Relationship between Permeability Glycoprotein (P-gp) Gene ...

This article was downloaded by: [Department Of Fisheries]On: 15 June 2014, At: 19:49Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Aquatic Animal HealthPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/uahh20

Relationship between Permeability Glycoprotein (P-gp)Gene Expression and Enrofloxacin Metabolism in NileTilapiaKun Hua, Gang Chengb, Haixin Zhanga, Huicong Wanga, Jiming Ruana, Li Chena, WenhongFangc & Xianle Yanga

a National Pathogen Collection Center for Aquatic Animals, Shanghai Ocean University, 999Hucheng Huan Road, Shanghai, 201306, Chinab South-Central College for Nationalities, 708 Nationalities Road, Wuhan, Hubei, 430074,Chinac East China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, 300Jungong Road. 200090, Shanghai, ChinaPublished online: 04 Jun 2014.

To cite this article: Kun Hu, Gang Cheng, Haixin Zhang, Huicong Wang, Jiming Ruan, Li Chen, Wenhong Fang & Xianle Yang(2014) Relationship between Permeability Glycoprotein (P-gp) Gene Expression and Enrofloxacin Metabolism in Nile Tilapia,Journal of Aquatic Animal Health, 26:2, 59-65, DOI: 10.1080/08997659.2013.860059

To link to this article: http://dx.doi.org/10.1080/08997659.2013.860059

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/uahh20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/08997659.2013.860059

http://dx.doi.org/10.1080/08997659.2013.860059

http://www.tandfonline.com/page/terms-and-conditions


Journal of Aquatic Animal Health 26:59–65, 2014C© American Fisheries Society 2014ISSN: 0899-7659 print / 1548-8667 onlineDOI: 10.1080/08997659.2013.860059

ARTICLE

Relationship between Permeability Glycoprotein(P-gp) Gene Expression and Enrofloxacin Metabolismin Nile Tilapia

Kun HuNational Pathogen Collection Center for Aquatic Animals, Shanghai Ocean University,999 Hucheng Huan Road, Shanghai 201306, China

Gang ChengSouth-Central College for Nationalities, 708 Nationalities Road, Wuhan, Hubei 430074, China

Haixin Zhang, Huicong Wang, Jiming Ruan, and Li ChenNational Pathogen Collection Center for Aquatic Animals, Shanghai Ocean University,999 Hucheng Huan Road, Shanghai 201306, China

Wenhong FangEast China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, 300 Jungong Road,Shanghai 200090, China

Xianle Yang*National Pathogen Collection Center for Aquatic Animals, Shanghai Ocean University,999 Hucheng Huan Road, Shanghai 201306, China

AbstractThe aim of this study was to analyze the influence of permeability glycoprotein (P-gp) gene expression on en-

rofloxacin (ENR) metabolism in aquatic animals. Nile Tilapia Oreochomis niloticus were fed different doses of ENRranging from 0 to 80 mg/kg. The P-gp gene expression levels were determined by quantitative real-time PCR (qRT-PCR) at indicated time points after drug administration. Drug metabolism was determined by HPLC. The P-gp geneexpression in liver and kidney was greatly enhanced 30 min after ENR administration at 40 mg/kg, peaked 3 h afterdrug administration, and then gradually decreased. Thirty minutes after a single oral administration of ENR (0, 20,40, or 80 mg/kg), the P-gp gene expression increased in a dose-dependent manner. The P-gp gene expression levelsin the kidney were significantly higher than those in the liver. Additionally, the metabolic rate of ENR in kidneywas more rapid than that in liver. Furthermore, a close correlation was found between P-gp gene expression andENR concentrations. These results suggest that P-gp may be involved in the ENR metabolism process in Nile Tilapia,providing a novel model for the potential utility of gene expression and drug metabolism studies in aquatic animals.

Permeability glycoprotein (P-gp), also known as transmem-brane glycoprotein, is widely expressed in normal cells, includ-ing cells that constitute the barrier, and is involved in metabolism

*Corresponding author: [email protected] July 9, 2013; accepted October 17, 2013

in the liver, kidney, intestines, pancreas, and brain microvascularendothelia (Saeki et al. 1993; Hemmer et al. 1995; Thomas et al.2007; Lee et al. 2013; Sousa et al. 2013; Wessler et al. 2013).

59

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4

60 HU ET AL.

This protein is present in certain tissue compartments, such asthe gastrointestinal tract, and plays an essential role in prevent-ing absorption and eliminating xenobiotics or preventing expo-sure of sensitive tissues to xenobiotic agents (Hochman et al.2001). Moreover, P-gp acts as a transmembrane efflux pump andis a member of the adenosine triphosphate (ATP)-binding cas-sette (ABC) transporter family (Hughes 1994; Rosenberg et al.2001). This efflux transporter has broad substrate specificity,including various structurally divergent drugs in clinical use to-day, and may reduce drug residues in tissues (Marzolini et al.2004). In addition, P-gp can mediate renal tubular secretion ofdrugs (Ito et al. 1993).

The expression of P-gp has been detected in a variety ofspecies, including Zebrafish Danio rerio (Bresolin et al. 2005),a killifish (Miller 1993), and a catfish (Kleinow et al. 2000),and has been proposed as a biomarker of pollution exposure.By using immunohistochemical analysis with a mammalian C-219 monoclonal antibody, Kleinow and colleagues detected P-gp protein expression in the liver and intestines of a catfish(Kleinow et al. 2000; Doi et al. 2001). In addition, P-gp pro-tein levels were measured seasonally in the gills of southeasternoysters Crassostrea virginica (Keppler and Ringwood 2001).Furthermore, P-gp in Zebrafish shares 49.4% and 49.7% nu-cleotide homology with that in human and mouse, respectively(Ma 2008). Similarly, a highly conserved sequence of the P-gp gene (over 78%) was found in major fish groups such asChondrichthyes and Osteichthyes (Ma 2008). Nile tilapia Oreo-chomis niloticus, which belongs to Osteichthyes, is an importantChinese aquatic product export.

Enrofloxacin (ENR) is a fluoroquinolone antibiotic that hasbeen widely applied to treat fish commercially (Dalsgaard andBjerregaard 1991). However, use of the primary metabolite ofENR, ciprofloxacin (CIP), may increase the risk of developingdrug resistance and chronic diseases (Hernandez et al. 2002).The European Community has established a maximum residuelimit (MRL) of 100 µg/kg for both ENR and CIP in the edibletissues of fish (DOCE 1990). Although the metabolism of ENRand CIP in fish has been given sufficient attention to ensure thesafety of human food production (Nouws et al. 1988; Bowseret al. 1992; Stoffregen et al. 1997; Intorre et al. 2000; della Roccaet al. 2004; Fang et al. 2012; Liang et al. 2012; Wack et al. 2012;Zhang et al. 2012), the in vivo metabolism of ENR and its poten-tial correlation with P-gp expression is still poorly understood.

In the present study, we investigated P-gp gene expressionlevels in the liver and kidney of Nile Tilapia after ENR admin-istration. In addition, the potential relationship between residueamounts of ENR and P-gp gene expression levels in these tis-sues was explored. These results may help to better understandthe mechanism of ENR disposition in aquatic animals.

METHODSReagents.—The antibiotics ENR and CIP (content ≥ 98.5%)

were purchased from Zhejiang Guobang Pharmaceutical, Zhe-

jiang, China. Acetonitrile (HPLC grade), sodium sulfate (chem-ically pure), hydrochloric acid (chemically pure), n-hexane(chemically pure), and tetrabutylammonium bromide (chemi-cally pure) were obtained from Shanghai Anpel Scientific In-strument, Shanghai, China.

Fish.—Nile Tilapia, weighing 50 ± 10.4 g (mean ± SD),were provided by a farm in Nantong, Jiangsu, China. All fishwere housed in fiberglass tanks (500 L) supplied with a constantflow (1 L/h) of aerated freshwater for 7 d before the experiments.The water quality was monitored daily and adjusted as neces-sary; the pH was maintained between 6.4 and 7.0, while watertemperature was adjusted to 25.0 ± 1.0◦C by means of coolingand heating devices. Fish (240 in total) were divided into fourgroups of 60 fish each. Enrofloxacin (10 mg) was treated with5 mL of acetic acid and then diluted with 95 mL of distilled wa-ter to give a stock solution of 0.1 mg ENR/mL. Flexible tubesof approximately 15 cm in length were connected with the in-jectors (1 mL). Based on the weights of the individual fish, theappropriate amount of ENR solution was loaded into the injec-tors. By inserting the flexible tube through the esophagus intothe fish’s stomach carefully, each group of fish was treated witha different dose of ENR (0, 20, 40, or 80 mg/kg body weight) byoral gavage. The procedure was finished within a few seconds.The fish not treated with ENR served as the control group. Tenfish were sacrificed at each time point (1 and 30 min, and 3, 6,12, and 48 h) after drug administration. All experiments wereperformed in accordance with the guidelines of the Regulationon Animal Experimentation and were approved by the StateScientific and Technological Commission.

Isolation of RNA and qRT-PCR.—At indicated time pointsafter drug administration, fish were sacrificed, and the totalRNA was extracted from liver or kidney tissues with a TRI-zol RNA extraction kit (Invitrogen, Carlsbad, California) ac-cording to the manufacturer’s protocol. To ensure the extractedRNA was integrated enough for the following experiments, theoptical density (OD) ratios, OD260/OD280, of extracted totalRNA were measured by a spectrophotometer (model DU800,Beckman Coulter, Indianapolis, Indiana.). The RNA purity wasevaluated by 1% gel electrophoresis. Single-stranded comple-mentary DNA (cDNA) was generated by reverse transcriptionusing ReverTra Ace-α-TM (Toyobo, Osaka, Japan). Quantita-tive real-time PCR (qRT-PCR) was performed with 0.4 µL ofcDNA, 0.2 µL of forward primer (10 µmol/L), 5 µL of reverseprimer (10 µmol/L), and 5 µL of SYBRGreen PCR Master Mix(QPK-201; Toyobo) in a Bio-Rad iQ5 real-time PCR instru-ment (Hercules, California). The PCR primers were designedby Primer Premier 5.0 software (PREMIER Biosoft, Palo Alto,California) asfollows: for the P-gp gene, P-gp forward (F) (5′-CACCTGGACGTTACCAAAGAAGATATA-3′) and P-gp re-verse (R) (5′-TCACCAACCAGCGTCTCATATTT-3′); for theβ-actin gene, β-actin F (5′-CCTCACCCTCAAGTACCCCAT-3′) and β-actin R (5′-TTGGCCTTTGGGTTGAGTG-3′). Theβ-actin gene was used as a control for the analysis. The PCRfor P-gp was performed in a final reaction volume of 10 µL

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4

P-GP AND ENROFLOXACIN METABOLISM 61

using the following conditions: a preheating cycle at 94◦C for3 min, followed by 35 cycles at 94◦C for 30 s, 51.5◦C for 30 s,and 72◦C for 30 s, and finally elongation at 72◦C for 10 min.The PCR conditions for β-actin were similar, except that the an-nealing temperature was 49.6◦C. The amplified products wereidentified by agarose gel electrophoresis. The amount of P-gpmessenger RNA (mRNA) was normalized to β-actin expres-sion. Data were quantified from three independent experimentsfor each group.

The qRT-PCR assays were run in triplicate for each cDNAsample, and melting curves were drawn to monitor the qualityof amplicons and reactions. Amplification products were quan-tified by comparison to experimental threshold cycle (Ct) levels(Ct is defined as the PCR cycle where an increase in fluores-cence over the background level first occurs). A standard curvewas generated for each primer pair based on known quantitiesof cDNA (10-fold serial dilutions corresponding to cDNA tran-scribed from 100 to 0.01 ng of total RNA). The relative mRNAexpression was calculated using the 2��Ct equation and β-actinwas used as an internal control for the PCRs.

Evaluation of tissue drug concentrations.—Tissue drug con-centrations were determined by HPLC. Tissue samples wereextracted with acetonitrile. After centrifugation at 4,500 rpmfor 5 min, the supernatants were carefully collected and com-pletely mixed with n-hexane. The aqueous phase of the mixturewas removed and the organic phase was evaporated using a ro-tary evaporator (Heidolph Instruments, Schwabach, Germany).The residues were dissolved in the mobile phase (acetonitrile: 0.01 M tetrabutyl ammonium bromide, pH 3.1 [5:95 v/v]),and the solution was then filtered through a 0.45-µm filter. AnHPLC (Agilent, Santa Clara, California) instrument equippedwith a fluorescence detector and a reversed-phase C18 column(150 × 4.6 mm) was used at a flow rate of 1.5 mL/min anda column temperature of 40◦C. The detection wavelength wasset at 280 nm, and 10 µL of sample solution was loaded. Stan-dard solutions of ENR and CIP (0.2–10 µg/mL) were prepared,and standard curves were drawn with the drug concentration(µg/mL) on the x-axis and peak area on the y-axis. Drug con-centrations in tissue samples were determined by the externalstandard method. Data were calculated from five independentexperiments for each group. To certify the validation of theHPLC method, ENR and CIP standard solutions (5 mg/L) wereeach added into blank tissue samples. The recoveries and relativeSD-values were determined in the intraday (five replicates in 1d) and the interday (5 d). The main pharmacokinetic parameterssuch as the peak concentration (Cmax) and the time for the drugto reach the peak level (Tmax) were examined. In addition, thearea under the curve (AUC0-t) was calculated by the computerprogram Drug and Statistics (DAS version 3.0; Center for DrugClinical Evaluation, Wannan Medical College, China).

Statistical analysis.—Statistical analyses were performedwith SPSS version 16.0 software (SPSS, Chicago, Illinois) anddata were presented as the mean ± SD. Significant differencesbetween means were determined using one-way ANOVA and

the Duncan’s multiple range test for comparison with signifi-cance level established at P < 0.05.

RESULTS

P-gp Gene Expression Levels in Liver and Kidney Tissuesafter ENR Administration

In the present study, P-gp fragments (127 bp) were success-fully amplified in both liver and kidney of Nile Tilapia by usingspecific primers (data not shown). The baseline P-gp gene ex-pression levels in liver or kidney of Nile Tilapia were relativelylow. However, administration of 40 mg/kg ENR significantlyenhanced the P-gp gene expression levels 30 min after drugadministration (P < 0.05) (Figure 1a, b). In addition, the P-gpgene expression levels in liver and kidney both peaked 3 h afterdrug administration and then gradually decreased (Figure 1a, b).Furthermore, 3 h after a single oral administration of ENR (0,20, 40, or 80 mg/kg), the P-gp gene expression levels increasedin a dose-dependent manner (Figure 2a, b). Moreover, the P-gp

FIGURE 1. Quantification of P-gp gene expression and ENR residue amountsin (upper) liver and (lower) kidney of Nile Tilapia at indicated time points (0 min[control], 30 min, 3 h, 6 h, 12 h, or 48 h) after a single oral administration of ENR(40 mg/kg). An asterisk indicates a significant difference (P < 0.05) comparedwith control.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4

62 HU ET AL.

FIGURE 2. Quantification of P-gp gene expression in (upper) liver and (lower)kidney of Nile Tilapia 3 h after a single oral administration of ENR (0 [control],20, 40, or 80 mg/kg). An asterisk (*) indicates a significant difference (P <

0.05) compared with control.

gene expression levels in the kidney were significantly higherthan those in the liver (P < 0.05).

Drug Concentration in Liver and Kidney TissuesTissue concentrations of ENR and CIP were determined by

HPLC analysis. The peaks of ENR and CIP could be sepa-rated properly with a stable baseline. The retention times forENR and CIP were 5.50 min and 6.38 min, respectively, and

the standard curves were y = 5.524x − 0.135 (r2 = 0.998)and y = 3.550x − 0.226 (r2 = 0.999), respectively, within aconcentration range of 0.2–10 µg/mL. By using the externalstandard method, ENR and CIP residue concentrations weredetermined with a limit of detection of 1 µg/kg. The averagerecoveries were 70.01–80.52% and the relative SD-values were1.20–7.92% for ENR, while the recoveries were 65.01–83.43%and SDs were 2.63–5.91% for CIP (Table 1). After a singleoral administration of ENR at different dosages (20, 40, or80 mg/kg), the concentrations of ENR and CIP in liver andkidney were examined. The residues in the two tissues werefitted to the open two-compartment model using DAS phar-macokinetic software. The pharmacokinetic parameters derivedfrom the compartment model are shown in Tables 2 and 3. Thedistribution and elimination half-lives (T1/2α and T1/2β), totalbody clearance (Cl), and apparent volume of distribution (Vd)of ENR in Nile Tilapia after a single oral dose (40 mg/kg) were0.14 h, 19.19 h, 636.44 kg·h−1·kg−1, and 4,012.21 kg/kg in liver,while the values were 0.36 h, 40.12 h, 213.82 kg·h−1·kg−1, and3,372.12 kg/kg in kidney. The distribution and elimination half-lives, total body clearance, and apparent volume of distributionof CIP showed the similar tendency in liver and kidney. Themetabolic rate of ENR and CIP at different concentrations inliver was more rapid than that in kidney. In addition, the Cmax

of ENR in Nile Tilapia after single oral doses of 20, 40, or80 mg/kg were 1.97, 4.47, and 39.44 mg/kg, respectively, inliver, and the respective values in kidney were 0.93, 2.56, and28.38 mg/kg. The Cmax of CIP residues in liver were 0.84, 1.33,and 2.80 mg/kg and in kidney were 0.63, 0.88, and 2.38 mg/kg.The metabolism of ENR and CIP in the liver and kidney gradu-ally increased in a dose-dependent pattern.

Relationship between P-gp Expression and ENRMetabolism

As mentioned above, the P-gp gene expression levels in liverand kidney were significantly increased 30 min after ENR ad-ministration of 40 mg/kg; they peaked at 3 h and then gradu-ally decreased. Normalized to β-actin, the P-gp gene expressionmaximum could reach 1.38 ± 0.02 in liver and 1.40 ± 0.01in kidney (Figure 1a, b). A similar trend was detected for ENRresidues in the kidney, whereas ENR residues in the liver peaked3 h after drug application, which was slightly earlier than in the

TABLE 1. Recoveries (mean ± SD) of ENR and CIP from the liver and kidney of Nile Tilapia. The values were the average of five batches of samples (n = 5)treated with ENR or CIP at concentrations of 5 mg/L.

Recovery (%) Relative SD (%)

Tissue Intraday Interday Intraday Interday

Liver ENR 72.76 ± 1.63 70.01 ± 1.43 5.67 7.92CIP 65.01 ± 4.93 70.27 ± 1.79 4.52 5.91

Kidney ENR 80.52 ± 1.03 78.92 ± 0.73 1.20 4.93CIP 83.43 ± 3.27 78.23 ± 2.03 2.63 3.73

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4


TABLE 2. Main pharmacokinetic parameters of ENR in Nile Tilapia after a single oral dose (0, 20, 40, or 80 mg/kg). Abbreviations are as follows: Cmax =maximum concentration; Tmax = peak time; T1/2α = distribution half-life of the drug; T1/2β = elimination half-life of the drug; AUC = area under the curve; Cl= total body clearance of the drug; Vd = extensive apparent volume of the central compartment; ND = not detected.

ENR dose (mg/kg)

0 20 40 80

Pharmacokinetic parameter Liver Kidney Liver Kidney Liver Kidney Liver Kidney

Cmax (mg/kg) ND ND 1.97 0.93 4.47 2.56 39.44 28.38Tmax (h) ND ND 2.11 2.89 2.00 3.02 2.33 3.12T1/2α (h) 0.10 0.25 0.14 0.36 0.45 0.61T1/2β (h) 8.22 19.82 19.19 40.12 39.03 54.33AUC0—240 (mg·h−1·kg−1) 30.32 49.34 70.33 163.22 180.32 282.98AUC0-∞ (mg·h−1·kg−1) 31.11 52.34 77.85 186.56 193.32 302.31Cl (kg·h−1·kg−1) 1,014.89 462.32 636.44 213.82 129.83 107.62Vd (g/g) 5,680.82 4,972.12 4,012.21 3,372.12 2,687.21 1,883.76

kidney (Figure 1a, b). At 3 h after drug application, the maxi-mum concentration of ENR reached 4.39 ± 0.10 mg/kg in liverand 2.65 ± 0.61 mg/kg in kidney, while the concentrations ofCIP in liver and kidney were 0.31 ± 0.08 mg/kg and 0.27 ±0.04 mg/kg, respectively.

DISCUSSIONPermeability glycoprotein, a plasma membrane ATP-binding

cassette transporter, is responsible for the elimination of drugcomponents in vivo, and the P-gp level in aquatic animals is as-sociated with the extent of chemical contamination (Bard et al.2002; Thomas et al. 2007). Using immunohistochemical anal-ysis, Kleinow et al. (2000) investigated the involvement of theP-gp in the liver and intestines during xenobiotic absorption andexcretion in catfish. In addition, Bard et al. (2002) demonstratedthat exposure to anthropogenic drugs and naturally occurringtoxins led to enhanced P-gp gene expression in the intertidal fish,

Anoplarchus purpurescens. Hochman et al. (2001) suggestedthat P-gp may share similar functions with the cytochrome P450enzyme. In mammals, P-gp regulates drug metabolism by me-diating the drug efflux process and thus influencing in vivo drugbioavailablility (Lown et al. 1997; Marzolini et al. 2004; Thomaset al. 2007). By comparing the main pharmacokinetic param-eters, such as Cmax, Tmax, and AUC(0-t), Thomas et al. (2007)indicated that P-gp activity in human blood could be modu-lated by grapefruit juice and herbal drugs. In addition, Lownet al. (1997) measured the pharmacokinetics of orally admin-istered cyclosporine on stable kidney transplant recipients andfound that intestinal P-gp played a significant role in the first-pass elimination of cyclosporine. With regard to aquatic animalssuch as fish, P-gp was activated along with cytochrome P450when animals were exposed to environmental contaminants, andtheir activities were related to drug resistance in fish (Bard et al.2002). A recent study reported that cytochrome P4501A partici-pates in the metabolism of difloxacin in hybridized Crucian Carp

TABLE 3. Main pharmacokinetic parameters of CIP in Nile Tilapia after a single oral dose (0, 20, 40, or 80 mg/kg). Abbreviation are as follows: Cmax =maximum concentration; Tmax = peak time; T1/2α = distribution half-life of the drug; T1/2β = elimination half-life of the drug; AUC = area under the curve; Cl= total body clearance of the drug; Vd = extensive apparent volume of the central compartment; ND = not detected.

CIP dose (mg/kg)

0 20 40 80

Pharmacokinetic parameter Liver Kidney Liver Kidney Liver Kidney Liver Kidney

Cmax (mg/kg) ND ND 0.84 0.63 1.33 0.88 2.80 2.38Tmax (h) ND ND 5.01 6.29 5.12 6.42 5.83 6.12T1/2α (h) 1.30 1.45 1.44 1.56 1.65 1.81T1/2β (h) 10.24 11.82 20.19 22.12 29.03 32.33AUC0—240 (mg·h−1·kg−1) 25.26 29.64 38.05 43.83 64.37 72.21AUC0-∞ (mg·h−1·kg−1) 25.67 27.83 39.04 42.42 80.10 92.01Cl (kg·h−1·kg−1) 874.21 352.45 531.11 202.12 109.38 96.32Vd (g/g) 4,261.72 3,852.32 3,281.11 3,012.32 2,012.32 1,473.65

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4

64 HU ET AL.

Carassius carassius (Fu et al. 2010). Nonetheless, the effect ofP-gp in the drug metabolism of aquatic animals is still poorlyunderstood. The liver and kidneys are vital organs that servecritical roles for drug accumulation and efflux. The drug residuein liver and kidney could give rise to the changes of some proteininvolved in the metabolism. In the present study, we found thatP-gp gene expression in liver and kidney was greatly enhanced30 min after 40 mg/kg ENR administration, peaked 3 h afterdrug feeding, and then gradually decreased, a trend that waspositively correlated with the amount of ENR and CIP residuein these organs (Figure 1). Moreover, the upregulation of P-gpgene expression levels in the liver and kidneys depended on thedose of ENR (Figure 2). The bioavailability (Cmax, Tmax, andAUC) of ENR and the residue of CIP seemed relative with theP-gp gene expression. Furthermore the oral dosage of ENR has astrong influence on the elimination efficiency (T1/2β, Cl, and Vd)and the gene expression of P-gp. These observations suggest thatP-gp gene expression may correlate with the ENR metabolismprocess in the liver and kidneys of Nile Tilapia. It is possiblethat P-gp alters the distribution of ENR in organs (Kemper et al.2004). As a similar example, difloxacin (a chemical agent of thefluoroquinone family) is confirmed as the substrate of CYP1Ain hybridized Prussian carp (Fu et al. 2010). On the other hand,P-gp may accelerate the efflux process of drugs (Pavek et al.2003). Permeability glycoprotein-modulated ENR eliminationcould serve two functions: (1) direct clearance of ENR so thatsystemic exposure to ENR is minimized, and (2) removal ofENR from the cytosol, preventing the accumulation of highconcentrations of ENR within cells (Hochman et al. 2001).

In summary, our study demonstrates that P-gp gene may beinvolved in the ENR metabolism process in Nile Tilapia whileENR may be a substrate for P-gp. Furthermore, CIP residue maybe indicated by P-gp gene expression. It offers a novel strategyfor promoting drug bioavailability and reducing the risks of drugresidues in animals.

ACKNOWLEDGMENTSThis study was supported by the 863 Program (Grant

2011AA10A216), the Special Fund for Agro-scientific Re-search in the Public Interest (Grant 201203085), Shanghai Uni-versity Knowledge Service Platform, and the Open Fund of theKey Laboratory of Marine and Estuarine Fisheries Resourcesand Ecology, Ministry of Agriculture (Grant Kai-09-06). Thereis no conflict of interest related to this manuscript.

REFERENCESBard, S. M., B. R. Woodin, and J. J. Stegeman. 2002. Expression of P-

glycoprotein and cytochrome p450 1A in intertidal fish (Anoplarchus pur-purescens) exposed to environmental contaminants. Aquatic Toxicology60:17–32.

Bowser, P. R., G. A. Wooster, J. St Leger, and J. G. Babish. 1992. Pharmacoki-netics of enrofloxacin in fingerling rainbow trout (Oncorhynchus mykiss).Journal of Veterinary Pharmacology and Therapeutics 15:62–71.

Bresolin, T., M. de Freitas Rebelo, and A. Celso Dias Bainy. 2005. Expres-sion of PXR, CYP3A and MDR1 genes in liver of Zebrafish. ComparativeBiochemistry and Physiology C 140:403–407.

Dalsgaard, I., and J. Bjerregaard. 1991. Enrofloxacin as an antibiotic in fish.Acta Veterinaria Scandinavica Supplementum 87:300–302.

della Rocca, G., A. Di Salvo, J. Malvisi, and M. Sello. 2004. The disposition ofenrofloxacin in Seabream (Sparus aurata L.) after single intravenous injectionor from medicated feed administration. Aquaculture 232:53–62.

DOCE (Diario Oficial de las Comunidades Europeas). 1990. Por el que se es-tablece un procedimiento comunitario de fijacion de los limites maximos deresiduos de medicamentos veterinarios en los alimentos de origen animal.[Laying down a community procedure for the establishment of maximumresidue limits of veterinary products in feed of animal origin.] DOCE Regu-laion 2377/90(18 August 1990):L224.

Doi, A. M., E. Holmes, and K. M. Kleinow. 2001. P-glycoprotein in the cat-fish intestine: inducibility by xenobiotics and functional properties. AquaticToxicology 55:157–170.

Fang, X., X. Liu, W. Liu, and C. Lu. 2012. Pharmacokinetics of enrofloxacin inallogynogenetic Silver Crucian Carp, Carassius auratus gibelio. Journal ofVeterinary Pharmacology and Therapeutics 35:397–401.

Fu, G. H., X. L. Yang, H. X. Zhang, W. J. Yu, and K. Hu. 2010. Effects ofcytochrome P450 1A substrate (difloxacin) on enzyme gene expression andpharmacokinetics in Crucian Carp (hybridized Prussian Carp). EnvironmentalToxicology and Pharmacology 31:307–313.

Hemmer, M. J., L. A. Courtney, and L. S. Ortego. 1995. Immunohistochem-ical detection of P-glycoprotein in teleost tissues using mammalian poly-clonal and monoclonal antibodies. Journal of Experimental Zoology 272:69–77.

Hernandez, M., C. Aguilar, F. Borrull, and M. Calull. 2002. Determination ofciprofloxacin, enrofloxacin and flumequine in pig plasma samples by capillaryisotachophoresis–capillary zone electrophoresis. Journal of ChromatographyB 772:163–172.

Hochman, J. H., M. Chiba, M. Yamazaki, C. Tang, and J. H. Lin. 2001. P-glycoprotein-mediated efflux of indinavir metabolites in Caco-2 cells ex-pressing cytochrome P450 3A4. Journal of Pharmacology and ExperimentalTherapeutics 298:323–330.

Hughes, A. L. 1994. Evolution of the ATP-binding-cassette transmembranetransporters of vertebrates. Molecular Biology and Evolution 11:899–910.

Intorre, L., S. Cecchini, S. Bertini, A. M. Cognetti Varriale, G. Soldani, andG. Mengozzi. 2000. Pharmacokinetics of enrofloxacin in the seabass (Dicen-trarchus labrax). Aquaculture 182:49–59.

Ito, S., C. Woodland, P. A. Harper, and G. Koren. 1993. P-glycoprotein-mediatedrenal tubular secretion of digoxin: the toxicological significance of the urine-blood barrier model. Life Sciences 53:25–31.

Kemper, E. M., W. Boogerd, I. Thuis, J. H. Beijnen, and O. van Tellingen. 2004.Modulation of the blood-brain barrier in oncology: therapeutic opportunitiesfor the treatment of brain tumours? Cancer Treatment Reviews 30:415–423.

Keppler, C., and A. H. Ringwood. 2001. Expression of P-glycoprotein in thegills of oysters, Crassostrea virginica: seasonal and pollutant related effects.Aquatic Toxicology 54:195–204.

Kleinow, K. M., A. M. Doi, and A. A. Smith. 2000. Distribution and inducibilityof P-glycoprotein in the catfish: immunohistochemical detection using themammalian C-219 monoclonal. Marine Environmental Research 50:313–317.

Lee, S. D., T-Osei, A.-Jo, and K. M. Wasan. 2013. Dose-dependent targeted sup-pression of P-glycoprotein expression and function in Caco-2 cells. MolecularPharmaceutics 10:2323–2330.

Liang, J. P., J. Li, F. Z. Zhao, P. Liu, and Z. Q. Chang. 2012. Pharmacokineticsand tissue behavior of enrofloxacin and its metabolite ciprofloxacin in Tur-bot Scophthalmus maximus at two water temperatures. Chinese Journal ofOceanology and Limnology 30:644–653.

Lown, K. S., R. R. Mayo, A. B. Leichtman, H. L. Hsiao, D. K. Turgeon, P.Schmiedlin-Ren, M. B. Brown, W. Guo, S. J. Rossi, L. Z. Benet, and P.B. Watkins. 1997. Role of intestinal P-glycoprotein (mdr1) in interpatient

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4


variation in the oral bioavailability of cyclosporine. Clinical Pharmacologyand Therapeutics 62:248–260.

Ma, Y. F. 2008. cDNA cloning and expression analysis of P-glycoprotein inZebrafish. Master’s thesis. Southwest University, Chongqing, China.

Marzolini, C., E. Paus, T. Buclin, and R. B. Kim. 2004. Polymorphisms in hu-man MDR1 (P-glycoprotein): recent advances and clinical relevance. ClinicalPharmacology and Therapeutics 75:13–33.

Miller, D. S. 1993. Daunomycin secretion by killifish renal proximal tubules.American Journal of Physiology 269:R370–R379.

Nouws, J. F., J. L. Grondel, A. R. Schutte, and J. Laurensen. 1988. Pharmacoki-netics of ciprofloxacin in carp, African catfish and Rainbow Trout. VeterinaryQuarterly 10:211–216.

Pavek, P., F. Staud, Z. Fendrich, H. Sklenarova, A. Libra, M. Novotna, M.Kopecky, M. Nobilis, and V. Semecky. 2003. Examination of the functionalactivity of P-glycoprotein in the rat placental barrier using rhodamine 123.Journal of Pharmacology and Experimental Therapeutics 305:1239–1250.

Rosenberg, M. F., G. Velarde, R. C. Ford, C. Martin, G. Berridge, I. D. Kerr,R. Callaghan, A. Schmidlin, C. Wooding, K. J. Linton, and C. F. Higgins.2001. Repacking of the transmembrane domains of P-glycoprotein during thetransport ATPase cycle. EMBO Journal 20:5615–5625.

Saeki, T., K. Ueda, Y. Tanigawara, R. Hori, and T. Komano. 1993. HumanP-glycoprotein transports cyclosporin A and FK506. Journal of BiologicalChemistry 268:6077–6080.

Sousa, E., A. Palmeira, A. S. Cordeiro, B. Sarmento, D. Ferreira, R. T. Lima,M. H. Vasconcelos, and M. Pinto. 2013. Bioactive xanthones with effect onP-glycoprotein and prediction of intestinal absorption. Medicinal ChemistryResearch 22:2115–2123.

Stoffregen, D. A., G. A. Wooster, P. S. Bustos, P. R. Bowser, and J. G. Babish.1997. Multiple route and dose pharmacokinetics of enrofloxacin in juve-nile Atlantic Salmon. Journal of Veterinary Pharmacology and Therapeutics20:111–123.

Thomas, V., K. Ramasamy, R. Sundaram, and A. Chandrasekaran. 2007. Ef-fect of honey on CYP3A4 enzyme and P-glycoprotein activity in healthyhuman volunteers. Iranian Journal of Pharmacology and Therapeutics 6:171–176.

Wack, A. N., B. KuKanich, E. Bronson, M. Mary Denver. 2012. Pharmacoki-netics of enrofloxacin after single dose oral and intravenous administrationin the African penguin (Spheniscus demersus). Journal of Zoo and WildlifeMedicine 43:309–316.

Wessler, J. D., L. T. Grip, J. Mendell, and R. P. Giugliano. 2013. The P-glycoprotein transport system and cardiovascular drugs. Journal of the Amer-ican College of Cardiology 61:2495–502.

Zhang, Y. Y., Y. Q. Huang, F. L. Zhai, R. Du, Y. D. Liu, and K. Q. Lai.2012. Analyses of enrofloxacin, furazolidone and malachite green in fishproducts with surface-enhanced Raman spectroscopy. Food Chemistry 135:845–850.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:49

15

June

201

4



Investigation of the Link between BroodstockInfection, Vertical Transmission, and Prevalence ofFlavobacterium psychrophilum in Eggs and Progeny ofRainbow Trout and Coho SalmonAmy Longac, Douglas R. Callb & Kenneth D. Caina

a Department of Fish and Wildlife Sciences and the Aquaculture Research Institute,University of Idaho, Post Office Box 441136, Moscow, Idaho 83844-1136, USAb Paul G. Allen School for Global Animal Health, Washington State University, Pullman,Washington 99164-7090, USAc Present address: Oregon Department of Fish and Wildlife, Department of Microbiology,Oregon State University, Nash Hall, Room 220. 97331-3804, Corvallis, Oregon, USAPublished online: 04 Jun 2014.

To cite this article: Amy Long, Douglas R. Call & Kenneth D. Cain (2014) Investigation of the Link between BroodstockInfection, Vertical Transmission, and Prevalence of Flavobacterium psychrophilum in Eggs and Progeny of Rainbow Trout andCoho Salmon, Journal of Aquatic Animal Health, 26:2, 66-77, DOI: 10.1080/08997659.2014.886632







http://dx.doi.org/10.1080/08997659.2014.886632




ARTICLE

Investigation of the Link between Broodstock Infection,Vertical Transmission, and Prevalence of Flavobacteriumpsychrophilum in Eggs and Progeny of Rainbow Troutand Coho Salmon

Amy Long1

Department of Fish and Wildlife Sciences and the Aquaculture Research Institute, University of Idaho,Post Office Box 441136, Moscow, Idaho 83844-1136, USA

Douglas R. CallPaul G. Allen School for Global Animal Health, Washington State University, Pullman,Washington 99164-7090, USA

Kenneth D. Cain*Department of Fish and Wildlife Sciences and the Aquaculture Research Institute, University of Idaho,Post Office Box 441136, Moscow, Idaho 83844-1136, USA

AbstractThe etiological agent of bacterial coldwater disease (BCWD), Flavobacterium psychrophilum, can be transmitted

both vertically and horizontally. Outbreaks of BCWD can result in significant losses in salmonid aquaculture.Reduction of outbreaks in fry may be possible through implementation of a management strategy in which progenyof heavily infected broodstock are culled from the general population. Diagnostic assays to quantify F. psychrophilumconcentrations in tissue samples and confirm presence of the bacterium in ovarian fluid have been previously validated.In the current study, these assays were used to screen 60 female Rainbow Trout Oncorhynchus mykiss and 60 femaleCoho Salmon O. kisutch broodstock at two aquaculture facilities. Eyed eggs from 10 female broodstock (five fishfrom each facility) exhibiting graded levels of infection were transferred to the University of Idaho and monitoredthrough early life stages for the presence of F. psychrophilum. Female Rainbow Trout broodstock were not positivefor F. psychrophilum by enzyme-linked immunosorbent assay (ELISA) and prevalence was low in these progeny.However, ELISA optical density values for kidney correlated to F. psychrophilum prevalence in progeny (r = 0.938,P < 0.05) of Coho Salmon. Nested PCR on ovarian fluid was not a reliable indicator of vertical transmission in eitherspecies as broodstock ovarian fluid results did not correlate to F. psychrophilum prevalence in eyed eggs. Furtherresearch with these assays is necessary; however, results from this study indicate that broodstock screening may bea potential tool for evaluating F. psychrophilum infection levels, which could become an important component fordisease management.

Flavobacterium psychrophilum, the causative agent of bacte-rial coldwater disease (BCWD) and rainbow trout fry syndrome(RTFS), is a pathogen of concern in the aquaculture industry

*Corresponding author: [email protected] address: Oregon Department of Fish and Wildlife, Department of Microbiology, Oregon State University, Nash Hall, Room 220,

Corvallis, Oregon 97331-3804, USA.Received December 19, 2012; accepted December 8, 2013

as outbreaks cause high mortality in economically importantspecies (Michel et al. 1999; Barnes and Brown 2011). Vaccineresearch is ongoing but there are no licensed vaccines available,

66

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4

SCREENING TO REDUCE BACTERIAL COLDWATER DISEASE IN FRY 67

which limits treatment options primarily to antibiotics. Reportsof resistance to oxytetracycline and increases in florfenicol min-imum inhibitory concentration have been reported for numer-ous F. psychrophilum strains (Bruun et al. 2000; del Cerro et al.2010; Henrıquez-Nunez et al. 2012). As such, there is a need forthe development of alternative management strategies to reduceBCWD outbreaks.

In addition to horizontal transmission (Madetoja et al. 2000),F. psychrophilum has been re-isolated from reproductive fluids,egg surfaces, and egg contents (Holt et al. 1993; Brown et al.1997a; Madetoja et al. 2002; Taylor 2004; Cipriano 2005; Mad-sen et al. 2005; Kumagai and Nawata 2011a), indicating thatvertical transmission may be an important factor in pathogendissemination. Quantification of F. psychrophilum infectionlevels in broodstock in relation to BCWD outbreaks in progenyhas only been attempted with ovarian fluid samples (Kumagaiand Nawata 2011a, 2011b), in part because of a lack of assaysthat can quantify infection levels in tissue samples. While apolyclonal enzyme-linked immunosorbent assay (ELISA) hasbeen developed for F. psychrophilum, cross-reactivity to otherFlavobacterium species has been reported (Rangdale and Way1995; Lorenzen and Olesen 1997; Mata and Santos 2001).With the recent development and commercial availability of amonoclonal antibody (MAb FL43) that is specific to a proteinon the outer surface of F. psychrophilum (Lindstrom et al.2009) and the validation of a capture ELISA that uses MAbFL43 (Long et al. 2012), quantification of infection levels inbroodstock tissues is now possible.

Renibacterium salmoninarum, the etiological agent of bacte-rial kidney disease (BKD), is another common fish pathogen thatcan be transmitted vertically. Historical losses in Pacific salmonstocks due to BKD have been as high as 80% of a given popu-lation (Evenden et al. 1993). Renibacterium salmoninarum canenter oogonia during oogenesis and survive in the yolk throughfertilization and hatch (Bruno and Munro 1986). Additionally,if R. salmoninarum levels are high in ovarian fluid, eggs canalso acquire the bacterium postovulation when eggs are held inthe body cavity prior to spawning (Evelyn et al. 1986a, 1986b;Pascho et al. 1991). By localizing in the egg, R. salmoninarumcan persist even after povidone-iodine disinfection similar toF. psychrophilum (Evelyn et al. 1984, 1986b; Lee and Evelyn1989; Brown et al. 1997a; Kumagai and Nawata 2010b).

