Screening of Lactobacillus rhamnosus-producing gamma aminobutyric acid (GABA) isolated from Sumbawa mare milk and its potential application to increase GABA content in fermented milk

Nursini, N.W., Antara, N.S., Sugitha, I.M. and Sujaya, I.N. Faculty of Health, Science and Technology Dhyana Pura University, Kuta Utara, Dalung, Bali 80361, Indonesia Doctoral Program of Agricultural Sciences, Faculty of Agriculture, Udayana University, JL. PB Sudirman, Badung Bali 80232, Indonesia Department of Food Science and Technology, Faculty of Agricultural Technology, Udayana University, Bukit Jimbaran Campus, Badung Bali 80362, Indonesia School of Public Health, Faculty of Medicine, Udayana University, JL. PB Sudirman, Badung Bali 80232, Indonesia Integrated Laboratory for Bioscience and Biotechnology, Udayana University, Bukit Jimbaran Campus, Badung Bali 80362, Indonesia


Introduction
Gamma-aminobutyric acid (GABA) is a non-protein amino acid produced through the irreversible αdecarboxylation reaction of the amino acid L-glutamic acid catalysed by the glutamic acid decarboxylase enzyme Zhuang et al., 2018). GABA has the general structure of H 2 N (CH 2 ) -CO 2 H (Anju et al., 2014). GABA is a neurotransmitter, a major inhibitor of the central nervous system, and plays important role in regulating physiological processes such as blood pressure and heart rate, increasing plasma concentrations, growth hormone, and protein synthesis in the brain (Villegas et al., 2016). Consumption of foods enriched with GABA can inhibit the proliferation of cancer cells (Park and Oh, 2007), and improve memory and learning abilities (Miura et al., 2006). On the other hand, GABA deficiency can cause several diseases such as Huntington's, Parkinson's, Alzheimer's, schizophrenia, and depression (Diana et al., 2014).
The functional effects of GABA on various aspects of health have led to increasingly interesting studies on how to produce GABA and increase GABA content in food. Due to the difficulty in isolating GABA from natural resources, hampering its content in foodstuffs, eISSN: 2550eISSN: -2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER recent studies are more focused on how to produce GABA using the activity of microorganisms (Dhakal et al., 2012) than isolating it from natural resources. Various types of microorganisms can produce GABA including bacteria, moulds, and yeasts. However, lactic acid bacteria (LAB) are of great interest in GABA production, because LAB plays an important role in the fermentation process of foods, which can improve aroma, taste, and its preservation. In addition, LAB has a long history of being used in traditional fermented food production so that it is considered generally recognized as safe (GRAS) for human consumption.
The ability of LAB for GABA synthesis and the fermentation parameters were reported to vary in several specific studies. Recent studies have reported that LAB can synthesize GABA in high concentrations in Mexican artisanal cheese (Gonzalez-Gonzalez et al., 2019). Lactobacillus rhamnosus YS9 has monosodium glutamate (MSG) concentration limit of 200 mM, an incubation temperature of 43 o C, 200 µM pyridoxal 5'phosphat (PLP), and a medium pH of 4.4 which can produce 1.87 mM of GABA (Lin, 2013). This is different from Lactobacillus rhamnosus GG which is used in the fermentation of adzuki beans and uses 2.27% MSG, 1.44% galactose, 0.20% PLP at 37 o C incubation temperature at a medium pH of 4.4 which can produce 1.12mg/mL GABA (Song and Yu, 2018). Lactobacillus rhamnosus has been isolated and identified from Sumbawa mare milk (Shi et al., 2012;Sujaya et al., 2008) and was shown to have the potential as a starter for fermented milk, with the main functional effect of improved lipid profiles (Nocianitri et al., 2017), the potential for this strain to produce GABA has never been elucidated. Considering the importance of GABA in various applications, it is important to screen GABA producing L. rhamnosus strain isolated from Sumbawa mare milk to improve GABA content in fermented milk. Therefore, this study aimed to identify L. rhamnosus from Sumbawa mare milk which is best capable to produce GABA when applied as a starter for fermented milk.

Bacterial strain and culture conditions
The L. rhamnosus strain used in this study was from the Unud Culture Collection stored in 30% glycerol at -80 o C ( Table 1). The strain was grown by transferring a loopful of stock strain and inoculated into 5 mL of MRS broth (OXOID TM , Thermo Fisher Scientific, USA) then incubated aerobically (in Gas Pack, Thermo, USA) at 37 o C for 24 hrs. After that, a confirmation test was FULL PAPER carried out to ensure the characteristics of the strain including gas production tests from glucose metabolism, catalase test, Gram staining, and cell morphology.

