The NF-κB Transcription Factor c-Rel Modulates Group 2 Innate Lymphoid Cell Effector Functions and Drives Allergic Airway Inflammation

Group 2 innate lymphoid cells (ILC2s) play a key role in the initiation and orchestration of early type 2 immune responses. Upon tissue damage, ILC2s are activated by alarmins such as IL-33 and rapidly secrete large amounts of type 2 signature cytokines. ILC2 activation is governed by a network of transcriptional regulators including nuclear factor (NF)-κB family transcription factors. While it is known that activating IL-33 receptor signaling results in downstream NF-κB activation, the underlying molecular mechanisms remain elusive. Here, we found that the NF-κB subunit c-Rel is required to mount effective innate pulmonary type 2 immune responses. IL-33-mediated activation of ILC2s in vitro as well as in vivo was found to induce c-Rel mRNA and protein expression. In addition, we demonstrate that IL-33-mediated activation of ILC2s leads to nuclear translocation of c-Rel in pulmonary ILC2s. Although c-Rel was found to be a critical mediator of innate pulmonary type 2 immune responses, ILC2-intrinsic deficiency of c-Rel did not have an impact on the developmental capacity of ILC2s nor affected homeostatic numbers of lung-resident ILC2s at steady state. Moreover, we demonstrate that ILC2-intrinsic deficiency of c-Rel alters the capacity of ILC2s to upregulate the expression of ICOSL and OX40L, key stimulatory receptors, and the expression of type 2 signature cytokines IL-5, IL-9, IL-13, and granulocyte-macrophage colony-stimulating factor (GM-CSF). Collectively, our data using Rel −/− mice suggest that c-Rel promotes acute ILC2-driven allergic airway inflammation and suggest that c-Rel may contribute to the pathophysiology of ILC2-mediated allergic airway disease. It thereby represents a promising target for the treatment of allergic asthma, and evaluating the effect of established c-Rel inhibitors in this context would be of great clinical interest.


INTRODUCTION
Group 2 innate lymphoid cells (ILC2s) mediate early type 2 immune responses and thereby exert key roles in the initiation and orchestration of anti-helminth immunity as well as allergic inflammation (1)(2)(3)(4)(5). ILC2s exhibit striking functional similarity to adaptive type 2 T helper (Th2) cells. Like Th2 cells, they depend on the transcription factor GATA3 and produce type 2 signature cytokines such as interleukin (IL)-5 and IL-13 upon activation driving eosinophil recruitment as well as goblet cell hyperplasia and mucus production, respectively (6). ILC2s are located at barrier surfaces including the lung and, contrary to Th2 cells, lack the expression of specific antigen receptors (6). They instead become activated in an antigen-independent fashion in response to environmental cues such as alarmins IL-25, IL-33, and/or thymic stromal lymphopoietin (TSLP) that are released upon tissue perturbation (6).
Among these alarmins, IL-33 has been described as the most potent activator of lung ILC2s (7). IL-33 signals through the heterodimeric IL-33 receptor (IL-33R) composed of the ligandbinding chain ST2 and IL-1 receptor accessory protein (IL-1RacP) (8). While it is known that activating IL-33R signaling results in downstream nuclear factor (NF)-kB activation, the underlying molecular mechanisms in ILC2s remain largely elusive (8). NF-kB signaling is mediated by homo-or heterodimers of proteins of the Rel/NF-kB transcription factor superfamily including NF-kB1 (p50), NF-kB2 (p52), RelA (p65), RelB, and c-Rel (9). Despite sharing a common DNA recognition motif, knockout mice lacking individual Rel/NF-kB family members exhibit non-redundant phenotypes (10,11). Moreover, differential expression patterns in tissues and responses to receptor signals as well as target gene specificity indicate that distinct NF-kB subunits exert unique physiological roles (10,11). While stimulation of lung ILC2s with IL-33 results in phosphorylation of RelA and treatment with a pan-NF-kB inhibitor impairs ILC2 effector functions (12)(13)(14)(15), RelB was shown to be an intrinsic repressor of ILC2s (16). On the contrary, the function of c-Rel in ILC2s remains undefined. c-Rel has been shown to promote type 2 immune responses upon ovalbumin (OVA)-induced allergic airway inflammation (17). Furthermore, inhibition of c-Rel in a mouse model of house dust mite-mediated allergic inflammation resulted in reduced levels of IL-13 and airway hyper-reactivity as well as lung inflammation (18,19) and inhibited eosinophil recruitment in an OVA model of chronic asthma (20). Since ILC2s are main drivers of allergic asthma, we aimed to investigate whether c-Rel exhibits similar effects during ILC2driven allergic airway inflammation.
In the present study, we found that deficiency in c-Rel severely diminished early pulmonary ILC2-driven type 2 responses to intranasal IL-33 administration. We further demonstrate that c-Rel expression and activation in ILC2s are positively regulated by IL-33.

