Viral hemorrhagic septicemia virus is a potentially lethal pathogen that infects fish species in wild and hatchery fish populations mainly in Europe and North America. This project was conducted as the second year of a two-year study of VHSV in Lake Erie yellow perch, in collaboration with Ohio Department of Natural Resources’ Fairport Fish Research Station. Taking advantage of a routine yellow perch survey by Ohio Department of Natural Resources (ODNR) off the mouth of the Chagrin River in the Central Basin of Lake Erie, we obtained biotic data on subsamples of yellow perch (sex, age, size, and density) and abiotic data (water temperature, dissolved oxygen saturation, and conductivity).  To obtain better estimates of prevalence of infection, we pooled yellow perch tissue samples of kidney, spleen, and brain tissues from two individuals per pool and sent samples to diagnostic testing laboratories. We also collected 0.5 to 1 cc of blood from pooled and untested individual fish. Blood samples were processed after return to port and stored the sera at -80°C.  Stored sera were used for subsequent obtained antibody analyses and viral RNA analyses.

Confirming our 2007 findings, we detected infection during a three-week period following spawning in 2008. Unlike results from the pools of five individuals in 2007, we obtained variable results from cell culture assays for VHSV from different testing labs. We also detected antibodies in yellow perch sera during the four-month period following detectable infections.  Because Ohio DNR observed no yellow perch mortality associated with the VHSV infection during either 2007 or 2008, we investigated potential morbidity effects by applying a bioenergetics model to simulate the effect of reduced consumption on the specific growth rate of yellow perch.  Depending upon the level of effect of VHSV infection on consumption, the VHSV incidence patterns of infection have a large enough potential influence on growth rate to justify additional morbidity research.  Finally, in a preliminary study, we were able to develop a diagnostic protocol based on reverse transcriptase-polymerase chain reaction (RT-PCR) coupled with nested polymerase chain reaction (PCR) technologies was developed for the detection of the nucleocapsid gene (N-gene) of the IVb sub-lineage of VHSV in blood samples.  Using this protocol on stored sera, we were able to confirm positive tests from cell culture assays.

VHSV is an invasive viral species in the Great Lakes that has the potential to effect wild fish populations and the fishing industry. Since its detection in Lake Erie in the spring of 2006, there has been a great concern from both American and Canadian agencies regarding the effect of the virus on the health of fish populations in the lake. The virus was first detected in North America in returning adult coho and chinook salmon in 1988 in Washington state (Meyers, 1995). Sequencing analysis confirmed that the North American isolates are a different genotype (IV) compared to the European genotypes (I-III) (Meyers, 1995 and Enisen-Jensen, 2001).  Fish that are infected with the virus exhibit a variety of clinical signs ranging from pale viscera to internal hemorrhaging.  High mortality rates are associated with some genotypes of the virus and in some environmental conditions with adult fish having death rates ranging from 25 to 75% and with juveniles having close to a 100% rate of death.  The virus is spread by urine and female sex products.  Gills and fin bases are thought to be the main portals of infection to the fish (Meyers, 1995, Lorenzen, 2000, and Harmache et al., 2006).

In 2003 VHSV was isolated from muskellunge in Lake St. Clair (Elsayed, 2006). In the spring 2006 there was a large die off of yellow perch and several other species in Lake Erie and the cause was attributed to VHSV. VHSV has also been attributed to large die offs in 2006 in Lake Ontario freshwater drum, smallmouth bass, and bluegill and in Lake St. Clair in gizzard shad, redhorse and blunt nose sucker, and yellow perch (USDA, 2006).  Sequencing analysis of the isolates obtained from morbid fish species in the Great Lakes has proven that these isolates are a sublineage of genotype IV and has subsequently been labeled as IVb.  The genotype IVb is believed to be endemic to the Canadian east coast (Elsayed, 2006 and Gange, 2007).

With support from the Lake Erie Protection Fund, we started a pilot study on the prevalence of viral hemorrhagic septicemia virus (VHSV) infection in yellow perch in coordination with the Ohio Department of Natural Resources (ODNR) ongoing yellow perch survey.  Our project focused on providing seasonal epidemiological descriptive data for VHSV infection in yellow perch by testing a sub-sample of fish from the ODNR's trawling effort off the mouth of the Chagrin River.  The Chagrin site (approximate area of 625 km2) is located in Management Unit 2 (approximate area of 7672.67 km2) in the Central Basin of Lake Erie.  We sampled 400 Yellow Perch from May 15 to October 29, 2007, and sorted the sample into 80 pools (five individuals per pool) of which 48 pools were adults, 16 pools were yearlings, and 16 were young of the year (YOY).  On two successive sampling dates (May 22 and June 6), all adult pools tested positive or VHSV.  Before and after this time period all adult pooled samples tested negative. We also never found any positive pools of YOY or yearling Yellow Perch.