Screening of GABA-producing Lactobacillus rhamnosus
A total of 50 μL of culture broth of L. rhamnosus that have been cultured in MRS broth for 24 hrs was inoculated into MRS broth containing 2% monosodium glutamate (MSG), a commercial product from Ajinomoto, purchased in the local market and was incubated at 37 o C for 48 hrs. Cultures were centrifuged at 10,000 rpm at 5 o C, for 10 mins. The supernatant was then taken and the transformation of NA glutamate to GABA was analysed using TLC. The supernatant (1 μL) was spotted on a TLC plate (Silica gel 60 F254, Merck No. 1.05554.0001) and the plate was eluted using a mobile phase containing butanol (Merck): glacial acetic acid (Merck): water (5:3:2 v/v). For colour development, 0.5% (w/v) ninhydrin solution in ethanol was used, followed by heating at 105 o C for 5 mins. The presence of GABA spots was observed under UV rays at 254 nm and 366 nm . The spots which have the same R f as the spot of standard GABA were then calculated quantitatively.

Quantitative analysis of GABA concentration produced by Lactobacillu rhamnosus
Quantitative GABA analysis was performed using UPLC (Waters) according to the procedure described by Wu and Shah (2015) with slight modifications. The details are described as follows. A total of 0.1-1.0 g of fermented milk sample was put into 20 mL vials headspace (Waters No.186000384C) and added with 5 mL of 5N HCL. It was then heated at 110 o C for 22 hrs. The entire sample was transferred into a 50 mL measuring flask, homogenized, and successively filtered using an ash-free filter paper No. 42 and using a GHP / RC 0.2 μm filter. The filtrate (500 μL) was transferred into 2 mL tubes, added with 400 μL of 2.5 mM ABAA solution as the internal standard, and 460 μL of distilled water. Derivatization of the sample was carried out by taking 10 μL of standard solution or sample solution that has been added to the internal standard into the insert vial, and then added with 70 μL of AccQ-Tag Flour Borate Buffer, and was mixed vigorously. After that, the solution was added with 20 μL of AccQ-Tag Reagent 2A followed by heating at 60 o C for 10 mins. After cooling, 1 μL of the sample was injected into the UPLC. The analytical conditions were as follows. The column was AccQ-Tag Ultra C18 1.7 μm (2.1×100 mm), which was run at 49 o C throughout the detection time. The mobile phase consisted of four solutions namely: solution A, AccQ-Tag Ultra Eluent A 100%; solution B, a mixture of AccQ-Tag Ultra Eluent B with double distilled water in a ratio of 90:1; solution C, double distilled water; and solution D, AccQ-Tag Ultra Eluent B 100%. The mobile phase flow rate was adjusted at 0.5 mL/min. with gradient pump system. The injection volume of the sample was 1 μL, and GABA was detected using a PDA detector at 260 nm.

Preparation of starter culture
The selected L. rhamnosus strains were grown in 5 mL of MRS broth at 37 o C for 24 hrs. Broth culture was vortexed and 1 mL was put into 2 mL Eppendorf tubes and centrifuged at 7,000 rpm for 5 mins at 5 o C to separate the cell mass and the supernatant. The supernatant was discarded and the cell mass obtained was washed twice using saline solution (0.85% NaCl) and then suspended using 5 mL of skim milk, henceforth referred to as inoculum. The population of L. rhamnosus strains (CFU/mL) in Culture broth and CFU/gr in starter culture on fermented milk, were enumerated by culturing them on MRS agar, incubated at 37 o C for 24 hrs anaerobically using a gas pack.

Fermentation of skim milk by Lactobacillus rhamnosus
The process of making fermented milk with GABAproducing L. rhamnosus using skim milk (Diamond brand, commercial skim milk purchased at a local supermarket) was preceded by making a starter culture and milk fermentation using an active starter culture. Liquid skim milk (Diamond brand commercial milk) was pasteurized at 80 o C for 30 mins, then as much as 100 mL was put into a glass jar and cooled down to 45 o C. Skim milk in a jar was inoculated with 2% v/v inoculum and incubated for 48 hrs at 43 o C. A total of 10 g of fermented milk in a glass beaker was added with 10 mL of distilled water, stirred, and then the pH was measured using a pH meter (Martini instruments Mi-1050), previously calibrated using pH 4.1 and 7.0 buffers. Then, the starter culture is ready to be used for making fermented milk.
Liquid skim milk was pasteurized at 80 o C for 30 mins, then as much as 100 mL was put into a glass jar and cooled down to 45 o C. Skim milk in a jar was inoculated with a 5% w/v starter and then incubated at 43 o C for 72 hrs (to reach the pH of fermented milk according to the pH of yoghurt pH <4.5). The pH of the fermented milk was measured as previously described.

