Thesis (Ph. D.)--University of Hawaii at Manoa, 1991. / Includes bibliographical references (leaves 159-175). / Microfiche. / xiii, 175 leaves, bound ill. (some col.) 29 cm

Nile tilapia, Oreochromis niloticus, is one of the most important farmed fish globally. One of the most serious bacterial diseases constraining global tilapia production is Francisellosis caused by Francisella noatunensis subsp. orientalis (Fno). Although outbreaks of Fno are increasing worldwide, there are no licenced commercial vaccines to prevent the disease for use on tilapia farms. Thus, the current treatment of choice is the use of antibiotics combined with increasing water temperature up to 30°C. Studies investigating the diversity of circulating Fno isolates and the immune response of tilapia elicited by vaccination against piscine francisellosis are lacking. In addition, the current conventional and molecular tools used for detection of Fno have many drawbacks, making detection of Fno a challenging process. In this study, five clinical isolates of Fno from diverse geographical locations (UK, Costa Rica, Mexico, Japan and Austria), previously characterised by morphology, genotype, antimicrobial susceptibility and virulence, were used in a proteomic study. The whole proteomic cell profile of the five isolates were homogenous by one-dimension sodium dodecyl polyacrylamide gel electrophoresis (1D-SDS-PAGE), while minor differences in the intensity of 15 proteins between the strains were observed by two-dimension SDS-PAGE (2DE), including some important virulence related proteins. The UK isolate was the most significantly different isolate when compared to the other Fno isolates in the current study. The Fno UK isolate had significantly higher abundance of 10/15 of the significantly expressed proteins including four of the essential pathogenicity and virulence related proteins (IglC, GroEL, DnaK, ClpB) compared to the other used Fno isolates. The antigenic profiles of the five Fno isolates were studied by 1D western blotting using tilapia hyper immune sera which recognised an immunodominant band of a molecular weight of ~ 17-28 kDa in all tested Fno isolates. Liquid chromatography-electrospray ionization-tandem mass spectrometry (LC/ESI/MS/MS) identified 47 proteins in this antigenic band. Some of the identified proteins are associated with Fno pathogenicity. 2D western blot analysis of the vaccine isolate (Fno UK) revealed differential antigen recognition between sera from vaccinated and non-vaccinated fish following experimental challenge (26 antigenic spots recognised by sera from vaccinated fish; 31 antigenic spots recognised by sera from vaccinated and challenged fish and 30 antigenic spots recognised by non-vaccinated and challenged fish). The identity of these proteins was determined by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) and some of them are known Francisella virulence related proteins. Bioinformatics analyses revealed diverse categories of proteins with high biological functions, however the vast majority of these proteins are involved in energy production and metabolic pathways of the bacteria. This detailed analysis will facilitate the development of cross-strain protective subunit Fno vaccines and antigen-targeted Fno diagnostics. The outer membrane proteins (OMPs) of the same five Fno isolates were extracted using the ionic detergent sarkosyl. The OMP fraction of the different isolates were separated via 1D-SDS PAGE and the digested peptides of the UK isolate were analysed by LC/ESI/MS/MS. High degree of similarity was observed in the OMP profile of the five Fno isolates with an abundant protein band at 17-28 kDa, which was found to be antigenic by 1D western blot using convalescent tilapia sera. LC/ESI/MS/MS analysis of the OMPs of the Fno UK isolate identified 239 proteins, including 44 proteins in the antigenic band (17-28 kDa). Comparison between the proteins identified in the immunogenic band of whole cell lysate and OMP fraction of the Fno UK isolate showed 30 common proteins between the two preparations, 17 proteins were identified only in the whole cell extract and 14 were identified only in OMP fraction. Outer membrane proteins (e.g. Omp-A), virulence related proteins such (e.g. IglC) and other stress related proteins (e.g. AhpC/TSA family peroxiredoxin) were more abundant in the OMP fraction than the whole cell lysate. In silico analysis enabled prediction of the function and location of the OMPs