Weissella confusa and its application in improving the microbial structure and flavor of green moldy tofu

Abstract

The present invention discloses Weissella confusa and its application in improving the microbial structure and flavor of green-preserved tofu, belonging to the field of food fermentation. The Weissella confusa, named M1, is preserved in the General Microbiological Center of the China General Microbiological Culture Collection Center, and the preservation number is CGMCC NO. 24526. The Weissella confusa of the present invention can be used as an auxiliary fermentation agent for the production of green-preserved tofu to improve the microbial structure and flavor of green-preserved tofu.

Description

Technical Field

[0001] The invention relates to the field of food fermentation, and more particularly to fused Weissella and application thereof in improving the microbial structure and flavor of green squares. Background Art

[0002] Fermented bean curd is a condiment made from soybeans through natural solid-state fermentation. It can be divided into three types: red, green and white, according to different production processes. The green type has the most distinctive flavor, which smells bad but tastes good, so it is also called stinky tofu. In addition to its strong odor and the mellow flavor of soy products, the green type is also rich in nutrients such as isoflavones, vitamin B12 and amino acids, and has antioxidant, blood pressure lowering, cholesterol lowering and bone loss reducing effects.

[0003] Flavor is an important indicator for evaluating the sensory quality of fermented foods, and the formation of flavor is closely related to microorganisms. In essence, the formation of Qingfang flavor is strictly regulated by the microbial community during the fermentation process, and abnormal flavor is the most common quality problem of Qingfang. Compared with the traditional natural fermentation process, modern fermentation technology based on microbial regulation can accurately control the fermentation process of Qingfang, which is of great significance for improving product quality and safety. However, the lack of high-quality fermentation agents is a prominent bottleneck restricting the industrialization process of the product. Therefore, how to provide a microorganism that can be used for Qingfang fermentation has become a technical problem that needs to be solved urgently in this field. Summary of the invention

[0004] In view of this, the present invention provides a fusion of Weissella and its application in improving the microbial structure and flavor of Qingfang fermented bean curd. By studying the mechanism of its influence on Qingfang fermentation, a theoretical basis is provided for the excellent starter of Qingfang fermented bean curd, laying a foundation for the modern industrial production of traditional Qingfang fermented bean curd.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] Weissella confusa, named M1, is deposited in the General Microbiology Center of China Culture Collection Administration, with the deposit number CGMCC NO.24526. The deposit address is No. 3, Yard No. 1, Beichen West Road, Chaoyang District, Beijing, and the deposit time is March 14, 2022.

[0007] It has been found through research that Weissella M1 contributes to the fermentation of blue-veined fermented bean curd, and makes important contributions to the quality characteristics of the product such as the fermentation cycle, flavor, texture, and nutritional components, indicating that M1 can be used as an excellent starter culture for the industrial production of blue-veined fermented bean curd. Regarding how Weissella drives microbial changes during the fermentation of blue-veined fermented bean curd and thus affects the flavor, the present invention characterized the effects of Weissella M1 on the key flavor compounds during the fermentation of blue-veined fermented bean curd, as well as the core microbial community structure and function related to flavor formation by combining metagenomics and flavoromics technologies, providing a theoretical basis for Weissella M1 as an excellent starter culture for blue-veined fermented bean curd and laying a foundation for the modern industrial production of traditional blue-veined fermented bean curd.

[0008] A blue-veined fermented bean curd starter culture, characterized in that it includes the above-mentioned fused Weissella.

[0009] Use of the above-mentioned fused Weissella or the above-mentioned blue-veined fermented bean curd starter culture in the preparation of blue-veined fermented bean curd.

[0010] The above-mentioned fused Weissella can improve the structure of the microorganisms in blue-veined fermented bean curd, adjust the microbial metabolic pathways in blue-veined fermented bean curd (including alanine-aspartate and glutamate metabolic pathways, arginine and ornithine metabolic pathways, etc.), and regulate the synthesis of flavor compounds.

[0011] A kind of blue-veined fermented bean curd prepared by fermenting with the above-mentioned fused Weissella or the above-mentioned blue-veined fermented bean curd starter culture.

[0012] A preparation method of blue-veined fermented bean curd, inoculating the above-mentioned fused Weissella or the above-mentioned blue-veined fermented bean curd starter culture into the fermentation broth, adding the fermented and salted tofu curd embryo, and fermenting at 28 ± 2 °C for 28 - 30 d.

[0013] The inoculation amount of the above-mentioned fused Weissella is 1×10 6 CFU / mL.

[0014] As can be seen from the above technical solutions, the Weissella confusa disclosed in the present invention, as an auxiliary fermenting agent for the production of Qing Fang, has an impact on the physical and chemical indexes, microbial succession and flavor substances of the product. Compared with natural fermentation, inoculation with M1 can accelerate the fermentation process and significantly improve the protease activity, amino acid nitrogen and amino acid content. The sensory evaluation results show that the addition of M1 can improve the flavor, color, texture and taste of the product. Combining metagenomics and flavoromics technologies reveals the key flavor compounds formed during the fermentation process of Qing Fang driven by M1, as well as the core microbial community structure related to flavor formation. These microbial communities are mainly involved in the transport and metabolism of amino acids and carbohydrates. Spearman correlation analysis shows that Enterobacter and Weissella that appear during the fermentation process of M1 are significantly positively correlated with 13 key flavor substances. In addition, Enterobacter, Escherichia and Tetragenococcus are related to the production of various amino acids.