Determination of GABA content in fermented milk
A total of 1 mL of fermented milk was centrifuged at eISSN: 2550-2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER 13,000 rpm at 5 o C for 10 mins, then the supernatant was taken and tested quantitatively by UPLC (Seo et al., 2013) according to the method described above.

Data analysis
The data were analysed using SPSS 16 package and results were expressed as mean ± standard deviation from three replications. Significant differences were assessed by one-way analysis of variance (ANOVA) at a p-value of ≤0.05.

Bacterial strain and culture conditions
A total of 34 strains of L. rhamnosus isolates from Sumbawa mare milk, retrieved from the Unud Culture Collection (UNCC) were screened in this study ( Table  1). The strains were previously identified, and their resistance to the gastrointestinal tract in vitro for probiotic application was characterized (Shi et al., 2012). Moreover, its potential as a starter of fermented milk has also been carried out (Dewi et al., 2012). Before using any strains for further study, the strains were cultured in MRS broth and the purity of the strains was checked by observing the shape and cell morphology by Gram staining, gas formation, and catalase. The results confirmed that all strains were long rods appearing in long-chains cell arrangement, Gram-positive, catalasenegative, and there was no gas production from glucose (homofermentative). All strains showed good growth performance and good purity and are ready to be used for further research.

Screening of GABA-producing Lactobacillus rhamnosus
Isolated L. rhamnosus was cultured in MRS broth media containing 2% MSG and incubated at 37 o C for 24 hrs. The results showed that only 17 out of 34 strains were able to produce GABA (R f = 0.61) at detectable concentrations (Table 1). The other strains did not result in detectable GABA under our experimental protocols, even though glutamate, a precursor of GABA, was added to the medium. This result implied that activities of glutamate decarboxylation (GAD) enzyme in those strains were diminished, which might be altered by the growth conditions. GABA is reported to be synthesized by LAB through decarboxylation of glutamic acid molecules catalyzed by the glutamate decarboxylase enzyme Diana et al., 2014;Zhuang et al., 2018). The differences in GABA production by different strains have been reported to be influenced by various factors. The series of previous research results show that GABA is produced depending on the concentration of glutamate (as a precursor) and the presence of pyridoxal 5'-phosphate (PLP) as a cofactor to catalyze α-decarboxylation of glutamic acid or its salts in growth media (Shi and Li, 2011;Xu et al., 2017). Therefore, in the early screening, L-glutamate is deliberately added to induce the activity of the GAD enzyme, although GABA may also be formed through the other metabolic pathways of microorganisms (Cui et al., 2020). We examined the L. rhamnosus population ( Table 2) and showed that the population of strains was not much different and the pH of fermented milk was at an optimum value of GAD activity (pH 3.8-4.6) (Capitani et al., 2003). This evidence suggested that the GABA production was merely due to the characteristic of individual strains investigated.
GABA is mainly formed by the α-decarboxylation reaction of L-glutamic acid or its salts, which is catalyzed by the glutamic acid decarboxylase enzyme (GAD; EC 4.1.1.15) Diana et al., 2014;Zhuang et al., 2018). The biochemical properties of GAD have been well characterized (Nomura et al., 1999). GAD enzymes are found in various types of microorganisms such as LAB (Bertoldi et al., 1999), Escherichia coli (Rice et al., 1993), Streptococcus, Aspergillus (Kato et al., 2002), and Neurospora (Kubicek and Hampel, 1979). GAD is also found in plants such as tea (Zhao et al., 2011), tomatoes, soybean (Serraj et al., 1998), mulberry leaves (Yang et al., 2012), germinated brown rice (Yong-qiang, 2008), and petunias (Johnson et al. 1997). Even GAD is also found in the brains of mammals (Nathan et al., 1994). However, for commercial purposes and its application in food and pharmaceuticals, bacterial groups such as the LAB group (Maras et al., 1992) and yeasts (Hao and Schmit, 1993), are very important producers of GABA because they can also be applied in the food fermentation process.
As given in Table 1, the activity of GAD to synthesize GABA without PLP addition was different from each of the screened strains. It was also seen that 17/34 (50%) of the L. rhamnosus strains screened did not show GABA formation at detectable concentrations. To optimize the formation of GABA, it is necessary to carry out further studies, especially in the optimization of growth conditions and the addition of PLP as an activator of the GAD enzyme in the L. rhamnosus strain used in this study.

