Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Low-cost effective culture medium optimization for d-lactic acid production by Lactobacillus coryniformis subsp. torquens under oxygen-deprived condition

Low-cost effective culture medium optimization for d-lactic acid production by Lactobacillus... Lactic acid is considered a commodity and its production is boosted by the synthesis of polylactic acid. D-lactic acid (DLA) isomer offers greater flexibility and biodegradability and it can only be obtained in its pure form through fermentation. The lactate dehydrogenase is stereospecific for homofermentative production of DLA isomer in the metabolic pathway of Lactobacillus coryniformis subsp. torquens, with optical purity of ≥ 99.9% under oxygen-deprived condition. A simple culture medium that increases DLA production and reduces fermentation costs is fundamental for industrial applicability. A central composite rotatable design was used to evaluate significant components influencing the DLA production. Concentrations were adjusted using the Design-Expert 7.0 optimization tool with a desirability coefficient of 0.693 and the best concentrations of each component were determined. Finally, an assay in the bioreactor with the modified culture medium resulted in a product yield of 0.95 g/g, volumetric productivity of 0.85 g/L.h and 95% of efficiency. . . . Keywords Fermentation Optical purity Central composite rotatable design Optimization tool Introduction et al. 2013). The world demand for lactic acid production is forecasted to reach over one million metric tons by the Lactic acid (LA) was discovered in 1780 by chemist Carl year 2020 with annual growth of 5–8% (Jem et al. 2010; Wilhelm Scheele. It is an important hydroxycarboxylic Abdel-Rahman et al. 2013). The application of lactic acid acid with applications in the food, pharmaceutical, cos- in the production of biodegradable biopolymer poly-lactic metics, textile, and chemical industries (Abdel-Rahman acid (PLA) is growing due to its biodegradability charac- teristics (Nampoothiri et al. 2010). Lactic acid has an asymmetric β-carbon as chiral cen- ter that produces two enantiomers: L(+) lactic acid (LLA) * Lizeth Jaramillo and D(−)lactic acid (DLA). While LLA and DLA are only lizethacevedo@eq.ufrj.br obtained by microbial fermentation processes, racemic DL- lactic is always produced via chemical synthesis (Li and Laboratory of Biocorrosion, Biodegradation and Biosynthesis, Cui 2010). Depending on the microorganism and the op- School of Chemistry, Federal University of Rio de Janeiro, Av. erating conditions employed, an optically pure product Horacio Macedo, 2030, Bloco E, Sala E-109, CT, Cidade L(+) or D(−) and in some cases racemic mixtures can be Universitária - Ilha do Fundao, Rio de Janeiro, RJ CEP: 21949-900, produced (Narayanan et al. 2004). Brazil The chemical, physical, and biological properties of Nucleus of Ecology and Socio-Environmental Development of PLA are determined by the isomeric composition, pro- Macaé, NUPEM, Federal University of Rio de Janeiro, Av. São José do Barreto, 764 -São José do Barreto, Macaé, RJ CEP: 27965-045, cessing temperature, annealing time, and molecular Brazil weight. Polymerization of LLA results in poly-L(+) lactic Department of Biochemical Engineering. School of Chemistry, acid (PLLA) and polymerization of DLA results in poly- Federal University of Rio de Janeiro, Av. Horacio Macedo, 2030, D(−) lactic acid (PDLA). Due to the stereo regular chain Bloco E, Sala E-203, CT, Cidade Universitária - Ilha do Fundao, Rio microstructure, optical pure PLLA and PDLA are semi de Janeiro, RJ CEP: 21949-900, Brazil 548 Ann Microbiol (2018) 68:547–555 crystalline. The majority of commercial PLA is poly production. Various studies have reported the use of cheap (meso-lactide), a mixture of LLA (> 95%) and DLA (< raw material for DLA production (Tanaka et al. 2006; 5%) (Nampoothiri et al. 2010). Nakano et al. 2012; Nguyen et al. 2013; Cingadi et al. Poly (meso-lactide) can be used in a wide range of appli- 2015). However, few studies have showed the effect of medi- cations. However, this type of PLA has no stereochemical um components in the production of DLA. Such studies are structure and is highly amorphous. Also, the end products important to establish a simple medium, with few components made from this PLA are not suitable for high temperatures and low concentrations that offers a good balance between the applications (Shen et al. 2010). The physicochemical proper- nutritional needs of the microorganism and the economic vi- ties of optically active PDLA and PLLA are nearly the same, ability of the process. whereas the racemic PLA has very different characteristics. Response surface methodology (RSM) is a collection of Properties of PLA biopolymers can be modified by varying statistical techniques and tools for constructing a functional the proportions of the enantiomers. For example, the melting relationship between a response variable and a set of design temperature of PLLA can be increased to 40–50 °C and its variables used in studies of developing, improving, and opti- heat deflection temperature can be increased from approxi- mizing processes (Myers et al. 2009). This experimental meth- mately 60–190 °C by physically blending the polymer with odology has been successfully used in optimization studies for PDLA (Nampoothiri et al. 2010). The crystallization can be different biotechnological processes (Dubey et al. 2011;Zhu decreased with the presence of DLA isomer, resulting in et al. 2012; Kongruang and Kangsadan 2015;Dasguptaetal. amorphous polymers with controlled degradation, increasing 2013; Cingadi et al. 2015; Abdelwahed et al. 2017;Almeida et biocompatibility and, therefore, better application in biomed- al. 2017, Imran et al. 2017; Ojha and Das 2018). There are ical fields. Hence, the production of PDLA has gained impor- different RSM experimental designs, and in this study, a sta- tance and the optical purity of DLA plays a crucial role in tistical optimization of the producing medium was performed determining the characteristics of the PLA biopolymer. by employing a central composite rotatable design (CCRD) in High-purity LLA or DLA monomers are necessary to order to obtain the maximum amount of reliable information achieve the desired polymer properties (Li and Cui 2010). with the fewest number of experiments as well as to achieve Nowadays, most lactic acid is produced by microbial fermen- high production of optically pure DLA using low concentra- tation and many studies have reported optical pure LLA pro- tions of components that influence the DLA production. duction at industrial scale. On the other hand, production of optically pure DLA is not well established (Cingadi et al. 2015). The fermentative production has the advantage of giv- Materials and methods ing the ability to choose a strain of homofermentative lactic acid bacteria (LAB) that will produce only one optical isomer Microorganism and inoculum preparation with high purity. Few wild type strains have been studied for the production of optical pure DLA (Cingadi et al. 2015). Lactobacillus coryniformis subsp. torquens (ATCC 25600) LAB are Gram-positive, non-motile, non-spore forming was employed in this study, and it was acquired from the bacteria, and most of the group are cocci. They are faculta- Collection of Institute Pasteur (CIP). This strain was selected tive anaerobes, catalase negative, with high acid tolerance among four Lactobacillus strains after preliminary tests where and are generally recognized as safe (GRAS). LAB are un- its homofermentative characteristic for production of DLA able to synthesize ATP by respiration and have complex with high optical purity was verified by chiral chromatogra- nutrient requirements due to their limited ability to synthe- phy (unpublished data). size B vitamins and amino acids (Hofvendahl and Hahn- Cells were activated in 100 mL sealed anaerobic bottles Hägerdal, 2000; Gao et al. 2011). The Lactobacillus strain containing 50 mL of MRS medium. The anaerobic bottles needs complex nitrogen sources, vitamins, and minerals were inoculated with 1 mL of the stock culture, which was (Narayanan et al. 2004; Wee et al. 2006;Liand Cui 2010). preserved in glycerol and stored at − 80 °C. The inoculum was Homo-fermentative Lactobacillus produces LA as the main prepared in anaerobic bottles containing 100 mL of MRS product of the glucose metabolism. Through the Embden- medium and the inoculation volume was 10% of exponential- Meyerhof-Parnas pathway, 1 mol of hexose is converted to ly growing cells from the activation step. 2 mol of lactic acid and 2 mol of ATP (Hofvendahl and The medium for activation and inoculum preparation Hahn-Hägerdal 2000;Li and Cui 2010). contained the follow components: glucose, 10 g/L; peptone, Due to the nutritional requirements of Lactobacillus 10 g/L; yeast extract, 5 g/L; meat extract, 1 g/L; ammonium strains, the fermentation process may have high costs associ- citrate, 2 g/L; sodium acetate, 5 g/L; dipotassium phosphate, ated to the raw materials used in the process. The use of 2 g/L; magnesium sulfate, 0.1 g/L; manganese sulfate, 0.05 g/ cheaper feedstock as carbon source along with higher produc- L; and Tween-80, 1 g/L. Glucose was aseptically added to the tivity is a promising strategy to reduce the cost of DLA medium after autoclaving. All culture media were flushed Ann Microbiol (2018) 68:547–555 549 with nitrogen gas for 15 min to remove oxygen and (CCRD). The experimental data allowed describe the relation- autoclaved at 121 °C for 15 min. All liquid cultures were ships the DLA production and the medium components by a incubated in rotary shaker at 37 °C and 120 rpm. mathematical model following