Host-Parasite Interaction of Atlantic salmon (Salmo salar) and the Ectoparasite Neoparamoeba perurans in Amoebic Gill Disease

Abstract

Marine farmed Atlantic salmon (Salmo salar) are susceptible to recurrent amoebic gill disease (AGD) caused by the ectoparasite Neoparamoeba perurans over the growout production cycle. The parasite elicits a highly localized response within the gill epithelium resulting in multifocal mucoid patches at the site of parasite attachment. This host-parasite response drives a complex immune reaction, which remains poorly understood. To generate a model for host-parasite interaction during pathogenesis of AGD in Atlantic salmon the local (gill) and systemic transcriptomic response in the host, and the parasite during AGD pathogenesis was explored. A dual RNA-seq approach together with differential gene expression and system-wide statistical analyses of gene and transcription factor networks was employed. A multi-tissue transcriptomic data set was generated from the gill (including both lesioned and non-lesioned tissue), head kidney and spleen tissues naïve and AGD-affected Atlantic salmon sourced from an in vivo AGD challenge trial. Differential gene expression of the salmon host indicates local and systemic upregulation of defense and immune responses. Two transcription factors, znfOZF-like and znf70-like, and their associated gene networks significantly altered with disease state. The majority of genes in these networks are candidates for mediators of the immune response, cellular proliferation and invasion. These include Aurora kinase B-like, rho guanine nucleotide exchange factor 25-like and protein NDNF-like inhibited. Analysis of the N. perurans transcriptome during AGD pathology compared to in vitro cultured N. perurans trophozoites, as a proxy for wild type trophozoites, identified multiple gene candidates for virulence and indicates a potential master regulatory gene system analogous to the two-component PhoP/Q system. Candidate genes identified are associated with invasion of host tissue, evasion of host defense mechanisms and formation of the mucoid lesion. We generated a novel model for host-parasite interaction during AGD pathogenesis through integration of host and parasite functional profiles. Collectively, this dual transcriptomic study provides novel molecular insights into the pathology of AGD and provides alternative theories for future research in a step towards improved management of AGD.

Keywords: Atlantic salmon (Salmo salar); Neoparamoeba perurans; amoebic gill disease; aquaculture; dual RNA-Seq; host-parasite interaction; immunity.

Figure 1

Schematic showing datasets and analytical approach to infer host-parasite interaction in amoebic gill disease (AGD). HOST, Atlantic salmon (Salmo salar); PARASITE, Neoparamoeba perurans; SP, spleen; A, AGD-affected; HK, head kidney; G, gill; D, distal to the lesion; L, lesion; C, cultured floating N. perurans trophozoites.

Figure 2

Summary of the host transcriptomic response during amoebic gill disease (AGD) progression. Differential gene expression and significantly enriched gene ontology (GO) terms are shown for the local (A–C) and systemic (D–I) response to AGD in Atlantic salmon. Heat maps show hierarchical clustering of differentially expressed genes (rows) with differential expression among replicates from a naïve (C) and AGD-affected (A) Atlantic salmon. Expression values are log₂-transformed and median-centered by gene. (A) The local response is characterized by differential gene expression (fold change > 2; corrected P-value < 0.01) and includes a gill (G) biopsy distal (D) to the lesion (L) of an AGD-affected fish. The systemic response among head kidney (HK) (D) and spleen (SP) (G) replicates is characterized by differential gene expression (fold change > 2; corrected P-value < 0.05). Enriched gene ontology (GO) terms (hypergeometric test, Bonferroni-adjusted P < 0.05) among the differentially expressed genes for the gills (B, C), head kidney (E, F) and spleen (G, H) along with the gene ratio for the genes that map to each term. The majority of the enriched terms are related to host defense and immune response.