[0015] In summary, the present invention for the first time proves that Weissella confusa M1 can be used as an auxiliary fermenting agent for the production of Qing Fang, and clarifies the mechanism of action between the microbial succession and flavor formation driven by M1 in the fermentation of Qing Fang, providing a theoretical basis for the modern industrial production of traditional Qing Fang fermented bean curd. Brief Description of the Drawings

[0016] Figure 1 Shows the test results of the physical and chemical properties of Qing Fang fermented bean curd;

[0017] Among them, A. pH, B. salt content, C. protease activity, D. TCA-soluble peptide content, E. amino acid nitrogen (AAN) content, F. chromaticity a* (green) value; G. sensory evaluation results (Appearance, appearance; Texture, texture; Taste, taste; Harmony: overall, Aroma: fragrance), H. changes in the appearance (Appearance) of Qing Fang fermented bean curd at different times (Period); *P≤0.05, **P≤0.01 and ***P≤0.001;

[0018] Figure 2 Shows the heat map of the amino acid content of groups M and N of Qing Fang fermented bean curd;

[0019] Figure 3 Shows the OPLS-DA analysis of the amino acid determination results of Qing Fang fermented bean curd;

[0020] Among them, A and B are the OPLS-DA score plots and model validation plots of the samples fermented for 14 days; C and D are the OPLS-DA score plots and model validation plots of the samples fermented for 28 days. OPLS-DA can evaluate the classification effect of the model using R2X, R2Y, and Q2. R2X represents the interpretability of the model (for the X variable dataset); R2Y represents the interpretability of the model (for the Y variable dataset); Q2 represents the predictability of the model.

[0021] Figure 4 Shown is the change in the microbial community during the fermentation of Qing Fang.

[0022] Among them, A. Relative abundance of bacteria at the genus level; B. Relative abundance of bacteria at the species level; C. Relative abundance of fungi at the genus level; D. Relative abundance of fungi at the species level.

[0023] Figure 5 Shown is the PLS-DA analysis of microorganisms during the fermentation of Qing Fang.

[0024] Among them, A. PLS-DA of bacteria in groups M and N; B. PLS-DA of fungi in groups M and N.

[0025] Figure 6 Shown is the LEfSE analysis of microorganisms during the fermentation of Qing Fang.

[0026] Figure 7 Shown is the LEfSE analysis of microorganisms during the fermentation of Qing Fang.

[0027] Figure 8 Shown is the correlation analysis between the main microorganisms during the fermentation process.

[0028] Figure 9 Shown is the relative abundance map of bacteria at the species level in group M.

[0029] Figure 10 Shown is the relative abundance map of fungi at the genus level in group M.

[0030] Figure 11 Shown is the Venn diagram of the change in the microbial species composition during the fermentation process in group M.

[0031] Figure 12 Shown is the KEGG pathway of microorganisms during the fermentation process.

[0032] Figure 13 Shown is the LEfSe of the KEGG pathway of microorganisms during the fermentation process.

[0033] Figure 14 Shown is the prediction of the amino acid metabolic pathway during the fermentation process.

[0034] Figure 15The content heatmap and clustering of volatile compounds in Qing Fang are shown; blue indicates the lowest content, and red indicates the highest content;

[0035] Figure 16 The number of volatile compounds in groups M and N is shown;

[0036] Figure 17 The OPLS-DA score plots of volatile compounds in groups M and N are shown; among them, A. Fermentation for 14 days; B. Fermentation for 28 days;

[0037] Figure 18 The VIP values of differential metabolites in groups M and N are shown; among them, A. Fermentation for 14 days; B. Fermentation for 28 days;

[0038] Figure 19 The content heatmap of 33 core flavor substances with ROAV≥0.1 during the fermentation process is shown;

[0039] Figure 20 The PCA plots of core flavor substances in groups M and N are shown; among them, A. Fermentation for 14 days; B. Fermentation for 28 days;

[0040] Figure 21 The flavor recombination of group M is shown; Fishy, fishy smell; Fruity, fruity smell; Gasoline-like, gasoline smell; Grassy, grassy smell; Musty, musty smell; Savory, savory taste; Sour, sour taste; Sulfurous, sulfur smell; Recombination1, recombination model 1; Recombination2, recombination model 2;

[0041] Figure 22 The RDA analysis of microorganisms and physicochemical properties during the fermentation process is shown;

[0042] Figure 23 The correlation analysis among core microorganisms, key flavor substances and free amino acids is shown;

[0043] The lines in the circles indicate the correlation between flavor substances and microorganisms (|r|>0.6, P<0.05); the red lines indicate positive correlation; the blue lines indicate negative correlation; Esters, esters; Aldehydes&Ketones, aldehydes and ketones; Others, other compounds; Genus, genus; Bitterness, bitterness; Sweetness, sweetness; Umami, umami. Detailed implementation methods

[0044] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] Embodiment

[0046] 1. Preparation of Qingfang fermented bean curd

[0047] The fermentation of Qingfang fermented bean curd is divided into the early fermentation of the tofu blank and the later fermentation of adding brine. The tofu blank used in this embodiment (fresh tofu is sprayed with a mixed suspension of Mucor and Rhizopus, placed on a sterile wooden tray, and fermented in a 28 °C fermentation chamber for 7 days, then dry-salted with salt for 1 day, and the ratio of the tofu blank to salt is 10:1; then soaked in 20% brine to make a fresh salt blank with a salt concentration of 12%-14%) and the fermentation broth (a mixture of brine and yellow slurry water, where the yellow slurry water is a by-product of tofu production) are both prepared and collected in local factories. Weissella confusa M1 is inoculated into the fermentation broth at a concentration of 1×10 6 CFU / ml. The salt blank is placed in a 350 mL glass jar, filled with the fermentation broth inoculated with M1, and the mass ratio of the salt blank to the fermentation broth is 7:3, which is used as the Weissella confusa M1 group (Add Weissella confusa M1, abbreviated as M). After the jar is sealed, it is fermented later at 28 °C for 28 d; a naturally fermented Qingfang is prepared as a control group (Natural, abbreviated as N) under the same conditions. Except for not adding M1, other conditions are the same as those of the Weissella confusa M1 group. Qingfang is taken on the 0th, 7th, 14th, 21st, and 28th days of fermentation and ground, and the samples are stored at -80 °C before analysis.

