A processing method for improving the quality and digestive characteristics of sourdough food by three-strain fermentation

Abstract

This invention belongs to the field of precision food processing technology, and provides a processing method for improving the quality and digestibility of sourdough foods through three-strain fermentation. It uses *Lactobacillus plantarum* (… Lactiplantibacillus plantarum Lp100), Dwarf Kazakhstan yeast ( Maudiozyma humilis Mh2) and brewer's yeast ( Saccharomyces cerevisiae Sc4) Inoculate the flour at an inoculation ratio of 1:1:1 or 10:1:1 to prepare sourdough. The inoculation ratio is calculated based on the dry weight of the wheat flour, where 1 represents 10 5 CFU / g; then, the prepared sourdough was mixed with flour and water to prepare dough, and after a second fermentation, sourdough food was prepared. This invention confirms that the community building principles established in a liquid culture system are equally applicable in a real dough system. The initial inoculation ratio, by regulating the microbial community assembly pattern, alters the metabolic network and nutrient transformation pathways, thereby shaping the nutritional quality and digestibility of sourdough food. Different inoculation ratios, by adjusting the metabolic intensity of microorganisms and changes in matrix structure during fermentation, influence the nutrient release characteristics of sourdough steamed buns during digestion to a certain extent.

Description

Technical Field

[0001] This invention belongs to the field of precision food processing technology, specifically relating to a processing method for improving the quality and digestibility of sourdough food through three-strain fermentation. Background Technology

[0002] Within a solid-state sourdough fermentation system, the raw materials exhibit significant spatial heterogeneity, humidity gradients, and differences in oxygen concentration. These factors collectively shape multiple unequal ecological niches. This environmental change may reshape the growth behavior and interaction intensity of strains, thereby affecting metabolic specialization patterns rather than a simple community coexistence state. Compared to homogenized simple liquid culture media, the in-situ environment of sourdough may reconstruct the original competition and synergy of strains, thus generating new community response patterns.

[0003] Community succession is influenced not only by strain interactions, but also by environmental selection, metabolic feedback, and spatial heterogeneity, all of which play crucial roles in community diversity and succession. The mechanisms regulating community assembly observed in simple liquid systems may be altered in complex matrices. Factors such as the physical structure of the gluten network, substrate spatial heterogeneity, and buffering effects in dough environments may regulate microbial growth and metabolic expression. Whether succession patterns based on liquid models still apply in complex and heterogeneous solid dough matrices, and the applicability of in vitro mechanisms to in-situ sourdough systems, warrants further investigation. Summary of the Invention

[0004] This invention aims to explore the influence of different inoculation ratios on the assembly pattern of microbial communities in a complex matrix-controlled environment, and further to investigate how to regulate the metabolic network and nutritional quality formation during the fermentation process of sourdough. It provides theoretical support for the synthesis of microbial communities to regulate the quality of fermented foods and offers a processing method for improving the quality and digestibility of sourdough foods through three-strain fermentation.

[0005] This invention is achieved by the following technical solution: a processing method for improving the quality and digestibility of sourdough foods through three-strain fermentation, utilizing *Lactobacillus plantarum* (… Lactiplantibacillus plantarum Lp100, Dwarf Kazakhstan yeast ( Maudiozyma humilis Mh2 and brewer's yeast ( Saccharomyces cerevisiae Sc4 is inoculated into flour at an inoculation ratio of 1:1:1 or 10:1:1 to prepare sourdough. The inoculation ratio is calculated based on the dry weight of wheat flour, where 1 represents 10. 5 CFU / g; then, the prepared sourdough is mixed with flour and water to prepare dough, and after a second fermentation, sourdough food is prepared.

[0006] The Lp100, Mh2 and Sc4 were inoculated into flour at an inoculation ratio of 10:1:1 to prepare sourdough.

[0007] The *Lactiplantibacillus plantarum* Lp100, *Maudiozyma humilis* Mh2, and *Saccharomyces cerevisiae* Sc4 used in this invention are derived from the literature: (Mechanistic insights into strain-specific lactiplantibacillus-yeast interactions in a wheat-based system: Acidification and metabolite dynamics, Food Research International, Yujuan Yu et al, 2025.11.26).

[0008] Lactiplantibacillus plantarum Lp100 is deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 39266 and deposit date of June 5, 2026.

[0009] Maudiozyma humilis Mh2 is deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 39267.

[0010] Saccharomyces cerevisiae Sc4 is deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 39268.

[0011] Furthermore, the specific method for preparing sourdough is as follows: Lp100, Mh2, and Sc4 are cultured in MRS or YPD liquid medium for 12 hours to obtain precultures, and 5000 g Centrifuge for 10 minutes to collect the precipitated bacterial cells, wash twice with sterile distilled water, and resuspend in sterile water; mix the bacterial suspension with flour and sterile water evenly according to the ratio of flour to water of 2:1 to obtain the initial dough. Ferment the initial dough at 30℃ and 80% humidity for 8-24 hours to obtain the mature sourdough.

[0012] The method for secondary fermentation is as follows: Mix 30g of mature sourdough with 60g of flour and 30g of water, knead repeatedly for 5-10 minutes to form a new dough, and then ferment the new dough for 1 hour at 30℃ and 80% humidity.

[0013] This invention, based on the ratio of three microbial communities (R1=1:1:1, R2=10:1:1), introduces a synthetic microbial community into a real sourdough fermentation system to explore the regulatory effects of the initial inoculation ratio on microbial community assembly, metabolic network, and nutritional quality. The main conclusions are as follows: (1) The inoculation ratio significantly modulates the fermentation kinetics and community selection intensity of sourdough, but has limited effect on the final fermentation endpoint. A high Lp100 ratio (R2) promotes the acidification rate, and the TTA is about 60% higher than that of R1 at 12 h. A high Lp100 inoculation ratio also strengthens the deterministic selection of the community, increasing the heterogeneous selection ratio to 45.45% in the middle and late stages of fermentation.

[0014] (2) The inoculation ratio modulates the intensity of microbial interactions and metabolic activity, thereby altering metabolic trajectories and influencing substrate utilization patterns. The microbial interaction network strength of the R2 system was significantly enhanced, with the average weighted interaction strength increasing by 89.8% and the average network node degree increasing to 9.29. At the same time, the proportion of negative correlations decreased from 18.03% to 3.65%. Metabolomics analysis showed that R2 exhibited higher metabolic activity in the early stage of fermentation (8-12h), mainly enriched in phenolic metabolism, antioxidant metabolism, and the pentose-glucuronic acid conversion pathway. It also significantly reduced the reducing sugar content of sourdough and the release of free phenols in the steamed bun digestion system.

[0015] This invention confirms that the community building principles established in liquid culture systems are equally applicable to real dough systems. The initial inoculation ratio can regulate the microbial community assembly pattern, alter the metabolic network and nutrient conversion pathway, thereby shaping the nutritional quality and digestibility of sourdough foods.

[0016] The Lactiplantibacillus plantarum Lp100 described in this invention is deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 39266 and deposit date of June 5, 2026.

[0017] Maudiozyma humilis Mh2 is deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 39267 on June 5, 2026.

