An asthma-associated IL4R variant exacerbates airway inflammation by promoting conversion of regulatory T cells to T_H17-like cells

Abstract

Mechanisms by which regulatory T (T_reg) cells fail to control inflammation in asthma remain poorly understood. We show that a severe asthma-associated polymorphism in the interleukin-4 receptor alpha chain (IL4RA R576) promotes conversion of induced T_reg (iT_reg) cells towards a T helper 17 (T_H17) cell fate. This skewing is mediated by the recruitment by IL-4Rα-R576 of the growth factor receptor-bound protein 2 (GRB2) adaptor protein, which drives IL-17 expression by activating a pathway involving extracellular signal-regulated kinase, IL-6 and STAT3. T_reg cell-specific deletion of Il6ra or Rorc, but not Il4 or Il13, prevented exacerbated airway inflammation in Il4ra^R576 mice. Furthermore, treatment of Il4ra^R576 mice with a neutralizing anti-IL-6 antibody prevented iT_reg cell reprogramming into T_H17-like cells and protected against severe airway inflammation. These findings identify a novel mechanism for the development of mixed T_H2-T_H17 cell inflammation in genetically prone individuals, and point to interventions that stabilize iT_reg cells as potentially effective therapeutic strategies.

Asthma is characterized by airway hyperresponsiveness (AHR), chronic inflammation and tissue remodeling ¹. Evidence points to reduced frequency and/or suppressive activity of FOXP3⁺ T_reg cells in asthma and atopic conditions ^2-4. Intense inflammation in experimental allergic asthma as well as other conditions, including rheumatoid arthritis and multiple sclerosis, may lessen T_reg cell suppression, or redirect T_reg cells to pro-inflammatory phenotypes ^5-9. However, the molecular basis for impaired T_reg cell function in asthma, and the impact of T_reg cell dysfunction on asthma severity, remain to be fully elucidated.

The cytokines IL-4 and IL-13, acting via two distinct heterodimeric IL-4 receptor complexes, play a pivotal role in asthma pathogenesis. IL-4, which binds to both the IL-4R type I (composed of the IL-4Rα and the common cytokine receptor γc chains) and II (composed of IL-4Rα and IL-13Rα1 chains) directs T_H2 cell differentiation. In contrast, IL-13, which exclusively binds to the IL-4R type II, promtes airway inflammation and remodeling (IL-13) ¹⁰. Increased signaling via IL-4rα and the downstream transcription factor STAT6 exacerbates allergic airway inflammation and promotes food allergy, the latter shown to involve suppressed differentiation of allergen-specific T_reg cells and their re-programming into T_H2 cell-like cells ^9,11.

Polymorphisms in IL-4Rα are associated with atopy and asthma ¹². In particular, a glutamine- (Q) to arginine (R) substitution at position 576 of IL-4Rα (IL-4Rα-Q576R) has been linked to asthma exacerbation and severity ^13-16. The Q576 residue is conserved in the murine IL-4Rα chain. Its replacement by R576 in ll4ra^tm2Tch mice (also known as Il4ra^R576 mice) resulted in augmented allergic airway inflammation, severe eosinophilic infiltration and increased T_H2 responses ^17,18. Q576 lies immediately downstream of a STAT6 binding site at Y575, one of three present in IL-4Rα ¹⁹. Yet no discernable effect of the R576 mutation on STAT6 activation has been noted, indicating that it mobilizes alternative signaling pathways ^17,20. We here show the IL-4Rα-R576 mutation enables recruitment of the adaptor GRB2 to promote airway inflammation by a novel mechanism involving IL-4-directed iT_reg cell differentiation towards the T_H17 cell lineage.

Results

Il4ra^R576 T_reg cell phenotype in allergic airway inflammation

Whereas allergic asthma has classically been associated with a skewed T_H2 cell response, a subgroup of patients with severe disease manifest a mixture of T_H2 and T_H17 cell responses in their airways ^21-24. To determine whether the IL-4Rα-R576 mutation elicits T_H17 airway responses in vivo, we employed a house dust mite (HDM)-induced model of airway inflammation ¹⁸. Il4ra^R576 mice, homozygous for the IL-4Rα-R576 mutation, exhibited severe airway inflammation as compared to control mice homozygous for the wild-type (WT) IL-4Rα-Q576 residue. The Il4ra^R576 mice had increased inflammatory infiltrates and mucus production, accentuated methacholine–induced AHR and robust lung tissue eosinophilia accompanied by neutrophilia, the latter masked in earlier studies by the use of alum as an adjuvant (data not shown) (Fig. 1a–e) ¹⁷. HDM treatment led to increased frequency and number of T_reg cells in lung tissues of WT mice, but this increase was tempered in Il4ra^R576 mice (Fig. 1f). In contrast, HDM-treated Il4ra^R576 mice exhibited substantially increased frequencies and numbers of IL-17, IL-13 and IL-6 expressing CD4⁺Foxp3⁻ conventional T (T_conv) and CD4⁺Foxp3⁺ T_reg cells as compared to WT controls (Fig. 1g–i). Expression of IL-17 and IL-6, but not IL-13, in Il4ra^R576 T_conv and T_reg cells was largely overlapping (Supplementary Fig. 1a,b). Expression of IL-17, IL-13 and IL-5 was increased, and that of IFN-γ was decreased, in the BAL fluid of HDM-treated Il4ra^R576 mice as compared to WT controls (Supplementary Fig. 1c). Furthermore, HDM exposure was associated with increased expression of the T_H17 and T_H2-promoting transcription factors RORγt and GATA3, respectively, in lung T_conv and T_reg cells (Supplementary Fig. 1d–g). The frequencies and numbers of IL-4 producing T_conv and T_reg cells in lung tissues of HDM-treated Il4ra^R576 mice trended to increase as compared to WT controls, while those of IFN-γ cells were unchanged (Supplementary Fig. 1h,i).

Figure 1.

Il4ra^R576 polymorphism promotes enhanced lung inflammation and AHR, associated with increased IL-17, IL-6 and IL-13 expression. (a) Representative PAS staining of lung sections isolated from WT and Il4ra^R576 mice either sham treated with PBS or immunized and challenged with HDM and analyzed 24 hr after the last challenge (20×, scale bar 50 μm) (n = 11 counts per field-of-view). (b,c) Inflammation and PAS scores from the group of mice shown in a. (d) Methacholine–induced AHR for the same groups shown in a (n = 8 mice per group). (e) Absolute numbers of lung tissue eosinophils, neutrophils and lymphocytes, in the respective mouse groups (n = 5 mice for PBS and 7 mice for HDM groups). (f) Flow cytometric analysis, frequencies and absolute numbers of CD4⁺Foxp3⁺ T_reg cells within lung tissue (n = 8 mice per group). (g–i) Flow cytometric analysis of IL-13 (g), IL-17 (h) and IL-6 (i) expression by CD4⁺Foxp3⁻ T_conv or CD4⁺Foxp3⁺ T_reg cells within CD90.2⁺ gated cells (representing all T lymphocytes) in lung tissues of WT and Il4ra^R576 mice treated with PBS or HDM (n = 5 mice for PBS and 7 mice for HDM groups). Results represent means ± s.e.m. from two independent experiments. *P < 0.05, **P < 0.01 and ***P < 0.001 by one-way ANOVA with Bonferroni posttest analysis. For AHR analysis, *P < 0.05 and **P < 0.01 by two-way repeated measures ANOVA.

