Eubacterium limosum ATCC 8486 was obtained from the Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). The strain was grown strictly under anaerobic conditions at 37°C in 100 mL of modified DSMZ 135 medium (pH 7.0), which comprised 1 g/L ammonium chloride, 2 g/L yeast extract, 10 g/L sodium bicarbonate, 0.1 g/L magnesium sulfate heptahydrate, 0.3 g/L cysteine-HCl, 10 mL vitamin solution (4 mg/L biotin, 4 mg/L folic acid, 20 mg/L pyridoxine-HCl, 10 mg/L thiamine-HCl, 10 mg/L riboflavin, 10 mg/L nicotinic acid, 10 mg/L pantothenate, 0.2 mg/L vitamin B12, 10 mg/L p-aminobenzoic acid, and 10 mg/L lipoic acid), 5.36 mM K₂HPO₄, 4.64 mM KH₂PO₄, 4 μM resazurin, and 20 mL trace element solution (1.0 g/L nitrilotriacetic acid, 3.0 g/L MgSO₄.7H₂O, 0.5 g/L MnSO₄.H₂O, 1.0 g/L NaCl, 0.1 g/L FeSO₄.7H₂O, 180 mg/L CoSO₄.7H₂O, 0.1 g/L CaCl₂.2H₂O, 180 mg/L ZnSO₄.7H₂O, 10 mg/L CuSO₄.5H₂O, 20 mg/L KAI(SO₄)₂.12H₂O, 10 mg/L H₃BO₃, 10 mg/L Na₂MO₄.2H₂O, 30 mg/L NiCl₂.6H₂O, 0.3 mg/L Na₂SeO₃.5 H₂O, 0.4 mg/L Na₂WO₄.2H₂O), at a pressure of 200 kPa and 50 mL of headspace filled with 0%, 20%, 40%, 60%, 80%, and 100% CO that is balanced using 100%, 80%, 60%, 40%, 20%, and 0% N₂, respectively, for autotrophic growth conditions. To enhance the autotrophic growth rate during adaptation, 40 mM NaCl was supplemented to the media, which couples with sodium dependent ATP synthase (Jeong et al., 2015; Song et al., 2017).

Adaptive laboratory evolution experiment was conducted with a syngas (44% CO, 22% CO₂, 2% H₂, and 32% N₂) described in Section “Bacterial Strains and Culture Conditions.” Before the ALE, the strain underwent preadaptation step by passing thrice through the syngas condition at mid-exponential phase. During the ALE experiment with four independent populations, the culture was transferred to a fresh medium at mid-exponential phase.

To construct whole genome resequencing libraries, the genomic DNA samples were extracted from the evolved populations and isolated single clone. Cell stocks were cultured in modified DSMZ 135 medium (pH 7.0) supplemented with glucose (5 g/L) and were incubated for 12 h at 37°C. The harvested cells were resuspended with 500 μL of lysis buffer containing Tris–HCl (pH 7.5), 5 M NaCl, 1 M MgCl₂, and 20% triton × 100. Thereafter, the cells were frozen using liquid nitrogen and ground using a mortar and pestle. The powder was resuspended with 600 μL of nuclei lysis solution (Promega, Madison, WI, United States), incubated at 80°C, and then cooled at 4°C. RNAs were removed from the cell lysate using RNase A solution. Proteins in the solution were precipitated using protein precipitation buffer (Promega). Following the protein precipitation, the samples were cooled at 4°C for 10 min, and were then centrifuged at 16,000 g for 10 min. The supernatant was transferred to a new tube, and 1 × volume of isopropanol was added. After centrifugation at 16,000 g for 5 min, the DNA pellet was obtained, which was then washed with 80% ethanol twice. The obtained DNA quality was determined by the A260/A280 ratio (>1.9) and inspected by gel electrophoresis, and the concentration was quantified via a Qubit 2.0 Fluorometer (Invitrogen, Carlsbad, CA, United States) with a Qubit dsDNA HS Assay kit (Invitrogen). The sequencing libraries were constructed using a TruSeq Nano DNA library prep kit (Illumina, La Jolla, CA, United States). The constructed libraries of evolved populations were sequenced with Illumina HiSeq2500 (rapid-run mode as 50 cycle-ended reaction) and the constructed libraries of isolated clone were sequenced with Illumina MiSeq (a 150 cycle-ended reaction).

