The E1 beta-subunit of pyruvate dehydrogenase is surface-expressed in Lactobacillus plantarum and binds fibronectin

G

M

Ti

VMa

b

a

ARRAA

KFLAPP

I

wimnStaatttesta2

(

0h

ARTICLE IN PRESS Model

ICRES-25596; No. of Pages 7

Microbiological Research xxx (2013) xxx– xxx

Contents lists available at ScienceDirect

Microbiological Research

jo ur nal ho me p age: www.elsev ier .com/ locate /micres

he E1 beta-subunit of pyruvate dehydrogenase is surface-expressedn Lactobacillus plantarum and binds fibronectin

aleria Vastanoa,1, Marzia Salzilloa,1, Rosa A. Sicilianob, Lidia Muscarielloa,argherita Saccoa, Rosangela Marascoa,∗

Dipartimento di Scienze e Tecnologie Ambientali, Biologiche, e Farmaceutiche, Seconda Università di Napoli, via Vivaldi 43, 81100 Caserta, ItalyCentro di Spettrometria di Massa Proteomica e Biomolecolare, Istituto di Scienze dell’Alimentazione, CNR, Avellino, Italy

r t i c l e i n f o

rticle history:eceived 4 February 2013eceived in revised form 8 July 2013ccepted 18 July 2013vailable online xxx

eywords:ibronectin

a b s t r a c t

Lactobacillus plantarum is among the species with a probiotic activity. Adhesion of probiotic bacteriato host tissues is an important principle for strain selection, because it represents a crucial step in thecolonization process of either pathogens or commensals. Most bacterial adhesins are proteins, and amajor target for them is fibronectin, an extracellular matrix glycoprotein. In this study we demonstratethat PDHB, a component of the pyruvate dehydrogenase complex, is a factor contributing to fibronectin-binding in L. plantarum LM3. By means of fibronectin overlay immunoblotting assay, we identified a L.plantarum LM3 surface protein with apparent molecular mass of 35 kDa. Mass spectrometric analysis

actobacillus plantarumdhesinsDHBrobiotics

shows that this protein is the pyruvate dehydrogenase E1 beta-subunit (PDHB). The corresponding pdhBgene is located in a 4-gene cluster encoding pyruvate dehydrogenase. In LM3-B1, carrying a null mutationin pdhB, the 35 kDa adhesin was not anymore detectable by immunoblotting assay. Nevertheless, the pdhBnull mutation did not abolish pdhA, pdhC, and pdhD transcription in LM3-B1. By adhesion assays, we showthat LM3-B1 cells bind to immobilized fibronectin less efficiently than wild type cells. Moreover, we show

egativ
that pdhB expression is n
ntroduction

The human gut microbiota is characterized by species diversity,hich plays a crucial role in host health, being either a source of

nfection or, conversely, a protection against disease. The attach-ent ability of a strain to the intestinal mucosa is one of the features

ecessary to define a bacterium as probiotic (Oelschlaeger 2010).ome lactobacilli are frequently isolated from the gastrointestinalract of healthy individuals and are classified as probiotics (Elmadfand Meyer 2010; Gareau et al. 2010). However, despite the fact that

considerable amount of data on the probiotic properties of lac-obacilli is available, little is known about mechanisms governingheir interaction with human intestinal cells. Lactobacilli adhesiono mucosal surface is dependent on specific binding of adhesins topithelial cell receptors. Studies on mechanisms mediating adhe-ions show that commensal as well as pathogenic bacteria, can bind

Please cite this article in press as: Vastano V, et al. The E1 beta-subunit of pyand binds fibronectin. Microbiol Res (2013), http://dx.doi.org/10.1016/j.mi

o mucin or to some components of the extracellular matrix (ECM),s fibronectin, collagen I, laminin, and fibrinogen (Kinoshita et al.008; Castaldo et al. 2009; Kline et al. 2009; Munoz-Provencio et al.

∗ Corresponding author. Tel.: +39 0823 274557; fax: +39 0823 274605.E-mail addresses: [email protected], [email protected]

R. Marasco).1 These authors contributed equally to this work.

944-5013/$ – see front matter © 2013 Published by Elsevier GmbH.ttp://dx.doi.org/10.1016/j.micres.2013.07.013

ely regulated by the CcpA protein and is induced by bile.© 2013 Published by Elsevier GmbH.

2009; von Ossowski et al. 2011; Glenting et al. 2013). Fibronectin(Fn) is a large and essential dimeric glycoprotein found in body flu-ids and in the ECM of loose connective tissue. Several pathogenicgram-positive bacteria, including species belonging to the gen-era Streptococcus, Staphylococcus, Mycobacterium, Borrelia, Listeria,Clostridium and Campylobacter, colonize mucous surfaces by bind-ing of specific bacterial adhesins to Fn (Secott et al. 2001; Dramsiet al. 2004; Brissette et al. 2009; Barketi-Klai et al. 2011; Rasigadeet al. 2011; Eucker and Konkel 2012; Maddocks et al. 2012).

