sign in sign up
<<
Home , Dongchimi, Kimchi, Weissella koreensis

rdeti aytpso icii hns abg ( Chinese is of types in- compestries common many A in . sodium for gredient 3% used before to are desalting (equilibrated) 2% or chloride of processing concentrations further salt Typically, require These consumption. not 1994). do in- Park brine and ingredients and flavoring be cluded not with (Cheigh prepared or can typically fermentation whether are which the on , to based added groups of was 2 food into traditional classified roughly a kimchi, vegetable o:10.1111/1750-3841.12848 doi: Introduction Cho, Hyong Joo Choi, Joo Jin Kim, Ho Breidt Kyong Frederick Lee, and Jae Barrangou, Yong Rodolphe Kim, Chung, Gun Ho Song Chang Pradas, Medina Eduardo Kyung, Hang Kyu Kimchi Watery of Ecology Microbial ute erdcinwtotpriso sprohibited is permission without reproduction Further nure oato rit(-al [email protected]). Direct U.S.A. (E-mail: Hall, 27695-7624, Breidt Schaub NC author 322 Raleigh, to Univ., Unit, inquiries State Authors Research Carolina U.S.A. Science North 27695-7624, Food 7624, NC Box SAA Allen USDA-ARS, Dan Raleigh, with Univ., 400 are Sciences, State Breidt Nutrition 98 Carolina Univ., and Bioprocessing Sejong North Food, Drive, Bar- Management, are of and Dept. Service Pradas Chung with Authors Food are and Korea. rangou and South Cho 143-747, Science Choi, Seoul, Pukyong Microbiology, Gwangjin-gu, Culinary Authors Gunja-dong, of of Korea. Dept. Dept. South with with 608-737, is Pusan Kim Univ., Author Natl. Re- Korea. Microbial Daejeon, of Biotechnology, and with Republic Padre Bioscience are Av. 305-806, of Lee Seoul, Inst. (CSIC), and Research Grasa Gwangjin-gu, Kim Korea la Authors Center/KCTC, Gunja-dong, Spain. source de Seville, Inst. 98 with 41012 Univ, 4, is Tejero Pradas Garcia Sejong Author Science, Korea. South Food 143-747, of Dept. with eeals otpclyece rmr ie h oueo the of volume the times more Varieties or vegetables. 2 exceed typically to vegetables, ( yeolmoo ( ( S2111 umte /121,Acpe /721.Ato yn is Kyung Author 2/17/2015. Accepted 9/11/2014, Submitted 20141519 MS C mul Raphanus tongbaechu) 05Isiueo odTechnologists Food of Institute 2015 hr r aydfeetkns(ehp udes ffermented of hundreds) (perhaps kinds different many are There o mrvdmtosfrietfigaddfeetaigcoeyrltdlci cdbcei pce sn high-throughput need using the species illustrate also the acid data the of lactic addition, related conditions In closely product. methods. differentiating fermentation the sequencing and of identifying and produced) for is method methods acid improved much preparation for too if from undesirable, be result can can fermentation ( safety that impact may ecology which microbial kimchi, and biochemistry in o h n 10 20 and and 10 4 at the fermentation for during similar were populations 4 atcadaei cd n eraigp o ahtmeaue u ifrne nmcoit eeaprn.Interestingly, apparent. were microbiota in differences but temperature, the each for from pH decreasing bacteria and acids acetic and lactic rcia Implication: Practical Keywords: Abstract: rdcsmyafc irboaadpoutquality. product and microbiota affect may products n ihtruhu N eunigo 6 rDNA. 16S of sequencing DNA high-throughput and 20 and 10, 4, at ( vegetables and radish unsliced ici sfretdwt ae o atbie de othe to added brine) salt (or water with fermented is kimchi) ° Cin woesalrdse)kmh,adohr.Wtr kimchi Watery others. and kimchi, radishes) small (whole sdi rdtoa chopped traditional in used ) p. icivreis including varieties, kimchi spp.) nabak ici te omntpso iciicueradish include kimchi of types common Other kimchi. ihtruhu eunig irba clg,wtr kimchi watery ecology, microbial sequencing, high-throughput h iceityadmcoileooyo iia ye fwtr ( watery of types similar 2 of ecology microbial and biochemistry The abg emnain oprdwith compared fermentations cabbage ° mul eeaaye ypaigo ifrniladslciemda ihpromnelqi chromatography, liquid high-performance media, selective and differential on plating by analyzed were C ° Proteobacterium ape fboth of samples C iciinclude kimchi hsrsac a i h nesadn fhwpoesn odtosfrfretdvegetable fermented for conditions processing how of understanding the aid may research This R hsatcehsbe otiue yU oenetepoesadterwr si h ulcdmi nteUSA. the in domain public the in is work their and employees Government US by contributed been has article This . nabak Enterobacteriaceae hlm including phylum, baechu biak and kakdugi ici(with kimchi nabak rwoecabbage whole or and cbd and (cubed) epciey,wr netgtd ape rmkmh eefermented were kimchi from Samples investigated. were respectively), , ouain a nld ahgncbcei)adqaiy(homolactic quality and bacteria) pathogenic include may populations dongchimi Brassica dongchimi. baechu Enterobacteriaceae Nabak ° nn pce,wt emnainocrigb hsogns at organism this by as occurring low as fermentation temperatures with species, inant 15 at fermentation 10 of At d 4 first the ing gasicomitatum 15 Leuconostoc and 10 at nlci cdbceilppltoswr ena 25 dongchimi at increase seen and were pH 5 populations fermentation to in of bacterial decline d acid rapid 7 lactic more A to in 2008). 3 others first and the (Park during 25 organism and dominant 5 the at showed isolates species but similar methods, 2 the methods between electrophoresis discrepancies gel showed gradient denaturing and based culture of study A that 2008). shown others has and Park 2006; baechu others others and and Lee of Cho 1984; Kwon rate 2005; and the de- (Mheen are that temperature populations evident on microbial pendent is and biochemistry, It pH, 2014). and in Jeong others reduction 2009; and others Jung and 2013; Park others 1995; others and Fleming Park 1994; and (Cheigh methods sequencing DNA high-throughput and dongchimi emnain(en n tes21) ait of variety A 2013). others and (Jeong fermentation that showed technology in included be also may ingredients vegetable mul other of variety A and radish), quartered ingredient), vegetable or main the as cabbage uigteeprmn.Teedt hwtedifferences the show data These experiment. the during ,tehmlci tg ffretto i o develop not did fermentation of stage homolactic the C, icisoe iia rnsas trends similar showed kimchi lhuhcagsfor changes Although irba clg td of study ecology microbial A h hne nmcoilppltosduring populations microbial in changes The icia io constituents. minor as kimchi ° .Asuyo h vlto fmcoilppltosduring populations microbial of evolution the of study A C. ° icifretto sltsuig1SrN sequencing rDNA 16S using isolates fermentation kimchi Corcolder, erae oerpdydrn emnainat fermentation during rapidly more decreased , emnaina 4 at fermentation emnainhv endcmne yisolate-based by documented been have fermentation o.8,N.5 2015 5, Nr. 80, Vol. ° uigteiiilsaeo emnain with fermentation, of stage initial the during C Leuconostoc esel koreensis − and 1 Leuconostoc mul Proteobacterium nabak ° Jogadohr 2013). others and (Jeong C p.and spp. ecnso citrium Leuconostoc ici otiigsie and sliced containing kimchi, ) ° ° ,with C, o 0duig44sequencing 454 using d 90 for C r ici(ihtil lcdradish). sliced thinly (with kimchi ora fFo Science Food of Journal dongchimi dongchimi pce rdmntdduring predominated species Weissella ° Jogadohr 2013). others and (Jeong C a on ob h dom- the be to found was ecnso mesenteroides Leuconostoc and dongchimi t5ad25 and 5 at ihincreasing with , Enterobacteriaceae rdmntn dur- predominating p.predominated spp. baechu ° compared C wt whole (with iciand kimchi Leuconostoc ° Cusing M1031 as