All process for mutation screening were performed using a CLC genomics Workbench v6.5.1 (CLC bio, Aarhus, Denmark). Adapter sequences of the sequencing reads were removed by using a trimming tool with the default parameters (quality limit and ambiguous nucleotides residues 2). The resulting reads were mapped to the E. limosum ATCC 8486 reference genome (NCBI accession NZ_CP019962.1) with mapping parameters (mismatch cost: 2, indel cost: 3, deletion cost: 3, length fraction: 0.9, and similarity fraction: 0.9). Variants were detected from the mapped reads using a Quality-based variant detection tool with the parameters (neighborhood radius: 5, maximum gap and mismatch count: 5, minimum neighborhood quality: 30, minimum central quality: 30, minimum coverage: 10, minimum variant frequency: 10%, maximum expected alleles: 4, non-specific matches: ignore and genetic code: bacterial and plant plastid).

The single clones were isolated from the evolved populations by streaking the culture onto RCM agar medium. To confirm the sequence of each mutation site, the genomic regions were amplified by PCR using primer pairs and sequenced by Sanger sequencing (Supplementary Table S1). The selected single clones with the mutations were cultured in DSMZ 135 medium supplemented with CO in the headspace to measure the growth rate and metabolite production.

All primers used to construct plasmid in this study are listed in Supplementary Table S2. Initially, the pJIR750ai plasmid was used as a shuttle vector, and the chemically competent Escherichia coli DH5α (Enzynomics, Inc., South Korea) was used for cloning the plasmid. Acetolactate synthase (alsS) and acetolactate decarboxylase (alsD) were obtained via gene synthesis from Bacillus subtilis and Aeromonas hydrophila, respectively (Oliver et al., 2013). The synthesized alsS and alsD were amplified using primer sets alsS_F-alsS_R and alsD_F-alsD_R, respectively. The pJIR750ai reduced by PvuI (named as pJIR750_PvuI_cut) was digested by BamHI and SalI, and was then assembled with the amplified alsD by using In-Fusion HD cloning Kit (TaKaRa, Japan). Subsequently, the assembled plasmid (pJIR750_alsD) was linearized by SacI and BamHI, and was then assembled with the amplified alsS by In-Fusion cloning Kit, generating the pJIR750_alsS_alsD plasmid. To control the gene expressions, promoters of ELIM_c2885 (pyruvate:ferredoxin oxidoreductase) and ELIM_c1121 ([Fe] hydrogenase) were selected as the two genes were constitutively expressed with high expression levels in E. limosum (Song et al., 2018). The native promoters were amplified from genomic DNA of E. limosum and inserted in the pJIR750_alsS_alsD plasmid, resulting in the construction of pJIR750_alsS_U_1121_P1121_P2885_U1121_alsD plasmid.

To prepare the electrocompetent strains, a previously modified protocol was used (Shin et al., 2019). The cells were cultured in 100 mL of DSM 135 medium supplemented with 5 g/L glucose. At the early exponential phase (OD₆₀₀ 0.3 ∼ 0.5), the cells were harvested by centrifuging at 10,000 rpm for 10 min at 4°C. The harvested cells were washed with 50 mL of 270 mM sucrose buffer (pH 6) and resuspended to achieve a final concentration of 10¹¹ cells/mL. About 1.5 ∼ 2 μg plasmid was added to the electrocompetent cells and then the solution was transferred to a 0.1-cm-gap Gene Pulser cuvette (Bio-Rad, Hercules, CA, United States). Thereafter, the cells were pulsed at 2.0 kV and immediately resuspended with 0.9 mL of reinforced clostridial medium (RCM). The cells were recovered on ice for 5 min, and incubated at 37°C for 16 h. The recovered cells were plated on an RCM plate (1.5% agar) containing 15 μg/mL thiamphenicol. A single colony was selected and cultured in DSM 135 medium supplemented with 5 g/L glucose.

Primary metabolites were measured via high performance liquid chromatography (Waters, Milford, MA, United States) equipped with refractive index detector and MetaCarb 87 H 300 × 7.8 mm column (Agilent, Santa Clara, CA, United States). The mobile phase used was 0.007 N sulfuric acid solution with 0.6 mL/min flow rate. The oven temperature was set at 37°C for acetate, lactate and butyrate and 50°C for acetoin.

