In the latest developments related to the ABE fermentation process, acetone was considered to be indesirable co-product, whereas butanol is the main product of interest. Over the past few decades, various strategies have been developed to decrease the production of acetone and increase the production of butanol (Nair et al.
1994; Nair and Papoutsakis
1994; Harris et al.
2000; Sillers et al.
2008; Jiang et al.
2009; Sillers et al.
2009; Han et al.
2011). The intracellular conversion of acetone into isopropanol was an attractive alternative to avoid acetone excretion and produce a valuable alcohol. The C. beijerinckii NRRL B593 strain reduced acetone naturally thanks to a secondary-alcohol dehydrogenase (s-Adh) but the final titres of solvents by the NRRL B593 (George et al.
1983; Survase et al.
2011) were lower than those of the best ABE producers (Monot et al.
1982; Qureshi and Blaschek
In this study, we have constructed four plasmids harbouring the adh gene from C. beijerinckii NRRL B593 and the genes from C. acetobutylicum ATCC 824 that are part of the metabolic pathway from acetoacetyl-CoA to acetone. The cloned genes were successfully expressed in both E. coli BW25113 and C. acetobutylicum ATCC 824. In E. coli, the expression of ctfA and ctfB along with adh and thl genes allowed for the production of isopropanol. The lack of the adc gene did not prevent the decarboxylation of acetoacetate by E. coli harbouring pFC006, probably because of the instability of the molecule in acidic conditions (Hay and Bond
1967). The final concentration of isopropanol in cultures of E. coli strains expressing ctfA and ctfB genes (pTHL and pFC006 or pTHL and pFC007) was lower than those previously reported by other groups (Hanai et al.
2007; Atsumi and Liao
2008; Jojima et al.
2008; Yoshino et al.
2008). In our study, E. coli cultures were not optimised, but carried out with the purpose of checking the validity of each construct.
The plasmids were electroporated in C. acetobutylicum ATCC 824. The expression of adh gene allowed transformants to reduce natively produced acetoin and acetone to 2,3-BD and isopropanol, respectively. Either the D or L forms of 2,3-BD or a combination of both but no meso-2,3-BD was produced. The achiral HPLC used in the present study did not differentiate between the D and L enantiomers. Since the activity of s-Adh on acetoin had never been described, this result extends the range of substrates known for this enzyme (Ismaiel et al.
1993). Recently, the production of 2,3-BD by C. acetobutylicum transformants expressing an acetoin reductase (acr) from C. beijerinckii NCIMB 8052 was reported (Siemerink et al.
2011). The resulting strains also produced 2,3-BD but did not produced isopropanol. For future applications, the production of 2,3-BD by ATCC 824 transformants is still very far from that of Klebsiella pneumoniae (up to 150 g/L of 2,3-BD) (Ma et al.
Each transformant of ATCC 824 was characterised in a batch culture either with pH regulation at 5.0 or without pH regulation. All transformants of ATCC 824 and the wild type displayed higher solvent production levels when grown without pH-regulation. The solvent yield based on glucose consumption did not depend on the genetic modifications, but rather on the culture conditions (pH control or not). Acid assimilation was improved in the cultures without pH regulation, as also suggested by the increase of the C3 compound (acetone or isopropanol) productions. When the pH was not regulated, the pH value of the culture dropped below 5.0, increasing the concentrations of the protonated form of the acids. This has been associated with the onset of solventogenesis (Monot et al.
1984; Hüsemann and Papoutsakis
1988). Therefore, the high level of protonated acid forms in pH not-regulated cultures of ATCC 824 transformants might trigger solventogenesis at a lower concentration of total acids (protonated plus ionized) and drive more the carbon flux towards butanol or ethanol formation.
The expression of only the adh gene lowered total solvent production by ATCC 824(pFC002) compared to the wild type and the transformant harbouring the empty vector (pMTL500E). The lower solvent excretion by ATCC 824(pFC002) could be explained by the higher toxicity of isopropanol compared to acetone, as suggested by the octane/water partition coefficients (logKow) values i.e. 0.05 for isopropanol and −0.25 for acetone (Yaws and Sachin
1999). The logKow was reported to be a good estimation for solvent toxicity (Vermue et al.
1993; Heipieper et al.
1994), high logKow compounds are generally more toxic than compounds with lower value. It has to be noted that the final IBE concentration of ATCC 824(pFC002) cultures (16 g/L) was still higher than that of NRRL B593 cultures (13 g/L) suggesting that the solvent sensitivity is a strain-dependent characteristic.
Under all culture conditions tested, the overexpression of all genes encoding enzymes of the acetone route (ctfA, ctfB and adc), along with expression of adh gene, conferred to ATCC 824(pFC007) high solvent production rate and high final solvent titres. The use of thl promoter to control the expression of ctfA and ctfB genes initiated excretion of isopropanol before those of other solvents. Recently, (Lee et al.
2012) have developed a transformant comparable with ATCC 824(pFC007) in which expression of isopropanol pathway genes were controlled by two adc promoters. In batch culture with pH regulation at 5.0, the maximal end-concentration of IBE was only 17.1 g/L of which 6.1 g/L was isopropanol. The difference in solvent productions observed in the two studies might result from the culture mode applied. Fed-batch with gas stripping was used and was found to improve IBE production by 35.6 g/L (Lee et al.
2012) but this type of process has never been scaled up.
The role of each gene involved in the pathway from acetoacetyl-CoA to acetone in the enhancement of ATCC 824(pFC007) fermentation performances was clarified by expressing two derivative plasmids. The overexpression of the ctfA and ctfB genes increased both the speed and the extent of acid assimilation while the overexpression of the adc gene had a little effect (Table
This result indicates that decarboxylation of acetoacetate is not the real bottleneck. In a previous study on ABE production by ATCC 824, the overexpression of ctfA, ctfB and adc genes controlled by the adc promoter was studied at pH 5.5 (Mermelstein et al.
1993). As with our results, the combined overexpression of ctfA, ctfB and adc increased the solvent production by transformants, whereas expression of adc gene alone had little effect. Unlike our results, the combined expression of ctfA and ctfB genes without adc was found to have a limited effect. Therefore, the impact of ctfA and ctfB overexpression observed in our study might have been supported by the chemically acid-calalysed decarboxylation of acetoacetate (Hay and Bond