Single-cell atlas of rainbow trout peripheral blood leukocytes and profiling of their early response to infectious pancreatic necrosis virus

The recent development of single cell sequencing technologies has revolutionized the state-of-art of cell biology, allowing the simultaneous measurement of thousands of genes in single cells. This technology has been applied to study the transcriptome of single cells in homeostasis and also in response to pathogenic exposure, greatly increasing our knowledge of the immune response to infectious agents. Yet the number of these studies performed in aquacultured fish species is still very limited. Thus, in the current study, we have used the 10x Genomics single cell RNA sequencing technology to study the response of rainbow trout (Oncorhynchus mykiss) peripheral blood leukocytes (PBLs) to infectious pancreatic necrosis virus (IPNV), an important trout pathogen. The study allowed us to obtain a transcriptomic profile of 12 transcriptionally distinct leukocyte cell subpopulations that included four different subsets of B cells, T cells, monocytes, two populations of dendritic-like cells (DCs), hematopoietic progenitor cells, non-specific cytotoxic cells (NCC), neutrophils and thrombocytes. The transcriptional pattern of these leukocyte subpopulations was compared in PBL cultures that had been exposed in vitro to IPNV for 24 h and mock-infected cultures. Our results revealed that monocytes and neutrophils showed the highest number of upregulated protein-coding genes in response to IPNV. Interestingly, IgM+IgD+ and IgT+ B cells also upregulated an important number of genes to the virus, but a much fainter response was observed in ccl4 + or plasma-like cells (irf4 + cells). A substantial number of protein-coding genes and genes coding for ribosomal proteins were also transcriptionally upregulated in response to IPNV in T cells and thrombocytes. Interestingly, although genes coding for ribosomal proteins were regulated in all affected PBL subpopulations, the number of such genes transcriptionally regulated was higher in IgM+IgD+ and IgT+ B cells. A further analysis dissected which of the regulated genes were common and which were specific to the different cell clusters, identifying eight genes that were transcriptionally upregulated in all the affected groups. The data provided constitutes a comprehensive transcriptional perspective of how the different leukocyte populations present in blood respond to an early viral encounter in fish.


Introduction
Teleost fish have a complex immune system with both innate and adaptive branches.For years, many of the advances made in fish immunology have been based on the identification of immune genes and analysis of their transcriptomic responses.These genes include homologues to multiple cytokines and their receptors, different classes of pattern recognition receptors (PPRs), immunoglobulins (Igs), T cell receptors (TCRs), major histocompatibility complexes (MHC I and MHC II), some specific markers of lymphocyte populations such as CD4 and CD8, and many transcriptional factors, among others, hence demonstrating an overall conservation of the fundamental components of the innate and adaptive immune system between teleost fish and other vertebrates (1).Despite this, the identification and characterization of leukocyte subsets has been a much harder task in most fish species, because in some cases the range of surface markers that define these subpopulations in mammals are not that well conserved in fish, and also due to difficulties in generating specific monoclonal antibodies against these markers when identified.
The explosion of the single-cell mRNA sequencing technology in recent years has drastically increased our capacity to analyze the transcriptome from thousands of individual cells, helping to the identification and characterization of specific immune populations at a transcriptional level in physiologic and pathogenic conditions without requiring their labeling with specific antibodies.These single-cell RNA sequencing technologies have been used in different species to dissect cellular heterogeneity in complex systems (2), including several fish species.For example, novel insights on T cell and natural killer (NK)-like cell functionality were revealed by single cell transcriptomics in zebrafish (Danio rerio) (3).Also in zebrafish, single-cell RNA sequencing was used to define cell heterogeneity within the kidney, revealing the existence of novel cell types including two types of NK-like cells and different hematopoietic progenitor subsets (4).Similarly, innate lymphoid cells (ILCs) have been identified and characterized in zebrafish following this methodology (5).In Atlantic cod (Gadus morhua), a species that lacks the MHC II system, several cell subpopulations were transcriptionally characterized among splenocytes and blood leukocytes, including cytotoxic T cells, B cells, erythrocytes, thrombocytes, neutrophils, and macrophages (6).Different subsets of non-specific cytotoxic cells (NCCs) were also identified in the Nile tilapia (Orepchromis niloticus) kidney using this technology (7).Single-nuclei RNA sequencing was similarly used to characterize different cell lineages from Atlantic salmon (Salmo salar) liver after challenge with Aeromonas salmonicida, identifying a subpopulation of defense-specialized hepatocytes and detecting the up-regulation of diverse immune genes (8).Regarding rainbow trout (Oncorhynchus mykiss), our group recently applied single cell transcriptomics to characterize peripheral blood B cells, recognizing several B cell clusters with specific gene profiles (9).Additionally, this technology was also used to establish that rainbow trout single B cells have the capacity to transcribe several immunoglobulin light chains (IgL) of different specificity (10).
Single-cell RNA sequencing technologies have also been widely applied to investigate the response to infectious diseases, increasing our knowledge on how different cell subsets specifically respond to pathogens.Thus, using this technology to investigate the response of complex cell populations to viruses has allowed the description of novel cell subpopulations involved in infectious responses; the determination of susceptible cell types and infection dynamics; the recognition of differentially expressed genes during the infection; or the discovery of biomarkers for infectious diseases.For example, knowing that CD4 + T cell permissiveness to Human Immunodeficiency Virus (HIV) infection is highly heterogeneous across individuals, single cell RNA-seq was applied to define markers of CD4 + T cell permissiveness (11).Similarly, a study performed with peripheral blood leukocytes (PBLs) from Dengue virus-infected patients identified Mx2 in naïve B cells and CD163 and IFIT1 in monocytes as predictive markers of disease onset (12).In another study, single-cell RNA sequencing was used to characterize the differentiation of plasmacytoid DCs during a cytomegalovirus infection in mice (13).Other studies have used this methodology to study virus variability such as that in which viral quasispecies were defined during hepatitis C virus (HCV) infection (14).Concerning fish, a single cell transcriptome of midbrain cells infected with Red spotted grouper nervous necrosis virus (RGNNV) performed in orange-spotted grouper (Epinephelus coiodes) revealed that the brain was enriched in macrophages in RGNNV-infected fish, and described the transcriptional profile of these macrophages in response to the virus.Interestingly, further analysis revealed that, upon infection, the microglia transformed into M1-type activated macrophages that produced cytokines to help reduce the damage caused by the virus in this tissue (15).Different leukocyte subsets were also identified in the zebrafish spleen, which showed differential responses to spring viremia of carp virus (SVCV) infection (1).Single cell transcriptomics was also used to study the response of the SHK-1 cell line derived from Atlantic salmon head kidney to infectious salmon anemia virus (ISAV) (16).
Infectious pancreatic necrosis virus (IPNV) is a non-enveloped double stranded RNA (dsRNA) virus belonging to the genus Aquabirnavirus within the family Birnaviridae.IPNV can infect a wide range of salmonid species, including Atlantic salmon and rainbow trout, thus causing huge economic losses to the global salmonid industry every year (17).Although the percentage of blood leukocytes that supports viral replication in newly infected fish is low, the virus seems strongly associated with blood leukocytes in fish persistently infected with the virus, in which a carrier state is established (18).In this context, IPNV seemed like an adequate model to investigate the early immune responses of rainbow trout peripheral blood leukocytes (PBLs) to a viral encounter.To this aim, trout PBLs were exposed in vitro to the virus or mock-infected.After 24 h of incubation, a single cell transcriptomic analysis was undertaken using the 10x Genomics technology previously applied to fish cells (9,10).Based on transcript heterogeneity and expression of known markers, PBLs were classified into 12 different clusters that included four different subsets of B cells, T cells, monocytes, two populations of dendritic-like cells (DCs), hematopoietic progenitor cells, NCCs, neutrophils and thrombocytes.Further differences in the expression of proteincoding genes, long non-coding RNAs, ribosomal and mitochondrial genes were described for each subpopulation providing us with an atlas of blood leukocyte subsets in rainbow trout.When the transcriptional profile of these PBL subpopulations was compared between IPNV-infected and mock-infected cultures, we established that monocytes, neutrophils, IgM + IgD + and IgT + B cells, T cells and thrombocytes were the main responding cell types.An analysis of common and specific regulated genes proteins identified a set of viral-regulated genes involved in different cellular functions such as antigen presentation, effector processes, cell surface receptor signaling, phagocytosis or regulation of gene expression, that are thoroughly discussed throughout the paper.These results contribute to a better understanding of how different innate and adaptive immune populations respond to viral encounter in teleost fish, providing us with useful information for the development of prophylactic measures in the future.

