ID	Reaction	\(\:{\varvec{\varDelta\:}\varvec{G}}^{0\varvec{{\prime\:}}}\) (kJ rxn^− 1)
Rxn 1	\(\:{C}_{6}{H}_{12}{O}_{6}\to\:{C}_{4}{H}_{7}{O}_{2}^{-}+2{CO}_{2}+2{H}_{2}+{H}^{+}\)	-248
Rxn 2	\(\:{C}_{4}{H}_{7}{O}_{2}^{-}+2{CO}_{2}+6{H}_{2}\to\:{C}_{6}{H}_{11}{O}_{2}^{-}+4{H}_{2}O\)	-159
Rxn 3	\(\:{C}_{6}{H}_{12}{O}_{6}+4{H}_{2}\to\:{C}_{6}{H}_{11}{O}_{2}^{-}+4{H}_{2}O\)	-407

Approximately twelve hours after inoculating the serum bottles, glucose consumption commenced with H₂-supplemented bottles experiencing a higher rate of glucose consumption (Fig. 4A). Lactate accumulated throughout the batch experiments under both conditions (Fig. 4B). While standard deviations across the three replicates were low for glucose consumption and lactate production, the carboxylic acid products had much higher standard deviations (Fig. 4C-F). Butyrate and hexanoate both increased initially (Fig. 4D-E), but then plateaued. Hexanoate increased by more than a factor of two during the batch experiments regardless of headspace condition. Octanoate concentrations remained low (Fig. 4F).

After 84 hours of incubation, the COD of the N₂ headspace incubations had a COD of 6.4 +/- 0.24 g COD L^− 1 which is a decrease of 1.1 g COD L^− 1 (Fig. 5). The H₂ headspace incubations had a COD of 7.3 +/- 1.22 g COD L^− 1 which is a decrease of 0.2 g COD L^− 1. Taken as a whole, while H₂ supplementation resulted in some replicates with higher COD and intermittent concentrations of desirable products (e.g., hexanoate), the differences were not sustained and the final COD of the target products in the H₂ supplemented incubations was not significantly different than those in the N₂ supplemented incubation (p = 0.256). The only fermentation product that exhibited a significantly higher concentration in H₂ incubations was butyrate (p = 0.012). This suggests that the microbial community was unable to reliably utilize H₂ as an electron donor to drive conversion of butyrate to hexanoate.

Metatranscriptome dominated by two chain elongators and one hydrogenic sugar fermenter

To gain greater insights into the activity of individual microbial species during the incubations, metatranscriptomic analyses were performed with samples from the N₂ and H₂ incubations after 84 hours. At this time, glucose levels had decreased substantially (Fig. 1A) but was not depleted. A mixture of VFAs were present, with butyrate and hexanoate being the most abundant fermentation products (Fig. 5). Transcript relative abundance was greater than 1% for 13 MAGs (Fig. 6A) and the most transcriptionally active organisms were similar across both incubations, with 31–32% of transcripts mapping to the genome of Caproicibacterium sp002399445, 20–21% to Caproicibacter sp002316805, and 16–17% to Tractidigestivibacter sp902834555. Together, these three organisms accounted for 68% of gene expression under both incubation conditions.

Glucose was consumed in both incubations. Therefore, transcript abundance for fermentation of glucose to pyruvate, acetate, and lactate were assessed (Fig. 6B). Specifically, two routes of glucose utilization were explored- the Embden-Meyerhoff (EM) pathway (aka, glycolysis) and the Bifid shunt pathway (Fig. 7). Caproicibacterium sp002399445 and Caproicibacter sp002316805 expressed genes for glucose consumption via the EM pathway. While Caproicibacterium sp002399445 contained a gene annotated to produce the phosphoketolase enzyme, no transcripts mapped to this gene under either headspace condition. Tractidigestivibacter sp902834555 expressed genes for both the EM and Bifid shunt pathways, suggesting this organism can produce lactate and acetate from glucose. Notably, genes encoding glucokinase were not present in the Tractidigestivibacter sp902834555 genome; however, transcripts encoding phosphotransferase system (PTS) proteins, which couple the transport and phosphorylation of sugars [48], were among the most abundant transcripts for this organism.