[0018] Saccharomyces cerevisiae Sc4 is deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 39268 and deposit date of June 5, 2026. Attached Figure Description

[0019] Figure 1 The growth dynamics of aseptic dough fermentation with Lp100 and yeast at inoculation ratios R1 and R2 are shown in the figure. (A) Growth dynamics of Mh2 over 12–72 h, (B) Growth dynamics of Sc4 over 12–72 h, and (C) Growth dynamics of Lp100 over 12–72 h. R1 is a 1:1:1 inoculation ratio, and R2 is a 10:1:1 inoculation ratio. Data are expressed as mean ± standard deviation (n = 3 biological replicates). Figure 2 The effect of different inoculation ratios on dough fermentation kinetics; in the figure: (A) pH value, TTA value, (B) fermentation power, expressed as dough volume growth rate. SY = active dry yeast, R1 = 1:1:1, R2 = 10:1:1. Data are expressed as mean ± standard deviation (n = 3 biological replicates); Figure 3 The effect of inoculation ratio on α-diversity during sourdough fermentation; in the figure: (A) and (E) are the Chao indexes for bacteria and fungi, respectively; (B) and (F) are the Shannon indexes for bacteria and fungi, respectively; (C) and (G) are the Simpson indexes for bacteria and fungi, respectively; (D) and (H) are the number of observed bacterial and fungal species, respectively. Data are expressed as mean ± standard deviation (n=3 biological replicates); Figure 4 The effect of inoculation ratio on the microbial composition during sourdough fermentation is shown in the figure. In the figure: (A) and (B) are the PCA results for bacteria and fungi, respectively. (C) and (D) are the compositions of different taxa of bacteria and fungi, respectively. Figure 5 The diagram shows the community assembly pattern during sourdough fermentation. In the figure: (A) Bacterial community βNTI; (B) Relative proportions of different ecological processes during bacterial fermentation; (C) Fungal community βNTI; (D) Relative proportions of different ecological processes during fungal fermentation. The relative proportions were calculated based on the number of comparisons. R1 and R2 were both compared 210 times, and R1-R2 was compared 450 times. Figure 6 The figures show the bacterial network in sourdough inoculated at ratios R1 and R2; in Figure 1, network structures for (A) and fungi (B) are shown. Node size represents the number of connections (degree), and node color represents different bacterial phyla. Green edges indicate negative correlations between taxa, and red edges indicate positive correlations. Figure 7 Metabolic trajectories were determined by the remodeling time at the initial inoculation ratio; in the figure: (A) PCA score plots during sourdough fermentation in R1 (1:1:1) and R2 (10:1:1); (B) Bray-Curtis mean distance based on the relative abundance of metabolites at adjacent time points; Figure 8Metabolite changes during sourdough fermentation at different Lp100 inoculum ratios are shown in the figure. In the figure, (A, F) represent the trends of metabolite ions during fermentation at R1 and R2 ratios, respectively. All ions were clustered into four groups using K-means clustering. The black solid line represents the median value of ions in each cluster. The four fermentation stages are labeled with Roman numerals. (BE, GJ) ions within each cluster were matched against the KEGG compound database to obtain metabolite classifications. The numbers on the bars represent the number of compounds annotated to that category. (R1=1:1:1, R2=10:1:1); Figure 9 Metabolic pathway remodeling and differential accumulation of flavor-related metabolites driven by inoculation ratio; (AC) Metabolic pathways that changed significantly with R2 inoculation ratio at 8h, 12h, and 24h compared to R1. (D) Temporal abundance changes of flavor-related metabolites with R1 and R2 inoculation ratios; Figure 10 Correlation analysis between microorganisms and flavor-related substances; Figure (A) shows the correlation heatmap between flavor-related substances and microorganisms at the R1 inoculation ratio; (B) shows the correlation heatmap between flavor-related substances and microorganisms at the R2 inoculation ratio; Figure 11 The effect of inoculation ratio on the nutritional quality of sourdough is shown in the figure. (A) Total free amino acid content, (B) Total free phenol content, (C) Reducing sugar content, (D) Total free amino acid release factor, (E) Total free phenol release factor, and (F) Sensory evaluation score of steamed buns. Indicates a significant difference. P < 0.01, ns indicates no significant difference. Data are expressed as mean ± standard deviation (n = 3 biological replicates). C0: unfermented, C24: fermented for 24 h without inoculation, R1: fermented for 24 h with Lp100:Mh2:Sc4 = 1:1:1, R2: fermented for 24 h with Lp100:Mh2:Sc4 = 10:1:1, SY: steamed buns prepared by fermenting dough with 2% active dry yeast. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains, and all materials publicly cited herein and cited by them are incorporated herein by reference.

[0022] Equivalent technologies of the specific embodiments described herein that are readily apparent to those skilled in the art through routine experimentation are included in this application.

[0023] Unless otherwise specified, the experimental methods used in the following examples are conventional methods. Unless otherwise specified, the instruments and equipment used in the following examples are all standard laboratory instruments and equipment; unless otherwise specified, the experimental materials used in the following examples were all purchased from regular biochemical reagent stores.

[0024] I. Main Reagents and Culture Media: The reagents required for the experiment included the EZNA® Soil DNA Extraction Kit (provided by Omega Biotechnology, USA), the PCR Purification Kit (purchased from Shanghai Yuhua Biotechnology Co., Ltd.), Folin-Ciocalteu (purchased from Shanghai Yuanye Biotechnology Co., Ltd.), gallic acid, glucose standard, pepsin (3000U / mg), and trypsin (250U / mg) (purchased from Shanghai Maclean Biotechnology Co., Ltd.), 3,5-dinitrosalicylic acid (DNS) (purchased from Beijing Solarbio Science & Technology Co., Ltd.), the Amino Acid Kit (BC1575) (purchased from Beijing Box Biotechnology Co., Ltd.), and the BCA Protein Quantification Kit (ZJ102) (from Shanghai Yamei Biomedical Technology Co., Ltd.).

[0025] Main instruments: NanoDrop2000 spectrophotometer, VanquishFlex ultra-high performance liquid chromatography-mass spectrometry system (Thermo Fisher Scientific); T100 real-time quantitative PCR instrument (Bio-Ray Biotech, Inc.); Wonbio-96c cryo-tissue homogenizer (Shanghai Wanbo Biotechnology Co., Ltd.)

[0026] The main culture media: Plate counting medium (HB0101), nutrient broth solid medium (HB0108), and crystal violet neutral red bile salt agar (HB0114) were purchased from Qingdao Haibo Biotechnology Co., Ltd.

[0027] II. Experimental Methods: 1. Aseptic Flour Treatment: 1 kg / bag of wheat flour was placed at room temperature and irradiated using an electron beam irradiator at a dose of 10 kGy. Unirradiated natural flour served as a control. After irradiation, the flour stock solution was coated with MRS, YPD, plate counting medium, nutrient broth solid medium, and crystal violet neutral red bile salt agar solid medium to verify the asepticity of the irradiated flour. The results showed sterility and were suitable for experimental use.

[0028] 2. Real-time PCR quantitative analysis of aseptic flour inoculated with Lp100, Mh2, and Sc4 in two inoculation ratios (R1=1:1:1; R2=10:1:1) were inoculated into aseptic flour for fermentation. The number 1 in the ratio is based on the dry weight of wheat flour and is 10. 5 CFU / g. Precultures were obtained by culturing in MRS / YPD liquid medium for 12 h, and then centrifuged (5000 rpm). g The precipitated bacterial cells were collected (after 10 min), washed twice with sterile distilled water, and resuspended in sterile water. The bacterial suspension was mixed with flour and sterile water (flour to water ratio 2:1) in a dough mixer to obtain the initial dough. To minimize the introduction of exogenous microorganisms from the environment, all equipment was wiped with 75% ethanol before the experiment. Fermentation was carried out at 30°C and 80% humidity. 0.5 g of fermented dough was weighed at 12, 24, 36, 48, and 72 h, and DNA was extracted using the EZNA® Soil DNA Extraction Kit. The mass and concentration of the extracted DNA were determined by 1.0% agarose gel electrophoresis and spectrophotometry.

