Invasive candidiasis has high mortality rates in immunocompromised patients, causing serious health problems. In mouse models, innate immunity protects the host by rapidly mobilizing a variety of resistance and tolerance mechanisms to systemic Candida albicans infection. We have previously demonstrated that exogenous IL-33 regulates multiple steps of innate immunity involving resistance and tolerance processes. In this study, we systematically analyzed the in vivo functions of endogenous IL-33 using Il33−/− mice and in vitro immune cell culture. Tubular epithelial cells mainly secreted IL-33 in response to systemic C. albicans infection. Il33−/− mice showed increased mortality and morbidity, which were due to impaired fungal clearance. IL-33 initiated an innate defense mechanism by costimulating dendritic cells to produce IL-23 after systemic C. albicans infection, which in turn promoted the phagocytosis of neutrophils through secretion of GM-CSF by NK cells. The susceptibility of Il33−/− mice was also associated with increased levels of IL-10, and neutralization of IL-10 resulted in enhanced fungal clearance in Il33−/− mice. However, depletion of IL-10 overrode the effect of IL-33 on fungal clearance. In Il10−/− mouse kidneys, MHC class II+F4/80+ macrophages were massively differentiated after C. albicans infection, and these cells were superior to MHC class II−F4/80+ macrophages that were preferentially differentiated in wild-type mouse kidneys in killing of extracellular hyphal C. albicans. Taken together, our results identify IL-33 as critical early regulator controlling a serial downstream signaling events of innate defense to C. albicans infection.
The innate immune system directs the host protection mechanisms against systemic Candida albicans infection in two main ways: elimination of invading fungi and inhibition of fatal infection-related immunopathology. Ly6G+ neutrophils play a pivotal role in killing fungi in the bloodstream and organs and thereby prevent fungal dissemination and growth (1–3). Mononuclear phagocytes are also indispensable for host defense against systemic C. albicans infection (4–6). For example, Ly6C+ monocytes and CX3CR1+ macrophages cannot only directly phagocytose fungi (4–6), but also help neutrophils eliminate fungi indirectly by mediating tissue inflammation (7, 8). Dysregulation of these inflammatory cells is associated with uncontrolled, lethal tissue inflammation independently of resistance mechanisms (8). Recent studies have well established immune cell–cytokine networks linked to neutrophils’ fungal control activities (8–11). A tissue inflammatory milieu lacking anti-inflammatory mediators is also favorable for effective elimination of fungi. IL-10 is a prototype of cytokines that control both tissue inflammation and antipathogen immunity (12). Although the protective role of IL-10 in candidiasis has been known for long time (13, 14), which factor regulates IL-10 during candidiasis has not been disclosed in detail.
IL-33 is a multifaceted cytokine released rapidly by injury and infection, which subsequently participates in initiation and amplification of immune responses (15). Using an IL-33 infusion model, IL-33 has been shown to play a critical role in many steps of the neutrophil-mediated resistance mechanisms to C. albicans, ranging from recruitment through fungal killing (16). IL-33 also induces polarization of renal macrophages toward an M2 type, which promotes resolution of inflammation caused by C. albicans infection (17). However, a physiological function of endogenous IL-33 remains to be clarified. In this study, we demonstrated that renal CD11b+ dendritic cells (DCs) of Il33−/− mice released lower levels of IL-23, a cytokine that stimulates GM-CSF production by NK cells. As a consequence, reduced levels of GM-CSF in Il33−/− mice resulted in lowering a neutrophils’ ability to inhibit proliferation of C. albicans. By contrast, Il33−/− mice secreted higher levels of IL-10, which played a role in impairing killing of extracellular fungi by neutrophils and macrophages. Our results reveal a hitherto unappreciated host protection mechanism of IL-33 during disseminated candidiasis, which may be of relevance to the understanding of human candidiasis.
Materials and Methods
C57BL/6 mice were purchased from Orient Bio–Charles River Laboratories. Il33−/− and Il10−/− mice with a C57BL/6 background were purchased from The Jackson Laboratory, maintained in a specific pathogen-free facility, and used at 7–10 wk of age. All experiments were conducted according to the regulations of the Animal Committee of the University of Ulsan.
Production of recombinant IL-33 protein
Fungal strains and growth conditions
C. albicans (ATCC26555) was grown in peptone dextrose extract at 30°C overnight, and aliquots were frozen at −80°C. To obtain hyphae, yeasts were suspended in RPMI 1640 medium at a final concentration of 5 × 106 cells/ml, and they were further cultured on a rotary shaker at 37°C for 120 min. To kill C. albicans hyphae, organisms were harvested by centrifugation, pellets were washed twice in sterile PBS, and they were resuspended at a density of 1 × 108 cells/ml before heat killing at 90°C for 30 min.
