β-Glucan-Induced IL-10 Secretion by Monocytes Triggers Porcine NK Cell Cytotoxicity

Beta-glucans are naturally occurring polysaccharides present in cell walls of fungi, yeast, bacteria, cereals, seaweed, and algae. These microbe-associated molecular patterns (MAMPs) possess immunomodulatory properties. In human, it has been suggested that NK cells can be activated by β-glucans. Here, we aimed to elucidate whether β-glucans modulate porcine NK cell responses in vitro and if so, how these effects are mediated. We investigated the effect of two β-glucans, Macrogard and Curdlan, which differ in solubility and structure. Direct addition of β-glucans to purified porcine NK cells did not affect cytotoxicity of these cells against K562 target cells. However, when using PBMC instead of purified NK cells, β-glucan addition significantly increased NK cell-mediated cytotoxicity. This effect depended on factors secreted by CD14+ monocytes upon β-glucan priming. Further analysis showed that monocytes secrete TNF-α, IL-6, and IL-10 upon β-glucan addition. Of these, IL-10 turned out to play a critical role in β-glucan-triggered NK cell cytotoxicity, since depletion of IL-10 completely abrogated the β-glucan-induced increase in cytotoxicity. Furthermore, addition of recombinant IL-10 to purified NK cells was sufficient to enhance cytotoxicity. In conclusion, we show that β-glucans trigger IL-10 secretion by porcine monocytes, which in turn leads to increased NK cell cytotoxicity, and thereby identify IL-10 as a potent stimulus of porcine NK cell cytotoxicity.


INTRODUCTION
Beta-glucans are naturally occurring glucose polymers that are important structural components of the cell wall of e.g., fungi, yeast, bacteria, cereals, seaweed, and algae (1). Thus, they originate from different sources and therefore are a heterogenous group, sharing a linear β-1,3-linked backbone structure, sometimes complemented with β-1,6-linked side chains. Other variable characteristics include branching frequency, length of the side branches, solubility, tertiary structure, and molecular weight. It is known that these variable features of β-glucans impact their immunomodulatory capacities (2). Recognition of β-glucans by the mammalian immune system occurs through pathogen-recognition receptors (PRR), such as dectin-1 and complement receptor 3 (CR3). Recognition of β-glucans and downstream signaling pathways were shown to be cell-type specific in humans (3) and mice (4), as well as in pigs (5). For instance, β-glucan-mediated effects in neutrophils occur mainly through recognition via the CR3 subunit CD11R3, whereas these effects in macrophages are mainly mediated by dectin-1 (3,5,6).
β-glucan-mediated effects have been studied in a wide variety of species (7). In mammals, they have been shown to possess important immunomodulatory capacities. For example, β-glucans enhance the phagocytotic capacity of macrophages (8) and oxidative burst of monocytes, macrophages and neutrophils (3,(9)(10)(11). They induce proliferation of peripheral blood mononuclear cells (PBMC) in vitro (11) and stimulate macrophages and dendritic cells to produce cytokines and chemokines, including TNF-α (11,12), IL-12 (12), CXCL2, IL-6 (13), IL-1β, and IL-10 (11). β-glucans thus have the ability to trigger innate immune responses and shape innate and potentially also adaptive immunity (14). Therefore, β-glucans are also considered as interesting alternatives to prophylactic in-feed antibiotics in animal husbandry. Indeed, growth and immunity promoting antibiotics are banned by an increasing number of countries, including all European Union member states, and βglucan-supplemented feed has been suggested to improve growth and overall health of commercially farmed pigs (15,16).
Natural Killer (NK) cells are innate immune cells particularly known for their ability to kill virus-infected and malignant cells as well as to produce large amounts of IFN-γ upon stimulation (17). Activation of NK cells depends on the critical balance between inhibitory and activating signals transmitted by ligands on potential target cells (18). For example, MHC class I on the cell surface delivers a potent inhibitory signal to NK cells, and MHC class I downregulation on malignant cells or virus-infected cells therefore contributes to NK cell activation and killing of these target cells (19). This mechanism is also exploited to investigate NK cell-mediated cytotoxicity in in vitro assays, in which cell lines expressing limited amounts of MHC class I, such as K562, are co-cultured with effector cells.
It has been claimed that β-glucans can activate human NK cells, but conflicting results have been described in literature. One report described a directly enhanced NK cell-mediated cytotoxicity toward different target cells after preincubation of human NK cells with laminarin (20). Another study showed only an indirect effect on human NK cell cytotoxicity toward K562 cells, an NK cell-susceptible cancer cell line, after stimulation of PBMC with lentinan (21). Furthermore, inhibitory effects of β-glucans on NK cell activity have also been reported. For example, a dose-dependent inhibition of human NK cell killing of different target cells by zymosan and a particulate yeast glucan was described (22). In pigs, an in vivo study suggested that oral delivery of yeast-derived β-glucans impacted on the number of proliferating NK cells (23). This might indicate that β-glucans also influence porcine NK cells. Therefore, we set out to investigate whether two β-1,3-glucan preparations, differing in origin and structure, Macrogard (yeast-derived, β-1,6-branched glucan) and Curdlan (bacterial-derived, unbranched glucan), modulate porcine NK cell responses in vitro. Since administration of β-glucans in vivo is generally not accompanied by side effects (14), further insights into NK cell modulation by β-glucans may not only yield important information with regard to the use of βglucans in animal husbandry, but also pave the way toward new strategies in vaccine development, anti-viral therapies and possibly cancer therapies.