[0048] 2. Detection of physical and chemical properties

[0049] 2.1 Measurement of acidity and alkalinity

[0050] According to the manufacturer's instructions, a digital pH meter (3C pH meter of Shanghai Youke Instrument Co., Ltd., China) is used to measure the acidity and alkalinity of the samples. The results are as Figure 1 shown in A. The pH values of the control group (N) and the Weissella confusa M1 group (M) increase with the prolongation of fermentation time.

[0051] 2.2 Measurement of salt content

[0052] The salt content is analyzed by the silver nitrate titration method in the Determination of Chloride in Foods (GB 5009.44-2016) of the Chinese national food standard. The results are as Figure 1As shown in Figure B, the salt content in the control group (N) and Weissella M1 group (M) increased with the prolongation of fermentation time.

[0053] 2.3 Determination of Amino Acid Nitrogen (AAN), TCA-Soluble Peptide Content and Protease Activity

[0054] Determination of amino acid nitrogen: After dissolving the sample in water, it was titrated with a standard sodium hydroxide solution for determination.

[0055] Determination of soluble peptides: The sample was mixed with 5% TCA solution, homogenized for 1 min, placed in an ice bath for 1 h, centrifuged at 4 °C for 5 min after the ice bath, and the absorbance of the supernatant was measured at 680 nm.

[0056] Determination of protease activity: The Qingfang sample was mixed with 0.1 M Tris-HCl buffer solution and allowed to stand for 30 min to obtain an enzyme extract. 2% casein solution and the enzyme extract were added to a test tube and mixed evenly, incubated in a water bath at 40 °C for 10 min, immediately mixed with 0.4 M TCA solution to terminate the hydrolysis reaction, and then centrifuged at 1914 g for 10 min. The supernatant was reacted with 0.4 M sodium carbonate and Folin-Ciocalteu reagent at 40 °C for 20 min, and the absorbance was measured at 680 nm. Protease activity was expressed as tyrosine equivalents in the supernatant.

[0057] Amino acid nitrogen, TCA-soluble peptides and proteases reflect the fermentation degree and taste characteristics of fermented foods, which are very important for the quality of fermented foods. Overall, the protease activity, TCA-soluble peptide and amino acid nitrogen concentrations in all samples showed a significant upward trend ( Figure 1 Figures C, 1D, 1E), indicating that soy protein was decomposed into amino acids under the action of microorganisms. Among them, at 28 days of fermentation in group M, the protease activity, TCA-soluble peptides and amino acid nitrogen were all significantly higher than those in the control group (P < 0.01; P < 0.05; P < 0.05), indicating that M1 promoted the fermentation process, produced more amino acids, and made the taste of Qingfang fermented bean curd more delicious.

[0058] 2.4 Chromaticity Evaluation

[0059] A color difference meter CR-400 (Konica Minolta, Shanghai, China) was used to evaluate the color change, and the brightness value (ΔL*), red value (Δa*), and yellow value (Δb*) were recorded. The Euclidean distance formula (ΔE* = √(ΔL*) 2 +(Δa*) 2 +(Δb*) 2 ) was used to evaluate the color difference during fermentation. The color changes of Qingfang fermented bean curd at different fermentation times are shown in Table 1.

[0060] Table 1

[0061]

[0062] Note: n = 3, different marked letters indicate significant differences (P < 0.05).

[0063] Overall, the total color difference (ΔE*) of the blue - mold fermented bean curd showed a downward trend during the fermentation process. As the fermentation proceeded, the a* value of the blue - mold fermented bean curd showed a downward trend (the green color deepened). Among them, the a* value of group M was significantly lower than that of the control group from 14 d (P < 0.01), indicating that M1 accelerated the fermentation process to a certain extent ( Figure 1 F, 1H).

[0064] 2.5 Sensory evaluation

[0065] Ten members of a panel who had received regular sensory evaluation training (five males and five females; average age 27 years) were selected to evaluate the appearance color, fragrance, taste (texture), flavor, and overall sensory properties of the blue - mold fermented bean curd. The flavor characteristics of the blue - mold fermented bean curd were evaluated through the average value of the final scores. The results are as Figure 1 shown in G, indicating that the quality of the blue - mold fermented bean curd with M1 added was better than that of the control group.

[0066] 3. Determination of amino acids

[0067] 3.1 Amino acid extraction

[0068] Accurately weigh an appropriate amount of the sample into a 2 - ml EP tube, accurately add 600 μL of 10% formic acid methanol solution - H2O (1:1, V / V) solution, add 2 steel beads, and vortex for 30 s; put it into a tissue grinder and grind at 60 Hz for 90 s; centrifuge at 12000 rpm at 4 °C for 5 min, take 10 μL of the supernatant, add 990 μL of 10% formic acid methanol - H2O (1:1, V / V) solution, vortex for 30 s, take 100 μL of the diluted sample, add 100 μL of an isotope internal standard with a concentration of 1000 ppb and vortex for 30 s, filter the supernatant through a 0.22 - μm membrane, and add the filtrate to the detection bottle.

[0069] 3.2 LC - MS detection

[0070] Analysis was performed using a Waters UPLC liquid chromatograph (Waters ACQUITY UPLC) and an AB 4000 triple quadrupole mass spectrometer (AB 5000). An ACQUITY UPLC BEH C18 chromatographic column (2.1×100 mm, 1.7 μm, Waters Corporation, USA) was used, with an injection volume of 5 μL, a column temperature of 40 °C, mobile phase A - 10% methanol in water (containing 0.1% formic acid), and B - 50% methanol in water (containing 0.1% formic acid). The gradient elution conditions were as follows: 0 - 6.5 min, 10 - 30% B; 6.5 - 7 min, 30 - 100% B; 7 - 14 min, 100% B; 14 - 17.5 min, 100 - 10% B. The flow rate was 0.3 ml / min from 0 - 8.0 min and 0.4 ml / min from 8.0 - 17.5 min. Mass spectrometry conditions: electrospray ionization (ESI) source, positive ion ionization mode. The ion source temperature was 500 °C, the ion source voltage was 5500 V, the collision gas was 6 psi, the curtain gas was 30 psi, and the nebulizing gas and auxiliary gas were both 50 psi. Multiple reaction monitoring (MRM) was used for scanning. The concentration of amino acids was determined according to the established standard curve.