Experimental systemic candidiasis
C. albicans yeasts were inoculated i.v. into the lateral caudal tail veins with 3 × 105 CFUs (lethal dose) or 1 × 105 CFUs (sublethal dose).
Mice were euthanized, and kidneys were removed aseptically to determine fungal burden. Harvested kidneys were homogenized in 2 ml PBS, serial dilutions of homogenates were plated on Sabouraud agar, and the plates were incubated at 37°C for 24 h. Colonies were counted, and results were expressed as log10(CFUs/organ).
Preparation of kidney cells
Kidneys were perfused, minced, and placed in DMEM (Life Technologies) containing 1 mg/ml collagenase IA and 100 ng/ml DNAse I (Sigma-Aldrich) at 37°C for 30 min. Digested kidney tissues were passed through a 40-µm cell strainer (BD Falcon), and the cell suspension obtained was centrifuged at 300 × g for 10 min. Cells were washed in PBS containing 2% BSA, suspended in 36% Percoll (Amersham Pharmacia Biotech), and gently overlaid onto 72% Percoll. After centrifugation at 900 × g at room temperature for 30 min, cells were retrieved from the Percoll interface and washed twice in DMEM and once with staining buffer (PBS containing 2% BSA and 0.1% sodium azide).
The following FITC-, PE-, PE-Cy5 (PE-cytochrome 5), PerCP-, or allophycocyanin-conjugated mAbs to mouse proteins were purchased from BD Biosciences or eBioscience and used for cell staining: anti-CD45, anti-CD11b, anti-Ly6G, anti-Ly6C, anti–Gr-1, anti-F4/80, anti–MHC class II (MHC II), anti-CD11c, anti–IL-23p19, and anti-rat IgG2a κ-chain. Prepared cells were blocked with 2.4G2 mAb in a staining buffer at 4°C for 20 min. After washing twice with staining buffer, cells were incubated with the relevant mAb at 4°C for 30 min. For intracellular cytokine staining, after staining of surface markers, cells were fixed and permeabilized using Cytofix/Cytoperm and Perm/Wash buffer (BD Biosciences), followed by staining with mAbs to mouse anti–GM-CSF and anti–IL-23p19 (eBioscience). Flow cytometric analysis was performed using an FACSCanto II (BD Biosciences) cytometer, and data were analyzed using FACSDiva software (BD Biosciences) and FlowJo software (Tree Star).
Bone marrow (BM) cells were collected from femurs and tibias, suspended in RPMI 1640 medium (Welgene) supplemented with 10% FBS, penicillin/streptomycin (100 U/ml), 2 mM l-glutamine, and 50 μM 2-ME (Life Technologies). To purify neutrophils from BM cells, anti-Ly6G MACS beads were used according to the manufacturer’s instructions (Miltenyi Biotec). Purified neutrophils were resuspended in RPMI 1640 medium and adjusted to a concentration of 2 × 106 cells/ml. The cells were preincubated with GM-CSF (10 ng/ml; PeptroTech) and/or IL-33 (150 ng/ml) at 37°C for 2 to 3 h. After the addition of either nonopsonized or serum-opsonized FITC-labeled heat-killed (HK) hyphal forms of C. albicans (multiplicity of infection [MOI] of 10), the mixtures were incubated with slow rotation at 37°C for the indicated times. Phagocytosis was stopped by the immediate transfer of cells onto ice, and the cells were washed thoroughly with cold FACS buffer. Extracellular fluorescence was quenched in a quenching solution containing 0.04% trypan blue and 1% formaldehyde, and the cells containing fungi were analyzed using flow cytometry. Phagocytosis was expressed as the percentage of neutrophils phagocytosing FITC-labeled C. albicans.
Kidney cells were harvested at 0, 1, or 3 d postinfection (PI). CD11b+ myeloid cells were enriched by MACS purification with anti–CD11b-PE/anti-PE microbeads (Miltenyi Biotec). Positively selected cells were stained with anti-CD45, anti-Ly6C, anti-Ly6G, anti–MHC II, anti-F4/80, and anti-CD11c mAbs. Cells were sorted using the BD FACSAria fusion cell sorter (BD Biosciences). Sorted cells were used for RNA extraction and killing assays. To purify tubular epithelial cells (TECs), anti-CD45 and anti–epithelial cell adhesion molecule (EpCAM) MACS beads were used according to the manufacturer’s instructions (Miltenyi Biotec). The purified CD45−EpCAM+ TECs reached >92% purity.