Transcript abundance for advanced fermentation processes, including reverse β-oxidation, the Wood Ljungdahl pathway, the RNF complex, and hydrogen metabolism (Fig. 7) were then assessed. Caproicibacterium sp002399445 and Caproicibacter sp. 002316805 transcribed all genes for pyruvate flavodoxin oxidoreductase (pfor), reverse β-oxidation, the RNF complex, and hydrogen production via an FeFe hydrogenase. Tractidigestivibacter sp902834555 expressed a gene encoding pfor and a FeFe hydrogenase, suggesting reduced ferredoxin produced by pyruvate flavodoxin oxidoreductase may be oxidized through hydrogen generation. Tractidigestivibacter sp902834555 did not contain genes for reverse β-oxidation. Surprisingly, two MAGs classified as Caproicbacter species– Caproicibacter (Unclassified II) and Caproicibacter sp002409675– did not contain genes for reverse β-oxidation. This suggests species-level differences in fermentation end products across the Caproicibacter genus. Only one organism expressed all genes needed for the Wood Ljungdahl pathway – Clostridium_B sp003497125, and all these genes were only expressed during the incubation with H₂. Clostridium_B sp003497125 also expressed some – but not all -- genes for reverse β-oxidation, suggesting this organism may be able to incorporate H₂ and CO₂ into acetyl-CoA and while it has the genomic potential to perform chain elongation, it may be producing acetate rather than butyrate or hexanoate. Despite the increased transcription of Wood Ljungdahl pathway genes, there was no significant increase in hexanoate production under these conditions (Fig. 4E). Therefore, it is expected that this low abundance organism may contribute to butyrate production with H₂ as an electron donor but not hexanoate production.

Pseudoclavibacter_A canei and Lactobacillus paracasei are both potential sugar degraders based on genome annotations and published species descriptions [49, 50], but they did not express genes for sugar degradation (Fig. 6). Interestingly, Pseudoclavibacter_A canei is classified as a strict aerobe[49]. For both organisms, transcriptomes were dominated by phage-related transcripts. For Pseudoclavibacter_A canei transcripts that mapped to genes annotated as “phage portal protein,” “phage protein,” and “prophage Clp protease protein” were very highly expressed (RPKM > 40,000; Supplement 1). Similarly, the Lactobacillus paracasei transcriptome was dominated by “phage terminase, large subunit,” “phage portal protein,” “phage protein,” and “prophage pi2 protein 38.” This suggests that rather than contributing to trophic interactions needed for chain elongation, these organisms may be involved in bacteriophage cycling.

Caproicibacterium and Caproicibacter differ in gene abundance and synteny

Caproicibacterium sp. 002399445 and Caproicibacter sp. 002316805 were the predominant chain elongators found during the current study. While both genera have previously been described as chain elongating organisms, key differences between the genera have not been explored thoroughly. The metagenomic and metatranscriptomic results suggest that Caproicibacterium sp. 002399445 contains multiple copies of genes encoding acyl-CoA dehydrogenases and electron transfer flavoproteins while Caproicibacter sp. 002316805 has only single copies of the acyl-CoA dehydrogenase, EtfA, and EtfB gene cluster (Fig. 7). Gene clusters containing reverse β-oxidation genes have been discovered across chain-elongating taxa, including Caproiciproducens, Pseudoramibacter, Candidatus Weimeria and related species [12, 51, 52].

The MAG for Caproicibacter sp. 002316805 contains a single reverse β-oxidation gene cluster and a single pair of electron transfer flavoproteins contained within this cluster (Fig. 7). Caproicibacterium sp. 002399445 contains a gene cluster containing the reverse β-oxidation genes, another cluster with acyl-CoA dehydrogenase and two electron transfer flavoproteins, and an additional set of electron flavoproteins adjacent to a gene annotated as (S)-2-hydroxy-acid oxidase. A search with BLASTP confirmed the annotation of this gene but found it had 99% coverage and 87% sequence identity with a gene annotated as “L-lactate, D-lactate, and/or glycolate dehydrogenase” from Ruminococcaceae bacterium BL-6, suggesting that this cluster may encode an electron-bifurcating lactate dehydrogenase complex and may consume lactate via an electron confurcating lactate dehydrogenase which has been described for other lactate consumers [53]. Taken together, these results suggest that there may be differences in the chain elongation process between the Caproicibacter and Caproicibacterium genera. Gene expression and genome topology suggest that Caproicibacterium expresses multiple sets of electron bifurcating acyl-CoA dehydrogenases, but further work is needed to determine if these acyl-CoA dehydrogenases have different specificity for different chain lengths.