second-order polynomial equation: Fermentation assays Y ¼ β þ ∑β X þ ∑β X þ ∑β X X ; i i j o i ii i ij ð1Þ i≠j; i; j ¼ 1; 2; 3;…:k Shake flask fermentation were performed for the development of experimental designs. All experiments were carried out in where Y is the predicted response (DLA production, g/L), β is 100 mL anaerobic flasks containing 50 mL of producing me- the interception coefficient, β is the linear term, β is the i ii dium with different concentrations according to the experi- quadratic term, β is the interaction term, and X , X represent ij i j mental design. The initial pH was adjusted to 6.8, and glucose the independent variables. Experimental data from the CCRD was aseptically added to the medium after autoclaving. were subjected to analysis of variance (ANOVA). Analysis of Anaerobic fermentations were performed at 37 °C and the model, calculation of the predicted responses, and the 120 rpm under uncontrolled pH condition for 12 h. All exper- plotting of surface plots were done using Design-Expert® iments were inoculated with cells at the exponential growing software version 7.0. The final adjustment of the model was phase using inoculum size 10% (v/v) with initial cells concen- done using the Design-Expert 7.0 optimization tool consider- tration around 0.13 g/L. ing as main criteria reduction of component concentrations The fermentation media resulted of the shake flask exper- and maximized DLA production. iments were used for production of DLA at bioreactor scale. Batch fermentation was carried out in a 2 L bioreactor (New Calculation of kinetic parameters Brunswick Bio Flo®) with a working volume of 800 mL and the same conditions from the shake flask experiments. The pH The yield of DLA was calculated as the amount of DLA pro- was automatically controlled at 6.8 by the addition of 4 M duced from 1 g of consumed glucose. Productivity was deter- NaOH and the fermentation time was 46 h considering that mined as the concentration of DLA per fermentation time. controlled pH condition maintained in the bioreactor may al- Efficiency was determined as the process yield divided by low the reduction of cell inhibition by product formation, and the theoretical yield, Y is 1.0 g of LA produced p/s theoretical it was possible to development a long-term fermentation with per 1.0 g of glucose consumed (Abdel-Rahman et al. 2011). total substrate consumption. Analytical methods Results and discussion The concentrations of organic acid and glucose were deter- mined by high-efficiency liquid chromatography (HPLC). Preliminary screening of components Glucose was analyzed using an Aminex HPX-87P column (300 mm × 7.8 mm, 9 μm; Bio-Rad) and a refractive index Plackett-Burman design was used for the screening of medi- detector (RI-410, Waters). The mobile phase was Milli-Q wa- um components that influence DLA production. Ten compo- ter at a flow rate of 1.0 mL/min, and the column was operated nents, such as glucose, peptone, meat extract, yeast extract, at 85 °C. Organic acid were analyzed using a column C18 ammonium citrate, sodium acetate, magnesium sulfate, man- (250 mm × 4.6 mm, 9 μm; StrodsIIPeek) and UV detector ganese sulfate, dipotassium phosphate, and Tween-80 were set at 210 nm. The mobile phase was phosphate buffer at a used in the formulation of the fermentation medium. All fac- flow rate of 0.9 mL/min, and the column was operated at tors were set up at two levels (− 1 or + 1), and the central point 50 °C. Enantiomer D(−) lactic acid was identified using a and 13 different medium conditions were evaluated in 15 trials chiral column (150 mm × 4.6 mm; Chirex® 3126) with (data not show). The results of these preliminary experiments 1mM CuSO as the mobile phase, flow rate of 0.8 mL/min, showed concentrations of DLA from 6.65 to 13.23 g/L and temperature 50 °C, and UV detector set at 254 nm. productivity from 0.30 to 1.10 g/L · h. These results were similar to the data reported by Chauhan et al. (2007) when Experimental design selecting components for Lactobacillus sp. KCP01, they ob- tained LA concentrations ranging from 5.775 to 16.860 g/L Response surface methodology was used to study the individ- after 48 h of fermentation. The DLA concentration and ual and interactive effects of the significant variables identi- productivity obtained in the present study were also fied through the screening design on DLA production. The compared to the results obtained by Slavica et al. (2015)with variables were studied at four levels of concentration and at L. coryniformis subsp. torquens. These authors reported DLA the central point using central composite rotatable design production with concentration between 10.09 and 14.03 g/L 550 Ann Microbiol (2018) 68:547–555 and productivity between 0.94 and 0.99 g/L · h using the MRS production. Dipotassium phosphate has also been reported as medium. Therefore, the results of the screening experiments a significant component to the process (Chauhan et al. 2007). were considered satisfactory. This could be due to the buffer effect of this substance. In The components were screened at the confidence level of these experiments, which were conducted without pH control, 95%. As shown in the Pareto chart (see Fig. 1), the factors the buffer effect reduced the metabolism inhibition by the that positively affected DLA production were sodium ace- acidification of the medium. tate, meat extract, yeast extract, glucose, and dipotassium Ammonium citrate had a negative effect on the fermenta- phosphate. Ammonium citrate had a negative effect. tion process. The citrate ion might have had an inhibitor effect Manganese sulfate, tween-80, peptone, and magnesium sul- on pyruvate formation at PDH complex. This negative effect fate were considered insignificant. In these experiments, pH was also reported by Chauhan et al. (2007). Magnesium sul- was not maintained at a constant value and decreased during fate and manganese sulfate did not show a significant impact the fermentation process due to the growth-associated DLA in the process and were excluded. Magnesium and manganese production (Abdel-Rahman et al. 2011). Therefore, glucose act as co-factors and were probably already present in the was not completely depleted from the medium and had a needed concentrations in the complex nitrogen sources. moderate effect. Tween-80 was also insignificant to the process, according to L. coryniformis subsp. torquens has complex nutrient re- the results. These five factors were not further studied. quirements due to their limited ability to synthesize B vitamins and amino acids (Abdel-Rahman et al. 2013). Complex nitro- gen sources such as meat extract and yeast extract contain Study of main components using the response vitamins, mineral salts, trace elements (magnesium, manga- surface methodology nese, zinc and selenium), B vitamins (B ,B ,B ), and amino 1 2 6 acids that can fulfill the microbial nutritional requirements and The response surface methodology was used to model and reduce the production time. Meat extract and yeast extract optimize the biotechnology process. The variables meat ex- were considered significant factors and selected as nitrogen tract, yeast extract, sodium acetate, and dipotassium phos- sources. In this study, peptone was not found to be a phate were studied using a central composite rotatable design significant factor. This result is different from the results keeping a fixed concentration of glucose. The factors were set reported by Chauhan et al. (2007) using a similar design for up at four levels (− 1, + 1, − 2, + 2) and the central point (see Lactobacillus sp. KCP01 and Naveena et al. (2005)for Table 1). Table 2 represents the experimental design, and the Lactobacillus amylophilis GV6. In both studies, peptone had results that were obtained. Different medium conditions were a significant impact on the acid production. Considering the evaluated in 30 trials, and the results showed DLA concentra- high cost of peptone and the desire to improve the economic tions ranging from 7.06 to 12.29 g/L (see Table 2). The vari- parameters of DLA production, the fact that peptone was not ation of DLA concentration in this study was higher than the significant to the process is a positive result. one reported by Bustos et al. (2004) using the response surface Sodium acetate was found to be significant. According to methodology for evaluate different organic nitrogen sources Hertzberger et al. (2013), acetate is used in the formation of (corn steep liquor, peptone, and yeast extract) in complex acetyl-CoA and enhances cell growth which influences DLA composition medium, which lactic acid ranged from 9.8 to Fig. 1 Pareto chart of standardized effects for ten-factor on DLA production by L. coryniformis subsp. torquens at 12 h fermentation time. Positive effect ■,negative effect for alpha = 0.05 Ann Microbiol (2018) 68:547–555 551 Table 1 Coded and real values of factors in the central composite analysis of the coefficients in Table 3 showed that the main rotatable experimental design factors were meat extract, yeast extract, and sodium acetate. Dipotassium phosphate did not show a significant effect as an Factor Level of factor individual factor, which could be related to the absence of the − 2 −10 1 2 buffer effect for such nutrient. On the other hand, dipotassium phosphate was kept in the model because it had a small effect Sodium acetate (A, g/L) 0.05 1.70 3.35 5.00 6.65 in the interaction with sodium acetate. Contrary to this study, Meat extract (B, g/L) 0.25 3.50 6.75 10.00 13.25 Bustos et al. (2004) reported that yeast extract did not show a Dipotassium phosphate (C, g/L) 0.05 0.70 1.35 2.00 2.65 significant effect as an individual factor although it presented Yeast extract (D, g/L) 0.05 1.70 