IL-33–Responsive Innate Lymphoid Cells Are an Important Source of IL-13 in Chronic Rhinosinusitis with Nasal Polyps

Patients

The research protocol for this study was approved by the Institutional Review Board of The University of Texas Health Science Center, Houston, TX. Ethmoid sinus mucosa removed during the course of medically indicated endoscopic sinus surgery was collected from patients providing informed consent. Patient characteristics are summarized in Table 1.

Flow Cytometry Analysis of ILC

Cells dissociated from sinonasal mucosa using Collagenase D and DNase I (Roche Diagnostics, Mannheim, Germany) were stained with fluorescein isothiocyanate (FITC)-labeled anti-human CD3, CD14, CD19, CD34, αβ-TCR, γδ-TCR, allophycocyanin (APC)-labeled anti-human CD117, and phycoerythrin (PE)-labeled anti-human CD127 (Beckman Coulter, Brea, CA), and ILC were identified as FITC^neg CD127⁺ events. In other experiments, cells were stained with FITC-labeled anti-human CD3, CD14, CD19, CD34, αβ-TCR, γδ-TCR, APC-labeled anti-human CD117, and PE-Cy7–labeled anti-human CD127 (BD Biosciences, San Jose, CA) together with PE-labeled anti-CD161 (Beckman Coulter), anti-CRTH2 (Miltenyi Biotec, Auburn, CA), anti-ST2 (R&D Systems, Minneapolis, MN), or relevant isotype controls. Data were collected using a FACS Calibur (Beckton Dickinson, San Jose, CA) and analyzed using FlowJo software (Tree Star, Ashland, OR). All antibodies are from eBioscience (San Diego, CA) unless otherwise noted.

Lymphoid Cell Culture and Stimulation

A CD45-enriched population was obtained from a single cell suspension from inflamed sinonasal mucosa by positive selection using human CD45 microbeads (Miltenyi Biotec). Cells were plated at a density of 10⁶ cells/ml and stimulated with 10 U/ml recombinant human IL-2 (Chiron, Emeryville, CA) or 100 ng/ml recombinant human IL-33 (Peprotech, Placentia, CA) alone or in combination. Supernatant was collected after 96 hours of stimulation, and IL-13 levels were determined by ELISA (eBioscience). For intracellular staining experiments, cells were stimulated with recombinant IL-2 alone or together with recombinant IL-33 for 6 hours, then stained with FITC-labeled anti-human CD3, CD14, CD19, CD34, αβ-TCR, γδ-TCR, APC-labeled anti-human CD117, and PE-Cy7–labeled anti-human CD127, then fixed and permeabilized and stained with PE-labeled anti-human IL-13 (BD Biosciences).

Epithelial Cell Culture and Stimulation

Epithelial cells were cultured for 24 hours in medium alone or containing extracts from Aspergillus fumigatus, Alternaria alternata, or Cladosporium herbarum (Greer, Lenoir, NC) at a concentration of 10 μg/ml; relative expression of IL-33 was then determined by quantitative real-time polymerase chain reaction (qPCR).

Statistics

Data were analyzed with Prism version 6 software (GraphPad, La Jolla, CA). Data are represented as mean ± SEM. Normally distributed data were analyzed by Student t test, paired t test, or one way analysis of variance with Tukey posttest as appropriate. Nonparametric data were analyzed by Mann-Whitney test or Wilcoxon signed rank test as appropriate. A P value less than 0.05 was considered significant.

Elevated Expression of ST2 in Inflamed Ethmoid Sinus Mucosa from Patients with CRSwNP Compared with Patients with CRSsNP and Healthy Control Subjects

Using qPCR, we examined the expression of IL-33, ST2, and IL1RAP in inflamed ethmoid sinus mucosa from patients with CRSwNP (n = 36), patients with CRSsNP (n = 20), or healthy control subjects (HC) (n = 8) with no history of CRS or allergic disease. Relative expression of ST2 was significantly elevated in inflamed ethmoid sinus mucosa from patients with CRSwNP compared with patients with CRSsNP (P < 0.001) and HC (P < 0.05) (Figure 1A). Expression of IL-33 and IL1RAP was observed in inflamed ethmoid sinus mucosa from patients with CRSwNP, patients with CRSsNP, and HC. Although no significant difference in the relative expression of IL-33 was observed in the inflamed ethmoid sinus mucosa from patients with CRSwNP compared with CRSsNP and HC (Figures 1B and 1C), a high level of expression of IL-33 was evident in tissue from all groups, raising the possibility that IL-33 could interact with an ST2⁺ cell population specifically present within the inflamed mucosa of patients with CRSwNP.

