in vitro inhibition of epithelial cell invasion by aeromonashydrophila vibrio species by fish...

© 1998

Blackwell Science Ltd 273

Journal of Fish Diseases (1998) 21, 273–280

In vitro inhibition of epithelial cell invasion by Aeromonashydrophila and Vibrio species by fish Aeromonashydrophila major adhesin

H M Fang, K C Ling , Y L Tan, R Ge and Y M Sin

School of Biological Sciences, National University of Singapore, Singapore

Abstract

A virulent strain of Aeromonas hydrophila (PPD 134/91) was obtained from the Primary ProductionDepartment, Singapore. Its major adhesin wasisolated and purified by potassium thiocyanateextraction and Bio-Gel P-100 gel filtration. Theability of the protein in peak 1, termed majoradhesin, to inhibit bacteria from adhering to andinvading host cells was studied in vitro usingepithelioma papillosum cells of carp (EPC). Resultsshowed that a concentration of 10 µg ml–1 of thismajor adhesin could competitively inhibit 28% ofA. hydrophila PPD 134/91 from invading EPC cellsin vitro. When the concentration was increased to40 µg ml–1, the major adhesin significantly cross-inhibited nine other virulent or weakly virulentstrains of A. hydrophila. In addition, the majoradhesin significantly inhibited not only anotherbacterial strain from the same family, Aeromonassobria, but also strains of Vibrio spp. tested.Therefore, we suggest that the major adhesin of thisvirulent A. hydrophila strain has the potential to beused as a vaccine against the heterogeneousAeromonas and Vibrio species.

Introduction

Aeromonas hydrophila is a Gram-negative bacteriumthat infects a wide range of hosts, includingamphibians, reptiles, birds and mammals (Popoff1984). It is also well-known as a fish pathogen whereit causes motile aeromonad septicaemia (MAS), a

Correspondence Dr Y M Sin, School of Biological Sciences,National University of Singapore, Singapore

common bacterial cause of fish mortality (Austin &Austin 1987). This disease can be fatal with nogross clinical signs. It can also result in the formationof deep dermal ulcers, abscesses in the peritonealcavity and organ haemorrhage (Thune, Stanley &Cooper 1993). A variety of factors may contributeto the overall virulence of this bacterium. Theseinclude extracellular products (ECPs) (Ljungh &Wadstrom 1982; Leung & Stevenson 1988a,b;Angka, Lam & Sin 1995) and the S-layer (Dooley& Trust 1988). It is believed that adhesion ofbacteria is a prerequisite for successful invasion ofhost tissues and is required for the maximal effectof toxins produced by adherent pathogens (Quinn,Wong, Atkinson & Flower 1993).

Adhesion refers to a relatively stable, irreversibleattachment mediated by specialized complementarymolecules of the bacterial and host surface (Arp1988). It is not just a static process, but ratherelicits a response to the target with an outcome. Itis an active process (Hoepelman & Tuomanen1992). It allows the microorganisms to subsequentlycolonize epithelial layers at the initiation of diseaseand facilitates the organism in gaining a footholdin a variety of tissues as the infection continues.The colonization factors of A. hydrophila are notvery clear. It has been suggested that the presenceof wavy pili (w-pili) or ‘flexible pili’ correlates withthe adherence of A. hydrophila (Hokama, Honma &Nakasone 1990; Ho, Mietzner, Smith & Schoolnik1990). Atkinson, Adams, Savvas & Trust (1987)reported that haemagglutination by A. hydrophilacorrelates with the presence of a 43 kDa outermembrane protein. This protein was confirmed tobe a carbohydrate-reactive outer membrane protein

© 1998


Journal of Fish Diseases 1998, 21, 273–280 H M Fang et al. Adhesin of A. hydrophila

(Quinn, Atkinson, Bretag, Tester, Trust, Wong &Flower 1994).

