Eudiplozoon nipponicum (Monogenea, Diplozoidae) and its adaptation to haematophagy as revealed by transcriptome and secretome profiling

Background Ectoparasites from the family Diplozoidae (Platyhelminthes, Monogenea) belong to obligate haematophagous helminths of cyprinid fish. Current knowledge of these worms is for the most part limited to their morphological, phylogenetic, and population features. Information concerning the biochemical and molecular nature of physiological processes involved in host–parasite interaction, such as evasion of the immune system and its regulation, digestion of macromolecules, suppression of blood coagulation and inflammation, and effect on host tissue and physiology, is lacking. In this study, we report for the first time a comprehensive transcriptomic/secretome description of expressed genes and proteins secreted by the adult stage of Eudiplozoon nipponicum (Goto, 1891) Khotenovsky, 1985, an obligate sanguivorous monogenean which parasitises the gills of the common carp (Cyprinus carpio). Results RNA-seq raw reads (324,941 Roche 454 and 149,697,864 Illumina) were generated, de novo assembled, and filtered into 37,062 protein-coding transcripts. For 19,644 (53.0%) of them, we determined their sequential homologues. In silico functional analysis of E. nipponicum RNA-seq data revealed numerous transcripts, pathways, and GO terms responsible for immunomodulation (inhibitors of proteolytic enzymes, CD59-like proteins, fatty acid binding proteins), feeding (proteolytic enzymes cathepsins B, D, L1, and L3), and development (fructose 1,6-bisphosphatase, ferritin, and annexin). LC-MS/MS spectrometry analysis identified 721 proteins secreted by E. nipponicum with predominantly immunomodulatory and anti-inflammatory functions (peptidyl-prolyl cis-trans isomerase, homolog to SmKK7, tetraspanin) and ability to digest host macromolecules (cathepsins B, D, L1). Conclusions In this study, we integrated two high-throughput sequencing techniques, mass spectrometry analysis, and comprehensive bioinformatics approach in order to arrive at the first comprehensive description of monogenean transcriptome and secretome. Exploration of E. nipponicum transcriptome-related nucleotide sequences and translated and secreted proteins offer a better understanding of molecular biology and biochemistry of these, often neglected, organisms. It enabled us to report the essential physiological pathways and protein molecules involved in their interactions with the fish hosts. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-021-07589-z.


Background
Monogeneans, typically ectoparasites of freshwater and marine fish, are the causative agents of major global fish diseases. Clinical symptoms of monogenean infections, such as tissue injury, anaemia, and respiratory and osmoregulatory dysfunctions, are often accompanied by secondary microbial infections [1,2] that can lead to increased fish mortality. Monogenean infestations in fish aquacultures result in significant economic losses; in Norway, for example, annual economic losses due to the presence of the monogenean Gyrodactylus salaris in commercial breeding stocks of Atlantic salmon (Salmo salar) are estimated at over 500 million USD [3].
Despite the economic importance of this group of ectoparasites, there is a shortage of data on their functional molecular biology and interactions with their fish hosts. To reveal phylogenetic relationships within the species-rich monogenean group, several studies focused on sequencing the mitochondrial genome. This research investigated, for instance, (a) polytypic parasites G. salaris [4], which infects salmon, and Gyrodactylus thymalli [5], which typically infect the grayling; (b) Neobenedenia melleni, a generalist parasite of marine fish [6]; (c) Benedenia hoshinai [7], a parasite of the striped knifejaw Oplegnathus fasciatus, and (d) Pseudochauhanea macrorchis [8], which typically infects the pickhandle barracuda Sphyraena jelloi.
E. nipponicum is an oviparous, blood-feeding ectoparasite which infests the gills of the common carp. During its unique lifecycle, two larvae (diporpae, post-oncomiracidial stage) permanently fuse to form the juvenile stage, which then develops into an adult individual [13,14]. E. nipponicum was introduced to Europe from Southeast Asia prior to 1983 [15] and has since become a common parasite of carp with negative impact on their populations [16]. It presents a particular problem for intensive pond carp farming in Europe, which produces over 187,000 tons of carp a year (based on data from 2018) [17].