Mice
C57BL/6J wild-type (WT) mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME) and bred in house.
Rel −/− mice have been previously described and were kindly provided by Dr. Steve Gerondakis (Monash University) (21). All animals were maintained on a C57BL/6J background and were bred and housed under specific pathogen-free conditions with ad libitum access to food and water. Experiments were conducted with adult female age-matched mice (8-12 weeks) in accordance with the guidelines and policies of the Canadian Council on Animal Care and those of McGill University.

In Vivo Stimulation
Mice were anesthetized with isoflurane followed by intranasal administration of either phosphate-buffered saline (PBS) as a control or 500 ng carrier-free recombinant murine IL-33 (rmIL-33, R&D Systems) per mouse in a total volume of 40 ml. Mice were challenged with PBS or IL-33 for three consecutive days (d0, d1, and d2), and lungs were isolated and analyzed 24 h after the last treatment.

Lung Histopathology
Lungs were inflated with 10% buffered formalin, fixed overnight, and transferred to 70% ethanol. Fixed lungs were embedded in paraffin, sectioned, and stained with hematoxylin and eosin following standard procedures. A histologic disease score from 0 to 4 was attributed based on peribronchial, perivascular, and parenchymal immune cell infiltration.

Preparation of Single-Cell Suspensions From Tissue
Lungs were isolated, finely minced, and digested in Roswell Park Memorial Institute (RPMI)-1640 containing 5% fetal bovine serum (FBS), 0.2 mg/ml of Liberase ™ TM (Roche), and 0.1 mg/ml of DNase I (Roche) at 37°C. Digested lungs were homogenized with a 5-ml syringe attached to an 18G needle, filtered through a 70 mM cell strainer, and washed with PBS. Red blood cells were lysed using Red Blood Cell Lysing Buffer Hybri-Max ™ (Sigma-Aldrich), and cells were washed with fluorescence-activated cell sorting (FACS) buffer (PBS + 2% FBS). Small intestines were isolated, opened longitudinally, cut in small pieces, and vortexed in gut buffer (Hanks' Balanced Salt Solution (HBSS) + 2% FBS + 15 mM of HEPES) to remove fecal matter. Intestinal epithelial cells were removed by incubation in gut buffer containing 5 mM of EDTA for 20 min. Lamina propria was subsequently digested in RPMI-1640 supplemented with 5% FBS, 15 mM of HEPES, 0.1 mg/ml of Liberase ™ TM (Roche), and 0.1 mg/ml of DNase I (Roche) at 37°C for 15 min. Digestion suspensions were filtered through a 70 mM cell strainer to obtain a single-cell suspension, and cells were washed with FACS buffer.

Flow Cytometry
Pelleted single-cell suspensions were resuspended in 2.4G2 hybridoma supernatant dilution and incubated for 15 min on ice to block Fc receptors. Cells were subsequently stained with antibody dilutions prepared in PBS supplemented with 2% FBS for 30 min on ice. Dead cells were excluded by staining with Fixable Viability Dye eFluor ™ 780 (eBioscience) following the manufacturer's instructions. Intracellular staining was performed using the FoxP3/Transcription Factor Staining Buffer Set (eBioscience) according to the manufacturer's protocol. Stained cell suspensions were acquired on a BD FACSCanto ™ II System (BD Biosciences), a BD LSRFortessa ™ Cell Analyzer (BD Biosciences) or sorted on a BD FACSAria ™ III Cell Sorter (BD Biosciences). Flow cytometry data were analyzed using FlowJo X (BD Biosciences). All antibodies used for flow cytometry analyses are listed in Table 1.

