Thermophilic fermentation of acetoin and 2,3-butanediol by a novel Geobacillus strain
© Xiao et al.; licensee BioMed Central Ltd. 2012
Received: 12 September 2012
Accepted: 30 November 2012
Published: 6 December 2012
Skip to main content
© Xiao et al.; licensee BioMed Central Ltd. 2012
Received: 12 September 2012
Accepted: 30 November 2012
Published: 6 December 2012
Acetoin and 2,3-butanediol are two important biorefinery platform chemicals. They are currently fermented below 40°C using mesophilic strains, but the processes often suffer from bacterial contamination.
This work reports the isolation and identification of a novel aerobic Geobacillus strain XT15 capable of producing both of these chemicals under elevated temperatures, thus reducing the risk of bacterial contamination. The optimum growth temperature was found to be between 45 and 55°C and the medium initial pH to be 8.0. In addition to glucose, galactose, mannitol, arabionose, and xylose were all acceptable substrates, enabling the potential use of cellulosic biomass as the feedstock. XT15 preferred organic nitrogen sources including corn steep liquor powder, a cheap by-product from corn wet-milling. At 55°C, 7.7 g/L of acetoin and 14.5 g/L of 2,3-butanediol could be obtained using corn steep liquor powder as a nitrogen source. Thirteen volatile products from the cultivation broth of XT15 were identified by gas chromatography–mass spectrometry. Acetoin, 2,3-butanediol, and their derivatives including a novel metabolite 2,3-dihydroxy-3-methylheptan-4-one, accounted for a total of about 96% of all the volatile products. In contrast, organic acids and other products were minor by-products. α-Acetolactate decarboxylase and acetoin:2,6-dichlorophenolindophenol oxidoreductase in XT15, the two key enzymes in acetoin metabolic pathway, were found to be both moderately thermophilic with the identical optimum temperature of 45°C.
Geobacillus sp. XT15 is the first naturally occurring thermophile excreting acetoin and/or 2,3-butanediol. This work has demonstrated the attractive prospect of developing it as an industrial strain in the thermophilic fermentation of acetoin and 2,3-butanediol with improved anti-contamination performance. The novel metabolites and enzymes identified in XT15 also indicated its strong promise as a precious biological resource. Thermophilic fermentation also offers great prospect for improving its yields and efficiencies. This remains a core aim for future work.
Acetoin (3-hydroxybutan-2-one) is widely used in food, flavour, cosmetics, and chemical synthesis . Its reduced form, 2,3-butanediol, is not just limited to the manufacture of butadiene, or to its use as an antifreeze agent and a precursor of many synthetic materials including polymers and resins . With a heating value of 27,200 J/g, 2,3-butanediol compares favourably with ethanol (29,100 J/g) and methanol (22,100 J/g) for use as a liquid fuel or fuel additive. Dehydration of 2,3-butanediol yields the industrial solvent methyl ethyl ketone, which can be hydrogenated to yield high octane isomers, suitable for high quality aviation fuels.
Acetoin and 2,3-butanediol can be produced by bacterial fermentation using renewable biomass. Such platform chemicals are regarded as potential alternatives of non-renewable fossil materials . Both acetoin and 2,3-butanediol are metabolites of the acetoin metabolic pathway in bacteria. They can be transformed to each other by 2,3-butanediol dehydrogenase in cells. Acetoin and 2,3-butanediol sometimes coexist in fermentation broth. Because these two compounds are quite different in boiling points, they can be easily separated from each other by distillation.
The scientific understanding and the technological development leading to the production of acetoin and 2,3-butanediol have undergone four periods, evident from the number of the main publications. During the first period (1920s-1981), research work focused on product determination, the basic characterization of microbial properties and development of fermentation processes. Period II (1982–2001) addressed the main biological questions concerning pathways, enzymes, genes, and regulations . This active phase was then followed by a few quiet years (Period III). Resurging from 2008, the current period features three words: cost-effective [5–7], efficient-recovery [8–10], genomics-guided [3, 11]. But in general, fermentation study is one of the main themes throughout the entire development process.
To date, however, almost all these studies have only reported the production of acetoin and 2,3-butanediol from mesophilic fermentation processes or enzymatic reactions, typically at the temperature range of 30-40°C (for recent reviews, see [2, 12]). To avoid bacterial contamination during these biotransformation processes, culture media must be strictly sterilized at high temperature and pressure, requiring the installation of specialized sterilization facilities and provision of steam and cooling water . Operation of the sterilization facilities requires trained manpower and consumes energy. To worsen the situation, long time sterilization under high temperature causes serious nutrient loss and generates harmful chemicals that may in turn undermine normal fermentation.
