- Email: [email protected]
Detection of aerolysin gene in Aeromonas strains isolated from drinking water, fish and foods by the polymerase chain reaction
Camp. Immun. Mhrobiol. infect. Dis. Vol. 18, No. I, pp. 17-26, 1995
Pergamon
Copyright0 1994ElsevierScienceLtd
0147-9571(94)EOOOl-A Printedin Great Britain. All rightsreserved 0147-9571/95$7.00+ 0.00
DETECTION OF AEROLYSIN GENE IN AEROMUNAS STRAINS ISOLATED FROM DRINKING WATER, FISH AND FOODS BY THE POLYMERASE CHAIN REACTION SURAJ B. BALODA,” KAREL KROVACEK,’ LAILA ERIKSSON,’ TOMMY LINNB* and INGMAR M~;NSSON’ ‘Section of Bacteriology and Epizootology and ZSection of Molecular Virology, Department of Veterinary Microbiology, Faculty of Veterinary Medicine, Swedish University of Agricultural Sciences, Biomedical Center, Box 583, S-751 23, Uppsala, Sweden (Received
for publication 14 January 1994)
Abstract-A polymerase chain reaction (PCR) technique was used to assay the presence of the aerolysin gene in a total of 89 Aeromonas hydrophifa and A. sobria strains isolated from drinking water, fish and foods. These strains were also characterized for the production of virulence factors such as haemolysin, protease and cytotoxin. The primers used in the PCR targeted a 209-bp fragment of the aer gene coding for the I-haemolysin and detected template DNA only in haemolytic A. hydrophila strains. The cell-free culture supematants of these aerolysin-positive A. hydrophila strains were also cytotoxic to the HeLa and McCoy cells. The haemolytic A. sobria and non-haemolytic A. hydrophila were consistently negative in the PCR assay. Primer specificity was determined in the PCR by using a control haemolytic Escherichia co/i, Strepfococcus pyogenes and a restriction endonuclease assay. The PCR clearly identified the aerolysin-producing strains of A. hydrophila and may have application as a rapid species-specific virulence test. Key words: Aeromonas, PCR, haemolysin, protease, cytotoxin, virulence.
R&stun&-Une technique de reaction polymtrasique en chaine (PCR) a it& utilisb pour mettre en evidence la presence du gene aerolysine dans un total de 89 souches d’deromonas hydrophila et d’deromonas sobria, isolees de l’eau de boisson, de poissons et d’aliments. Ces souches se caracterisent aussi par la production de facteurs virulents tels que hemolysine, protease et cytotoxine. Les premieres utilisations de la PCR ont localist un fragement de 209-bp du gene aer codant pour la p-hbmolysine et ont d&e& un ADN calibre seulement dans les souches Aeromortas hydrophila hbmolytiques. Les cultures supernaissantes de cellules libres de ces souches d’deromonas hydrophila B aerolysine cytotoxique se sont aussi revelees cytotoxiques pour Ies cellules HeLa et McCoy. Les Aeromonas sobria hbmolytiques et les Aeromonas hydrophila non hemolytiques ont et& constamment negatives a la technique PCR. La specificite premiere de la PCR a CtC dtterrnin6e en la mettant a l’epreuve de la technique de restriction a l’endonucltase et de l’effet htmolytique d’Escherichia co/i, et de Streprococcus pyogenes. La PCR a clairement identifiee les souches productrices d’aerolysine d’deromonas hydrophi/a et pourrait se voir appliquee en tant que test rapide de virulence specifique et d’especes. Mets-clefs:
Aeromonas, PCR, htmolysine,
protease, cytotoxine, virulence.
*Author for correspondence. 17
SUKAJ B. BALODAet al
18
INTRODUCTION
The involvement of Aeromonas spp in the incidences of human diarrhoeal disease and other extra-intestinal infections has been on the increase in recent years [l-3]. It has also been reported that Aeromonas infection in immunocompromized hosts can lead to sepsis or meningitis and such patients as well as immunocompetent hosts can potentially suffer from Aeromonas wound infections [4]. A majority of human isolates of Aeromonas spp are Aeromonas infechaemolytic and soft-tissue necrosis is characteristic of extra-intestinal tions, thereby suggesting that haemolysin may be a virulence factor. Some strains of Aeromonas spp appear to be an important cause of acute diarrhoea in children and in adults, taking at times a fatal course in especially those older than 60 years. It has been suggested that both antibiotic therapy and drinking of untreated water are significant risk factors for susceptible hosts [5]. Our recent studies have reported on the isolation of different Aeromonas spp from various sources e.g. drinking water, fish, foods and marine environments. All these strains when tested in various in vitro assays, showed that they were capable of producing a variety of potential virulence factors [6,7]. These organisms, could thus act as pathogens involved in human disease as well as in fish disease. Several extracellular products of A. hydrophilu which have been suggested as possible contributory factors in the pathogenesis of these diseases include a b-haemolysin (Aerolysin) also known as Asao toxin [S], cytotoxic enterotoxin [9] and a cholera-toxin cross-reactive cytolytic and haemolytic enterotoxin [lo]. It has been suggested that all of these extracellular products are similar to the aerolysin which is an extracellular, soluble, hydrophilic protein exhibiting both, haemolytic and cytolytic properties [ 11, 121. By use of specific primers, the detection of aerolysin gene in strains of A. hydrophila isolated from various sources will help distinguish between fl-haemolysins from A. hydrophila and other Aeromonas spp [12, 131. The study by Pollard et al. [12] describes the use of a PCR technique to detect the Aerolysin gene in faecal isolates of Aeromonas strains isolated from human patients with diarrhoea. Our current study was extended to evaluate the PCR method, and to rapidly and specifically detect aerolysin genes in strains of A. hydrophila isolated from drinking water, fish and various foods. The isolates of A. sobriu from similar sources, were also included in this study to evaluate the specificity of the PCR assay. The results obtained with the PCR assay were also compared with the biological assays for haemolytic and cytotoxic activity. MATERIALS Bacterial
AND
METHODS
strains and culture media
A total of 89 Aeromonas strains (60 A. hydrophila and 29 A. sobria) isolated from drinking water, fish, foods and seawater were included in this study (Table 2). The strains were streaked on blood agar (5% horse blood in brain heart infusion (BHI) agar) and incubated at 22°C for 24 hr [14]. The respective primary cultures were harvested and further inoculated into 100 ml Trypticase Soya Broth (TSB) (Difco) and incubated on a shaker (60 rpm) for 24 hr at 22°C. The cultures were then harvested by centrifugation (16,000 g) at 4°C for 30 min and the supernatants were membrane filtered (Millipore filter, 0.22 pm). Sterile cell-free culture filtrates were tested for the presence of haemolysin, cytotoxin and protease.