Isolation of PBLs from rainbow trout
Rainbow trout (Oncorhynchus mykiss) of approximately 70-100 g were obtained from Piscifactorı ́a Cifuentes (Guadalajara, Spain).Fish were maintained at the Animal Health Research Centre (CISA-INIA) laboratory at 14°C in a re-circulating water system with 12:12 h light:dark photoperiod.Fish were fed twice a day with a commercial diet (Skretting, Spain).Prior to sampling, fish were acclimatized to laboratory conditions for 2 weeks and during this period no clinical signs were ever observed.At this point, two rainbow trout were killed by benzocaine (Sigma) overdose.Blood was extracted with a heparinized needle from the caudal vein and diluted 10 times with Leibovitz medium (L-15, Thermo Fisher Scientific) supplemented with 100 IU/ml penicillin and 100 mg/ml streptomycin (P/S, Thermo Fisher Scientific), 5% fetal calf serum (FCS, Thermo Fisher Scientific) and 10 IU/ml heparin (Sigma).PBLs were obtained by centrifugation (500 x g for 30 min at 4°C) of diluted blood on 51% continuous Percoll (GE Healthcare) density gradients.The interface cells were collected, washed twice in L-15 containing antibiotics and 5% FCS and adjusted to 1 x 10 6 cells/ml.Dye exclusion test using trypan blue (Sigma) was used to determine the number of viable cells.

In vitro infection of PBLs with IPNV
IPNV (Sp strain, ATCC VR 1318) was propagated in the RTG-2 (Rainbow Trout Gonad-2) cell line, an established line of fibroblasts from rainbow trout gonads (19).RTG-2 cells were routinely grown in 75 cm 2 culture flasks (ThermoFisher Scientific) at 19°C, and split 1:2 when confluent, after detaching the cells with 0.25% trypsin-EDTA in PBS (Phosphate Buffer Saline) (Gibco).To produce viral stocks, complete culture media was removed from 80% confluent cell cultures in 75 cm 2 culture flasks and IPNV (200 µl of 1x10 7 TCID 50 /ml) inoculated.Immediately after, L-15 medium with antibiotics and 2% FCS was added and the cells cultured at 14°C for approximately 5-7 days.When cytopathic effect was extensive, the supernatant was harvested and centrifuged to eliminate cell debris (2000 x g for 15 min at 4°C).Clarified supernatants were used for the experiments.All virus stocks were titrated in 96-well plates according to the procedure described by Reed and Müench (20).
For the in vitro infection of rainbow trout PBLs, cells resuspended in L-15 containing antibiotics and 5% FCS were disposed in 24-well plates (500 µl per well).Immediately after, 10 µl of IPNV (5 x 10 7 TCID 50 /ml) were added to each well, to achieve a multiplicity of infection (MOI) of 1.At this point, the cells were incubated at 18°C for 24 h, before they were harvested to perform the single-cell transcriptomic analysis.Non-infected control cells from each fish were included and treated in the same conditions.

Construction of the 5´single cell library and sequencing
After the 24 h of culture, cell viability was checked on a FACS Celesta ™ flow cytometer (BD Biosciences) after staining the cells with DAPI (0.2 mg/mL; Sigma-Aldrich).After confirming cell viabilities of approximately 85%, cells gently pipetted and diluted in L-15 medium supplemented with antibiotics and 5% FCS to a concentration of 1,000 cells/µl were used for cell isolation on a 10x Genomics Chromium Controller instrument.Cell suspensions from each sample were loaded into the chips of the Chromium ™ Single Cell 5′ Gel Beads Kit (10x Genomics) and subjected to the Chromium Controller instrument to generate single cell Gel Bead-In Emulsions (GEMs) following manufacturer's instructions.Next, GEMs were subjected to library construction using the Chromium ™ Single Cell 5′ Library Kit v1 (10x Genomics).As a first step, reverse transcription was performed, resulting in cDNA tagged with a cell-specific barcode and unique molecular index (UMI) per transcript.Fragments were then size selected using SPRIselect magnetic beads (Beckman Coulter).Next, Illumina sequencing adapters were ligated to the size-selected fragments and cleaned up using SPRIselect magnetic beads (Beckman Coulter).Finally, sample indices were selected and amplified, followed by a double sided size selection using SPRIselect magnetic beads (Beckman Coulter).The quality of the libraries was assessed using an Agilent 2100 Bioanalyzer (Agilent technologies) and samples were then sequenced using a NextSeq instrument (Illumina) with 150PE chemistry.

