Characterization of an acquired dps -containing gene island in the lactic acid bacterium Oenococcus...

ORIGINAL ARTICLE

Characterization of an acquired dps-containing gene islandin the lactic acid bacterium Oenococcus oeniA. Athane1, E. Bilhere1, E. Bon1,2, G. Morel1, P. Lucas1, A. Lonvaud1 and C. Le Marrec1

1 UMR 1219, INRA-Universite Victor Segalen Bordeaux 2 – Universite Bordeaux 1, Talence Cedex, France

2 Universite Victor Segalen Bordeaux 2 – LaBRI (Universite de Bordeaux – CNRS) 351, Talence Cedex, France

Introduction

Bacteria have evolved a complex network of stress response

pathways to promote their growth and ⁄ or survival during

environmental challenges. One important mechanism is

the DNA binding protein from starved cells (Dps), first

described in Escherichia coli, where its expression was

found to be activated when the bacterium finds itself in

nutritionally limiting conditions or under oxidative stress,

regardless of whether cells are actively growing or not

(Almiron et al. 1992). Recent evidence expands the role of

Dps in tolerance to acid and osmotic stresses (Choi et al.

2000; Weber et al. 2006), metals, UV-irradiation (Nair and

Finkel 2004), and high pressure (Malone et al. 2006).

The Dps protein from E. coli has a molecular mass of

19 kDa, and is known to be a member of the Fe-binding

protein family. This family forms large (approx.

150 kDa), hexameric complexes in cells that reduce the

intracellular level of iron (Fe2+), thereby preventing the

deleterious effect of the superoxide anion through the

Fenton reaction (Almiron et al. 1992). This combined

action of hydrogen peroxide and intracellular iron is of

paramount importance in vivo since aerobic growth

generates sufficient H2O2 to create toxic levels of DNA

damage in E. coli (Park et al. 2005). Other Dps proteins

such as the cyanobacterial Synechococcus DpsA proteins

also display catalase activity, and can therefore degrade

hydrogen peroxyde (Pena and Bullerjahn 1995). The

DNA binding properties of some members of the Dps

family constitute another protective mechanism. No

apparent sequence specificity has been attributed to them.

By coating the DNA, Dps provides a physical shield from

damaging agents, and prevents mutagenesis (Nair and

Finkel 2004; Park et al. 2005). In addition, the reduction

of the nucleoid surface by Dps-dependent-compaction

was recently suggested as an explanation for the reduced

challenge of reactive oxygen species (Morikawa et al.

2006; Ceci et al. 2007). By modulating DNA structure,

Keywords

Dps, horizontal transfer, malolactic

fermentation, Oenococcus oeni, stress.

Correspondence

Claire Le Henaff-Le Marrec, UMR Œnologie,

Universite Victor Segalen Bordeaux 2 –

Universite Bordeaux 1 (ISTAB)- INRA, 351,

Cours de la Liberation, 33405 Talence Cedex,

France. E-mail: [email protected]

2007 ⁄ 1692: received 22 October 2007,

revised 8 April 2008 and accepted 8 May

2008

doi:10.1111/j.1365-2672.2008.03967.x

Abstract

Aims: To identify novel actors responsible for the marked adaptation of the

Oenococcus oeni species to its environment.

Methods and Results: Genomic surveillance of the available genome sequences

from O. oeni indicated the presence of a small ORF, encoding a protein named

DpsA. The cloned gene complemented the dps) mutant of Escherichia coli and

conferred resistance to hydrogen peroxide, wine, and metals. The dpsA gene

was flanked by IS-related elements. The entire region was characterized by an

anomalously high GC content compared to those reported for oenococcal

genomes. The dpsA gene was present in 15 of the 38 tested isolates. Positive

strains originated from different geographical areas and sources. No change in

tolerance to wine or to oxidative stress was observed between O. oeni strains

harbouring dpsA and those not harbouring this gene.

Conclusions: Some O. oeni have acquired a functional homologue to the dps

gene from E. coli as part of a mobile element.

Significance and Impact of the Study: DpsA probably increases the bacterial

fitness in response to environmental challenges. However, the physiological

condition under which it adds a selective advantage to O. oeni during wine-

making remains to be found.

Journal of Applied Microbiology ISSN 1364-5072

1866 Journal compilation ª 2008 The Society for Applied Microbiology, Journal of Applied Microbiology 105 (2008) 1866–1875

ª 2008 The Authors

Dps may also regulate gene expression (Nair and Finkel

2004).

Dps and homologous proteins have been identified in

distantly-related bacteria facing different challenges in their

environment, suggesting that this protein plays an essential

role in bacterial vitality. This is highlighted by the fact that

Dps is considered one of the protective mechanisms dis-

played by bacterial pathogens against oxidative damage

during phagocytosis (Hasley et al. 2004; Olsen et al. 2005).

Dps has been shown to play a role in bacteriophage toler-

ance (Lacqua et al. 2006) and has recently been proposed

as a candidate protein for adaptation to epiphytic life in

Methylobacterium extorquens (Gourion et al. 2006). Consis-

tent with these pivotal roles, protein levels are thought to

be tightly regulated, with regulatory modes that may differ

according to the bacteria. This is illustrated by the observa-

tion that various transcriptional factors (OxyR ⁄ IHF, Per)

are responsible for dps gene expression in Proteobacteria

and Firmicutes (Morikawa et al. 2006). Dps also undergoes

protease-dependent regulation involving ClpXP and ClpAP

in E. coli (Stephani et al. 2003; Weichart et al. 2003), and

post-translational modifications including phosphorylation

in wall-less E. coli cells (Freestone et al. 1998) and glycosyl-

ation in Salmonella enterica (Hanna et al. 2007). The signifi-

cance of such modifications in vivo remains unknown.

