Expression of amplified synthetic ethanol pathway integrated using Tn7-tool and powered at the expense of eliminated pta, ack, spo0A and spo0J during continuous syngas or CO2/H2 blend fermentation

Authors


Correspondence

Michael Tyurin, Syngas Biofuels Energy, Inc., 2441 Del Monte, Houston, TX 77019, USA. E-mail: michael@syngasbiofuelsenergy.com

Abstract

Aims

To engineer acetogen biocatalyst selectively overproducing ethanol from synthesis gas or CO2/H2 as the only liquid carbonaceous product.

Methods and Results

Ethanol-resistant mutant originally capable of producing only acetate from CO2/CO was engineered to eliminate acetate production and spore formation using our proprietary Cre-lox66/lox71-system. Bi-functional aldehyde/alcohol dehydrogenase was inserted into the chromosome of the engineered mutant using Tn7-based approach. Recombinants with three or six copies of the inserted gene produced 525 mmol l−1 and 1018 mmol l−1 of ethanol, respectively, in five independent single-step fermentation runs 25 days each (P < 0·005) in five independent repeats using syngas blend 60% CO and 40% H2. Ethanol production was 64% if only CO2 + H2 blend was used compared with syngas blend (P < 0·005).

Conclusions

Elimination of genes unnecessary for syngas fermentation can boost artificial integrated pathway performance.

Significance and Impact of the Study

Cell energy released via elimination of phosphotransacetylase, acetate kinase and early-stage sporulation genes boosted ethanol production. Deletion of sporulation genes added theft-proof feature to the engineered biocatalyst. Production of ethanol from CO2/H2 blend might be utilized as a tool to mitigate global warming proportional to CO2 fermentation scale.

Introduction

Sources of inorganic carbon have become attractive targets for microbial biocatalysis (Fischer et al. 2008; Demain 2009; Köpke et al. 2011). Acetogens are known to reduce inorganic carbon of carbon monoxide (CO) and carbon dioxide (CO2) to organic compounds with two or more carbons (Gaddy et al. 2001; Fischer et al. 2008; Köpke et al. 2011; Abubackar et al. 2012).

Carbonaceous materials for the production of synthesis gas (syngas – blend of CO, CO2 and H2) are various kinds of coal, acid-hydrolysed lignocellulosic biomass or natural gas available at c. $2·50 per 1 GJ in the United States and some other countries. CO and CO2 can be selectively fermented to specialty, commodity chemicals, food components or fuels using metabolically engineered biocatalysts with high product selectivity (Berzin et al. 2012a) as opposed to chemical syngas catalysis offering low product selectivity (Yang and Parr 1985). Engineered biocatalysts can ferment CO2-rich syngas or captured CO2 with zero CO2 process vent gas emissions to the target product when inexpensive H2 source is available (water hydrolysis using electricity produced by DOE recommended high-efficiency solar panels in situ) (Berzin et al. 2012a).

Ethanol as a functional gasoline extender replaces up to 10% in commercial gasoline blends. Despite the existence of scaling up start-ups with syngas fermentation to mixtures of liquid carbohydrates using described processes (Gaddy et al. 2001; Paul et al. 2010), there is no commercial manufacture of ethanol fuel from syngas. Problems associated with the use of existing naturally occurring biocatalysis for ethanol manufacture from syngas are its concentration below 1·5% in fermentation broth and co-production of other liquid organic components along with CO2 emission in single-step continuous fermentations as a part of the process vent gas.

As no known commercial biocatalysis utilizes naturally occurring strains, metabolic engineering renders new synthetic pathways stably functional in engineered biocatalyst micro-organisms where the proper management of cell energy pool has been elaborated (Berzin et al. 2012a,b).

We have used lox66/lox71-Cre-based system to eliminate phosphotransacetylase (pta) in Clostridium sp. MT1243. The accumulated data supported our hypothesis that elimination of some genes in acetogens not essential for a particular process releases the cell energy previously used for replication, expression and other maintenance needs of the eliminated genes to make available to power artificial pathways.

In this report, we describe elimination of pta and acetate kinase (ack) to direct the released acetyl-CoA carbon flow to ethanol using engineered amplified ethanol biosynthesis pathway powered at the expense of the cell energy made available due to elimination of both acetate biosynthesis genes. The loss of the ATP fraction rendered via eliminated acetate from acetyl-CoA pathway did not affect performance of the engineered biocatalyst overproducing ethanol. Additional boost of the engineered cell energy pool towards ethanol path was achieved via elimination of early sporulation spo0A and spo0J genes using the same approach (Berzin et al. 2012b).

The amplified synthetic ethanol pathway was integrated using optimized for acetogens Tn7-based system for insertion of the predetermined copy number of bi-functional aldehyde/alcohol dehydrogenase (al-adh) into created de novo Tn7 integration site (Craig 1991). The site was created in the chromosome of the ethanol-resistant mutant strain Clostridium sp. MT1243 during the step of ack elimination.

The resulted engineered biocatalysts produced 490 or 928 mmol l−1 ethanol in the fermentation broth in single-step continuous fermentation if three or six copies of al-adh were inserted, respectively. Elimination of early spore formation genes spo0A and spo0J made the engineered biocatalysts partially theft-proof to prevent unauthorized use and improve ethanol yields in recombinants carrying three or six copies of integrated al-adh, respectively.

The technology described does not intend to compete with the existing commercial technologies for fuel ethanol production. Our technology complements the existing energy production/recovery technologies emitting CO2 as the process waste, as our process might use CO2 as the raw material carbon source. The engineered biocatalysts were capable of CO2 fermentation in the presence of hydrogen and rendered about 64% of the fermentation performance for the syngas blend under conditions comparable to those of syngas fermentation. The technology opens new frontiers for unlimited scaling up of CO2 biocatalysis directly and selectively to valuable chemicals and food components as an alternative to photosynthetic pathway of inorganic carbon reduction to carbohydrates. The technology will become the crucial technological approach to CO2 reduction in the atmosphere in the immediate future because the planet already experiences global warming with growing consequences getting out of control.

