This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

Author's personal copy

Expanding the diversity of oenococcal bacteriophages:Insights into a novel group based on the integrase sequence

Fety Jaomanjaka, Patricia Ballestra, Marguerite Dols-lafargue, Claire Le Marrec ⁎University of Bordeaux, ISVV, Unit of Enology, EA 4577, Villenave d'Ornon, France

a b s t r a c ta r t i c l e i n f o

Article history:Received 3 February 2013Received in revised form 14 June 2013Accepted 25 June 2013Available online 7 July 2013

Keywords:Oenococcus oeniMalolactic fermentationLysogenyProphage induction

Temperate bacteriophages are a contributor of the genetic diversity in the lactic acid bacterium Oenococcus oeni.We used a classification scheme for oenococcal prophages based on integrase gene polymorphism, to analyze acollection of Oenococcus strains mostly isolated in the area of Bordeaux, which represented the major lineagesidentified through MLST schemes in the species. Genome sequences of oenococcal prophages were clusteredinto four integrase groups (A to D) which were related to the chromosomal integration site. The prevalence ofeach group was determined and we could show that members of the intB- and intC-prophage groups wererare in our panel of strains. Our study focused on the so far uncharacterized members of the intD-group. VariousintD viruses could be easily isolated fromwine samples, while intD lysogens could be induced to produce phagesactive against two permissive O. oeni isolates. These data support the role of this prophage group in the biologyof O. oeni. Global alignment of three relevant intD-prophages revealed significant conservation and highlighteda number of unique ORFs that may contribute to phage and lysogen fitness.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

In the context of an increasing worldwide interest in improvingthe quality and safety of wines, significant effort has been made toprovide exhaustive inventories of the complex microbial communi-ties associated with grapes and wine. Grapes were reported to havea complex microbial ecology including filamentous fungi, yeasts andbacteria with different physiological characteristics and effects uponwine production. These communities were found to be particularlyprone to fluctuations in composition, and a complexity and diversityof the consortium have been found at all stages of the winemakingprocess: on grape berries, in must during fermentation, and in wineduring aging (Renouf et al., 2007; Barata et al., 2012; Martins et al.,2012). Presence of viruses infecting bacteria (called bacteriophages)in this changing ecosystem has been widely neglected, and it remainsunknown whether they may play a significant role in shaping themicrobial communities, and influence the dynamics of fermentingbacteria, or spoilage microorganisms which lead to depreciationof wine. Yet, increasing evidence is being found in other ecosystems(such as aquatic systems, and mammalian gut microbiota) indicatingthat propagation of bacteriophages not only controls host diversityand abundance, but can also in some cases benefit bacterial host

fitness under specific environmental conditions (Weinbauer andRassoulzadegan, 2004; Mills et al., 2012).

During winemaking, a major source of concern is the progress ofmalolactic fermentation (MLF) which reduces acidity and developsaromas in the final product, and is largely driven by the lactic acidbacterium Oenococcus oeni. The process can be spontaneous, relyingon endogenous O. oeni strains, but winemakers often prefer to controlMLFwith predictable starters, available under ready-to-use concentrat-ed cultures. Pioneering studies have focused on the possible predationof O. oeni by specific viruses, and diverse oenococcal phages have beensuccessfully isolated from wines with sluggish or delayed MLF (Sozziet al., 1982; Tenreiro et al., 1993; Davis et al., 1985; Henick-Klinget al., 1986a, 1986b). Some have been characterized in terms of mor-phology, lytic spectra, restriction enzyme analysis of the genome, DNAhomology, genome size and protein structure (Nel et al., 1987; Arendtet al., 1991; Boizet et al., 1992; Tenreiro et al., 1993; Davey et al.,1995; Santos et al., 1998). The lack of characterization at the molecularlevel has considerably limited the assessment of oenococcal phagediversity. On the other hand, a high prevalence of lysogeny has beenlargely documented in O. oeni (Cavin et al., 1991; Tenreiro et al., 1993;Poblet-Icart et al., 1998) and lysogenic strains have been proposed toserve as a reservoir for phages, protecting them from the aggressiveconditions prevailing in wine (Gindreau, 1998; Henick-Kling et al.,1986b). More recent comparisons of complete genome data confirmedthe presence of numerous phage-related sequences in the genomes ofstrains, making prophages a major contributor of the genetic diversityin the O. oeni species (Zé-Zé et al., 1998; Bon et al., 2009; Bornemanet al., 2012a, 2012b).

International Journal of Food Microbiology 166 (2013) 331–340

⁎ Corresponding author at: EA OEnologie, Institut des Sciences de la Vigne et du Vin(ISVV) (Université Bordeaux Segalen-Institut Polytechnique de Bordeaux), 210 Cheminde Leysotte, 33882 Villenave d'Ornon Cedex, France. Tel.: +33 5 57 575 831; fax: +335 57 575 813.

E-mail address: clehenaff@enscbp.fr (C. Le Marrec).

0168-1605/$ – see front matter © 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.ijfoodmicro.2013.06.032

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j ourna l homepage: www.e lsev ie r .com/ locate / i j foodmicro

Author's personal copy

In order to gain insights in the significance of lysogeny in O. oeni,the integration modules comprising the integrase-encoding genes (int)and the phage attachment sites from different temperate oenococcalphages, namely fog30, fog44, fogPSU1 andΦ10MC have been identifiedand sequenced (Gindreau et al., 1997; Santos et al., 1998; Parreira et al.,1999; São-José et al., 2004). Phages segregated in three groups regardingthe tRNA site used for site-specific recombination. In addition, a linkingof these tRNA sites and int sequences has been evidenced (Ballestraet al., 2011). These datawere recently confirmed in an in silico evaluationof prophages carriage in strains of O. oeni (Borneman et al., 2012b).Interestingly, this study also revealed the existence of putativecomplete prophages harboring a distinct tRNA integration siteand cognate int sequence, corresponding to a probable fourthgroup. However, whether these sequences produce active phageon induction has not been documented. The present study wasconducted to gain more information on this so far unexploredgroup of prophages. Their prevalence was analyzed in a set of28 strains of O. oeni mostly collected in the Bordeaux area andreflecting the overall genetic diversity in the species. We assessedwhether (i) these prophages were able to be induced for self-replication and to be excised from the genome, and (ii) free bacte-riophages belonging to this group were present in wine samples.Global alignments of three active prophages belonging to this novelgroup are presented.

2. Material and methods

2.1. Bacterial strains and phages

All O. oeni strains were routinely grown in liquid or solid modi-fied MRS (Difco, Fischer Bioblock Scientific, Illkirch, France) adjustedto pH 4.8 at 25 °C. Their characteristics have been described earlier(Bilhère et al., 2009; Favier et al., 2012). The strains were stored in30% (vol/vol) glycerol at−80 °C.

