Introduction

Gyrodactylid monogeneans are widespread parasites that inhabit the skin, fins, and gills of various teleost fishes. These helminths are viviparous and contain several generations of embryos in utero in the “Russian-doll” manner (Bakke et al. 2007). They are considered as one of the most invasive fish parasites due to their superior dispersal ability and viviparous mode of reproduction, and infection with these parasites results in high host morbidity and mortality (Cable and Harris 2002). To reduce economic losses caused by Gyrodactylus, many useful strategies are employed including the breeding of new disease-resistant strains and frequent use of various insecticides (Schelkle et al. 2009). However, these practices have disadvantages, such as lengthy breeding cycles and adverse environmental impacts. It has been indicated that fish can mount an effective immune response against gyrodactylids and finally remove these parasites (Jorgensen et al. 2009). Hence, understanding the underlying immune mechanisms and the relationship between host and pathogen will facilitate the development of effective methods to prevent infection with this parasite.

Studies have been conducted in recent decades to elucidate the immune mechanisms behind the ability of fish to control the infection of Gyrodactylus (Collins et al. 2007; Matejusova et al. 2006). A hypothetical model for the intricate immune mechanisms in fish against Gyrodactylus was proposed in which the expression of cytokines in the host epidermis could initiate host immune responses and then decrease the parasite population (Buchmann 1999). In our previous study, a similar result was also observed, where the expression of IL-1β2 significantly increased in goldfish (Carassius auratus) in response to G. kobayashii infection (Zhou et al. 2018). Tu et al. (2019) also shown that G. kobayashii infection induced significant upregulation of inflammatory cytokines including IL-1β, TNF-α, TGF-β, and IL-4 in goldfish. Although, the primary focus of these research is the role of cytokines during gyrodactylid infection. There is still a need for comprehensive analysis of the underlying molecular mechanisms governing the ability of host to remove these parasites.

Recently, high-throughput sequencing technology has already been employed in aquatic animals to clarify host defense mechanisms against parasite infection. For example, some immune-related molecules such as chemokines, chemokine receptors, toll-like receptors, and complement components were significantly upregulated in Epinephelus coioides skin after Cryptocaryon irritans infection, indicating that the skin’s local immune response is crucial for host defense against ectoparasite infection (Hu et al. 2017). Similarly, the complement and coagulation cascade pathway has been found playing an important role in the early stages of C. irritans infection in Larimichthys crocea using transcriptome analysis (Yin et al. 2016). Also, the myxozoan parasite Tetracapsuloides bryosalmonae infection could induce significant upregulation of cytokines and associated genes including chemokines, interferons, interleukins, tumor necrosis factors, transforming growth factors, and colony-stimulating factors in the posterior kidney of Salmo trutta (Sudhagar et al. 2019). Transcriptome analysis is thus a formidable tool to provide a comprehensive view of host immunity and understand the underlying immune mechanisms against parasite infection. However, less information is available on the transcriptome analysis of aquatic organisms in response to gyrodactylid monogenean.

To comprehensively elucidate the immune mechanisms against G. kobayashii, next-generation sequencing herein was used to identify DEGs at the transcriptome level in goldfish after G. kobayashii infection. The fish skin is the main parasitic site of G. kobayashii and a vital tissue for communicating with the external surrounding. Hence, the skin was selected for transcriptome analysis to clarify the molecular mechanisms in response to gyrodactylid infection. In addition, several randomly selected DEGs were verified by qRT-PCR to determine the reliability of the RNA-Seq results. Also, some serum biochemical factors including alkaline phosphatase (ALP), superoxide dismutase (SOD), malondialdehyde (MDA), and nitric oxide (NO) were determined. Our results revealed significant changes in gene expression levels and serum biochemical factors in goldfish after infected with G. kobayashii, which would promote a better understanding of the molecular mechanism underlying the goldfish-G. kobayashii interaction.

