Contents lists available at ScienceDirect

Fish and Shellfish Immunology

journal homepage: www.elsevier.com/locate/fsi

Full length article

Identification of immune-related genes in gill cells of Japanese eels (Anguillajaponica) in adaptation to water salinity changes

Jie Gua,1, Shuya Daia,1, Haitao Liua, Quanquan Caob, Shaowu Yinb, Keng Po Laic,William Ka Fai Tsed,∗, Chris Kong Chu Wonge, Haifeng Shia,∗∗

a Institute of Life Science, Jiangsu University, Zhenjiang, Jiangsu, 212000, Chinab College of Life Sciences, Key Laboratory of Biodiversity and Biotechnology of Jiangsu Province, Nanjing Normal University, Nanjing, Jiangsu, 210023, Chinac Department of Chemistry, City University of Hong Kong, Kowloon, Hong Kongd Faculty of Agriculture, Kyushu University, Fukuoka, Japane Department of Biology, Hong Kong Baptist University, Kowloon, Hong Kong

A R T I C L E I N F O

Keywords:Immune-related genesAdaptive immune responseEnvironmental salinityRNA-seq

A B S T R A C T

The changes in ambient salinity influence ion and water homeostasis, hormones secretion, and immune responsein fish gills. The physiological functions of hormones and ion transporters in the regulation of gill-osmor-egulation have been widely studied, however the modulation of immune response under salinity changes is notdetermined. Using transcriptome sequencing, we obtained a comprehensive profile of osmo-responsive genes ingill cells of Japanese eel (Anguilla japonica). Herein, we applied bioinformatics analysis to identify the immune-related genes that were significantly higher expressed in gill pavement cells (PVCs) and mitochondrial-rich cells(MRCs) in freshwater (FW) than seawater (SW) adapted fish. We validated the data using the real-time qPCR,which showed a high correlation between the RNA-seq and real-time qPCR data. In addition, the im-munohistochemistry results confirmed the changes of the expression of selected immune-related genes, includingC-reactive protein (CRP) in PVCs, toll-like receptor 2 (TLR2) in MRCs and interleukin-1 receptor type 2 (IL-1R2)in both PVCs and MRCs. Collectively our results demonstrated that those immune-related genes respond tosalinity changes, and might trigger related special signaling pathways and network. This study provides newinsights into the impacts of ambient salinity changes on adaptive immune response in fish gill cells.

1. Introduction

Environmental salinity is an important factor for aquaculture andcould affect physiological responses of aquatic organisms. In fish, en-vironmental salinity changes influences osmoregulation, hormonalcontrol, energy metabolism and growth [1–3]. Japanese eel is culturedin aquaculture ponds, and is an important commercial fish in East Asia.In addition to dietary consumption, some proteins isolated from the fishare used for medical purposes. For example, a fluorescent protein calledUnaG isolated from muscles of Japanese eel could be applied forscreening toxins that can trigger liver disease [4]. Although the salinityis an important factor considered for the growth of Japanese eel [5], itis difficult to avoid salinity alterations in freshwater of open-air culturedue to natural (e.g raining or infiltration of groundwater) or man-made(e.g feeding or drugs) causes.

It suggests that environmental salinity not only affects these

physiological processes (i.e osmoregulation, ion transport and hor-monal control), but also alters the immune system in fishes [6,7]. Ingilthead seabream (Sparus aurata L.), a significant increase in plasmaIgM level was observed after acclimation at high salinity condition for14 days, while enhancement peroxidase content and complement ac-tivity in plasma were detected after a longer term of acclimation (100days) [6]. In brown trout (Salmo trutta), phagocytic activities of thepronephric leucocytes and the lysozyme concentrations were sig-nificantly increased after transfer from freshwater to seawater [8]. Inaddition, acute exposure of tilapia to hyperosmotic conditions wasfound to have immunostimulatory effects on its cellular immune reac-tions (phagocytosis and respiratory burst activity) and humoral re-sponses (lysozyme activity and complement activity) [9]. Moreover,fresh water to seawater transfer in rainbow trout showed sustainedelevation in total white blood cell counts, increased plasma but de-creased mucus lysozyme, and enhanced head kidney macrophage

https://doi.org/10.1016/j.fsi.2017.12.026Received 24 August 2017; Received in revised form 6 December 2017; Accepted 17 December 2017

∗ Corresponding author. Attached Promotive Center for International Education and Research of Agriculture, Faculty of Agriculture, Kyushu University, Fukuoka, Japan.∗∗ Corresponding author. Xuefu Road 301, Zhenjiang, Jiangsu, China.

1 These authors contributed equally to this work.E-mail addresses: kftse@agr.kyushu-u.ac.jp (W.K.F. Tse), shihf@ujs.edu.cn (H. Shi).

Fish and Shellfish Immunology 73 (2018) 288–296

Available online 18 December 20171050-4648/ © 2017 Elsevier Ltd. All rights reserved.

T

respiratory burst activity [10]. Increased activity and proliferation ofimmune cells were found in pipefish (Syngnathus typhle) after acutesalinity change, but the expression of the immune genes granulocytecolony-stimulating factor precursor (GRCSF) and interleukin 10 (IL10)were significantly decreased with increasing salinity on day one [11].In kidney of striped catfish (Pangasianodon hypophthalmus, Sauvage),chronic hyperosmotic stress inhibited toll-like receptors (TLRs) ex-pression, and the down-regulations of TLRs were aggravated whenexposed to bacterial infection [52]. Taken together, acute and chronicsalinity change could stimulate both cellular and humoral immuneactivities; but also suppress the expressions of some immune-relatedgenes.

In this report, we would like to understand if the immune-relatedgenes are responsive to salinity changes. Recently, our group had ob-tained the transcriptome profiles of gill cells from Japanese eel (Anguillajaponica) adapted in freshwater and seawater environments [12,13].Besides identifying the new osmo-responsive genes and pathways in thestudy, we also identified several enhanced immune-related pathwayssuch as cytokine-related pathways such as IL-6, IL-8 and IL-9 signalingpathways under different osmotic environment [12,13]. In this study,we combined transcriptomics approach, RT-qPCR and im-munohistochemistry to declare the effect of environmental salinitychanges on immune-related genes involved in adaptive immune re-sponses in gill cells of Japanese eels. The results of this study provide animportant resource for future investigations on the mechanisms ofadaptive immune response of gill cells.

2. Materials and methods

2.1. Animals and isolation of gill cells

Japanese eels (Anguilla japonica) weighing 500–600 g were pur-chased and kept in a 40 L glass tanks supplied with charcoal-filteredaerated fresh water (FW) at 18–20 °C under a 12 L:12D photoperiod for3 weeks before the experiments. The fish were then kept in FW (n= 4)or acclimated to seawater (SW) (34 ppt) (n=4) for another 3 weeks.Ten liters of water was changed every 3 days. The experiment was re-peated at least three times. The fishes were anesthetized by 0.1%MS222 in a plastic container and the gills were perfused with bufferedsaline (130mM NaCl, 2.5mM KCl, 5 mM NaHCO3, 2.5 mM glucose,2 mM EDTA, and 10 mM Hepes, pH 7.7) to remove blood cells fromgills. Gill arches were excised and washed. The gill arches were cut intosmall fragments and subjected to two cycles of trypsin digestion (0.5%trypsin + 5.3 mM EDTA), each for 20 min at room temperature in arotator (300 rpm). The cell suspension was then filtered, washed, andunderwent a three-step Percoll gradient of 1.09, 1.06, and 1.03 g/ml inPBS and centrifuged at 2000 g at 15 °C for 20 min. Cells at the interfaceof 1.03 and 1.06 g/ml Percoll solution were regarded as pavement cells(PVCs), while at the interface of 1.06 and 1.09 g/ml were mitochondriarich cells (MRCs). The identity of MRCs was confirmed by mitochondriastaining (Mitotracker, CMTMRos-H2, Molecular Probes) and Na+/K+-ATPase staining (mouse anti-Na+/K+-ATPase α-subunit antiserum)(1:50, Developmental Studies Hybridoma Bank, the University of Iowa).

