WO2023022662A2 - Culture for improvement of the quality of soy sauce moromi - Google Patents

Abstract

Disclosed herein are cultures comprising bacterial strains and uses thereof. Also disclosed herein are processes of fermentation, as well as soy sauce products, methods using the bacterial strains disclosed herein.

Description

CULTURE FOR IMPROVEMENT OF THE QUALITY OF SOY SAUCE MOROMI

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority of Singapore provisional application no. 10202109142R, filed 20 August 2021, the contents of it being hereby incorporated by reference in its entirety for all purposes.

FIELD OF THE INVENTION

[0002] The present disclosure relates to the field of food science. In particular, the present disclosure relates to the use of microbes to manipulate flavour.

BACKGROUND OF THE INVENTION

[0003] Demand for healthy fermented food is growing worldwide, with soy sauce being one of the primary condiments in Asian countries. China’s annual production of soy sauce was shown to have reached more than 8 million tons per annum, as reported in 2022. While there are range of various products on the market covering customers with specific flavour preferences, there is still a demand for natural food products with enhanced taste characteristic. This enhanced taste characteristic coincides with a higher content of peptides with rich umami (meaty) taste and kokumi mouthfeel experience, without employing artificial additives such as sodium glutamate. Since food allergies are rising within the growing, urbanized population, further concerns of potential customers would be being access gluten free foods and products with depleted histamine content.

[0004] Thus, there is an unmet need for an additive-free method of increasing of taste peptides present in soy sauce.

SUMMARY OF THE INVENTION

[0005] In one aspect, the present disclosure refers to a culture comprising Lactobacillus pobuzihii WZ3 (DSM 33648), or Lactobacillus pobuzihii WZ5 (DSM 33658), or a combination thereof.

[0006] In another aspect, the present disclosure refers to a process of fermentation comprising a starter culture comprising at least one Lactobacillus pobuzihii strain.

[0007] In yet another aspect, the present disclosure refers to a soy sauce product made using the culture as disclosed herein or the process as disclosed herein.

[0008] In a further aspect, the present disclosure refers to a method of reducing and/or suppressing growth of undesired bacteria during fermentation, the method comprising the use of the culture as disclosed herein.

[0009] In one aspect, the present disclosure refers to a method of lowering undesirable taste components in fermented black bean soy sauce, the method comprising use of the culture as disclosed herein. [0010] In yet another aspect, the present disclosure refers to a method of increasing desirable taste components in fermented black bean soy sauce, the method comprising use of the culture as disclosed herein, wherein the increase in desirable taste components is characterised by the increase in taste conferring peptides and/or taste conferring amino acids in relation to total protein content of a sample.

[0011] In yet another aspect, the present disclosure refers to a method of reducing the concentration of Weissella sp. during fermentation, the method comprising the use of the culture disclosed herein.

[0012] In one aspect, the present disclosure refers to Lactobacillus pobuzihii WZ3 (DSM 33648).

[0013] In another aspect, the present disclosure refers to Lactobacillus pobuzihii WZ5 (DSM 33658).

BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The invention will be better understood with reference to the detailed description when considered in conjunction with the non-limiting examples and the accompanying drawings, in which:

[0015] Fig. 1 shows a histogram depicting the results of a comparison of taste grades between various soy sauce types and brands. Specifically, the quality of Taiwanese black bean Wuan Chuang soy sauce is shown compared to other products, thereby showing an overall quality of Taiwanese black bean Wuan Chuang soy sauce compared to other soy sauces available on the Chinese market.

[0016] Fig. 2 is a column graph showing the accumulation of taste 3-6 mer peptides over time during a black bean moromi fermentation.

[0017] Fig. 3 is a column graph showing the development of the level of tasty peptides over 4 months. This shows that the abundance of Lactobacillus pobuzihii correlates with level of tasty peptides over 4 months. Lactobacillus pobuzihii and Staphycoccus bacteriophages were the only ones to be positively associated with taste peptides.

[0018] Fig. 4 is a column graph showing the concentration of valine-proline-proline (VPP) in black bean (BB) compared to yellow bean (V) and wheat (YW) soy sauce moromi made with WZ. VPP is known to have a blood pressure lowering effect by inhibiting angiotensin converting enzymes (ACS). The valine-proline-proline (VPP) peptide content is indicative of being an inhibitor of antiotension converting enzymes in Wuang Zhong (WZ) soy sauces.

[0019] Fig. 5 is a heatmap showing the results of a comparison of the volatile organic compounds obtained from traditional black bean moromi (BB) with yellow bean (YB), industrial yellow bean, and wheat (YW) moromi for the final 5 months of fermentation. Black bean final moromi accumulated pyrazines, various organic compounds with butane chain, acetic acid, acetone, propanoic acid. Levels of octanoic acid are high at the first month, and disappeared at later stages of fermentation.

[0020] Fig. 6 is a heatmap showing the chemical product profile of fermentation with various bacterial strains. Lactobacillus pobuzihii produces 3 hydroxy-2-butanone, butanoic acid, pentyl octanoate and acetic acid. Weissella paramensteroides is responsible for octanoic acid, ethyl acetate, methylbutanols. Tetragenococcus halophilus releases benzaldehydes, pentanedione, toluene and butenal. Bacillus amyloliquefaciens produces pyrazines compounds. [0021] Fig. 7 is a column graph showing that bacteria and bacteriophages are negatively associated with taste peptides. Shown are the bacteria Bacillus amyloliquefaciens , Escherichia coli, Weissella cibaria, Enterococcus casseliflavus, and Ochrobactrum.

[0022] Fig. 8 is a line graph depicting the optical density of various bacterial strains over time. Strains of Lactobacillus pobuzihii WZ3 and WZ5 are shown to have different growth characteristics, which affects characteristics, such as stress tolerance and the ability to grown under 17% salt during moromi fermentation.

[0023] Fig. 9 shows the relative abundances of microbial members responsible for the quality of koji and moromi of black bean soy sauce. The sampling event had taken place in March 2016 Fig. 9A is a horizontal stacked bar chart showing the microbial composition based on shotgun metagenomics of black bean soy sauce fermentation during 4 months of fermentation. Samples were taken from vats at timepoints of 0, 0.5, 1, 2, 3 and 4 months in March 2016 (Spring) in three replicates (rep 1- vatl, rep2 - vat2, rep3 - vat3). Bioinformatics analysis was conducted with Metaphlan2 pipeline as input metagenomics reads. Viral composition is omitted. Fig. 9B is another stacked bar chart horizontal stacked bar chart showing the results of amplicon sequencing output on microbial composition of black bean soy sauce fermentation during 4 months of fermentation. Samples were taken from vats back to 0, 0.5, 1, 2, 3 and 4 months starting in March 2016 (Spring) in three replicates (rep 1- vatl, rep2 - vat2, rep3 - vat3). Fig. 9C shows a line graph showing a summary of the growth dynamic of major bacterial species during koji maturation of black beans. Fig. 9D shows a line graph depicting a summary of the growth dynamic of four most abundant bacterial species in black bean moromi.

[0024] Fig. 10 shows the relative abundances of microbial members together with the profiles showing the bacteriophage diversity in koji and moromi of black bean soy sauce. The sampling event had taken place in March 2016. Fig. 10A is a heatmap depicting the total population of bacterium present in the initial sampling of traditional black bean soy sauce. Bacteriophages of Weissella, Enterobacteria and Cronobacter had an abundance of almost 60%. Fig. 10B shows a stacked bar chart showing the identity and relative amounts of DNA as determined by direct DNA extraction from 5g of koji and 500 mL of sample. This figure shows the relative abundances of microbial taxa, such as viruses, bacteria and eukaryota in koji and moromi of black bean soy sauce. The sampling event had taken place in March 2016. Fig. 10C shows a stacked bar chart depicting the relative abundance of archaea, eukaryotic, viral, and bacterial DNA present in a pellet obtained from the 500 mL of sample in a subsequent DNA extraction step.

[0025] Fig. 11 shows the dynamics of main microbial functions (based on shotgun metagenome reads) changing over period of soy sauce maturation from koji until end of 4 months of moromi maturation. The sampling event had taken place in March 2016. Fig. 11A shows a heatmap of the metabolic pathways activated during koji and moromi fermentations, based on HUMANN2 output. Fig. 11B shows a cladogramm depicting the pathway contribution of Lactobacillus pobuzihii, Bacillus amyloliquefaciens, Tetragenococcus halophilus, Weisella paramensteroides, and Staphylococcus scuiri at koji (time point 0) and during 4 months of black bean moromi fermentation in with following timepoints: 0.5 month - timepoint 1, 1 month - timepoint 2, 2 months - timepoint 3, 3 months - timepoint 4, 4 months - timepoint 5. Months of fermentation are coloured. Folate production happened in koji stage with help of Bacillus and Staphylococcus. Production of methionine, threonine at 2 months; isoleucine and arginine at 5 months. V pentose phospate cycle is solely shared between 4 species.

[0026] Fig. 12 shows the relative abundances of microbial members responsible for the quality of koji and moromi of black bean soy sauce (BB), yellow soy sauce (Yellow) and yellow wheat soy sauce (YW). These sampling events had taken place in October 2016 and February 2017. Fig. 12A shows a stacked bar chart showing the microbial composition of black soy sauce fermentation during 5 months of fermentation. Samples were taken from vats back to 0, 0.5, 1, 2, 3, 4 and 5 months on October 2016 (Fall) in two replicates (A - vatl, B - vat2). Viral composition is included. Lactobacillus pobuzihii maximum abundance was 33% at month 2. Fig. 12B shows a stacked bar chart showing the microbial composition of black soy sauce fermentation during 5 months of fermentation. In total, 19 samples were analysed. Samples were taken from vats with 0, 0.5, 1, 2, 3, 4 and 5 months starting in February 2017 (winter) in three replicates (A - vatl, B - vat2). Viral composition is included. Halococcus spp was only detected in the 5^th month vat 2 (BB-5-2m). Lactobacillus pobuzihii population was its highest at month 3 with 66% abundance. Fig. 12C shows a stacked bar chart depicting the metagenomic profiles of industrial process of wheat soy sauce. Koji samples were taken from koji chamber at the beginning of koji maturation (koji_6 hours) and in the end koji_32 hours) and moromi (YW) at 1st month maintained at 15 °C and 5 months of fermentation. No Lactobacillus pobuzihii was detected. Fig. 12D is a stacked bar chart depicting the abundance and identity of various bacterial strains in yellow (Y) bean soy sauce, sampled in February 2017. Lactobacillus pobuzihii abundance was 7.2%. The amount of Klebsiella pneumoniae was shown to be increased up to 6.3% in contrast to black bean.

[0027] Fig. 13 shows two line graphs depicting the summary of dynamics of major microbial species in black bean (BB; top) versus wheat (YW; bottom). These figures show the comparison between microbial population of black bean moromi (top figure) and yellow and wheat soy sauce moromi (bottom picture).

[0028] Fig. 14 shows growth curves of microbial species isolated from black bean and yellow and wheat moromi at 0, 10, 14 and 18% NaCl in TSB broth. Fig. 14A shows a line graph depicting the growth of Lactobacillus pobuzihii WZ3 under various concentration of NaCl in TSB broth. Fig. 14B shows a line graph depicting the growth of Tetragenococcus halophilus #6 under various concentration of NaCl in TSB broth. Fig. 14C shows a line graph depicting the growth of Bacillus amyloliquefaciens #8 under various concentration of NaCl in TSB broth. Fig. 14D shows a line graph depicting the growth of Weissella paramenstroides under various concentration of NaCl in TSB broth. Fig. 14E shows a line graph depicting the growth of Staphylococcus sciuri under various concentration of NaCl in TSB broth. Fig. 14F shows a line graph depicting the growth of Enterococcus f aecium under various concentration of NaCl in TSB broth. [0029] Fig. 15 shows a line graph depicting the concentrations of Weissella (top) and Enterococcus (bottom), both of which can maintain growth in up to 14% NaCl. At 18% NaCl, their growth is inhibited. Measurements taken after 61 days were affected, possibly due to cross contamination (possibly due to salt-resistant Staphylococcus) as cups started to crack due to CO2 pressure. The negative control was also contaminated after 2 months of sampling.

[0030] Fig. 16 shows primary metabolite concentrations and pH changes over 25 days growth of Lactobacillus pobuzihii WZ3, Tetragenococcus halophilus #6, Bacillus amyloliquefaciens and Staphylococcus scuiri under 18% NaCl in TSB medium. Fig. 16A shows a line graph depicting Lactobacillus pobuzihii WZ3 metabolites under 18% NaCl in TSB. Fig. 16B shows a line graph depicting Tetragenococcus halophilus #6 metabolites 18% NaCl in TSB. Fig. 16C shows a line graph depicting Bacillus amyloliquefaciens metabolites 18% NaCl in TSB. Fig. 16D shows a line graph depicting Staphylococcus scuiri metabolites 18% NaCl in TSB.

[0031] Fig. 17 shows the dynamics of microbial relative abundances in pilot trials. Fig. 17A is a horizonal stacked bar chart showing the results of the pilot trial al. Tank 1 was with crushed beans + oat bran. Tank 2 was with crushed beans without bran. Fig. 17B is a horizonal stacked bar chart showing the microbial composition of Trial 2 and Trial 3. Weissella (green coloured) dominates up to 70% in moromi from 1 to 7 months of fermentation at 17.5% sat. Tetragenococcus halophilus developed without need of addition of it as starter culture. Pediococcus pentosaceus developed and took up 17% of total abundance. C5 - Control, no starter cultures; F10, F12 - Lactobacillus pobuzihii WZ3 added without koji washing; C2 - Lactobacillus pobuzihii WZ3 added; C4 - Weissella added; B1-B5 - Bacillus amyloliquefaciens added in koji with Tetragenococcus halophilus and Lactobacillus pobuzihii WZ3, Al - Bacillus amyloliquefaciens added in moromi with Tetragenococcus halophilus and Lactobacillus pobuzihii WZ3. Hl - 45 kg of 1 -month moromi was used as starter culture, H2 - 45 kg of 2-month moromi as starter, H3 - 45 kg of 3-month moromi as starter, H4 - 45 kg of 4 month moromi as starter, Gl-12, G2-12 - washed koji, E10 - Tetragenococcus halophilus no washed koji, E12 - Bacillus amyloliquefaciens no washed koji, F10 - Lactobacillus pobuzihii no washed koji, F12 - Lactobacillus pobuzihii no washed koji. Fig. 17C is a horizonal stacked bar chart showing the results of trial 4a performed at Xiluo, Taiwan, according to experimental design of Table 8 from August 2018 - January 2019. Fig. 17D is a horizonal stacked bar chart showing the results of the Quindao trial B conducted in February 2018. Vat inoculated with: 2 - Lactobacillus pobuzihii WZ3, 4 - Bacillus amyloliquefaciens # 8, 6 - Tetragenococcus halophilus # 6, 11126 - 6 month moromi inoculum, koji composition. Koji in Qingdao was also contained Weissella spp. as back in the Wuang Zhong (WZ) factory. Vats with starter culture are failed from the beginning. 6 months moromi without culture has less than 0.1% abundance of Lactobacillus pobuzihii. Wild species Weissella spp. and Eeuconostoc spp. were most abundant genus in Qingdao trial experiments. Fig. 17E is a horizonal stacked bar chart showing the results of Trial C. 16r RNA amplicon bacterial identification. Trial C in Tai Zhou. Tanks with samples A80_l, A80_2, A80_3, A80_4: A86_0, A86_l, A86_2, A86_3, A86_4 added 1% Lactobacillus pobuzihii', A83_0, A83_l, A83_2, A83_3, A83_4, control. The highest content of Lactobacillus was 0.5% in 3-4 months of fermentation. 16s rRNA amplicon sequencing picked up only a few hundred Lactobacillus pobuz.ihii at 1 month, accounting for less than 1% of the total bacteria.

[0032] Fig. 18 shows line graphs depicting the total nitrogen (TN) content and pH values taken during trial a3 over 5 months. Only vats with Bacillus inoculated into moromi together with Tetragenococcus and Lactobacillus. TN content was the highest in experiment with lowest drop in pH for AL

[0033] Fig. 19 shows a stacked bar chart showing the relative abundance of microbes in traditional BB moromi samples used for metatranscriptomic study. An abundance of 60% of Lactobacillus pobuz.ihii were observed at the first month versus 18% at fourth month of moromi fermentation.

[0034] Fig. 20 is a heatmap showing the top 20 most abundant peptides of soy sauce maturation in black bean (3^rd_0 - 3^rd_5), yellow bean only, trial al (tank 1 and Tank 2) and wheat (YW) starting from 6-hour koji until 5-month moromi.

[0035] Fig. 21 shows column graphs representing the profile of taste peptides. Left to right: black bean (BB) moromi (0-5) vs yellow bean (YB) (/ and //) and wheat (YW) (1-5B) over time of fermentation. Colours in (grey) green - umami, (black background) blue - kokumi, pink (white) - sweet, (slash line) yellow - sour, (white dots on grey background) grey - salty, (vertical lines) red - bitternesssuppressing taste.

[0036] Fig. 22 shows the protease activity of Aspergillus oryzae in moromi and koji samples. Fig. 22A is a column graph depicting the protease activity of WZ and DX starters in YPD medium at 117 hours. Laboratory experiment compared to Pediococcus acidilactici isolated from black soy sauce moromi. Protease activity on starter cultures, salt tolerant bacteria by NaCl. Casein as substrate according to the Folin method as described herein. Fig. 22B shows a scatter plot depicting protease activity measured by FITC-casein express method in the field (Wuang Zhong factory, Silo, Taiwan). Assays were applied at the point of collection taking koji protease activity from bamboo trays.

[0037] Fig. 23 shows images of the appearance of bacterial isolates from black soy sauce moromi on solid agar plates.

