Enterococcus seriolicida Is a Junior Synonym of Lactococcus garvieae , a Causative Agent of Septicemia and Meningoencephalitis in Fish

CURRENT MICROBIOLOGY Vol. 32 (1996), pp. 85–88

An International Journal

R Springer-Verlag New York Inc. 1996

Enterococcus seriolicida Is a Junior Synonym of Lactococcus garvieae, a Causative Agent of Septicemia and Meningoencephalitis in Fish Avi Eldar,1,2 Claudio Ghittino,3 Liliana Asanta,2 Elena Bozzetta,3 Mariella Goria,3 Marino Prearo,3 Herve Bercovier1 1Department

of Clinical Microbiology, The Hebrew University–Hadassah Medical School, POB 12272, Jerusalem, 91120, Israel of Poultry and Fish Diseases, The Kimron Veterinary Institute, POB 12, Beit Dagan, Israel 3Experimental Institute for Zooprophylaxis, via Bologna 148, 10154, Torino, Italy 2Department

Abstract. The reference strains of Enterococcus seriolicida (ATCC 49156T) (T 5 type strain) and of Lactococcus garvieae (ATCC 43921T) and 30 field strains of Gram-positive cocci isolated from diseased rainbow trout in Italy were found to be phenotypically (API 20 STREPT and API 50 CH) and genetically (DNA-DNA hybridization) similar. The high DNA-DNA homologies (70–100%) and the low DTm(e) (less than 1.1°C) among these strains showed that Enterococcus seriolicida and Lactococcus garvieae are synonyms, describing a single bacterial species. E. seriolicida strains should be classified as L. garvieae, which must be considered as a major pathogen of freshwater and salt water fish with a world-wide distribution.

More than 35 years have passed since the first description of streptococci as the etiologic agents of a fish disease [14]. Over the years, septicemic diseases and meningoencephalitis due to streptococcus or streptococcus-like organisms (Gram-positive cocci) became one of the major threats to aquaculture. ‘‘Streptococcosis’’ in fish has been reported from various parts of the world: Japan [16, 18], Taiwan [25], Korea [15], South Africa [1], USA [22, 23], Australia [3], and Israel [8]. In Europe the disease appeared only recently, but has already been described in many countries, including Spain [20], England [26], and Italy [11]. In Italy, from sporadic outbreaks observed 3 years ago, the disease has spread to the entire northern part of the country, killing as much as 50% of the crop, thus becoming the most important risk factor to the trout industry during summer [12]. It is now clear that streptococcosis in fish is not due to a single bacterial species. Various microorganisms, pathogenic for fish, have been isolated and characterized, such as Streptococcus iniae, S. diffıcile [8, 21], Lactococcus piscium [26], Vagococcus salmoninarum [6], and Enterococcus seriolicida [17]. E. seriolicida was first described in 1991 as a new species

Correspondence to: H. Bercovier

that was responsible for considerable economic loss in Japanese aquaculture. E. seriolicida was known as a pathogen of saltwater fish species, and, with the exception of eels, was never described in freshwater species, nor was it described elsewhere than in Japan. The results of the present work show that Enterococcus seriolicida [17] is in fact a synonym of Lactococcus garvieae [4], another fish pathogen, and is responsible for fatal septicemia and meningoencephalitis in trout. Materials and Methods Bacterial strains. The bacterial strains used in this study are listed in Table 1. The 30 Italian strains were all field isolates collected from 1991 to 1993 from diseased rainbow trout presenting symptoms of septicemia and meningoencephalitis (‘‘Pop-eyes’’). Diseased fish were sacrificed, and samples from the kidney, aseptically drawn, were streaked onto blood agar plates and incubated at 28°C. All isolates were stored in storage medium (Sanofi Diagnostics Pasteur) at 270°C. The reference strains (T stands for type strain) E. seriolicida ATCC 49156T, L. garvieae ATCC 43921T, S. iniae ATCC 29178T, S. diffıcile CIP 103769T, L. lactis NCFB 604T, L. piscium NCFB 2778T, and V. salmoninarum NCFB 2777T were stored under identical conditions. Media, biochemical, and physiological tests. Bacteria were routinely grown on Columbia agar (GIBCO) containing 5% defibrinated sheep blood. For large quantities, bacteria were grown in BHI broth (DIFCO). The bile (‘‘ox-gall’’) and the skimmed milk (methylene blue milk test) were also obtained from DIFCO. All