Despite the economic importance of E. nipponicum, only a handful of studies so far investigated the genes of this parasite, so that to date, only 38 nucleotide and 10 amino acid sequences are deposited in NCBI databases. Key studies focused on (a) genetics and molecular biology [16,[18][19][20][21][22][23][24][25][26], in particular molecular identification and characterisation of key peptidases and their inhibitors, namely cathepsins L, B, and D [21,22], cystatin [20], serpin [23], and a Kunitz-type inhibitor [24]; (b) cytogenetics [27] and phylogenetics, with the aim to further our understanding of monogenean evolution; (c) morphological adaptations to ectoparasitism [13,14,28]; (d) involvement of surface carbohydrates [29] during the fusion process between diporpae and in interaction with the fish host, and finally, (e) the effect of somatic fusion between the two diporpae on the neural system [30][31][32]. In the present study, we report and make publicly available the first global investigation into the biology of E. nipponicum using an integrated transcriptomic and proteomic approach and escribe certain important new aspects of the ectoparasite-host relationship.
The transcripts were annotated using seven databases, resulting in 53% (n = 19,644) of transcripts annotated by at least one database (Table 2; Additional file 2: Table S2).
Analysis of transcript abundance revealed that adult E. nipponicum parasites are transcriptionally active, with the top 100 transcripts representing approximately 39% of total transcription, represented by TPM (Additional file 2: Table S2). Within these abundantly expressed transcripts, uncharacterised proteins prevail. Among the annotated transcripts, ribosome-associated and ubiquitin-related transcripts are highly transcribed. Consistently with the haematophagous strategy used by E. nipponicum, key genes associated with blood feeding and digestion also belong to the most transcribed: they include a number of ferritins (iron storage proteins), Kunitz-type inhibitor KT1 (anticoagulation properties), and a CD59-like transcript (inhibition of the complement cascade) (Additional file 3: Table S3).
Analysis by gene ontology (GO) of the total transcriptome revealed that abundant transcription is associated with GO terms related to binding, protein synthesis, and catalytic activity. In particular, key GO terms associated with the ribosome ( A recent in-depth analysis of proteins secreted by the E. nipponicum had shown that extracellular vesicles (EVs) play a key role at the host-parasite interface. In particular,   Table S6). An analysis of enzymes involved in glycogen synthesis and catabolism revealed low levels of transcription of phosphofructokinase (related to glycogen breakdown; 4.30 TPM), which is in keeping with the fact that in oxygen-rich environment, E. nipponicum energy metabolism relies on its glycogen stores. In comparison, a higher transcription was observed for an enzyme involved in glycogen synthesis, namely fructose 1,6-bisphosphatase (86.34 TPM), which may be required for vitelline cell development during the parasite's egg formation process [35].
Other key pathways with high levels of transcription included pathways associated with signal transduction, which regulate a range of cell function processes and play a critical role in cellular development (Fig. 3

Abundant transcription of peptidases and peptidase inhibitors
Consistently with recent somatic proteomic analysis of adult E. nipponicum by Roudnický and colleagues [12], peptidases and their inhibitors are highly transcribed in the adult transcriptome data. Transcripts associated with  Further in-depth analysis of these transcripts that used the MEROPS peptidase database identified 555 proteases and 149 inhibitors, which were classified into 62 peptidase and 15 inhibitor families (Table 3).
Peptidase classification was consistent with the GO analysis, reflecting a predominance of threonine, Fig. 3 Assignment of the E. nipponicum transcripts to individual KEGG Orthology categories. KEGG pathways [34] (axis y, top 10 for each category) sorted (in ascending order) by their abundance (expressed by the sum of TPM values for all included transcripts, axis x) in the main categories: brown (A09100 Metabolism), blue (A09120 Genetic information processing), black (A09130 Environmental information processing), red (A09140 Cellular processes), orange (A09150 Organismal systems), and purple (A09180 Brite hierarchies). Categories A09160 (Human diseases) and A09190 (Not included in pathway or brite) were excluded. A logarithmic scale was used to display the relative expression of each pathway metallo, serine, and cysteine peptidases (Fig. 1). Of the three peptidase families with threonine peptidase activity, the most abundantly transcribed genes belong to proteasome-related threonine T1 family (PB clan), which reflects intensive protein turnover during this parasite stage. Supporting the protein degradation role played by the proteasome, our analysis shows that genes associated with the metallo peptidase family M67, which plays a critical role in deubiquitination of proteins, are also highly transcribed. Abundant transcription of genes associated with families Serine S1 (chymotrypsin family; PA clan) and Cysteine C2 (calpain family; CA clan) similarly support critical processes such as development and digestion [36] (family Serine S1), as well as signal transduction, cellular differentiation, cytoskeletal remodelling, and vesicular trafficking (family Cysteine C2) [37].