[0029] Based on the 16S rRNA gene sequence of Lp100 and the ITS gene sequences of Mh2 and Sc4, sequence alignment was performed using BioEdit software. Species-specific primers were manually designed for the variable regions of the genes to ensure that the amplified fragment length was approximately 300 bp. Primer sequences are shown in Table 1.

[0030] Table 1: Strain-specific primers qPCR analysis was performed using the SYBR Green I fluorescent dye kit on an ABI 7300 real-time PCR instrument. A standard curve was plotted based on the linear relationship between CFU and amplified Ct values ​​(Table 2) to calculate the viable cell count in the three-strain co-culture.

[0031] Table 2: Standard Curves for Strains 3. Preparation of sourdough: Lp100, Mh2, and Sc4 were inoculated into natural flour fermented dough in two ratios (R1=1:1:1; R2=10:1:1). The preparation method was the same as described above. Fermentation was carried out at 30℃ and 80% humidity, and samples were taken at 0, 4, 8, 12, and 24 hours for subsequent analysis. Samples for metabolomics and high-throughput sequencing analysis were immediately stored at -80℃ until extraction.

[0032] 4. Preparation of sourdough steamed buns: Take 30g of mature sourdough dough, mix it with 60g of flour and 30g of water, knead repeatedly for 5-10 minutes, form a new dough, divide it into 50g portions, and let it ferment a second time for 1 hour in a proofing box at 30℃ and 80% humidity. Place boiling water in a steamer, steam for 20 minutes, turn off the heat and let it stand for 5 minutes, cool for 1 hour, and then place it at -80℃ overnight. The next day, freeze-dry it, grind it into powder, sift it (100 mesh), and package it for storage at room temperature.

[0033] 5. Detection of physicochemical indicators: Mix 10g of dough with 90mL of sterile distilled water and homogenize for 3min. Measure the pH value of the slurry at 0, 4, 8, 12, 24, and 48h. Determine the total titratable acid (TTA) according to the national standard method (GB / T 5756-2022), i.e., titrate with 0.1mol / L NaOH standard solution until the pH value of the slurry is 8.2.

[0034] The leavening power of dough was determined using the dough volume expansion method. Freshly prepared dough samples were placed in a graduated cylinder and fermented at 30°C. Dough volume was recorded at set time intervals. Leavening power was expressed as a percentage increase in volume (VI), calculated using the following formula: Where V0 is the initial dough volume, V t Let t be the volume of the dough at time t.

[0035] 6. Microbial DNA Extraction: Total microbial genomic DNA was extracted from the sourdough samples using the EZNA® Soil DNA Extraction Kit according to the instructions. The extracted DNA was analyzed by 1.0% agarose gel electrophoresis and spectrophotometry to determine its quality and concentration. Primer pairs 338F (5'-ACTCCTACGGGAGGCAGCAG-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3') were used to amplify the hypervariable regions V3-V4 of the bacterial 16S rRNA gene, and primers (5'-CTTGGTCATTTAGAGGAAGTAA-3' / 5'-GCTGCGTTCTTCATCGATGC-3') were used to amplify the fungal ITS region. All reactions were performed on a PCR instrument. The PCR reaction system contained 10 μL of 2×Fast... Pfu buffer, 0.8 μL each of primers (5 μM), and 10 ng template DNA were added, and the volume was brought to 20 μL with ddH2O. The PCR amplification cycling conditions were as follows: initial denaturation at 95℃ for 3 min, followed by 27 cycles, each cycle consisting of 95℃ denaturation for 30 s, 55℃ annealing for 30 s, and 72℃ extension for 45 s, with a final extension at 72℃ for 10 min, and termination at 4℃. PCR products were separated by 2% agarose gel electrophoresis, purified using a PCR purification kit according to the manufacturer's instructions, and quantified using Qubit 4.0. The purified amplicon samples were equimolarly mixed and then processed by Shanghai Meiji Biopharmaceutical Technology Co., Ltd. according to standard procedures using an Illumina analyzer. Paired-end sequencing was performed on the Nextseq2000 platform. The raw sequencing data were uploaded to the NCBI Sequence Reading Archive (SRA) database (PRJNA1414783). Subsequently, operational taxonomic units (OTUs) were further analyzed using the MajorBio cloud platform (https: / / cloud.majorbio.com). A correlation was considered statistically significant if |r|>0.6 and p<0.05; this correlation analysis was performed using OmicStudio (https: / / www.omicstudio.cn).

[0036] 7. Community Assembly Index Calculation: The β-Nearest Taxonomic Unit Index (βNTI) was calculated on the MajorBio Cloud Platform (https: / / cloud.majorbio.com) to assess community assembly patterns. +2 < βNTI > -2 indicates a random microbial community assembly pattern, while the opposite is considered deterministic assembly. We applied the community assembly calculation to time series samples.

[0037] 8. Metabolomics Analysis: First, metabolite extraction was performed using 400 μL of extraction buffer (methanol:water = 4:1 (v:v)) containing 0.02 mg / mL internal standard (L-2-chlorophenylalanine). The samples were homogenized using a cryo-tissue homogenizer for 6 min (-10℃, 50 Hz), followed by low-temperature ultrasonic extraction for 30 min (5℃, 40 kHz). The samples were then incubated at -20℃ for 30 min, followed by centrifugation for 15 min (4℃, 13000 rpm). g The supernatant was transferred to a sample vial for LC-MS / MS analysis. Ultra-high performance liquid chromatography-mass spectrometry (UHPLC-MS) was used for analysis. The chromatographic column was an ACQUITY HSS T3 column (100 mm × 2.1 mm, 1.8 μm). Mobile phase A was 0.1% formic acid aqueous solution: acetonitrile (95:5, v / v), and mobile phase B was 0.1% formic acid acetonitrile-isopropanol aqueous solution (47.5:47.5, v / v). The flow rate was 0.40 mL / min, and the column temperature was 40 °C.

[0038] Mass spectrometry data were acquired using a mass spectrometer equipped with an electrospray ionization (ESI) source, capable of operating in both positive and negative ion modes. Three independent biological replicates were used for each experimental group. Optimized instrument parameters were as follows: ion source temperature 425 °C; sheath gas flow rate 50 arb; auxiliary gas flow rate 13 arb; ion spray voltage fluctuating to -3500 V in negative ion mode and 3500 V in positive ion mode; normalized collision energies of 20 / 40 / 60 eV, with stepwise MS / MS fragmentation analysis performed. Full-scan mass spectrometry had a resolution of 60,000, while tandem mass spectrometry had a resolution of 7500. Data were collected in a data-dependent acquisition mode, with a mass range of m / z 70–1050. Raw data were processed using Progenesis QI software (Waters, Milford, USA), including peak alignment, normalization, deconvolution, and removal of internal standards and artifacts. Metabolite identification was based on the HMDB, METLIN, and Majorbio databases. Matrix data, including sample metadata, metabolite names, and peak intensities, was exported in CSV format. Multivariate statistical analysis and data visualization were performed using MetaboAnalyst 6.0 (http: / / www.metaboanalyst.ca), including principal component analysis (PCA) and KEGG pathway enrichment analysis. Differential metabolite selection was based on fold change (FC>1.5), importance of projected variables (VIP>1), and statistical significance (P<0.05, corrected for FDR using the Benjamini-Hochberg method).

