Optimization of Media Composition for Maximum Growth of Probiotic Lactobacillus fermentum NBC-08 Using Response Surface Methodology

Article Info Abstract: In this study, it was aimed to determine the medium where Lactobacillus fermentum microorganisms with probiotic properties grow most efficiently by using Response Surface Method (RSM). Studies have been conducted on a 500 ml erlenmeyer scale, the medium optimization of the Lactobacillus fermentum strain was carried out according to the Central Composite Design (CCD) included in RSM. The effects of glucose, yeast extract, inorganic salts, and Tween 80 were investigated on the growth rate of the L. fermentum NBC-08 strain. Samples were taken at regular intervals from the erlenmeyer flask and the number of viable cells was measured by planting them in petri agar medium. In the study, the number of viable cells log10 (cfu/ml) was chosen as the response variable. As a result of the study, it was concluded that glucose and yeast extract are absolutely essential components in the medium. The optimum medium composition was found as 96.06 g/L glucose, 40.76 g L-1 yeast extract, 19.43 g L-1 inorganic salts, and 11.01 ml/L Tween 80. The production of the maximum L. fermentum strain was determined as 10.75 log10 (cfu ml-1). It is predicted that this study will make positive contributions to the fermentation conditions and medium optimization studies for production of lactic acid bacteria. Received: 10.09.2021 Accepted: 10.02.2022 Online Published: 15.03.2022 DOI: 10.29133/yyutbd.993781


Introduction
Probiotics are defined as living microorganisms that cause positive effects on host health when consumed in certain amounts (Anvari et al., 2014). Probiotics affect the host by regulating mucosal and systemic immunity and are thus beneficial microorganisms. Most probiotic bacteria are gram-positive, and their main functions are related to the maintenance of intestinal system health (Marco et al., 2006). Lactic acid bacteria constitute the most important group of probiotic microorganisms. Among the lactic acid bacteria (LAB), Bifidobacterium and Lactobacillus species stand out as the most widely used as a probiotic (Iranmanesh et al., 2014). LAB is an important group of fermentative bacteria that are widely used as starter cultures for the production of fermented foods both at home and in industry, and it is also part of the normal intestinal microflora (Pithva et al., 2014). Numerous studies have shown that LAB has a wide application area as a probiotic organism and provides beneficial effects to humans and animals (Lee et al., 2010;Shamekhi et al., 2012;Kaewnopparat et al., 2013). The beneficial effects of LAB on human health and intestinal microbial balance stand out as valuable probiotic properties. Lactobacillus strains are characterized by their ability to adhere to and colonize the intestinal mucous layer and to produce antimicrobial agents (Belova et al., 2016;Boricha et al., 2019;Ozkan et al., 2020;Teame et al., 2020). These properties allow probiotic strains to compete with gastrointestinal tract pathogens and food degrading pathogens (Bove et al., 2013). LAB species were listed in the qualified presumption of safety published by the European Food Safety Authority (EFSA) and generally recognized as safe (GRAS) by the US Food and Drug Administration (FDA) (Ricci et al., 2017).
Lactobacillus fermentum is a bacterium with probiotics including health-promoting effects, and it widely inhabits the human gastrointestinal and urogenital tracts. It is present naturally in raw milk, dairy products, and many traditional fermented foods and beverages. L. fermentum bacteria is a microorganism whose main task is to produce lactic acid (Ramos et al., 2013). It is considered as beneficial for developing various methods to ensure that large quantities of L. fermentum biomass can be produced effectively and cost-effectively. Optimization of the culture medium to obtain high-quality probiotic products is one of the most important steps in improving the efficiency of the production process in terms of both economic and yield-related aims. (Gao et al., 2009).
Many studies have been conducted related to L. fermentum strain isolated from breast milk. Also, there have been studies that showed L. fermentum strain has various benefits on the gastrointestinal system and cholesterol-lowering effects (Gil-Campos et al., 2012;Maldonado et al., 2012;López-Huertas, 2015;Asan-Ozusaglam and Gunyakti, 2019).
Response Surface Methodology (RSM) is a set of mathematical and statistical techniques useful for analyzing the relationship between several independent variables and one or more responses. (Liew et al., 2005;Kanmani et al., 2012). Compared to traditional methods, RSM has many advantages, such as requiring a minimum number of experiments, lower processing costs, and using fewer medium components (Rodrigues et al., 2006). Additionally, unlike traditional methods, RSM shows the combined effects of all independent variables on response variables (Anvari et al., 2014). Recently, RSM has been widely used in many bioprocessing studies, including modeling and optimization of culture media (Stephenie et al., 2007;Shang et al., 2013;Thite et al., 2020).
The L. fermentum NBC-08 strain used in this study was isolated from breast colostrum milk. L. fermentum NBC-08 strain is a local probiotic strain originating from Turkey. In addition, the low-cost medium developed in this study can be used on an industrial scale. This aspect makes the study original and significant. In this study, optimization of media components for the maximum growth of L. fermentum NBC-08 probiotic strain isolated from breast milk colostrum was performed using the RSM. olduğu sonucu ortaya çıkmıştır. Optimum besi yeri bileşimi 96.06 g/L glikoz, 40.76 g/L yeast ekstrakt, 19.43 g/L inorganik tuzlar ve 11.01 ml/L Tween 80 olarak bulunmuştur. Maksimum L. fermentum suşunun üretimi ise 10.75 log10 (cfu/ml) şeklinde bulunmuştur. Bu çalışmanın laktik asit bakterilerinin fermantasyon yolu ile üretilmesi ve probiyotik suşların besi yeri optimizasyonu çalışmalarına olumlu katkılar yapacağı ön görülmektedir.