Elevated Percentage of ST2⁺ ILC in Inflamed Sinonasal Mucosa from Patients with CRSwNP Compared with Patients with CRSsNP

Because we previously identified elevated expression of ST2 in inflamed ethmoid sinus mucosa from patients with CRSwNP as compared with patients with CRSsNP and HC, we next sought to identify the cells present within inflamed sinonasal mucosa of patients with CRSwNP expressing ST2. We examined the percentage of ILC, an important cell type expressing ST2, in inflamed sinonasal mucosa of patients with CRS using flow cytometry. We stained cells from inflamed ethmoid sinus mucosa from patients with CRSsNP (n = 4) and patients with CRSwNP (n = 6) with a cocktail of FITC-conjugated antibodies recognizing several lineage markers, including CD3, αβ-TCR, and γδ-TCR, which are expressed on T cells; CD14, which is expressed on monocytes, macrophages, and some dendritic cell subsets; CD19, which is expressed on B cells; and CD34, which is expressed by hematopoietic progenitor cells. The cells were also stained with a PE-conjugated antibody recognizing CD127, the α chain of the interleukin 7 receptor, and an APC-conjugated antibody recognizing CD117, also known as c-kit. We gated on cells lacking lineage markers (lineage^neg) and examined expression of CD117 and CD127 within this population. Results are expressed as the percentage of CD127⁺ cells within the total lineage^neg population. We considered all lineage^neg CD127⁺ cells to be ILC, and observed donor-dependent variable expression of CD117. Representative data from one patient with CRSsNP and one patient with CRSwNP are shown in Figure 2A, and in Figure 2B data from patients with CRSsNP (n = 4) and CRSwNP (n = 6) are compiled. The percentage of lineage^neg CD127⁺ ILC was significantly higher in inflamed ethmoid sinus mucosa from patients with CRSwNP (mean = 1.953 ± 1.11%) compared with patients with CRSsNP (mean = 0.35 ± 0.25%) (P = 0.0095). We then analyzed the surface phenotype of these ILC in greater detail. We stained cells from inflamed mucosa of patients with CRSwNP (n = 4) with the cocktail of FITC-conjugated lineage antibodies described previously, APC-conjugated anti-human CD117, PE-Cy7–conjugated anti-human CD127, and PE-conjugated antibodies recognizing CD161 and CRTH2, markers previously shown to be expressed by ILC (21), and ST2. As shown in Figure 2C, lineage^neg CD127⁺ ILC present in inflamed ethmoid sinus mucosa of patients with CRSwNP expressed ST2 as well as CRTH2 and CD161 in agreement with previously published data (21). Lineage^neg CD127⁺ CD117⁺ cells within the lineage^neg CD127⁺ population could include mast cells, which also express CD117. However, the percentage of lineage^neg CD127⁺ CD117^neg cells was also significantly elevated in inflamed ethmoid sinus mucosa from patients with CRSwNP compared with CRSsNP (Figure 2D). The percentage of lineage^neg CD127⁺ CRTH2⁺ cells was significantly higher in inflamed ethmoid sinus mucosa from CRSwNP as compared with CRSsNP (Figure 2E). We also observed expression of ST2 on both CD4⁺ T cells and mast cells within inflamed ethmoid sinus mucosa of patients with CRSwNP (data not shown). We previously demonstrated increased frequency of both CD4⁺ T cells and mast cells in inflamed sinonasal mucosa from patients with CRSwNP compared with patients with CRSsNP independent of atopy (28), and these cells likely contribute to the elevated expression of ST2 observed in inflamed ethmoid sinus mucosa from patients with CRSwNP.