M: & Safety a c Lof µ g; Roche, -CRRCA- µ ´ values of 3.86 ): S-D-Bact- Cfor30s,and a ° K buffer, 0.5 0 6515 × M of each primer. Reac- C) Time (d) No. Reads ° µ Escherichia coli b C for 30 s, 60 ° L FastStart 10 µ m filter paper. The filter paper was µ ; Klindworth and others 2013). Sequenc- -kimchi brine, including solid particles, ´ mul L of FastStart HIFI Polymerase (5 U/ µ C followed by 30 cycles of 94 ° Ten milliliters of NA, not applicable, fresh cabbage sample. ND, not determined. No. Reads, number of DNA sequences used for analysis. C0407C0414 Sliced Sliced 4 4 7 14 6628 ND (approximate base 1061): S-D-Bact-1061-a-A-17CGAGCTGACGAC-3 (5 ing was donebarcode. The unidirectionally, PCR reactions so containedplate, 5 the 0.25 to 10 reverse ngMannheim, primer of Germany), DNA had tem- 2.5 dNTP no mix (10 mM each), and 0.4 were filteredground using with 0.2- glasstracted beads according under to liquidDNA the Isolation nitrogen. instructions Kit DNA inU.S.A.). (MoBio Hypervariable was the Laboratories, regions ex- MoBio (V3 Inc., to Poweramplified Carlsbad, V6) Soil of by Calif., the Polymerase 16s rDNAterial chain were DNA reaction using (PCR)Klindworth the from and forward total others and (2013). bac- reversequences PCR primers primers and described included barcodes, by leaderstructions and se- from were the WM designed Kecktions Center according for sequencing to 454 facility sequencing instruc- the (http://www.biotech.uiuc.edu/centers). The in- forward primerbacterial included a 16S-specific leader341 primer sequence, of barcode, starting0341-b-S-17 and the at (5’-CCTACGGGNGGCWGCAG-3’) rDNAreverse approximately and primer gene base contained the (from a leader and 16S primer sequence b c eluent, and a flowwere rate calculated based of on 0.5 the pH mL/min.the and Protonated Henderso–Hasselbalch acid concentration organic equation, data acids based using on p Bacterial 16S rDNA gene amplification and pyrosequencing Table 1–Experimental design and sequence data. SampleC0000 Sliced, Nabak TypeC0421C0430 SlicedC1003 SlicedC1007 Temp ( Sliced NA C1014 SlicedC1021 SlicedC2001 SlicedC2003 SlicedC2005 SlicedC2007 SlicedN0000 4 SlicedN0407 4 Unsliced, Dongchimi 10N0414 Unsliced 10N0421 Unsliced 10N0430 NA 21 Unsliced 10N1003 30 Unsliced 20N1007 3 Unsliced 20N1014 7 Unsliced 20N1021 14 4696 0 Unsliced 20N2001 21 4189 UnslicedN1003 4 1 ND UnslicedN2005 4 3 6895 Unsliced 7332 N2007 4 5 2396 Unsliced 8235 4 7a Unsliced 10 5782 7 10 4306 14 10 6005 21 10 30 ND 20 3 8502 20 7 8748 20 14 3305 20 21 5183 1 4792 3 8827 3828 5 4354 7 6112 3320 4283 5868 and 4.76 for lactic and acetic acids, respectively. tion conditions consisted of95 an initial denaturation for 2 min at , C × × C. ° )at ° C, as ,were ° C, but kimchi ° dongchimi mul dongchimi Le. gelidum and CFU/mL) after the kimchi preparations, and kimchi using selective spp. (Kong and others 6-7 nabak ) ecology . . . nabak kimchi were prepared with nabak Vol. 80, Nr. 5, 2015 and r mul spp. were able to grow slowly at Lactobacilli Dongchimi Le. Le. gasicomitatum and did not. The growth rates of these species dongchimi spp. and were evident during the first 3 d of fermen- Le. Nabak C to determine how changes in radish prepara- ° Lactobacilli Weissella m membrane), brine samples were injected in a high- µ Journal of Food Science kimchi. For both C before biochemical analysis and microbial DNA sampling. ° Cwhereas Total aerobic plate count, total LAB, and dextran-producing Laboratory-scale batches of In this study, a survey of the microbial ecology of 70 ° tigated in this study, ourfor work determining represents how an processing important conditionstemperature) initial (slicing, may step influence fermentation the chemistry and microbiota. 10, 20, and 30 tion and fermentation temperature affectedand both the microbiota. biochemistry Although quality factors were not directly inves- of 0.1 N sodiumcalculated as hydroxide percent to an equivalent. Aftertion dilution end and (0.2- filtra- point of pH 8.2; TA was azide, and peptone-yeastyeast (PY) extract, sucrose 20 agar gazide sucrose, (10 (respectively), and followed g by 15 incubation peptone, gMeasurements for agar/L) 5 of 1 with titratable to g 0.05% acidity 4 sodium (TA) d were at done 30 using aliquots LAB were estimatedCount by plating Agar dilutions (DifcodeMan, of Laboratories Rogosa, the Inc., and brine Detroit, on Sharpe Mich., Plate (MRS) U.S.A.), agar with 0.05% sodium − were placed in jars,indicated in which Table were 1. placedprocessed Brine for at samples traditional 4, (10 microbiological 10, mL) methods, and then were 20 collected stored at and or further slicednabak into about 0.5-green to onion, 1-cm-thick , thin and pieces for were added. Prepared materials M1032 Microbial and biochemical analyses Materials and Methods Preparation of watery kimchi and sampling species and tation; however, 2 species, Watery kimchi ( were proportional to temperature, increasing attheir 10 growth and rate 20 decreased when acid levels increased. was conducted with sliced and unsliced ( initiation of fermentation. 5 acids, samples were4.2 run on mm, an Phenomenex,with Luna Torrance, a C18 Calif., UV column U.S.A.)with (250 0.05 detector and M mm analyzed (Waters monopotassium phosphate 2487, adjusted to at pH 210 2.8 as nm) the run at 40 dex detector (RI-410,0.6 Bio-Rad). mL/min with The aacids samples 0.01 concentrations were M were eluted potassium also sulfate measured at by solution. HPLC. Organic For organic performance liquid chromatography (HPLC) systemysis for of the sugars, anal- anddone organic acids. by Sugar HPLC and7.8 using ethanol mm, Aminex analyses Bio-Rad, were HPX-87C Hercules, column Calif., (300 U.S.A.) mm and a refractive in- trimmed, and cut into quarters along the long axis for with 1.5 kg of distilled water,green 1 onion, kg 5 of g radish, of 50.36 garlic, andlength g 3 and of g salt, with of 10 a ginger. g Radish diameter of (17 of to 8 23 to cm in 10 cm) was washed with water, 3.6-L glass containersused with in lids. this Ingredients studySouth were Korea, for in purchased 2013. watery from Each a kimchi batch local of watery market kimchi in was Seoul, prepared 2005). Both groups had high numbers (10 the predominant species forperiod. the A remainder study ofmedia the of 90 the showed d ecology changes sampling (LAB) of in isolates of the number of