2.2. Isolation and propagation of phages

Phages were isolated from 22 red wine samples collected atthe beginning or at the end of MLF from diverse wineries (Medoc,Bordeaux region, Aquitaine, France). Wine samples were ten-folddiluted in MRS liquid medium, pH 4.8 and incubated for 5 days at25 °C. Cultures were centrifuged and supernatants were filter-sterilizedthrough 0.2-μm membranes. Wine samples were screened for the pres-ence of phages using the classical double-layer plating technique, usingMRS agar supplemented with MgSO4 (3.75 g/l) and CaCl2 (2.375 g/l)(MRSΦ). Eleven different indicator O. oeni strains mostly collectedin the Bordeaux area were used. Eight strains have been isolated fromred wine samples during spontaneous MLF (Sarco S14, S15, S24, S28,S51, S161, B10, IOEB 89006). One strain has been isolated from a whitesparkling wine (Sarco S11). The others (IOEB-Sarco 277 and 450PreAc®, Laffort) are commercial starters. Plates were incubatedunder anaerobic conditions at 25 °C for 4 to 7 days. Plaques (seeSupplementary Fig. S1) were carefully excised, suspended in 0.5 mlof sterile MRSΦ medium and stored at 4 °C. Two successive roundsof purification were carried out for each phage.

2.3. High-titers phage lysates

For each purified phage/host, ten confluent lysis plateswere prepared.Soft agars were collected, centrifuged, and the supernatant was filter-sterilized. Phage titers ranging from 106 to 1010 plaque-forming unitsper ml (PFU/ml) were obtained. Phage lysates were stored at 4 °C untiluse.

2.4. Host range determination

Bacteria were grown to exponential phase, and 200 μl of the bac-terial culture was mixed in 5 ml of top MRSΦ agar and poured ontoMRSΦ agar plates. Spot assays were performed using 8 μl of the puri-fied phage. Plates were incubated at 25 °C 72h-96h, and presence ofplaques was recorded as a positive test. The minimum detectable titerwas 1.2 × 102 PFU/ml.

2.5. Induction with MC

For the induction of phages, mitomycin C was used as the inducingagent (0.5 to 1 μg/ml), with bacterial culture on modified MRS. Over-night cultures were diluted 10-fold in 10 ml of fresh broth, grownto an optical density at 600 nm (OD600) of 0.2 to 0.3 prior to theaddition of inducing agent, and incubated for 24 h. OD600wasmeasuredperiodically. Supernatants were filtered (0.2-μm membrane), and100 μl of the filtrate (or of dilutions when necessary) was mixed with200 μl of host cells (Sarco S25 and IOEB-Sarco 277) and added to 5 mlof top agar (0.6% [wt/vol] agar) at 40 to 45 °C and poured on agar plates.The protocol allowed the obtention of a functional phage lysate from theestablished lysogen B10 containing the integrated intB-Φ10MC phage(Gindreau et al., 1997).

2.6. Comparative bioinformatic analysis of prophages

Raw sequences were assembled using Newbler and annotated usingRAST (Rapid Annotation using Subsystem Technology) (Aziz et al.,2008). Homology searches were performed using NCBI BLAST2 soft-ware. Multiple alignments of protein sequences were constructedusing COBALT (Papadopoulos, and Agarwala, 2007). Phylogeneticanalyses of the phage proteins based on amino acid sequences werecarried out using Neighbor–Joining (NJ) methods.

2.7. PCR and DNA sequencing of PCR products

Standard PCR reactions were performed in a 25 μl volume using theBiorad Taq PCRMasterMix kit, 0.2 μMof each primer, and 0.5 μl of tem-plate DNA (50 ng). Amplifications were also carried out on free phageparticles. Diluted phage lysates (~103 PFU per reaction) were first sub-mitted to a control PCR usingmalolactic enzyme gene targeted primers(Divol et al., 2003). Absence of amplicon was considered as an absenceof contaminating bacterial DNA in the phage sample. Primer designwasachieved by using the eprimer3 and Oligo analyser 1.0.3 software.Oligonucleotides, purchased from Qiagen Operon, are listed in Table 1.Sequencing reactions were performed with 1 μl of purified PCR pro-duct, using the BigDye Terminator v3.1 Cycle Sequencing kit (AppliedBiosystems).

2.8. Phage and prophage molecular typing

Based on the comparative genomic analysis of the available phagesequences, a PCR strategy for typing the four different integrase-typesequences was developed (Ballestra et al., 2011). A high level ofnucleotide sequence identity was observed between int genes fromgroup A (97 to 100% identity), and those from group D (99–100%identity). PCR primers amplified conserved 273-bp (IntA f/r) and343-bp fragments (IntD f/r), respectively (Table 1). Group B containeda unique sequence, obtained from strain B10 (Φ10MC) (Gindreauet al., 1997). Group C comprised the integrase sequences fromΦFogPSU-1 and that of a prophage found in strain S28. The primercouples IntB f/r and IntC f/r were designed (Table 1) leading to theamplification of 443-bp and 406-bp, amplicons, respectively. Allcouples yielded amplicons with the expected size using lysogenicstrains described by Borneman et al. (2012b).

332 F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

PCR tests specific for the four bacterial insertion sites were developed:attBA (OEOE_0506), attBB (OEOE_0851), attBC (sequence downstreamof OEOE_0685) and attBD (OEOE_1359). The ability to amplify these attBregions indicated the presence of prophage cured cells within the lyso-genic population, or nonlysogenic strains. Primers for genes adjacent toeach attB site were designed and sizes of the amplicons were 145-bp,730-bp, 337-bp and 342-bp, respectively (Table 1). The genomic environ-ment of attBC and attBD in strains PSU-1 and ATCC BAA1163 slightlydiffered. We first observed a deletion of the region downstream of attBCdeleting the attBC reverse primer in ATCC BAA 1163. Consequently, asecond couple of primer was designed (reg1f/r) and used to assess thepresence of attBC in strain ATCC BAA1163, aswell as in the strains yieldingno signal with primers attBC f/r (Bon et al., 2009). A similar feature wasobserved at the vicinity of attBD site. A set of eight genes was insertedupstream of the tRNALeu gene in strains IOEB 0501 and 0502, deletingthe attBD f primer. Presence of this modified attBD site was assessedwith the primer couple attBD f bis/r (Table 1).

Presence of prophages in all strains (including the 12 unsequencedisolates) was tested as follows. First, chromosomal DNA was amplifiedusing primers for each type of integrase sequence, and each emptybacterial attachment site. Strains positive for int sequence were testedusing a PCR assay targeting two short conserved sequences in theendolysin genes (endoF/R and endoF/R′). These primer couples ampli-fied a 447-bp or a 449-bp amplicon in all strains. Next, amplificationwas performed across the phage junction sites. The attLwas consideredas the junction located upstream of the prophagic int sequence. To ob-tain these attL sequences (A to D), four primer couples were designed,each corresponding to sequences associated with the predicted bacteri-al attB region and respective integrase gene (see Table 1). Sizes of theamplicons were 1123-pb, 1755-pb, 848-pb and 1294-pb, respectively.To obtain attR sequences from intD-lysogens, a distinct primer flankingthe attBD region was combined with a primer located in a conservedregion of the distal module found in all intD-prophages (joncD).A 361-bp amplicon was obtained in all intD lysogens strains, exceptfor Sarco S13 which gave a 876-bp amplicon. Lastly, the attP site wasamplified from intD-phage particles using primers joncD and intDF,yielding a 1314-bp fragment.

