Role of IL-18 transformed CD274-expressing eosinophils in promoting airway obstruction in experimental asthma

Abstract

Background:

IL-5-dependent residential and IL-18-transformed pathogenic eosinophils have been reported; however, the role of IL-18-transformed CD274 expressing pathogenic eosinophils compared to IL-5-generated eosinophils in promoting airway obstruction in asthma has not yet been examined.

Methods:

Eosinophils are detected by tissue anti-MBP and anti-EPX immunostaining, CD274 expression by flow cytometry, and airway resistance using the Buxco FinePointe RC system.

Results:

We show that A. fumigatus -challenged wild-type mice, different gene-deficient mice including naïve CC10-IL-18-transgenic mice accumulates mostly peribrochial and perivascular CD274-expressing eosinophils except naïve CD2-IL-5-transfgenic mice. Additionally, we show that CD2-IL-5 mice following rIL-18 treatment accumulate high number of CD274-expressing perivascular and peribronchial eosinophils with induced collagen, goblet cell hyperplasia, and airway resistance compared to saline challenged CD2-IL5 mice. Furthermore, we also show that even A. fumigatus -challenged IL-5^−/− mice and rIL-18 given ΔdblGATA mice accumulate CD274-expressing eosinophil-associated asthma pathogenesis including airway obstruction. Most importantly, we provide evidence that neutralization of CD274 and IL-18 in A. fumigatus -challenged mice ameliorate experimental asthma. Taken together, the data presented are clinically significant in establishing that anti-IL-18 neutralization is a novel immunotherapy to restrict asthma pathogenesis.

Conclusions:

We demonstrate that IL-18 is critical for inducing asthma pathogenesis and neutralization of CD274 is a potential immunotherapeutic strategy for asthma.

Graphical Abstract

• Expression of CD274 in the BALF of IL-18^−/− mice is significantly reduced compared to IL-5^−/− and WT in A. fumigatus-challenged mice.

• IL-18 treatment induces transformation of CD274⁻ eosinophil to pathogenic CD274⁺ eosinophils.

• In vivo neutralization of CD274 and neutralization of IL-18 reduce airway resistance in A. fumigatus-challenged mice.

Abbreviations: A. fumigatus, Aspergillus fumigatus; BALF, bronchoalveolar lavage fluid; CD2-IL-5-tg, transgenic mice overexpressing IL-5 under the control of the Cd2 promoter; ΔdblGATA, eosinophil-deficient mice; Eos, eosinophil;Ig, immunoglobulin; Mch, methacholine; WT, wild type; IL-5^−/−, IL-5 knockout;IL-18^−/−, IL-18 knockout

INTRODUCTION

Asthma is a chronic inflammatory lung disease characterized by activation of inflammatory cells and mediators, variable obstruction, hyperresponsiveness, and remodeling of the airway.^1-3 In the United States, nearly 25 million people (8% of adults and 9% of children) suffer from asthma, and it is responsible for over 5,000 deaths per year. Several studies have provided evidence that asthma is a Th2-type malady that is induced in response to food or environmental antigens.^4-7 The lung inflammatory response in asthma is characterized by induced expression of multiple gene-encoding cytokines, chemokines, and adhesion molecules that are associated with the recruitment of eosinophils and Th2 lymphocytes.⁸ Asthma patients show evidence of food and aeroallergen hypersensitivity.⁹ Chronic asthma is also characterized by structural changes in the lung through a Th2-mediated eosinophilic inflammatory response known as airway remodeling.^8,10 Clinical studies and experimental modeling have established that Th2 cytokine (IL-4, IL-5, IL-13) signaling is essential for disease pathogenesis.^11-13 These cytokines have a proven role in inducing eosinophilic inflammation in a number of tissues in several diseases.^5,14-16 Numerous studies in mice^17,18 and humans¹⁹ have supported an important role for eosinophilic inflammation in allergen-induced asthma. It is believed that asthma pathogenesis is prompted by the induction of IL-5-associated eosinophilia in response to several types of allergens. IL-5 is a well-established eosinophil differentiation and growth factor which is induced in several allergic diseases including asthma.^20,21 Of note, the intrinsic molecular pathway of IL-5-regulated eosinophilia is largely unknown and remains to be defined.⁹ Most recently, IL-5 responsive pathogenic eosinophil subsets have been reported in mice; however, they have not been validated in a human allergic diseases.²² Anti-IL-5 therapy causes significant improvement in eosinophilia in asthma patients, but the eosinophilia returns to the tissue once therapy is withdrawn.^15,19,23-25 Additionally, a clinical report demonstrated no optimal response to anti-IL-5 therapies in substantial prototype patients,²⁶ and functional tissue eosinophils were detected following post 750 mg mepolizumab treatment.²⁷ In addition, the effects of anti-IL-5 immunotherapy on several characteristics of asthma pathogenesis, like improved mucus production and airway resistance, have not been fully studied. It is possible that the long-term effects of anti-IL-5 may compromise innate immunity of the gastrointestinal tract, as eosinophils reside in the gastrointestinal tract and are involved in parasite exclusion. Recently, we presented our novel finding that IL-18 can generate IL-5-independent subsets of eosinophils and transform IL-5-responsive naïve eosinophils into CD274-expressing pathogenic eosinophils.²² Several clinical reports show that IL-18 is induced in allergic patients, including asthmatics.^28-33 CD274 (PDL1) is known to be expressed on T cells, B cells, macrophages, and dendritic cells (DCs).³⁴ Previous reports indicated that CD274 gene-deficient mice showed reduced levels of airway hyperreactivity and reduced asthmatic inflammation by enhancing production of IFN-γ by iNKT cells.³⁵ However, we previously reported that iNKT cells are not critical in promoting asthma pathogenesis, as A. fumigatus -challenged iNKT cell-deficient mice are not protected from eosinophilic inflammation.³⁶ Thus, the current study provides insight into the critical role of IL-18 and IL-18-responsive pathogenic CD274+ eosinophils in asthma pathogenesis, including promoting airway hyperresponsiveness in A. fumigatus - or cytokine-induced asthma. Herein, we provide mechanistic evidence for the significance of IL-18-associated CD274 in experimentally induced asthma. Accordingly, we propose neutralization of IL-18 or CD274 as a potential target therapy for human asthma.

