IL-6 promotes NK cell production of IL-17 during toxoplasmosis

Abstract

Previous studies have implicated T cell production of IL-17 in resistance to T. gondii as well as the development of immune mediated pathology during this infection. Analysis of C57BL/6 and C57BL/6 RAG^-/- mice challenged with T. gondii identified NK cells as a major innate source of IL-17. The ability of soluble toxoplasma antigen to stimulate NK cells to produce IL-17 was dependent on the presence of accessory cells and the production of IL-6, IL-23 and TGF-β. In contrast, these events were inhibited by IL-2, IL-15 and IL-27. Given that IL-6 was one of the most potent enhancers of NK cell production of IL-17, further studies revealed that only a subset of NK cells expressed both chains of the IL-6R, IL-6 upregulated expression of the Th17 associated transcription factor RORγt, and IL-6^-/- mice challenged with T. gondii had a major defect in NK cell production of IL-17. Together, these data indicate that many of the same cytokines that regulate Th17 cells are part of a conserved pathway that also control innate production of IL-17 and identify a major role for IL-6 in the regulation of NK cell responses.

Introduction

The Th1-Th2 paradigm has dominated the study of immune-mediated resistance to infection for almost 20 years (1, 2). With the initial description of IL-17 as a cytokine produced by T cells (2-4), it was not clear how this pro-inflammatory factor related to the biology of established T helper subsets (5, 6). Early reports identified a role for IL-17 in the development and recruitment of neutrophils to sites of inflammation (7, 8), but studies by Cua and colleagues highlighted the ability of IL-17 producing CD4⁺ T cells, termed Th17 cells, to contribute to the development of inflammation in models of rheumatoid arthritis and EAE (9, 10). It is now appreciated that there are additional cellular sources of IL-17 that contribute to a variety of inflammatory conditions, including CD8⁺ T cells (11, 12), γδ T cells (13) and more recently iNKT cells (14-16).

While IL-17 has been implicated in the pathology associated with various forms of infection-induced inflammation (11, 17), it has also been linked to resistance to multiple pathogens including Klebsiella pneumoniae (18), Candida albicans (19) and Toxoplasma gondii (20). The development of a protective host response to intracellular pathogens requires the coordinated action of the innate and adaptive immune system. For many pathogens, the interaction between dendritic cells, macrophages, NK cells and neutrophils provide a limited mechanism of innate resistance during the early stages of infection, and also influence the development of adaptive immunity required for long term protection (21-23) these are mostly toxo refs and there are other reviews that would be better. Relatively little is known about the role of IL-17 in these early events, but IL-17R^-/- mice challenged with T. gondii have an early defect in neutrophil recruitment to the local site of infection and an increased parasite burden (20). Given reports that neutrophils play a important role in innate immunity to toxoplasmosis (24, 25), studies were performed to characterize the early cellular sources of IL-17 during toxoplasmosis. The comparison of C57BL/6 and C57BL/6 RAG^-/- mice, which lack T and B cells, revealed that NK cells were major contributors to the production of IL-17. This observation led to further experiments that established that many of the events that promote T cell secretion of IL-17 also induce the secretion of this cytokine from NK cells. Additionally, these findings highlight the prominent role of IL-6 in this regulatory pathway and describe a new role for this pleotropic cytokine during innate responses to an intracellular parasitic pathogen.

Materials and Methods

Parasites and infection

C57BL/6 (National Cancer Institute or from the Jackson Laboratory, Bar Harbor Maine), C57BL/6 RAG1^-/- (C57BL/6 12957- RAG1 ^{tm 1} Mom/j) (Jackson Laboratory) or C57BL/6 IL-6^-/- mice (Jackson Laboratory) were infected with 20 cysts of the ME49 strain of T. gondii intraperitoneally (i.p.) as previously described (26). For in vitro studies, a soluble Toxoplasma antigen – (STAg) (50µg/ml) that was prepared from the RH strain of T. gondii was used as previously described (27, 28) and tested for the absence of LPS and Mycoplasma contaminants by the University of Pennsylvania Cell Center (Philadelphia, PA).

