Probiotic Properties and Optimization of Gamma-Aminobutyric Acid Production by Lactiplantibacillus plantarum FBT215

Gamma-aminobutyric acid (GABA) improves various physiological illnesses, including diabetes, hypertension, depression, memory lapse, and insomnia in humans. Therefore, interest in the commercial production of GABA is steadily increasing. Lactic acid bacteria (LAB) have widely been reported as a GABA producer and are safe for human consumption. In this study, GABA-producing LAB were preliminarily identified and quantified via GABase assay. The acid and bile tolerance of the L. plantarum FBT215 strain were evaluated. The one-factor-at-a-time (OFAT) strategy was applied to determine the optimal conditions for GABA production using HPLC. Response surface methodology (RSM) with Box-Behnken design was used to predict the optimum GABA production. The strain FBT215 was shown to be acid and bile tolerant. The optimization of GABA production via the OFAT strategy resulted in an average GABA concentration of 1688.65 ± 14.29 μg/ml, while it was 1812.16 ± 23.16 μg/ml when RSM was applied. In conclusion, this study provides the optimum culture conditions for GABA production by the strain FBT215 and indicates that L. plantarum FBT215 is potentially promising for commercial functional probiotics with health claims.


16S rDNA Sequencing and Identification of Adhesion-Related Gene Sequence
The universal primer set 27F (5'-AGAGTTTGATCMTGGCTCAG-3') and 1492R (5'-GGTTACCTTGTT ACGACTT-3') were used to identify the GABA producers [23]. SrtAF (5'-ATGAAGTCCAAGCAACA-3') and SrtAR (5'-TTAATATTTGTTATTAAAATGACTTG-3') were used for the amplification of the sortase A (SrtA). The PCR cycling conditions were: initial denaturation at 95 o C/5 min, 35 cycles of denaturation at 95 o C/30 sec, annealing at 55 o C/30 sec, and extension at 72 o C/30 sec, and a final extension at 72 o C/7 min. DNA sequencing was performed at Macrogen (Republic of Korea). Sequence comparisons were performed using the Basic Local Alignment Search Tool (BLAST) available at the National Center for Biotechnology Information (NCBI; National Institutes of Health, USA). Sequence alignments were performed using BioEdit (Ibis Biosciences, USA).

Acid and Bile Tolerance
L. plantarum FBT215 was cultured in MRS broth at 37°C for 24 h. The culture broth was centrifuged at 6,000 ×g for 5 min, and the bacterial pellet was resuspended in saline. The acid tolerance was determined after adjusting the pH of the suspensions to pH 2.5, 3.0, and 6.0 with 1% (w/v) pepsin (Junsei Chemical Co., Japan). The viable cell count was measured every 1 h for 2 h on MRS agar. The bacterial culture was incubated in pH-adjusted MRS broth (pH 3.0) at 37°C for 2 h to determine bile tolerance. Subsequently, the culture broth was centrifuged at 6,000 ×g for 5 min; the medium was replaced with an equal volume of MRS broth with 0.3% (w/v) bile (oxgall, Sigma-Aldrich). The viable cell count was measured every 3 h for 6 h on MRS agar.

HPLC
GABA production was analyzed using HPLC (1260 Infinity series; Agilent Technologies) and a Poroshell120 HPH-C18 (4.6 mm × 150 mm × 4 μm; Agilent technologies). The pre-column ortho-phthalaldehyde derivatization method was used for analyzing GABA production. The automated injection program was performed as described by Agilent Technologies (USA), with slight modifications. In brief, solvent A contained 10 mM Na 2 HPO 4 , 10 mM Na 2 B 4 O 7 , and 5 mM NaN 3 ; solvent B contained acetonitrile: methanol: water (45:45:10, v:v:v). The solutions were filtered through a 0.20 μm membrane (Hyundai Micro Co., LTD, Republic of Korea) using vacuum filtration. Sonication (JAC-1505; Kodo, Hwaseong-si, Republic of Korea) was performed for the bubble decay process. The isocratic elution conditions were: 2% solvent B from 0 min to 0.35 min, 57% solvent B from 0.35 min to 13.4 min, 100% solvent B from 13.4 min to 20.3 min, and 0% mobile phase B from 20.3 min to 23.0 min.

