Nutrigenomics and immune function in fish: new insights from omics technologies

1Both energy and nutrients consumed with food are essential for maintaining optimal immune function. Without adequate nutrition, the immune system is deprived of the resources that are needed to defend the host against bacteria, viruses and parasites. According to the epidemiological and clinical studies, nutritional deficiencies substantially alter immunocompetence and increase the risk of infection. Human malnutrition is usually a complex syndrome of multiple nutrient deficiencies, caused by insufficient intakes of energy, macronutrients and/or micronutrients. However, even a single nutrient deficiency (such as lack of specific vitamins, minerals or trace elements) may have detrimental effects on functioning of immune system, as demonstrated in laboratory rodents during the feeding trials. Understanding the impacts of adequate nutrition on immunity is important not only for human health, but also for the increasing number of animals maintained on formulated diets.

2Recent interest in fish nutrition has been fuelled by the rapid expansion of aquaculture industry, with worldwide fish production continuing to increase at approximately 5% per year. Despite nearly 370 fish species (including hybrids) being registered by FAO as cultured commercially, diet formulations for many of these species are based on limited information about their nutrient requirements. Farmed fish are also commonly subjected to short-term food deprivation (fasting) as part of a seasonal feeding pattern and in response to overproduction or disease outbreaks, yet fasting-induced impacts on fish immunity are largely unknown and differ between species. Further complexity is added by the global decrease in the availability of high quality marine ingredients for aquaculture feeds, such as fish meal as a protein source and fish oil as a lipid source. These two ingredients are of great importance as they supply essential amino acids (such as lysine and methionine) that are often deficient in plant proteins and fatty acids (eicosapentaenoic acid (EPA) and docosahexanoic acid (DHA)) that are not found in vegetable oils. With the wild fish supplies being unable to meet the demand from aquaculture, farmed fish are increasingly exposed to dietary plant materials, often without a comprehensive understanding of their impacts on fish health and ability to resist pathogens.

3Fish have considerably higher exposure to pathogens than non-aquatic vertebrates, with typically a million of bacteria and 10 million of viruses per millilitre of seawater, including both pathogenic and non-pathogenic forms. The pathogen exposure in fish starts immediately after hatching from their protective chorions and is further enhanced during the mouth and gut opening stages and at the onset of exogenous feeding. Furthermore, many fish species are exposed to different and unfamiliar pathogens when they switch between fresh and salt water environments. Evidence is also growing that some fish, including non-migratory species, are being exposed to novel pathogens as a result of climate change. Finally, outbreaks of fish diseases commonly occur when fish are stressed due to a variety of factors associated with the aquaculture environment and management procedures such as high stocking densities, transport or handling. Such stress may induce a variety of physiological responses grouped broadly as primary (release of adrenal catecholamines and corticosteroids) and secondary, which include changes in energy metabolism and hydromineral balance, coupled with alterations in cardiovascular, respiratory and immune functions. Both primary and secondary stress responses may then contribute to the changes in whole-animal performance (called tertiary stress responses), most likely by redirecting energy and other resources from one set of physiological processes to another, resulting for example in impaired resistance to disease and enhanced pathology during infection. Among the most common causative agents of infectious diseases in aquaculture are bacteria (54.9%), followed by viruses (22.6%), parasites (19.4%) and fungi (3.1%). Because the ability of fish to resist pathogens and cope with stress depends to a large extent on their nutritional status, the need for dietary interventions that would improve fish health has become widely recognised as central to sustainable aquaculture and the future of industry.

4In fish as in other vertebrates, immunity is typically divided into two distinct components: the innate immune response and the adaptive immune response. Innate immunity is the first line of defence against infection and it includes both physical barriers as well as humoral and cellular responses. The adaptive immune response also relies on humoral and cellular mechanisms and is characterised by a specific antigen recognition that drives a stronger and faster secondary pathogen-specific immune response. Recent advances in general and fish immunology have demonstrated that many of the cells and molecules considered unique to either the innate or adaptive systems play specific roles in both of them, making the cross-talk between the innate and adaptive systems more complex than previously thought. The cellular components of fish immunity include T and B lymphocytes, natural killer cells, monocytes, macrophages, neutrophils, eosinophils, mast cells and thrombocytes as well as dendritic-like cells described recently in Atlantic salmon (Salmo salar), rainbow trout (Oncorhynchus mykiss) and zebrafish (Danio rerio). While in some tissues (such as gonads) the population of leukocytes is relatively sparse, substantially more immune cells are found in mucosal tissues such as gut, gills, and skin. Consequently, these structures are classified as gut-associated lymphoid tissue (GALT), gill-associated lymphoid tissue (GIALT) and skin-associated lymphoid tissue (SALT).