In order to resolve these problems, thermophilic fermentation processes, typically operated at 50 – 60°C, have attracted attention over the past few years . Recently Wang et al. reported the first high temperature fermentation of 2,3-butanediol using a genetically engineered bacterial strain, showing its great potential . Because normal environmental microorganisms cannot reproduce above 45°C and vegetative cells are killed at higher temperature by pasteurization, thermophilic fermentation can thus reduce the risk of bacterial contamination. It is in fact possible to operate thermophilic fermentation even without sterilization , making it more efficient and cost-effective. In addition, thermophilic fermentation matches the simultaneous saccharification process which has the best efficiency around 55°C .
In this work, we report the isolation and characterization of a novel thermophilic bacterium for the production of acetoin and 2,3-butanediol and the related investigation of its novel metabolites and enzymes, aiming at developing a more efficient and effective fermentation process utilizing the thermophilic strain.
Physiological characteristics of strain XT15
Hydrogen sulfide production
The effect of DO on bacterial growth and product yield was tested by either changing the medium loading volume in shaking flasks or comparing aerobic and anaerobic cultivation. Loading 30 or 20 mL medium per flask did not alter the yield of acetoin. This was because when DO was above a certain level (the critical DO concentration), changes in DO concentration would not affect production any longer. However, under anaerobic condition, no growth or product was detected during a one-week period cultivation of XT15.
Peaks 1, 2, 7, and 9 could be easily recognized as diacetyl, acetoin, 2,3-butanediol, and 2,3-butanediol, respectively, in gas chromatography–mass spectrometry (GC-MS) identification. There are 3 stereoisomers of 2,3-butanediol: (2R, 3R)-, (2S, 3S)-, and meso-2,3-butanediol. The first two kinds are racemic mixtures and they can only be separated from each other in a chiral column . But the racemic mixtures (or just one of the two racemic stereoisomers) can be easily separated from the meso form by ordinary non-chiral GC capillary columns.
Ammonia was generated in the medium by the oxidative deamination of amino acids. Nitrogen-containing heterocyclic compounds could be formed from acetoin and ammonia by non-biological processes . Control experiments without inoculum (but adding 10 g/L of acetoin and 10 g/L of diammonium phosphate in the medium) were performed. Identical peaks of 2,3,5,6-tetramethylpyrazine (TTMP) and an oxazoline (1-(2,4,5-trimethyl-2,5-dihydrooxazol-2-yl)ethanol, OXZ) were detected as Shu and Lawrence reported  in both experiments (with or without inoculum). OXZ produced multiple peaks because it has 3 chiral carbon atoms in its molecule. Therefore, peak 6 and peaks 10–13 in Figure 8 were identified as TTMP and OXZ, respectively. Peak 4 was identified as 2,3,5-trimethylpyrazine (TMP), a common by-product of TTMP . As mentioned by Larroche et al. , 1-aminopropan-2-one could be generated from L-threonine. Then 1-aminopropan-2-one, acetoin, and ammonia would synthesize TMP spontaneously in aqueous solution.
Identification results of volatile products
Common name or abbr.
IUPAC systematic name
The spontaneous cleavage (SC) of α-acetolactate should not be neglected as shown in Figure 10, especially when the temperature was above 50°C. When the activity of SC was taken into account, α-acetolactate could be transformed into acetoin with maximum efficiency at 50°C. Therefore, the total cleavage activity of α-acetolactate remained high in the temperature range from 28 to 65°C in Geobacillus sp. XT15.
By isolation from different environments  or genetic manipulations [14, 22, 23], dozens of species from various genera, such as Bacillus[24, 25], Klebsiella, Enterobacter, Serratia, and Paenibacillus[29, 30] have been obtained for potential industrial production of acetoin and/or 2,3-butanediol. However, none of these strains is a naturally occurring thermophile. To our best knowledge, XT15 is the first naturally occurring thermophilic acetoin and/or 2,3-butanediol producer. It is thus of great value to both practical application and theoretical study.
During enhanced oil recovery process, additional water is often injected into reservoirs to help force the crude oil to the surface. Both the injected water and the natural water lying under the hydrocarbons are eventually produced along with the oil. Therefore, the bacteria collected from PW may come from ground or underground. In recent years, Bacillus strains, which were known as acetion producers, were also isolated from the PW of Shengli Oilfield, China. Geobacillus species were regarded as a branch of the Bacillus genus before Nazina et al. reclassified them in 2001 . Strain XT15 shared most physiological characteristics with common Geobacillus species, including the ability of hydrocarbon utilization. But XT15 is also capable of acetoin excretion (positive in Voges-Proskauer test), which is unusual in the genus Geobacillus. The record product yield achieved was at 82% of the theoretical value. In other words, 82% of glucose was converted into acetoin, indicating that the acetoin metabolic pathway is a dominant pathway in Geobacillus sp. XT15. This would distinguish XT15 from other Geobacillus species. The substrate conversion rate of XT15 was equivalent to that of Bacillus subtilis CICC 10025 .
