Probiotic fermentation of soybeans and use of fermented soy in cookies

TL;DR

This study explores using probiotic fermentation of soybeans to boost GABA content, which may help with mental health, and shows that fermented soy can be used in cookies without compromising texture.

Contribution

The study demonstrates that Lactococcus lactis fermentation enhances GABA content and preserves cookie quality better than unfermented soy.

Findings

Co-fermentation with Lactococcus lactis and Lactiplantibacillus plantarum significantly reduced phytic acid levels.

Cookies with Lactococcus lactis-fermented soybean flour had higher GABA content and better hardness than those with unfermented soy.

Fermentation with Lactococcus lactis preserved cookie texture comparable to 100% wheat flour cookies.

Abstract

With increasing prevalence of mental health issues such as insomnia and anxiety, γ‐aminobutyric acid (GABA) has gained attention for its neurological benefits. The development of GABA‐rich functional foods has emerged as a promising research direction. Soybeans, rich in glutamate, are excellent substrates for GABA biosynthesis using microbial fermentation. In this study, Lactococcus lactis (LL) and Lactiplantibacillus plantarum (LP) were used to ferment soybeans, aiming to enhance GABA production. The effects of fermentation on pH value, yield, colony‐forming unit, phytic acid content, total phenol content, and ABTS+ free radical scavenging activity were investigated. Additionally, soybean flour fermented by LL was incorporated into cookie formulations to evaluate its effects on cookie spread factor and hardness. The GABA content of the cookies was also analyzed.

Co‐fermentation with LL and LP significantly reduced phytic acid levels. Mono‐fermentation with LL showed higher probiotic viability, better retention of total phenolics and antioxidant capacity, and superior GABA accumulation. Cookies supplemented with LL‐fermented soybean flour better maintained the hardness and exhibited a higher GABA content compared to those prepared with unfermented soybean flour (P < 0.05).

Fermentation with LL effectively enhanced GABA content (P < 0.05) and preserved cookie texture comparable to that of 100% wheat flour cookies. These results can support the potential of LL fermentation in developing GABA‐enriched plant‐based functional foods. © 2025 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Figures

Colony‐forming units of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. Different lower‐case letters indicate significant differences (P < 0.05).

Phytic acid content of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. An asterisk indicates significant difference compared with the soybean control group (P < 0.05).

Total phenolic content (A), and ABTS+ radical clearing capacity (B) of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. Different lower‐case letters indicate significant differences (P < 0.05). GAE, gallic acid equivalents; TE, Trolox equivalents.

γ‐Aminobutyric acid (GABA) content of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. Different lower‐case letters indicate significant differences (P < 0.05).

Effect of the addition of soybeans fermented by Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP on the diameter (A), height (B), spread index (C) and hardness (D) of cookies. Different lower‐case letters indicate significant differences (P < 0.05).

Effect of addition of fermented soybean on γ‐aminobutyric acid (GABA) content of cookies. Different lower‐case letters indicate significant differences (P < 0.05).

Tables

Table 1. Yield and pH of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP) and a mixture of LL and LP

Bacteria	Dry temperature (°C)	Time (h)	Yield (%)	pH
LL	37	0	100.00 ±  0.00a	7.00
LL	37	24	98.71 ±  0.07cdef	4.50
LL	37	48	98.48 ±  0.67defg	4.30
LL	37	72	98.28 ±  0.53efg	4.20
LL	65	0	100.00 ±  0.00a	7.00
LL	65	24	98.67 ±  0.07cdef	4.50
LL	65	48	98.03 ±  0.56fg	4.30
LL	65	72	97.84 ±  1.01gh	4.20
LP	37	0	100.00 ±  0.00a	6.50
LP	37	24	99.2 ±  0.3abcd	4.20
LP	37	48	99.13 ±  0.52bcd	4.10
LP	37	72	98.86 ±  0.5bcde	4.00
LP	65	0	100.00 ±  0.00a	6.50
LP	65	24	99.58 ±  0.23ab	4.20
LP	65	48	99.58 ±  0.17ab	4.10
LP	65	72	99.17 ±  0.07bcd	4.00
LL and LP	37	0	100.00 ±  0.00a	7.00
LL and LP	37	24	98.83 ±  0.24bcdef	4.00
LL and LP	37	48	98.1 ±  0.26efg	4.10
LL and LP	37	72	97.24 ±  1.14h	4.10
LL and LP	65	0	100.00 ±  0.00a	7.00
LL and LP	65	24	99.47 ±  0.29abc	4.00
LL and LP	65	48	98.75 ±  0.26cdef	4.10
LL and LP	65	72	98.56 ±  0.29defg	4.10

