- Email: [email protected]
Characterization of four Flavobacterium columnare (Flexibacter columnaris) strains isolated from tropical fish
Veterinary Microbiology 62 (1998) 35±45
Characterization of four Flavobacterium columnare (Flexibacter columnaris) strains isolated from tropical fish A. Decostere*, F. Haesebrouck, L.A. Devriese Laboratory of Veterinary Bacteriology and Mycology, University of Gent, Salisburylaan 133, B 9820 Merelbeke, Belgium Received 15 August 1997; accepted 17 March 1998
Abstract Four Flavobacterium columnare strains (AJS 1±4) were isolated from black mollies (Poecilia sphenops) and platies (Xiphophorus maculatus), showing white spots on the back, head and skin ulcers. The isolates developed characteristic rhizoid yellow pigmented colonies on Shieh agar and typical growth in Shieh broth. They were Gram-negative, filamentous bacteria exhibiting flexing movements. When compared to F. columnare strains isolated from temperate fish, it was noted that the four strains originating from tropical aquarium fish are more capable of growing at higher temperatures, the opposite being true for the strains isolated from temperate fish. Biochemical characterization and agglutination tests proved that the isolated strains could be classified as F. columnare. Low minimal inhibitory concentration (MIC) values were found for chloramphenicol, erythromycin, furazolidone, kanamycin, lincomycin, nalidixic acid, oxytetracycline and streptomycin. MIC values were high for colistin, sulfamethoxazole and neomycin. Pathogenicity studies were performed on black mollies. When these animals were submersed in an infective solution of the F. columnare strains, a marked difference in virulence was noted among the four isolated strains, strain AJS 1 being the most virulent one and strain AJS 4 being of low virulence. # 1998 Elsevier Science B.V. All rights reserved. Keywords: Flavobacterium columnare; Fish; Diagnosis-bacteria; Virulence
* Corresponding author. Tel.: +32 9 2647431; fax: +32 9 2647494; e-mail: [email protected] 0378-1135/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved. PII S 0 3 7 8 - 1 1 3 5 ( 9 8 ) 0 0 1 9 6 - 5
36
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
1. Introduction Flavobacterium columnare has been recognized as a worldwide-occurring pathogen of numerous freshwater fish species (Anderson and Conroy, 1969; Austin and Austin, 1987). It was formerly called Flexibacter columnaris, but in 1996 it was transferred to the genus Flavobacterium (Bernardet et al., 1996). In temperate fish, columnaris disease is recognized by the appearance of greyish white or yellow areas of erosion, usually surrounded by a reddish hyperemic zone, on the body surfaces or the gills of fish. When these lesions occur around the dorsal fin, they are called saddle-back lesions, which are typical for columnaris disease of coldwater fish (Austin and Austin, 1987). Outbreaks of columnaris disease are rarely spontaneous, but are influenced by a combination of environmental (temperature) and other factors stressful to the host, such as high stocking density, high levels of ammonia and organic load (Wakabayashi and Egusa, 1972; Chen et al., 1982; Wakabayashi, 1991). F. columnare is mentioned in most books of tropical aquarium fish diseases as the cause of cotton-wool disease. Post (1987) reported fish exhibiting greyish-white discoloration especially on the head and around the mouth. These may evolve into whitish filaments emerging from the lesions, hence the name mouth fungus. Although F. columnare is described as being an important pathogen (Bassleer, 1983; Tomey, 1985; Butcher, 1992), as far as we know, it has never been actually isolated from tropical aquarium fish, nor has its significance and pathogenicity been studied in any way. In the present paper the isolation and characterization of four F. columnare strains obtained from tropical fish with typical lesions is described. The pathogenicity of these strains was determined by immersion and injection experiments of intact black mollies. Additionally, the role of mechanical injury in transmission of F. columnare was determined. 2. Materials and methods 2.1. Isolation of F. columnare strains from diseased fish Samples of skin mucus were collected from 50 moribund fish originating from 12 different aquaria. The fish species included black mollies (Poecilia sphenops), platies (Xiphophorus maculatus), guppies (Poecilia reticulata), tetras (Cheirodon axelrod). These samples were streaked on Trypticase Soy Agar (Gibco, Paisley, Scotland), Shieh agar (Shieh, 1980) without any additives and Shieh agar supplemented with tobramycine at a concentration of 1 mg/ml (Decostere et al., 1997). For the supplemented Shieh agar, a membrane-filtered solution of tobramycin was added after sterilisation. To ensure sufficient moisture content, plates with and without tobramycin were used within 24 h. Plates were incubated at 308C and examined after 24, 36 and 48 h. 2.2. Identification and characterization of strains isolated from diseased fish For the strains isolated from diseased fish morphology, growth, biochemical and antigenic characteristics were determined. These characteristics were compared with
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
37
Table 1 Viable counts of the F. columnare strains AJS 1±4 and the five test strains determined at 168C, 258C, 308C and 378C Strain
Origin
CFU/ml determined at 168C
F. columnare F. columnare F. columnare F. columnare F. columnare Strain AJS 1 Strain AJS 2 Strain AJS 3 Strain AJS 4
IC8XHD/88 TG37/87 82-3035P GA325V CDI363061
Catfish Black bullhead Steelhead trout Goldfish Carp Black molly Platy Black molly Black molly
258C 2
210 1102 3104 3103 210 <10 <10 <10 <10
308C 7
210 2108 9106 3106 1106 1107 1107 3106 3106
378C 7
410 5106 9107 9107 1107 1108 9107 1108 2107
<10 <10 <10 <10 <10 3102 3102 2102 3103
those of the F. columnare reference strain NCMB 2248 and/or five F. columnare strains isolated from temperate fish. The latter strains and their sources are given in Table 1. Colonies of the strains isolated from diseased fish were checked on colour, adherence to the agar and rhizoid edges. The suspected colonies were transferred to both Shieh agar and Shieh broth. The following characteristics of the broth cultures were determined: (1) colour and size of Gram-stained bacteria (2) gliding motility of the bacteria (3) growth aspect after 24 h. Growth at different temperatures was tested as follows. The strains isolated from diseased fish along with the five strains isolated from temperate fish were prepared by growing them for 24 h at 308C in Shieh broth. From each culture ten-fold serial dilutions were made in phosphate buffered saline (PBS) and 0.1 ml of each dilution was inoculated on four different Shieh plates. These plates were incubated for 24 h at 168C, 258C, 308C or 378C and viable counts were determined. Further examination of the phenotypic and biochemical traits (Table 2) of the strains isolated from the diseased fish was carried out using the methods described by Bernardet and Grimont (1989) and compared with the reference strain NCMB 2248. Table 2 Biochemical characteristics of the F. columnare strains isolated from tropical fish and reference strain NCMB 2248 Oxidase Catalase production Flexirubin pigment Congo red absorpton NO3 reduction H2S prodution
