L. brevis CGMCC 1306 strain was streaked on fresh MRS agar plates for single colony growth, transferred to MRS broth, and grown anaerobically at 37°C for 48 h. Streptococcus thermophilus ATCC 14485 was transferred from commercial agar plates to M17 medium and grown anaerobically at 37°C for 48 h. Agar (1.8% w/v) was added to the corresponding medium if a solid medium was required.

The method used was based on national food safety standards with modifications (52). A 25 g yogurt sample was mixed with 225 mL of normal saline for dilution. Next, 10 mL of the mixture was added to 90 mL of normal saline, and this step gradient was continued for normal dilution. The MRS agar plates were spread with 100 μL of the corresponding diluents and anaerobically grown at 37°C for 48 h. Single colony numbers were then obtained, and the unit of viable count was described as cfu/mL.

A 10% (w/v) milk powder, with 1.5 g/L L-MSG as a substrate, and 8%(w/v) sucrose were dissolved in sterilized distilled water, homogenized and pasteurized in a water bath at 95°C for 15 min, and then cooled to 43°C. Subsequently, 8% (v/v) L. brevis CGMCC 1306 was added to the milk mixture and incubated at 43°C for 40 h. Multiple factors were tested to optimize the GABA yield in the yogurt, and the yogurt was stored at 4°C for 3 weeks continuously to evaluate yogurt storage stability based on commercial yogurt shelf life and storage conditions in the market. The water holding capacity (WHC), viscosity, and GABA stability tests were then conducted. The designed factors were as follows: fermentation time was from 8 to 48 h; fermentation temperature was 31, 34, 37, 40, 43, and 46°C; L-MSG concentration was 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 g/L; sucrose content was 2, 4, 6, 8, and 10%; and the inoculation amount was 2, 4, 6, 8, and 10%. Mixed fermentation was studied using a 0.25:1, 0.5:1, 1:1, 2:1, 3:1, and 4:1 ratio between L. brevis CGMCC 1306 and S. thermophilus ATCC 14485.

GABA concentration was determined using high-performance liquid chromatography (HPLC; LC-2030, SHIMADZU, Japan). The sample was pretreated before the HPLC test via centrifugation at 8000 rpm for 3 min (Centrifuge 5,424, Eppendorf, Germany). The supernatant was mixed with 0.5 mol/L NaHCO₃ and 8 g/L dansyl chloride–acetone solution and then reacted for 1 h at 30°C in the dark. The mixtures were filtered and measured with a chromatographic column (Hypersil ODS2 C18, Elite, China) using a gradient elution procedure with mobile phases A and B. The mobile phase A was ethanol, and mobile phase B was a complex of tetrahydrofuran, methanol, and 0.05 mol/L sodium acetate (5:75:420, v/v/v) (53).

The WHC assesses casein coagulate compactness and represents the yogurt texture. The yogurt samples were gently stirred to homogeneity and centrifuged for 10 min at 1250 × g at 4°C. The WHC (% w/w) was evaluated by weight change after whey was expelled according to Equation (1) (54) as follows:

where Y is the weight of the yogurt sample, and W is the weight of whey expelled.

The viscosity of the yogurt was measured at different fermentation periods using a viscometer (DV2T; Brookfield, United States). The samples previously incubated at 4°C were adapted to 25°C for 30 min and fitted with LV-1 rotor at 30 rpm.

The appearance, odor, taste, acidity, texture of the fermented milk, and overall acceptance were estimated by 40 professionally trained panelists, with scores ranging from 1 to 9. The two control groups were commercial yogurts purchased from the market. The quantitative scoring rules are described in Table 1. The method was modified based on Han’s method, and a commercial yogurt was used as a control group (55).

All data are represented as mean ± SD. Data analysis was performed using Origin 8.5. Statistical significance was determined using IBM SPSS Statistics 27. Duncan’s test was used to analyze the statistical significance variability between samples using analysis of variance with a significance level of 95% (α = 0.05). The experiments were conducted in triplicate.