Reduction of BKD outbreaks in recent years has been ac-complished through a management strategy that incorporateserythromycin injections of individual female broodstock priorto spawning (Brown et al. 1990; Moffitt 1991; Haukenes andMoffitt 2002) and measuring antigen levels in individual brood-stock by a polyclonal ELISA (Pascho and Mulcahy 1987). Eggsfrom broodstock with high antigen levels are either culled orprogeny are reared in isolation to reduce the spread of disease inthe general population. Due to the direct relationship betweenantigen concentration and ELISA optical density (OD) valuesfor a kidney tissue sample (Pascho and Mulcahy 1987; Turagaet al. 1987), this strategy has been applied routinely and is ef-

fective at reducing disease outbreaks (Elliott et al. 1989; Paschoet al. 1991; Meyers et al. 2003; Munson et al. 2010). With thisin mind, we surmise that a similar management strategy maybe successful for the control of BCWD in populations wherevertical transmission is suspected.

To implement such a strategy, it is important to demonstrate arelationship between F. psychrophilum infection levels in femalebroodstock and the incidence of disease outbreaks in progeny. Inthis study, broodstock infection levels were measured at two dif-ferent hatcheries: a commercial Rainbow Trout Oncorhynchusmykiss facility and a coastal Coho Salmon O. kisutch hatcherywhere broodstock returned naturally following 1–3 years in thePacific Ocean. We hypothesized that antigen levels in brood-stock can predict disease outbreaks or pathogen prevalence inprogeny. Therefore, the objectives of this study were to (1)screen female broodstock at each facility and select egg groupsto monitor the total prevalence of F. psychrophilum in eggs andfry from each family and (2) correlate progeny prevalence to fe-male broodstock infection levels during both early rearing anda controlled stress experiment.

METHODS

Experimental DesignFemale broodstock were screened for F. psychrophilum at

two facilities. Broodstock at hatchery 1, a private RainbowTrout facility in Washington, were sampled in winter 2010.Broodstock are reared on site at this facility and BCWD out-breaks are common in fingerlings. Broodstock at hatchery 2, acoastal Coho Salmon hatchery in Washington, were sampled infall 2010. Broodstock naturally return to this facility from thePacific Ocean. Outbreaks of BCWD are more likely to occurwithin 1 to 2 weeks of fish going on feed. Broodstock are notinjected with antibiotics prior to spawning at either facility. Kid-ney, spleen, and ovarian fluid samples were collected from 60female broodstock at each facility (120 fish total).

The experimental design for this study is shown in Figure 1.While broodstock samples were being evaluated for F. psy-chrophilum, fertilized eggs from each sampled broodstock weremaintained in separate incubators at the facility. Five femalebroodstock with varying infection levels were identified at eachfacility, and eyed eggs from the selected females were trans-ported to the University of Idaho. Progeny were monitored forthe presence of F. psychrophilum through hatch and early rear-ing by bacteriological culture and nested PCR. A controlledstress was then applied to progeny from each family to inducea BCWD outbreak with subsequent monitoring by nested PCRand bacteriological culture. All animal handling and experimen-tal procedures were approved by the University of Idaho Insti-tutional Animal Care and Use Committee (protocol 2010–38).

Broodstock TestingTo determine the infection level of F. psychrophilum in in-

dividual broodstock, kidney and spleen samples were tested by

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4

68 LONG ET AL.

FIGURE 1. Experimental design. Female broodstock at each facility werescreened for F. psychrophilum by ELISA (kidney and spleen tissue), nestedPCR (ovarian fluid), and bacteriological culture (kidney tissue, spleen tissue,and ovarian fluid). Fertilized eggs from each sampled broodstock were main-tained in separate incubators at the facility. Five female broodstock with varyinginfection levels were identified at each facility and eyed eggs from the selectedfemales were transported to the University of Idaho. Progeny were monitoredfor F. psychrophilum presence through hatch and early rearing by bacteriolog-ical culture and nested PCR. A controlled stress was then applied to progenyfrom each family to induce a BCWD outbreak with subsequent monitoring bynested PCR and bacteriological culture.

ELISA and bacteriological culture while ovarian fluid was testedby nested PCR and bacteriological culture using methodologiespreviously described by Long et al. (2012). Different assayswere used for tissue and ovarian fluid samples as the ELISA is

not suitable for ovarian fluid because of high background levels(data not shown). For bacteriological culture, kidney and spleentissue and ovarian fluid (100 µL) were inoculated on tryptoneyeast extract salts (TYES; 0.4% tryptone, 0.04% yeast extract,0.05% calcium chloride, 0.05% magnesium sulfate, pH 7.2) agar(Holt 1988).

Ovarian fluid samples were obtained by pouring off approxi-mately 3 mL of ovarian fluid from eggs collected in a paper cup.Samples were either inoculated immediately or transported onice to the University of Idaho within 24 h of collection wherethey were then inoculated. Ovarian fluid samples were thenstored at −80◦C.

Approximately 1 g each of head kidney and spleen was asep-tically collected from fish from each broodstock. As with ovarianfluid, all tissue samples were either inoculated immediately onTYES agar or transported to the University of Idaho within 24 hof collection and then inoculated. Tissue samples were split intosubsamples of 0.5 g each and stored at −80◦C until tested byELISA. All ELISA plates were washed by hand. Samples wererun in duplicate and any samples with a CV (100 × SD/mean)greater than 10% were rerun. The positive–negative thresholdwas set at the OD of the negative control plus 2 SDs.

Family SelectionProgeny from five female broodstock fish were chosen from

each facility for further study by reviewing the assay results forindividual broodstock (Table 1). As none of the kidney samplesfrom fish from the Rainbow Trout broodstock were positive byELISA [threshold value: 0.066 + (2 × 0.0077 SD) = 0.081], weselected three families (F61, F70, and F74) from broodstockthat were positive [threshold value: 0.084 + (2 × 0.014 SD) =0.112] by the spleen ELISA. The fourth family selected wasfrom a broodstock that had F. psychrophilum re-isolated fromboth spleen and ovarian fluid (F87). Finally, one family (F54)was selected to serve as a putative negative control becausethe female fish tested negative by all assays. In contrast to theRainbow Trout, all kidney samples from female Coho Salmonbroodstock were positive [threshold value: 0.077 + (2 × 0.002SD) = 0.081]. All ovarian fluid samples from this populationwere positive by the nested PCR. As such, we selected progenyfrom the females with the lowest (F6) and highest (F20) kidneyELISA OD value. The third family was taken from the onlyfemale broodstock in which F. psychrophilum was isolated fromovarian fluid by bacteriological culture (F27). The remainingtwo (F19 and F30) families were chosen because the femaleshad ELISA OD values for kidney that were near the medianOD value (0.105) for the population. Spleen tissue from F20,F27, and F30 was also positive by ELISA [threshold value:0.126 + (2 × 0.002) = 0.130].

Fish and Rearing ConditionsApproximately 1,000 eyed eggs from each family were

shipped in separate containers to the University of IdahoCollege of Natural Resources wet laboratory facility. Fish

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4


TABLE 1. Assay results for individual broodstock selected for further studies. The average ELISA OD value and SD are reported for each female. Asterisks (*)denote tissue samples that were greater than the positive–negative threshold (negative control plus two SDs). Minus (−) and plus ( + ) symbols indicate negativeand positive PCR result, respectively. The pound symbol (#) after a family number indicates the designated negative control for each trial.

Kidney ELISA Spleen ELISA Ovarian fluidTrial Family ODa (SD) ODb nested PCR Culture confirmed

Rainbow Trout F54# 0.071 (0.005) 0.082 (0.001) −F61 0.077 (0.006) 0.115* (0.016) − KidneyF70 0.058 (0.001) 0.124* (0.015) −F74 0.064 (0.002) 0.117* (0.013) −F87 0.066 (0.008) 0.09 (0) − Spleen, ovarian fluid

Coho Salmon F6# 0.094* (0.004) 0.109 (0.001) +F19 0.100* (0.004) 0.127 (0.009) +F20 0.133* (0.006) 0.131* (0.005) +F27 0.096* (0.001) 0.143* (0.007) + Ovarian fluidF30 0.127* (0.013) 0.140* (0.009) +

aPostive–negative threshold for kidney tissue samples in both trials was 0.081.bPositive–negative threshold for spleen tissue samples in the Rainbow Trout trial was 0.112. The threshold in the Coho Salmon trial was 0.130.

remained separated by family in all experiments. Upon arrival,eggs were disinfected in 100 ppm Ovadine (Western Chemicals,Ferndale, Washington) for 10 min. After disinfection, eggswere transferred to 6-L McDonald-type hatching jars (AquaticEcosystems, Apopka, Florida) and incubated until hatch. Afterhatching, fry were reared in 100-L troughs until fish reached1 g (Rainbow Trout) or 0.3 g (Coho Salmon). Fish were main-tained in flow-through systems using dechlorinated, specificpathogen-free municipal water in both trials during this lifestage. Eggs and fry were reared at the standard environmentaltemperature recommended for each species (13◦C for RainbowTrout, 8◦C for Coho Salmon). During the stress experiments,water temperature was maintained at 15◦C (Rainbow Trout) or11◦C (Coho Salmon).

Feeding began when approximately 50% of the fry hadreached the swim-up stage. Rainbow Trout were fed RangenTrout Production Starter Diet (Rangen, Buhl, Idaho) #0 crum-ble and feed size was increased based on manufacturer’s rec-ommendations. Coho Salmon were fed EWOS micro #0 feed(EWOS, Surrey, British Columbia) and feed size was increasedbased on manufacturer’s recommendations. During the progenymonitoring trials, fry were fed to satiation. In the Rainbow Troutstress experiments, fish were fed at a rate of 5% body weight/dthroughout the experiment. In the Coho Salmon stress trials,fish were fed at a rate of 2.4% body weight/d throughout theexperiment.

Progeny Monitoring TrialsTrial 1 (Rainbow Trout).—To determine the prevalence of

F. psychrophilum in Rainbow Trout eggs upon arrival at theUniversity of Idaho, a convenience sample of five eggs wasselected from each family immediately after the initial disinfec-tion in 100 ppm Ovadine. The selected eggs were disinfectedin 400 ppm Ovadine for 15 min, rinsed five times in sterile de-

ionized water, and material was then aseptically extracted usinga sterile 21-gauge needle (Brown et al. 1997a; Taylor 2004). Foreach family, material from within the five eggs was pooled ina 1.5-mL tube, 1 mL of sterile phosphate-buffered saline (PBS;pH 7.0) was added, and samples were vortexed briefly using avortex mixer (VWR, Radnor, Pennsylvania) set to 3,200 rpm.

A convenience sample of five fry was taken from each familyonce a week. Fish were euthanized in a lethal overdose of tri-caine methanesulfonate (MS-222; Argent Chemicals, Redmond,Washington). As with egg sampling, fish were first disinfected in400 ppm Ovadine for 15 min and then rinsed five times in sterilede-ionized water. Samples from each family were pooled priorto DNA extraction and homogenized in 1 mL of sterile PBSwith a Stomacher 80 Lab-Blender (Seward, Port Saint Lucie,Florida). In all, there were eight pools per family. By samplingday 57 (50 d posthatch), fry had reached a size where kidneyand spleen could be sampled directly for bacteriological culture.

Trial 2 (Coho Salmon).—In the second trial with CohoSalmon, sample size and frequency were increased and sam-ples were not pooled to allow us to better track changes inprevalence over time. Eggs were sampled from each of the fivefamilies immediately upon arrival using the same methodologyas in trial 1. Twice a week eight individuals were taken from eachfamily until fish weighed approximately 0.3 g (50 d posthatch).Eggs, sac fry, and swim-up fry were disinfected and sampledin the same manner as described above with the exception thatindividual samples were homogenized in 1 mL of sterile PBS.In all, 120 fish were sampled per family.

Bacteriological culture.—A subsample of egg material orfish homogenate (100 µL) was spread-plated onto TYES agarwith 5 µg/mL tobramycin (TYES-TB). Tobramycin can inhibitthe growth of yellow-pigmented bacteria (YPB) but does notinhibit F. psychrophilum (Kumagai et al. 2004). In trial 1, kidneyand spleen tissue from sampled fry was streaked on TYES-TB

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4

70 LONG ET AL.

using a sterile inoculating loop. All plates were incubated at15◦C for 72–96 h after which plates were checked for the pres-ence of YPB. As individual fish could have numerous yellow-pigmented colonies, four yellow-pigmented colonies wereselected, re-isolated from each plate, and subcultured for confir-mation as F. psychrophilum by Gram stain and PCR (see below).

DNA extraction.—To extract DNA from pooled egg materialin trial 1, samples (900 µL) were first centrifuged at 9,300 × gfor 5 min, the supernatant decanted, and the pellets stored at−20◦C until extraction (Taylor 2004). For the extraction ofDNA from egg material in trial 2, resuspended samples (900 µL)were centrifuged at 5,000 × g for 10 min, the supernatant wasdecanted, and DNA was extracted from pellets. Homogenizedfry were centrifuged at 380 × g for 3 min to pellet the spinal cordand skin, which would otherwise clog the column. All extrac-tions were carried out using the Qiagen DNeasy DNA extractionkit (Qiagen, Valencia, California) following the manufacturer’sinstructions for Gram-negative bacteria. To ensure complete tis-sue lysis, samples were incubated overnight at 55◦C. DNA wasstored at −20◦C until needed.

Stress ExperimentsTrial 1 (Rainbow Trout).—Stress experiments were con-

ducted at the University of Idaho’s Cold Water Research Lab-oratory (Moscow, Idaho) when Rainbow Trout fry were ap-proximately 1 g. Two stressors were selected for use: (1) gassupersaturation (chronic stress) and (2) gas supersaturation plusnet stress (acute stress). Fish were exposed to the two stressorsfor 56 consecutive days. For each family, there were triplicatechronic and acute stress tanks with 50 fish per tank. Tank volumewas approximately 60 L and the water source was well watertreated with ultraviolet light.

Net stress was accomplished by collecting fish in a net andholding the net in air for 20 s, two times per day. Well water wasused at this facility which, although naturally supersaturated,was not consistently supersaturated or at the levels desired. Toproduce supersaturated water with nitrogen gas (N2) concentra-tions consistently greater than 100%, well water was initiallypassed through a packed column to reduce the baseline nitrogensupersaturation and introduce oxygen. Water was then pumpedthrough a Venturi to the top of a 2-m column. Total gas pressurewas increased by pressurizing the column with an air blower to50 PSI (345 kPa). Water exited at the bottom of the column andwas distributed via a manifold to tanks. The supersaturated waterwas introduced to the tanks below the surface with a short lengthof vinyl tubing. Gas levels in the tanks were initially measureddaily using a PT4 Tracker total dissolved gas pressure meter(Point Four Systems, Coquitlam, British Columbia) and oncethe system was stabilized, measurements were taken weekly.The percent N2 plus argon saturation was back calculated usingbarometric pressure, temperature, and dissolved oxygen, whichwere measured concurrently.

As F. psychrophilum can be present on fish that show noclinical signs of disease, three fish per treatment (six per family)

were randomly sampled without replacement (one from eachtank) once a week for the duration of the experiment (n = 48).Fish were euthanized with a lethal overdose of MS-222. Mor-talities were monitored on a daily basis for 56 d and a minimumof 20% of the daily mortality was processed in the same manneras the weekly samples. For both random samples and mor-talities, kidney, liver, and spleen were streaked on TYES-TB.Nested PCR was also carried out on DNA extracted from wholefish homogenates using the same techniques as those used forscreening fry.

Trial 2 (Coho Salmon).—One week after the start of feeding,stress experiments were initiated with fry (mean weight, 0.3 g).Fry were stocked into circular flow-through tanks containingdechlorinated, pathogen-free municipal water. For each family,there were triplicate control and treatment tanks with 75 fish pertank. In this trial, two stressors were chosen to reflect conditionsin hatcheries thought to be linked to BCWD outbreaks in CohoSalmon: low flow and high rearing density. The values for lowflow and high rearing densities were modeled after conditionsobserved at the originating hatchery (C. Olson, Northwest IndianFisheries Commission, personal communication). To change thedensity but keep the same number of fish in each tank, the tankvolume was manipulated. Tank volume in the control group was17.6 L (density, 1.3 g/L) and flow rates were set at 2.5 L/min.In the treatment group, initial tank volume was 4.5 L (density,5 g/L) and flow rates were 0.2 L/min. Tank volume was increasedto 7 L and flow rates increased to 0.4 L/min in treatment tanks40 d after the start of the experiment to account for the increasein fish biomass. Inventory occurred weekly and feed rate wasadjusted accordingly.

To check populations for the presence of F. psychrophilum,two fish were removed randomly from each tank twice a weekwithout replacement and euthanized with a lethal overdose ofMS-222 (n = 180 fish per family). Fish samples were processedusing the same techniques as those used in trial 1. Mortalitieswere monitored on a daily basis for 64 d and a minimum of 20%of the daily mortality was sampled and processed in the samemanner as the biweekly samples.

DNA extraction.—In both stress experiments, fish were ho-mogenized in sterile PBS at a 1:1 (w/v) ratio. In trial 1, fishwere too large to homogenize by 28 d, and kidney and spleenwere dissected from each fish for all samples collected after thattime point. In trial 2, fish were prepared for DNA extractionin the same manner as for progeny monitoring samples. To ex-tract DNA, the Qiagen DNeasy blood and tissue extraction kitwas used following the manufacturer’s instructions for Gram-negative bacteria or tissue samples depending on sample type.

PCR AnalysisNested PCR.—Nested PCR of fish samples was performed

using the protocol published by Taylor (2004) with universal16S rDNA primers, Fp1 and Rp2, in the first round and primersspecific to F. psychrophilum, PSY1 and PSY2, in the secondround. Positive–negative reactions were confirmed by agarose

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4


gel electrophoresis. The DNA extracted from F. psychrophilum(strain CSF 259–93) served as the positive control. The risk ofcross-contamination was minimized by use of a dedicated hoodand pipettes. Furthermore, a no-template control (NTC) was runduring each round of PCR. If the NTC was positive, results werediscarded and the nested PCR was carried out again.

PCR of bacterial isolates.—For bacterial isolates from trial1, isolates were confirmed as F. psychrophilum by nested PCRusing DNA extracted from YPB as the template in the first round.For bacterial isolates from trial 2, isolated colonies from purecultures were used as the template in the PCR. The same runningconditions and concentrations as those used in the second roundof the nested PCR were used in a single-round PCR with theexception that the volume of water was increased to 29.75 µL.

Statistical AnalysisThe software, Minitab 16 (Minitab, State College, Pennsylva-

nia), was used for all statistical analysis (significance level, α =0.05). Pearson’s correlations were used to compare ELISA ODof broodstock kidney or spleen with the total F. psychrophilumprevalence in progeny for each family (number of nested PCRpositive samples / total number of samples) in the progeny mon-itoring trials. Normal distributions of the data were confirmedusing the Ryan–Joiner normality test. For the stress experi-ments we used Pearson’s correlations to compare total F. psy-chrophilum prevalence in progeny for each family as determinedby nested PCR with ELISA OD of broodstock kidney or spleen.

We used chi-square tests (two-way table) to test whetherthe proportion of positive nested PCR samples was differentbetween treatments and families. To determine whether differ-ences in cumulative percent mortality (CPM) during the stressexperiments were significant, a two-way ANOVA with treat-ment and family as the factors was used on ranked CPM values.Finally, differences in weights between the control and treat-ment tanks in the Coho Salmon stress experiment for the firstthree inventory dates were evaluated by a Student’s t-test. Forthe remaining two inventory dates, the data were not normally

TABLE 2. Nested PCR results from pooled samples taken during the progenymonitoring experiment in trial 1 (Rainbow Trout). Each pool consisted of fivefish that were externally disinfected and homogenized prior to DNA extraction.Pooled samples that tested positive are denoted by the plus ( + ) sign., negativesamples are denoted by the minus (−) sign. Hatch began on day 7 of samplingand was complete by day 9, and feeding began on day 18.

Sampling day

Family 1 9 16 23 29 36 44 51

F54 + − + + + − − +F61 + + + + − − − −F70 + − + + + − − −F74 + + − + + − − −F87 + + − + − − − −

distributed and differences in weight between the control andtreatment tanks were evaluated by a Mann–Whitney U-test.

RESULTS

Progeny MonitoringTrial 1 (Rainbow Trout).—There were eight pools per family

tested by nested PCR in this experiment. Pooled egg sampleswere positive for F. psychrophilum by nested PCR upon arrivalat the University of Idaho and prevalence varied throughoutthis period (Table 2). Although YPB growth was recorded forsamples taken on three different sampling days (days 36, 44,and 51), these isolates were not F. psychrophilum. There wasno correlation between prevalence of F. psychrophilum fromprogeny and broodstock kidney or spleen ELISA OD values(kidney, r = 0.25; spleen, r = −0.15).

Trial 2 (Coho Salmon).—Disinfected eggs sampled from fourfamilies (F6, F20, F27, and F30) were positive for F. psy-chrophilum by nested PCR upon arrival at the University ofIdaho. The proportion of positive samples tapered off throughhatch before increasing as the yolk sac was absorbed and feed-ing was initiated (Figure 2). There was a significant correlation

FIGURE 2. Proportion of sampled fish that tested positive for F. psychrophilum as determined by nested PCR in the Coho Salmon progeny monitoring trial. Aconvenience sample of eight fish was taken from each family at every sampling point. The asterisk (*) denotes the start of hatch and the pound symbol (#) denotesinitiation of feeding.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4

72 LONG ET AL.

FIGURE 3. Correlation between the five female broodstock infection levelsbased on kidney tissue ELISA OD values and the total progeny prevalence inthe Coho Salmon trial (n = 120). Error bars denote the SD of the ELISA ODvalue.

between the total proportion of positive samples and ELISAOD values for kidneys (r = 0.938; Figure 3), but there was nocorrelation between the total proportion of positive samples andELISA OD values for spleens (r = 0.335).

Even though YPB isolates with colony morphology similar tothat of F. psychrophilum, i.e., a “fried egg” with spreading mar-gins appearance, were observed on TYES-TB plates as early asthe first sampling day, F. psychrophilum was not re-isolated fromany samples throughout this period. There were 80 YPB isolatesfrom this trial, many of which had the typical colony morphol-ogy of F. psychrophilum. It is possible that F. psychrophilum didgrow on agar plates, but we were unsuccessful in our re-isolationattempts due to the high number of yellow-pigmented coloniespresent on the plates.

Stress ExperimentsTrial 1 (Rainbow Trout).—The percent N2 and argon satura-

tion of the tank water ranged between 101% and 111% through-out the experiment. Fish in both the chronic and acute treatmentsexhibited signs of chronic stress including frayed fins, petechialhemorrhaging, scale loss, and overinflated swim bladders. Theinteraction between family and treatment was significant as de-termined by two-way ANOVA (F4, 20 = 2.90). However, CPMvalues were low overall in each family and treatment. The high-est CPM (6%) was recorded in the F87 acute treatment. NestedPCR results showed that F. psychrophilum was present in 29.6%of all mortalities. Although 26% of the mortalities in this trialhad YPB growth on TYES-TB plates, none of the re-isolatedYPB could be confirmed as F. psychrophilum.

For the weekly random samples, no bacterial growth wasrecorded. Flavobacterium psychrophilum was detected bynested PCR in all families in both chronic and acute treatmentsthroughout the experiment and detection increased in individualfish over time (Figure 4). However, any differences in the totalnumber of positive samples in each family and treatment were

not significant (χ2 = 10.90, df = 12). There was no signifi-cant correlation between OD values for broodstock tissue andprogeny prevalence (kidney, r = −0.772; spleen, r = 0.190)

Trial 2 (Coho Salmon).—Lesions on the dorsal surface anderosion of the caudal peduncle were observed in three mortali-ties from treatment tanks, but a full-scale disease outbreak didnot occur. At every inventory date, the average weight of fishin the stress treatment was significantly lower than that of thecontrol treatment. Flow was interrupted on day 40 causing alow dissolved oxygen event that resulted in increased mortalityonly in treatment tanks, specifically for families F6 and F20.These families also had the highest CPM, 14% and 21%, re-spectively. The mean CPM in other families was less than 6%for both the control and stress treatment. The CPM was signif-icantly greater (two-way ANOVA: F1, 20 = 18.69) in treatmenttanks, which was likely due to the low dissolved oxygen event.Differences between families in either treatment were not sig-nificant (two-way ANOVA: F4, 20 = 1.78). Of the 41 mortalitiesthat could be tested by nested PCR, 29 (71%) were positive forF. psychrophilum.

Differences in the total number of samples that tested pos-itive by nested PCR between families in both treatments werenot significant (χ2 = 12.28, df = 12). In general, variation inthe proportion of samples that tested positive by nested PCRwas low throughout the experiment (Figure 5). The highest totalprevalence was observed in F20, which corresponds to the high-est ELISA OD value for broodstock kidneys. Family F6, whichhad the lowest ELISA OD value for kidney, had the lowest totalprevalence of F. psychrophilum in the stress experiments. Totalprevalence for each family is given in Table 3. There was no cor-relation between OD values for broodstock tissue and progenyprevalence (kidney, r = 0.611; spleen, r = 0.778).

Due to the large number of YPB isolates that were re-isolatedfrom individual fish sampled throughout this trial (>500), wewere unable to confirm them as F. psychrophilum by PCR.Yellow-pigmented bacterial growth was more frequently ob-served for individual samples taken from treatment tanks (117fish) than from samples taken from control tanks (62 fish).

DISCUSSIONFry infected with a pathogen acquired via vertical transmis-

sion can serve as the initial infection source in an otherwisehealthy population and disseminate the pathogen to uninfectedfish resulting in an epizootic (Reno 1998). Reducing the rateof vertical transmission has become a critical aspect of diseasemanagement for fish species susceptible to vertically transmittedpathogens such as R. salmoninarum (Pascho et al. 1991; Munsonet al. 2010). Vertical transmission of F. psychrophilum has beendemonstrated (Brown et al. 1997a; Cipriano 2005; Kumagai andNawata 2010a), although not conclusively, and specific factorsinfluencing transmission are still unknown. In this study, weevaluated the relationship between F. psychrophilum infectionlevels in female salmonid broodstock and disease outbreaks in

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4


FIGURE 4. Proportion of sampled fish positive for F. psychrophilum by nested PCR in the (A) chronic and (B) acute groups in Rainbow Trout stress trials. Ineach group, three fish from each family were collected at every time point.

FIGURE 5. Proportion of sampled fish positive for F. psychrophilum by nested PCR in the (A) control and (B) treatment groups in Coho Salmon stress trials. Ineach group, six fish from each family were sampled at every time point.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4

74 LONG ET AL.

TABLE 3. Broodstock kidney and spleen ELISA OD values, F. psychrophilum prevalence in progeny during progeny monitoring experiments, and F. psy-chrophilum prevalence in progeny during the stress experiments. Prevalence in progeny is based on results of nested PCR. For the Rainbow Trout trial, the progenymonitoring prevalence refers to the percentage of positive pooled samples (eight pools for each family) and the stress trial prevalence refers to the number ofpositive individual samples out of 48 total samples. For the Coho Salmon trial, the progeny-monitoring prevalence is the total number of positive samples duringthe monitoring period (n = 120). The stress trial prevalence is the total number of positive samples from each family (n = 180).

Kidney Spleen Progeny monitoring Stress trialTrial Family ELISA OD ELISA OD prevalence prevalence

Rainbow Trout F54 0.071 0.082 0.71 0.62F61 0.077 0.115 0.57 0.54F70 0.058 0.124 0.57 0.73F74 0.064 0.117 0.57 0.57F87 0.066 0.09 0.43 0.60

Coho Salmon F6 0.094 0.109 0.40 0.71F19 0.100 0.127 0.43 0.78F20 0.133 0.131 0.48 0.83F27 0.096 0.143 0.41 0.79F30 0.127 0.140 0.45 0.77

progeny. The study was structured such that prevalence in eggsand fry could be determined by nested PCR on samples takenduring hatch and early rearing stages. Progeny were then sub-jected to varying stressors in an attempt to induce a naturaloutbreak of BCWD and monitor infection prevalence. Whilevertical transmission of F. psychrophilum was confirmed in thisstudy, we were unable to induce a BCWD outbreak. Neverthe-less, the results of this study suggest that ELISA OD values ofkidney samples from individual female broodstock may be cor-related to the presence of F. psychrophilum in progeny, which inthe future could provide an important tool with the potential todecrease vertical transmission through the culling of eggs fromheavily infected broodstock.

In both progeny monitoring experiments, F. psychrophilumprevalence in individual families increased throughout the yolk-sac absorption process and immediately after as fry transitionedto exogenous feeding. It is possible that maternal transfer ofimmunity kept F. psychrophilum levels low during egg devel-opment and hatching. Although broodstock and egg-associatedantibody levels were not monitored in this study, transfer of an-tibodies from mother to progeny via the yolk sac has been docu-mented in several different teleost species (Brown et al. 1997b;Swain and Nayak 2009). Brown et al. (1997b) determined theseantibodies disappear by the time the yolk-sac absorption is com-plete or shortly thereafter in Coho Salmon. In Rainbow Trout,maternal antibodies specific for F. psychrophilum have beendetected in eggs and alevins up to 2 d posthatch (K. D. Cain, un-published data). Antimicrobial compounds (Yousif et al. 1994a,1994b), neutralizing antibodies (Shors and Winston 1989), andimmunoglobulin M (IgM) (Yousif et al. 1995) are also presentin yolk. In short, eggs and larvae rely on immune factors in theyolk, some of which are maternally derived, to protect againstpathogens through the end of the sac fry stage. Flavobacteriumpsychrophilum can survive in the perivitelline space and in yolk

(Brown et al. 1997a; Ekman et al. 2003; Vatsos et al. 2006;Kumagai and Nawata 2010a). Hence, increased detection of F.psychrophilum during this time period may be due to depletionof immunological factors resulting in increased growth of thebacterium.

When the total prevalence for the progeny monitoring ex-periment was calculated in trial 2, there was a significant pos-itive relationship between total F. psychrophilum prevalence inprogeny and ELISA OD values for female broodstock kidneysamples. This trend was not observed in the Rainbow Trout trial,but it is important to realize that broodstock infection was min-imal and they were not considered positive by ELISA. Whilethe results of trial 2 should be viewed with caution as both sam-ple size and range in OD values are small, they do support alink between broodstock and progeny infection. Further workthat incorporates a large-scale broodstock monitoring programat various facilities would provide a better understanding ofvertical transmission.

Ovarian fluid was used in this study as other studies haveused it to estimate broodstock infection levels and risk of verti-cal transmission with seemingly promising results (Taylor 2004;Kumagai and Nawata 2010a, 2011a, 2011b) although no attemptwas made to quantify tissue infection levels in those studies. Re-sults from the progeny monitoring experiments in the currentstudy indicate that ovarian fluid assays are not reliable tools forpredicting disease outbreaks in progeny in Rainbow Trout orCoho Salmon. Egg material from nine families (five RainbowTrout and four Coho Salmon) was positive for the bacteriumafter two successive rounds of disinfection. However, F. psy-chrophilum was only detected by means of nested PCR in ovar-ian fluid from five females, all of which were Coho Salmon.Detection of the bacterium in ovarian fluid by bacteriologicalculture was minimally successful as samples from four RainbowTrout and one Coho Salmon broodstock were positive by this

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4


method. Although contamination from other sources is possible,this is not the first study to report conflicting results for the detec-tion of F. psychrophilum in ovarian fluid and eggs (Taylor 2004).

As previously mentioned, the bacterium may have enteredthe eggs through other sources, specifically water and milt.Flavobacterium psychrophilum is capable of entering the eggduring the water hardening process (Kumagai et al. 2000). Asthere is no way of knowing whether the bacterium was present inthe water used in this study, it may have been an external sourceof contamination. Another potential source of contaminationnot evaluated in this study is milt. However, this seems unlikelyas male salmonids are not a factor in vertical transmission of R.salmoninarum. Evelyn et al. (1986a) found that eggs from unin-fected female broodstock fertilized with heavily contaminatedmilt did not become infected with the bacterium. Moreover,studies with F. psychrophilum have shown that bacterial levelsin reproductive fluids need to be high, i.e., >106 CFU/mL, forthe bacterium to enter the egg during the spawning process(Kumagai and Nawata 2010a). While there have been no studieslooking at the concentration of F. psychrophilum in milt in theUnited States, the average concentration of F. psychrophilumcells in salmonid milt in Japan is less than 104.5 CFU/mL(Kumagai and Nawata 2011a). For more information about therole of milt in vertical transmission of F. psychrophilum andto determine whether milt needs to be screened as part of abroodstock management plan, an experiment similar to thatused in Evelyn et al. (1986a) would need to be carried out.

These results do not rule out ovarian fluid as a means of trans-mitting the bacterium from the parent to the progeny. Rather,they indicate F. psychrophilum is not limited to entering the eggspostovulation when eggs are held in the body cavity and bathedin ovarian fluid. Taking into account the relationship betweenantigen values and prevalence in the progeny monitoring andstress experiments as well as the results from ovarian fluid as-says and other studies, we propose that F. psychrophilum canlocalize in reproductive tissue when fish are exposed to the bac-terium during early rearing. The bacterium could then enter theegg either during oogenesis, similar to R. salmoninarum, or per-haps prior to egg release into the body cavity, and localize inthe chorion or perivitelline space (Brown et al. 1997a; Kuma-gai et al. 1998, 2000; Vatsos et al. 2006; Kumagai and Nawata2010a).

Bacterial coldwater disease outbreaks have been linked tocommon stressors such as handling (Barton et al. 1986; Ryceand Zale 2004), high rearing density (Iguchi et al. 2003), waterflows, and decreased water quality including increased organicload (Wood 1974; Holt 1988; Garcia et al. 2000; Nematollahiet al. 2003; Cipriano and Holt 2005). We hypothesized thatprogeny from females with high antigen levels as measured byELISA would be more likely to experience a BCWD outbreakwhen reared under stressful conditions. We were unable to in-duce a disease outbreak in either trial, although several fish ex-hibited clinical symptoms of the disease, and F. psychrophilumwas detected in a proportion of mortalities. In the Rainbow

Trout trial, prevalence measured prior to the start of the stressexperiments was low in all families and tanks. As fish weremaintained in disinfected water at all stages, the only source ofF. psychrophilum would be the fish themselves. The proportionof positive samples in both the chronic and acute treatmentsincreased with time in this experiment and there was no dif-ference between families. Because fish exhibited signs of stressand F. psychrophilum can be shed from live fish (Madetoja et al.2000), we suspect that the increase in detection was a result ofincreased horizontal transmission of F. psychrophilum in bothtreatments. In contrast, F. psychrophilum was detected in themajority of the Coho Salmon samples taken immediately priorto the start of the stress experiments in trial 2, and prevalenceremained high throughout the experiment.

Given that F. psychrophilum was present in the RainbowTrout and Coho Salmon populations, our inability to cause aBCWD outbreak in the stress experiments may be due to otherfactors. Flavobacterium psychrophilum virulence can vary fromstrain to strain (Nematollahi et al. 2003) and it is possible thatthe detected strains were not virulent. The more likely expla-nation, however, is the differences between the laboratory andhatchery setting. It is impossible to mimic every biotic and abi-otic factor at a hatchery in the laboratory setting, including tankshape (circular versus raceway), water source, biofilm, organicmaterial in the raceways, as well as the interactions betweenthese factors, which can influence growth of the bacterium andoverall fish health.

Finally, there were noticeable differences in prevalence be-tween Rainbow Trout and Coho Salmon progeny suggestingthat vertical transmission may be influenced by the life cycle ofthe fish, i.e., iteroparous versus semelparous life history. Due togreater control and management possibilities at facilities withresident broodstock and the likelihood of constant exposureto ubiquitous environmental F. psychrophilum, it seems likelythat a vertical transmission risk may be higher in anadromousspecies that undergo a prespawn migration. During this migra-tion, fish are immunologically compromised due to stress andhigh levels of circulating cortisol as well as the start of senes-cence. It is not uncommon for these broodstock to have systemicbacterial infections as a result (Schreck et al. 2001). This is incontrast to broodstock at private commercial facilities that mayhave increased circulating cortisol levels during ovulation andspawning but still have a functional immune system and arenot entering senescence. As such, horizontal transmission maybe a greater risk factor at these facilities. Additionally, studiesby Soule et al. (2005a, 2005b) have demonstrated that F. psy-chrophilum isolates can be classified into two genetic lineages,trout and salmon, which may have implications for prevalencein broodstock and vertical transmission.

In summary, results from this study have implications formanagement and prevention of BCWD outbreaks. This is thefirst study to use ELISA to quantify F. psychrophilum levels inbroodstock tissue samples and relate this information to infec-tion prevalence in progeny. Ovarian fluid samples may indicate

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4

76 LONG ET AL.

broodstock infection levels at the time of spawning but do notprovide an accurate indication that the bacterium will be de-tected in progeny. While positive ELISA OD values appear tobe linked to increased infection prevalence in progeny in CohoSalmon, further studies that would have a larger sample sizeand take into account the many potential disease factors at rear-ing facilities as well as differences between species are needed.Without such studies, it will be difficult to prove that reductionof BCWD outbreaks in fry may be possible by implementing amanagement strategy that incorporates monitoring and cullingof eggs or progeny from families with high ELISA OD values.

ACKNOWLEDGMENTSThe authors acknowledge the Lummi tribe for their dona-

tion of Coho Salmon used in this study and Bill Finkbonner(Skookum Creek Fish Hatchery) for egg care. The EWOS feedwas provided by EWOS Canada, Ltd. The following individu-als assisted in sample collection: Mary Elliott and Sean Nepper;and Craig Olson, Marcia House, and Jim Bertolini of North-west Indian Fisheries Commission. Assistance was provided byScott Williams and the staff at the Aquaculture Research Insti-tute Coldwater Research Laboratory as well as by Neil Ashton,David Burbank, Tyson Fehringer, Kali Turner, and Jenna Davisfrom the University of Idaho. Statistical assistance was providedby Renae Shrum at the University of Idaho. The authors thankChristine Moffitt for a critical review of this manuscript. Fund-ing for this project was provided in part by the U.S. Departmentof Agriculture through a grant to the Western Regional Aqua-culture Center (grants 715623, 603298, 464697, and 688010),Idaho Chapter of the American Fisheries Society Graduate Stu-dent Scholarship and Susan B. Martin Scholarship (to A.L.),and the Washington State Agricultural Research Center.

REFERENCESBarnes, M. E., and M. L. Brown. 2011. A review of Flavobacterium psy-

chrophilum biology, clinical signs, and bacterial cold water disease preventionand treatment. Open Fish Science Journal [online serial] 4:40–48.

Barton, B. A., C. B. Schreck, and L. A. Sigismondi. 1986. Multiple acute distur-bances evoke cumulative physiological stress responses in juvenile Chinooksalmon. Transactions of the American Fisheries Society 115:245–251.

Brown, L. L., L. J. Albright, and T. P. T. Evelyn. 1990. Control of verticaltransmission of Renibacterium salmoninarum by injection of antibiotics intomaturing female Coho Salmon Oncorhynchus kisutch. Diseases of AquaticOrganisms 9:127–131.

Brown, L. L., W. Cox, and R. P. Levinel. 1997a. Evidence that the causal agentof bacterial cold-water disease Flavobacterium psychrophilum is transmittedwithin salmonid eggs. Diseases of Aquatic Organisms 29:213–218.

Brown, L. L., T. P. T. Evelyn, and G. K. Iwama. 1997b. Specific protective activ-ity demonstrated in eggs of broodstock salmon injected with rabbit antibodiesraised against a fish pathogen. Diseases of Aquatic Organisms 31:95–101.

Bruno, D. W., and A. L. S. Munro. 1986. Observations on Renibacteriumsalmoninarum and the salmonid egg. Diseases of Aquatic Organisms 1:83–87.

Bruun, M. S., A. S. Schmidt, L. Madsen, and I. Dalsgaard. 2000. Antimicro-bial resistance patterns in Danish isolates of Flavobacterium psychrophilum.Aquaculture 187(3–4):201–212.

Cipriano, R. C. 2005. Intraovum infection caused by Flavobacterium psy-chrophilum among eggs from captive Atlantic Salmon broodfish. Journalof Aquatic Animal Health 17:275–283.

Cipriano, R. C., and R. A. Holt. 2005. Flavobacterium psychrophilum, cause ofbacterial cold-water disease and Rainbow Trout fry syndrome. U.S. Geolog-ical Survey, Fish Disease Leaflet 86, Kearneysville, West Virginia.

del Cerro, A., I. Marquez, and J. M. Prieto. 2010. Genetic diversity and antimi-crobial resistance of Flavobacterium psychrophilum isolated from culturedRainbow Trout, Oncorhynchus mykiss (Walbaum), in Spain. Journal of FishDiseases 33:285–291.

Ekman, E., G. Akerman, L. Balk, and L. Norrgren. 2003. Nanoinjection as a toolto mimic vertical transmission of Flavobacterium psychrophilum in RainbowTrout Oncorhynchus mykiss. Diseases of Aquatic Organisms 55:93–99.