5/5 5/5 5/5 5/5 5/5 5/5
Hydrolysis of Starch Gelatin Lecithin Tyrosin
0/5 5/5 5/5 0/5
38
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
The strains isolated from the diseased fish were checked for the presence of F. columnare common antigen (Anacker and Ordal, 1959) by slide agglutination using rabbit antiserum against type strain F. columnare NCMB 2248. The organisms were grown for 24 h at 308C in Shieh broth. The bacteria were inactivated by heating at 568C for 30 min and then harvested by centrifugation (3000 G for 15 min). They were washed twice in PBS solution. After washing, the organisms were resuspended in PBS to obtain a final concentration of 108 CFU/ml. The tests were performed by mixing a drop of bacterial antigen suspension with a drop of undiluted serum. An immediately occurring agglutination was regarded positive. Controls were made with saline and serum from nonimmunized rabbits (Toranzo et al., 1987). 2.3. Antibiotic sensitivity of isolated strains The agar dilution procedure was used to determine Minimal inhibitory concentrations (MIC) of chloramphenicol, colistin, erythromycin, furazolidone, kanamycin, lincomycin, nalidixic acid, neomycin, oxytetracycline, streptomycin and sulfamethoxazole. Substances with known potency were dissolved in appropriate solvents and diluted with distilled water to make up stock solutions of 1000 mg/ml. These were filter-sterilized and two-fold dilutions were added to molten Shieh agar to prepare plates containing 64± 0.12 mg/ml active product. The strains isolated from the diseased fish were grown overnight in Shieh broth and inoculated on the antibiotic-containing plates as well as on antibiotic-free control plates by means of a Denley multipoint inoculator (Mast, UK). MIC were recorded after incubation at 308C for 24 h. 2.4. Experimental infections In total 330 black mollies with an average weight and length of 5 g and 6 cm, respectively, were used. These fish belonged to three different batches. All animals used in the studies appeared to be healthy before experiments were carried out. From each batch, ten animals were sacrificed. F. columnare was not isolated from samples of skin, gills and internal organs of these animals and parasite infestations were not noted upon examination of wet mount preparations made from skin and gills. All animals were maintained in glass aquaria containing 50 l well water at 308C and fed daily with a commercial diet. Twenty percent water was exchanged per day. The animals were divided into 30 groups of ten animals each. Experimental infections were carried out with four F. columnare strains (nrs AJS 1, 2, 3 and 4) isolated from diseased fish. Stock suspensions of the strains were stored at ÿ708C. After thawing, the bacteria were grown for 24 h at 308C in Shieh broth. Subsequently, subcultures were made on Shieh broth. This medium was incubated for 20 h at 258C with gentle shaking. In a first experiment, the bacteria were harvested by centrifugation (3000 G, 20 min) and washed twice in PBS. Five groups of ten fish were anaesthetized with a solution containing 1 g of benzocaine (ethyl aminobenzoate) in 10 ml ethanol. Four groups were intramuscularly inoculated with 0.03 ml PBS containing 106 CFU of strains AJS 1, 2, 3 and 4. A fifth group was inoculated with 0.03 ml PBS and used as a control. The experiment was repeated once.
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
39
In a second experiment, four groups of ten fish were immersed in aquaria filled with 150 ml cultured Shieh broth of strains AJS 1, 2, 3 or 4 and 1350 ml well water during 30 min. The final concentration of F. columnare was 106 CFU/ml. A fifth group served as a control and was immersed in an aquarium with 150 ml Shieh broth and 1350 ml well water but without F. columnare. The third experiment was carried out as was the second experiment except for the fact that a piece of the caudal fin was clipped off after anaesthesia. The experiment was repeated once more. All groups were examined clinically 14 days before, until 15 days after inoculation. Moribund or dead fish were autopsied and samples of the skin, gills, spleen and liver were streaked onto Shieh plates supplemented with tobramycin at a concentration of 1 mg/ml and incubated at 308C. Wet mount preparations were made from the gills and lesions of the moribund or dead fish and examined by a phase contrast microscope. 2.5. Statistical analysis The Fischer exact test was used to determine whether there was a significant difference between the first and second time the experiments were carried out. Difference in the observed mortality and morbidity of the used strains was assessed by means of the Yates' corrected chi-square test (computer program STATISTIX 4.1, Analytical software). The numbers of the first and second experiment were added and a significance level of 0.05 was used. 3. Results 3.1. Isolation of F. columnare strains from diseased fish Four strains were isolated. F. columnare AJS 1 was isolated from a black molly with locally bleached and ulcerated skin, kept in an aquarium with an appreciable daily mortality. Strain F. columnare AJS 2 was isolated from a platy in an aquarium where chronic mortality was present and most fish showed discoloration on their back and finrot, especially of the caudal fin. No ulceration was present. Strain F. columnare AJS 3 originated from an outbreak of mouth-dot among black mollies. That is, the skin around the mouth and opercula was blanched. Again, no ulceration could be noted. No significant mortality was recorded. Strain F. columnare AJS 4 was found among black mollies whose skin around the dorsal fin was covered with grey patches. No significant mortality was present among these animals. 3.2. Identification and characterization of F. columnare strains isolated from diseased fish The colonies produced by the four isolated F. columnare strains were yellow-orange, flat, rather small and had rhizoid edges. The colonies were adherent to the agar so that small pieces of agar had to be removed along with the colony to ensure growth of the bacterium when transmitted in broth.