Oenococcus oeni is a lactic acid bacterium responsible for

the malolactic fermentation of wines (MLF). This bacte-

rium is able to cope with a hostile environment including

acid, ethanol, toxic phenolic compounds and nutrient lim-

itation. Activation of membrane-bound H+-ATPases,

modification of membrane fluidity, stress protein synthesis

and exclusion of different cytoplasmic stressors through

multi-drug resistance proteins have been described as

essential mechanisms helping the strain to cope with the

harsh environment (See for review Spano and Massa

2006). The genomes of two strains of O. oeni have been

recently completed (Mills et al. 2005; GenBank accession

number AAUV01000000) and a dps-like gene was identi-

fied in both strains. We present evidence that the encoded

DpsA protein from O. oeni is a functional homologue of

dps from E. coli protecting the cell from H2O2, and suggest

its involvement in the defence of the lactic acid bacterium

against wine and metal stresses. We also posit the role of

dpsA during plasmid establishment. We further show that

dpsA is a mobile gene acquired as part of a mobile element.

Materials and methods

Bacterial strains, plasmids and growth conditions

The 38 O. oeni isolates used in this study are presented in

Table 1. Species identification was carried out by PCR,

using primers OO1 and OO2 targeting the gene encoding

the malolactic enzyme which are specific to O. oeni (Divol

et al. 2003). All strains were cultured in MRS broth pH 4Æ8.

The collection includes 27 wild strains, isolated from vari-

ous types of wines undergoing malolactic fermentation,

each isolate representing a specific vineyard of France

(Delaherche et al. 2006). These strains have been characte-

rized regarding their tolerance to wine, corresponding to

the ability to induce malolactic fermentation after inocula-

tion to wine (oenological potential). Briefly, precultures

grown in modified acidic grape medium were used to ino-

culate three different wines. Microvinifications were done

in duplicate using independent precultures. Assays were

conducted at 20�C and monitored for a three-week period

for their residual malic acid content. Consumption of malic

acid was checked weekly using a thin layer chromatography

method (Roche, Meylan, France). The strains which com-

pleted MLF in the three tested wines (n = 8) were further

considered as isolates of high tolerance (Table 1). Isolates

which did not perform MLF in any assay formed the low

performing group (n = 19). The group of highly perform-

ing strains was enriched with 11 commercial starters (see

Table 1). To avoid redundancy in our collection, the 38 iso-

lates were analysed by pulsed-field gel electrophoresis

(PFGE) using NotI, as previously described (Gindreau et al.

1997). All isolates exhibited distinct PFGE banding pattern

types.

Plasmid pBBR1 Kan carries a replicon from Bordetella

bronchiseptica, and replicates at a low-copy number in

E. coli (Kovach et al. 1995). It was used with the XL1-

Blue strain of E. coli for cloning purposes. The E. coli

dps) ZK1058 (ZK126 dps::kan) and SF2043 (ZK1058

pBR322), and the E. coli dps+ SF2042 (ZK1058 pJE106)

were kindly provided by Dr S. Finkel (Nair and Finkel

2004). They were grown in LB medium and cultures were

incubated at 37�C in rotary shakers for 16 h. When nec-

essary, the following antibiotic concentrations were used:

ampicillin 150 lg ml)1, kanamycin 50 lg ml)1.

Extraction of genomic DNA and subtractive

hybridization

Genomic DNA from O. oeni was isolated as previously

described (Le Marrec et al. 2007). Subtractive hybridiza-

tion was conducted using an adapted protocol of the PCR-

Select Bacterial Genome Subtraction Kit (Clontech), with

DNA from strain IOEB-SARCO 444 as the driver, and an

equimolar mixture of DNAs from strains IOEB-SARCO

277, IOEB-SARCO 384, IOEB-SARCO 450 and VP41 as

the tester. All DNAs were digested with RsaI. Restricted

tester DNA was marked by ligation with oligonucleotide

adapters and hybridized at 63�C with restricted driver

DNA in excess. Self-hybrid tester fragments were amplified

by PCR and nested PCR, using adapter-specific primers.

A. Athane et al. Acquired dps-containing gene island in lactic acid bacterium O. oeni

ª 2008 The Authors

Journal compilation ª 2008 The Society for Applied Microbiology, Journal of Applied Microbiology 105 (2008) 1866–1875 1867

Amplicons were cloned in the pGEM-T Easy Vector (Pro-

mega) and transformed into competent DH5 a cells. Posi-

tive clones were screened on LB medium supplemented

with ampicillin, X-Gal and IPTG. Inserted fragments were

sequenced. The specificity of the selected sequences was

assessed through a PCR assay in driver and tester DNAs.

Analysis of the distribution of dpsA by PCR

Primer design was achieved using the eprimer3 and Oligo

Analyser 1Æ0Æ3 softwares, using data from the genomes of

O. oeni PSU-1 and ATCC BAA-1163. Oligonucleotides

were purchased from Sigma-Aldrich. They are listed in

Table 2. PCR amplifications were carried out using the

primers AX1 and AX2, which are internal to the dpsA

gene from O. oeni PSU-1 (OEOE_1750) and ATCC BAA-

1163 (OENOO_64063). Genetic organization of the dpsA

region was first assessed using the primers AX3 and AX4,

whose sequences are part of the flanking OEOE_1749 and

OEOE_1751 genes, respectively, in O. oeni PSU-1. These

primers are 100% identical to the corresponding regions

in strain ATCC BAA-1163. The following additional sets

Table 1 PCR distribution of dpsA sequences

in a collection of 38 O. oeni strains, which

have been characterized regarding their toler-

ance to wine (high or low) (see Materials and

Methods)

Strains Origin

PCR amplicon

AX 1-2 AX 3-4

High tolerance

IOEB-SARCO 347 Red wine, MLF, La Roquille, France ) )IOEB-SARCO 384 White wine, MLF, Savoie, France ) )IOEB-SARCO 396 White wine, MLF, Jura, France ) )IOEB-SARCO 422 White wine, MLF, Jura, France + +

