I

Vd

SHa

Rb

T

a

ARR1A

KVVTCI

1

t(n

CEtimRpsj

eS

h1

ARTICLE IN PRESSG ModelJMM-51011; No. of Pages 10

International Journal of Medical Microbiology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Medical Microbiology

j ourna l h o mepage: www.elsev ier .com/ locate / i jmm

vpE mediates the intestinal colonization of Vibrio vulnificus by theisruption of tight junctions

ei-Jung Leea, Young Hyun Junga, Jung Min Ryua, Kyung Ku Jangb, Sang Ho Choib,o Jae Hana,∗

Department of Veterinary Physiology, College of Veterinary Medicine, Research Institute for Veterinary Science, and BK21 PLUS Creative Veterinaryesearch Center, Seoul National University, Seoul, South KoreaNational Research Laboratory of Molecular Microbiology and Toxicology, Department of Agricultural Biotechnology, and Center for Food Safety andoxicology, Seoul National University, Seoul, South Korea

r t i c l e i n f o

rticle history:eceived 16 July 2015eceived in revised form7 September 2015ccepted 26 October 2015

eywords:ibrio vulnificusvpEight junctionsolonization

a b s t r a c t

The disruption of gastrointestinal tight junctions and their colonization evoked by enteric pathogens arehallmarks of the pathogenesis. Vibrio (V.) vulnificus, VvpE, is an elastase which is responsible for hostsurface adherence and vascular permeability; however, the functional roles of VvpE in the pathogenesisof V. vulnificus (WT) are poorly understood. In the present study, we have investigated the role of VvpE inregulation of intestinal tight junctions and the colonization of WT. We found that mutation of the vvpEgene from V. vulnificus (vvpE mutant) prevents intestinal tight/adherens junction dysregulation due toa WT infection and maintains the physiological level of the epithelial paracellular permeability. Inter-estingly, the vvpE mutant exhibited defective intestinal colonization abilities, whereas WT colonizationwas significantly elevated in the ileum in a time-dependent manner. Finally, the vvpE mutant negatedthe enterotoxicity, the breakdown of red blood cells, and pro-inflammatory responses, all of which are

ntestinal epithelial cells induced by the WT infection. In addition, the results of a LC–MS/MS analysis showed that VvpE con-tributes to WT pathogenesis in multiple ways by interacting with intestinal proteins, including �-globin,Annexin A2, Annexin A4, F-actin, and intelectin-1b. These results demonstrate that VvpE plays importantrole in promoting the tight junction disruption and intestinal colonization of V. vulnificus and that it alsohas the ability to interact with the intestinal proteins responsible for microbial pathogenesis.

. Introduction

Enteric bacterial pathogens have various bacterial infec-

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

ious stratagems to circumvent the epithelial barrier of the gutAshida et al., 2012). Specifically, impairment of the intesti-al tight/adherens junction (TAJ) during bacterial infection

Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase;FU, colony-forming unit; Cont, control; EDTA, ethylenediaminetetraacetic acid;GTA, ethylene glycol tetraacetic acid; FITC, Fluorescein isothiocyanate; H&E, hema-oxylin and eosin; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; Ig,mmunoglobulin; LB, Luria Bertani; LC–MS/MS, liquid chromatography–tandem

ass spectrometry; PBS, phosphate-buffered saline; PVDF, polyvinylidene fluoride;OD, relative optical density; ROS, reactive oxygen species; rVvpE, recombinantrotein VvpE; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophore-is; SD, standard deviation; TBST, tris-buffer solution-tween 20; TAJ, tight/adherensunction; TLRs, toll-like receptors; Veh, vehicle.∗ Corresponding author at: Department of Veterinary Physiology, College of Vet-rinary Medicine, Seoul National University, Gwanak-ro, Gwanak-gu, Seoul 08826,outh Korea.

E-mail address: hjhan@snu.ac.kr (H.J. Han).

ttp://dx.doi.org/10.1016/j.ijmm.2015.10.006438-4221/© 2015 Elsevier GmbH. All rights reserved.

© 2015 Elsevier GmbH. All rights reserved.

facilitates the invasion of bacteria to promote colonization andenhance epithelial cell-microbiota interactions, evoking hostprotective/stress responses such as multiple pro-inflammatoryresponses (Ashida et al., 2012). Vibrio (V.) vulnificus is an anaerobicGram-negative bacterium which often causes lethal infections inhumans who are immunocompromised or who have underlyingdiseases such as cirrhosis of the liver (Blake et al., 1979; Park et al.,1991). The most prominent aspect of V. vulnificus pathogenesisis its ability to infect a host via the gastrointestinal tract, afterwhich it rapidly spreads from the small intestine to the bloodstream (Jeong and Satchell, 2012). Thus, studies regarding thefactors leading to the dysregulation of TAJ are likely to be criticalfor uncovering the mechanisms of the pathogenesis of V. vulnificusin the intestine. Many pathogens have been found to cause TAJdysregulation either as a consequence of host infections or byproducing toxic products (Berkes et al., 2003). Fragilysin is an

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

example of toxin produced by Bacteroides fragilis that disruptsthe TAJ barrier by the proteolytic degradation of E-cadherin (Wuet al., 1998). The pathogenic proteins known as toxins A and Bof Clostridium difficile have also been shown to regulate protein

ING ModelI

2 f Med

kTtsa

apm(aatVfPd1aoVnrV

hFmhtefltBetscitv

inpv

2

2

pTEB(TapIRCor

ARTICLEJMM-51011; No. of Pages 10

S.-J. Lee et al. / International Journal o

inase C and the Rho GTPase pathway, respectively, in promotingAJ dysregulation (Berkes et al., 2003). Therefore, the identifica-ion of functional toxins of bacterial pathogens which alter thetructure and function of the intestinal TAJ barrier could providen important therapeutic strategy for bacterial infections.

Although the majority of the virulence effects of V. vulnificusre known to derive from secreted toxins encoded by cytolyticore-forming hemolysin (VvhA) (Jeong and Satchell, 2012) andultifunctional autoprocessing repeats in the toxin (MARTXVv)

Jeong and Satchell, 2012; Lee et al., 2007a), several other secretednd cell-associated factors of V. vulnificus have also been proposeds potential virulence determinants which are also involved inhe pathogenesis of V. vulnificus. A 45 kDa elastase designated as. vulnificus VvpE is considered to be another possible virulence

actor of V. vulnificus (Kothary and Kreger, 1987; Miyoshi, 2006).urified VvpE was proven to induce hemorrhagic damage andermonecrosis (Kothary and Kreger, 1987; Miyoshi and Shinoda,992; Molla et al., 1989) and was shown to cause tissue necrosisnd increased vascular permeability, thus facilitating the invasionf this bacterium (Jones and Oliver, 2009; Miyoshi, 2006). WhilevpE is the major elastase produced by V. vulnificus, there areo previous reports related to the functional role of VvpE in theegulation of intestinal TAJ dysregulation and the colonization of. vulnificus.

On the other hand, many types of pathogenic bacteria mimicost cell ligands to transduce signals into host cellular responses.or example, internalin, a bacterial surface protein of Listeriaonocytogenes, is known to interact with E-cadherin to invadeepatocytes (Mengaud et al., 1996). Invasin, a pathogenic pro-ein which is the major adhesion and invasion factor of Yersinianterocolitica (Isberg and Leong, 1990), along with IpaB of Shigellaexneri (Skoudy et al., 2000), have also been shown to interact withhe mammalian receptors �5�1 integrin and CD44, respectively.y binding to the host receptors, these pathogens have profoundffects on host cellular processes. Interestingly, VvpE plays impor-ant roles in the surface adherence of V. vulnificus by facilitatingwarming, and it regulates the invasiveness of V. vulnificus byleaving the host IgA and lactoferrin (Kim et al., 2007). Thus, it ismportant to determine whether VvpE interacts with the essen-ial host cell surface proteins responsible for the adherence andirulence effect of V. vulnificus within cells.

