Technical Paper Microbial Community Characteristics in Industrial Matured

Home , Pediococcus claussenii, Synergistetes

http://www.jsfst.or.jp

Technical paper

Microbial Community Characteristics in Industrial Matured Chinese paocai, a Fermented Vegetable Food, from Different Factories

1 1 1 1 1 1,2* Huipeng Liang , An Zhang , Zhengyun Wu , Shupin Cheng , Wenping Yu and Wenxue Zhang

1College of Light Industry, Textile and Food Engineering, Sichuan University, Chengdu, 610065, China 2School of Liquor-Making Engineering, Sichuan University Jinjiang College, Meishan, 620860, China

Received March 11, 2016 ; Accepted June 6, 2016

Paocai was a traditional fermented food in China and produced by spontaneous fermentation of various vegetables. The microbial community occurring in industrial matured Chinese paocai (IMCP) from different factories was investigated by PCR-denaturing gradient gel electrophoresis (DGGE) and quantitative PCR (qPCR). The dominant bacteria were different in IMCP from different factories. Three genera including Debaryomyces, Pichia and Candida were identified and Debaryomyces was mutual and dominant yeast in all IMCP. Fifteen bands from DGGE profiles of lactic acid bacteria (LAB) were identified and four genera were detected. Alkalibacterium and Lactobacillus were dominant in all IMCP. In conclusion, the IMCP from different factories possessed different dominant bacteria. The community of LAB in IMCP from different factories was different, and Lactobacillus and Alkalibacterium were main LABs in all IMCP. The quantity of LAB was quantitated and shown to exist at 107 _ 109 copies in the IMCP from different factories.

Keywords: Chinese paocai, DGGE, qPCR, microbial community, lactic acid bacteria

Introduction the yield and quality of paocai, it is necessary to understand fully Fermented vegetables are usually produced by spontaneous the microbial community in various paocai. As far as we know, the fermentation of the raw materials with various naturally-occurring conventional culture method is often time-consuming and bacteria in brine water (Tanganurat et al., 2009). Paocai is one of laboursome. In recent years, plenty of powerful molecular the typical representatives of traditional fermented vegetable foods ecological methods, such as PCR-DGGE, qPCR, Amplified in China, especially in Sichuan province. It was recorded that ribosomal DNA restriction analysis (ARDRA), fluorescent in situ Chinese made paocai as early as more than 3000 years ago in Zhou hybridization (FISH), etc., have been widely used to analyze the Dynasty (Xiong et al., 2012). Paocai, commonly served as side microbial community in the field of food (Yeun et al., 2014; Iwobi dish or used as an appetizer, is a popular food items prepared as a et al., 2015; Luo et al., 2014; Ding et al., 2014), which could result of spontaneous fermentation and in recent years it has been overcome the shortcomings of culture approaches and detect the recognized as a healthy functional food because of its healthcare uncultured microbes as well. PCR-DGGE, a classical molecular function such as antibiosis, anticancer, anti-senescence, anti- ecological technic, has been widely applied to directly reveal and obesity and so on (Li et al., 2005; Yan et al., 2008). Nowadays rapidly monitor the microbial community in the process of there is an increasing consumer demand for various paocai on fermentation (Nie et al., 2015, 2013). Yeun et al. (2014) identified account of its convenience, tasty, nutrition and health. To improve the most prevalent LAB in salted Chinese cabbage samples by

