- Email: [email protected]
Microbial community analysis of Korean soybean pastes by next-generation sequencing
International Journal of Food Microbiology 155 (2012) 36–42
Contents lists available at SciVerse ScienceDirect
International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro
Microbial community analysis of Korean soybean pastes by next-generation sequencing Young-Do Nam, So-Young Lee, Seong-Il Lim ⁎ Fermentation and Functionality Research Group, Korea Food Research Institute, Sungnam 463-746, Republic of Korea
a r t i c l e
i n f o
Article history: Received 30 November 2011 Received in revised form 16 January 2012 Accepted 16 January 2012 Available online 24 January 2012 Keywords: Doenjang Korean fermented soybean food 16S rRNA 454 pyrosequencing Microbial community
a b s t r a c t Soybean pastes, doenjang, have long been consumed as a fortified protein source in Korea. The quality of doenjang is determined by fermentation and Bacillus subtilis is suspected to be the main microorganism responsible. In the current culture-independent analysis, 17,675 bacterial sequences were derived from nine local and two commercial brands of doenjang samples by a barcoded pyrosequencing method targeting the hyper-variable regions V1/V2 of the 16S rRNA gene. In contrast to what has previously been found using plating or conventional molecular biology based methods, doenjang contains a diversity of bacterial species (total 208 species) and each doenjang reflects a region-specific bacterial community. While the Bacillus species was thought to be dominant in soybean pastes, we found that they were in high abundance (58.3–91.6%) only in samples from the central region of Korea, whereas lactic acid bacteria (LAB) (39.8–77.7%) were the dominant bacterial members of other doenjang samples. Compared to local brands of doenjang, commercial brands contain simple microbial communities dominated by Tetragenococcus and Staphylococcus that resemble the microbial communities of Japanese miso; this suggests that artificial inoculation was used for the quality control and standardization of doenjang. In this study, a massive sequencing approach was applied for the first time to analyze the microbial communities of different doenjang samples. Thus, we have determined that massive sequencing is a valid approach for assessing the overall microbial community of Korean fermented soybean pastes. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Doenjang is a traditional fermented soybean food product that has been consumed for centuries as a protein source and flavoring ingredient in Korea, like that of miso in Japan and tempeh in Indonesia (Golbitz, 1995). Doenjang is traditionally prepared by mixing and fermenting brine with moldy cooked soybeans (meju) in which naturally transferred microorganisms degrade soy proteins and produce many nutritious compounds (Cho and Seo, 2007). Doenjang is also recognized as a nutritious food that provides essential amino acids lacking in cereal protein diets, as well as fatty acids, organic acid, minerals and vitamins (Namgung et al., 2010). Recently, doenjang has gained attention not only because of its excellent nutritional value but also because of its health-promoting properties, such as its antioxidant (Park et al., 2008), fibrinolytic (Choi et al., 2009), antimutagenic (Kim, 2004), and anticancer effects (Jung et al., 2006). The quality and functionality of doenjang is affected by microbes, by the fermentation process, and by basic ingredients such as soybeans or grains (Yoo et al., 2000). ⁎ Corresponding author at: Korea Food Research Institute, 516, Baekhyun-dong, Bundang-gu, Sungnam-si, Gyeonggi-do 463-746, Republic of Korea. Tel.: + 82 31 780 9277; fax: + 82 31 709 9876. E-mail address: [email protected] (S.-I. Lim). 0168-1605/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.ijfoodmicro.2012.01.013
Diverse microorganisms participate in doenjang fermentation and produce the unique flavors and tastes of doenjang by decomposing soybean protein during fermentation. While it has been reported that Bacillus subtilis and B. licheniformis are the dominant organisms in doenjang and play important roles during fermentation, these previous studies were mainly based on culture-dependent methods limited to only identifying isolated strains growing on specific nutrient media (Shin et al., 1985; Yoo et al., 1999). Recently, 16S rRNA cloning and sequencing was used to analyze the bacterial community of doenjang and revealed that Staphylococcus equorum and some lactic acid bacteria (LAB) were the dominant microbes, rather than the previously identified Bacillus species (Cho and Seo, 2007). Kim et al. (2009) also determined the diversity of microbes involved in doenjang fermentation using PCR-DGGE (denaturing gradient gel electrophoresis) as a culture-independent analysis method. In that study, LAB, including Leuconostoc mesenteroides, Tetragenococcus halophilus, and Enterococcus faecium, were predominant, and Bacillus species could only be identified by nested PCR-DGGE using Bacillus specific primers. Recent advances in sequencing technology, such as the 454 pyrosequencing approach, are changing the way that microbial communities are analyzed. These sequence-by-synthesis methods represent a simple and rapid way of studying the microbial community by permitting the analysis of hundreds of thousands of nucleotide
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
sequences at one time (Cardenas and Tiedje, 2008) and have been successfully used to characterize the microbial diversity in various fermented foods such as Ben-saalga (Humblot and Guyot, 2009), jeotgal (Roh et al., 2010), makgeolli (Jung et al., 2011) and meju (Kim et al., 2011). In the current study, detailed analysis of community structure and comparative analysis of bacterial communities were carried out. Nine regional and two commercial brands of doenjang were analyzed by 454 pyrosequencing using sample-specific barcoded primers targeting the hyper-variable regions V1/V2 of the 16S rRNA gene. To our knowledge, this is the first molecular ecological investigation into the microbial community of doenjang using a deep sequencing approach. The results of this work will broaden our knowledge about microorganisms participating in the fermentation process and will help improve the quality of fermented soy foods.
37
Table 2 Number of sequences analyzed, observed diversity richness (OTUs), estimated OTU richness (Chao 1), diversity index (Shannon), and estimated sample coverage for 16S rRNA libraries of doenjang samples. Sample Barcode ID
Reads OTUs Chao1
Shannon
D1 D2 D3 D4 D5 D6 D7 D8 D9 R1 R2
434 3136 2932 1388 2369 1459 1297 869 308 1013 2470
2.88 3.39 3.83 3.30 2.83 3.10 4.15 4.10 3.20 1.78 2.33
AAGGTTGG AATAGCGG AATAGGCG ACACAGAG ACACCTGA ACAGACAG ACAGCAGA ACAGTCTG ACAGTGAG ACCACATG ACCACTAG
66 124 204 113 136 107 167 128 59 70 81
112 187 303 146 211 195 222 198 158 121 198
(85, 177) (152, 265) (260, 377) (128, 187) (174, 283) (149, 291) (195, 274) (162, 274) (95, 329) (91, 192) (128, 367)
(2.73, (3.34, (3.77, (3.22, (2.76, (3.02, (4.08, (4.03, (3.05, (1.65, (2.27,
Coverage (%) 3.03) 3.44) 3.88) 3.39) 2.90) 3.18) 4.22) 4.18) 3.34) 1.90) 2.39)
92.2 98.7 97.1 97.2 97.6 96.4 95.7 94.1 89.6 96.6 98.3
Abbreviations: OTU, operational taxonomic unit; ESC, estimated sample coverage. OTUs, Chao 1, Shannon and ESC were calculated with MOTHUR at the 3% distance level. Values in brackets represent 95% confidence intervals.
2. Materials and methods 2.1. Sample collection and DNA extraction Nine samples (D1 to D9) made via the traditional Korean process and two commercially manufactured samples (R1 and R2) were collected from manufacturers and immediately stored at − 20 °C. Total DNA was extracted from doenjang samples, as previously described (Jung et al., 2011). The extracted DNA was subsequently purified with the UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories, CA, USA). The DNA concentration and quality was determined by 1% agarose gel electrophoresis and with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, DE, USA).
Purification kit (Qiagen, CA, USA). An equal quantity (100 ng) of each PCR amplicon tagged with the sample-specific barcode sequences was pooled and the quantity and quality of DNA was further assessed on a BioAnalyzer 2100 microfluidics device (Agilent, CA, USA) using a DNA1000 lab chip (Agilent, CA, USA). The pooled DNA was amplified by emulsion PCR before sequencing by synthesis using the massive parallel pyrosequencing protocol (Margulies et al., 2005) and sequencing was performed using a 454 pyrosequencing Genome Sequencer FLX Titanium (Life Sciences, CT, USA), according to the manufacturer's instructions, by a commercial sequencing facility (Macrogen, Seoul, Korea).
