Tissue eosinophils express the IL‐33 receptor ST2 and type 2 cytokines in patients with eosinophilic esophagitis

CONFLICT OF INTEREST

AMU, PJL, PV, KCM, MVY, LQ, LAC, and LRA have no conflicts of interest to disclose. MD receives research support from Novartis and Eli Lilly, consulting fees from Partner Therapeutics, Tillotts Pharma, ORIC Pharmaceuticals, AzurRx, SQZ, Moderna, WebMD, and Experts at Your Fingertips, and Genentech, and is a member of the Scientific Advisory Board for Neoleukin Therapeutics. JJG reports no conflicts of interest related to this study but has received research support from Takeda and the American Partnership for Eosinophilic Disorders. SKD reports no conflicts of interest related to this study, but has received research support from Novartis, BMS, and Eli Lilly, is a co‐founder and SAB member for Kojin, and received consulting fees from GSK.

AUTHOR CONTRIBUTIONS

AMU performed and analyzed flow cytometry, immunofluorescence, and animal experiments. PV, PJL, MVY, LQ, and LRA assisted with acquiring data and conducting experiments. KCM and LAC screened and recruited subjects from the patient cohort and obtained samples for analysis. MD provided patient samples and also managed patient cohort, in addition to providing intellectual guidance. JJG provided patient samples, managed patient cohort, supervised experiments, and analyzed all data providing intellectual guidance of the project. SKD supervised experiments, analyzed all data, and provided intellectual guidance of the project. All authors approved the final version of the manuscript.

Funding information

AMU received funding through NIH T32DK007191 grant. PJL was funded by NIH T32CA207021. JJG has received research award support from the American Partnership for Eosinophilic Disorders (APfED). LQ was funded by a SITC‐Bristol‐Myers Squibb Postdoctoral Cancer Immunotherapy Translational Fellowship. SKD was funded by the Emerson Collective Cancer Research Fund, NIH U01 CA224146, the Ludwig Center at Harvard and is a Pew‐Stewart Scholar in Cancer Research. MD was funded by a Mentored Clinical Scientist Development Award 1K08DK114563, the MGH Center for the Study of Inflammatory Bowel Disease DK043351, and an American Gastroenterology Association Research Scholars Award.

To the Editor,

Eosinophilic esophagitis (EoE) is a type 2 inflammatory disease of the esophagus largely driven by food antigens and is characterized by esophageal eosinophilia, fibrosis, and clinically by dysphagia and food impactions ¹ . The alarmin interleukin (IL)‐33 is elevated in the esophageal epithelium and endothelium of patients with EoE ² , ³ . Exogenous IL‐33 causes esophageal eosinophilia and basal hyperplasia in mice ³ , implicating IL‐33 in EoE pathogenesis. The IL‐33 receptor, suppression of tumorigenicity 2 (IL1RL1/ST2), induces type 2 cytokines IL‐4, IL‐5, and IL‐13. As blockade of IL‐4/IL‐13 improves both histologic and clinical EoE ⁴ , and trials of anti‐IL‐5R are ongoing, we aimed to identify cells that can respond to IL‐33 and the cellular sources of pathogenic Type 2 cytokines in EoE. T‐helper 2 (Th2) CD4 lymphocytes are known producers of IL‐4, IL‐5, and IL‐13, as are group 2 innate lymphoid cells (ILC2), but whether tissue eosinophils, mast cells, or basophils contribute type 2 cytokines in active EoE is under explored ⁵ , ⁶ .

We examined biopsies and blood samples from patients undergoing esophagogastroduodenoscopy at our institution (Table S1). We found eosinophils, mast cells, basophils, and Th2 cells; however, ILC2s were not reliably detected in esophageal tissue and were withheld from further analyses (Figure 1A‐B). ST2 was robustly detected on tissue eosinophils from patients with active EoE (Figure 1C, 1E), whereas blood eosinophils from healthy control, remission, and active EoE were uniformly low for ST2 (Figure 1D, 1F‐G; Figure S1). Similarly, serum levels of soluble ST2 (sST2), which is shed as a decoy receptor, were not significantly different among healthy controls, patients with EoE in remission or active disease (Figure 1H, 1I).

FIGURE 1.