CO and CO₂ concentrations were measured via gas chromatography (Shimadzu, Japan) equipped with thermal conductivity detector and ShinCarbon ST Micropacked column (1 mm × 2 m, 1/16′′, 100/120 mesh; Restek, Bellefonte, PA, United States). Helium was used as the carrier gas at a flow rate of 30 mL/min. The initial oven temperature was 30°C for 1 min, programmed with a ratio 5°C/min until it reached 100°C. The temperature for injector and detector was 100°C.

To confirm the CO tolerance of E. limosum ATCC 8486, the cell growth was determined by culturing the strain in 100 mL of the modified DSMZ 135 medium with 0%, 20%, 40%, 60%, 80%, and 100% CO concentrations (Figure 1A). In the absence of CO, E. limosum proliferated to a maximum optical density at 600 nm of 0.062 ± 0.003, which likely due to the presence of sodium bicarbonate and yeast extract in the medium served as the carbon source for the E. limosum (Figure 1B). Under 20% CO growth condition, the cell reached a growth rate of 0.063 ± 0.011 h^–1; whereas, a growth rate of 40% CO growth condition was 0.035 ± 0.002 h^–1, which increased by 1.80 folds. In contrast, maximum optical densities (600 nm) of 20% CO and 40% CO culture conditions were 0.193 and 0.438, respectively, which decreased by 2.27 folds (Figure 1B). The cell cultivated under CO concentrations exceeding 60% demonstrated insignificant proliferation under the conditions, indicating that cell growth was inhibited with increasing CO concentrations in the growth medium. We performed an additional experiment to confirm the effect of CO on the cell growth. Initially, the cells were cultured in the DSMZ 135 medium containing 5 g/L glucose, then incubated until reaching the mid-exponential phase, optical density (600 nm) of 1.320. When the cell reached the mid-exponential phase, 0% (100% N₂, as control), 20%, 40%, 60%, 80%, or 100% CO was purged to each sample with the same pressure, then measured the optical density (600 nm) of the cells (Supplementary Figure S1). Two hours after the CO injection, the optical densities (600 nm) were measured, which decreased compared to the control at all CO concentrations (Supplementary Figure S1).

To further understand the phenotypical effect of CO, metabolites produced by E. limosum were investigated (Figure 1C). Among the identified metabolites, acetate was the dominant metabolic product, which is known as a key end product for acetogens under autotrophic growth conditions. In general, acetate production is well correlated with the cell growth, because acetate synthesis generates ATP that is required for cellular function of acetogens. Under 20% and 40% CO conditions, the acetate production was increased compared to the other CO conditions; moreover, the acetate production patterns revealed growth-dependent profiles. Under high CO conditions, acetate production decreased as the concentrations of CO increased, and insignificant changes were determined (P-value > 0.05) compared to the control experiment. Based on these results, the acetate productions were correlated with the growth pattern, which was influenced by the amount of CO concentration in the culture medium, suggesting that the increase of CO tolerance potentially enhances E. limosum acetate production.

To improve the CO tolerance of E. limosum, we applied the robust ALE method to the organism, which is used as a tool to engineer organism for overcoming target stress conditions. For designing the ALE experiment, establishing an appropriate stress condition is an essential factor for an organism to reach the desired phenotype. In this study, syngas was selected for the ALE experiment. Syngas is generated from gasification, and the composition of the gas is determined by a source of gasifier type or biomass, which typically comprises CO and H₂ with 14% ∼ 67% and 5% ∼ 32%, respectively (Munasinghe and Khanal, 2010). To decide the precise composition of syngas for ALE, previous studies on syngas fermentations of acetogens were investigated, and a syngas composition of 44% CO, 22% CO₂, 2% H₂, and balanced N₂ was selected, which was widely observed in the syngas generated by industries (Köpke et al., 2011). Using the syngas composition, E. limosum was cultivated in the same basal media, which was utilized for determining the CO tolerance, to determine the transfer point for ALE (Figure 2A). The growth rate under the condition was 0.070 ± 0.002 h^–1 with a maximum optical density (600 nm) of 0.486 ± 0.021, which slightly differs compared to the growth rate (0.058 ± 0.000 h^–1) obtained from the 40% CO condition, because a presence of H₂ with CO in the environment enhances the autotrophic growth of acetogens (Bertsch and Müller, 2015; Valgepea et al., 2018). According to the growth profile, the mid-exponential phase is between 42 and 54 h after the initial inoculation; thus, we selected 48 h for the transfer point for ALE. For ALE, four separate populations of E. limosum were adapted to confirm the reproducibility, labeling as ALE1, 2, 3, and 4. Initially, at the 40^th generation, the growth rates of all populations were increased to 0.085 h^–1, which then maintained the enhanced growth rate up to 120^th generation (Figure 2B). After 120 generations of adaptation, slight variations in the growth rates were observed in the entire population, with growth rate around 0.086 h^–1; however, no growth rate changes were observed after 150 generations (Figure 2B); thus, the ALE was stopped at 150^th generation.