[0038] Fig. 24 shows a schematic representation of traditional black bean moromi fermentation.

[0039] Fig. 25 is a line graph showing the concentration of valine -proline -proline (VPP) peptide in 3 types of moromi: black bean 3^rd (0-5 months), yellow bean only moromi, trials with black beans (Tank 1 and Tank 2) and wheat moromi (YW).

[0040] Fig. 26 is a heatmap showing the correlation of volatile organic compounds (VOC) obtained by microbial fermentation in moromi from black bean (BB), yellow bean (Y), and wheat.

[0041] Fig. 27 is a heatmap showing the distinctive volatile organic compounds found in black bean (BB) moromi moths of fermentation BB, yellow beans moromi (Y) and wheat moromi (YW).

[0042] Fig. 28 is a schematic plan of a koji maturation room for inoculation of Bacillus amyloliquefaciens starter culture by sprinkling. [0043] Fig. 29 shows a horizontal column graph depicting the protease activity in an ex vivo laboratory trial 500 mL experiment. Timepoints taken are 1 day and 10 days.

[0044] Fig. 30 shows a horizontal column graph depicting the amylase activity in an ex vivo laboratory trial 500 mL experiment. Timepoints taken are 1 day and 10 days.

[0045] Fig. 31 shows a horizontal column graph depicting the lipase activity in an ex vivo laboratory trial 500 mL experiment. Timepoints taken are 1 day and 10 days.

[0046] Fig. 32 shows a line graph depicting the microbial abundance in Pilot 1 trial al.

[0047] Fig. 33 shows a LefSe analysis of pilot 1 large scale trial. 1A - traditional black bean moromi at month 1. In contrast, tank with pilot 1 showed the list of wild strains playing the main role in moromi. Most of species belong to Weissella sp strains.

[0048] Fig. 34 shows the results of adaptation of Lactobacillus pobuzihii strains at various concentration NaCl in De Man, Rogosa and Sharpe broth (MRS) and with supplementation of soy lecithin and tween-80. Fig. 34A shows line graphs depicting growth curves of WZ3 and WZ5 under 10-18% NaCl (v/w) in De Man, Rogosa and Sharpe broth and with additional of choline. Fig. 34B shows line graphs depicting ranges of soy lecithin concentrations under 0, 16, 17% of NaCl (w/v) at De Man, Rogosa and Sharpe broth. Fig. 34C shows line graphs depicting the cell concentration after addition of tween 80 to lecithin in 16 % and 17 % NaCl in De Man, Rogosa and Sharpe broth.

[0049] Fig. 35 shows the lag phase reduction of Lactobacillus pobuzihii live cells supplied with soy lecithin at 16 -17% of NaCl (concentration of salt used at factory manufactures) and down/upregulated genes at 17% of NaCl vs MRS and with/without soy lecithin. Fig. 35A shows line graphs depicting ranges of soy lecithin concentrations under 0, 16, 17% NaCl (w/v) at De Man, Rogosa and Sharpe broth. Fig. 35B shows column graphs depicting cell counts taken of Lactobacillus pobuzihii WZ3 during growth on De Man, Rogosa and Sharpe broth with soy lecithin (SL). Fig. 35C shows a volcano plot of Lactobacillus pobuzihii WZ3 gene expression at 2 mM lecithin with salt vs Salt without lecithin added. Four genes were significantly upregulated (red) with addition of lecithin: manX_2, manX_3, manZ_4 and sorA. First three genes are associated with mannose transport and sorA is responsible for sorbose utilization. Fig. 35D shows a volcano plot of up-regulated genes under salt stress - argF Ornithine carbamoyltransferase responsible for production of citrulline. Fig. 35E is a schematic showing the pathway of terminal reactions of degradation of amino acid L-citrulline upregulated in Lactobacillus pobuzihii WZ3 under salt conditions.

[0050] Fig. 36 shows a column graph depicting the relative abundance of 3- to 6-mer peptides. The relative abundance of 3- to 6-mer peptides shown in Tank 1 and Tank 2 were obtained after inappropriate fermentation (after 3.5 months of brewing). It is shown that Tanks 1 and 2 have same amount of taste peptides as present in black bean moromi at the beginning of moromi fermentation (at one month). This data indicates that the presence of L. pobuzihii in the black moromi is necessary to enrich amount of taste peptides during the time of brewing. [0051] Fig. 37 shows a schematic of pilot trial with a Bacillus amyloliquefaciens strain having been introduced at beginning of koji stage. Koji room was split into 3 zones which were planned for conducting of experiment for assessment of impact of introduced B. amyloliquefaciens on the inhibition of Weissella spp. population. For the convivence of the experimental trial, the left zone contained only A. oryzae spores without any additional bacterial strain (negative control); the middle zone served as a buffer zone containing trace amounts of B. amyloliquefaciens', and the right zone of koji was sprayed with B. amyloliquefaciens culture.

TABLES [0052] Table 1. Flavour characteristics of compounds which compose district flavour profiles of black bean moromi.

[0053] Table 2. Most abundant microbes in WZ traditional black bean moromi compared to industrial yellow bean and wheat in final months of moromi over 3 independent batches repeats.

[0054] Table 3. Relative abundance in % of Lactobacillus pobuzihii in soy black bean moromi.

[0055] Table 4. Summary of most prevalent microbes in high-grade black bean soy sauce moromi vs generic wheat soy sauce moromi. Both soy sauces are produced in Wuang Zhong (WZ) factory, Silo, Taiwan.

[0056] Table 5. Fermentation products produced with pure cultures on De Man, Rogosa and Sharpe (MRS) broth with 14% NaCl.

[0057] Table 6. Trial a3 with rational to test washed vs non-washed koji and utilization of 1-4 moromi as starter culture mix for new batches of soy sauce.

[0058] Table 7. Taste panel results for Trial a2 samples of 4th month moromi.

[0059] Table 8. Scheme started cultures addition for trial a4 with aim to enhance taste of industrial black bean (BB), yellow bean (YB), and wheat (YW) soy sauces.

[0060] Table 9. Metatranscriptome output of moromi fermentation at 1 and 4 months. The most abundant species including Lactobacillus pobuzihii and Tetragenococcus halophilus are shown at 1 month (1A, IB, 1C) and 4 month (4A, 4B, 4C) of moromi fermentation.

[0061] Table 10. Microorganisms which had a high abundance in Wuang Zhong (WZ) soy sauce moromi with some have been already isolated on the agar plate. Metagenome results provide guidance for key microorganisms responsible for moromi quality. Isolates were identified by 16s rRNA i

[0062] Table 11. Sources of current isolates/cultures origin

[0063] Table 12. Experimental design for laboratory trial with 600 mL moromi with various combinations of starter wild cultures. B - Bacillus amyloliquefaciens #8, W - Weisella paramensteroides #1, Ec - Enterobacter cloacea, Ef - Enterococcus faecalis, Sc - Staphylococcus sciuri, Sg - Streptococcus gangivalis, M - Micrococcus luteus. Tetragenococcus halophilus #6 - T. halophilus #6.

[0064] Table 13. B1-B5: Bacillus amyloliquefaciens added with aim to test Bacillus amylolysin for Weissella inhibition; C1-C2 only addition of Tetragenococcus halophilus and Lactobacillus pobuzihii WZ3, C3-C4: Weissella starter, C5: no starter cultures added, A1-A2: Bacillus amyloliquefaciens was added into moromi together with Tetragenococcus halophilus and

Lactobacillus pobuzihii WZ3.

[0065] Table 14. pH at laboratory scale experiment in moromi on 31August 2017

[0066] Table 15. Diffusion disk assay results of bacteriocins screening from wild soy moromi strains Bacillus amyloliquefaciens #8 (B), Weissella paramensteroides #1 (W), and nisin producing type strains Lactobacillus lactis B-978 and Lactobacillus lactis B-1948. No culture - negative control without addition any bacteriocins, x - indicates no growth, number indicates diameter of halo spread around disk.

[0067] Table 16. - Table of correlations (Pearson and R2) between microbial abundances (Metaphlan2 relative numbers) of 3 types of soy sauce moromi (BB, Y and YW) and concentrations of main volatile organic compounds (VOC) measured in the same samples. Bacterial species and related VOC are ranked by the strongest correlations from top to the bottom. Microbial values are given at left hand side and VOC values are given at the right-hand side of the table.

[0068] Table 17. -List of class Ila bacteriocines found in the public NCBI database based on original weissellin A bacteriocin amino acid sequence from Weissella paramensteroides

[0069] Table 18. - Bacteriocins identified from whole genome sequences of various microbial species deposited in NCBI, Pfam and InterProScan protein databases, which were found to be similar to Bacillus amyloliquefaciens bacteriocin obtained from Wuang Zhong (WZ) samples.

[0070] Table 19. - Diffusion disk assay of potential bacteriocins produced by soy sauce isolates Weissella paramensteroides and Bacillus amiloliquefaciens and their impact (halo spread size, in mm) on other members of soy sauce microbial community. Legend: 6 - Bacillus amyloliquefaciens #S; 16 - Weissella paramensteroides #1; N - bacteriocin nisin; 43 - Lactococcus lactis B-978 (nisin producer); 44 - Lactococcus lactis B-1948 (nisin producer) ; ? - unclear; x - no effect.

[0071] Table 20. - List of volatile organic compounds (VOC) their flavour and taste profiles, as well as their detection threshold. Also shown is the concentration found in black bean (BB), yellow bean (Y), and yellow wheat (YW) moromi at 5, or 5.5, months fermentation.

[0072] Table 21. - Comparative analysis of proteins from whole genome sequencing (WGS) of L. pobuzihii WZ3 and WZ5 indicating the number of proteases, peptidase and amylases related to cleavage of protein resulting in release of taste peptides and sugars from the starch.

Table 22. -List of upregulated genes under salt stress for L. pobuzihii WZ3. Output was generated as results of RNA-seq on De Man, Rogosa and Sharpe broth (MRS) versus MRS + 17 % NaCl (SALT). Id - gene abbreviation; log2FC - fold change in log2 value; p-adj - FDR > 0.01; description - functions of upregulated genes.

DEFINITIONS

[0073] As used herein, the term “umami” is defined as the taste of the amino acid L-glutamate and 5’- ribonucleotides, such as glutamates (salts of glutamic acid), the amino acid L-glutamate, guanosine monophosphate (GMP) and inosine monophosphate (IMP). It can be described as a pleasant "brothy" or "meaty" taste with a long-lasting, mouth-watering and coating sensation over the tongue.

[0074] As used herein, the term “kokumi”, which translates as “heartiness”, “full flavour” or “rich”, describes compounds in food that do not have their own taste but enhance the characteristics when combined. Alongside the five basic tastes of sweet, sour, salty, bitter and savoury, kokumi has been described as something that may enhance the other five tastes by magnifying and lengthening the other tastes, or “mouthfulness”. Garlic is a common ingredient to add flavour used to help define the characteristic kokumi flavours. Calcium-sensing receptors (CaSR) are receptors for "kokumi" substances. Kokumi substances, applied around taste pores, induce an increase in the intracellular Ca concentration in a subset of cells. This subset of CaSR-expressing taste cells is independent from the influenced basic taste receptor cells. CaSR agonists directly activate the CaSR on the surface of taste cells and integrated in the brain via the central nervous system. However, a basal level of calcium, corresponding to the physiological concentration, is necessary for activation of the CaSR to develop the kokumi sensation.

[0075] As used herein, the term “3- to 6-mer peptides” refers to oligomers consisting of 3, 4, 5, or 6 amino acids in length.

[0076] As used herein, the term “high salt fermentation condition” refers to fermentation conditions using a salt content of up to 17%. In other words, the Lactobacillus strains disclosed herein are able to tolerate conditions including up to 17% salt content.

[0077] As used herein, the term “culture” refers to a cell culture or a bacterial culture. Such cell or bacterial culture also include amounts of bacteria used to inoculate clean substrate, such as, but not limited to, starter cultures for inoculating fermentation.

[0078] As used herein, the term “microaerophilic” refers to an environment containing lower levels of dioxygen than that are present in the atmosphere (such as, less than 21% O2; this can range between 2 to 10% O2) for bacterial growth. DETAILED DESCRIPTION OF THE PRESENT INVENTION

[0079] While main progress in assessing quality of soy sauce have been seen, enhancement in flavour have been mainly achieved by genetic engineering of fungi and yeasts to shorten the time of fermentation. Despite huge variations in soy sauce manufacturing, there are only few starter cultures used for solid and liquid fermentation of soybeans and wheat used in ratio defined by Japanese or Chinese style. Solid fermentation or koji is controlled mainly by using spores of specific Aspergillus oryzae strains or substrains, the methods of which vary little between various manufactures.

[0080] Japanese-style liquid fermentation requires addition of a specific starter culture of Zygosaccharomyces rouxii, which in the presence of 55% wheat will produce a smell and alcohol content that is characteristic of this type of moromi. Sometimes, the starter culture Tetragenococcus halophilus can be added as well. Wild species Candida etchellsii and Candida versatilis have been reported for their specific flavour formation with 4-ethylphenol, along with main bacterial species such as Weissella cibaria, Bacillus spp., Lactobacillus fermentum, Streptococcus gallinarum, and Staphylococcus saprophyticus.

[0081] Korean-style soy sauce fermentation can be considered as a modification of Japanese-style soy sauce fermentation, which includes addition of Zygosacchramyces spp. Co-abundant wild yeasts, such as, but not limited to, Candida temnochilae, Pichia guilliermondii, Pichia sorbitophila, Pichia triangularis, Absidia corymbifera, Rhodotorula mucilaginosa, and bacteria, such as, but not limited to, Enterococcus durans, Bacillus subtilis, and Enterococcus faecium had been previously reported as part of a moromi flavour formation.

[0082] Chinese-style soy sauce moromi is characterised by presence of wild species of Candida spp., Kluyveromyces marxianis, Pichia fabianii, Weissella cibaria, Weissella confusa, Bacillus spp., Corynebacterium qlutamicum, Staphylococcus gallinarum, Staphylococcus saccharolyticus, and Enterococcus faecalis, without addition of specific Zygosacchramyces rouxii.

[0083] Taiwanese soy sauce, characterised by a lack of wheat and yeasts in liquid fermentation, can be considered a modification of Chinese-style soy sauce. Few wild bacterial species are involved at the moromi stage, such as, but not limited to, Staphylococcus sciuri, Klebsiella pneumoniae, Enterobacter cloacae, Salmonella enterica, Enterococcus faecium, Weissella confusa, Bacillus amyloliquefaciens , and Staphylococcus gallinarum.

[0084] While yeast Zygosaccharomyces rouxii and some lactic acid bacteria, such as, but not limited to, Tetragenococcus halophilus, are widely used for mass-production of wheat-based soy sauce, other fermenting species are wild origin which bring very generic taste characteristics to the final product.

[0085] Disclosed herein is the use of Lactobacillus pobuzihii (also referred to as L. pobuzihii) as a starter culture for the production of Taiwanese-style, gluten-free black bean soy sauce with shows an enriched umami and kokumi taste and has an increased content of valine-proline-proline (VPP) peptide. The nutritional value and health benefits of such soy sauce, which can diminish risks of cardiovascular disease and prevent metabolic conditions like diabetes, are highly desirable. [0086] Lactobacillus pobuzihii can be used in gluten-free soy sauce to obtain a “meaty taste”. In wheat-based soy sauces, main driver of taste formation and smell is a starter culture of Zygosaccharomyces rouxii yeast. In situations where gluten free soy sauce is to be produced, the fermentation process is driven by lactic acid bacteria resistant to salt. Lactobacillus pobuzihii is one of the bacteria that can be added to enrich taste in soy moromi (without wheat). Common wild bacteria used to drive fermentation in gluten free sauces are Tetragenococcus halophilus, Weissella spp. and Staphylococcus spp. These bacteria are widely spread at factory environment and do not need to be added resulting in very generic taste of factory produced moromi. In the present disclosure, Bacillus amyloliquefacies is a desired bacterium from koji fermentation stage.

[0087] Two salt resistant strains Lactobacillus pobuzihii WZ3 (accession no. DSM 33648) and WZ5 (accession no. DSM 33658) were isolated from moromi, characterized, and used in studies for the processing of traditional Taiwanese black bean soy sauce.

[0088] Thus, in one example, there is disclosed a culture or starter culture comprising Lactobacillus pobuzihii WZ3 (accession number DSM 33648), or Lactobacillus pobuzihii WZ5 (DSM 33658), or a combination thereof. In another example, the culture or starter culture disclosed herein further comprises wild-type Lactobacillus pobuzihii. In another example, the Lactobacillus pobuzihii strain is selected from the group consisting of wild-type Lactobacillus pobuzihii; Lactobacillus pobuzihii WZ3 (DSM 33648); Lactobacillus pobuzihii WZ5 (DSM 33658); a combination of Lactobacillus pobuzihii WZ3 (DSM 33648) and Lactobacillus pobuzihii WZ5 (DSM 33658); a combination of wild-type Lactobacillus pobuzihii and Lactobacillus pobuzihii WZ3 (DSM 33648); a combination of wild-type Lactobacillus pobuzihii, Lactobacillus pobuzihii WZ5 (DSM 33658); and combination of a wild-type Lactobacillus pobuzihii, Lactobacillus pobuzihii WZ3 (DSM 33648), and Lactobacillus pobuzihii WZ5 (DSM 33658).

[0089] In yet another example, there is disclosed a process of fermentation comprising a starter culture comprising at least one Lactobacillus pobuzihii strain.

[0090] In one example, the culture disclosed herein is for use in fermentation. In one example, the fermentation is soybean fermentation, or moromi fermentation. Exemplary steps of such a fermentation process include, but are not limited to, the steps of a koji maturation, one or more optional washing steps, moromi fermentation and maturation, moromi pressing, and pasteurisation. In one example, the process disclosed herein comprises all of the steps described above.

