Suitability of Different PCR-DGGE Primer Sets for the Monitoring of Lactic Acid Bacteria in Wine

Suitbility of Different PCR-DGGE Primer Sets for the Monitoring of Lctic Acid Bcteri in Wine S. Myrhofer*, R. Filipp, D. Lehner, C. Reiterich, W. Kneifel, K.J. Domig BOKU - University of Nturl Resources nd Life Sciences Vienn, Deprtment of Food Science nd Technology, Institute of Food Science, Muthgsse 18, A-1190 Vienn, Austri Submitted for publiction: October 2013 Accepted for publiction: My 2014 Key words: Mlolctic fermenttion, lctic cid bcteri, wine, PCR-DGGE, primer-set Lctic cid bcteri (LAB) ply dul role in winemking s they re the min effectors of mlolctic fermenttion, but some members cn lso cuse wine spoilge. PCR-DGGE hs proved to be quick tool to study the LAB community nd their fluctution in wine. For detecting wine-ssocited LAB by PCR- DGGE, the primer sets WLAB1/WLAB2 GC, WBAC1/WBAC2 GC, Lc1/Lc1o/Lc2 GC, 341f GC /518r nd rpob1/ rpob1o/rpob2 GC were tested nd evluted in this study. The primer systems were ssessed by the seprtion of LAB reference strins on DGGE gels nd by ttributing the resulting mplicons to defined species. Subsequently, the detection of LAB in wine smples nd enrichments thereof ws compred. While the primer systems WBAC1/WBAC2 GC nd 341f GC /518r were not pproprite, the Lc1/Lc1o/Lc2 GC primer set performed well. However, multiple bnds complicted the evlution. The rpob1/rpob1o/rpob2 GC set seemed to be promising for the detection of LAB in wine, lthough further improvements in terms of the detection limit need to be done. Due to the pronounced sensitivity nd the sufficient discrimintion of LAB t species level, the WLAB1/WLAB2 GC primer system ws found to be most suitble for studying the occurrence of LAB in wine. INTRODUCTION Winemking is complex microbil process in which primrily yests, but lso lctic cid bcteri (LAB), ply pivotl roles (Lonvud-Funel, 1999). Mlolctic fermenttion (MLF) cn occur t the end of the lcoholic fermenttion conducted by yests (Lonvud-Funel, 1999). This fermenttion is usully desirble in most of the red wines, some white cultivrs, including Chrdonny, some sprkling wines nd lso in cool-climte Riesling wines (Lerm et l., 2010; Knoll et l., 2012). MLF is the bcteril conversion of L-mlic cid to L-lctic cid nd CO 2 (Bousbours & Kunkee, 1971) nd exerts significnt influences on wine cidity, flvour nd microbiologicl stbility (Moreno-Arribs & Polo, 2005). The min effectors of MLF re LAB. Of these, Oenococcus oeni is the species minly responsible for MLF (Dvis et l., 1986), s it is the species tht hs ccommodted the best to the difficult fermenttion conditions, such s low ph vlues nd high ethnol concentrtions (Wibowo et l., 1988). Due to its prticulr role, this species is commonly used s strter culture to promote MLF (Mills et l., 2005). Furthermore, it is not very frequently reported to be ssocited with off flvours like voltile cidity nd mousiness, spoilge like ropiness, or the formtion of undesirble metbolites such s ethyl crbmte nd biogenic mines, which cn be cused by other wine-relted LAB (Mills et l., 2005). Owing to these observtions there is need to control MLF to enhnce the positive ttributes or to reduce potentil negtive impcts on the prticulr wine (Mills et l., 2005). Trditionl culture-bsed techniques re often used to detect LAB in wine smples (Cho et l., 2011) but, especilly in cse of the min effector O. oeni, up to 14 dys re required to yield results. Such long cultivtion periods, however, do not llow the crrying out of possible oenologicl prevention or opertion in wine production (Pinzni et l., 2004). Thus, severl culture-independent methods (e.g. PCR-DGGE, qpcr) hve been developed becuse they overcome the problems described bove. PCR-denturing grdient gel electrophoresis (DGGE) is commonly used culture-independent fingerprinting technique for the rpid nlysis of microbil communities nd hs been used to nlyse LAB in food (Cocolin et l., 2001). This technique is pplied to seprte mixture of PCR mplicons of the sme size but of different sequences (Ercolini, 2004). Double-strnded PCR mplicons in the gel re subjected to n incresingly denturing environment. The migrtion is stopped when the DNA frgments re completely dentured (Renouf et l., 2007), yielding ptterns tht visulise the genetic diversity of the investigted microbil community (Ercolini, 2004). Implementing new method in the own lbortory lwys requires intensive literture reserch. In terms of PCR-DGGE pplied to study wine LAB microbiot, severl *Corresponding uthor: sigrid.myrhofer@boku.c.t Other uthors: konrd.domig@boku.c.t; wolfgng.kneifel@boku.c.t Acknowledgements: This study ws funded by the Austrin Economic Chmbers (WKO; Wirtschftskmmerpreis 2009) 185