Expression of the transcription factor Helios differentiates between natural T_reg (nT_reg) cells, which develop in the thymus and are biased towards recognition of self-antigens, from iT_reg cells that arise de novo in the peripheral tissues and are biased towards foreign antigens ²⁵. Analysis of lung tissue T_reg cells revealed decreased Foxp3⁺Helios^low T_reg cells in HDM-treated Il4ra^R576 as compared to WT mice, indicative of existence of a reduced iT_reg cell population (Supplementary Fig. 2a,b). Expression of IL-17 and the T_H17 cell-associated chemokine receptor CCR6 was largely overlapping and highly enriched in T_reg and T_conv cells of HDM-treated Il4ra^R576 as compared to WT mice (Supplementary Fig. 2c,d). In Il4ra^R576 T_reg cells, IL-17 and CCR6 were almost exclusively restricted to the Foxp3⁺Helios^low iT_reg cell population (Supplementary Fig. 2e,f) ^26,27.

Il4ra^R576 iT_reg cells express IL-13, IL-17 and IL-6

TGFβ and T cell receptor (TCR) stimulation converts naïve CD4⁺Foxp3⁻ T cells into Foxp3⁺ T^reg cells ²⁸. Given the increased IL-17 and IL-13 production in lung T_reg cells of allergen-exposed Il4ra^R576 mice, we examined in vitro-generated Il4ra^R576 iT_reg cells for aberrant cytokine expression and functional competency to suppress allergic airway inflammation. WT and Il4ra^R576 naïve CD4⁺DO10.11⁺Rag2^−/−Foxp3^EGFP T cells were purified by cell sorting and differentiated into T_reg cells in culture by stimulation with anti-CD3 and anti-CD28 mAbs and TGF β1. Il4ra^R576 CD4⁺ T cells exhibited decreased conversion to Foxp3⁺ T_reg cells as compared to WT controls, a deficit exacerbated by the addition of IL-4 to the cell cultures (Fig. 2a,b). Importantly, a sizeable fraction of Il4ra^R576, but not WT, CD4⁺Foxp3⁺ iT_reg cells and CD4⁺Foxp3⁻ T_conv cells expressed IL-17, which further increased by IL-4 (Fig. 2c,d). Anti-CD3 and anti-CD28 stimulation of Il4ra^R576 T cells induced elevated expression of IL-13 as compared to WT controls, independent of added TGF β1. IL-13 and IL-17 expression was largely non-overlapping. Addition of IL-4 induced increased IL-17 and IL-13 production in both CD4⁺Foxp3⁺ and CD4⁺Foxp3⁻ populations (Fig. 2e,f). In line with the above studies, expression of RORγt and GATA3 was upregulated in Il4ra^R576 iT_reg cells relative to WT controls (Fig. 2e,f and Supplementary Fig. 3a,b). IL-6 was co-expressed with IL-17 in Il4ra^R576 T_reg cells (Fig. 2g,h); both cytokines were IL-4-dependent, and IL-17 was also IL-6-dependent, consistent with autocrine and paracrine IL-4 production driving IL-6-dependent IL-17 expression (Fig. 2g,h). In contrast, IL-13 and IL-4 were increased in Il4ra^R576 T_reg cells in an IL-4 but not IL-6 dependent manner (Supplementary Fig. 3c–e). IL-4 + TGF β1 treatment induced a modest, equivalent increase in IL-9 expression in WT and Il4ra^R576 iT_reg cells (Supplementary Fig. 3f–h).

Figure 2.

Defective formation and impaired suppressive function of Il4ra^R576 induced-T_reg cells. (a,b) Flow cytometric plot (a) and bar graph (b) demonstrating in vitro generation of iT_reg cells form DO11.10⁺Rag2^−/−Foxp3^EGFP and DO11.10⁺Il4ra^R576Rag2^−/−Foxp3^EGFP naïve CD4⁺ T_conv cells in the presence of anti-CD3 and anti–CD28 mAbs and TGF β1 in the absence or presence of IL-4 for 5 d (n = 6 replicates per group). (c,d) Flow cytometric analysis of IL-17 and IL-13 expression by converted Foxp3⁺ iT_reg cells (c) and CD4⁺Foxp3⁻ T_conv cells (d) in culture. (e,f) Bar graphs demonstrating the frequencies of converted Foxp3⁺ iT_reg and CD4⁺Foxp3⁻ T_conv cells IL-17 and RORγt (e) and IL-13 and GATA3 expression (f) (n = 6 replicates for IL-17 and IL-13 and 6 replicates for RORγt and GATA3 expression). (g) Flow cytometric analysis of dual IL-6 and IL-17 expression by converted iT_reg cells. (h) Bar graph demonstrating the frequencies of double IL-6 and IL-17 expression within converted iT_reg cells (n = 6 replicates per group). Each dot represents one replicate. Data represent means ± s.e.m. from two independent experiments. ***P < 0.001 by one-way ANOVA with Bonferroni posttest analysis.

The cell surface protein neuropillin1 (Nrp1) is highly expressed on nT_reg cells but not iT^reg cells ^29,30. To determine the impact of IL-4 signaling on Il4ra^R576 nT_reg cells, splenic Nrp1^high WT and Il4ra^R576 T_reg cells, which are overwhelmingly nT_reg cells ^29,30, were isolated to high purity by cell sorting (Supplementary Fig. 4a), and were examined for their suppressive capacity in an in vitro T cell proliferation assay. IL-4 treatment did not impact the suppressive function of either WT or Il4ra^R576 nT_reg cells (Supplementary Fig. 4b,c). Treatment with anti-CD3 and anti-CD28 mAbs in the presence of IL-4 induced IL-17, but not IL-13, expression in a minor population of Il4ra^R576 but not WT nT_reg cells that was on average 10 fold less in frequency than that noted for IL-4-treated Il4ra^R576 iT_reg cells (Supplementary Fig. 4d,e). Also, treatment of naïve Il4ra^R576 T_conv cells with IL-4 failed to induce IL-17 expression in the absence of TGF β1 (Supplementary Fig. 4f,g). These results indicated that the capacity of Il4ra^R576 T_reg cells to express IL-17 in response to IL-4 treatment is largely restricted to the iT_reg cell population.

Impaired function of Il4ra^R576 T_reg cells

To assess suppressive capacities of Il4ra^R576 vs. WT iT_reg cells, we employed an adoptive transfer model. iT_reg cells were derived from naïve splenic CD4⁺ T cells of WT and Il4ra^R576Thy1.1⁺DO10.11⁺Rag2^−/−Foxp3^EGFP mice and adoptively transferred to OVA-sensitized Il4ra^R576Thy1.2⁺ mice, which were then challenged with aerosolized OVA and analyzed (Supplementary Fig. 5a). WT iT_reg cells almost completely abrogated OVA–induced tissue inflammation, goblet cell hyperplasia, AHR, eosinophilia neutrophilia and lymphocytosis in lungs of recipient Il4ra^R576 mice. They also suppressed IL-17 and IL-13 expression in the recipient lung T_conv and T_reg cells. In contrast, Il4ra^R576 iT_reg cells failed to do so (Supplementary Fig. 5b–f). Unlike the transferred WT iT_reg cells, up to a third of donor Il4ra^R576 iT_reg cells lost their EGFP expression to become exiT_reg cells that expressed either IL-17 or, less frequently, IL-13, consistent with heightened instability (Supplementary Fig. 5g–j).

