Kimchi

Kimchi, which was used for microbial community analysis [14], was used in this study. In brief, baechu-kimchi was home-made kimchi in Sangju, Gyeongsangbuk-do, and which was used as a control group. Baechu-kimchi was made by adding carrot, chives, garlic, ginger, jeotgal, leaf mustard, and red pepper into salted baechu and jogi—baechu—kimchi with approximately 5% jogi added to the total volume of the control group was used as an experimental group. Kimchi was immediately stored at 10 °C for 20 days, sampled every 5 days, and stored in a deep freezer at −80 °C. In order to reduce the prejudice of kimchi collected part, all samples were ground and used in the experiment.

Color analysis

Kimchi sample for color difference analysis was ground with a blender, and 1 g was taken to measure the values of lightness (L), redness (a), and yellowness (b), which are the components of the Hunter color system, and ΔE values were derived. The chromaticity was measured by colorimeter (Chroma Meter CR-400, Konica Minolta, Tokyo, Japan) calibrated with a standard white plate (L = 93.28, a = −0.58, and b = 3.50). All experiments were conducted in triplicate.

Reduction sugar analysis

For measurement of the reducing sugar, kimchi sample was homogenized, filtered through sterilized cheesecloth, and a filtrate was used. Reduction sugars were measured by a colorimetric method using dinitrosalicylic acid (DNS) [3, 15]. The DNS reagent of 3 mL was added to 1 mL of a 50-fold diluted filtrate, stirred well, reacted in boiling water for 5 min, and cooled to measure the absorbance at 550 nm by UV/VIS spectrophotometry (Optizen POP, Mecasys Co., Ltd, Daejeon, Republic of Korea), and this measurement was then converted into glucose. All experiments were conducted in triplicate.

Organic acid analysis

Ground samples (5 g) were mixed with 45 mL of sterilized water. The mixture was sonicated for 10 min, and then filtered through a 0.2 μm syringe filter (Sartorius Stedim Biotech GmbH, Göttingen, Germany). The filtrate was analyzed by a custom service provided by the National Instrumentation Center for Environmental Management in Korea (http://nicem.snu.ac.kr/). The analyses were performed using the Dionex Ultimate 3000 HPLC system (Thermo Scientific, Waltham, MA, USA). Chromatographic separation was conducted with an Aminex 87H column (300 mm × 10 mm; Bio-Rad, Hercules, CA, USA). Gradient elution was carried out with 0.01N H₂SO₄. Injection volumes were 10 μL, and the column temperature and flow rate were 40 °C and 0.5 mL/min, respectively. The detection was performed using a refractive index (RI) detector (RefractoMAX520, ERC Inc., Saitama, Japan) at 210 nm. The organic acid contents of two samples prepared in the same conditions were analyzed in triplicate.

Analysis of free amino acids

The same filtrate that was used for the analysis of organic acids was analyzed using a JASCO LC-2000 Plus Series HPLC system (JASCO, Tokyo, Japan). Chromatographic separation was conducted with an AccQ-Tag Amino Acids C18 Column (3.9 mm × 150 mm, 4 μm; Waters Corporation, Milford, MA, USA). Gradient elution was carried out with AccQ-Tag/water (solvent A; 10:90, v/v) and acetonitrile/water (solvent B; 60:40, v/v). The following binary mobile phase linear gradients were used: 100% A at 0 min, 98% A at 5 min, 93% A at 15 min, 90% A at 19 min, 67% A at 32–33 min, 0% A at 34–37 min, and 100% A at 38 min. Injection volumes were 10 μL, and the column temperature and flow rate were 37 °C and 1 mL/min, respectively. The detection was performed using a fluorescent detector. AccQ-Fluor Reagent Kit (WAT052880, Waters Corporation, Milford, MA, USA) was simultaneously used as derivatizing agents, according to the manufacturer’s instructions. The photodiode array (PDA) detector (MD-2018 Plus, JASCO, Tokyo, Japan) wavelength was set to 254 nm for the determination of the AccQ-Fluor Reagent-derivatized amino acids. The concentrations of individual free amino acids were determined using five-point calibration curves of the amino acid standard (WAT088122, Waters Corporation, Milford, MA, USA). The free amino acid contents of two samples prepared in the same conditions were analyzed twice.