[0071] Calculation of 3.3 TAV

[0072] The calculation of the taste activity value of free amino acids is as follows:

[0073]

[0074] Where C is the concentration of a single compound; threshold is the corresponding taste threshold of the single compound.

[0075] The amino acid content is a key indicator in Qingfang fermentation. As Figure 2 shown, a total of 21 amino acids were detected in two groups, including 16 free amino acids. OPLS analysis showed that there were significant differences in the physical and chemical indexes of the groups fermented with M1 for 14 d and 28 d compared with the control group ( Figure 3 A, 3C). Figure 3 Figures B and 3D show that the principal components can explain 99.6% and 100% of the variation in the Y variable, and the predictability of the model is 96.1% and 99.7%, indicating that the OPLS model was successfully established and can be used to distinguish differences between groups. Among them, the content of free amino acids in the M group was significantly higher than that in the control group, and the content was the highest at 28 d (Table 2).

[0076] Table 2

[0077]

[0078] Note: n = 3, different marked letters indicate significant differences (P < 0.05); "*", essential amino acids; TAA, total free amino acid content; EAA, essential amino acid content; DAA (umami amino acids), Glu + Asp + Gly + Ala; SAA (sweet amino acids), Gly + Ala + Ser + Thr + Pro + Arg; BAA (bitter amino acids), Lys + Met + Val + Ile + Leu + Tyr + His + Phe; sAA (salty amino acids), Glu + Asp; AAA (sour amino acids), Glu + Asp + His.

[0079] According to the taste characteristics of free amino acids, they were divided into umami amino acids (DAA), sweet amino acids (SAA), bitter amino acids (BAA), salty amino acids (sAA), and sour amino acids (AAA). The results in Table 2 showed that the contents of all taste amino acids increased with the progress of fermentation time. In addition, substances with TAV ≥ 1 contributed greatly to the taste. As shown in Table 2, except for arginine (TAV = 0), other free amino acids had high TAV values, endowing the Qingfang sufu with rich taste characteristics. After the fermentation with the addition of M1 ended, the taste amino acids in the Qingfang were all higher than those in the control group, especially Asp and Glu (P < 0.05), strengthening the umami taste of the Qingfang.

[0080] 4. Metagenomic sequencing of the microbial community in Qingfang sufu

[0081] 4.1 Extraction of metagenomic DNA and shotgun sequencing

[0082] The total microbial genomic DNA of each sample was extracted using the OMEGA soil DNA kit (D5625-01) according to the manufacturer's instructions and stored at -20 °C for further evaluation. The quantity and quality of the extracted DNA were measured using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis, respectively. The extracted microbial DNA was processed using the Illumina TruSeq Nano DNA LT Library Preparation Kit to construct a metagenomic shotgun sequencing library with an insert size of 400 bp. Each library was sequenced using the PE150 strategy on the Illumina HiSeq X-ten platform (Illumina, USA) by Shanghai Personal Biotechnology Co., Ltd.

[0083] 4.2 Metagenomic analysis

[0084] The raw sequencing reads were processed to obtain quality-filtered reads for further analysis. First, the sequencing adapters were removed from the sequencing reads using Cutadapt (v1.2.1). Second, the low-quality reads were trimmed using the sliding window algorithm in fastp. Once the quality-filtered reads were obtained, the metagenomic sequencing reads of each sample were taxonomically classified using Kraken2 against a database from RefSeq, which includes genomes from archaea, bacteria, viruses, fungi, protozoa, metazoa, and green plants. The reads assigned to metazoa or green plants were removed for downstream analysis. Megahit (v1.1.2) was used to assemble for each sample using the super-large preset parameters. Then the generated contigs (longer than 200 bp) were pooled together and clustered using mmseqs2 with the "easy-linclust" mode, setting the sequence identity threshold to 0.95 and covering the residues of the shorter contigs to 90%. The lowest common ancestor classification of the non-redundant contigs was obtained by aligning them with the NCBI-nt database using mmseqs2 in the "classify" mode, and the contigs assigned to Viridiplantae or Metazoa were removed in the following analysis. The CDS sequences of all samples were clustered by mmseqs2 in the "easy-cluster" mode, setting the protein sequence identity threshold to 0.90 and covering the residues of the shorter contigs to 90%. To evaluate the abundance of these genes, the high-quality reads from each sample were mapped to the predicted gene sequences using "salmon" based on the quasi-mapping mode, "--meta–minScoreFraction = 0.55", and CPM (copy per million mapped reads per kilobase) was used to normalize the abundance values in the metagenome. The functions of the non-redundant genes were obtained by annotating them against the protein databases of the KEGG, EggNOG, and CAZy databases respectively using mmseqs2 in the "search" mode. EggNOG and GO were obtained using EggNOG-mapper (v2). The GO ontology was obtained using map2slim (www.metacpan.org). KO was obtained using KOBAS.

