In-vivo analysis of Protec ™ and β -glucan supplementation on innate immune performance and intestinal health of rainbow trout

.


Introduction
Functional feeds are now used as part of health management and aquaculture in response to increased intensification of rearing, for prevention of disease outbreaks and improving general health in farmed fish.These feeds aim to not only to sustain normal physiological function but also to improve responses to pathogenic threats which subsequently aids/improves survival.These functional feed ingredients can be split into several main groups: prebiotics, probiotics, immunostimulants and micronutrients.Prebiotics are often added to support the growth of a favourable microbial communities in the intestine by acting as a nutrient source for bacteria, these nutrients often being indigestible by the animal itself and are commonly oligosaccharides made up of repeating monomers of sugars (Mannose, Galactose, Fructose) with a variety of different lengths and structures.The molecules can also act directly upon the immune cells present in the gut with effects on macrophage activation, lysozyme activity, respiratory burst and leucocyte activity [1].The addition of these oligosaccharides can alter gene expression in the intestinal tissue with indication there is enhancement of the innate immune response [2].
β-glucans are commonly used prebiotics, they can have a variety of different structures that can induce different immune responses [3,4], they are comprised of a D-glucose backbone with side chains connected by β-glycosidic bonds.These naturally occurring compounds are found Abbreviations: VHSV, Viral haemorrhagic septicaemia virus; PAMP, pathogen associated molecular pattern; GALT, Gut associated lymphoid tissue; rIL-1β, recombinant Interleukin-1β.
in the cell wall of yeast, algae, and many other species of plants and fungi.The innate immune system acts on a broad range of molecules and is thought to bind to β-glucans through C-type lectins [4], these are receptors that bind carbohydrate molecules and trigger immune pathways through the C-type lectin signalling pathways which induces both innate and adaptive immune responses [5].C-type lectin 4 has been previously characterised in rainbow trout (Oncorhynchus mykiss), where the gene is induced following infection with the bacterial pathogen Yersinia ruckeri [6].Upregulation of the C-type lectin pathway can lead to amplification of a panel of genes known to be drivers of inflammation including COX-2, IL-1β, TNFα, IL-6, and IFNγ as demonstrated in a number of studies [3,4,[7][8][9][10].The increase in expression of these key proinflammatory genes suggests β-glucans can prime the innate immune system and hence protect the fish against disease, leading to higher survival against pathogens as previously demonstrated for viral haemorrhagic septicaemia virus (VHSV) [8], Aeromonas salmonicida [11,12] and A. hydrophilia challenges [7].
Immunostimulants are of great research interest in health management for both production animals and human health, and as a result has led to the discovery of many different immunostimulants; polysaccharides, vitamins, hormones, and bacterial components (PAMPs) [13].Protec™ diet is a commercial aquafeed (Skretting) which contains several functional compounds including β-glucans, vitamin E, vitamin C and zinc.The Protec™ diet enhances the innate and adaptive immune response of rainbow trout through the enhancement of respiratory burst, leucocyte proliferation and increased production of specific IgM against Lactococcus garvieae [14] and can be considered to have both immunostimulatory and prebiotic effects in the host.A feeding trial using Protec™ in rainbow trout following an infection with L. garvieae showed increased leucocyte respiratory burst activity and increased levels of L. garvieae antibodies occurred at both 4-and 8-weeks post-infection compared to a control diet before and after vaccination against L. garvieae.Protec™ has also been demonstrated to protect against viral pathogen infections such as viral haemorrhagic septicaemia virus (VHSV) [8], where fish fed the Protec™ diet mounted a greater immune gene response (MX, IgM, IgT and IgD) to VHSV and showed increased levels of natural IgM in sera after being fed for 30 days on the functional diet.Further research using in-vitro approaches in rainbow trout demonstrated improvements in head kidney leucocyte proliferation and respiratory burst activity after stimulation with sonicated Protec™ immune support pack which is a combination of glucans, vitamin C, vitamin E and zinc [14].
To measure the effectiveness of functional ingredients, both in-vitro and in-vivo methods can be used.Traditionally, feeding trials are commonly used, followed by assessment of growth and immune parameters.To complement such studies, recent developments of cell culture techniques such as the use of cell lines, RTS11 [9] and primary cell cultures such as head kidney macrophages [4], have been used to further elucidate the role of functional ingredients on specific cell types.This use of cell lines gives added information to mechanisms of response, but also reduces the number of fish being used in feeding trials.Gut associated lymphoid tissue (GALT) leucocytes have been used in assays to identify responses to PAMPs [15,16] and have also been used in previous research to assess nutritional ingredients [10].GALT leucocytes extracted from the gut contain multiple different cell types and enable greater interpretation of responses to whole tissue or specific cell lines used in cell culture, where the cell lines are restricted to a single phenotype.These immune cells can interact directly with the functional ingredients and may demonstrate how they are likely to respond in the gut at the interface between the host and the feed sources.
In this study, the responses in rainbow trout following a 6-week feeding trial of two diets containing β-glucans compared to fish fed control diet was examined.The three diets were a commercial functional feed diet containing β-glucan and other components (Protec™, Skretting), a diet with β-glucan alone as prebiotic, and a commercial diet.To assess the in-vivo effects on the fish; growth parameters, histology of the distal intestine, gene expression of a panel of immune cell markers and expression of innate immune genes in primary cultures of GALT leucocytes were used.To identify how the functional diets modulated the innate immune response, GALT leucocytes were extracted and stimulated with Poly I:C and recombinant IL-1β followed by gene expression analysis.To further understand how diets, modulate the innate immune response to pathogenic challenges an immune challenge was carried out using A. salmonicida and was followed by targeted gene expression analysis of innate immune related genes.The data collected indicates the effectiveness of β-glucans and the Protec™ diet in modulating the gut innate immune response potentially allowing the fish to be better equipped to fight disease during stressful life events.

