(Plecoglossus altivelis) in Japan

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 1992, p /92/ $02.00/0 Copyright 1992, American Society for Microbiology Vol. 58, No. 9 Vibrio cholerae Non-O1 Isolated from Ayu Fish (Plecoglossus altivelis) in Japan CIIRA KIIYUKIA,l* A. NAKAJIMA,2 T. NAKAI,2 K. MUROGA,2 H. KAWAKAMI,' AND H. HASHIMOTO' Department of Food Microbiology and Hygiene' and Department of Fish Pathology, 2 Faculty of Applied Biological Science, Hiroshima University, Higashi-Hiroshima 724, Japan Received 6 April 1992/Accepted 9 July 1992 A fish pathogen, Vibrio cholerae non-ol, was isolated from diseased ayu fish (Plecoglossus altivelis) collected from rivers in eight prefectural districts of Japan. This organism was found to have biochemical characteristics similar to those of V. cholerae non-o1, except that our isolates were negative for ornithine decarboxylase. Antiserum against an ayu isolate did not agglutinate with the majority of environmental V. cholerae non-ol isolates, but a major 0 antigen was common among the ayu isolates. All strains were hemolytic to sheep erythrocytes, and oral administration of culture supernatants induced fluid accumulation in suckling mice. However, the crude toxin was not lethal to adult mice, and no cholera toxin-like enterotoxins were detected. Vibriosis is a well-described devastating fish disease which occurs in both marine and freshwater environments. The causative agents include Vibno anguillarum, V. cholerae non-o1, V. damsela, V. vulnificus, and others (1). V. cholerae non-o1 was first reported as the causative agent of vibriosis in the freshwater fish ayu (Salmoniformes: Plecoglossidae; Plecoglossus altivelis) in Japan by Muroga et al. (8), but V. cholerae non-o1 infection of ayu was considered an uncommon occurrence (5). However, mass mortality in wild ayu with an epizootiology similar to that of the case mentioned above (8) has recently occurred in various parts of Japan. Several bacterial strains were isolated from moribund ayu caught in different rivers of Japan from 1987 to 1991, and most strains were identified as V. cholerae non- 01. Thus, infection of wild ayu with this organism has been established as a common and widely distributed fish disease in Japan. Various exotoxins produced by environmental V. cholerae non-o1 have been reported, including cholera toxin (CT)-like enterotoxin (4), a fluid-accumulating factor (7), and various cytotoxins (11). We have also examined the nature of exotoxins produced by V. cholerae non-o1 isolated from ayu in order to elucidate the significance of this epizootic with regard to public health. Antigenic relationships among strains isolated from ayu and the environment and sensitivities to commonly used antibiotics were also studied. MATERIALS AND METHODS Test strains. Bacterial strains (54 strains from ayu and 6 strains from the environment [Table 1]) used in this study were either isolated in Shiga Prefecture or received from other areas of Japan. The strains from diseased ayu had been isolated by direct plating onto nutrient agar (Eiken Chemical Co., Ltd.) from the livers, spleens, or kidneys of moribund or dead fish. Those strains from apparently healthy fish were isolated from whole-body homogenates after enrichment in alkaline peptone water (1% peptone, 3% NaCl; ph 9.4) by selective isolation on thiosulfate-citrate-bile salts-sucrose (TCBS) agar (Eiken). Strains PS-7702 and PS-7705 were among the first isolates reported to be associated with * Corresponding author. diseased ayu (8) and were used here as reference strains. The DNA homology between PS-7702 and V. cholerae NIH 35A3 was 86% (8). Six environmental strains isolated from Lake Biwa (Shiga Prefecture) or its influent rivers or from the pond water of an ayu farm were also included in our study for comparison. The strains