Abstract

   Fish red blood cells (RBCs), are integral in several biologic processes
   relevant to immunity, such as pathogen recognition, pathogen binding
   and clearance, and production of effector molecules and cytokines. So
   far, one of the best strategies to control and prevent viral diseases
   in aquaculture is DNA immunization. DNA vaccines (based on the
   rhabdoviral glycoprotein G [gpG] gene) have been shown to be effective
   against fish rhabdoviruses. However, more knowledge about the immune
   response triggered by DNA immunization is necessary to develop novel
   and more effective strategies. In this study, we investigated the role
   of fish RBCs in immune responses induced by DNA vaccines. We show for
   the first time that rainbow trout RBCs express gpG of viral hemorrhagic
   septicaemia virus (VHSV) (GVHSV) when transfected with the DNA vaccine
   ex vivo and modulate the expression of immune genes and proteins.
   Functional network analysis of transcriptome profiling of RBCs
   expressing GVHSV revealed changes in gene expression related to
   G-protein coupled receptor (GPCR)-downstream signaling, complement
   activation, and RAR related orphan receptor α (RORA). Proteomic profile
   functional network analysis of GVHSV-transfected RBCs revealed proteins
   involved in the detoxification of reactive oxygen species,
   interferon-stimulated gene 15 (ISG15) antiviral mechanisms, antigen
   presentation of exogenous peptides, and the proteasome. Conditioned
   medium of GVHSV-transfected RBCs conferred antiviral protection and
   induced ifn1 and mx gene expression in RTG-2 cells infected with VHSV.
   In summary, rainbow trout nucleated RBCs could be actively
   participating in the regulation of the fish immune response to GVHSV
   DNA vaccine, and thus may represent a possible carrier cells for the
   development of new vaccine approaches.

   Keywords: rainbow trout, erythrocytes, red blood cells, VHSV
   glycoprotein G, DNA vaccine, transcriptome, proteome, immune response

Introduction

   The fish immune system is characterized by an active innate immune
   system that is of primary importance in combating infections ([37]1).
   However, differences between the fish immune system and that of mammals
   complicate the extrapolation of knowledge, thus limiting our ability to
   control infectious diseases in fish. Piscine infectious diseases,
   especially viral infections, cause significant losses globally, which
   implies a negative impact on aquaculture industry. For that reason,
   efforts have been made to understand the fish antiviral immune response
   over the last few decades. A particular area of interest is the role
   that nucleated red blood cells (RBCs) play as immune cell mediators
   ([38]2–[39]4).

   In contrast to mammalian RBCs, fish RBCs are nucleated and contain
   organelles in their cytoplasm ([40]5). In addition, RBCs contain
   transcriptome machinery that plays an active role in the immune
   response against viral infections ([41]2). The involvement of nucleated
   RBCs in the immune response has been demonstrated both in vivo and in
   vitro. Moreover, RBCs can act as phagocytes and antigen-presenting
   cells and release cytokine-like factors such as interferon gamma that
   could affect macrophage function ([42]6). Nucleated RBCs also are able
   to develop specific responses to different pathogen-associated
   molecular patterns (PAMPs) ([43]4) and can modulate leukocyte activity
   by producing soluble factors ([44]4, [45]7). Recently, it has been
   demonstrated that fish RBCs are also involved in the immune response to
   viral infections ([46]7–[47]10).

   Currently, the prevention of viral diseases is only possible through
   vaccination or immune stimulation. In aquaculture, one of the best
   strategies for controlling and preventing viral diseases is DNA
   immunization ([48]11). This method triggers long-term protection
   against viral infections ([49]11). However, more knowledge about the
   immune response triggered by DNA immunization is necessary in order to
   develop new and more effective viral prevention strategies. It is
   necessary to understand both the viral molecules involved in triggering
   the host immune responses (immunogenicity and antigenicity) and the
   viral-induced immune host responses in efforts to improve current and
   develop new vaccination strategies. In addition, new applications
   targeting immune cells are being developed to improve the immune
   response of DNA vaccines ([50]12).

   At the present time, only DNA vaccines based on the rhabdoviral
   glycoprotein G (gpG) gene have been found to be effective for all fish
   rhabdoviruses tested ([51]13, [52]14). However, the immune mechanisms
   responsible for their efficacy remain largely unknown. Therefore, this
   topic is of special importance to supplement the knowledge of how DNA
   vaccines confer immune protection, especially given the recent
   revelations of the importance of RBCs in the piscine immune response.

   In this study, we show for the first time that rainbow trout RBCs can
   express the gpG of VHSV (GVHSV) and modulate the expression of genes
   related to interferon, such as interferon-inducible myxovirus
   resistance gene (mx), interferon-inducible RNA-dependent protein kinase
   (pkr), and interferon-induced protein with tetratricopeptide repeats 5
   (ifit5). Transcriptomic and proteomic profiles network analyses
   revealed genes and proteins involved in G-protein coupled receptor
   (GPCR)-downstream signaling, complement activation, RORA-activates gene
   expression, ISG15 antiviral mechanisms, and detoxification of reactive
   oxygen species. Further, conditioned medium (CM) of GVHSV-transfected
   RBCs conferred protection to RTG-2 cell line against VHSV infection.
   Our results lead us to suggest that RBCs are immune cell mediators that
   play an active role in GVHSV DNA vaccine immune stimulation. Therefore,
   RBCs could be considered promising target or carrier cells in the
   development of new vaccine approaches.

Materials and methods

Animals

   Rainbow trout (Oncorhynchus mykiss) of ~7–10 cm were obtained from a
   VHSV-free commercial farm (PISZOLLA S.L., CIMBALLA FISH FARM, Zaragoza,
   Spain), and maintained at the University Miguel Hernandez (UMH)
   facilities at 14°C. Prior to experiments, fish were acclimatized to
   laboratory conditions for 2 weeks.

Cell cultures

   RBCs were obtained from the peripheral blood of fish sacrificed by
   overexposure to tricaine (tricaine methanesulfonate, Sigma-Aldrich,
   Madrid, Spain) (0.2 g/L). Peripheral blood was collected from the
   caudal vein using insulin syringes (NIPRO, Bridgewater, NJ, USA). RBCs
   were purified by two density gradient centrifugations (1,600 rpm,
   Ficoll 1.007; Lymphoprep, Reactiva, Sigma-Aldrich) as previously
   described ([53]8). Purified RBCs were placed in RPMI-1640 medium (Dutch
   modification) (Gibco, Thermo Fisher Scientific Inc., Carlsbad, CA)
   supplemented with 10% gamma irradiated fetal bovine serum (FBS)
   (Cultek, Madrid, Spain), 1 mM pyruvate (Gibco), 2 mM L-glutamine
   (Gibco), 50 μg/mL gentamicin (Gibco), 2 μg/mL fungizone (Gibco), 100
   U/mL penicillin (Sigma-Aldrich), and 100 μg/mL streptomycin
   (Sigma-Aldrich). The cells were cultured at 14°C.