Analysis of the 15 peptidase inhibitory families showed a predominance of inhibitors of serine and cysteine peptidases, specifically inhibitors belonging to family I29, which consists of inhibitors of C1 papain-like cysteine peptidases (Table 4). This is consistent with our previous biochemical characterisation of three key peptidase inhibitors highly expressed in the adult parasite secretome, namely a type I cysteine peptidase inhibitor (EnStef) [20] and two inhibitors of serine peptidases, namely serpin EnSerp1 [23] and Kunitz-type inhibitor EnKT1 [24].
Key E. nipponicum molecules important for blood feeding in adult stage parasites Members of clade Neodermata synthesise haem via haem biosynthesis pathway consisting of eight enzymatic steps [38][39][40]. E. nipponicum, like other haematophagous parasites which after adopting a blood feeding strategy lost the ability to synthesise haem de novo, relies during egg production solely on host blood as a rich source of carbohydrates for energy metabolism, amino acids, and fatty acids. Still, physiological haem plays a critical role in a wide range of this parasite's biological processes [41]. Interestinglyand in contrast to many blood-feeding parasites that lost most enzymes belonging to the haem biosynthesis pathwayhomologues of seven such enzymes have been identified in E. nipponicum transcriptome (Table 5).
In fact, only the transcript for 5-Aminolevulinic acid synthase (ALAS), enzyme responsible for initiation of the pathway, was absent. Further in-depth genomic analysis is required to confirm the absence of this crucial gene and to determine whether this is a feature which E. nipponicum shared with other haematophagous monogeneans. Similarly, functional analysis of the seven identified homologous enzymes would determine whether these enzymes evolved other functions important for the parasite.
Based on the results of structural and histochemical analyses [42,43], one can conclude that the blood digestion process in E. nipponicum resembles intracellular digestion that takes place inside digestive cells and is usually found in ectoparasitic haematophagous mites, such as ticks. Nevertheless, a recent study [21] seems to indicate that an extracellular phase of digestion, in the lumen of the gut, is also present and blood digestion in E. nipponicum thus more closely resembles digestive processes in liver flukes [44] rather than in ticks. Erythrocytes are probably lysed within the monogenean gut lumen, releasing haemoglobin tetramers which are then hydrolysed in specialised haematin cells of the phagolysosome [45]. This process releases iron-rich haem which plays an important role in a number of biological processes [41] and is a crucial component required for egg production [46]. To protect the parasite from haem toxicity and related effects of oxidative stress, iron ions are stored in intracellular iron storage proteins, ferritins [47]. Analysis of E. nipponicum transcriptome shows that ferritins are amongst the most transcribed genes, represented by 36 transcripts (14,024.50 TPM). This finding is consistent with studies by Galay and colleagues which show that multiple ferritins are critical for successful blood feeding and reproduction in hard ticks [48].
Residual free haem is removed by conversion to haematin crystals which are expelled back into the gut lumen and regurgitated by the worm into the outer environment [45]. Similarly to the nematodes and the tick Ixodes ricinus [49], E. nipponicum does not encode a gene for haem oxygenase. The process of haem detoxification by catabolism is therefore mediated by high affinity haem-binding proteins, glutathione S-transferases (GSTs) [49,50], which are abundantly transcribed within the E. nipponicum transcriptome. We identified 29 transcripts, including 24 mu class and two mitochondrial kappa class GSTs (3554.0 TPM).
Cathepsin cysteine peptidases are essential for degradation of host haemoglobin and are abundantly transcribed in the E. nipponicum transcriptome. Consistent with a study by Jedličková and colleagues [21], adult E. nipponicum transcribe mainly cathepsin L peptidases, specifically cathepsin L1 and L3 (n = 36; 3789.31 TPM) at a ratio of We also found a number of transcripts encoding calcium-dependent, non-lysosomal calpain-like proteases (n = 10; 354.97 TPM), which cleave the blood-clotting fibronectin and thereby facilitate parasite feeding [37,51]. Cathepsin D aspartic endopeptidase (n = 5; 314.49 TPM) and aminopeptidases that use a metal ligand within their active site (n = 12; 103.12 TPM), such as aminopeptidase P3 (also known as Xaa-Pro aminopeptidase) and aminopeptidase A (or glutamyl aminopeptidase), also probably play a role in blood digestion, although not to the same extent as similar molecules in other haematophagous ectoparasites, such as ticks.