Materials and methods

Bacterial strain and antibiotics

All manipulations with cell, media/solutions, glass and plastic laboratory ware for microbial cultivation were performed under syngas blend (v/v 60% CO + 40% H2) in customized Anaerobe Chamber AS-580 (Anaerobe Systems, Morgan Hill, CA, USA) with an extra section and neoprene sleeves and gloves added to maintain anaerobic conditions with <1 ppm of oxygen. The syngas blend composition was elaborated based on reported earlier empirical stoichiometry of the process (Berzin and Tyurin 2012; Berzin et al. 2012a; Tyurin et al. 2012). It was revealed that each mol of CO2 produced during reduction of CO by carbon monoxide dehydrogenase in continuous syngas fermentations required four moles of hydrogen to ensure reduction of all produced CO2 to organic carbon. That molar gas ratio has been used in continuous fermentation tests if only CO2/H2 blend was used.

Wild-type Clostridium sp. MT1230 was isolated from the oil well flooding water in Saratov region (Russian Federation) in 2008. The Gram (+) strictly anaerobic catalase (−) spore-forming rod utilized CO/CO2 as the carbon source. The purified culture was grown in liquid (SFB) or on solidified SFB with 1·2% agar (SFA) medium (Tanner et al. 1993). Wheaton serum bottles with capped full-size butyl rubber stoppers were used for liquid cultures and 100-mm Petri dishes for plating. Plates were incubated in vented anaerobic Vacu-Quick Jars (Almore International, Inc., Portland, OR, USA). The syngas pressurized at 1·76 kg cm−2 in the bottles and 0·42 kg cm−2 in the jars was changed every 6 h to ensure proper microbial feed. The incubation temperature 36°C was used. Clostridium sp. MT1230 was naturally tolerant to 650 mmol l−1 ethanol in SFB during single-step continuous syngas fermentation.

Clostridium sp. MT1230 was UV-mutated resulting in selection of a mutant tolerating 1·5 mol l−1 ethanol in SFB with the same 48- to 50-min cell duplication time during syngas continuous fermentation as the wild type (data not shown). For mutagenesis (Bates et al. 1989), cell suspensions (c. 9 × 109 cells in 5 ml of SFB) were subjected to 254 nmol l−1 UV light (9 W tube) from the 5 cm distance over the Petri dish with the 18-h-old broth culture at the cell killing rate 99% after 65-s exposure. The resulted ethanol-resistant UV-mutant tolerating 1·5 mol l−1 ethanol was named Clostridium sp. MT1243.

Recombinant acetogens were grown in SFB or on SFA containing Chloramphenicol (Cm) (Sigma, St. Louis, MO, USA) 35 μg ml−1 or without the antibiotic. Certain DNA accumulation was performed in Escherichia coli JM109 [New England BioLabs, Ipswich, MA, USA (NEB, MA)] using Brain Heart Infusion (Becton Dickinson, Franklin Lakes, NJ, USA) in the presence of 40 μg ml−1 Cm.

Continuous syngas fermentation

BioFlo 2000 (2·5 l) vertical bioreactors (New Brunswick Scientific, Edison, NJ, USA) with two Rushton impellers 3·5 cm apart and 2·5 cm from the bottom of the vessel were used for single-stage continuous syngas fermentation. Ten centimetre flat dick Pall spargers with pore size 0.5 μm (Pall Corporation, Jersey Village, TX, USA) covered bottoms of each vessel. Sterile filtered syngas blend as above was used in all experiments. The impeller speed was set at 250 rpm. The fermentation broth volume was maintained at 1·8 l at pH 6·4 ± 0·1 adjusting it with autoclaved 20% NaOH. Each bioreactor with 1·5 l of the prereduced sterile SFB was purged with syngas for 24 h with the pH and temperature controls set P-I-D until resazurin changed colour, indicating the state of anaerobiosis was reached.

Each bioreactor was inoculated with 250 ml of overnight seed culture with OD600 3·65 ± 0·15 either of ethanol-resistant UV-mutant Clostridium sp. MT1243 or of its recombinants. The seed batch cultures were obtained from BioFlo 2000 bioreactors. Inoculated bioreactors for continuous syngas fermentations were kept running with no liquid flow and syngas flow set 25 ml min1 until the OD600 reached 6·60 ± 0·15 (the zero time point at the fermentation plots). At that point, the liquid flow was gradually increased from 0 to 3 ml min1 at maintaining the OD600 6·60 ± 0·16 at the gas flow constant for the 25-day-long runs. Each bioreactor was gravity fed with the sterile prereduced SFB in two connected in parallel 38-l vented bottles per each bioreactor. The waste culture broths were gravity flown to two similar sterile bottles. Vents of all bottles were connected to the reservoir with oxygen-free nitrogen maintained at normal pressure to balance the change in liquid volumes. Waste culture broths were sampled every 15 min to monitor the OD600, detection of ethanol, and acetate using HPLC and also for collecting fresh cells for DNA extraction and electrotransformation experiments.

Vent of each bioreactor was connected via a sterile filter to the carousel mechanism distributing samples from various bioreactors to the Portable NDIR Syngas Analyzer Gasboard 3100P (Wuhan Cubic Optoelectronics CO., Ltd, China) (http://www.gassensor.com.cn/product_detail_cn/typeid/3/id/10.html) via 6·3-cm copper line. The line and the carousel mechanism were purged with oxygen-free sterile nitrogen after each sampling. Vent gas components CO, CO2 and H2 were analysed in real-time sampled every 15 min (96 data points daily) for each bioreactor. The data were collected in five independent 25-day-long continuous syngas fermentation runs in five repeats each and analysed based on significance of differences between means.

In continuous fermentation experiments, syngas blend (v/v 60% CO + 40% H2) was used. In addition to that, the process efficiency by the recombinant with six copies of integrated al-adh (with or without early sporulation genes) was additionally evaluated using v/v 20% CO2 + 80% H2 blend to determine whether CO2 alone can be used as the carbon source for the engineered biocatalysts.

Electrotransformation cuvettes, electrodes and electrotransformation

Benchtop Centrifuge EBA™ 21 (Cardinal Health, Dublin, OH, USA) was placed inside anaerobic chamber to concentrate cells under anaerobic conditions from the waste culture broth flow (see continuous fermentation section) collected using sterile prereduced syringes aseptically and to preserve anoxic conditions for the samples and the fermentation set-ups. Samples (10 ml each) were collected to sterile sealed purged with nitrogen, chilled on ice serum bottles and transferred to anaerobic chamber. There the cells were transferred to 15-ml disposable polypropylene centrifuge tubes, chilled on ice and concentrated via centrifugation. Resulting pellets were washed once with sterile cold 0·1 mol l−1 prereduced sucrose and resuspended in the same solution to make 0·1 ml from each originally 10-ml sample of the waste culture broth.