3.35 5.00 6.65 significant interactions with other two organic nitrogen sources considered the main factors. For this case, yeast ex- tract was of less importance for the metabolic activity, since in 9.9 g/L after 20 h of shake flask fermentations using calcium the conditions studied by the authors the requirements of B carbonate (100 g/L) to neutralize the acid production. vitamins and proteins were provided by corn steep liquor and Table 3 shows the analysis of variance (ANOVA) for the peptone, which were more easily assimilable. The authors model and its statistics. Values of Prob > Fisher inferior to maintained the three nitrogen sources in the model for fermen- 0.0500 indicate that the model terms are significant. The tation of 20 h, and in this way, we could consider that the characteristics of that medium are more complex than those Table 2 Central composite rotatable experimental design matrix with presented in this study. experimental values of DLA produced by L. coryniformis subsp. torquens The regression equation was obtained through the analysis Run A B C D DLA (g/l) of variance, which gave the response (DLA concentration g/L) as a function of four variables. A second-order polynomial 1 − 1.00 − 1.00 − 1.00 − 1.00 8.42 ± 0.08 (Eq. 2) was obtained using the terms that were considered 2 − 1.00 − 1.00 1.00 1.00 10.45 ± 0.06 significant to the process. The equation in terms of coded 31.00 − 1.00 − 1.00 1.00 9.62 ± 0.07 factors can be used to make predictions about the response 40.00 − 2.00 0.00 0.00 8.69 ± 0.01 for a given level of each factor. Even though the model coef- 5 − 2.00 0.00 0.00 0.00 9.26 ± 0.06 ficients were obtained experimentally, the model can be useful 6 0.002.000.000.00 12.29±0.06 to predict results of untested conditions. 7 0.000.000.000.00 9.84±0.121 DLAðÞ g=L ¼ 10:56 þ 0:45A þ 0:86 B−0:09 C 8 1.001.001.00 − 1.00 10.92 ± 0.03 þ 0:68 D þ 0:28 AC−0:21 D 9 0.000.000.000.00 10.59±0.05 10 0.00 0.00 − 2.00 0.00 11.30 ± 0.08 ðA; sodium acetate; B; meat extract 11 − 1.00 1.00 1.00 − 1.00 10.12 ± 0.07 ; C; dipotassium phosphate; D; yeast extract Þð2Þ 12 0.00 0.00 2.00 0.00 10.85 ± 0.09 13 0.00 0.00 0.00 0.00 10.99 ± 0.03 The F-value of 23.11 implies that the model is significant 14 2.00 0.00 0.00 0.00 11.42 ± 0.07 and that the model was accurate in describing the experimen- 15 1.00 − 1.00 − 1.00 − 1.00 9.27 ± 0.09 tal data. The Lack of Fit F-value was 1.49. A non-significant 16 0.00 0.00 0.00 0.00 10.82 ± 0.06 Lack of Fit value means that the model fits. The R coefficient indicates good agreement between experimental and predicted 17 − 1.00 1.00 − 1.00 1.00 11.92 ± 0.07 18 − 1.00 1.00 1.00 1.00 10.56 ± 0.04 data for a microbiological process and suggests that the model is reliable for depicting DLA production by L. coryniformis 19 1.00 − 1.00 1.00 − 1.00 9.16 ± 0.03 subsp. torquens. 20 0.00 0.00 0.00 2.00 10.90 ± 0.08 Figure 2 shows the surface plots obtained using the model 21 1.00 1.00 − 1.00 1.00 11.92 ± 0.07 equation. The main factors were meat extract and yeast extract 22 0.00 0.00 0.00 − 2.00 8.67 ± 0.04 concentrations (Fig. 2a–c) while sodium acetate and 23 − 1.00 − 1.00 − 1.00 1.00 10.35 ± 0.03 dipotassium phosphate had little effect (Fig. 2b–d). The results 24 0.00 0.00 0.00 0.00 10.88 ± 0.06 showed that increasing concentrations of meat extract and 25 1.00 1.00 − 1.00 − 1.00 11.19 ± 0.04 yeast extract resulted in increased DLA concentration. 26 1.00 − 1.00 1.00 1.00 11.42 ± 0.04 However, high concentrations of these products represent 27 − 1.00 − 1.00 1.00 − 1.00 7.06 ± 0.06 higher production cost. Therefore, in order to make the pro- 28 0.00 0.00 0.00 0.00 10.28 ± 0.05 cess economically feasible, a lower concentration of these 29 − 1.00 1.00 − 1.00 − 1.00 10.34 ± 0.03 products must be used without, however, affecting the perfor- 30 1.00 1.00 1.00 1.00 12.17 ± 0.07 mance of the microorganism. 552 Ann Microbiol (2018) 68:547–555 Table 3 Analysis of variance Source Sum of squares df Mean square F value P value Prob > F (ANOVA) for the selected model Model 36.46 6 6.08 23.11 < 0.0001 A–sodium acetate 4.83 1 4.83 18.38 0.0003 B–meat extract 17.66 1 17.66 67.19 < 0.0001 C–dipotassium phosphate 0.18 1 0.18 0.68 0.4184 D–yeast extract 11.19 1 11.19 42.58 < 0.0001 AC 1.27 1 1.27 4.84 0.0382 D 1.32 1 1.32 5.02 0.0350 Residual 6.05 23 0.26 Lack of fit 5.09 18 0.28 1.49 0.3508 Pure error 0.95 5 0.19 < 0.0001 Cor total 42.51 29 0.0003 2 2 [R =0.858; Adj R =0.821] From a technical point of view, the highest DLA con- optimization. However, the economic aspect must also be centration should be considered the main objective of this considered. In order to optimize the medium (considering Fig. 2 Surface plots for interactions between the different components of acetate and meat extract, (c) dipotassium phosphate and meat extract, the media that were optimized to increase the DLA production by L. and (d) dipotassium phosphate and sodium acetate. The effect of two coryniformis subsp.torquens at 12h fermentation time where (a) variables with the other two variables maintained at their respective zero represents the effect of yeast extract and meat extract, (b) sodium Ann Microbiol (2018) 68:547–555 553 Table 4 Medium modified using Constraints the Design Expert 7.0 optimization tool for DLA Factor Goal Lower limit Upper limit Importance production by L. coryniformis Sodium acetate Minimize −22 3 subsp. torquens Meat extract Minimize −22 5 Dipotassium phosphate Minimize −22 3 Yeast extract Minimize −22 4 D(−) lactic acid Maximize 7.06 12.29 5 Solution-optimization Factor Coded values Real values (g/L) Desirability Sodium acetate − 2.00 0.05 0.715 Meat extract − 1.08 3.23 Dipotassium phosphate − 2.00 0.05 Yeast extract − 0.43 2.63 Response Prediction SE Pred 95% PI low 95% PI high D-lactic acid (DLA, g/L) 9.70 0.80 8.04 11.35 both the technical and economic point) two aspects were In order to validate the model, DLA was produced by targeted: increasing DLA concentration and reducing the L. coryniformis subsp. torquens using MRS medium and concentration of components, mainly meat extract. Under the modified medium under the same conditions of the these conditions, the concentrations were optimized using shake flask experiments. The concentration of DLA using the quadratic model of the Design Expert 7.0 optimization the modified medium was 10.72 g/L with 0.89 g/L.h of tool. The parameters and results were summarized in productivity. This result corroborates the model prediction Table 4. The final modified medium contained 3.23 g/L (Table 4) and shows the adequacy of the model. The con- of meat extract, 2.63 g/L of yeast extract, 0.05 g/L of centration of DLA using MRS medium was 12.69 g/L. sodium acetate, and 0.05 g/L of dipotassium phosphate. This concentration was 1.19-fold higher than the concen- The differences are the reductions in concentrations of tration obtained when the modified medium was used meat extract from 10 to 3.23 g/L, yeast extract from 5.0 (10.72 g/L). This difference is a result of the reduction to 2.63 g/L, sodium acetate from 5.0 to 0.05 g/L, and in the number and concentrations of medium components. dipotassium phosphate from 2.0 to 0.05 g/L. The desir- From the technical-economic point of view, the modified ability coefficient of 0.715 was obtained, indicating that medium showed a satisfactory result since the decrease in approximately 72% of the production and cost reduction the DLA concentration was only 20%, especially when requirements were achieved. considering the higher cost of the MRS medium. Fig. 3 Profile of glucose consumption (□), DLA production (○) and acetic acid production (Δ)in batch fermentation by L. coryniformis subsp. torquens using the modified medium at 37 °C, 120 rpm and pH 6.8 controlled with NaOH (4 M) 554 Ann Microbiol (2018) 68:547–555 Bustos et al. (2004) reported a model with higher produc- the carbon source and obtained a lactic acid production of tion using 5 g/L of CSL, 3.6 g/L of yeast extract, and 10 g/L of 57 g/L and yield factor of 0.63 g/g of reducing sugar after peptone for LA production by L. coryniformis subsp. torquens. 48 h at pH 6.2 and Wang et al. (2016) reported lactic acid The authors obtained a concentration of 58.9 g/L and 0.61 g/ concentration of 36 g/L and yield factor of 0.69 g/g at pH L.h of productivity using 100 g/L of glucose after 96 h. The 6.25 by L. rhamnosus LA-04-1. productivity reported by these authors are lower than the one The results of this study showed that the modified medium, obtained in this study, where lower concentrations of expen- which has a lower concentration of nutrients, lead to the total sive nitrogen sources were employed. The DLA concentration consumption of the glucose by L. coryniformis subsp. and productivity (10.72 g/L and 0.89 g/L) obtained using the torquens. The DLA production showed no inhibitory effects modified medium was also comparable to the results obtained by nutrient limitation or by-products formation, and a 0.95 by Slavica et al. (2015). These authors obtained 12.95 g/L of yield factor was obtained. DLA and productivity of 0.99 g/L.h by L. coryniformis subsp. torquens using MRS medium after 13 h of fermentation. Özcelik et al. (2016) found LA concentrations ranging from Conclusions 0.27 to 0.56 g/L for eight lactic acid bacteria (LAB) strains in MRM medium after 4 days of fermentation at 37 °C. Cingadi In this study, the preliminary screening showed the significant et al. (2015) reported that homo-fermentative LAB strains effect sodium acetate, meat extract, yeast extract, glucose, and studied for DLA production by batch culture in MRS medium dipotassium phosphate had on the DLA production. These produced between 0.22 and 11.32 g/L by L. coryniformis bac- components were studied using a central