M: Food Microbiology & Safety h pnsymbols. the open and the symbols, filled the by cdt B r hw,wt h aafr4(ice) 0(rage) and (triangles), 10 (circles), 4 for data the with shown, 20 are (B) acidity ihcso yhnsrpst xrc eetdOUpopulations OTU selected extract to done scripts was python analysis data custom ma- Additional with distance QIIME. by UniFrac Principal visualized the and QIIME. on trices performed in was (UPGMA) analysis mean component group jackknifed arithmetic pair with UniFrac-based unweighted using method A constructed 2006). was cluster others sam- hierarchical diver- of and levels beta pairs (Lozupone all identity For between ples later. 99% determined were described and distances as 97% UniFrac sity, 2011), default others the and at (Kuczynski QI- scripts using python (http://rdp.cme.msu.edu/) Green- classifier IME the BLAST using or the determined with RDP 13_5) was version (http://greengenes.lbl.gov, OTU database reads. each genes the units for among identity similarity taxonomic barcodes sequence The Operational by the identified removed. and were barcode, pa- (OTUs) sequences by QIIME primer binned score default were and quality sequences using sequencing The length 454 scripts 25). and (minimum program including scores Python rameters, quality edited pro- the of on were initially relating pipeline based were Files prepared QIIME Sequences identifiers. was the (http://qiime.org/index.html). sample file with to sequenc- mapping cessed data DNA A barcode score) files. sequence (quality .qual analysis and ing format) (FASTA .fna iue1Pyiceia aa(,1,ad20 and 10, (4, data 1–Physiochemical Figure euniganalysis Sequencing for Center Keck sequencing. for WM U.S.A.) platform Ill., 454 the (Chicago, at Genomics Functional Laboratory and Comparative Biotech submitted then and Carver were samples, the products 3 to into PCR concentrations barcoded equimolar The at on mixed reader. U.S.A.) plate Calif., 96-well Hayward, a (Biotium, kit ds- AccuClear Quantification using DNA DNA U.S.A.). quantified MinE- were Calif., a amplicons Valencia, using of (Qiagen, extracted concentrations Kit and Extraction gel Gel agarose lute 1% gel a by in purified electrophoresis were nucleotides 800 approximately with products 72 ( kimchi Watery aafie bandfo h avrLbrtr included Laboratory Carver the from obtained files Data ° ° surs.Dt o the for Data (squares). C