Elliott, D. G., R. J. Pascho, and G. L. Bullock. 1989. Developments in thecontrol of bacterial kidney disease of salmonid fishes. Diseases of AquaticOrganisms 6:201–215.

Evelyn, T. P. T., J. E. Ketcheson, and L. Prosperi Porta. 1984. Further evidencefor the presence of Renibacterium salmoninarum in salmonid eggs and forthe failure of povidone iodine to reduce the intra ovum infection rate in waterhardened eggs. Journal of Fish Diseases 7:173–182.

Evelyn, T. P. T., L. Prosperi-Porta, and J. E. Ketcheson. 1986a. Experimentalintra-ovum infection of salmonid eggs with Renibacterium salmoninarum andvertical transmission of the pathogen with such eggs despite their treatmentwith erythromycin. Diseases of Aquatic Organisms 1:197–202.

Evelyn, T. P. T., L. Prosperi-Porta, and J. E. Ketcheson. 1986b. Persistenceof the kidney-disease bacterium, Renibacterium salmoninarum, in CohoSalmon, Oncorhynchus kisutch (Walbaum), eggs treated during and afterwater-hardening with providone-iodine. Journal of Fish Diseases 9:461–464.

Evenden, A. J., T. H. Grayson, M. L. Gilpin, and C. B. Munn. 1993. Reni-bacterium salmoninarum and bacterial kidney disease–the unfinished jigsaw.Annual Review of Fish Diseases 3:87–104.

Garcia, C., F. Pozet, and C. Michel. 2000. Standardization of experimentalinfection with Flavobacterium psychrophilum, the agent of Rainbow TroutOncorhynchus mykiss fry syndrome. Diseases of Aquatic Organisms 42:191–197.

Haukenes, A. H., and C. M. Moffitt. 2002. Hatchery evaluation of erythromycinphosphate injections in prespawning spring Chinook Salmon. North Ameri-can Journal of Aquaculture 64:167–174.

Henrıquez-Nunez, H., O. Evrard, G. Kronvall, and R. Avendano-Herrera. 2012.Antimicrobial susceptibility and plasmid profiles of Flavobacterium psy-chrophilum strains isolated in Chile. Aquaculture 354–355:38–44.

Holt, R. A. 1988. Cytophaga psychrophila, the causative agent of bacterialcold-water disease in salmonid fish. Oregon State University, Corvallis.

Holt, R. A., J. S. Rohovec, and J. L. Fryer. 1993. Bacterial coldwater disease.Pages 2–23 in V. Inglis, R. J. Roberts, and N. R. Bromage, editors. Bacterialdiseases of fish. Blackwell Scientific Publications, Oxford, UK.

Iguchi, K., K. Ogawa, M. Nagae, and F. Ito. 2003. The influence of rearingdensity on stress response and disease susceptibility of Ayu (Plecoglossusaltivelis). Aquaculture 220(1–4):515–523.

Kumagai, A., C. Nakayasu, and N. Oseko. 2004. Effect of tobramycin supple-mentation to medium on isolation of Flavobacterium psychrophilum fromAyu Plecoglossus altivelis. Fish Pathology 39:75–78.

Kumagai, A., and A. Nawata. 2010a. Mode of the intra-ovum infection ofFlavobacterium psychrophilum in salmonid eggs. Fish Pathology 45:31–36.

Kumagai, A., and A. Nawata. 2010b. Prevention of Flavobacterium psy-chrophilum vertical transmission by iodophor treatment of unfertilized eggsin salmonids. Fish Pathology 45:164–168.

Kumagai, A., and A. Nawata. 2011a. Concentration of Flavobacterium psy-chrophilum in the ovarian fluid and milt of cultured salmonids. Fish Pathology46:116–119.

Kumagai, A., and A. Nawata. 2011b. Low likelihood of intra-ovum infectionwith Flavobacterium psychrophilum in cultured salmonids in Japan. FishPathology 46:123–125.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4


Kumagai, A., K. Takahashi, S. Yamaoka, and H. Wakabayashi. 1998. Inef-fectiveness of iodophore treatment in disinfecting salmonid eggs carryingCytophaga psychrophila. Fish Pathology 33:123–128.

Kumagai, A., S. Yamaoka, K. Takahashi, H. Fukuda, and H. Wakabayashi. 2000.Waterborne transmission of Flavobacterium psychrophilum in Coho Salmoneggs. Fish Pathology 35:25–28.

Lee, E. G. H., and T. P. T. Evelyn. 1989. Effect of Renibacterium salmoninarumlevels in the ovarian fluid of spawning Chinook Salmon on the prevalenceof the pathogen in their eggs and progeny. Diseases of Aquatic Organisms7:179–184.

Lindstrom, N. M., D. R. Call, M. L. House, C. M. Moffitt, and K. D. Cain. 2009.A quantitative enzyme-linked immunosorbent assay and filtration-based fluo-rescent antibody test as potential tools to screen broodstock for infection withFlavobacterium psychrophilum. Journal of Aquatic Animal Health 21:43–56.

Long, A., M. P. Polinski, D. R. Call, and K. D. Cain. 2012. Validation ofdiagnostic assays to screen broodstock for Flavobacterium psychrophiluminfections. Journal of Fish Diseases 35:407–419.

Lorenzen, E., and N. J. Olesen. 1997. Characterization of isolates of Flavobac-terium psychrophilum associated with coldwater disease or Rainbow Trout frysyndrome II: serological studies. Diseases of Aquatic Organisms 31:209–220.

Madetoja, J., I. Dalsgaard, and T. Wiklund. 2002. Occurrence of Flavobacteriumpsychrophilum in fish-farming environments. Diseases of Aquatic Organisms52:109–118.

Madetoja, J., P. Nyman, and T. Wiklund. 2000. Flavobacterium psychrophilum,invasion into and shedding by Rainbow Trout Oncorhynchus mykiss. Diseasesof Aquatic Organisms 52:109–118.

Madsen, L., J. D. Møller, and I. Dalsgaard. 2005. Flavobacterium psychrophilumin Rainbow Trout, Oncorhynchus mykiss (Walbaum), hatcheries: studies onbroodstock, eggs, fry and environment. Journal of Fish Diseases 28:39–47.

Mata, M., and Y. Santos. 2001. An improved enzyme-linked immunosorbentassay (ELISA) for detection of Flavobacterium psychrophilum isolated fromsalmon and Rainbow Trout. Bulletin European Association of Fish Patholo-gists 21:195–199.

Meyers, T. R., D. Korn, K. Glass, T. Burton, S. Short, K. Lipson, and N.Starkey. 2003. Retrospective analysis of antigen prevalences of Renibac-terium salmoninarum (Rs) detected by enzyme-linked immunosorbent assayin Alaskan Pacific salmon and trout from 1988 to 2000 and management ofRs in hatchery Chinook and Coho salmon. Journal of Aquatic Animal Health15:101–110.

Michel, C., D. Antonio, and R. P. Hedrick. 1999. Production of viable cul-tures of Flavobacterium psychrophilum: approach and control. Research inMicrobiology 150:351–358.

Moffitt, C. M. 1991. Oral and injectable applications of erythromycin insalmonid fish culture. Veterinary and Human Toxicology 33:49–53.

Munson, A. D., D. G. Elliott, and K. Johnson. 2010. Management of bacterialkidney disease in Chinook Salmon hatcheries based on broodstock testingby enzyme-linked immunosorbent assay: a multiyear study. North AmericanJournal of Fisheries Management 30:940–955.

Nematollahi, A., A. Decostere, F. Pasmans, R. Ducatelle, and F. Haesebrouck.2003. Adhesion of high and low virulence Flavobacterium psychrophilumstrains to isolated gill arches of Rainbow Trout Oncorhynchus mykiss. Dis-eases of Aquatic Organisms 55:101–107.

Pascho, R. J., D. G. Elliott, and J. M. Streufert. 1991. Brood stock segregationof spring Chinook Salmon Oncorhynchus tshawytscha by use of the enzyme-

linked immunosorbent assay (ELISA) and the fluorescent antibody technique(FAT) affects the prevalence and levels of Renibacterium salmoninarum in-fection in progeny. Diseases of Aquatic Organisms 12:25–40.

Pascho, R. J., and D. Mulcahy. 1987. Enzyme-linked immunosorbent assayfor a soluble antigen of Renibacterium salmoninarum, the causative agent ofsalmonid bacterial kidney disease. Canadian Journal of Fisheries and AquaticSciences 44:183–191.

Rangdale, R. E., and K. Way. 1995. Rapid identification of C. psychrophilafrom infected spleen tissue using an enzyme-linked-immunosorbent as-say (ELISA). Bulletin European Association of Fish Pathologists 15:213–216.

Reno, P. W. 1998. Factors involved in the dissemination of disease in fishpopulations. Journal of Aquatic Animal Health 10:160–171.

Ryce, E. K. N., and A. V. Zale. 2004. Bacterial cold water disease in West-slope Cutthroat Trout: hatchery epidemiology and control. Montana StateUniversity, Bozeman.

Schreck, C. B., W. Contreras-Sanchez, and M. S. Fitzpatrick. 2001. Effects ofstress on fish reproduction, gamete quality, and progeny. Aquaculture 197(1–4):3–24.

Shors, S. T., and V. Winston. 1989. Neutralizing antibodies for infectioushematopoietic necrosis virus in eggs of steelhead trout (Salmo gairdneri).American Journal of Veterinary Research 50:232–234.

Soule, M., K. D. Cain, S. LaFrentz, and D. R. Call. 2005a. Combining suppres-sion subtractive hybridization and microarrays to map the intraspecies phy-logeny of Flavobacterium psychrophilum. Infection and Immunity 73:3799–3802.

Soule, M., S. LaFrentz, K. D. Cain, S. LaPatra, and D. R. Call. 2005b. Polymor-phisms in 16S rRNA genes of Flavobacterium psychrophilum correlate withelastin hydrolysis and tetracycline resistance. Diseases of Aquatic Organisms65:209–216.

Swain, P., and S. K. Nayak. 2009. Role of maternally derived immunity in fish.Fish and Shellfish Immunology 27:89–99.

Taylor, P. W. 2004. Detection of Flavobacterium psychrophilum in eggs andsexual fluids of Pacific salmonids by a polymerase chain reaction assay:implications for vertical transmission of bacterial coldwater disease. Journalof Aquatic Animal Health 16:104–108.

Turaga, P., G. Wiens, and S. Kaattari. 1987. Bacterial kidney disease: the poten-tial role of soluble protein antigen(s). Journal of Fish Biology 31:191–194.

Vatsos, I. N., K. D. Thompson, and A. Adams. 2006. Colonization of Rain-bow Trout, Oncorhynchus mykiss (Walbaum), eggs by Flavobacterium psy-chrophilum, the causative agent of Rainbow Trout fry syndrome. Journal ofFish Diseases 29:441–444.

Wood, J. W. 1974. Diseases of Pacific salmon, their prevention and treatment.Washington Department of Fisheries, Olympia.

Yousif, A. N., L. J. Albright, and T. P. T. Evelyn. 1994a. In vitro evidencefor the antibacterial role of lysozyme in salmonid eggs. Diseases of AquaticOrganisms 19:15–19.

Yousif, A. N., L. J. Albright, and T. P. T. Evelyn. 1994b. Purification and char-acterization of a galactose-specific lectin from the eggs of Coho SalmonOncorhynchus kisutch and its interaction with bacterial fish pathogens. Dis-eases of Aquatic Organisms 20:127–136.

Yousif, A. N., L. J. Albright, and T. P. T. Evelyn. 1995. Immunological evidencefor the presence of an IgM-like immunoglobulin in the eggs of Coho SalmonOncorhynchus kisutch. Diseases of Aquatic Organisms 23:109–114.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:50

15

June

201

4



Susceptibility of Koi and Yellow Perch to InfectiousHematopoietic Necrosis Virus by ExperimentalExposureAlexander D. Palmera & Eveline J. Emmeneggera

a U.S. Geological Survey, Western Fisheries Research Center, 6505 Northeast 65th Street,Seattle, Washington 98115, USAPublished online: 04 Jun 2014.

To cite this article: Alexander D. Palmer & Eveline J. Emmenegger (2014) Susceptibility of Koi and Yellow Perch toInfectious Hematopoietic Necrosis Virus by Experimental Exposure, Journal of Aquatic Animal Health, 26:2, 78-83, DOI:10.1080/08997659.2014.886634







http://dx.doi.org/10.1080/08997659.2014.886634



Journal of Aquatic Animal Health 26:78–83, 2014American Fisheries Society 2014ISSN: 0899-7659 print / 1548-8667 onlineDOI: 10.1080/08997659.2014.886634

COMMUNICATION

Susceptibility of Koi and Yellow Perch to InfectiousHematopoietic Necrosis Virus by Experimental Exposure

Alexander D. Palmer and Eveline J. Emmenegger*U.S. Geological Survey, Western Fisheries Research Center, 6505 Northeast 65th Street,Seattle, Washington 98115, USA

AbstractInfectious hematopoietic necrosis virus (IHNV) is a novirhab-

doviral pathogen that originated in western North America amonganadromous Pacific salmonids. Severe disease epidemics in the late1970s resulting from IHNV’s invasion into farmed Rainbow TroutOncorhynchus mykiss in North America, Asia, and Europe empha-sized IHNV’s ability to adapt to new hosts under varying rearingconditions. Yellow Perch Perca flavescens and Koi Carp Cyprinuscarpio (hereafter, “Koi”) are aquaculture-reared fish that are highlyvalued in sport fisheries and the ornamental fish trade, respectively,but it is unknown whether these fish species are vulnerable to IHNVinfection. In this study, we exposed Yellow Perch, Koi, and steelhead(anadromous Rainbow Trout) to IHNV by intraperitoneal injection(106 PFU/fish) and by immersion (5.7 × 105 PFU/mL) for 7 h, andmonitored fish for 28 d. The extended immersion exposure andhigh virus concentrations used in the challenges were to determineif the tested fish had any level of susceptibility. After experimentalexposure, Yellow Perch and Koi experienced low mortality (<6%)compared with steelhead (>35%). Virus was found in dead fish ofall species tested and in surviving Yellow Perch by plaque assayand quantitative reverse transcription polymerase chain reaction(qPCR), with a higher prevalence in Yellow Perch than Koi. Infec-tious virus was also detected in Yellow Perch out to 5 d after bathchallenge. These findings indicate that Yellow Perch and Koi arehighly resistant to IHNV disease under the conditions tested, butYellow Perch are susceptible to infection and may serve as possiblevirus carriers.

Infectious hematopoietic necrosis virus (IHNV) is consid-ered one of the greatest viral threats to salmonid populationsin the Pacific Northwest. The virus causes acute and systemicdisease in both wild and farmed salmonid populations (Wolf1988). Infectious hematopoietic necrosis virus isolates canbe genetically separated into three genogroups in NorthAmerica—U, M, and L—which typically correlate geograph-ically with host species specificities (Kurath et al. 2003). Inthe Pacific Northwest, U-genogroup isolates range from the

*Corresponding author: [email protected] July 29, 2013; accepted November 22, 2013

Columbia River north to Alaska and are highly virulent forSockeye Salmon Oncorhynchus nerka (Garver et al. 2006;Penaranda et al. 2009). L-genogroup isolates of IHNV arefound off the southern coast of Oregon and in California water-sheds, and cause severe mortalities in populations of ChinookSalmon O. tshawytscha (LaPatra et al. 1990, 1993). Infectioushematopoietic necrosis virus isolates of the M-genogroupare found in the Columbia River basin. M-genogroup IHNVstrains are particularly lethal to Rainbow Trout O. mykiss andsteelhead (anadromous Rainbow Trout) and considered themost genetically diverse (Kurath et al. 2003; Garver et al. 2006).

Disease was first observed in cultured Sockeye Salmonon the west coast of North America in the 1950s (Ruckeret al. 1953; Watson et al. 1954). It was not until 1969 that thedisease was isolated and described in Rainbow Trout (Amendet al. 1969) and later proved to be the same virus that alsoinfected Chinook Salmon (Amend et al. 1970). In addition tothe salmonid species mentioned above, IHNV is also a reportedpathogen of Chum Salmon O. keta, Amago O. rhodurus, MasuO. masou, Coho Salmon O. kisutch, and Atlantic Salmon Salmosalar (Wolf 1988). Though not considered principal hosts ofIHNV, the virus has been isolated from other salmonids, includ-ing Brown Trout S. trutta, Cutthroat Trout O. clarkii, Lake TroutSalvelinus namaycush, Arctic Char S. alpinus, Brook Trout S.fontinalis, White-Spotted Char S. leucomaenis, Ayu Plecoglos-sus altivelis, and nonsalmonids European Eel Anguilla anguilla,Pacific Herring Clupea pallasii, Atlantic Cod Gadus morhua,White Sturgeon Acipenser transmontanus, Northern Pike Esoxlucius, Shiner Perch Cymatogaster aggregata, and TubesnoutAulorhynchus flavidus in the wild or after waterborne virusexposure (Wolf 1988; Bootland and Leong 1999). Infectioushematopoietic necrosis virus has been isolated from manyspecies of fish, but the host range is relatively smaller thanthat of other novirhabdoviral pathogens like viral hemorrhagic

78

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

COMMUNICATION 79

septicemia virus (VHSV), a virus that has been isolated from82 different species (OIE 2009). Although native to the PacificNorthwest, IHNV has spread to Asia and Europe, particularlyimpacting Rainbow Trout aquaculture. The ever-increasingworld aquaculture production of finfish, 1.5 million tons in1970–41.6 million tons in 2011 (FAO 2013), indicates that viralpathogens like IHNV that evolve relatively rapidly are of majorconcern (Bandın and Dopazo 2011; Crane and Hyatt 2011).

Yellow perch Perca flavescens and Koi Carp Cyprinus carpio(hereafter, “Koi”) are aquaculture-reared fish that are highly val-ued in sport fisheries and the ornamental fish trade, respectively,and it is unknown whether these fish species are vulnerable toIHNV infection. Yellow Perch is native to much of northernNorth America east of the Rockies but can also be found inmost Pacific Northwest bodies of fresh water where exposure toIHNV is possible (Brown et al. 2009). Koi originated in CentralAsia and has moved east to China and Japan, and west to Eu-rope where IHNV is also found (Balon 1995; Crane and Hyatt2011). Also, Koi are present in outdoor ponds at productionfacilities and on private lands in the Pacific Northwest whereIHNV is endemic. We challenged Koi and Yellow Perch withIHNV 220-90, a highly virulent M-genogroup isolate (Wargoet al. 2010), by intraperitoneal injection and immersion to assessinfection levels as well as cumulative percent mortality over a28-d period.

METHODSStudy fish.—All fish species transferred to the Western Fish-

eries Research Center (WFRC) in Seattle, Washington, werechallenged and held in tanks receiving single-pass, sand-filtered,UV-treated freshwater. Fish health history is well documented,and none of the stockfish used had any prior exposures to IHNV.The protocols for experimental use of live animals were ap-proved by the Institutional Animal Care and Use Committeeof the WFRC under the guidelines provided by the Guide forthe Care and Use of Laboratory Animals (NRC 2011). Yel-low Perch were shipped to the WFRC from the University ofWisconsin Great Lakes Water Institute in Milwaukee, Wiscon-sin. Yellow Perch fry were held at 12◦C for the first year, andthen water temperature was lowered to 9◦C until challenged.Nineteen-month-old perch used for the intraperitoneal (IP) in-jection portion of the challenge weighed on average 3.6 g, whilethose used for the immersion portion of the challenge averaged1.1 g. These fish were held at a lower temperature relative tostandard Yellow Perch rearing procedures in order to increasesusceptibility. A domestic stock of Koi fry was obtained fromPan Intercorp in 2006 (Kenmore, Washington). Koi were heldat temperatures ranging from 15◦C to 19◦C. At the time ofchallenge, Koi were ∼ 6 years old and weighed on average23.8 g. Steelhead fry received from the Quinault National FishHatchery (Humptulips, Washington) were 11 months old andweighed 3.0 g at the start of challenge. All fish were fed areduced diet to restrict growth; this could have led to an inhib-

ited immune system, affecting susceptibility to IHNV. Steelheadare known to be vulnerable to IHNV strain 220-90 (G. Kurath,WFRC, personal communication) and served as susceptible hostcontrols.

Study design.—The IHNV 220-90 strain used in challengeswas originally isolated from acutely infected juvenile RainbowTrout during routine examinations of hatchery-reared fish inHagerman Valley, Idaho (LaPatra et al. 1991). Monolayers ofepithelioma papulosum cyprini (EPC) cells, which are com-posed of cells of Fathead Minnow Pimephales promelas (Fijanet al. 1983; Winton et al. 2010), were grown until conflu-ent following standard IHNV culture propagation procedures(LaPatra et al. 1994). Infected cells were incubated at 15◦C un-til 90% cytopathic effect was observed. Cells and media werespun at 1,000 × g for 10 min to pellet debris. The supernatantwas collected and frozen at −80◦C until challenge.

Bath challenge.—Groups of fish species were challengedby immersion in 3 L of aerated water at a virus concentrationof 5.7 × 105 PFU/mL for 7 h, after which water was turnedback on. High virus concentrations and a longer exposure timewere utilized to maximize the likelihood of infection. Triplicategroups of 20 Koi, 35 Yellow Perch, and 35 steelhead were used toassess cumulative mortality. Time point samples were collectedfrom a fourth replicate tank of Koi and a fourth replicate tankof Yellow Perch to assess infection, both of which were alsoexposed to virus by immersion. Three Yellow Perch and twoKoi were euthanized on each sampling day following protocolsoutlined in Kell et al. (2013) and stored at −80◦C. Fish samplingwas performed 1, 3, 5, 7, 10, 13, 16, 19, 21, 24, and 28 dpostexposure.

Injection challenge.—All three fish species were challengedin triplicate groups of 35 fish each by IP injection after be-ing anesthetized with 0.12 g of tricaine methanesulfonate (MS-222; Argent Chemical Laboratories, Redmond, Washington)and 0.60 g of NaHCO3 (Sigma-Aldrich, St. Louis, Missouri)in 2 L of freshwater. After injection with 100 µL containing1.0 × 106 PFU of IHNV, fish were placed into a recovery bucketcontaining 2 L of freshwater and then to their respective hold-ing tanks containing flow-through water. Mock treatments ofeach fish species, also in triplicate groups of 35 fish per tank,were injected with virus-free cell culture media of the samevolume. During injection and immersion challenges, the watertemperature ranged from 10◦C to 12◦C. All challenged fish weremonitored daily, and all dead fish were removed. Any sampledfish were frozen at −80◦C until processing.

Analysis of challenge data.—Susceptibility comparisons be-tween the tested fish species were based on the mortality chal-lenge data, which were statistically analyzed using GraphPadInStat version 3.1a. Cumulative percent mortality from the trip-licate treatment groups of the fish species challenges were arc-sine square root transformed prior to statistical assessment byANOVA and a Tukey–Kramer multiple comparison posttest, ifapplicable. A significant relationship was designated for com-parisons yielding P-values of ≤0.05.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

80 PALMER AND EMMENEGGER

Processing of whole-fish samples.—Time point samples,dead fish, and a subset of survivors and mock fish were initiallydiluted in a volume of minimum essential medium containingFungizone, penicillin, and gentamicin, and homogenized us-ing a Stomacher-80 (Seward Laboratory Systems, Biomaster).Whole Yellow Perch and steelhead were diluted at a ratio of 1 gof fish tissue to 7 mL of medium (1:8). The larger Koi werediluted at a ratio of 1 g of fish tissue to 1 mL of medium (1:2).The homogenate was then spun down for 10 min at 1,000 × g.A 210-µL portion of this homogenate was processed immedi-ately for RNA extraction following manufacturer’s protocols forthe Qiagen RNeasy kit (Qiagen, Hilden, Germany). To confirmthe quality and quantity of RNA, samples were tested using aNanoDrop ND-1000 spectrophotometer (NanoDrop Technolo-gies, Wilmington, Delaware). The initial homogenate super-natant was further diluted in the same medium to a 1:40 dilutionin order to quantify viable virus particles, and extracted RNAwas stored at −80◦C until further processing.

Detection of infectious hematopoietic necrosis virus.—To de-tect virus, two methods were used: plaque assay and quantitativereverse transcription PCR (qPCR). For plaque assay, the numberof plaque forming units per gram of fish was determined usingthe method previously described by Batts and Winton (1989).The limit of detection of the plaque assay was 2.3 log10 PFU/g.The qPCR assay was performed using RNA extracted from fishtissue homogenate supernatant as described by Purcell et al.(2013) to quantitate IHNV N-gene copies. Results are shown asmean viral load, and the limit of detection of the qPCR reactionis 3.4 log10 copies/g.

RESULTSMortality levels of virus in IHNV-exposed steelhead reached

30% and 96% cumulative percent mortality (CPM) in immer-sion and IP injection treatments, respectively. In contrast, virus-exposed Koi and Yellow Perch mortality was less than 6% inany treatment (Figure 1). Mean CPM of fish from mock treat-ments was 1% in Koi, 0% in Yellow Perch, and 2% in steel-head. All mock-treated fish tested for virus were negative viaqPCR and plaque assay (data not shown). Susceptibility compar-isons between each fish species tested determined that IHNV-challenged steelhead were significantly different (P < 0.001)from mock-exposed steelhead and both mock or virus-exposedKoi or Yellow Perch by either exposure method (e.g., injectionor immersion).

Koi injected with IHNV suffered two deaths postexposure—one on day 2 and another on day 15, the latter showing hem-orrhaging on the abdomen and eye. One Koi from the virusimmersion treatments died on day 15 with minor hemorrhagingaround the eye. There was no significant difference in mortalitybetween Koi tested by either challenge method (P > 0.05). Ofthe two dead Koi exposed to IHNV through IP injection, onetested positive for virus by plaque assay and qPCR at 3.1 log10

PFU/g and 5.5 log10 IHNV N-gene copies copies/g, respectively,

FIGURE 1. Mean CPM of Koi (triangle), Yellow Perch (square), and steelhead(circle) after exposure to IHNV using two methods. Immersion (Imm) treatments(shaded symbols) were at a dose of 5.7 × 105 PFU/mL and intraperitoneal-injected (IP) fish each receive 1 × 106 PFU. Error bars represent SD from themean CPM of three replicate tanks for each treatment group.

a titer lower than the injection dose (Figure 2). All other virus-exposed Koi were negative for virus, including time point sam-ples, survivors, and the remaining dead fish (data not shown).

No deaths occurred in Yellow Perch in IHNV immersiontreatment tanks, but a total of six dead fish were found in theinjection treatment tanks between days 10 and 18 postexposure.Mortality levels of Yellow Perch injected with virus were barelysignificant over mocks and Yellow Perch exposed to the virus byimmersion (P < 0.05). Clinical signs in dead virus-exposed Yel-low Perch included exophthalmia; hemorrhaging along the base

FIGURE 2. Virus concentrations of dead fish. Mean virus titers (bars) byplaque assay (PA) listed on left axis and viral loads (solid squares) determinedby qPCR displayed on right axis for each fish species (Koi [Koi], Yellow Perch[Perch], or steelhead [Trout]) challenged with IHNV by immersion or IP injec-tion. The limit of detection of the plaque assay is 2.3 log10 PFU/g while thatof the qPCR reaction is 3.4 log10 copies/g. IHNV prevalence (n = number ofvirus positive samples/number of dead fish tested) for each challenge method ispresented above each treatment group listed on the x-axis.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

COMMUNICATION 81

FIGURE 3. (a, b) External clinical signs of exophthalmia; hemorrhaging inoperculum, pectoral fins, anal vent, and eyes; and distended abdomen observedon a Yellow Perch that died 10 d after IHNV injection. [Figure available onlinein color.]

of the pectoral fins, operculum, and anal vent; and a distendedabdomen (Figure 3).

Dead Yellow Perch from the IHNV-injection treatment tanksthat were positive for virus (n = 6/6) had higher viral titers(5.5–6.3 log10 PFU/g and viral loads ranging from 6.4 to 8.7log10 copies/g) relative to Koi and comparable virus levels toimmersion-treated dead steelhead (5.2–6.4 log10 PFU/g and 7.6–8.1 log10 copies/g), but lower than that of dead trout injected withvirus (5.9–7.6 log10 PFU/g and 7.8–9.3 log10 copies/g; Figure 2).Yellow Perch survivors exposed to IHNV via immersion testednegative for virus (data not shown). In contrast, Yellow Perchsurvivors injected with IHNV were positive for virus, they hadplaque assay titers ranging from 3.3 to 5.1 log10 PFU/g and viralloads ranging from 5.8 to 8.1 log10 copies/g (Figure 4). Onesteelhead survivor injected with virus tested positive for IHNVwith viral titers and viral loads of 5.9 PFU/g and 7.8 log10

copies/g, respectively. Though no bath-exposed Yellow Perchdied in this challenge, time point samples indicated that bath-exposed Yellow Perch were positive for IHNV 220-90 throughday 5. Virus titers in these fish ranged from 2.3 to 4.4 log10

PFU/g, and N-gene copies ranged from 4.4 to 7.0 log10 copies/g(Figure 5).

FIGURE 4. Virus concentrations in fish surviving 28 d after injection withIHNV. Mean virus titers (bars) by plaque assay (PA) listed on left axis and viralloads (solid squares) determined by qPCR displayed on right axis for each fishspecies (Koi, Yellow Perch, or steelhead). The limit of detection of the plaqueassay is 2.3 log10 PFU/g, while that of the qPCR reaction is 3.4 log10 copies/g;IHNV prevalence (n = number of virus positive samples/number of survivor fishsamples tested) for each challenge method is presented above each treatmentgroup listed on the x-axis.

DISCUSSIONKoi have previously been shown to have low sus-

ceptibility to another aquatic Novirhabdovirus, VHSV, inother experimental exposure studies (Cornwell et al. 2013;Emmenegger et al. 2013). Our findings indicate that Koi arehighly resistant to infection and disease by IHNV in the

FIGURE 5. Virus concentrations in Yellow Perch tested at time points afterIHNV immersion exposure. Mean virus titers (bars) by plaque assay (PA) listedon left axis and viral loads (solid squares) determined by qPCR displayed onright axis for Yellow Perch. The limit of detection of the plaque assay is 2.3 log10

PFU/g, while that of the qPCR reaction is 3.4 log10 copies/g. Three fish weresampled at each time point and tested. Through day 5, all fish tested positive forIHNV N-gene. A single fish was positive for infectious viral particles on day 1postexposure, 3/3 were positive on day 3, and 2/3 were positive on day 5.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

82 PALMER AND EMMENEGGER

conditions tested, and along with previous studies, this couldsuggest that Koi may have a general resistance to novirhab-doviral infection. Although fish were challenged with highconcentrations of virus and exposed to virus both by injectionand immersion, for an extended period of 7 h, Koi showed neg-ligible susceptibility. However, caveats in experimental designmay have contributed to a decrease in Koi’s susceptibility toIHNV. Koi were only challenged at a single water temperatureof 10◦C. In another VHSV experiment challenging BluegillsLepomis macrochirus, peak mortalities were observed at 10◦C;cumulative mortality decreased as temperature increased (Good-win and Merry 2011), suggesting lower temperatures may beideal for novirhabdoviral replication in vivo. A low temperatureof 10◦C was selected to optimize viral introduction and prop-agation in a thermally stressed host, although challenges usinghigher temperatures like 15◦C would be more ideal for acceler-ating host metabolism and in turn IHNV replication. However,at this higher temperature there would be less thermal stresson Koi, potentially strengthening its immune system. Both fac-tors could influence mortality outcomes. Another parameter thatmay have affected mortality was the size and age of challengedKoi. Because Koi were larger and older, averaging 23.9 g and6 years of age, their immune systems may have been well de-veloped, causing mortality to be particularly low. An interestingobservation was that viable virus and clinical signs of diseasewere present from a Koi that died 15 d after being injected, indi-cating that IHNV infection and disease fulmination is possiblein Koi.

Yellow Perch demonstrated low mortality to IHNV, but perchappear more vulnerable to IHNV infection than Koi at the testedtemperature. The prevalence of the virus measured in both viablevirus and IHNV N-gene copies was greater in Yellow Perch thanin Koi. Injected Yellow Perch had viral titers and viral loads onlyslightly lower than that of the tested positive control species,steelhead, which is known to be susceptible to IHNV strain 220-90 (Kurath, personal communication). Detection of virus out to5 d after initiation of immersion infection indicates that onceIHNV gains entry, replication can occur in this species. The factthat surviving IP-injected Yellow Perch were also positive forvirus suggests that a long-term carrier state can be established.Yellow Perch share habitats with IHNV-susceptible salmon andtrout species in the Pacific Northwest (Bonar et al. 2005; Brownet al. 2009). Yellow Perch carriers may shed virus, infectingvulnerable cohabitants. Yellow Perch may also be at greaterrisk of developing IHN disease themselves if environmentalstressors (such as high fish densities, poor water quality, orclimate change) lead to a suppressed immune system.

This is the first study to test the susceptibility levels of Yel-low Perch and Koi after experimental exposure to IHNV. Futuretesting with younger fish using various IHNV strains from dif-ferent genogroups and a range of challenge temperatures wouldbe beneficial to further assessing species susceptibility to IHNV.Confirmatory screening of virus shedding levels and antibodytiters in the survivors would also be a valuable measure of in-

fectivity levels. Koi appear highly resistant to IHNV infectionand disease. Yellow Perch also appear to be resistant to diseasein these experimental conditions but have a higher incidence ofinfection and persistence. This suggests that although the prob-ability is extremely low, Yellow Perch may serve as potentialvirus carriers or hosts for adaptation if exposed continuously,similar to what occurred in Rainbow Trout. Yellow Perch couldbe at greater risk of developing IHN disease.

ACKNOWLEDGMENTSWe thank Rachel Thompson and Maureen Purcell for tech-

nical assistance, Gael Kurath for sharing steelhead generouslydonated by the Quinault National Fish Hatchery, Rick Goetzand Wendy Olson with the University of Wisconsin–Milwaukeefor sending extra Yellow Perch, Joel Burkard of Pan Intercorpfor providing Koi, and WFRC Fish Health team members BillBatts, Maureen Purcell, and Rachel Breyta for their editorialprowess in reviewing this paper. This work was funded by theU.S. Geological Survey (USGS) and has been peer reviewedand approved for publication consistent with USGS Funda-mental Science Practices (http://pubs.usgs.gov/circ/1367/). Thismanuscript is submitted for publication with the understandingthat the U.S. Government is authorized to reproduce and dis-tribute reprints for governmental purposes. Any use of trade,firm, or product names is for descriptive purposes only and doesnot imply endorsement by the U.S. Government.

REFERENCESAmend, D. F., and V. C. Chambers. 1970. Morphology of certain viruses of

salmonid fishes. I. In vitro studies of some viruses causing hematopoieticnecrosis. Journal of Fisheries Research Board of Canada 27:1285–1295.

Amend, D. F., W. T. Yasutake, and R. W. Mead. 1969. A hematopoietic virusdisease of Rainbow Trout and Sockeye Salmon. Transactions of the AmericanFisheries Society 98:796–804.

Balon, E. K. 1995. Origin and domestication of the wild carp, Cyprinus carpio:from Roman gourmets to the swimming flowers. Aquaculture 129:3–48.

Bandın, I., and C. P. Dopazo. 2011. Host range, host specificity and hypothesizedhost shift events among viruses of lower vertebrates. Veterinary Research42:67.

Batts, W. N., and J. R. Winton. 1989. Enhanced detection of infectioushematopoietic necrosis virus and other fish viruses by pretreatment of cellmonolayers with polyethylene glycol. Journal of Aquatic Animal Health1:284–290.

Bonar, S. A., B. D. Bolding, M. Divens, and W. Meyer. 2005. Effects of intro-duced fishes on wild juvenile Coho Salmon in three shallow Pacific Northwestlakes. Transactions of the American Fisheries Society 134:641–652.

Bootland, L. M., and J. C. Leong. 1999. Infectious hematopoietic necrosis virus.Fish Diseases and Disorders 3:57–112.

Brown, T. G., B. Runciman, M. J. Bradford, and S. Pollard. 2009. A biologicalsynopsis of Yellow Perch (Perca flavescens). Canadian Manuscript Report ofFisheries and Aquatic Sciences 2883.

Cornwell, E. R., S. L. LaBuda, G. H. Groocock, R. G. Getchell, and P. R.Bowser. 2013. Experimental infection of koi carp with viral hemorrhagicsepticemia virus type IVb. Journal of Aquatic Animal Health 25:36–41.

Crane, M., and A. Hyatt. 2011. Viruses of fish: an overview of significantpathogens. Viruses 3:2025–2046.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

COMMUNICATION 83

Emmenegger, E. J., H. M. Chong, P. K. Hershberger, and G. Kurath. 2013.Virulence of viral hemorrhagic septicemia virus (VHSV) genotypes Ia, IVa,IVb, IVc, in five fish species. Diseases of Aquatic Organisms 107:99–111.

FAO (Food and Agriculture Organization of the United Nations). 2013.Global aquaculture production statistics for the year of 2011. Available:http://www.fao.org/fishery/publications/en. (March 2014).

Fijan, N., E. Sulimanovic, M. Bearzotti, D. Muzinic, L. O. Zwillenberg, S.Chilmonczyk, J. F. Vautherot, and P. de Kinkelin. 1983. Some properties ofthe epithelioma papulosum cyprini (EPC) cell line from carp Cyprinus carpio.Annales de l’Institut Pasteur Virologie 134:207–220.

Garver, K. A., W. N. Batts, and G. Kurath. 2006. Virulence comparisons ofinfectious hematopoietic necrosis virus U and M genogroups in SockeyeSalmon and Rainbow Trout. Journal of Aquatic Animal Health 18:232–243.

Goodwin, A. E., and G. E. Merry. 2011. Mortality and carrier status of Bluegillsexposed to viral hemorrhagic septicemia virus genotype IVb at differenttemperatures. Journal of Aquatic Animal Health 23:85–91.

Kell, A. M., A. R. Wargo, and G. Kurath. 2013. The role of virulence in in vivosuperinfection fitness of the vertebrate RNA virus infectious hematopoieticnecrosis virus. Journal of Virology 87:8145–8157.

Kurath, G., K. A. Garver, R. M. Troyer, E. J. Emmenegger, K. Einer-Jensen, andE. D. Anderson. 2003. Phylogeography of infectious haematopoietic necrosisvirus in North America. Journal of General Virology 84:803–814.

LaPatra, S. E., W. J. Groberg, J. S. Rohovec, and J. L. Fryer. 1990. Size-relatedsusceptibility of salmonids to two strains of infectious hematopoietic necrosisvirus. Transactions of the American Fisheries Society 119:25–30.

LaPatra, S. E., J. L. Fryer, and J. S. Rohovec. 1993. Virulence comparison of dif-ferent electropherotypes of infectious hematopoietic necrosis virus. Diseasesof Aquatic Organisms 16:121–131.

LaPatra, S. E., K. A. Lauda, and G. R. Jones. 1994. Antigenic variants of infec-tious hematopoietic necrosis virus and implications for vaccine development.Diseases of Aquatic Organisms 20:119–126.

LaPatra, S. E., K. A. Lauda, and A. W. Morton. 1991. Antigenic and virulencecomparisons of eight isolates of infectious hematopoietic necrosis virus from

the Hagerman Valley, Idaho, USA. Pages 125–132 in J. L. Fryer, editor.Proceedings of the Second International Symposium on viruses of lowervertebrates. Oregon State University Press, Corvallis.

NRC (National Research Council). 2011. Guide for the care and use of labora-tory animals, 8th edition. Available: http://www.nap.edu/catalog/12910.html.(March 2014).

OIE (World Organization for Animal Health). 2009. Viral haemorrhagicsepticaemia. Pages 279–298 in Aquatic animal health standards com-mission, 6th edition. OIE, Paris. Available: http://web.oie.int/eng/normes/fmanual/2.3.09.VHS.pdf. (March 2014).

Penaranda, M. D., M. K. Purcell, and G. Kurath. 2009. Differential virulencemechanisms of infectious hematopoietic necrosis virus in Rainbow Trout(Oncorhynchus mykiss) include host entry and virus replication kinetics.Journal of General Virology 90:2172–2182.

Purcell, M. K., R. L. Thompson, K. A. Garver, L. M. Hawley, W. N. Batts, L.Sprague, C. Sampson, and J. R. Winton. 2013. Universal reverse-transcriptasereal-time PCR for infectious hematopoietic necrosis virus (IHNV). Diseasesof Aquatic Organisms 106:103–115.

Rucker, R. R., W. J. Whipple, J. R. Parvin, and C. A. Evans. 1953. A contagiousdisease of salmon possibly of virus origin. U.S. Fish and Wildlife ServiceFishery Bulletin 54:35–46.

Wargo, A. R., K. A. Garver, and G. Kurath. 2010. Virulence correlates withfitness in vivo for two M group genotypes of infectious hematopoietic necrosisvirus (IHNV). Virology 404:51–58.

Watson, S. W., R. W. Guenther, and R. R. Rucker. 1954. A virus disease of Sock-eye Salmon, interim report. U.S. Fish and Wildlife Service Special ScientificReport of Fisheries 138.

