Abstract

   Koi herpesvirus (KHV, CyHV-3) causes severe economic losses in carp
   farms. Its eradication is challenging due to the establishment of
   latency in blood leukocytes and other tissues. To understand the
   molecular mechanisms leading to KHV infection in leukocytes, common
   carp were bath-exposed to KHV at 17 °C. After confirming the presence
   of viral transcripts in blood leukocytes at ten days post infection,
   RNA-Seq was performed on peripheral blood leukocytes on the Illumina
   NovaSeq. KHV infection triggered a robust immune response mediated by
   pattern recognition receptors, mainly toll-like receptors (tlr2, tlr5,
   tlr7, and tlr13), urokinase plasminogen activator surface
   receptor-like, galectin proteins, and lipid mediators such as
   leukotriene B4 receptor 1. Enriched pathways showed increased
   mitochondria oxidative phosphorylation and the activation of signalling
   pathways such as mitogen-activated protein kinases (MAPKs) and vascular
   endothelial growth factor (VEGF). KHV-infected leukocytes showed low
   production of reactive oxygen species (ROS) and glutathione metabolism,
   high iron export and phagocytosis activity, and low autophagy.
   Macrophage polarization was deduced from the up-regulation of genes
   such as arginase non-hepatic 1-like, macrophage mannose receptor-1,
   crem, il-10, and il-13 receptors, while markers for cytotoxic T cells
   were observed to be down-regulated. Further work is required to
   characterise these leukocyte subsets and the molecular events leading
   to KHV latency in blood leukocytes.

   Keywords: koi herpesvirus (KHV, CyHV-3); common carp; transcriptomics;
   blood leukocyte; M2 macrophage; toll-like receptors; immunosuppression;
   pattern recognition receptors; inflammation

1. Introduction

   Koi herpesvirus (KHV), also known as cyprinid herpesvirus 3 (CyHV-3) or
   carp interstitial nephritis gill necrosis virus, is the causative agent
   of koi herpesvirus disease (KHVD), a severe disease responsible for
   causing high mortality rates in ornamental koi and common carp
   (Cyprinus carpio L.). The disease was first reported in Israel in 1997
   and was soon observed worldwide, except in Australia, due to the
   intense global trade of koi and common carp [[38]1]. In 2006, KHVD was
   made notifiable by the World Organisation for Animal Health (WOAH).

   KHV primarily infects epithelial cells of skin and gills, after which
   it becomes viraemic, causing systemic infection and finally infecting
   cells of the central nervous system [[39]2]. The onset and severity of
   the disease are highly influenced by water temperature, with an optimal
   water temperature between 16 and 25 °C. Infections at lower
   temperatures often result in asymptomatic infections and reduced
   pathology [[40]3].

   KHV belongs to the family Alloherpesviridae within the order
   Herpesvirales. Two main lineages, European and Asian, have been
   identified based on a variable number of tandem repeats (VNTR) analysis
   [[41]4]. With a large, enveloped, double-stranded linear DNA molecule
   of 298 kbp in length [[42]5], the KHV genome encodes for at least 156
   predicted genes in open reading frames (ORFs) [[43]6]. The virion is
   formed by forty structural proteins, including the immunogenic type 3
   membrane protein expressed by ORF81 [[44]6]. Non-structural viral
   proteins such as interleukin-10 (IL10, ORF134), lipoprotein (ORF68),
   tumour necrosis factor receptor (TNFR-1, ORF4L) and TNF-2 (ORF12) have
   been suggested to play a role in the virus survival within the host
   [[45]7]. However, very little is known about the interaction of these
   viral proteins with the host immune system.

   During the productive phase of the infection, KHV triggers the
   activation of signalling pathways mainly mediated by interleukin family
   members and interferon-gamma (IFN)-γ2 [[46]8]. Despite initiating
   inflammatory responses, KHV subversion of the host’s complement system
   in mucosal sites allows for long periods of viral shedding [[47]9] and
   facilitates secondary infections by compromising the mucosal barrier of
   the skin and the epithelium tight junctions by down-regulating mucin,
   defensin, and claudin proteins [[48]10]. Breeding selection for KHV
   resistance has resulted in differences in the immune response between
   resistant and susceptible families. In particular, resistance to KHV
   has been linked to autophagy, phagocytosis, cytotoxicity, virus
   blockage by lectins and mucin, higher expression of immune-related
   genes, complement activation, MHC class I-restricted antigen
   presentation, and the development of adaptive mucosal immunity
   [[49]11,[50]12,[51]13]. Interestingly, differences in the leukocyte
   population have also been observed among resistant and susceptible carp
   families, with a higher presence of cxcl8/il8 chemokines and
   neutrophils in resistant carp and a higher presence of T-cells and
   macrophages in susceptible families [[52]14]. KHV infection elicits a
   specific humoral response characterised by high-affinity B-cells
   [[53]15].

   Like many herpesviruses, KHV can lead to life-long infection through
   latency [[54]7]. Peripheral blood leukocytes, such as IgM-positive B
   lymphocytes, and other tissues, such as neurons, have been identified
   as cell targets in which KHV becomes latent [[55]16,[56]17]. In carp,
   latent KHV can be reactivated and enter productive replication after
   the fish experience stress, such as when exposed to increased water
   temperature [[57]16]. Thus, despite the key role of blood leukocytes in
   initiating inflammatory responses and maintaining KHV infection through
   latency, the molecular mechanisms leading to KHV infection in
   leukocytes remain poorly understood. This study aimed to identify
   critical molecular processes occurring during KHV infection in blood
   leukocytes before a humoral response occurs. We hypothesise that KHV
   infection may promote leukocyte subset differentiation involved in the
   inflammatory response and establishment of latency. To address this
   hypothesis, we infected common carp with KHV and used RNA-Seq to
   document the molecular responses to infection in isolated blood
   leucocytes; the information delivered from this study enhances our
   understanding of herpesvirus infection in aquatic species and provides
   valuable evidence to develop therapeutic and vaccinology strategies.

2. Materials and Methods

2.1. KHV Inoculum Preparation

   A European KHV isolate, K250, was obtained from a crucian carp KHVD
   outbreak in England in 2007. KHV was isolated in KF1 cells (ECACC
   10072801, Merck, Gillingham, UK), and the presence of the virus was
   confirmed via Sanger sequencing of a specific PCR product targeting the
   viral thymidine kinase gene [[58]18]. K250 was propagated on a common
   carp brain (CCB) cell line (ECACC 10072802) at 20 °C in EMEM media
   supplemented with two mM glutamine, 1% non-essential amino acids, 2%
   foetal bovine serum (FBS), and 10 mM HEPES (Merck, Gillingham, UK).
   Viral particles were harvested from the supernatant of cells showing
   cytopathic effects via centrifugation at 4000× g for 15 min in a Falcon
   6/300 centrifuge ( MSE, Cholet, France) and the viral titre stock
   inferred via the 50% tissue culture infective dose (TCID50) [[59]19].

2.2. Carp Experimental Infection and Blood Samples

   Twenty-four specific pathogen-free (SPF) immature mixed-sex common
   carp, weighing ~400 g, were split into four tanks of six fish each and
   acclimated to 17 °C. Two tanks were exposed to 15 mL of KHV isolate
   K250 at ~10^8 TCID[50]/mL (maximum titre, harvested from cell culture)
   diluted in 30 L of tank water, giving a viral titre on the immersion
   tanks of ~5 × 10^4 TCID[50]/mL, on an aerated static bath for four
   hours. The other two groups were mock-infected with culture media. The
   challenge was conducted at 17 °C to allow for a slow progression of the
   disease following a well-established challenge model [[60]9]. After the
   bath infection, each group was kept in a 600 L tank with a flow rate of
   4–6 L min^−1 for a maximum of 10 days, equivalent to 170-degree days
   (DD). Four control and KHV-infected carp were sampled on day four
   (pre-clinical fish), on day seven (fish showing mild symptoms such as
   loss of appetite), and on day ten (carp developing skin lesions typical
   of KHV) ([61]Supplementary Figure S1). These sampling points (four,
   seven, and ten days) were selected to identify the presence of viral
   transcripts in white blood cells before a humoral response occurred
   [[62]9]. Blood samples were collected from the caudal vein of MS222
   (400 ppm solution, Merck, Gillingham, UK),) anaesthetised fish, placed
   in heparinised tubes (BD vacutainer) containing potassium
   ethylenediaminetetraacetic acid (EDTA) and stored at 4 °C until further
   use. Blood smears were left to air dry and stained with Giemsa (Sigma)
   following the manufacturer’s standard protocols. In parallel, kidney
   sections were (Merck, Gillingham, UK) collected and stored in RNAlater
   ^® (Merck, Gillingham, UK) to confirm KHV infection via real-time qPCR,
   as described below.