were maintained at 25 C in T1N1 (1% tryptone, 1% NaCl) semisolid agar (0.7%). Biochemical tests. Selected biochemical tests were performed and the results were evaluated by the criteria of Baumann and Schubert (2). The ability to grow at a NaCl concentration of 0, 3, 6, or 8% was determined in 1% tryptone broth, and turbidity (which indicates growth) was scored visually after 18 and 24 h. Sugar utilization was tested in T1N1 broth with 1% carbohydrates, and Andrade solution was used as an indicator of acid production. Arginine dihydrolase and lysine and ornithine decarboxylase tests were carried out in a medium containing the following (per liter of distilled water): 3 g of yeast extract, 5 g of tryptone, 10 g of the respective amino acids, 1 g of glucose, 0.5 g of agar, 0.02 g of bromocresol purple. The ph of the medium was adjusted to 6.5. The test was carried out under anaerobic conditions in anaerobic GasPak jars (GasPak Systems [BBL]). Hemolytic activity was scored as a zone of hemolysis on sheep blood agar (BBL) after 18 to 24 h of incubation at 37 C. Antibody preparation. Two strains, PSH-9020 and PI- 8701, both from diseased ayu, were used to raise antibodies in rabbits. Formalin-killed cells were suspended in 0.01 M phosphate-buffered saline (PBS) at a concentration of 10 mg/ml and were thoroughly emulsified in an equivalent volume of Freund's complete adjuvant. One-milliliter portions of the mixture were injected twice hypodermically into the back of a rabbit at several different points. Subsequent booster injections without adjuvant were administered in the marginal ear vein until constant values for titers were achieved. The antiserum obtained was heated for 30 min at 56 C to destroy heat-labile components of the complement system and stored at -20 C until use. Antigen preparation. Heat-killed (100 C for 150 min) cells of all the strains were used as antigens in agglutination tests. The freshly prepared antigens were diluted to give an A530 of 0.5 Ȧgglutination titer. In a microtiter plate, 25,ul of 10-fold- 3078

2 VOL. 58, 1992 V. CHOLERAE NON-O1 ISOLATED FROM AYU FISH IN JAPAN 3079 TABLE 1. Locations (prefecture) and dates of isolation of V. cholerae non-o1 strains from ayu and the environment Strain Date (day-mo-yr) Location in Japan Sourcea Strain Date (day-mo-yr) Location in Japan Sourcea N Tochigi D PSH Shiga D N Tochigi D PSH Shiga D N Tochigi D PSH Shiga D N Tochigi D PSH Shiga D N Tochigi D psh Shiga H PI Ibaraki D psh Shiga H PI Ibaraki D psh Shiga H SH Shizuoka D psh Shiga H SH Shizuoka D psh Shiga H SH Shizuoka D psh Shiga H SH Shizuoka D psh Shiga H SH Shizuoka D psh Shiga H SH Shizuoka D psh Shiga H SH Shizuoka D psh Shiga H PI Ishikawa D psh Shiga H AA Aichi D psh Shiga H AA Aichi D psh Shiga E AA Aichi D psh Shiga E AA Aichi D psh Shiga E AA Aichi D PSH Shiga D AA Aichi D psh Shiga H AA Aichi D psh Shiga H AA Aichi D psh Shiga H PS Shiga D psh Shiga H PS Shiga D psh Shiga H PSH Shiga D PT Tokushima D PSH Shiga D PT Tokushima E PSH Shiga D POI Oita D psh Shiga E POI Oita D psh Shiga E PO Oita D a Strains isolated from diseased (D) or apparently healthy (H) ayu fish or from the environment (E). diluted antiserum was serially diluted in PBS. The same volume of antigen was added into each well, and the plate was thoroughly shaken for 5 min. After incubation at 37 C for 2 h and at 4 C overnight, agglutination titers were determined. To eliminate nonspecific agglutination, a titer of less than 80 was regarded as negative. Pathogenicity to ayu. The pathogenicity to ayu of a strain (PSH-9020) was tested by an immersion challenge method. Three groups of 25 ayu (each ayu weighs 10 to 15 g) were immersed in bacterial suspensions of 2.5 x 107, 2.5 x 105, or 0 CFU/ml for 5 min, kept at about 20 C, and monitored for 1 week. Toxin preparation. Eleven strains (seven