Winton, J., W. Batts, P. deKinkelin, M. LeBerre, M. Bremont, and N. Fijan.2010. Current lineages of the epithelioma papulosum cyprini (EPC) cell lineare contaminated with Fathead Minnow, Pimephales promelas, cells. Journalof Fish Diseases 33:701–704.

Wolf, K. 1988. Infectious hematopoietic necrosis. Pages 83–114 in K. Wolf,editor. Fish viruses and fish viral diseases. Cornell University Press, Ithaca,New York.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4



Risk Factors Associated with Enteric Septicemia ofCatfish on Mississippi Commercial Catfish FarmsFred L. Cunninghama, S. W. Jackb, David Hardinc & Robert W. Willsb

a Wildlife Services-National Wildlife Research Center, Mississippi Field Station, ScalesBuilding, 125 Stone Boulevard, Mississippi State, Mississippi 39762, USAb Department of Pathobiology and Population Medicine, College of Veterinary Medicine, PostOffice Box 6100, Mississippi State, Mississippi 39762, USAc School of Veterinary Medicine and Biomedical Sciences, University of Nebraska–Lincoln,120B VSB, Lincoln, Nebraska 68583, USAPublished online: 04 Jun 2014.

To cite this article: Fred L. Cunningham, S. W. Jack, David Hardin & Robert W. Wills (2014) Risk Factors Associated withEnteric Septicemia of Catfish on Mississippi Commercial Catfish Farms, Journal of Aquatic Animal Health, 26:2, 84-90, DOI:10.1080/08997659.2014.886635







http://dx.doi.org/10.1080/08997659.2014.886635




ARTICLE

Risk Factors Associated with Enteric Septicemiaof Catfish on Mississippi Commercial Catfish Farms

Fred L. Cunningham*Wildlife Services-National Wildlife Research Center, Mississippi Field Station, Scales Building,125 Stone Boulevard, Mississippi State, Mississippi 39762, USA

S. W. JackDepartment of Pathobiology and Population Medicine, College of Veterinary Medicine,Post Office Box 6100, Mississippi State, Mississippi 39762, USA

David HardinSchool of Veterinary Medicine and Biomedical Sciences, University of Nebraska–Lincoln,120B VSB, Lincoln, Nebraska 68583, USA

Robert W. WillsDepartment of Pathobiology and Population Medicine, College of Veterinary Medicine,Post Office Box 6100, Mississippi State, Mississippi 39762, USA

AbstractA gram-negative bacterium, Edwardsiella ictaluri, is the cause of enteric septicemia of catfish (ESC), which is one

of the most prevalent bacterial diseases in farm-raised catfish. The objective of this study was to identify risk factorsassociated with ESC mortalities and are reported by farm personnel. To identify risk factors a catfish managementdatabase was developed. The odds ratios (OR) of the final multivariable logistic regression model were: (1) volumeof the pond (OR, 0.56), (2) interval from harvest until a mortality event (OR, 1.49), (3) interval from stocking untila mortality event (OR, 0.52), (4) nitrite measured within 14 d of a mortality (OR, 3.49), (5) total ammonia measuredwithin 14 d of a mortality (OR, 20.48), and (6) sum of feed fed for 14 d prior to the disease outbreak (OR, 1.02),all of which were significantly (P ≤ 0.05) associated with ESC occurrence. This study showed that some commonlyrecorded production variables were associated with ESC outbreaks and if monitored could help identify “at risk”ponds prior to disease outbreaks.

Enteric septicemia of catfish (ESC) is one of the most preva-lent bacterial diseases in commercial catfish production (USDA2003). It is caused by the gram-negative bacterium Edwardsiellaictaluri (Hawke et al. 1981). The epidemiology of ESC can bemultifactorial. Outbreaks usually occur in the spring (April–June) and fall (September–November), months when water tem-peratures are 21–29◦C (70–85◦F) (Tucker et al. 2004). Entericsepticemia of catfish occurs in three forms: acute, subacute, andchronic. In the acute phase catfish show few clinical signs but go

*Corresponding author: [email protected] September 18, 2013; accepted December 17, 2013

off their feed and swim listlessly or become motionless. Infectedfish can have exophthalmia and distended abdomens. The sub-acute phase is characterized by a slower onset but cumulativemortalities may be high (Hawke and Khoo 2004). Catfish willhave petechial and ecchymotic cutaneous hemorrhage on theabdomen and around the head along with small, shallow, whiteor red ulcers. Fish will go off feed more slowly than in acuteESC. Chronic-phase clinical signs may include hemorrhagic ar-eas around the mouth and on the ventral sides of fish. Small

84

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

RISK FACTORS ASSOCIATED WITH ENTERIC SEPTICEMIA OF CATFISH 85

white ulcers may be found on the fish’s skin. Ulcers on thetop of the head and between the eyes are considered pathog-nomonic for the disease and give it one of its common names,hole-in-the-head disease (Tucker et al. 2004). Fish suffer centralnervous system involvement expressed as spinning, spiraling,and tail chasing (Hawke and Khoo 2004). Stress plays a keyrole in outbreaks. Stress factors such as handling, poor diet,poor water quality, overcrowding, and water temperature fluc-tuations can lead to an outbreak (Wise et al. 1993; Plumb andShoemaker 1995). Culturing fish in mixed-age populations orstocking young naive fish in a pond with older fish also plays akey role in the spread of the disease to healthy fish. Survivingfish can carry the pathogen for up to 200 d in their kidney, liver,or brain. Stress may increase the susceptibility to infection andlosses, but it is not a prerequisite for the disease. Immune statusof the individual fish may also determine the outcome (Hawkeand Khoo 2004).

Enteric septicemia of catfish is widespread throughout theU.S. catfish industry. The spread of the disease may be relatedto the shipment of infected but asymptomatic fingerlings. Thesefingerlings may be asymptomatic carriers outside of the temper-ature ranges in which the disease usually occurs (Klesius 1993).Bacteria may be maintained in a multibatch culture environmentwith the introduction of naıve fingerlings to a pond containingolder exposed catfish.

Transmission of ESC between fish is from fecal sheddingfrom sick fish or from the carcasses of dead fish (Earlix 1995).The bacterium can cross the intestinal epithelium, enter theblood stream, and migrate to the kidneys within 15 min ofexperimental intestinal infection (Baldwin and Newton 1993).Vertical transmission from infected broodstock to fry has notbeen demonstrated (Hawke and Khoo 2004). The presence ofviable E. ictaluri in the intestinal contents of cormorants andherons suggest the fecal wastes from piscivorous birds are apotential source of infection (Taylor 1992). However, Waterstratet al. (1999) found no viable E. ictaluri in feces, gastrointestinaltracts, or feathers in experimentally infected great blue heronsArdea herodias and concluded that great blue herons do not playa role in the transmission of ESC between catfish ponds.

Many agricultural industries use production databases to helpimprove production. Feed costs are the largest expense in ic-talurid catfish production. Catfish are fed daily as much as theywill eat during warm months. Catfish are fed to maximize growthand minimize waste because overfeeding can have a negativeeffect on water quality. Monitoring feed intake is an impor-tant management tool. Some catfish producers use a database,FISHY that was developed by the Mississippi State UniversityAgricultural Economics Department to help catfish producersimprove their production-management decision making. TheFISHY database concentrates on feeding and projecting fishgrowth. The catfish management database developed for thisresearch includes data from 2004 to 2007. Not only does thisdatabase allow the farmer to manage feed, stocking, harvesting,and mortality, it can also be used to generate user-defined re-

ports designed to (1) incorporate production data already beingrecorded for generating reports for use at managerial meetingsfocused on feeding rates, feed conversion ratios, mortalities, andharvesting events; (2) be easily used by a catfish farmer to col-lect management data in order to analyze production efficiency;(3) provide the farm with easy access to management reports;and (4) improve a pond’s efficiency and cost of production. Ad-ditional customized reports were generated as requested by thefarm management.

The objective of this study was to identify risk factors associ-ated with ESC mortalities. Of particular interest was determin-ing whether the production variables collected from the farmscould be used to predict the occurrence of ESC mortality events.

METHODSSampling and data collection.—A large commercial catfish

enterprise agreed to share their production records. Over 500ponds from five farms from multiple counties in the MississippiDelta and dedicated to food fish production were included inthis analysis. These ponds had an average size of 5.0 ± 1.66 ha(mean ± SD).

The catfish management database was programmed in Mi-crosoft Access (Microsoft, Redmond Virginia). A permanentunique pond identification number (ID) greater than 1 was as-signed to each pond. Data recorded by the producer from 2004 to2007was imported into the newly constructed database. Briefly,when a mortality event occurred, the date, pond ID, and rea-son or cause of the mortality event as well as biomass, averagesize, and number of fish lost were recorded. Occasionally, af-fected fish were submitted to the Mississippi State UniversityCollege of Veterinary Medicine Diagnostic Laboratory locatedin Stoneville, Mississippi, for laboratory confirmation of ESC.

Pond and production information, later used as explanatoryvariables in statistical models, were recorded or classified intofour main groups: (1) physical characteristics of the ponds,(2) time interval between fish handling and a mortality event,(3) daily feed consumption, and (4) water quality. Physicalcharacteristics of each pond included the surface area (hectares)and average depth (meters), which were used to calculate thevolume in hectare-meters (ha-m). Surface area was determinedfrom GIS files (ESRI, San Antonio, Texas) while pond depthwas recorded as a single point measurement supplied by farmmanagement. Disease-pond age was defined as the age of thepond at the time of a mortality event and was calculated fromthe time of original construction or from the time when thepond was rebuilt. A second group of variables included twocalculated variables. The disease-stocking interval was definedas the interval from the time fish were stocked into the ponduntil a mortality event occurred. Stocking event informationwas recorded and included the source of the fish stocked, datethe pond was stocked, and the number, size, and weight offish stocked. The disease-to-harvest interval was defined as thetime from a harvesting event until a mortality event occurred.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4

86 CUNNINGHAM ET AL.

Harvesting event information was recorded and included thedate of the harvest, the weight, and the number of fish harvested.

The third group of variables dealt with feed consumption.The catfish management database contained the feeding recordsin terms of total kilograms of feed fed for each pond on adaily basis. The total feed was then aggregated for periods of7, 14, 21, and 30 d prior to the ESC-related mortality event.These values did not take into account the varying sizes of theponds. To compare feed usage the aggregate totals were dividedby pond area to calculate feed per hectare, divided by ponddepth to calculate feed per meter of depth, and divided by pondvolume to calculate feed per hectare-meter. These calculationsallowed comparisons between ponds of differing sizes, depth,and volumes.

The fourth group of variables involved water quality mea-surements. A separate water quality database was developedand located in a water quality testing laboratory on the farm.All testing for pond water quality variables was performed inthe central laboratory by one technician. Pond water was testedfor total ammonia nitrogen (TAN), nitrite, and chloride. Chlo-ride was measured only if the TAN level was considered high(>6 mg/L). The database was designed to automatically gen-erate a report of ponds that exceeded the management-definedtotal chloride-to-nitrite ratio. The water quality database wasconstructed in 2005 so data from 2005 to 2007 were included inthe analysis. Water quality data were collected on a weekly orbiweekly basis during the growing season (March–November)and monthly during the nongrowing season.

An observation was defined as a pond with a positive mortal-ity event due to ESC as defined by the farm management. Pondswithout a history of mortality events associated with ESC wereselected as negative ponds and served as controls. A negativepond was defined as a pond that did not have a mortality event60 d before and 60 d after the mortality event date beinganalyzed.

Statistical procedures: risk-factor modeling, variableselection.—Logistic regression was used to assess the strengthof association between the dichotomous outcome of interest,ESC occurrence in ponds, and various independent variablesthat represented potential risk factors for the disease. The datain the study was hierarchically structured, which called for mul-tilevel modeling (Guo and Hongxin 2000) with ponds (level 1)nested in farms (level 2) that are nested in the catfish enterprise(level 3). Biases in variable estimates could result from ignoringobservations that are more highly correlated and within clustersor levels. Linear and binary models underestimate standard er-rors when clustering is not taken into account and the assumptionof independence is violated. Multilevel modeling provides moreaccurate standard errors, confidence intervals, and significancetests by correcting these biases (Guo and Hongxin 2000). Gener-alized linear mixed models designating a binomial distributionwith a logit link function were fitted using the GLIMMIX pro-cedure in SAS 9.2 software for Windows (SAS Institute, Cary,North Carolina) to conduct the logistic regression analysis. Ran-

dom effects were incorporated to account for the repeated mea-sures of ponds and variability among the participating farms andpossible intrafarm correlation. In the screening process, eachrisk factor was evaluated in the basic model as a single fixed-effects factor, and if associated with the outcome (P ≤ 0.20)was retained for further analysis. In the second step, all contin-uous variables considered as risk factors were retained from thescreening step and investigated for pairwise collinearity usingthe CORR procedure in SAS for Windows version 9.2. Eachcase of collinearity, defined at R ≥ 0.6, detected was evaluatedseparately on the significance of the association with the occur-rence of ESC and relationship with other explanatory variables.

To build the final multivariable model, the fixed-effects riskfactors retained from the screening and collinearity investiga-tions were offered to the basic model all at once as fixed-effectsfactors. After each model run, the fixed-effects factor with thehighest P-value was removed until a final model with all thefixed-effects variables significant at P ≤ 0.050 was developed.Further refinement of the developed full final model was pur-sued to obtain the most parsimonious model while preservingits explanatory ability. A limited number of tools are availableto evaluate the performance of generalized linear mixed modelswith a different set of predictors for a given outcome. There isno best way to estimate goodness of fit for multilevel models.In nonmultilevel logistic regressions, the chi-square goodness-of-fit test is appropriate when an assumption of independenceof observations and data are not very sparse (Schukken et al.2003). These assumptions are not met in multilevel modelingwith clustering so the chi-square test of goodness of fit is notappropriate.

The models were compared using the Akaike informationcriterion (AIC) score. Models that had AIC score differences ofgreater than 2 from the model with the minimum AIC score wereeliminated from the analysis (Burnham and Anderson 2001).Models with the lowest AIC scores were selected as the finalmodel. In the descriptive statistics, means were reported withtheir SD values. A strength of association between variables wasreported as the odds ratio (OR).

RESULTSIn ESC-related mortality events, mean losses were

4,053 ± 224 individuals, at 0.6 ± 0.01 kg per fish for a totalweight of 2,303 ± 120 kg per mortality event. Enteric septicemiaof catfish accounted for 18.91% of the observed mortalities from2004 to 2007. On a monthly basis, farm-recorded ESC mortali-ties peaked in September and October (Figure 1). Average pondsurface area was 4.65 ± 1.52 ha (range, 4.4–4.8 ha) and ponddepth was 1.76 ± 0.33 m (range, 1.71–1.81 m). This farm hasundergone a very aggressive pond rebuilding program over thelast 3 years with newer rebuilt ponds being deeper. Averagepond volume was 8.48 ± 3.84 ha-m (range, 7.9–9.1 ha-m).

The screening process for the data set identified 27 variablesthat had an association (P ≤ 0.2) with the occurrence of

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4


0% 0% 0% 0% 3%

9%

4%

5%

25%

46%

2%

5%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec

FIGURE 1. Farm-reported percent of ESC mortalities in catfish (number of cases per month, 2004–2007).

ESC and were considered as candidates in a multivariablemodel (Table 1). During the variable screening process 1,215observations were identified.

The most parsimonious multivariable model included six ef-fects in the final model: (1) pond volume, (2) interval fromstocking until a mortality event, (3) interval from harvest until amortality event, (4) nitrite measured within 14 d of a mortalityevent, (5) TAN measured within 14 d of a mortality event, and(6) sum of feed fed for 14 d prior to the disease outbreak. Aspond volume and the interval from stocking to a mortality eventincreased, the odds of a mortality event associated with ESCoccurring decreased. As the interval from harvest to a mortal-ity event, nitrite and ammonia levels within 14 d, and sum offeed fed for 14 d prior to a mortality event increased, the oddsof a mortality event associated with ESC occurring in a pondincreased (Table 2). For the ponds included in this analysis, themean ± SD pond surface area was 4.8 ± 1.60 ha and pond depthwas 1.84 ± 0.364 m.

DISCUSSIONThe objective of this study was to identify potential risk fac-

tors associated with mortality events that the farm managersattributed to ESC. Wagner et al. (2002) using the 1997 NationalAnimal Health Monitoring System (NAHMS) survey of cat-fish farmers found that the most frequently reported (35.7%)average loss per outbreak of ESC and columnaris combinedwas 90–900 kg (200–2,000 lb) per outbreak. Only 18.3% ofthe operations reported losses classified as severe (>900 kg). Incontrast, the ESC-related mortality events in this study resultedin losses (mean ± SD) of 3,156 ± 3,317 fish, at 0.5 ± 0.02 kg

per fish for a total weight of 1,615 ± 1,835 kg per mortalityevent. Enteric septicemia of catfish accounted for 18.91% of theobserved mortalities from 2004 to 2007. Wagner et al. (2002)found that 78.1% of the farms surveyed and 42.1% of all pondsexperienced ESC or columnaris problems.

In the model pond volume was significantly associated withESC occurrence. Ponds with more volume had reduced oddsof a mortality event associated with ESC. Since depth is a keycomponent of volume (area × depth) this result in not unex-pected. Hanson et al. (2008) found that as pond depth increased,catfish losses from weather-related causes decreased, becausethe deeper ponds were not as sensitive to windstorms, droughts,and freezing. In contrast, Cunningham et al. (2012) found thatthe incidence of columnaris increased as pond depth increased.Greater pond depth offers more habitat for the catfish; shallowerponds or older ponds that have filled in (Steeby et al. 2004) pro-vide less space and may lead to crowding and increased stresson the catfish. Pond depth data were a single measurement re-ported by the farm. However, catfish ponds are sloped and areshallower at the margins and/or at one end and deeper at theopposite end. Pond depth can be influenced by the age of thepond and sediment accumulation. Increased stress can lead togreater chance of disease occurring; deeper ponds may reducethis stress leading to reduced odds of a disease occurring.

As the stocking-to-disease interval increased, the odds of adisease outbreak associated with ESC decreased. Therefore, adecreased stocking-to-disease interval would be associated withincreased odds of ESC, suggesting that contaminated equip-ment used in stocking or stress due to the stocking event couldhave contributed to disease occurrence. The odds of a mortalityevent due to ESC increased as the harvest-to-disease interval

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4


TABLE 1. Logistic regression analysis results for variables associated with the occurrence of ESC in catfish (P-values < 0.20). ha-m = hectare-meter.

Measured unit Odds ConfidenceVariable or comparison N ratio interval P-value

Depth m 1,215 0.11 0.03–0.47 0.0027Volume ha- 1,215 0.78 0.66–0.91 0.0019Size ha 1,216 0.68 0.51–0.92 0.0124Year year 1,216 <0.0001

2005 versus 2007 0.02 0.01–0.052006 versus 2007 0.37 0.28–0.49

Disease-pond age d 1,215 1.23 1.15–1.32 <0.0001Prediseasefeed sum kg 1,215 3.14 2.54–3.86 <0.0001Disease-harvest interval d 1,215 1.47 1.40–1.53 <0.0001Disease-stock interval d 1,215 0.44 0.30–0.67 <0.0001Nitrites, 8–14 d mg/L 1,215 3.44 2.06–5.74 <0.0001Ammonia, 8–14 d mg/L 1,215 7.41 4.13–13.28 <0.0001Nitrites, 14–21 d mg/L 192 7.48 0.95–58.82 0.056Feed, 0–7 d kg 1,215 1.00 1.00–1.00 0.1058Feed, 0–14 d kg 1,215 1.00 1.00–1.00 0.0127Feed, 0–21 d kg 1,215 1.00 1.00–1.00 <0.0001Feed, 0–30 d kg 1,215 1.00 1.00–1.001 <0.0001Feed, 0–7 d/ha kg/ha 1,215 1.00 0.99–1.00 0.1230Feed, 0–14 d/ ha kg/ha 1,215 1.00 1.00–1.001 0.0167Feed, 0–21 d/ha kg/ha 1,215 1.00 1.00–1.00 <0.0001Feed, 0–30 d/ha kg/ha 1,215 1.001 1.001–1.001 <0.0001Feed, 0–7 d/ha-m kg/ha-m 1,215 0.999 0.998–1.000 0.1297Feed, 0–14 d/ha-m kg/ha-m 1,215 1.001 1.000–1.001 0.0058Feed, 0–21 d/ha-m kg/ha-m 1,215 1.001 1.000–1.001 <0.0001Feed, 0–30 d/ha-m kg/ha-m 1,215 1.002 1.002–1.002 <0.0001Feed, 0–7 d/ha-m kg/m 1,215 1.000 1.000–1.000 0.1606Feed, 0–14 d/ha-m kg/m 1,215 1.000 1.000–1.000 0.0022Feed, 0–21 d/ha-m kg/m 1,215 1.000 1.000–1.000 <0.0001Feed, 0–30 d/ha-m kg/m 1,215 1.000 1.000–1.000 <0.0001

increased. This could be caused by fewer fish in the pond afterharvest leading to decreased fish density, which would lowerthe odds of an ESC break. As the pond is restocked and thefish density is increased, stress will also increase, increasing therisk of an ESC break. These intervals could be used as indirectindicators of fish handling stress or the use of contaminatedequipment.

Increased total feed fed increased the odds of a disease out-break associated with ESC. Catfish ponds have a finite capacityto process waste without affecting water quality. Water qualityproblems including low dissolved oxygen (DO) will increasein severity and frequency if feed exceeds the waste-processingcapacity of a pond. Catfish ponds fed at a high rate, defined asa maximum of 78 kg/ha, had lower DO levels at dawn, reduced

TABLE 2. Odds ratio (OR) of the final logistic regression model.

Variables Measured unit OR Confidence interval P-value

Volume ha-m 0.56 0.42–0.74 <0.0001Disease-stocking interval d 0.52 0.34–0.81 0.0035Disease-to-harvest interval d 1.49 1.41–1.57 <0.0001Nitrites, 14 d mg/L 3.49 1.66–7.33 0.0010Total ammonia, 14 d mg/L 20.48 9.96–42.11 <0.0001Feed, 0–14 d 100 kg 1.02 1.01–1.03 <0.0001

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4


growth rate, poorer feed conversion, and increased mortalitycompared with medium (56 kg/ha) and low (34 kg/ha) feedingrates (Tucker and Boyd 1979). In 50% of the ponds fed at thehigher rate the mortality rate ranged from 7% to 32%. Cole andBoyd (1986) found that net fish production increased in propor-tion to feed fed up to 112 kg/ha per day but then decreased athigher feeding rates. Feed conversion (increased feed fed perunit of weight gain) was constant when feed fed was between28 and 112 kg/ha but quickly increased at higher feeding lev-els (>112 kg/ha per day) at which point fish did not consumeall of their feed, resulting in increased waste accumulation anddecreased water quality

The odds of a pond having an ESC outbreak were 3.49 timesgreater for each one unit increase in nitrite measured 14 d prior toa disease event and 20.48 times greater for each one unit increasein TAN levels measured for this same time period. Water qualitymeasures that potentially affect fish health include nitrite, am-monia, and oxygen levels. High nitrite can result from overfeed-ing and decomposition of organic materials. Therefore, routinemonitoring of these levels in ponds is considered to be an es-sential best management practice (BMP) towards the preventionof mortalities due to toxic levels. Water quality data were col-lected on a weekly or biweekly basis during the growing season(March–November) and monthly during the nongrowing sea-son. Weekly measurements give the farm time to identify pondsat risk. The addition of salt (NaCl) to ponds is a common man-agement practice aimed towards the treatment and prevention ofdisease. Elevated ammonia levels can cause physiological, bio-chemical, histological, and behavioral effects in fish. Un-ionizedammonia (NH3) is excreted by passive diffusion from the gillsof Channel Catfish Ictalurus punctatus. The pH and temperatureof the blood determine the proportion of TAN (NH3 + NH4

+ )that is partitioned between ionized and un-ionized forms. Gillepithelium diffusion of NH3 is a function of water pH, plasmapH, and total ammonia concentration (Hargreaves and Tomasso2004). High levels of NH3 in water will cause decreased diffu-sion resulting in increased levels in plasma. Low DO levels canincrease the effect of high ammonia levels. Although ammoniaconcentrations that cause death are seldom observed in catfishponds, sublethal effects such as compromised immune statusand reduced growth rate are observed. As oxygen levels de-crease even low levels of total ammonia (0.43 mg/L) can reducevoluntary feed consumption by 68% (Hargreaves and Tomasso2004). The chloride-to-nitrite ratio is important to determinemethemoglobin levels in catfish. Nitrite from the pond water isactively transported to the catfish circulatory system producinga life threatening condition known as brown blood disease ormethemoglobinemia (Durborrow et al. 1997). Nitrite : chlorideratios of 20:1 or greater are recommended. Lower ratios canresult in brown blood disease (Hargreaves and Tomasso 2004).The database was designed to automatically generate a reportof ponds that do not meet the management-defined chloride :nitrite ratio. This ratio is important because the potential toxicityof nitrite is reduced by increasing the chloride concentration of

the pond water by adding salt. Chloride : nitrite–nitrogen ratiosof 30:1 allow little nitrite to enter the catfish’s blood stream,but producers routinely maintain pond water chloride concen-trations at 100–150 mg/L to maintain a safety margin (Tuckerand Hargreaves 2004).

The data were used in the management of the catfish farmand assumed accurate. Compared with a prospective study adisadvantage of this retrospective study was that key variablessuch as pond DO levels, temperature, and pH were not recorded.The farm recorded DO content on each pond up to eight timesper night. They did not record these observations (>3,500 pernight) due to their high number. They responded to any low DOevents by aeration. Pond water temperature was available froma nearby weather station, but we did not use it because the watertemperature would be the same in ponds with and without ESCoutbreaks. The variables described in this study are associatedwith ESC mortality events but do not necessarily cause ESC.They are, however, good variables to consider when designingcontrolled experiments to determine which risk factors actuallypredispose a pond to ESC-associated mortalities. The modeland methodology developed for this study may be useful for theinvestigation of additional economically important catfish dis-eases. This study showed some commonly recorded productionvariables (feed consumption, pond size and depth, nitrite lev-els, and stocking events) were associated with ESC-associatedmortalities and if monitored could help identify “at risk” pondsprior to ESC outbreaks.

ACKNOWLEDGMENTSRichard Randle, Al Camus, and Terry Hanson are thanked

for their support of this project. Mention of trade names or com-mercial products is solely for the purpose of providing specificinformation and does not imply recommendation or endorse-ment by the U.S. Department of Agriculture.

REFERENCESBaldwin, T. J., and J. C. Newton. 1993. Early events in the pathogenesis of en-

teric septicemia of Channel Catfish caused by Edwarsiella ictulari: light andelectron microscope and bacteriologic findings. Journal of Aquatic AnimalHealth 5:189–198.

Burnham, K., and D. Anderson. 2001. Kullback–Leibler information as a basisfor strong inference in ecological studies. Wildlife Research 28:111–119.

Cole, B. A., and C. E. Boyd. 1986. Feeding rate, water quality and ChannelCatfish production in ponds. Progressive Fish-Culturist 48:25–29.

Cunningham, F. L., S. W. Jack, D. Hardin, and R. W. Wills. 2012. Pond-levelrisk factors associated with columnaris disease on Mississippi commercialcatfish farms. Journal of Aquatic Animal Health 24:178–184.

Durborrow, R. M., D. M. Crosby, and M. W. Bruson. 1997. Ammonia in fishponds. Southern Regional Aquaculture Center, Publication 463, Stoneville,Mississippi.

Earlix, D. J. 1995. Host, pathogen, and environmental interactions of en-teric septicemia of catfish. Doctoral dissertation. Auburn University, Auburn,Alabama.

Guo, G., and Z. Hongxin. 2000. Multilevel modeling for binary data. AnnualReview of Sociology 26:441–462.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4


Hanson, T., S. Shaik, K. Coble, S. Edwards, and J. Miller. 2008. Identifying riskfactors affecting weather and disease related losses in the U.S. farm raisedcatfish Industry. Agriculture and Resource Economic Review 37:27–40.

Hargreaves, J., and J. R. Tomasso. 2004. Environmental Biology. Pages 52–57in J. A. Hargreaves and C. S. Tucker, editor. Biology and culture of ChannelCatfish. Elsevier, Amsterdam.

Hawke, J. P., and L. H. Khoo. 2004. Infectious disease. Pages 387–443 in C. S.Tucker and J. A. Hargreaves, editors. Biology and culture of Channel Catfish.Elsevier, Amsterdam.

Hawke, J. P., A. C. McWhorter, A. G. Steigerwalt, and D. J. Brenner. 1981.Edwardsiella icataluri sp. nov., the causative agent of enteric septicemia ofcatfish. International Journal of Systematic Bacteriology 31:396–400.

Klesius, P. H. 1993. Rapid enzyme-linked immunosorbent tests for detectingantibodies to Edwardsiella ictaluri using exoantigen. Veterinary Immunologyand Immunopathology 36:359–368.

Plumb, J. A., and C. Shoemaker. 1995. Effects of temperature and salt concen-tration on latent Edwarsiella ictulari infections in Channel Catfish. Diseasesof Aquatic Organisms 21:171–175.

Schukken, Y. H., Y. T. Grohn, B. McDermott, and J. J. McDermott 2003. Anal-ysis of correlated discrete observations: background, examples and solutions.Preventive Veterinary Medicine 59:223–240.

Steeby, J. A., J. Hargreaves, C. S. Tucker, and S. Kingsbury. 2004. Accumula-tion, organic carbon and dry matter concentration of sediment in commercialChannel Catfish ponds. Aquaculture Engineering 30:115–126.

Taylor, P. W. 1992. Fish eating birds as potential vectors of Edwardsiella ictulari.Journal of Aquatic Health 4:240–243.

Tucker, C., J. Avery, C. Engle, and A. Goodwin. 2004. Industry profile: pond-raised Channel Catfish. Mississippi State University, Mississippi State.

Tucker, C., and J. Hargreaves. 2004. Pond water quality. Pages 271 in C. S.Tucker and J. A. Hargreaves, editors. Biology and culture of Channel Catfish.Elsevier, Amsterdam.

Tucker, L., and C. E. Boyd. 1979. Effects of feeding rate on water quality,production of Channel Catfish and economic returns. Transactions of theAmerican Fisheries Society 108:389–396.

USDA (U.S. Department of Agriculture). 2003. NAHMS catfish 2003 part II:reference of food size catfish health and production practices in the UnitedStates. USDA, Animal and Plant Health Inspection Service, National AnimalHealth Monitoring System, Washington, D.C.

Wagner, B., D. Wise, L. Khoo, and J. Terhune. 2002. The epidemiology ofbacterial diseases in food-size Channel Catfish. Journal of Aquatic AnimalHealth 14:263–272.

Waterstrat, P. R., B. Dorr, J. F. Glahn, and M. E. Tobin. 1999. Recovery andviability of Edwardsiella ictaluri from great blue herons Ardea herodias fedE. ictaluri-infected Channel Catfish Ictalurus punctatus fingerlings. Journalof the World Aquaculture Society 30:115–122.

Wise, D. J., T. E. Schwedler, and D. L. Otis. 1993. Effects of stress on suscep-tibility of naıve Channel Catfish in immersion challenge with Edwardsiellaictaluri. Journal of Aquatic Animal Health 5:92–98.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:51

15

June

201

4



Development of a Nonlethal Health Assessment for WildRed Drum Using a Health IndexCarla M. Bourtisa, Ruth Francis-Floydb, Eric A. Reyiera, Roy P. Yanongc & Louis J. GuilletteJr.d

a Kennedy Space Center Ecological Program, InoMedic Health Applications, Mail CodeIHA-300, Kennedy Space Center, Florida 32899, USAb Department of Large Animal Clinical Sciences, College of Veterinary Medicine, Universityof Florida, Post Office 100136, Gainesville, Florida 32610-0136, USAc Tropical Aquaculture Laboratory, Program in Fisheries and Aquatic Sciences, School ofForest Resources and Conservation, University of Florida, 1408 24th Street Southeast,Ruskin, Florida 33570, USAd Department of Obstetrics and Gynecology and the Hollings Marine Laboratory, MedicalUniversity of South Carolina, 221 Fort Johnson Road, Charleston, South Carolina 29412, USAPublished online: 04 Jun 2014.

To cite this article: Carla M. Bourtis, Ruth Francis-Floyd, Eric A. Reyier, Roy P. Yanong & Louis J. Guillette Jr. (2014)Development of a Nonlethal Health Assessment for Wild Red Drum Using a Health Index, Journal of Aquatic Animal Health,26:2, 91-95, DOI: 10.1080/08997659.2014.886633







http://dx.doi.org/10.1080/08997659.2014.886633




COMMUNICATION

Development of a Nonlethal Health Assessmentfor Wild Red Drum Using a Health Index

Carla M. Bourtis*Kennedy Space Center Ecological Program, InoMedic Health Applications, Mail Code IHA-300,Kennedy Space Center, Florida 32899, USA

Ruth Francis-FloydDepartment of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida,Post Office 100136, Gainesville, Florida 32610-0136, USA

Eric A. ReyierKennedy Space Center Ecological Program, InoMedic Health Applications, Mail Code IHA-300,Kennedy Space Center, Florida 32899, USA

Roy P. YanongTropical Aquaculture Laboratory, Program in Fisheries and Aquatic Sciences,School of Forest Resources and Conservation, University of Florida,1408 24th Street Southeast, Ruskin, Florida 33570, USA

Louis J. Guillette Jr.Department of Obstetrics and Gynecology and the Hollings Marine Laboratory,Medical University of South Carolina, 221 Fort Johnson Road, Charleston, South Carolina 29412, USA

AbstractNonlethal methods are needed to assess the health of wild fish

and quantify the robustness of the broader population. Resultscould be used to indicate exposure to various stressors, such as con-taminants, infectious disease, external parasite loads, and fishingpressure, to monitor changes in fish population health over time.The wild Red Drum Sciaenops ocellatus population in the KennedySpace Center Reserve of Merritt Island National Wildlife Refugewas used to develop a protocol to define the health of free-rangingfish using nonlethal techniques. This health index incorporatedmorphometric measurements, weight, an evaluation for externalparasite fauna, notation of physical deformities, and the presenceof lesions. A total of 126 adult Red Drum were collected using hook-and-line angling during prespawning (May), spawning (Septemberand October), and postspawning (December) periods. All fish werereleased alive back into their environment. The nonlethal healthassessment scored fish in the “healthy” range of the health indexduring the prespawning and spawning periods. Fish caught duringthe postspawning period scored slightly below this range. Parasiteload contributed to the depressed score during the postspawning

*Corresponding author: [email protected] June 19, 2013; accepted January 9, 2014

period. Fish collected in all sampling periods were rated on averageas “excellent” for condition factor, which suggests that the sampledpopulation in the reserve were thriving.

Wild fish health can be impacted by anthropogenic factorssuch as stress from catch-and-release fishing and exposure toand accumulation of toxins, as well as by natural factors such asinfectious agents, suboptimal habitat, lack of food, or overpre-dation. The definition of “health” for wild populations of fishis a nebulous term. Methods of nonlethal health assessment areneeded to improve our understanding of the biology and overallhealth status of wild populations. Monitoring health trends ofmarine animal populations as sentinel species is a useful tool forthe evaluation of the well-being of aquatic ecosystems (Bossart2011). Nonlethal assessments allow for long-term sampling ofthe same animal, when tagged with an identification number,facilitating detection of changes over time. Representative sam-ple sizes can be used so the health status of the population

91

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

92 BOURTIS ET AL.

can be extrapolated from data collected on individual animals.The goal of this study was to develop nonlethal techniques forthe assessment of the health of a population of wild adult RedDrum Sciaenops ocellatus inside NASA’s Kennedy Space Cen-ter (KSC) Reserve of Merritt Island National Wildlife Refuge(MINWR). The objectives of this project were to evaluate thepotential use of the health index developed by the Stock En-hancement Research Facility (SERF) on a wild population ofRed Drum inside the KSC Reserve. Health can be defined as thelevel of functional or metabolic efficiency of a living organism.The health data collected includes presence or absence of dis-ease, morphometric measurements, and physiological metrics.Red Drum is an economically important sport fish in all of itsrange. They have a sedentary lifestyle from juvenile throughadult stages, and can remain in the same estuarine habitat for upto 5 years (Evans et al. 1997). Nonlethal health assessmentshave not been carried out on Red Drum in Florida despitethe presence of the significant recreational catch-and-releasefishery.

Development of field techniques for the assessment of fishhealth is needed for fisheries management and environmentalmonitoring. An assessment of fish health can provide informa-tion used to link the effects of human activities to ecologicalchange (Schlacher et al. 2007). To determine the health statusof an individual or a population, normal functional physiol-ogy and disturbed or pathological physiology must be distin-guished (Aguirre and Lutz 2004). Baseline data from unfishedstocks can vastly improve estimates of population parametersfor harvested species by having a baseline data source withwhich to compare data (Murray et al. 1999). Data obtainedfrom a nonimpacted location study site, such as a reserve withno fishing allowed, could function as a standard for evaluat-ing other sites (Leamon et al. 2000). The KSC Reserve hasbeen a defacto reserve due to security restrictions in the Ba-nana River Lagoon since 1962, making it an ideal study sitefor the development of baseline data for sport fish such as RedDrum.

Stock assessments in Florida are currently conducted by theFlorida Fish and Wildlife Conservation Commission (FFWCC).Independent monitoring assessments of the fisheries consist ofmonthly sampling of Red Drum from both coasts using a varietyof fishing gear and targeting different size-classes. These assess-ments have been conducted in the northern Indian River Lagoonsystem since 1996. Nonlethal assessments are conducted on fishthat do not have obvious gross lesions. These fish are measuredand released; however, minimal data are collected to evaluatehealth. Fish with gross external abnormalities are culled fornecropsy, determination of reproductive status, age and growthanalysis, and evaluation of pathology. Red Drum stock assess-ments can be improved to collect more information while de-creasing the need for lethal harvest. Current methods allow ageestimates from length and weight relationships, and blood hor-mone analysis can be used for sex determination (Kucherkaet al. 2006), decreasing the need to cull fish for the collection ofbasic biological data.

The SERF, operated by FFWCC, has developed a health in-dex for cultured juvenile Red Drum. This health index couldbe completed without lethal harvest; however, as presently con-ducted, fish are euthanized so that complete necropsies are doneon all fish tested. The health index metric assesses the exter-nal condition of the fish using quantitative and qualitative databased on criteria developed by A. K. Dukeman, C. M. Arm-strong, and C. M. Stephenson (abstract of a poster presented atthe 57th Annual Gulf and Caribbean Fisheries Institute, 2006).This health index score ranges from 0 to 50 based upon externalcondition. When a Red Drum scores a value below 45, its healthis considered compromised; by contrast, a score of 45 and aboveis considered in the healthy range.

METHODSFish were caught inside the KSC Reserve (28◦32′59.76′′N,

80◦35′40.38′′W) using a hook and line with barbless circlehooks and pieces of cut mullet (Mugil spp.) as bait. Only fishover 650 mm SL were used to target mature adults. Three re-productive periods were targeted for sample collection during2011: prespawning (May), spawning (September and October),and postspawning (December). Measurements of the SL (cm),TL (cm), and weight were recorded. Unanesthetized fish weremanually restrained briefly out of the water on an inverted V-tray as a small section of gill was cut from the outer branchialarch and the assessment was carried out. The eyes, mouth, exter-nal mucous layer, scales, and the fins of the fish were evaluatedgrossly for pathogen presence. Physical deformities were noted,along with the presence of any lesions that could indicate bacte-rial, fungal, or other infections (e.g., presence of ulcers) and weresubsequently ranked according to severity. A skin mucus samplewas collected using a plastic microscope slide cover to lightlyscrape along the lateral line starting at the anterior end of the fish.Wet mounts of these tissues were immediately examined under amicroscope at a magnification of 100 × for the presence of exter-nal parasite fauna, and the number of organisms per sample wasrecorded. Microscopic parasites were not observed during fieldexaminations. External parameters were recorded and scaled ac-cording to severity using the altered SERF health index metricsmodified for wild fish; a total score of 27 was possible. For ourhealth index, a fish with a score of 23 was considered healthy(Table 1). A PIT tag (Biomark; 12 mm, 125 kHz) was insertedusing a sterile12-gauge needle placed subcutaneously near thecheek on the left side. All fish were released immediately af-ter examination was completed. Examinations took less than15 min from hooking time and no known mortality occurred.

Field attempts to identify the sex of collected fish includedlistening for the fish’s drumming sound and applying pressure tothe abdomen to determine whether milt was exuded during thefall spawning period. Both techniques would indicate a male.Suspected females were probed in the vent area with a pipetteto locate the urinary orifice, anus, and ovipositor. If the sexwas not clearly distinguished, then the fish was classified as“undetermined.”

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

COMMUNICATION 93

TABLE 1. Health index developed for wild Red Drum. A score of 23 or greater is considered representative of “healthy” fish. A standard condition factorequation was used for the calculation (see Methods).