Full text

INTRODUCTION

With the acceleration of the pace of life, insomnia and anxiety have become increasingly prevalent, imposing a burden on public health.1 Sleep disturbances are implicated as a key contributor to the pathophysiological mechanisms underlying mental disorders.2 However, medications for insomnia and anxiety often cause relatively serious harm to the body.3, 4 Therefore, health regulation through dietary intervention has emerged as a safer alternative.5, 6, 7

Among the bioactive substances that support such dietary intervention, γ‐aminobutyric acid (GABA) stands out due to its prominent regulatory roles. As a major inhibitory neurotransmitter, GABA exerts significant effects on sleep promotion, memory enhancement, anxiety relief, blood glucose regulation, and cancer cell inhibition.8, 9, 10 Although the GABA supplemented through ingestion is difficult to pass through the blood–brain barrier under normal circumstances,11 studies show that the GABA supplemented in the gastrointestinal tract can improve the health status of the nervous system through the gut microbiota–gut–brain axis.12 Currently, in the production of GABA, microbial fermentation has emerged as the optimal strategy for the preparation of food‐grade GABA due to its simplicity of operation, safety, and low cost.10, 13 The creation of GABA‐enriched functional foods via microbial fermentation has emerged as a pivotal research focus in food science and nutritional health disciplines.

Soybeans are one of the world's major food crops, and their annual output exceeds 350 million tons.14 Meanwhile, with a high glutamic acid content of up to 7.58 g 100 g^−1^, soybeans serve as an excellent substrate for the fermentative synthesis of GABA.10 Soybeans are rich in high‐quality proteins and amino acids. Incorporating fermented soybeans into cereal‐based foods can effectively enhance the nutritional value of cereal foods.15, 16 With the advancement of bio‐fermentation technology, the development prospects of fermented soybean products are becoming increasingly promising.17

Traditional cookies have relatively limited nutritional profiles. Specifically, they lack functional active ingredients and high‐quality proteins. To address this limitation, lactic acid bacteria‐fermented soybeans can be incorporated into cookie production. This modification enhances the nutritional value of cookies, as fermented soybeans provide high‐quality plant‐based proteins, GABA, and dietary fiber. It not only boosts the product's competitiveness in the market but also contributes to the upgrading of the traditional baking industry. However, the specific effects of Lactococcus lactis (LL) and Lactiplantibacillus plantarum (LP) on soybean fermentation remain to be explored.8, 17, 18, 19

Over recent years, the utilization of beneficial microorganisms for fermenting soybeans or okara to improve the nutritional composition of food raw materials and facilitate their integration into food processing has emerged as a prominent research focus.20 For example, Zhang et al. improved GABA production in okara using Kluyveromyces marxianus.21 Additionally, GABA‐enriched soy sauce was fabricated through co‐fermentation with a multi‐strain starter culture in another study.22 Wang et al. increased GABA content in soybean milk through fermentation with Lactiplantibacillus plantarum Lp3.13 In a randomized controlled crossover trial, Lee et al. demonstrated that consuming biscuits supplemented with fermented okara significantly increased circulating short‐chain fatty acid levels in human blood.23 However, the effects of single fermentation using LL, single fermentation using LP, and co‐fermentation with LL and LP on the GABA content, phytic acid level, and antioxidant capacity of soybean flours and the GABA content of cookies were unclear.

Therefore, soybeans were fermented by LL, LP, and their co‐culture in this investigation. The impacts of these fermentations on soybean phytic acid content, total phenolics, ABTS^+^ radical scavenging capacity, and GABA accumulation were systematically evaluated. Subsequently, the fermented soybean flour was added into cookies, and the impacts of adding fermented soybean flour on the spread factor, hardness, and GABA content of cookies were further investigated. This research is expected to provide useful insights for the development of novel fully plant‐based foods.

MATERIALS AND METHODS

Reagents and materials

Lactococcus lactis and Lactiplantibacillus plantarum were purchased from the American Type Culture Collection (ATCC). Yeast extract and Trolox (6‐hydroxy‐2,5,7,8‐tetramethylchroman‐2‐carboxylic acid) were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Peptone from potatoes was obtained from Sigma‐Aldrich (St Louis, MO, USA). Glucose was procured from Aldon Corporation (Avon, NY, USA). Soybeans, sugar, granulated brown sugar, non‐fat dry milk, salt, sodium bicarbonate and other food grade materials were purchased from Costco (East Lansing, MI, USA). Phytic acid assay kit was purchased from Megazyme Company (Chicago, IL, USA). GABA standard was purchased from Acros Organics (Geel, Belgium).

