Despite teleost fish being the first animal group in which all elements of adaptive immunity are present, the lack of follicular structures, as well as the fact that systemic Ab responses rely exclusively on unswitched low-affinity IgM responses, strongly suggests that fish B cell responses resemble mammalian B1 cell responses rather than those of B2 cells. In line with this hypothesis, in the current study, we have identified a homolog of CD5 in teleost fish. This pan-T marker belonging to the scavenger receptor cysteine-rich family of receptors is commonly used in mammals to distinguish a subset of B1 cells. Subsequently, we have demonstrated that a very high percentage of teleost IgM+ B cells express this marker, in contrast to the limited population of CD5-expressing B1 cells found in most mammals. Furthermore, we demonstrate that fish IgM+ B cells share classical phenotypic features of mammalian B1 cells such as large size, high complexity, high surface IgM, and low surface IgD expression, regardless of CD5 expression. Additionally, fish IgM+ B cells, unlike murine B2 cells, also displayed extended survival in cell culture and did not proliferate after BCR engagement. Altogether, our results demonstrate that although fish are evolutionarily the first group in which all the elements of acquired immunity are present, in the absence of follicular structures, most teleost IgM+ B cells have retained phenotypical and functional characteristics of mammalian B1 cells.
In mammals, three main subsets of B cells have been identified: B1 cells, marginal zone (MZ) B cells, and follicular B2 cells. Follicular B2 cells, which constitute the largest B cell compartment in adult mammals, are generally activated in response to T-dependent Ags within the lymphoid follicles and trigger the formation of germinal centers (GCs). These sites promote the close collaboration between proliferating Ag-specific B cells, T follicular helper cells, and specialized follicular dendritic cells that constitutively occupy the central follicular zones of secondary lymphoid organs. In this environment, B cells divide in response to Ags and acquire the capacity to differentiate into Ab-secreting cells, plasmablasts, terminally-differentiated plasma cells, or memory B cells. Both plasma cells and memory B cells have the capacity to secrete high-affinity Abs. Within the GC, isotype switching recombination takes place to remove the variable H chain segment (VDJ) of IgM and associate it with a different constant (CH) region, consequently secreting Ab isotypes with increased Ag affinity such as IgG, IgE, or IgA (1). In parallel, the genes encoding the variable domains of the heavy and light chains undergo a high rate of point mutations by a process called somatic hypermutation (SHM). SHM results in an increased diversity of the Ab pool after which only the cells with higher affinity are selected by the Ag. This diversification of the Ig genes is critical for the generation of an adequate specific immune protection (2).
Apart from these conventional B cells stand B1 cells and MZ B cells, usually classified as elements of the innate immune system (3). MZ B cells (∼15% of the splenic B cell population) are strategically located in a highly vascularized area of the spleen, and they are specialized in responses to bloodborne pathogens and are thought to represent some type of memory B cell (4). B1 cells, in contrast, are responsible for the production of more than 80% of the natural serum Abs (5). These Abs have a low affinity and wide reactivity, are spontaneously and constitutively expressed, and form a preexisting shield to infection during the early stages of pathogen replication until a specific B2 response is mounted (3). In mice, B1 cells are predominantly located in the peritoneal cavity but are also found in the spleen (accounting for <2% of the spleen B cell population), the pleural cavity, and the bone marrow and are either absent or very poorly represented in lymph nodes or blood (6, 7). B1 cells were first identified on the basis of their surface expression of CD5, a type I glycoprotein, which belongs to the scavenger receptor cysteine-rich (SRCR) family of receptors and is one of the main T cell surface molecules (6). In mice, B1 cell expression of CD5 correlates with the expression of other distinctive markers (IgMhiIgDloCD45loCD23lo/− and CD9+ shared with MZ cells) (3). Further analysis, however, revealed that CD5 surface expression is not a consistent characteristic of B1 cells, as two distinct populations were found among B1 cells. B1a cells (∼2% of the total splenic population) exhibit the phenotype described above, whereas B1b cells do not express CD5 on their surface but bear a similar expression pattern of all other surface markers (<1% of the total splenic population) (8). A further distinction based on surface markers can be made depending on their location, as peritoneal and pleural cavity B1 cells express CD11b but not splenic B1 cells or conventional B2 cells (8). Depending on their origin and phenotype, these different B1 cell subsets also exhibit some functional differences (9, 10). Although CD5 has been used as a B1 cell marker in other mammalian species (11, 12), CD5 does not seem to be an exclusive marker of B1 cells in humans, in which CD5 is expressed on some activated B2 cell populations (3). Thus, in humans, B1 populations with similar characteristics to murine B1 cells have been identified as CD3−CD19+CD20+CD27+CD43+CD69−CD70− cells (13).
Although evolutionarily jawed fish constitute the first group of animals in which adaptive immunity based on Ig receptors is present (14), many structural immune peculiarities anticipate important functional differences between fish and mammalian B cells. In the absence of lymph nodes, the teleost spleen constitutes the main secondary immune organ. However, the splenic white pulp is poorly developed in comparison with mammals, and no GCs have ever been visualized (15), pointing to an extrafollicular origin of all fish B cell responses. Moreover, fish comprise only three Ig classes: IgM, IgD, and IgT (designated as IgZ in some species). Because IgT is a teleost fish–specific Ig that seems specialized in mucosal immunity (16, 17) and IgT+ B cells constitute a linage of B cells independent from IgM+ B cells (16), no class switch recombination has ever been reported in fish. Additionally, although the existence of specific IgMs has been demonstrated in all teleost species studied so far (18–20), the role of these specific IgMs in protection against infectious agents, especially viruses, is controversial. Teleosts seem to have a more limited repertoire than mammals (21), and although SHM has been reported, the Ab response only reached a 2–3-fold higher affinity after 90 d (22), demonstrating that affinity maturation is much lower in fish than that seen in mammals. Thus, although the protective role of serum Abs has been established in passive immunization with either sera or purified IgM from rainbow trout surviving infections (23, 24), we know that in the context of a natural viral infection, trout virus-specific IgMs peak after 6–10 wk (23, 25), far after mortalities cease (usually 1 wk postinfection). These observations indicate that the specific neutralizing Abs produced as response to an infection in fish appear too late to play any role in protection and suggest that the natural resistance to a viral disease is dependent on innate responses, which might include the production of natural Abs, known to interfere with fish viruses (26). Moreover, neutralizing Abs can only be detected in a fraction of the survivor fish, and there is no detectable increase in their levels after a posterior encounter with the Ag (25, 27), revealing a limited secondary response.
