Macrophages participate in IL17-mediated inflammation

Abstract

The involvement of macrophages in Th17 responses is still poorly understood. While neutrophils are thought to be the predominant effector of Th17 responses, IL17 is also known to induce myelotropic chemokines and growth factors. Other T cell-derived cytokines induce nonclassical functions, suggesting that IL17 signaling may similarly elicit unique macrophage functions. We characterized the expression of subunits of the IL17 receptor on primary murine macrophages from different anatomical compartments. The greatest expression of IL17 receptors was observed on mucosal Ly6C^hi “inflammatory” macrophages. We further observed upregulation of IL17 receptors in vitro on bone marrow-derived macrophages in response to peptidoglycan or CpG oligonucleotide stimuli, and in vivo, upon CFA administration. Macrophages expressing IL17 receptors were observed infiltrating the hearts of mice with myocarditis, and genetic ablation of IL17RA altered macrophage recruitment. Treating primary macrophages from a wide variety of different anatomic sources (as well as cell lines) with IL17A induced the production of unique profiles of cytokines and chemokines, including GM-CSF, IL3, IL9, CCL4/MIP1β and CCL5/RANTES. IL17A also induced production of IL12p70; IL17-signaling deficient macrophages elicited diminished IFNγ production by responding DO11.10 CD4⁺ cells when used as APCs. These data indicate that macrophages from different anatomic locations direct IL17-mediated responses.

Introduction

The prototypic proinflammatory cytokine of the IL17 family, IL17A, defines a novel CD4⁺ subset, the Th17 lineage [1, 2]. Associated mainly with fungal and extracellular bacterial infection, data have recently emerged implicating Th17-mediated responses in autoimmune disease [3, 4]. Inflammatory effector functions mediated by IL17 are thought to be predominantly mediated through host tissues, particularly mesenchymal cells [5]. The major hematopoietic contributor to IL17-induced inflammation is thought to be the neutrophil, through the induction of G-CSF and ELR-subfamily CXC chemokines in host tissue [6]. IL17A has also been reported to induce cytokines and chemokines that are tropic for monocyte-lineage cells, including GM-CSF, CCL2/MCP1, CCL5/RANTES, and CCL20/MIP3α [7]. These reports suggest plausible mechanisms by which macrophages participate in IL17-mediated inflammatory responses.

Concurrently, a number of studies have revealed the exceptional diversity and heterogeneity of cells of the monocyte and macrophage lineages. Alternative activation of macrophages, under the influence of Th2 cytokines, tolerogenic IL10, or the tumor microenvironment, contrasts with classical activation by IFNγ [8, 9]. Circulating monocytic precursor subpopulations have defined progenitors for “resident” (Ly6C^lo) macrophages at steady state, or are recruited to sites of inflammation (Ly6C^hi) under the control of CCR2 signaling [10, 11]. The increasing appreciation of macrophage heterogeneity further suggests that macrophages may participate in Th17-mediated inflammatory responses in ways that are unique and non-redundant with already described “classically-activated” M1 macrophages, or “alternatively-activated” M2 macrophages [8, 9].

Our intention is to identify the mechanisms by which macrophages act as effectors of Th17 immunity. Our first goal was to describe the expression of the receptor subunits for IL17A, IL17RA and IL17RC, on primary macrophages from different anatomic compartments; we observed substantial heterogeneity of expression. We further found that IL17 receptors could be upregulated on macrophages in vitro, as well as in vivo, in response to different inflammatory stimuli. We next examined cytokine and chemokine production by multiple macrophage populations in response to IL17 itself, and report the induction of unique cytokine and chemokine profiles. These data indicate that IL17A induces macrophage activation in a unique manner that differs from other T cell-derived cytokines.

Materials and Methods

Mice

DO11.10 and BALB/cJ mice were obtained from the Jackson Laboratory (Bar Harbor, ME) and maintained in the Johns Hopkins University School of Medicine specific-pathogen free vivarium. Act1^−/− founder mice were the generous gift of Dr. Xiaoxia Li (Cleveland Clinic), and maintained on a Rag2^−/− background [12]. All rights, title, and interest in Act1^−/− mice are owned by The Cleveland Clinic Foundation. IL17RA^−/− founder mice were generously provided by Dr. Jay K. Kolls (Louisiana State University Health Sciences Center), under MTA with Amgen [13]. Experiments were conducted on 6–12 week old female mice. For CFA experiments, 100 μL of complete Freund’s adjuvant (Sigma, St. Louis, MO) emulsion was administered sc into the axillae of mice. For elicitation of peritoneal Mφ, mice were injected with 1.0 mL sterile 3% proteose peptone (Difco Labs, Detroit, MI) in 1× PBS ip, 5 days prior to harvest. All methods and protocols involving mice were approved by the Animal Care and Use Committee of Johns Hopkins University.