2.2. Library construction and illumina RNA-seq

The overall workflow described as before [12,13]. Briefly, total RNAextracted from FW PVCs, FW MRCs, SW PVCs and SW MRCs (eachn=2) by mirVanaTM miRNA isolation kit (Applied Biosystems). RNAquality with a RNA Integrity Number (RIN) > 8 (assessed using theAgilent 2100 Bioanalyzer system) were used for RNA library con-struction. The cDNA libraries were constructed according to manufac-turer's instruction and index codes were ligated as identification to in-dividual samples as previous described. Briefly, 8 cDNA libraries wereconstructed (2×4 groups), each prepared from 300 ng total RNA. Thenpurified mRNA from the total RNA using poly-T oligo-attached

magnetic beads (Illumina, San Diego, USA) to remove the ribosomalRNA. Then the mRNA was fragmented by divalent cations in Illuminaproprietary fragmentation buffer at 94 °C for 1min. First strand cDNAswere synthesized using random oligonucleotides and SuperScript II, andthe second cDNAs were synthesized using DNA polymerase I and RNaseH. Overhangs were blunted by using exonuclease/polymerase, followedby 3′ end adenylation. After adenylation, DNA fragments Illumina wereligated with PE adapter oligonucleotides. DNA fragments that ligatedwith adaptor molecules on both ends were selectively enriched by Il-lumina PCR Primer Cocktail in a 15 cycles PCR reaction. Libraries werepurified using AMPure XP system and quantified using the KAPA Li-brary Quantification Kits. Before start sequencing, the libraries werenormalized and pooled together in a two single lane on an IlluminaMiSeq platform and 150 bp paired-end reads were generated. Adaptersand reads containing poly-N were first trimmed and the sequence-readswere dynamically trimmed according to BWA's-q algorithm [14].Briefly, a running sum algorithm was executed in which a cumulativearea-plot is plotted from 3′-end to the 5′-end of the sequence reads andwhere positions with a base-calling Phred quality lower than 30 causean increase of the area and vice versa. Such plot was built for each readindividually and each read was trimmed from the 3′-end to the positionwhere the area was greatest. Read-pairs were then synchronized suchthat all read-pairs with sequence on both sides longer than 35 bp afterquality trimming were retained. Any singleton read resulting from readtrimming was removed [14]. All the downstream analyses were basedon quality trimmed reads.

2.3. De novo transcriptome assembly and annotation

Obtained reads (Forward and reverse) from all the libraries/sampleswere pooled and subjected to transcriptome de novo assembly usingTrinity (version r20140413p1) with “min_kmer_cov” set to 2;“SS_lib_type” set to RF, and all other parameters set to default [15].Trinity uses fixed kmer to generate an assembly and it is efficient inrecovering full-length transcripts as well as spliced isoforms. The openreading frames (ORF) were identified by Transdecoder [16] using thefollowing criteria: (1) the longest ORF was identified within eachtranscript; (2) from the longest ORFs extracted, a subset of the longestones was identified and randomized to provide a sequence compositioncorresponding to non-coding sequences before being used to para-meterize a Markov model based on hexamers; and (3) all the longestORFs were scored according to the Markov Model to identify thehighest scoring reading-frame out of the six possible reading-frames.These ORFs were then translated to protein sequences and subjected to(1) BLASTp search against UniProtKB/Swiss-Prot with a cut-off e-value[17,18] of 1.0× 10−6, (2) protein domain search via HMMScan, (3)transmembrane helicase prediction by TMHMM, and (4) signal peptideprediction by SignalP.

2.4. Differential expression, gene ontology (GO) and pathway enrichmentanalysis

Differential gene expression and TMM-normalized FPKM gene ex-pression were calculated by RSEM pipe-line with edgeR package [19].Samples from identical salinity condition were considered to be biolo-gical replicates. Genes with B&H corrected p-value< .05 and log2 (foldchange) > 1 were considered be statistically significant differentiallyexpressed (FW PVCs Vs SW PVCs and FW MRCs Vs SW MRCs). TheDatabase for Annotation, Visualization and Integrated Discovery(DAVID) was used for functional annotation clustering analysis with theclassification stringency as Benjamini-Hochberg corrected P-value(P < .05) [20]. The dysregulated transcripts with human homologs ofthe assembled contigs were identified by the IPA software to findfunctional canonical pathways and functions (www.qiagen.com/ingenuity) with the significance level set at Benjamini-Hochberg cor-rected P-value (P < .05) [20]. The dysregulated transcripts (FW PVCs

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

289

Vs SW PVCs MRCs Vs SWMRCs) involved in “immune response” relatedbiological process GO assignments were identified and classified.

2.5. Real-time qPCR analysis

To validate the sequencing data, RNA extract from the fish accli-mated in FW (n=4) and seawater (SW) (n=4) for RT-qPCR, and re-peated for three times. All the differentially expressed genes involved in“immune response” biological process GO assignments were detectedfor RT-qPCR analysis. The isolated enriched PVCs or MRCs were dis-solved in Tri-Reagent (Gibco-BRL) for total RNA extraction. Total RNAwith a ratio of 1.8–2.0 at A260/A280 was used for cDNA synthesis.Briefly, 0.5 μg total cellular RNA was reverse transcribed by VILO(Invitrogen). PCR was conducted using the Applied Biosystems 7500real-time PCR detection system using KAPA SYBR® Green Supermix(KAPA). Primers used in the real-time qPCR assay were designed on thebasis of transcriptome sequence through Primer design (NCBI).Sequences of the real-time qPCR primers and amplicon sizes were listedin Table 1. The data were then normalized using the expression levels ofgapdh mRNA since it stably expressed in gill cells under different os-motic stress (i.e. hypo- and hyper-) [12,13,21,22]. The existence ofprimer–dimers and secondary products was checked using meltingcurve analysis. Our data indicated that the amplification was specific.Only one PCR product was amplified for each individual primer set. Therelative expression ratio was calculated according to the method de-scribed by Pfaffl.

Expression ratio=EtargetCPtarget (FW–SW)/EgapdhCPgapdh (FW–SW), whereE= 10(−1/slope) and CP is the crossing point at which fluorescence risesabove the background level. Statistical significance of differential geneexpression between FW and SW was assessed with Student's t-test withp-value< .05. And we did a Pearson correlation analysis of the results(log2 fold change) of Illumina RNA-seq and RT-qPCR.

