Lactic acid bacteria (LAB) can produce γ-aminobutyric acid (GABA) with antioxidant properties and sedative effects when it binds to GABA receptor in human’s brain. LAB can also produce bacteriocin-like inhibitory substances (BLIS) with antimicrobial capabilities during carbohydrate fermentation. GABA and BLIS are natural compounds with potential health benefits and food preservation properties. Lactobacillus brevis C23 was co-cultured with three different LAB as inducers, which produced the highest GABA content and BLIS activity. They were cultured in various plant-based media to obtain an edible and better-tasting final product over commercially available media like MRS broth. A coconut-based medium with additives was optimized using Response Surface Methodology (RSM) to increase GABA and BLIS production. The optimized medium for maximum GABA production (3.22 ± 0.01mg/mL) and BLIS activity (84.40 ± 0.44%) was a 5.5% coconut medium containing 0.23% glucose, 1.44% Tween 20, 0.48% L-glutamic acid and 0.02% pyridoxine. Due to the presence of GABA, the cell-free supernatant (CFS) as a postbiotic showed a higher antioxidant activity than other food preservatives like nisin and potassium sorbate. Finally, a microbiological test on food samples showed that the postbiotic was more effective than other preservatives at combating the growth of LAB, molds, and coliform bacteria, making it a possible food preservative.
Research Article
Production of Postbiotic Metabolites γ-Aminobutyric Acid (GABA) and Bacteriocin-Like Inhibitory Substances (BLIS) by Lactobacillus brevis C23 Co-culture in Coconut-Based Medium for Potential Psychobiotics and Food Preservatives
https://doi.org/10.21203/rs.3.rs-2836748/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
γ-Aminobutyric Acid (GABA)
Bacteriocin-Like Inhibitory Substances (BLIS)
Co-culture
Optimization
Lactic Acid Bacteria (LAB)
The World Health Organization (WHO) acknowledges that food security, nutrition, and safety are interconnected facets that can have significant implications for human health. It is well-established that consuming contaminated food is a major contributor to global disease burden, with approximately one in every ten individuals experiencing food-borne illness due to inadequate food handling practices. The prevalence of such conditions is alarmingly high, with an estimated 600 million cases of food-borne diseases occurring each year, leading to over 420,000 deaths (Pires and Devleesschauwer 2021). These statistics underscore the critical importance of ensuring food safety to mitigate the risk of food-borne illness. Beyond the realm of food safety, mental health disorders and psychiatric diseases represent a significant global health burden especially after COVID-19 pandemic (Chatterjee and Mukherjee 2020). Hence, this study aims to explore the interconnected nature of food safety and mental health, with a particular focus on the potential of postbiotics with bio-preservative and psychobiotics properties to address these global health challenges.
Lactic acid bacteria (LAB) are a diverse group of microorganisms widely used in food fermentation due to their ability to produce a variety of flavor and texture-enhancing compounds. In recent years, there has been a growing interest in the potential health benefits of LAB-produced compounds, particularly γ-aminobutyric acid (GABA) and bacteriocins (Liu et al. 2020).
γ-Aminobutyric acid (GABA) is a non-proteinogenic amino acid and an antioxidant agent produced by LAB (Diez-Gutiérrez et al. 2022). LAB produces γ-aminobutyric acid (GABA), an antioxidant agent and non-proteinogenic amino acid that is deemed safe by the FDA. GABA can prevent the formation of lipoxidation end-products and oxidation of essential food components, which are the primary culprits of food quality degradation. By increasing antioxidant enzyme activity, GABA can reduce oxidative damage in fruits and prolong their storage time by boosting citrate and amino acid accumulation (Liang et al. 2022). Besides, GABA acts as a health supplement or "psychobiotic" that binds to GABA receptors in the human brain, promoting sleep, reducing anxiety, and exerting calming effects (Yunes et al. 2020). Bacteriocin-like inhibitory substances (BLIS) constitute a significant metabolite synthesized at ribosomes by LAB. They are antimicrobial peptides that have antibacterial properties against closely related bacterial strains. BLIS are used as food bio-preservative; they can inhibit foodborne pathogens in solid or liquid form and eliminate the selected target pathogens (Ng et al. 2020).
A multitude of studies have reported the presence of additional bacteria as a stress signal in co-culture, which can enhance bacteriocin production. Previous investigations have demonstrated that Lactobacillus brevis is capable of producing higher quantities of γ-aminobutyric acid (GABA) than other strains (Jin et al. 2021). However, the potential use of induction systems utilizing lactic acid bacteria (LAB)remains unexplored, representing a novel avenue for investigation. Furthermore, cultivation of LAB in commercially available media can lead to increased production cost due to the need for purification after fermentation to remove the color, smell, and taste of broth. Utilization of natural media, such as plant-based materials, for LAB growth to produce GABA-enriched nutrition could circumvent these issues.
The current study focused on the effect of adding different LAB as inducers by co-culturing with Lactobacillus brevis C23 in different types of plant-based medium on the production of postbiotics that contain GABA and BLIS. The interaction effects of medium supplements on the concentration of GABA and BLIS activity were studied via response surface methodology. Finally, the application of the postbiotic as a bio-preservative in meat and fruit was also examined and discussed.
2.1 Preparation of culture media and agar
De Man, Rogosa, and Sharpe (MRS) broth (HiMedia, Mumbai, India) were used for the cultivation of LAB strains, Brain Heart Infusion (BHI) broth (Merck, Germany) was used for the cultivation of pathogenic strain. MRS agar (HiMedia, Mumbai, India), Potato Dextrose Agar (PDA) (HiMedia, Mumbai, India), Violet Red Bile Agar (VRBA) (Merck, Germany) was used in the microbiological test to check the growth of LAB, mold, fungus, and coliform bacteria counts. All the broth and agar powder were prepared by following the directions labeled on the bottle and autoclaved at 121°C, 15 psi for 15 min.
For plant-based medium preparation, different types of commercial fruits / vegetable powder (coconut, almond, barley, hawthorn, green papaya, pumpkin, bitter gourd, mung bean and yam) were purchased, which are 100% pure without any additives. A known amount (5.5 g) of each powder was added to 100 mL of distilled water. The Plant-based medium was prepared in 5.5% following the standard medium percentage as MRS medium. The mediums were boiled to 100°C for 5 min to homogenize them, filtrated, and autoclaved at 121°C, 15 psi for 15 min.
2.2 Preparation of LAB strains
Culture stocks of L. brevis C23, L. paracasei FD1, L. plantarum CAM4, and Lactococcus lactis GH1 were obtained from a laboratory in the School of Industrial Technology, Universiti Sains Malaysia, Pulau Pinang. All LAB strains were preserved in glycerol at -20°C. The LAB culture stocks were cultured in MRS broth at the ratio of 1: 9. 1 mL LAB culture was added to 9 mL of MRS broth. The bacteria culture was incubated at 37°C for 24 h in the incubator (Binder Redline Incubator, RI 115-U, America) and stored at 4°C.