Isolation of primary tissue macrophages

Spleen, lung, liver, and CNS were resected from mice perfused with 1× PBS under Avertin anesthesia. Small intestines were collected, and generously rinsed with 1× HBSS + 2% BSA + 2 mM EDTA. Tissues were minced and digested for <1h in 50 U/mL Collagenase II + 50 U/mL Collagenase VIII (Sigma), followed by generous washing. Single cell suspensions were prepared, and RBCs lysed by <2 min incubation in ACK buffer (Biofluids, Gaithersburg, MD), followed by generous washing. Lavage cells from the bronchoalveolar tract or peritonea were collected in 1× HBSS.

Bone marrow derivation of macrophages

Femurs and tibiae were resected from mice, and HSCs collected from bone marrow. RBCs were optionally briefly lysed by <2 minute incubation in ACK lysis buffer (Biofluids). Cells were washed generously, and seeded in 48-well format in complete DMEM + 10% FBS, supplemented with 50 ng/mL rmM-CSF or 10 ng/mL rmGM-CSF (R&D Systems). Cells were incubated for 6–7 days, with one refeed. Non-adherent cells were thoroughly depleted before use by generous rinsing.

In vitro stimulation of macrophages

Primary cells from above were enriched for macrophages by paramagnetic positive selection with CD11b⁺ microbeads, according to manufacturer’s instructions (Miltenyi), and seeded at a density of 5 × 10⁵ cells in 48-well format, or 10⁶ in 24-well format in complete DMEM + 10% FBS. Adherent macrophages were rinsed in 1× HBSS. Cells were stimulated for 24h with recombinant murine IL17A (R&D Systems) at concentrations of 0–10 ng/mL. For PRR experiments, cells were stimulated with 1.0 μg/mL E. coli LPS (Sigma), 1.0 μg/mL peptidoglycan (Biochemika), 100 ng/mL CpG ODN 1668 (Alexis/Enzo Life Sciences, Farmingdale, NY), 1.0 μg/mL zymosan (Sigma), 1.0 μg/mL CM-curdlan (Wako), 100 ng/mL poly(I:C) (Sigma), 1.0 μg/mL TSLP (R&D Systems).

Induction of experimental autoimmune myocarditis

Previously well-established methods were used to induce EAM [14]. Briefly,100 μg commercially fMOC synthesized MyHCα_614–629 [15] was emulsified 1:1 in CFA (Sigma, St. Louis, MO) supplemented with 5 mg/mL of heat-killed M. tb strain H37Ra (Difco, Detroit, MI), and installed sc in mice twice on days 0 and 7. On day 0, mice also received 500 ng of pertussis toxin ip (List Biologicals, Campbell, CA).

Flow cytometry

Single cell suspensions were extracted into 1× PBS + 0.5% FBS, and RBCs lysed by <2 minute incubation in ACK lysis buffer (Biofluids). For intracardiac cytometry, animals were perfused and single cell suspensions prepared as previously described [16], with modifications as detailed by manufacturer’s recommendations for the GentleMACS™ tissue dissociation system (Miltenyi). Cells were washed, and FcγRII/III blocked with αCD16/32 (eBiosciences). Surface markers were stained with fluorochrome-conjugated Abs to CD4, CD11b, CD11c, F4/80, Ly6C, Gr1 (Ly6C/Ly6G), IL17RA, and IL17RC (eBiosciences, Serotec AbD, R&D Systems). For intracellular staining, cells were fixed and permeabilized, stained for IFNγ and IL17 (eBiosciences), and washed. Samples were acquired on a LSR II quad-laser cytometer running FACSDiva (BD Immunocytometry). Data were analyzed with FlowJo 7.5 (Treestar Software).

Confocal immunofluorescent microscopy

Tissues were resected from naïve BALB/c mice, briefly fixed in 5% paraformaldehyde, and preserved in an increasing sucrose gradient prior to snap-freezing in OCT (Takura). 10 μm frozen sections were cut, fixed in acetone, and stained with fluorochrome-conjugated Abs to F4/80 Pacific Blue (Invitrogen), IL17RA PE, and IL17RC APC (R&D Systems). Images were acquired on the Zeiss Meta 5100 confocal microscope at the JHUSOM Microscopy Core Facility.

Multiplex cytokine array

Linco multiplex cytokine arrays were employed, according to manufacturer’s instructions (Millipore). Briefly, supernatants from IL17A-stimulated Mφ were collected and stored at −80° C. Plates were acquired on a Luminex XMAP reader.

