Modifying the product pattern of Clostridium acetobutylicum
Clostridial acetone–butanol–ethanol (ABE) fermentation is a natural source for microbial n-butanol production and regained much interest in academia and industry in the past years. Due to the difficult genetic accessibility of Clostridium acetobutylicum and other solventogenic clostridia, successful metabolic engineering approaches are still rare. In this study, a set of five knock-out mutants with defects in the central fermentative metabolism were generated using the ClosTron technology, including the construction of targeted double knock-out mutants of C. acetobtuylicum ATCC 824. While disruption of the acetate biosynthetic pathway had no significant impact on the metabolite distribution, mutants with defects in the acetone pathway, including both acetoacetate decarboxylase (Adc)-negative and acetoacetyl-CoA:acyl-CoA transferase (CtfAB)-negative mutants, exhibited high amounts of acetate in the fermentation broth. Distinct butyrate increase and decrease patterns during the course of fermentations provided experimental evidence that butyrate, but not acetate, is re-assimilated via an Adc/CtfAB-independent pathway in C. acetobutylicum. Interestingly, combining the adc and ctfA mutations with a knock-out of the phosphotransacetylase (Pta)-encoding gene, acetate production was drastically reduced, resulting in an increased flux towards butyrate. Except for the Pta-negative single mutant, all mutants exhibited a significantly reduced solvent production.
KeywordsBiofuels Butanol Butyrate ABE fermentation Metabolic engineering
Clostridia represent a diverse group of Gram-positive, strictly anaerobic bacteria which are able to form endospores for long-term survival of unfavorable environmental conditions. Besides several pathogenic species, a large number of terrestrial non-pathogenic species show biotechnologically interesting metabolic properties, including cellulose degradation and the biosynthesis of solvents. The model organism for acetone–butanol–ethanol (ABE) fermentation is Clostridium acetobutylicum, whose genome sequence was already published 10 years ago (Nölling et al. 2001). The fermentative metabolism includes several branchpoints and typically the acids acetate and butyrate are formed during exponential growth, whereas the ABE solvents are produced in the stationary phase. Tightly associated to the metabolic changes, the cells initiate the endospore formation, and the later release of mature spores terminates the characteristic life cycle of C. acetobutylicum (Paredes et al. 2005; Jones et al. 2008). The physiological relationship and the regulatory circuits are still not well understood, although some new insights into the role of different sigma factors and phosphorylation mechanisms were obtained recently (Tracy et al. 2011; Jones et al. 2011; Bi et al. 2011; Steiner et al. 2011).
Since native n-butanol synthesis comprises a significant byproduct formation, different strategies to improve the butanol yield were considered (Lee et al. 2008; Papoutsakis 2008). However, an important drawback was the fact that clostridia were genetically inaccessible for a long time while other model organisms such as Escherichia coli offer a great portfolio of suitable molecular techniques for metabolic engineering. Thus, recombinant butanol production in E. coli and other organisms was conducted by expression of the respective clostridial genes (Inui et al. 2008; Atsumi et al. 2008; Steen et al. 2008; Berezina et al. 2010; Nielsen et al. 2009; Fischer et al. 2010). However, initial attempts revealed only very low butanol titers because of the limited function of the clostridial butyryl-CoA dehydrogenase complex in E. coli. Substitution of this enzyme with an NADH-dependent enoyl-CoA reductase in the heterologous pathway significantly increased the butanol level to the gram per liter scale (Bond-Watts et al. 2011; Shen et al. 2011). Comparably high butanol amounts, i.e. 10 g/l, were obtained by using an acetate kinase-negative mutant of C. tyrobutyricum as a host, naturally equipped with the butyryl-CoA biosynthetic pathway similar to solventogenic clostridia (Yu et al. 2011).