PCR-DGGE for LAB Detection 186 different primer sets nd PCR conditions hve lredy been introduced by reserchers. Of these, five primer sets were shortlisted, s they seemed to be pproprite (Lopez et l., 2003; Rntsiou et l., 2004; Endo & Okd, 2005; Be et l., 2006; Renouf et l., 2006; Spno et l., 2007). The objective of this study ws to prtly modify nd extensively test nd evlute these primer systems regrding their suitbility to monitor LAB in wine. The results of this study cn be consulted to investigte the presence of LAB in wine by PCR-DGGE. MATERIALS AND METHODS Strins nd growth conditions Bcteril reference strins nd their corresponding growth conditions (medium nd temperture) used in this study re listed in Tble 1. The strins were grown nerobiclly (85% N 2, 10% CO 2, 5% H 2 ) using MACS VA 500 microerophilic worksttion (Don Whitley Scientific, Shipley, U.K.). In ddition, further LAB (Enterococcus feclis LMG 7937 T, Enterococcus fecium LMG 11423 T ; Streptococcus thermophilus LMG 6897 T nd Tetrgenococcus muriticus LMG 18498 T ) were included in the tests in order to evlute the specificity of primer Lc1o. Except for Tetrgenococcus muriticus, ll strins were grown on MRS medium (demn, Rogos, Shrpe; Merck, Drmstdt, Germny): Streptococcus thermophilus nerobiclly t 37 C nd Enterococcus feclis s well s Enterococcus fecium TABLE 1 LAB reference strins nd growth conditions. Genus Species Subspecies Source Growth conditions Lctobcillus brevis - LMG 6906 T MRS, 30 C Lctobcillus buchneri - LMG 6892 T MRS, 37 C Lctobcillus csei - LMG 6904 T MRS, 30 C Lctobcillus collinoides - LMG 9194 T MRS, 30 C Lctobcillus coryniformis torquens LMG 9197 T MRS, 30 C Lctobcillus curvtus curvtus LMG 9198 T MRS, 30 C Lctobcillus delbrueckii delbrueckii LMG 6412 T MRS, 37 C Lctobcillus frciminis - LMG 9200 T MRS, 30 C Lctobcillus fermentum - LMG 6902 T MRS, 37 C Lctobcillus fructivorns - LMG 9201 T MRS, 30 C Lctobcillus hilgrdii - LMG 6895 T MRS, 30 C Lctobcillus lindneri - LMG 14528 T MRS, 30 C Lctobcillus mli - LMG 6899 T MRS, 30 C Lctobcillus ngeli - LMG 21593 T MRS, 37 C Lctobcillus prcsei prcsei LMG 13087 T MRS, 30 C Lctobcillus pentosus - LMG 10755 T MRS, 30 C Lctobcillus plntrum - LMG 6907 T MRS, 30 C Lctobcillus rhmnosus - LMG 6400 T MRS, 37 C Lctobcillus zee - LMG 17315 T MRS, 37 C Lctococcus lctis lctis LMG 6890 T MRS, 30 C Leuconostoc mesenteroides mesenteroides LMG 6893 T MRS, 30 C Oenococcus oeni - LMG 9851 T MLO, 30 C Pediococcus cidilctici - LMG 11384 T MRS, 30 C Pediococcus dmnosus - LMG 11484 T MRS, 30 C Pediococcus inopintus - LMG 11409 T MRS, 30 C Pediococcus prvulus - LMG 11486 T MRS, 30 C Pediococcus pentosceus - LMG 11488 T MRS, 30 C Weissell confus - LMG 9497 T MRS, 30 C Weissell prmesenteroides - LMG 9852 T MRS, 30 C LMG: BCCM/LMG Bcteri Collection, Ghent University, Ghent, Belgium

187 PCR-DGGE for LAB Detection erobiclly t 37 C. Tetrgenococcus muriticus ws cultivted on GYP sodium cette minerl slts medium with 5% sodium chloride (BCCM/LMG Bcteri Collection, Ghent University, Ghent, Belgium; medium 244) under erobic conditions t 30 C. Isoltes of the ccompnying bcteril flor (Bcillus cogulns MSB 29W, Gluconobcter jponicus MSB 32W, Gluconobcter oxydns MSB 107W nd Acetobcter ceti MSB 109W) grown nerobiclly on MLO medium (medium for Leuconostoc oeni; Germn Collection of Microorgnisms nd Cell Cultures (DSMZ); medium 59) t 30 C were lso included. Wine smples nd corresponding microbil enrichment cultures In ddition to the reference strins described bove, wine smples contining n unknown vriety of microorgnisms were investigted. However, s the detection limit of PCR- DGGE is round 10 4 cells/ml or even higher (Andorrà et l., 2008), enrichment cultures of wine smples were lso used for this investigtion. For this purpose, 1 ml of wine smple ws inoculted in 9 ml MRS nd MLO medium nd incubted for seven dys t 30 C under nerobic conditions to enhnce nturlly occurring wine LAB. Initilly, the DNA of 16 wine smples ws isolted for the investigtion. As the first PCR-DGGE results of these smples displyed no or only fint bnds, their enrichments were used minly to compre the performnce of the different primer systems. DNA extrction DNA extrction from 2 ml of pure or enriched cultures ws performed with the Archive Pure DNA Yest & Grm+ Kit (5 Prime, Hmburg, Germny), ccording to the mnufcturer s instructions. The