In general, both the surface and bottom temperature of the sample locations increased from early spring to the end of the summer. Surface temperature varied little between each depth station, but bottom temperature displayed more variation with much lower temperatures in the 20m depth station.  On VHSV positive dates, May 22nd and June 6th, surface and bottom temperatures varied at all depth stations.  On May 22nd the surface temperatures ranged from 11-12.7°C and bottom temperatures from 6.6-12.4°C compared June 6th temperatures, 17.1-19.25°C at the surface and 16.2-19.3°C at the bottom.  There is no correlation seen between VHSV infection and DO, pH and conductivity.

Fish density appeared not to have a significant effect on infection.  According to the catch per hectare (density) data sampling dates on August and September had higher catch per hectare means for off shore compared to the positive dates.  On June 18th catch per hectare had a higher near shore mean.  Infection with VHSV did not correlate with the sex of the fish. On the two days that we detected the virus in fish, all adult males and females tested positive for the virus (data not shown).  Overall temperature appears to have the strongest association with infection.

Obtaining all positive pools for two sample periods was unexpected and implies that all spawning Yellow Perch were infected.  These results raise two important questions.  Because this was the second year VHSV was detected in Lake Erie, what is the reservoir for the virus?  Second, what is the actual level of infection in the spawning population?  Laboratory studies of VHSV infection (Isshiki et al. 2003) indicate that survival of infection is low.  At the levels of infection we are observing,  Yellow Perch mortality due to VHSV cannot be as high as observed in salmonid populations.  Furthermore, preliminary models of the dynamics of the epizootic indicate that substantial numbers of individuals must have latent infections.  Other Rhabdoviruses are known to produce latent infections in nervous tissues, and some studies have suggested that VHSV infection persists longer in brain tissue (Mazarakis, 2001).  Protocols established prior to our study with testing facilities required use of tissue from posterior kidney and spleen.  Results from our pilot work indicate that we need to use more flexible protocols with respect to pool size and sampled brain tissue during periods in which we did not expect active infection.

For 2008, we decreased the sample size to two individuals per pool to improve the estimate of proportion of individuals infected during the expected period of active infection.  The proportion of susceptible individuals after an epizootic episode is an important statistic for estimation of parameters in epizootic models.  We estimated the probability of individual infection (p) as binomial function of the number of fish per pool (eqn 1).  
     p = 1- (1-P)^(1/v)   (1)
where P is the probability of a positive pools and ν is the number of individuals per pool.  The lower bound for a 95% confidence interval on the probability that all pools are positive (L1) is determined by the standard confidence interval for a binomial population parameter: (eqn 2).
     L = ε/(ε + F[0.025,2,2*ε])  (2)
where ε is the number of replicate pools.

With 100% positive pools of five individuals per pool, equations 1 and 2 indicate that the lower bound of the probability of a positive pool at 95% confidence on twelve replicate pools of five individuals each is 0.735, which corresponds to a probability of infection range of  [0.23, 1.00].  The 95% confidence interval for twelve replicates of two individuals per pool is [0.486, 1.000].

To better estimate prevalence of VHSV during the peak of infection when temperatures were between 10 and 15°C, we decreased the number of individuals per pool from five to two.  For each individual we collected kidney and spleen tissue and prepared the samples for shipment according to the instructions from the LaCrosse Fish Health Center and the Ohio Department of Agriculture Diagnostic Lab where pools were tested for the presence of VHSV.  We also collected sera from adult yellow perch by caudal vein or cardiac puncture during sampling, stored on ice before centrifuging for serum separation, and stored the sera at -80°C for future analysis.  Sera were subsequently used in both preliminary analyses of antibody presence and detection of viral RNA.

Antibody analyses were conducted through a collaboration arrangement with Dr. Scott LaPatra (Clear Fork Springs Food, Inc. Research Division). The samples processed by Dr. LaPatra followed a complement-dependent neutralization test developed for IHNV in rainbow trout.  Viral RNA was extracted from whole serum samples using the Qaigen RNeasy mini RNA extraction kit.  We obtained a positive control viral RNA (MI03GL, accession #427105) isolate from the U.S. Geological Survey Western Fisheries Research Center. This isolate was isolated from Muskellunge in Lake St, Claire, MI and was identified as a member of the IVb genotype. The primer sequences used in the RT-PCR portion of this protocol were designed from the American Fisheries Society Blue Book for the diagnostic method for VHSV detection. An RT-PCR was set up using a series of dilutions (1:100, 1:200, 1:500, 1:1,000, and 1:10,000) on the MI03GL isolate of VHSV using primers VHSV-1 and VHSV-2 and a nested PCR was preformed on the RT-PCR products using primers nest-1a and nest-2.  These experiments were conducted to confirm a working protocol for the detection of VHSV n-gene sequences. Subsequent RT-PCR and nested PCR experiments using primers VHSV-1 and VHSV-2, and nect-1a and nest2 respectively, were performed on RNA extracted from five yellow perch samples and the MI03GL isolate as a positive control.