Materials and methods

Fish and parasites

The goldfish-G. kobayashii model was established according to the methods used in our previous study (Zhou et al. 2017a). In short, goldfish were obtained from a local fish farm and subjected to continuous baths with formalin solutions to remove all ectoparasites. Subsequently, ectoparasite-free goldfish was placed near the caudal fin of an infected fish, allowing an individual parasite to transfer. Ten days later, two worms from newly infected fish were collected for species identification to confirm they are G. kobayashii. Ectoparasite-free goldfish were introduced into the aquarium with newly infected goldfish to obtain more G. kobayashii-infected goldfish. All animal experiments were approved and conducted in compliance with the experimental practices and standards developed by the Animal Welfare and Research Ethics Committee of Yangtze River Fisheries Research Institute.

Experimental infection and tissue sampling

Thirty-six healthy and ectoparasite-free goldfish with body weight of 79.6 ± 7.7 g were randomly divided into six tanks, each tank containing six goldfish and 45 L of dechlorinated tap water. One week prior to processing, three tanks of fish were randomly selected for G. kobayashii infection and termed as infected groups. Each goldfish from the infected groups was anesthetized with 0.02% MS-222 (tricaine methane sulfonate) and put in contact with the caudal fin of an infected fish reared in our laboratory, allowing 30 G. kobayashii to transfer between hosts (Zhou et al. 2017b). Subsequently, these fish were released into tanks and maintained in static dechlorinated tap water with a 12-h light-dark cycle at 20 ± 1 °C and fed to satiation with commercial pellet feed daily. The remaining three tanks of fish were termed as control groups and set up under the same condition.

Our previous study indicated that goldfish are capable of mounting an immune response against gyrodactylid 7 days after infection (Zhou et al. 2018). Thus, to understand the immune mechanisms of goldfish against G. kobayashii during the early infection stage, 9 days post-infection was selected as the sampling time point. Nine days after infection, the number of gyrodactylids on the caudal fins of all experimental fish was examined for parasitological analysis. Three individuals from infected and control groups were selected for blood collection from the caudal vein of fish, and the serum was obtained and stored at − 20 °C for biochemical analysis. Then, these fish were killed by an overdose of MS222; skin sample from the left side near the caudal fin was dissected and immediately frozen in liquid nitrogen and then stored at − 80 °C before use for RNA extraction.

Serum biochemical analysis

At the termination of the experiment, the enzymatic activities (ALP and SOD) as well as the contents of MDA and NO in fish serum were determined by using the commercial kits (Nanjing Jiancheng Bioengineering Institute, China) according to the manufacturer’s instructions.

RNA extraction, library construction, and sequencing

Total RNA from control and infected groups, termed as C1, C2, C3, I1, I2, and I3, respectively, was extracted using TRlzol Reagent (Life Technologies, USA) according to the instruction manual. The RNA concentration and integrity were examined using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., USA). Subsequently, the mRNA was isolated by NEBNext Poly(A) mRNA Magnetic Isolation Module, and the cDNA libraries were constructed using the NEBNext Multiplex Oligos and NEBNext Ultra RNA Library Prep Kit for Illumina according to the manufacturer’s instructions. Finally, six constructed cDNA libraries were sequenced using an Illumina HiSeqX sequencing platform with a paired-end sequencing strategy.

Reads mapping and sequence annotation

The raw reads were firstly filtered using fastp v 0.20.0 by eliminating adaptor sequences and low-quality reads, such as unknown nucleotides > 5% or Q20 < 20% (Chen et al. 2018). Then, all clean reads were mapped to goldfish (C. auratus) reference genome (https://www.ncbi.nlm.nih.gov/genome/?term=goldfish) using Tophat2 software (Kim et al. 2013).

Differential expression analysis and enrichment analysis

Gene expression levels were analyzed using FPKM values (fragments per kilobase of exon per million fragments mapped values) by the Cufflinks software v 2.1.1 (Trapnell et al. 2010). The significance of DEGs between control and infected groups was calculated using the package DESeq with a corrected p value ≤ 0.05 as a threshold for significance (Trapnell et al. 2012). Also, the false discovery rate (FDR) control method was introduced, and the genes with an absolute value of log2 ratio ≥ 1 and FDR < 0.05 were selected for further analysis. Subsequently, Gene Ontology (GO) functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of DEGs were conducted using TopGO v.3.10 and Fisher’s exact test in R v3.6 (http://www.geneontology.org and https://www.kegg.jp/). Bonferroni correction was used to adjust p values, and pathways with corrected p values ≤ 0.05 were considered significantly enriched in DEGs (Benjamini and Hochberg 1995).