Nevertheless, the robustness of C. acetobutylicum and related production strains has been demonstrated on the industrial scale, but the unique metabolism leaves many scientific questions open. Hence, appropriate metabolic engineering tools were developed to broaden the physiological knowledge on this interesting and technically relevant microorganism (Ezeji et al. 2010; Green 2011; Lütke-Eversloh and Bahl 2011). For unknown reasons, recombination in C. acetobutylicum and other solventogenic clostridia is extremely difficult, and knock-out mutants described until recently comprised exclusively single-crossover events of homologous recombination obtained by non-replicative plasmids. Therefore, early studies on specifically decreasing gene expression in C. acetobutylicum employed antisense RNA for targeted gene "knock-down", although with limited success (Desai and Papoutsakis 1999). A reliable and reproducible method, the ClosTron system, which is based on the mobile group II intron of Lactococcus lactis and therefore recombination-independent, works well in various clostridial species, including C. acetobutylicum (Heap et al. 2007, 2010; Shao et al. 2007). Group II introns are catalytically active RNA molecules able to self-splice out of the RNA transcript via a mechanism similar to that of eukaryotic nuclear introns. The mobile group II intron also encodes a multi-domain intron-encoded protein, which posesses several catalytic activities to allow the spliced-out intron RNA to be re-targeted, i. e. used as a template for the specific insertion into the respective gene. To accomplish the latter, only a minimal consensus sequence which determines the recognition site has to be exchanged (Karberg et al. 2001). To our best knowledge, a total of 21 TargeTron-based or ClosTron-based mutants of C. acetobutylicum plus one double knock-out have been published until today. These include proof-of-concept mutants (Heap et al. 2007, 2010; Jia et al. 2011; Shao et al. 2007), methodical mutants for genetic engineering purposes (Dong et al. 2010, 2011; Jia et al. 2011), glucose utilization mutants (Ren et al. 2010; Xiao et al. 2011) and histidine kinase mutants (Steiner et al. 2011), but only three mutants with defects in the fermentative metabolism (Shao et al. 2007; Jiang et al. 2009; Lehmann and Lütke-Eversloh 2011). In this study, five ClosTron-mediated knock-out mutants of C. acetobutylicum ATCC 824 with defects in the acetate and acetone biosynthetic pathways, as well as combinations thereof, were generated and physiologically characterized.
Materials and methods
Strains and cultivation conditions
Strains and plasmids used in this study
Strain or plasmid
C. acetobutylicum ATCC 824
Amercian Type Culture Collection
C. acetobutylicum pta::int(80)
Group II intron inserted at 80./81. bp of pta (CAC1742), ErmR
C. acetobutylicum pta::int(80)ΔRAM
Group II intron inserted at 80./81. bp of pta, RAM cassette excised, ErmS
C. acetobutylicum adc::int(180)
Group II intron inserted at 180./181. bp of adc (CAP0165), ErmR
C. acetobutylicum ctfA::int(352)
Group II intron inserted at 352./353. bp of ctfA (CAP0163), ErmR
C. acetobutylicum pta::int(80)ΔRAM-adc::int(180)
Group II intron without RAM inserted at 80./81. bp of pta, group II intron inserted at 180./181. bp of adc, ErmR
C. acetobutylicum pta::int(80)ΔRAM-ctfA::int(180)
Group II intron without RAM inserted at 80./81. bp of pta, group II intron inserted at 180./181. bp of adc, ErmR
E. coli DH5α
F−, φ80lacZΔM15, Δ(lacZYA), recA1, endA1, hsdR17 (rk−, mk+), phoA, supE44 thi1, gyrA96, relA1, λ−
Grant et al. 1990
E. coli ER2275
mcrA, ΔmcrBC, hsdR, recA1
Mermelstein and Papoutsakis 1993
φ3TI, p15A, TetR
Heap et al. 2007
ltrA, Ll.ltrB intron, pCB102, ColE1, CmR
Heap et al. 2010
pMTL007C-E2 re-targeted for pta
pMTL007C-E2 re-targeted for adc
pMTL007C-E2 re-targeted for ctfA
FLP1, repL, ColE1, CmR
Cultures were inoculated with spores in clostridial growth medium (CGM) containing per liter 0.75 g KH2PO4, 0.75 g K2HPO4, 0.71 g MgSO4 × 7 H2O, 0.01 g MnSO4 × 7 H2O, 0.01 g FeSO4 × 7 H2O, 1 g NaCl, 2 g (NH4)2SO4, 5 g yeast extract, 2 g asparagin and 1 mg resazurin; the pH was adjusted to 6.6 with NH4OH, and 2.5 g glucose was added after autoclaving (Roos et al. 1985). Prior to cultivation, the CGM cultures were pasteurized for 10 min at 80 °C to inactivate vegetative cells. For C. acetobutylicum mutants, erythromycin was added at a concentration of 20 μg/ml to the CGM cultures, but was omitted in the second precultures and main cultures.