Alignment and initial processing of sequencing data
The Cell Ranger software (10x Genomics, v3.1) was used to process the sequenced libraries.The complementary DNA reads from each sample were mapped against a previously adapted Oncorhynchus mykiss reference transcriptome (9) using the "Cell Ranger count" tool.Through this system, filtered UMI expression matrices from each sample were generated.As a result, raw expression data was obtained containing transcriptomes for individual cells, both for control non-infected leukocytes and for leukocytes exposed to IPNV during 24 h, from each of the two fish included in analysis.

Quality filtration of cells
A quality control of datasets was performed in order to filter out contaminants such as abnormal cells in all datasets.For this purpose, filtered UMI expression matrices were conducted to Seurat package (v3.1) for successive analysis.Cells with at least 200 detected genes, and only those genes that appeared in at least three cells were included in an initial matrix for each fish.A cell was considered to be abnormal if any of the following criteria were met: (i) detected gene number >3,500; (ii) detected count number >15,000 or (iii) >25% of reads in a cell mapped to mitochondrial genes.Additionally, cells containing >1% of reads mapped to hemoglobin genes were considered contaminant red blood cells or doublets involving these undesired cells, thus cells achieving this threshold were also filtered out at this step.

Sample integration and data reduction analysis
The SCTransform method from the Seurat software was applied in order to normalize the four filtered single-cell datasets that were then integrated using the PrepSCTIntegration tool to avoid a batch effect between samples.The merged data was subjected to dimensionality reduction using the principal component analysis (PCA) followed by uniform manifold approximation and projection (UMAP) using 35 dimensions.

Marker identification and functional analysis
The identification of genes showing differential expression associated to a specific cluster was performed using the Findcluster tool from the Seurat software, considering a significant association for those genes showing an adjusted p < 0.001 and log2FC ≥ 0.25.The information relative to the gene description contained in the O. mykiss genome v1.0 Omyk_1.0 (GCF_002163495.1)was taken into account for gene name association.In order to obtain an actualized functional annotation, the nucleotide sequences from the genes identified as markers were compared with proteins from a set of model species (Homo sapiens, Mus musculus, Danio rerio, Macata mulata, Drosophila melanogaster, and Xenopus tropicalis) using the Blastx software applying as threshold a minimum E value of 10 -5 .Blast results were subjected to the Blast2GO software for GO term mapping.Sequences were also compared against domain databases using the InterProScan tool implemented in Blast2GO.GO term annotations were inferred for rainbow trout transcripts.Single enrichment analysis was performed by comparing the functions associated to genes from each cluster taking into account differences with an adjusted p < 0.05.

Differential gene expression analysis of cell subsets in response to IPNV interaction
The differential expression analysis between leukocytes exposed to IPNV and non-infected cells was conducted globally and within each cell subset using the MAST package and the FindMarkers function from the Seurat package (version 3.1.5).Multiple tests corrections were applied to the p values to control the FDR using the procedure described by Benjamini and Hochberg (21), as implemented in MAST.

Single-cell transcriptome sequencing of rainbow trout PBLs
To explore the heterogeneity of rainbow trout PBLs and their responses to a virus infection, a transcriptomic profile at single-cell resolution was constructed for all isolated cells, both control and IPNV-infected cells from each fish (Figure 1A).Single cell cDNA libraries were sequenced using Illumina HiSeq 150PE obtaining a total number of 145,325,945 and 166,384,599 raw reads for control and IPNV-infected PBLs from fish 1, respectively, as well as 287,908,169 and 284,601,127 for control and IPNV-infected PBLs from fish 2, respectively.Using the Cell Ranger software around 75-85% of raw reads per sample were maintained after the identification of valid cell barcodes, that when mapped to the transcriptome resulted in approximately 65-75% of mapped reads.Filtered data from Cell Ranger was loaded to the Seurat software (v3.0) for successive analysis.Finally, cells showing more than 1% of reads mapping to hemoglobin genes were excluded as potential contaminant erythrocytes.After filtering, all datasets showed a similar distribution (Figure 1B).All datasets were integrated together using the SCTransform function, and the resulting matrix was dimensionally reduced by applying PCA, followed by UMAP.The cell projection reflected a homogeneous separation of cells in the four samples included in the analysis (Figure 1C).