Material and Methods

Mice

Specific pathogen-free BALB/c, ΔdblGATA (GATA1deficiency) BALB/c background mice, C57BL6 and IL-18^−/− C57BL6 background mice were obtained from the Jackson Laboratory. IL-5^−/− and CD2-IL-5 transgenic (CD2-IL-5 Tg) BALB/c background mice were provided by Marc Rothenberg, MD, PhD (Cincinnati Children’s Hospital Medical Center, Cincinnati OH).³⁷ CC10-IL-18 transgenic (CC10-IL-18 Tg) C57BL6 background mice were obtained from the laboratory of Dr. Jack Elias, MD., PhD, Yale University.³⁸ The mice were maintained in a pathogen-free barrier facility. All experiments were performed on age- and gender-matched 6–8-week-old mice. The Tulane Institutional Animal Care and Use Committee (IACUC) approved the animal protocols, which were employed in accordance with NIH guidelines. We used both C57BL6 and BALB/c strains of mice based on the availability of respective gene-deficient and transgenic mice. The CD2-IL-5 Tg, IL-5^−/−, and ΔdblGATA mice were BALB/c background and CC10-IL-18 Tg and IL-18^−/− were C57BL6. The double knockout IL-5^−/−/IL-18^−/− mice were generated by mating individual IL-5^−/− with IL-18^−/− mice that provided IL-5^−/−/IL-18^−/− double gene-deficient mice in F2 generation. Both IL-5^+/+ and IL-18^+/+ mice were used as littermate controls, as described previously.^39,40 Experimental asthma was induced using established methods.^5,41,42

Measurement of airway resistance (RI)

The trachea was surgically exposed in anesthetized mice and a cannula was inserted and tied with a suture to prevent air leakage.^43,44 Mice were placed in the Buxco Finepointe RC system chamber (Data Sciences International [DSI] St. Paul, MN) and connected to a ventilator. Mechanical ventilation was started with the appropriate respiratory rate and tidal/stroke volume. After two baseline measurements lasting 3 min, mice were exposed to phosphate-buffered saline (PBS) or methacholine (MCh) (3.125 to 50 mg/ml) and the airway resistance (RI) was measured as per the manufacturer's protocol.

Additional details regarding the method are provided in the supporting information.

Statistical analysis.

Data are expressed as mean ± standard deviation (SD). Statistical significance comparing different groups of mice was determined using unpaired InStat GraphPad Prism5 Version 5.03 (San Diego, CA) software and the nonparametric one-way ANOVA Kruskal-Wallis test followed by Dunn’s corrections for multiple comparisons. A p-value < 0.05 was considered statistically significant.

RESULTS:

Analysis of eosinophil subtypes in the lungs of A. fumigatus -induced experimental asthma

Previously, we reported that rIL-5-generated naïve eosinophils from bone marrow precursors in response to IL-18 transform into CD274⁺ eosinophils, and both CD274⁺ and CD274⁻ eosinophils are present in mice and humans.^22,33 Therefore, in this study we examined the eosinophil subtypes in allergen-challenged experimental asthma. Experimental asthma was induced by challenging BALB/c mice with A. fumigatus extract as per established protocol.^5,42 Histopathological analysis of major binding protein (MBP) and morphometric quantification of lung sections showed that most of the eosinophils in A. fumigatus -challenged WT mice accumulated in the peribronchial and perivascular epithelium compared to very few eosinophils in the lung parenchyma of saline-challenged wild-type (WT) mice (Fig. 1A, B). Our analysis detected ~97% CD274⁺CCR3⁺Seglec-F⁺ eosinophils in the BALF of A. fumigatus -challenged WT mice, based on the isotype controls (Fig. 1C-F). Further, we showed that CD2-IL-5 Tg mice accumulate eosinophils in the lung parenchyma (Fig. 1G) compared to the peribronchial and perivascular eosinophils accumulation in CC10-IL-18 Tg (DOX) mice (Fig. 1H). Eosinophil accumulation in the peribronchial and perivascular regions is observed in all A. fumigatus -challenged IL-5^−/− (Fig. 1L) and IL-18^−/− mice (Fig. 1N), but in low numbers compared to A. fumigatus -challenged WT mice (Fig. 1B, P). Saline-challenged mice do not show any eosinophils in the lungs (Fig. 1I, J, K, M, O). BALF analysis of CCR3⁺Siglec-F⁺ eosinophils of A. fumigatus challenged WT (Fig. 1C,D) and IL-5^−/− (Fig. 1Q,S) mice showed that more lung eosinophils express CD274 compared to the A. fumigatus -challenged IL-18^−/− mice (Fig. 1R,T). The absolute number of BALF eosinophils (Fig. 1U,V) along with CD274⁺ and CD273⁻ eosinophils in A. fumigatus -challenged WT, IL-5^−/− (Fig. 1Q,S), and IL-18^−/− mice is presented in Fig. 1R,T. More BALF eosinophils in IL-18^−/− mice are CD274⁻ eosinophils (Fig. 1X) compared to IL-5^−/− mice (Fig.1W). Very few eosinophils show the expression of CD274 in IL-18^−/− mice. This may be caused by antibody cross-reactivity, as CD273 and CD274 have 70% amino acid homology,^45,46 or due to compensatory mechanism by which IL-1 family of some inflammatory cytokine bind to IL-18R and induce CD274 expression on cells.⁴⁷ The schematic representation of A. fumigatus -induced experimental asthma protocol is presented in Suppl. Fig. 1A with flow cytometry gating strategy for CCR3⁺ Siglec-F⁺ eosinophil analysis in Suppl. Fig. 1B. In addition, we show quantification of MBP⁺ tissue eosinophils in saline- or A. fumigatus -challenged WT, IL-5^−/− and IL-18^−/−, CD2-IL-5 Tg and CC10-IL-18 Tg mice (Suppl. Fig. 1C, D, E).