Pre‐treatment of soybeans

The dried soybeans were crushed and then ground and sieved using a Thomas Wiley Laboratory Mill Model 4 (Thomas Scientific, Swedesboro, NJ, USA) with a pore size of 3 mm or less before being stored for future use.24

Fermentation process

Yeast extract–peptone–dextrose medium

The yeast extract*–*peptone–dextrose (YPD) medium consisted of 10 g L^−1^ yeast extract, 20 g L^−1^ peptone, and 20 g L^−1^ glucose. To prepare the medium, 20 g peptone and 10 g yeast extract were first mixed in 950 mL distilled water, while 20 g glucose was dissolved separately in 50 mL distilled water. Both solutions were then sterilized at 121 °C for 15 min using a Thermo Scientific ST75925 SterileMax benchtop steam sterilizer (Thermo Fisher Scientific, MA, USA). After sterilization of the glucose solution, it was aseptically combined with the yeast extract and peptone solution in a biosafety cabinet. The complete YPD medium was thus obtained.

Strain activation

Lactococcus lactis and Lactiplantibacillus plantarum were routinely maintained on YPD agar plates at 4 °C. Under aseptic conditions, single colonies were inoculated into fresh 10 mL YPD broth for activation. Every 24 h, their morphology was examined using a microscope, and potential contamination was assessed via plating and microscope. Their pH was measured using pH test strips. Subsequently, the cultures were subcultured, and the third‐generation strains were used as inocula for sample fermentation.25, 26

Sterilization of soybeans and soybean

Soybeans (500 g) were combined with 1500 mL distilled water in beakers. The beakers were hermetically sealed with aluminum foil and subjected to autoclaving at 121 °C for 15 min using a Thermo Scientific ST75925 SterileMax benchtop steam sterilizer. The autoclaving process was repeated twice.

Fermentation, drying, and grinding

Under aseptic conditions, 200 mL of the third‐generation25, 26 activated bacterial culture was introduced into each beaker containing sterilized soybeans. The upper opening of the beaker was sealed with plastic wrap and aluminum foil. The fermentation process was conducted at 35 °C for 24, 48, and 72 h.27, 28, 29, 30 Following fermentation, the samples were subjected to drying at either 65 or 37 °C for 48 h. Finally, the dried samples were pulverized and then sieved using a Thomas Wiley laboratory equipped with a 40‐mesh sieve.31 Following this processing step, the treated sample was preserved in a 4 °C refrigerator for use in subsequent experimental analyses.

Determination of yield, pH value, and colony‐forming unit

The yield was determined by the weighing method and calculated according to the formula presented as Eqn (1). The pH value was measured using pH test strips, and colony‐forming unit (CFU) was determined by the plate count method:

$$[eqn]$$

where WP is dry weight of fermented product and WB is dry weight of sample before fermentation.

Determination of phytic acid content

Phytic acid levels were quantified via the Megazyme phytic acid assay kit (Megazyme Ltd, Wicklow, Ireland). Precisely 1.0 g of each sample was weighed into a 50 mL conical flask, followed by the addition of 20 mL of 0.66 mol L^−1^ hydrochloric acid. The mixture was then agitated vigorously on an orbital shaker at ambient temperature for 3 h. Following extraction, 1.0 mL of the homogenate was transferred to a 1.5 mL microcentrifuge tube and centrifuged at 13 000 × g for 10 min. Subsequently, 0.5 mL of the supernatant was pipetted into a new tube and neutralized with 0.5 mL of 0.75 mol L^−1^ sodium hydroxide solution. Reaction reagents were then added according to the kit's protocol. Finally, absorbance values were measured at 655 nm using a spectrophotometer.32

Total phenolic content analysis

Total phenolic content (TPC) was quantified using the Folin–Ciocâlteu colorimetric assay.33 Precisely 0.1 g of each sample was weighed into a 1.5 mL microcentrifuge tube, and 1 mL distilled water was added. The mixture was vortexed for 30 s to ensure homogeneity. Following centrifugation at 3000 × g for 10 min, 50 μL of the supernatant was carefully transferred to a new tube. Subsequently, 25 μL Folin–Ciocâlteu reagent, 400 μL of 0.071 g mL^−1^ sodium carbonate solution, and 425 μL deionized water were added in sequence, with thorough mixing via vortexing after each addition. Absorbance was recorded at 760 nm using a spectrophotometer. A standard calibration curve was established using gallic acid solutions, and TPC was expressed as micrograms of gallic acid equivalents per gram of sample.