40
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
Table 3 Antibiotic sensitivity of four F. columnare strains isolated from tropical fish Antibiotics
MIC (mg/ml)
Chloramphenicol Colistin Erythromycin Furazolidone Kanamycin Lincomycin Nalidixic acid Neomycin Oxytetracycline Streptomycin Sulfamethoxazole
0.5±1 >128 0.031 1±2 2 0.125 2 10±20 0.5 1±2 64
All strains were Gram-negative, filamentous bacteria exhibiting flexing movements. The size varied from 0.2±0.5 mm by 5±8 mm. When grown in a flask a filamentary deposit was formed on the base and several thread-like structures were fixed at the sides of the flask forming a ring at the upper level of the Shieh medium. Viable counts of F. columnare strains 1±4 and the strains isolated from temperate fish as determined at different temperatures are given in Table 1. Optimal growth of all strains occurred at 25±308C. Strains isolated from temperate fish were able to grow at 168C and strains isolated from tropical fish at 378C. None of the F. columnare strains was able to grow in the presence of 1% NaCl. Biochemical characteristics of F. columnare strains 1±4 as well as from the reference strain F. columnare NCMB 2248 were identical. Results are given in Table 2. Carbohydrate metabolism was found negative in all strains. All strains isolated from the diseased fish were agglutinated by serum against F. columnare common antigen. 3.3. Antibiotic sensitivity tests MIC of antimicrobials were identical for all strains isolated during this study. They are given in Table 3. It can be seen that MIC of colistin and sulfamethoxazole were very high indicatory of intrinsic resistance typical of the species. The MIC values for neomycin were also high. 3.4. Experimental infections No significant difference in mortality and morbidity was seen between the two times the experiment was carried out and was therefore combined (Table 4). Disease signs or mortality were not observed in control animals. In animals intramuscularly inoculated mortality occurred between 8 h and 6 days post infection (p.i.). The mortality among the group infected with strain AJS 3 was significantly higher than among the group inoculated with strain AJS 4 and the control group
1 2 3 4 5
1 2 3 4 5
II
III
Mortality
4/20 4/20 8/20c 2/20 0/20 18/20c 6/20c 2/20 0/20 0/20 18/20c 8/20c 0/20 2/20 0/20
Treatment
AJS 1 IM(106 CFU) AJS 2 IM(106 CFU) AJS 3 IM(106 CFU) AJS 4 IM(106 CFU) Buffered saline IM
AJS 1 106 CFUa AJS 2 106 CFUa AJS 3 106 CFUa AJS 4 106 CFUa Sterile shieh broth
AJS 1 106 CFUb AJS 2 106 CFUb AJS 3 106 CFUb AJS 4 106 CFUb Sterile shieh broth
18/20c 8/20c 2/20 2/20 0/20
18/20c 8/20c 2/20 0/20 0/20
14/20c 6/20c 10/20c 2/20 0/20
Morbidity
NA NA NA NA NA
NA NA NA NA NA
14/20 6/20 10/20 2/20 0/20
18/20 8/20 2/20 2/20 0/20
16/20 8/20 2/20 0/20 0/20
0/20 0/20 0/20 0/20 0/20
b
NA NA NA NA NA
NA NA NA NA NA
2/4 0/4 4/8 2/2 0/0
14/18 8/8 0/0 2/2 0/0
18/18 4/6 2/2 0/0 0/0
2/4 2/4 4/8 0/2 0/0
Gills
16/18 8/8 0/0 2/2 0/0
16/18 6/6 2/2 0/0 0/0
NT NT NT NT NT
Mucus
2/18 2/8 0/0 0/2 0/0
2/18 0/6 0/2 0/0 0/0
0/4 0/4 4/8 0/2 0/0
Spleen
Injection site
With swelling at injection site
Swimming at the water surface
Number of animals/total number with reisolation of F. columnare from
Number of animals/total number
Fish were left intact and immersed in a bath containing 106 CFU/ml. Fish were clipped off a piece of the caudal fin before they were immersed in a bath containing 106 CFU/ml. c Significantly different (p<0.05) from the control group as determined by the Yate's corrected chi-square test. IM: intramuscularly; NA: not applicable; NT: not tested.
1 2 3 4 5
I
a
Group number
Exp. number
Table 4 Results of experimental infection of black mollies (Poecilia sphenops) with F. columnare strains AJS 1±4
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45 41
42
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
Fig. 1. Fish following intramuscular injection with F. columnare strain AJS 3. Extensive necrosis exposing the underlying musculature and the spine is evident.
(p<0.05). No significant differences were found between the mortality of the AJS 1, AJS 2, AJS 4 infected group and the control group. Morbidity was significantly higher in the AJS 1 and AJS 3 infected groups than in the control and AJS 4 infected groups. In animals that developed disease, first clinical signs consisted of a swelling at the injection site which evolved into a blanched spot on the back region covered with thread-like structures, resembling lesions caused by Saprolegnia. In rare cases, gradually a small ulcer developed, progressing into extensive necrosis exposing the underlying musculature and the spine (Fig. 1). Wet mount preparations of the lesions revealed masses of filamentous bacteria exhibiting gliding motility. During the first stages of disease, F. columnare could be reisolated from the lesions but was reduced as the disease progressed. From 50% of the fish that died following injection with F. columnare strain AJS 3, F. columnare could be reisolated from the spleen, especially when the fish died at the beginning of the experiment. Results of contact infection of intact fish and fish with injury of the caudal fin were very similar. Morbidity and mortality in groups infected with strain AJS 1 was significantly higher than in all other groups. In groups infected with strain AJS 2 morbidity and mortality were significantly higher than in control fish and fish infected with strains AJS 3 or AJS 4. Approximately 10 h after infection with strain AJS 1, the fish began swimming near the water surface. Acute mortality occurred 10 h after infection and the gills of the autopsied fish often showed yellow-orange spots. Microscopic examination revealed the formation of a column composed of long, gliding rods in the majority of the cases. As the disease progressed the fish often remained motionless on their sides at the water surface until, after a few seconds of agitation, they died. Mortality stopped 60 h
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
43
after infection. Thirty-six hours after infection, fish showed areas of greyish discoloration. From almost every autopsied fish F. columnare could be reisolated from the gills. F. columnare was reisolated from the spleen of only six animals. 4. Discussion Four F. columnare strains were isolated from black mollies and platies. Biochemical and morphological characteristics of the strains were identical to those of the reference F. columnare strain. To our knowledge, isolation of this bacterium from a tropical freshwater aquarium fish has not been described. However, columnaris disease seems to be a problem among many freshwater tropical aquarium fishes, probably because of the high temperature of aquaria (25±308C), which favours growth of many F. columnare strains (Post, 1987). Nevertheless, there is scant information available on its actual occurrence, pathogenesis or severity in tropical aquarium fish. In the majority of the studies dealing with the disease, coldwater and temperate fish species have been used such as salmon, trout (Fish and Rucker, 1943; Holt et al., 1975), eel (Chen et al., 1982) and carp (Bootsma and Clerx, 1976). Our F. columnare isolates from tropical fish species quite distinctively differed from temperate isolates, when tested for growth at different temperatures. It was noted that strains isolated from tropical fish were able to grow at 378C but not at 168C. The opposite was found for strains from temperate fish. Results of experimental infections indicate that strain AJS 1 was highly virulent whereas strain AJS 4 was of low virulence. In the aquaria from which these strains were originally isolated, mortality was high for strain AJS 1 and low for strain AJS 4. Differences in virulence have also been described among F. columnare strains isolated from coldwater fish (Pacha and Ordal, 1963) and temperate fish (Wakabayashi et al., 1970). Further studies are necessary to determine factors involved in virulence. After contact infection with strain AJS 1, acute mortality was recorded from 10 h after infection. In the aquarium from which strain AJS 1 was originally isolated an appreciable daily mortality was recorded. In experimentally infected fish, areas of greyish discoloration were noted. The black molly from which the strain AJS 1 was isolated also displayed bleached skin. No ulceration could be noted among the black mollies from which strain AJS 3 was originally isolated. However, in experimentally infected fish, ulcers were recorded after intramuscular injection of strain AJS 3. This was not the case after contact infection. Contact infection represents a more natural way of infection whereas intramuscular injection bypasses the natural defense mechanisms such as skin and mucus. Morbidity and mortality rates were significantly lower in fish immersed in a bath with strain AJS 3 than in fish immersed in a bath with strain AJS 1. This was not the case when fish were inoculated intramuscularly, indicating that the capacity to pass the mucus and skin barrier is lower for strain AJS 3 than AJS 1. Pacha and Ordal (1970) reported a similar finding. They stated that contact with highly virulent strains induced infection and disease contraction more than intramuscular injection, but injection of low virulent bacteria more readily induced infection and disease contraction than did contact.