Fish
All procedures were carried out under the UK Animals (Scientific Procedures) Act 1986 and Home Office code of practice guidance, under Home Office project licence PFF8CC5BE.Rainbow trout (~75 g) were obtained from a commercial rainbow trout farm and kept in 1 m diameter fiberglass tanks with recirculating freshwater at 14 ± 1 • C, within the aquarium facilities in the School of Biological Sciences, University of Aberdeen, UK.

Feeding trial
For the feeding trial 270 fish were distributed randomly between nine tanks; with tanks being run in triplicate for each diet giving 90 fish per diet group, a diagram of the experimental set up can be seen in Fig. 1.After a two-week acclimatization period (during this acclimation fish were feed on standard commercial diet, Skretting), a six-week feeding trial was carried out.Three diets were used; acclimatization as a control diet and two experimental diets one with β-glucan supplementation (0.1%) and another the Protec™ diet produced by Skretting ARC (commercially available).The composition and formulation of experimental diets are shown in Table 1.All diets were prepared by Skretting Aquaculture Innovation.Fish were fed continuously throughout daylight hours (using automated Arvotec feeders) with fish being fed an amount equal to 1.5% of their average body weight per day.All fish were weighed at the beginning of the feeding trial and ten fish per tank were weighed every two weeks until the end of the trial with the amount of feed adjusted accordingly.

Growth parameters
Specific Growth Rate (SGR), Fulton's Condition Factor (K), Feed Conversion Ratio (FCR), Weight Gain (WG), weights and fish lengths were calculated as biometric data.
SGR was calculated using the following formula to calculate the percentage gain per day: where W1 is starting weight and W2 is final weight.FCR was calculated using the following formula: Fulton's Condition Factor was calculated using the following formula:

Histological examination of distal intestine
Following week 6 of the feeding trial, fish were randomly selected from each diet (20 fish per diet, 60 total) and killed by overdose on anaesthetic before confirmation of death by the destruction of the brain (schedule 1 killing) before distal intestine tissue samples were taken (approximately 1 cm in length).The tissue samples were rinsed with PBS (Sigma) to remove excess digesta before being stored in 10% neutral buffered formaldehyde (CellPath).Sample preparation, and image acquisition were carried out at the University of Aberdeen Microscopy and Histology Core Facility.Distal intestine samples were routinely dehydrated in ethanol, equilibrated in xylene, and embedded in paraffin wax according to standard protocols [17].Cross sections (3 μm) of the distal intestine were cut with a microtome and mounted onto microscope slides.These sections where subjected to haematoxylin and eosin (H&E) staining with alcian blue.Sections were imaged using the Zeiss AxioScan Z1 slide scanner.The resultant images were randomised to ensure blinded assessment and then scored using a semi-quantitative criterion using the Zen Blue software (V3.4).The criteria measured 6 variables (goblet cells, eosinophilic granulocytes, lamina propria, mucosal folds, supranuclear vacuoles and sub-epithelial mucosa) and each were given a score between 1 and 5 as previously described by Uran et al [18] and can be seen in Table 2.An increasing score represents increasing inflammation.The morphology of the mucosal folds (MF) is an indicator of gut health, as MFs are important for increasing surface area for the absorption of nutrients.The presence and size of supranuclear vacuoles (SNV) is measured as SNVs contains lysosomal proteins.The abundance of goblet cells was measured due to their roles in mucous production and are a characterised inflammatory marker.The widening of the lamina propria (LP) and thickness of the sub-epithelial mucosa (SEM) were measured to monitor the health of mucosal membranes with a larger LP or thicker SEM suggesting higher levels of inflammation.The infiltration of eosinophilic granulocytes (EGs) into the SEM or LP was measured as an indicator of the immune capacity/migration of immune cells to the gut (site of inflammation).These scores were analysed using a two-way Anova in R (V4.1.0).