Syngas Biofuels Energy, Inc. electrotransformation generator and their proprietary nickel-titanium alloy flat parallel electrodes were used for electric treatment with single 10 ms square pulses at 6250 V with digital recording of the pulse current as described (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin and Tyurin 2012; Tyurin et al. 2012). The cell viability was monitored by counting colonies grown on SFA inoculated using aliquots of decimal sample dilutions (Berzin and Tyurin 2012).

DNA isolation

DNA isolation from acetogens was performed as described by Tyurin et al. (2012) with modifications. Cell concentration, cold wash with sterile 0.1 mol l−1 sucrose in HPLC-grade water, treatment with 5 mg ml1 lysozyme and SDS lysis were performed in anaerobic chamber to preserve the DNA preps from the damage with the products of membrane oxidation.

All subsequent steps were performed on the bench: RNA digestion, Proteinase K treatment to decrease the viscosity of the aqueous phase and deactivate endo- and exo-nuclease activities, deproteinization by Tris-equilibrated (pH 8·0) phenol/chloroform/isoamyl alcohol mixture (25:24:1 v/v/v) in the presence of 1 mol l−1 NaCl and precipitation of DNA with three volumes of ethanol in 0.1 mol l−1 sodium acetate (pH 5·2).

All the enzymes and chemicals were purchased from Sigma, MO. Upon the extraction, the DNA from Clostridium sp. MT1243 and its recombinants was additionally purified using Qiagen Gel Extraction Kit (Valencia, CA, USA) to decrease degradation with endogenous endo- and exo-nucleases.

Below, all the gene elimination steps are listed for cat, ack, pta, spo0A and spo0J in the order they were performed over the time of this 36-month-long project. Partial documentation of the results presented is provided and discussed to reduce the volume of the manuscript

Integration vector for pta elimination in Clostridium sp. MT1243

Vector pMTloxApta: Integration vector pMTlox1pta for elimination of pta in Clostridium sp. MT1243 comprised a cassette inside multiple cloning site of pUC19 (New England BioLabs, MA) serving as a backbone for DNA accumulation in E. coli. The first component of the cassette was terminator 1 (region 3491…3541 of NC_014328) flanked by a 450-bp fragment (1375865…1376315) just upstream of the pta (NC_014328, region 1376316…1377317). The second component was lox66 (Leibig et al. 2008; Berzin et al. 2012b) upstream of a synthetic Cm-resistance gene comprising cat (region 13456…14106 of FM201786) under the promoter sequence (region 2156 to 2233 of NC_014328) flanked with lox71 (Leibig et al. 2008; Berzin et al. 2012b) downstream. The third component of the cassette was a 456-bp fragment (1377318…1377774) just downstream of the pta (NC_014328, region 1376316…1377317) flanked by terminator 2 (NC_014328, region 8965…9061).

For electrotransformation, the amount of transforming DNA (pMTloxApta) was 10 μg per 0·2-ml samples with selection of recombinants using resistance to Cm.

Elimination of pta from chromosome of by excision

Cmr recombinant cells were prepared as for the regular electrotransformation and suspended in Cre reaction buffer (NEB, MA). Each sample was electrotransformed in the presence of 15 U of Cre-recombinase (NEB, MA) using parameters as for the DNA electrotransformation (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin and Tyurin 2012; Tyurin et al. 2012). Decimal dilutions of samples were plated on nonselective SFA to obtain single colonies. Only colonies sensitive to Cm were further tested for the absence of pta expression using rtPCR with the primers listed in the Table 1.

Table 1. Primers for rtPCR
GeneProduct size, bpPrimers
cat 3535′-gcatgataaccatcacaaac-3′; 5′-tgggataagttagagccact-3′
ack 3855′-tgggaactagatgcggtgat-3′; 5′-agtgcctctccccttttctt-3′
pta 3555′-gggaagtgcaaaacacgaat-3′; 5′-agcttgtgcttgaactgctg-3′
spo0A11345′-ataataatgcctcaccttga-3′; 5′-cctaatgttatagccctttg-3′
spo0A21565′-aagttttcaagagatgatcc-3′; 5′-ttttgtaactgccgataata-3′
spo0J11515′-attctctgcaaatcttcttg-3′; 5′-aaggcagctaagaatgataa-3′
spo0J21575′-tgctcttgtattgactgaga-3′; 5′-gagcattgatacctgaaagt-3′
al-adh 3945′-ataaagccttgacattcttt-3′; 5′-tacgtcaggattattaccaa-3′

Integration vector for elimination of ack and creation of the attTn7 insertion site for targeted al-adh integration in Clostridium sp. MT1243 chromosome.

Vector pMTloxAack.attTn7: Integration vector pMTlox1ack.attTn7 for elimination ack in Clostridium sp. MT1243 comprised a cassette inside multiple cloning site of pUC19 (NEB, MA) serving as a backbone for DNA accumulation in E. coli. The first component of the cassette was the 61-bp recognition site for Tn7 (attTn7): 5′-gcggcgggagtaccgctcggcgcaccgatccggccttcggatcgatgcgcctgccaacgaa-3′ as per Craig (1991) with added downstream terminator 1 (region 3491…3541 of NC_014328) flanked by a 323-bp fragment (NC_014328, region 1376994…1377317) just upstream of the ack (NC_014328, region 1377362…1378558). The second component was lox66 (Leibig et al. 2008; Berzin et al. 2012b) upstream a synthetic Cm-resistance gene comprising cat (region 13456…14106 of FM201786) under the promoter sequence (region 2156 to 2233 of NC_014328) flanked with lox71 (Leibig et al. 2008; Berzin et al. 2012b) downstream. The third component of the cassette was a 456-bp fragment (1378559…1379015) just downstream of the ack (NC_014328, region 1377362…1378558) flanked by terminator 2 (NC_014328, region 8965…9061). Colonies resistant to Cm were further tested for the presence of attTn7 using PCR with the primers for the presence of region comprising attTn7: 5′-gcctgccaacgaaatgtat-3′; 5′-ttgttcttggtgacttgccta-3′ with the expected product size of 224 bp. The same colonies were rtPCR tested for expression of synthetic cat using primers indicated in Table 1.