pH 72 of extension final a and s, 60 for C 3.0 3.5 4.0 4.5 5.0 5.5 6.0 0 Nabak 5 and dongchimi 10 nabak Dongchimi Time (Days) Time sie)smlsaerpeetdby represented are samples (sliced) 15 usie)smlsaerepresented are samples (unsliced) ° ) h H()adtitratable and (A) pH The C). ° clg . . . ecology ) o i.TePCR The min. 7 for C 20 20 10 4 o C o o 25 C C 30 = usie)smlsaerpeetdb h le ybl,adthe symbols. and open symbols, the by filled represented the are samples by (sliced) represented are samples (unsliced) h aafr4(ice) 0(rage) n 20 and with (triangles), shown, 10 are (circles), (B) 4 concentrations for 20 acid data and acetic 10, the (4, and data (A) acid concentrations acetic and 2–Lactic Figure Results numbers: Accession text. the ( in database GenBank indicated NCBI as analysis BLAST for h ubr fDAsqecsfrte2tpsof types 2 the for sequences dongchimi DNA of numbers the rprtos hstedwsas paetfraei cd u was but both acid, for acetic for 3.3 apparent approximately also was was pH trend This the preparations. although 2), (Figure chi of because difference less tempera- 20 or with unit varied pH ( pH 0.2 type a in showed decrease only the but ture, kimchi, of types both oprdwt 7m ( mM 17 with compared umr ftesmln ie n eprtrs swl as well as temperatures, and times sampling the of summary A ° Acetic Acid (mM) Lactic Acid (mM) C, dongchimi 20 25 30 10 15 10 12 14 nabak 0 5 0 2 4 6 8 usie)and (unsliced) 01 02 30 25 20 15 10 5 0 o.8,N.5 2015 5, Nr. 80, Vol. icihd2 M( mM 26 had kimchi oprdwith compared nabak ± ceso ubr ple for applied numbers accession . M atcai for acid lactic mM) 0.2 sie) speetdi al .For 1. Table in presented is (sliced), Time (Days) Time r nabak ora fFo Science Food of Journal ± ,a hw nFgr .At 1. Figure in shown as ), .3m)lci cda d 7 at acid lactic mM) 0.03 ° (squares). C ° ) h atcacid lactic The C). dongchimi 20 10 4 o ). C o o mul C C Dongchimi -kimchi, M1033 nabak kim-

M: Food Microbiology & Safety ± kim- dongchimi 72.5 bp for the and ± C). The PCA (A) and MRS ° 4 degrees C 10 degrees C 20 degrees C nabak C C o o C o kimchi resulted in 3000 and 4 10 20 , and 720.6 Time (Days) (unsliced) samples are represented by the dongchimi (sliced) samples are represented by the open samples, and 15341 representative OTUs dongchimi and nabak Dongchimi nabak dongchimi samples. The average sequence length was 722.2 0 5 10 15 20 25 30 3 9 8 7 6 5 4 6 5 4 3 9 8 7

nabak C (squares).