DNA from the wine smple ws extrcted directly using protocol described by Renouf et l. (2009), with minor modifictions. In brief, microbil cells were collected from 50 ml of wine by centrifugtion (4 500 x g, 15 min, 4 C) nd the pellet ws wshed in 600 µl TE buffer (10 mm Tris, 1 mm EDTA). After the next centrifugtion step (10 000 x g, 7 min, 4 C), the superntnt ws discrded nd the pellet resuspended in 300 µl TE buffer. Furthermore, 300 µl of sterile glss beds were dded nd the smples were vortexed for 10 min t 4 C. The superntnt ws mixed with 300 µl cell lysis solution (5 Prime). Subsequently, 200 µl of protein precipittion solution (5 Prime) were dded nd mixed. Precipittion of cellulr frgments ws done on ice for 5 min, followed by centrifugtion step t 10 000 x g for 3 min t 4 C. The superntnt ws then trnsferred to new 1.5 ml micro-centrifuge tube lredy contining 100 µl of 10% polyvinyl-pyrrolidone solution (PVP, Sigm-Aldrich, St. Louis, Missouri, USA) for the elimintion of tnnins. After vortexing nd further centrifugtion (10 000 x g, 10 min, 4 C), the superntnt ws once more trnsferred to new micro-centrifuge tube contining 300 µl of isopropnol. The tube ws gently mixed by inversion nd centrifuged t 10 000 x g for 3 min t 4 C. Subsequently, the superntnt ws discrded, nd 300 µl of 70% ethnol were dded to the pellet nd mixed by inversion. A finl centrifugtion step (10 000 x g, 3 min, 4 C) followed, nd the ethnol ws removed crefully. The tube ws dried for 15 min. To rehydrte the DNA, 25 µl TE buffer nd 0.5 µl RNse (4 mg/ml) were dded to the smple overnight t 4 C. The DNA of the wine smples ws stored t -20 C until use. DNA mplifiction nd primers The primer systems WLAB1/WLAB2 GC (Lopez et l., 2003), WBAC1/WBAC2 GC (Lopez et l., 2003), Lc1/Lc2 GC /Lc3 (Wlter et l., 2001; Endo & Okd; 2005), Lc1/Lc1o/Lc2 GC (Wlter et l., 2001; this study) nd 341f GC /518r (Be et l., 2006; Muyzer et l., 1993) were used for the mplifiction of frgments of the bcteril 16S ribosoml RNA (rrna) gene (Tble 2). In ddition, the primer system rpob1/rpob1o/rpob2 GC ws pplied for the dupliction of frgments of the RNA polymerse bet subunit, rpob (Renouf et l., 2006b; Spno et l., 2007). PCR mplifiction ws performed t finl volume of 25 µl with thermocycler (Mstercycler, Eppendorf, Hmburg, Germny) contining combintion of the corresponding primers nd templte DNA, s indicted by the uthors (Muyzer et l., 1993; Wlter et l., 2001; Lopez et l., 2003; Spno et l., 2007), nd 2.5 µl 10 x PCRbuffer (Finnzymes, Vnt, Finlnd), 0.5 µl dntp-mix (10 mm), nd 0.5 µl DNA polymerse (2U/µL, Dynzyme II; Finnzymes). The remining volume ws filled up with sterile distilled wter. Amplicons were run on 2% grose gels, stined with ethidium bromide nd photogrphed under UV trnsillumintion. DGGE The Dcode universl muttion detection system TM (Bio- Rd, Hercules, Cliforni, USA) ws used for the sequencespecific seprtion of PCR products. These were run on 8% (w/v) polycrylmide gels in TAE-buffer (40 mm Triscette; 2 mm N 2 -EDTA x H 2 O, ph 8.5) nd denturing grdient s described originlly or modified ccording to Tble 2. The electrophoresis ws performed t 85 V for 16 h in 1 x TAE-buffer t constnt temperture of 60 C. Bnd-mtching nlysis Using the BioNumerics softwre, version 6.6 (Applied Mths, Sint-Mrtens-Ltem, Belgium), wine-ssocited LAB species were identified by performing bnd mtching. Accordingly, bnds were utomticlly ssigned to bnd clsses defined by the progrm. Uncertin bnds were ignored. The ssignments were corrected mnully, leding to n optimistion of 0% (WBAC, Lc, rpob primer set) or 0.5% (WLAB, 341f GC /518r primer set) nd position tolernce of 0.5% (WLAB primer set) or 1% (WBAC, Lc, 341f GC /518r, rpob primer set). Sequence nlysis After stining the DGGE gel, bnds of interest were excised directly from the gels with sclpel, mixed with 100 µl of 1x PCR buffer, nd incubted overnight t 4 C. Two microlitres of this solution were used to re-mplify the PCR product. The PCR products were purified with the PCRExtrct Mini Kit (5 Prime) nd subjected to commercil sequencing (Eurofins