[0039] 9. Bray-Curtis distance calculation: For each biological replicate (i=1,2,3), calculate the Bray-Curtis distance between samples at adjacent time points: Where n is the total number of metabolites, These are the log10-transformed relative abundances of the k-th metabolite in the i-th replicate at times T1 and T2, respectively. The average of the three replicates is taken as the final result. .

[0040] 10. Determination of Total Free Phenolic Content: The total free phenol content (FPC) was determined using the Folin-Ciocalteu method. After freeze-drying the mature sourdough sample, a portion of the dried sample was mixed with 80% (v / v) methanol solution and extracted by constant temperature shaking for 30 min. The extraction was performed at 8000... g Centrifuge for 10 min and collect the supernatant. Repeat the above operation twice, and combine the supernatants from both tests for determination. The determination procedure is shown in Table 3. Using gallic acid as the standard, the total free phenol content is expressed as milligrams of gallic acid equivalent (GAE) per 100 grams of dry weight (DW). The gallic acid standard curve is y = 0.0065x + 0.0056 (R²). 2 =0.9995).

[0041] Table 3: Reaction system for determining total free phenols 11. Reducing sugar determination: The reducing sugar (RS) content was determined using the DNS method. The dried sample was mixed with distilled water and extracted at 50℃ with constant temperature stirring for 30 min. The extract was then further processed at 4000 mL / min. g After centrifugation for 10 min, collect the supernatant. Take 1.0 mL of the extract, add 1.0 mL of DNS reagent, mix well, boil in a water bath for 5 min, cool to room temperature, dilute with water to 10 mL, and measure at a wavelength of 540 nm. Distilled water was used as a blank, and glucose was used as a standard solution for simultaneous operation. The reducing sugar content is expressed as glucose equivalent (mg) per gram of sample. The glucose standard curve is y = 3.8413x - 0.0054 (R²). 2 =0.9994).

[0042] 12. Determination of Total Free Amino Acid Content: The total free amino acid (FAA) content in the mature sourdough sample was determined using an amino acid kit, following the manufacturer's instructions. The sample was homogenized, extracted, and reacted with the provided reagents to generate a colorimetric product. The absorbance was measured at 570 nm using a microplate reader. Leucine was used as the standard, and the FAA content was expressed as milligrams of leucine equivalent per 100 g of sample. The leucine standard curve was calculated as y = 0.0026x - 0.0092 (R²). 2 =0.9995).

[0043] 13. Simulated Gastrointestinal Digestion: A gastrointestinal digestion model of sourdough steamed buns was constructed. Simulated gastric digestive fluid (SGF) was prepared by adjusting the pH to 2.0 with deionized water using HCl. Pepsin stock solution was prepared by dissolving 0.25 g of pepsin in 150 mL of SGF. Severe intestinal digestive fluid (SIF) was prepared using 0.1 mol / L PBS at pH 7.5, and trypsin stock solution was prepared by dissolving 1.2 g of trypsin in 300 mL of PBS. All solutions were prepared fresh and used immediately.

[0044] Gastric phase: First, add 0.5g of sourdough freeze-dried bread powder (100 mesh) to SGF (10mL), preheat at 37℃ for 5min, add pepsin stock solution to make the final concentration 2000U / mL, and then shake at 200rpm for 2h at constant temperature. Intestinal phase: After the gastric digestion stage, mix SIF (20mL) with the above gastric juice, adjust the pH of the final solution to 7.0 with NaOH, shake at 200rpm for 2h at constant temperature, terminate the reaction by water bath at 95℃ for 10min, cool to room temperature, centrifuge at 10000g for 10min to collect the sample solution, and obtain the gastrointestinal digestive products.

[0045] Determination of physicochemical indicators of gastrointestinal digestive fluids: The total free phenols in the digestive fluid of sourdough steamed buns were determined using the same method as described above, calculated based on the release multiple; the FAA was determined using the same method as described above, calculated based on the release multiple.

[0046] 14: Sensory evaluation and texture analysis of steamed buns: A sensory evaluation team of 10 evaluators with some sensory evaluation experience (male to female ratio of 1:1) was selected to evaluate the appearance, elasticity, stickiness, taste, internal structure and overall acceptability of the steamed buns. The scoring criteria are shown in Table 4.

[0047] After the steamed buns were cooled at room temperature for 1 hour, a 20 mm thick central slice was cut as a test sample. The sample was measured using a texture analyzer with the following parameters: force range 25 N, probe height 25 mm, deformation 50%, interval between two compressions 5 s, detection speed 60 mm / min, and initial force 0.05 N.

[0048] Table 4 Sensory Evaluation Scoring Table for Steamed Buns Statistical Analysis: Statistical analysis and data visualization were performed using GraphPad Prism 10.5.0 software. One-way or two-way ANOVA was used according to the experimental design. Post-hoc multiple comparisons were performed using the Tukey multiple comparison test. Metabolite comparisons were performed using the Kruskal-Walli test. Unless otherwise specified, statistical significance was defined as P < 0.05.

[0049] III. Experimental Results: 1. Growth dynamics of different inoculation ratios in aseptic dough fermentation: To analyze the stability of the synthetic communities of the three strains in dough, qPCR was performed to quantify the growth of Lp100 and yeast at initial ratios of R1 and R2 inoculated into aseptic flour, and the growth dynamics of Lp100 and yeast were evaluated. During fermentation, the Sc4 cell count continuously increased over time at both ratios. Figure 1 The peak value of Sc4 in R2 was reached at 72 h. The final value of Sc4 in R2 (8.64 log10 CFU / mL) was slightly higher than that in R1 (8.52 log10 CFU / mL), and the difference became significant from 48 h onwards. Sc4 in R2 maintained strong growth from 48 to 72 h, while the growth in R1 slowed down. The Mh2 count in R1 fluctuated slightly after 12 h, but significantly increased to 7.04 log10 CFU / mL at 72 h, showing a slow growth pattern. In R2, Mh2 essentially stagnated after 12 h, or even slightly decreased, reaching 6.78 log10 CFU / mL at 72 h. The Mh2 count in R1 was higher than that in R2 at all time points. Figure 1 B). This indicates that high initial Lp100 (R2) has a significant inhibitory effect on Mh2. Lp100 reached an extremely high density within 12 hours ( Figure 1 The Lp100 count in R1 decreased slowly, while it remained at a higher level in R2, only slightly declining. The Lp100 count in R2 was significantly higher than in R1 from 24 to 72 hours (average 0.3–0.45 log10 CFU / mL higher). Despite an initial inoculum size difference of only 10-fold, Lp100 maintained its dominance in R2 for a longer period, possibly due to its rapidly established niche and stronger competition for nutrients at higher densities. The viable cell ratios of the three strains during fermentation in aseptic flour were similar to the relative abundances of strains detected in most mature sourdough.