10 ° CFU/ml) 10 10 MRS Cell Counts (Log Counts Cell MRS kimchi. CFU/ml) PCA Data (Log Data PCA Sequencing of Bacterial population profiles between Figure 4–Microbial cell counts (4, 10, and 20 were achieved when the calculated protonated acid(data concentrations not shown) were aroundfor 3.0 lactic (C2003) acid, and and 3.5 the3.3 mM protonated (C2003) acetic (C1014) and acid 8.3 concentrations mM were (C1014). nabak chi preparations are shown in Figure 5. Comparisons at each (B) cell count data areand shown, 20 with the data for 4 (circles), 10 (triangles), symbols. 9000 qualified readsreads/sample. for There each were a sample,resentative total set with of of an 19988 OTUs sequences defined averageidentity) by in for of the the the QIIME 5657 rep- softwarefor (for 97% the 68.3 base pair (bp) for filled symbols, and the 8 C C C o o o 4 10 20 fermen- C, and the ° nabak C). The glucose (A) and kimchi. The diffusion ° ) ecology . . . Vol. 80, Nr. 5, 2015 (unsliced) samples are represented r , as indicated by both the acid (sliced) samples are represented by dongchimi Dongchimi Time (Days) nabak and Dongchimi dongchimi had LAB cell counts that exceeded 10 had higher free sugar concentrations than compared to nabak C. In both cases, the maximum CFU/mL values Nabak ° nabak C (squares). ° nabak 0 5 10 15 20 25 30 for most of the time. Journal of Food Science

0 5 0

10 20 15 10 20 40 30

Fructose (mM) Fructose Glucose (mM) Glucose The total aerobic plate count and LAB plate count data were similar (Figure 4),substantially and from the the MRSpling PY data times, (data cell only not count shown). data For all did sam- not differ and B), where dongchimi production data (Figure 2A and B) and the sugar data (Figure 3A rate for nutrients andtation sugars compared may be to higher the in the M1034 angles), and 20 by the filled symbols,the and open the symbols. Figure 3–Sugar concentrations (4,fructose (B) 10, concentrations are and shown, with 20 the data for 4 (circles), 10 (tri- 14 d sample at 10 CFU/mL, which was recorded for the 3 d samples at 20 apparent for less pronounced, with only atypes 2 of to kimchi. 5 The mM difference release between of the nutrients 2 by slicing the radish was Watery kimchi (

M: Food Microbiology & Safety otycnitdo eune dnie nya ersnaieof remainder representative The as only samples. identified nabak sequences of for same consisted found the mostly by sequences represented database primarily were 3 and analysis, BLAST the order nabak as: level genus the OTUs), to identified was remainder the database sequence For rDNA 13_5). identity algorithm level (version Megablast 99% Greengenes Greengenes the the and using analysis subjected BLASTN and unpublished) to Breidt, (F. script ex- python were a they using sequences, tracted these the identity by phylogenetic ( the fermentations vestigate the fermentation dominated of level) day order 5th the to only order fied the of representative sequences OTU 20 the In kimchi. as level genus family the of members other family the of representative Enterobacteriaceae sequences that showed temperature ( kimchi Watery piaiyt pcfi eune ntedtbs) n 0OTUs 20 family and the database), as the only in identified sequences were specific 3 to (primarily Lactobacillales the for temperature same in abundance genus total the greater from sequences Similarly, kimchi. Lactobacillales ape,14o 9 eune eeietfidol othe to only identified were sequences 195 of 104 samples, Leuconostoc Tsrmie dnie oteodrlvlonly level order the to identified remained OTUs Leuconostoc eerdcdin reduced were Nabak ° hs eune eentfrhrcasfidby classified further not were sequences These . ape fboth of samples C 2 Ts,or OTUs), (22 dongchimi and dongchimi > a rae bnac in abundance greater a had nabak 0,Fgr n ) ofrhrin- further To 6). and 5 Figure 60%, Leuconostocaceae Dongchimi ape N07,4 f151 of 42 (N2007), samples d 7 emnain oprdt the to compared fermentations nabak emnainsml,although sample, fermentation Lactococcus nabak oprdwith compared clg . . . ecology ) and Lactobacillales o dnie othe to identified not Leuconostoc 7OU) o the For OTUs). (7 h aoiyof majority The . dongchimi dongchimi dongchimi eein were kimchi, (identi- (51 PM reaayi niae la ifrnefrteunfer- the for difference clear a indicated analysis tree UPGMA 20 and 10 at similar 4 for between seen difference was a Interestingly, 7. Figure Discussion fermented the to both compared for products 8A) (Figure ingredients fresh mented irba clg a eetybe pae rmtaiinlmi- the traditional products, from updated vegetable been fermented recently has many ecology For microbial popularity. in ing rba tde yavreyo oeua ehius including techniques, molecular of variety a by studies crobial lhuhteewsn la rn o l ape,priual for particularly samples, all for 10 trend and time, clear the (C0000 fermentation no was samples with there decreased radish although diversity fresh great- general, The the In 2. with Table N0000). seen in was shown diversity are est values Shannon and Simpson, ing OTUs), (14 family the rtmeaueo fermentation. of temperature compo- or either for principal 8B) or (Figure tree analysis UPGMA nent by either samples of clustering o ra itr ftecagsi irba clg during ecology microbial phyla in the changes of the representatives of fermentation, picture broad a For rdtoal emne ntrl eeal rdcsaegrow- are products vegetable “natural” fermented Traditionally h siaosfrbceilapadvriy nldn Chao1, including diversity, alpha bacterial for estimators The Enterobacteriaceae ° ape o both for samples C Lactobacillaceae o.8,N.5 2015 5, Nr. 80, Vol. Leuconostoc ° ,btpten o h hnigppltoswere populations changing the for patterns but C, nabak )and ° bnac for abundance for microbiota of data abundance 5–Relative Figure ahpplto r sidctdi h legend. the in indicated as are population each 20 (C) and 10, C. and nabak 2 Ts,and OTUs), (29 2 Ts,a ela genera as well as OTUs), (28 dongchimi nabak sie)Kmh.Terelative The Kimchi. (sliced) r ora fFo Science Food of Journal ° nabak nabak r hw.Teclrcdsfor codes color The shown. are C icuigLB r hw in shown are LAB) (including and oee,teewsn clear no was there however, ; emnain t()4 (B) 4, (A) at fermentations dongchimi oprdwith compared Lactococcus Proteobacteria nabak lseigby Clustering . and 1 OTUs). (10 Lactobacillus dongchimi dongchimi (includ- M1035