Variable Variable condition Point value

Condition factor × 5 10Eyes Normal 1

Abnormal 0Mouth Normal 1

Abnormal 0Mucous Normal thick layer, edges of scales not detected 1

Edges of scales protruding, fish feels “dry” 0Skin lesions None 1

One or more pathogens 0Gill parasites No parasites present 1

Parasites present 0Gill parasite number None present 2

One species present 1Two or more species present 0

Skin/fin parasite No parasites present 1Parasites present 0

Skin/fin parasite number None present 2One species present 1Two or more species present 0

Skin/fin parasite load None 3Low 2Moderate 1Heavy 0

Scale loss None 1Minor loss 0

Wounds None 2Skin ulcer or erosion, not affecting muscle tissue 1Skin ulcer or erosion penetrating into muscle tissue 0

Other physical abnormalities None 1Skeletal, fin deformity, missing opercula, other 0

Health index score 27

Treatments consisted of the three sampling periods (pres-pawning, spawning, and postspawning). Condition factor wascalculated for all fish using the equation: [fish weight (g) × 100]/SL3 (Fulton 1904) to determine the total for the health index.Health index data did not pass the normality test (Shapiro–Wilk:P = 0.03, 0.02, 0.02), so data were log transformed, and aone-way ANOVA was used to test for differences among thethree sampling periods. A P-value was considered significantat P < 0.05. Statistical analyses were performed with SigmaPlot version 11.0 (Systat Software, San Jose, California).

RESULTS

Morphometrics and Condition FactorA total of 126 fish were caught during the three sampling pe-

riods. Using barbless circle hooks, we were able to catch trophy-

size Red Drum with ease, suggesting that barbed hooks maybe unnecessary for catching these fish. Males used in this studywere all above the 50% maturity length of 511 mm FL since thesampling cutoff was set at 650 mm SL; however, Atlantic coastfemales have a 50% maturity length of 900 mm FL, and only11% of the females sampled in this study were over 900 mm TL.Sex was undetermined in only 35 fish. Fish caught outside of thespawning season were more difficult to determine without theuse of a blood sample for hormone analysis or internal gonadidentification. Fish (n = 38) collected in the prespawning periodin May 2011 had a TL of 774 ± 96.2 mm (mean ± SD), an av-erage weight of 5.39 ± 2.72 kg, and an average condition factorof 1.6 ± 0.10. This placed them in the excellent category. Fish(n = 45) collected during the spawning period in September andOctober 2011 had an average TL of 797 ± 79 mm, an averageweight of 6.12 ± 2.79 kg, and an average condition factor of

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

94 BOURTIS ET AL.

0

5

10

15

20

25

30

Pre-Spawning Spawning Post- Spawning

Hea

lth In

dex

Sco

re

Reproductive Period

FIGURE 1. Health index score (mean ± SD) of wild-caught Red Drum in2011 presented by sampling period: prespawning, May (n = 38); spawning,September–October (n = 46), and postspawning, December (n = 42).

1.72 ± 0.18. They were classified in the excellent category aswell. Fish (n = 42) collected during the postspawning periodin December 2011 had a TL of 847 ± 79 mm, an averageweight of 7.64 ± 2.71 kg, and an average condition factor of1.77 ± 0.17. They were also placed in the excellent category.

Health IndexUsing the modified SERF health index, the Red Drum ob-

tained during this study exhibited average scores above 23 dur-ing the prespawning and spawning periods, which is consideredthe threshold for healthy (Figure 1). The average score decreasedto 19 in postspawning fish. A one-way ANOVA was used to testfor differences among the three periods of reproduction (F2, 123

= 36.58, P = 0.14). A post hoc Tukey’s test showed the repro-ductive periods, spawning versus postspawning, differed sig-nificantly at P < 0.05, and prespawning versus postspawningdiffered significantly at P < 0.05. The spawning versus pres-pawning periods were not significantly different at P = 0.25.

ParasitesThe majority of fish (90%) sampled in this study had three

types of grossly visible external parasites identified as Argulussp., Caligus sp., and Ergasilus sp. One Argulus sp. per fishon average was observed on the ventral surface. Caligus sp.averaged 15 per fish and were found on the ventral surface.A mean of one Ergasilus sp. per fish was found inside theoperculum of each fish.

DISCUSSIONThis study defined parameters that can be used to assess the

health of wild Red Drum in the field. Fish scored in the healthyrange of the health index during the prespawning and spawningperiods. Fish caught in the postspawning period scored slightlybelow this range. A factor that contributed to the depressedscore during the postspawning period was parasite load. TheSERF health index was developed to assess cultured juvenile

Red Drum. These fish are not expected to have parasites; bycontrast, over 90% of wild fish sampled had external parasites.This is not surprising when evaluating wild fish, and Landsberget al. (1998) suggested that parasites on wild fish are indicativeof healthy ecosystems. Wild fish are able to sustain a light loadof parasites, and that is expected in a natural setting. However,excessive parasite loads on a fish suggest a compromised im-mune system and can be indicative of compromised health. Allfish with the three types of parasites observed in this study werestill robust in length and weight and had excellent conditionfactor scores.

Fish collected during the postspawning sampling period hadthe highest average condition factor suggesting that there wasminimal spawning condition deterioration in this population.Fish collected in all sampling periods averaged in the excellentcategory, i.e., condition factor over 1.6, which gives the im-pression the sampled population in the reserve were thriving.Johnson (1999) demonstrated that sport fish within the KSCReserve achieve a larger mean size and exist in higher densitiesthan those in adjacent public areas.

Future ResearchOne effective way to understand the impacts of human activ-

ities on natural systems is to have reference areas that have hadminimal human impact (Bohnsack 1993) for comparison. Ref-erence areas help resource managers detect changes in healthand also help them distinguish whether these changes are nat-ural or caused by human actions. The KSC Reserve serves asan excellent reference area and provides baseline data for thehealth assessments of adult Red Drum. The next step for a betterunderstanding Red Drum health is to use this modified healthindex to compare fish from inside the KSC reserve with thosein surrounding areas, including other nearby heavily fished wa-ters, such as Mosquito Lagoon, which is part of the Indian RiverLagoon system but open to fishing by the public year-round.Red Drum are known to tolerate repeated catch-and-releaseevents from recreational fishing, although there may be somepostrelease mortality; but how these events affect overall healthand reproductive potential of a fish is unknown. Fish seem torecover more slowly during the spawning season and mightbe more susceptible to higher predation or decreased spawn-ing success (Vecchio and Wenner 2007). An acoustic telemetrystudy on adult Red Drum in Mosquito Lagoon had a conserva-tively estimated angler recapture rate of 41%, which far exceedsmost other mark–recapture studies (Reyier et al. 2011). Bohn-sack (2011) reported Red Drum had 55% of recreational worldrecords near the KSC Reserve.

This proposed approach could positively impact the fisheryby decreasing the need for FFWCC to cull fish for collectionof data needed to manage it. There is an existing database onlength–weight relationships relative to both the age of Red Drumand gonadosomatic index data. The health parameters presentedhere provide some of the baseline data necessary to monitor

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

COMMUNICATION 95

the health of Red Drum in a nonlethal manner in the future.Implementing nonlethal methods for health assessments wouldbe a mutually suitable arrangement for fisheries managementand recreational anglers.

ACKNOWLEDGMENTSWe thank Doug Scheidt, Russ Lowers, Karen Holloway-

Adkins, Shanon Gann, Carlton Hall, Donna Oddy, and TimKozusko for their assistance in field sampling. This research wasfunded in part by InoMedic Health Applications, the AquaticAnimal Health Program of the University of Florida, the MedicalUniversity of South Carolina and Hollings Marine Laboratory.All research was performed under the authorization of the spe-cial activity license SAL-09-0512A0SR. All fish handling wasin accordance with the Merritt Island National Wildlife RefugeSpecial Use Permit 2011SUP001 and NASA’s Kennedy SpaceCenter Institutional Animal Care and Use Committee ProjectGRD-11-077.

REFERENCESAguirre, A. A., and P. L. Lutz. 2004. Marine turtles as sentinels of ecosystem

health: is fibropapillomatosis an indicator? EcoHealth 1:275–283.Bohnsack, J. A. 1993. Marine reserves—they enhance fisheries, reduce conflicts,

and protect resources. Oceanus 36(3):63–71.Bohnsack, J. A. 2011. Impacts of Florida coastal protected areas on recreational

world records for Spotted Seatrout, Red Drum and Common Snook. Bulletinof Marine Science 87:939–970.

Bossart, G. D. 2011. Marine mammals as sentinel species for oceans and humanhealth. Veterinary Pathology 48:676–690.

Evans, M. R., S. J. Larsen, G. H. M. Riekerk, and K. G. Burnett. 1997. Patternsof immune response to environmental bacteria in natural populations of the

Red Drum, Sciaenops ocellatus (Linnaeus). Journal of Experimental MarineBiology and Ecology 208:87–105.

Fulton, T. W. 1904. The rate of growth of fishes. Pages 141–241 in Twenty-second annual report of the Fishery Board for Scotland, being for theyear 1903. Part III—scientific investigations. James Hedderwick and Sons,Glasgow, UK.

Johnson, D. R., N. A. Funicelli, and J. A. Bohnsack. 1999. Effectivenessof existing estuarine no-take fish sanctuary within the Kennedy SpaceCenter, Florida. North American Journal of Fish Management 19:436–453.

Kucherka, W. D., P. Thomas, and I. A. Khan. 2006. Sex differences in circulatingsteroid hormone levels in the Red Drum, Sciaenops ocellatus L. AquacultureResearch 37:1464–1472.

Landsberg, J. H., B. A. Blakesley, R. O. Reese, G. McRae, and P. R. Forstchen.1998. Parasites of fish as indicators of environmental stress. EnvironmentalMonitoring and Assessment 51:211–232.

Leamon, J. H., E. T. Schultz, and J. F. Crivello. 2000. Variation among fourhealth indices in natural populations of the estuarine fish, Fundulus heterocli-tus (Pisces, Cyprinodontidae), from five geographically proximate estuaries.Environmental Biology of Fishes 57:451–458.

Murray, S. N., R. F. Ambrose, J. A. Bohnsack, L. W. Botsford, M. H. Carr,G. E. Davis, P. K. Dayton, D. Gotshall, D. R. Gunderson, M. A. Hixon,J. Lubchenco, M. Mangel, A. MacCall, D. A. McArdle, J. C. Ogden, J.Roughgarden, R. M. Starr, M. J. Tegner, and M. M. Yoklavich. 1999. No-take reserve networks: sustaining fishery populations and marine ecosystems.Fisheries 24(11):11–25.

Reyier, E. A., R. H. Lowers, D. M. Scheidt, and D. H. Adams. 2011. Movementpatterns of adult Red Drum, Sciaenops ocellatus, in shallow Florida lagoonsas inferred through autonomous acoustic telemetry. Environmental Biologyof Fishes 90:343–360.

Schlacher, T. A., J. A. Mondon, and R. M. Connolly. 2007. Estuarine fish healthassessment: evidence of wastewater impacts based on nitrogen isotopes andhistopathology. Marine Pollution Bulletin 54:1762–1776.

Vecchio, J. L., and C. A. Wenner. 2007. Catch-and-release mortality in subadultand adult Red Drum captured with popular fishing hook types. North Amer-ican Journal of Fisheries Management 27:891–899.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4



Comparative Susceptibility of Channel Catfish, BlueCatfish, and their Hybrid Cross to ExperimentalChallenge with Bolbophorus damnificus (Digenea:Bolbophoridae) CercariaeMatt J. Griffina, Stephen R. Reichleya, Lester H. Khooa, Cynthia Warea, Terrence E.Greenwayb, Charles C. Mischkeb & David J. Wiseb

a Thad Cochran National Warmwater Aquaculture Center, Aquatic Research and DiagnosticLaboratory, College of Veterinary Medicine, Mississippi State University, 127 ExperimentStation Road, Stoneville, Mississippi 38776, USAb Thad Cochran National Warmwater Aquaculture Center, Mississippi Agricultural andForestry Experiment Station, Mississippi State University, 127 Experiment Station Road,Stoneville, Mississippi 38776, USAPublished online: 04 Jun 2014.

To cite this article: Matt J. Griffin, Stephen R. Reichley, Lester H. Khoo, Cynthia Ware, Terrence E. Greenway, CharlesC. Mischke & David J. Wise (2014) Comparative Susceptibility of Channel Catfish, Blue Catfish, and their Hybrid Cross toExperimental Challenge with Bolbophorus damnificus (Digenea: Bolbophoridae) Cercariae, Journal of Aquatic Animal Health,26:2, 96-99, DOI: 10.1080/08997659.2014.886636







http://dx.doi.org/10.1080/08997659.2014.886636




COMMUNICATION

Comparative Susceptibility of Channel Catfish,Blue Catfish, and their Hybrid Cross to ExperimentalChallenge with Bolbophorus damnificus(Digenea: Bolbophoridae) Cercariae

Matt J. Griffin,* Stephen R. Reichley, Lester H. Khoo, and Cynthia WareThad Cochran National Warmwater Aquaculture Center, Aquatic Research and Diagnostic Laboratory,College of Veterinary Medicine, Mississippi State University, 127 Experiment Station Road, Stoneville,Mississippi 38776, USA

Terrence E. Greenway, Charles C. Mischke, and David J. WiseThad Cochran National Warmwater Aquaculture Center, Mississippi Agricultural and ForestryExperiment Station, Mississippi State University, 127 Experiment Station Road, Stoneville,Mississippi 38776, USA

AbstractThe digenetic trematode Bolbophorus damnificus has been im-

plicated in significant losses in catfish aquaculture since the late1990s. The complex life cycle sequentially involves the Americanwhite pelican Pelecanus erythrorhynchos, the marsh rams hornsnail Planorbella trivolvis, and Channel Catfish Ictalurus punctatus.Research supports anecdotal reports from the industry, suggestingthat the hybrid of Channel Catfish × Blue Catfish I. furcatus is lesssusceptible to disease agents that have been historically prohibitiveto Channel Catfish production, namely the gram-negative bacte-ria Edwardsiella ictaluri and Flavobacterium columnare, as well asthe myxozoan parasite Henneguya ictaluri. This current researchcompared the susceptibility of Channel Catfish, Blue Catfish, andtheir hybrid cross to an experimental challenge by B. damnificus.Fish were exposed to 0, 100, 200, and 400 B. damnificus cercariaeper fish, and the numbers of metacercariae per fish were deter-mined 14 d postchallenge. Metacercariae were recovered from allchallenged fish. There were no significant differences among fishgroups challenged with the same dose, suggesting Channel andBlue Catfish and their hybrid are comparably susceptible to B.damnificus infection. As such, it is recommended that producersraising hybrid catfish remain diligent in controlling populations ofthe snail intermediate host to prevent production losses attributedto B. damnificus, especially when loafing pelicans have been ob-served at the aquaculture operation.

*Corresponding author: [email protected] October 22, 2013; accepted January 5, 2014

The digenetic trematode Bolbophorus damnificus (Digenea:Bolbophoridae) is one of the most significant parasites affect-ing commercial catfish production in the southeastern UnitedStates. The parasite has been attributed to significant reductionsin overall production in Louisiana, Arkansas, and Mississippi(Terhune et al. 2002). Even subclinical infections with limitedmortality can result in significant reductions in overall produc-tion, with moderate to heavy infections resulting in negative netreturns (Wise et al. 2008).

The complex life cycle sequentially involves the Americanwhite pelican Pelecanus erythrorhynchos, the marsh rams hornsnail Planorbella trivolvis, and Channel Catfish Ictalurus punc-tatus (Levy et al. 2002; Overstreet et al. 2002; Yost 2008).Metacercariae remain viable in the Channel Catfish up to 36months postinfection (Mitchell et al. 2011). Although metacer-cariae can persist in the fish for extended periods, the parasiteis most limiting during the acute stages of infection. Once thesource of infectious cercariae is removed, disease resistance andproduction characteristics of infected fish are comparable withnoninfected fish (Labrie et al. 2004; Wise et al. 2013). As such,current management strategies primarily focus on breaking thelife cycle and eliminating the source of infection through erad-ication or limiting the presence of the intermediate snail host

96

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

COMMUNICATION 97

(Mitchell and Hobbs 2003; Mischke et al. 2005; Wise et al.2006; Mitchell et al. 2007). These management strategies haveproven effective, provided operations are diligent in implement-ing a snail control program. However, the parasite still poses asignificant problem for producers.

Due to their favorable production characteristics, there hasbeen a trend towards increased stocking of hybrids of Chan-nel Catfish × Blue Catfish I. furcatus as a production animal(Bosworth et al. 2004; Green and Rawles 2010). Because hybridcatfish production has increased, it is essential that pathogenshistorically associated with Channel Catfish production are eval-uated in hybrid catfish to identify disease risks and develop ef-fective disease management programs. Several pathogens highlydetrimental to Channel Catfish production, namely Edward-siella ictaluri, Flavobacterium columnare, and Henneguya ic-taluri, do not induce the same degree of pathology in hybrid andBlue Catfish (Wolters et al. 1996; Bosworth et al. 2003; Beechamet al. 2010; Griffin et al. 2010a; Arias et al. 2012). Although in-fection in the hybrid is not as common as it is in Channel Catfish,a review of disease case submissions to the Aquatic Researchand Diagnostic Laboratory in Stoneville, Mississippi, shows anincreased incidence of B. damnificus infections in hybrid catfishin recent years (http://tcnwac.msstate.edu/publications.htm).

The research described here focused on the comparative sus-ceptibility of Channel Catfish, Blue Catfish, and their hybridcross to experimental challenge with B. damnificus cercariae toidentify any differences among these fish groups.

METHODSMarsh rams horn snails (n = 800) were collected from a

commercial catfish pond infested with B. damnificus cercariae.Snails were individually placed in glass scintillation vials con-taining 10 mL of reservoir pond water passed through a 20-µmmesh and incubated overnight at 27◦C. After 24 h, the waterin each vial was examined for the presence of cercariae. Snailsshedding only Bolbophorus sp. cercariae were initially identi-fied morphologically (Flowers et al. 2005), and identities wereconfirmed using molecular protocols outlined in Griffin et al.(2010b).

Water was then removed from snail vials containing onlyB. damnificus cercariae and replaced with filtered pond wa-ter. Snails were reincubated at 27◦C. After an 18-h incubationperiod, water from each vial was pooled into a 250-mL Erlen-meyer flask. A magnetic stir bar was added to the flask, andpooled cercariae were gently stirred. While stirring, individual100-µL aliquots were collected and placed as separate dropletson a plastic petri dish. Cercariae were visualized using a NikonSMZ-U stereoscopic microscope (Nikon Instruments, Melville,New York) and the number of cercariae present in each dropletwas counted. These counts were used to approximate the con-centration of cercariae (number of cercariae/mL) in the flask.Based on this approximate concentration, aliquots representingeach individual challenge dose (0, 100, 200, and 400 cercariae

per fish) were placed in respective 250-mL Erlenmeyer flasksand immediately used to infect fish.

Fingerlings (mean, 18.2 g; range, 10–25 g) of Channel Cat-fish, Blue Catfish, and their hybrid were obtained from the U.S.Department of Agriculture’s Warmwater Aquaculture ResearchUnit in Stoneville, Mississippi. Before the challenge, fish wereheld in 2,000-L fiberglass holding tanks containing 1,000 L ofwell water that had a flow rate of a 1 L/min and was constantlyaerated. For the challenge, individual fish were placed in 2-Lcontainers containing 1 L of filtered, constantly aerated pondwater. Three catfish from each species and the hybrid wererespectively exposed to 0, 100, 200, and 400 B. damnificus cer-cariae per fish. After 1 h, fish from each treatment group werepooled in respective 80-L aquaria containing 20 L of well waterand held under flow-through conditions (1 L/min) with constantaeration. Fish were fed to satiation on every third day and after 14d all fish were euthanized by an overdose of tricaine methane-sulfonate (MS-222; Tricaine-S, Western Chemical, Ferndale,Washington). Fish were examined for the presence of trematodemetacercariae using sequential incisions, 1–2 mm apart, throughthe musculature, beginning rostrally and proceeding caudally onboth sides of the sagittal plane. All excised metacercariae werepooled, and five metacercariae from each treatment group weresubjected to molecular analysis to confirm identity. Briefly, ge-nomic DNA (gDNA) from individual metacercariae was isolatedusing the PureGene DNA isolation kit, following the manufac-turer’s suggested protocol for frozen animal tissues (Qiagen,Valencia, California). Isolated gDNA was then used for PCRidentification according to procedures outlined by Griffin et al.(2010b).

Differences in total metacercariae recovered, based on doseand fish group, were identified using a two-way ANOVA. Anal-ysis was carried out using the PROC ANOVA and Duncan’smultiple range test in SAS version 9.3 (SAS Institute, Cary,North Carolina) at a significance level of 0.05.

RESULTS AND DISCUSSIONMetacercariae developed in all fish exposed to cercariae.

There were no significant differences in metacercariae excisedfrom Channel Catfish, Blue Catfish, and hybrid catfish that re-ceived the same challenge dose, and no interactive effects wereobserved. For all groups, the number of metacercariae presentincreased with the challenge dose; no metacercaria were foundin control fish (Table 1) All metacercariae analyzed by PCRwere identified as B. damnificus.

Although small in scale, the data presented here suggestB. damnificus has the potential to be a significant deterrent tohybrid catfish production, as transmission rates and metacer-cariae development are similar to those seen in Channel Catfish.This is supported by case studies and diseased fish submissionsto the Aquatic Research and Diagnostic Laboratory (ARDL) atMississippi State University College of Veterinary Medicine inStoneville, Mississippi. In 2013, there were 80 B. damnificus

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

98 GRIFFIN ET AL.

TABLE 1. Mean (SD) metacercariae excised from catfish (n = 3) challenged with 0, 100, 200, and 400 Bolbophorus damnificus cercariae per fish, 14 dpostchallenge. No interactive effects were observed between challenge dose and fish group; there were no significant differences in mean metacercariae excisedfrom Channel and Blue Catfish and their hybrid receiving the same challenge dose. For all groups, the number of metacercariae present significantly increasedwith challenge dose. Within columns and within rows, pooled means followed by different letters are significantly different (P < 0.05).

Number of cercariae per fish Blue Catfish Channel Catfish Hybrid catfish Pooled means

0 0 0 0 0100 4.0 (1.0) 4.3 (0.6) 4.0 (0.0) 4.1 (0.6) x200 9.0 (2.0) 9.3 (4.0) 12.6 (3.5) 10.3 (3.4) y400 15.0 (1.0) 15.7 (2.1) 14.3 (0.6) 15.0 ( ± 1.3) zPooled means 9.3 (4.9) 9.8 (5.4) 10.3 (5.1)

Results of two-way ANOVA: df P > F Pooled SE

Fish group 2 0.607 0.7Dose 2 <0.001 0.7Interaction 4 0.3050

cases diagnosed at the ARDL. Of these, 11.3% were in hybridcatfish, nearly half of which (44.4%) had B. damnificus as thesole infectious agent (L. H. Khoo, unpublished). The remainderhad concomitant bacterial infections (Edwardsiella ictaluri andFlavobacterium columnare), which may have been secondary,a function of increased susceptibility to bacterial infections fol-lowing B. damnificus exposure (Labrie et al. 2004). This repre-sents a marked increase in B. damnificus cases involving hybridssince 2009, when the ARDL began reporting hybrid catfish as aseparate category (http://tcnwac.msstate.edu/publications.htm).

Initial work has demonstrated an association betweenB. damnificus and fish kills on Channel Catfish fingerlingoperations in Louisiana and Mississippi (Venable et al. 2000;Levy et al. 2002). This was supported by Yost (2008), whocharacterized the acute pathology and mortality in ChannelCatfish fingerlings exposed to B. damnificus cercariae. Recenteconomic analysis of farm-level impacts of B. damnificus onChannel Catfish production showed that trematode infectionsreduce fish production and economic returns due to parasite-induced inappetence and increases in mortality (Hawke andKhoo 2004; Wise et al. 2008). Even outbreaks with lowprevalence and low parasite intensity, which often go unnoticedby producers, can seriously affect farm profitability. Wise et al.(2008) demonstrated low levels of B. damnificus infestationin a culture operation significantly reduced overall production,and ponds categorized as having a light infection rate exhibiteda 61% reduction in net returns compared with uninfectedpopulations. Moreover, their work showed ponds carryingmoderate to heavy infections were operating at a net loss. Inaddition to these negative impacts on overall catfish production,B. damnificus infection can also increase susceptibility tobacterial infection, further exacerbating potential losses (Labrieet al. 2004). While the long-term effect of B. damnificus onhybrid catfish production has not been adequately evaluated, wespeculate it will be comparable to what is observed in ChannelCatfish.

Hybrid catfish have many desirable production characteris-tics, including improved disease resistance and lower suscepti-bility to several Channel Catfish pathogens (Wolters et al. 1996;Bosworth et al. 2003; Beecham et al. 2010; Griffin et al. 2010a;Arias et al. 2012). However, based on this current work, Channeland Blue Catfish and their hybrid appear comparably suscep-tible to B. damnificus infection. To prevent the life cycle ofB. damnificus from becoming established in fish culture ponds,producers raising hybrid catfish are encouraged to remain dili-gent in reducing the number of snails that act as an intermediatehost. Fortunately, there are several management strategies avail-able to eradicate or dramatically reduce snail populations in cat-fish production ponds (Mitchell and Hobbs 2003; Mischke et al.2005; Wise et al. 2006; Mitchell et al. 2007). Eliminating thesnail host breaks the parasite’s life cycle and removes the sourceof cercariae from the system. In Channel Catfish production fishrecover quickly once the source of infection is removed, and theproduction of fish that have survived an outbreak is similar tothat of cohorts that have never been exposed to the parasite(Wise et al. 2013).

ACKNOWLEDGMENTSThe authors thank Holly Whitehead and Kyle Christopher for

their technical assistance. This work was supported by the U.S.Department of Agriculture Catfish Health Initiative (Project6402-31320-002-02), the Mississippi State University Collegeof Veterinary Medicine, and the Mississippi Agriculture andForestry Experiment Station. This is Mississippi Agriculturaland Forestry Experiment Station publication J-12513.

REFERENCESArias, C. R., W. Cai, E. Peatman, and S. A. Bullard. 2012. Catfish hybrid Ictalu-

rus punctatus × I. furcatus exhibits higher resistance to columnaris diseasethan the parental species. Diseases of Aquatic Organisms 100:77–81.

Beecham, R. V., M. J. Griffin, S. B. LaBarre, D. Wise, M. Mauel, L. M. W. Pote,and C. D. Minchew. 2010. The effects of proliferative gill disease on the blood

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4

COMMUNICATION 99

physiology of Channel Catfish, Blue Catfish and Channel Catfish × BlueCatfish hybrid fingerlings. North American Journal of Aquaculture 72:213–218.

Bosworth, B. G., D. J. Wise, J. S. Terhune, and W. R. Wolters. 2003. Family andgenetic group effects for resistance to proliferative gill disease in ChannelCatfish, Blue Catfish and Channel Catfish × Blue Catfish backcross hybrids.Aquaculture Research 34:569–573.

Bosworth, B. G., W. R. Wolters, J. L. Silva, R. S. Chamul, and S. Park.2004. Comparison of production, meat yield, and meat quality traits ofNWAC103 line Channel Catfish, Norris line Channel Catfish, and femaleChannel Catfish × male Blue Catfish F1 hybrids. North American Journal ofAquaculture 66:177–183.

Flowers, J. R., M. F. Poore, L. M. Pote, R. Wayne Lidaker, and M. G.Levy. 2005. Cercariae of Bolbophorus damnificus and Bolbophorus sp. withnotes on North American bolbophorids. Comparative Parasitology 72:220–226.

Green, B. W., and S. D. Rawles. 2010. Comparative growth and yield of Chan-nel Catfish and channel × blue hybrid catfish fed a full or restricted ration.Aquaculture Research 41:109–119.

Griffin, M. J., A. C. Camus, D. J. Wise, T. E. Greenway, M. J. Mauel, and L.M. Pote. 2010a. Variation in susceptibility to Henneguya ictaluri infectionby two species of catfish and their hybrid cross. Journal of Aquatic AnimalHealth 22:21–35.

Griffin, M. J., D. J. Wise, M. C. Yost, C. M. Doffitt, L. M. Pote, T. E. Greenway,and L. H. Khoo. 2010b. A duplex real-time polymerase chain reaction assayfor differentiation between Bolbophorus damnificus and Bolbophorus typeII species cercaria. Journal of Veterinary Diagnostic Investigations 22:615–622.

Hawke, J. P., and L. H. Khoo. 2004. Infectious disease. Pages 387–443 inC. S. Tucker and J. A. Hargreaves, editors. Biology and culture of the ChannelCatfish. Elsevier, Amsterdam.

Labrie, L., C. Komar, J. Terhune, A. Camus, and D. Wise. 2004. Effect ofsublethal exposure to the trematode Bolbophorus spp. on the severity ofenteric septicemia of catfish in Channel Catfish fingerlings. Journal of AquaticAnimal Health 16:231–237.

Levy, M. G., J. R. Flowers, M. F. Poore, J. E. Mullen, L. H. Khoo, L. M. Pote,I. Paperna, R. Dzikowski, and R. W. Litaker. 2002. Morphologic, pathologic,and genetic investigations of Bolbophorus species affecting cultured ChannelCatfish in the Mississippi Delta. Journal of Aquatic Animal Health 14:235–246.

Mischke, C. C., D. J. Wise, and L. M. Pote. 2005. Acute toxicity of chemicalsto the marsh rams-horn snail Planorbella trivolvis. Journal of the WorldAquaculture Society 36:560–563.

Mitchell, A. J., and M. S. Hobbs. 2003. Effect of citric acid, copper sulfateconcentration, and temperature on a pond shoreline treatment for control ofthe marsh rams-horn snail Planorbella trivolvis and the potential toxicity ofthe treatment to Channel Catfish. North American Journal of Aquaculture65:306–313.

Mitchell, A. J., S. Snyder, D. J. Wise, and C. C. Mischke. 2007. Evaluating pondshoreline treatments of slurried hydrated lime for reducing marsh rams-hornsnail populations. North American Journal of Aquaculture 69:313–316.

Mitchell, A., M. Yost, L. Pote, B. Farmer, and C. Panuska. 2011. Longevityof Bolbophorus damnificus infectious in Channel Catfish. Journal of AquaticAnimal Health 23:103–109.

Overstreet, R. M., S. S. Curran, L. M. Pote, D. T. King, C. K. Blend, and W.D. Grater. 2002. Bolbophorus damnificus n. sp. (Digenea: Bolbophoridae)from the Channel Catfish Ictalurus punctatus and American white pelicanPelecanus erythrorhynchos in the USA based on life-cycle and moleculardata. Systematic Parasitology 52:81–96.

Terhune, J. S., D. J. Wise, and L. H. Khoo. 2002. Bolbophorus confusus in-fections in Channel Catfish in northwestern Mississippi and effects of watertemperature on emergence of cercariae from infected snails. North AmericanJournal of Aquaculture 64:70–74.

Venable, D. L., A. P. Gaude III, and P. L. Kerks. 2000. Control of the trematodeBolbophorus confusus in Channel Catfish Ictalurus punctatus ponds usingsalinity manipulation and polyculture with Black Carp Mylopharyngodonpiceus. Journal of the World Aquaculture Society 31:158–166.

Wise, D. J., T. R. Hanson, and C. S. Tucker. 2008. Farm-level economic impactsof Bolbophorus infections of Channel Catfish. North American Journal ofAquaculture 70:382–387.

Wise, D. J., M. H. Li, M. J. Griffin, E. H. Robinson, L. H. Khoo, T. E. Greenway,T. S. Byars, J. R. Walker, and C. C. Mischke. 2013. Impacts of Bolbophorusdamnificus (Digenea: Bolbophoridae) on production characteristics of Chan-nel Catfish, Ictalurus punctatus, raised in experimental ponds. Journal of theWorld Aquaculture Society 44:557–564.

Wise, D. J., C. C. Mischke, T. E. Greenway, T. S. Byars, and A. J. Mitchell.2006. Uniform application of copper sulfate as a potential treatment forcontrolling snail populations in Channel Catfish production ponds. NorthAmerican Journal of Aquaculture 68:364–368.

Wolters, W. R., D. J. Wise, and P. H. Klesius. 1996. Survival and antibody re-sponse of Channel Catfish, Blue Catfish, and Channel Catfish female × BlueCatfish male hybrids after exposure to Edwardsiella ictaluri. Journal ofAquatic Animal Health 8:249–254.

Yost, M. C. 2008. The study of the life cycle of Bolbophorus damnificus and itspathology in the Channel Catfish (Ictalurus punctatus). Doctoral dissertation.Mississippi State University, Mississippi State.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:53

15

June

201

4



Association of Mitochondrial Dysfunction withOxidative Stress and Immune Suppression in BluntSnout Bream Megalobrama amblycephala Fed a High-Fat DietKang-Le Lua, Wei-Na Xua, Wen-Bin Liua, Li-Na Wanga, Chun-Nuan Zhanga & Xiang-Fei Liaa Key Laboratory of Aquatic Animal Nutrition and Feed Science of Jiangsu Province, Collegeof Animal Science and Technology, Nanjing Agricultural University, Number 1 Weigang Road,Nanjing 210095, ChinaPublished online: 04 Jun 2014.

To cite this article: Kang-Le Lu, Wei-Na Xu, Wen-Bin Liu, Li-Na Wang, Chun-Nuan Zhang & Xiang-Fei Li (2014) Association ofMitochondrial Dysfunction with Oxidative Stress and Immune Suppression in Blunt Snout Bream Megalobrama amblycephalaFed a High-Fat Diet, Journal of Aquatic Animal Health, 26:2, 100-112, DOI: 10.1080/08997659.2014.893460







http://dx.doi.org/10.1080/08997659.2014.893460




ARTICLE

Association of Mitochondrial Dysfunction with OxidativeStress and Immune Suppression in Blunt Snout BreamMegalobrama amblycephala Fed a High-Fat Diet

Kang-Le Lu, Wei-Na Xu, Wen-Bin Liu,* Li-Na Wang, Chun-Nuan Zhang,and Xiang-Fei LiKey Laboratory of Aquatic Animal Nutrition and Feed Science of Jiangsu Province, College of AnimalScience and Technology, Nanjing Agricultural University, Number 1 Weigang Road, Nanjing 210095,China

AbstractHigh-fat diets may have favorable effects on growth, partly based on protein sparing, but high-fat diets often

lead to fatty liver (excessive fat deposition in the liver), which may be deleterious to fish growth and health. Thegoal of this study was therefore to investigate possible adverse effects and how they develop. Juvenile Blunt SnoutBream Megalobrama amblycephala (initial weight ± SE = 17.70 ± 0.10 g) were fed two diets (5% fat [control] or15% fat). After 8 weeks, fish that were fed the 15% fat diet showed a high rate of mortality and poor growth. Thehistological results clearly showed that the high fat intake resulted in fat and glycogen accumulation and structuralalterations of the hepatocytes, mitochondria, and nuclei. In the high-fat group, impairments of the mitochondriaincluded mitochondrial swelling and the loss of cristae and matrix. Fish that were given the 15% fat diet exhibited lowsuccinate dehydrogenase and Na+ ,K+ -ATPase activities and increased cytochrome-c release from the mitochondria.Expression of genes for complex I and III subunits of the mitochondrial respiratory chain were down-regulatedin fish that received the high-fat diet. Increases in malondialdehyde level and the ratio of oxidized glutathioneto reduced glutathione suggested oxidative stress in the livers of fish from the high-fat diet group. Moreover, thelower leukocyte count, lysozyme and alternative complement activities, and globulin level in fish that received thehigh-fat diet indicated suppressive immune responses. Overall, the intake of excessive fat impaired mitochondrialbioenergetics and physiological functions. The dysfunction of the mitochondria subsequently mediated oxidativestress and hepatocyte apoptosis, which in turn led to the reduced efficacy of the immune system.

High-fat diets may have favorable effects on growth for somefish because of the protein-sparing effect, with dietary lipid be-ing used to provide energy in replacement of protein (Watanabe2002; Du et al. 2006; Li et al. 2012a). Dietary lipids play an im-portant role in fish nutrition as a source of energy and essentialfatty acids for maintaining the biological structure and normalfunction of cell membranes (Sargent and Tacon 1999). Indeed,increasing dietary lipid levels within certain limits will supportthe higher growth rates of fish. However, administering dietarylipid near an upper limit often leads to unwanted fat depositionin the liver (or other tissues), inducing a condition referred to

*Corresponding author: [email protected] July 15, 2013; accepted January 30, 2014

as fatty liver (Du et al. 2006, 2008; Lu et al. 2013). Accordingto previous studies, fatty liver is often closely correlated witha high rate of mortality or poor growth (Roberts 1989). There-fore, it is necessary to acquire additional knowledge about fattyliver, as such knowledge benefits the development of efficientand sustainable aquaculture.

Previous work has shown that histological examination of theliver provides an index of general fish condition (Coz-Rakovacet al. 2005). The understanding of organelle ultrastructure couldprovide more information on metabolic condition. Mitochondriaare organelles found in all eukaryotic cells, and their major

100

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

MITOCHONDRIAL DYSFUNCTION IN BLUNT SNOUT BREAM 101

function is to generate cellular energy in the form of adenosinetriphosphate (ATP). In the presence of oxygen, the mitochondriaoxidize nutrients (glucose, fatty acids, and amino acids) to ATPvia oxidative phosphorylation. Therefore, feed efficiency andmitochondrial function are fundamentally linked through theconcept of energy allocation (Eya et al. 2012). Numerous studieshave already provided evidence of a relationship between feedefficiency or growth and mitochondrial function in poultry, rats,and cattle (Bottje et al. 2002; Iqbal et al. 2004; Sandelin et al.2005; Bottjee and Carstens 2009). However, to our knowledge,there has been no prior study on the condition of mitochondriain fish with fatty livers.

The Blunt Snout Bream Megalobrama amblycephala is anherbivorous freshwater fish that is native to China (Zhou et al.2008). Due to its fast growth, tender flesh, and high disease resis-tance, this species has been widely favored for use in aquaculturein China. However, similar to a number of other commerciallyproduced fishes, the artificial rearing of Blunt Snout Bream isoften associated with the occurrence of fatty liver, which cor-relates closely with a high rate of mortality or poor growth.Therefore, studies focusing on the mechanisms of fatty liver areimperative not only to clarify the scientific nature of this condi-tion but also for the benefit of commercial culture. In mammals,the mitochondrial abnormalities associated with fatty liver in-clude ultrastructural lesions and decreased activity of respiratorychain complexes (Wei et al. 2008). Mitochondria are consideredto be the most important subcellular organelles for mediatingoxidative stress (Mignotte and Vayssiere 1998). On the otherhand, several more recent studies support the concept that ox-idative stress is a significant factor underlying a dysfunctionalimmune response (Sordillo and Aitken 2009). Fatty liver condi-tion has been observed in fish that were fed high-carbohydrate orlipid-rich formulated feeds (Szlaminska et al. 1991; Catacutanet al. 1997; Nanton et al. 2001). The widespread use of lipid-rich diets due to the protein-sparing potential of lipid is the mainreason for the occurrence of fatty liver in farmed fishes (Nantonet al. 2001; Li et al. 2012a). Based on our previous studies, thelivers of Blunt Snout Bream that were fed a 5% fat diet showeda normal structure (Lu et al. 2013); Blunt Snout Bream thatwere given a 7% fat diet had the best weight gain and proteinutilization (Li et al. 2010). Moreover, fat accumulation in theliver occurs when the dietary lipid level reaches 15% (Lu et al.2013). The principal goal of this study was therefore to comparethe effects of two diets (5% fat [control] and 15% fat) on mito-chondrial condition, oxidative stress, and immune response inBlunt Snout Bream. The result may have implications for ourunderstanding of fatty liver and the possible adverse effects ofthis condition in fish.

METHODSExperimental fish and feeding trial.—Juvenile Blunt Snout

Bream were obtained from the fish hatchery of Wuhan (Hubei,China). The experiment was performed in a recirculating aqua-

TABLE 1. Formulation and proximate composition of the experimental dietsthat were administered to Blunt Snout Bream (control = 5% fat; high fat = 15%fat).