Isolation of primary GALT leucocytes
GALT cells were isolated as previously outlined by Porter et al [10].Fig. 1.Diagram of the experimental set-up.A 6-week feeding trial was carried out with a range of immunological sampling taken at the end of the 6 weeks.After which 12 fish per diet were injected with Aeromonas salmonicida for 48hrs before whole tissue samples of head kidney and distal intestine were collected to measure the acute response to bacterial infection.
The distal intestine was removed from fish and stored in cold PBS (Sigma) until washing.Samples were rinsed in PBS three times before being cut longitudinally and then into 1 cm 2 sections to aid with rinsing.Samples were then rinsed for 30 min in an orbital shaker at 40 rpm in 30 ml of PBS.Following one final rinse in PBS samples were added to falcon tubes containing 25 ml of predigestion solution (HBSS (Gibco, 14025092) + 0.145 mg/ml DTT (Sigma, D9779) + 0.37 mg/ml EDTA (Fisher, D/0700/53)) and put in an orbital shaker at 40 rpm for 20 min.
The supernatant (S1) was filtered into 50 ml falcon tubes using 100 μm cell strainers (Greiner).before being rinsed with PBS twice, the pelleted cells were stored in growth media (Leibovitz L-15 media (Gibco) + 10% FBS + 1% P/S) in a 20 • C incubator.The gut segments were rinsed using washing solution (HBSS + 0.05 mg DNase1/ml (Sigma, DN25) + 5% FBS + 1% Penicillin/Streptomycin) to remove EDTA and DTT.After which the gut segments were added to 25 ml of digestion solution (washing solution + 0.37 mg collagenase IV/ml (Gibco, 17104-019) in 50 ml falcon tubes before being placed in an orbital shaker for 2 h at 40 rpm.The supernatant was filtered into the tube containing S1 using 100 μm cell strainers (Greiner).The mixture was washed three times with PBS +1% P/S to remove any collagenase before being resuspended in 5 ml of growth media.The cell suspension was slowly layered on top of a discontinuous percoll gradient (25% and 75%) (GE healthcare, 17-0891-01) and centrifuged for 30 min at 400 g, at 4 • C. Cells were then collected from the interface between the 25 and 75% percoll gradients and washed twice in PBS +1% P/S before being resuspended in stimulation media (Leibovitz-15 media +1% FBS + 1% P/S).Collected cells were adjusted to 1 × 10 6 cells/ml before 1 ml was added to 6 well plates to which stimulants were added.

Immune stimulation of GALT leucocytes
Two different stimuli were used to stimulate the GALT leucocytes; poly I:C (100 μg/ml, Sigma, P1530) to induce an antiviral response and, recombinant IL-1β (20μgml − 1 , Provided by Dr. Tiehui Wang (Scottish Fish Immunology Research Centre)) [19].A stock solution for poly I:C was diluted in L-15 media with aliquots stored in − 20 • C until use.rIL-1β was stored in − 80 • C in aliquots and diluted in L-15 media.The stimulants were then added to wells containing 1 × 10 6 cells diluted in 1 ml of stimulation media to give the working solutions.All experiments were performed in a randomized design with sextuplicate wells that were treated independently through the entire protocol (n = 6).

Immune challenge with Aeromonas salmonicida
After week 6 of the feeding trial, 24 fish were randomly selected for immune challenge from each diet.These fish were then split into two challenge tanks each containing 12 fish for each diet.Fish were injected interperitoneally with 1 ml of live Aeromonas salmonicida (2 × 10 6 CFU/ ml) or PBS for control fish.Head kidney and distal intestine tissues (~1 cm) were harvested 48-hrs after infection.Samples were stored in RNA later and stored in the − 80 • C freezer until RNA extraction.

