Figures
Abstract
In fish, the skin is a multifunctional organ and the first barrier against pathogens. Salmonids differ in their susceptibility to microorganisms due to varied skin morphology and gene expression patterns. The brown trout is a salmonid species with important commercial and ecological value in Europe. However, there is a lack of knowledge regarding the genes involved in the immune response and mucus secretion in the skin of this fish. Thus, we characterized the skin transcriptome of anadromous brown trout using next-generation sequencing (NGS). A total of 1,348,306 filtered reads were obtained and assembled into 75,970 contigs. Of these contigs 48.57% were identified using BLAST tool searches against four public databases. KEGG pathway and Gene Ontology analyses revealed that 13.40% and 34.57% of the annotated transcripts, respectively, represent a variety of biological processes and functions. Among the identified KEGG Orthology categories, the best represented were signal transduction (23.28%) and immune system (8.82%), with a variety of genes involved in immune pathways, implying the differentiation of immune responses in the trout skin. We also identified and transcriptionally characterized 8 types of mucin proteins–the main structural components of the mucosal layer. Moreover, 140 genes involved in mucin synthesis were identified, and 1,119 potential simple sequence repeats (SSRs) were detected in 3,134 transcripts.
Citation: Malachowicz M, Wenne R, Burzynski A (2017) De novo assembly of the sea trout (Salmo trutta m. trutta) skin transcriptome to identify putative genes involved in the immune response and epidermal mucus secretion. PLoS ONE 12(2): e0172282. https://doi.org/10.1371/journal.pone.0172282
Editor: Patrick Prunet, Institut National de la Recherche Agronomique, FRANCE
Received: August 3, 2016; Accepted: February 2, 2017; Published: February 17, 2017
Copyright: © 2017 Malachowicz et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The sea trout raw transcriptome sequences from this study were deposited in the NCBI Sequence Read Archive (SRA) under the accession number [SRP075946].
Funding: This study was partially funded by project: No. 397/N-cGRASP/2009/0 of the Ministry of Science and Higher Education in Poland to RW and statutory topic IV.1 in the IO PAS.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Brown trout (Salmo trutta) is a salmonid fish species of significant ecological and commercial value that is, naturally distributed throughout all of Europe and in adjacent waters [1]. This species represent two different life strategies: resident and migratory [2]. An anadromous form, called sea trout (Salmo trutta m. trutta) migrates to the sea for feeding and returns to its natal rivers for spawning [3]. The highest total (commercial and recreational) sea trout catches in the Baltic Sea, above 1000 tons, were recorded in the early and late 1990s. Since then, the total sea trout catches have been decreasing and are now below 500 tons [4]. Several factors have contributed to this decrease, including: pollution, environmental degradation and diseases [5–7]. One of the most common diseases that deplete sea trout populations is furunculosis, which is caused by the gram-negative bacteria Aeromonas salmonicida [8]. This disease is transmitted by close contact between infected and healthy fish, it begins with hypertrophy of the epidermis, followed by ulcers spreading into the subcutaneous tissue, and leads to general infection and septicemia. Teleostean fish have evolved innumerable host-pathogen relationships, some of which are unique to only a few or a single species [9]. Even related species can show differences in susceptibility to pathogens due to varied skin morphology [9] and gene activity [10]. Coho salmon (Oncorhynchus kisutch) is less susceptible to infectious hematopoietic necrosis virus (IHNV) than rainbow trout (Oncorhynchus mykiss) or Atlantic salmon (Salmo salar) [11]. Moreover, rainbow trout and Atlantic salmon are more resistant against A. salmonicida, while brown trout is the most susceptible [12].
For a pathogen such as A. salmonicida to infect the host, innate barriers must first be crossed, including a shield of scales, skin and mucus [13, 14]. Fish skin traps and washes away pathogens and contains many immune factors, such as: antimicrobial peptides [15, 16], C-reactive proteins [17], complement components [18], immunoglobulins [19], lectins [20], lysozymes [21] and proteases [22]. In addition to fish skin functioning as a mechanical barrier, it is a multi-functional organ (with roles in ion regulation, excretion, ion and thermal regulation) and metabolically active tissue [23–25]. These functions are possible due to the complex structure and cell composition of the skin [23]. The important defensive role of the skin in the innate immune system is well known and has been studied in several fish species [22, 26–29]. However, there is a lack of genomic information such as the skin transcriptome of sea trout.
One of the most distinctive features of fish skin is the production of mucus by the unicellular glands of the epidermis, mainly goblet cells and club cells [28]. Mucus protects the underlying epithelium from chemical, enzymatic and mechanical damage [30]. Bacterial infection or a change in diet can increase mucus secretion and change the glycosylation pattern and thus the bacterial adherence to the skin, potentially increasing pathogen removal [14, 31]. The complexity of glycan epitopes and sialylation seem to be key factors in the host–pathogen interaction between Atlantic salmon and A. salmonicida [14]. The goblet cells in the skin epidermis are responsible for the production and maintenance of the mucosal layer via the synthesis and secretion of mucins [27, 32]. Mucins are high molecular weight glycoproteins with a large number of tandem repeats that are rich in proline, threonine and serine (the PTS domain) [33]. These PTS regions tends to be poorly conserved [30]. In vertebrates, over 20 mucin types have been characterized thus far. Mucins can be divided into two structurally distinct subfamilies: large secreted gel- forming mucins (SGFM) and membrane-bound mucins [34]. The SGFM subfamily include Muc2, Muc5AC, Muc5B, Muc6, and Muc19. These mucins contain a von Willebrand factor type D (VWD) domain, cysteine-rich C8 (C8) domain and C-terminal cysteine-knot (CK) domain, which are all involved in the oligomerization of mucins [35]. By contrast, membrane-bound mucins (Muc1, Muc3, Muc4, Muc7-9, and Muc12–22) do not oligomerize and are characterized by specific domains such as the C-terminal sea urchin sperm protein-enterokinase-agrin (SEA) domain, an epidermal growth factor (EGF) or EGF-like (EGF-L), and a transmembrane domain (TM) [36]. Mucins constitute an important part of the mucosal defense against infection [37, 38]. Furthermore, based on mucin studies of Atlantic salmon, skin and intestinal mucins have different biochemical characteristics, which in turn affect their binding of the pathogens [14]. The density of the skin mucins was lower and glycans were shorter and less branched [14]. Recently, function of fish mucins was confirmed in few studies. Pérez-Sánchez et al. [39] challenged in the study of sea bream, specific expression of mucins under different nutritional conditions, parasite infections and the tissue-specific expression. Mucin structure and tissue-expression were also studied in carp and zebrafish [40, 41]. Moreover, stressed rainbow trout had greater proportions (two times higher) of goblet cells that contained mucins with sialic acid compared to control [42]. Peatman et al. [43] used utilized high-density Affymetrix microarrays to examine gene expression profiles in channel and blue catfish skin upon A. hydrophila infection. Several recent studies have described mucin sequences in teleost fish [27, 44] but mucin genes have not been characterized in the transcriptomes of different tissues from brown trout, such as the liver, kidney, gut, gill, embryo, testis, intestine, bones, muscle, heart, brain and ovary [45, 46].
In this study, we assembled and characterized the skin transcriptome of sea trout using the Roche GS-FLX 454 pyrosequencing system. A large number of genes involved in immune reactions and mucus secretion were identified. Moreover, this study is the first to present and characterize trout mucins. The generated data add to the growing sequence database for brown trout and provide the first representation of the trout skin transcriptome and valuable resources for functional genomics.
Materials and methods
Ethics statement
This study was performed in accordance with the three Rs for the humane use of animals in scientific research and was approved by the Local Ethics Committee on Animal Experimentation of the Inland Fisheries Institute of Olsztyn, Poland (Nr 51/N/2005).