We utilized CCR6 as a marker of T_reg cells committed towards the T_H17 cell lineage to examine their functional, epigenetic and transcriptional profiles. CCR6⁺ Il4ra^R576 T_reg cells isolated from OVA-sensitized and challenged mice exhibited decreased methylation of the Foxp3 CNS2 locus, indicative of decreased T_reg cell phenotypic stability (Fig. 3a,b). They also exhibited profoundly decreased suppressive function in an in vitro T cell proliferation assay as compared to CCR6⁻ WT and CCR6⁻ Il4ra^R576 counterparts (Fig. 3c). Transcriptional profiling revealed increased expression in CCR6⁺ Il4ra^R576 T_reg cells of genes associated with a T_H17 cell signature, including Rorc, Ccr6, Il23r, Il17a, Il17f, Il1r1, Nr1d1, Cstl, and Ahr (Fig. 3d and Supplementary Data Set 1) ^26,31-33. To determine whether the T_H17 cell-like T_reg cells in the lungs of allergen treated Il4ra^R576 mice gave rise to Foxp3⁻ T_H17 cells, we employed a lineage tracing approach using a Rosa26 Stop-flox YFP reporter (R26^YFP) crossed with a Foxp3-directed Cre recombinase (Foxp3^EGFPCre) on WT and Il4ra^R576 background. Il4ra^R576Foxp3^EGFPCreR26^YFP and control Foxp3^EGFPCreR26^YFP mice were either sham or OVA sensitized than challenged with aerosolized OVA. CCR6 and cytokine expression was examined in T_reg (EGFP⁺YFP⁺), exT_reg (EGFP⁻YFP⁺) and CD4⁺ T^conv cells (EGFP⁻ YFP⁻). We found markedly increased frequencies of EGFP⁻YFP⁺ exT_reg cells in the lungs of OVA-sensitized and challenged Il4ra^R576Foxp3^EGFPCreR26^YFP relative to controls, indicative of heightened T_reg cell instability (Fig. 3e). Furthermore, about half of those exT_reg cells expressed IL-17 and CCR6, consistent of their differentiation into T_H17 cells. In contrast, expression of IL-13 by exT_reg cells was modest (Fig. 3f,g). Together, IL-17 producing T_reg and exT_reg cells accounted for 35-40% of the total CD4⁺IL-17⁺ and CD4⁺CCR6⁺ T cells in the lungs of OVA-treated Il4ra^R576Foxp3^EGFPCreR26^YFP mice but less than 3% in similarly treated WT controls (Fig. 3h). These results indicated that Il4ra^R576–driven T_reg cell instability and T_H17-cell-like programming in the context of allergic airway inflammation ultimately leads to T_reg cell transformation into T_H17 cells.

Figure 3.

CCR6⁺ IL-17⁺ Il4^R576 T_reg cells exhibit instability and compromised suppressive activity (a) Methylation status of CpG motifs in T_reg cells isolated from lung tissue of OVA–sensitized and challenged mice. Numbers on the left side indicate the position of the respective motifs. (b) Global methylation status of Foxp3 CNS2 in the respective T_reg cell populations (n = 3 mice per group with 7-12 clones per mouse). (c) In vitro suppression of the proliferation of WT responder CD4⁺ T cells (T_eff) by the respective T_reg cell populations (n = 3 replicates per group) (d) Gene expression profiles (volcano plot) of EGFP⁺CCR6⁻ versus EGFP⁺CCR6⁺ T_reg cells isolated by FACS from lung digests of OVA-sensitized and challenged Foxp3^EGFP Il4^R576 mice (n = 3–4 mice). FDR: false discovery rate; Log₂FC: Log₂ fold change. (e) Flow cytometric analysis and frequencies of exT^reg (GFP⁻YFP⁺) cells, plotted as a fraction of exT_reg to total T_reg cells in lung tissue. (f,g) Flow cytometric analysis and frequencies of CCR6 producing (f) and IL-17 and IL-13 producing (g) exT_reg cells in lung tissues. (h) Flow cytometric analysis and frequencies of exT_reg and T_reg cells among CD4⁺IL-17⁺ T_conv cells in lung tissues of the respective mouse groups (n = 6 mice for PBS- and 9 mice for OVA-treated groups for e–h). Data represent means ± s.e.m. from two independent experiments. *P < 0.05, **P < 0.005 and ****P < 0.0001 by one-way ANOVA with Bonferroni posttest analysis. For suppression assay ****P < 0.0001 by repeated measures two-way ANOVA.

Recruitment of GRB2 to IL-4Rα-pY575 activates MAPK

We noted that the R576 substitution rendered the sequence at Y575 (574-GpYREF-578) homologous to a previously reported consensus sequence for high specificity binding of the src homology 2 (SH2) domain of the adaptor protein GRB2 (pY-K/R-N-I/L) ³⁴. Consistent with this prediction, GRB2 and the GRB2-associated binding protein 2 (GAB2) were detected by immunoblotting in IL-4Rα immunoprecipitates derived from IL-4–treated Il4ra^R576 but not WT T cells (Fig. 4a). Furthermore, GRB2 was recovered from mouse splenocyte and human PBMC lysates using biotinylated 15 amino acid murine and human IL-4Ra peptides containing phosphorylated Y575 residue and R576 (pY575R576) but not the Q576 residue (pY575Q576) (Fig. 4b and Supplementary Fig. 6a). Both murine and human pY575R576 peptides demonstrated a modest decrease in STAT6 binding compared to pY575Q576. Together, these data indicated that the R576 substitution altered the specificity of pY575 towards binding GRB2 versus STAT6.

Figure 4.

IL-4Rα-R576 activates GRB2–coupled MAPK. (a) Immunoprecipitates (IP) from medium and IL-4-treated WT and Il4ra^R576 splenocytes using IgG or anti–IL4Rα mAb were probed with anti-GRB2, -GAB2, and -IL-4Rα mAbs. Lys: lysates (b) Binding of GRB2 and STAT6 in splenocyte lysates to biotinylated murine IL-4Rα Y575Q576 (YQ), pY575Q576 (pYQ) and pY575R576 (pYR) peptides. Immunoblotted proteins (upper panel) were quantified by densitometry (lower panel). (c) Immunoblots of phospho- and total ERK1/2 and p38, and of GRB2 in medium or IL-4–treated WT and Il4ra^R576 BMDM transfected with scrambled or GRB2–specific siRNA. (d) RT-PCR analysis of Il6 transcripts in the same groups as c (e) Il6 transcripts in splenocytes treated with medium or IL-4 and the indicated concentrations of MEK-Inh. (f) ChIP analysis of C/EBP-β, NF-κB and AP-1 binding at the Il6 promoter in medium (Un-Stim) or IL-4–treated WT and Il4ra^R576 splenocytes. (g) Flow cytomeric analysis (left) and mean fluorescence intensity (MFI, right) of pSTAT3 in WT and Il4ra^R576 iT_reg cells differentiated with anti–CD3, anti–C28 mAbs and TGF β1 without or with IL-4, anti-IL-4 or anti–IL-6 mAbs (n = 3–6 replicates per group for b–g). (h–j) Frequencies of iT_reg cells (h) IL-17 and IL-13 expressing Foxp3⁺ iT_reg cells (i) and CD4⁺Foxp3⁻ T_conv cells (j) treated with the indicated combinations of cytokines, inhibitors and mAbs (n = 3–6 replicates per group). Data represent means ± s.e.m. from 2–3 independent experiments. *P < 0.05, ***P < 0.001 and ****P < 0.0001 by one-way ANOVA with Bonferroni posttest analysis.

Receptor-bound GRB2 activates the guanine nucleotide exchange factor Son of Sevenless (SOS), which in turn activates the mitogen-activated protein kinase (MAPK) kinase cascade ³⁵. Consistent with GRB2 recruitment and activation, IL-4 stimulation of Il4ra^R576 but not WT splenocytes increased the phosphorylation of MAPKs, including the c-Raf, MKK3/6, ERK1/2 and p38 MAPK (Fig. 4c and Supplementary Fig. 6b,c). Activation of these kinases was inhibited by GRB2-specific but not scrambled small interfering RNA (siRNA) (Fig. 4c). ERK1/2 but not p38 MAPK activation was suppressed by treatment with a c-Raf inhibitor, pointing to two downstream cascades branching out from GRB2, one involving c-Raf–ERK1/2 and the other P38 MAPK and its activator MKK3/6 (Supplementary Fig. 6d). IL-4 activated Il6 transcription in several cell types, including bone marrow derived macrophages (BMDM), splenocytes and purified T cells in a GRB2-ERK1/2 cascade-dependent manner, evidenced by inhibition of Il6 transcripts by Grb2 siRNA and a MAPK kinase (MEK) inhibitor, which inhibits ERK1/2 activation, but not by a P38 MAPK inhibitor (Fig. 4d,e and Supplementary Fig. 6e). IL-4 induced ERK1/2-dependent phosphorylation of NF-κB p65, known transcriptional activators of Il6, and mobilized p65 and C/EBPβ to the Il6 promoter ^36,37 (Fig. 4f and Supplementary Fig. 6f). The activation of NF-κB was inhibited by MEK inhibitor treatment (Supplementary Fig. 6g).

