View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Institute of Hydrobiology, Chinese Academy Of Sciences

Aquaculture Research, 2010, 41, 1835 ^ 1844 d o i: 10.1111/ j.1365 -2109.2010.02571. x

Isolation and characterization of ictaluri strains as pathogens from diseased yellow catfish Pelteobagrus fulvidraco (Richardson) cultured in China

Jin Yu Liu1,2,AiHuaLi1,DongRenZhou3, Zhou Rui Wen4 &XuePingYe3 1Institute of Hydrobiology,Chinese Academy of Sciences,Wuhan, China 2Graduate University of Chinese Academy of Sciences, Beijing, China 3Zhejiang Institute of Freshwater Fisheries, Huzhou Zhejiang, China 4Hubei Fisheries Science Research Institute,Wuhan, China

Correspondence: A H Li, Institute of Hydrobiology, Chinese Academy of Sciences,Wuhan, China 430072. E-mail: [email protected]

Abstract Introduction Yellow cat¢sh Pelteobagrus fulvidraco (Richardson) is a , a Gram-negative, rod-shaped, commercially important ¢sh generally distributed in oxidase-negative, peritrichous, fermentative bacter- Southeast Asian countries. The well-known aetiologi- ium, has been reported almost exclusively as an obli- cal agent of enteric septicaemia of cat¢sh, Edward- gate pathogen of ictalurids ¢sh and causes enteric siella ictaluri, was isolated from diseased yellow cat¢sh septicaemia of cat¢sh (ESC), especially in channel P. fulvidraco (Richardson) reared at two commercial cat¢sh Ictalurus punctatus (Hawke, Mc-Whorter, Stei- ¢sheries in China.The economic losses due to the high gerwalt & Brenner 1981). In the United States, enteric mortalities (about 50%) caused by this bacterium septicaemia of cat¢sh (ESC) was recognized as the have been increasing annually.The a¡ected ¢sh pre- most prevalent disease a¡ecting farm-raised channel sented two di¡erent, typical symptoms: pale gills, cat¢sh (Hawke 1979), costing approximately 50 mil- slight exophthalmia and a‘hole in the head’, and hae- lion dollars in losses annually (Mitchell 1997). Some morrhage on the opercula, in the skin under the jaw, other ictalurids (i.e. the ictaluridae family) ¢sh could creating a ‘hole under the jaw’. These diseases were be infected by E. ictaluri under natural conditions, found frequently in cultured yellow cat¢sh through- such as blue cat¢sh Ictalurus furcatus, white cat¢sh out China. The isolates from both outbreaks were all Ictalurus catus (Newton, Bird, Blevins, Wilt & Wolfe Gram negative, facultatively anaerobic and short rod. 1988) and brown bullhead Ictalurus nebulosus (Iwa- Morphological and biochemical tests and phyloge- nowicz, Gri⁄n, Cartwright & Blazer 2006). netic analysis based on the 16S rDNA sequences all Isolation of E. ictaluri from non-ictalurids has been strongly indicated that these yellow cat¢sh isolates reported in natural disease outbreaks in green knife were highly identical to the known E. ictaluri. In addi- ¢sh Eigenmannia virescens (Kent & Lyons1982), danio tion, the isolates possessed the typical plasmid pro¢le Danio devario (Blazer, Shotts & Waltman 1985; Walt- of E. ictaluri. Experimental infection assays were con- man, Shotts & Blazer 1985; Petrie-Hanson, Romano, ducted and pathogenicity (by an intraperitoneal injec- Mackey, Khosravi, Hohn & Boyle 2007), walking cat- tion) was demonstrated in yellow cat¢sh and channel ¢sh Clarias batrachus L. (Kasornchandra, Rogers & cat¢sh Ictalurus punctatus. The results showed that Plumb 1987), rosy barb Puntius conchonius (Hum- yellow cat¢sh isolates were quite conservative pheno- phrey, Lancaster, Gudkovs & McDonald 1986), rain- typically and genetically, and were able to cause two bow trout Oncorhynchus mykiss (Keskin, Secer, di¡erent, typical symptoms in this ¢sh under un- Izgur,Turkyilmaz & Mkakosya 2004), freshwater cat- known conditions and mechanism. ¢sh Pangasius hypophthalmus (Sauvage) (Crumlish, Dung, Turnbull, Ngoc & Ferguson 2002), Tra cat¢sh Keywords: isolation, characterization, Edwardsiel- Pangasianodon hypophthalmus (Sauvage) (Truong, la ictaluri, yellow cat¢sh Areechon, Srisapoome & Mahasawasde 2007), harle-