Transcriptional analysis
RNA was extracted from cells in individual wells or tissue samples using 750 μL or 1 ml of TRI Reagent (Sigma) respectively following the manufacturer's instructions.The RNA pellet was washed using 80% ethanol, dissolved in RNase-free water, and stored at − 80 • C until use.Quality control of the RNA was determined by nanodrop spectrophotometer for quantity and Agilent bioanalyzer 2100 to determine RNA integrity.RNA (500 ng for cells and 1 μg for tissues) was used as template for reverse transcription using the Qiagen QuantiTect Reverse Transcription Kit following manufacturer's instructions.The generated cDNA was diluted 10x with RNase-free water and stored at − 20 • C until use.

Statistical analysis
Data handling, calculation of fold change and statistical analysis was performed in Microsoft Excel 2016.Statistical analysis carried out in Excel was the two-tailed students T-test.Graphical representation and one-way and two-way Anova's of the data were completed using GraphPad Prizm 5 and R (V4.1.0).

Growth parameters
A feeding trial was carried out using a β-glucan supplemented diet, Protec™ diet and a commercial control diet, all fish survived and remained healthy throughout the trial.No significant differences were found in the biometric parameters measured (Table 4) or between tanks for each diet.Overall, the fish had an FCR of 0.95 ± 0.17, 0.96 ± 0.11, 1.23 ± 0.12 for control, β-glucan and Protec™ diets respectively .

Histological analysis of dietary modulation of the distal intestine
Histological analysis revealed all three diets had healthy levels of all parameters measured (examples of intestinal sections for each diet see Fig. 2).The number of goblet cells was significantly different (p < 0.05) across all three diets with fish fed the Protec™ showing the highest number of goblet cells followed by the fish fed the β-glucan diet with the lowest numbers were observed in the fish fed the control diet.Scoring of the mucosal folds indicated a significant difference (p < 0.05) between all three diets with the β-glucan diet showing the highest scoring followed by the Protec™ diet with the lowest scoring found in the control diet.These results suggest that both the β-glucan and Protec™ had increased bloating and shrinkage of the mucosal folds compared to the control diet.The supranuclear vacuoles demonstrated significant decreases in size (p < 0.001) in both the Protec™ and the β-glucan diet compared to the control diet, 2.35 (Table 5).The lamina propria showed some significant increases (p < 0.001) in size and thickness in both the β-glucan diet, 1.87, and Protec™ diets, 1.79 compared to the control diet, 1.56 (Table 5).There were no significant differences observed for sub-epithelial mucosa (Table 5) or the eosinophilic granulocytes () Table 5 between the different diet treatments.

Identification of GALT leucocyte cell markers
Gene expression for specific cellular markers were used to identify Tcell (CD4-1, CD8α), dendritic cell (MHCIIβ, CD108) and B-cell (sIgM and sIgT) populations to determine if there were any differences between dietary groups which may signify changes in cellular abundance/activity of these cell types within the distal intestine (Fig. 3).T-cell marker CD4 was expressed at a significantly more abundant level (p < 0.05) in the fish fed the β-glucan diets compared to the control diet, however no significant differences were observed for CD8α.sIgM as marker for Bcells was expressed at significantly higher abundance in the β-glucan diet, additionally the fish fed Protec™ diet showed an increase but did not reach significance.Expression of sIgT was not altered as result of diet.No differences were detected for the dendritic cell markers between diet treatments.

Dietary modulation of key immune markers in GALT
Expression of key anti-viral and proinflammatory response genes were examined in the GALT leucocytes (Fig. 4).Targeted gene expression analysis highlighted differences in the expression of proinflammatory genes studied and clustering of the β-glucan and control diets (Fig. 5).GALT cells from fish fed the Protec diet had significantly higher expression of the two proinflammatory cytokines, IL-1β and TNFα, these cells also had significantly higher expression of the antimicrobial peptide HAMP and the acute phase protein SAA in comparison to both cells from the control and β-glucan fed fish.Analysis of the genes associated with anti-viral activity indicated an increased expression of both MX and γIP in the cells from both β-glucan and Protec™ fed fish but was not significantly increased.However, IFN-1α, RSAD2 and PSMB9 were not modulated by diet (Fig. 4).

