The commercial corn flour was produced from a common yellow corn variety (Zea mays L.) purchased from China Neihe Tianfeng Flour Co., Ltd., and obtained through grinding and sieving. α-Amylase (4,000 U/g), neutral protease (100 U/mg), cellulase (400 U/mg), compound protease (120 U/mg), papain (800 U/mg), and transglutaminase (TG, 200 U/g) were purchased from Nanning Dongheng Huadao Biotechnology Co., Ltd. (Nanning, China). Lactobacillus HR is a research strain preserved by the Culture Collection Center of the College of Food and Biological Engineering, Qiqihar University (Qiqihar, Heilongjiang, China). All other chemicals and reagents were purchased from Tianjin Kaitong Chemical Reagent Co., Ltd. (Tianjin, China) and were of analytical grade.

Regarding inoculum size (3 × 10⁷–1.5 × 10⁸ CFU/mL), an inoculum of 1.2 × 10⁸ CFU/mL led to a substantial reduction in GI (51.8) (Figure 1A1), and a significant increase in RS content (p < 0.05), attributed to the activity of Lactobacillus HR (Table 1) while the soluble protein content also showed a corresponding change (Figure 1A3). In contrast, a higher inoculum of 1.5 × 10⁸ CFU/mL resulted in overfermentation (Figure 1A2). RS is metabolized through fermentation, which not only leads to a decrease in its content but also produces an unpleasant flavor. Therefore, the inoculum of 1.2 × 10⁸ CFU/mL was more favorable for reducing the GI.

Under the above-mentioned inoculum conditions, drastic starch breakdown occurred at 60 h (Figure 1B1), reducing the GI to approximate value of 51.3 due to indigestible components (Table 1). RS content and soluble protein levels peaked between 48 and 60 h, whereas fermentation beyond 60 h led to a decline in Lactobacillus HR populations, and the RS and soluble proteins might be further hydrolyzed (Figures 1B2,B3).

Moisture content analysis showed that 45% moisture achieved the lowest GI and highest RS levels (Table 1 and Figure 1C2) with “Moisture content analysis showed that starch hydrolysis was influenced by different moisture levels (Figure 1C1). A moisture content of 45% achieved the lowest GI and highest RS levels (Table 1 and Figure 1C2).

Temperature optimization studies found that 33–37 °C supported microbial metabolism and enzymatic activity, decreasing GI, and increasing RS content (Figure 1D1). Temperatures above 37 °C had negative effects on the fermentation process, thereby increasing GI and decreasing RS formation (Table 1 and Figure 1D2). Soluble protein production peaked at about 39 °C before a decline (Figure 1D3).

Comprehensive analysis determined the optimal fermentation parameters as follows: 1.2 × 10⁸ CFU/mL inoculum, 60-h fermentation, 45% moisture, and 37 °C temperature. These conditions balanced bacterial activity, starch modification, and enzyme stability to maximize RS and minimize GI.

Figure 2 shows the effects of six different enzymes (α -amylase, neutral protease, cellulase, complex protease, papain, and TG enzyme) on the starch hydrolysis rate within a reaction time of 180 min. α -amylase exhibits significant hydrolytic activity, with its reaction curve rising rapidly, achieving a starch hydrolysis rate of approximately 35% within 60 min. In contrast, the starch hydrolysis rates catalyzed by the other five enzymes remained at an extremely low level (below 5%) throughout the entire reaction process, and their reaction curves were gentle with no obvious upward trend. Neutral protease, cellulase, complex protease, papain, and TGase mainly target cereal proteins. Moderate proteolysis might improve the texture of food, while slowing down the enzymatic digestion rate of starch in the food matrix through synergistic means such as physical embedding, structural reorganization, and cross-linking with dietary fiber.