   The rainbow trout cell line RTG-2 (Rainbow Trout Gonad-2) was purchased
   from the American Type Culture Collection (ATCC 50643) and maintained
   at 21°C in MEM medium (Sigma-Aldrich) containing 10% FBS, 1 mM
   pyruvate, 2 mM L-glutamine, 50 μg/mL gentamicin, and 2 μg/mL fungizone.
   RTS11, a rainbow trout monocyte/macrophage-like cell line (donated by
   Dr. Niels Bols) ([54]15) isolated from a spleen hematopoietic culture
   was maintained at 21°C in Leibovitz's medium (L-15) (Sigma-Aldrich)
   supplemented with 20% FBS, 1 mM pyruvate, 2 mM L-glutamine, 50 μg/mL
   gentamicin, and 2 μg/mL fungizone.

Antibodies

   Primary antibodies used in the manuscript included rabbit polyclonal
   antibody against Mx protein produced at the laboratory of Dr. Amparo
   Estepa ([55]16, [56]17), and mouse polyclonal antibodies against IL1β
   (interleukin 1 beta) ([57]18, [58]19), IL8 (interleukin 8) ([59]20),
   and TNFα (tumor necrosis factor alpha) ([60]21) produced at the
   laboratory of Dr. Luis Mercado. A mouse monoclonal 2C9 antibody
   produced at laboratory of Dr. Julio Coll against the N protein of VHSV
   was used for VHSV labeling ([61]22). For GVHSV labeling, we used a
   mixed of anti-GVHSV monoclonal antibodies (MAbs) (C10, 3F1A2, and I16)
   ([62]23) produced at Dr. Julio Coll's laboratory. Secondary antibodies
   used in these studies included anti-rabbit IgG CF^TM 647 and anti-mouse
   IgG CF^TM 647 (Sigma-Aldrich, Madrid, Spain) produced in goat.

Plasmids

   Plasmid pmTFP1 (Allele Biotechnology, ABP-FP-TCNCS), encoding the teal
   fluorescent protein 1 (mTFP1) ([63]24), used as control plasmid, and
   pmTFP1GVHSV, encoding mTFP1 fused to the C-terminus of the membrane gpG
   of VHSV (GVHSV) (GenBank accession [64]A10182.1), described previously
   ([65]25), were used for transfection assays.

Cell transfection assays

   RBC transfection assays were performed by electroporation using the
   Neon™ Transfection System (Life Technologies, Thermo Fisher Scientific,
   Inc.) one day after Ficoll purification. For each electroporation
   reaction, we used 4 μg of plasmid construct (pmTFP1 or pmTFP1GVHSV
   plasmid) per 1 × 10^6 cells resuspended in Buffer T (Neon™ Transfection
   System Kit, Life Technologies). RBCs were electroporated at 1600 V, 30
   ms, and 1 pulse and incubated at 14°C for one to six days in RPMI 10%
   FBS.

   The RTS11 cell line was transfected by electroporation with 4 μg of
   plasmid construct (pmTFP1 or pmTFP1GVHSV) per 1·10^6 cells using the
   Neon™ Transfection System and resuspended in Buffer R (Neon™
   Transfection System Kit). RTS11 was electroporated at 1,600 V, 30 ms,
   and 1 pulse and incubated at 21°C for one to six days in L-15 20% FBS.

Transcriptome analysis of FACS single-cell sorted GVHSV-expressing RBCs

   Ficoll-purified RBCs from 24 fish were transfected as described above
   with pmTFP1 or pmTFP1GVHSV (Figure [66]1). At six days
   post-transfection, TFP1- or GVHSV-expressing RBCs (6–10 cells per fish)
   were sorted by FACS single-cell sorting using the BD FACSJazz™ cell
   sorter (BD Biosciences, Madrid, Spain). FACS single-cell sorted RBCs
   were visualized in the IN Cell Analyzer 6000 Cell Imaging system (GE
   Healthcare, Little Chalfont, UK) (Figure [67]S1). Each sample was
   resuspended in 9.5 μL of 10 × lysis buffer (Clontech, Takara Bio,
   Mountain View, CA, USA) and 0.5 μL of RNase inhibitor (Invitrogen,
   ThermoFisher Scientific, Waltham, MA, USA). Twenty-four fish samples
   were grouped in three pools of eight individuals for each condition
   (pmTFP1 or pmTFP1GVHSV) (Figure [68]1) and preserved at −80°C until
   cDNA library construction. Then, cDNA was directly produced from pooled
   lysed cells using SMART-Seq v4 Ultra Low Input RNA Kit (Clontech,
   Takara Bio) ([69]26). Sequence reads are available at SRA-NCBI,
   SRA-NCBI Accession SRP133501. RNA-Seq library preparation and
   sequencing were carried out by STABVida Lda (Caparica, Portugal).

Figure 1.

   [70]Figure 1
   [71]Open in a new tab

   General workflow of experimental steps from sample collection to data
   analysis.

Proteome analysis of transfected RBCs

   Ficoll-purified RBCs from 16 fish were transfected as described above
   with pmTFP1 or pmTFP1GVHSV (Figure [72]1). At six days
   post-transfection, RBCs were pelletized by centrifugation (1,600 rpm),
   the supernatant was removed, and the cell pellet was washed three times
   with PBS, digested, and cleaned-up/desalted as previously described
   ([73]26). Samples were pooled in two pools of eight individuals for
   each condition (pmTFP1 or pmTFP1GVHSV) (Figure [74]1). Then, samples
   were subjected to liquid chromatography and mass spectrometry analysis
   (LC-MS) as previously described ([75]26), except that High pH
   Reversed-phase Peptide Fractionation Kit (Pierce, Thermo Fisher
   Scientific Inc.) was used for pH reversed-phase peptide fractionation,
   and four peptide fractions were collected. Progenesis QI v4.0
   (Nonlinear Dynamics, Newcastle, UK) was used to analyze differential
   protein expression according to the “between-subject design.” Log[2]
   peptide ratios followed a normal distribution that was fitted using
   least squares regression. Mean and standard deviation values derived
   from the Gaussian fit were used to estimate P-values and false
   discovery rates (FDR) at a quantitation level. The confidence interval
   for protein identification was set to ≥95% (P-value ≤ 0.05), and only
   peptides with an individual ion score above the 1% FDR threshold were
   considered to be correctly identified. Only proteins having at least
   two peptide spectrum matches (PSMs) were considered in the
   quantitation.