Antibodies and Flow cytometry

The FITC-labeled Abs used were: NK1.1, CD3, CD4, CD8, CD11b, CD11c and CD19. PE-labeled Abs were: Gr-1, CD126 (anti-mIL-6Rα), RORγt and IL-17. Biotinylated Abs were: IL-6Rα and Rat IgG2α. Allophycocyanin-labeled Abs were: F4/80, CD130 (anti-mgp130), IL-12Rβ2 and IFN-γ. An Alexa 647-labeled antibody was used to detect T-bet while a PerCP-labeled Ab was used to assay expression of CD3, and a pacific blue labeled Ab was used to detect NK1.1. All antibodies were obtained from eBioscience, (San Diego, CA) or R&D Systems (Minneapolis, MN). To assay cytokine production, cells were pulsed with PMA (50 ng/ml; Sigma-Aldrich) and ionomycin (500ng/ml; Sigma-Aldrich). Two hours later cells were treated with brefeldin A (BFA) (10 ng/ml; Sigma-Aldrich), incubated for an additional 2 h, and then used for flow cytometry. For intracellular staining, surface-stained samples were washed, fixed and then permeabilized with 0.1% saponin in FACS buffer before staining for cytokine production as routinely performed in this laboratory (29). Flow cytometric analysis was performed on a FACSCalibur or FACSCantoII (BD Biosciences – San Jose, CA) instrument and analyzed using FlowJo software (Tree Star, Inc).

Analysis of NK cell production of cytokines

For depletion of NK cells in vivo, 50 µg of a rabbit anti-asialo GM1 antiserum (WAKO - Richmond, VA.) was administered i.p. 1 day prior to infection and thereafter every 3 days in a volume of 0.2 ml for the duration of the experiment. The depletion of NK cells were confirmed by flow cytometry on day 5 post-infection.

To assess NK cell responses in the presence of accessory cells, splenocytes from uninfected or infected RAG^-/- mice (day 5 post-infection) were prepared as previously described (30). These cells were cultured in RPMI 1640, (Life Technologies -Gaithersburg, MD) supplemented as previously described (30), and stimulated with STAg alone (50 µg/ml) or in the presence of the following cytokine neutralizing antibodies: α-IL-12/23p40 (10µg/ml), α-IL-12p19 (10µg/ml), α-TGF-β (10µg/ml), or various cytokines, including IL-2 (100U/ml) (R&D), IL-6 (50ng/ml) (eBioscience), IL-15 (10ng/ml) (eBioscience), IL-23 (10ng/ml) (eBioscience), IL-27 (100ng/ml) (Amgen), or TGF-β (1ng/ml) (eBioscience). After 20 hours in culture, IL-17A and IFN-γ were assayed by ELISA (eBioscience). Enriched populations of NK cells were obtained from splenocytes by negative selection using sheep anti-FITC magnetic beads (Polysciences -Niles, IL) to deplete CD3⁺, CD4⁺, CD8⁺, CD11b⁺, CD11c⁺ and CD19⁺ cells. The spleen cells were incubated for 20 minutes with FITC conjugated antibodies, before the addition of anti-FITC magnetic beads and then applied to a magnetic column (Polysciences, Inc.). After elution, the resulting cell preparation was washed 2 times with RPMI 1640, (Life Technologies - Gaithersburg, MD) supplemented as previously described (30) and routinely provided cells with > 90% purity, based on staining with NK1.1.

Real-Time quantitative PCR Analysis

Quantitative real-time PCR of total cellular RNA from splenocytes or purified NK cells isolated from uninfected or infected (day 5 post-infection) RAG^-/- mice was performed using standard procedures (31). Expression of RORγt, IL-6, IL-23 and TGF-β were determined with primers obtained from Qiagen (Valencia, CA) using an AB7500 fast real-time PCR thermal cycler and power SYBR green reagents (Applied Biosystems – Foster City, CA). The ‘housekeeping’ gene encoding β-actin was used as a normalization control.

Statistical Analysis

The student's t-test for two samples with unequal variance was used to determine statistical significance.