Determination of ABTS

• radical clearing capacity

The ABTS^+^ radical scavenging capacity of fermented soybeans under different fermentation protocols was determined according to published methods.34, 35 A volume of 100 mL of 7 mmol L^−1^ ABTS^+^ radical solution was combined with an equivalent 100 mL of 2.5 mmol L^−1^ K_2_S_2_O_8_. The mixture was kept in the dark for 12–16 h before use, yielding the ABTS^+^ radical stock solution. A portion of the ABTS^+^ radical stock solution was diluted to achieve an absorbance of 0.7 at 732 nm, as per the experimental requirements, to prepare the ABTS^+^ working solution. The absorbance was adjusted to be as close to 0.7 as possible, and the actual absorbance was recorded as A 1. Then, 0.8 g of each sample was accurately weighed into a 1.5 mL microcentrifuge tube, and 10 mL deionized water was added. The mixture was vortexed for 30 s to ensure homogeneity. Following centrifugation at 3000 × g for 10 min, 10 μL of the supernatant was transferred to a new tube containing 990 μL ABTS^+^ working solution. The reaction mixture was thoroughly vortexed and incubated in the dark for 30 min. Subsequently, the absorbance (A 2) of the sample was measured. A calibration curve was established using Trolox as the standard, with Trolox concentration plotted on the abscissa and the value of A 1 − A 2 on the ordinate. The Trolox equivalent of the ABTS^+^ scavenging capacity of the sample was calculated based on the standard curve.

Quantification of GABA content

The GABA content in soybeans fermented by different schemes was determined with reference to the published methods.21 1.0 g of each sample (particle size < 100 μm) was suspended in 10 mL deionized water and agitated at ambient temperature for 10 min. Following incubation, the suspension was centrifuged at 13 000 × g for 10 min. The resultant supernatant was then decanted into a fresh microcentrifuge tube. To the supernatant, 200 μL was introduced into a solution containing 400 μL of 0.1 mol L^−1^ borate buffer (pH 9, prepared by adjusting 0.1 mol L^−1^ boric acid with sodium hydroxide), 200 μL of 6% (w/v) phenol, and 1 mL of 10% (v/v) sodium hypochlorite. The mixture was left to stand at room temperature for approximately 3 min, followed by incubation in a 98 °C water bath for 10 min. Immediately following this step, the solution was quenched in an ice bath for 10 min. Subsequently, 4 mL of 60% (v/v) ethanol was added to each sample, and the absorbance was measured at 645 nm. Quantification was performed using a standard calibration curve, with GABA standards prepared at concentrations of 0, 0.05, 0.1, 0.2, 0.4, 0.6, and 0.8 g L^−1^.

Production process of cookies

Cookie preparation followed the procedure outlined in AACC Method 10‐54,36 with minor modifications (Supporting Information, Fig. S1). Fermented soybean flour was employed to partially substitute wheat flour at replacement levels of 25% and 50%. Baking was conducted at 350 and 400 °F, respectively. To maintain the quality of the cookies, baking times were adjusted based on the proportion of fermented soybean flour added and the baking temperature used. The specific baking times are presented in Supporting Information, Table S1.

Determination of diameters, heights, spread indexes, and hardness of cookies

The diameters and heights of the cookies were measured using a vernier caliper. Two cookies were placed together for each measurement, and the result was then divided by two. During the measurement, the vernier caliper was passed through the center of the cookies. After each measurement, the vernier caliper was rotated by 90° and the measurement was repeated. The experiment was replicated four times, and the arithmetic mean was computed. The spread ratio was defined as the quotient of cookie diameter and height. Cookie hardness was evaluated using an FSC431 texture analyzer (Food Technology Corp., Sterling, VA, USA).36 Because texture properties may be affected by product temperature and geometric characteristics, the texture and geometric features of cookies were measured 1 h after baking to allow product cooling. Following the assessment of geometric and textural properties, the cookies were ground and further analyzed.36

Statistical analysis

All experiments were performed in triplicate. Figures were generated using Origin 2018 (OriginLab Corp., Northampton, MA, USA), while statistical analyses were conducted using SPSS 20 (IBM Corp., Armonk, NY, USA). Statistical significance was determined via Duncan's multiple range test, with significance set at P < 0.05.

RESULTS

Effect of fermentation on yield and pH

The changes in yield and pH during the fermentation process are of great significance for monitoring the fermentation process. The yield and pH of soybean after fermentation with LL, LP, and a mixture of LL and LP are shown in Table 1. As the fermentation time extended from 24 to 48 and 72 h, the yields of soybeans fermented by LL, LP, and the mixture of LL and LP all decreased with time, but remained above 97.24%. This experimental phenomenon indicates that the fermentation strategy can achieve high yields. Table 1 also shows the changes in pH during the fermentation process. The pH values of soybeans fermented by LL, LP, and the mixture of LL and LP initially decreased over time and then tended to stabilize. Before fermentation, the soybean treated with LP had the lowest pH value. After 24 h of fermentation, the pH values in groups inoculated with the LL–LP co‐culture and LP monoculture were significantly lower than those in the LL monoculture group. As fermentation proceeded, the differences among the three groups gradually diminished. After 3 days of fermentation, the LP‐treated group had the lowest pH value, followed by the group treated with the mixture of LL and LP, and the LL‐treated group had the highest pH value.