Pathway enrichment analysis

   Differentially expressed genes (DEGs) and differentially expressed
   proteins (DEPs) pathway enrichment analysis were performed using ClueGO
   ([76]27) CluePedia ([77]28), and Cytoscape ([78]29). The GO Immune
   System Process, Reactome Pathway, and Reactome Reactions databases were
   used. A P-value ≤ 0.05 and Kappa score of 0.4 were used as threshold
   values. Genes and proteins were identified by sequence homology with
   Homo sapiens using Blast2GO version 4.1.9 Gotz ([79]30).

RTG-2 cell line immune response to conditioned medium from transfected RBCs

   In order to evaluate the immune response elicited by GVHSV-transfected
   RBCs on RTG-2 cells, RTG-2 cell monolayers in 96-well plates were
   treated with CM from pmTFP1- or pmTFP1GVHSV-transfected RBCs. First, CM
   of transfected RBCs were collected at three and six days
   post-transfection, recovered by centrifugation (1,600 rpm), and
   filtered with 0.2 μm filters (Cultek). The CM was diluted 1/5 in MEM
   10% FBS, and RTG-2 cell monolayers were treated with diluted CM for
   three days at 14°C. Finally, RTG-2 cell were stored at −80°C in lysis
   buffer until RNA extraction and RT-qPCR.

   To evaluate the protection conferred by GVHSV-transfected RBC CM on
   RTG-2 cells against VHSV infection, pmTFP1- and pmTFP1GVHSV-transfected
   RBC CM was collected at three and six days post-transfection as
   described above. Then RTG-2 cell monolayers were pre-treated with the
   CM, diluted 1/5 and 1/125 in MEM 10% FBS, and incubated for 24 h at
   14°C. Then, CM was removed and RTG-2 cell monolayers were infected with
   VHSV at a multiplicity of infection (MOI) of 10^−2 in RPMI 2% FBS, for
   2 h at 14°C. Medium was removed and fresh medium (RPMI 2% FBS) was
   added. The cells were incubated for an additional 24 h at 14°C. After
   that, VHSV infectivity was evaluated by means of focus forming units
   (FFU)/mL as previously described ([80]9). N-VHSV antibody (2C9) was
   used as primary antibody. Immunofluorescence images were taken with the
   IN Cell Analyzer 6000 cell imaging system.

Co-cultures of transfected RBCs with RTS11 cells

   Ficoll-purified RBCs were transfected as indicated above. Transfected
   RBCs were co-cultured with RTS11 cells using Transwell® polyester
   membrane cell culture inserts (0.4 μm pore size, Costar, Corning,
   Sigma-Aldrich) on 24-well plates for three days at 14°C. Then, RTS11
   samples were stored at −80°C in lysis buffer until RNA extraction and
   RT-qPCR.

RNA extraction, cDNA synthesis, and RT-qPCR gene expression

   RNA extraction, cDNA synthesis and RT-qPCR analyses were performed as
   previously described ([81]8). Briefly, E.Z.N.A.® Total RNA Kit (Omega
   Bio-Tek, Inc., Norcross, GA) was used together with DNAse (TURBO™
   DNase, Ambion, Thermo Fisher Scientific, Inc.) for RNA extraction. RNA
   was quantified with a NanoDrop® Spectrophotometer (Nanodrop
   Technologies, Wilmington, DE). After cDNA synthesis ([82]31), RT-qPCR
   was performed using the ABI PRISM 7300 System (Applied Biosystems,
   Thermo Fisher Scientific, Inc.). Specific primers and probes are listed
   in Table [83]1. The eukaryotic 18S rRNA gene (Applied Biosystems,
   Thermo Fisher Scientific, Inc.) or the gene encoding EF1α were used as
   endogenous controls.

Table 1.

   Table of primers used in RT-qPCR.
   Gene Forward primer Reverse primer Probe Reference or accession number
   arrb2 GTGGAGCTGCCCTTTGTCTTA TGAATGTGGGCGGGATATG
   TGCACCCCAAACCCACAGAACTGC [84]NM_001171899.1
   cat TGCAAGACACCCCGTTCATA TGGCGTGTACCACCCTCTCT [85]XM_021557350.1
   dnm2 GTCAACAAGTCCATCAGGGATCT CAACTCAGAATGGATGAAGTCTTTAGC
   [86]XM_021596596.1
   ef1α ACCCTCCTCTTGGTCGTTTC TGATGACACCAACAGCAACA GCTGTGCGTGACATGAGGCA
   ([87]32)
   gvhsv GGGCCTTCCTTCTACTGGTACTC CGGAATCCCGTAATTTGGAAT
   CTGTTGCTGCAAGGCGTCCCCT ([88]31)
   ifit5 CCCTGCCCTCATCTTTCTTCT CCCTCAATGACTCTGACAAGCA
   CCAGCTTCGGCCTGTTTCTGTTCCA [89]AF483530.1
   ifn1 ACCAGATGGGAGGAGATATCACA GTCCTCAAACTCAGCATCATCTATGT
   AATGCCCCAGTCCTTTTCCCAAATC ([90]8)
   il10 CTGCTGGACGAAGGGATTCTA TAAAGTCGTTGTTGTTCTGTGTTCTG
   AAGTTCTATCTCGACACGGTGCTGCCC [91]NM_001245099.1
   il12β TGACAGCCAGGAATCTTGCA GAAAGCGAATGTGTCAGTTCAAA
   ACCCAACGACCAGCCTCCAAGATG ([92]33)
   inos TCAGAACCTCCTCCACAA GTGTACTCCTGAGAGTCCTTT GCACCGACAGCGTCTA ([93]33)
   jak2 CCTGCTCTACGCCTCACAGATC GCCAAGTCACGGTGGATGTA
   CAAGGGCATGGACTACCTAGCGACCA [94]XM_021622657.1
   mhcI GACAGTCCGTCCCTCAGTGT CTGGAAGGTTCCATCATCGT ([95]34)
   mhcII TGCCATGCTGATGTGCAG GTCCCTCAGCCAGGTCACT
   CGCCTATGACTTCTACCCCAAACAAAT ([96]35)
   mx1-3 TGAAGCCCAGGATGAAATGG TGGCAGGTCGATGAGTGTGA
   ACCTCATCAGCCTAGAGATTGGCTCCCC ([97]36)
   nkef CGCTGGACTTCACCTTTGTGT ACCTCACAACCGATCTTCCTAAAC ([98]8)
   nup107 GCTGTCGCCTATTGTACGAGATG TGAGCCTTCTTCTGAACTGAACTCT
   [99]XM_021564152.1
   pkr ACACCGCGTACCGATGTG GGACGAACTGCTGCCTGAAT CACCACCTCTGAGAGCGACACCACTTC
   ([100]8)
   prdx6 GGACCCTGATGAGCTTGACAA CTTATCTGGACCAATCACAAACACA
   [101]NM_001165132.2
   rab7a GTTGCGTGCTGGTGTTTGAC ACTCGTCCCTCCAGCTGTCTAG
   TGACCGCCCCCAACACCTTCAA [102]XM_021609589.1
   rora AGGTGGTGTTCATCAGGATGTG CGTCGGTCCCAGCGTACTT
   CGTGCCTTTGACTCTCAGAACAGCACC [103]XM_021608048.1
   sec13 GCAGTGATCCAGGCACAGAA CTGGGACTAGGATAGATGGTAGAAGTG
   ATTCCACTCCTCCTCCTACCCCCACA [104]XM_021610740.1
   socs1 GATTAATACCGCTGGGATTCTGTG CTCTCCCATCGCTACACAGTTCC ([105]37)
   sod1 GCCGGACCCCACTTCAAC CATTGTCAGCTCCTGCAGTCA ([106]8)
   trx AGACTTCACAGCCTCCTGGT ACGTCCACCTTGAGGAAAAC ([107]8)
   [108]Open in a new tab