Detection of aerolysin gene in Aeromonasusing
19
PCR
Detection of haemolytic, cytotoxic and proteolytic activities
All the strains included in this study were tested for the production of haemolysin and cytotoxin production as described by Krovacek et al. [ 15, 161and for protease production by the method of Brown and Foster [ 171. Preparation of the bacterial samples for the PCR
A small-sized colony from a respective pure culture grown on blood agar was suspended in an Eppendorf tube containing 100~1 of sterile distilled water. The suspension was heated to 96”-100°C for 5 min and later centrifuged for 2 min (3000g) in an Eppendorf centrifuge. The samples were either used immediately or were stored at -20°C until processed. PCR protocol
The PCR was performed using the GeneAmp PCR Reagent kit reagents including the AmpliTaq DNA polymerase enzyme from Perkin-Elmer Cetus. The reaction mixture (including AmpliTaq DNA polymerase) was prepared according to the manufacturer’s guidelines. A final “master mix” was obtained by adding 1 ~1 each of the aerolysin-specific oligonucleotide primers (1 mM Aer 1 and 1 mM Aer2) [12 and Table l] to 65 ~1 of the reaction mixture. PCR was performed in an Eppendorf tube containing 2 ~1 of respective bacterial sample and 18 ~1 of master mix which was finally overlaid with 20 ~1 of mineral oil (Sigma Chemical Co., St Louis, MO.). The following amplification cycles were used: denaturation at 92°C for 1 min, annealing of primers for 1 min at 52°C and the primer extension at 72°C for 1 min. A total of 30 cycles were used followed by a 72°C incubation for 5 min. The samples were analyzed on the submarine 2% agarose gel electrophoresis. The PCR product was visualized by staining the agarose gel with ethidium bromide (5 pg/ml). Restriction endonuclease digestions 5 ~1 of the respective (either aerolysin positive or negative samples of A. hydrophila and A. sobria; one control strain each of haemolytic E. coli and S. pyogenes) samples were
added to 45 ~1 of master mix and amplified by PCR method (described above). This 50 ~1 mixture was then divided into two aliquots of 25 ~1 each. One of the aliquots of the respective samples was subjected to restriction endonuclease digestion with iVci1 enzyme (GIBCO-BRL Inc., Gaithersburg, Md) as recommended by the manufacturer. The original and respective digested samples were analyzed as above.
Table I. Base sequences, locations and predicted sizes of amplified products for the aerolysin olinonucleotide primers Primer AWI Aer2
Oligonucleotide seouence 1%3’ . . 1” , CCAAGGGGTCTGTGGCGACA TTTCACCGGTAACAGGATTG
Location within geneb
Size of amplified uroduct Ibo) .
645-664 834-853
209
‘Aerolysin gene sequence published by Howard er al. [13]. bin nucleotides.
SURAJB. BALODAet al
20
Instruments and chemicals used Ecosyn D300 and Mastercycler from Eppendorf-Netheler-Hinz GmbH, Hamburg, Germany were used to synthesize the oligonucleotide primers and performing the PCR, respectively. GeneAmp PCR Reagent kit and AmpliTaq DNA polymerase were purchased from Perkin-Elmer Cetus (Norwalk, Conn.). The chemicals and reagents for electrophoresis, and molecular weight standards were purchased from Bio-Rad Laboratories AB, Solna, and Pharmacia, Uppsala, Sweden, respectively.
RESULTS
Detection
qf haemolytic, cytotoxic and proteolytic activities of dtyerent strains
All the A. hydrophila and A. sohria strains included in this study were characterized for the production of the above-mentioned extracellular products. Of the 29 A. sobria strains, 16 strains (55%) produced protease. Eight strains produced /I-haemolysin while only one strain produced cr-haemolysin. Majority of these strains were cytotoxin positive as approx. 78% (22/29) of the A. sobria strains produced cytotoxic activity in the HeLa and McCoy cells. One A. sobria strain isolated from pork did not produce any of the enzymes. Of the 60 A. hydrophila strains of different origin, 41 strains produced protease while 16 were “weak positive” for the production of protease and the remaining three were negative for the assay. Majority of the strains (58/60) produced cytotoxin as detected by the assay on HeLa and McCoy cells in tissue culture. Furthermore, 52 of the 60 strains tested produced /?-haemolytic activity on horse blood agar. Only eight strains were haemolysin negative. Finally, the strains did not show any typical pattern of production of different virulence determinants with respect to their source of isolation since none of the isolates from a particular source produced all the extracellular products or showed a typical pattern of virulence profile. Aerolysin speccfic PCR Aerolysin-specific nucleotide probes were prepared according to Pollard et al. [12]. The pairs of these synthetic primers targeted a 209-bp fragment of the aer gene coding for the hole-forming aerolysin toxin were used in the PCR. The results obtained are summarized in Table 2 showing the presence or absence of the amplified product when the representative A. hydrophila and A. sobria were used as templates. An amplification of the expected size of 209-bp was observed mainly in the b-haemolytic strains of A. hydrophila. In majority of the cases, no similar fragments were observed in the PCR when non-haemolytic A. hydrophila (8 strains) or haemolytic and non-haemolytic A. sobria, haemolytic E. coli and S. pyogenes were used. Results from selected strains are shown in Fig. 1. However, four A. hydrophila strains isolated from drinking water (VLA 49, VLA 51, VLA 55 and VLA 71) did not show the presence of 209-bp amplification fragment despite the fact that these strains were fl-haemolytic. Spectficity of oligonucleotide probes .for detection of aerolysin gene sequences The primer specificity was determined in the PCR by using selected strains of A. hydrophila, A. sobria, E. coli and Str. pyogenes and the PCR clearly identified all 6
Detection of aerolysin gene in
Aeromonas
21
using PCR
Table 2. Summary of protease, cytotoxin and haemolysin production and PCR probe results of the Aeromonas strains isolated from different sources Production of Strain no.
