Molecular indices of viral disease development in wild migrating salmon†

Abstract

Infectious diseases can impact the physiological performance of individuals, including their mobility, visual acuity, behavior and tolerance and ability to effectively respond to additional stressors. These physiological effects can influence competitiveness, social hierarchy, habitat usage, migratory behavior and risk to predation, and in some circumstances, viability of populations. While there are multiple means of detecting infectious agents (microscopy, culture, molecular assays), the detection of infectious diseases in wild populations in circumstances where mortality is not observable can be difficult. Moreover, if infection-related physiological compromise leaves individuals vulnerable to predation, it may be rare to observe wildlife in a late stage of disease. Diagnostic technologies designed to diagnose cause of death are not always sensitive enough to detect early stages of disease development in live-sampled organisms. Sensitive technologies that can differentiate agent carrier states from active disease states are required to demonstrate impacts of infectious diseases in wild populations. We present the discovery and validation of salmon host transcriptional biomarkers capable of distinguishing fish in an active viral disease state [viral disease development (VDD)] from those carrying a latent viral infection, and viral versus bacterial disease states. Biomarker discovery was conducted through meta-analysis of published and in-house microarray data, and validation performed on independent datasets including disease challenge studies and farmed salmon diagnosed with various viral, bacterial and parasitic diseases. We demonstrate that the VDD biomarker panel is predictive of disease development across RNA-viral species, salmon species and salmon tissues, and can recognize a viral disease state in wild-migrating salmon. Moreover, we show that there is considerable overlap in the biomarkers resolved in our study in salmon with those based on similar human viral influenza research, suggesting a highly conserved suite of host genes associated with viral disease that may be applicable across a broad range of vertebrate taxa.

Keywords: aquaculture; disease biomarkers; host transcriptome; salmon; viral disease; wild populations.

Figure 1:

Schematic of viral disease development (VDD) discovery, refinement, validation and application datasets. The VDD discovery dataset was identified from published microarray viral challenge studies that included five RNA virus species. In house (MGL) IHNv challenge microarray studies across four salmon species were used to refine the VDD panel. Analytical validations of the qRT-PCR assays developed to 45 biomarkers within the VDD panel was performed using independent in-house studies that tested discrimination abilities of the proposed VDD between latent and disease-associated viral infections across tissues, salmon and viral species, as well as differentiation of fish undergoing viral, bacterial, and parasitic diseases. The VDD panel was then applied to wild migrating Sockeye salmon smolts to discern whether wild fish infected with IHNv could be identified in a VDD state.

Figure 2:

Time-course of expression of VDD genes post IHNv ip challenge, by tissue. (A) Sockeye, (B) Atlantic and (C) Chum salmon. Only samples from IHNv infected fish are included in the displayed post controls time course data. For Sockeye this included all 45 samples, while one Day 1 sample was excluded in the Atlantic salmon time course, and 11 samples from different time points were excluded in the Chum salmon time course.

Figure 3:

Expression of VDD genes on a time-course post IHNv waterborne challenge, by tissue. (A) Sockeye salmon (B) Atlantic salmon (C) Chum salmon. Post controls time course samples represent IHNv infected fish only (26 for Sockeye, 37 for Atlantic and 18 for Chum salmon). Only time points with data for at least two samples are displayed.

Figure 4:

PCA classification of salmon post IHNv challenge, by species, challenge-type and tissues, as visualized by principle component analysis. (A) Sockeye salmon: (i) ip-challenge by tissue (head kidney, liver, gill, respectively) and (ii) waterborne-challenge by tissue (head kidney, gill, respectively). (B) Atlantic salmon by challenge-type and tissue (ip: kidney, waterborne: head kidney, gill, respectively). (C) Chum salmon by challenge-type and tissue (ip: head kidney, waterborne: head kidney, gill, respectively).

Figure 5:

PCA classification of five tissues from farmed Chinook salmon undergoing an outbreak of jaundice/anemia. Analysis based on (A) a 40 biomarker VDD panel, (B) a 17 biomarker VDD panel identified through gene-shaving and (C) a 7 biomarker VDD panel derived from gene shaving. In each plot, samples with viral jaundice are shown in blue and healthy controls in peach, with shades and shapes within each depicting different tissues, as illustrated in the panel legend under (B). The single viral jaundice sample not properly classifying showed weak lesions and low viral loads, and is suspected to represent a fish in recovery.