2.6. Immunohistochemistry (IHC)

Paraformaldehyde-fixed gill sections were dewaxed, rehydrated ingraded ethanol, and rinsed in PBS + 0.1% Tween 20 (PBST). The

staining procedure involved pretreatment of tissue sections with 3%normal goat serum in PBST to reduce non-specific staining, followed byan overnight incubation at 4 °C with rabbit anti-C Reactive Protein(Y284) (1:50, Abcam, Cat. No. 32412) rabbit anti-IL-1RII (H-71) (1:50,santacruz, Cat. No. sc-292522), rabbit anti-TLR2 (H-175) (1:50, santa-cruz, Cat. No. sc-10739) and mouse anti-Na+/K+-ATPase-subunits(Naα5, 1:50, Developmental Studies Hybridoma Bank, the University ofIowa) antibodies. Then the slides incubated with a cocktail of AlexaFluor 488 goat anti-rabbit IgG (1:200, Molecular Probes, Invitrogen)and Alexa Fluor 568 goat anti-mouse IgG (1:200, Molecular Probes,Invitrogen) for 1 h at room temperature. The tissues were stained by theDAPI mounting medium (Vector Labotorary) for several minutes, andwere examined by Olympus IX73 microscopy (Japan). The slides werewashed 3×15min in PBST after each antiserum application. Thecontrol procedure included the application of non-immune rabbitserum (Sigma). Image J software (National Institutes of Health, USA)was used to determine the fluorescence intensity signals and countpositively stained cells. Representative images collected under the samemicroscope settings were compiled without any modification, and ad-justed for brightness and contrast using Image J. Background of eachsummed image was subjected to threshold filtering, and thereby thepixels brighter than a given threshold background value were con-sidered as the signals attributable to the fluorescent proteins indicators.The positively stained cells were counted per fixed length (250 μm) of afilament and at least three filaments were selected. The relative fluor-escence intensity= total fluorescence intensity/cell numbers.

2.7. Statistical analysis

All treatments were performed in quadruplicate in each experimentand every experiment was repeated at least three times. All data arerepresented as means ± s.e.m. Statistical significance was assessedwith Student's t-test. Groups were considered significantly different ifP < .05.

Table 1List of sequences of primers for real-time PCR.

Gene Primers Amplicon Size (bp)

BLNK F: 5′-TACAACATCCCGATACGCCA-3′ R: 5′-TCGATCATACCGGAGACACT-3′ 101CRP F: 5′-ACTCGATATGCAACACCCAC-3′ R: 5′-AAGGATCTCAGGCAGAGTGT-3′ 71CD79A F: 5′-GACCAATTACACCACGACGA-3′ R: 5′-GAGCCCAGTGTCATTCATGT-3′ 139FYB F: 5′-CCCCAATTATCAGCAGGCTC-3′ R: 5′-TGGATCTCGCCATCGAATTT-3′ 133GCH1 F: 5′-AAGCTGCAGAAATGAACGGA-3′ R: 5′-CGAAGGATGCTGGTATAGGC-3′ 190C1QBP F: 5′-AATCTCTTGCCAGTGTGGTG-3′ R: 5′-GTAGACCCGCTATTTCGCAT-3′ 172CLNK F: 5′-GGGCAGAAAGATGGAACACA-3′ R: 5′-AGGCTGCATCAGACATCAAG-3′ 134DMBT1 F: 5′-GTAGAGGGCAACATCACAGG-3′ R: 5′-CGTACACGGGGTAGTTGAAG-3′ 85EOMES F: 5′-GCCTATCAACTTGCCGAAGA-3′ R: 5′-TCTCACCTTCTCCCTCGTTT-3′ 182IL12B F: 5′-TTACTCAGTGCCGTGCTTTT-3′ R: 5′-TATGTCCATGCCCTCGTACT-3′ 153IL8 F: 5′-ATGCGCACATCAATCTTCCT-3′ R: 5′-TTAGCCTGCGTTCATGGTTT-3′ 118LAT2 F: 5′-GTCCGACTCCAACAACGG-3′ R: 5′-CTATAATGCCACAGGCGCTC-3′ 83LY9 F: 5′-ATCTCCGTACAACAGTCCCA-3′ R: 5′-ACCTGTCCCCAAATCTCTCA-3′ 113MYO1F F: 5′-TATCAGTATGTGGGGCAGGA-3′ R: 5′-CGGAAGCGTCTTCACTGATA-3′ 82PRDX1 F: 5′-GTTCAGGAAGATCGACTGCG-3′ R: 5′-GTAGGCCTTGGAGATCGAGT-3′ 154LYN F: 5′-TGCACACGATCGGAAGAAAA-3′ R: 5′-ATGTGGTTTACTGGTGGTGG-3′ 82CD276 F: 5′-CATTGCGGTAGTCTTCGTCA-3′ R: 5′-CCGTCGTATCGTTCAAACCT-3′ 149CNPY3 F: 5′-CATGTTCTGAAAGGCCAGGA-3′ R: 5′-AGTGGGGCTTTCTTCTGAAC-3′ 293CCR9 F: 5′-GGCCTATTAGCTCTCCCAGA-3′ R: 5′-AGGATCTTTGTGCGGTTGTT-3′ 104CCL8 F: 5′-GGCCTATTAGCTCTCCCAGA-3′ R: 5′-AGGATCTTTGTGCGGTTGTT-3′ 104GBP1 F: 5′-ATAGAAAGGGCTCGCAAGGA-3′ R: 5′-AATTCATCCTCGGTTTCCGT-3′ 100IL1R2 F: 5′-GAACAGGAGCTTCGTAGAGC-3′ R: 5′-CTGGAAAACCAGGTCCCTTC-3′ 112LYST F: 5′-TCTACCTGTACCTGCTCCTG-3′ R: 5′-TCCGCCTCGTATTTCAACTG-3′ 181PRKDC F: 5′-CCTTGTTGCTGGATGCCTAA-3′ R: 5′-AGGCAACTGCTAAACTGACC-3′ 204SBNO2 F: 5′-AGGGGACCCCCTTCAAAATC-3′ R: 5′-AGCTTTGTTGTAGACGAGCTT-3′ 197TLR2 F: 5′-TGAAGACACCCTGGGATTCA-3′ R: 5′-GATAAACTGGTGAGGCGGTT-3′ 195ZEB1 F: 5′-GAGCACAGAGACACTAACGG-3′ R: 5′-TCTGTGTCCTTCTCGGACTT-3′ 138GAPDH F: 5′-GCGCCAGCCAGAACATCATC-3′ R: 5′-GTTAAGCTCGGGGATGACC-3′ 74

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

290

2.8. Availability of supporting data

The sequencing data from this study have been submitted to theNCBI Sequence Read Archive (SRA) (https://www.ncbi.nlm.nih.gov/sra) under the accession number SRP049703.