44
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
F. columnare was isolated from the internal organs of four and two fish that died after immersing them in an infective solution of strains AJS 1 and AJS 2, respectively. This indicates that these strains can enter the blood stream possibly through the gill lesions and cause septicaemia. Septicaemia caused by F. columnare has indeed been reported (Hawke and Thune, 1992; Koski et al., 1993). After intramuscular inoculation, F. columnare could easily be reisolated from the lesions during the first stages of the disease, but as it progressed, a decrease in the recovery of F. columnare from the ulcera was noted. This could be due to overgrowth by other bacteria. Indeed, F. columnare is easily overgrown by commensal bacteria (Decostere et al., 1997), the latter taking advantage of the damage caused by the former. This has been reported for Aeromonas hydrophila which invades the host after it has been compromised, making the condition worse and being responsible for ultimate death (Roberts, 1993). Chowdhury and Wakabayashi (1989) also noted that the progression of infection of F. columnare may be mediated by other bacteria like Aeromonas hydrophila and Citrobacter freundii. Acknowledgements We would like to thank J.-F. Bernardet for providing the antiserum against the reference strain of F. columnare and several strains used in the test for growth at different temperatures. The Fund for Scientific Research ± Flanders (FWO) is acknowledged for providing a grant to Annemie Decostere. References Anacker, R.L., Ordal, E.J., 1959. Studies on the myxobacterium Chondrococcus colmnaris. I. Serological typing. J. Bacteriol. 78, 25±32. Anderson, J.I.W., Conroy, D.A., 1969. The pathogenic myxobacteria with special reference to fish diseases. J. Appl. Bact. 32, 30±39. Austin, B., Austin, D.A., 1987. Bacterial Fish Pathogens. Ellis Horwood, West Sussex, England, pp. 225±247. Bassleer, G., 1983. Wegwijs In Visziekten. B.V.W.J. Thieme & Cie, Zutphen, pp. 42±43. Bernardet, J.F., Grimont, P.A.D., 1989. Deoxyribonucleic acid relatedness and phenotypic characterization of Flexibacter columnaris sp. nov., nom. rev., Flexibacter psychrofilus sp. nov., nom. rev., and Flexibacter maritimus. Int. J. Syst. Bacteriol., 346±354. Bernardet, J.F., Segers, P., Vancanneyt, M., Berthe, M., Kersters, K., Vandamme, P., 1996. Cutting a Gordian knot: Emended classification and description of the genus Flavobacterium, emended description of the family Flavobacteriaceae, and proposal of Flavobacterium hydatis nom. nov. (basonym, Cytophaga aquatilis Strohl and Tait 1978). Int. J. Syst. Bacteriol. 46, 128±148. Bootsma, R., Clerx, J.P.M., 1976. Columnaris disease of cultured carp, Cyprinus carpio L. : Characteristics of the causative agent. Aquaculture 7, 371±384. Butcher, R.L., 1992. Manual of Ornamental Fish. British Small Animal Veterinary Association, Gloucestershire, pp. 92±93. Chen, C.-R.L., Chung, Y.-Y., Kuo, G.-H., 1982. Studies on the pathogenicity of Flexibacter columnaris. I. Effect of dissolved oxygen and ammonia on the pathogenicity of Flexibacter columnaris to eel (Anguilla japonica). CAPD Fisheries Series No. 8, Reports on Fish Disease Research, vol. 4, pp. 57±61. Chowdhury, B.R., Wakabayashi, H., 1989. Effects of competitive bacteria on the survival and infectivity of Flexibacter columnaris. Fish Pathol. 24, 9±15.
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
45
Decostere, A., Haesebrouck, F., Devriese, L.A., 1997. Shieh medium supplemented with tobramycin for selective isolation of Flavobacterium columnare (Flexibacter columnaris) from diseased fish. J. Clin. Microb. 35, 322±324. Fish, F.F., Rucker, R.R., 1943. Columnaris as a disease of coldwater fishes. Trans. Am. Fish. Soc. 73, 32±36. Hawke, J.P., Thune, R.L., 1992. Systemic isolation and antimicrobial susceptibility of Cytophaga columnaris from commercially reared channel catfish. J. Aquat. Animal Health 4, 109±113. Holt, R.A., Sanders, J.E., Zinn, J.L., Fryer, J.L., Pilcher, K.S., 1975. Relation of water temperature to Flexibacter columnaris infection in steelhead trout (Salmo gairdneri), coho (Oncorhynchus kisutch) and chinook (O. Tschowytscha) salmon. J. Fish. Res. Board Can. 32, 1553±1555. Koski, P., HiervelaÈ-Koski, V., Bernardet, J.F., 1993. Flexibacter columnaris infection in arctic char (Salvelinus alpinus L.). First isolation in Finland. Bull. Eur. Ass. Fish Pathol. 31, 66±69. Pacha, R.E., Ordal, E.J., 1963. Epidemiology of columnaris disease in salmon. Bacteriological Proceedings 63, 3±4. Pacha, R.E., Ordal, E.J., 1970. Myxobacterial diseases of salmonids. In: Snieszko, S.F. (Ed.), A Symp. on Diseases of Fishes and Shellfishes, American Fisheries Society, Special Publication No. 5, Washington DC, pp. 243±257. Post, G., 1987. Textbook of Fish Health. T.F.H. Publications, Ascot, Berkshire, pp. 51±56. Roberts, R.J., 1993. Introduction. In: Inglis, V., Roberts, R.J., Bromage, R.N. (Eds.), Bacterial Diseases of Fish. Blackwell Scientific Publications, Oxford, pp. I±IV. Shieh, H.S., 1980. Studies on the nutrition of a fish pathogen, Flexibacter columnaris. Microbios Lett. 13, 129± 133. Tomey, W.A., 1985. Ziekten bij aquariumvissen. Zuid Boekprodukties B.V., Best, Nederland, pp. 53±54. Toranzo, A.E., Baya, A.M., Roberson, B.S., Baya, J.L., Grimes, D.J., Hetrick, F.M., 1987. Specificity of slide agglutination test for detecting bacterial fish pathogens. Aquaculture 61, 81±97. Wakabayashi, H., 1991. Effect of environmental conditions on the infectivity of Flexibacter columnaris to fish. J. Fish Diseases 14, 279±290. Wakabayashi, H., Egusa, S., 1972. Preliminary experiments on environmental factors influencing the prevalence of columnaris disease. Fish Pathol. 7, 58±63. Wakabayashi, H., Kira, K., Egusa, S., 1970. Studies on columnaris disease of pond-cultured eel ± I. Characteristics and pathogenicity of Chondrococcus columnaris isolated from pond-cultured eels. Bull. Jpn. Soc. Scientific Fisheries 36, 147±155.