from diseased fish, two from healthy fish, and two from the environment) were preenriched for 6 to 8 h at 37 C in 2 ml of brain heart infusion (BHI) broth (Eiken), and the young cultures were transferred into 10-ml portions of BHI broth in L-shaped tubes. These tubes were incubated at 30 C by shaking in a water bath for 16 to 18 h and centrifuged (9,000 x g for 15 min at 4 C), and the culture supernatant was filtered (0.22-,um-pore-size filter) and treated as crude toxin for hemolysin analysis. For the detection of CT-like enterotoxin, the strains were grown in medium consisting of Casamino Acids (1.5%) and yeast extract (0.3%) and supplemented with 0.2% glucose and 1% sodium chloride. Hemolysin assay. One-milliliter portions of a 2% washed, defibrinated sheep erythrocyte (Nippon Bio-Test Laboratories, Inc.) suspension in 1 mm Tris-hydrochloric acid buffer (0.85% NaCl plus Tris-HCl, ph 7.2) were added to 1-ml portions of twofold serial dilutions of the crude toxin. The mixture was incubated at 37 C for 2 h and then kept at 4 C for 15 h to allow whole cells to settle. The hemolysin titer was the reciprocal of the highest dilution showing complete erythrocyte hemolysis. CT-like enterotoxin assay. The presence of CT-like toxin was tested using anti-ct sensitized latex (VET-RPLA) from a commercial kit (Denka Seiken). Twofold serial dilutions of the crude toxin were prepared in U-shaped microtiter plates and added to the same amount of sensitized latex. The mixture was shaken by a micromixer and incubated at 37 C for 1 h. Agglutination was observed after 20 to 24 h at 25 C, using unsensitized latex as a negative control and the purified CT (supplied with the kit) as a positive control. Enterotoxin assay in infant mice. Crude toxin (0.1 ml) with Evan's blue dye (0.01%) was given orally to 3-day-old mice, and fluid accumulation (FA) in the stomachs of mice was assessed. Sterilized culture medium was used as a negative control. Test animals were kept at 22 to 25 C for 4 h on white filter paper for easier observation of diarrheal feces. The mice were then sacrificed, and the FA ratios were determined by ascertaining the ratio of the weight of the stomach and intestine to the weight of the rest of the body (9). Four animals were used to test each bacterial strain, and each result represents the average FA ratio of four animals. In addition to FA, enterotoxin production was determined by production of Evan's blue dye-stained watery feces (an indicator of diarrhea) and mortality of the suckling mice. Mortality was recorded as negative when no mice or one mouse died within 48 h and was recorded as positive when two to four mice died within 48 h. Toxicity to adult mice. The toxicity of crude extract to adult mice was examined by injecting 0.5-ml samples of the filtered extract intraperitoneally into 4-week-old male mice. Four animals were used per strain, and the culture medium

3 3080 KIIYUKIA ET AL. was used as a negative control. The number of dead mice was recorded after 48 h. Antibiotic susceptibility tests. The following antibiotics were tested against 31 selected isolates (20 from diseased fish and 11 from healthy fish); ampicillin (10,ug), cephalothin (30,ug), chloramphenicol (30,ug), colistin (10,ug), erythromycin (15 p,g), kanamycin (30,ug), nalidixic acid (30,ug), polymyxin B (300 IU), and tetracycline (30,ug). Well-isolated colonies were transferred into trypto-soy broth and incubated for 6 to 8 h at 37 C. The cultures were diluted in PBS and spread plated on the drug sensitivity test agar (Eiken). The plates were allowed to stand at 4 C for 30 min after the antibiotic disks (KB disk, [Eiken]) were placed on the plates. Incubation was carried out at 37 C overnight, after which the inhibition zones were measured. The susceptibilities of the TABLE 2. Biochemical characteristics of the