Figure 1. Lung-accumulated eosinophils express CD274 following the induction of experimental asthma.

A representative photomicrograph of anti-MBP immunostained lung sections shows peribroncheal and perivascular eosinophil accumulation of MBP+ cells in an A. fumigatus -challenged WT (BALB/c) murine model of experimental asthma (A, B). A representative flow cytometry analysis of BALF live CCR3⁺Siglec-F⁺ eosinophils (C) and CD274⁺ eosinophils (D) based on isotype controls (E, F). A representative lung photomicrograph indicated most MBP⁺ eosinophils accumulate in the lung parenchyma of CD2-IL-5 Tg mice (G) compared to the peribroncheal eosinophil accumulation in DOX-inducible CC10-IL-18 Tg mice (N). Saline-challenged WT and CC10-IL-18 (no DOX) Tg mice show no baseline eosinophils (I-J). Similar peribroncheal eosinophils were observed in A. fumigatus -challenged IL-5^−/− (L) and IL-18^−/− (N) mice compared to no eosinophils in respective saline-challenged mice (K, M). Flow cytometry analysis detected CD274 expression in the BALF of all CCR3⁺Siglec-F⁺ eosinophils of A. fumigatus -challenged IL-5^−/− mice (Q, S) compared to mostly CD274⁻ eosinophils in A. fumigatus -challenged IL-18^−/− mice (R, T). BALF eosinophil quantitation shows a significantly increased number of eosinophils in A. fumigatus -challenged WT, IL-5^−/− and IL-18^−/− mice compared to the respective saline-challenged WT (BALB/c, C57BL/6) mice (U, V). Significantly reduced CD274 expression in the BALF of A. fumigatus -challenged IL-5^−/− (W) mice compared to increased CD274 expression in A. fumigatus -challenged IL-18^−/− (X) with their respective A. fumigatus -challenged WT (BALB/c, C57BL/6) mice. Data are expressed as mean ± SD, n=8–10 mice/group. *p<0.05; **p<0.001; ***p<0.0001. Photomicrographs presented are 100X original magnification.

Comparative analysis of IL-5, IL-13 and IL-18 in A. fumigatus -challenged WT and IL-5^−/− mice

BALF eosinophils and active cytokines IL-18, IL-5, and IL-13 were analyzed in Aspergillus-challenged WT and IL-5^−/− mice by performing ELISA analysis. The levels of IL-18 were highly elevated compared to IL-5 and IL-13 (Fig. 2A-C). This data indicates that IL-5 may negatively regulate IL-18, which promotes eosinophil accumulation even in IL-5^−/− mice following A. fumigatus challenge. The anti-NLRP3, anti-F4/80 and anti-IL-18 immunofluorescence staining shows that NLRP3 is induced in the F4/80⁺ lung accumulated macrophages (Supp Fig. 2A i-ii) and NLRP3-IL-18 colocalization further indicates that NLRP3 activation produce IL-18 in A. fumigatus -challenged mice (Supp Fig. 2B i-ii). These data provide the evidence that lung-accumulated macrophages are the source of IL-18 in A. fumigatus -challenged mice. Additionally, eosinophil-induced airway resistance was analyzed in A. fumigatus -challenged WT and IL-5^−/− mice using the Buxco Finepointe RC system. This analysis indicated that both A. fumigatus -challenged WT and IL-5^−/− mice have increased airway resistance compared to respective saline-challenged mice (Fig. 2D). Further, goblet cell hyperplasia was detected in A. fumigatus -challenged IL-5^−/− mice and WT mice, but no goblet cells were observed in saline-challenged mice (Fig. 2E). Quantitative analysis showed reduced goblet cells in IL-5^−/− mice compared to WT mice (Fig. 2F). We also observed significantly reduced collagen accumulation in A. fumigatus -challenged IL-5^−/− mice compare to WT mice (Fig. 2G). The reduced airway resistance in A. fumigatus -challenged IL-5^−/− mice compared to WT mice is consistent with reduced levels of IL-13 protein (Fig. 2B), mRNA levels (Suppl. Fig. 1F), goblet cell hyperplasia, and collagen accumulation in the lungs of A. fumigatus -challenged IL-5^−/− mice compared to WT mice. Interestingly, we observed induced MUC1 and MUC5AC protein levels in both A. fumigatus -challenged WT and IL-5^−/− mice (Fig. 2H). These data indicate that IL-5 is redundant for the generation of the CD274 expressing eosinophils that is responsible for promoting airway obstruction even in allergen challenged IL-5^−/− mice.

Figure 2. Induced IL-18 is associated with airway resistance and goblet cell hyperplasia in A. fumigatus challenged WT and IL-5^−/− mice.

ELISA analysis detected elevated levels of IL-18 (A), decreased levels of IL-13 (B), and no IL-5 (C) in the BALF of A. fumigatus -challenged IL-5^−/− mice compared to A. fumigatus -challenged WT (BALB/c) mice. Airway resistance (RI) in response to different concentrations of methacholine is shown in A. fumigatus -challenged IL-5^−/− and WT mice (D). A representative photomicrograph of goblet cell hyperplasia (E) and quantitation of PAS-positive goblet cells in WT, IL-5^−/−, A. fumigatus -challenged WT and A. fumigatus -challenged IL-5^−/− mice (F). Representative photomicrographs of lung sections with Masson’s trichrome staining show perivascular and peribronchiolar collagen accumulation in the A. fumigatus and WT mice compared to IL-5^−/− and A. fumigatus -challenged IL-5^−/− mice (G). Immunoblot analysis of MUC1, MUC5AC and GAPDH protein levels in IL-5^−/− and WT (BALB/c) A. fumigatus -challenged mice (H). Data are expressed as mean ± SD, n=8–10 mice/group. *p<0.01, **p<0.001; ***p<0.0001 (WT + A. fumigatus vs IL-5^−/− + A. fumigatus); ^##p<0.001, ^###p<0.0001 (WT + A. fumigatus vs WT + saline); ^$p<0.04 (IL-5^−/− + A. fumigatus vs IL-5^−/− + saline). Photomicrographs presented are 100X original magnification.