Characterization of four Flavobacterium columnare (Flexibacter columnaris) strains isolated from tropical fish A. Decostere*, F. Haesebrouck, L.A. Devriese Laboratory of Veterinary Bacteriology and Mycology, University of Gent, Salisburylaan 133, B 9820 Merelbeke, Belgium Received 15 August 1997; accepted 17 March 1998
Abstract Four Flavobacterium columnare strains (AJS 1±4) were isolated from black mollies (Poecilia sphenops) and platies (Xiphophorus maculatus), showing white spots on the back, head and skin ulcers. The isolates developed characteristic rhizoid yellow pigmented colonies on Shieh agar and typical growth in Shieh broth. They were Gram-negative, filamentous bacteria exhibiting flexing movements. When compared to F. columnare strains isolated from temperate fish, it was noted that the four strains originating from tropical aquarium fish are more capable of growing at higher temperatures, the opposite being true for the strains isolated from temperate fish. Biochemical characterization and agglutination tests proved that the isolated strains could be classified as F. columnare. Low minimal inhibitory concentration (MIC) values were found for chloramphenicol, erythromycin, furazolidone, kanamycin, lincomycin, nalidixic acid, oxytetracycline and streptomycin. MIC values were high for colistin, sulfamethoxazole and neomycin. Pathogenicity studies were performed on black mollies. When these animals were submersed in an infective solution of the F. columnare strains, a marked difference in virulence was noted among the four isolated strains, strain AJS 1 being the most virulent one and strain AJS 4 being of low virulence. # 1998 Elsevier Science B.V. All rights reserved. Keywords: Flavobacterium columnare; Fish; Diagnosis-bacteria; Virulence
* Corresponding author. Tel.: +32 9 2647431; fax: +32 9 2647494; e-mail: [email protected] 0378-1135/98/$19.00 # 1998 Elsevier Science B.V. All rights reserved. PII S 0 3 7 8 - 1 1 3 5 ( 9 8 ) 0 0 1 9 6 - 5
36
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
1. Introduction Flavobacterium columnare has been recognized as a worldwide-occurring pathogen of numerous freshwater fish species (Anderson and Conroy, 1969; Austin and Austin, 1987). It was formerly called Flexibacter columnaris, but in 1996 it was transferred to the genus Flavobacterium (Bernardet et al., 1996). In temperate fish, columnaris disease is recognized by the appearance of greyish white or yellow areas of erosion, usually surrounded by a reddish hyperemic zone, on the body surfaces or the gills of fish. When these lesions occur around the dorsal fin, they are called saddle-back lesions, which are typical for columnaris disease of coldwater fish (Austin and Austin, 1987). Outbreaks of columnaris disease are rarely spontaneous, but are influenced by a combination of environmental (temperature) and other factors stressful to the host, such as high stocking density, high levels of ammonia and organic load (Wakabayashi and Egusa, 1972; Chen et al., 1982; Wakabayashi, 1991). F. columnare is mentioned in most books of tropical aquarium fish diseases as the cause of cotton-wool disease. Post (1987) reported fish exhibiting greyish-white discoloration especially on the head and around the mouth. These may evolve into whitish filaments emerging from the lesions, hence the name mouth fungus. Although F. columnare is described as being an important pathogen (Bassleer, 1983; Tomey, 1985; Butcher, 1992), as far as we know, it has never been actually isolated from tropical aquarium fish, nor has its significance and pathogenicity been studied in any way. In the present paper the isolation and characterization of four F. columnare strains obtained from tropical fish with typical lesions is described. The pathogenicity of these strains was determined by immersion and injection experiments of intact black mollies. Additionally, the role of mechanical injury in transmission of F. columnare was determined. 2. Materials and methods 2.1. Isolation of F. columnare strains from diseased fish Samples of skin mucus were collected from 50 moribund fish originating from 12 different aquaria. The fish species included black mollies (Poecilia sphenops), platies (Xiphophorus maculatus), guppies (Poecilia reticulata), tetras (Cheirodon axelrod). These samples were streaked on Trypticase Soy Agar (Gibco, Paisley, Scotland), Shieh agar (Shieh, 1980) without any additives and Shieh agar supplemented with tobramycine at a concentration of 1 mg/ml (Decostere et al., 1997). For the supplemented Shieh agar, a membrane-filtered solution of tobramycin was added after sterilisation. To ensure sufficient moisture content, plates with and without tobramycin were used within 24 h. Plates were incubated at 308C and examined after 24, 36 and 48 h. 2.2. Identification and characterization of strains isolated from diseased fish For the strains isolated from diseased fish morphology, growth, biochemical and antigenic characteristics were determined. These characteristics were compared with
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
37
Table 1 Viable counts of the F. columnare strains AJS 1±4 and the five test strains determined at 168C, 258C, 308C and 378C Strain
Origin
CFU/ml determined at 168C
F. columnare F. columnare F. columnare F. columnare F. columnare Strain AJS 1 Strain AJS 2 Strain AJS 3 Strain AJS 4
IC8XHD/88 TG37/87 82-3035P GA325V CDI363061
Catfish Black bullhead Steelhead trout Goldfish Carp Black molly Platy Black molly Black molly
258C 2
210 1102 3104 3103 210 <10 <10 <10 <10
308C 7
210 2108 9106 3106 1106 1107 1107 3106 3106
378C 7
410 5106 9107 9107 1107 1108 9107 1108 2107
<10 <10 <10 <10 <10 3102 3102 2102 3103
those of the F. columnare reference strain NCMB 2248 and/or five F. columnare strains isolated from temperate fish. The latter strains and their sources are given in Table 1. Colonies of the strains isolated from diseased fish were checked on colour, adherence to the agar and rhizoid edges. The suspected colonies were transferred to both Shieh agar and Shieh broth. The following characteristics of the broth cultures were determined: (1) colour and size of Gram-stained bacteria (2) gliding motility of the bacteria (3) growth aspect after 24 h. Growth at different temperatures was tested as follows. The strains isolated from diseased fish along with the five strains isolated from temperate fish were prepared by growing them for 24 h at 308C in Shieh broth. From each culture ten-fold serial dilutions were made in phosphate buffered saline (PBS) and 0.1 ml of each dilution was inoculated on four different Shieh plates. These plates were incubated for 24 h at 168C, 258C, 308C or 378C and viable counts were determined. Further examination of the phenotypic and biochemical traits (Table 2) of the strains isolated from the diseased fish was carried out using the methods described by Bernardet and Grimont (1989) and compared with the reference strain NCMB 2248. Table 2 Biochemical characteristics of the F. columnare strains isolated from tropical fish and reference strain NCMB 2248 Oxidase Catalase production Flexirubin pigment Congo red absorpton NO3 reduction H2S prodution