2.9. Nucleotide sequence accession number

DNA sequence data of the prophages found in strains IOEBSarco 9805,Sarco S11 and Sarco S13have beendeposited inGenBankunder accessionnumbers no. KF147927, KF183314 and KF183315, respectively.

3. Results and discussion

3.1. Survey of prophages among O. oeni strains collected in Bordeaux

A survey for prophageswas conducted on a panel of 28O. oeni strains,which all presented distinct DNA banding patterns by NotI-PFGE analysis,or MLST (Bridier et al., 2010; Favier et al., 2012).

3.1.1. Integrase protein sequencesThe databases of 16 draft sequences of O. oeni strains isolated during

spontaneous fermentationwere exploredwith regard to their prophagecontent and respective predicted integrase genes (Table 2A). Oursurvey identified a total of 12 distinct prophage loci containing clus-tered genes encoding phage-related proteins. Their size and organiza-tion suggested they may correspond to putative active prophages. Theevolutionary relations of the integrases found in bacteriophages andprophages infecting O. oeni were recently reported. Integrase proteinswere shown to correspond to tyrosine recombinases, and segregatedinto four groups (hereafter named intA, intB, intC, and intD), providinga first classification scheme for oenococcal phages (Ballestra et al.,2011; Borneman et al., 2012b). Accordingly, all integrases found in ourpanel of 12 prophages could be assigned to one of the four describedgroups. Within each group, the nucleotide sequence identity was97% to 100% between members. Owing to this high conservation, fourintegrase-specific PCR tests were designed, and used to type an addi-tional panel of 12 unsequencedMLF performing strains, including com-mercial starters. Presence of defective prophages harboring integrase-related sequences has been documented in the past in O. oeni (Bonet al., 2009; Borneman et al., 2012b) and consequently, amplificationof the integrase gene types does not necessarily mean that the bacterialstrains analyzed are true lysogens. To extend the accuracy of our

Table 1Primers used in this study.

Name Purpose Sequence (5′-3′)

Oo1 Control for absence of bacterial genomic fragments (malolactic enzyme gene) in phage samples GTGCCGCTTTTTTGGATATTAOo2 AGCAATTTTATCTTTATAGCTIntAf Detection of intA-type integrases (Φfog30/44) CGAAGTTTTGACTGGAAAAGAAAIntAr TTGAGCGAAGCTGCTATAAGAACIntBf Detection of intB-type integrases (Φ10MC) AGTTACCACCAAAGGCCATAAACIntBr GGCTATGACGCAGGGCGTGGIntCf Dectection of intC-type integrases (Φfog PSU-1) TGACGGGACGTGCTGGCAAGIntCr GCTCTGACGACTTACCAGCTTTAIntDf Detection of intD-type integrases CGGAAAATATTATCAAGCACGAGIntDr TTCAGCGTGATCTTTACCAAAATattBA f Amplification of the 145-bp empty attB site (OEOE_0506- tRNAGlu) used by IntA-related phages (Fog30/44) CTCCATGGTCAAGTGGCTAAGattBA r AAAGCATCCACCCTTAATTGGattBB f Amplification of a 730-bp empty attB site used by the IntB type CTTTAGGCTTTGTTCGTGGAReg1R phage Φ10MC in strain B10 (OEOE_0851- tRNALeu) TATCAAAATCAAGATCTT TTTCCAAattBC f Amplification of a 337-bp empty attB site (downstream of OEOE_0685- tRNALys) used by IntC-related phages (FogPSU1) AGCAATGCCCCTTTAACTCAattBc r TTGCGCTATACGCACCATAGReg1R Alternative primer used with attBc f to amplify the empty attBC site in strain ATCC BAA1163 TACGAAGATGGTATTAAAGCGGTTAattBD f Amplification of a 342-bp empty attB site (OEOE_1359- tRNALeu) used by the IntD-related phages GGCTAATGTTTGCCGGATTAattBD r CAAGCAGTTCGTCCGTTTTTattBD f bis Alternative primer used with attBD r to amplify the empty attBD site in strain ATCC BAA1163 TGCTGCCAGCTAATAGCAAAjuncD Primer used with attBD f or attBD f bis to amplify the attR junction from lysogens, and with IntDf to amplify the attP

site from intD bacteriophagesATGTGCCCGATAAAACTGTG

Endo F Amplification of a conserved 447–bp in the endolysin sequence GCTTCTAAATGGTCGACTGGEndo R GGCATTTTTTAATCCATTTAAEndoR′ ATAGGGTGKTTTGATTGAATT

333F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

method, primers were designed to detect additional conserved regionsof temperate oenococcal bacteriophages (endolysin, and left attachmentjunction, see materials and methods). Results are shown in Table 2A.Combining the results of our two panels of strains, we observed that22 O. oeni strains out of the 28 analyzed were lysogenic (78.6%),reinforcing the view of awidespread lysogeny in the species. Four strainswere poly-lysogenic and a total of 26 putative O. oeni prophages weretherefore identified. The majority clustered in groups intA (54%) andintD (27%), while intB-associated prophage sequences only accountedfor around 15%. These data are roughly in agreement with Bornemanet al. (2012b) (Table 2B). In contrast, we found a unique intC-related pro-phage among the 26 sequences identified in our study (~4%),while threeout of 12 prophages (25%) harbored the corresponding integrase gene inthe Australian study. Noteworthy, these three identified intC-lysogens(namely AWRIB202, AWRIB304, AWRIB318) have been isolatedfrom natural wine ferments in Australia (Coonawarra, AdelaideHills, and Merbein) and a maximum-likelihood phylogeny basedon whole-genome alignment of 14 O. oeni strains revealed that allthree strains had a very close genetic relationship. The relatednessbetween the intC-lysogens in the panel of 14 strains analyzed mighttherefore explain the value of 25% (Borneman et al., 2012b).

A phylogenetic-tree analysis with site-specific integrases of phagesof related lactic acid bacteria indicated that the four oenococcalintegrase proteins type belong to the same cluster of integrases,which also comprises sequences originating from the Leuconostoc andFructobacillus genera (Fig. 1). Classification in the Leuconostoc genusis rapidly evolving and the four former Leuconostoc species L. fructosum,L. durionis, L. ficulneum, and L. pseudoficulneum have been recentlytransferred to the novel genus Fructobacillus (fructose-loving lacticacid-producing bacillus) (Endo and Okada, 2008; Chelo et al., 2010).As can be seen in Fig. 1, two distinct sub-groups were observed.Integrases from groups A and D were clustered with a prophagesequence from a Leuconostoc isolate recently isolated from Kimchi(Lee et al., 2011). Sequences from groups B and C, which are rare inour panel of strains, were grouped with a prophagic integrase foundin a Fructobacillus fructosus isolate.