composite rotatable teria (NCDC367, NCDC368, NCDC369) produced DLA acid design, and a quadratic model was obtained to describe the ranging from 9.76 to 11.32 g/L in shake flask experiments. relationship between the DLA production and the medium components. The optimization of model employing the DLA production in bioreactor Design-Expert 7.0 tool resulted in the modified medium with the following composition: 2.3 g/L of meat extract, 3.58 g/L Batch fermentation was carried out in a bioreactor to study the of yeast extract, 0.05 g/L of sodium acetate, and 0.05 g/L of DLA production using the modified medium under controlled dipotassium phosphate. The scale-up from flask to bioreactor operating conditions. Figure 3 shows the profiles of glucose uti- under controlled pH condition using the modified medium lization and organic acid production during the fermentation at showed the total consumption of glucose by L. coryniformis pH 6.8 and initial glucose of 33.85 g/L. Under controlled pH subsp. torquens without inhibitory effects, little formation of conditions, glucose was completely consumed by L. by-product acetic acid, and yield factor of 0.95 g/g. The mod- coryniformis subsp. torquens after 38 h—and DLA production ified medium obtained in this work could contribute to the was increased due to the reducing of inhibition by acid products development of processes economically viable for the optical- formation. DLA with an optical purity of ≥ 99.0% was produced ly pure DLA production. with concentration of 33.6 g/L, yield factor of 0.95 g/g, produc- tivity of 0.88 g/L · h, and 95% of fermentation efficiency. A small Acknowledgements The authors gratefully acknowledge financial sup- quantity of acetic acid was also produced but other products were port by Leopoldo Américo Miguez de Mello Research and Development not identified. The acetic acid production was 0.31 g/L at the end Center (Cenpes), Petrobras. We also would like to acknowledge the sup- port of the Laboratories of Bioprocess Development (Ladebio), from the of the fermentation. Acetic acid was still being produced while school of chemistry of Federal University of Rio de Janeiro, UFRJ. DLA concentration remained constant in the stationary phase. These results of yield factor (0.95 g/g) and productivity Compliance with ethical standards (0.88 g/L · h) were observed to be high in contrast to DLA production by Lactobacillus delbrueckii IFO3202 reported for Conflict of interest The authors declare that they have no conflict of Tanaka et al. (2006) with an estimated yield factor of 0.9 g/g interest. and productivity of 0.75 g/L · h using the MRS medium and after 36 h from 30 g/L of glucose at pH 6.0. In our present study, yield factor was similar to that obtained by Zhao et al. (2010) who reported conversion of glucose to DLA of 0.94– References 0.99 g/g for repeated production in one-reactor system at Abdel-Rahman MA, Tashiro Y, Sonomoto K (2011) Lactic acid produc- pH 5.6–5.8 by Sporolactobacillus sp. CASD. Others studies tion from lignocelluse-derived sugars using lactic acid bacteria: for production and optimization of lactic acid in batch overview and limits. J Biotechnol 156:286–301 fermentation have presented lower yield factor values, for Abdel-Rahman MA, Tashiro Y, Sonomoto K (2013) Recent advances in example Bernardo et al. (2016)used Lactobacillus rhamnosus lactic acid production by microbial fermentation processes. B103 and lactose (approximately 90 g/L of reducing sugar) as Biotechnol Adv 31:877–902 Ann Microbiol (2018) 68:547–555 555 Abdelwahed NAM, Gomaa EZ, Hassan AA (2017) Statistical modelling Sources of Renewable Energy. Springer, Dordrecht, pp 211–228. https://doi.org/10.1007/978-90-481-3295-9_11 and optimization of fermentation medium for lincomycin production by Streptomyces lincolnensis immobilized cells. Braz Arch Biol Myers RH, Montgomery DC, Anderson-Cook CM (2009) Response sur- Technol 60:e160210 face methodology: process and product optimization using designed Almeida DG, Soares da Silva RCF, Luna JM, Rufino RD, Santos VA, experiments, 3rd edn. Wiley, New York Sarubbo LA (2017) Response surface methodology for optimizing Nakano S, Ugwu CU, Tokiwa Y (2012) Efficient production of D-(-)- the production of biosurfactant by Candida tropicalis on industrial lactic acid from broken rice by Lactobacillus delbrueckii using waste substrates. Front Microbiol 8:157 Ca(OH) as a neutralizing agent. Bioresour Technol 104:791–794 Bernardo MP, Coelho LF, Sass DC, Contiero J (2016) L-(+)-Lactic acid Nampoothiri KM, Nair NR, John RP (2010) An overview of the recent production by Lactobacillus rhamnosus B103 from dairy industry developments in polytactide (PLA) research. Bioresour Technol waste. Braz J Microbiol 47:640–646 101:8493–8501 Bustos G, Moldes AB, Alonso JL, Vazquez M (2004) Optimization of D- Narayanan N, Roychoudhury PK, Srivastava A (2004) L (+) lactic acid lactic acid production by Lactobacillus coryniformis using response fermentation and its product polymerization. Electron J Biotechnol. surface methodology. Food Microbiol 21:43–148 https://doi.org/10.2225/vol7-issue2-fulltext-7 Chauhan K, Trivedi U, Patel KC (2007) Statistical screening of medi- Naveena BJ, Altaf MD, Bhadriah K, Reddy G (2005) Selection of medi- um components by Plackett-Burman design for lactic acid pro- um components by Plackett–Burman design for production of L(+) duction by Lactobacillus sp, KCP01 using date juice. Bioresour lactic acid by Lactobacillus amylophilus GV6 in SSF using wheat Technol 98:98–103 bran. Bioresour Technol 96(4):485–490 Cingadi S, Srikanth K, Arun EVR, Sivaprakasam S (2015) Statistical Nguyen CM, Choi GJ, Choi YH, Jang KS, Kim JC (2013) D- and L-lactic optimization of cassava fibrous waste hydrolysis by response sur- acid production from fresh sweet potato through simultaneous sac- face methodology and use of hydrolysate based media for the pro- charification and fermentation. Biochem Eng J 81:40–46 duction of optically pure D-lactic acid. Biochem Eng J 102:82–90 Ojha N, Das N (2018) A statistical approach to optimize the production of Dasgupta D, Suman SK, Pandey D, Ghosh D, Khan R, Agrawal D, Jain Polyhydroxyalkanoates from Wickerhamomyces anomalus VIT- RK, Vadde VT, Adhikari DK (2013) Design and optimization of NN01using response surface methodology. IJBM 107:2157–2170 ethanol production from bagasse pith hydrolysate by a Özcelik S, Kuley E, Özogul F (2016) Formation of lactic, acetic, succinic, thermotolerant yeast Kluyveromyces sp. IIPE453 using response propionic, formic and butyric acid by lactic acid bacteria. LWT Food surface methodology. Springerplus 2:159 Sci Technol 73:536–542 Dubey S, Singh A, Banerjee UC (2011) Response surface methodology Shen L, Worrel E, Patel M (2010) Present and future development in of nitrilase production by recombinant Escherichia coli.Braz J plastics from biomass. Biofuels Bioprod Biorefin 4:25–40 Microbiol 42:1085–1092 Slavica A, Trontel A, Jelovac N, Kosovec Z, Santek B, Novak S (2015) Gao C, Ma C, Xu P (2011) Biotechnological routes based on lactic acid Production of lactate and acetate by Lactobacillus coryniformis production from biomass. Biotechnol Adv 29:930–939 subsp. torquens DMS 20004T in comparison with Lactobacillus Hertzberger RY, Pridmore RD, Gysler CH, Kleerebezem M, Texeira de amylovorus DMS 20531T. 2015. J Biotechnol 202:50–59 Mattos MJ (2013) Oxygen relieves the CO and acetate dependency Tanaka T, Hoshina M, Tanabe S, Sakai K, Ohtsubo S, Taniguchi M of Lactobacillus johnsonii NCC533. PLoS One. https://doi.org/10. (2006) Production of D-lactic acid from defatted rice bran by simul- 1371/journal.pone.0057235 taneous saccharification and fermentation. Bioresour Technol 97: Hofvendahl K, Hahn-Hägerdal B (2000) Factors affecting the fermenta- 211–217 tive lactic acid production from renewable resources. Enzym Microb Wang Y, Chan C, Cai D, Wang Z, Qin P, Tan T (2016) The optimization Technol 26:87–107 of L-lactic acid production from sweet sorghum juice by mixed Imran M, Anwar Z, Irshad M, Javid A, Hussain A, Ali S (2017) fermentation of Bacillus coagulans and Lactobacillus rhamnosus Optimization of cellulase production from a novel strain of under unsterile conditions. Bioresour Technol 218:1098–1105 Aspergillus Tubingensis IMMIS2 through response surface method- Wee YJ, Kim JN, Ryu HW (2006) Biotechnological production of lactic ology. Biocatal Agric Biotechnol 12:191–198 acid and its recent applications. Food Technol Biotechnol 44(2): Jem KJ, van der Pol JF, de Vos S (2010) Microbial lactic acid, its polymer 163–172 poly(lactic acid), and their industrial applications. In: Chen GQ (ed) Zhao B, Wang L, Li F, Hua D, Ma C, Ma Y, Xu P (2010) Kinetics of D- Plastics from Bacteria, vol. 14. Springer, Berlin/Heidelberg, pp 323– lactic acid production by Sporolactobacillus sp. strain CASD using repeated batch fermentation. Bioresour Technol 101:6499–6505 Kongruang S, Kangsadan T (2015) Optimization of succinic acid produc- tion from crude glycerol by encapsulated Anaerobiospirillum ZhuLW, Wang CC, LiuRS, Li HM,Wan DJ,TangYJ succiniciproducens using response surface methodology. IJBBB (2012) Actinobacillus succinogenes ATCC 55618 fermentation me- 5(1):11–25 dium optimization for the production of succinic acid by response Li Y, Cui F (2010) Microbial lactic acid production from renewable re- surface methodology. J Biomed Biotechnol. https://doi.org/10.1155/ sources. In: Singh OV, Harvey SP (eds) Sustainable Biotechnology: 2012/626 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Microbiology Springer Journals

Low-cost effective culture medium optimization for d-lactic acid production by Lactobacillus coryniformis subsp. torquens under oxygen-deprived condition

Download PDF
9 pages

Loading next page...
