Allergic airway recall responses require IL-9 from resident memory CD4+ T cells

Abstract

Asthma is a chronic inflammatory lung disease with intermittent flares predominately mediated through memory T cells. Yet, the identity of long-term memory cells that mediate allergic recall responses is not well defined. In this report, using a mouse model of chronic allergen exposure followed by an allergen-free rest period, we characterized a subpopulation of CD4⁺ T cells that secreted IL-9 as an obligate effector cytokine. IL-9-secreting cells had a resident memory T cell phenotype, and blocking IL-9 during a recall challenge or deleting IL-9 from T cells significantly diminished airway inflammation and airway hyperreactivity. T cells secreted IL-9 in an allergen recall-specific manner, and secretion was amplified by IL-33. Using scRNA-seq and scATAC-seq, we defined the cellular identity of a distinct population of T cells with a proallergic cytokine pattern. Thus, in a recall model of allergic airway inflammation, IL-9 secretion from a multicytokine-producing CD4⁺ T cell population was required for an allergen recall response.

Figure 1.

IL-9-producing CD4+ T cells in a memory allergen recall response. (A) Schematic of the recall allergen chronic/rest/challenge (C/R/C) model. Mice were sensitized intranasally to A.f. 3 times per week for 6-weeks to develop a chronic response. To generate a memory response, mice were rested for either 6 or 12-weeks without additional manipulation. Recall responses were induced with 2 doses of A.f. 24 and 48 hours prior to analysis. (B-C) Flow analysis of isolated lung CD4+ T cells without stimulation (B) and with PMA-Ionomycin stimulation (C). (D) Total lung cell number by hemocytometer and eosinophil and CD4+ T cell numbers per lung determined using flow cytometry (n = 4–5). (E-F) Flow analysis of IL-9 protein from isolated and stimulated lung CD4+ T cells (n = 3–4). (G-L) Mice were treated with FTY720 for 21 days prior to recall challenge (n = 4–5). (G) Flow analysis of peripheral blood CD4+ T cells two days prior to the recall challenge. (H), Total Lung and BAL cell number by hemocytometer, (I) CD4+ T cell number by flow cytometry, and (J) eosinophil cell number by flow cytometry. (K), Flow analysis of isolated stimulated lung CD4+ T cells. (L) IL-9+ CD4+ T cell numbers per lung. (M) Flow analysis of stimulated lung ILC2 and CD4+ T cells (n = 4). (N) Distribution of lung IL-9-reporter INFER+ cells after C/R/C (n = 3). (O-P) Quantification of INFER+CD4+CD90.2- cells by immunofluorescence of intact lungs from the recall allergen model. Data are presented as mean +/− SEM are representative of 2–3 independent experiments. Student’s unpaired two-tailed t test was used for comparison to generate p values in B, F, G, H, J, K, I, L, M and O. One-way ANOVA with a post hoc Tukey test was used to generate p values in N. *, p<0.05, **, p<0.01, ***, p<0.001, ****p<0.0001.

Figure 2.

IL-9 is a memory cytokine in a recall challenge time course. (A) Schematic of the recall allergen time course model. Cells from the chronic/rest/challenge (C/R/C) model described in Fig. 1A were taken for analysis at the indicated time points. (B) Total lung cell numbers by hemocytometer throughout the time course (n = 3–4). (C) Lung granulocyte frequency by flow cytometry from hematopoietic CD45+ cells (n = 3–4). (D-E) Flow analysis of isolated lung CD4+ T cells (D) stimulated with PMA/Ionomycin or (E) unstimulated (n = 3–4). (F) IL-9 protein from ex vivo cytokine assay with A.f. The 50,000 purified CD4+ T cells were cocultured with an equal number of CD11c+ antigen presenting cells and A.f. extract for 72 hours. Supernatants were analyzed for IL-9 by ELISA. (n = 3–4). (G-I) mRNA expression of (G-H) cytokine and (I) transcription factors in unstimulated CD4+ T cells isolated from lung(n = 3). Data are presented as mean +/− SEM from a representative experiment from 2 independent experiments. One-way ANOVA with a post hoc Tukey test was used to generate p values for all comparisons. *, p<0.05, **, p<0.01, ***, p<0.001, ****p<0.0001.

