Molecular Interaction Between Fish Pathogens and Host Aquatic Animals

Home , Aeromonas salmonicida

K. Tsukamoto, T. Kawamura, T. Takeuchi, T. D. Beard, Jr. and M. J. Kaiser, eds. Fisheries for Global Welfare and Environment, 5th World Fisheries Congress 2008, pp. 277–288. © by TERRAPUB 2008.

Molecular Interaction between Fish Pathogens and Host Aquatic Animals

Laura L. Brown* and Stewart C. Johnson

National Research Council of Canada Institute for Marine Biosciences 1411 Oxford Street Halifax, NS, B3H 3Z1, Canada

Present address: Fisheries and Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, NS, V9T 6N7, Canada

*E-mail: [email protected]

We have studied the host-pathogen interactions between Atlantic salmon (Salmo salar L.) and Aeromonas salmonicida. Sequencing the genome of the bacterium allowed us to investigate virulence factors and other gene products with potential as vaccines. Using knock-out mutants of A. salmonicida, we identified key virulence factors. Proteomics studies of bacterial cells grown in a variety of media as well as in an in vivo implant system revealed differential protein production and have shed new light on bacterial proteins such as superoxide dismutase, pili and flagellar proteins, type three secretion systems, and their roles in A. salmonicida pathogenicity. We constructed a whole genome DNA microarray to use in comparative genomic hybridizations (M-CGH) and bacterial gene expression studies. Carbohydrate analysis has shown the variation in LPS between strains and reveals the importance of LPS in virulence. Salmon were challenged with A. salmonicida and tissues were taken to construct suppressive subtractive hybridization libraries to investigate differential host gene expression. We constructed an Atlantic salmon cDNA microarray to investigate the host response to A. salmonicida. Real-Time qPCR and NMR-based metabolomics have revealed important information about host responses to infection and to chronic stress. By linking genome sequencing, functional genomics, proteomics, carbohydrate analysis, metabolomics, and whole animal assays, we took integrated and innovative approach to pathogenesis research.

KEYWORDS host-pathogen interactions; genomics; proteomics; furunculosis; DNA microarrays; knock-out mutants 278 L. L. BROWN AND S. C. JOHNSON