Immunofluorescence and flow cytometer assays

   Transfected RBCs were fixed, permeabilized, and incubated with primary
   and secondary antibodies as described in Nombela et al. ([109]9). Flow
   cytometry was done using a FACS Canto II (BD Biosciences, Madrid,
   Spain) flow cytometer. RBC populations were selected by forward scatter
   (FSC) and side scatter (SSC) (Figure [110]S2). Immunofluorescence
   images were taken with the IN Cell Analyzer 6000 cell imaging system.

Statistical analysis

   GraphPad Prism 6 ([111]www.graphpad.com) software was used for
   statistical analysis. Flowing Software ([112]www.flowingsoftware.com)
   was used to analyze flow cytometry experiments.

Results

GVHSV expression in rainbow trout RBCs

   TFP1 (Figure [113]2A) and GVHSV (Figure [114]2B) expression in
   transfected RBCs was monitored through fluorescent microscopy.
   Perinuclear expression of GVHSV was observed in pmTFP1GVHSV-transfected
   RBCs (Figure [115]2B), which is in contrast to the nuclear and
   cytoplasmic expression observed in pmTFP1-transfected RBCs (Figure
   [116]2A).

Figure 2.

   [117]Figure 2
   [118]Open in a new tab

   Monitoring transfection of rainbow trout red blood cells (RBCs).
   Fluorescent micrographs of RBCs transfected with (A) pmTFP1 and (B)
   pmTFP1GVHSV at six days post-transfection at 14°C, monitored by teal
   fluorescent protein (TFP). Fluorescent images taken with the IN Cell
   6000 imaging system, augmentation 60x. (C) RBC transfection with 2, 4,
   and 8 μg of plasmid pmTFP1GVHSV was confirmed by GVHSV gene RT-qPCR, at
   one day (white bars), three days (gray bars) and six days (black bars)
   post-transfection. The eukaryotic 18S rRNA gene was used as an
   endogenous control. Data are displayed as mean ± SD (n = 3). Two-way
   ANOVA with Tukey's multiple comparisons test was performed between
   plasmid concentrations (black lines and asterisks) and times
   post-transfection (gray lines and asterisks). (D) Time course of
   transfected RBCs (black bars) and transfected RTS11 (gray bars) with 4
   μg of pmTFP1GVHSV at one, three and six days post-transfection
   monitored by GVHSV RT-qPCR. The eukaryotic 18S rRNA gene was used as an
   endogenous control. Data are displayed as mean ± SD (n = 3 for RBCs and
   n = 2 for RTS11). Two-way ANOVA with Sidak's multiple comparisons test
   was performed between cell types at the different times
   post-transfection. ^*, ^**, ^***, and ^****, represent the P values <
   0.05, < 0.01, < 0.001, and < 0.0001, respectively. Overlay flow
   cytometry histogram of representative GVHSV immunostaining of (E) RBCs
   and (F) RTS11 cells transfected with 4 μg of pmTFP1 or pmTFP1GVHSV at
   three days post-transfection (gray histogram, pmTFP1-transfected; green
   histogram, pmTFP1GVHSV-transfected). Representative immunofluorescence
   of GVHSV intracellular immunostaining of (G) RBCs and (H) RTS11 cells
   transfected with 4 μg of pmTFP1 or pmTFP1GVHSV at three days
   post-transfection (protein [APC, red] and nuclei [DAPI, blue]).
   Fluorescence images were taken with 60 × magnification.

   Time-course and dose-response assays were performed to establish the
   optimal conditions of pmTFP1GVHSV transfection. RBCs achieved the
   maximum expression of the GVHSV gene at six days post-transfection with
   4 μg per 1 × 10^6 RBCs evaluated by RT-qPCR (Figures [119]2C,D). These
   conditions were used for the following assays.

   GVHSV gene expression in rainbow trout RBCs was compared with that in
   RTS11, another rainbow trout cell line. The RTS11
   monocyte/macrophage-like cell line had higher levels of GVHSV gene
   expression at 24 h post-transfection (Figure [120]2D) than the RBCs and
   decreased over time. This is in contrast to pmTFP1GVHSV-transfected
   RBCs, which reached the maximum level of GVHSV expression at six days
   post-transfection. GVHSV gene expression levels were not significantly
   different between RBCs and RTS11 at three and six days
   post-transfection, although GVHSV gene expression was lower in RBCs
   than RTS11 at all-time points analyzed.