Optimization of GABA Production via OFAT Strategy
The OFAT strategy was used to determine the optimal conditions for GABA production by L. plantarum FBT215; the optimal temperature, initial pH, and fermentation time in MRS broth were determined. The effect of culture temperature in modified-MRS broth containing 50 mM MSG was evaluated at 25°C, 30°C, 37°C, 40°C, and 45°C. The influence of initial pH on GABA production was investigated at pH 3.5, 4.5, 5.5, 6.5, 7.5, 8.5, and 9.5. The optimal fermentation time was measured every 3 h for 12 h and then every 12 h until 96 h. The 2% glucose in MRS broth (pH 7.5) was replaced with different carbon sources (2% w/v; fructose, sucrose, lactose, galactose, xylose, maltose, mannitol, and lactulose). The 1% proteose-peptone No. 3 and 1% beef extract in MRS broth (pH 7.5) were replaced with different nitrogen sources (2% w/v; peptone, tryptone, soytone, proteose-peptone No.3, malt extract, and beef extract). The 2% glucose in MRS broth (pH 7.5) was replaced with three fructosecontaining polymers (2% w/v; fructooligosaccharides, inulin, and raffinose) to evaluate the availability of prebiotics as a carbon source. The MSG concentrations (25-250 mM) and PLP concentrations (0-50 mM) were individually optimized.

Optimization of GABA Production via RSM
The RSM package in R software, version 4.1.2 (The R Foundation, Austria), was used to predict the optimal conditions for GABA production. RSM using the BBD model was performed, as described previously [7], with slight modifications. In brief, carbon concentration, nitrogen concentration, and initial pH were chosen as major factors influencing GABA production. Each factor was transformed into 3 coded levels (-1, 0, and 1). A secondorder model was used to fit the data for the Y responses (GABA concentration). The Quadratic equation is as follows: β 0 is constant; β i , β j , and β ii are coefficients of variables. X i and X j indicate the levels of independent variables. Analysis of variance and regression analysis were performed using R.

Sequence Accession Number
The 16S rRNA gene sequence of L. plantarum FBT215 (accession no. OL587487) is available on GenBank (NCBI).

Statistical Analysis
Assays were performed in triplicate, and the results were analyzed using IBM SPSS Statistics 25 (IBM Corp., USA). Data are presented as mean ± standard deviation in bar charts. One-way ANOVA and Tukey's multiple range tests were used to evaluate the significant differences (p < 0.01) among the groups.

Isolation and Identification of GABA-Producing LAB
Potential GABA-producing LAB strains were isolated from 10 varieties of Korean fermented foods. GABA production from the isolates was evaluated using a GABase assay. Ten isolates produced high amounts of GABA (54.37 ± 5.44 μg/ml to 144.02 ± 14.40 μg/ml, Fig. 1); among them, isolate FBT215, which is the highest GABA producer at 144.02 ± 14.40 μg/ml, was selected for further experiments. 16S rRNA gene sequencing identified isolate FBT215 as L. plantarum.

Optimization of GABA Production via OFAT Strategy
GABA production was analyzed using the OFAT strategy and HPLC to determine the optimal culture conditions for L. plantarum FBT215. The optimal culture temperature for GABA production was 37°C, at which the GABA concentration was 103.67 ± 1.65 μg/ml (Fig. 3A). The GABA content significantly decreased at temperatures greater than or less than 37°C. The optimal pH for GABA production was measured in the pH range of 4.5-9.5. The optimal initial pH was 7.5 and 8.5 (121.76 ± 1.14 μg/ml and 114.75 ± 0.56 μg/ml, respectively, Fig. 3B); and was not detected at pH 3.5. The GABA production increased steadily from 9 h (33.28 ± 0.43 μg/ml) to 72 h (151. 42 ± 1.96 μg/ml); however, it remained stable after 72 h (Fig. 3C).