[0050] 2. Effect of different inoculation ratios on dough fermentation kinetics: R1 and R2 were inoculated into natural flour to investigate the effect of inoculation ratio on the fermentation behavior of sourdough. pH, TTA, and dough expansion were continuously monitored throughout the fermentation process. In the initial stage (0-4h), the pH and TTA levels of the R1 and R2 sourdoughs showed similar trends (…). Figure 2 A). At 8 h of fermentation, the pH of R2 (4.9) was significantly lower than that of R1 (pH=5.9). By 12 h, the TTA of R2 was significantly higher than that of R1 (R2: 6.54; R1: 3.90), indicating that the acid accumulation rate may be accelerated at higher Lp100 inoculum sizes. Despite these kinetic differences, both systems gradually approached similar acidification endpoints within 24–48 h, with pH stabilizing in the 3.8–4.0 range and TTA approaching 9.0. The data suggest that the inoculum ratio primarily affects the acidification rate, while its impact on the final acidification level is relatively limited.

[0051] Further investigation was conducted to determine whether acidification kinetics translated into functional fermentation results. As a control, active dry yeast (SY) showed a rapid increase in fermentation activity within the first 4 hours, reaching 170% VI, followed by a significant decrease after 12 hours. Figure 2 B). In contrast, the R1 and R2 sourdough systems showed delayed but sustained expansion. At 12 h, there was no significant difference in the dough volume growth rate between R1 and R2, with VI values ​​of approximately 167% and 140%, respectively. The results indicate that the inoculation ratio primarily determines the time distribution of fermentation performance and has a relatively small impact on fermentation capacity.

[0052] 3. Effects of inoculation ratio on microbial diversity: The α-diversity of bacterial communities generally decreased during fermentation, with significant differences observed at different inoculation ratios. The Chao index and observed species gradually decreased with the progress of fermentation. Figure 3 A, D). In the early stages of fermentation (0-4h), there was no significant difference between R1 and R2. As fermentation progressed, R1 was significantly higher than R2 at 8h and 24h, indicating that a higher inoculum ratio may lead to a more rapid decline in community richness. The Shannon index remained at a high level in the early stages of fermentation, then decreased significantly. Figure 3 B). At 4h and 8h, R1 was significantly higher than R2, while at 12h there was no significant difference between the two groups, and at 24h R1 was again higher than R2. The Simpson index showed the opposite trend to Shannon's ( Figure 3 (c) At multiple time points, R2 was significantly higher than R1, indicating that community evenness decreased under high Lp100 inoculation conditions. This suppressed the growth of other bacterial groups through acidification and resource competition, leading to a decline in community diversity. Under R1 conditions, the initial selection pressure was relatively weak, and the community maintained a higher level of species coexistence over a longer period.

[0053] Unlike bacterial communities, the α-diversity of fungal communities did not differ significantly with changes in inoculum ratio. The Chao index and observed species, characterizing richness, were high in the early stages of fermentation, then rapidly decreased during 0-8 hours. Figure 3 E, H), indicating that the richness of the fungal community may be reduced by factors such as acidic environmental disturbance and nutrient competition, and remained at a low level during the 12-24h period. The Shannon index was high in the early stage of fermentation, decreased during 4-8h, and rebounded in 12-24h. Figure 3 F), while the Simpson index shows the opposite trend to Shannon's (F). Figure 3 G). The inoculation ratio has a limited regulatory effect on fungal species richness.

[0054] 4. Effect of inoculation ratio on microbial community composition: To compare the changes in the community structure of sourdough with different inoculation ratios, PCA analysis was performed on the samples based on OTU levels. The results showed significant differences in bacterial community structure among different samples (R=0.78436, P =0.001). The sample distribution indicates that fermentation time is the driving factor for changes in bacterial community structure. On the PC1 axis, the samples show a time-series change with fermentation time ( Figure 4 A). As fermentation progressed, the samples gradually moved towards the positive direction of PC1, indicating a significant succession of the bacterial community structure during fermentation. In the early stages of fermentation (0h), the distribution of R1 and R2 samples in the bacterial and fungal communities was relatively similar, indicating small differences in the initial community structure under the two inoculation ratios. From 4-8h, the R2 group samples showed greater dispersion in the PC2 direction, indicating that a higher Lp100 inoculation ratio accelerated the remodeling process of the community structure. In the later stages of fermentation (12-24h), R1 and R2 samples gradually clustered towards the positive direction of PC1, and the distance between samples at different time points decreased significantly, indicating that the bacterial community gradually stabilized. The fungal community results showed ( Figure 4 (B) The fungal community structures of groups R1 and R2 were quite similar at the beginning of fermentation. As fermentation progressed, the two groups showed a temporal succession trend, indicating that the fermentation process was the driving factor for changes in fungal community structure. Between 8 and 24 hours, the clusters of samples from groups R1 and R2 highly overlapped, indicating that the fungal community structure tended to stabilize at this stage, and different inoculation ratios had little impact on the overall structure of the fungal community.

[0055] Taxonomic features at the genus level support the observed changes in community structure. The changes in microbial composition were similar in R1 and R2, with *Lactobacillus* genus (*Lactobacillus*) showing a higher proportion during fermentation. Lactiplantibacillus The largest proportion () Figure 4 C), whose relative abundance continued to increase during fermentation, reaching 92.15% and 97.34% at 24 h, respectively. In the fungal community ( Figure 4 D), Kazakhstani bacteria ( ) in R1 and R2 doughs Kazachstania ) is the dominant genus, followed by yeast ( Saccharomyces ). Kazachstania The percentage of R1 gradually increased within 0-8 hours, reaching a peak at 8 hours (R1: 90.91%; R2: 91.16%), and then decreased within 8-24 hours; while Saccharomyces The number of these species increases within 12-24 hours. Simultaneously, some potentially harmful environmental fungi, including *Alternaria* (…), are also present. Alternaria ), Powdery mildew ( Blumeria ) and Aspergillus ( Aspergillus It was only detected in the early stages of fermentation, indicating that fermentation-related acidification has a strong inhibitory effect on environmental fungi.

[0056] 5. Effects of different inoculation ratios on community assembly patterns: To assess the relative importance of determinism and randomness in microbiome assembly, a null model (999 randomizations) was used to calculate βNTI. During sourdough fermentation, the community was influenced by both random processes (|βNTI|<2) and deterministic processes (|βNTI|>2). In the R1 bacterial community, random processes dominated (…). Figure 5 A) The βNTI values ​​at 0, 12, and 24 h remained within the range of 0-2, while at 4 h and 8 h, some βNTI values ​​were >2, indicating that the bacterial community was simultaneously influenced by both stochastic and deterministic processes at this stage. In contrast, the bacterial community established under R2 conditions had βNTI values ​​ranging from -1 to +2 at 0 h and 12 h, dominated by stochastic processes. The βNTI values ​​at 4 h, 8 h, and 24 h were between 0 and 7, indicating that the bacterial community in R2 was simultaneously influenced by both stochastic and deterministic processes at this stage. Based on βNTI, the Raup-Crick index (RCbary value) was used to divide the deterministic and stochastic processes into five ecological processes and determine the relative importance of each ecological process. Figure 5 According to B, within R1, the number of pairwise comparisons for 8h and 12h is relatively small. Heterogeneous selection accounts for 100% of the 12h comparisons, while for 24h, undetermined processes account for 66.67%, and heterogeneous selection accounts for 33.33%. Within R2, at 8h, undetermined processes account for 62.50%, heterogeneous selection for 25.00%, and diffusion restriction for 12.50%; at 12h, heterogeneous selection accounts for 100%; and at 24h, undetermined processes account for 51.52%, heterogeneous selection for 45.45%, and homogeneous selection for 3.03%. In the comparisons between R1 and R2, heterogeneous selection accounts for 100% at 4h; heterogeneous selection accounts for 62.50% at 8h, and undetermined processes for 37.50%; heterogeneous selection accounts for 56.25% at 12h, and undetermined processes for 43.75%; and heterogeneous selection accounts for 45.83% at 24h, undetermined processes for 52.08%, and homogeneous selection for 2.08%. The results showed that the proportion of deterministic selection (especially heterogeneous selection) within R2 and between R1 and R2 was significantly higher than that within R1 in the mid-to-late stages of fermentation. This indicates that a higher initial Lp100 inoculum enhances the shaping effect of environmental filtration on the bacterial community and drives the structural differentiation of the community under both proportions.