5The mucosal layers of GALT, GIALT and SALT interfere with pathogens not only by trapping them, but also through the action of a variety of antimicrobial factors present in the mucus like lectins, lysozymes, pentraxins, complement proteins, antibacterial peptides and immunoglobulins. If, however, the pathogen succeeds to penetrate the mucosal epithelium, it encounters the innate cellular machinery, triggered by the cell types equipped with invariable receptors called pattern recognition receptors (PRRs), able to recognize common conserved molecules (PAMPs) characteristic of many microbial agents. The uptake of the antigen then leads to 1) release of cytokine mediators and attractants for different cell types to initiate the inflammatory process, and 2) antigenic presentation through the major histocompatibility complex (MHC) in the lymphoid tissues for the activation of the primary responses of antigen-specific lymphocytes expressing variable receptors able to recognize molecules specific to the pathogen, thus contributing to the development of secondary responses and memory. Due to the extensive interplay between innate and adaptive immunities, both these systems are according to the human studies equally sensitive and responsive to nutritional deficiencies, with the most consistent abnormalities seen in cell-mediated immunity, complement system, phagocyte function, cytokine production, mucosal secretory antibody response and antibody affinity.

6In this review, we will present the current use of high-throughput omics approaches to investigate the interplay between nutrition and immunity in fish (Fig. 1). Specifically, we will focus on recent advances in omics technologies, which include high-throughput transcriptomics by both microarray and RNA-seq approaches, but also proteomics examining the final products of gene expression. The advantages and disadvantages of the transcriptomic approaches have been discussed previously. The use of high-throughput techniques allows gene networks and molecular pathways to be identified that would not be possible when investigating only a small number of genes at any one time. There are advances being made using other technologies for deeper understanding of gene regulation including non-coding RNAs, epigenetics and metabolomics, however at the time of writing this review, these approaches have not been used to examine the relationship between nutrition and immune function and will not be covered here.

7One of the central criteria for successful high-throughput genomics is the availability of genomic resources. With the dramatic increase in availability of whole genomes and the reduced cost of deep sequencing, the landscape is changing at an increasing speed. Until several years ago, there were only genome sequences for a limited number of model species such as zebrafish and pufferfish (Fugu rubripes), from which inferences to non-model aquaculture species could be made. However at the time of writing this review, there are 82 fish genomes available on NCBI, 14 of these genomes are now anchored to chromosomes, and the remaining being at various stages draft sequences yet to be fully assembled. The implication is that there is a vast resource for gene and protein expression studies available for farmed fish. The key aquaculture species including Atlantic salmon, rainbow trout, Atlantic cod (Gadus morhua), common carp (Cyprinus carpio), European sea bass (Dicentrarchus labrax), tilapia (Oreochromis niloticus), grass carp (Ctenopharyngodon idellus) and channel catfish (Ictalurus punctatus) are examples of completed genomes, which have also gene models and RNA-seq resources to define the transcriptome. These resources are being continually annotated to a greater depth and allow for direct mapping of RNA-seq and gel-free proteomic mass spectrometry outputs to be achieved.