3.1.2. Integration sites used by prophagesIntegrase identification can allow prediction of the chromosomal

location of the prophage in O. oeni (Ballestra et al., 2011; Bornemanet al., 2012b). Prominent loci for prophage integration have so far in-volved sequences in, or adjacent to tRNA genes. Hence, intA-intB- andintC-temperate phages have been shown to integrate into a tRNAGlu gene(attBA, OEOE_0506 in PSU-1), a tRNALeu gene (attBB, OEOE_0851), andupstream of a tRNALys gene (attBC, OEOE_0685), respectively (São-Joséet al., 2004; Gindreau et al., 1997; Bon et al., 2009). More recently,prophages harboring intD integrases were found at a tRNALeu gene(attBD, OEOE_1359) (Borneman et al., 2012b) (Fig. 2).

Examination of the sequences flanking the different prophagesidentified in the study confirmed the linking of the tRNA sites andint sequences. Our analysis also revealed a plasticity of the regionssurrounding the attBC and attBD sites in some strains (Fig. 2).Hence, a phage remnant (~ 5.2 kb) flanked by two large DR repeatswas found immediately downstream of the attBC site in seven strains,as previously observed in strain IOEB Sarco 1491 (Bon et al., 2009)(Fig. 2). The attBD site was also preceded by a 7-kb fragment (calledthe gtf region in this paper) in strains ATCC BAA-1163, IOEB 0501 and0502, which consisted of eight putative open reading frames (orfs).Only three orfs displayed a Best Blast Hit (BBH) in the databases, andthe closest homologs were found in Lactobacillus plantarum. Deducedproteinswere related to a putative glycosyltransferase, a transcriptionalregulator, as well as a major facilitator of the EmrB-QacB family (Fig. 2).The coexistence of the gtf region and the phage remnant, with aprophage was observed in strains Sarco S13 and S28, respectively,which suggests that these regions do not exclude any superinfectingphage at the attBC or attBD site.

Table 2Distribution of bacteriophages infecting O. oeni into four groups based on integrasesequence and integration site.

Presence of prophages in a panel of 28 O. oeni strains was assessed by analysis ofincomplete genome data (GD) or PCR analysis (PCR) using primers specific for phagesequences. The distribution is compared to previous data obtained on a panel of 14strains (Borneman et al., 2012b).Prevalence of the different phage groups found in lysogens and in wine samples.

334 F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

3.1.3. Alternative tRNA as potential targets for site-specific recombinationof phages in O. oeni

Previous studies suggested the existence of alternative attB sitesputatively used by Φ10MC (intB) in MCW (described as a tRNALeu,OEOE_0923) and by ΦFog44 (intA) in strains MCW and ML34-C10(tRNAGlu, OEOE_0698), respectively (São-José et al., 2004). Despitethe wealth of genomic data analyzed, our study did not supportthese predictions. At the latter position (OEOE_0698), all strains

analyzed contained a small integrated 400 bp-intA relic, which proba-bly corresponds to traces of ancient infection in the O. oeni species.Another remarkable phage genomic signature corresponding to a13-kb remnant was observed at the tRNASer gene (OEOE_t1213in PSU-1) (Fig. 2). The remnant still encoded a putative functionalintegrase, and the protein sequence formed a separate divergentlineagewithin the phylogenetic tree, demonstrating that it is not closelyrelated to the four major integrase sequences associated with complete

Fig. 2. Location of phage-related sequences in the genome of O. oeni. A) Schematic representation of the genomic environments of the tRNA genes associated with phage remnant(dashed line arrows) and complete bacteriophages (solid line arrows) in the chromosome of O. oeni (adapted from Borneman et al., 2012b). Numbering of genes in strain O. oeniPSU-1 has been used. rRNA operons are also represented (⨀). attBA was found downstream of a tmRNA sequence, which often represents a target for mobile DNAs; attBC and attBBwere downstream of genes encoding s catabolite control protein A homologue (ccpA), and a large mechanosensitive channel (mcsl) which possibly contributes to protection againstextreme turgor in LAB (Folgering et al., 2005), respectively. Lastly, attBD was located between genes encoding a NADPH-quinone reductase and a ribosomal protein. Two differentgenomic environments were observed for attBC and attBD due to the presence of an integron in the vicinity of the bacterial attachment site. Positions of PCR primers used to amplifythe different attB sites in nonlysogenic strains are indicated. Gtf, glycosyl tranferase; metR, putative methicillin resistance protein; MFS, major facilitator (EMrB-Qacb family); qox,NADPH-quinone reductase; ribSU, ribosomal protein; treg, transcriptional regulator (TetR family).

Fig. 1. Phenogram of the oenococcal phage integrases (A to D), and integrases of bacteriophages of closely related lactic acid bacteria. The GenBank accession numbers of thesequences are given in parentheses (max seq difference, 0.95). The Cobalt program package was used to construct a tree by Neighbor–Joining method, using an XerD recombinasefrom L. mesenteroides as an outgroup. The partially annotated draft corresponding to strain AWRIB-429 has the accession number ACSE00000000.

335F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

bacteriophages ofO. oeni (Fig. 1). The remnant targeted thefirst of a clus-ter of 15 tRNA genes. Strikingly, Borneman et al. (2012b) also identified asimilar remnant in this region in six strains, but its integration was pro-posed to occur in the last tRNA (OEOE_1227) (Fig. 2). Sequences flankingthe 13-kb remnant in the six AWRIB strains (202, 304, 318, 429, 568 and576) were examined and data suggest that integration has also occurredin the tRNAser gene (corresponding toOEOE_1213 fromPSU-1) (See Sup-plementary Fig. S2). Borneman et al. (2012b) also mentioned a remnantin a tRNASer (OEOE_0530) in a single strain, namely AWRIB422, whileno phage-related sequenceswere found at this site in the strains analyzedin the present paper (Fig. 2). As a conclusion, seven distinct tRNA lociwere found to contain phage-related sequences in the pangenome of O.oeni, of which only four contain putative complete prophages. No alterna-tive attB sites were found for these complete prophages.

3.2. Replicating intD-bacteriophages

Previous in silico analysis of lysogens did not evaluate in detail thegenomic structure of the resident prophages (Borneman et al.,2012b). In particular, it remains unknown whether intD prophages

can be regarded as simple phage remnants, or if they can excise andplay an active role in O. oeni biology. Moreover, due to the lack of clas-sification scheme until recently, there's so far no indication for thepresence of intD viruses among the various phages infecting O. oenithat have been isolated from wine samples in the past.

We first decided to analyze different red wine samples collected inthe Bordeaux area for the presence of bacteriophages (see Materialsand methods). As often reported, phages were easily detected, since20 out of 22 tested wine samples yielded plaques on at least one ofthe eleven indicator strains used. A total of 17 phages from 12 winesamples were selected based on distinct plaque morphology and hostspectra, and purified. PCR targeting the integrase sequences was usedto rapidly assign newly isolated phages to one of the four main groups(Table 2). We found that all groups of integrase sequences were repre-sented, and seven bacteriophages harbored the intD sequences con-firming that intD-replicating bacteriophages are widely encountered inwine. When the seven intD phages were spotted on the five int-D-lysogens isolated in the study, no plaque formation was observed onthe bacterial lawns. These results suggest that superinfection immunityoccurs during intD phages-prophages crosses.