Our biochemical characterisation of key secreted peptidase inhibitors had shown that they play a critical role during blood feeding [20,23,24]. Kunitz-type inhibitor KT1 is among the most abundantly transcribed genes of the E. nipponicum transcriptome (n = 68; 16,683.0 TPM). This interesting serine protease inhibitor has anticoagulation properties and can impair the host complement  system [24]. The adult parasite also transcribes other serine protease inhibitors, especially serpins (n = 20; 414.49 TPM), though at lower levels of transcription. These serpins have been shown to play a role in the suppression of blood coagulation (by targeting mainly factor Xa), complement activation, and fibrinolysis [23]. Similarly, we identified a number of transcripts encoding a type I cysteine peptidase inhibitor (cystatin/stefin; n = 6; 574.20 TPM), which has been shown to be involved in the regulation of haemoglobin degradation [20].
Proteins probably acting at the host-parasite interface To date, only a handful of proteomic studies have been carried out for E. nipponicum, namely a gel-based analysis of secreted proteins that focused on cathepsin peptidases [21,22] and a study of microdissected tissuespecific somatic proteins [12]. In our study, we conducted a gel-free proteomic analysis of excreted-secreted proteins (ESP), otherwise known as secretome of adult E. nipponicum, which identified 721 proteins with at least two unique peptides (Additional file 7: Table S7). Consistent with the transcriptome analysis, most of these proteins have not been characterised previously. We identified several key proteins involved in blood feeding and digestion, namely a number of ferritins (n = 7; 1.25 NSAF) and glutathione S-transferases (n = 7; 1.95 NSAF), which are abundantly transcribed in the adult transcriptome and consequently also abundantly expressed.
We have previously reported that cathepsin L peptidases dominate the proteolytic activity of adult E. nipponicum secretome, where they in conjunction with cathepsin B peptidases play a critical role in haemoglobin degradation. Both these peptidases were identified in our current secretome analysis (cathepsin L1: 1.26 NSAF; cathepsin B: 0.14 NSAF). Additionally, we observed other key peptidases involved in blood digestion, including (a) calpain, which cleaves the blood-clotting protein fibronectin and therefore has an anticoagulation effect (0.19 NSFA) [51]; (b) aspartic endopeptidase cathepsin D (0.04 NSFA), which probably cleaves haemoglobin [52]; (c) saposin, involved in red cell lysis (0.01 NSFA) [53]; and (d) cathepsin C, also known as dipeptidyl peptidase I (0.13 NSFA), which in schistosomes may in conjunction with leucine aminopeptidases play a role in the terminal hydrolysis of haemoglobin-derived peptides [54].
Despite the high levels of gene transcription, the Kunitz-type inhibitor KT1 previously characterised by us [24] was not highly abundant in the secretome (0.02 NSAF). In fact, marginally higher levels based on NSAF values were observed for serpin (0.24 NSAF) and stefin (0.12 NSAF) inhibitors. In situ hybridisation studies localised the KT1 gene transcript to haematin (or digestive) cells and not the digestive tract [24], which is consistent with its lower expression in the secretome.
Analysis of the most abundant proteins in the secretome of adult E. nipponicum revealed a predominance of proteins involved in immunomodulation, which is critical for parasite survival ( Table 6). The second most abundant single protein was a peptidyl-prolyl cis-trans isomerase (1.37 NSFA), which, as previous studies have shown, is involved in the modulation of dendritic and T cell responses in other parasitic platyhelminths [56,57]. In addition to their role in haemoglobin degradation, cathepsin L peptidases (1.11 NSFA) also play a key role in immunomodulation [21,22,54]. Other important abundant molecules present in the E. niponnicum secretome which probably play a role in the modulation/suppression of host immune response are fatty acid binding proteins (n = 3; 1.05 NSAF) and CD59-like molecules (n = 4; 1.08 NSAF) ( Table 6).
As shown above, adaptation to a blood feeding strategy leads among other things to increased levels of oxidative stress resulting from the release of iron from haemoglobin. To neutralise or reduce the levels of free radicals, the parasite produces a range of antioxidants, including superoxide dismutase (SOD; 0.70 NSAF), thioredoxin (TRX; 0.30 NSAF), and peroxiredoxin (PRX; 0.26 NSAF), which are then found in the secretome.