Analysis of CD274+ eosinophils induced upon administration of rIL-18 in CD2-IL-5 transgenic mice

We previously showed that most of the eosinophils present in CD2-IL-5 transgenic mice do not express CD274.²² Therefore, we next examined whether rIL-18-challenged mice can transform all eosinophils into the CD274⁺ pathogenic eosinophils that exacerbate airway resistance and other characteristics of asthma pathogenesis. Accordingly, both naïve WT mice and CD2-IL-5 transgenic mice were challenged with 10μg of rIL-18 in 50μl saline or saline alone via intranasal (i.n.) route as in the presented schematic protocol (Suppl. Fig. 2C). We used the i.n. route to examine local overexpression of IL-18-induced responses in eosinophilic asthma, which has been previously reported.^38,48,49 CCR3⁺Siglec-F⁺ eosinophil samples in the BALF of saline and rIL-18-challenged WT mice and CD2-IL-5 transgenic mice were analyzed for CD274 expression using flow cytometry. BALF analysis indicated that rIL-18-challenged WT mice showed a larger number of CCR3⁺ and anti-Siglec-F⁺ eosinophils compared to saline-challenged (Fig. 3A-B), which was further increased by ~3-fold in rIL-18- challenged CD2-IL-5 transgenic mice (Fig. 3C-D). The rIL-18-treated CD2-IL-5 transgenic mice had a significantly higher number of CD274⁺ eosinophils compared to rIL-18-challenged WT mice (Fig. 3F-H). The quantification of eosinophils in BALF showed that rIL-18-challenged CD2-IL-5 transgenic mice have ~3-fold higher number of induced CD274⁺ eosinophils compared to rIL-18-challenged WT mice (Fig. 3I-J). rIL-18 treatment induced peribronchial and intraepithelial accumulation of eosinophils in the lungs of CD2-IL-5 transgenic mice compared to rIL-18-challenged WT mice. Saline-treated respective controls showed fewer baseline eosinophils (anti-MBP) in the lungs of compared to rIL-18 challenged WT and CD2-IL-5 transgenic mice (Fig. 3K). Additionally, the anti-EPX staining revealed very high number of perivascular and peribronchial intact and degranulated eosinophils (extracellular eosinophilic granules) in rIL-18-challenged CD2-IL-5 transgenic mice compared to saline-challenged mice (Fig. 3L). In comparison, anti-MBP staining revealed a low number of mostly intact eosinophils in rIL-18-challenged CD2-IL-5 transgenic mice (Fig. 3M). Further, we observed increased PAS-stained goblet cell hyperplasia in rIL-18-treated CD2-IL-5 transgenic mice compared to WT mice, and very few to none were detected in saline-treated CD2-IL-5 transgenic mice and WT mice, respectively (Fig. 3N, O). Further, to establish that rIL-18- challenged CD274⁺ eosinophils have a critical role in promoting airway resistance, we examined airway resistance in rIL-18- challenged WT and CD2-IL-5 transgenic mice in response to different concentrations of methacholine. The rIL-18- challenged CD2-IL-5 transgenic mice showed significantly increased airway resistance compared to rIL-18-challenged WT mice (Fig. 3P). Saline-challenged CD2-IL-5 transgenic mice showed similar airway resistance as rIL-18-challenged WT mice; this may be due to the presence of baseline endogenous IL-18 in CD2-IL-5 transgenic mice (Fig. 3Q). Furthermore, a significant increase in the levels of eosinophil-active cytokines IL-5 and IL-13 was observed in rIL-18- challenged CD2-IL-5 transgenic mice and WT mice compared to saline-challenged mice (Suppl. Fig. 2D, E). We also observed induced collagen accumulation in the lungs of rIL-18-challenged CD2-IL-5 transgenic mice compared to WT mice (Suppl. Fig. 2F i-iv). The morphometric quantitation of MBP⁺ and EPX⁺ cells in the lungs of saline- and rIL-18-challenged WT and CD2-IL-5 transgenic mice is presented in Suppl. Fig. 2G, H. These data establish that IL-18 is indeed critical in transforming IL-5-responsive eosinophils to CD274⁺ pathogenic eosinophils that promote asthma pathogenesis.

Figure 3. Asthma pathogenesis in IL-18-induced CD274⁺ pathogenic eosinophils in CD2-IL-5 transgenic mice.