2.3. Leukocyte Isolation and Blood Cell Population Count

   Blood leukocytes were separated using density gradient centrifugation
   following [[63]20]. Briefly, equal volumes of blood and
   phosphate-buffered saline (PBS) (Merck, Gillingham, UK) were mixed and
   layered into a freshly made discontinuous (from the top layer down)
   20%, 30%, 40%, and 50% density gradient of 1.130 g/mL Percoll^®, pH
   8.5–9.5 (Merck, UK) at a ratio of 1 mL of diluted blood in 6 mL. The
   density gradient was obtained via centrifugation at 2000× g for 30 min
   in a swinging bucket rotor Falcon 61,300 centrifuge (MSE, Cholet,
   France)) with the brake OFF at 4 °C. Leukocytes, mainly located between
   the 30 and 40% interface, were collected and washed with five volumes
   of PBS. Cells in PBS were split into three equal aliquots; one aliquot
   was pelleted via centrifugation at 500× g for 5 min and stored in 500
   uL of RNAlater ^® (Merck, Gillingham, UK) at −20 °C for RNA extraction;
   the second aliquot was resuspended in staining buffer (eBioscience,
   Paisley, UK) and the size and granularity of the isolated leukocytes
   analysed in a flow cytometer (CytoFlex, Beckman, Wycombe, UK); and the
   third aliquot was used for the flow cytometry detection of KHV, as
   described below.

2.4. Flow Cytometry Detection of KHV in Peripheral Blood Leukocytes

   The average size and granularity of common carp blood leukocytes were
   observed in a CytoFlex (Beckman, Wycombe, UK) flow cytometer. Forward
   scatter (FSC) and side scatter (SSC) properties of ten thousand events
   were measured and analysed in CytExpert version 2.3.0.84 software
   (Beckman, Wycombe, UK). The leukocyte population gating was conducted
   according to previous studies on carp [[64]21].

   To identify the percentage of KHV-positive leukocytes, approximately
   10^6 leukocytes per mL were fixed in 100 µL of IC fixation buffer
   (eBioscience, Paisley, UK) for 30 min at room temperature (~18–20 °C),
   washed with a staining buffer (eBioscience, Paisley, UK), centrifuged
   at 500× g for 5 min in an Eppendorf centrifuge 5417R (Eppendorf,
   Arlington, UK), and then resuspended in 500 µL of permeabilisation
   buffer (eBioscience, Paisley, UK) containing 1% of bovine serum albumin
   (BSA) (Merck, Gillingham, UK) for 15 min. A KHV monoclonal antibody
   (Aquatic Diagnostics, Argyll, UK) was conjugated with Alexa Flour 488
   using the Zip Alexa Fluor™ Rapid Antibody Labelling Kit (Invitrogen,
   Gillingham, UK) per the manufacturer’s instructions. Fixed and
   permeabilised leukocytes were incubated with a solution containing
   permeabilisation buffer, 1% of BSA, and 1 µg of the Alexa
   Fluor-conjugated KHV antibody for 30 min in the dark at room
   temperature. The cells were washed twice with the staining buffer and
   analysed on a CytoFlex. Significant differences among groups were
   determined through Student’s t-tests for unequal variance, given that
   normality and equal variance assumptions for parametric testing were
   unmet. Data are presented as mean ± standard deviation of the mean.

2.5. Library Prep and RNA Sequencing

   Total RNA was extracted from leukocytes and kidney sections using the
   Maxwell^® RSC simplyRNA Cells Kit or the Maxwell RSC simplyRNA Tissue
   Kit in the Maxwell^® RSC 48 extraction robot (Promega, Southampton,
   UK), following the manufacturer’s instructions. RNase-free DNase I
   (Promega, Southampton, UK) was added to remove DNA contamination during
   the RNA extraction, following the manufacturer’s guidelines. RNA-Seq
   was conducted on leukocytes harvested from control carp and
   KHV-infected carp sampled ten days post infection (dpi), and three
   replicate samples were analysed for each group. RNA quality was
   visualised on a TapeStation with an RNA ScreenTape & Reagents (Agilent,
   Cheadle, UK). Illumina-indexed libraries were prepared with
   concentrations ranging from 50 to 200 ng RNA with rRNA depletion as
   part of the lllumina Stranded Total RNA Prep Ligation with Ribo-Zero
   Plus protocol (Illumina, Great Abington, UK). Samples were pooled in
   equimolar concentrations before denaturing, diluting, and 150
   paired-end (PE) sequencing on the Illumina NovaSeq 6000 platform over
   0.4 lanes of an SP flow cell (Illumina, Great Abington, UK).

2.6. Annotation of the Common Carp Genome

   The carp reference genome (ASM1834038v1; RefSeq Assembly Accession
   GCF_018340385.1) and corresponding protein sequences were downloaded
   from the RefSeq database [[65]22]. The carp protein sequences were
   subjected to sequence similarity searches using Diamond v2.0.15
   [[66]23] and the full NCBI nr database (downloaded on 3 April 2022).
   Functional annotations were assigned to the sequences using
   InterProScan v5.55-88.0 and all default analysis modules [[67]24]. All
   annotations were imported into OmicsBox v1.4.12 [[68]25], which was
   used to add gene ontology (GO) annotations to each of the proteins
   (Gene Ontology Annotation database accessed on 1 August 2022;
   [69]ftp://ftp.ebi.ac.uk/pub/databases/GO/goa/).

2.7. Sequence Pre-Processing, Read Mapping, and Abundance Estimation

   Raw data quality control and the removal of adapter sequences and
   low-quality bases were conducted using Fastp 0.20.0 [[70]26]. Ribosomal
   RNA (rRNA) sequences were removed with SortMeRNA v4.3.6 [[71]27] using
   default parameters and a selection of SILVA SSU, LSU [[72]28], and RFAM
   [[73]29] databases, including rfam-5.8s-database-id98,
   rfam-5s-database-id98, silva-arc-16s-id95, silva-arc-23s-id98,
   silva-bac-16s-id90, silva-bac-23s-id98, silva-euk-18s-id95, and
   silva-euk-28s-id98. The nf-core/RNAseq v3.0 pipeline [[74]30] was used
   to map the reads to both the common carp reference genome and the KHV
   genome ([75]NC_009127.1) using STAR v2.7.6a [[76]31] and to quantify
   gene expression abundances with RSEM v1.3.1 [[77]32] and samtools v1.10
   [[78]33]. The resulting gene expression count matrices were loaded into
   the DEBrowser v1.18.3 [[79]34] for differential gene expression
   analysis. Genes with less than one count per million in at least two
   samples were removed, and expression values were normalised across the
   samples using median ratio normalisation (MRN). Differences in gene
   expression between the various treatments were determined using DESeq2
   v1.30.1 [[80]35] with parametric fit dispersion, the likelihood ratio
   test (LRT) and the Wald test. Genes were considered differentially
   expressed when adjusted p values < 0.05 (Benjamini-Hochberg correction)
   and fold change (FC) > 2. Principal component analysis (PCA) plots,
   heatmaps (based on hierarchical clustering: Euclidean as the distance
   measure and complete for clustering linkage), and scatter plots of read
   counts were generated from within the DEBrowser using default
   parameters.

2.8. Functional Annotation Analysis

   To identify enriched biological processes and pathways, gene set
   enrichment analysis (GSEA) was performed on a ranked list of genes,
   based on their associated log2 FC values (high to low), using the
   clusterProfiler package (v3.18.1) [[81]36] within Rstudio v1.3.1056
   [[82]37] running R v4.0.2 [[83]38]. Gene ontology (GO) and Kyoto
   Encyclopaedia of Genes and Genomes (KEGG) pathway gene sets were used
   for the analysis. The GO enrichment analyses were conducted using the
   GSEA function of clusterProfiler, with minimum and maximum gene set
   sizes of 10 and 500, respectively. KEGG pathway enrichment analysis was
   conducted using the gseKEGG function, with a minimum gene set size of
   10 and the KEGG dataset for Cyprinus carpio (ccar). Processes and
   pathways were considered to be enriched when adjusted p values < 0.05
   (Benjamini–Hochberg correction). Dot plots for the GO terms and KEGG
   pathways were generated using the dot plot function in clusterProfiler.