PCR-DGGE for LAB Detection 188 MWG Operon, Ebersberg, Germny). Sequence compiltion nd comprison were performed with the BLASTn progrm. RESULTS AND DISCUSSION PCR-DGGE is fst method for bcteril nlysis, enbling the survey of LAB during winemking (Renouf et l., 2006b). Therefore, selection of primer systems proposed in the literture were checked ginst ech other by testing set of reference strins s well s wine smples nd their enrichments. Lopez et l. (2003) hve lredy shown tht number of primers re not suited, s they lso mplify nonbcteril DNA, resulting in msking of bcteril popultions in DGGE profiles. They therefore developed two new primer sets specificlly for the mplifiction of bcteril 16S rrna gene in wine fermenttion smples. One primer set, termed WLAB1/WLAB2 GC, mplifies LAB, while the other one, termed WBAC1/WBAC2 GC, mplifies LAB nd cetic cid bcteri (AAB). The primer set WLAB1/WLAB2 GC trgets the V4 nd V5 regions of the 16S rrna gene nd produces frgment of pproximtely 400 bp (Lopez et l., 2003). Pure reference cultures were exmined using this primer system. It ws found tht severl LAB species exhibited similr electrophoretic mobilities, becuse ll mplicons were only TABLE 2 Primers tested for PCR-DGGE. Primer Sequence (5 3 ) Trget region Reference Modified PCR-DGGE conditions WLAB1 WBAC1 Lc1 341f GC 518r TCCGGATTTATTGGGCG- TAAAGCGA WLAB2 GC CGCCCGCCGC- GCCCCGCGCCCGGCCC- GCCGCCCCCGCCCCTC- GAATTAAACCACAT- GCTCCA GTCGTCAGCTCGTGTC- GTGAGA WBAC2 GC CGCCCGCCGC- GCCCCGCGCCCGGCCC- GCCGCCCCCGCCCCCCC- GGGAACGTATTCACCGCG AGCAGTAGGGAATCTTC- CA Lc2 GC CGCCCGGGGC- GCGCCCCGGGCGGCCC- GGGGGCACCGGGGGAT- TYCACCGCTACACATG Lc3AGCAG- TAGGGAATCTTCGG Lc1oTGCAG- TAGGGAATTTTCCG CGCCCGCCGC- GCGCGGCGGGC- GGGGCGGGGGCAC- GGGGGGCCTACGGGAG- GCAGCAG ATTACCGCGGCTGCTGG 16S rrna gene (V4 V5) 16S rrna gene (V7 V8) 16S rrna gene (V3) 16S rrna gene (V3) Lopez et l., 2003 Lopez et l., 2003 Wlter et l., 2001 Endo & Okd, 2005 this study PCR mplifiction: Lopez et l., 2003 DGGE denturing grdient: 30 to 55% of ure nd formmide PCR mplifiction: Lopez et l., 2003 DGGE denturing grdient: Lopez et l., 2003 PCR mplifiction: Wlter et l., 2001 DGGE denturing grdient: 35 to 55% of ure nd formmide Muyzer et l., 1993 PCR mplifiction: Be et l., 1993, except touchdown: 0,5 C/cycle DGGE denturing grdient: 35 to 60% of ure nd formmide rpob1 rpob1o rpob2 GC ATTGACCACTTGGGTA- ACCGTCG ATCGATCACTTAG- GCAATCGTCG CGCCCGCCGC- GCGCGGCGGGC- GGGGCGGGGGCAC- GGGGGGGCACGATCAC- GGGTCAAACCACC Modified nucleobses (bold nd itlic letters) rpo gene Renouf et l., 2006b; Spno et l., 2007 (bet-subunit) PCR mplifiction: Spno et l., 2007 DGGE denturing grdient: Renouf et l., 2006b

189 PCR-DGGE for LAB Detection displyed in smll rnge of the denturnt concentrtion. Vrying the concentrtion of denturnt of the electrophoresis gel did not improve the seprtion of the tested reference strins. Due to severl copies of the trgeted gene, some species even resulted in multiple bnds, complicting the lloction of bnds to certin LAB species. However, ll of the tested LAB could be detected nd differentited, except for Lb. csei nd Lb. prcsei (Fig. 1). Primer set WBAC1/WBAC2 GC trgets the V7 to V8 regions of the 16S rrna gene nd produced n pproximtely 320 bp mplicon with ll tested reference strins. With reference to Lopez et l. (2003), this primer system works prticulrly well to resolve AAB strins on DGGE. Anyhow, ccording to our results, this primer pir ws not cpble for nlysing the LAB diversity in wine, s the seprtion of the tested reference strins ws not sufficient nd mny species migrted to the sme position (Fig. 1). Nevertheless, it ws possible to discriminte between Lb. csei nd Lb. prcsei. Compred to the WLAB primer system, more multiple nd stronger bnds were obtined with the reference strins, except for Lb. buchneri nd Lb. fructiovorns, which resulted in wek bnds. Along with ll the LAB reference strins, the Bcillus sp. isolte ws lso detected. This isolte, s well s the AAB, produced mplicons t the sme gel positions s LAB. The primer pir Lc1/Lc2 GC ws designed for nlysing the diversity of fecl or vginl LAB nd is specific for the gener Lctobcillus, Pediococcus, Weissell nd Leuconostoc. The primer pir forms 340 bp frgment of the V3 region of the 16S rrna gene (Wlter et l., 2001). An dditionl primer ws constructed by Endo nd Okd (2005) to extend the rnge of detectble LAB for the investigtion of fermented foods. This Lc3 primer ttches t the sme position s Lc1 nd mplifies the 16S rrna gene of Lctococcus, Streptococcus, Enterococcus, Vgococcus nd Tetrgenococcus Testing different primer combintions, Endo nd Okd (2005) observed tht the use of ll three primers in PCR t the sme time ws useful to nlyse LAB diversity. Applying the primer mixture Lc1/Lc2 GC /Lc3, the most relevnt wine LAB, O. oeni, ws not mplified (dt not shown). To overcome this problem, the primer Lc1 or Lc3 ws modified in this work (Lc1o, Tble 2). The specificity of the new primer set Lc1/Lc1o/Lc2 GC ws nlysed using BLASTn nd evluted by performing PCR-DGGE with reference