Recombinant DNA techniques
Standard protocols to isolate, amplify, manipulate and transform DNA were employed according to Sambrock and Russell (2001). DNA-modifying enzymes, kits and biochemicals were purchased from Fermentas GmbH (St. Leon-Rot, Germany), New England Biolabs GmbH (Schwalbach, Germany), Roche Diagnostics GmbH (Mannheim, Germany), Sigma-Aldrich Chemie (Deishofen, Germany), Macherey & Nagel GmbH (Düren, Germany), Quiagen GmbH (Hilden, Germany) or Peqlab Biotechnologie GmbH (Erlangen, Germany) and were used according to the manufacturer's description. Other chemicals were obtained from Sigma-Aldrich Chemie (Deishofen, Germany) or Applichem GmbH (Darmstadt, Germany). DNA sequencing was done by GATC Biotech AG (Konstanz, Germany).
Oligonucleotide sequences for PCR primers are listed in Table S1. For recombinant DNA techniques, E. coli DH5α (Grant et al. 1990) and E. coli ER2275 pAN2 (Heap et al. 2007) were cultivated in Luria Bertani (LB) medium comprising per liter 10 g trypton, 5 g yeast extract and 10 g NaCl; solid LB contained 15 g agar. Antibiotics were added as required (Sambrock and Russell 2001).
The generation of C. acetobutylicum mutants was conducted as previously described in detail (Heap et al. 2007, 2010; Lehmann and Lütke-Eversloh 2011). Briefly, for the disruption of the genes pta, adc and ctfA, plasmid pMTL007C-E2 was retargeted using the respective primers as listed in Table S1. After in vivo methylation in E. coli ER2275 pAN2, the plasmids were tranformed in C. acetobutylicum and selected on RCA containing 15 μg/ml thiamphenicol (Mermelstein and Papoutsakis 1993; Riebe et al. 2009). After 48 h, single colonies were resuspended in 1 ml CGM with 7.5 μg/ml thiamphenicol, incubated for 6 h at 37 °C and spread on RCA containing 20 μg/ml erythromycin. Single colonies were streaked on fresh selective plates and analyzed by PCR screening. For this, cell material was resuspended in 25 μl H2O and cooked for 10 min prior to its use as PCR template. To construct double mutants, the RAM cassette was excised by transformation with plasmid pCLF1 harboring an FLP recombinase (Soucaille 2008). Thiamphenicol-resistant and erythromycin-resistant colonies were selected after 48 h at 37 °C and transferred to 4 ml CGM containing 15 μg/ml thiamphenicol, but no erythromycin. After overnight incubation at 30 °C, cell suspensions were plated on RCA plus thiamphenicol, and the plates were further incubated for 3 days at 30 °C. Single colonies were streaked on RCA with and in parallel without erythromycin. Subsequently, erythromycin-sensitive clones were streaked on RCA with and in parallel without thiamphenicol to select clones which lost plasmid pCLF1. Positive clones were analyzed by PCR screening.
The correct insertion sites of the introns were validated by Southern hybridization with gene-specific and intron-specific DNA probes as described previously (Lehmann and Lütke-Eversloh 2011).
Fermentation experiments and product analyses
Cultivation experiments for mutant characterization were performed in 200 ml MS-MES medium in serum bottles (Müller & Krempel AG, Bülach, Switzerland) with the following composition per liter: 0.55 g KH2PO4, 0.55 g K2HPO4, 0.22 g MgSO4 × 7 H2O and 0.011 g FeSO4 × 7 H2O; 2.3 ml of acetic acid was omitted if not stated otherwise in the text. After the pH was adjusted to 6.6 with ammonia solution, 40 μg p-aminobenzoic acid, 0.32 μg biotin, 1 mg resazurin and 21.3 g 2-(N-morpholino) ethanesulfonic acid (MES) were added (Monot et al. 1982). Glucose as carbon sources was added at a concentration of 60 g/l. The glucose concentration was measured enzymatically as previously described (Lehmann and Lütke-Eversloh 2011).
Precultures were inoculated with spore suspensions as described previously in the article and incubated overnight. The second and third precultures were conducted in MS-MES in Hungate tubes (10 ml medium) and serum bottles (50 ml medium), respectively. The inoculation volume was 10% for the precultures and for the main cultures; the initial optical density (OD600) was adjusted to 0.02–0.1. The OD600 and pH values were monitored during fermentation; cell-free supernatant samples were stored at −20 °C for further analyses.