2.2. Pyrosequencing of bacterial 16S rRNA fragments 2.3. Sequence processing To analyze the bacterial community, regions V1/V2 of the 16S rRNA gene of DNA extracted from doenjang samples were amplified, sequenced and analyzed as previously described (Nam et al., 2011). To amplify the bacterial 16S rRNA gene fragments, 20 ng of purified DNA (determined by a NanoDrop 1000 spectrophotometer) was amplified with a TOPsimple™ DryMIX solution (Enzynomics, Daejeon, Korea). The V1 and V2 hyper-variable regions of the bacterial 16S rRNA gene (Baker et al., 2003) were amplified with the primer pair 8F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 338R (5′-TGCTGCCTCCCGTAGG AGT-3′) containing eight base sample-specific barcode sequences (Table 2) and common linker (TC for forward and CA for reverse primer) sequences in the 5′ end (Hamady et al., 2008). Thermocycling was conducted in a C1000 Thermal Cycler (Bio-Rad, CA, USA) under the following conditions: initial denaturation at 94 °C for 2 min; 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min; and a final extension at 72 °C for 10 min. The quality of the amplified PCR products was verified by electrophoresis in a 1% agarose gel and purified using the QIAquick PCR
The sequences generated from pyrosequencing were mainly analyzed with the software MOTHUR for pre-processing (quality-adjustment, barcode split), identification of operational taxonomic units (OTUs), taxonomic assignment, community comparison, and statistical analysis (Schloss et al., 2009). Sequences were filtered to minimize the effects of poor sequence quality and sequencing errors by removing sequences with more than one ambiguous base call and retaining only sequences that were 300 nt or longer. Samplespecific sequences were collected according to their barcode sequences tagged to each sample and barcode, forward and reverse primer sequences were trimmed from initial sequences. By the preprocessing of pyrosequences, sequences that were shorter than 300 nt, had one or more ambiguous base calls, or had multiple barcode or primer motifs were excluded from the analysis. We included only sequences with the forward primer motif to ensure that the highly informative V1/V2 region was available for taxonomic assignment. The sequences obtained in this study were uploaded and made available through the DDJB database under the accession numbers DRS001177 to DRS001187. 2.4. Operational taxonomic unit (OTU) determination and taxonomic classification
Table 1 List of the analyzed doenjang samples. Sample ID
Region
Salt concentration (%)
pH
D1 D2 D3 D4 D5 D6 D7 D8 D9 R1 R2
Cheorwon, Gangwon Anseong, Gyeonggi Yangpyeong, Gyeonggi Gyeongsan, Gyeongbuk Damyang, Jeonam Sunchang, Jeonbuk Cheongyang, Chungnam Jeonju, Jeonbuk Jocheon, Jeju Commercial Commercial
11.7 17.1 12.4 12.0 12.7 10.8 14.5 14.9 14.2 11.6 11.8
4.81 4.89 5.27 6.67 5.31 5.48 4.95 5.77 5.22 5.46 5.67
The trimmed sequences from each barcode bin were aligned using Infernal, and associated covariance models were obtained from the Ribosomal Database Project Group (Cole et al., 2009) and further trimmed to encompass the same V1/V2 regions for accurate analysis by using the same regions. Then, the sequences spanning the same region were realigned with the SILVA compatible alignment database that contains 50,000 columns and 14,956 aligned bacterial sequences (http://www.mothur.org/w/images/9/98/Silva.bacteria.zip). A distance matrix was calculated from the aligned sequences, and operational taxonomical units (OTUs; 90 to 100% sequence similarity)
38
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
were assigned by using the furthest neighbor clustering algorithm. The OTUs defined by a 3% distance level were phylogenetically classified with a modified bacterial RDP II database containing 164,517 almost full-length 16S rRNA gene sequences prepared using TaxCollector (http://www.microgator.org) and a taxonomy file containing the complete taxonomic information of each sequence in the database from domain to species 2.5. Community comparison analysis To compare the overall bacterial community of local and commercial brands of doenjang, OTU-based and phylogeny-based analysis were archived. For the OTU-based analysis, pyrosequencing reads from each sample were assigned as an OTU with 97% sequence identity and the OTU information from each sample was then transferred into a dendrogram. The distances between microbial communities from each sample were calculated using the Yue and Clayton θ coefficient and represented as an Unweighted Pair Group Method with an Arithmetic Mean (UPGMA) clustering tree describing the dissimilarity (1-similarity) between multiple samples. A Newick-formatted tree file was generated using this analysis. For the phylogeny-based analysis, the genetic distance between all sequences was calculated and a phylogenetic tree was constructed. Then, unifrac-based metrics inferring the similarity between the two communities' membership and structure were generated by computing the fraction of the branch length on the tree unique to each sample. These metrics were used to generate PCoA plots describing whether communities of analyzed samples exhibit the same structure. 2.6. Calculation of species richness and diversity indices Shannon's diversity (H′ = −∑piln(pi) where pi is the proportion of taxon i) (Gotelli, 2002), Chao I richness indices (Chao and Bunge, 2002), rarefaction curves (Colwell and Coddington, 1994) and Good's coverage (G = 1 − n/N, where n is the number of phylotypes that have been sampled once and N is the total number of individuals in the sample) were generated with a 3% sequence dissimilarity cut-off value. 2.7. Calculation of pH and salt concentration For the pH measurement, 25 ml of deionized distilled water was added to 5 g of doenjang samples, homogenized and then filtered with whatman filter paper (No. 2). The pHs of the samples were measured using a digital pH meter (Orion 3-Star Plus pH Meters, Thermo Scientific, USA). For the determination of salt content, doenjang samples were analyzed by Volhard titration method. 3 g of samples were diluted with 50 ml of deionized distilled water, transferred to a conical flask, 20 ml of 0.1 N of silver nitrate solution was added to the flask and then, 5 ml of 3 N of nitric acid solution and 0.5 ml of iron alum solution were added to the sample solution. The total solution was titrated with 0.1 N of potassium thiocyanate solution and concentration of sodium chloride content in doenjang was measured by calculating the added volume of potassium thiocyanate solution.
sample, as estimated by the Chao1 estimator, was considerably higher than the observed number of OTUs, covering an average of 60.4% (±11.9) of the estimated richness, which suggests that there could be on average 72.7 (±24.7) additional bacterial phylotypes. Additionally, the Shannon diversity index varied from 1.78 to 4.15 (average, 3.17 ± 0.72) (Table 2). When a rarefaction analysis was carried out to determine whether all the OTUs present in the datasets had been sufficiently recovered in the pyrosequencing study, individual rarefaction curves showed a similar pattern of becoming gradually stable without reaching a saturation phase, suggesting that a large number of unseen OTUs still existed in the original samples and that more sequencing efforts may be required to detect additional phylotypes (Fig. 1). However, Good's coverage for the samples, which provides an estimate of sampling completeness using a probability calculation with randomly selected amplicon sequences, was an average of 95.8% (±2.7) with 97% species level phylotypes, suggesting that the majority of bacterial phylotypes present in doenjang samples were identified. 3.2. Bacterial communities of doenjang samples All the doenjang samples was dominated by the phylum Firmicutes, Actinobacteria, Bacteroides, Proteobacteria and unclassified bacteria but the bacterial communities of each doenjang sample at this level were not significantly different between samples (Fig. 2A). As shown in Fig. 2B, all doenjang samples were populated by nine families, comprising more than 90.8% of the bacterial population, but the most abundant family differed significantly in each sample. For example, samples D1, D4, D6, D9, and R2 showed the highest abundance of Enterococcaceae family; samples D2, D3, D7, and D8 displayed the highest abundance of Bacillaceae family; and samples D5 and R2 showed the highest abundance of Staphylococcaceae. To analyze the bacterial community associated with doenjang fermentation in detail, a heat-map representing the relative abundance of each species level phylotype was constructed (Fig. 3). Two hundred and five species level phylotypes were identified and bacterial species were classified into three different population groups depending on the relative abundance, where b1% were rare, 1% to b10% were subdominant, and above 10% were considered to be a predominant group. One hundred and sixty one species, constituting an average of 78.9% of the total species, were considered rare, and the integrated relative abundance of these rare species accounted for 4.7% of the total population, while those of the 11 predominant species accounted for 75.1%. Therefore, the current study showed high resolution in presenting a large
3. Results and discussion 3.1. Comparison of phylotypes and diversity estimates of bacterial communities After quality control processes filtered out reads containing incorrect primer or barcode sequences and sequences that were shorter than 300 nucleotides or with more than one ambiguous base, a total of 17,675 high quality sequences were obtained. Each doenjang sample was covered by an average of 1606 (±977) reads and the average number of OTUs was 144 (±44). The total number of OTUs for each
Fig. 1. Rarefaction analysis of V1/V2 pyrosequencing reads of the bacterial 16S rRNA gene from elevendoenjang samples. Rarefaction curves were constructed at a 97% sequence similarity cut-off value.
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
Fig. 2. Relative abundance of bacterial 16S rRNA genes from doenjang samples at the phylum level (A) and family level (B) identified with a modified 16S rRNA database from the Ribosomal Database Project (RDP).
diversity of rare microorganisms along with predominant members of doenjang fermentation as compared to previous studies using culturedependent or -independent methods (Cho and Seo, 2007; Kim et al., 2009; Yoo et al., 1999, 2000). Previous studies reported that Korean fermented soybean paste doenjang is dominated by B. subtilis and B. licheniformis, but only three of eleven samples (samples D2, D3, and D7) were dominated by these two Bacillus species. In addition, previously isolated Bacillus species such as B. amyloliquefaciens and B. pumilis were only detected in some samples at low levels, and B. atrophaeus was not even detected by this massive sequencing approach. Eight samples contained less than 25% of Bacillus species (average: 8.4%, SD: 8.5%). In contrast, the majority of doenjang samples were dominated by sample-specific Bacillus, Staphylococcus and LAB (Table 3): sample D1 contained mainly Enterococcus faecalis (22.4%), E. faecium (24.2%) and Tetragenococcus halophilus (19.8%); sample D2 contained mainly B. licheniformis (57.4%) and B. subtilis (21.9%); sample D3 contained mainly B. licheniformis (15.4%), B. subtilis (23.1%), E. faecalis (11.2%) and E. faecium (12.3%); sample D4 contained mainly E. faecalis (19.0%), E. faecium (20.2%) and T. halophilus (28.2%); sample D5 Fig. 3. Relative abundance of species level taxa. Each column in the heat-map represents one doenjang sample, while each row represents a species level phylotype. The color intensity of the panel is proportional to the abundance of OTUs (max 10%). The staggered bars on the right indicate the specific genus groups.
