Previously, four putative novel NK-lysin-like peptides from S. salar were identified based on the EST database. The four transcripts identified were named SsNK-lysin 1 (GenBank accession no.: NM_001141110.1), SsNK-lysin 2 (GenBank accession no.: EG932844.1), SsNK-lysin 3 (GenBank accession no.: EG810337.1) and SsNK-lysin 4 (GenBank accession no.: EG819316.1). The present study searched Gene Databases in GenBank (http://www.ncbi.nlm.nih.gov/) for additional putative NK-lysin sequences. All gene and protein sequences were obtained from the National Center for Biotechnology Information (NCBI). The signal peptide cleavage site was predicted with SignalP 5.0 (http://www.cbs.dtu.dk/services/SignalP/). A multiple sequence alignment was performed using the Clustal Omega tool (https://www.ebi.ac.uk/Tools/msa/clustalo/) from the NK-lysin protein sequences of O. mykiss and S. salar. Besides, another multiple sequence alignment was performed using the ClustalW tool (https://www.genome.jp/tools-bin/clustalw) from NK-lysin protein sequences without signal peptides from mammals, avian, and teleosts. With this last alignment, a phylogenetic tree was constructed with the Molecular Evolutionary Genetic Analysis 11 (MEGA11) program, using the Neighbor-Joining method and the bootstrap test with 1000 replicates. In addition, all ambiguous positions were removed for each pair of sequences (pairwise deletion option), and the Poisson distance correction method was used. The NK-lysin sequences used for phylogenetic analysis and sequence alignment are listed in Table 1 .

SsNK-lysin sequences were modeled using the SWISS-MODEL server (Swiss Bioinformatics Institute, Basel, Switzerland), available at https://swissmodel.expasy.org/interactive (42). A search for templates that could fit the target sequence was performed. From a comprehensive list of more than 100 templates, a heuristic filter was applied to select the 50 most promising models based on coverage and sequence identity criteria. The resulting models were obtained from a variety of sources, including the UniProt database (European Bioinformatics Institute, Cambridge, UK) (43), the Protein Data Bank (PBD) (44), and AlphaFold (DeepMind, London, UK) (44, 45). Subsequently, the models generated for each NK-lysin were subjected to a validation phase, in which tools such as PROCHECK (Laboratory of Molecular Biology, Medical Research Council, Norwich, United Kingdom) (46), VERIFY3D (47), ERRAT (48) and QMEANDisCo (Institute of Bioinformatics, University of Zurich, Zurich, Switzerland) (49) were used. These steps were performed to ensure the integrity and quality of the models obtained.

To predict disulfide bridge formation, the Disulfide by Design 2 server (Boston University School of Medicine, Boston, USA) was used (50). Specific setup parameters included the definition of an angle χ3 with values of -87° or +97°, with a ±30 variation range, and an angle Cα-Cβ-Sγ set at 114.6°, with a tolerance of ±10. The various models generated from the SsNK-lysin sequences were loaded, and the energetic parameters expressed in kcal/mol were evaluated. It is relevant to note that a lower energy reflects a higher probability of disulfide bridge formation. Subsequently, amino acid positions with a higher likelihood of forming disulfide bonds were identified using PyMOL 2.5.5 (DeLano Scientific LLC, San Carlos, USA), and the presence of these bonds with the lowest energy in the analyzed structure was confirmed.

Because the increase in length makes it difficult to obtain NK-lysin by chemical synthesis, short peptides derived from these molecules have been designed and studied. We previously designed and synthesized two small peptides, NK1 and NK2, derived from SsNK-lysin 1 and SsNK-lysin 2, based on alignments between NK-lysin sequences identified in S. salar and NKLP27, a peptide derived from C. semilaevis NK-lysin (17). NK1 and NK2 comprise 27 residues that form the H2 and H3 α-helices of the SapB domain of SsNK-lysin 1 and SsNK-lysin 2, respectively (17). NK3, NK4, NK5, and NK6 derived from SsNK-lysin 3, 4, 5, and 6 were designed based on sequence alignment between NK-lysin sequences previously identified in S. salar. NK3 peptide resulted in 96% identical to NK2, and NK5 and NK6 were identical to NK1. Therefore, we synthesized and characterized three NK-lysin-derived peptides: NK1, NK2, and NK4. The Hydrophobicity, Hydrophobic Moment, total net charge, theoretical isolelectric point, and molecular weights of each peptide were calculated using the Database of Antimicrobial Activity and Structure of Peptides (https://dbaasp.org/home) and Compute pI/Mw from ExPASy (http://web.expasy.org/compute_pi/). The physicochemical properties of the three NK-lysin-derived peptides selected are listed in ( Supplementary Material, Table 1 ).