T cell stimulations

T cells were isolated from spleens of naïve DO11.10 TcR transgenic mice, and purified by CD4⁺CD62L⁺ naïve T cell isolation kit, according to manufacturer’s instructions (Miltenyi Biotec). 2 × 10⁵ CD4⁺CD62L⁺ cells were cocultured with BMMφ, in the presence of 10 μg/mL OVA_323–339 (Anaspec) in complete DMEM + 10% FBS in 48-well format for 96h @ 37° C, 5% CO₂. At the end of stimulation, cells were pulsed with GolgiStop/GolgiPlug (BD Pharmingen) for 6h, prior to intracellular cytokine staining.

Statistics

Normally-distributed data were analysed by two-tailed Students’ t-test (Microsoft). Analyses involving multiple tissues and cell populations involved an x-immediate multiple linear regression model, treating tissue and subpopulation as interdependent dimensions (Stata); for these data, p values are reported for individual tissues compared to spleen, and between Ly6C^hi and Ly6C^loGr1⁻ macrophages.

Results

IL17 receptors are variably expressed on murine macrophages

In order to compare expression of IL17 receptors on primary murine macrophages, we prepared single cell suspensions from a variety of hematopoietic and non-hematopoietic tissues including spleen, peripheral blood, peritoneal cavity, lamina propria of the small intestine, lung, liver, and central nervous system. Cells were stained with antibodies to CD45, CD11b, and F4/80 to verify their identities as macrophage-lineage cells (Supplemental Figure 1). Cells were also stained with Gr1 and Ly6C to categorize them as “inflammatory”-type (Ly6C^hi) or “resident”-type (Ly6C^loGr1⁻) macrophages [10]. We have previously observed mutually exclusive staining of Ly6C and the homeostatic chemokine receptor CX3CR1, allowing us to rely on Ly6C expression as a surrogate for chemokine receptor expression (unpublished observations).

Representative histogram overlays of IL17RA and IL17RC expression on Ly6C^hi inflammatory monocytes and macrophages from several tissues, including spleen, peritoneum, lamina propria, and lung, are shown in Supplemental Figure 2, compared to control staining. Staining for IL17RA and IL17RC was observed in peritoneal and gut macrophages, while staining of spleen macrophages was comparable to negative controls. Lung macrophages stained for IL17RC, but not for IL17RA. In separate experiments, we isolated lung macrophages from bronchoalveolar lavage, and observed high levels of IL17RA expression (data not shown), in contrast to the intermediate levels of receptor expression shown here on predominantly parenchymal lung macrophages.

Our use of a standard digestion protocol across tissues allowed for lateral comparisons between tissues for receptor expression by fluorescence intensity analysis for IL17RA and IL17RC (Figure 1). By statistical analysis of the data using an x-immediate multiple linear regression model, we were able to assess the relative contributions of anatomical location and macrophage subtype to differences in IL17 receptor expression. Generally, Ly6C^hi inflammatory Mφ stained more brightly for IL17RA than Ly6C^loGr1⁻ resident macrophages (p < 10⁻³). This distinction was significant (p < 10⁻³), but less pronounced for IL17RC staining. In general, the highest levels of IL17RA and IL17RC staining were observed in macrophages associated with mucosal interfaces, particularly gut and peritoneum. The lowest levels of staining were observed in cells isolated from lymphoid tissues, particularly spleen and blood. We observed substantially more differential staining by compartment for IL17RC than for IL17RA; macrophages from mucosal surfaces stained several logfold more brightly for IL17RC than macrophages from spleen or peripheral blood.

Figure 1.

Expression of IL17 receptors on primary murine macrophages. Quantitative analysis of fluorescence intensities of A) IL17RA and B) IL17RC staining on primary murine Ly6C^hi inflammatory (filled) and Ly6C^loGr1⁻ resident (open) macrophages, by compartment. Statistics are by x-immediate multiple linear regression of mean fluorescence intensity against tissue and subpopulation. Statistics denote comparisons to lymphoid tissues or comparison of Ly6C^hi to Ly6C^loGr1⁻ cells. Data indicate means of individual animals plus standard deviation, n = 5. Representative of three independent experiments.

Expression of both IL17RA and IL17RC is required for ligand-mediated cell signaling [17, 18]. Therefore, we determined coexpression of these receptors at the single-cell level, and observed appreciable costaining of IL17RA and IL17RC on macrophages in peritoneum and gut, but not from spleen (Figure 2a). Gating based on Ly6C and Gr1 for inflammatory- and resident-type Mφ showed the highest proportions of IL17RA⁺IL17RC⁺ double-positive cells in peritoneal and gut macrophages (p < 10⁻³) (Figure 2b). Generally, a greater proportion of Ly6C^hi inflammatory Mφ were IL17RA⁺IL17RC⁺ double-positive, compared to Ly6C^loGr1⁻ resident Mφ (p < 0.001). In separate experiments, bone marrow-derived macrophages costained for low levels of IL17RA and IL17RC expression (data not shown).