RNA extraction and pyrosequencing
Six specimens of sea trout were obtained from the Department of Salmonid Fish Breeding of the Inland Fishery Institute of Rutki, Poland in 2010. Extraction of total RNA from the deep frozen skin tissue upstream the anal fin was performed using a GenElute Mammalian Total RNA Miniprep Kit (Sigma) according to the manufacturer’s protocol, with minor modifications. Homogenization was performed in custom buffer composed of 100 mM Tris, 1.4 M NaCl, 20 mM EDTA, 2% CTAB, proteinase K (20 mg/ml) and 0.03 mM 2-mercaptoethanol. The incubation time was 30 minutes. The DNA was removed following the protocol provided by the manufacturer. The RNA isolate was stored at -20°C.
Total RNA was used for cDNA synthesis using the SMART (Switching Mechanism At 5' end of RNA Template) kit from BD Biosciences Clontech. The cDNA library, was later normalised using the duplex-specific nuclease (DSN) method [47]. Approximately 15 μg of cDNA were used for sequencing (Roche GS-FLX) at the Max Planck Institute for Molecular Genetics (Berlin, Germany), following described procedures [48].
Assembly and annotation
The raw reads produced from sea trout sequencing were cleaned by removing adapters and low quality sequences. The quality score limit was set to 0.05 and the reads shorter than 50 bp were discarded using CLC Genomics Workbench software (v.7.5.5, CLC Bio, Qiagen, Aarhus, Denmark). The quality reads were assembled into contigs using de-Bruijn graphs in CLC Genomics Workbench. The generated contigs were searched against the National Center of Biotechnology Information non-redundant protein database (NCBI nr), Swiss-Prot (ftp://ftp.ncbi.nlm.nih.gov/blast/db), Clusters of Eukaryotic Orthologous Groups (KOG; ftp://ftp.ncbi.nih.gov/pub/COG/KOG) and the Teleost Ensembl database. The teleost fish database was created by joining reference protein sequences of 11 teleost fish species (Amazon molly—Poecilia formosa, cave fish—Astyanax mexicanus, cod—Gadus morhua, fugu—Takifugu rubripes, medaka- Oryzias latipes, southern platyfish—Xiphophorus maculatus, spotted gar—Lepisosteus oculatus, three-spined stickleback—Gasterosteus aculeatus, green spotted puffer—Tetraodon nigroviridis, Nile tilapia—Oreochromis niloticus, zebrafish—Danio rerio) obtained from the Ensembl database (http://www.ensembl.org). Annotations were completed using the BLASTX tool implemented in BLAST+ (v.2.2.29) [49], with an E-value cut-off of 10−5. A list of the top BLAST hits was obtained using Blast2GO software [50] and was based on the best-aligned results. Sequences with positive matches in the NCBI nr protein database were further searched for Gene Ontology (GO) terms [51] with Blast2GO. The level 2 GO terms were retrieved and classified into three categories: cellular component, biological process and molecular function. Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were assigned using the single directional best-hit (SBH) method in the KEGG Automatic Annotation Server (KAAS) [52]. This tool assigns KEGG Orthology (KO) identifiers to sequences based on their similarity, which is directly linked to an object in the KEGG pathway map or the BRITE functional hierarchy [52]. Moreover, sequences without significant BLAST hits were searched against the noncoding RNA database using Rfam 12.0 (http://rfam.xfam.org) [53].
The results from this study were also compared to the previously published trout multi-tissue transcriptome from the River Hayle and River Teign in southwest England, which was obtained from the authors [45]. Raw data are available in the Gene Expression Omnibus (GEO) under accession number GSE45637. Contig identification and GO term annotation were completed by same methodology as previously described, and the GO results from both transcriptomes were compared. To find enriched GO terms between skin and other tissues, Fisher’s exact test was performed with Blast2GO, applying a False Discovery Rate (FDR) <0.05 [54]. We also investigated nucleotide-level comparisons by mapping skin transcriptome reads to the multi-tissue transcriptome using CLC Genomics Workbench with default parameters.
Mucin bioinformatics
Based on the top BLAST hit descriptions, contigs annotated as mucin genes were extracted from skin and multi-tissue trout transcriptomes and used for further analysis. Sequences were checked and confirmed for homologs in the GenBank nr database and Salmo salar genome [55], using the program BLASTX and BLASTN (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The coding regions (CDS) of each contig sequence were predicted and translated using GENSCAN (http://genes.mit.edu/GENSCAN.html) [56]. The functional domains and motifs were identified using Simple Modular Architecture Research Tool (SMART) software, which allows the identification and annotation of domain architectures (http://smart.embl-heidelberg.de) [57]. In addition, transmembrane domains (TM) were predicted with TMHMM v. 2.0 software [58]. To further scrutinize I-mucin (I-Muc) genes, a representative reference teleost database was created. All protein sequences described as intestinal-mucin, mucin-5AC, mucin-5B and mucin-2, which are available at the GenBank database (www.ncbi.nlm.nih.gov) [59], were downloaded (771 sequences), including the Baltic cod transcriptome [60]. Next, trout sequences were searched against the created database (E-value cut-off = 10−10). Furthermore, all homologous bony fish proteins were analyzed with SMART to identify protein domains. Only sequences with annotated VWD and C8 domains, which are typical for SGFM, were selected, with one representative species per family, and each type of mucin was represented. A total of 15 mucin sequences from six species (five families) were used in the reconstruction of I-Muc relationships, and 18 sequences from four species (three families) were used for phylogenetic analysis of SGFM (Muc5, Muc2 and I-Muc). For phylogenetic analysis, VWD and C8 domain sequences were extracted and aligned using MUSCLE via MEGA v.7.0.14 [61] with default parameters. The best substitution model was chosen using MEGA. The phylogenetic trees were inferred using a maximum likelihood approach with the Whelan and Goldman (WAG) substitution model and gamma-distributed rates. The reliability of the tree was assessed by bootstrapping [62] using 1000 bootstrap replicates. Furthermore, genes involved in mucin synthesis were identified in the sea trout skin transcriptome based on gene name searches against previous studies in mammals and fish [44, 63] and the KEGG mucin type O-glycan biosynthesis pathway.
SSR detection
Potential microsatellite markers were detected using MSDB software [64]. We identified all SSRs in our data set with a repeat length ranging from 2 to 6 and a minimum sequence length of 12 nt. Mononucleotide repeats were ignored because it was difficult to distinguish them from polyadenylation products.
Results
Sequencing and assembly
In this study, 1,440,373 raw reads with an average read length of 333 bp were generated using pyrosequencing. After the low-quality sequences and short reads were trimmed, 1,348,306 (93.61%) high-quality reads were used for de novo assembly in CLC Genomics Workbench (Table 1). A total of 75,970 contigs with a mean length of 668 bp were produced (Table 1). The transcriptome data analysis workflow is displayed in S1 Fig. The sea trout raw transcriptome sequences from this study were deposited in the NCBI Sequence Read Archive (SRA) under accession number SRP075946. Assembled contigs and annotation information are available from the authors upon request.
Transcriptome annotation
All the assembled 75,970 contigs were annotated by BLASTX searches against four protein databases (Table 1). In total, 36,902 (48.57%) contigs had hits in public databases, and 19,299 (52.30%) had hits in all four databases (Fig 1A). The E-value distribution of the top hits in the NCBI nr database indicated that 6.09% contigs showed excellent matches (E-value = 0) and that 47.27% had highly and moderately significant homology (≤1e-50, Fig 1B). Most of the matched contigs had hits in the salmonid family: rainbow trout (O. mykiss, 57.27%) and Atlantic salmon (S. salar, 15.47%), as expected. Moreover, 27.01% exhibited similarity to other organisms, mainly eukaryotes: fishes, mammals and birds (Fig 1C).