Experimental validation using quantitative real-time PCR

To investigate the reliability of the RNA-Seq results, 12 DEGs were selected and verified by qRT-PCR (Table 1). The primers were designed based on the transcriptome sequences using the Primer3 software v. 4.1.0 (http://primer3.ut.ee/), and the housekeeping gene β-actin was selected as the internal control. The qRT-PCR was performed on a QuantStudio 5 real-time PCR system (Applied Biosystems, USA). Each PCR reaction contained 10 μl ChamQ™ Universal SYBR qPCR Master Mix (Vazyme), 0.4 μl of each primer (10 μM), 2 μl of cDNA template, and 7.2 μl of dH2O. The PCR program included an initial denaturation at 95 °C for 30 s, and 40 cycles of 95 °C for 10 s and 60 °C for 30 s, followed by a melting curve analysis to verify the PCR specificity. The relative expression levels of each gene were calculated using the 2−ΔΔCt method, and the differences in each gene expression between control and infected groups were analyzed by Student’s t test using SPSS 20.0 (IBM Corp 2011), and the statistical significance is highlighted with asterisks in the figures as follows: p > 0.05, not significant; p < 0.05 (*), significant; p < 0.01 (**), extremely significant.

Table 1 The primers of genes used for quantitative real-time PCR analysis
Full size table

Results

Parasite infection and serum biochemical analysis

In the infected group, mean abundance of G. kobayashii on the caudal of goldfish increased to 198.7 ± 109.5 worms/fish from 30 worms/fish whereas no record of infection with gyrodactylids was observed in the control group. Serum samples from infected and control groups were obtained and used for biochemical analysis, and the results are shown in Table 2. G. kobayashii infection induced significant increases in the ALP activity and NO content in goldfish serum. Slight but not statistically significant changes in the SOD activity and MDA content between infected and control groups were detected.

Table 2 Effect of experimental infection of Gyrodactylus kobayashii on alkaline phosphatase (ALP), superoxide dismutase (SOD) as well as the contents of malondialdehyde (MDA) and nitric oxide (NO) in goldfish (Carassius auratus) serum. Values are expressed as mean ± SE (n = 3) in the columns with different superscripts differ significantly (p < 0.05)
Full size table

Sequencing and reads mapping

Six cDNA libraries were sequenced from goldfish skin from control and infected groups, and the characteristics of these libraries are summarized in Table 3. Approximately 7.53, 6.34, 6.29, 6.31, 6.31, and 6.2 G raw reads were produced from the C1, C2, C3, I1, I2, and I3 samples, respectively. After removing low-quality reads, the number of clean reads per library ranged from 6.03 to 7.43 G with a Q30 percentage of over 91.4%. Among the clean reads, the number of sequences mapped to the reference genome ranged from 16,702,308 to 19,114,800, and the percentage of total mapped reads ranged from 77.12 to 85.05% in different libraries. Raw sequencing reads data have been submitted to Sequence Read Archive in NCBI under the accession numbers SAMN12726145–SAMN12726150, respectively.

Table 3 Summary statistics of the transcriptome sequences from skin tissues of goldfish (Carassius auratus)
Full size table

Differential expression analysis and GO enrichment analysis

A total of 556 genes were differently expressed, including 344 upregulated and 212 downregulated genes (Fig. 1). To better understand the biological functions of DEGs, all DEGs were first mapped to the GO database and GO analysis indicated that DEGs were classified into three major functional categories including biological process, cellular component, and molecular function. In the biological process, a total of 287 DEGs were categorized as single-organism process, cellular process, and metabolic process. In the cellular component category, the major category represented was cell, cell part, and membrane part. In the case of molecular function, the category of binding and catalytic activity was the most strongly represented (Fig. 2).