Figure 2.

Coexpression of IL17RA and IL17RC on primary murine macrophages. A) Representative bivariate psuedocolor plot of IL17RA and IL17RC staining on spleen, peritoneal, and lamina propria F4/80⁺-gated macrophages. IL17RA⁺IL17RC⁺ double-positive gate is based on negative staining for isotype control (not shown). B) Quantitative analysis of IL17RA⁺IL17RC⁺ double-positive staining on F4/80⁺Ly6C^hi inflammatory (left) and F4/80⁺Ly6C^loGr1⁻ resident (right) macrophages, by compartment. Statistics are by x-immediate multiple linear regression of MFI against tissue and subpopulation. Arrows denote comparisons to spleen; printed p values denote comparison of Ly6C^hi versus Ly6C^loGr1⁻ cells. Data indicate means of individual animals plus standard deviation, n = 5. Representative confocal immunofluorescent staining of IL17RA (green), IL17RC (red), and F4/80 (blue) in C) duodenum and D) spleen of naïve mice. Data are representative of two independent experiments.

To confirm these data, we stained sections of primary tissues from naïve mice and imaged by confocal microscopy. In duodenum, IL17RA and IL17RC costained with F4/80 in the lamina propria (Figure 2c). In contrast, spleen F4/80⁺ macrophages did not appreciably costain with IL17RA or IL17RC; interestingly, a population of small IL17RA⁺IL17RC⁺ cells was observed in lymphoid regions of white pulp (Figure 2d). Controls showed minimal background staining of IL17RA or IL17RC (Supplemental Figure 3).

IL17 receptor expression on macrophages upregulates in response to inflammatory stimuli

To test the possibility that macrophages tune their responsivity to IL17 signaling by regulating receptor expression in response to inflammatory stimuli or proinflammatory cytokines [17], we differentiated macrophages from proliferating bone marrow hematopoietic precursors with M-CSF. Bone marrow-derived macrophages (BMMφ), representing a highly homogenous population, were stimulated for 24h in the presence of model agonists for various Toll-like receptors (TLRs). These included LPS for TLR4, peptidoglycan (PGN) for TLR2/6 and NOD2, the CpG oligonucleotide ODN1668 for TLR9, poly(I:C) for TLR7, zymosan for TLR1/2 and Dectin-1, CM-curdlan for Dectin-1, and the Th2-associated alarmin TSLP. As shown in Figure 3a–b, BMMφ are largely unresponsive to TLR stimulation in terms of IL17RA expression. However, PGN and CpG stimulation dramatically upregulated the expression of IL17RC on BMMφ (Figure 3a,c).

Figure 3.

IL17 receptor expression on murine BMMφ, in response to inflammation in vitro. A) Representative histograms of IL17RA (left) and IL17RC (right) expression on BMMφ in response to TLR ligands; isotype control (filled), no stimulus (dashed), 6.0 μg/mL CpG ODN1668 (solid, thick). Quantitative analysis of B, D) IL17RA and C, E) IL17RC expression on BMMφ in response to (B, C) TLR ligands alone, and (D, E) TLR ligands in combination with proinflammatory cytokines TNFα (filled) and IFNγ (diagonal hatching) after 24h of stimulation. Data indicate means of duplicate wells plus standard error. Asterisks denote (p < 0.05) by two-tailed Students’ t-test, compared to control. Representative of three independent experiments.

We also examined the interaction of TLR pathways with other proinflammatory pathways, including the cytokines TNFα and IFNγ. The addition of TNFα or IFNγ did not substantially affect the expression of IL17RA. TNF signaling diminished the expression of IL17RA, but increased expression of IL17RC. IFNγ had no effect on IL17RA expression, but upregulated IL17RC (Figure 3d–e). There was little additive or synergistic effect of these cytokines on concurrent TLR signaling; for the most part, TLR signaling overrode TNF or IFNγ, with one exception – TNF enhanced the upregulation of IL17RC induced by CpG ligation of TLR9 (Figure 3e).

As a corollary to TLR regulation of IL17 receptors, we examined regulation of components of the IL17 signaling pathway, including Act1, IκBζ, and TRAF6. We observed upregulation of IκBζ by all PRR stimuli tested, though this effect was most potent for LPS at the early timepoint, and CpG at the later timepoint (Supplemental Figure 4).