Results

NK cells are an early source of IL-17 during toxoplasmosis

To evaluate the early source of IL-17 during toxoplasmosis, C57BL/6 and RAG^-/- mice were infected with the ME49 strain of T. gondii and serum levels of IL-17 were measured at days 5 and 7 postinfection by ELISA. In uninfected mice, there were low basal levels of IL-17A, but following challenge, there was a marked increase in the level of circulating IL-17 in both sets of mice (Fig 1A). It should be noted that, across multiple experiments, infected RAG^-/- mice had 15-20% less IL-17 than T cell-sufficient wild-type C57BL/6 mice. Nevertheless, since RAG^-/- mice do not have T or B cells, this finding indicated that infection with T. gondii led to the production of IL-17 from an innate source.

Figure 1. NK cells are an early source of IL-17 during toxoplasmosis.

A) Kinetics of IL-17A production in the serum from wild-type C57BL/6 (n=3) and RAG^-/- mice (n=3) infected with 20 cysts from the ME49 strain of T. gondii measured by ELISA. Data shown from a representative experiment of three performances. B) Flow cytometry of IL-17A versus IFN-γ from splenocytes isolated from naive C57BL/6 mice (n=3) or from mice infected for five days with 20 cysts from the ME49 strain of T. gondii (n=3). Plots are gated on NK1.1⁺ CD3^- or CD3⁺ cells. Numbers in quadrants indicate the frequency of cells in each. Representative profiles are from three individual experiments.

Given current models of innate resistance to T. gondii in RAG^-/- mice (21, 32), likely sources of IL-17 would include macrophages, dendritic cells and/or NK cells. Therefore, intracellular staining for IL-17 was combined with flow cytometry of these immune populations to identify the cellular source of IL-17. Analysis of multiple macrophage and DC populations from infected mice failed to identify any IL-17⁺ cells (data not shown). When lymphocyte subsets were evaluated in naïve C57BL/6 mice, following a 4 h stimulation with PMA and ionomycin, there was a significant percentage of IFN-γ⁺ CD3⁺ T cells and NK1.1⁺ CD3NK cells, but minimal production of IL-17 by these cell types. By day 5 post-infection the frequency of IFN-γ⁺ CD3⁺ T cells was decreased but a small population of IL-17+ T cells was detected. However, at this early time point, infection resulted in an increased frequency of IFN-γ- and IL-17-producing NK cells, but dual cytokine producers were not detected (Fig 1B). To directly test the contribution of NK cells to the production of IL-17, C57BL/6 and RAG^-/- mice were treated with anti-asialo GM1 starting one day prior to infection to deplete this population, and serum levels of IL-17 were measured on day 5 post-infection. In this study, the depletion of NK cells led to a 95% reduction in NK1.1⁺ cells as evaluated by flow cytometry (data not shown), and led to a 65% and 77% reduction in serum IL-17 in C57BL/6 or RAG^-/- mice, respectively (Fig. 2A). When splenocytes from C57BL/6 or C57BL/6 RAG^-/- mice were re-stimulated on day 5 post-infection with STAg in vitro, IL-17 production was detected, and furthermore, the depletion of NK cells, reduced IL-17 production by 40% and 88% respectively (Fig. 2B and 2C). The differences in the relative contribution of NK cells in these experiments likely reflects the ability of IL-17 producing T cells to expand in vitro whereas NK cell expansion in these cultures is not observed. Consistent with previous studies in IL-17R^-/- mice (17), the administration of an anti-IL-17 antibody to infected RAG^-/- mice resulted in an approximate 40% decrease in the infection-induced infiltration of neutrophils into the peritoneum cavity (supplemental data Figure 1). Taken together, the flow cytometric data and the depletion studies identify NK cells as a physiologically significant source of IL-17 during toxoplasmosis.

Figure 2. Depletion of NK cells decreases IL-17A production during T. gondii infection.

A) Serum from wild-type C57BL/6 and RAG^-/- mice (n=4) treated in vivo with α-asialo GM1 to deplete NK cells (50 μg ip 1 day prior to infection and thereafter every 3 days in a volume of 0.2 ml for the duration of the experiment) was used to measure IL-17A by ELISA. Data shown from a representative experiment of five performances. *p<0.01 and **p<0.01. B and C) ELISA of IL-17A in the supernatants of splenocytes isolated at day 5 post-infection from wild-type C57BL/6 mice (n =4) or RAG^-/- mice (n=3) treated with or without a-asialo GM1 stimulated for 20 h in the presence or absence of STAg. Data shown from a representative experiment of three performances. *p<0.01.