In this study, therefore, we investigate the essential role of VvpEn the regulation of the intestinal TAJ dysregulation and the colo-ization of V. vulnificus and further identify the potential intestinalroteins interacting with VvpE which are responsible for the acti-ation of multiple cellular signaling pathways.

. Material and methods

.1. Chemicals

Fluorescein isothiocyanate (FITC)-labeled 4-kDa dextran wasurchased from Sigma Chemical Company (St. Louis, MO, USA).he following antibodies were purchased: Claudin 1/2, Occludin,-cadherin, �-actin and Connexin43 antibodies (Santa Cruziotechnology, Paso Robles, CA, USA); intelectin-1b antibodyAbcame Cambridge, MA, USA); F-actin antibody (Cell Signalingechnology, Danvers, MA, USA); Annexin A2 and Annexin A4ntibodies (BD Biosciences, Franklin Lakes, NJ, USA); Horseradisheroxidase (HRP)-conjugated goat anti-rabbit and goat anti-mouse

gG antibodies (Jackson Immunoresearch, West Grove, PA, USA);

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

abbit anti-VvpE antibody was kindly provided by Prof. Sang Hohoi (Seoul National University, Korea). All other reagents weref the highest purity commercially available and were used aseceived.

PRESSical Microbiology xxx (2015) xxx–xxx

2.2. Ethics statement

All animal procedures were performed following the NationalInstitutes of Health Guidelines for the Humane Treatment of Ani-mals, with approval from the Institutional Animal Care and UseCommittee of Seoul National University (SNU-140108-4). Animalsof only male were used in this study. All surgery was performedunder zoletil-xylazine, and all efforts were made to minimize suf-fering.

2.3. Bacterial strains, plasmids, and culture media

All V. vulnificus strains (M06-24/O WT and M06-24/O vvpE) areisogenic and naturally resistant to polymyxin B (Table 1). Unlessotherwise noted, V. vulnificus strains were grown in Luria Bertani(LB) medium supplemented with 2.0% (wt/vol) NaCl (LBS) at 30 ◦C.All media components were purchased from Difco (Difco Labora-tories Inc, Detroit, MI). V. vulnificus were grown to mid-log phase(A600 = 0.500) corresponding to 2 × 108 CFU/mL and centrifuged at6000 × g for 5 min. The pellet was washed with phosphate bufferedsaline (PBS) and adjusted to desired colony-forming unit (CFU)/mLbased on the A600 determined using a UV–VIS spectrophotometer(UV-1800, Shimadzu, Japan) to estimate culture density.

2.4. Intestinal paracellular permeability and colonization assay

The intestinal paracellular flux was determined by examiningthe apical-to-basolateral flux of fluorescein isothiocyanate-labeled4-kDa dextran (FITC-dextran, 1 mg/mL). Seven-week-old ICR mice(n = 10) inoculated intragastrically (i.g.) with 100 �L of the boiledV. vulnificus at 100 ◦C for 30 min (Cont), V. vulnificus (WT), anda mutant deficient in vvpE gene in V. vulnificus (vvpE mutant)at 1.1 × 109 CFU/mL, and killed at 4 h, 8 h, and 16 h later. Theimmunogenicity of Cont in TAJ disruption, bacterial coloniza-tion, enterotoxicity, and intestinal inflammatory responses wasnot found for 16 h, compared to the non-treated mice (data notshown). It is noted that the mice were given FITC-labeled 4 kDadextran in PBS (20 mg/100 �L) by oral injection for 4 h prior tothe sacrifice. A blood sample was collected from caudal venacava, and the plasma was taken for measuring the concentra-tion of FITC-dextran. The fluorescence intensity of the plasmawas examined with a Victor3 luminometer (Perkin-Ehmer Inc.,Waltham, MA, USA) using 488 nm excitation and 515 nm emission.Standard curves were generated by serial dilution of FITC-dextranin PBS. On the other hand, mouse colonization assays were per-formed essentially as described in earlier work (Kim et al., 2013).Mice given an i.g. inoculation of WT, boiled WT (Cont), and vvpEmutant were sacrificed and intestine, colon, spleen, and liverof each mouse were collected, washed, and homogenized. Thehomogenates of each organ were serially diluted and spread on LBagar containing polymyxin B (100 U/mL). CFUs were normalized tograms of intestinal tissues (CFU/g) to represent superficial bacterialcounts.

2.5. Complementation of the vvpE mutant

To complement the vvpE mutation, an open reading frame(ORF) of vvpE was amplified from the genomic DNA of V. vulnificusMO6-24/O by PCR with the primer pair VVPE001F and VVPE001R(Supplementary Table 1) and then digested with BamHI. The ampli-fied vvpE ORF was subcloned into the broad-host-range vector

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

pRK415 (Keen et al., 1988) linearized with the same enzyme toresult in pKK1450 (Table 1). Escherichia coli S17-1�pir, tra strain(Simon et al., 1983) containing pKK1450 was used as a conjugaldonor to vvpE mutant. The plasmid pKK1450 was delivered into

ARTICLE IN PRESSG ModelIJMM-51011; No. of Pages 10

S.-J. Lee et al. / International Journal of Medical Microbiology xxx (2015) xxx–xxx 3

Table 1Plasmids and bacterial strains used in this study.

Strain or plasmid Relevant characteristicsa Reference or source

Bacterialstrains

V. vulnificus M06-24/O Clinical isolate; virulent; WT Laboratory collectionCMM111 MO6-24/O vvpE::pKC9844; elastase deficient; vvpE mutant Jeong et al., 2000

E. coli S17-1�pir �-pir lysogen; thi pro hsdR hsdM+ recA RP4-2Tc::Mu-Km::Tn7;Tpr Smr; host for �-requiring plasmids;conjugal donor

Simon et al., 1983

PlasmidspRK415 IncP ori; broad-host-range vector, oriT of RP4; Tcr Keen et al., 1988

ith vv r

tant.

t2

2

iwBataT

2

(wip(VPRar5

2s

wls(utiswDCWaoUresSt

pKK1450 pRK415 w

a Tpr, trimethoprim resistant; Smr, streptomycin resistant; Tcr, tetracycline resis

he vvpE mutant by conjugation as described previously (Lim et al.,014).

.6. Hematological analysis and general toxicity tests

Routine complete blood count (CBC) in mice (n = 6) given an.g. inoculation of WT, boiled WT (Cont), and vvpE mutant for 16 h

as analyzed by using a hematological autoanalyzer (AVIDA 2120,ayer Diagnostics, Giessen, Germany). The serum test for aspartateminotransferase (AST). alanine aminotransferase (ALT), and lac-ate dehydrogenase (LDH) were measured by using a biochemistryutoanalyzer (Hitachi 7180 autoanalyser, High-Technologies Corp.,okyo, Japan).