IL-33 Induces IL-13 Production from a Lineage^neg ILC Population within Sinonasal Mucosa of Patients with CRSwNP

Having identified an elevated percentage of ST2⁺ ILC within inflamed ethmoid sinus mucosa from patients with CRSwNP compared with patients with CRSsNP, we next sought to determine whether ST2⁺ ILC present within inflamed sinonasal mucosa of patients with CRSwNP could respond to IL-33 stimulation. In initial experiments, we examined the production of IL-13 from total lymphoid cells isolated from inflamed mucosa of patients with CRSwNP (n = 10) in response to stimulation with recombinant IL-33 alone or together with recombinant IL-2, a cytokine necessary for lymphoid cell activation. We isolated cells expressing CD45, a marker present on all lymphoid cells, from dissociated sinonasal mucosa from patients with CRSwNP and cultured them for 4 days in medium alone or with recombinant IL-2, IL-33, or both. After stimulation, supernatant was collected and IL-13 production was assessed by ELISA. As shown in Figure 3A, we did not observe a significant increase in IL-13 production from CD45⁺ cells from patients with CRSwNP in response to IL-2 (mean = 173.0 ± 48 pg/ml) or IL-33 (mean = 34.5 ± 12.6 pg/ml) compared with medium control (mean = 2.9 ± 2.1 pg/ml). However, we observed a significantly elevated level of IL-13 production in response to stimulation with IL-2 and IL-33 together (609.0 ± 170.6 pg/ml) compared with stimulation with IL-2 (P < 0.01) or IL-33 (P < 0.001) alone or medium control (P < 0.001). In similar experiments with CD45⁺ cells isolated from inflamed sinonasal mucosa from patients with CRSsNP (n = 3), no significant increase in IL-13 production was observed in response to stimulation with IL-2 and IL-33 together compared with stimulation with IL-2 alone (Figure 3B). Also, the amount of IL-13 produced by CD45⁺ cells from patients with CRSwNP stimulated with IL-2 and IL-33 (range = 80.0–1,620.0 pg/ml) was significantly higher than the level of IL-13 produced by CD45⁺ cells from patients with CRSsNP (range = 32.7–127 pg/ml) (Figure 3C). We did not observe IL-13 production from CD45^neg cells from patients with CRSwNP or CRSsNP in response to stimulation with recombinant IL-2, IL-33, or both (data not shown). This indicates that a CD45⁺ cell population within the inflamed sinonasal mucosa of patients with CRSwNP can produce IL-13 in response to IL-33 stimulation in the presence of IL-2.

To identify the cells producing IL-13 in response to stimulation with IL-2 and IL-33, CD45⁺ cells from inflamed sinonasal mucosa from patients with CRSwNP were cultured for 6 hours with IL-2 alone or together with IL-33, and IL-13⁺ ILC were subsequently detected by intracellular staining. Briefly, cells were stained, as described previously, with a cocktail of FITC-conjugated lineage antibodies, APC-conjugated anti-human CD117, PE-Cy7–conjugated anti-human CD127, then fixed and permeabilized and stained with PE-conjugated anti-human IL-13. Representative data from one patient are shown in Figure 4A, and data from six independent experiments, each with one patient, are compiled in Figure 4B. Lineage^neg CD127⁺ CD117⁺ cells were not observed after in vitro stimulation; therefore, the percentage of IL-13⁺ cells within the lineage^neg CD127⁺ CD117^neg population is shown. After stimulation with IL-2 together with IL-33, the percentage of lineage^neg CD127⁺ ILC that are IL-13⁺ (mean = 28.9 ± 6.6%) is significantly higher than after stimulation with IL-2 alone (mean = 14.8 ± 2.4%) (P = 0.0313). As shown in Figure 4C, sorted ILC produce IL-13 in response to stimulation with recombinant IL-2 and IL-33. These data indicate that ILC present within diseased mucosa in CRSwNP produce IL-13 in response to IL-33 stimulation.