DietProximate component (%)or ingredient (%) Control High fat

Proximate compositionMoisture 9.80 9.70Crude protein 31.1 31.3Crude lipid 4.90 14.7Crude fiber 13.8 4.10Ash 7.80 8.00Carbohydratea 32.6 32.3Energyb 14.9 18.8

IngredientsFish meal 15.0 15.0Casein 15.0 15.0Soybean meal 20.0 20.0Corn starch 25.0 25.0α-starch 5.00 5.00Fish oil 1.90 6.90Soybean oil 1.90 6.90Cellulose 10.4 0.40Calcium biphosphate 1.80 1.80Premixc 1.00 1.00Carboxymethyl cellulose 3.00 3.00

aCarbohydrate (nitrogen-free extract) was calculated by difference (carbohydrate =100 − moisture − crude protein − crude lipid – ash − crude fiber).

bEnergy (kJ/g of diet) = (% crude protein × 23.6) + (% crude lipid × 39.5) + (%carbohydrate × 17.3).

cPremix supplied the following minerals (g/kg) and vitamins (IU or mg/kg):CuSO4·5H2O, 2.0 g; FeSO4·7H2O, 25 g; ZnSO4·7H2O, 22 g; MnSO4·4H2O, 7 g; Na2SeO3,0.04 g; KI, 0.026 g; CoCl2·6H2O, 0.1 g; vitamin A, 900,000 IU; vitamin D, 200,000 IU;vitamin E, 4,500 mg; vitamin K3, 220 mg; vitamin B1, 320 mg; vitamin B2, 1,090 mg;niacin, 2,800 mg; vitamin B5, 2,000 mg; vitamin B6, 500 mg; vitamin B12, 1.6 mg; vitaminC, 5,000 mg; pantothenate, 1,000 mg; folic acid, 165 mg; and choline, 60,000 mg.

culture system in the laboratory. After a 1-week acclimationperiod, 192 juveniles of similar size (average weight ± SE =17.7 ± 0.10 g) were randomly distributed into twelve 100-Ltanks (16 fish/tank). Water temperature, dissolved oxygen con-centration, and pH were monitored daily. During the feedingperiod, fish were reared under the following conditions: wa-ter temperature of 25–27◦C; dissolved oxygen concentration of5.0–6.0 mg/L; pH of 7.2–7.6; and a 12-h light : 12-h dark pho-toperiod. Fish were hand fed to apparent satiation three timesdaily (0800, 1200, and 1600 hours) using two experimental diets(5% fat and 15% fat). Formulation and proximate compositionof the diets are presented in Table 1. The amount of feed of-fered was quantified daily for each tank. Each diet treatmentwas tested in six replicates, and the trial lasted 8 weeks.

Sample collection.—At the end of the feeding trial, fish werestarved overnight prior to sampling. Total number and weight offish in each tank were determined. Eight fish were randomly se-lected from each tank and were immediately euthanized with an

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

102 LU ET AL.

overdose of tricaine methanesulfonate (MS-222; Sigma, USA)at 100 mg/L of water. A blood sample from the caudal vessel wasrapidly collected into heparinized Eppendorf tubes. One part ofthe blood sample was used for determination of total erythrocytecount and total leukocyte count. The remainder was centrifuged(850 × g for 10 min at 4◦C), and the plasma was stored at−70◦C until analysis. After blood collection, the liver was re-moved, placed on ice, and then stored at −70◦C until analysis.Additionally, the liver samples for the histology observationswere fixed in the relevant buffer.

Histology study.—Samples used in light microscopy obser-vation were fixed in 10% buffered formalin, dehydrated in agraded ethanol series, embedded in paraffin, and sectioned to5-µm thickness with a rotary microtome. The sections werethen stained with hematoxylin and eosin for observation undera Nikon 50i microscope (Nikon, Tokyo, Japan).

Samples for electron microscopy observation were fixed in2.5% glutaraldehyde for 24 h, postfixed in 1% osmium tetroxide(OsO4) for 1 h, and stored at 4◦C. Sections were embedded inepoxy resin (Epon812), cut to 70-nm thickness with an RMCPowerTome XL microtome, stained with uranyl acetate andlead citrate, and examined under a Hitachi H-7650 transmissionelectron microscope (Hitachi, Tokyo, Japan).

For hepatocyte apoptosis determination, the 5-µm-thick sec-tions were further treated using the terminal deoxynucleotidyltransferase (TdT)-mediated deoxyuridine triphosphate (dUTP)–biotin nick end-labeling (TUNEL) assay with the Apoptosis De-tection Kit (Nanjing Jiancheng Bioengineering Institute [NJJC],China) via the manufacturer’s protocol. Briefly, the sectionswere deparaffinized, hydrated, and incubated with freshly pre-pared 3% hydrogen peroxide (H2O2) in phosphate-bufferedsaline for 10 min to block the endogenous peroxidase activ-ity. Each slide was permeated with a 10-µg/mL solution ofproteinase K at 37◦C for 10 min. The tissues were immersed in20 µL of TdT, which was marked with biotin on dUTP and in-cubated at 37◦C for 2 h. The slide was covered with horseradishperoxidase-labeled streptavidin at 37◦C for 30 min and thenwas stained with 3,3′-diaminobenzidine for 15 min. Finally, thenucleus was counterstained with hematoxylin for 30 s. Brown-yellow granules were found in the nucleus of TUNEL-positivecells. Tissue treated with deoxyribonuclease I (DNase1) wasused as the positive control; the reaction without TdT enzymewas used as the negative control. For the quantitative measure-ment of the number of hepatocytes, 100 cells were counted indifferent areas (five fields in each histological slide) at 200 ×magnification, and the percentage of cells with DNA damagewas calculated.

Isolation of mitochondrial and cytoplasmic fractions.—Theisolation of liver mitochondria was performed in a medium con-taining 10-mM KH2PO4, 250-mM sucrose, and 5-mM EDTAadjusted to pH 7.4. Liver tissue (0.1–0.2 g) was excised, washedin cold phosphate-buffered saline, and then homogenized in10 volumes of cold medium. The homogenate was centrifugedtwice at 800 × g for 10 min at 4◦C. The superficial lipid layer

was removed, and the remaining supernatant was centrifuged at15,000 × g for 15 min at 4◦C. The supernatant of the 15,000× g spin was considered to be the cytoplasmic fraction, and theresidual pellet was the mitochondrial fraction. The pellet ob-tained from the second spin was washed three times in mediumand was resuspended in a small volume of medium plus 1-mg/mL fatty-acid-free bovine serum albumin.

Detection of mitochondrial cytochrome-c release.—An in-crease in cytochrome c in the cytosol with a concomi-tant decrease in mitochondrial cytochrome c is indicative ofcytochrome-c release from the mitochondria (Sun et al. 2012).Cytochrome-c contents in mitochondrial and cytosol fractionswere prepared by treatment with sodium dithionite and thenwere measured with a spectropolarimeter according to themethod of Sun et al. (2012). The optical density of the samplewas obtained at 520 nm and was compared to a standard curve.The concentration of cytochrome c was calculated according tothe standard curve of the samples.

Measurements of enzyme activities.—For enzymatic anal-ysis, the mitochondrial pellets were suspended in the isola-tion medium. Succinate dehydrogenase (SDH; enzyme number1.3.5.1 [IUBMB 1992]) was estimated with sodium succinateas substrate according to the method of Philip et al. (1995). TheNa+ ,K+ -ATPase (3.6.3.9) activity was measured according toMcCormick et al. (1995). Mitochondrial protein concentrationwas determined using the method of Lowry et al. (1951). En-zymatic activities were expressed in units (U) per milligram ofmitochondrial protein.

Total RNA extraction, reverse transcription, and real-timePCR.—Total RNA was extracted from the liver tissue by us-ing RNAiso Plus (Takara Co. Ltd, Japan). To avoid genomicDNA amplification, RNA samples were treated by RQ1 RNase-free DNase prior to reverse transcription (RT)-PCR (Takara).Complementary DNA (cDNA) was generated from 500 ng ofDNase-treated RNA using the ExScript RT-PCR Kit (Takara);the mixture consisted of 500 ng of RNA, 2 µL of buffer(5 × ), 0.5 µL of deoxynucleotide triphosphate mixture (10 mMeach), 0.25 µL of RNase inhibitor (40 U/µL), 0.5 µL of dT-APprimer (50 mM), 0.25 µL of ExScript RTase (200 U/µL), anddiethylpyrocarbonate-treated H2O, with a total volume of up to10 µL. The reaction conditions were 42◦C for 40 min, 90◦C for2 min, and 4◦C thereafter.

Real-time PCR was employed to determine messenger RNA(mRNA) levels based on the SYBR Green I Fluorescence Kit.The primers (Table 2) for mitochondrial genes that encode sub-units of the respiratory chain (complex I: NADH dehydroge-nase subunit 1 [ND1]; complex III: cytochrome b [CYTB]; com-plex IV: cytochrome-c oxidase subunit 1 [COX1] and subunit2 [COX2]; complex V: ATP synthase subunit 6 [ATP6]; heatshock protein 70 [HSP70], liver-expressed antimicrobial peptide2 [LEAP2], and β-actin [ACTB]) were designed using Primer 5software based on the cDNA sequences available in GenBank.Real-time PCR was performed in a Mini Option Real-TimeDetector (Bio-Rad, USA). The fluorescent quantitative PCR

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4


TABLE 2. Nucleotide sequences of the primers used for real-time PCR assay of gene expression in samples of Blunt Snout Bream liver tissue (ND1 = NADHdehydrogenase subunit 1; CYTB = cytochrome b; COX1 = cytochrome-c oxidase subunit 1; COX2 = cytochrome-c oxidase subunit 2; ATP6 = ATP synthasesubunit 6; HSP70 = heat shock protein 70; LEAP2 = liver-expressed antimicrobial peptide 2; ACTB = β-actin). Annealing temperature was 59◦C for each reaction.

Target GenBank accession PCR productgene number Forward (5′–3′) Reverse (5′–3′) length (bp)

ND1 NC010341 CTGACCACTAGCCGCAATA GGAAGAAGAGGGCGAAGG 140CYTB NC010341 CATACACTATACCTCCGACAT TCTACTGAGAAGCCACCT 357COX1 NC010341 CATACTTTACATCCGCAACA TCCTGTCAATCCACCCAC 158COX2 NC010341 AACCCAGGACCTTACACCC CCCGCAGATTTCAGAACA 232ATP6 NC010341 TGCTGTGCCACTATGACT ATTATTGCCACTGCGACT 315HSP70 EU884290 CTTGAGATCGACTCGCTGTA AGGTCTGGGTCTGTTTGG 439LEAP2 JQ344324 TATGGAGGATCATGGGTAC CTTTGTGGCTTTAGAGGA 278ACTB AY170122 CGGACAGGTCATCACCATTG CGCAAGACTCCATACCCAAGA 196

solution consisted of 12.5 µL of SYBR premix Ex Taq (2 × ),0.5 µL of PCR forward primer (10 µM), 0.5 µL of PCR reverseprimer (10 µM), 2.0 µL of RT reaction (cDNA solution), and9.5 µL of distilled H2O. The reaction conditions were as fol-lows: 95◦C for 3 min, followed by 45 cycles of 95◦C for 10 s and60◦C for 20 s. The fluorescent flux was then recorded, and thereaction continued at 72◦C for 3 min. The dissolution rate wasmeasured between 65◦C and 90◦C. Each 0.2◦C increase wasmaintained for 1 s, and the fluorescent flux was recorded. Allamplicons were initially separated by agarose gel electrophore-sis to ensure that they were the correct size. A dissociation curvewas determined during the PCR program to verify that specificproducts were obtained in each run. The gene expression levelswere normalized to the mean of the reference gene (ACTB).Normalized gene expression for the control group was set to1.0, and the expression of each target gene for the high-fat dietgroup was expressed relative to that of the control group.

Assays for antioxidant status in liver.—For the determina-tion of antioxidant status, liver samples were prepared as de-scribed by Lygren et al. (1999). Total superoxide dismutase(SOD; 1.15.1.1) activity was measured with a commercial kit(NJJC) according to Nakano (1990). The method of Maral et al.(1977) for catalase (CAT; 1.11.1.6) assay was adapted as de-scribed by Rueda-Jasso et al. (2004). Glutathione peroxidase(GPX; 1.11.1.9) activity was measured using the method ofDabas et al. (2012). Thiobarbituric-acid-reactive substances as-says were performed with a malondialdehyde (MDA) kit (NJJC)as described by Rueda-Jasso et al. (2004). Oxidized glutathione(GSSG) and reduced glutathione (GSH) were determined enzy-matically with a commercial kit (NJJC) based on the recyclingreaction of GSH with 5,5’-dithio-bis-(2-nitrobenzoic acid) in thepresence of an excess of glutathione reductase. Measurementswere made in a microplate reader, and the GSSG/GSH ratio wascalculated as the quotients of reduced and oxidized GSH. Pro-tein concentration in liver homogenates was determined usingthe method of Lowry et al. (1951).

Measures for hematological–immunological parameters.—Total erythrocytes and leukocytes were counted in a hemocy-

tometer by using erythrocyte- and leukocyte-diluting fluids, re-spectively, according to the method of Johnson et al. (2002).Plasma lysozyme activity was determined by using turbidimet-ric assay according to Obach et al. (1993). Alternative comple-ment (ACH50) activity was estimated as described by Monteroet al. (1998); ACH50 activity represented the volume of plasmathat was necessary to produce lysis of 50% of sheep red bloodcells under standard conditions. Plasma total protein concen-tration was estimated by the Biuret and bromocresol green dyebinding method (Reinhold 1953), and the albumin concentra-tion was estimated by the bromocresol green binding method.Globulin was calculated by subtracting the albumin value fromtotal plasma protein (Li et al. 2012b). Cortisol in plasma was es-timated with a validated radioimmunoassay from Winberg andLepage (1998) as modified by Li et al. (2012b).

Statistical analysis.—Data were analyzed with SPSS version10.0. Student’s t-test was used to analyze differences betweenthe two treatments. The level of significance α was set at 0.05.All data are presented as mean ± SE.

RESULTS

Growth PerformanceGrowth, feed utilization, and survival results are presented

in Table 3, showing that final body weight and survival of BluntSnout Bream in the high-fat diet group were significantly lower(P < 0.05) than those of the control group. Feed conversionratio was significantly higher in fish that were given the high-fatdiet than in fish that were offered the control diet (P < 0.05).Moreover, the liver fat level and mesenteric fat index of fish inthe high-fat diet group were also significantly higher than thoseof control fish (P < 0.05).

Histological ExaminationsThe livers of fish that were fed the control diet were ob-

served to have a normal structure, showing clear, complete, andwell-arranged hepatic lobules with hepatocytes having roundnuclei centrally located in the cytoplasm (Figure 1a). However, a

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

104 LU ET AL.

TABLE 3. Growth performance and feed utilization (mean ± SE; n = 6)in Blunt Snout Bream that were fed the experimental diets (control = 5%fat; high fat = 15% fat). Feed conversion ratio (FCR) was calculated as (to-tal diet fed, kg)/(total wet weight gain, kg). Mesenteric fat index was cal-culated as [(mesenteric fat weight)/(fish weight)] × 100. Asterisks indicatesignificant differences between the high-fat diet group and the control group(P < 0.05).

Variable Control diet High-fat diet

Initial body weight (g) 17.6 ± 0.21 17.8 ± 0.12Final body weight (g) 48.4 ± 1.72 40.8 ± 1.29*Survival (%) 93.8 ± 2.5 82.2 ± 3.2*FCR 1.49 ± 0.03 1.99 ± 0.09*Liver fat level (%) 6.93 ± 0.71 11.9 ± 1.69*Mesenteric fat index (%) 2.67 ± 0.21 3.69 ± 0.21*

number of abnormalities were found in the livers of fish that re-ceived the high-fat diet, including diffuse lipid vacuolization, ex-cessive glycogen droplets, irregular arrangement of hepatocytesand endothelial cells of the central vein, inordinate and obscurehepatic cords, and stricture of the hepatic sinus (Figure 1b).

Livers of fish that were fed the control diet demonstrated anormal ultrastructure. Each hepatocyte had a large, round nu-cleus that was centrally located within a moderate cytoplasm(Figure 2a). The nucleus was ovoid and contained a promi-nent nucleolus (Figure 2c). The endoplasmic reticulum waswell developed and often adjacent to the nucleus (Figure 2c).Hepatocytes displayed dark, slender mitochondria with well-developed cristae and matrix (Figure 2e). However, a number ofabnormalities were observed in livers of fish that received thehigh-fat diet. Hepatocytes exhibited many large, electron-densefat droplets, some of which were even larger than the nucleus(Figure 2b). These extensive intracellular lipid droplets resultedin the displacement of the nucleus to the cell margin as well asa loss of cytoplasm (Figure 2b). The nucleus was atrophic andpresented as a polygon (Figure 2d). The endoplasmic reticulumwas poorly developed and lost most of its ribosomes (Figure 2d).All of the cellular membranes (plasma, mitochondria, and nu-clear membranes) were disrupted. The mitochondria exhibited aloss of cristae, matrix, and matrix density with highly hydropicchanges (Figure 2f, g).

The TUNEL-stained hepatocytes showed the characteristicfeatures of DNA damage. Normal hepatocytes were stainedamethyst by hematoxylin, whereas apoptotic cells were stainedbrown (Figure 3a, b). Apoptotic cells made up just 1% of thetotal number of hepatocytes in control fish, whereas apoptoticcells constituted about 7% of the hepatocytes in fish that were fedthe 15% fat diet. Thus, there were significantly more TUNEL-positive cells in fish from the high-fat diet group than in controlfish (P < 0.05). Many of the apoptotic cells in fish given thehigh-fat diet were distributed around the central vein. Moreover,steatotic hepatocytes seemed more likely to be apoptotic.

Mitochondrial StatusFish that were fed the high-fat diet exhibited significantly

lower SDH and Na+ ,K+ -ATPase activities than control fish(P < 0.05; Table 4). Mitochondrial cytochrome-c concentrationwas also significantly lower in the high-fat diet group than inthe control fish (P < 0.05); however, the variation in cytosoliccytochrome c was the opposite. Moreover, the mitochondrialprotein content per gram of liver was significantly decreased inthe high-fat diet group (P < 0.05).

The expression of mitochondrial genes coding for subunitsof the respiratory chain is presented in Table 5. Expression ofthe ND1 and CYTB genes was significantly lower in fish thatwere fed the high-fat diet than in the control group (P < 0.05).However, the expression of COX1, COX2, and ATP6 was notsignificantly altered by dietary treatment.

FIGURE 1. Photomicrographs (400 × ) of liver tissue from Blunt Snout Bream(scale bar = 50 µm): (a) liver from a control fish (5% fat diet), exhibitingnormal structure; and (b) liver from a fish that received the high-fat (15% fat)diet, presenting extensive intracellular lipid and glycogen droplets. [Color figureavailable online.]

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4


Hepatic Oxidative StatusVariables indicating hepatic oxidative status are presented

in Table 6. Hepatic SOD, GPX, and CAT activities in fish thatreceived the high-fat diet were significantly higher than thosein control fish (P < 0.05). The liver GSH level in fish that weregiven the high-fat diet was significantly lower than the liverGSH level in fish that received the control diet; however, thevariations in GSSG level were just the opposite. Moreover, theGSSG/GSH ratio for the high-fat diet group was significantlygreater than that of the control group. A considerably increased

(approximately twofold) liver MDA content was observed in thehigh-fat diet group relative to the control fish.

Non-specific Immune ResponseHematological–immunological variables are presented in Ta-

ble 7; these results show that the leukocyte count, ACH50 andlysozyme activities, and globulin content in fish that were fedthe high-fat diet were significantly lower than those in controlfish (P < 0.05). However, the erythrocyte count and cortisollevel were both significantly elevated in the high-fat diet group

FIGURE 2. Transmission electron microscope images of hepatocyte, nucleus (N), mitochondria (M), and endoplasmic reticulum (ER) ultrastructure in BluntSnout Bream (scale bar = 1 µm): (a) hepatocytes of a control fish (5% fat diet), showing normal structure; (b) hepatocytes of a fish that received the high-fat (15%fat) diet, presenting extensive intracellular lipid droplets (L); (c) nucleus from a control fish, displaying clear ultrastructure; (d) nucleus from a fish in the high-fatdiet group, exhibiting atrophy (arrow); (e) mitochondria from a control fish, showing well-developed cristae and matrix; and (f), (g) mitochondria from a fish thatreceived the high-fat diet, exhibiting the loss of cristae and matrix with highly hydropic changes (arrows).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

106 LU ET AL.

FIGURE 2. Continued

compared with the control group (P < 0.05). There was no sig-nificant difference in total protein content or albumin contentbetween the two groups.

Results of relative mRNA expression analysis of HSP70 andLEAP2 in the liver are shown in Figure 4. The HSP70 mRNAexpression was significantly elevated by the high-fat diet (P <

0.05), and the increase was about fourfold (3.84 versus 1.00).However, LEAP2 mRNA expression in fish that were fed thehigh-fat diet was significantly lower than that in control fish(0.67 versus 1.00; P < 0.05).

DISCUSSIONIncreases in dietary lipid level support higher growth rates,

partly based on protein sparing in many fish species, includ-

ing the Blunt Snout Bream (Watanabe 2002; Li et al. 2012a).This has driven the use of high-fat diets to become increasinglywidespread in aquaculture. However, some reports have shownthat high-fat diets lead to excess lipid deposition in the liver,inducing negative consequences for fish health and growth—consequences that correlate closely with a high rate of mortalityor poor growth (Du et al. 2008; Lu et al. 2013). In general, fattyliver is documented as a significant problem in cultured fish. Pre-vious researchers have shown that structural alterations of theliver can provide information on metabolic and nutritional sta-tus and liver lesions in fish (Caballero et al. 1999; Coz-Rakovacet al. 2008; Bolla et al. 2011). In our study, many abnormalitieswere observed in the livers of fish that were given the high-fatdiet, including vacuolization due to a combination of increased

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4


FIGURE 3. Hepatocyte apoptosis in Blunt Snout Bream: (a) normal cells (stained amethyst; blue arrow) and apoptotic cells (stained brown; brown arrow) froma control fish (200 × ; scale bar = 50 µm); (b) normal cells and apoptotic cells from a fish that was fed the high-fat (15% fat) diet (200 × ; scale bar = 50 µm);and (c) percentages of TUNEL-positive cells (% of total cell number; mean ± SE) in the two diet groups (TUNEL is described in Methods; asterisk indicates asignificant difference between the high-fat diet group and the control: P < 0.05). [Color figure available online.]

TABLE 4. Mitochondrial status indices (mean ± SE; n = 6) in Blunt SnoutBream that were fed the experimental diets (control = 5% fat; high fat = 15%fat; SDH = succinate dehydrogenase; U/mg mt prot = units per milligram ofmitochondrial protein). Asterisks indicate significant differences between thehigh-fat diet group and the control group (P < 0.05).


Mitochondrial proteincontent (mg/g of wettissue)

2.95 ± 0.05 2.05 ± 0.13*

SDH (U/mg mt prot) 0.12 ± 0.01 0.06 ± 0.01*Na+ ,K+ -ATPase (U/mg

mt prot)6.64 ± 0.13 5.58 ± 0.21*

Mitochondrial cytochromec (µg/g protein)

87.1 ± 3.0 71.0 ± 2.1*

Cytosolic cytochrome c(µg/g protein)

16.2 ± 1.7 27.8 ± 1.0*

TABLE 5. Relative gene expression (mean ± SE; n = 6) for mitochondrialgenes in the livers of Blunt Snout Bream that were fed the experimental diets(control = 5% fat; high fat = 15% fat; see Table 2 for definitions of gene codes).Expression of the target genes is presented relative to the control (set to 1.0).Expression levels were normalized to the mean of the reference gene (β-actin).Asterisks indicate significant differences between the high-fat diet group andthe control group (P < 0.05).

Target gene Control diet High-fat diet

ND1 1.00 ± 0.13 0.15 ± 0.03*CYTB 1.00 ± 0.17 0.68 ± 0.06*COX1 1.00 ± 0.10 1.19 ± 0.16COX2 1.00 ± 0.29 1.26 ± 0.23ATP6 1.00 ± 0.27 1.13 ± 0.10

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

108 LU ET AL.

FIGURE 4. Relative messenger RNA expression of the heat shock protein 70 (HSP70) and liver-expressed antimicrobial peptide 2 (LEAP2) genes in the liversof Blunt Snout Bream fed the control (5% fat) and high-fat (15% fat) diets (values are mean ± SE; n = 6). Expression of the target genes is presented relativeto the control (set to 1.0). Expression levels were normalized to the mean of the reference gene (β-actin). Asterisk indicates a significant difference between thehigh-fat diet group and the control (P < 0.05). [Color figure available online.]

lipid and glycogen, nucleus migration, nuclear atrophy, andapoptosis of hepatocytes. Based on these results, we considerthe liver alterations observed in our fish to be hepatic lesionsdue to fat overload. The liver is the main organ of metabolism,and the growth of fish can be affected by the lesions.

The alteration of mitochondrial structure was very profoundin Blunt Snout Bream that received the 15% fat diet. Du et al.(2008) suggested that the normal structure of mitochondriacould be destroyed by serious hepatic pathology symptoms in-duced mainly by fat deposition and that the structural changesdecreased the function of the mitochondria. Mitochondria playa vital role in cellular homeostasis, and over 90% of cellularenergy is generated in the mitochondria (Manoli et al. 2007).

TABLE 6. Oxidative status (mean ± SE; n = 6) in livers of Blunt SnoutBream that were fed the experimental diets (control = 5% fat; high fat = 15%fat; U/mg prot = units per milligram of protein; SOD = superoxide dismutase;GPX = glutathione peroxidase; CAT = catalase; GSH = reduced glutathione;GSSG = oxidized glutathione; MDA = malondialdehyde). Asterisks indicatesignificant differences between the high-fat diet group and the control group(P < 0.05).


SOD (U/mg prot) 131 ± 9 172 ± 4*GPX (U/mg prot) 53.2 ± 5.5 69.1 ± 4.3*CAT (U/mg prot) 53.7 ± 2.5 66.7 ± 1.4*GSH (µg/mg prot) 13.0 ± 0.7 8.91 ± 0.30*GSSG (µg/mg prot) 8.21 ± 0.70 10.7 ± 0.97*GSSG/GSH ratio 0.63 ± 0.07 1.20 ± 0.09*MDA (nmol/mg prot) 8.14 ± 0.15 16.1 ± 1.4*

Some previous studies have highlighted possible relationshipsbetween mitochondrial function or biochemistry and growthperformance characteristics that are important for aquacultureproduction (Eya et al. 2012). Based on the above, we sus-pected that mitochondrial dysfunction would induce harm tofish growth and feed utilization.

Mitochondrial condition can be assessed by both ultrastruc-ture histology and biochemistry (Bonnard et al. 2008; Bottjeeand Carstens 2009). According to data on mitochondrial struc-ture presented in this study, distinct differences existed betweenthe two dietary treatment groups of Blunt Snout Bream. In fishthat were fed the 15% fat diet, the mitochondria lost cristae, ma-trix, and matrix density with highly hydropic changes. Du et al.

TABLE 7. Hematological–immunological variables (mean ± SE; n = 6) inBlunt Snout Bream that were fed the experimental diets (control = 5% fat; highfat = 15% fat; ACH50 = alternative complement; U = units). Asterisks indicatesignificant differences between the high-fat diet group and the control group(P < 0.05).


Leukocyte count (105/µL) 3.95 ± 0.14 3.45 ± 0.03*Erythrocyte count (106/µL) 1.15 ± 0.03 1.76 ± 0.03*ACH50 (U/mL) 251 ± 15 171 ± 8*Lysozyme (U/mL) 170 ± 10 120 ± 9*Total protein (g/L) 25.5 ± 1.6 22.3 ± 0.8Albumin (g/L) 12.8 ± 0.2 14.5 ± 0.3Globulin (g/L) 12.6 ± 0.5 7.83 ± 0.05*Cortisol (ng/mL) 242 ± 20 310 ± 12*

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4


(2008) suggested that the normal structure of the mitochon-dria could be destroyed mainly by fat overload, thus decreasingthe enzyme activities of the mitochondria. Mitochondria havea central role in energy metabolism of cells and usually pro-vide most of the ATP by oxidative phosphorylation. Ultrastruc-tural lesions in mitochondria would therefore impair the energymetabolism of the cell. Mitochondrial enzymes have been re-ported to exhibit higher activities in fast-growing fish (Houlihanet al. 1993), and enzyme activity can be used as an indicator ofaerobic metabolism in fish muscle (Goolish and Adelman 1987).However, in the present study, SDH and Na+ ,K+ -ATPase ac-tivities in fish that were fed the high-fat diet were lower thanthose in control fish. The two enzymes play important rolesfor energy metabolism; therefore, abnormal activities of theseenzymes could imply disorders in energy metabolism. More-over, there is also a relationship between growth characteris-tics and mitochondrial gene expression, as a down-regulationof some mitochondrial genes was observed in slow-growingRainbow Trout Oncorhynchus mykiss (Eya et al. 2011). In thepresent study, the expression of ND1 and CYTB was down-regulated. According to the literature, reduced expression ofND1 suggests that the assembly of the entire complex I andthe catalytic activity of this complex are impaired (Eya et al.2011). In addition, the lower activities of complex I could causeoxidative damage to the respiratory chain proteins due to theincreased generation of reactive oxygen species (ROS; Bottjeet al. 2002).

As mentioned above, the mitochondrion is also recognizedas a major source of ROS production and endogenous oxidativestress (Qian et al. 2005). Between 2% and 4% of the total O2

consumed by mitochondria is converted to ROS by univalentreduction of O2 to form superoxide (O2

−). When mitochondrialpermeability becomes elevated, ROS produced by the mitochon-dria are released into the cytosol and cause oxidative stress. Inour results, an increase in cytochrome c in the cytosol and aconcomitant decrease in cytochrome c in the mitochondria offish from the high-fat diet group were indicative of cytochrome-c release from the mitochondria, suggesting the onset of themitochondrial permeability transition. Thus, the alteration ofmitochondria may be an important contributor to the leakageof ROS and subsequent oxidative damage. In antioxidant sys-tems, the SOD, CAT, and GPX enzymes play important roles.Superoxide dismutases constitute a group of metalloenzymesthat catalyze the dismutation superoxide radicals (O2

−) into O2

and hydrogen peroxide (H2O2; Winston and Di Giulio 1991).Thus, the elevated SOD activity in the livers of Blunt SnoutBream fed the high-fat diet was an adaptive response to theabnormal oxidative status. In addition, CAT and GPX catalyzethe reduction of both H2O2 and lipid peroxides (Winston andDi Giulio 1991). In the high-fat diet group, the elevated SODactivity resulted in increased H2O2 levels and consequently arise in CAT and GPX activities. Malondialdehyde is an indica-tor that is commonly used to evaluate lipid peroxidation (Parvezand Raisuddin 2005). In the present study, the elevated MDA

level in fish that were fed the high-fat diet indicated an imbal-ance between the generation and removal of ROS. Reduced glu-tathione is a main non-protein thiol and is a primary reductantin cells. The increase in the GSSG/GSH ratio was also inter-preted as an index of oxidative stress in fish that were fed thehigh-fat diet. The decrease in GSH with a concomitant increasein GSSG level was caused by the utilization of nonenzymaticantioxidants to control the enormous amount of free radicals.In some previous studies, high dietary fat resulted in lowersurvival of the farmed fish (Chou et al. 2001; Ai et al. 2008;Zheng et al. 2010). Those results could be due to increasedmalonaldehyde levels in oxidized lipid, as malonaldehyde istoxic to fish (Baker and Davies 1997). We observed no para-sitic diseases or other infections upon necropsy of Blunt SnoutBream that died during the experiment. Therefore, the causeof death in the present study may also be related to oxidativestress.

Another disadvantage of mitochondrial dysfunction is theinduction of apoptosis. When mitochondrial permeability in-creases, proteins located in the intermembrane space of the mi-tochondria are subsequently released into the cytosol. Releaseof cytochrome c from the mitochondrion into the cytosol is aprimary event leading to apoptosis (Sun et al. 2012). Our resultsshowed an increased number of TUNEL-positive cells in BluntSnout Bream that were given the high-fat diet. Apoptosis is acharacteristic feature of liver damage triggered by viral infec-tion, drug abuse, fat overload, or autoimmune disease in mam-mals (Schulze-Osthoff and Haussinger 2007). In our results, theTUNEL-positive cells were often observed with a high degreeof vacuolation, indicating that excess dietary lipid would placea metabolic burden on the liver.

Recently, a link has been established between stress and im-mune functional capacity in fish (Tort 2011). Cortisol is used asgeneral stress indicator and is typically released when fish un-dergo stressful conditions (Stasiack and Bauman 1996). BecauseHSP70 is a stress response protein, expression of the HSP70gene is also used as an indicator of stress response (Kagawaand Mugiya 2002). In the present study, both plasma cortisollevel and liver HSP70 expression were increased in Blunt SnoutBream that received the high-fat diet, thereby indicating a stressresponse. The stress response in fish comprises a cascade ofreactions, including primary, secondary, and tertiary responses(Tahmasebi-Kohyani 2012). It is initiated after the perceptionof stress in the hypothalamus, resulting in a hormonal responseto the release of cortisol (Tahmasebi-Kohyani 2012). The sec-ondary stress response includes metabolic, hematological, andimmunological changes due to the action of cortisol (Bartonand Iwama 1991). The tertiary response is the final stage, whichleads to disease or exhaustion, growth retardation, and finallydeath (Barton and Iwama 1991). According to previous stud-ies, longer-term stress normally shows suppressive effects onimmune function (Tort 2011). The leukocyte count, ACH50 ac-tivity, lysozyme activity, and globulin level are commonly usedas nonspecific immune response indicators in fish. The LEAP2

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

110 LU ET AL.

is an important component of animal innate immunity againstantibiotic-resistant bacterial strains (Liu et al. 2010). All ofthe immune response variables examined in the present studywere lower in the high-fat diet group than in the control group,thus demonstrating immunosuppression in the high-fat treat-ment fish. In a previous study, immune suppression in rats wasassociated with increased oxidative damage and mitochondrialdysfunction (Hao et al. 2009). Oxidative stress induces mito-chondrial dysfunction as well as apoptosis. Apoptosis can causeimmune suppression by (1) inducing depletion of various im-mune cells, resulting in the loss of key antimicrobial functions;and (2) inducing immunosuppressive effects in the survivingcells (Hotchkiss and Nicholson 2006). Therefore, the decreasein leukocyte counts for the high-fat diet group may be relatedto apoptosis. The reduced lysozyme activity might be ascribedto the relatively low leukocyte number, as fish serum lysozymeis believed to be of leukocytic origin (Lie et al. 1989). Liver isconsidered the main source of complement proteins, and hepaticdamage due to oxidative stress and mitochondrial dysfunctionmight therefore negatively affect complement proteins. Emerg-ing knowledge suggests that mitochondria function as centrallypositioned hubs in the human innate immune system (West et al.2011). A reduced ATP supply due to mitochondrial dysfunctioncould be one reason for a decrease in protein synthesis, therebycausing attenuation in RNA synthesis and calcium transport ofimmune cells and in their functions (Buttgereit et al. 2000). Onthe whole, reductions in immune system efficacy occurred as aconsequence of oxidative stress and mitochondrial dysfunctionin Blunt Snout Bream. However, further experiments on howmitochondrial dysfunction suppresses immunity in fish are stillneeded.

In summary, Blunt Snout Bream showed a high rate of mor-tality and poor growth after receiving the high-fat diet. The histo-logical results clearly showed that the high fat intake resulted infat accumulation and structural alterations of the hepatocytes,mitochondria, and nuclei. Impairment of the mitochondria inthe high-fat diet group was indicated by mitochondrial swellingand the loss of cristae and matrix. An increase in cytosoliccytochrome c with a concomitant decrease in mitochondrial cy-tochrome c for fish that were given the high-fat diet suggestedthe onset of the mitochondrial permeability transition, whichlargely contributes to oxidative stress and apoptosis. Elevatedplasma cortisol levels and liver HSP70 expression provide evi-dence of increased stress among fish in the high-fat diet group.Moreover, nonspecific immune responses were also suppressedin fish that were given the high-fat diet. Overall, the intake ofexcessive fat impaired mitochondrial bioenergetics and phys-iological functions. The dysfunction of the mitochondria sub-sequently mediated oxidative stress and hepatocyte apoptosis,which in turn led to reductions in immune system efficacy. Thepresent results have implications for our understanding of the ad-verse effects of fatty liver and the development of those effects,and such information will be helpful in decreasing metabolicdiseases among cultured fish.

ACKNOWLEDGMENTSThis project was supported by the National Natural Science

Foundation of China (31172418) and National Natural ScienceFoundation of China for Young Scholars (31202005). We thankHuang Guo-Qing (College of Veterinary Medicine, NanjingAgricultural University) for his assistance with the photomi-crographs.

REFERENCESAi, Q. H., J. Z. Zhao, K. S. Mai, W. Xu, B. P. Tan, H. M. Ma, and Z. G. Liufu.

2008. Optimal dietary lipid level for Large Yellow Croaker (Pseudosciaenacrocea) larvae. Aquaculture Nutrition 14:515–522.

Baker, R. T. M., and S. J. Davies. 1997. Modulation of tissue α-tocopherolin African Catfish, Clarias gariepinus (Burchell), fed oxidized oils, and thecompensatory effect of supplemental dietary vitamin E. Aquaculture Nutri-tion 3:91–97.

Barton, B. A., and G. K. Iwama. 1991. Physiological changes in fish from stressin aquaculture with emphasis on the response and effects of corticosteroids.Annual Review of Fish Diseases 1:3–26.

Bolla, S., O. Nicolaisen, and A. Amin. 2011. Liver alterations induced bylong term feeding on commercial diets in Atlantic Halibut (Hippoglossushippoglossus L.) females: histological and biochemical aspects. Aquaculture312:117–125.

Bonnard, C., A. Durand, S. Peyrol, E. Chanseaume, M.-A. Chauvin, B. Morio, H.Vidal, and J. Rieusset. 2008. Mitochondrial dysfunction results from oxidativestress in the skeletal muscle of diet-induced insulin-resistant mice. Journal ofClinical Investigation 118:789–800.

Bottjee, W., and G. E. Carstens. 2009. Association of mitochondrial function andfeed efficiency in poultry and livestock species. Journal of Animal Science87:E48–E63.

Bottje, W., M. Iqbal, Z. X. Tang, D. Cawthon, and R. Okimoto. 2002. Associationof mitochondrial function and feed efficiency within a single genetic line ofmale broilers. Poultry Science 81:546–555.

Buttgereit, F., G. R. Burmester, and M. D. Brand. 2000. Bioenergetics of immunefunctions: fundamental and therapeutic aspects. Immunology Today 21:194–199.

Caballero, M. J., G. Lopez-Calero, J. Socorro, F. J. Roo, M. S. Izquierdo, and A.J. Fernandez. 1999. Combined effect of lipid level and fish meal quality onliver histology of Gilthead Seabream (Sparus aurata). Aquaculture 179:277–290.

Catacutan, M. R., and R. M. Coloso. 1997. Growth of juvenile Asian Seabass,Lates calcarifer, fed varying carbohydrate and lipid levels. Aquaculture149:137–144.

Chou, R. L., M. S. Su, and H. Y. Chen. 2001. Optimal dietary protein and lipidlevels for juvenile Cobia (Rachycentron canadum). Aquaculture 193:81–89.

Coz-Rakovac, R., I. Strunjak-Perovic, M. Hacmanjek, P. N. Topic, Z. Lipej, andB. Sostaric. 2005. Blood chemistry and histological properties of wild andcultured Sea Bass (Dicentrarchus labrax) in the north Adriatic Sea. VeterinaryResearch Communications 29:677–687.

Coz-Rakovac, R., I. Strunjak-Perovic, N. Topic Popovic, M. Hacmanjek, T.Smuc, M. Jadan, Z. Lipej, and Z. Homen. 2008. Cage culture effects onMullet (Mugilidae) liver histology and blood chemistry profile. Journal ofFish Biology 72:2557–2569.

Dabas, A., N. S. Nagpure, R. Kumar, B. Kushwaha, P. Kumar, and W. S. Lakra.2012. Assessment of tissue-specific effect of cadmium on antioxidant defensesystem and lipid peroxidation in freshwater murrel, Channa punctatus. FishPhysiology and Biochemistry 38:469–482.

Du, Z. Y., P. Clouet, L. M. Huang, P. Degrace, W. H. Zheng, J. G. He, L. X. Tian,and Y. J. Liu. 2008. Utilization of different dietary lipid sources at high levelin herbivorous Grass Carp (Ctenopharyngodon idella): mechanism related tohepatic fatty acid oxidation. Aquaculture Nutrition 14:77–92.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4


Du, Z. Y., P. Clouet, W. H. Zheng, P. Degrace, L. X. Tian, and Y. J. Liu. 2006.Biochemical hepatic alterations and body lipid composition in the herbivorousGrass Carp (Ctenopharyngodon idella) fed high-fat diets. British Journal ofNutrition 95:905–915.

Eya, J. C., M. F. Ashame, and C. F. Pomeroy. 2011. Association of mitochondrialfunction with feed efficiency in Rainbow Trout: diets and family effects.Aquaculture 321:71–84.

Eya, J. C., M. F. Ashame, C. F. Pomeroy, B. B. Manning, and B. C. Peterson.2012. Genetic variation in feed consumption, growth, nutrient utilizationefficiency and mitochondrial function within a farmed population of ChannelCatfish (Ictalurus punctatus). Comparative Biochemistry and Physiology PartB 163:211–220.

Goolish, E. M., and I. R. Adelman. 1987. Tissue specific cytochrome c oxi-dase activity in Largemouth Bass: the metabolic cost of feeding and growth.Physiological Zoology 60:454–464.