.7. Immunoprecipitation

Mice (n = 6) were given i.g. inoculation of WT, boiled WTCont), and vvpE mutant at 1.1 × 109 CFU/mL for 16 h. The miceere sacrificed and ileal tissue was lysed with denaturing radio-

mmunoprecipitation assay (RIPA) buffer containing proteases andhosphatases inhibitors at 4 ◦C as described in detail previouslyYun et al., 2014). The lysates (400 mg) were mixed with anti-vpE antibody. The samples were incubated for 4 h, mixed withrotein A/G PLUS-agarose immunoprecipitation reagent (Pierce,ockford, IL, USA), and then incubated for an additional 12 h. Thegarose beads were washed five times, and the bound proteins wereeleased from the beads by boiling in SDS-PAGE sample buffer for

min. Samples were analyzed by LC–MS/MS or Western blotting.

.8. Proteomic analysis by liquid chromatography-tandem masspectrometry (LC–MS/MS)

Immunoprecipitated intestinal proteins with VvpE antibodyere subjected to SDS-PAGE and stained with Coomassie Bril-

iant Blue. The indicated stained protein bands were excised andubjected to in-gel tryptic digestion as described by Heo et al.2007). Eluted peptide mixture was purified using C18 spin col-mn (Pierce Biotechnology, Thermo Scientific, IL, USA) according tohe manufacturer’s instruction. Each sample was then resuspendedn 10 �L of water with aqueous 0.1% trifluoroacetic acid for masspectrometric analysis. The peptide separations were performedith a 1200 series HPLC system (Agilent Technologies, Wilmington,E, USA) using an HPLC-chip (large capacity chip, 150 mm, 300 A,18 chip, with a 160 nL trapping column) (Agilent Technologies,ilmington, DE, USA) with a nano-flow pump. Mass spectrometry

nalysis was performed using 6550 iFunnel Quadrupole Time-f-Flight (Q-TOF) LC–MS (Agilent Technologies, Wilmington, DE,SA) with HPLC-chip cube source. Data were acquired in the mass

ange 100–3000 m/z with positive ion polarity as described by Han

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

t al. (2013). MS/MS data from the Agilent Q-TOF instrument wasearched against the UNIPROT mouse proteomic database usingEQUEST (Sorcerer 3.5; Sage-N Research, La Jolla, CA, USA). Allhe peptide identifications were validated using PeptideProphet

pE; Tc This study

(Institute for Systems Biology, Seattle, WA, USA). A cut-off proba-bility score of 0.95 was used for this study. In the trans-proteomicspipeline (TPP), ProteinProphet (Institute for Systems Biology, Seat-tle, WA, USA), infers the simplest list of proteins consistent withthe peptides identified.

2.9. Western blot analysis

Western blotting was performed as previously described (Leeet al., 2014). Briefly, intestinal tissue was lysed with buffer (20 mMTris [pH 7.5], 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 1 mg/mLaprotinin, and 1 mM phenylmethylsulfonylfluoride) for 30 min onice. The lysates were then cleared by centrifugation (15,000 × g at4 ◦C for 30 min). Protein concentration was determined by the Brad-ford method (Bradford, 1976). Equal amounts of protein (20 �g)were resolved by 10% sodium dodecyl sulfate–polyacrylamidegel electrophoresis and transferred to a polyvinylidene fluoridemembranes. The membranes were washed with TBST solution(10 mM Tris–HCl [pH 7.6], 150 mM NaCl, and 0.05% Tween-20),blocked with 5% skim milk for 1 h, and incubated with appro-priate primary antibody at 4 ◦C for overnight. The membranewas then washed and detected with a horseradish peroxidase-conjugated secondary antibody. The bands were visualized byenhanced chemiluminescence (Amersham Pharmacia Biotech Inc.,Buckinghamshire, UK). The relative optical density (ROD) of thebands was quantified using Scion Imaging Software (Scion ImageBeta 4.02, MD, USA).

2.10. Immunofluorescence and immunohistochemical analysis

Mice (n = 6) were given i.g. inoculation of WT, boiled WT (Cont),and vvpE mutant at 1.1 × 109 CFU/mL for 16 h. The ileum tissueswere embedded in O.C.T. compound and stored at −70 ◦C. Sampleswere then cut into 6-�m-thick frozen sections. For immunofluo-rescence analysis, sections were fixed in 4% paraformaldehyde inPBS for 10 min at room temperature, permeabilized in 0.1% Tri-ton X-100 in PBS for 5 min, and blocked in PBS containing 5% (v/v)normal goat serum (NGS) for 30 min at room temperature. Tissueswere then stained with primary antibody for overnight at 4 ◦C. Fol-lowing three washes with PBS, the tissues were incubated withAlexa 488-conjugated goat anti-rabbit/mouse IgM (Invitrogen Co.,Carlsbad, CA, USA), and counterstained with PI in PBS containing 5%(v/v) NGS for 2 h. After washing with PBS, samples were mountedon slides and visualized with an Olympus FluoViewTM 300 confocalmicroscope with 400× objective. For immunohistochemical anal-ysis, tissues incubated with primary antibody for overnight at 4 ◦Cwere treated with biotinylated secondary antibody solution (Vec-tastain Elite ABC kit, Vector Laboratories, CA, USA) for 1 h at roomtemperature. Sections were washed with PBS, incubated in the ABC

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

reagent for 1 h at room temperature, washed again and incubatedin a peroxidase solution. Sections were then counterstained withhematoxylin, dehydrated, and coverslipped. Images were acquiredusing an Axioskop 2 plus microscope equipped with an AxioCam

ING ModelI

4 f Med

Mt

2

awpwlaat0pacamowd

2r

g2ocptubKudc

2

emirBua(omiat(

2

ebBn

ARTICLEJMM-51011; No. of Pages 10

S.-J. Lee et al. / International Journal o

Rc5 CCD camera (Zeiss). Other samples were subjected to hema-oxylin and eosin (H&E) staining for histological examinations.

.11. Histologic damage score

Mice (n = 6) were given i.g. inoculation of WT, boiled WT (Cont),nd vvpE mutant at 1.1 × 109 CFU/mL for 16 h. The ileum tissuesere embedded in O.C.T. compound and stored at −70 ◦C. Sam-les were then cut into 6-�m-thick frozen sections. Tissue samplesere subjected to hematoxylin and eosin (H&E) staining for histo-

ogical examinations. Histological parameters were determined in blinded fashion by two experienced gastrointestinal pathologistss previously described (Clark et al., 2006) with some modifica-ion (Schulz et al., 2013). Briefly, scores were assigned as follows:

= no damage (normal); 1 = slight submucosal and/or lamina pro-ria separation (mild); 2 = moderate separation of the submucosand/or lamina propria and/or edema in the submucosa and mus-ular layers (moderate); 3 = severe separation of the submucosand/or lamina propria and/or severe edema in the submucosa anduscular layers with regional villous sloughing (severe); or 4 = loss

f villi and necrosis (necrosis). Intermediate scores of ×.0 or ×.5ere also used to more precisely assign the degree of intestinalamage.

.12. Reverse transcription (RT) and real-time polymerase chaineaction (PCR)

RNA was prepared by using the RNeasy Plus Mini Kit (Quia-en, Valencia, CA, USA). Reverse transcription (RT) was done with

�g of RNA using a Maxime RT premix kit (iNtRON Biotechnol-gy, Sungnam, Korea). The cDNAs (5 �L) for pro-inflammatoryytokines and toll-like receptors (TLRs) were amplified using therimers described in Supplementary Table 2. The real-time quan-ifications of pro-inflammatory cytokines and TLRs were performedsing a Rotor-Gene 6000 real-time thermal cycling system (Cor-ett Research, New South Wales, Australia) with a QuantiMix SYBRit (PhileKorea Technology, Daejeon, Korea) according to the man-facturer’s instructions with minor modifications as previouslyescribed (Lee et al., 2014). �-Actin was used as an endogenousontrol.