M: Food Microbiology & Safety fermentations at kimchi, a high- C are shown. The color ° nabak dongchimi after a similar time of fer- and samples (Figure 3). Sugar concen- (unsliced) Kimchi. The relative dongchimi dongchimi nabak kimchi had 26 mM lactic acid compared Figure 6–Relative abundance data for dongchimi abundance data for (A) 4, (B) 10, and (C)codes for 20 each population are aslegend. indicated in the nabak C, the ° For further analysis of At 20 with 17 mM lacticmentation acid (7 d), for but3.3. surprisingly It these is samples both possibledifferent had that preparation a buffering methods pH and in of buffering rates the compounds at brine diffused which was into acidstion affected between the and sugar by brine. other concentration the and Similarly, fermentation thea time more rela- indicated rapid diffusion for tration changes did not show aof consistent pattern free because sugars diffusion fromwere the concurrent. radish Metabolism and of consumption theter sugar of the continued time sugar to when by occur the LAB as maximum af- cell indicated concentration was recorded, by(Figure 3). the The continued protonatedwere lactic change presumably and responsible in acetic forand acid sugar preventing the subsequent further concentrations decline concentration cell inwas cell division still numbers present of at LAB, because these sugar time-points. throughput 16S rDNAknown sequencing to was have used. similara Because 16S 454 LAB sequences pyrosequencing are (Singhgenerate sequencing and strategy 700 others was tonext 2009), used generation 800 that sequencing could bp(Quail technologies and generate or shorter others greaterdiscriminate reads 2012), sequencing closely which reads. related would decrease Other LAB the species. ability PCR to primers were C. ° Pesudomon- kimchi that , C) the com- ° dongchimi ) ecology . . . and Enterobacteriaceae C, although they had a similar C(Figure1and2)gavediffer- Vol. 80, Nr. 5, 2015 ° ° ) and other epiphytic bacteria r nabak Dongchimi ,suchas and Firmicutes kimchi, which differ by slicing method for Nabak Proteobacteria dongchimi C samples of both ° Journal of Food Science and , and others may be affected by relatively small changes in C, differences in microbiota were less apparent that at 4 ° nabak M1036 These data indicate thatpetition at between colder LAB temperatures (4 ( biochemistry. It is interestinglactic that and acetic similar acids, biochemical and pHent values at results for 4 for the microbiota20 (Figure 5A, 6A, and 7A). At 10 and of the main (radish) ingredientin showed their some microbiota for interesting samples differences at 4 hya and others 2007; Kim andHowever, often others overlooked 2012; in Jung these and studies others ising 2014). the conditions effect of and process- ingredientsinfluence on both microbial the ecology, quality which and may safety of these products. Our study various types of kimchiPark 1994; and Lee cabbage and fermentations others 2005; (Cheigh Cho and and others 2006; Plengvid- Watery kimchi ( tic fermentation can resultproduct. in a higher quality, lightly fermented 4and10 the homolactic stageby of the fermentation end of was the not experiment. fully The developed delayed onset of homolac- adaceae environment brought about by slicingthe compared radish with vegetable not material, slicing rates possibly of due the to competing the organisms. slower It growth is also apparent that for the in the phylum

M: Food Microbiology & Safety ybl)and symbols) triangles. ape t4() 0() n 20 and (B), 10 (A), 4 at samples and triangles) ward le n h cr lto rnia opnn nlss()aeson ,gensurs 4 squares; green d, 30% 0 to shown. orange 100% are of (B) level analysis confidence continuous component indicates principal nodes of of plot color where score (A) the tree and UPGMA The blue, samples. kimchi watery 8–Clustering Figure ( kimchi Watery iue7Cagsi hl o aeykmh ape.The samples. kimchi watery for phyla in 7–Changes Figure C B A