IOEB-SARCO 433b Red wine, MLF, Cahors, France + +

IOEB-SARCO 438a Red wine, MLF, Arveyres, France + +

IOEB-SARCO 450 Commercial strain, Laffort + +

IOEB-SARCO 451 Red wine, Gironde, France ) )IOEB-SARCO1491 Red wine, France ) )IOEB-SARCO 277 Commercial strain, Laffort + +

B1 Microenos MBR B1, Laffort ) )B16 Microenos B16 Standard, Laffort + +

ExpS Commercial strain, Oeno France ) )ExpC Commercial strain, Oeno France + +

MBO Commercial strain, USA ) )PSU-1 Commercial strain, California, USA + +

VF Vitilactic F, Martin Vialatte ) )VP41 Commercial strain, Lalvin ) )VO Viniflora oenos, CHR Hansen ) )

Low tolerance

IOEB-8413 Red wine, MLF, Gironde, France + +

ATCC BAA-1163 Red wine, MLF, Gironde, France + +

IOEB-SARCO 37 Red wine, MLF, Gironde, France + +

IOEB-SARCO 39a Red wine, MLF, Gironde, France + +

IOEB-SARCO 171 Red wine, MLF, Sadirac, France + +

IOEB-SARCO 399 Red wine, MLF, Vic Fezensac, France ) )IOEB-SARCO 409 White wine, MLF, Charentes, France ) )IOEB-SARCO 425 White wine, MLF, Jura, France ) )IOEB-SARCO 428 Red wine, MLF, Libourne, France ) )IOEB-SARCO 434 Red wine, Portugal + +

IOEB-SARCO 436b Red wine, MLF, Arveyres, France ) )IOEB-SARCO 440 Red wine, MLF, St Emilion, France ) )IOEB-SARCO 441 Floc de Gascogne, France ) )IOEB-SARCO 444 Pineau, France ) )IOEB-SARCO 445 Banyuls, France ) )IOEB-SARCO 447 Red wine, MLF, Bordeaux, France ) )IOEB-SARCO 454 White wine, Chardonnay, Chavannes, France ) )IOEB-SARCO 455 White wine, Chardonnay, Chavannes, France + +

IOEB-SARCO 462 White wine, Bordeaux, France ) )

IOEB, Institut d’Oenologie de Bordeaux.

Acquired dps-containing gene island in lactic acid bacterium O. oeni A. Athane et al.


ª 2008 The Authors

were also used: AX5-AX6, AX5-AX8, AX7-AX8 and AX7-

AX9. PCR with these primers amplified DNA fragments

of 6031-bp, 3600-bp, 8124-bp and 12000-bp, respectively

from O. oeni PSU-1 (Table 2 and Fig. 3). Amplifications

were carried out in a 20 ll reaction volume comprising

1 X iProof PCR Master Mix (Bio-Rad), 1 lmol l)1 for-

ward primer, 1 lmol l)1 reverse primer and 50 ng DNA

template using a iCycler thermal cycler (Bio-Rad).

Cycling parameters comprised an initial denaturation at

95�C for 3 min, 35 cycles of denaturation at 95�C for

30 s, annealing at x�C for 30 s and extension at 72�C for

y min (where annealing temperature x and extension time

y depend on the Tm of the primers and the expected

amplicon size, respectively), and a final extension of 72�C

for 10 min.

Cloning of dpsA and transformation in the E. coli dps)

mutant strain

The dpsA gene was PCR-amplified from the chromosome

of strain O. oeni PSU-1 using the primer pair AX3 and

AX4. The amplicon was cloned using the TOPO-XL

vector in TOP10 cells (Invitrogen). Three clones were

randomly selected and their plasmid DNA was extracted

using the Qiagen kit. The same sequence was obtained for

all three recombinant plasmids, and pAX1 was retained

for investigation. The 1-kb insert was subcloned as a

BamHI–XbaI fragment in the low copy AmpR vector

pBBR1-MCS4, yielding pAX19, which was introduced in

E. coli XL1-Blue using the classical transformation

method with calcium.

Plasmid DNAs from pAX19 and pBBR1-MCS4 were

also introduced in E. coli ZK1058 as described above. To

compare their transformation efficiencies in the dps)

mutant strain, pure plasmid DNAs were first isolated

from E. coli XL1-Blue and photometrically quantified.

Samples showing A260 ⁄ 280 ratio between 1Æ9 and 2 were

used for further investigations. The same quantities

(0Æ1 lg of DNA in 1–2 ll) were introduced in E. coli

XL1-blue and E. coli ZK1058. The ratio between the

transformation efficiencies (number of transformants per

lg DNA) of pAX19 and pBBR1-MCS4 were calculated in

each strain and compared. All assays were performed in

triplicate.

E. coli dps::kan mutant complementation assay and stress

tolerance assay in O. oeni

The parameters of the lethal shocks (duration and inten-

sity of the stressor) were defined using the two E. coli

ZK1058 derivatives containing pBR322 and pJE106 (dps

from E. coli). We selected stress conditions yielding differ-

ences in survival rates greater than 1 log unit after 15 to

30 min in the presence of the stressor. Sensitivity of the

strains containing pBBR1-MCS4 and pAX19 (dpsA from

O. oeni) were then assessed. Oxidative-stress assays were

conducted as follows. First, growth with a suboptimal

H2O2 concentration was tested. After overnight growth,

samples were inoculated in LB with or without H2O2

(4Æ85 mmol l)1) to an initial OD600 of 0Æ1, and were incu-

bated at 37�C for 6 h. Absorbance was recorded every

30 min. Survival to lethal oxidative stress was also

assessed. For this purpose, overnight LB cultures were

diluted 1 : 50 into 10 ml of fresh medium and incubated

at 37�C with shaking for 3h. The stationary growing

E. coli cells were exposed to a concentration of

40 mmol l)1 in H2O2 (obtained from Carlo Erba as a 9Æ7mol l)1 stock solution) in LB. After addition of the stres-

sor (time zero in the experiment), viable-cell counts were

periodically determined. Samples were recovered, serially

diluted in LB medium and immediately plated on LB

plates to determine the number of CFU and to calculate

survivors.