IL-6-dependent STAT3 activation in Il4ra^R576 cells

IL-6 skews iT_reg cell differentiation towards the T_H17 cell lineage in a STAT3-dependent manner ^38,39. Consequent to IL-4Rα-R576-mediated IL-6 production, differentiating Il4ra^R576 iT_reg cells exhibited increased STAT3 activation as compared to WT controls, and this STAT3 activation that was IL-4 and IL-6 dependent (Fig. 4g). Analysis of IL-4 induced STAT3 activation in Il4ra^R576 naïve CD4⁺ T cells revealed delayed phosphorylation of STAT3 that started at 30 min post stimulation and progressively increased over time. It was completely abrogated by treatment with an anti–IL-6 mAb, indicating that IL-4–induced STAT3 phosphorylation proceeded by IL-4-mediated IL-6 production. In contrast, activation by IL-4 of STAT6, and by IL-6 of STAT3, was indistinguishable between WT and Il4ra^R576 T cells (Supplementary Fig. 7a–c).

We next analyzed the role of GRB2-activated MAPK cascades in iT_reg cell skewing towards a T_H17 versus T_H2 cell phenotype. Treatment of differentiating Il4ra^R576 iT_reg cells with MEK inhibitor or anti–IL-6 mAb, but not p38 MAPK inhibitor, rescued the decreased iT_reg cell differentiation, decreased IL-6 levels in culture, and blocked IL-17 expression in differentiating iT_reg cells and in CD4⁺ T_conv cells present in the same co-cultures while sparing IL-13 (Fig. 4h–j and Supplementary Fig. 6h–j). Treatment with p38 MAPK inhibitor blocked IL-13 but not IL-17 expression, consistent with the role of p38 in the upregulation of Il13 gene transcription by a GATA3-dependent mechanism ⁴⁰, while treatment with anti–IL-4 mAb blocked both (Fig. 4i,j and Supplementary Fig. 6i,j). Thus the two branches of MAPK signaling downstream of IL-4Rα–R576 mediate activation of distinct T_H cell cytokine programs.

Whereas T cells express only the type I IL-4R, which responds to IL-4, bone marrow derived macrophages (BMDM) also express the type II receptor, which responds to IL-4 and IL-13. Stimulation of Il4ra^R576 but not WT BMDM with IL-4 or IL-13 induced a similar profile of delayed, sustained STAT3 activation. In contrast, STAT3 was similarly activated in IL-6–treated WT and Il4ra^R576 BMDM (Supplementary Fig. 7d,e).

Expression of transcripts encoding CCL11, which mediates tissue eosinophil recruitment, is markedly upregulated in the airways of allergen-challenged Il4ra^R576 mice ¹⁷. Ccl11 transcripts were greatly increased in IL-4-treated Il4ra^R576 BMDM as compared to WT controls (Supplementary Fig. 7f). Ccl11 transcriptional activation by IL-4/IL-13 is normally mediated by STAT6, while that by IL-17 and IL-9 is mediated by STAT3 ^41-43. Chromatin immunoprecipitation (ChIP) assays revealed that whereas STAT6 bound equally well to the Ccl11 promoter in IL-4-treated WT and Il4ra^R576 BMDM, STAT-3 was selectively recruited to the Ccl11 promoter in Il4ra^R576 BMDM in an IL-6-dependent manner (Supplementary Fig. 7g). Knockdown of STAT3 by siRNA treatment abrogated the increase in Ccl11 transcription in IL-4-treated Il4ra^R576 BMDM, reducing it to levels similar to those of WT controls (Supplementary Fig. 7h,i). Thus dual STAT6-STAT3 activation by IL-4Rα-R576 may promote tissue eosinophilia by super-inducing Ccl11 transcription.

iT_reg cells of humans expressing IL4R^R576 produce IL-17

We examined in vitro iT_reg cell differentiation in prospectively recruited healthy and asthmatic subjects who were either WT (IL4R^Q576/Q576), heterozygous (IL4R^Q576/R576) or homozygous for the mutant allele (IL4R^R576/R576). The gating strategy in these studies is outlined in Figure 5a. In healthy subjects, whose demographics are detailed in Supplementary Table 1, naïve heterozygous and homozygous mutant T_conv cells differentiated to iT_reg cells at lower frequencies as compared to WT T cells, especially in the presence of IL-4 (Fig. 5b). The deficit in iT_reg conversion was rescued by treatment with an anti–IL-6 mAb. The differentiated iT_reg cells and CD4⁺ T_conv cells present in the same co-cultures produced increased amounts of IL-17 as assessed by flow cytometric analysis and by ELISA, which was further upregulated by IL-4 treatment but abrogated by anti–IL-6 mAb (Fig. 5b,d). IL-13 expression was significantly increased in activated T cells of subjects carrying the mutant allele independent of their differentiation into iT_reg cells, mirroring the findings in murine Il4ra^R576 T cells (Fig. 5b,d).

Figure 5.

Naïve CD4⁺ T_conv cells from asthmatic subjects bearing R576 mutation show defective induction of iT_reg cells and their skewing towards a T_H17-like phenotype. (a) Flow cytometric analysis of CD4⁺Foxp3⁺ iT_reg cells induced from purified naïve CD4⁺ T_conv in the presence of anti-CD3 + anti-CD28 mAbs and TGF β1 + IL-4 in the absence or presence of anti-IL-6 mAb. (b,c) Cumulative frequencies of CD4⁺Foxp3⁺ iT_reg cells and IL-17 and IL-13 producing Foxp3⁺ iT_reg cells within converted iT_reg cells of non-asthmatic (n = 3–13) (b) and asthmatic (n = 4–19) (c) subjects with respective genotypes differentiated in the absence or presence of TGF β1 or TGF β1 + IL-4 or TGF β1 + IL-4 + anti-IL-6 mAb. (d) IL-17 and IL-13 cytokine levels in the supernatant of differentiated iT_reg cells in the treated groups mentioned in b (n = 2–5). (e-g) Frequencies of total CD4⁺Foxp3⁺ T_reg cells (n = 5–14) (e) CD4⁺Foxp3⁺Helios^low and Foxp3⁺CCR6⁺Helios^low T_reg cells (n = 4) (f) and frequencies of IL-17 and IL-13 producing Foxp3⁺ T_reg and CD4⁺ T_conv cells in respective groups within PBMC of asthmatic subjects (n = 4–14). Each dot represents one subject. Data represent means ± s.e.m. For subjects see Supplementary Tables 2 and 3. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P <0.001 by one-way ANOVA with Bonferroni posttest analysis.