[0057] In the fungal community, both inoculation ratios exhibited a dominant stochastic process during fermentation (|βNTI|<2). Figure 5 C), the classification and statistics of ecological processes further confirm ( Figure 5D). In group R1, the proportion of undetermined processes remained between 17.14% and 19.05%, with extremely low rates of heterogeneous selection and homogeneous diffusion (0%-2.38%). In group R2, the proportion of undetermined processes ranged from 16.19% to 18.10%, with a low rate of heterogeneous selection (0.95%-3.81%), and occasional occurrences of homogeneous selection and homogeneous diffusion (0.48%-1.43%). Comparing the two groups, the proportion of undetermined processes ranged from 16.44% to 18.89%, the rate of heterogeneous selection ranged from 0% to 2.44%, homogeneous selection occurred only at 8 hours (2.44%), and diffusion restriction also occurred only at 8 hours (0.67%). The proportion of definitive selection at each time point and in each group did not exceed 4%, and no significant differentiation was observed among different vaccination ratios.

[0058] The assembly patterns of bacterial and fungal communities differ significantly. Bacterial communities are jointly regulated by deterministic selection and stochastic processes, with the initial inoculation ratio significantly influencing the intensity and duration of deterministic selection. Under R2 conditions, the proportion of deterministic selection (especially heterogeneous selection) is higher and its duration is longer. Fungal communities, on the other hand, are primarily dominated by stochastic processes, showing a conservative response to the initial inoculation ratio, with limited contribution from environmental filtration. This difference may stem from the rapid metabolic response of bacteria (especially lactic acid bacteria) to environmental changes such as acidification, while fungi, as eukaryotic microorganisms, are more influenced by stochastic factors such as dispersal and ecological drift in their community dynamics.

[0059] 6. The effect of inoculation ratio on microbial interaction networks: Bacterial co-occurrence network analysis revealed that adjusting the initial inoculation ratio significantly altered the intensity of interactions within the microbial community and the overall network structure. Figure 6 A, Table 5). Under R2 conditions, the average weighted interaction strength increased significantly from 3.33 in R1 to 6.32, an increase of nearly 90%, indicating a general increase in the strength of bacterial community associations. This shift to positive correlations was accompanied by a decrease in negative correlations, with the proportion of negative associations decreasing from 18.03% in R1 to 3.65% in R2. The interaction patterns of most community members changed from mixed in R1 to almost entirely positive correlations in R2. For example, Acinetobacter (… Acinetobacter In R1, a certain proportion of negative correlations remained, while in R2, positive correlations approached 100%. This reorganization was accompanied by a significant enhancement in network connectivity, with the average node degree increasing from 6.13 to 9.29. These results indicate that the initial inoculation ratio did not change the centrality hierarchy of the core taxa, but rather promoted the reorganization of the community from a network dominated by negative correlations to an association network structure dominated by positive correlations by enhancing the strength of interactions.

[0060] A similar positive correlation enhancement trend was observed in fungal co-occurrence networks. Figure 6B). Although the fungal network showed a positive correlation in R1 (approximately 97.3% positive correlation), this trend was further strengthened in R2 (96.3%). At both inoculation ratios, g_ Saccharomyces The negative correlation axis centered on the inoculation ratio shows negative correlations with various fungal taxa in both R1 and R2. Since microbial interaction networks are a crucial basis for metabolic exchange, such changes in network structure driven by inoculation ratios may influence community metabolic patterns.

[0061] Table 5: Topological properties of bacterial and fungal community co-occurrence network analysis of sourdough at two inoculation ratios 7. Changes in metabolic trajectory driven by inoculation ratio: To assess whether the inoculation ratio affects the metabolic pattern throughout the fermentation process, metabolomics analysis was performed. PCA results revealed that the metabolic dynamics of both R1 and R2 sourdough fermentation were time-dependent, with fermentation time explaining 60.0% of PC1. Figure 7 A). R1 ​​and R2 exhibited non-overlapping pathways throughout the fermentation process. In the early stages of fermentation (0-4h), the samples with both inoculation ratios showed low dispersion in PCA space. As fermentation progressed, the dispersion gradually increased, and even after fermentation, the two inoculation systems failed to converge, indicating a historical dependence of metabolite accumulation and persistent differences in metabolite profiles formed under different inoculation ratios. Notably, the metabolic state of R1 at 12h was very close to that of R2 at 8h in PCA space, suggesting that different inoculation ratios can achieve similar metabolic conformations at different fermentation time points.

[0062] To further quantify the dynamic process of metabolic changes, the Bray-Curtis distance between adjacent time points was calculated. Figure 7 B). The distance gradually increases during fermentation, indicating that metabolic turnover is continuous throughout the process. Compared with R2, the metabolic composition of R1 changes more significantly over time, highlighting the significant differences in the rhythm and pace of metabolic evolution under the two inoculation ratios.

[0063] 8. Metabolite Dynamic Cluster Analysis: To analyze the dynamic change patterns of metabolites, Fuzzy C-Means Clustering was used for Mfuzz time series cluster analysis, classifying metabolites into four dynamic patterns (…). Figure 9 Both R1 and R2 exhibit clear temporal clustering structures, but the temporal variation characteristics of each cluster and the types of metabolites they contain differ significantly. In the R1 system ( Figure 8 A), cluster 1 is mainly rich in organic acids and phospholipids, which continuously increase during fermentation. Figure 8B); Cluster 2 mainly consists of amino acids and peptides, and its content continues to decline, consistent with the gradual consumption trend of nitrogen. Figure 8 C); Cluster 3 is mainly composed of carbohydrate-related metabolites, which show an increasing trend in the later stages of fermentation (C). Figure 8 D); while cluster 4 is rich in oligosaccharides and vitamins ( Figure 5-8 E) decreased during the same period. Overall, the metabolic dynamics in R1 showed a smooth pattern and a consistent direction of change. In contrast, the metabolic dynamics in R2 were more complex. Figure 8 F). Cluster 1 is rich in carboxylic acids, fatty acids, and oligosaccharides, which begin to decrease after 8 hours, and the rate of decrease accelerates after 12 hours. Figure 8 G). Conversely, cluster 2 is mainly composed of carboxylic acids, monosaccharides, and phospholipids, which continuously increase throughout the fermentation process (G). Figure 8 H). Cluster 4, rich in amino acids and peptides ( Figure 8 J) decreased significantly within 4-8 hours, and then partially recovered within 12-24 hours, showing a clear trajectory of first decreasing and then increasing. These patterns indicate that there are differences in the metabolic patterns of the R1 and R2 systems.