Clustering and assignment of cell identities
The application of the FindCluster algorithm from Seurat software resulted in the detection of 9 different clusters of cells showing clear differences in gene expression patterns.Meanwhile, analyzing UMAP projection, a set of 88 cells were catalogued as doublets and filtered out at this step due to discordances between cluster assignment and location in cell projection.After this last filtering step, a total of 3,108, 3,724, 3,584 and 4,167 cells were retained for successive analysis in the samples from fish 1 (control and IPNV-infected) and fish 2 (control and IPNV-infected), respectively.Three small groups of cells, which contained cells from the four samples analyzed, were found to clearly separate in cell projection.Thus, these three groups, which included irf4 + B cells, dendritic cells (DCs), and DC-like cells, were manually defined as new clusters achieving a total of 12 final clusters (Figure 2A).These clusters were visualized both in control and infected samples (Supplementary Figure S1).
Once the main cell clusters were defined, a global identification of differentially expressed genes for each cell cluster was performed.Differentially expressed genes were analyzed to identify known markers commonly associated to different cell types which allowed the assignation of putative identities (Figure 2B, Supplementary Table S1).Four clusters were clearly associated with B cells according to the expression of the constant regions of Ig heavy (H) and light (L) chains as well as different genes encoding CD79 (Figure 2B).The first cluster of B cells is preferentially represented by IgM + IgD + B cells (known to be the main subset of B cells in systemic compartments), whereas the second one is mainly represented by IgT + B cells.The third group of B cells identified also preferentially contained IgM + IgD + B cells, but with a differential expression pattern defined mainly by the expression of ccl4, a chemokine previously defined as a marker for a specific subset of B cell in salmonids (9).Finally, a reduced group of 30 cells was identified as plasma-like cells based on the reduced expression of IgD, the increased expression of IgM and the higher expression of irf4, a common marker of plasma cells (22).Only one cluster was clearly identified as a T cell population according to the expression of the constant region for TCRa and other common markers such as CD3d, CD3e, CD4 or ZAP-70 (Figure 2B).Three clusters seemed to correspond to myeloid cells according to the expression of different genes annotated as CD209.The biggest of these groups, represented by ~5% of total cells, was identified as monocytes attending to the expression of CD209 and lysozyme II (lyz2) genes, common markers for this subset in several species (23) (Figure 2B).The other two clusters were represented by only 26 and 22 cells.The first one was annotated as a DC cluster, based on the higher expression of different genes encoding components of the MHC II complex (LOC11049409, oncmyk-dab, oncmyk-dbb) and an homologue of the zinc finger protein 366 (LOC110536334), a DC-specific marker in mammals (DC-SCRIPT) (24) (Figure 2B).The last one of these small myeloid groups represents a rare population of cells similar to the DC population, which also transcribes other genes encoding CD209 antigen-like protein C (LOC110533211) and E (LOC110534781) (Figure 2B).A small cell population of approximately 2% of total cells was defined as hematopoietic progenitor cells (HPCs) attending to the specific expression of an FLT3-like gene, also known as CD135, a common marker of this cell type in mice and human (25).Interestingly, Single-cell analysis of PBLs infected or not in vitro with IPNV for 24 h and cell projection.(A) Workflow diagram.(B) Quality check of single-cell results.Each two rows correspond to PBL samples from one fish, being odd rows the control samples and the even, the infected ones.On each column, one characteristic is represented: number of genes, number of reads, percentage of reads mapping ribosomal genes, and percentage of reads mapping mitochondrial genes.(C) UMAP visualization of the cells from the four samples after the integration.several cells within this population also express CD209 antigen-like protein C and E, which may indicate they preferentially correspond to myeloid progenitor cells.In fact, it has been shown that in mammals most lymphoid tissue DCs derive from Flt3 + progenitors whereas a few derive from Flt3 − monocytes and these DCs seem to maintain Flt3 in the steady-state (25).Therefore, it is difficult to unequivocally establish whether these cells correspond to a progenitor cell type or to yet another DC subset, but because they are clustered independently from other DC subtypes and given they lack some other DC markers, in this paper, we will refer to them as HPCs.The identity of the last three clusters identified during the analysis was easily assigned according to differentially expressed genes.The first one was identified as NCCs, considered as evolutionary precursors of natural killer cells (NK) in teleost (7), defined by the expression of perforin-1-like gene (LOC110531658).
The second one was clearly identified as a neutrophil population, attending to the expression of an eosinophil peroxidase-like gene (LOC110509901), previously associated to fish neutrophils (6).The final of these clusters is a highly represented one, corresponding to approximately 25% of total cells.The cells in this cluster corresponded to thrombocytes according to the expression of platelet glycoprotein Ib beta chain-like gene (LOC110536200) (Figure 2B).Once the different clusters were defined, an enrichment analysis was performed to identify functional Clustering of integrated data and group identities.annotations associated with each PBL population (Figure 2C, Supplementary Figure S2, Supplementary Table S2).

Identification of novel markers and functions associated to rainbow trout PBL populations
In addition to the well-known markers previously mentioned, clear differences among rainbow trout PBL populations were detected during the analysis, both in the expression of coding genes and long non-coding RNAs (Supplementary Table S1) and in the enrichment of specific features or functions (Figure 2C, Supplementary Figure S2, Supplementary Table S2).For instance, monocytes showed higher values in total number of reads and in total number of expressed genes than other populations (Supplementary Figures S3A, B).On the other hand, both T and B cells show a clear enrichment in the percentage of reads mapping to ribosomal proteins (Supplementary Figure S3C), which also results in a higher expression of these genes, identifying several of them as markers in these populations (Supplementary Table S1).The expression of mitochondrial genes also showed an irregular distribution, with a higher percentage of reads mapping these genes in monocytes, DC-like cells, HPCs and neutrophils (Supplementary Figure S3D).The differential expression analysis also highlighted a set of coding genes and long non-coding RNAs which effectively differentiate at a transcriptional level among PBL populations (Tables 1, 2).These genes were differential taking into account transcripts from control and infected samples altogether, and also in control samples exclusively.
Analyzing B cells globally (the 4 clusters), the expression of a gene encoding a phospholipase A and acyltransferase 1-like (LOC110505920) was identified in approximately 85% of these cells and was only identified in 0.04% of other leukocyte subsets.Also, transcriptional activity of zinc finger proteins 34-like (LOC110500012) and 35-like (LOC110490446) was identified in 53 and 66% of B cells, respectively, whereas it was only expressed in 0.05-0.06% of other PBL populations (Table 1).In addition, an uncharacterized gene (LOC110533868), that codes for a protein that shows similarities with a domain from the ectropic viral integration site 2A protein (EVI2A), was also identified as preferentially expressed in B cells, both in IgM + IgD + and IgT + B cells.Specific markers differentiating B cell subpopulations were also identified.For instance, ~83% of cells included in the IgT + B cell cluster express a prosaposin-like coding gene (LOC110509903).Finally, the irf4 + subpopulation identified as a plasma-like population showed differential expression of several enzymes, s u c h a s d o l i c h y l -d i p h o s p h o o l i g o s a c c h a r i d e -p r o t e i n glycosyltransferase 48 kDa subunit-like (LOC110492237), protein disulfide-isomerase A6-like (LOC110529378) and A4-like (LOC110489602), expressed by approximately 90% of the cells from this group but only expressed in a 0.07-0.09% of cells from all other PBL subsets, including other B cell populations (Table 1).Different long non-coding RNAs also seemed strongly associated with this plasma-like cell subpopulation, such as LOC110499933 and LOC110526128, which are expressed in ~87% and 60% of cells, respectively (Table 2).
Concerning T cells, the expression of a gene encoding the protein S100-A5-like (LOC110520885) was observed in 82% of these cells at very high transcript levels.Other genes with expression highly associated with T cells encode an SH2 domain-containing protein 1A-like (LOC110530635), a PLAC8-like protein 1 (LOC110489105) or a granzyme K-like (LOC110523798) identified in at least 62% of T cells but transcribed at very low levels by other cells subsets (Table 1).Surprisingly, the percentage of cells expressing some common markers used for identity assignment were only identified in 30-50% of cells.Hence, for example, the transcription for the TCRa constant region was observed only in ~31% of cells and ZAP70 in ~47% of cells (Supplementary Table S1).The transcription of other known T cell markers such as CD8 or TCRb was practically residual (Supplementary Table S1).
Cells from the HPC population defined by the differential expression of an FLT3-like gene as previously mentioned, widely transcribed other interesting genes such as two genes encoding beta-galactoside-binding lectins (leg and LOC110512958), three genes encoding tetraspanin-8-like proteins (LOC110499719, LO C 11 04 9 97 18 an d L O C1 10 49 97 1 7) , a l o w a ffi n i t y immunoglobulin gamma Fc region receptor II-like protein (LOC110515346) and an osteoclast stimulatory transmembrane protein-like (LOC110505982) (Table 1, Supplementary Table S1).Additionally, cells included in this leukocyte population were also characterized by the transcription of one long non-coding RNA, LOC110524566, whose transcription was observed in 100% of cells while residually expressed in other PBL populations (0.3% of cells) (Table 2).
The gene encoding for the macrophage expressed 1 (mpeg1) was the most predominantly transcribed in DCs (Table 1).Other characteristic gene products from this PBL population were the m y r i s t o y l a t e d a l a n i n e -r i c h C -k i n a s e s u b s t r a t e -l i k e (LOC110529520), two ictacalcin-like (LOC110536275 and LOC110520890), and one uncharacterized gene (LOC110523838).In the other population of DC-like cells, genes encoding reductase subunits (LOC110529451 and rrm1) stood up, together with genes mapping topoisomerase DNA II alpha (top2a) or tubulin beta-1 chain (LOC110535859) (Table 1).
The genes that most distinguished neutrophils from the rest were those encoding choriolytic enzymes (hce23 and LOC110526217) and complement factor D (cfd), while the most enriched gene products in thrombocytes were integrins (LOC110491676 and LOC110538030) and the coagulation factor XIII A chain-like (LOC110500457) (Table 1).After having defined the transcriptional profile of the different leukocyte populations in rainbow trout blood, we proceeded to analyze the differential expression between control and IPNV infected cells, which allowed the identification of several genes that significantly altered their transcript levels in response to viral exposure.A first analysis was conducted treating the expression data like a bulk RNA-seq (without differentiating among leukocyte subsets) with an adjusted p value < 0.001 and an average log2FC > 0.25.This initial analysis identified a total of 497 transcripts that were significantly modified their transcript levels in response to IPNV (Figure 3A, Supplementary Table S3).Among them, approximately 67% corresponded with protein-coding transcripts whereas approximately 29%, 3% and 1.6% represent ribosomal proteins, long non-coding RNAs and mitochondrial genes,  3B).This distribution was also observed in the protein-protein interaction network obtained (Figure 3C), where a condensed cluster (red nodes) corresponds to different components of ribosomes together with other proteins mainly annotated as elongation factors or proteasome subunits.Another cluster, represented by light blue nodes, grouped the different mitochondrial genes together with nuclear genes involved in mitochondrial functions, for instance ATP-synthases or cytochrome oxidase subunits.The dark blue cluster grouped proteins associated with immune response related to functions like antigen processing and presentation (mhc I and mhc II) and also genes involved in immune effector processes (stat1, ifi44 or ptpn6, among others).The green cluster, also involved in the immune response, grouped several genes involved in cell surface receptors signaling pathways (f.i.stat3, irf1, ccr9, ilr6r or lcp1) as well as a set of genes involved in phagocytosis including several family members of actin-related proteins.Finally, the yellow cluster joined several proteins involved in the regulation of gene expression as well as other metabolic processes (Figure 3C).In a second step, the differential expression was studied within different leukocyte subpopulations.These studies rendered some interesting results concerning both shared and specific molecular responses.Monocytes and neutrophils showed the highest number of upregulated protein-coding genes in response to IPNV, with 267 and 233, respectively (Figure 4A).Both IgM + IgD + and IgT + B cells also experienced an important upregulation of genes (152 and 154,