Isolation of Lung and Small Intestinal Group 2 Innate Lymphoid Cells
Single-cell suspensions from lung and small intestine were stained with respective surface antibodies and viability dye as described above and sorted on a BD FACSAria ™ III Cell Sorter (BD Biosciences) as live CD45 + Lin − Thy-1 + ST2 + CD25 + or live CD45 + Lin − Thy-1 + CD127 + KLRG1 + cells, respectively. Isolated cells were cultured for 18 h in complete ILC2 medium containing recombinant murine IL-7 (10 ng/ml; R&D Systems), washed, and rested for 4 h in complete ILC2 medium without cytokines before being used in experiments.

Western Blotting Analysis
Bone marrow-derived ILC2s were stimulated for the indicated time points with rmIL-33, and sub-cellular fractionation was performed as previously described (23). Briefly, pelleted cells were resuspended in 900 ml of PBS/0.1% NP-40 containing protease inhibitor and triturated with a micropipette to lyse the cell membranes. Three hundred microliters of lysate was collected (WC=whole-cell lysate). The remaining 600 ml were centrifuged at 13,000 g for 10 s, and supernatant was collected (C=cytoplasmic fraction). Finally, the remaining pellet, containing intact nuclei, was resuspended in 300 ml of PBS/ 0.1% NP-40 with protease inhibitor. All fractionated samples were probe-sonicated for 15 s at 60% amplitude. Protein was quantified using the Bio-Rad Protein Assay (#500-0006), as per manufacturer's instructions. Prior to loading samples were denaturated in 3× Laemmli buffer containing sodium dodecyl sulfate (SDS) and 15% b-mercaptoethanol at 95°C for 5 min. 15 µg of protein for whole-cell lysate and cytoplasmic fraction samples, or 30 ml/one-tenth of the total nuclear fractionated samples, was separated on 8% polyacrylamide gels, with a 3% polyacrylamide stacking gel. Proteins were wet-transferred onto nitrocellulose membranes. For immunoblotting, membranes were probed with anti-c-Rel; anti-GAPDH or anti-H2A (all Cell Signaling Technology) in 0.1% T-TBS + 5% milk powder followed by incubation with anti-rabbit IgG-HRP (Cell Signaling Technology). The details of the antibodies are listed in Table 1.

ImageStream Analysis
Isolated lung ILC2s were left for 4 h in medium without cytokines and left untreated or stimulated with rmIL-33 (10 ng/ml; R&D Systems) for the indicated time points. Cells were fixed and permeabilized using the FoxP3/Transcription factor staining kit (eBioscience) according to the manufacturer's instructions and stained with anti-c-Rel, anti-CD25, and DAPI (NucBlue ™ Fixed Cell ReadyProbes ™ Reagent, Life Technologies). Samples were run on an ImageStreamX Mark II imaging flow cytometer (Amnis), and nuclear translocation of c- Rel in CD25 + c-Rel + DAPI + cells was analyzed using the similarity feature of the IDEAS software (Amnis). All antibodies are listed in Table 1.

RNA Extraction and Quantitative
Real-Time PCR  Table 2.

Intracellular Cytokine Staining
To determine intracellular cytokine production by isolated murine lung ILC2s, sorted cells in complete ILC2 medium (see above) were stimulated in 96-well round-bottom plates (15,000 cells/well) for 48 h with medium only; recombinant murine IL-2, IL-7, IL-9, and IL-33 (10 ng/ml; R&D Systems); or combinations thereof. GolgiPlug (BD Biosciences) was added for the last 6 h of culture.
To analyze cytokine production following in vivo stimulation, lung cell suspensions (1 × 10 6 cells/well in a 96-well plate) in complete ILC2 medium were stimulated with Cell Stimulation Cocktail (eBioscience) according to the manufacturer's instructions in the presence of GolgiPlug (BD Biosciences). Cells were stained with respective surface antibodies and viability dye as described above, and intracellular cytokine staining was performed using the FoxP3/Transcription Factor Staining Buffer Set (eBioscience) according to the manufacturer's protocol. Stained cells were acquired on a BD LSRFortessa ™ Cell Analyzer (BD Biosciences) and analyzed using FlowJo X (BD Biosciences). All antibodies used for flow cytometry analyses are listed in Table 1.