[0064] 9. Effect of inoculation ratio on metabolic pathway enrichment: Key metabolite (FC>1.5) analysis showed that the initial inoculation ratio may have shaped unique metabolic trajectories at different stages. No effective metabolite data were obtained from 4h samples. At 8h ( Figure 9 A), R2 showed a significant metabolic shift compared to R1, with 28 metabolites significantly upregulated and only 2 metabolites downregulated (Table 6). KEGG enrichment analysis of these differentially expressed metabolites identified 5 significantly enriched pathways ( P <0.05, mainly involving phenolic compound metabolism and redox processes, including phenylpropanoid biosynthesis, ascorbate and aldarate metabolism, pentose and glucuronate interconversions, and cofactor biosynthesis. Maximum metabolic difference is reached at 12 hours. Figure 9B), at this point, 52 metabolites were significantly upregulated in R2 compared to R1, while only 1 metabolite was downregulated (Table 7). Although the number of differentially expressed metabolites increased significantly, pathway enrichment was limited to three core metabolic pathways: pentose and glucuronate interconversions, ascorbate and aldarate metabolism, pentose and glucuronate interconversions metabolism, and biosynthesis of cofactors. 24 hours later ( Figure 9 C), metabolic differences were significantly reduced, with only three upregulated metabolites and one downregulated metabolite detected in R2 (Table 8), characterized by only minor adjustments to nucleotide-related pathways.

[0065] Table 6: Metabolites showing significant changes in R2 compared to R1 at 8 h Table 7: Metabolites showing significant changes in R2 compared to R1 at 12 h Table 8: Metabolites showing significant changes in R2 compared to R1 at 24 h Compared to R1, R2 exhibited higher metabolic activity and a wider range of metabolic pathway responses during 8–12 h, particularly in phenolic compound metabolism, antioxidant-related metabolism, and the pentose-glucuronide conversion pathway. However, this difference gradually diminished after 24 h, suggesting that a higher Lp100 ratio may primarily influence the overall metabolic trajectory of the sourdough system by enhancing metabolic activity in the early stages of fermentation.

[0066] To further analyze the potential flavor differences caused by different inoculation ratios, 15 flavor-related substances were screened from the differential metabolites, including phenolic acids, organic acids, esters, nucleosides, and various flavor peptides. The results showed that during the 8-24 h fermentation process, most flavor metabolites exhibited an increasing trend in both R1 and R2 (…). Figure 9(D). However, different inoculation ratios (R2:R1) significantly altered flavor-related metabolites. Specifically, aromatic amino acid-related small peptides such as Ile-Phe and Phe-Leu reached higher levels in R2 during the early stages of fermentation, while their accumulation was relatively delayed in R1, indicating that R2 enhanced the turnover of aromatic amino acids and the formation of flavor precursors. Simultaneously, glutamate-related peptides such as Glu-Ile, Glu-Leu, and Glu-Val-Leu were significantly enriched in R2, suggesting that a high Lp100 ratio may promote the formation of umami-related metabolites, thereby enhancing the umami characteristics of the system. Furthermore, the accumulation level of the phenolic metabolite caffeoyl quinic acid was generally higher in the R2 system than in R1. Notably, some pyroglutamate-related metabolites showed a decreasing trend in the later stages of R2, indicating that they may have undergone further metabolic transformation, thereby reducing the accumulation of bitter-related substances. Overall, a higher Lp100 inoculation ratio not only alters the overall metabolic trajectory of the sourdough system, but also drives the sourdough to form a potential flavor profile of high aroma, strong umami, and low bitterness by promoting the early accumulation of aromatic amino acids, umami peptides, and phenolic metabolites.

[0067] 10. Correlation analysis between core strains and key flavor precursors: To analyze the changes in the association patterns between core strains and flavor-related metabolites under different inoculation ratios, the potential flavor-related metabolites screened were correlated with the taxonomic units corresponding to the three inoculated strains. Lactiplantibacillus plantarum , Maudiozyma humilis and Saccharomyces Spearman correlation analysis was performed on sp. The results showed that both vaccination ratios resulted in a correlation between the two vaccination ratios. Lp. plantarum and Saccharomyces The flavor association module is based on sp., but there are significant differences in the association patterns between different bacterial species and flavor metabolites.

[0068] In the R1 system ( Figure 10 A), Lp. plantarum It showed a significant positive correlation with a variety of flavor-related metabolites, including the umami-related peptide Glu-Val-Leu (r=0.875, P<0.001), organic acids (2-Hydroxycaproic acid), phenolic acid metabolites (Caffeoyl quinic acid), and esters (Methionine ethyl ester), indicating that it may be involved in the release and accumulation of flavor precursors. Saccharomyces sp. in R1 and Lp. plantarum It exhibits a similar correlation, showing a significant positive correlation with Glu-Val-Leu, 2-Hydroxycaproic acid, and Phe-Leu. In contrast, M. humilisThere are few significantly associated metabolites in R1, and they are mainly negatively associated with Val-Ile, Glu-Leu and Pyroglutamylisoleucine, with a relatively weak overall association strength.

[0069] Compared to R1, the strain-flavor metabolite association pattern in the R2 system changed significantly. Figure 10 B). Lp. plantarum It maintained a significant positive correlation with caffeoyl quinic acid and 2-hydroxycaproic acid, but its correlation with Methyl 2-(2-Hydroxypropanoylamino)-3-Methylbutanoate changed from insignificant in R1 to significant positive (r=0.532, P<0.05). Meanwhile, Saccharomyces The increased positive correlation between *Lactobacillus* sp. and ester-related metabolites such as *Methionine ethyl ester* (r=0.875, P<0.001) and *Methyl 2-(2-Hydroxypropanoylamino)-3-Methylbutanoate* (r=0.761, P<0.01) suggests that a higher Lp100 proportion may promote synergistic ester flavor metabolism between lactic acid bacteria and yeast. It is noteworthy that... M. humilis In R2, there was a significant negative correlation with multiple flavor-related metabolites, including caffeoylquinic acid, 2-hydroxyhexanoic acid, Glu-Val-Leu, and Phe-Leu, indicating that... M. humilis It may serve as a potential core microbial species for flavor regulation.

[0070] 11. Effect of Inoculation Ratio on the Quality Characteristics of Sourdough and Fermented Steamed Buns: To evaluate the effects of different inoculation ratios on substrate utilization and nutrient metabolism, total free amino acids (FAA), total free phenols (FPA), and reducing sugars (RS) were measured, reflecting physicochemical indicators of protein degradation, phenolic release, and carbohydrate metabolism, respectively. There was no significant difference in FAA content between uninoculated dough and fermentation (C0 vs. C24). The FAA content of sourdough inoculated with both R1 and R2 ratios was significantly reduced, but there was no significant difference between R1 and R2. Figure 11 A). This result indicates that, in the inoculation system, the continuous uptake and metabolism of amino acids by the strains may be the main reason for the decrease in FAA content. Compared with uninoculated fermented dough (C24), R1 and R2 significantly increased the FPA content in sourdough. Figure 11B). This indicates that the inoculation of Lp100 produces acid, lowering the pH of the system, and through the action of glycosidases, hydrolyzes bound phenols and increases the content of free phenols, thereby significantly increasing the FPA level in the sourdough. There was no significant difference between R1 and R2, which is consistent with the similarity of acidification levels at the fermentation endpoint. Furthermore, both the R1 and R2 inoculation ratios significantly reduced the reducing sugar (RS) content in the sourdough, with R2 being significantly lower than R1. Figure 11 C). Reducing sugars are rapidly consumed during fermentation for energy metabolism and cell growth, and are converted into metabolites such as organic acids, ethanol, and CO2. Under R2 conditions, a higher initial inoculum size of Lp100 may accelerate the rate of sugar metabolism, leading to more significant consumption of reducing sugars, further indicating that a higher initial inoculum ratio can enhance the utilization of carbon sources by microorganisms.