5/5 5/5 5/5 5/5 5/5 5/5
Hydrolysis of Starch Gelatin Lecithin Tyrosin
0/5 5/5 5/5 0/5
38
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
The strains isolated from the diseased fish were checked for the presence of F. columnare common antigen (Anacker and Ordal, 1959) by slide agglutination using rabbit antiserum against type strain F. columnare NCMB 2248. The organisms were grown for 24 h at 308C in Shieh broth. The bacteria were inactivated by heating at 568C for 30 min and then harvested by centrifugation (3000 G for 15 min). They were washed twice in PBS solution. After washing, the organisms were resuspended in PBS to obtain a final concentration of 108 CFU/ml. The tests were performed by mixing a drop of bacterial antigen suspension with a drop of undiluted serum. An immediately occurring agglutination was regarded positive. Controls were made with saline and serum from nonimmunized rabbits (Toranzo et al., 1987). 2.3. Antibiotic sensitivity of isolated strains The agar dilution procedure was used to determine Minimal inhibitory concentrations (MIC) of chloramphenicol, colistin, erythromycin, furazolidone, kanamycin, lincomycin, nalidixic acid, neomycin, oxytetracycline, streptomycin and sulfamethoxazole. Substances with known potency were dissolved in appropriate solvents and diluted with distilled water to make up stock solutions of 1000 mg/ml. These were filter-sterilized and two-fold dilutions were added to molten Shieh agar to prepare plates containing 64± 0.12 mg/ml active product. The strains isolated from the diseased fish were grown overnight in Shieh broth and inoculated on the antibiotic-containing plates as well as on antibiotic-free control plates by means of a Denley multipoint inoculator (Mast, UK). MIC were recorded after incubation at 308C for 24 h. 2.4. Experimental infections In total 330 black mollies with an average weight and length of 5 g and 6 cm, respectively, were used. These fish belonged to three different batches. All animals used in the studies appeared to be healthy before experiments were carried out. From each batch, ten animals were sacrificed. F. columnare was not isolated from samples of skin, gills and internal organs of these animals and parasite infestations were not noted upon examination of wet mount preparations made from skin and gills. All animals were maintained in glass aquaria containing 50 l well water at 308C and fed daily with a commercial diet. Twenty percent water was exchanged per day. The animals were divided into 30 groups of ten animals each. Experimental infections were carried out with four F. columnare strains (nrs AJS 1, 2, 3 and 4) isolated from diseased fish. Stock suspensions of the strains were stored at ÿ708C. After thawing, the bacteria were grown for 24 h at 308C in Shieh broth. Subsequently, subcultures were made on Shieh broth. This medium was incubated for 20 h at 258C with gentle shaking. In a first experiment, the bacteria were harvested by centrifugation (3000 G, 20 min) and washed twice in PBS. Five groups of ten fish were anaesthetized with a solution containing 1 g of benzocaine (ethyl aminobenzoate) in 10 ml ethanol. Four groups were intramuscularly inoculated with 0.03 ml PBS containing 106 CFU of strains AJS 1, 2, 3 and 4. A fifth group was inoculated with 0.03 ml PBS and used as a control. The experiment was repeated once.
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
39
In a second experiment, four groups of ten fish were immersed in aquaria filled with 150 ml cultured Shieh broth of strains AJS 1, 2, 3 or 4 and 1350 ml well water during 30 min. The final concentration of F. columnare was 106 CFU/ml. A fifth group served as a control and was immersed in an aquarium with 150 ml Shieh broth and 1350 ml well water but without F. columnare. The third experiment was carried out as was the second experiment except for the fact that a piece of the caudal fin was clipped off after anaesthesia. The experiment was repeated once more. All groups were examined clinically 14 days before, until 15 days after inoculation. Moribund or dead fish were autopsied and samples of the skin, gills, spleen and liver were streaked onto Shieh plates supplemented with tobramycin at a concentration of 1 mg/ml and incubated at 308C. Wet mount preparations were made from the gills and lesions of the moribund or dead fish and examined by a phase contrast microscope. 2.5. Statistical analysis The Fischer exact test was used to determine whether there was a significant difference between the first and second time the experiments were carried out. Difference in the observed mortality and morbidity of the used strains was assessed by means of the Yates' corrected chi-square test (computer program STATISTIX 4.1, Analytical software). The numbers of the first and second experiment were added and a significance level of 0.05 was used. 3. Results 3.1. Isolation of F. columnare strains from diseased fish Four strains were isolated. F. columnare AJS 1 was isolated from a black molly with locally bleached and ulcerated skin, kept in an aquarium with an appreciable daily mortality. Strain F. columnare AJS 2 was isolated from a platy in an aquarium where chronic mortality was present and most fish showed discoloration on their back and finrot, especially of the caudal fin. No ulceration was present. Strain F. columnare AJS 3 originated from an outbreak of mouth-dot among black mollies. That is, the skin around the mouth and opercula was blanched. Again, no ulceration could be noted. No significant mortality was recorded. Strain F. columnare AJS 4 was found among black mollies whose skin around the dorsal fin was covered with grey patches. No significant mortality was present among these animals. 3.2. Identification and characterization of F. columnare strains isolated from diseased fish The colonies produced by the four isolated F. columnare strains were yellow-orange, flat, rather small and had rhizoid edges. The colonies were adherent to the agar so that small pieces of agar had to be removed along with the colony to ensure growth of the bacterium when transmitted in broth.