According to Nazina et al. , Geobacillus species are chemoorganotrophic, obligately thermophilic, aerobic or facultatively anaerobic, Gram positive, rod-shaped bacteria. Most species are widely distributed in nature. Oxygen is the electron acceptor, replaceable in some species by nitrate. The growth-temperature range is 37-75°C, with an optimum at 55-65°C. Growth occurs in a pH range of 6.0 to 8.5, with an optimum at pH 6.2-7.5. Therefore, the optimum temperature of XT15 is relatively lower, but the pH range is a bit higher.
Optimization of culture medium components, especially carbon source and nitrogen source, is not only yield-relevant but also cost-effective. Acetoin and 2,3-butanediol are important primary metabolites like ethanol, acetate, and lactate in various microorganisms and the excretion of them has vital physiological implications . Since production of acetoin and 2,3-butanediol is generally accepted to be a growth-associated phenomenon, the conditions for the maximum product formation must approximately be the same as those for the maximum biomass yield .
The influences of fermentation ages on acetoin and 2,3-butanediol production were also studied (unpublished data). The maximum yield could usually be achieved 1–3 h after the time point when the carbon source was depleted in the culture medium. But if the carbon source was excessive and could not be used up, the yield usually remained relatively stable after it grew to a certain extent. However, these finding were not always repeatable. The quality of seed culture had great influence on the maturation age of a fermentation process. But generally the time point of the maximum product yield was approximately the same as that for the maximum biomass yield. In Figures 4, 5, and 6, when 20 g/L of glucose was used as the carbon source, glucose could usually be depleted in 8 h and the maximum product yield and biomass yield could be achieved at about 10 h. In Figure 7 when 220 g/L of glucose was used, maximum yields could be achieved at about 48 h. All these findings again supported the notion that bacterial production of acetoin and 2,3-butanediol is a growth-associated phenomenon.
A noticeable observation is that the ratios between acetoin and 2,3-butanediol in the cultivation broth were variable. This could be explained by the physiological state of the bacteria, which would be substantially influenced by the culture medium, especially the content of the carbon source. When the concentration of glucose was low (as shown in Figures 4, 5, 6), the product would be acetoin. As the concentration of glucose was elevated to a certain higher level, 2,3-butanediol would arise as another important product. This phenomenon could be linked to the cellular NADH/NAD levels. The reversible transformation between acetoin and 2,3-butanediol couples with the NADH/NAD conversion, participating in the regulation of the cellular NADH/NAD levels [32, 33]. Glucose not only provides energy and anabolic materials but also sustains a reducing cellular environment, protecting cells from oxidative stress. When glucose was abundant and the system was maintained under a highly reducing environment, XT15 cells yielded 2,3-butanediol to store the reduced force. As glucose was kept at a low level or depleted, the product would be acetoin and even the accumulated 2,3-butanediol would be oxidized into acetoin to generate NADH and maintain NADH/NAD homeostasis .
Including metabolites and non-enzymatic products, a total of 14 substances (lactic acid and 13 volatile products) were identified from the fermentation broth of Geobacillus sp. XT15 when this strain was cultivated in media containing glucose, peptone and yeast extract. TMP, TTMP, and OXZ were believed to be non-enzymatic products. As for metabolites, XT15 shares great similarity with certain Bacillus species. In addition to acetoin, 2,3-butanediol, diacetyl and acetylbutanediol, some Bacillus strains can produce mixtures of lactic, isovaleric, isobutyric and acetic acids . The trace amount of phenylethyl alcohol detected in the cultivation broth would be the product of L-phenylalanine, as in certain microorganisms [35, 36]. As for the novel product butyrylbutanediol, the interpretation of the formation process could be inspired by the formation of acetybutanediol . Butyryl-CoA is readily generated from butanoic acid by CoA-transferase or hexanoic acid by the beta-oxidation pathway. Then it is postulated that butyryl-CoA and acetoin produced butyrylbutanediol by AoDH ES E1, which was also detected in this study.
