Piscirickettsia salmonis, the aetiological agent of piscirickettsiosis, has recently been responsible for significant disease outbreaks in a variety of economically important freshwater and seawater fish species cultured worldwide. The development of effective control strategies for the disease has been limited due to a lack of knowledge about the physiology, intracellular growth, transmission and pathogenesis of the organism. It is also notknown how P. salmonis isolates differ when isolated from different locations and from different fish species. The aim of the present study was to characterise phenotypic and serological differences between various P. salmonis isolates. The growth of these isolates in vitro was also examined together with their virulence and their pathogenesis. The antibody response of fish to live and killed P. salmonis was examined in an attempt to understand the antigenicity of the organism, and various diagnostic techniques were developed as a means of controlling the disease.
A simple and effective method for the purification of P. salmonis from fish cells to examine the antigenic properties of P. salmonis was developed. P. salmonis purified using differential pelleting and 30% Percoll (v/v) gradient gave yields with the highest purity and the highest infectivity. Fish cell lines were used to examine growth characteristics of P. salmonis isolates in vitro. It appears that P. salmonis has a 3-5 days (d) lag-phase and an 8 d log-phase of exponential growth in CHSE-214 cells. When the susceptibility of different fish cell lines to different P. salmonis isolates were compared, the highest TCID50 ml-1 was obtained in CHSE- 214, SHK-1 and EPC cells.
The antigenic structure of the ten P. salmonis isolates was analysed using SDS-PAGE. P. salmonis isolates shared many bands but did appear to have differences in the low molecular weight regions of their profiles. Silver staining of Proteinase-K digested P. salmonis isolates showed that all isolates contained carbohydrate moieties below 30 kDa. Further characterisation of these was performed using a glycoprotein determination kit and a number of different biotin-labelled lectins. The SDS-PAGE profile of isolate R-29 appeared different when passaged three times through fish cell lines. The bacteria grown in CHSE-214, EPC and SBL cells, which showed a higher susceptibility to the bacterium, had more material in the lower region of the gel than bacteria recovered from less susceptible BF-2 and RTG-2 cells where the bands in this region had disappeared. A nested polymerase chain reaction (N-PCR) was used to differentiate different P. salmonis isolates. Isolate R-29 was the only isolate which resulted in a PCR product different to that of the other isolates examined.
A whole cell preparation of purified P. salmonis (type strain LF-89) was used to prepare a rabbit polyclonal antibody and six monoclonal antibodies. These antibody probes were used to compare the antigenicity between P. salmonis isolates. Major antigens of P. salmonis were observed at 95, 72, 60, 36, 32 and 20 kDa in Western blot (WB) analysis. The antibodies were also used to develop a variety of antibody-based tests for the more specific determination of the pathogen. A number of staining methods was also tried in an attempt to establish a rapid diagnostic method for piscirickettsiosis.
The virulence of P. salmonis isolates was compared in an experimental infection of Atlantic salmon. It was possible to re-isolate the bacterium from infected kidney tissue during the first 4 weeks post-inoculation (wpi) and to detect by ELISA in the kidney of fish sampled 8 wpi. The histopathology observed was similar to that seen during natural outbreaks of piscirickettsiosis. No obvious difference was seen in the pathology between the different isolates used in the experimental infection. Atlantic salmon were immunised with heat-killed preparation of various P. salmonis isolates. The highest antibody response was obtained in sera raised against P. salmonis isolate R-29. This was also found to be the case in fish infected with live R-29 compared with other P. salmonis isolates. Sera from fish either challenged with live bacteria or immunised with heat-killed bacteria were examined in WB analysis. Proteins between 30-60 kDa appeared to be recognised with sera raised against live P. salmonis, unlike the sera raised against heat-killed P. salmonis, except for a strong reaction with the band seen at 60 kDa with sera from fish immunised with isolate R-29. Other bands at 95,72 and 20 kDa were recognised by the fish sera raised against either heat-killed P. salmonis or the live bacteria.
For the first time, phage particles were observed to be associated with rickettsia infecting fish. Attempts were made to identify and characterise the phages in this study with a view to using them in the control of P. salmonis. The significance of the characterisation of P. salmonis isolates and the development of tests to detect and identify the pathogen carried out in this study is discussed within this thesis, together with further research which may lead to the development of successful control strategies for piscirickettsiosis.