Further enzyme screening is carried out based on an optimal fermentation process for selecting enzymes that have a positive impact on GI value when combined with Lactobacillus HR (Table 2). Different enzyme treatments have significant effects on the hydrolysis index (HI) and GI of corn flour. α -amylase and neutral protease significantly increased HI and GI values, indicating that they promote starch digestion. Cellulase and complex protease also cause a slight increase in the GI value. It is worth noting that papain and TGase significantly reduced HI and GI values. The mechanism may lie in their action on protein components, delaying digestion by altering the state of starch coating or forming a protein network. Moreover, although there is no significant difference between papain and TGase at the overall level (p > 0.05), TGase has a relatively low HI and GI index. This indicates that TGases are enzymes with the potential to reduce the GI of corn flour. However, during the fermentation of corn flour, the pH value decreased (37), which likely inhibited the activity of neutral protease and composite protease. This also indicates that these two enzymes are not suitable for co-fermentation with acid-producing bacterial strains.

Regarding the inoculum amount, the glycemic index (GI) of corn flour initially decreased and subsequently increased with varying inoculum levels (Table 3). In contrast, the content of RS and soluble protein contents showed an inverse trend (Figures 3A2,A3). An intermediate inoculum of 9 × 10⁷ CFU/mL was identified as optimal for maximizing microorganism-enzyme synergy and was selected for subsequent experiments. Exceeding this threshold led to nutrient insufficiency per cell, undermining synergy and reducing fermentation efficiency.

With respect to enzyme dosage, the observed trend was analogous to inoculum effects: GI initially decreased and then increased with rising enzyme concentrations, while RS contents first increased and then declined. In contrast, soluble protein contents rapidly increased initially before stabilizing at higher doses (Figures 3B2,B3 and Table 3). Excessive enzyme addition beyond saturation caused substrate saturation, which interfered with microbial proliferation and metabolism, disrupted the enzyme–microbe balance, and impaired product quality and yield.

With respect to temperature variations, GI decreased initially and then increased as temperature rose, while RS and soluble protein concentrations showed opposite trends (Figures 3C2,C3 and Table 3). Temperatures exceeding the physiological optimum reduce microbial viability and enzymatic activity and, in some cases, cause complete thermal inactivation, alter metabolic by-product profiles, and affect starch digestibility (Figure 3C1) (38). A comprehensive analysis determined that 37 °C is the optimal temperature for maintaining biological activity and product quality.

As for the time course, GI first increased and then decreased over time, while RS and soluble protein contents showed opposite trends (Figures 3D2,D3 and Table 3). The 48-h mark was identified as the optimal processing window, achieving minimum GI, maximum RS accumulation, and high soluble protein levels simultaneously. This duration is 12 h shorter than conventional single-strain fermentation, establishing it as the most efficient co-processing duration that balances product quality and economic feasibility.

The optimal results of enzyme dosage, inoculum level, fermentation time, and fermentation temperature obtained from single-factor experiments were used to establish the level ranges for each factor in subsequent optimization. The fermentation process was optimized with GI value as the response variable. The results are shown in Tables 4, 5.

Using the experimental data, Design-Expert 3.7 was utilized to perform multiple regression fitting. A quadratic multiple regression equation was established to describe the relationship between the influencing factors—enzyme addition amount, inoculation amount, time, temperature, and the GI value. The general form of the regression equation is:

Y = 48.71–0.1242 × A-0.3883 × B-0.1533 × C + 0.2075 × D-0.1925 × A × B-0.0475 × A × C + 0.1375 × A × D-0.0175 × B × C-0.2200 × B × D + 0.2350 × C × D + 1.31 × A2 + 0.3390 × B2 + 0.7715 × C2 + 0.4752 × D2.

where Y represents the GI value, A represents the amount of enzyme added, B represents the number of bacteria inoculated, C represents the time, and D represents the temperature.

