Accurate Detection and Quantification of the Fish Viral Hemorrhagic Septicemia virus (VHSv) with a Two-Color Fluorometric Real-Time PCR Assay

Viral Hemorrhagic Septicemia virus (VHSv) is one of the world's most serious fish pathogens, infecting >80 marine, freshwater, and estuarine fish species from Eurasia and North America. A novel and especially virulent strain – IVb – appeared in the Great Lakes in 2003, has killed many game fish species in a series of outbreaks in subsequent years, and shut down interstate transport of baitfish. Cell culture is the diagnostic method approved by the USDA-APHIS, which takes a month or longer, lacks sensitivity, and does not quantify the amount of virus. We thus present a novel, easy, rapid, and highly sensitive real-time quantitative reverse transcription PCR (qRT-PCR) assay that incorporates synthetic competitive template internal standards for quality control to circumvent false negative results. Results demonstrate high signal-to-analyte response (slope = 1.00±0.02) and a linear dynamic range that spans seven orders of magnitude (R2 = 0.99), ranging from 6 to 6,000,000 molecules. Infected fishes are found to harbor levels of virus that range to 1,200,000 VHSv molecules/106 actb1 molecules with 1,000 being a rough cut-off for clinical signs of disease. This new assay is rapid, inexpensive, and has significantly greater accuracy than other published qRT-PCR tests and traditional cell culture diagnostics.


Introduction
Molecular diagnostic tools have facilitated the early detection, prevention, and spread of many important pathogens [1], led by the speed, sensitivity, and accuracy of Polymerase Chain Reaction (PCR)-based assays [2]. Their ability to diagnose targeted genetic sequences and quantify levels of infectious agents with hybridization probes has advanced screening technology for multiple human diseases, including influenza, hepatitis, and HIV [3,4]. Use of these approaches to elucidate and characterize plant and animal pathogens likewise is growing at a rapid pace [5,6]. Viral Hemorrhagic Septicemia virus (VHSv) causes one of the world's most serious finfish diseases, infecting .80 species across the Northern Hemisphere [7], yet there remains a need for a fast, sensitive, accurate, and inexpensive diagnostic test. VHSv is a negative-sense, single stranded RNA Novirhabdovirus of , 12,000 nucleotides, with six open reading frames of 39N-P-M-G-Nv-L'5 [8]. Infected fishes often swim erratically and have bulging eyes, distended abdomens, and extensive external/internal hemorrhaging [9]. The virus survives for up to 13 days in the water [10], and can be spread via ballast water, boating, equipment, and aquatic animals (e.g. birds, turtles, leeches, and amphipod crustaceans) [11][12][13][14]. It is transmitted most readily during the spring spawning season through fish waste, reproductive fluids, and skin secretions [11].
VHSv first was described in the late 1930s as ''Nierenschwellung'' in aquacultured rainbow trout (Oncorhynchus mykiss) from Europe [15]. It now occurs across the Northern Hemisphere as four genetically and geographically distinct strains (I-IV) and substrains, whose evolutionary and biogeographic patterns recently were analyzed by Pierce and Stepien [16]. Strains I-III primarily occur in Europe, where they infect a wide variety of marine, estuarine, and freshwater fishes. Strain IV (now classified as IVa; [17]) first was discovered in 1988 from North American Pacific coastal fishes, including salmonids [18,19], and now also occurs in Japan [20]. In 2000, another IV substrain (now designated as IVc per [16]) was discovered off the coast of New Brunswick, Canada, infecting the estuarine mummichog (Fundulus heteroclitus) and three-spined stickleback (Gasterosteus aculeatus) [21]. In 2003, a new and especially virulent substrain, IVb, was described from a moribund muskellunge (Esox masquinongy) in Lake St. Clair of the freshwater Laurentian Great Lakes system [17]. Substrain IVb since has spread throughout all five of the Great Lakes, infecting .30 species, including many commercially and ecologically important fishes, such as muskellunge, drum (Aplodinotus grunniens), walleye (Sander vitreus), yellow perch (Perca flavescens), and round goby (Neogobius melanostomus). Substrain IVb now contains at least 16 glycoprotein (G)-gene sequence variants [22], whose rapid spread and diversification in a quasispecies mode suggest that this strain mutates rapidly and may be highly adaptable (see [16]).