Effect of fermentation on CFU

CFU is an FDA‐recognized indicator for measuring the number of probiotics in fermented foods. The changes in CFU during the fermentation process are shown in Fig. 1. The CFU of soybeans inoculated with LL, LP, and their co‐culture increased over time and stabilized by day 2 of fermentation. The LL‐inoculated group exhibited the highest CFU at day 2. From day 2 to day 3, no significant difference was observed between the LL and co‐culture groups, and both showed higher CFU than the LP‐inoculated group. This suggests that after 1 day (24 h) of fermentation, the samples that had not been dried after fermentation in the LL treatment group contained higher levels of probiotics than the LP group. Live LL bacteria would confer greater probiotic benefits. After hot‐air drying at 37 °C, CFU decreased significantly, but the abundance in each treatment group remained above 8 × 10^6^ CFU g^−1^, indicating that soybeans fermented by LL, LP, and the mixture of LL and LP are probiotic foods with potential.

Colony‐forming units of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. Different lower‐case letters indicate significant differences (P < 0.05).

Influence of fermentation on phytic acid content in soybeans

Phytic acid is an important chemical component in soybeans. It can affect mineral elements; meanwhile, it also possesses functional activities such as anti‐inflammatory and antioxidant properties.37 Therefore, paying attention to the content of phytic acid during the fermentation process is of great significance. Figure 2 illustrates the impact of fermentation on soybean phytic acid levels. Experimental data showed that mono‐fermentation with LL did not result in a significant reduction of phytic acid in soybeans. When using combined fermentation of LL and LP, significant decreases in phytic acid content were observed on both the second and third days of fermentation, regardless of whether drying at 37 or 65 °C was applied. Overall, fermentation with the mixture of LL and LP was more conducive to reducing phytic acid content in soybeans. Although single use of LL did not exhibit degradation of phytic acid, it appeared to have the ability to enhance the phytic acid‐degrading effect of LP. For soybeans fermented with LP alone or the LL–LP co‐culture, phytic acid content in soybeans exhibited a decreasing trend as fermentation time extended, irrespective of drying at 37 or 65 °C.

Phytic acid content of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. An asterisk indicates significant difference compared with the soybean control group (P < 0.05).

Effects of fermentation on TPC in soybeans

The TPC largely reflects the antioxidant activity of fermented soybeans. Therefore, the TPC of soybeans during the fermentation process was determined. Variations in total phenol levels of soybeans under different fermentation protocols are presented in Fig. 3(A). The raw soybeans and the control group exhibited higher total phenolic levels, primarily due to the rich phenolic antioxidants naturally present in unprocessed soybeans. The TPC levels of soybeans fermented with LL, LP, and the mixture of LL and LP showed no significant changes with the prolongation of fermentation time (P > 0.05). Notably, the TPC in LL‐inoculated soybeans significantly exceeded that in the LP‐inoculated group and LL–LP co‐culture group (P < 0.05).

Total phenolic content (A), and ABTS+ radical clearing capacity (B) of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. Different lower‐case letters indicate significant differences (P < 0.05). GAE, gallic acid equivalents; TE, Trolox equivalents.

Effects of fermentation on ABTS

• radical scavenging activity

Merely determining the TPC is insufficient to directly characterize the antioxidant capacity of post‐fermentation soybean. Therefore, the ABTS^+^ free radical scavenging rate of the fermented soybean was measured. The ABTS^+^ radical scavenging capacity of soybeans under three distinct fermentation protocols is depicted in Fig. 3(B). ABTS^+^ radical scavenging capacity represents a classic assay for evaluating the antioxidant potential of compounds, foods, or other samples.38 Experimental data indicated that unprocessed soybeans displayed stronger ABTS^+^ radical scavenging ability. This phenomenon can be ascribed to the abundant heat‐labile antioxidants in soybean materials, including soy isoflavones and soybean saponins.39 At the starting point of fermentation, the higher ABTS^+^ radical scavenging capacity in the control group was associated with the rich content of antioxidants in the seed medium. Both peptone and yeast extract in YPD medium contain abundant components with antioxidant activity.40, 41 The reduction in radical scavenging capacity during fermentation, when compared to unprocessed soybeans, can be attributed to the decrease in total phenols. Additionally, experimental data revealed that the ABTS^+^ radical scavenging ability of soybeans fermented with LL, LP, and their co‐culture remained relatively stable throughout the fermentation period from day 1 to day 3 (P > 0.05). However, the LL‐inoculated group demonstrated significantly higher ABTS^+^ radical scavenging capacity than both the LP group and the LL–LP co‐culture group (P < 0.05). This finding aligned with the trend in TPC across treatment groups.