2.3 Co-culture of L. brevis C23 with LAB inducers
The ability of L. brevis C23 to produce higher γ-aminobutyric acid (GABA) was investigated through co-culturing with three other lactic acid bacteria (LAB) strains, namely L. paracasei FD1, L. plantarum CAM4, and L. lactis GH1, which acts as inducers. These co-cultures were prepared with one, two, and three inducers. For the first set of co-cultures, 1 mL of L. brevis C23 culture was added to 18 mL of MRS broth, along with 1 mL of single inducer strain. For the second set of co-cultures, 1 mL of L. brevis C23 culture was added to 27 mL of MRS broth, along with 2 mL of the combination of two inducer strains. Finally, for the third set of co-cultures, 1 mL of L. brevis C23 culture was added to 36 mL of MRS broth, along with 3 mL of the combination of three inducer strains. The co-cultures were incubated at 37°C for 24 h and stored at 4°C for further analysis.
2.4 Cell-free supernatant (CFS) postbiotic preparation
LAB strains were cultured overnight and centrifuged at 10000 rpm, 4°C for 15 min via using Tabletop Refrigerated Centrifuge, 5500, Kubota, Japan. The cell-free supernatant (CFS) was syringe filtered by a sterile 0.22 µm pore size membrane (PES Sterile Syringe Filter, Jinteng, China) and stored at 4°C.
2.5 Antimicrobial assay
Broth dilution method was used for antimicrobial assay by following the technique reported by Kuhan Sreedharan et al. (2021). Pathogen Listeria monocytogenes was obtained from School of Industrial Technology, Universiti Sains Malaysia, Pulau Pinang. L. monocytogenes is a pathogenic strain that is commonly present in food. L. monocytogenes was preserved in glycerol at -20°C. 1 mL pathogenic strain culture stock was added to 9 mL of BHI broth, and incubated at 37°C for 24 h °C. The pathogen was diluted with BHI broth to 0.02 (108 CFU/mL) at the absorbance range of 600 nm by using visible microplate reader, HALO MPR-96, Dynamica. The diluted pathogen, CFS and BHI broth were mixed together at the ratio of 1: 1: 1 to a total volume of 3 mL. 200 µL of the mixture was pipetted into a 96-well microplate. The microplate was incubated at 37°C for 24 h, concurrently, the turbidity was measured at the absorbance of 600 nm. In this assay, the control used was MRS broth instead of CFS. The percentage of pathogen growth inhibition of CFS was calculated by the following equation:
$$Inhibition \left(\%\right)=\left( \frac{{OD}_{c}-{OD}_{s}}{{OD}_{c}} \right)\times 100$$
Where, ODC represents the difference in the optical density of the growth control, and ODS represents the difference in the optical density (OD) of the sample.
2.6 GABA determination HPLC
GABA determination and quantification was carried out via the method reported by Kanklai, Somwong, Rungsirivanich, et al., (2020). UV-vis detector UFLC (Shimadzu, LC-20AD, Japan), equipped with ODS C18 reverse-phase column (5 µm, 250 mm x 4.6 mm, Thermo Fisher Scientific Co., Waltham, MA, USA) was employed for this purpose. In short, CFS (1 mL) was added to 1 mL of a mixture of ethanol: water: triethylamine (2: 1: 1) and next added to 80 µL mixture of ethanol: water: triethylamine: phenyl isothiocyanate (7: 1: 1: 1). The mixtures were left for 20 min at a room temperature, then filtered through 0.22 µm pore size membrane. Mobile phase A consists of 1.4 mM sodium acetate, 0.1% triethylamine, 6% acetonitrile, and the pH were adjusted to 6.1. Mobile phase B consists of 60% acetonitrile. Both mobile phases were filtered through a 0.22 µm pore size membrane and degassed before used. 10 µL sample (of what) was injected into the HPLC system with a linear gradient system of buffer A (100-0%) for 70 min. The flow rate was 1 mL/min. The sample peak was detected at 254 nm with a UV detector.
2.7 Analysis of sugar composition using HPLC
Isocratic mode reverse-phase HPLC identification and quantification of fructose, glucose and sucrose were performed following the method (Saad et al. 2020).The sugars composition were separated using NH2 column (5 µm, 250 x 4.6 mm, Inertsil, Japan) connected to a refractive index HPLC system with refractive index detector (Shimadzu, RID-10A, Japan), an auto-sampler injector (Shimadzu, SIL-20A HT, Japan) and oven system (Shimadzu, CTO-10AS VP, Japan). 20 µL of CFS was loaded into the HPLC system and ran at 35°C with isocratic elution at the flow rate of 1.0 mL/min for 20 mins with 75% acetonitrile as mobile phase. The CFS and mobile phases were filtered with 0.45 µm membrane and degassed before run. Sugar compounds were identified and quantified by retention times. All samples were analyzed in duplicates. To make the standards of sugars, 10 mg of fructose, glucose and sucrose were dissolved in 1.5 mL of acetonitrile and water solution (2: 1), and then filtered through 0.45 µm polyethersulfone (PES) syringe filter. A series of working solutions were prepared by diluting stock solution 2-folds in ranges of 0.31 to 5.00 mg/mL.
2.7 Cell viability test
Cell viability test was done to determine the growth of LAB strains in the plant-based medium. Co-cultured LAB (1 mL) was added to 9 mL of each medium and incubated at 37°C for 24 h. A 10-fold dilution of LAB culture was diluted to a dilution factor of 10− 7 with autoclaved distilled water as diluent. First, 1 mL of LAB culture was added to 9 mL of diluent to make the total volume of 10 mL, which makes a dilution factor of 10− 1. The dilution was mixed by pipetting the pipette a few times. Next, 1 mL of the mixture with dilution factor of 10− 1 was added to 9 mL of diluent, this makes the dilution factor of 10− 2. The same procedure was repeated until the final dilution factor of 10− 7. After the serial dilution, 100 µL of each diluted culture were pipetted and spread on MRS agar. The agar plates were incubated at 37°C for 24 h and plate count was done.
2.8 Optimization of the fermentation medium composition for the production of BLIS and GABA by LAB using response surface methodology (RSM)
A preliminary study on the production of BLIS and GABA by LAB was conducted at a different range of tween 20 (0.5-3.0%), glucose (0.5-3.0%), L-glutamic acid (0.5-3.0%) and pyridoxine (0.05–0.3%). The selected plant-based medium was added with a different range of each supplement. Selected ranges obtained from the preliminary study were applied in a 44 full factorial central composite design (CCD) (Design Expert 11, Stat-Ease Inc., USA) to evaluate the interaction effects of the independent variables, along with optimizing the composition of fermentation medium. A set of 30 experiment runs was generated by CCD. There are 16 runs for factorial design, 8 runs for axial points, and 6 runs for replications for central points. BLIS antimicrobial activity and GABA concentration were represented as a response (Y) to be analyzed. The experimental data were fitted from the following polynomial Eq. (1):
$$Y=Intercept+A+B+C+D+AB+AC+AD+BC+BD+CD+{A}^{2}+{B}^{2}+{C}^{2}+{D}^{2}$$
where Y is the response (BLIS antimicrobial activity and GABA concentration); A = tween 20, B = glucose, C = L-Glutamic acid and D = pyridoxine was represented the coded values of the independent variables; Intercept represents the constant regression coefficient; A, B, C, D are the linear regression coefficients; AB, AC, AD, BC, BD, CD are the cross-product regression coefficients; A2, B2, C2, D2 are the quadratic regression coefficient.