Statistical Analyses
All data were analyzed with GraphPad Prism software (GraphPad Software). p-Values below 0.05 were defined as statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001). Unless otherwise indicated, figures display means ± standard deviation (SD). Experiment sample sizes (n), experiment replicate numbers, and statistical tests used are included in the respective figure legends.

c-Rel Is a Critical Mediator of Innate Type 2 Immune Responses
To investigate the role of c-Rel during ILC2-driven allergic airway inflammation, we challenged WT and c-Rel-deficient (Rel −/− ) mice intranasally with either PBS or IL-33 for three consecutive days and analyzed parameters of airway inflammation ( Figure 1A).
Pulmonary tissue histology showed that, when exposed to IL-33, lack of c-Rel resulted in a significant reduction in perivascular and peribronchial immune cell infiltration into lung tissue as compared with WT mice ( Figures 1B, C). These findings could be recapitulated by flow cytometry, showing markedly decreased numbers of pulmonary CD45 + leukocytes in c-Rel-deficient animals after IL-33 treatment ( Figure 1D). Moreover, while challenge of WT mice with IL-33 resulted in a significant increase in total numbers of type 2 immunity-associated cell populations including eosinophils, type 2 dendritic cells (DC2s), and ILC2s, numbers were significantly diminished in Rel −/− mice ( Figure 1D and Supplementary Figures 1A-D). Furthermore, c-Rel deficiency led to reduced expression of lung Il5 transcripts with a trend towards lower Il13 levels ( Figure 1E). In addition, the type 2-associated chemokines Ccl17 and Ccl22 were greatly reduced in Rel −/− compared with WT mice ( Figure 1F). In summary, compared with WT animals, c-Rel-deficient mice mount a reduced innate type 2 immune response with a significant reduction in pulmonary leukocyte infiltration, eosinophilia, DC2s, and ILC2s. Our results thus suggest that c-Rel is critical for the development of IL-33-driven allergic lung inflammation. To assess whether lack of c-Rel results in ILC2-intrinsic alterations that may explain why Rel −/− animals mount a strongly reduced innate type 2 immune response, we analyzed ILC2 progenitor numbers in the bone marrow of c-Rel sufficient (Rel +/+ ), heterozygous (Rel +/− ), and c-Rel-deficient (Rel −/− ) littermate control animals. We observed that mice of all three genotypes harbored comparable numbers of common lymphoid progenitors (CLPs), as well as the more downstream common helper ILC progenitors (CHILPs), and ILC2 progenitors (ILC2Ps) (Figure 2A and Supplementary Figure 2A). In addition, there was no difference in total numbers of mature lung ILC2s ( Figure 2B). Together, these results indicate that c-Rel is dispensable for the development of ILC2s in the bone marrow and that c-Rel deficiency does not affect homeostatic numbers of lung-resident ILC2s.

IL-33-Driven Ex Vivo Activation of Group 2 Innate Lymphoid Cells Induces c-Rel Expression
Although it is well established that IL-33 signaling via ST2 triggers downstream activation of NF-kB pathways (8), the specific roles of the NF-kB subunit c-Rel in ILC2s remain unknown. We therefore analyzed the direct effect of IL-33 on c-Rel expression and activation in ILC2s. Ex vivo culture of bone marrow-derived (D) Total numbers of CD45 + leukocytes, eosinophils, type 2 conventional dendritic cells (cDC2s), and group 2 innate lymphoid cells (ILC2s) in lungs of WT (gray) or Rel −/− (red) mice were determined by flow cytometric analysis. Expression levels of (E) Il5 and Il13 as well as (F) Ccl17 and Ccl22 in whole lung tissue of WT (gray) and Rel −/− (red) animals were assessed by qRT-PCR. Data are representative of two independent experiments with n = 3-4 mice per group. Data are shown as mean ± SD with *p < 0.05, **p < 0.01, ***p < 0.001 as determined by one-way ANOVA followed by Tukey's multiple comparisons test.
( Figure 3A and Supplementary Figure 3A) as well as lung ILC2s ( Figure 3B and Supplementary Figure 2B) with IL-33 resulted in a rapid increase in c-Rel transcripts levels compared with unstimulated cells. Consistently, lung c-Rel protein levels increased in a dose-dependent manner and remained elevated for at least 72 h of culture with IL-33 ( Figures 3C, D). Furthermore, c-Rel protein expression in lung, small intestine, and bone marrow-derived ILC2s was significantly elevated over levels found in unstimulated cells as early as 4 h following stimulation with IL-33 and was further induced 24 h post stimulation ( Figures 3E, F and Supplementary Figure 3B). Taken together, these findings show that IL-33 potently induces c-Rel expression in ILC2s derived from different anatomical locations, indicating a tissue-spanning role of the IL-33-c-Rel axis.
c-Rel Translocates to the Nucleus Upon Ex Vivo Activation of Group 2 Innate Lymphoid Cells by IL-33 To determine whether c-Rel is activated upon IL-33 stimulation, we assessed nuclear translocation in ILC2s upon ex vivo IL-33 stimulation by Western blotting using cytoplasmic and nuclear ILC2 lysates ( Figure 4A) as well as by ImageStream analysis (Figures 4B, C). With both methods, we observed that c-Rel remained confined to the cytoplasm in unstimulated ILC2s, while activation with IL-33 resulted in translocation of c-Rel from the cytoplasm to the nucleus, indicating a potential role of c-Rel in transcriptional regulation of ILC2 effector functions.