To assess sensitivity to wine, log phase cells grown in

LB broth were diluted to 1 ⁄ 10 in LB diluted with 50%

wine (Saint-Emilion, France). Stationary-phase cultures

were stressed similarly with pure wine. Metal stress assays

were conducted on stationary phase cells in LB containing

FeSO4 100 mmol l)1 or CuSO4 10 mmol l)1 as previously

reported (Nair and Finkel 2004). All assays were

performed in triplicate.

Survival of log phase cells of O. oeni to a lethal oxidative

stress was assessed in MRS containing H2O2 40 mmol l)1.

Survival was determined after 60 min as described above,

except that diluted samples were plates on MRS plates.

Results

DpsA from O. oeni is a functional homologue of Dps

from E. coli

Blast searches using the deduced amino acid sequence of

the Dps protein from E. coli revealed a significant identity

Table 2 Primers used for PCR amplification

Primer Sequence Location

AX1 5¢-TGCCTAGATAAGCACTGATC-3 ¢ OEOE_1750 (dpsA)

AX2 5¢-ATCAACTGATTGCCGATATC-3¢ OEOE_1750 (dpsA)

AX3 5¢-CGCCAGGTTCAAAATGTCTT-3¢ OEOE_1749 (IS1480)

AX4 5¢-TCAATTCGTATTCCCGAAGC-3¢ OEOE_1751 (IS1165)

AX5 5¢-ATCAATCCGGTCACGATAGTTC-3¢ OEOE_1748

AX6 5¢-TCTGGATACAGAAAGGGATCGT-3¢ OEOE_1755

AX7 5¢-CTGATAGTCTTCCGGATGATGC-3¢ OEOE_1745

AX8 5¢-CAGGCAGAGCTTACCTACACACT-3¢ OEOE_1753

AX9 5¢-GGTAGTATCAAGGACCGACTGG-3¢ OEOE_1759


ª 2008 The Authors


(29% identity, 48% similarity) to the OEOE_1750 and

OENOO_64063 gene products from O. oeni PSU-1 and

ATCC BAA-1163, respectively. Both products corre-

sponded to a putative polypeptide of 160 amino acids

(18Æ7 kDa, pI 4Æ56) henceforth named DpsA. This oeno-

coccal sequence displayed best BLAST matches (49% to

67% identity) with dps-like gene products from the lactic

acid bacteria Pediococcus pentosaceus, Lactobacillus fermen-

tum, Lactobacillus plantarum, Enterococcus faecalis, Lacto-

bacillus reuteri and Streptococcus thermophilus. Analysis

also revealed 46% identity to the Listeria monocytogenes

fri gene product, while a more remote homology was

found with MrgA from Bacillus subtilis (59% similarity,

33% identity).

Oenococcus oeni is not amenable to genetic transforma-

tions enabling dpsA gene inactivation. Therefore, to assess

the role of dpsA in-vivo, the gene was amplified from the

chromosome of O. oeni PSU-1, and cloned in the pBBR1-

MCS4 vector. The resulting plasmid was introduced into

the dps::kan mutant (ZK1058) of E. coli (Nair and Finkel

2004). The strain was also transformed with pBBR1-

MCS4 DNA to verify that the vector alone had no effect

on the growth ⁄ survival of the dps::kan mutant strain.

Surprisingly, the two plasmid derivatives (one harbouring

the dpsA gene and one that did not) differed in their

transformation efficiency of the E. coli ZK1058 strain. The

efficiency of the dpsA-containing plasmid was 750 (±50)-

fold higher than that observed for the empty vector,

although both values observed in E. coli XL1-Blue were

similar, yielding a ratio of 1Æ16 ± 0Æ2.

Given the role of Dps-proteins in resistance to oxida-

tive stress, strains ZK1058 (whether harbouring the dpsA

gene or not) were tested for their sensitivity to hydrogen

peroxyde in liquid cultures (Fig. 1a). Growth of the Dps-

deficient strain was not distinguishable from that of strain

ZK1058 harbouring dpsA in the absence of hydrogen per-

oxide. In contrast, its growth was completely impaired in

the presence of 4Æ85 mmol l)1 H2O2, but was restored by

providing dpsA on a plasmid in trans. dpsA also protected

the E. coli cells from a lethal oxidative shock (Fig. 1b).

Taken together, these results imply that dpsA is a func-

tional homologue to the dps gene from E. coli.

Among the most relevant stresses encountered by

O. oeni is the direct inoculation in wine, as practised in

order to increase control over malolactic fermentation.

Therefore, we decided to examine the survival rates to

wine shock of the E. coli dps::kan mutants, both harbou-

ring the O. oeni dpsA gene and without this gene. Results

of assays performed on stationary phase cells are pre-

sented Fig. 2a. A rapid loss of culturability was observed

in the dps::kan mutant containing the empty vector in

wine, since we noted a 5-log unit decrease in survival

after 30 min. In contrast, mutant death by contact with

wine was inhibited in the presence of dpsA. After 30 min,

the viable cells were increased 1000-fold when compared

to the control. Protection did not depend on the growth

phase, since similar results were obtained after log phase

cells were shocked (Fig. 2b).

