Appl Microbiol Biotechnol (2009) 82:49–57 DOI 10.1007/s00253-008-1732-7
BIOTECHNOLOGICAL PRODUCTS AND PROCESS ENGINEERING
Enhanced 2,3-butanediol production by Klebsiella pneumoniae SDM Cuiqing Ma & Ailong Wang & Jiayang Qin & Lixiang Li & Xulu Ai & Tianyi Jiang & Hongzhi Tang & Ping Xu
Received: 21 August 2008 / Revised: 25 September 2008 / Accepted: 25 September 2008 / Published online: 24 October 2008 # Springer-Verlag 2008
Abstract Enhanced 2,3-butanediol (BD) production was carried out by Klebsiella pneumoniae SDM. The nutritional requirements for BD production by K. pneumoniae SDM were optimized statistically in shake flask fermentations. Corn steep liquor powder and (NH4)2HPO4 were identified as the most significant factors by the two-level Plackett– Burman design. Steepest ascent experiments were applied to approach the optimal region of the two factors and a central composite design was employed to determine their optimal levels. The optimal medium was used to perform fed-batch fermentations with K. pneumoniae SDM. BD production was then studied in a 5-l bioreactor applying different fed-batch strategies, including pulse fed batch, constant feed rate fed batch, constant residual glucose concentration fed batch, and exponential fed batch. The maximum BD concentration of 150 g/l at 38 h with a diol productivity of 4.21 g/l h was obtained by the constant residual glucose concentration feeding strategy. To the best
Cuiqing Ma and Ailong Wang contributed equally to this work. C. Ma : A. Wang : J. Qin : L. Li : X. Ai : T. Jiang State Key Laboratory of Microbial Technology, Shandong University, Jinan 250100, People’s Republic of China H. Tang : P. Xu Key Laboratory of Microbial Metabolism, Ministry of Education, School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, People’s Republic of China P. Xu (*) School of Life Sciences and Biotechnology, Shanghai Jiao Tong University, Shanghai 200240, People’s Republic of China e-mail:
[email protected]
of our knowledge, these results were new records on BD fermentation. Keywords Klebsiella pneumoniae . 2,3-Butanediol . Medium optimization . Fed batch
Introduction Gradual exhaustion of natural resources has led to the search for renewable resources for sustainable development and there has been an increasing interest in 2,3-butanediol (BD) because of its extensive application in varied fields such as fuel, chemical industry, food industry, and so on (Xiu and Zeng 2008). BD is a chiral compound with a high boiling point and a low freezing point, which is a colorless and odorless liquid at room temperature (Garg and Jain 1995). As an important starting material, BD can be used to produce valuable derivatives such as methyl ethyl ketone and 1,3-butadiene (Syu 2001). BD has been shown to have the potential to be used in the manufacture of printing inks, perfumes, fumigants, moistening and softening agents, explosives and plasticizers, and as a carrier for pharmaceuticals (Garg and Jain 1995). BD is produced via a mixed acid fermentation pathway and differs qualitatively and quantitatively depending on the strains and species of bacteria and their culture conditions. Acetoin (AC) is the main by-product during the fermentation. Many bacterial species such as Klebsiella pneumoniae (Lee and Maddox 1986), Klebsiella oxytoca (Afschar et al. 1993), Enterobacter cloacae (Saha and Bothast 1999), Enterobacter aerogenes (Zeng et al. 1990), and Bacillus polymyxa (de Mas et al. 1988) can secrete BD. Among all these strains, K. pneumoniae is one of the best organisms that have shown the potential for industrial BD production. K. pneumoniae mainly produces meso-BD and
50
is often used for the production of BD because of its more complete fermentation, broad substrate spectrum and cultural adaptability (Garg and Jain 1995). Several strategies have been widely used to enhance BD production, such as introducing superproductive strains, optimizing fermentation operating conditions, and building mathematical models (Garg and Jain 1995; Lee and Maddox 1986; Zeng et al. 1990; Syu et al. 1993). Afschar et al. (1993) achieved a BD concentration of 118 g/l using recycled cells and the diol (AC + BD) productivity amounted to 2.4 g/l h by K. oxytoca. Yu and Saddler (1983) obtained a diol concentration of 113 g/l using fed-batch operation with K. pneumoniae but the diol productivity was relatively low (0.94 g/l h). Although BD production has been improved, the concentration and productivity are not high enough for economical industrial production. Therefore, it is essential to further improve BD production by selecting high-yield strains or systematically optimizing the fermentation condition. In this study, a bacterium designated as K. pneumoniae SDM was isolated from orchard soil samples and demonstrated good potential for BD production. A sequential statistical experimental design was used to develop a suitable fermentation medium. Fed-batch fermentations were conducted using different feeding strategies in a 5l bioreactor and remarkable BD yield and productivity were obtained.