Hao, J., W. Shen, and C. Tian, et al. 2009. Mitochondrial nutrients improveimmune dysfunction in the type 2 diabetic Goto-Kakizaki rats. Journal ofCellular and Molecular Medicine 13:701–711.

Hotchkiss, R. S., and D. W. Nicholson. 2006. Apoptosis and caspases regulatedeath and inflammation in sepsis. Nature Reviews Immunology 6:813–822.

Houlihan, D. F., E. M. Mathers, and A. Foster. 1993. Biochemical correlates ofgrowth rate in fish. Pages 45–67 in J. C. R. Rankin and F. B. Jensen, editors.Fish ecophysiology. Chapman and Hall, London.

Iqbal, M., N. R. Pumford, Z. X. Tang, K. Lassiter, T. Wing, M. Cooper, and W.Bottje. 2004. Low feed efficient broilers within a single genetic line exhibithigher oxidative stress and protein expression in breast muscle with lowermitochondrial complex activity. Poultry Science 83:474–484.

IUBMB (International Union of Biochemistry and Molecular Biology). 1992.Enzyme nomenclature 1992. Academic Press, San Diego, California.

Johnson, C. W., D. L. Timmons, and P. E. Hall. 2002. Essential laboratory math-ematics: concepts and applications for the chemical and clinical laboratorytechnician. Cengage Learning, Florence, Italy.

Kagawa, N., and Y. Mugiya. 2002. Brain HSP70 mRNA expression is linkedwith plasma cortisol levels in Goldfish (Carassius auratus) exposed to apotential predator. Zoological Science 19:735–740.

Li, X. F., Y. Y. Jiang, W. B. Liu, and X. P. Ge. 2012a. Protein-sparing effectof dietary lipid in practical diets for Blunt Snout Bream (Megalobrama am-blycephala) fingerlings: effects on digestive and metabolic responses. FishPhysiology and Biochemistry 38:529–541.

Li, X. F., W. B. Liu, Y. Y. Jiang, H. Zhu, and X. P. Ge. 2010. Effects of dietaryprotein and lipid levels in practical diets on growth performance and bodycomposition of Blunt Snout Bream (Megalobrama amblycephala) fingerlings.Aquaculture 303:65–70.

Li, X. F., W. B. Liu, K. L. Lu, W. N. Xu, and Y. Wang. 2012b. Dietary car-bohydrate/lipid ratios affect stress, oxidative status and non-specific immuneresponses of fingerling Blunt Snout Bream, Megalobrama amblycephala.Fish and Shellfish Immunology 33:316–323.

Lie, Ø., Ø. Evensen, A. Sorensen, and E. Froysadal. 1989. Study on lysozymeactivity in some fish species. Diseases of Aquatic Organisms 6:1–5.

Liu, F., J. L. Li, G. H. Yue, J. J. Fu, and Z. F. Zhou. 2010. Molecular cloning andexpression analysis of the liver-expressed antimicrobial peptide 2 (LEAP-2)gene in Grass Carp. Veterinary Immunology and Immunopathology 133:133–43.

Lowry, O. H., N. J. Rosebrough, and A. L. Farr. 1951. Protein measurementwith the folin-phenol reagent. Journal of Biological Chemistry 193:265–275.

Lu, K. L., W. N. Xu, J. Y. Li, X. F. Li, G. Q. Huang, and W. B. Liu. 2013.Alterations of liver histology and blood biochemistry in Blunt Snout BreamMegalobrama amblycephala fed high-fat diets. Fisheries Science 79:661–671.

Lygren, B., K. Hamre, and R. Waagbø. 1999. Dietary pro- and antioxidants andsome health parameters in Atlantic Salmon (Salmo salar). Journal of AquaticAnimal Health 11:211–221.

Manoli, I., S. Alesci, M. R. Blackman, Y. A. Su, O. M. Rennert, and G. P.Chrousos. 2007. Mitochondria as key components of the stress response.Trends in Endocrinology and Metabolism 18:190–198.

Maral, J., K. Puget, and A. M. Michelson. 1977. Comparative study of superox-ide dismutase, catalase and glutathione peroxidase levels in erythrocytes ofdifferent animals. Biochemical and Biophysical Research Communications77:1525–1535.

McCormick, S. D., B. T. Bjornsson, M. Sheridan, C. Eilertson, J. B. Carey,and M. O’Dea. 1995. Increased day length stimulates plasma growth hor-mone and gill Na+ -K+ -ATPase in Atlantic Salmon (Salmo salar). Journalof Comparative Physiology B 165:245–254.

Mignotte, B., and J. L. Vayssiere. 1998. Mitochondria and apoptosis. EuropeanJournal of Biochemistry 252:1–15.

Montero, D., L. Tort, L. Izquierdo, and J. M. Vergara. 1998. Depletion ofserum alternative complement pathway activity in gilthead seabream causedby α-tocopherol and n-3 HUFA dietary deficiencies. Fish Physiology andBiochemistry 18:399–407.

Nakano, M. 1990. Assay for superoxide dismutase based on chemiluminescenceof luciferin analog. Methods in Enzymology 186:227–232.

Nanton, D. A., S. P. Lall, and M. A. McNiven. 2001. Effects of dietary lipid levelon liver and muscle lipid deposition in juvenile Haddock, Melanogrammusaeglefinus L. Aquaculture Research 32:225–234.

Obach, A., C. Quentel, and F. B. Laurencin. 1993. Effects of alpha-tocopheroland dietary oxidized fish oil on the immune response of Sea Bass Dicentrar-chus labrax. Diseases of Aquatic Organisms 15:175–185.

Parvez, S., and S. Raisuddin. 2005. Protein carbonyls: novel biomakersof exposure to oxidative stress-inducing pesticides in freshwater fishChanna punctata (Bloch). Environmental Toxicology and Pharmacology 20:112–117.

Philip, G. H., P. M. Reddy, and G. Sridevi. 1995. Cypermethrin induced in vivoalterations in the carbohydrate metabolism of freshwater fish, Labeo rohita.Ecotoxicology and Environmental Safety 31:173–178.

Qian, W., M. Nishikawa, A. M. Haque, M. Hirose, M. Mashimo, E. Sato, andM. Inoue. 2005. Mitochondrial density determines the cellular sensitivityto cisplatin-induced cell death. American Journal of Physiology 289:1466–1475.

Reinhold, J. G. 1953. Manual determination of serum total protein, albumin andglobulin fractions by Biuret method. Standard method of clinical chemistry.Academic Press, New York.

Roberts, R. J. 1989. Fish pathology, 2nd edition. Bailliere Tindall, London.Rueda-Jasso, R., L. E. C. Conceicao, J. Dias, W. De Coen, E. Gomes, J. F. Rees,

F. Soares, M. T. Dinis, and P. Sorgeloos. 2004. Effect of dietary non-proteinenergy levels on condition and oxidative status of Senegalese Sole (Soleasenegalensis) juveniles. Aquaculture 231:417–433.

Sandelin, B. A., A. H. Brown Jr., M. A. Brown, C. Ojano-Dirain, and R. T.Baublits. 2005. Association of mitochondrial function with feed efficiencyin Angus cattle. Arkansas Agricultural Experiment Station Research Series535:118–121.

Sargent, J. R., and A. G. J. Tacon. 1999. Development of farmed fish: a nutri-tionally necessary alternative to meat. Proceedings of the Nutrition Society58:377–383.

Schulze-Osthoff, K., and D. Haussinger. 2007. Apoptosis in the liver: a matterof ion fluxes and oxidative stress. Liver International 27:1039–1044.

Sordillo, L. M., and S. L. Aitken. 2009. Impact of oxidative stress on thehealth and immune function of dairy cattle. Veterinary Immunology andImmunopathology 128:104–109.

Stasiack, A. S., and C. P. Bauman. 1996. Neutrophil activity as a potent indicatorfor concomitant analysis. Fish and Shellfish Immunology 37:539.

Sun, X. D., H. Pan, H. W. Tan, and Y. R. Yu. 2012. High free fatty acids levelrelated with cardiac dysfunction in obese rats. Diabetes Research and ClinicalPractice 95:251–259.

Szlaminska, M., A. M. Escaffre, and P. Bergot. 1991. Utilization of dietarypregelatinized starch by Common Carp (Cyprinus carpio L.) larvae. Journalof Animal Physiology and Animal Nutrition 65:65–71.

Tahmasebi-Kohyani, A., S. Keyvanshokooh, A. Nematollahi, N. Mahmoudi,and H. Pasha-Zanoosi. 2012. Effects of dietary nucleotides supplementa-tion on Rainbow Trout (Oncorhynchus mykiss) performance and acute stressresponse. Fish Physiology and Biochemistry 38:431–440.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4

112 LU ET AL.

Tort, L. 2011. Stress and immunomodulation in fish. Developmental and Com-parative Immunology 35:1366–1375.

Watanabe, T. 2002. Strategies for further development of aquatic feeds. FisheriesScience 68:242–252.

Wei, Y., R. S. Rector, J. P. Thyfault, and J. A. Ibdah. 2008. Nonalcoholic fattyliver disease and mitochondrial dysfunction. World Journal of Gastroenterol-ogy 14:193–199.

West, A. P., G. S. Shadel, and S. Ghosh. 2011. Mitochondria in innate immuneresponses. Nature Reviews Immunology 11:389–402.

Winberg, S., and O. Lepage. 1998. Elevation of brain 5-HT activity, POMCexpression and plasma cortisol in socially subordinate Rainbow Trout and

subordinate juvenile Arctic Charr. Journal of Experimental Biology 179:181–195.

Winston, G. W., and R. T. Di Giulio. 1991. Prooxidant and antioxidant mecha-nisms in aquatic organisms. Aquatic Toxicology 19:137–161.

Zheng, K., X. Zhu, D. Han, Y. Yang, W. Lei, and S. Xie. 2010. Ef-fects of dietary lipid levels on growth, survival and lipid metabolismduring early ontogeny of Pelteobagrus vachelli larvae. Aquaculture 299:121–127.

Zhou, Z., Z. Ren, H. Zeng, and B. Yao. 2008. Apparent digestibility of vari-ous feedstuffs for Bluntnose Black Bream, Megalobrama amblycephala Yih.Aquaculture Nutrition 14:153–165.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:54

15

June

201

4



The Endemic Copepod Calanus pacificus californicus asa Potential Vector of White Spot Syndrome VirusFernando Mendoza-Canoa, Arturo Sánchez-Paza, Berenice Terán-Díazb, Diego Galván-Alvareza, Trinidad Encinas-Garcíaa, Tania Enríquez-Espinozac & Jorge Hernández-Lópeza

a Centro de Investigaciones Biológicas del Noroeste S. C., Laboratorio de Referencia, Análisisy Diagnóstico en Sanidad Acuícola, Calle Hermosa 101, Col. Los Ángeles, Hermosillo, SonoraC. P. 83260, Mexicob Universidad Autónoma de Baja California Sur Dirección, Carretera al Sur KM 5.5, La Paz,Baja California Sur C.P. 23080, Mexicoc Departamento de Investigaciones Científicas y Tecnológicas, Universidad de Sonora, Av.Colosio s/n, entre Sahuaripa y Reforma, Hermosillo, Sonora 83000, MexicoPublished online: 04 Jun 2014.

To cite this article: Fernando Mendoza-Cano, Arturo Sánchez-Paz, Berenice Terán-Díaz, Diego Galván-Alvarez, TrinidadEncinas-García, Tania Enríquez-Espinoza & Jorge Hernández-López (2014) The Endemic Copepod Calanus pacificuscalifornicus as a Potential Vector of White Spot Syndrome Virus, Journal of Aquatic Animal Health, 26:2, 113-117, DOI:10.1080/08997659.2013.852635







http://dx.doi.org/10.1080/08997659.2013.852635




COMMUNICATION

The Endemic Copepod Calanus pacificus californicusas a Potential Vector of White Spot Syndrome Virus

Fernando Mendoza-Cano and Arturo Sanchez-PazCentro de Investigaciones Biologicas del Noroeste S. C., Laboratorio de Referencia,Analisis y Diagnostico en Sanidad Acuıcola, Calle Hermosa 101, Col. Los Angeles, Hermosillo,Sonora C. P. 83260, Mexico

Berenice Teran-DıazUniversidad Autonoma de Baja California Sur Direccion, Carretera al Sur KM 5.5, La Paz,Baja California Sur C.P. 23080, Mexico

Diego Galvan-Alvarez and Trinidad Encinas-GarcıaCentro de Investigaciones Biologicas del Noroeste S. C., Laboratorio de Referencia,Analisis y Diagnostico en Sanidad Acuıcola, Calle Hermosa 101, Col. Los Angeles, Hermosillo,Sonora C. P. 83260, Mexico

Tania Enrıquez-EspinozaDepartamento de Investigaciones Cientıficas y Tecnologicas, Universidad de Sonora, Av. Colosio s/n,entre Sahuaripa y Reforma, Hermosillo, Sonora 83000, Mexico

Jorge Hernandez-Lopez*Centro de Investigaciones Biologicas del Noroeste S. C., Laboratorio de Referencia,Analisis y Diagnostico en Sanidad Acuıcola, Calle Hermosa 101, Col. Los Angeles, Hermosillo,Sonora C. P. 83260, Mexico

AbstractThe susceptibility of the endemic copepod Calanus pacificus

californicus to white spot syndrome virus (WSSV) was establishedby the temporal analysis of WSSV VP28 transcripts by quantita-tive real-time PCR (qRT-PCR). The copepods were collected froma shrimp pond located in Bahia de Kino Sonora, Mexico, andchallenged per os with WSSV by a virus–phytoplankton adhesionroute. Samples were collected at 0, 24, 48 and 84 h postinoculation(hpi). The VP28 transcripts were not detected at early stages (0and 24 hpi); however, some transcript accumulation was observedat 48 hpi and gradually increased until 84 hpi. Thus, these resultsclearly show that the copepod C. pacificus californicus is suscepti-ble to WSSV infection and that it may be a potential vector for thedispersal of WSSV. However, further studies are still needed to cor-relate the epidemiological outbreaks of WSSV with the presence ofcopepods in shrimp ponds.

*Corresponding author: [email protected] May 22, 2013; accepted October 2, 2013

Aquaculture is a valuable source of food, an especially im-portant economic activity, and provides employment to a signif-icant number of people in several countries around the world.One of the challenges in contemporary aquaculture is to produceorganisms successfully at the lowest cost; however, the shrimpfarming industry has been severely affected by the emergenceof viral diseases, causing significant economic losses (Lightner1996). The white spot syndrome virus (WSSV) is considered tobe the most devastating and virulent viral agent threatening thepenaeid shrimp culture industry (Moser et al. 2012), and underfarming conditions, mortalities can reach 100% within 3–10 dafter the onset of clinical signs (Chou et al. 1995). To date,more than 90 species of arthropods, including shrimp, lobsters,crayfish, and crabs, and a number of nonarthropod species asrotifers, polychaete worms, and mollusks, have been reportedto be potential carriers or hosts of WSSV (Sanchez-Paz 2010).

113

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

114 MENDOZA-CANO ET AL.

Furthermore, some of these organisms are commonly used aslive feed in the culture of penaeid shrimp larvae (Reymond andLagardere 1990). Thus, the identification of susceptible speciesto infection with WSSV should be considered an essential topicfor the use of sanitary programs and management strategiesto reduce the negative impact of this disease in shrimp farms.Copepods, rotifers, and Artemia may act as mechanical vec-tors for WSSV dispersion, but only few reports have demon-strated their susceptibility to this lethal virus (Zhang et al. 2006,2008, 2010; Chang et al. 2011). Copepods are a diverse groupof crustaceans, and over 12,000 marine species have been de-scribed. They constitute the biggest source of animal protein inthe oceans and provide food for many fishery species (Olsen2007). Some calanoid copepod species are regularly used asaquaculture feeds, and a number of species are highly abundantin the estuarine coastal waters surrounding shrimp grow-outponds in Mexico, where they are an important part of the natu-ral diet of penaeid shrimp species, at least in their postlarval andjuvenile phases (Martınez-Cordova et al. 2011). Thus, the aimof this study was to investigate the susceptibility of the endemiccopepod Calanus pacificus californicus to the WSSV.

METHODSSample collection.—In April 2012, copepods samples were

collected from a shrimp pond located in Bahia de KinoSonora, Mexico, (Figure 1) using a 100-µm-mesh planktonnet. The specimens were reared in 60-L plastic containerscontaining 35‰ seawater, at 28◦C with continuous aeration,and were fed daily with three species of algae (Chaetocerossp., Dunaliella sp., and Tetraselmis sp.). Partial nucleotidesequences of the mitochondrial 16S rRNA were used foridentification of the organisms according to the method de-scribed by Lindeque et al. (1999). The primers used were16SAR (5′-CGCCTGTTTAACAAAAACAT-3′) y 16SBR (5′-ATTCAACATCGAGGTCACAAAC-3′), and the PCR thermo-cycling conditions were set as follows: 95◦C for 5 min, 35 cyclesof 95◦C for 1 min, 55◦C for 30 s, 72◦C for 1 min, and a finalextension step at 72◦C for 10 min. The purified PCR ampliconswere directly sequenced using a Perkin Elmer/Applied Biosys-tems automatic sequencer, model 3730, at the facilities of theInstituto de Biotecnologıa, Universidad Nacional Autonoma deMexico. The obtained 16S rRNA gene sequence was comparedwith sequences previously submitted to GenBank by othersby using the program BLASTN with a minimum expect-valuethreshold of 1 × 10–10.

Previous to the above mentioned procedure, copepods weremanually separated from the bulk sample and tested for WSSVdetection by quantitative PCR (qPCR) according to the methoddescribed by Mendoza-Cano and Sanchez-Paz (2013) using iQSYBR Green Supermix (Bio-Rad) and the primers VP28-140Fw(5′-AGGTGTGGAACAACACATCAAG-3′) and VP28-140Rv(5′-TGCCAACTTCATCCTCATCA-3′) by using the followingprotocol: 95◦C for 5 min, 30 cycles of 95◦C for 30 s, 61◦C for

FIGURE 1. The arrow represents the location of the shrimp pond located inBahia de Kino Sonora, Mexico, where samples were collected.

30 s, and 72◦C for 30 s (signal acquisition). For WSSV detectionDNA was isolated from copepods following the GeneClean Spinprotocol (MP Biomedicals).

Inoculum preparation and infection assay.—Virus inoculumwas prepared from WSSV-infected shrimp tissue according tothe method described by Escobedo-Bonilla et al. (2005) withsome modifications. The tissue was homogenized in sterilephosphate-buffered saline (PBS), pH 7.4 (1:6 w/v), and clarifiedby centrifugation at 3,000 × g for 20 min at 4◦C. The super-natant was then removed and centrifuged once more at 15,000× g for 20 min. The recovered supernatant was then filteredthrough a 0.45-µm membrane (Millipore) and used for exper-imental assays. For the susceptibility study on the copepod C.pacificus californicus, organisms were challenged with freshlyprepared WSSV inoculum according to the methodology (virus–phytoplankton adhesion route) proposed by Zhang et al. (2008).Accordingly, 5 mL of the inoculum were mixed for 30 min with50 mL of a mixed algal culture of Chaetoceros sp., Dunaliellasp., and Tetraselmis sp., and subsequently the copepods werefed once with this mix. About 30 specimens of each treatmentwere sampled at 0, 24, 48, and 84 h postinoculation (hpi) andpreserved in 96% ethyl alcohol for further analysis. The control

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

COMMUNICATION 115

FIGURE 2. Nucleotide sequence alignment of samples collected from Bahia de Kino Sonora, Mexico, and that reported for C. pacificus californicus (AF295333)in GenBank. Letters in bold text represent nucleotide differences.

copepods were treated in the same manner as in the suscepti-bility study, except that the algal culture was first mixed withsterile PBS (0.9% NaCl) as a substitute for the WSSV inoculum.

Temporal analysis of WSSV VP28 transcription.—Total RNAwas purified with TRIzol Reagent (Invitrogen) from copepodscollected at each sampling time according to the manufacturer’sinstructions and then treated with DNase I (Invitrogen) to elim-inate any residual DNA. The concentration of total RNA wascalculated by measuring its optical density (OD) at 260 nm us-ing a Nanodrop ND-2000 spectrophotometer. For monitoringWSSV temporal transcription in challenged specimens and thecontrol treatment, gene fragments of the VP28 viral envelopeprotein were amplified from the RNA isolated with RT-PCR us-ing iScript One-Step RT-PCR Kit With SYBR Green (Bio-Rad)following the methodology described by Mendoza-Cano andSanchez-Paz (2013) with an initial cDNA synthesis of 10 minat 50◦C and following the protocol described above. As qualitycontrol, DNA contamination of the RNA isolates was confirmedby PCR. Equal RNA concentrations (30 ng/µL) were used formolecular tests.

RESULTS AND DISCUSSIONCopepods are an important part of the penaeid shrimp diet

and the most common and abundant zooplankton species found

in shrimp culture ponds (Martınez-Cordova and Pena-Messina2005). In Bahia de Kino, Sonora, copepods often represent>90% of the zooplankton abundance during spring and fall,while its relative abundance decreases progressively to an abun-dance of 39% throughout the ecosystem in summer and winter(Salas 2011).

Sequence analysis of the partial 16S rRNA gene obtainedfrom the copepods samples (456 bp) showed a 99% similarityto the best-scoring reference sequence (C. pacificus californicus,GenBank accession number AF295333), confirming the identityof the sample as C. pacificus californicus (Figure 2).

It has been recently proposed that WSSV may be dispersedto neighboring ponds or farms through the water as viral parti-cles suspended in water and/or by means of particulate fractionsas viral particles carried by zooplankton or adhered to microal-gae (Esparza-Leal et al. 2009). Copepods collected from theshrimp pond were tested for WSSV by qPCR and diagnosed asWSSV-free. However, this does not imply that these organismscould not be vectors of this disease. Very low level infections incrustaceans can occur, sometimes at undetectable levels, evenby highly sensitive PCR procedures (Walker and Winton 2010).The temporal expression of VP28 in C. pacificus californicusafter WSSV exposure was analyzed by real-time PCR (RT-PCR)(Figure 3). No amplification for VP28 transcripts was observedat 0 and 24 hpi; however, transcripts were continuously detected

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

116 MENDOZA-CANO ET AL.

FIGURE 3. Detection and quantification of WSSV VP28 transcripts by qPCRin experimentally infected C. pacificus californicus. Standard curves for RT-PCR are shown based on 10-fold serial dilutions of a 141-bp WSSV VP28 DNAfragment. The letters a, b, c, d, and e represent 6.57 × 105, 6.57 × 104, 6.57× 103, 6.57 × 102, and 657 copies/ng of total DNA, respectively. The solidlines represented by the numbers 1 and 2 are WSSV VP28 transcripts at 84 hpiand 48 hpi, respectively. The shaded box represents the end-point amplificationof VP28 as follows: lane 1: low mass DNA ladder; lane 2: negative control; lane3: 0 hpi; lane 4: 24 hpi; lane 5: 48 hpi; lane 6: 84 hpi; lanes 7–11: 10-fold serialdilutions (from 6.57 × 105 to 657 copies/ng) of total DNA.

from 48 to 84 hpi. Based on the quantification cycle (Cq) (Bustinet al. 2009), at 48 hpi 1.25 × 103 viral copies/ng of DNA weredetected, and this amount increased to 1.13 × 105 copies/ngof DNA at 84 hpi. These results are in agreement with thosereported by Chang et al. (2011) in a different copepod species(Apocyclops royi) in which transcripts of the viral protein weredetected. Differences between this study and that of Chang et al.(2011) must be noted, as transcripts for VP28 were detected inA. royi at 24 hpi, while in C. pacificus californicus collected inBahıa de Kino, transcripts were not detected until 48 hpi. Thisdifference in the detection of transcripts of VP28 may be dueeither to the susceptibility of both species to WSSV or to thetotal number of genomic copies inoculated to each species. Re-ports in penaeid shrimp indicate that expression of this protein(VP28) occurs in the first 6 hpi (Zhang et al. 2002). This virushas been reported in at least 93 WSSV vectors or hosts, includingpenaeid shrimps, lobsters, crabs, crayfish, shrimp, polychaetes,and different components of the zooplankton as brine shrimp,cladocerans, rotifers, and copepods (Sanchez-Paz 2010).

It is important to note that a positive detection of WSSVby PCR does not imply that the copepod is susceptible toWSSV because this technique is based on the presence or ab-sence of fragments of the viral genome that could be localized

intracellularly, on the surface, or in the organism’s intestinalcontents.

To confirm viral infection, several techniques, such as histol-ogy, in situ hybridization, or transmission electron microscopy,could be used. In this case, the gene expression profile of WSSVusing RT-PCR may be considered a useful tool to define whetherthe virus is in an active (infective) or an inactive (noninfective)form, and it is applicable not only to WSSV, but also to otherviruses (Chang et al. 2011).

Water may be a major pathway for WSSV dispersion intoa shrimp farm (Lotz and Lightner 1999), and the virus can bealso carried by zooplankton or can be attached to microalgae(Esparza-Leal et al. 2009). Thus, WSSV outbreaks have beenrelated to ocean currents across the Gulf of California.The re-sults obtained in the present study clearly show that the copepodC. pacificus californicus is susceptible to WSSV and that it islikely virus dispersion may occur through infected copepods.However, more tests are needed to correlate the epidemiologicaloutbreaks of white spot syndrome and the presence of copepodsin shrimp ponds.

ACKNOWLEDGMENTSThe authors thank Consejo Nacional de Ciencia y Tec-

nologıa (CONACyT), Mexico, for partially funding this project(102744), project 960-1 from CIBNOR, and Acuıcola la Bor-bolla. We also thank our colleagues from Instituto de Biotec-nologıa, of the Universidad Nacional Autonoma de Mexico forsequencing the samples.

REFERENCESBustin, S. A., V. Benes, J. A. Garson, J. Hellemans, J. Huggett, M. Kubista, R.

Mueller, T. Nolan, M. W. Pfaffl, G. L. Shipley, J. Vandesompele, and C. T.Wittwer. 2009. The MIQE guidelines: minimum information for publicationsof quantitative real-time PCR experiments. Clinical Chemistry 55:611–622.

Chang, Y. S., T. C. Chen, W. J. Liu, JS Hwang, G. H. Kou, and C. F. Lo. 2011.Assessment of the roles of copepod Apocyclops royi and bivalve molluskMeretrix lusoria in white spot syndrome virus transmission. Marine Biotech-nology 13:909–917.

Chou, H. Y., C. Y. Huang, C. H. Wuang, H. C. Chiang, and C. F. Lo. 1995Pathogenicity of a baculovirus infection causing white spot syndrome incultured penaeid shrimp in Taiwan. Diseases of Aquatic Organisms 23:165–173.

Escobedo-Bonilla, C. M., M. Wille, V. A. Sanz, P. Sorgeloos, M. B. Pensaert,and H. J. Nauwynck. 2005. In vivo titration of white spot syndrome virus(WSSV) in specific pathogen-free Litopenaeus vannamei by intramuscularand oral routes. Diseases of Aquatic Organisms 66:163–170.

Esparza-Leal, H. M., C. M. Escobedo-Bonilla, R. Casillas-Hernandez, P.Alvarez-Ruız, G. Portillo-Clark, R. C. Valerio-Garcıa, J. Hernandez-Lopez, J.Mendez-Lozano, N. Vibanco-Perez, and F. J. Magallon-Barajas. 2009. Detec-tion of white spot syndrome virus in filtered shrimp-farm water fractions andexperimental evaluation of its infectivity in Penaeus (Litopenaeus) vannamei.Aquaculture 292:16–22.

Lightner, D. V. 1996. Epizootiology, distribution and the impact on internationaltrade of two penaeid shrimp viruses in the Americas. Revue Scientifique etTechnique 15:579–601.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

COMMUNICATION 117

Lindeque, P. K., R. P. Harris, M. B. Jones, and G. R. Smerdon. 1999. Simplemolecular method to distinguish the identity of Calanus species (Copepoda:Calanoida) at any developmental stage. Marine Biology 133:91–96.

Lotz, J. M., and D. V. Lightner. 1999. Shrimp biosecurity: pathogen andpathogen exclusion. Pages 67–74 in R. A. Bullis and G. D. Pruder, edi-tors. Controlled and biosecure production systems: evolution and integrationof shrimp and chicken models. World Aquaculture Society, Sidney, Australia.

Martınez-Cordova, L. R., A. Campana-Torres, and M. Martınez-Porchas. 2011.Effect of supplying four copepod densities (Acartia sp. and Calanus pacificus)on the productive response of Litopenaeus vannamei pregrown intensively atmicrocosm level. Ciencias Marinas 37:415–423.

Martınez-Cordova, L. R., and E. Pena-Messina. 2005. Biotic communitiesand feeding habits of Litopenaeus vannamei (Boone 1931) and Litopenaeusstylirostris (Stimpson 1974) in monoculture and polyculture semi-intensiveponds. Aquaculture Research 36:1075–1084.

Mendoza-Cano, F., and A. Sanchez-Paz. 2013. Development and validation ofa quantitative real-time polymerase chain assay for universal detection of thewhite spot syndrome virus in marine crustaceans. Virology Journal 10:186.

Moser, J. R., D. A. Galvan Alvarez, F. Mendoza Cano, T. Encinas Garcıa, D. E.Coronado Molina, G. Portillo Clark, M. R. Marques, F. J. Magallon Barajas,and Hernandez Lopez. 2012. Water temperature influences viral load anddetection of white spot syndrome virus (WSSV) in Litopenaeus vannameiand wild crustaceans. Aquaculture 326:9–14.

Olsen, Y. 2007. Live food technology of cold-water marine fish larvae. Pages73–128 in E. Moksness, E. Kjørsvik, and Y. Olsen, editors. Culture of cold-water marine fish. Blackwell Scientific Publications, Oxford, UK [ebook].DOI: 10.1002/9780470995617.ch4

Reymond, H., and J. P. Lagardere. 1990. Feeding rhythms and food of Penaeusjaponicus Bate (Crustacea, penaeidae) in salt marsh ponds; role of halophilicentomofauna. Aquaculture 84:125–143.

Salas, D. R. 2011 Abundancia estacional y composicion isotopica de C y Ndel zooplancton relacionados a efluentes de granjas camaronıcolas, en Bahıade Kino Sonora, Mexico. [Seasonal abundance and composition of isotopicC and N of zooplankton related to effluents from shrimp farms, in Bahiade Kino, Sonora, Mexico.] Bachelor’s thesis. Universidad de Sonora, Her-mosillo, Sonora, Mexico.

Sanchez-Paz, A. 2010. White spot syndrome virus: an overview on an emergentconcern. Veterinay Research 41:43.

Walker, P. J., and J. R. Winton. 2010 Emerging viral diseases of fish and shrimp.Veterinary Research 41:51.

Zhang, J. S., S. L. Dong, Y. W. Dong, X. L. Tian, Y. C. Cao, Z. J. Li, and D.C. Yan. 2010. Assessment of the role of brine shrimp Artemia in white spotsyndrome virus (WSSV) transmission. Veterinary Research Communications34:25–32.

Zhang, J. S., S. L. Dong, Y. W. Dong, X. L. Tian, and C. Q. Hou. 2008. Bioassayevidence for the transmission of WSSV by the harpacticoid copepod Nitocrasp. Journal of Invertebrate Pathology 97:33–39.

Zhang, J. S., S. L. Dong, X. L. Tian, Y. W. Dong, X. Y. Liu, and D. C. Yan.2006. Studies on the rotifer (Brachionus urceus Linnaeus, 1758) as a vec-tor in white spot syndrome virus (WSSV) transmission. Aquaculture 261:1181–1185.

Zhang, X., C. Huang, X. Xu, and C. L. Hew. 2002. Identification and localizationof a prawn white spot syndrome virus gene that encodes an envelope protein.Journal of General Virology 83:1069–1074.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4



Susceptibility of Fish and Turtles to Three RanavirusesIsolated from Different Ectothermic Vertebrate ClassesRoberto Brenesa, Debra L. Millerbc, Thomas. B. Waltzekd, Rebecca P. Wilkesc, Jennifer L.Tuckere, Jordan C. Chaneyb, Rebecca H. Hardmanc, Mabre D. Brandc, Rebecca R. Huetherc &Matthew J. Grayb

a Department of Biology, Carroll University, 100 North East Avenue, Waukesha, Wisconsin53186, USAb Center for Wildlife Health, University of Tennessee, 274 Ellington Plant Sciences Building,Knoxville, Tennessee 37996, USAc College of Veterinary Medicine, University of Tennessee, 2407 River Drive, Knoxville,Tennessee 37996, USAd College of Veterinary Medicine, University of Florida, Building 1379, Mowry Road,Gainesville, Florida 32610, USAe Department of Biological Sciences, Humboldt State University, 1 Harpst Street, Arcata,California 95521, USAPublished online: 04 Jun 2014.

To cite this article: Roberto Brenes, Debra L. Miller, Thomas. B. Waltzek, Rebecca P. Wilkes, Jennifer L. Tucker, Jordan C.Chaney, Rebecca H. Hardman, Mabre D. Brand, Rebecca R. Huether & Matthew J. Gray (2014) Susceptibility of Fish andTurtles to Three Ranaviruses Isolated from Different Ectothermic Vertebrate Classes, Journal of Aquatic Animal Health, 26:2,118-126, DOI: 10.1080/08997659.2014.886637







http://dx.doi.org/10.1080/08997659.2014.886637




ARTICLE

Susceptibility of Fish and Turtles to Three RanavirusesIsolated from Different Ectothermic Vertebrate Classes

Roberto Brenes*Department of Biology, Carroll University, 100 North East Avenue, Waukesha, Wisconsin 53186, USA

Debra L. MillerCenter for Wildlife Health, University of Tennessee, 274 Ellington Plant Sciences Building, Knoxville,Tennessee 37996, USA; College of Veterinary Medicine, University of Tennessee, 2407 River Drive,Knoxville, Tennessee 37996, USA

Thomas. B. WaltzekCollege of Veterinary Medicine, University of Florida, Building 1379, Mowry Road, Gainesville,Florida 32610, USA

Rebecca P. WilkesCollege of Veterinary Medicine, University of Tennessee, 2407 River Drive, Knoxville,Tennessee 37996, USA

Jennifer L. TuckerDepartment of Biological Sciences, Humboldt State University, 1 Harpst Street, Arcata,California 95521, USA

Jordan C. ChaneyCenter for Wildlife Health, University of Tennessee, 274 Ellington Plant Sciences Building, Knoxville,Tennessee 37996, USA

Rebecca H. Hardman, Mabre D. Brand, and Rebecca R. HuetherCollege of Veterinary Medicine, University of Tennessee, 2407 River Drive, Knoxville,Tennessee 37996, USA

Matthew J. GrayCenter for Wildlife Health, University of Tennessee, 274 Ellington Plant Sciences Building, Knoxville,Tennessee 37996, USA

AbstractRanaviruses have been associated with mortality of lower vertebrates around the world. Frog virus 3 (FV3)-like

ranaviruses have been isolated from different ectothermic vertebrate classes; however, few studies have demonstratedwhether this pathogen can be transmitted among classes. Using FV3-like ranaviruses isolated from the Americanbullfrog Lithobates catesbeianus, eastern box turtle Terrapene carolina carolina, and Pallid Sturgeon Scaphirhynchus

*Corresponding author: [email protected] October 22, 2013; accepted January 8, 2014

118

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

SUSCEPTIBILITY OF FISH AND TURTLES TO RANAVIRUS 119

albus, we tested for the occurrence of interclass transmission (i.e., infection) and host susceptibility (i.e., percentmortality) for five juvenile fish and three juvenile turtle species exposed to each of these isolates. Exposure wasadministered via water bath (103 PFU/mL) for 3 d and survival was monitored for 28 d. Florida softshell turtlesApalone ferox experienced no mortality, but 10% and 20% of individuals became infected by the turtle and fishisolate, respectively. Similarly, 5% of Mississippi map turtles Graptemys pseudogeographica kohni were subclinicallyinfected with the turtle isolate at the end of the experiment. Channel Catfish Ictalurus punctatus experienced 5%mortality when exposed to the turtle isolate, while Western Mosquitofish Gambusia affinis experienced 10% mortalitywhen exposed to the turtle and amphibian isolates and 5% mortality when exposed to the fish isolate. Our resultsdemonstrated that interclass transmission of FV3-like ranaviruses is possible. Although substantial mortality did notoccur in our experiments, the occurrence of low mortality and subclinical infections suggest that fish and aquaticturtles may function as reservoirs for FV3-like ranaviruses. Additionally, our study is the first to report transmissionof FV3-like ranaviruses between fish and chelonians.

Transmission of viruses among vertebrate classes (hereafterreferred to as interclass transmission) is uncommon. Viral infec-tion is a complex process that involves several steps and exploitsa variety of cellular activities (Su et al. 2008; Cronin et al. 2010;Jackson et al. 2010; Paull et al. 2012). The first and perhapsquintessential challenge a virus has to overcome after enteringa new host is its replication. Once inside the new cell, a virushas to uncoat, transport its genetic material to the appropriatecellular compartment, gather all the necessary replication ma-chinery, produce copies of its genome and virion components,and package the genome into the capsids (Webby et al. 2004;Acheson 2007). If a virus successfully replicates in the new hostcell, there are other obstacles that limit it from infecting its newhost. The virus must exit the cell (i.e., exocytosis or lysis ofthe cell), overcome or avoid the host’s immunological response,infect other cells quickly, and be shed from the host so transmis-sion to other hosts can occur (Webby et al. 2004; Bandin andDopazo 2011; Crispe et al. 2011; Starick et al. 2011).

This complex process of host establishment makes interclasstransmission unlikely in most cases. However, several viruseshave found ways to overcome these obstacles, and examplesof viruses transmitting between species have been recorded(Webby and Kalmakoff 1998; Keesing et al. 2010; Boelle et al.2011; Swayne 2011). For example, some large double-strandedDNA (dsDNA) viruses in the family Iridoviridae are known toinfect multiple amphibian species (Hoverman et al. 2011). Iri-doviruses enter the cell carrying start-up proteins that are usedto initiate genome replication and protein production, therebyfacilitating virus replication in the host cell (Chinchar 2002;Chinchar et al. 2011). The highly conserved major capsid pro-tein of the virus and widely distributed cell receptors targetedby the pathogen likely contribute to the wide host range of iri-doviruses. Currently, five genera within the family Iridoviridaeare recognized (King et al. 2012): two genera, Iridovirus andChloriridovirus, infect arthropods (Camazine and Liu 1998;Hunter et al. 2001; Marina et al. 2003; Gregory et al. 2006),two genera, Lymphocystivirus and Megalocytivirus, infect fish(Sudthongkong et al. 2002; Palmer et al. 2012; Rimmer et al.2012; Waltzek et al. 2012), and one genus, Ranavirus, has been

isolated from amphibians, fish, and reptiles (Chinchar et al.2009; Cinkova et al. 2010; Vesely et al. 2011; Nazir et al. 2012;Robert and Chinchar 2012).

Ranaviruses have been associated with disease and mortal-ity in numerous lower vertebrate species, including amphibians,fishes, and reptiles and are considered a pathogen of ecologicaland economic importance (Chinchar 2002; Keesing et al. 2010;Robert and Chinchar 2012; Gray and Miller 2013). Currently,the International Committee on Taxonomy of Viruses recog-nizes six species of ranaviruses (Jancovich et al. 2010; Kinget al. 2012). Three of the species infect fish exclusively: the epi-zootic hematopoietic necrosis virus, European catfish virus, andSantee–Cooper ranavirus (Bigarre et al. 2008; Chinchar et al.2009; Whittington et al. 2010; Bang-Jensen et al. 2011a; Veselyet al. 2011). The other species—Frog Virus 3 (FV3), Ambystomatigrinum virus (ATV), and Bohle iridovirus (BIV)—have beenisolated most frequently from amphibian hosts, but might infectand cause disease in other ectothermic vertebrates. For exam-ple, ATV is known to cause high mortality in tiger salamandersAmbystoma tigrinum (Jancovich et al. 2003; Collins et al. 2004)and has been reported to cause infection in the LargemouthBass Micropterus salmoides (Picco et al. 2010). Also, BIV wasoriginally isolated from an amphibian (Speare and Smith 1992;Cullen et al. 1995; Cullen and Owens 2002; Weir et al. 2012), butcan infect fish and turtles (Moody and Owens 1994; La Fauceet al. 2012). Recently, transmission of FV3-like ranaviruses wasdemonstrated in fish (Bang-Jensen et al. 2009, 2011a; Bayleyet al. 2013), chelonians (Allender et al. 2006, 2013; Johnsonet al. 2010), and multiple amphibian species (Hoverman et al.2011).

Despite these findings, the host range of FV3-like ranavirusesremains unclear, especially with North American fish and chelo-nian species (Gray et al. 2009). Also, the possibility of interclasstransmission of FV3-like ranaviruses has not been investigatedextensively (Bayley et al. 2013). Our objective was to deter-mine whether three FV3-like ranaviruses isolated from hostsof three different ectothermic classes—amphibians (Amphibia),reptiles (Reptilia), and bony fishes (Osteichthyes)—were able tocause infection and mortality in fish and turtle species known to

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

120 BRENES ET AL.

coexist with amphibians or that are important to the aquacultureindustry in North America. If interclass transmission is possi-ble, fish and turtles may be important reservoirs of FV3-likeranaviruses (Gray et al. 2009), particularly in habitats whereamphibians are not present yearlong.