Induced detection of CCR3⁺Siglec-F⁺ eosinophils (A-D) and CD274 expression (E-H) in the BALF of CD2-IL-5 Tg and WT (BALB/c) rIL-18 treated mice compared to respective controls. The quantification of CD274⁺ eosinophils in BALF of saline- and rIL-18-treated WT mice and CD2-IL-5 Tg mice (I, J). A representative photomicrograph lung section of anti-MBP immunostained lung sections of saline- and rIL-18-treated WT and CD2-IL-5 Tg (K) detected high numbers of MBP⁺ eosinophils. The anti-EPX stained tissue sections show intact and degranulated (extracellular granules) eosinophils (indicated by black arrows) (L) compared to intact anti-MBP stained eosinophils (indicated by black arrows) in tissue sections of saline- and rIL-18-treated CD2-IL-5 Tg mice (M, original magnification 400X). A representative photomicrograph of PAS-stained goblet cell hyperplasia and quantitation of rIL-18-treated CD2-IL-5 Tg mice compared to WT mice (N, O). Airway resistance (RI) in response to methacholine was measured in rIL-18-or saline-treated CD2-IL-5 Tg and WT mice (P). ELISA analysis detected highly induced protein levels of IL-18 in the serum of rIL-18-treated CD2-IL-5 Tg mice compared to WT mice (Q). Data are expressed as mean ± SD, n=10–12 mice/group. *p<0.02, **p<0.001; ***p<0.0001 (CD2-IL-5 Tg + saline vs CD2-IL-5 Tg + rIL-18); ^#p<0.02, ^##p<0.001 (WT+ saline vs WT + rIL-18); ^$$ p<0.001 (WT + rIL-18 vs CD2-IL-5 Tg +rIL-18); ^§§p<0.001 (WT-saline vs CD2-IL-5 Tg + saline). Photomicrographs presented are 100X (K,N) and 400X (L,M) original magnification , respectively.

IL-18-induced eosinophilic asthma pathogenesis in ΔdblGATA mice.

Next, to establish that IL-18 is indeed critical for the in vivo maturation of CD274⁺ eosinophils that promotes asthma pathogenesis, WT and ΔdblGATA mice were challenged intranasally with 10μg of rIL-18 in 50μl saline or saline alone. The ΔdblGATA mice are deficient in mature eosinophils but have bone marrow eosinophil precursors.⁵⁰ We present direct evidence that rIL-18 is critical for the in vivo generation of CD274⁺ pathogenic eosinophils. The rIL-18 challenge to ΔdblGATA mice showed induction of CD274⁺CCR3⁺SiglecF⁺ eosinophils in the BALF, blood, and bone marrow (Fig. 4A i-viii) and peribronchial accumulation of lung eosinophils compared to saline-challenged ΔdblGATA mice (Fig. 4B). The morphometric quantitation of MBP⁺ cells showed a comparable number of tissue-accumulated eosinophils in the lungs of rIL-18-challenged WT and ΔdblGATA mice (Fig. 4C). We also found that induction of IL-18-generated CD274⁺ eosinophils accumulation resulted in induced IL-13 protein with goblet cell hyperplasia, collagen accumulation, and airway obstruction in rIL-18-challenged ΔdblGATA mice compared to saline-challenged ΔdblGATA mice (Fig. 4D, E, F, G). The morphometric quantitation of goblet cells in the lungs of saline- and rIL-18-challenged WT and ΔdblGATA mice are presented in Fig. 4H.

Figure 4. IL-18-induced eosinophilic asthma pathogenesis in ΔdblGATA (GATA1 gene-deficient) mice.

Induced detection of CD274 expression in CCR3⁺Siglec-F⁺ eosinophils in the BALF (A i-ii), blood (A iii-iv), and bone marrow (A v-vi) of rIL-18- challenged ΔdblGATA mice. The quantification of CD274⁺ eosinophils in BALF, blood, and bone marrow of saline- and rIL-18- challenged ΔdblGATA mice (A vii-viii). A representative photomicrograph of anti-MBP immunostained lung sections with quantitation of eosinophils in saline- and rIL-18- challenged WT (BALB/c) and ΔdblGATA mice (B,C). Airway resistance (RI) in response to methacholine was measured in ΔdblGATA and WT (BALB/c) mice challenged with or without rIL-18 (D). ELISA analysis detected comparable induced IL-13 protein levels in the BALF of rIL-18-challenged WT and ΔdblGATA mice compared to the saline-challenged mice (E). A representative photomicrograph of PAS-stained goblet cell hyperplasia with quantitation of rIL-18-challenged WT and ΔdblGATA mice (F, H). Masson’s trichrome stained lung sections showed collagen accumulation in rIL-18-challenged ΔdblGATA mice and WT (BALB/c) mice compared to saline-challenged mice (G). Data are expressed as mean ± SD, n=6–10 mice/group. *p<0.01, **p<0.001, ***p<0.0001 (WT+ saline vs WT + rIL-18); ^#p<0.01, ^##p<0.001, (ΔdblGATA + saline vs ΔdblGATA + rIL-18). Photomicrographs presented are 100X original magnification.

Analysis of asthma pathogenesis in IL-5^−/−/IL-18^−/− mice

Since both IL-5 and IL-18 can generate, proliferate, and transform naïve and pathogenic eosinophils, we further tested the hypothesis that IL-5 and IL-18 synergy is critical in promoting eosinophil-induced asthma pathogenesis. Accordingly, we generated endogenous IL-5 and IL-18-deficient (IL-5^−/−/IL-18^−/−) mice. Experimental asthma was induced in IL-5^−/−/IL-18^−/− mice along with littermate-matched control mice following the schematic protocol (Suppl. Fig. 3A).⁵ Anti-MBP immunohistochemical analysis detected no eosinophils in the lung sections of A. fumigatus -challenged IL-5^−/−/IL-18^−/− mice compared to high levels in A. fumigatus -challenged littermate-matched control mice. Saline-challenged mice did not show any lung eosinophilia (Fig. 5A-D). Like tissue eosinophilia, very few baseline eosinophils (with no CD274⁺ eosinophils) were detected in the BALF of IL-5^−/−/IL-18^−/− mice by flow cytometry analysis compared to high levels in A. fumigatus -challenged littermate-matched control mice (Fig. 5E-H). The quantification of absolute eosinophils in BALF and accumulation in lung tissue showed they were highly induced in A. fumigatus -challenged littermate matched control mice compared to almost none in IL-5^−/−/IL-18^−/− mice (Fig. 5I-J). Additionally, we observed highly reduced baseline bone marrow eosinophils (Fig. 5K) and A. fumigatus challenge induced blood eosinophilia (Fig. 5L) in the IL-5^−/−/IL-18^−/− mice compared to IL-5^−/− and IL-18^−/−, and littermate-matched control mice. Lastly, airway resistance analysis indicated that A. fumigatus -challenged IL-5^−/−/IL-18^−/− mice showed significantly improved airway resistance compared to A. fumigatus -challenged littermate-matched control mice (Fig. 5M). Some airway resistance was still observed in A. fumigatus -challenged IL-5^−/−/IL-18^−/− mice that may be due to the A. fumigatus -induced IL-13 in IL-5^−/−/IL-18^−/− mice that is an established enhancer of mucus production. In addition, asthmatic IL-5^−/−/IL-18^−/− mice had significantly lower levels of eosinophil-active inflammatory cytokines than their A. fumigatus -challenged littermate matched control mice; IL-5 and IL-13 levels were almost undetectable in saline- or A. fumigatus -challenged IL-5^−/−/IL-18^−/− mice (Suppl. Fig. 3B-C). The morphometric quantitation of MBP⁺ cells showed only a few baseline eosinophils in saline- and A. fumigatus -challenged lungs of IL-5^−/−/IL-18^−/− mice compared to high levels in littermate-matched control WT mice (Suppl. Fig. 3D). This data further confirms that the eosinophils observed in IL-5^−/− mice are due to the presence of endogenous IL-18, which is consistent with the report that IL-5^−/− mouse bone marrow eosinophil precursors develop into mature eosinophils in response to rIL-18.²²