The strains used to study A. hydrophila adhesionhave come mostly from human or mammalianisolates. The amount of information concerning theadhesins from strains that infect fish is limited.Merino, Rubires, Aguilar & Tomas (1997) reportedthat the presence of flagella is essential for somefish A. hydrophila strains in adhesion and invasion.Lipopolysaccharide (LPS) has also been reported asan adhesin in some virulent strains isolated fromfish (Merino, Rubires, Aguilar & Tomas 1996). Inour previous studies (Lee, Yin, Ge & Sin 1997),major and minor adhesins were isolated from avirulent strain of A. hydrophila PPD 134/91. Wedemonstrated that the major adhesin is most likelya 43 kDa outer membrane protein which cancompetitively inhibit bacterial invasion into hostcells in vitro. The aim of this study is to investigatethe possible cross-inhibition of major adhesin onserologically similar or different strains ofA. hydrophila, and on strains of the genera Aeromonasand Vibrio.

Materials and methods

Bacterial strains and media

Aeromonas hydrophila PPD 134/91 used for isolatingadhesin and three other strains of A. hydrophila(PPD11/90, PPD122/91 and PPD 70/91) and threeVibrio spp. [G/virus/5 (3), S1/7/93 (1) and W618]were all obtained from the Primary ProductionDepartment, Singapore. They were mostly isolatedfrom infected fish originating in Malaysia andSingapore. The sources of other strains of Aeromonasand Vibrio used are shown in Table 1: A. hydrophilastrains 58–20–9 and ST-78–3–3 and A. sobria strainCR-79–1–1 were from the Institute of Hydro-biology, Chinese Academy of Sciences, Luojiashan,Wuhan, China; A. hydrophila strain AH 94063 wasfrom the School of Biological Sciences, Singapore;and A. hydrophila strains L31 and L38 were fromthe South-East Asian Regional Centre for TropicalBiology, Bogor, Indonesia (Angka et al. 1995). Themajority of the above bacteria were isolated fromdiseased fish, except for the type strain ATCC7966(American Type Culture Collection, Rockville, MD,USA) which was isolated from milk. All these strainswere tested using standard biochemical diagnostickits (Microbact 24E System, Adelaide, Australia)and their identities were confirmed according to the

criteria of Popoff (1984). Cultures were routinelymaintained in tryptic soy broth (TSB, Difco, USA)or tryptic soy agar (TSA, Difco, USA) at 25 °C andstock cultures were kept in a suspension of TSBwith 25% glycerol and stored at –70 °C. For strainsof Vibrio, TSB or TSA was supplemented with0.5% NaCl.

Extraction and purification of adhesin

A volume of 30 ml of stationary overnight bacterialculture (PPD 134/91) was added to 2.4 l of freshTSB and cultured with gentle shaking on a BalmarGFL model 3005 shaker for 4 h until late log phasewas reached (Leung, Yeap, Lam & Sin 1995).Bacterial cells were then harvested by centrifugationat 6000 g for 10 min and washed twice withphosphate-buffered saline (PBS) (0.12 M NaCl,0.01 M Na2HPO4, 3.16 mM KH2PO4, pH 7.2–7.3)and pelleted before use. Pelleted bacteria were treatedwith 120 ml of 3 M KSCN (Sigma, USA) at roomtemperature with gentle swirling (Altmann, Pyliotis& Mukkur 1982). Cells were then spun down at12 000 g for 30 min with the resultant supernatantfiltered through a 0.22 µm pore size filter to removeany remaining bacteria or debris. Dialysis againstPBS was subsequently carried out at 4 °C to removeKSCN before flagella were pelleted and removedby ultracentrifugation at 40 000 g for 3 h. Thesupernatant was ultrafiltered through YM 1 filterpaper (Amicon, USA) according to themanufacturer’s instructions at 4 °C to desalt andfurther concentrate the supernatant. The lyophilizedsupernatant was redissolved in a minimal amountof PBS and immediately chromatographed on a1.5 3 120 cm column of Bio-Gel P 100 (fine) (Bio-Rad, USA). Fractionation was carried out at a flowrate of 0.1 ml min–1 and fractions were collectedevery hour at room temperature using a Pharmacia(Uppsala, Sweden) LKB Frac-100.