strins nd isoltes of the unwnted, ccompnying bcteril wine micro-flor (for detils see Mterils nd Methods). DGGE bnds were obtined for ll strins of the gener Lctobcillus, Pediococcus, Leuconostoc, Weissell nd O. oeni. In contrst, no bnds were chieved for the gener Enterococcus, Streptococcus, Tetrgenococcus (dt not shown) nd Lctococcus, for which the Lc3 primer ws designed by Endo nd Okd (2005). However, these gener generlly re not relevnt for the fermenttion processes of wines. Due to the genertion of multiple bnds for mny reference strins, the evlution of the results ws difficult. Except for Lb. prcsei, the identifiction of species of the Lctobcillus csei nd Lb. plntrum group, s well s the Pediococcus genus, ws often only possible t genus or species group level (Fig. 1), lthough good seprtion ws chieved for ll other reference strins. No bnds on DGGE gel were obtined for the non-lab bcteri Bcillus cogulns, Acetobcter ceti nd Gluconobcter (Fig. 1). The universl primer set 341f GC /518r, designed by Muyzer et l. (1993), ws pplied successfully by Be et l. (2006) to detect LAB ssocited with wine grpes. It mplifies frgment of the V3 region of the 16S rrna gene, forming 233 bp product. The PCR product from the Lb. fructivorns reference strin ws wek when using this primer set without GC-clmp, nd resulted in no visible bnd on the DGGE gel. However, the distribution of the bnds of ll other LAB reference strins ws good, lthough multiple bnds per strin ppered (Fig. 1). In ddition, different LAB of one genus or species group (e.g. the Pd. dmnosus, Pd. prvulus, Pd. inopintus, Lctobcillus csei nd Lb. plntrum group) showed identicl results on the DGGE gel. As this primer set is universl, fint bnds of AAB nd the Bcillus isolte were displyed, but not in the concentrtion rnge of LAB. As ribosoml genes re present in severl copies with different sequences (Rntsiou et l., 2004), ll primer systems considered generted diverse mplicons, resulting in multiple bnds on the gel. Thus, nother primer set trgeting the RNA polymerse bet subunit gene rpob, which is only present s single copy (Rntsiou et l., 2004), ws lso included in the tests. This primer set, originlly developed by Renouf et l. (2006), hd lredy been used to study the effect of different oenologicl prctices on LAB popultions nd their evolution during winemking. Bsed on the rpob1/rpob1o/rpob2 GC primer system, the bnds of the reference strins were well seprted within this study (Fig. 2). However, the optimlly expected single bnd per strin ppered s min bnd with wek double bnds in its neighbourhood (Renouf et l., 2006). Amongst others, these bnds my be due to n enzymtic process involving the TAQ polymerse (Jnse et l., 2004). Nevertheless, the min bnds were clerly seprted nd visible. In the cse of wine smples, the unmbiguous detection of min bnds could even be improved with mixture of vrious species. Otherwise, reference strins of some LAB species lredy found in wine could not be detected beside the ccompnying bcteril flor (e.g. Lc. lctis, Lb. brevis, Lb. csei, Lb. coryniformis, Lb. curvtus, Lb. delbrueckii, Lb. fermentum, Lb. fructivorns, Lb. hilgrdii, Lb. lindneri, Lb. ngeli, Lb. zee nd Pd. inopintus), or generted only fint bnds (e.g. Lb. prcsei nd Pd. prvulus). Although Renouf et l. (2006) could determine the wine-relevnt species Lb. brevis nd Lb. hilgrdii by PCR-DGGE, we could not produce mplicons for the used reference strins of these two species in our study when pplying this primer system. However, Lb. brevis could be identified in one of the subsequently tested smples. As PCR-DGGE ptterns obtined with the reference strins should llow the tenttive identifiction of DNA frgments in ech smple, wine-ssocited LAB species were identified by mtching their bnd distnces to those of the reference strins using the BioNumerics softwre. O. oeni, Lb. brevis, members of the Pediococcus genus, the Lb. buchneri, Lb. csei nd Lb. plntrum group were

PCR-DGGE for LAB Detection 190 FIGURE 1 Digitised PCR-DGGE ptterns of ctive rnges from 29 reference strins, four ccompnying flor isoltes, one wine smple nd 26 wine enrichments with primer sets WLAB1/WLAB2 GC (30% to 80%), WBAC1/WBAC2 GC (45% to 75%), Lc1/Lc1o/ Lc2 GC (25% to 85%) nd 341f GC /518r (10% to 80%). Verticl lines indicte the specified bnd clsses. The bnds lbelled 1 to 56 re described in Tble 4. bnd clss ssignment