Discussion
Monogeneans are the most species-rich group of fishinfecting parasites within the phylum Platyhelminthes. They evolved unique morphological adaptations associated with parasitism, including varying shape and size of their attachment organs and hooks, which are widely used for species identification. These traditional morphological methods are nowadays often combined with molecular sequencing technologies to provide more robust determination approaches [70]. On the other hand, despite the advanced methods for classifying these parasites, the amount of molecular and biochemical data pertaining to them is meagre. In this study, we conducted in-depth transcriptome and secretome analyses to further our understanding of adult E. nipponicum parasites, which helped us provide novel insights into this blood-feeding parasite's feeding strategy.
E. nipponicum is an obligate blood-feeding ectoparasite. As a consequence, fish infected with it suffer from decreased levels of haemoglobin [71] and hypochromic microcytic anaemia, characterised by increased ratio of immature red blood cells [2]. The process of blood feeding is initiated following attachment to the fish host, where the gill tissue is mechanically damaged by pressure created by suckers located in parasite's oral cavity, which leads to bleeding from host's superficial capillaries. There is currently no evidence that this process is supported by peptidases secreted by the parasite that would digest the host tissue, as has been reported for mucus-feeding monogeneans which most employ use elastase-like serine peptidases to that purpose [21]. In the case of E. nipponicum, some proteins it secretes are involved in preventing blood coagulation and digestion in its gut lumen and specialised digestive (haematin) cells [45].
In this respect, the E. nipponicum resembles other blood-feeding platyhelminths which initiate haemoglobin processing extracellularly. For instance, the schistosomes digest the bloodmeal extracellularly in the gut lumen using a range of cathepsin peptidases secreted from their gastrodermis [72], while the liver fluke Fasciola hepatica combines extracellular digestion in the gut facilitated by cathepsin L peptidases with intracellular digestion by cathepsin C and aminopeptidases following absorption of haemoglobin peptides in its gastrodermal epithelial cells [44]. Intriguingly, the process of fully intracellular blood digestion resembles most closely the strategy of blood-feeding tick I. ricinus [73], which presents a clear contrast to other blood-feeding arthropods, mainly insects, that rely solely on extracellular digestion [74].
Despite the different location and use of various digestive enzymes that play a role in haemoglobin digestion, the biochemistry E. nipponicum digestion resembles that of other blood-feeding platyhelminths which use a range of cathepsin peptidases and peptidase inhibitors. Our analysis of adult parasite transcriptome and secretome revealed that the parasite abundantly expresses several key peptidases, including cathepsins B, D, L1, and L3, which play a critical role in haemoglobin processing [22]. We have previously shown that serine peptidase inhibitors, Kunitz-type inhibitor EnKT1, and serpin EnSerp1 target host peptidases belonging to the coagulation cascade, such as factors IIa (thrombin) and Xa [23,24]. Our current analysis of the adult transcriptome and secretome had moreover revealed that the genes which encode these peptidase inhibitors are highly transcribed and that the abovementioned inhibitors are indeed secreted by adult parasites, although they are not present in the ESP in large quantities. This secretome analysis is consistent with our previous characterisation of the EnSerp1 [23]. Similarly, in situ hybridisation studies localised the EnKT1 gene transcript to the haematin (or digestive) cells and not the digestive tract [24], which is consistent with its low expression in the secretome.