Total fermentation time is essential to yogurt production because it determines the microbial growth stage and GABA production. It also influences the yogurt texture as the coagulation level of casein in milk is affected by the acidifying procedure during the fermentation ending point. GABA production, microbe growth dynamics, and acidifying procedure compared with fermentation time are shown in Figure 1. The data of the present study indicated that GABA accumulated continuously in the early stage of fermentation (< 24 h), whereas GABA yield was comparatively stable at 32–40 h with 33.3 mg/100 g. Microbial growth demonstrated a similar trend to GABA production before 24 h, as cells began to decay from 24 h to 48 h. The acidifying procedure was maintained throughout fermentation, and the final yogurt pH decreased to 4.5 at the end of fermentation. Therefore, fermentation time significantly affected GABA yield (p < 0.05).

In the early fermentation period, the resources were intensively utilized for biomass increase, and GABA yield was limited. However, it plateaued after 32 h of fermentation as shown in Figure 1A, as the residual microbes aimed to maintain basic physiological activity instead of synthesizing GABA, which led to slight increases in GABA yield. Yogurt pH decreased persistently but slowly owing to the LAB growth, and the final pH was 4.5 shown in Figure 1A. This pH was higher than that of commercial yogurts, which normally ranges from 4.1 to 4.3. This result was attributed to GABA production, which is H⁺ consuming, thus neutralizing the pH decrease resulting from the growth of the LAB. This characteristic could help prevent yogurt over-acidification and, thus, improve shelf life. As expected, L. brevis CGMCC 1306 displayed good growth characteristics, even in the initial stage (< 8 h) of fermentation, the viable cell number reached 1.65 × 10⁷ cfu/mL (Figure 1B), which is on par with commercial yogurt. After 24 h, the nutrient levels in the milk gradually decreased and could not support the high level of cell growth; consequently, the number of microorganisms began to decline. Meanwhile, cell growth and decreased pH mutually promote each other. That is, LAB converts the lactose in milk to lactic acid, decreasing the pH and creating a relatively acidic environment to help LAB grow. The procedure stimulated and induced glutamic acid decarboxylase (GAD) activity to gradually catalyze L-MSG to GABA, verifying the GABA accumulation trend. Therefore, 40 h of fermentation was recommended to achieve sufficient GABA yields and an appropriate pH during yogurt production.

Fermentation temperature influences microbial growth and exopolysaccharide (EPS) formation, affecting the acidity procedures of different LAB stains, which may influence the yogurt texture and GABA yield (56). The GABA yield, yogurt pH, and microbe growth are shown in Figure 2. GABA yield was the highest at 43°C and microbial growth was limited by high temperature. The pH at the different fermentation temperatures ranged from 4.4 to 4.9. The results indicate that fermentation temperature significantly affected GABA yield (p < 0.05).

The GABA yield increased at <43°C and decreased at >43°C (Figure 2A). Two reasons reliably explain the climbing trend. First, the increased temperature aided microbial growth and proliferation. Second, the optimum catalytic temperature for glutamic acid decarboxylase (GAD), which transforms L-MSG into GABA, is >40°C. The higher temperatures caused cell aging and increased the portion of crosslinking between the casein micelle branched chains within the microcavity inner surface of yogurt, reducing the WHC (57, 58). Moreover, the irregular inner structure could lead to the irregular distribution of microbes, potentially increasing GABA mass transfer resistance. This also explains the microbial growth trend (Figure 2B). The final pH is a delicate balance between GABA production which is H⁺ consuming and microbial physiological activity which is H⁺ generating. The influence of fermentation temperature on yogurt pH was not obvious (Figure 2A). The optimal yogurt production temperature selected was 43°C, which is the same as that for commercial yogurt, indicating that GABA-enriched yogurt could potentially be manufactured in dairy plants without additional equipment and energy costs.

L-MSG, a substrate of GAD, can be catalyzed into GABA by LAB. Although some LAB can synthesize GABA using a precursor substance within the cell, the resulting GABA production is limited. The addition of L-MSG into the fermentation liquid is necessary to achieve sufficient GABA production. GABA production, microbial growth, and pH compared with L-MSG concentration are shown in Figure 3. The data indicated that GABA yield increases sharply with an L-MSG concentration of <1.5 g/L, and there is a slight decrease with >1.5 g/L. Therefore, the L-MSG concentration significantly affected GABA yield (p < 0.05).