To further understand the difference in gene synteny across the genera Caproicbacterium and Caproicbacter, representative genomes from NCBI GenBank[54] were compared (Fig. 8). Caproicibacter sp. 002316805 and the genomes from the representative species GCA_002316805.1, GCA_002411605.1 and GCA_003539715.1, all have a reverse β-oxidation gene cluster but no additional gene clusters were present in the Caproicibacterium sp. 002399445. Further, Caproicibacterium sp. 002399445 and genomes from the representative species GCA_002399445.1, GCA_002398045.1, GCA_002397165.1, GCA_002397745.1, GCA_002409805.1, GCA_003535075.1, GCA_003519705.1 and GCA_903789565.1, have the reverse β-oxidation gene cluster, a separate cluster with acyl-CoA dehydrogenase and two electron transfer flavoproteins, and additional electron flavoproteins adjacent to a (S)-2-hydroxy-acid oxidase gene. Also, an additional cluster with (S)-2-hydroxy-acid oxidase and acyl-CoA dehydrogenase and two electron transfer flavoproteins genes was found in the GCA_903789565.1. Together, these results show key differences between the genomes of Caproicibacterium and Caproicibacter related to reverse b-oxidation and potentially lactate consumption.

Another key difference was that Caproicibacter sp. 002316805 expressed hydrogenase enzyme components that were not present in the MAG of Caproicibacterium sp. 002399445 (Supplement 1). NiFe hydrogenases are characterized as multi-enzyme complexes that can conserve energy through electron bifurcation [55], while the FeFe hydrogenase (E.C. 1.12.7.2) is typically a single component enzyme that reduces protons to H₂ with ferredoxin as the electron donor [56]. Besides the “assembly proteins” (HypCDEF), no other components of known NiFe hydrogenases were found in the genome of Caproicibacter sp. 002316805. This suggests that these proteins may be involved in the assembly of other nickel-containing metallocenter catalytic sub-units, but the precise role of these enzymes in Caproicibacter sp. 002316805 remains unknown.

Glucose was converted to hexanoate and butyrate, and the most abundant and transcriptionally active organisms were Caproicibacterium sp002399445, Caproicibacter sp002316805, and Tractidigestivibacter sp902834555. The current study and studies by others suggest that H₂ augmentation is not a reliable way to select for production of hexanoate over butyrate during chain elongation from sugars [17]. Clostridium_B sp003497125, however, increased gene expression related to the Wood Ljungdahl pathway, suggesting the potential for some sugar-consumers to also be capable of fixing CO₂ with H₂ as an electron donor. Ways to exploit these metabolic processes to increase hexanoate and octanoate production with H₂ supplementation require further attention. Further, differences between the genomes of Caproicibacterium and Caproicibacter suggest differences in the metabolic processes of lactate consumption, reverse b-oxidation, and hydrogen generation. The ways these genomic features and their expression impact final products of chain elongation should be investigated further.

Ethics approval and consent to participate

Not Applicable

Consent for publication

Not Applicable

Availability of data and materials

The raw DNA and RNA sequencing data for this project is available through the NCBI under PRJNA1105885. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was funded by the United States National Science Foundation Division of Chemical, Bioengineering, Environmental and Transport Systems award 2143446.

Authors’ contributions

PS and MS conceived of the project, designed the work, analyzed and interpreted the data, and drafted and revised the manuscript.

Acknowledgements

The authors acknowledge the Vermont Integrative Genomics Resource for DNA and RNA sequencing services and the Vermont Advanced Computing Core for high performance computing support.