Factors that regulate NK cell production of IL-17

To better understand the events that promote NK cell production of IL-17, the factors that influence the ability of splenocytes from uninfected or infected RAG^-/- mice to secrete IL-17 in response to STAg were examined. When cells from uninfected RAG^-/- mice were stimulated with STAg, low basal levels of IL-17 were detected, whereas splenocytes from infected mice produced 6 fold higher amounts (Fig 3A). This latter effect was dependent on the presence of accessory cells, as the ability of STAg to induce IL-17 was reduced when only enriched populations of NK cells were used (Fig. 3B). This finding suggested that NK cells require additional signals from accessory cells to produce IL-17. Previous studies have established that multiple accessory cell derived cytokines such as IL-6, IL-23 and TGF-β promote the development and maintenance of Th17 cells (10, 33-36). To determine the possible role of these cytokines in this experimental system, splenocytes from RAG^-/- mice infected for 5 days were cultured with STAg and the effects of specific neutralizing antibodies on the production of IL-17 were assessed. In these experiments, treatment with anti-IFN-γ had no effect (data not shown), but the addition of α-IL-6 resulted in a marked decrease in IL-17 production (Fig 3C). The use of α-IL-12/23p40 or α-IL-23p19 also resulted in a decrease in IL-17 secretion, implicating IL-23 in these events. Similarly, the blockade of TGF-β also led to a reduction in IL-17 levels. Stimulation of RAG splenocytes with STAG resulted in secretion of IL-6 (media = 51±7 pg/ml; STAg = 123±31 pg/ml) and to increased levels of IL-6 and IL-23 mRNA but not TGF-β (Supplemental data Figure 2). These findings are consistent with previous studies (37, 38) and while TGF-β production is regulated post-transcriptionally this stimulus does lead to production of biologically active TGF-β (39). These results suggest that at least three cytokines (IL-6, IL-23 and TGF-β) serve to promote NK cell IL-17 production.

Figure 3. Factors that regulate NK cell production of IL-17.

A) Levels of IL-17A measured by ELISA in the supernatant from uninfected or infected (day 5 pi) RAG^-/- mice (n=3) cultured with media or STAg for 20 h. B) IL-17A levels measured by ELISA in supernatants from cultured splenocytes isolated from infected RAG^-/- mice (day 5 pi) and NK cells purified from infected RAG^-/- mice, stimulated with or without STAg for 20 h. C) ELISA for IL-17A production was performed using supernatants from infected RAG^-/- mice cultured with media alone or STAg in the presence or absence of α-IL-6, α-IL-12/23p40, α-IL-23p19 or α-TGF-β for 20 h. The data represent means (±SD) from 3 independent experiments. *p<0.01 and **p<0.01. D) Supernatants of splenocytes from uninfected or infected RAG^-/- mice (n= 3) cultured with media or IL-6, IL-23, TGF-β or the combination of IL-6, IL-23 and TGF-β for 20 h. The data represent means (±SD) from 3 independent experiments. E) IL-17A ELISA was performed with the supernatants of purified NK cells from uninfected or infected RAG^-/- mice cultured for 20 h with media alone or in the presence of IL-6, IL-23, TGF-β or all three cytokines together. The data represents two independent experiments. F) Supernatants from cultured splenocytes derived from infected RAG^-/- mice (n=3) cultured with media, IL-6 or IL-6 plus IL-2 (ν), IL-15 (Δ) or IL-27 (O) at different concentrations for 20 h. The data represent means (±SD) from three independent experiments.