Protein assay

The protein concentration of fractions isolated wasdetermined essentially by the Bradford (1976)method using a Bio-Rad protein kit according tothe manufacturer’s instructions. A protein standardcurve was prepared using bovine serum albumin(BSA) provided in the kit.

SDS-PAGE

SDS-PAGE was performed according to the methodof Laemmli (1970) using a 12.5% separating gel

© 1998



Table 1 Sources of bacterial strains used in cross-inhibition studies

Strain Fish condition Source Virulence

Aeromonas hydrophilaPPD 134/91 Diseased fish Singapore VPPD 11/90 Diseased fish Singapore VPPD 122/91* Diseased fish Singapore VPPD 70/91* Diseased fish Singapore VAH 94063 Diseased fish Singapore VL 31 Diseased fish Indonesia VL 38 Diseased fish Indonesia WATCC 7966 Milk USA W58–20–9 Fish China VST-78–3–3 Fish China V

Aeromonas sobriaCR-79–1–1 Fish China V

Vibrio anguillarum811218–5 W Milkfish pondwater Taiwan VG/Virus/5 (3) Unknown Singapore V

Vibrio harveyiW 618 Diseased fish Singapore V

Vibrio vulnificusS1/7/93 (1) Diseased fish Singapore V

V, virulent (LD50 , 107 CFU bacteria); W, weakly virulent (LD50 5 107–108 CFU bacteria).*, similar serogroup (Leung et al. 1995).

and a 5% stacking gel. Samples were heated for5 min at 90 °C in loading buffer before loading.Prestained low-molecular-weight protein markerswere loaded in parallel.

Cell culture

All tissue culture reagents were obtained from GibcoLaboratories (Grand Island, NY, USA). Epitheliomapapillosum of carp (EPC) was grown in minimalessential medium (MEM) with Hanks’ salts, 2 mM

glutamine, 10 mM HEPES (pH 7.3), 0.23%NaHCO3 and 10% heat-inactivated foetal calf serum(Lee et al. 1997). Cells were grown in 75 cm2 flasks(Corning, NY, USA) at 25 °C with 5% carbondioxide. These were maintained by weeklysubculture following trypsin/EDTA treatment(0.05% trypsin, 0.53 mM EDTA) and diluted 1:10in MEM.

Competitive inhibition assay

The invasion and intracellular replication studieswere performed as described in Leung & Finlay(1991) with the following modifications. First,medium from 3-day-old tissue culture cells wasremoved and replaced with fresh medium. Fractionsor purified adhesin (100 µl) to be tested were addedinto their respective wells and incubated for 30 min

at 25 °C. In the two control wells, either BSAor no fractions were added. Two duplicates wereperformed for each sample in any experimental test.Next, 100 µl bacterial suspension (108 CFU ml–1)was added to each well and incubated for 30 minat 25 °C. The inoculum was later removed andwashed three times with HBSS (Hank’s balancedsalt solution). The cells were then overlaid with1 ml of medium and 100 µg gentamicin (Sigma,USA) in each well and incubated for another 30 minat 25 °C. This was to kill any extracellular bacteriathat had not penetrated EPC cells and were notremoved during the washing. The medium was thenremoved and washed twice with PBS. One millilitreof 1% Triton X-100 (Bio-Rad, USA) was thenadded to lyse the cells and release the intracellularbacteria. The number of bacteria invading the EPCcells was quantitated by plating aliquots on TSAplates. Experiments were performed in duplicatewith a total of eight TSA plates for each sampletested. At least two separate experiments wereperformed for each test.