Differential gene expression between IPNV-infected and non-infected PBL cultures. (A) Heatmap comparing gene expression between the control and IPNV-infected cells. (B)
Percentage of long non-coding RNAs, protein coding transcripts and genes mapping ribosomal proteins or mitochondrial DNA among differentially regulated genes.(C) Protein-protein interaction network of the whole data.Red: ribosomal genes, light blue: mitochondrial genes, dark blue & green: immune response genes, and yellow: metabolic regulation genes.
respectively) in response to virus exposure (Figure 4A), whereas the B cells defined by ccl4 transcription only up-regulated 13 proteincoding genes in response to IPNV.In T cells and thrombocytes, also a moderate number of protein-coding genes (94 and 112, respectively) experienced upregulation upon the in vitro infection (Figure 4A).The number of down-regulated genes in response to IPNV in all these populations was much lower, with no genes down-regulated in ccl4 + B cells (Figure 4A).Interestingly, plasmalike cells (irf4 + B cells), DC-like cell populations, NCC and HPCs did not experience regulation of any or a reduced number of genes in response to the virus within the established threshold (Figure 4A).Remarkably, the populations that up-regulated the higher number of protein-coding genes, also experienced important changes in the transcription levels of a high number of ribosomal proteins in response to the virus, with IgM + IgD + and IgT + B cells experiencing changes in a higher number of genes (Figure 4A).Finally, some long non-coding RNAs were also modulated in response to IPNV in these same populations (Figure 4A).
A functional enrichment analysis of the genes modulated in response to the virus pointed to several functionalities within upregulated genes identified in the PBL populations that were mainly affected by IPNV (Figure 4B, Supplementary Table S4).Interestingly, genes related to response to cytokines (GO:0034097) are among the most regulated genes for all the cell types mentioned above.Additionally, the genes regulated in both IgM + IgD + and IgT + B cells share some functions, such as peptide metabolic process (GO:0006518), amide biosynthetic process (GO:0043604) and positive regulation of gene expression (GO:0010628).Nonetheless, IgT + B cells show a high expression of genes related to myeloid leukocyte mediated immunity (GO:0002444) or myeloid cell activation involved in immune response (GO:0002275).On the other hand, some of the genes regulated in ccl4 + B cells are connected to the natural killer cell chemotaxis (GO:0035747, GO:2000501, GO:2000503).The functional adscription of genes regulated in T cells in response to the virus suggest an important role in the antiviral defense, with genes ascribed to response to virus (GO:0051607), type I-II interferon (IFN) (GO:0034340, GO:0032481, GO;0034341) and response to symbiont (GO:0140546).In addition, genes related to the regulation of defense response (GO:0031347) and innate immune response (GO:0045088) are also upregulated in T cells.Monocytes stand out because of their high expression of genes involved in the cellular response to organic substance (GO:0071310), but also those related to innate immune responses (GO:0045088, GO:0045089, GO:0002758).As expected, most of the genes regulated in neutrophils are associated with inflammatory response (GO:0006954), as well as with neutrophil degranulation (GO:0043312).Granulocyte activation (GO:0036230) and exocytosis (GO:0006887) genes are also strongly represented among the regulated genes in neutrophils.Finally, the genes regulated in thrombocytes in response to the virus are functionally related to hemostasis (GO:0007599), negative regulation of peptidase activity (GO:0010466) and integrin activation (GO:0033622, GO:0033623, GO: 003625).Other functions shared among the genes regulated by several cell types include myeloid leukocyte mediated immunity (GO:0002444, GO:0002275) and immune response-activating cell surface receptor signaling pathway (GO:0002768) (Supplementary Table S4).
When we performed a focused analysis on upregulated proteincoding genes among the most responsive populations, we identified a set of transcripts that was significantly upregulated at the same time in several PBL subpopulations indicating that the virus is able to induce a common molecular pathway (Figures 5A, B).These eight genes widely induced in cells from all affected PBL subpopulations included genes coding for two helicase with zinc finger 2 (helz2 and LOC110528534), a CD9 antigen-like (LOC110486460), a E3 ubiquitin-protein ligase rnf213-beta-like (LOC110502787), a tripartite motif-containing protein 16-like (LOC110513115), a sacsin-like (LOC110505631), a galectin-9-like (LOC110533869) and an IFN-induced very large GTPase 1 (LOC110486424) (Figure 5B).Another set of genes was widely stimulated in response to virus in at least three different PBL populations.These genes were grouped in important gene families such as, for instance, VHSV-induced proteins (LOC100135997 and LOC100136003), polyubiquitin-like proteins (LOC110536126, LOC110536134 and LOC100135966), different E3 ubiquitin-protein ligases rnf213-type alpha-like (LOC110528546, LOC110508243 and LOC110485301), RNF144A-A (LOC110500498), E3 ubiquitin/ISG15 ligase TRIM25-like (LOC110485965, LOC110516270 and LOC110518678) and additional genes encoding tripartite motif-containing protein 16like (LOC110516150 and LOC110521165) and 47-like (LOC110522087), among others (Supplementary Table S5).
Analyzing each PBL subpopulation, some of the genes significantly upregulated at transcriptional level in response to IPNV were exclusive to only one or two of them (Figures 5A, C).In an overview, a set of 34 genes was exclusively upregulated in IgM + IgD + B cells, whereas a total of 43 genes were singularly upregulated in IgT + B cells.Both B cell groups share 31 upregulated genes.T cells show only 14 unique, upregulated genes, increasing this quantity up to 36 together with monocytes.Additionally, monocytes upregulated 129 distinct genes.A similar number of unique, upregulated genes were found in neutrophils.Finally, 52 of these genes were uniquely upregulated in thrombocytes (Figure 5A).
Several genes previously identified as markers for B cells showed a significant increase at transcriptional level in response to IPNV, such as the Ig mu and delta heavy chains in IgM + IgD + B cells, and the immunoglobulins tau 1, tau 2 and tau 3 heavy chains in IgT + B cells, or the B cell receptor (LOC110537828) and CD79b (cd79b) in both.Similarly, different Ig light chains increased their levels of transcription in response to IPNV, such as kappa G1, kappa G3 and kappa F2 in both groups of B cells and sigma in IgT + B cells.Interestingly, kappa F1 and kappa G2 showed an opposite trend in these B cell populations, with the first one increasing its transcriptional level in IgM + IgD + B cells and decreasing it in IgT + B cells upon virus exposure, and the second one showing an opposite tendency.The interferon inducible protein (iip) also stands up in both groups while the interferon alpha/beta receptor 2 (LOC110501424) only in IgM + IgD + B cells.Related to elongation factor activity, several elongation factors (LOC110496143, LOC110538825, LOC110489228, LOC110524284, LOC110508425, LOC110521428, LOC110530688) were upregulated in both B cell types by the virus (Figure 5C, Supplementary Tables S5, S6).Regarding ccl4 + B cells, the most significant markers for this B cell population, CCL4 protein (ccl4) and C-C motif chemokine 4 (LOC110494096) also experienced a remarkable transcriptional upregulation in response to IPNV (Table 3).The interferoninduced protein 44-like was also highly expressed in this group, as in monocytes.In this last group, several genes encoding interferon-induced proteins (LOC110498001, LOC110538600, mx, LOC110494492, LOC110494493, LOC110488243, LOC110522342, and LOC110522343), E3 ubiquitin-protein ligases (DTX3L -LOC110494241, TRIM39 -LOC110500407, and LOC110531526), and tripartite motif-containing protein 16 (TRIM16 -LOC110489932, LOC110499822, LOC110513989, LOC110518413, and LOC110521260) or 47 (TRIM47 -LOC110503178) were also transcriptionally regulated (Supplementary Table S5).In a similar Enrichment analysis after IPNV infection.(A) Number of upregulated and downregulated genes in each blood leukocyte subpopulation.(B) Top 10 biological functions at level 6 of GO-terms after the IPNV infection in each cluster with upregulated genes.Bars represent the percentage of expressed genes associated with the GO term in the specific subpopulation analyzed (colored bars) regarding the percentage identified in all the transcriptome used as reference (grey bars).