For electrotransformation, we used 10 μg of pMTloxApta DNA per 0·2-ml sample with selection of recombinants using resistance to Cm.

Elimination of ack from chromosome of Clostridium sp. MT1243 by excision

Cells of recombinants resistant to Cm were prepared as for regular electrotransformation and suspended in Cre reaction buffer (NEB, MA). Each sample was electrotransformed in the presence of 15 U of Cre-recombinase (NEB, MA) using parameters as for DNA electrotransformation (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin and Tyurin 2012; Tyurin et al. 2012). Decimal dilutions of samples were plated on nonselective SFA to obtain single colonies. Only colonies sensitive to Cm were further rtPCR tested for the absence of ack expression using primers indicated in Table 1.

Elimination of spo0A gene from the chromosome Clostridium sp. MT1243

Elimination of targeted spo0A gene from the chromosome was followed by the gene removal event similar to that described in detail above for the elimination of pta but aimed at elimination of spo0A. For that purpose, integration vector pMTloxAspo0A similar to pMTloxApta was engineered with the only difference that, instead of pta fragments, the 640-bp fragment (region 1223000…1223640) upstream of spo0A from C. ljungdahlii DSM13528 (region 1223641…1224456 of NC_014328) and 460 bp (1224457…1224917) fragment downstream of spo0A (NC_014328, region 1223641…1224456) was used to eliminate the ORFs of this early sporulation gene. Only colonies sensitive to Cm were further tested using rtPCR for the absence of spo0A expression using primers listed in Table 1.

Elimination of spo0J gene from the chromosome Clostridium sp. MT1243

Elimination of targeted spo0J gene from the chromosome was followed by the gene removal event similar to the above but aimed at the elimination of the spo0J gene. For that purpose, integration vector pMTloxAspo0J similar to pMTloxApta was engineered with the only difference that, instead of pta fragments, the 425-bp fragment (4620133…4620558) upstream of spo0J (region 4620559…4621419 of NC_014328) and 425-bp fragment downstream of spo0J (4621420…4621845) from C. ljungdahlii DSM13528 (NC_014328, region 4620559…4621419) were used to remove the ORFs of this early sporulation gene. Only colonies sensitive to Cm were further tested using rtPCR for the absence of spo0J expression using primers listed in Table 1.

Phenotypic confirmation of spo recombinants

In addition to the rtPCR, we confirmed the Spo phenotype using physiological test for sporulation (Tyurin et al. 2004, 2006, 2012). The phenotypic confirmation of Spo recombinants was confirmed via detection of zero cell viability after exposure of the samples to 50% ethanol for 1 h at 37°C under strictly anaerobic conditions.

Integration of bi-functional aldehyde/alcohol dehydrogenase using Tn7 assistance in the presence of transposases ABCDE

Rationale: Acetogen cells with eliminated acetate biosynthesis pathway would overproduce phosphorylated products of acetyl-CoA condensation, which render certain cell toxicity if accumulated above 80 mmol l−1 (Sivy et al. 2011). Therefore, addition of the amplified ethanol biosynthesis pathway would relieve the toxic effect for the recombinant cells shortening cell duplication time of recombinants producing ethanol as compared to nontransformed cell. The decrease in cell duplication time will be noticed by the difference in the size of the colonies with no need for antibiotic resistance–based recombinant selection: recombinant colonies producing ethanol will be larger. The ethanol-specific staining was also used to confirm ethanol-producing phenotype for quick and easy identification of ethanol-producing colonies. Below we describe vectors for integration of three and six copies of the synthetic al-adh into the specific integration site for Tn7.

Integration vectors

Vector pMTal-adhTn7.3: Synthesized and supplied in circular form integration vector pMTal-adhTn7.3 for integration of synthetic ethanol biosynthesis genes in ethanol-resistant mutant comprised the left end of Tn7 (Tn7L) (JQ429758, region 1893…2058) and then all the subsequent components in inverted orientation relative to Tn7L: terminator (NC_014328, region 8965…9061), bi-functional aldehyde/alcohol dehydrogenase (al-adh) (NC_014328, region 1791269..1793881) and the promoter sequence (region 2156–2233 of NC_014328), in triplicate separated with 5′-aaa-3′ spacers as usually, and then the right end of Tn7 (Tn7R) (JQ429758, region 189..387).

Vector pMTal-adhTn7.6: Synthesized and supplied in circular form integration vector pMTal-adhTn7.3 for integration of synthetic ethanol biosynthesis genes in ethanol-resistant mutant comprised the left end of Tn7 (Tn7L) (JQ429758, region 1893…2058) and then all the subsequent components in inverted orientation relative to Tn7L: terminator (NC_014328, region 8965…9061), bi-functional aldehyde/alcohol dehydrogenase (al-adh) (NC_014328, region 1791269..1793881) and the promoter sequence (region 2156–2233 of NC_014328) in six identical copies separated with 5′-aaa-3′ spacers as usually and then the right end of Tn7 (Tn7R) (JQ429758, region 189..387).

Electrotransformation procedure described by Tyurin 1992; Tyurin et al. 2004, 2005; Berzin and Tyurin 2012; Tyurin et al. 2012) was used. The amounts of transforming pMTal-adhTn7.3 or pMTal-adhTn7.6 DNAs were 10 μg per 0·2-ml sample. In addition, each sample was loaded with 15 U of transposases ABCDE custom synthesized by NEB (New England BioLabs, MA). Therefore, each sample was electrotransformed with the mixture of vector DNA and transposases. Decimal dilutions of samples were plated on nonselective SFA to obtain single colonies.

Detection of ethanol-producing recombinant colonies followed by detection of expression of al-adh

Detection of ethanol-producing recombinants was performed as follows. Each Petri dish with single colonies was replica-plated using sterile velveteen squares (http://www.amazon.com/Bel-Art-Scienceware-378480000-Replica-Plating-Velveteen/dp/B002VBW7W0) to sterile dishes. The dishes with plated replicas were marked in the way to trace the original colonies for subsequent testing using rtPCR and then in fermentation vessels as appropriate.