In E. coli and some other bacterial species, Dps has

been shown to accumulate during the stationary phase

and protect cells from the deleterious effects of various

metals (Nair and Finkel 2004). We chose to test the abi-

lity of dpsA to protect E. coli from lethal concentrations

of CuSO4 and FeSO4 because both compounds are pres-

ent in the environment of O. oeni. The repeated use of

copper ⁄ ferrous fungicides to control downy mildew

(caused by Plasmopara viticola) has indeed been responsi-

ble for the heavy increase of metal concentration in the

upper layers of vineyard soils (Brun et al. 2003; Robinson

10

100(b)

(a)

1·61·8

2

0·001

0·01

0·1

1

Sur

viva

l (%

)

0·20·40·60·8

11·21·4

0·000120151050

06 5 4 3

Time (h)

Time (min)

2 1 0

* * *A

600

Figure 1 The dpsA gene complements the dps::kan mutant of E. coli

during oxidative stress. (a) Growth of the ZK1058 strain of E. coli in

LB medium in the presence (closed symbols) or absence (open sym-

bols) of H2O2 4Æ85 mmol l)1. (¤,)) ZK1058 (pBBR1-MCS4); ( , 4),

ZK1058 (pAX19). The figure represents one experiment out of three.

(b) The presence of the dpsA gene increases the survival of the

dps::kan mutant strain in LB containing H2O2 40 mmol l)1. Stationary

phase cells of E. coli ZK1058 (pBBR1-MCS4) (open symbols) or

ZK1058 (pAX19) (closed symbols) were treated as described in the

materials and methods part of the study. Asterisks indicate no detect-

able cells (limit of detection, <1000 CFU ml)1). The experiment was

conducted in triplicate with less than 10% variation. Representative

data are shown.



ª 2008 The Authors

et al. 2006) and grapes (Mirlean et al. 2005). Results pre-

sented in Fig. 2c,d show to what extent the dpsA-positive

strain’s survival is improved in the presence of both com-

pounds, more so with FeSO4 than CuSO4.

Characterization of the dpsA region in O. oeni

The 480-bp dpsA genes from PSU-1 and ATCC-BAA-

1163 are both characterized by an anomalously high GC

content (40Æ6%) compared to that reported for complete

oenococcal genomes (38% and 37%, respectively). A simi-

lar organization of the dpsA region is found in both

strains (Fig. 3). Hence, dpsA is surrounded by sequences

with a high GC content (41Æ8% and 44Æ1% for the 905-bp

and 1243-bp flanking regions, respectively) which are

similar to transposable elements (99% nucleotide identity)

corresponding to IS1480 from Xanthomonas campestris

(AAD00098) and IS1165 from Leuconostoc mesenteroides

subsp. cremoris (CAA44487). It is noteworthy that

OEOE_1749 and OENOO_64062 both correspond to a

truncated IS1480 copies (Fig. 3). The Blastx analysis of

the upstream sequence suggests the presence of a com-

plete IS1480 which is inactive because of two frameshift

mutations in PSU-1, and one in ATCC BAA-1163. In

contrast, the size of the IS1165 copy present in strains

PSU-1 and ATCC BAA-1163 is identical to that reported

in Leuc. mesenteroides (315 aa), which suggests that the

copy may be functional in both O. oeni isolates.

Members of the IS1480 and IS1165 families have been

described for other lactic acid bacteria. Hence, the 905-bp

sequence corresponding to the IS1480 element (encom-

passing OEOE_1749) shares extensive identity (99%) with

an intergenic plasmid-borne sequence found in p1 from

Lb. casei, as well as a region from plasmid pRH45II from

Lactobacillus brevis. Part of the sequence was recently iden-

tified upstream of the mobile clpL gene in Lactobacillus

rhamnosus (Suokko et al. 2005) (Fig. 3). IS1165-related

elements have been reported in Pediococcus sp., Lactobaci-

llus helveticus, and Lactobacillus casei (Mills et al. 2005),

and more recently upstream of the mobile odc ⁄ potE genes

in O. oeni RM83 (Marcobal et al. 2006) (Fig. 3).

Distribution of the dpsA gene in O. oeni

The logical procedure was to assess how the dpsA

sequence was distributed in the species. Because our labo-

ratory explores the genomic diversity in O. oeni, different

librairies of specific sequences obtained by suppressive

subtractive hybridizations (SSH) between different isolates

were available (Bilhere et al., unpublished data). Their

screening led to the identification of two fragments corre-

sponding to dpsA (OEOE_1750) and its downstream gene

(OEOE_1751) as obtained in an SSH experiment compa-

ring a mixture of four strains (IOEB-SARCO 277, 384,

450 and VP41) vs strain IOEB-SARCO 444. The distribu-

tion of dpsA in the five strains was tested by PCR, using

the internal primers AX1 and AX2. Consistent with the

use of DNA from strain IOEB-SARCO 444 as the driver

in the SSH experiment, no amplicon was observed with

its DNA. This was also true for strains VP41 and IOEB-

SARCO 384. In contrast, strains IOEB-SARCO 277 and

450 tested positive. These results prompted us to test

additional strains, originating from different geographical

areas and sources (Table 1). All isolates exhibited distinct

pulsotypes in PFGE, limiting redundancy in our collec-

tion. A total of 15 O. oeni strains out of 38 (39Æ5%) were

found positive for dpsA internal sequences. The same

scores and distribution were observed using an additional

10

100 100

0·0001

0·001

0·01

0·1

1

0 15 30 45 60

Sur

viva

l (%

)S

urvi

val (

%)

Sur

viva

l (%

)

Sur

viva

l (%

)

* *0·1

1

10

0 20 40 60Time (min)

110

100(c) (d)

(a) (b)

Time (min)

110

100

0·00010·001

0·010·1

0 20 40 60Time (min)

* **

0·00010·001

0·010·1

0 15 20 30Time (min)

*

Figure 2 The presence of the dpsA gene

increases the survival of the dps::kan mutant

strain in wine (a, b), FeSO4 (c) and CuSO4 (d)

stresses. Stationary phase (a–d) or log phase

(b) cells of E. coli ZK1058 (pBBR1-MCS4)

(open symbols) or ZK1058 (pAX19) (closed

symbols) were treated as described in the

materials and methods part of the study.