With GI as the response variable, the quadratic regression model was extremely significant (p < 0.01) (Table 5). The lack-of-fit terms of the regression equation showed no statistical significance (p = 0.0781), indicating that unknown factors had minimal interference with the experimental results, as p > 0.05. The coefficient of determination (R² = 0.9823) and the adjustment coefficient of determination R²_Adj = 0.9586 indicate that the GI value of corn flour processed by the co-treatment with bacteria and enzymes primarily depends on the amount of enzyme dosage, inoculation amount, fermentation time, and fermentation temperature. The signal-to-noise ratio (S/N) = 13.8982 > 4 also indirectly confirms that this model has a certain degree of credibility.

Through the analysis of variance of the regression equation, the influence degrees of the four factors on the GI value were B > D > C > A, that is, the inoculation amount > temperature > time > enzyme addition amount. Among these, factor B (inoculation amount) had an extremely significant impact on the result (p < 0.01), while factors C (time) and D (temperature) had significant impacts on the result (p < 0.05). Factor A (enzyme dosage) showed no significant impact on the result (p > 0.05). Interaction terms such as AB, AC, AD, BC, BD, and CD had no significant influence on the results (p > 0.05), whereas all quadratic terms had an extremely significant impact on the results (p < 0.01). The 3D curve of the response surface of the interaction of enzyme dosage, inoculation amount, temperature, and time on the GI value, as determined by the regression equation, is shown in Figure 4.

At a fixed time of 48 h and temperature of 37 °C, the GI value shows a trend of first decreasing and then increasing with the increase in the inoculation amount and the enzyme dosages (Figure 4A). When the amount of enzyme added is 3‰ (90 U/g), the GI value gradually decreases as the inoculation amount increases. The lowest GI value was observed when the enzyme dosage was 2.581‰ (77.43 U/g), and the inoculation amount was 1.03 × 10⁸ CFU/mL. The elliptical contour lines and concave response surface indicate that the interaction between these two factors was statistically significant and yielded a clear minimum value.

As shown in Figure 4B, the value of GI decreased with the increase in fermentation time at a fixed inoculum size of 9 × 10⁷ CFU/mL, temperature of 37 °C, and the enzyme dosage of 3‰, but began to increase after reaching the threshold. This might be due to the prolonged fermentation, leading to the depletion of nutrients in the substrate, which was not conducive to bacterial growth and metabolic activities. The interaction between enzyme dosage and fermentation time was significant, and the lowest GI value was observed at an enzyme dosage of 2.581‰ and a fermentation time of 49.251 h, representing the valley point of the response surface.

Figure 4C shows that when the inoculum size was fixed at 9 × 10⁷ CFU/mL, and the enzyme dosage was 3‰, the GI value showed a trend of first decreasing and then increasing with an increase in temperature and time. At a fixed temperature of 37 °C, the GI value gradually decreased with an increase in time. The lowest GI value was obtained when the time was 49.25, and the temperature was 36.83 °C. The inclination angle of the time surface was smaller than that of the temperature surface. Therefore, it is inferred that the influence of temperature on the results is greater than that of time.

At a fixed inoculation size of 9 × 10⁷ CFU/mL and fermentation time of 48 h, the GI value exhibited a trend of first decreasing and then increasing with the increase of enzyme dosages and temperature, as shown in Figure 4D. The lowest GI value was observed when the enzyme dosage was 2.58‰, and the temperature was 36.83 °C, representing the valley point of the response surface. The elliptical contour map and the presence of concave points suggest that there is the strongest interaction between enzyme dosage and temperature among the tested factor combinations, exerting a significant influence on GI reduction.

The optimal fermentation process conditions obtained by the response surface experiment were: enzyme dosage of 2.581‰ (77.43 U/g), fermentation time of 49.25 h, inoculation amount of 1.03 × 10⁸ CFU/mL, and temperature of 36.831 °C. Under these conditions, the predicted value of GI was 48.91. Fermentation was carried out under these conditions, and 3 to 5 groups of parallel experiments were conducted. The final result was 48.87, which was close enough to the model prediction value. Therefore, the regression model can effectively reduce the GI value. Therefore, the co-fermentation of bacteria and enzymes adopted in this study is a comprehensive bioprocessing strategy. Beyond achieving GI reduction, this method is more likely to simultaneously improve the flavor and nutritional quality of the product, offering a dual benefit compared to single-mechanism interventions.