After replica-plating, each original Petri dish was sprayed with the commercially available reagent for colorimetric ethanol detection ALCO-SCREEN (Chematics, Inc., North Webster, IN, USA). The method was based on the colorimetric detection of ethanol in the colony residues still left on the agar surface and ethanol diffused into the agar underneath such colonies formed by cells expressing al-adh. Such ethanol-containing agar spots were stained darker compared with the plate background in the areas with no growth or with colonies with no ethanol production. The method was quick, inexpensive and easy to use to distinguish the larger colonies of recombinants with no need for antibiotic selection to use as the selective marker.

In addition to colorimetric ethanol detection, the al-adh+ recombinants using rtPCR and primers indicated in Table 1 were confirmed.

Promoter and terminator sequences for the components of all vectors

Promoter and terminator sequences for the components of all vectors were identified using SoftBerry Bacterial Promoter, Operon and Gene Finding tool (http://linux1.softberry.com/).

rtPCR

Qiagen OneStep RT-PCR Kit and mRNAs isolated from the chosen Clostridium sp. MT1234 and respective ethanol-producing recombinant of that mutant were used as described (Paul et al. 2010; Berzin et al. 2012a,b). Primers specific to detect the expression of cat, ack, pta, spoA1, spoA2, spo0J1, spo0J2 and al-adh (Table 1) in the respective intermediate recombinants or the finalized biocatalysts were used where appropriate.

PCR

PCR was performed using PCR Kit (Qiagen, CA) DNA templates isolated from Clostridium sp. MT1243ack::aatTn7, Clostridium sp. MT1243al-adhTn7.3 and Clostridium sp. MT1243al-adhTn7.6 and ack::aatTn7 primers: 5′-gcctgccaacgaaatgtat-3′; 5′-ttgttcttggtgacttgccta-3′.

DNA sequencing and synthesis

DNA sequencing (both DNA strands) including all primers, the synthesized constructs and DNA inserts (for pUC19 backbone) and DNA synthesis were performed by Integrated DNA Technologies, Inc. (San Diego, CA, USA).

HPLC Analysis

For HPLC analysis, which was performed as described by Berzin and Tyurin 2012; Berzin et al. 2012a,b), standard positive control mixture containing formate, acetate, ethanol, acetaldehyde, acetone, acetoacetate, ethyl acetoacetate, butyraldehyde, butyrate and n-butanol was used on Aminex HPX-87H column (Bio-Rad Laboratories, Inc., Hercules, CA, USA) at 55°C with 5 mmol l−1 sulphuric acid as the mobile phase at 0·6 ml min1 flow rate. Detections were performed via refractive index using Waters 2414 Infra Red Detector (Milford, MA). The minimal detection limit was set at 0·1 mmol l−1. The samples were prepared by mixing of 0·5 ml of 0·1 μm filtered fermentation broth with 0·5 ml 1·76 mol l−1 H3PO3 and filtering again the resulted solution through 0·1 μmol l−1 filter to HPLC vials, both done no later than 10 min prior to the HPLC detection.

Statistical analysis

The significance of differences between means for all fermentation experiments was evaluated based on a one-sided t-test as described by Ross (2000). Daily sampling comprised 96 points for the listed liquid and gas components of the process for each vessel in operation for 25 days of continuous fermentation runs in BioFlo2000 in five independent replicas using syngas or carbon dioxide-hydrogen blends.

The recombinant clones subjected to continuous fermentations studies were of two major groups: Group 1 included recombinants with eliminated pta and ack and integrated three and six copies of al-adh, and all clones were capable of sporulation; and Group 2 included recombinants with eliminated pta, ack, spo0A, spo0J and integrated three or six copies of al-adh. In separate experiments, the last type of recombinants was additionally tested to compare the process efficiency if the syngas blend was replaced with gas blend comprising 1 mol of CO2 and 4 moles of H2.

Results

Integration of pMTloxApta

Syngas Biofuels Energy, Inc. generator rendered frequency of pMTloxApta integration into the chromosome of Clostridium sp. MT1234 8·2 ± 0·3 × 10−3 per the number of recipient cells. The obtained Cmr recombinants were analysed for the presence of pta and cat using rtPCR, and the genotype of Clostridium sp. MT1243cat+ was confirmed (Fig. 1).

Figure 1.

rtPCR: expression of cat in Clostridium sp. MT1243cat+, band size marker is 100-bp Ladder.

Elimination of pta and cat from chromosome of Clostridium sp. MT1243cat+ by excision

After electrotransformation of five independent clones Clostridium sp. MT1243pta+cat+ in five repeats with Cre-recombinase (NEB, MA) using parameters as for DNA electrotransformation (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin et al. 2012b; Berzin and Tyurin 2012; Tyurin et al. 2012), we randomly tested 91 340 clones grown on the Petri dishes inoculated with the third dilution of the transformed samples as single colonies of the size suitable further clone testing for sensitivity to Cm. Five per cent of such clones (4567) were Cm-sensitive and thus were further tested for the absence of pta and cat expression using rtPCR. All the tested Cms clones were found catpta (Clostridium sp. MT1243ptacat).

Integration of pMTloxAack.attTn7 in Clostridium sp. MT1243ptacat

Syngas Biofuels Energy, Inc. generator rendered frequency of pMTloxAack.attTn7 integration into the chromosome of Clostridium sp. MT1243ptacat 7·8 ± 0·2 × 10−3 per the number of recipient cells. The obtained Cmr recombinants were analysed using rtPCR for the presence of cat, ack and absence of pta expression. The genotype of Clostridium sp. MT1243ack+ptacat+ was confirmed. In addition to that, we have confirmed the presence of the attTn7 using PCR in the DNA of such clones.

Elimination of ack from chromosome of Clostridium sp. MT1243ptacat+attTn7 by excision

After electrotransformation of Clostridium sp. MT1243ptacat+attTn7 with Cre-recombinase (NEB, MA) using parameters as for the DNA electrotransformation, we have obtained 96 865 clones grown on the Petri dishes inoculated with the third dilution of the transformed samples as single colonies of the size suitable for clone analysis. Five per cent of such clones (4844) revealed sensitivity to Cm. All tested clones were found ptaack via rtPCR using primers indicated in the Table 1. The presence of aatTn7 was re-confirmed using PCR. The resulted clones had genotype Clostridium sp. MT1243ptaackattTn7.

