- Open Access
Enrichment of syngas-converting mixed microbial consortia for ethanol production and thermodynamics-based design of enrichment strategies
© The Author(s) 2018
- Received: 25 March 2018
- Accepted: 27 June 2018
- Published: 19 July 2018
The production of ethanol through the biochemical conversion of syngas, a mixture of H2, CO and CO2, has been typically studied using pure cultures. However, mixed microbial consortia may offer a series of benefits such as higher resilience and adaptive capacity, and non-sterile operation, all of which contribute to reducing the utility consumption when compared to pure culture-based processes. This work focuses on the study of strategies for the enrichment of mixed microbial consortia with high ethanologenic potential, investigating the effect of the operational conditions (pH and yeast extract addition) on both the ethanol yield and evolution of the microbial community along the enrichment process. The pH was selected as the main driver of the enrichment as it was expected to be a crucial parameter for the selection of carboxydotrophic bacteria with high ethanologenic potential. Additionally, a thermodynamic analysis of the network of biochemical reactions carried out by syngas-converting microbial consortia was performed and the potential of using thermodynamics as a basis for the selection of operational parameters favoring a specific microbial activity was evaluated.
All enriched consortia were dominated by the genus Clostridium with variable microbial diversity and species composition as a function of the enrichment conditions. The ethanologenic potential of the enriched consortia was observed to increase as the initial pH was lowered, achieving an ethanol yield of 59.2 ± 0.2% of the theoretical maximum in the enrichment at pH 5. On the other hand, yeast extract addition did not affect the ethanol yield, but triggered the production of medium-chain fatty acids and alcohols. The thermodynamic analysis of the occurring biochemical reactions allowed a qualitative prediction of the activity of microbial consortia, thus enabling a more rational design of the enrichment strategies targeting specific activities. Using this approach, an improvement of 22.5% over the maximum ethanol yield previously obtained was achieved, reaching an ethanol yield of 72.4 ± 2.1% of the theoretical maximum by increasing the initial acetate concentration in the fermentation broth.
This study demonstrated high product selectivity towards ethanol using mixed microbial consortia. The thermodynamic analysis carried out proved to be a valuable tool for interpreting the metabolic network of microbial consortia-driven processes and designing microbial-enrichment strategies targeting specific biotransformations.
- Carbon monoxide
- Mixed culture
- Microbial consortia
Over the past decades, the rising concerns about climate change and the depletion of fossil fuels, along with the ever increasing demand of transportation fuels, have led to the global implementation of policies fostering biofuel production. As a result of these policies, the biofuel market underwent a rapid expansion rising from less than 20 billion liters/year in 2001 to over 100 billion liters/year in 2011 . However, the rapid growth of this market, strictly based on first generation biofuels, brought along a series of environmental and socioeconomic impacts derived from the competition with food crops such as land use change, rising food and feed prices, and poor greenhouse gas emission savings [2, 3]. Developing second-generation biofuel technologies is thus considered an important step forward as these are based on the use of non-food biomasses and waste streams as feedstock, and are expected to overcome the limitations of first-generation biofuels in terms of environmental impacts and range of exploitable feedstocks.
Among the different approaches within second-generation biofuel technologies, syngas fermentation is one of the most promising as it combines the benefits of both thermochemical and biochemical biomass conversion processes. Typically, this process comprises thermochemical conversion of the biomass through gasification into synthesis gas, a mixture of mainly H2, CO2 and CO, followed by its biological conversion into a variety of chemicals and fuels [4–7]. The fermentation of syngas is carried out by anaerobic acetogenic bacteria, which are able to use both CO and H2/CO2 as the sole carbon and energy source through the Wood–Ljungdahl pathway. So far, mostly pure cultures have been employed in syngas fermentation processes, with the most common wild-type strains being C. autoethanogenum , C. ljungdahlii , C. ragsdalei  and C. carboxidivorans . However, several studies have reported that microbial growth on CO and H2/CO2 can be significantly inhibited by the impurities of syngas [12, 13], which may result in decreased productivity of syngas fermentation systems and/or higher raw syngas clean-up requirements. Another limitation is the fact that sterile operation is necessary to avoid a possible microbial contamination of the monoculture, which increases the energy input requirements.
A possible alternative to overcome these limitations is to use open-mixed microbial consortia. Developing microbial consortia-driven processes may allow reducing the cleaning requirements of raw syngas as they present a higher resilience and adaptive capacity due to their microbial diversity . Additionally, sterilization is not necessary when using microbial consortia , which contributes to reducing the utility consumption. Nonetheless, their higher complexity often entails an inadequate understanding of the microbial interactions within the consortium, resulting in limited process control. Low product selectivity is another challenge generally encountered in microbial consortia-driven processes, ultimately lowering the yield of the desired product.
Although the use of open-mixed microbial consortia in syngas fermentation processes is still rather limited compared to pure cultures, the potential of microbial consortia for producing H2 , CH4 , carboxylic acids [18–20] and higher alcohols  has been demonstrated in a number of studies. The production of solvents through the reduction of carboxylic acids using either syngas or H2 as electron donor has also been studied [22, 23]. However, scientific literature on the production of ethanol by mixed microbial consortia using syngas as the sole carbon source is scarce, with only two studies showing a significant ethanologenic potential. Singla et al.  were the first demonstrating a high ethanol production using an enriched mixed culture from the culture collection of TERI University. In their study, the operating conditions were optimized for maximizing the ethanol yield in batch experiments, obtaining a maximum concentration of 1.54 and 0.9 g/l of ethanol and acetic acid, respectively. In turn, Ganigué et al.  studied the enrichment of a pre-acclimatized open-mixed microbial consortium using 2-bromoethanesulfonate for inhibiting methanogenic activity, where significant amounts of ethanol, butanol and hexanol were produced through syngas fermentation and chain elongation. Nevertheless, a dedicated study on enrichment of mixed microbial consortia aiming at enhancing the product selectivity towards ethanol has not been conducted yet.
This work focuses on the study of selective enrichment strategies for developing microbial consortia with high ethanologenic potential, laying special emphasis on the effect of the enrichment conditions on both the ethanol yield and selectivity, and the evolution of the microbial community along the enrichment process. Additionally, the thermodynamics of the network of biochemical reactions of microbial consortia is evaluated based on the Gibbs free energy change at different enrichment conditions with the perspective of using thermodynamics as a tool for developing selective enrichment strategies targeting specific biotransformations.
The inoculum used was a combination of two different types of anaerobic sludge collected from the Lundtofte wastewater treatment plant (Denmark) and from a lab-scale anaerobic digester fed with manure (at Chemical and Biochemical Engineering Department, Technical University of Denmark). The inoculum was prepared by mixing the two anaerobic sludges in equal amounts (50/50 v/v) and adjusting the pH to 6 with 1 M HCl while flushing with N2 to ensure anaerobic conditions.
Unless otherwise stated, the starting inoculum used in the enrichments (see “Enrichment experiments and conditions”) underwent a heat-shock treatment to suppress the methanogenic activity. The heat-shock treatment was carried out by heating the mixture of anaerobic sludges up to 90 °C for 15 min while flushing with N2.