39
40
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
Table 3 Relative abundance (%) of 25 major species from each doenjang sample. Taxon name
D1
D2
D3
D4
D5
D6
D7
D8
D9
R1
R2
Bacillus licheniformis Bacillus subtilis Bacillus amyloliquefaciens Bacillus sonorensis Bacillus sp. BCL23-2 Bacillus sp. BSS2 Bacillus sp. R-30903 Bacillus sp. YIM KMY7 Brochothrix sp. NJ-25 Enterococcus faecalis Enterococcus faecium Enterococcus casseliflavus Lactobacillus halophilus Leuconostoc mesenteroides Leuconostoc sp. HBB8 Pediococcus acidilactici Staphylococcus gallinarum Staphylococcus lentus Staphylococcus saprophyticus Staphylococcus sciuri Staphylococcus nepalensis Tetragenococcus halophilus Weissella hellenica Weissella salipiscis Pseudomonas sp. NJ-22
1.8% 7.1% 1.8% 2.1% 0.0% 1.8% 0.0% 0.5% 0.0% 22.4% 24.2% 0.0% 1.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 19.8% 0.0% 0.2% 0.0%
57.4% 21.9% 2.0% 0.9% 4.1% 3.4% 0.1% 1.8% 0.0% 0.5% 0.4% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.3% 0.1% 0.5% 0.0% 0.0% 0.0% 0.0% 0.0%
15.4% 23.1% 3.2% 6.1% 0.9% 6.0% 0.5% 3.0% 0.0% 11.2% 12.3% 0.4% 0.0% 0.0% 0.0% 0.0% 0.2% 5.3% 1.0% 2.2% 0.1% 0.0% 0.0% 0.4% 0.0%
0.2% 3.7% 0.4% 0.0% 0.0% 0.7% 0.0% 0.2% 0.0% 19.0% 20.2% 1.2% 3.6% 0.0% 0.1% 4.4% 0.1% 0.5% 0.2% 0.5% 3.2% 28.2% 0.0% 1.0% 0.0%
0.9% 3.9% 0.4% 0.0% 0.0% 0.4% 0.0% 0.4% 0.0% 8.3% 10.0% 0.5% 0.0% 0.3% 0.0% 0.2% 1.0% 20.4% 10.3% 27.8% 0.2% 4.4% 0.0% 1.3% 0.0%
0.2% 1.4% 0.1% 0.2% 0.0% 0.3% 0.0% 0.0% 0.0% 1.4% 3.1% 0.3% 0.0% 0.0% 0.1% 0.0% 0.1% 0.0% 0.5% 0.5% 1.0% 69.8% 0.0% 0.4% 0.1%
26.4% 22.0% 1.5% 0.9% 1.9% 1.9% 4.2% 0.2% 0.5% 0.2% 1.0% 0.2% 0.2% 4.5% 4.2% 0.0% 0.2% 0.3% 0.3% 0.4% 0.7% 4.4% 0.2% 0.0% 6.2%
9.4% 9.7% 1.0% 0.6% 0.1% 2.3% 0.5% 1.5% 3.2% 0.7% 1.7% 0.1% 4.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 6.1% 28.3% 4.9% 0.0% 0.2%
0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.6% 13.3% 4.2% 23.7% 15.3% 1.6% 0.0% 0.0% 3.2% 5.8% 0.3% 5.8% 9.7% 0.0% 0.0% 0.0%
1.3% 6.9% 0.6% 1.9% 0.0% 0.8% 0.0% 0.3% 0.1% 0.0% 0.0% 0.0% 0.1% 0.4% 0.0% 0.0% 73.0% 0.0% 0.0% 0.0% 0.0% 1.5% 0.0% 9.7% 0.0%
0.3% 0.9% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 30.8% 0.0% 0.0% 0.0% 0.0% 53.9% 0.0% 0.4% 0.0%
contained mainly E. faecium (10.0%), Staphylococcus lentus (20.4%), S. saprophyticus (10.3%) and S. sciuri (27.8%); sample D6 contained mainly T. halophilus (69.8%); sample D7 contained mainly B. licheniformis (26.4%) and B. subtilis (22.0%); sample D8 contained mainly B. licheniformis (9.4%), B. subtilis (9.7%) and T. halophilus (28.3%); and sample D9 contained mainly E. faecium (13.3%), Lactobacillus halophilus (23.7%), and Leuconostoc mesenteroides (15.3%). Compared to local brands, commercial doenjang brands showed somewhat reduced bacterial diversities (75, average of other samples: 122) and have distinctive community structures; S. gallinarum (73.0%) and Weissella salipiscis (9.7%) were the major species in R1, while S. gallinarum (30.8%) and T. halophilus (53.9%) were the major species in R2. In the current study, all doenjang samples contained various lactic acid bacteria and many of the identified LAB overlapped with those in Japanese miso. While Zhao et al. (2009) also reported that LAB were the predominant bacteria contributing to the fermentation of Chinese soybean pastes, the LAB types differed significantly from Korean or Japanese soybean pastes. Therefore, the overall microbial community of Korean fermented soybean pastes resembles that of Japanese fermented soybean pastes rather than Chinese ones. Among the LAB, the Enterococcus species were broadly distributed in local doenjang samples (average: 17.6%; SD: 17.1%). Kim et al. (2009) reported that E. faecium, along with Leu. mesenteroides and T. halophilus, was broadly distributed among fermented Korean soybean pastes and Onda et al. (2002) revealed the widespread nature of E. faecium in miso. Proliferation of LAB not only contributes to the generation of flavor compounds but also inhibits the growth of pathogenic bacteria by producing acids and bacteriocins. It is known that enterocin, a bacteroicin produced by E. faecium, can prevent food spoilage and inhibit pathogenic bacteria such as Listeria sp., Staphylococcus aureus, Vibrio cholera, Clostridium sp. and even Bacillus species (Yoon et al., 2008). In fact, the levels of Bacillus species in samples with a high abundance of E. faecium were relatively low on average (6.6%), except for one sample (D3). Therefore, it is reasonable to assume that bacteriocinproducers associated with doenjang such as E. faecium may be important in maintaining the fermentation conditions by preventing the growth of contaminants. In the current study, nine of eleven doenjang samples contained T. halophilus and, among these samples, five samples primarily contained T. halophilus at more than 19.8% of the total population
(average: 40%; SD: 21%). T. halophilus are a halophilic lactic acid bacterial species that require high NaCl content for growth, and the salt concentration of doenjang samples used in this study range from 10.8 to 14.9% (Table 1). T. halophilus (previously Pediococcus halophilus) was isolated from salted anchovies (Villar et al., 1985) and is known to be a predominant component of fermented foods with a high salt concentration such as soy sauce, fish sauce (Tanasupawat et al., 2002) and miso (Onda et al., 2003). T. halophilus is known to play a role in developing appealing flavors and in masking offensive ones during the ripening of miso (Kim et al., 2010b) and was also identified as one of the predominant lactic acid bacteria in doenjang (Kim et al., 2009). It is interesting that none of the previous studies using molecular methods such as DGGE and pyrosequencing detected T. halophilus in meju samples (Kim et al., 2011; Lee et al., 2010), since doenjang is manufactured by mixing and fermenting brine with moldy cooked soybeans (meju), which is used as a natural starter of doenjang fermentation. Therefore, it could be assumed that T. halophilus is included in the mixing process of brine or that T. halophilus remained dormant in the meju at extremely low levels. The commercial samples R1 and R2 showed higher levels of Staphylococcus gallinarum at 73.0% and 30.8%, respectively, compared to the local brands of doenjang (average 0.2%). In the previous study analyzing doenjang by the DGGE method, S. gallinarum was also detected in a commercial brand of doenjang (Kim et al., 2009) and S. gallinarum is also one of the most dominant bacterial species identified in Japanese miso (Kim et al., 2010b). S. gallinarum was also isolated from fermented sausages (Drosinos et al., 2007) While it is not clear where the S. gallinarum comes from and how this microorganism makes up such a large proportion, high productivity of useful enzymes and antibacterial potential against undesirable microorganisms of S. gallinarum (Nychas and Arkoudelos, 1989; Ottenwaelder et al., 1995) might play important roles in the fermentation of doenjang. 3.3. Comparison of bacterial communities of doenjang samples The other objective of this study was to compare the microbial community of local and commercial brands of doenjang collected from different regions. To compare the bacterial composition of each doenjang sample collected from various regions of Korea, phylotype-based and phylogeny-based comparisons were carried
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
out. A UPGMA tree constructed by the dissimilarity of phylotypes concatenated with 97% sequence identity in each sample revealed three different clusters (Fig. 4A): group 1 — D1, D3, D4, D5 and D9; group 2 — D6, D8, R1 and R2; and group 3 — D2 and D7. In the clustering analysis based on the comparison of phylotypes, only two samples collected from the same province, Jeonbuk, were grouped together and the other samples showed mixed features that did not correspond to their collected regions. Samples D2 and D7, which were collected from the Gyeonggi and Chungnam provinces, respectively, clustered together as a separate branch from the other samples. Interestingly, these two samples are in close proximity geographically. In addition, commercial samples of doenjang, R1 and R2, clustered together with D6 and D8, which were collected from Jeonbuk, the same province where R2 was manufactured. UniFrac analysis data that took into account the phylogenetic distances of each bacterium in doenjang samples showed similar but distinctive clustering features compared to UPGMA clustering analysis (Fig. 4B). The maximum variations in bacterial communities of doenjang samples were found to be 38.3% (PC1) and 20.5% (PC2) with a strong separation by region. PCoA also revealed three different groups: group 1– D1, D3, D4, D5, D9 and R1; group 2 – D6, D8 and R2; and group 3 – D2 and D7. In group 1, geographically distant samples D1 (Gangwon), D4 (Gyengbuk) and D9 (Jeju) showed a close relationship, along with the R1 sample. While samples D2 and D7 and samples D6, D8 and R2 were separately grouped together, similar to the results from the UPGMA analysis, the bacterial communities of D2 and D7 differed significantly and the bacterial community of D6 was more closely related to the commercial brand R2 than the local brand D8. The local brand D6 and the commercial brand R2 were manufactured in the same city of the Jeonbuk province, suggesting that the microbial community of doenjang was influenced by the manufacturing location.
41
Doenjang is made from meju fermented in a natural environment so that various region-specific microorganisms including fungi, yeast, LAB and Bacillus species are transferred during the manufacturing process. These microbes participate in doenjang fermentation, and the growth of each microbial group is influenced by other microbes. Therefore, elucidating the overall microbial community in doenjang is important for understanding the fermentation process and for improving the quality of fermented soybean foods. Previous studies using culture-dependent and independent methods only revealed bacteria such as B. subtilis, B. licheniformis and B. amyloliquefaciens as the predominant members of Korean fermented soybean paste (Kim et al., 2010a), while the current study showed that the above Bacillus species were only the dominant species of a few doenjang samples and that a variety of doenjang samples contained less than 10% of Bacillus species (sum of three Bacillus species). In addition, a large diversity of rare species participating in doenjang fermentation (each consisting of less than 1% of the total population) accounted for approximately 80% of the total number of species. The current study found that 51 Bacillus species level phylotypes, including 33 unclassified Bacillus, were related to doenjang fermentation, and doenjang samples from different regions had a sample-specific LAB composition constituted by 45 different lactic acid bacteria species. While we could not address how these rare microspheres interact with the predominant microorganisms within the community, rare species may play an important role in maintaining the stability of the microbial community and might have potential to become the predominant microbial source when the appropriate conditions are presented. Therefore, it should be noted that the microbial community of doenjang not only consists of a few predominant species but also includes many other rarer species. In conclusion, the current study showed that massive sequencing approach can be successfully applied for the analysis of the microbial community of fermented food. While additional studies will be needed to fully analyze the microorganisms that participate in fermentation and to determine the biochemical mechanisms underlying Korean fermented soybean paste, the insight gained from current study will give us an improved knowledge for understanding the function of microorganisms in Korean traditional fermented foods.