2.9 Antioxidant assay
The antioxidant assay was conducted using DPPH radical scavenging assay. The scavenging effect of the CFS product cultured in optimized medium, CFS product cultured in MRS medium, nisin solution, and potassium sorbate solution on the free radical DPPH was measured using the method of Xing, Wang, Zhang, et al., (2015). Later, 1 mL of sample (what sample – either give a label or acronym) and 1 mL of freshly prepared DPPH solution (0.2 mM, Sigma Aldrich, St. Louis, MO) were mixed. The mixture was shaken vigorously and left to react for 30 min in the dark at room temperature. The control was performed using deionized water instead of CFS. The scavenged DPPH was then measured by using a microplate reader with an absorbance at 517 nm. The radical scavenging activity was quantified as units/mL (U/mL) by using the following formula:
$$DPPH activity (U/mL)=\left( \frac{{Abs}_{c}-{Abs}_{s}}{{Abs}_{c}} \right)\times 100$$
Where ABSC and ABSS represent the absorbance of control and samples, respectively.
2.10 Microbiological test
2.10.1 Preparation of food samples and treatments
Fresh chicken meat and strawberries were purchased and cut into small pieces. The chicken meat and strawberries were separated into 4 groups, 20 pieces for each. The samples were dipped in sterile water (control), 0.1% nisin solution, 0.1% potassium sorbate solution, and CFS product that were cultured in an optimized medium for 2 min, respectively. The treated chicken meat was stored in sterile plastic containers at 4°C for 12 days. The strawberries were air-dried and stored in sterile plastic containers at room temperature for 6 days.
2.10.2 Microbiological analyses of food samples
The microbial analyses for chicken meat were conducted at days 0, 4, 8, and 12 during storage at 4°C. While for strawberries were conducted at days 0, 2, 4, and 6 during storage at room temperature. A piece of sample was transferred into a sterile centrifuge tube, and then 10 mL of sterile water was added. After 2 min of homogenization, 1 mL aliquots from the mixture were 10-fold serially diluted. 100 µL of the dilutions was spread onto different selective agars including MRS agar for LAB counts, PDA for molds and fungus counts and VRBA for coliform bacteria counts. The agar plates were incubated at 37°C for 24 h and plate count was done.
2.11 Statistical analysis
The results obtained from Design Expert 11 software (Stat-Ease Inc., USA) were subjected to analysis of variance (ANOVA). The regression coefficients for GABA concentration and BLIS antimicrobial activity in a confidence level above 95% were considered significant (p < 0.05) due to large intrinsic variability in biological processes.
3.1 Effect of co-culturing L. brevis C23 with different LAB as inducers on the production of BLIS and GABA
Table 1 displays the results of BLIS antimicrobial activity and gamma-aminobutyric acid (GABA) concentration produced by the monoculture of L. brevis C23 and co-cultures of L. brevis C23 with LAB inducers. Co-culture of L. brevis C23 with inducers exhibited significantly higher antimicrobial activity against pathogens than the monoculture of L. brevis C23. Notably, the monoculture of L. brevis C23 possessed the lowest antimicrobial activity of 75.94 ± 0.73% among the co-cultures. Further, co-culture of L. brevis C23 with L. paracasei FD1 and L. lactis GH1 demonstrated the highest antimicrobial activity of 87.83 ± 0.89%. The presence of multiple microorganisms in the substrate can create a stressful environment that induces bacteria to overexpress bacteriocin-like antimicrobial proteins, as documented by Gutiérrez-Cortés et al. (2018). Co-cultures can have the effect of inducing the activity of a previously inactive yet bacteriocin-competent strain (Chanos and Mygind 2016). For instance, in the mixture (L. brevis, L. plantarum and Weissella cibaria) during the first 12 h of fermentation, the antimicrobial activity was higher (2.12–2.28 cm) than the antimicrobial activity of the monoculture L. brevis (1.66–2.23 cm) (Serna-Cock et al. 2019). Hence, co-cultivation of LAB was utilized as induction of BLIS of L. brevis C23 in this study.
|
LAB strains |
BLIS Antimicrobial activity, Mean ± Standard Deviation (%) |
GABA Concentration, Mean ± Standard Deviation (mg/mL) |
|---|---|---|
|
Lactobacillus brevis C23 |
75.94 ± 0.73 |
0.50 ± 0.02 |
|
Lactobacillus brevis C23 + Lactobacillus paracasei FD1 |
84.64 ± 0.65 |
0.67 ± 0.02 |
|
Lactobacillus brevis C23 + Lactobacillus plantarum CAM4 |
81.01 ± 0.73 |
0.70 ± 0.01 |
|
Lactobacillus brevis C23 + Lactococcus lactis GH1 |
77.49 ± 0.56 |
0.38 ± 0.01 |
|
Lactobacillus brevis C23 + Lactobacillus paracasei FD1 + Lactobacillus plantarum CAM4 |
79.31 ± 0.20 |
0.28 ± 0.02 |
|
Lactobacillus brevis C23 + Lactobacillus paracasei FD1 + Lactococcus lactis GH1 |
87.83 ± 0.89 |
0.67 ± 0.01 |
|
Lactobacillus brevis C23 + Lactobacillus plantarum CAM4 + Lactococcus lactis GH1 |
86.15 ± 0.28 |
0.67 ± 0.01 |
|
Lactobacillus brevis C23 + Lactobacillus paracasei FD1 + Lactobacillus plantarum CAM4 + Lactococcus lactis GH1 |
85.97 ± 0.42 |
0.70 ± 0.01 |
Regarding GABA production, some co-cultures of L. brevis C23 enhanced GABA production compared to monoculture of L. brevis C23, while other co-cultures showed a decrease in GABA production. Co-culture of L. brevis C23, L. paracasei FD1, and L. plantarum CAM4 yielded the lowest GABA concentration of 0.28 ± 0.02 mg/mL. In contrast, co-cultures of L. brevis C23, L. paracasei FD1, L. plantarum CAM4, and L. lactis GH1 produced the highest GABA concentration of 0.70 ± 0.01 mg/mL. The monoculture of L. brevis C23 exhibited GABA production within a moderate range of 0.50 ± 0.02 mg/mL. LAB co-cultures can increase lactate fermentation and decrease pH, thereby inducing GABA synthesis, according to (2020). However, some LAB co-cultures may deactivate the GABA-competent strain. In contrast, other LAB strains may prohibit different strains from expressing their functionalities when they are dominant in the co-culture (Canon et al. 2020). Therefore, co-culture of L. brevis C23 with L. paracasei FD1, L. plantarum CAM4, and L. lactis GH1 was selected due to their high antimicrobial activity and GABA production.