To complement the findings described above, additional experiments were performed to assess the effects of exogenous cytokines on the production of IL-17. Supernatants of cultured splenocytes from uninfected or infected RAG^-/- mice incubated alone or in the presence of various cytokines for 20 h were used to measure the secretion of IL-17. The inclusion of exogenous IL-1β or IL-18 did not promote IL-17 (data not shown), but the addition of IL-6, IL-23 or TGF-β alone to RAG^-/- splenocytes from uninfected mice led to low level IL-17 production compared to splenocytes incubated without any cytokines and the combination of all three cytokines led to the highest levels of IL-17 (Fig. 3D). When similar conditions were used for splenocytes from infected RAG^-/- mice, each of these stimuli led to approximately 3 times more IL-17 been produced than observed with uninfected splenocytes (Fig. 3D). When IFN-γ levels were measured under these same conditions IL-6 and IL-23 both enhanced the levels of IFN-γ in these culture supernatants whereas TGF-β had no effect (supplemental data Figure 3). Similar conditions were used with purified NK cells and while these cytokines alone can promote IL-17, a difference between cells from uninfected or infected mice was not apparent (Fig. 3E). Moreover, when bone marrow derived macrophages or dendritic cells were pre-incubated with STAG and the co-cultured with purified NK cells these conditions also led to increased secretion of IL-17 with the most significant effect observed with dendritic cells (Supplemental Figure 4). Taken together, these results are consistent with a model in which the ability of STAg to stimulate accessory cells to produce IL-6, IL-23 and TGF-β leads to NK cell secretion of IL-17.

While the data described above identified cytokines that promote NK cell secretion of IL-17, experiments were performed to assess whether known antagonists (IL-2 and IL-27) of Th17 cells (17, 40, 41)also influence NK cells. For these studies, IL-6 was selected as the primary stimulus as it was the most potent single cytokine that promoted NK cell synthesis of IL-17. As shown in figure 3F, the addition of IL-6 alone to splenocytes from infected RAG^-/- mice increased levels of IL-17 and the addition of IL-2, or its close relative IL-15, had a profound dose-dependent suppressive effect. A similar activity was observed when IL-27 was used (Figure 3F). Thus, it appears that the same factors that limit Th17 activity are also relevant to NK cells. To have an access to accessory cells stimulated with STAg restores NK cell response, bone marrow driven macrophages and dendritic cells where cultured with NK cells from infected RAG^-/- mice alone or with STAg plus NK cells from infected RAG^-/- mice for 6 h. ELISA for IL-17 reveled that dendritic cells, more than macrophages, are able to restore NK cell IL-17 production (supplemental data Figure 4).

The role of IL-6 in the regulation of the NK cell response

The results described in the previous section identified NK cells as a source of IL-17 during acute toxoplasmosis and indicated a key role for IL-6 in these events. To better understand how IL-6 promoted a fraction of NK cells to produce IL-17, the surface levels of the two components of the IL-6R, IL-6Rα and gp130 (42) on NK cells from uninfected and infected mice was assessed. In uninfected mice, almost all NK cells expressed gp130, but only a subset expressed the α-chain (Fig. 4A). Following challenge, gp130 expression remained constant, but there was an approximate two fold increase in the frequency of cells which express the IL-6Rα chain which was observed across multiple experiments (Fig. 4A and B). Interestingly, although the IL-17⁺ and IFN-γ⁺ NK cells appear to be distinct populations (Fig. 1B), the IL-6Rα⁺ population of NK cells also express the IL-12Rβ2 chain (Fig. 4C).

Figure 4. Examination of IL-6R expression NK cells during toxoplasmosis.

A) Flow cytometry of gp130 and IL-6 receptor alpha on gated NK1.1⁺ CD3^- cells from splenocytes of uninfected and infected RAG^-/- mice (n=5). Representative profiles are shown. Numbers in quadrants indicate the frequency of cells in each. The data represents three independent experiments. B) Frequency of IL-6Rα positive cells from splenocytes of uninfected or infected RAG^-/- mice (n=5). Data from a representative experiment. C) Flow cytometry of IL-6Rα and IL-12Rβ2 on NK1.1⁺CD3^- cells from splenocytes of uninfected or infected RAG^-/- mice. Representative profiles from 5 mice are shown. Numbers in quadrants indicate the frequency of cells in each.

When NK cell expression of the transcription factor RORγt, a master regulator of Th17 cells (43), was examined NK cells from uninfected or infected RAG^-/- mice displayed basal levels of RORγt mRNA transcripts (Fig. 5A). Stimulation of these cells with IL-23 or TGF-β did not alter RORγt expression, but the addition of IL-6 resulted in a 100-fold elevation in RORγt mRNA levels (Fig. 5A). When intracellular staining was used to compare levels of T-bet and RORγt in NK cells from uninfected mice, this analysis revealed that there is a sub-population of NK cells that are RORγt + but which express low levels of T-bet and these cells expand in response to infection. Although intracellular staining for these transcription factors resulted in a reduced ability to detect IL-6Rα (data not shown) by gating on the IL-6Rα+ NK cells these data showed this sub-population expressed the highest levels of RORγt but the lowest levels of T-bet (Figure 5C). While these data have to be interpreted with care, they suggest a model in which the ability of NK cells to produce IL-17 is associated with expression of the IL-6Rα, RORγt and the absence of T-bet.