*To whom correspondence should be addressed. E-mail: [email protected] 596 H. Liang et al. using PCR-DGGE and SDS-PAGE. Yeun et al. (2013) analyzed GCG GGC GGG GCG GGG GCA CGG GGG GAG CAG TAG the changes in the bacterial microflora of two commercial kimchi, GGA ATC TTC CA-3’) and LacR (5’-ATT YCA CCG CTA CAC salted cabbage, and ingredient mix samples during 30 days of ATG-3’) targeting the region of 16S rRNA gene (Walter et al., fermentation at 4℃ and 10℃ by using PCR-DGGE. And for all we 2001). All the primers were synthesized by the Sangon Biotech know, LABs are important in many food fermentations because Company (Shanghai, China). The PCR reaction contained 50 μL of they contribute to sensory characteristics and preservative effects 2×PCR Mix (TIANGEN Biotech, Beijing, China), 20 pmol (Holzapfel, 1995). During the fermentation, LAB utilise primers, template DNA and distilled water. The PCR program for carbohydrate substrates available in the fermentation system and bacteria was 95℃ for 5 min, followed by 30 cycles of 95℃ for 60 s, produce organic acids, especially lactic acid, which not only play annealing at 67℃ for 30 s and elongation at 72℃ for 60 s, and an important role in the taste and aroma of the product but also finally a 5 min elongation step at 72℃. The PCR of 18S rRNA lower the product’s pH to ensure quality and safety (Zhou et al., gene was initiated at 94℃ for 5 min, followed by 30 cycles of 2014). As a complementary method, qPCR technology which has denaturation at 94℃ for 60 s, annealing at 53℃ for 60 s and been used to quantitatively determine microbe in various extension at 72℃ for 60 s, and a final extension at 72℃ for 5 min. environments (Mokhtari et al., 2013; Kao et al., 2007) was used to PCR amplification of LAB-group was performed as follows: 94℃ quantitate the quantity of LAB in this study. for 2min, followed by 35 cycles of 94℃ for 30 s, annealing at 61℃ The major objective of this study, therefore, was to investigate for 60 s and elongation at 68℃ for 60 s, and finally a 7 min the characteristic and the differences of the microbial community elongation step at 68℃. The PCR products were checked by structures in IMCP from different factories. For the best of our electrophoresis on a 2% agarose gel. knowledge, this is the first report to reveal the diversity and (3) PCR-DGGE analysis of microbial communities The PCR differences of microbial community in traditional IMCP by using products were analyzed by DGGE using the D-CodeTM Universal the combined PCR-DGGE and qPCR methods. Mutation Detection System (Bio-Rad, Hercules, CA, USA). DGGE was performed with 80% polyacrylamide gel (Acrylamide/ Materials and Methods Bisacrylamide, 37.5:1) with a linear gradient of 30 _ 55% (1) Sampling and chemical analysis Samples of IMCP were denaturant for the bacterial community, 20 _ 40% for the fungal collected from three paocai Condiment Company limited (C1, C2 community and 25 _ 45 for LAB-group (a 100% denaturant and C3), located in Meishan, Sichuan province of China, in March corresponds to 7 M urea and 40% [v/v] formamide). 2014. The varieties of paocai samples included Qingcai (P1, Electrophoresis was run for 4 h at 200 V for bacteria, 5h at 160 V Brassica chinensis var chinensis) and Zhacai (P2, Brassica juncea for fungi and 4.5 h at 130 V for the LAB at 60℃ in 1×TAE buffer var.tumida). The simples were stored at _20℃ for further analysis. respectively. After electrophoresis, the gel was stained with SYBR The pH value of the samples from different factories was measured Green I and visualized using a Gel DocTM XR (Bio-Rad, USA). with a pH meter (PHS-3C, China). Total titratable acidity (TTA) Bands patterns of the DGGE profiles were analyzed by Quantity was titrated using 0.1 N NaOH to final pH 8.2 with phenolphthalein One software (Bio-Rad, USA). The band Richness (S), Shannon as the indicator and expressed in lactic acid. Salinity was measured (H) and Pielou (E) index were determined based on the number and by a digital salt meter (ATAGO, Japan). All analyses were relative quantity of the DGGE bands (Shannon, 1949; Pielou, conducted in duplicate and the average values are presented. 1966). The S showed that the number of bands in DGGE profiles. (2) DNA extraction and PCR amplification The total DNA The H gives the proportional abundance of species and reacts was extracted using the Ezup Column Genomic DNA Purification sensitively to rare species and when all species are represented by Kit (Sangon Biotech, China) and stored at _20℃ until used. the same number of individuals, the H reaches its maximum value. The PCR was carried out in a MyCycler™ Thermal (Bio Rad, The E was presented to describe the uniformity of the distribution USA). The primers of WBAC1 (5’-GTC GTC AGC TCG TGT of the individuals on the number of bands. Principal component CGT GAG A-3’) and WBAC2-GC (5’-CGC CCG CCG CGC analysis (PCA) was performed using Canoco for Windows v4.5 GCG GCG GGC GGG GCG GGG GCA CGG GGG GCC CGG software (Wageningen UR, Netherlands). GAA CGT ATT CAC CGC G-3’) were used to amplify the (4) Excision of representative DGGE bands and sequencing bacterial 16S rRNA gene (Lopez et al., 2003). The GC clamp Bands of interest from DGGE profiles were excised. The DNA sequence attached to primer was underlined. For analysis of fungal eluted from excised bands was amplified using the primers diversity, PCR amplification of the 18S rRNA gene was performed mentioned above with no GC-clamp and the amplified products using primers NS3-GC (5’-CGC CCG CCG CGC GCG GCG GGC were sent to company for cloning and sequencing (Sangon, GGG GCG GGG GCA CGG GGG GGC AAG TCT GGT GCC Shanghai, China). The sequence information was acquired by AGC AGC C-3’) and YM951r (5’-TTG GCA AAT GCT TTC GC- aligning the results with the sequences in Gen Bank using the 3’) (Haruta et al. 2006). LAB group-specific PCR was performed BLAST search program at the National Center for Biotechnology using the primers of LacF-GC (5’-CGC CCG CCG CGC GCG Information (NCBI) (i) and the Classifier search program at the Microbial Community Differences in Industrial Matured Chinese Paocai 597

Table 1. The diversity indices of microbial community in IMCP samples from different factories.

Bacteria Fungi LAB-group Salinity Samples* pH TTA (NaCl%) Richness Shannon Pielou Richness Shannon Pielou Richness Shannon Pielou (S) (H) (E) (S) (H) (E) (S) (H) (E) P1 C1 4.14 0.20 4.60 14 2.17 0.82 6 1.20 0.67 12 1.77 0.77 C2 4.04 0.34 8.00 16 1.96 0.71 5 0.69 0.43 9 1.51 0.69 C3 4.23 0.30 6.70 22 2.32 0.75 6 1.38 0.77 10 1.83 0.79 P2 C1 4.01 0.17 3.60 19 2.41 0.82 6 1.67 0.93 10 1.63 0.71 C2 3.67 0.67 8.80 15 1.78 0.66 7 1.53 0.79 10 1.99 0.86 C3 4.00 0.40 6.50 16 2.16 0.78 7 1.33 0.68 11 2.01 0.84

*P1 and P2 denoted two kinds of vegetables samples. C1, C2 and C3 represented the samples collected from three different factories respectively.