2.9. Real-Time PCR Validation of RNA-Seq Differentially Expressed Genes

   Total RNA was reverse transcribed at 37 °C for 1 h in a 20 µL reaction
   containing 200 U of M-MLV RT, M-MLV RT 5× reaction buffer (250 mM
   Tris-HCl, pH 8.3; 375 mM KCl; 15 mM MgCl[2]; 50 mM DTT); 1 mM dNTP mix;
   500 ng of random primers; and 25 units of RNasin^® Ribonuclease
   Inhibitor (Promega, Southampton, UK). KHV transcripts were analysed
   using a diagnostic real-time PCR (RT-PCR) targeting a fragment of the
   viral orf90 gene [[84]39]. The relative expression of eleven carp genes
   found differentially expressed (DEG) via the RNA-Seq analysis were also
   assayed using RT-PCR. Primers and probes were de novo designed using
   Primer Express v3.0 (ThermoFisher, Ringway, UK) (primers sequences in
   [85]Table S1 in the Supplemental Materials). A previously described
   RT-PCR assay for the carp actin beta 1 (actb1) gene was used as a
   reference gene for normalisation purposes [[86]9,[87]40].
   Reverse-transcribed cDNA from five control and four KHV-infected carps
   sampled at ten dpi, which included the same RNA extracted from the
   samples used for generating the sequencing library, were tested with
   the RT-PCR assays. Each sample was analysed in duplicate in a master
   mix containing 2 µL of the cDNA, 500 nM of each primer, and 250 nM of
   the probe labelled with 6-FAM 5′ and 3′ BHQ1, in a total volume of 20
   µL, by using the Taqman^® Universal Master Mix II with UNG (Life
   Technologies, Paisley, UK). PCR reactions were performed on a StepOne
   Plus™ Real-Time PCR System V2.3 software (Applied Biosystem, UK) with
   40 cycles of PCR amplification and fluorescence detection as
   recommended by the manufacturer. Negative control samples containing
   molecular-grade water were used. PCR efficiency was calculated for each
   assay using serial dilutions of a control cDNA. Cycle thresholds (Ct)
   were used to calculate the FC (2∆∆Ct) expression of a target gene
   between samples using the PCR package [[88]41] in the R Bioconductor
   within Rstudio v1.3.1056 running R v4.0.2 [[89]37,[90]38].

3. Results

3.1. Detection of KHV in White Blood Cells

   Flow cytometry analysis of blood leukocytes showed the enrichment of
   two different populations of leukocytes (P1 and P2). Based on the size
   and granularity properties, these populations were hypothesised to be
   lymphocytes (P1) and either enlarged lymphocytes or monocytes (P2)
   ([91]Figure 1A,B and [92]Figure S2). Similarly, in blood smears, an
   apparent higher proportion of lymphocytes and monocytes was observed in
   infected carp sampled at ten dpi, but this was not formally quantified
   ([93]Figure 1C).

Figure 1.

   [94]Figure 1
   [95]Open in a new tab

   Peripheral blood leukocyte analysis in common carp infected with koi
   herpesvirus (KHV). (A,B) Pseudo-colour density dot plot showing the
   forward scatter (FSC) and side scatter (SSC) properties of blood
   leukocytes from (A) a control carp and (B) a KHV-infected carp sampled
   at ten days post infection (dpi). Two populations were gathered: P1 and
   P2. (C) Giemsa-stained blood smear photomicrograph of an infected carp
   at ten dpi. Black arrows: lymphocytes-like; open arrows:
   monocytes-like; asterisk: neutrophil-like. (D) Box plots showing the
   percentage of KHV-positive leukocytes, P1 and P2 populations in blood
   leukocytes, and the percentage of KHV positive cells in P1 (P1 and KHV
   positive) and P2 (P2 and KHV positive). Control group: 12 fish;
   KHV-infected group: four fish per each sampling point (four, seven, and
   ten dpi). Asterisk (*) denotes a significant difference (t-test p-value
   < 0.05) in the number of cells between control and infected carp. (E)
   Flow cytometry analysis of peripheral leukocytes of four control fish
   (C9, C10, C11, C12) and four KHV-infected fish (KHV9, KHV10, KHV11,
   KHV12) sampled at ten dpi stained with an anti-KHV antibody conjugated
   with Alexa Fluor 488. Leukocytes from KHV-infected carp showed
   increased fluorescence on the FITC channel in both P1 and P2
   populations.

   A significant increase in the number of cells representing the P1
   population was observed in KHV-infected carp sampled at ten dpi,
   coinciding with the development of skin lesions. In contrast, for the
   P2 population, a significant increase in the number of cells was
   observed at seven dpi ([96]Figure 1C). From the three sampling points
   (four, seven, and ten dpi), flow cytometry analysis identified
   KHV-positive leukocytes at ten dpi ([97]Figure 1D,E and [98]Figure S3)
   in both P1 and P2. The number of KHV-positive leukocytes increased to
   56% ± 21% compared with the basal fluorescence reading of uninfected
   carp leukocytes (0.3% ± 0.5%) ([99]Figure 1D). The detection of KHV via
   flow cytometry was corroborated by the detection of KHV transcripts in
   leukocytes at ten dpi and in kidney sections at 7 and 10 dpi by a
   KHV-specific Taqman RT-PCR ([100]Supplementary Figure S4). The
   detection of KHV in blood leukocytes at ten dpi supported the decision
   to conduct the transcriptomics analysis of isolated leukocytes at ten
   dpi.

3.2. Transcriptome Mapping of Carp Blood Leukocytes

   The average GC content of the sequencing data was 45.2% for the
   controls and 50.7% for the infected samples, and the average Q20 and
   Q30 scores were 98% and 94%, respectively. The raw paired reads
   generated for the various samples ranged from 29.64 to 34.71 million,
   averaging 32.28 ± 1.48 million. After the quality trimming and removal
   of rRNA sequences, the average number of read pairs per sample was
   32.18 ± 1.47 (ranging from 29.55 to 34.58 million) and 7.35 ± 3.58
   (ranging from 2.20 to 12.36 million), respectively ([101]Supplementary
   Table S2). There were no significant differences between the numbers of
   reads for any treatment groups, both before and after quality trimming
   (t-test p-value > 0.05). However, the infected groups had significantly
   more reads than the controls after removing the rRNA reads (p = 0.03).
   The NCBI RefSeq Cyprinus carpio genome assembly (ASM1834038v1) consists
   of 43,833 genes, which represent 80,673 transcripts that encode for
   80,686 protein sequences. GO annotations were assigned to 69,323
   reference proteins (85.9% of the total proteins) or 38,698 reference
   genes (88.3%). In our study, 24,773 genes (56.5% of the total number of
   reference genes) or 47,482 proteins were found to be expressed across
   the samples, of which 21,215 genes (85.6%) had GO annotations assigned
   to them. A list containing all expressed genes and their annotations is
   presented in [102]Supplementary Dataset S1.

3.3. Koi Herpesvirus Transcripts in Carp Leukocytes

   We mapped the sequence reads to the KHV genome to assess viral
   replication in leukocytes ([103]Supplementary Figure S5). The mapped
   viral transcripts represented only ~0.001% of the transcripts found in
   the leukocytes after read normalisation and rRNA removal. KHV
   transcripts found in infected blood leukocytes covered 61% (100 out of
   164 KHV ORFs) of gene-encoding proteins in the viral genome. The more
   abundant viral transcripts were gene products of the ORF68, which
   encodes for a myosin-like protein; the ORF35, encoding for an unknown
   viral protein; the ORF92, which encodes for the major capsid protein;
   and the ORF62, encoding for a putative polypeptide substrate binding
   site. Transcripts of the viral ORF90, the target of the currently
   recommended RT-qPCR on the WOAH diagnostic manual [[104]39], were
   expressed but were lower than those of other viral genes.