86

CURRENT MICROBIOLOGY Vol. 32 (1996)

Table 1. DNA–DNA homologies between L. garvieae, E. seriolicida, Italian field isolates, and miscellaneous Gram-positive cocci Strain designationa ITP 1010 ITP 1039 ITP 1055 ITP 1061 ITP 1072 ITP 1313 ITP 1316 ITP 1371 ITP 1375 ITP 1457 ITP 1486 ITP 1545 ITP 1547 ITP 1676 ITP 1794 ITP 1814 ITP 1815 ITP 1833 ITP 1873 ITP 1877 ITP 1897 ITP 1917 ITP 1920 ITP 1921 ITP 1964 ITP 1965 ITP 1992 ITP 1995 ITP 2001 ITP 2035 L. garvieae ATCC 43921T E. seriolicida ATCC 49156T L. lactis NCFB 604T S. iniae ATCC 29178T S. diffıcile CIP 103769T L. piscium NCFB 2778T V. salmoninarum NCFB 2777T

Place of isolation

Date of

isolationb

RBRc (DTm(e))d to strain ITP 2001

RBR (DTm(e)) to L. garvieae ATCC 43921T

Cassolnovo Cassalnovo Calcio Cerano T. Pallavicina T. Pallavicina Torre Pallav. Treviso Istrana Cassalnovo Cerano T. Pallavicina Calcio Centallo Cassolnovo Cerano Cassolnovo Cassalnovo Cerano T. Pallavicina Cassalnovo Cassalnovo Trecate Calcio Cassalnovo Cerano Arco Beinette Cassalnovo Cerano England

8/91 9/91 9/91 10/91 10/91 5/92 5/92 6/92 6/92 8/92 9/92 11/92 11/92 3/93 5/93 5/93 8/93 6/93 7/93 7/93 9/93 7/93 7/93 7/93 8/93 8/93 9/93 9/93 9/93 10/93 1981

96 (0.5) 94 (0.9) 76 (0.7) 94 76 100

Kochi, Japan

1974

79 (0.6)

82 (0.7)

Denmark San Fransisco, USA Ein Hanatziv, Israel

1940 1976 9/89

21 (7.3) 9 (12)

39 (6.0) 32 (8.0) 16 (14)

USA

1989

31

27 (4.5)

USA

1989

26

30 (8.4)

97 97 96 76 (0.4) 77 83 85 (1.0) 82 (0.2) 80 86 (1.0) 96 (0.5) 94 96 87 (0.3) 98 86 (0.1) 91 93 (0.8) 100 91 (1.1) 85 (0.5)

74 72

73 (0.2)

85 (0.3) 100

a ITP is the Experimental Institute for Zooprophylaxis in Piedmont, Liguria and Aosta Valley, Torino. When strains were isolated in Italy, the location is given. b Date (month and year of the isolation of the strain). c The relative binding ratio (RBR) is the mean of four determinations and is expressed in percentage. d DT m(e) is the difference in melting points and reflects the percentage of divergence in related DNA sequences. When not indicated in parentheses, it was not determined.

isolates were tested twice on the API 20 STREP system; the reference strains plus 15 randomly selected Italian strains were tested also on the API 50 CH system (La Balmes Les Grottes, France). The instructions of the manufacturer were followed. The tests were done at 28°C. Results were recorded after 24, 48, and 96 h.