Growth medium composition
A modified basal anaerobic (BA) medium was used in all experiments, which consisted of six stock solutions containing phosphate buffer, vitamins, trace elements, salts, chelating agents and reducing agents. The stock solutions had the following composition: solution A (NH4Cl, 100 g/l; NaCl, 10 g/l; MgCl2∙6H2O, 10 g/l; CaCl2∙2H2O, 5 g/l), vitamins solution according to Wolin et al. , trace metal solution (FeCl2∙4H2O, 2000 mg/l; H3BO3, 50 mg/l; ZnCl2, 50 mg/l; CuCl2, 30 mg/l; MnCl2∙4 H2O, 50 mg/l; (NH4)6Mo7O24∙4 H2O, 50 mg/l; AlCl3, 50 mg/l; CoCl2∙6H2O, 50 mg/l; NiCl2, 50 mg/l; Na2SeO3∙5H2O, 100 mg/l; Na2WO4∙2H2O, 60 mg/l), chelating agent solution (Nitrilotriacetic acid, 1 g/l) and reducing agent solution (Na2S∙9H2O, 25 g/l).
The medium was prepared by adding 10 ml/l of solution A, 1 ml/l of trace metal solution, 10 ml/l of vitamins solution, 10 ml/l of reducing agent solution, 20 ml/l of chelating agent solution and distilled water up to 1 l. The pH was adjusted with phosphate buffer (50 mM) using three stock solutions (K2HPO4∙3H2O, 200 g/l; KH2PO4, 136 g/l; H3PO4, 98 g/l). When relevant, yeast extract (YE) and sodium acetate were added to the medium with a final concentration of 0.5 g/l and 20 mM, respectively.
Enrichment experiments and conditions
All enrichment experiments were performed in 330 ml serum flasks with an active volume of 100 ml and an inoculum size of 15% v/v (15 ml). The medium (85 ml) was added to the flasks and was flushed with H2 to create anaerobic conditions. After the flasks were sealed with rubber stoppers and screw plugs, the remaining gases (CO and CO2) were added up to a total pressure of 2 atm prior to inoculation using a precision pressure indicator (model CPH6400, WIKA, Germany). All gases used had purity above 99.9%, and were purchased from AGA (Denmark). After inoculation, the total initial pressure increased to 2.14 atm and the final gas composition of the headspace corresponded to approximately 10.1 mmol of H2, 4.5 mmol of CO and 5.5 mmol of CO2 at 25 °C. The fermentation flasks were incubated in the dark at 37 °C and 100 rpm. Control experiments were performed at the same incubation conditions with no addition of gaseous substrates, adjusting the gaseous composition of the headspace to 1.44 atm of N2 and 0.56 atm of CO2 prior to inoculation.
Enrichment conditions based on initial pH and initial addition of acetate
Syngas composition (%H2, %CO, %CO2)
5.95 ± 0.06
50%, 22.2%, 27.8%
5.51 ± 0.09
50%, 22.2%, 27.8%
5.05 ± 0.02
50%, 22.2%, 27.8%
5.43 ± 0.13
50%, 22.2%, 27.8%
5.02 ± 0.10
50%, 22.2%, 27.8%
5.04 ± 0.13
50%, 22.2%, 27.8%
4.99 ± 0.08
50%, 22.2%, 27.8%
NaCH3COO (20 mM)
The enrichment series HT5YE was interrupted at transfer T4 and presented a significant loss of solventogenic activity upon resuming the enrichment. Thus, this enrichment was extended for one more transfer to confirm the recovery of the previous activity.
Samples for the determination of the metabolites concentration and yield in each transfer were taken at the beginning of the experiments and after the culture reached the stationary phase. Samples for microbial growth determination were taken from transfer T3, when the solids from the anaerobic sludge were completely diluted and did not interfere with the absorbance of the fermentation broth. As it was not possible to measure microbial growth during the first transfers of the enrichments, the distinction between exponential and stationary growth phase was based on the profile of the consumption of H2 and CO over time.
DNA extraction, sequencing and microbial population analysis
For analysis of microbial composition, selected fermentation steps during enrichments were sampled at both exponential and stationary growth phase. 5 ml of culture was spun down and genomic DNA was isolated using DNeasy Blood & Tissue Kit, following manufacturer recommendations for Gram-positive bacteria (Qiagen, Copenhagen).
The DNA sample was submitted to Macrogen Inc. (Korea) for 16S rRNA amplicon library preparation and sequencing using Illumina Miseq instrument (300 bp paired-end sequencing). Amplification of V3 and V4 region of 16S rRNA gene was done with Pro341F 5′-CCTACGGGNBGCASCAG-3′ and Pro805R 5′-GACTACNVGGGTATCTAATCC-3′ . Sequences containing primers were trimmed with cutadapt and all other reads filtered out . Subsequently, filtering, generation of operational taxonomic units (OTUs) and mapping of reads to OTUs were performed using the UPARSE/unoise3 pipeline . Taxonomy was assigned to OTUs using SINTAX and NCBI database of 18421 16S ribosomal RNA sequences from NCBI RefSeq Targeted Loci Project . Subsequently, OTU table was corrected using the UNCROSS algorithm , normalized with respect to 16S copy number and primer mismatches with UNBIAS algorithm . Each sample was normalized to the depth of the sample with least counts. OTUs with overall count less than 100 were filtered out and sample data from available replicate runs have been collapsed based on the mean count. Downstream analyses were performed with Phyloseq R  package and MicrobiomeAnalyst web service, available at http://www.microbiomeanalyst.ca/ . Fitting of environmental variables onto ordination plots was performed with R Vegan package .
The gaseous composition of the headspace (H2, CO, CO2 and CH4) was determined using a gas chromatograph (model 8610C, SRI Instruments, USA) equipped with a thermal conductivity detector and two packed columns, a 6′ × 1/8″ Molsieve 13× column and a 6′ × 1/8″ silica gel column connected in series through a rotating valve. The column temperature was maintained at 65 °C for 3 min, followed by a temperature ramp of 10 °C/min to 95 °C and 24 °C/min from 95 to 140 °C. Gaseous samples of 50 µl were collected with a gas-tight syringe (model 1750SL, Hamilton). Volatile fatty acids (VFA) (acetate, propionate, iso-butyrate, butyrate and caproate) and alcohols (ethanol and 1-butanol) were determined using a high performance liquid chromatograph (Shimadzu, USA) equipped with a refractive index detector and an Aminex HPX-87H column (Bio-Rad, USA) at 63 °C. A solution of 12 mM H2SO4 was used as eluent at a flow rate of 0.6 ml/min. Volatile Suspended Solids (VSS) concentration in the fermentation broth was determined according to standard methods . Microbial biomass growth was monitored by measuring the absorbance of liquid samples at 600 nm using a spectrophotometer (DR2800, Hach Lange), and was correlated to the volatile suspended solids (VSS) concentration of the fermentation broth.