We also examined the differentiation and cytokine expression of iT_reg cells in prospectively recruited asthmatics as a function of IL4R genotype. The demographics, IL4R genotype and disease severity classification of asthmatic subjects are described in Supplementary Table 2. Whereas there were no differences in age and sex distribution between asthmatics bearing the WT allele versus those heterozygous and homozygous for the mutant allele, the latter subjects exhibited increased asthma severity, in agreement with previously published reports (Supplementary Table 2) ^13-16. Heterozygous and homozygous mutant, but not WT, iT_reg cells of asthmatics also exhibited IL-6–dependent decreased differentiation and increased IL-17 production (Fig. 5c,d). IL-13 expression was upregulated in activated T_conv and in iT_reg cells of asthmatics regardless of their genotype, although ELISA assays revealed its increased production by cells carrying the mutant allele (Fig. 5c,d). Analysis of peripheral blood mononuclear cells (PBMC) revealed a modest but significant decrease in circulating total T_reg cells in asthmatics homozygous for the IL4R^R576 allele. This decrease was restricted to the Foxp3⁺Helios^low (iT_reg) cell population, which also manifested a selective increase in CCR6 expression, echoing the results found in Il4r^R576 mice (Fig.5 e,f; Supplementary Fig. 8). Importantly, the frequency of IL-17–expressing CD4⁺ T_conv and T_reg cells was selectively increased in subjects either heterozygous or homozygous for IL4R^R576. IL-13 expression was modestly increased in CD4⁺ T_conv and T_reg cells of PBMC of subjects homozygous for IL4R^R576 consistent with heightened T_H17 and T_H2 responses (Fig. 5g). These results established that IL-4–driven acquisition by iT_reg cells of a T_H17 cell-like phenotype is an attribute that segregates with the IL4-Rα-R576 mutation.

Il4ra^R576 T_reg cell phenotype change impacts airway inflammation

To determine whether the acquisition by lung Il4ra^R576 T_reg cells of a T_H17 cell-like phenotype contributed to exaggerated airway inflammation, we examined the consequences of T_reg cell-specific deletion of Rorc, which encodes the T_H17-promoting transcription factor RORγt, on airway inflammation ⁴⁴. T_reg cell specific deletion was effected by crossing mice expressing a floxed Rorc allele with mice expressing Foxp3^YFPCre, a Cre recombinase under control of the Foxp3 locus (Supplementary Fig. 9a,b). T_reg cell-specific deletion of Rorc in Il4ra^R576 mice (Il4ra^R576Foxp3^YFPCreRorc^Δ/Δ) decreased airway inflammation, mucus production and AHR upon OVA sensitization and challenge to an extent sufficient to resemble WT controls that were either sufficient or deficient in Rorc (Foxp3^YFPCre and Foxp3^YFPCreRorc^Δ/Δ, respectively), whose aforementioned responses were not significantly different (Fig. 6a–c). It also led to a sharp decrease in lung tissue neutrophilia, lymphocytosis and to a lesser extent eosinophilia. In contrast, total and OVA-specific serum IgE concentrations were marginally decreased (Fig. 6d,e). Rorc deletion rescued the decrease in lung tissue T_reg cells in OVA-sensitized and challenged Il4ra^R576 mice (Supplementary Fig. 9c). It abolished IL-17 expression and normalized CCR6 expression in Il4ra^R576 T_reg cells, and normalized IL-17 and CCR6 expression in CD4⁺ T_conv cells. In contrast, IL-13 expression was unaffected in Il4ra^R576Foxp3^YFPCreRorc^Δ/Δ T_reg cells and marginally so in CD4⁺ T_conv cells (Fig. 6f–i and Supplementary Fig. 9d–h). Virtually identical results were obtained upon Il4ra^R576 T_reg cell-specific deletion of Il6ra, encoding IL-6Rα chain, consistent with the requirement for IL-4-induced IL-6 production in programming T_H17 cell-like T_reg cells (Supplementary Fig. 10). Similarly, treatment with anti–IL-6 mAb suppressed airway inflammation and AHR in HDM-treated Il4ra^R576 mice, abrogated airway neutrophilia and profoundly inhibited the airway eosinophilia. It normalized lung tissue T_reg cell numbers and prevented iT_reg cell reprograming into T_H17-like cells, and it inhibited T_H2 and T_H17 cytokine production and the increased IgE response (Supplementary Fig. 11). In contrast, treatment with an anti–IL-17 mAb was less effective in suppressing HDM-induced airway inflammation in Il4ra^R576 mice. It partially decreased tissue inflammation and AHR, normalized airway neutrophilia and inhibited IL-17 production. However, it only marginally decreased airway eosinophilia. It failed to suppress IgE or T_H2 cytokine production or to rescue reduced T_reg cell numbers in lung tissues (Supplementary Fig. 12). Treatment with an anti-CCL11 mAb, while inhibiting dysregulated airway eosinophilia and CCL11 production in HDM-treated Il4ra^R576 mice, did not suppress tissue inflammation, airway neutrophilia, T_H2 and T_H17 cytokine or IgE production, nor did it rescue ineffective T_reg cell generation in lung tissues (Supplementary Fig. 13). Unlike the case of Rorc and Il6ra, combined T_reg cell-specific deletion of Il4 and Il13 had no significant effect on airway inflammation, AHR, or lung tissue neutrophilia in OVA-sensitized and challenged Il4ra^R576 mice despite decreased total eosinophilia and total and OVA-specific IgE responses. It failed to rescue the decrease in lung tissue T_reg cells in inflamed Il4ra^R576 mice. While it completely suppressed IL-13 expression in T_reg cells, and decreased it in CD4⁺ T_conv cells, it left unaltered IL-17 expression in either cell population (Supplementary Fig. 14).

Figure 6.

T_reg cell lineage-specific deletion of Rorc reverses the aggravated airway inflammation in Il4ra^R576 mice. (a) PAS staining of lung sections isolated from the mouse groups shown and immunized and challenged with OVA (20×, scale bar 50 μm). (b) Inflammation and PAS scores in the mouse groups shown (n = 8–9 counts per field of view). (c) Methacholine–induced AHR in the respective mouse groups (n = 4 mice for PBS- and 6 mice for OVA–treated mouse groups). (d) Total and OVA–specific serum IgE concentrations in the mouse groups shown in b (n = 8). (e) Frequencies of lung tissue eosinophils, neutrophils and lymphocytes (n = 8 mice for PBS- and 9 mice for OVA-treated mouse groups). (f–i) Frequencies of lung tissue Foxp3⁺ T_reg cells (f), and IL-17 (g), CCR6 (h) and IL-13 (i) producing CD4⁺ T^conv and Foxp3⁺ T_reg and CD4⁺ T_conv cells (n = 5–9 mice per group). Each dots represents one animal. Data represent means ± s.e.m. from two to three independent experiments. *P <0.05, **P <0.01, ***P <0.001 and ****P <0.0001 by one-way ANOVA with Bonferroni posttest analysis. For AHR studies, *P < 0.05 and **P < 0.01 by repeated measures two-way ANOVA.

For comparison, we employed Il4ra^F709 mice, homozygous for a Y709F substitution that inactivates the immunotyrosine inhibitory motif of the IL-4Rα chain and consequently augments STAT6 activation ¹¹. This mutation, which models human IL-4Rα polymorphisms that promote STAT6 activation, results in exaggerated allergic airway inflammation due to dysregulated T_H2 immunity ^11,45,46. Il4ra^F709 T_reg cells undergo T_H2 cell-like reprogramming that plays a fundamental role in mediating the exaggerated allergic phenotype ⁹. Unlike Il4ra^R576 mice, T_reg cell-specific deletion of Il4/Il13 in Il4ra^F709 mice normalized their airway inflammation, AHR, and tissue eosinophilia and decreased their total and OVA-specific IgE responses. It reversed T_reg cell T_H2 cell-like reprogramming, evidenced by abolishment of their heightened IL-13 expression, and reduced IL-13 production by CD4⁺ T_conv cells, without affecting the low IL-17 expression in CD4⁺ T_conv and T_reg cells (Supplementary Fig. 15). Thus, the Il4ra^R576 and Il4ra^F709 augment airway inflammation by two distinct mechanisms: one eliciting a T_H17 and the other a T_H2 cell-like reprograming of T_reg cells.