GABA yield exhibited a continuous and rapid increase when the L-MSG addition was <1.5 g/L (Figure 3A). This result was attributed to the abundance of microbes compared with the substrate in the microenvironment, and GAD and microbes could adequately leverage the substrate (Figure 3B). As GABA synthesis precursor, L-MSG could improve carbon flux in the GABA metabolic pathway, thus regulating catalysis. The increase in substrate concentration stimulated a GAD catalysis reaction to move in the forward direction, converting L-MSG to GABA. When the L-MSG addition reached 1.5 g/L, the substrate became excessive compared to GAD. When the GAD is fully loaded, feedback inhibition reduced enzyme activity, resulting in slight decreases in GABA yield (59). Therefore, 1.5 g/L L-MSG was identified as the optimal concentration for yogurt production.

Sucrose can be used by microbes as an additional energy source and provides flavor to yogurt. Adding sucrose could potentially improve the robustness and persistence of LAB fermentation. The effects of 2–12% sucrose additions on GABA yield, microbial growth, and pH were analyzed (Figure 4). The data revealed that 6–8% sucrose additions are suitable for GABA production, as the GABA yield reached 56.9 mg/100 g, and microbial growth exhibited a similar trend. Sucrose content significantly affected GABA yield (p < 0.05). The exogenous carbohydrate supplement could help LAB growth and GABA production, and this is in line with previous studies (60). Sucrose is an easy-to-use carbohydrate for bacteria that can be decomposed into glucose, readily supplying carbon resources. The addition of sucrose promoted LAB growth, thus complementing the rapid lactose consumption in milk. Although the GABA yield increased notably with 6–8% sucrose additions, it declined with additions exceeding >8%. This could attributed to the inhibition of microbial growth by high osmotic pressure caused by the high sugar concentration. Consequently, 6–8% sucrose additions were deemed optimal for yogurt production.

The inoculation amount determined the viable microorganism cell numbers during fermentation, particularly within the initial fermentation period. Meanwhile, the viable LAB abundance reflected the GABA production capacity. To determine the optimum microbial quantity for GABA production, different inoculation amounts were evaluated, and microbial growth and acidity were monitored (Figure 5). GABA yield increased with the inoculation amounts, reaching a peak with the 8% addition. The microbial growth increased within the tested range, and the inoculation amount significantly affected GABA yield (p < 0.05).

The increase in biomass resulted in notable increases in GABA yield with low inoculation levels (< 8%; Figures 5A,B). This result was attributed to the fact that increasing the inoculation amount could improve the conversion efficiency during this period, with sufficient viable cells leveraging L-MSG and transforming it to GABA. However, when the inoculation amount reached 8%, a slight decrease was observed. This result was attributed to the high cell concentration, which led to L-MSG being outcompeted by the excessive microbes in the microenvironment, hindering its conversion within the inner cells. Moreover, the behavior of the excessive microbes shifted from GABA production to competing for nutrition to ensure survival. Consequently, 8% was selected as the optimum inoculation amount for yogurt production.

In commercial yogurt, mixed fermentation is often used to improve the sensory and nutritional characteristics of the yogurt. Thus, the GABA production of L. brevis CGMCC 1306 under mixed fermentation conditions was evaluated. It was fermented with different proportions of S. thermophilus ATCC 14485. The GABA yield and acidity results are shown in Figure 6A. The data demonstrated that L. brevis CGMCC 1306 exhibited good GABA production when the ratio between S. thermophilus ATCC 14485 and L. brevis CGMCC 1306 was 0.5:1. GABA production was significantly enhanced by mixed fermentation, reaching nearly 75.3 mg/100 g. The pH at the fermentation endpoint was 4.1–4.4, which is lower than with pure fermentation. Therefore, mixed fermentation significantly affected GABA yield (p < 0.05).

The GABA yield with mixed fermentation may be closely related to the symbiotic effects between S. thermophilus and L. brevis (61). Ratios of 0.5:1 and 0.25:1 were deemed suitable for GABA production, as the GABA yields achieved maximum values of 75.3 mg/100 g (Figure 6A). At these ratios, the S. thermophilus cells grew rapidly and produced formic acid, folic acid, and carbon dioxide. The acidic environment created by S. thermophilus cells could accelerate the speed of L. brevis growth. Simultaneously, L. brevis could massively hydrolyze proteins in milk into peptides or various amino acids, aiding S. thermophilus growth and thus promoting GABA production. In this way, the two strains maintained a valuable symbiotic relationship. When the ratio was >1:1, the biomass in the fermentation liquid became too high, and the cells began to compete for the nutrition in the milk, resulting in reduced GABA production. Owing to the strong acidification ability of S. thermophilus, the yogurt pH with mixed fermentation was lower than with pure fermentation with one strain at the fermentation ending point. The large, strong casein coagula agglomerated by lower pH also increased the mass transfer resistance for GABA diffusion, which reduced the GABA yield. In conclusion, a ratio between S. thermophilus ATCC 14485 and L. brevis CGMCC 1306 of 0.5:1 was identified as optimal for GABA-enriched yogurt mixed fermentation.