r 2010 The Authors Aquaculture Research r 2010 Blackwell Publishing Ltd 1835 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. Aquaculture Research, 2010, 41,1835^1844 ququin tetra Rosbara heteromorpha (Reid & Boyle en ponds of two commercial ¢sheries in Wuhan 1989) and tadpole madtoms Noturus gyrinus (Klesius, (located in the middle of China) and Huzhou (located Lovy, Evans, Washhuta & Arias 2003). Additionally, in eastern China) during August and September other species, such as chinook salmon Oncorhynchus in 2007 and 2008 respectively. During the out- tshawytscha, tilapia Sarotherodon aureus and Eur- breaks, the water temperatures were approximately opean cat¢sh Silurus gianis,havebeeninfectedex- 25^27 1C. perimentally with E. ictaluri (Plumb & Sanchez 1983; In the two areas, the behavioural signs of a¡ected Plumb & Huge 1987; Baxa, Gro¡, Wishkovsky & ¢sh included reduced feeding activity, lethargy in re- Hdrick 1990), but natural outbreaks in these species sponse, listless swimming on the surface with a‘head have not been reported. Actually,Plumb and Sanchez up^tail down’ posture, sometimes spiralling or occa- (1983) reported that a number of warmwater species, sional rapid swimming, usually followed by death. including largemouth bass Micropterus salmoides, Clinical signs of the diseased ¢sh in Huzhou included golden shiner Notemigonus crysoleucas and bighead ulcerated haemorrhages on the oral area, on the op- carp Aristichthys nobilis, were highly resistant to ercula, at the base of all ¢ns and in the skin under the E. ictaluri experimental intraperitoneal infections. jaw, even creating a ‘hole-under-the-jaw’ condition White sturgeon Acipenser transmontanus, striped bass (Fig. 1b). Upon necropsy, numerous petechial hae- Morone saxatilis and rainbow trout Oncorhynchus my- morrhages were found on the liver, intestines and kiss were resistant to E. ictaluri in experimental im- adipose tissue, and accumulation of ascites in the mersion exposures as well (Baxa et al.1990). abdomen, whereas clinical signs of the a¡ected Yellow cat¢sh Pelteobagrus fulvidraco is a species of ¢sh fromWuhan included pale gills, slight exophthal- order Siluriformes, family Bagridae, genus Pelteoba- mia, with the most serious symptom being an open grus, placed in the same order with channel cat¢sh, necrotic lesion in the skull between the eyes, namely but in a di¡erent family.Yellow cat¢sh is a commer- a‘hole in the head’ (Fig.1a). Five ¢sh showing typical cially important ¢sh generally distributed and clinical signs representative of each outbreak were cultured in Southeast Asian countries and especially collected and sent to the laboratory alive for fur- in China. However, the problem of bacterial diseases ther examination in plastic transportation bags with related to this ¢sh has drawn little attention. The pre- oxygen supply. sent report documents the ¢rst isolation of E. ictaluri strains from the moribund cultured yellow cat¢sh Bacterial isolation with di¡erent clinical symptoms and reared at two di¡erent farms in China. The organisms from the Using aseptic techniques, samples taken from the two sources were proved to be identical and patho- brain, kidneyand liver of the moribund yellow cat¢sh genic to yellow cat¢sh, channel cat¢sh ¢ngerlings by were streaked on brain^heart infusion (BHI, BD, an experimental intraperitoneal injection. Franklin Lakes, NJ, USA) agar plates and incubated at 26 1C for 48 h. Colonies on the plates were pure and single colonies were selected and re-streaked on Materials and methods the same media. Six bacterial strains were collected from the di¡erent moribund ¢sh: two strains (HSN-1 Diseased yellow cat¢sh and characterization and HSS-1) from Wuhan and four strains (HSA-1, Epizootics occurred in adult yellow cat¢sh (body HSA-2, HSX-1 and HSX-2) from Huzhou. HSN-1 and weight100 130 g approximately) cultured in earth- HSS-1 were isolated from the brain and the kidney

Figure 1 Yellow cat¢sh a¡ected by Edwardsiella ictaluri, showing the most typical external clinical signs of the disease: a‘hole in the head’ found inWuhan (a) and a ‘hole under the jaw’ found in Huzhou (b).

r 2010 The Authors 1836 Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 Aquaculture Research, 2010, 41,1835^1844 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. respectively; HSA-1, HSA-2, HSX-1 and HSX-2 all Marcy l’Etoile, France) according to the manufac- were isolated from the liver. All isolates were main- turer’s instructions. tained separately in BHI (BD) broth with15% glycer- ol at 80 1C. Antimicrobial susceptibility testing

Morphological studies The susceptibility patterns of isolates to 25 antimicrobial agents (Oxoid, Mercers Row, Cam- Isolates were inoculated on BHI (BD) agar overnight at bridge, UK) including ampicillin (10 mg), bacitracin 1 26 C; colony morphology was observed directly and (0.04 IU), cephalothin V (30 mg), chloramphenicol using light microscopy.Cell morphology was examined (30 mg), cipro£oxacin (5 mg), clarithromycin (15 mg), using scanning electron microscopy (Hitachi, Japan) as doxycycline (30 mg), enoxacin (10 mg), erythromycin described previously (Goldstein, Newbury, Joy, Lyman, (15 mg), £orfenicol (30 mg), furadantin (300 mg), fura- Echlin, Lifshin, Sawyer & Michael 2003). Cell mobility zolidone (300 mg), gentamicin (10 mg), kanamycin was determined in wet mounts using a phase-contrast (30 mg), neomycin (30 mg), nor£oxacin (10 mg), o£oxa- microscopy (Olympus, Shinjuku-ku, Tokyo, Japan) ex- cin (5 mg), oxytetracycline (30 mg), piperacillin amination ( 1000) and in semisolid glucose moti- (100 mg), rifampicin (5 mg), spectinomycin (100 mg), lity deep cultures (Walters & Plumb1978). streptomycin (10 mg), sulphamethoxazole/trimetho- prim (23.75/1.25 mg), tetracycline (30 mg) and tobra- Biochemical analysis mycin (10 mg) were tested and determined using the Biochemical tests were performed at 30 1Cunless standard method of Kirby^Bauer (Bauer, Kirby, otherwise speci¢ed. Before being tested, the isolates Sherris & Turck 1966) on Mueller^Hinton agar 1 were subcultured twice overnight in BHI (BD) broth (MHA, BD) and incubated for 48 h at 26 C. The at 30 1C. For cultural and morphological examina- reference strain ATCC 25922 was tions, Gram staining, oxidation/fermentation (O/F) used as quality control (Miller, Walker, Baya, Clem- ens, Coles, Hawke, Henricson, Hsu, Mathers, Oaks, reaction, indole and hydrogen sulphide (H2S) pro- duction were investigated using conventional meth- Papapetropoulou & Reimschuessel 2003), but incu- 1 ods. The growth on MacConkey (BD) agar was tested bation occurred for 24 h at 37 C. Results were inter- for 7 days. Catalase activity was determined by trans- preted as susceptible, mid-susceptible or resistant, ferring fresh colonies from BHI agar to a drop of 5% based on zone diameters of inhibition, including the diameter of the disc (mm) (Liu, Li, Ji & Yang 2009). (v/v) H2O2 (SCRC, Shanghai, China) in a slide glass. Oxidase activity was determined using 1% (w/v) di- methyl r-phenylenediamine chloride. The optimum Molecular analysis growth temperature for each isolate was tested in BHI broth after incubation at 4, 10, 15, 20, 25, 30, 37 All isolates were incubated into 5 mL of BHI broth se- 1 and 42 1C for 15 days. Growth in the presence of parately for 24 h at 30 C. Genomic DNA was ex- 1.0%, 2.0% and 3.0% (w/v) of NaCl was determined tracted using the silica gel ¢lm genomic DNA in BHI broth until growth was observed or otherwise extraction kit (SBS, Shanghai, China) according to at least for15 days. Hydrolysis of esculin and gelatine the manufacturer’s protocol. DNA was stored at 1 was detected using the method described previously 20 C until use. The PCR reaction contained 0.5 U (Hsu, Shotts & Waltman1985). Reduction in1% (w/v) of Taq polymerase (TaKaRa, Shimogyo-ku, Kyoto, nitrate was determined as described previously after Japan), 5 mLof10 PCR bu¡er,3 mLof25mMMgCl2, incubation in BHI broth with 1% nitrate for 7 days. 2 mL of each 10mM dNTP (TaKaRa), 1 mL of each Acid formation from carbohydrates was tested using 10mM primer and 500 ng of template DNA, in a ¢nal a basal medium containing 2.0 g tryptone,5.0 g NaCl, volume made up to 50 mL with sterile double-distilled water. Two sets of universal primers U8f: 50-AGAG 0.2 g K2HPO4, 6.0 g agar, 10.0 g carbohydrate and 0 0 1000mL distilled water, adjusting to pH 7.0. Bro- TTGATCATGGCTCAG-3 and U1492r: 5 -GGTTCAC 0 mothymol blue was used as an indicator. TTGTTACGACTT-3 (Weisburg, Barns, Pelletier & Lane 1991) were used to amplify the 16S rRNA gene, which was synthesized by Invitrogen (Shanghai, Chi- Enzymatic activity pro¢les na). The ampli¢cations were carried out in a thermal Enzyme production of the isolates was determined cycler (BioRad, Alfred Nobel Drive, Hercules, CA, using the API ZYM kit (API ZYM 25200, bioMe¤rieux, USA) with the following parameters: an initial dena-