Discussion

Our studies elucidate a novel mechanism for the development of mixed T_H2 and T_H17 cell responses relevant to asthma and other allergic inflammatory diseases. By activating an additional branch of IL-4R signaling initiated by the adaptor protein GRB2, an IL-4Rα polymorphism associated with asthma exacerbations and severity potentiates T_H17 and to a lesser extent T_H2 responses. Importantly, IL-4-induced GRB2 recruitment mediates MAPK-dependent autocrine IL-6 signaling in T cells, which redirects TGF β-dependent iT_reg cell differentiation towards the T_H17 cell lineage. Deletion of Il6ra or Rorc in T_reg cells or treatment with an anti–IL-6 mAb protected against the deleterious effects of IL-4Rα-R576 on airway inflammation, prevented iT_reg cell reprogramming towards the T_H17 lineage and normalized the T_H17 cell response, confirming the centrality of iT_reg cell acquisition of a T_H17 cell-like phenotype in mediating disease exacerbation by this polymorphism.

In mouse models, IL-17 induces steroid-resistant airway inflammation and neutrophilic infiltration ^27,47,48. However, while anti–IL-17 mAb treatment suppressed lung tissue neutrophilia and neutralized IL-17 in the airways of Il4ra^R576 mice, it only partially ameliorated the augmented AHR and tissue inflammation and failed to normalize the T_reg cell response or suppress T_H17/T_H2 polarization. Thus, the efficacy of interventions such as T_reg cell-specific deletion of Rorc or Il6ra or treatment with anti–IL-6 mAb relates not simply to their suppression of IL-17 production but more fundamentally to their stabilization of the iT_reg cell response, resulting in better control of airway inflammation. Studies on Il4ra^R576 and Il4ra^F709 mice revealed two distinct modes of pathogenic iT_reg cell programming relevant to airway inflammation. The former involved GRB2-dependent iT_reg cell programming towards a T_H17-cell like phenotype, while the latter involved STAT6-dependent programming towards a T_H2 cell-like phenotype ⁹. T_reg cell-specific deletion of Il6ra or Rorc in Il4ra^R576 mice and Il4-Il13 in Il4ra^F709 mice reduced airway inflammation in the respective strain back to levels associated with the WT Il4ra allele. Il4ra^R576 and Il4ra^F709 mice model mixed T_H2-T_H17 and T_H2 cell-high asthma inflammatory mechanisms, or endotypes, respectively ^49-51. Both models would indicate that acquisition of iT_reg cells of a T_H17 or a T_H2 cell-like program plays a fundamental role in the pathogenicity of the respective allele in airway inflammation. Accordingly, pro-asthmatic IL-4R polymorphisms may promote distinct asthma endotypes (T_H2 cell-high versus mixed T_H2-T_H17 cell inflammation) by directing iT_reg cells differentiation towards the respective T_H cell program.

Our results indicate that IL-4-directed and IL-6-dependent subversion of IL4R^R576 iT_reg cells into a T_H17 like cells may play a critical role in asthma severity associated with this allele. Blocking the IL-4R or IL-6R pathways may prevent the subversion of allergen-specific iT_reg cells into T_H2 and/or T_H17 like cells, respectively, and provide personalized therapeutic approaches to tolerance re-establishment in asthma ^52,53.

Online Methods

Animals

BALB/cByJ (WT) and all the following strains, except where indicated, were obtained from or rederived at the JAX lab. C.129X1-Il4ra^tm2.1Tch (Il4ra^R576), C.129X1-Il4ra^tm3.1Tch (Il4ra^F709), and C.Cg-Foxp3^tm2Tch/J (Foxp3^EGFP) have been previously described ^11,17,54. NOD/ShiLt-Tg(Foxp3^EGFP-cre)1cJbs/J (Foxp3^EGFPCre) were backcrossed 12 generations on BALB/cBYJ ⁵⁵. C.129P2(Cg)-Il4-Il13^tm1.1Lky (Il4-Il13^fl/fl) mice were crossed with Il4ra^R576 and Il4ra^R576Foxp3^EGFPCre mice ⁵⁶. DO11.10⁺Rag2^−/− mice were obtained from Taconic Farms (Hudson, NY) and were crossed with Thy1.1⁺Foxp3^EGFP and Thy1.1⁺Il4ra^R576 mice to generate Thy1.1⁺DO11.10⁺Rag2^−/−Foxp3^EGFP and Thy1.1⁺Il4ra^R576DO11.10⁺Rag2^−/−Foxp3^EGFP mice. ROR-gtf/fB6(Cg)-Rorc^tm3Litt/J (Rorc^fl/fl), B6;SJL-Il6ra^tm1.1Drew/J (Il6ra^fl/fl), B6.129(Cg) Foxp3tm4(YFP/cre)Ayr/J) (Foxp3^YFPCre) and Il4ra^R576 C57BL/6 congenic were crossed to generate Il4ra^Q576Foxp3^YFPCre, Foxp3^YFPCreRorc^Δ/Δ and Il4ra^R576Foxp3^YFPCreRorc^Δ/Δ. BALB/c Il4ra^R576 mice were crossed to congenic Foxp3^EGFPCreRosa26^YFPCre mice to generate Il4ra^R576 Foxp3^EGFPCreRosa26^{YFPCre 57,58}. Male and female mice, age 5-7 weeks, were employed in experiments in equal proportions. Mice were bred and maintained under specific pathogen-free conditions and used under protocols approved by the Institutional Animal Care and Use Committee at Boston Children's Hospital.

Flow cytometry

Antibodies against the following murine antigens were used for flow cytometric analyses: IL-4 (clone 11B11, 1:300 dilution), Siglec-F (E50-2440, 1:300), Phospho-Stat3 (pY705, 1:50), Phospho-Stat6 (pY641, 1:50) (BD Biosciences), Thy1.2 (30-H12, 1:500), Foxp3 (FJK-16S, 1:300), DO11.10 (KJ1-26, 1:500), IFN-γ (XMG1.2, 1:300), IL-13 (eBio13a, 1:300), GATA3 (TWAJ, 1:200), RORγt (AFKJS-9, 1:200), Helios (22F6, 1:200), CD25 (PC61.5, 1:500), CTLA4 (UC10-4B9, 1:200), CD11c (30-f11, 1:500), CD11b (M1/70, 1:500) (eBioscience), CD4 (RM4-5, 1:500), CD3 (145-2C11, 1:500), IL-17 (TC11-18H10.1, 1:200), IL-6 (MP5-20F3, 1:200), GR-1 (RB6-8C5, 1:500), CCR6 (29-2L17, 1:300) and CD45 (30-F11, 1:300) and CD304 (Neuropilin-1; 3E12, 1:200) (Biolegend). Antibodies against the following human antigens were used: CD4 (RPA-T4, 1:300), Foxp3 (236A/E7, 1:200), IL-13 (JES10-5A2, 1:200), IL4 (MP4-25D2, 1:400) (eBioscience), IL-17A (BL168, 1:200), Helios (22F6, 1:200), CCR6 (G034E3, 1:200) (Biolegend). The specificity and optimal dilution of each antibody was validated by testing on appropriate negative and positive controls or otherwise provided on the manufacturer's website. Intracellular cytokine staining was performed as previously described ⁹. Cells were routinely excluded from the analysis based on the staining of eFluor 780 Fixable Viability Dye (eBioscience). Stained cells were analyzed on a BD LSRFortessa cell analyzer (BD Biosciences) and data were processed using Flowjo (Tree Star Inc.).

Cell purification

Cell suspensions were enriched in CD4⁺ T cells by Magnetic-activated cell sorting (MACS) negative selection (Miltenyi Biotec) and further sorted by a fluorescence-activated cell sorting (FACS) on a FACSAria II (BD Biosciences) based on CD4 and EGFP (Foxp3) expression.