it is notable that several problems are asso- 1. Introduction ciated with commercially available vaccines for fish. Oil-based and formalin-based The development of genomics, proteomics vaccines have been associated with adhe- and metabolomics, have enabled the transi- sions and other pathologies including weight tion from sequential studies of single genes, loss (Midtlyng and Lillehaug 1998), second- proteins or metabolites to the simultaneous ary infections (Bruno and Brown 1999), in- study of many components and their inter- jury and inflammation (Mutoloki et al. actions at a variety of biological levels (from 2004). Therefore, research into vaccines and pathways to cells, tissues and whole organ- vaccine delivery systems is ongoing. isms). The development of these fields has Similarly, effective husbandry tools and been supported by the concurrent develop- practices have been used since the inception ment of new technologies and methods such of aquaculture. Fallowing farm sites, sepa- as quantitative PCR or Real Time PCR, RNA ration of year classes, culling of infected interference (RNAi) assays, and in situ hy- broodstock, and reduction of anthropogenic bridization. These fields and their support- stressors are all examples of successful hus- ing technologies and methods are now bandry techniques. However, these are widely used in the study of disease of higher largely based on trial and error and on em- vertebrates as well as in the development of pirical observations. new therapeutants and vaccines. The field In order to devise effective aquatic ani- of aquatic animal health has only recently mal health management tools in a timely and begun to benefit from their application. cost-effective fashion, it is essential that we It is well known that farmed aquatic ani- identify meaningful biological targets, by mals are susceptible to a wide variety of in- knowing which antigens, host immune fac- fectious diseases. Direct and indirect costs tors, and stress response genes to target. In of preventing and controlling infectious dis- order to do that, it is necessary to increase eases in aquaculture are estimated to exceed our understanding of the interactions be- 10% of total production costs. This repre- tween host aquatic animals and pathogens. sents an appreciable loss of profits to the Genomics, proteomics, and other biotechno- aquaculture industry. Coupled with this is an logical approaches help us to do that. increasing global concern about the use of Aeromonas salmonicida is a non-motile, therapeutants (antibiotics and chemical treat- Gram-negative bacterium that is the ments) in the production of food animals and etiological agent of furunculosis. Furuncu- issues related to animal welfare. Develop- losis has been recognized as a widespread ment of more effective vaccines for aqua- important disease in wild and farmed fish culture is necessary to reduce losses due to since the 1890s (Emmerich and Weibel 1894). disease, to enable further reductions in the Due to its early recognition as an impor- use of therapeutants, and to improve animal tant pathogen it is arguably the best studied welfare. bacterial pathogen of fish. Several excellent Vaccines against infectious diseases in reviews of earlier research on A. salmonicida fish have been effective and available for and furunculosis are available (Bernoth years, For example, multi-valent vaccines 1997; Wiklund and Dalsgaard 1998; Hiney based on whole bacterins in formalin- or oil- and Olivier 1999). based preparations have provided protection In this chapter we will demonstrate how to salmon against Aeromonas salmonicida, a multi-disciplinary approach, combining Listonella anguillarum. Vaccines are also genomics, proteomics and metabolomics available against viral pathogens. However, with more traditional methods was used Molecular interaction between fish pathogens and host aquatic animals 279 within the National Research Council in infectious disease research. Genomic Genomics and Health Initiative Program sequencing has facilitated the development “Genomics Approaches to Aquatic Animal of new methods of pathogen detection, our Health Management,” to understand the bi- understanding of genetic relationships be- ology of the bacterial pathogen Aeromonas tween species and investigations into the salmonicida and its interactions with one of nature of host-pathogen interactions. its hosts, the Atlantic salmon (Salmo salar Through the use of bioinformatics methods, Linnaeus). We will also show how this ap- which allow for genomic comparisons be- proach was used to identify candidate pro- tween species, putative virulence factors, as tein and carbohydrate antigens for use in well as genes that encode proteins that may vaccines, in conjunction with whole animal be suitable for use in vaccines can be identi- assays. fied. Genomic sequencing also facilitates the use of other technologies such as the devel- 2. Development of an in-vivo Growth opment of knock-out mutants, and other System for Aeromonas salmonicida genomics (e.g., microarray), and proteomic studies, all of which serve to increase our It is well known that the conditions under understanding of host-pathogen interactions which the bacterium are grown influence and aid in the rational design of effective their gene expression and ultimately their vaccines phenotype (Dalsgaard et al. 1998). This was At the initiation of our research program originally demonstrated for A. salmonicida there was little genomic information avail- grown within chambers in the host by Garduño able for A. salmonicida or closely related et al. (1993). These authors demonstrated species. In order to understand the infection that cells grown under in vivo growth condi- process and possible interactions between tions had increased resistance to bacterioly- A. salmonicida and its hosts, as well as to sis, phagocytosis and oxidative killing. We aid in the rational design of new vaccines developed a method for the in vivo growth the whole genome of wild-type strain A449 of large volumes of A. salmonicida (Dacanay was sequenced and assembled (GenBank Ac- et al. 2003). Briefly this method involves the cession #s NC_009348-009350). The A449 use of growth chambers made of autoclaved genome consists of a circular chromosome 12 to 14-kDa molecular-mass-cutoff dialy- and five plasmids that encode more than sis tubes that are filled with A. salmonicida 4700 genes (Boyd et al. 2003; Reith et al. suspensions and surgically implanted into the unpublished). During this program a full abdominal cavity of Atlantic salmon for a genomic sequence became available for period of 24 hours, then harvested. Bacteria Aeromonas hydrophila (Seshadri et al. grown within these implants have been used 2006). Comparison of the Aeromonas in studies of gene expression, for proteomics, salmonicida genome with genomes of related carbohydrate analysis and for studying the organisms such as A. hydrophila enabled the host response to infection as described be- identification of large numbers of genes en- low. coding potential virulence factors (Boyd et al. 2003; Nash et al. 2006; Reith et al. un- 3. Development of Genomics Resources published). Virulence genes were targeted for and Tools for Aeromonas salmonicida future study and for use as vaccine candi- subsp. salmonicida (A449) dates. Table 1 summarizes some of the results of the projects described above. The acquisition of full genomic sequences Genome arrays which contain large for pathogens has facilitated rapid advances numbers of genes or whole genomes have 280 L. L. BROWN AND S. C. JOHNSON

2003

2006

2008

et al.

npublished

Reference

Dacanay (unpublished) Dacanay (unpublished)

Unpublished

Dacanay

e also elucidated

tered macrophage

— Increased enzyme levels in virulent strains — Increased enzyme levels in virulent strains —

—

No change Boyd

cytokine profiles

mmersion

No change in virulence

Reduced Reduced virulence virulence Reduced virulence

No change in virulence

——

Reduced virulence No change in virulence

No change in virulence

In vivovivo Ex

Injection I

Reduced virulence Avirulent Avirulent

in virulence

No change in virulence No change in virulence No change in virulence

No change in virulence

in virulence

in virulence Avirulent Avirulent Al

Reduced virulence

genes identified via genomic sequencing and knock-out mutants. W genes identified via genomic

published

Secreted effector

Anguibactin-like Secreted effectorsiderophore receptor change No

Superoxide anion Haemoprotein detoxification receptor Superoxide anion detoxification Hydrogen peroxide detoxification

Amonabactin-like siderophore receptor

Type I pilus subunit, tip and usher proteins

Type IV pilus subunitchange No

TTSS secretome protein Secreted effector

Aeromonas salmonicida

Gene(s) Role

aopH

fstB aopO

sodA hupA sodB

katA

fstC

fimABCD

No change in virulence

tapA

flp1

ascC

aexT

Summary of

Table 1. Table the role of the expressed proteins in virulence proteins of the expressed the role