Effects of fermentation on GABA

GABA content is one of the most nutritionally distinctive functional indicators of soybean fermented by LL and LP GABA exhibits abundant biological activities and plays a crucial role in improving sleep, mood, and mental health. Therefore, the GABA contents in fermented soybeans were determined. Figure 4 illustrates the influence of fermentation on soybean GABA levels. The GABA content in soybeans fermented with LL, LP, and their co‐culture increased over time, demonstrating significantly higher values than those in unprocessed soybeans. Relatively high levels of GABA were also detected at the beginning of fermentation, which can be attributed to the production of GABA by LL, LP, and their mixture in the seed medium. Following comparison of the three fermentation groups, the LL‐inoculated soybeans exhibited significantly higher GABA levels than both the LP‐inoculated and LL–LP co‐culture groups. Importantly, this advantage persisted during the drying process of the samples, indicating that LL‐fermented soybeans are a rich source of GABA.

γ‐Aminobutyric acid (GABA) content of soybean after fermentation with Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP. Different lower‐case letters indicate significant differences (P < 0.05).

Effect of adding soybean flour fermented by LL on the height, diameter, and spread factor of cookies

Based on the experimental results of antioxidant activity and GABA content, soybeans fermented by LL alone exhibited higher total phenolic acid content, better antioxidant activity, and greater GABA production compared to LP alone or LL combined with LP. Therefore, soybean flour fermented by LL (3 days) was selected for application in cookie processing. The height, diameter, and spread factor are key indicators for evaluating the quality of cookies after baking.36, 42, 43, 44, 45 As shown in Fig. 5(A), with the increase in the addition amount of LL‐fermented and unfermented soybean flour, the diameter of cookies showed a significant expansion (P < 0.05). At the same addition level, the height of cookies made with fermented soybean flour was higher than that of cookies made with unfermented soybean flour (Fig. 5(B)). The cookie spread factor exhibited a significant increase (P < 0.05) with higher inclusion of LL‐fermented and unfermented soybean flour (Fig. 5(C)). When the addition level of soybean flour fermented by LL was 25% and the baking temperature was 35 °F, the spread factor of the resulting cookies was most similar to that of cookies made with 100% wheat flour.

Effect of the addition of soybeans fermented by Lactococcus lactis (LL), Lactiplantibacillus plantarum (LP), and a mixture of LL and LP on the diameter (A), height (B), spread index (C) and hardness (D) of cookies. Different lower‐case letters indicate significant differences (P < 0.05).

Effect of adding soybean flour fermented by LL on the hardness of cookies

Hardness is another key indicator affecting cookie quality. The addition of untreated soybeans often causes an excessive decrease in cookie hardness. Therefore, we aimed to explore whether fermenting soybean with LL could minimize this impact. Figure 5(D) depicts the influence of incorporating different soybean flours on cookie hardness. Experimental results demonstrate that the inclusion of both LL‐fermented soybean flour and unfermented soybean flour led to a reduction in cookie hardness. For the fermented soybean flour, at a constant baking temperature, cookie hardness decreased with increasing addition levels of the fermented soybean flour. For unfermented soybean flour, cookie hardness similarly decreased as its addition level increased when baked at 400 °F. In contrast, no significant variation in cookie hardness was detected when the baking temperature was 350 °F. Notably, when the addition level was 25% and the baking temperature was 350 °F, the hardness of cookies prepared with fermented soybean flour was significantly higher than that of cookies made with unfermented soybean flour, and was more similar to that of cookies made with 100% wheat flour.

Effect of adding soybean flour fermented by LL on GABA content in cookies

Finally, whether the addition of fermented soybean actually increases the GABA content in cookies still requires verification after the cookie manufacturing process. Therefore, the GABA content in cookies supplemented with fermented soybean was determined. Figure 6 illustrates the influence of soybean flour inclusion on GABA levels in cookies. Experimental data indicate that incorporating fermented soybean flour by LL significantly elevated GABA levels in cookies. With higher soybean flour inclusion, GABA levels in cookies showed a progressive increase.

Effect of addition of fermented soybean on γ‐aminobutyric acid (GABA) content of cookies. Different lower‐case letters indicate significant differences (P < 0.05).