Lactobacillus plantarum is usually adopted in food industry asstarter in various fermentations and is among the species witha high probiotic activity, being able to adhere to and to colo-nize intestinal mucosa, thus competing with potentially pathogenicbacteria (van Baarlen et al. 2009; Liu et al. 2010; Lönnermark et al.2010). For this reason much attention has been addressed to thisbacterial species and the complete genome sequence of four L. plan-tarum probiotic strains is now available (Kleerebezem et al. 2003;Zhang et al. 2009; Wang et al. 2011; Axelsson et al. 2012). In L.plantarum, protein-dependent binding to mucus has been reported.Indeed, L. plantarum LA318 glyceraldehyde-3-phosphate dehydro-genase (GAPDH) adheres to human colonic mucin (Kinoshita et al.

ruvate dehydrogenase is surface-expressed in Lactobacillus plantarumcres.2013.07.013

2008) and new insight into the understanding of its cell wall loca-tion has been reported (Saad et al. 2009). Additional proteins ofvarious strains of L. plantarum, involved in binding to mucin andFn, have been identified (Sánchez et al. 2009). Ramiah et al. (2008)

dx.doi.org/10.1016/j.micres.2013.07.013


http://www.sciencedirect.com/science/journal/09445013

http://www.elsevier.com/locate/micres

mailto:[email protected]

mailto:[email protected]


ING Model

M

2 gical R

hgit

L(pdtbgaapsp

M

B

LwMcE8rt

Re

(oap3Cnafr

F

Be(PTdc

P

ard

ARTICLEICRES-25596; No. of Pages 7

V. Vastano et al. / Microbiolo

ave shown that cytoplasmatic enzymes as elongation factor Tu,lyceraldehyde-3-phosphate dehydrogenase and triosephosphatesomerase (TPI) are surface-bound proteins with a role in L. plan-arum 423 adhesion to Caco-2 cells.

In a previous work we demonstrated the involvement of the. plantarum LM3 surface displaced alfa-enolase in Fn-bindingCastaldo et al. 2009). Here we show the identification of another L.lantarum moonlighting protein interacting with Fn, the pyruvateehydrogenase E1 beta-subunit (PDHB), which is a component ofhe pyruvate dehydrogenase enzyme complex (PDH) and is codedy the pdhB gene. Expression of this gene was studied in a simulatedastrointestinal transit (presence of bile) and in the presence orbsence of the CcpA master regulator, which was firstly identifieds the specific repressor of carbon catabolite regulation in gram-ositive bacteria (Titgemeyer and Hillen 2002). The pdhB gene washown to be negatively regulated by CcpA and up-regulated in theresence of bile.

aterials and methods

acterial strains and culture conditions

L. plantarum LM3 (wt), LM3-2 (ccpA1) (Muscariello et al. 2001),M3-CC1 (�enoA1) (Castaldo et al. 2009), and LM3-B1 (this study)ere used throughout this study. L. plantarum was grown inRS medium at 30 ◦C. When needed, erythromycin (5 �g/ml) or

hloramphenicol (10 �g/ml) was added to the MRS medium. The. coli Top10 F′{F [(lacIq Tn10 (TetR)] mcrA �(mrr-hsdRMS-mcrBC)0lacZ�M15 �lacX74 recA1 araD139 �(ara-leu)7697 galU galKpsL (StrR) endA1 nupG} strain was used for plasmid cloning (Invi-rogen).

esolution of Lactobacillus cell wall proteins by two-dimensionallectrophoresis (2DE)

The cell wall fraction was obtained as described previouslyCastaldo et al. 2009). Briefly, 500 �g (wet wt) of cell wall material,btained by ultracentrifugation of French-pressed L. plantarum LM3nd LM3-B1 cells, were suspended in 100 �l of 100 mM Tris–HCl,H 7.4, containing 1% SDS and boiled for 10 min. After cooling,0 �l solubilization buffer, consisting of 7 M urea, 2 M thiourea, 4%HAPS, 1% DTT, 0.2% Bio-Lytes pH 3–10, and 0.002% bromophe-ol blue, was added. The mixture was gently agitated for 30 mint room temperature, followed by TCA precipitation. The cell wallraction was solubilized and 2DE was carried out as previouslyeported (Castaldo et al. 2009).

n overlay assay

The proteins resolved by 2DE, were transferred to an Immuno-lot PVDF membrane (BIORAD Inc.) using the Mini Trans-Blotquipment (BIORAD Inc.), at 35 V, 4 ◦C, for 16 h in Towbin Buffer20 mM Tris base, 192 mM glycin, 20% methanol, pH 8.3). TheVDF membrane was reversibly stained with Ponceau S (Sigma).o detect fibronectin binding, overlay assays were performed asescribed (Castaldo et al. 2009). In parallel, 2DE gels were stained byoomassie brilliant blue, and used for mass spectrometry analysis.

rotein identification by peptide mass fingerprinting


For protein identification mass spectrometry data was searchedgainst the NCBI nr database using the MASCOT search algo-ithm (http://www.matrixscience.com/), and the parameters wereescribed previously (Castaldo et al. 2009).

PRESSesearch xxx (2013) xxx– xxx

Construction of the L. plantarum LM3-B1 mutant strain

Based on the high percentage of identity between the LM3 andthe WCFS1 strains, DNA fragments were amplified from the LM3chromosome using the WCFS1 sequence data (Siezen et al. 2006).