System

Iron sequestration

Oxidative stress

Type I pili

Reduced covert infection? Reduced adherence Un

Type IV pili

Type III Secretion (TTSS) Molecular interaction between fish pathogens and host aquatic animals 281 been used to study genomic variability was confirmed with knock-out mutants as among strains of a variety of bacterial patho- described below. gens through a process known as microarray- based comparative genomic hybridization 4. Transcriptional, Proteomics and Bio- (M-CGH) (Ong et al. 2004; Taboada et al. chemical Responses of A. salmonicida 2004). From the genomic sequence Nash et Grown under Selected Conditions: al. (2006) developed a DNA microarray of Understanding Virulence Mecha- 2024 genes for A. salmonicida subsp. nisms of A. salmonicida Using Mutants salmonicida and used M-CGH to compare a and Live Challenges variety of A. salmonicida subspecies, strains and isolates obtained from different hosts and In order to investigate effects of host re- geographical locations. Using this technique sponses and other growth conditions on these authors were able to show variable car- A. salmonicida we conducted studies of gene riage of the virulence-associated genes and expression and protein expression, as well generally increased variation in gene content as carbohydrate analysis. These studies ex- across sub-species and species boundaries. amined how the phenotype of A. salmonicida Variable genes included those encoding outer changed in response to the host factors (in membrane proteins, pili, and flagella. Their vivo growth condition) by comparing in vivo- data showed little correlation between geo- grown bacteria to bacteria grown under vari- graphic region and degree of variation for able in vitro conditions such as high tempera- all isolates tested. The greatest variation was tures, and within iron-restricted media. Such observed among genes associated with data can aid understanding the host-patho- plasmids and transposons. However, many gen interaction as well as in the selection and known virulence genes were common to all development of antigens for use in vaccines. strains tested, including superoxide These data also highlight that interpretation dismutase, and those associated with iron- of data from studies of expression single restriction and sequestration. genes or gene sets, proteins or carbohydrates Wang et al. (2007) examined the LPS must be taken within the context of the strain, structures of the same strains and isolates that subspecies, or isolate used. were used by Nash et al. (2006) in their For example, the A. salmonicida ge- M-CGH studies. The carbohydrate analysis nomic DNA microarray was used to exam- showed three distinct structural types. The ine changes in transcription levels and separation of the isolates into the three LPS proteomics was used to examine changes in types corresponded to the clustering of the the translation levels for A. salmonicida in gene variation seen in the M-CGH studies response to iron-restricted in vitro and in vivo (Nash et al. 2006). Wang et al. (2007) specu- growth conditions (Brown et al. unpub- lated that an LPS-based classification sys- lished). Based upon this work it was possi- tem of A. salmonicida sub-species consist- ble to identify genes that are differentially ing of two serologically distinct types: type expressed when the organism is under these A and non-type A, could be developed. two conditions. Not surprisingly many of the Identification of common proteins, car- genes that were identified as differentially bohydrates, and other antigens is an impor- expressed were involved in the utilization of tant first step in identifying genes important iron or heme. However, there were several in virulence, as well as vaccine candidates genes which were upregulated only in vivo, that are conserved across all strains known including genes homologous to an ABC-type to cause disease. In addition the importance galactoside transporter and components of of some of these genes as virulence factors the lateral flagellar system. Comparison of 282 L. L. BROWN AND S. C. JOHNSON the expression data to proteomic data ob- regulation of two iron regulation systems tained from the same samples showed that (ferric siderophore and heme acquisition sys- for many of the genes and proteins there was tems). Based on their results these authors a strong positive correlation between gene felt that with respect to the OMPs that growth expression level and protein expression, thus in low iron media was a good model system confirming the conclusions on the roles of for in vivo growth. Microarray analysis of the proteins and genes in virulence and mRNA expressed in bacteria grown under pathogenicity. These results demonstrate the same conditions confirmed the results how microarray analysis of A. salmonicida obtained in the proteomics experiments de- under varied growth conditions can be used scribed above (Brown et al. unpublished). to determine how different host factors in- As these proteins were also seen to be ex- fluence gene expression, as well as empha- pressed in avirulent strains of A. salmonicida size the importance that interpretation of data under these conditions it is suggested that from studies of expression single genes or they may be necessary but not sufficient for gene sets, proteins or carbohydrates must be bacterial virulence (Ebanks et al. 2004). taken within the context of the growth con- Proteomics and RT-qPCR were also used to ditions. study the type III secretion system of A. Within our research program two dimen- salmonicida grown under selected conditions sional gel electrophoresis (2DE)-based (Ebanks et al. 2006). In that study a variety proteomics methods were used to identify of strains of A. salmonicida were grown in outer membrane proteins (OMP) of A. vitro at 16, 17 and 28°C, as well as in vivo. salmonicida, as well as to describe changes These authors demonstrated the increased in the OMP profile resulting from in vitro expression, as well as assembly of, a func- culture in low iron media and in vivo culture tionally competent type III secretion system (Ebanks et al. 2004, 2005). Ebanks et al. in cells grown at 28°C and in the presence of (2005) were able to identify 76 proteins from 0.19 to 0.39 M NaCl. Cells that were grown a carbonate-enriched A. salmonicida mem- in vivo, in low iron media, at low pH or at brane preparation which corresponded to high cell densities did not show increased approximately 60% of all of the protein spots expression of the type III secretion system. that could be visualized by 2DE (pI 4–7). Proteins are not the only virulence fac- Within these were a number of proteins such tors tested within