The first step of the acetoin biosynthesis pathway is the reaction from pyruvate to α-acetolactate catalysed by two enzymes: the catabolic α-acetolactate-forming enzyme and the anabolic α-acetolactate-forming enzyme . However, there is only one enzyme catalysing the transformation from α-acetolactate to acetoin, i.e. ALDC. Therefore, ALDC was regarded as the key enzyme and the marker of the acetoin biosynthesis pathway. Compared to the published work [37, 38], the ALDC of XT15 obtained a higher optimum temperature and a wider active range at higher temperatures. When glucose was depleted in the medium, AoDH ES could be induced by acetoin and then acetoin would be reutilized as a substitutional carbon source. The initial step of acetoin breakdown was catalysed by AoDH ES E1 in the catabolic pathway of acetoin in bacteria . Only AoDH ESs from mesophilic bacteria were characterized before Payne et al.  reported the AoDH ES from a hyperthermophilic archaeon Sulfolobus solfataricus with an optimum temperature of 80°C. This study thus provided an example of a moderately thermophilic AoDH ES. ALDC and AoDH ES E1 were both active in Geobacillus sp. XT15, indicating that the acetoin metabolic pathway was active in this strain. The optimum temperatures and active temperature ranges of the two key enzymes were consistent with those of growth temperatures, respectively, indicating that the acetoin metabolic pathway originally existed in this bacterium. The acetoin metabolic genes in XT15 could have experienced similar evolutionary processes or even affected with each other with Bacillus by horizontal gene transfer, in consideration of the great similarities between Geobacillus sp. XT15 and some Bacillus species.
A novel thermophilic acetoin and 2,3-butanediol producer XT15 was isolated. It is an aerobic, gram positive, rod-shaped Geobacillus bacterium. At its optimum growth temperature range of 45-55°C, XT15 cells would reproduce every 2 h. The optimum initial pH of the culture medium is 8.0. XT15 prefers organic nitrogen sources like CSLP and has the potential to utilize cellulose hydrolysates. 14 products were identified in its cultivation broth, including a novel metabolite. Its two key enzymes in acetoin biosynthesis pathway and catabolism pathway shared identical optimum temperature of 45°C and wide active temperature range. As a novel naturally occurring thermophilic strain for the production of acetoin and/or 2,3-butanediol, there is potential to enhance its yields in further studies.
Composition of CSLP used in this study
GPY (glucose 20 g/L, peptone 20 g/L, yeast extract 10 g/L, deionized water, pH 7.0) or modified GPY media were used in this study. Glucose was autoclaved separately. The oil well PW from Shengli Oilfield (Dongying, China) was used to replace deionized water for the preparation of screening culture medium (PW-GPY medium). The PW-GPY medium was loaded into Erlenmeyer flasks and heated to 100°C rapidly. Without delay the flasks were then transferred into a shaking bath controlled at 60°C. After 2 d enrichment, the culture was streaked onto GPY agar and incubated at 60°C for strain isolation. Colonies larger in diameters and with deeper red colour in Voges-Proskauer tests were picked up and inoculated into GPY medium for final screening.
Physiological characterization was done using a bacterial biochemical identification kit (Qingdao Hope Bio-Technology Co., Ltd., China) according to the manufacturer’s instructions. The partial 16S rRNA gene sequence of the isolate was routinely amplified using universal primers 27 F and 1492R. The PCR products were purified and sequenced using an ABI 3730 DNA Analyzer (Applied Biosystems) by Shanghai Sangon (China). Searches of the GenBank database were conducted by using BLAST . 16S rRNA sequences of representative homologous species were retrieved from GenBank to generate the phylogenetic tree (1,000 bootstrap replicates) using the maximum composite likelihood method and the neighbour-joining method in the MEGA5 program .
Bacterial growth of XT15 on crude oil was tested in a MS medium containing 500 mg/L of crude oil (from Shengli Oilfield, Dongying, China), 2.28 g/L of K2HPO4·3H2O, 0.47 g/L of NaH2PO4·2H2O, 1.32 g/L of (NH4)2SO4, 0.12 g/L of MgSO4, and 1 ml/L of trace element solution (2.63 g/L of CaCl2, 0.72 g/L of FeSO4·7H2O, 0.46 g/L of ZnSO4·7H2O, and 0.22 g/L of MnSO4·H2O). To avoid the entrainment of nutrients from inoculum, the cells in the seed culture were harvested by centrifugation and washed with sterile saline before inoculated into the MS medium. Control experiments were performed using heat-inactivated XT15 cells.