Construction of the LM3-B1 mutant strain by homologousrecombination was carried out with the pUCB1 suicide plasmid,which was constructed as follows. Two fragments of 920 bp and977 bp, localized upstream and downstream the pdhB gene, respec-tively, were amplified from L. plantarum LM3 chromosomal DNAby PCR and cloned into the BamHI–SalI and PstI–HindIII sites ofpUC18Ery (van Kranenburg et al. 1997). These two cloning sitesare placed, respectively, at the 3′-end and at the 5′-end of theery gene, which confers the erythromycin-resistance phenotype.The 920 bp upstream fragment includes the pdhA gene plus thefirst 11 codons belonging to the pdhB coding region, and the977 bp downstream fragment corresponds to the 3′-end of thepdhB gene (last 45 codons) plus the pdhC gene up to the 268thcodon. In the PCRs the oligonucleotides HA1 and HA2 were usedto amplify the 920 bp upstream fragment; HC1 and HC2 wereused to amplify the 977 bp downstream fragment (Table 1). Theresultant pUCB1 plasmid was used to transform L. plantarum asdescribed previously (Muscariello et al. 2001). Integration eventswere selected for growth in the presence of erythromycin. To ana-lyze the recombination events leading to the construction of theLM3-B1 mutant strain, PCR analyses were performed with differ-ent pairs of oligonucleotides (Table 1): HA4, complementary to thesequence spanning from codon 57 to codon 63 of the pdhA gene,and Ery1, complementary to a sequence located downstream theery gene; Ery2, complementary to a sequence located upstreamthe ery gene, and HC2, complementary to the sequence spanningfrom codon 268 to codon 262 of the pdhC gene. With these pairsof oligonucleotides, only PCR amplification of recombinants deriv-ing from a double crossing over event, replacing the pdhB codingregion with the ery gene, would yield both fragments of 1053 bp(HA4/Ery1), and 1076 bp (Ery2/HC2) (data not shown). The samepairs of oligonucleotides were used to PCR amplify chromosomalDNA from the LM3 chromosomal DNA as negative control (data notshown).

RT-PCR analysis

RT-PCR analysis was performed on total RNAs extracted fromLM3-B1, by using the RNeasy Midi Kit (Qiagen, Valencia, California,USA). Total RNAs (0.5 �g) was reverse transcribed to cDNA usingthe QuantiTect reverse transcription kit (Qiagen), which includesa DNase I treatment step. PCR analysis was performed with differ-ent pairs of oligonucleotides: HA3, complementary to the sequencespanning from codon 22 to codon 29 of the pdhA gene, and Ery1were used to verify pdhA expression (Table 1); Ery2 and HC2(described above) were used to verify pdhC expression (Table 1);HC5, complementary to the sequence spanning from codon 404to codon 410 of the pdhC gene, and HDC, complementary to thesequence spanning from codon 258 to codon 251 of the pdhD gene,were used to verify pdhD expression (Table 1). PCR amplificationfrom LM3-B1 chromosomal DNA was also performed as positivecontrol.

Adhesion of bacteria to Fn-coated surfaces

Adhesion assays were performed as described previously(Castaldo et al. 2009). Briefly, L. plantarum LM3 and LM3-B1 were


grown in MRS broth for 12 h at 30 ◦C. Cells were harvested bycentrifugation at 3000 × g for 15 min at 4 ◦C for three times andthe pellets were resuspended in 50 mM Tris–HCl at pH 7.5 andincubated at 30 ◦C for 1 h. Cells were harvested by centrifugation


http://www.matrixscience.com/

ARTICLE IN PRESSG Model

MICRES-25596; No. of Pages 7

V. Vastano et al. / Microbiological Research xxx (2013) xxx– xxx 3

Table 1Primer used in this study.

Primer set Sequence 5′–3′ Size of amplicon (bp)

Primers used for construction and analysis of LM3-B1 mutant strainHA1 CGCGGATCCGAACGGACGTTACATCAACG 920HA2 AACATGGTCGACCGGTGATCGCTTGAATATACG

HC1 AAAACTGCAGGGTGCGCTTTATCTCAGTGC 977HC2 AACCGCAACCAAAGCTTTCACG

HA4 GAACGGACGTTACATCAACG 1053Ery1 TAAATTTGGAAAGTTACACG

Ery2 CACGAACCGTCTTATCTCCC 1076HC2 AACCGCAACCAAAGCTTTCACG

Primers used in RT-PCR analysesHA3 GAGCCCGTGCAAGTATTGGATG 1145Ery1 TAAATTTGGAAAGTTACACG

Ery2 CACGAACCGTCTTATCTCCC 1076HC2 AACCGCAACCAAAGCTTTCACG

HC5 CTGATTGATGGTGCAACGGCCC 867HDC GCGTAGGTAACCGTAACACCAC

Primers used in real-time PCR analysesRecA1 CGCTGCCCAGACGACTATTT 100RecA2 GATTTCCACGATCCGACCAC

R

awwswCtwq2

T

aepiMtwstmu(saaiwcac(cc

HB2 TGCTGAAAAAGGTGCGCTTTHB3 GTAGACCGTATCTGGTGCGTACAC

estriction endonuclease sites are in bold.