References
Fig. 4. Community comparison of microbiota in each doenjang sample. Each community was clustered by the phylotype-based UPGMA method (A) and phylogeny-based weighted unifrac method (B).
Baker, G.C., Smith, J.J., Cowan, D.A., 2003. Review and re-analysis of domain-specific 16S primers. Journal of Microbiological Methods 55, 541–555. Cardenas, E., Tiedje, J.M., 2008. New tools for discovering and characterizing microbial diversity. Current Opinion in Biotechnology 19, 544–549. Chao, A., Bunge, J., 2002. Estimating the number of species in a stochastic abundance model. Biometrics 58, 531–539. Cho, K.M., Seo, W.T., 2007. Bacterial diversity in a Korean traditional soybean fermented foods (doenjang and ganjang) by 16S rRNA gene sequence analysis. Food Science and Biotechnology 16, 320–324. Choi, N.-S., Chung, D.-M., Han, Y.-J., Kim, S.-H., Song, J.J., 2009. Purification and characterization of a subtilisin D5, a fibrinolytic enzyme of Bacillus amyloliquefaciens DJ-5 isolated from doenjang. Food Science and Biotechnology 18, 500–505. Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., Kulam-Syed-Mohideen, A.S., McGarrell, D.M., Marsh, T., Garrity, G.M., Tiedje, J.M., 2009. The Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Research 37, D141–D145. Colwell, R.K., Coddington, J.A., 1994. Estimating terrestrial biodiversity through extrapolation. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 345, 101–118. Drosinos, E.H., Paramithiotis, S., Kolovos, G., Tsikouras, I., Metaxopoulos, I., 2007. Phenotypic and technological diversity of lactic acid bacteria and staphylococci isolated from traditionally fermented sausages in Southern Greece. Food Microbiology (London) 24, 260–270. Golbitz, P., 1995. Traditional soyfoods: processing and products. Journal of Nutrition 125, 570S–572S. Gotelli, N.J., 2002. Ecology: biodiversity in the scales. Nature 419, 575–576. Hamady, M., Walker, J.J., Harris, J.K., Gold, N.J., Knight, R., 2008. Error-correcting barcoded primers for pyrosequencing hundreds of samples in multiplex. Nature Methods 5, 235–237. Humblot, C., Guyot, J.-P., 2009. Pyrosequencing of tagged 16S rRNA gene amplicons for rapid deciphering of the microbiomes of fermented foods such as pearl millet slurries. Applied and Environmental Microbiology 75, 4354–4361.
42
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
Jung, K.-O., Park, S.-Y., Park, K.-Y., 2006. Longer aging time increases the anticancer and antimetastatic properties of doenjang. Nutrition 22, 539–545. Jung, M.-J., Nam, Y.-D., Roh, S.W., Bae, J.-W., 2011. Unexpected convergence of fungal and bacterial communities during fermentation of traditional Korean alcoholic beverages inoculated with various natural starters. Food Microbiology 30 (1), 112–123. Kim, J.-G., 2004. Antigenotoxic effects of water extract from Korean fermented soybean paste (doen-jang). Journal of Food Protection 67, 156–161. Kim, T.-W., Lee, J.-H., Kim, S.-E., Park, M.-H., Chang, H.C., Kim, H.-Y., 2009. Analysis of microbial communities in doenjang, a Korean fermented soybean paste, using nested PCR-denaturing gradient gel electrophoresis. International Journal of Food Microbiology 131, 265–271. Kim, T.-W., Kim, Y.-H., Kim, S.-E., Lee, J.-H., Park, C.-S., Kim, H.-Y., 2010a. Identification and distribution of bacillus species in doenjang by whole-cell protein patterns and 16S rRNA gene sequence analysis. Journal of Microbiology and Biotechnology 20, 1210–1214. Kim, T.W., Lee, J.-H., Park, M.-H., Kim, H.-Y., 2010b. Analysis of bacterial and fungal communities in Japanese- and Chinese-fermented soybean pastes using nested PCR-DGGE. Current Microbiology 60, 315–320. Kim, Y.-S., Kim, M.-C., Kwon, S.-W., Kim, S.-J., Park, I.-C., Ka, J.-O., Weon, H.-Y., 2011. Analyses of bacterial communities in Meju, a Korean traditional fermented soybean bricks, by cultivation-based and pyrosequencing methods. Journal of Microbiology 49, 340–348. Lee, J.H., Kim, T.W., Lee, H., Chang, H.C., Kim, H.Y., 2010. Determination of microbial diversity in meju, fermented cooked soya beans, using nested PCR-denaturing gradient gel electrophoresis. Letters in Applied Microbiology 51, 388–394. Margulies, M., Egholm, M., Altman, W.E., Attiya, S., Bader, J.S., Bemben, L.A., Berka, J., Braverman, M.S., Chen, Y.J., Chen, Z., Dewell, S.B., Du, L., Fierro, J.M., Gomes, X.V., Godwin, B.C., He, W., Helgesen, S., Ho, C.H., Irzyk, G.P., Jando, S.C., Alenquer, M.L., Jarvie, T.P., Jirage, K.B., Kim, J.B., Knight, J.R., Lanza, J.R., Leamon, J.H., Lefkowitz, S.M., Lei, M., Li, J., Lohman, K.L., Lu, H., Makhijani, V.B., McDade, K.E., McKenna, M.P., Myers, E.W., Nickerson, E., Nobile, J.R., Plant, R., Puc, B.P., Ronan, M.T., Roth, G.T., Sarkis, G.J., Simons, J.F., Simpson, J.W., Srinivasan, M., Tartaro, K.R., Tomasz, A., Vogt, K.A., Volkmer, G.A., Wang, S.H., Wang, Y., Weiner, M.P., Yu, P., Begley, R.F., Rothberg, J.M., 2005. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380. Nam, Y.-D., Jung, M.-J., Roh, S.W., Kim, M.-S., Bae, J.-W., 2011. Comparative analysis of Korean human gut microbiota by barcoded pyrosequencing. PLoS One 6, e22109. Namgung, H.-J., Park, H.-J., Cho, I.H., Choi, H.-K., Kwon, D.-Y., Shim, S.-M., Kim, Y.-S., 2010. Metabolite profiling of doenjang, fermented soybean paste, during fermentation. Journal of the Science of Food and Agriculture 90, 1926–1935. Nychas, G.J., Arkoudelos, J., 1989. Staphylococci: their role in fermented sausages. Journal of Applied Bacteriology 67, vi.
Onda, T., Yanagida, F., Uchimura, T., Tsuji, M., Ogino, S., Shinohara, T., Yokotsuka, K., 2002. Widespread distribution of the bacteriocin-producing lactic acid cocci in Miso-paste products. Journal of Applied Microbiology 92, 695–705. Onda, T., Yanagida, F., Uchimura, T., Tsuji, M., Ogino, S., Shinohara, T., Yokotsuka, K., 2003. Analysis of lactic acid bacterial flora during miso fermentation. Food Science and Technology Research 9, 17–24. Ottenwaelder, B., Kupke, T., Brecht, S., Gnau, V., Metzger, J., Jung, G., Goetz, F., 1995. Isolation and characterization of genetically engineered gallidermin and epidermin analogs. Applied and Environmental Microbiology 61, 3894–3903. Park, J.-S., Park, H.Y., Kim, D.H., Kim, D.H., Kim, H.K., 2008. Ortho-dihydroxyisoflavone derivatives from aged Doenjang (Korean fermented soypaste) and its radical scavenging activity. Bioorganic & Medicinal Chemistry Letters 18, 5006–5009. Roh, S.W., Kim, K.H., Nam, Y.D., Chang, H.W., Park, E.J., Bae, J.W., 2010. Investigation of archaeal and bacterial diversity in fermented seafood using barcoded pyrosequencing. ISME Journal 4, 1–16. Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski, R.A., Oakley, B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G., Van Horn, D.J., Weber, C.F., 2009. Introducing mothur: open-source, platformindependent, community-supported software for describing and comparing microbial communities. Applied and Environmental Microbiology 75, 7537–7541. Shin, S.Y., Kim, Y.B., Yu, T.J., 1985. Flavour improvement of soybean pastes by the addition of Bacillus licheniformis and Saccharomyces rouxii. Korean Journal of Food Science and Technology 17, 8–14. Tanasupawat, S., Thongsanit, J., Okada, S., Komagata, K., 2002. Lactic acid bacteria isolated from soy sauce mash in Thailand. Journal of General and Applied Microbiology 48, 201–209. Villar, M., Ruiz Holgado, A.P., Sanchez, J.J., Trucco, R.E., Oliver, G., 1985. Isolation and characterization of Pediococcus halophilus from salted anchovies (Engraulis anchoita). Applied and Environmental Microbiology 49, 664–666. Yoo, S.-K., Cho, W.-H., Kang, S.-M., Lee, S.-H., 1999. Isolation and identification of microorganisms in Korean traditional soybean paste and soybean sauce. Korean Journal of Applied Microbiology and Biotechnology 27, 113–117. Yoo, S.-K., Kang, S.-M., Noh, Y.-S., 2000. Quality properties on soy bean pastes made with microorganisms isolated from traditional soy bean pastes. Korean Journal of Food Science and Technology 32, 1266–1270. Yoon, M.Y., Kim, Y.J., Hwang, H.-J., 2008. Properties and safety aspects of Enterococcus faecium strains isolated from Chungkukjang, a fermented soy product. LWT- Food Science and Technology 41, 925–933. Zhao, J., Dai, X., Liu, X., Chen, H., Tang, J., Zhang, H., Chen, W., 2009. Changes in microbial community during Chinese traditional soybean paste fermentation. International Journal of Food Science and Technology 44, 2526–2530.