Ribosomal Database Project (RDP) (ii). 2012; Liu et al., 2015). The pH values of P1 were higher than that (5) qPCR analysis of LAB in IMCP samples The total quantity of P2. The pH values were 3.67 _ 4.23 for all IMCP samples. The of LAB was quantitated by employing the qPCR with SYBR Green sample C2 of P1 and P2 had the lowest pH (4.04 and 3.67) and the I. The primers of LacF and LacR mentioned above with no GC- highest TTA (0.34 and 0.67) and salinity (8.00 and 8.80) (Table 1). clamp were used to perform the qPCR in a Light cycler Nano The pH values of IMCP samples from different factories were System (Rocha, Switzerland) (Walter et al., 2001). The different and this was potentially connected with their differences amplification reactions were carried out in a total volume of 20 μL in microbial community. containing SYBR® Green PCR Real time PCR Master mix (2) Diversity of microbial community in IMCP samples The (Toyobo, Japan), 2 pmol each primer, DNA template and distilled microbial community DGGE profile of IMCP from different water. According to the protocol described by Walter, a modified factories was shown in Fig. 1. A total of 27 bands were observed form of amplification program was as follows: 50℃ for 2 min, within all the samples from bacterial DGGE profile and the S of initial denaturation at 95℃ for 10 min, followed by 40 cycles of each lane were shown in Table 1. The S of C3 and C1 were the denaturation at 95℃ for 15 s, annealing at 60℃ for 30 s, and highest and lowest in P1 samples. While the S of C1 and C2 was extension at 72℃ for 15 s, with collection of fluorescence signal at the highest and lowest in P2 samples. Based on DGGE profile of the end of each cycle. bacterial community, the ranges of H in P1 and P2 samples were The melting curve analysis was carried out after amplification 1.96 _ 2.32 and 1.78 _ 2.41 respectively (Table 1). The H of C2 by heating gradually from 60℃ to 95℃ at 0.1℃/s with continuous was the lowest in P1 and P2 samples. This may be due to their fluorescence monitoring. The standard curve, evaluated by lowest pH and highest salinity. In fungal DGGE profile, the S in P1 correlation coefficient (R2), was constructed using plasmids, which and P2 samples are comparable and the ranges of H were were prepared from the 16S rRNA gene library using Mini Plasmid 0.69 _ 1.38 and 1.33 _ 1.67 respectively (Table 1). The microbial Kit (Tiangen, Beijing, China). All samples were performed in groups shifted sequentially from less acid and salt tolerant to more triplicate. acid and salt tolerant groups adapted to the acidic and salt environmental condition (Fierer and Jackson, 2006; Plengvidhya et Results and Discussion al., 2007). Therefore, the relatively low pH of the samples may In China, paocai is a traditional fermented food which drives predict a low microbial diversity. And the bacterial and fungal H of by spontaneous fermentation microorganism. Although many C3 (2.32 and 1.38) and C1 (2.41 and 1.67) were the highest in P1 studies have revealed that a variety of microbial species contribute and P2 samples respectively. The E of P1 samples showed no to the production of kimchi products (Yeun et al., 2013; Jeong et significant differences with those of P2 samples (Table 1). al., 2013; Park et al., 2012), little is known about the microbial (3) Microbial community analysis of IMCP samples In order communities in IMCP samples made in southwest China. to understand the dominant microbe in IMCP samples from (1) Physico-chemical analyses of IMCP samples The pH, different factories, a total of 27 representative bands of bacterial which is considered as fundamental variables determining the PCR-DGGE indicated in Fig. 1 were sequenced and the results success of vegetable fermentation, is an important indicator for were shown in Table 2. All the bands were identified and fell into determining the maturity of paocai (Penas et al., 2010). The pH, four phyla including Firmicutes (15 bands), Proteobacteria TTA and salinity of IMCP brine samples from different factories (10 bands), Synergistetes (1 band) and Chloroflexi (1 band). The were shown in Table 1. The pH values of IMCP in our similarity of all band sequences was nearly ≥ 99% comparing with investigation were in a common range and the pH profile was those available in the GenBank database. After the fermentation, similar with those researches on fermented vegetables (Yu et al., IMCP samples of different kinds or from different factories 598 H. Liang et al.