The peptides derived from SsNK-lysin 1, 2, and 4, NK1 (TLKQKLLSVCDKVGFLKSMCKGLMKKH), NK2 (EIKQKLLSYCGKLPLVKSTCEDLVKKH), and NK4 (EIKQKLLSVCDKMGLLKSLCKGMVKKH) were chemically synthesized by GenScript Company (https://www.genscript.com/). Cysteines 10 and 20 were linked by disulfide bonding in the three peptides. The peptides were purified by high-performance liquid chromatography to 90% of purity. Lyophilized peptides were stored at −20°C and dissolved in DMSO before use.

Unvaccinated Atlantic salmon (Salmo salar) were obtained and maintained in the Marine Biotechnology unit, Faculty of Natural and Oceanography Science, University of Concepcion. The fish occupied in the experiments were certified as free of the most prevalent pathogens. The animals were maintained under a 12: 12 h light: dark cycle and fed daily until satiety with a commercial diet (Micro 200, EWOS).

All the animals used in this study were treated under the Biosecurity Regulations and Ethical Protocols approved by the University of Concepcion Ethics Committee, as required by Chilean Regulatory Entities: National Research and Development Agency (ANID) and National Fisheries and Aquaculture Service (SERNAPESCA).

The minimal inhibitory concentration (MIC) of synthetic NK-lysin-derived peptides (NK1, NK2, and NK4) was measured for P. salmonis (LF89) and F. psychrophilum by a broth microdilution method (51). The MIC is defined as the lowest concentration of an antimicrobial agent at which bacterial growth is not detected. P. salmonis was grown in Tryptic Soy Broth (TSB) (Merck, Darmstadt, Germany) supplemented with NaCl 3 g/L, fetal bovine serum (FBS) 2.5%, L-cysteine 0.05%, and FeCl3 0.01 g/L (52) and F. psychrophilum was grown in a broth containing Tryptone 4 g/L, MgSO₄ 0.5 g/L, CaCl₂ 0.5 g/L, Yeast extracts 0.4 g/L, pH 7.2. Briefly, logarithmic phase microorganism cultures were diluted in the broth according to the microorganism to an optical density at 600 nm (OD₆₀₀) of 0.001, approximately equivalent to 10⁵ cfu mL⁻¹. Diluted microorganisms (90 μL) were mixed with 10 μL of water (negative control) or peptides in wells of polypropylene microtiter plates (Greiner Bio-One, Germany). The peptides were two-fold serially diluted. The growth was monitored by measuring the change in the absorbance of the culture at 600 nm using a microplate reader after 2 and 5 incubation days at 18 °C for F. psychrophilum and P. salmonis, respectively. The MIC was determined by visual verification of microbial sedimentation and absorbance reading (600 nm). In addition, the half maximal inhibitory concentration (IC50) was determined, which is a measure of the efficacy of a compound in inhibiting a biological or biochemical function and indicates how much of a given drug or other substance (inhibitor) is needed to inhibit a given biological process by half. The IC50 for each peptide was calculated from concentration-effect curves after non-linear regression analysis using GraphPad prism10 software. MIC and IC50 were expressed as the absolute value of the mean of at least two determinations in triplicate.

The hemolytic activity of NK-lysin-derived peptides (NK1, NK2, and NK4) was determined using human and fish erythrocytes. Briefly, fresh human or fish erythrocytes (5 mL) were washed with PBS and resuspended in PBS (50 mL) supplemented with glucose (0.2%, w/v). Synthetic peptides (serially diluted in PBS) were added to 90 μL of 1% erythrocyte suspension. Samples were incubated for 30 min at 37°C and centrifuged for 10 min at 3500 rpm at room temperature. The supernatants (70 μL) were transferred to a microtiter plate, and the optical density was determined at 405 nm. The percentage of hemolysis was defined relative to the hemolysis obtained with the erythrocyte suspension treated with 0.1% SDS (100% hemolysis). Two individual experiments were performed using duplicate samples for each peptide.

Five Salmo salar weighting 100-150 g were sacrificed by overexposure to benzocaine (20%), and the gill, muscle, intestine, liver, spleen, stomach, head kidney, heart, and skin were aseptically removed for evaluation of constitutive expression of SsNK-lysin 1, 2 and 4 transcripts. The tissues were kept in RNAlater (Invitrogen) at -80°C until use.