nd resuspended in Dulbecco’s modified Eagle’s medium (DMEM)ith 2% fetal bovine serum (FBS, Invitrogen). Microtiter plates (96ells) were coated with Fn (0.25 �g/well) at 4 ◦C overnight and

ubsequently blocked with 2% BSA for 1 h at 37 ◦C. After threeashes with PBST a 100 �l bacterial suspension containing 5 × 108

FU in DMEM was added, and after 2 h of incubation at 37 ◦Che wells were washed three times with PBST. Adherent bacteriaere detached from the wells by adding 100 �l of 10% trypsin and

uantified by real-time PCR as previously reported (Castaldo et al.009).

ranscriptional analysis of the pdhB gene

Total RNAs were extracted and reverse transcribed as describedbove. Transcriptional analysis of PDHB was performed in differ-nt growth conditions: LM3 cells were grown to mid-exponentialhase in MRS medium containing 0.2% glucose, in the absence or

n the presence of bile (3 g/l); LM3 and LM3-2 cells were grown inRS medium containing 2% glucose. We used ox gall as one substi-

ute for human bile because of their similarity. Total RNA (0.5 �g)as reverse transcribed to cDNA using the QuantiTect reverse tran-

cription kit (Qiagen), which includes a DNase I treatment step. Realime PCRs were performed in 7500 Real Time PCR System instru-

ent (Applied Biosystem), using SYBR green technology. Primerssed in this study were designed from the WCFS1 genome sequenceTable 1). The recA gene was used as internal standard for expres-ion analysis. The PCR conditions included an initial denaturationt 95 ◦C for 10 min, followed by 40 cycles of amplification of 15 st 95 ◦C, 1 min at 60 ◦C, and 30 s at 72 ◦C. Fluorescence was mon-tored during each extension phase, and a melting-curve analysis

as performed after each run to confirm the amplification of spe-ific transcripts. Real time PCR analysis was performed in triplicate


nd each assay included standard curves for both the internal-ontrol and target gene, obtained by amplifying serial dilutionsratio 1:10) of the samples. Data were analyzed using the standardurve quantification method (Ramiah et al. 2007). Expression ofontrol samples was normalized to one.

67

Results

Identification of L. plantarum fibronectin-binding surface proteins

To identify putative Fn-binding adhesins, we performed animmunoblot overlay assay of LM3 surface proteins with humanfibronectin. Surface proteins were extracted, resolved by twodimensional electrophoresis, blotted onto a PVDF membrane andtested for Fn binding (Fig. 1). Besides the signal correspondingto the EnoA1 protein, previously identified (Castaldo et al. 2009),we also detected a signal corresponding to a protein with anapparent molecular mass of 35 kDa (Fig. 1B). Using in-gel trypticdigestion and MALDI-TOF analysis, this protein was identified asthe E1 beta subunit (PDHB) of pyruvate dehydrogenase (Fig. 1C).Sequence analysis of the L. plantarum WCFS1 genome (GenBank:A2935263) shows that the pdhB gene is the second gene of a 4-gene cluster probably organized in an operon, which includes theE1 alfasubunit (pdhA), the E2 component (pdhC), and the E3 compo-nent (pdhD) of the pyruvate dehydrogenase multienzyme complex(PDH) (Kleerebezem et al. 2003).

Isolation of the LM3-B1 mutant strain

A strain carrying a null mutation in the pdhB gene was con-structed, using oligonucleotides designed on the L. plantarumWCFS1 genome sequence, on the basis of the high sequence identitybetween LM3 and WCFS1 (Siezen et al. 2006). The pUCB1 plas-mid was constructed by cloning two LM3 chromosomal fragments,flanking the pdhB coding region (see “Materials and methods”),upstream and downstream the ery gene in the pUC18Ery plas-mid. LM3 cells were transformed, and double recombination eventswere screened to identify clones carrying a null mutation in thepdhB gene by PCR analysis. In clones where a double recombinationevent had occurred, leading to substitution of pdhB with ery, ampli-


fication of chromosomal DNA with HA4 and Ery1 oligonucleotidesyielded a fragment of 1053 bp, and amplification with Ery2 andHC2 oligonucleotides yielded a fragment of 1076 bp (Table 1) (see“Materials and methods”); no amplification was detected with




4 V. Vastano et al. / Microbiological Research xxx (2013) xxx– xxx

F d MALDI-TOF analysis. (A) 2DE of surface proteins stained with coomassie brilliant blue.( ed, Castaldo et al. 2009) and PDHB are circled. (C) Identification of PDHB by MALDI-TOFa

eLnOcanwocU(mptapaff

Fig. 3. pdhABCD transcription analysis by RT-PCR in LM3-B1. Amplification was per-formed using specific pair primers (Table 1): HA3-Ery1 for pdhA (lane 2); Ery2-HC2for pdhC (lane 5); HC5-HDC for pdhD (lane 8). Negative controls (PCR without theRT step) for pdhA, pdhC, and pdhD in lanes 3, 6, and 9, respectively. Positive controls(PCR amplifications from LM3-B1 chromosomal DNA) for pdhA, pdhC, and pdhD in