Our analysis of the transcriptome and secretome had also revealed that ferritin proteins and GSTs play an important role in the life of adult E. nipponicum, in particular its iron and haem processing. Ferritins are globular proteins which store and transport iron ions in a soluble and non-toxic form. Although ferritins are essential for all blood-feeding parasites [75][76][77][78][79], their abundance in the E. nipponicum transcriptome and secretome seems more akin to the blood-feeding strategy of ticks [48]. Also abundantly expressed in the secretome, at comparable levels, are GST molecules which are part of the phase II detoxification system. Studies of blood-feeding ticks and nematodes had shown that in addition to being involved in drug metabolism, these molecules also play an essential role in haem detoxification. In ticks, gene transcription of detoxification enzymes such as the GSTs is upregulated following blood feeding [38,49]. Similarly, it has been shown that nematodes Haemonchus contortus and Caenorhabditis elegans express Nu-class GSTs which can bind the haem [50]. Further investigation is required to determine the class and function of the E. nipponicum GSTs to determine what role they might play in haem detoxification. Both free-living and parasitic nematodes have lost the ability to synthesise haem de novo, a fact reflected in the absence of haem biosynthesis pathway [38]. Platyhelminthes, on the other hand, retained the genes associated with haem biosynthesis pathway, although several parasitic platyhelminths adopted a blood-feeding strategy, which may indicate that the haem biosynthesis pathway may be of importance during nonblood-feeding stages as well [38]. There is currently little information on the potential for de novo haem biosynthesis in monogenean parasites. For instance, the G. salaris genome encodes all enzymes involved in this pathway [38], which is consistent with this parasite's feeding strategy that relies on mucus and skin rather than blood. Our analysis of the E. nipponicum transcriptome shows that the first enzyme in the haem biosynthesis pathway, 5-aminolevulinic acid synthase (ALAS), is absent in its transcriptome. Moreover, all other enzymes belonging to this pathway are transcribed at low levels, which implies that haem biosynthesis is not active in this parasite, at least not during the adult stage. Further analysis is required to determine whether all components of the pathway are present in the E. nipponicum genome, whether they are differently regulated during its distinct lifecycle stages, and to test whether these enzymes are functional. Similarly, further investigation of monogeneans is required to determine whether they all use a conservatived strategy (and keep on synthetising the haem themselves) or whether some of those rely on absorption of host haem.
As an ectoparasite, E. nipponicum is not subjected to the same host immune response that endoparasitic platyhelminths which migrate throughout the host must deal with. Still, E. nipponicum must employ evasion strategies to target molecules within the host blood. Key molecules identified in the E. nipponicum secretome that likely play a role in the modulation/suppression of host immune response include several fatty acid-binding proteins (FABPs) and CD59-like molecules. The function of FABPs in monogeneans is currently unknown, but in F. hepatica these molecules play a role in fatty acid uptake from host blood and in immunomodulation, where they suppress Toll-like receptor (TLR) stimulation and signalling [62,63,80,81]. CD59-like molecules are abundantly expressed in E. nipponicum secretome and they may play a role in modulating/inhibiting the host complement system by molecular mimicry in a fashion similar to that described in other platyhelminths [82]. It has been shown that infections with ectoparasites, such as Ichthyophthirius multifiliis, stimulate the expression of the carp complement system [83] and in salmonid fish, the host complement has a lethal effect on monogeneans Gyrodactylus derjavini and G. salaris [84,85].

Conclusions
In the present study, we explored the transcriptome and secretome of adult E. nipponicum worms using bioinformatic analyses of RNA-seq and LC-MS/MS data. The datasets and results we obtained are unique for parasitic monogeneans because a similarly comprehensive transcriptomic/secretomic study has not been undertaken before. The reported primary dataset can be used for further monogenean research as well as for identification of protein molecules involved in host-parasite interactions. Our insight into transcribed molecules of E. nipponicum revealed a machinery of highly expressed proteins critical for (a) the suppression of anticoagulation processes of the fish host by deployment of protease inhibitors (Kunitztype inhibitors and serpins, CD59-like proteins); (b) digestion of blood proteins (cathepsins) and iron processing (ferritin), and (c) modulation of immune reaction (peptidyl-prolyl cis-trans isomerase, fatty acid binding proteins, and tetraspanin).

Parasite material
E. nipponicum adults were collected during the summer periods from the gills of naturally infected and freshly sacrificed carp (C. carpio) bred in the ponds of a local commercial fishery in southwestern Czech Republic (Rybářství Třeboň, Plc., Třeboň basin, South Bohemia, Czech Republic).

Collection of excretory-secretory proteins (ESP)
ESP were collected from 100 adult worms. Worms were gently washed in sterile tap water and incubated in 10 mM PBS, pH 7.2, for three hours at room temperature in Eppendorf tubes. ESP were purified and concentrated 20 times by centrifugation using an Amicon Ultra 3 kDa column (Merck Millipore) to a final volume of 5 ml. Protein concentration (0.01 μg · μl − 1 ) was determined using Quaint-iT Protein Assay Kit (Life Technologies) and SpectraMax i3 fluorometer (Molecular Devices). The ESP sample was stored at − 80°C until used.