Pauly M, Keegstra K: Cell‐wall carbohydrates and their modification as a resource for biofuels. The Plant Journal 2008, 54(4):559-568.
Walker LP, Wilson DB: Enzymatic hydrolysis of cellulose: An overview. Bioresource Technology 1991, 36(1):3-14.
Sarker A, Ahmmed R, Ahsan S, Rana J, Ghosh MK, Nandi R: A comprehensive review of food waste valorization for the sustainable management of global food waste. Sustainable Food Technology 2024.
Angenent LT, Richter H, Buckel W, Spirito CM, Steinbusch KJJ, Plugge CM, Strik DPBTB, Grootscholten TIM, Buisman CJN, Hamelers HVM: Chain Elongation with Reactor Microbiomes: Open-Culture Biotechnology To Produce Biochemicals. Environmental Science & Technology 2016, 50(6):2796-2810.
Seedorf H, Fricke WF, Veith B, Brüggemann H, Liesegang H, Strittmatter A, Miethke M, Buckel W, Hinderberger J, Li F et al: The genome of Clostridium kluyveri, a strict anaerobe with unique metabolic features. Proceedings of the National Academy of Sciences 2008, 105(6):2128-2133.
Reddy MV, Chang Y-C: Production of biofuel precursor molecules (monocarboxylic acids, biohydrogen) from apple and pumpkin waste through an anaerobic fermentation process. Sustainable Energy & Fuels 2021, 5(16):4133-4140.
Stamatopoulou P, Malkowski J, Conrado L, Brown K, Scarborough M: Fermentation of Organic Residues to Beneficial Chemicals: A Review of Medium-Chain Fatty Acid Production. Processes 2020, 8(12).
Scarborough MJ, Lynch G, Dickson M, McGee M, Donohue TJ, Noguera DR: Increasing the economic value of lignocellulosic stillage through medium-chain fatty acid production. Biotechnol Biofuels 2018, 11:200.
Li X, Jensen RL, Højberg O, Canibe N, Jensen BB: Olsenella scatoligenes sp. nov., a 3-methylindole- (skatole) and 4-methylphenol- (p-cresol) producing bacterium isolated from pig faeces. International Journal of Systematic and Evolutionary Microbiology 2015, 65(Pt_4):1227-1233.
Scarborough MJ, Myers KS, Donohue TJ, Noguera DR: Medium-Chain Fatty Acid Synthesis by “Candidatus Weimeria bifida” gen. nov., sp. nov., and “Candidatus Pseudoramibacter fermentans” sp. nov. Applied and Environmental Microbiology 2020, 86(3):e02242-02219.
Kim H, Choi O, Jeon BS, Choe W-S, Sang B-I: Impact of feedstocks and downstream processing technologies on the economics of caproic acid production in fermentation by Megasphaera elsdenii T81. Bioresource technology 2020, 301:122794.
Esquivel-Elizondo S, Bağcı C, Temovska M, Jeon BS, Bessarab I, Williams RB, Huson DH, Angenent LT: The isolate Caproiciproducens sp. 7D4C2 produces n-caproate at mildly acidic conditions from hexoses: genome and rBOX comparison with related strains and chain-elongating bacteria. Frontiers in microbiology 2021, 11:594524.
Flaiz M, Baur T, Brahner S, Poehlein A, Daniel R, Bengelsdorf FR: Caproicibacter fermentans gen. nov., sp. nov., a new caproate-producing bacterium and emended description of the genus Caproiciproducens. International Journal of Systematic and Evolutionary Microbiology 2020, 70(7):4269-4279.
Wang H, Zhou W, Gao J, Ren C, Xu Y: Revealing the Characteristics of Glucose- and Lactate-Based Chain Elongation for Caproate Production by Caproicibacterium lactatifermentans through Transcriptomic, Bioenergetic, and Regulatory Analyses. mSystems 2022, 7(5):e00534-00522.
Scarborough Matthew J, Lawson Christopher E, Hamilton Joshua J, Donohue Timothy J, Noguera Daniel R: Metatranscriptomic and Thermodynamic Insights into Medium-Chain Fatty Acid Production Using an Anaerobic Microbiome. mSystems 2018, 3(6):10.1128/msystems.00221-00218.
Fushinobu S: Unique Sugar Metabolic Pathways of Bifidobacteria. Bioscience, Biotechnology, and Biochemistry 2010, 74(12):2374-2384.
Baleeiro FC, Kleinsteuber S, Sträuber H: Hydrogen as a co-electron donor for chain elongation with complex communities. Frontiers in Bioengineering and Biotechnology 2021, 9:650631.
Chen W-S, Strik DP, Buisman CJ, Kroeze C: Production of caproic acid from mixed organic waste: an environmental life cycle perspective. Environmental Science & Technology 2017, 51(12):7159-7168.
Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, Alexander H, Alm EJ, Arumugam M, Asnicar F et al: Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nature Biotechnology 2019, 37(8):852-857.
Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP: DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods 2016, 13(7):581-583.
Glöckner FO, Yilmaz P, Quast C, Gerken J, Beccati A, Ciuprina A, Bruns G, Yarza P, Peplies J, Westram R et al: 25 years of serving the community with ribosomal RNA gene reference databases and tools. Journal of Biotechnology 2017, 261:169-176.
Krueger F, James F, Ewels P, Afyounian E, Weinstein M, Schuster-Boeckler B, Hulselmans G, sclamons: FelixKrueger/TrimGalore: v0.6.10 - add default decompression path (0.6.10). In.: Zenodo.
Wick R: Porechop. In.: GitHub; 2018.
Koren S, Walenz BP, Berlin K, Miller JR, Bergman NH, Phillippy AM: Canu: scalable and accurate long-read assembly via adaptive k-mer weighting and repeat separation. Genome Res 2017, 27(5):722-736.
Nurk S, Meleshko D, Korobeynikov A, Pevzner PA: metaSPAdes: a new versatile metagenomic assembler. Genome Res 2017, 27(5):824-834.
Li H, Durbin R: Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 2009, 25(14):1754-1760.
Walker BJ, Abeel T, Shea T, Priest M, Abouelliel A, Sakthikumar S, Cuomo CA, Zeng Q, Wortman J, Young SK et al: Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS One 2014, 9(11):e112963.
Wu Y-W, Simmons BA, Singer SW: MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics 2015, 32(4):605-607.
Kang DD, Li F, Kirton E, Thomas A, Egan R, An H, Wang Z: MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ 2019, 7:e7359.
Olm MR, Brown CT, Brooks B, Banfield JF: dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. The ISME Journal 2017, 11(12):2864-2868.
Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW: CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res 2015, 25(7):1043-1055.
Aziz RK, Bartels D, Best AA, DeJongh M, Disz T, Edwards RA, Formsma K, Gerdes S, Glass EM, Kubal M: The RAST Server: rapid annotations using subsystems technology. BMC genomics 2008, 9(1):1-15.
Parks DH, Chuvochina M, Rinke C, Mussig AJ, Chaumeil P-A, Hugenholtz P: GTDB: an ongoing census of bacterial and archaeal diversity through a phylogenetically consistent, rank normalized and complete genome-based taxonomy. Nucleic Acids Research 2021, 50(D1):D785-D794.
Stamatakis A: RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 2014, 30(9):1312-1313.
Quan J, Langelier C, Kuchta A, Batson J, Teyssier N, Lyden A, Caldera S, McGeever A, Dimitrov B, King R et al: FLASH: a next-generation CRISPR diagnostic for multiplexed detection of antimicrobial resistance sequences. Nucleic Acids Res 2019, 47(14):e83.
Kopylova E, Noé L, Touzet H: SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data. Bioinformatics 2012, 28(24):3211-3217.
Anders S, Pyl PT, Huber W: HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics 2015, 31(2):166-169.
Zhao Y, Li M-C, Konaté MM, Chen L, Das B, Karlovich C, Williams PM, Evrard YA, Doroshow JH, McShane LM: TPM, FPKM, or Normalized Counts? A Comparative Study of Quantification Measures for the Analysis of RNA-seq Data from the NCI Patient-Derived Models Repository. Journal of Translational Medicine 2021, 19(1):269.
Scarborough MJ, Lynch G, Dickson M, McGee M, Donohue TJ, Noguera DR: Increasing the economic value of lignocellulosic stillage through medium-chain fatty acid production. Biotechnology for Biofuels 2018, 11(1):200.
Li L, Liu C, Xu L, Zhuang H, He J, He Q, Zhang J: Acclimation of anaerobic fermentation microbiome with acetate and ethanol for chain elongation and the biochemical response. Chemosphere 2023, 320:138083.
Nzeteu Co, Coelho F, Trego AC, Abram F, Ramiro-Garcia J, Paulo L, O'Flaherty V: Development of an enhanced chain elongation process for caproic acid production from waste-derived lactic acid and butyric acid. Journal of Cleaner Production 2022, 338:130655.
Myers KS, Ingle AT, Walters KA, Fortney NW, Scarborough MJ, Donohue TJ, Noguera DR: Comparison of metagenomes from fermentation of various agroindustrial residues suggests a common model of community organization. Frontiers in Bioengineering and Biotechnology 2023, 11.
Myers KS, Ingle AT, Walters KA, Fortney NW, Scarborough MJ, Donohue TJ, Noguera DR: Comparison of metagenomes from fermentation of various agroindustrial residues suggests a common model of community organization. Frontiers in Bioengineering and Biotechnology 2023, 11:1197175.