DNA extraction and DNA hybridizations. The method used to lyse streptococci and to extract their DNA has been described previously [8]. Briefly, the pellet obtained from growth in 1 liter BHI was washed twice in PBS, resuspended in the same buffer, and lysed by adding 2 mg/ml of lysozyme plus 5 units/ml of mutanoly-

A. Eldar et al.: Enterococcus seriolicida Is a Synonym of Lactococcus garvieae

sin (Sigma). After 2 h, SDS (0.5% wt/vol) and 10 mg/ml of Pronase E (Sigma) were added, and incubation was allowed to proceed for an additional 12 h. Final lysis was achieved by raising the final SDS concentration to 2% (wt/vol). The DNAs were purified by phenol and chloroform extraction, and precipitated by adding two volumes of ice-cold ethanol supplemented with NaCl (final concentration, 100 mM). Spooled DNA was suspended in TE (10 mM Tris-HCl; 1 mM EDTA) and stored at 4°C. DNA-DNA hybridizations were done as described previously [2] with modifications regarding the volumes employed. In a final volume of 100 µl, 15 µg of sonicated DNA was hybridized with 5 ng of nick-translated 32P-labeled DNA. After boiling, the reassociation of the single-stranded DNA was allowed to proceed in 280 mM PBS for 18 h at 60°C. Single- and double-stranded DNA were separated by the hydroxyapatite method [2]. Reactions were done in duplicate, and each run was performed twice. RBR (relative binding ratio) and the differences in melting points DTm(e) reflecting the percentage of divergence in related DNA sequences were calculated as published previously [2, 8]. The ratio of self-reassociation of the labeled DNA was routinely 2–4%, which was subtracted from the absolute ratios of the hybridizations.

Results and Discussion Biochemical and physiological reactions. With the API 20 STREP and the CH 50 systems, no differences were seen between the Italian field isolates and the reference strains of E. seriolicida and L. garvieae. All were Vogues-Proskauer (VP), esculin (ESC), pyrrolidonylarylamidase (PYRA), leucine arylamidase (LAP), and arginine dehydrolase (ADH) positive. Ribose, galactose, D-glucose, D-fructose, D-mannose, mannitol, N-acetyl glucosamine, amygdalin, arbutin, salicin, cellobiose, maltose, trehalose and b-gentobiose were acidified by fermentation. All the other tests were negative including lactose and sucrose. In the 9th edition of the Bergey’s Manual of Determinative Bacteriology [13], L. garvieae is described as lactose positive and E. seriolicida as lactose and sucrose negative. The API 20 STREP identification code identified these strains, with a probability of 79.8%, as Lactococcus lactis. After 24 h of incubation at 37°C, all the strains grew in 0.3% methylene blue milk and in 4% NaCl. They were able to grow also at 10° and 42°C. These results suggested that they should be classified in the genus Lactococcus [13]. However, strain variability was seen when the various cultures were grown at 6.5% NaCl, at pH 9.6, or at 45°C. When tested in broth, positive results for these tests were rarely recorded with a weak and late growth (after 3–5 days’ incubation) not always reproducible. These tests are considered important criteria to differentiate between Lactococcus, which do not grow under these conditions, and Enterococcus, which do grow under these conditions. All the Italian field isolates, L. garvieae ATCC strain 43921 and E. seriolicida ATCC

