Lactic Acid Bacteria (LAB) play an essential role in biotechnology. They have a wide range application in food, cosmetic and pharmaceutical industry. These are the most important groups of microorganisms used in the food industry. The LAB have contributed in the increased volume of fermented foods worldwide to their roles in biopreservation and in modulating the health of their hosts by their bacteriocin and lactic acid production (Soomro et al., 2002).
Therefore, their growth requires rich and complex media based on yeast extract or meat and animal peptones (Ashraf and Shah, 2011). They are also characterized by high polyauxotrophic requirements, for various amino acids, peptides, nucleotides, vitamins and fatty acids, owing to their limited ability to biosynthesize Bvitamins and amino acids (Desmazeaud, 1983; Fitzpatrick and O'keeffe, 2001).
In order to choose appropriate growth medium, different aspect have to be considered: Cost, the ability to reach a high number of cells and harvesting methods (Georgieva et al., 2009), Owing to the fastidious growth factor requirements of LAB and since their sensitivity to inhibitors, selective culture media are very difficult to formulate (DjeghriHocine et al., 2006). Many animal supplementations have been tested such as: ram horn peptone (Kurbanoglu, 2004) and vegetal supplementations as Cassava Bagasse (John et al., 2008), potato extracts (Gaudreau et al., 2002), molasses (Coelho et al., 2011), oilseed crop peaand chickpea (DjeghriHocine et al., 2007) cereals (Charalampopoulos et al., 2002). Plant extracts are generally more economical than meat extracts and there are reports of their use with LAB (Gaudreau et al., 2002).
Nutritional requirement can be determined by the statistical methods. These methods offer several advantages; they are fast, reliable and reduce the number of experiments. They have been used in many optimization processes of fermentation (Coelho et al., 2011; Naveena et al., 2005). The conventional experimental approaches for the optimization of media by the method "Onevariableatatime" requires a large number of experiments and not take account the interactions between variables while statistical approach can determine the effect of variables and their interactions with a limited number of essays and the results are achieved in an economical manner.
Lactobacillus plantarum is one of the most of lactic acid bacteria used traditionally in fermented food such as meats, vegetables and dairy products (Hwang et al., 2012). There are few published literatures undertaking biomass production of LAB. In the present study, we report optimization of growth medium basis on split pea and dry figs for the growth of Lactobacillus plantarum BH14 using statistical approach.
MATERIALS AND METHODS
Microorganism: Lactobacillus plantarum BH14 isolated from camel milk and preidentified by the team of Karem NE KaremZADI: Laboratory of Biology of Microorganisms and Biotechnology, Faculty of Science, Department of Biotechnology, University of Essenia.
Substratebased fermentation media
Split pea juice: Seeds were manually sorted and rinsed, before being powdered. The 50 g of the powder were then dissolved in 450 mL of distilled water at pH 9, magnetically stirred during 30 min, pH of the resulting slurry were adjusted at 6 to add papaine. This slurry was heated in 60°C water bath for 2 h then centrifuged at 3500 rpm min1 for 15 min. The supernatant was heated in a boiling water bath for 20 min and recentrifuged 4000 rpm min1 for 15 min. The supernatant used as a basis for the preparation.
Dry figs juice: The 250 g of dry figs were rinsed and added to 500 mL of distilled water. The mixture was boiled at 90°C h1. 100 g of the result was dissolved in 450 mL of distilled water, magnetically stirred during 30 min and centrifuged at 4000 rpm min1 for 20 min. The supernatant was added to split pea juice (v/v) and used as a basis for medium reconstitution.
Inoculum preparation: Colonies isolated from MRS agar were precultured in MRS broth for 18 h at appropriate temperature. A precultured cells with an optical density 0.6 were used to inoculate the fermentation media. Incubation lasted 18 h at 37°C.
Cell enumeration: The growth of the microorganism was estimated by the determination of colonyforming units (CFU mL1). This procedure involves making decimal serial dilutions of the sample in sterile physiological water. The growth was recorded as Colony Forming Units (CFU) in various media 1 mL of each fermentation sample was decimally diluted in sterile saline solution (9 g L1 of NaCl). One milliliter of dilution 78 were pourplated in the MRS agar media contained 0.05 g of cysteine, to reduce the fraction of oxygen present in the media, thereby promoting higher survival bacteria (Dave and Shah, 1996). After its solidification, petri dishes were incubated for 48 h at 37°C. After incubation, Colony Forming Units (CFU) were counted.
Screening of nutrients and physical parameters using PlackettBurman design: PlackettBurman design is an efficient way to screen the important factors "critical factors" among a large number of variables with minimal number of experiments. It allows the investigation of up to N1 variables with N experiments. This design was used to determine the most influential of 9 variables which includes chemical parameters (glucose, lactulose, tween80, K2HPO4,, sodium acetate, and MgSO4) and physical parameters (pH, shaking, inoculum size ) at two levels (+1 and 1). The low and high levels were shown in Table 1 (Plackett and Burman, 1946). Twelve experiments were generated from those factors. The variables with a confidence level (1α) ≥70% were considered to have a significant influence on the growth of L. plantarum. All the ingredients were dissolved in the basis media.