Mouse allergic sensitization and antibody treatment

Mice were sensitized to ovalbumin (OVA) (Grade V; Sigma Aldrich) by intraperitoneal injection of 100 μg of OVA in 100 μL of PBS and then boosted 2 weeks later with a second intraperitoneal injection of OVA in PBS. Control mice were sham sensitized and boosted with PBS alone. Starting on day 29, both OVA- and sham-sensitized mice were challenged with aerosolized OVA at 1% delivered through a Schuco 2000 nebulizer (Allied Health Care Products) for 30 minutes daily for 3 days. Mice were euthanized on day 32 post sensitization and analyzed ¹⁸. For House Dust mite (HDM)– induced allergic airway inflammation, mice received 20 μg of lyophilized Dermatophagoides pteronyssinus (DP) extract in 100 μl of PBS intranasally for 3 days at the start of the protocol and then challenged with the same dose of DP extract on days 15 to 17. Mice were euthanized on day 18 and analyzed for measures of airway inflammation ¹⁸. In some experiments at day 15 to 17 monoclonal anti-IL-6 (MP5-20F3, BioXCell) or anti-IL-17 (17F3, BioXCell) antibodies or the isotype-matched control antibodies (BioXCell) were administered intra-tracheally (100 μg/day), 30 min prior to allergen challenge.

Preparation of cell suspensions from lung tissues

Lungs were removed, minced, and incubated for 45 minutes at 37’C in collagenase (Sigma-Aldrich) (0.5 mM in PBS) and passed through a 40-mm strainer. Single-cell suspensions were resuspended in RPMI medium (Invitrogen)

Measurement of airway tissue and functional responses

Paraffin-embedded lungs sections were stained with periodic acid–Schiff (PAS) as previously described ⁵⁹. Lung inflammation was scored for cellular infiltration around the airways: 0, no infiltrates; 1, few inflammatory cells; 2, a ring of inflammatory cells 1 cell layer deep; 3, a ring of inflammatory cells 2 to 4 cells deep; and 4, a ring of inflammatory cells greater than 4 cells deep ¹¹. The number and distribution of goblet cells was assessed by using PAS staining of mucin granules. Individual airways (bronchi/bronchioles) were scored for goblet cell hyperplasia according to the following scale: 0, no PAS-positive cells; 1, less than 5% PAS-positive cells; 2, 5% to 10% PAS-positive cells; 3, 10% to 25% PAS-positive cells; and 4, greater than 25% PAS-positive cells ¹¹. Allergen-induced airway hyperreactivity (AHR) was measured, as previously described ¹⁸. Anesthetized mice were exposed to doubling concentrations of aerosolized acetyl-β-methacholine (Sigma-Aldrich) by using a Buxco small-animal ventilator (Data Sciences International). The relative peak airway resistance for each methacholine dose, normalized to the saline baseline, was calculated.

In Vitro iT_reg cell differentiation

Sorted naive CD4⁺CD62L⁺Foxp3^EGFP− T cells (1 × 10⁶/ml) were cultured with plate-bound anti-CD28 (5 μg/ml, Biolegend), anti-CD3 (5 μg/ml, Biolegend), recombinant TGF β1 (5 ng/ml, R&D Systems), with or without murine recombinant IL-4 (10 ng/ml) (Perpotech), anti-IL-4 (11B11, 10 μg/ml) (Biolegend) or anti-IL-6 mAb (MP5-20F3, 10 μg/ml) (Biolegend), mitogen activated protein kinase kinase (MEK) inhibitor PD98059 (50 μM, Sigma-Aldrich) or P38 inhibitor IV (10 μM, Sigma-Aldrich). After 4 days, the induced iT_reg cells were analyzed by flow cytometry for Foxp3 expression and intracellular cytokines production and/or re-sorted on the basis of EGFP fluorescence.

Differentiation of Bone marrow derived macrophages (BMDM)

BMDM were generated from bone marrow cells collected from femurs were cultured in complete medium supplemented with 10 ng/mL recombinant murine M-CSF (PeproTech) for 10 days. The medium was changed every other day. Adherent BMDM were harvested at day 10. Purity of BMDM was verified by using flow cytometry and was greater than 90%, as determined by using CD11c and F/480 staining.

Quantitative real-time PCR

RNA was extracted from treated cells, as mentioned in the figure legends, using Quick-RNA MiniPrep kit (Zymo Research) according to the manufacturer protocol. Reverse transcription was performed with the SuperScript III RTPCR system (Invitrogen) and quantitative real-time reverse transcription (RT)-PCR with Taqman® Fast Universal PCR master mix, internal house keeping gene mouse (GAPDH VIC-MGB dye) and specific target gene primers for murine Il6, Ccl11, Rorc or Il4 genes, as indicated (FAM Dye) (Applied Biosystems) on Step-One-Plus machine. Relative expression was normalized to GAPDH and calculated as fold change compared to un-stimulated WT cells.

Immunoblotting and immunoprecipitation

Cells were treated with medium alone or medium with IL-4 (10 ng/ml) for the indicated times, lysed with the lysis buffer (25 mM Tris-HCl (pH 7.5), 150 mM NaCl, 5 mM EDTA (pH 8), 1 mM Na₃VO₄, 0.5% NP-40, 1 × diluted protease inhibitor cocktail (Sigma-Aldrich). Whole cell lysates were resolved by SDS/PAGE and transferred to PVDF membrane (Thermo Scientific) for immunoblotting analysis ⁵⁹. Immunoprecipitates were derived from cell lysates using the indicated antibodies and Protein G superparamagnetic beads (Life Technologies). Following overnight incubation, the immunoprecipitates were eluted from the beads by boiling in gel loading buffer and subjected to immunoblotting as previously described ⁵⁹. The following antibodies were used: GRB2 (clone 4C6-H6, 1:1000 dilution), p-GAB2 (Ser159, EP369Y, 1:1000) p-ERK1/2 (Thr202/Tyr204, D13.14.4E, 1:1000), ERK1/2 (p44/42 MAP Kinase, 137F5, 1:1000), STAT6 (D3H4, 1:1000), STAT3 (D3Z2G, 1:1000) p-c-Raf (Ser338, 56A6, 1:1000), c-Raf (D5X6R, 1:1000), p-P38 (Thr180/Tyr182, D3F9, 1:1000), P38 (D13E1, 1:1000), p-MKK3/6 (Ser189/Ser207, 22A8, 1:1000), MKK3 (D4C3, 1:1000), NF-κB p-p65 (Ser536, 93H1, 1:1000), NF-κB p65 (D14E12, 1:1000) (all obtained from Cell Signaling Technology), C/EBPβ (C-190, 1:1000) (Santa Cruz Biotechnology) and anti-IL-4Rα (CD124) mAb (mIL-4R-M1, 1:1000) (BD Pharmingen). The specificity and optimal dilution of each antibody was validated by testing on appropriate negative and positive controls or otherwise provided on the manufacturer's website.