Microorganism
The Lactobacillus fermentum NBC-08 strain used in this study was isolated from breast colostrum milk by Orzax Pharmaceuticals and Chemical Industry Trade Corporation in a project within the scope of the R&D Support Program supported by the General Directorate of Agricultural Research and Policies (TAGEM). The characterization and reliability tests of this strain were performed by Orzax Company. L. fermentum NBC-08 strain was identified on the basis of genus and species by the MALDI-TOF MS method. The L. fermentum NBC-08 strain bile tolerance, gastric acid resistance, lysozyme tolerance, and intestinal adhesion tests were applied for the determination of the probiotic character of the bacteria.

Substrates and Chemicals
In this study, glucose as carbon source, yeast extract as organic nitrogen source, and inorganic salts (sodium acetate, potassium dihydrogen phosphate, magnesium sulphate, and manganese (II) sulphate) which are thought to be needed for the growth of microorganisms and Tween 80, were used to provide a rich culture growth medium. All chemicals used in the experiments are of analytical purity.

Microorganism growth medium and conditions
Lactobacillus fermentum strain was taken from 20% (v/v) glycerol stock and first inoculated into MRS broth (Table 1) medium, and its growth was achieved in a shaking incubator at 37 °C and 200 rpm agitation speed for 24 hours. Bacterial cultures that completed their pre-development were cultivated in MRS Agar (Table 1) culture storage medium by spreading with drigalski spatula and kept in an incubator at 37 °C for 48 hours. L. fermentum strain in petri dishes was inoculated into 500 ml erlenmeyer flasks containing 100 ml of bacteria growth medium with the help of a sterile spatula (loop). At the end of the incubation period, petri dishes were kept at +4 °C, and the cultures were renewed once every 15 days. The L. fermentum strain used in each stage of the study was inoculated in two stages. First, the L. fermentum strain taken from the culture storage medium was inoculated in the bacteria growth medium and then L. fermentum strain taken from bacteria growth medium was inoculated in the production medium. MRS broth medium of 100 ml was put into 500 ml erlenmeyer flasks. The initial pH of the bioprocess media was adjusted to a value of 6.0 by using 0.1 N HCl and 0.1 N NaOH. The brim of the erlenmeyer flasks was sealed with a plug made of cotton and gauze, and the top of the plug was covered with aluminum foil. The medium prepared in this way was sterilized with an autoclave (Alp CL40, Japan) for 20 minutes at 121 °C and 1.2 bar pressure. The sterilized media was cooled to ambient temperature in laminar flow. 1% of the L. fermentum strain in the growth medium was inoculated into growth media of different compositions. The growth of bacteria was ensured for 18 hours in the culture inoculated into the erlenmeyer flasks at 37 °C and in a shaker with an orbital incubator (Stuart SI600, United Kingdom) with an agitation speed of 200 rpm.