Optimization by response surface methodology: The next step in the formulation of the medium was to determine the optimum levels of significant variables giving the optimal growth. For this purpose, the Response Surface Methodology (RSM), using a Central Composite Design (CCD) of Box and Wilson (1951), was adopted. Three important variables (glucose, lactulose and MgSO4) from the result of PlackettBurman design were selected for further optimization. Using the CCD method, a total of 20 experiments with various combinations of glucose, lactulose and MgSO4 were conducted. Table 2 displays the range and levels of the variables investigated.
The variables of the experiments were coded according to the following equation:
where, xi is the coded value of an independent variable, Xi is the real value of an independent variable, XCP is the real value of an independent variable at the center point and ΔXi is the step change value.
|Table 1:||Variables and levels used in PlackettBurman design|
|Table 2:||Experimental codes, ranges and levels of the independent variables for response surface methodological experiment|
The behavior of the system was described by the following quadratic equation:
where, Y is the predicted response ( log10 CFU mL1) ; b0 is the offset term, bi is the linear effect, bii is the squared effect; bij is the interaction effect and xi is the independent variable. After determining the composition of the optimum fermentation medium, L. plantarum was grown in our final medium and in the reference medium (vegetal MRS) for comparison.
Data analysis: Minitab essay version 16 was used to fit the experimental PlackettBurman design and also the quadratic response surface model to the experimental data through multiple regressions analysis.
Screening of significant variables using PlackettBurman design: A total of nine variables were analyzed with regard to their effects on growth of L. plantarum using a PlackettBurman design. This last is an efficient way to screen for the important factors "Critical factors" among a large number of variables with minimal number of experiments. According to Paratoes law, initial screening of the ingredients is done to understand the significance of their effect on the product formation and then a few better ingredients are selected for further optimizations (Naveena et al., 2005). The results were analyzed using the software MINITAB 16. Table 3 displays the PlackettBurman design matrix (real and coded values) of the 12 experiments and the respective results (log 10 CFU mL1). Whereas, Table 4 represents the coefficient, tvalue and pvalue of each variable. Factors evidencing pvalues of less than 0.3 were considered to have significant and were therefore, selected for further optimization studies effects on the response (cell number).
As can be seen from Table 3, the highest cell number of 10.16 log10 CFU mL1 was obtained in combination N°5.
Glucose, lactulose, MgSO4 and shaking was the influential variables on the growth of L. plantarum BH14, those variables had a significant effect on growth of this strain at a 70% confidence level. Glucose, lactulose and MgSO4 showed positive coefficient but shaking had a negative coefficient, these results suggest that L. plantarum strain is an anaerobic microorganism. Thus shaking was avoided in subsequent experiments.
While the confidence level of tween 80, K2HPO4, sodium acetate, MnSO4, pH and inoculum size was below 70% in the growth, hence, these variables were considered insignificant. Figure 1 (Pareto chart) illustrates the effects of these variables.
The addition of phosphate to the culture medium increases the growth of the microorganism and it maintains the pH near the optimal growth value, thereby allowing the conduction of fermentation for a longer time. (Coelho et al., 2011) so, K2HPO4 was used at 2 g L1 in the rest of media formulation. The above results indicated that the PlackettBurman design is a powerful tool for identifying factors, which had significant influence on L. plantarum growth.
The critical factors were finding by PlackettBurman design, the next step was to determine the optima of those variables using a Composite Central Design (CCD).
Response surface methodology: To determine the optimum conditions of fermentation of L. plantarum in split pea and dry figs basis medium, the RSM was employed. Table 5 displays the design matrix of the variables in coded units and real values with the respective results. The highest production of biomass was 8.46 log10 CFU mL1, obtained from combination number 3:20 g L1 glucose, 20 g L1 lactulose and 0.15 g L1 MgSO4.
The application of multiple regression analysis methods yielded the following regression (Eq. 3) for the experimental data and explained the role of each variable and their second order interactions in the growth:
|Table 3:||PlackettBurman design (real and coded values) with the respective results|
|*D1 and D2 represent dummy variables|
|Table 4:||Estimated effects of the tested variables on growth of Lactobacillus plantarum using PlackettBurman design|
|Fig. 2:||Response surface of log10 CFU mL1 of Lactobacillus plantarum BH14 showing the interaction between glucose and lactulose|
|Table 5:||Central composite design and results|
|Table 6:||Analysis of variance for Lactobacillus plantarum growth|
|DF: Degrees of freedom, Seq SS: Sum of squares, Adj SS: Adjusted sum of squares, Adj MS: Adjusted mean sum of squares, F: Variance ratio; p: Probability|
The quadratic model in Eq. 3 contains three linear terms, three quadratic terms and three factorial interactions in which Y is the predicted response (log 10 CFU mL1) and X1, X2 and X3 are the coded values of glucose, lactulose and MgSO4 respectively. Table 6 displays the Students tdistribution and the probability pvalues that used as a tool to check the significance of each of the coefficients.