Dietary mediated responses of GALT leucocytes to poly I:C
Poly I:C was used to generate anti-viral activity in GALT cells fed the different diets and were compared against non-stimulated cells (Fig. 6a,   b, c).The GALT cells isolated from fish fed all three diets showed significant increases to the poly I:C with IFN-1α, RSAD2 and γIP all being significantly increased compared to the unstimulated cells.The GALT cells from the fish fed the β-glucan and Protec™ diets both had significantly increased the magnitude of MX gene expression compared to unstimulated cells.Finally, the GALT cells from the Protec™ fed fish also had a significant increase expression of PSMB9 compared to unstimulated cells, which was not observed in the other diet groups.Although the GALT cells all had capacity to respond to the diets, further analysis of the magnitude of these changes suggested the diet did not affect the response between diets other than for the PSMB9 gene that only responded to the poly I:C in the Protec™ fed fish (Fig. 6d).

Dietary mediated responses of GALT leucocytes to rIL-1β
The rIL-1β induced expression of IL-1β, IL-8, and TNFα in GALT cells derived from fish fed all diets (Fig. 7a, b, c).SAA, an acute phase response gene was significantly upregulated in both the β-glucan and Protec™ fed fish but not in the fish fed the control diet compared to their respective unstimulated samples.No significant difference in fold change was observed in HAMP in the fish fed the β-glucan and control diets, however there was a significant decrease in expression of the fish fed the Protec™ diet.When the magnitude of the response to the rIL-1β was examined in the GALT cells, the cells from the fish fed the Protec™ had a significant larger increase (fold change) in TNFα compared to the other diets (Fig. 7d).

Dietary mediated responses to Aeromonas salmonicida
To further assess the impact of the diets on the immune response of rainbow trout, fish were challenged with A. salmonicida, a bacterial pathogen of salmonid fish.Proinflammatory genes were measured in both the head kidney and the distal intestine alongside genes relating to mucins and secretory antibodies in the distal intestine (Fig. 8).
Immune Responses Head Kidney following bacterial infection: IL-1β was significantly upregulated in the head kidney of fish fed all three diets in compared to PBS treated controls.Fish the β-glucan diet had a significantly lower induction, (2-fold change), compared to the control diet and Protec™ diet (6.7-and 5.6-fold respectively).IL-8 was upregulated compared to unstimulated samples in fish fed both the control diet and the Protec™ diets (1.8-and 2.2-fold change respectively) but was not changed in the fish fed β-glucan diet.TNFα was significantly  upregulated in all three diets with no significant differences between the control and Protec™ fed fish.However, the fish fed β-glucan diet showed significantly less magnitude of response (p < 0.05) compared to both control and Protec™ fed fish with only a 2.8-fold increase.SAA and HAMP showed a similar trend in both antimicrobial genes with all diets showing a significant upregulation in response to A. salmonicida infection.However, for both genes, both SAA and HAMP showed significantly lower magnitude of response in both the β-glucan and Protec™ diets compared to the control diets.

Distal Intestine Immune Responses following bacterial infection:
IL-1β was significantly upregulated in all three diets compared to PBS treated samples.The fish fed β-glucan diet showed a significantly higher magnitude of response (16.3-fold change) compared to fish fed both the control and Protec™ diets (4.5-and 2.8-fold change respectively).The HAMP gene followed the same trend as observed in the head kidney with both the fish fed β-glucan and Protec™ diets, 6.2-and 3.8-fold change, showing a much lower magnitude of response compared to the control diet, 34.1-fold change.Conversely, SAA had a higher magnitude of response in both the fish fed β-glucan and Protec™ diets (17.5-and 21.4fold change) compared to the control diet (6.4-fold change).The genes relating to secretory immunoglobulins T and M both exhibited a similar trend with the fish being fed the β-glucan diet having a higher magnitude of response ~3.4-fold for both genes compared to both the control   and Protec™ fed groups.Fish fed the Protec™ diet also showed a significantly lower magnitude of response to both the control diet and β-glucan diet in sIgT.CLEC4T was significantly upregulated in both the control and β-glucan diet groups after the immune challenge, no change in expression was observed in the fish fed the Protec™ diet.Both the control and β-glucan diets showed a higher magnitude of response for CLEC4T, (3.0-and 3.1-fold change) compared to the Protec™ diet (0.8fold change) after immune stimulation.A significantly higher relative expression of CLEC4T was observed in the Protec™ diet in PBS treated samples, compared to both the control and β-glucan diets.Two mucin genes MUC5AC and MUC2A were examined, MUC5AC had a higher magnitude of response in both the control and β-glucan diet groups (~2.4-fold change) compared to those fed Protec™ diet (0.1-fold change).However, fish fed the Protec™ diet showed a significant increase in relative expression in the PBS samples compared to fish fed both the control and β-glucan groups.The MUC2A gene showed no significant difference to PBS treated samples across all three diets.The proinflammatory response and antimicrobial genes IL-1β, SAA and HAMP were measured across both the head kidney and the distal intestine.Interestingly IL-1β showed a higher magnitude of response in the fish fed the β-glucan diet in the distal intestine compared to the head kidney, 16.3-compared to 2-fold change respectively.For fish fed both the control and Protec™ diets a lower magnitude of response was observed in the distal intestine compared to the head kidney.For both anti-microbial genes SAA and HAMP, a higher magnitude of response was observed in the head kidney compared to the distal intestine.