METHODSRanaviruses and hosts.—The FV3-like ranaviruses used

in our study were isolated from a morbid Pallid SturgeonScaphirhynchus albus in Missouri (T. B. Waltzek, unpublisheddata), eastern box turtle Terrapene carolina carolina in Ken-tucky (Ruder et al. 2010), and American bullfrog Lithobatescatesbeianus in Georgia (Miller et al. 2007). We tested fivefish species: Nile Tilapia Oreochromis niloticus, ChannelCatfish Ictalurus punctatus, Western Mosquitofish Gambusiaaffinis, Bluegill Lepomis macrochirus, and Fathead MinnowPimephales promelas. All fish species were fingerlings (about5–10 cm in length) and were obtained from commercialhatcheries (Table 1). Fish were reared from fry in independent,outdoor, concrete troughs with constant water flow and had nocontact with other species. Upon purchase and arrival at theUniversity of Tennessee, a random sample of five individualswas humanely euthanized by immersion in a solution ofbenzocaine hydrochloride (100 mg/L: Iwama and Ackerman1994) and tested for ranavirus infection using quantitativereal-time PCR (qPCR; see methods below); all qPCR resultswere negative. Prior to the start of the experiments, fishes wereacclimated in the laboratory for 1 week in separate 1,200-Ltanks with flow-through, dechlorinated water (75.7 L/s) at 26◦Cwith 12 h light : 12 h dark photoperiod. During the acclimationperiod, fish were fed a commercial high protein fish food(TetraMin, Blacksburg, Virginia) daily ad libitum.

We tested three aquatic turtle species: Florida softshell turtleApalone ferox, eastern river cooter Pseudemys concinna, andMississippi map turtle Graptemys pseudogeographica kohni.Turtles were purchased as 15-d-old hatchlings (approximately5 cm in length) from commercial retailers (Table 1). All specieswere raised in captivity and in isolation from other species priorto shipment to the University of Tennessee. Turtles were housed

under identical conditions as were the fish, except floating plat-forms were added to the 1,200-L tanks and specialized lampswere provided for thermal and ultraviolet (UV) light exposure(Zoo Med Powersun UV Self-Ballasted Mercury Vapor UVBLamp, San Luis Obispo, California). A random sample of fiveindividuals per species was collected and euthanized to verifyindividuals were not infected with ranavirus prior to exper-imentation; all individuals tested negative by qPCR. Turtleswere fed live crickets and bloodworms once daily ad libitum.

Fish challenges.—Each experimental trial consisted of fourtreatments with 20 replicate fish per treatment, totaling 80 ex-perimental units. The treatments were three ranavirus isolatesand a negative control. Eighty fish were randomly selected fromthe 1,200-L tank and placed individually into 4-L (17.7 × 17.7× 28.5 cm) tubs filled with 2 L of dechlorinated, aged tap water;the tubs were placed on 122 × 244-cm shelving units. Prior toadding the fish, each container was randomly assigned to a viralor control treatment in a randomized block design, in whichtwo shelf heights were the blocking variables. Viral treatmentswere inoculated with 103 PFU/mL of the appropriate virus iso-late, and the controls were inoculated with the same quantity ofvirus-free media (i.e., MEM Eagle, Sigma-Aldrich, Seelze, Ger-many). We used 103 PFU/mL because it has been suggested thatthis concentration is ecologically relevant (Gray et al. 2009).Rojas et al. (2005) reported this titer of ranavirus in water shedby an infected salamander. Dose-dependent studies (e.g., Brun-ner et al. 2005) show that mortality is low typically when am-phibian larvae are exposed to <100 PFU/mL of ranavirus. Lessinformation is available on the dose-dependent relationships ofranavirus and fish hosts, but the viral titer we used is known tocause ranaviral disease in fish (Moody and Owens 1994; Grizzleet al. 2002; Gobbo et al. 2010). Individuals were exposed to thevirus in stagnant water for 3 d, which has become standard forranavirus challenges with amphibians due to no apparent effectof exposure duration on pathogenicity of ranavirus (Hovermanet al. 2010, 2011). Given that fish were negative for ranavirus atthe beginning of each experiment, the inoculations likely repre-sented a first-time exposure to the pathogen, which is standardin ranavirus-challenge experiments (Bang-Jensen et al. 2011a;Jaramillo et al. 2012).

TABLE 1. Vendors for specimens during the challenge experiments.

Species Vendor

Nile Tilapia Oreochromis niloticus Greenwater Fish Farm, Milan, TennesseeChannel Catfish Ictalurus punctatus Greenwater Fish Farm, Milan, TennesseeWestern Mosquitofish Gambusia affinis Alabama Aquarium and Pond Services, Birmingham, AlabamaBluegill Lepomis macrochirus Bell Springs Fish Hatchery, Riceville, TennesseeFathead Minnow Pimephales promelas Bell Springs Fish Hatchery, Riceville, TennesseeFlorida softshell turtle Apalone ferox JP Pets, Sanford, FloridaEastern river cooter Pseudemys concinna JP Pets, Sanford, FloridaMississippi map turtle Graptemys pseudogeographica kohni Backwater Reptiles, Sacramento, California

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4


During experiments, fish were fed high-protein commercialfood every day at a ratio of 3% of body mass, which is suffi-cient for normal growth and development (Budy et al. 2011).The amount of food required was calculated based on the bodymass of a separate sample of five nonexperimental fish that weretreated in a manner identical to the controls. Fish were monitoredtwice daily for survival and morbidity. Dead individuals wereremoved from their containers, necropsied, and any gross signsof ranaviral infection recorded. Fish that exhibited morbidityconsistent with ranaviral disease (i.e., petechial hemorrhages,edema, and loss of equilibrium) for >24 h during the experi-ment were humanely euthanized by immersion in benzocainehydrochloride solution (100 mg/L). Water was changed (100%of volume) every 3 d to maintain water quality during the exper-iment (Hoverman et al. 2010), and laboratory temperature wasmonitored and maintained at 26◦C. Duration for all trials was 28d, which is sufficient duration for morbidity to be observed fromranavirus infection (Bang-Jensen et al. 2009, 2011a; Jaramilloet al. 2012). At the end of each experiment, all surviving indi-viduals were humanely euthanized by immersion in 100 mg/Lof benzocaine hydrochloride (Iwama and Ackerman 1994).

Turtle challenges.—Turtle experiments followed the sameprocedures as the fish challenges with three exceptions. First,the turtles were housed in 15.5-L containers (41.6 × 28.6 ×18.7 cm) containing 2 L of dechlorinated, aged tap water (ap-proximately 3 cm depth). This amount of water was sufficient toallow the turtle to fully immerse its body while maintaining itshead above water. Second, during the experiments, turtles werefed two live crickets per day, which was sufficient for normalgrowth and development (Teece et al. 2001). Lastly, individualsthat exhibited gross signs of ranaviral disease (e.g., cutaneousabscessation, oral ulceration or abscessation, respiratory dis-tress, anorexia, and lethargy: Allender et al. 2006; Johnson et al.2006) and survivors at the end of the experiment were humanelyeuthanized via intravenous injection of 60–100 mg/kg of sodiumpentobarbital. All procedures followed approved University ofTennessee Institutional Animal Care and Use Committee proto-col 2052.

Ranavirus testing.—Genomic DNA (gDNA) was extractedfrom a tissue homogenate of the kidney and liver collectedduring necropsy using the DNeasy Blood and Tissue Kit(Qiagen, Valencia, California).We used a Qubit fluorometer andthe Quant-iT dsDNA BR Assay Kit to quantify the concentra-tion of gDNA in each sample (Invitrogen, Carlsbad, California).Quantitative real-time PCR (qPCR) was used to amplify a 70-bpsequence of the ranavirus major capsid protein using primers andprotocol identical to Picco et al. (2007). The extracted DNA sam-ples were run in duplicate, and an individual was declared posi-tive if the qPCR cycle threshold (CT) was <30 for both samples.This CT decision rule was determined for our PCR system (ABI7900 Fast Real-Time PCR System; Life Technologies Corpora-tion, Carlsbad, California) by developing a standard curve with95% CIs using known quantities of ranavirus. For qPCR-positiveindividuals, we reported the predicted PFU per 0.25 µg of host

gDNA as a relative indicator of viral load. Four controls wereincluded in each qPCR assay: DNA extracted from a ranavirus-positive animal, DNA extracted from a ranavirus-negative ani-mal, DNA extracted from cultured ranavirus, and water.

Statistical analyses.—We summarized the results as individ-uals that died and were infected (case mortality), survived andwere infected (subclinical infection), and died but were not in-fected (natural mortality). For our study, we defined infectionas qPCR positive according to our CT decision rule, which iscommon in transmission studies (e.g., Brunner et al. 2005; Hov-erman et al. 2011). Given that quiescent infections are possiblewith ranaviruses (Robert et al. 2011) and active replication is un-necessary for qPCR to amplify viral DNA (Green et al. 2009),it is possible that our qPCR positive results did not representactive infections. Nonetheless, detection of ranavirus DNA viaqPCR is evidence of transmission in our study considering thatprescreening resulted in no positive results, and our experimentwas designed based on independent water bath challenges. Foreach species, we tested for the difference in case mortality andinfection prevalence (i.e., qPCR positive) among the ranavirusisolates using a G-test of maximum likelihood (Sokal and Rohlf1995). All analyses were performed using SAS 9.3 (SAS 2012)at α = 0.05.

RESULTSTwo fish species experienced case mortality: Channel Catfish

and Western Mosquitofish (Figure 1). The catfish experienced5% mortality when exposed to the fish isolate, while themosquitofish experienced 10, 10, and 5% mortality whenexposed to the turtle, amphibian, and fish isolates, respectively.Average viral load for infected fish tissue (0.25 µg) was 8.9PFU (Table 2). No statistical differences were detected in casemortality (G = 5.71, df = 12, P = 0.28) or infection prevalence(G = 18.94, df = 12, P = 0.13) among the three isolates. Catfishdied between 16 and 24 d postexposure, while mosquitofishbegan to die after 4 d postexposure to the virus (Figure 3).

No deaths were documented in turtles exposed to ranavirus;however, infection occurred in two species (Figure 2). Afterexposure to the turtle and fish isolates, 10% and 20% of Floridasoftshell turtles, respectively, were infected. The Mississippimap turtle experienced 5% infection when exposed to thebox turtle isolate. Average viral load for infected turtle tissue(0.25 µg) was 228 PFU, and was greatest for Florida softshellturtles exposed to the box turtle isolate (Table 2). No statisticaldifferences were detected in infection prevalence (G = 7.32,df = 12, P = 0.19) among the three isolates.

DISCUSSIONOur study documented two new cases of interclass trans-

mission: (1) transmission of a FV3-like ranavirus isolated froma fish to a turtle species, and (2) transmission of a FV3-likeranavirus isolated from a turtle to a fish species. We also doc-umented transmission of a FV3-like ranavirus isolated from an

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

122 BRENES ET AL.

0102030405060708090

100

Indi

vidu

als (

%)

Died not infectedSurvived, infectedDied infected

0102030405060708090

100

Indi

vidu

als (

%)

0102030405060708090

100

Indi

vidu

als (

%)

0102030405060708090

100

Indi

vidu

als (

%)

0102030405060708090

100

Control Turtle Fish Amphibian

Indi

vidu

als(

%)

Isolate

Channel Cat�ish

Nile Tilapia

Western Mosquito�ish

Fathead Minnow

Bluegill

FIGURE 1. Percent mortality and infection of five fish species (Channel Cat-fish, Nile Tilapia, Western Mosquitofish, Fathead Minnow, and Bluegill) ex-posed to three ranavirus isolates from different ectothermic vertebrate hosts:turtle, fish, and amphibian. Results are based on exposure of 20 individuals perfish species per ranavirus isolate for 28 d. Infection was determined via qPCRand may not represent occurrence of active virus replication in the host.

amphibian to a fish species, which has been reported by oth-ers (e.g., Bang-Jensen et al. 2009, 2011b; Gobbo et al. 2010;Picco et al. 2010). These results provide additional evidencethat FV3-like ranaviruses can be transmitted among ectother-mic vertebrate classes.

We documented 5% mortality of Channel Catfish exposed tothe turtle isolate, and 5–10% mortality of Western Mosquitofishexposed to fish, turtle, or amphibian isolates. Although thislevel of mortality is low, these results suggest that ranaviruscould negatively impact aquaculture industries (Prasankok et al.

TABLE 2. Viral load (PFU) in a homogenate of liver and kidney tissue(0.25 µg) of infected individuals exposed to three FV3-like ranavirus isolatesfrom a morbid turtle (eastern box turtle), a fish (Pallid Sturgeon), and an am-phibian (American bullfrog).

Species Isolate PFU

Western Mosquitofish Turtle 11.7Fish 10.2Amphibian 12.1

5.2Channel Catfish Turtle 5.2Florida softshell turtle Turtle 760

606Fish 1.6

1.11.40.8

Mississippi map turtle Turtle 2.6

2002; Bang-Jensen et al. 2011b; Vesely et al. 2011). Bang-Jensen et al. (2011b) reported that ranaviruses were a concern tothe aquaculture industry in the European Union, and the occur-rence of subclinically infected individuals in international fishtrade could result in the emergence of ranavirus. Production ofChannel Catfish and Western Mosquitofish are major industriesin the United States (Mischke et al. 2013; Torrans et al. 2013).Additionally, mosquitofish are commonly released as biologi-cal control agents into natural aquatic systems containing nativepopulations of ectothermic vertebrates (Griffin and Knight 2012;Samidurai and Mathew 2013). The fact that mosquitofish canbe subclinically infected with FV3-like ranaviruses is a conser-vation concern.

The species of ranaviruses that are found exclusively in fishhosts (i.e., epizootic hematopoietic necrosis virus, Europeancatfish virus, and Santee–Cooper ranavirus) are known to causesignificant morbidity and mortality in several fish species aroundthe world (Bigarre et al. 2008; Picco et al. 2010; Whittingtonet al. 2010; Bang-Jensen et al. 2011b; Vesely et al. 2011). Theranavirus BIV can cause significant mortality in BarramundiLates calcarifer (Moody and Owens 1994). However, FV3-likeand ATV ranaviruses appear to cause subclinical infections andlow mortality in fish (Bang-Jensen et al. 2009; 2011a; Gobboet al. 2010; Picco et al. 2010). The reduced susceptibility offish to ATV and FV3-like ranaviruses could be a result of hostspecificity for cell entry and replication, or an inability to bypassthe fully functional immune system of fish (Grayfer et al. 2012).

The low susceptibility of the turtles that we tested to ranaviruswas unexpected, as cases of ranavirus infection and diseasehave been reported in at least 11 tortoise and box turtle species(Marschang et al. 1999; De Voe et al. 2004; Benetka et al. 2007;Johnson et al. 2007, 2010; Marschang 2011), red-eared sliderturtle Trachemys scripta elegans (Johnson et al. 2006, 2010;Allender et al. 2013) and Chinese softshell turtle Pelodiscus

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4


0102030405060708090

100

Indi

vidu

als (

%)

Survived, infected

Died, not infected

Died, infected

0102030405060708090

100

Indi

vidu

als (

%)

0102030405060708090

100

Control Turtle Fish Amphibian

Indi

vidu

als

(%)

Isolate

Eastern river cooter

Florida softshell turtle

Mississippi map turtle

FIGURE 2. Percent mortality and infection of three turtle species (easternriver cooter, Florida softshell turtle, and Mississippi map turtle) exposed to threeranavirus isolates from different ectothermic hosts: turtle, fish, and amphibian.Results are based on exposure of 20 individuals per turtle species per ranavirusisolate for 28 d. Infection was determined via qPCR and may not representoccurrence of active virus replication in the host.

sinensis (Chen et al. 1999) in both natural and laboratory en-vironments (Chen et al. 1999; De Voe et al. 2004; Allenderet al. 2006; Johnson et al. 2008). However, most of these re-ports were diagnostic cases on a single individual or challengeexperiments via isolate injection, which may be an unrealistictransmission route (Gray et al. 2009). Allender et al. (2013)reported greater susceptibility of adult red-eared slider turtlesinjected with ranavirus at 21◦C compared with 28◦C. Giventhat our experiment was performed at 26◦C, the lower infec-tion we observed could have been influenced by temperature.More information is needed on the susceptibility of cheloniansto ranavirus, and the role of temperature.

Our susceptibility results likely reflect a best-case scenarioinasmuch as our experiments were conducted under controlledconditions with food provided ad libitum. Additionally, fac-tors that contribute to ranavirus emergence such as density-dependent transmission were controlled. In wild or captive

0102030405060708090

100

1 4 7 10 13 16 19 22 25 28

Surv

ival

(%) Control

TurtleFishAmphibian

0102030405060708090

100

1 4 7 10 13 16 19 22 25 28

Surv

ival

(%)

Western Mosquito�ish

Channel Cat�ish

FIGURE 3. Survival curves of fish species (Channel Catfish and West-ern Mosquitofish) that experienced mortality when exposed to ranavirus iso-lates from three different ectothermic vertebrate classes (i.e., turtle, fish, andamphibian).

populations, multiple infected and morbid individuals can bepresent, which might increase the likelihood of transmissionto other ectothermic vertebrates, particularly those that predate(e.g., fish) or scavenge (e.g., turtles) other hosts.

The majority of individuals in our study tested negative forranavirus DNA in liver and kidney tissue 28 d following expo-sure to an isolate. It is possible that individuals became infectedand cleared the virus prior to the end of the experiment. For ex-ample, Fathead Minnow cells have been used to replicate FV3in the laboratory for many decades (Green et al. 2009), yet noindividuals of this species were positive after 28 d in our study.Short-duration infection could play a role in the epidemiologyof ranaviruses, especially where host densities are high. Futuretransmission studies should consider euthanizing individuals atdifferent postexposure durations to document host susceptibil-ity and improve our understanding of short- versus long-termreservoirs.

Our results indicate that fish and aquatic turtles could func-tion as reservoirs for FV3-like ranaviruses and, through com-mercial trade, contribute to pathogen pollution (Cunninghamet al. 2003). In the United States, 662 million tons of catfish(Hanson 2012) were produced in 2012, and 31.8 million turtlesincluding 17.5 million individual red-eared slider turtles weresold between 2004 and 2005 (Brown et al. 2011; WCT 2013).Our results suggest that fish and turtles infected with ranavirusshould be included in the World Organization for Animal Health(OIE) standards for notifiable diseases (Schloegel et al. 2010).Currently, amphibians infected with ranaviruses are the only

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

124 BRENES ET AL.

taxonomic group listed in the OIE regulations (Schloegel et al.2010).

Although our results showed that some fishes (Channel Cat-fish and Western Mosquitofish) and turtles (Florida softshelland Mississippi map turtles) are suitable hosts for FV3-likeranaviruses, additional research is needed on other species inNorth America. Additionally, experiments are needed to deter-mine whether an infected individual of one vertebrate class cantransmit ranavirus through water to a different class. The ca-pacity of fish and turtle species to transmit ranavirus to highlysusceptible hosts that inhabit aquatic environments seasonally(e.g., amphibians) will help us understand the reoccurrences ofoutbreaks in ecosystems with fluctuating species composition(Pearman and Garner 2005; Teacher et al. 2010). This infor-mation could be essential for the planning and execution ofconservation strategies for areas that exhibit recurrent ranavirusoutbreaks, as well as the identification of areas with risk ofranaviral disease.

ACKNOWLEDGMENTSFunding for this research was provided by The University of

Tennessee (UT) Institute of Agriculture, UT College of Agri-cultural Sciences and Natural Resources via Hazelwood andthe UT-ESPN Scholarships, and the Society of Wetland Scien-tists, Student Research Award. We thank Bobby Simpson andRoger Long of the UT East Tennessee Research and EducationCenter for providing laboratory space. Sean Roon and LaurenHenderson provided technician support for the experiments.

REFERENCESAcheson, N. H. 2007. Fundamentals of molecular virology. Wiley, Danvers,

Massachusetts.Allender, M. C., M. M. Fry, A. R. Irizarry, L. Craig, A. J. Johnson, and M. Jones.

2006. Intracytoplasmic inclusions in circulating leukocytes from an easternbox turtle (Terrapene carolina carolina) with iridoviral infection. Journal ofWildlife Diseases 42:677–684.

Allender, M. C., M. A. Mitchell, T. Torres, J. Sekowska, and E. A. Driskell. 2013.Pathogenicity of frog virus 3-like virus in red-eared slider turtles (Trachemysscripta elegans) at two environmental temperatures. Journal of ComparativePathology 149:356–367.

Bandin, I., and C. P. Dopazo. 2011. Host range, host specificity and hypothesizedhost shift events among viruses of lower vertebrates. Veterinary Research42:67.

Bang-Jensen, B., A. K. Ersboll, and E. Ariel. 2009. Susceptibility of pike Esoxlucius to a panel of ranavirus isolates. Diseases of Aquatic Organisms 83:169–179.

Bang-Jensen, B., R. Holopainen, H. Tapiovaara, and E. Ariel. 2011a. Sus-ceptibility of pike-perch Sander lucioperca to a panel of ranavirus isolates.Aquaculture 313:24–30.

Bang-Jensen, B., S. Reschova, K. Cinkova, E. Ariel, and T. Vesely. 2011b.Common Carp (Cyprinus carpio) and Goldfish (Carassius auratus) were notsusceptible to challenge with ranavirus under certain challenge conditions.Bulletin of the European Association of Fish Pathology 31:112–118.

Bayley, A. E., B. J. Hill, and S. W. Feist. 2013. Susceptibility of the Europeancommon frog Rana temporaria to a panel of ranavirus isolates from fish andamphibian hosts. Diseases of Aquatic Organisms 103:171–183.

Benetka, V., E. Grabensteiner, M. Gumpenberger, C. Neubauer, B. Hirschmuller,and K. Mostl. 2007. First report of an iridovirus (Genus Ranavirus) infectionin a Leopard tortoise (Geochelone pardalis pardalis). Wiener TierarztlicheMonatsschrift 94:243–248.

Bigarre, L., J. Cabon, M. Baud, F. Pozet, and J. Castric. 2008. Ranavirusesassociated with high mortalities in catfish in France. Bulletin of the EuropeanAssociation of Fish Pathologists 28:163–168.

Boelle, P.-Y., S. Ansart, A. Cori, and A.-J. Valleron. 2011. Transmission param-eters of the A/H1N1 (2009) influenza virus pandemic: a review. Influenza andOther Respiratory Viruses 5:306–316.

Brown, D. J., V. R. Farallo, J. R. Dixon, J. T. Baccus, T. R. Simpson, and M. R.J. Forstner. 2011. Freshwater turtle conservation in Texas: harvest effects andefficacy of the current management regime. Journal of Wildlife Management75:486–494.

Brunner, J. L., K. Richards, and J. P. Collins. 2005. Dose and host characteristicsinfluence virulence of ranavirus infections. Oecologia 144:399–406.

Budy, P., M. Baker, and S. K. Dahle. 2011. Predicting fish growth potentialand identifying water quality constraints: a spatially-explicit bioenergeticsapproach. Environmental Management 48:691–709.

Camazine, S., and T. P. Liu. 1998. A putative iridovirus from the honey bee miteVarroa jacobsoni Oudemans. Journal of Invertebrate Pathology 71:177–178.

Chen, Z. X., J. C. Zheng, and Y. L. Jiang. 1999. A new iridovirus isolated fromsoft-shelled turtle. Virus Research 63:147–151.

Chinchar, V. G. 2002. Ranaviruses (family Iridoviridae): emerging cold-bloodedkillers—brief review. Archives of Virology 147:447–470.

Chinchar, V. G., A. Hyatt, T. Miyazaki, and T. Williams. 2009. Family Iri-doviridae: poor viral relations no longer. Current Topics in Microbiology andImmunology 328:123–170.

Chinchar, V. G., K. H. Yu, and J. K. Jancovich. 2011. The molecular biology offrog virus 3 and other iridoviruses infecting cold-blooded vertebrates. Viruses3:1959–1985.

Cinkova, K., S. Reschova, P. Kulich, and T. Vesely. 2010. Evaluation of apolyclonal antibody for the detection and identification of ranaviruses fromfreshwater fish and amphibians. Diseases of Aquatic Organisms 89:191–198.

Collins, J. P., J. L. Brunner, J. K. Jancovich, and D. M. Schock. 2004. A modelhost-pathogen system for studying infectious disease dynamics in amphib-ians: tiger salamanders (Ambystoma tigrinum) and Ambystoma tigrinum virus.Herpetological Journal 14:195–200.

Crispe, E., D. S. Finlaison, A. C. Hurt, and P. D. Kirkland. 2011. Infectionof dogs with equine influenza virus: evidence for transmission from horsesduring the Australian outbreak. Australian Veterinary Journal 89:27–28.

Cronin, J. P., M. E. Welsh, M. G. Dekkers, S. T. Abercrombie, and C. E.Mitchell. 2010. Host physiological phenotype explains pathogen reservoirpotential. Ecology Letters 13:1221–1232.

Cullen, B. R., and L. Owens. 2002. Experimental challenge and clinical casesof Bohle iridovirus (BIV) in native Australian anurans. Diseases of AquaticOrganisms 49:83–92.

Cullen, B. R., L. Owens, and R. J. Whittington. 1995. Experimental infection ofAustralian anurans (Limnodynastes terraereginae and Litoria latopalmata)with Bohle iridovirus. Diseases of Aquatic Organisms 23:83–92.

Cunningham, A. A., P. Daszak, and J. P. Rodriguez. 2003. Pathogen pollu-tion: defining a parasitological threat to biodiversity conservation. Journal ofParasitology 89(Supplement):78–83.

De Voe, R., K. Geissler, S. Elmore, D. Rotstein, G. Lewbart, and J. Guy.2004. Ranavirus-associated morbidity and mortality in a group of captiveeastern box turtles (Terrapene carolina carolina). Journal of Zoo and WildlifeMedicine 35:534–543.

Gobbo, F., E. Cappellozza, M. R. Pastore, and G. Bovo. 2010. Susceptibility ofBlack Bullhead Ameiurus melas to a panel of ranavirus isolates. Diseases ofAquatic Organisms 90:167–174.

Gray, M. J., and D. L. Miller. 2013. Rise of ranavirus: an emerging pathogenthreatens ectothermic vertebrates. Wildlife Professional 7:51–55.

Gray, M. J., D. L. Miller, and J. T. Hoverman. 2009. Ecology and pathology ofamphibian ranaviruses. Diseases of Aquatic Organisms 87:243–266.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4


Grayfer, L., F. D. Andino, G. C. Chen, G. V. Chinchar, and J. Robert. 2012.Immune evasion strategies of ranaviruses and innate immune responses tothese emerging pathogens. Viruses 4:1075–1092.

Green, D. E., M. J. Gray, and D. L. Miller. 2009. Disease monitoring andbiosecurity. Pages 481–505 in C. K. Dodd, editor. Amphibian ecology andconservation: a handbook of techniques. Oxford University Press, Oxford,UK.

Gregory, C. R., K. S. Latimer, K. E. Pennick, K. Benson, and T. Moore. 2006.Novel iridovirus in a nautilus (Nautilus spp.). Journal of Veterinary DiagnosticInvestigation 18:208–211.

Griffin, L. F., and J. M. Knight. 2012. A review of the role of fish as biologicalcontrol agents of disease vector mosquitoes in mangrove forests: reducinghuman health risks while reducing environmental risk. Wetlands Ecologyand Management 20:243–252.

Grizzle, J. M., I. Altinok, W. A. Fraser, and R. Francis-Floyd. 2002. Firstisolation of Largemouth Bass virus. Diseases of Aquatic Organisms 50:233–235.

Hanson, T. R. 2012. U.S. farm-raised catfish industry 2010 review and 2011outlook. Auburn University, Auburn, Alabama.

Hoverman, J. T., M. J. Gray, N. A. Haislip, and D. L. Miller. 2011. Phylogeny,life history, and ecology contribute to differences in amphibian susceptibilityto ranaviruses. Ecohealth 8:301–319.

Hoverman, J. T., M. J. Gray, and D. L. Miller. 2010. Anuran susceptibilities toranaviruses: role of species identity, exposure route, and a novel virus isolate.Diseases of Aquatic Organisms 89:97–107.

Hunter, W. B., C. P. Patte, X. H. Sinisterra, D. S. Achor, C. J. Funk, andJ. E. Polston. 2001. Discovering new insect viruses: whitefly iridovirus (Ho-moptera: Aleyrodidae: Bemisia tabaci). Journal of Invertebrate Pathology 78:220–225.

Iwama, G. K., and P. A. Ackerman. 1994. Anaesthetics. Pages 1–17 in P. W.Hochachka and T. P. Mommsen, editors. Biochemistry and molecular biologyof fishes: analytical techniques. Elsevier, Amsterdam.

Jackson, J. O., E. Lin, P. G. Spear, and R. Longnecker. 2010. Insertion mutationsin herpes simplex virus 1 glycoprotein H reduce cell surface expression, slowthe rate of cell fusion, or abrogate functions in cell fusion and viral entry.Journal of Virology 84:2038–2046.

Jancovich, J. K., M. Bremont, J. W. Touchman, and B. L. Jacobs. 2010. Evi-dence for multiple recent host species shifts among the ranaviruses (familyIridoviridae). Journal of Virology 84:2636–2647.

Jancovich, J. K., J. Mao, V. G. Chinchar, C. Wyatt, S. T. Case, S. Kumar,G. Valente, S. Subramanian, E. Davidson, J. P. Collins, and B. L. Jacobs.2003. Genomic sequence of a ranavirus (family Iridoviridae) associated withsalamander mortalities in North America. Virology 316:90–103.

Jaramillo, D., A. Tweedie, J. A. Becker, A. Hyatt, S. Crameri, and R. J. Whitting-ton. 2012. A validated quantitative polymerase chain reaction assay for thedetection of ranaviruses (family Iridoviridae) in fish tissue and cell cultures,using EHNV as a model. Aquaculture 356:186–192.

Johnson, A. J., A. P. Pessier, A. L. Childress, and E. R. Jacobson. 2006. Red-eared sliders (Trachemys scripta elegans) as a model of ranavirus infectionin chelonians. Pages 73–74 in C. K. Baer, editor. Proceedings of the Associa-tion of Reptilian and Amphibian Veterinarians. Association of Reptilian andAmphibian Veterinarians, Lawrence, Kansas.

Johnson, A. J., A. P. Pessier, and E. R. Jacobson. 2007. Experimental trans-mission and induction of ranaviral disease in western ornate box turtles (Ter-rapene ornata ornata) and red-eared sliders (Trachemys scripta elegans).Veterinary Pathology 44:285–297.

Johnson, A. J., A. P. Pessier, J. F. X. Wellehan, A. Childress, T. M. Norton, N.L. Stedman, D. C. Bloom, W. Belzer, V. R. Titus, R. Wagner, J. W. Brooks,J. Spratt, and E. R. Jacobson. 2008. Ranavirus infection of free-ranging andcaptive box turtles and tortoises in the United States. Journal of WildlifeDiseases 44:851–863.

Johnson, A. J., L. Wendland, T. M. Norton, B. Belzer, and E. R. Jacobson.2010. Development and use of an indirect enzyme-linked immunosorbentassay for detection of iridovirus exposure in gopher tortoises (Gopherus

polyphemus) and eastern box turtles (Terrapene carolina carolina). VeterinaryMicrobiology 142:160–167.

Keesing, F., L. K. Belden, P. Daszak, A. Dobson, C. D. Harvell, R. D. Holt, P.Hudson, A. Jolles, K. E. Jones, C. E. Mitchell, S. S. Myers, T. Bogich, andR. S. Ostfeld. 2010. Impacts of biodiversity on the emergence and transmis-sion of infectious diseases. Nature 468:647–652.

King, A. M. Q., E. Lefkowitz, M. J. Adams, and E. B. Carstens, editors. 2012.International commitee on taxonomy of viruses, 9th edition. Elsevier, London.

La Fauce, K., E. Ariel, S. Munns, C. Rush, and L. Owens. 2012. Influenceof temperature and exposure time on the infectivity of Bohle iridovirus, aranavirus. Aquaculture 354:64–67.

Marina, C. F., J. E. Ibarra, J. I. Arredondo-Jimenez, I. Fernandez-Salas, J. Valle,and T. Williams. 2003. Sublethal iridovirus disease of the mosquito Aedesaegypti is due to viral replication not cytotoxicity. Medical and VeterinaryEntomology 17:187–194.

Marschang, R. E. 2011. Viruses infecting reptiles. Viruses 3:2087–2126.Marschang, R. E., P. Becher, H. Posthaus, P. Wild, H. J. Thiel, U. Muller-Doblies,

E. F. Kaleta, and L. N. Bacciarini. 1999. Isolation and characterization of aniridovirus from Hermann’s tortoises (Testudo hermanni). Archives of Virol-ogy 144:1909–1922.

Miller, D. L., S. Rajeev, M. J. Gray, and C. A. Baldwin. 2007. Frog virus3 infection, cultured American bullfrogs. Emerging Infectious Diseases 13:342–343.

Mischke, C. C., M. J. Griffin, T. E. Greenway, and D. Wise. 2013. Effects ofMosquitofish, Gambusia affinis, on Channel Catfish, Ictalurus puncatatus,production ponds. Journal of the World Aquaculture Society 44:288–292.

Moody, N. J. G., and L. Owens. 1994. Experimental demonstration of thepathogenicity of a frog virus, bohle iridovirus, for a fish species, BarramundiLates calcarifer. Diseases of Aquatic Organisms 18:95–102.

Nazir, J., M. Spengler, and R. E. Marschang. 2012. Environmental persistence ofamphibian and reptilian ranaviruses. Diseases of Aquatic Organisms 98:177–184.

Palmer, L. J., N. S. Hogan, and M. R. van den Heuvel. 2012. Phylogeneticanalysis and molecular methods for the detection of lymphocystis diseasevirus from Yellow Perch, Perca flavescens (Mitchell). Journal of Fish Diseases35:661–670.

Paull, S. H., S. J. Song, K. M. McClure, L. C. Sackett, A. M. Kilpatrick, andP. T. J. Johnson. 2012. From superspreaders to disease hotspots: linking trans-mission across hosts and space. Frontiers in Ecology and the Environment10:75–82.

Pearman, P. B., and T. W. J. Garner. 2005. Susceptibility of Italian agile frogpopulations to an emerging strain of ranavirus parallels population geneticdiversity. Ecology Letters 8:401–408.

Picco, A. M., J. L. Brunner, and J. P. Collins. 2007. Susceptibility of the en-dangered California tiger salamander, Ambystoma californiense, to ranavirusinfection. Journal of Wildlife Diseases 43:286–290.

Picco, A. M., A. P. Karam, and J. P. Collins. 2010. Pathogen host switching incommercial trade with management recommendations. Ecohealth 7:252–256.

Prasankok, P., M. Chutmongkonkul, and S. Kanchanakhan. 2002. Characterisa-tion of iridovirus isolated from diseased Marbled Sleepy Goby, Oxyeleotrismarmoratus. Pages 197–206 in P. Walker, R. Lester, and M. G. Bondad-Reantaso, editors. Diseases in Asian aquaculture 5th symposium. Asian Fish-eries Society, Fish Health Section, Manila.

Rimmer, A. E., J. A. Becker, A. Tweedie, and R. J. Whittington. 2012. De-velopment of a quantitative polymerase chain reaction (qPCR) assay for thedetection of dwarf gourami iridovirus (DGIV) and other megalocytivirusesand comparison with the Office International des Epizooties (OIE) referencePCR protocol. Aquaculture 358:155–163.

Robert, J., and V. G. Chinchar. 2012. “Ranaviruses: an emerging threat toectothermic vertebrates” report of the first international symposium onranaviruses, Minneapolis MN July 8, 2011. Developmental and Compara-tive Immunology 36:259–261.

Robert, J., E. George, F. D. Andino, and G. C. Chen. 2011. Waterborne infectivityof the ranavirus frog virus 3 in Xenopus laevis. Virology 417:410–417.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

126 BRENES ET AL.

Rojas, S., K. Richards, J. K. Jancovich, and E. W. Davidson. 2005. Influ-ence of temperature on ranavirus infection in larval salamanders Ambystomatigrinum. Diseases of Aquatic Organisms 63:95–100.

Ruder, M. G., A. B. Allison, D. L. Miller, and M. K. Keel. 2010. Pathology inpractice. Journal of the American Veterinary Medical Association 237:783–785.

Samidurai, K., and N. Mathew. 2013. Mosquito larvicidal and ovicidal activityof puffer fish extracts against Anopheles stephensi, Culex quinquefasciatusand Aedes aegypti (Diptera: Culicidae). Tropical Biomedicine 30:27–35.

SAS. 2012. SAS, version 9.3. SAS Institute, Cary, North Carolina.Schloegel, L. M., P. Daszak, A. A. Cunningham, R. Speare, and B. Hill. 2010.

Two amphibian diseases, chytridiomycosis and ranaviral disease, are nowglobally notifiable to the World Organization for Animal Health (OIE): anassessment. Diseases of Aquatic Organisms 92:101–108.

Sokal, R. R., and F. J. Rohlf. 1995. Biometry, 3rd edition. Freeman, New York.Speare, R., and J. R. Smith. 1992. An iridovirus-like agent isolated from the

ornate burrowing frog Limnodynastes ornatus in northern Australia. Diseasesof Aquatic Organisms 14:51–57.

Starick, E., S. R. Fereidouni, E. Lange, C. Grund, T. Vahlenkamp, M. Beer, andT. C. Harder. 2011. Analysis of influenza A viruses of subtype H1 from wildbirds, turkeys and pigs in Germany reveals interspecies transmission events.Influenza and Other Respiratory Viruses 5:276–284.

Su, Y., H. Y. Yang, B. J. Zhang, H. L. Jia, and P. Tien. 2008. Analysis ofa point mutation in H5N1 avian influenza virus hemagglutinin in relationto virus entry into live mammalian cells. Archives of Virology 153:2253–2261.

Sudthongkong, C., M. Miyata, and T. Miyazaki. 2002. Viral DNA sequencesof genes encoding the ATPase and the major capsid protein of tropical iri-dovirus isolates which are pathogenic to fishes in Japan, South China Sea andSoutheast Asian countries. Archives of Virology 147:2089–2109.

Swayne, D. E. 2011. Mechanisms of interspecies transmission of avian influenzaviruses at the human-animal interface. Influenza and other respiratory viruses5:44–45.

Teacher, A. G. F., A. A. Cunningham, and T. W. J. Garner. 2010. Assessing thelong-term impact of ranavirus infection in wild common frog populations.Animal Conservation 13:514–522.

Teece, M. A., C. W. Swarth, N. Tuross, and M. L. Fogel. 2001. Investigationsof turtle metabolism and hatchling development using stable carbon isotopeanalysis of individual amino acids. Abstracts of Papers American ChemicalSociety 221:U534–U534.

Torrans, L., B. Ott, R. Jones, and R. Jones. 2013. Channel Catfish hatcheryproduction efficiency Using a vertical-lift incubator (the see-saw) at variousegg loading rates. North American Journal of Aquaculture 75:235–243.

Vesely, T., K. Cinkova, S. Reschova, F. Gobbo, E. Ariel, M. Vicenova, D.Pokorova, P. Kulich, and G. Bovo. 2011. Investigation of ornamental fishentering the EU for the presence of ranaviruses. Journal of Fish Diseases34:159–166.

Waltzek, T. B., G. D. Marty, M. E. Alfaro, W. R. Bennett, K. A. Garver, M.Haulena, E. S. Weber, and R. P. Hedrick. 2012. Systemic iridovirus fromThreespine Stickleback Gasterosteus aculeatus represents a new megalocy-tivirus species (family Iridoviridae). Diseases of Aquatic Organisms 98:41–56.

WCT (World Chelonian Trust). 2013. Turtle survival alliance. Available:http://www.chelonia.org./ (March 2014).

Webby, R., E. Hoffmann, and R. Webster. 2004. Molecular constraints to inter-species transmission of viral pathogens. Nature Medicine 10:S77–S81.

Webby, R., and J. Kalmakoff. 1998. Sequence comparison of the major capsidprotein gene from 18 diverse iridoviruses. Archives of Virology 143:1949–1966.

Weir, R. P., N. J. G. Moody, A. D. Hyatt, S. Crameri, R. Voysey, J. Pallister,and I. V. Jerrett. 2012. Isolation and characterisation of a novel Bohle-likevirus from two frog species in the Darwin rural area, Australia. Diseases ofAquatic Organisms 99:169–177.

Whittington, R. J., J. A. Becker, and M. M. Dennis. 2010. Iridovirus infectionsin finfish—critical review with emphasis on ranaviruses. Journal of FishDiseases 33:95–122.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 1

9:55

15

June

201

4

Relationship between Permeability Glycoprotein (P-gp) Gene ...

Documents

Transcript of Relationship between Permeability Glycoprotein (P-gp) Gene ...