Fd

ig. 1. Identification of Fn-binding surface proteins of L. plantarum LM3 by 2DE anB) Chemifluorescent detection. Spots corresponding to EnoA1 (previously identifinalysis.

ither couple of oligonucleotides when chromosomal DNA fromM3 was used as control (data not shown). The double recombi-ation event occurred in 15% of the erythromycin resistant clones.ne of these clones, hereby named LM3-B1, showing a growth rateomparable to its isogenic wild type strain, was chosen to perform

fibronectin-overlay immunoblotting assay. In this mutant the sig-al corresponding to the 35 kDa adhesin, namely the PDHB protein,as not anymore detected (Fig. 2). To test a putative polar effect

f the pdhB null mutation on pdhC and pdhD expression, a trans-riptional study by RT-PCR was performed on the LM3-B1 strain.sing the oligonucleotide pairs HA3-Ery1, Ery2-HC2, and HC5-HDC

Table 1), PCR amplifications of c-DNA from LM3-B1 yielded frag-ents of 1145, 1076 and 867 bp, corresponding to pdhA, pdhC, and

dhD transcripts, respectively (Fig. 3, lanes 2, 5 and 8). No amplifica-ion was detected when the reverse transcriptase enzyme was notdded to the reaction mixture (Fig. 3, lanes 3, 6 and 9). Expression of


dhA is expected from its own promoter; expression of both pdhC,nd pdhD is likely achieved by read-through transcription, eitherrom the promoter of the pdh operon, or from a putative promoteround on the non-coding strand of the 5′-end ery gene.

ig. 2. Identification of Fn-binding surface proteins of L. plantarum LM3-B1 by 2DE. (A) 2etection. The spot corresponding to EnoA1 is circled.


lanes 4, 7, and 10, respectively. Lane 1: 1-kb ladder (Fermentas).

DE of surface proteins stained with coomassie brilliant blue. (B) Chemifluorescent




V. Vastano et al. / Microbiological Research xxx (2013) xxx– xxx 5

Fit

Afi

pBbrstmbmve

T

tauppesh(asterAbffiC

D

pfLs

Fig. 5. Real time PCR analysis of pdhB expression, normalized to recA expression.

ig. 4. Binding of LM3 (w.t.), LM3-B1 (�pdhB) and LM3-CC1 (�enoA1) to fibronectinmmobilized on microtiter plate wells. Error bars represent ±standard deviation ofhe mean values.

dhesion of wild type and mutant strains to immobilizedbronectin

In order to prove the significance of the E1 beta-subunit for L.lantarum adhesion to Fn, the extent of binding of LM3 and LM3-1 cells was analyzed on Fn-coated microtiter plates. Adhesion ofoth strains was evaluated by real-time PCR, amplifying the 16Sibosomal DNA with species-specific primers (Heilig et al. 2002). Aignificant difference was found in the ability of LM3 and LM3-B1o bind Fn, with wild type cells about 100-fold more efficient than

utant cells (P < 0.05) (Fig. 4). Moreover, LM3-B1 was also found toind less efficiently than the mutant strain LM3-CC1, carrying a nullutation in the enoA1 gene, which codes for EnoA1 �-enolase, pre-

iously demonstrated to be a Fn-binding protein (Fig. 4) (Castaldot al. 2009).

ranscriptional analysis of the pdhB gene

In the human intestine, bile is an important factor, which affectshe viability of microorganisms (Begley et al. 2005). In order tonalyze pdhB expression when cells are exposed to conditions sim-lating the gastro-intestinal transit, a real-time PCR analysis waserformed using total RNAs extracted from LM3 cells grown in theresence of the physiological bile concentration of 0.3%. The pdhBxpression was up-regulated when cells were grown in mediumupplemented with bile; indeed the pdhB expression was 3-foldigher (P < 0.05) in the presence of bile compared to the controlgrown without bile) (Fig. 5, panel A). The pdhB transcription waslso monitored under normal grown condition in the LM3-2 (ccpA1)train, which carries a null mutation in the ccpA gene, encodinghe catabolite control protein A (Muscariello et al. 2001). The pdhBxpression was massively increased (66.7-fold; P < 0.05) in LM3-2espect to LM3 cells, as quantified by real-time PCR (Fig. 5, panel B).

putative and highly conserved carbon catabolite element (CRE),inding site for the CcpA protein (Muscariello et al. 2001), was alsoound upstream of pdhA, in the same position previously identi-ed in L. plantarum WCFS1 (Mazzeo et al. 2012), confirming thecpA-mediated down regulation of the pdhABCD operon.

iscussion

In this report we investigated the adhesive properties of L.


lantarum, a member of the human microbiota, whose probioticeature have been extensively described (van Baarlen et al. 2009;önnermark et al. 2010; Liu et al. 2010). In particular we demon-trated that a L. plantarum surface protein of apparent molecular

The pdhB expression values of LM3 (control samples) are arbitrarily fixed to1. pdhBexpression: (A) in absence and presence of 0.3% bile; (B) in wild type and ccpA nullmutant strain.