Cross-inhibition assay

All the procedures used were similar to thecompetitive inhibition assay except that, for eachstrain, 100 µl of the respective bacterial suspension(108 CFU ml–1) was added to one well with

© 1998



purified adhesin or without adhesin (control). Forstrains of Vibrio spp., the cells were incubatedfor 2 h with 1 ml of medium and gentamicin(100 µg), and not for 30 min as was the case forAeromonas spp. This increased time was requiredfor the bacteria to be killed by gentamicin. Resultswere expressed as the percentage of bacteriapresent in the well with adhesin compared withbacteria in the control well.

Statistical analysis

All results from bacterial invasion assays wereexpressed as means 6 SD. The data fromthese assays were analysed using the Student’s t-test. A value of P , 0.05 was considered to besignificant.

Results and discussion

Purification of the major adhesin anddetermination of its biological activity

After KSCN extraction and Bio-Gel P-100 gelfiltration of A. hydrophila PPD 134/91, six peakswere collected and were used for the competitiveinhibition test of bacterial invasion into EPC cellsin vitro. The results showed that peak 1 had thehighest biological ability to inhibit the homologousbacterium, A. hydrophila PPD 134/91, frominvading EPC cells. When the fraction in peak 1was analysed by SDS-PAGE, it showed three proteinspecies of molecular weights 43, 28 and 22 kDa,respectively. The most prominent protein band wasthat of 43 kDa. These results agree well with thereport by Lee et al. (1997). The fraction of peak 1was then ultrafiltered, lyophilized and used for thein vitro assay for cross-inhibition of bacterial invasioninto fish epithelial cells.

Figure 1 shows the comparative studies ofbiological activity of the major adhesin fromA. hydrophila PPD 134/91 with BSA and the heat-denatured (56 °C, 30 min) major adhesin as thecontrols. Both the heat-denatured adhesin and BSAshow no significant inhibitory activity in EPC cellswhen treated with the virulent strain (PPD 134/91). Since the heat-denatured major adhesin showsno inhibitory activity, the inhibitory action of themajor adhesin may be proteinaceous in nature. Asshown in Fig. 1, 20 µg ml–1 of the major adhesincould exert a significant (P , 0.05) inhibitory effectby competitively inhibiting 56.5% of the

homologous bacterial invasion into EPC cellsin vitro. The relationship between the amount ofmajor adhesin used and the inhibitory effect wasalso studied (Fig. 2). A concentration of 10 µgml–1 of major adhesin showed a significantinhibitory effect.

The mechanism of bacterial invasion into fishepithelial cells is not very clear. According to thegeneral consensus, adherence should be an importantstep before bacteria can invade fish cells. Severalkinds of adhesins have been reported to be involvedin bacterial adhesion and invasion of fish cellsin vitro. Merino et al. (1997) reported that thepresence of a flagellum is essential for adhesion andinvasion in some A. hydrophila strains. However,our previous studies (Lee et al. 1997) found thatthe flagellum could not inhibit A. hydrophila frominvading EPC cells in vitro. Therefore, we suggestthat the flagellum of A. hydrophila may only providethe motility necessary for bacterial cells to contactfish epithelial cells. Lipopolysaccharide (LPS) hasalso been reported as an adhesin by Merino et al.(1996). They found that mutants of A. hydrophiladevoid only of the O-antigen lipopolysaccharideshowed significantly lower levels of adhesion toHep-2 cells than the wild-type bacteria. Therefore,it is possible that adhesin is a term applied to variouscomponents that contribute to the adherence ofvirulent A. hydrophila strains. The major adhesinisolated in this study is most likely an outermembrane protein as pointed out by Lee et al.(1997). It is quite similar to the 43 kDacarbohydrate-reactive outer membrane proteinpurified from a human A. hydrophila isolate whichQuinn et al. (1993) claimed plays an important rolein the initial colonization of bacterial pathogens ofthe human enteric tract.