the program. Wang et al. as endolase, which due to their lack of a clas- (2004) compared the structures of the sical export sorting signals were not pre- capsular polysaccharide and lipopolysaccha- dicted to be OMPs, i.e., based on motif se- ride O-antigen of A. salmonicida grown on quence. Ebanks et al. (2005) determined that TBS and for 72 hours in vivo. This study they were in fact, OMPs, based on protein demonstrated the presence of capsular analysis. Being able to identify these pro- polysaccharide and novel LPS O-chain teins within the OMPs of A. salmonicida polysaccharide in cells grown in vivo when demonstrates the value of combining ge- compared to the in vitro grown bacterial in- nomic and proteomic approaches. In another ocula. These authors suggested that the use study in our research program, outer mem- of in vivo-cultured cells for isolation and brane protein profiles of A. salmonicida were structural analysis of A. salmonicida compared between in vitro-grown bacteria polysaccharides leads to the development of and bacteria grown in low iron media and in more meaningful biological data. vivo (Ebanks et al. 2004). With respect to The conclusions of Ebanks et al. (2004), the OMPs, growth under both iron-restricted based on proteomics data, were supported conditions and in vivo caused a marked up- by the knock-out mutant results of Dacanay Molecular interaction between fish pathogens and host aquatic animals 283 et al. (2006). Those authors examined the When challenged by injection there was no contribution of the type III secretion system significant change in virulence when com- (TTSS) to virulence of A. salmonicida. The pared to wild type, however the sodA and TTSS consists of an injection system (mem- sodB mutants when used to challenge fish brane proteins and a needle like structure) by immersion had reduced virulence (Table 1). and a number of effector proteins that are More recently, Boyd et al. (2008) exam- transmitted into host cells, leading to modu- ined the contribution of Type IV pili to the lation of components of both the innate and virulence of A. salmonicida. Pili are struc- adaptive immune response that are advan- tures that allow bacteria to attach to surfaces tages to the bacterium. In their study dele- such as host tissues and for this reason are tion mutant strains in the genes of the outer important virulence factors for many patho- bacterial transmembrane pore (ΔascC) and genic bacteria. In this work the authors con- three effector genes (ΔaexT, ΔaopH, ΔaopO) structed three mutant strains that were defi- were generated and the virulence of these cient in: a gene involved in encoding Tap pili strains tested by challenge of Atlantic (tapA), a gene involved in encoding FLP pili salmon. These authors demonstrated that (flpA) and a double (flpA-tapA) mutant. Im- the TTSS was essential for virulence of mersion challenge with the tapA and flpA- A. salmonicida as the ΔascC mutant strain tapA mutants resulted in reduced cumulative was avirulent when challenged by both im- mortality when compared to wild type. How- mersion and intraperitoneal (i.p.) injection. ever, when challenged i.p. there was no sig- None of the three effector mutants showed nificant difference between these mutants significantly decrease virulence when com- and the wild type. In contrast the Flp pili pared to wild type in i.p. challenges. Inter- appear to contribute little to virulence as estingly, significantly lower morbidity of two immersion and i.p. challenge with the flapA of the mutant strains (ΔaexT and ΔaopH) mutant resulted in mortalities that were not when compared to wild type in immersion significantly different when compared to the trials suggested a possible role of these ef- wild-type (Boyd et al. 2008). The authors fectors in host colonization. Furthermore fish concluded that pili were essential factors of that were i.p. challenged with the ΔascC the initial adhesion and colonization processes. mutant were not protected upon re-challenge Results within our research program with the wild type strain. The authors con- showed that when determining the role of cluded that the type III secretion system is putative virulence factors of pathogens it is absolutely essential for virulence of the bac- equally important to test various methods of terium, and that removal of individual effec- presentation of the bacterial antigens, as seen tors has little effect on virulence but has a by the studies cited above. significant effect on colonization. Knock-out mutants were also used to As described above for the type III se- investigate pathways involved in host re- cretion systems, using information provided sponses. For example, Fast et al. (2008, and by the genomic sequence a number of mu- unpublished data) using enriched cultures of tant strains of A. salmonicida have been con- head kidney macrophages exposed to the all structed and tested for virulence by live ani- of the type III secretion mutants, showed the mal challenge (Table 1). For example, same pattern of expression of Interleukin knock-out mutants for genes (sodA, sodB and (IL)-8 and Tumour Necrosis Factor (TNF) katA) involved in protection against when compared to cells exposed to wild oxidative stress have been created and used type (parent) strain of A. salmonicida. With in challenge trials with Atlantic salmon (J. M. respect to expression of IL-10 and IL-12 Boyd and A. Dacanay, unpublished data). the ΔascC mutant strain (avirulent when 284 L. L. BROWN AND S. C. JOHNSON challenged by both immersion and i.p. in- subtracted against infected tissue, as well as jection) did not stimulate their expression. the reciprocal subtractions, i.e., infected tis- Exposure to the mutant Aop3Δ (lacking 3 of sue was subtracted against that of the con- the type III secretory system effectors) only trol fish. Tissues were kept separate, i.e., dis- stimulated IL-12 expression. Taken together tinct libraries were each created for liver, these data suggest that the type III secretion head kidney, and spleen. This resulted in the system is not only essential for the virulence identification of 1778 expressed sequence of the bacterium, but components of the sys- tags (ESTs) which were used to create a cus- tem are also important for stimulation of host tom cDNA microarray that has been used to immune responses. IL-12 is an important describe the transcriptional responses of At- cytokine released by infected macrophages, lantic salmon following cohabitation chal- which stimulate T-helper type I cells driving lenge and Atlantic salmon macropahges ex- towards cell-mediated immunity. As AscCΔ posed in vitro to in vivo and in vitro cultured does not stimulate this cytokine, it