Nuclear and cytoplasmic extraction
The nuclear extraction was prepared using an NE-PER Nuclear and Cytoplasmic Extraction Reagent Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Briefly, TECs were washed twice with cold PBS and centrifuged at 500 × g for 5 min. The cell pellet was suspended in 200 μl cytoplasmic extraction reagent I containing proteinase and phosphatase inhibitor cocktail (Thermo Fisher Scientific) by maximum-speed vortexing. The suspension was incubated on ice for 10 min, and 11 μl of a second cytoplasmic extraction reagent II was added. The suspension was vortexed, incubated on ice for 1 min, and centrifuged at 16,000 × g for 5 min. The supernatant fraction (cytoplasmic extract) was transferred immediately to a prechilled tube. The insoluble pellet fraction was resuspended in 100 μl nuclear extraction reagent containing proteinase and phosphatase inhibitor mixture by the highest-speed vortexing for 15 s. The sample was placed on ice and vortexed for 15 s every 10 min for a total of 40 min. After centrifuging at 16,000 × g for 10 min, the supernatant (nuclear extract) fraction was immediately transferred to a clean prechilled tube. Protein concentrations were determined by the BCA Protein Assay Kit (Thermo Fisher Scientific).
Proteins (10 µg) were separated by electrophoresis using 12% SDS-PAGE and transferred onto nitrocellulose membranes (GE Healthcare). Anti–IL-33 (Abcam), anti–α-tubulin (Cell Signaling Technology), and anti–poly(ADP-ribose) polymerase 1 (Cell Signaling Technology) rabbit polyclonal Abs were used at 1:1000 dilution and incubated overnight at 4°C. Anti–α-tubulin and anti–poly(ADP-ribose) polymerase 1 were used as the loading control for cytoplasmic and nuclear extract, respectively. HRP-conjugated anti-rabbit IgG secondary Ab (Abcam) was incubated at room temperature for 1 h. Immunoreactivity was detected using the ECL detection system (Bio-Rad Laboratories).
Measurement of reactive oxygen species
Sorted CD45+CD11b+Ly6G+ neutrophils were seeded in 96-well plates (3 × 105 cells/well), and opsonized C. albicans hyphae (MOI of 10) was added. After a 1-h incubation, 20 μM 2′,7′-dichlorodihydrofluorescein diacetate (Sigma-Aldrich) was added, and the cells were further incubated for 0, 10, 30, and 60 min. After washing with FACS buffer, fluorescence was measured using FACS.
Sorted CD45+CD11b+Ly6G+ neutrophils, CD45+CD11b+Ly6G−Ly6Chi monocytes, and CD45+CD11b+Ly6C−Ly6G− macrophage-enriched cells were incubated with C. albicans hyphae (MOI of 20) for 3 h. CFUs were determined as described previously (18). The percentage of killing (%) was calculated at (1 − [CFUs after incubation/CFUs at the start of incubation]) × 100. In some experiments, 10 μM diphenyleneiodonium chloride (Sigma-Aldrich), a reactive oxygen species (ROS) inhibitor, was added to wells containing neutrophils 30 min before exposure to C. albicans.
Kidneys were fixed in 10% (v/v) formalin, embedded in paraffin, sectioned (5 μm), stained with H&E and periodic acid-Schiff (PAS), and analyzed.
Kidneys were harvested, rapidly embedded in O.C.T. compound (Sakura Finetek), and frozen in liquid nitrogen. Frozen sections (8 μm) were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.25% Triton X-100 for 10 min, and blocked with 2% BSA in PBS for 1 h. Sections were then stained for 1:100 diluted FITC-conjugated anti-CD326 (EpCAM; eBioscience) and anti–IL-33 Abs (Santa Cruz Biotechnology) at room temperature for 2 h and then treated with PE-conjugated donkey anti-rabbit IgG secondary Ab (1:100; Santa Cruz Biotechology) at room temperature for 1 h. All specimens were mounted with Prolong Antifade reagent (Molecular Probes). Slides were examined under a laser-scanning confocal microscope (Olympus).