Experimental Design
Four of the medium components (glucose concentration (g L -1 ), yeast extract concentration (g L -1 ), inorganic salts (sodium acetate (50%, w/w), potassium dihydrogen phosphate (20%, w/w), magnesium sulphate (20%, w/w), and manganese (II) sulphate (10%, w/w mixture) and Tween 80 (ml L -1 ) were selected as an independent parameter for the RSM. Above mentioned medium components are also present in the MRS broth medium and are thought to be effective in the production of Lactobacillus fermentum strain. The value ranges of the independent variables were determined by preliminary experiments. In the study, all experiments were carried out in 3 parallel, and the arithmetic mean of the obtained results was used.
The experimental design was carried out according to Central Composite Design (CCD), which is included in RSM and is widely used in optimization processes, and the number of live cells log10 (cfu/ml) was selected as the response variable. The independent variables selected for the medium composition and examined ranges are given in Table 2.

Independent variables
Code In the full factorial design for CCD, an experimental matrix consisting of 16 factorials, 8 endpoints, and 6 center points were used, and the total of the experiments was equal to 30. The total number of experiments was calculated using the equation given below.
(2) was used (Liu et al., 2020); (2) Where Y; the number of viable cells (log10 (cfu ml -1 )), b0; a constant coefficient, Xi and Xj are independent variables that affect the response, bi; linear effect, bii; second degree effect, bij; binary interaction between independent variables. Analysis of variance (ANOVA) and three-dimensional surface graphs for CCD were obtained using Design Expert 10.0 (Stat-Ease) software.