To avoid outbreaks of the virus, the Aquatic Invasive Species Action Plan [23] requires that aquaculture and baitfish vendors from U.S. states (Illinois, Indiana, Michigan, Minnesota, New York, Ohio, Pennsylvania, and Wisconsin) and Canadian provinces (Ontario and Quebec) have their fish products certified as VHSv-free prior to interstate or international transport. Cell culture is the VHSv diagnostic that is approved by the World Organization for Animal Health [24], along with the joint Fish Health Section of the U.S. Fish and Wildlife Service and the American Fisheries Society [25]. The cell culture process takes a month or longer for cell growth, cell confluency, viral growth, and confirmation PCR. It moreover lacks the sensitivity to detect low viral concentrations in carrier fish, and results in false negative levels reported as 43-95% [26][27][28].
To avoid those issues, the present research describes and evaluates a new, accurate, fast, and highly reliable assay to diagnose and quantify VHSv. This assay uses Standardized Reverse-Transcriptase Polymerase Chain Reaction, i.e. StaRT-PCR, which is a form of competitive template RT-PCR that yields rapid, reproducible, standardized, and quantitative measurement of data for many genes simultaneously [36]. StaRT-PCR uniquely incorporates internal standards (IS) in the rxn mixture to improve accuracy and circumvent false negative results. Our new assay is based on real-time PCR equipment that is readily available in most diagnostic laboratories, markedly improving on a previous version of our VHSv test [37], which also used StaRT-PCR, but relied on less common capillary electrophoresis equipment (i.e. Agilent; Agilent Technologies, Santa Clara, CA). In the present study, results from both assays are evaluated by us to determine the presence or absence of VHSv and measure concentration of the virus from fish samples in targeted field and laboratory studies. We assay the VHSv nucleoprotein (N)-gene and the fish reference beta-actin 1 (actb1) gene, assessing amplification relative to known numbers of their respective competitive IS molecules. Our new approach uses sequence specific fluor-labeled hydrolysis probes that can be used on a variety of real-time PCR thermal cyclers, on which positive VHSv results are visible as two colors on the realtime PCR plot (see Figure 1a; green = native template (NT), red = IS).
Results of this 2-color fluorometric assay are compared to those from our previously reported Agilent capillary electrophoresis test [37], SYBRH green qRT-PCR, and cell culture [24], using the same biological samples. The numbers of VHSv molecules are quantified from field-caught and laboratorychallenged VHSv-infected fish samples with the new 2-color fluorometric assay in comparison to the Agilent capillary electrophoresis test [37].

VHSv Assay Development
All primers and NT probes were matched to homologous sequences of the VHSv N-gene, based on all VHSv strains and substrain variants from NIH GenBank (http://www.ncbi.nlm.nih. gov/genbank/) and the literature, using Biosearch Technologies Real Time Design software (Novato, CA; http://www. biosearchtech.com/). The original muskellunge isolate MI03GL from the Great Lakes (GenBank Accession no. DQ427105) served as the reference for VHSv, and actb1 mRNA from the yellow perch Perca flavescens (AY332493) was used as the fish reference gene sequence. Selection criteria included: product lengths that were ,100 bp, with optimal melting temperatures of 65-68uC for primers and 68-72uC for probes. NT probes for the target and reference genes were labeled with FAM (fluorescein amidite dye).
The competitive template IS probes for the VHSv N-gene and the fish reference actb1 gene each were constructed by altering 5-6 bp of the NT probe sequences, and were labeled with Quasar dye having 670 nm maximum absorbance (Biosearch Technologies). The IS probes were designed to: minimize cytosine and thymine (30%), maximize adenine (50%), avoid guanine at the 59 end, have lengths ,24 bp, and have predicted melting temperatures 60.02uC of the NT probe. Synthetic NT and IS templates for VHSv and actb1 were assembled by combining the forward primer, probe, and connecting sequence through the reverse primer (Table 1), and were synthesized by Life Technologies (Grant Island, NY; http://www.lifetechnologies.com/us/en/ home.html). The BLAST procedure (http://blast.ncbi.nlm.nih. gov/Blast.cgi) was employed to verify that all primers, probes, and IS did not recognize other viral or fish DNA sequences.