V. cholerae non-o1 strains isolated from ayu and the environment Result for V. cholerae non-01 strains Result for Biochemical characteristic isolated froma: V. choleror test eab Ayu (54 Environment a strains) (6 strains) Motility Gram stain Cytochrome oxidase Nitrate reduction Sensitivity to 0/129 (150,ug) Swarming Arginine dihydrolase Lysine decarboxylase Ornithine decarboxylase Esculin hydrolysis Growth at 42 C Growth at NaCl: 0% +(W) + + 3% % - - v 8% - Fermentation or oxidation of F F F glucose Gas from glucose fermenta tion Acid production from: L-Arabinose - - m-inositol - - D-Mannose - - v Sucrose Cellobiose 1.9 Maltose + + +* Lactose - - 1* Salicin - - 3* Galactose * Gelatinase Voges-Proskauer reaction Methyl red 16.7 Indole test * Hydrogen sulfide production - - Simmons citrate test * ONPGc + + +* Sheep blood hemolysis * a The numbers are the percentages of strains positive for the characteristic or test. Values of 100 and 0% have been replaced by + and - signs, respectively. (W), weak reaction, v, variable reactions; F, fermentation. b The data indicated by asterisks were obtained by Bryant et al. (3); all the other data were obtained from Baumann and Schubert (2). c ONPG, o-nitrophenyl-13-d-galactopyranoside. TABLE 3. APPL. ENVIRON. MICROBIOL. Serological cross-reactions of two V. cholerae non-01 strains PSH-9020 antiserum PI-8701 antiserum adsorbed Antigen absorbed with strain: with strain: None PSH-9020 PI-8701 None PI-8701 PSH-9020 PSH PI isolates to the drugs were determined according to the manufacturer's instructions. RESULTS Identification. All the ayu isolates were gram-negative, motile, short rods forming smooth colonies on nutrient agar and yellow colonies on TCBS agar. Although the examined strains differed from V. cholerae in being negative for ornithine decarboxylase (49 of 54) and exhibiting poor growth (54 of 54) in 1% tryptone water without NaCl (Table 2), all 54 strains were classified as V. cholerae. Six environmental isolates were also classified as V. cholerae, but they were all positive for ornithine decarboxylase except for one strain isolated from a pond stocked with ayu. Serology. All 60 strains did not agglutinate with anti-v. cholerae 01 serum (Denka Seiken). The homologous agglutination titers of anti-psh-9020 and PI-8701 sera were both 1,280. Cross-absorption tests of antisera with the two strains showed identical antigenicity patterns (Table 3). The agglutination titers of anti-psh-9020 serum against heat-killed cells of the 54 ayu strains and 6 environmental strains of V. cholerae non-o1 are shown in Table 4. Of the 54 ayu strains (including PSH-9020 and PI-8701), 33 (61.1%) strains had homologous 0 antigens (titers of >640), 15 (27.8%) strains had common antigens (titers, 160 to 320), and 6 (11.1%) strains did not react with the antiserum. Of the six nonagglutinable strains, four were from healthy fish and were positive for ornithine decarboxylase, one was from a healthy fish but was negative for ornithine decarboxylase, and one was a reference strain, PS Most of the environmental strains (four of six) were nonagglutinable or agglutinated at low titers, and only two strains shared common antigens. Antibiotic sensitivity. All the tested strains (31 strains) were sensitive to chloramphenicol, nalidixic acid, and tetracycline, while a few displayed partial resistance to cephalothin, erythromycin, colistin, kanamycin, and ampicillin (Table 5). Of the 31 strains tested, 14 (45%) were resistant to both polymyxin B and colistin and 29 (94%) were resistant to TABLE 4. Agglutination titers of anti-v. cholerae non-ol PSH serum with the heat-killed cells of V. cholerae non-o1 strains Agglutination No. of strains from: titer Ayu Environment <80 6a > 1, a Five of the six strains were from healthy fish and one was isolated from a diseased ayu in 1977.