(A) Venn diagram illustrating the distribution of matches against four public databases (NCBI nr, Swiss-Prot, Teleost Ensembl and KOG databases). In total, 36,902 (48.57%) transcripts were annotated. (B) E-value and (C) species distribution of the best hits based on the NCBI nr database.
Based on the NCBI nr database best hits imported into Blast2Go, a total of 164,515 GO terms were assigned to 26,263 contigs, and on average, one contig was assigned to six GO terms. Of these, the biological process category was the most prevalent (89,576, 54.45%), followed by the cellular component category (40,896, 24.86%) and the molecular function category (34,043, 20.69%). Within the biological process category, contigs involved in cellular process (GO: 0009987) and metabolic process (GO: 0008152) were highly represented. In the cellular component category, cell (GO: 0005623) and organelle (GO: 0043226) were the most represented subcategories. Binding (GO: 0005488) and catalytic activity (GO: 0003824) were dominant groups within the molecular function category (Fig 2).
For functional prediction and classification, all assembled contigs were compared with the KOG database. In total, 19,515 (25.68%) contigs were classified into 24 functional categories. The seven largest categories were follows: general function prediction only (3,600, 16.28%), followed by signal transduction mechanisms (3,038, 13.74%); posttranslational modification, protein turnover, and chaperones (2,186, 9.89%); transcription (1,515, 6.85%); function unknown (1,396, 6.31%); cytoskeleton (1,380, 6.24%) and intracellular trafficking, secretion, and vesicular transport (1,353, 6.12%; Table 2).
Pathway-based analysis revealed that 10,182 (13.40%) of the contigs were assigned to 344 pathways in the KEGG database. The five highly represented pathways include cancer (724), PI3K-Akt signaling (646), focal adhesion (568), endocytosis (543), proteoglycans in cancer (485), MAPK signaling (458) and regulation of actin cytoskeleton (451; S1 Table). Within the top five KO categories including metabolism, genetic information processing, environmental information processing, cellular process and organismal system, the subcategory signal transduction (6,197) was the most highly represented, followed by immune system (2,347), nervous system (1,542), transport and catabolism (1,320), translation (1,222), cellular community (1,197), and folding, sorting and degradation (1,196; Table 3).
Given the existence of contigs that were not annotated to known proteins in the four public databases by BLAST searches, we attempted to identify them by searching against the Rfam database [53]. As shown in S2 Table, a total of 526 transcripts without significant blast hits were identified as putative ncRNAs. UnaL2 LINE 3' element (UnaL2, 68.08%), ribosomal RNA (rRNA, 8.17%) and transfer RNA (tRNA, 7.22%) were the most abundant ncRNA families (S2 Table).
We compared the trout skin transcriptome to the multi-tissue transcriptome using BLASTn. While 68,697 (90.43%) contigs of the skin transcriptome were homologous with one of the 202,994 multi-tissue transcripts, 9.57% did not show any similarity and thus represent new sequence information. Of these 7,273 novel genes, 1,429 (19.65%) were annotated in at least one of the searched database. Additionally, 58,700 (85%) of homologous contigs have high similarity (>95%). We also investigated nucleotide-level comparisons by mapping skin transcriptome reads to the multi-tissue transcriptome. A total of 52,522 (69.14%) sequences were mapped to the reference transcriptome; of these sequences, 2.51% mapped perfectly and unambiguously. Moreover, GO annotation results were compared between these trout transcriptomes (Fig 2). In general, the main GO category assignment was similar: the biological process category was the most highly represented, followed by cellular component and molecular function categories. However, there were some minor differences in the distribution of the genes in these three categories. The most noticeable differences were observed in the biological process category. In this category the highest numbers of transcripts were associated with cellular process (19.11%), metabolic process (17.76%) and single-organismal process (15.77%) in the skin transcriptome and with cellular process (15.74%), single-organismal process (14.03%) and metabolic process (12.83%) in the multi-tissue transcriptome. Fisher’s exact test was applied to discover significantly over-represented GO terms in skin transcriptome. GO enrichment analysis revealed that the highest enrichment was associated with the molecular function category (51.28%), followed by the biological process category (32.48%) and cellular component category (16.24%; S3 Table). The over-represented biological process categories were metabolic process (e.g. methylation, GO:0032259, p-value = 1.9×10−7), cellular component and biogenesis (e.g. ribosome biogenesis, GO:0042254; p-value = 2.4×10−5), cellular process (e.g. cellular metabolic compound salvage, GO:0043094, p-value = 4.6×10−9) and single organismal process (e.g. oxidation reduction process, GO:0055114, p-value = 3×10−11). In molecular function category, over-represented were terms associated with catalytic activity (e.g. oxidoreductase activity, GO:0016491.p-value = 1×10−27; hydrolase activity, GO:0016787, p-value = 1.6×10−6), molecular transducer activity (e.g. scavenger receptor activity, GO:0005044;P-Value = 2×10−7), structural molecule activity (e.g. structural constituent of ribosome GO:0003735, p-value = 1.2×10−17) and binding (e.g. heterocyclic compound binding GO:1901363, p-value = 2.9×10−8; S3 Table).
Mucins
The BLASTX results identified numerous contigs as showing significant homology to fish mucin genes. Searches in the created sea trout database identified 14 contigs that exhibited homology to mucins: 11 contigs showed homology to the SGFM family (Muc2, Muc5AC, Muc5B, I-Muc), and 3 showed homology to membrane-bound mucins (Muc12, Muc15, Muc17) (Table 4). Additionally, 14 contigs were obtained from the multi-tissue database and were annotated as Muc2, Muc5AC, Muc5B, I-Muc, Muc13 and Muc17 (S4 Table). Generally, the top hits of the BLASTX results were Atlantic salmon mucins, whereas of the top hits for ST_7108 (Muc15), ST_2730 were: northern pike (Esox lucius), brichardi cichlid (Neolamprologus brichardi), respectively. Furthermore, all contigs showed homology to Salmo salar genome (Table 4 and S4 Table).
The SMART analysis revealed PTS domains in all annotated sequences typical for all mucins. As depicted in Fig 3, the sequences annotated as I-Muc, Muc2, Muc5AC, and Muc5B share VWD and C8 domain, which are typical for the SGFM subfamily. S2 Fig shows the deduced amino acid sequence of I-Muc together with sequence and domain alignments with orthologs from other species. A von Willebrand factor type C (VWC) domain was found in the sequences recognized as Muc5B and I-Muc, and a C-terminal cysteine knot-like domain (CT) was found in the sequences recognized as Muc5B. The sequences recognized as Muc12, Muc13 and Muc17 shared an extracellular proteolytic cleavage site (SEA domain) characteristic of membrane-bound mucins. The Muc13 sequences contained an EGF domain (before the SEA domain), and Muc17 sequences contained a TM domain. The sequence recognized as Muc15 possessed a Muc15 domain (Fig 3).
Based on ST_51843, ST_271, ST_5641, ST_39, ST_35504, ST_27547, ST_7108, and ST_28361.
The phylogenetic analysis of the mucins of salmonidae and other teleost fish revealed three major clades (Muc5, Muc2 and I-Muc) (Fig 4 and S3 Fig). Of note, the I-Muc genes are classified as a SGFM group with Muc2 and Muc5, in contrast to previous studies [39]. Contigs annotated as Muc5AC and Muc5B are part of the Muc5 protein family, although it cannot be definitely concluded whether they are type 5AC or 5B.