Production of GABA in fermented milk by Lactobacillus rhamnosus
A total of 17 strains capable of producing GABA (Table 1) were screened first by observing their growth rate on the MRS broth medium. Based on the screening results, 6 strains of L. rhamnosus (Table 2)  The pH value of fermented milk produced by these 6 strains ranged from 3.91-4.14 ( Table 2). In the context of GABA production, the optimal activity of GAD to produce GABA is at pH 4.4 (Huang et al., 2016), which means that GABA may be produced during the milk fermentation process in this study. Furthermore, it was reported that in using Lactococcus spp. in the manufacture of fermented milk, GABA was produced at a pH range of 4-6 at 37 o C (Santos-Espinosa et al., 2020).
The production of GABA by L. rhamnosus in this study on MRS media ranged from 0.24 to 2.27 mg/mL. Meanwhile, the addition of 2% glutamate in MRS increased GABA production, ranging from 0.24 to 0.29 mg/mL. In addition, the GABA production in milk ranged from 0.07-2.01 mg/mL (Tables 1 and 2). The production of GABA, which is different from each strain, is influenced by the fermentation activity. The fermentation process is reported to cause various LAB stresses, especially due to exposure to low pH, changes in osmotic pressure, and the diminishing presence of nutrients. The conversion activity of primary acid to GABA by GAD is accompanied by intracellular consumption of protons (hydrogen ions) (Feehily and Karatzas, 2013), which causes a decrease in pH (Le Vo et al., 2012).
Lactobacillus is the most frequently reported as GABA-producing LAB, which is widely applied in several food products such as raw milk, cheese, fermented milk, pickles, tea, and vegetable drinks with their GAD well characterized (Lin, 2013;Nejati et al., 2013;Franciosi et al., 2015;Song and Yu, 2018;Santos-Espinosa et al., 2020). The fermented milk made in this study used skim milk and L. rhamnosus without the addition of MSG or PLP. However, the activity of GAD on L. rhamnosus used may have also been stimulated by the presence of PLP in skim milk which may exist in a low concentration of approximately 3 μM as reported by Li et al. (2016) and Schmidt et al. (2017). From the results, it was also found that the highest GABA was produced by L. rhamnosus SMM37 at about 2.01 mg/ mL. Even though the produced GABA concentration was lower than that produced by other L. rhamnosus strains, it is still an opportunity to optimize GABA yield through optimizing the growth condition.
Various studies have been conducted to evaluate the production of GABA on various growth factors such as substrates in culture media, as well as in food products such as vegetable drinks, meat products, seafood, and dairy products (Ratanaburee et al., 2013;Shan et al., 2015;Taoka et al., 2019). All of these study reports state that natural GABA production depends mainly on glutamate concentration in the food matrix, the LAB strain used, and the fermentation conditions. It was reported that the yoghurt added with Lactobacillus plantarum NDC75017 affects GABA production, which reaches 3.15 g/kg (Shan et al., 2015). In another study, Lactobacillus brevis 877G produced 1.95 g/L GABA in skim milk. Using a mixture of L. brevis 877G and Lactobacillus sakei 795 (in the proportion of 1:1) GABA production was increased to 2.32 g/L (Seo et al., 2013). On the other hand, Lactobacillus rhamnosus GG used in the fermentation of adzuki beans can produce 1.12 mg/ mL GABA (Song and Yu, 2018). In this study, however, the production of GABA by L. rhamnosus is lower than the L. rhamnosus GG, which was probably caused by the growth conditions of the strain characteristics (Minervini et al., 2009) or the expression of the GAD enzyme (Li et al., 2013). In addition, some strains may have deficiencies in the glutamate-GABA transport system to the inside and outside of the cell (Feehily and Karatzas, 2013) or the assimilation of glutamate in other metabolic pathways, which will limit GABA production (Fernández and Zúñiga, 2006). Therefore, these results need to be further investigated.

Conclusion
It can be concluded from the results that 17 L. rhamnosus strains isolated from Sumbawa mare milk were able to produce GABA in MRS without and with the addition of monosodium glutamate. Further studies on 6 selected strains to ferment skim milk showed that the highest GABA was produced by L. rhamnosus SMM37 at about 2.01 mg/mL. Further studies are needed