Product yield and efficiency calculations
Biochemical reactions, ATP yield and average stoichiometric number used in the thermodynamic potential factor calculation
Stoichiometry of biochemical reactions
ATP yield (mol per reaction)
H2/CO2 conversion into acetate/ethanol
4 H2 + 2 CO2 → CH3COO− + H+ + 2 H2O
6 H2 + 2 CO2 → CH3CH2OH + 3 H2O
CO conversion into acetate/ethanol
4 CO + 2 H2O → CH3COO− + H+ + 2 CO2
6 CO + 3 H2O → CH3CH2OH + 4 CO2
VFA reduction to corresponding alcohols
CH3COO− + H+ + 2 H2 → CH3CH2OH + H2O
CH3(CH2)2COO− + H+ + 2 H2 → CH3(CH2)2CH2OH + H2O
CH3COO− + H+ + 2 CO + H2O → CH3CH2OH + 2 CO2
CH3(CH2)2COO− + H+ + 2 CO + H2O → CH3(CH2)2CH2OH + 2 CO2
5 CH3CH2OH + 3 CH3COO− → 4 CH3(CH2)2COO− + H+ + 3 H2O + 2 H2
Thermodynamic potential factor (F T)
The calculations of the FT for the direct conversion of H2/CO2 and CO into acetate and ethanol were carried out using ATP yields given by Bertsch and Müller . The chain elongation reaction was assumed to yield 1 mol ATP per reaction, obtained through substrate level phosphorylation, according to Angenent et al. . The reduction of acetic acid into ethanol with H2 as electron donor was assumed to follow the acetate activation pathway via acetyl-CoA at the expense of 1 ATP, which would be compensated through the translocation of four protons across the membrane resulting in 0.33 mol ATP per reaction, as suggested by González-Cabaleiro et al. . The analogous acetic acid reduction to ethanol using CO was assumed to follow the same pathway; however, a tentative yield of 0.66 mol of ATP per reaction was used, as the ATP yield of this reaction would be expected to be higher due to the oxidation of the additional reduced ferredoxin produced by CO dehydrogenase. The stoichiometry of the ATP synthesis through ion translocation used for the calculated ATP yields corresponded to a fixed ratio of 3 mol of H+/Na+ per mol of ATP. The average stoichiometric number (χ) was determined by the number of ions translocated across the membrane for all reactions, except for the chain elongation, in which the substrate level phosphorylation was used instead. It should be noted that the ATP yields used here were not determined experimentally, and thus, are subject to uncertainties derived from the assumptions made in each case. To account partially for these uncertainties and give an idea of the sensitivity of the thermodynamic potential factor (FT) to the energy conservation parameter, a rather broad range of Gibbs free energy of phosphorylation was used for the calculations corresponding to 45 kJ/mol ATP , 57.5 kJ/mol ATP and 70 kJ/mol ATP . The values used for the calculations are summarized in Table 2.
Enrichment of syngas-converting microbial consortia based on pH
Ethanologenic potential of enriched consortia
A number of enrichment strategies using the pH as the main selective driver were designed to study the evolution of the ethanologenic activity during the enrichment of microbial consortia at different initial conditions. Different initial pH conditions were tested using a heat-shock-treated inoculum (pH 6, 5.5 and 5) and a non-treated inoculum (pH 5).
All enrichment strategies successfully suppressed the methanogenic activity of the anaerobic sludge as methane was not detected in any transfer of the enrichments. This was expected in the enrichments using a heat-shock-treated inoculum, since spore-forming bacteria should be predominant in the anaerobic sludge as a result of the heat treatment . In turn, small amounts of methane were expected when using the non-treated inoculum due to the abundance of methanogenic archaea in the anaerobic sludge, as observed by Steinbusch et al. . However, in this study, the low initial pH (pH 5) of experiments using the non-treated inoculum inhibited both methanogenic and acetogenic growth when YE was not added to the growth medium, whereas experiments with YE addition presented exclusively acetogenic growth. This indicates that the methanogenic activity of the sludge was inhibited by the combination of low pH and toxicity of CO. Thus, open-mixed cultures could be used in syngas fermentation processes with no need of heat treatment or methanogenic inhibitors, just by operating at harsh conditions for methanogenic archaea.
The enrichment with initial pH 5 using the non-treated inoculum (NT5YE) presented a similar trend with enrichment HT5YE, as the solventogenic activity was rapidly boosted along the successive transfers (Fig. 2e). However, the maximum ethanol yield obtained in enrichment NT5YE at transfer T6, namely, 0.033 mol/e-mol and 39.7% of the stoichiometric maximum (average of 0.032 ± 0.002 mol/e-mol), was not as high as that of enrichment HT5YE. An explanation for this difference in the product yields could be based on the heat treatment of the inoculum, as the enrichment HT5YE presented a more abrupt response upon exposure to the enrichment conditions (from T0 to T1) compared to enrichment NT5YE, where changes in the product profile took place gradually (from T0 to T2). Thus, it is possible that the higher degree of sporulation derived from the heat treatment of the initial inoculum favored a faster microbial selection process ultimately resulting in a different ethanologenic activity in these two enriched consortia. Besides the quantitative difference in the final ethanol yield, the high similarity in the behavioral traits of these microbial consortia is also supported by their common response upon reactivation of the cultures. As shown in Fig. 2d, e, a noticeable decrease of solventogenic activity was observed in both enrichments after stopping them for 2 months at transfer T4 (HT5YE) and T2 (NT5YE). Additionally, a very similar or even higher solventogenic activity was recovered after two transfers upon resuming the enrichments in both microbial consortia.
The apparent biomass yield was observed to be affected by both the initial pH of the enrichments and the addition of YE. Generally, the average biomass yield along the enrichments varied between 1.7 and 2.8 mg VSS/e-mol, with enrichment HT5.5 presenting the lowest biomass yield and enrichments HT5YE and NT5YE exhibiting the highest biomass yields (Additional file 1: Figure S5). No statistically significant differences were found between the average biomass yield of the enrichments at different pH, with P values above 0.05 in all cases when comparing HT6 to the rest of the enrichments (Additional file 1: Figure S5 and Table S1). However, the fact that the enrichment HT5.5 could not be reactivated at transfer T6 and that enrichment HT5 did not present any growth indicates a clear negative effect of the pH on microbial growth. In turn, the addition of YE was observed to improve the biomass yield of the enrichment cultures as statistically significant differences with P values below 0.05 were found between enrichment HT5.5 and all enrichments with YE addition, namely, HT5.5YE, HT5YE and NT5YE (Additional file 1: Table S1).
A low pH is commonly applied in syngas fermentation studies [47, 48] based on the hypothesis that the higher diffusion of VFAs through the cell membrane at acidic pH triggers solventogenesis as a means of preventing a further intracellular pH drop [49, 50]. The observations made in this study are in agreement with this hypothesis as (i) the highest ethanol yields were obtained in the enrichments at the lowest initial pH tested, and (ii) the final pH of the fermentations oscillated around 4.3–4.6 in most enrichment conditions and seemed not to be related to the initial pH conditions (Fig. 2c, d). Thus, it is likely that intracellular pH homeostasis may have driven a higher ethanol production by the enriched consortia.