Contents lists available at SciVerse ScienceDirect
International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro
Microbial community analysis of Korean soybean pastes by next-generation sequencing Young-Do Nam, So-Young Lee, Seong-Il Lim ⁎ Fermentation and Functionality Research Group, Korea Food Research Institute, Sungnam 463-746, Republic of Korea
a r t i c l e
i n f o
Article history: Received 30 November 2011 Received in revised form 16 January 2012 Accepted 16 January 2012 Available online 24 January 2012 Keywords: Doenjang Korean fermented soybean food 16S rRNA 454 pyrosequencing Microbial community
a b s t r a c t Soybean pastes, doenjang, have long been consumed as a fortified protein source in Korea. The quality of doenjang is determined by fermentation and Bacillus subtilis is suspected to be the main microorganism responsible. In the current culture-independent analysis, 17,675 bacterial sequences were derived from nine local and two commercial brands of doenjang samples by a barcoded pyrosequencing method targeting the hyper-variable regions V1/V2 of the 16S rRNA gene. In contrast to what has previously been found using plating or conventional molecular biology based methods, doenjang contains a diversity of bacterial species (total 208 species) and each doenjang reflects a region-specific bacterial community. While the Bacillus species was thought to be dominant in soybean pastes, we found that they were in high abundance (58.3–91.6%) only in samples from the central region of Korea, whereas lactic acid bacteria (LAB) (39.8–77.7%) were the dominant bacterial members of other doenjang samples. Compared to local brands of doenjang, commercial brands contain simple microbial communities dominated by Tetragenococcus and Staphylococcus that resemble the microbial communities of Japanese miso; this suggests that artificial inoculation was used for the quality control and standardization of doenjang. In this study, a massive sequencing approach was applied for the first time to analyze the microbial communities of different doenjang samples. Thus, we have determined that massive sequencing is a valid approach for assessing the overall microbial community of Korean fermented soybean pastes. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Doenjang is a traditional fermented soybean food product that has been consumed for centuries as a protein source and flavoring ingredient in Korea, like that of miso in Japan and tempeh in Indonesia (Golbitz, 1995). Doenjang is traditionally prepared by mixing and fermenting brine with moldy cooked soybeans (meju) in which naturally transferred microorganisms degrade soy proteins and produce many nutritious compounds (Cho and Seo, 2007). Doenjang is also recognized as a nutritious food that provides essential amino acids lacking in cereal protein diets, as well as fatty acids, organic acid, minerals and vitamins (Namgung et al., 2010). Recently, doenjang has gained attention not only because of its excellent nutritional value but also because of its health-promoting properties, such as its antioxidant (Park et al., 2008), fibrinolytic (Choi et al., 2009), antimutagenic (Kim, 2004), and anticancer effects (Jung et al., 2006). The quality and functionality of doenjang is affected by microbes, by the fermentation process, and by basic ingredients such as soybeans or grains (Yoo et al., 2000). ⁎ Corresponding author at: Korea Food Research Institute, 516, Baekhyun-dong, Bundang-gu, Sungnam-si, Gyeonggi-do 463-746, Republic of Korea. Tel.: + 82 31 780 9277; fax: + 82 31 709 9876. E-mail address: [email protected] (S.-I. Lim). 0168-1605/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.ijfoodmicro.2012.01.013
Diverse microorganisms participate in doenjang fermentation and produce the unique flavors and tastes of doenjang by decomposing soybean protein during fermentation. While it has been reported that Bacillus subtilis and B. licheniformis are the dominant organisms in doenjang and play important roles during fermentation, these previous studies were mainly based on culture-dependent methods limited to only identifying isolated strains growing on specific nutrient media (Shin et al., 1985; Yoo et al., 1999). Recently, 16S rRNA cloning and sequencing was used to analyze the bacterial community of doenjang and revealed that Staphylococcus equorum and some lactic acid bacteria (LAB) were the dominant microbes, rather than the previously identified Bacillus species (Cho and Seo, 2007). Kim et al. (2009) also determined the diversity of microbes involved in doenjang fermentation using PCR-DGGE (denaturing gradient gel electrophoresis) as a culture-independent analysis method. In that study, LAB, including Leuconostoc mesenteroides, Tetragenococcus halophilus, and Enterococcus faecium, were predominant, and Bacillus species could only be identified by nested PCR-DGGE using Bacillus specific primers. Recent advances in sequencing technology, such as the 454 pyrosequencing approach, are changing the way that microbial communities are analyzed. These sequence-by-synthesis methods represent a simple and rapid way of studying the microbial community by permitting the analysis of hundreds of thousands of nucleotide
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
sequences at one time (Cardenas and Tiedje, 2008) and have been successfully used to characterize the microbial diversity in various fermented foods such as Ben-saalga (Humblot and Guyot, 2009), jeotgal (Roh et al., 2010), makgeolli (Jung et al., 2011) and meju (Kim et al., 2011). In the current study, detailed analysis of community structure and comparative analysis of bacterial communities were carried out. Nine regional and two commercial brands of doenjang were analyzed by 454 pyrosequencing using sample-specific barcoded primers targeting the hyper-variable regions V1/V2 of the 16S rRNA gene. To our knowledge, this is the first molecular ecological investigation into the microbial community of doenjang using a deep sequencing approach. The results of this work will broaden our knowledge about microorganisms participating in the fermentation process and will help improve the quality of fermented soy foods.
37
Table 2 Number of sequences analyzed, observed diversity richness (OTUs), estimated OTU richness (Chao 1), diversity index (Shannon), and estimated sample coverage for 16S rRNA libraries of doenjang samples. Sample Barcode ID
Reads OTUs Chao1
Shannon
D1 D2 D3 D4 D5 D6 D7 D8 D9 R1 R2
434 3136 2932 1388 2369 1459 1297 869 308 1013 2470
2.88 3.39 3.83 3.30 2.83 3.10 4.15 4.10 3.20 1.78 2.33
AAGGTTGG AATAGCGG AATAGGCG ACACAGAG ACACCTGA ACAGACAG ACAGCAGA ACAGTCTG ACAGTGAG ACCACATG ACCACTAG
66 124 204 113 136 107 167 128 59 70 81
112 187 303 146 211 195 222 198 158 121 198
(85, 177) (152, 265) (260, 377) (128, 187) (174, 283) (149, 291) (195, 274) (162, 274) (95, 329) (91, 192) (128, 367)
(2.73, (3.34, (3.77, (3.22, (2.76, (3.02, (4.08, (4.03, (3.05, (1.65, (2.27,
Coverage (%) 3.03) 3.44) 3.88) 3.39) 2.90) 3.18) 4.22) 4.18) 3.34) 1.90) 2.39)
92.2 98.7 97.1 97.2 97.6 96.4 95.7 94.1 89.6 96.6 98.3
Abbreviations: OTU, operational taxonomic unit; ESC, estimated sample coverage. OTUs, Chao 1, Shannon and ESC were calculated with MOTHUR at the 3% distance level. Values in brackets represent 95% confidence intervals.
2. Materials and methods 2.1. Sample collection and DNA extraction Nine samples (D1 to D9) made via the traditional Korean process and two commercially manufactured samples (R1 and R2) were collected from manufacturers and immediately stored at − 20 °C. Total DNA was extracted from doenjang samples, as previously described (Jung et al., 2011). The extracted DNA was subsequently purified with the UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories, CA, USA). The DNA concentration and quality was determined by 1% agarose gel electrophoresis and with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, DE, USA).
Purification kit (Qiagen, CA, USA). An equal quantity (100 ng) of each PCR amplicon tagged with the sample-specific barcode sequences was pooled and the quantity and quality of DNA was further assessed on a BioAnalyzer 2100 microfluidics device (Agilent, CA, USA) using a DNA1000 lab chip (Agilent, CA, USA). The pooled DNA was amplified by emulsion PCR before sequencing by synthesis using the massive parallel pyrosequencing protocol (Margulies et al., 2005) and sequencing was performed using a 454 pyrosequencing Genome Sequencer FLX Titanium (Life Sciences, CT, USA), according to the manufacturer's instructions, by a commercial sequencing facility (Macrogen, Seoul, Korea).