Elimination of spo0A

Elimination of spo0A gene from the chromosome of Clostridium sp. MT1243ptaackattTn7 was followed by the gene removal event similar to the above but aimed at elimination of the spo0A gene using integration vector pMTloxAspo0A. The frequency of cat (pMTloxAspo0A) integration into the chromosome was 3·1 ± 0·1 × 10−3 per the number of recipient cells. The obtained Cmr recombinants were confirmed for the expression of cat using rtPCR.

After electrotransformation of the Cmr recombinants with Cre-recombinase (NEB, MA) using parameters as for the DNA electrotransformation, we have obtained 66 809 clones grown on the Petri dishes inoculated with the third dilution of the transformed samples as single colonies of the size suitable for clone analysis. Five per cent of such clones (3340) were sensitive to Cm. All the tested clones were found spo0A using rtPCR and primers indicated in the Table 1. One of such tested clones is shown in Fig. 2. The presence of aatTn7 was re-confirmed using PCR.

Figure 2.

rtPCR: absence of spo0A expression in Clostridium sp. MT1243catspo0A, band size marker is 100-bp Ladder.

Elimination of spo0J

Elimination of spo0J gene from the chromosome of Clostridium sp. MT1243ptaackspo0A attTn7 was followed by the gene removal event similar to the above but aimed at elimination of the spo0A gene using integration vector pMTloxAspo0J. The frequency of cat (pMTloxAspo0J) integration into the chromosome was 4 × 10−3 per the number of recipient cells. The obtained Cmr recombinants were confirmed for the presence of cat using rtPCR.

After electrotransformation of the Cmr recombinants with Cre-recombinase (NEB, MA) using parameters as for the DNA electrotransformation, we have obtained 76 934 clones grown on the Petri dishes inoculated with the third dilution of the transformed samples as single colonies of the size suitable for clone analysis. Five per cent of such clones (3847) were sensitive to Cm and thus were further tested for the absence of spo0J. All the tested clones were found spo0J using rtPCR and primers indicated in the Table. One of such tested clones is shown in Fig. 3. The presence of aatTn7 was re-confirmed using PCR.

Figure 3.

rtPCR: absence of spo0J expression in Clostridium sp. MT1243catspo0Aspo0J, band size marker is 100-bp Ladder.

Additional tests confirmed loss of sporulation capability in the spo0Aspo0J recombinants. All the recombinants did not survive exposure to 50% ethanol for 1 h at 37°C under strictly anaerobic conditions.

Integration of al-adh: easy detection of ethanol production by the recombinant clones on original Petri dishes and characterization of the replica-printed recombinant clones

After replica-plating, each original Petri dish was sprayed with the reagent for colorimetric ethanol detection ALCO-SCREEN. The ethanol-containing areas of agar underneath recombinant colonies stained darker compared with the background and in the areas with colonies with no ethanol production.

In addition to colorimetric ethanol detection, we confirmed the al-adh+ in the replica-plated recombinants using rtPCR and primers indicated in Table 1.

The frequency of pMTal-adhTn7.3 integration was 2·3 ± 0·1 × 10−6 per total number of recipient cells. We had to analyse batches of fifteen independent experiments with 267 Petri dishes inoculated with the third sample dilutions plated in every batch to select total 83 clones producing ethanol. The frequency of pMTal-adhTn7.6 integration was 1·2 ± 0·05 × 10−6 per total number of recipient cells. We had to analyse batches of eighteen independent experiments with 263 Petri dishes inoculated with the third sample dilutions plated in every batch to select total 57 clones producing ethanol. Analysis using rtPCR of such clones did not show any deviation: all clones were proven as expressing al-adh.

Integration of three or six copies of al-adh in Clostridium sp. MT1243ptaackattTn7 prior to elimination of early-stage sporulation genes was performed as above and is not described in detail because the frequencies of events were essentially of the same orders as above. The clone testing and molecular characterization was performed as appropriate based on the guidelines described.

Continuous fermentation

Syngas blend

Recombinants of Group 1 with three copies of al-adh (Clostridium sp. MT1243ptaackal-adh.3) produced 490 mmol l−1 ethanol by the seventh day of 25-day-long continuous fermentations (P < 0·005). Recombinants with six copies (Clostridium sp. MT1243ptaackal-adh.6) established 928 mmol l−1 ethanol levels by the same time (P < 0·005).

Elimination of both spo0A and spo0J improved ethanol yields by 35 (Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.3) and 90 (Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.6) mM in the respective recombinants of Group 2 (P < 0·005) (Fig. 4).

Figure 4.

Continuous syngas blend fermentation by recombinant biocatalysts. Ethanol: Clostridium sp. MT1243ptaackal-adh.3 (triple solid line); Clostridium sp. MT1243ptaackal-adh.6 (double thin-thick solid line); Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.3 (double thick-thin solid line); Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.6 (solid line); vent gases: H2 (double dotted line); CO (double broken line) and CO2 (double solid line).

Carbon dioxide – hydrogen blend

Recombinants of Group 1 with three copies of al-adh (Clostridium sp. MT1243ptaackal-adh.3) produced 314 mmol l−1 ethanol by the seventh day of 25-day-long continuous fermentations (P < 0·005). Recombinants with six copies (Clostridium sp. MT1243ptaackal-adh.6) established 594 mmol l−1 ethanol levels by the same time (P < 0·005). Elimination of both spo0A and spo0J improved ethanol yields by 22 (Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.3) and 58 (Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.6) mM in the respective recombinants of Group 2 (P < 0·005).

Ethanol production from CO2/H2 blend by Clostridium sp. MT1243ptaackal-adh.6 and Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.6 is presented in Fig. 5.

Figure 5.

Continuous CO2/H2 blend fermentation by recombinant biocatalysts. EtOH: Clostridium sp. MT1243ptaackal-adh.6 (double thin-thick solid line); Clostridium sp. MT1243ptaackspo0Aspo0Jal-adh.6 (solid line); vent gases: H2 (double dotted line); CO (double broken line) and CO2 (double solid line).