4 VOL. 58, 1992 TABLE 5. Antibiotic Antibiotic susceptibilities of V. cholerae non-o1 strains isolated from ayu % of V. cholerae non-01 strains (n = 31) that are: Resistant Intermediate Susceptible Ampicillin Cephalothin Chloramphenicol Colistin Erythromycin Kanamycin Nalidixic acid Polymyxin B Tetracycline polymyxin B. There was no apparent difference in sensitivity between the strains from diseased and healthy fish. Pathogenicity. The mortality rates of the challenged ayu were 72, 16, and 0% at the doses of 2.5 x 107, 2.5 x 105, and 0 (control) CFU/ml, respectively. The pathogenicity to mice and toxin production of nine ayu isolates and two environmental strains are presented in Table 6. Seven strains had an FA ratio of 0.09 or higher. Five strains (two of them environmental) were capable of causing diarrhea within 4 h, but this was not fatal to the suckling mice. CT-like enterotoxin was not detected in the crude extracts of any tested strains by the method used. Hemolysin production was high for some strains but did not correlate with either FA ratio or diarrhea production. Intraperitoneal injection of the filtered crude extract into adult mice did not have a lethal effect, although one of the four mice injected with the PI-8701 extract died within 20 h (Table 6). DISCUSSION Ayu is one of the most important species for Japanese fisheries. Wild populations of this anadromous fish spend most of their 1-year life in freshwater environments. After spending their first few months in the sea or Lake Biwa, where landlocked stocks of ayu are found, juvenile fish enter rivers in the spring and gradually migrate upstream. By late summer, they begin to migrate back to the lower reaches of the river for spawning. In order to augment declining popu- TABLE 6. V. CHOLERAE NON-O1 ISOLATED FROM AYU FISH IN JAPAN 3081 lations of ayu in many rivers, about 200 tons of ayu fingerlings caught in Lake Biwa are transplanted into several rivers every year. The present investigation revealed that V. cholerae non-o1 infection of ayu, which was originally found more than 10 years ago in the rivers running into Lake Biwa in Shiga Prefecture (8), is now widely distributed in many rivers in various parts of Japan. This infection was supposedly spread by transplantation of juvenile ayu from Lake Biwa into these rivers. This is supported by the finding that all strains isolated from diseased ayu in different rivers share a major common 0 antigen with strains PS-7702 and PSH- 9020, which were isolated from diseased ayu found near or in Lake Biwa in 1977 and 1990, respectively. Environmental isolates of V. cholerae non-o1 generally consist of a variety of 0 serotypes (10), and only two environmental strains, including one from water in an ayu pond, of the six strains tested in this study shared the common antigen with ayu isolates. There was another piece of evidence which supports our speculation that the disease agent was spread from a common source. Almost all the strains (36 of 37) isolated from diseased ayu had a negative ornithine decarboxylase reaction. This is another unique property of ayu isolates by which they can be differentiated from environmental or clinical isolates of V. cholerae which are ornithine decarboxylase positive (2). Other representative fish-pathogenic vibrios, e.g., V. anguillarum, V. damsela, V. ordalii, and V. vulnificus, are also all ornithine decarboxylase negative (1). V. cholerae non-o1 infections in ayu occurred only in summer and early autumn, when the water temperature of the rivers was 22 C or higher. The high temperature is favorable for the growth of the pathogen and unfavorable for the physiology of the host, ayu. Transmission of this pathogen among the fish is not clear, because the pathogen is rarely isolated in aquatic environments. However, some healthy ayu caught in Lake Biwa or in the rivers running into the lake harbored the pathogen without showing any signs of the disease; thus, these fish could act as carriers. V. cholerae non-o1 infection in fish should be investigated not only from the standpoint of fish resources but also from the standpoint of food hygiene. It was demonstrated by Yamanoi et al. (12) that mice were killed by intraperitoneal injection of live cells of V. cholerae non-o1 isolated from diseased ayu. However, filtered culture extracts of similar Toxicity of extracellular products of V. cholerae non-o1 strains isolated from ayu and the environment Result in suckling mice No. of dead adult Strain Sourcea Hemolysin titerb mice after intraperito- CT-like enterotoxin FA ratio Diarrhea Mortality neal injection (n = 4) PSH-9020 D PI-8701 D PS-7702 D PS-7705 D N-5 D AA-9030 D PSH-9028 D psh-9034 H psh-9047 H psh-9026 E psh-9044 E None (control) a Strains isolated from diseased (D) or apparently healthy (H) ayu or from the environment (E). b Reciprocal of the highest dilution showing complete hemolysis of sheep erythrocytes.