[0085] The results showed that the addition of M1 could enhance the flavor-related physicochemical indexes in the blue cheese because M1 drove the changes in the microbial community structure during fermentation. As Figure 4As shown in Figure A, at the genus level, the microorganisms in the control group were mainly Leuconostoc (36.04%), Enterococcus (13.00%), Enterobacter (5.63%), Tetragenococcus (3.26%), Weissella (0.77%), etc. The main genera in the Qing Fang with added M1 were Leuconostoc (34.79%), Enterococcus (10.45%), Enterobacter (8.66%), Lactococcus (6.84%), Tetragenococcus (3.57%), etc. Compared with the control group, Enterobacter in the Qing Fang with added M1 was significantly up-regulated (P < 0.01), while Weissella and Bacillus were significantly down-regulated (P < 0.05). Enterobacter, Weissella, and Bacillus are the core flora during the natural fermentation of soy products and Daqu, have a close relationship with various flavor substances, and play an important role in maintaining community balance. Therefore, it can be speculated that the introduction of M1 is an important reason for the difference in the flora compared with the control group. At the species level, the microorganisms in the control group were mainly Leuconostoc citreum (14.50%), Enterococcus faecalis (8.68%), Leuconostoc pseudomesenteroides (7.17%), Leuconostoc lactis (7.79%), Leuconostoc garlicum (4.97%), etc.; the dominant bacteria in the Qing Fang with added M1 were Leuconostoc citreum (11.69%), Leuconostoc pseudomesenteroides (8.33%), Leuconostoc lactis (7.78%), Enterococcus faecalis (6.24%), Lactococcus lactis (5.44%), etc.( Figure 4 B). The abundance of Enterococcus faecalis in the Qing Fang with added M1 was significantly lower than that in the control group (P < 0.05); Enterococcus faecalis has been proven to produce a large amount of methoxy-phenyl oxime during the fermentation of soybeans, which is the main source of musty smell. Therefore, the introduction of M1 may reduce the musty smell of the Qing Fang. From the fungal community perspective( Figure 4 C, 4D), the Qing Fang with added M1 was similar to the control group. At the genus level, it was mainly Trichosporon (99.08%), and at the species level, it was mainly Trichosporon asahii (99.10%).

[0086] Figure 5 A and 5B are PLS-DA ordination plots based on bacteria and fungi respectively. It can be seen that M1 significantly changed the bacterial composition but had no effect on the fungal composition. Only 7 species of fungi were detected in the Qing Fang, and the relative abundance of Trichosporonasahii ranked 108th among all detected microbial species, indicating that bacteria are the main reason for the product quality differences between the M group and the control group.

[0087] LEfSe analysis also showed significant differences in the species composition between M1 and the control group ( Figure 6 ). Bacteria with significant changes after adding M1 were determined according to the LDA score ≥ 2, including Enterococcus_sp_6C8_DIV0013, Shewanella_chilikensis, Tetragenococcus_osmophilus, etc. at the species level, and they were determined as biomarkers after adding M1 ( Figure 7 ).

[0088] Interestingly, it was found through analysis that Weissella confuse is a biomarker of the control group, not the M group. As Figure 8 can be seen, Weissella confuse has both positive correlations and competitive relationships with Leuconostoc citreum and Leuconostocpseudomesenteroides, etc., which may thus affect the abundance of Weissellaconfuse. In addition, it was also found that Leuconostoc citreum, Leuconostoc pseudomesenteroides and Weissella confuse are related to multiple species, thus affecting the entire microbial community structure. In summary, it shows that the introduction of M1 has changed the microbial structure of the Qing Fang differently from that of the control group.

[0089] To further explore the dynamic changes of microorganisms in the M1-fermented Qing Fang, the microbial structures at different fermentation stages after adding M1 were characterized. As Figure 9As shown, the main microorganisms on the 0th day were Leuconostoc lactis (13.19%), Lactococcus lactis (10.42%), and Enterococcus faecalis (7.99%). After 7 days of fermentation, the microbial structure of the Qingfang changed significantly, mainly Leuconostoc citreum (23.77%) and Leuconostoc pseudomesenteroides (14.70%). The abundance of lactic acid bacteria increased significantly at this stage, which may be due to the higher acidic conditions in the initial stage of fermentation promoting their growth. However, with the progress of fermentation, the infiltration of salt and anaerobic environment inhibited their growth. Therefore, starting from the 14th day of fermentation, the abundances of Leuconostoc citreum and Leuconostoc pseudomesenteroides began to gradually decline. On the 14th day of fermentation, the abundances of these two genera were 21.68% and 14.70% respectively; on the 21st day of fermentation, the abundances decreased to 14.27% and 15.94%; on the 28th day of fermentation, the abundances reached the lowest, 1.70% and 10.20% respectively. Figure 11 The Venn diagram shown indicates the species composition during the fermentation of M1. A total of 1213 microorganisms were identified, including 1159 bacteria, 7 fungi, and 47 viruses. Among them, the species common throughout the fermentation period were 394, and the unique species were 204 (0 days), 34 (7 days), 42 (14 days), 88 (21 days), and 24 (28 days) respectively, indicating that with the progress of fermentation, the abundance of species in the Qingfang generally showed a downward trend, which may be due to the inhibition of the growth of many species by the infiltration of salt into the Qingfang.

[0090] 3.2.2 Microbial Functional Annotation and Metabolic Pathway Prediction in Fermented Qingfang with M1

[0091] Furthermore, the KEGG database was used to annotate the functions of microorganisms in the Qingfang under the intervention of M1. The results showed that during the fermentation of the Qingfang, microorganisms participated in a variety of metabolic and energy conversion processes, mainly amino acid and carbohydrate transport and metabolism ( Figure 12 ). In addition, according to the LDA score ≥ 2, the significantly up-regulated metabolic pathways after adding M1 were mainly the alanine - aspartate and glutamate metabolic pathways, and the arginine and ornithine metabolic pathways ( Figure 13 ), indicating that there is a close connection between the differential microorganisms formed by M1 and free amino acid metabolism. Protein degradation and the synthesis of volatile compounds are crucial for improving the flavor and taste of fermented soy products. Therefore, based on the amino acid detection results ( Figure 2) and the KEGG annotation results were used to predict and restore the upstream and downstream relationships of amino acid metabolism during the M1 fermentation process. The anabolic pathways of 21 amino acids were included in these pathways, and the synthesis pathways of glutamic acid and aspartic acid were the most ( Figure 14 ), which also explained the high content of glutamic acid and aspartic acid at the end of the M1 fermentation. The synthesis pathways of glutamic acid and aspartic acid mainly include: (1) Glutamic acid synthase [EC:1.4.1.13]gltB, [EC:1.4.1.13]gltD, and [EC:1.4.1.13]GLT1 produced by Escherichia coli act on glutamine to synthesize glutamic acid, etc.; (2) After phenylalanine generates hydrocyanic acid, it generates β-cyanoalanine under the action of [EC:2.5.1.47 4.4.1.9]ATCYSC1, and then further generates asparagine under the action of [EC:3.5.5.44 2.1.65]NIT4, and finally synthesizes aspartic acid under the action of [EC:3.5.1.13 4.19.5]ASRGL1. Leucine (Leu), isoleucine (Ile), methionine (Met), and phenylalanine (Phe) will be degraded by transaminases and decarboxylases produced by microorganisms such as Enterobacter and Leuconostoc, and then produce volatile flavor substances such as ketones, aldehydes, and alcohols.