r 2010 The Authors Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 1837 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. Aquaculture Research, 2010, 41,1835^1844 turation step of 94 1Cfor3min;35serialcyclesof lates was extracted using the GenExtractTM plasmid 94 1Cfor1min,561C for 30 sec and extension at DNA extraction kit (DouPson, Beijing, China) accord- 72 1C for 90 sec; and a ¢nal extension step of 72 1C ing to the manufacturer’s protocol. Plasmid DNA ex- for 10min. A negative control (no template DNA) tracts were analysed by 0.7% (w/v) agarose gel was included in the PCR.The PCR products were ana- (containing ethidium bromide) electrophoresis in lysed by 1% (w/v) agarose gel (containing ethidium 1% Tris^acetic acid^EDTA bu¡er. A supercoiled bromide) electrophoresis in 1% Tris^acetic acid^ DNA ladder marker (TaKaRa) was used as a plasmid EDTA bu¡er. Gels were visualized and photographed DNA molecular marker. Gels were visualized and under UV illumination. photographed under UV illumination. The PCR products were puri¢ed and cloned into pMD18-T (TaKaRa) in order to transform E. coli Pathogenicity test (DH5a) competent cells. The positive clones were se- In order to test the pathogenic potential of E. ictaluri quenced by Bigdye-Terminator in Unigene (Shang- isolates, two strains HSN-1 and HSA-1 were selected hai, China). The 16S rRNA sequence obtained were for the experimental infection of healthy yellow cat- BLAST (Altschul, Gish, Miller, Myers & Lipman 1990) ¢sh or channel cat¢sh. Bacteria isolates were cul- in GeneBank database and analysed by the Riboso- tured on BHI agar under 30 1C overnight and made mal Database Project (RDP-II) (Cole, Chai, Farris, into bacterial suspensions with sterile phosphate- Wang, Julam, McGarrel, Garrity & Tiedje 2005). Pre- bu¡ered saline (PBS) bu¡er. Yellow cat¢sh with a viously published 16S rDNA sequences of related or- mean weight of12.5 1.2 g and channel cat¢sh with ganisms were obtained from the GeneBank database. a mean weight of10.5 0.3g were used inthe infec- The nucleotide sequences were aligned using the tion trials. CLUSTAL X program, version 1.8 (Thompson, Gibson, For the HSN-1 isolate, two species of ¢sh, yellow Plewniak, Jeanmougin & Higgins 1997). Genetic dis- cat¢sh and channel cat¢sh, were challenged. Each tances were obtained using Kimura’s two-parameter species of ¢sh had three groups (eight ¢sh in each model (Kimura 1980) and evolutionary trees were group), and each ¢sh was injected intraperitoneally constructed using the neighbour-joining method with 0.2 mL of bacterial suspension containing (Saitou & Nei 1987) with MEGA 3 program (Kumar,Ta- 5.8 107,5.8 10 6 or 5.8 105 CFUmL 1 in PBS. mura & Nei 2004) through 1000-replicate bootstrap In the control group, each ¢sh injected with 0.2 mL analysis. GeneBank Accession numbers of the nu- of PBS was kept under the same conditions. For the cleotide sequences used for the phylogenetic analysis HSA-1isolate, only yellow cat¢sh, which were divided are shown in Fig. 2. into ¢ve groups (each consisting of seven ¢sh), were used. Each ¢sh was injected intraperitoneally with Plasmid pro¢le analysis 0.2 mL of bacterial suspension containing from All isolates were incubated in 5 mL of BHI broth 1.0 10 8 to 1.0 10 4 CFUmL 1 in PBS, and seven overnight at 30 1C and 1.5 mL aliquots were centri- ¢sh were injected with 0.2 mL of PBS as a control. fuged at17000 g for 1min. Then plasmid DNA of iso- All assays were conducted in 100L aquaria with