.13. Ileal-ligated mouse model

To determine the functional role of VvpE, we performed furtherxperiment by using the ileal-ligated mouse model. Before surgery,ice were fasted for 24 h and anesthetized by intraperitoneal

njection of a 2:1 mixture of ZoletilTM (20 mg/kg, Virbac Labo-atories, Carros, France) and Xylazine HCl (10 mg/kg, Rompun®,ayer, Germany). While maintaining the body temperature at 37 ◦Csing a heating pad, a small abdominal incision was made and

loop of middle ileum of intestine was isolated by silk suture2–3 cm in length). The closed ileal loop was instilled with 100 �Lf phosphate-buffered saline (PBS) containing either WT or vvpEutant at 1.1 × 109 CFU/mL for 2 h. After putting the ileal loop back

nto the peritoneal cavity, the cavity was closed with suture. At 2 hfter the inoculation of both strains, the mice were sacrificed andhe intestinal loops were removed for the hematoxylin and eosinH&E) staining and the real-time PCR.

.14. Statistical analysis

Results are expressed as means ± standard deviation (S.D.). All

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

xperiments were analyzed by ANOVA, followed in some casesy a comparison of treatment means with a control using theonferroni-Dunn test. Differences were considered statistically sig-ificant at P < 0.05.

PRESSical Microbiology xxx (2015) xxx–xxx

3. Results

3.1. VvpE is required by V. vulnificus to disrupt the intestinalbarrier

We first determined the paracellular permeability of intestinalepithelial cells with respect to the proteolytic activity of VvpE inpromoting of the invasion of V. vulnificus. Seven-week-old ICR miceinoculated intragastrically with boiled V. vulnificus (Control), V. vul-nificus (WT), and a mutant deficient in vvpE gene in V. vulnificus(vvpE mutant) at 1.1 × 109 CFU/mL, and killed at 4 h, 8 h, and 16 hlater. It is noted that the mice were given FITC-labeled 4 kDa dextranby oral injection for 4 h prior to the sacrifice to evaluate the intesti-nal leakage and translocation of 4 kDa FITC-dextran to the bloodstream. There were no statistically significant differences on thelevels of FITC-dextran after treatment with WT for 4 h, compared tothe control (Fig. 1A). The serum level of FITC-dextran in mice givenWT was significantly up-regulated by 0.25 and 0.52 �g/mL at 8 and16 h, respectively, compared to the control (Fig. 1B and C). However,the level of FITC-dextran was significantly down-regulated by 0.22and 0.48 �g/mL at 8 h and 16 h after inoculation with vvpE mutant,respectively, compared to the WT. In addition, complementationof vvpE mutant with a functional vvpE gene resulted in increasedparacellular permeability of intestinal epithelial cells. The levelof FITC-dextran in mice given vvpE complementation strain wasalmost similar to that of WT. This result suggests that VvpE playsan important role in the regulation of epithelial barrier disruptioninduced by WT. The barrier dysfunction after WT infection mightin part arise from an alteration in the stability of tight/adherensjunction (TAJ) proteins. The results in Fig. 1D revealed that inocu-lation with WT for 16 h significantly decreased the expression ofproteins related to the tight junction (claudin1/2 and occludin) andadherens junction (E-cadherin), but not the gap junction (Connexin43). However, the vvpE mutant inoculation negated intestinal bar-rier dysfunction and maintained the physiological expression levelsof TAJ proteins.

3.2. VvpE contributes to the intestinal colonization of V. vulnificus

Damage and impairment of the tight junctions facilitate theinvasion of pathogenic microorganisms to promote the bacte-rial colonization and to obtain the nutrients. To evaluate therole of VvpE in intestinal colonization of V. vulnificus, mice inoc-ulated intragastrically with Control, WT, and vvpE mutant at1.1 × 109 CFU/mL for 4 h, 8 h, and 16 h, after which the level of bac-terial colonization were monitored. At 4 h and 8 h, most of the WTand vvpE mutant appeared to colonize in mainly the mice colons,where the vvpE mutant exhibited significantly defective coloniza-tion ability as compared to WT (Fig. 2A and B). The WT colonizationin the PC and DC increased by 2.8 ± 0.5 and 3.3 ± 0.4 (×106 CFU/gtissue) at 4 h and by 10.8 ± 3.0 and 16.7 ± 2.9 (×106 CFU/g tis-sue) at 8 h, respectively. However, when the mice inoculatedwith vvpE mutant, the levels were diminished by 2.0 ± 0.5 and1.4 ± 0.3 (×106 CFU/g tissue) at 4 h and by 9.1 ± 2.6 and 14.7 ± 3.2(×106 CFU/g tissue) at 8 h, respectively, compared to the WT. Inter-estingly, there was a trend toward increased colonization in thesmall intestine, especially in the ileum by a WT infection at 16 h(Fig. 2C). However, the colonization level of vvpE mutant wassignificantly lower than that elicited by WT. Here, the levels of col-onization were augmented by 32.7 ± 7.5 (×106 CFU/g tissue) in theileum after treatment with WT for 16 h, whereas the colonizationlevels of vvpE mutant were decreased by 25.3 ± 2.1 (×106 CFU/g

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

tissue). The trend of intestinal colonization of WT is summarizedin Fig. 2D. Interestingly, neither of the bacterial strains led to liverand spleen colonization. It was noted that all of the mice given oraladministrations of WT had survived by 16 h post injection (data not

ARTICLE IN PRESSG ModelIJMM-51011; No. of Pages 10

S.-J. Lee et al. / International Journal of Medical Microbiology xxx (2015) xxx–xxx 5

Fig. 1. VvpE is required by V. vulnificus to disrupt the intestinal barrier. Mice inoculated with WT, boiled WT (Cont), and vvpE mutant at 1.1 × 109 CFU/mL, and sacrificedat 4 h (A) and 8 h (B) later. (C) Mice inoculated with WT, boiled WT (Cont), vvpE mutant, and vvpE complement (comp) at 1.1 × 109 CFU/mL, and sacrificed 16 h later. TheFITC-dextran was orally administered into mice for 4 h prior to the sacrifice. Translocation of FITC-dextran into blood stream was measured by using a luminometer. Errorbars represent the means ± S.D. from six independent experiments (n = 10). *, P < 0.01 versus Cont. #, P < 0.01 versus WT. a, P < 0.01 versus vvpE mutant. (D) The ileal expressionof tight/adherens junction (TAJ) proteins in the mice inoculated intragastrically with Cont, WT, and vvpE mutant at 1.1 × 109 CFU/mL for 16 h was determined by Western blotw imags ContR

st

3

nSmr1ataati(Wl

ith Claudin 1/2, Occludin, E-cadherin, and Connexin 43 antibodies. Representativehown (n = 6). The numbers (1 and 2) indicates two different mice. *, P < 0.01 versusOD, relative optical density.

hown). Taken together, our results indicate that VvpE contributeso the intestinal colonization of V. vulnificus in vivo.