We further examined the regulation of IL17 receptor expression by inflammation in vivo. We immunized mice with complete Freund’s adjuvant (CFA), a potent proinflammatory adjuvant containing heat-killed Mycobacteria critical for the induction of animal models of autoimmune disease (Cihakova D., et al., manuscript in preparation). Staining for IL17RA and IL17RC on macrophages from multiple compartments was determined 48 hours later, and compared to unimmunized controls. The most dramatic increases in IL17RA expression were observed on liver Mφ. Representative histogram overlays show acquisition of IL17RA expression by a subpopulation of both inflammatory and resident-type liver Mφ, in response to CFA administration. Küpffer cells from CFA-treated animals exhibited increased staining for IL17RA, compared to unimmunized controls (Figure 4a–b). Interestingly, CFA did not induce changes in IL17RC expression (data not shown).

Figure 4.

IL17 receptor expression on primary murine macrophages, in response to inflammation in vivo. A) Representative histograms of IL17RA expression on primary liver macrophages 48h following sc CFA administration (open, thick), compared to control naïve mice (filled). Quantitative analysis of IL17RA expression on B) liver macrophages and C) peripheral blood monocytes in response to CFA (filled). Control animals were given sc sterile PBS (open). D) Quantitative analysis of IL17RA and IL17RC expression on CD11b^hiF4/80⁺Ly6C^hi inflammatory (left) and CD11b^hiF4/80⁺Ly6C^loGr1⁻ resident (right) peritoneal macrophages in response to sterile elicitation by proteose peptone (filled). Control animals were administered vehicle PBS alone (open). Data represent individual animals (diamonds), and mean of each group (bars). Statistics are by two-tailed Students’ t-test. Representative of three independent experiments.

We also detected increased IL17RA expression on peripheral blood resident-type Ly6C^loGr1⁻ macrophages (Figure 4c), but did not observe substantial changes in IL17 receptor expression on Mφ from other locations, including CNS, spleen, lung or gut at this timepoint (data not shown). Sterile elicitation of peritoneal macrophages represents a commonly used source of macrophages; therefore, we examined IL17 receptor expression on peritoneal Mφ five days after sterile elicitation with ip proteose peptone, and observed no differences compared to naïve resident Mφ (Figure 4d).

IL17 receptor expression and signaling in autoimmune heart disease

We further sought to examine whether inflammation-dependent regulation of IL17 receptor expression within the macrophage compartment held any bearing on immunopathologic states. To study this question, we turned to a model of autoimmune heart disease in mice, in which we have previously described that macrophages and monocytes comprise the majority of infiltrating inflammatory leukocytes [16, 19]. Heart-infiltrating macrophages on days 0, 14, and 21 were interrogated by cytometric methods. As shown in Figure 5a, inflammatory cardiac infiltrates were comprised of an increased number of CD11b⁺F4/80⁺ macrophages expressing IL17RA and IL17RC, compared to cardiac-resident macrophages in naïve mice.

Figure 5.

IL17 receptor expression and signaling in experimental murine autoimmune myocarditis. A) Enumerative FACS analysis of expression of IL17RA (left), IL17RC (middle) and double-positive (right) CD11b⁺F4/80⁺CD45⁺ intracardiac macrophages on days 0, 14, and 21 of EAM. B) Proportion and C) absolute enumerative cardiac macrophage infiltration in IL17RA^−/− (filled) and WT (open) mice at day 21 of EAM. Data represent individual animals (diamonds), and mean of each group (bars). Statistics are by two-tailed Students’ t-test. D) Cytometric analysis of M-CSF-differentiated bone marrow-derived macrophages from IL17RA^−/− (dotted pink) and WT (thick blue) mice. Isotype controls are tinted grey.

We have demonstrated that IL17A-deficient mice develop similar EAM to wild type BALB/c control mice, but are protected from progression from EAM to dilative cardiomyopathy at later timepoints [20]. IL17RA-deficient mice develop a similar disease phenotype (unpublished). We examined cardiac infiltrates in IL17RA^−/− at day 21 of EAM cytometrically to determine a role for IL17 signaling upon the expansion, recruitment, or differentiation of macrophages to the inflamed heart. We observed a relative decrease in Ly6C^hi “inflammatory” macrophages and an increase in Ly6C^lo “resident”-type macrophages (Figure 5b). When these proportions were calculated to absolute values, it appeared that this shift in macrophage infiltration was largely due to enhanced recruitment, expansion, or differentiation of Ly6C^lo macrophages (Figure 5c).

In order to examine cell-intrinsic effects of IL17 signaling upon macrophage differentiation, macrophages were derived from bone marrow precursors with M-CSF. IL17RA^−/− BMMφ definitively displayed no expression of IL17RA and minimal expression of IL17RC (Figure 5d). We further observed diminished expression of MHC Class II, as well as CD206/macrophage mannose receptor, a marker of alternative activation [8]. Together, these data indicate that IL17 signaling participates in not only macrophage recruitment in vivo, but also in the ability of macrophages to acquire antigen presentation and alternative activation profiles in vitro.