c-Rel Expression Is Induced in Lung Group 2 Innate Lymphoid Cells Following Intranasal Challenge With IL-33
To assess if c-Rel expression is also increased upon in vivo activation of ILC2s, we induced allergic airway inflammation by treating mice intranasally with IL-33 for 3 days. c-Rel-deficient mice that underwent the same treatment were used as negative controls. c-Rel transcript levels were significantly higher in lung tissues of IL-33-challenged mice compared with mice that received PBS, while no transcripts were detected in the lungs of control Rel −/− mice ( Figure 5A). Consistent with ILC2s stimulated ex vivo, intranasal administration of IL-33 in WT mice resulted in a 10-fold induction of c-Rel expression in lung ILC2s over PBS-treated mice ( Figures 5B, C). Taken together, these findings show that c-Rel expression is induced upon ex vivo stimulation of ILC2s as well as during IL-33-and allergen-induced allergic airway inflammation.

c-Rel Deficiency Limits Group 2 Innate Lymphoid Cell Effector Functions Following Ex Vivo Activation
Besides activating signals provided by alarmins, ILC2s require cues from STAT5-activating cytokines such as IL-2, IL-7, or TSLP, as well as costimulatory signals for optimal activation and exertion of effector functions (6,26). We therefore assessed whether observed reduction in Rel −/− ILC2 numbers during allergic airway inflammation may stem from lowered sensitivity to activating signals due to diminished expression of the respective surface receptors. To this end, we assessed surface expression levels of CD25 (IL-2Ra), CD127 (IL-7R), and other key activating ILC2 receptors as well as the ILC2 master transcription factor GATA3 in isolated pulmonary WT and Rel −/− ILC2s at steady state and after ex vivo stimulation with IL-33 in the presence or absence of IL-2 or IL-7. We found that c-Rel deficiency did not result in diminished expression of surface CD25, CD127, or GATA3 by ILC2s at steady state or upon IL-33-mediated activation under the tested conditions (Figures 6A and Supplementary Figure 4A). Moreover, the IL-33R chain ST2 and the costimulatory receptor ICOS ( Figure 6B) were slightly enhanced upon ILC2 activation in c-Rel-deficient ILC2s when compared with WT cells. Besides ICOS, murine ILC2s also express ICOS ligand (ICOSL), whose interaction with ICOS promotes ILC2 survival and effector cytokine production upon IL-33-mediated activation (27,28). Furthermore, constitutive as well as activation-induced expression of other costimulatory ligands and receptors including DR3, GITR, and OX40L on ILC2s have been reported (13,(29)(30)(31)(32). We therefore assessed whether c-Rel regulates the expression of costimulatory ligands and/or receptors that promote ILC2 functions directly as well as are expressed by ILC2s to aid in instructing adaptive immune responses. While no changes in DR3 , or bone-marrow-derived (right) ILC2s quantified as gMFI fold change over unstimulated cells. Asterisks indicate significance over respective untreated controls. Data are representative of at least two independent experiments including two or more biological replicates. Data are shown as mean ± SD with *p < 0.05, **p < 0.01, and ***p < 0.001 as determined by two-tailed t-test (unpaired). gMFI, geometric mean fluorescence intensity. or GITR expression were observed (data not shown), deficiency in c-Rel resulted in significantly reduced expression of surface ICOSL as well as OX40L ( Figure 6C), indicating that c-Rel might further enhance ILC2 functions by driving ICOSL expression and also potentially drive adaptive type 2 responses by positively regulating surface OX40L levels.
Earlier studies demonstrated that c-Rel promotes cell cycle progression and survival in activated lymphocytes (21,(33)(34)(35)(36). We therefore analyzed the proliferative capacity of lung ILC2s at steady state or upon IL-33-driven activation by Ki-67 staining after 48 h of activation. Ki-67 expression levels in lung ILC2s isolated from WT and Rel −/− mice were comparable upon IL-33 stimulation, indicating that c-Rel does not impact early expansion of IL-33-activated ILC2s (Supplementary Figure 4B).
ILC2s mediate their effector mechanisms by rapidly secreting large amounts of type 2 signature cytokines upon activation. To determine whether c-Rel directly regulates lung ILC2 type 2 signature cytokine production, we determined the concentrations of IL-5, IL-13, GM-CSF, and IL-9 in culture supernatants of ILC2s that were left unstimulated or were cultured with IL-2, IL-7, or IL-33 alone, or with indicated combinations (Figures 6D-G). We additionally assessed frequencies of cytokine-producing ILC2 and intracellular cytokine expression levels. All measured type 2 cytokines were induced upon activation of ILC2s with IL-33, and expression was further elevated when ILC2s were stimulated with IL-33 together with IL-2 or IL-7. Importantly, effector cytokine levels in the culture supernatant were significantly reduced in the absence of c-Rel ( Figures 6D-G, left panel), which may be a direct result of significantly lower frequencies of cytokine-producing ILC2s ( Figures 6D-G, middle panel) as well as significantly diminished production of cytokines within those cells ( Figures 6D-G, right panel). Collectively, our data show that c-Rel positively regulates the expression of costimulatory ligands as well as ILC2 effector cytokines, thereby driving innate and potentially also adaptive type 2 immune responses.

