I. Introduction
Kimchi is fermented by lactic acid bacteria (LAB) at low temperatures, ensuring proper ripening and preservation. For the manufacturing of commercial kimchi products, natural fermentation with unsterilized raw materials leads to the growth of various microorganisms, making it difficult to standardize the fermentation process. Natural fermentation often results in final products exhibiting inconsistent product quality (Lee et al. 2015;Lee et al. 2018). Therefore, the use of starter cultures is considered as an alternative to address these problems (Jang et al. 2015;Kim et al. 2016).
Leuconostoc mesenteroides strains are some of the most widely studied and commercialized strains in the kimchi manufacturing process. Although the inoculation of some starter strains can control kimchi fermentation, starter dominance might change depending on the fermentation conditions and based on various raw materials of kimchi. Consequently, a strain-specific identification system for the control and tracking of specific microorganisms (starter), including verification of the presence of the strain in starter kimchi, is a critical issue in terms of quality control and maintenance. The majority of bacteria in kimchi can be cultured (Kim and Chun 2005); therefore, agar-plate culturing is typically used as a culture-dependent method for identifying and monitoring bacterial flora in samples. However, this method (plate culture method) requires a prolonged incubation time, uses a selective medium, and has the disadvantage that culture conditions must be specified. Moreover, because the nutritional requirements and culture conditions of LAB strains are very similar, it is difficult to distinguish them from one another. Strain-specific identification was originally performed based on analysis of specific culture properties and has subsequently been achieved using DNA fingerprinting or enzyme-linked immunosorbent assays with monoclonal antibodies. However, these methods are time-consuming and require experienced technicians (Endo et al. 2012).
Polymerase chain reaction (PCR)-based methods for strainspecific detection include random amplified polymorphic DNA-PCR, PCR-restriction fragment-length polymorphism, and multiplex PCR, which are sensitive, selective, rapid, economical, and convenient (Juvonen et al. 2008;Maruo et al. 2006). Among these methods, multiplex PCR can be used to measure a large number of samples simultaneously, making it possible to distinguish between similar microorganisms (Kwon et al. 2005).
A variety of Leuconostoc spp. have been identified in kimchi samples using molecular methods (Choi et al. 2003;Kim et al. 2000). However, these methods are insufficient for detection of particular strains in kimchi. We have developed Leu. mesenteroides WiKim33, which is used for the kimchi fermentation control. In this study, a strain-specific primer set was evaluated based on the complete genome sequence of Leu. mesenteroides WiKim33 to provide future use in the kimchi industry.
II. Material and Methods
1. Leu. mesenteroides WiKim33-specific primer design
The complete genome sequence was obtained using a PacBio RSII instrument (Pacific Biosciences, Menlo Park, CA, USA). A total of 189,111 reads (mean read length: 11,305 bp) were obtained, and four contigs were formed (1 chromosomal DNA and 3 plasmid DNA). Four pairs of primer sets were designed based on the plasmid DNA sequence of strain WiKim33 using the National Center for Biotechnology Information Primer-Blast Tool (http://www. ncbi.nlm.nih.gov/tools/primer-blast/), and Macrogen, Ltd. (Seoul, Korea) synthesized all primers used in this study. The primer sequences are shown in Table 2 along with their corresponding GenBank accession numbers (CP021492- CP021494) and predicted product sizes.
2. DNA extraction and amplification by multiplex PCR
The kimchi samples were ground and filtered twice through sterilized gauze, and genomic DNA was extracted from the kimchi soup using a genomic DNA prep kit for bacteria (Qiagen, Valencia, CA, USA) according to manufacturer instructions. The 20-μL reaction volume contained 10 ng genomic DNA, 1×PCR buffer, dNTP mixture (2.5 mM each), and 0.5 U of AccuPower Multiplex PCR DNA polymerase (Bioneer, Daejeon, Korea). All multiplex PCRs were performed in a thermal cycler (Bio-Rad, Hercules, CA, USA) with the following protocol: pre-incubation at 95°C for 10 min, followed by 35 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for 1 min, with a final elongation at 72°C for 5 min. PCR products were analyzed by agarose gel electrophoresis, which was performed on 1.2% (w/v) UltraPure Agarose gels (Invitrogen, Carlsbad, CA, USA) containing 0.1 μg/mL ethidium bromide. A 100-bp DNA ladder (Bioneer) was used as a size marker.
3. Examination of sensitivity and specificity
The sensitivity of multiplex PCR was tested using a serial dilution of Leu. mesenteroides WiKim33 prepared from overnight cultures in MRS agar. The bacteria were serially diluted in phosphate-buffered saline (PBS) or kimchi filtrate (early fermentation; pH 5.80; acidity: 0.32%). DNA was isolated from 1 mL of each dilution and subsequently tested using multiplex PCR. Specificity tests were performed by colony PCR with 45 Leuconostoc spp. and 30 other strains.
4. Preparation of kimchi
Vegetables, including cabbage, and other seasonings were obtained from local grocery stores. Kimchi samples were categorized depending on the addition of Leu. mesenteroides WiKim33 (1×107 CFU/g kimchi) as a starter culture, and each sample was labeled as CTL (kimchi without starter) or S (kimchi with starter). The temperature during the entire fermentation process was maintained at 2.0±0.5, 0.0±0.5, 6.0±0.5, or 10.0±0.5°C for 84 days.
5. pH and acidity
Kimchi samples (500 g) were blended using a hand blender (HR1372/90; Philips, Amsterdam, Netherlands) and filtered through sterilized gauze before their chemical characteristics were analyzed. The pH values of the kimchi filtrates were determined using a pH meter (Orion 3-Star; Thermo Fisher Scientific, Waltham, MA, USA). The kimchi filtrate was then titrated with 0.1 N NaOH to pH 8.3 to determine the titratable total acidity, which was estimated in previous study (Lee et al. 2018).