[0091] It is of note that, for example, soymilk fermentation involves aqueous extraction of soybeans, which is completely different processing compared to soy sauce fermentation. In soy sauce fermentation, mash of whole soybeans gets immersed into 18% salt bran solution. Therefore, bacteria which are responsible for formation of taste peptides have to be extremely salt tolerant. Usually, common Lactobacillus species found in milk cannot survive and grow under salt concentration higher than 5%.

[0092] In one example, the starter culture disclosed herein is a moromi fermentation starter culture.

[0093] In another example, the method disclosed herein applies to the strains disclosed herein which are used as starter cultures for fermentation. Each batch of moromi is to be inoculated with IxlO⁷ of salt tolerant Lactobacillus pobuzihii cells. These Lactobacillus pobuzihii cells had been prepared in a fermenter with supplementation of 2 mM soy lecithin in MRS (De Man, Rogosa and Sharpe broth) medium or equivalent as osmoprotector, prior to application in fermentation. Addition of soy lecithin is essential for faster adaptation of Lactobacillus pobuzihii to high salt stress.

[0094] Application of whole genome sequencing (WGS) and RNA sequencing of Lactobacillus pobuzihii revealed enzymatic degradation of L-citrulline with the release of ammonia, and producing mannitol as an osmoprotectant under high salt concentrations. Shotgun metagenomics analysis performed after several months of fermentation revealed the presence of bacteriophages (phage) in the composition of moromi, which have not been previously reported.

[0095] Fermentation of traditional Taiwanese, Wuan Chuang-style black bean soy sauce was subjected by intensive multi-omics analysis. Metagenomics analysis of several months of fermentation revealed a composition of bacteria and bacteriophages of moromi which have not been reported before. Based on genomic results, Lactobacillus pobuzihii was targeted for isolation as a lactic acid bacteria shown to play a role in improving the taste and flavour formation of black bean moromi. The strains were isolated under anaerobic conditions and characterized for functioning pH, growth rate, and key metabolite production under anaerobic conditions. Several trials for production of black bean soy sauce were conducted with these Lactobacillus pobuzihii strains as starter culture.

[0096] Without being bound by theory, it is thought that taste is characterized by relative abundance of taste peptides. A heatmap of taste peptides (as shown herein, for example) reflects as quality (name of short peptides), as well as the relative quantity (relative numbers) of such peptides. It has been shown herein that the amount of taste peptides accumulates within maturation of moromi, as can be seen, for example, in case of black bean Wuang Zhong (WZ) soy sauce, while Lactobacillus pobuzihii abundance increase over the time. In contrast, amount of taste peptides, especially those responsible for umami and kokumi taste, did not change in yellow and wheat during 5 months of moromi maturation (whereby no Lactobacillus pobuzihii was present). Taste peptides of yellow and wheat soy sauce formed mainly during the koji stage due to the contribution of Aspergillus oryzae present in the standard culture. Once the koji is immersed in salt brine, the activity of fungal protease (thought to be responsible for cleavage protein) is inhibited. Thus, only osmotolerant bacterial cultures will contribute to the number of unique taste peptides present in the resulting product.

[0097] Without being bound by theory, it is thought that flavour is the result of the presence of volatile organic compounds (VOC). In the field of soy sauce manufacturing, the term “volatile organic compounds” usually refers to classes of volatile organic compounds, such as, but not limited to, pyrazines, phenols, acetate, ethanol, and the like, present in soy sauce. About 200 volatile organic compounds have been detected during GC/MS analysis (gas chromatography coupled mass spectrometry analysis). The concentration of these compounds was measured in moromi samples of three different types of soy sauces. The resulting heatmap shows the concentrations of all volatile organic compounds. The concentration of each measured compound is divergent by nature (e.g., pyrazines can be smelled by a human at a concentration of less than 0.001 mg/mL, while ethanol can only be detected/smelled by humans at a concentration of 1 mg/mL), thus the resulting values had been normalised in order to be able to build the heatmap. Due to this normalisation, all volatile organic compounds concentrations can be visualized at the same figure. Examples of absolute concentrations for each sample are provided in Table 20. The purpose of the heatmap disclosed herein is to provide a representation of district features of black bean moromi compared to standard yellow and wheat moromi. As known in the art, overall flavour characteristics can be subjective (e.g., flavour preferences for soy sauce are formed and influenced based on geographical locations within China). Thus, focus was concentrated on the improvement of taste and flavour characteristics of the soy sauce disclosed herein.

[0098] Also disclosed herein is a method for providing taste and flavour characteristic of a traditional Taiwanese black bean soy sauce by using an inoculating mix of Lactobacillus pobuzihii during moromi fermentation. The addition of Lactobacillus pobuzihii has been shown to reduce the number of undesirable wild species of Weissella spp. and Staphylococci spp., thus lowering ethanol and acetaldehyde content, and resulting in the production of valine -proline -proline (VPP) peptide. The lowering of alcohol and acetaldehyde has been shown to improve kokumi and umami taste profiles. Valine -proline-proline (VPP) peptide has been shown to possess angiotensin inhibiting properties.

[0099] Thus, in one example, there is disclosed a method of lowering undesirable taste components in fermented black bean soy sauce, the method comprising use of the culture as disclosed herein.

[00100] Examples of undesirable taste components are, but are not limited to, ethanol, aldehydes, octanoic acid, ethyl acetate, methyl-butanol, benzene acetaldehyde, 4-ethyl-2-methoxy phenol, benzaldehyde, acetaldehyde, octanoic acid ethyl ester, 2-furanmethanol, propanoic acid, 2-hydroxyethyl ester, 3-methyl-l -butanol, and combinations thereof. In another example, the undesirable taste components as disclosed herein can be considered to undesirable taste components only if a respective odour threshold is passed. Thus, in another example, undesirable taste components are, but are not limited to, benzeneacetaldehyde, 4-ethyl-2-methoxy phenol, benzaldehyde, acetaldehyde, octanoic acid ethyl ester, 2-furanmethanol, 2-hydroxyethyl ester, 3-methyl-l -butanol, ethanol, octanoic acid, ethyl acetate, and methyl-butanol. For many taste components (for example, those that impart bitter or sour tastes), these may only be considered undesirable if a certain odour threshold is exceeded. The following is a non-exhaustive and exemplary list of taste components and their odour thresholds in brackets: benzeneacetaldehyde - 0.05 mg/ml (0.004 mg/ml); 4-ethyl-2-methoxy phenol,- 0.07 mg/ml (0.05 mg/ml); benzaldehyde - 0.27 mg/ml (0.35 mg/ml); acetaldehyde - 0.03 mg/ml (0.015 mg/ml); ethanol - 9.5 mg/ml (10 mg/ml); octanoic acid ethylester - 2.9 mg/ml (0.015 mg/ml); 2-furanmethanol - 0.43 mg/ml (0.008 mg/ml);, 2-hydroxyethyl ester - 0.5 mg/ml (0.2 mg/ml); 3-methyl-l -butanol - 0.57 mg/ml (0.02 mg/ml).

[00101] In another example, there is disclosed a method of increasing desirable taste components in fermented black bean soy sauce, the method comprising use of the culture disclosed herein. In one example, the increase in desirable taste components is characterised by the increase in taste conferring peptides and/or taste conferring amino acids in relation to total protein content of a sample.

[00102] Examples of taste conferring peptides and/or taste conferring amino acids are, but are not limited to, VPP, EV, EE, DES, ED, K, EEDGK, KGSLADEE, D, DE, DD, DEE, VE, VD, KGDEE, ADE, EGS, ES, EDD, EED, DDE, DED, DDD, EEE, EDE, E, SLAKGDEE, SLADEEKG, KGDEESLA, DA, VG, VV, LE, EL, EG, EY, V, P, SPE, EEN, EPAD, VGV, FFRPFFRPFF, GP, RRPFF, VYPFGGGINH, PR, LK, WP, FFPG, RGPPF, GGP, RKE, RPGGFF, YGY, RGPPGGF, RGPPGIG, PK, RGPPFIVRGPPFIV EM, GGFFGG, VF, DLL, YGG, PGI, APPPPAEVHEV, RGPPFIVGG, FPK, AED, GFF, RPFFRPFF, VL, RGGFIV, FP, AE, PPF, GGLGG, VVGET, EVG, FIV, VW, GR, AAA, RPFF, GGV, KPK, SAEQK, WW, LLLL, RGPFPIV, YPF, FPF, GLGG, GGLG, RG, PGP, AA, GLGGG, KP, FGGF, VVAPFPEV, WWW, LGG, FGFG, LVYPFPGPIHN, LT, WL, GGA, RGPPGFF, LF, PGPGPG, RGPFPIIV, EH, PP, YPFPGPI, FPPFIV, GGY, VYP, GGFF, EQ, LLL, GV, VY, FFF, R, YY, GPFP, GLL, GAA, RRPP, FGG, PSE, IL, IV, QEEL, RRPPFF, GGRGPPFIVGG, DG, LG, KGD, RPFG, AD, LLG, RGFF, GE, EGG, VYPFPPGI, PFIV, PIP, FF, APPPPAEVHEVV, RPFFGG, RGPPFFF, RGPEPIIV, RR, LI, GPPFIV, AGA, YQEPVLGPVRGPFPIIV, PVRGPFPIIV, FFPP, GGGGL, APPPPAEVHEVVE, YG, VVAPFP, VGG, PGG, GPFPI, RRPPGF, LGGGG, VAPFP, FGF, GPVRGPFP, FV, VIFPPG, PL, WF, EINEL, EF, GY, FFPGG, PFPGPI, KF, GAG, INEL, FW, AF, VAPFPE, EEL, RGPPFIV, PFPIIV, VYPFPPGIGG, FFPR, RGPPFGG, RGPPGGGFF, APFPE, VYPFPPG, GGVVV, VIIPFPGR, W, RGP, FL, ELL, RP, L, YYY, RGPPGFG, PPP, RGPKPIIV, RGPPGGV, VVAPFPE, RRPPPFFF, PVLGPV, RPPFIV, VIF, PPG, GW, FFGG, RRR, RPG, RGPPFF, GLG, FFG, RPPPFFF, G, VPPFL, RGPPGGFF, VYPFPPGINH, PPFIV, GF, LGL, YYG, ECG, KPF, VIIPFPG, GGGLG, YF, RF, AAG, GPG, RGPPFIIV, LA, GGRGPPFIV, DV, DL, EQEEL, FG, F, APPPPAEVHEEVH, PF, FPP, LGGG, AGG, GL, GGGL, NWGET, GGF, PFPP, RPFFRPFFRPFF, PFPGPIP, KPPFIV, IF, SLA, A, LQ, VI, YP, GYG, GRP, RGPFPI, RPGF, VIFPPGR, GGRPFF, RPF, LEQL, YPFPGPIHNS, VIPFPGR, ECA, PFP, PGR, VYPFPPIGNH, AL, GGRPFFGG, GPPF, FI, IG, GI, GPFF, FFPE, II, RGPGPIIV, El, LL, RPGFF, PGPIP, IW, VRGPFP, VYPF, RGPFIV, LV, GYY, LD, LW, PVLGPVRGPFPIIV, GFG, ENINEL, GGL, RGPPFI, and RGPPGF.

[00103] In another example, the taste conferring peptides and/or taste conferring amino acids is/are, but are not limited to, EV, EE, DES, ED, EEDGK, KGSLADEE, D, DE, ECG, LT, EH, EQ, ECA, LA, LQ, VGV, AAA, AA, GGA, GAA, KGD, AGA, VD, KGDEE, ADE, E, SLAKGDEE, SLADEEKG, KGDEESLA, SPE, DD, DEE, EDD, VE, DA, VG, VV, LE, and RPGGFF.

Sequencing results summary

[00104] Based on annotation against a Lactobacillus database, 2217 and 2156 coding sequences (CDSs) were found between the Lactobacillus pobuzihii strain WZ3 (DSM33648) and the reference strain Lactobacillus pobuzihii KCTC 13174, respectively. Among those, 769 and 777 genes (not counting multiple copies of the same gene) were annotated respectively. [00105] As a result of bioinformatics analysis, 20 genes were found only to be present in Lactobacillus pobuzihii strain WZ3 (DSM33648). The gene names and their respective product names are: xtmB (phage terminase large subunit), gspA-1 (universal stress protein A), uxaC (glucuronate isomerase), uxuB (mannitol dehydrogenase), uxuA (mannonate dehydratase); uxaA (altronate hydrolase); iclR (IclR family transcriptional regulator); bglG (transcriptional antiterminator); yorL (putative DNA polymerase YorL); yxaB (general stress protein 30); ggaB (minor teichoic acid biosynthesis protein GgaB); arcR (ArcR family transcriptional regulator); malA (maltodextrose utilization protein malA); wcaJ (UDP-phosphate galactose phosphotransferase); vapl (toxin-antitoxin system antitoxin subunit); sacA (sucrose-6- phosphate hydrolase); relB (DNA-damage-inducible protein J); argF (ornithine/putrescine carbamoyltransferase); arcC (carbamate kinase); and cheY (chemotaxis protein CheY).

Shotgun metagenomics revealed prevalent abundance of Lactobacillus pobuzihii in black bean (BB) moromi

[00106] Fermentation of traditional Taiwanese Wuan Chuang-style black bean soy sauce was subjected by intensive multi omics analysis. In total, three independent sampling events were undertaken on 6 batches on traditional black bean (BB) soy fermentation (Figs 9A to 9D, Figs. 12A to 12B). In total, 10 independent koji trays and 49 moromi vats were analysed by shotgun metagenomics.

[00107] In addition, shotgun metagenomics was undertaken to analyse microbial communities obtained from yellow bean soy sauce fermentation (Fig. 12D) and wheat soy sauce fermentation (Fig. 12C), to aid in comparing bacterial compositions of key bacteria between different type of soy sauces produced at the same factory. As result, metagenomics analysis of several months of fermentation revealed unique composition of bacteria and bacteriophages of black been moromi which have not been reported before for soy sauce (Table 4).

[00108] Along with previously described Tetragenococcus halophilus, Weissella spp and Bacillus amyloliquefaciens, Lactobacillus pobuzihii were observed to be the most abundant bacteria (70%) at 3- month and 4-month fermentation of moromi (Fig. 9A). The second most abundant bacterium was the salt resistant Tetragenococcus halophilus. These results were confirmed by two independent sampling events undertaken at different batches in other weather seasons (Fig. 12B and Fig. 12C). Since sampling time points were taken from independent vats prepared manually at various batches, it was concluded that presence of Lactobacillus pobuzihii is consistent within the black bean moromi, and cannot be considered as accidental contamination. Therefore, it is shown that Lactobacillus pobuzihii is a contributor for the formation of specific flavour and taste formation of traditional black bean moromi.

[00109] Surprisingly, a high abundance of DNA bacteriophages (around 90% of all metagenomic reads), corresponding to main bacterial species presented during koji maturation, was observed. Among them, the most abundant were bacteriophages with hosts such as Staphylococcus, Weissella and Bacillus (Fig. 10A) The presence of viral reads was detected not only in liquid of moromi but was also seen in solid material after adjusting method of DNA extraction (from working on mash sample to solid pellet) and varied between 5-40% of total DNA (Fig. 10B). [00110] In contrast, the moromi of wheat soy sauce (YW) did not contain any traceable Lactobacillus pobuzihii present in tanks examined with same shotgun metagenomics method. The wheat soy sauce (YY) microbial population consisted of Weissella spp., Bacillus amyloliquefaciens, Staphylococcus spp., and Tetragenococcus halophilus species dominated in liquid fermentation (moromi). It was also observed that only the amount of B. amyloliquefaciens bacteriophages correlated with the population of bacterial B. amyloliquefaciens present in moromi (data not shown). However, bacteriophages of Staphylococcus and Weissella were not detected in the moromi of yellow bean (YB) compared to black bean (BB) moromi. Thus, without being bound by theory, it is thought that very low abundance of salt-resistant Staphylococcus spp. and Weissella spp., and the presence of Lactobacillus pobuzihii in black bean moromi may be due to high pressure of lytic bacteriophages proliferating Staphylococci and Weisseilla bacteria. This results in Lactobacillus pobuzihii having an advantage to compete for nutrients with other bacteria.

[00111] Attempts to isolate bacteriophages from wild Staphyloccus scuiri and Weisseilla paramenteroides strains isolated from salt bran collected at the bottom of unwashed vats (after 2 weeks of moromi fermentation) and factory wastewaters failed to show any plaques of the lytic bacteriophages. Without being bound by theory, this is thought to indicate that identified reads from shotgun metagenomics belong to temperate bacteriophages and/or to highly specific bacteriophages that did not infect the isolated bacterial hosts. It is thought that presence of these bacteriophages may be one of the further conditions that play a role in the successful, traditional fermentation of soybeans. Introducing of bacteriophages from traditional fermentation into moromi can be one of the strategies used to control the presence of undesirable strains of wild strains of Staphylococcus and Weissella. This can also apply when up-scaling fermentation. Thus, in one example, the method or fermentation process disclosed herein includes the introduction/addition of bacteriophages. In another example, the bacteriophages introduced into the method are bacteriophages that infect Bacillus amyloliquefaciens. In another example, the bacteriophages do not infect Staphyloccus scuiri and Weisseilla paramenteroides strains. In another example, the bacteriophages infect species that compete with Weissella spp. and Staphylococcus spp.

[00112] Bacteriophages can be introduced, for example, by inoculation. Thus, in one example, bacteriophages are present in the starter culture. In another example, bacteriophages are introduced at a concentration of at least 1 x 10⁶ PFU/L. In another example, the bacteriophages are introduced during the first month of moromi fermentation.