Chromatin immunoprecipitation (ChIP)

ChIP was performed with Agarose ChIP Kit (Pierce) and antibodies toward STAT3 (D3Z2G, 1:100), STAT6 (D3H4, 1:100), c-Jun (AP-1, 1:100) and NF-κB p65 (D14E12, 1:100) (all obtained from Cell Signaling Technology), C/EBPβ (C-190, 1:100) (Santa Cruz Biotechnology) and the respective isotype control antibodies on 2 × 10⁶ cells starved for 3 hr and treated with medium alone or with IL-4 (10 ng/ml, 2 hrs). The DNA from sample and input was amplified by RT-PCR (40 cycles), using the following primers specific for the Il6 promoter; 5’-AGTGGTGAAGAGACTCAGTG-3’ (forward) and 5’-GGCAGAATGAGCCTCAGA-3’ (reverse) and specific primers for the Ccl11 promoter; 5’-CTTCATGTTGGAGGCTGAAG-3’ (forward) and 5’-GGATCTGGAATCTGGTCAGC-3’ (reverse). Results presented as the ratio of the cycling threshold value of immunoprecipitated DNA to that of input DNA.

siRNA–mediated knockdown of GRB-2

Cells were transfected for 72 hr with scrambled small interfering RNA (siRNA) or Grb2 siRNA I (Cell Signaling Technology) for inhibition of GRB2 expression (50 nM /10⁶ cells). Stat3 siRNA I (Cell Signaling Technology) was used for STAT3 inhibition. Lipofectamine 3000 reagent (Life Technologies) was used to increase the efficiency of transfection (0.5 nmol/L).

IgE and cytokine ELISA assays

Total mouse serum IgE was measured by ELISA using Mouse IgE ELISA Ready-SET-Go kit (eBioscience). OVA-specific IgE was measured by a modified assay such that the plates were coated overnight with 200 μg/ml OVA rather than anti-IgE capture antibody and the rest of the test was similar to the total serum IgE ELISA assay. Mouse IL-4, IL-5, IL-13, IL-17 IFN-γ and Eotaxin cytokines were analyzed in BAL fluids collected from mice using eBioscience ELISA kits, according to the manufacturer's protocol. Human IL-13 and IL-17 were analyzed in culture supernatants of human peripheral blood cells activated with anti-CD3 + anti-CD28 mAbs in the absence or added presence of TGF β or TGF β + IL-4 using eBioscience ELISA kits.

Peptide affinity pull-down experiment

The following biotinylated peptides, corresponding to amino acids 569-583 and 569-584 of the human and mouse IL-4Ra, respectively, and encompassing either WT (Q) or mutant (R) at position 576 were used: Y575Q576: Biotin-SAPTSGYQEFVHAVE (human), Biotin-PAPAGGYQEFVQAVKQ; pY575Q576 (mouse); pY575R576: Biotin-SAPTSGpYQEFVHAVE (human), PAPAGGpYQEFVQAVKQ (mouse); pY575R576: Biotin-SAPTSGpYREFVHAVE (human), PAPAGGpYREFVQAVKQ (mouse). The peptides were incubated with whole splenocyte lysates and avidin-conjugated magnetic beads overnight. The beads were then separated by magnetic sorting and washed. The bound proteins were eluted and analyzed using SDS-PAGE and immunoblotting ^11,60.

Methylation analysis

The methylation status of the Foxp3 T_reg cell–specific demethylation region (CNS2) of WT Foxp3⁺, Il4ra^R576 Foxp3⁺CCR6⁻ and Il4ra^R576 Foxp3⁺CCR6⁺ lung T_reg cells of OVA-sensitized and challenged mice was assessed by bisulfite sequence analysis, as described ⁶¹. The T_reg cell-specific demethylation region of converted DNA was amplified with methylation-specific primer sequences: Foxp3 CNS2, 5’-TATTTTTTTGGGTTTTGGGATATTA-3’ (forward) and Foxp3 CNS2 5’-AACCAACCAACTTCCTACACTATCTAT-3’ (Reverse). The PCR product was subcloned and sequenced. Blast analyses were done by comparison of the resulting sequences with converted Foxp3 sequences.

In vitro suppression assays

Total CD4⁺ T cells were isolated using a CD4 negative isolation kit (Miltenyi Biotec) followed by cell sorting on FACSAria. Isolated cells were labeled with CellTrace Violet Cell Proliferation dye according to the manufacturer's instructions (Life Technologies) and were used as responder cells. T_reg cells were isolated on a FACSAria on the basis of CD4, EGFP and/or CCR6 expression and were used as suppressor cells. Responder cells were co-cultured with T_reg cells, at the indicated ratios, and stimulated for 3 days with 2 μg/ml of soluble anti-CD3 and 5 μg/ml of soluble anti-CD28 in 96-well, round-bottomed plates in triplicates. The cells were then analyzed for CellTrace dye dilution by flow cytometry.

Transcriptome profiling

CCR6⁻ or CCR6⁺ CD4⁺EGFP⁺T_reg cells were isolated by FACS from total lung digests of OVA-sensitized and challenged Il4ra^R576 mice. Total RNA was extracted using TRIzol reagent (Invitrogen) and converted into Double-stranded DNA (dsDNA), using SMART-Seq v4 Ultra Low Input RNA kit (Clontech). dsDNA was then fragmented to 200-300 bp size, using M220 Focused-ultrasonicator (Covaris), and utilized for construction of libraries for Illumina sequencing using KAPA Hyper Prep Kit (Kapa Biosystems). Libraries were then quantified using Qubit dsDNA HS (High Sensitivity) Assay Kit on Agilent High Sensitivity DNA Bioanalyzer.

RNA sequencing data was demultiplexed using perfect matches to indices and quality inspected using FastQC. The sequencing data was aligned to the mm10 build (Gencode annotation) of the mouse genome using STAR ⁶², and counts were quantified using HTSeq ⁶³. Raw counts were filtered for non-mitochondrial protein-coding genes with at least 3 counts in one sample, and were normalized using the DESeq2 package in R ⁶⁴. Pairwise comparisons of differential gene expression were computed using DESeq2.

Human cell studies

Asthmatic subjects and healthy controls were recruited under protocols approved by the Institutional Review Board at Boston Children's Hospital. The demographic details of the respective groups are described in Supplemental Tables 2 and 3. Asthmatic subjects, age range 6-47 years, were identified as having intermittent asthma or persistent mild, moderate to severe asthma following the classification of The National Asthma Education and Prevention Program: Expert Panel Report 3 ⁶⁵. Inclusion Criteria for Asthma included current doctor's diagnosis of asthma, and ≥ 12 % FEV1 bronchodialtor reversibility or airway hyperresponsiveness reflected by a methacholine PC20 ≤16 mg/ml. Healthy control subjects were recruited with no medical history of asthma. Informed consent was obtained from all participants prior to blood collection. Blood samples were used for isolation of DNA using DNeasy Blood & Tissue Kit (Qiagen) for genotyping. Mutation detection was carried out using amplification resistance mutation screen (ARMS) PCR method with the following primers: Common Oligo: 5'-CCAGAGTCCAGACAACCTGACTTGCAAGAGAC-3'; Q576-specific: 5'-TGGGTGCCACCCTGCTCCACCGCATGTACAAACTCAT-3'; R576-specific: 5'-TGGGTGCCACCCTGCTCCACCGCATGTACAAACTCAC-3' ⁶⁶. For iT_reg cell differentiation, peripheral blood mononuclear cells (PBMC) were isolated and naïve CD4⁺CDRA⁺CDRO⁻T cells were derived by magnetic separation (Miltenyi biotech). The Isolated T cells were differentiated into iT_reg differentiation as described for murine study and directly used for intracellular cytokine and FOXP3 expression analysis, using flow cytometry.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 6 software (GraphPad), and employing the statistical tests indicated in the individual figure legends. Samples size was selected based on previous experiments. No samples were excluded. The investigators were blinded as to scoring the lung histopathology in individual experiments. P values of <0.05 were considered significant; n.s.: not significant. All error bars represent s.e.m. as noted in the individual figure legends. Unless otherwise stated, ≥2 independent experiments were carried out for all assays, and the displayed figures are representative of the obtained results. Comparisons between the average severity scores and ages between asthmatics subjects homozygous for IL4RA Q576 versus those heterozygous and homozygous for the IL4RA R576 were carried out using Student's unpaired two-tailed t-test. The sex distribution between those two groups was compared using Fisher's exact test.