Biographical correlation analysis between bacterial communities and free amino acids

Bacterial communities at species level in the two types of kimchi were analyzed for correlation with the free amino acids using the Pearson correlation coefficient. Correlation analysis was performed using the Paleontological Statistics software package (PAST) version 4.13 [16].

Comparative genomics

Comparative genomic analysis was performed to explain the production of amino acids produced in kimchi via genomic insights. Ten species assumed to be correlated with amino acid production were selected. Genome information was prioritized for species registered with complete genomes and type strain, and next selected as species registered with complete genomes, except for type strain and draft genomes. The genomic information for the 10 species was retrieved from the National Center for Biotechnology Information database (NCBI; http://ncbi.nlm.nih.gov/genomes): Leuconostoc (Leu.) mesenteroides ATCC 8293^T (GenBank accession no. CP000414), Leu. lactis DMLL10 (GenBank accession no. CP116456), Bacillus (B.) subtilis NCIB 3610^T (GenBank accession no. CP020102), B. velezensis DMB06 (GenBank accession no. CP083763), B. siamensis B28 (GenBank accession no. CP066219), B. amyloliquefaciens DSM 7^T (GenBank accession no. FN597644), B. aerophilus KJ82 (GenBank accession no. CP091093), B. paralicheniformis 14DA11 (GenBank accession no. CP023168), Klebsiella (K.) grimontii SS141 (GenBank accession no. CP044527), and Mammaliicoccus (M.) sciuri NCTC12103^T (GenBank accession no. LS483305) Amino acid biosynthetic pathways were predicted using Rapid Annotations using the Subsystems Technology (RAST) server for SEED-based automated annotation [17], then confirmed using the iPath (ver. 3) module [18] and CLgenomics ver. 1.55 software.

Statistical analysis

Principal component analysis (PCA) was performed on the content of amino acids detected in both of the kimchi samples to visually differentiate between the bacterial communities of each kimchi sample during fermentation, using SPSS software v.27 (SPSS Inc., Chicago, IL, USA).

Effects of jogi addition on color value

The effect of jogi addition on the color of kimchi was confirmed. In the case of BK without jogi, the L, a, and b values were 37.50 ± 0.22, 15.31 ± 0.18, and 22.78 ± 0.24, respectively. In the case of JBK with jogi, the L, a, and b values were 38.77 ± 0.17, 16.21 ± 0.10, and 22.80 ± 0.19, respectively, showing similar results to BK (Table 1). During the fermentation, L and b values decreased slightly, while in the case of BK, the a value increased slightly. However, in the case of JBK, the L, a, and b values all increased. These results assumed that the increase in L, a, and b values was due to jogi addition, but in the case of kimchi with Alaska Pollack, the a and b values decreased [19], and thus we suggest this to be a change by case, rather than a change due to seafood addition. The a/b ratio, which is the ratio of red to yellow, increased from 0.67 to 0.72 on the 20^th day of fermentation, and from 0.71 to 0.74. In the case of the overall change in color (ΔE), as the storage period increased, ΔE tended to increase, and all groups were in the numerical range 0 − 2.62, making the change difficult to distinguish with the naked eye. As the kimchi fermented, all samples, such as kimchi cabbage, seasoning, ingredients, and jogi, were mixed and aged, and the light yellow—white color of the jogi itself affected the L, a, and b values of kimchi, but the degree was weak, so there was no significant difference in color overall.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Change of color of Baechu—kimchi and Jogi—baechu—kimchi during fermentation.