Figure 5. IL-5^−/−/IL-18^−/− double gene-deficient mice are deficient in bone marrow, blood and lung eosinophils following allergen induced asthma.

A representative photomicrograph of anti-MBP antibody immunostained lung sections shows no accumulation of eosinophils in asthmatic IL-5^−/−/IL-18^−/− mice compared to A. fumigatus -challenged littermate matched control mice (A-D). Representative flow cytometry analysis of CCR3⁺Siglec-F⁺ eosinophils (E-F) and CD274 expression (G, H) in the BALF of A. fumigatus -challenged littermate matched control mice and IL-5^−/−/IL-18^−/− mice. Eosinophil quantitation of BALF in A. fumigatus -challenged littermate matched control and IL-5^−/−/IL-18^−/− mice compared to saline-challenged littermate matched control mice (I, J). Baseline bone marrow and A. fumigatus -induced blood eosinophils in IL-5^−/−/IL-18^−/− mice (K, L). Airway resistance (RI) in response to methacholine was measured in A. fumigatus -challenged littermate matched control and IL-5^−/−/IL-18^−/− mice (M). Data is expressed as mean ± SD, n=6–10 mice/group. *p<0.01, **p<0.001, (control vs A. fumigatus); ^#p<0.01, ^###p<0.0001 (Littermate-matched control + A. fumigatus vs IL-5^−/−/IL-18^−/− + A. fumigatus); Photomicrographs presented are 100X original magnification.

Neutralization of CD274 ameliorates A. fumigatus -induced experimental asthma.

Next, we aimed to establish whether IL-18-induced CD274+ eosinophils are responsible for the induction of experimental asthma. Accordingly, we neutralized CD274⁺ cells in vivo in an experimental mouse model of asthma by intraperitoneal injection of neutralizing anti-CD274 antibody (1 mg two times per week for 3 weeks) as in the presented schematic protocol (Suppl. Fig. 4A). Neutralization of CD274 in A. fumigatus -challenged mice resulted in reduced airway eosinophils in BALF compared to isotype control-treated A. fumigatus -challenged mice (Fig. 6A). A quantification of CD274⁺ eosinophils in BALF by flow cytometry showed a ~3-fold reduction upon neutralization of CD274 in A. fumigatus -challenged mice compared to isotype control-treated A. fumigatus -challenged mice (Fig. 6B). A similar reduction in perivascular and peribronchial tissue eosinophils was detected upon neutralization of CD274 in A. fumigatus -challenged mice compared to isotype control-treated A. fumigatus -challenged mice; no eosinophils were detected in saline-challenged mice (Fig. 6C). Further, lung function tests (airway resistance analysis) indicated that A. fumigatus -challenged mice upon CD274 neutralization show significantly improved airway resistance compared to A. fumigatus -challenged isotype control-treated mice (Fig. 6D). A significant reduction in PAS-stained mucus-producing goblet cells was detected in upon CD274 neutralization in Aspergillus-challenged mice compared to isotype control-treated A. fumigatus s-challenged mice (Fig. 6E), and a morphometric analysis indicated significantly reduced goblet cells upon neutralization of CD274 in A. fumigatus challenged mice compared to isotype control-treated A. fumigatus -challenged mice (Fig. 6F). Flow cytometry analysis of BALF cells showed reduced expression of CD274 in the CCR3⁺SiglecF⁺ eosinophils of A. fumigatus -challenged mice upon neutralization of CD274 compared to isotype control-treated A. fumigatus -challenged mice (Suppl. Fig. 4B,C). Reduced collagen accumulation was observed in allergen-challenged mice upon neutralization of CD274 compared to A. fumigatus -challenged isotype control-treated mice (Suppl. Fig. 4D i-iv). ELISA analysis showed significantly reduced levels of IL-18 and IL-13 in A. fumigatus -challenged mice treated with anti-CD274 mice compared to isotype control-treated and A. fumigatus -challenged mice (Suppl. Fig. 4E, F). Morphometric quantitation of anti-MBP staining showed a statistically significant reduction of lung tissue-accumulated eosinophils in A. fumigatus -challenged mice treated with anti-CD274 antibody compared to isotype control-treated and A. fumigatus -challenged mice (Suppl. Fig. 4G). The presented data also indicate that CD274 neutralization downregulates also T cell function in A. fumigatus -challenged mice, suggesting that CD274⁺ eosinophil accumulation-induced responses are indeed responsible for promoting asthma pathogenesis.