40
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
Table 3 Antibiotic sensitivity of four F. columnare strains isolated from tropical fish Antibiotics
MIC (mg/ml)
Chloramphenicol Colistin Erythromycin Furazolidone Kanamycin Lincomycin Nalidixic acid Neomycin Oxytetracycline Streptomycin Sulfamethoxazole
0.5±1 >128 0.031 1±2 2 0.125 2 10±20 0.5 1±2 64
All strains were Gram-negative, filamentous bacteria exhibiting flexing movements. The size varied from 0.2±0.5 mm by 5±8 mm. When grown in a flask a filamentary deposit was formed on the base and several thread-like structures were fixed at the sides of the flask forming a ring at the upper level of the Shieh medium. Viable counts of F. columnare strains 1±4 and the strains isolated from temperate fish as determined at different temperatures are given in Table 1. Optimal growth of all strains occurred at 25±308C. Strains isolated from temperate fish were able to grow at 168C and strains isolated from tropical fish at 378C. None of the F. columnare strains was able to grow in the presence of 1% NaCl. Biochemical characteristics of F. columnare strains 1±4 as well as from the reference strain F. columnare NCMB 2248 were identical. Results are given in Table 2. Carbohydrate metabolism was found negative in all strains. All strains isolated from the diseased fish were agglutinated by serum against F. columnare common antigen. 3.3. Antibiotic sensitivity tests MIC of antimicrobials were identical for all strains isolated during this study. They are given in Table 3. It can be seen that MIC of colistin and sulfamethoxazole were very high indicatory of intrinsic resistance typical of the species. The MIC values for neomycin were also high. 3.4. Experimental infections No significant difference in mortality and morbidity was seen between the two times the experiment was carried out and was therefore combined (Table 4). Disease signs or mortality were not observed in control animals. In animals intramuscularly inoculated mortality occurred between 8 h and 6 days post infection (p.i.). The mortality among the group infected with strain AJS 3 was significantly higher than among the group inoculated with strain AJS 4 and the control group
1 2 3 4 5
1 2 3 4 5
II
III
Mortality
4/20 4/20 8/20c 2/20 0/20 18/20c 6/20c 2/20 0/20 0/20 18/20c 8/20c 0/20 2/20 0/20
Treatment
AJS 1 IM(106 CFU) AJS 2 IM(106 CFU) AJS 3 IM(106 CFU) AJS 4 IM(106 CFU) Buffered saline IM
AJS 1 106 CFUa AJS 2 106 CFUa AJS 3 106 CFUa AJS 4 106 CFUa Sterile shieh broth
AJS 1 106 CFUb AJS 2 106 CFUb AJS 3 106 CFUb AJS 4 106 CFUb Sterile shieh broth
18/20c 8/20c 2/20 2/20 0/20
18/20c 8/20c 2/20 0/20 0/20
14/20c 6/20c 10/20c 2/20 0/20
Morbidity
NA NA NA NA NA
NA NA NA NA NA
14/20 6/20 10/20 2/20 0/20
18/20 8/20 2/20 2/20 0/20
16/20 8/20 2/20 0/20 0/20
0/20 0/20 0/20 0/20 0/20
b
NA NA NA NA NA
NA NA NA NA NA
2/4 0/4 4/8 2/2 0/0
14/18 8/8 0/0 2/2 0/0
18/18 4/6 2/2 0/0 0/0
2/4 2/4 4/8 0/2 0/0
Gills
16/18 8/8 0/0 2/2 0/0
16/18 6/6 2/2 0/0 0/0
NT NT NT NT NT
Mucus
2/18 2/8 0/0 0/2 0/0
2/18 0/6 0/2 0/0 0/0
0/4 0/4 4/8 0/2 0/0
Spleen
Injection site
With swelling at injection site
Swimming at the water surface
Number of animals/total number with reisolation of F. columnare from
Number of animals/total number
Fish were left intact and immersed in a bath containing 106 CFU/ml. Fish were clipped off a piece of the caudal fin before they were immersed in a bath containing 106 CFU/ml. c Significantly different (p<0.05) from the control group as determined by the Yate's corrected chi-square test. IM: intramuscularly; NA: not applicable; NT: not tested.
1 2 3 4 5
I
a
Group number
Exp. number
Table 4 Results of experimental infection of black mollies (Poecilia sphenops) with F. columnare strains AJS 1±4
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45 41
42
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
Fig. 1. Fish following intramuscular injection with F. columnare strain AJS 3. Extensive necrosis exposing the underlying musculature and the spine is evident.
(p<0.05). No significant differences were found between the mortality of the AJS 1, AJS 2, AJS 4 infected group and the control group. Morbidity was significantly higher in the AJS 1 and AJS 3 infected groups than in the control and AJS 4 infected groups. In animals that developed disease, first clinical signs consisted of a swelling at the injection site which evolved into a blanched spot on the back region covered with thread-like structures, resembling lesions caused by Saprolegnia. In rare cases, gradually a small ulcer developed, progressing into extensive necrosis exposing the underlying musculature and the spine (Fig. 1). Wet mount preparations of the lesions revealed masses of filamentous bacteria exhibiting gliding motility. During the first stages of disease, F. columnare could be reisolated from the lesions but was reduced as the disease progressed. From 50% of the fish that died following injection with F. columnare strain AJS 3, F. columnare could be reisolated from the spleen, especially when the fish died at the beginning of the experiment. Results of contact infection of intact fish and fish with injury of the caudal fin were very similar. Morbidity and mortality in groups infected with strain AJS 1 was significantly higher than in all other groups. In groups infected with strain AJS 2 morbidity and mortality were significantly higher than in control fish and fish infected with strains AJS 3 or AJS 4. Approximately 10 h after infection with strain AJS 1, the fish began swimming near the water surface. Acute mortality occurred 10 h after infection and the gills of the autopsied fish often showed yellow-orange spots. Microscopic examination revealed the formation of a column composed of long, gliding rods in the majority of the cases. As the disease progressed the fish often remained motionless on their sides at the water surface until, after a few seconds of agitation, they died. Mortality stopped 60 h
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
43
after infection. Thirty-six hours after infection, fish showed areas of greyish discoloration. From almost every autopsied fish F. columnare could be reisolated from the gills. F. columnare was reisolated from the spleen of only six animals. 4. Discussion Four F. columnare strains were isolated from black mollies and platies. Biochemical and morphological characteristics of the strains were identical to those of the reference F. columnare strain. To our knowledge, isolation of this bacterium from a tropical freshwater aquarium fish has not been described. However, columnaris disease seems to be a problem among many freshwater tropical aquarium fishes, probably because of the high temperature of aquaria (25±308C), which favours growth of many F. columnare strains (Post, 1987). Nevertheless, there is scant information available on its actual occurrence, pathogenesis or severity in tropical aquarium fish. In the majority of the studies dealing with the disease, coldwater and temperate fish species have been used such as salmon, trout (Fish and Rucker, 1943; Holt et al., 1975), eel (Chen et al., 1982) and carp (Bootsma and Clerx, 1976). Our F. columnare isolates from tropical fish species quite distinctively differed from temperate isolates, when tested for growth at different temperatures. It was noted that strains isolated from tropical fish were able to grow at 378C but not at 168C. The opposite was found for strains from temperate fish. Results of experimental infections indicate that strain AJS 1 was highly virulent whereas strain AJS 4 was of low virulence. In the aquaria from which these strains were originally isolated, mortality was high for strain AJS 1 and low for strain AJS 4. Differences in virulence have also been described among F. columnare strains isolated from coldwater fish (Pacha and Ordal, 1963) and temperate fish (Wakabayashi et al., 1970). Further studies are necessary to determine factors involved in virulence. After contact infection with strain AJS 1, acute mortality was recorded from 10 h after infection. In the aquarium from which strain AJS 1 was originally isolated an appreciable daily mortality was recorded. In experimentally infected fish, areas of greyish discoloration were noted. The black molly from which the strain AJS 1 was isolated also displayed bleached skin. No ulceration could be noted among the black mollies from which strain AJS 3 was originally isolated. However, in experimentally infected fish, ulcers were recorded after intramuscular injection of strain AJS 3. This was not the case after contact infection. Contact infection represents a more natural way of infection whereas intramuscular injection bypasses the natural defense mechanisms such as skin and mucus. Morbidity and mortality rates were significantly lower in fish immersed in a bath with strain AJS 3 than in fish immersed in a bath with strain AJS 1. This was not the case when fish were inoculated intramuscularly, indicating that the capacity to pass the mucus and skin barrier is lower for strain AJS 3 than AJS 1. Pacha and Ordal (1970) reported a similar finding. They stated that contact with highly virulent strains induced infection and disease contraction more than intramuscular injection, but injection of low virulent bacteria more readily induced infection and disease contraction than did contact.