 
/lp/springer-journals/low-cost-effective-culture-medium-optimization-for-d-lactic-acid-88UHn741Cm

References (26)

Publisher
Springer Journals
Copyright
Copyright © 2018 by Springer-Verlag GmbH Germany, part of Springer Nature and the University of Milan
Subject
Life Sciences; Microbiology; Microbial Genetics and Genomics; Microbial Ecology; Mycology; Medical Microbiology; Applied Microbiology
ISSN
1590-4261
eISSN
1869-2044
DOI
10.1007/s13213-018-1362-y
Publisher site
See Article on Publisher Site

Abstract

Lactic acid is considered a commodity and its production is boosted by the synthesis of polylactic acid. D-lactic acid (DLA) isomer offers greater flexibility and biodegradability and it can only be obtained in its pure form through fermentation. The lactate dehydrogenase is stereospecific for homofermentative production of DLA isomer in the metabolic pathway of Lactobacillus coryniformis subsp. torquens, with optical purity of ≥ 99.9% under oxygen-deprived condition. A simple culture medium that increases DLA production and reduces fermentation costs is fundamental for industrial applicability. A central composite rotatable design was used to evaluate significant components influencing the DLA production. Concentrations were adjusted using the Design-Expert 7.0 optimization tool with a desirability coefficient of 0.693 and the best concentrations of each component were determined. Finally, an assay in the bioreactor with the modified culture medium resulted in a product yield of 0.95 g/g, volumetric productivity of 0.85 g/L.h and 95% of efficiency. . . . Keywords Fermentation Optical purity Central composite rotatable design Optimization tool Introduction et al. 2013). The world demand for lactic acid production is forecasted to reach over one million metric tons by the Lactic acid (LA) was discovered in 1780 by chemist Carl year 2020 with annual growth of 5–8% (Jem et al. 2010; Wilhelm Scheele. It is an important hydroxycarboxylic Abdel-Rahman et al. 2013). The application of lactic acid acid with applications in the food, pharmaceutical, cos- in the production of biodegradable biopolymer poly-lactic metics, textile, and chemical industries (Abdel-Rahman acid (PLA) is growing due to its biodegradability charac- teristics (Nampoothiri et al. 2010). Lactic acid has an asymmetric β-carbon as chiral cen- ter that produces two enantiomers: L(+) lactic acid (LLA) * Lizeth Jaramillo and D(−)lactic acid (DLA). While LLA and DLA are only lizethacevedo@eq.ufrj.br obtained by microbial fermentation processes, racemic DL- lactic is always produced via chemical synthesis (Li and Laboratory of Biocorrosion, Biodegradation and Biosynthesis, Cui 2010). Depending on the microorganism and the op- School of Chemistry, Federal University of Rio de Janeiro, Av. erating conditions employed, an optically pure product Horacio Macedo, 2030, Bloco E, Sala E-109, CT, Cidade L(+) or D(−) and in some cases racemic mixtures can be Universitária - Ilha do Fundao, Rio de Janeiro, RJ CEP: 21949-900, produced (Narayanan et al. 2004). Brazil The chemical, physical, and biological properties of Nucleus of Ecology and Socio-Environmental Development of PLA are determined by the isomeric composition, pro- Macaé, NUPEM, Federal University of Rio de Janeiro, Av. São José do Barreto, 764 -São José do Barreto, Macaé, RJ CEP: 27965-045, cessing temperature, annealing time, and molecular Brazil weight. Polymerization of LLA results in poly-L(+) lactic Department of Biochemical Engineering. School of Chemistry, acid (PLLA) and polymerization of DLA results in poly- Federal University of Rio de Janeiro, Av. Horacio Macedo, 2030, D(−) lactic acid (PDLA). Due to the stereo regular chain Bloco E, Sala E-203, CT, Cidade Universitária - Ilha do Fundao, Rio microstructure, optical pure PLLA and PDLA are semi de Janeiro, RJ CEP: 21949-900, Brazil 548 Ann Microbiol (2018) 68:547–555 crystalline. The majority of commercial PLA is poly production. Various studies have reported the use of cheap (meso-lactide), a mixture of LLA (> 95%) and DLA (< raw material for DLA production (Tanaka et al. 2006; 5%) (Nampoothiri et al. 2010). Nakano et al. 2012; Nguyen et al. 2013; Cingadi et al. Poly (meso-lactide) can be used in a wide range of appli- 2015). However, few studies have showed the effect of medi- cations. However, this type of PLA has no stereochemical um components in the production of DLA. Such studies are structure and is highly amorphous. Also, the end products important to establish a simple medium, with few components made from this PLA are not suitable for high temperatures and low concentrations that offers a good balance between the applications (Shen et al. 2010). The physicochemical proper- nutritional needs of the microorganism and the economic vi- ties of optically active PDLA and PLLA are nearly the same, ability of the process. whereas the racemic PLA has very different characteristics. Response surface methodology (RSM) is a collection of Properties of PLA biopolymers can be modified by varying statistical techniques and tools for constructing a functional the proportions of the enantiomers. For example, the melting relationship between a response variable and a set of design temperature of PLLA can be increased to 40–50 °C and its variables used in studies of developing, improving, and opti- heat deflection temperature can be increased from approxi- mizing processes (Myers et al. 2009). This experimental meth- mately 60–190 °C by physically blending the polymer with odology has been successfully used in optimization studies for PDLA (Nampoothiri et al. 2010). The crystallization can be different biotechnological processes (Dubey et al. 2011;Zhu decreased with the presence of DLA isomer, resulting in et al. 2012; Kongruang and Kangsadan 2015;Dasguptaetal. amorphous polymers with controlled degradation, increasing 2013; Cingadi et al. 2015; Abdelwahed et al. 2017;Almeida et biocompatibility and, therefore, better application in biomed- al. 2017, Imran et al. 2017; Ojha and Das 2018). There are ical fields. Hence, the production of PDLA has gained impor- different RSM experimental designs, and in this study, a sta- tance and the optical purity of DLA plays a crucial role in tistical optimization of the producing medium was performed determining the characteristics of the PLA biopolymer. by employing a central composite rotatable design (CCRD) in High-purity LLA or DLA monomers are necessary to order to obtain the maximum amount of reliable information achieve the desired polymer properties (Li and Cui 2010). with the fewest number of experiments as well as to achieve Nowadays, most lactic acid is produced by microbial fermen- high production of optically pure DLA using low concentra- tation and many studies have reported optical pure LLA pro- tions of components that influence the DLA production. duction at industrial scale. On the other hand, production of optically pure DLA is not well established (Cingadi et al. 2015). The fermentative production has the advantage of giv- Materials and methods ing the ability to choose a strain of homofermentative lactic acid bacteria (LAB) that will produce only one optical isomer Microorganism and inoculum preparation with high purity. Few wild type strains have been studied for the production of optical pure DLA (Cingadi et al. 2015). Lactobacillus coryniformis subsp. torquens (ATCC 25600) LAB are Gram-positive, non-motile, non-spore forming was employed in this study, and it was acquired from the bacteria, and most of the group are cocci. They are faculta- Collection of Institute Pasteur (CIP). This strain was selected tive anaerobes, catalase negative, with high acid tolerance among four Lactobacillus strains after preliminary tests where and are generally recognized as safe (GRAS). LAB are un- its homofermentative characteristic for production of DLA able to synthesize ATP by respiration and have complex with high optical purity was verified by chiral chromatogra- nutrient requirements due to their limited ability to synthe- phy (unpublished data). size B vitamins and amino acids (Hofvendahl and Hahn- Cells were activated in 100 mL sealed anaerobic bottles Hägerdal, 2000; Gao et al. 2011). The Lactobacillus strain containing 50 mL of MRS medium. The anaerobic bottles needs complex nitrogen sources, vitamins, and minerals were inoculated with 1 mL of the stock culture, which was (Narayanan et al. 2004; Wee et al. 2006;Liand Cui 2010). preserved in glycerol and stored at − 80 °C. The inoculum was Homo-fermentative Lactobacillus produces LA as the main prepared in anaerobic bottles containing 100 mL of MRS product of the glucose metabolism. Through the Embden- medium and the inoculation volume was 10% of exponential- Meyerhof-Parnas pathway, 1 mol of hexose is converted to ly growing cells from the activation step. 2 mol of lactic acid and 2 mol of ATP (Hofvendahl and The medium for activation and inoculum preparation Hahn-Hägerdal 2000;Li and Cui 2010). contained the follow components: glucose, 10 g/L; peptone, Due to the nutritional requirements of Lactobacillus 10 g/L; yeast extract, 5 g/L; meat extract, 1 g/L; ammonium strains, the fermentation process may have high costs associ- citrate, 2 g/L; sodium acetate, 5 g/L; dipotassium phosphate, ated to the raw materials used in the process. The use of 2 g/L; magnesium sulfate, 0.1 g/L; manganese sulfate, 0.05 g/ cheaper feedstock as carbon source along with higher produc- L; and Tween-80, 1 g/L. Glucose was aseptically added to the tivity is a promising strategy to reduce the cost of DLA medium after autoclaving. All culture media were flushed Ann Microbiol (2018) 68:547–555 549 with nitrogen gas for 15 min to remove oxygen and (CCRD). The experimental data allowed describe the