Development of methods to determine prevalence of Flavobacterium psychrophilum in farm systems

Abstract

ABSTRACT Flavobacterium psychrophilum (Fp), the aetiological agent of rainbow trout fry syndrome (RTFS) and bacterial cold-water disease (BCWD), is responsible for significant mortalities and economic losses in the salmonid aquaculture industry worldwide. Currently, there is no effective commercial vaccine against RTFS available, and the treatment of the disease depends on the oral administration of a wide range of anti-microbial compounds, some of which have proven ineffective. With unsuccessful disinfection procedures, possibilities of antibiotic resistance developing and no commercial vaccine available, there is an increased need to rapidly detect Fp and reduce mortalities in the industry by improving control measures in the farm system. The aim of this thesis was to investigate possible sources of Fp in a rainbow trout fry farm system and to use this data to develop strategies to reduce the prevalence of the pathogen with this farming system. Novel assays to detect Fp (loop-mediated isothermal amplification; LAMP), quantify Fp (quantitative PCR; qPCR) and to detect the fishes’ host response to Fp (Luminex™) were developed, and then used alongside bacterial culture and nested PCR to determine the prevalence of Fp on a commercial fish farm. Four batches of eggs from 3 different geographic sources were collected on arrival at the farm and tested for the prevalence of Fp. Fry from these batches were monitored as they grew and were moved to different sites at the farm. Kidney, spleen and blood were collected at 3 different life stages from the fry, until they were sold for ongrowing by the farm. Water samples from the inlet, outlet and fry tanks were collected at each sampling point. PCR analysis and bacteriology were the two main methods selected for screening the eggs and fry tissue for Fp. All sources of eggs were found to be positive for Fp with prevalences ranging from 1.1 % - 1.9 % and there was a significant increase in prevalence over time for all 4 batches of eggs ranging from 19.8 % - 34.6 % by the final life stage sampled. There was also a substantial difference in the numbers of fry samples positive for Fp depending on whether nested PCR or bacterial culture were used, as well as the organ (kidney or spleen) tested. This highlighted the importance of sampling both organs rather than just the one. Nested PCR was more sensitive than culture with 13 % of the fry samples reported as Fp positive, by sampling both the kidney and spleen collectively, while only 5 % were Fp positive by bacteriology. The levels of Fp in all samples could not be quantified by qPCR due to limits in the sensitivity of the assay. For those samples that were quantified at the levels of Fp detected by qPCR ranged from 3.38 x 104 well-1 - 2.07 x 106 well-1 genome copies in egg samples; from 3.38 x 103 well-1 – 3.07 x 107 well-1 genome copies well-1 in tissue samples (spleen or kidney), and from 7.89 x 103 – 7.22 x 104 genome copies well-1 in water samples. The sensitivity of the standard curve was limited to 103 copies well-1 and following optimisation of the assay the annealing temperature was decreased by 1˚C to 62°C to reduce the cross-reactivity to negligible levels, though this reduced the sensitivity of the assay even further to 104 copies well-1. The detection limits by qPCR obtained by spiking samples with known amounts of Fp were 192 CFU mg-1 from egg samples, 184 CFU mg-1 from fry tissue samples, and 220 CFU ml-1 from water samples,. The sensitivity of the LAMP assay determined by spiking egg, kidney, spleen and water samples was 18 CFU mg-1, 22 CFU mg-1, 25 CFU mg-1 and 16 CFU ml-1, respectively. The latter was similar to, though not as sensitive as nested PCR. Nested PCR limits determined by spiking egg, kidney, spleen and water samples were 14 CFU mg-1, 11 CFU mg-1, 13 CFU mg-1 and 11 CFU ml-1. No cross-reactivity was found with any bacteria including other Flavobacterium species with nested PCR but cross-reactivity with other Flavobacterium species were found with both qPCR (1.51 % with Flavobacterium aquatile and 0.30 % with Flavobacterium johnsoniae) and LAMP. The LAMP assay showed slight cross-reactivity with Flavobacterium columnare and Flavobacterium branchiophilum. A novel Luminex™ assay was also developed and optimised, using microspheres coated with Fp, to detect antibodies to Fp in the serum of the fry. The Luminex™ allowed small volumes of serum from individual fry to be used to evaluate antibody response as an indirect method to determine exposure to and infection by Fp. A large number of fry from all 4 batches (88% - 94%) of eggs were found to contain anti-Fp antibodies though it still remains to be determined whether the antibodies were specific to Fp. From the work carried out in this study, it can be concluded that whether eggs are carrying Fp inside and/or on their surface, it should be possible to reduce the prevalence of Fp in farm systems by regularly screening the broodstock, eggs and fry. Supply of Fp-free eggs and milt is essential to reduce the reservoir of Fp on farms. Both the qPCR and LAMP assay require further optimisation but they do offer potential for the future screening of Fp at farm sites and in the laboratory. Future control measures for RFTS should include the screening of broodstock and eggs by sensitive methods so that Fp-free seed can be supplied to farms. This, alongside effective disinfection procedures, rigorous husbandry practices and future vaccine development will all be required to manage this very significant fish disease.