https://doi.org/10.1371/journal.pone.0300249.t001

Effects of jogi addition on reducing sugar production

In kimchi fermentation, reducing sugar is used as a carbon source for microorganisms, and is decomposed into organic acids and CO₂; thus as fermentation proceeds, the content of reducing sugar decreases [20]. Since this decomposition of reducing sugar gives kimchi its unique taste and flavor, the degree of aging of kimchi, the degree of growth of microorganisms, and changes in flavor are also evaluated by examining the changes in reducing sugar [3]. To confirm the effect of jogi addition on the change in reducing sugar, the reducing sugar was analyzed. In the case of BK, on the 5^th day of fermentation, the reduction sugar content increased; but as fermentation progressed, it decreased (Fig 1). In addition, in the case of JBK, as fermentation progressed, it decreased, and the content of reducing sugar was less on 20^th day of fermentation, compared to the BK without jogi. As a result of the following literature survey to check if this was the result of jogi addition, the kimchi group with pollack added had a higher content of reducing sugar than the kimchi group without Pollack [19]. However, kimchi groups with octopus or squid were found to have lower levels of reducing sugar than kimchi groups without octopus or squid [7]. As a result of difficulty in affirming that by adding seafood, the reduction sugar has decreased or increased, these results are assumed to derive different results for each case. However, as fermentation progresses, reducing sugar clearly decreases.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Effects of jogi addition of baechu—kimchi on reducing sugar production.

BK, baechu—kimchi; JBK, jogi—baechu—kimchi; the following number indicates the fermentation time (in days). All values are presented as the mean ± SD of triplicate determination. Different letter indicates significant difference at p<0.05 using Duncan’s multiple range test.

https://doi.org/10.1371/journal.pone.0300249.g001

Effects of jogi addition on organic acid production

Five organic acids and ethanol were identified in the two types of kimchi during fermentation (Fig 2). Citric acid was a major organic acid in the early stages of fermentation, but as fermentation progressed, it tended to gradually decrease. As fermentation progressed, acetic acid, ethanol, and lactic acid increased significantly, of which lactic acid increased the most. In comparison, the contents of formic acid and fumaric acid were not high during the fermentation period. Above all, formic acid showed no pattern of increase or decrease during the fermentation period. Accordingly, these results assumed that those two organic acids did not have a significant effect on kimchi fermentation. To determine the effect of jogi addition, we tried to confirm the change in organic acids between baechu—kimchi and jogi–baechu—kimchi, but it was confirmed that the effect on the production of organic acids was not significant.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Effects of jogi addition of baechu—kimchi on organic acid production.

BK, baechu—kimchi; JBK, jogi—baechu—kimchi; the following number indicates the fermentation time (in days). All values are presented as the mean ± SD of triplicate determination. Different letter indicates significant difference at p<0.05 using Duncan’s multiple range test.

https://doi.org/10.1371/journal.pone.0300249.g002

Kimchi is a representative lactic acid fermented food, and the results of our experiments show that the amount of lactic acid increases over time, indicating that fermentation is proceeding well. What is interesting is that the sum of the acetic acid and ethanol contents is approximately same as the content of lactic acid. Ethanol, as well as lactic acid, were produced from the same molecule from one molecule of glucose via the hetero-fermentative pathway (or phosphoketolase pathway) under anaerobic condition [21]. On the other hand, the homo-fermentative pathway is known to produce lactic acid and acetic acid from glucose in the presence of oxygen [22]. Kimchi is not a perfectly controlled anaerobic fermentation. Therefore, based on the results, it was estimated that hetero-lactic fermentative LAB in contact with oxygen produces acetic acid, and in its absence, ethanol. To confirm this, the same kimchi used in this experiment was compared with the results of bacterial communities, which results are shown in previous studies [14]. It was found that the amount of bacterial communities was related to organic acids; Acetic acid, EtOH, and lactic acid increased from day 10 of fermentation (Fig 2), while Leu. mesenteroides, a heterolactic-fermentative LAB, was found to increase from day 10 and 5 of fermentation in the NGS results and culture-dependent results, respectively [14] (S1 Fig). These results suggest that acetic acid, EtOH, and lactic acid are produced by the influence of Leu. mesenteroides. However, in the culture-independent method (S1 Fig), Lactobacillus sakei, homo-lactic fermentative LAB, was more predominately detected than Leu. mesenteroides. Lb. sakei only produces two molecules of lactic acid from one molecule of glucose, so the lactic acid content should have been higher than the sum of acetic acid or ethanol [23]. Based on the content of organic acids (Fig 2), we assumed that during kimchi fermentation, Leu. mesenteroides could have influenced the production of organic acids, not Lb. sakei. To date, it is known that there is a difference in microbial aggregation according to culture-dependent and non-dependent methods [12, 14, 24], but considering the relationship with fermented products, we judged that it was more correlated with the results of culture-dependent methods.