Figure 6:

Principle component analysis depicting the differentiation of (A) Atlantic salmon farm audit fish (based on mixed tissue cDNA) diagnosed with viral (HSMI) versus bacterial diseases (mouth rot, winter ulcer, rickettsiosis, and vibriosis) based on the full 40 biomarker VDD panel (left), 15 biomarker VDD panel (mid) and 9 biomarker VDD panel (right) derived from gene shaving. (B) Chinook salmon farm audit fish diagnosed with viral (jaundice/anemia) versus bacterial (rickettsiosis and vibriosis) and parasitic (Loma) diseases based on the full 36 biomarker VDD panel (left), 25 biomarker VDD panel (mid), and 9 biomarker VDD panel (right) derived from gene shaving.

Figure 7:

Gene expression box plots of top 11 biomarkers in the VDD panel for Atlantic (left) and Chinook (right) salmon from the farm audit study, contrasting median expression levels between viral (HSMI or jaundice) and bacterial (rickettsiosis, vibriosis, mouth rot, winter ulcer, and bacterial kidney disease [BKD]) or parasitic (Loma) diseases.

Figure 7:

Gene expression box plots of top 11 biomarkers in the VDD panel for Atlantic (left) and Chinook (right) salmon from the farm audit study, contrasting median expression levels between viral (HSMI or jaundice) and bacterial (rickettsiosis, vibriosis, mouth rot, winter ulcer, and bacterial kidney disease [BKD]) or parasitic (Loma) diseases.

Figure 7:

Gene expression box plots of top 11 biomarkers in the VDD panel for Atlantic (left) and Chinook (right) salmon from the farm audit study, contrasting median expression levels between viral (HSMI or jaundice) and bacterial (rickettsiosis, vibriosis, mouth rot, winter ulcer, and bacterial kidney disease [BKD]) or parasitic (Loma) diseases.

Figure 7:

Gene expression box plots of top 11 biomarkers in the VDD panel for Atlantic (left) and Chinook (right) salmon from the farm audit study, contrasting median expression levels between viral (HSMI or jaundice) and bacterial (rickettsiosis, vibriosis, mouth rot, winter ulcer, and bacterial kidney disease [BKD]) or parasitic (Loma) diseases.

Figure 8:

Principle component analysis of 39-biomarker VDD panel applied to non-destructive gill tissue from 213 wild migrating Sockeye salmon smolts. Coloring depicts VDD separation of most fish carrying high IHNv loads (log copy number >2).

Figure 9:

Heatmaps for the IHNv and farm audit validation datasets showing the up-regulation of VDD biomarkers in fish tissues under viral challenge (A) and in natural disease outbreaks (B). Heatmap depicting relative gene express (2^−ΔΔCt method) is scaled from brown (down-regulated) to teal (up-regulated) with darker colors indicating higher expression differentials as indicated by the color key on the top right. Grey rows indicate that no working assay was available for the corresponding genes. The Sockeye, Atlantic and Chum IHNv datasets depicted in (A) include heatmaps for multiple tissues (head kidney, gill, liver and spleen) from fish that were injected with IHNv (top), exposed to IHNv in waterbath (bottom), and controls that were not injected or exposed (both). The Jaundice Chinook dataset (B, left) includes heapmaps for head kidney, gill, liver, heart and spleen samples and Farm Audit Salmon datasets (B, right) show heatmaps for mixed-tissue samples.

Figure 10:

Gene network constructed based a 27-gene VDD panel identified to mammalian genes based on their gene symbols, showing key regulators (IFNG highlighted in green) and nearest neighbors (no highlight) overlaid onto the cell to show localization of protein activity. Analysis performed in Pathway Studio (Elsevier, Amsterdam). Fourteen VDD genes mapping to mammalian genes (IFH1, HERC6, RSAD2, DDX58, CD9, MX1, IFIT5, STAT1, XAF1, MX1, RNF213, TRIM21, USP18) are found within this gene network (highlighted in blue).