Mast cells control lung type 2 inflammation via prostaglandin E2-driven soluble ST2

Abstract

Severe asthma and sinus disease are consequences of type 2 inflammation (T2I), mediated by interleukin (IL)-33 signaling through its membrane-bound receptor, ST2. Soluble (s)ST2 reduces available IL-33 and limits T2I, but little is known about its regulation. We demonstrate that prostaglandin E₂ (PGE₂) drives production of sST2 to limit features of lung T2I. PGE₂-deficient mice display diminished sST2. In humans with severe respiratory T2I, urinary PGE₂ metabolites correlate with serum sST2. In mice, PGE₂ enhanced sST2 secretion by mast cells (MCs). Mice lacking MCs, ST2 expression by MCs, or E prostanoid (EP)₂ receptors by MCs showed reduced sST2 lung concentrations and strong T2I. Recombinant sST2 reduced T2I in mice lacking PGE₂ or ST2 expression by MCs back to control levels. PGE₂ deficiency also reversed the hyperinflammatory phenotype in mice lacking ST2 expression by MCs. PGE₂ thus suppresses T2I through MC-derived sST2, explaining the severe T2I observed in low PGE₂ states.

Keywords: IL33; ST2; aspirin exacerbated respiratory disease; asthma; mast cells; nasal polyps; prostaglandin E(2); type 2 inflammation.

Figure 1.. PGE₂ is necessary for Df-induced sST2 production in vivo.

(A) H and E staining of lung tissue from representative saline and Df-treated Ptges^−/− mice and WT C57BL/6 controls. Scale bar is 350 microns. Quantitative inflammation score and goblet cell counts (based on PAS staining) are shown. (B) BAL fluid total cell counts, percentages of eosinophils, and total eosinophil counts from Df- and saline-treated Ptges^−/− mice and WT C57BL/6 controls. (C) Whole lung IL-33 protein concentrations from the indicated groups. (D) Total, proliferating (Ki67+), and activated (KLRG1+) lung ILC2 counts from the same groups as in B. (E) Total, tissue resident β7^lo, and recruited β7^hi MC counts in lung tissues from the indicated groups. (F) Representative intracellular staining for mMCP-1 in recruited β₇^hi and resident β₇^lo lung tissue MCs in saline and Df-challenged WT mice (G) Percentages of mMCP-1⁺ MCs in both strains. (H) sST2 protein detected in in BAL fluids and whole lung from the indicated groups. Results in B-H are from 3–16 mice per group pooled from three independent experiments.

Figure 2.. PGE₂-driven sST2 production limits lung T2I in Df-treated Ptges^−/− mice.

(A) Schematic of protocol used to administer intratracheal vehicle/PGE₂ (10 nmols) or intraperitoneal recombinant IgG1/sST2-Fc. (B) BAL fluid and lung sST2 in Ptges^−/− mice receiving intratracheal PGE₂ or vehicle. (C) Total BAL fluid eosinophil counts for the indicated groups of Ptges^−/− mice treated with PGE₂ or vehicle. (D) Total, proliferating, and activated ILC2 counts for the same groups as in B. (E) Lung MCs, subsets, and mMCP-1 percent positive MCs. (F) Whole lung IL-33. (F) BAL fluid and lung sST2. (G) Total cell BAL fluid eosinophil counts for Df- and saline-treated Ptges^−/− following IgG1 Fc or sST2-Fc. (H) Total, proliferating, and activated lung ILC2 counts for the indicated groups of Ptges^−/− mice. (I) Lung MC counts and subgroups from the indicated groups. (J) Whole lung IL-33 concentrations in the indicated groups of Ptges^−/− mice.

Figure 3.. Serum sST2 correlates with urinary PGE metabolite and disease severity in humans.

Spearman correlations between serum sST2 concentrations and urinary PGE-M, sinonasal outcome test (SNOT-22), asthma control questionnaire (ACQ6), forced vital capacity (FVC), forced expiratory volume (FEV1), and total blood eosinophils. Data are combined from two independent cohorts of patients with AERD.