Functional genes analysis of microbial population during maturation of black bean soy sauce

[00113] Based on shotgun metagenomics reads, differential functional genes abundances present during koji and moromi fermentations were analysed (Fig. 11B). As expected, the differences were observed between koji maturation samples and moromi fermentation (Fig. 11 A). Koji maturation showed characteristic pathways of phytol degradation as a main component, along with and a phosphate-storing complex of soybean bran and a main inhibitor of animal enzymes. The first upregulated pathway clusters during koji maturation included branch chained amino acid formation with L-valine and L-isoleucine biosynthesis, sulphur containing essential amino acids such as L-methionine and L-homoserine biosynthesis, aromatic amino acids anabolism with L-tryptophan degradation pyruvate fermentation of isobutanol, and leukotriene biosynthesis during aerobic respiration related to cytochrome c. Sulphate reduction, peptidoglycan maturation, fatty acid biosynthesis, palmitoleate biosynthesis, fatty acid elongation oleate biosynthesis formed another cluster which was shown to be upregulated during koji maturation. Polysaccharide degradation was detected based on stachyose degradation.

[00114] Moromi fermentation pathways showed few specific upregulated clusters with nucleotide synthesis: adenosine, guanosine and pyrimidine biosynthesis, GDP-mannose biosynthesis. Presence of lactic acid bacteria (LAB) provided upregulation of oligosaccharides involving sucrose, lactose and galactose degradation. Together with hexose sugar, the pentose phosphate pathway was also upregulated, indicating that lignin fractions of soybean bran (for example, arabinoxylan, xylan) can be actively involved in fermentation. These pathways may result in specific flavours due to the hulls of whole black beans which are present in soy flakes, which are a main source in industrial-scale process. Active acetate production during anaerobic fermentation was mirrored by acetyl-CoA biosynthesis. Fatty acids synthesis was shifted towards specific cis-vaccinate biosynthesis, which can be related to desaturation of linoleic acid. Adenine, purine and pyrimidine degradation can be indication of autolysis of koji. Specific gondoate anaerobic biosynthesis can results in the formation of specific MUFA C20:l, n-9 fatty acid. Urate biosynthesis/inosine -5 -phosphate degradation, which was pathway contributing the utilization of carbamide, was upregulated in the batch.

[00115] Metatranscriptome analysis with Humann2 of 1 -month vs 4-month fermented moromi with prevalent abundances of Lactobacillus pobuzihii at the 1-month (62%; Fig. 19) revealed non-annotated reads mainly from this species. The few annotated metabolic pathways showed to be upregulated were pyruvate fermentation to acetate and lactate, and D-galactose and L-rhamnose degradation. Tetragenococcus halophilus was responsible for acetic and lactate production at the 4-month moromi (Table 9).

Isolation of key fermentation bacteria and strain characterisation

[00116] Based on genomic results, Lactobacillus pobuzihii was chosen as primary culture for the isolation, as this lactic acid bacterium (LAB) was found to play a role in unique taste and flavour formation of black bean moromi.

[00117] Despite increased presence of Lactobacillus pobuzihii in 4-month of fermentation, two cultures were isolated from vat 1A and IB of the 1-month, WZ3 and WZ4 together with Bacillus amyloliquefaciens. While Bacillus amyloliquefaciens abundance shown to be around 5%, Lactobacillus pobuzihii abundance shown to be less than 1%. Nevertheless, the presence of live Lactobacillus pobuzihii in the moromi indicated that wild strains had been present at the beginning of fermentation, and which can be used as strategy to inoculate this bacterium right after the beginning of moromi stage. Lactobacillus pobuzihii WZ5 (DSM 33658) was isolated from the 4A vat which showed the highest abundance of Lactobacillus pobuzihii. Two strains of Tetragenococcus halophilus were also isolated from month 4 old moromi. It was noted that Lactobacillus pobuzihii WZ5 (DSM 33658) showed less resistance to salt solution compared to Lactobacillus pobuzihii WZ3 (DSM 33648). However, the growth rate of Lactobacillus pobuzihii WZ5 was shown to be faster than Lactobacillus pobuzihii WZ3 on De Man, Rogosa and Sharpe broth medium without salt. Lactobacillus pobuzihii strain WZ3 was resistant up to 17% of salt while growing on De Man, Rogosa and Sharpe broth and was chosen as starter culture for laboratory and pilot trials at industrial scale.

[00118] Thus, in one example, in the process disclosed herein, the at least one Lactobacillus pobuzihii strain is adapted to tolerate high salt concentrations in an adaptation step. Prior inoculation into moromi, Lactobacillus pobuzihii cells were inoculated into growth media (1 m³) with 2 mM soy lecithin at 17% salt for 5 days until OD600 of 0.6 was reached. Thereafter, the adapted starter culture was added to a tank (capacity of about 100 tons).

[00119] In one example, the Lactobacillus pobuzihii strain is adapted to tolerate high salt concentrations prior to its use in fermentation. In other words, the Lactobacillus pobuzihii strain is made to tolerate high salt concentrations. In one example, this is done prior to inoculation and/or fermentation using the Lactobacillus pobuzihii strain disclosed herein. Such an adaptation step can include use of an osmoprotector, for example, soy lecithin.

[00120] Since Weissella and Staphylococcus strains were found to be mainly present in yellow bean (Y) and yellow wheat (YW) tanks, and were present at less than 5% in black bean (BB) moromi, these strains were also targeted for isolation. However, it was only possible to isolate these strains from large pilot trials of crushed beans with bran with addition of antibiotics together with vancomycin-resistant Bacillus subtilus. After isolation of Weisseilla paramenstoroides , Staphylococcus saprophyticus and Staphylococcus scuiri were also found to be resistant to up to 14% NaCl (Fig. 14D, Fig. 14F).

[00121] Bacillus amyloliqiefaciens was shown to produce higher amounts of lactate (0.124 g/L), while E. faecium released higher amounts of acetate (0.259 g/L) compared to other lactic acid bacteria (LAB) isolated from moromi (Fig 14F and Table 5). Both Lactobacillus pobuzihii WZ3 and Tetragenococcus halophilus #6 were found to reduce pH to 6.0 in De Man, Rogosa and Sharpe broth, but glutamic acid (0.027 g/L) was only detected in the presence of Lactobacillus pobuzihii WZ3. Thus, Lactobacillus pobuzihii WZ3 should not be considered as spoilage agent of soy sauce.

[00122] In one example, the process disclosed herein includes use of a starter culture further comprises bacteria selected from the group consisting of Tetragenococcus halophilus, Bacillus amyloliquifaciens, and combinations thereof.

[00123] Traditional fermentation is usually conducted under anaerobic conditions. However, in massscale production, the moromi is mixed with air (with the purpose of adjusting the pH and boosting growth of the Z. rouxii yeast) from time to time, which brings in oxygen and disturbs anaerobic conditions. As a result of that mixing, Weissella and Staphylococcus might be overrepresented in fermentation. Regarding the strains disclosed herein, successful fermentation of soy sauce with application of Lactobacillus pobuzihii is dependent on avoiding (excessive) incorporation of oxygen into the moromi. Weisella spp. is not a desirable bacterium in the claimed method. Typically, it is present in Chinese soy sauce of modest quality. The aim of the presently claimed method is to replace the wild Weissella population with the Lactobacillus pobuzihii strains disclosed herein, thereby improving the taste and flavour of the soy sauce.

[00124] One example of a specific condition for inoculation is: inoculation at 1% of inoculum (for example, with IxlO⁷ cells) into 17% salt solution with koji immersed in it. Temperature of water is 25- 28°C and initial pH range from 5.6-6.0. Mixing with air should be avoided, otherwise Weissella and Staphylococcus will overgrow.

[00125] In another example, the fermentation process as disclosed herein is performed under microaerophilic conditions, or under conditions which are sufficiently anaerobic to enable preferential growth of Lactobacillus pobuzihii over Weissella and Staphylococcus.

Genomic characterization of Lactobacillus pobuzihii revealed unique features compared to type strain KCTC 13174

[00126] Both genomes of Lactobacillus pobuzihii WZ3 and WZ5 strains were sequenced, assembled, and compared to the strain Lactobacillus pobuzihii E100301 which had been deposited as KCTC 13174. Lactobacillus pobuzihii El 00301 has been described as homofermentative lactobacilli, which were unable to utilize mannose, mannitol, xylose, sorbose, dulcitol, inositol, sorbitol, fucose, arabitol, raffinose and inulin.

[00127] Comparative genomic analysis of the Lactobacillus pobuzihii WZ3 strain revealed the following unique genes distinctive compared to KCTC 13174: xtmB phage terminase large subunit; gspA-1 universal stress protein A; uxaC glucuronate isomerase; uxuB mannitol dehydrogenase; uxuA mannonate dehydratase; uxaA galactate dehydratase/altronate hydrolase; iclR IclR family transcriptional regulator; bglG transcriptional antiterminator; yorL putative DNA polymerase YorL; yxaB general stress protein 30; ggaB minor teichoic acid biosynthesis protein GgaB; arcR ArcR family transcriptional regulator; malA maltodextrose utilization protein malA; wcaJ UDP-phosphate galactose phosphotransferase; vapl toxin-antitoxin system antitoxin subunit; sacA sucrose-6-phosphate hydrolase; relB DNA-damage-inducible protein J; argF ornithine/putrescine carbamoyltransferase; arcC carbamate kinase; cheY chemotaxis protein CheY. In other words, these genes were genes that were found in Lactobacillus pobuzihii WZ3 but not found in KCTC 13174.

[00128] One of feature of the Lactobacillus pobuzihii WZ3 strain is its ability to utilize pentose sugars by altronate hydrolase (catalyzing the conversion of D-altronate into 2-dehydro-3-deoxy-D-gluconate) and glucuronate isomerase, in contrast to common hexose sugar fermentation. Another feature distinguishing Lactobacillus pobuzihii WZ3 from KCTC 13174 is the presence of mannitol dehydrogenase and mannonate dehydratase, which results in the classification of Lactobacillus pobuzihii WZ3 to obligatory heterofermentative lactobacilli group type 3. Type 3 lactobacilli ferment hexoses to lactate, ethanol and/or acetate, and carbon dioxide. These species exclusively use the 6- phosphogluconate/phosphoketolase (6-PG/PK) pathway. Lactobacillus pobuzihii WZ3 also has the advantage of utilising fructose sugar not only as a substrate but also as an electron acceptor (as a component of sucrose) to produce mannitol as a polyol (with the help of NAD⁺ mannitol dehydrogenase), instead of producing ethanol which is toxic for the culture. As result, one ATP molecule can be generated and only one molecule of acetate is formed from acetyl-CoA, which can explain the reduced acetate formation in Lactobacillus pobuzihii WZ3 compared to the Lactobacillus pobuzihii KCTC13174 strain. This is metabolic function is specific to Lactobacillus, as only few species were described before with this metabolic function.

[00129] Regarding the question of its ability to cleave proteins, whole genome sequencing (WGS) analysis of Lactobacillus pobuzihii WZ3 identified a higher number of proteases for producing oligopeptides and free amino acids compared to the Lactobacillus pobuzihii WZ5 strain and Lactobacillus pobuzihii KCTC13174 (see Table 21). This can have an impact on the production of valine-proline- proline peptides, which are thought to have a positive effect on, for example, cardiovascular disease.

Comparative genomic characterisation of two Aspergillus oryzae strains revealed no difference in functions impacting quality affinal soy sauce product

[00130] Since koji maturation is a key stage for later black bean (BB) moromi, the genomes of a Japanese Aspergillus oryzae \NA (obtained from Japan) and a Chinese produced Aspergillus oryzae DX strain were analysed. Based on the purity of acid and alkaline protease activity, and the glucoamylase/amylase of different Chinese koji, it was found that the properties of Aspergillus oryzae DX koji produced in He Bei province is similar to Japanese koji used in Xiluo. The results of whole genome sequencing of Aspergillus oryzae DX koji were identical to the results of whole genome sequencing of the Aspergillus oryzae in koji used in Xiluo. The only difference was that the Aspergillus oryzae DX koji sample showed trace amounts of the Dasheen mosaic virus, which can be an indicator of local production in China.

[00131] Based on genomic comparison, it was concluded that Aspergillus oryzae DX can be used as complete substitute of Aspergillus oryzae \NA strain. In other words, the solid fermentation (koji) will not affect the quality of black bean moromi. Koji fermentation is controlled by Aspergillus oryzae. Since koji maturation is stable and fixed, it is thought that the improvement of taste is based on addition of the Lactobacillus pobuzihii WZ3 and/or WZ5 strains disclosed herein to the liquid fermentation stage or to moromi.

[00132] When protease activity of WZ koji on whole beans was measured from bamboo tray at the Wuang Zhong (WZ) factory site, only 1.2 days were required to reach 60% of the maximum protease activity. The protease activity of koji was its highest on the fourth day with 600 U/g (Fig. 22B). In the laboratory, the protease activity of same fungi in liquid was lower than 2.5 U/mL, and could be compared to the bacterial protease of Pediococcus acidilactici (Fig. 22A) isolated from moromi. Thus, Pediococcus acidilactici protease has 100 times less activity than overall koji protease activity. According to protease activity during black bean koji making in traditional process, the time of koji making could thereby be shortened from 7 days to 4 days. However, protease activity of Aspergillus oryzae was fully inhibited under 17% salt, leaving salt-resistant bacteria as the primary peptide contributors during a 5-month moromi fermentation.

[00133] Thus, the moromi maturation can be potentially shortened from 6 months to 4 months with use of the Lactobacillus pobuzihii strains disclosed herein. At 4 months post-inoculation, the number of tasty peptides in the sample reached a maximum. It was shown that moromi fermentation for longer than 4 months did not provide further enrichment in taste peptides.

[00134] Therefore, in one example, the moromi fermentation and maturation as disclosed herein is shortened due to the use of Lactobacillus pobuzihii compared to a moromi maturation without the use of Lactobacillus pobuzihii.

Taste and functional peptides of black bean were enhanced due to growth Lactobacillus pobuzihii

[00135] Black bean moromi brewed with Lactobacillus pobuzihii has been shown to have exceptional taste quality, compared to yellow wheat (YW) and yellow bean (YB). moromi, which can be simply expressed as a “chicken broth” taste. To reveal these taste characteristics, the high-resolution liquid chromatography/mass spectrometry (LC/MS) was applied to identify the oligopeptides and their 3-6 mer composition (Fig. 20). Later, these profiles were matched using public databases containing various taste characteristics of peptides. It was found that black bean (BB) moromi was enriched by kokumi and salty peptides over the time of liquid fermentation, while yellow bean (YB) and wheat (YW) peptides has not changed peptides profile from the time of koji maturation (Fig. 21).

[00136] Over the time, the enrichment of short peptides was seen together with a growing abundance of Lactobacillus pobuzihii and Staphylococcus bacteriophages (Fig. 3). In trials which lacked Lactobacillus pobuzihii with a domination of Weissella, the short peptide content (as disclosed herein in Fig. 7) remained unchanged from the first month of moromi - an accumulation of 3- to 6-mer peptides was not seen (Fig. 7). Thus, without being bound by theory, it is thought that a higher abundance of Lactobacillus pobuzihii increases kokumi and umami taste by contributing up to 20% of total abundance of 3-6 peptides (Fig. 4). In contrast, while conducting fermentation trials, it was found that higher amount of Weissella were not contributing to taste peptides abundance (Fig. 2).

[00137] Thus, in one example, the fermentation process disclosed herein results in an increase in umami, and/or an increase in taste-conferring components. Also disclosed herein is the use of Lactobacillus pobuzihii for taste enhancement of moromi. In one example, as Lactobacillus pobuzihii is used as a main culture component.

[00138] Valine-proline-proline (VPP) peptide had higher abundance in black bean moromi Based on profiling of 3-6 mer peptides, higher levels of valine-proline-proline (VPP) peptides were found in soy sauce. It was further found that the enrichment of valine-proline-proline (VPP) peptides coincided with the enrichment of other 3-mer peptides over the period of fermentation time, with a maximum at the third month (Fig. 25). Moromi with an increased abundance of Lactobacillus pobuzihii in black bean (BB) soy sauce had significantly higher content of valine-proline-proline (VPP) peptides over the time of fermentation, compared to wheat (YW) moromi (Fig. 4). [00139] Thus, in one example, there is disclosed a soy sauce product comprising at least one peptide comprising the sequence of Val-Pro-Pro (VPP). That is to say, at least one of enriched peptides valineproline -proline (VPP) disclosed herein has anti-angiotensin properties. Such anti-angiotensin properties (such as, an increase is vasodilation) have been shown to result in a lowering of blood pressure in, for example, but not limited to, humans.

Black bean moromi had distinct volatile organic compound (VOC) profiles and associated with lack of Weis sella, Bacillus, Kurthia, and Enterococcus, and the presence of Enterobacteria bacteriophages [00140] Gas chromatography/mass spectrometry (GC/MS) analysis found over 150 various volatile organic compounds (VOCs) when examining three types of moromi. A comparison of volatile organic compounds is presented in Fig. 5, Fig. 6, and Fig. 27.

[00141] Black bean (BB) 5-month moromi (that is, moromi just before pasteurisation) had following distinctive flavour characteristics: benzene acetaldehyde - honey, floral rose, sweet; 4-ethyl phenol - smoky, phenolic, creosote and savoury; 2-methyl-propanoic acid - acidic sour cheese dairy; 2-hydrooxy- ethyl ester propanoic acid,- sweet, fruity, acidic; 2-methyl-, methyl ester propanoic acid - sweet, fruity;

2-methyl-, ethyl ester propanoic acid, (or ethyl isobutyrate) - citrus, cherry, strawberry; isopropyl alcohol (or isopropanol) - no flavour, trimethyl pyrazine - nutty, musty, cocoa, potato; 1 -hexanol - pungent, fusel oil; 2-methyl-, ethyl ester butanoic acid - fruity, berry with fresh tropical nuances; 2-heptenal - no odour; 1 -pentanol, 4-methyl- not found; furfural - sweet, brown, woody, bready; 3-(methylthio) 1 -propanol - onion like with a sweet and savoury flavour. At the beginning of the first months of fermentation, the flavour compounds which had distinct profiles for black bean moromi included, but were not limited to, hydroxy-2-butanone, butanoic acid, pentyl octanoate and acetic acid.