relation- autoclaved at 121 °C for 15 min. All liquid cultures were ships the DLA production and the medium components by a incubated in rotary shaker at 37 °C and 120 rpm. mathematical model following second-order polynomial equation: Fermentation assays Y ¼ β þ ∑β X þ ∑β X þ ∑β X X ; i i j o i ii i ij ð1Þ i≠j; i; j ¼ 1; 2; 3;…:k Shake flask fermentation were performed for the development of experimental designs. All experiments were carried out in where Y is the predicted response (DLA production, g/L), β is 100 mL anaerobic flasks containing 50 mL of producing me- the interception coefficient, β is the linear term, β is the i ii dium with different concentrations according to the experi- quadratic term, β is the interaction term, and X , X represent ij i j mental design. The initial pH was adjusted to 6.8, and glucose the independent variables. Experimental data from the CCRD was aseptically added to the medium after autoclaving. were subjected to analysis of variance (ANOVA). Analysis of Anaerobic fermentations were performed at 37 °C and the model, calculation of the predicted responses, and the 120 rpm under uncontrolled pH condition for 12 h. All exper- plotting of surface plots were done using Design-Expert® iments were inoculated with cells at the exponential growing software version 7.0. The final adjustment of the model was phase using inoculum size 10% (v/v) with initial cells concen- done using the Design-Expert 7.0 optimization tool consider- tration around 0.13 g/L. ing as main criteria reduction of component concentrations The fermentation media resulted of the shake flask exper- and maximized DLA production. iments were used for production of DLA at bioreactor scale. Batch fermentation was carried out in a 2 L bioreactor (New Calculation of kinetic parameters Brunswick Bio Flo®) with a working volume of 800 mL and the same conditions from the shake flask experiments. The pH The yield of DLA was calculated as the amount of DLA pro- was automatically controlled at 6.8 by the addition of 4 M duced from 1 g of consumed glucose. Productivity was deter- NaOH and the fermentation time was 46 h considering that mined as the concentration of DLA per fermentation time. controlled pH condition maintained in the bioreactor may al- Efficiency was determined as the process yield divided by low the reduction of cell inhibition by product formation, and the theoretical yield, Y is 1.0 g of LA produced p/s theoretical it was possible to development a long-term fermentation with per 1.0 g of glucose consumed (Abdel-Rahman et al. 2011). total substrate consumption. Analytical methods Results and discussion The concentrations of organic acid and glucose were deter- mined by high-efficiency liquid chromatography (HPLC). Preliminary screening of components Glucose was analyzed using an Aminex HPX-87P column (300 mm × 7.8 mm, 9 μm; Bio-Rad) and a refractive index Plackett-Burman design was used for the screening of medi- detector (RI-410, Waters). The mobile phase was Milli-Q wa- um components that influence DLA production. Ten compo- ter at a flow rate of 1.0 mL/min, and the column was operated nents, such as glucose, peptone, meat extract, yeast extract, at 85 °C. Organic acid were analyzed using a column C18 ammonium citrate, sodium acetate, magnesium sulfate, man- (250 mm × 4.6 mm, 9 μm; StrodsIIPeek) and UV detector ganese sulfate, dipotassium phosphate, and Tween-80 were set at 210 nm. The mobile phase was phosphate buffer at a used in the formulation of the fermentation medium. All fac- flow rate of 0.9 mL/min, and the column was operated at tors were set up at two levels (− 1 or + 1), and the central point 50 °C. Enantiomer D(−) lactic acid was identified using a and 13 different medium conditions were evaluated in 15 trials chiral column (150 mm × 4.6 mm; Chirex® 3126) with (data not show). The results of these preliminary experiments 1mM CuSO as the mobile phase, flow rate of 0.8 mL/min, showed concentrations of DLA from 6.65 to 13.23 g/L and temperature 50 °C, and UV detector set at 254 nm. productivity from 0.30 to 1.10 g/L · h. These results were similar to the data reported by Chauhan et al. (2007) when Experimental design selecting components for Lactobacillus sp. KCP01, they ob- tained LA concentrations ranging from 5.775 to 16.860 g/L Response surface methodology was used to study the individ- after 48 h of fermentation. The DLA concentration and ual and interactive effects of the significant variables identi- productivity obtained in the present study were also fied through the screening design on DLA production. The compared to the results obtained by Slavica et al. (2015)with variables were studied at four levels of concentration and at L. coryniformis subsp. torquens. These authors reported DLA the central point using central composite rotatable design production with concentration between 10.09 and 14.03 g/L 550 Ann Microbiol (2018) 68:547–555 and productivity between 0.94 and 0.99 g/L · h using the MRS production. Dipotassium phosphate has also been reported as medium. Therefore, the results of the screening experiments a significant component to the process (Chauhan et al. 2007). were considered satisfactory. This could be due to the buffer effect of this substance. In The components were screened at the confidence level of these experiments, which were conducted without pH control, 95%. As shown in the Pareto chart (see Fig. 1), the factors the buffer effect reduced the metabolism inhibition by the that positively affected DLA production were sodium ace- acidification of the medium. tate, meat extract, yeast extract, glucose, and dipotassium Ammonium citrate had a negative effect on the fermenta- phosphate. Ammonium citrate had a negative effect. tion process. The citrate ion might have had an inhibitor effect Manganese sulfate, tween-80, peptone, and magnesium sul- on pyruvate formation at PDH complex. This negative effect fate were considered insignificant. In these experiments, pH was also reported by Chauhan et al. (2007). Magnesium sul- was not maintained at a constant value and decreased during fate and manganese sulfate did not show a significant impact the fermentation process due to the growth-associated DLA in the process and were excluded. Magnesium and manganese production (Abdel-Rahman et al. 2011). Therefore, glucose act as co-factors and were probably already present in the was not completely depleted from the medium and had a needed concentrations in the complex nitrogen sources. moderate effect. Tween-80 was also insignificant to the process, according to L. coryniformis subsp. torquens has complex nutrient re- the results. These five factors were not further studied. quirements due to their limited ability to synthesize B vitamins and amino acids (Abdel-Rahman et al. 2013). Complex nitro- gen sources such as meat extract and yeast extract contain Study of main components using the response vitamins, mineral salts, trace elements (magnesium, manga- surface methodology nese, zinc and selenium), B vitamins (B ,B ,B ), and amino 1 2 6 acids that can fulfill the microbial nutritional requirements and The response surface methodology was used to model and reduce the production time. Meat extract and yeast extract optimize the biotechnology process. The variables meat ex- were considered significant factors and selected as nitrogen tract, yeast extract, sodium acetate, and dipotassium phos- sources. In this study, peptone was not found to be a phate were studied using a central composite rotatable design significant factor. This result is different from the results keeping a fixed concentration of glucose. The factors were set reported by Chauhan et al. (2007) using a similar design for up at four levels (− 1, + 1, − 2, + 2) and the central point (see Lactobacillus sp. KCP01 and Naveena et al. (2005)for Table 1). Table 2 represents the experimental design, and the Lactobacillus amylophilis GV6. In both studies, peptone had results that were obtained. Different medium conditions were a significant impact on the acid production. Considering the evaluated in 30 trials, and the results showed DLA concentra- high cost of peptone and the desire to improve the economic tions ranging from 7.06 to 12.29 g/L (see Table 2). The vari- parameters of DLA production, the fact that peptone was not ation of DLA concentration in this study was higher than the significant to the process is a positive result. one reported by Bustos et al. (2004) using the response surface Sodium acetate was found to be significant. According to methodology for evaluate different organic nitrogen sources Hertzberger et al. (2013), acetate is used in the formation of (corn steep liquor, peptone, and yeast extract) in complex acetyl-CoA and enhances cell growth which influences DLA composition medium, which lactic acid ranged from 9.8 to Fig. 1 Pareto chart of standardized effects for ten-factor on DLA production by L. coryniformis subsp. torquens at 12 h fermentation time. Positive effect ■,negative effect for alpha = 0.05 Ann Microbiol (2018) 68:547–555 551 Table 1 Coded and real values of factors in the central composite analysis of the coefficients in Table 3 showed that the main rotatable experimental design factors were meat extract, yeast extract, and sodium acetate. Dipotassium phosphate did not show a significant effect as an Factor Level of factor individual factor, which could be related to the absence of the − 2 −10 1 2 buffer effect for such nutrient. On the other hand, dipotassium phosphate was kept in the model because it had a small effect Sodium acetate (A, g/L) 0.05 1.70 3.35 5.00 6.65 in the interaction with sodium acetate. Contrary to this study, Meat extract (B, g/L) 0.25 3.50 6.75 10.00 13.25 Bustos et al. (2004) reported