RNA extraction, library preparation, and sequencing
Total RNA was isolated from ten E. nipponicum adults (two independent replicates of five worms) using TriPure Isolation Reagent (Roche) according to manufacturer's instructions. This was followed by DNase I treatment as previously described [86]. RNA concentrations were quantified spectrophotometrically (Nano-Drop 8000, Thermo Fisher Scientific) and fluorometrically (Qubit 2.0, Life Technologies), and integrity was verified by gel electrophoresis using 2100 BioAnalyser (Agilent). GS FLX Titanium Rapid library was prepared from one replicate of five worms using 1.2 μg of total RNA according to manufacturer's instructions (GS FLX Titanium Rapid library preparation kit v. 3.0, 454 Life Sciences). Illumina TruSeq RNA library (non-stranded TruSeq RNA Library Prep Kit v. 2, Illumina) was prepared from 1 μg of total RNA extracted from the second replicate of five worms as previously described [87]. The libraries were sequenced using appropriate sequencing platforms, namely GS FLX Titanium Roche 454 (single-end sequencing) and MiSeq Illumina (short-insert paired-end sequencing, 2 × 100 bp long reads). Sequencing was carried out by BGI Group, Hong Kong (Illumina sequencing) and by the Faculty of Medicine in Hradec Králové, Charles University, Czech Republic (Roche 454 sequencing).

Processing of raw reads, de novo assembly, and annotation
The quality of raw sequencing paired-end Illumina reads, exported in FASTQ format, was evaluated using FastQC v. 0.11.3 [88]. Sequencing adaptors and nucleotides with Phred quality score below 28 were trimmed using Trimmomatic v. 0.33 [89] and sequencing errors and mismatches corrected using SPAdes v. 3.6.0 [90] (BayesHammer tool; −-only-error-correction and --careful modes). Contaminating reads from the fish host were removed using TopHat v. 2.0.14 [91] by aligning RNAseq reads to the carp genome (NCBI Genome ID 10839). Processed reads were finally assembled by Oases v. 0.2.08 [92] with coverage cut-off ranging from 2 to 26 (increasing by one) and k-mers values ranging from 19 to 67 (increasing by two). All assembled transcriptomic datasets were statistically evaluated in the following steps: (a) Basic transcriptome assembly quality analysis was carried out by Transrate v. 1.0.3 [93]; (b) Highly conserved eukaryotic core genes were classified using CEGMA v. 2.5 [94] and BUSCO v. 3.0.1. (Metazoa dataset) [95]; (c) The raw sequencing reads were mapped to the assembled contigs by Burrows-Wheeler Aligner (BWA-backtrack algorithm) v. 0.7.13 [96], and (d) Transdecoder v. 3.0.1 [97] was used to calculate the representation of nucleotide sequences encoding a protein (from each nucleotide contig only the longest protein-coding part was selected, with minimal protein length 30 amino acids).
Read information from the SFF file generated by Roche 454 sequencer was extracted and converted into a FASTQ format using tool sff2fastq v. 0.9.2 [98]. The quality of raw reads was evaluated using FastQC v. 0.11.5 [88]. Adaptors and nucleotides with Phred quality score below 18 were discarded by Trimmomatic v. 0.36 [89] and sequencing errors corrected by Pollux v. 1.0.2 [99]. Contaminating reads from the fish host were identified by Burrows-Wheeler Aligner (BWA-SW algorithm) v. 0.7.13 [100] by mapping to the carp genome. Final assembly was performed by SPAdes v. 3.9.0 [90] (rnaSPAdes tool) based on three datasets generated from the Illumina data (k-mer value 53 and coverage cut-off value 8; k-mer value 57 and coverage cut-off value 9; kmer value 55 and coverage cut-off value 10) and four datasets generated from the Roche 454 reads (k-mer values 99, 95, 87, and 83). Statistical evaluation of assemblies was performed as above. Duplicate sequences were removed after clustering using CD-HIT-EST (nucleotide identity threshold 95%) [101]. The final transcript dataset used for further analysis was based on sequences representing the longest open reading frames encoding at least 30 amino acids, which were selected by Transdecoder.