In this context, we hypothesized that fish B cells might not be the equivalent of mammalian B2 cells but closely resemble extrafollicularly activated mammalian B1 cells, as already suggested by other authors (28, 29). To provide additional evidence that might help us establish to what extent this hypothesis can be confirmed, we identified in rainbow trout (Oncorhynchus mykiss) a homolog of CD5 and demonstrated that a major fraction of trout splenic B cells express CD5 on the cell surface. Additionally, we identified a homolog of CD6, another member of the SRCR family of receptors recently described to be selectively expressed on mammalian splenic B1a cells (30) and T cells (31), demonstrating that fish B cells also transcribe CD6. Further phenotypical and functional characterization of fish B cells in parallel to murine splenic B2 cells clearly established a resemblance of fish IgM+ B cell responses to mammalian B1 cell responses regardless of CD5 surface expression. As teleost fish represent one of the earliest groups to exhibit adaptive immune responses, our results collectively demonstrate that in the early stages of adaptive immunity, IgM+ B cells exhibited a phenotypic and functional profile that was retained in mammalian B1 cells, which govern the B cell responses in the early phases of mammalian development.
Materials and Methods
Sequence retrieval and analysis
Using the basic local alignment search tool (BLAST) (32) to search the nonredundant protein sequence database at National Center for Biotechnology Information (NCBI), a trout sequence that coded for a CD5 (accession number, GBTD01068436.1; https://www.ncbi.nlm.nih.gov/genbank/) and a CD6 homolog (accession number, CDR01152.1; https://www.ncbi.nlm.nih.gov/genbank/) were identified. Subsequently, primers were designed to confirm the open reading frame (ORF) of the trout CD5 gene. The multiple sequence alignment was generated using ClustalX v1.81 (33), and the presence of a transmembrane domain and a signal peptide were predicted using TMHMM Server v2.0 (34) and SignalP v4.1 (35), respectively. For the phylogenetic analysis, CD5, CD6, and CD5 Ag-like (CD5L) amino acid sequences identified from mammals, reptiles, birds, amphibians, and fish were used. Sequences that were not previously published were found using the FASTA (36) and BLAST (32) suite of programs to search the nonredundant protein sequence database at NCBI or by searching available genomes within Ensembl (37) and using GeneID (38) and Genscan (39) to predict gene coding regions. Phylogenetic relationships were constructed from ClustalX v1.81 (33) generated alignments of amino acid sequences using the neighbor-joining method (40). The tree was drawn using TreeView v1.6.1 (41) and bootstrapped 1000 times (42).
CD5 gene organization and chromosome synteny
Ensembl (37) was used to locate regions of the human (genome assembly: GRCh37.p13), mouse (genome assembly: GRCm38.p5), wallaby (genome assembly: Meug_1.0), and chicken (genome assembly: Gallus gallus-5.0) genomes containing the CD5 and CD6 genes to analyze the gene organizations of the CD5 genes and determine the gene order around these genes for the synteny analysis. Regions of the salmon and trout genomes containing CD5 and CD6 were obtained using BLAST (32) to search the whole-genome shotgun (WGS) contigs sequence database at NCBI. The trout and salmon CD5 cDNA was aligned with the gDNA using Splign (43) to obtain the gene organizations. Genscan (39), BLAST (32), and FASTA (36) were used to annotate the gene order around the CD5 and CD6 genes in the genomes of the wallaby, chicken, salmon, and trout and compared with human chromosome 11 to determine the level of synteny conserved.
Healthy specimens of rainbow trout (O. mykiss) of ∼50–70 g were obtained from Centro de Acuicultura El Molino (Madrid, Spain). Fish were maintained at the Animal Health Research Center (Centro de Investigación en Sanidad Animal–Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria [INIA]) laboratory at 16°C with a recirculating water system and 12:12 h light/dark photoperiod. Fish were fed twice a day with a commercial diet (Skretting, Spain). Prior to any experimental procedure, fish were acclimatized to laboratory conditions for 2 wk, and during this period, no clinical signs were ever observed. Inbred FVB mice were maintained under pathogen-free conditions at the Centro de Investigación en Sanidad Animal–INIA animal facility until they became 40 d old. All the experiments described comply with the guidelines of the European Union Council (2010/63/EU) for the use of laboratory animals and were previously approved by the ethics committee from INIA (Code CEEA 2011/044).
Tissue collection from perfused trout
Fish were killed by benzocaine (Sigma-Aldrich) overdose. Blood was extracted from the caudal vein of freshly killed rainbow trout using a heparinized needle/syringe. Transcardial perfusion using teleost Ringer’s solution (pH 7.4) containing 0.1% procaine was undertaken to remove all blood from fish tissues as previously described (44). Thymus, head kidney, gills, spleen, intestine, skin, gonad, muscle, heart, and liver samples were then collected and placed in TRIzol (45, 46).
Gene expression in trout tissues
DNase I–treated total RNA was prepared from tissue samples using a combination of TRIzol (Invitrogen) and an RNeasy Mini Kit (Qiagen) as described previously (47). Total RNA was eluted from the columns in RNase-free water, quantified using a NanoDrop 1000 spectrophotometer (Thermo Scientific), and stored at −80°C. For each sample, 2 μg of total RNA was reverse transcribed using BioScript Reverse Transcriptase (Bioline Reagents) primed with oligo(dT)12–18 (0.5 μg/ml) following the manufacturer’s instructions. cDNA was diluted in nuclease-free water and stored at −20°C.