[0071] To further evaluate the effect of different inoculation ratios on the digestibility of sourdough steamed buns, mature sourdough was made into steamed buns, and an in vitro gastrointestinal digestion model was constructed. The release folds of FAA and FPA in the in vitro digestive fluid were measured. Compared with uninoculated fermented steamed buns (C0), the FAA release folds of commercial yeast (SY) steamed buns and R2 sourdough steamed buns were significantly increased (P<0.05), while R1 was at an intermediate level between the two, but there was no significant difference compared with SY and R2. Figure 11 D), indicating that different inoculation ratios have a relatively limited regulatory effect on the release of protein digestion products. R1 significantly increased the fold increase in FPA release during the digestion of steamed buns ( Figure 11 E). This phenomenon may be related to the release mechanism of phenolic substances during sourdough fermentation. A moderately acidified environment can promote the breakdown of grain cell wall structure and facilitate the release of bound phenols from polysaccharide or protein complexes, increasing the amount of phenols available for release during digestion and thus improving health-related functions. However, under R2 conditions, due to the higher Lp100 inoculum, the fermentation system acidifies more rapidly, leading to the degradation or transformation of some phenolic substances, or they may form new binding forms with proteins and polysaccharides during fermentation. Metabolomics data show a significant upregulation of bound phenols (such as tachioside and 5-O-Feruloylquinic acid) in R2 (Table 9), supporting the hypothesis that phenols undergo binding transformation during fermentation, resulting in a relatively lower release level during digestion.

[0072] Table 9: Changes in bound phenols and related metabolites between R2 and R1 (24h) Sensory evaluation results showed that ( Figure 11The sensory scores of the sourdough steamed buns in groups R1 and R2 were higher than those in group SY, indicating that sourdough fermentation effectively improves the sensory characteristics of the steamed buns. Groups R1 and R2 scored similarly in terms of structure, color, elasticity, and aroma, with group R2 scoring highest in stickiness (16.64), showing better palatability and texture. In contrast, group SY scored lower overall in all sensory indicators, especially in elasticity (8.90), stickiness (13.90), and aroma (13.23), significantly lower than the sourdough group. This suggests that sourdough fermentation not only improves the appearance and internal structure of the product but also significantly enhances its texture and flavor quality. The textural analysis results were largely consistent with the sensory scores (Table 10). Compared to group SY, sourdough fermentation significantly reduced the hardness and chewiness of the steamed buns while improving elasticity and resilience. Overall, sourdough fermentation significantly improved the textural properties and sensory quality of the steamed buns. R1 emphasizes softness and uniform texture, while R2 excels in elasticity and delicate taste. Both have comparable overall sensory appeal and are significantly superior to traditional active dry yeast fermentation.

[0073] Table 10 Effects of two inoculation ratios on the textural properties of steamed buns Note: Different letters indicate significant differences under the same parameter.

[0074] Community assembly theory posits that varying initial abundances of the same species alter interspecific competition, resource utilization, and stochastic drift, thereby changing the assembly pathway of microbial communities and the relative contributions of deterministic and stochastic processes. Increasing the initial abundance of *Lactobacillus plantarum* Lp100 (R2) allows Lp100 to establish population dominance earlier through acidification, thus inhibiting the colonization of other species and altering community structure. Conversely, when Lp100 abundance is low (R1), resource utilization is shared by multiple species, and stochastic fluctuations more readily change the relative abundance of species. Under R1 conditions, due to the high coexistence levels of various bacteria in the early stages of fermentation, resource competition and stochastic drift jointly influence community dynamics. Furthermore, community size affects interspecific competition; species entering the community with higher initial abundance have stronger resource occupation capabilities, thus establishing a competitive advantage in the early stages and influencing subsequent species colonization. The positive correlation of bacterial networks in the R2 system was significantly enhanced, while the proportion of negative correlations decreased, indicating that the competitive relationship within the community may be replaced by the synergistic relationship of resource utilization. This further illustrates that the initial inoculation ratio not only affects the community structure but also the ecological interaction patterns within the community.

[0075] Community metabolic patterns and functional outputs are influenced by complex interaction network structures. The metabolic trajectories of the R1 and R2 systems differed during fermentation. Correlation analysis showed a significant association between changes in relative amino acid abundance and changes in the abundance of various bacterial and fungal taxa, but the strength of these associations varied significantly with the inoculation ratio, particularly evident in the R2 system. This suggests that a higher Lp100 inoculation ratio may reshape nitrogen metabolism interactions within the community by enhancing microbial metabolic activity. This finding aligns with the role of cross-feeding of amino acids in the construction of the community metabolic network and the promotion of microbial co-metabolism, indicating that early community structure may influence long-term metabolic pathways by altering metabolic network interactions.

[0076] This invention indicates that the inoculation ratio may be an important ecological factor regulating community interaction networks. By altering the initial community structure and the intensity of metabolic interactions, the inoculation ratio shapes different metabolic network structures in the early stages of fermentation and continuously influences the metabolic dynamics throughout the entire fermentation process through a history-dependent effect.

[0077] This invention confirms that the community building principles established in liquid culture systems are equally applicable to real dough systems. The initial inoculation ratio can regulate the microbial community assembly pattern, alter the metabolic network and nutrient conversion pathway, thereby shaping the nutritional quality and digestibility of sourdough foods.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A processing method for improving the quality and digestibility of sourdough foods through three-strain fermentation, characterized in that: Lactobacillus plantarum ( Lactiplantibacillus plantarum Lp100, Dwarf Kazakhstan yeast ( Maudiozyma humilis Mh2 and brewer's yeast ( Saccharomyces cerevisiae Sc4 is inoculated into flour at an inoculation ratio of 1:1:1 or 10:1:1 to prepare sourdough. The inoculation ratio is calculated based on the dry weight of wheat flour, where 1 represents 10. 5 CFU / g; then, the prepared sourdough is mixed with flour and water to prepare dough, and after a second fermentation, sourdough food is prepared.

2. The processing method according to claim 1, characterized in that: The Lp100, Mh2 and Sc4 were inoculated into flour at an inoculation ratio of 10:1:1 to prepare sourdough.

3. The processing method according to claim 1 or 2, characterized in that: The specific method for preparing sourdough is as follows: Lp100, Mh2, and Sc4 are cultured in MRS or YPD liquid medium for 12 hours to obtain precultures, and 5000 g Centrifuge for 10 minutes to collect the precipitated bacterial cells, wash twice with sterile distilled water, and resuspend in sterile water; mix the bacterial suspension with flour and sterile water evenly according to the ratio of flour to water of 2:1 to obtain the initial dough. Ferment the initial dough at 30℃ and 80% humidity for 8-24 hours to obtain the mature sourdough.

4. The processing method according to claim 1, characterized in that: The method for secondary fermentation is as follows: Mix 30g of mature sourdough with 60g of flour and 30g of water, knead repeatedly for 5-10 minutes to form a new dough, and then ferment the new dough for 1 hour at 30℃ and 80% humidity.

Images