Fig. 1
figure 1

Volcano plot of distribution trends for differentially expressed genes in goldfish (Carassius auratus) skin after challenge with Gyrodactylus kobayashii. The log2 fold (infected group/control group) indicates the expression level for each gene. Red dots and gray dots represent the differentially expressed genes and nondifferentially expressed genes, respectively

Full size image
Fig. 2
figure 2

Gene ontology (GO) annotation of the differentially expressed genes (DEGs) in the three main GO categories: biological process, cellular component, and molecular function. The x-axis means most represented GO enriched categories of the genes, and the y-axis shows the percentage of genes annotated to a GO term in all the GO annotation genes. The red column represented all the genes annotated to a GO term, and blue column represented DEGs annotated to a GO term

Full size image

KEGG pathway analysis and DEGs associated with immune response

All DEGs were further mapped to the KEGG database, and 380 DEGs were successfully annotated to 95 signaling pathways in KEGG, and the top 30 enriched KEGG pathways are shown in Fig. 3. Fourteen pathways related to immune response were identified, including mTOR signaling pathway (ko04150, 17 genes); cytokine-cytokine receptor interaction (ko04060, 13 genes); glycine, serine, and threonine metabolism (ko00260, 5 genes); intestinal immune network for IgA production (ko04672, 4 genes); toll-like receptor signaling pathway (ko04620, 4 genes); phagosome (ko04145, 5 genes); peroxisome (ko04146, 3 genes); RIG-I-like receptor signaling pathway (ko04622, 2 genes); apoptosis (ko04210, 5 genes); MAPK signaling pathway (ko04010, 9 genes); TGF-beta signaling pathway (ko04350, 3 genes); NOD-like receptor signaling pathway (ko04621, 1 gene); salmonella infection (ko05132, 1 gene); and herpes simplex infection (ko05168, 1 gene). A few vital immune-related DEGs are shown in Table 4. In addition to immune-related pathways, the DEGs were also assigned to metabolic pathways (ko01100, 28 genes), endocytosis (ko04144, 15 genes), focal adhesion (ko04510, 11 genes), regulation of actin cytoskeleton (ko04810, 11 genes), neuroactive ligand-receptor interaction (ko04080, 11 genes), and calcium signaling pathway (ko04020, 10 genes) (Table S2).

Fig. 3
figure 3

Scatter chart of pathway enrichment analysis. The x-axis is the enrichment factor ((the number of DEGs annotated in this pathway / the number of all DEGs) / (the number of genes annotated in this pathway / the number of genes annotated in all pathways)), and the y-axis means various pathways. The magnitude of the dots displays the number of DEGs annotated to the pathway, and the colors of dots represent corrected p values

Full size image
Table 4 Representative immune-related differentially expressed genes after Gyrodactylus kobayashii infection
Full size table

Validation of RNA-Seq results by qRT-PCR

Twelve DEGs were selected and validated using qRT-PCR, and the statistical results were as follows: CXCR4 (t = − 8.93, p = 0.012), TNF-10 (t = 46.6, p = 0), CCL-25 (t = 18.26, p = 0.003), Ig (t = 8.23, p = 0.014), DDIT4 (t = − 5.57, p = 0.031), HSP70 (t = − 2.53, p = 0.127), C7 (t = − 5.34, p = 0.033), IL-17RD (t = 15.81, p = 0.004), P450 (t = 51.41, p = 0), TNFRSF19 (t = 7.75, p = 0.016), IL8 (t = − 5.8, p = 0.028), and PIK3R (t = − 6.41, p = 0.023). Fold changes of these genes between RNA-Seq and qRT-PCR were compared, and the results are listed in Fig. 4. A similar trend of these genes between RNA-Seq and qRT-PCR was observed, indicating that RNA-Seq data was reliable.

Fig. 4
figure 4

Validation of differentially expressed genes by qRT-PCR. The x-axis displays 12 DEGs including CXCR4 (C-X-C chemokine receptor type 4-B), TNF-10 (tumor necrosis factor ligand superfamily member 10), CCL-25 (C-C motif chemokine 25), Ig (immunoglobulin superfamily member 5), DDTT4 (DNA damage-inducible transcript 4 protein), HSP70 (heat shock 70 kDa protein), C7 (complement component C7), IL-17RD (interleukin-17 receptor D), P450 (cytochrome P450 26A1), TNFRSF19 (tumor necrosis factor receptor superfamily member 19), IL-8 (interleukin-8), and PIK3R (phosphatidylinositol 3-kinase regulatory subunit gamma). The y-axis represents relative fold change, and the data represent the means for three independent experiments and error bars indicate SD (*p < 0.05; **p < 0.01)