Gas chromatography was employed to measure acetate, butyrate, acetone, ethanol and butanol in cell-free supernatant samples as described by Thormann et al. (2002) using an Agilent 7890A gas chromatograph (Agilent Technologies, Böblingen, Germany) equipped with a Chromosorb 101 (80/100 mesh, 2.0 m × 3.0 mm × 1.6 mm) glass column. Lactate was determined spectrophotometrically using a commercial d-/l-lactate kit according to the manual (K-DLATE kit, Megazyme International Ireland Ltd., Wicklow, Ireland).
HPLC (Thermo Fisher Scientific GmbH, Dreieich, Germany) was used to analyze acetoacetate (modified from Rumsby et al. 1987) using an Aminex HPX-87H column (300 × 7.8 mm; BioRad Labs, Hercules, USA) at 40 °C with a flowrate of 0.6 ml/min of 0.005 M sulfuric acid. Standards were prepared from Lithiumacetoacetate (Applichem Lifescience; A1941.0001). Peaks were detected with a diode array detector, and acetoacetate was quantified at a wavelength of 210 nm.
Generation and verification of C. acetobutylcium mutants
In silico restriction analyses of genomic DNA from C. acetobutylicum ATCC 824 and its mutants generated in this study
C. acetobutylicum strain
Lane Fig. 1
For the construction of double knock-out mutants, the retrotransposition-activated marker (RAM) conferring erythromycin resistance was removed from C. acetobutylicum pta::int(80) by FLP recombinase-mediated excision, resulting in strain C. acetobutylicum pta::int(80)ΔRAM. This strain was then used for mutagenesis employing plasmids pMTL007-adc and pMTL007-ctfA, respectively, and the mutants C. acetobutylicum pta::int(80)ΔRAM-adc::int(180) and pta::int(80)ΔRAM-ctfA::int(352) were subjected to Southern hybridization analyses. EcoRV/PaeI-restricted genomic DNA of the double mutants exhibited a 2.8-kbp restriction fragment with the pta-specific probe, whereas no corresponding signal was obtained with the RAM-specific probe due to the loss of the RAM cassette (Fig. 1d–e, Table 2). Hybridization with adc-specific, ctfA-specific and RAM-specific probes verified the effective disruption of adc and ctfA, respectively (Fig. 1f–i, Table 2). Moreover, hybridization with the RAM-specific probe did not result in two hybridization signals, confirming that the re-targeted plasmids were lost in the mutant strains (Fig. 1e, g, i).
To see whether the succession of the mutagenesis or the removal of the RAM cassette had an influence on the phenotype, a vice versa experiment was conducted: first, the RAM cassette was excised from C. acetobutylicum adc::int(180), and second, the pta gene was mutagenized, leading to strain C. acetobutylicum adc::int(180)ΔRAM-pta::int(80). The experimental details are summarized in the "Supplementary materials". This strain showed the same phenotype as C. acetobutylicum pta::int(80)ΔRAM-adc::int(180), indicating that the aforementioned influences were not relevant (Figs. S1 and S2).
It should be noted at this point that ClosTron-generated mutants are very stable; revertants have not been observed so far (Heap et al. 2007, 2010). For example, a ClosTron mutant of C. acetobutylicum was monitored by PCR, and the intron was shown to be firmly integrated for at least 10 days in continuous cultures without erythromycin addition (H. Janssen and R.-J. Fischer, personal communication).
Disruption of the acetate pathway
The generation of distinct C. acetobutylicum knock-out mutants was motivated by two major issues: (i) investigation of metabolic effects when one pathway is blocked, and (ii) evaluation of reduced byproduct formation to increase the flux towards butanol as a desired product. Therefore, the acetate pathway was targeted, and the phosphotransacetylase-encoding pta gene was disrupted.