In vivo administration of cytokines and Abs
IL-33 (1 μg/mouse) was i.p. injected into mice 1 d before infection and 1 d PI. IL-23 (1 μg/mouse; BioLegend) was injected daily starting immediately PI, and GM-CSF (5 μg/mouse) was injected at 24 h and 36 h PI. Neutralizing anti–IL-23 (100 μg/mouse; clone G23-8), anti–GM-CSF (100 μg/mouse; clone MP1-22E9), and anti–IL-10 (100 μg/mouse; clone 20C2) Abs were purchased from BioXCell and injected at different times (anti–IL-23, days 0 and 2; anti–GM-CSF, day 2; and anti–IL-10, immediately before infection). Anti-NK1.1 mAb was purified from PK136 hybridoma and i.p. injected into mice 1 d before and 2 d PI to deplete NK cells (200 μg/mouse).
Total RNA was extracted from kidneys or cultured cells using TRIzol reagent (Invitrogen), according to the manufacturer’s instructions. Whole tissues were homogenized with a TissueLyzer tissue homogenizer (Qiagen), and cDNA was synthesized using SuperScript reverse transcriptase (Invitrogen). Real-time PCR was performed using SYBR Green PCR Master Mix (Qiagen) on an ABI 7500 Fast Real-Time PCR System (Applied Biosystems). The primers used in the experiments were as follows: Il33: 5′-CCTCCCTGAGTACATAACATGACC-3′ (forward) and 5′-GTAGTAGCACCTGGTCTTGCTCTT-3′ (reverse); Csf2: 5′-CGTTCCCCTGGTCAGTGTC-3′ (forward) and 5′-CCGCTGGCCTGGATCTTC-3′ (reverse); Il23a: 5′-CCAGCAGCTCTCTCGGAATC-3′ (forward) and 5′-TCATAGTCCCGCTGGTGC-3′(reverse); Il12b: 5′-CCTGGTTTGCCATCGTTTTG-3′ (forward) and 5′-TCAGAGTCTCGCCTCCTTTGTG-3′(reverse); 18S rRNA: 5′-AGACAAATCGCTCCACCAAC-3′ (forward) and 5′-CTAAACACGGGAAACCTCAC-3′ (reverse); and Il10: 5′-AGGGTTACTTGGGTTGCCAA-3′ (forward) and 5′-CACAGGGGAGAAATCGATGA-3′ (reverse). All PCRs were performed in triplicate and normalized to internal control 18S rRNA mRNA. Relative expression was presented using the 2−ΔΔCT method.
Culture and stimulation of BM-derived dendritic cells
BM cells were collected from Il33−/− mouse femurs and tibias and suspended in RPMI 1640 medium (Welgene) supplemented with 10% FBS, 100 U/ml penicillin/streptomycin, 2 mM l-glutamine (Life Technologies), 50 μM 2-ME, and FLT3 ligand (150 ng/ml; PeptroTech). Cells were cultured in six-well plates at a density of 2 × 106 cells/ml for 7 d. GM-CSF (10 ng/ml) was added at day 7, and cells were cultured for an additional 2 d. Mature BM-derived DCs (BMDCs) were stimulated with a combination of IL-33 (150 ng/ml) and HK C. albicans hyphae (MOI of 10) for 12 h. For adoptive transfer, mature Il33−/− BMDCs were primed with a combination of IL-33 and HK C. albicans for 24 h. Primed or unprimed Il33−/− BMDCs (1 × 106 cells/mouse) were i.v. injected into wild-type (WT) and Il33−/− mice 2 h before infection.
Primary TEC culture
TECs were cultured as previously described (19). In brief, TECs were obtained by digesting kidney with collagenase IA for 30 min. Cells were grown in complete DMEM/Ham’s F-12 (50:50) culture medium (Invitrogen), supplemented with 5% (v/v) FBS, 100 μg/ml streptomycin, 100 U/ml penicillin, hormone mix (5 μl/ml insulin, 5 μl/ml transferrin, 1.25 ng/ml PGE1, and 1.73 ng/ml sodium selenite), and 25 ng/ml epidermal growth factor. Cells were trypsinized before each passage. Three-passage TECs (1 × 106 cells/well) were used for experiments. Cells were incubated in six-well plates for 6 h prior to being challenged by HK C. albicans yeasts (MOI of 10) or hyphae (MOI 10). After culture for 24 h, harvested TECs and culture medium were used for RT-PCR and ELISA, respectively.
Measurement of cytokines and chemokines
Cytokines and chemokines present in total kidney homogenate and cell culture supernatant were measured using a Cytometric Bead Array kit (BD Biosciences) or ELISA kits (eBioscience or R&D Systems), according to the manufacturers’ protocols.