have been reported to produce antimicrobial compounds and exhibited antimicrobial activity against several pathogens (Romanenko et al., 2008).The genus Alkalibacterium dominated in C2 of P1 samples especially comprises marine LAB and they produce lactic acid as the main product of glucose fermentation (Ishikawa et al., 2011, 2013). Tetragenococcus halophilus dominated in C2 of P1 samples is a moderately halophilic Gram- positive LAB and has been widely found in salted products (Fukuda et al., 2002; Taira et al. 2007). Pediococcus dominated in C2 of P2 samples were LAB found on plants and in many fermented foods (Dobson et al., 2002; Simpson et al., 2002). The existence of Alkalibacterium, Tetragenococcus and Pediococcus maybe lead to a relative low pH in C2 of P1 and P2 samples (Table 1). For fungal PCR-DGGE (Fig. 2), a total of 9 bands were excised and sequenced. The results from the sequencing of the highlighted bands are shown in Table 2. All sequenced bands were affiliated with three genera, including Debaryomyces (bands 1, 4 and 5), Pichia (bands 2 and 3) and Candida (bands 6, 7, 8 and 9). Fig. 1. The PCR-DGGE profile of 16S rRNA gene extracted Obviously, PCR-DGGE profile analyses showed that from bacterial community in IMCP samples from different Debaryomyces predominated in all IMCP samples. Debaryomyces factories. P1 and P2 denoted two kinds of vegetables samples. C1, C2 and C3 represented the samples collected from three hansenii reportedly produces vitamins, amino acid and so on and different factories respectively. The bands indicated by the contributes to the development of flavor in cheese and fermented arrows and numbers were excised and sequenced, and the sausages (Breuer et al. 2006; Cano-García et al. 2014). Some alignment results were listed in Table 2. strains, pertaining to the species D. hansenii, have the ability to produce aroma compounds, such as aldehydes, ketones, alcohols, possessed the different dominated bacteria. In samples P1, the esters and sulphur compounds (Cano-García et al. 2014). dominant bacteria of C1 sample were Brevundimonas (bands 12), Inferentially, D. hansenii is also involved in the development of Clostridium (band 13) and Sphingomonas (band 18). While in C2 flavor in IMCP. sample Alkalibacterium (bands 5 and 6), Brevundimonas and (4) Multivariate analysis of DGGE profiles PCA was Tetragenococcus (band 15) were prevalent. Uncultured bacterium performed based on the relative quantity of microbial DGGE band (band 9, 10 and 16) and Tetragenococcus (bands 11) were profile to analyze the correlation between microorganisms and the dominant in C3 sample. In samples P2, Uncultured bacterium (band IMCP samples from different factories. The PCA plots were shown 9, 12, 16 and 18), Clostridium and Lactobacillus (band 14) were in Fig. 3. The principal components PC1 and PC2 of IMCP dominant bacteria in C1 sample. In C2 sample of P2, Pediococcus samples accounted for 50.30% and 24.90% of the variance. The (band 7) and uncultured bacterium (band 12) were preponderant. PCA ordination of the genera and sample variables demonstrated While in C3 sample, uncultured bacterium (band 12) and that Brevundimonas, Alcanivorax and Pediococcus were correlated Clostridium were prevalent. Among these genera Lactobacillus is with C1 of P1, C2 and C3 of P2 samples (Fig. 3). Alkalibacterium regarded as one of main bacteria that produce plenty of lactic acid and Tetragenococcus were correlated with C2 of P1 samples while and one of dominant LAB in vegetable fermentations (Xiong et al., Pediococcus was correlated with C2 of P2 samples. This may be 2013). Some researchers have demonstrated that Lactobacillus is due to the higher salinity of C2 samples. Halovibrio, unclassified_ contributed to the depletion of nitrite accumulated during Rhodobacteraceae, Acinetobacter and Candida were correlated fermentation in paocai (Yan et al., 2008; Rai et al., 2010). Then with C3 of P1 samples, and Lactobacillus and Alkalibacterium Lactobacillus is not only beneficial to the taste and aroma but also were correlated with C3 of P2 samples (Fig. 3). to ensure the safety of the IMCP. Brevundimonas detected in all Plants provide a nutrient-rich niche for the growth of samples was found in various environments but its functions on the microorganisms, particularly bacteria (Pereira et al., 2012). The fermentation need further researches (Ryu et al., 2007). bacteria which attached to the raw vegetables present significant Clostridium which could metabolize organic substances to organic differences associated with regions, seasons, vegetable varieties as acids such as caproic acid, butylic acid, etc. (Zhang et al., 2005), well as cultivation patterns, resulting in the differences of was possible involved in the flavor components in paocai. microflora in different fermented vegetable products (Xiong et al., Sphingomonas dominated in C1 of P1 and P2 sample especially 2012; Islam et al., 2010; Zhang et al., 2013; Haque et al., 2015). Microbial Community Differences in Industrial Matured Chinese Paocai 599

Table 2. The identities of 16S and 18S rRNA gene sequences of bands excised from DGGE gel by using the BLAST and RDP search tools.

Identity Alignment Banda Closest sequence/microorganismb Phylogenetic affiliation Accession No. (%) length (bp) Bacteria 1 Clostridium butyricum strain CB8 Clostridium 99 327 KJ558433.1 2 Lactobacillus zymae strain KCC-14 Lactobacillus 99 328 KC625331.1 3 Uncultured Lactobacillaceae bacterium clone PC-B50 Lactobacillus 99 328 JQ809294.1 4 Uncultured Pediococcus sp. clone PC-B14 Pediococcus 99 328 JQ809283.1 5 Alkalibacterium kapii strain MGR70 Alkalibacterium 97 327 KF151857.1 6 Alkalibacterium gilvum strain 5AE-1 Alkalibacterium 100 328 AB690572.1 7 Pediococcus ethanolidurans strain lao3-6-1 Pediococcus 100 243 KJ690914.1 8 Lactobacillus hammesii strain Kw2S11L1 Lactobacillus 99 329 JF427720.1 9 Uncultured bacterium clone Ll142-2L4 Acinetobacter 100 330 FJ671990.1 10,16 Uncultured bacterium clone Woods-Hole_a1567 Rhodobacteraceae 99 307, 302 KF798445.1 11 Tetragenococcus halophilus strain JCM 20256 Tetragenococcus 99 329 AB911552.1 12 Uncultured bacterium clone TV_B8 Brevundimonas 100 289 JX575614.1 13 Clostridium sp. S11-3-10 Clostridium 99 329 AB838978.1 14 Lactobacillus sp. NBRC 101665 Lactobacillus 99 328 AB681518.1 15 Tetragenococcus halophilus strain 7-8 Tetragenococcus 98 329 HQ384302.1 17 Lactobacillus sp. B4(2014) Lactobacillus 100 328 KM259929.1 18 Uncultured bacterium isolate DGGE gel band A28 Sphingomonas 100 328 KC736689.1 19 Idiomarina loihiensis strain DY032-2 Idiomarina 98 330 KJ466005.1 20,21 Lactobacillus farciminis strain Y124D Lactobacillus 98 328, 329 AB889729.1 22,26 Uncultured bacterium clone C184 Alcanivorax 99 330 KC523578.1 23 Pseudomonas halophila isolate P-halo Halovibrio 98 329 HG964480.1 24 Brevundimonas sp. BBDP1026 Brevundimonas 98 306 EF471236.1 25 Uncultured bacterium DGGE gel band L18 Aminobacterium 99 331 AB856317.1 27 Uncultured Chloroflexi bacterium clone SL73 Sphaerobacter 98 316 JX240517.1 Fungi 1,4 Debaryomyces hansenii isolate LJ-2-Y-1 Saccharomycetaceae 100 402 KJ781291.1 2 Pichia triangularis strain BC 211 Saccharomycetaceae 100 400 AY227018.1 3 Pichia farinosa strain CO-2 Saccharomycetaceae 99 402 EU106163.1 5 Uncultured Debaryomyces clone: ZQC98 Saccharomycetaceae 100 401 AB986207.1 6,7,8,9 Uncultured Candida clone: ZLB6 Saccharomycetaceae 99 389-390 AB986197.1 LAB 1,13 Lactobacillus ginsenosidimutans strain EMML 3041 Lactobacillus 99,100 347 HQ389549.1 2,4,5,7, Uncultured Lactobacillaceae bacterium clone PC-B50 Lactobacillus 99,100 347 JQ809294.1 11,12 3 Uncultured Lactobacillus sp. clone: 2X7 Lactobacillus 99 347 LC002932.1 6 Pediococcus ethanolidurans strain MF14 Pediococcus 100 347 KJ994507.1 8 Leuconostoc sp. P Leuconostoc 99 347 DQ061074.1 9 Lactobacillus alimentarius gene strain: JCM 1095 Lactobacillus 99 347 LC055606.1 10 Alkalibacterium gilvum gene strain: 5AE-1 Alkalibacterium 99 347 AB690572.1 14,15 Lactobacillus sp. NBRC 107199 Lactobacillus 96 347 AB682498.1