may ex- A. salmonicida (Douglas et al. 2003; Ewart plain why infection with this knock-out does et al. 2005, 2007). Ewart et al. (2005) iden- not provide protection from future A449 in- tified differentially expressed genes in head fection as immunological memory has not kidney, spleen and liver samples obtained been triggered by the initial exposure. Fur- from fish after 7 and 13 days of co-habita- thermore, AscCΔ does not survive/replicate tion challenge. Using this microarray and RT- within macrophages as well as A449 (80% qPCR validation they were able to identify reduced). in addition to known immune-related genes Production of IL-10 by the wild type some unusual genes that were highly up- strain may assist in prolonging infection and regulated. These genes were homologous to down regulation of pro-inflammatory pro- pufferfish plasma high-affinity saxitoxin- duction and secretion, as IL-10 is inhibitory binding protein and snake peptide neurotoxin towards TH-1 responses and generally to- (also know as differentially regulated trout wards inflammation, in mammals. Protection protein). They were also able to identify a by Aop3Δ therefore, would be expected due large number of unknown genes that were to stimulation of IL-12, but possibly in the also differentially regulated. This same absence of significant virulence (no IL-10), microarray was used to examine differences as seen in the wild type strain. in the transcriptional profile of primary isolates of Atlantic salmon macrophages in- 5. Transcriptional Responses of Atlantic fected with in vivo and in vitro cultured Salmon to Infection with A. salmonicida A. salmonicida (Ewart et al. 2007). and Chronic Stress As mentioned previously there are significant and biologically meaningful differ- At the beginning of this research program ences in the phenotype of A. salmonicida there were relatively few genomic sequences culture under these two conditions (Ebanks available for Atlantic salmon. In order to et al. 2004; Dacanay et al. 2006; Brown et characterize genes involved in the immune al. unpublished). Examination of gene tran- response of Atlantic salmon to infection with scription patterns at 0.5, 1 and 2 hours post- A. salmonicida reciprocal suppression infection revealed commonalities, as well as subtractive hybridization (SSH) six cDNA significant differences in gene expression libraries were constructed for liver, head kid- patterns between macrophages exposed to ney and spleen tissues following i.p. chal- in vivo and in vitro grown cells. Differences lenge (Tsoi et al. 2004). Control (uninfected, in expression of genes such as JunB and injected i.p. with saline only) cDNA was TNF-α that were evident 30 minutes after Molecular interaction between fish pathogens and host aquatic animals 285 infection led these authors to suggest that handling stress resulted in increased consti- these differences may be due to differential tutive expression of IL-β in head kidney production by the bacteria of diffusible prod- macrophages from Atlantic salmon, followed ucts (Ewart et al. 2007). As in the earlier by decreased stimulation of leukocytes by study a number of unknown genes were extracellular antigens (LPS) and decreased also identified to be differentially regulated leukocyte survival. This is important infor- making them interesting candidates for fur- mation and highlights the need for further ther study. Overall the use of microarrays studies. Aquaculture fish species can be sub- to examine the transcriptional response of jected to acute and chronic stressors and it is A. salmonicida has led to the development essential to understand the effect of those of more questions than answers. stressors on the animals’ ability to mount Fast et al. (unpublished data) isolated protective immune responses against patho- macrophages from Atlantic salmon and in- gens. If the efficacy of vaccines is compro- cubated them with the wild-type parent strain mised because of the hosts’ immune state, and with two type III secretion system knock- this will add to the cost of production to the out mutants, Aop3Δ (deficient in three of the farmer and will also lead to increased losses type III secretion system effector genes) and due to infection. If this can be avoided by AscCΔ (deficient in the gene encoding the avoiding stressors at the time of vaccination outer-membrane pore of the type III secre- (for example), this would prove to be a low- tion apparatus). The authors also incubated cost approach to aquatic animal health man- Salmon Head Kidney (SHK) cell culture agement. with the same wild-type and mutant strains. Gene expression within salmon cells was 6. Metabolomics Responses of Atlantic measured using Real-Time qPCR. Their re- Salmon to Infection with A. salmonicida sults showed A. salmonicida infection stimu- and Vaccination lates inflammatory gene expression and that expression of IL-8 and TNF are not affected Metabolomics is the “systematic study of the by mutations within the type III secretion unique chemical fingerprints that specific system. However, they did observe that the cellular processes leave behind”—specifi- outer-pore membrane secretion apparatus cally, the study of their small-molecule seems to be essential for expression of IL-12 metabolite profiles (Davis 2005). The and IL-10, and the three effector genes are metabolome represents the collection of all also essential for IL-10. metabolites in a biological organism, which While investigating the mechanisms of are the end products of its gene expression. host-pathogen interactions it is extremely Solanky et al. (2005) used NMR-based important to determine the effect of the metabolomics to show that metabolites physiological state of the host on those in- within the plasma of Atlantic could clearly teractions. The fish species involved in these separate into distinct groups: those fish that investigations are important to aquaculture, had been vaccinated with killed Aeromonas and are subjected to anthropogenic and natu- salmonicida bacterin, those that had been ral stressors, such as handling, grading, vac- challenged, and those that had survived the cinations, and fluctuations in temperature challenge. Dacanay et al. (2006) showed that and water chemistry. These and other fac- this technique can be used to investigate host tors, contribute to the physiological state of response to specific virulence factors in the animal, and can directly or indirectly af- pathogens, using plasma from fish chal- fect their ability to respond to pathogens. lenged with selected knock-out mutants, and Fast et al. (2008) showed that repeated with the wild-type parent strain. Using 286 L. L. BROWN AND S. C. JOHNSON