3. Results

3.1. Differentially expressed genes (DEGs) in PVCs and MRCs in FW andSW conditions

From our obtained transcriptome profiles of gill cells of Japanese eeladapted in freshwater and seawater environments [12,13] and bioin-formatics analysis of differentially expressed genes (p < .05) in pave-ment cells (PVCs) and mitochondria rich cells (MRCs) in FW and SWconditions using DAVID, we found a group of differentially expressedgenes (p < .05) involved in “immune” related biological processes(Fig. 1A and B). The top three enriched biological process (BP) terms inPVCs are “immune response”, “immune system development”, and“positive regulation of immune system process (Fig. 1A); while inMRCs, “immune response” and “macrophage activation during immuneresponse”, were identified (Fig. 1B). It should be noted that these 20“immune response” related genes in PVCs and the 11 immune relatedgenes in MRCs were expressed significantly higher in FW than in SW(p < .05). Among these differentially expressed genes, there are 4genes both expressed in PVCs and MRCs, including Interferon-inducedguanylate-binding protein 1 (GBP1), C-C motif chemokine 8 (CCL8), C-

C chemokine receptor type 9 (CCR9), and Interleukin-1 receptor type 2(IL1R2) (Fig. 2).

3.2. Validation of differentially expressed immune-related genes by RT-qPCR

To validate additional mRNA transcript levels of RNA-seq result, weperformed RT-qPCR on these 20 immune-related genes in PVCs and 11immune-related genes in MRCs in fish adapted in FW and SW condi-tions. The results of the RT-qPCR analysis matched with the IlluminaRNA-seq data (Tables 2 and 3). It showed these immune response geneswere suppressed in SW compared with in FW. And a comparison of theexpression change of these immune response genes from RNA-seqanalysis and RT-qPCR showed a statistically significant correlation(R2= 0.70922 in FW PVCs vs SW PVCs and R2=0.70481 in FW MRCsvs SW MRCs respectively) (Fig. 3A and B). It suggested that our Illu-mina RNA-seq data was reliable, although there were some deviationsbetween Illumina RNA-seq and RT-PCR (the difference of log 2 foldchange of these two test > 1), such as GBP1, FYB in PVCs and CNPY3in MRCs. Two potential sources of error that could explain the largedeviations were the intra-experimental variations (i.e., enzyme activity)and technical variability (pipetting). However, the trend was consistentand the results of these two methods were highly correlated.

3.3. Protein expression of CRP, TLR2 and IL-1R2

According to the mRNA expression detected by real-time qPCR, C-reactive protein (CRP) was the topmost significantly differentially ex-pressed genes between FW and SW in PVCs, Toll-like receptor 2 (TLR2)was the topmost significantly differentially expressed genes betweenFW and SW in MRCs and Interleukin-1 receptor type 2 (IL-1R2) was thetopmost significantly differentially expressed genes between FW andSW in both PVCs and MRCs. In addition, CRP, TLR2 and IL-1R2 areusually used for investigating immunocytokines in fish. Therefore, thesethree cytokines were selected as representatives in PVCs and/or MRCsfor protein level expressions detection in gill section by IHC (Fig. 4).The IHC results showed that all these proteins expressed in both PVCsand MRCs, and all of them co-localized with Na+/K+-ATPase-subunits(Naα5) in MRCs, which we used Naα5 as the MRCs specially expressedmarker to labeled them. These results matched with our previous Illu-mina RNA-seq data in Japanese eel gills that CRP, TLR2 and IL-1R2were detected in both PVCs and MRCs [21]. Fig. 4A showed a sig-nificantly higher expression of CRP protein levels in the FW PVCs thanthe SW PVCs; while there was no difference in MRCs between FW andSW condition (Fig. 4A). In contrast, significantly higher expression ofTLR2 protein levels were detected in the FW MRCs compared with inSW MRCs, but there was no significant difference of TLR2 expression inPVCs between FW and SW (Fig. 4B). Lastly, IL-1R2 protein had a

Fig. 1. “Immune response” Gene ontology of DEGs identified FW PVCs vs SW PVCsand FW MRCs vs SW MRCs. (A) Under the biological process category, 20 differentiallyexpressed genes (DEGs) in PVC cells involved in “immune response”, followed by 9 DEGsinvolved “immune system development”, and 8 DEGs were involved “positive regulationof immune system process” in the GO assignments. (B) 11 differentially expressed genes(DEGs) in MRC cells involved in “immune response” and 2 DEGs were involved “mac-rophage activation during immune response” in the GO assignments. The p-value wasgenerated by Student's t-test and further corrected by Benjamini correction.

Fig. 2. Comparison of DEGs in PVCs and MRCs in eel gills. The 20 “immune response”related genes in PVCs and the 11 “immune response” related genes in MRCs were ex-pressed significantly higher in FW than in SW (p < .05). Among these differentiallyexpressed genes, there are 4 genes both expressed in PVCs and MRCs, including GBP1,CCL8, CCR9, and IL1R2.

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

291

significantly higher expression in both the PVCs and MRCs in eelsadapted in FW than SW (Fig. 4C). All the results of these proteins ex-pression in gill sections matched with Illumina RNA-seq and RT-qPCRresults.

4. Discussion

The immune response is critical to all species and the adaptiveimmune system has conserved some important features over millions ofyears of evolution [23,24]. It is suggested that the immune-relatedgenes play an important role for the plasticity of the adaptive immunesystem and in animal diversification in teleost fishes [25]. Previousstudies on fish or shellfish immune responses mainly focused on thehead kidney, thymus and spleen [26–29]. Recently, some inflammatorycytokines, such as interleukins (ILs) and tumor necrosis factor α (TNFα)were detected in gills, indicating the possibilities of their roles in im-mune responses [30]. However, the relationship between immune-re-lated genes and salinity changes in fish gill is not yet known. Presentstudies combined transcriptomics approach, RT-qPCR and im-munohistochemistry to declare high environmental salinity conditiondown-regulated immune related genes expression in gill cells of Japa-nese eels. However, it should be noted that since the bioinformaticsanalysis is based on mammalian and human database, our annotationmethod by conserved genes may result in the loss of non-conservedregions in fish genes, which is a limitation on the fishes' studies.

Gill epithelial cells are directly facing the external environment.Response to salinity changes, the endocrine osmoregulatory hormonessuch as growth hormone (GH), cortisol and insulin-like growth factor I(IGF-I) targeted on gill cells will modulate the mRNA expressions of iontransporters and osmotic stress respond genes [31]. Interestingly, GHand IGF-I were found to stimulate the mRNA expression of lysozyme (animmune-related gene) in gill filament of Atlantic salmon (Salmo salar)under high salinity condition [32]. In addition, cytokines such as ILsand TNFα were involved in the osmoregulatory signaling network ingill cells of leopard sharks (Triakis semifasciata) and tilapia (Oreochromismossambicus) [33–35]. Thus it is reasonable to assume that there shouldbe a group of immune-related genes and related signaling networks ingill cells, which can directly responds to environmental salinitychanges, and some genes that could be regulated by endocrine osmor-egulatory hormones to respond to the salinity changes. Our previousproteomics studies annotated the acute responsive proteins during os-motic stress, including transfer FW to SW and SW to FW [21,22]. Short-term hyperosmotic challenge (FW to SW, 6h) induced 10 up-regulated(e.g. alox5, trim16, ckb) and 9 down-regulated proteins (e.g. samhd1,mr1, astl). On the other hand, the acute hypo-osmotic stress (SW to FW,6h) induced 15 proteins up-regulated (e.g. aprt, vwa5a, lgals1) and 9down-regulated proteins (e.g. cldn7, hal, scel). Among those differentialexpressed proteins, eight proteins (e.g. nt5c1a, serpinb1) were responseto both hyper- and hypo-osmotic stress [21,22]. These results suggestedthat the response of short-term transfer (6h) FW to SW is different from

Table 2Relative mRNA expressions of immune response genes for comparison of the SW PVCs versus FW PVCs, in respect to Illumina RNA-Seq and real-time qPCR.