Figure 6. CD274-expressed eosinophils are indeed responsible for asthma pathogenesis in mice including airway obstruction.

A significant difference was detected in CD274+ BALF eosinophil numbers in the A. fumigatus -challenged WT mice upon neutralization of CD274 and A. fumigatus -challenged WT mice treated with matching isotype control (A, B). A representative photomicrograph of anti-MBP antibody immunostained eosinophils in lung sections of A. fumigatus -challenged anti-CD274-treated and isotype control-treated WT mice (C). Airway resistance (RI) in saline- and A. fumigatus -challenged anti-CD274-treated and anti-isotype control-treated mice are shown (D). A representative photomicrograph of goblet cell hyperplasia (E) and respective cell numbers in A. fumigatus -challenged anti-CD274-treated and isotype control-treated mice (F). Data is expressed as mean ± SD, n=8–12 mice/group. * p<0.01, ** p<0.001 (WT + A. fumigatus + anti-CD274 vs WT + A. fumigatus + isotype control). Photomicrographs presented are 100X original magnification.

Neutralization of IL-18 protects Aspergillus-induced CD274-expressing eosinophil-mediated induction of experimental asthma.

We further examined the effectiveness of IL-18 neutralizing antibody in restricting the generation and transformation of CD274+ pathogenic eosinophils that induce asthma in A. fumigatus -challenged mice. Accordingly, we intraperitoneally (IP) injected anti-IL-18 neutralizing antibody (200 μg two times per week for 3 weeks) and isotype control in WT mice following the schematic protocol presented in Suppl. Fig. 5A. Previous studies have shown that the IP injection route is pharmacologically relevant for slow absorption and long-term body retention of neutralizing antibodies.^51,52 BALF eosinophils (Fig. 7A,C) and CD274+ eosinophils were analyzed by flow cytometry. The neutralization of IL-18 in Aspergillus-challenged mice showed significantly reduced numbers of CD274-⁺ eosinophils compared to isotype control-treated A. fumigatus -challenged mice (Fig. 7B,D). Further, a similar reduction in peribronchial accumulation of tissue eosinophils was detected by anti-MBP immunostaining (Fig. 7E i-iv), and morphometric analysis indicated that MBP⁺ cells are significantly reduced upon neutralization of IL-18 in A. fumigatus -challenged mice compared to isotype control-treated A. fumigatus -challenged mice (Fig. 7F). Lung functional analysis indicated that A. fumigatus -challenged mice treated with anti-IL-18 show significantly improved airway resistance compared to A. fumigatus -challenged isotype control-treated mice (Fig. 7G), which correlates with the reduced number of CD274⁺ eosinophils. We also detected goblet cell hyperplasia via PAS staining in A. fumigatus -challenged mice (Fig. 7H) and morphometric analysis indicated that PAS positive goblet cells are significantly reduced in upon neutralization of IL-18 in A. fumigatus -challenged mice compared to isotype control-treated A. fumigatus -challenged mice (Fig. 7I). Further, a similar reduction was also observed in collagen accumulation upon neutralization of IL-18 in A. fumigatus -challenged mice compared to isotype control-treated A. fumigatus -challenged mice (Suppl. Fig. 5B i-iv). IL-18 neutralization does not give complete protection against disease pathogenesis, including CD274⁺ eosinophils, as these mice still show induced IL-18 in the blood and BALF (Suppl. Fig 5C, D). These data establish that anti-IL-18-treated A. fumigatus -challenged mice have restricted differentiation and lung accumulation of CD274⁺ pathogenic eosinophils, resulting in improved asthmatic characteristics including mucus production and airway obstruction, and that anti-IL-18 neutralizing antibody may be used a novel strategy to restrict asthma pathogenesis without compromising gastrointestinal innate immunity.

Figure 7. Neutralization of IL-18 ameliorates A. fumigatus -induced experimental asthma.

A flow cytometry analysis of CCR3⁺Siglec-F⁺ eosinophils and CD274⁺ eosinophils in the BALF of A. fumigatus -challenged anti-IL-18-treated and isotype control-treated mice (A,B). A quantification of flow cytometry analysis of CCR3⁺Siglec-F⁺ BALF eosinophils and CD274⁺ BALF eosinophil numbers in A. fumigatus -challenged anti-IL-18-treated and isotype control-treated mice (C,D). A representative photomicrograph of anti-MBP antibody immunostained lung sections of A. fumigatus -challenged anti-IL-18-treated and isotype control-treated mice (E). Morphometric quantification of MBP⁺ cells, expressed as cells/mm² (F). Airway resistance (RI) of all four groups of A. fumigatus - and saline-challenged anti-IL-18-treated and isotype control-treated mice was measured (G). A representative photomicrograph of goblet cell hyperplasia (H) with respective cell numbers in A. fumigatus -challenged anti-IL-18-treated and isotype control-treated mice are shown (I). Data is expressed as mean ± SD, n=8–12 mice/group. *p<0.01, **p<0.001, ***p<0.0001 (WT + A. fumigatus + anti-IL-18 vs WT + A. fumigatus + isotype control). Photomicrographs presented are 100X original magnification.