Figure 3

Sub-networks for the top differentially connected genes (DCGs) likely to regulate the transcriptomic response during amoebic gill disease (AGD) in Atlantic salmon. (A) AGD network of 8 transcription factors (TFs) among the top 20 DCGs using the PCIT algorithm. All nodes are represented by ellipses except for genes coding key regulators (TFs) which are diamond shaped. Nodes are orange for gill, green for head kidney and purple for spleen. The size of the nodes is relative to the normalized mean expression values in all samples. (B) Subnetworks of top differentially connected TFs. The networks created with the most differentially connected genes between naïve and AGD-affected networks with zinc finger protein OZF-like (OZF) as the key regulator with the greatest number of connections in the AGD-affected network, while zinc finger 135 (znf135) lost the majority of its connections in the AGD-affected network. (C) Heat map shows hierarchical clustering of differential expression of connected genes (rows) in the OZF network among replicates from a naïve (C) and AGD-affected (A) Atlantic salmon in the head kidney (HK), spleen (SP) and gill (G). The AGD-affected gill data is represented by the lesion (L), a biopsy distal to the lesion (D) and naïve gill (C). Expression values (CPM) are log₂-transformed and median-centered by gene.

Figure 4

Bar plot of the bacterial taxa identified from sequence reads. (A) Bacterial taxa identified in the unmapped Neoparamoeba perurans transcriptome. While the community diversity of 142 species was plotted to show visual diversity, only the top 12 candidates based on proportional reads (>1% TPM) are denoted in the legend. (B) Bacterial taxa identified in the unmapped Atlantic salmon gill transcriptome from naïve fish, biopsies distal to AGD lesions, and the AGD lesion data. The top 15 candidates based on proportional reads (>1% TPM) are denoted in the legend.

Figure 5

A theoretical model for the host-parasite interaction between Neoparaomoeba perurans candidate genes and Atlantic salmon during the pathogenesis of amoebic gill disease (AGD). The pathway was generated from the differential gene expression and network analyses together with the key KEGG pathways mapped and visualized (sasa04310, sasa05168, sasa05132, hsa04151, hsa05146) for the host and the parasite. Pathogen invasion is facilitated by degradation of host mucus and epithelia. To maintain a pathogen friendly environment on the gill N. perurans releases factors to decrease ammonia and reactive oxygen species released by the host. Actin rearrangement of the pathogen and the host facilitates attachment. Virulence factors are modulated by the pathogen master two component PhoPQ virulence regulatory system. Downregulation of the host wnt, Ap-1 and PI3K/Akt signaling pathways supports pathogen immune evasion, proliferation, and survival. The Th1/17 cell differentiation pathway is upregulated inducing innate and adaptive immune responses in the host. NP signifies N. perurans genes, upregulated genes are green, downregulated brown, not differentially expressed are white, italicized genes were not identified in our dataset. CASP3, caspase-3; COL, collagen (various); cox2, cyclooxygenase 2; ctnnb1, β-catenin; FN, fibronectin; FOS, fos; FZD, frizzled; HAAF, hemagglutinin/amebocyte aggregation factor-like; HSPB1, heat shock protein beta-1-like; HVEM, tumor necrosis factor receptor superfamily; IFNGR1, interferon gamma receptor 1; il10, interleukin 10; IL10RB, interleukin-10 receptor subunit beta-like; IL12, interleukin-12; IL1R1, interleukin-1 receptor type 1-like; IL1β, interleukin-1 beta; IL6, interleukin-6; IL8, interleukin 8; ITGAM, integrin alpha-X-like; ITGB2, integrin beta-2; JUNB, junb; LAM, laminin; LIGHT, tumor necrosis factor ligand superfamily; mhci, major histocompatibility complex class I; MHCII, major histocompatibility complex class II; MUC2, mucin-2-like; NFκB1, nuclear factor NF-κ-β p105 subunit-like; NOS2, nitric oxide synthase 2; NpACA1, prokumamolisin activation domain containing protein; NpAPRA, AprA protease; NpCBP, cathepsin-B; NpHCP, hybrid cluster protein; NpMADS, MADS-box transcription factor; NpPHB, prohibitin; NpPhoPQ, PhoPQ-activated pathogenicity-related protein; NpSNF7, vacuolar sorting protein SNF7; NpSPP, signal peptide peptidase; NpV-ATPase, vacuolar proton ATPase; NpPAK, p-21 activated kinase; Nppka, protein kinase A; NpPKC, protein kinase C; pge2, prostaglandin E2; PI3K, phosphatidylinositol 3’-kinase; RELA, putative transcription factor p65 homolog; SFRP, secreted frizzled-related protein; TGFβ, transforming growth factor beta; TIGIT, T-cell immunoreceptor with Ig and ITIM domains; TLR2/4, toll-like receptor 2 and 4; tnfa, tumor necrosis factor alpha; WNT, protein Wnt; ZAP, zinc finger antiviral protein.