Adoptive transfer of neutrophils
Neutrophils were isolated from BM cells using anti-Ly6G microbeads. Purified neutrophils were resuspended in RPMI 1640 medium (Welgene) and adjusted to a concentration of 2 × 106 cells/ml. The cells were preincubated with a combination of GM-CSF (10 ng/ml) and IL-33 (150 ng/ml) at 37°C for 2 to 3 h. Cells were harvested, washed, resuspended in PBS, and i.v. injected into mice (1 × 107 cells/mouse) 2 h before infection.
All data were analyzed using GraphPad Prism Software version 5. Unpaired Student t test or one- or two-way ANOVA with post hoc analysis was used to compare differences between the groups. The log-rank test was used to analyze survival curves. Error bars represent the SEM. A p value <0.05 was considered statistically significant.
TECs release IL-33 after systemic C. albicans infection
The kidney is a main site for proliferation of C. albicans. We first investigated the kinetics of IL-33 production in the kidney after systemic C. albicans infection. Levels of Il33 transcripts increased from 1 d PI, reached a peak at 3 d PI, and declined thereafter (Fig. 1A). Production of IL-33 protein followed a similar pattern as expression of Il33 mRNA (Fig. 1B). Immunohistochemical and FACS analysis showed that IL-33 was detected mainly in EpCAM+ TECs at 3 d PI (Fig. 1C, 1D). Endothelial cells, mast cells, and inflammatory monocytes barely expressed IL-33 at 3 d PI, but levels of IL-33 were mildly increased in renal macrophages at 3 d PI (Fig. 1E). Isolated TECs also increased transcription and release of IL-33 in response to yeast or hyphal forms of C. albicans (Fig. 1F). To further confirm that TECs release IL-33 at 3 d PI, we investigated the cellular localization of IL-33 in TECs. At 3 d PI, full-length IL-33 was detected in the nucleus of purified TECs, whereas either full-length or cleaved mature IL-33 was localized in the cytoplasm (Fig. 1G). These results suggest that IL-33 is processed in the cytoplasm and released after cleavage.
Il33−/− mice are susceptible to systemic C. albicans infection
To define the role of endogenous IL-33 during systemic candidiasis, Il33−/− mice were challenged with C. albicans and monitored for survival. Il33−/− mice succumbed to a sublethal dose of C. albicans infection rapidly, showing a higher mortality rate than WT mice (Fig. 2A). Consistently, Il33−/− mice experienced more severe body weight loss at 5 d PI (Fig. 2B), and their kidneys appeared more swollen and had more severe edema with a greater weight gain than WT mouse kidneys at 3 and 5 d PI (Fig. 2C, 2D). Gross observations also indicated that Il33−/− kidneys had many more distinguishable nodules than WT kidneys (Fig. 2C). All of these gross observations indicated that fungi proliferated more rapidly and induced more extensive abscess formation in Il33−/− versus WT kidneys. Indeed, Il33−/− mice had markedly increased fungal burden in the kidney at 3 d and 5 d PI compared with WT mice (Fig. 2E). Histopathological analysis showed that Il33−/− kidneys had numerous multifocal areas of abscess formation compared with WT kidneys (Fig. 2F). PAS staining clearly revealed more prominent hyphae within abscesses of Il33−/− versus WT kidneys (Fig. 2F).