Organism
Protease
Cytotoxin
Haemolysinb
1. Isolated from drinking water 62 81 82 VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA
8 21 28 31 35 40 42 54 56 57 58 59 66 68 72 73 71 78 79 Al A4 A6 Al 2 3 9 IO II 12 24 25 26 29 30 36 38 39 43 44 45 46 47 48 49 50 51 52 52 55 60 61 61 70 71 75 16 83
A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A.
sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophiia hydrophila hydrophilu hydrophila hydrophila hydrophila hydrophila hydrophiia hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila
+ + + _ _ _ _ _ (k)” + + + + + + -
(k) + +
A+ + _ (+) +
(1, (+) (+)
(+)
(1, (+)
A (+) (+) (+) (+) (+)
(+) (+) + + + + + + + + + + + + + + +
+ + + + + + + + + + + + _ _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
PCR amplification product (209 bp)
_ -
_ _ _
-a -
+ + -
_ _ _ _ _ _ _ _ _ _ _
+ + + + + + + -
+ + + + + + + _
+ + + + + + + + + + + + + + + + •t -
+ + + + + + + + + + + + + + + + _ _ _
+ -
+ + + + + + + + + + + +
+ + _ + + + + _ + + +
Continued overeleqf‘Weak positive reaction. bHaemolysin production by the strains was either OLor fl or the strains were non-haemolytic.
SURAJ B. BALOVA er al.
22 Table 2.-Conrinuud
Organism
stram no. VLA VLA VLA VLA M800 M855
84 85 86 87
Protease
A. A. A. A. A. A.
hydrophila hydrophila hydrophila hydrophila hydrophila hydrophilu
+ + + + t +
A. A. A. A. A. A. A. A. A. A. A. A.
sobrta sobriu hydrophilu hydrophila hydrophila hydrophila hydrophila hydrophila hydrophilu hydrophilo hvdrophila hydrophila
+ + + + t + +
Production of -~- ----~ Haemoly& Cytotoxin
+ + t t + t II.
3&74 39-74 Ba5 LLI 54 55-74 70-4 83 174 472 650 I865
I + + t
Is&ted + t + t + + + + + t t +
PCR amplification product (209 bp)
+ + + + + + from fish + + + f + + _ + + + +
t t t + + + +
111. Isolated from foods (0) Fish 2D 1.1 3.1 (h) Pork I I.D I2 A (c) Chicken 14.8 15.6 I6 D 14.1 IS.4 I6 A
A. sohria A. hydrophilo A. hydrophila
+
A. sobriu A. hvdrophila
t
A. A. A. A. A. A.
sobrlu sobriu sobriu hydrophilu hpdrophila hydrophilo
i-
t +
t t t
_
t t t t t
+ + + t
t
+
+
_
dWeak positive reaction. “Haemolysin production by the strains was either * or /j or the strains were non-haemolytic
aerolysin-positive strains of A. hydrophilu (Fig. 1) The non-haemolytic strains hydrophila did not show amplification (Fig. 2) as did the A. sobriu strains. Non-speciJic ampl$cation
of A.
of the aerolysin gene
Only one A. hydrophilu strain (VLA 50) out of a total of 89 Aeromonas strains tested produced non-specific amplification of the aer gene in the PCR assay as was judged by the analysis on 2% agarose gels (Fig. 1, Lane 9) and restriction enzyme analysis of the PCR products (Fig. 2, lanes 6 and 7). DISCUSSION This study evaluates the PCR method to detect the aerolysin genes in strains of A. hydrophilu. While Pollard et ul. [12] tested the PCR assay system using human isolates of A. hydrophilr, our study extends it further to detect the aerolysin gene in Aeromonus strains isolated from drinking water, fish and foods, in Sweden and Italy. It is known that the distinction between /j’-haemolysins of strains of A. hydrophilu and A. sohriu is not possible using biological assays but the PCR assay targeting the aer gene only, distinguished the p-haemolytic A. hydrophilu strains from the A. sohria strains, isolated from different sources. None of the 8 fl-haemolytic strains of A. sohria showed any primer specificity.
Fig. 1. Detection of 209-bp amplification fragment in selected strains of Aeromonas and control strains using PCR. Lanes I and 14; 100 bp marker ladder. Lane 2; A. .sohria 1 ID. Lanes 3%7and IO show /I-haemolytic A. hydrophila strains. Lane 3; Strain VLA 12. Lone 4; VLA 26. Lane 5: VLA 36. Lane 6; VLA 44. Lane 7; VLA 60. Lane 8; VLA 10 (b-haemolysin (/I-Hly) negative). Lane 9; VLA 50 (/I-Hly negative) and Lane IO; Strain 12 A. Lane II; S. pyogenes44-1032. Lane 12: Haemolytic E. co/i K88263 and Lane 13; Blank. Fig. 2. Determination of specificity of the 209-bp amplification fragment by using restriction endonuclease enzyme Neil digestion. Lanes I and 13; IOO-bp marker ladder. Undigested PCR amplification fragments appear in Lane 2 (containing A. sobria strain I l.D; haemoiysinand aer-negative), Lanes 4 and 6 (containing A. hydrophila strains VLA 10 and VLA 50; haemolysinand aer-negative). Lanes Sand locontain undigested PCR amplification products of A. hydrophilu strains VLA 12 and VLA ‘36 while their respective Neil digests appear in Lanes 9 and Il. Lane 12 is a blank.