The tree was constructed using multiple alignment of the VWD and C8 domains of the translated (ST_9981, ST_36280, ST_39) and selected teleost families: Salmonidae, Esocidae, and Cichlidae. The maximum likelihood phylogeny in MEGA 7 [61] was selected. The tree was bootstrapped 1,000 times. Accession numbers: Atlantic salmon I-Muc* (XP_014041914), northern pike I-Muc* (XP_012994242.1), northern pike I-Muc** (XP_012993966.1), Atlantic salmon I-Muc** (XP_013982567.1), Astatotilapia burtoni I-Muc* (XP_005941718.1), A. burtoni I-Muc** (XP_005946303.1); A. burtoni I-Muc*** (XP_005952623.1), Atlantic salmon I-Muc*** (XP_014038548.1), Atlantic salmon Muc5B (XP_014031349.1), Atlantic salmon Muc5AC (XP_014036802.1), northern pike Muc5AC (XP_010867519.2), A. burtoni Muc5AC (XP_005946314.1), A. burtoni Muc2 (XP_005952624.2), northern pike Muc2 (XP_012994223.1), Atlantic salmon Muc2 (XP_014040158.1).
Mucin biosynthesis
A total of 40 contigs from the sea trout skin transcriptome were assigned to the 7 enzymes in the mucin type O-glycan biosynthesis KEGG pathway (Fig 5). Of these 40 contigs, 29 contigs were mapped to GALNT enzymes (2.4.1.41), which includes 11 members of the N-acetylgalactosaminyltransferase family (GALNT 3, 4, 5, 6, 7, 8, 12, 14, 15 and 18). Moreover, 6 contigs were assigned to glycoprotein-N-acetylgalactosamine 3-beta-galactosyltransferases (2.4.1.122), and 2 contigs were mapped to glucosaminyl (N-acetyl) transferases (GCNT 3 and 4; 2.4.1.102/148). Furthermore, 3 contigs were annotated as sialyltransferases: SIAT4A (ST3GAL1), SIAT4B (ST3GAL2)–(2.4.99.3) and SIAT7 (ST6GALNAC2)–(2.4.99.4). In total, 140 genes that are implicated in synthesis, intracellular transport, and postsecretory modifications of mucins were identified in the sea trout skin transcriptome (S5 Table). In this study, 5 members of the protein disulfide isomerase (PDI) family and 45 members of the glycosyltransferase family were identified. Furthermore, transcripts of SAM-pointed domain-containing ETS transcription factor (SPDEF) and epidermal growth factor receptor (EGFR) were recognized. Two functional groups of genes encoding proteins associated with intracellular transport, RAB and SNARE (36 and 18 members respectively), were identified. Three functional groups of genes encoding proteins associated with postsecretory modifications of mucus gels were also identified. These proteins included 2 galectins, 4 aquaporins, and 13 ion channels, ion pumps and transporters. Six members of the C2 domain containing protein kinases family (PKC) and myristoylated alanine rich protein kinase C substrate (MARCKS) were also recognized.
Genes identified from the sea trout skin transcriptome are shown in yellow.
Immune pathways
Although brown trout multi-tissue transcriptomes have been previously characterized using RNA-seq [45], the immune response genes and pathways in the sea trout skin have not been characterized. GO classification identified 456 genes related to the immune response (GO: 0045087). Moreover, KEGG annotation identified 310 genes related to the immune system: these genes were distributed in 15 pathways. The dominant pathways included chemokine signaling (271), leukocyte transendothelial migration (264) and Fc gamma R-mediated phagocytosis (204) (Table 5). The highest number of retrieved genes in each pathway were found in the B cell receptor signaling pathway (74.55%), Fc gamma R-mediated phagocytosis (72.41%) and leukocyte transendothelial migration (72%) (Table 5). Combining annotations from GO, KEGG and BLAST searches revealed 954 genes potentially involved in immune reactions in the sea trout skin transcriptome (S6 Table).
SSR detection
In total, 1,119 potential SSRs with an average length of 64.28 bp were identified in 3,134 (4.13%) contigs, and 130 of these contigs contained more than one SSR (Table 6). Among the identified repetitive elements, the most highly represented were di-nucleotide repeats, and AC/GT (703; 62.82%) was the dominating motif. Among the tri-nucleotide and tetra-nucleotide repeats, ACT/AGT and ACAT/ATGT (26 and 30, respectively) accounted for the most frequent motifs (S7 Table).
Discussion
The innate immune system is the only known defense weapon of invertebrates and a fundamental defense mechanism of fish [65]. Skin and the mucosal layer form the first barrier against infection [29, 66]. The defensive role of mucus is well known and has been studied in several fish species [21, 26 and 28]. In the present study, we characterized the skin transcriptome of the sea trout using the Roche GS-FLX 454 pyrosequencing system, in order to identify putative immunogens, including mucins. We obtained over 479 Mb of raw sequence data and were able to assemble these reads into 75,970 contigs, 36,902 (48.57%) of which were annotated by BLAST searches against four public databases (Table 1), which is consistent with previous studies [45]. The high number of contigs (98%) were annotated by using the NCBI nr database, which suggest that this database is very reliable for annotation of non-model species. On the other hand this database exhibit high level of misannotation [67], so to ameliorate this problem other conservative and verified databases (Swiss-Prot, Ensembl and KOG) were used to confirm annotation. Moreover, these databases provide additional sequence information for 741 contigs e.g. RAB38b (putatively involved in mucin synthesis), vesicle transport through interaction with t-SNAREs 1A (VTI1A, involved in vesicular transport) [68], peroxiredoxin (antioxidant). These results provide a valuable resource for further brown trout genome studies. Furthermore, most of the species on the top hits list of annotated genes were fishes, suggesting the high quality of the assembled transcripts and high levels of sequence conservation and homology with other fish species, mainly salmonids (Fig 1C). A large number of contigs were assigned to GO terms in three main categories, including biological process, molecular function and cellular component (Fig 2). GO analysis of the trout skin was compared with a transcriptome from a previous study [45]. The proportion of annotated contigs and GO terms were comparable, despite the different number of sequences used to retrieve GO terms (202,994 versus 75,970) and the different sequencing system (Illumina versus 454). This result suggests that the dataset from this study includes a wide variety and full view of the diverse structural, regulatory and metabolic proteins expressed in the sea trout skin. According to the GO enrichment results, terms associated with biological processes (e.g. metabolic process) were enriched, indicating the basic functions of trout skin (S3 Table). Moreover, ribosomal proteins (40s and 60s) were over-expressed, similarly with the skin of channel catfish [69]. In the level two GO category of molecular function, enriched were terms involved in catalytic activity and molecular transducer activity, including oxidoreductase activity (e.g. superoxide dismutase), hydrolase activity (e.g. leukotriene A-4 hydrolase) and scavenger receptor activity (e.g. scavenger receptor class A). All these genes are involved in stress reaction and immune response against pathogens in Atlantic salmon and other teleost [29, 70, 71]. Furthermore, 7,273 contigs did not match any sequences identified in the multi-tissue transcriptome. These contigs may represent novel or specific genes for sea trout or for the skin transcriptome.