8In nature, fish can undergo extended periods of fasting, often associated with spawning migrations, during which some species of salmonids can go without food for many months. Yet, there are only few studies that have specifically examined the relationship between fasting and immune function using omics approaches (Table 1). The Atlantic salmon liver transcriptome was examined following infection with a bacterial pathogen Aeromonas salmonidica, with groups of fish being either fed or starved prior to infection. Microarray analysis clearly demonstrated that in the uninfected starved fish compared to the uninfected fed fish, there was a clear decrease in many components of the immune system, suggesting down-regulation of constitutive immune genes, most likely as an energy conserving mechanism. However, when the fish were infected, the magnitude of response of many acute phase response proteins was greater in the starved fish than in fed fish. The likely reason for this is that the starved fish were compensating and attempting to produce sufficient mRNAs for immune protein transcription. A more recent study involved the impact of 7-day starvation on channel catfish fingerlings. These fish are known to become highly vulnerable to Flavobacterium columnare, when undergoing food deprivation. The skin and gills were subjected to RNA-seq analysis to gain insights into mucosal immune system following starvation. Overall, the RNA-seq showed the majority of transcripts to be down-regulated and within these, a core group could be associated with immune function, including components of the Toll pathway, iNOS, lysozyme and peptidoglycan recognition protein 6. A number of complement genes were decreased in expression as were chemokines and interleukins 17 and 22. Together, these two studies show how transcriptomics reveals the immune components that are affected even by a short-term food withdrawal and as such feeding regimes need to be carefully assessed not only during commercial production but also when laboratory-based challenge experiments are being performed.

9Functional feeds are defined as feeds which are supplemented to enhance the health benefits to the fish, with the additives being beyond the basal requirements of the fish for normal growth and health. Such diets encompass many different components including micro ingredients such as selenium, zinc and vitamins, other components that act as immunostimulants such as algal and plant extracts are now often utilized. Prebiotics which are often derived from yeast extracts are believed to enhance the functionality of the microbiome in the fish intestine, whereas probiotics are bacterial species added directly to the diet with the aim of these becoming established in the intestine. Additional functional ingredients include oils such as high marine oil diets or phospholipid rich diets derived from krill that may enhance the immune function. Such diets are under continued development and increasingly viewed as important complementary component of fish health management in aquaculture. The expectation is that these diets will enhance the innate immune system enabling the fish to better resist pathogens such as sea lice in Atlantic salmon, or enable faster clearance of pathogen such as viruses once infected. There is growing literature on such feed additives (e.g., ), however the mechanisms by which these feeds alter metabolism and immune capacity is poorly understood. Nevertheless, evidence is growing that these dietary interventions do have an impact on both local (intestinal) and systemic immune systems. A further use for functional feeds in aquaculture has been to alleviate the impacts of high levels of antinutritional factors (ANFs) that are not fully removed from plant-derived proteins. Omics approaches have the potential to define some of the responses to such diets by allowing a holistic view of the transcriptome and proteome to be examined at once (Table 2).

10Sea lice are a major parasitic pest of Atlantic salmon and are represented by two main species 1) Lepeophtheirus salmonis, which is currently the most abundant species in the northern hemisphere salmon aquaculture, and 2) Caligus rogercresseyi, which is the predominant species impacting salmon farming in Chile. Varying levels of resistance to diets containing emamectin benzoate have brought a new emphasis on lice control, as costs for treatments are continually growing. Lice cause immunomodulation in the skin and mucosal surface by immunosuppression, resulting in a TH2 type of response. Resistant salmon show a greater Th1 and Th17 response, which is believed to drive early protection to lice infection. Both management and control of sea lice numbers have direct and indirect impacts on the salmon health, and functional feeds are viewed as a potential part of integrated pest management for these parasites in parallel to fallowing, choice of farm site, chemotherapeutic and selection for resistance lines of salmon. Dietary additives have impacts on lice settlement, however to improve and understand the mechanisms by which these diets cause reduced infestation, the use of omics approaches would be beneficial to employ, in particular to determine if there are associated alteration in the immune system, or if the effects are exerted by other routes.