Salmo salar weighting 150-200 g were sacrificed by overexposure to benzocaine (20%), and the head kidneys were aseptically removed to isolate head-kidney leukocytes (HKL) for in vitro assays.

HKLs were isolated from S. salar following the method previously described (53). The head kidney removed aseptically was homogenized through a 40 μm nylon mesh using Leibovitz medium (L-15, Gibco, USA) supplemented with 100 IU/mL penicillin (Gibco, USA), 100 μg/mL streptomycin, 2% heparin and 2% fetal bovine serum (FBS, Hyclone, USA). The resulting cell suspension was placed onto Percoll gradients with a density of 51%/34% and then was centrifuged at 800 g for 40 min at 15°C. The fraction corresponding to the leukocytes was collected, washed twice, and centrifuged at 800 g for 5 min at 15°C in an L-15 medium supplemented with 10% FBS. Viable cell concentration was determined by the Trypan blue exclusion method, and the cells were resuspended in an L-15 medium supplemented with 10% FBS. For each experiment in which HKLs were used, each experimental replicate corresponded to cells from different animals.

96-well plates were seeded at a cell density of 500,000 cells per well for HKLs. After 24 hours, the cells were incubated with 50 μM of NK-lysin-derived peptides (NK1, NK2, and NK4) for 24 and 48 hours. After incubation, the cells were incubated with 1 mg/ml MTT at 20°C for 6 hours. Finally, all MTT was removed, and the formazan salts were resuspended in 100 μl isopropanol. Absorbance was then read at 550 nm. Cells without peptides were used as controls for 100% cell viability.

S. salar HKLs were seeded into 24-well culture plates at a concentration of 10⁶ cells/well in L-15 medium with 10% FBS and cultured at 18°C. The cells were incubated with culture media containing 50 μM of the SsNK-lysin-derived peptides (NK1, NK2, and NK4) or culture media alone as negative control. The cells were harvested 6, 12, and 48 hours after the stimulation. For all treatments, cells were kept in incubation at 18°C. Expression of TNF-α, IL-1β, IL-8, IFN-γ, IL-10, IL-18, Mx, and TGF-β was determined by qRT-PCR.

HKLs (4x10⁶ cells/well) were pretreated for 2 hours with 10 μM of SB202190 inhibitor (p38 inhibitor) (Sigma Aldrich), U0126 inhibitor (mitogen-activated protein kinase 1/2 (MEK1/MEK2) inhibitor) or SP600125 inhibitor (JNK inhibitor) (Abcam). As a negative control, cells were incubated with 0.1% DMSO as a vehicle for 2 hours. Cells were then incubated with 50 μM of NK1, NK2, NK4, or medium (negative control) for 12 hours. For all treatments, cells were kept in incubation at 18°C. Cell lysates were harvested, RNA was extracted, and IL-1β relative expression was analyzed by qRT-PCR.

SHK-1 cells were seeded in L-15 medium (supplemented with 10% FBS) at 3x10⁶ cells/well. After 24 hours, the cells were co-stimulated with 50 μM of SsNK-lysin-derived peptides (NK1, NK2, and NK4) and 1 µg/mL of lipopolysaccharide (LPS from E. coli, 0111:B4) (Sigma Aldrich) or 1 µg/mL of poly(I:C) (Sigma Aldrich) for 6 and 12 hours. Besides, cells were treated only with LPS, poly(I:C), or peptides. Negative control cells with culture media alone were included. For all treatments, cells were kept in incubation at 18°C. Expression of TNF-α, IL-1β, and IL-8 as an LPS-induced response was determined by qRT-PCR. For poly(I:C), the expression of IFN-1α and Mx involved in the antiviral response was determined by qRT-PCR.

Twenty-five S. salar per group of approximately 50 g of body weight will be acclimatized for two weeks at 10-12°C. After this period, animals were intraperitoneally injected with SsNK-lysin-derived peptides (NK1, NK2, and NK4) (20 μg per fish). At 1-, 3-, 7-, 14- and 21 days post-injection, fish (n = 5 per treatment group per time point) were euthanized, and head kidneys were collected. Immediately after tissue extraction, tissues were stored in RNAlater (Thermo Fisher Scientific) at -80°C until use. Expression of IL-1β, IL-8, IFN-γ, Mx, IL-4/13 and IL-22 was determined by qRT-PCR.