To evaluate the levels of transcription of CD5 and CD6, real-time PCR was performed in a LightCycler 96 System instrument (Roche) using FastStart Essential DNA Green Master reagents (Roche) and specific primers. The following primers sets were used: 5′-GATCCAGAAGGAGGTCACCA-3′ and 5′-TTACGTTCGACCTTCCATCC-3′ for EF-1α, 5′-ATCTTAGCCTTGAACTGGCAGAATC-3′ and 5′-ACAGGAGTTATTGGTAACCCTTCCA-3′ for the amplification of CD5, and 5′-GGACGTGTAGAGCTGTGGAGAGA-3′ and 5′-CCCTGTCACACTCAAAGCATACC-3′ for the amplification of CD6. The efficiency of the amplification was determined for each primer pair using serial 10-fold dilutions of pooled cDNA, and only primer pairs with efficiencies between 1.95 and 2 were used. Each sample was measured in duplicate under the following conditions: 10 min at 95°C followed by 40 amplification cycles (30 s at 95°C and 1 min at 60°C). The expression of individual genes was normalized to relative expression of trout EF-1α, and the expression levels were calculated using the 2−ΔCt method, in which Δ threshold cycle (Ct) is determined by subtracting the EF-1α value from the target Ct as described previously (48, 49). Negative controls with no template were included in all the experiments. A melting curve for each PCR was determined by reading fluorescence every degree between 60 and 95°C to ensure only a single product had been amplified.
Rainbow trout were killed by benzocaine overdose, and blood was extracted from the caudal vein using a heparinized needle and diluted 10 times with Leibovitz medium (L-15; Life Technologies) supplemented with 100 IU/ml penicillin, 100 μg/ml streptomycin, 10 U/ml heparin, and 5% FCS (all supplements also obtained from Life Technologies). Peritoneal lavage was achieved by gently pipetting of medium in the peritoneal cavity to collect the resident leukocytes. Spleen and kidney were collected, and single-cell suspensions were generated using 100-μm nylon cell strainers (BD Biosciences). Blood cell suspensions were placed onto 51% Percoll (GE Healthcare) cushions, whereas kidney, spleen, and peritoneal suspensions were placed onto 30/51% discontinuous density gradients. All suspensions were then centrifuged at 500 × g for 30 min at 4°C. The interface cells were collected and washed twice with L-15 containing 5% FCS.
Mice were killed by cervical dislocation, and the spleen was collected. Single-cell suspensions were prepared by mechanical disruption through 100-μm nylon cell strainers. After lysis of erythrocytes with distilled water, splenocytes were cultured in RPMI medium, 10% FCS, 2% glutamine, 1% sodium pyruvate, 1% HEPES, and 0.1% 2-ME.
Cell sorting of trout leukocyte populations
Trout IgM+ B cells were FACS isolated from spleen, peritoneum, blood, or kidney leukocyte suspensions using a BD FACSAria III (BD Biosciences) cell sorter. For this, leukocytes were incubated for 30 min on ice with an anti-trout IgM mAb (1.14) coupled to PE (R-PE) in FACS staining buffer (PBS containing 1% FCS and 0.5% sodium azide) to prevent cell activation. Following two washing steps, cells were resuspended in FACS buffer, and IgM+ B cells were isolated based on their forward scatter (FSC)/side scatter (SSC) profiles (to exclude the granulocyte gate) and then on the basis of the fluorescence emitted by the anti-trout IgM Ab.
To determine whether CD5 and CD6 were also expressed in T cells, we used T cell–enriched splenocyte and PBL cultures, given that no Abs are available against extracellular pan-T cell markers in rainbow trout. These T cell–enriched cultures were obtained by depleting all IgD+, IgM+, and MHC class II+ lymphoid cells through cell sorting as previously described (50). The enrichment in T cells was assessed by intracellular staining with an anti-CD3 Ab (51), which showed more than 90% of CD3+ cells in the cultures.
Real-time PCR analysis of FACS-isolated leukocyte populations
Total RNA was isolated from FACS-isolated leukocyte populations from the different tissues using the Power SYBR Green Cells-to-Ct Kit (Life Technologies) following manufacturer’s instructions. For comparative purposes, RNA was also isolated from the RTS11 rainbow trout macrophage–monocyte cell line (52). RNAs were treated with DNase during the process to remove genomic DNA that might interfere with the PCR reactions. Reverse transcription was also performed using the Power SYBR Green Cells-to-Ct Kit (Invitrogen) following manufacturer’s instructions. To evaluate the levels of transcription of the different genes, real-time PCR was performed with a LightCycler 96 System instrument using SYBR Green PCR Core Reagents (Applied Biosystems) and specific primers as described above. Each sample was measured in duplicate under the following conditions: 10 min at 95°C, followed by 45 amplification cycles (15 s at 95°C and 1 min at 60°C). A melting curve for each PCR was also included to ensure that only a single product had been amplified. The expression of individual genes was normalized to the relative expression of trout EF-1α, and the expression levels were calculated as described above.
Anti-CD5 Ab production
A customized mAb was produced by Naxo (Tartu, Estonia) using a selected specific peptide from the trout CD5 sequence as an Ag (PLTPSDSTPATPA). To verify the specificity of the Ab, rainbow trout spleen protein lysates were used to test the specificity of the Ab through Western blotting as previously described (47). A further verification of the Ab specificity was performed blocking the anti-CD5 mAb with the specific synthetic peptide used for immunization. For this, a 30-min blocking preincubation of the mAb with the specific peptide at 150:1, 75:1, and 15:1 molar ratios was performed before cell staining for flow cytometry.
Flow cytometry analysis
To analyze the percentage of CD5+ IgM+ B cells, leukocytes were incubated with anti-CD5 (hybridoma supernatant 1:1) and anti-trout IgM (1.14 mAb mouse IgG1 coupled to allophycocyanin, 1 μg/ml) (53) for 30 min. Cells were then washed three times with staining buffer (PBS containing 1% FCS and 0.5% sodium azide) and stained for 20 min with a secondary Ab for CD5 (Alexa Fluor 488 goat anti-mouse IgM [Life Technologies]). After incubation, cells were washed three times with staining buffer and analyzed on a FACSCalibur flow cytometer (BD Biosciences) equipped with CellQuest Pro software.