Genes identified in the KEGG database provide a resource for functional information and a better understanding of the physiology of organisms [72]. The annotated contigs represent a wide array of KEGG-linked pathways (Table 3 and S1 Table). The predominantly represented KO category was organismal system (7,216), with highly represented subcategories including immune system and nervous system pathways and environmental information processing (6,831), with signal transduction pathway being predominant (Table 3). The most highly represented pathways were associated with signal transduction and involved in the immune system, such as the PI3K-Akt signaling pathway (secretion of proinflammatory cytokines) [73], MAPK signaling pathway (production of immunomodulatory cytokines) [74] and Rap1 signaling pathway (essential for changes in B cell morphology) [75]. Other highly represented pathways included those involved in various cellular functions, including cell proliferation, differentiation, motility, and regulation of gene expression, suggesting high cellular process activity in the trout skin, which is comparable with previous studies of fish skin and which suggest the importance of this pathway in fish skin [28]. Moreover, many genes associated with the immune system were expressed in the trout skin. The identified immunogens represent between 51% to 75% of the known genes in each pathway, except natural killer cell-mediated cytotoxicity, hematopoietic cell lineage and intestinal immune network for IgA production (Table 5). The highest numbers of sequences were assigned to chemokine signaling and leukocyte transendothelial migration pathways, which are both differentially activated in other fish (Pseudosciaena crocea) infected with Aeromonas hydrophila [76]. These results seem to be consistent with previous skin studies of mud loach [28], with two exceptions. In sea trout, more genes were restored in the complement and coagulation cascades (55% versus 31% in mud loach), and in contrast, antigen processing and presentation was less represented (54% versus 75%). Both GO and KEGG analyses revealed significantly enriched GO terms and pathways associated with the immune system (S6 Table). In addition to KEGG pathways, our high-throughput sequencing effort revealed the presence of numerous (954) molecules involved in the immune response: cytokines, antimicrobial peptides (AMPs), lectins, galectins, heat shock proteins, granzymes, immunoglobulins, complement components and cathepsins.
Antimicrobial peptides are oligopeptides with a varying number (from five to over a hundred) of amino acids [77], playing crucial role in innate immune defences against pathogen. They display a broad spectrum of activity against numerous pathogenic organisms including Gram-positive and Gram-negative bacteria, yeast, fungi, and parasites with little or no toxicity to host cells [25]. AMPs activity have been tested against few bacterial pathogens [78, 79]. In present study, we identified several contigs presenting homology to cathelicidins, β-defensin, hepcidin and H2A.
The complement component system is a cascade of a large number of distinct plasma proteins that react with one another to opsonize pathogens and induce a series of inflammatory responses [80] essential in innate immune system. In vertebrates, the three pathways of activation (classical, alternative and lectin pathway) lead to generate C3-convertase, which activated the central component C3 [81]. We have identified several sequences presenting homology to C1, C3, C4, C5, C6, C7, and C8. Only the C2 and C9 (component seems not to be represented. Moreover, we identified receptors and other proteins involved in the complement component cascade, including the complement receptor type 1and type 2 (CR1, CR2), the complement factor H, D.
Immunoglobulins are Y-shaped glycoproteins produced mainly by plasma cells and play key roles in the mucosal adaptive system, specifically recognizing and binding to particular antigens, such as bacteria [25]. There are three major types of immunoglobulins in teleost fish: IgM, IgD and IgT/IgZ, the latter of which is unique for teleost fish [82, 83]. In present study, we identified several contigs presenting similarity to IGHD and IGHM.
Among other immune-related proteins, several proteins were identified as being related to immune mechanisms in Atlantic salmon and rainbow trout infected by A. salmonicida. These proteins include ferritin, voltage-dependent anion channel 2 (VDAC2), GTPase IMAP family member 7 (GiMAP7), protease subunit alpha type (PSMA1), annexin A11 (ANXA11), lysozyme, and alkaline phosphatase (AKP) [28, 29] (S6 Table), which were not previously reported for the sea trout transcriptome.
Mucins are high molecular weight glycoproteins and are the main constituents of fish skin mucus, which can produce a protective gel to bind a range of bacteria and which constitutes an important part of the mucosal defense against infection [37]. The mucins and other genes involved in mucus synthesis remain poorly characterized. In this study, 7 types of mucins (14 contigs) were identified (Table 4). The molecular identity of these mucins was established on the basis of BLAST annotations, domain searches and phylogenetic analysis. All annotated mucins contain a PTS domain (Fig 3). The number of repeats rich in proline, threonine and serine and the amino acid sequence depend on the mucin type [30]. In contigs annotated as intestinal mucin, a VWD domain was found, indicating a relationship to the SGFM mucin family (Fig 3 and S3 Fig), although this domain can also be found in other non-mucin proteins [84]. However, the sea trout sequences found in the present study also share cysteine-rich domains with other bony fishes (Fig 3 and S3 Fig). The phylogenetic analysis suggested that the VWD and C8 domains have been conserved throughout the evolution of vertebrates, consistent with previous studies [39, 85]. It is noteworthy that I-Muc was categorized as a SGFM based on phylogenetic analysis and domain characteristics which is in contrast to the results of previous studies [39] (Figs 3 and 4). Moreover, phylogenetic analysis of intestinal mucins in a representative teleost group suggested that I-Muc family might consist of three subfamilies (S3 Fig). Notably, the identified mucins are the first reported trout sequences in the GenBank database. Furthermore, we reported the first protein sequence identified as Muc15 for the Salmonidae family. The presence of characteristic domains such as VWD, C8, SEA and TM strongly supports conclusion that the sea trout mucins are real mucin genes, instead of genes that only reassemble mucins (Muc-like genes). Mucins are characterized by their extensive O-glycosylation [86]. The biosynthesis of O-glycans starts from the transfer of N-acetylgalactosamine (GalNAc) to serine or threonine [87]. The first GalNAc may be extended with sugars such as galactose, N-acetylglucosamine, fucose, or sialic acid, but not mannose, glucose, or xylose. Depending on the sugars added, there are four common O-glycan core structures, cores 1 through 4, and an additional four, cores 5 through 8 [87]. In this study, 7 enzymes assigned to the mucin type O-glycan biosynthesis KEGG pathway were identified (Fig 5), which is similar to the findings of a previous study [28], suggesting that only cores 1 through 4 are activated in fish. Moreover, 11 members of the polypeptide N-acetyl-galactosaminyltransferase family (EC2.4.1.41) were expressed in the skin of sea trout. These enzymes belong to a large enzyme family (with at least 15 members) with tissue-specific expression and different substrate specifies [88]; these enzymes are active at the first step of glycosylation. Furthermore, 129 other genes associated with mucin synthesis were identified (S5 Table). The results of this study provide a valuable resource for characterizing mucus secretion in sea trout.
Transcriptome- and EST-based markers are important resource for determining functional genetic variation [89]. Moreover, because transcriptome-based SSRs mainly occur in the protein-coding regions of annotated genes, they are better at identifying phenotypic associations [90]. Among the various molecular markers, SSRs are highly polymorphic, are relatively easy to develop and serve as a rich source of diversity [91]. In this study, we identified 1,119 potential SSRs with predominant di-nucleotide repeats (Table 6).
Conclusions
This study investigated the skin transcriptome of the sea trout using the Roche GS-FLX 454 pyrosequencing system. We generated 75,970 contigs, 48.57% of which were annotated by BLAST searches. A large number of contigs were classified according to GO and KO terms, which represent multiple signaling pathways and processes. Many of these terms are related to the immune response including mucins and mucus secretion. To our knowledge, this study is the first to describe the skin transcriptome of sea trout.
Supporting information
S1 Fig. Sea trout skin transcriptome analysis pipeline.
https://doi.org/10.1371/journal.pone.0172282.s001
(TIF)
S2 Fig. Multiple alignment of the amino acid sequences of sea trout I-Muc with other teleost.
https://doi.org/10.1371/journal.pone.0172282.s002
(TIF)
S3 Fig. Unrooted phylogenetic tree showing the evolutionary relationship of the I-Muc in teleost.