3.4. Hierarchical Clustering Revealed Dramatic Differential Transcription
Profiles in KHV-Infected Carp Leukocytes

   Clustering analysis of all differentially expressed genes after quality
   control and normalisation revealed that KHV-infected common carp
   leukocytes clustered together and separated from the controls
   ([105]Figure 2A). Principal component analysis (PCA) showed low
   replicate variability among the controls and KHV-infected, although a
   higher variation was observed among the KHV samples. From all the
   detected expressed genes, 7228 genes were found to be significantly
   up-regulated (p < 0.05, FC > 2) and 6818 genes down-regulated (p <
   0.05, FC < 2) in the KHV-infected group ([106]Figure 2B,C).

Figure 2.

   [107]Figure 2
   [108]Open in a new tab

   (A) Heatmap showing hierarchical clustering of all differentially
   expressed genes in common carp naïve leukocytes (samples C1, C2, and
   C3) and leukocytes infected with koi herpesvirus (KHV) (samples T1, T2,
   and T3). (B) Volcano plot showing up-regulated (up) and down-regulated
   (down) genes (p < 0.05, FC > 2) and genes found expressed but not
   significantly different from controls (NS) (p > 0.05). (C) Principal
   component analysis (PCA) plot of control (C1, C2, and C3) and
   KHV-infected (T1, T2, T3) common carp leukocytes sampled at ten days
   post infection.

3.5. Gene Set Enrichment Analyses and Over-Representation of KEGG Pathways in
KHV-Infected Leukocytes

   Gene set enrichment analysis (GSEA) identified 139 unique GO terms
   enriched in KHV-infected leukocytes ([109]Figure 3A–C,
   [110]Supplementary Dataset S2), and 38 over-represented KEGG pathways
   among the differentially expressed gene lists in KHV-exposed leukocytes
   ([111]Figure 3D, [112]Supplementary Dataset S3). Examples of genes that
   play a crucial role in those molecular pathways and that were found
   differentially expressed can be found in [113]Table 1.

Figure 3.

   [114]Figure 3
   [115]Open in a new tab

   Enriched gene ontology (GO) terms and Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathways in Koi herpesvirus-infected common carp blood
   leukocytes. Enrichment of GOs belonging to (A) cellular components, (B)
   molecular function, (C) biological processes, and (D) enriched KEGG
   pathways. The size and colour of each dot represent the number of genes
   and adjusted p values, as depicted in the legend (Benjamini–Hochberg
   correction; FDR < 0.05), respectively.

Table 1.