Our results showed that the effects of glucose (X1), lactulose (X2) and MgSO4 (X3) on the cell number of L. plantarum were significant based on pvalues lower than 0.05. Also, the squared variable X12 and the X1X2 interaction, X1X3 interaction were also significant for the growth of L. plantarum. The Eq. 3 model was modified to the reduced Eq. 4 fitted model:
Analysis of variance for log10 CFU mL1 was done by MINITAB16. The regression equation obtained from the ANOVA showed that the R2 (multiple correlation coefficient) was 94.38% (avalue >0.75 indicates fitness of the model). This is an estimate of the fraction of overall variation in the data accounted by the model and thus the model is capable of explaining 95% of the variation in response. In addition a very low pvalue (0.000) demonstrated a very high significance for the regression model and smaller pvalues denote a more significant corresponding coefficient (Coelho et al., 2011).
Graphical representation of response surface shown in Fig. 24 help to understand the effect of glucose, lactulose and MgSO4 on the growth, visualize their interactions and locate the optimal level of each variable for maximal response. Threedimensional response surface plots were constructed by plotting the response (log10 CFU mL1) on the Zaxis against any two independent variables, while maintaining other variables at their optimal levels.
|Fig. 3:||Response surface of log10 CFU mL1 of Lactobacillus plantarum BH14 showing the interaction between lactulose and MgSO4|
|Fig. 4:||Response surface of log10 CFU mL1 of Lactobacillus plantarum BH14 showing the interaction between glucose and MgSO4|
Figure 2 explains that decrease in concentration of both glucose and lactulose can increase the growth of L. plantarum. Figure 3 shows that the response (log10 CFU mL1) varied significantly when glucose concentration increased and MgSO4 concentration decreased. Figure 4 explains that a maximum cell number was obtained at a high level of MgSO4 and intermediate level of lactulose. An optimized formulation of nutrition levels was suggested from the software at the following concentrations: 11.59 g L1 glucose, 11.59 g L1 lactulose and 0.23 g L1 MgSO4.
Verification experiment: Growth of L. plantarum on our final medium resulted in a higher final cell number than that recorded on vegetal MRS medium, 3.9 109 and 2.5 109 CFU mL1, respectively.
Lactic acid bacteria are characterized by high polyauxotrophic requirements. Their growth is possible only on rich and complex culture media which must essentially comprise a source of nitrogen (amino acid, peptide), a carbon source, vitamins, minerals and trace elements. These nutrients must be made to optimal concentrations (DjeghriHocine et al., 2010). Therefore, determining a suitable growth media appears to be difficult.
The use of statistical models to optimize culture medium components and conditions has increased in presentday biotechnology, due to its ready applicability and aptness (Reddy et al., 2008).
Most studies on optimizing medium composition for biomass production have been reported. In our study, PlackettBurman design was used to screen the important factors and central composite design was used to determine their optimum concentrations.
The result showed that glucose, lactulose and MgSO4 were the most significant variables on growth of L. plantarum BH14 and their optimal concentrations in the medium were 11.59, 11.59 and 0.23 g L1, respectively.
The strains of Lactobacillus genus are chemotrophic, they acquire the energy necessary to their growth from the oxidation of sugars and other chemical compounds. As shown in Fig. 1, glucose was the most variable influencing the growth of L. plantarum BH14. Glucose presented the best carbon source for biomass production of L. plantarum YJG (Han et al., 2011). It has been reported that a good growth rate of L. plantarum was achieved in MRS media supplemented with glucose as carbon source (Georgieva et al., 2009). However, a low glucose concentration increased the growth of L. plantarum BH14, that can be explained by: A higher glucose concentration may have resulted a cell growth inhibition. Recently, the effect of lactulose on the growth of LAB has been widely studied. Lactobacillus plantarum BH14 is able to use lactulose as carbon source, our results are in an agreement with the results of Alejandra CardelleCobas et al. (2011), who found that L. plantarum CLB7 cell density reached a maximum OD600 value using lactulose as substrate. Saarela et al. (2003) are reported that lactulose was the favored lactose derivative used by Lactobacillus strains. Also, MgSO4 had a significant effect on growth of L. plantarum BH14, it represent a sources of oligoelements serving as cofactors of enzymes involved in the growth of this strain.
Both juices, originated from split pea and dry fig, appeared to be convenient basis for the formulation of specific media for L. plantarum growth with low cost. The results showed that viable counts in the optimized culture medium were significantly higher than those in the vegetal MRS medium. This indicates that our newly developed culture medium is more adaptable to the growth of L. plantarum than reference medium. Thats due to the richness of chemical composition of split pea in protein and starch (Arntfield and Maskus, 2011). Also, dry figs are rich on vitamins, mineral elements and fats. This work confirmed the feasibility of the use of a vegetal substrate to replace the expensive usual nitrogen supplements like yeast extract, meat extract and peptones used on the formulation of culture media for lactic acid bacteria growth.
The optimized medium in this study is more economical and good for the growth of L. plantarum. For these reasons, it could be useful in large scale application.
The authors would like to thank Prof. Ladjama Ali director of laboratory of applied biochemistry and microbiology and Dr M. Mechakra for their helpful collaboration during the experimental work.