Gene symbol Gene Name Transcript ID Illumina RNA-seq (log2 fold change) Real-time qPCR (log2 fold change)

DMBT1 Deleted in malignant brain tumors 1 protein c38209_g2 −6.66 −5.42 ± 0.85*CRP C-reactive protein c88246_g2 −6.12 −5.69 ± 0.28*GBP1 Interferon-induced guanylate-binding protein 1 c94302_g3 −5.46 −4.00 ± 0.68*GCH1 GTP cyclohydrolase 1 c80288_g1 −5.37 −4.69 ± 1.64*LAT2 Large neutral amino acids transporter small subunit 2 c95958_g3 −5.23 −4.24 ± 0.34*LY9 T-lymphocyte surface antigen Ly-9 c89727_g2 −4.92 −4.43 ± 0.52*CCL8 C-C motif chemokine 8 c91708_g1 −4.90 −5.52 ± 0.28*EOMES Eomesodermin c85649_g1 −4.65 −5.06 ± 0.51*CLNK Cytokine-dependent hematopoietic cell linker c82390_g1 −3.85 −4.32 ± 0.50*FYB FYN-binding protein c68403_g1 −3.83 −4.95 ± 1.13*IL12B Interleukin-12 subunit beta c78952_g2 −3.43 −4.19 ± 0.96*CCR9 C-C chemokine receptor type 9 c27146_g1 −3.19 −3.54 ± 0.46*IL1R2 Interleukin-1 receptor type 2 c87055_g1 −2.78 −2.70 ± 0.46*CD79A B-cell antigen receptor complex-associated protein alpha chain c86012_g1 −2.42 −3.60 ± 0.76*LYN Tyrosine-protein kinase Lyn c69819_g1 −2.36 −3.06 ± 0.34*IL8 Interleukin-8 c76038_g1 −2.17 −3.95 ± 0.78*C1QBP Complement component 1 Q subcomponent-binding protein c82445_g1 −1.96 −1.52 ± 0.29*MYO1F Unconventional myosin-If c91930_g1 −1.86 −1.94 ± 0.10*BLNK B-cell linker protein c79280_g1 −1.82 −2.46 ± 0.26*PRDX1 Peroxiredoxin-1 c82360_g1 −1.66 −2.54 ± 0.35*

Note: The expression ratio calculated as described in materials and methods. The experiments were repeated for three times. Values were presented as the mean ± S.E.M. (n = 4 fish pergroup). *Statistical significance between FW PVCs and SW PVCs, *P < .05, using Student's t-test.

Table 3Relative mRNA expression of immune response genes for comparison of the SW MRCs versus FW MRCs, in respect to RNA-Seq and real-time PCR.

Gene symbol Gene Name Transcript ID Illumina RNA-seq (log2 fold change) Real-time PCR (log2 fold change)

CNPY3 Protein canopy homolog 3 c77950_g1 −6.45 −3.11 ± 0.70*TLR2 Toll-like receptor 2 c94925_g1 −4.57 −5.78 ± 0.21*GBP1 Interferon-induced guanylate-binding protein 1 c94302_g3 −4.39 −4.05 ± 1.07*CCL8 C-C motif chemokine 8 c91708_g1 −4.10 −4.62 ± 0.04*CCR9 C-C chemokine receptor type 9 c27146_g1 −3.48 −4.02 ± 0.43*IL1R2 Interleukin-1 receptor type 2 c87055_g1 −3.48 −4.81 ± 0.11*ZEB1 Zinc finger E-box-binding homeobox 1 c93234_g4 −3.40 −4.24 ± 0.71*LYST Lysosomal-trafficking regulator c96814_g1 −2.28 −4.07 ± 0.75*SBNO2 Protein strawberry notch homolog 2 c98802_g1 −2.00 −3.84 ± 0.48*CD276 CD276 antigen c100067_g1 −1.73 −3.11 ± 0.70*PRKDC DNA-dependent protein kinase catalytic subunit c100046_g1 −1.68 −2.59 ± 0.71*

Note: The expression ratio calculated as described in materials and methods. The experiments were repeated for three times. Values were presented as the mean ± S.E.M. (n = 4 fish pergroup). *Statistical significance between FW MRCs and SW MRCs, *P < .05, using Student's t-test.

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

292

SW to FW. The present study aimed to understand the response ofimmune-related genes of FW and SW acclimated eel (3 weeks) ac-cording to our published transcriptome profile of gill cells [12,13].

Like TNFα and the interleukins, C-reactive protein (CRP), TLR, IL-1receptor type 1 (IL-1R1), IL-1 receptor type 2 (IL-1R2) and IL-1 receptorantagonist (IL-1RA) are the most investigated immunocytokines to de-termine the immunity in colorectal cancer patients or fishes[36–40,52]. CRP and TLR mediated the activation of innate andadaptive immunity [41–43]. Intracellular CRP-positive reaction hadbeen detected in rainbow trout hepatocytes, head kidney macrophages,spleen lymphocytes and peripheral blood lymphocytes (PBL), aftermaking the cells permeable with methanol [44]. Intracellular CRP ischaracterized as a Ca2+ binding protein and perturbation of cellularcalcium blocked the secretion of CRP from the rough endoplasmic re-ticulum (ER) [45,46]. Under high salinity condition, the fish reduceCa2+ uptake/influx through the gill epithelia cells to maintain thewhole body Ca2+ content [47], which might result in lower expressionof CRP in gill PVCs of eels adapted in SW compared with FW. CRP

interacts with Fcγ receptor (FcγR) to induce the production of TNFα, IL-1β, and IL-1RA in human blood monocyte [41,48,49]. CRP interactswith FcγRI to stimulate the production of anti-inflammatory cytokineIL-10, a potent inhibitor of IL-12 production. Moreover, binding of CRPto inhibitory receptor FcγRII suppresses the IL-12 response to lipopo-lysaccharide (LPS), which activates toll-like receptor (TLR) to produceIL-12 [50]. TLR is structurally similar to IL-1R1, and induces immuneresponses during inflammation and host defense [37,51]. In kidney ofstriped catfish (Pangasianodon hypophthalmus, Sauvage), chronic hy-perosmotic stress inhibited TLRs expression, and the down-regulationof TLRs were aggravated when exposed to bacterial infection [52]. Inmacrophages and dendritic cells, activations of intracellular TLR sig-naling pathways depend on trafficking and processing of intracellularTLRs, which requires intact ER Ca2+ stores [53]. The ER Ca2+ storesreuptake the overload Ca2+ to maintain intracellular Ca2+ home-ostasis. When the extracellular Ca2+ concentration changes, such asfrom FW (0.2mM) to SW (10mM), it might alter the TLRs expressionand their relative signaling pathways [54]. In macrophages, TLR alsoshowed to mediate the production of IL-1RA [55–57]. Although ourIllumina RNA-seq data in Japanese eel did not detect the expression ofIL-1RA in gill cells; CRP, TLR2, IL-1R1 and IL-1R2 were all detected inboth gill PVCs and MRCs [21]. Blocking the IL-1/IL-1R1 interaction byIL-1R2 in neutrophils and monocytes inhibited the IL-1-induced IL-6,IL-8 and TNF-α productions [40,58], which might explain the low ex-pression of IL-6, IL-8 and enriched signaling pathways in PVCs of eelsadapted in SW compared with FW [12,13]. In this study, the resultsshowed expression of CRP in gill PVCs, IL-1R2 in both PVCs and MRCs,and TLR2 in MRCs were significantly down-regulated in higher salinitycondition (SW). In lower salinity condition (FW), CRP might regulateIL-1RA expression in PVCs, which was also regulated by CRP-modu-lated TLR2 signal in MRCs (Fig. 5). In addition, IL-1RA and IL-1R2 inboth PVCs and MRCs could inhibit IL-1/IL-1R1 signal (Fig. 5). Cer-tainly, the existence of IL-1RA or IL-1RA homologous and the indeedexisting network of CRP, IL1-R1, IL1-R2 and TLR2 in gill cells, andinteraction between PVCs and MRC under different salinity conditionstill need further study.