To further investigate the changes in cell growth at the population level, we selected the evolved population (ECO), ALE4, which demonstrated the highest growth rate with 0.089 h^–1 at the 150^th generation. Moreover, the other growth rates were comparable to that of the ECO, which were 0.086, 0.087, and 0.088 h^–1 for the ALE1, ALE2, and ALE3, respectively. Growth and CO consumption profiles of the ECO and the parental strain were compared under the syngas growth conditions (Figure 2C). The two strains completely consumed CO presented in the headspace by 84 and 60 h, respectively, resulting in CO consumption rates of 0.043 ± 0.019 and 0.058 ± 0.003 mmol h^–1, respectively, which increased the carbon consumption rate of the adapted strain by 1.35 folds (Figure 2C). In addition, the stationary phases of the parental strain and the ECO were attained at 72 and 48 h, respectively, and the maximum optical densities (600 nm) were 0.498 ± 0.028 and 0.639 ± 0.016, respectively, which increased by 1.28 folds (Figure 2C).

Following the investigation of its phenotypical changes under the syngas condition, CO tolerance of the ECO was measured by cultivating the strain under different CO concentrations ranging from 0% to 100% CO (Figure 2D). Under 20% and 40% CO conditions, the growth rates were 0.076 ± 0.001 and 0.089 ± 0.001 h^–1, respectively. Compared to the parental strain, the growth rates of the ECO under 20% and 40% CO conditions were enhanced by 1.21 and 1.65 folds higher. In addition, the growth rates of the inhibitory CO conditions were 0.069 ± 0.001, 0.056 ± 0.001, and 0.048 ± 0.001 h^–1 under 60%, 80%, and 100% CO conditions, respectively (Figure 2D). In addition, the maximum optical density at 600 nm of the ECO were higher than that of the parental strain by 7.49 folds under 60% CO condition (Figure 2E). Acetate productions of the ECO were identified under the given conditions, resulting that 0.274 ± 0.060, 3.835 ± 0.215, 6.819 ± 0.457, 8.311 ± 0.254, 0.866 ± 0.217, and 0.264 ± 0.043 mM of acetate were produced under 0%, 20%, 40%, 60%, 80%, and 100% CO conditions (Figure 2F). In general, the acetate productions of the ECO were growth dependent, similar to those of the parental strain; however, they were increased compared to the production by the parental strain. Such results indicate that the adaptively evolved E. limosum enhanced CO consumption, growth, and tolerance under the autotrophic conditions, thus questioning about genetic mutations causing the phenotypic changes.

Adaptive laboratory evolution allows a target organism to adapt to the desired condition by modifying the genotype, rewiring and changing the metabolic pathways, and altering enzyme kinetics in the organism. Identification of mutations in the genome throughout the ALE is important to understand the fitness landscape of the organism under the growth condition. Although the ALE revealed that the fitness, tolerance, and acetate production were increased under the 44% CO condition, the causal genotypic changes and their collateral effects on the phenotype remain unclear. To address this, the four evolved populations at initial, 50^th, 100^th, and 150^th generation were proceeded for whole genome resequencing. We identified 39 mutations in all adapted strains, locating 33 and 6 mutations in the genic and intergenic regions, respectively (Supplementary Table S3).