The addition of YE appeared to have no effect on the final yield of ethanol in enrichments at an initial pH of 5.5 (Fig. 2b, c), but triggered the production of butyrate, butanol and small amounts of caproate leading to a broader product spectrum in all YE-supplemented enrichments (Fig. 2c–e). The production of butyrate and butanol was observed to take place in a two-step reaction, which indicates that they were produced through chain elongation and reduction of VFAs (Additional file 1: Figures S6, S7). The potential of microbial consortia for producing medium-chain fatty acids (MCFAs) through chain elongation has been shown in a number of studies [18, 51]. However, when it comes to ethanol production, the chain elongation process is often regarded as a major drawback as it reduces the selectivity of the mixed culture towards ethanol due to the conversion of ethanol and VFAs into MCFAs, as found in El-gammal et al. . In this study, a significant chain-elongating activity appeared to be prevented by the low pH of the fermentations (generally ranging between 4.3 and 4.6 at the end of the experiments), since both acetate and ethanol remained as the major end products at all enrichment conditions. Ganigué et al.  reached similar conclusions in a study targeting the production of higher alcohols, in which low pH was found to affect negatively the chain elongation process. Nevertheless, the reduced chain-elongating activity found in the present study allowed achieving high ethanol yield in enrichments at pH 5.
A maximum ethanol yield of 0.050 mol/e-mol (59.5% of e-mol recovery) and an ethanol-to-acetate ratio of 1.58 g/g was achieved in enrichment HT5YE. The maximum ethanol-to-acetate ratio obtained was significantly higher than those often reported in other batch experiments using pure cultures such as C. ragsdalei (1.30 g/g) , C. autoethanogenum (0.39 g/g)  and C. ljungdahlii (0.70 g/g) , and in other mixed-culture studies with ratios below 0.4 g/g [20, 52]. Yet, higher ethanol-to-acetate ratios were reported by Singla et al. (2014) using the enriched culture TERI-SA1 (2.46 g/g).
Efficiency calculated in terms of e-mol and Cmol recovery and product yields for all enriched microbial consortia
92.92 ± 0.54
85.83 ± 2.46
89.84 ± 1.80
88.57 ± 1.77
95.01 ± 3.29
83.16 ± 0.70
77.33 ± 2.31
93.18 ± 1.69
80.29 ± 0.83
91.60 ± 6.41
Product yield (% e-mol/e-mol)
83.32 ± 0.81
61.15 ± 7.41
41.38 ± 1.75
29.29 ± 0.63
27.68 ± 1.67
0.00 ± 0.00
0.11 ± 0.20
0.01 ± 0.08
2.11 ± 0.38
0.26 ± 0.04
1.55 ± 0.09
0.00 ± 0.00
1.82 ± 0.78
0.00 ± 0.00
1.17 ± 0.31
0.00 ± 0.00
0.00 ± 0.00
11.59 ± 4.82
2.05 ± 0.74
16.51 ± 1.07
8.16 ± 0.49
25.20 ± 5.32
29.40 ± 5.36
59.15 ± 0.18
34.81 ± 2.27
0.00 ± 0.00
0.00 ± 0.00
15.37 ± 1.55
3.50 ± 0.64
21.82 ± 1.45
0.00 ± 0.00
0.00 ± 0.00
1.39 ± 0.03
3.33 ± 0.20
1.09 ± 0.09
Biomass yield (g VSS/e-mol)
2.10 ± 0.11
1.83 ± 0.42
3.21 ± 0.51
3.13 ± 0.20
2.85 ± 1.06
Microbial characterization of enrichment cultures
A total of 49,736,621 sequences were obtained from all investigated samples after quality checking and data filtering, with an average of 956,473 reads per sample (range 309.337–8,985,816 reads per sample). Replication, error correction, denoising with unoise algorithm and filtering of OTU table resulted in 6183 OTUs with all but one OTUs belonging to bacteria domain. Considering all sequences retrieved in the present study, Firmicutes accounted for the largest fraction (78% of the total), mainly represented by the classes Clostridia, Tissierella and Bacilli.
Despite the clear dominance of OTUs belonging to the genus Clostridium in all enrichments, the abundance of individual OTUs found in the enrichment samples varied depending on the enrichment conditions (Additional file 2: Table S3). The mean relative frequencies between the OTU abundances were compared across all samples with and without addition of YE. It was found that representative sequences of abundant OTUs identified solely in enrichments without addition of YE (HT6 and HT5.5), aligned with high identity (98.4–100%) to 16S rRNA genes from Clostridium autoethanogenum and Clostridium ljungdahlii. Abundances of these OTUs (2;1302;1233;1249;1983) were 81.5% and 41.4% in samples HT5.5T5SP and HT6T6SP, respectively (Additional file 2: Table S3). In turn, reads mapping to these OTUs were negligible in YE-supplemented enrichments. Moreover, identified OTUs exclusively present in YE-supplemented enrichments and their representative sequences exhibited the best alignment (97–100% identity) with 16S rRNA sequences of Clostridium drakei and Clostridium carboxidivorans. Abundance of these OTUs (1;337;434;359) was up to 87.5% in sample NT5YET6SP (Additional file 2: Table S3). These differences indicate that the addition of YE, besides promoting better growth conditions for the entire microbial consortium, also played a determining role as a selection factor in addition to the pH conditions and the substrate composition.
The range of metabolites found in each of the enrichment series was in agreement with the product portfolio of the putative dominant species identified in the enrichment samples. As mentioned above, acetate and ethanol were the main metabolites in enrichments HT6 and HT5.5 where the putative dominant species were C. autoethanogenum and C. ljungdahlii, while longer carbon chain products such as butyrate and butanol were also found in YE-supplemented enrichments (HT5.5YE, HT5YE and NT5YE) with C. drakei and C. carboxidivorans as putative dominant species. Both C. autoethanogenum and C. ljungdahlii have been reported to produce only acetate and ethanol when fermenting syngas in batch cultures [55, 56]. In turn, C. drakei and C. carboxidivorans present a broader product spectrum including butyrate, butanol, and even caproate and hexanol in the case of C. carboxidivorans [57, 58], which are produced through re-assimilation, chain elongation and reduction of the primary metabolites . Additionally, the optimum growth conditions for all these species vary between a pH of 5.5 and 6.2 and temperatures around 37 °C, which explains the dominance of these species during the enrichments. Therefore, most likely the dominant species identified were the major contributors to the formation of products observed during the enrichments.
The ethanol yield seemed to be independent of the microbial composition of the enrichment cultures at genus level. Enrichments HT6 and HT5.5YE resulted in a similar microbial composition (Fig. 3) and presented a maximum ethanol yield along the enrichment of 0.015 and 0.028 mol/e-mol, respectively (Fig. 2). Similarly, enrichments HT5.5, HT5YE and NT5YE also presented similar microbial composition with a clear dominance of the genus Clostridium (Fig. 3) and resulted in a maximum ethanol yield of 0.025 mol/e-mol, 0.050 mol/e-mol and 0.034 mol/e-mol, respectively (Fig. 2). On the contrary, enrichments at similar operating conditions and different microbial composition (HT5.5 and HT5.5YE) resulted in similar maximum ethanol yields. Thus, despite the clear effect of the pH on the microbial composition of the enriched consortia, it can be concluded that the shift towards ethanol observed in the enrichment experiments was probably the result of the metabolic response to the different initial pH conditions and not so dependent on the microbial composition of the enrichment cultures.