Single-cell atlas or rainbow trout PBLs
In the current study, we have used total rainbow trout PBLs to establish a transcriptional atlas of a range of rainbow trout leukocyte subsets, to then study the early response of all these leukocyte subtypes to a viral encounter.Because in previous singlecell sequencing experiments in which we analysed blood B cells in rainbow trout by means of single cell transcriptomics, we sorted cells by means of small size, low complexity and high MHC II expression to focus our analysis on B cells (9,10), these previous studies never identified transcriptional patterns associated with plasma-like cells.In the current study, a small population identified clearly corresponds to this cell type, showing a high expression of Ig mu heavy chain (IgM) and irf4, but no expression of the Ig delta heavy chain (IgD).We had previously determined by means of flow cytometry that fish B cells loose surface IgD upon differentiation (26), similarly to mammalian B cells (27).Remarkably, an enrichment analysis highlighted the transcription of genes related with GO terms associated with endoplasmic reticulum development and differentiation.Again, flow cytometry studies have performed in rainbow trout recently have also established endoplasmic reticulum expansion as a characteristic trait of differentiated plasmablasts/plasma cells (28).An additional B cell subset was identified among PBLs, characterized by the transcription of ccl4.This ccl4 + B cell subset had been also reported in previous single-cell studies performed with rainbow trout blood B cells by our group (9) and in carp (29).As CCL4 seems to be closely related to mammalian CCL5, which is known to have the capacity to attract Th cells, it has been suggested that this population might specifically be focused on doing this, also in teleost fish.Regarding IgT + B cells, the majority of cells expressed a prosaposin-like coding gene (LOC110509903).The protein synthesized, prosaposin, is a precursor of four different saposins.In the tongue sole (Cynoglossus semilaevis), a peptide derived from saposins was shown to exhibit a significant antimicrobial activity (30), suggesting a possible antimicrobial role for IgT + B cells.Additionally, these cells are also characterized by the transcription of myocyte enhancer factor 2C (mef2c).In mice, the protein synthesized is important for B cell proliferation after BCR engagement, upon phosphorilation by p38 MAPK (31).
Among the different markers identified in B cells in this study, one gene stands up for its broad transcription in all B cells in comparison to other populations, the phospholipase A and acyltransferase 1-like.The role of this gene in the synthesis of Nacylethanolamines (NAEs) has been studied in humans and mice (32), yet their role in B cell function is still not clear.Other genes expressed at significantly higher levels in B cells when compared to other cell types include the zinc finger proteins 34-like (znf34) and 35-like (znf35).The first one has been shown to play an important role in the immune response against viruses by promoting the expression of type I IFN.Thus, for example, PRRSV (Porcine Reproductive and Respiratory Syndrome Virus) was shown to inhibit a long non-coding RNA and, consequently ZNF34, not allowing type I IFN expression and facilitating viral replication (33).ZNF35, on the other hand, binds to the rs10924104-A allele and promotes the expression of CD58 in B lymphocytes (34).Finally, the uncharacterized gene (LOC110533868), another highly expressed transcript in rainbow trout B cells, codes for EVI2A.Even though the function of this protein is still not clearly defined, it is thought to be related to BCR function as a lymphocyte-specific tumor suppressor (35).
In T cells, the most highly expressed gene was the one encoding the S100-A5-like protein.In humans, this protein has been identified in epithelial cells as a target for immunotherapy against bladder cancer.It is known to inhibit the secretion of inflammatoryrelated chemokines and the recruitment of CD8 + T cells (36).However, this protein was residually expressed in human peripheral blood cells (36), therefore, the biological role it plays in rainbow trout T cells remains to be defined.Another gene, which synthesizes the SH2 domain-containing protein 1A-like, is also overexpressed in T cells when compared to other cell subsets.This protein is present in activated T cells and NK cells in humans, regulating their immune functions (37).The gene encoding PLAC8 protein, also highly expressed in rainbow trout T cells, has been reported to play a role in inflammatory response by Th cells (38).Granzymes are known to play an important role in cell-mediated cytotoxic capacities of T cells and NK cells, in both mammals and fish (39).
In this study, a granzyme K-like was also highly transcribed in T cells.Rainbow trout monocytes highly transcribed olfactomedin-4, which has been reported to be present in a specific subset of human and murine neutrophils (40).Additionally, these cells transcribe lysozyme C-II, a c-type lysozyme with described roles in the immune response against a wide range of pathogens in fish (41,42), as well as the KRT79 protein, reported in monocytes and macrophages in humans (43).
The most characteristic gene transcribed by DCs was the macrophage expressed 1 (mpeg1), used as a marker for macrophages in vertebrates and for which antimicrobial activity has been reported in fish (44).In the other DC-like subpopulation, some additional genes are detected, such as the ribonucleosidediphosphate reductase subunit M2 (rrm2) or the ribonucleotide reductase catalytic subunit M1 (rrm1), known to alter the DNA damage response and the p53 pathway in humans (45).
HPCs were characterized by the high expression of genes encoding beta-galactoside-binding lectins, tetraspanin-8-like proteins, a low affinity immunoglobulin gamma Fc region receptor II-like protein and an osteoclast stimulatory transmembrane protein.Among beta-galactoside-binding lectins, we can find galectins, which have an important role in innate immunity against infections also in fish (46).Tetraspanin-8 has also been shown to act as a key component of the pro-inflammatory signaling cascade in cancer and viral infections due to its interaction with other proteins (47).On the other hand, in humans, the low affinity immunoglobulin gamma Fc region receptor IIa is related to phagocytosis and has been identified in neutrophils, monocytes, macrophages and DCs (48).Therefore, this group of HPC seems to represent a precursor of the myeloblastic line, as it shares additional markers also found in monocyte-macrophage precursors, such as for example the osteoclast stimulatory transmembrane protein.
In NCCs, the transcription of an eotaxin-like gene was the most outstanding characteristic.Its product is the cytokine CCL11, which interestingly, has been shown to inhibit the differentiation of DCs and promote the development of Th2 cells in humans (49).Neutrophils, on the other hand, were characterized by a high expression of complement factor D (cfd), also designated as adipsin, a well-known adipokine (a cytokine secreted by adipose tissue), also identified in fish (50).Although its precise role in the functionality of fish neutrophils is unknown, in mammals, adipsin secreted from adipose tissue seemed associated to the neutrophil infiltration that takes place during inflammatory arthritis (51), suggesting a potential autocrine role of this molecule.Finally, thrombocytes were characterized by the transcription of genes coding integrin aIIb and b3, which were also reported as thrombocyte-markers in fish in a previous study (52).Additionally, the gene coding for the coagulation factor XIII A chain was also identified as a thrombocyte marker in this study and before (53).