RNA extraction from tissue samples, SHK-1 cells, and HKLs was performed using TRIzol reagent (Thermo Fisher Scientific) according to the manufacturer’s instructions. Tissue samples were homogenized in TRIzol reagent before RNA extraction. For all assays, RNA concentration, and purity were assessed using a Sinergy^® HTK Take3 microplate reader (BioTek, Agilent Technologies), and RNA integrity was verified by 1% agarose gel electrophoresis. RNA samples were stored at -80°C until use.

For the analysis of tissue expression of SsNK-lysin transcripts, a pool was prepared from each tissue (5 samples each). The samples were treated with DNAse I # M0303 (New England Bio Labs), using 2 μg of RNA in a final volume of 11 μL, according to the manufacturer’s protocols. Reverse transcription was performed using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific) according to the following protocol: 11 μL of DNAse I treatment reaction, 1 μL of random hexamer primer, 4 μL of 5X reaction buffer, 2 μL of dNTP mix (10 mM), 1 μL of Ribolock RNase inhibitor (20 U/μL), 1 μL of RevertAid RT (200 U/μl) in a final volume of 20 μL. The reactions were incubated in the TProfessional Basic Thermocycler (Biometra) at 25°C for 5 minutes, then at 42°C for 60 minutes, and finally at 70°C for 5 minutes. A negative control without retrotranscriptase (No-RT control) was performed for each tissue.

The Primer-BLAST tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) was used for primer design. Specific oligonucleotides, which did not amplify transcript variants and were aligned with an exon-exon junction (not applicable for predicted sequences), were selected for the desired products. Sequences and information for the primers used in all RT-qPCR assays are shown in Table 2 .

RT-qPCR reactions were performed with cDNA samples to analyze the expression of SsNK-lysin 1, 2, and 4. The KAPA SYBR FAST One-Step qRT-PCR Master Mix (Kapa Biosystems, USA) was used according to the following protocol: 5 μL of qPCR Master Mix (2X), 0.2 μL of forward primer (10 μM), 0.2 μL of reverse primer (10 μM), 1 μL of cDNA and 3.6 μL of water, in 10 μL final volume. The thermal profile for all genes was as follows: 90°C for 3 minutes, 90°C for 10 seconds (40 cycles), and 58°C for 20 seconds. A no reverse transcriptase (No-RT) control was included for each tissue, and a no cDNA control (No Template Control, NTC) was included for each gene. Each sample measurement was repeated three times.

RT-qPCR reactions were performed on RNA samples for inhibitor treatment, PAMPs-induced response, and in vitro and in vivo stimulation with SsNK-lysin-derived peptides. In all cases, the Brilliant II SYBR^® Green qRT-PCR Master Mix, 1-Step kit (Agilent, USA) was used according to the following protocol: 2 μL of RNA (amounts described in Table 2 ), 0.4 μL of RT/RNase block enzyme mixture, 5 μL of 2X Brilliant II SYBR Green qRT-PCR Master Mix, 2.36 μL of water, 0.24 μL of mix forward and reverse primers (2.5 μM each) in 10 μL final volume. The cycling conditions for all genes and assays were as follows: 50°C for 30 minutes (RT reaction), 95°C for 10 minutes, 95°C for 15 seconds, and 58°C for 30 seconds (40 cycles). No RNA controls (NTC) were included for all genes and assays, and each sample measurement was repeated three times.

The EF-1α reference gene was used as a normalizer for all RT-qPCR assays. For the tissue SsNK-lysin expression assay, stomach tissue was used as a control or calibrator. For the other assays, untreated cells or tissues from untreated animals were used as calibrators. The AriaMx real-time PCR system (Agilent, USA) was used for all RT-qPCR reactions. In addition, to standardize and validate the primers used and the reactions, dynamic ranges, efficiency calculation, melting curve evaluation, and visualization of amplicons in 1% agarose gel electrophoresis were performed. The results were analyzed using the comparative Ct (2^-ΔΔCt) method (54), and GraphPad Prism 10 software was used for graph generation and statistical analysis. Specifically for constitutive tissue expression, relative expression was determined according to Paff’s mathematical model (2001) (55).