c-Rel Drives Type 2 Effector Responses During IL-33-Induced Allergic Airway Inflammation
To further decipher the in vivo impact of c-Rel on ILC2 function, WT and Rel −/− mice were challenged intranasally for three consecutive days with PBS as a control or IL-33 ( Figure 7A). Mice were sacrificed and lung ILC2 surface expression levels of CD25, CD127, and ST2 ( Figure 7B), ICOS as well as ICOSL and OX40L were determined. Additionally, we assessed intracellular expression levels of GATA3 ( Figure 7F) and frequencies of Ki-67 + proliferating lung ILC2s following in vivo challenge. In accordance with our ex vivo stimulation results, no significant differences were observed in the expression of the key stimulatory ILC2 receptors CD25, CD127, and ST2 and ICOS (Figures 7B, C). Importantly, both ICOSL and OX40L were significantly downregulated in the absence of c-Rel, suggesting an in vivo role of c-Rel in the regulation of ILC2 costimulation ( Figures 7D, E). While GATA3 expression was comparable in ILC2s of WT and c-Rel-deficient mice ( Figure 7F), frequencies of proliferating ILC2s were significantly reduced in the absence of c-Rel ( Figure 7G).
To determine the effect of c-Rel on ILC2 effector cytokine production, lung cells from challenged animals were activated with phorbol myristate acetate (PMA)/ionomycin; and intracellular levels of IL-5 ( Figure 7H), IL-13 ( Figure 7I), GM-CSF ( Figure 7J), and IL-9 ( Figure 7K) in ILC2s were analyzed. While production of all cytokines was induced, frequencies of cytokine-expressing ILC2s were significantly reduced in Rel −/− mice following in vivo challenge ( Figures 7H-K, top panel). Furthermore, intracellular cytokine expression levels, measured as gMFI, were markedly lower in the absence of c-Rel. Together, these data show that c-Rel is critical for ILC2 proliferation as well as type 2 effector cytokine production during allergic airway inflammation and might potentially drive subsequent adaptive immune responses by promoting the expression of OX40L.