23
Detection of aerolysin gene in Aeromonas using PCR
25
The control strains of E. coli, defined in the terms of production of virulence-associated cytotoxins, verotoxins, enterotoxins etc. were consistently negative in the PCR assays. Furthermore, the non-haemolytic A. hydrophila and haemolytic A. sobria which produced either cytotoxin or protease (or both) were negative in the assay indicating that the presence or absence of these toxins did not affect the detection of aerolysin gene. Aerolysin has been established to be a virulence factor contributing to the pathogenesis of A. hydrophila [I 81. The prevalence of this virulence factor in the majority of isolates of A. hydrophila from drinking water, fish and foods reported in this study and in our previous studies [6,7, 15, 161indicates that the factor may be a quite common trait of this pathogen since a significant correlation between the production of cytotoxic haemolysin and the presence of diarrhoea has been reported [19]. Aerolysin might thus be quite an important virulence factor since 58 of the 60 A. hydrophila strains tested produced cytotoxin. This observation corroborates our previous reports [6, 7, 15, 161 and the study by Palumbo et al. [20]. This study clearly indicates the potential of the PCR method in detecting the aerolysin gene in Aeromonas strains isolated from different sources. In contrast with the method used by Pollard et al. [12] who used purified nucleic acids for the PCR assay, we simplified the method by using bacterial colonies directly from the pure culture plates for the preparation of DNA. It thus makes the PCR protocol to be simpler and faster. Furthermore, the method may act as an important tool in rapidly identifying the species-specific virulence tests. Acknowledgments-These studies were supported by grants from the Swedish Council for Forestry and Agricultural Research (Grant Nos 98.0485/92 (to SBB), 0926/86 V5l:l and 50.0477/93 to KK). We would also like to gratefully acknowledge the equipment grants from the FRN, Stockholm to the authors.
REFERENCES I. Baman S. J. Aeromonas hydrophila as the etiologic agent in severe gastroenteritis: report of a case. Am. J. Med. Technol. 46, 179-181 (1980). 2. Janda J. M. and Duffey P. S. Mesophilic aeromonads in human disease: current taxonomy, laboratory identification. and infectious disease spectrum. Rev. Infecf. Dis. 10. 980-997 (1988). 3. Altwegg M. and Geiss H. K. Aeromonas as a human pathogen. Crir. Rev. Gicrobiol. 16, 253-286 (1989). 4. Rolston K. V. I., Zandvliet S. E., Rodrigues S., Nguyen H. T. and Bodey G. P. Spectrum of Aeromonas and Plesiomonas infections in patients with cancer and AIDS. Experienfiu 47, 437439 (1991). 5. Moyer N. P. Clinical significance of Aeromonas species isolated from patients with diarrhoea. J. C/in. Microbial. 25, 20442048 (1987). 6. Krovacek K., Faris A., Baloda S. B., Peterz M., Lindberg T. and Mansson I. Prevalence and characterization of Aeromonus spp. isolated from foods in Uppsala, Sweden. Fd Microbial. 9, 29-36 (1992). 7. Krovacek K., Faris A., Baloda S. B., Lindberg T., Peterz M. and Mansson I. Isolation and virulence profiles of Aeromonus spp. from different municipal drinking water supplies in Sweden. Fd Microbial. 9, 215-222 (1992). 8. Asao T., Kinoshita Y., Kozaki S., Uemura T. and Sakaguchi G. Purification and some properties of Aeromonas hydrophila hemolysin. Infecr. Immun. 46, 122-127 (1984). 9. Burke V., Robinson J., Berry R. J. and Gracey M. Detection of enterotoxins of Aeromonas hydrophilu by a suckling mouse test. J. Med. Microbial. 14, 401408 (198 I). 10. Rose J. M., Houston S. W., Coppenhaver D. H., Dixon J. D. and Kurosky A. Purification and chemical characterization of a cholera-cross-reactive cytolytic enterotoxin produced by a human isolate of Aeromonas hydrophila. Infect. Immun. 57, 1165-I 169 ( 1989).
I I. Howard S. P. and Buckley J. T. Protein export by a gram-negative bacterium: production of aerolysin by Aeromonas hydrophila. J. Eucrer. 161, I1 18-1124 (1985). 12 Pollard D. R., Johnson W. M., Lior H., Tyler S. D. and Rozee K. R. Detection of the aerolysin gene in Aeromonas hydrophila by the polymerase chain reaction. J. Clin. Microbial. 28, 2477-2481 (1990).
26
SUKAJB. BALODA et al
13. Howard S. P., Garland W. J., Green M. J. and Buckley J. T. Nucleotide sequence of the gene for hole-forming toxin Aerolysin of Aeromonas hydrophila. J. Bact. 169, 2869-2871 (1987). 14. Krovacek K, Faris A. and Mansson I. Growth of and toxin production by Aeromonns hydrophila and Aeromonas sobria at low temperatures. Int. J. Fd Microbial. 13, 1655176 (1991). 15. Krovacek K., Faris A., Eriksson L., Jansson E., Ljungberg 0. and M&son I. Virulence, cytotoxic and inflammatory activities of Vibrio unguillarum and Aeromonas salmonicida isolated from cultivated salmonid fish in Sweden. Acta Vet. &and. 28, 47-54 (1987). 16. Krovacek K., Peterz M., Faris A. and Mansson I. Enterotoxigenic and drug sensitivity of Aeromonas hydrophila isolated from well water in Sweden: a case study. Int. J. Fd Microbial. 8, 149-154 (1989). 17. Brown M. R. W. and Foster J. H. S. A simple diagnostic milk medium for Pseudomonas aeruginosa. J. Clin. Path. 23, 171-177 (1970). 18. Stelma G. N., Johnson C. H. and Spaulding P. Evidence for the direct involvement of beta-hemolysin in Aeromonas hydrophila enteropathogenicity. Curr. Microbial. 14, 71-77 (1986). 19. Brauer C.. Scheftel J. M.. Rihn B. and Monte11 H. Isolement de Aeromonas hvdroohila dans les diarrhees: caractbrisation des souches enterotoxinogenes et relations cliniques Ann. Biol ‘Clin: 43, 725-73 I (1985). 20. Palumbo S. A., Bencivengo M. M., Del Corral F., Williams A. C. and Buchanan R. L. Characterization of the Aeromonus hydrophila group isolated from retail foods of animal origin. J. Clin. Microbial. 27, 854-859 (1989).