Phagocytic activity was analyzed by a microplate fluorometric assay using pHrodoTM Green E. coli Bioparticles™ conjugate (Molecular Probes/Thermo Fisher Scientific). Particles were suspended in Live Cell Imaging Solution buffer (Molecular Probes/Thermo Fisher Scientific) supplemented with 0.2% (w/v) glucose (LCIS-glu) at a density of 1 mg/mL, vortexed for 2 minutes and sonicated for 5 minutes at room temperature. HKLs were seeded at 1x10⁶ cells/well density in 96-well plates in L-15 medium supplemented with 2% FBS and incubated at 18°C for 24 hours. Cells were washed with HBSS, and the phagocyte-enriched population was incubated for 30 minutes at 18°C with 50 μL of LCIS-Glu and then with 50 μL of bioparticles. SsNK-lysin-derived peptides (NK1, NK2, and NK4) were evaluated at 0, 10, and 50 μM. Control wells without cells containing bioparticles were included in each plate to allow subtraction of background fluorescence from the particles at neutral pH. Cells plus peptide and bioparticles were incubated for 3 hours at 18°C. Fluorescence was recorded on the Synergy HTX microplate reader (BioTek Instruments, USA) in well area scanning mode, with excitation and emission wavelengths of 488 and 528 nm, respectively. Data were calculated as percentage phagocytosis relative to control cell samples incubated with bioparticles without stimuli after subtracting background particle fluorescence (cell-free control). Each assay was performed in triplicate and repeated with at least two independent samples.

All data were assessed for homogeneity of variances and normal distribution before data were analyzed by one-way ANOVA. The Dunnett post hoc test compared means from experimental groups against a control group mean. Šidák’s multiple comparisons test was used to perform simultaneous joint pairwise comparisons for all possible pairwise combinations of means. Statistical analysis was done using GraphPad PRISM version 10.00 (GraphPad Software, San Diego, CA). p<0.05 was considered a significant difference.

Based on the Gene database, this study identified two additional putative novel NK-lysin-like peptides from Salmo salar. The two transcripts identified were named SsNK-lysin 5 (GenBank accession nos.: XM_014125754.1) and Ss-NK-lysin 6 (GenBank accession nos.: XM_014129907.1). Besides, we identified the genes for the six SsNK-lysin-like peptides ( Table 3 ). SsNK-lysin 1, 2, 3, 5, and 6 are located on chromosome 1, whereas the gene for SsNK-lysin 4 is located on chromosome 9. Table 3 also shows the orientation of these genes on chromosome arms 1q and 9q of the Salmo salar genome.

The SsNK-lysin 5 open reading frames (ORF) consisted of 402 bp and encoded a protein of 133 amino acid residues. It contains an N-terminal signal peptide (1–22 aa) and a SapB domain (49–123 aa). The theoretical molecular mass of the mature SsNK-lysin 5 is 12629.84 Da, and the isoelectric point (pI) is 8.18 ( Table 3 ). The SsNK-lysin 6 ORF consisted of 402 bp and encoded a protein of 133 amino acid residues. It contains an N-terminal signal peptide (1–22 aa) and a SapB domain (49–123 aa). The theoretical molecular mass of the mature SsNK-lysin 6 is 12615.85 Da, and the pI is 8.14 ( Table 3 ).

The SsNK-lysin 5 encoded protein sequence shared 84.3%, 74.2%, 75%, and 77.4% identity with the Ss-NK-lysin 1, 2, 3, and 4 (reported in 2019), respectively (17). The SsNK-lysin 6 encoded protein sequence shared 86.6%, 71.9%, 72.7%, and 76.7% identity with the Ss-NK-lysin 1, 2, 3 and 4, respectively (17). Besides, the SsNK-lysin 5 and SsNK-lysin 6 encoded protein sequences shared 90.98% identity. On the other hand, the percent of the identity of SsNK-lysin peptides with NK-lysin (NKL) peptides from O. mykiss (determined without signal peptide) ranged from 71% to 85% (22). However, the percent of the identity of SsNK-lysin peptides with Nkl-like a and four variants of Nkl-like b from O. mykiss ranged from 27% to 31% (22) ( Figure 1 ).

The inferred phylogenetic tree showed that the mammalian and avian NK-lysin sequences cluster within the clade furthest from the rest of the phylogenetic tree and are related to the NK-lysin sequences of the L. crocea species and the evaluated sequences of the Oreochromis genus. On the other hand, all S. salar sequences and most NK-lysin sequences of other salmonids are grouped in the same clade and subdivided into sub-branches. One sub-branch groups the SsNK-lysin 4 sequence with those of other salmonids, suggesting that this sequence may be less related to the other S. salar sequences. On the other hand, SsNK-lysin 2 and SsNK-lysin 3 cluster in another sub-branch closer to OmNK2, while SsNK-lysin 1, SsNK-lysin 5 and SsNK-lysin 6 cluster together with OmNK3 in another sub-branch of this clade. The rest of the salmonid NK-lysin sequences, including the NK-lysin-like sequences from O. mykiss and its variants, cluster in a distant clade, closer to sequences from species such as C. carpio, L. crocea, O. niloticus, O. aureus, and avian and mammalian sequences ( Figure 2 ).