Edwardsiella ictaluri JCM 1680T (AB050826) 55 HSX-2 (GU176615) HSN-1 (EF015475) 98 HSA-1 (GU176612) HSA-2 (GU176613) Figure 2 Phylogenetic relation- 51 Edwardsiella ictaluri 93-146 (EU285521) ships between yellow cat¢sh 61 HSS-1 (EF015476) Edwardsiella ictaluri isolates and HSX-1 (GU176614) other related bacteria based on ATCC 15947T (AB050827) 16S rRNA gene sequences. The phylogenetic tree was generated Edwardsiella hoshinae JCM 1679T (AB050825) using the neighbour-joining 95 Enterobacter cloacae ATCC 13047T (AJ251469) method. The numbers indicate Citrobacter freundii DSM 30039T (AJ233408) 100 bootstrap values1000. Gene- 100 Escherichia coli ATCC 11775T (X80725) Bank accession numbers are Shigella dysenteriae ATCC 13313T (X96966) given in parentheses.

r 2010 The Authors 1838 Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 Aquaculture Research, 2010, 41,1835^1844 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al.

50 L still water at a temperature ranging from 25 to but not at 37 1C. All could grow at10,15,20,25,30 and 27 1C, with aeration. The water was exchanged half 37 1C but failed to grow at 4 1C. All could grow on with aeration water once in 2 days. Clinical signs MacConkey agar and were not tolerant of 42% NaCl. and mortality were recorded daily for 14 days post Gas was produced from glucose by isolates at 25 1C challenge. but not at 37 1C. All isolates were positive for nitrate reduction, urease and phenylalanine ammonialyase

and negative for H2S, indole, MR,VP, dihydrolysis of Results arginine, hydrolysis of esculin and gelatine, or- Morphologic, phenotypic and biochemical nithine decarboxylase, lysine decarboxylase or utili- characteristics zation of malonate, citrate and acetate. All produced acid from alantin, D-galactose, D-glucose, D-maltose, Colonies of isolates on the BHI agar plate were usual- D-ribose and mannitose but did not produce acid ly opaque, smooth, circular and slightly convex, with from D-cellobiose, D-mannitol, D-melezitose, D-ra⁄- a diameter of 0.7^1.0 mm. All isolates were found to nose, D-sorbitol, esculin, galactitol, inositol, lactose, be phenotypically homogeneous and formed a laetrile, L-arabinose, melibiose, rhamnose, salicin, small (change ok??) rod shape, with a width and ÃÃ starch, sucrose, trehalose and xylose. In addition, all a length of 0.9^1.0 and 2.0^3.0 mm respectively. strains produced acid weakly from glycerol. The biochemical characteristics of isolates are summarized in Table 1 and compared with the known E. ictaluri and E. tarda strains. All isolates Enzyme production pro¢le in our research yielded identical biochemical test results. They were Gram-negative, catalase-positive In the API ZYM test, six strains presented identical and oxidase-negative. Motility was observed at 25 1C enzymatic properties. All showed positive reactions

Table 1 Characteristics of yellow cat¢sh Edwardsiella ictaluri isolates compared with known Edwardsiella ictaluri and Edward- siella tarda isolates

Edwardsiella Edwardsiella Edwardsiella Six ictaluri tarda Six Edwardsiellaictaluri tarda Characteristic isolates ATCC33202Ã ATCC15947w Characteristic isolates ATCC33202Ã ATCC15947 w

Gram stain 0 Urease 0 Growth at 4 1C0Phe deaminase 0 Growth at 37 1C611Acid production from Motility at 25 1C611Amygdalin 0 Motility at 37 1C0 1 Arabinose 0 Gas from glucose at 25 1C6 11Dulcitol 0 Gas from glucose at 37 1C0 1 Galactose 0 1 Maconkey agar 6 11Glucose 6 11 Oxidase 0 Inositol 0 Catalase 6 11Lactose 0 Oxidation/fermentation 6/6 1/11/1 Maltose 6 11 Nitrate reduction 6 11Mannitol 0 Gelatine hydrolysis 0 Mannose 6 11 Esculin hydrolysis 0 Mellibiose 0 Malonate 0 Raffinose 0 Simmons citrate 0 Rhamnose 0

H2S production 0 1 Ribose 6 11 Indole test 0 1 Salicin 0 Methyl red 0 1 Sorbitol 0 Voges–Proskauer test 0 Starch 0 Lysine decarboxylase 6 11Sucrose 0 Ornithine decarboxylase 6 11Trehalose 0 Arginine double hydrolase 0 Xylose 0

ÃCombined results from Hawke et al.1981,Waltman et al. (1986) and Holt and Noel (1994). wCombined results from Amandi, Hiu, Rohovec and Fryer (1982) and Holt and Noel (1994).1, 90% or more of strains positive; ,90% or more strains negative; ND, not determined.

r 2010 The Authors Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 1839 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. Aquaculture Research, 2010, 41,1835^1844

Table 2 Enzymatic pro¢les of yellow cat¢sh Edwardsiella M 123456M ictaluri isolates bp