IL17 induces production of RANTES, IL3, IL9 and IL12p70 from Mφ

To determine cytokines induced by IL17 signaling in macrophages, we interrogated supernatants of IL17-stimulated macrophages by multiplex cytokine array. Primary macrophages were paramagnetically isolated from spleen, peritoneum, peripheral blood, bronchoalveolar lavage, or derived from bone-marrow; and were stimulated in vitro with recombinant IL17A for 24–48 hours. Additionally, peritoneal macrophages were elicited with proteose peptone, and spleen macrophages elicited by CFA administration.

We focused on cytokines reliably induced in macrophages from multiple sources, particularly macrophages from homogenous sources including RAW264.7 cells, and macrophages derived from bone marrow hematopoietic precursors in the presence of either M-CSF or GM-CSF [21]. IL17-induced dose-dependent production of CCL4/MIP1β, CCL5/RANTES, GM-CSF, IL3, and IL9 from several macrophage populations (Figure 6a–e).

Figure 6.

Physiologic responses of primary murine macrophages, in response to IL17A signaling. Production of A) CCL4/MIP1β, B) CCL5/RANTES, C) GM-CSF, D) IL3, and E) IL9 by resident peritoneal (Per), proteose peptone-elicited peritoneal (ePer), naïve resident spleen (Spl), CFA-elicited spleen (eSpl), bronchoalveolar lavage (BAL), peripheral blood (PB), and bone marrow-derived (BMMφ) macrophages, in dose-response to IL17A, determined by multiplex cytokine assay. Data indicate means of triplicate cultures, plus/minus standard error. Asterisks indicate p < 0.05 by two-tailed Students’ t-test, compared to unstimulated control.

Strikingly, production of IL12p70 was the most common macrophage response to IL17A signaling. IL17-induced dose-dependent production of IL12p70 was detected from macrophages from peritoneum, spleen, bronchoalveolar lavage, bone marrow-derived macrophages, and the cell line RAW264.7 (Figure 7a). This effect was specific for IL12; no induction of IL23 was detected (data not shown).

Figure 7.

Physiologic responses of primary murine macrophages, in response to IL17A signaling. A) Production of IL12p70 in dose-response to IL17A from multiple macrophage cell systems. B) IFNγ production by DO11.10 (KJ1-26⁺-gated) CD4⁺ cells responding to 10 μg/mL OVA_323–339-pulsed BMMφ. BMMφ were preconditioned with 10 μg/mL α-mouse IL17A mAb 50104, or derived from Act1-deficient mice. Data indicate means of triplicate cultures, plus/minus standard error. Asterisks indicate p < 0.05 by two-tailed Students’ t-test. Representative of three independent experiments. C) Representative bivariate plots of viable CD4⁺CD3ε⁺-gated cells from B, median well of each condition.

To confirm the specificity of this profile of cytokine and chemokine regulation for IL17 signaling, we repeated these experiments with IL17RA-deficient mice. In the absence of IL17RA, primary murine macrophages expressed substantially diminished CCL4/MIP1β, CCL5/RANTES, GM-CSF, IL3, IL9, and IL12p70 in response to IL17A (Supplemental Figure 5). Several cytokines were induced by IL17 in only a few macrophage populations, including IL1β, TNFα, CXCL1/GROα, and CXCL9/Mig (data not shown). We cannot entirely exclude the possibility that these cytokines were induced in minor contaminating fractions of cultured cells, perhaps including granulocytes, dendritic cells, lymphocytes, or stroma; hence, we have chosen to focus on cytokines induced in multiple populations of macrophages.

IL17-induced production of IL12p70 from Mφ enhances Th1 differentiation

The induction of IL12p70 production by macrophages by IL17 suggested a role in potentiating IFNγ production by CD4⁺ cells. To assess the bioactivity of IL17-elicited IL12p70, we ablated IL17 signaling in BMMφ by blockade of IL17A with monoclonal antibody, or utilizing BMMφ derived from Act1-deficient mice, in which IL17 signal transduction is ablated [12]. These BMMφ were used as antigen-presenting cells for naive TCR transgenic DO11.10 CD4⁺ T cells, which were then interrogated for IFNγ production by intracellular cytokine staining following stimulation with OVA_323–339. We observed diminished IFNγ production in KJ1-26⁺ T cells responding to BMMφ pretreated with αIL17A mAb, or Act1^−/− BMMφ (Figure 7b–c). Similar results were obtained using elicited peritoneal macrophages as APCs (data not shown). No differences were detected in the production of IL17A, indicating this effect was specific for IFNγ (data not shown). As an internal control, antigen-nonspecific KJ1-26⁻CD4⁺CD3ε⁺ T cells in these cultures did not show substantive production of IFNγ under any circumstances (Figure 7c), further reinforcing that the Th1-tropism of IL17-conditioned macrophages is not exerted in a bystander fashion.