Among these, five common mutations were identified across the populations, resulting in five key mutations with top 15% mutation frequency in the evolved populations. The five mutations were located in ELIM_c1038, ELIM_c1073, and ELIM_c1653 as single nucleotide variations (SNVs) and ELIM_c1031 and ELIM_c1654 as nucleotide insertion mutations (Table 1). For SNVs, E⁴⁸K in ELIM_c1038, Y¹³⁶X in ELIM_c1073, and A⁹⁷E in ELIM_c1653 were identified, which are responsible for putative ATPase, DNA methylase (dam), and CODH catalytic subunit (acsA), respectively. For the insertional mutations, N¹¹⁹KfsX133 in ELIM_c1031 encoding integrase protein and A⁷²AfsX92 in ELIM_c1654 encoding CODH accessory protein (cooC2) were observed. Of the five mutated genes, two mutations were located in the putative genes, and interestingly the other three mutations were located in the genes with certain functional roles, of which two mutations occurred in CODH/ACS complex coding genes, acsA and cooC2, that were reported to play a pivotal role for the active site and maturity of the complex (Morton et al., 1991; Kerby et al., 1997; Bender et al., 2011). The mutation in acsA was not located at the activate sites of the enzyme; however, substituting a small non-polar amino acid into a large polar amino acid potentially altered the protein structure that affects the enzymatic activity. The other mutation, which occurred in cooC2, introduced an early stop codon at 20^th amino acid downstream of the mutation site, which revealed synonymous substitution at the mutation site. The early termination often leads to a loss of function, suggesting that, despite the importance of cooC under the autotrophic growth condition, the functional role of the gene in the evolved strain may be ineffective. Collectively, the five mutations were the dominant variants, and three mutations were located in the genes with functional roles, of which acsA was hypothesized to be the driving mutation for the altered phenotype in adapted strain under autotrophic condition.

In order to validate the hypothesis, obtaining a strain from ECO with the mutation on acsA is essential. For obtaining a single clone with the acsA mutation from the evolved populations, 20 colonies were isolated, which then confirmed the presence of acsA mutation in 17 of 20 colonies. Of the obtained 17 colonies, a single clone with the acsA mutation and without the other key mutations was selected, labeling as ECO_acsA strain (Supplementary Figure S2). The obtained strain, ECO_acsA, underwent genome resequencing to identify the mutations embedded in the genome, resulting in six mutations with four in the genic regions and two in the intergenic regions (Supplementary Table S4). The four mutations were associated with ELIM_c0006, ELIM_c1653, ELIM_c2214, and ELIM_c2227. Of the four mutations, only ELIM_c1653, which encodes acsA, is associated with the autotrophic growth condition.

Initially, growth profile of the strain under syngas condition with 44% CO was measured and compared to the parental strain, resulting in a growth rate of 0.095 ± 0.000 h^–1 and 0.050 ± 0.001 h^–1 and a maximum optical density (600 nm) of 0.703 ± 0.023 and 0.498 ± 0.028 for ECO_acsA and parental strain, respectively (Figure 3A). These results indicate that ECO_acsA strain proliferates more rapidly by 1.90 folds with higher optical densities (600 nm) by 1.41 folds than the parental strain under CO condition.

To further understand the phenotypic changes, CO consumed and metabolites produced by the ECO_acsA strain were measured under the syngas condition with 44% CO. In accordance to the previous analysis on metabolites, acetate was the dominant end product produced by the ECO_acsA strain, with 6.889 mM. Increase in the acetate production by the strain indicates that the altered C-cluster of the active site of CODH subunit encoded by the mutated acsA enhanced CO utilization (Figure 3A). Consistent with acetate production, the CO consumption rates by ECO_acsA and parental strain were 0.059 ± 0.002 and 0.043 ± 0.019 mmol h^–1, respectively, which differed by 1.37 folds (Figure 3A).

Although the growth rate under CO condition was enhanced, the increase in CO tolerance of ECO_acsA strain compared to the parental strain remains unclear. To measure the CO tolerance, the strain was cultured under different CO concentrations. The growth rates of ECO_acsA strain under 20% and 40% CO conditions were 0.070 ± 0.001 and 0.079 ± 0.001 h^–1, and these decreased for the strains cultured under high CO conditions with 0.059 ± 0.001, 0.043 ± 0.002, and 0.040 ± 0.001 h^–1 for 60%, 80%, and 100% CO, respectively (Figure 3B). Although the growth rates decreased with increasing CO concentrations, compared to the parental strain, the growth rates of the ECO_acsA strain were increased by 2.25 folds under 40% CO, and the maximum optical density (600 nm) under all CO concentrations (Supplementary Figure S3), representing CO tolerance of the strain was greatly enhanced. Based on these results, our hypothesis, in which the mutation on acsA affects phenotype of the E. limosum under CO growth conditions, was tested through growth profiling and metabolite measurement experiments. We conclude that, indeed, the speculation was correct with enhanced growth rate, CO consumption rate, acetate production, and CO tolerance under CO growth conditions, suggesting that mutation on acsA is essential to engineer the strain for CO utilization.