Materials and methods Bacterial identification Classic physiological characteristics were tested according to Bergey’s Manual of Determinative Bacteriology (Holt et al. 1994). 16S rRNA gene sequence was amplified according to standard procedures (Sambrook and Russell 2001) and compared to the sequences in the GenBank database through BLAST sequence analysis. Inoculum preparation K. pneumoniae SDM was maintained on agar slants containing the following medium: glucose 15 g/l, peptone 10 g/l, yeast extract 5 g/l, KCl 5 g/l, and agar at pH 7.0. The slants were incubated at 37°C for 12 h and then stored at 4°C. The seed culture was prepared by inoculating a full loop of cells from freshly prepared slants into 50 ml of the following medium: glucose 20 g/l, (NH4)2HPO4 5 g/l, MgSO4 0.3 g/l, KCl 1 g/l, pH 7.0. The cultivation was conducted in 300-ml shake flasks for 10 h with agitation (200 rpm, reciprocal shaker) at 37°C.
Appl Microbiol Biotechnol (2009) 82:49–57
Statistical experiment design and data analysis A three-step experimental design based on statistical methods was used to optimize the medium for BD production. The Plackett–Burman (PB) design, an efficient technique for medium-component optimization (Reddy et al. 2008), was firstly used to pick factors that significantly influenced BD production, and insignificant ones were eliminated to obtain a smaller, more manageable set of factors. It was based on the first-order model: X Y ¼ b0 þ bi Xi ð1Þ where Y is the response; β0 is the model intercept and βi is the linear coefficient, and Xi is the level of the independent variable. Each variable is represented at two levels, high and low, which are denoted by (+) and (−), respectively. Table 1 illustrates the levels of each variable used in the experimental design. SAS package (version 9.0, SAS Institute, Cary, USA) was used to direct the effects and statistical analysis of the variables. The fit of the regression model obtained was checked using the adjusted coefficient of determination Rsquared. The significance of each variable was determined using Student’s t test. In our experiments, the variables with confidence levels above 95% were considered as influencing BD production significantly. Following that, the steepest ascent was generated by the first-order empirical equation obtained by the PB design to move rapidly towards the neighborhood of the optimum response. The center point of the PB design was taken as the origin for the path of steepest ascent. Response surface methodology (RSM) was used to optimize the screened variables for enhanced BD production based on central composite design (CCD, Kennedy and Krouse 1999) with five coded levels. For statistical calculations, the relationships between the coded values and actual values are described by the following equation: Xi ¼
xi x0 i ¼ 1; 2; . . . ; k Δxi
ð2Þ
where Xi is the dimensionless coded value of the independent variable xi; xi is the actual value of that independent Table 1 The PB design for screening variables in BD production Factors (g/l)
Variables
Glucose CSLP (NH4)2HPO4 Sodium acetate KCl MgSO4 FeSO4·7H2O MnSO4·7H2O
X1 X2 X3 X4 X5 X6 X7 X8
Low level (−1)
High level (+1)
60 2 1 1 0.4 0.1 0 0
80 6 3 3 0.8 0.3 0.02 0.01
Appl Microbiol Biotechnol (2009) 82:49–57
51
variable; x0 is the real value of the independent variable xi at the center point and Δxi is the step change. The role of each variable, their interactions, and statistical analysis to obtain predicted yields is explained by applying the following quadratic equation: Y ¼ b0 þ
X
bi Xi þ
X
bij Xi Xj þ
X
bii Xi2
ð3Þ