Cross-inhibition of serologically different strainsof A. hydrophila by the major adhesin

The cross-inhibition effect of the isolated majoradhesin was tested in vitro with nine other strainsof A. hydrophila, seven of which were virulent andtwo of which were weakly virulent (Table 1). Theybelonged to different serological groups except forPPD 122/91 and PPD 70/91 (Leung et al. 1995).Equal amounts of major adhesin (40 µg ml–1) wereincubated with EPC cells before the respectivebacterial cells were added to the mixture. Forcontrols of respective strains, no adhesin was added.The data showed that the major adhesin from

© 1998



Figure 1 Competitive inhibitory effect ofmajor adhesin from Aeromonas hydrophilaPPD 134/91 on the invasion of EPCcells in vitro by the homologous strain(PPD 134/91). The percentage ofbacteria that invaded EPC cells in theabsence of adhesin was taken to be100%. Values shown are the mean 6 SDof duplicates plated on eight TSA platesfor each sample: Group I, no adhesinadded; II, 20 µg ml–1 crude extractpurified from KSCN extraction; III,20 µg ml–1 adhesin added; IV, 20 µg ml–1

BSA added; V, denatured adhesin at aconcentration of 20 µg ml–1 added.*, P , 0.05. Note that the major adhesin(group III) exerted the greatest inhibitoryeffect on bacterial invasion into theEPC cells.

Figure 2 Relationship between theamount of major adhesin fromAeromonas hydrophila PPD 134/91 usedand its inhibitory effect on the invasionof EPC cells by the homologous strain(PPD134/91). The percentage of bacteriathat invaded EPC cells in the absence ofadhesin was taken to be 100%. Valuesshown are the mean 6 SD of duplicateson eight TSA plates for each sample.Note that the higher the dosage ofadhesin used, the greater was thepercentage of inhibition of bacterialinvasion into fish epithelial cells.*, P , 0.05.

A. hydrophila PPD 134/91 significantly (P , 0.05)cross-inhibited bacterial invasion of all the virulentor weakly virulent strains of A. hydrophila tested

in vitro (Fig. 3). Therefore, we conclude that thismajor adhesin could be a very conserved componentin virulent A. hydrophila strains, although the

© 1998



Figure 3 Cross-inhibition of majoradhesin from Aeromonas hydrophila PPD134/91 with other virulent and weaklyvirulent strains of A. hydrophila. Samplesof the major adhesin were used at aconcentration of 40 µg ml–1. Resultswere quantitated as in Fig. 1. Group I,strain PPD134/91; II, strain PPD 11/90;III, strain PPD 122/91; IV, strain PPD70/91; V, strain AH 94063; VI, strainL31; VII, strain L38; VIII, strain ATCC7966; IX, strain 58–20–9; X, strain ST-78–3–3. Note that the major adhesinfrom A. hydrophila PPD 134/91significantly cross-inhibited other virulentor weakly virulent strains ofA. hydrophila. *, P , 0.05.

Figure 4 Cross-inhibition of majoradhesin from Aeromonas hydrophila PPD134/91 with other virulent strains ofAeromonas and Vibrio. Samples ofadhesin used were 20 µg ml–1. Resultswere quantitated as in Fig. 1. Group I,A. sobria CR 79–1–1; II, V. anguillarum811218–5 W; III, V. harveyi W618; IV,V. vulnificus S1/7/93; V, V. anguillarumG/Virus/5 (3). Note that the majoradhesin significantly cross-inhibited thevarious virulent strains of both Aeromonasand Vibrio. *, P , 0.05.

© 1998



biochemical properties of this protein are stillunclear.

Cross-inhibition with Aeromonas sobria andstrains of Vibrio

The same cross-inhibition tests were carried outwith one related member of the Aeromonas family,A. sobria, two strains of Vibrio anguillarum, onestrain of Vibrio harveyi and one strain of Vibriovulnificus using 20 µg ml–1 of major adhesin foreach test. Results showed that the major adhesin alsosignificantly cross-inhibited these virulent strains(Fig. 4) from invading EPC cells. This is the firstreport that adhesin isolated from A. hydrophila canprevent not only A. sobria from invading EPC cellsin vitro, but also other virulent strains of Vibrio spp.This strongly suggests that this major adhesin is avery conserved outer membrane structural proteinpresent in all the enterobacteria.