44
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
F. columnare was isolated from the internal organs of four and two fish that died after immersing them in an infective solution of strains AJS 1 and AJS 2, respectively. This indicates that these strains can enter the blood stream possibly through the gill lesions and cause septicaemia. Septicaemia caused by F. columnare has indeed been reported (Hawke and Thune, 1992; Koski et al., 1993). After intramuscular inoculation, F. columnare could easily be reisolated from the lesions during the first stages of the disease, but as it progressed, a decrease in the recovery of F. columnare from the ulcera was noted. This could be due to overgrowth by other bacteria. Indeed, F. columnare is easily overgrown by commensal bacteria (Decostere et al., 1997), the latter taking advantage of the damage caused by the former. This has been reported for Aeromonas hydrophila which invades the host after it has been compromised, making the condition worse and being responsible for ultimate death (Roberts, 1993). Chowdhury and Wakabayashi (1989) also noted that the progression of infection of F. columnare may be mediated by other bacteria like Aeromonas hydrophila and Citrobacter freundii. Acknowledgements We would like to thank J.-F. Bernardet for providing the antiserum against the reference strain of F. columnare and several strains used in the test for growth at different temperatures. The Fund for Scientific Research ± Flanders (FWO) is acknowledged for providing a grant to Annemie Decostere. References Anacker, R.L., Ordal, E.J., 1959. Studies on the myxobacterium Chondrococcus colmnaris. I. Serological typing. J. Bacteriol. 78, 25±32. Anderson, J.I.W., Conroy, D.A., 1969. The pathogenic myxobacteria with special reference to fish diseases. J. Appl. Bact. 32, 30±39. Austin, B., Austin, D.A., 1987. Bacterial Fish Pathogens. Ellis Horwood, West Sussex, England, pp. 225±247. Bassleer, G., 1983. Wegwijs In Visziekten. B.V.W.J. Thieme & Cie, Zutphen, pp. 42±43. Bernardet, J.F., Grimont, P.A.D., 1989. Deoxyribonucleic acid relatedness and phenotypic characterization of Flexibacter columnaris sp. nov., nom. rev., Flexibacter psychrofilus sp. nov., nom. rev., and Flexibacter maritimus. Int. J. Syst. Bacteriol., 346±354. Bernardet, J.F., Segers, P., Vancanneyt, M., Berthe, M., Kersters, K., Vandamme, P., 1996. Cutting a Gordian knot: Emended classification and description of the genus Flavobacterium, emended description of the family Flavobacteriaceae, and proposal of Flavobacterium hydatis nom. nov. (basonym, Cytophaga aquatilis Strohl and Tait 1978). Int. J. Syst. Bacteriol. 46, 128±148. Bootsma, R., Clerx, J.P.M., 1976. Columnaris disease of cultured carp, Cyprinus carpio L. : Characteristics of the causative agent. Aquaculture 7, 371±384. Butcher, R.L., 1992. Manual of Ornamental Fish. British Small Animal Veterinary Association, Gloucestershire, pp. 92±93. Chen, C.-R.L., Chung, Y.-Y., Kuo, G.-H., 1982. Studies on the pathogenicity of Flexibacter columnaris. I. Effect of dissolved oxygen and ammonia on the pathogenicity of Flexibacter columnaris to eel (Anguilla japonica). CAPD Fisheries Series No. 8, Reports on Fish Disease Research, vol. 4, pp. 57±61. Chowdhury, B.R., Wakabayashi, H., 1989. Effects of competitive bacteria on the survival and infectivity of Flexibacter columnaris. Fish Pathol. 24, 9±15.
A. Decostere et al. / Veterinary Microbiology 62 (1998) 35±45
45
Decostere, A., Haesebrouck, F., Devriese, L.A., 1997. Shieh medium supplemented with tobramycin for selective isolation of Flavobacterium columnare (Flexibacter columnaris) from diseased fish. J. Clin. Microb. 35, 322±324. Fish, F.F., Rucker, R.R., 1943. Columnaris as a disease of coldwater fishes. Trans. Am. Fish. Soc. 73, 32±36. Hawke, J.P., Thune, R.L., 1992. Systemic isolation and antimicrobial susceptibility of Cytophaga columnaris from commercially reared channel catfish. J. Aquat. Animal Health 4, 109±113. Holt, R.A., Sanders, J.E., Zinn, J.L., Fryer, J.L., Pilcher, K.S., 1975. Relation of water temperature to Flexibacter columnaris infection in steelhead trout (Salmo gairdneri), coho (Oncorhynchus kisutch) and chinook (O. Tschowytscha) salmon. J. Fish. Res. Board Can. 32, 1553±1555. Koski, P., HiervelaÈ-Koski, V., Bernardet, J.F., 1993. Flexibacter columnaris infection in arctic char (Salvelinus alpinus L.). First isolation in Finland. Bull. Eur. Ass. Fish Pathol. 31, 66±69. Pacha, R.E., Ordal, E.J., 1963. Epidemiology of columnaris disease in salmon. Bacteriological Proceedings 63, 3±4. Pacha, R.E., Ordal, E.J., 1970. Myxobacterial diseases of salmonids. In: Snieszko, S.F. (Ed.), A Symp. on Diseases of Fishes and Shellfishes, American Fisheries Society, Special Publication No. 5, Washington DC, pp. 243±257. Post, G., 1987. Textbook of Fish Health. T.F.H. Publications, Ascot, Berkshire, pp. 51±56. Roberts, R.J., 1993. Introduction. In: Inglis, V., Roberts, R.J., Bromage, R.N. (Eds.), Bacterial Diseases of Fish. Blackwell Scientific Publications, Oxford, pp. I±IV. Shieh, H.S., 1980. Studies on the nutrition of a fish pathogen, Flexibacter columnaris. Microbios Lett. 13, 129± 133. Tomey, W.A., 1985. Ziekten bij aquariumvissen. Zuid Boekprodukties B.V., Best, Nederland, pp. 53±54. Toranzo, A.E., Baya, A.M., Roberson, B.S., Baya, J.L., Grimes, D.J., Hetrick, F.M., 1987. Specificity of slide agglutination test for detecting bacterial fish pathogens. Aquaculture 61, 81±97. Wakabayashi, H., 1991. Effect of environmental conditions on the infectivity of Flexibacter columnaris to fish. J. Fish Diseases 14, 279±290. Wakabayashi, H., Egusa, S., 1972. Preliminary experiments on environmental factors influencing the prevalence of columnaris disease. Fish Pathol. 7, 58±63. Wakabayashi, H., Kira, K., Egusa, S., 1970. Studies on columnaris disease of pond-cultured eel ± I. Characteristics and pathogenicity of Chondrococcus columnaris isolated from pond-cultured eels. Bull. Jpn. Soc. Scientific Fisheries 36, 147±155.