Effects of jogi addition on amino acid production

The effect of jogi addition on the microbial community was weak [14], and there was little effect on reducing sugar, organic acids, and color (Figs 1 and 2, Table 1). In general, kimchi with seafood, such as oysters, is delicious right after being made, and it is recommended to speed up consumption. Above all, it has been reported that the texture is related to the rich taste immediately after making kimchi [8, 19]. The amino acids that affect the umami flavor are glutamic acid and aspartic acid, and thus check the free amino acid containing two amino acids (Table 2). The increasing amino acids in fermentation for BK are alanine, arginine, leucine, lysine, phenylalanine, proline, tyrosine, and valine. The decreasing amino acids are aspartic acid, cysteine, glutamic acid, histidine, isoleucine, methionine, and threonine. In addition, JBK with jogi also shows similar results. Interestingly, the contents of glutamic acid and aspartic acid, which are rich taste components, in the early stages of fermentation, are higher in JBK than in BK (Table 2). Above all, on day 0, aspartic acid was three times higher than BK, while at day 15 of the fermentation, it was about 1.5 times higher. On day 0, glutamic acid was 1.2 times higher than BK, while on day 15 of fermentation, it tended to become similar. In addition, throughout the fermentation, alanine, arginine, glycine, isoleucine, leucine, lysine, methionine, phenylalanine, tyrosine, and valine showed higher levels than BK. In contrast, histidine, serine, and threonine were detected high in BK, and in the early stages of fermentation, showed higher values. This indicates that jogi addition is related to the content of amino acids, rather than color, reducing sugar, and organic acids. However, it is considered necessary through additional experiments to verify whether in this experiment, an effective amino acid is produced by the addition of seafood, such as premature fish, or an amino acid was produced only in this fermentation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Effects of jogi addition of baechu—kimchi on amino acid production.

BK, baechu—kimchi; JBK, jogi—baechu—kimchi; the following number indicates the fermentation time (in days). All values are presented as the mean ± SD of duplicate determination. Different letter indicates significant difference at p<0.05 using Duncan’s multiple range test.

https://doi.org/10.1371/journal.pone.0300249.t002

PCA analysis was performed through the SPSS statistical program using the measured amino acid content (Fig 3). In the PCA factor loading plot, aspartic acid, cysteine, histidine, threonine, proline, and valine were located in the negative direction based on PC1, while all other amino acids were located in the positive direction (Fig 3A). In the PCA score, each kimchi sample was divided by the fermented day (Fig 3B). On day 0 of fermentation, both JBK and BK were located in the negative direction, based on PC1 (Fig 3B). However, as fermentation progressed, both JBK and BK migrated in the positive direction, based on PC1. In particular, on day 20, JBK is in the positive direction, based on PC1; it is located in the first quadrant, which is the positive direction based on PC2; and the rich content of amino acids can be visually predicted from the location of this point. As can be seen in Table 2, these results show that on day 20, the content is high, except for the six amino acids: aspartic acid, cysteine, histidine, threonine, proline, and valine, out of 17 amino acids. It was visually confirmed that aspartic acid and glutamic acid of the JBK were also higher in content than in the BK. These results suggest that jogi addition has a greater impact on amino acid content than on color, reducing sugar, organic acid, and microbial community. In particular, the results clearly show that in the early stages of fermentation, there was a significant difference in the content of savory components, such as glutamic acid and aspartic acid.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Principal component analysis based on amino acid production.