.3. VvpE mediates enterotoxicity of V. vulnificus

Elevated serum iron levels and tissue destruction are promi-ent features of V. vulnificus infection (Jones and Oliver, 2009).erum lactate dehydrogenase (LDH) release was determined as aarker for cytotoxicity. As shown in Fig. 3A, the levels of serum LDH

elease significantly augmented by 805 U/L in mice given WT for6 h, compared to the control, whereas it was reduced by 567 U/Lfter treatment with the vvpE mutant, compared to the WT. Consis-ent with the data on liver and spleen colonization, WT did not showny effect on the level of alanine aminotransferase (ALT)/aspartateminotransferase (AST) related to liver damage (Fig. 3B). In addi-ion, the results of a routine complete blood count (CBC) of mice

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

nfected with WT showed reduced levels of red blood cells (RBC)Fig. 3C) as well as hematocrit (HCT) (Fig. 3D), suggesting that

T infection caused the destruction of RBC to elevate serum ironevel. Despite the decrease in RBC, the levels of white blood cells

es of western blot of total protein lysates isolated from ileal tissue in sixplicate are. #, P < 0.05 versus WT. Error bars represent the means ± S.D. (n = 6). Abbreviations:

(WBC), hemoglobin (HGB), mean corpuscular volume (MCV), meancorpuscular hemoglobin (MCH), mean corpuscular hemoglobinconcentration (MCHC), cellular hemoglobin concentration mean(CHCM), cellular hemoglobin content (CH), red blood cell distribu-tion width (RDW), hemoglobin distribution width (HDW), platelets(PLT), and mean platelet volume (MPV) showed no significantdifferences from the control groups (Supplementary Table 3). How-ever, the virulent effects of WT were not evident in mice infectedwith the vvpE mutant. These results indicate that VvpE plays animportant role in the enterotoxicity of WT as well as in the break-down of the iron-containing hemoglobin of RBC.

3.4. VvpE is required for pro-inflammatory responses of V.vulnificus

WT induced severe inflammation of the intestine, where

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

it caused shortened villi heights accompanied by increasedinflammatory cells at 16 h infection, resulting in increased levelof histopathological damage score, compared to control mice(Fig. 4A). However, vvpE mutant almost completely prevented

ARTICLE IN PRESSG ModelIJMM-51011; No. of Pages 10

6 S.-J. Lee et al. / International Journal of Medical Microbiology xxx (2015) xxx–xxx

Fig. 2. VvpE contributes to the intestinal colonization of V. vulnificus. Colonization activities in mice given an i.g. inoculation of WT, boiled WT (Cont), and vvpE mutant at1 intest( distal

tcrsdiclCccpuwtmirTtcFabobwmtrc(t

Specifically, �-globin expressed in red blood cells (RBC), AnnexinA2 and Annexin A4 in intestinal epithelial cells, intelectin-1b ingoblet cells, and F-actin were co-immunoprecipitated with anti-VvpE antibody in mice inoculated with WT for 16 h, but not

Table 2Intestinal proteins that interact with VvpE.

Protein name Accession number

�-globin A8DUK4Annexin A2 P07356Annexin A4 P97429Actin, aortic smooth muscle P62737Actin, cytoplasmic 1 P60710

.1 × 109 CFU/mL for 4 h (A), 8 h (B), and 16 h (C) were determined. (D) The trend of

n = 10). Abbreviations: D, duodenum; J, jejunum; I, ileum; PC, proximal colon; DC,

heir intestinal villi structures and the inflammation. In contrast,omplementation of vvpE mutant with a functional vvpE geneeversed the effect of vvpE mutant on histopathological damagecore (Supplementary Fig. 1A). In order to achieve effective hostefensive mechanism, we thought the inflammatory process dur-

ng infection must be coupled with the elevation of specific immuneells and evaluated the role of VvpE in the regulation of the level ofeukocytes. There were no changes between the mice treated withontrol, WT, and vvpE mutant in the levels of lymphocyte, mono-yte, eosinophil, and basophil in differential counts of white bloodells (WBC) (Supplementary Table 4). Interestingly, however, theercentage of neutrophils in mice inoculated with WT for 16 h wasniquely elevated by 19.6%, compared to the control, whereas itas blocked by 15.1% after treatment with vvpE mutant, compared

o the WT (Fig. 4B). In addition, the inoculation of mice with vvpEutant for 16 h failed to elevate the level of expression of pro-

nflammatory cytokine (IL-6, IL-8, and TNF-�) (Fig. 4C) and toll-likeeceptor (TLR-4,-5, and -9) (Fig. 4D) as caused by the WT infection.he results of complementation of vvpE mutant with a func-ional vvpE gene resulted in increased level of pro-inflammatoryytokines level similar to those of WT infection (Supplementaryig. 1B). To eliminate the possibility that the lower enterotoxicitynd lower induction of cytokines are simply explained by lessacteria in the mutant infected mice, independent of the presencer absence of vvpE, we further performed additional experimenty using the ileal-ligated mouse model where the closed ileal loopas instilled with 100 �L of PBS containing either WT or vvpEutant at 1.1 × 109 CFU/mL for 2 h. We anticipated that 2 h was

he minimum duration for enterotoxicity and pro-inflammatory

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

esponses based on our recent report which showed that WTolonized the mouse small intestine as early as 1 h after infectionKim et al., 2013). At 2 h after inoculation, we have further inves-igated whether VvpE mediates enterotoxicity (Supplementary

inal colonization of WT. *, P < 0.01 versus WT. Error bars represent the means ± S.D.colon; S, spleen; L, liver; CFU, colony forming unit

Fig. 1C) and pro-inflammatory responses (Supplementary Fig. 1D).Consistently, the injection of mice with vvpE mutant for 2 h failedto elevate the level of histopathological damage score as well asthe expression of pro-inflammatory cytokine as caused by theWT infection. These results indicate that VvpE is a critical elastasewhich mediates intestinal inflammatory responses of V. vulnificus.

Impairment of the tight junctions also enhances epithelialcell-pathogen interactions (Parlato and Yeretssian, 2014). Hav-ing shown that the effect of VvpE in mediating virulence effectof V. vulnificus on TAJ dysregulation and colonization, we antic-ipated that there are potential host binding partners of VvpEin intestine. At 16 h after inoculation, mouse ileal protein lysatewere prepared, immunoprecipitated with anti-VvpE antibody, andthen analyzed by duplicate LC–MS/MS. Table 2 shows that sev-eral important proteins co-immunoprecipiated with the anti-VvpEantibody were identified through a search of the UNIPROT database.

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

Intelectin 1b Q80ZA0

Mice inoculated intragastrically (i.g.) with V. vulnificus (WT) at 1.1 × 109 CFU/mL,and killed at 16 h later. Intestinal proteins were prepared, immunoprecipitated withanti-VvpE antibody, and then analyzed by the duplicate LC–MS/MS.

ARTICLE IN PRESSG ModelIJMM-51011; No. of Pages 10

S.-J. Lee et al. / International Journal of Medical Microbiology xxx (2015) xxx–xxx 7

F leveW ed. n =L ansfe

wAlswptoSAbdopFndAl

4

jcictt

ig. 3. VvpE mediates enterotoxicity of V. vulnificus. Effect of vvpE mutant on serumT, boiled WT (Cont), and vvpE mutant at 1.1 × 109 CFU/mL for 16 h was determin

DH, lactate dehydrogenase; ALT, alanine aminotransferase; AST, aspartate aminotr

ith the vvpE mutant. The abilities of the two strains to bind tonnexin A2/A4, intelectin-1b, and F-actin in mouse ileal protein

ysate were further confirmed by immunoprecipitation analy-es (Fig. 4E). As expected, the VvpE antibody specifically reactedith Annexin A2/A4, intelectin-1b, and F-actin. We have furthererformed additional immunoprecipitation assay with ileal pro-ein lysate extracted from mouse inoculated with either Controlr WT as a cross reacting control. As shown in left panel ofupplementary Fig. 1E, the co-immunoprecipitation of VvpE withnnexin A2/A4, F-Actin, and Intelectin 1b was markedly inducedy WT infection the ileal tissue lysate. There were no significantifferences on the expression level of potential binding partnersf VvpE in total protein lysates after treatment with WT, com-ared to the control. In addition, as shown in right panel ofig. 4E, the level of potential binding partners of VvpE was alsoot changed by vvpE mutant infection, suggesting that VvpE pro-uced by V. vulnificus possibly interacts with Annexin A2, Annexin4, F-Actin, and intelectin 1b, without affecting their expression

evels.