11Among the dietary additives important for sea lice control are glucosinolates, When these plant derived-phytochemicals are metabolised by fish intestinal microbiota to isthiocynates, they can generate anti-inflammatory responses and promote antioxidant status and detoxification, but they may also promote pro-inflammatory responses in species such as mice. Diets containing glucosinolates were shown to reduce sea lice burden by up to 25% in comparison to control diets. The transcriptome analysis of skin showed that interferon and related genes were increased in expression by the supplemented diet prior to infection, and that supplementation induced higher expression profiles of Type 1 immune genes late into the infection. It was of interest that many of the transcripts associated with the lice burden were antiviral, but also chemokines and genes associated with acute phase response were involved in preventing the immunosuppressive effect of the lice. The fish however may also perceive glucosiolates as toxins and further transcriptomics investigations to liver, kidney and muscle showed significant enrichment for detoxification genes in the liver, but consistent with skin discussed above the kidney showed increased in interferon genes, with other processes including iron metabolism being altered, which has been associated with lice resistance. Another plant-derived functional feed with anti-lice properties has been described by, with lice counts significantly reduced following the feeding trial. The RNA-seq was carried out on both skin and head kidney, however no details of the dietary component have been given, so it is impossible to interpret the results in the context of potential mechanisms.

12Beta-glucans (β-G) and mannan oligosaccharides (MOS) are prebiotics commonly used in fish and are naturally occurring indigestible carbohydrates found in the yeast cell wall (YCW). Both β-G and MOS have been implicated in reduced sea lice load as well as improved response to other pathogens. Several proteomics studies that have examined skin and mucus proteomes suggest that the supplemented diets are modulating mucus to prevent lice attachment, with a variety of different proteins being associated with immune function such as lysozyme and galactins, which bind carbohydrates. A further study using YCW indicated calreticulin, (a multifunctional protein that occurs in the endoplasmic reticulin and plays a critical role in protein folding and degradation of glycoproteins) as a putative key mucosal protein induced by the YCW feed, however a direct link to how this protein affects development of mucosal immune response is not yet established. Diets containing YCW are also associated with increased survival of channel catfish when challenged with Flavobacterium columnare, the causative agent of columaris disease. During pre-feeding phase, the impact of these diets on gill, the major route of infection, was examined by RNA-seq showing the changes in the mannose receptor and genes related to IL4 signalling, which may modify the host response to the infection. These results suggested that IL4 signalling could reduce the pro-inflammatory response and drive tissue repair and enable resolution of the infection, with a Th2/M2 macrophage response being associated with the increased survival in fingerling catfish. This finding corroborates the hypothesis that lectins (especially the rhamnose binding lectins) and cytokines associated with IL4 R are implicated in the resistance to bacterial attachment. Lactocellulose, another indigestible carbohydrate synthetized from sucrose and lactose, has been used as a prebiotic in grass carp (Ctenopharyngodon idella), after which the liver transcriptome was assessed using a zebrafish microarray. It was suggested that immune function in liver was modulated by increased expression of novel immune receptors, leucocyte derived chemotaxins and α 2 macroglobulin, yet the specificity of hybridization to the array is uncertain even though both species are cyprinids.

13With the diets of farmed piscivourous fish changing dramatically in recent years, attempts have been made to manipulate the lipid sources to modulate the outcome of infection and post-infection tissue repair. Eicosanoids derived from arachidonic acid can promote inflammation, whereas eicosanoids formed from docosahexaenoic acid and eicosapentaenoic acid generally reduce inflammation and hence reduce the overall damage caused by the inflammatory process. Although fish fed high levels of vegetable oil grow well, these oils may have negative impacts on the ability to mount the correct response to infection under stressful condition and exposure to multiple pathogens. Two viral diseases in Atlantic salmon, heart and skeletal muscle inflammation (HSMI) and cardiac myopathy syndrome (CMS) are chronic diseases that result in viral accumulation in the heart and subsequent cardiac damage. Diets rich in fish oil and also krill oil (which is especially rich in phospholipids) were compared to a standard salmon diet containing high levels of rapeseed oil. Following infection with piscine reovirus (PRV), fish fed the diets rich in marine oils showed lower viral replication and subsequently less pathology in the heart, probably due to a dampened inflammatory response. Microarray analysis of heart tissue from selected time points following infection clearly demonstrated the reduced expression of key groups of immune genes until 12 weeks post-infection, after which there was a marginal rise in antiviral genes and antigen presentation, suggesting improved clearing of the virus. Another trial examining marine lipid functional feeds (specifically enriched with EPA) and responses to CMS showed similar mechanisms occurring, with antiviral and interferon genes generally suppressed as well as reduced stimulation of inflammatory genes following infection as compared to fish fed the reference diet. These experiments show that the lipid source can have major implications for disease response, which is typically not observed during normal growth. Further evidence for DHA having direct impacts on immune transcripts was observed following a graded feeding trial in salmon. This study demonstrated an increased expression of genes associated with immune system in the liver with gene sets enriched for chemokine signalling and natural killer cell mediated toxicity, but no disease challenge was performed and the data were not fully interrogated to examine immune function. Functional feeds containing phospholipids have also been shown to improve the survival outcomes for sea bream during cold water conditions that cause “winter disease”, a non-specific decrease in metabolic function and suppressed immunocompetence and performance. Furthermore, diets rich in fish oil and krill oil have also been shown to improve the survival rates of fish suffering from winter disease, while recent proteomic studies in both plasma and liver identified proteins indicative of an improved immune status, improved cellular stress response and altered lipid metabolism.