where Y is the predicted response; β0 is the offset term; βi is the linear effect; βii is the squared effect; βij is the interaction effect, and Xi is the dimensionless coded value of xi. For the flask experiments, 5 ml of the seed culture were inoculated into 100 ml of the basal medium of fermentation in 500-ml flasks. The flasks were incubated at 37°C on a reciprocal shaker at 200 rpm. All experiments were repeated at least three times. Statistical and numerical analyses were carried out by means of the analysis of variance (ANOVA). The 3-D response surface and contour presentations were plotted using the SAS package. Batch and fed-batch fermentations Batch and fed-batch fermentations were conducted in a 5-l bioreactor (BIOSTAT B, B. Braun Biotech International GmbH, Germany) with 3-l initial medium. The seed culture prepared previously was inoculated (10%, v/v) into the optimized fermentation medium with initial pH 7.0. The cultivation was carried out at 37°C, stirring at 500 rpm, and airflow at 1.5 vvm. When the pH decreased to 6.0, it was maintained at 6.0 by automatic addition of 4 M H3PO4 or 6 M KOH using a program-controlled peristaltic pump. Feed rate of substrate was designed according to the modes of operation used and the glucose consumption rate. Samples were collected periodically to determine the biomass, glucose, and diol concentrations. Batch fermentation was carried out with a glucose concentration of 70 g/l. It was used to confirm the suitability of the model equation for predicting maximum BD production, to study the kinetic parameters for growth of K. pneumoniae SDM, and to understand its physiology in the bioreactor in order to optimize the feed strategy. All fed-batch fermentations were conducted with an initial glucose concentration of 70 g/l and the feeding substrate was pumped into the bioreactor using a computercoupled peristaltic pump. In pulse fed-batch fermentation, glucose was fed into the bioreactor at different pulse feeding times, when the residual glucose concentration decreased to 15–20 g/l. In constant feed rate fed-batch fermentation, when the residual glucose concentration decreased to 15–20 g/l, glucose solution (800 g/l) was pumped into the bioreactor at a feeding rate of 20 or 30 ml/
h. In constant residual glucose concentration fed-batch, the residual glucose concentration was maintained at 0–10, 20– 30, or 40–50 g/l by feeding glucose solution (800 g/l). In exponential fed-batch fermentation, the nutrient feeding rate was determined by Eq. 4 (Nor et al. 2001): F¼
m ðVX Þ0 expðmt Þ YX =S ðSi S Þ
ð4Þ
where F is the feeding rate; μ is the specific growth rate; (VX)0 is the biomass at start of feed; t is the culture time; YX/S is the theoretical cell yield on substrate; Si and S are substrate concentrations in the feeding solution and in the reactor, respectively. Analytical methods The glucose concentration of the samples was measured enzymatically by a bioanalyzer (SBA-40C, Shandong Academy of Sciences, P. R. China) after centrifuging and diluting to appropriate concentration. Dry cell mass concentration was estimated by measuring the optical density of the sample at 620 nm in a spectrophotometer (Ultrospec 2100 pro UV/Visible Spectrophotometer, Amersham Biosciences, Piscataway, USA) and by its correlation with the dry cell weight (DCW), obtained gravimetrically. Diol in the broth was extracted by ethyl acetate with the addition of isoamyl alcohol as the internal standard and then quantified using a GC system (Varian 3800, Palo Alto, USA) equipped with a flame ionization detector and a 30-m SPB-5 capillary column (0.32-mm inside diameter, 0.25μm film thickness; Supelco, Bellefonte, USA). The operation conditions were as follows: nitrogen was used as the carrier gas; the injector temperature and the detector temperature were both 280°C; and the column oven temperature was maintained at 40°C for 3 min, then raised to 240°C at a rate of 20°C/min. The software Star 6.0 Chromatography Workstation was used for data acquisition and evaluation. The concentration of the products was determined by calibration curves.