. Discussion

In this study, we demonstrated that VvpE is required for tightunction disruption and the intestinal colonization of V. vulnifi-us and that it has an important potential to interact with the

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

ntestinal proteins responsible for the bacterial pathogenesis. Con-erning the pathogenic mechanism of VvpE, we showed that VvpEargets the gut junctional barrier by reducing the expression ofight/adherens junction (TAJ) proteins such as claudin1/2, occludin,

ls of (A) LDH, (B) AST/ALT, (C) RBC, and (D) HCT in mice given an i.g. inoculation of 6. *, P < 0.01 versus Cont. #, P < 0.01 versus WT. Abbreviations: n.s., not significant;rase RBC, red blood cells; HCT, hematocrit

and E-cadherin, thereby playing a critical role in promoting theepithelial paracellular permeability of V. vulnificus. In addition, thetight junction has been characterized as a platform for intercellularreceptor complex including Annexin A2 in mediating physiologi-cal signal transduction (Lee et al., 2004; Sears, 2000). Specifically,Annexin 2A was known to mediate the tight junction assembly pos-sibly through linking juxtaposed exoplasmic leaflets to build up thelipid platform (Lee et al., 2004). It is not clear whether the functionalrole of VvpE in the repression of TAJ proteins in vivo is a directeffect on TAJ proteins as a bacterial-derived protease or alterna-tively, a sequential process involving other cellular signaling eventsrelated to the integrity of the cytoskeleton. However, researchclearly shows that the disruption of the TAJ barrier as a consequenceof a bacterial infection results in various modes of perturbations ofepithelial functions, such as fluid/electrolyte secretion and effectson the intestinal barrier integrity (Berkes et al., 2003). In addi-tion, VvpE was shown to have proteolytic activity such that itdegrades type IV collagen in promoting the vascular permeabil-ity that is necessary for the invasion of this bacterium (Miyoshiet al., 1998). Similarly, several pathogens that produce extracellu-lar toxins, including B. fragilis, have been shown to exert proteolyticactivity to disrupt the TAJ barrier, thus having a virulent effecton intestinal pathogenesis (Wu et al., 1998). Because TAJ proteinsare widely used as a target for pathogens to regulate the proteinkinase C and Rho GTPase pathways responsible for actin cytoskele-

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

ton rearrangement (Kim et al., 2010), our data here suggest thatV. vulnificus efficiently infects within the gut mucosa by produc-ing VvpE with modes of action that disrupt the stability of the TAJbarrier.

ARTICLE IN PRESSG ModelIJMM-51011; No. of Pages 10

8 S.-J. Lee et al. / International Journal of Medical Microbiology xxx (2015) xxx–xxx

Fig. 4. VvpE is required for pro-inflammatory responses of V. vulnificus. Mice were given an i.g. inoculation of WT and vvpE mutant at 1.1 × 109 CFU/ml and killed at 16 h later.(A) Representative ileum tissues stained with H&E are shown (left panel). Scale bars represent 100 �m. Average scores of histopathologic damage index from mouse ileumare shown (right panel). (B) The results of neutrophil level in white blood cells (WBC) differential count in mouse blood infected with WT and vvpE mutant are shown. Theexpression levels of (C) pro-inflammatory cytokines (IL-1�, IL-6, and TNF-�) and (D) TLRs (TLR-2,-4,-5, and -9) in ileum tissue are shown. (E) Total protein lysates isolatedfrom ileal tissue infected by WT and vvpE mutant were immunoprecipitated with anti-VvpE antibody, and then co-immunoprecipitated Annexin A2, Annexin A4, F-actin, andi nteleci presen( ersus

roapirtftW

ntelectin-1b were detected by using anti-Annexin A2, -Annexin A4, -F-actin, and -intelectin-1b in total protein lysates are shown in the right panels (n = 3). Lysate reA-D) Error bars represent the means ± S.D. n = 6. *, P < 0.01 versus Cont. #, P < 0.01 v

It has been clearly reported that the impairment of the TAJ bar-ier enhances the colonization of pathogens with the establishmentf an appropriate portal of entry, where it promotes pathogen–hostdherence mechanisms and thereby initiates a pro-inflammatoryrocess. In the present study, we uncovered the importance of VvpE

n the regulation of intestinal colonization and in inflammatoryesponses during V. vulnificus infection. Mice appeared to be able

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

o rid themselves of V. vulnificus in the colon at early time pointsollowing an oral administration of WT, as the bacterial popula-ion was highest in the distal colon. At 16 h, however, a shift in

T colonization was observed in the ileum, whereas a loss of the

tin-1b antibodies (left panels). Expressions of Annexin A2, Annexin A4, F-actin, andts the total protein lysates isolated from ileal tissue before applying IP procedure.

WT. Abbreviations: n.s., not significant.

VvpE function exhibited highly reduced colonization ability, sug-gesting that VvpE is the relevant elastase responsible for mouseileal colonization. The late time point for the ileum growth of WTis potentially due to a secondary infection caused by grooming andthe consumption of feces via the cohousing of similarly infectedanimals. These results are in contrast to those of a previous reportwhich revealed that VvpE appeared to be dispensable for a viru-

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

lence effect of V. vulnificus (ATCC29307) in a mouse wound infectionmodel induced by subcutaneous inoculation (Jeong et al., 2000). Thediscrepancy with regard to the functional role of VvpE may due todifferences in the pathogenic isolate of V. vulnificus (MO6-24/O), the

ING ModelI

f Med

aetbetsMic

rtrtyotttphptttibtitttani2si

tptVAwtteai2m(tpa2prtaa

ai

ARTICLEJMM-51011; No. of Pages 10

S.-J. Lee et al. / International Journal o

nesthesia method, or the injection location used in this study (Leet al., 2007b; Olivier et al., 2009). In fact, it has been shown thathere are significant differences in the biofilm formation abilitiesetween the V. vulnificus strains ATCC29307 and MO6-24/O (Leet al., 2007b). In addition, many relevant reports have suggestedhat VvpE plays critical roles in hemorrhagic damage, dermonecro-is, and vascular permeability in mice (Kothary and Kreger, 1987;iyoshi and Shinoda, 1992). Thus, we suggest that VvpE is another

mportant virulence factor of V. vulnificus responsible for mouseolonization.