A direct comparison with the literature is not possible since the only quantitative analysis of the composition of a syngas-converting microbial consortium using a 16S rRNA amplicon-based sequencing method was performed under higher pH conditions (pH 7.5), and thus, resulted in significantly different microbial composition . However, the species identified in other studies carried out at a comparable pH range (pH 6.2–6.0) which were entirely consistent with the results found here [21, 24]. Ganigué et al. (2016) studied the composition of a microbial consortium during the conversion of syngas into higher alcohols using PCR-DGGE analysis and identified both of the putative dominant species reported here (C. autoethanogenum/C. ljungdahlii and C. drakei/C. carboxidivorans). In their study, C. autoethanogenum/C. ljungdahlii were found to be the main species carrying out the carbon fixation. In turn, Singla et al.  found that C. drakei and C. scatalogenes were either the main or possibly the only members of the enriched microbial consortium TERI-SA1. Interestingly, YE was not added in the medium used by Ganigué et al.  during the enrichment, while Singla et al.  added 1 g/l of YE to the medium. Taking this into consideration, the findings reported in the literature and the results reported here follow the same trend, with the putative dominant species of the enrichment cultures being C. autoethanogenum/C. ljungdahlii when YE was absent in the growth medium and C. drakei/C. carboxidivorans when YE was added to the medium.
Thermodynamic analysis of the metabolic network of the enriched consortia
The enriched consortia developed through pH-based enrichments were observed to produce ethanol with relatively high selectivity. However, the production of ethanol took place during the exponential phase of the fermentations, and thus, it was not possible to distinguish between direct production of ethanol and reduction of acetic acid to ethanol. Therefore, a thermodynamic analysis of the metabolic network of net biochemical reactions taking place during the activity of the enriched consortia was performed to identify possible bioenergetic drivers of the metabolic shift observed under different enrichment conditions. Based on experimental observations, the reactions considered for evaluating the ∆rG′310 K and the thermodynamic potential factor (FT) under changing process conditions were the production of ethanol and acetic acid from H2/CO2 and CO, the production of butyric acid through chain elongation, and the reduction of both acetic and butyric acid into their corresponding alcohols using either H2 or CO as electron donor.
According to the thermodynamic analysis, the direct conversion of H2/CO2 and CO into either acetate or ethanol is not expected to be thermodynamically controlled until these substrates become severely depleted. However, the higher thermodynamic driving force (lower ∆rG′310 K) and ATP yield per mol of substrate for acetate-producing reactions suggest that these would prevail over direct ethanol-producing reactions under kinetic control. Besides, calculations carried out by Bertsch and Müller  for the model organism Acetobacterium woodii indicate that the production of ethanol from H2/CO2 might not be possible, as this reaction would require a net input of 0.1 mol of ATP per mol of ethanol.
As opposed to direct conversion route from H2/CO2 and CO to liquid products, VFA-reducing reactions are subject to thermodynamic control under the operating conditions considered and are clearly favored upon decreasing the pH. The FT values for all VFA-reducing reactions show that changes in ∆rG′310 K at the pH range studied have a strong effect on the rates of these reactions, which could explain the higher ethanol yield obtained in the enrichment experiments at an initial pH of 5. As illustrated in Fig. 4b, a high pH in the fermentation broth renders the reduction of acetate with H2 unfeasible (negative FT values). However, the boundaries of feasibility for this reaction cannot be accurately delimited due to the high sensitivity of FT to the values of ATP yield and Gibbs free energy of phosphorylation (∆Gp) used in the calculations. Considering an ATP yield of 0.33 mol per reaction and a ∆Gp of 45 kJ/mol of ATP, the reduction of acetate would be feasible below a pH of 5.6, whereas using a ∆Gp of 70 kJ/mol of ATP would render this reaction unfeasible at any pH resulting in a maximum FT of − 0.23 at pH 3. Thus, detailed conclusions on whether this reaction is possible as a function of pH cannot be drawn, although it is obvious that this reaction is more likely to occur at the lower range of pH studied. On the other hand, using CO as electron donor for the reduction of acetate provides a much lower ∆rG′310 K, reducing the uncertainties on the activity of this reaction. In this case, the use of CO as electron donor is possible at all conditions regardless of the ∆Gp considered (Fig. 4b). Furthermore, decreasing the pH from 6 to 5 causes the FT of this reaction to increase from 0.68 to 0.79 (Fig. 4b), indicating that the acetate-reducing activity is significantly boosted as the pH decreases. Thus, it can be concluded that acetate-reducing reactions played an important role on the solventogenic activity observed in the enrichment experiments. Additionally, based on the FT for these reactions, the acetate-reducing activity using CO rather than H2 as electron donor was probably more significant during the enrichments. This is in line with the observations made by Hu et al.  while studying the thermodynamics of the oxidation of CO and H2, where it was concluded that the use of CO as a source of electrons is thermodynamically more favorable than H2 at all conditions.
In “Ethanologenic potential of enriched consortia”, it was hypothesized that the chain elongation reaction was inhibited by the decrease of pH during the fermentation, as both acetate and ethanol remained as the main products of the fermentation and were only partially converted into butyrate (Additional file 1: Figures S6, S7). However, the results of the FT for this reaction at different pH conditions show that the chain elongating activity is negatively affected by the decrease of pH when considering a ∆Gp of 70 kJ/mol, with FT values corresponding to 0.98, 0.86 and 0.68 at pH 6, 5 and 4, respectively (Fig. 4b). Therefore, the inhibition due to pH drop observed experimentally could be grounded on a limitation in the thermodynamic driving force for this reaction to proceed forward.
The thermodynamic analysis carried out here suggests that a significant amount of ethanol was produced via a two-step reaction, where direct production of acetic acid was initially favored followed by its reduction into ethanol in a second step. This is consistent with the distinction between acidogenic and solventogenic growth phases commonly applied in syngas fermentation processes [48, 62]. Furthermore, the limited thermodynamic drive for chain-elongating activity found when decreasing the pH could explain the high selectivity towards ethanol observed in enrichments at pH 5. Thus, the methodological approach used here proved to be useful for a qualitative interpretation of how the metabolic network of mixed microbial consortia responds when changing operational conditions. Of course, microbial growth inhibition phenomena due to high VFAs/solvents concentration are not taken into account in this method and need to be considered from a microbiological perspective; however, the enrichment experiments took place at low substrate and product concentration and were not expected to present such inhibition phenomena. This method presented low accuracy when attempting to draw definite boundaries on the feasibility of the acetate reduction with H2 due to the broad range of ∆Gp used in the calculations. Other limitations identified were the fact that the energy conservation requirements, determined by the ATP yield and the ∆Gp, were assumed to be constant regardless of the reaction and operating conditions considered. The stoichiometry of ATP synthesis was also assumed to have a fixed ratio of 3 ions translocated per mol of ATP. Nevertheless, both the ∆Gp and the stoichiometry of ATP synthesis have been shown to be subject to variation depending on several factors such as intracellular ATP/ADP ratio, electrochemical membrane potential, electron donors and acceptors considered, or even the species carrying out the reaction .
Despite the limitations outlined above, the thermodynamic analysis allowed for interpretation of the effects of operating conditions on the network of biochemical reactions prevailing in mixed microbial consortia. Thus, this method could be used for the selection of operational conditions with the aim of boosting specific reactions. To test the validity of this method for predicting changes in the microbial activity of enriched consortia and improving further the ethanol yield obtained previously, an additional experiment series was performed.