Previously, TNFα and IL-8 had been reported as hyperosmolar stressinducible genes in human epithelial cells [59–61]. In fish, studies byKültz et al. suggested that these cytokines were involved in the osmotic-stress signaling network in regulate epithelial responses to salinitychanges [33–35]. In addition, we also reported that IL-6, IL-8 signalingpathway and IL-9 signaling pathway were identified to respond to theosmotic stress in eel gill PVCs and MRCs respectively [12,13]. In thisstudy, we further confirm that the expression of IL-8 was higher in PVCsof eels adapted in FW compared with SW using RT-qPCR. It indicatesthat IL-8 was down-regulated by higher salinity condition in eel gills,and suggest that an osmosensor might specifically exist in gill epithelialcells.

There are limited studies on the action of these immune-relatedgenes in fish that related to environmental salinity changes. The nextgeneration sequencing method offers a new platform of comparativetranscriptome, and it has been applied to reveal immune-related genesin shellfish gills response to heavy metal (copper) exposure [62] and infish gills for rapid evolution under environment selection [63]. Here,we further analyzed our transcriptome data of gill cells [12,13] andclustered a group of immune-related genes that respond to environ-mental salinity changes. The results indicated that high environmentalsalinity significantly reduced expression of the immune-related genes,including CRP, TLR2 and IL1-R2 in the gill cells of Japanese eels.Furthermore, transcriptions of the above-mentioned genes and theirprotein levels in the gill cells were suppressed in SW-acclimated fishcompared with FW-acclimated fish. Different from humoral immuneresponses (e.g. increased lysozyme activity and complement activity),these immune-related genes mediated cellular adaptive immune re-sponses in gill cells which were found to be suppressed when fisheswere acclimated to high environmental salinity [11,52]. The results of

Fig. 3. Pearson correlation analysis between Illumina RNA-seq and RT-qPCR log2fold change for the DEGs identified in FW PVCs vs SW PVCs and FW MRCs vs SWMRCs. Pearson correlation analysis of the results (log2 fold change) of Illumina RNA-seqand RT-qPCR of (A) the 20 differentially expressed “immune response” related genes inPVCs; and (B) The 11 differentially expressed “immune response” related genes in MRCs.The R2 values indicate the correlation coefficient of the results of these two methods.

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

293

this study show direct relationship between immune-related genes andenvironmental salinity in fish gill cells and could help further under-standing in the functions of immune-related genes in adaptive immuneresponse to ambient salinity changes.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgment

The work in Gu's laboratory is supported by the Start-Up Research

Fig. 4. IHC staining of CRP, TLR2 and IL-1R2 in eel gills. Expressions of CRP (A), TLR2 (B) and IL-1R2 (C) in the gills of Japanese eels adapted in different water. The cell nucleusappears blue, CRP, TLR2 and IL-1R2 were expressed in both PVCs and MRCs and appeared green; Na+/K+-ATPase (Naα5) used as a marker to label MRCs and appears red, co-localizations of CRP, TLR2 or IL-1R2 with Na+/K+-ATPase appears yellow. And higher CRP expression in PVCs, higher TLR2 expression in MRCs, and higher IL-1R2 expression in bothPVCs and MRCs of FW adapted eels were detected. The relative fluorescence intensity in PVCs and MRCs was calculated as described in materials and methods. The experiments wererepeated for three times. Values were presented as the mean ± S.E.M. (n = 4 fish per group). *Statistical significance between FW and SW, *P < .05, using Student's t-test. Scalarbar= 50 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

294

Funding of Jiangsu University for Distinguished Scholars (5501330001)and the National Natural Science Foundation of China (31271272 and31600952).

References

[1] R. Laizcarrion, J.M. Guzman, R. Mpmdel, J.L. Soengas, J.M. Mancera, Growthperformance of gilthead sea bream Sparus aurata in different osmotic conditions:implications for osmoregulation and energy metabolism, Aquaculture 250 (2005)849–861.

[2] J. Maetz, Aspects of adaptation to hypo-osmotic and hyper-osmotic environments,in: D.C. Malins, J.R. Sargent (Eds.), Biochemical and Biophysical Perspectives inMarine Biology, Academic Press, New York, US, 1974, pp. 1–167.

[3] M. SD, Endocrine control of osmoregulation in teleost fish, Am. Zool. 41 (2001)781–794.

[4] A. Kumagai, R. Ando, H. Miyatake, P. Greimel, T. Kobayashi, Y. Hirabayashi,T. Shimogori, A. Miyawaki, A bilirubin-inducible fluorescent protein from eelmuscle, Cell 153 (2013) 1602–1611.

[5] W.N. Tzeng, Y. Iizuka, J.C. Shiao, Y. Yamada, H.P. Oka, Identification and growthrates comparison of divergent migratory contingents of Japanese eel (Anguilla ja-ponica), Aquaculture 216 (2003) 77–86.

[6] A. Cuesta, R. Laiz-Carrión, M.P. Del Río, J. Meseguer, J.M. Mancera, M.A. Esteban,Salinity influences the humoral immune parameters of gilthead seabream (Sparusaurata L.), Fish Shellfish Immunol. 18 (2005) 255–261.

[7] L. Tort, Stress and immune modulation in fish, Dev. Comp. Immunol. 35 (2011)1366–1375.

[8] A.M. Marc, C. Quentel, A. Severe, P.Y.L. Bail, G. Boeuf, Changes in some en-docrinological and non-specific immunological parameters during seawater ex-posure in the brown trout, J. Fish. Biol. 46 (2010) 1065–1081.

[9] I.F. Jiang, V.B. Kumar, D.N. Lee, C.F. Weng, Acute osmotic stress affects Tilapia(Oreochromis mossambicus) innate immune responses, Fish Shellfish Immunol. 25(2008) 841.

[10] J.F. Taylor, M.P. Needham, B.P. North, A. Morgan, K. Thompson, H. Migaud, Theinfluence of ploidy on saltwater adaptation, acute stress response and immunefunction following seawater transfer in non-smolting rainbow trout, Gen. Comp.Endocrinol. 152 (2007) 314.

