Pan-genome analysis of the fish pathogen Aeromonas salmonicida with focus on genetic determinants for host specificity.

Salmon lice are one of the most important threats in Norwegian salmonid aquaculture and cleaner-fish of several species are used for biological control of salmon lice. The health situation in cleaner fish species is both alarming and unsustainable, with almost total loss of cleaner-fish during the salmon culture cycle being relatively common. Infection with the bacterium Aeromonas salmonicida represents one of the most important causes of death among cleaner-fish in salmon farms. This bacterium is known to cause disease in many different fresh- and saltwater fish species worldwide. Historically, focus has however been upon strains belonging to the sub-species A. salmonicida subsp. salmonicida (also known as 'typical' A. salmonicida), which primarily infects salmonid fish. Other fish species, such as cleaner fish, are more prone to infection with other A. salmonicida sub-species/-types (collectively referred to as 'atypical' A. salmonicida), and this highly diverse group of bacteria remains very poorly described and systematized. This has e.g. hampered vaccine development and made reproducible studies on 'atypical' A. salmonicida difficult to conduct. A previous study carried out by our group found that A. salmonicida ('typical' and 'atypical') can be divided into different sub-types (A-layer types) based on sequencing of the gene encoding a protein layer (A-layer) exposed on the surface of the bacterial cell. Furthermore, each of the identified A-layer types seem to cause disease only in a few specific fish species, and primarily two such A. salmonicida A-layer types are found in cleaner fish, both in Norway and the British Isles. By expanding the collection spectrum of A-layer typed A. salmonicida isolates further with regard to origin (host, geography time), we sought to enable selection ofcan make a globally representative sub-selectioncollection of isolates for whole genome sequencing (WGS), with. Ssubsequent bioinformatics analysis towill illuminate the overall population structure of this bacterium., while p Possibly, this could also facilitateing a better understanding of factors underlying the observed host specificity. Knowledge arising from this project could be potentially valuable e.g. for optimization of mitigating measures against A. salmonicida in farmed fish, including cleaner fish species. As of October 2018, our A. salmonicida collection has been extended to now include roughly 700 isolates, recovered from 50 different fish species in 26 countries. A-layer typing performed on this extensive collection has identified over twenty A-layer types, and strongly support the theory of host specificity among subtypes of A. salmonicida. A publication on this topic is currently being prepared. Furthermore, a selection of representative isolates from each A-layer type have now been subjected to whole-genome sequencing, and we now have just under 200 assembled A. salmonicida genomes available from bioinformatics investigation, which is ongoing. Through the project period and up to its end, we have managed to assemble a large collection of approximately 700 A. salmonicida isolates from 50 different fish species in 26 countries, which were all A-layer typed. These could be divided amongst 23 distinct A-layer types, most of which were only recovered from a limited number of fish species. This work may prove highly relevant for future efforts to produce vaccines against this bacterium. Moreover, representative isolates for each of the respective A-layer types were selected for whole genome sequencing and subsequent bioinformatics analysis. Genomes from a total of 196 A. salmonicida isolates were included in the final dataset, thus forming the basis for the most extensive population study that has been conducted on this bacterium to date. From this it becomes immediately clear that the current systematic subdivision of this species is inexpedient and should be revised. Further, the results e.g. provide insights into how some A. salmonicida strains may have become spread around the world through transport of cultivated fish, while others presumably remain geographically confined together with their natural wild-living fish hosts. On mobile genetic elements, we identify and describe genes of importance for the bacterium?s ability to cause disease, and we find that strains which have been in close contact with anthropogenic activity comprise a much higher number of genes conferring antibiotic resistance.

A-lags typing, i kombinasjon med det globalt dekkende datasettet med A-lags typede Aeromonas salmonicida isolater publisert gjennom prosjektet, benyttes i dag av Veterinærinstituttet samt kommersielle aktører (f.eks. vaksineprodusenter og diagnostiske laber) i Norge, og trolig utenlands, til å karakterisere denne bakterien. De 151 sekvenserte genomene vil (ved publisering av artikkel) tilgjengeliggjøres gjennom en offentlig internasjonal database. Den enorme mengden informasjon som ligger i dette datasettet vil kunne vise seg svært verdifull for andre forskere som jobber med denne bakterien, og vårt arbeid med det har bidratt til å heve den bioinformatiske kompetansen ved Veterinærinstituttet. Samarbeidet med verdensledende internasjonale forskere innen fagfeltet har videre gitt spin-offs i form av nyoppstartede forskningsprosjekter. MLVA genotyping metoden planlegges anvendt til å utforske epidemiologien rundt furunkulose i villaks i Norge generelt og i Namsen-vassdraget spesielt.

Salmon lice are one of the most important threats in Norwegian salmonid aquaculture and cleaner fish of several species are used for biological control of salmon lice. The health situation in cleaner fish species is both alarming and unsustainable, with almost total loss of cleaner-fish during the salmon culture cycle being relatively common. Acute bacterial disease, particularly infection with Aeromonas salmonicida, is the most common diagnostic finding. A. salmonicida is a very important and almost ubiquitous pathogen in many species of fish. Without exception, all marine fish species introduced to aquaculture in recent years in Norway, including lumpsucker, ballan wrasse, turbot, halibut, cod and wolffish, have been identified as susceptible to A. salmonicida infection. There is a need for development of prophylactic measures including vaccination against this bacterium. The species A. salmonicida comprises a wide range of genetically and phenotypically diverse strains however, and a lack for comprehensive, resolute and unambiguous subtyping systems have hampered research on the bacterium. This particularly applies to so-called ‘atypical’ A. salmoncida strains, which are highly diverse and frequently infect non-salmonid fish species, including wrasse and lumpsucker. Our group (Gulla et al. 2015) recently published the first systematic typing system for A. salmonicida, based on sequence variation in the vapA gene (coding for the extracellular A-layer in A. salmonicida). This A-layer typing system separates all five recognised subspecies, identifies several new clusters which may well represent additional, as yet undescribed subspecies, and identifies a significant degree of host specificity in many of these clusters. By further expanding the collection (with regard to origin) of A. salmonicida strains investigated by A-layer typing we will make a selection for whole genome sequencing. This will allow in-depth investigations into the pan-genome and overall population structure of this important fish pathogen, while also facilitating identification of additional elements associated with its apparent host specificity. The project may well result in spin-offs relating to vaccine optimisation.