Pergamon
Copyright0 1994ElsevierScienceLtd
0147-9571(94)EOOOl-A Printedin Great Britain. All rightsreserved 0147-9571/95$7.00+ 0.00
DETECTION OF AEROLYSIN GENE IN AEROMUNAS STRAINS ISOLATED FROM DRINKING WATER, FISH AND FOODS BY THE POLYMERASE CHAIN REACTION SURAJ B. BALODA,” KAREL KROVACEK,’ LAILA ERIKSSON,’ TOMMY LINNB* and INGMAR M~;NSSON’ ‘Section of Bacteriology and Epizootology and ZSection of Molecular Virology, Department of Veterinary Microbiology, Faculty of Veterinary Medicine, Swedish University of Agricultural Sciences, Biomedical Center, Box 583, S-751 23, Uppsala, Sweden (Received
for publication 14 January 1994)
Abstract-A polymerase chain reaction (PCR) technique was used to assay the presence of the aerolysin gene in a total of 89 Aeromonas hydrophifa and A. sobria strains isolated from drinking water, fish and foods. These strains were also characterized for the production of virulence factors such as haemolysin, protease and cytotoxin. The primers used in the PCR targeted a 209-bp fragment of the aer gene coding for the I-haemolysin and detected template DNA only in haemolytic A. hydrophila strains. The cell-free culture supematants of these aerolysin-positive A. hydrophila strains were also cytotoxic to the HeLa and McCoy cells. The haemolytic A. sobria and non-haemolytic A. hydrophila were consistently negative in the PCR assay. Primer specificity was determined in the PCR by using a control haemolytic Escherichia co/i, Strepfococcus pyogenes and a restriction endonuclease assay. The PCR clearly identified the aerolysin-producing strains of A. hydrophila and may have application as a rapid species-specific virulence test. Key words: Aeromonas, PCR, haemolysin, protease, cytotoxin, virulence.
R&stun&-Une technique de reaction polymtrasique en chaine (PCR) a it& utilisb pour mettre en evidence la presence du gene aerolysine dans un total de 89 souches d’deromonas hydrophila et d’deromonas sobria, isolees de l’eau de boisson, de poissons et d’aliments. Ces souches se caracterisent aussi par la production de facteurs virulents tels que hemolysine, protease et cytotoxine. Les premieres utilisations de la PCR ont localist un fragement de 209-bp du gene aer codant pour la p-hbmolysine et ont d&e& un ADN calibre seulement dans les souches Aeromortas hydrophila hbmolytiques. Les cultures supernaissantes de cellules libres de ces souches d’deromonas hydrophila B aerolysine cytotoxique se sont aussi revelees cytotoxiques pour Ies cellules HeLa et McCoy. Les Aeromonas sobria hbmolytiques et les Aeromonas hydrophila non hemolytiques ont et& constamment negatives a la technique PCR. La specificite premiere de la PCR a CtC dtterrnin6e en la mettant a l’epreuve de la technique de restriction a l’endonucltase et de l’effet htmolytique d’Escherichia co/i, et de Streprococcus pyogenes. La PCR a clairement identifiee les souches productrices d’aerolysine d’deromonas hydrophi/a et pourrait se voir appliquee en tant que test rapide de virulence specifique et d’especes. Mets-clefs:
Aeromonas, PCR, htmolysine,
protease, cytotoxine, virulence.
*Author for correspondence. 17
SUKAJ B. BALODAet al
18
INTRODUCTION
The involvement of Aeromonas spp in the incidences of human diarrhoeal disease and other extra-intestinal infections has been on the increase in recent years [l-3]. It has also been reported that Aeromonas infection in immunocompromized hosts can lead to sepsis or meningitis and such patients as well as immunocompetent hosts can potentially suffer from Aeromonas wound infections [4]. A majority of human isolates of Aeromonas spp are Aeromonas infechaemolytic and soft-tissue necrosis is characteristic of extra-intestinal tions, thereby suggesting that haemolysin may be a virulence factor. Some strains of Aeromonas spp appear to be an important cause of acute diarrhoea in children and in adults, taking at times a fatal course in especially those older than 60 years. It has been suggested that both antibiotic therapy and drinking of untreated water are significant risk factors for susceptible hosts [5]. Our recent studies have reported on the isolation of different Aeromonas spp from various sources e.g. drinking water, fish, foods and marine environments. All these strains when tested in various in vitro assays, showed that they were capable of producing a variety of potential virulence factors [6,7]. These organisms, could thus act as pathogens involved in human disease as well as in fish disease. Several extracellular products of A. hydrophilu which have been suggested as possible contributory factors in the pathogenesis of these diseases include a b-haemolysin (Aerolysin) also known as Asao toxin [S], cytotoxic enterotoxin [9] and a cholera-toxin cross-reactive cytolytic and haemolytic enterotoxin [lo]. It has been suggested that all of these extracellular products are similar to the aerolysin which is an extracellular, soluble, hydrophilic protein exhibiting both, haemolytic and cytolytic properties [ 11, 121. By use of specific primers, the detection of aerolysin gene in strains of A. hydrophila isolated from various sources will help distinguish between fl-haemolysins from A. hydrophila and other Aeromonas spp [12, 131. The study by Pollard et al. [12] describes the use of a PCR technique to detect the Aerolysin gene in faecal isolates of Aeromonas strains isolated from human patients with diarrhoea. Our current study was extended to evaluate the PCR method, and to rapidly and specifically detect aerolysin genes in strains of A. hydrophila isolated from drinking water, fish and various foods. The isolates of A. sobriu from similar sources, were also included in this study to evaluate the specificity of the PCR assay. The results obtained with the PCR assay were also compared with the biological assays for haemolytic and cytotoxic activity. MATERIALS Bacterial
AND
METHODS
strains and culture media
A total of 89 Aeromonas strains (60 A. hydrophila and 29 A. sobria) isolated from drinking water, fish, foods and seawater were included in this study (Table 2). The strains were streaked on blood agar (5% horse blood in brain heart infusion (BHI) agar) and incubated at 22°C for 24 hr [14]. The respective primary cultures were harvested and further inoculated into 100 ml Trypticase Soya Broth (TSB) (Difco) and incubated on a shaker (60 rpm) for 24 hr at 22°C. The cultures were then harvested by centrifugation (16,000 g) at 4°C for 30 min and the supernatants were membrane filtered (Millipore filter, 0.22 pm). Sterile cell-free culture filtrates were tested for the presence of haemolysin, cytotoxin and protease.