Results and Discussion
One of the most important parameters affecting the production of primary and secondary metabolites produced by microorganisms is the medium composition. Among the medium components, the carbon source accounts for approximately 60-80% of the medium cost (Rehm et al., 1987). The maximum availability of valuable metabolites obtained as a result of microbial activities is closely related to the carbon/nitrogen (C/N) ratio in the growth medium where the production takes place (Farezvidal et al., 1992). In previous studies conducted it is stated that microbial cells and metabolite i n n n -1 n 2 Y= β + β X + β X + β X X å å åå production will reach maximum values when the C/N ratio is selected as optimum (Ülger, 1997). In addition, with the thought that inorganic salts and Tween 80 would positively affect the microbial cell formation, experiments were carried out within the framework of the experimental plan given in Table  3. Cultivation was carried out using the medium compositions in Table 3, and at the end of the 18th hour, the samples taken from the Erlenmeyer flasks were diluted as required and inoculated in petri dishes containing MRS broth solid medium with agar. After the petri dishes were incubated for 24 hours in an incubator at 37 °C, the colonies formed ( Figure 1) were counted, and the results were multiplied by the dilution factor, and thus the number of viable cells formed under each experiment was determined. Actual numbers of living cells obtained in the experiment and the living cell numbers estimated by the model are given in Table 3 as cfu/ml on a log10 basis.  In order to understand which model (linear, 2FI, quadratic, and cubic) better fit the experimental data, Design Expert 10 software was used in the analysis of statistical results, and the findings are given in Table 4. By looking at both the program and statistical coefficients, it was observed that the experimental data fit the quadratic model equation better. As a result of the ANOVA, the quadratic model Probe > F value was found to be significant with <0.0001 at a 95% confidence level. It has been determined that the model terms A, B, C, D, BC, CD, A2, B2, C2, D2 where the probe>F value is less than 0.05 are effective medium composition parameters in the production of Lactobacillus fermentum strain. Probe>F value is used to control the effect of coefficients on the result of each component. By using these coefficients, the interaction between independent variables can be determined. Probe> F values of linear, quadratic, and interaction terms and other statistical results are presented in Table 5.
The  The compatibility of the model with experimental data is evaluated based on the value of the multiple correlation coefficient, R2 . R2 shows that suggested linear models can define the changes in the response variable as high as 97.5% and only 2.5% of the total variation cannot be explained by the model. Similarly, the realistically corrected R2Adj term value was determined as 95.2% (Table 5).
High R 2 and R 2 Adj values are an indicator of a close interaction between the experimental results and the values obtained from the model (Wu and Ahn, 2018). From the high R 2 and R 2 Adj values obtained in the study, it is possible to say that experimental and predicted values are very close to each other, indicating the success of the established model. , The adequate precision value was found to be 22.43. This value shows the signal to noise ratio, and this value is expected to be above 4 (Venkateswarulu et al., 2017). Lack of Fit value (0.3679) was found to be insignificant for the model. The low CV (CV = 0.690%) value clearly demonstrates that the deviations between experimental and predicted values are low, and there is a high degree of precision in the experiments performed, and also the model is reproducible (Long et al., 2018). All descriptive statistical data obtained show that the models proposed in this study are significant.
Studentized residues and a normal % probability plot are given in Figure 2.a. Studentized residues measure the number of standard deviations that distinguish actual and estimated values. The normal % probability checks whether the residues are in a normal distribution. The straight line between these two variables shown in Figure 2.a determines whether a response transformation is required (Mona et al., 2011). Experimentally obtained log 10 (cfu ml -1 ) values and estimated log10 (cfu ml -1 ) values are given in Figure 2.b. As it can be understood from Figure 2.b, it is clearly seen that the estimated log10 (cfu ml -1 ) values suggested by the created model and the experimentally obtained log10 (cfu ml -1 ) values conform and can be defined by the model. Four independent parameters, namely, glucose concentration (g L -1 ), yeast extract concentration (g L -1 ), inorganic salts, and Tween 80 (ml L -1 ), appear to significantly affected the production of the Lactobacillus fermentum strain. It is extremely important to determine the interaction of components with each other while modeling studies when parameters are affected by each other. RSM can reveal these interactions between parameters with the help of stratified graphics and 3D graphics. The 3D response surface graphics obtained as a result of the study are given in Figure 3. As can be seen from Figures 3.a, 3.b, 3.c, 3.d and 3.e, the production level of Lactobacillus fermentum strain decreases in media with low and high concentrations of glucose and low and high yeast extract concentrations. In other current studies, this observation has been expressed as a decrease in microbial activity due to the repression effect of some carbon sources at high concentrations in the growth medium (Yun and Ryu, 2001;Myers and Montgomery, 2002;Geckil et al., 2004). It is reported that microbial activity will be negatively affected due to the absence of a sufficient concentration of nitrogen source in the growth medium. However, it is stated that if the nitrogen source at high concentrations is in the growth medium, the microbial product production efficiency will be inhibited (Avonts et al., 2004). In this context, a similar decrease was observed in low and high yeast extract concentrations. In Figure 3.b, 3.d, and 3.f it is seen that the production decreases in the medium where inorganic salts at higher and lower concentrations are present. This can be interpreted as a requirement for ensuring the stability of enzymes that catalyze the reactions necessary for microbial cell formation (Djekrif-Dakhmouche et al., 2006). The same is also valid for Tween 80 containing growth medium (Figures 3.c, 3.e and 3.f).

Conclusion
As a result, optimization of the media components for the production of Lactobacillus fermentum strain by using Design Expert 10 software has revealed the conditions by which the maximum production of L. fermentum NBC-08 strain biomass is possible. Confirmation experiments were carried out under these conditions, and the production levels of L. fermentum strain in the medium composition suggested by the model were examined. The optimum medium composition suggested by the model is 96.06 g L -1 glucose, 40.76 g L -1 yeast extract, 19.43 g L -1 inorganic salts, and 11.01 ml L -1 Tween 80. This medium composition created as a result of this study was coined as NBC media. The production of the maximum L. fermentum strain is 10.75 log10 (cfu ml -1 ). By using the medium composition suggested by model 3 parallel validation experiments were performed, and the production of L. fermentum strain as the result of these experiments was 10.79 log10 (cfu ml -1 ). These values, which are quite close, indicate that the optimization work is successful and reliable. As a controlled study, the L. fermentum NBC-08 strain was grown in MRS Broth medium. As a result of this study, cell viable count in MRS Broth was reached 9.43 log10 (cfu ml -1 ). A significant difference was found when comparing viable cell counts between MRS broth and the medium suggested by the model. In this case, it can be achieved a more viable cell count by using less medium composition.
It is concluded that this study will make important contributions to the literature and industry for optimizing nutrient composition media for producing maximum lactic acid bacteria.