[00142] In contrast to black bean (BB) moromi, final yellow bean (YB) moromi had a prevalence of volatile organic compounds (VOCs) with stringent flavours such as, but not limited to, ethanol (phenyl ethanones, benzyne alcohol), aldehydes (2-methyl butanal), ethyl acetate.

[00143] The abundance Lactobacillus pobuzihii was correlated to higher concentration of isopropyl alcohol, isopropyl octanoate, 4-ethyl phenol, 2-methoxy phenol and 2,6-dimethyl pyrazine (Fig. 26). Bacillus amyloliquefacies had similar correlation values as shown for Lactobacillus pobuzihii.

[00144] In contrast, koji maturation using the fungus Aspergillus oryzae contributed mainly to unpleasant smelling pelargonic acid or nonanoic acid (C9:0), a neutral smell of octanoic acid (C8:0), and

3-methylbutyl ester of octanoic acid, in the first 2 months of moromi fermentation. It is also thought that the presence of Weissella (unclassified) was associated with increased 3-methyl, 2-butanol concentration in the first 3 months of black bean (BB) moromi fermentation. Weisella confusa, Kurthia_sp_Dielmo , and Bacillus subtilus were shown to contribute to the presence of 2-methyl, 2-butenoic acid in final wheat (YW) moromi. Taken together, this shows the impact of uncontrolled growth of wild cultures, such Weissella, Kurthia, and Bacillus, on the final quality of moromi flavour.

[00145] High concentrations of l-(2-hydroxy-5-methylphenyl)-ethanone (10 mg/L) were associated with the presence of Staphylococcus xylosus and Staphylococcus lentus in the final months of wheat (YW) moromi fermentation, while Enterococcus faecium was associated with ethylbenzene in wheat (YW) and yellow bean (YB) moromi. Only trace amounts of these volatile organic compounds (VOCs) were shown to be present in black bean (BB) moromi.

[00146] Concentrations of 2,6-dimethyl-pyrazine were correlated with the presence of Enterococcus faecium, Bacillus subtilus, Weissella confuse, and Kurthia sp Dielmo.

[00147] Distinctive volatile organic compounds (VOCs) for black bean (BB) moromi, such as, for example, 4-ethyl phenol and benzeneacetaldehyde, were associated with an abnormal increase of Halococcus unclassified in the final month of moromi fermentation at 13% relative abundance (to the overall microbial population), with Halococcus thailandensis and Halococcus morrhuae present in very low abundance (0.005 - 0.26%).

[00148] Interestingly, a caramel flavour found in the second month of black bean (BB) moromi was represented by the presence of maltol, 3-methyl-3-buten-l-ol, 3-octanone, butyrolactone, n-Decanoic acid and ethyl ester of decanoic acid. These fragrances were found to be lacking in other types of moromi analysed, and were associated with an abundance of 1.5% of Cronobacter bacteriophage vB CsaP GAP52 and 1.5% Enterobacteria bacteriophage CC31. Since the population of Enterobacteria was high in yellow bean (YB) moromi, the high abundance of bacteriophages was thought to contribute to Enterobacteria lysis, thereby contributing to the specific flavour profile.

Bacteriocins of wild strains were identified in moromi could contribute to their survival

[00149] Since the pilot trial al showed that Weissellla had the most impact on spoiled 1 month fermentation (Fig 17 A), it was attempted to elucidate which factors aid Weissella in competing with other species, thereby allowing this genus to be more abundant. One of strategy used was identifying bacteriocins from bacteria causing spoilage of the moromi and testing their biological activity against beneficial strains including Lactobacillus pobuz.ihii.

[00150] While blasting the sequence of Weisellin A against shotgun metagenome reads, the results identified 91 bacteriocins of heat stable class Ila with up to 70% identity to pediocin of Pediococcus acidilactici and Enterococcus faecium (see Table 17). Indeed, the only surviving lactic acid bacterial strain isolated from old moromi was Pediococcus acidilactici, Thus, without being bound by theory, it is thought that previous attempts to isolate Lactobacillus pobuzihii may have hampered by the presence of pediocin in Pediococcus acidilactici, favours the growth of Pediococcus acidilactici over other lactic acid bacteria.

[00151] While blasting the sequence of Bacillus amyloliquefaciens bacteriocins, including amylolisin, 25 genes were found to have a similarity to bacteriocins of Bacillus subtilis, Enterococcus faecium, Staphylococcus sciuri and Staphylococcus gallinarum (Table 18). Indeed, these bacteria were found to be major species present in wheat (YW) fermentation and represented a minority in black bean (BB) moromi.

[00152] While searching for a Bacillus specific bacteriocin, a sequence derived from a bacteriophage which was similar to the SPBc2 prophage-derived protein BhlA was found (MEMDITQYLSTQGPFAVLFCWLLFYVMKTSKERESKLYNQIDSQNEVLGKFSEKYDVVIEKLD KIEQNFK - SEQ ID NO: 131). Considering the amount of bacteriophages found in moromi, it is thought that the presence of bacteriocins can be due to horizontal transfer taking place between the bacteriophages.

[00153] Despite of the presence of bacteriocin sequences having been identified in the samples, diffusion disk assays with concentrated Bacillus, amyloliquefaciens #8 and Weissella paramensteroides #1 did not shown any visual inhibition of all key black bean (BB) isolates examined (Table 15) in contrast to the strains of Lactobacillus lactis B-978 and Lactobacillus lactis B-1948.

[00154] Thus, in on example, there is disclosed a method of reducing and/or suppressing growth of undesired bacteria during fermentation, the method comprising the use of the culture. Such undesired bacteria can be, but are not limited to, bacteria known to contribute to product spoilage, as well as bacteria which result in an unpalatable flavour of the product. In one example, the undesired bacterium is, but is not limited to, Weissella sp., Weissella cibaria, Weissella paramenesteroides, and Enterococcus casseliflavus.

[00155] Also disclosed herein is a method of reducing the concentration of Weissella sp. during fermentation, the method comprising the use of the culture disclosed herein.

Phage isolation for Weissella, Bacillus and Staphilococcus

[00156] Lytic bacteriophages of wild strains of Weissellla paramensteroides, Bacillus spp and Staphiloccus scuiri were not obtainable from soy sauce wastewaters or soy sauce brine, compared to positive T2 proliferation on an E. coli lawn. Previously, Weissellla cibaria bacteriophages phi YS60 had been reported in kimchi spoilage and were found to be important in controlling the dynamic of lactic acid bacteria (LAB) in kimchi fermentation. Traditional black bean (BB) moromi contains markers of a similar Weissella bacteriophage phi YS65 and presence of such phage perhaps is thought to protect fermentation from Weissella spoilage.

Trials with starter cultures revealed highly resistant population of Weissella, Pediococcus and Staphylococcus in moromi

[00157] During initial laboratory trials conducted with black beans (Table 12), it was observed that it was important, if not critical, to avoid Bacillus spore proliferation during koji maturation as it could spoil the development of the Aspergillus oryzae fungi.

[00158] While protease and amylase activity were highest during koji maturation (Fig. 29 and Fig. 30), lipase activity was significantly higher in moromi stage at day 10 (Fig. 31). Considering that conventional method of industry scale soy sauce production utilises defatted soy flakes, bacterial lipolytic activity is important for fermentation of whole beans, which still have a large amount of fats present. Therefore, specific lipase activity of starter cultures can be important for their survival, and subsequently for the formation of taste and flavour formation based on lipid transformation. Distinct flavour characteristics were found to be present in some samples, for example, that moromi fermented with Staphylococcus scuiri had a distinct “salami” smell on the second day of fermentation. It is known in the art that non- coagulase Staphyloccocus xylosus, Staphyloccocus scuiri and Staphylococcus saprophiticus are the second major group found during the ripping of salami in France, Spain and Greece.

[00159] While working on the laboratory trial, several full-scale trials for production of black bean soy sauce were undertaken to support the optimisation of black bean soy sauce production at a new factory in China. Five of a total 6 trials involved the addition of Lactobacillus pobuzihii WZ3 strain as starter culture. Further strategies were applied to enhance growth of this species in moromi.

[00160] Results of a first trial al without any starter cultures showed that there was no Lactobacillus pobuzihii or other lactobacilli reads identified. Instead, it was found that the first month moromi using 7 tons of crushed beans, with oat bran added and without bran, contained major microbes such as, but not limited to, Tetragenococcus halophilus (30%), Weissellla cibaria (23%), Weissellla paramesenteroides (20%), Kurthia sp (11%), Enterococcus faecium (5%), and Staphylococcus saprophyticus (1.5%), and lacked Lactobacillus pobuzihii (Fig. 32). In contrast, traditional moromi in vats of the first month of fermentation showed a higher abundance of Tetragenococcus halophilus (96%) with Bacillus amyloliquefaciens (1.5 %) including Bacillus bacteriophage and Lactobacillus pobuzihii (< 1%) (Fig. 17B). It appeared that due to the conditions of the trial al, during solid fermentation, Weissella spp grew without concurrent growth Bacillus amyloliquefaciens during koji maturation (as it was observed in traditional black bean (BB) moromi). Subsequently, the amount of Weissella spp increased up 50% during moromi stage. The uncontrolled growth of Weissella spp caused a poor quality of crushed black beans moromi in the pilot tank 1 and tank 2. Usage of oat bran did not protect Weissella overgrowth during koji maturation stage. Overgrowth of Weissella spp could be also controlled by the Weissella bacteriophage phiYS61, which was detected in a high abundance in vats at month 4 of traditional black bean (BB) moromi.

[00161] Staphylococcus spp and Enterococcus spp were presented only in Tank 1 with oat bran, where Weissella was reduced by 10%. Staphylcoccus. saprophyticus presented one third of total microbial population in wheat (YW) moromi during initial month under controlled 15 °C temperature. However, after 5 months of fermentation, population of staphylococci decreased down to 3%, together with an increase of Tetragenococcus halophilus (40%), Bacillus amyloliquefaciens (27%) and Weissella spp. (20%; Fig. 12C). Lactic acid bacteria, including but not limited to, Weissella spp., Tetragenoccocus spp., Lactobacillus spp., can produce antimicrobials against Staphylococcus spp. and others. Therefore, it was thought to keep a number of these lactobacteria in the moromi to prevent staphylococci overgrowth and other potential foodborne pathogenic bacteria. However, in traditional black bean (BB) moromi, staphylococci bacteriophages GH15 were observed for several months. Therefore, high titres of these bacteriophages is thought to be another type of biocontrol of staphylococci in the fermentation tanks. However, usage of bacteriophages in soy sauce is not regulated by food regulatory standards. Thus, other ways of reducing staphylococci and Weissella were undertaken during next trials. LefSe analysis confirmed that main 3 main subjects, Tetragenoccocus halophilus, Bacillus amyloliquefaciens and Lactobacillus pobuzihii play a role in initial black bean (BB) moromi formation. In contrast, in a population of trial al at the same month, the population was very diverse with Weissella, Enterococcus faecium and Staphylococcus as main contributors responsible for fermentation (Fig. 33).

[00162] Trials a2 (Table 13) and a3 (Table 12) were planned based on comparative results between black bean (BB), yellow bean (Y), and wheat (YW) moromi. More Bacillus spp. were observed in the wheat (YW) tank compared to black bean (BB) vats with total colony count more than 10⁶cfu/g. In vitro experiments showed that even at a high concentration of around 18% NaCl in the medium (v/w), Bacillus can still exist in the form of spores, which has a high risk due to proliferation once aerobic conditions are formed. However, the protease activity of Bacillus amyloliquefaciens together with Tetragenoccocus halophilus and Pediococcus spp. are known in the art, and such protease activity can contribute to taste formation of the final product during month 5 of the moromi fermentation. In trials a2 and a3, Bacillus amyloliquefaciens addition into the moromi did not improve the taste of the product. In contrast, Lactobacillus pobuzihii inoculation improved the taste and flavour reported by sensory panel (Table 13).

[00163] Shotgun metagenomics of trial a2 and trial a3 demonstrated the prevalence of Weissella, as in trial al, and showed a lack of Lactobacillus in all vats (Fig. 17B). Weissella was present for up to 70% in moromi from 1 to 7 months of fermentation under 17.5% salt conditions. Tetragenoccocus halophilus developed up to 20%, without need of its addition of as starter culture. Pediococcus pentosaceus, which was not sampled at the Xiluo factory moromi, was shown to be present in an abundance of up to 17% of the total amounts.

[00164] After these trials, Weissella cultures were isolated. It was found in vitro that strains of Weissella paramenstoides had higher growth rate compared to Lactobacillus pobuzihii strains under aerobic conditions, resulting in its higher cell count (Fig. 14D). Therefore, trial a4 was designed with the aim of favouring anaerobic conditions, providing the competitive advantage of Lactobacillus pobuzihii over Weissella. Upon completion of trial 4a, Lactobacillus pobuzihii was detected only in trace amounts where it had been added (Fig. 17C). No differences were seen in microbial composition between stirred or unstirred vats. Instead, Staphylococcus developed up to an abundance of 50% with Weissella and Kurthia, similar as to the results of trial al.

[00165] Trial b was started in February 2018 in Qingdao (China) and was conducted under more moderate weather conditions (ambient temperature in February in Quindao was around 5°C, compared to 22°C in Taiwan) than the trials in Taiwan. Vats were inoculated with monocultures of Lactobacillus pobuzihii WZ3, Bacillus amyloliquefaciens #8, or Tetragenoccocus halophilus #6 at the beginning of the moromi stage and kept at temperatures not exceeding 25°C. In addition, 6-month moromi of black beans (BB) from the Xiluo factory was used as starter culture. Initial koji microbial composition at the Qingdao factory was different from Xiluo factory, and mainly consisted of Corynebactrium spp. (20%), Staphylococcus vitulinus (20%) and Weissella spp. (10%). During the moromi stage, only a vat inoculated with Tetragenoccocus halophilus had a distinctive microbial composition, compared with the rest of vats examined. Despite of Tetragenoccocus halophilus as starter culture, this species was not detected. However, the abundance of Weissella confusa remained low (5%) with the most abundant microorganism being Macrococcus caseolyticus (up to 40%) detected in Trial a4, indicating a possible contamination by wild airborne bacteria during fermentation. Streptococcus infantarius, Enterococcus faecium, and Entrobacter cloacea represented another 15% of total abundance. Vats with Bacillus amyloliquefaciens and Lactobacillus pobuzihii lacked the enriched starter cultures. Instead, 30% of metagenomic reads came from Aspergillus oryzae, which is thought to be the result of a poor autolysis of koji mould during the moromi stage, which can be an effect of the lower temperature during winter season at Qingdao, compared to the warmer climate in Taiwan. Microbial profiles were similar between 6 months moromi inoculum, Bacillus amyloliquefacies and Lactobacillus pobuzihii. However, Weissella cibaria and Weissella confusa species (15% abundance), together with Leuconostoc pseudomomeneteroides, Leuconostoc citreum and Leuconostoc lactis, represented 22% of the total microbial abundance. A wild Lactobacillus lactis was observed to be present at around 1 % abundance, which was previously reported to be resistant to up to 7% salt and ethanol. A vat with Lactobacillus pobuzihii was shown to contain 7% of Methanosaeta concilii, a methanogen usually found in low acetate environment with strict anaerobic conditions. It is of note that these archaea Methanosaeta concilii (with an abundance of 5%) were also found in one of the traditional Xiluo vats, together with Lactobacillus pobuzihii (at 62.4% abundance; Fig. 19). This confirmed that anaerobic conditions favour Lactobacillus pobuzihii enrichment in the open environment. In other words, the presence of strictly anaerobe archae (such as, but not limited to, Methanosaeta concilii) can be used as indicators or markers of anaerobic conditions favouring the growth of Lactobacillus pobuzihii, thereby indicating the presence of microaerophilic conditions in traditional fermentation. Wild species Weissella spp. and Leuconostoc spp. were the most abundant genus found during the Qingdao trial experiments. In summary, the trial in Qingdao showed that wild species of Weissella spp, Leuconostoc spp and Staphylococcus spp were prevalent in all vats, with the exception of the vat inoculated with a Tetragenoccocus halophilus starter, where its abundance increased during the moromi stage. Vats inoculated with Lactobacillus pobuzihii and Bacillus amyloliquefaciens remained populated with the wild species.

[00166] Finally, trial C was conducted in Tai Zhou, China, with 1.3 m³ inoculum of Lactobacillus pobuzihii WZ3 (DSM 33648) in large tanks. Trial C revealed an abundance of Lactobacillus spp. of 0.5- 1% at 4 months of moromi fermentation, while wild Tetragenoccocus halophilus was present in up to 70% abundance. Since this trial was analysed using amplicon sequencing (which is a type of targeted next generation sequencing that uses PCR to create sequences of DNA called amplicons, and thereby allowing for analysis of specific genomic regions), it was understood that this lactobacilli species belonged to Lactobacillus pobuzihii. Abundances of Staphylococcus and Weissella were shown to be low at 10% and 5%, respectively. However, relative abundances of Corynebacterium, and Enterobacter, including Pantoea, were found to be high at 10-20% abundance, compared to traditional black bean (BB) moromi, where these bacteria were reported in trace amounts. Thus, while started cultures survived 4 months of moromi fermentation, they could not dominate over the entire fermentation time frame. As shown below, this issue was addressed by adapting Lactobacillus pobuzihii using osmoprotectors to protect starter cultures from salt stress prior its inoculation into brine.