Detection of aerolysin gene in Aeromonasusing
19
PCR
Detection of haemolytic, cytotoxic and proteolytic activities
All the strains included in this study were tested for the production of haemolysin and cytotoxin production as described by Krovacek et al. [ 15, 161and for protease production by the method of Brown and Foster [ 171. Preparation of the bacterial samples for the PCR
A small-sized colony from a respective pure culture grown on blood agar was suspended in an Eppendorf tube containing 100~1 of sterile distilled water. The suspension was heated to 96”-100°C for 5 min and later centrifuged for 2 min (3000g) in an Eppendorf centrifuge. The samples were either used immediately or were stored at -20°C until processed. PCR protocol
The PCR was performed using the GeneAmp PCR Reagent kit reagents including the AmpliTaq DNA polymerase enzyme from Perkin-Elmer Cetus. The reaction mixture (including AmpliTaq DNA polymerase) was prepared according to the manufacturer’s guidelines. A final “master mix” was obtained by adding 1 ~1 each of the aerolysin-specific oligonucleotide primers (1 mM Aer 1 and 1 mM Aer2) [12 and Table l] to 65 ~1 of the reaction mixture. PCR was performed in an Eppendorf tube containing 2 ~1 of respective bacterial sample and 18 ~1 of master mix which was finally overlaid with 20 ~1 of mineral oil (Sigma Chemical Co., St Louis, MO.). The following amplification cycles were used: denaturation at 92°C for 1 min, annealing of primers for 1 min at 52°C and the primer extension at 72°C for 1 min. A total of 30 cycles were used followed by a 72°C incubation for 5 min. The samples were analyzed on the submarine 2% agarose gel electrophoresis. The PCR product was visualized by staining the agarose gel with ethidium bromide (5 pg/ml). Restriction endonuclease digestions 5 ~1 of the respective (either aerolysin positive or negative samples of A. hydrophila and A. sobria; one control strain each of haemolytic E. coli and S. pyogenes) samples were
added to 45 ~1 of master mix and amplified by PCR method (described above). This 50 ~1 mixture was then divided into two aliquots of 25 ~1 each. One of the aliquots of the respective samples was subjected to restriction endonuclease digestion with iVci1 enzyme (GIBCO-BRL Inc., Gaithersburg, Md) as recommended by the manufacturer. The original and respective digested samples were analyzed as above.
Table I. Base sequences, locations and predicted sizes of amplified products for the aerolysin olinonucleotide primers Primer AWI Aer2
Oligonucleotide seouence 1%3’ . . 1” , CCAAGGGGTCTGTGGCGACA TTTCACCGGTAACAGGATTG
Location within geneb
Size of amplified uroduct Ibo) .
645-664 834-853
209
‘Aerolysin gene sequence published by Howard er al. [13]. bin nucleotides.
SURAJB. BALODAet al
20
Instruments and chemicals used Ecosyn D300 and Mastercycler from Eppendorf-Netheler-Hinz GmbH, Hamburg, Germany were used to synthesize the oligonucleotide primers and performing the PCR, respectively. GeneAmp PCR Reagent kit and AmpliTaq DNA polymerase were purchased from Perkin-Elmer Cetus (Norwalk, Conn.). The chemicals and reagents for electrophoresis, and molecular weight standards were purchased from Bio-Rad Laboratories AB, Solna, and Pharmacia, Uppsala, Sweden, respectively.
RESULTS
Detection
qf haemolytic, cytotoxic and proteolytic activities of dtyerent strains
All the A. hydrophila and A. sohria strains included in this study were characterized for the production of the above-mentioned extracellular products. Of the 29 A. sobria strains, 16 strains (55%) produced protease. Eight strains produced /I-haemolysin while only one strain produced cr-haemolysin. Majority of these strains were cytotoxin positive as approx. 78% (22/29) of the A. sobria strains produced cytotoxic activity in the HeLa and McCoy cells. One A. sobria strain isolated from pork did not produce any of the enzymes. Of the 60 A. hydrophila strains of different origin, 41 strains produced protease while 16 were “weak positive” for the production of protease and the remaining three were negative for the assay. Majority of the strains (58/60) produced cytotoxin as detected by the assay on HeLa and McCoy cells in tissue culture. Furthermore, 52 of the 60 strains tested produced /?-haemolytic activity on horse blood agar. Only eight strains were haemolysin negative. Finally, the strains did not show any typical pattern of production of different virulence determinants with respect to their source of isolation since none of the isolates from a particular source produced all the extracellular products or showed a typical pattern of virulence profile. Aerolysin speccfic PCR Aerolysin-specific nucleotide probes were prepared according to Pollard et al. [12]. The pairs of these synthetic primers targeted a 209-bp fragment of the aer gene coding for the hole-forming aerolysin toxin were used in the PCR. The results obtained are summarized in Table 2 showing the presence or absence of the amplified product when the representative A. hydrophila and A. sobria were used as templates. An amplification of the expected size of 209-bp was observed mainly in the b-haemolytic strains of A. hydrophila. In majority of the cases, no similar fragments were observed in the PCR when non-haemolytic A. hydrophila (8 strains) or haemolytic and non-haemolytic A. sobria, haemolytic E. coli and S. pyogenes were used. Results from selected strains are shown in Fig. 1. However, four A. hydrophila strains isolated from drinking water (VLA 49, VLA 51, VLA 55 and VLA 71) did not show the presence of 209-bp amplification fragment despite the fact that these strains were fl-haemolytic. Spectficity of oligonucleotide probes .for detection of aerolysin gene sequences The primer specificity was determined in the PCR by using selected strains of A. hydrophila, A. sobria, E. coli and Str. pyogenes and the PCR clearly identified all 6
Detection of aerolysin gene in
Aeromonas
21
using PCR
Table 2. Summary of protease, cytotoxin and haemolysin production and PCR probe results of the Aeromonas strains isolated from different sources Production of Strain no.