β-Glucans
Macrogard was kindly provided by Biotec Pharmacon ASA (Tromsø, Norway) and Curdlan was purchased from Sigma (Bornem, Belgium). All material coming into contact with the β-glucans was made pyrogen-free. To this end, clean glassware and material was heated at 250 • C for 4 h. β-glucan solutions were made in LAL reagent water (Catalog number W50-100; Lonza, Basel, Switzerland). Both types of β-glucan were solubilized via sonication. Quantification of endotoxin concentrations present in the β glucan preparations, was done by the Chromogenic Limulus Amebocyte Lysate (LAL) Test (Catalog number 50-647U, Lonza) and were 0.0362 EU/µg MG and 4.22 EU/µg CL.

Isolation of PBMC
Heparinized blood samples (50 units/ml blood, LEO Pharma, Lier, Belgium) were obtained from the external jugular vein of pigs (8-29 weeks old) that were kept as blood donors at the Faculty of Veterinary Medicine, Merelbeke, Belgium. The blood sampling procedure was approved by the Ethical Committee of the Faculty of Veterinary Medicine (EC2017/121). PBMCs were isolated from whole blood by lymphoprep (Axis-Shield, Dundee, UK) density centrifugation. Red blood cells were lysed by a 10 min osmotic shock at room temperature by incubating the collected buffy coat in lysis buffer [composed of 90% NH4Cl 0.83% (w/v) and 10% TRIS 2.06% (w/v) in distilled water at a final pH of 7.2]. After a final washing step, PBMC were resuspended in 1 ml PBS-EDTA buffer and counted.

RT-PCR
Total RNA was isolated was isolated from freshly isolated PBMC and NK cells using the RNeasy mini kit (catalog number 74104; Qiagen, Hilden, Germany) according to the manufacturer's procedure. Concentration and purity of the RNA was measured by microvolume UV-Vis spectrophotometry (DeNovix, Wilmington, DE, USA). Afterwards, RNA was subjected to DNase treatment (RQ1 RNase-Free DNase; catalog number M6101; Promega, Madison, WI, USA) according to the manufacturer's protocol. Subsequently, RNA was reverse transcribed into cDNA using the SuperScript III Reverse Transcriptase kit (catalog number 18080093; Invitrogen) in the presence of a recombinant ribonuclease inhibitor (RNase OUT; catalog number 10777019; Invitrogen) according to the manufacturer's instructions to obtain cDNA in a single-step reaction consisting of 5 min at 25 • C (priming), 50 min at 50 • C (reverse transcription) and 15 min at 70 • C (RT inhibition). The resulting cDNA served as a template for the polymerase chain reaction (PCR) assay. Primer oligonucleotides were synthesized by Integrated DNA Technologies (IDT, Coralville, IA). The sequences of the forward and reverse primers used for the amplification of the genes of interest and final primer concentrations can be found in Table 1. For non-preexisting primer oligonucleotide sequences, the primer design tool Primer-BLAST (NIH, USA) was used. The PCR reaction mix consisted of the Ready Mix (ReadyMix Taq PCR Reaction Mix with MgCl2, Catalog number P4600, Sigma-Aldrich, Saint Louis, MO, USA), sense and antisense primer and 100 ng of cDNA. The cycle conditions were 1 cycle of 10 min at 95 • C followed by 40 cycles of 15 s at 95 • C, 30 s at 60 • C and 30 s at 70 • C and finally 1 cycle of 10 min at 70 • C. PCR products were visualized by GelRed (catalog number 41003, Biotium, Landing Parkway Fremont, CA, USA) agarose gel electrophoresis.