3.2 Effect of fermentation of co-cultured LAB in different types of plant-based medium on the production of BLIS and GABA by LAB
In this study, we investigated selected LAB's cell viability in various plant-based media types. Results presented in Table 2 showed that LAB cultured in mung bean, coconut, and MRS broth demonstrated higher cell viability compared to other mediums. Notably, the highest cell viability was observed in mung bean broth (10.90 ± 0.10 × 108 CFU/mL), followed by MRS broth (10.50 ± 0.40 × 108 CFU/mL) and coconut broth (9.10 ± 0.20 × 108 CFU/mL). Conversely, LAB cultured in almond broth exhibited the lowest cell viability (0.90 ± 0.02 × 108 CFU/mL). The superior performance of mung bean and coconut media can be attributed to their higher carbohydrate content than other plant-based media (Hossain et al. 2020).
|
Medium |
Cell Viability, Mean ± Standard Deviation (× 108 CFU/mL) |
BLIS Antimicrobial activity, Mean ± Standard Deviation (%) |
GABA Concentration, Mean ± Standard Deviation (mg/mL) |
|---|---|---|---|
|
Almond |
0.90 ± 0.02 |
23.30 ± 0.31 |
0.43 ± 0.02 |
|
Barley |
1.00 ± 0.04 |
43.72 ± 0.31 |
0.38 ± 0.02 |
|
mung bean |
10.90 ± 0.10 |
74.00 ± 0.22 |
0.33 ± 0.01 |
|
Green papaya |
2.68 ± 0.06 |
80.04 ± 0.20 |
0.28 ± 0.02 |
|
Coconut |
9.10 ± 0.20 |
81.21 ± 0.34 |
0.28 ± 0.01 |
|
Bitter gourd |
3.10 ± 0.30 |
69.32 ± 1.40 |
0.21 ± 0.01 |
|
pumpkin |
2.00 ± 0.08 |
81.13 ± 1.09 |
0.14 ± 0.01 |
|
Yam |
4.60 ± 0.20 |
77.69 ± 0.53 |
0.09 ± 0.01 |
|
Hawthorn |
1.50 ± 0.03 |
56.44 ± 0.59 |
0.05 ± 0.01 |
|
MRS |
10.50 ± 0.40 |
85.97 ± 0.42 |
0.70 ± 0.01 |
Furthermore, the ability of LAB cultured in plant-based media to produce bacteriocins with high antimicrobial activity was evaluated. Our results demonstrated that LAB cultured in mung bean and coconut media produced bacteriocins that were comparable in activity to those produced in MRS broth. LAB cultured in MRS broth displayed the highest antimicrobial activity, while LAB cultured in coconut medium exhibited the highest activity against pathogens, with an antimicrobial activity of 81.21 ± 0.34% among the nine types of plant-based media. Conversely, LAB cultured in barley medium exhibited the lowest antimicrobial activity (13.72 ± 0.31%). Previous research has shown that phytochemicals in vegetable media, especially phenolic compounds and flavonoids, can enhance antimicrobial activity (Hochma et al. 2021).
Additionally, our results showed that GABA produced by LAB cultured in these media was lower than in MRS broth. Notably, LAB cultured in almond broth exhibited the highest GABA concentration (0.43 ± 0.02 mg/mL), followed by LAB cultured in coconut medium, which produced a moderate amount of GABA concentration (0.28 ± 0.01 mg/mL). In contrast, LAB cultured in hawthorn broth produced the lowest amount of GABA (0.05 ± 0.01 mg/mL). Naumenko et al. (2022) have suggested that grains and seeds contain glutamate decarboxylase enzyme, which is essential in GABA synthesis. As such, GABA concentration is typically higher in grains and seed media. Magnesium is also known to enhance GABA production, and MRS medium, which contains higher levels of magnesium than fruit and vegetable-based media, demonstrated the highest GABA production.
Based on these findings, we selected coconut medium for further investigation due to its efficient LAB growth, high antimicrobial activity, and moderate GABA production levels compared to the other plant-based media. Moreover, compared to MRS medium, coconut medium had a clear colour, and better taste, reducing the cost for downstream processing decolorization and exerting minimal effects on the original colour of food when applied as a preservative. Additionally, coconut medium had no toxic effects on the human body. It represented a good alternative to chemically defined medium, at a cost approximately 30 times lower than MRS medium. The colour of the different plant-based media is shown in Fig. 1.
3.3 Optimization of the composition of supplements in the plant-based medium for the production of BLIS and GABA by LAB using response surface methodology (RSM)
The present study investigates the main and interaction effects of four supplements, namely tween 20, glucose, L-glutamic acid, and pyridoxine, on the production of BLIS and GABA by LAB cultured in coconut medium. The study utilized a central composite design (CCD) statistical model to evaluate the effect of different supplement concentrations on BLIS antimicrobial activity and GABA concentration. Prior to CCD, a preliminary study was conducted using the one-factor-at-a-time (OFAT) method.
The results included the BLIS’s antimicrobial activity and GABA’s concentration produced by LAB with varying supplement concentrations, are presented in Table 3. The findings indicate that 2% of tween 20, a surfactant, had the highest antimicrobial activity (85.17 ± 0.62%) and GABA concentration (0.60 ± 0.02 mg/mL). However, when the concentration of tween 20 increased to 3%, there was a decrease in antimicrobial activity (83.93 ± 0.93%) and GABA concentration (0.20 ± 0.06 mg/mL). The use of surfactants can increase bacteriocin concentration due to accelerated cell growth. Furthermore, surfactants may alter the surface tension of the producer cell, facilitating bacteriocin discharge from the cell surface (Abbasiliasi et al. 2017). The study found that 2% of L-glutamic acid resulted in the highest GABA concentration (5.77 ± 0.17mg/mL) but also possessed the lowest antimicrobial activity (33.70 ± 0.52%). The highest antimicrobial activity of 80.29 ± 1.00% was achieved when the percentage of L-glutamic acid was 0.5%. Cui et al. (2020) stated that glutamate acts as the precursor for GABA, so the increasing L-glutamic acid stimulates the production of GABA by glutamate decarboxylase (GAD). Similarly, 2% of glucose resulted in the highest antimicrobial activity (86.53 ± 0.06%) and GABA concentration (0.61 ± 0.02 mg/mL). When the glucose percentage increased to 3%, the antimicrobial activity decreased to 85.36 ± 1.20%, and the GABA concentration decreased to 0.08 ± 0.01 mg/mL. Studies have shown that high bacteriocin yield was associated with glucose in growth conditions. Finally, 1% of pyridoxine resulted in the highest antimicrobial activity (82.59 ± 0.74%) and GABA concentration (0.65 ± 0.02 mg/mL). When there was 2% of pyridoxine in the medium, the LAB achieved the highest antimicrobial activity of 84.12 ± 0.19%, but the GABA concentration decreased to 0.4718 ± 0.02 mg/mL. According to Lim et al. (2022), GAD uses pyridoxal phosphate (PLP) as a coenzyme. As PLP is the biologically active form of pyridoxine, adding pyridoxine to the medium increases GAD activity and GABA synthesis.