To ensure that the probes did not bind to non-homologous template, their specificities were tested using synthetic templates for the VHSv Nand actb1 genes. Both synthetic templates (NT and IS) were serially diluted 10-fold from 10 211 M to 10 215 M and tested with all probes in PCR amplification experiments, following the directions for ''Performing the VHSv Assay'', as detailed below. For example, the VHS N-gene IS synthetic template was evaluated with the VHS N-gene IS probe, as well as the VHS N-gene NT probe, and vice versa. The same was done for actb1. Cycle thresholds (C t ) from the homologous and nonhomologous templates were compared at each dilution, and the non-homologous amplifications were calculated with formula 2 (2DCt) and multiplied by the known number of input copies. If the resulting numbers of molecules were .10% of the known input copy number, then the probe was re-designed, and the process was repeated.
After synthesis, the NT and IS for each gene were PCR amplified (Table 1) in 10 individual 10 ml rxns, containing 1 ml 750 nM of each primer, 0.5 U Go-TAQ polymerase (Promega, Madison, WI), 1 ml 10X MgCl 2 PCR buffer, 0.2 mM dNTPs, and RNAse-free water on a Rapid Cycler 2 (Idaho Technology, Inc., Salt Lake City, UT; www.biofiredx.com/). Rxns were run for 35 cycles of 5 sec at 94uC, 10 sec at 58uC, and 15 sec at 72uC, with a slope of 9.9. To purify the NT and IS, all 10 replicate PCR products per template were combined into a single tube, loaded onto individual 2% low melting pre-cast agarose gels from E-Gel iBase (Invitrogen, Grand Island, NY; www.invitrogen.com/), separated by electrophoresis, and visualized on a UV transilluminator. The NT and IS bands for the VHSv Nand actb1 genes were harvested from their respective gel collection wells. Mean concentrations (ng/ml) of each were calculated from 1 ml of their purified products as measured with an Agilent 2100 Bioanalyzer in triplicate, and converted into molarities according to the formula (1): Molarity tconcentration (mg=ml)s = tmolecular wt: (mg=mmole)s To control for inter-sample and inter-experimental pipetting variation, a synthetic internal standard mixture (ISM) was created with the purified IS described above. To prepare the original stock ISM ''A'', we estimated the relative concentrations of the VHSv N-gene and actb1 IS needed to achieve a 1:1 cDNA NT:IS relationship in a variety of samples (Table 2). Briefly, we mixed  An External Standardized Mixture -ESM (comprised of the synthesized NT and IS for the VHSv Nand actb1 genes) -was made to control for inter-lot and inter-experimental variation in probe fluorescence intensity, guard against inter-experimental variation in C t selection, and normalize the probe (see equation (2), ''Correction for variation in fluorescence among probes'').

Fish Samples used to Evaluate the VHSv Assay
Spleen tissues from a variety of fish samples were used to test our assay for the presence and concentration of VHSv (and to compare our results to other assays, using the same samples). Fish were obtained, maintained, anesthetized, and sacrificed following the Institutional Animal Care and Use Committee (IACUC) approved protocols from the University of Toledo (#106419), Michigan State University (MSU; East Lansing, MI) (#AUF 07/ 07-123-00), and the U.S. Geological Survey's (USGS) Western Fisheries Research Center Challenge Facility (WFCCF; Seattle, WA) (#2008-17). Fish were euthanized with an overdose of 25 mg/ml tricaine methanesulfonate (MS-222; Argent Chemical Lab, Redmond, WA) and decapitated to ensure death. To remove any external viral particles, each fish was washed separately by submerging it 3X in double distilled H 2 O. The surgical site (anus to operculum) was disinfected with 100% ethanol and betadine using sterile equipment. Spleen tissue was removed, placed into individual 1.5 mL eppendorf tubes, flash frozen in liquid nitrogen or stored in RNAlater (Qiagen), and kept at 280uC until further processing. Gloves and all equipment were changed between each fish to ensure sterile conditions. Specimens were disposed of following the respective approved biohazard protocols of the University of Toledo, MSU, and USGS.
Samples tested for VHSv included cDNA from 23 wild-caught Great Lakes fishes, including 10 infected and 13 negatives: two A series of laboratory challenge experiments were conducted by MSU-AAHL on certified VHSv-free juvenile muskellunge (Rathburn National Fish Hatchery, Moravia, Iowa) under MSU IACUC protocols AUF 07/07-123-00. Muskellunge were challenged via water immersion for 90 min with VHSv-IVb (isolate MI03GL) at 4.0610 3 pfu/ml, and the negative controls with 1 ml sterile maintenance minimum essential media. Fish then were placed into clean VHSv-free water, and later randomly sacrificed at pre-determined intervals, as previously described.