[0092] In summary, it can be seen that the dominant microorganisms formed during the M1 fermentation process are closely related to amino acid metabolism and flavor formation. Although the abundance of fungi is not high during the fermentation process, it is known that fungi such as Trichosporon are related to various flavors. Therefore, 20 bacterial genera ( Figure 9 ) and 8 fungal genera ( Figure 10 ) were defined as the key microorganisms in the M1 fermented blue moldy tofu.

[0093] 5. Detection of Volatile Flavor Substances during the Fermentation Process of Blue Moldy Tofu

[0094] 5.1 Extraction of Volatile Metabolites

[0095] The frozen samples were ground, 1 g was weighed and mixed with internal standards (30 μL of 40 μg / ml 2-octanol and 10 μL of 100 μg / ml dichlorobenzene), and transferred to a headspace vial and shaken well. A DVB / CAR / PDMS fiber head (2 cm) was used, incubated at 40 °C for 10 min, extracted for 30 min, and desorbed in the injection port for 5 min. Then, GC×GC-TOFMS analysis was performed according to the set parameters.

[0096] 5.2 GC×GC-TOFMS Analysis

[0097] Volatile compounds were analyzed using an Agilent 7890B gas chromatograph (Santa Clara, CA) and a LECO Pegasus 4D mass spectrometer. The first - dimension column was DB - WAX (30 m × 250 μm × 0.25 μm); the injection temperature was 250 °C; the initial temperature was 40 °C and held for 3 min, then increased to 250 °C at a rate of 5 °C / min and held for 5 min; helium (99.9999%) flow rate was 1.0 ml / min; splitless mode; the second - dimension column was DB - 17MS (2 m × 100 μm × 0.10 μm); the column temperature was 5 °C higher than that of the first - dimension column; the modulator temperature was always 5 °C higher than that of the second - dimension column. Mass spectrometry conditions: modulation period was 6.0 s for comprehensive two - dimensional analysis; interface temperature was 270 °C; ion source temperature was 250 °C; electron impact source was 70 eV; detector voltage was 1680 v; acquisition rate was 50 spectra / s; mass spectrometry scan range was 33 - 500 m / z. The identified volatile compounds were compared with the aroma compounds in the National Institute of Standards and Technology (NIST) database, and volatile compounds with a similarity ≥80% to the NIST database were considered. The concentration of volatile compounds was determined based on the ratio of their peak areas to the peak area of the internal standard, and the differences of all volatile compounds relative to the internal standard were standardized. Each sample was measured 3 times, and the average value was calculated to two decimal places.

[0098] As Figure 15 shown, a total of 272 volatile compounds were identified in all Qing - Fang samples, including 66 esters, 56 alcohols, 35 hydrocarbons, 32 ketones, 29 aldehydes, 10 acids, 10 furans, 5 phenols, and 29 other types of volatile compounds ( Figure 16 ). It can be seen that the main volatile compounds in Qing - Fang are esters and alcohols. Esters are mainly formed during microbial metabolism through enzymatic or non - enzymatic reactions of glucose and amino acids to produce alcohols and acids, and then through esterification reactions. They mostly have fruity aromas. Alcohol compounds may be produced during the metabolism of carbohydrates and the decarboxylation and dehydrogenation of amino acids during the maturation of soybeans, giving Qing - Fang a grassy, fruity, and sweet taste. In addition, many other types of compounds, such as indole, dimethyl sulfide, and dimethyl trisulfide, were also detected in Qing - Fang, presenting pungent odors such as burnt, fishy, and sulfurous smells, which are the main sources of the stinky smell in Qing - Fang. They mainly come from the degradation of methionine and tryptophan in the middle stage of fermentation ( Figure 2 ). Secondly, as can be seen from Figure S15, the flavor substances of the three parallel samples in the naturally fermented Qing - Fang (control group) fluctuated greatly, while after adding M1, the changes in flavor substances among the parallel groups were less different, indicating that M1 is beneficial to maintaining the stability of flavor during the fermentation process of Qing - Fang.

[0099] As Figure 16As shown, 55 alcohols, 65 esters, 31 ketones, 29 aldehydes, 10 acids, 9 furans, 4 phenols, 34 hydrocarbons and 25 other compounds were detected at 28 d after adding M1. Seven out of these eight classes of compounds were higher than those in the control group (except for phenols), indicating that inoculating M1 can make the flavor of Qing Fang more abundant. In addition, OPLS analysis showed that there were significant differences in volatile compounds between Qing Fang inoculated with M1 at 14 d and 28 d and the control group ( Figure 17 A, 17B). The variable importance in the projection (VIP) represents the contribution of each metabolite to the model. Generally, volatile compounds with VIP≥1 are selected as signature flavor compounds, and the larger the VIP value of a compound, the greater its contribution to the flavor difference between the two groups; at 14 d and 28 d, there were 48 and 42 volatile compounds with VIP values≥1, respectively ( Figure 18 A, 18B). Among them, the VIP values of methanol, methyl 4-methyl-4-nitropentanoate, 2-butanol, ethanol, 4-oxopentanoate, dimethyl trisulfide and ethyl acetate were≥3, so they were considered as key differential metabolites.