Results Bacterial identification The bacterium used in this study was isolated from orchard soil in Shandong Province, P. R. China. The strain with highest BD production was obtained by ion beam mutation and deposited at the China Center for Type Culture Collection (M 208097). Strain SDM was found to be a nonmotile, gram-negative, and straight rod-shaped bacteria. It did not grow at 10°C. Voges–Proskauer test was positive,
52
Appl Microbiol Biotechnol (2009) 82:49–57
as well as urease and lysine decarboxylase. However, oxidase reaction was negative, so were tests for methyl red test, liquefaction of gelatin, and indole production. It was capable of growing on glucose, lactose, sucrose, xylose, maltose, arabinose, rhamnose, mannitol, and KCN and could utilize citrate and malonate. The organism was also identified through BLAST analysis of the partial sequences of 16 S rRNA gene. It was 99% identical with some sequence of K. pneumoniae according to its 16 S rDNA sequence (1,506 bp). The sequence was deposited in the GenBank database with accession no. EU872412. Based on these results, strain SDM was identified as a strain of K. pneumoniae and designated as K. pneumoniae SDM.
Medium optimization Plackett–Burman experimental design PB design for a total of eight variables was used to identify which variables have significant effects on BD production (Table 1). The medium includes phosphate, acetate, Fe2+, Mn2+, and Mg2+, which significantly affect BD production (Garg and Jain 1995; Syu 2001; Qin et al. 2006). The medium also contained glucose as carbon source and corn steep liquor powder (CSLP) as nitrogen source according to the pre-experiments. The upper and lower limits of each variable were chosen according to the preliminary investigation of the limits of the variables. Table 2 represents the PB experimental design for 12 trials with two levels of each variable and the corresponding BD production. Table 3 shows the effects of the variables on the response and the significant levels. Based on the statistical analysis, the factors having the Table 2 The PB design variables (in coded levels) with BD production as response Run
1 2 3 4 5 6 7 8 9 10 11 12
Variable levels
BD (g/l)
X1
X2
X3
X4
X5
X6
X7
X8
+1 +1 −1 +1 +1 +1 −1 −1 −1 +1 −1 −1
−1 +1 +1 −1 +1 +1 +1 −1 −1 −1 +1 −1
+1 −1 +1 +1 −1 +1 +1 +1 −1 −1 −1 −1
−1 +1 −1 +1 +1 −1 +1 +1 +1 −1 −1 −1
−1 −1 +1 −1 +1 +1 −1 +1 +1 +1 −1 −1
−1 −1 −1 +1 −1 +1 +1 −1 +1 +1 +1 −1
+1 −1 −1 −1 +1 −1 +1 +1 −1 +1 +1 −1
+1 +1 −1 −1 −1 +1 −1 +1 +1 −1 +1 −1
23.68 21.41 26.89 23.80 22.14 27.52 28.67 25.61 11.22 10.18 22.37 9.20
Table 3 Effects and statistical analysis of variables Variable
Coefficient
Standard error
t value
P value
Intercept X1 X2 X3 X4 X5 X6 X7 X8
21.0575 0.3975 3.7758 4.9708 1.0842 −0.4642 −0.4308 1.0508 0.9108
0.6658 0.6658 0.6658 0.6658 0.6658 0.6658 0.6658 0.6658 0.6658
31.63 0.60 5.67 7.47 1.63 −0.70 −0.65 1.58 1.37
