Interaction between dietary fatty acids and genotype on immune response in Atlantic salmon (Salmo salar) after vaccination: A transcriptome study

Abstract

A pivotal matter to aquaculture is the sourcing of sustainable resources as ingredients to aquafeeds. Levels of plant delivered oils as source of fatty acids (FA) in aquafeeds have reached around 70% resulting in reduced levels of long-chain omega-3 polyunsaturated fatty acids (LC n-3 PUFA), such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), in salmon fillet composition. EPA and DHA can modulate inflammation and immune response, so it is crucial to understand how fish immune response is affected by low LC n-3 PUFA diet and if this diet can have a detrimental effect on vaccine response. Atlantic salmon (Salmo salar) can produce EPA/DHA from α-linolenic acid (ALA) and this endogenous capacity can be explored to develop families with higher tolerance to low LC n-3 PUFA diets. Here we analyze innate and adaptive immune response in Atlantic salmon to a commercial vaccine after being fed low levels of EPA and DHA, and we also compare three strains of salmon selected by their endogenous capacity of synthesizing LC- n-3 PUFA. A total of 2,890 differentially expressed genes (DEGs) were identified (p-value adjusted < 0.1) when comparing vaccinated fish against control non-vaccinated. Gene ontology (GO) and KEGG analysis with 442 up/downregulated genes revealed that most DEGs were both related to immune response as well as part of important immune related pathways, as "Toll-like receptor" and "Cytokine-Cytokine interaction". Adaptive response was also addressed by measuring antigen specific IgM, and titers were significantly higher than in the pre-immune fish at 62 days post-immunization. However, diet and strain had no/little effect on vaccine-specific IgM or innate immune responses. Atlantic salmon therefore display robustness in its response to vaccination even when feed low levels of LC n-3 PUFA.

Fig 1. P-value distribution histogram.

Ballgown stattest function used to specify effect of treatment and covariates. (A) Control fish (n = 6) versus vaccinated fish (n = 24). (B) Vaccinated fish only and strain as covariate (n = 24). (C) Vaccinated fish only and diet as covariate (n = 24). (D) Interaction analysis of diet and strain in vaccinated fish (n = 24).

Fig 2. Number of genes differentially expressed in head kidney induced by vaccination, one day after immunization.

Control fish (n = 6) versus vaccinated fish (n = 24).

Fig 3. Gene ontology enrichment analysis of the differentially expressed genes from vaccinated fish comparing to control fish.

GO annotation based on Salmo salar OrgDb object. BP—Biological Processes, MF—Molecular Function and CC—Cellular Components. Color gradient from red to blue, where red indicates high enrichment (low p.adjust) and blue indicates low enrichment (high p.adjust). Dot size corresponds to count (number of DEGs in each term).

Fig 4. KEGG analysis of the differentially expressed genes from vaccinated fish comparing to control fish.

Color gradient from red to blue, where red indicates high enrichment (low p.adjust) and blue indicates low enrichment (high p.adjust). Dot size corresponds to count (number of DEGs in each pathway).

Fig 5. KEGG pathway map.

Cytokine-cytokine receptor interaction network map for DEGs of vaccinated fish. Red squares represent upregulated genes, while yellow shows downregulated genes. Grey are genes with lower expression levels. Fold change values are represented in log2 scale.

Fig 6. UpSet plot showing overlapping of genes identified by each of the different analyses.

The bars show the overlap between the indicated motifs below: Expressed_Genes (all genes expressed in head kidney), Vaccine (DEGs in vaccinated fish), Strain (only vaccinated fish with strain as covariate) and Interaction (interaction of strain and diet in vaccinated fish only).

Fig 7. ELISA plot.

Comparison of IgM titers in fish 62 days after immunization. Plasma from preimmune (n = 36) and vaccinated fish (immune, n = 36) were used to perform ELISA assay. Data expressed as OD adjusted for background. T-test used to calculate difference between immune and preimmune and ANOVA for multiple comparison between each variable (strain and diet).

Fig 8. Comparison of qPCR and RNA-seq.

Bar plot showing similar pattern of down/upregulation for tested genes on both analyses. Fold change values are represented in log2 scale from both RNA-seq and qPCR. 18s and ef1a were used as reference genes to calculate gene expression levels for qPCR data from three biological and two technical replicates.

Fig 9. qPCR of liver samples from day 1 and day 62 after immunization.

Box plot showing expression levels of three desaturase genes, fadsd5, fadsd6_a and fadsd6_b. This assay was performed with three biological and two technical replicates. TukeyHDS test was used to compare all the variables (strain/diet/time) against each other, and qval < 0.05 was considered significant (significant comparison shown in Table 4).