NMR-based metabolomics, the authors com- and homology to known genes (e.g., GO pared the plasma from vaccinated salmon annotation). If useful tools and knowledge immersion-challenged with the wild-type is to be derived from the use of these genomics parent strain with plasma from fish chal- technologies, then it must be validated by lenged with the avirulent mutant strain complementary studies of the biological AscCΔ. Their results showed that the plasma function of the genes and proteins. Knock- metabolite profiles correlated with the pro- out mutants, transformative cell lines, and tective immune response, i.e., where a pro- recombinant protein technologies are all use- tective immune response was observed in ful in this sense. Ultimately, however, whole whole animal challenge, the plasma profiles animal trials validate any assumptions made clustered separately. When there was no sig- via the genomics and proteomics approaches nificant protection, the plasma profiles were described above. We have taken this systems indistinguishable (Dacanay et al. 2006). This approach in our research program. study clearly demonstrates that metabolomics Also required is an understanding of the technologies have potential to provide non- physiological responses of the host, within invasive assays to monitor host immune re- the context of nutrition, stress, environmen- sponses. tal effects. Equally important is knowledge of the biology of the pathogen: its genetic 7. Summary and Future Directions variability, its gene expression within the context of its environment. In using these and other techniques, we have As can be seen in describing the early been able to identify virulence proteins and and then later work of our team and others, carbohydrates from Aeromonas salmonicida, these genomics and proteomics technologies and have been able to develop more effec- can also be used to push earlier studies much tive vaccine candidates against the pathogen. further. Thus the techniques will allow us to Moreover, in examining host immune re- achieve a greater understanding of the sponses at the molecular level, we have de- mechanisms of systems and processes that veloped antigen delivery strategies that tar- had been previously described at an higher get the immunomodulatory mechanisms of organismal level. the host. Our investigations into the molecular Acknowledgements mechanisms of salmonid stress-related genes The authors would like to thank all of the NRC mem- and disease susceptibility will allow us to bers of the research program team: Andrew Dacanay, develop more effective tools for husbandry. Devanand Pinto, Michael Reith, Jessica Boyd, Kelly These and other studies will also aid in tar- Soanes, Vanya Ewart, John Walter, Evelyn Soo, Luis geting specific disease-resistance markers for Afonso, Tony Manning, Sandra Sperker, Jason family selection and comprehensive breed- Williams, Dave O’Neil, Cindy Leggiadro, Sue Penny, ing programs. Roger Ebanks, Tiago Hori, Ken Chisholm, Bob However, a word of caution is war- Richards, JC Achenbach, Debbie Plouffe, Joseph Hui, ranted. The tools and technologies as de- Elizabeth Huenupi, Ian Burton, Tobias Karakach, Leah scribed above are very powerful, yet remain Knickle, Bruce Curtis, Lisa Cook, Rama Singh, Sathish only tools. Genome sequencing, cDNA Prasad, Colleen Murphy, Pauline Murphy, Mark Fast, Jane Osborne, Darren Sarty, John Nash, Eleonora microarrays, EST libraries, etc., generate Altman, Jianjun Li, Anne Bouevitch, Oksana huge data sets, usually with hundreds, if not Mykytzcuk, Simon Foote, Wendy Findlay, Jacek thousands of unknown genes and proteins. Stupak, Xin Liu, Zhan Wang, Vandana Chandan, as Currently, gene function is often tentatively well as the members of the team from Dalhousie assigned on the basis of BLAST annotation University: Rafael Garduño, Elizabeth Garduño. The Molecular interaction between fish pathogens and host aquatic animals 287 authors would also like to thank those colleagues who Dr. Mark Strom, NOAA, USA, Dr. Rocco Cipriano, kindly contributed Aeromonas salmonicida strains and USGS, Fish Health Research Program, USA, Dr. A. isolates, advice and assistance: Dr. Gilles Olivier, Fish- Chopra, Department of Microbiology, University of eries and Oceans Canada, Dr. William Davidson, Simon Texas, Dr. Steve Kaattari, Virginia Institute of Marine Fraser University, Dr. William Kay, University of Vic- Science, Dr. Kenneth Cain, University of Idaho, Mos- toria, BC, Dr. Joanne Constantine, BC Ministry of Ag- cow, Idaho, USA, Dr. Duncan Colquhoun, National riculture, Food and Fisheries, Dr. S. Peter Howard, Veterinary Institute, Norway, Dr. Vera Lund, University of Regina, Saskatchewan, Dr. Oksana Fiskeriforskning, Norway, Dr. Bjarnheidur Bjornsdottir, Yarosh, Canadian Food Inspection Agency, Ottawa, University of Iceland, Reykjavik, Iceland, Dr. Ranneveig Dr. David Sims, University of Prince Edward Island, Bjornsdottir, University of Akureyri, Iceland.