Figure 3.

IL-9-secretion is antigen-specific and enhanced in the presence of IL-33. (A) Flow analysis of isolated lung CD4+ T cells 8hrs after recall challenge (n = 3). (B) mRNA expression in unstimulated lung CD4+ T cells (n = 6–8). (C) Flow staining of isolated lung CD4+ T cells stimulated with PMA/Ionomycin. (D) mRNA expression in flow sorted lung CD4+ T cells at the chronic time point (n = 3). (E) IL-33 protein levels in the BAL fluid determined using ELISA (n = 3). (F) IL-9 protein from T cells stimulated ex vivo with A.f. (left) or without A.f (right) (n = 4). (G-J) Mice were treated with either A.f. or IL-33 during the recall challenge (n = 4–5). (G) Total lung cell cell number by hemocytometer and (H) CD4+ cell numbers using flow cytometry. (I) Frequency and number per lung of IL-9+ from CD4+ using flow cytometry. (J) IL-9 protein from T cells stimulated ex vivo with allergen extract. (K-O) The recall allergen model was used to compare wild type and Il1rl1-deficient mice (n = 4–5). (K) Lung CD4+ T cell, (L) ILC2 cell and Treg cell numbers by flow cytometry. (M-N) Flow analysis of isolated CD4+ lung cells stimulated with PMA/Ionomycin. (O) IL-9 protein from ex vivo cytokine assay. (P) IL-9 protein from ex vivo T cell stimulation with either BSA or A.f (n = 3–4). (Q) IL-9 protein from ex vivo T cell stimulation with A.f., A.a. or HDM allergen (n = 3–4). (R-V), Mice were either treated i.v. with anti-MHC-II or isotype control antibody prior to and during the recall response or examined without recall challenge (PBS treated) (n = 4–5). (R) Total lung cell, mediastinal lymph node and (S) lung CD4+ T cell number. (T), Flow analysis of isolated lung CD4+ T cells stimulated with PMA/Ionomycin. (U) IL-9+CD4+ T cell number. Data are presented as mean +/− SEM and are representative of 2 experiments. Student’s unpaired two-tailed t test was used for comparison to generate p values in A, B, D (INFER +/− comparison), E, G, H, I, J, K, L, N, and P. One-way ANOVA with a post hoc Tukey test were used to generate p values in F (groups with same innate treatment), J, O, Q, R, S, T, and U. Two-way ANOVA with a post hoc Tukey test were used to generate p values in D, and F. *, p<0.05, **, p<0.01, ***, p<0.001, **** and ####, p<0.0001. In B, D, and F * indicates a comparison between time points or conditions and # designates a comparison between innate cytokine or receptors.

Figure 4.

IL-9-secreting CD4+ Trm cells represent a unique differentiated population. (A-B) Flow cytometry analysis of isolated lung CD4+ T cells following recall allergen and stimulated with PMA/Ionomycin in vitro. (B) Percentages of multi-cytokine positive cells in the IL-9+ population. (C-H) scRNA-seq analysis of lung CD4+ T cells from A.f. recall allergen challenged mice, 5 mice were pooled per group before FACS sorting of ST2+/−CD4+. (C) Unsupervised UMAP with clustering based on gene expression. (D) Log-fold gene expression over UMAP of the indicated genes. (E) Pseudotime trajectory analysis over naïve (right circle) and activated T cell populations (left circle). (F) Plot of Log-fold expression of the indicated genes as a function of pseudotime. (G) Plot of unsupervised UMAP clustering of Th2, Th9rm, and Th9 cell populations. (H) Violin plots of Log-fold expression of the indicated genes across naïve, Th2, Th9 and Th9rm cell populations.

Figure 5.