A)

B)

Fig. 3. Nucleotide sequences promoting the site-specific recombination of intD bacteriophages in O. oeni. A) Sequence of the whole tRNAleu gene. The anticodon is underlined, andthe 11 terminal bases found in the all attachment sites are boxed in grey. B) Alignment of the attP, left and right attachment sites (attL and attL), and attB sequences found in phage,lysogens and nonlysogens, respectively. Nucleotide variations between strains are underlined and numbered (1 to 12). The attP was obtained from phage Φ34-S51 and the MCinduced phages. Left and right attachment sequences were obtained from the genomes of lysogenic strains IOEB-8417, Sarco S11 and S13. The attB sequence was obtainedfrom ten nonlysogenic strains: a identical sequence was present in two strains IOEB 9803 and IOEB 9805; b a reduced 73-bp attBD site was also found in the nonlysogenic strainIOEB 0502 (nonhomologous bases are in bold italics); c sequences also found in strain Sarco S25; d sequences also found in four additional strains IOEB-9517, B16, IOEB-Sarco277 and Sarco S28.

336 F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

We next assessed whether lysogenic oenococcal strains could serveas a reservoir for phages by testing the inducibility of the intD-prophagesdetected in silico. We took advantage of an observation made duringthe survey for phages in wine, indicating that two particular strains(IOEB-Sarco277 and Sarco S25) were both permissive to infection byall isolated phages, and also by Φ10MC (intB group), previously isolatedin our laboratory. Prophage excision was monitored upon mitomycin C(MC) induction of the five intD-lysogens through plaque formation ofboth indicators. Strain Sarco S11 lysed, and this occurred 12–15 h afterMC addition. The growth curves obtained with the four other strains dif-fered slightly, and cells entered a plateau after 12–15 h (SupplementaryFig. S3). The five culture supernatants were tested for the presence ofplaques on strains IOEB-Sarco 277 and Sarco S25, using samples collectedbefore MC addition and after 40 h of contact. No plaques were observedin the samples before addition of MC, showing that lysogeny was stablein all tested strains and that no spontaneous excision occurred. Particleswere detected upon MC addition in all strains and titers were rangingfrom 1 × 102 (Sarco S13/S11) to 1.2 × 103 PFU/ml (IOEB 9803/9805/8417) after 40 h. Subsequent propagation of the MC lysates on the per-missive strain IOEB-Sarco 277 produced high titer lysates containing108 PFU/ml. In view of these results, we concluded that IntD-prophagescan excise themselves, replicate, and be released fromO. oeni cells as par-ticulate DNA. However, all excised phages did not retain their ability tore-lysogenize O. oeni. Hence viruses released from strains IOEB9803/9805/8417 proved to form clear plaques on strain IOEB-Sarco 277, andno lysogens could be isolated. This point warrants attention since the ex-istence of purely lytic phages infecting O. oeni has not been documented,and a careful assessment of the intrinsic inducing factors should now beconducted since lytic phagesmay represent amajor threat in the contextof starter production and stabilization.

3.3. Site-specific recombination of intD prophages in the bacterial genome

Based on in silico analysis, prophages found in IOEB 8417, IOEB9803, and IOEB 9805 were found to share the same architecture andsequence. Consequently only prophages found in strains IOEB-8417,Sarco S11 and Sarco S13 were kept for comparative analysis.

We first focused on the attachments sites which have notbeen so far characterized in detail, because of the nonavailabilityof intD-bacteriophage-related sequences in databanks, preventingcomparisons between attL, attR, attB and attP sequences. Site-specificintegration of intD phages has been previously proposed to occur in atRNALeu sequence (OEOE_t1359), which is flanked by genes encodinga putative quinone oxido-reductase (qox) and a ribosomal protein(ribSU) (Borneman et al., 2012b) (Fig. 2). We found that prophagesfrom strains IOEB 8417 and Sarco S11were flanked by large 140-bp im-perfect repeats, overlapping the 3′ extremity of the tRNALeu sequence,but not the anticodon loop. The integrity of the tRNA gene was pre-served after recombination, and overlapped the attL junction (Fig. 3A).The 140-bp repeat was also found in the genome of ten strains contain-ing no presumptive prophage, nor remnant at that site, and in thegenome of free phages (Φ34–51, an intD-phage isolated from wine,and the MC induced intD-phages). The 140-bp sequence therefore pre-sumably represents the conserved core sequence of the intD-phage at-tachment sites (attB, attP, attL and attR). As shown in Fig. 3B, discretevariations were found in the respective attL, attR and attB sequencesbetween the different strains (noted 1 to 12). However, no allele wasstrictly conserved between one of the junction (left or right) andattP, or attB, and the site where breakage and reunion occur duringInt-dependent recombination could not be localized at this pointof the analysis.

Interesting observations came from the examination of strain SarcoS13 which displayed a half right attachment junction (attR), since onlythe first 73 bp of the 140 bp sequence were present (Fig. 3B). The samesequence also corresponded to the putative attBD site found in thenonlysogenic strains IOEB 0501 and IOEB 0502 (Figs. 2 and 3B). We

speculate that the shorter 73-bp attachment site observed resultedfrom a deletion event at the attB site following the integration of thegtf region (yielding IOEB 0501 and IOEB 0502). This event has been pos-sibly followed by site-specific integration of an intD-phage at the shorterattBD site (yielding Sarco S13). This hypothesiswould suggest that a halfattB site is still functional for phage integration. This scenario could havebeen assessed by testing whether infection of strains IOEB 0501 andIOEB 0502 by an intD-phage yielded turbid plaques, suggesting lysogenyestablishment. Unfortunately, both strains were resistant to all isolatedintD-phages. We can not exclude a second scenario where integrationof the gtf region would have occurred in the 140-bp attR of anintD-lysogen, deleting the attachment sequence (yielding the finalorganization found in Sarco S13). Since excision of the intD prophagewas observed upon MC addition to a culture of strain Sarco S13, aminimal substrate size of 73-bp for attR would then still be efficientfor recombination.

We conclude that site-specific recombination involves a 140-bpsequence in the phage and the bacterial genomes. Experiments withstrain Sarco S13 suggest that the site where breakage and reunionoccur during Int-dependent recombination is localized within the first73 bp of the sequence, and that the reduction of the size of the attBsite does not prevent recombination. Future studies will be importantto assess whether the efficiency of the process is impaired in this case.It is noteworthy that the recombination system found in intD-phagesshows a striking characteristic compared to those found in other tem-perate phages infecting O. oeni, as the core is significantly longer thanthose reported for intA- (17-bp) intB- (15-bp) and intC-phages(20 bp), and LAB phages in general.

3.4. General characteristics of intD prophages

The total length of the prophageswith the two attachment sites was46 kb (IOEB 9805 and Sarco S11), and 43.4 kb (strain Sarco S13). TheirGC content was in the range 38.6–38.9%, which is characteristic of theO. oeni genome as a whole. The genomes of the three distinct intDprophages were annotated after comparison to sequences in currentdatabases. They were predicted to encode 55, 59 and 54 gene products,respectively. Significant matches to orfs contained in databases wereobtained, but they were often restricted to sequences from O. oeni.Biological functions were assigned to only 63 to 69% of these orfs. NotRNA genes were found.