Optimization of GABA Production via RSM
The optimal nutritional factors for GABA production were determined using the OFAT strategy; fructose and tryptone were the best carbon and nitrogen sources, respectively. Subsequently, the initial pH and the two nutritional factors were optimized using the RSM with the BBD model. The three variables were coded to three levels, and GABA production was calculated based on the average of triplicates. Data were analyzed using the quadratic regression model, and the critical variables were identified through backward elimination. As a result, the following equation was used. X 1 , X 2 , and X 3 indicate the concentration of fructose, tryptone, and the initial pH, respectively. This model's multiple R-squared and adjusted R-squared value was 0.9903 and 0.9730, respectively. The value of Pr (>F) of lack of fit was 0.057 (   The responses considering the interaction of two variables with the other variable fixed at the optimal point were illustrated using a three-dimensional response surface graph. The effect of fructose and tryptone concentrations on GABA production at an initial pH of 7.18 was analyzed (Fig. 4A). The relationship between fructose concentrations and initial pH was analyzed at a tryptone concentration of 2.94% (w/v, Fig. 4B). Additionally, the effect of tryptone concentration and initial pH was analyzed at a fructose concentration of 1.02% (w/v, Fig. 4C). The optimum conditions for maximum GABA production were determined using the quadratic regression model equation and response surface graph; they were 1.02% fructose, 2.94% tryptone, and an initial pH of 7.18. The GABA content was 2239.07 μg/ml at the optimum point.