   List of selected differentially expressed genes (p-values < 0.05) in
   koi herpesvirus-infected common carp blood leukocytes. The full list of
   genes can be found in [116]Supplementary Dataset S1. The number of
   reads (transcripts) per million (TPM) mapped reads is shown for each
   sample. FC = fold change in control versus KHV-infected leukocytes. C:
   control samples; KHV: KHV-infected carp blood leukocytes. * Carp
   macrophage polarisation signatures according to Wentzel et al., 2020
   [[117]42].
   Gene Accession C (1) C (2) C (3) KHV (1) KHV (2) KHV (3) log2FC Gene
   Innate immune response
   [118]XM_042741157.1 0 0 0 62.82 25.63 346.84 10.09 toll-like receptor 2
   [119]XM_042711149.1 1.97 8.18 7.71 570.11 636.4 1111.85 7.05 toll-like
   receptor 13
   [120]XM_042778173.1 17.08 12.65 6.17 1927.08 1064.39 1366.38 6.91
   toll-like receptor 5
   [121]XM_042764380.1 4.6 0.74 0.77 180.61 298.7 372.01 7.06 toll-like
   receptor 7
   [122]XM_042710956.1 24.97 41.66 17.72 4765.08 2668.21 3978.87 7.08
   complement factor B-like
   [123]XM_042750556.1 5.91 2.98 6.17 293.69 98.08 275.51 5.46 antiviral
   innate immune response receptor RIG-I-like
   [124]XM_042713940.1 103.16 107.86 145.65 2461.07 2047.41 2216.7 4.24
   antiviral innate immune response receptor RIG-I
   [125]XM_042772231.1 26.94 19.34 16.18 1204.62 560.61 966.4 5.44 tumour
   necrosis factor receptor superfamily member 1A
   [126]XM_019090552.2 0 0 0 20.42 10.03 20.98 7 interferon-induced
   protein with tetratricopeptide repeats 5-like
   [127]XM_042724080.1 21.02 14.87 23.89 243.43 91.39 271.31 3.38 grass
   carp reovirus (GCRV)-induced gene 2o
   [128]XM_019104238.2 4.6 1.49 0 8259.58 4073.65 7973.12 11.63 C-C motif
   chemokine 19a.1
   [129]XM_019081222.2 7.225.2 5.20 10.78 1367.95 1747.60 1747.60 8.04
   interferon-induced GTP-binding protein MxB-like
   [130]XM_042768955.1 0 0 0 111.51 14.49 117.48 9.25 interferon-induced
   protein 44-like isoform X1
   [131]XM_042747598.1 17.08 14.13 18.5 67491.73 18,406.65 61,035.41 11.53
   urokinase plasminogen activator surface receptor-like
   [132]XM_019072681.2 0 2.98 0 144.49 71.33 114.68 6.83 polymeric
   immunoglobulin receptor-like
   [133]XM_042759636.1 1.97 0 0 161.77 37.89 246.14 7.72 interleukin-13
   receptor subunit alpha-1-like
   [134]XM_019084892.2 58.48 51.33 48.55 675.34 624.14 609.77 3.59
   interferon regulatory factor 3
   [135]XM_042746991.1 231.93 215.72 237.36 1397.8 1259.43 2093.63 2.79
   interferon regulatory factor 4-like
   [136]XM_042711994.1 7.88 11.9 10.79 0 1.11 0 −4.4 complement C3-like
   [137]XM_019097650.2 45.99 88.52 73.21 3.14 12.26 1.4 −3.54 vascular
   endothelial growth factor A-A isoform X2
   [138]XM_019068368.2 11.17 9.67 16.18 1.57 3.34 8.39 −1.47 transcription
   factor Maf-like
   Attraction of mononuclear phagocytes
   [139]XM_019080073.2 14.45 23.06 5.39 5500.11 3708.08 3164.91 8.17
   interleukin-1 beta
   [140]XM_042733011.1 413.28 678.41 315.19 26,566.05 17,916.25 16,020.36
   5.43 collagenase 3a
   [141]XM_019074825.2 93.96 125.71 91.71 2272.6 1176.96 1724.41 4.05
   ras-related C3 botulinum toxin substrate 2
   [142]XM_019099675.2 0 0 0 175.9 215.11 40.56 10.08 interleukin-8
   [143]XM_042763296.1 0 0.74 0 86.38 32.32 46.15 7.72 tumour necrosis
   factor ligand superfamily member 11 isoform X1
   [144]XM_042744022.1 0 0.74 0 257.57 53.5 160.83 9.24 leukotriene B4
   receptor 1-like
   Mitochondrial and iron metabolism
   [145]XM_042767134.1 4.6 2.23 6.17 650.21 283.09 351.04 6.63
   fructose-bisphosphate aldolase A-like
   [146]XM_042772236.1 0.66 0 0 28.27 0 13.99 5.75
   glyceraldehyde-3-phosphate dehydrogenase
   [147]XM_042758635.1 126.81 52.81 129.47 518.29 779.06 949.61 2.86
   vascular endothelial growth factor C-like
   [148]XM_042729971.1 264.79 237.29 190.35 87.95 81.36 86.71 −1.44
   hydroxyacylglutathione hydrolase, mitochondrial isoform X1
   [149]XM_042768274.1 13.8 5.95 10.02 0 1.11 0 −4.36 5-oxoprolinase
   [150]XM_019069346.2 2540.11 2938.27 2714.99 4.71 34.55 8.39 −7.36
   microsomal glutathione S-transferase 1
   [151]XM_042714822.1 15.77 11.16 6.94 237.15 112.57 303.48 4.26
   cytochrome c oxidase subunit 6B1-like
   [152]XM_042713191.1 3.29 2.23 6.17 80.1 37.89 67.13 3.99 succinate
   dehydrogenase [ubiquinone] iron-sulphur subunit, mitochondrial-like
   [153]XM_019067779.2 201.05 278.95 181.87 3.14 1.11 0 −7.36 ferritin,
   lower subunit-like
   [154]XM_042720263.1 3.29 46.86 53.95 0 0 0 −7.13 ferritin, middle
   subunit-like
   [155]XM_042777542.1 316.69 288.62 220.41 1.57 11.15 12.59 −4.97
   glutaredoxin-related protein 5, mitochondrial-like
   [156]XM_042730488.1 1854.81 1651.38 1301.62 10,618.57 5679.7 13,834.43
   2.65 solute carrier family 40 member 1
   Autophagy and protein degradation
   [157]XM_019066281.2 56.51 60.25 68.59 4.71 8.92 1.4 −3.56 BTB/POZ
   domain-containing adapter for CUL3-mediated RhoA degradation protein
   1-like
   [158]XM_042763993.1 21.03 14.13 15.41 0 1.11 0 −5.13 FUN14
   domain-containing protein 1-like
   [159]XM_042714562.1 74.25 75.13 81.69 7.85 8.92 2.8 −3.54 sorting
   nexin-7-like isoform X2
   Epigenetic and transcriptional regulation
   [160]XM_019116784.2 9.86 5.95 11.56 0 0 0 −5.2 sclerostin
   domain-containing protein 1a
   [161]XM_042722790.1 239.16 196.38 201.91 7.85 14.49 5.59 −4.46
   C2H2-type zinc finger protein
   [162]XM_042767739.1 204.34 190.43 171.85 36.12 49.04 54.54 −2.01
   male-specific lethal 1 homolog
   [163]XM_042729995.1 19.05 17.11 17.72 4.71 0 0 −3.69 histone-lysine
   N-methyltransferase SETD1B-like isoform X2
   Phagocytosis and actin network
   [164]XM_042742454.1 0 1.49 0 48.69 14.49 227.96 7.63 spectrin beta
   chain, non-erythrocytic 5 isoform X1
   [165]XM_019088805.2 26.94 24.55 21.58 331.39 254.12 1026.53 4.46
   allograft inflammatory factor 1-like isoform X1
   [166]XM_042737753.1 0.66 0 0 15.71 24.52 12.59 6.1 smoothelin-like 1
   isoform X1
   [167]XM_042718436.1 25.62 23.06 20.04 207.31 177.21 22.38 2.56
   macrophage mannose receptor 1-like
   [168]XM_042772413.1 53.88 43.14 70.9 373.79 345.51 346.84 2.67
   septin-2B isoform X2
   [169]XM_042725715.1 34.17 39.42 35.45 480.59 558.39 383.2 3.71 septin
   9b isoform X1
   [170]XM_042732942.1 5.26 0.74 7.71 36.12 25.63 64.33 3.2 septin-9-like
   Extracellular trafficking of proteins
   [171]XM_019107912.2 2.63 1.49 0 1444.92 503.77 991.57 9.42
   galectin-2-like
   [172]XM_042731380.1 0 1.49 0 739.73 292.01 567.81 10.11
   galectin-3-binding protein A-like
   [173]XM_042734258.1 11.17 15.62 11.56 1176.35 499.31 2032.09 6.6
   SLIT-ROBO Rho GTPase-activating protein 2 isoform X1
   [174]XM_042726165.1 0 0 0 26.7 8.92 27.97 7.31 rho guanine nucleotide
   exchange factor 25
   [175]XM_042727426.1 1.31 0 0 119.36 25.63 247.54 8.11
   serine/threonine-protein phosphatase 2A 56 kDa regulatory subunit beta
   isoform
   [176]XM_042764581.1 0 0 0 53.4 22.29 54.54 8.34 tyrosine-protein kinase
   transmembrane receptor ROR2-like
   [177]XM_042764258.1 0.66 0 0.77 424.05 179.44 155.24 9.04 G-protein
   coupled bile acid receptor 1
   [178]XM_019081835.2 0 0 0 39.26 15.6 26.57 7.66 psychosine
   receptor-like
   [179]XM_042717793.1 39.42 55.05 8.48 3726.94 2631.43 1974.75 6.34
   epidermal growth factor receptor-like
   Interleukins
   [180]XM_019113158.2 1.31 0.74 1.54 42.41 43.47 22.38 4.91 tumour
   necrosis factor-like
   [181]XM_019116350.2 3.29 5.21 8.48 560.69 270.83 507.67 6.32
   transforming growth factor beta-1 proprotein-like
   [182]XM_042734270.1 0.66 0.74 0 29.84 10.03 19.58 5.36 interleukin-10
   [183]XM_042721949.1 11.83 8.93 10.79 0 0 0 −5.41 interleukin-22
   [184]XM_042721950.1 34.17 14.88 12.33 0 0 0 −6.37 interleukin-26
   [185]XM_019098159.2 1081.49 851.73 756.78 1.57 1.11 0 −9.98 C-X-C motif
   chemokine 14
   [186]XM_042721947.1 7.88 6.69 10.79 1.57 3.34 0 −2.3 interferon gamma 1
   isoform X2
   Cluster of differentiation
   [187]XM_019082792.2 34.82 54.3 36.99 2091.99 1338.56 1496.45 5.29 CD9
   antigen isoform X1
   [188]XM_042721119.1 1.31 2.23 0.77 1551.71 626.37 1288.06 9.65 CD9
   antigen-like
   [189]XM_019125028.2 103.16 122.74 81.69 5396.45 2833.16 2206.91 5.08
   CD59 glycoprotein-like
   [190]XM_042765175.1 159 87.03 107.12 1380.52 802.47 1314.63 3.31 CD166
   antigen homolog isoform X2
   [191]XM_042716436.1 804.22 794.45 945.58 18.85 44.58 25.17 −4.81 CD99
   molecule
   [192]XM_042778750.1 24.31 17.11 31.6 3.14 2.23 2.8 −3.18 T-cell surface
   glycoprotein CD8 alpha chain
   [193]XM_019112429.2 17.74 8.92 15.41 1.57 5.57 12.58 −1.08 T-cell
   surface glycoprotein CD5
   [194]XM_019075258.2 63.07 87.03 53.17 9.42 26.74 12.58 −2.02 T-cell
   surface glycoprotein CD3 zeta chain-like
   Carp M1-type macrophage *
   [195]XM_019114762.2 0 0 0 138.21 51.27 222.37 10.01 C-X-C motif
   chemokine 13-like
   [196]XM_042759493.1 5.91 14.88 8.48 6129.9 3212.11 19460.79 9.95 serum
   amyloid A
   [197]XM_042731330.1 0.65 2.23 0 40.83 60.18 40.55 5.62 cis-aconitate
   decarboxylase-like
   [198]XM_019068715.2 15.76 10.41 4.62 284.27 21.17 258.73 4.18
   olfactomedin-4-like
   Carp M2-type macrophage *
   [199]XM_042725334.1 0 0 0 496.3 24.52 288.1 10.99 transglutaminase 5,
   like
   [200]XM_042737121.1 72.93 83.31 29.28 5699.57 3166.41 1065.69 5.74
   arginase, non-hepatic 1-like
   [201]XM_019101597.2 6.57 2.97 5.39 207.31 140.43 180.41 5.12
   metalloreductase steap4
   [202]XM_019074086.2 132.72 38.68 87.08 1157.50 863.76 1064.29 3.57
   heparin-binding EGF-like growth factor a
   [203]XM_042714579.1 44.68 43.89 59.34 713.03 590.71 209.78 3.36
   cAMP-responsive element modulator-like isoform X1
   [204]XM_042777120.1 97.24 26.77 54.71 238.72 317.64 339.84 2.32
   angiopoietin-1-like
   [205]XM_042777566.1 76.21 36.44 47.00 194.74 215.10 290.89 2.12
   phospholipid phosphatase 3-like isoform X1
   [206]Open in a new tab

   Based on the list of enriched GO and KEGG pathways, our data showed
   that KHV infection in carp triggered strong immune and proinflammatory
   responses, as seen by the enrichment of numerous immune-related GO
   terms such as inflammatory response, innate immune response, and
   defence response to virus. In particular, we observed the production of
   pro-inflammatory chemokines and cytokines responsible for cell-to-cell
   communication and recognition of infection. Enriched signalling
   pathways included the chemokine-mediated signalling pathway and the
   cytokine-mediated signalling pathway, and among the KEGG pathways were
   the toll-like receptor (TLR) signalling pathway, C-type lectin receptor
   pathway, NOD-like receptor signalling pathway, and the antiviral
   specific RIG-I-like receptor signalling pathway. The attraction of
   mononuclear phagocytes (monocytes and macrophages) was deduced from a
   strong enrichment of the GO terms macrophage chemotaxis and neutrophil
   chemotaxis.