References

Bernoth E-M. Furunculosis: the history of the diseases and of disease research. In: Bernoth E-M, Ellis AE, Midtlyng PJ, Olivier G, Smith P (eds). Furunculosis—Multidisciplinary Fish Disease Research. Academic Press, London. 1997; 1–20. Boyd J, Williams J, Curtis B, Kozera C, Singh R, Reith M. Three small, cryptic plasmids from Aeromonas salmonicida subsp. salmonicida A449. Plasmid 2003; 50(2): 131–144. Boyd JM, Dacanay A, Knickle LC, Touhami A, Brown LL, Jericho MH, Johnson SC, Reith M. Contribution of type IV pili to the virulence of Aeromonas salmonicida subsp. salmonicida in Atlantic salmon (Salmo salar L.). Infection & Immunity 2008; 76: 1445–1455. Bruno DW, Brown LL. The occurrence of Renibacterium salmoninarum within vaccine adhesion components from Atlantic salmon, Salmo salar L. and coho salmon, Oncorhynchus kisutch Walbaum. Aquaculture 1999; 170: 1–5. Dacanay A, Johnson SC, Bjornsdottir R, Reith M, Ross, NW, Singh R, Brown LL. Molecular cloning and quantitative analysis of iron and manganese co-factored superoxide dismutases in Aeromonas salmonicida subsp. salmonicida: re-evaluation of their role in virulence. J. Bacteriol. 2003; 185: 4336– 4344. Dacanay A, Knickle L, Solanky KS, Boyd JM, Walter JA, Brown LL, Johnson SC, Reith M. Contribution of the type three secretion system (TTSS) to virulence of Aeromonas salmonicida subsp. salmonicida. Micro- biology 2006; 152: 1847–1856. Dalsgaard I, Gudmundsdottir BK, Helgason S, Hoie S, Thoresen IF, Wichardt, UP, Wiklund T. Identification of atypical Aeromonas salmonicida: inter-laboratory evaluation and harmonization of methods. J. Appl. Microbiol. 1998; 84: 999–1006. Davis B. Growing pains for metabolomics. The Scientist 2005; 19: 25–28. Douglas SE, Tsoi SCM, Penny S, Melville K, Reith ME, Ewart KV. Expressed sequence tags—a snapshot of the fish genome. In: Shimizu N, Aoki T, Hirono I, Takashima F (eds). Aquatic Genomics: Steps towards a Great Future. Springer, Berlin. 2003; 115–127. Ebanks RO, Dacanay A, Goguen M, Pinto DM, Ross NW. Differential proteomic analysis of Aeromonas salmonicida outer membrane protein in response to low iron and in vivo growth conditions. Proteomics 2004; 4: 1074—1085. Ebanks RO, Goguen M, McKinnon S, Pinto DM, Ross NW. Identification of the major outer membrane proteins of Aeromonas salmonicida. Diseases of Aquatic Organisms 2005; 68: 29–38. Ebanks RO, Knickle LC, Goguen M, Boyd JM, Pinto DM, Reith M, Ross NW. Expression of and secretion through the Aeromonas salmonicida type III secretion system. Microbiology 2006; 152: 1275–1286. Emmerich R, Weibel E. Über eine durch Bakterien erzengte Seuche unter den Forellen. Archives für Hygiene und Bakteriologie 1894; 21; 1–21. Ewart KV, Belanger JC, Williams J, Karakach T, Penny S, Tsoi SCM, Richards RC, Douglas SE. Identification of genes differentially expressed in Atlantic salmon (Salmo salar) in response to infection by Aeromonas salmonicida using cDNA microarray technology. Dev. Comp. Immunol. 2005; 29: 333–347. Ewart KV, Williams J, Richards RC, Gallant JW, Melville K, Douglas SE. The early response of Atlantic salmon (Salmo salar) macrophages exposed in vitro to Aeromonas salmonicida cultured in broth and in fish. Developmental and Comparative Immunology 2007; doi:10:1016/j.dci.2007.07.005. 288 L. L. BROWN AND S. C. JOHNSON