Disruption of the acetone pathway
C. acetobutylicum typically shows an ABE ratio of 3:6:1, acetone being the second most fermentation product after butanol (Lee et al. 2008). Thus, modifying the acetone pathway should have a major impact on the cell's central metabolism. Constructing antisense RNA against the responsible genes, i. e. adc and ctfAB coding for the acetoacetate decarboxylase and the acetoacetyl-CoA:acyl-CoA transferase, respectively, indicated that CtfAB is the rate-limiting enzyme, because the knock-down of adc did not change acetone production in C. acetobutylicum (Tummala et al. 2003c). However, a TargeTron-based knock-out of the adc gene in the Chinese industrial strain C. acetobutylicum EA 2018 exhibited a drastically reduced acetone production, resulting in acetate accumulation and reduced butanol formation (Jiang et al. 2009). Fermentation experiments with C. acetobutylicum adc::int(180) and C. acetobutylicum ctfA::int(352) showed a similar phenotype, whereas the latter was more profound, i. e. acetone-negative in contrast to the acetone-leaky phenotype of the adc mutant. Both mutants showed elevated acetate levels, whereas butyrate was re-assimilated in the stationary growth phase (Fig. 2). Using MS-MES without previously added acetate, which triggers endospore formation (Monot et al. 1982), all strains including the wildtype produced acetate, but no significant re-uptake was observed. It should be noted that C. acetobutylicum ATCC 824 exhibited a clearly decreasing acetate concentration in the course of fermentation in MS-MES comprising 40 mM acetate (Lehmann and Lütke-Eversloh 2011). Cultivation of C. acetobutylicum adc::int(180) and C. acetobutylicum ctfA::int(352) in MS-MES with acetate showed a steadily increasing acetate level, but no re-utilization was detected (data not shown). All fermentation experiments with both adc and ctfA mutants revealed a distinct butyrate peak at the end of the exponential growth phase, demonstrating that butyrate is actually re-assimilated via an Adc/CtfAB-independent pathway. In accordance with previous results (Jiang et al. 2009), alcohol production was significantly lowered, although the ABE ratio was improved in favor of butanol, i. e. a final molar BE/A ratio of 17 for C. acetobutylicum adc::int(180) was achieved in comparison to 2.5 for the wildtype.
Combined disruption of the acetate and acetone pathways
A very traditional approach to get insights into microbial physiology is the generation of defective mutants and subsequent analyses of the mutations, which requires either a suitable selection system or a method for specific gene inactivation. Therefore, we used the ClosTron mutagenesis for targeted gene inactivation coding for central metabolic enzymes and analyzed the physiological effects in C. acetobutylicum.
Biosynthesis of acetate comprises the conversion of acetyl-CoA to acetate via acetyl phosphate and is a favored pathway of C. acetobutylicum, because it yields one additional ATP per acetyl-CoA in comparison to the analogous butyrate pathway. However, the accumulation of acids and the concomitant pH decrease forces the organism to counteract the increasing environmental toxicity. Thus, the metabolism alters the carbon flux towards the neutral solvents butanol and acetone, gaining some extra time to finish the sporulation process. Conceivably, batch cultures of mutant strains without pH control might struggle more with the rapid pH decrease than the wildtype, preventing the regular switch to solventogenesis. Except for the pta single mutant, all C. acetobutylicum mutants lacked the typical pH increase at the beginning of the stationary growth phase (Figs. 2 and 3). Therefore, high-throughput optimization of the fermentation conditions for the mutant strains is currently under investigation (Schmidt and Weuster-Botz 2012; Schmidt et al., manuscript in preparation).
Employing non-replicative plasmid integration, a pta as well as butyrate kinase (encoded by the buk gene)-negative mutants were constructed to study the acid biosynthetic pathways in C. acetobutylicum ATCC 824 (Green et al. 1996). Interestingly, neither of those mutants were clearly negative for the respective acids, although the amounts of acetate or butyrate were reduced with a shift towards the opposite acid pathways, respectively. In a subsequent publication, the pta mutant was reported to produce increased amounts of butanol (Zhao et al. 2005). C. acetobutylicum pta::int(80) generated in this study did not exhibit an increased butanol production in a defined mineral salts medium; the fermentation profiles were usually similar to those of the wildtype (Fig. 2). To monitor gene expression in the pta mutant, reverse transcription (RT) PCR was performed. Whereas the wildtype exhibited strong signals for both genes of the bicistronic operon coding for Pta and acetate kinase (Ack), cDNA samples of C. acetobutylicum pta::int(80) resulted in a faint PCR product which was enlarged by the intron size. The mutant also revealed significant but lower expression of the downstream located ack gene, indicating that Ack activity might still be present (Fig. S3).
Thus, the fact that the pta mutants did not show an acetate-negative phenotype is most likely due to compensating phosphotransbutyrylase (Ptb) and probably Buk activities in addition to residual Ack activity. Both enzymes were purified from C. acetobutylicum and exhibited a broad substrate range and significant activities with acetyl-CoA and acetate, respectively (Wiesenborn et al. 1989; Hartmanis 1987).