identified by Mass-spectrometry. The findings of this study provide preliminary data on bacterial surface proteins that exist in direct contact with the host immune defence during infection and offering an insight into their potential role as novel targets for Fno diagnostics and vaccine development. The efficacy of an injectable whole cell oil-adjuvanted vaccine was evaluated against challenge with heterologous Fno isolates in Nile tilapia, Oreochromis niloticus. Three duplicate groups of 130 healthy Nile tilapia (~15 g) were intraperitoneally (i.p.) injected with the vaccine, adjuvant-alone or PBS followed by an i.p. challenge with three Fno isolates from geographically distinct locations. The vaccine provided significant protection to all immunised tilapia groups with a significantly higher relative percent survival (RPS) of 82.3% against homologous challenge, compared to 69.8% and 65.9% after heterologous challenge. Protection correlated with significantly elevated specific antibody responses and western blot analysis demonstrated cross-isolate antigenicity with sera from fish post-vaccination and post-challenge. Moreover, a significantly lower bacterial burden was detected by quantitative real-time polymerase chain reaction (qPCR) in conjunction with significantly greater expression of IgM, IL-1β, TNF-a and MHCII 72 hours post-vaccination (hpv) in spleen samples from vaccinated tilapia compared to those of adjuvant-alone and control fish. The latter results suggested stimulation of protective immune responses following vaccination. In addition, a whole cell formalin killed autogenous immersion vaccine against Fno was developed using the same isolate used for the injectable vaccine. Duplicate tanks of 35 tilapia fry were immersed in the vaccine or in sterile Modified Muller Hinton broth (MMHB) diluted in tank water (1:10 dilution) for 30 s and at 30 days post-vaccination (dpv), all fish groups were immersion challenged with the homologous Fno isolate and monitored for 21 days. A moderate RPS of 43.7% was provided by the vaccine. Serum IgM levels were below the threshold in 30 % of the vaccinated fry 30 dpv. Also, the IgM levels of the vaccinated fry were not significantly different from control fry 21 days-post challenge. A recombinase polymerase amplification (RPA) assay was developed and validated for rapid detection of Fno. The RPA reaction was performed at a constant temperature of 42°C for 20 min. The RPA assay was performed using a quantitative plasmid standard containing a unique Fno gene sequence. Validation of the assay was performed not only by using DNA from Fno, closely related Francisella species and other common bacterial pathogens in fish farms, but also by screening 78 Nile tilapia and 5 water samples collected from UK and Thailand. All results were compared with those obtained by previously established real-time qPCR. The developed RPA showed high specificity in detection of Fno with no cross-detection of either the closely related Francisella spp. or the other species of bacteria tested. The Fno-RPA performance was highly comparable to the published qPCR with detection limits at 15 and 11 DNA molecules detected, respectively. The Fno-RPA was rapid, giving results in approximately 6 min in contrast to the qPCR that required approximately 90 min to reach the same detection limits. Moreover, the RPA was more tolerant to reaction inhibitors than qPCR when tested with field samples. The fast reaction, simplicity, cost-effectiveness, sensitivity and specificity make the RPA an attractive diagnostic tool that will contribute to control the infection through prompt on-site detection of Fno. The overall results of this study indicated that Fno isolates from different origins share a high degree of homology in their proteomic and antigenic profile. Proteomic characterisation data of Fno isolates has contributed to understanding the diversity of Fno isolates and assisted in identifying suitable candidates for developing an effective Fno vaccine. / Moreover, this study has proven the efficacy of a cross protective Fno injection vaccine in tilapia fingerlings, with further optimisation needed for immersion vaccination of fry, and given insights into the immune response of tilapia to vaccination against francisellosis. In addition, it provided a rapid, sensitive, specific and robust molecular tool for detection of Fno that can assist surveillance and control of piscine francisellosis on tilapia farms.