Fast MD, Hosoya S, Johnson SC, Afonso LOB. Cortisol response and immune-related effects of Atlantic salmon (Salmo salar Linnaeus) subjected to short- and long-term stress. Fish Shellfish Immunol. 2008; 24: 194–204. Garduño RA, Thornton JC, Kay WW. Fate of the fish pathogen Aeromonas salmonicida in the peritoneal cavity of rainbow trout. Can. J. Microbiol. 1993; 39: 1051–1058. Hiney M, Olivier G. Furunculosis (Aeromonas salmonicida). In: Woo PTK, Bruno DW (eds). Fish Diseases and Disorders III: Viral, Bacterial and Fungal Infections. CAB Publishing, Oxon, England. 1999; 341– 425. Midtlyng PJ, Lillehaug A. Growth of Atlantic salmon Salmo salar after intraperitoneal administration of vaccines containing adjuvants. Diseases of Aquatic Organisms 1998; 32: 91–97. Mutoloki S, Alexandersen S, Evensen Ø. Sequential study of antigen persistence and concomitant inflammatory reactions relative to side-effects and growth of Atlantic salmon (Salmo salar L.) following intraperitoneal injection with oil-adjuvanted vaccines. Fish Shellfish Immunol. 2004; 16: 633–644. Nash JHE, Findlay WA, Luebbert CC, Mykytczuk OL, Foote SJ, Taboada EN, Carrillo CD, Boyd JM, Colquhoun DJ, Reith ME, Brown LL. Comparative genomics profiling of clinical isolates of Aeromonas salmonicida using DNA microarrays. BMC Genomics 2006; 7: 43. Ong C, Ooi CH, Wang D, Chong H, Ng KC, Rodrigues F, Lee MA, Tan P. Patterns of large-scale genomic variation in virulent and avirulent Burkholderia species. Gen. Res. 2004; 14: 2295–2307. Seshadri R, Joseph SW, Chopra AK, Sha J, Shaw J, Graf J, Haft D, Wu M, Ren Q, Rosovitz MJ, Madupu R, Tallon L, Kim M, Jin S, Vuong H, Stine OC, Ali A, Horneman AJ, Heidelberg JF. Genome sequence of Aeromonas hydrophila ATCC 7966: Jack of all trades. J. Bacteriol. 2006; 188: 8272–8282. Solanky KS, Burton IW, MacKinnon SL, Walter JA, Dacanay A. Metabolic changes in Atlantic salmon exposed to Aeromonas salmonicida detected by 1H-nuclear magnetic resonance spectroscopy of plasma. Diseases of Aquatic Organisms 2005; 65(2): 107–114. Taboada EN, Acedillo RR, Carrillo CD, Findlay WA, Medeiros DT, Mykytczuk OL, Roberts MJ, Valencia CA, Farber JM, Nash JHE. Large-scale comparative genomics meta-analysis of Campylobacter jejuni isolates reveals low level of genome plasticity. J. Clinic. Microbiol. 2004; 42: 4566–4576. Tsoi SCM, Ewart KV, Penny S, Melville K, Liebscher RS, Brown LL, Douglas SE. Identification of immune- relevant genes from Atlantic salmon using suppression subtractive hybridization. Marine Biotechnol. 2004; 6: 199–214. Wang Z, Larocque S, Vinogradov E, Brisson J-R, Dacanay A, Greenwell, M, Brown, LL, Li J, Altman E. Struc- tural studies of the capsular polysaccharide and lipopolysaccharide O-antigen of Aeromonas salmonicida strain 80204-1 produced under in vitro and in vivo growth conditions. Eur. J. Biochem. 2004; 271: 4507–4516. Wang Z, Liu X, Dacanay A, Harrison BA, Fast M, Colquhoun DJ, Lund V, Brown LL, Li J, Altman E. Carbohy- drate analysis and serological classification of typical and atypical isolates of Aeromonas salmonicida: a rationale for the lipopolysaccharide-based classification of A. salmonicida. Fish Shellfish Immunol. 2007; 23: 1095–1106. Wiklund T, Dalsgaard I. Occurrence and significance of atypical Aeromonas salmonicida in non-salmonid and salmonid fish species: a review. Diseases of Aquatic Organisms 1998; 32: 49–69.