Adaptation of Lactobacillus pobuzihii to osmotic stress and genes expressed with addition of soy lecithin as potential osmoprotectant

[00167] There are limited number of osmoprotectors described in the art for lactobacilli which involve Tetragenoccocus halophilus, Lactobacillus plantarum, and Enterococcus faecalis, with glycine betaine shown to be the most efficient osmoprotectant for lactic acid bacteria (LAB). Since no reports are available for the use of Lactobacillus pobuzihii with osmoprotectors, glycine betaine was tested as a primary compound, together with choline, and other chemicals containing choline molecules, such as, but not limited to, raw soy lecithin containing phosphatidyl choline (which is a by-product of defatting of soybean of the company’s factories). Ability of Tetragenoccocus halophilus to metabolise choline and accumulate the choline metabolite, glycine betaine, enabled Tetragenoccocus halophilus to be adapted to high salt conditions in soy moromi.

[00168] Successful fermentation under high salt conditions is dependent on proper adaptation of the starter culture of Lactobacillus pobuzihii strains to salt prior to their use as an inoculum for fermentation. The adaptation step must take place prior to the beginning of moromi stage, since an inappropriately adapted inoculum (or inoculum that is not adapted at all) will result in the death of living Lactobacillus cells once they immersed into a bran solution of moromi containing 17% salt. For example, food grade soy lecithin as osmoprotector, since it allowed a shortening of the lag phase of Lactobacillus pobuzihii under salt stress from 1 month to 3 days.

[00169] Lactobacillus pobuzihii WZ3 (DSM 33648) actively utilises glycerol with gldA (COG0371 Glycerol dehydrogenase and related enzymes) and the accumulation of glycerol can help Lactobacillus pobuzihii to survive under salt stress.

[00170] Thus, in one example, an adaptation step is carried out prior to inoculation. In another example, the adaptation step comprises use of an osmoprotector. In another example, the culture disclosed herein is adapted to tolerate high salt fermentation conditions.

[00171] In one example, the osmoprotector disclosed herein is selected from soy lecithin, glycerol, mannose, and mannitol or combinations thereof. In another example, the osmoprotector is soy lecithin. In another example, the osmoprotector is glycerol.

[00172] At concentrations of salt exceeding 16% (v/w) in De Man, Rogosa and Sharpe broth, growth of both Lactobacillus pobuzihii WZ3 (DSM 33648) and Lactobacillus pobuzihii WZ5 (DSM 33658) was completely inhibited (Fig. 34A). Addition of glycine betaine was not shown to protect the strains from salt stress (data not shown). However, with application of choline in same medium, Lactobacillus pobuzihii WZ3 (DSM 33648) showed a gradual shortage of lag-phase on various concentrations of salt (Fig. 34B). Various concentrations of choline did not improve the growth of Lactobacillus pobuzihii WZ3 (DSM 33648) on the filtered black bean (BB), 1-month moromi with a salt content of 17% (data not shown), indicating that other nutrients or microbial metabolites might be necessary to promote the growth of Lactobacillus pobuzihii in sterile moromi.

[00173] After choline was replaced with raw soy lecithin, a growth at the concentration of 2 mM or 4 mM of soy lecithin was observed at 5 days fermentation (Fig. 35A). Under conditions of De Man, Rogosa and Sharpe broth, Lactobacillus pobuzihii WZ3 (DSM 33648) was able to initiate growth on 16% and 17% of salt (w/v) on 10 days and 20 days, respectively (Fig. 35A and Fig. 35B). Therefore, it was able to adapt the starter culture strain Lactobacillus pobuzihii WZ3 (DSM 33648) on 5 and 10 days, which could be indicative of a shortening of lag-phage by half. Addition of tween-80 to soy lecithin did not improve the growth on salt (Fig. 35D).

[00174] Since soy lecithin provided some turbidity, the content of MRS broth was plated on De Man, Rogosa and Sharpe broth agar to confirm the growth of culture taken by optical density measurement in the presence of lecithin to avoid interference (Fig. 35B). At the highest salt concentration of 17% NaCl used in moromi with 2 mM lecithin, Lactobacillus pobuzihii WZ3 (DSM 33648) reached 2 x 10⁶ CFU/mL on 14 day and 1.6 x 10⁷ CFU/mL on 21 day of cultivation, whereas at a salt concentration of 16% NaCl, 1 x 10⁷ CFU/mL was reached on 14 day with the same concentration of lecithin. Viable counts of Lactobacillus pobuzihii WZ3 (DSM 33648) at 17% salt without lecithin were equal to the initial inoculation number of 2 x 10⁶ CFU/mL after 21 days of incubation on De Man, Rogosa and Sharpe broth. This data had been reported during trial c, in which the abundance of Lactobacillus pobuzihii at 3-4 months was equal to the abundance at the day of starter inoculation. Thus, the addition of soy lecithin was shown to support Lactobacillus pobuzihii WZ3 (DSM 33648) in tolerating the osmotic stress caused by 17% NaCl in tank with brine.

[00175] Thus, in one example, the concentration of soy lecithin is 0.5 to 4 mM. In another example, the concentration of soy lecithin is between 1 to 2 mM, between 1.5 to 2.5 mM, between 2 to 3 mM, between

2.5 to 3.5 mM, between 3 to 4 mM, between 3.5 to 4 mM. In yet another example, the concentration of soy lecithin is 1 mM, 1.25 mM, 1.5 mM, 1.75 mM, 2 mM, 2.25 mM, 2.5 mM, 2.75 mM, 3 mM, 3.25 mM,

3.5 mM, 3.75 mM, and 4 mM.

[00176] Which genes were involved in salt adaptation of Lactobacillus pobuzihii WZ3 (DSM 33648) was further investigated to find substrates might be missing in raw moromi. Comparison of RNA sequencing for Lactobacillus pobuzihii WZ3 (DSM 33648) at 2 mM lecithin with salt versus salt without lecithin added showed 4 genes which were significantly upregulated compared to 28 genes which were downregulated (Fig. 35C). The four upregulated genes on lecithin included manX_2, manX_3, manZ_4 and sorA. The first three genes (manX_2, manX_3, manZ_4) are associated with mannose transport, while sorA is responsible for sorbose transport into cell for further utilization (for example, the Pts system, sorbose-specific iic component; Sorbose-permease PTS system IIC component). Active mannose transport inside of the cell can lead to the production of mannitol, a polyol which serves as solute capable of accumulating water inside the cell during osmotic shock. Whole genome sequencing (WGS) analyses identified that the genes of mannitol dehydrogenase specifically for this strain support mannitol’s role in water accumulation during the salt stress. Soy lecithin can therefore help to facilitate mannose transport and mannitol can serve as water absorbent during salt stress.

[00177] Without lecithin addition, 138 genes were upregulated, and 145 genes were downregulated under salt stress (Fig. 35D). An upregulated enzyme (6.7 log2 fold-change compared to control) was ornithine carbamoyltransferase argF, which is responsible for L-citrulline degradation. L-citrulline is known to accumulate during soy sauce fermentation by Pediococcus acidilactici and Weissella confusa and can react with ethanol under formation of carcinogenic ethyl-carbamate during pasteurization of the final product. Pathway of degradation in Lactobacillus pobuzihii WZ3 (DSM 33648) was shown to be complete since arcC carbamate kinase co-expressed with the transcription regulator arcR. It has been reported that carbamate kinase is a terminal enzyme of the citrulline pathway ending with release CO2 and ammonia (Fig. 35H). Thus, addition of Lactobacillus pobuzihii into fermentation can contribute to utilising the pre-cursor molecule L-citrulline, and therefore deplete concentrations of cancerogenic ethylcarbamate.

[00178] In addition, Lactobacillus pobuzihii WZ3 (DSM 33648) was shown to express the antilisterial bacteriocin subtilisin biosynthesis protein AlbC, which is thought to inhibit other salt tolerant bacteria such as, but not limited to, Weissella and Staphylococcus. Another advantage of this bacteriocin can be the prevention of an overgrowth of common food pathogen Listeria monocytogens during the storage of soy sauce. Thus, in one example, the Lactobacillus pobuzihii disclosed herein is capable of synthesising an antimicrobial compound.

[00179] Monosugar sorbose is actively transported with sorA (as seen with soy lecithin as well) and converted into sorbitol with sorbitol dehydrogenase gutB showing as upregulated. A 3.3 log2 FC change in fumarate reductase was shown, and fumarate hydratase is indicative of production L-malate from fumarate.

[00180] Peptidase activity was upregulated with Xaa-Pro dipeptidase pepQ activated under salt stress which is evidence that Lactobacillus pobuzihii is actively involved in catabolism of peptides and in formation of moromi taste. Lactobacillus pobuzihii WZ3 (DSM 33648) is thought to actively integrate free amino acid histidine into its proteins with hisS (Histidine— tRNA ligase; histidyl-tRNA synthetase). This metabolism of Lactobacillus pobuzihii WZ3 (DSM 33648) can contribute to consumption of the histamine which is a known precursor of potentially allergenic histamine in moromi.

[00181] Thus, fermentation under high salt conditions is dependent on proper adaptation of the starter culture of Lactobacillus pobuzihii strains to salt prior to their use as an inoculum for fermentation. The adaptation step must take place prior to the beginning of moromi stage, since an inappropriately adapted inoculum (or inoculum that is not adapted at all) will result in the death of living Lactobacillus cells once they immersed into a brine solution of moromi containing 17% salt. Among others, food-grade soy lecithin can therefore be used as an osmoprotector, since it allows a shortening of the lag phase of Lactobacillus pobuzihii under salt stress from 0.5-1 month to 3-5 days. [00182] Each batch of moromi was therefore inoculated with IxlO⁷ of salt tolerant Lactobacillus pobuzihii cells. These Lactobacillus pobuzihii cells had been prepared in a fermenter with supplementation of 2 mM soy lecithin in De Man, Rogosa and Sharpe broth (MRS) medium or equivalent medium with an osmoprotector, prior to application in fermentation.

[00183] Until now, Lactobacillus pobuzihii species were reported only in a few studies, and it is role as a microbial strain associated with a specific taste of the fermented product was not clearly defined.

[00184] A Taiwanese study reported negative role of Lactobacillus pobuzihii, which was one of few lactic acid bacteria (LAB) associated with spoilage of canned soy sauce. The microbial community of normal and swollen canned soy sauce was investigated using PCR-denaturing gradient gel electrophoresis (DGGE), whose profiles showed that four lactic acid bacteria, including Lactobacillus pobuzihii, Lactobacillus acidipiscis, Lactobacillus piscium and Lactobacillus sp. were involved in the swollen canned samples. In addition, isolation on solid agar showed that three other diverse species of Bacillus (B. subtilis, B. oleronius and B. flexus) were also present in the swollen canned samples. Lactic acid concentrations of these swollen canned samples were significantly higher than in the normal samples with a smaller number of these species. Based on these results, it was concluded that lactic acid bacteria (LAB) play a role in contributing to the acidization of the swollen canned soy sauce products. Results confirmed the existence of Bacillus sp. and lactic acid bacteria in the packaged fermented soy sauce. It has been shown that Bacillus amyloliquefaciens can produce twice the amount of lactic acid released by Lactobacillus pobuzihii under the same conditions. Thus, the Lactobacillus pobuzihii strains disclosed herein are not spoilage agents of soy sauce product.

[00185] In another study, Lactobacillus pobuzihii reported as one of major bacterial species involved into fermentation of traditional non-salted fish sauce or ngari. Ngari is the most popular traditionally processed non-salted fish product, prepared from sun-dried small cyprinid fish Puntius sophore (Ham.) in Manipur state of Northeast India. PCR-denaturing gradient gel electrophoresis (DGGE) analysis was used for microbial profiling identified few bands of Tetragenoccocus halophilus, Lactobacillus pobuzihii, Staphylococcus carnosus, and Bacillus indicus, which are thought to be the main agents of ngari fermentation. Lactobacillus pobuzihii DNA marker band on the DGGE was only detected once, as a spike at 6^th month of fermentation, where total bacterial load was estimated to be its highest (10⁶ CFU/mL) of all 9 months of fermentation. The pH of studied ngari was between 6.2-6.7 during all fermentation period, which supports the indication that Lactobacillus pobuzihii is a non-spoilage bacterium. Also, according to previous studies, Lactobacillus pobuzihii was isolated only with addition of 5% NaCl into De Man, Rogosa and Sharpe (MRS) broth agar, despite the ngari sample coming from nonsalted environment, but required incubation for 15 days. This is in contrast to the data disclosed herein indicating 3 days of incubation to be sufficient. Thus, without being bound by theory, it is thought that Lactobacillus pobuzihii as disclosed herein is more adaptive to salt stress than previously reported. The present disclosure shows that the introduction of Lactobacillus pobuzihii required specific adaptation to a high salt environment with addition of osmoprotectors prior inoculation into large tanks. It has been demonstrated in the present disclosure that in large-scale trials with aim for industrialization, utilisation of a traditional fermented product enriched with wild Lactobacillus pobuz.ihii as surrogate for starter culture failed, and that only a pure starter culture stock can serve as the initial stock of Lactobacillus pobuz.ihii. It has further been shown that the lack of Lactobacillus pobuzihii in the trials leads to a significant decrease in accumulation of taste peptides and generic volatile organic compound (VOC) profiles similar to conventional middle quality wheat soy sauce.

[00186] Also disclosed herein is a soy sauce product made using the culture disclosed herein or the process as disclosed herein. In one example, the soy sauce product is Taiwanese-style black bean soy sauce. In another example, the soy sauce product is gluten-free.

[00187] The invention illustratively described herein may suitably be practiced in the absence of any element or elements, limitation or limitations, not specifically disclosed herein. Thus, for example, the terms "comprising", "including", "containing", etc. shall be read expansively and without limitation. Additionally, the terms and expressions employed herein have been used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the inventions embodied therein herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention.

[00188] As used in this application, the singular form “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. For example, the term “a genetic marker” includes a plurality of genetic markers, including mixtures and combinations thereof.

[00189] As used herein, the term “about”, in the context of concentrations of components of the formulations, typically means +/- 5% of the stated value, more typically +/- 4% of the stated value, more typically +/- 3% of the stated value, more typically, +/- 2% of the stated value, even more typically +/- 1% of the stated value, and even more typically +/- 0.5% of the stated value.

[00190] Throughout this disclosure, certain embodiments may be disclosed in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the disclosed ranges. Accordingly, the description of a range should be considered to have specifically disclosed all the possible sub-ranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed sub-ranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

[00191] Certain embodiments may also be described broadly and generically herein. Each of the narrower species and sub-generic groupings falling within the generic disclosure also form part of the disclosure. This includes the generic description of the embodiments with a proviso or negative limitation removing any subject matter from the genus, regardless of whether or not the excised material is specifically recited herein.

[00192] The invention has been described broadly and generically herein. Each of the narrower species and sub-generic groupings falling within the generic disclosure also form part of the invention. This includes the generic description of the invention with a proviso or negative limitation removing any subject matter from the genus, regardless of whether or not the excised material is specifically recited herein.

[00193] Other embodiments are within the following claims and non- limiting examples. In addition, where features or aspects of the invention are described in terms of Markush groups, those skilled in the art will recognize that the invention is also thereby described in terms of any individual member or subgroup of members of the Markush group.

EXPERIMENTAL SECTION

Sample site and manufacturing process description

[00194] Samples of soy sauce fermentation were collected at three different seasons: spring, fall and winter between March 2016 and August 2018, during traditional and industrial process of manufacturing of soy sauce at Wuan Chuang factory, Xiluo, Yunlin Country, Taiwan. Samples were also received on Spring 2018 to further validate the previous observations.

[00195] Traditional manufacturing is a labour-intensive procedure and used for production of high-grade black bean soy sauce and have not been changed much from date of the factory founded in 1909. Main stages of soy making involve sterilization of raw material, koji maturation, koji washing, moromi fermentation and maturation, moromi pressing, and pasteurisation. Sterilization of raw material starts with soaking dry beans in cold water (1/3 of tank volume) for 1 hour with subsequent release of remaining water from the tank and treatment with hot steam at 115°C for 15 min at 1.2 atm. This preliminary step aiming not so much to make material sterile (beans are not free from bacterial spores after the heating) but rather to facilitate the denaturation of beans proteins for ease of fungi penetration during the next stage - koji maturation. Beans must be still warm for koji maturation where Aspergillus oryzae spores are distributed evenly with trace amount of wheat flour, mixed with beans, and placed onto circular bamboo or steel trays (1 m in diameter with 5 cm side walls). Koji maturation is conducted at stacks of 15 trays stored at well ventilated room under 30°C. Since koji maturation is fermentation with exogenic reaction, temperature inside of beans reaches 36°C with maximum at the second day of incubation and slowly returned to 30°C at the end of maturation on day 5. Longer koji maturation time might cause fungi to degrade proteins and consume nutrients further and not recommended.

[00196] Once koji maturation is completed, koji is washed with 40°C tap water to clean fungi conidia from the beans surface. Within 4 hours, 44 kg of still warm clean koji is slowly cooled down to 35°C and solid content is loaded into metal or ceramic vats containing 20 litres of water with 25% NaCl (w/v). Obtained koji mash is covered by 4-5 kg of koji mixed with 2 kg of sea salt and top up by 3-5 kg of sea salt with resulting in 75 kg of net mash mass with 14-19 % NaCl (w/v) in fermentation vat (see Figure S1A). Liquid fermentation or moromi lasts for 4-6 months depending on the weather seasons and characterized by declining pH from 5.8 down to 4.8 during first two months of fermentation. At the end of fermentation, salt layer is discarded and at first, moromi is collected using perforated plastic tube with steel mesh filter allowing raw liquid moromi to be concentrated. This first filtration is marketed as a premium grade soy sauce. Remaining soy mash is distributed manually on the cotton sheets (1x1 m) forming stacks of 300 sheets and moromi is filtered by applying industry press to get the 70% of liquid and sold out as a second-grade black bean soy sauce. Remaining solid content is removed from sheets and used as animal feed or fertilizer on local farms. Filtered moromi is kept at 16°C up to 7 days before following two stage pasteurization at 90°C for 30 minutes. Upon pasteurization, necessary sugar and seasonings are added to form the final product.