14Feed supplements can include vitamins and minerals, however to date there has been little research using vitamins in fish and employing omics approaches. As for micronutrients, recent work on selenium (Se) has demonstrated that this element can impact on immune responses. Selenium is an essential micronutrient and different chemical forms can have differing bioavailability. Organic selenium such as yeast derived Sel-plex® (manufactured by Alltec) has higher bioavailability than inorganic Se and may be used to increase Se levels in diets without inducing any toxic effects. Sel-plex enriched diets were feed to rainbow trout, after which the antiviral responses in head kidney and liver were assessed following a poly I:C stimulation (mimicking a viral infection). Transcriptomic profiling was performed by microarray and this clearly showed an enhance response to the stimulant in the fish fed Se rather than the basal diet. The response was much greater in the head kidney than in liver, in particular interferon γ and downstream proteins involved in cell mediated immunity appeared to be enhanced in expression by the Se-enriched diets. These results show that the capacity and components of the immune system can only be fully assessed by high-throughput transcriptomics rather than the single gene approach.

15Most fish need relatively high levels of protein and oil in their diet as they do not utilise carbohydrates efficiently enough to rely on them as a source of energy. For this reason, carnivorous, herbivorous and omnivorous fish all require similar quantities of dietary protein and oil per body weight, but it is the group of carnivorous fish such as salmonids that need the highest levels of fish meal and fish oil to ingest, as these marine ingredients closely resemble the natural feeding habits of predatory fish. As a result, farmed salmonids are highly sensitive to the dietary changes that involve replacement of marine ingredients with alternative plant materials, but these dietary shifts are inevitable and are expected to continue into the future. Indeed, the global contribution of fish meal to salmon feeds has decreased from ∼45% in 1995 to ∼20% in 2012 and is predicted to reduce further to ∼12% by 2020. Similarly, the global use of fish oil in salmon feeds has decreased from ∼25% in 1995 to ∼12% in 2012 and is expected to drop to ∼8% by 2020. Understanding the impacts of plant materials on fish health has been greatly advanced by employing the omics technologies to determine the molecular and cellular aspects of immune responses in organs directly affected by diet, such as intestine and liver.

16Sources of dietary protein have changed dramatically in the last decade with reduced reliance on marine derived proteins and greater use of terrestrial protein sources. This change although ensuring good growth is not without its complications and complex interaction with host microbiome and immune system. There are two key factors regarding plant proteins in fish feed. Firstly, there is imbalance of essential amino acids with lysine and methionine often being below the required levels for fish. Secondly, plant defence molecules (ANFs) can be co-purified with the proteins. Essential amino acids in the form of crystalline amino acids can be added to the diet at a cost, and also “functional amino acids” can be used to enhance specific aspects of physiology and health such as arginine. However, the problems generated by the presence of ANFs in aquaculture feeds are typically more difficult to address. When ingested, these ANFs impact on digestion, absorption and utilisation of nutrients and negatively modify intestinal physiology. The major plant materials used in aquaculture feeds are legumes such as beans, lupins and peas, which are rich in digestible proteins and have favourable amino acid profiles. The ANF content varies between species and depends on the method used for protein extraction. Soybean meal (SBM), a relativity crude protein extract can induce gut inflammation (enteritis) in the distal intestine. Intestinal transcriptome profiling has been carried out independently by a number of groups that indicate conserved responses to these diets (Table 3). Early alterations of the gut transcriptome were examined by, indicating a rapid response to the dietary components. However, many of these changes were still present after 8–10 weeks of feeding with SBM, indicating that clear immune responses were followed by dysfunction of intestinal barrier and unresolved gut inflammation.