Discussion

The participation of macrophages in Th17 responses has been proposed, though little is known about the specifics of their involvement [22]. In addition, phenotypic and functional diversity in the macrophage lineages have only recently begun to be widely appreciated [8, 9, 22]. Importantly, these patterns of macrophage differentiation are regulated in no small part by T cell-derived cytokines: IFNγ, IL4, IL13 and IL10 are all important regulators of macrophage function. We provide here evidence that IL17A elicits macrophage activation in a manner that is not redundant or synonymous with other T cell-derived cytokine patterns.

Thus, IL17 signaling induces the production of GM-CSF, IL3, IL9, and IL12p70 from multiple macrophage-lineage cells, as well as the chemokines CCL4/MIP1β and CCL5/RANTES (Figure 6). Further, the induction of IL12p70 favors the induction of Th1 differentiation in responding transgenic CD4⁺ T cells (Figure 7). We also provide the first evidence that anatomical and physiologic specialization of macrophage-lineage cells regulates their responsivity to IL17 through differential expression of IL17 receptor subunits of macrophages obtained from different anatomical compartments.

The recent structural solution of the IL17 receptor complex offers insight into the IL17 cytokine family; IL17RA appears to be a common, shared signaling subunit, analogous to the common gamma chain [17]. Recruitment of additional IL17 receptor subunits tunes the ligand specificity of the receptor complex [23, 24]. For example, upregulation of IL17RA on CNS microglia in EAE attunes these macrophage-lineage cells to IL17 signaling [25]. In this context, our findings predict that inflammatory regulation of IL17 signaling may involve modular assembly of different IL17 receptor complexes, potentially transducing qualitatively or quantitatively different signals. Although comprehensive understanding of IL17 signaling remains to be elucidated, it seems likely that inflammatory pathways regulate expression of other components of the IL17 signaling pathway, including Act1, TRAF3, TRAF6, or C/EBP [18]. We have provided evidence here that TLR signaling enhances expression of the downstream signaling transducer IκBζ.

The precise signals that induce tissue-specific specialization in macrophage-lineage cells are largely unknown; perhaps the best-studied example is the critical role of TNFSF11/RANK-L signaling in the differentiation of osteoclasts [26]. We have observed that in steady-state non-inflammatory conditions, IL17 receptor expression appears to be highest on inflammatory-type monocytes and macrophages, especially those associated with mucosae. However, inflammatory stimulus, in this case subcutaneous CFA administration, rapidly upregulated IL17RA expression in peripheral blood and liver.

To determine a possible mechanism underlying this inflammation-dependent regulation of IL17 receptor expression, we measured the expression of IL17RA and IL17RC in BMMφ in response to PRR and other inflammatory stimuli [27]. In vitro, we observed the greatest upregulation of IL17 receptor expression in response to stimulation with PGN or CpG oligonucleotide, particularly upon the expression of IL17RC. IFNγ signaling partly counteracted this effect, suggesting that Th1 and Th17-mediated effector responses in macrophages may be mutually inhibitory. In vivo, the mycobacterial component of complete Freund’s adjuvant upregulated IL17 receptors, most notably on macrophages and monocytes in the liver, as well as in peripheral blood.

We further observed recruitment of IL17 receptor-expressing macrophages into cardiac infiltrates in a mouse model of autoimmune myocarditis. Monocytes and macrophages comprise the most numerous component of the inflammatory infiltrate in this CD4⁺ T cell-dependent disease, implicating them as critical effectors of cardiac damage [16, 19]. Macrophage recruitment is perturbed in mice deficient in IL17 signaling, potentially through dysregulation of myelotropic inflammatory chemokines that are known to be IL17-controlled transcriptional targets [7]. However, we observed enhanced recruitment of Ly6C^lo “resident”-type macrophages, which have been reported to express chemokine receptors more consistent with wound repair and homeostasis, rather than inflammatory processes [10]. These chemokines have not been described as IL17-driven transcriptional targets, suggesting that IL17 signaling impacts macrophage infiltration in EAM by mechanisms other than effects on the production of CCR2 ligands. We have further observed that IL17RA^−/− BMMφ acquire an aberrant phenotype upon differentiation in M-CSF, particularly diminished expression of MHC Class II and the macrophage mannose receptor, a marker of alternative activation. We have observed similar defects in GM-CSF-derived BMMφ (data not shown). These data provide evidence that IL17 signaling directly impacts macrophage maturation, activation, and differentiation.