Asterisks (*) indicate no detectable cells (limit

of detection, <1000 CFU ml)1). The experi-

ment was conducted in triplicate with less

than 10% variation. Representative data are

shown.


ª 2008 The Authors


primer pair (AX3 and AX4) designed in the upstream

and downstream ORFs in O. oeni PSU-1 and ATCC

BAA-1163 (Fig. 3). These results support the ideas (i) that

dpsA flanking sequences are conserved among strains

positive for this trait and (ii) that dpsA is part of a mobile

region encompassing at least OEOE_1749, OEOE_1750

and OEOE_1751. We next tried to assess the outer limits

of the dpsA containing gene island, using primers at

upstream and downstream of OEOE_1749 and

OEOE_1752 (see Fig. 3). Amplicons with the expected

sizes were obtained with all dpsA+ strains. In contrast,

dpsA) isolates yielded no products. These results suggest

that the dpsA variable region is larger than 12-kb.

The distribution of dpsA sequences in the collection

indicated no particular association with specific geograph-

ical areas, nor the type of samples used for bacterial isola-

tion. Lastly, we evaluated the possibility of an association

with enhanced oenological potential. Our collection con-

tained 19 strains previously known as highly tolerant to

wine (HP), including commercial starters, and 19 isolates

which failed at inducing malolactic fermentation after

direct inoculation in wines (LP) (Table 1). An equal dis-

tribution of the dpsA sequences in the HP and LP groups

was observed (8 and 7 strains were dpsA+, respectively).

This is suggesting that dpsA is not critical for survival in

wine. We could draw a similar conclusion regarding the

tolerance to oxidative conditions. Survival to H2O2 was

assessed for 11 O. oeni strains, five dpsA+ and six dpsA)

(Fig. 4). dpsA positive strains were not significantly more

resistant to an oxidative environment.

AX5AX3 AX1/2 AX4

AX6AX7 AX8 AX9

OE

OE

_174

8

OE

OE

_175

0

OE

OE

_174

9

OE

OE

_175

1

OE

OE

_175

2*

OE

OE

_175

3

OE

OE

_175

4

OE

OE

_175

5

OE

OE

_175

6

OE

OE

_175

7

OE

OE

_175

8

OE

OE

_175

9

OE

OE

_174

7

OE

OE

_174

6

OE

OE

_174

5

OE

OE

_174

4LPS synthesisglycosyltransferase

MethioninetRNAsynthase

Permease IS1165 mrr Cysteinesynthase

HP IS1480

dpsA Methyltransferase

Hydrolase Phosphoglyceratemutase

Cystathioninebeta-lyase

HP Pseudogene Serineacetyltransferase

+3

O. oeni ATCCBAA-1163

O. oeni PSU-1+2

+1

+1+3

O. oeni RM83

3128 4497 8866 11103 OE

NO

O_6

4063

OdcIS1165

Plasmid p1(Lb. casei)

Plasmid pRH45II(Lb. brevis)

IS1480

14335 15192

16703 17557 18466 19926ABC

transporterHP

IS1480

Lb. rhamnosusClpL2

167 462 839 2953

IS1480

Figure 3 Schematic overview of the dpsA chromosomal region in O. oeni PSU-1 and ATCC-BAA 1163. The dpsA gene is represented as a

hatched arrow, and flanking IS as grey arrows. Frameshifts in the IS1480 and IS1165 sequences are indicated. The sequence of the leftward

IS-related transposase in strain PSU-1 (OEOE_1752) has been incorrectly predicted, due to an error when locating the translation start site. The

correction of this error is indicated with an asterisk (OEOE_1752*). The positions of the nine primers are also indicated. Gene sequence conserva-

tion between PSU-1 genome and plasmid or cloned chromosomal regions from other lactic acid bacteria is represented by shaded areas. The

nucleotide position of the homologous regions in these sequences is also indicated.



ª 2008 The Authors

Discussion

The mechanisms underlying the marked adaptation of

O. oeni are intriguing in that a wide diversity of resistance

is encountered within the species. Their elucidation may

lead to a better understanding of the evolution of the

species, and possibly to the discovery of new molecular

determinants for strain selection. In this study, we

demonstrated that dpsA is a functional homologue to the

dps gene in E. coli, protecting the cells from the deleterious

effects of hydrogen peroxide (Nair and Finkel 2004). Pro-

tection of E. coli cells from wine stress was also

demonstrated. This work identifies DpsA as a novel candi-

date protein contributing to resistance to wine, which

adds to the complex mechanisms so far described in

O. oeni, including membrane-bound H+-ATPases, FtsH

protease and chaperones or multi-drug resistance proteins

(Spano and Massa 2006). Interestingly, the E. coli dps::kan

mutant was also protected from the presence of metals,

such as copper and ferric ions, in the presence of dpsA

from O. oeni. High concentrations of these ions can

remain on grapes and in must as the result of the repeated

use of copper ⁄ ferrous fungicides to control downy mil-

dew. These metals are also present in wines. However their

concentrations (generally < 10 mg l)1) are not high

enough to inhibit bacterial growth or to affect survival.

Hence, DpsA may be necessary for O. oeni in its survival

to stressful conditions as part of the stress response on

grape berries. Therefore, future studies should address the

role of DpsA as a component for adaptation to the

epiphytic life of O. oeni, and ⁄ or the transition from grapes

to must.

Nair and Finkel (2004) previously observed that the

pBR322 plasmid was cured from an E. coli dps-deficient

strain more rapidly than the plasmid carrying dps during

the stationary phase. Dps was therefore attributed a role

in plasmid maintenance. In the present study, the

presence of dpsA in a plasmid resulted in the higher trans-

formation efficiency of the E. coli dps::kan mutant than

that observed for the plasmid alone. A similar event was

observed, although to a lesser extent, with plasmid pJ06

DNA, a pBR322 derivative harbouring the dps gene from

E. coli (data not shown). Although we cannot explain the

difference in transformation efficiencies, our data may

indicate that Dps proteins also modulate the establish-

ment of plasmid replicons in bacteria, possibly via inter-

actions between Dps and plasmid components (DNA or

proteins involved in replication).