The effects of various treatments on the functional properties and macroscopic structure of corn flour are comprehensively shown in Figure 5 (CP: control; HCP: Lactobacillus HR fermentation; TCP: TGase treatment; THCP: Lactobacillus HR combined with TGase). As shown in Figure 5, the water-holding capacity of corn flour treated by the synergy of Lactobacillus HR fermentation and TGase modification decreased by 10.3% (p < 0.05), and the gelation force increased by 29.7% (p < 0.05). Notably, the TCP group exhibited the highest water-holding capacity, significantly surpassing all other groups. This was followed by the control group (CP), the THCP group, and finally the HCP group (Figure 5A). Regarding gel strength, THCP showed the highest value, which was significantly greater than that of TCP, and both enzyme-treated groups outperformed the singly Lactobacillus HR fermented (HCP) and control (CP) samples (Figure 5B).

The change in water-holding capacity and gel properties can be attributed to the protein cross-linking action of TGase, which promotes the formation of a robust three-dimensional network. Lactobacillus HR fermentation may further contribute by moderately hydrolyzing proteins and exposing additional cross-linking sites. Under the combined action of Lactobacillus HR and TGase, the network structure is optimized, leading to the greatest improvement in gel performance. These results clarify the role of TGase in strengthening the water retention and gelling properties of protein matrices and highlight the synergistic potential of combined enzymatic modification.

As summarized in Table 6, regarding mineral content, the ash values remained highly stable across all treatment groups, with a narrow range from 1.09 ± 0.31% to 1.58 ± 0.12%. This consistency suggests that neither bacterial fermentation, TGase enzymatic modification, nor their combined application had a significant impact on the mineral retention capacity of the corn flour. In contrast, the protein content indicated substantial improvements through processing interventions. Both TGase-modified corn flour (5.59 ± 0.53%) and bacteria-enzyme co-processed corn flour (5.51 ± 0.84%) exhibited statistically higher crude protein levels compared to the control group (CP) (4.53 ± 0.89%).

The starch composition analysis yielded notably intriguing findings. The CP group possessed the highest starch content (78.05 ± 2.51%), while the TGase-modified group showed a significant reduction to 71.91 ± 2.73% (p < 0.05). In contrast, the HCP (69.69 ± 2.38%) and THCP (65.19 ± 1.99%) groups demonstrated the lowest starch concentrations. This significant consumption reveals that Lactobacillus HR and enzymatic treatment preferentially utilize readily hydrolysable starch fractions as carbon sources. Correspondingly, dietary fiber content showed the most significant improvement in the THCP group (7.98 ± 1.25%), representing a 60% increase compared to the CP group (5.04 ± 0.76%).

Total sugar analysis revealed consistent reductions across all processed samples, with the bacteria-enzyme co-modification group achieving the lowest level at 23.40 ± 0.04% compared to the CP group (25.16 ± 0.09%).

The above results clearly underscore the nutritional advantages of co-processing corn flour with bacteria and enzymes. The observed reduction in total sugar content effectively lowers the product’s glycemic index (GI). This is achieved not only through the direct delay in glucose release due to a decrease in available sugar content, but also via the elevated crude protein levels, which might physically encapsulate starch granules and interact with them, thereby impeding α-amylase hydrolysis to some extent and further slowing carbohydrate digestion and absorption.