In some experiments, the levels of expression of surface IgD were determined using an anti-trout IgD Ab (mAb mouse IgG1 coupled to R-PE, 5 μg/ml) previously characterized (54) or an allophycocyanin anti-mouse IgD Ab (BioLegend) in the case of mice. In mice, all phenotypic analyses were performed differentiating between the IgM+CD9− cell population, which excludes B1 and MZ cells that express CD9 on the membrane, and IgM+CD9+ cells. For this, anti-IgM (Alexa Fluor 488 anti-mouse IgM; Life Technologies) and anti-CD9 (Biotin Rat Anti-Mouse CD9; BD Biosciences) Abs were combined. After 30 min of incubation, cells were washed three times and PE Streptavidin (BD Biosciences) was added.
In all cases, isotype controls for mAbs (BD Biosciences) were tested in parallel to discard unspecific binding of the Abs. All the incubations were performed at 4°C. During the setting up of the experiments, cell viability was checked by propidium iodide (PI) staining. Cell viability was always higher than 95% in our experimental conditions.
B cell survival
Trout and mice splenocytes dispensed in 24-well plates at a density of 2 × 106 cells/ml were cultured for 3 d with the corresponding medium (described above) and at the corresponding temperature (20°C for fish and 37°C for mice). IgM+ B cell survival was evaluated at time 0 h and days 1, 2, and 3. To assess cell death, trout splenocytes were stained with anti-trout IgM-Alexa Fluor 488 (1.14). In some experiments, cells were also stained with the anti-trout CD5 Ab as described above. To assess B2 cell survival in murine splenocyte cultures, an anti-mouse IgM-Alexa Fluor 488 was combined with a Biotin Rat Anti-Mouse CD9 as described above. In all cases, after the staining procedures, 1 μg of PI (Invitrogen) was added, and cells were immediately analyzed by flow cytometry (BD FACSCalibur; BD Biosciences).
B cell proliferation
IgM+ B cell proliferation was assessed in response to anti-IgM (10 μg/ml) or LPS (100 μg/ml) at concentrations previously optimized. Nonstimulated controls were always included. The BrdU Flow Kit (BD Biosciences) was used to measure the proliferation of IgM+ B cells following manufacturer’s instructions. Trout or mice splenocytes dispensed in 24-well plates at a density of 2 × 106 cells/ml were incubated for 3 d at 20 and 37°C, respectively, with the different stimuli or with media alone in the case of controls. Bromodeoxyuridine (10 μM) was then added to the cultures, and the cells were incubated for an additional 24 h. After that time, cells were collected and stained with anti-trout anti-IgM–PE (1.14) and anti-mouse biotin-conjugated anti-IgM (Thermo Scientific) followed by Streptavidin-PE in the case of murine cells and then fixed and permeabilized with Cytofix/Cytoperm Buffer for 15 min on ice. Afterwards, cells were incubated with Cytoperm Permeabilization Buffer Plus for 10 min on ice and refixed with Cytofix/Cytoperm Buffer for 5 min at room temperature (RT). Cells were then incubated with DNase (30 μg/106 cells) for 1 h at 37°C to expose the incorporated BrdU. Finally, cells were stained with FITC anti-BrdU Ab. After 20 min of incubation at RT, the cells were analyzed by flow cytometry to determine the percentage of BrdU+ cells.
For calcium flux analysis, the calcium indicator Fluo-3 AM (Life Technologies) was used following the manufacturer’s instructions. Briefly, Fluo-3 was dissolved in DMSO and further diluted in an equal volume of 20% (w/v) Pluronic F-127 (Life Technologies). Trout and mice splenocytes were diluted in their specific culture media without FCS and incubated with Fluo-3 AM at a final concentration of 5 μM for 1 h. Cells were then collected and washed, and a baseline reading for 60 s was acquired in a FACSCalibur flow cytometer. Thereafter, 10 μg/ml anti-IgM was added to each sample, and the emission of fluorescence (525 nm) was determined for 360 s.
Six- to eight-week-old female C57BI/6 mice and rainbow trout were euthanized, and spleens were processed to obtain splenocytes. Murine splenocytes were isolated and seeded into complete RPMI medium (RPMI supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, 100 μg/ml gentamicin, and 10% newborn calf serum; Life Technologies). Trout splenocytes were isolated and seeded in complete MGFL-15 medium (MGFL-15 supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, 100 μg/ml gentamicin, 10% newborn calf serum [Life Technologies], and 5% carp serum) (55). Splenocytes were incubated for 30 min or 3 h in the presence of anti-mouse or anti-trout IgM (10 μg/ml). Incubation with Salmonella spp. was used as a positive control for cell activation. Following stimulation, cells were fixed in 1% formaldehyde and washed twice in PBS with 2% calf serum and 0.1% saponin (permeabilization buffer). In murine splenocyte cultures, cells were stained with a Biotin Rat Anti-Mouse CD9 to determine the NF-κB nuclear translocation in murine B2 cells exclusively. In some experiments, trout cells were also stained with the anti-trout CD5 Ab as described above. In all cases, to determine NF-κB nuclear translocation, splenocytes were stained with unlabeled rabbit IgG anti-mouse p65 (Santa Cruz Biotechnology) for 30 min at 4°C followed by 20 min at RT. Following the primary staining, cells were washed and stained with a rabbit anti-mouse FITC secondary Ab (Jackson ImmunoResearch). Prior to acquisition, DRAQ5 nuclear stain (Molecular Probes) was added as per manufacturer’s recommendations. Data were collected on an ImageStream MK II and analyzed using IDEAS software (Amnis), as previously described (56–58).