CR3 Cell Surface Expression
Freshly isolated porcine NK cells were washed with Dulbecco's Modified Eagle Medium (DMEM) (Thermo Fisher Scientific) + 1% fetal calf serum (FCS) (Thermo Fisher Scientific) and labeled with a primary mouse mAb for 20 minutes at 4 • C. The following primary mAbs were used to identify the presence of CR3 on porcine NK cells: mouse anti-pig CD11R1 (clone MIL-4; Bio-Rad), mouse anti-pig CD11R3 (clone 2F4/11; Bio-Rad), and mouse anti-pig CD18a (clone PNK-I; Bio-Rad). Cells stained with isotype-matched irrelevant mAbs were used as a negative control. After incubation, cells were washed and stained with goat anti-mouse IgG1-PE. The cells were washed and Sytox blue (Catalog number S34857, Thermo Fisher Scientific) was added to discriminate live and dead cells. Doublets were excluded based on FSC-H/FSC-A and SSC-H/SSC-A and a minimal event count of 15,000 NK cells was used. Flow cytometry was performed using a Beckman Coulter Cytoflex and samples were analyzed using CytExpert software (Beckman Coulter).

Direct NK Cell Cytolytic Assays
Cytolytic assays were performed as described before (31)  condition (Medium). Reference genes were selected based on geNorm analysis using qBase+ software.

Depletion of Cytokines via Immunoprecipitation
Freshly isolated PBMC (1 × 10 6 ) were cultured in sterile 96well-flat bottomed plates (Nunc, Thermo Fisher Scientific) in porcine NK medium with or without β-glucan supplementation (Macrogard or Curdlan at 10 µg/ml) for 16 h. Afterwards, SN was collected and frozen at−80 • C and used for immunoprecipitation of TNF-α, IL-6, or IL-10 afterwards. Briefly, 400 µl of SN was incubated in the presence of an antibody specific for TNF-α, IL-10, or IL-6 (listed in Table 2) or without antibody as negative control for 4 h at 4 • C on a rotating wheel. Meanwhile, protein A/G magnetic beads (Catalog number 88802, Thermo Fisher) were washed according to the manufacturer's recommended protocol. Thereafter, the incubated SN was added to a 1.5 mL microcentrifuge tube containing pre-washed magnetic beads and incubated for 4 h at 4 • C on a rotating wheel. Finally, beads were removed with a magnetic stand and the SN was collected and stored at −80 • C. SN was analyzed for immunoprecipitation efficiency with commercially available porcine-specific ELISA kit (see section ELISA) and used for CA (see section Indirect NK cell cytolytic assays).

Statistical Analysis
Statistical analysis was performed using Graphpad Prism 6. Data were analyzed for statistical differences with a Friedman test (in the case of one independent variable) and two-way ANOVA (in the case of two independent variables) at the 5% significance level. Post hoc analysis was performed using Dunn's test and Sidak's test, respectively. The synergistic effect of LPS on IL-10 secretion by monocytes upon Macrogard stimulation was tested for statistical significance using a one-tailed paired t-test.