6133 Two isolates Four isolates 3977 Biochemical test from Wuhanà from Huzhou 3049 2087 Alkaline phosphatase 11 Esterase (C4) 11 Figure 3 Plasmid pro¢les of yellow cat¢sh Edwardsiella ic- Esterase lipase (C8) 11 taluri isolates M, supercoiled DNA ladder marker;1, HSN-1; Lipase (C14) 2, HSS-1;3, HSA-1;4, HSA-2;5, HSX-1; and 6, HSX-2. Leucine arylamidase 11 Valine arylamidase 11 Cystine arylamidase Trypsin Table 3 Antimicrobial susceptibility testing of yellow cat- Chymotrypsin 11 ¢sh Edwardsiella ictaluri isolates Acid phosphatase 11 Naphthol-AS-BI-phosphohdrolase 11 Two isolates Four isolates a -galactosidase Antimicrobial agents from Wuhanà from Huzhou b-galactosidase b-glucuronidase Acetylspiramycin R R a-glucosidase Ampicillin R R b-glucosidase Bacitracin R R N-acetyl-b-glucosaminidase 11 Cephalothin V S S a-mannosidase Chloramphenicol S S a-fucosidase Ciprofloxacin S S Clarithromycin R R Ã1, positive; , negative. Clindamycin R R Doxycycline S S Enoxacin S S for the following enzymes: alkaline phosphatase, es- Erythromycin M R terase (C4), esterase lipase (C8), leucine arylamidase, Florfenicol S S valine arylamidase, chymotrypsin, acid phosphatase, Fosfomycin S S naphthol-AS-BI-phosphohdrolase and N-acetyl-b- Furadantin S S glucosaminidase, and negative reactions for lipase Gentamicin S S Kanamycin S S (C14), cystine arylamidase, trypsin, a-galactosidase, Neomycin S S b-galactosidase, b-glucuronidase, a-glucosidase, b-glu- Norfloxacin S S cosidase, a-mannosidase and a-fucosidase (Table 2). Ofloxacin S S Oxytetracycline S S Piperacillin R S 16rDNA sequences analysis Rifampicin R M Spectinomycin R R Nucleotide sequences of 16S rRNA genes of six iso- Streptomycin M M Sulphamethoxazole/trimethoprim R S lates were determined. The GeneBank Accession Tetracycline S S Numbers were EF015475, EF015476, GU176612, Tobramycin S S GU176613, GU176614 and GU176615 respectively. à They had almost identical 16S rRNA sequences R, resistant; S, susceptible; M, medium susceptible. (99.9% similarity), implying that they were members of one species. A phylogenetic tree was constructed and is shown in Fig. 2. They formed a single cluster lates was found to harbour two plasmids. By interpo- with the E. ictaluri type stain JCM1680 and E. ictaluri lation from the reference plasmid molecular marker, 93-146 (GeneBank Accession no. AB050826, the estimated size of the two plasmids in each of the EU285521respectively). isolates was about 4100 and 5600 base pairs (bp).

Plasmid pro¢le Antimicrobial susceptibility result

Plasmid DNA was prepared from the infected yellow The susceptibility pattern of the isolates from 25 anti- cat¢sh isolates and is shown in Fig.3. Each of the iso- bacterial agents is shown in Table 3. All isolates were

r 2010 The Authors 1840 Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 Aquaculture Research, 2010, 41,1835^1844 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. susceptible to cephalothin V, cipro£oxacin, doxycy- appears that the higher the increase in ¢sh produc- cline, enoxacin, £orfenicol, furadantin, furazolidone, tion, the higher the risk of disease outbreak. The neomycin, nor£oxacin, o£oxacin, oxytetracycline main reason may be the high density of cultured ¢sh and tetracycline. Additionally, isolates from Wuhan in the intensive system. The waste from unconsumed were susceptible to chloramphenicol, gentamicin, feed, excretion and poor pond management lead to kanamycin and tobramycin, while Huzhou isolates the occurrence of ¢sh disease. Among the diseases were susceptible to ampicillin, piperacillin and found commonly in yellow cat¢sh, bacterial infec- sulphamethoxazole/trimethoprim. Wuhan isolates tion is worth discussing and considering due to the showed medium susceptibility to erythromycin and high mortality rate and the loss of income (Deng, streptomycin, while Huzhou isolates showed med- Luo, Tan, Qiu & Chen 2008). ium susceptibility to chloramphenicol, gentamicin, In the present study, morphological, biochemical kanamycin and rifampicin. According to the result, and molecular assays con¢rmed that the yellow cat- the appropriate drug was recommended to the ¢sh- ¢sh isolates were E. ictaluri, the same species as that eries, resulting in the successful control of the out- found from other species of aquatic animals by others breaks. This test was carried out not only for this (Hawke 1979; Hawke et al. 1981; Waltman, Shotts & purpose; the drug resistance pro¢le of isolates also Hsu 1986). The six strains, isolated from two regions, re£ects the characteristics of this species. had identical physiological and biochemical charac- teristics, 16S rRNA gene sequences, plasmid pattern and enzymatic pro¢le, with a few di¡erent antimicro- Pathogenicity bial susceptibility patterns. E. ictaluri was placed in the genus Edwardsiella In challenge trials, the results showed that E. ictaluri within the family Enterobacterieae based on biochem- isolates HSN-1 and HSA-1 were virulent for yellow ical characterization and DNA^DNA homology cat¢sh and channel cat¢sh ¢ngerlings. For the HSN- (Hawke et al.1981). Biochemically, E. ictaluri has been 1 isolate, clinical signs of a¡ected yellow cat¢sh in- described as a very homogenous organism (Plumb & cluded haemorrhages on the oral area, at the base of Vinitnantharat 1989) and shares many common fea- the pectoral ¢n and accumulation of ascites in the ab- tures with another member of this genus, E. tarda,ex- domen. Mortality occurred at the rates of 75%,100% cept for some di¡erent characteristics including the and 100% via the low, medium and high doses re- production of indole, the production of H SinSIM spectively. For channel cat¢sh, ulcer or perforation 2 media and motility at 37 1C (Waltman, Shotts & Hsu on pars cephalica, haemorrhages in the oral area 1986). The yellow cat¢sh isolates were negative for and on the base of the pectoral ¢n, and accumulation the production of indole or H S, and motility was ob- of ascites in the abdomen were observed. Mortality 2 served at 25 1C and not at 37 1C, which was the same occurred at the rates of 0%, 50% and 62.5% via the as in E. ictaluri,butnotE. tarda. low, medium and high doses respectively. For the According to16S rDNA sequence analysis, the iso- HSA-1 isolate, the symptoms of infected ¢sh were lates had 16S rDNA gene sequences (99% similarity) identical to those of the naturally infected ¢sh, and that were almost identical to those of the validated the LD value in yellow cat¢sh was determined to 50 E. ictaluri species. The phylogenetic tree of 16S rDNA be 0.2 10 4.7 CFUmL 1from the dose^response re- showed that the isolates formed a single cluster and lationships using the improved Karber method. A were closely related to the type strain of E. ictaluri pure culture of E. ictaluri was re-isolated and identi- (Fig. 2). Furthermore, plasmid pro¢le analysis has ¢ed from the organs (brain, liver and kidney) of the been applied to identify the isolates of E. ictaluri ra- dying ¢sh. No control ¢ngerlings developed clinical pidly. Several studies indicated that most E. ictaluri signs or died and no bacteria were re-isolated from isolates, regardless of origin, harboured homologous control ¢sh. cryptic plasmids of similar, but not identical sizes that are genetically related (Lobb & Rhoades 1987; Speyerer & Boyle 1987; Newton et al. 1988; Reid & Discussion Boyle 1989; Lobb, Gha¡ari, Hayman & Thompson Yellow cat¢sh production in China has increased in 1993).These plasmid classes include a 4.0^4.7kb plas- recent years. The abundance of freshwater, the cli- mid and a 5.6 kb plasmid (Reger, Mockler & Miller mate and the consumption market have led to in- 1993; Fernandez, Pittman-Cooley & Thune 2001).This creased cultivation of yellow cat¢sh. However, it suggests that these plasmids are extremely stable and