Enrichment strategies based on thermodynamics of the metabolic network
Thermodynamic predictions of the microbial activity
From a thermodynamic perspective, the metabolic network of microbial consortia can be affected by several operating parameters such as the partial pressure of gases, the concentration of products or the pH. Several of these parameters could potentially enhance the production of ethanol due to their distinct effect on different reactions such as the pH already discussed, the partial pressure of CO2 given the distinct stoichiometric CO2 formation in acetate- and ethanol-producing reactions, the partial pressure of H2 and CO affecting acetate-reducing reactions, and the initial concentration of acetate and ethanol affecting the whole metabolic network. In this case, given the effect of the pH on the acetate-reducing activity discussed in “Thermodynamic analysis of the metabolic network of the enriched consortia”, it was decided to study the effect of the initial acetate concentration in the medium to boost these reactions even further. However, this can be regarded as a proof-of-concept since this method could be used to evaluate the effect of the abovementioned parameters on any reactions taking place under thermodynamic control, e.g., in systems operating in continuous mode under substrate-limiting conditions.
Comparing the effect of the pH and the initial acetate concentration on acetate-reducing reactions, the analysis of the changes in ∆rG′310 K and FT shows that the concentration of acetate has a stronger effect on the activity rates of these reactions. An increase in acetate concentration from 1 mM to 20 mM would significantly boost the acetate-reducing activity as the FT would increase from 0.79 to 0.88 when using CO as electron donor, and from − 0.13 to 0.45 when using H2. In this case, at an initial acetate concentration of 20 mM, both reactions would be clearly feasible regardless of the ∆Gp considered, even when using the more conservative ∆Gp of 70 kJ/mol of ATP (Fig. 5b). Thus, according to these calculations, an enrichment at pH 5 and 20 mM of initial concentration of acetate would be expected to boost the ethanologenic potential of the microbial consortium by increasing its acetate-reducing activity.
As opposed to lowering the pH, increasing the initial concentration of acetate would favor the chain-elongating activity. In this case, the FT for chain elongation would remain constant at values approaching 1 by increasing the initial concentration of acetate from 1 mM to 20 mM when using a ∆Gp of 57.5 kJ/mol of ATP (Fig. 5b). Considering the most conservative ∆Gp of 70 kJ/mol of ATP, the FT would increase from 0.86 to approx. 1. Therefore, the rate of this reaction would be clearly boosted by the increase of initial acetate concentration in the fermentation broth. This would theoretically decrease the ethanologenic potential of the microbial consortium, as also shown experimentally by El-gammal et al. . However, as the pH was anticipated to decrease during the course of the fermentation, the inhibition of the chain-elongating activity due to the low pH observed in this and other studies  was expected to play an important role in such enrichment conditions by preventing a significant activity.
Enrichment with acetate addition: ethanol yield and microbial community
Effect of enrichment conditions on microbial diversity
The non-metric multidimensional scaling (nMDS) analysis (Fig. 9b) illustrates the degree of microbial composition similarity between enrichment samples based on their relative distance. The results show that the samples from each enrichment are grouped together as a result of their higher similarity, which indicates that the enrichment cultures had reached a stable microbial composition at transfer T3. Comparing across enrichment conditions, it can be seen that enrichments HT6 and HT5.5YE, on one hand, and enrichments HT5YE, NT5YE and HT5YE-Ac on the other, developed closely related microbial communities, although a more widespread distribution can be observed in the latter group (Fig. 9b). In turn, the microbial consortium from enrichment HT5.5 was less related to other enrichments, probably due to the drastic drop in alpha diversity as compared to HT5.5YE. A statistically significant correlation was found between the initial pH conditions of each enrichment series and their microbial composition with a R2 corresponding to 0.90 (P value < 0.001), which indicates that the microbial composition found in the enrichment cultures was pH-dependent (Fig. 9b). Similarly, the ethanol and acetate yields were also found to be correlated with the ordination of the samples (with a R2 of 0.73 and 0.65, respectively, P value < 0.001) and followed a similar gradient direction with the pH (Fig. 9b). Thus, these results show that both the microbial composition and the yield of the main products were affected by the pH conditions of each enrichment series.
The enrichment strategies studied resulted in the successful selection of acetogenic bacteria from both untreated and heat-shock-treated anaerobic sludge, obtaining a number of enriched mixed microbial consortia with variable ethanologenic potential and microbial diversity as a function of the enrichment conditions applied. The composition of the microbial community was shown to shift rapidly along the enrichments, reaching a stable microbial composition dominated by the genus Clostridium in all cases, with a single dominant species in most of the enrichments. Both pH and nutrient supplements (YE) were found to be determinant operational parameters affecting the specific composition of the consortia and their microbial diversity. The ethanologenic potential of the enriched consortia was strongly dependent on the pH conditions applied, where an ethanol yield of 59.15 ± 0.18% of the stoichiometric maximum was achieved in pH-based enrichments at the lowest pH tested (pH 5). On the other hand, the addition of YE triggered the production of C4 compounds, opening the way for the production of MCFAs and higher alcohols. The thermodynamic approach used for the analysis of the metabolic network of reactions carried out by syngas-converting microbial consortia proved to be highly useful for assisting the design and interpretation of enrichment strategies. Based on the qualitative predictions of the thermodynamic analysis, it was possible to improve the product selectivity and enhance the maximum ethanol yield obtained in pH-based enrichments by 22.5% (72.44 ± 2.11% of the stoichiometric maximum) through an increase of the initial acetate concentration (enrichment HT5YE-Ac). Thus, this work demonstrated that a highly selective microbial activity towards the production of ethanol is possible using open-mixed microbial consortia. However, given the experimental observations made here, the ethanol yield obtained in enrichment HT5YE-Ac in batch mode cannot be extrapolated to processes in continuous mode as the long-term exposure of the enriched consortium to elevated ethanol and acetate concentrations would likely promote a high chain-elongating activity, lowering the product selectivity towards ethanol. Further work in this area is still needed to develop operational strategies able to control the chain-elongation reaction in syngas-converting microbial consortia.
AG performed the enrichment experiments, analyzed the data from enrichment experiments, carried out the thermodynamic analysis and wrote the manuscript. ML analyzed the microbial composition of enrichment samples, wrote the corresponding section and reviewed the manuscript. LL reviewed the microbial composition data and the manuscript. HNG and IVS guided the experimental design and interpretation of the data and reviewed the manuscript. All authors read and approved the final manuscript.
This work was financially supported by the Technical University of Denmark (DTU) and Innovation Foundation-DK in the frame of SYNFERON project.
The authors declare that they have no competing interests.