[00197] In contrast to traditional way, industrial manufacturing processing has no washing step and few automatic steps was employed for mass scale production. Production of medium grade yellow whole bean (Y) soy sauce resulted in combination of traditional koji maturation (described above) with moromi stage at 20-ton tanks for fermentation in analogy to wheat soy sauce production (described below).

[00198] Production of wheat soy sauce (YW) is carried out under typical industrial way using koji maturation facility (koji room) where grinded and roasted wheat is mixed with defatted soy flakes in ratio of 55:45 and Aspergillus oryzae spores added. After maturation at 30°C for 2 days with intensive ventilation, koji is slowly mixed by dipping cold brine solution with 25% NaCl (w/v). Obtained semisolid slurry is further bypassed into a tank kept at a temperature of 17-20°C for the first month of fermentation. Salt-resistant Zygosaccharomyces rouxii is added to the cold tank to reduce the pH value to combat undesirable growth of Bacillus and Micrococcus. After, one-month moromi is transferred into 85-ton tank, where it is mixed weekly by pressured air for 2 minutes until moromi fully matured at month 5. To obtain the final soy sauce, moromi is filtered using same press as described above, heat pasteurized while adding seasoning.

Laboratory small scale trial with cultures isolated from moromi

[00199] Koji stage simulation was conducted in 800 mL Falcon tissue culture flasks with vented cap (#353138, Corning) and moromi stage was carried out in 500 mL Scott bottles with screw cap with two hose connectors (GL45, Duran) to purge air for mixing in order to reproduce factory conditions for industry scale soy sauce.

[00200] First, black beans (Wuan Chuang factory, Taiwan) were autoclaved at 121°C for 15 minutes at 1.2 atm pressure. After cooling down, 100 g of wet sterile beans were transferred aseptically into tissue culture flask and 50 mg of Aspergillus oryzae WZ spores (Japan). Tissue culture flasks were placed under 30°C incubator with ventilation (TC135S, Tintometer, Lovibond) for 3 days. Koji maturated with spores developed white mycelia after 2 days and remained dry without a beany smell was used for next stage. In case if mycelia did not develop and beans became sticky, such koji was considered as spoiled and was discarded. For liquid stage of fermentation, 230 g of matured beans were transferred into bottles with 370 mL of 23.1% of NaCl (w/v) in sterile tap water to get final salt concentration of 17-18%. Bacterial cultures were added in order according to Table 12 at a final concentration of log9 in each bottle. The amount of prepared inoculum was calculated based on the standard conversion ratio for E. coli cells as OD600 1.0 is equal to 1 x 10⁹/mL cells.

Large scale pilot trials

[00201] In total, six pilot trials were undertaken between 2017 and June 2020 at three geographical locations: (a) Xiluo (Taiwan); (b) Qindao (China) and (c) Tai Zhou (China).

[00202] Trial al started at February 2017 with crushed black beans without addition of starter culture. Rational of this trial was to move from whole beans as substrate to soy flakes as used in industrial scale wheat soy sauce production. Oat bran was chosen as porous material for more efficient ventilation of sticky crushed beans in the koji room. Moromi stages were sampled from Tank 1 with crushed beans with oat bran and from Tank 2 where koji of crushed beans without bran was fermented. Pilot tanks had 7 tons of crushed beans each.

[00203] Trial a2 was conducted from 28 July until November 2017 and involved isolated cultures from moromi as starters (see Table 6 for details). Crushed black beans were socked in NK unit, steamed and mixed with Aspergillus oryzae spores. After, koji was divided into 250 kg batches one part was sprayed with B. amyloliquefaciens culture, the second part left untreated according to scheme (Fig. 37). After 3 days of koji maturation, 72 kg of koji was transferred to individual vats. After, starter cultures Lactobacillus pobuzihii WZ3, Weissella paramentsteroides #1, Tetragenococcus halophiles #6 were added according to scheme on Table 13. Every month, samples of 100-200 mL were taken and kept at - 20°C for further genomic analysis. Inoculum of starter cultures were prepared in advance, 50 mL of stock culture was inoculated in 6L of sterile TSB medium consisted of 5g NaCl, 2.5 g K2HPO4, 17 g tryptone, 2.5 glucose in IL and incubated at 30°C for 48 hours at shaker. Content was centrifuged at 5000 g for 10 minutes, pellet was reconstituted in 25L of PBS buffer (Sigma) and around 20L was spread onto 250 kg of koji, remaining koji was sprinkled with PBS as control. In experiments Al and A2, 1.5L of same inoculum was added to each vat. First sampling was taken at 2 weeks of moromi on 15 August, T. halophilus was inoculated on 28 August, Lactobacillus pobuzihii VFZJ was inoculated on 28 September in 2 months of moromi. Samples were taken on 28 October at 3-month moromi fermentation and final sampling was performed on 28 November at the 4^th month moromi.

[00204] Trial a3 was initiated in parallel of trial a2, as excess of koji was remaining for preparation of a2. Forty-five kg of each moromi from previous batches with various months of fermentation was used as seed. Following abbreviations used: F10, F12 - Lactobacillus pobuzihii WZ3 added without koji washing; Hl - 45 kg of 1 month moromi was used as starter culture, H2- 45 kg of 2 month moromi as starter, H3 - 45 kg of 3 month moromi as starter, H4 - 45 kg of 4 month moromi as starter, Gl-12, G2-12 - washed koji, E10 - Tetragenococcus halophiles #6 no washed koji, E12 - B. amyloliquefaciens #8 no washed koji, F10 - Lactobacillus pobuzihii WZ3 no washed koji, F12 - Lactobacillus pobuzihii WZ3 no washed koji.

[00205] For trial a4 started on August 2018 and completed on February 2019, Lactobacillus pobuzihii WZ3 were grown as adapted on 10% salt in De Man, Rogosa and Sharpe (MRS) broth and introduced into moromi and kept under anaerobic conditions for the moromi stage. For series 1, 2 and 5 (A and B as duplicates) content was not stirred and closed tightly with lid covered by salt to minimize amount of oxygen; for series 3, 4 and 6 (A and B as duplicates) content was stirred weekly. Lactobacillus pobuzihii WZ3 were prepared as follows: 9 L of batch culture of Lactobacillus pobuzihii was prepared a week prior trial, max OD600 with 0.6 was reached after 10 days. Considering previous trials, the purity of starter culture was examined using 16s rRNA PCR to exclude cross-contamination of Staphylococcus or other bacteria. Cultures were reconstituted in warm PBS and IL of each culture were 1.5 x 10⁶ cells per vat. WZ3 were added after soaking of koji (3 hours) into 23% salt water to avoid the salt stress, koji were diffused into salt water and lower salt concentration down to 19%. B. amyloliquefaciens #8 were reconstituted from lyophilized powder and added in 1 x 10⁵ cells per vat.

[00206] For trial b carried out between November 2017 and March 2018, the following vats were examined: 6-month moromi used as inoculum, starter cultures T. halophilus #6, B. amyloliquefacies #8, Lactobacillus pobuzihii WZ3. Starter cultures were prepared according to the protocol described for trial a2.

[00207] Trial c was conducted between February and June 2020 and involved mashed black bean koji processed in a round machine. Lactobacillus pobuzihii WZ3 was used as starter culture at 1% inoculum as follows: Fermentation tank TK086 with total koji weight of 179 tons with pH 5.54 was added 1.79 m3 of Lactobacillus pobuzihii WZ3. TK080 with koji total weight of 157 tons with pH 5.84 mixed with 1.57 m3 of Lactobacillus pobuzihii WZ3. TK083 was control tank without addition of any culture. The temperature inside fermenters was around 30°C ±2 and the salt content was between 15.8-16.5%. Content was stirred by air at 1, 24 and 48 hours for 10 minutes, after was stirred for 5 minutes every 2-5 days. Samples were taken monthly. Samples were examined using 16s RNA amplicon sequencing for examination of microbial composition.

Samples collection and storage

[00208] Koji were collected aseptically from each site at the sterile plastic bags containing 100 g of beans, stored at -20°C at the factory overnight, delivered next day on ice to airport, and transferred within 12 hours to the laboratory in Singapore. On delivery, bags were immediately stored at -80°C.

[00209] Moromi of 400 mL were collected in the middle of vat or tank, filtered through sterile cheese cloth and remaining semi-solid content was placed into sterile 50 mL falcon tubes and stored at -20°C until air delivery within days on ice packs. On arrival, the samples were stored at -80°C. Liquid moromi supernatant was collected in another 50 mL falcon tube for GC/MS and stored at -80°C.

[00210] For metatranscriptomics, 10 mL of moromi samples were filtered through cheese cloth, collected mash with 2-3 g were further centrifuged at 5000 G for 15 minutes for removal of remaining salt solution, fixed with 5 volumes of RNAlater (Thermo Fisher Scientific, Massachusetts, USA) and delivered at room temperature (RT) to Singapore next day. On arrival, samples were stored at -80°C.

[00211] For bacterial isolation, 50 mL of mash moromi in falcon tube were kept at 4°C overnight, placed on ice and were processed next day of after sample arrived at the laboratory in Singapore.

Bacterial isolation and identification

[00212] One millilitre of moromi mesh was serially diluted to 10⁴ and 10⁵ serial dilutions and 100 pL of the final dilutions were spread on agar plates under microaerophilic conditions of anaerobic jar (Oxoid, UK) supplied with CampyGen sachet (Oxoid, UK) or under aerobic conditions. Following commercial agar medium were used and prepared according to manufacture protocols with addition 5-6% NaCl: de Man Rogosa Sharpe agar (MRS) (Oxoid, UK), Soy tryptone broth (TSB) (Oxoid, UK), Columbia Blood agar (CBA) (Oxoid, UK) supplied antibiotics (50 mg/L of vancomycin; 50 mg/L novobiocin; 50 mg/L nalidixic acid; 50 mg/L colistin). After incubation at 28°C for 2-3 days, 8-12 single colonies per plate appeared. Colonies were picked up with a sterile bacteriological loop into 1 mL sterile water and DNA templates were prepared according InstaGene Matrix protocol (Bio-Rad, USA). Bacteria were further identified by 16s rRNA Sanger sequencing of PCR products with 27F and 1492R primers at Axil Scientific, Singapore and further identified with 16s rRNA gene matching with use of blastN online tool (https://blast.ncbi.nlm.nih.gov/).

[00213] It is of note that the strains of Lactobacillus disclosed herein are distinct from, for example, Lactobacillus pobuzihii strain E100301, based on 16s rRNA analysis.

Biochemical characteristics of isolates

[00214] Pure cultures were grown in 10 mL in De Man, Rogosa and Sharpe broth (MRS) or TSB broth with 0-18% concentrations of NaCl under microaerophilic conditions formed by tightly closed caps at 220 rpm under 28°C for period up to 1 month. Growth rate was measured at 600 nm (GD600 DiluPhotometer, Implen, Germany) and pH were monitored (SevenDirect SD20, Mettler Toledo). Concentrations (g/L) of acetic acid, L-lactic acid, ethanol, AN ('ammonia') and glutamic acid were measured and assessed using enzymatic assays (Megazyme, UK).

DNA extraction

[00215] Koji and moromi samples of 500 mg each were thawed on ice, resuspended in 500 pL of lysis buffer (1% SDS v/v, 0.2 M lithium acetate, 20 mM EDTA) containing 20 pL of 5 pg/mL lysozyme (12650-88-3, Sigma) and incubated at 37°C for 30 min. After, 20 pL of Proteinase K (Maxwell 16 FFS kit, Promega) were added, tubes were incubated at 56°C for 30 minutes. Content was transferred to beadbids tubes (Lysing Matrix E, MP Biomedicals) and homogenized twice with 6.0 m/sec for 40 seconds (Fast-Prep 24 5G, MP Biomedicals). Tubes were centrifuged for 5000 x g for 10 minutes to reduce foaming, supernatant was mixed with 300 pL of lysing buffer (Maxwell 16 FFS kit, Promega, USA), loaded into cartridge and processed automatically at Maxwell 16 according to manufacture protocol (Promega). DNA quality was assessed by Nanodrop 2000 (Thermo Fisher Scientific, Massachusetts, USA) and concentrations were measured with QuantiFluor® dsDNA system on Quantus (Promega). Samples with quality 260/280 = 1.8 - 2.0 and total concentrations above 800 ng were passed to shotgun sequencing.

RNA extraction

[00216] Moromi or bacterial pellets were thawed on ice, centrifuged at 13 000 x g for 3 minutes and supernatant containing RNA was later discarded. Pellets of approximately 500 mg were processed with RNA extraction kit (RiboPure -Yeast, Invitrogen) according to the manufacturers’ protocol with slight modification. Briefly, pellet was mixed with 480 pL ice-cold lysis buffer, 48 pL of 10% SDS and 480 pL ice-cold phenol, chloroform, isoamyl Alcohol (25:24:1), vortexed for 20 seconds and homogenised with 750 pL of ice-cold Zirconia beads at 6.0 m/sec (Fast-Prep 24 5G, MP Biomedicals), placed on ice for Iminute between homogenization steps with total of 5 cycles. After, foamy material was settled at 15 000 x g for 5 minutes, aqueous phase was collected and mixed with 1.9 mL of ice-cold binding buffer. Top layer was added to 1.25 mL of fresh 100% ethanol and all content was placed on filter cartridge and centrifuged at 15 000 x g for 20 seconds. After couple of washing steps, 25 pL of RNA containing solution was eluted and treated with DNAse I (Sigma). RNA quality was assessed with RNA Screen tape with use of 4200 TapeStation (Agilent). Samples with RIN > 5 and total RNA amount more than 100 ng were processed for library generation for RNA-seq.

Shotgun metagenomics

[00217] Libraries from koji and moromi DNA extracts were prepared according to TruSeq DNA PCR- Free Library Prep (Illumina). Briefly, about 1000 ng of genomic DNA was first fragmented with acoustic shearing for 350 bp with use of Covaris E220 focused-ultrasonicator (Covaris Inc, USA) according Sonolab 7.2 software settings to Illumina LS350 bp profile. After subsequent End repair, A-tailing and adapter ligation followed by size selection with SPB beads, libraries were normalized at 4 nM and pooled with 5% of phiX for pair-end 2X150bp sequencing on NextSeq500 v2 (Illumina) with NextSeq 500 Mid Output Kit v2 300 cycles (Illumina) targeting ~ 2Gb reads per sample resulting in 20-30 mln reads per sample at TLL facility (Singapore). Demultiplexing from raw reads, and conversion to FASTQ files was performed using Bcl2Fastq v2 (Illumina).

[00218] Taxonomic and functional profiles were processed with use of BioBakery tools. Briefly, all sequenced samples were trimmed and mapped to the human genome with filtering relevant microbial reads using KneadData (https://github.com/biobakery/kneaddata). Samples with less than 20 mln reads were excluded from further analysis. Taxonomic profiles of shotgun metagenomes were produced with MetaPhlan2 pipeline using a recent library of species with clade-specific markers. Functional profiling of metagenomes reads were generated with Humann2, where Humann2 builds a sample-specific reference database from pangenomes on species identified in the same samples by MetaPhlan2.

Amplicon sequencing

[00219] Amplicon sequencing of v3-v4 region was outsourced to EHBIO gene technology (Beijing, China) and conducted only with for first sampling event to perform comparative analysis vs in-house shotgun metagenomics. Amplicon sequencing was also applied to the trial c. Metatranscriptome of moromi samples

[00220] Triplicates of samples from 1 -month (1A, IB, 1C) and 4-month (4A, 4B, 4C) conditions with distinct abundance of Lactobacillus pobuzihii confirmed by shotgun metagenomics were taken for subsequent RNA extraction described above. RNA libraries from moromi RNA extracts samples were prepared according to protocol for Universal Prokaryotic RNA-Seq with Prokaryotic AnyDeplete (Tecan, USA). Briefly, 100 ng RNA was used to generate the first and second cDNA consequently following acoustic fragmentation under Covaris E220 focused-ultrasonicator (Covaris Inc, USA) and corresponding conditions targeting 200 bp size (Duty Cycle 10%, intensity 5%, cycles/burst 200, time 180 seconds, temperature of water bath 6-8°C). After subsequent steps of End-Repair and Adapter ligation, samples were stranded by Strand Selection I and II steps followed by ribosomal depletion with RD1 enzyme mix (Tecan). The final library was amplified according to recommended protocol with initial denaturation at 95°C for 2 minutes, initial 2 cycles of amplification (95°C for 30 seconds, 60°C for 90 seconds) following 18 cycles (95°C for 30 seconds, 65°C for 90 seconds) with final extension at 65°C for 5 minutes. Libraries were submitted to Macrogen, Korea for sequencing at NovaSeq platform 150PE (Illumina) to get 6Gb data per library which corresponded ~ 40 mln reads. Functional profiling was generated using Humann2 (Franzosa, McIver et al. 2018).

Whole genome sequencing of Aspergillus oryzae and Lactobacillus pobuzihii strains

[00221] Short reads for Lactobacillus pobuzihii WZ3 and Lactobacillus pobuzihii WZ5 were generated using the True Seq protocol described above. In addition, long reads were produced by MinlON (Oxford Nanopore, UK) using the standard protocol. Assembly with single contigs and error correction was processed with SPAdes via Unicycler 0.4.9b. Annotation was performed with Prokka. Assembled genomes were deposited in the JGI database https://gold.jgi.doe.gov/ with under study ID Gs0157280 and live bacteria were deposited in DSMZ collection under accession numbers DSM 33648 (Lactobacillus pobuzihii WZ3) and DSM 33658 (Lactobacillus pobuzihii WZ5).