   Also, GVHSV protein expression was lower in RBCs than RTS11 at three
   days post-transfection by flow cytometry (Figures [121]2E,G) compared
   to RTS11 (Figures [122]2F, H).

RNA sequencing of FACS single-cell sorted GVHSV-expressing RBCs

   In order to evaluate the immune response triggered by GVHSV DNA vaccine
   in RBCs, we analyzed the transcriptome of FACS single-cell sorted
   GVHSV-expressing RBCs, exclusively (Figure [123]1; Figure [124]S1).

   RNA-sequencing of FACS single-cell sorted GVHSV-expressing RBCs
   (compared with FACS single-cell sorted TFP1-expressing RBC) revealed
   3249 DEGs (FDR < 0.05) from a total of 137,444 transcripts. Among these
   3,249 DEGs, 1,786 were upregulated, and 1,463 were downregulated (Table
   [125]S1). Functional pathway enrichment evaluation in FACS single-cell
   sorted GVHSV-expressing RBCs showed upregulation of GPCR downstream
   signaling and RORA-activates gene expression pathways using the
   Reactome Pathways Database (Figure [126]3A, Table [127]S2), and the
   complement activation pathway using GO Immune Process Database (Figure
   [128]3C, Table [129]S3). On the other hand, transcriptional regulation
   by RUNX3 and eukaryotic translation elongation pathways appeared to be
   downregulated using the Reactome Pathways Database (Figure [130]3B,
   Table [131]S2).

Figure 3.

   [132]Figure 3
   [133]Open in a new tab

   Functional pathway analysis of differentially expressed genes (DEGs) in
   the FACS single-cell sorted GVHSV-expressing RBC transcriptome profile.
   DEGs in pmTFP1GVHSV with FDR Q value < 0.05 and fold change P value <
   0.05. Overrepresented GO terms were identified by the Cytoscape ClueGO
   app with Reactome and GO Immune Process terms. The Reactome Pathways
   multilevel pie chart shows (A) upregulated and (B) downregulated
   overrepresented terms in pmTFP1GVHSV-expressing RBCs. The GO Immune
   Process multilevel pie chart shows (C) upregulated overrepresented
   terms in pmTFP1GVHSV-expressing RBCs. Asterisks denote GO-term
   significance (^**p < 0.01).

   Among all DEGs identified, we identified modulation of genes related to
   interferon alpha/beta signaling in antiviral mechanisms. Particularly,
   genes encoding suppressor of cytokine signaling 3 (socs3); adenosine
   deaminase, RNA specific (adar); radical s-adenosyl methionine
   domain-containing 2 (rsad2); tripartite motif-containing 25 (trim25);
   and nucleoporins such as nucleoporin 58 (nup58), karyopherin subunit
   beta 1 (kpnb1), and nucleoporin 210 (nup210) were upregulated (Table
   [134]S1). Conversely, interferon regulatory factor 1 (irf1), interferon
   alpha and beta receptor subunit 1 (ifnar1), Janus kinase 1 (jak1), and
   major histocompatibility complex class I (mhcI) genes were
   downregulated. We also found that several DEGs related to interleukin
   signaling were upregulated. These included interleukin 2 receptor
   subunit beta (il2rb), socs3, Janus kinase 2 (jak2), interleukin 16
   (il16), interleukin 15 (il15), interleukin 12a (il12a), tumor necrosis
   factor (tnf), rar-related orphan receptor α (rora), and interleukin 8
   (il8) (Table [135]S1). The complement cascade was represented by the
   upregulation of genes encoding complement c3 (c3), carboxypeptidase b2
   (cpb2), coagulation factor II, thrombin (f2), and complement c1q b
   chain (c1qb) (Table [136]S1).

Proteome sequencing of GVHSV-transfected RBCs

   We evaluated the proteome of pmTFP1GVHSV-transfected RBCs, a sample
   composed of few GVHSV-expressing RBCs and mostly non–GVHSV-expressing
   RBCs, in order to evaluate the immune response of non–GVHSV-expressing
   RBCs to the signal triggered by GVHSV-expressing RBCs. Proteomic
   profiling identified 1,750 proteins (Table [137]S4). After applying a
   filter of FDR < 0.001 and [−1.5 < Log[2]fold change (FC)> 1.5], for
   pmTFP1GVHSV-transfected RBCs compared to pmTFP1-transfected RBCs, 199
   DEPs were identified, of which 75 were upregulated and 124 were
   downregulated (Table [138]S4). ClueGO analysis using the Reactome
   Pathways Database revealed upregulated terms related to ISG15 antiviral
   mechanism, detoxification of reactive oxygen species (ROS), mRNA
   splicing, host interactions of HIV, CLEC7A signaling, interleukin1
   family signaling, and FCERI-mediated NF-κβ factors in
   pmTFP1GVHSV-transfected RBCs (Figure [139]4A, Table [140]S5).
   Conversely, downregulated terms in pmTFP1GVHSV-transfected RBCs
   appeared to be related to DNA replication and cell cycle regulation,
   Orc1 removal from chromatin, synthesis of PIPs at the plasma membrane,
   and COPI-mediated anterograde transport (Figure [141]4B, Table
   [142]S5). ClueGO analysis using Reactome Reactions Database showed
   upregulated terms related to proteosomal cleavage of exogenous antigen,
   snRNP nuclear import and release, and formation of the AT-AC C complex
   in pmTFP1GVHSV-transfected RBCs (Figure [143]4C, Table [144]S6). On the
   other hand, downregulated terms included degradation of ubiquitinated
   p27/p21 by the 26S proteasome, Orc1 phosphorylation by cyclin A/CDK2,
   association of phospo-L13a with GAIT element of ceruloplasmin mRNA and
   exocytosis of secretory granulate lumen proteins (Figure [145]4D, Table
   [146]S6). ClueGO analysis using the GO Immune Process Database
   identified antigen processing and presentation of exogenous peptide
   terms in pmTFP1GVHSV-transfected RBCs (Figure [147]4E, Table [148]S7).

Figure 4.

   [149]Figure 4
   [150]Open in a new tab

   Functional pathway analysis of differentially expressed proteins (DEPs)
   in the GVHSV-transfected RBC proteome profile. DEPs in pmTFP1GVHSV with
   (−1.5 < log[2] fold change < 1.5) and an FDR P-value < 0.001.
   Overrepresented GO terms were identified by the Cytoscape ClueGO app,
   with Reactome Pathways, Reactome Reactions and GO Immune Process terms.
   The Reactome Pathways multilevel pie chart shows (A) upregulated and
   (B) downregulated overrepresented terms in pmTFP1GVHSV-transfected
   RBCs. The Reactome Reactions multilevel pie chart shows (C) upregulated
   and (D) downregulated overrepresented terms in pmTFP1GVHSV-transfected
   RBCs. The GO Immune Process multilevel pie chart shows (E) upregulated
   overrepresented terms in pmTFP1GVHSV-transfected RBCs. Asterisks denote
   GO-term significance (^**p < 0.01).