The microscopic morphology of corn particles under magnification of 500×, 1,000×, 2,000×, and 4,000× was observed during bacteria-enzyme co-processing (Figure 6). The particle surfaces of the group CP were smooth, and the distribution was relatively uniform. The particles in group HCP were unevenly distributed, varying in shape and size. The surface was rough, with flocculent substances, depressions, protrusions, and distinct edges and corners. Under the action of Lactobacillus HR, extensive hydrolysis causes a reduction in particle size, producing irregular shapes and, in some cases, fragmented starch granules. TGase-modified corn flour (TCP) exhibited a rough, textured surface with depressions and flocculent substances, showing a phenomenon of particle breakage and creating irregular spherical shapes. Moreover, Figure 6D shows that after the co-treatment of Lactobacillus HR and TGase (THCP), starch particles aggregate and cross-link with proteins, which can effectively prevent the contact between starch molecules and hydrolases, increase the content of RS, and reduce the GI value.

Fourier transform infrared spectroscopy (FTIR) was utilized to analyze the structural changes of functional groups in processed corn flour (Figure 7). No new absorption peaks emerged in the corn flour after modification, indicating that no novel chemical functional groups were formed. There were some spectral changes between 500 and 1,500 cm⁻¹, corresponding to elongation vibrations of C-C, C-OH, and C-H bonds in starch (39). The absorption peak at 2,929 cm⁻¹ is related to the C-H stretching vibration of the system. It became flatter after treatment, indicating starch molecular chain breakage under bacterial-enzyme co-processing. Peaks near 1,047 cm⁻¹ and 1,022 cm⁻¹ reflected the crystalline and non-crystalline regions of starch, respectively, highlighting alterations in molecular order between ordered and disordered structures. Additionally, the absorption peak near 995 cm⁻¹ is caused by the bending vibration of C-OH. Corresponding to the hydrogen bond structure formed among starch molecules, further indicating structural reorganization that may influence digestibility and glycemic response.

The ratio of the absorption peak intensities at 1,047/1,022 cm⁻¹ and 1,022/995 cm⁻¹ serves as an indicator of the ordered structure of starch (40). As shown in Tables 7, a larger 1,047/1,022 cm⁻¹ ratio corresponds to a higher degree of molecular ordering within starch. After the co-treatment of bacteria and enzymes, the ratio of 1,047:1,022 cm⁻¹ indicates an increase in the orderliness of the starch. Consequently, the structurally modified starch exhibits anti-digestive properties, reducing susceptibility to hydrolysis and thereby contributing to the observed decline in GI.

As shown in Table 8, compared to the CP group, the total amino acid content showed an increase after treatment, while a small portion of the amino acid content experienced a slight decrease. Notably, the glycine (Gly) content increased significantly by 57.79% through bacterial-enzyme collaborative processing. Among the essential amino acids, particularly Lysine increased by 38.11% after fermentation, 14.23% after TGase enzymatic hydrolysis, and 12.45% after synergy. The content of Thr increased by 17.53% after fermentation, 3.94% after enzymatic hydrolysis, and 2.27% after bacterial-enzyme synergy. The changes in the contents of Tyr, Phe, Lys, and other amino acids are mainly attributed to the Lactobacillus HR metabolism. The hydrolysis of corn proteins by TGase releases amino acids. These amino acids are subsequently harnessed by Lactobacillus HR to support its metabolic activities and growth. However, the synergistic intervention of Lactobacillus HR and TGase still yielded a significant overall increase in TAA. The increase in lysine, methionine, and alanine contents enhances the nutritional value of the corn flour, while phenylalanine and tryptophan showed slight decreases. Therefore, the impact on the properties of the corn flour is relatively positive. The general increase in content may be attributed to the hydrolysis of proteins into smaller molecules, where peptide bonds in proteins or peptide chains are broken.

The Lactobacillus HR initially hydrolyzes starch and protein, providing more terminal and defect sites for the starch molecular chains. Meanwhile, TGase can indirectly stabilize these starch fragments by promoting the formation of protein cross-linking networks, thereby reorganizing them into a more ordered and compact arrangement. Therefore, the resulting resistant-to-digestion starch is a molecularly ordered and structurally enhanced starch. This starch exhibits higher resistance to enzymatic hydrolysis, reduces sensitivity to enzymatic hydrolysis, and thus contributes to the observed GI reduction.