Only highest homology matches are presented. a the number of bands was in accordance with that in Fig. 1 and Fig. 3. b sequences were compared with known sequences in NCBI database.

Potentially, this is the main reason for the difference of bacteria in (5) PCR-DGGE analysis of LAB in IMCP samples The DGGE different varieties IMCP samples from different factories. In pattern of LAB was showed in Fig. 4, and a total of 15 bands were addition, another reason for the difference of microflora appeared observed. The S of C1 and C2 in P1 samples was the highest and among various fermented vegetable products were probably related lowest, while the S in P2 samples was comparative (Table 1). The to chemical and physical factors including substrates, NaCl H of C3 (1.83 and 2.01) was the highest in P1 and P2 samples concentration and fermentation temperature (Xiong et al., 2012; respectively. In addition, the E range of LAB in IMCP samples Cho et al., 2006). community was from 0.69 to 0.86. 600 H. Liang et al.

Fig. 2. The PCR-DGGE profile of 18S rRNA gene extracted from fungal community in IMCP samples from different factories. P1 and P2 denoted two kinds of vegetables samples. Fig. 4. The PCR-DGGE profile of LAB-group community in C1, C2 and C3 represented the samples collected from three IMCP samples from different factories. P1 and P2 denoted two different factories respectively. The bands indicated by the kinds of vegetables samples. C1, C2 and C3 represented the arrows and numbers were excised and sequenced, and the samples collected from three different factories respectively. alignment results were listed in Table 2. The bands indicated by the arrows and numbers were excised and sequenced, and the alignment results were listed in Table 2.

LAB in IMCP. And Lactobacillus alimentarius was detected more abundant in P1 samples than that in P2 samples. Pediococcus sp. was found to be dominant LAB in mukeunji aged for at least 2 years (Hong et al., 2015). In addition, Leuconostoc was detected while it was not in bacterial DGGE profiles, and it was more abundant in P2 samples than that in P1 samples. Some species affiliated to Leuconostoc was detected in many fermented vegetables such as Chinese sauerkraut, leek and kimchi (Xiong et al., 2012; Hong et al., 2015; Wouters et al., 2013). Marshall (1987) has pointed out that Leuconostoc mesenteroides ssp. cremoris and Leuconostoc lactis are two important organisms for aroma production. Then Leuconostoc was possibly contributed to the Fig. 3. Principal component analysis (PCA) of microbial composition aroma of IMCP. in IMCP samples from different factories. P1 and P2 denoted two kinds of vegetables samples. C1, C2 and C3 represented the samples (6) Quantitative of LAB in IMCP samples from different collected from three different factories respectively. The directions factories LABs play an important role in IMCP fermentation from of the arrows indicate the relative loading on the first and second the results of DGGE analysis. The quantity of LAB was monitored principal components. using qPCR. Compared with the conventional quantitative method, the qPCR approach can quickly quantitate both cultivable and non- In Fig. 4, fifteen bands were sequenced and four genera cultivable microorganisms in samples. The amplification curves including Alkalibacterium, Lactobacillus, Pediococcus, were of sigmoidal form. Melting curves, as an additional quality Leuconostoc were identified (≥ 96%). Lactobacillus (band 7) and control step for the analysis of qPCR data to verify the specificity Alkalibacterium (band 10) were mutual and preponderant in all of amplified products, can distinguish false positive signals due to IMCP samples. Many previous studies have shown that LAB non-specific amplification or primer-dimers (Soares et al., 2013). including Leuconostoc, Lactobacillus, Weissella, Lactococcus, and The melting curves of LAB in IMCP samples only had one peak, Pediococcus, are key players responsible for kimchi fermentation indicating that the specificity of primers which were used was (Chang and Chang, 2010; Lee and Lee, 2010; Lee et al., 2010). good. No primer dimers or non-specific amplification products Alkalibacterium, Lactobacillus and Pediococcus were already were visible for all samples. The standard curve of LAB for P1 and detected in the bacterial DGGE profiles. But the genus P2 samples was as Eq. 1 and Eq. 2. Alkalibacterium was found for the first time to be a dominated Microbial Community Differences in Industrial Matured Chinese Paocai 601

_ Table 3. The quantity of LAB in IMCP samples from 1 Cq = 3.34 log10 (q) + 48.74 ······Eq. different factories by qPCR.