Relative Abundance (%) Relative Abundance (%) Relative Abundance (%) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 dongchimi 0 0 0 Nabak Proteobacteria 3 Proteobacteria Proteobacteria Firmicutes Firmicutes 5 fildsymbols). (filled 1 and 6 ( ( nabak dongchimi 10 ° C emnain o both for fermentations (C) C Dongchimi ( ( nabak dongchimi ) 2 dwwr rage)aesonfor shown are triangles) (downward Time (d) Time (d) Time (d) 9 ) ) 15 ) 12 3 clg . . . ecology ) 20 15 4 25 Firmicutes 18 nabak 30 5 21 (open (up- efudrltvl e eune ersnaieo h genus the of representative sequences few relatively found we fermentation. the was during fermentations species numbers se- from in in DNA and decline extracellular present that PCR of nuclease degradation by the that amplified for likely responsible been that is have species of It could cells quenced. it dead the decline, from present in still were was DNA If 6). and the For bac- 2007). others for and accuracy (Wang level classification genus 20 greater the at by or shown species 85% been terial give has region to This variable analysis 2013). covering the others on fragment and (based a worth 5 amplify to to 3 bac- chosen regions domain were the and of coverage teria, phylogenetic optimum for selected fseiswt iia 6 eune.Frhrrsac a also may research Further sequences. 16S similar be with with may communities microbial species approach define of metagenomics precisely more a developed to way but been best the 2009), have others LAB dif- for and of methods (Singh species of variety related A closely 2012). ferentiating others and Kim 2007; others (order heterolactic Leuconostocaceae) identified have kimchi as from fermentations isolates cultures vegetable isolated related with and studies Previous analysis. our ( level order or or family the to OTUs some identify to Weissella ai sulkl eas ftedciei pce bevdduring observed species the in sce- of decline time-course This the the from of 2007). amplified because others unlikely was and is that (Plengvidhya nario abun- DNA cells relative by nonviable the biased or on be dead may data OTUs that of is dance fermentations vegetable in ogy appar- was the level in sequences family database. uncharacterized or matching OTUs of order because ently the beyond sequences identify (order level order the beyond tobacillales OTUs of identification obtain nareetwt rvosrpr Jogadohr 2013), others and (Jeong report previous a with agreement In n rwako sn N-ae ehd o irba ecol- microbial for methods DNA-based using of drawback One Lactobacillales ° ape for samples C Lactobacillales oee,te70b 6 eune eeol sufficient only were sequences 16S 700-bp the however, , o h aoiyo eune.Telmtdaiiyto ability limited The sequences. of majority the for ) .mesenteroides L. o.8,N.5 2015 5, Nr. 80, Vol. epciey t9%iett,wihwsue for used was which identity, 97% at respectively) , n ooatcioae as isolates homolactic and ° nabak ,rdtinls 10 triangles; red C, MenadKo 94 lnvdy and Plengvidhya 1984; Kwon and (Mheen ) nabak and , .citrium L. dongchimi and .coli E. r dongchimi ora fFo Science Food of Journal ,and ° 6 DApstos Klind- positions; rDNA 16S ,however,wewereunableto ,bu rage;20 triangles; blue C, Weissella emnain Fgr 5 (Figure fermentations atbclu plantarum Lactobacillus Leuconostocaceae pce (family species ° ,orange C, M1037 nsilico in Lac-