The characterized intD prophages were next compared withthree available complete intD-prophage-sequences found in strainsAWRIB-418, AWRIB 568 and AWRIB-576 (it should be stressed thattheir inducibility has not been not documented yet) (Borneman etal., 2012b). All sequences shared extensive nucleotide identity andgeneral organization (Fig. 4). The genomes were organized into sixmodules. Gene cluster 1 encoded signature proteins implicated in ly-sogeny establishment and control of the lysogeny-lytic switch(integrase, putative repressor). The following replication cluster 2 wasnot easily recognizable. The only protein showing homology withphage transcription regulation proteinswas gp11 (Fig. 4). Gene clusters3 and 4 encoded gene products predicted to play a role in DNA packag-ing (terminase, portal protein), and viral particle formation and assem-bly (tail and head proteins), respectively. Cluster 5was the lysis cassetteinvolved in the host lysis for viral particle release (holin and endolysin).The sixth distal genemodulewasmore flexible (see below). However, itcontained conserved sequences related to a possible superinfection im-munity protein associated with the Lactobacillus prophage Lj965 (Ex-pect 6e-06), as well as a member of the glycosyltransferase family 2(Expect 4e-101).

Three major regions of variability between the prophage genomeswere detected along the genomes, and comprised a set of 34 genes(Table 3). Eleven were orphans and six deduced sequences didnot generate significant BLAST matches to proteins other than those

337F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

of O. oeni. Searches based on PSI-BLAST gave functional clues about asubset of the genes.

The first variable region was localized between the integrationand replication-packaging modules. In strain IOEB 9805, the regionencoded a putative member of the ATP-binding cassette superfamily.Best Blast Hit (BBH) was found with a protein from the lactic acidbacterium Lactobacillus vini, a contaminant of the sugarcane-basedethanol fermentation (De Souza et al., 2012). The sequence was differ-ent in other strains, suggesting exchanges of genes between distinctphages.

The second variable sequencewas localizedwithin the tail-specificityfiber gene. The sequences found in strains Sarco S11, S13 and AWIRB418differed since an internal 500-bp was missing, with no modificationsof the frame. We are currently investigating whether this is altering thephage host range specificities.

The distal genemodule was themost hypervariable region amongstthe five prophages. Of interest within this module was a gene in strainIOEB 9805 (orf52) previously found in the 7-kb phage remnant foundin O. oeni IOEB-Sarco 1491 (Bon et al., 2009). In strains Sarco S11 andAWRIB 418, the distal module contained three genes encoding distinctpossible DNA restriction systems and accompanying modification en-zymes (cytosine methylase). The presence of RM systems in phages isfrequent, playing a relevant role in protection of phage DNA fromdegradation during phage subversion of the host resources and/or thestabilization of the mobile element within the host genome. Thesegene cassettes are frequent targets of horizontal gene transfer betweenbacteria. Accordingly, an internal 740-bp sequence had 67% identitywith a cognate sequence encoding a type II restriction-modificationsystem associated with a 22.4 kb plasmid in Lactococcus lactis (Fallicoet al., 2012). Strain Sarco S13 displayed an intriguing 1700-bp orphansequence in its distal module (between orfs 49 and 50) (Fig. 4). Shorthomologous sequences (113 bp to 283 bp) to different internal

segments were found in four distinct positions of the O. oeni PSU-1 ge-nome. In particular three such regions contained a tRNA gene in the vi-cinity. Comparisons of intD prophages with members of the three othergroups are now needed to determine whether phages represent a re-pository for bacteria DNA in the O. oeni species, and assess whetherthese cargo genes might have been acquired through phage-mediatedtransduction events.

4. Conclusion

We used a classification scheme for oenococcal prophages basedon integrase gene polymorphism and integration site to analyze alarge collection of Oenococcus strains. We confirmed the high preva-lence of lysogeny in the O. oeni species and the existence of fourdistinct groups of temperate bacteriophages. We demonstrated thatthe so far uncharacterized intD group contains inducible phages, andthat members can be isolated from wine samples, supporting therole of these intD bacteriophages in the biology of O. oeni.

Doria et al. (in press) recently described a method for the detec-tion and identification of O. oeni lysogenic strains by amplificationof the endolysin gene sequence. Bacteriophage typification was sub-sequently achieved by RAPD-PCR. Strikingly, this PCR assay basedon the whole lys gene only targeted 25% of the lysogens included inour study, and in particular no intD-phages could be detected. Duringthe course of our study, we also had to design such a PCR assay ona shorter internal region of the endolysin gene, and observed thatabout 90% of the strains were positive. There's therefore a strongpolymorphism of the endolysin gene among O. oeni bacteriophagesand any exhaustive inventory should use additional regions of thephage genomes.

We observed that the prevalence of the four different groupsdiffered markedly. Hence, prophages belonging to the intB- and

Fig. 4. Comparative analysis of the genome content of intD prophages. Alignment of the genetic maps of the intD prophages found in strains O. oeni IOEB 9805 (IOEB 9803 and IOEB8417), Sarco S11, Sarco S13, AWRIB 568/576 and 418. Genes sharing aa sequence similarity are linked by grey shading. Organization of the genome is broadly categorized into sixgenetic modules and the modular structure is indicated by the color code: red-lysogeny/genetic switch genes; blue-DNA replication genes; green-DNA packaging; violet-assemblygenes; red-lysis genes; black-distal module. Probable gene functions identified by bioinformatic analysis are noted next to the orfs. The asterisk represents the attL and attRsequences. The 1700-bp orphan sequence found in the distal module of the prophage from strain Sarco S13 (between orfs 49 and 50, hatched bar) contains four short homologoussequences (113-bp to 283-bp) which are present at different locations in the chromosome of strain PSU-1.

338 F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

intC-groups were less abundant in our panel of strains. Strikingly, wealso observed that free intC-bacteriophages were rare in the winesamples analyzed in our study. These observations raise the questionwhether prophage groups differ regarding stability of lysogeny, and/orsensitivity of released particles towards wine conditions. We have

to keep in mind that our conclusions may have been influenced bymethodological bias because only red wines were tested. More exten-sive work is necessary to analyze a broader sampling size, includingdifferent wine types (red wines, white wines, champagne) and cidersobtained from distinct geographical origins.

To date, bacteriophages are often a neglected reason for sluggishor stuck MLF. Gindreau (1998) introduced an oenophage during aspontaneous fermentation and observed no delay. It was thereforeproposed that the diversity of O. oeni in must may compensate thedestruction of hosts by bacteriophages, preventing any delay or failureof MLF. However, the study has been conducted using a member ofthe intB group (Φ10MC), although such phages were not prevalent inour survey. Considering the unexpected diversity of phages infectingO. oeni, we propose that a reassessment of their impact during winemaking should be now carefully conducted. The major world wineregions have developed various innovations, and in particular manywinemakers now practice inoculated fermentations. This is leading toa paradoxbecause proposed criteria for selection ofMLF starters includethe use of nonlysogens (Torriani et al., 2010) whereas various commer-cial O. oeni strains prove to be lysogens. Another recent development isthe simultaneous inoculation of must with yeast andMLF-bacteria. Thisis expected to allow amore successful induction ofMLF due to a gradualadaptation of bacteria to increasing alcohol concentrations and tothe benefit from higher nutrient availability present in grape musts,compared to the conditions at the end of alcoholic fermentation. It islikely that such conditions may also benefit bacteriophages.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ijfoodmicro.2013.06.032.