The acetone pathway is associated with the acid pathways, because re-assimilation of acetate and butyrate begins with the CoA activation catalyzed by CtfAB; the subsequent exergonic and irreversible decarboxylation of acetoacetate has been described as a driving force during solventogenesis (Jones and Woods 1986; Gheshlaghi et al. 2009). In contrast to the alcohol forming pathways, acetone synthesis does not involve a reduction step to regenerate NAD(P)+, a requirement to maintain the glycolytic flux. Thus, butanol and small amounts of ethanol are synthesized to compensate the redox balance. In theory, C. acetobutylicum can be transformed to a "homobutanol" producer, considering redox stoichiometry without the formation of molecular hydrogen (Lütke-Eversloh and Bahl 2011).
Targeting the acetone metabolic pathway for reduced byproduct formation constitutes an obvious intuitive strategy and has been performed previously. Tummala et al. generated antisense RNA constructs against adc and ctfB: whereas adc downregulation did not reduce acetone production, the ctfB antisense RNA plasmid conferred a drastically decreased solventogenesis (Tummala et al. 2003b, c). Additional overexpression of the bifunctional aldehyde/alcohol dehydrogenase gene adhE1 (aad, CAP0162) restored butanol production and led to very high ethanol titers (Tummala et al. 2003a; Sillers et al. 2009). In this study, C. acetobutylicum ATCC 824 adc and ctfA mutants generated by the ClosTron technology exhibited an acetone-leaky, i. e. 10% of the wildtype level, and an acetone-negative phenotype revealing 54% and 53% of the wildtype butanol titer (Fig. 4b). The product pattern of C. acetobutylicum adc::int(180) matches the results of a comparable recent study very well, in which a TargeTron-generated adc mutant of C. acetobutylicum EA 2018 produced 7% acetone and 54% butanol, respectively, of the parental strain in P2 medium (Jiang et al. 2009).
Regarding the acetone-leaky phenotype, one seeks an explanation for a tenth acetone production level although the adc gene was disrupted. Whereas the wildtype did not produce any detectable amounts of acetoacetate, cultures of C. acetobutylicum adc::int(180) exhibited a maximum of 2 mM acetoacetate during the course of fermentation, indicating that this intermediate is not accumulated (data not shown). Acetoacetate is a chemically instable compound, and its non-enzymatic decarboxylation occurs spontaneously; the rate increases with a concomitant decrease of the environmental pH value. This was recently demonstrated under clostridial cultivation conditions, including a putative adc mutant of C. beijerinckii generated by an integrational non-replicative plasmid targeted for the adc gene. However, this adc mutant of C. beijerinckii NCIMB 8052 did not exhibit a reduced acetone formation, and the phenotype was unexpectedly very similar to its parental strain, although the total ABE yield was lower than that of the parental strain (Han et al. 2011).
As a completely new approach, adc and ctfA mutations were combined with a pta disruption, yielding the first examples of metabolically engineered double knock-out mutants of solventogenic clostrdia with the goal of rational phenotypic combination and thus increasing the carbon flux towards butyryl-CoA. Unfortunately, none of the mutants exhibited an increased butanol production under the conditions applied in this work, indicating a complex relation between the branched fermentative pathways of C. acetobutylicum (Fig. 4). The fact that the double mutants still produced some butanol (and C. acetobutylicum pta::int(80)ΔRAM-adc::int(180) traces of acetone) indicated that the mutants did not degenerate, i. e. the majority of cells did not lose the megaplasmid pSOL1. Referring to the drastic pH decrease in the mutant cultures, optimization of the fermentation conditions might therefore improve the butanol production.
Nevertheless, the fermentation data of the five knock-out mutants presented here provide new insights into the complex physiology of C. acetobutylicum according to the metabolic changes of the respective mutants. Acetone-negative and acetone-leaky mutations combined with a pta gene disruption resulted in fact in a severely decreased acetate formation, although the pta mutation alone did not significantly alter the phenotype of C. acetobutylicum (Fig. 4b).