Continuous fermentation was kept for 25 days each after establishing constant flow rates for the broth flows in five independent repeats for each of five alike clones of each type tested and could last longer with no detectable problems (data not shown). The 25-day-long fermentation runs were chosen to illustrate stability of the process and to make the bioreactors available for the largest number of continuous runs at particular laboratory space and time frame dedicated for the project.

Discussion

Acetate pathway comprising pta and ack was the first target for the gene elimination effort. As that was learned experimentally, Clostridium sp. MT1243 shared enough DNA homology of the target genes with the published sequences of the functionally similar genes in acetogen (NC_014328) to make the gene elimination efforts successful. Acetate production in acetogens is one of the ways of energy conservation in the form of ATP (Gaddy et al. 2001; Köpke et al. 2011). Based on the presented continuous fermentation data and from the past experience (Berzin et al. 2012a,b), elimination of acetate pathway so far did not render substantial negative effect on the engineered acetogen biocatalyst strains performances. The data suggest that the eliminated acetate pathway in Clostridium sp. MT1243 should not be the major carbon sink during syngas fermentation.

Elimination of unnecessary genes improves strain overall performance and shortens cell division time thus rendering more biomass and/or product compared with the same strain with a full-size genome. One of the best ever practical examples of that is Clean Genome® E. coli (Scarab Genomics, LLC, Madison, WI, USA) with the outstanding cell growth performance.

After elimination of some early sporulation genes in Clostridium MT1234, a boost of the recombinant ethanol production was detected. These data support our hypothesis that elimination of some genes in acetogens not essential for a particular process released the cell energy previously used for replication, expression and other maintenance needs of the eliminated genes. There are many other genes in acetogens not used during syngas fermentation, which might be eliminated. This might be one of the future ways to consider for further efforts in biocatalyst strain improvement. Surprisingly, despite of the various genotypes of the recombinants engineered using Clostridium sp. MT1243, there were no significant differences (P < 0·005) in the gas consumption profiles for all the recombinants tested. In case of syngas blend, the rates of CO and H2 consumption did not shift much if three or six copies of al-adh were expressed. The same gas consumption patterns were detected for recombinants either with or without early-stage sporulation genes.

Similar patterns of gas consumption were revealed if CO2/H2 blend was used. This fact suggests that the overall biocatalyst metabolic cell machinery in the studied engineered biocatalyst strains consumed so much of the gas components that any changes associated with the carbon flux fraction devoted solely to ethanol production had the only effect reflected in changes in the amounts of ethanol produced by different kinds of the respective recombinants.

Sporulation machinery in acetogens is closely linked to availability of inorganic carbon. Starving in CO/CO2 triggers cell to sporulate to preserve the energy accumulated so far in inorganic carbon reduction process (M. Tyurin – personal communication). Sporulation machinery is common for many clostridia and incorporates about ten genes associated with sporulation in and subsequent germination of spores when inorganic carbon source deems available as per the cell sensor mechanisms. Similar processes were found in variety of micro-organisms including Streptomyces producing antibiotics. Elimination of sporulation genes there and genes encoding spore germination speeded up the new antibiotic discovery process in the past. We may think of elimination all of the genes encoding sporulation and spore germination in acetogens to preserve the cell energy pool for metabolic engineering endeavours.

Tn7 approach for targeted integration worked well rendering stable performance of the integrated genes. The product yield was in direct proportion to the number of the gene copies integrated (as per the construct design), while we did not measure the actual gene copy number per recombinant genomes at various points of the stable continuous fermentation processes presented.

The quick test for ethanol detection allowed us to spray the original plates inoculated with proper dilutions of the electrotransformed samples to identify exactly single recombinant colonies producing ethanol. The convenience and time-saving feature allowed us to directly single out a number of colonies for further testing. That would be virtually impossible if random colony pick up would be used for the subsequent HPLC and rtPCR analysis.

The engineered biocatalysts were capable of CO2 fermentation in the presence of hydrogen and rendered about 64% of the continuous fermentation performance of that recorded for the syngas blend under other comparable conditions. This fact opens the frontiers for unlimited scaling up of CO2 biocatalysis directly and selectively to valuable chemicals and food components bypassing photosynthetic pathway of inorganic carbon reduction to carbon of carbohydrates existing in nature. This might become the crucial global technological approach in the immediate future because the planet experiences global warming with growing consequences.

From the practical point, it is difficult to propose the actual mass balance for syngas fermentations due to limitations of gas solubility in water phase and the contribution of carbon-containing components of nutrient media in the carbon flux. We suspect that gaseous substrates might be consumed directly by the biomass thus bypassing the water phase barrier when certain cell densities were reached. The empirical modification of the water–gas shift reaction describes the approximate carbon flux in the engineered biocatalyst fermenting syngas blend (60% CO + 40% H2) to ethanol with zero CO2 emissions (Equation 1):

display math(1)

If the gas blend comprised only CO2 + 4H2, then the acetogen cells enforced functioning of both methyl and carbonyl branches of the acetyl-CoA pathway rendering another empirical stoichiometry (Drake 1994):

display math(2)

Modern solar panels with efficiency of light energy recovery > six times of that of photosynthetic path would require just a fraction of surface area required to grow plants as the source of carbohydrates to produce hydrogen. For instance, a 3000 ft2 solar panel roof over a 20 000 gallon horizontal bioreactor would produce 300 kW every second at the peak of the day light time. Solar panels would enable commercial hydrogen production in situ (bioreactor roof, etc.) for the process needs when electrolysis of water solutions would be integral part of the technology process (http://www.nrel.gov/hydrogen/proj_production_delivery.html). Vent gas (100% CO2) of >100 MW power plants utilizing Integrated Gasification Combined Cycle (IGCC) for clean coal burning would make the fermentation process sustainable.

Electrotransformation has been proven as the ultimate tool for metabolic engineering with strictly anaerobic micro-organisms when it is essential to keep cell samples as anoxic as possible (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin et al. 2012b; Berzin and Tyurin 2012; Tyurin et al. 2012). The integration frequencies detected were in a good agreement with the numbers reported earlier for the acetogens and other anaerobic micro-organism with the Gram (+) cell wall when Syngas Biofuels Energy, Inc. electrotransformation generators were used (Tyurin et al. 2004, 2005; Berzin et al. 2012b; Berzin and Tyurin 2012; Tyurin et al. 2012).

