Graham Wilcox

In 1995 and again in 1998/9, millions of pilchards (Sardinops sagax neopilchardus) were found dead or dying off the coast of Australia. Some pilchards in New Zealand were also affected during the 1995 epizootic. The epizootics moved progressively in a “bushfire-like” manner against the prevailing currents at a rapid speed. An uncharacterised herpesvirus, pilchard herpesvirus (PHV), was associated with both mortality events and, based on its occurrence in affected fish and the disease pathogenesis, generally accepted as the causative agent, although experimental evidence of its role was not established. Until recently, rapid and sensitive methods to detect PHV were not available and so research into the characteristics of this virus were limited. New methods developed and optimised for the detection of PHV include Polymerase Chain Reaction (PCR), in situ hybridization (IS H) and real-time PCR. These methods were then applied to a number of pilchard samples that were obtained at various times during the two outbreaks. Some sequence data for PHV was obtained. These data are the first step in obtaining information at the molecular level as to how this virus fits within the Herpesviridae. Since acquiring this initial sequence data, techniques have been applied to try to further characterise PHV. At the time of starting this project, opportunities to increase the known sequence were limited due to the inability to grow the virus in culture, the limited virus stocks and the best virus material only being part purified and of unknown viability. The additional sequence data obtained is still preliminary at this stage and requires further analysis. Three molecular methods have recently become available for detecting PHV (ISH, PCR and Real-time PCR) and were used to detect PHV in samples from fish collected pre-, during and post-outbreak. Samples collected before 1995 were not available to test. The real-time detected virus in fish collected 4 days before the epizootic front, during the front, and 8 days after the front had passed. The samples that were available for this stud y were archival and limited the information that could be obtained on the length of the virus incubation period and persistence of the virus after the epizootic front had passed. However, the information obtained was consistent with modelling by Murray et al. (2003) that predicted an incubation period of up to 12 days, depending on the wave speed of the epidemic. The detection of PHV in samples after mortalities had ceased, that appeared to be healthy by histology and were not positive by ISH or conventional PCR, indicates that virus persisted in the population of survivors at low levels. In 1999 an experiment was carried out to transmit PHV to apparently healthy pilchards in an attempt to (1) show that the virus could be transmitted to uninfected pilchards, (2) demonstrate that the infected pilchards develop gill les ions characteristic of the disease and (3) show that the virus can be re-isolated from experimentally infected fish. At the time of the experiment the only available methods for detection of PHV were histology and electron microscopy. As attempts to grow the virus in vitro were unsuccessful the lack of virus detected in experimentally exposed pilchards suggested that virus transmission was unsuccessful. Because the technology is now available to apply molecular techniques to detect PHV, the transmission trial samples have been re-analysed. Re-examination of archived transmission trial samples produced some interesting results. All samples were negative by ISH and PCR, yet 3 of 6 samples were positive by real-time PCR. Real-time PCR analysis indicated that these positives had low levels of virus present. The results suggest that PHV was transmitted to healthy pilchards during the trial, and results were also consistent with samples that showed some histological changes suggestive of early stages of disease, however disease was not produced within the time frame of the trial. It maybe that only low levels of virus were transmitted and were not enough to cause disease. The positives detected by real-time PCR still confirm that sub-clinical infection was achieved. Wild pilchard populations were surveyed in 2004 as part of the project, to determine whether the virus is currently detectable and causing disease. Results of this survey show that PHV is significantly present within the current pilchard population: PHV is now endemic. The re does appear to be variation between sub-populations of Australian pilchards, with the possibility that NSW pilchards are not infected at all. Further studies are required to prove this, using more randomised samples. The suggestion that PHV is now endemic makes sense with the lack of further epizootics seen after the 1998/9 epizootic. Mortalities of the scale seen in 1995 and 1998 are not likely to be witnessed again du e to the continued presence of PHV within the population and the likelihood of a degree of immunity in pilchards in the current population. It is expected, however, that small-scale mortality events will occur, but will either not be seen or not be reported. It is feasible that another PHV epizootic will never occur in Australian waters and further studies should include the testing of pilchards from overseas waters for the presence of PHV. If PHV is present elsewhere overseas, especially with no large-scale mortalities having been reported, then this m ay give an insight in to how PHV was introduced into Australian waters. PHV from the 1995 outbreak was compared over 373 bp of sequence data to PHV from the 1998 outbreak. The viruses are considered to be the same over this region. However, this is only a small amount of sequence data considering that the expected size of the genome is approximately 200 kbp. This should be interpreted as a preliminary investigation of sequence similarities between the 1995 and 1998 virus(es). Based on the current limited data, PHV appears to be related to other fish herpesviruses but bears no resemblance to mammalian or avian herpesviruses. This sequenced region is known to be highly conservative in herpesviruses, suggesting that PHV is indeed unique, although still classed morphologically as a herpesvirus.