The question why the carbon flux goes towards butyrate and not towards butanol production remains to be elucidated. The regulatory circuits of the solventogenic shift are still not understood, but some redox-dependent mechanisms must be involved. As a recent example, a butyrate/butanol-negative mutant of C. acetobutylicum, which was generated by ClosTron-mediated disruption of the 3-hydroxybutyryl-CoA dehydrogenase gene, produced large amounts of ethanol, whereas the acetate and acetone pathways operated unchanged (Lehmann and Lütke-Eversloh 2011). This mutant indicated that butyryl-CoA or butyryl-phosphate are not necessarily required for high alcohol production (Desai et al. 1999; Harris et al. 2000; Zhao et al. 2005), and identification of the triggering mechanism to increase alcohol production will be a challenging scientific objective.
A different strategy to increase the butanol yield was conducted by using C. acetobutylicum M5, a degenerated strain lacking the megaplasmid-encoded genes ctfAB, adhE1, adhE2 and adc. Introduction of a plasmid harboring adhE1 into the solvent-negative M5 strain resulted in ethanol and butanol formation without acetone (Nair and Papoutsakis 1994; Sillers et al. 2008). Additional expression of the ctfAB genes along with adhE1 in strain M5 led to a higher alcohol production with only 20% of the wildtype acetone concentrations (Lee et al. 2009). All these recombinant M5 strains exhibited high amounts of acetate as a final product. The parental C. acetobutylicum M5 produces only acetate and butyrate, and an acetate kinase-negative mutant, M5 AKKO, showed reduced acetate formation. Interestingly, overexpression of adhE1 in M5 AKKO again significantly increased the acetate level, confirming a complex relationship between acid and alcohol biosynthetic pathways according to ATP formation and redox balance (Sillers et al. 2008; Lee et al. 2009). The lack of ctfAB and adc genes und thus the acetone pathway in the M5 strain is in agreement with the suggestion that acetate is re-utilized via the Adc/CtfAB-catalyzed pathway (Fig. 4a). Controlling the oxidoreduction potential of C. acetobutylicum DSM 1731 cultures was recently employed to reduce acetone and increase alcohol formation, and the lower acetone biosynthetic flux was accompanied by an inhibition of the acetate re-assimilation (Wang et al. 2011).
Finally, the results presented here provided experimental evidence that butyrate is not re-assimilated via the Adc/CtfAB-dependent pathway in C. acetobutylicum, in contrast to acetate. Our data clearly show a distinct re-uptake of butyrate in the fermentation experiments of both adc-negative and ctfA-negative mutants (Fig. 2). Early metabolic flux analyses of the acid formation pathways revealed that butyrate uptake fluxes did not correlate with acetone production, indicating that Ptb and Buk might be responsible for both butyrate synthesis and re-assimilation (Desai et al. 1999). In addition, overexpression of the ptb and buk genes in C. acetobutylicum resulted in decreased, not elevated butyrate concentrations (Walter et al. 1994). The reason for this phenomenon is unclear, because it does not seem reasonable for the organism to go a metabolic pathway back and forth, considering the energy consumption for butyrate uptake. We speculate that the acid accumulation and concomitant sharp pH decrease triggers a priority stress response to immediately reduce the butyrate concentration, even at the expense of ATP.
In conclusion, generation and fermentation profiles of C. acetobutylicum mutants with defects in the central fermentative metabolism provided new information on the physiology of solventogenic clostridia. However, disruption of competing biosynthetic pathways did not yield an improved butanol production, indicating strong relations of the complex metabolism. Intuitive pathway engineering in order to reduce byproduct formation and therefore enabling an increased carbon flux towards the desired product clearly shows naturally set limits. Thus, gaining more insights into the metabolism and its regulation constitutes a general requirement and provokes more basic research on solventogenic clostridia. As a contribution to this, we are currently performing global transcriptional analyses of the C. acetobutylicum mutants.
The authors thank Nigel P. Minton and John T. Heap, University of Nottingham for kindly providing the ClosTron plasmids, and P. Soucaille, Institut nationale des sciences appliquées de Toulouse for kind provision of plasmid pCLF1. Furthermore, experimental support by M. Schmidt and M. Klipp for conducting some of the fermentation experiments is gratefully acknowledged. This study was financially supported by the Süd-Chemie AG, Munich and the German Federal Ministry of Education and Research (grant no. 0315419A).