All transcripts with high sequence similarity to potential contaminating sequences (virus, bacteria, cyanobacteria, yeast, fungi, algae, green plants, or carp) were removed from the dataset prior to further analysis. Additionally, we also excluded any sequences with a potential open reading frame of less than 50 amino acids with no putative annotation. Further in silico analysis was conducted using the following tools: (a) KAAS (KEGG Automatic Annotation Server [34,109]); (b) Gene Ontology (GO) prediction was performed using InterProScan v. 5.30-69.0 [110] with default search parameters; (c) transcript abundance was quantified using RSEM v. 1.3.1 [111] by mapping trimmed and corrected Roche 454 and Illumina reads onto the final transcriptome. The resulting TPM values were averaged, which was necessitated by the varied character of sequential data (single-end and paired-end), while Roche 454 and Illumina reads have to be mapped separately. Transcription abundance was measured by the sum of the TPM values of all the participating transcripts in a given set. Additionally, we calculated a ratio between the TPM value and the number of participating transcripts was calculated to provide information regarding how many transcripts were responsible for transcription abundance. GO terms and KEGG pathways with less than 10 transcripts were excluded from this analysis.
Identification of the excretory-secretory proteins by mass spectrometry ESP sample was digested using filter-aided sample preparation with 1 μg of trypsin (sequencing grade, Promega). For peptide separation for MS/MS analysis, we used UltiMat 3000 RSLCnano liquid chromatography (LC) system (Thermo Fisher Scientific). Separation was achieved using a capillary column filled with the nonpolar stationary phase (500 mm × 75 μm, C18 anchors, 3 μm particles, Acclaim PepMap, Thermo Fisher Scientific) during a 135 min gradient elution (0.5 μl · min − 1 ). Mobile phase consisted of a polar (0.1% formic acid (FA)) and nonpolar phase (80% acetonitrile (ACN), 0.1% FA). Eluted peptides (2 μg) were ionised by a nanospray (PicoView 550 nano source) and analysed in a mass spectrometer (Orbitrap Elite, Thermo Fisher Scientific). MS data were acquired in a data-dependent mode, selecting up to top ten precursors based on precursor abundance in the survey scan (resolution 60,000 in the range 350-2000 m/z). Maximum accumulation time for MS/MS spectra acquisition was 500 ms (resolution 15, 000 at 400 m/z) and isolation window for fragmentation was set to 2 m/z. The resulting MS data were recalibrated using 445.120028 signal from the first 10 min and used for identification of proteins. Mass spectrometric raw data files were analysed using Proteome Discoverer software (Thermo Fisher Scientific; v. 1.4) with in-house Mascot search engine (Matrixscience; v. 2.5.1.3) set up to search an in-house protein database containing 37, 062 protein sequences derived from the E. nipponicum transcriptome sequencing, carp-specific proteins derived from C. carpio genome (63,928 sequences, NCBI Genome 10839), and cRAP contaminants (110 sequences). Modifications for all database searches were set as follows: oxidation (M), deamidation (N, Q), and acetylation (Protein N-term) as variable modifications, with carbamidomethylation (C) as a fixed modification. Enzyme specificity was semitryptic with one allowed miscleavage. Percolator was used for postprocessing of search results. Only peptides with q-value < 0.05, rank 1, and with at least six amino acids were considered. LC-MS/MS analysis was conducted at Proteomics Core Facility, CEIT EC, Masaryk University, Czech Republic. Availability of data and materials This Transcriptome Shotgun Assembly project was deposited at DDBJ/ENA/ GenBank under the accession GFYM00000000. The version described in this paper is the first version, GFYM01000000. Raw sequence data were deposited in the NCBI SRA database under accession numbers SRX2995121 (Roche 454) and SRX2995122 (Illumina). Mass spectrometry proteomics data were deposited to the ProteomeXchange Consortium via the PRIDE partner repository under dataset identifier PXD017293. Transcriptomic sequences and annotation are available on GitHub repository Eudiplozoon-nipponicumtranscriptome-secretome (https://github.com/jirivorel/Eudiplozoonnipponicum-transcriptome-secretome).

Declarations
Ethics approval E. nipponicum is a species that is not protected under the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). All procedures performed in studies involving animals were carried out in accordance with European Directive 2010/63/EU and relevant Czech legislation (Act no. 246/1992 Coll. and Act no. 359/2012 Coll.) which regulates research involving animals. All experiments were performed with legal consent of the Animal Care and Use Committee of Masaryk University and Research and Development Section of the Ministry of Education, Youth and Sports of the Czech Republic.