Availability of data
The datasets used and analyzed in the current study are included either in the main text of the article, in the additional files, or are available upon reasonable request to AG and ML. Raw sequences were submitted to NCBI Sequence Read Archive (SRA) database and are available under the project ID PRJNA439372 (BioSample accessions SAMN08765035-SAMN08765086).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- HLPE. Biofuels and food security. A report by the High Level Panel of Experts on Food Security and Nutrition of the Commitee on World Food Security. Rome, 2013.Google Scholar
- Havlík P, et al. Global land-use implications of first and second generation biofuel targets. Energy Policy. 2011;39:5690–702.View ArticleGoogle Scholar
- Rosegrant MW, Msangi S. Consensus and contention in the food-versus-fuel debate. Annu Rev Environ Resour. 2014;39:271–94.View ArticleGoogle Scholar
- Henstra AM, Sipma J, Rinzema A, Stams AJM. Microbiology of synthesis gas fermentation for biofuel production. Curr Opin Biotechnol. 2007;18:200–6.View ArticleGoogle Scholar
- Kennes D, Abubackar HN, Diaz M, Veiga MC, Kennes C. Bioethanol production from biomass: carbohydrate vs syngas fermentation. J Chem Technol Biotechnol. 2016;91:304–17.View ArticleGoogle Scholar
- Grimalt-Alemany A, Skiadas IV, Gavala HN. Syngas biomethanation: state-of-the-art review and perspectives. Biofuels Bioprod Biorefining. 2018;12:139–58.View ArticleGoogle Scholar
- Wainaina S, Horváth IS, Taherzadeh MJ. Biochemicals from food waste and recalcitrant biomass via syngas fermentation: a review. Bioresour Technol. 2018;248:113–21.View ArticlePubMedGoogle Scholar
- Abubackar HN, Veiga MC, Kennes C. Biological conversion of carbon monoxide to ethanol: effect of pH, gas pressure, reducing agent and yeast extract. Bioresour Technol. 2012;114:518–22.View ArticlePubMedGoogle Scholar
- Mohammadi M, Younesi H, Najafpour G, Mohamed AR. Sustainable ethanol fermentation from synthesis gas by Clostridium ljungdahlii in a continuous stirred tank bioreactor. J Chem Technol Biotechnol. 2012;87:837–43.View ArticleGoogle Scholar
- Kundiyana DK, Wilkins MR, Maddipati P, Huhnke RL. Effect of temperature, pH and buffer presence on ethanol production from synthesis gas by “Clostridium ragsdalei”. Bioresour Technol. 2011;102:5794–9.View ArticlePubMedGoogle Scholar
- Shen Y, Brown R, Wen Z. Syngas fermentation of Clostridium carboxidivorans P7 in a hollow fiber membrane biofilm reactor: evaluating the mass transfer coefficient and ethanol production performance. Biochem Eng J. 2014;85:21–9.View ArticleGoogle Scholar
- Xu D, Tree DR, Lewis RS. The effects of syngas impurities on syngas fermentation to liquid fuels. Biomass Bioenergy. 2011;35(7):2690–6.View ArticleGoogle Scholar
- Ramachandriya KD, et al. Critical factors affecting the integration of biomass gasification and syngas fermentation technology. AIMS Bioeng. 2016;3:188–210.View ArticleGoogle Scholar
- Kleerebezem R, van Loosdrecht MC. Mixed culture biotechnology for bioenergy production. Curr Opin Biotechnol. 2007;18:207–12.View ArticleGoogle Scholar
- Redl S, Diender M, Ølshøj T, Sousa DZ, Toftgaard A. Exploiting the potential of gas fermentation. Ind Crop Prod. 2017;106:21–30.View ArticleGoogle Scholar
- Liu Y, Wan J, Han S, Zhang S, Luo G. Selective conversion of carbon monoxide to hydrogen by anaerobic mixed culture. Bioresour Technol. 2016;202:1–7.View ArticlePubMedGoogle Scholar
- Navarro SS, Cimpoia R, Bruant G, Guiot SR. Biomethanation of syngas using anaerobic sludge: shift in the catabolic routes with the CO partial pressure increase. Front Microbiol. 2016;7:1–13.Google Scholar
- Vasudevan D, Richter H, Angenent LT. Upgrading dilute ethanol from syngas fermentation to n-caproate with reactor microbiomes. Bioresour Technol. 2014;151:378–82.View ArticleGoogle Scholar
- Ganigué R, Ramió-Pujol S, Sánchez P, Bañeras L. Conversion of sewage sludge to commodity chemicals via syngas fermentation. Water Sci Technol. 2015;72(3):415–20.View ArticlePubMedGoogle Scholar
- Woo C, Jung KA, Moon J. Biological carbon monoxide conversion to acetate production by mixed culture. Bioresour Technol. 2016;211:478–85.View ArticleGoogle Scholar
- Ganigué R, Sánchez-Paredes P, Bañeras L, Colprim J. Low fermentation pH is a trigger to alcohol production, but a killer to chain elongation. Front Microbiol. 2016;7(702):1–11.Google Scholar
- Steinbusch KJJ, Hamelers HVM, Buisman CJN. Alcohol production through volatile fatty acids reduction with hydrogen as electron donor by mixed cultures. Water Res. 2008;42:4059–66.View ArticlePubMedGoogle Scholar
- Liu K, Atiyeh HK, Stevenson BS, Tanner RS, Wilkins MR, Huhnke RL. Mixed culture syngas fermentation and conversion of carboxylic acids into alcohols. Bioresour Technol. 2014;152:337–46.View ArticlePubMedGoogle Scholar
- Singla A, Verma D, Lal B, Sarma PM. Enrichment and optimization of anaerobic bacterial mixed culture for conversion of syngas to ethanol. Bioresour Technol. 2014;172:41–9.View ArticlePubMedGoogle Scholar
- Wolin EA, Wolin MJ, Wolfe RS. Formation of methane by bacterial extracts. J Biol Chem. 1963;238(8):2882–6.PubMedGoogle Scholar
- Takahashi S, Tomita J, Nishioka K, Hisada T, Nishijima M. Development of a prokaryotic universal primer for simultaneous analysis of bacteria and archaea using next-generation sequencing. PLoS ONE. 2014;9(8):e105592.View ArticlePubMedPubMed CentralGoogle Scholar
- Martin M. Cutadapt removes adapter sequences from high-thoughput sequencing reads. EMBnet J. 2011;17(1):10–2.View ArticleGoogle Scholar
- Edgar RC. UNOISE2: improved error-correction for Illumina 16S and ITS amplicon sequencing. BioRxiv. 2016. https://doi.org/10.1101/003723.View ArticleGoogle Scholar
- Edgar RC. SINTAX: a simple non-Bayesian taxonomy classifier for 16S and ITS sequences. BioRxiv. 2016. https://doi.org/10.1101/074161.View ArticleGoogle Scholar
- Edgar RC. UNCROSS: filtering of high-frequency cross-talk in 16S amplicon reads. BioRxiv. 2016. https://doi.org/10.1101/088666.View ArticleGoogle Scholar
- Edgar RC. UNBIAS: an attempt to correct abundance bias in 16S sequencing, with limited success. BioRxiv. 2017. https://doi.org/10.1101/124149.View ArticleGoogle Scholar
- McMurdie PJ, Holmes S. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE. 2013;8(4):e61217.View ArticlePubMedPubMed CentralGoogle Scholar
- Dhariwal A, Chong J, Habib S, King IL, Agellon LB, Xia J. MicrobiomeAnalyst: a web-based tool for comprehensive statistical, visual and meta-analysis of microbiome data. Nucleic Acids Res. 2018;45:180–8.View ArticleGoogle Scholar