- Amador-Noguez D, Brasg IA, Feng XJ, Roquet N, Rabinowitz JD (2011) Metabolome remodeling during the acidogenic-solventogenic transition in Clostridium acetobutylicum. Appl Environ Microbiol 77:7984–7997Google Scholar
- Desai RP, Papoutsakis ET (1999) Antisense RNA strategies for metabolic engineering of Clostridium acetobutylicum. Appl Environ Microbiol 65:936–945Google Scholar
- Hartmanis MGN (1987) Butyrate kinase from Clostridium acetobutylicum. J Biol Chem 262:617–621Google Scholar
- Jones DT, Woods DR (1986) Acetone–butanol fermentation revisited. Microbiol Rev 50:484–524Google Scholar
- Mermelstein LD, Papoutsakis ET (1993) In vivo methylation in Escherichia coli by the Bacillus subtilis phage phi3T I methyltransferase to protect plasmids from restriction upon transformation of Clostridium acetobutylicum ATCC 824. Appl Environ Microbiol 59:1077–1081Google Scholar
- Monot F, Martin JR, Petitdemange H, Gay R (1982) Acetone and butanol production by Clostridium acetobutylicum in a synthetic medium. Appl Environ Microbiol 44:1318–1324Google Scholar
- Nair RV, Papoutsakis ET (1994) Expression of plasmid-encoded aad in Clostridium acetobutylicum M5 restores vigorous butanol production. J Bacteriol 176:5843–5846Google Scholar
- Nölling J, Breton G, Omelchenko MV, Makarova KS, Zeng Q, Gibson G, Hong Mei L, Dubois J, Qiu D, Hitti J, Aldredge T, Ayers M, Bashirzadeh R, Bochner H, Boivin M, Bross S, Bush D, Butler C, Caron A, Caruso A, Cook R, Daggett P, Deloughery C, Egan J, Ellston D, Engelstein M, Ezedi J, Gilbert K, Goyal A, Guerin J, Ho T, Holtham K, Joseph P, Keagle P, Kozlovsky J, LaPlante M, LeBlanc G, Lumm W, Majeski A, McDougall S, Mank P, Mao JI, Nocco D, Patwell D, Phillips J, Pothier B, Prabhakar S, Richterich P, Rice P, Rosetti D, Rossetti M, Rubenfield M, Sachdeva M, Snell P, Spadafora R, Spitzer L, Shimer G, Thomann HU, Vicaire R, Wall K, Wang Y, Weinstock K, Lai Peng W, Wonsey A, Xu Q, Zhang L, Wolf YI, Tatusov RL, Sabathe F, Doucette-Stamm L, Soucaille P, Daly MJ, Bennett GN, Koonin EV, Smith DR (2001) Genome sequence and comparative analysis of the solvent-producing bacterium Clostridium acetobutylicum. J Bacteriol 183:4823–4838CrossRefGoogle Scholar
- Sambrock J, Russell DW (2001) Molecular cloning: a laboratory manual, 3rd edn. Cold Spring Harbor Laboratory Press, Cold Spring HarborGoogle Scholar
- Schmidt M, Weuster-Botz D (2012) Reaction engineering studies of acetone-butanol-ethanol fermentation with Clostridium acetobutylicum. Biotechnol J. doi:10.1002/biot.201100418
- Sillers R, Chow A, Tracy B, Papoutsakis ET (2008) Metabolic engineering of the non-sporulating, non-solventogenic Clostridium acetobutylicum strain M5 to produce butanol without acetone demonstrate the robustness of the acid-formation pathways and the importance of the electron balance. Metab Eng 10:321–332CrossRefGoogle Scholar
- Soucaille P (2008) Process for chromosomal integration and DNA sequence replacement in Clostridia. International patent WO 2008/040387Google Scholar
- Wang S, Zhu Y, Zhang Y, Li Y (2011) Controlling the oxidoreduction potential of the culture of Clostridium acetobutylicum leads to an earlier initiation of solventogenesis, thus increasing solvent productivity. Appl Microbiol Biotechnol. doi:10.1007/s00253-011-3570-2
- Wiesenborn DP, Rudolph FB, Papoutsakis ET (1989) Phosphotransbutyrylase from Clostridium acetobutylicum ATCC 824 and its role in acidogenesis. Appl Environ Microbiol 55:317–322Google Scholar
- Xiao H, Gu Y, Ning Y, Yang Y, Mitchell WJ, Jiang W, Yang S (2011) Confirmation and elimination of xylose-metabolic bottlenecks in glucose-PTS-deficient Clostridium acetobutylicum to realize simultaneous utilization of glucose, xylose and arabinose. Appl Environ Microbiol. doi:10.1128/AEM.00644-11