However, other CD45⁺ cells present within diseased mucosa, including T cells and mast cells, can produce IL-13 in response to IL-33 stimulation. We reexamined the data from the intracellular staining experiments described previously and determined the total percentage of IL-13⁺ cells within the lineage^neg or lineage⁺ cell populations after stimulation with IL-2 alone or together with IL-33. There is a significant increase in the percentage of total IL-13⁺ cells within the lineage^neg population in response to stimulation with IL-2 and IL-33 (mean = 6.0 ± 3.1%) compared with IL-2 alone (mean = 1.75 ± 0.5%) (P = 0.0313). However, no significant difference is observed within the lineage⁺ population (Figure 4D). Mast cells are present within the lineage^neg population; however, published data indicate that they respond to IL-33 stimulation without a requirement for concurrent IL-2 (12), and no significant increase in IL-13 production from CD45⁺ cells from patients with CRSwNP was observed in response to stimulation with IL-33 alone (Figure 3A). To confirm that T cells present within diseased mucosa are not a significant source of IL-13 in response to IL-33 stimulation, we cultured CD45⁺ cells from CRSwNP (n = 3) with IL-2 and IL-33 alone or in combination or with plate-bound anti-human CD3 and soluble anti-human CD28, to specifically activate T cells, alone or together with IL-2, IL-33, or both. After 4 days of culture, supernatant was collected and IL-13 levels were determined by ELISA. As shown in Figure 4E, there is little difference in IL-13 production from CD45⁺ cells from patients with CRSwNP in response to stimulation with plate-bound anti-CD3 and soluble anti-CD28 alone or together with IL-2, IL-33, or IL-2 and IL-33, suggesting that T cells present within diseased mucosa are not a significant source of IL-13 in response to IL-33 stimulation. Finally, we stained dissociated cells from CRSwNP mucosa with PE-labeled anti-human CD45, the cocktail of FITC-labeled lineage antibodies described previously, and FITC-labeled anti-human FcεR1α, which is expressed on mast cells but not ILC. We sorted total or lineage-depleted CD45⁺ cells, cultured them for 4 days with recombinant IL-2 alone or with recombinant IL-33, then measured IL-13 in the supernatant by ELISA. The amount of IL-13 produced by total and lineage-depleted CD45⁺ cells is similar (Figure 4F), suggesting that ILC rather than mast cells or T cells are the primary source of IL-13 in response to IL-33 stimulation.

Induction of IL-33 Expression in Primary Sinonasal Epithelial Cells from Patients with CRSwNP in Response to Stimulation with an Extract of Aspergillus fumigatus

We have demonstrated an elevated percentage of ST2⁺ ILC in inflamed ethmoid sinus mucosa from patients with CRSwNP compared with patients with CRSsNP and determined that these cells produce IL-13 in response to stimulation with recombinant IL-2 and IL-33. This raises the question of which cells are producing IL-33 and how its expression is regulated. Numerous cell types, including epithelial cells, express IL-33, and its expression can be up-regulated in response to stimulation with microbial antigens. Numerous environmental fungi are associated with CRS (5), so we investigated their ability to induce IL-33 expression in sinonasal epithelial cells from patients with CRS. We cultured primary sinonasal epithelial cells from patients with CRSsNP (n = 7) and patients with CRSwNP (n = 12) and stimulated them for 24 hours with extracts of several fungi commonly associated with CRS, including Aspergillus fumigatus, Alternaria alternata, and Cladosporium herbarum (5), then determined the relative expression of IL-33 in the cultured epithelial cells using qPCR. We observed significant induction of IL-33 expression in epithelial cells from patients with CRSsNP in response to stimulation with an extract of A. fumigatus (P = 0.0374) but not A. alternata (P = 0.7154) or C. herbarum (P = 0.9212) (Figure 5A). We also observed significant induction of IL-33 expression in epithelial cells from patients with CRSwNP in response to all fungal antigens tested including A. fumigatus (P = 0.0002), A. alternata (P = 0.0444), and C. herbarum (P = 0.0148) (Figure 5B). We subsequently compared the fold induction of IL-33 expression in response to fungal antigen stimulation in epithelial cells from CRSsNP and patients with CRSwNP. In response to stimulation with A. fumigatus extract, induction of IL-33 expression was significantly higher in epithelial cells from CRSwNP as compared with CRSsNP (P = 0.0404) (Figure 5C). However, there was no significant difference in the induction of IL-33 expression in epithelial cells derived from patients with CRSsNP and patients with CRSwNP in response to A. alternata (P = 0.1565) or C. herbarum (P = 0.1325) extracts. We observed a trend toward increased IL-33 protein production in epithelial cells from CRSwNP in response to fungal antigen stimulation, but this was not statistically significant and was variable between patients (see Figure E1 in the online supplement). Collectively, these data indicate that epithelial cells from patients with CRSwNP react to stimulation with A. fumigatus extract with increased induction of IL-33 expression compared with epithelial cells from patients with CRSsNP.