β-Glucans Are Unable to Directly Stimulate Porcine NK Cell Cytotoxicity
First, the ability of the β-glucans Macrogard and Curdlan to directly stimulate NK cell cytotoxic activity was investigated. Results showed that Macrogard nor Curdlan at 10 µg/ml were able to modulate the cytotoxic capacity of IL-2 primed NK cells against NK-susceptible K562 cells ( Figure 1A). In addition, further increasing the concentration of these β-glucans (20, 50, or 200 µg/ml) also did not affect the cytolytic activity of IL-2 primed NK cells (Supplementary Figure 1).
To understand the reason for the lack of response of NK cells toward β-glucans, the expression of the two major β-glucan receptors, dectin-1, and Complement Receptor 3 (CR3), on porcine NK cells was investigated. Unfortunately, neither porcine-specific nor human cross-reactive anti-dectin-1 antibodies are available, hampering flow cytometric analysis of dectin-1 expression on porcine cells (32). In addition, different dectin-1 isoforms have been described in pigs (33), further complicating antibody-mediated detection of dectin-1, as such an approach may possibly miss certain isoforms. Therefore, we opted for mRNA detection via RT-PCR and primers that recognize these isoforms were designed. As shown in Figure 1B, dectin-1 mRNA was not detectable in freshly isolated porcine NK cells, while PBMCs express different isoforms of this βglucan receptor. Expression of CR3 on porcine NK cells was investigated by flow cytometry. In pigs, this β 2 -integrin is composed of an alpha-subunit (CD11R1 or CD11R3) and a beta-subunit (CD18a). A functional study in pigs showed that recognition of β-glucans by CR3 solely depends on CD11R3 and is independent of CD11R1. As displayed in Figure 1C, this analysis revealed that most NK cells express CD11R1 (97.82% ± 1.18) in combination with the β-subunit CD18a (99.87% ± 0.11). The β-glucan recognizing subunit CD11R3 (0.31% ± 0.11), on the other hand, was only present on a very minor subset of NK cells. We therefore conclude that porcine NK cells do not or do only minimally express either of the two major β-glucan receptors, in line with the observed lack of a direct effect of β-glucan addition on NK cell cytotoxicity.

β-Glucans Indirectly Stimulate NK Cell Cytotoxicity via Monocytes
It is well-established that addition of β-glucans to PBMC may result in a variety of responses, including secretion of cytokines, which in turn may lead to modulation of other immune cells. To investigate whether such downstream effects may affect NK cell cytotoxicity, PBMC were primed overnight with the β-glucans Macrogard and Curdlan and the cytotoxicity of the PBMC toward K562 cells was analyzed. Cytolytic assays showed a statistically significant increased NK-mediated lysis of target cells by PBMC primed with β-glucans (Figure 2A). This increase in cytotoxic activity against K562 cells was abrogated when NK cells were removed from the PBMC population via FACS (Supplementary Figure 2), confirming that cytotoxicity was indeed exerted by NK cells. Although NK cells are capable of producing large amounts of IFN-γ upon activation, intracellular cytokine staining analyzed by flowcytometry showed that no detectable levels of IFN-γ were produced by NK cells in β-glucantreated PBMC (Supplementary Figure 3).
To further investigate whether the activating effect of βglucan-treated PBMC on NK cell cytotoxicity was mediated by secreted molecules or depended on cell-cell contacts, the supernatant (SN) of β-glucan primed PBMC was collected and added to freshly isolated and purified NK cells. After overnight incubation with the SN, NK cells showed an enhanced cytolytic activity toward K562 cells (Figure 2B), indicating that the NK cell-activating factors present in β-glucan-primed PBMC consist of secreted molecules, likely cytokines.