that yeast extract did not show a Dipotassium phosphate (C, g/L) 0.05 0.70 1.35 2.00 2.65 significant effect as an individual factor although it presented Yeast extract (D, g/L) 0.05 1.70 3.35 5.00 6.65 significant interactions with other two organic nitrogen sources considered the main factors. For this case, yeast ex- tract was of less importance for the metabolic activity, since in 9.9 g/L after 20 h of shake flask fermentations using calcium the conditions studied by the authors the requirements of B carbonate (100 g/L) to neutralize the acid production. vitamins and proteins were provided by corn steep liquor and Table 3 shows the analysis of variance (ANOVA) for the peptone, which were more easily assimilable. The authors model and its statistics. Values of Prob > Fisher inferior to maintained the three nitrogen sources in the model for fermen- 0.0500 indicate that the model terms are significant. The tation of 20 h, and in this way, we could consider that the characteristics of that medium are more complex than those Table 2 Central composite rotatable experimental design matrix with presented in this study. experimental values of DLA produced by L. coryniformis subsp. torquens The regression equation was obtained through the analysis Run A B C D DLA (g/l) of variance, which gave the response (DLA concentration g/L) as a function of four variables. A second-order polynomial 1 − 1.00 − 1.00 − 1.00 − 1.00 8.42 ± 0.08 (Eq. 2) was obtained using the terms that were considered 2 − 1.00 − 1.00 1.00 1.00 10.45 ± 0.06 significant to the process. The equation in terms of coded 31.00 − 1.00 − 1.00 1.00 9.62 ± 0.07 factors can be used to make predictions about the response 40.00 − 2.00 0.00 0.00 8.69 ± 0.01 for a given level of each factor. Even though the model coef- 5 − 2.00 0.00 0.00 0.00 9.26 ± 0.06 ficients were obtained experimentally, the model can be useful 6 0.002.000.000.00 12.29±0.06 to predict results of untested conditions. 7 0.000.000.000.00 9.84±0.121 DLAðÞ g=L ¼ 10:56 þ 0:45A þ 0:86 B−0:09 C 8 1.001.001.00 − 1.00 10.92 ± 0.03 þ 0:68 D þ 0:28 AC−0:21 D 9 0.000.000.000.00 10.59±0.05 10 0.00 0.00 − 2.00 0.00 11.30 ± 0.08 ðA; sodium acetate; B; meat extract 11 − 1.00 1.00 1.00 − 1.00 10.12 ± 0.07 ; C; dipotassium phosphate; D; yeast extract Þð2Þ 12 0.00 0.00 2.00 0.00 10.85 ± 0.09 13 0.00 0.00 0.00 0.00 10.99 ± 0.03 The F-value of 23.11 implies that the model is significant 14 2.00 0.00 0.00 0.00 11.42 ± 0.07 and that the model was accurate in describing the experimen- 15 1.00 − 1.00 − 1.00 − 1.00 9.27 ± 0.09 tal data. The Lack of Fit F-value was 1.49. A non-significant 16 0.00 0.00 0.00 0.00 10.82 ± 0.06 Lack of Fit value means that the model fits. The R coefficient indicates good agreement between experimental and predicted 17 − 1.00 1.00 − 1.00 1.00 11.92 ± 0.07 18 − 1.00 1.00 1.00 1.00 10.56 ± 0.04 data for a microbiological process and suggests that the model is reliable for depicting DLA production by L. coryniformis 19 1.00 − 1.00 1.00 − 1.00 9.16 ± 0.03 subsp. torquens. 20 0.00 0.00 0.00 2.00 10.90 ± 0.08 Figure 2 shows the surface plots obtained using the model 21 1.00 1.00 − 1.00 1.00 11.92 ± 0.07 equation. The main factors were meat extract and yeast extract 22 0.00 0.00 0.00 − 2.00 8.67 ± 0.04 concentrations (Fig. 2a–c) while sodium acetate and 23 − 1.00 − 1.00 − 1.00 1.00 10.35 ± 0.03 dipotassium phosphate had little effect (Fig. 2b–d). The results 24 0.00 0.00 0.00 0.00 10.88 ± 0.06 showed that increasing concentrations of meat extract and 25 1.00 1.00 − 1.00 − 1.00 11.19 ± 0.04 yeast extract resulted in increased DLA concentration. 26 1.00 − 1.00 1.00 1.00 11.42 ± 0.04 However, high concentrations of these products represent 27 − 1.00 − 1.00 1.00 − 1.00 7.06 ± 0.06 higher production cost. Therefore, in order to make the pro- 28 0.00 0.00 0.00 0.00 10.28 ± 0.05 cess economically feasible, a lower concentration of these 29 − 1.00 1.00 − 1.00 − 1.00 10.34 ± 0.03 products must be used without, however, affecting the perfor- 30 1.00 1.00 1.00 1.00 12.17 ± 0.07 mance of the microorganism. 552 Ann Microbiol (2018) 68:547–555 Table 3 Analysis of variance Source Sum of squares df Mean square F value P value Prob > F (ANOVA) for the selected model Model 36.46 6 6.08 23.11 < 0.0001 A–sodium acetate 4.83 1 4.83 18.38 0.0003 B–meat extract 17.66 1 17.66 67.19 < 0.0001 C–dipotassium phosphate 0.18 1 0.18 0.68 0.4184 D–yeast extract 11.19 1 11.19 42.58 < 0.0001 AC 1.27 1 1.27 4.84 0.0382 D 1.32 1 1.32 5.02 0.0350 Residual 6.05 23 0.26 Lack of fit 5.09 18 0.28 1.49 0.3508 Pure error 0.95 5 0.19 < 0.0001 Cor total 42.51 29 0.0003 2 2 [R =0.858; Adj R =0.821] From a technical point of view, the highest DLA con- optimization. However, the economic aspect must also be centration should be considered the main objective of this considered. In order to optimize the medium (considering Fig. 2 Surface plots for interactions between the different components of acetate and meat extract, (c) dipotassium phosphate and meat extract, the media that were optimized to increase the DLA production by L. and (d) dipotassium phosphate and sodium acetate. The effect of two coryniformis subsp.torquens at 12h fermentation time where (a) variables with the other two variables maintained at their respective zero represents the effect of yeast extract and meat extract, (b) sodium Ann Microbiol (2018) 68:547–555 553 Table 4 Medium modified using Constraints the Design Expert 7.0 optimization tool for DLA Factor Goal Lower limit Upper limit Importance production by L. coryniformis Sodium acetate Minimize −22 3 subsp. torquens Meat extract Minimize −22 5 Dipotassium phosphate Minimize −22 3 Yeast extract Minimize −22 4 D(−) lactic acid Maximize 7.06 12.29 5 Solution-optimization Factor Coded values Real values (g/L) Desirability Sodium acetate − 2.00 0.05 0.715 Meat extract − 1.08 3.23 Dipotassium phosphate − 2.00 0.05 Yeast extract − 0.43 2.63 Response Prediction SE Pred 95% PI low 95% PI high D-lactic acid (DLA, g/L) 9.70 0.80 8.04 11.35 both the technical and economic point) two aspects were In order to validate the model, DLA was produced by targeted: increasing DLA concentration and reducing the L. coryniformis subsp. torquens using MRS medium and concentration of components, mainly meat extract. Under the modified medium under the same conditions of the these conditions, the concentrations were optimized using shake flask experiments. The concentration of DLA using the quadratic model of the Design Expert 7.0 optimization the modified medium was 10.72 g/L with 0.89 g/L.h of tool. The parameters and results were summarized in productivity. This result corroborates the model prediction Table 4. The final modified medium contained 3.23 g/L (Table 4) and shows the adequacy of the model. The con- of meat extract, 2.63 g/L of yeast extract, 0.05 g/L of centration of DLA using MRS medium was 12.69 g/L. sodium acetate, and 0.05 g/L of dipotassium phosphate. This concentration was 1.19-fold higher than the concen- The differences are the reductions in concentrations of tration obtained when the modified medium was used meat extract from 10 to 3.23 g/L, yeast extract from 5.0 (10.72 g/L). This difference is a result of the reduction to 2.63 g/L, sodium acetate from 5.0 to 0.05 g/L, and in the number and concentrations of medium components. dipotassium phosphate from 2.0 to 0.05 g/L. The desir- From the technical-economic point of view, the modified ability coefficient of 0.715 was obtained, indicating that medium showed a satisfactory result since the decrease in approximately 72% of the production and cost reduction the DLA concentration was only 20%, especially when requirements were achieved. considering the higher cost of the MRS medium. Fig. 3 Profile of glucose consumption (□), DLA production (○) and acetic acid production (Δ)in batch fermentation by L. coryniformis subsp. torquens using the modified medium at 37 °C, 120 rpm and pH 6.8 controlled with NaOH (4 M) 554 Ann Microbiol (2018) 68:547–555 Bustos et al. (2004) reported a model with higher produc- the carbon source and obtained a lactic acid production of tion using 5 g/L of CSL, 3.6 g/L of yeast extract, and 10 g/L of 57 g/L and yield factor of 0.63 g/g of reducing sugar after peptone for LA production by L. coryniformis subsp. torquens. 48 h at pH 6.2 and Wang et al. (2016) reported lactic acid The authors obtained a concentration of 58.9 g/L and 0.61 g/ concentration of 36 g/L and yield factor of 0.69 g/g at pH L.h of productivity using 100 g/L of glucose after 96 h. The 6.25 by L. rhamnosus LA-04-1. productivity reported by these authors are lower than the one The results of this study showed that the modified medium, obtained in this study, where lower concentrations of expen- which has a lower concentration of nutrients, lead to the total sive nitrogen sources were employed. The DLA concentration consumption of the glucose by L. coryniformis subsp. and productivity (10.72 g/L and 0.89 g/L) obtained using the torquens. The DLA production showed no inhibitory effects modified medium was also comparable to the results obtained by nutrient limitation or by-products formation, and a 0.95 by Slavica et al. (2015). These authors obtained 12.95 g/L of yield factor was obtained. DLA and productivity of 0.99 g/L.h by L. coryniformis subsp. torquens using MRS medium after 13 h of fermentation. Özcelik et al. (2016) found LA concentrations ranging from Conclusions 0.27 to 0.56 g/L for eight lactic acid bacteria (LAB) strains in MRM medium after 4 days of fermentation at 37 °C. Cingadi In this study, the preliminary screening showed the significant et al. (2015) reported that homo-fermentative LAB strains effect sodium acetate, meat extract, yeast extract, glucose, and studied for DLA production by batch