DISCUSSION

The development of probiotic‐fermented soybean‐based functional foods has gained increasing attention in recent years, driven by the growing demand for plant‐derived products with enhanced GABA content and health benefits.12, 14, 46 However, optimizing fermentation strategies to improve GABA content and ensuring compatibility with food processing remains a key challenge. Against this backdrop, in this study, LL single fermentation, LP single fermentation, and mixed fermentation of LL and LP were employed to ferment soybeans. The effects of fermentation on GABA accumulation, phytic acid content, TPC, and antioxidant activity of soybeans were evaluated. Subsequently, the fermented soybean flour was added to the cookie formula, and its effects on the GABA content, spread factor, and hardness of cookies were investigated, aiming to provide a theoretical basis for the development of full plant‐based functional cookies with high GABA content.

First, the fermentation process was monitored, including the yield of fermented soybean products, as well as the pH value and CFU during fermentation. The yield of fermented products was high (>97.24%), indicating its promising potential to exhibit advantages in the industrial production of related products. The pH value during fermentation showed a trend of decreasing first and then stabilizing. The decrease in pH is attributed to organic acids by LL and LP during fermentation.47, 48, 49, 50, 51 Microorganisms exhibit a finite tolerance to low pH;51 therefore, the number of CFU gradually stabilizes as fermentation proceeds.

The number of viable bacteria is a crucial evaluation index for probiotic products. The health benefits of consuming foods rich in live probiotics have been widely documented.27, 52, 53 During food processing, better survival rates are conducive to the processing procedures. International organizations such as the FAO/WHO stipulate that probiotic foods should maintain a viable count of at least 1 × 10^6^ CFU g^−1^.27 The experimental results demonstrated that both the LL single‐fermentation group and the LL–LP co‐culture group retained relatively high CFU counts before and after drying, while the drying process induced a reduction in CFU numbers, but the count remained above 8 × 10^6^ CFU g^−1^ (dry weight basis), thereby conferring favorable probiotic properties to the product.27 After drying, the LL treatment group exhibited the highest bacterial survival rate, indicating that the LL strain was superior to LP and the LL–LP mixture in maintaining probiotic viability.

Phytic acid, a typical antinutrient in soybeans, forms insoluble chelates by binding divalent metal ions, including calcium, iron, and zinc. This interaction significantly decreases mineral bioavailability, thereby impairing human nutrient absorption.54 In this study, it was observed that LP fermentation displayed stronger phytic acid‐reducing activity than LL fermentation, but LL fermentation in turn enhanced the phytic acid‐reducing activity of LP fermentation. These results reveal that the LL–LP co‐culture group exhibits higher efficiency in phytic acid reduction in soybeans. Specifically, single‐strain fermentation shows limited efficacy in phytic acid reduction, whereas the synergy between the two strains significantly improves phytic acid degradation efficiency through the combined metabolic effects and complementary enzyme system mechanisms. The phytic acid‐degrading capability of LP has been documented in previous investigations. Srivastava et al. found that LP fermentation significantly decreased phytic acid content in millet.55 Additionally, research has shown that non‐recombinant LL exhibited no detectable phytase activity.56 Although LL is less effective than LP in reducing phytic acid activity, maintaining appropriate levels of phytic acid in foods may not necessarily be detrimental to enhancing overall food nutritional value. Notably, phytic acid does not exhibit only negative effects.57 Phytic acid also exerts positive effects in scavenging free radicals, preventing type 2 diabetes, exhibiting anti‐inflammatory properties, and demonstrating anticancer activities.37 Balancing the beneficial properties and antinutrient effects of phytic acid remains a key challenge in the food industry.37 Therefore, although the effect of LL on reducing phytic acid activity is not as significant as that of LP, maintaining an appropriate level of phytic acid in food may not necessarily have an adverse impact on improving the overall nutritional value of food.

Phenolic compounds typically demonstrate remarkable antioxidant activities, enabling them to scavenge free radicals and mitigate oxidation reactions.58 An increase in TPC is generally associated with stronger antioxidant capacity. Additionally, phenolic compounds also exhibit anti‐inflammatory and immunomodulatory, cardiovascular protective, antitumor, and microbiota‐regulating activities.59 The experimental results showed that, under different fermentation schemes, the total phenolic levels in the raw soybean and control groups were higher than those in the LL and LP fermentation groups. This is because soybeans are rich in phenolic substances,60 and studies have shown that the fermentation process of LP can lead to a decrease in TPC and antioxidant activity. LP has been demonstrated to possess the ability to transform polyphenols;61 another study also found that the antioxidant capacity of skim milk decreased after fermentation with LP for more than 16 h.62 After fermentation, the TPC of soybeans inoculated with LL was significantly higher than that of soybeans inoculated with LP and those in the LL–LP co‐culture group (P < 0.05). The experimental results showed that single fermentation with LL could effectively maintain the TPC in soybeans without reduction, thereby being more conducive to preserving the antioxidant properties and nutritional performance of fermented soybean products. The ABTS^+^ free radical scavenging assay further verified the regularity results of changes in total phenolic content.