The targets we have identified as IL17-regulated in macrophages may not be canonical C/EBP-driven transcriptional targets characterized in other tissue types [7, 28]. The functional specialization of macrophages may indicate that not only does IL17 regulate unique targets, independent of other T cell-derived cytokines, but it may also induce a transcriptional program that is unique to macrophages. We have undertaken pathfinder microarray experiments, which have indicated that IL17-driven transcription is largely non-redundant with IFNγ- or IL4/IL13-driven transcription in macrophages (data not shown). Canonical IL17-driven targets were largely not detected in these experiments, including CXCL1/GROα (data not shown), underscoring the specialization of macrophage-lineage cells. We have also not observed nitric oxide production, or increased phagocytic uptake in macrophages stimulated with IL17 (data not shown).

It is notable that in these experiments, the capacity of primary tissue macrophages to produce cytokines in response to IL17A in vitro did not necessarily correlate with expression of IL17 receptor subunits in vivo. This may be dismissed as an artifact – isolation, enrichment, and culture of primary macrophages may lead to a more substantial degree of activation than the relatively minimal tissue processing methods used for cytometric measurements of IL17 receptor expression [29, 30]. Several of the cytokines and chemokines reported here were produced by bone marrow-derived macrophages, a relatively homogenous population with the advantage of maintaining relative quiescence in vitro, prior to the application of exogenous stimuli [31, 32], even though BMMφ did not express high levels of IL17 receptor.

Finally, it is likely that IL17 acts as a modulator of inflammatory pathways in macrophage-lineage cells, such as those induced by TLR/IL1 signaling. The known IL17 signaling pathway appears to intersect TLR/IL1 signaling through TRAF3/6, as well as further downstream in the activation of NFκB and C/EBP [18]. TRAF3 signaling has recently been reported to mediated suppression of IL17 signaling in experimental autoimmune encephalomyelitis (EAE) [33]. Therefore, concurrent IL17 signaling may actually limit the induction of inflammatory cytokines upon TLR stimulation. In pilot experiments, we have observed IL17-dependent modulation of TLR-induced production of inflammatory cytokines (data not shown).

We have further demonstrated that the elicitation of IL12p70 production is bioactive and relevant to the potentiation of IFNγ production by responding CD4⁺ T cells. It seems surprising that this component of a Th17 response may elicit a Th1 response, though we may also interpret this finding to mean that macrophage involvement in Th17 responses serves to modulate adaptive CD4⁺ T cell responses by diversification. We must also take into account the finding that CD4⁺ αβ T cells are not necessarily the predominant producer of IL17 in vivo; gamma-delta (γδ) T cells have recently been described to be major producers of IL17 [34–36], as have NKT cells [37]. Mucosal γδ T cells or NKT cells perhaps contribute to the development of subsequent systemic Th1 responses by CD4⁺ T cells, or mixed Th1/Th17 responses through IL17 signaling to macrophages.

Macrophage responses have been demonstrated in response to chitin, in a manner that coupled TLR2/MyD88 to IL17 signaling [38]. IL17A signaling in macrophages has recently been demonstrated to be a requisite intermediary in the development of Th1 defenses to Francisella tularensis, through a mechanism involving IL12p70 and IFNγ [39]. However, in experimental autoimmune encephalomyelitis, Kang and colleagues found that ablation of IL17 signaling specifically in the mononuclear phagocyte lineage had no effect on the development of disease [12]. These disparities may be due to any number of reasons, from simple differences in the pathophysiology of these models, to discrete differences in the coupling of innate recognition pathways to IL17 signaling.

In these experiments, we have not extensively investigated the role of IL17F signaling in the macrophage lineages. IL17F has been shown to be a secondary product of IL17A-producing Th17 CD4⁺ cells [3, 40], and signals through the same receptor complex [41–43]. Detailed binding studies of IL17A, IL17F, and IL17A/F heterodimers to components of the receptor complex have indicated preferential affinities of each of these ligands to various assemblies of the IL17RA/IL17RC complex [17], potentially resolving nonredundant physiologic functions ascribed to IL17A and IL17F in vivo [5, 40].

Together, our findings demonstrate novel interactions between IL17 signaling and inflammation-dependent regulation of macrophage function in vitro and in vivo. We have provided evidence that anatomical specialization may be a critical regulator of the participation of macrophages in Th17 responses, in addition to TLR2/6 and TLR9 signaling. IL17 receptor expressing macrophages were observed tracking with disease in EAM, and ablating the receptor enhanced the recruitment of macrophages to inflamed hearts. We have also shown that IL17 signaling in macrophages elicits a novel cytokine profile, characterized by production of GM-CSF, IL3, IL9, CCL4/MIP1β, CCL5/RANTES and notably, IL12p70. These data indicate that IL17A elicits macrophage activation in a manner that is distinct from other T cell-derived cytokines. Moreover, it demonstrates that the responses of macrophages to IL17 signaling appear to be distinct and specialized, compared to other described IL17-responding cell types.

Abbreviations

Mφ

macrophage