[11] S.C. Birrer, T.B.H. Reusch, O. Roth, Salinity change impairs pipefish immune de-fence, Fish Shellfish Immunol. 33 (2012) 1238–1248.

[12] K.P. Lai, J.W. Li, J. Gu, T.F. Chan, W.K.F. Tse, C.K.C. Wong, Transcriptomic analysisreveals specific osmoregulatory adaptive responses in gill mitochondria-rich cellsand pavement cells of the Japanese eel, BMC Genom. 16 (2015) 1072.

[13] W.K. Tse, J. Sun, H. Zhang, K.P. Lai, J. Gu, A.Y. Sheung Law, B.H. Yee Yeung,S. Ching Chow, J.W. Qiu, C.K. Wong, Data for transcriptomic and iTRAQ proteomicanalysis of Anguilla japonica gills in response to osmotic stress, Data Brief 3 (2015)120–125.

[14] J.W. Li, K.P. Lai, A.K. Ching, T.F. Chan, Transcriptome sequencing of Chinese andCaucasian population identifies ethnic-associated differential transcript abundanceof heterogeneous nuclear ribonucleoprotein K (hnRNPK), Genomics 103 (2014) 56.

[15] M.G. Grabherr, B.J. Haas, M. Yassour, J.Z. Levin, D.A. Thompson, I. Amit,X. Adiconis, L. Fan, R. Raychowdhury, Q. Zeng, Full-length transcriptome assemblyfrom RNA-Seq data without a reference genome, Nat. Biotechnol. 29 (2011) 644.

[16] B.J. Haas, A. Papanicolaou, M. Yassour, M. Grabherr, P.D. Blood, J. Bowden,M.B. Couger, D. Eccles, B. Li, M. Lieber, De novo transcript sequence reconstructionfrom RNA-seq using the Trinity platform for reference generation and analysis, Nat.Protoc. 8 (2013) 1494.

[17] B. Ewen-Campen, N. Shaner, K.A. Panfilio, Y. Suzuki, S. Roth, C.G. Extavour, Thematernal and early embryonic transcriptome of the milkweed bug Oncopeltus fas-ciatus, BMC Genom. 12 (2011) 61.

[18] H. Du, Z. Bao, R. Hou, S. Wang, H. Su, J. Yan, M. Tian, Y. Li, W. Wei, W. Lu,Transcriptome sequencing and characterization for the sea cucumber Apostichopusjaponicus (Selenka, 1867), PLoS One 7 (2012) e33311.

[19] B. Li, C.N. Dewey, RSEM: accurate transcript quantification from RNA-Seq datawith or without a reference genome, BMC Bioinf. 12 (2011) 323.

[20] D.W. Huang, B.T. Sherman, R.A. Lempicki, Systematic and integrative analysis oflarge gene lists using DAVID bioinformatics resources, Nat. Protoc. 4 (2009) 44.

[21] W.K. Tse, J. Sun, H. Zhang, A.Y. Law, B.H. Yeung, S.C. Chow, J.W. Qiu, C.K. Wong,Transcriptomic and iTRAQ proteomic approaches reveal novel short-term hyper-osmotic stress responsive proteins in the gill of the Japanese eel (Anguilla japonica),J. Proteonomics 89 (2013) 81–94.

[22] W.K. Tse, J. Sun, H. Zhang, K.P. Lai, J. Gu, J.W. Qiu, C.K. Wong, iTRAQ-basedquantitative proteomic analysis reveals acute hypo-osmotic responsive proteins inthe gills of the Japanese eel (Anguilla japonica), J. Proteonomics 105 (2014) 133.

[23] M.F. Flajnik, M. Kasahara, Origin and evolution of the adaptive immune system:genetic events and selective pressures, Nat. Rev. Genet. 11 (2010) 47.

[24] G.W. Litman, J.P. Rast, S.D. Fugmann, The origins of vertebrate adaptive immunity,Nat. Rev. Immunol. 10 (2010) 543.

[25] M. Malmstrøm, M. Matschiner, O.K. Tørresen, B. Star, L.G. Snipen, T.F. Hansen,H.T. Baalsrud, A.J. Nederbragt, R. Hanel, W. Salzburger, Evolution of the immunesystem influences speciation rates in teleost fishes, Nat. Genet. 48 (2016) 1204.

[26] B.H. Dannevig, A. Lauv, Receptor-mediated endocytosis and phagocytosis byrainbow trout head kidney sinusoidal cells, Fish Shellfish Immunol. 4 (1994) 3–18.

[27] W.D. Jiang, K. Hu, Y. Liu, J. Jiang, P. Wu, J. Zhao, Y.A. Zhang, X.Q. Zhou, L. Feng,Dietary myo -inositol modulates immunity through antioxidant activity and theNrf2 and E2F4/cyclin signalling factors in the head kidney and spleen followinginfection of juvenile fish with Aeromonas hydrophila, Fish Shellfish Immunol. 49(2015) 374–386.

[28] P.R. Rauta, B. Nayak, S. Das, Immune system and immune responses in fish andtheir role in comparative immunity study: a model for higher organisms, Immunol.Lett. 148 (2012) 23.

[29] T.J. Bowden, P. Cook, J.H. Rombout, Development and function of the thymus inteleosts, Fish Shellfish Immunol. 19 (2005) 413–427.

[30] S. Reyes-Cerpa, K. Maisey, F. Reyes-Lpez, D. Toro-Ascuy, A. Mara, M. Imarai, Fishcytokines and immune response, in: H. Türker (Ed.), New Advances andContributions to Fish Biology, InTech Press, Rijeka, Croatia, 2012, pp. 3–57.

[31] D.H. Evans, P.M. Piermarini, K.P. Choe, The multifunctional fish gill: dominant siteof gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenouswaste, Physiol. Rev. 85 (2005) 97–177.

[32] T. Yada, S.D. Mccormick, S. Hyodo, Effects of environmental salinity, biopsy, andGH and IGF-I administration on the expression of immune and osmoregulatorygenes in the gills of Atlantic salmon (Salmo salar), Aquaculture 362–363 (2012)177–183.

[33] W.W. Dowd, B.N. Harris, J.J. Cech, D. Kültz, Proteomic and physiological responsesof leopard sharks (Triakis semifasciata) to salinity change, J. Exp. Biol. 213 (2010)210.

[34] D.F. Fiol, S.Y. Chan, D. Kültz, Identification and pathway analysis of immediatehyperosmotic stress responsive molecular mechanisms in tilapia (Oreochromismossambicus) gill, Comp. Biochem. Physiol. Genom. Proteonomics 1 (2006)344–356.

[35] D. Kültz, The combinatorial nature of osmosensing in fishes, Physiology 27 (2012)259.

[36] H. Bessler, H. Salman, M. Bergman, Y. Alcalay, M. Djaldetti, In vitro effect of ly-copene on cytokine production by human peripheral blood mononuclear cells,Immunol. Invest. 37 (2008) 183.

[37] A. Dunne, L.A. O'Neill, The interleukin-1 receptor/Toll-like receptor superfamily:signal transduction during inflammation and host defense, Sci. Stke SingalTransduct. Knowl. Environ. 2003 (171) (2003) re3.