Organism
Protease
Cytotoxin
Haemolysinb
1. Isolated from drinking water 62 81 82 VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA VLA
8 21 28 31 35 40 42 54 56 57 58 59 66 68 72 73 71 78 79 Al A4 A6 Al 2 3 9 IO II 12 24 25 26 29 30 36 38 39 43 44 45 46 47 48 49 50 51 52 52 55 60 61 61 70 71 75 16 83
A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A.
sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria sobria hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophiia hydrophila hydrophilu hydrophila hydrophila hydrophila hydrophila hydrophiia hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila hydrophila
+ + + _ _ _ _ _ (k)” + + + + + + -
(k) + +
A+ + _ (+) +
(1, (+) (+)
(+)
(1, (+)
A (+) (+) (+) (+) (+)
(+) (+) + + + + + + + + + + + + + + +
+ + + + + + + + + + + + _ _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
PCR amplification product (209 bp)
_ -
_ _ _
-a -
+ + -
_ _ _ _ _ _ _ _ _ _ _
+ + + + + + + -
+ + + + + + + _
+ + + + + + + + + + + + + + + + •t -
+ + + + + + + + + + + + + + + + _ _ _
+ -
+ + + + + + + + + + + +
+ + _ + + + + _ + + +
Continued overeleqf‘Weak positive reaction. bHaemolysin production by the strains was either OLor fl or the strains were non-haemolytic.
SURAJ B. BALOVA er al.
22 Table 2.-Conrinuud
Organism
stram no. VLA VLA VLA VLA M800 M855
84 85 86 87
Protease
A. A. A. A. A. A.
hydrophila hydrophila hydrophila hydrophila hydrophila hydrophilu
+ + + + t +
A. A. A. A. A. A. A. A. A. A. A. A.
sobrta sobriu hydrophilu hydrophila hydrophila hydrophila hydrophila hydrophila hydrophilu hydrophilo hvdrophila hydrophila
+ + + + t + +
Production of -~- ----~ Haemoly& Cytotoxin
+ + t t + t II.
3&74 39-74 Ba5 LLI 54 55-74 70-4 83 174 472 650 I865
I + + t
Is&ted + t + t + + + + + t t +
PCR amplification product (209 bp)
+ + + + + + from fish + + + f + + _ + + + +
t t t + + + +
111. Isolated from foods (0) Fish 2D 1.1 3.1 (h) Pork I I.D I2 A (c) Chicken 14.8 15.6 I6 D 14.1 IS.4 I6 A
A. sohria A. hydrophilo A. hydrophila
+
A. sobriu A. hvdrophila
t
A. A. A. A. A. A.
sobrlu sobriu sobriu hydrophilu hpdrophila hydrophilo
i-
t +
t t t
_
t t t t t
+ + + t
t
+
+
_
dWeak positive reaction. “Haemolysin production by the strains was either * or /j or the strains were non-haemolytic
aerolysin-positive strains of A. hydrophilu (Fig. 1) The non-haemolytic strains hydrophila did not show amplification (Fig. 2) as did the A. sobriu strains. Non-speciJic ampl$cation
of A.
of the aerolysin gene
Only one A. hydrophilu strain (VLA 50) out of a total of 89 Aeromonas strains tested produced non-specific amplification of the aer gene in the PCR assay as was judged by the analysis on 2% agarose gels (Fig. 1, Lane 9) and restriction enzyme analysis of the PCR products (Fig. 2, lanes 6 and 7). DISCUSSION This study evaluates the PCR method to detect the aerolysin genes in strains of A. hydrophilu. While Pollard et ul. [12] tested the PCR assay system using human isolates of A. hydrophilr, our study extends it further to detect the aerolysin gene in Aeromonus strains isolated from drinking water, fish and foods, in Sweden and Italy. It is known that the distinction between /j’-haemolysins of strains of A. hydrophilu and A. sohriu is not possible using biological assays but the PCR assay targeting the aer gene only, distinguished the p-haemolytic A. hydrophilu strains from the A. sohria strains, isolated from different sources. None of the 8 fl-haemolytic strains of A. sohria showed any primer specificity.
Fig. 1. Detection of 209-bp amplification fragment in selected strains of Aeromonas and control strains using PCR. Lanes I and 14; 100 bp marker ladder. Lane 2; A. .sohria 1 ID. Lanes 3%7and IO show /I-haemolytic A. hydrophila strains. Lane 3; Strain VLA 12. Lone 4; VLA 26. Lane 5: VLA 36. Lane 6; VLA 44. Lane 7; VLA 60. Lane 8; VLA 10 (b-haemolysin (/I-Hly) negative). Lane 9; VLA 50 (/I-Hly negative) and Lane IO; Strain 12 A. Lane II; S. pyogenes44-1032. Lane 12: Haemolytic E. co/i K88263 and Lane 13; Blank. Fig. 2. Determination of specificity of the 209-bp amplification fragment by using restriction endonuclease enzyme Neil digestion. Lanes I and 13; IOO-bp marker ladder. Undigested PCR amplification fragments appear in Lane 2 (containing A. sobria strain I l.D; haemoiysinand aer-negative), Lanes 4 and 6 (containing A. hydrophila strains VLA 10 and VLA 50; haemolysinand aer-negative). Lanes Sand locontain undigested PCR amplification products of A. hydrophilu strains VLA 12 and VLA ‘36 while their respective Neil digests appear in Lanes 9 and Il. Lane 12 is a blank.