References

Arendt, E.K., Lonvaud, A., Hammes, W.P., 1991. Lysogeny in Leuconostoc oenos.Journal of General Microbiology 137, 2135–2139.

Aziz, R.K., Bartels, D., Best, A.A., DeJongh, M., Disz, T., Edwards, R.A., Formsma, K.,Gerdes, S., Glass, E.M., Kubal, M., Meyer, F., Olsen, G.J., Olson, R., Osterman, A.L.,Overbeek, R.A., McNeil, L.K., Paarmann, D., Paczian, T., Parrello, B., Pusch, G.D.,Reich, C., Stevens, R., Vassieva, O., Vonstein, V., Wilke, A., Zagnitko, O., 2008. TheRAST server: rapid annotations using subsystems technology. BMC Genomics 9, 75.

Ballestra, P., Jaomanjaka, F., Dols-Lafargue, M., Le Marrec, C., 2011. Oenophages: revisitingan old problem with new tools. In: Darriet, P., Geny, L., et al. (Eds.), OENO2011:Proceedings of the IX International Symposium in Enology in Bordeaux, Dunod,Paris, pp. 523–528.

Barata, A., Malfeito-Ferreira, M., Loureiro, V., 2012. The microbial ecology of wine grapeberries. International Journal of Food Microbiology 153, 243–259.

Bilhère, E., Lucas, P.M., Claisse, O., Lonvaud-Funel, A., 2009. Multilocus sequence typing ofOenococcus oeni:detection of two subpopulations shaped by intergenic recombination.Applied and Environmental Microbiology 75, 1291–1300.

Boizet, B., Mata, M., Mignot, O., Ritzenthaler, P., Sozzi, T., 1992. Taxonomic characteri-zation of Leuconostoc mesenteroides and Leuconostoc oenos bacteriophage. FEMSMicrobiology Letters 90, 211–216.

Bon, E., Delaherche, A., Bilhère, E., de Daruvar, A., Lonvaud-Funel, A., Le Marrec, C.,2009. Oenococcus oeni genome plasticity is associated with fitness. Applied andEnvironmental Microbiology 75, 2079–2090.

Borneman, A.R., McCarthy, J.M., Chambers, P.J., Bartowsky, E.J., 2012a. Functional divergencein the genus Oenococcus as predicted by genome sequencing of the newly-describedspecies, Oenococcus kitaharae. PloS One 7, e29626.

Borneman, A.R., McCarthy, J.M., Chambers, P.J., Bartowsky, E.J., 2012b. Comparative analysisof the Oenococcus oeni pan genome reveals genetic diversity in industrially-relevantpathways. BMC Genomics 13, 373.

Bridier, J., Claisse, O., Coton, M., Coton, E., Lonvaud-Funel, A., 2010. Evidence of distinctpopulations and specific subpopulations within the species Oenococcus oeni.Applied and Environmental Microbiology 76, 7754–7764.

Cavin, J.F., Drici, F.Z., Prevost, H., Divies, C., 1991. Prophage curing in Leuconostoc oenosby mitomycin C induction. American Journal of Enology and Viticulture 42,163–166.

Chelo, I.M., Zé-Zé, L., Tenreiro, R., 2010. Genome diversity in the genera Fructobacillus,Leuconostoc andWeissella determined by physical and geneticmapping.Microbiology156, 420–430.

Davey, G.P., Ward, L.J.H., Brown, J.C.S., 1995. Characterisation of four Leuconostocbacteriophages isolated from dairy fermentations. FEMS Microbiology Letters128, 21–25.

Table 3Specific sequences associated with intD-prophages.

Strain Gene Deduced protein

Nr Size(bp)

GC%

Size(aa)

BBH (% id; e-value; % residue identity)

IOEB-9805 15 1335 32.4 444 HP from Lb. vini (WP_010579731.1);43% id; 8e−111; 443/444

16 450 29.5 149 HP from Lb. vini (WP_010579732.1);48% id; 3e−25; 122/149

45 1881 28.1 626 AAA domain protein from S. mitis SK579(gb|EID31554.1|);36% id; 3e−98; 544/626

47 810 29.5 269 HP from Lb. casei (WP_003603662.1);29% id; 3e−11; 132/269

52 1020 36.9 339 HP OEOE_0686 O. oeni PSU-1 (YP_810287);97% id; 3e−66; 107/339

53 615 32.3 204 prophage P1 protein 5, superinfectionexclusion (cell surface N-anchored) fromL. plantarum WCFS1 (YP_004888620.1);41% id; 6e−26; 174/204

54 129 25.6 42 None55 201 34.3 66 HP from W. paramesenteroides ATCC 33313

(gb|EER74952.1|); 42% id; 9e−07; 53/66Sarco S11 14 480 35.6 159 None

15 117 23.1 38 None16 951 31.0 316 HP from S. suis D12 (YP_006081640.1);

31% id; 5e−31; 306/31617 564 29.8 187 HP from S. agalactiae (WP_000427215.1);

35%; 1e−15; 155/18718 117 24.8 38 None47 777 32.0 258 HP from C. hathewayi (WP_006780132.1);

40% id; 4e−42;257/25849 327 28.7 108 HP from L. vaginalis (WP_003717196.1);

43% id; 5e−05; 49/10850 858 28.8 285 None51 1260 34.6 419 Putative uncharacterized protein from

Lactobacillus sp.66c (WP_009557716.1);53% id; 6e−149; 419/419

52 1194 36.2 397 Cytosine-specific methyltransferase L. fallax(WP_010007034.1); 64% id, 0; 397/397

53 339 35.7 112 DNA mismatch endonuclease VsrLb. parafarraginis F0439 (gb|EHL95521.1);73% id; 6e−53; 112/112

59 126 32.5 41 NoneSarco S13 15 789 31.3 262 HP from E. faecalis (WP_010709725.1);

55% id; 3e−95; 256/26216 300 31.0 99 HP from Dehalococcoides mccartyi

DCMB5(YP_007483051.1); 30% id;1e−06; 99/99

17 177 27.7 58 None46 432 36.3 143 Glyoxalase/bleomycin resistance protein/

dioxygenase from Lb. acidipiscis(WP_010495368.1); 61% id; 2e−50;132/143

48 759 29.4 252 None49 294 26.5 97 None

AWRIB418 15 720 28.2 239 None16 213 28.6 70 HP from Oenococcus phage fOg44

(CAD19133.1); 87% id; 2e−36; 70/7045 1194 27.5 397 Abi family protein from O. oeni

(WP_002822293.1); 100% id; 0; 397/39746 714 23.8 237 None48 1218 31.0 405 restriction endonuclease StsI from O. oeni

(WP_002822290.1);100% id; 0; 405/40549 921 28.0 306 Modification methylase from O. oeni

(WP_002822289.1); 100% id; 0; 306/30650 1992 28.8 663 Type Is restriction endonuclease from O. oeni

(WP_002822288.1); 100% id; 0; 663/663

Best Blast Hit (BBH) is indicated; nr represents the gene number according to Fig. 4;C, Clostridium; E, Enterococcus; L, Leuconostoc; Lb, Lactobacillus; S, Streptococcus; W,Weissella.