_ Samples* Concentration of LAB (copies/mL) Cq = 3.41 log10 (q) + 49.40 ······Eq. 2 P1 C1 (1.20 ± 0.02) × 109 2 The R (0.99 and 0.99) suggested a good correlation between C2 (2.73 ± 0.13) × 108 Cq and the logarithm of template concentrations. The efficiency of C3 (8.30 ± 0.39) × 107 amplification calculated according to the slope of the standard P2 C1 (1.78 ± 0.01) × 108 curve was 99.30% and 96.60%, which indicated an optimum PCR C2 (1.63 ± 0.00) × 108 9 efficiency. The quantity of LAB in IMCP samples from different C3 (1.10 ± 0.01) × 10 factories was shown in Table 3. The quantities of LAB were shown *P1 and P2 denoted two kinds of vegetables samples. to exist at 107 _ 109 copies in the IMCP samples from different C1, C2 and C3 represented the samples collected from three different factories respectively. factories. The results obtained by qPCR showed that the quantity of LAB in C1 and C3 was the highest in P1 and P2 samples respectively (Table 3). The quantity of LAB in IMCP of different Acknowledgment The research was financially supported by kinds or from different factories was different. The difference of grants from the National Science and Technology Project (NO. the quantity of LAB in P1 samples was more obvious than that in 2012BAD31B04) of the Ministry of Science and Technology of P2 samples (Table 3). The factors that affected the quantity of LAB the People’s Republic of China, and the Key Science and in IMCP included the vegetable varieties, naturally occurring Technology Project (NO. 2013NZ0055) of the Sichuan Science microbial populations in the raw materials as well as the and Technology Department. environmental conditions such as pH, temperature and salt concentration. Reference Breuer, U. and Harms, H. (2006). Debaryomyces hansenii – an Conclusion extremophilic yeast with biotechnological potential. Yeast, 23, 415-437. Combined PCR-DGGE and qPCR analysis may be a valuable Cano-García, L., Rivera-Jiménez, S., Belloch, C., and Flores, M. (2014). and mutual complementary way to elucidate the microbial Generation of aroma compounds in a fermented sausage meat model community in IMCP samples and this is the first report about the system by Debaryomyces hansenii strains. Food Chem., 151, 364-373. microbial community structure and diversity in IMCP from Chang, J.Y. and Chang, H.C. (2010). Improvements in the quality and shelf different factories by combined PCR-DGGE and qPCR analyses. life of kimchi by fermentation with the induced bacteriocin-producing The microbial diversity was different in IMCP of different kinds or strain, Leuconostoc citreum GJ7 as a starter. J. Food Sci., 75, 103-110. from different factories. The results of bacterial DGGE profiles Cho, J., Lee, D., Jeon, J., Kim, J., and Han, H. (2006). Microbial population showed that a total of 27 bands were sequenced and four phyla, dynamics of kimchi, a fermented cabbage product. FEMS Microbiol. including Firmicutes, Proteobacteria, Synergistetes and Chloroflexi, Lett., 257, 262-267. were identified. Fourteen genera including Lactobacillus, Ding, X.F., Wu, C.D., Huang, J., Li, H., and Zhou, R.Q. (2014). Eubacterial Alkalibacterium, Pediococcus, Clostridium, Tetragenococcus, and archaeal community characteristics in the man-made pit mud Acinetobacter, Idiomarina, Alcanivorax, Halovibrio, revealed by combined PCR-DGGE and FISH analyses. Food Res. Int ., Brevundimonas, Sphingomonas, Aminobacterium, Sphaerobacter 62, 1047-1053. and unclassified_Rhodobacteraceae were detected. The IMCP Dobson, C.M., Deneer, H., Lee, S., Hemmingsen, S., Glaze, S., and Ziola, samples of different kinds or from different factories possessed B. (2002). Phylogenetic analysis of the genus Pediococcus, including different dominant bacteria. For fungal PCR-DGGE (Fig. 2), a total Pediococcus claussenii sp. nov., a novel lactic acid bacterium isolated of 9 bands were sequenced and affiliated with three genera, from beer. Int. J. Syst. Evol. Micr., 52, 2003-2010. including Debaryomyces, Pichia and Candida. Obviously, Fierer, N. and Jackson, R.B. (2006). The diversity and biogeography of soil Debaryomyces was mutual and dominant yeast in all IMCP bacterial communities. Proc. Natl. Acad. Sci. USA, 103, 626-631. samples. Fifteen bands from LAB DGGE profiles were identified Fukuda, D., Watanabe, M., Sonezaki, S., Sugimoto, S., Sonomoto, K., and and four genera including Alkalibacterium, Lactobacillus, Ishizaki, A. (2002). Molecular Characterization and regulatory analysis Pediococcus, Leuconostoc were detected. Lactobacillus and of dnaK operon of halophilic lactic acid bacterium Tetragenococcus Alkalibacterium were mutual and preponderant in all IMCP halophile. J. Biosci. Bioeng., 4, 388-394. samples. And the genus Alkalibacterium was found for the first Haque, M.A., Lee, J.H., and Cho, K.M. (2015). Endophytic bacterial time to be a dominated LAB in IMCP. The quantity of LAB was diversity in Korean kimchi made of Chinese cabbage leaves and their 7 _ 9 monitored by qPCR and shown to exist at 10 10 copies in the antimicrobial activity against pathogens. Food Control, 56, 24-33. IMCP samples from different factories The quantity of LAB in Haruta, S., Ueno, S., Egawa, I., Hashiguchi, K., Fujii, A., Nagano, M., Ishii, IMCP of different kinds or from different factories was different. M., and Igarashi, Y. (2006). Succession of bacterial and fungal 602 H. Liang et al.