17The early response at days 3 and 5 of dietary manipulation were dominated by immune-related transcripts, with gut function related genes being enriched from 5 days onwards. For immune genes, GTPase IMAP family members, regulators of T cell and B cell function and NF-kB-related genes were clearly increased in expression. The transcripts with reduced expression were related to functions including endocytosis, transport and metabolic processes, suggesting dysfunction of the intestinal epithelium and altered intestinal function.

18Long-term feeding on high SBM diet was characterised by activation of T cell mediated processes via up-regulation of the CD86 antigen, cytotoxic T lymphocyte-associated protein 4, interleukin-18 (IL-18) and IL-22, while inflammation and respiratory burst were indicated by the increased expression of genes in the TNF- and NF-kB-mediated responses. These responses could reflect the increased translocation of luminal bacteria, viruses and antigens across the intestinal epithelium, as indicated by increased expression of pathways for clathrin-mediated endocytosis signalling, macropinocytosis signalling and virus entry via endocytic pathways.

19The SBM is now commonly used as a model for inducing gut inflammation in fish intestine (e.g.,,, ). For use in aquaculture feeds (especially for piscivourous fish), SBM is further purified to soy protein concentrate (SPC) by alcohol washing, which removes saponins, some of the major ANFs in raw soy material.

20SPC-rich diets do not typically modify intestinal histology or enrich genes associated with SBM-induced enteritis. The direct effect of soysaponins is complex as when fed to fish alone or with combination with corn gluten, sunflower, rapeseed or horsebean proteins - no or little inflammation was observed, whereas when soysaponins were used with pea protein, significant histological and transcriptional changes were observed. These results suggest that the interactions between different AFNs are also important and the mechanisms by which different ANFs affect gut health are not fully understood. One possible explanation could be the increased intestinal permeability and response to various antigens associated with different types of plant materials. Other plant protein concentrates used at high levels such as faba bean (bean protein concentrate, BPC) do not generally contain saponins. Instead, BPC is characterised by high levels of condensed tannins and the presence of faba bean-specific glucosides such as vicine and convicine. Importantly, both diets with high levels of SBM or BPC induced the gut inflammatory disease, but the transcriptomic analysis of the tissue indicated substantially different diet-specific responses, with a small core group of genes representing generic inflammation responses. The shared response genes represent putative markers for inflammation across different inflammatory inducers.

21Development of effective strategies to replace fish oil with vegetable oils has been hindered by the difference in their fatty acid profiles, especially in the content of n-3 long-chain polyunsaturated fatty acids (n-3 LC-PUFAs) such as EPA and DHA, which are abundant in fish oil but not synthesised by terrestrial plants. These fatty acids are essential or conditionally essential for all vertebrates and have well-established effects on immune and inflammatory processes in humans, including decreased production of pro-inflammatory cytokines and eicosanoids, enhanced phagocytosis and reduced leucocyte-endothelial adhesive interactions. These effects have been interpreted in the context of reducing inflammation that would enable inflamed tissues to return to homeostasis. In human nutrition, the most widely available dietary source of EPA and DHA is oily fish, such as salmon, herring, mackerel and sardines. The fish used as sources do not actually produce EPA or DHA, but instead accumulate them by consuming either marine phytoplankton (the primary source of n-3 LC-PUFAs), prey fish with EPA/DHA already accumulated or an aquaculture feed containing sufficient levels of fish oil as an ingredient. In contrast, vegetable oils shape the tissue fatty acid profile differently, generating fish that are typically low in EPA/DHA and high in n-6 LC-PUFAs, thus making them less desirable for human consumption. These low levels of dietary EPA/DHA may also have adverse effects on fish themselves, but the magnitude of these effects remains largely unknown. Research into the impacts of vegetable oils on fish health and immunity has recently gained momentum by combining feeding trials with large-scale analyses of gene expression performed on organs most likely affected by changes in the dietary fatty acid profile (Table 4).