Figure 5. Expression of the transcription factor RORyt and T-bet in NK cells.

A) Total RNA was prepared from purified NK cells from splenocytes of uninfected and infected RAG^-/- mice (n=3) cultured with media, IL-6, IL-23 or TGF-β alone for 20 h. The mRNA expression was measure by quantitative real-time PCR. The data is represented as fold induction relative to the house-keeping gene β- actin. Data shown representative of 3 independent experiments. B) Flow cytometry of RORγt and T-bet gated on NK1.1⁺ CD3^- cells from splenocytes of uninfected or 5 days post-infection infected RAG^-/- mice (n=6). Numbers in quadrants indicate the frequency of cells in each. The data are representative of four experiments. C) Flow cytometry of IL-6Rα expression on RORyt⁺ or T-bet⁺ NK1.1⁺ cells (dotted line) or RORγt^- or T-bet^- NK1.1⁺ cells (black line). The shaded histogram represents the isotype control. Representative profiles are shown. The data represents four independent experiments.

To evaluate the biological significance of the effects of IL-6 on NK cells, wild-type C57BL/6 mice and IL-6^-/- mice were infected and on days 3, 5 and 7 post-infection serum levels of IL-17 were analyzed (Fig. 6A). In these experiments, the absence of IL-6 led to decreased serum levels of IL-17. In addition, the frequency of NK and T cells producing IL-17 were analyzed by flow cytometry in naive C57BL/6 mice and IL-6^-/- mice infected for six days (Fig. 6B). There was a sizeable decrease in the frequency of IL-17-producing CD3⁺ T cells cells as well as by NK1.1⁺ cells in comparing wild-type C57BL/6 mice to IL-6^-/- mice (Fig. 6B). Together, these data confirm the importance of IL-6 for Th17 differentiation, and provide new insights into its role in the regulation of NK cell IL-17 production.

Figure 6. IL-6 has a critical role in the production of IL-17 by NK cells.

A) Serum from wild-type C57BL/6 or IL-6^-/- mice (n=4) was used to measure the production of IL-17A by ELISA on days 3, 5 and 7 post-infection. Data shown are from one representative experiment. B) Flow cytometry of intracellular IL-17 gated on NK1.1⁺ CD3^- cells or CD3⁺ T cells from splenocytes of day six infected wild-type C57BL/6 mice or IL-6^-/- mice (n=5). Numbers in quadrants indicate the frequency of cells in each. Representative profiles are shown. The data presented are from three independent experiments.

Discussion

The initial report in which the IL-17R^-/- mice were challenged with T. gondii indicated a role for IL-17 in the mobilization of neutrophils required for resistance to this pathogen (20-22). In those studies, the source of IL-17 was not addressed, but subsequent work indicated that CD4⁺ and CD8⁺ T cells produced this factor during chronic toxoplasmosis (17). In other experimental models, γδ and iNKT cells also secrete IL-17 (14, 44) but, to the best of our knowledge, there have been no reports that NK cells produce IL-17. Consequently, the identification of NK cells as a relevant source of IL-17 during acute toxoplamosis was unexpected. Studies from multiple laboratories identified NK cells as an important source of IFN-γ and necessary for innate resistance to T. gondii (45-48) as well as other intracellular pathogens, including Listeria monocytogenes (49) and T. cruzi (50). This led to a body of work that defined the events that promoted this pathway of resistance, and the development of a model in which the recognition of microbial antigens by macrophages and dendritic cells leads to the production of a number of proinflammatory cytokines including IL-12, which promote NK cell synthesis of IFN-γ (51) (48, 52). The data presented here suggest a similar series of events, whereby infection with T. gondii and/or parasite-derived antigens lead to accessory cell production of IL-6 and IL-23, two cytokines related to IL-12 (53), which in combination with TGF-β promote IL-17 production by a distinct population of NK cells. It is notable that splenocytes derived from infected RAG mice routinely produced more IL-17 in response to STAg than those from uninfected mice. Perhaps the simplest explanation is that there are higher basal levels of pro-inflammatory factors present in the cultures from infected mice. Another possibility is that during toxoplasmosis, the splenic accessory cells produce additional cytokines/co-factors that co-operate with IL-6, IL-23 and TGF-β to promote IL-17 secretion.