Tissue eosinophils express ST2. (A) Flow cytometry of esophageal biopsies. (B) Cell type as percent of CD45⁺ leukocytes in patients with active (n=8–9) and remission (n=4–7) EoE. Flow cytometry showing ST2 expression across esophageal immune cells (C) and blood eosinophils (D). Quantification of ST2 in tissue (E) and blood eosinophils (F). (G) ST2 on blood versus tissue eosinophils. (H) Concentration of sST2 in healthy non‐EoE, remission, and active EoE patients shown overall (H) and by sex (I). N=34 EoE (16 active, 16 remission), n=34 controls [Color figure can be viewed at wileyonlinelibrary.com]

We next measured IL‐4, IL‐5, and IL‐13 by flow cytometry and immunofluorescence from esophageal tissue in patients with active EoE (Figure 2A‐H). Eosinophils (which unlike T cells were not stimulated with PMA/ionomycin) were frequent expressors of these cytokines (Figure 2A‐E). IL‐4 was detected in most unstimulated eosinophils (60%) compared to mast cells (8.0%) and basophils 3.8% (Figure 2C). We found IL‐13 in 74% of eosinophils compared to 17% of stimulated CRTH2+ Th2 cells, 3.5% of mast cells, and 3.4% of basophils (Figure 2D‐F). Although the proportion of unstimulated mast cells positive for type 2 cytokines was relatively low, mast cells may be pathogenic in EoE and warrant further investigation. Tissue eosinophil expression of type 2 cytokine protein was further verified by immunofluorescent staining for IL‐4, IL‐13, and eosinophil peroxidase (EPX). A majority of the IL‐4⁺ cells in the esophageal epithelium were eosinophils (80.7%); this was similar for IL‐13⁺ cells (82.7%) (Figure 2G‐H; Figure S2). To assess whether eosinophils were necessary for type 2 cytokine induction, we administered IL‐33 to eosinophil‐sufficient (BALB/c WT) and deficient (BALB/c ΔdblGATA‐1) mice using a previously established model of EoE ³ (Figure 2I). Immunofluorescence revealed eosinophil infiltration in IL‐33‐treated WT esophagi but not PBS‐treated WT or ΔdblGATA‐1 mice, as expected (Figure 2J). Although IL‐13 was below the limit of detection in this model, multiplex cytokine analysis of esophagi showed IL‐33‐induced IL‐4 was critically dependent on the presence of eosinophils (Figure 2K).

FIGURE 2.

Human esophageal eosinophils express type 2 cytokines IL‐4, IL‐5, and IL‐13 by flow cytometry and immunofluorescence. (A) Flow cytometry of intracellular cytokine IL‐4, IL‐5, and IL‐13 on esophageal immune cells, quantification shown in (B) with IL‐5 and IL‐13 co‐expression, (C) IL‐4, (D) IL‐13, and (E) IL‐13 MFI. T‐cell cytokines shown in (F). (G‐H) Immunofluorescence of human esophageal tissue identifies eosinophils by eosinophil peroxidase (EPX), IL‐4‐ and IL‐13‐expressing cells with quantification shown to the right. (I) IL‐33 EoE mouse model; n=5/group. (J) Immunofluorescence of mouse esophagi. (K) Esophageal cytokine concentrations for IL‐4, IL‐5, and IL‐13 from various treatment groups [Color figure can be viewed at wileyonlinelibrary.com]

Here, we report frequent expression of the IL‐33 receptor ST2 on esophageal‐infiltrating eosinophils compared to blood eosinophils, other granulocytes, and Th2 cells. Noting eosinophils lack antigen‐specific T‐cell receptors, their ability to respond to IL‐33 may explain continued inflammation observed in patients after removal of dietary antigen. Future studies are warranted to identify factors influencing ST2 and IL‐33 expression and determine whether modulation of this pathway may be a novel therapeutic approach to this increasingly diagnosed yet poorly understood allergic disease.

1.

National Institutes of Health; American Partnership for Eosinophilic Disorders; Massachusetts General Hospital; Harvard Medical School; Bristol‐Myers Squibb; American Gastroenterological Association

Supporting information

Figure S1

Figure S2

Figure S3

Table S1

Supplementary Material