Chromatin structure distinguishes Th2 and Th9rm populations in the memory recall model. scACTC-seq analysis for lung CD4+ T cells from A.f. recall allergen challenged mice, 5 mice were pooled per group before FACS sorting of ST2+CD4+. (A) UMAP of ST2+ CD4+ T cells clusters derived from integrative analysis of single-cell RNA and ATAC sequencing, showing Th9rm, Th2 and T-regulatory cell populations. (B-E) ATAC-seq profiles with peaks indicating chromatin accessibility at each loci generated by using CoveragePlot function in Seurat. (F) Differentially enriched motifs for transcription factors in Th9rm and Th2 populations determined by performing ChromVar analysis. (G)Gene activity determined by using Findmarkers function in Seurat based on open promoter regions and + 2kb upstream of the promoter of transcription factors associated with IL-9-secreting CD4+ T cells in tissue resident Th9rm and Th2 cells.

Figure 6.

Anti-IL-9 at the recall challenge phase blocks cell expansion and inflammation (A-G) Mice were either intranasally treated with IL-9 neutralizing or isotype control antibody prior to and during the recall allergen response (n = 5–7). (A) Schematic of the recall antigen antibody blockade model. Similar to the recall allergen chronic/rest/challenge (C/R/C) model described in Fig. 1A, in the recall response mice were intranasally treated with IL-9 neutralizing or isotype control antibody prior to and during the recall allergen response. (B) Total lung cell numbers by hemocytometer. (C) Paraffin-embedded lung lobe sections were stained with Masson’s Trichome. (D) Airway resistance 24 hours after the last intranasal A.f. treatment was measured following intubation and intratracheal challenge with increasing doses of methacholine. Student’s unpaired two-tailed t test comparing isotype vs naïve represented by * and anti-IL-9 vs isotype by ^#. (E) Total lung and BAL cell numbers by hemocytometer. (F) Stacked bar graph of lung cell numbers based on flow cytometry. (G) Cell number of lung cell populations by flow cytometry. (H-L), Mice were either i.n. treated with IL-9 neutralizing or isotype control antibody in the last week of the chronic allergen challenge (n = 6–7). (H), Schematic of the chronic antibody blockade model. Mice were sensitized intranasally to A.f. 3 times per week for 6-weeks to develop a chronic response. During the last week mice were intranasally treated with IL-9 neutralizing or isotype control antibody during allergen challenge. (I) Total lung cell numbers by hemocytometer. (J) Airway resistance 24 hours after the last intranasal A.f. treatment was measured following intubation and intratracheal challenge with increasing doses of methacholine. Student’s unpaired two-tailed t test comparing isotype vs naïve represented by * and anti-IL-9 vs naive by ^$. (K) Stacked bar graph of lung cell numbers based on flow cytometry. (L-N), B cell and CD4+ cell number per lung. (M-P) Recall allergen model was used to compare Cre- and Cd4-cre+ Il9 fl/fl mice (n = 4–5). (M) Cell number of lung cell populations by flow cytometry. (N), Flow analysis of lung ILC2 cells stimulated with IL-33. (O) Flow analysis of isolated lung CD4+ T cells stimulated with PMA/Ionomycin. (P) IL-9 protein from ex vivo cytokine assay stimulated with A.f. in the presence or absence of IL-33. Data are presented as mean +/− SEM and are representative of 2 independent experiments. Data in D and J are mean +/− SEM from pooled data from 3 independent experiments. Student’s unpaired two-tailed t test was used for comparison to generate p values in D, I, J, L, M, N, and O. One-way ANOVA with a post hoc Tukey test were used to generate p values in B, E, G, and P. *, p<0.05, **, p<0.01, ***, p<0.001, ****p<0.0001.

Figure 7.

Blockade of IL-9 in the recall model alters gene expression in multiple lung cell types. (A) UMAP of unsupervised clustering defining cell types in the lung. (B) UMAP feature plot of comparing Scgb1a1 expression across all cell types with/without anti-IL-9 and isotype treatments. (C) Seurat DotPlot of cytokine receptor expression in lymphoid populations. (D) Heatmap of differentially expressed genes in Il1rl1 expressing lymphoid populations. (E) Heatmap of differentially expressed genes in B cells, neutrophils, and interstitial macrophages. (F) Volcano plot showing differentially expressed genes based on fold change in interstitial macrophages. (G) Dot plot of Il1rl1, Il33 and Tslp expression. (H-I) Seurat DotPlot of gene expression patterns in lung structural cells. (J) Volcano plot showing differentially expressed genes based on fold change in ATII cells.