Statistical analyses were performed using a two-tailed Student t test with Welch’s correction when the F test indicated that the variances of both groups differed significantly. The differences between the mean values were considered significant on different degrees, in which *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.005 (GraphPad Prism 4 software).
Analysis of the trout CD5 amino acid sequence
Using BLAST to search the nonredundant protein sequence database at NCBI, a trout sequence that coded for a CD5 (accession number, GBTD01068436.1) and a CD6 homolog (accession number, CDR01152.1) were identified. Subsequently, primers were designed to confirm the ORF of the trout CD5 gene. PCR confirmed the ORF of trout CD5 to be 798 bp encoding a predicted protein of 265 aa. Multiple alignment of the trout and salmon protein sequences with the known mammalian and bird sequences indicated the presence of a signal peptide region, no proline-rich region, one SRCR domain, a transmembrane region, and a cytoplasmic region (Fig. 1). Nine matching cysteines are found within the trout and salmon SRCR domain, with three of them matching some of the SRCR domains found within the mammal and bird domains. The transmembrane and cytoplasmic regions were the most conserved portions of the trout and salmon amino acid sequence when compared with mammalians and birds. Within the cytoplasmic region there are a number of residues conserved between all sequences important in forming potential signaling motifs.
Phylogenetic analysis clearly grouped the trout and salmon CD5 amino acid sequence with other identified and predicted fish CD5 sequences and indicated a close relationship with the mammalian, reptile, and bird sequences (Fig. 2). Trout CD6 formed its own group with other identified and predicted fish sequences away from CD5 and showed a closer relationship to mammalian, reptile, bird, and amphibian CD6 sequences. Last, CD5L identified in mammals, reptiles, and amphibians (59) was also included, but no fish or bird homolog currently exists. This gene codes for a soluble protein mostly produced by macrophages during inflammatory processes (59).
Structure of the trout CD5 gene and synteny analysis
The human, mouse, and wallaby CD5 gene organization consists of 11 exons and 10 introns, whereas the chicken consists of 10 exons and 9 introns but is missing the extra intron found within the 3′ untranslated region in mammals (Fig. 3). The trout CD5 gene organization was predicted using the cDNA sequence obtained by PCR aligned with two trout WGS sequences (CCAF010188581.1; CCAF010006860.1), and the salmon gene organization was predicted by aligning a predicted cDNA sequence (XM_014195396) with a salmon WGS sequence (NC_027303). Both were found to have a gene organization of six exons and five introns and are missing two exons that code for SRCR domains; an exon that codes for a proline-rich domain; and, similar to the chicken organization, the extra intron found within the 3′ untranslated region in mammals (Fig. 3).
Analysis of the genes surrounding CD5 in the salmon and selected fish genomes shows there are a number of genes, which include TMEM132A, VPS32C, and MS4A12, that are also found surrounding CD5 of the mammalian, reptile, and bird genomes observed (Fig. 4). CD6, which is found next to CD5 in the human, anole, and chicken genomes, is found at a completely different region of the genome in the majority of fish genomes analyzed, with no clear synteny, as was seen for the CD5 (Supplemental Fig. 1). However, among a number of the fish, some synteny can be seen (Supplemental Fig. 1). Interestingly, the coelacanth is an exception, having both the CD5 and CD6 genes located next to each other in the genome and showing good conservation of surrounding genes with the human, anole, and chicken genomes (Fig. 4, Supplemental Fig. 1).
CD5 and CD6 transcription in trout tissues and IgM+ B cells
The constitutive transcription of CD5 was studied in different tissues from healthy nonstimulated adult rainbow trout. To avoid contamination of tissues with circulating blood, the fish were transcardially perfused prior to tissue removal. CD5 expression was highest in the thymus and lowest in the liver under basal conditions (Fig. 5A). High CD5 transcription levels were also found in head kidney, gills, spleen, intestine, and peripheral blood leukocytes, whereas intermediate transcription levels were observed in skin, gonad, muscle, and heart (Fig. 5A). Next, we analyzed CD5 transcription in FACS-isolated IgM+ B cells from different sources, observing very high CD5 mRNA levels in B cells from spleen, blood, kidney, and peritoneum (Fig. 5A). As expected, CD5 transcription was also high in T cells from spleen or blood (Fig. 5A), whereas CD5 mRNA levels were nearly undetectable in the RTS11 monocyte–macrophage cell line that was included as a negative control (Fig. 5A).
In the case of CD6, a slightly different transcription pattern was obtained in tissues, supporting the fact that in fish as in mammals there is not a perfect overlap between the expression of CD6 and CD5 (60). The highest CD6 transcription levels were observed in thymus and spleen, whereas intermediate mRNA levels were found in head kidney, gills, intestine, and peripheral blood leukocytes (Fig. 5B). Lower CD6 mRNA levels were obtained in skin, gonad, muscle, and heart (Fig. 5B). CD6 mRNA levels remained undetected in liver but were high in IgM+ B cells from spleen, blood, kidney, and peritoneum (Fig. 5B). Spleen and blood T cells also showed high levels of CD6 transcription, whereas undetectable CD6 mRNA levels were found in the RTS11 monocyte–macrophage cell line (Fig. 5B).