r 2010 The Authors Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 1841 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. Aquaculture Research, 2010, 41,1835^1844 highly conserved in almost all E. ictaluri isolates (Ab- The experimental infection trials using yellow cat- bott & Janda 2006), indicating that the plasmid pro¢le ¢sh and channel cat¢sh ¢ngerlings all clearly demon- is valuable for the identi¢cation of E. ictaluri. In con- strated the pathogenic potential of the E. ictaluri trast to the small-size plasmids detected in E. ictaluri isolates, which could develop acute septicaemia un- strains, much larger extrachromosomal elements der experimental conditions. The ‘hole-under-the- have been isolated from E. tarda strains (Abbott & Jan- jaw’ symptom was only reproduced slightly in a few da 2006). Plasmids in the E. tarda species range in mo- yellow cat¢sh, but not in channel cat¢sh. And the lecular mass from 2 to120 MDa, and plasmid carriage pathogenic mechanism was worth studying clearly. appears to vary signi¢cantly from strain to strain Nevertheless, the ‘hole-in-the-head’ symptom was a (Janda, Abbott, Kroske-Bystrom, Cheung, Powers, chronic disease caused by many reasons, and could Kokka & Tamura 1991). In our studies, the estimated not be replicated in our challenge experiments. size of the two plasmids in each of the isolates was 4.1 and 5.6 kb, which was similar to other E. ictaluri strains as discussed above. Acknowledgments The owner of the ¢shery had su¡ered a severe loss of up to 30^40% pond ¢sh in the outbreaks. In com- The project was supported by the National Basic pliance with our suggestion, the owner treated the Research Program of China (973 Program) ¢sh in some of the a¡ected ponds by oral administra- (2009CB118705) and the Open Fund of Guangdong tion of oxytetracycline, and within 2 weeks, there Provincial Key Lab of Pathogenic Biology and Epide- were no longer any deaths of ¢sh. Interestingly, the miology for Aquatic Economic Animals. antimicrobial susceptibility pattern of yellow cat¢sh isolates in the present study was quite similar to that recorded by other researchers (Stock & Wiedemann References 2001; McGinnis, Gaunt, Santucci, Simmons & Endris Abbott S.L. & Janda J.M. (2006) The Genus Edwardsiella 2003; Dung, Haesebrouck, Tuan, Sorgeloos, Baele & Chapter 3.3.4. 6. In Prokaryotes. Springer, New York, NY, Decostere 2008), indicating that E. ictaluri possesses USA pp.72^89. a relatively stable antimicrobial susceptibility pattern Altschul S.F., Gish W., Miller W., Myers E.W. & Lipman D.J. regardless of its geographical source. (1990) Basic local alignment search tool. Journal of Mole- In channel cat¢sh, E. ictaluri can cause acute septi- cular Biology 215,403^410. caemia or a chronic open lesion between the frontal Amandi A., Hiu S.F., Rohovec J.S. & Fryer J.L. (1982) Isolation bones of the skull, posterior to or between the eyes and characterizationof Edwardsiella tarda from fall Chi- (Plumb 1993). Concerning the pathogenesis, E. icta- nook salmon (Oncorhynchus tsawytscha). Applied Environ- luri can infect ¢sh by two di¡erent routes. Water- ment Microbiology 43,1380^1384. borne bacteria can invade the olfactory organ via Bauer A.W., Kirby W.M., Sherris J.C. & Turck M. (1966) Anti- biotic susceptibility testing by a standardized single the nasal opening of the ¢sh and migrate into the ol- disk method. American Journal of Clinical Pathology 45, factory nerve and then into the brain (Miyazaki & 493^496. Plumb 1985; Shotts, Blazer & Waltman 1986). The in- Baxa D.V., Gro¡ J.M.,WishkovskyA. & Hdrick R.P.(1990) Sus- fection spreads from the meninges to the skull and ceptibility of nonictalurid ¢shes to experimental infection skin, thus creating a ‘hole-in-the-head’ condition. with Edwardsiella ictaluri. Disease Aquatic Organisms 8, E. ictaluri can also be ingested and can enter the 113 ^ 117. blood through the intestine and result in septicaemia Blazer V.S., Shotts E.B. & WaltmanW.D. (1985) Pathology as- (Shotts, Blazer & Waltman 1986). By this route, the sociated with Edwardsiella ictaluri in channel cat¢sh, bacteria apparently colonize capillarities in the der- Ictalurus punctatus (Ra¢nesque) and Danio devario mis, causing necrosis and haemorrhage. (Hamilton-Buchanan, 1822). Journal of Biology 27, In the present study, E. ictaluri was found to cause 167^ 175. Cole J.R., Chai B., Farris R.J.,Wang Q., Julam S.A., McGarrel diseases in the yellow cat¢sh population naturally D.M., Garrity G.M. & Tiedje J.M. (2005) The ribosomal da- and caused two kinds of clinical signs, ‘hole-under- tabase project (RDP-II): sequences and tools for high- the-jaw’ and ‘hole-in-the-head’ symptoms, which throughput rRNA analysis. Nucleic Acids Research 33, were similar to those of the a¡ected channel cat¢sh 294^296. (Plumb1993), except that severe haemorrhage occurs Crumlish M., Dung T.T., Turnbull J.F., Ngoc N.T.N. & Fergu- in the skin under the jaw, creating a ‘hole under the son H.W. (2002) Identi¢cation of Edwardsiella ictaluri from jaw’ in yellow cat¢sh. diseased freshwater cat¢sh, Pangasius hypophthalmus