BK, baechu—kimchi; JBK, jogi—baechu—kimchi; numbers beside symbols indicate fermentation time (in days). The curved arrow represents the path of amino acid change with fermentation time.

https://doi.org/10.1371/journal.pone.0300249.g003

Correlation of amino acids and microbial communities

In summary, based on the results so far, the jogi addition affects the change in amino acid content from the beginning of fermentation. In addition, although there is no significant difference in the production of organic acids, it was judged that the production of bacterial community and organic acids derived by the culture-dependent method was more correlated than the results of the non-culture-dependent method. Therefore, the results were derived by analyzing the correlation between bacterial community analyzed by the culture-dependent method and the amino acids with a heat map (Fig 4). Most species have been found to exhibit some correlation with amino acids (Fig 4). However, 6 out of the 17 amino acids, namely alanine, glycine, isoleucine, phenylalanine, tyrosine and valine, were not correlated with microorganisms. Leu. mesenteroides, which accounted for 53% of the total microbial flora, formed a positive correlation between arginine and proline. B. subtilis, which accounted for 17.8%, showed a positive correlation with cysteine, glutamic acid, histidine, and threonine. Unusually, in the case of the genus Weissella, there was no correlation with amino acid production. The addition of jogi in the early fermentation period differed in the content of aspartic acid and glutamic acid, and the microorganisms that showed a positive correlation were B. amyloliquefaciens, B. subtilis, B. velezensis, B. aerophilus, and K. grimontii, all of which at the beginning of fermentation are species that dominate, and then decrease. Interestingly, the amino acids that are positively correlated with the genus Bacillus encoded the bio-synthetic genes for production of those amino acids via Bacillus genome analysis (S2 Fig) [25], whereas LAB, including Leu. mesenteroides, did not encode the synthetic genes (S2 Fig) [26]. This mainly suggests that during fermentation, amino acids were degraded by proteolytic enzymes from protein. Bacillus species and K. grimontii, which are positively correlated with a number of amino acids, possessed more protease genes, which degrade protein, than genomes of Leu. mesenteroides and Leu. lactis (S1 Table). Therefore, we assumed that the effect on the amino acid content is different, depending on the enzyme produced by the microorganism. Consequently, these results assumed that the amino acids contents were affected by bacteria biosynthesis, and that protease activity during fermentation could also have an effect.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Correlation of amino acid and microbial community in kimchi.

Only significant correlations (p < 0.05) are shown with a heat map; blue (positive) and red (negative), using Pearson correlation coefficient between the bacterial community at species level, and the amino acid contents of kimchi.

https://doi.org/10.1371/journal.pone.0300249.g004

It was expected that the addition of jogi to kimchi would increase the protein content and the content of bacteria that break down the protein, but contrary to our expectations, previous studies did not show much difference in microbial flora from the control groups without jogi addition [14]. In the current experiments, the results of confirming changes in physical and chemical components through jogi addition did not confirm dramatic changes in color, reducing sugar, and organic acid components (Figs 1 and 2, Table 1). However, in the case of amino acids, at the beginning of fermentation, aspartic acid and glutamic acid related to umami ingredients showed a significant difference in the jogi addition group, while on 20^th day of fermentation, most amino acids, including these two amino acids, were higher than in the control group (Table 2). Although the paper could not confirm that in the early stages of fermentation, the addition of seafood made excellent nutritional or sensual differences, word of mouth or universally known content recommends the consumption of seafood-added kimchi as quickly as possible, as kimchi immediately after manufacture is delicious. This general knowledge is probably understood to be due to the significant difference in taste when fish is added at the beginning of the fermentation. It is not just the difference in taste that encourages consumption within a short period of time, but it may also have included concerns that degrading enzymes from fish cause rapid fermentation, and shorten the period of shelf life. The results of this experiment can be meaningful, in so far as it is a scientifically proven fact that immediately after the preparation of kimchi, the jogi addition differs in the composition of the umami. However, these results are based on homemade kimchi and experiments conducted within 20 days after production. Therefore, it is suggested necessary to conduct experiments in the future to explore variations in factors such as the amount of jogi, extended fermentation periods, and fermentation temperatures.