   Changes in the metabolism of macrophages were suggested by the
   enrichment of biological processes associated with increased metabolic
   glycolysis, such as glycolytic process, glycolysis/gluconeogenesis,
   mitochondrial oxidative phosphorylation, the mitogen-activated protein
   kinases (MAPK) signalling pathway, and the vascular endothelial growth
   factor (VEGF) receptor signalling pathway. Metabolic processes observed
   to be suppressed included the glutathione metabolic process,
   glutathione transferase activity, and the KEGG pathways glutathione
   metabolism and peroxisome proliferator-activated receptor (PPAR)
   signalling pathway. Changes in the iron metabolism in KHV-exposed
   leukocytes were also inferred by the suppression of biological
   processes such as iron ion transport and intracellular iron ion
   homeostasis and the suppression of KEGG pathways such as ferroptosis
   and porphyrin metabolism.

   We observed the inhibition of molecular pathways related to
   transcription and autophagy as seen by strong suppression of the KEGG
   pathway basal transcription factors and autophagy, and GO terms related
   to protein degradation such as Cul4-RING E3 ubiquitin ligase complex
   and Cul3-RING ubiquitin ligase complex. The suppression of the GO terms
   also suggested a disruption of epigenetic regulation in KHV-exposed
   leukocytes. We observed the activation of phagocytosis in KHV-infected
   leukocytes, aided by a strong activation of the actin cytoskeleton.
   Some of the molecular pathways found activated were the KEGG phagosome
   and regulation of actin cytoskeleton, stress fibre, filamentous actin,
   and regulation of actin filament polymerisation. An interesting feature
   from the list of enriched GO terms is the key role of septins, a
   component of filamentous polymers, in connecting the leukocyte’s
   cytoskeleton network. We observed a strong activation of septin-related
   GO terms, such as septin complex and septin ring, and biological
   processes involving septins, such as the integrin-mediated signalling
   pathway. We also observed activated pathways related to the
   extracellular trafficking of proteins, which included the Arp2/3
   protein complex, extracellular space, and extracellular region. The
   dynamic trafficking of proteins was also supported by the activation of
   biological processes involved in cell signalling, enzyme activity, and
   protein binding, as observed by the enrichment of GO terms such as
   protein dephosphorylation and G-protein-coupled receptor activity.

3.6. Carp Blood Leukocyte Immunophenotype in Response to Koi Herpesvirus
Infection

   The gene expression of typical markers of polarised macrophages was
   observed in KHV-infected leukocytes ([207]Table 1). Examples of
   up-regulated M2-type macrophage markers were transglutaminase 5-like
   (tgm5l) (Log2FC = 10.98), arginase non hepatic 1-like (Log2FC = 5.74),
   and cAMP-responsive element modulator-like (crem) (Log2FC = 3.35),
   while examples of M1-type markers up-regulated were C-X-C motif
   chemokine 13-like (Log2FC = 10.01) and serum amyloid A (Log2FC = 9.94).
   Interestingly, a typical marker for M1-type, nitric oxide synthase 2b
   was not differentially regulated (log2F = 0.65). Cell surface receptors
   found strongly up-regulated included the urokinase plasminogen
   activator surface receptor-like (upar) (Log2FC = 11.53), TLRs such as
   tlr2 (Log2FC = 10.09), tlr5 (Log2FC = 6.91), tlr7 (Log2FC = 7.06), and
   tlr13 (tlr22) (Log2FC = 7.05), several G-protein coupled receptors such
   as g-protein coupled bile acid receptor 1 (gpbar1) (Log2FC = 9.04), the
   leukotriene B4 receptor 1-like (btl1) (log2FC = 9.24), a polymeric
   immunoglobulin receptor-like (Log2FC = 6.83), and the interleukin-13
   receptor subunit alpha-1-like (Log2FC = 7.72) ([208]Figure 4A).
   Up-regulated interleukins in carp leukocytes included the tumour
   necrosis factor (tnf)-like (Log2FC = 5.44), transforming growth factor
   beta-1 proprotein-like (Log2FC = 6.32), il-8 (cxcl8a) (Log2FC = 10.08),
   il1b (Log2FC = 8.17), il-10 (Log2FC = 5.36), and the chemokine C-C
   motif chemokine 19a.1 (ccl19a.1) (Log2FC = 11.63). Among the
   interleukins observed to be down-regulated were il-22 (Log2FC = −5.41),
   il-26 (Log2FC = −6.37), and chemokine C-X-C motif chemokine 14 (cxcl14)
   (Log2FC = −9.98) ([209]Figure 4B). Example of CDs observed to be highly
   up-regulated were different isoforms of the tetraspanin cd9 (Log2FC =
   5.29) and cd59 (Log2FC = 5.08), whether examples of CDs strongly
   down-regulated were cd99 (Log2FC = −4.81), cd8 (Log2FC = −3.18), T-cell
   surface glycoprotein cd5 (Log2FC = −1.08), and T-cell surface
   glycoprotein CD3 zeta chain-like (Log2FC = −2.02) ([210]Figure 4C).
   Other genes of interest for macrophage immunotyping were
   galectin-3-binding protein A-like (Log2FC = 10.11), interferon
   regulatory factors such as irf3 (Log2FC = 3.59), irf4 (Log2FC = 2.79),
   complement c3-like (Log2FC = −4.4), and the vascular endothelial growth
   factor A-A isoform X2 (vegfaa) (Log2FC = −3.54) ([211]Figure 4D).

Figure 4.

   [212]Figure 4
   [213]Open in a new tab

   Carp macrophage phenotype in response to koi herpesvirus infection
   (KHV). Gene expression (log2 fold change) of either up- or
   down-regulated (A) cellular receptors, (B) interleukins, (C) cluster of
   differentiation (cd), and (D) other genes informative for macrophage
   phenotyping.

3.7. Real-Time PCR Validation of RNA-Seq Data

   Eleven selected genes that were observed differentially expressed on
   the RNA-Seq data also showed significant differences (FC>2, p-values <
   0.05) when analysed by RT-PCR. The gene expression of carp il10 was
   detected by RT-PCR only in KHV-infected leukocytes. The FC direction of
   the genes tested via RT-PCR agreed with the RNA-Seq data. Moreover, the
   FC values calculated from both methods were highly correlated
   (Pearson’s correlation = 0.8254054) ([214]Figure 5).

Figure 5.

   [215]Figure 5
   [216]Open in a new tab

   Bar plot showing a comparison of the RT-PCR and RNA-Seq fold change
   values obtained for ten selected genes. Error bars represent 95%
   confidence interval (fold change ± 1.96 × SD), number of fish were
   three KHV-infected fish for the RNA-Seq analysis and four KHV-infected
   and control carps for the RT-qPCR tests, all sampled at ten days post
   infection. The inset scatter plot highlights the high correlation found
   between both methods. Genes ID: LOC109089066 urokinase plasminogen
   activator surface receptor-like, GenBank accession number
   [217]XM_042747598.1; ccl19a.1 C-C motif chemokine 19a.1,
   [218]XM_019104238.2; tgm5l transglutaminase 5-like, XM042725334.1;
   LOC122139442 arginase, non-hepatic 1-like, [219]XM_042737121.1;
   LOC109112007 CD59 glycoprotein-like, [220]XM_019125028.2; LOC109074181
   CD44 antigen-like isoform X2, [221]XM_042752900.1; LOC109072573 CD83
   antigen-like, [222]XM_019119505.2; LOC109058299 solute carrier family
   40 member 1, [223]XM_042730488.1; LOC109075423 macrophage mannose
   receptor 1-like, [224]XM_042718436.1; and LOC109106145 allograft
   inflammatory factor 1-like isoform X1, [225]XM_019088805.2. Note the
   plot does not show the fold change of the carp interleukin-10 (il10,
   [226]XM_042734270.1) which was only detected in KHV-infected samples.