Given the common cytokine and signaling pathways that drive NK and T cell production of IFN-γ, it is not surprising that some of the same elements that are relevant to Th17 differentiation are also pertinent to NK cells. Thus, IL-6, IL-23 and TGF-β are all associated with promoting or maintaining Th17 responses, and the ability of IL-6 to induce expression of mRNA transcripts for RORγt, the Th17 master regulator, in NK cells highlights the common molecular pathways that are relevant to IL-17 production. Interestingly, while we have focused on the effects of IL-6 to support an IL-17 response, a recent report concluded that IL-6 does not promote the ability of iNKT cells to produce IL-17 (16). This latter observation is consistent with the description of IL-6-independent circumstances that promote Th17 cells (54), and suggests that the relative contribution of different cell populations to the pool of IL-17 is profoundly influenced by the local cytokine response. Additionally, the shared biology of T and NK cell populations is also relevant to antagonists of Th17 cells. Thus, IL-2 and IL-27 are potent antagonists of T cell production of IL-17 (17, 41, 55), and, along with IL-15, have a similar inhibitory effect on NK cells. Taken together, these data indicate that the same pathways that have been shown to regulate T cell production of IL-17 are also relevant to the TCR-independent process in NK cells.

It is relevant to note that while others have previously shown that NK cells express the IL-6R (56), there is remarkably little known about the functional effects of IL-6 on these cells. Previous reports have highlighted that while IL-6 has no effect on cytotoxicity, it does enhance NK cell adhesion and proliferation when combined with IL-15 in vivo (57). However, the ability of IL-6 to promote NK cell expression of RORγt and IL-17 and the defect in the numbers of IL-17⁺ NK cells in IL-6^-/- mice challenged with T. gondii suggests that promoting innate IL-17 may be one of the main biological functions of IL-6. This in turn may impact on the interpretation of the phenotype of the IL-6^-/- mice infected with T. gondii (or other pathogens). For instance, IL-6^-/- mice have been reported to have deficient neutrophil responses (58) and an increased susceptibility to toxoplasmosis (59, 60), One possible explanation for this outcome is that the increased susceptibility of these mice to T. gondii may be a result of the absence of IL-17 and consequently reduced neutrophils, a notion that has yet to be tested directly.

In the last 15 years there has been an increased appreciation of the functional role of NK cells in promoting resistance to infection and tumor surveillance. The data presented here add the production of IL-17 to that list. These studies also contribute to the idea that there are distinct subsets of NK cells that can be defined based on their capacity to produce distinct patterns of cytokines (61-63). The characterization of the expression patterns of activating and inhibitory receptors (64) as well as the IL-12Rβ2 chain (65) has indicated the existence of multiple NK cell subsets. The observation that only a portion of resting NK cells express both chains of the IL-6R, and that IL-17⁺ IFN-γ⁺ double producers were not observed as well as the distinct pattern of T-bet and RORγt expression in NK cells supports this idea. Thus, in contrast to T cells, in which a naïve T cell can adopt a multitude of cell fates (Th1, Th2, Th17, TFH, Treg), it seems possible that NK cells may be hard wired to produce certain combinations of cytokines that provide the innate equivalent of distinct T cell subsets. This may be most apparent during challenge with pathogens like T. gondii where a prominent population of IFN-γ⁺ NK cells is present, (48) whereas during helminth infection these lymphocytes are a potent source of IL-13 (66). Those previous studies, and those presented here, have not yet distinguished whether a mature NK cell has the potential to produce either IFN-γ, IL-13 or IL-17 or acquires a more exclusive phenotype through education or development. Regardless, NK cells provide a model to study the events that influence cytokine production and may provide new insights into the mechanisms that regulate this pathway without the confounding factor of TCR and co-stimulation requirements.

Abbreviations used in this paper

NK

natural killer cells

STAg

soluble Toxoplasma Ag