M: Food Microbiology & Safety ). levels of nabak kimchi at different temperatures. Korean 2 in Korean fermentation based on 16S rRNA sequence. J Korean Soc Appl Biol Chem , with a focus on discrimination between closely related species: a review. Mulkimchi Lactobacillus Soc Food Sci Nutr 34(5):707–14 ( rRNA sequences into the new bacterial . Appl Environ Microbiol 73(16):5261–7. real-time PCR for enumerationJ of Microbiol total (Korea) bacterial, 47(6):682–5. archaeal, and populations inradish kimchi. kimchi, at the early and mid-phase fermentation. Foodin Sci Biotechnol fermentations 17(4): by 892–4. methods. Appl Environ Microbiol 73(23):7697–702. HP, Gu Y. 2012.Torrent, Pacific A Biosciences tale and of Illumina three MiSeq next sequencers. BMC generation Genomics sequencingfor 13:341–54. platforms: comparison of Ion (Korean fermented vegetable products). Crit Rev Food Sci Nutra 34(2):175–203. fermented cabbage product. FEMS Microbiol Lett 257:262–7. editors. Biotechnology, Vol. 9. Enzymes,Publishers, biomass, Inc. food p and 629–61. feed. 2nd ed. Newchanges York: VCH during fermentationMicrobiol of 164:46–53. dongchimi, traditional Korean watery kimchi.for Int industrial J kimchi production. Food Appl10.1007/s00253-014-5513-1 Microbiol Biotechnol. 2014 Mar;98(6):2385–93. doi: during 55(6):787–92. 2013. Evaluation ofnext-generation general sequencing-based 16S diversitydoi:10.1093/nar/gks808. ribosomal studies. Nucelic RNA Acids gene Res PCRing 41(1), primers QIIME e1, for to 1–11; rent classical protocols analyze and in 16S10.1002/0471250953.bi1007s36. bioinformatics. rRNA New York, gene N.Y.: John sequences Wiley from µflora microbial using Sons. denaturing Ch gradient communities. gel 10:10.7; electrophoresis. Cur- Int doi: J Food Microbiol 102(2):143–50. community diversity in a phylogenetic context.7-371. BMC Bioinf 7:371. doi:10.1186/1471-2105- tation. Korean J Food Sci Technol 16(4):443–50. LWT-Food Sci Technol 42:448–57. lactic acid bacterial growth, and CO Wang Q, Garrity GM, Tiedje JM, Cole JR. 2007. Naive Baysian classified for rapid assignment of Park EJ, Chang HW, Kim KH, Nam YD, Roh SW, Bae JW. 2009.Park Application SJ, of quantitative Chang JH, Cha SK, Moon GS. 2008.Plengvidhya V, Microbial analysis Breidt of F, dongchimi, Lu Korean Z, watery Fleming HP.Quail 2007. MA, DNA fingerprinting Smith M, of lactic Coupland acid P, bacteria Otto TD, Harris SR,Singh Connor S, TR, Goswami Bertoni P, A, Shigh Swerdlow R, Heller KJ. 2009. Application of molecular identification tools the Spanish Government (MECD) forsupport the postdoctoral for fellowship Dr. E. Medina-Pradas. References Cheigh HS, Park KY. 1994. Biochemical, microbiological,Cho J, and Lee nutritional D, aspects Yang C, of J, kimchi KimFleming J, Han HP, H. Kyung 2006. Microbial KH, population Breidt dynamics of F. kimchi, 1995. Vegetable fermentations.Jeong In SH, Rehm Jung HJ, JY, Lee Reed SH, G, Jin HM, JeonJung CO. JY, 2013. Lee Microbial SH, succession Jeon CO. and 2014. metabolite Kimchi microflora: history, currentKim status, B, Seo and WT, perspectives Kim MG, YunHD, Cho KM. 2012. Metagenomic lactic acid bacterial diversity Klindworth A, Pruesse E, Schweer T, PepliesKuczynski J, J, Quast Stombaugh J, C, Walters Horn WA, Gonzalez M, A, Glockner Caporaso JG, FO. KnightLee JS, R. Heo GY, 2011. Lee JW, Us- Oh YJ, Park JA, Park YH,Lozupone Pyun C, YR, Ahn Hamady JS. M, 2005. Knight Analysis of R. kimchi 2006. UniFrac—an online tool forMheen comparing TI, microbial Kwon TW. 1984. Effect of temperature and salt concentration on kimchi fermen- Kong CS, Seo JO, Bak SS, Rhee SH, Park KY. 2005. Standardization of manufacturing methods, c mul Shannon . c dongchimi ) ecology . . . = Simpson Vol. 80, Nr. 5, 2015 NNNN r c ,N Dongchimi nabak Chao1 = and kimchi fermentations and vegetable fer- NNNN b Nabak OTUs dongchimi and Journal of Food Science a nabak This work was carried out as part of the international collab- Number of OTUs, based on 1000 random reads for each coded sample. Diversity indices, as described in Materials and Methods. Sample, Coded samples: C ingredients (garlic, ginger, andkimchi and other) related on products. the microbiota of mentations in general and to investigate the effects of other food Packers Intl. Inc., Washington, D.C., U.S.A. The authors thank orative R&D programFood funded Cluster (2013), by and supported the in part Agency by for a grant Korea from Pickle Natl. M1038 Acknowledgments be needed toin characterize the consistency of microbial changes b c N2003N2005N2007a 109.4 82.8 69.2 265.1 161.3 156.9 0.799 0.775 0.614 3.63 3.37 2.53 N1003N1007N1014N1021N2001 99.5 79.5 117.5 108.2 239.5 62.6 183.5 268.1 248.7 146.0 0.876 0.620 0.801 0.821 0.862 3.90 2.65 3.93 3.81 3.63 N0407N0414N0421N0430 125.7 121.7 85.3 114.3 293.1 296.1 211.6 224.1 0.865 0.840 0.647 0.815 4.17 4.08 2.87 3.77 C2001C2003C2005N0000 108.1 104.3 65.8 159.7 271.2 250.2 173.1 353.2 0.802 0.746 0.463 0.877 3.75 3.34 2.08 4.59 C0421C0430C1007C1014C1021 86.3 94.7 79.9 69.5 167.9 141.9 202.9 210.7 183.0 375.4 0.786 0.726 0.542 0.382 0.785 3.33 3.37 2.37 1.84 4.05 Sample C0000C0407 167.3 97.6 378.2 236.0 0.874 0.802 4.71 3.50 Table 2–Species diversityquences randomly estimators chosen calculated from the from reads 1000 of kimchi se- samples. Watery kimchi (

M: Food Microbiology & Safety