culture in MRS medium dipotassium phosphate had on the DLA production. These produced between 0.22 and 11.32 g/L by L. coryniformis bac- components were studied using a central composite rotatable teria (NCDC367, NCDC368, NCDC369) produced DLA acid design, and a quadratic model was obtained to describe the ranging from 9.76 to 11.32 g/L in shake flask experiments. relationship between the DLA production and the medium components. The optimization of model employing the DLA production in bioreactor Design-Expert 7.0 tool resulted in the modified medium with the following composition: 2.3 g/L of meat extract, 3.58 g/L Batch fermentation was carried out in a bioreactor to study the of yeast extract, 0.05 g/L of sodium acetate, and 0.05 g/L of DLA production using the modified medium under controlled dipotassium phosphate. The scale-up from flask to bioreactor operating conditions. Figure 3 shows the profiles of glucose uti- under controlled pH condition using the modified medium lization and organic acid production during the fermentation at showed the total consumption of glucose by L. coryniformis pH 6.8 and initial glucose of 33.85 g/L. Under controlled pH subsp. torquens without inhibitory effects, little formation of conditions, glucose was completely consumed by L. by-product acetic acid, and yield factor of 0.95 g/g. The mod- coryniformis subsp. torquens after 38 h—and DLA production ified medium obtained in this work could contribute to the was increased due to the reducing of inhibition by acid products development of processes economically viable for the optical- formation. DLA with an optical purity of ≥ 99.0% was produced ly pure DLA production. with concentration of 33.6 g/L, yield factor of 0.95 g/g, produc- tivity of 0.88 g/L · h, and 95% of fermentation efficiency. A small Acknowledgements The authors gratefully acknowledge financial sup- quantity of acetic acid was also produced but other products were port by Leopoldo Américo Miguez de Mello Research and Development not identified. The acetic acid production was 0.31 g/L at the end Center (Cenpes), Petrobras. We also would like to acknowledge the sup- port of the Laboratories of Bioprocess Development (Ladebio), from the of the fermentation. Acetic acid was still being produced while school of chemistry of Federal University of Rio de Janeiro, UFRJ. DLA concentration remained constant in the stationary phase. These results of yield factor (0.95 g/g) and productivity Compliance with ethical standards (0.88 g/L · h) were observed to be high in contrast to DLA production by Lactobacillus delbrueckii IFO3202 reported for Conflict of interest The authors declare that they have no conflict of Tanaka et al. (2006) with an estimated yield factor of 0.9 g/g interest. and productivity of 0.75 g/L · h using the MRS medium and after 36 h from 30 g/L of glucose at pH 6.0. In our present study, yield factor was similar to that obtained by Zhao et al. (2010) who reported conversion of glucose to DLA of 0.94– References 0.99 g/g for repeated production in one-reactor system at Abdel-Rahman MA, Tashiro Y, Sonomoto K (2011) Lactic acid produc- pH 5.6–5.8 by Sporolactobacillus sp. CASD. Others studies tion from lignocelluse-derived sugars using lactic acid bacteria: for production and optimization of lactic acid in batch overview and limits. J Biotechnol 156:286–301 fermentation have presented lower yield factor values, for Abdel-Rahman MA, Tashiro Y, Sonomoto K (2013) Recent advances in example Bernardo et al. (2016)used Lactobacillus rhamnosus lactic acid production by microbial fermentation processes. B103 and lactose (approximately 90 g/L of reducing sugar) as Biotechnol Adv 31:877–902 Ann Microbiol (2018) 68:547–555 555 Abdelwahed NAM, Gomaa EZ, Hassan AA (2017) Statistical modelling Sources of Renewable Energy. Springer, Dordrecht, pp 211–228. https://doi.org/10.1007/978-90-481-3295-9_11 and optimization of fermentation medium for lincomycin production by Streptomyces lincolnensis immobilized cells. Braz Arch Biol Myers RH, Montgomery DC, Anderson-Cook CM (2009) Response sur- Technol 60:e160210 face methodology: process and product optimization using designed Almeida DG, Soares da Silva RCF, Luna JM, Rufino RD, Santos VA, experiments, 3rd edn. Wiley, New York Sarubbo LA (2017) Response surface methodology for optimizing Nakano S, Ugwu CU, Tokiwa Y (2012) Efficient production of D-(-)- the production of biosurfactant by Candida tropicalis on industrial lactic acid from broken rice by Lactobacillus delbrueckii using waste substrates. Front Microbiol 8:157 Ca(OH) as a neutralizing agent. Bioresour Technol 104:791–794 Bernardo MP, Coelho LF, Sass DC, Contiero J (2016) L-(+)-Lactic acid Nampoothiri KM, Nair NR, John RP (2010) An overview of the recent production by Lactobacillus rhamnosus B103 from dairy industry developments in polytactide (PLA) research. Bioresour Technol waste. Braz J Microbiol 47:640–646 101:8493–8501 Bustos G, Moldes AB, Alonso JL, Vazquez M (2004) Optimization of D- Narayanan N, Roychoudhury PK, Srivastava A (2004) L (+) lactic acid lactic acid production by Lactobacillus coryniformis using response fermentation and its product polymerization. Electron J Biotechnol. surface methodology. Food Microbiol 21:43–148 https://doi.org/10.2225/vol7-issue2-fulltext-7 Chauhan K, Trivedi U, Patel KC (2007) Statistical screening of medi- Naveena BJ, Altaf MD, Bhadriah K, Reddy G (2005) Selection of medi- um components by Plackett-Burman design for lactic acid pro- um components by Plackett–Burman design for production of L(+) duction by Lactobacillus sp, KCP01 using date juice. Bioresour lactic acid by Lactobacillus amylophilus GV6 in SSF using wheat Technol 98:98–103 bran. Bioresour Technol 96(4):485–490 Cingadi S, Srikanth K, Arun EVR, Sivaprakasam S (2015) Statistical Nguyen CM, Choi GJ, Choi YH, Jang KS, Kim JC (2013) D- and L-lactic optimization of cassava fibrous waste hydrolysis by response sur- acid production from fresh sweet potato through simultaneous sac- face methodology and use of hydrolysate based media for the pro- charification and fermentation. Biochem Eng J 81:40–46 duction of optically pure D-lactic acid. Biochem Eng J 102:82–90 Ojha N, Das N (2018) A statistical approach to optimize the production of Dasgupta D, Suman SK, Pandey D, Ghosh D, Khan R, Agrawal D, Jain Polyhydroxyalkanoates from Wickerhamomyces anomalus VIT- RK, Vadde VT, Adhikari DK (2013) Design and optimization of NN01using response surface methodology. IJBM 107:2157–2170 ethanol production from bagasse pith hydrolysate by a Özcelik S, Kuley E, Özogul F (2016) Formation of lactic, acetic, succinic, thermotolerant yeast Kluyveromyces sp. IIPE453 using response propionic, formic and butyric acid by lactic acid bacteria. LWT Food surface methodology. Springerplus 2:159 Sci Technol 73:536–542 Dubey S, Singh A, Banerjee UC (2011) Response surface methodology Shen L, Worrel E, Patel M (2010) Present and future development in of nitrilase production by recombinant Escherichia coli.Braz J plastics from biomass. Biofuels Bioprod Biorefin 4:25–40 Microbiol 42:1085–1092 Slavica A, Trontel A, Jelovac N, Kosovec Z, Santek B, Novak S (2015) Gao C, Ma C, Xu P (2011) Biotechnological routes based on lactic acid Production of lactate and acetate by Lactobacillus coryniformis production from biomass. Biotechnol Adv 29:930–939 subsp. torquens DMS 20004T in comparison with Lactobacillus Hertzberger RY, Pridmore RD, Gysler CH, Kleerebezem M, Texeira de amylovorus DMS 20531T. 2015. J Biotechnol 202:50–59 Mattos MJ (2013) Oxygen relieves the CO and acetate dependency Tanaka T, Hoshina M, Tanabe S, Sakai K, Ohtsubo S, Taniguchi M of Lactobacillus johnsonii NCC533. PLoS One. https://doi.org/10. (2006) Production of D-lactic acid from defatted rice bran by simul- 1371/journal.pone.0057235 taneous saccharification and fermentation. Bioresour Technol 97: Hofvendahl K, Hahn-Hägerdal B (2000) Factors affecting the fermenta- 211–217 tive lactic acid production from renewable resources. Enzym Microb Wang Y, Chan C, Cai D, Wang Z, Qin P, Tan T (2016) The optimization Technol 26:87–107 of L-lactic acid production from sweet sorghum juice by mixed Imran M, Anwar Z, Irshad M, Javid A, Hussain A, Ali S (2017) fermentation of Bacillus coagulans and Lactobacillus rhamnosus Optimization of cellulase production from a novel strain of under unsterile conditions. Bioresour Technol 218:1098–1105 Aspergillus Tubingensis IMMIS2 through response surface method- Wee YJ, Kim JN, Ryu HW (2006) Biotechnological production of lactic ology. Biocatal Agric Biotechnol 12:191–198 acid and its recent applications. Food Technol Biotechnol 44(2): Jem KJ, van der Pol JF, de Vos S (2010) Microbial lactic acid, its polymer 163–172 poly(lactic acid), and their industrial applications. In: Chen GQ (ed) Zhao B, Wang L, Li F, Hua D, Ma C, Ma Y, Xu P (2010) Kinetics of D- Plastics from Bacteria, vol. 14. Springer, Berlin/Heidelberg, pp 323– lactic acid production by Sporolactobacillus sp. strain CASD using repeated batch fermentation. Bioresour Technol 101:6499–6505 Kongruang S, Kangsadan T (2015) Optimization of succinic acid produc- tion from crude glycerol by encapsulated Anaerobiospirillum ZhuLW, Wang CC, LiuRS, Li HM,Wan DJ,TangYJ succiniciproducens using response surface methodology. IJBBB (2012) Actinobacillus succinogenes ATCC 55618 fermentation me- 5(1):11–25 dium optimization for the production of succinic acid by response Li Y, Cui F (2010) Microbial lactic acid production from renewable re- surface methodology. J Biomed Biotechnol. https://doi.org/10.1155/ sources. In: Singh OV, Harvey SP (eds) Sustainable Biotechnology: 2012/626

Journal

Annals of Microbiology – Springer Journals

Published: Aug 14, 2018

There are no references for this article.