The tree was constructed using multiple alignment of the VWD and C8 domains of the translated (ST_10480 and ST_39) and selected teleost families: Salmonidae, Esocidae, Poeciliidae, Cyprinodontidae, and Cichlidae. The maximum likelihood phylogeny in MEGA 7 [61] was selected. The tree was bootstrapped 1000 times. Accession numbers: Atlantic salmon* (XP_014041914.1), northern pike* (XP_012994242.1), A. burtoni* (XP_005941718.1), cave molly* (XP_014867637.1), northern pike** (XP_012993966.1), Atlantic salmon** (XP_013982567.1), A.burtoni** (XP_005946303.1), sheepshead minnow* (XP_015252020.1), cave molly** (XP_014863832.1), sheepshead minnow** (XP_015259098.1), cave molly*** (XP_014832181.1), A. burtoni*** (XP_005952623.1), Atlantic salmon*** (XP_014038548.1).
https://doi.org/10.1371/journal.pone.0172282.s003
(TIF)
S1 Table. Summary of sequences involved in the KEGG pathways and KO categories.
https://doi.org/10.1371/journal.pone.0172282.s004
(PDF)
S2 Table. Classification of identified non-coding RNAs.
https://doi.org/10.1371/journal.pone.0172282.s005
(PDF)
S3 Table. Significantly overrepresented GO terms in the sea trout skin transcriptome (FDR<0.05).
https://doi.org/10.1371/journal.pone.0172282.s006
(PDF)
S4 Table. Classification of the mucins identified in multi-tissue transcriptome, according to the BLASTX and BLASTN searches.
https://doi.org/10.1371/journal.pone.0172282.s007
(PDF)
S5 Table. Genes involved in mucin biosynthesis and mucus production in sea trout skin transcriptome.
https://doi.org/10.1371/journal.pone.0172282.s008
(PDF)
S6 Table. List of immunity-related genes found in sea trout skin transcriptome.
https://doi.org/10.1371/journal.pone.0172282.s009
(PDF)
S7 Table. Distribution of EST-SSRs based on motif types.
https://doi.org/10.1371/journal.pone.0172282.s010
(PDF)
Author Contributions
- Conceptualization: RW MM.
- Data curation: MM AB RW.
- Formal analysis: MM AB RW.
- Funding acquisition: RW.
- Investigation: MM AB RW.
- Methodology: MM AB RW.
- Project administration: RW.
- Resources: MM RW.
- Supervision: RW.
- Validation: RW MM.
- Writing – original draft: MM.
- Writing – review & editing: MM AB RW.
References
- 1. Klemetsen A, Amundsen PA, Dempson J B, Jonsson B, Jonsson N, O'Connell M F, et al. Atlantic salmon Salmo salar L., brown trout Salmo trutta L. and Arctic charr Salvelinus alpinus (L.): a review of aspects of their life histories. Ecology of Freshwater Fish. 2003; 12(1): 1–59.
- 2. Poćwierz-Kotus A, Bernaś R, Dębowski P, Kent MP, Lien S, Kesler M, et al. Genetic differentiation of southeast Baltic populations of sea trout inferred from single nucleotide polymorphisms. Anim Genet. 2014; 45(1): 96–104. pmid:24236931
- 3.
Elliott JM. Quantitative ecology and the brown trout. USA: Oxford University Press; 1994.
- 4.
ICES. Report of the Baltic Salmon and Trout Assessment Working Group (WGBAST). 23–31 March 2015. Rostock, Germany. ICES CM 2015\ACOM:08. 362 pp.
- 5. Wąs A, Wenne R. Genetic differentiation between hatchery and wild sea trout (Salmo trutta) in the Southern Baltic at microsatellite loci. Aquaculture. 2002; 204(3–4): 493–506.
- 6. Wąs A, Wenne R. Microsatellite DNA polymorphism in intensely enhanced populations of sea trout (Salmo trutta) in the Southern Baltic. Marine Biotechnology. 2003; 5(3): 234–243. pmid:14502395
- 7. Jonsson B, Jonsson N. A review of the likely effects of climate change on anadromous Atlantic salmon Salmo salar and brown trout Salmo trutta, with particular reference to water temperature and flow. J Fish Biol. 2009; 75(10): 2381–447. pmid:20738500
- 8. Janda JM, Abbott SL. The genus Aeromonas: taxonomy, pathogenicity, and infection. Clin Microbiol Rev. 2010; 23(1): 35–73. pmid:20065325
- 9. Fast MD, Sims DE, Burka JF, Mustafa A, Ross NW. Skin morphology and humoral non-specific defence parameters of mucus and plasma in rainbow trout, coho and Atlantic salmon. Comp Biochem Physiol A Mol Integr Physiol. 2002; 132(3): 645–57. pmid:12044774
- 10. Freedman SJ. The role of a-2-macroglobulin in the furunculosis: a comparison of rainbow trout and Brook trout. Comp Biochem Physiol. 1991; 98(4): 533–549.
- 11. LaPatra SE, Fryer JL, Wingield WH, Hedrick RP. Infectious Hematopoietic Necrosis Virus (IHNV) in Coho Salmon. J Aquat Anim Health. 1989; 1(4): 277–280.
- 12. Ellis AE, Stapleton KJ. Differential susceptibility of salmonid fishes to furunculosis correlates with differential serum enhancement of Aeromonas salmonicida extracellular protease activity. Microbial Pathogenesis. 1988; 4(4): 299–304. pmid:3059141
- 13.
Anderson DP, Siwicki AK, Dixon OW, Lizzio EF. Immunostimulation by levamisole in rainbow trout (Salmo gairdneri). In: Ahne W, Kurstak E. editors. Viruses of lower vertebrates. Berlin: Springer; 1989. pp. 469–478.
- 14. Padra JT, Sundh H, Jin C, Karlsson NG, Sundell K, Lindén SK. Aeromonas salmonicida binds differentially to mucins isolated from skin and intestinal regions of Atlantic salmon in an N-acetylneuraminic acid-dependent manner. Infect Immun. 2014; 82(12): 5235–45. pmid:25287918
- 15. Bergsson G, Agerberth B, Jornvall H, Gudmundsson GH. Isolation and identification of antimicrobial components from the epidermal mucus of Atlantic cod (Gadus morhua). FEBS J. 2005; 272(19): 4960–4969. pmid:16176269
- 16. Saurabh S, Sahoo PK. Lysozyme: an important defence molecule of fish innate immune system. Aquaculture research. 2008; 39(3): 223–239.
- 17. Alexander JB, Ingram GA. Noncellular nonspecific defence mechanisms of fish. Annu Rev Fish Dis. 1992; 2: 249–279.
- 18. Gonzalez SF, Buchmann K, Nielsen ME. Complement expression in common carp (Cyprinus carpio L.) during infection with Ichthyophthirius multifiliis. Dev Comp Immunol. 2007; 31(6): 576–86. pmid:17107712
- 19. Salinas I, Zhang Y-A, Sunyer JO. Mucosal immunoglobulins and B cells of Teleost fish. Dev Comp Immunol. 2011; 35(12): 1346–1365. pmid:22133710
- 20. Tsutsui S, Komatsu Y, Sugiura T, Araki K, Nakamura O. A unique epidermal mucus lectin identified from catfish (Silurus asotus): first evidence of intelectin in fish skin slime. Journal of Biochemistry. 2011; 150(5): 501–514. pmid:21757471
- 21. Subramanian S, Ross NW, MacKinnon SL. Myxinidin, a novel antimicrobial peptide from the epidermal mucus of hagfish, Myxine glutinosa L. Mar Biotechnol. 2009; 11(6): 748–757. pmid:19330556
- 22. Aranishi F, Nakane M. Epidermal proteinases of the Japanese eel. Fish Physiol Biochem. 1997; 16(6): 471–478.
- 23.
Elliott DG. Integumentary system. In: Ostrander GK, editor. The Laboratory Fish. London: Academic Press; 2000. pp. 95–108.
- 24. Marshall WS, Bellamy D. The 50 year evolution of in vitro systems to reveal salt transport functions of teleost fish gills. Comp Biochem Physiol. 2010; 155(3): 275–280.