Multiple alignments revealed six cysteine residues in the two SsNK-lysin mature peptides reported here that are highly conserved among fish species. Nevertheless, all SsNK-lysins contain at least an additional cysteine residue conserved among them ( Figure 1 ). SsNK-lysin 1 and SsNK-lysin 6 have eight cysteine residues in the mature peptide conserved among them, and SsNK-lysin 5 possesses eight cysteine residues; one of them is not conserved among SsNK-lysin sequences.

To predict the structures and the formation of disulfide bridges, we used the SWISS-MODEL server (Swiss Bioinformatics Institute, Basel, Switzerland) (42) and the Disulfide by Design 2 server (Boston University School of Medicine, Boston, USA) (50), respectively. As a result, we found that three disulfide bridges are predicted to be generated for all sequences, with one or two cysteine residues not forming bridges. The positions of the bridging cysteines are shown in Table 4 . Figure 3 shows the structures corresponding to the SsNK-lysin. Visualization of the protein revealed a predominance of α-helices, which are extensively distributed along the molecule. No folded β-sheets were detected in the current protein conformation. Additionally, segments known as loops or turns were identified. These segments could act as flexible connections facilitating continuity between ordered structures. Loops and unstructured areas indicate sites of flexibility and movement capability within the protein.

The antimicrobial activity of peptides derived from SsNK-lysin against P. salmonis and F. psychrophilum was determined. Bacteria were chosen to represent important fish pathogens. Results show that NK1 and NK4 are active against P. salmonis. In contrast, the three peptides were active against F. psychrophilum. SsNK-lysin-derived peptides over a range of concentrations up to 200 μM displayed variable degrees of antimicrobial activity against the microorganisms tested ( Table 5 ). NK1 and NK4 were the most active peptides against these critical fish pathogens. In contrast, NK2 exhibited a more limited spectrum of activity against these pathogens.

Synthetic SsNK-lysin-derived peptides were assays for hemolytic activity in human and fish red blood cells. NK2 and NK4 were not hemolytic for fish red blood cells at concentrations below 100 μM. In contrast, NK1 was hemolytic in fish erythrocytes, reaching 50% of hemolysis at approximately 30 μM ( Figure 4A ). In addition, hemolytic activity in human red blood cells was observed for NK1 in a dose-dependent manner, reaching 50% of hemolysis at 25 μM approximately. On the other hand, NK4 reached 33% of hemolysis at the highest concentration analyzed ( Figure 4B ). In contrast, NK2 was not hemolytic for human erythrocytes at concentrations below 100 μM ( Figure 4B ).

We analyzed the constitutive expression of the three Salmo salar NK-lysin transcripts, SsNK-lysin 1, SsNK-lysin 2, and SsNK-lysin 4. The high sequence homology of the salmon NK-lysin transcripts does not allow us to design specific oligonucleotides for NK-lysin variants 3, 5, and 6.

When the constitutive expression of SsNK-lysin 1, 2, and 4 were analyzed in different S. salar tissues, different expression levels were observed for each transcript analyzed ( Figure 5 ). SsNK-lysin 1 was highly expressed in skin, head kidney, spleen, and gills, showing its highest expression level in gills. Besides, SsNK-lysin 1 was also expressed in muscle, intestine, liver, and heart; SsNK-lysin 2 was expressed in all tissues analyzed but showed a lower expression level than the other two transcripts analyzed. On the other hand, SsNK-lysin 4 showed its highest expression level in the gills, spleen, and head kidney. In addition, SsNK-lysin 4 was also expressed in muscle, skin, liver, heart, and intestine. The intestine showed the lowest expression level for all transcripts analyzed.

The cytotoxicity of the NK-lysin-derived peptides on S. salar HKLs was assessed by a standard MTT assay conventionally used to measure cell viability. As shown in Figure 6 , none of the peptides were significantly toxic to S. salar HKLs at 50 μM, at 24 and 48 hours of treatment, with cell viability greater than 87% at this concentration.