|
Supplement |
BLIS Antimicrobial activity, |
GABA Concentration, |
|---|---|---|
|
(%) |
(mg/ml) |
|
|
Tween 20 |
||
|
0.00% |
78.34 ± 0.73 |
0.19 ± 0.01 |
|
0.50% |
81.96 ± 0.88 |
0.59 ± 0.03 |
|
1.00% |
83.22 ± 0.73 |
0.59 ± 0.39 |
|
2.00% |
85.17 ± 0.62 |
0.60 ± 0.02 |
|
3.00% |
83.93 ± 0.93 |
0.20 ± 0.06 |
|
Glucose |
||
|
0.00% |
78.34 ± 0.73 |
0.19 ± 0.01 |
|
0.50% |
80.71 ± 1.11 |
0.36 ± 0.01 |
|
1.00% |
83.77 ± 0.98 |
0.47 ± 0.03 |
|
2.00% |
86.53 ± 0.96 |
0.61 ± 0.02 |
|
3.00% |
85.36 ± 1.20 |
0.08 ± 0.01 |
|
L-Glutamic Acid |
||
|
0.00% |
78.34 ± 0.73 |
0.19 ± 0.01 |
|
0.50% |
80.29 ± 1.00 |
3.87 ± 0.14 |
|
1.00% |
51.04 ± 0.31 |
4.42 ± 0.18 |
|
2.00% |
33.70 ± 0.52 |
5.77 ± 0.17 |
|
3.00% |
50.07 ± 0.34 |
1.65 ± 0.05 |
|
Pyridoxine |
||
|
0.00% |
78.34 ± 0.73 |
0.19 ± 0.01 |
|
0.50% |
80.15 ± 0.46 |
0.60 ± 0.36 |
|
1.00% |
82.59 ± 0.74 |
0.65 ± 0.02 |
|
2.00% |
84.12 ± 0.19 |
0.47 ± 0.02 |
|
3.00% |
77.67 ± 0.80 |
0.03 ± 0.01 |
The experimental and predicted data of the BLIS antimicrobial activity and GABA concentration conditions designed by CCD were shown in Table 4. The highest BLIS antimicrobial activity of 83.51% was achieved in run 12 with 3% tween 20, 3% glucose, 1% L-glutamic acid and 0.2% of pyridoxine. Run 23 had the lowest BLIS antimicrobial activity of 60.2% with 2% tween 20, 2% glucose, 2% L-glutamic acid and 0% of pyridoxine. The highest GABA concentration produced by LAB of 7.21 mg/mL was achieved in run 14 with 3% tween 20, 1% glucose, 3% L-glutamic acid, and 0.2% of pyridoxine. Run 21 had the lowest BLIS antimicrobial activity of 0.47 mg/mL with 2% tween 20, 2% glucose, 0% L-glutamic acid, and 0125% pyridoxine.
|
Std |
Independent variables |
Response values |
||||||
|---|---|---|---|---|---|---|---|---|
|
Tween 20 (%) |
Glucose (%) |
L-glutamic acid (%) |
Pyridoxine (%) |
BLIS antimicrobial activity (%) |
GABA concentration (mg/mL) |
|||
|
Actual |
Predicted |
Actual |
Predicted |
|||||
|
1 |
1 |
1 |
1 |
0.05 |
78.67 |
79.20 |
5.47 |
5.75 |
|
2 |
3 |
1 |
1 |
0.05 |
80.98 |
79.88 |
5.58 |
5.31 |
|
3 |
1 |
3 |
1 |
0.05 |
76.98 |
76.00 |
4.48 |
4.24 |
|
4 |
3 |
3 |
1 |
0.05 |
81.12 |
81.28 |
4.41 |
4.41 |
|
5 |
1 |
1 |
3 |
0.05 |
65.68 |
65.21 |
6.85 |
6.85 |
|
6 |
3 |
1 |
3 |
0.05 |
66.18 |
67.86 |
6.52 |
6.52 |
|
7 |
1 |
3 |
3 |
0.05 |
62.32 |
62.82 |
4.28 |
4.24 |
|
8 |
3 |
3 |
3 |
0.05 |
71.79 |
70.07 |
4.31 |
4.52 |
|
9 |
1 |
1 |
1 |
0.2 |
81.75 |
81.67 |
3.51 |
3.22 |
|
10 |
3 |
1 |
1 |
0.2 |
80.49 |
80.39 |
3.92 |
3.96 |
|
11 |
1 |
3 |
1 |
0.2 |
80.14 |
78.86 |
4.52 |
4.30 |
|
12 |
3 |
3 |
1 |
0.2 |
83.51 |
82.18 |
2.52 |
2.44 |
|
13 |
1 |
1 |
3 |
0.2 |
65.26 |
65.51 |
4.96 |
4.83 |
|
14 |
3 |
1 |
3 |
0.2 |
67.02 |
66.20 |
7.21 |
7.37 |
|
15 |
1 |
3 |
3 |
0.2 |
64.21 |
63.51 |
3.10 |
3.59 |
|
16 |
3 |
3 |
3 |
0.2 |
68.91 |
68.79 |
5.03 |
4.75 |
|
17 |
0 |
2 |
2 |
0.125 |
67.09 |
67.50 |
2.98 |
2.99 |
|
18 |
4 |
2 |
2 |
0.125 |
72.49 |
73.47 |
3.93 |
4.00 |
|
19 |
2 |
0 |
2 |
0.125 |
73.75 |
73.10 |
2.70 |
2.84 |
|
20 |
2 |
4 |
2 |
0.125 |
70.46 |
72.50 |
3.54 |
3.62 |
|
21 |
2 |
2 |
0 |
0.125 |
82.32 |
83.71 |
0.47 |
0.71 |
|
22 |
2 |
2 |
4 |
0.125 |
69.75 |
69.46 |
4.29 |
4.12 |
|
23 |
2 |
2 |
2 |
0 |
60.21 |
61.02 |
5.08 |
5.22 |
|
24 |
2 |
2 |
2 |
0.275 |
78.25 |
79.64 |
5.81 |
5.89 |
|
25 |
2 |
2 |
2 |
0.125 |
71.37 |
72.55 |
3.68 |
3.59 |
|
26 |
2 |
2 |
2 |
0.125 |
73.38 |
72.55 |
3.58 |
3.59 |
|
27 |
2 |
2 |
2 |
0.125 |
72.39 |
72.55 |
3.48 |
3.59 |
|
28 |
2 |
2 |
2 |
0.125 |
72.40 |
72.55 |
3.68 |
3.59 |
|
29 |
2 |
2 |
2 |
0.125 |
74.36 |
72.55 |
3.76 |
3.59 |
|
30 |
2 |
2 |
2 |
0.125 |
71.41 |
72.55 |
3.38 |
3.59 |
BLIS Antimicrobial Activity results were analyzed by analysis of variance (ANOVA) using CCD and presented in Table 5. The model F-value is 71.98, which suggests that the model is significant. There is only a 0.0001% chance that model F-value could occur due to human or environmental error during the experiment. The “lack of fit F-value” is 0.2048, indicating that the model fits due to a lack of fit is not significant compared to the pure error. The p-value less than 0.05 indicates model terms A, C, AB, D² are significant. Below is the polynomial model Eq. (2):
|
Analysis |
Source |
Sum of squares |
df |
Mean square |
F-value |
P-value (Prob > F) |
|
|---|---|---|---|---|---|---|---|
|
BLIS Activity |
Model |
1007.8 |
14 |
71.98 |
31.51 |
< 0.0001 |
significant |
|
A |
53.37 |
1 |
53.37 |
23.36 |
0.0003 |
||
|
B |
0.549 |
1 |
0.549 |
0.2403 |
0.6321 |
||
|
C |
793.25 |
1 |
793.25 |
347.23 |
< 0.0001 |
||
|
D |
1.51 |
1 |
1.51 |
0.6631 |
0.4301 |
||
|
AB |
21.09 |
1 |
21.09 |
9.23 |
0.0095 |
||
|
AC |
3.87 |
1 |
3.87 |
1.69 |
0.2156 |
||
|
AD |
3.85 |
1 |
3.85 |
1.69 |
0.2167 |
||
|
BC |
0.6521 |