- 25. Esteban MA. An overview of the immunological defenses in fish skin. ISRN Immunology Volume 2012. Article ID 853470, 29 pages, 2012.
- 26. Hjelmeland K, Christie M, Raa J. Skin mucus protease from rainbow trout, Salmo gairdneri Richardson, and its biological significance. J Fish Biol. 1983; 23(1): 13–22.
- 27. Micallef G, Bickerdike R, Reiff C, Fernandes JM, Bowman AS, Martin SA. Exploring the transcriptome of Atlantic salmon (Salmo salar) skin, a major defense organ. Mar Biotechnol. 2012; 14(5): 559–69. pmid:22527268
- 28. Long Y, Li Q, Zhou B, Song G, Li T, Cui Z. De Novo assembly of mud loach (Misgurnus anguillicaudatus) skin transcriptome to identify putative genes involved in immunity and epidermal mucus secretion. PLoS One. 2013; 8(2): e56998. pmid:23437293
- 29. Du Y, Yi M, Xiao P, Meng L, Li X, Sun G, et al. The impact of Aeromonas salmonicida infection on innate immune parameters of Atlantic salmon (Salmo salar L). Fish Shellfish Immunol. 2015; 44(1): 307–15. pmid:25725402
- 30. Desseyn JL, Aubert JP, Porchet N, Laine A. Evolution of the large secreted gel-forming mucins. Mol Biol Evol. 2000; 17(8): 1175–1184. pmid:10908637
- 31. Neuhaus H, Van der Marel M, Caspari N, Meyer W, Enss M, Steinhagen D. Biochemical and histochemical effects of perorally applied endotoxin on intestinal mucin glycoproteins of the common carp Cyprinus carpio. Dis Aquat Org. 2007; 77(1): 17–27. pmid:17933394
- 32. Specian RD, Oliver MG. Functional biology of intestinal goblet cells. Am J Physiol. 1991; 260 (2 Pt 1): C183–93.
- 33. Hollingsworth MA, Swanson BJ. Mucin in cancer: protection and control of the cell surface. Nat Rev Cancer. 2004; 4: 45–60. pmid:14681689
- 34. Moniaux N, Escande F, Porchet N, Aubert JP, Batra SK. Structural organization and classification of the human mucin genes. Front Biosci. 2001; 6: D1192–206. pmid:11578969
- 35. Perez-Vilar J, Eckhardt AE, DeLuca A, Hill RL. Porcine submaxillary mucin forms disulfide-linked multimers through its amino-terminal D-domains. J Biol Chem. 1998; 273(23): 14442–14449. pmid:9603957
- 36. Hoorens PR, Rinaldi M, Li RW, Goddeeris B, Claerebout E, Vercruysse J, et al. Genome wide analysis of the bovine mucin genes and their gastrointestinal transcription profile. BMC Genomics. 2011; 12: 140. pmid:21385362
- 37. Linden SK, Sutton P, Karlsson NG, Korolik V, McGuckin MA. Mucins in the mucosal barrier to infection. Mucosal Immunol. 2008; 1(3): 183–97. pmid:19079178
- 38. Yan HY. A histochemical study on the snout tentacles and snout skin of bristlenose catfish Ancistrus triradiatus. J Fish Biol. 2009; 75(4): 845–861. pmid:20738583
- 39. Pérez-Sánchez J, Estensoro I, Redondo MJ, Calduch-Giner JA, Kaushik S, Sitjà-Bobadilla A. Mucins as diagnostic and prognostic biomarkers in a fish-parasite model: transcriptional and functional analysis. PLoS One. 2013; 8: e65457. pmid:23776483
- 40. Marel Mv, Adamek M, Gonzalez SF, Frost P, Rombout JH, Wiegertjes GF, et al. Molecular cloning and expression of two β-defensin and two mucin genes in common carp (Cyprinus carpio L.) and their up-regulation after β-glucan feeding. Fish Shellfish Immunol. 2012; 32(3): 494–501. pmid:22227003
- 41. Jevtov I, Samuelsson T, Yao G, Amsterdam A, Ribbeck K. Zebrafish as a model to study live mucus physiology. Sci Rep. 2014; 17(4): 6653.
- 42. Tacchi L, Lowrey L, Musharrafieh R, Crossey K, Larragoite ET, Salinas I. Effects of transportation stress and addition of salt to transport water on the skin mucosal homeostasis of rainbow trout (Oncorhynchus mykiss). Aquaculture. 2015; 435: 120–127. pmid:25705060
- 43. Peatman E, Lange M, Zhao H, Beck BH. Physiology and immunology of mucosal barriers in catfish (Ictalurus spp.). Tissue Barriers. 2015; 3(4): e1068907. pmid:26716071
- 44. Ao J, Mu Y, Xiang L-X, Fan D, Feng M, Zhang S, et al. Genome Sequencing of the Perciform Fish Larimichthys crocea Provides Insights into Molecular and Genetic Mechanisms of Stress Adaptation. PLoS Genet. 2015; 11(4): e1005118. pmid:25835551
- 45. Uren Webster TM, Bury N, van Aerle R, Santos EM. Global transcriptome profiling reveals molecular mechanisms of metal tolerance in a chronically exposed wild population of brown trout. Environ Sci Technol. 2013; 47(15): 8869–77. pmid:23834071
- 46. Uren Webster TM, Santos EM. Global transcriptomic profiling demonstrates induction of oxidative stress and of compensatory cellular stress responses in brown trout exposed to glyphosate and Roundup. BMC Genomics. 2015; 16:32. pmid:25636363
- 47. Zhulidov PA, Bogdanova EA, Shcheglov AS, Vagner LL, Khaspekov GL, Kozhemyako VB, et al. Simple cDNA normalization using kamchatka crab 694 duplex-specific nuclease. Nucleic Acids Res. 2004; 32(3):e37. pmid:14973331
- 48. Margulies M, Egholm M, Altman WE, Attiya S, Bader JS, Bemben LA, et al. Genome sequencing in microfabricated high-density picolitre reactors. Nature. 2005, 437: 376–380. pmid:16056220
- 49. Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos P, Bealer K, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009; 10: 421. pmid:20003500
- 50. Conesa A, Götz S, García-Gómez JM, Terol J, Talón M, Robles M. Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics. 2005; 21(18): 3674–6. pmid:16081474
- 51. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. Nat Genet. 2000; 25(1): 25–9. pmid:10802651
- 52. Moriya Y, Itoh M, Okuda S, Yoshizawa AC, Kanehisa M. KAAS: an automatic genome annotation and pathway reconstruction server. Nucleic Acids Res. 2007; 35: 182–5.
- 53. Nawrocki EP, Burge SW, Bateman A, Daub J, Eberhardt RY, Eddy SR, et al. Rfam 12.0: updates to the RNA families database. Nucleic Acids Res. 2014; 43: D130–7. pmid:25392425
- 54. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B Methodol. 1995; 57: 289–300.
- 55. Lien S, Koop BF, Sandve SR, Miller JR, Kent MP, Nome T, et al. The Atlantic salmon genome provides insights into rediploidization. Nature. 2016; 533(7602): 200–5. pmid:27088604
- 56. Burge C, Karlin S. Prediction of complete gene structures in human genomic DNA. J Mol Biol. 1997; 268(1): 78–94. pmid:9149143
- 57. Letunic I, Doerks T, Bork P. SMART: recent updates, new developments and status in 2015. Nucleic Acids Res. 2014; 43: D257–60. pmid:25300481
- 58. Moller S, Croning MDR, Apweiler R. Evaluation of methods for the prediction of membrane spanning regions. Bioinformatics. 2001; 17(7): 646–653. pmid:11448883
- 59. Benson DA, Cavanaugh M, Clark K, Karsch-Mizrachi I, Lipman DJ, Ostell J, et al. GenBank. Nucleic Acids Res. 2013; 41: 36–42.