CD5 expression in fish B cells
Our discovery that CD5 was constitutively transcribed in B cell populations from the central immune tissues at very high levels strongly suggested that a major percentage of fish B cells express CD5 on the cell membrane. To confirm this hypothesis, we raised a murine mAb against rainbow trout CD5 and used it to assess CD5 expression in fish B cells. The Ab generated was an IgM isotype Ab. It recognized the peptide used to immunize mice via ELISA (titer higher than 1:512000) and a single protein corresponding to the predicted size through Western blot (Supplemental Fig. 2). The Ab was tested in flow cytometry using blood leukocytes recognizing 20–30% of total lymphocytes (Supplemental Fig. 2). The binding specificity of the Ab to rainbow trout CD5 was also assessed in flow cytometry by blocking the Ag recognition site with the peptide used for the immunization. After the blockage of the Ab, only a residual CD5+ population could be detected (Supplemental Fig. 2), indicating a specific recognition. Additionally, when CD5+ cells and CD5− cells were FACS isolated, the levels of transcription of CD5 were at least 20-fold higher in the former than in the latter. Once the specificity of the Ab had been verified, we combined this Ab with an anti-IgM mAb to determine by flow cytometry whether fish B cells express CD5 on their cell surface. These studies were performed with spleen and blood leukocytes, and we observed that a high percentage of the IgM+ B cells in these tissues express CD5 on the cell membrane (Fig. 6). As expected, CD5+ cells with no IgM on the cell membrane were also identified (Fig. 6). These cells probably correspond to T cells because they account for 5.86 ± 1.41 of the leukocyte population in blood and 12.6% ± 0.96 in the spleen, numbers which are in concordance with the percentages of T cells reported in rainbow trout in these tissues (61). These results demonstrate that unlike the situation in mice in which CD5+ B1 cells represent only a minor percentage of the naive B cells, especially in central immune organs such as spleen (where they account for <2%), ∼50% of splenic IgM+ B cells and 25% of blood IgM+ B cells express CD5 on the cell membrane in trout.
Trout IgM+ B cells have higher complexity and are larger than murine B2 cells
Having established that a large fraction of fish B cell populations expresses CD5 on the cell surface, we analyzed additional phenotypical traits of fish B cells, comparing them to murine splenic B2 cells. To study the phenotype of murine B2 cells, we focused on the murine IgM+CD9− cell population found in the spleen, as CD9 is specifically expressed in the surface of MZ and B1 cells (3). Taking into account that not all B1 populations are expected to express CD5 (8), initially we used total IgM+ B cells and not just CD5+ cells for this phenotypic characterization. As described before (12, 62), murine CD9+ cells (comprising B1 and MZ cells) are larger and more granular than murine B2 cells (Fig. 7A, 7B). Trout B cells showed levels of FSC that refer to size and levels of SSC that refer to complexity; these levels were similar to those of murine CD9+ cells but significantly higher than murine B2 cells (Fig. 7A, 7B). When we compared the size and complexity of trout IgM+CD5+ and IgM+CD5− cells, no significant differences were found (Supplemental Fig. 3).
Trout IgM+ B cells express more membrane IgM and less IgD than murine B2 cells
Other basic features of murine B1 cells include a high IgM and a very low IgD expression on the cell surface. Our experiments revealed that trout IgM+ B cells were IgMhiIgDlo, as murine CD9+ B cells, and in contrast to murine CD9− B2 cells that were IgMloIgDhi (Fig. 7). These results were confirmed by real-time PCR because trout B cells contained higher levels of membrane IgM mRNA than the murine B cell subpopulations (Fig. 7C). When the levels of transcription of membrane IgD were studied, trout B cells had transcription levels lower than murine B2 cells but higher than those observed in murine CD9+ cells (Fig. 7C). Interestingly, although there were no significant differences in the levels of surface IgM between trout IgM+CD5+ and IgM+CD5− B cells, the levels of surface IgD were significantly lower in trout IgM+CD5− cells than in IgM+CD5+ cells (Supplemental Fig. 3).
IgM+ B cells from trout exhibit extended survival in cell culture
In mammals, B1 cells have an extended survival in cell culture, whereas the percentage of B2 cells decreases in cell culture by ∼80% after 3–4 d (63). Thus, we compared the survival of trout IgM+ B cells and mice B2 (IgM+CD9−) cells in splenocyte cultures, assessing the percentage of living (PI−) B2 (mice) or IgM+ (trout) cells among total B cells at days 0, 1, 2, and 3. In mice, initially ∼52% of the cells in the culture were B2 cells. At this moment, ∼96% of the B2 cells in the culture were alive (Fig. 8). This percentage sharply decreased after 3 d in culture, as at this point only ∼17% of the B2 cells in the culture were alive, accounting for an 80% decrease in the number of viable IgM+ B2 cells throughout these 3 d (Fig. 8). In rainbow trout, ∼19% of the cells in the splenocyte cultures were IgM+ B cells initially, being ∼97% of IgM+ B cells alive. The percentage of IgM+ B cells in these cultures only decreased ∼45% in 3 d, with ∼48% of IgM+ B cells alive in the cultures at this moment, values significantly higher than in mice (Fig. 8). The results show that rainbow trout IgM+ B cells have an extended survival in culture such as that seen for mammalian B1 cells. Remarkably, significant differences were found when the survival in the cell culture was compared in trout IgM+CD5+ and IgM+CD5− B cells (Supplemental Fig. 3) because the cells that were not expressing CD5 on the cell membrane had an extended survival when compared with the cells that had CD5 on the surface.
Trout IgM+ B cells do not proliferate after BCR activation
Another property of mammalian B1 cells is that they do not proliferate in response to BCR cross-linking, unlike mammalian B2 cells. Thus, to compare the proliferative responses of trout and mice IgM+ B cells upon BCR activation, we stimulated splenocytes with the corresponding anti-IgM Abs and assessed the percentage of BrdU+ proliferating cells after 4 d. As expected, we observed a significant increase on the proliferation of anti-IgM–stimulated splenocytes in comparison with the control cells in mice (Fig. 9A). In rainbow trout, BCR activation showed no proliferative effects, as the percentage of BrdU+ cells was not increased in cell cultures after 4 d (Fig. 9A), not even when a range of anti-IgM doses (0.1–10 μg/ml) was assayed (data not shown) or when combinations of different anti-IgM Abs were used (data not shown). To confirm that the stimulation with anti-IgM in trout was sufficient to provoke BCR cross-linking, we assessed the mobilization of intracellular calcium in response to anti-IgM, comparing it with that observed in mice. Trout B cells responded to BCR cross-linking with a strong calcium mobilization, as previously described (58), confirming that proliferative responses were not induced in fish B cells despite the effective BCR engagement. In fact, the levels of intracellular calcium mobilization observed in response to BCR cross-linking were higher in fish than in mice (data not shown). To further verify that fish B cells could proliferate in response to stimuli different from BCR cross-linking, we assessed the proliferative response to LPS in parallel, comparing it to that of mice B cells. LPS induced the proliferation of IgM+ B cells in both mice and trout B cells (Fig. 9B).