1 |
0.6521 |
0.2854 |
0.6022 |
||
|
BD |
0.1502 |
1 |
0.1502 |
0.0657 |
0.8017 |
||
|
CD |
4.74 |
1 |
4.74 |
2.08 |
0.1733 |
||
|
A² |
6.98 |
1 |
6.98 |
3.06 |
0.104 |
||
|
B² |
0.102 |
1 |
0.102 |
0.0447 |
0.8359 |
||
|
C² |
5.96 |
1 |
5.96 |
2.61 |
0.1303 |
||
|
D² |
39.42 |
1 |
39.42 |
17.26 |
0.0011 |
||
|
Residual |
29.7 |
13 |
2.28 |
||||
|
Lack of fit |
22.99 |
8 |
2.87 |
2.14 |
0.2084 |
not significant |
|
|
Pure Error |
6.71 |
5 |
1.34 |
||||
|
Cor Total |
1037.5 |
27 |
|||||
|
R2 |
0.9714 |
||||||
|
Adjusted R2 |
0.9405 |
||||||
|
Predicted R2 |
0.7843 |
||||||
|
Adeq precision |
18.882 |
||||||
|
GABA Concentration |
Model |
50.47 |
14 |
3.61 |
57.33 |
< 0.0001 |
significant |
|
A |
1.31 |
1 |
1.31 |
20.85 |
0.0008 |
||
|
B |
12.75 |
1 |
12.75 |
202.76 |
< 0.0001 |
||
|
C |
13.62 |
1 |
13.62 |
216.64 |
< 0.0001 |
||
|
D |
4.65 |
1 |
4.65 |
73.88 |
< 0.0001 |
||
|
AB |
0.2585 |
1 |
0.2585 |
4.11 |
0.0676 |
||
|
AC |
0.0089 |
1 |
0.0089 |
0.1419 |
0.7135 |
||
|
AD |
1.1 |
1 |
1.1 |
17.42 |
0.0016 |
||
|
BC |
0.9598 |
1 |
0.9598 |
15.26 |
0.0024 |
||
|
BD |
0.2696 |
1 |
0.2696 |
4.29 |
0.0627 |
||
|
CD |
3.4 |
1 |
3.4 |
54.1 |
< 0.0001 |
||
|
A² |
0.0153 |
1 |
0.0153 |
0.2431 |
0.6317 |
||
|
B² |
3.71 |
1 |
3.71 |
59.05 |
< 0.0001 |
||
|
C² |
2.22 |
1 |
2.22 |
35.23 |
< 0.0001 |
||
|
D² |
10.97 |
1 |
10.97 |
174.48 |
< 0.0001 |
||
|
Residual |
0.6918 |
11 |
0.0629 |
||||
|
Lack of fit |
0.5902 |
6 |
0.0984 |
4.84 |
0.0522 |
not significant |
|
|
Pure Error |
0.1016 |
5 |
0.0203 |
||||
|
Cor Total |
51.17 |
25 |
|||||
|
R2 |
0.9865 |
||||||
|
Adjusted R2 |
0.9693 |
||||||
|
Predicted R2 |
0.8809 |
||||||
|
Adeq precision |
34.938 |
From the fit statistic result, response of BLIS antimicrobial activity obtained a determination coefficient (R2) of 0.9714, indicating that the model well interprets 97.14% of the total deviation of experimental results. The predicted R² of 0.7843 is in reasonable agreement with the adjusted R² of 0.9405, since the difference between the two is less than 0.2. As a result, the fit statistic indicates that Eq. (2) is adequately used as a predictive model for this experiment.
The interaction terms of A, C, AB, and D² are significant (p < 0.05). The response surfaces graph in Fig. 2A illustrates the interaction between A (Tween 20) and B (Glucose) relative to BLIS antimicrobial activity by keeping the other component at their central level. The interaction terms of AC, AD, BC, BD, and CD were non-significant (p > 0.05).
GABA concentration results were analyzed by ANOVA using CCD and also presented in Table 5. The model F-value is 57.33 suggests that the model is significant. There is only a 0.0001% chance that model F-value could occur due to human or environmental error during the experiment. The “lack of fit F-value” is 4.84, it indicates that the model fits due to a lack of fit is not significant compared to the pure error. The p-value less than 0.05 indicates model terms A, B, C, D, AD, BC, CD, B², C², and D² are significant. Below is the polynomial model Eq. (3):
From the fit statistic result, the response of GABA concentration obtained a determination coefficient (R2) of 0.9865, indicating that the model well interprets 98.65% of the total deviation of experimental results. The predicted R² of 0.8809 is in reasonable agreement with the adjusted R² of 0.9693, since the difference between the two is less than 0.2. As a result, the fit statistic indicates that Eq. (3) is adequately used as a predictive model for this experiment.
The interaction terms of A, B, C, D, AD, BC, CD, B², C², and D² are significant (p < 0.05). The response surfaces graph in Fig. 2B, 2C and 2D illustrating the interaction between A (Tween 20) and D (Pyridoxine), B (Glucose) and C (L-glutamic acid) and C (L-glutamic acid), and D (Pyridoxine) relative to GABA concentration by keeping the other component at their central level. The interaction terms of AB, AC, and BD were non-significant (p > 0.05).
3.4 Validation of optimal conditions for GABA and BLIS production
The present study reports on the results of validation tests, as presented in Table 6. There were 4 different conditions of the constraints in the tests designed to examine the efficacy of the RSM. First, simultaneously maximizing both constraints of GABA concentration and BLIS antimicrobial activity. Second, maximizing constraint of GABA concentration, while remaining within the range of the constraint of BLIS antimicrobial activity. Third, remaining within the range of the constraint of GABA concentration while maximizing constraint of BLIS antimicrobial activity. Fourth, maximizing constraint of GABA concentration while none for constraint of BLIS antimicrobial activity.