87

strain 49156, were a-hemolytic on sheep blood agar in opposition to the description of Domenech et al. [7], who found these species to be b-hemolytic but in agreement with the description of E. seriolicida [13]. In conclusion, the Italian Gram-positive cocci isolated from diseased rainbow trout were found phenotypically identical with E. seriolicida and L. garvieae. Therefore, we pursued the study of these strains to define their genetic relatedness. DNA relatedness between the various isolates. All the Italian isolates as well as E. seriolicida and L. garvieae (Table 1) formed a homogeneous genetic group when hybridized with the labeled ITP 2001 strain (RBR of 76% to 100%). The DTm(e) of the Italian field isolates, hybridized to the field isolate ITP 2001, was consistently less than 1°C. When L. garvieae ATCC strain 43921T was labeled and hybridized with Italian field isolates including strain ITP 2001 and E. seriolicida ATCC 49156T, the RBR values varied from 73% to 85%, with DTm(e) of 0.3–0.7°C. The biochemical, physiological and genetic data indicated that the Italian field isolates should be classified as Lactococcus garvieae. In the 9th edition of the Bergey’s Manual [13], the lack of growth at 45°C and at pH 9.6 are considered essential characteristics to distinguish lactococci from enterococci. However, it is accepted that the physiological characteristics of a bacterial species should be expressed in terms of percentage rather than in a qualitative manner (plus or minus) [5, 10]. Murray [19] found that 56% of the lactococci tested were able to grow in 6.5% NaCl and that 25% of them were also able to grow at pH 9.6. Strain variability might be, at least in part, due to technical reasons such as the type, the batch, and the manufacturer of the media, the hardness of the water used to prepare the media, the type of culture (broth or agar plates), or the density of the inoculum. It is now accepted that taxonomy based on a molecular methods, mainly on DNA homology (DNA-DNA homology of more than 70% with a low DTm(e) defines a bacterial species) and on 16S rRNA sequences, should be used to establish the phylogenetic and taxonomic position of bacterial species [24]. Since Enterococcus seriolicida and Lactococcus garvieae were found identical in their biochemical properties and had a DNA-DNA relatedness of more then 70% with a low DTm(e), they should be considered as synonyms describing a single bacterial species. Our data confirm and strengthen the data of Domenech et al. [7], who showed that E. seriolicida and L. garvieae 16S rRNA sequences were identical. L. garvieae was first charac-

88

terized by Collins et al. [4], who described its cultural and serological properties and analyzed the composition of the cell-wall peptidoglycan along with its fatty acids and isoprenoid quinone contents. Enterococcus seriolicida was described some 10 years later [17] without data on 16S rRNA sequence. Enterococcus seriolicida strains should therefore be reclassified as Lactococcus garvieae. Our data and the data of other authors [7, 11, 17, 20] show clearly that Lactococcus garvieae is a major fish pathogen of rainbow trout in Europe as well as a pathogen of saltwater fish in the Far East. Lactococcus garvieae has been isolated also from humans and cattle [4, 9]. The finding that the bacterium is highly virulent in rainbow trout, with an LD50 of 10 bacteria per fish (Ghittino C. and Eldar A., unpublished data) must alarm the fish producers who should recognize that diseases due to Lactococcus garvieae are likely to become frequent in the future, with increasing export and import of fish and fish products between countries and continents. ACKNOWLEDGMENTS This research was supported in part by grant No. IS 2307-93 from the US-Israel BARD foundation.

Literature Cited 1. Boomkar J, Imes GD, Cameron CM, Naude TW, Schoonbee HJ (1979) Trout mortalities as a result of Streptococcus infection. Onderestepoort J Vet Med 46:71–78 2. Brenner DJ, Fanning GR, Rake AV, Johnson KE (1969) Batch procedure for thermal elution of DNA from hydroxyapatite. Anal Biochem 28:447–459 3. Carson J, Gudkovs N, Austin B (1993) Characteristics of an Enterococcus-like bacterium from Australia and South Africa, pathogenic for rainbow trout, Onchorynchus baum. J Fish Dis 16:381–388 4. Collins MD, Farrow JAE, Phillips BA, Kandler O (1983) Streptococcus garvieae sp. nov. and Streptococcus plantarum sp. nov. J Gen Microbiol 129:3427–3431 5. Collins MD, Farrow JAE, Katic V, Kandler O (1984) Taxonomic studies on streptococci of serological Groups E, P, U and V; description of Streptococcus porcinus sp. nov. Syst Appl Microbiol 5:402–413 6. Collins MD, Ash C, Farrow JAE, Wallbanks S, Williams AM (1989) 16S ribosomal ribonucleic acid sequence analyses of lactococci and related taxa. Description of Vagococcus fluvialis gen. nov. sp. nov. J Appl Bacteriol 67:453–460 7. Domenech A, Prieta J, Fernandez-Garayzabal JF, Collins MD, Jones D, Dominguez L (1993) Phenotypic and phylogenetic evidence for a close relationship between Lactococcus garvieae and Enterococcus seriolicida. Microbiologia 9:63–68