2.2. Pyrosequencing of bacterial 16S rRNA fragments 2.3. Sequence processing To analyze the bacterial community, regions V1/V2 of the 16S rRNA gene of DNA extracted from doenjang samples were amplified, sequenced and analyzed as previously described (Nam et al., 2011). To amplify the bacterial 16S rRNA gene fragments, 20 ng of purified DNA (determined by a NanoDrop 1000 spectrophotometer) was amplified with a TOPsimple™ DryMIX solution (Enzynomics, Daejeon, Korea). The V1 and V2 hyper-variable regions of the bacterial 16S rRNA gene (Baker et al., 2003) were amplified with the primer pair 8F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 338R (5′-TGCTGCCTCCCGTAGG AGT-3′) containing eight base sample-specific barcode sequences (Table 2) and common linker (TC for forward and CA for reverse primer) sequences in the 5′ end (Hamady et al., 2008). Thermocycling was conducted in a C1000 Thermal Cycler (Bio-Rad, CA, USA) under the following conditions: initial denaturation at 94 °C for 2 min; 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min; and a final extension at 72 °C for 10 min. The quality of the amplified PCR products was verified by electrophoresis in a 1% agarose gel and purified using the QIAquick PCR
The sequences generated from pyrosequencing were mainly analyzed with the software MOTHUR for pre-processing (quality-adjustment, barcode split), identification of operational taxonomic units (OTUs), taxonomic assignment, community comparison, and statistical analysis (Schloss et al., 2009). Sequences were filtered to minimize the effects of poor sequence quality and sequencing errors by removing sequences with more than one ambiguous base call and retaining only sequences that were 300 nt or longer. Samplespecific sequences were collected according to their barcode sequences tagged to each sample and barcode, forward and reverse primer sequences were trimmed from initial sequences. By the preprocessing of pyrosequences, sequences that were shorter than 300 nt, had one or more ambiguous base calls, or had multiple barcode or primer motifs were excluded from the analysis. We included only sequences with the forward primer motif to ensure that the highly informative V1/V2 region was available for taxonomic assignment. The sequences obtained in this study were uploaded and made available through the DDJB database under the accession numbers DRS001177 to DRS001187. 2.4. Operational taxonomic unit (OTU) determination and taxonomic classification
Table 1 List of the analyzed doenjang samples. Sample ID
Region
Salt concentration (%)
pH
D1 D2 D3 D4 D5 D6 D7 D8 D9 R1 R2
Cheorwon, Gangwon Anseong, Gyeonggi Yangpyeong, Gyeonggi Gyeongsan, Gyeongbuk Damyang, Jeonam Sunchang, Jeonbuk Cheongyang, Chungnam Jeonju, Jeonbuk Jocheon, Jeju Commercial Commercial
11.7 17.1 12.4 12.0 12.7 10.8 14.5 14.9 14.2 11.6 11.8
4.81 4.89 5.27 6.67 5.31 5.48 4.95 5.77 5.22 5.46 5.67
The trimmed sequences from each barcode bin were aligned using Infernal, and associated covariance models were obtained from the Ribosomal Database Project Group (Cole et al., 2009) and further trimmed to encompass the same V1/V2 regions for accurate analysis by using the same regions. Then, the sequences spanning the same region were realigned with the SILVA compatible alignment database that contains 50,000 columns and 14,956 aligned bacterial sequences (http://www.mothur.org/w/images/9/98/Silva.bacteria.zip). A distance matrix was calculated from the aligned sequences, and operational taxonomical units (OTUs; 90 to 100% sequence similarity)
38
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
were assigned by using the furthest neighbor clustering algorithm. The OTUs defined by a 3% distance level were phylogenetically classified with a modified bacterial RDP II database containing 164,517 almost full-length 16S rRNA gene sequences prepared using TaxCollector (http://www.microgator.org) and a taxonomy file containing the complete taxonomic information of each sequence in the database from domain to species 2.5. Community comparison analysis To compare the overall bacterial community of local and commercial brands of doenjang, OTU-based and phylogeny-based analysis were archived. For the OTU-based analysis, pyrosequencing reads from each sample were assigned as an OTU with 97% sequence identity and the OTU information from each sample was then transferred into a dendrogram. The distances between microbial communities from each sample were calculated using the Yue and Clayton θ coefficient and represented as an Unweighted Pair Group Method with an Arithmetic Mean (UPGMA) clustering tree describing the dissimilarity (1-similarity) between multiple samples. A Newick-formatted tree file was generated using this analysis. For the phylogeny-based analysis, the genetic distance between all sequences was calculated and a phylogenetic tree was constructed. Then, unifrac-based metrics inferring the similarity between the two communities' membership and structure were generated by computing the fraction of the branch length on the tree unique to each sample. These metrics were used to generate PCoA plots describing whether communities of analyzed samples exhibit the same structure. 2.6. Calculation of species richness and diversity indices Shannon's diversity (H′ = −∑piln(pi) where pi is the proportion of taxon i) (Gotelli, 2002), Chao I richness indices (Chao and Bunge, 2002), rarefaction curves (Colwell and Coddington, 1994) and Good's coverage (G = 1 − n/N, where n is the number of phylotypes that have been sampled once and N is the total number of individuals in the sample) were generated with a 3% sequence dissimilarity cut-off value. 2.7. Calculation of pH and salt concentration For the pH measurement, 25 ml of deionized distilled water was added to 5 g of doenjang samples, homogenized and then filtered with whatman filter paper (No. 2). The pHs of the samples were measured using a digital pH meter (Orion 3-Star Plus pH Meters, Thermo Scientific, USA). For the determination of salt content, doenjang samples were analyzed by Volhard titration method. 3 g of samples were diluted with 50 ml of deionized distilled water, transferred to a conical flask, 20 ml of 0.1 N of silver nitrate solution was added to the flask and then, 5 ml of 3 N of nitric acid solution and 0.5 ml of iron alum solution were added to the sample solution. The total solution was titrated with 0.1 N of potassium thiocyanate solution and concentration of sodium chloride content in doenjang was measured by calculating the added volume of potassium thiocyanate solution.
sample, as estimated by the Chao1 estimator, was considerably higher than the observed number of OTUs, covering an average of 60.4% (±11.9) of the estimated richness, which suggests that there could be on average 72.7 (±24.7) additional bacterial phylotypes. Additionally, the Shannon diversity index varied from 1.78 to 4.15 (average, 3.17 ± 0.72) (Table 2). When a rarefaction analysis was carried out to determine whether all the OTUs present in the datasets had been sufficiently recovered in the pyrosequencing study, individual rarefaction curves showed a similar pattern of becoming gradually stable without reaching a saturation phase, suggesting that a large number of unseen OTUs still existed in the original samples and that more sequencing efforts may be required to detect additional phylotypes (Fig. 1). However, Good's coverage for the samples, which provides an estimate of sampling completeness using a probability calculation with randomly selected amplicon sequences, was an average of 95.8% (±2.7) with 97% species level phylotypes, suggesting that the majority of bacterial phylotypes present in doenjang samples were identified. 3.2. Bacterial communities of doenjang samples All the doenjang samples was dominated by the phylum Firmicutes, Actinobacteria, Bacteroides, Proteobacteria and unclassified bacteria but the bacterial communities of each doenjang sample at this level were not significantly different between samples (Fig. 2A). As shown in Fig. 2B, all doenjang samples were populated by nine families, comprising more than 90.8% of the bacterial population, but the most abundant family differed significantly in each sample. For example, samples D1, D4, D6, D9, and R2 showed the highest abundance of Enterococcaceae family; samples D2, D3, D7, and D8 displayed the highest abundance of Bacillaceae family; and samples D5 and R2 showed the highest abundance of Staphylococcaceae. To analyze the bacterial community associated with doenjang fermentation in detail, a heat-map representing the relative abundance of each species level phylotype was constructed (Fig. 3). Two hundred and five species level phylotypes were identified and bacterial species were classified into three different population groups depending on the relative abundance, where b1% were rare, 1% to b10% were subdominant, and above 10% were considered to be a predominant group. One hundred and sixty one species, constituting an average of 78.9% of the total species, were considered rare, and the integrated relative abundance of these rare species accounted for 4.7% of the total population, while those of the 11 predominant species accounted for 75.1%. Therefore, the current study showed high resolution in presenting a large
3. Results and discussion 3.1. Comparison of phylotypes and diversity estimates of bacterial communities After quality control processes filtered out reads containing incorrect primer or barcode sequences and sequences that were shorter than 300 nucleotides or with more than one ambiguous base, a total of 17,675 high quality sequences were obtained. Each doenjang sample was covered by an average of 1606 (±977) reads and the average number of OTUs was 144 (±44). The total number of OTUs for each
Fig. 1. Rarefaction analysis of V1/V2 pyrosequencing reads of the bacterial 16S rRNA gene from elevendoenjang samples. Rarefaction curves were constructed at a 97% sequence similarity cut-off value.
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
Fig. 2. Relative abundance of bacterial 16S rRNA genes from doenjang samples at the phylum level (A) and family level (B) identified with a modified 16S rRNA database from the Ribosomal Database Project (RDP).
diversity of rare microorganisms along with predominant members of doenjang fermentation as compared to previous studies using culturedependent or -independent methods (Cho and Seo, 2007; Kim et al., 2009; Yoo et al., 1999, 2000). Previous studies reported that Korean fermented soybean paste doenjang is dominated by B. subtilis and B. licheniformis, but only three of eleven samples (samples D2, D3, and D7) were dominated by these two Bacillus species. In addition, previously isolated Bacillus species such as B. amyloliquefaciens and B. pumilis were only detected in some samples at low levels, and B. atrophaeus was not even detected by this massive sequencing approach. Eight samples contained less than 25% of Bacillus species (average: 8.4%, SD: 8.5%). In contrast, the majority of doenjang samples were dominated by sample-specific Bacillus, Staphylococcus and LAB (Table 3): sample D1 contained mainly Enterococcus faecalis (22.4%), E. faecium (24.2%) and Tetragenococcus halophilus (19.8%); sample D2 contained mainly B. licheniformis (57.4%) and B. subtilis (21.9%); sample D3 contained mainly B. licheniformis (15.4%), B. subtilis (23.1%), E. faecalis (11.2%) and E. faecium (12.3%); sample D4 contained mainly E. faecalis (19.0%), E. faecium (20.2%) and T. halophilus (28.2%); sample D5 Fig. 3. Relative abundance of species level taxa. Each column in the heat-map represents one doenjang sample, while each row represents a species level phylotype. The color intensity of the panel is proportional to the abundance of OTUs (max 10%). The staggered bars on the right indicate the specific genus groups.