WHC demonstrates the ability of macromolecular substances to combine with water in the yogurt, which indicates the stability of the system. The WHC was tested throughout the fermentation and storage periods for the GABA-enriched yogurt to estimate yogurt syneresis. The relationship between WHC and time is shown in Figure 6B. Throughout the fermentation period, the WHC was affected by fermentation time. However, during the storage period, there were no significant results with time.

The WHC began to increase when casein precipitation was observed, rising from 43.9% at 24 h of fermentation to 48.1% at 40 h of fermentation, and remained relatively stable during the 3-week storage period. With increased fermentation time, LAB growth and the pH of bulk liquid decreased, and the weight of the casein coagulates increased, causing an increase in the WHC. After fermentation, the yogurt was stored at 4°C, and the whey was slightly expelled but remained stable during the storage period (62). The WHC was comparable to that of commercial yogurt, indicating that the GABA-enriched yogurt exhibited good stability.

Viscosity is another characterization of the yogurt texture. The casein in milk coagulates due to the acidification process during milk fermentation, which alters yogurt viscosity (Figure 7A). The data showed an increase in viscosity during yogurt fermentation and storage time, ultimately reaching 620 cp.

The coagulum in the yogurt consisted mainly of casein, and the connections between coagulated casein contributed to the increased viscosity. Moreover, EPS produced by microbes during fermentation also contributed to the increased viscosity. The viscosity of the GABA-enriched yogurt was comparable to that of some commercially available yogurts and showed stability during the storage period.

Sensory tests can provide an overall description of a yogurt. Appearance, odor, taste, acidity, texture, and overall acceptance were tested in this study (Table 2).

The data showed that overall acceptance was not significantly different between GABA-enriched yogurt and commercial yogurt. The appearances and textures of the three types of yogurt tested received similar rankings. The odor of commercial yogurt was slightly better than that of GABA yogurt, but not significantly. The taste and acidity of the two commercial yogurts were rated equally and were considered better than those of the GABA-enriched yogurt. The chosen commercial yogurts were common and popular in the market. These results indicated that GABA-enriched yogurt was competitive with commercial yogurt in terms of its sensory characteristics, and it is expected that the health benefits conferred by GABA would further improve its ranking.

The storage stability of the GABA-enriched yogurt was tested (Figure 7B). The data indicated robust stability, as during the 3-week storage period, the GABA content was maintained at 75 mg/100 g. This provides further evidence of the robustness of the GABA-enriched yogurt.

Overall, the target yogurt with a pH of 4.37 ± 0.06 and acidity 89 ± 2 °T, fermented with naturally accumulated GABA, exhibited competitive sensory quality and stability compared to commercial yogurt. This yogurt might be effective for improving sleep quality, hypertension, and anxiety. Moreover, the potential applications for GABA-enriched yogurt could be expanded, given its inability to develop drug resistance.

The GABA-enriched yogurt provides a comprehensive nutrition product for the dairy industry. This novel yogurt provides triple benefits for human health: viable LAB benefits intestinal health, GABA provides additional functions, and the yogurt is nutrient-rich. Moreover, for traditional dairy manufacturers, the optimized yogurt production condition described in this study could be achieved using existing equipment without extra processing investment. The only additional material required for fermentation compared with conventional yogurt is L-MSG, the cost of which is negligible compared with other materials. Furthermore, GABA-enriched yogurt as a new competitive functional food with broad applications could provide manufacturers with a wider range for independent pricing to achieve optimum returns.

In this study, mixed fermentation was beneficial to GABA production, suggesting an opportunity for further GABA yield increases in future investigations. In addition, mixed fermentation could improve the taste of yogurt by increasing the flavor substances under the symbiotic effects between two or even three strains. Furthermore, additional sugar resources could be considered as extra energy sources for LAB cell growth and GABA production since sucrose has proven beneficial for GABA yield.