[0100] 5.3 Identification and screening of key aroma compounds

[0101] The relative odor activity value was calculated by the following equation to determine the contribution of each volatile compound to the aroma characteristics:

[0102]

[0103]

[0104] where, OAV i is the odor activity value (OAV) of any volatile compound, and OAV max is the maximum odor activity value among all volatile compounds in each sample group; C i is the concentration of the volatile compound; T i is the odor threshold of the volatile compound. Compounds with 0.1≤ROAV<1 are considered to contribute to the aroma, while compounds with ROAV≥1 are considered as key aroma compounds. All independent experiments were carried out at least 3 times. The heatmap was plotted for the data set using the pheatmap package in R (v3.3.2).

[0105] A total of 27 volatile compounds with ROAV≥0.1 were identified in the samples added with M1 and their control group, including 2 alcohols, 3 ketones, 9 aldehydes, 6 esters and 7 other compounds, as shown in Figure 19Among them, the threshold of dimethyl trisulfide is relatively low, so it is considered to contribute the most to the flavor of Qing Fang during the fermentation process. It mainly presents an unpleasant odor and is the main source of the stench of Qing Fang. At the beginning of fermentation (day 0), Qing Fang contained 9 main volatile compounds (0.1 ≤ ROAV < 1) and 8 core volatile compounds (ROAV ≥ 1), which restored the aroma characteristics of Qing Fang before fermentation. Among them, 2-methylbutyraldehyde (ROAV = 100), dimethyl trisulfide (ROAV = 98.4), and 2,3-butanedione (ROAV = 9.18) contributed the most to its flavor, providing Qing Fang with grassy, salty and buttery aromas. 14 days after the start of fermentation, Qing Fang supplemented with M1 contained 12 main volatile compounds and 4 core volatile compounds, and the control group contained 13 main volatile compounds and 4 key volatile compounds. At this time, there were differences in the main flavor compounds between the two groups, and the PCA analysis also showed a separation trend in the composition of flavor compounds between the two groups (the cumulative interpretation amount of PC1 and PC2 was 75.45%). Figure 20 A). By the end of fermentation on day 28, the M group contained 11 main volatile compounds and 7 core volatile compounds, and the control group contained 12 main volatile compounds and 4 core volatile compounds, indicating that the introduction of M1 increased the variety of core flavor substances. At the same time, Figure 20 As can be seen in B, on the 28th day of fermentation, there was an obvious separation trend in the flavor composition of the two groups in the PCA score plot (the cumulative interpretation amount of PC1 and PC2 was 96.24%), indicating that at the end of fermentation, there were significant differences in the core flavor substances in Qing Fang supplemented with M1 compared with the control group. Specifically, the core compounds in the control group were dimethyl trisulfide (ROAV = 100), 2-methylbutyraldehyde (ROAV = 20.91), 3-methylbutyraldehyde (ROAV = 2.4), and methyl butyrate (ROAV = 1.17), which gave Qing Fang salty, grassy and fruity aromas. After fermentation with M1 added, on the basis of retaining the above flavor substances, Qing Fang added special flavor substances such as hexanal (ROAV = 1.01), ethyl 3-methylbutyrate (ROAV = 1.47), ethyl butyrate (ROAV = 4.78), and methyl butyrate (ROAV = 2.51). These compounds added pleasant aromas such as leaves, apples and pineapples to Qing Fang, making the product flavor more rich and palatable.

[0106] 5.4 Recombination and deletion of key aroma compounds

[0107] To verify the identification and screening of aroma compounds by OAV analysis, its results were recombined and compared with the aroma characteristics of the original samples.

[0108] First, freeze-dry the samples of the M groups fermented for 28 days, grind them, and then gradually extract them with methanol, ether, dichloromethane, and pentane respectively. After filtration and drying, a recombinant matrix is obtained.

[0109] Secondly, add 7 core flavor substances with ROAV≥1 in the M groups fermented for 28 days to the matrix at their actual concentrations to obtain Recombinant Model 1; at the same time, add the top five of the M groups with the highest aroma compound content to the matrix at their actual concentrations to obtain Recombinant Model 2. Then, equilibrate the test samples at room temperature for 10 min, and conduct sensory evaluation on the recombinant samples.

[0110] The results show that compared with Recombinant Model 2, the aroma characteristics of Recombinant Model 1 are the closest to those of the product ( Figure 21 ), indicating that the screening, identification, and quantification of core flavor substances by OAV analysis and GC-MS technology are successful.

[0111] 5.5 Omission test

[0112] Use the triangle test to conduct an omission test on Recombinant Model 1 to evaluate the contribution of a single compound to the aroma characteristics of the entire M-28d. The samples include complete recombinant samples and missing samples with one or a group of aroma-active compounds removed, which are randomly numbered for the triangle test. The sensory panel members (10 members), and the experiment is repeated three times.

[0113] As shown in Table 3, it is found that 2 omission models show extremely significant differences in aroma characteristics (P≤0.001), 1 omission model shows highly significant differences (P≤0.01), and the remaining 5 omission models show significant differences (P≤0.05). First, this experiment shows that the removal of 7 core flavor substances (Model 1: dimethyl trisulfide, 2-methylbutyraldehyde, n-valeraldehyde, n-hexanal, ethyl 3-methylbutyrate, ethyl butyrate, and methyl butyrate) leads to extremely significant differences in their aroma characteristics from the original samples (P≤0.001), indicating the key role of core flavor substances in the formation of the overall flavor of fermented tofu. Secondly, when there is no n-hexanal or no dimethyl trisulfide, there are also highly significant (P≤0.01) and extremely significant (P≤0.001) differences between Model 1-3 and 1-7 and the original samples. This also shows the great contribution of n-hexanal and dimethyl trisulfide to the flavor of fermented tofu.