Full size image

Discussion

Species of Gyrodactylus are common monogenean parasites, and infection with these parasites results in high mortalities and significant economic losses, being a significant obstacle for the long-term sustainability of aquaculture (Bakke et al. 2007). In the present study, the transcriptome profiles of goldfish skin after challenge with G. kobayashii were investigated using RNA-Seq. To our knowledge, this is the first comprehensive analysis of the global transcriptome profile of goldfish in response to gyrodactylid monogenean infection. These results will help us better understand the host-parasite interactions and the potential molecular mechanism in response to monogenean infection.

In this study, a total of 556 DEGs were obtained and 380 DEGs were successfully annotated to 95 signaling pathways including metabolic pathways, mTOR signaling pathway, endocytosis, cytokine-cytokine receptor interaction, and focal adhesion, which indicated that complicated molecular mechanisms were involved in G. kobayashii infection in goldfish. Some DEGs displayed involvement in biological functions, such as metabolic pathways, insulin signaling pathway, fructose and mannose metabolism, adipocytokine signaling pathway, and arginine and proline metabolism, suggesting that G. kobayashii infection had a considerable effect on metabolic processes of goldfish. Similar results were also found in Chinese grass shrimp (Palaemonetes sinensis) infected with isopod parasite Tachaea chinensis (Li et al. 2018). Also, abundant immune-related genes including TNF-10, C-C motif chemokine 25, HSP70, C7, IL-17RD, P450, IL-8, and PIK3R were identified and several immune-related pathways such as mTOR signaling pathway, cytokine-cytokine receptor interaction, toll-like receptor signaling pathway, and phagosome were enriched. To clarify the immune defense mechanisms in response to gyrodactylid infection, as discussed below, several key pathways likely regulating goldfish immune response against G. kobayashii infection were emphasized.

The mechanistic target of rapamycin (mTOR) is an evolutionarily conserved serine/threonine kinase that plays an essential role in host immune defense against invading pathogens (Keating and McGargill 2016; Shapira and Zinoviev 2011). In our trial, frizzled-3-like isoform X1 and several Wnt protein genes including Wnt-3a-like, Wnt-7a-like, Wnt-4a isoform X1, Wnt-3a, and Wnt-7a were significantly downregulated after G. kobayashii infection. These genes were enriched to mTOR signaling pathway, and the downregulation of these genes might trigger a series of cascade reactions to inhibit mTORC1, thereby resulting in the suppression of type I IFN expression, an important cytokine combating pathogens (Bogdan et al. 2004). The result was supported by the result that IFN-γwas significantly downregulated in goldfish skin during the early stages of G. kobayashii infection (Zhou et al. 2018). These reflected gyrodactylids might suppress host immune response to establish infection by regulating mTOR signal pathway. In addition, gyrodactylids are epidermal browsers and feed on epidermal cells and mucus (Bakke et al. 2007). mTOR signaling pathway is a key regulator of protein synthesis, and in this study, the inhibition of mTOR pathway could decrease host protein synthesis, thereby suppressing epidermal cell growth and mucus secretion. These results suggested that host might inhibit protein synthesis by mTOR signaling pathway to reduce the available nutrients for gyrodactylids, thereby inhibiting the population growth of this parasite. Thus, the arms races in host-parasite interaction occurred and mTOR signaling pathway might play a vital role in this interaction.

Cytokines are soluble extracellular proteins or glycoproteins that play crucial roles in host immune defenses against invasive pathogens (Sedger et al. 2014). Previous studies suggested that many cytokines were activated in fish infected with some parasites such as C. irritans and T. bryosalmonae (Hu et al. 2017; Sudhagar et al. 2019). Our previous study also indicated that expression levels of several cytokines such as IL-1β2, TNF, and TGFβ were significantly changed in goldfish after challenge with G. kobayashii (Zhou et al. 2018). In this trial, 10 immune-related DEGs were annotated to cytokine-cytokine receptor interaction signaling pathways, and most of them such as TNFRSF1B, TNFRSF19, and TNFRSF10 were significantly downregulated. TNF receptor superfamily (TNFRSF) proteins are a class of cell surface receptors that regulating the immune system (Silke and Brink 2009). For example, TNFRSF10 has been proved as a tumor suppressor which induced cell apoptosis in various tumors (Wen et al. 2016). Here, several TNFRSF-encoding genes were significantly downregulated, and this might reflect a parasite-induced downregulation of cytokines and corresponding receptors as an immune evasion strategy during the early stage of infection.