339F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340

Author's personal copy

Davis, C., Silveira, N.F.A., Fleet, G.H., 1985. Occurrence and properties of bacteriophagesof Leuconostoc oenos in Australian wines. Applied and Environmental Microbiology50, 872–876.

de Souza, R.B., dos Santos, B.M., de Fátima Rodrigues de Souza, R., da Silva, P.K., Lucena,B.T., de Morais, M.A. Jr, 2012. The consequences of Lactobacillus vini and Dekkerabruxellensis as contaminants of the sugarcane-based ethanol fermentation. Journalof Industrial Microbiology and Biotechnology 39, 1645–1650.

Divol, B., Tonon, T., Morichon, S., Gindreau, E., Lonvaud-Funel, A., 2003. Molecularcharacterization of Oenococcus oeni genes encoding proteins involved in argininetransport. Journal of Applied Microbiology 94, 738–746.

Doria, F., Napoli, C., Costantini, A., Berta, G., Saiz, J.C., Garcia-Moruno, E., 2013. Developmentof a new method for the detection and identification of O. oeni bacteriophages basedon endolysin gene sequence and RAPD. Applied and Environmental Microbiology.http://dx.doi.org/10.1128/AEM.01307-13 (in press).

Endo, A., Okada, S., 2008. Reclassification of the genus Leuconostoc and proposalsof Fructobacillus fructosus gen. nov., comb. nov., Fructobacillus durionis comb. nov.,Fructobacillus ficulneus comb. nov. and Fructobacillus pseudoficulneus comb. nov. In-ternational Journal of Systematic and Evolutionary Microbiology 58, 2195–2205.

Fallico, V., Ross, R.P., Fitzgerald, G.F., McAuliffe, O., 2012. Novel conjugative plasmidsfrom the natural isolate Lactococcus lactis subspecies cremoris DPC3758: a repositoryof genes for the potential improvement of dairy starters. Journal of Dairy Science 95,3593–3608.

Favier, M., Bilhère, E., Lonvaud-Funel, A., Moine, V., Lucas, P.M., 2012. Identification ofpOENI-1 and related plasmids in Oenococcus oeni strains performing the malolacticfermentation in wine. PloS One 7, e49082.

Folgering, J.H., Moe, P.C., Schuurman-Wolters, G.K., Blount, P., Poolman, B., 2005.Lactococcus lactis uses MscL as its principal mechanosensitive channel. Journal ofBiological Chemistry 280, 8784–8792.

Gindreau, E., 1998. Etudes sur les bactériophages d'O.oeni. phD thesis UniversityBordeaux 2.

Gindreau, E., Torlois, S., Lonvaud-Funel, A., 1997. Identification and sequence analysis of the re-gion encoding the site-specific integration system from Leuconostoc oenos (Oenococcusoeni) temperate bacteriophage phi 10MC. FEMS Microbiology Letters 147, 279–285.

Henick-Kling, T., Lee, T.H., Nicholas, D.J.D., 1986a. Inhibition of bacterial growth andmalolactic fermentation in wine by bacteriophage. Journal of Applied Bacteriology61, 287–293.

Henick-Kling, T., Lee, T.H., Nicholas, D.J.D., 1986b. Characterization of the lytic activity ofbacteriophages of Leuconostoc oenos isolated fromwine. Journal of Applied Bacteriology61, 525–534.

Lee, S.H., Jung, J.Y., Lee, S.H., Jeon, C.O., 2011. Complete genome sequence of Leuconostockimchii strain C2, isolated from Kimchi. Journal of Bacteriology 193, 5548.

Martins, G., Miot-Sertier, C., Lauga, B., Claisse, O., Lonvaud-Funel, A., Soulas, G., Masneuf-Pomarède, I., 2012. Grape berry bacterial microbiota: impact of the ripening processand the farming system. International Journal of Food Microbiology 158, 93–100.

Mills, S., Shanahan, F., Stanton, C., Hill, C., Coffey, A., Ross, R.P., 2012. Movers andshakers: influence of bacteriophages in shaping the mammalian gut microbiota.Gut Microbes 4, 1–13.

Nel, L., Wingfield, B.D., van der Meer, L.J., van Vuuren, H.J.J., 1987. Isolation and charac-terization of Leuconostoc oenos bacteriophages from wine and sugarcane. FEMSMicrobiology Letters 44, 63–67.

Papadopoulos, J.S., Agarwala, R., 2007. COBALT: constraint-based alignment toolfor multiple protein sequences. Bioinformatics 23, 1073–1079.

Parreira, R., São-José, C., Isidro, A., Domingues, S., Vieira, G., Santos, M.A., 1999.Gene organization in a central DNA fragment of Oenococcus oeni bacteriophagefOg44 encoding lytic, integrative and non-essential functions. Gene 226, 83–93.

Poblet-Icart, M., Bordons, A., Lonvaud-Funel, A., 1998. Lysogeny of Oenococcus oeni(syn. Leuconostoc oenos) and study of their induced bacteriophages. CurrentMicrobi-ology 36, 365–369.

Renouf, V., Claisse, O., Lonvaud-Funel, A., 2007. Inventory and monitoring of winemicrobial consortia. Applied Microbiology and Biotechnology 75, 149–164.

Santos, R., São-José, C., Vieira, G., Paveia, H., Santos, M.A., 1998. Genome diversityin temperate bacteriophages of Oenococcus oeni. Archives of Virology 143, 523–536.

São-José, C., Santos, S., Nascimento, J., Brito-Madurro, A.G., Parreira, R., Santos,M.A., 2004. Di-versity in the lysis-integration region of oenophage genomes and evidence for multipletRNA loci, as targets for prophage integration inOenococcus oeni. Virology 325, 82–95.

Sozzi, T., Gnaegi, F., D'Amico, N., Hose, H., 1982. Difficultés de fermentationmalolactique du vin dues à des bactériophages de Leuconostoc oenos. Revue Suissede Viticulture Arboriculture et Horticulture 14, 17–23.

Tenreiro, R., Santos, R., Brito, L., Paveia, H., Vieira, G., Santos, M.A., 1993. Bacteriophagesinduced by mitomycin C treatment of Leuconostoc oenos strains from portuguesewines. Letters in Applied Microbiology 16, 207–209.

Torriani, S., Felis, G.E., Fracchetti, F., 2010. Selection criteria and tools for malolacticstarters development: an update. Annals of Microbiology 61, 33–39.

Weinbauer, M.G., Rassoulzadegan, F., 2004. Are viruses driving microbial diversificationand diversity? Environmental Microbiology 6, 1–11.

Zé-Zé, L., Tenreiro, R., Brito, L., Santos, M.A., Paveia, H., 1998. Physical map of the genomeof Oenococcus oeni PSU-1 and localization of genetic markers. Microbiology 144,1145–1156.

340 F. Jaomanjaka et al. / International Journal of Food Microbiology 166 (2013) 331–340