Trout IgM+ B cells activate NF-κB upon BCR activation
Because murine B1 cells fail to activate NF-κB upon BCR activation (64), we also analyzed the translocation of the NF-κB p65 subunit to the nucleus of splenocytes in response to BCR cross-linking in both murine and fish B cells. We performed this using a quantitative imaging flow cytometry–based approach previously used in rainbow trout (57, 58). In mice, anti-IgM provoked a significant increase in the percentage of CD9− B cells, with p65 translocated to the nucleus after 30 min (Fig. 10). In trout, an equivalent anti-IgM incubation was also capable of significantly increasing the percentage of splenocytes, with p65 translocated to the nucleus after 30 min at levels similar to those observed in mice (Fig. 10). These results were not improved in rainbow trout or mice with longer stimulation periods (data not shown). These results reveal that trout B cells behave similarly to murine B2 cells in what concerns their capacity to activate NF-κB in response to BCR engagement. Furthermore, the capacity of translocating p65 to the nucleus was compared between trout IgM+CD5+ and IgM+CD5− B cells, and no significant differences were found between the two subsets (Supplemental Fig. 3).
Despite teleost fish being the first animal group in which all elements of adaptive immunity are present, the lack of follicular structures and GCs as well as the fact that systemic Ab responses rely exclusively on unswitched low-affinity IgM responses prompted us to hypothesize that fish B cell responses resemble mammalian B1 cell responses rather than those of B2 cells. Recent findings by our group further support this statement, as we have established that fish IgM+ B cells, unlike mammalian B2 cells, constitutively express many pattern recognition receptors (46, 57), strongly respond to proinflammatory signals (65), and transcribe B1 cell markers not expressed in B2 cells such as CD9 (66). Furthermore, a high phagocytic capacity has been demonstrated for teleost B cells (67) and later on for mammalian B1 cells (68). All these results have also led other authors to hypothesize that the mammalian B1 lineage evolved from ectotherm IgM+ B cells and that it was later in evolution that the B2 lineage emerged as a more efficient subset that gradually acquired a dominant role in adaptive immunity (28, 29), a hypothesis that would perfectly correlate with the idea that postulates that evolution has created a layered immune system in which successive progenitors reach predominance during development, giving rise to differentiated cells responsible for progressively more complex immune functions (69). In this context, our main goal was to provide additional evidence to establish to what degree fish IgM+ B cells resemble mammalian B1 cells.
First, we focused on establishing whether fish IgM+ B cells expressed CD5 on the cell membrane, given that CD5 is one of the main markers used for identification of B1a populations in mice and B1 cells in other mammalian species (11, 12, 70). In the current study, we identified a homolog to mammalian CD5 in rainbow trout that constitutes the first report, to our knowledge, of CD5 in teleost fish. CD5 and CD6 belong to the SRCR superfamily (SRCR-SF), which is an ancient and highly conserved group of membrane-bound and/or soluble proteins found from low invertebrates to mammals as well as in some aquatic plants, such as unicellular green algae (71–73). In mammals, SRCR-SF members can be expressed by hematopoietic and nonhematopoietic cells at embryonic and adult developmental stages, depending on species and tissue type, and are reported to be involved in the regulation of innate and adaptive immune responses as well as in the development of the immune system (71, 72). They are characterized by the presence of one or several repeats of highly conserved cysteine-rich extracellular SRCR domains (∼100–110 aa in size). CD5 is a type I membrane glycoprotein whose extracellular region contains three SRCR domains in mammals and birds (71, 74, 75). The 265 aa trout CD5 sequence consists of a putative leader sequence, an extracellular domain, a transmembrane domain, and a cytoplasmic domain. Although, at first glance, the CD5 found within salmon and trout are very different, as they contain only one SRCR domain and no proline-rich like region, there are a number of important features that indicate they are related to CD5 from other species. There is good conservation of synteny found between the fish, mammalian, and bird genomes, and phylogenetic analysis groups the fish CD5 and CD6 sequences with their mammalian, bird, reptile, and amphibian homologs. Members of the SRCR-SF can be divided into two mutually exclusive groups, depending on the number of cysteine residues present in the SRCR domains and the number of exons coding for each domain. Group A SRCR domains are encoded by two or more exons and include six cysteines that form three disulphide bonds, whereas group B SRCR domains are encoded by a single exon and contain eight cysteines forming four disulphide bonds (71, 72). Human and chicken CD5 belong to group B (74, 76), and multiple alignment of the chicken, trout, and salmon CD5 amino acid sequences indicates they also belong to this group, as the SRCR domains are each coded for within one exon and are cysteine rich, containing at least eight cysteines. However, not all the cysteines found in the salmon and trout match well with the mammal and bird sequences, leading to the extracellular domain having the lowest identity. In contrast, the transmembrane and cytoplasmic domains are the most conserved portion of the CD5 amino acid sequence. In humans, the cytoplasmic domain is devoid of any intrinsic catalytic activity but contains conserved motifs involved in signal transduction (77). There are three tyrosine residues (Y429, Y441, and Y463) that become phosphorylated, leading to the recruitment of signaling molecules, such as PI3K, c-Cbl, and RasGAP (78–80). Two threonine residues (T410 and T412) phosphorylated by protein kinase C (81) and a cluster of three serine residues (S458, S459, and S461) phosphorylated by casein kinase II (82) have also been described. All of these are conserved in the cytoplasmic region of the chicken sequence, whereas only some of them appear conserved in the salmon and trout. In mammals, CD5 is physically associated with the BCR (83) and was revealed as a negative regulator of the BCR signaling, as CD5-null mice have restored the capacity of B1 cells to proliferate in response to BCR cross-linking (84). Posterior studies further demonstrated that the CD5 region containing Y429 and Y441 is specifically responsible for antagonizing the early signaling events triggered through the BCR (85), whereas only Y429 is mandatory for the inhibition by CD5 of the calcium response activated via the BCR (86). Interestingly, fish CD5 have conserved Y441 and Y463 but not Y429, and this could be the reason why fish B cells show a strong calcium response upon BCR cross-linking that goes along with a significant NF-κB activation, as this transcription factor highly depended on the mobilization of intracellular calcium (87). Hence, the implications that all the structural differences found between fish and mammalian CD5 have on the capacity of this molecule to modify the BCR responses in fish are complex and deserve further investigation.