mass of 35 kDa is involved in binding to human Fn. Using in-geltryptic digestion and MALDI-TOF analysis, this protein was iden-tified as the E1 beta subunit (PDHB) of pyruvate dehydrogenase,a component of the pyruvate dehydrogenase enzyme complex(PDH). The presence of housekeeping cytoplasmatic enzymes oncell surfaces has been observed in a variety of pathogenic bacteria(Dallo et al. 2002; Bergmann and Rohde 2004) and in some Lacto-bacillus species. Antikainen et al. (2007) demonstrated that GAPDHand ENO are anchored on the cell surface of Lactobacillus crispa-tus, where they interact with lipoteichoic acid. More recently thesurface localization of GAPDH was reported also in some strainsof L. plantarum (Izquierdo et al. 2009; Glenting et al. 2013). In aprevious work, we showed the involvement of the L. plantarumLM3 surface displaced EnoA1 protein in Fn-binding and demon-strated the significance of its surface localization in the mechanismof adhesion to fibronectin (Castaldo et al. 2009). Concerning PDHB,previous studies report its surface localization in Archoleplasmalaidlawii, a non-sterol requiring mollicute (Wallbrandt et al. 1992).Moreover, PDHB acts as a fibronectin and plasminogen bindingprotein in Mycoplasma pneumoniae (Dallo et al. 2002; Thomaset al. 2012). However, similar to other housekeeping enzymesshowing moonlighting functions, no signal peptide responsiblefor protein secretion or hydrophobic membrane-spanning regions,was identifiable in the PDHB sequence (Munoz-Provencio et al.2009).


We isolated the LM3-B1 strain, carrying a null mutation in thepdhB gene. By adhesion assays, we demonstrated that LM3-B1adheres much less efficiently than the wild type strain to Fn-coatedwells. Moreover, this mutant adheres less efficiently than the


ING Model

M

6 gical R

La

paeobpntocsftsPMppiswt2

iaog(crmgspect

nfscnsmoift

A

R

A

A



M3-CC1 mutant strain, carrying a null mutation in enoA1, reporteds a gene coding an Fn-binding protein (Castaldo et al. 2009).

Despite the fact that several data on in vitro adhesion ability ofrobiotics to matrix components are available, very little is knownbout the expression of genes involved in these processes. Knowl-dge about regulatory factors involved in the expression pathwayf adhesins may be important to select strains with improved pro-iotic characteristics. To this aim we analyzed expression of thedhB gene in L. plantarum LM3 and LM3-2 (ccpA1), which carries aull mutation in the ccpA gene, encoding the catabolite control pro-ein A. The pdhB differential expression in LM3 and LM3-2, grownn glucose, indicates that this gene is regulated by a CcpA-mediatedarbon catabolite repression. Analysis of the pdhB regulatory regionhows the presence of a CRE sequence, which is the binding siteor the CcpA protein (Muscariello et al. 2001). Although a puta-ive pdhABCD operon was identified in the L. plantarum genomeequence, it was reported that this bacterium has no detectableDH activity under various growth conditions (Lorquet et al. 2004).oreover, PDH activity was never found in our enzyme assays

erformed either on LM3 or on LM3-2 (data not shown). Indeedroteomic data, previously carried out on L. plantarum WCFS1 and

ts isogenic ccpA null mutant (WCFS1-2), indicated that A, B and Cubunits of PDH were not detectable in the wild type, while theyere expressed in WCFS1-2 (Mazzeo et al. 2012). On the contrary,

he D subunit of PDH was expressed in both strains (Mazzeo et al.012).

During the transit through and the persistence in the gastro-ntestinal tract, bacteria are exposed to several stress conditionss the presence of bile salts in the duodenum. It has been previ-usly demonstrated that expression of the mapA mucus adhesionene was up-regulated by bile and pancreatin in L. plantarum 423Ramiah et al. 2007). Moreover, DNA micro-array-based identifi-ation of bile-responsive genes in L. plantarum WCFS1 has beeneported (Bron et al. 2006). These authors observed that severalembrane-associated proteins were down-regulated by bile sug-

esting prominent changes in cell envelope architecture during biletress. To simulate a stress condition of the upper digestive tract, weerformed transcriptional analysis also in cells grown in the pres-nce of bile salts. We found a 3-fold higher expression of pdhB whenells were grown in medium supplemented with bile compared tohe control.

To our knowledge, this is the first study to show transcriptio-al analysis of the L. plantarum pdhABCD operon and to attribute a

unction to the PDHB subunit. Adhesive properties for the E1 betaubunit of pyruvate dehydrogenase in a bacterium with probioticharacteristics are of interest for the understanding of gut colo-ization processes. Indeed, identification and characterization ofpecific adhesins are crucial to elucidate the molecular mechanismsediating competition between probiotic and pathogenic bacteria

ccurring in the gut. Moreover, identification of regulation factors,nvolved in the expression pathway of adhesion proteins, may beunctional to select strains with improved probiotic characteris-ics.

cknowledgement

This work was partially supported by MIUR-PRIN 2008.

eferences

ntikainen J, Kuparinen V, Lahteenmaki K, Korhonen TK. pH-dependent associa-tion of enolase and glyceraldehyde-3-phosphate dehydrogenase of Lactobacillus


crispatus with the cell wall and lipoteichoic acids. J Bacteriol 2007;189:4539–43.

xelsson L, Rud I, Naterstad K, Blom H, Renckens B, Boekhorst J, et al. Genomesequence of the naturally plasmid-free Lactobacillus plantarum strain NC8 (CCUG61730). J Bacteriol 2012;194:2391–2.