The ability of SsNK-lysin-derived peptides (NK1, NK2, and NK4) to induce cytokine expression in the S. salar HKL was evaluated by RT-qPCR. The mRNA expression of IL-10, TGF-β, TNF-α, IL-8, IL-1β, Mx-1, IFN-γ and IL-18 was evaluated after 6, 12, and 48 h post-treatment with synthetic peptides. NK1 at 50 μM induced a significant increase in the expression of pro-inflammatory cytokines IL-1β (6- and 12-hours; 1.2- and 18.3-fold change, respectively), IL-8 (12- and 48-hours; 35- and 2.4-fold change, respectively), IL-18 (6- and 48-hours; 1.4- and 2.3-fold change, respectively) and TNF-α (12 hours; 5.4 folds respect to control). Besides, NK1 at 50 μM induced a significant increase in the expression of T_H1 cytokine IFN-γ at 48 hours post-treatment, reaching 9.6-fold concerning control. NK1 also induced a significant increase in the expression of IFN-γ inducible cytosolic protein Mx-1 at 6 hours (3.4-fold change). Respect to anti-inflammatory cytokines, NK1 at 12 hours post-treatment induced a significant increase in the expression of IL-10 (4.5-fold change) and induced a significant increase in the expression of TGF-β at 12- and 48-hours post-treatment (1.2- and 2.2- fold change, respectively) ( Figure 7 ).

On the other hand, NK2 at 50 μM induced a significant increase in the expression of pro-inflammatory cytokines IL-8 (6- and 48-hours; 2.3- and 1.6-fold change, respectively) and IL-18 (6-, 12- and 48-hours; 1.4-, 1.6- and 1.8-fold change, respectively) and TNF-α (6 hours; 2.5-folds respect to control) ( Figure 7 ). Besides, NK2 at 50 μM induced a significant increase in the expression of IFN-γ at 12- and 48-hours post-treatment reaching 3- and 3.7-folds respect to control, respectively. NK2 also induced a significant increase in the expression of Mx-1 at 12 hours (1.9-fold change). Concerning anti-inflammatory cytokines, NK2 at 12- and 48-hours post-treatment caused a significant increase in the expression of IL-10 (3.7- and 5.1-fold change, respectively) and TGF-β (1.7- and 2.9-fold change, respectively) ( Figure 7 ).

NK4 at a dose of 50 μM induced the expression of all cytokines analyzed. At 6 hours post-treatment, NK4 induced the expression of IL-8 and IL-10, reaching 1.9- and 1.3-fold concerning control, respectively. The higher expression levels were observed for IFN-γ at 12- and 48-hours, IL-1β at 12 hours, IL-8 at 12 hours, IL-10 at 12 hours, and TNF-α at 12 hours (4-, 9.5-, 4.5-, 7.1-, 5.6- and 9.8- fold change, respectively). Besides, NK4 induced a significant increase in the expression of IL-1β, IL-8, IL-10 and TGF-β at 48 hours post-treatment (1.8-, 2.7-, 4.6- and 3.7- fold change, respectively) ( Figure 7 ).

To investigate the relationship between the activation of the MAPK pathways and IL-1β induction, specific inhibitors of the MAPK pathways were used. The expression of IL-1β was significantly abrogated when HKLs were treated with NK4 in the presence of inhibitors, including SB202190, U0126, and SP600125 (p38, ERK1/2, and JNK inhibitors, respectively) ( Figure 8 ). In contrast, leukocytes treated with NK1 only in the presence of SB202190 inhibitor significantly abolished the expression of IL-1β ( Figure 8 ). On the other hand, the expression of IL-1β was significantly abrogated when HKLs were treated with NK2 in the presence SB202190 and SP600125 inhibitors ( Figure 8 ). The results suggest that activation of the MAPK pathways is not only induced by NK-lysin-derived but is also necessary for the peptides-induced immune response.