communities during a traditional pot fermentation of rice vinegar D., Zhang, J., and Zhang, W.X. (2014). Monitoring of the prokaryotic assessed by PCR-mediated denaturing gradient gel electrophoresis. Int. J. diversity in pit mud from a Luzhou-flavor liquor distillery and evaluation Food Microbiol., 109, 79-87. of two predominant archaea using qPCR assays. J. Inst. Brew., 120, 253- Holzapfel, W. (1995). Use of starter cultures in fermentation on a household 261. scale. Food Control, 8, 241-258. Marshall, V.M. (1987). Lactic acid bacteria: Starters for flavour. FEMS Hong, Y., Li, J.M., Qin, P., Lee, S.Y., and Kim, H.Y. (2015). Predominant Microbiol. Lett., 46, 327-336. lactic acid bacteria in Mukeunji, a long-term-aged kimchi, for different Mokhtari, W., Nsaibia, S., Gharbi, A., and Aouni, M. (2013). Real-time aging periods. Food Sci. Biotechnol., 24, 545-550. PCR using SYBR Green for the detection of Shigella spp. in food and Ishikawa, M., Nakajima, K., Ishizaki, S., Kodama, K., Okamoto-Kainuma, stool samples. Mol. Cell. Probe., 27, 53-59. A., Koizumi, Y., Yamamoto, Y., and Yamasato, K. (2011). Nam, Y.D., Chang, H.W., Kim, K.H., Roh, S.W., and Bae, J.W. (2009). Alkalibacterium subtropicum sp. nov.a slightly halophilic marine lactic Metatranscriptome analysis of lactic acid bacteria during kimchi acid bacterium isolated from decaying marine algae. Int. J. Syst. Evol. fermentation with genome-probing microarrays. Int. J. Food Microbiol., Micro., 61, 2996-3002. 130, 140-146. Ishikawa, M., Yamasato, K., Kodama, K., Yasuda, H., Matsuyama, M., Nie, Z.Q., Zheng, Y., Du, H.F., Xie, S.K., and Wang, M. (2015). Dynamics Okamoto-Kainuma, A., and Koizumi, Y. (2013). Alkalibacterium gilvum and diversity of microbial community succession in traditional sp. nov., slightly halophilic and alkaliphilic lactic acid bacterium isolated fermentation of Shanxi aged vinegar. Food Microbiol., 47, 62-68. from soft and semi-hard cheeses. Int. J. Syst. Evol. Micro., 63, 1471- Nie, Z.Q., Zheng, Y., Wang, M., Han, Y., Wang, Y.N., Luo, J.M., and Niu, 1478. D.D. (2013). Exploring microbial succession and diversity during solid- Islam, S.M.A., Math, R.K., Kim, J.M., Yun, M.G., Cho, J.J., Kim, E.J., state fermentation of Tianjin duliu mature vinegar. Bioresource Technol., Lee, Y.H., and Yun, H.D. (2010). Effect of plant age on endophytic 148, 325-333. bacterial diversity of balloon flower Platycodon( grandiflorum) root and Park, E.J., Chun, J., Cha, C.J., Park, W.S., Jeon, C.O., and Bae, J.W. their antimicrobial activities. Curr. Microbiol., 61, 346-356. (2012). Bacterial community analysis during fermentation of ten Iwobi, A., Sebah, D., Kraemer, I., Losher, C., Fischer, G., Busch, U., and representative kinds of kimchi with barcoded pyrosequencing. Food Huber, I. (2015). A multiplex real-time PCR method for the Microbiol., 30, 197-204. quantification of beef and pork fractions in minced meat. Food Chem., Penas, E., Frias, J., Sidro, B., and Vidal-Valverde, C. (2010). Chemical 169, 305-313. evaluation and sensory quality of sauerkrauts obtained by natural and Jeong, S.H., Jung, J.Y., Lee, S.H., Jin, H.M., and Jeon, C.O. (2013). induced fermentations at different NaCl levels from Brassica oleracea Microbial succession and metabolite changes during fermentation of Var. capitata cv. Bronco grown in eastern Spain. Effect of storage. J. dongchimi, traditional Korean watery kimchi. Int. J. Food Microbiol., Agr. Food Chem., 58, 3549-3557. 164, 46-53. Pereira, G.V.D.M., Magalhaes, K.T., Lorenzetii, E.R., Souza, T.P., and Kao, Y.T., Liu, Y.S., and Shyu, Y.T. (2007). Identification of Lactobacillus Schwan, R.F. (2012). A multiphasic approach for the identification of spp. in probiotic products by real-time PCR and melting curve analysis. endophytic bacterial in strawberry fruit and their potential for plant Food Res. Int., 40, 71-79. growth promotion. Microb. Ecol., 63, 405-417. Lee, K. and Lee, Y. (2010). Effect of Lactobacillus plantarum as a starter Pielou, E.C. (1966). Species-diversity and pattern-diversity in the study of on the food quality and microbiota of kimchi. Food Sci. Biotechnol., 19, ecological succession. J. Theor. Biol ., 10, 370-383. 641-646. Plengvidhya, V., Breidt, F., Lu, Z., and Fleming, H.P. (2007). DNA Lee, K.W., Park, J.Y., Chun, J., Han, N.S., and Kim, J.H. (2010). fingerprinting of lactic acid bacteria in sauerkraut fermentations. Appl. Importance of Weissella species during kimchi fermentation and future Environ. Microb., 73, 7697-7702. works. Korean J. Microbiol. Biotechn., 38, 341-348. Rai, K.P., Zhang, C., and Xia, W.S. (2010). Effects of pure starter cultures Li, S.H., Pu, B., and Chen. F.Z. (2005). Research progress of function and on physico-chemical and sensory quality of dry fermented Chinese-style shelf-life of pickles. China Brew., 4, 6-8 (in Chinese). sausage. J. Food Sci. Tech. Mys., 47, 188-194. Liu, W.L., Zhang, L.W., Shi, J., Yi, H.X., Zhang, Y.C., Zhang, S., Gao, W., Romanenko, L.A., Uchino, M., Kalinovskaya, N.I., and Mikhailov, V.V. Du, M., Han, X., and Yu, X.J. (2015). Assessment of the safety and (2008). Isolation, phylo-genetic analysis and screening of marine applications of bacteriocinogenic Enterococcus faecium Y31 as an mollusk-associated bacteria for antimicrobial, hemolytic and surface adjunct culture in North-eastern Chinese traditional fermentation paocai. activities. Microbiol. Res., 163, 633-644. Food Control, 50, 637-644. Ryu, S.H., Park, M., Lee, J.R., Yun, P.Y., and Jeon, C.O. (2007). Lopez, I., Ruiz-Larrea, F., Cocolin, L., Orr, E., Phister, T., Marshall, M., Brevundimonas aveniformis sp. nov., a stalked species isolated from VanderGheynst, J., and Mills, D.A. (2003). Design and evaluation of activated sludge. Int. J. Syst. Evol. Micro., 57, 1561-1565. PCR primers for analysis of bacterial populations in wine by denaturing Shannon, C.E. (1949). Communication theory of secrecy systems. Bell Syst. gradient gel electrophoresis. Appl. Environ. Microb., 69, 6801-6807. Tech. J., 28, 656-715. Luo, Q.C., Liu, C.L., Wu, Z.Y., Wang, H.Y., Li, W.F., Zhang, K.Z., Huang, Simpson, P.J., Stanton, C., Fitzgerald, G.F., and Ross, R.P. (2002). Microbial Community Differences in Industrial Matured Chinese Paocai 603