[0114] Table 3

[0115]

[0116] a. Number of correct judgments of 10 panel members evaluating flavor differences according to the triangle test

[0117] b. *** (P≤0.001, extremely significant); ** (P≤0.01, highly significant); * (P≤0.05, significant)

[0118] Correlation Analysis between Physicochemical Properties, Key Microflora and Flavor Compounds in 5.6M1 Fermented Blue Mould Tofu

[0119] 5.6.1 Correlation Analysis between Physicochemical Properties and Key Microflora during M1 Blue Mould Tofu Fermentation

[0120] RDA was used to analyze the correlation between physicochemical properties and key microflora during the fermentation of M1 blue mould tofu. As Figure 22 shown, it was found that Enterobacter and Citrobacter were positively correlated with pH, salt content, protease activity, TCA-soluble peptides and amino acid nitrogen, especially Enterobacter, which indicated that Enterobacter played a crucial role in amino acid metabolism during fermentation. In addition, color difference was also an important index to judge the fermentation degree of blue mould tofu, and it was strongly positively correlated with Lactococcus, showing its important role during the fermentation of blue mould tofu. To sum up, the main microorganisms in blue mould tofu had their unique physiological characteristics, which could change the fermentation environment and continuously adapt to it.

[0121] In addition, the microflora on the 0th day and 28th day of M1 fermentation were relatively similar, and the microflora structures on the 7th day, 14th day and 21st day were relatively similar( Figure 22 ), which was consistent with the results of metagenomic sequencing. Among the main microbial genera, Enterobacter, Enterococcus, Lactococcus, and Leuconostoc were negatively correlated, while positively correlated with Citrobacter, which was consistent with the results Figure 5 shown.

[0122] 5.6.2 Correlation Analysis between Free Amino Acids, Key Flavor Compounds and Key Microflora in M1 Fermented Blue Mould Tofu

[0123] To clarify how M1 drives the changes in microorganisms and thus affects the flavor during the fermentation of blue mould tofu, Spearman correlation analysis was performed on the key bacterial genera (20 bacterial genera and 8 fungal genera), 16 free amino acids and 20 key flavor compounds (18 ROAV≥0.1 and 2 high-content volatile compounds) at the end of fermentation during the M1 fermentation process. The results showed that there were significant correlations between 23 genera (19 bacterial genera and 4 fungal genera) and 19 key flavor compounds and 15 free amino acids (|r|>0.6, P<0.05)( Figure 23 ), 20 genera (17 bacterial genera and 3 fungal genera) were related to key flavor substances, and 18 genera (14 bacterial genera and 4 fungal genera) were significantly related to free amino acids, indicating that bacteria contributed more to the flavor during the fermentation of blue mould tofu.

[0124] Among them, Lactococcus, Enterobacter, and Tetragenococcus were significantly positively correlated with 17 key flavor substances, mainly including esters and aldehyde compounds such as methyl acetate, methyl butyrate, and benzaldehyde, indicating that they were mainly related to lipid metabolism and carbohydrate metabolism during the fermentation of green brined tofu. Consistent with this, previous studies have shown that Tetragenococcus expressed aldehyde dehydrogenase [EC: 1.2.1.10], alcohol dehydrogenase [EC: 1.1.1.1], and branched-chain amino acid transaminase [EC: 2.6.1.42] related to fatty acid metabolism, and participated in the synthesis of aldehydes and ketones, which had a great impact on the flavor of fermented foods. In contrast, Staphylococcus and Pichia were significantly negatively correlated with 17 key flavor compounds, so they might play a direct or indirect inhibitory role in the flavor formation process.

[0125] Fourteen bacterial genera and three fungal genera were significantly positively correlated with free amino acids in green brined tofu. In particular, sweet and bitter amino acids such as proline, serine, threonine, and arginine were significantly correlated with a variety of microorganisms (|r|>0.6, P<0.05). In addition, Enterobacter, Escherichia, and Tetragenococcus were related to various amino acids, indicating that they were more involved in the synthesis and metabolism of free amino acids during fermentation. Among them, Enterobacter, Escherichia, and Tetragenococcus were significantly positively correlated with umami amino acids, which contributed to the formation of umami in green brined tofu. Similarly, studies have shown that Tetragenococcus had genes related to the metabolism and transport of proline, glutamate, and glutamine. In addition, although Escherichia, which might have potential pathogenic risks, was detected during fermentation, its abundance was low during fermentation and decreased over time, and would completely disappear after 45 days of fermentation. And in the previous functional annotation study, few genes related to human diseases and viruses were annotated ( Figure 12 ), indicating that the fermented green brined tofu produced under these fermentation conditions was safe.

[0126] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. Weissella confusa, characterized in that, Named M1, it is deposited in the General Microbiology Center of the China Committee for Culture Collection of Microorganisms, with the deposit number of CGMCC NO. 24526.

2. The blue-square starter culture, characterized in that, It includes Weissella confusa as described in claim 1.

3. Use of Weissella confusa as described in claim 1 or the blue cheese starter as described in claim 2 in the preparation of blue cheese.

4. According to the use described in claim 3, characterized in that the Weissella confusa is used to improve the microbial structure of blue cheese.

5. According to the use described in claim 3, characterized in that the Weissella confusa is used to adjust the microbial metabolic pathways of blue cheese.

6. According to the use described in claim 5, characterized in that the microbial metabolic pathways include alanine - aspartate and glutamate metabolic pathways, and the metabolic pathways of arginine and ornithine.

7. According to the use described in claim 3, characterized in that the Weissella confusa is used to regulate the synthesis of flavor compounds.

8. A Qing Fang, characterized in that, It is prepared by fermentation using Weissella confusa as described in claim 1 or the blue cheese starter as described in claim 2.

9. A method for preparing blue cheese, characterized in that the Weissella confusa as described in claim 1 or the blue cheese starter as described in claim 2 is inoculated into the fermentation broth, and the fermented and salted tofu embryo is added, and fermented at 28 ± 2 °C for 28 - 30 days.

10. According to the method for preparing blue cheese described in claim 9, characterized in that The inoculum size of Weissella confusa is 1×10 6 CFU / mL.

Images