   Among the most upregulated DEPs in pmTFP1GVHSV-transfected RBCs (Table
   [151]S4), we identified proteins related to: (i) the nuclear pore
   complex, such as importin-8 (IPO8), nuclear pore complex protein Nup107
   (NUP107), and translocated promoter region nuclear basket protein
   (TPR); (ii) inhibitor of nuclear factor kappa-B kinase subunit alpha
   (CHUK), and (iii) protection against oxidative stress such as
   thioredoxin (TRX), peroxiredoxin 4 (PRDX4), superoxide dismutase 1
   (SOD1), and thioredoxin like 1 (TXNL1).

Validation of upregulated pathways by RT-qPCR

   Upregulated pathways in pmTFP1GVHSV-transfected RBCs were validated via
   RT-qPCR analysis. Genes and proteins were selected from each pathway
   for validation. For the GPCR-downstream signaling term (which was
   upregulated in the RNA sequencing results of FACS single-cell sorted
   GVHSV-expressing RBCs), we measured gene expression levels of the
   arrestin beta 2 (arrb2) gene, which was significantly upregulated, and
   the Janus kinase 2 (jak2) gene, which was significantly downregulated
   in our RT-qPCR results in contrast to RNA sequencing results (Figure
   [152]5). The rora gene, a representative gene of RORA-activates gene
   expression pathway, an overrepresented term in RNA sequencing results
   of FACS single-cell sorted GVHSV-expressing RBCs, was significantly
   upregulated in RT-qPCR results (Figure [153]5). For the detoxification
   of reactive oxygen species pathway (found to be upregulated in the
   proteome analysis of pmTFP1GVHSV-transfected RBCs), the superoxide
   dismutase 1 (sod1), peroxiredoxin 6 (prdx6), natural killer enhancing
   factor (nkef), and thioredoxin (trx) genes were significantly
   upregulated (Figure [154]5). For the ISG15 antiviral mechanism pathway
   (which was upregulated in the proteome analysis of
   pmTFP1GVHSV-transfected RBCs), the nucleoporin 107 (nup107),
   interferon-induced protein with tetratricopeptide repeats 5 (ifit5),
   interferon-inducible Mx (mx), and interferon-inducible RNA-dependent
   protein kinase (pkr) genes were significantly upregulated. However, the
   interferon type 1 (ifn1) gene was significantly downregulated (Figure
   [155]5). For the antigen presentation of exogenous peptide pathways
   (upregulated in proteome analysis of pmTFP1GVHSV-transfected RBCs), the
   major histocompatibility complex class I (mhcI) and II (mhcII), SEC13
   homolog-nuclear pore and COPII coat complex component (sec13), and
   dynamin 2 (dnm2) genes were significantly upregulated, but the
   RAB7A-member RAS oncogene family (rab7a) appeared to be slightly
   downregulated in contrast to proteomic results (Figure [156]5).

Figure 5.

   [157]Figure 5
   [158]Open in a new tab

   Gene expression evaluation to validate functional pathway analysis of
   differentially expressed genes (DEGs) and differentially expressed
   proteins (DEPs) in GVHSV-transfected RBCs transcriptome and proteome
   profiles, respectively. RBCs were transfected with pmTFP1 and
   pmTFP1GVHSV plasmids for 6 days at 14°C. Afterwards, gene expression
   was evaluated by RT-qPCR. Data are displayed as mean ± SD (n = 5). The
   EF1α gene was used as an endogenous control. The Wilcoxon test was
   performed between pmTFP1GVHSV- and pmTFP1-transfected RBCs (control,
   red line). ^*and ^**represent P-values < 0.05 and < 0.01, respectively.

   However, at a protein level, we confirmed the upregulation of
   interferon inducible Mx protein, and interleukins interleukin 1 beta
   (IL1β), interleukin 8 (IL8), and tumor necrosis factor alpha (TNFα)
   measured by flow cytometry (Figures [159]6A, B).

Figure 6.

   [160]Figure 6
   [161]Open in a new tab

   Immune proteins expression in GVHSV-transfected RBCs. (A) Immune
   protein expression measured by flow cytometry and calculated by the
   formula MRFI (mean relative fluorescence intensity) = fluorescence in
   pmTFP1GVHSV- transfected RBCs MRFI (mean relative fluorescence
   intensity) = fluorescence in pmTFP1GVHSV- transfected RBCs
   /fluorescence in pmTFP1-transfected RBCs. (B) Representative
   immunofluorescence of pmTFP1- and pmTFP1GVHSV-transfected RBCs (protein
   [APC, red] and nuclei [DAPI, blue]). Fluorescence images were taken
   with 60 × magnification.

Protection conferred by pmtfp1gvhsv-transfected RBC CM on RTG-2 cells

   In order to evaluate the capacity of RBCs to propagate the immune
   response elicited by GVHSV to other cell types, we measured the
   protection conferred by pmTFP1GVHSV-transfected RBC CM to RTG-2 cells
   against VHSV infection. Treatment of RTG-2 cells with three or six days
   pmTFP1GVHSV-transfected RBC CM significantly decreased VHSV compared
   with pmTFP1-transfected RBC CM (Figures [162]7A,B, for three and six
   days transfected RBC CM, respectively).

Figure 7.

   [163]Figure 7
   [164]Open in a new tab

   Protection conferred by conditioned medium (CM) from GVHSV-transfected
   RBCs against VHSV infection in RTG-2 cells. RTG-2 cells pretreated with
   CM from RBCs transfected with pmTFP1 (black bars) and pmTFP1GVHSV (gray
   bars) plasmids (diluted 1/5 and 1/125 in MEM 10% FBS), at three (A) and
   six (B) days post-transfection. Cells were then infected with VHSV (MOI
   1 × 10^−2) for 24h at 14°C. The positive control is non-pretreated
   RTG-2 cells infected with VHSV. VHSV infectivity was evaluated by means
   of focus forming units (FFU)/mL. FFUs were stained with 2C9 antibody
   against the N protein of VHSV. Data are displayed as mean ± SD (n = 3).
   The Mann Whitney test was performed between treatments at each
   dilution. (C) ifn1 and mx gene expression in RTG-2 cells quantified by
   RT-qPCR after treatment with CM from pmTFP1- (black bars) and
   pmTFP1GVHSV- (gray bars) transfected RBCs (diluted 1/5 in MEM 10% FBS)
   at three and six days post-transfection. The eukaryotic 18S rRNA gene
   was used as an endogenous control. Data are displayed as mean ± SD (n =
   3). The Mann Whitney test was performed between conditions. ^*, ^**,
   and ^****, represent P-values < 0.05, < 0.01, and < 0.0001,
   respectively.