[38] J. Kamińska, M.M. Kowalska, M.P. Nowacki, M.G. Chwaliński, A. Rysińska,M. Fuksiewicz, CRP, TNF-alpha, IL-1ra, IL-6, IL-8 and IL-10 in blood serum ofcolorectal cancer patients, Pathol. Oncol. Res. 6 (2000) 38–41.

[39] A.G. Lacour, A. Gervaix, S.A. Zamora, L. Vadas, P.R. Lombard, J.M. Dayer, S. Suter,Procalcitonin, IL-6, IL-8, IL-1 receptor antagonist and C-reactive protein as identi-ficators of serious bacterial infections in children with fever without localisingsigns, Eur. J. Pediatr. 160 (2001) 95–100.

[40] V.A. Peters, J.J. Joesting, G.G. Freund, IL-1 receptor 2 (IL-1R2) and its role in im-mune regulation, Brain Behav. Immun. 32 (2013) 1–8.

[41] C.T. Du, C. Mold, C-reactive protein: an activator of innate immunity and a mod-ulator of adaptive immunity, Immunol. Res. 30 (2004) 261.

[42] I. Bekeredjianding, S. Inamura, T. Giese, H. Moll, S. Endres, A. Sing, U. Zähringer,G. Hartmann, Staphylococcus aureus protein A triggers T cell-independent B cellproliferation by sensitizing B cells for TLR2 ligands, J. Immunol. 178 (2007)2803–2812.

[43] F. Re, J.L. Strominger, Heterogeneity of TLR-induced responses in dendritic cells:from innate to adaptive immunity, Immunobiology 209 (2004) 191.

[44] Y. Liu, T. Iwasaki, S. Watarai, H. Kodama, Effect of turpentine oil on C-reactiveprotein (CRP) production in rainbow trout (Oncorhynchus mykiss), Fish ShellfishImmunol. 17 (2004) 203–210.

[45] H.F. Lodish, N. Kong, Perturbation of cellular calcium blocks exit of secretoryproteins from the rough endoplasmic reticulum, J. Biol. Chem. 265 (1990)10893–10899.

[46] T. Christopeit, T. Gossas, U.H. Danielson, Characterization of Ca2+ and phos-phocholine interactions with C-reactive protein using a surface plasmon resonance

Fig. 5. A schematic to illustrate immune-related genes CRP, TLR2 and IL-1R2 in gillPVCs and MRCs in response to salinity changes. In lower salinity condition (FW),higher expression of CRP was detected in gill PVCs. CRP might regulate IL-1RA expressionin PVCs and modulated TLR2 signal in MRCs. Higher expression of IL-1R2 in both PVCsand MRCs in lower salinity condition, which together with IL-1RA could inhibit IL-1/IL-1R1 signal.

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

295

biosensor, Anal. Biochem. 391 (2009) 39–44.[47] J.G. Hoenderop, B. Nilius, R.J. Bindels, Calcium absorption across epithelia,

Physiol. Rev. 85 (2005) 373–422.[48] C.A. Pue, R.F. Mortensen, C.B. Marsh, H.A. Pope, M.D. Wewers, Acute phase levels

of C-reactive protein enhance IL-1 beta and IL-1ra production by human bloodmonocytes but inhibit IL-1 beta and IL-1ra production by alveolar macrophages, J.Immunol. 156 (1996) 1594–1600.

[49] H. Tilg, E. Vannier, G. Vachino, C.A. Dinarello, J.W. Mier, Antiinflammatoryproperties of hepatic acute phase proteins: preferential induction of interleukin 1(IL-1) receptor antagonist over IL-1 beta synthesis by human peripheral bloodmononuclear cells, J. Exp. Med. 178 (1993) 1629–1636.

[50] T.W.D. Clos, C. Mold, C-reactive protein, Immunol. Res. 30 (2004) 261–277.[51] Y. Xu, X. Tao, B. Shen, T. Horng, R. Medzhitov, J.L. Manley, L. Tong, Structural

basis for signal transduction by the Toll/interleukin-1 receptor domains, Nature 408(2000) 111–115.

[52] M. Schmitz, S. Baekelandt, S. Bequet, P. Kestemont, Chronic hyperosmotic stressinhibits renal Toll-Like Receptors expression in striped catfish (Pangasianodonhypophthalmus, Sauvage) exposed or not to bacterial infection, Dev. Comp.Immunol. 73 (2017) 139–143.

[53] A.L. Blasius, B. Beutler, Intracellular toll-like receptors, Immunity 32 (2010) 305.[54] H. Plattner, A. Verkhratsky, Inseparable tandem: evolution chooses ATP and Ca2+

to control life, death and cellular signalling, Phil. Trans. Roy. Soc. Lond. 371 (2016)20150419.

[55] C. TO, M.C. Cleophas, M. Oosting, H. Lemmers, T. H, M.G. Netea, T.L. Jansen,L.A. Joosten, Soluble uric acid primes TLR-induced proinflammatory cytokineproduction by human primary cells via inhibition of IL-1Ra, Ann. Rheum. Dis. 75(2016) 755–762.

[56] J. Darragh, O. Ananieva, A. Courtney, S. Elcombe, J.S. Arthur, MSK1 regulates thetranscription of IL-1ra in response to TLR activation in macrophages, Biochem. J.425 (2010) 595–602.

[57] K. Rehani, H. Wang, C.A. Garcia, D.F. Kinane, M. Martin, Toll-like receptor-medi-ated production of IL-1Ra is negatively regulated by GSK3 via the MAPK ERK1/2, J.Immunol. 182 (2009) 547.

[58] J.E. Sims, M.A. Gayle, J.L. Slack, M.R. Alderson, T.A. Bird, J.G. Giri, F. Colotta,F. Re, A. Mantovani, K. Shanebeck, Interleukin 1 signaling occurs exclusively viathe type I receptor, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 6155.

[59] D.Q. Li, L. Luo, Z. Chen, H.S. Kim, X.J. Song, S.C. Pflugfelder, JNK and ERK MAPkinases mediate induction of IL-1β, TNF-α and IL-8 following hyperosmolar stress inhuman limbal epithelial cells, Exp. Eye Res. 82 (2006) 588.

[60] S.M. Loitsch, M.C. Von, S. Kippenberger, D. Steinhilber, T.O. Wagner, J. Bargon,Reactive oxygen intermediates are involved in IL-8 production induced by hyper-osmotic stress in human bronchial epithelial cells, Biochem. Biophys. Res. Commun.276 (2000) 571–578.

[61] Z.H. Németh, E.A. Deitch, C. Szabó, G. Haskó, Hyperosmotic stress induces nuclearfactor-κb activation and Interleukin-8 production in human intestinal epithelialcells, Am. J. Pathol. 161 (2002) 987–996.

[62] X. Meng, X. Tian, M. Liu, G. Nie, K. Jiang, B. Wang, L. Wang, The transcriptomicresponse to copper exposure by the gill tissue of Japanese scallops (Mizuhopectenyessoensis) using deep-sequencing technology, Fish Shellfish Immunol. 38 (2014)287–293.

[63] J. Xiao, H. Zhong, Z. Liu, F. Yu, Y. Luo, X. Gan, Y. Zhou, Transcriptome analysisrevealed positive selection of immune-related genes in tilapia, Fish ShellfishImmunol. 44 (2015) 60–65.

J. Gu et al. Fish and Shellfish Immunology 73 (2018) 288–296

296