Gu Y, Zhu X, Lin F, Shen C, Li Y, Ao L, Fan W, Ren C, Xu Y: Caproicibacterium amylolyticum gen. nov., sp. nov., a novel member of the family Oscillospiraceae isolated from pit clay used for making Chinese strong aroma-type liquor. International Journal of Systematic and Evolutionary Microbiology 2021, 71(4):004789.
Parks DH, Rinke C, Chuvochina M, Chaumeil P-A, Woodcroft BJ, Evans PN, Hugenholtz P, Tyson GW: Recovery of nearly 8,000 metagenome-assembled genomes substantially expands the tree of life. Nature Microbiology 2017, 2(11):1533-1542.
Parks DH, Chuvochina M, Waite DW, Rinke C, Skarshewski A, Chaumeil PA, Hugenholtz P: A standardized bacterial taxonomy based on genome phylogeny substantially revises the tree of life. Nat Biotechnol 2018, 36(10):996-1004.
Schoch CL, Ciufo S, Domrachev M, Hotton CL, Kannan S, Khovanskaya R, Leipe D, Mcveigh R, O’Neill K, Robbertse B: NCBI Taxonomy: a comprehensive update on curation, resources and tools. Database 2020, 2020:baaa062.
Jeckelmann J-M, Erni B: Transporters of glucose and other carbohydrates in bacteria. Pflügers Archiv-European Journal of Physiology 2020, 472(9):1129-1153.
Srinivasan S, Kim HS, Kim MK, Lee M: Pseudoclavibacter caeni sp. nov., isolated from sludge of a sewage disposal plant. International Journal of Systematic and Evolutionary Microbiology 2012, 62(Pt_4):786-790.
Collins MD, Phillips BA, Zanoni P: Deoxyribonucleic Acid Homology Studies of Lactobacillus casei, Lactobacillus paracasei sp. nov., subsp. paracasei and subsp. tolerans, and Lactobacillus rhamnosus sp. nov., comb. nov. International Journal of Systematic and Evolutionary Microbiology 1989, 39(2):105-108.
Scarborough Matthew J, Myers Kevin S, Donohue Timothy J, Noguera Daniel R: Medium-Chain Fatty Acid Synthesis by “Candidatus Weimeria bifida” gen. nov., sp. nov., and “Candidatus Pseudoramibacter fermentans” sp. nov. Applied and Environmental Microbiology 2020, 86(3):e02242-02219.
Liu B, Popp D, Müller N, Sträuber H, Harms H, Kleinsteuber S: Three novel Clostridia isolates produce n-caproate and iso-butyrate from lactate: comparative genomics of chain-elongating bacteria. Microorganisms 2020, 8(12):1970.
Kayastha K, Katsyv A, Himmrich C, Welsch S, Schuller JM, Ermler U, Müller V: Structure-based electron-confurcation mechanism of the Ldh-EtfAB complex. Elife 2022, 11.
Benson DA, Cavanaugh M, Clark K, Karsch-Mizrachi I, Lipman DJ, Ostell J, Sayers EW: GenBank. Nucleic Acids Res 2013, 41(Database issue):D36-42.
Feng X, Schut GJ, Haja DK, Adams MWW, Li H: Structure and electron transfer pathways of an electron-bifurcating NiFe-hydrogenase. Science Advances, 8(8):eabm7546.
Rangarajan ES, Asinas A, Proteau A, Munger C, Baardsnes J, Iannuzzi P, Matte A, Cygler M: Structure of [NiFe] hydrogenase maturation protein HypE from Escherichia coli and its interaction with HypF. J Bacteriol 2008, 190(4):1447-1458.

No competing interests reported.

Supplement1.xlsx

Download PDF

Reviews received at journal
21 Aug, 2024
Reviewers agreed at journal
19 Aug, 2024
Reviewers agreed at journal
14 Aug, 2024
Reviewers invited by journal
14 Aug, 2024
Editor assigned by journal
08 Aug, 2024
Submission checks completed at journal
08 Aug, 2024
First submitted to journal
08 Aug, 2024

You are reading this latest preprint version

Microbiome-based conversion of glucose to butyrate and hexanoate: roles and limitations revealed through genome-centered metatranscriptomics

Status:

Version 1

Abstract

Background

Results

Conclusions

Figures

Background

Methods

Bioreactor configuration and operation

Batch Experiments

Inoculum and media preparation

Chemical analyses

DNA and RNA Extraction and Sequencing

Bioinformatic Analyses

Results and Discussion

Glucose is converted to butyrate and hexanoate

A simple microbial community of Actinobacteria and Firmicutes was enriched

Genome-scale taxonomic classification reveals mostly uncultured species

Hydrogen supplementation does not increase hexanoate yield

Metatranscriptome dominated by two chain elongators and one hydrogenic sugar fermenter

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1