Discussion
In this current study, we have demonstrated that the administration of immunostimulant in the form of either purified β-glucan or β-glucan with additional factors (Protec™) can modulate the innate immune activity in the intestinal cells of rainbow trout.The data indicates that the diets are modulating the immune response through both the localised response within the distal intestine and the systemic response within the head kidney.The results of this trial elucidate that diets containing β-glucans are able to alter the magnitude of the response against bacterial challenges.
Supplementation with immunostimulants had no significant effect on growth performance as assessed by growth rate, food conversion and condition factor, indicating that all three diets were performing well regarding biometric performance.Detailed examination of the distal intestine by histological analysis revealed that the fish fed the β-glucan supplemented and Protec™ diets had a higher number of goblet cells distal intestine compared to the control fed fish.This result may reflect the immunostimulatory molecules found in both experimental diets that can cause an increase in the proinflammatory response which can lead to upregulation of mucins which are secreted by this cell type [23].Goblet cells are specialised epithelial cells that maintain barrier function through secreting mucous, and secrete important antimicrobial peptides, mucins, and cytokines [24].Goblet cells aid the innate immune system by creating goblet cell-associated antigen passages which allow the delivery of microbial products and other antigens to be presented to antigen presenting cells in the lamina propria and trigger the innate immune response.Increased goblet cell numbers have been implicated in a wide number of different detrimental conditions such soybean induced enteritis caused by the supplementation of high concentrations of soybean meal to fish diets although such high concentrations are not commonly used in commercial diets [18,[25][26][27].However, with increased levels of goblet cells, if associated enteritis, often, higher scores for lamina propria and other variables measured [27].As all other variables measured demonstrated either no change or an improvement in gut health in the supplemented diets the increased goblet cell number may be due to the immunomodulation caused by β-glucans.The higher levels of goblet cells seen by the supplemented diets may offer increased protection against pathogens but not necessarily representing an inflamed status on the distal intestine.
The analysis of transcriptional cell markers that can be used as a proxy to represent the abundance of immune cell types and they indicate the two function ingredients can alter the leucocyte composition in the distal intestine.The fish fed the β-glucan diet had higher expression of the B-cell marker, sIgM and the T-cell marker, CD4, compared to both the control and Protec™ fed fish.The Protec™ diet also showed nonsignificant increased relative expression levels in genes relating to Band T-cell markers.Higher levels of CD4 transcripts would suggest there are higher numbers of CD4 + T-cells, a key cell in mediating the adaptive Fig. 5. Heatmap of basal expression between diets.qPCR was used to quantify gene expression where samples were normalised against three housekeeping genes.Genex Multi-iD 5 was then used to calculate relative expression values.Values are mean expression levels relative to each other.response [28].Multiple subsets of CD4 + T-cells have been characterised in teleost fish and are the dominant subtype of T-cells in the lamina propria [29].Levels of sIgM transcripts were also observed to be higher in the fish fed the β-glucan supplemented diet.IgM secreting B-cells are the prominent B-cell subtype in the intestine and as such play a key role in barrier maintenance and mucosal immunity [29].IgM is responsible for helping maintain homeostasis in the microbiome and is secreted during the first stages of the immune response to infection.IgM is released by B-cells where it can bind to antigens found within the gut microbiome and acts as a marker for phagocyte recruitment, activation of the complement pathways and destroy antigens through opsonisation [30,31].Supplementation with β-glucans has previously shown to cause increased expression of IgM in serum in rainbow trout [8] before infection and also increased expression in head kidney RNA in gilthead sea bream (Sparus aurata) [32] after one week of being fed β-glucan supplemented diets.Further studies may include the use of immunoglobulin D to identify specific B-cell responses to functional diets.
GALT leucocytes are a useful tool in identifying specific localised responses of the distal intestine immune cells between the host and external factors [33].Isolated GALT leucocytes gene expression was examined after 4hrs with no stimulation to ensure cells were healthy and the responses observed were not just due to the extraction method.The GALT leucocytes extracted from fish fed the Protec™ diet had higher expression levels of the proinflammatory cytokines IL-1β, SAA, HAMP, and TNFα after 6 weeks of feeding compared to fish fed both the control and β-glucan diets potentially confirming the immunostimulatory nature of β-glucans, which is in agreement in previous research [8,9,12,34].However, further research is needed to identify if these higher expression levels were directly due to β-glucans or other components within the Protec™ diet.Regarding the modulation of anti-viral genes in the GALT following the feeding trial we find there is minimal impact on these genes with only a non-significant increase in relative expression of MX transcripts observed in both the β-glucan and Protec™, suggesting neither the β-glucans nor the Protec™ diets are not triggering an anti-viral response.Viral responses also showed similar responses to previous research [8] with MX showing no significant difference between control and Protec™ diets with all other genes tested showing similar responses across all diets.