From electrophoresis results, Lee et al. (1997)suggested that this major adhesin is most likely a43 kDa outer membrane protein. Amino terminalsequence analysis showed substantial homology witha 40 kDa pore-forming carbohydrate-reactive outermembrane protein isolated from human isolateA. hydrophila A6 by Quinn et al. (1994) and a43 kDa outer membrane protein (OMP) isolatedfrom a fish A. salmonicida isolate (Darveau,MacIntyre, Buckley & Hancock 1983).

Beachery (1981) suggested that it might bepossible to prevent serious bacterial infections byinterrupting the interactions of bacterial adhesinswith host receptors. Since then, a plethora of studieshave been carried out in this field, especially in thearea of human infectious diseases (Bisno 1995).A. hydrophila is an important bacterial pathogen offish (Austin & Austin 1987), but no vaccine basedon interrupting bacterial adherence to host tissueshas been developed. This is because A. hydrophilauses multiple, often serologically variable, adhesinssuch as pili and lipopolysaccharides to bind tohost tissues. Vaccines developed from one kind ofadhesin, such as pili, cannot prevent serologicallydifferent virulent strains from invading fish tissues.However, due to the high conservation level of theOMP, it might be possible to develop a vaccineagainst the heterogeneous A. hydrophila and otherbacteria of the family with OMP. Lutwyche, Exner,Hancock & Trust (1995) reported a conserved A.salmonicida porin isolated from an outer membranepreparation with an apparent molecular weight of

28 000. However, they showed that animmunologically cross-reactive protein, althoughpresent in other Aeromonas strains, was not presentin strains of Vibrio. It is interesting that rainbowtrout immunized intraperitoneally with the purifiedporin protein from A. salmonicida were significantlyprotected from heterologous A. salmonicidachallenge. In our present studies, we have shown thatthe purified major adhesin cross-inhibited virulentstrains of both Aeromonas and Vibrio from adhesionand invasion in vitro. Based on this finding, weconclude that the major adhesin obtained fromA. hydrophila PPD 134/91 could be developed intoa vaccine to protect fish from infections caused byheterologous strains of A. hydrophila and Vibrio spp.

Acknowledgments

The authors are grateful to the National Universityof Singapore for providing a research grant for thiswork. We also wish to thank Dr Kini of theBioscience Center, National University of Singapore,for his technical assistance in protein purification.

References

Altmann K., Pyliotis N.A. & Mukkur T.K.S. (1982) A newmethod for the extraction and purification of K99 pili fromenterotoxigenic Escherichia coli and their characterization.Biochemical Journal 201, 505–513.

Angka S.L., Lam T.J. & Sin Y.M. (1995) Some virulencecharacteristics of Aeromonas hydrophila in walking catfish(Clarius gariepinus). Aquaculture 130, 103–112.

Arp L.H. (1988) Bacterial infection of mucosal surfaces: anoverview of cellular and molecular mechanism. In: VirulenceMechanisms of Bacterial Pathogens (ed. by J.A. Roth), pp. 3–27. American Society for Microbiology, Washington, DC.

Atkinson H.M., Adams D., Savvas R.S. & Trust T.J. (1987)Aeromonas adhesin antigen. Experientia 43, 372–374.

Austin B. & Austin D.A. (1987) Bacterial Fish Pathogens, Diseasein Farmed and Wild Fish. Ellis Horwood, Chichester.

Beachery E.H. (1981) Bacterial adherence: adhesin–receptorinteractions mediating the attachment of bacteria to mucosalsurfaces. Journal of Infectious Diseases 143, 325–345.

Bisno A.L. (1995) Molecular aspects of bacterial colonization.Infection Control and Hospital Epidemiology 16, 648–657.

Bradford M. (1976) A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Analytical Biochemistry72, 248.

Darveau R., MacIntyre S., Buckley J.T. & Hancock R.E.W.(1983) Purification and reconstitution in lipid bilayers ofan outer membrane pore-forming protein of Aeromonassalmonicida. Journal of Bacteriology 156, 1006–1011.