DISCUSSION
ILC2s are critical mediators of early type 2 immune responses and allergic airway disease. They depend on an intricate network of transcriptional regulators to ensure efficient activation and execution of effector functions. Rel/NF-kB transcription factors modulate immune responses by regulating the expression of hundreds of genes involved in lymphoid cell development, proliferation, survival, and immune cell effector functions (37).
We here observed that the NF-kB transcription factor c-Rel promotes type 2 immunity in an ILC2-driven mouse model of allergic airway inflammation. Challenge of c-Rel-deficient mice resulted in reduced lung immune cell infiltration and diminished numbers of key type 2 immune cell populations including ILC2s, cDC2s, and eosinophils. Consistent with reduced ILC2 numbers, markedly lower transcript and protein levels of the type 2 signature cytokine IL-5 were observed, which is critical for the recruitment of eosinophils to the lung during allergic airway inflammation (38). Moreover, c-Rel deficiency led to lower pulmonary IL-13 transcript and protein levels, a hallmark cytokine of type 2 immunity. Attenuated induction of the type 2 chemokine ligands CCL17 and CCL22 in the lungs of Rel −/− mice following exposure to IL-33 was observed as well, which may be a direct result of reduced numbers of CCL17-and CCL22-producing DC2s (39). Moreover, CCL17 and CCL22 have been described previously to promote ILC2 chemotaxis, and reduced levels might affect positioning of ILC2s within the airways, which could contribute to the observed reduction of type 2 responses (40). In addition, we demonstrate that c-Rel deficiency in ILC2s leads to markedly reduced IL-9 and GM-CSF production in vitro and in vivo. IL-9 has been shown to increase ILC2 fitness (41) and reinforces innate and adaptive type 2 immune responses (42,43). In addition, GM-CSF has been shown to drive type 2 immunity (44) and allergic sensitization (45). This suggests that the attenuated response of c-Rel-deficient mice to IL-33 may be the result of defective activation and cytokine production by ILC2s.
Consistent with earlier reports showing that c-Rel is dispensable for normal hemopoiesis and lymphocyte development, the absence of c-Rel did not result in altered numbers of ILC2 precursors (21). In addition, numbers of lung ILC2s were comparable in naive WT and Rel −/− mice, indicating no homeostatic defects due to c-Rel deficiency that might contribute to the inability of c-Rel-deficient mice to mount an efficient type 2 response upon IL-33 challenge.
Our data show that c-Rel expression is potently induced in a dose-dependent manner in pulmonary ILC2s at both transcript and Data are shown as mean ± SD with *p < 0.05, **p < 0.01, and ***p < 0.001 as determined by two-tailed t-test (unpaired). gMFI, geometric mean fluorescence intensity.
protein levels following activation with IL-33. Upregulation of c-Rel protein levels was also observed in IL-33-stimulated ILC2s isolated from other tissues, indicating a general role of c-Rel in ILC2 function. Importantly, c-Rel expression was induced in lung ILC2s following intranasal challenge with IL-33, suggesting a physiological function during allergic airway inflammation. We furthermore demonstrated nuclear translocation of c-Rel in lung ILC2s upon IL-33 stimulation, and since c-Rel harbors a transcription transactivation domain, it can thereby act as a direct transcriptional activator once in the nucleus (37). Our data suggest that the attenuated response of c-Rel-deficient mice to IL-33 may be the result of defective activation and expansion of ILC2s. IL-33 acts in concert with signals provided by costimulatory STAT5-activating cytokines IL-2 and IL-7, and costimulatory surface molecules to promote ILC2 survival, expansion, and exertion of effector functions (6,26). Previous work demonstrated that c-Rel upregulates the expression of IL-2Ra (CD25) in T cells (21,33,46,47), and it is known that IL-2 enhances ILC2 survival as well as activation-mediated expansion and type 2 cytokine production (48)(49)(50). Moreover, earlier studies demonstrated that c-Rel positively modulates cell cycle progression and the expression of anti-apoptotic proteins in activated T and B cells (21,(33)(34)(35)(36). In addition, genes encoding the type 2 signature cytokines IL-4 and IL-13 were proposed as putative c-Rel target genes based on differential transcript expression observed in gainof-function and loss-of-function experiments in T cells (51). While our work showed no effect of c-Rel on ILC2-intrinsic expression of the cytokine receptors CD25 and CD127, we observed a strong reduction of IL-5, IL-9, IL-13, and GM-CSF production upon IL-33 stimulation in vitro and in vivo. Similar to T cells, ILC2s express activating costimulatory molecules, including DR3, GITR, and ICOS, that promote ILC2 function in disease settings as well as during homeostasis (13, 29-31, 52, 53). While no reduction in ICOS expression was observed, ICOSL and OX40L expression was found to be significantly lower in c-Rel deficient ILC2 after activation. This could explain the markedly reduced ILC2 response in the absence of c-Rel during experimental acute allergic airway inflammation.
Interestingly, WT and c-Rel deficient ILC2s proliferated equally well in vitro. In contrast, the proliferative potential was found to be altered in vivo. Since c-Rel is expressed in a variety of cells within the hematopoietic compartment (10,54), it cannot be excluded that c-Rel deficiency impacts effector functions of other innate as well as adaptive pulmonary immune cell populations such as IL-33reactive macrophages, DCs, or CD4 + T helper cells, which could potentially promote ILC2 activation by providing costimulation and/or activating cytokines. IL-33 was shown to induce IL-2 expression in DCs (55) and IL-2 is known to strongly enhance ILC2 survival as well as activation-mediated expansion and type 2 cytokine production (48)(49)(50). Moreover, interaction of ILC2 with CD4 + T cells has been suggested to enhance the ability of CD4 + T cells to produce IL-2, thereby reinforcing ILC2 activation (49). Interestingly, c-Rel deficiency was shown to alter the ability of CD4 + T cells to produce IL-2 (46,56). It is therefore possible that c-Rel deficiency also alters the capacity of CD4 + T cells to produce IL-2 in response to IL-33, which could contribute to the altered proliferative capacity of ILC2 observed in vivo.
Our data using Rel −/− mice suggest that c-Rel promotes acute ILC2-driven allergic airway inflammation and suggests that c-Rel may contribute to the pathophysiology of ILC2-mediated allergic airway disease. It thereby represents a promising target for the treatment of allergic asthma, and evaluating the effect of established c-Rel inhibitors in this context would be of great clinical interest. In addition, while RelA deficiency results in embryonic lethality and RelA −/− cells exhibit severe defects regarding survival, proliferation, and effector functions (57), mice lacking c-Rel are viable and only show limited immunological defects (21). Thus, targeting c-Rel specifically may avoid the adverse side effects that have halted advancement of pan-NF-kB inhibitors for therapeutic applications (19,58). IL-33-dependent ILC2 activation has also been reported during cancer (59,60), viral infections (22,61,62), and murine models of autoimmune diseases (63)(64)(65)(66); and given the observed induction of c-Rel expression in ILC2s from different tissues, modulation of c-Rel function in these conditions may be worth evaluating.