In order to examine the if the different diets impacted the capacity of the GALT cells to responds to immunological stimulation a panel of stimulants (rIl-1β or Poly I:C) known to drive proinflammatory or antiviral responses were as previously described to work well on GALT cells from rainbow trout [16].All cells no matter which diet the fish had been fed responded in the predicted way to the stimulants after 4hrs.For the anti-viral activity driven by poly I:C the only significant difference in response between diets was for PSMB9 [35] which encodes a proteasome subunit involved in antigen presentation and known to be increased under type 1 interferon stimulation [36].The increased magnitude of response of this gene may suggest an accelerated response to the poly I:C in the GALT cells isolated from the fish fed the Protec™ diet, which could improve the response to viral pathogens as shown by Leal et al [8].Following the exposure to rIL-1β, there was a strong increase in the three cytokines examine at 4hrs.TNFα was the only gene that was found to respond differently depending on the diet the fish had been fed prior to GALT extraction and stimulation, where we found that the fish fed the Protec™ diet showed significantly increased expression of this gene.TNFα is a key proinflammatory cytokine and like Il-1β drives the proinflammatory response [37].The protein is secreted by macrophages and monocytes [38,39].This increased expression in response to the proinflammatory stimulus would suggest the fish being fed the Protec™ diet have an accelerated response and may increase their capacity to deal with immunological insults in the intestine quicker than fish fed the β-glucan or control diet.TNFα was also found to be expressed at significantly higher levels in the distal intestine.
To understand how the changes in magnitude of these immune genes (seen to be upregulated in the GALT leucocytes) may impact the immune response, fish were challenged with A. salmonicida for 48hrs.A. salmonicida was chosen due to the known effects of β-glucan on the proinflammatory response [40] and has previously shown to improve survival A. salmonicida with higher levels of inclusion into the diet.However, in future studies a titrated dose may provide a greater resolution in teasing out key changes in datasets when exploring the immune modulatory capacity of feed additives.Our results show that in RNA extracted from head kidney the key antimicrobial genes SAA and HAMP show a lower magnitude of response to A. salmonicida infection in the fish fed the β-glucan and Protec™ diets compared to those fed control diet.A similar result was also observed in the distal intestine for HAMP whilst SAA was conversely affected.These results suggest that fish are offered some protection due to these immunostimulatory molecules found in the diets, due to their innate immune response already being modulated in PBS-injected fish.IL-1β was observed to be upregulated in all diets and in both the head kidney and distal intestine showing similar levels of relative expression in both the head kidney and distal intestine after infection with A. salmonicida.IL-1β has been previously demonstrated to be modulated by supplementation with β-glucans [3,7,9,40].Genes relating to mucins, which are likely to be involved in mucus production [41] were observed to show a much lower magnitude of response in the Protec™ fed fish, however this is likely due to these genes being expressed at much higher levels in PBS treated samples.Significantly higher numbers of goblet cells as observed in the histology which may offer increased mucosal immunological activity.CLEC4T has a lower magnitude of response in the Protec™ diets after infection, however, this gene was expressed at much higher levels in PBS-treated samples compared to the control and β-glucan diets.C-type lectins have previously been identified in carp as potential receptors for β-glucans in carp [4] and have been linked to upregulation of the C-type lectin signalling pathways.Relative expression of CLEC4T in A. salmonicida infection suggested there was significant upregulation of this gene in both the control and β-glucan diets, generating similar relative expression as observed in the Protec™ diets, indicating the C-type lectin pathways is also triggered by A. salmonicida infection.Secretory immunoglobulin M had significantly higher relative expression compared to fish fed both the control and β-glucan samples in PBS treated samples which has previously been shown to be increased in the Protec™ diet [8].To identify potential benefits for in terms of disease resistance and immunity a mortality trial would need to be carried out.
Research into gut immune function is limited in fish, with responses unique to individual species [42] more research is needed to identify exactly how functional feeds modulate the gut and subsequently cause a systemic immune response in each species.Traditionally, key immune organs such as the spleen and the head kidney are the focus of immune challenges due to their immunological importance in fish.However, the use of GALT leucocytes in this experiment has elucidated that GALT is able to mount a sizeable immune response to stimulation and this response can be modulated by diet and as such may be a useful tool to measure immune response against pathogens.One method to reduce economic loss due to disease is to train immunity through the modulation of GALT tissue and may provide a viable alternative to allow for the reduction in the use of antibiotics in aquaculture.Previous research by Petit et al [43] has already demonstrated that macrophages can be trained to protect against bacterial stimuli in-vitro through the use of immunostimulants (β-glucans and peptidoglycans) however further research would be needed to confirm that the same would be true in-vivo.Research into zebrafish injected with β-glucans, revealed short lived protection against spring viremia of carp virus (SVHV) after 7 days but not after 35 days post infection [44].The data presented demonstrates the potential of functional diets and more studies are required to understand how gut immunity can be modulated to better protect against infection.Fig. 8.Immune challenge with A. salmonicida for the head kidney and distal intestine.All results were quantified by qPCR.qPCR results were normalised against three housekeeping genes.Graphs show fold change relative to samples stimulated with PBS for each diet.a, b, and c denote significant difference where p < 0.05.Significance was identified using an Anova followed by a Tukey post hoc test.