In addition to the abovementioned cytokines and cytokine receptors, various immune-related DEGs were also annotated to intestinal immune network for IgA production and toll-like receptor signaling pathway. For example, pro-inflammatory cytokine interleukin 8 (IL-8), which is essential to neutrophil recruitment and neutrophil degranulation (Zou et al. 2015), was significantly upregulated after gyrodactylid infection. Likewise, phosphatidylinositol 3-kinase regulatory subunit alpha (PIK3Rα) and PIK3Rγ were also upregulated, possibly promoting host inflammatory response (Wang et al. 2016). These results demonstrated that the pro-inflammatory cytokines were involved in host immune response against gyrodactylids at the early phase of infection. However, in this study, C-C motif chemokine 25 (CCL25), which is homeostatic chemokine and plays a critical role in recruiting leukocytes to the sites of inflammation (Comerford et al. 2006; Silva-Santos 2012), was downregulated in infected goldfish. In general, inflammation must be controlled to reduce the tissue damage caused by overactivation of immune cells (Boneberg and Hartung 2002). In this trial, gyrodactylid infection caused the upregulation of pro-inflammatory cytokines and downregulation of anti-inflammatory cytokines and then induced excessive inflammation, thereby leading to damage of goldfish skin. The results were supported by earlier histological observations that gyrodactylid infection induced obvious tissue damage on the skin surface (Bakke et al. 2007; Sterud et al. 1998).

Besides, there were also numerous DEGs participating in other immune-related pathways. Integrins have been proved to be involved in phagocytosis and play key roles in modulating immunity including the proPO system, phagocytosis, and the antioxidant system (Lin et al. 2013). In this study, integrin β and integrin α were significantly upregulated after G. kobayashii infection. Heat shock proteins (HSPs) have been reported to serve as molecular chaperones, which participated in the process of cytoprotection by repairing damaged protein (Roberts et al. 2010). So, increased expression of HSP70 in this trial suggested the repairs of damaged proteins caused by G. kobayashii occurred. These might reflect that goldfish had initiated a complicated immune response to resist gyrodactylid infection.

Several serum biochemical factors of goldfish were also determined in our trial. ALP has still considered an important component of innate immunity, and significant increases of the ALP activity indicated that goldfish can initiate an innate immune response against G. kobayashii (Lallès 2019). Also, ALP is capable of dephosphorylating pro-inflammatory components, and increased ALP activity has been viewed as clear signs of higher systemic inflammation, which is consistent with transcriptomic result that excessive inflammation occurred (Lalles 2014). NO is an intercellular messenger that governs a variety of biological processes, and many studies have indicated that parasite infections trigger increased production of NO, which can protect host by killing parasites or limiting growth of parasites (Brunet 2001). Similarly, evident increases in NO were observed in our study. Nevertheless, increased NO levels might be accompanied by detrimental effects against host due to its significant cytotoxicity (Keita et al. 2000). Therefore, further studies are still necessary to clarify the specific mechanisms of NO against G. kobayashii.

In the present study, the transcriptome profiles of goldfish skin after challenge with G. kobayashii were first investigated to explore the molecular mechanisms against gyrodactylid monogenean using next-generation sequencing. A total of 556 DEGs were identified, and KEGG analysis indicated that many pathways were involved in the host immune response mainly including mTOR signaling pathway, cytokine-cytokine receptor interaction, intestinal immune network for IgA production, toll-like receptor signaling pathway, and phagosome. In summary, this study provided better understandings of immune defense mechanisms of goldfish against G. kobayashii, which would support future molecular research on gyrodactylids and facilitate the prevention and treatment of gyrodactylosis in aquaculture.