RNA-seq of Lactobacillus pobuzihii WZ3 under salt stress

[00222] Cultures of Lactobacillus pobuzihii WZ3 were grown in triplicates at De Man, Rogosa and Sharpe broth (MRS) only, MRS supplemented with 17% NaCl, MRS supplemented with 17% NaCl and 2 mM soy lecithin (Wilmar) at 30°C for 10 days. The supernatant was removed, and samples were stored immediately at -80°C. RNA extraction and sequencing reads were generated followed by protocol for metatranscriptome of moromi samples (see above). Forward and reverse raw reads were processed by ProkSeq with reference files for Lactobacillus pobuzihii reference genome. Countfile transcript reads were further normalised and processed with SARTools 1.7.4 scripts utilised DESeq2 and EdgeR under R studio with generation of expressed files and generation of graphs including VolcanoPlot.

Salt measurement

[00223] Salt was measured using Mohr’s method. Briefly, 5 mL of sample was dissolved with 100 mF water and mixed with 2 mL of 5% potassium chromate indicator solution (Sigma). Content was titrated with 0.1N silver nitrate (Sigma). Percentage of salt was calculated whereby “(%) = (58.443 x A x (Vs- Vo))/m”, wherein A = concentration of silver nitrate solution (mol/L); Vs - volume of silver nitrate solution used in titration of test sample (mL); Vo = volume of silver nitrate solution used in titration of blank (mL); M = Mass of test sample (mL); 58.443 = Molecular weight of NaCl.

Taste peptides analysis

[00224] Soy sauce moromi (1 ml) was passed through 0.2 pm syringe filter to remove particles, diluted lOx with mobile phase A and centrifuged for 10 minutes at 14,000 rpm. The supernatant of 20 pl was injected directly into the LC-MS. Analysis was done on Vanquish LC coupled to Q Exactive Plus MS (Thermo Scientific). LC parameters are as followed: mobile phase A is 0.1% formic acid in water; mobile phase B is 0.1% formic acid in methanol, flow rate is 50 pl/min, gradient is 2% B at 0-3 min, 2-30% B at 3-20 minutes, 30-99% B at 20-21 min, 99% B at 21-25 minutes, 99-2% B at 25-26 minutes and hold at 2% B until 30 minutes. The column was ACE AQ 1.0 x 150 mm, 3.0 pm (ACE, UK), column temperature was 45°C. Sampler temperature was 4°C. MS was set in fullMS-ddMS2 mode which performs a full scan followed by data-dependent MS/MS for top 10 most abundant masses. Each sample was run in triplicates. Data was analysed using PEAKS software (Bioinformatics Solutions Inc., Canada) de novo sequencing to generate peptide lists.

[00225] Taste-conferring components, taste-conferring peptides, and/or taste-conferring amino acids: 3-6 amino acid peptides are categorised into umami, kokumi, sweet, salty, sour and bitter groups, which are reflected in BIOPEP database http://www.uwm.edu.pl/biochemia/index.php/en/biopep. Thus, depending on what type of peptides are prevalent in a specific moromi, one can expect the taste characteristics defined by these groups of peptides.

[00226] In one example, the taste-conferring components are, but are not limited to, glutamate, inosine monophosphate (IMP), guanosine monophosphate (GMP), hydroxy-2-butanone, butanoic acid, pentyl octanoate and acetic acid, as well as the compound listed in the heatmap figure presented herewith.

Volatile organic compounds (VOC) analysis

[00227] One mL sample was centrifuged with 10 000 G for 15 minutes and 200 pL of sample were later used for GC/MS analysis for untargeted VOC content. Briefly, for Gerstel Dynamic Headspace (DHS) analysis, samples were pre-incubated in DHS module of the Gerstel MPS unit, at 40°C for 3 minutes to release the VOC. The volatiles were trapped in Tenax tubes (Tenax TA, Gerstel) at flow rate of 50 mL/min and desorbed from Tenax tubes into the Thermal Desorption Unit (TDU) by heating to 280°C with a hold time of 5 minutes before injecting into the GC column. The volatiles were separated in GC (Agilent 7890B) on a DB-FFAP column (Agilent 60 m x 250 pm x 0.25 pm). The column was held at 40°C for 2 minutes, temperature was increased at 5°C/min to 150°C, followed by 10°C/min to 240°C and held for 10 minutes. Helium was used as carrier gas at a constant rate of 1.8 mL/min. The MS (Agilent 5977B) scans from 40 - 350 amu with ionization energy 70eV and transfer line temperature of 250°C. Protease, amylase and lipase activity

[00228] Freshly inoculated cultures were re-inoculated into fresh media in a 96-deep well plate at an initial OD600 of 0.1. After 24 hours, the plates were centrifuged at 3000xg for 10 m minutes in and protease, amylase and lipase activities were measured as following.

[00229] Protease kit uses the conjugated protein FITC-casein as substrate. Protease-catalysed hydrolysis releases highly fluorescent-labelled peptides; the accompanying increase in fluorescence is proportional to protease activity and can be conveniently measured in a continuous assay format using Infinite M Nano+ plate reader (Tecan, Zurich, Switzerland) with EX/EM= 490/520 nm.

[00230] Amylase activity in the supernatant was measured according to the EnzChek™ Amylase Assay Kit (E33651, Invitrogen, Carlsbad, US). The assay is based on the detection of highly fluorescent BIODIPY FL dye-labelled peptides released by amylase-catalysed hydrolysis. Amylase activities in the culture were expressed in fluorescence intensity units measured with an Infinite M Nano-i- plate reader (Tecan, Zurich, Switzerland) with a filter fluorometer (excitation wavelength -505 nm, emission wavelength 512 nm). During the second round of screening, protease activity for each of the analysed secretion tags was evaluated in triplicate. Lipase activity was evaluated with 405 nm with p-nitrophenol palmitate according to standard method.

Phage isolation

[00231] Single and double agar layer methods were applied according to standard protocols (protocol numbers 2001a 2012, 2001b 2012, EPA (USA)). Briefly, for double layer, wild strains Lactobacillus pobuzihii, T. halophilus, B. cereus, W. paramensteroides, B. amyloliquefaciens, Staphylococcus scuiri were grown at MRS on 30°C until OD600 0.8. After, 100 pL of fresh cultures (4-6 hours) were mixed with 25 mL of warm 0.5% MRS agar with 0.5 mM CaCL and poured into 50 mL of pre-warmed sample of Xiluo factory wastewater or vat brine (14% NaCl w/v) or 0.5 - 5-month fresh moromi were added and mixed briefly with palms hand. After 2 minutes of incubation at water bath with 30°C, content was plated on 2% MRS agar.

[00232] For single layer, 1.5% MRS agar mixed with 100 pL of 0.22 pm cellulose acetate membrane (Millipore) filtrate of sample free of bacteria was mixed with 0.5% MRS agar included potential host. Plaques were examined in parallel with reference positive control lytic phage T4 mixed with E. coli K12as a host.

Bacteriocins screening

[00233] To assess the main bacteriocin producers, such as Bacillus subtilus and B. amyloliquefacies (amylolisin) and Weissella paramensteroides (weissellin-A) species on other microbes in koji/moromi, metagenomes and WGS sequencies from black been fermentation were searched against conservative bacteriocins in approach previously used.

[00234] For bacteriocin screening we used method described before. Nisin (Sigma- Aldrich) was used as a reference stock solution. A standard stock solution of nisin containing 1x105 lU/mL was prepared, by dissolving 100 mg of nisin in 0.02 M HC1 (1 mL) and adding 9 mL of distilled water. Nisin was added at concentrations of 100 lU/g and 500 lU/g, respectively to the bacterial solution. Diffusion disk (6 mm in diameter) assays was used was to test colonies of interest on solid agar MRS.

Folin method

[00235] Briefly, 1 mL of supernatant was taken into 5 mL 1% casein solution, incubated at 37dC for 10 min. After, reaction was stopped by TCA solution (Sigma) and content was filtered with 0.45 um using syringe. 1 mL of filtrate was added to 5 mL of sodium carbonate solution to neutralise pH and 1 mL of Folin’s reagent (Sigma) was added to bind free tyrosine with incubation at 37dC for 30 min. Product of the reaction was analysed at 660 nm by spectrophotometry. Amount of protease was calculated based on calibration curve with free tyrosine reacted with Folin’s reagent.

SEQUENCE LISTING

Claims

89

CLAIMS A culture comprising Lactobacillus pobuzihii WZ3 (DSM 33648), or Lactobacillus pobuzihii WZ5 (DSM 33658), or a combination thereof. The culture according to claim 1 , further comprising wild-type Lactobacillus pobuzihii. The culture according any one of the preceding claims, where the culture is adapted to tolerate high salt fermentation conditions. The culture of any of the preceding claims for use in fermentation. The culture of claim 4, wherein the fermentation is soybean fermentation, or moromi fermentation. A process of fermentation comprising a starter culture comprising at least one Lactobacillus pobuzihii strain. The process of claim 6, wherein the Lactobacillus pobuzihii strain is selected from the group consisting of wild-type Lactobacillus pobuzihii; Lactobacillus pobuzihii WZ3 (DSM 33648); Lactobacillus pobuzihii WZ5 (DSM 33658); a combination of Lactobacillus pobuzihii WZ3 (DSM 33648) and Lactobacillus pobuzihii WZ5 (DSM 33658); a combination of wild-type Lactobacillus pobuzihii and Lactobacillus pobuzihii WZ3 (DSM 33648); a combination of wildtype Lactobacillus pobuzihii, Lactobacillus pobuzihii WZ5 (DSM 33658); and combination of a wild-type Lactobacillus pobuzihii, Lactobacillus pobuzihii WZ3 (DSM 33648), and Lactobacillus pobuzihii WZ5 (DSM 33658). The process according to any one of claims 6 to 7, wherein the starter culture further comprises bacteria selected from the group consisting of Tetragenococcus halophilus, Bacillus amyloliquifaciens, and combinations thereof. The process of claim 6, wherein the fermentation comprises the steps of

■ a koji maturation,

■ one or more optional washing steps,

■ moromi fermentation and maturation,

■ moromi pressing, and

■ pasteurisation. 90

10. The process of claim 9, wherein the starter culture is the moromi fermentation starter culture.

11. The process according to claim 9, wherein the moromi fermentation and maturation of step c is shortened due to the use of Lactobacillus pobuzihii compared to a moromi maturation without the use of Lactobacillus pobuzihii.

12. The process according to any one of claims 6 to 11, wherein the at least one Lactobacillus pobuzihii strain is adapted to tolerate high salt concentrations in an adaptation step.

13. The process according to claim 12, wherein the adaptation step is carried out prior to inoculation.

14. The process according to any one of claims 12 to 13, where the adaptation step comprises use of an osmoprotector.

15. The process of claim 14, wherein the osmoprotector is of soy lecithin.

16. The process according to any one of claims 6 to 15, wherein fermentation results in an increase in umami, and/or an increase in taste-conferring components.

17. The process according to claim 16, wherein the taste-conferring components are selected from the group consisting of glutamate, inosine monophosphate (IMP), guanosine monophosphate (GMP), hydroxy-2-butanone, butanoic acid, pentyl octanoate and acetic acid.

18. The process according to any one of claims 9 to 17, wherein the moromi fermentation is performed under microaerophilic conditions, or under conditions which are sufficiently anaerobic to enable preferential growth of Lactobacillus pobuzihii over Weissella and Staphylococcus.

19. The process according to any one of claims 6 to 18, wherein bacteriophages are present in the starter culture.

20. A soy sauce product made using the culture disclosed in any one of claims 1 to 5 or the process disclosed in any one of claims 6 to 19.

21. The soy sauce product of claim 20, wherein the soy sauce is Taiwanese-style black bean soy sauce. 91 The soy sauce product of any one of claims 20 to 21, wherein the soy sauce product comprises at least one peptide comprising the sequence of Val-Pro-Pro (VPP). The soy sauce product of any one of claims 20 to 22, wherein the soy sauce product is gluten- free. A method of reducing and/or suppressing growth of undesired bacteria during fermentation, the method comprising the use of the culture according to any one of claims 1 to 5. The method of claim 24, wherein the undesired bacteria is selected from the group consisting of Weissella sp. , Weissella cibaria, Weissella paramenesteroides, and Enterococcus casseliflavus. A method of lowering undesirable taste components in fermented black bean soy sauce, the method comprising use of the culture according to any one of claims 1 to 5. The method of claim 26, wherein the undesirable taste components are selected from the group consisting of ethanol, aldehydes, octanoic acid, ethyl acetate, methyl-butanol, benzene acetaldehyde, 4-ethyl-2-methoxy phenol, benzaldehyde, acetaldehyde, octanoic acid ethyl ester, 2-furanmethanol, propanoic acid, 2-hydroxyethyl ester, 3-methyl-l -butanol. The method of claim 27, wherein the undesirable taste components are only considered to be undesirable taste components if a respective odour threshold is surpassed. A method of increasing desirable taste components in fermented black bean soy sauce, the method comprising use of the culture according to any one of claim 1 to 5, wherein the increase in desirable taste components is characterised by the increase in taste conferring peptides and/or taste conferring amino acids in relation to total protein content of a sample. The method of claim 29, wherein the taste conferring peptides and/or taste conferring amino acids are selected from the group consisting of VPP, EV, EE, DES, ED, K, EEDGK, KGSLADEE, D, DE, DD, DEE, VE, VD, KGDEE, ADE, EGS, ES, EDD, EED, DDE, DED, DDD, EEE, EDE, E, SLAKGDEE, SLADEEKG, KGDEESLA, DA, VG, VV, LE, EL, EG, EY, V, P, SPE, EEN, EPAD, VGV, FFRPFFRPFF, GP, RRPFF, VYPFGGGINH, PR, LK, WP, FFPG, RGPPF, GGP, RKE, RPGGFF, YGY, RGPPGGF, RGPPGIG, PK, RGPPFIVRGPPFIV EM, GGFFGG, VF, DLL, YGG, PGI, APPPPAEVHEV, RGPPFIVGG, FPK, AED, GFF, RPFFRPFF, VL, RGGFIV, FP, AE, PPF, GGLGG, VVGET, EVG, FIV, VW, GR, AAA, RPFF, GGV, KPK, SAEQK, WW, LLLL, RGPFPIV, YPF, FPF, GLGG, GGLG, RG, PGP, AA, GLGGG, KP, FGGF, 92

VVAPFPEV, WWW, LGG, FGFG, LVYPFPGPIHN, LT, WL, GGA, RGPPGFF, LF, PGPGPG, RGPFPIIV, EH, PP, YPFPGPI, FPPFIV, GGY, VYP, GGFF, EQ, LLL, GV, VY, FFF, R, YY, GPFP, GLL, GAA, RRPP, FGG, PSE, IL, IV, QEEL, RRPPFF, GGRGPPFIVGG, DG, LG, KGD, RPFG, AD, LLG, RGFF, GE, EGG, VYPFPPGI, PFIV, PIP, FF, APPPPAEVHEVV, RPFFGG, RGPPFFF, RGPEPIIV, RR, LI, GPPFIV, AGA, YQEPVLGPVRGPFPIIV, PVRGPFPIIV, FFPP, GGGGL, APPPPAEVHEVVE, YG, VVAPFP, VGG, PGG, GPFPI, RRPPGF, LGGGG, VAPFP, FGF, GPVRGPFP, FV, VIFPPG, PL, WF, EINEL, EF, GY, FFPGG, PFPGPI, KF, GAG, INEL, FW, AF, VAPFPE, EEL, RGPPFIV, PFPIIV, VYPFPPGIGG, FFPR, RGPPFGG, RGPPGGGFF, APFPE, VYPFPPG, GGVVV, VIIPFPGR, W, RGP, FL, ELL, RP, L, YYY, RGPPGFG, PPP, RGPKPIIV, RGPPGGV, VV APFPE, RRPPPFFF, PVLGPV, RPPFIV, VIF, PPG, GW, FFGG, RRR, RPG, RGPPFF, GLG, FFG, RPPPFFF, G, VPPFL, RGPPGGFF, VYPFPPGINH, PPFIV, GF, LGL, YYG, ECG, KPF, VIIPFPG, GGGLG, YF, RF, AAG, GPG, RGPPFIIV, LA, GGRGPPFIV, DV, DL, EQEEL, FG, F, APPPPAEVHEEVH, PF, FPP, LGGG, AGG, GL, GGGL, NVVGET, GGF, PFPP, RPFFRPFFRPFF, PFPGPIP, KPPFIV, IF, SLA, A, LQ, VI, YP, GYG, GRP, RGPFPI, RPGF, VIFPPGR, GGRPFF, RPF, LEQL, YPFPGPIHNS, VIPFPGR, EC A, PFP, PGR, VYPFPPIGNH, AL, GGRPFFGG, GPPF, FI, IG, GI, GPFF, FFPE, II, RGPGPIIV, El, LL, RPGFF, PGPIP, IW, VRGPFP, VYPF, RGPFIV, LV, GYY, LD, LW, PVLGPVRGPFPIIV, GFG, ENINEL, GGL, RGPPFI, and RGPPGF. The method of any one of claims 29 to 30, wherein the taste conferring peptides and/or taste conferring amino acids is/are selected from the group consisting of EV, EE, DES, ED, EEDGK, KGSLADEE, D, DE, ECG, LT, EH, EQ, ECA, LA, LQ, VGV, AAA, AA, GGA, GAA, KGD, AGA, VD, KGDEE, ADE, E, SLAKGDEE, SLADEEKG, KGDEESLA, SPE, DD, DEE, EDD, VE, DA, VG, VV, LE, and RPGGFF. A method of reducing the concentration of Weissella sp. during fermentation, the method comprising the use of the culture according to any one of claim 1 to 5. Lactobacillus pobuzihii WZ3 (DSM 33648). Lactobacillus pobuzihii WZ5 (DSM 33658).