   To determine whether this protection was due to the stimulation of type
   1 interferon signaling in RTG-2 cells, we evaluated the expression of
   ifn1 and interferon-inducible mx genes in RTG-2 cells incubated with
   pmTFP1- or pmTFP1GVHSV-transfected RBC CM (Figure [165]7C). We observed
   significant upregulation of ifn1 gene expression in RTG-2 cells treated
   with pmTFP1GVHSV-transfected RBC CM at six days post-transfection and
   of mx gene expression at three and six days post-transfection.

Crosstalk between transfected RBCs and RTS11

   In order to evaluate whether pmTFP1GVHSV-transfected RBC CM could
   induce monocyte/macrophage differentiation, we co-incubated transfected
   RBCs with RTS11, a monocyte/macrophage-like cell line. At three days
   post-transfection, pmTFP1-and pmTFP1GVHSV-transfected RBCs were
   co-cultured with RTS11 cells for three days. Using RT-qPCR, we observed
   slight, but not significant, upregulation of RTS11 differentiation
   markers The slight upregulation of the interleukin 10 (il10) gene, a
   marker of M2 macrophages ([166]38) was accompanied by a slight
   downregulation of interleukin 12 subunit beta (il12β) and inducible
   nitric oxide synthase (inos), which are markers of M1 macrophages
   (Figure [167]8).

Figure 8.

   Figure 8
   [168]Open in a new tab

   M1 and M2 macrophage markers in RTS11 cells co-cultured with
   GVHSV-transfected RBCs. Rainbow trout RBCs transfected with pmTFP1 or
   pmTFP1GVHSV for three days at 14°C. Afterwards, transfected RBCs were
   co-cultured with the RTS-11 cell line at 14°C for three days. Then, the
   il12β, inos, and il10 gene expression profiles were quantified by
   RT-qPCR. Gene expression was normalized against eukaryotic 18S rRNA and
   compared to control cells (RTS11 co-cultured with pmTFP1-transfected
   RBCs, red line) (fold-change). Data represent the mean ± SD (n = 6). A
   Wilcoxon test was performed between both conditions.

Discussion

   Recent studies have implicated nucleated RBCs in the immune response to
   viral infections in fish, as these cells are able to actively
   transcribe and transduce signaling molecules in response to viral
   attack ([169]2). Moreover, although DNA vaccines are mainly delivered
   via intramuscular injection, non-nucleated RBCs are thought to be
   promising drug and vaccine carriers ([170]39, [171]40, [172]41,
   [173]42) by eliciting humoral immune responses comparable or superior
   to those obtained via the subcutaneous vaccination route ([174]39). In
   this study our aim was to elucidate the role of nucleated RBCs in the
   immune response to DNA vaccines in order to explore their usefulness in
   improving immune response to DNA vaccines in fish.

   As far as we know, this is the first report of fish nucleated RBCs
   expressing the antigen encoded by a DNA vaccine in vitro. Besides, we
   have not found any report showing nucleated RBCs expressing the protein
   encoded by a DNA vaccine in vivo. GVHSV-transfected RBCs showed a
   characteristic perinuclear expression of GVHSV protein that appeared in
   the perinuclear region of stressed RBCs or on the membrane of
   shape-shifted RBCs (shRBC) ([175]33). shRBCs are small, round cells
   with a thin membrane derived from RBCs subjected to stressful
   conditions. Translation of the GVHSV gene into protein and
   translocation to the cell membrane induces RBC transformation into
   stressed RBC and shRBC ([176]33). Like RBCs, shRBCs have been shown to
   participate in roles related to immune response and homeostasis
   ([177]33).

   Transcriptome profiling of single-cell sorted GVHSV-expressing RBCs
   demonstrated the capacity of RBCs to modulate the expression of genes
   related to innate and adaptive immune responses in response to the DNA
   vaccine. Among the pathways upregulated in single-cell sorted
   GVHSV-expressing RBCs, GPCR-downstream signaling and RORA activates
   gene expression were the most represented. GPCRs constitute a large
   protein family of receptors that detect molecules outside the cell and
   activate internal signal transduction pathways ([178]43). GPCRs are
   known to play a role in inflammation and are widely targeted in drug
   discovery ([179]44). Immune cells such as monocytes and macrophages
   express a large number of GPCRs for classic chemoattractants and
   chemokines ([180]44, [181]45). These receptors are critical for
   phagocyte migration and accumulation at sites of inflammation, where
   these cells can intensify inflammation or contribute to its regulation
   ([182]44, [183]45). Among the molecules found downstream of GPCRs,
   β-arrestins are known to be key regulators of GPCR signaling through
   interaction with the Iκβα component of the NF-κβ signaling complex
   ([184]44, [185]45). Consistent with these effects on signaling,
   knockdown of ARRB2 has been described to enhance the expression of the
   NF-κβ target proteins IL6 and IL8 in response to proinflammatory
   stimulus ([186]46). After RNA sequencing and RT-qPCR, we observed that
   the arrb2 gene was significantly upregulated in GVHSV-transfected RBCs.
   We also observed that the IL1β, IL8, and TNFα proteins were slightly
   upregulated in response to GVHSV transfection. However, the il8 and
   tnfα genes appeared to be highly upregulated in single-cell sorted
   GVHSV-expressing RBCs, and the il6 gene appeared to be highly
   downregulated. In this regard, it would be interesting to study the
   implication of β-arrestins in RBCs regulation of the proinflammatory
   response. Considering the RORA activates gene expression pathway
   (upregulated in single-cell sorted GVHSV-expressing RBCs), it has been
   described that RORA is a nuclear receptor highly expressed in Th17
   cells that regulates differentiation of Th17 cells ([187]47). Moreover,
   RORA depletion has been reported to attenuate cytokine production
   ([188]48) and has thus demonstrated its involvement in inflammatory
   responses. The signaling paradigms of GPCRs and RORA in inflammatory
   regulation and immune cell differentiation in nucleated RBCs remain to
   be studied and are part of our ongoing research.