To identify the cell type responsible for the production of these secreted molecules, PBMC were MACS-sorted into CD14 + and CD14 − cell fractions and stimulated overnight with either Macrogard or Curdlan. In porcine PBMC, CD14 is solely present on monocytes and a small subset of B cells, and absent from e.g., dendritic cells (34)(35)(36). After β-glucan priming of the different populations, SN was collected and incubated overnight with freshly isolated NK cells to analyze their potential to affect NK cell cytotoxic activity. The results showed that the β-glucan primed CD14 + population secreted factors that enhance the cytolytic activity of porcine NK cells ( Figure 2C). Depletion of CD14 + cells from the PBMC fraction (=CD14 − ) completely abrogated the indirect stimulatory effect of β-glucans on porcine NK cell cytotoxicity. These results point to monocytes as being both essential and sufficient to trigger NK cell-activating factors upon β-glucan priming of PBMC. To assure that the observed effects using CD14 + populations were not in part mediated by the small population of CD14 + B cells, these assays were repeated with CD172a + and CD172a − cell populations. CD172a is expressed by monocytes and dendritic cells, but not by B cells (34,37). Upon stimulation of NK cells with SN of β-glucan-primed CD172a − cells, no stimulation of NK cell cytotoxicity was observed. In contrast, priming of NK cells with SN of β-glucan-primed CD172a + cells increased the NK cell cytotoxicity to similar levels as seen with SN from the CD14 + population ( Figure 2C). Therefore, we conclude that monocytes play a cardinal role in the production of secreted molecules that increase NK cell cytotoxicity upon β-glucan priming of PBMC.
were investigated in β-glucan-stimulated PBMC (4 h) by RT-qPCR. Results showed that TNF-α, IL-6, and IL-12p40 were upregulated upon β-glucan stimulation (Figure 3A). Although expression levels of IL-10 were not substantially upregulated, previous research has shown that IL-10 at the protein level is significantly upregulated upon β-glucan stimulation of PBMC (11). Therefore, we decided to include IL-10 in our further analysis at the protein level by ELISA. An upregulation at the protein level of TNF-α, IL-6, and IL-10 was confirmed, whereas no increase of IL-12p40 protein was detected upon β-glucan stimulation of PBMC or CD14+ cells (Figures 3B-E).
To analyze the potential contribution of each of the upregulated cytokine proteins (TNF-α, IL-6, and IL-10) in the increased NK cell mediated cytotoxicity observed upon βglucan stimulation of PBMC, NK cell cytotoxicity assays were performed. First, the effects of recombinant porcine TNF-α, IL-10, or IL-6 on cytotoxicity of freshly isolated pNK cells were investigated. Both recombinant TNF-α ( Figure 4A) and IL-10 ( Figure 4D) were able to directly stimulate NK cell cytotoxicity in a concentration-dependent manner, whereas recombinant IL-6 ( Figure 4G) did not modulate NK cell cytotoxicity. To further assess the role of these cytokines in activating NK cell cytotoxicity, TNF-α, IL-10, and IL-6 were depleted by immunoprecipitation from the SN of β-glucan-primed PBMC (Figures 4B,E,H, respectively). Upon stimulation of NK cells with IL-10-depleted SN, we observed a complete abrogation of the β-glucan-triggered increase in NK cell cytolytic activity (Figure 4F), while depletion of TNF-α ( Figure 4C) or IL-6 ( Figure 4I) did not alter NK cell cytotoxicity. Since involvement of LPS, present in bacterial-derived CL, in IL-10 secretion by PBMC could not be completely excluded, we determined IL-10 levels in supernatant of PBMC stimulated with 10 ng/ml LPS, corresponding to the upper level of endotoxin concentration found in CL. The results (Supplementary Figure 4) show that LPS alone induces only minimal levels of IL-10 secretion by PBMC, whereas PBMC stimulated with MG do secrete substantial levels of IL-10. When MG is supplemented with LPS (10 ng/ml), a synergistic effect on IL-10 production by PBMC was observed. We therefore conclude that the β-glucan structure of MG alone is able to induce IL-10 production by PBMC, but that LPS is able to synergistically enhance this response. This might explain the slightly larger cytotoxic response of NK cells that is generally seen after stimulation with bacterial-derived CL compared to yeast-derived MG. To further confirm the role of IL-10 in the β-glucan mediated NK cell activation, expression of both the alfa-and beta-subunit of the IL-10 receptor by porcine NK cells was confirmed by RT-PCR ( Figure 5). These results not only highlight a pivotal role for IL-10 in the stimulation of NK cell cytotoxic capacity upon β-glucan priming of monocytes, but also identify IL-10 as a potent stimulus of porcine NK cell cytotoxicity.