r 2010 The Authors 1842 Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 Aquaculture Research, 2010, 41,1835^1844 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al.

(Sauvage), cultured in the Mekong delta,Vietnam. Journal (Oncorhynchus mykiss). TurkishJournal ofVeterinaryAnimal of Fish Disease 25, 733^736. Sciences 28,649^653. DengX.Y.,LuoW.,TanS.H.,QiuS.H.&ChenK.G.(2008)Iso- Kimura M. (1980) A simple method for estimating evolution- lation and identi¢cation of bacteriosis pathogen ^ Edwra- ary rates of base substitutions through comparative stu- siella tarda from yellow cat¢sh (Pelteobagrus fulvidraco) dies of nucleotide sequences. Journal of Molecular with red head disease. Oceanologia and Limnologia Sinica Evolution16, 111 ^ 120. 39,511^515. Klesius P., Lovy J., Evans J., Washhuta E. & Arias C. (2003) DungT.T., Haesebrouck F.,Tuan N.A., Sorgeloos P., Baele M. & Isolation of Edwardsiella ictaluri from tadpole madtom in Decostere A. (2008) Antimicrobial susceptibility pattern of a southwestern New Jersey River. Journal of Aquatic Ani- Edwardsiella ictaluri isolates from natural outbreaks of ba- mal Health15,295^301. cillary necrosis of pangasianodon hypophthalmus inViet- Kumar S., Tamura K. & Nei M. (2004) MEGA3: integrated nam. Microbial Drug Resistance14,311^316. software for molecular evolutionary genetics analysis Fernandez D.H., Pittman-Cooley L. & Thune R.L. (2001) Se- and sequence alignment. Brie¢ngs in Bioinformatics 5, quencing and analysis of the Edwardsiella ictaluri plas- 150^163. mids. Plasmid 45, 52^56. Liu J.Y., Li A.H., Ji C. & YangW.M. (2009) First description of a Goldstein J., Newbury D., Joy D., Lyman C., Echlin P., Lifshin novel Weissella species as an opportunistic pathogen for E., Sawyer L. & Michael J. (2003) Scanning Electron Micro- rainbow trout Oncorhynchus mykiss (Walbaum) in China. scopy and X-ray Microanalysis ISBN: 0306472929,3rd Veterinary Microbiology136,314^320. edn. Kluwer Academic/Plenum Publishers, Spring Street, Lobb C.J. & Rhoades M. (1987) Rapid plasmid analysis for NY,USA. identi¢cation of Edwardsiella ictaluri from infected chan- Hawke J.P. (1979) A bacterium associated with disease of nel cat¢sh (Ictalurus punctatus). Applied Environment Mi- pond-cultured channel cat¢sh, Ictalurus punctatus. Journal crobiology 53, 1267^1272. of the Fisheries Research Board of Canada 36, 1508^1512. Lobb C.J., Gha¡ari S.H., HaymanJ.R. & Thompson D.T. (1993) Hawke J.P., Mc-WhorterA.C., Steigerwalt A.G. & Brenner D.J. Plasmid and serological di¡erences between Edwardsiella (1981) Edwardsiella ictaluri sp. nov., the causative agent of ictaluri strains. Applied Environment Microbiology 59, enteric septicemia of cat¢sh. International Journal of Sys- 2830^2836. tematic Bacteriology 31,396^400. McGinnis A., Gaunt P., Santucci T., Simmons R. & Endris R. Hsu T.C., Shotts E.B. & Waltman W.D. (1985) Action of Aero- (2003) In vitro evaluation of the susceptibilityof Edwardsiel- monas hydrophila on carbohydrate substrate. Journal of ¢sh la ictaluri, etiological agent of enteric septicemia in chan- Pathology 20,245^251. nel cat¢sh, Ictalurus punctatus (Ra¢nesque), to £orfenicol. Holt J.G. & Noel R.K. (1994) Bergey’s Manual of Systematic Journal ofVeterinary Diagnostic Investigation15,576^579. Bacteriology, 9th edn. The Williama and Wilkins, Balti- Miller R.A., Walker R.D., Baya A., Clemens K., Coles M., more, MD, USA. Hawke J.P., Henricson B.E., Hsu H.M., Mathers J.J., Oaks Humphrey J.D., Lancaster C., Gudkovs N. & McDonald W. J.L., Papapetropoulou M. & Reimschuessel R. (2003) Anti- (1986) Exotic bacterial pathogens Edwardsiella tarda and microbial susceptibility testing of aquatic bacteria: qual- Edwardsiella ictaluri from imported ornamental ¢sh Betta ity control disk di¡usion ranges for Escherichia coli ATCC splendens and Puntius conchonius, respectively: isolation 25922 and Aeromonas salmonicida subsp. Salmonicida and quarantine signi¢cance. AustralianVeterinaryJournal ATCC 33658 at 22 and 28 1C. Journal of Clinical Microbiol- 63, 369^371. ogy 41, 4318^ 4323. Iwanowicz L.R., Gri⁄n A.R., Cartwright D.D. & Blazer V.S. Mitchell A.J. (1997) Fish disease summaries for the south- (2006) Mortality and pathology in brown bullheads eastern United States from1976^1995. Aquaculture Maga- Amieurus nebulosus associated with a spontaneous Ed- zine 23,87^93. wardsiella ictaluri outbreak under tank culture condi- Miyazaki T. & Plumb J.A. (1985) Histopathology of Edward- tions. DiseaseAquatic Organisms 70,219^225. siella ictaluri in channel cat¢sh Ictalurus punctatus (Ra¢n- Janda J.M., Abbott S.L., Kroske-Bystrom S., Cheung W.K.W., esque). Journal of Fish Diseases 8,389^392. Powers C., Kokka R.P. & Tamura K. (1991) Pathogenic Newton J.C., Bird R.C., Blevins W.T., Wilt G.R. & Wolfe L.G. properties of Edwardsiella species. Journal of Clinical Mi- (1988) Isolation, characterization, and molecular cloning crobiology 29,1997^2001. of cryptic plasmids from Edwardsiella ictaluri. American Kasornchandra J., Rogers W.A. & Plumb J.A. (1987) Edward- Journal ofVeterinary Research 49,1856^1860. siella ictaluri from walking cat¢sh, Clarias batrachus L., in Plumb J.A. (1993) Edwardsiella septicemia. In: Bacteria Dis- Thailand. Journal of Fish Disease10, 137^ 138. eases of Fish (ed. byV.Inglis, R.J. Roberts & N.R. Bromage), Kent M.L. & Lyons J.M. (1982) Edwardsiella ictaluri in the pp. 67^72. Institute of Aquaculture, , Oxford Blackwell green knife ¢sh, Eigemannia virescens. Fish Health News Scienti¢c Publications, London, UK. 2,2. Plumb J.A. & Huge V. (1987) Susceptibility of European cat- Keskin O., Secer S., Izgur M.,Turkyilmaz S. & Mkakosya R.S. ¢sh (Silurus gianis)toEdwardsiella ictaluri. Journal of Ap- (2004) Edwardsiella ictaluri infection in rainbow trout plied Ichthyology 3,45^48.