DATA AVAILABILITY STATEMENT
The datasets presented in this article are not readily available because no datasets were generated as listed in 2. Requests to access the datasets should be directed to jorg.fritz@mcgill.ca.

ETHICS STATEMENT
The animal study was reviewed and approved by the Canadian Council on Animal Care, McGill University, ethics committee.

AUTHOR CONTRIBUTIONS
JF and BM designed the intellectual framework of the study and the layout of the manuscript. BM, SK, and CD performed in vivo allergic airway inflammation models, and BM and SK conducted ex vivo cell culture experiments. BM, SK, CD, and DL analyzed data. MM carried out cellular fractionation and Western blotting analysis. LR performed histological scoring. SV and SG provided critical reagents and mouse strains. BM and JF designed experiments and wrote the manuscript. All authors contributed to the article and approved the submitted version. Multicolor flow cytometry gating strategies to identify murine pulmonary (B) eosinophils (single, live CD45 + SSC hi Siglec-F + CD11c -), (C) cDC2s (single, live CD45 + CD64 -Lin -MHCII + CD26 + CD11c + XCR1-CD172a + ) as well as (D) ILC2s (single, live CD45 + Lin -Thy-1 + CD127 + KLRG1 + ). cDC2, conventional type 2 dendritic cell.
Supplementary ILC2s. Data points are representative of two independent experiments with two biological replicates for each stimulation condition. Data are shown as mean ± SD with **p < 0.01 as determined by two-tailed t test (unpaired). gMFI, geometric mean fluorescence intensity.