191 PCR-DGGE for LAB Detection TABLE 3 Detection of LAB in wine smples nd/or their enrichment cultures bsed on bnd position, bnd pttern of species, sequencing results nd comprison between different primer systems. Smple Primer set WLAB1/WLAB2 GC WBAC1/WBAC2 GC Lc1/Lc1o/Lc2 GC 341f GC /518r rpob1/rpob1o/rpob2 GC 1 b O. oeni, Pd. dmnosus Pd. prvulus O. oeni, Pediococcus Pediococcus O. oeni, Pd. dmnosus 2 c O. oeni O. oeni O. oeni O. oeni, Pediococcus O. oeni 3 c O. oeni, Pd. dmnosus O. oeni, Pd. dmnosus O. oeni, Pediococcus Pediococcus O. oeni, Pd. dmnosus 4 c Pd. dmnosus Pd. dmnosus Pediococcus Pediococcus Pd. dmnosus 5 c O. oeni O. oeni O. oeni O. oeni, Pediococcus O. oeni 6 c Le. mesenteroides, Pd. dmnosus 7 c Le. mesenteroides, Pd. prvulus, Lb. csei-group Pd. dmnosus O. oeni, Pediococcus Pediococcus O. oeni, Pd. dmnosus Lb. prcsei O. oeni, Pd. prvulus, Lb. prcsei Pediococcus, Lb. cseigroup O. oeni, Pd. prvulus 8 c Le. mesenteroides, Pd. prvulus Pd. prvulus Pediococcus Pediococcus Pd. prvulus 9 c Lb. buchneri/hilgrdii Lb. prcsei Lb. buchneri, Lb. prcsei Lb. buchneri, Lb. csei-group O. oeni, Lb. buchneri, Lb. rhmnosus 10 c Lb. csei-group, Pd. prvulus Lb. prcsei Lb. prcsei, Pediococcus Lb. csei-group, Pediococcus O. oeni, Pd. prvulus 11 c Lb. csei-group Lb. prcsei Lb. prcsei Lb. csei-group Lb. rhmnosus 12 c Lb. csei-group, Pd. prvulus Lb. prcsei Lb. prcsei, Pediococcus Lb. csei-group, Pediococcus Lb. buchneri, Lb. rhmnosus 13 c Lb. csei-group Lb. prcsei Lb. prcsei Lb. csei-group Lb. rhmnosus 14 c Lb. buchneri Lb. buchneri Lb. buchneri Lb.buchneri Lb. buchneri 15 c O. oeni, Pd. prvulus O. oeni O. oeni, Pediococcus Pediococcus O. oeni, Pd. prvulus 16 c Pd. prvulus Pd. prvulus O. oeni, Pediococcus Pediococcus O. oeni, Pd. prvulus 17 c Pd. prvulus Pd. prvulus O. oeni, Pediococcus Pediococcus Pd. prvulus, Pd. dmnosus 18 c O. oeni O. oeni O. oeni O. oeni, Geminicoccus O. oeni 19 c Lb. brevis Lb. brevis Lb. brevis Lb. brevis, Enterobcter Lb. brevis 20 c Pd. prvulus Pd. prvulus, Gluconobcter 21 c Lb. buchneri/hilgrdii Acetobcter, Gluconobcter Pd. prvulus Pediococcus Pd. prvulus Lb. diolivorns Lb. diolivorns/hilgrdii Lb. buchneri

PCR-DGGE for LAB Detection 192 TABLE 3 (CONTINUED) Primer set Smple WLAB1/WLAB2 GC WBAC1/WBAC2 GC Lc1/Lc1o/Lc2 GC 341f GC /518r rpob1/rpob1o/rpob2 GC 22 c Lb. csei-group Lb. prcsei Lb. prcsei Lb. csei-group Lb. prcsei 23 c O. oeni O. oeni, Gluconobcter O. oeni O. oeni - Lb. prcsei Lb. diolivorns/hilgrdii, Lb. csei-group 24 c Lb. buchneri/hilgrdii Lb. prcsei Lb. diolivorns/hilgrdii, Lb. prcsei 25 c Lb. plntrum Lb. plntrum-group Lb. plntrum-group Lb. plntrum-group Lb. plntrum 26 c Lb. plntrum Lb. plntrum-group Lb. plntrum-group Lb. plntrum-group Lb. plntrum 27 c Pd. prvulus Pd. prvulus Pd. prvulus Pediococcus Pd. prvulus The highest bnd ppered in ll strin mixtures (used s mrker, dt not shown) nd smples with strong bnd signls. It therefore ws not used for differentition. b wine smple, c enrichment smple FIGURE 2 Digitised PCR-DGGE ptterns of the ctive rnge (10% to 75%) from 29 reference strins, four ccompnying flor isoltes, one wine smple nd 26 wine enrichments with primer set rpob1/rpob1o/rpob2 GC. Verticl lines indicte the specified bnd clsses. The bnds lbelled 57 to 67 re described in Tble 4. bnd clss ssignment

193 PCR-DGGE for LAB Detection TABLE 4 Similrities of sequenced bnds. Primer system Bnd Closest reltives GenBnk ccession no. % sequence similrity WLAB 1, 10 O. oeni NR_075030.1 99% 2, 3 Pd. dmnosus NR_042087.1 99% 4, 9 Lb. csei, Lb. prcsei, Lb. zee NR_075032.1, NR_041054.1, NR_037122.1 99% 5-8 Lb. buchneri NR_102772.1 99% 11 Lb. plntrum, Lb. prplntrum, Lb. pentosus NR_075041.1, NR_025447.1, NR_029133.1 99% 12 Pd. prvulus NR_029136.1 99% WBAC 13-15 Pd. dmnosus NR_042087.1 99% 16, 18 Pd. prvulus, Pd. ethnolidurns NR_029136.1, NR_043291.1 99% 17 Lb. prbuchneri, Lb. buchneri NR_041293.1, NR_041294.1 99% 19 Lb. brevis NR_044704.1 99% 20, 24 Gluconobcter NR_041047.1, NR_041050.1, NR_041049.1, NR_026118.1 99% 21 Acetobcter NR_028614.1, NR_025513.1, NR_025512.1 99% 22 Lb. prcsei, Lb. csei NR_041054.1, NR_043408.1 99% 23 O. oeni NR_040810.1 99% Lc 25 Pd. dmnosus NR_042087.1 99% 26, 27, 39, 42 Pd. prvulus NR_029136.1 99% 28, 29, 34-37 Lb. buchneri NR_041293.1 99% 30-33 Lb. prcsei, Lb. csei, Lb. zee NR_041054.1, NR_041893.1, NR_037122.1 99% 38 Lb. brevis NR_044704.1 99% 40 Lb. diolivorns NR_037004.1 99% 41 Lb. plntrum, Lb. prplntrum, Lb. pentosus NR_042394.1, NR_025447.1, NR_029133.1 99% 341f 43-45 Pd. prvulus, Pd. dmnosus, Pd. inopintus NR_029136.1, NR_042087.1, NR_025388.1 99% 46-48 Lb. buchneri NR_041293.1 99% 49, 55 Lb. rhmnosus, Lb. csei, Lb. prcsei, Lb. zee NR_102778.1, NR_075032.1, NR_041054.1, NR_037122.1 97%, 99% 50 O. oeni NR_075030.1 100% 51 Geminicoccus sp. NR_042567.1 82% 52 Lb. brevis NR_075024.1 99% 53 Klebsiell sp., Erwini sp., Enterobcter sp. NR_102982.1, NR_102820.1, NR_024640.1 100% 54 Lb. diolivorns, Lb. hilgrdii NR_044708.2, NR_037004.1 99% 56 Lb. plntrum, Lb. prplntrum, Lb. pentosus NR_075041.1, NR_025447.1, NR_029133.1 99% rpob 57, 58 Pd. dmnosus DQ176043.1 99%, 100% 59, 60, 64 O. oeni CP000411.1 98% 61-63 Pd. prvulus AY875850.1 99%-100% 65 Lb. brevis AP012167.1 98% 66 Lb. prcsei, Lb. csei CP007122.1, HE970764.1 99% 67 Lb. plntrum CP006033.1 99% sme similrity for ll sequenced bnds