IL-33 is crucial for IL-23 production by CD11b+ DCs
DCs play a critical role in clearance of C. albicans through IL-23, which mediates sequential downstream signaling events involving NK cell production of GM-CSF and consequent enhancement of neutrophil’s phagocytic and fungicidal activities (9–11). We examined the involvement of IL-33 in this process. Il33 expression in the kidney was transiently increased at 12 h PI (Fig. 3A). By contrast, expression of Il23a, Il12b, and Csf2 showed a continuous increasing pattern from 12 h PI (Fig. 3B–D). These gene expression patterns suggest that IL-33 may trigger expression of the downstream genes including Il23a, Il12b, and Csf2 during the early phase of C. albicans infection. As IL-23 is produced by CD11b+ DCs, we investigated whether there are changes in cell numbers of renal CD11b+ DCs PI. We did not observe a difference in the absolute number of CD103−CD11b+ DCs in the 3 d PI kidney between Il33−/− and WT mice (Fig. 4A). Analysis for intracellular staining of IL-23p19 showed that there were lower levels of intracellular IL-23p19 in CD11b+ DCs of 3 d PI Il33−/− mouse kidneys (Fig. 4B). In a similar context, injection of IL-33 resulted in no change in the absolute number of CD11b+ DCs (Fig. 4C), but it increased intracellular levels of IL-23p19 (Fig. 4D). Consistent with this, total concentrations of IL-23 in the kidney were significantly lower in Il33−/− mice at 3 d PI, while being significantly greater in IL-33–injected WT and Il33−/− mice (Fig. 4E). IL-33–mediated increase of IL-23 was associated with more effective inhibition of fungal proliferation in either WT or Il33−/− mice (Fig. 4F). Next, we explored whether CD11b+ DCs can produce IL-23 directly in response to IL-33 and/or C. albicans. CD11b+ DCs were isolated from BMDCs and stimulated with a combination of IL-33 and HK C. albicans hyphae for 24 h. We then measured levels of Il12a and Il23a transcripts in cultured BMDCs and protein levels of IL-23 in culture supernatant. Although IL-33 alone barely induced expression of either Il12b or Il23a, HK C. albicans hyphae markedly increased their expression (Fig. 4G). A combination of IL-33 and C. albicans further enhanced expression of Il12b or Il23a (Fig. 4G). Measurement of IL-23 in culture supernatant highlights the potentiating effect of IL-33 on production of IL-23 by CD11b+ DCs (Fig. 4H). To further support the involvement of IL-23 in the IL-33–mediated defense against C. albicans infection, we infused IL-23 into Il33−/− mice before C. albicans infection to see whether IL-23 can lower fungal burden. Indeed, injection of IL-23 significantly reduced fungal burden in Il33−/− mice (Fig. 4I). By contrast, neutralization of IL-23 resulted in increase in fungal burden in WT mice (Fig. 4J). In aggregate, our results indicate that IL-33 acts on CD11b+ DCs to increase their ability to produce IL-23 in response to C. albicans and subsequently inhibit fungal proliferation more effectively.
GM-CSF is indispensable for IL-33–mediated fungal clearance
We next investigated whether IL-33 affects GM-CSF production after systemic C. albicans infection. Il33−/− mouse kidneys contained significantly lower levels of GM-CSF at 3 d PI (Fig. 5A), whereas its renal levels in WT mice were significantly increased by injection of IL-33 (Fig. 5B). Consistent with this, infusion of GM-CSF into Il33−/− mice significantly decreased CFUs of C. albicans in the kidney (Fig. 5C), but neutralization of GM-CSF had an opposite effect on fungal burden in WT mice (Fig. 5D). In addition, adoptive transfer of GM-CSF–primed neutrophils into Il33−/− mice augmented fungal clearance (Fig. 5E). Priming of neutrophils by IL-33 also increased their effector function for fungal clearance, as previously shown (16, 17), but IL-33–primed neutrophils did so less potently than those primed by GM-CSF (Fig. 5E). There was no synergistic effect of IL-33 and GM-CSF on the neutrophil activity for fungal clearance (Fig. 5E). This result seems to indicate that IL-33–mediated increase of GM-CSF is largely responsible for fungal clearance by neutrophils. In support of this hypothesis, our in vitro analysis showed that priming of neutrophils with GM-CSF promoted their phagocytic activity for either opsonized or nonopsonized C. albicans (Fig. 5F), but priming of neutrophils with IL-33 did so for only opsonized fungi (Fig. 5F). There was no synergistic effect of IL-33 and GM-CSF on the phagocytic activity of neutrophils for either opsonized or nonopsonized C. albicans (Fig. 5F). In sum, our results suggest that although IL-33 is able to directly activate neutrophils, IL-33–mediated fungal clearance occurs largely through GM-CSF.
As expected, injection of IL-23 into Il33−/− mice increased the concentration of GM-CSF at 3-d PI kidneys (Fig. 5G), whereas neutralization of IL-23 led to reduced levels of GM-CSF (Fig. 5H). To further investigate whether lower levels of GM-CSF in Il33−/− mouse kidneys are linked to CD11b+ DC production of lower levels of IL-23 during systemic C. albicans infection, Il33−/− mice were adoptively transferred with IL-33 plus HK C. albicans–primed DCs. Primed WT DCs, but not unprimed DCs, increased renal production of GM-CSF equally in either WT or Il33−/− mice (Fig. 5I).