Interestingly, the virulence effect of VvpE was connecting toeductions in the levels of RBC and HCT in mice. It has been shownhat hemoglobin, methemoglobin, and hematin play importantoles as iron donors for V. vulnificus (Helms et al., 1984). Givenhe finding that �-globin interacts with VvpE in a LC–MS/MS anal-sis, our results suggest that VvpE is important for the bindingf V. vulnificus to the local hemoglobin of RBC in the intestine,hereby obtaining nutrients such as serum iron by destroyinghe RBC. These results also indicate that VvpE plays an impor-ant role in the RBC targeting of V. vulnificus. Colonization byathogenic microorganisms destabilizes the homeostasis of theost immune responses and thereby initiates an inflammatoryrocess (Ivanov et al., 2010). In the present study, we foundhat VvpE is necessary for inflammation-related tissue injury andhat this leads to the elevation of pro-inflammatory-related fac-ors, including pro-inflammatory cytokines and TLRs. As revealedn a WBC test, the inflammation mediated by VvpE may haveeen caused by the unique elevation of neutrophils, which ishe first inflammatory responder to an acute bacterial infectionn the intestinal epithelium, where it delays bacterial dissemina-ion by limiting the bacterial infection to the intestine as part ofhe host defense mechanism (Queen and Satchell, 2012). In con-rast to the host, several pathogens, including V. vulnificus, havelso been reported to obtain necessary nutrients by recruitingeutrophils and thereby promoting bacterial replication in the

ntestinal lumen and at sites of colonization (Queen and Satchell,013). Thus, these results suggest that VvpE uniquely regulatesevere inflammation by controlling the level of neutrophils in thentestine.

To find another important function of VvpE in mediatinghe pathogenesis of V. vulnificus, we focused closely on theathogen–host interactions between VvpE and mouse ileal pro-ein, as revealed by a LC–MS/MS analysis. Our results show thatvpE is an important elastase which co-immunoprecipitated withnnexin A2/A4, F-actin, and intelectin-1b in the ileum. Althoughe did not investigate whether the functional role of these interac-

ion is connected to the phenotypes observed in the present study,he importance of these intestinal proteins in microbial pathogen-sis is underscored by the findings that Annexin A2 regulates thedherence of E. coli (EPEC) at an early stage of infection, wheret connects to the actin cytoskeleton of host cells (Zobiack et al.,002); the expression of Annexin A4 is responsible for the inflam-atory process induced by Helicobacter pylori (H. pylori) infections

Lin et al., 2008); V. parahaemolyticus VopV binds directly to F-actino promote enterotoxicity (Hiyoshi et al., 2011); and intelectin-1b,roduced mainly by goblet cells recognizes the parasite infections a decoy pathogen receptor (Pemberton et al., 2004; Tsuji et al.,001; Wrackmeyer et al., 2006). Thus, these results present theossibility that VvpE influences the pathogenesis of V. vulnificus byegulating the above moue ileal proteins. This interaction is uniqueo VvpE, as no such defect was reported from the absence of VvhAnd MARTXVv, that are known as major virulence proteins which

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

ct during WT infection.Collectively, the results of this study suggest that VvpE plays

n essential role in promoting the tight junction disruption andntestinal colonization of V. vulnificus and that it also has the ability

PRESSical Microbiology xxx (2015) xxx–xxx 9

to interact with the intestinal proteins responsible for microbialpathogenesis.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgements

This research was supported by grants to both HJ Han and SHChoi from the Agriculture, Food and Rural Affairs Research CenterSupport Program, Ministry of Agriculture, Food and Rural Affairs,Republic of Korea (Grant 710002-07-5).

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.ijmm.2015.10.006.

References

Ashida, H., Ogawa, M., Kim, M., Mimuro, H., Sasakawa, C., 2012. Bacteria and hostinteractions in the gut epithelial barrier. Nat. Chem. Biol. 8, 36–45.

Berkes, J., Viswanathan, V.K., Savkovic, S.D., Hecht, G., 2003. Intestinal epithelialresponses to enteric pathogens: effects on the tight junction barrier, iontransport, and inflammation. Gut 52, 439–451.

Blake, P.A., Merson, M.H., Weaver, R.E., Hollis, D.G., Heublein, P.C., 1979. Diseasecaused by a marine Vibrio. Clinical characteristics and epidemiology. N. Engl. J.Med. 300, 1–5.

Bradford, M.M., 1976. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding.Anal. Biochem. 72, 248–254.

Clark, J.A., Doelle, S.M., Halpern, M.D., Saunders, T.A., Holubec, H., Dvorak, K.,Boitano, S.A., Dvorak, B., 2006. Intestinal barrier failure during experimentalnecrotizing enterocolitis: protective effect of EGF treatment. Am. J. Physiol.Gastrointest. Liver Physiol. 291, G938–G949.

Han, N.Y., Choi, W., Park, J.M., Kim, E.H., Lee, H., Hahm, K.B., 2013. Label-freequantification for discovering novel biomarkers in the diagnosis andassessment of disease activity in inflammatory bowel disease. J. Dig. Dis. 14,166–174.

Helms, S.D., Oliver, J.D., Travis, J.C., 1984. Role of heme compounds andhaptoglobin in Vibrio vulnificus pathogenicity. Infect. Immun. 45, 345–349.

Heo, S.H., Lee, S.J., Ryoo, H.M., Park, J.Y., Cho, J.Y., 2007. Identification of putativeserum glycoprotein biomarkers for human lung adenocarcinoma bymultilectin affinity chromatography and LC–MS/MS. Proteomics 7, 4292–4302.

Hiyoshi, H., Kodama, T., Saito, K., Gotoh, K., Matsuda, S., Akeda, Y., Honda, T., Iida, T.,2011. VopV, an F-actin-binding type III secretion effector, is required for Vibrioparahaemolyticus-induced enterotoxicity. Cell Host Microbe 10, 401–409.

Isberg, R.R., Leong, J.M., 1990. Multiple �1 chain integrins are receptors for invasin,a protein that promotes bacterial penetration into mammalian cells. Cell 60,861–871.

Ivanov, A.I., Parkos, C.A., Nusrat, A., 2010. Cytoskeletal regulation of epithelialbarrier function during inflammation. Am. J. Pathol. 177, 512–524.

Jeong, H.G., Satchell, K.J., 2012. Additive function of Vibrio vulnificus MARTX(Vv)and VvhA cytolysins promotes rapid growth and epithelial tissue necrosisduring intestinal infection. PLoS Pathog. 8, e1002581.

Jeong, K.C., Jeong, H.S., Rhee, J.H., Lee, S.E., Chung, S.S., Starks, A.M., Escudero, G.M.,Gulig, P.A., Choi, S.H., 2000. Construction and phenotypic evaluation of a Vibriovulnificus vvpE mutant for elastolytic protease. Infect. Immun. 68, 5096–5106.

Jones, M.K., Oliver, J.D., 2009. Vibrio vulnificus: disease and pathogenesis. Infect.Immun. 77, 1723–1733.

Keen, N.T., Tamaki, S., Kobayashi, D., Trollinger, D., 1988. Improvedbroad-host-range plasmids for DNA cloning in gram-negative bacteria. Gene70, 191–197.

Kim, C.M., Park, R.Y., Chun, H.J., Kim, S.Y., Rhee, J.H., Shin, S.H., 2007. Vibriovulnificus metalloprotease VvpE is essentially required for swarming. FEMSMicrobiol. Lett. 269, 170–179.

Kim, M., Ashida, H., Ogawa, M., Yoshikawa, Y., Mimuro, H., Sasakawa, C., 2010.Bacterial interactions with the host epithelium. Cell Host Microbe 8, 20–35.

Kim, S.M., Park, J.H., Lee, H.S., Kim, W.B., Ryu, J.M., Han, H.J., Choi, S.H., 2013. LuxRhomologue SmcR is essential for Vibrio vulnificus pathogenesis and biofilmdetachment, and its expression is induced by host cells. Infect. Immun. 81,3721–3730.

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

Kothary, M.H., Kreger, A.S., 1987. Purification and characterization of an elastolyticprotease of Vibrio vulnificus. J. Gen. Microbiol. 133, 1783–1791.