6. Total viable bacteria, LAB counts, and WiKim33 population ratio in kimchi samples
Microorganisms in the kimchi liquids were counted according to a standard plate-count procedure and the microbial counts in the kimchi samples were determined as the number of CFUs per milliliter. Total viable bacteria, and LAB enumeration were carried out as described previously (Lee et al. 2017). The colonies formed on the agar plates were applied to a WiKim33-specific detection system to count WiKim33 colonies. The ratio of WiKim33 cells to total viable bacterial is shown as the starter domination value.
III. Results and Discussion
1. Design of the multiplex primer set for Leu. mesenteroides WiKim33 detection
Multiplex PCR primers (4 sets) for Leu. mesenteroides WiKim33 were designed from plasmid base sequences. The strain-specific primer sets yielded four amplicons at each of the expected sizes (585, 703, 884, and 1217 bp) <Figure 1>.
Generally, Multiplex PCR targets the chromosomal 16S rRNA gene (Grahn et al. 2003). However, in the case of LAB, the homology between chromosomal DNA sequences among species is so high that it is difficult to identify a specific base sequence capable of distinguishing a specific strain (Collins et al. 1991). Plasmids are DNA sequences that can be propagated independently from the chromosome in a bacterial cell (Cohen et al. 1973;Tenover et al. 1995). Although plasmids are more unstable than chromosomes, passage stability of the plasmid was confirmed by PCR with the primer set after 70 consecutive subcultures, and four strain-specific amplicons were clearly detected (data not shown).
The specificity of the designed primer set was validated against 45 Leuconostoc spp. and 30 other LAB strains. Although seven Leuconostoc spp. produced single amplicons in the PCR product of 884 bp or 585 bp, none of the strains produced four amplicons simultaneously. Moreover, 30 non- Leuconostoc spp. strains did not produce any amplicons <Table 1 and Supplementary Figure S1>. These results indicated that the designed primers were highly specific for the target species.
2. Detection limit of the designed primer set
The detection limit of the WiKim33-specific PCR method was measured using serial dilutions in PBS or kimchi filtrate. Leu. mesenteroides WiKim33 (undiluted solution: 1.8×109 CFU/mL) was serially diluted 10-fold, and each dilution was subjected to multiplex PCR to determine whether the four amplicons were present. As shown in <Figure 2>, strain WiKim33 could be detected clearly after 0- to 5-fold dilutions (with PBS) or 4- to 5-fold dilutions (with kimchi filtrate). The sensitivity of the detection primer set for the WiKim33 strain was estimated at 1.8×104 CFU/mL (2.77 pg DNA) for the PBS-diluted solution and 1.8× 105 CFU/mL for the diluted kimchi filtrate. Therefore, we assumed that strain WiKim33 was present at more than the detection limit when applied to kimchi. Table 2
3. Application of WiKim33-specific PCR for sensing starter-inoculated kimchi
LAB successions can differ depending on fermentation temperature, despite the use of the same raw materials and preparation methods (Jung et al. 2014). Moreover, kimchi is generally prepared at room temperature, immediately cooled to below 8°C, and then distributed. Fermentation occurs during distribution or the period in which the food is kept under refrigeration at the market (4°C) (Eom et al. 2007), and consumers can store kimchi at low temperatures (2 or 0°C) after purchase. Therefore, it is necessary to accurately detect only the starter strain in various microbial-community structures according to various temperature conditions.
To evaluate the possibility of applying the strain-specific detection system in the field, kimchi was manufactured and stored at different temperatures. Specific PCR was performed for kimchi, including the WiKim33 strain, and products were stored at 2, 0, 6, and 10°C for 8 weeks. The physicochemical (pH and acidity) microbial enumeration (total viable bacteria and LAB) showed different fermentation aspects according to fermentation temperature <Supplementary Figure S2>. The starter WiKim33 strain-population ratios varied according to temperature conditions, but more than 105 CFU/ mL was detected during the entire fermentation period <Figure 3A>. Moreover, in Figure 3A, which shows the number of WiKim33 strain against the total viable of bacteria, the population ratio of WiKim33 strain was gradually decreased as the fermentation progresses. Our multiplex PCR system did not show amplicons for kimchi without the starter in each temperature condition (control kimchi) <Figure 3B>; however, for kimchi supplemented with the WiKim33 starter strain, the four different PCR amplicons were clearly detected at different fermentationstorage temperatures during the entire fermentation period despite the complex microbial composition <Figure 3B>. Although, species specific detection methods have been reported in food products (Sheng et al. 2018), the specialty of our assay was strain-specific and was sufficiently sensitive to kimchi products.
In addition, when this detection system is applied to the plate culture, precise quantitative analysis of living WiKim33 strain actually in kimchi is possible and it is useful to monitor the inoculated starter strain by calculating the dominance rate for the total viable bacteria. Supplementary Table S1
IV. Summary and Conclusion
In this study, we developed a strain-specific primer set based on the complete genome sequence of Leu. mesenteroides WiKim33 for applications in the kimchi industry. Our Multiplex PCR detection system allows rapid identification of and high specificity for only specific strain in kimchi in a single reaction. This primer set could be used to detect the WiKim33 strain specifically in kimchi samples, establishing a novel tool for quality management through monitoring of the starter. The results of this study suggested that future studies are required to quantify populations of the WiKim33 strain in complex microbiota compositions.