The ECO_acsA strain demonstrated the enhanced CO consumption and growth rates, thereby querying whether the evolved strain is a better platform to produce biochemical than the parental strain. To investigate the strain capacity to produce biochemicals, acetoin was selected as a target chemical, which is widely utilized in various industries from cosmetics, food flavoring, and pharmaceuticals (Werpy and Petersen, 2004; Bao et al., 2015). Interestingly, in E. limosum, the acetoin biosynthetic pathway coding genes are located, but the production was not detected under CO and other conditions, such as glucose and H₂/CO₂ conditions (Song et al., 2017, 2018). In the presence of the genes, E. limosum is capable of synthesizing acetoin; however, the insignificant transcriptional and translational expressions under the heterotrophic and autotrophic conditions prevent the production, according to a previous study (Song et al., 2018). Thus, activation of the biosynthetic pathway coding genes using novel bio-parts potentially produce acetoin in E. limosum, and then determine whether ECO_acsA is a superior platform to produce the biochemical compounds.

To produce acetoin, a plasmid with alsS, which encodes α-acetolactate synthase that condenses two molecules of pyruvate to one acetolactate, and alsD, which encodes acetolactate decarboxylase that converts acetolactate to acetoin, was constructed (Figure 4A and see “Materials and methods” for more details). For activating the biosynthesis, bio-parts associated with highly expressed genes that are functionally related to CO metabolism were selected, locating promoters of ELIM_c2885 encoding pyruvate:ferredoxin oxidoreductase and ELIM_c1121 encoding hydrogenase subunit for alsS and alsD, respectively, with 5′ untranslated region that was obtained from ELIM_c1121 (Figure 4A; Song et al., 2018). Following the construction, the plasmid was transformed into the parental strain, which was then cultured for biological triplicates under the syngas condition with 44% CO, proliferating with cell density of 0.053 ± 0.002 gDW and producing acetoin of 14.6 ± 0.8 mM/gDW, along with 23.3 ± 1.5 mM/gDW acetate, 4.0 ± 0.0 mM/gDW butyrate, and 6.2 ± 1.7 mM/gDW lactate, which recovered 96.2% of carbons (Figure 4B). After confirming acetoin production, the same plasmid was introduced into ECO_acsA strain, and was then cultured under similar CO condition; thereafter, the acetoin production was measured, resulting in 19.6 ± 1.3 mM/gDW, which significantly increased by 1.34 folds (P-value ≤ 0.015). In addition, ECO_acsA strain with cell density of 0.034 ± 0.001 gDW produced 23.6 ± 1.1 mM/gDW acetate, 6.4 ± 0.2 mM/gDW butyrate, and 0.7 ± 0.5 mM/gDW lactate, which recovering 86.7% of carbons. Regarding CO consumption, the WT_ACT strain consumed with 429.8 ± 131.6 mM/gDW and ECO_acsA_ACT strain consumed 642.7 ± 317.8 mM/gDW, indicating higher CO consumption value by ECO_acsA_ACT strain compared to that by WT_ACT strain. Less consumption of CO leads to needing for the reduced ferredoxin, which is essential for E. limosum under autotrophic growth condition by generating a chemiosmosis gradient that synthesizes ATP. Depletion of reduced ferredoxin can be replenished by producing lactate from pyruvate via lactate dehydrogenase reaction, which oxidizes NADH and reduces ferredoxin. Based on the assumption, compared to ECO_acsA_ACT strain, the WT_ACT strain consumed less CO that caused less available reduced ferredoxin. To overcome the needs, the WT_ACT strain produced more lactate by catalyzing lactate dehydrogenase that produces reduced ferredoxin. Whereas, with the increased CO consumption capacity and more reduced ferredoxin available, ECO_acsA_ACT strain does not need to catalyze lactate dehydrogenase and alter the carbon flux to produce acetoin. Therefore, the authors hypothesize that the ECO_acsA strain produced significantly enhanced the amount of acetoin, using increased capacity with increased CO consumption and CO tolerance. Further studies on metabolic engineering are needed to verify the effects of the mutations on the change of the production by the acetogens.