DpsA is a mobile gene in O. oeni and about 40% of the

tested strains harboured the corresponding sequences.

The anomalous GC content of the dpsA region, the varia-

tion in codon usage content and the presence of mobile

sequences in the vicinity strongly suggest a foreign origin

of the locus, possibly through a transposition event. Inter-

estingly, homologues to both IS have been recently associ-

ated with the horizontal transfer of stress response

mechanisms in other LAB (Suokko et al. 2005; Marcobal

et al. 2006). Other Dps-like proteins have been identified

on mobile elements such as the MP1 mega plasmid from

Deinococcus radiodurans (White et al. 1999) or the LP65

bacteriophage from Lb. plantarum (Chibani-Chennoufi

et al. 2004). In contrast, the genetic organization of the

DpsA ortholog-containing regions of other LAB was

different compared to O. oeni. In particular, the absence

of surrounding IS sequences suggests that dpsA are not

mobile genes in these species, or that extended recombi-

nation took place after transfer.

Although the dpsA gene protected E. coli from the

deleterious effects of wine, we could not associate the

presence of the oenococcal sequence with increased fitness

in response to a wine environment. Hence dpsA sequences

were not prevalent in isolates with remarkable oenological

properties (i.e. high survival and growth in wine after

inoculation, accompanied by malate decarboxylation).

This apparent discrepancy can be explained by the fact

that we assessed the presence of a dpsA sequence, which

indicates the presence ⁄ absence of the gene. However,

positive strains may exhibit different DpsA activities

because of variations in the expression level and ⁄ or in the

stability of the protein. These aspects have now to be

examined. The second point is that like in other bacteria,

0·1

1

10

100S

urvi

val

(%)

IOE

B-S

AR

CO

438

aV

OIO

EB

841

3IO

EB

-SA

RC

O14

91IO

EB

-SA

RC

0 43

3b B1

PS

U-1

IOE

B-S

AR

CO

396 VF

IOE

B-S

AR

CO

384

IOE

B-S

AR

CO

422

Figure 4 Tolerance of various O. oeni isolates to oxidative stress. Sur-

vival of log phase cells after exposure to MRS containing H2O2

40 mmol l)1 for 60 min was determined as described in the materials

and methods part. The experiment was conducted in triplicate. dpsA+

and dpsA) strains are represented with hatched and white bars,

respectively.


ª 2008 The Authors


the stress response system to wine or oxidative conditions

in O. oeni is probably complex and can be decomposed

into functional (and maybe compensating) modules. dpsA

may be one of them, even though its contribution to the

overall resistance can be expected to be weak, since it

does not appear critical to survival. Alternatively, a dpsA

deficiency could be complemented by the presence

and ⁄ or enhanced activity of other resistance protein(s), or

by a Dps-paralog. The latter possibility is supported by

the observation that the O. oeni PSU-1 genome appears

to encode a second Dps-like protein. Paired Dps proteins

have been found in Bacillus and Lactococcus lactis but are

rare in other bacteria (Liu et al. 2006). Although these

hypothesis require further investigation, our results stress

the need in the future to address the role of ferritin-like

proteins in helping O. oeni to cope with its environment.

Acknowledgements

E. Bilhere received a grant from the French Ministry of

Education.

References

Almiron, M., Link, A.J., Furlong, D. and Kolter, R. (1992)

A novel DNA-binding protein with regulatory and protec-

tive roles in starved Escherichia coli. Genes Dev 6, 2646–

2654.

Brun, L.A., Le Corff, J. and Maillet, J. (2003) Effects of ele-

vated soil copper on phenology, growth and reproduction

of five ruderal plant species. Environ Poll 122, 361–368.

Ceci, P., Mangiarotti, L., Rivetti, C. and Chiancone, E. (2007)

The neutrophil-activating Dps protein of Helicobacter

pylori HP-NAP adopts a mechanism different from

Escherichia coli Dps to bind and condense DNA. Nucleic

Acid Res 35, 2247–2256.

Chibani-Chennoufi, S., Dillmann, M.L., Marvin-Guy, L.,

Rami-Shojaei, S. and Brussow, H. (2004) Lactobacillus

plantarum bacteriophage LP65 : a new member of the

SPO1-like genus of the family Myoviridae. J Bacteriol 186,

7069–7083.

Choi, S.H., Baumler, D.J. and Kaspar, C.W. (2000) Contribu-

tion of dps to acid stress tolerance and oxidative stress

tolerance in Escherichia coli O157:H7. Appl Environ Micro-

biol 66, 3911–3916.

Delaherche, A., Bon, E., Dupe, A., Lucas, M., Arveiler, B., De

Daruvar, A. and Lonvaud-Funel, A. (2006) Intraspecific

Diversity of Oenococcus oeni Strains Determined by

Sequence Analysis of Target Genes. Appl Microbiol Technol

73, 394–403.

Divol, B., Tonon, T., Morichon, S., Gindreau, E. and Larvaud-

Funel, A. (2003) Molecular characterization of Oenococcus

oeni genes encoding proteins involved in arginine trans-

port. J Appl Microbiol 94, 738–746.

Freestone, P., Grant, S., Trinei, M., Onada, T. and Norris, V.

(1998) Protein phosphorylation in Escherichia L form

NC-7. Microbiology 144, 3289–3295.

Gindreau, E., Joyeux, A., De Revel, G., Claisse, O. and

Lonvaud-Funel, A. (1997) Evaluation of the settling of

malolactic starters within the indigenous microflora of

wines. J Inter des Sciences de la Vigne et du Vin 31, 197–202.

Gourion, B., Rossignol, M. and Vorholt, J.A. (2006) A proteo-

mic study of Methylobacterium extorquens reveals a

response regulator essential for epiphytic growth. Proc Natl

Acad Sci USA 103, 13186–13191.