We next investigated whether production of GM-CSF by NK cells is a signaling event downstream of IL-33 signaling. Injection of anti-NK1.1 mAb (PK136) effectively depleted NK1.1+ PBMCs (Fig. 6A) and subsequently abrogated the effect of IL-33 on GM-CSF production and fungal burden in WT mice (Fig. 6B). Taken together with other studies (9–11), these results indicate that NK cells produce GM-CSF after systemic C. albicans infection, and injection of IL-33 further elevated levels of GM-CSF. In addition, IL-33–mediated increment of fungal clearance was significantly abolished when NK cells were depleted (Fig. 6C). In aggregate, our data suggest that the IL-33→IL-23→GM-CSF axis is critical in fungal clearance. It is worthwhile to notice that there were a greater number of neutrophils infiltrating into the kidney of Il33−/− mice at 3 d PI (Fig. 6D), which was associated with more severe renal inflammation (Fig. 6F), Taken together, our results indicate that levels of IL-23 or GM-CSF were not reduced sufficiently to severely impair survival of neutrophils in Il33−/− mice (20).
IL-10 can control fungal clearance in an IL-33–independent way
We detected higher levels of IL-10 in Il33−/− mouse kidneys 1 d PI (Fig. 7A). In WT mice, Ly6C−Ly6G−CD11b+MHC II−F4/80+ macrophages expressed higher levels of Il10 mRNA, whereas its expression levels were very low in Ly6C+Ly6G−CD11b+ monocytes (Fig. 7B). There were medium levels of Il10 mRNA in WT Ly6G+CD11b+ neutrophils and MHC II+F4/80+ macrophages. Deficiency of Il33 resulted in increased transcription of Il10 in Ly6G+CD11b+ neutrophils and MHC II+F4/80+ macrophages (Fig. 7B), but expression of Il10 was mildly reduced in Il10−/− MHC II−F4/80+macrophages (Fig. 7B). As expected, neutralization of IL-10 lowered fungal burden in either WT or Il33−/− mice (Fig. 7C). Unlike WT mice, Il33−/− mice failed to decrease fungal burden in response to injected IL-33 (Fig. 7D). These results seem to suggest that IL-33 no longer affects defense mechanisms in the renal microenvironment created by the absence of IL-10. To test this hypothesis, we first investigated whether a change in the renal microenvironment occurs in Il10−/− mice PI. The kidneys of Il10−/− mice displayed a couple of characteristic features of myeloid cell compositions (Fig. 8A, 8B): 1) they contained significantly reduced percentages of Ly6G+ neutrophils and Ly6C+ monocytes; and 2) a F4/80+ macrophage phenotype was mainly MHC II+ in Il10−/− kidneys, while being MHC II− in WT kidneys. Except for MHC II+F4/80+ macrophages, injection of IL-33 did not induce a change in percentages of myeloid cells either in WT or Il10−/− mice, including monocytes, neutrophils, MHC II− macrophages, and CD11c+F4/80−MHC II+ DCs (Fig. 8A, 8B). Finally, we assayed the killing activities of Ly6C+ monocytes, Ly6G+ neutrophils, and Ly6C−Ly6G− macrophage-enriched myeloid cells of infected WT or Il10−/− mice that received PBS or IL-33. Il10−/− neutrophils and macrophages were superior to WT counterparts in killing extracellular hypha-phase C. albicans, regardless of IL-33 injection (Fig. 8C, 8D). Monocytes were equal in their killing activity in all four groups (Fig. 8E). Treatment of Il10−/− neutrophils with opsonized C. albicans induced higher levels of ROS (Fig. 8E). Il10−/− neutrophils also killed extracellular C. albicans hyphae more effectively (Fig. 8G). However, ROS inhibitor diminished the killing activity of neutrophils of WT and Il10−/− mice to a similar baseline (Fig. 8G), indicating that more active production of ROS is linked to effective fungal killing of Il10−/− neutrophils. Taken together, these results supported the hypothesis that IL-33 did not further augment the antifungal activity of Il10−/− neutrophils and macrophages.
In this study, we characterized the mechanism of IL-33–mediated protection to systemic C. albicans infection. IL-33 is produced from early on PI and takes part in two critical points of resistance processes. First, IL-33 action on CD11b+ DCs is a critical point at which the IL-33→IL-23→GM-CSF resistance axis is initiated. Second, IL-33 seems to suppress expression of Il10, an antifungal cytokine gene, in Ly6G+ neutrophils and MHC II+F4/80+ macrophages. IL-10 is required for IL-33–mediated fungal clearance, as the IL-33→IL-23→GM-CSF resistance axis functions only in an IL-10–replete renal microenvironment. Il10−/− neutrophils and MHC II+F4/80+ macrophages infiltrating into the infected kidney display a superior fungicidal activity that is not affected by IL-33. Considering cell numbers, MHC II+F4/80+ macrophages are likely to be the major effector cell in restricting fungal growth in Il10−/− mice. This interpretation may provide an adequate explanation for why IL-33 cannot co-op the IL-33→IL-23→GM-CSF axis in Il10−/− mice.