- Oksanen J et al. Vegan: community ecology package., R package version 2. 4-6. 2018.Google Scholar
- APHA. Standard methods for the examination of water and wastewater. Washington DC: American Public Health Association/American Water Works Association/Water Pollution Control Federation; 2005.Google Scholar
- Alberty RA. Thermodynamics of biochemical reactions. Hoboken: Wiley; 2003.View ArticleGoogle Scholar
- Alberty RA. Effect of temperature on standard transformed Gibbs energies of formation of reactants at specified pH and ionic strength and apparent equilibrium constants of biochemical reactions. J Phys Chem B. 2001;105:7865–70.View ArticleGoogle Scholar
- Amend JP, Shock EL. Energetics of overall metabolic reactions of thermophilic and hyperthermophilic Archaea and bacteria. FEMS Microbiol Rev. 2001;25(2):175–243.View ArticlePubMedGoogle Scholar
- Jin Q, Bethke CM. The thermodynamics and kinetics of microbial metabolism. Am J Sci. 2007;307:643–77.View ArticleGoogle Scholar
- Bertsch J, Müller V. Bioenergetic constraints for conversion of syngas to biofuels in acetogenic bacteria. Biotechnol Biofuels. 2015;8(210):1–12.Google Scholar
- Angenent LT, et al. Chain elongation with reactor microbiomes: open-culture biotechnology to produce biochemicals. Environ Sci Technol. 2016;50:2796–810.View ArticleGoogle Scholar
- González-Cabaleiro R, Lema JM, Rodríguez J, Kleerebezem R. Linking thermodynamics and kinetics to assess pathway reversibility in anaerobic bioprocesses. Energy Environ Sci. 2013;6:3780–9.View ArticleGoogle Scholar
- Jin Q. Energy conservation of anaerobic respiration. Am J Sci. 2012;312:573–628.View ArticleGoogle Scholar
- Jackson BE, McInerney MJ. Anaerobic microbial metabolism can proceed close to thermodynamic limits. Nature. 2002;415:454–6.View ArticlePubMedGoogle Scholar
- Varrone C, et al. Comparison of different strategies for selection/adaptation of mixed microbial cultures able to ferment crude glycerol derived from second-generation biodiesel. Hindawi. 2015;2015:1–14.Google Scholar
- Steinbusch KJJ, Arvaniti E, Hamelers HVM, Buisman CJN. Selective inhibition of methanogenesis to enhance ethanol and n-butyrate production through acetate reduction in mixed culture fermentation. Bioresour Technol. 2009;100(13):3261–7.View ArticlePubMedGoogle Scholar
- Phillips JR, Klasson KT, Clausen EC, Gaddy JL. Biological production of ethanol from coal synthesis gas. Appl Biochem Biotechnol. 1993;39–40:559–71.View ArticleGoogle Scholar
- Abubackar HN, Fernández-Naveira Á, Veiga MC, Kennes C. Impact of cyclic pH shifts on carbon monoxide fermentation to ethanol by Clostridium autoethanogenum. Fuel. 2016;178:56–62.View ArticleGoogle Scholar
- Gottwald M, Gottsehalk G. The internal pH of Clostridium acetobutylicum and its effect on the shift from acid to solvent formation Matthias. Arch Microbiol. 1985;143:42–6.View ArticleGoogle Scholar
- Huang LI, Forsberg CW, Gibbins LN. Influence of external pH and fermentation products on Clostridium acetobutylicum intracellular ph and cellular distribution of fermentation products. Appl Environ Microbiol. 1986;51:1230–4.PubMedPubMed CentralGoogle Scholar
- Grootscholten TIM, Steinbusch KJJ, Hamelers HVM, Buisman CJN. Chain elongation of acetate and ethanol in an upflow anaerobic filter for high rate MCFA production. Bioresour Technol. 2013;135:440–5.View ArticlePubMedGoogle Scholar
- El-gammal M, Abou-shanab R, Angelidaki I, Omar B. High efficient ethanol and VFA production from gas fermentation: effect of acetate, gas and inoculum microbial composition. Biomass Bioenergy. 2017;105:32–40.View ArticleGoogle Scholar
- Younesi H, Najafpour G, Mohamed AR. Ethanol and acetate production from synthesis gas via fermentation processes using anaerobic bacterium, Clostridium ljungdahlii. Biochem Eng J. 2005;27(2):110–9.View ArticleGoogle Scholar
- Esquivel-Elizondo S, Delgado AG, Rittmann BE, Brown RK. The effects of CO2 and H2 on CO metabolism by pure and mixed microbial cultures. Biotechnol Biofuels. 2017;10(220):1–13.Google Scholar
- Najafpour GD, Younesi H. Ethanol and acetate synthesis from waste gas using batch culture of Clostridium ljungdahlii. Enzym Microb Technol. 2006;38:223–8.View ArticleGoogle Scholar
- Abrini J, Naveau H, Nyns E. Clostridium autoethanogenum, sp. nov., an anaerobic bacterium that produces ethanol from carbon monoxide. Arch Microbiol. 1994;161:345–51.View ArticleGoogle Scholar
- Liou JS, Balkwill DL, Drake GR, Tanner RS. Clostridium carboxidivorans sp. nov., a solvent-producing clostridium isolated from an agricultural settling lagoon, and reclassification of the acetogen Clostridium scatologenes strain SL1 as Clostridium drakei sp. nov. Int J Syst Evol Microbiol. 2005;55:2085–91.View ArticlePubMedGoogle Scholar
- Ramió-pujol S, Ganigué R, Bañeras L, Colprim J. Incubation at 25 °C prevents acid crash and enhances alcohol production in Clostridium carboxidivorans P7. Bioresour Technol. 2015;192:296–303.View ArticlePubMedGoogle Scholar
- Zhang J, Taylor S, Wang Y. Effects of end products on fermentation profiles in Clostridium carboxidivorans P7 for syngas fermentation. Bioresour Technol. 2016;218:1055–63.View ArticlePubMedGoogle Scholar
- Singla A, et al. Optimization and molecular characterization of syngas fermenting anaerobic mixed microbial consortium TERI SA1. Int J Renew Energy Dev. 2017;6(3):241–51.View ArticleGoogle Scholar
- Hu P, Bowen SH, Lewis RS. A thermodynamic analysis of electron production during syngas fermentation. Bioresour Technol. 2011;102(17):8071–6.View ArticlePubMedGoogle Scholar
- Richter H, Martin ME, Angenent LT. A two-stage continuous fermentation system for conversion of syngas into ethanol. Energies. 2013;6(8):3987–4000.View ArticleGoogle Scholar
- Puig S, Ganigué R, Batlle-Vilanova P, Balaguer MD, Bañeras L. Tracking bio-hydrogen-mediated production of commodity chemicals from carbon dioxide and renewable electricity. Bioresour Technol. 2017;228:201–9.View ArticlePubMedGoogle Scholar
- Diender M, Stams AJM, Sousa DZ. Production of medium-chain fatty acids and higher alcohols by a synthetic co-culture grown on carbon monoxide or syngas. Biotechnol Biofuels. 2016;9(82):1–11.Google Scholar
- Xu S, Fu B, Zhang L, Liu H. Bioconversion of H2/CO2 by acetogen enriched cultures for acetate and ethanol production: the impact of pH. World J Microbiol Biotechnol. 2015;31:941–50.View ArticlePubMedGoogle Scholar