We evaluated the immunomodulatory properties of the SsNK-lysin-derived peptides through their modulation of LPS and poly(I:C)-induced response in the SHK-1 cell line. SHK-1 cells were stimulated with LPS or poly(I:C) in the presence or absence of NK1, NK2, or NK4. The effect of SsNK-lysin-derived peptides on genes related to inflammatory and antiviral responses is shown in Figure 9 . We found that the LPS challenge significantly elevated the expression levels of inflammatory cytokines such as the IL-1β, TNF-α, and IL-8 in the SHK-1 cells at 6 and 12 hrs post-treatment. However, co-treatment of LPS and NK-lysin-derived peptides significantly decreased the IL-1β and TNF-α expression levels at 12 hrs post-treatment. At 12 hrs, only the co-treatment of LPS and NK2 showed a significant decrease in the IL-8 expression level. NK1 and NK4 showed no significant change in the IL-8 expression level in the presence of LPS. At 6 hrs, the co-treatment of LPS and NK-lysin-derived peptides increased the IL-1β and IL-8 expression levels compared to the group treated with LPS alone. The co-treatment of NK1 or NK2 with LPS for 6 hrs didn’t show significant differences in the TNF-α expression level. However, the co-treatment of NK4 and LPS for 6 hrs significantly reduced the TNF-α expression level compared with LPS alone.

Besides, we determine if co-treatment with NK-lysin-derived peptides modulated the antiviral immune response to poly(I:C), a synthetic ligand of TLR3, and viral mimic in SHK-1 cells. The expression levels of the IFN-α1 and Mx genes were analyzed at 6- and 12 hours post-treatment. Poly(I:C) upregulated the expression of IFN-α1 and Mx only at 6 hrs post-treatment ( Figure 9 ). Co-incubation with NK-lysin-derived peptides for 12 hrs significantly increased the transcription of IFN-α1 and Mx compared to poly(I:C) treated cells ( Figure 9 ). At 6 hrs, the co-treatment of NK2 and poly(I:C) significantly reduced the Mx expression level. NK1 and NK4 didn’t affect the Mx expression level in the presence of poly(I:C). In addition, NK1 and NK2 reduced the IFN-α1 transcription, while NK4 significantly increased this level in the presence of poly(I:C).

The phagocytic activity of the head kidney cells treated with Salmo salar NK-lysin-derived peptides was investigated ( Figure 10 ). The phagocytosis rate of the head kidney cells treated with NK1 and NK4 at 50 μM was significantly upregulated at 3 and 6 hrs post-stimulation compared to that of the control ( Figure 10 ). The NK1 showed the best results in the stimulation of phagocytic activity.

In the in vivo experiment, the immunomodulatory effects of synthetic NK-lysin-derived peptides were evaluated. To this end, Salmo salar (50 g) were ip injected with NK-lysins-derived peptides (20 μg per fish). The control group was only injected with 100 μL PBS. At 1-, 3-, 7-, 14- and 21 days post-injection, fish (n = 5 per treatment group per time point) were euthanized, and head kidneys were collected for analysis of immune responses by RT-qPCR. The immune-related genes (IL-8, Mx, IL-4, IL-22, and IFN-γ) were upregulated in the head kidney of the NK1-injected group when compared with the control one ( Figure 11 ). IL-8 expression was increased at day 1, 14 and 21 (2.7-, 1.9- and 1.6-fold change, respectively). Mx expression was increased at day 3 (2.2-fold change). IL-4 expression was increased at days 3 and 14 (3.4- and 2.4-fold change, respectively). IL-22 expression was increased at days 3 and 14 (3.8- and 2.6-fold change, respectively). IFN-γ expression was increased at day 14 (2.2-fold change) ( Figure 11 ). When the fish were injected with NK2, the immune-related genes (IL-1β, IL-8, IL-4, IL-22, and IFN-γ) were upregulated in the head kidney ( Figure 11 ). IL-1β expression was increased only at day 1 (1.4-fold change). IL-8 expression was increased at days 7 and 21 (2.6- and 1.6-fold change, respectively). IL-4 expression was increased at days 3 and 7 (1.5- and 1.8-fold change, respectively). IL-22 expression was increased at day 14 (2.4-fold change). IFN-γ expression was increased at day 7 (2.6-fold change) ( Figure 11 ). The immune-related genes (IL-1β, IL-8, Mx, IL-4, IL-22, and IFN-γ) were upregulated in the head kidney of NK4-injected group when compared with the control one ( Figure 11 ). IL-1β expression was increased at days 14 and 21 (2.2- and 1.9-fold change, respectively). IL-8 expression was increased on days 1, 14, and 21 (1.4-, 2.4- and 1.6-fold change, respectively). Mx expression was increased at day 21 (1.9-fold change). IL-4 expression was increased at days 1 and 21 (2- and 3.4-fold change, respectively). IL-22 expression was increased at day 21 (2.2-fold change). IFN-γ expression was increased at days 14 and 21 (1.5- and 1.3-fold change, respectively) ( Figure 11 ).