Lee, D.B., Jamgotchian, N., Allen, S.G., Kan, F.W., Hale, I.L., 2004. Annexin A2heterotetramer: role in tight junction assembly. Am. J. Physiol. Renal Physiol.287, F481–F491.

ING ModelI

1 f Med

L

L

L

L

L

M

M

M

M

M

O

P

P

ARTICLEJMM-51011; No. of Pages 10

0 S.-J. Lee et al. / International Journal o

ee, J.H., Kim, M.W., Kim, B.S., Kim, S.M., Lee, B.C., Kim, T.S., Choi, S.H., 2007a.Identification and characterization of the Vibrio vulnificus rtxA essential forcytotoxicity in vitro and virulence in mice. J. Microbiol. 45,146–152.

ee, J.H., Rhee, J.E., Park, U., Ju, H.M., Lee, B.C., Kim, T.S., Jeong, H.S., Choi, S.H.,2007b. Identification and functional analysis of Vibrio vulnificus SmcR, a novelglobal regulator. J. Microbiol. Biotechnol. 17, 325–334.

ee, S.J., Jung, Y.H., Oh, S.Y., Yun, S.P., Han, H.J., 2014. Melatonin enhances thehuman mesenchymal stem cells motility via melatonin receptor 2 couplingwith G�q in skin wound healing. J. Pineal Res. 57, 393–407.

im, J.G., Bang, Y.J., Choi, S.H., 2014. Characterization of the Vibrio vulnificus 1-Cysperoxiredoxin Prx3 and regulation of its expression by the Fe-S clusterregulator IscR in response to oxidative stress and iron starvation. J. Biol. Chem.289, 36263–36274.

in, L.L., Chen, C.N., Lin, W.C., Lee, P.H., Chang, K.J., Lai, Y.P., Wang, J.T., Juan, H.F.,2008. Annexin A4: a novel molecular marker for gastric cancer withHelicobacter pylori infection using proteomics approach. Proteomics Clin. Appl.2, 619–634.

engaud, J., Ohayon, H., Gounon, P., Mege, R.M., Cossart, P., 1996. E-cadherin is thereceptor for internalin, a surface protein required for entry of L. monocytogenesinto epithelial cells. Cell 84, 923–932.

iyoshi, S., 2006. Vibrio vulnificus infection and metalloprotease. J. Dermatol. 33,589–595.

iyoshi, S., Nakazawa, H., Kawata, K., Tomochika, K., Tobe, K., Shinoda, S., 1998.Characterization of the hemorrhagic reaction caused by Vibrio vulnificusmetalloprotease, a member of the thermolysin family. Infect. Immun. 66,4851–4855.

iyoshi, S., Shinoda, S., 1992. Activation mechanism of human Hagemanfactor-plasma kallikrein-kinin system by Vibrio vulnificus metalloprotease.FEBS Lett. 308, 315–319.

olla, A., Yamamoto, T., Akaike, T., Miyoshi, S., Maeda, H., 1989. Activation ofhageman factor and prekallikrein and generation of kinin by various microbialproteinases. J. Biol. Chem. 264, 10589–10594.

livier, V., Queen, J., Satchell, K.J., 2009. Successful small intestine colonization ofadult mice by Vibrio cholerae requires ketamine anesthesia and accessorytoxins. PLoS ONE 4, e7352.

Please cite this article in press as: Lee, S.-J., et al., VvpE mediates the injunctions. Int. J. Med. Microbiol. (2015), http://dx.doi.org/10.1016/j.ijm

ark, S.D., Shon, H.S., Joh, N.J., 1991. Vibrio vulnificus septicemia in Korea: clinicaland epidemiologic findings in seventy patients. J. Am. Acad. Dermatol. 24,397–403.

arlato, M., Yeretssian, G., 2014. NOD-like receptors in intestinal homeostasis andepithelial tissue repair. Int. J. Mol. Sci. 15, 9594–9627.

PRESSical Microbiology xxx (2015) xxx–xxx

Pemberton, A.D., Knight, P.A., Gamble, J., Colledge, W.H., Lee, J.K., Pierce, M., Miller,H.R., 2004. Innate BALB/c enteric epithelial responses to Trichinella spiralis:inducible expression of a novel goblet cell lectin, intelectin-2, and its naturaldeletion in C57BL/10 mice. J. Immunol. 173, 1894–1901.

Queen, J., Satchell, K.J., 2012. Neutrophils are essential for containment of Vibriocholerae to the intestine during the proinflammatory phase of infection. Infect.Immun. 80, 2905–2913.

Queen, J., Satchell, K.J., 2013. Promotion of colonization and virulence by choleratoxin is dependent on neutrophils. Infect. Immun. 81, 3338–3345.

Schulz, S., Wong, R.J., Jang, K.Y., Kalish, F., Chisholm, K.M., Zhao, H., Vreman, H.J.,Sylvester, K.G., Stevenson, D.K., 2013. Heme oxygenase-1 deficiency promotesthe development of necrotizing enterocolitis-like intestinal injury in anewborn mouse model. Am. J. Physiol. Gastrointest. Liver Physiol. 304,G991–G1001.

Sears, C.L., 2000. Molecular physiology and pathophysiology of tight junctions V.assault of the tight junction by enteric pathogens. Am. J. Physiol. Gastrointest.Liver Physiol. 279, G1129–G1134.

Simon, R., Priefer, U., Pühler, A., 1983. A broad host range mobilization system forin vivo genetic engineering: transposon mutagenesis in gram negative bacteria.Nat. Biotechnol. 1, 784–791.

Skoudy, A., Mounier, J., Aruffo, A., Ohayon, H., Gounon, P., Sansonetti, P., Tran VanNhieu, G., 2000. CD44 binds to the Shigella IpaB protein and participates inbacterial invasion of epithelial cells. Cell. Microbiol. 2, 19–33.

Tsuji, S., Uehori, J., Matsumoto, M., Suzuki, Y., Matsuhisa, A., Toyoshima, K., Seya, T.,2001. Human intelectin is a novel soluble lectin that recognizesgalactofuranose in carbohydrate chains of bacterial cell wall. J. Biol. Chem. 276,23456–23463.

Wrackmeyer, U., Hansen, G.H., Seya, T., Danielsen, E.M., 2006. Intelectin: a novellipid raft-associated protein in the enterocyte brush border. Biochemistry 45,9188–9197.

Wu, S., Lim, K.C., Huang, J., Saidi, R.F., Sears, C.L., 1998. Bacteroides fragilisenterotoxin cleaves the zonula adherens protein, E-cadherin. Proc. Natl. Acad.Sci. U. S. A. 95, 14979–14984.

Yun, S.P., Lee, S.J., Oh, S.Y., Jung, Y.H., Ryu, J.M., Suh, H.N., Kim, M.O., Oh, K.B., Han,H.J., 2014. Reactive oxygen species induce MMP12-dependent degradation ofcollagen 5 and fibronectin to promote the motility of human umbilical

testinal colonization of Vibrio vulnificus by the disruption of tightm.2015.10.006

cord-derived mesenchymal stem cells. Br. J. Pharmacol. 171, 3283–3297.Zobiack, N., Rescher, U., Laarmann, S., Michgehl, S., Schmidt, M.A., Gerke, V., 2002.

Cell-surface attachment of pedestal-forming enteropathogenic E. coli induces aclustering of raft components and a recruitment of annexin 2. J. Cell Sci. 115,91–98.