Early transcriptional response of PBLs to IPNV exposure
The increase observed in the number of genes transcribed in PBLs after the IPNV infection demonstrates that these cells respond to the viral encounter.Although the major percentage of transcripts upregulated corresponds to protein coding genes, the amount of transcripts mapping ribosomal genes is also strongly regulated, especially in some specific PBL subsets.
Interestingly, plasmablasts/plasma cells (irf4 + B cells), DC-like cell populations, NCC and HPCs were not significantly affected by virus encounter at this very early time point, while the other B cell subsets, T cells, DCs, thrombocytes and monocytes were the main cell types that responded to IPNV recognition.Surprisingly, some of the upregulated genes were common to all these affected PBL subpopulations after the IPNV infection, being all of them related to the immune system.These included genes like the sacsin-like, which codes for a protein that regulates the Hsp70 chaperone machinery (54) and has been previously seen to be upregulated in fish response to reovirus (55).Two other commonly regulated genes were those coding for helicases with zinc finger 2 (helz2 and LOC110528534), known to be involved in antiviral responses to the viral mimic polyriboinosinic polyribocytidylic acid (pIC) (56).This protein is also a transcriptional co-activator of several nuclear receptors, such as the peroxisome proliferator-activated receptor a (PPARA) and g (PPARG), being this a mechanism by which it exerts antiviral effects (57).A gene coding for the IFN-induced very large GTPase 1 (GVIN1) protein, which participates in the IFNmediated immunity (58), was also up-regulated by IPNV in all affected subsets.Interestingly, a recent study identified GVIN1 among the transcripts differentially expressed in bacterial coldwater disease (BCWD) resistant rainbow trout (59).A gene coding for TRIM16 was also in the group of commonly up-regulated genes.This protein is part of the TRIM superfamily, greatly involved in the cellular response to type I and II IFN-mediated immunity (60).A gene coding for a galectin-9-like protein was also up-regulated by IPNV in all affected subsets.Similarly, in red sea bream (Pagrus major), galectin-9-like transcription was up-regulated in response to either a viral or a bacterial pathogen (61).Interestingly, galectin-9 has been shown to induce the apoptosis of activated T cells (62) and to work as a chemoattractant for eosinophils (63).In mammals, galectin-9 has been demonstrated to induce platelet adhesion (64).The CD9 antigen-like, an abundant transcript in rainbow trout PBLs that seems to participate in the immune response to pathogenic encounter or vaccination (65, 66), was also among the genes commonly modified by IPNV.Yet, in contrast to what we have observed in the current study, a previous work from our group determined that the levels of transcription of CD9 decreased in IgM + B cells in response to viral hemorrhagic septicemia virus (VHSV) (66).Another commonly regulated gene is that coding for the E3 ubiquitin protein ligase rnf213-beta-like, a key factor for ubiquitin-dependent autophagy that usually shows higher levels of expression in cell types with a phagocytic capacity (67).Interestingly, Nombela et al. demonstrated that VHSV induced protein ubiquitination in rainbow trout (68).Therefore, it might be possible that IPNV stimulates a similar pathway.
Nonetheless, in this study, we have also detected a series of genes that are exclusively regulated in each of these specific leukocyte subsets in response to the virus, pointing to differences in how these cells respond to the virus.For example, B cells showed a response to the virus focused on both type-I IFN and cytokinemediated signaling, yet the overexpression of genes encoding different types of Igs, revealed that these cells also increase Ig production in response to the virus.The ccl4 + B cells, thought to be involved in attracting Th cells (9), upregulated a group of genes related to chemotaxis, in addition to genes related to the IFN response.In T cells, one of the genes exclusively regulated was that coding for the IFN-induced protein 44-like, known to be induced by IFNg in teleosts (69).Although its precise function is still unknown in fish species, it has been suggested to act as a viral replication suppressor, like in mammals (69).T cells also seem to regulate the transcription of different types of E3 ubiquitin-protein ligases upon IPNV encounter.Many of these ligases are well-known IFN-activated proteins that regulate the NF-kB signaling pathway (70).Some TRIM ligases were also specifically regulated in monocytes by the virus, in addition to IRF-3 and IRF-8 which also control the NF-kB signaling pathway in teleost fish (71).In neutrophils, the up-regulation of the levels of transcription of the NF-kB p100 subunit seems to indicate an activation of the noncanonical NF-kB pathway, which was also shown to be activated in human neutrophils (72).In concordance, the up-regulation of IL-1RII transcription should avoid the docking between IL-1b and IR-1RI, negatively regulating the canonical NF-kB pathway in these cells (73).Finally, in addition to the genes commonly regulated in all affected leukocyte subsets, thrombocytes specifically upregulates genes that code for proteins related to the adhesion and mobility.
In conclusion, in this study, we have obtained a single cell transcriptome of rainbow trout PBLs in control culture conditions and in cultures exposed to IPNV.This set of samples has allowed us to obtain a transcriptional atlas of rainbow trout PBL populations, identifying several novel genes for each of them.When the transcriptional profile of these cells was compared in control and infected conditions, we found that plasmablasts/plasma cells (irf4 + B cells), DC-like cell populations, NCC and HPCs were not significantly affected by the virus at this early time point, while the other B cell subsets, T cells, DCs, thrombocytes and monocytes were the main cell types that responded.Interestingly, eight genes were transcriptionally up-regulated by the virus in all of these cell types, genes mostly related to IFN signaling.Nonetheless, each of these populations also regulated a set of specific genes that revealed an exclusive response to the viral infection.Thus, for example, B cells regulated Ig production, while thrombocytes regulated genes associated with adhesion.These exclusive profiles also were in many cases associated to activation of the NF-kB pathway, yet revealed specific ways to achieve this activation in some cases.The data generated will be useful to better understand the functionality of teleost leukocyte subpopulations, and how they respond to a viral encounter.