Genomic diversity within the genus Pediococcus as revealed by Yan, P.M., Xue, W.T., Tan, S.S., Zhang, H., and Chang, X.H. (2008). randomly amplified polymorphic DNA PCR and pulsed-field gel Effect of inoculating lactic acid bacteria starter cultures on the nitrite electrophoresis. Appl. Environ. Microb., 68, 765-771. concentration of fermenting Chinese paocai. Food Control, 19, 50-55. Soares, S., Amaral, J.S., Oliveira, M.B.P.P., and Mafra, I. (2013). A SYBR Yeun, H., Yang, H.S., Chang, H.C., and Kim, H.Y. (2013). Comparison of Green real-time PCR assay to detect and quantify pork meat in processed bacterial community changes in fermenting kimchi at two different poultry meat products. Meat Sci., 94, 115-120. temperatures using a denaturing gradient gel electrophoresis analysis. J. Taira, W., Funatsu, Y., Satomi, M., Takano, T., and Abe, H. (2007). Microbiol. Biotechn., 23, 76-84. Changes in extractive components and microbial proliferation during Yeun, H., Yang, H.S., Li, J.M., Han, S., Chang, H.C., and Kim, H.Y. fermentation of fish sauce from underutilized fish species and quality of (2014). Identification of lactic acid bacteria in salted Chinese cabbage by final products.Fisheries Sci., 73, 913-923. SDS-PAGE and PCR-DGGE. J. Sci. Food Agric., 94, 296-300. Tanganurat, W., Quinquis, B., Leelawatcharamas, V., and Bolotin, A. Yu, J., Gao, W., Qing, M.J., Sun, Z.H., Wang, W.H., Liu, W.J., Pan, L., (2009). Genotypic and phenotypic characterization of Lactobacillus Sun, T., Wang, H.M., Bai, N., and Zhang, H.P. (2012). Identification and plantarum strains isolated from Thai fermented fruits and vegetables. J. characterization of lactic acid bacteria isolated from traditional pickles in Basic Microb., 49, 377-385. Sichuan, China. J. Gen. Appl. Microbiol., 58, 163-172. Walter, J., Hertel, C., Tannock, G.W., Lis, C.M., Munro, K., and Hammes, Zhang, W.X., Qiao, Z.W., Shigematsu, T., Tang, Y.Q., Hu, C., Morimura, W.P. (2001). Detection of Lactobacillus, Pediococcus, Leuconostoc, and S., and Kida, K. (2005). Analysis of the bacterial community in Zaopei Weissella Species in human feces by using group-specific PCR primers during production of Chinese Luzhou-flavor liquor. J. Inst. Brew., 111, and denaturing gradient gel electrophoresis. Appl. Environ. Microbiol., 215-222. 67, 2578-2585. Zhang, X.X., Gao, J.S., Cao, Y.H., Ma, X.T., and He, J.Z. (2013). Long- Wouters, D., Bernaert, N., Conjaerts, W., Van Droogenbroeck, B., De term rice and green manure rotation alters the endophytic bacterial Loose, M., and De Vuyst, L. (2013). Species diversity, community communities of the rice root. Microb. Ecol., 66, 917-926. dynamics, and metabolite kinetics of spontaneous leek fermentations. Zhou, F., Zhao, H.F., Bai, F.L., Dziugan, P., Liu, Y., and Zhang, B.L. Food Microbiol., 33, 185-196. (2014). Purification and characterisation of the bacteriocin produced by Xiong, T., Guan, Q.Q., Song, S.H., Hao, M.Y., and Xie, M.Y. (2012). Lactobacillus plantarum, isolated from Chinese pickle. Czech J. Food Dynamic changes of lactic acid bacteria flora during Chinese sauerkraut Sci., 32, 430-436. fermentation. Food Control, 26, 178-181. Xiong, T., Song, S.H., Huang, X.H., Feng, C., Liu, G.Q., Huang, J.Q., and URL cited Xie, M.Y. (2013). Screening and identification of functional i) www.ncbi.nlm.nih.gov/BLAST/ (Oct. 06, 2015) Lactobacillus specific for vegetable fermentation. J. Food Sci., 78, 84- ii) http://rdp.cme.msu.edu/index.jsp (Oct. 06, 2015) 89. 604 H. Liang et al.

Supplementary Fig. 1. The standard curve for LAB community in IMCP samples from different factories (C1, C2 and C3). P1 and P2 denoted two kinds of vegetables samples.

Supplementary Fig. 2. The melting curves for LAB community in IMCP samples from different factories (C1, C2 and C3). P1 and P2 denoted two kinds of vegetables samples.