   Interestingly, GVHSV-transfected RBC CM could switch RTS11
   monocyte/macrophage differentiation markers, upregulating the il10 gene
   [a marker of M2 macrophages ([189]38)] and downregulating the il12β and
   inos genes [markers of M1 macrophages ([190]38)]. M1 macrophages are
   known to be activated by LPS and IFNγ and secrete high levels of IL12
   and low levels of IL10. On the other hand, M2 macrophages are
   alternatively activated by certain cytokines such as IL4, IL10, or IL13
   and produce high levels of IL10 and TGFβ and low levels of IL12. Their
   function is implicated in constructive processes like wound healing and
   tissue repair and in anti-inflammatory responses. The cytokines or
   molecules secreted by GVHSV-transfected RBCs responsible for inducing
   M2 macrophages markers in RTS11 are unknown and should be further
   investigated.

   Genes related to the complement pathway were also overrepresented in
   single-cell sorted GVHSV-expressing RBCs. The complement system is an
   essential part of the innate immune response and acts as a connection
   between innate and acquired immunity [reviewed in Nesargikar et al.
   ([191]49)]. The complement system is known to mediate responses to
   inflammatory triggers, leading to clearance of foreign cells through
   pathogen recognition, opsonization and lysis ([192]50). On the other
   hand, genes and proteins related to proteosomal cleavage of exogenous
   antigen and antigen presentation of exogenous peptides were also
   upregulated in GVHSV-transfected RBCs (via MHCI or MHCII pathways),
   indicating that RBCs may have the capacity to present DNA vaccine
   antigens as has been recently reported ([193]26). Further research is
   needed to determine whether RBCs are functionally capable of inducing T
   cell activation upon antigen presentation on their membrane.

   Proteomic sequencing of GVHSV-transfected RBCs, a sample containing
   both few GVHSV-expressing RBCs and non–GVHSV-expressing RBCs, revealed
   the upregulation of ISG15 antiviral mechanisms. ISG15 is a member of
   the ubiquitin-like (UBL) family. ISG15 conjugates with several target
   proteins in a process termed ISGylation. Hundreds of target proteins
   have been identified in ISGylation. Among them, several proteins that
   are part of antiviral signaling pathways, such as Mx1 or PKR, have been
   identified as targets for ISGylation ([194]51). Upregulation of the
   ISG15 antiviral mechanism pathway was confirmed by gene expression
   analysis of effector molecules within the pathway such as mx, pkr, and
   ifit5. The Mx, PKR, and IFIT proteins are known interferon-inducible
   antiviral effectors ([195]51). Surface expression of the GVHSV protein
   by GVHSV-transfected cells has been reported to be a major mechanism of
   interferon induction ([196]52), and VHSV infection and GVHSV
   vaccination have been demonstrated to induce ISGs such as isg15
   ([197]53) and mx ([198]16, [199]17). However, ifn1 appeared to be
   downregulated in GVHSV-transfected RBCs by RT-qPCR and in single-cell
   sorted GVHSV-expressing RBCs by RNA sequencing. Also, RNA sequencing
   data showed the downregulation of genes related to interferon
   alpha/beta signaling such as irf1, ifnar1, and jak1. However, VHSV has
   been reported to induce ifn1 downregulation in rainbow trout RBCs
   ([200]8). These differences between ifn1 and ISG gene expression could
   be due to the effort of the immune system to maintain homeostasis or to
   the differential regulation of these genes. Alternatively, it has been
   reported that Mx induction could be independent of interferon in HIV
   infection ([201]54, [202]55). In addition, infectious salmon anemia
   (ISA) virus could trigger mx and isg15 stimulation but not ifn1 gene
   expression, suggesting ISG stimulation independent of interferon
   ([203]56). Despite the fact that ifn1 gene expression was downregulated
   in GVHSV-transfected RBCs, the IFN protein, which could be differently
   expressed to inf gene, or other cytokines or molecules secreted by
   GVHSV-transfected RBCs, were able to stimulate ifn1 and mx gene
   expression as well as induce protection against VHSV infection in RTG-2
   cells.

   Another interesting pathway identified during the proteomic profiling
   of GVHSV-transfected RBCs was detoxification of reactive oxygen species
   (ROS). Gene expression of antioxidant enzymes such as sod1, nkef ,
   prdx6, and trx appeared to be upregulated in GVHSV-transfected RBCs.
   This mechanism has been reported in rainbow trout RBCs exposed to VHSV
   ([204]8), where protective antioxidant enzymes were implicated in the
   response of RBCs to the induction of ROS after viral exposure. However,
   in this study, it is important to note that the antigen GVHSV encoded
   by this DNA vaccine is able to induce ROS signaling and homeostasis.

   In summary, rainbow trout nucleated RBCs were able to induce immune
   responses to the DNA vaccine and send signals to neighboring cells or
   other cell types. This reveals a new approach to explore the function
   of RBCs in the complex teleost immune system and could prompt
   development in the field of vaccination with RBCs as targets or carrier
   cells for immunostimulation. Future studies will be focused on the
   molecules of interest produced by GVHSV-expressing RBCs in order to
   identify future vaccination targets.

Ethics statement

   Experimental protocols and methods of the experimental animals were
   reviewed and approved by the Animal Welfare Body and the Research
   Ethics Committee at the University Miguel Hernandez (approval number
   2014.205.E.OEP; 2016.221.E.OEP) and by the competent authority of the
   Regional Ministry of Presidency and Agriculture, Fisheries, Food and
   Water supply (approval number 2014/VSC/PEA/00205). Besides, all methods
   were carried out in accordance with the Spanish Royal Decree RD 53/2013
   and EU Directive 2010/63/EU for the protection of animals used for
   research experimentation and other scientific purposes.

Author contributions

   SP-M performed experiments, analyzed data, and wrote the manuscript. IN
   and VC performed experiments. SC and MM performed proteomic sequencing.
   LM provided valuable antibodies for the experiments. MO-V conceived
   ideas, oversaw the research, and co-wrote the manuscript. VC and JC
   contributed to the preparation of the manuscript.

Conflict of interest statement

   The authors declare that the research was conducted in the absence of
   any commercial or financial relationships that could be construed as a
   potential conflict of interest.

Acknowledgments