Cells from slant cultures were inoculated into a pH modified GPY medium (initial pH changed from 7.0 to 8.0) to prepare the seed culture. After 6–8 h incubation at 55°C with agitation, the grown seed culture was inoculated (5%, v/v) into appropriate media for optimization experiments including temperature, medium initial pH, DO, carbon source, and nitrogen source. Unless otherwise noted, all the optimization experiments were performed in triplicates using 100 mL Erlenmeyer flasks (loading 30 mL medium per flask) in a reciprocal shaking water bath controlled at 170 rpm and 55°C. For anaerobic cultivation tests, 200 mL serum bottles fully filled with medium were applied and stirrer bars were placed into the serum bottles to keep the bacterial cells well suspended.
Daily cell observation was performed using a light microscope (DMI3000B Microscope, Leica). Scanning electron microscopy analysis was performed routinely  using JSM-6390LV (JEOL). Bacterial growth was measured spectrophotometrically at 600 nm (OD600) and dry cell weight (DCW) was calculated from OD600 with a linear correlation factor (1 OD600 = 0.39 DCW g/L). The pH values of the cultivation broth were determined using pH 211 (HANNA instruments). Glucose and lactic acid concentrations were enzymatically determined using an SBA-40C biochemical analyser (Biology Institute of Shandong Academy of Sciences, China). Acetoin and 2,3-butanediol concentrations in the fermentation broth were analysed periodically with an Agilent 7890A GC equipped with a flame ionization detector and a 30-m HP-5 capillary column (HP Agilent). Prior to GC analysis, the samples were extracted using ethyl acetate with the addition of isoamyl alcohol as the internal standard. The operating conditions were as follows. Nitrogen was used as the carrier gas. The injector temperature and detector temperature were both set at 280°C. The column oven was maintained at 40°C for 2 min and then programmed to increase to 130°C at a rate of 20°C/min.
In GPY medium fewer kinds of volatile products were detected. In order to make more metabolites detectable, XT15 was cultured in an enhanced GPY medium with elevated carbon substrate and nitrogen substrate concentrations (glucose 220 g/L, peptone 80 g/L, yeast extract 10 g/L). After 48 h of cultivation at 170 rpm and 55°C, 150 mL of fermentation broth was centrifuged at 8000 g for 5 min and then 30 g of NaCl was added into the supernatant. Then the supernatant was adjusted to pH 2.0 using 6 M hydrochloric acid before extraction with 20 mL of dichloromethane. The organic phase was concentrated 10 times before GC-MS analysis using Agilent 7890A-5975C equipped with a 30 m HP-INNOWax Polyethylene Glyco column (0.25 mm inside diameter, 0.25 μm film thickness). Helium was used as the carrier gas. The column oven temperature was initially maintained at 40°C for 5 min and then ramped to 200°C at 15°C min-1, maintained at 200°C for 1 min and then ramped to 230°C at 20°C min-1. Mass spectra in the electron impact mode were generated at 70 eV and scan mode in the range of 20–400 amu. The compounds were tentatively identified by comparing their mass spectra with those contained in the MS data system (NIST08.L library). For the further identification of unknown products, high-resolution GC-MS in CI mode was also performed using Waters GCT Premier.
Strain XT15 was cultured in GPY medium (20 g/L of glucose) and samples were collected at different stages for the analysis of different enzymes. After culturing for 5 h (glucose was not depleted at this time), the cells were harvested by centrifugation for the determination of ALDC activity. About 5 h after glucose was depleted in the culture medium, the cells were harvested for the determination of AoDH ES E1. The harvested cells were resuspended in 50 mM phosphate buffered saline (pH 7.0) and disrupted by sonication. The disrupted cells were centrifuged at 10,000 g for 20 min at 4°C, and the supernatant was used as the crude cell extract. ALDC activity was assayed by measuring the amount of acetoin generated from α-acetolactate according to Stormer . The creatine-naphthol colorimetric method was applied for acetoin measurement at 522 nm using a UV/visible spectrophotometer (DR 5000, HACH). The formation of acetoin in the enzymatic reactions was also confirmed by the GC method. AoDH ES E1 was assayed by measuring the reduction of DCPIP at 578 nm according to Steinbüchel et al. . All the enzymatic experiments were performed in triplicates and the reactions without the crude extracts were set as blanks.
Scanning electron microscopy
Corn steep liquor powder
Gas chromatography–mass spectrometry
Total ion chromatogram
DCPIP OR or AoDH ES E1: Acetoin:2,6-dichlorophenolindophenol oxidoreductase
We thank the National and the Shandong Province Science Foundation (30800012, ZR2011CQ048) and the Fundamental Research Funds for the Central Universities of China for funding support and the Royal Society (London) for a Sino-British Fellowship.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.