The results obtained might be outlined by several unique factors specifically rendered by Syngas Biofuels Energy, Inc. electroporation generators.

  1. Circuit design and sample connection to the power circuit: Commercial Square Wave Electroporators comprise insulated gate bipolar transistor (IGBT) with the voltage output rated to >3·0 kV. IGBT is a combination of insulated gate low power transistor with input impedance around 1 MΩ powering low impedance input silicon collector-emitter p-n threshold of a power bipolar transistor rated for high pulse currents to survive hundreds of amperes of the pulse current at the output voltages >3·0 kV. The resulting n-p-n ‘sandwich’ has its transient capacitance >1500 pF to ensure it sustains repetitive pulsing high voltage high currents without detectable thermal disintegration for >5500 h of operation life. The cell sample is connected in parallel to the IGBT with no opportunity to monitor the actual pulse current through the sample to optimize electric treatment conditions as described (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin et al. 2012b; Berzin and Tyurin 2012; Tyurin et al. 2012). Regular low internal resistance electrolytic capacitor is used to power the high voltage output of the generator.

Syngas Biofuels Energy, Inc. electrotransformation generator comprises a power tetrode as the active power circuit element, which is a power vacuum tube with the transient capacitance below 13 pF and operation life 100 000 h at the anode current above 52 A. Only such combination of low transient capacitance of the power circuit and its tremendous capability to sustain extreme pulse voltages and currents during pulse application renders a recordable electric pulse application causing electropermeabilization of cell membranes in samples with high internal cell sample capacitance [3–8 μF in acetogens compared with 150 pF for E. coli cells (Tyurin 1992)]. The power tetrode actively amplifies nonlinear pulse current changes in the circuit due to its low transient capacitance thus facilitating generation of high-frequency pulse current oscillations generated by the cell sample connected in series with the power tetrode and the power capacitor (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin et al. 2012b; Berzin and Tyurin 2012; Tyurin et al. 2012).

  1. The difference in the electroporation cuvette/electrodes assembly and in thermal exposure of the electropermeabilized cells: Commercial Square Wave Electroporators use standard 0·1-mm plastic disposable cuvettes with aluminium plate parallel electrodes rendering uniform pattern of the pulse electric filed applied and a 2-mm-thick polystyrene wall preserving a great portion of the pulse heat regardless of pre- and postpulse cooling on ice. Syngas Biofuels Energy, Inc. electrotransformation generator uses 0·5-ml disposable sterile PCR tubes as the electroporation cuvettes imbedded in ice block during pulse application as described with the assembly of re-usable titanium alloy VK-2 polished electrodes creating a nonuniformic electric field applied to the samples (Tyurin 1992; Tyurin et al. 2004, 2005; Berzin et al. 2012b; Berzin and Tyurin 2012; Tyurin et al. 2012).
  2. Effect of oxygen accumulation in the cell sample during pulse application due to water electrolysis: mitigating role of high-frequency pulse current oscillations in the proprietary Syngas Biofuels Energy, Inc. circuit used: In the typical case of electroporation with a single DC pulse, if we might possibly assume the cells have been electroporated using Bio-Rad Gen Pulser with the generator output set at ‘infinity’ and the voltage at 2500 V, then the recorded pulse constant would be around 4·7 ms rendering the pulse decaying oxygen produced during the pulse caused water electrolysis as somewhat around of 3500 ppm in the sample immediately after the pulse application. The negative electroporation outcome for a strict anaerobe becomes obvious due to killing most of the cells and triggering initiation of sporulation event in the rest. That was noted if sporulation was triggered on, then the introduced DNA was never included into the spores. With the use of Syngas Biofuels Energy, Inc. electrotransformation generators, there is still some production of oxygen due to electrolysis of water phase in the sample. However, generation of the pulse current oscillations at high frequency around 20–25 MHz (Fig. 6) holds the produced oxygen ions in place and prevents them from reaching the core of the sample where the cells are still subjected to the electric field in the presence of the transforming DNA (and proteins when appropriate) under anoxic conditions. All the above factors make Syngas Biofuels Energy, Inc electrotransformation generators unique and render high electrotransformation efficiency of electroporating cells of strict anaerobes.
Figure 6.

Generation of the pulse current oscillations at high frequency around 20–25 MHz at the cell sample of acetogen electrocompetent cells during electrotransformation using Syngas Biofuels Energy, Inc. electrotransformation generator: pulse current at the current sensor at oscilloscope horizontal deflection 20 ns/division.

Conclusion

The report describes development and application of standard molecular biology tools for engineering of industrial grade strain biocatalyst suitable for CO2 reduction to valuable chemical and food components. The detailed approach for targeted gene elimination to release fraction of cell energy pool to power introduced synthetic ethanol biosynthesis pathway is described. The synthetic ethanol biosynthesis pathway was integrated into a designed integration site in the chromosome without disrupting genes essential for the process. The integration was performed in the predetermined number of copies to test whether the released, due to elimination of certain genes, cell energy was sufficient enough to power the engineered ethanol pathway.

It was proven that gene elimination rendered beneficial effect on the biocatalyst performance. Targeted integration of three or six copies of the synthetic ethanol pathway using Tn7 integration tool was achieved. Based on the previously reported experience, it is feasible to have the reported ethanol concentration doubled if the second step is added to the continuous fermentation process with no actual increase in the total volume of the liquid to process for the product recovery, just by changing the cell retention time in the presence of feeding gas blend and the rational commercial bioreactor design (not shown). This is the first report on the targeted gene integration in the predetermined number of copies in acetogen biocatalyst for CO2 reduction to ethanol fuel. The technology might become the crucial global technological approach to CO2 reduction in the atmosphere in the immediate future because the planet already experiences global warming with growing consequences getting ready to get out of control.

Acknowledgements

The research was supported by the funds of Syngas Biofuels Energy, Inc. Syngas Biofuels Energy, Inc. is the sole distributor of the electroporation and electrofusion equipment: www.syngasbiofuelsenergy.com.

Conflict of interest

The authors declare that they have no conflict of interest. The authors warrant that they have the authority to publish the material.

Ancillary