To study potential morbidity of VHSV infection in yellow perch populations, we implemented a version of the Wisconsin bioenergetics model for yellow perch in Lake Erie.  We simulated the effects of reduced consumption during periods of infection and explored predicted changes in specific growth rates and growth of age IV yellow perch.

Our research during 2008 accomplished our objectives, and produced an unexpected result.  Coupled to the 2007 results, we have made the following observations:  1) Active VHSV infection occurs during a three-week period at the end of spawning with a very high proportion of the population infected; 2) Infection occurs with a lack of evidence of external or internal clinical signs of infection and there was no observed mortality associated with VHSV infection; 3) Morbidity effects on consumption rate during the period of infection should cause an observable decline in growth rates and size at age; 4) Antibodies to VHSV are detectable for up to 4 months following infection period; 5) We can isolate VHSV RNA from blood samples with a protocol based on reverse transcriptase-polymerase chain reaction (RT-PCR) coupled with nested polymerase chain reaction (PCR) technologies to detect the nucleocapsid gene (N-gene) of the IVb sub-lineage of VHSV; and 6) Sequencing shows the N-gene from VHS in yellow perch is identical to an N-gene isolate from Muskellunge in Lake St, Claire.

The unexpected result was the lack of consistent VHSV detection with the cell culture methods used by two testing laboratories (the Ohio Department of Agriculture’s Diagnostic Laboratories and the U.S. Fish and Wildlife Service’s La Crosse Fish Health Center).  In 2008, we collected a total of 36 pools of adult fish age-2 and older during expected pre-infection and 24 post-infection periods, and 82 pools during expected infection period.  La Crosse received samples during the expected pre-infection period, ODA and La Crosse received samples during the expected infection period, and ODA received samples during the expected post-infection period.  During the expected pre- and post-infection periods, all pools tested negative for infection.  During the expected infection period, results varied.  All pools on 14 May sent to La Crosse and ODA tested negative.  On 20 May 10 out of 11 pools sent to the ODA tested positive, but 12 pools sent to La Crosse tested negative.  On 28 May 10 out of 11 pools sent to ODA tested positive and 12 pools sent to La Crosse tested negative.  While on 2 June all 12 pools sent to ODA tested positive and all 12 pools sent to La Crosse tested negative .  A very small percentage of sampled fish exhibited clinical signs of infection (external skin lesions and/or kidney and spleen hemorrhaging), but the vast majority of fish appeared to be healthy.

The main difference in methodology used in 2007 and 2008 was the number of fish per pool during the active infection period.  We reduced the pool size to two fish in 2008 from five in 2007.  Because sample preparation protocols did not vary and samples were randomly assigned to the two testing labs, it appears that some aspect of the sensitivity of the cell culture protocols used by the two labs may be the cause of the discrepancy.  Our successful detection of VHSV in blood sera from individuals in positive pools from ODA testing confirms the presence of the virus at testing dates that the La Crosse method showed negative results.  It thus appears that cell culture detection of VHSV should be supplemented with RT-PCR amplification of virus from sera of tested fish.

This project addressed several important priorities of Great Lakes agencies and organizations.  The priorities list includes: the Great Lakes Governors' Priorities to stop invasive species; the Health and Disease Issue of the Sea Grant's Thematic Areas of Aquaculture, Ecosystems and Habitats;  the Fisheries, Marine and Aquatic Science Literacy priority; Aquatic Invasive Species priority, and the Healthy Coastal Ecosystems priority of the National Sea Grant Program. VHSV can possibly be a potential risk of infecting cultured fish and a considerable impact on the yellow perch population.  This project also addressed the Thematic Area of Fisheries because of its current collaboration with Fairport Harbor Fisheries Division of the ODNR, and the collaboration contributed to the Marine and Aquatic Science Literacy by providing results to researchers and public.  This research has added important biological characterization of the threats to fish health and economic foundations of fisheries in the Great Lakes of introduced viruses.

The preliminary work has produced two undergraduate honors theses, an M.S. Thesis, and currently there is one manuscript under review for the Journal of Great Lakes Research.