Fig. 3 .
Fig. 3. Cellular markers for B-Lymphocytes (sIgM & sIgT), T-Lymphocytes (CD4 & CD8α) and Dendritic Cells (CD108 & MHCIIβ) in isolated GALT leucocytes from rainbow trout fed two different functional diets.The marker gene expression was determined by qPCR and is presented as relative expression after normalisation against three reference genes.Bars are means plus SEMs of 6 fish.

Fig. 4 .
Fig. 4. Basal gene expression of key immune markers in isolated GALT leucocytes from Rainbow Trout fed two functional diets.Basal gene expression of proinflammatory and viral genes was determined by qPCR and is presented as relative expression after normalisation against three reference genes.Bars are means plus SEMs of 6 fish.Significance was tested via two-way Anova, followed by a post hoc Tukey test where a and b denote significance where p < 0.05.

Fig. 6 .
Fig. 6.Response to poly I:C.All results were quantified by qPCR.qPCR results were normalised against three housekeeping genes.A, B, C are relative expression of stimulated samples verses their unstimulated samples for each diet.D represents the fold change of each diet calculated from the relative expression.a and b denote significant difference where p < 0.05.Significance was identified using an Anova followed by a Tukey post hoc test.

Fig. 7 .
Fig. 7. Response to rIL-1β.All results were quantified by qPCR.qPCR results were normalised against three housekeeping genes.A, B, C are relative expression of stimulated samples verses their unstimulated samples for each diet.D represents the fold change of each diet calculated from the relative expression.a and b denote significant difference where p < 0.05.Significance was identified using an Anova followed by a Tukey post hoc test.

Table 1
Dietary formulation and proximate composition of the experimental diets (% dry matter).

Table 3
qPCR primer sequences, product size and GenBank accession numbers.

Table 4
Growth parameters from the feeding trial for each diet with means and standard deviations given.a and b denote a significant difference where p < 0.05.

Table 5
Means and standard deviations of scoring parameters for histology, a,b and c denote significant difference.