4. Discussion

   CyHV3, or KHV, is a highly contagious virus that can cause significant
   mortality in carp species [[227]1]; its eradication is challenging due
   to the establishment of lifelong latency in white blood cells and other
   tissues [[228]7]. In this study, peripheral blood was collected during
   the active phase of infection, and the presence of KHV confirmed in
   leukocytes sampled at ten dpi. The bath immersion approach was chosen
   to mimic a natural route of infection [[229]12]. KHVD outbreaks are
   highly influenced by water temperature, typically occurring between
   16-25°C [[230]3]. On previous experimental challenges conducted at
   17°C, mortalities were recorded from day 10 to 27 with a survival rate
   of 40% [[231]9]; other parameters, such as fish size/age, genetic
   background, and previous exposure to KHV can influence the outcome of
   the infection [[232]14].

   In this study, the detection of KHV in blood samples was accompanied by
   an increased number of leukocytes in the P1 and P2 populations. In a
   previous study, a rapid prominent leukocytosis in the blood of
   KHV-infected carp occurred during the initial two dpi and throughout
   the 14 days of a clinical infection at 25 °C [[233]43]. Flow cytometric
   identification of leukocyte populations was based on previous studies
   on carp blood leukocytes [[234]20]. Here, leukocytosis was observed as
   a significant increase in P1 at ten and P2 at seven dpi coinciding with
   the detection of KHV in blood at 10 dpi. P1 was believed to be
   lymphocytes and P2 either enlarged lymphocytes or monocytes. Enlarged
   lymphocytes in carp have been previously described in parasitised blood
   samples [[235]44]. Currently, the identification and sorting of
   different subsets of carp leukocytes is difficult due to the lack of
   commercially available cell markers [[236]45].

   To uncover the molecular changes in carp blood leukocytes in response
   to KHV infection before a humoral reaction occurs, we analysed the
   enrichment of molecular pathways on unsorted blood leukocytes. The
   RNA-Seq analysis of carp blood leukocytes showed clear differences in
   the gene expression profile between uninfected and KHV-infected
   samples. However, a higher inter-group variation was observed among the
   infected samples, which might be attributed to differences in the level
   of infection or in host response. We confirmed the presence of KHV in
   leukocytes sampled at ten dpi and estimated that, on average, 56% ± 21%
   of the blood leukocytes were KHV-positive in the infected group. The
   variation in KHV-positive leukocytes within the infected samples might
   explain the inter-group variability observed in the RNA-Seq data.

   The analysis of the RNA-Seq data showed low levels of viral transcripts
   in the KHV-infected samples, with only 61% of the KHV ORFs identified.
   The RNA extraction and sequencing method used in this study was
   optimised to study the host response, and thus a specific enrichment
   for viral transcripts was not undertaken during the library
   preparation. Moreover, the presence of uninfected leukocytes in the
   bulk RNA might jeopardise the sequencing of low-expressed viral
   transcripts [[237]46]. Similar to other herpesviruses, the expression
   of KHV genes is classified into three groups of genes: immediate–early
   IE, early E, and late L [[238]47]. The top 15 KHV transcribed genes
   belonged to the E and L groups. Two major antigenic proteins, the ORF
   62 encoding for a putative polypeptide substrate binding site and the
   ORF 68 encoding for a myosin-like protein [[239]48], and a structural
   protein, the ORF92 encoding for the major capsid protein [[240]49],
   were among the top five most abundant KHV transcripts found in carp
   leukocytes. The viral ORF6, identified as the primary viral transcript
   expressed [[241]50] and translated [[242]51] in KHV-infected IgM+ B
   cells during latency, was ten times lower than those top-expressed
   viral genes.

   KHV infection triggered the recruitment of leukocytes via TLR and
   C-type lectin signalling. The molecular analysis suggested that KHV
   triggers a robust immune response, which is mediated by pattern
   recognition receptors (PRRs), mainly TLRs and C-type lectin
   stimulation, with the activation of several signalling pathways
   involved in the immune response and antiviral responses, consistent
   with that reported in the literature [[243]8]. Our data also suggested
   that KHV infection attracted macrophages which is in agreement with
   published data noting a higher presence of T-cells and macrophages in
   KHV-susceptible carps [[244]14]. Interestingly, we found that the genes
   tlr2, tlr5, tlr7, and tlr13 were highly expressed in KHV-exposed
   leukocytes, suggesting that KHV proteins may act as ligands of those
   receptors. Similarly, the gene expression of tlr2, tlr5, and tlr7 was
   previously found up-regulated in the spleen of KHV-infected carp
   [[245]8] strongly suggesting an important role in recognising KHV
   proteins. Little is known about TLR ligands in fish species despite
   their importance in initiating immune responses. The carp tlr2 is an
   orthologue of the mammalian TLR2 with lipopeptides acting as its ligand
   [[246]52], and it has been associated with immunosuppression and
   M2-type macrophages [[247]53]. Tlr13 and the soluble form of tlr5 are
   so far described exclusively in teleost species [[248]54,[249]55]. In
   some fish species, flagellin is a ligand for soluble tlr5 [[250]56],
   while the teleost ligand tlr7 is still unknown despite its orthologous
   in mammals playing an essential role in antiviral immunity [[251]56].

   Among the activated PRRs, the cellular receptor that showed the higher
   up-regulation in KHV-exposed blood leukocytes was an upar-like gene
   (log2FC > 10). High blood protein levels of their soluble form have
   been linked to immunosenescence and systemic chronic inflammation
   [[252]57]. Interestingly, in human macrophages, the uPAR system
   participates in the maintenance of intra-cytoplasmic vesicles that
   actively accumulate virions, thus sustaining the presence of viral
   reservoirs of macrophage origin [[253]58]. Other PPRs that were highly
   up-regulated were galectin proteins, such as galectin 2-like and
   galectin-3-binding proteins, and lipid mediators or G-protein-coupled
   receptors, such as blt1. Galectin family proteins have been previously
   described as involved in immune responses in different species of koi
   and glass carp [[254]59,[255]60]; in particular, the genes expression
   of galectin-3 was observed up-regulated in the kidney of KHV-infected
   carp [[256]12], and it has been associated with KHV-resistant common
   carp [[257]11]. The expression btl1 has been shown to activate several
   kinases and downstream immune signalling pathways and play a role in
   both acute and chronic inflammation [[258]61]. Thus, our dataset has
   identified specific PRRs activated in KHV-exposed leukocytes, in
   particular glycan-binding proteins and TLRs, and the subsequent
   initiation of specific signalling pathways that can aid in our
   understanding of carp innate immune responses to KHV and other cyprinid
   herpesviruses. Following KHV-induced inflammation, activation of the
   interferon (IFN)-I system has been described in KHV infection
   [[259]10,[260]62]. Similarly, we observed increased gene expression of
   an array of interferon-stimulated genes (ISG) in KHV-infected
   leukocytes, and amongst the antiviral proteins, the interferon-induced
   protein 44-like showed higher up-regulation. This gene has been
   observed highly up-regulated in virally infected zebrafish models
   [[261]63]; thus, it might be a useful marker for studying
   interferon-induced antiviral responses in cyprinid species.

   Enriched molecular pathways suggest macrophage polarisation in
   KHV-infected carp. Macrophage polarisation states range from
   classically activated “pro-inflammatory” M1 macrophages on one extreme
   to activated “anti-inflammatory or regulatory” M2 macrophages on the
   opposite extreme [[262]64]. After PPR stimulation, macrophage
   activation and differentiation typically occur through changes in cell
   metabolism [[263]64]. In this regard, we observed a marked increase in
   both metabolic glycolysis and oxidative phosphorylation (OXPHOS) and
   the activation of MAPK and VEGF signalling pathways. The enrichment of
   these transduction pathways were previously noted in the spleen of
   KHV-infected carps [[264]8] and in orange-spotted grouper (Epinephelus
   coioides) infected with Singapore grouper iridovirus (SGIV) [[265]65].
   The activation of the MAPK signalling pathway contributes to
   pro-inflammatory cytokine production, and it has been linked to
   enhanced virus replication [[266]66]. We also observed that the
   glutathione metabolism and the PPAR signalling pathway decreased in
   KHV-exposed leukocytes. In mammals, M1 macrophages are typically
   characterised by glycolytic metabolism and the expression of inducible
   nitric oxide synthase (iNOS), reactive nitrogen intermediates, and
   reactive oxygen species (ROS), promoting Th1 responses and strong
   microbicidal activity. Conversely, M2 macrophages are characterised by
   increased phagocytic activity and mitochondrial respiration, with a
   high expression of scavenging, mannose, and galactose receptors through
   the arginase pathway [[267]67]. Interestingly, our data show that
   KHV-exposed leukocytes exhibited increased mitochondrial respiration
   through activation of the OXPHOS pathway, low production of ROS
   compared with naïve leukocytes, low glutathione metabolism, and the
   up-regulation of arginase non-hepatic 1-like, macrophage mannose
   receptor-1, and scavenger receptors. We also observed up-regulation of
   the gene expression of crem and il-13 receptors and down-regulation of
   ifng1. In fish, macrophage polarisation towards M1- and M2-like states
   has been previously demonstrated [[268]68,[269]69]. In common carp,
   macrophage polarization of kidney-derived macrophages showed
   differential activity of inducible NO synthase and arginase depending
   on the type of stimulation [[270]70]. In particular, carp macrophage
   polarisation into an M2-type phenotype was driven by il-4/13 or cAMP
   stimulation [[271]42].