PCR-DGGE for LAB Detection 194 detectble when investigting the smples using ech selected primer system (Tble 3). In ddition to these LAB, wek bnds corresponding to the species Le. mesenteroides were displyed by the WLAB set. This species, s well s the species detected by ll primer sets, belong to the min LAB isolted from must nd wines (Pozo-Byón et l., 2009). Compred to the other primer systems used in our experiment, O. oeni ws rrely detected by the 341f GC /518r primer set, wheres this species ws frequently identified by the Lc nd rpob primer systems with specific primers for O. oeni. Furthermore, species of the Lb. buchneri group nd the Pediococcus genus were rrely found by the WBAC primer set. The detection of the diversity of species by vrious primer sets my be influenced by their differing ffinity to different species (Be et l., 2006). According to our observtions, the primer pir used itself lso ffects the detection limit of PCR-DGGE. Thus, O. oeni could not or hrdly be identified in smple 23 by the rpob nd 341f GC /518r primer sets respectively, wheres its presence ws clerly detected by ll other primer systems (Tble 3). This detection limit even increses when competitive templte DNAs re present (Andorrà et l., 2008). Furthermore, Be et l. (2006), s well s Renouf et l. (2006b), concluded independently tht their pplied primer sets were only ble to revel the predominnt species. When pplying the primer sets described bove to investigte 27 wine smples nd enrichment cultures, the sme species were mostly detected by the Lc1/Lc1o/Lc2 GC nd WLAB1/WLAB2 GC primer sets, followed by the rpob1/rpob1o/rpob2 GC primer systems. The biggest diversity of LAB species ws lso verified by these primer sets. The WBAC1/WBAC2 GC nd 341f GC /518r primer systems showed the poorest complince. Bsed on the sequence nlysis performed for selected bnds (Tble 4), it turned out tht the similrity of the sequences of LAB bnds generted by primers trgeting the 16S rrna gene (WLAB, WBAC, Lc nd 341f GC /518r systems) with those vilble in the dtbse ws 97%, wheres ll tested rpob sequences corresponded to dtbse sequences with similrity of 98%. Furthermore, sequences generted by the rpob primer set were more discrimintive for the identifiction of relted LAB species thn those produced by 16S rrna gene primer systems, which is in ccordnce with the literture (Renouf et l., 2006; Lv et l., 2012). CONCLUSIONS Due to poor complince with the other primer sets, the WBAC1/WBAC2 GC nd 341f GC /518r primer systems re not suitble to investigte the diversity of LAB involved in winemking. In ddition, multiple bnds were frequently produced for the reference strins tested, complicting the lloction of prticulr bnd to defined species. Compred to these primer systems, the Lc set with the modified primer Lc1o exhibited slightly better performnce, lthough multiple bnds lso were obtined. Due to the proper seprtion of different species on the gel, s well s their verifiction by distinctive sequence nlysis, the rpob1/rpob1o/rpob2 GC primer system seems to be promising tool for monitoring the evolution of wine LAB. However, improvements should be mde, s the detection limit of this set seems to be higher thn tht of the other primer sets. Owing to its pronounced sensitivity nd its cpbility of discriminting to species level, the WLAB1/WLAB2 GC primer set turned out to be dvntgeous for LAB detection purposes in wine. LITERATURE CITED Andorrà, I., Lndi, S., Ms, A., Guillmón, J.M. & Esteve-Zrzoso, B., 2008. Effect of oenologicl prctices on microbil popultions using culture-independent techniques. Food Microbiol. 25, 849-856. Be, S., Fleet, G.H. & Herd, G.M., 2006. Lctic cid bcteri ssocited with wine grpes from severl Austrlin vineyrds. J. Appl. Microbiol. 100, 712-727. Bousbours, G.E. & Kunkee, R.E., 1971. Effect of ph on mlo-lctic fermenttion in wine. Am. J. Enol. Vitic. 22, 121-126. Cho, G.