39
40
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
Table 3 Relative abundance (%) of 25 major species from each doenjang sample. Taxon name
D1
D2
D3
D4
D5
D6
D7
D8
D9
R1
R2
Bacillus licheniformis Bacillus subtilis Bacillus amyloliquefaciens Bacillus sonorensis Bacillus sp. BCL23-2 Bacillus sp. BSS2 Bacillus sp. R-30903 Bacillus sp. YIM KMY7 Brochothrix sp. NJ-25 Enterococcus faecalis Enterococcus faecium Enterococcus casseliflavus Lactobacillus halophilus Leuconostoc mesenteroides Leuconostoc sp. HBB8 Pediococcus acidilactici Staphylococcus gallinarum Staphylococcus lentus Staphylococcus saprophyticus Staphylococcus sciuri Staphylococcus nepalensis Tetragenococcus halophilus Weissella hellenica Weissella salipiscis Pseudomonas sp. NJ-22
1.8% 7.1% 1.8% 2.1% 0.0% 1.8% 0.0% 0.5% 0.0% 22.4% 24.2% 0.0% 1.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 0.0% 19.8% 0.0% 0.2% 0.0%
57.4% 21.9% 2.0% 0.9% 4.1% 3.4% 0.1% 1.8% 0.0% 0.5% 0.4% 0.0% 0.0% 0.0% 0.0% 0.1% 0.0% 0.3% 0.1% 0.5% 0.0% 0.0% 0.0% 0.0% 0.0%
15.4% 23.1% 3.2% 6.1% 0.9% 6.0% 0.5% 3.0% 0.0% 11.2% 12.3% 0.4% 0.0% 0.0% 0.0% 0.0% 0.2% 5.3% 1.0% 2.2% 0.1% 0.0% 0.0% 0.4% 0.0%
0.2% 3.7% 0.4% 0.0% 0.0% 0.7% 0.0% 0.2% 0.0% 19.0% 20.2% 1.2% 3.6% 0.0% 0.1% 4.4% 0.1% 0.5% 0.2% 0.5% 3.2% 28.2% 0.0% 1.0% 0.0%
0.9% 3.9% 0.4% 0.0% 0.0% 0.4% 0.0% 0.4% 0.0% 8.3% 10.0% 0.5% 0.0% 0.3% 0.0% 0.2% 1.0% 20.4% 10.3% 27.8% 0.2% 4.4% 0.0% 1.3% 0.0%
0.2% 1.4% 0.1% 0.2% 0.0% 0.3% 0.0% 0.0% 0.0% 1.4% 3.1% 0.3% 0.0% 0.0% 0.1% 0.0% 0.1% 0.0% 0.5% 0.5% 1.0% 69.8% 0.0% 0.4% 0.1%
26.4% 22.0% 1.5% 0.9% 1.9% 1.9% 4.2% 0.2% 0.5% 0.2% 1.0% 0.2% 0.2% 4.5% 4.2% 0.0% 0.2% 0.3% 0.3% 0.4% 0.7% 4.4% 0.2% 0.0% 6.2%
9.4% 9.7% 1.0% 0.6% 0.1% 2.3% 0.5% 1.5% 3.2% 0.7% 1.7% 0.1% 4.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 6.1% 28.3% 4.9% 0.0% 0.2%
0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 1.6% 13.3% 4.2% 23.7% 15.3% 1.6% 0.0% 0.0% 3.2% 5.8% 0.3% 5.8% 9.7% 0.0% 0.0% 0.0%
1.3% 6.9% 0.6% 1.9% 0.0% 0.8% 0.0% 0.3% 0.1% 0.0% 0.0% 0.0% 0.1% 0.4% 0.0% 0.0% 73.0% 0.0% 0.0% 0.0% 0.0% 1.5% 0.0% 9.7% 0.0%
0.3% 0.9% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 30.8% 0.0% 0.0% 0.0% 0.0% 53.9% 0.0% 0.4% 0.0%
contained mainly E. faecium (10.0%), Staphylococcus lentus (20.4%), S. saprophyticus (10.3%) and S. sciuri (27.8%); sample D6 contained mainly T. halophilus (69.8%); sample D7 contained mainly B. licheniformis (26.4%) and B. subtilis (22.0%); sample D8 contained mainly B. licheniformis (9.4%), B. subtilis (9.7%) and T. halophilus (28.3%); and sample D9 contained mainly E. faecium (13.3%), Lactobacillus halophilus (23.7%), and Leuconostoc mesenteroides (15.3%). Compared to local brands, commercial doenjang brands showed somewhat reduced bacterial diversities (75, average of other samples: 122) and have distinctive community structures; S. gallinarum (73.0%) and Weissella salipiscis (9.7%) were the major species in R1, while S. gallinarum (30.8%) and T. halophilus (53.9%) were the major species in R2. In the current study, all doenjang samples contained various lactic acid bacteria and many of the identified LAB overlapped with those in Japanese miso. While Zhao et al. (2009) also reported that LAB were the predominant bacteria contributing to the fermentation of Chinese soybean pastes, the LAB types differed significantly from Korean or Japanese soybean pastes. Therefore, the overall microbial community of Korean fermented soybean pastes resembles that of Japanese fermented soybean pastes rather than Chinese ones. Among the LAB, the Enterococcus species were broadly distributed in local doenjang samples (average: 17.6%; SD: 17.1%). Kim et al. (2009) reported that E. faecium, along with Leu. mesenteroides and T. halophilus, was broadly distributed among fermented Korean soybean pastes and Onda et al. (2002) revealed the widespread nature of E. faecium in miso. Proliferation of LAB not only contributes to the generation of flavor compounds but also inhibits the growth of pathogenic bacteria by producing acids and bacteriocins. It is known that enterocin, a bacteroicin produced by E. faecium, can prevent food spoilage and inhibit pathogenic bacteria such as Listeria sp., Staphylococcus aureus, Vibrio cholera, Clostridium sp. and even Bacillus species (Yoon et al., 2008). In fact, the levels of Bacillus species in samples with a high abundance of E. faecium were relatively low on average (6.6%), except for one sample (D3). Therefore, it is reasonable to assume that bacteriocinproducers associated with doenjang such as E. faecium may be important in maintaining the fermentation conditions by preventing the growth of contaminants. In the current study, nine of eleven doenjang samples contained T. halophilus and, among these samples, five samples primarily contained T. halophilus at more than 19.8% of the total population
(average: 40%; SD: 21%). T. halophilus are a halophilic lactic acid bacterial species that require high NaCl content for growth, and the salt concentration of doenjang samples used in this study range from 10.8 to 14.9% (Table 1). T. halophilus (previously Pediococcus halophilus) was isolated from salted anchovies (Villar et al., 1985) and is known to be a predominant component of fermented foods with a high salt concentration such as soy sauce, fish sauce (Tanasupawat et al., 2002) and miso (Onda et al., 2003). T. halophilus is known to play a role in developing appealing flavors and in masking offensive ones during the ripening of miso (Kim et al., 2010b) and was also identified as one of the predominant lactic acid bacteria in doenjang (Kim et al., 2009). It is interesting that none of the previous studies using molecular methods such as DGGE and pyrosequencing detected T. halophilus in meju samples (Kim et al., 2011; Lee et al., 2010), since doenjang is manufactured by mixing and fermenting brine with moldy cooked soybeans (meju), which is used as a natural starter of doenjang fermentation. Therefore, it could be assumed that T. halophilus is included in the mixing process of brine or that T. halophilus remained dormant in the meju at extremely low levels. The commercial samples R1 and R2 showed higher levels of Staphylococcus gallinarum at 73.0% and 30.8%, respectively, compared to the local brands of doenjang (average 0.2%). In the previous study analyzing doenjang by the DGGE method, S. gallinarum was also detected in a commercial brand of doenjang (Kim et al., 2009) and S. gallinarum is also one of the most dominant bacterial species identified in Japanese miso (Kim et al., 2010b). S. gallinarum was also isolated from fermented sausages (Drosinos et al., 2007) While it is not clear where the S. gallinarum comes from and how this microorganism makes up such a large proportion, high productivity of useful enzymes and antibacterial potential against undesirable microorganisms of S. gallinarum (Nychas and Arkoudelos, 1989; Ottenwaelder et al., 1995) might play important roles in the fermentation of doenjang. 3.3. Comparison of bacterial communities of doenjang samples The other objective of this study was to compare the microbial community of local and commercial brands of doenjang collected from different regions. To compare the bacterial composition of each doenjang sample collected from various regions of Korea, phylotype-based and phylogeny-based comparisons were carried
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
out. A UPGMA tree constructed by the dissimilarity of phylotypes concatenated with 97% sequence identity in each sample revealed three different clusters (Fig. 4A): group 1 — D1, D3, D4, D5 and D9; group 2 — D6, D8, R1 and R2; and group 3 — D2 and D7. In the clustering analysis based on the comparison of phylotypes, only two samples collected from the same province, Jeonbuk, were grouped together and the other samples showed mixed features that did not correspond to their collected regions. Samples D2 and D7, which were collected from the Gyeonggi and Chungnam provinces, respectively, clustered together as a separate branch from the other samples. Interestingly, these two samples are in close proximity geographically. In addition, commercial samples of doenjang, R1 and R2, clustered together with D6 and D8, which were collected from Jeonbuk, the same province where R2 was manufactured. UniFrac analysis data that took into account the phylogenetic distances of each bacterium in doenjang samples showed similar but distinctive clustering features compared to UPGMA clustering analysis (Fig. 4B). The maximum variations in bacterial communities of doenjang samples were found to be 38.3% (PC1) and 20.5% (PC2) with a strong separation by region. PCoA also revealed three different groups: group 1– D1, D3, D4, D5, D9 and R1; group 2 – D6, D8 and R2; and group 3 – D2 and D7. In group 1, geographically distant samples D1 (Gangwon), D4 (Gyengbuk) and D9 (Jeju) showed a close relationship, along with the R1 sample. While samples D2 and D7 and samples D6, D8 and R2 were separately grouped together, similar to the results from the UPGMA analysis, the bacterial communities of D2 and D7 differed significantly and the bacterial community of D6 was more closely related to the commercial brand R2 than the local brand D8. The local brand D6 and the commercial brand R2 were manufactured in the same city of the Jeonbuk province, suggesting that the microbial community of doenjang was influenced by the manufacturing location.
41
Doenjang is made from meju fermented in a natural environment so that various region-specific microorganisms including fungi, yeast, LAB and Bacillus species are transferred during the manufacturing process. These microbes participate in doenjang fermentation, and the growth of each microbial group is influenced by other microbes. Therefore, elucidating the overall microbial community in doenjang is important for understanding the fermentation process and for improving the quality of fermented soybean foods. Previous studies using culture-dependent and independent methods only revealed bacteria such as B. subtilis, B. licheniformis and B. amyloliquefaciens as the predominant members of Korean fermented soybean paste (Kim et al., 2010a), while the current study showed that the above Bacillus species were only the dominant species of a few doenjang samples and that a variety of doenjang samples contained less than 10% of Bacillus species (sum of three Bacillus species). In addition, a large diversity of rare species participating in doenjang fermentation (each consisting of less than 1% of the total population) accounted for approximately 80% of the total number of species. The current study found that 51 Bacillus species level phylotypes, including 33 unclassified Bacillus, were related to doenjang fermentation, and doenjang samples from different regions had a sample-specific LAB composition constituted by 45 different lactic acid bacteria species. While we could not address how these rare microspheres interact with the predominant microorganisms within the community, rare species may play an important role in maintaining the stability of the microbial community and might have potential to become the predominant microbial source when the appropriate conditions are presented. Therefore, it should be noted that the microbial community of doenjang not only consists of a few predominant species but also includes many other rarer species. In conclusion, the current study showed that massive sequencing approach can be successfully applied for the analysis of the microbial community of fermented food. While additional studies will be needed to fully analyze the microorganisms that participate in fermentation and to determine the biochemical mechanisms underlying Korean fermented soybean paste, the insight gained from current study will give us an improved knowledge for understanding the function of microorganisms in Korean traditional fermented foods.