   Interestingly, we found an up-regulation of il-10 in KHV-exposed
   leukocytes. IL10 is a typical marker of M2 macrophages [[272]64], and
   in carp, it has been shown to promote anti-inflammatory activities on
   phagocytes [[273]71]. We also observed a robust up-regulation of
   transglutaminase proteins, a conserved marker for mammalian M2
   macrophages [[274]72], which has been previously found to be expressed
   in carp M2 macrophages [[275]42]. Among the cytokines, we observed the
   gene up-regulation of both pro-inflammatory cytokines, such as
   tnf-like, and anti-inflammatory cytokines and chemokines, such as tgfb1
   and il-10, characteristic of an M2 type [[276]64]. In addition, we
   observed a strong up-regulation of cxcl8/il8, which has been shown to
   attract neutrophils in carp [[277]73], and ccl19a.1, a chemokine that
   attracts T-cells and macrophages and that has been linked to carp
   susceptibility to KHV [[278]14]. Our dataset also showed significant
   changes in the regulation of iron metabolism in carp leukocytes. In
   particular, we observed the down-regulation of ferritin-related genes,
   which relates to low iron storage, and high expression of ferroportin,
   which suggests high iron export, suggesting that KHV-exposed leukocytes
   are set in an iron-export mode which correlates with typical M2
   polarised macrophages [[279]67]. Based on the current knowledge on
   herpesviruses, the balance of M1 and M2 macrophages play a role in
   viral clearance and latency [[280]74]. Overall, our dataset suggests
   both pro- and anti-inflammatory responses in KHV-infected carps. This
   finding and its impact on KHV pathogenicity require further
   investigation.

   Evidence of altered epigenetic and transcriptional regulation in carp
   leukocytes. Interestingly, we observed a strong suppression of
   epigenetics and regulation of transcription pathways, such as histone
   H3-K4 methylation. Recent studies have shown the important role of
   histone-modifying enzymes, particularly histone acetyltransferases
   (HATS) and histone methyltransferases (HMTs), in regulating macrophage
   function. However, the exact role of epigenetic regulation on
   inflammatory processes is not yet well understood [[281]75].
   Remarkably, the inhibition of H3-K4 methylation has been shown to
   compromise immune responses in murine macrophages [[282]76]. Taken
   together, the down-regulation of histone H3-K4 methylation and other
   critical epigenetic regulation in KHV-exposed leukocytes might suggest
   an immunocompromised response [[283]77]. KHV establishes a long-lasting
   life infection through latency with reactivation episodes and the
   appearance of secondary infections [[284]7]. The disruption of
   epigenetic regulation may have contributed towards immunosuppression
   associated with chronic KHV infection in carp. In grass carp,
   epigenetic mechanisms such as DNA methylation have been linked to
   immune responses and carp susceptibility to viral infection [[285]78].
   The role of epigenetic regulation on carp inflammatory responses
   requires further research.

   High phagocytosis but low autophagy of KHV-infected leukocytes might
   support KHV latency. We observed a strong regulation of phagocytosis
   supported by a hyperactive rearrangement of the actin cytoskeleton,
   extracellular matrix, and numerous molecular pathways denoting high
   macrophage mobility, with up-regulation of Rho GTPase activator
   enzymes, G-protein-coupled receptors, and protein phosphatases
   [[286]79]. Among the carp genes highly up-regulated were allograft
   inflammatory factor 1-like, which has been associated with enhanced
   macrophage phagocytic activity in mice [[287]80], the macrophage
   mannose receptor 1-like, previously proposed as a marker of carp M2
   macrophages [[288]42], numerous septin genes involved in molecular
   pathways of the leukocyte’s cytoskeleton network, and several isoforms
   of the tetraspanin cd9 which play an essential role in activating the
   leukocyte’s motility and adhesion [[289]81].

   Our dataset also showed molecular evidence of low autophagy activity in
   KHV-exposed leukocytes, suggesting poor clearance of phagocytised viral
   particles. Indeed, the activation of autophagy has been observed in
   resistant carp families to KHVD [[290]11]. Several mammal herpesviruses
   have developed strategies to escape the autophagy of immune cells
   either by expressing specific anti-autophagic proteins or hijacking
   autophagy [[291]82]. Recent studies suggest a role of the complement
   system in autophagy [[292]83]. In the present study, the activation of
   complement pathways was not observed. An increase in complement
   haemolytic activity in blood plasma [[293]62] as well as C-reactive
   proteins in infected tissues such as the liver, spleen, and kidney
   [[294]84] have been described in KHV infection. However, the gene
   expression of c3, a central component of the complement system, was not
   observed at mucosal sites, such as gills and skin [[295]9]. Similarly,
   cyprinid herpesvirus 2 (CyHV2) infection in gibel carp (Carassius
   auratus gibelio) induced the up-regulation of c3 transcription in
   internal organs but not at mucosal sites [[296]85].The role of the
   complement system and autophagy in clearing KHV infection requires
   further research.

   Although this study has provided valuable data on carp immune responses
   against KHV infection, the mechanisms by which KHV infects and
   establishes latency in blood leukocytes, mainly B cells, are still to
   be elucidated. Our RNA-Seq sequencing approach used pooled isolated
   leucocytes and therefore did not provide data on transcription for
   individual cell types. Additionally, the risk of some contamination
   (with erythrocytes, for example) during the process of isolation of
   leucocytes cannot be excluded. Future work on identifying specific
   subsets of blood leukocytes in KHV-infected carp should involve either
   cell sorting if antibodies were available or single-cell sequencing
   [[297]86]. To facilitate this, we suggest genes that could be developed
   to be used as cellular markers for further characterisation and cell
   sorting, such as cd83, cd59, and cd44. In mammals, CD83 is induced
   explicitly in IL-4-stimulated M2-type macrophages [[298]87], CD44 has
   been shown to promote the accumulation of M2 macrophages [[299]88], and
   CD59 prevents complement activation by inhibiting the formation of the
   membrane attack complex [[300]88] These markers require validation
   prior to use in cell sorting studies. Interestingly, our data also
   showed down-regulation of cytotoxic T cell markers, such as cd8, cd5,
   and cd3, which suggests a T cell dysregulation in KHV-infected carps
   likely mediated by the up-regulation of il10, an anti-inflammatory
   cytokine [[301]71], which could contribute to the lack of clearance of
   KHV. The role of cytotoxic T cells in KHV chronic infection and latency
   requires further research.

5. Conclusions

   This study describes the molecular responses to KHV in carp blood
   leucocytes following infection at a relatively low temperature (17 °C).
   We identified several PRRs and cytokines involved in KHV recognition
   and immune signalling. We suggested candidate genes that can be used
   for further characterisation of carp leukocytes following viral
   infection. We also identified molecular pathways supporting macrophage
   polarisation in blood leukocytes exposed to KHV, characterised by high
   phagocytosis, low autophagy, and increased iron export. Further work is
   required to fully characterise the response to KHV in different
   leukocyte subsets and the molecular events leading to the establishment
   of KHV latency. Overall, this work contributes to a better
   understanding of the innate immune response to herpesvirus infection in
   cyprinids. Our results point towards new research priorities on
   understanding the response to KHV at the single-cell level and open new
   avenues for the development of treatments in the future.

Acknowledgments