Subsequently, we focused on the GABA‐enriching capacity of different fermentation protocols. Our study found that LL fermentation demonstrated the highest GABA‐enriching activity. The stronger GABA production by LL than LP might be attributed to the absence of vitamin B_6_ (a coenzyme for glutamic acid decarboxylase (GAD)) in the experiment; LP could be limited by its dependence on exogenous B_6_, whereas LL could synthesize sufficient B_6_ through its own metabolism, with its GAD enzyme having a lower requirement for B_6_. Additionally, the rapid acid production by LP causes a sharp pH drop, which may inhibit the GAD activity of both LL and LP. The combined effect of these possibilities might result in LL‐fermented soybean flour having the highest GABA content.

The increase in GABA content during fermentation suggests that soybean flour fermented by LL can be applied to cookie processing to improve the nutritional value of cookies. Interest in developing nutritionally enhanced products is growing globally. Additionally, the scientific community and food industries aim to promote the consumption of legume‐based products. However, incorporating high levels of fiber powders without altering the textural properties of final products remains challenging.36 Therefore, we applied the soybean flour fermented by LL to the processing of cookies.

The spread factor reflects the shape retention capability of cookies after baking.36, 42, 43, 44, 45 During baking, the melting of fats and dissolution of sucrose promote the expansion of dough, thereby increasing the spread factor.45 Under general conditions, a lower viscosity of the dough corresponds to a larger spread factor.44 A lower spread factor typically indicates that the cookies are more elastic, whereas a higher spread factor renders the cookies crispier when baking.36, 42, 43, 44, 45 On the whole, the addition of soybean flour increased the spread factor of the cookies, which implies that the incorporation of soybean flour made the cookies more prone to breaking – this is also confirmed by the decrease in their hardness. A comparison between unfermented soybean flour and fermented soybean flour revealed that when the addition level was 25% and the baking temperature was 350 °F, the soybean flour fermented by LL better maintained the hardness of 100% wheat flour cookies. The experimental results indicate that fermented soybean flour is more suitable for cookie processing than unfermented soybean flour.

In addition to the effects of fermented soybean flour on the height, diameter, spread factor, and hardness of cookies, we focused more on the effect of fermented soybean flour addition on the GABA content of cookies with increasing amounts of soybean flour. The experimental results showed that the soybean flour fermented by LL significantly increased the GABA content in cookies, and the GABA level in cookies showed a gradual upward trend with the increase in the amount of fermented soybean flour added. This further confirms that fermentation with LL contributed to enhancing the nutritional properties of soybean flour for cookie production.

Overall, the data demonstrate that soybean flour fermented by LL is more suitable for incorporation into cookies than unfermented soybean flour. It can maintain the texture and hardness of cookies and, more importantly, significantly increases the GABA content in cookies.

CONCLUSIONS

Co‐fermentation with LL and LP exhibited potent efficacy in reducing phytic acid content in soybeans. In contrast, single‐strain fermentation with LL demonstrated greater advantages in maintaining probiotic viable counts, preserving total phenolic content and antioxidant activity, and promoting GABA accumulation. Through cookie‐baking experiments, we further confirmed that cookies prepared with LL‐fermented soybean flour showed significantly improved processing characteristics compared to those added with unfermented soybean flour, accompanied by a notable increase in GABA content. This study provides empirical data for the innovative development of fully plant‐based foods and the exploitation of fermented soybean products. In future research, we will endeavor to further enhance the GABA‐enriching capacity of LL via genetic engineering approaches and explore the in vivo health benefits of the GABA‐enhanced cookies.

AUTHOR CONTRIBUTIONS

Xin Chen: investigation, writing – original draft, data curation, writing – review and editing. Zhenjia Chen: investigation, formal analysis, validation. Perry K. W. Ng: validation and supervision. Yan Liu: conceptualization, writing – review and editing, supervision, project administration, funding acquisition.

FUNDING INFORMATION

Funding for this study was provided by the Michigan Soybean Committee (St Johns, MI).

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest regarding the publication of this paper. All authors approved the paper.

Supporting information

Table S1. Baking temperature and time of cookies with different amounts of soybean flour added. Figure S1. Process flow for the production of cookies with the addition of fermented soya beans.