We also analyzed RNA from a series of juvenile yellow perch laboratory challenge experiments, using six-month-old (VHSvcertified-free) Choptank broodstrain [42] from the University of Wisconsin-Milwaukee's Great Lakes WATER Institute (Milwaukee, WI), which were conducted at USGS-WFRCCF under their 2008-17 IACUC protocol. Perch were challenged either via intraperitoneal injection of 1.0610 5 pfu/ml VHSv-IVb (strain MI03GL) or with immersion for two hours in the same dosage, while control groups had a dose of minimum essential media. Fish were selected randomly in days 0-6 for euthanization with 240 mg/L MS-222 and 1.2g/L NaHCO 3 . Dissection followed protocols described above.

Performing the VHSv Assay
Spleen tissue (0.25-0.50 g) was ground using a sterile mortar and pestle under liquid nitrogen, and its RNA was extracted with   A set of PCR rxns was run per each cDNA sample to determine the appropriate concentrations of NT and IS for actb1 to achieve a ratio of .1:10 and ,10:1 of amplified products. Once the NT:IS products were in balance, the VHSv Nand actb1 target genes were pre-amplified simultaneously to increase the signal (i.e. lower C t ) relative to non-specific background.
For each pre-amplification, a 10 ml volume of a master mixture was prepared with 5 ml TaqManH Universal Master Mix II (without uracil N-glycosylase; Applied Biosystems International (ABI), Grand Island, NY), 1 ml of 10X primer solution (final concentration: 75 nM) of the forward and reverse primers for the VHS N and actb1 genes (mixed together), and RNAse-free water. Eight ml of this master mixture was dispensed into individual wells of 0.1 mL 96-well TempPlateH (USA Scientific, Inc.; www. usascientific.com/) containing 1 ml cDNA and 1 ml of the appropriate ISM concentration (Table 2). This was done in triplicate to allow calculation of the mean and standard error (S.E.) of the relative VHSv N-gene concentration/10 6 actb1 molecules per fish sample. The plates then were sealed with a TempPlateH RT Optical Film and centrifuged for 2 min at 2000 rpm. PCR rxns were conducted on an ABI 7500 Fast using standard mode cycling conditions: 10 min at 95uC, followed by 13 cycles of 15 sec at 95uC and 1 min at 60uC. For the Poisson distribution experiments, 25 pre-amplification cycles were used due to lower amount of starting template. Three no-template controls per rxn, located on separated areas on the plate, were used to control for possible contamination.
A second round of amplification was performed, in which each pre-amplified sample was diluted 1000-fold with TE buffer (10 mM Tris-Cl, 0.1 mM EDTA, pH 7.4); 2 ml of each diluted product was placed into each well of a new 0.1 mL 96-well TempPlateH, along with 18 ml of a master mixture containing 10 ml TaqManH Universal Master Mix II (without uracil Nglycosylase), 2 ml of each 10X primer solution (final concentration: 750 nM), 2 ml of each NT and IS probe (final concentration: 200 nM), and RNAse-free water. This second amplification was conducted as described above, except run for 40 cycles. The number of VHSv molecules/10 6 actb1 molecules was calculated using equations (2) and (3) Specificity, True Accuracy, and Linearity Our assay was tested for non-specific amplification using two human viruses (Encephalomyocarditis virus and Vesicular Stomatitis virus) and five fish viruses related to VHSv (Hirame rhabdovirus, Infectious Hematopoietic Necrosis virus, Infectious Pancreatic Necrosis virus, Spring Viremia of Carp virus, and Snakehead rhabdovirus). The Snakehead rhabdovirus is the nearest relative to VHSv, with 62% sequence similarity [8,16]. Twenty-five VHSv isolates were tested to evaluate amplification across a range of European, Asian, and North American variants (Table 3), encompassing all four strains. All samples were assayed in triplicate.
To measure true accuracy -the agreement between a measurement and its known value [38] -the relationship between the observed versus expected numbers of VHSv N-gene and actb1 molecules based on Poisson analysis was determined [39]. Ten replicates were amplified for the VHSv Nand actb1 genes over a series of limiting PCR dilutions, which were predicted to contain 40, 20, 10, 7, 6, 5, 4, 2, 1, 0.7, 0.4, and 0.1 molecules. Linear regression analysis was performed in the R statistical software suite v2.15.2 [40]. A x 2 test (in Microsoft Excel) compared the number of molecules measured with the 2-color fluorometric assay versus those from the Agilent 2100 Bioanalyzer, at the same dilutions.