- 60. Małachowicz M, Kijewska A, Wenne R. Transcriptome analysis of gill tissue of Atlantic cod Gadus morhua L. from the Baltic Sea. Mar Genomics. 2015; 27: 37–40.
- 61. Kumar S, Stecher G, Tamura K. MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Mol Biol Evol. 2016. In Press.
- 62. Felsenstein J. Confidence limits on phylogenies: an approach using the bootstrap. Evolution. 1985; 39(4): 783–791.
- 63. Pluta K, McGettigan PA, Reid CJ, Browne JA, Irwin JA, Tharmalingam T, et al. Molecular aspects of mucin biosynthesis and mucus formation in the bovine cervix during the periestrous period. Physiol Genomics. 2012; 44(24): 1165–78. pmid:23092952
- 64. Du L, Li Y, Zhang X, Yue B. MSDB: a user-friendly program for reporting distribution and building databases of microsatellites from genome sequences. J Hered. 2013; 104(1): 154–157. pmid:23144492
- 65. Magnadóttir B. Innate immunity of fish (overview). Fish Shellfish Immunol. 2006; 20(2): 137–51. pmid:15950491
- 66. Ellis AE. Innate host defense mechanisms of fish against viruses and bacteria. Dev Comp Immunol. 2001; 25(8–9): 827–39. pmid:11602198
- 67. Schnoes AM, Brown SD, Dodevski I, Babbitt PC. Annotation error in public databases: misannotation of molecular function in enzyme superfamilies. PLoS Comput Biol. 2009; 5(12): e1000605. pmid:20011109
- 68. Xu Y, Wong SH, Tang BL, Subramaniam VN, Zhang T, Hong W. A 29-kilodalton Golgi soluble N-ethylmaleimide-sensitive factor attachment protein receptor (Vti1-rp2) implicated in protein trafficking in the secretory pathway. J. Biol. Chem. 1998; 273(34): 21783–9. pmid:9705316
- 69. Karsi A, Cao D, Li P, Patterson A, Kocabas A, Feng J, et al. Transcriptome analysis of channel catfish (Ictalurus punctatus): initial analysis of gene expression and microsatellite-containing cDNAs in the skin. Gene. 2002; 285: 157–168. pmid:12039042
- 70. Meng Z, Zhang XY, Guo J, Xiang LX, Shao JZ. Scavenger receptor in fish is a lipopolysaccharide recognition molecule involved in negative regulation of NF-κB activation by competing with TNF receptor-associated factor 2 recruitment into the TNF-α signaling pathway. J Immunol. 2012; 189(8): 4024–39. pmid:22988031
- 71. Krasnov A, Skugor S, Todorcevic M, Glover KA, Nilsen F. Gene expression in Atlantic salmon skin in response to infection with the parasitic copepod Lepeophtheirus salmonis, cortisol implant, and their combination. BMC Genomics. 2012; 13:130. pmid:22480234
- 72. Kanehisa M, Goto S, Kawashima S, Okuno Y, Hattori M. The KEGG resource for deciphering the genome. Nucleic Acids Res. 2004; 32: D277–80. pmid:14681412
- 73. Xie S, Chen M, Yan B, He X, Chen X, Li D. Identification of a role for the PI3K/Akt/mTOR signaling pathway in innate immune cells. PLOS One. 2014; 9(4): e94496. pmid:24718556
- 74. Soares-Silva M, Diniz FF, Gomes GN, Bahia D. The mitogen-activated protein kinase (MAPK) pathway: role in immune evasion by Trypanosomatids. Front Microbiol. 2016; 7: 183. pmid:26941717
- 75. Lin KBL, Freeman SA, Zabetian S, Brugger H, Weber M, Lei V, et al. The Rap GTPases regulate B cell morphology, immune-synapse formation, and signaling by particulate B cell receptor ligands. Immunity. 2008; 28(1): 75–87. pmid:18191594
- 76. Mu Y, Ding F, Cui P, Ao J, Hu S, Chen S. Transcriptome and expression profiling analysis revealed changes of multiple signaling pathways involved in immunity in the large yellow croaker during Aeromonas hydrophila infection. BMC Genomics. 2010; 11:506. pmid:20858287
- 77. Bahar AA, Ren D. Antimicrobial Peptides. Pharmaceuticals (Basel). 2013; 6(12): 1543–1575.
- 78. Pan CY, Chen JY, Cheng YSE, Chen CY, Ni IH, Sheen JF, et al. Gene expression and localization of the epinecidin-1 antimicrobial peptide in the grouper (Epinephelus coiodes), and its role in protecting fish against pathogenic infection. DNA Cell Biol. 2007; 26: 403–413. pmid:17570764
- 79. Noga EJ, Silphaduang U, Park NG, Seo JK, Stephenson J, Kozlowicz S. Piscidin 4, a novel member of the piscidin family of antimicrobial peptides. Comp Biochem Physiol, Part B. 2009; 152: 299–305.
- 80.
Janeway CA Jr, Travers P, Walport M, Shlomchik MJ. Immunobiology: The Immune System in Health and Disease. 5th edition. The complement system and innate immunity. New York: Garland Science; 2001.
- 81. Gasque P. Complement: a unique innate immune sensor for danger signals. Mol Immunol. 2004; 41(11): 1089–98. pmid:15476920
- 82. Hordvik I. Immunoglobulin isotypes in Atlantic salmon, Salmo salar. Biomolecules. 2015; 5(1): 166–77. pmid:25734583
- 83. Hikima JI, Jung TS, Aoki T. Immunoglobulin genes and their transcriptional control in teleosts. Dev. Comp. Immunol. 2011; 35: 924–936. pmid:21078341
- 84. Dekker J, Rossen JWA, Buller HA, Einerhand AWC. The MUC family: an obituary. Trends Biochem Sci. 2002; 27(3): 126–31. pmid:11893509
- 85. Lang T, Alexandersson M, Hansson GC, Samuelsson T. Bioinformatic identification of polymerizing and transmembrane mucins in the puffer fish Fugu rubripes. Glycobiology. 2004; 14(6): 521–527. pmid:15044386
- 86. Straus GJ, Dekker J. Mucin-type glycoproteins. Critical Reviews in Biochemistry and Molecular Biology. 1992; 27: 57–92. pmid:1727693
- 87. Hashimoto M, Tan S, Mori N, Cheng H, Cheng PW. Mucin biosynthesis: molecular cloning and expression of mouse mucus-type core 2 beta1,6 N-acetylglucosaminyl-transferase. Glycobiology. 2007; 17(9): 994–1006. pmid:17591617
- 88. Hagen FK, Wuyckhuyse Van B, Tabak LA. Purification, cloning, and expression of a bovine UDP-GalNAc: popypeptide N-acetyl-galactosaminyltransferase. J Biol Chem. 1993; 268: 18960–18965. pmid:8360184
- 89. Varshney RK, Graner A, Sorrells ME. Genic microsatellite markers in plants: features and applications. Trends Biotechnol. 2005; 23(1): 48–55. pmid:15629858
- 90. Zalapa JE, Cuevas H, Zhu H, Steffan S, Senalik D, Zeldin E, et al. Using next-generation sequencing approaches to isolate simple sequence repeat (SSR) loci in the plant sciences. Am J Bot. 2012; 99(2): 193–208. pmid:22186186
- 91. Li D, Deng Z, Qin B, Liu X, Men Z. De novo assembly and characterization of bark transcriptome using Illumina sequencing and development of EST-SSR markers in rubber tree (Hevea brasiliensis Muell. Arg.) BMC Genomics. 2012; 13: 192. pmid:22607098