CURRENT MICROBIOLOGY Vol. 32 (1996)

8. Eldar A, Bejerano Y, Bercovier H (1994) Streptococcus shiloi and Streptococcus diffıcile: two new streptococcal species causing a meningoencephalitis in fish. Curr Microbiol 28:139– 143 9. Elliott JA, Collins MD, Piggott NE, Facklam RR (1991) Differentiation of Lactococcus lactis and Lactococcus garvieae from humans by comparison of whole cell protein profile patterns. J Clin Microbiol 29:2731–2734 10. Facklam RR, Hollis D, Collins MD (1989) Identification of gram-positive coccal and coccobacillary vancomycin-resistant bacteria. J Clin Microbiol 27:724–730 11. Ghittino C, Prearo M (1992) Report of streptococcosis in rainbow trout (Onchorynkus mykiss) in Italy: preliminary note. Boll Soc Ital Patol 8:4–11 12. Ghittino C, Prearo M (1993) Comparison of some strains isolated from rainbow trout affected by streptococcosis. Boll Soc Ital Patol 11:30–43 13. Gram-positive cocci. (1994) In: Holt JG, Krieg NR, Sneath PHA, Williams S.T. (eds), Bergey’s manual of determinative bacteriology, IX edition. Baltimore: Williams and Wilkins, pp 527–558 14. Hoshina T (1956) An epidemic disease affecting rainbow trout in Japan. J Tokyo Univ Fish 42:35–46 15. Kim YG, Lee KK (1994) Isolation, characterization and pathogenicity of a Streptococcus strain in the flounder (Paralichthys olivaceus) cultured in Korea. Bull Eur Ass Fish Pathol 14:8–11 16. Kitao T, Aoki T, Sakoh R (1981) Epizootic in culturing freshwater fish caused by beta haemolytic Streptococcus spp. Fish Pathol 15:301–307 17. Kusuda K, Kawai K, Salati F, Banner CR, Freyer JL (1991) Enterococcus seriolicida sp. nov., a fish pathogen. Int J Syst Bacteriol 41:406–409 18. Kusuda R, Komatsu I (1978) A comparative study on fish pathogenic Streptococcus isolated from freshwater and saltwater fishes. Bull Jpn Soc Sci Fish 44:1073–1078 19. Murray BB (1990) The life and times of the Enterococcus. Clin Microbiol Rev 3:46–65 20. Palacios MA, Zamora MJ, Velazques J, Zamora E, Duran A (1993) Streptococcosis in rainbow trout (Oncorhynchus mykiss) in Spain. Boll Soc Ital Patol 13:11–16 21. Pier GB, Madin SH (1976) Streptococcus iniae sp. nov., a beta hemolytic streptococcus isolated from an Amazon freshwater dolphin, Inia geofrrensis. Int J Syst Bacteriol 26:545–553 22. Rasheed V, Plumb JA (1984) Pathogenicity of a non haemolytic group B Streptococcus sp. in Gulf killfish (Fundulus grandis). Aquaculture 37:97–105 23. Robinson JA, Meyer FP (1966) Streptococcal fish pathogen. J Bacteriol 92:512 24. Stackenbrandt E, Teuber M (1998) Molecular taxonomy and phylogenetic position of lactic acid bacteria. Biochemie 70:317– 324 25. Tung MC, Chen SC, Tsai SS (1985) General septicemia of streptococcal infection in cage-cultured tilapia, Tilapia mossambica, in Southern Taiwan. COA Fish Dis Res 7:95–105 26. Williams AM, Freyer JL, Collins MD (1990) Lactococcus piscium sp. nov., a new Lactococcus species from salmonid fish. FEMS Microbiol Lett 68:109–114