5 3082 KIIYUKIA ET AL. isolates had no lethal effect on mice in our experiments, and CT-like enterotoxin was not detected from these isolates. The results of this study support the hypothesis that these ayu isolates are not highly toxic, and the potential for human disease seems not to be serious, at least from the viewpoint of food hygiene. Nevertheless, the danger of food poisoning cannot be completely ruled out, because some isolates produced hemolysin and enterotoxins, as demonstrated by FA ratios in infant mice. Antibiotic resistance patterns displayed were typical to those of V. cholerae (6), except for the high resistance to polymyxin B and colistin. ACKNOWLEDGMENTS We are indebted to M. Endo and the other members of Shiga Prefectural Fisheries Experimental Station for their help during sampling. We thank A. Kumamaru (Ibaraki), H. Kanogi (Tochigi), M. Tanaka (Shizuoka), M. Miyakawa (Aichi), Y. Sugimoto (Ishikawa), K. Sawada (Tokushima) and Y. Yokomatsu (Oita) for supplying the indicated bacterial isolates from diseased ayu. We are thankful to K. Venkateswaran and J. S. Rohovec for reading the manuscript and making helpful suggestions. REFERENCES 1. Austin, B., and D. A. Austin Vibrios, p In Bacterial fish pathogens: disease in farmed and wild fish. Ellis Horwood Ltd., Publisher, Chichester, England. 2. Baumann, P., and R. H. W. Schubert Family II. Vibrionaceae Veron 1965, 5245AL, p In N. R. Krieg and J. G. Holt (ed.), Bergey's manual of systematic bacteriology, vol. 1. The Williams & Wilkins Co., Baltimore. 3. Bryant, T. N., V. J. Lee, P. A. West, and R. R. Colwell APPL. ENVIRON. MICROBIOL. Numerical classification of species of Vibrio and related genera. J. Appl. Bacteriol. 61: Craig, J. P., K. Yamamoto, Y. Takeda, and T. Miwatani Production of cholera-like enterotoxin by a Vibrio cholerae non-o1 strain isolated from the environment. Infect. Immun. 34: Egusa, S Fish disease and public health/food hygiene, p In A treatise on fish diseases. Koseisha-Koseikaku, Tokyo. (In Japanese.) 6. Farmer, J. J., III, F. W. Hickman-Brenner, and M. T. Kelly Vibrio, p In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and H. J. Shadomy (ed.), Manual of clinical microbiology, 4th ed. American Society for Microbiology, Washington, D.C. 7. Gyobu, Y., H. Kodama, and H. Uetake Production and partial purification of a fluid-accumulating factor of non-o1 Vibrio cholerae. Microbiol. Immunol. 32: Muroga, K., S. Takahashi, H. Yamanoi, and M. Nishibushi Non-cholera vibrio isolated from diseased ayu. Bull. Jpn. Soc. Sci. Fish. 45: Nishibushi, M., and R. M. Seidler Responses in suckling mice induced by vibrio virulence factor(s), p In R. R. Colwell (ed.), Vibrios in the environment. John Wiley & Sons, Inc., New York. 10. Sakazaki, K., and T. Shimada Serovars of Vbrio cholerae identified during Jpn. J. Med. Sci. Biol. 30: Venkateswaran, K., H. Horiuchi, H. Nakano, H. Matsuda, and H. Hashimoto Vibrio virulence factors and the quantitative analysis of cytotoxicity elaborated by environmental isolates. Cytobios 66: Yamanoi, H., K. Muroga, and S. Takahashi Physiological characteristics and pathogenicity of NAG vibrio isolated from diseased ayu. Fish Pathol. 15: Downloaded from on July 1, 2018 by guest