r 2010 The Authors Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844 1843 E. ictaluri isolates from diseased yellow cat¢sh JYLiuet al. Aquaculture Research, 2010, 41,1835^1844

Plumb J.A. & Sanchez D.J. (1983) Susceptibility of ¢ve species Stock I. & Wiedemann B. (2001) Natural sus- of ¢sh to Edwardsiella ictaluri. Journal of Fish Disease 6, ceptibilities of Edwardsiella tarda, E. ictaluri, and E. hoshi- 261^266. nae. Antimicrobial Agents and Chemotherapy 45,2245^ Plumb J.A. & Vinitnantharat S. (1989) Biochemical, biophy- 2255. sical, and serological homogenicity of Edwardsiella icta- Thompson J.D., Gibson T.J., Plewniak F., Jeanmougin F. & luri. Journal of Aquatic Animal Health1, 51^56. Higgins D.G. (1997) The CLUSTAL X windows interface: Petrie-Hanson L., Romano C.L., Mackey R.B., Khosravi P., £exible strategies for multiple sequence alignment aided Hohn C.M. & Boyle C.R. (2007) Evaluation of zebra¢sh Da- by quality analysis tools. Nucleic Acids Research 25,4876^ nio rerio as a model for enteric septicemia of cat¢sh (ESC). 4882. Journal of Aquatic Animal Health19,151^158. Truong T.H., Areechon N., Srisapoome P.& Mahasawasde S. Reid W.S. & Boyle J.A. (1989) Plasmid homologies in (2007) Identi¢cation and antibiotic sensitivity test of the Edwardsiella ictaluri. Applied Environment Microbiology bacteria isolated from Tra cat¢sh (Pangasianodon hy- 55,3253^3255. pophthalmus [Sauvage,1878]) in pond cultured inVietnam. Reger P.J., Mockler D.F. & Miller M.A. (1993) Comparison of KasetsartJournal Natural Science 42,54^60. antimicrobial susceptibility, b-lactamase production, Walters G.R. & Plumb J.A. (1978) Modi¢ed oxidatiod/fermen- plasmid analysis and serum bactericidal activity in Ed- tation medium for use in identi¢cation of bacterial ¢sh wardsiella tarda, E. ictaluri,andE. hoshinae. Journal of Med- pathogens. Journal of the Fisheries Research Board of Cana- ical Microbiology 39,273^281. da 35, 1629^1630. Saitou N. & Nei M. (1987) The neighbour-joining method: a Waltman W.D., Shotts E.B. & Blazer V.S. (1985) Recovery of new method for reconstructing phylogenetic trees. Mole- Edwardsiella ictaluri from danio (Danio devario). Aquacul- cular Biology Evolution 4,406^425. tureAmsterdam 46,63^66. Shotts E.B., Blazer V.S. & Waltman W.D. (1986) Pathogenesis Waltman W.D., Shotts E.B. & Hsu T.C. (1986) Biochemical of experimental Edwardsiella ictaluri in channel cat¢sh characteristics of Edwardsiella ictaluri. Applied Environ- (Ictalurus punctatus). Canadian Journal of Fisheries and ment Microbiology 51,101^104. Aquatic Sciences 43,36^42. Weisburg W.G., Barns S.M., Pelletier D.A. & Lane D.J. (1991) Speyerer P.D. & Boyle J.A. (1987) The plasmid pro¢le of 16S ribosomal DNA ampli¢cation for phylogenetic study. Edwardsiella ictaluri. Journal of Fish Disease10,461^469. Journal of Bacteriology173,697^703.

r 2010 The Authors 1844 Aquaculture Research r 2010 Blackwell Publishing Ltd, Aquaculture Research, 41,1835^1844