Hanna, E.S., Roque-Barreira, M.C., Bernardes, E.S., Panunto-

Castelo, A., Sousa, M.V., Almeida, I.C. and Brocchi, M.

(2007) Evidence for glycosylation on a DNA-binding

protein of Salmonella enterica. Microb Cell Fact 6, 11.

Hasley, T., Vasquez-Torres, A., Gravdahl, D., Fang, F. and

Libby, S. (2004) The Ferritin-Like Dps Protein Is Required

For Salmonella enterica Serovar Typhimurium Oxidative

Stress Resistance and Virulence. Infect Immun 72, 1155–

1158.

Kovach, M.E., Elzer, P.H., Hill, D.S., Robertson, G.T., Farris,

M.A., Roop, R.M. and Peterson, K.M. (1995) Four new

derivatives of the broad-host-range cloning vector

pBBR1MCS, carrying different antibiotic-resistance

cassettes. Gene 166, 175–176.

Lacqua, A., Wanner, O., Colangelo, T., Martinotti, M.G. and

Landini, P. (2006) Emergence of biofilm-forming subpop-

ulations upon exposure of Escherichia coli to environmen-

tal bacteriophages. Appl Environ Microbiol 72, 956–959.

Le Marrec, C., Bon, E. and Lonvaud-Funel, A. (2007)

Tolerance to high osmolality of the lactic acid bacterium

Oenococcus oeni and identification of potential osmo-

protectants. Int J Food Microbiol 115, 335–342.

Liu, X., Kim, K., Leighton, T. and Theil, E.C. (2006) Paired

Bacillus anthracis Dps (mini-ferritin) have different reacti-

vities with peroxide. J Biol Chem 281, 27827–27835.

Malone, A.S., Chung, Y.K. and Yousef, A.E. (2006) Genes of

Escherichia coli O157:H7 that are involved in high-pressure

resistance. Appl Environ Microbiol 72, 2621–2671.

Marcobal, A., de Las rivas, B., Moreno-Arribas, M.V. and

Munoz, R. (2006) Evidence for horizontal gene transfer as

origin of putrescine production in Oenococcus oeni RM83.

Appl Environ Microbiol 72, 7954–7958.

Mills, D., Rawsthorne, H., Parker, C., Tamir, D. and Makar-

ova, K. (2005) Genomic Analysis of Oenococcus oeni PSU-1

and its relevance to winemaking. FEMS Microbiol Rev 29,

465–475.

Mirlean, N., Roisenberg, A. and Hies, J.O. (2005) Copper-

based fungicide contamination and metal distribution in

Brazilian grape products. Bull Environ Cont Tox 75,

968–974.

Morikawa, K., Ohniwa, R.L., Kim, J., Maruyama, A., Ohta, T.

and Takeyasu, K. (2006) Bacterial nucleoid dynamics:

oxidative stress response in Staphylococcus aureus. Genes

Cells 11, 409–423.



ª 2008 The Authors

Nair, S. and Finkel, S. (2004) Dps Protects Cells Against

Multiple Stresses During Stationary Phase. J Bacteriol 186,

4192–4198.

Olsen, K.N., Larsen, M.H., Mac Gahan, C., Kallipolitis, B.,

Wolf, X.A., Rea, R., Hill, C. and Ingmer, H. (2005) The

Dps-like Fri of Listeria monocytogenes promotes stress

tolerance and intracellular multiplication in macrophage-

like cells. Microbiology 151, 925–933.

Park, S., You, X. and Imlay, J.A. (2005) Substantial DNA

damage from submicromolar intracellular hydrogen

peroxide detected in Hpx) mutants of Escherichia coli.

Proc Natl Acad Sci USA 102, 9317–9322.

Pena, M.M. and Bullerjahn, G.S. (1995) The DpsA protein of

Synechococcus sp. Strain PC C7942 is a DNA-binding

hemoprotein. Linkage of the Dps and bacterioferritin

protein families. J Biol Chem 38, 22478–22482.

Robinson, B., Greven, M., Green, S., Sivakumaran, S.,

Davidson, P. and Clothier, B. (2006) Leaching of copper,

chromium and arsenic from treated vineyard posts in

Marlborough, New Zealand. Sci Tot Environ 364, 113–123.

Spano, G. and Massa, S. (2006) Environmental stress response

in wine lactic acid bacteria: beyond Bacillus subtilis. Crit

Rev Microbiol 32, 77–86.

Stephani, K., Weichart, D. and Hengge, R. (2003) Dynamic

Control of Dps Protein Levels by ClpXP and ClpAP

Proteases in Escherichia coli. Mol Biol 49, 1605–1614.

Suokko, A., Savijoki, K., Malinen, E., Palva, A. and Varmanen,

P. (2005) Characterization of a mobile clpL gene from

Lactobacillus rhamnosus. Appl Environ Microbiol 71, 2061–

2069.

Weber, A., Kogl, S.A. and Jung, K. (2006) Time-dependent

proteome alterations under osmotic stress during aerobic

and anaerobic growth in Escherichia coli. J Bacteriol 188,

7165–7175.

Weichart, D., Ouerfurth, N., Dreger, M. and Hengge-Aronis,

R. (2003) Global control for ClpP-containing proteases in

stationary-phase adaptation of Escherichia coli. J Bacteriol

185, 115–215.

White, O., Eisen, J.A., Heidelberg, J.F., Hickey, E.K., Peterson,

J.D., Dodson, R.J. and Haft, D.H. et al. (1999) Genome

sequence of the radioresistant bacterium Deinococcus

radiodurans R1. Science 286, 1571–1577.


ª 2008 The Authors


Characterization of an acquired dps -containing gene island in the lactic acid bacterium Oenococcus...

Documents

Transcript of Characterization of an acquired dps -containing gene island in the lactic acid bacterium Oenococcus...