-S., Kruß, S., Huch, M., Du Toit, M. & Frnz C.M.A.P., 2011. Development of quntittive PCR for detection of Lctobcillus plntrum strters during wine mlolctic fermenttion. J. Microbiol. Biotechnol. 21, 1280-1286. Cocolin, L., Mnzno, M., Cntoni, C. & Comi, G., 2001. Denturing grdient gel electrophoresis nlysis of the 16S rrna gene V1 region to monitor dynmic chnges in the bcteril popultion during fermenttion of Itlin susges. Appl. Environ. Microbiol. 67, 5113-5121. Dvis, C.R., Wibowo, T.H.L. & Fleet, G.H., 1986. Growth nd metbolism of lctic cid bcteri during nd fter mlolctic fermenttion of wines t different ph. Appl. Environ. Microbiol. 51, 539-545. Endo, A. & Okd, S., 2005. Monitoring the lctic cid bcteri diversity during shochu fermenttion by PCR-denturing grdient gel electrophoresis. J. Biosci. Bioeng. 99, 216-221. Ercolini, D., 2004. PCR-DGGE fingerprinting: Novel strtegies for detection of microbes in food. J. Microbiol. Methods. 56, 297-314. Jnse, I., Bok, J. & Zwrt, G., 2004. A simple remedy ginst rtifctul double bnds in denturing grdient gel electrophoresis. J. Microbiol. Methods. 57, 279-281. Knoll, C., Fritsch, S., Schnell, S., Grossmnn, M., Krieger-Weber, S., Du Toit, M. & Ruhut, D., 2012. Impct of different mlolctic fermenttion inocultion scenrios on Riesling wine rom. World J. Microbiol. Biotechnol. 28, 1143-1153. Lerm, E., Engelbrecht L. & Du Toit, M., 2010. Mlolctic fermenttion: The ABC s of MLF. S. Afr. J. Enol. Vitic. 31, 186-212. Lonvud-Funel, A., 1999. Lctic cid bcteri in the qulity improvement nd deprecition of wine. Antonie vn Leeuwenhoek 76, 317-331. Lopez, I., Ruiz-Lrre, F., Cocolin, L., Orr, E., Phister, T., Mrshll, M., Vndergheynst, J. & Mills, D.A., 2003. Design nd evlution of PCR primers for nlysis of bcteril popultions in wine by denturing grdient gel electrophoresis. Appl. Environ. Microbiol. 69, 6801-6807. Lv, X.-C., Weng, X., Zhng, W., Ro, P.F. & Ni, L., 2012. Microbil diversity of trditionl fermenttion strters for Hong Qu glutinous rice wine s determined by PCR-medited DGGE. Food Control 28, 426-434. Mills, D.A., Rwsthorne, H., Prker, C., Tmir, D. & Mkrov, K., 2005. Genomic nlysis of Oenococcus oeni PSU-1 nd its relevnce to winemking. FEMS Microbiol. Rev. 29, 465-475.

195 PCR-DGGE for LAB Detection Moreno-Arribs, M.V. & Polo, M.C., 2005. Winemking microbiology nd biochemistry, current knowledge nd future trends. Crit. Rev. Food Sci. nd Nutr. 45, 265-286. Muyzer, G., De Wl, E.C. & Uitterlinden, A.G., 1993. Profiling of complex microbil popultions by denturing grdient gel electrophoresis nlysis of polymerse chin rection mplified genes coding for 16S rrna. Appl. Environ. Microbiol. 59, 695-700. Pinzni, P., Boncini, L., Pzzgli, M., Orlndo, C., Guerrini, S. & Grnchi, L., 2004. Rpid detection of Oenococcus oeni in wine by rel-time quntittive PCR. Let. Appl. Microbiol. 38, 118-124. Pozo-Byón, M.A., Prdo, I., Ferrer, S. & Moreno-Arribs, M.V., 2009. Moleculr pproches for the identifiction nd chrcteristion of oenologicl lctic cid bcteri. Afr. J. Biotechnol. 8, 3995-4001. Rntsiou, K., Comi, G. & Cocolin, L., 2004. The rpob gene s trget for PCR-DGGE nlysis to follow lctic cid bcteril popultion dynmics during food fermenttions. Food Microbiol. 21, 481-487. Renouf, V., Clisse, O. & Lonvud-Funel, A., 2006. RpoB gene: A trget for identifiction of LAB cocci by PCR-DGGE nd melting curves nlysis in rel time PCR. J. Microbiol. Methods 67, 162-170. Renouf, V., Clisse, O. & Lonvud-Funel, A., 2007. Inventory nd monitoring of wine microbil consorti. Appl. Microbiol. Biotechnol. 75, 149-164. Renouf, V., Clisse, O., Miot-Sertier, C. & Lonvud-Funel, A., 2006b. Lctic cid bcteri evolution during winemking: Use of rpob gene s trget for PCR-DGGE nlysis. Food Microbiol. 23, 136-145. Renouf, V., Vyssieres, L.C., Clisse, O. & Lonvud-Funel, A., 2009. Genetic nd phenotypic evidence for two groups of Oenococcus oeni strins nd their prevlence during winemking. Appl. Microbiol. Biotechnol. 83, 85-97. Spno, G., Lonvud-Funel, A., Clisse, O. & Mss, S., 2007. In vivo PCR-DGGE nlysis of Lctobcillus plntrum nd Oenococcus oeni popultions in red wine. Curr. Microbiol. 54, 9-13. Wlter, J., Hertel, C., Tnnock, G.W., Lis, C.M., Munro, K. & Hmmes, W.P., 2001. Detection of Lctobcillus, Pediococcus, Leuconostoc, nd Weissell species in humn feces by using group-specific PCR primers nd denturing grdient gel electrophoresis. Appl. Environ. Microbiol. 67, 2578-2585. Wibowo, D., Fleet, G.H., Lee, T.H. & Eschenbruch, R.E., 1988. Fctors ffecting the induction of mlolctic fermenttion in red wines with Leuconostoc oenos. J. Appl. Bcteriol. 64, 421-428.