PRESSesearch xxx (2013) xxx– xxx

Barketi-Klai A, Hoys S, Lambert-Bordes S, Collignon A, Kansau I. Role of fibronectin-binding protein A in Clostridium difficile intestinal colonization. J Med Microbiol2011;60:1155–61.

Begley M, Gahan CG, Hill C. The interaction between bacteria and bile. FEMS Micro-biol Rev 2005;29:625–51.

Bergmann S, Rohde M, Hammerschmidt S. Glyceraldehyde-3-phosphate 21 dehy-drogenase of Streptococcus pneumoniae is a surface-displayed plasminogen-22binding protein. Infect Immun 2004;72:2416–9.

Brissette CA, Bykowski T, Cooley AE, Bowman A, Stevenson B. Borrelia burgdorferiRevA antigen binds host fibronectin. Infect Immun 2009;77:2802–12.

Bron PA, Molenaar D, de Vos WM, Kleerebezem M. DNA micro-array-based iden-tification of bile-responsive genes in Lactobacillus plantarum. J Appl Microbiol2006;100:728–38.

Castaldo C, Vastano V, Siciliano RA, Candela M, Vici M, Muscariello L, et al. Surfacedisplaced alfa-enolase of Lactobacillus plantarum is a fibronectin binding protein.Microb Cell Fact 2009;8:14–6.

Dallo SF, Kannan TR, Blaylock MW, Baseman JB. Elongation factor Tu and E1 betasubunit of pyruvate dehydrogenase complex act as fibronectin binding proteinsin Mycoplasma pneumoniae. Mol Microbiol 2002;46:1041–51.

Dramsi S, Bourdichon F, Cabanes D, Lecuit M, Fsihi H, Cossart P. FbpA, anovel multifunctional Listeria monocytogenes virulence factor. Mol Microbiol2004;53:639–49.

Elmadfa I, Meyer AL. Importance of food composition data to nutrition and publichealth. Eur J Clin Nutr 2010;3:S4–7.

Eucker TP, Konkel ME. The cooperative action of bacterial fibronectin-binding pro-teins and secreted proteins promote maximal Campylobacter jejuni invasion ofhost cells by stimulating membrane ruffling. Cell Microbiol 2012;14:226–38.

Gareau MG, Sherman PM, Walker WA. Probiotics and the gut microbiota in intestinalhealth and disease. Nat Rev Gastroenterol Hepatol 2010;7:503–14.

Glenting J, Beck HC, Vrang A, Reimann H, Ravn P, Hansen AM, et al. Anchorless surfaceassociated glycolytic enzymes from Lactobacillus plantarum 299V bind to epithe-lial cells and extracellular matrix proteins. Microbiol Res 2013;168:245–53.

Heilig HG, Zoetendal EG, Vaughan EE, Marteau P, Akkermans AD, de Vos WM. Molec-ular diversity of Lactobacillus spp. and other lactic acid bacteria in the humanintestine as determined by specific amplification of 16S ribosomal DNA. ApplEnviron Microbiol 2002;68:114–23.

Izquierdo E, Horvatovich P, Marchioni E, Aoude-Werner D, Sanz Y, Ennahar S. 2-DE and MS analysis of key proteins in the adhesion of Lactobacillus plantarum,a first step toward early selection of probiotics based on bacterial biomarkers.Electrophoresis 2009;30:949–56.

Kinoshita H, Uchida H, Kawai Y, Kawasaki T, Wakahara N, Matsuo H, et al. Cell surfaceLactobacillus plantarum LA 318 glyceraldehyde-3-phosphate dehydrogenase(GAPDH) adheres to human colonic mucin. J Appl Microbiol 2008;104:1667–74.

Kleerebezem M, Boekhorst J, van Kranenburg R, Molenaar D, Kuipers OP, Leer R, et al.Complete genome sequence of Lactobacillus plantarum strain WCFS1. Proc NatlAcad Sci USA 2003;100:1990–5.

Kline KA, Fälker S, Dahlberg S, Normark S, Henriques-Normark B. Bacterial adhesinsin host-microbe interactions. Cell Host Microbe 2009;5:580–92.

Liu Z, Zhang P, Ma Y, Chen H, Zhou Y, Zhang M, et al. Lactobacillus plantarum preventsthe development of colitis in IL-10-deficient mouse by reducing the intestinalpermeability. Mol Biol Rep 2010;38:1353–61.

Lönnermark E, Friman V, Lappas G, Sandberg T, Berggren A, Adlerberth I. Intakeof Lactobacillus plantarum reduces certain gastrointestinal symptoms duringtreatment with antibiotics. J Clin Gastroenterol 2010;44:106–12.

Lorquet F, Goffin P, Muscariello L, Baudry JB, Ladero V, Sacco M, et al. functional anal-ysis of the poxB gene, which encodes pyruvate oxidase in Lactobacillus plantarum.J Bacteriol 2004;186:3749–59.

Maddocks SE, Lopez MS, Rowlands RS, Cooper RA. Manuka honey inhibits the devel-opment of Streptococcus pyogenes biofilms and causes reduced expression of twofibronectin binding proteins. Microbiology 2012;158:781–90.