|
Constraints of GABA concentration |
Constraints of Antimicrobial Activity |
Tween 20 (%) |
Glucose (%) |
L-glutamic acid (%) |
Pyridoxine (%) |
Predicted |
Experimental |
Difference percentage, % |
Predicted |
Experimental |
Difference percentage, % |
|||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
(mean ± Standard Deviation) |
(mean ± Standard Deviation) |
|||||||||||||||||||||||
|
antimicrobial activity (%) |
antimicrobial activity (%) |
antimicrobial activity (%) |
GABA concentration (mg/mL) |
GABA concentration (mg/mL) |
GABA concentration (mg/mL) |
|||||||||||||||||||
|
Maximize |
Maximize |
1.44 |
0.23 |
0.48 |
0.02 |
84.5 |
84.40 ± 0.44 |
0.1 |
7.82 |
3.22 ± 0.01 |
58.9 |
|||||||||||||
|
Maximize |
In range |
0.48 |
0.7 |
1.3 |
0.02 |
78.2 |
77.04 ± 0.37 |
1.51 |
8.09 |
4.16 ± 0.02 |
48.6 |
|||||||||||||
|
In range |
Maximize |
1.69 |
1.03 |
0.21 |
0.06 |
83.5 |
85.32 ± 0.56 |
2.11 |
4.13 |
2.15 ± 0.02 |
48 |
|||||||||||||
|
Maximize |
None |
3 |
1 |
3 |
0.2 |
66.2 |
67.02 ± 0.36 |
1.26 |
7.37 |
7.21 ± 0.03 |
2.16 |
|||||||||||||
The experimental values of BLIS antimicrobial activity and GABA concentration were compared to their respective predicted values, with close similarity observed in the former case, validating the use of the central composite design (CCD) model for predicting these values. However, the experimental and predicted values of GABA concentration did not exhibit similar results, indicating the CCD model's limitations in precisely predicting this response. Nonetheless, when the constraint of BLIS antimicrobial activity is set to none, and the constraint of GABA concentration is maximized, the experimental and predicted values of GABA concentration exhibit close similarity (< 2.16% differences). Therefore, the CCD model cannot precisely predict BLIS antimicrobial activity and GABA concentration values simultaneously. The sugar composition of optimized coconut-based medium was determined and found that it contains 13.05 ± 0.23 g/L glucose, 6.29 ± 0.04 g/L fructose and 9.40 ± 0.13 g/L sucrose as shown in Online Resource 1. Following co-cultured LAB fermentation, the sugars in the medium were consumed, resulting in 3.86 ± 0.54 g/L glucose, 0.77 ± 0.06 g/L fructose and 6.00 ± 0.92 g/L sucrose.
3.5 Antioxidant capacity
The result of the antioxidant assay was shown in Table 7. Antioxidant capacity of the CFS postbiotic cultured in optimized medium has the highest antioxidant activity followed by the CFS of L. brevis C23 cultured in MRS medium. Nisin, potassium sorbate, and sterile water show no antioxidant activity. The CFS product cultured in an optimized medium produced a higher concentration of GABA than the CFS of L. brevis C23 cultured in MRS medium; it has higher antioxidant activity. The antioxidant capacity of CFS product cultured in optimized medium increased by about 2.42 times compared to the CFS of L. brevis C23 cultured in MRS medium in terms of DPPH radical scavenging activity.
|
Preservatives |
Antioxidant capacity (%) |
Food samples |
Storage Time (Days) |
Yeast and molds count (CFU/mL) |
Lactic acid bacteria count (CFU/mL) |
Coliform bacteria count (CFU/mL) |
|||
|
CFS postbiotic that cultured in optimized plant-based medium |
43.55 ± 0.01 |
Chicken meat |
0 |
0 |
4.80 ×101 |
0 |
|||
|
4 |
1.90 ×101 |
7.90 ×101 |
1 |
||||||
|
8 |
1.40 ×102 |
2.00 ×102 |
2 |
||||||
|
12 |
1.66 ×104 |
2.32 ×102 |
9 |
||||||
|
Strawberries |
0 |
0 |
0 |
0 |
|||||
|
2 |
0 |
0 |
0 |
||||||
|
4 |
1.10 ×101 |
1.37 ×102 |
2 |
||||||
|
6 |
3.00 ×103 |
2.94 ×102 |
3 |
||||||
|
CFS of L. brevis C23 that cultured in MRS medium |
17.99 ± 0.01 |
Chicken meat |
0 |
0 |
2.6 ×101 |
0 |
|||
|
4 |
2.5 ×101 |
6.2 ×101 |
6 |
||||||
|
8 |
1.93 ×102 |
1.52 ×102 |
7 |
||||||
|
12 |
1.78 ×103 |
2.33 ×102 |
4.6 ×101 |
||||||
|
Strawberries |
0 |
0 |
0 |
0 |
|||||
|
2 |
3 |
4 |
0 |
||||||
|
4 |
2.26 ×102 |
1.94 ×102 |
5 |
||||||
|
6 |
9.70 ×104 |
5.4 ×102 |
7 |
||||||
|
Potassium sorbate |
0% |
Chicken meat |
0 |
2.33 ×102 |
1.16 ×102 |
2.00 ×104 |
|||
|
4 |
4.60 ×105 |
5.00 ×104 |
9.30 ×104 |
||||||
|
8 |
9.00 ×105 |
2.40 ×105 |
9.00 ×107 |
||||||
|
12 |
7.70 ×106 |
6.50 ×106 |
1.05 ×109 |
||||||
|
Strawberries |
0 |
5.40 ×101 |
3 |
2.00 ×104 |
|||||
|
2 |
3.10 ×105 |
8 |
1.03 ×106 |
||||||
|
4 |
2.13 ×106 |
3.00 ×103 |
2.31 ×107 |
||||||
|
6 |
1.39 ×107 |
2.10 ×104 |
1.28 ×108 |
||||||
|
Nisin |
0% |
Chicken meat |
0 |
3.22 ×102 |
8.30 ×101 |
1.00 ×104 |
|||
|
4 |
1.80 ×105 |
1.40 ×104 |
6.20 ×104 |
||||||
|
8 |
1.50 ×106 |
2.50 ×105 |
8.10 ×107 |
||||||
|
12 |
1.89 ×107 |
7.60 ×106 |
6.80 ×108 |
||||||
|
Strawberries |
0 |
1.60 ×102 |
1 |
1.40 ×104 |
|||||
|
2 |
7.60 ×105 |
9 |
2.16 ×106 |
||||||
|
4 |
1.24 ×107 |
3.50 ×103 |
1.65 ×107 |
||||||
|
6 |
3.20 ×107 |
5.00 ×103 |
7.50 ×107 |
||||||
|
Sterile water (Control) |
0% |
Chicken meat |
0 |
1.00 ×104 |
1.80 ×103 |
2.00 ×104 |
|||
|
4 |
1.60 ×106 |
2.50 ×105 |
1.01 ×105 |
||||||
|
8 |
2.88 ×107 |
7.70 ×105 |
9.4 ×107 |
||||||
|
12 |
2.45 ×108 |
1.26 ×107 |
1.36 ×109 |
||||||
|
Strawberries |
0 |
2.00 ×105 |
6 |
1.50 ×104 |
|||||
|
2 |
2.91 ×106 |
2.35 ×102 |
3.70 ×106 |
||||||
|
4 |
3.50 ×107 |
2.18 ×104 |
1.02 ×108 |
||||||
|
6 |
2.29 ×109 |
7.00 ×104 |
1.67 ×108 |
||||||
3.6 Microbiological test
The findings of this study demonstrate that chicken meat and strawberries treated with the CFS postbiotic cultured in an optimized medium exhibit lower bacterial growth than other preservatives (Table 7). The efficacy of the CFS postbiotic in controlling the growth of LAB, mold, fungus, and coliform bacteria in food samples was found to be superior to that of nisin and potassium sorbate. The results indicate that after 12 days of storage, the yeast and mold count was 1.66 ×104 CFU/mL, lactic acid bacteria count was 2.32 ×102 CFU/mL, and the coliform bacteria count was 9 CFU/mL in chicken meat treated with the CFS postbiotic. Similarly, the CFS postbiotic was observed to inhibit foodborne pathogens in strawberries. The surface of minimally processed fruits and vegetables is the most contaminated part with foodborne pathogens (Mendoza et al. 2022). The CFS postbiotic, containing bacteriocins and other antimicrobial substances or stressors, has been shown to enhance its antimicrobial efficacy against spoiling and pathogenic microorganisms.