After having identified rainbow trout CD5, we verified that trout B cells from different sources (spleen, kidney, blood, and peritoneum) constitutively transcribe CD5 as well as another B1-related marker such as CD6. Altogether, these results suggest that the percentage of B cells expressing CD5 in central lymphoid organs is much higher in fish than in mammals, and to verify this, we obtained a specific mAb against rainbow trout CD5. Through the use of this Ab, we demonstrated that, unlike the situation in mice in which CD5+ B1 cells represent only a minor percentage of the naive B cells in central immune organs such as spleen (6, 7), ∼44% of splenic IgM+ B cells and 30% of blood IgM+ B cells express CD5 on the cell membrane. In mammals, it has been demonstrated that CD5 exerts negative signals on the BCR, probably contributing to set the threshold levels for activation signals (79) and consequently being responsible for the maintenance of tolerance (88). Interestingly, although in mice only 2% of splenic and almost no blood B cells express CD5, CD5+ B cells are more common in other species such as sheep (12) or rabbit, as almost all peripheral B cells express CD5 in the latter (70), suggesting differences in how B cells from these species respond to self- and nonself-antigens.
Having established that a large fraction of trout B cells express CD5, our next step was to compare fish IgM+ B cells to murine splenic B2 cells both phenotypically and functionally. These assays that would allow us to establish to what degree fish B cells resemble mammalian B1 cells were initially performed with the complete population of trout IgM+ B cells because not all B1 populations express CD5 on the cell surface (8). However, in a second step, most of these studies were performed comparing trout IgM+CD5+ and IgM+CD5− B cells. When we phenotypically compared fish IgM+ B cells to murine B2 cells, we verified that fish IgM+ B cells are bigger, are more complex, and express higher levels of surface IgM and lower amounts of surface IgD than murine B2 cells, making them similar to murine B1 cells (12, 62). These differences between murine B2 cells and fish IgM+ B cells were maintained regardless of CD5 surface expression, suggesting that in fish, all splenic IgM+ B cells phenotypically resemble mammalian B1 cells. However, we did find a significant difference regarding IgD surface expression levels between trout IgM+CD5+ and IgM+CD5− B cells, as IgM+CD5− B cells had significantly lower levels of surface IgD than IgM+CD5+ cells. This, together with the fact that IgM+CD5− cells had an extended survival in cell culture when compared with IgM+CD5+ cells, demonstrates that CD5 is not a requisite for IgM+ B cells to have a B1-like profile in fish.
At a functional level, B1 and B2 cells are mainly distinguished by their differential response to BCR engagement (64). Thus, we first determined the proliferative responses of fish B cells to BCR cross-linking, as it is known that B1 cells, unlike B2 cells, do not proliferate when their BCR is cross-linked (84). Fish IgM+ B cells were found not to proliferate in response to BCR cross-linking, pointing to an important functional similarity between fish B cells and mammalian B1 cells. Because it could have been possible that this lack of responsiveness to anti-IgM was a consequence of a poor cross-linking capacity of the Ab, we confirmed that the anti-trout IgM mAb was capable of inducing the mobilization of intracellular calcium in fish B cells. In mammals, there is some controversy regarding the calcium response to BCR cross-linking; although some authors have revealed impaired responses (89), others have reported calcium mobilizations similar to those of B2 cells (90). Despite these differences, it is well accepted that murine B1 cells do not activate NF-κB upon BCR cross-linking (64). Interestingly, fish B cells significantly translocated p65 to the nucleus in response to anti-IgM, revealing a functional difference between fish B cells and murine B1 cells, despite their similarity in many other functional and phenotypical aspects. Whether this functional difference is a consequence of teleost CD5 not having conserved Y429 (86) warrants further investigation. In response to other stimuli such as LPS, mammalian B1 cells activate NF-κB as effectively as B2 cells (91), and our results have also shown the capacity of fish B cells to induce NF-κB in response to LPS (58) or bacteria.
Thus, fish B cells seem to share with mammalian B1 cells their extensive life in cell culture, the inability to proliferate in response to BCR stimulation, a higher size and complexity, and the expression of high surface IgM and low surface IgD levels. These characteristics together with those previously reported, such as a high phagocytic capacity, (67) strongly suggest that fish B cells are more closely related to mammalian B1 than to mammalian B2 cells. Of course, some functional differences with mammalian B1 cells may be possible, such as, for example, the capacity of fish B cells to activate NF-κB in response to BCR cross-linking. Some of these similarities and differences could be explained by the conservation of specific motifs in the fish CD5 molecule, which is responsible for dampening the BCR responses in mammalian B1 cells (84). Thus, the identification of novel B1 markers in teleost fish will not only allow us to characterize B cell subpopulations in fish only designated up to date by the expression of surface Igs but will also help us understand the response of fish B cells to cognate Ags, which we will address in our future studies.
We thank Lucia Gonzalez, Rosario Castro, and Maria Sanz for technical assistance.
This work was supported by European Research Council (ERC) Starting Grant 2011 (280469) and ERC Consolidator Grant 2016 (725061) and by the European Commission under the 7th Framework Programme for Research and Technological Development of the European Union Grant Agreement (311993 TargetFish).
The online version of this article contains supplemental material.
Abbreviations used in this article:
basic local alignment search tool
Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria
National Center for Biotechnology Information
open reading frame
scavenger receptor cysteine-rich
The authors have no financial conflicts of interest.