© 1998



Dooley J.S.G. & Trust T.J. (1988) Surface protein compositionof Aeromonas hydrophila strains virulent for fish: identificationof a surface array protein. Journal of Bacteriology 170,499–506.

Ho A.S.K., Mietzner T.A., Smith A.J. & Schoolnik G.K.(1990) The pili of Aeromonas hydrophila: identification of anenvironmentally regulated ‘mini pilin’. Journal ofExperimental Medicine 172, 795–806.

Hoepelman A.I.M. & Tuomanen E.I. (1992) Minireview:consequences of microbial attachment: directing host cellfunctions with adhesins. Infection and Immunity 60, 1729–1733.

Hokama A., Honma Y. & Nakasone N. (1990) Pili of anAeromonas hydrophila strain as a possible colonization factor.Microbiology and Immunology 34, 901–915.

Laemmli U.K. (1970) Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227,680–685.

Lee S.Y., Yin Z., Ge R. & Sin Y.M. (1997) Isolation andcharacterization of fish Aeromonas hydrophila adhesinsimportant for in vitro epithelial cell invasion. Journal of FishDiseases 20, 169–175.

Leung K.Y. & Finlay B.B. (1991) Intracellular replication isessential for the virulence of Salmonella typhimurium.Proceedings of the National Academy of Sciences, USA 88,11 470–11 474.

Leung K.Y. & Stevenson R.M.W. (1988a) Characteristics anddistribution of extracellular proteases from Aeromonashydrophila. Journal of General Microbiology 134, 151–160.

Leung K.Y. & Stevenson R.M.W. (1988b) Tn5-induce proteasedeficient strains of Aeromonas hydrophila with reducedvirulence for fish. Infection and Immunity 56, 2639–2644.

Leung K.Y., Yeap I.V., Lam T.J. & Sin Y.M. (1995) Serumresistance as a good indicator for virulence in Aeromonashydrophila strains isolated from diseased fish in South-EastAsia. Journal of Fish Diseases 18, 511–518.

Ljungh A. & Wadstrom T. (1982) Aeromonas toxins.Pharmacology and Therapeutics 15, 339–354.

Lutwyche P., Exner M.M., Hancock R.E.W. & Trust T.J. (1995)A conserved Aeromonas salmonicida porin provides protectiveimmunity to rainbow trout. Infection and Immunity 63,3137–3142.

Merino S., Rubires X., Aguilar A. & Tomas J.M. (1996) TheO:34-antigen lipopolysaccharide as an adhesin in Aeromonashydrophila. FEMS Microbiology Letters 139, 97–101.

Merino S., Rubires X., Aguilar A. & Tomas J.M. (1997) Therole of flagella and motility in the adherence and invasionto fish cell lines by Aeromonas hydrophila serogroup O: 34strains. FEMS Microbiology Letters 151, 213–217.

Popoff M. (1984) Aeromonas. In: Bergey’s Manual of SystematicBacteriology, Vol. 1 (ed. by N.R. Krieg), pp. 545–548.Williams & Wilkins, Baltimore.

Quinn D.M., Atkinson H.M., Bretag A.H., Tester M., TrustT.J., Wong C.Y.F. & Flower R.L.P. (1994) Carbohydrate-reactive, pore-forming outer membrane proteins of Aeromonashydrophila. Infection and Immunity 62, 4054–4058.

Quinn D.M., Wong C.Y.F., Atkinson H.M. & Flower R.L.P.(1993) Isolation of carbohydrate-reactive outer membraneprotein of Aeromonas hydrophila. Infection and Immunity 61,371–377.

Thune R.L., Stanley L.A. & Cooper R.K. (1983) Pathogenesisof Gram-negative bacterial infections in warmwater fish.Annual Review of Fish Diseases 3, 145–185.

in vitro inhibition of epithelial cell invasion by aeromonashydrophila vibrio species by fish...

Documents