Discussion
GABA is a major inhibitory neurotransmitter in the CNS; its beneficial effects are widely recognized. In this study, we investigated the basic probiotic properties of GABA-producing L. plantarum FBT215. In addition, the optimal conditions for GABA production by L. plantarum FBT215 were determined using the OFAT strategy and the RSM method; the optimal conditions determined through the two methods were compared.
The survival of probiotic bacteria in the gastrointestinal tract is incumbent on their acid and bile tolerance [24]. L. rhamnosus GG, a probiotic bacterium, was unable to grow at pH < 3.0 and could tolerate 0.15% (w/v) bile salts [25]. Similarly, L. plantarum FBT215 was vulnerable at low pH (< 3.0) but had tolerance to bile salts (Fig. 2). The pH in the human stomach ranges from 1.0 to 2.0; however, it could be 3.0 or higher in the presence of food [26]. Therefore, the application of L. plantarum FBT215 as probiotics could be more efficient when coated with an acidtolerance agent or consumed immediately after a meal. srtA codes sortase A; it recognizes LPxTG sorting motifs and influences the adhesion of L. acidophilus ATCC 4356. [27]. SrtA deletion significantly decreases the surface exposure of mannose-specific adhesins; this could explain the effects on mannose-specific adhesion [28]. The complete sequence of L. plantarum FBT215 SrtA was identified and was consistent with that of SrtA from L. plantarum ATCC 202195 (Accession No. QVG76613.1) and L. plantarum ATCC 14917 (Accession No. EFK27984.1). This work has opened new avenues to study the association between SrtA and the cell adhesion capacity of L. plantarum FBT215; the LPxTG sorting motifs and mannose-specific adhesin-related genes can be explored further.
The optimal conditions for efficient GABA production by L. plantarum FBT215 under various environmental and nutritional conditions were determined using HPLC. L. plantarum is a mesophilic bacterium; the optimal growth temperature is 37 o C in MRS broth [29]. Microbial GABA production is efficient at a low pH; however, GABA production is optimal at neutral pH in some Lactobacillus spp. [30]. The severe growth retardation of LAB at low pH could inadequate GABA production. The GABA production in L. plantarum FBT215 was decreased under acidic conditions and was maximum at pH 7.5 and 8.5 (Fig. 3B). The preferred carbon source for GABA production varies with the LAB strain: 1% (w/v) glucose plus 1% fructose for L. brevis CRL 2013 [31], 3% glucose for L. plantarum N5 [32], 2% maltose for L. brevis HYE1 [33], and 2% glucose for L. plantarum KCTC 3103 [13]. L. plantarum FBT215 produced the highest amounts of GABA with 1% fructose (Fig. 3E). Using this strain would be industrially advantageous due to the lower carbon source consumption than other LAB. The optimal nitrogen source for GABA production by L. plantarum FBT215 was 2% tryptone (Fig. 3G). These results contradict the earlier finding that yeast extract was the optimal nitrogen source [16,17,32]. However, yeast produces various biologically active ingredients containing GABA and glutamate [34,35]. Therefore, yeast extract was considered a growth factor for the precise analysis of GABA production by L. plantarum FBT215. Consequently, the optimum carbon and nitrogen sources for GABA production by L. plantarum FBT215 were 1% fructose and 2% tryptone, respectively. Fructose-based carbon sources, except inulin, were metabolized for GABA synthesis in L. plantarum FBT215. FOS and raffinose lyase could be encoded in the putative probiotics; further studies are warranted. The addition of PLP, a co-factor of glutamate decarboxylase (GAD), increased GABA production (Fig. 3J) and is a major factor influencing GABA production [36,37].
The OFAT strategy is traditionally employed to optimize the responses of interest. Several studies aimed to optimize the conditions for enhancing GABA production by LAB [15][16][17][18]. However, the method has limitations because the interactions of individual factors cannot be considered in a complex system; it can only be evaluated at specific points [33]. Therefore, RSM with the CCD model was assessed for optimizing GABA production in an L. fermentum isolate. Three variables were analyzed, i.e., glucose concentration, MSG concentration, and incubation time [38]. GABA production by Lactobacillus sp. Makhdzir Naser-1 was optimized with a focus on temperature, pH, glutamic acid concentration, and incubation time [39]. In this study, optimization of GABA production by L. plantarum FBT215 was performed using RSM with a BBD model. The actual GABA content at the optimal conditions predicted using the conventional and statistical methods was significantly different (p < 0.05), at 1688.65 ± 14.29 μg/ml and 1812.16 ± 23.16 μg/ml, respectively (data not shown). The actual GABA content via RSM with the BBD model was 80.93% of the predicted value. Therefore, RSM with the BBD model is more reliable for optimizing the GABA production in L. plantarum FBT215.
In the beginning, GABA was chemically synthesized to meet demands, but this synthesis method was replaced due to the higher yields and lower costs of the biosynthetic process [40]. In this study, optimization of GABA production by L. plantarum FBT215 was investigated under various conditions. LAB have widely been studied as GABA producers not only because of their GRAS status but also the difficulty of human consumption of other microorganisms. Although optimization of GABA-producing LAB has already been reported in previous studies, it is vital to screen new resources because of the differences in the fermentation profiles of individual LAB [41]. Among LAB, L. plantarum has been investigated as a key species for GABA production [42]. L. plantarum FBT215 produced a high amount of GABA in MRS-based medium compared to that by others: 0.74 g/l for L. plantarum Taj-Apis362 [43], 1.5 g/l for L. plantarum K154 [44], 0.1 g/l for L. plantarum NMZ [45], and 0.6 g/l for L. plantarum NTU 102 [46]. GABA has long been thought to be unable to cross the blood-brain barrier; however, it is suggested that GABA exerts veridical effects on the brain possibly via the enteric nervous system [4]. Studies in animals have reported that the gut microbiota can regulate GABAergic neurotransmission through the vagus nerve, which is the main pathway from the abdominal cavity to the brain [47,48]. Currently, the number of patients affected by physiological disorders is gradually increasing, and it has been suggested that GABA could potentially support a role for dysregulated GABAergic functioning in neural circuits [49]. Therefore, further studies are needed on GABA production by L. plantarum FBT215, which could exert a therapeutic effect on physiological illnesses in mammalian models.
In summary, the basic probiotic properties of the LAB strain L. plantarum FBT215 were investigated. GABA production was optimized using the OFAT strategy and RSM with a BBD model. This study can provide culture conditions for commercial functional probiotics with health claims in vitro. Further studies are warranted to optimize the various culture conditions to enhance GABA production and understand the in vivo physiological effects of GABA produced by L. plantarum FBT215.