Limitations of the study
The results obtained in this paper regarding the classification of leukocyte subsets is based on the transcriptional profiles obtained and should be further validated in future studies by complementary imaging, functional or flow cytometry experiments.Furthermore, it should be taken into account that small changes in the data could be obtained if the results were compared with an updated reference genome.Finally, the differential response to the virus was obtained studying only two fish and incubating the leukocytes with the virus in vitro.Slightly different results might be obtained if the infection was to be performed in vivo.Finally, it might be possible that throughout the incubation period, control leukocytes slightly modify their transcriptional pattern to the one that may had been obtained in freshly isolated leukocytes.
(A) UMAP visualization showing the different clusters obtained at a resolution of 0.15 and the percentage of cells in each one.(B) Dotplot representation of known markers for each cell type.(C) Top 5 GO-term showing significant enrichment at level 6 of 'Biological process' category in each cell cluster identified.

5
FIGURE 5 Upregulated gene distribution through those clusters with different gene expression after the infection.(A) Venn diagram comparing the number of genes unique to each cluster and those shared between two or more subpopulations.(B) Dotplot showing the transcription levels of genes commonly upregulated in response to IPNV in IgM + IgD + B cells, IgT + B cells, ccl4 + B cells, T cells, monocytes, neutrophils, and thrombocytes.(C) Protein-protein interaction network of genes upregulated in one or more clusters.

TABLE 1
Top 5protein coding genes showing the highest differences in percentage of cells expressing these genes between PBL populations.

TABLE 2
Long non-coding RNAs showing significant differences between populations.

TABLE 3
Top 5 protein coding genes differentially expressed in response to IPNV specifically in one cell type.

TABLE 3 Continued
Fold change (FC) values are obtained dividing control by infected values, therefore negative logFC indicate overexpression of genes in infected samples.