23
Detection of aerolysin gene in Aeromonas using PCR
25
The control strains of E. coli, defined in the terms of production of virulence-associated cytotoxins, verotoxins, enterotoxins etc. were consistently negative in the PCR assays. Furthermore, the non-haemolytic A. hydrophila and haemolytic A. sobria which produced either cytotoxin or protease (or both) were negative in the assay indicating that the presence or absence of these toxins did not affect the detection of aerolysin gene. Aerolysin has been established to be a virulence factor contributing to the pathogenesis of A. hydrophila [I 81. The prevalence of this virulence factor in the majority of isolates of A. hydrophila from drinking water, fish and foods reported in this study and in our previous studies [6,7, 15, 161indicates that the factor may be a quite common trait of this pathogen since a significant correlation between the production of cytotoxic haemolysin and the presence of diarrhoea has been reported [19]. Aerolysin might thus be quite an important virulence factor since 58 of the 60 A. hydrophila strains tested produced cytotoxin. This observation corroborates our previous reports [6, 7, 15, 161 and the study by Palumbo et al. [20]. This study clearly indicates the potential of the PCR method in detecting the aerolysin gene in Aeromonas strains isolated from different sources. In contrast with the method used by Pollard et al. [12] who used purified nucleic acids for the PCR assay, we simplified the method by using bacterial colonies directly from the pure culture plates for the preparation of DNA. It thus makes the PCR protocol to be simpler and faster. Furthermore, the method may act as an important tool in rapidly identifying the species-specific virulence tests. Acknowledgments-These studies were supported by grants from the Swedish Council for Forestry and Agricultural Research (Grant Nos 98.0485/92 (to SBB), 0926/86 V5l:l and 50.0477/93 to KK). We would also like to gratefully acknowledge the equipment grants from the FRN, Stockholm to the authors.
REFERENCES I. Baman S. J. Aeromonas hydrophila as the etiologic agent in severe gastroenteritis: report of a case. Am. J. Med. Technol. 46, 179-181 (1980). 2. Janda J. M. and Duffey P. S. Mesophilic aeromonads in human disease: current taxonomy, laboratory identification. and infectious disease spectrum. Rev. Infecf. Dis. 10. 980-997 (1988). 3. Altwegg M. and Geiss H. K. Aeromonas as a human pathogen. Crir. Rev. Gicrobiol. 16, 253-286 (1989). 4. Rolston K. V. I., Zandvliet S. E., Rodrigues S., Nguyen H. T. and Bodey G. P. Spectrum of Aeromonas and Plesiomonas infections in patients with cancer and AIDS. Experienfiu 47, 437439 (1991). 5. Moyer N. P. Clinical significance of Aeromonas species isolated from patients with diarrhoea. J. C/in. Microbial. 25, 20442048 (1987). 6. Krovacek K., Faris A., Baloda S. B., Peterz M., Lindberg T. and Mansson I. Prevalence and characterization of Aeromonus spp. isolated from foods in Uppsala, Sweden. Fd Microbial. 9, 29-36 (1992). 7. Krovacek K., Faris A., Baloda S. B., Lindberg T., Peterz M. and Mansson I. Isolation and virulence profiles of Aeromonus spp. from different municipal drinking water supplies in Sweden. Fd Microbial. 9, 215-222 (1992). 8. Asao T., Kinoshita Y., Kozaki S., Uemura T. and Sakaguchi G. Purification and some properties of Aeromonas hydrophila hemolysin. Infecr. Immun. 46, 122-127 (1984). 9. Burke V., Robinson J., Berry R. J. and Gracey M. Detection of enterotoxins of Aeromonas hydrophilu by a suckling mouse test. J. Med. Microbial. 14, 401408 (198 I). 10. Rose J. M., Houston S. W., Coppenhaver D. H., Dixon J. D. and Kurosky A. Purification and chemical characterization of a cholera-cross-reactive cytolytic enterotoxin produced by a human isolate of Aeromonas hydrophila. Infect. Immun. 57, 1165-I 169 ( 1989).
I I. Howard S. P. and Buckley J. T. Protein export by a gram-negative bacterium: production of aerolysin by Aeromonas hydrophila. J. Eucrer. 161, I1 18-1124 (1985). 12 Pollard D. R., Johnson W. M., Lior H., Tyler S. D. and Rozee K. R. Detection of the aerolysin gene in Aeromonas hydrophila by the polymerase chain reaction. J. Clin. Microbial. 28, 2477-2481 (1990).
26
SUKAJB. BALODA et al
13. Howard S. P., Garland W. J., Green M. J. and Buckley J. T. Nucleotide sequence of the gene for hole-forming toxin Aerolysin of Aeromonas hydrophila. J. Bact. 169, 2869-2871 (1987). 14. Krovacek K, Faris A. and Mansson I. Growth of and toxin production by Aeromonns hydrophila and Aeromonas sobria at low temperatures. Int. J. Fd Microbial. 13, 1655176 (1991). 15. Krovacek K., Faris A., Eriksson L., Jansson E., Ljungberg 0. and M&son I. Virulence, cytotoxic and inflammatory activities of Vibrio unguillarum and Aeromonas salmonicida isolated from cultivated salmonid fish in Sweden. Acta Vet. &and. 28, 47-54 (1987). 16. Krovacek K., Peterz M., Faris A. and Mansson I. Enterotoxigenic and drug sensitivity of Aeromonas hydrophila isolated from well water in Sweden: a case study. Int. J. Fd Microbial. 8, 149-154 (1989). 17. Brown M. R. W. and Foster J. H. S. A simple diagnostic milk medium for Pseudomonas aeruginosa. J. Clin. Path. 23, 171-177 (1970). 18. Stelma G. N., Johnson C. H. and Spaulding P. Evidence for the direct involvement of beta-hemolysin in Aeromonas hydrophila enteropathogenicity. Curr. Microbial. 14, 71-77 (1986). 19. Brauer C.. Scheftel J. M.. Rihn B. and Monte11 H. Isolement de Aeromonas hvdroohila dans les diarrhees: caractbrisation des souches enterotoxinogenes et relations cliniques Ann. Biol ‘Clin: 43, 725-73 I (1985). 20. Palumbo S. A., Bencivengo M. M., Del Corral F., Williams A. C. and Buchanan R. L. Characterization of the Aeromonus hydrophila group isolated from retail foods of animal origin. J. Clin. Microbial. 27, 854-859 (1989).