DISCUSSION
NK cells are crucial players in the elimination of both virusinfected and malignant cells. Increasing evidence indicates that stimulation of NK cell activity may be a promising strategy in targeting viral diseases and cancer [reviewed in (38) and (39)]. Since swine husbandry is economically important worldwide as well, findings in porcine NK research are not only relevant with regard to their translational potential toward human medicine and research, but may also be implemented in swine industry. β-glucans have potent immunomodulatory properties [reviewed in (40)]. For example, they are able to alter the gut microbiota and immune system of pre-weaned piglets when supplemented to the feed (23). Furthermore, in-feed supplementation of β-glucans after weaning led to a reduced susceptibility to F4 + ETEC infection in vivo (41). Hence, β-glucans are looked at as an alternative to in-feed antibiotics to preserve animal health in commercially farmed pigs and as novel health promotor and therapeutic tool in human (para)medicine. Although the effects of β-glucans on immune cells have been investigated in pigs (11) and humans (42), their effects on NK cells remain largely unclear. One report suggested a direct effect of barley β-1,3/1,4-glucans on human NK cell cytotoxicity toward K562 cells (20). However, in that study, NK cells were not FACS-purified and a potential contribution of contaminating cell populations can therefore not be excluded. On the other hand, another study claimed that both zymosan and a particulate yeast β-1,3-glucan showed a direct and dose-dependent inhibition of human NK cell cytotoxicity. In this study, NK cells were also not FACS-purified and β-glucan concentrations were represented as relative particle:NK ratios, which complicates comparisons (22). Here, we aimed to elucidate whether β-glucans directly stimulate FACS-purified NK cells, using porcine primary NK cells and two different types of β-1,3-glucans, Macrogard and Curdlan.
The two tested β-glucans were unable to modulate NK cell cytotoxicity in a direct manner. This observation was also supported by the finding that the major β-glucan receptors, Dectin-1 and CD11R3, appear to be largely absent on porcine NK cells. Therefore, we conclude that there is no evidence for direct effects of β-glucans on NK cell cytotoxicity, at least in pig.
β-glucans have nonetheless been shown to activate innate immune cells such as monocytes, macrophages and neutrophils, including increased cytokine production by these cells (11)(12)(13). However, it was unclear if and how such increased cytokine production could alter NK cell cytotoxicity. We found that β-glucan treatment of PBMC resulted in a monocytedependent production of secreted factors that enhanced NK cell cytotoxicity in vitro. This is in line with a report describing indirect increased cytotoxicity of human NK cells after a 4day stimulation of PBMC with Lentinan, a β-1,3/1,6-branched glucan (21).
By depleting potentially involved cytokines from the supernatant, we could confirm the pivotal role of IL-10 in the activation of porcine NK cells upon β-glucan priming of monocytes. Although traditionally classified as an immunosuppressive cytokine, IL-10 has been described as an enhancer of NK cell cytotoxicity in human and mouse (43)(44)(45)(46)(47). Furthermore, IL-10-mediated increased NK cell cytotoxicity has been suggested to facilitate antigen uptake from dead cells by antigen presenting cells and thereby enhancing the cross-talk between the innate and the adaptive immune system during infection (48). Interestingly, Holder and Grant (49) described that cmvIL-10, a virokine homologous to human IL-10 encoded by human cytomegalovirus (HCMV), increased NK cell cytotoxicity in vitro as well. The authors could not determine whether this activity could benefit the virus in some way, but it appears to be in stark contrast with the generally acknowledged evasion strategies of herpesviruses, and cytomegaloviruses in particular, to slow down and suppress NK cell activity (50). Moreover, in acute murine cytomegalovirus (MCMV) infection in mouse, inhibition of the IL-10 receptor increased cell death in NK cells leading to a reduced number of cytotoxic NK cells in the spleen and lung. Together with our current data, this reinforces the hypothesis that IL-10 may contribute to antiviral innate immunity mediated by NK cells (51). Our current data that IL-10 increases NK cell cytotoxicity in pig are thus in line with similar observations in human (43)(44)(45), and further support the idea that experimental work on porcine NK cells may generate insights that are also relevant toward a better understanding of human NK cell biology (24,52,53).
In cancer biology, the exact role of IL-10 is insufficiently clear and both pro-tumorigenic and anti-tumorigenic functions have been described (54,55). However, the role of indirect NK cell activation in anti-tumor responses and the underlying mechanisms involved are increasingly studied. It has been shown that N-glycans expressed by tumor cells are recognized by tumor-associated macrophages (TAM) through Dectin-1. This recognition leads to induction of Dectin-1 signaling and is critical to the NK cell-mediated anti-tumor innate immunity (56,57). An in vivo study showed that there was a marked enhancement of metastasis of B16 tumor cells in the lungs of Dectin-1-deficient mice as compared to wild-type mice and that this was indeed due to lower NK cell activity. The exact molecules involved in the cross-talk between TAM and NK cells remain to be further characterized, although requirement of direct cell-cell contact was suggested (57). Based on our current results, however, IL-10 might be an interesting candidate molecule to study in this regard.
Taken together, these results show that β-glucans can significantly enhance NK cell effector functions by triggering NK cell-stimulating IL-10 secretion by monocytes. These findings highlight again the broad influence that β-glucans can exert on the immune response and therefore also the potential of these molecules in the development of novel therapeutic strategies, both in human as in veterinary medicine (42,58). Since administration of β-glucans is generally not accompanied by side effects, further insights into immune modulation and NK cell activation by these molecules may pave the way toward new strategies in vaccine development, anti-viral therapies and possibly cancer therapies.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by Ethical Committee of the Faculty of Veterinary Medicine (EC2017/121).

AUTHOR CONTRIBUTIONS
LH, SP, and SD performed the experiments. LH, EC, HF, and BD designed the research, made the figures, wrote the manuscript and analyzed the results. All authors contributed to the article and approved the submitted version.