3.7 CONCLUSIONS
L. brevis C23 co-culture with inducers can induce higher GABA concentrations and BLIS activity than L. brevis C23 monoculture. The optimised coconut growth medium could produce GABA and BLIS by LAB co-culture and is approximately 30 times less expensive than MRS medium. When L-glutamic acid is added to the growth medium, GABA production increases. Fermentation of LAB at optimum conditions with 1.435% tween 20, 0.23% glucose, 0.479% L-glutamic acid, and 0.021% pyridoxine showed 5 times increase in GABA concentration compared to LAB that cultured in MRS medium. The CFS product can control the growth of LAB, molds, fungus, and coliform bacteria on food samples, making it a potential food preservative.
6.1 Funding
This study is financially supported by the Ministry of Higher Education Malaysia for Fundamental Research Grant Scheme with Project Code: FRGS/1/2022/STG01/USM/02/21.
6.2 Conflict of interest
The authors declare no competing interests.
6.3 Authors contributions
W.W.C. contributed in investigation, validation and writing original draft of experiments, J.S.T contributed in conceptualization, supervision and reviewing draft, S.N.H.O. contributed in reviewing draft, P.B.R. contributed in reviewing draft.
6.4 Data Availability
All data generated or analysed during this study are included in this published article and its supplementary information file.
- Abbasiliasi S, Tan JS, Ibrahim TAT, Bashokouh F, Ramakrishnan NR, Mustafa S & Ariff AB (2017) Fermentation factors influencing the production of bacteriocins by lactic acid bacteria: A review. Rsc Advances. 7(47), 29395–29420.
- Canon F, Mariadassou M, Maillard M-B, Falentin H, Parayre S, Madec M-N, Valence F, Henry G, Laroute V & Daveran-Mingot M-L (2020) Function-driven design of lactic acid bacteria co-cultures to produce new fermented food associating milk and lupin. Frontiers in Microbiology. 11, 584163.
- Chanos P & Mygind T (2016) Co-culture-inducible bacteriocin production in lactic acid bacteria. Applied microbiology and biotechnology. 100(10), 4297–4308.
- Chatterjee SS & Mukherjee A (2020) Impact of COVID-19 pandemic on pre-existing mental health problems. Asian journal of psychiatry. 51, 102071.
- Cui Y, Miao K, Niyaphorn S & Qu X (2020) Production of gamma-aminobutyric acid from lactic acid bacteria: A systematic review. International Journal of Molecular Sciences. 21(3), 995.
- Diez-Gutiérrez L, San Vicente L, Sáenz J, Barron LJR & Chávarri M (2022) Characterisation of the probiotic potential of Lactiplantibacillus plantarum K16 and its ability to produce the postbiotic metabolite γ-aminobutyric acid. Journal of Functional Foods. 97, 105230.
- Gutiérrez-Cortés C, Suarez H, Buitrago G, Nero LA & Todorov SD (2018) Enhanced bacteriocin production by Pediococcus pentosaceus 147 in co-culture with Lactobacillus plantarum LE27 on cheese whey broth. Frontiers in Microbiology. 9, 2952.
- Hochma E, Yarmolinsky L, Khalfin B, Nisnevitch M, Ben-Shabat S & Nakonechny F (2021) Antimicrobial effect of phytochemicals from edible plants. Processes. 9(11), 2089.
- Hossain MN, Akter A, Humayan S, Mohanto LC, Begum S & Ahmed MM (2020) Edible growth medium: a new window for probiotic research. Advances in Microbiology. 10(2), 39–51.
- Jin YH, Hong JH, Lee J-H, Yoon H, Pawluk AM, Yun SJ & Mah J-H (2021) Lactic acid fermented green tea with Levilactobacillus brevis capable of producing γ-aminobutyric acid. Fermentation. 7(3), 110.
- Kuhan Sreedharan D, Abbasiliasi S, Mohamed MS, Ng ZJ, Ariff AB, Lee CK & Tan JS (2021) Fermentation strategies for improving the production of bacteriocin-like inhibitory substances by Lactobacillus brevis C23 with nutrient supplementation, pH, and temperature variations. Journal of Food Processing and Preservation. 45(11), e15914.
- Liang J, Guo F, Cao S, Zhao K, Zhao K, Wang H, Shao X, Wei Y, Zhang C & Zheng Y (2022) γ-Aminobutyric acid (GABA) alleviated oxidative damage and programmed cell death in fresh-cut pumpkins. Plant Physiology and Biochemistry. 180, 9–16.
- Lim HJ, Jung D-H, Cho E-S & Seo M-J (2022) Expression, purification, and characterization of glutamate decarboxylase from human gut-originated Lactococcus garvieae MJF010. World Journal of Microbiology and Biotechnology. 38(4), 69.
- Liu J, Meng F, Du Y, Nelson E, Zhao G, Zhu H, Caiyin Q, Zhang Z & Qiao J (2020) Co-production of nisin and γ-aminobutyric acid by engineered Lactococcus lactis for potential application in food preservation. Frontiers in Microbiology. 11, 49.
- Mendoza IC, Luna EO, Pozo MD, Vásquez MV, Montoya DC, Moran GC, Romero LG, Yépez X, Salazar R & Romero-Peña M (2022) Conventional and non-conventional disinfection methods to prevent microbial contamination in minimally processed fruits and vegetables. LWT. 165, 113714.
- Naumenko N, Potoroko I & Kalinina I (2022) Stimulation of antioxidant activity and γ-aminobutyric acid synthesis in germinated wheat grain Triticum aestivum L. by ultrasound: Increasing the nutritional value of the product. Ultrasonics Sonochemistry. 86, 106000.
- Ng ZJ, Zarin MA, Lee CK, Phapugrangkul P & Tan JS (2020) Isolation and characterization of Enterococcus faecium DSM 20477 with ability to secrete antimicrobial substance for the inhibition of oral pathogen Streptococcus mutans UKMCC 1019. Archives of oral biology. 110, 104617.
- Pires SM & Devleesschauwer B (2021) Estimates of global disease burden associated with foodborne pathogens. In: Foodborne Infections and Intoxications. p^pp 3–17. Elsevier.
- Saad WMM, Salin NSM, Ramzi AS & Salim F (2020) Identification and quantification of fructose, glucose and sucrose in watermelon peel juice. Malays J Anal Sci. 24, 382–389.
- Serna-Cock L, Rojas-Dorado M, Ordoñez-Artunduaga D, García-Salazar A, García-González E & Aguilar CN (2019) Crude extracts of metabolites from co-cultures of lactic acid bacteria are highly antagonists of Listeria monocytogenes. Heliyon. 5(9), e02448.
- Yunes R, Poluektova E, Vasileva E, Odorskaya M, Marsova M, Kovalev G & Danilenko V (2020) A multi-strain potential probiotic formulation of GABA-producing Lactobacillus plantarum 90sk and Bifidobacterium adolescentis 150 with antidepressant effects. Probiotics and antimicrobial proteins. 12, 973–979.
No competing interests reported.