Linearity was measured over two series of dilution experiments to: 1) determine the maximum and minimum ratio of NT to IS that produced reproducible results, and 2) verify that our test followed a linear trend in calculating the expected number of molecules per dilution. The first dilution set was made by mixing a constant amount of synthetic NT with decreasing amounts of IS to generate dilutions of: 1:1 (6610 4 molecules), 1:2, 1:3, 1:4, 1:5, 1:6, 1:7, 1:8, 1:9, 1:10, 1:12, 1:14, 1:16, 1:18, and 1:20 (3610 3 molecules) for both genes. Identical procedures were performed by holding the IS constant, while varying the NT. The second dilution series evaluated linearity for the VHSv Nand actb1 genes using 10-fold serial dilutions of the ESM at a 1:1 ratio, with dilutions of 6610 6 , 6610 5 , 6610 4 , 6610 3 , 6610 2 , 60, 6, and 0.6 molecules. Regression analyses were conducted to determine correlation (R 2 ), slope (linearity), and relation to a linear trend (F) among the various dilutions of NT:IS and IS:NT for each gene. Imprecision was reported as the coefficient of variation (CV), calculated as the standard deviation divided by the mean of triplicate measurements at each dilution (reported in %) (in Microsoft Excel) [41]. In addition, S.E. was calculated for each sample. For these linearity experiments, PCR was done as specified above in ''Performing the VHSv Assay'', but substituting the cDNA and ISM with either 2 ml of the ESM (from dilution 1) or a concentration of 1:1 NT/IS (dilution 2). Each dilution was run in triplicate, with a negative/no template control for each run.

VHSv Detection Comparisons of Our Assay to Others
Results from the new 2-color fluorometric test are compared to those from our prior Agilent capillary electrophoresis assay [37], conventional SYBRH green qRT-PCR, and cell culture to evaluate their relative abilities to detect VHSv in 63 fish samples (see ''Fish Samples used to Evaluate the VHSv Assay''). All samples were analyzed in triplicate and all runs had positive and negative controls. Each PCR rxn included a known cell culture positive, a negative VHSv cDNA, and a reagent negative control (nuclease-free H 2 O). PCR products were visualized on 1% agarose gels to confirm positive/negative results. The amount of yellow perch fish tissue available from the USGS laboratory challenge experiments precluded analysis with cell culture. x 2 tests [43] were used to compare the results among the approaches.
Cell culture was performed at MSU's AAHL by M. Faisal and R. Kim, following standard OIE [24] procedures. If results were positive, RNA was extracted from infected cells as described above, reverse transcribed with Affinity Script Multiple Temperature Reverse Transcriptase PCR (Stratagene, La Jolla, CA), and amplified following OIE [24].

VHSv Quantification using Our Assay
Positive samples were quantified with our new 2-color fluorometric real-time PCR assay and compared to our earlier results from the Agilent capillary electrophoresis procedure [36] for the 63 test fish, with linear regression in R and an F-test [43]. Numbers of VHSv/10 6 actb1 molecules were measured in triplicate, from which means and S.E. were calculated. Relative numbers of VHSv molecules were compared between laboratory challenged muskellunge showing clinical signs of infection (e.g. external hemorrhages; N = 9) versus those without signs (N = 9). A x 2 test (Microsoft Excel) was used to determine if a threshold number of VHSv molecules characterized the appearance of the clinical signs. Due to limited sample size, a power analysis (G*Power2; [44]) was used to estimate the number of fish needed to achieve 95% confidence, with an effect size of 0.50 [45].

Performance of Our 2-Color Fluorometric Assay for VHSv
Our test results are negative for all other viruses (i.e. did not result in amplification; Table 3), including human viruses (Encephalomyocarditis virus and Vesicular Stomatitis virus) and fish viruses that are related to VHSv (i.e. Hirame rhabdovirus, Infectious Hematopoietic Necrosis virus, Infectious Pancreatic Necrosis virus, Spring Viremia of Carp virus, and Snakehead rhabdovirus). All four VHSv strains (I-IV) and all substrains evaluated (I, Ia, II, III, IVa, IVb, and IVc) yield positive amplification results with our test (Figure 1a; Table 3). Thus this new assay is specific for VHSv.