References
Fig. 4. Community comparison of microbiota in each doenjang sample. Each community was clustered by the phylotype-based UPGMA method (A) and phylogeny-based weighted unifrac method (B).
Baker, G.C., Smith, J.J., Cowan, D.A., 2003. Review and re-analysis of domain-specific 16S primers. Journal of Microbiological Methods 55, 541–555. Cardenas, E., Tiedje, J.M., 2008. New tools for discovering and characterizing microbial diversity. Current Opinion in Biotechnology 19, 544–549. Chao, A., Bunge, J., 2002. Estimating the number of species in a stochastic abundance model. Biometrics 58, 531–539. Cho, K.M., Seo, W.T., 2007. Bacterial diversity in a Korean traditional soybean fermented foods (doenjang and ganjang) by 16S rRNA gene sequence analysis. Food Science and Biotechnology 16, 320–324. Choi, N.-S., Chung, D.-M., Han, Y.-J., Kim, S.-H., Song, J.J., 2009. Purification and characterization of a subtilisin D5, a fibrinolytic enzyme of Bacillus amyloliquefaciens DJ-5 isolated from doenjang. Food Science and Biotechnology 18, 500–505. Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., Kulam-Syed-Mohideen, A.S., McGarrell, D.M., Marsh, T., Garrity, G.M., Tiedje, J.M., 2009. The Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Research 37, D141–D145. Colwell, R.K., Coddington, J.A., 1994. Estimating terrestrial biodiversity through extrapolation. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 345, 101–118. Drosinos, E.H., Paramithiotis, S., Kolovos, G., Tsikouras, I., Metaxopoulos, I., 2007. Phenotypic and technological diversity of lactic acid bacteria and staphylococci isolated from traditionally fermented sausages in Southern Greece. Food Microbiology (London) 24, 260–270. Golbitz, P., 1995. Traditional soyfoods: processing and products. Journal of Nutrition 125, 570S–572S. Gotelli, N.J., 2002. Ecology: biodiversity in the scales. Nature 419, 575–576. Hamady, M., Walker, J.J., Harris, J.K., Gold, N.J., Knight, R., 2008. Error-correcting barcoded primers for pyrosequencing hundreds of samples in multiplex. Nature Methods 5, 235–237. Humblot, C., Guyot, J.-P., 2009. Pyrosequencing of tagged 16S rRNA gene amplicons for rapid deciphering of the microbiomes of fermented foods such as pearl millet slurries. Applied and Environmental Microbiology 75, 4354–4361.
42
Y.-D. Nam et al. / International Journal of Food Microbiology 155 (2012) 36–42
Jung, K.-O., Park, S.-Y., Park, K.-Y., 2006. Longer aging time increases the anticancer and antimetastatic properties of doenjang. Nutrition 22, 539–545. Jung, M.-J., Nam, Y.-D., Roh, S.W., Bae, J.-W., 2011. Unexpected convergence of fungal and bacterial communities during fermentation of traditional Korean alcoholic beverages inoculated with various natural starters. Food Microbiology 30 (1), 112–123. Kim, J.-G., 2004. Antigenotoxic effects of water extract from Korean fermented soybean paste (doen-jang). Journal of Food Protection 67, 156–161. Kim, T.-W., Lee, J.-H., Kim, S.-E., Park, M.-H., Chang, H.C., Kim, H.-Y., 2009. Analysis of microbial communities in doenjang, a Korean fermented soybean paste, using nested PCR-denaturing gradient gel electrophoresis. International Journal of Food Microbiology 131, 265–271. Kim, T.-W., Kim, Y.-H., Kim, S.-E., Lee, J.-H., Park, C.-S., Kim, H.-Y., 2010a. Identification and distribution of bacillus species in doenjang by whole-cell protein patterns and 16S rRNA gene sequence analysis. Journal of Microbiology and Biotechnology 20, 1210–1214. Kim, T.W., Lee, J.-H., Park, M.-H., Kim, H.-Y., 2010b. Analysis of bacterial and fungal communities in Japanese- and Chinese-fermented soybean pastes using nested PCR-DGGE. Current Microbiology 60, 315–320. Kim, Y.-S., Kim, M.-C., Kwon, S.-W., Kim, S.-J., Park, I.-C., Ka, J.-O., Weon, H.-Y., 2011. Analyses of bacterial communities in Meju, a Korean traditional fermented soybean bricks, by cultivation-based and pyrosequencing methods. Journal of Microbiology 49, 340–348. Lee, J.H., Kim, T.W., Lee, H., Chang, H.C., Kim, H.Y., 2010. Determination of microbial diversity in meju, fermented cooked soya beans, using nested PCR-denaturing gradient gel electrophoresis. Letters in Applied Microbiology 51, 388–394. Margulies, M., Egholm, M., Altman, W.E., Attiya, S., Bader, J.S., Bemben, L.A., Berka, J., Braverman, M.S., Chen, Y.J., Chen, Z., Dewell, S.B., Du, L., Fierro, J.M., Gomes, X.V., Godwin, B.C., He, W., Helgesen, S., Ho, C.H., Irzyk, G.P., Jando, S.C., Alenquer, M.L., Jarvie, T.P., Jirage, K.B., Kim, J.B., Knight, J.R., Lanza, J.R., Leamon, J.H., Lefkowitz, S.M., Lei, M., Li, J., Lohman, K.L., Lu, H., Makhijani, V.B., McDade, K.E., McKenna, M.P., Myers, E.W., Nickerson, E., Nobile, J.R., Plant, R., Puc, B.P., Ronan, M.T., Roth, G.T., Sarkis, G.J., Simons, J.F., Simpson, J.W., Srinivasan, M., Tartaro, K.R., Tomasz, A., Vogt, K.A., Volkmer, G.A., Wang, S.H., Wang, Y., Weiner, M.P., Yu, P., Begley, R.F., Rothberg, J.M., 2005. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380. Nam, Y.-D., Jung, M.-J., Roh, S.W., Kim, M.-S., Bae, J.-W., 2011. Comparative analysis of Korean human gut microbiota by barcoded pyrosequencing. PLoS One 6, e22109. Namgung, H.-J., Park, H.-J., Cho, I.H., Choi, H.-K., Kwon, D.-Y., Shim, S.-M., Kim, Y.-S., 2010. Metabolite profiling of doenjang, fermented soybean paste, during fermentation. Journal of the Science of Food and Agriculture 90, 1926–1935. Nychas, G.J., Arkoudelos, J., 1989. Staphylococci: their role in fermented sausages. Journal of Applied Bacteriology 67, vi.
Onda, T., Yanagida, F., Uchimura, T., Tsuji, M., Ogino, S., Shinohara, T., Yokotsuka, K., 2002. Widespread distribution of the bacteriocin-producing lactic acid cocci in Miso-paste products. Journal of Applied Microbiology 92, 695–705. Onda, T., Yanagida, F., Uchimura, T., Tsuji, M., Ogino, S., Shinohara, T., Yokotsuka, K., 2003. Analysis of lactic acid bacterial flora during miso fermentation. Food Science and Technology Research 9, 17–24. Ottenwaelder, B., Kupke, T., Brecht, S., Gnau, V., Metzger, J., Jung, G., Goetz, F., 1995. Isolation and characterization of genetically engineered gallidermin and epidermin analogs. Applied and Environmental Microbiology 61, 3894–3903. Park, J.-S., Park, H.Y., Kim, D.H., Kim, D.H., Kim, H.K., 2008. Ortho-dihydroxyisoflavone derivatives from aged Doenjang (Korean fermented soypaste) and its radical scavenging activity. Bioorganic & Medicinal Chemistry Letters 18, 5006–5009. Roh, S.W., Kim, K.H., Nam, Y.D., Chang, H.W., Park, E.J., Bae, J.W., 2010. Investigation of archaeal and bacterial diversity in fermented seafood using barcoded pyrosequencing. ISME Journal 4, 1–16. Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., Lesniewski, R.A., Oakley, B.B., Parks, D.H., Robinson, C.J., Sahl, J.W., Stres, B., Thallinger, G.G., Van Horn, D.J., Weber, C.F., 2009. Introducing mothur: open-source, platformindependent, community-supported software for describing and comparing microbial communities. Applied and Environmental Microbiology 75, 7537–7541. Shin, S.Y., Kim, Y.B., Yu, T.J., 1985. Flavour improvement of soybean pastes by the addition of Bacillus licheniformis and Saccharomyces rouxii. Korean Journal of Food Science and Technology 17, 8–14. Tanasupawat, S., Thongsanit, J., Okada, S., Komagata, K., 2002. Lactic acid bacteria isolated from soy sauce mash in Thailand. Journal of General and Applied Microbiology 48, 201–209. Villar, M., Ruiz Holgado, A.P., Sanchez, J.J., Trucco, R.E., Oliver, G., 1985. Isolation and characterization of Pediococcus halophilus from salted anchovies (Engraulis anchoita). Applied and Environmental Microbiology 49, 664–666. Yoo, S.-K., Cho, W.-H., Kang, S.-M., Lee, S.-H., 1999. Isolation and identification of microorganisms in Korean traditional soybean paste and soybean sauce. Korean Journal of Applied Microbiology and Biotechnology 27, 113–117. Yoo, S.-K., Kang, S.-M., Noh, Y.-S., 2000. Quality properties on soy bean pastes made with microorganisms isolated from traditional soy bean pastes. Korean Journal of Food Science and Technology 32, 1266–1270. Yoon, M.Y., Kim, Y.J., Hwang, H.-J., 2008. Properties and safety aspects of Enterococcus faecium strains isolated from Chungkukjang, a fermented soy product. LWT- Food Science and Technology 41, 925–933. Zhao, J., Dai, X., Liu, X., Chen, H., Tang, J., Zhang, H., Chen, W., 2009. Changes in microbial community during Chinese traditional soybean paste fermentation. International Journal of Food Science and Technology 44, 2526–2530.