22Most studies so far were conducted on Atlantic salmon (besides Atlantic cod and gilthead sea bream), using vegetable oil (either singly or as blends) from Camelina sativa (both wild-type and genetically modified) and focusing on either intestine (pyloric caeca and mid or distal guts) or liver, with one study targeting spleen. The replacement of fish oil with vegetable oils varied from 40 to 100% and the feeding trials lasted from 67 days to 55 weeks. Two of these studies investigated also the interaction between diet and genotype (fish selected for high and low lipid contents), while all other experiments were performed on non-selected fish. On two occasions, the feeding trails were followed by the disease challenge, using either viral mimic polyriboinosinic polyribocytidylic acid (pIC) or Enteromyxum leei (myxosporean parasite).

23The tissues for the transcriptomic evaluation of the effects of vegetable oils on fish health were selected based on their contribution to digestion, metabolism and immune system, including gut as the main site of the direct exposure to nutrients and foreign antigenic material, liver as the main metabolic organ and spleen as the important lymphoid organ. Although none of the studies listed in Table 4 compared the expression of immune genes between different tissues, the majority of changes induced by the exposure of non-challenged fish to wild-type vegetable oils were described in liver rather than in gut or spleen. These findings are consistent with the involvement of the liver in detoxification, modulation of immune responses, as well as the production of inflammatory mediators. Importantly, the replacement of 66% of fish oil with vegetable oil in the diet of gilthead sea bream had no effects on the gut transcriptome in unchallenged fish, but substantially altered the expression of immune genes in infected fish, highlighting the need for testing the novel diets in the context of exposure to pathogens.

24Although the interplay between nutrition and immune system is well recognised, basic and applied research on the interactions between diet and health in fish is lagging behind the mammalian studies. To fully understand the repercussions of aquaculture feeds on fish physiology, a shift in approach is required to determine the molecular and cellular pathways that regulate responses to different diets. The new omics technologies, especially transcriptomics coupled with full genome sequences, offer enormous potential to investigate the complex relationship between fish nutrition and immunity, both in health and disease. However even though the field is advancing rapidly, there are a number of major gaps in the knowledge that need to be addressed. One of the major challenges is for example the relationship between the nutritional content of aquaculture feeds, fish intestinal microbiota and the resultant metabolites, and how these metabolites modified differently by different diets impact fish heath and their resistance to pathogens. Fish microbiota studies are advancing with deep sequencing approaches, but as yet there is little interpretation of the findings of such studies in the context of fish immune status and health. Understanding the mechanisms that underpin the links between diet, intestinal microbiota and fish health will almost certainly become a major focus in the next few years.

25Early life development and the dietary impacts on development of the intestinal and immune system are still far from clear. Such research is important as it may lead to designing strategies for nutritional programming and intestinal tolerance at the time of first feeding. The relationship between diet and ontogeny of the immune system will also require the knowledge of both trans-generational epigenetic control of immune gene expression as well as life-long epigenetic control of immune genes expression established during the time of first feeding. Such experiments could expand the use of zebrafish as a model species and also exploit the lines of transgenic fish, which allows the development and homing of specific immune cell types to be examined. Other models include intestinal cell culture, either primary or stable cell cultures, that will help to explore the direct impacts of dietary components (or microbial metabolites) on intestinal immune capacity.

26Many studies described in this review provide strong evidence for the interplay between nutrition and immunity at the level of innate immune components. However, little is known about the impacts of dietary manipulations on the adaptive immune responses, such as activation of T and B lymphocytes. The adaptive immune responses are central to long-term effects of diet on fish health and resistance to pathogens. Furthermore, understanding the link between nutrition and adaptive immune system is essential for developing strategies for oral vaccination and the improvement of antigen uptake and memory.

27The final future perspective is how these omics technologies can be integrated with the ambition of generating predictive models for diet, immune system and health outcomes. Such work requires improved genome annotation, the knowledge of immune cell type-specific responses and mathematical computational expertise, which can then be combined and used to dissect the molecular mechanisms underlying the diet-immunity interactions, leading to improved health of farmed fish and sustainable aquaculture.