Numbers of VHSv molecules/10 6 actb1 molecules measured from the spleen tissues of positive fish are higher in the new assay, ranging to 1.21610 6 VHSv molecules/10 6 actb1 ( = 1.90610 4 VHSv molecules) than for the Agilent capillary-based test, which range to 8. All data and analyses are publically accessible on the University of Toledo Lake Erie Center's VHSv webpage (http://www. utoledo.edu/nsm/lec/research/glgl/VHS/VHS_main.html).

Discussion
Disease diagnostic laboratories depend on rapid, sensitive, and accurate detection methods, which are easy to employ and costeffective. Cell culture is the VHSv diagnostic approved by the World Organization of Animal Health [24], which takes up to a month to perform in clinical settings and often results in substantial false negatives -as revealed here and by other studies [26][27][28]. Compared with traditional cell culture, all PCR-based assays [32] -including the present one -show enhanced ability to detect VHSv since they amplify both the negative-strand nonreplicating genomic RNA and the positive-strand replicating mRNA transcripts. Amplification of both transcripts may be beneficial since positive results may denote new spread of VHSv or latent cases in the geographic region where the samples are taken. This can aid in diagnosis of viral infections.
The present assay detects and quantifies VHSv-IVb in fishes from the Great Lakes using primers and probes that are homologous to the N-gene sequence of the widespread IVb isolate MI03GL and matches conserved sequence regions among all VHSv strains and substrains. Results demonstrate cross-reaction with all other VHSv strains and substrains tested. Other human and fish viruses do not amplify. Thus, our assay is VHSv-specific and detects all of its known variants.
Specifically, exogenous (IS) and endogenous controls (the commonly used reference gene actb1) facilitate optimal detection of true positives and act to normalize the quantification of viral molecules. Use of IS is recommended by the International Organization for Standardization [46], the U.S. Environmental Protection Agency [47], and the U.S. Food and Drug Administration [48]. Tests for Hepatitis C virus [49] and Human Immunodeficiency virus [50,51] already have implemented IS in their assays.
Our assay is sensitive, follows a linear relationship with increasing viral concentration, and is highly reproducible. It detects down to five VHSv N-gene molecules with 100% accuracy, based on Poisson distribution. Other real-time PCR assays for VHSv had much higher detection thresholds. Notably, Liu et al.'s [29] test required $140 viral copies of VHSv, and assays by Hope et al. [32] and Garver et al. [33] needed $100 viral copies. Our results are consistent for samples containing six to 6,000,000 VHSv molecules. Stochastic variation is evident only at extremely low dilutions (, five molecules). Results confirm reliability from concentrations of 1:1 to 1:20 NT:IS, with some slight increase in CV at dilutions .1:10. We thus recommend adjusting the relative concentrations of NT:IS to maximize accuracy, following recommendations in the Materials and Methods section ''Performing the VHSv Assay''. All quantification values reported here fall within this 1:10 range, which allows us to distinguish , a 1.25-fold C t difference. Our assay also should work well with highly degraded samples (e.g. dead fish in the field), as described for human cancer samples using this type of approach by some of our team members [unpublished data]. This 2-color fluorometric real-time assay is highly accurate and free of the size separation steps required for our previouslypublished Agilent capillary electrophoresis approach [37]. Here we determine higher numbers of VHSv molecules for the same fish samples, due to the re-design of primers and use of fluorescentlabeled probes. Results from both methods have a linear relationship and are readily cross-calibrated.
Laboratory challenged muskellunge showing clinical signs of infection have a greater mean number of viral molecules than those without. It is estimated that , 1610 3 VHSv/10 6 actb1 molecules ( = 3.6610 2 VHSv molecules) appears to mark a clinical threshold for signs of VHSv. However, exhibition of clinical signs at this biomarker could be species-specific, and may differ between fish in the laboratory and those in the field. Further experiments are warranted to validate this assumption.

Conclusions
Our assay is highly sensitive and accurate, free of false negatives, and reliably quantifies a wide range of VHSv in fish tissue samples. Other PCR-based methods and cell culture techniques had high proportions of false negatives since they lacked intrinsic quality control, which could lead to spread of the virus. This new test will aid rapid, accurate, and low-cost diagnosis of the disease. It has wide applicability across the geographic range of the virus, and should be highly successful in elucidating new occurrences and circumventing spread.