Figure 4.. PGE₂ preferentially facilitates the expression and production of sST2 by MCs.

(A) Volcano plot of differentially regulated transcripts by PGE₂ including ST2L and sST2. (B) Time dependent changes in sST2 secretion and expression of cell surface ST2L expression by human CBMCs stimulated with IL-33 (10ng/mL) and PGE₂ (1 μM). Secretion of sST2 protein, cell surface ST2L expression, and IL-13 protein concentrations in the supernatants from mouse CBMCs stimulated with IL-33, PGE₂, or both together. (C) The same measurements in mouse BMMCs. (D) Impact of stimulation with PGE₂ with and without IL-33 on expression of transcripts reflecting proximal and distal Il1rl1 promoters by BMMCs. (E) sST2 production and changes in surface ST2L expression by β7^lo and β7^hi MCs sorted from the lungs of Df-treated WT C57BL/6 mice stimulated with IL-33 + PGE₂ (10 ng/mL and 1 uM, respectively) for 24 hours ex vivo. (F) sST2 production, surface ST2L expression, and IL-13 generation by ILC2s sorted from the lungs of Alternaria challenged WT mice. Cells were cultured for 48 hours in the presence of IL-2 + IL-7 and stimulated for 24 hours with IL-33, PGE₂, or both. Results are from 3–7 biologically independent samples per group pooled from at least two independent experiments.

Figure 5.. Complete MC deficiency, but not selective deficiency of constitutive resident MCs, increases lung T2I and suppresses sST2 production.

The impact of MC deficiencies in (A) Cpa3-Cre “Cre-master” mice and (B) Mcpt5-Cre-DTA mice on total BAL fluid eosinophil counts, lung tissue ILC2s, MCs, and MC subset counts, and sST2 concentrations in BAL fluid, whole lung, and serum compared with control strains. Results are from 4–13 mice per group pooled from three independent experiments.

Figure 6.. MC-associated EP₂ receptors are necessary to drive sST2 generation in vivo.

(A) Breeding schematic for the generation of Ptger2 ^fl/flCpa3^Cre/+ (Ptger2 mcKO) mice and littermate Ptger2 ^fl/fl controls. (B) Total BAL fluid cell counts, percentage of eosinophils, and total eosinophils in the indicated groups. (C) Total, proliferating, and activated lung tissue ILC2s (D) Lung tissue MCs, MC subsets, and percentages of mMCP-1+ MCs. (E) Lung tissue concentrations of alarmins. (F) sST2 concentrations in BAL fluid and lung tissue from the indicated groups. Results are from 3–7 mice per group in two experiments.

Figure 7.. PGE₂ concentrations determine the impact of MC-specific deletion of Il1lr1 on lung T2I.

(A) Schematic of breeding scheme used to generate Il1rl1^fl/fl-Cpa3-Cre/+ (Il1lr1 mcKO) and control mice. (B) Representative histograms of ST2L expression in Il1rl1^fl/fl (orange shading) and Il1rl1 mcKO mice (yellow shading, superimposed on blue shaded isotype controls) following saline and Df treatments for β7^lo and β7^hi mast cells. Percent positive is displayed for the Il1rl1^fl/fl controls. (C) Histograms of surface ST2L protein expression by BAL fluid eosinophils, lung basophils, and lung ILC2s from Il1rl1^fl/fl and Il1rl1 mcKO mice following saline and Df treatments for compared to isotype controls. (D) H and E staining of lung tissue from Il1rl1 mcKO mice and littermate control mice following saline and Df treatments. Scale bar is 350 microns. Inflammation score and goblet cell counts (based on PAS staining) are displayed for the indicated groups. (E) Total BAL fluid eosinophil counts, lung tissue ILC2s and MCs, lung homogenate IL-33 protein concentrations, and BAL fluid and lung tissue sST2 concentrations for Il1rl1 mcKO mice and controls following saline and Df treatments. (F) The same parameters for P^−/− Il1rl1 mcKO mice and P^−/− Il1rl1^flfl controls following saline and Df treatments. Results are from 3–16 mice per group pooled from two independent experiments.