The releasing mode of IL-33 from TECs PI is a typical feature of alarmin (21). Early target cells of IL-33 are myeloid cells, including DCs (in this study), monocytes, macrophages, and neutrophils (16, 17). Our results indicate that IL-33 costimulates C. albicans–recognizing cell surface receptors to induce expression of Il23a and Il12b in CD11b+ DCs (Fig. 4G). Although IL-33 has been shown to augment the phagocytic activity of neutrophils and macrophages directly (16, 17), our results from this study indicate that the IL-33→IL-23→GM-CSF axis lies in a central position for defense mechanisms against systemic C. albicans infection. If this axis is impaired, as seen in Il33−/− mice, excess fungal growth results in uncontrolled inflammation and fatal immunopathology (Fig. 6D, 6E).
IL-33 promotes DC maturation. IL-33–activated DCs induce differentiation of Th2 cells (22), although the mechanism behind this is still unknown. They also drive naive CD4+ T cells or regulatory T cells to differentiate into Th17 cells through release of IL-6 and IL-1β (23, 24). However, little has been known regarding the involvement of IL-33–activated DCs in activating cells of the innate arm of immunity. In this study, IL-33 is shown to be required for production of IL-23 by DCs during C. albicans infection. Although IL-23 is a key player for Th17 expansion and maintenance, CD4+ T cells are dispensable for protection from systemic C. albicans infection (25, 26). Instead, the protective role of IL-23 is linked to the augmentation of neutrophils’ phagocytosis through production of GM-CSF by NK cells. Another study has demonstrated that IL-23 protects the host from C. albicans infection by increasing neutrophils’ survival (20). As IL-23 action on myeloid cells occurs in an autocrine but not cell-intrinsic manner (20), and GM-CSF modulates a broad range of neutrophil activities, it is likely that the IL-33→IL-23→GM-CSF axis increases both survival and phagocytosis of neutrophils. However, a greater number of neutrophils infiltrate into the kidney of Il33−/− mice after C. albicans infection and more severe renal inflammation occurs in these mice (Fig. 6D, 6E). This seems to indicate that impaired fungal clearance is linked directly to severe renal inflammation that overcomes impaired survival of neutrophils caused by decreased levels of IL-23 and GM-CSF in Il33−/− mice.
Ly6C+ inflammatory monocytes infiltrating into the kidney are generally believed to be polarized to M1 macrophages during the acute phase of renal inflammation (27). Il10−/− macrophages were mostly MHC II+ in the infected kidney, whereas WT macrophages were MHC II− (Fig. 8B). MHC II+F4/80+ macrophages were superior in killing extracellular C. albicans. As MHC II+F4/80+ macrophages are a major cell population that produces IL-10, they may have an anti-inflammatory feature, as seen in the intestine (28, 29). Taken together, renal macrophages seem to become potent phagocytes in the absence of IL-10 during C. albicans infection, presumably due to loss of negative-feedback regulation for renal inflammation.
IL-33 seems to be unique in immune responses to systemic C. albicans infection, as it simultaneously induces sustained host mechanisms of defense and tolerance. In general, proinflammatory mediators released during acute inflammation results in severe tissue inflammation and repel pathogens. In addition, their negative-feedback regulation for inflammation frequently sacrifices infection clearance function (30, 31). A well-known example is type I IFNs. Similar to IL-33, IFN-β stimulates production of GM-CSF by NK cells through monocyte secretion of IL-15 (8, 32). Thus, signaling via IL-33 and IFN-β is converged on NK cells to exert their protection mechanism to C. albicans infection. In summary, our results provide insight into protective innate immune networks in which IL-33 plays a central role.
We are grateful to Dr. Seo’s and Dr. Kwon’s laboratory members for the help.
This work was supported by National Research Foundation of Korea Grants funded by the Korean government (NRF-2020M33A9D30378911 to B.K. and 2018R1A5A2021242 to S.-K.S.).
Abbreviations used in this article:
bone marrow–derived dendritic cell
- EpCAM, epithelial cell adhesion molecule; HK
- MHC II
MHC class II
multiplicity of infection
reactive oxygen species
tubular epithelial cell
The authors have no financial conflicts of interest.