Mazzeo MF, Cacace G, Peluso A, Zotta T, Muscariello L, Vastano V, et al. Effect ofinactivation of ccpA and aerobic growth in Lactobacillus plantarum: a protomicperspective. J Proteomics 2012;75:4050–61.

Munoz-Provencio D, Llopis M, Antolin M, de Torres I, Guarner F, Perez-MartinezG, et al. Adhesion properties of Lactobacillus casei strains to resected intesti-nal fragments and components of the extracellular matrix. Arch Microbiol2009;191:153–61.

Muscariello L, Marasco R, De Felice M, Sacco M. The functional ccpA gene is requiredfor carbon catabolite repression in Lactobacillus plantarum. Appl Environ Micro-biol 2001;67:2903–7.

Oelschlaeger TA. Mechanisms of probiotic actions. Int J Med Microbiol2010;300:57–62.

Ramiah K, van Reenen CA, Dicks LM. Expression of the mucus adhesion genesMub and MapA, adhesion-like factor EF-Tu and bacteriocin gene plaA of Lac-tobacillus plantarum 423, monitored with real-time PCR. Int J Food Microbiol2007;116:405–9.

Ramiah K, van Reenen CA, Dicks LM. Surface-bound proteins of Lactobacillusplantarum 423 that contribute to adhesion of Caco-2 cells and their role in com-petitive exclusion and displacement of Clostridium sporogenes and Enterococcusfaecalis. Res Microbiol 2008;159:470–5.

Rasigade JP, Moulay A, Lhoste Y, Tristan A, Bes M, Vandenesch F, et al. Impact of sub-


inhibitory antibiotics on fibronectin-mediated host cell adhesion and invasionby Staphylococcus aureus. BMC Microbiol 2011;11:263–71.

Saad N, Urdaci M, Vignoles C, Chaignepain S, Tallon R, Schmitter JM, et al. Lactobacil-lus plantarum 299V surface-bound GAPDH: a new insight into enzyme cell wallslocation. J Microbiol Biotechnol 2009;19:1635–43.


http://refhub.elsevier.com/S0944-5013(13)00124-9/sbref0005






















































































































































































































































































































































































































































































































































































































































ING Model

M

gical R

S

S

S

T

T

v



ánchez B, Schmitter JM, Urdaci MC. Identification of novel proteins secreted byLactobacillus plantarum that bind to mucin and fibronectin. J Mol MicrobiolBiotechnol 2009;17:158–62.

ecott TE, Lin TL, Wu CC. Fibronectin attachment protein homologue mediatesfibronectin binding by Mycobacterium avium subsp. paratuberculosis. InfectImmun 2001;69:2075–82.

iezen R, Boekhorst J, Muscariello L, Molenaar D, Renckens B, Kleerebezem M.Lactobacillus plantarum gene clusters encoding putative cell-surface proteincomplexes for carbohydrate utilization are conserved in specific gram-positivebacteria. BMC Genomics 2006;7:126–38.

itgemeyer F, Hillen W. Global control of sugar metabolism: a gram-positive solu-tion. Antonie Van Leeuwenhoek 2002;82:59–71.


homas C, Jacobs E, Dumke R. Characterization of pyruvate dehydrogenase sub-unit B and enolase as plasminogen binding proteins in Mycoplasma pneumoniae.Microbiology 2013;159:352–65.

an Baarlen P, Troost FJ, van Hemert S, van der Meer C, de Vos WM, de Groot PJ,et al. Differential NF-kappaB pathways induction by Lactobacillus plantarum in

PRESSesearch xxx (2013) xxx– xxx 7

the duodenum of healthy humans correlating with immune tolerance. Proc NatlAcad Sci USA 2009;106:2371–6.

van Kranenburg R, Marugg JD, van Swam II, Willem NJ, de Vos WM. Molec-ular characterization of the plasmid-encoded eps gene cluster essentialfor exopolysaccharide biosynthesis in Lactococcus lactis. Mol Microbiol1997;24:387–97.

von Ossowski I, Satokari R, Reunanen J, Lebeer S, De Keersmaecker SC, Vander-leyden J, et al. Functional characterization of a mucus-specific LPXTG surfaceadhesin from probiotic Lactobacillus rhamnosus GG. Appl Environ Microbiol2011;77:4465–72.

Wallbrandt P, Tegman V, Jonsson BH, Wieslander A. Identification and analysis ofthe genes coding for the putative pyruvate dehydrogenase enzyme complex in


Acholeplasma laidlawii. J Bacteriol 1992;174:1388–96.Wang Y, Chen C, Ai L, Zhou F, Zhou Z, Wang L, et al. Complete genome sequence of

the probiotic Lactobacillus plantarum ST-III. J Bacteriol 2011;193:313–4.Zhang ZY, Liu C, Zhu YZ, Zhong Y, Zhu YQ, Zheng HJ, et al. Complete genome sequence

of Lactobacillus plantarum JDM1. J Bacteriol 2009;15:5020–1.






















































































































































































































Documents

The E1 beta-subunit of pyruvate dehydrogenase is surface-expressed in Lactobacillus plantarum and binds fibronectin