Discussion:

The burden of immune-related diseases, including asthma, is increasing globally, particularly in the Western world^53,54 Most asthma studies are focused largely on analyses of the cellular and molecular events induced by allergen exposure in sensitized animals and humans.^55-59 Studies have identified elevated production of IgE and several inflammatory cytokines that lead to lung function abnormalities.^58,60-64 Previous findings from our group and others have indicated that eosinophil recruitment into inflammatory tissue is a complex process regulated by Th2 cytokines (IL-5, IL-13, IL-15) and eosinophil-associated chemokines in asthma pathogenesis.^12,13,65,66 Asthma patients tend to have both food and aeroallergen hypersensitivity.^11-13 It is believed that asthma pathogenesis is prompted by the induction of IL-5-associated eosinophilia in response to allergen exposure. IL-5 is a well-known eosinophil differentiating and growth factor that is induced in several allergic diseases including asthma,^20,21 and its effector functions and roles in various diseases are under renewed scrutiny. ^67,68 The intrinsic molecular pathway of IL-5-regulated eosinophilia is largely unknown and remains to be defined.⁹ A recent report indicated that allergic lungs have two eosinophil subsets, CD101⁺ and CD101⁻, and that CD101-expressing eosinophils are IL-5-responsive inflammatory eosinophils.⁶⁹ However, we reported that IL-5 transgenic naïve mice have no CD101⁺ eosinophils, which is inconsistent with a recent report showing CD101⁺ eosinophils as IL-5-dependent. ⁶⁹ In addition, we found that CD2-IL-5 transgenic mice eosinophils become both CD101⁺ and CD274⁺ eosinophils kinetically only following ex vivo exposure to IL-18, and that eosinophils have two naïve and pathogenic (CD274⁺) eosinophil subsets.²² Further, we also showed that while healthy humans have blood eosinophils, high levels of CD101-expressing and CD274⁺ eosinophils are associated with allergic diseases. Notably, IL-18 is increased in a number of food and aeroallergen-induced allergic diseases, ^28-30 including asthma.^31,33,70 Several clinical reports show that IL-18 is induced in allergic patients including asthmatics,^28-33 and experimental modeling shows that IL-18 overexpression in the lung promotes the asthma phenotype in mice.^71,72 Several reports have recognized the role of IL-18 in asthma and allergic diseases; however, the significance and functional mechanism of IL-18 in eosinophil biology and in promoting allergic diseases have never been fully examined. Herein, we provide the first evidence that IL-18 is involved in differentiation of CD274⁺ eosinophils in vivo, which induces pathogenesis of the disease. Previous reports indicated that PD-1 and its ligands CD274 and CD273 play a role in allergic asthma by involving T-cell development, maintenance, function and tolerance;^34,35,73,74 however, the factor that regulate CD274⁺ eosinophils was not known. The current study illustrates the significance of IL-18 in regulating CD274 expression on inflammatory cells in asthma. Additionally, CD274^−/− mice show clearing of adenovirus infection, Leishmania mexicana-mediated infection, and reduced airway hyperactivity.^35,75,76 We also show that despite different genotypes, allergen-challenged mice have induced IL-18 levels and accumulate eosinophils in the epithelial layer around the airway, and BALF analysis indicates that these eosinophils all express CD274.²² Studies have shown increased IL-18 and IL-18R in lung tissues of patients with fatal asthma.^77,78 The role of IL-18 in IgE production and mast cell biology has previously been explored,⁷⁹ but the direct role of IL-18 and IL-18-differentiated eosinophil subsets in asthma pathogenesis has not been established. Thus, the precise mechanism of IL-18-induced pulmonary eosinophilic or non-eosinophilic inflammation is not clearly understood. Based on these reports, we hypothesize that the IL-18-differentiated CD274⁺ eosinophil subset is critical in promoting asthma pathogenesis, including mucus production and collagen deposition in the lung. In this study, we provide important evidence on the significance of IL-18-differentiated eosinophils in promoting asthma pathogenesis following rIL-18 delivery to IL-5 transgenic mice. The rIL-18-treated CD2-IL-5 transgenic mice showed most of the characteristic features observed in human asthma, like peribronchial and perivascular eosinophilia-induced accumulation of collagen, goblet cell hyperplasia, and increased airway hyperactivity. This study is the first to provide direct evidence of the critical role of IL-18-differentiated pathogenic CD274⁺ eosinophil subsets in asthma pathogenesis. We used different background strain (BALB/c and C57BL6) mice because all gene-deficient and transgenic mice were not available from one background strain. Since BALB/c, C57BL6 and littermate-matched IL5^−/−IL18^−/− are crossbred mice that show similar baseline airway hyperactivity, using both mice in different experimental setups does not pose any risk of bias in our results.

Furthermore, our experiments provide evidence on the critical role of CD274 in promoting asthma pathogenesis. A. fumigatus -challenged anti-CD274-treated and anti-IL-18-treated mice show significantly improved asthma pathogenesis, including airway hyperreactivity. Additionally, we showed clinically relevant supportive data illustrating the effectiveness of aCD274 and IL-18 neutralization in protecting induction of CD274⁺ pathogenic eosinophils and mucus-producing goblet cell hyperplasia in asthmatic mice. We show that IL-13, goblet cells, and airway hyperactivity are reduced in A. fumigatus -challenged IL-5^−/− mice compare to WT mice. Of note, IL-13, goblet cells, and airway hyperactivity reduction are similar to the reduced number of IL-18-responsive CD274⁺ eosinophils in IL-5^−/− mice. Additionally, the anti-CD274-treated mice show reduced eosinophils, IL-13, and airway hyperactivity after A. fumigatus challenge. Most importantly, this study also shows that even rIL-18-treated ΔdblGATA mice show the accumulation of CD274⁺ eosinophils, induced airway obstruction, and IL-13-associated mucus-producing goblet cells. The ΔdblGATA mice are deficient in eosinophils but have eosinophil stem cell precursors,⁵⁰ and these precursors generate mature eosinophils upon rIL-18 challenge; this is consistent with our earlier report that IL-18 is also capable of generating and maturing eosinophils from bone marrow precursors.²² Of note, this is the first report that presents an in vivo study showing direct evidence of the generation of eosinophils in response to IL-18 from eosinophil stem cell precursors in eosinophil-deficient ΔdblGATA mice. Thus, the presented data strongly support our hypothesis that IL-18 responsive CD274⁺ eosinophils are responsible for eosinophilic asthma pathogenesis, including the induction of airway obstruction. These investigations also provide a strong preclinical rationale to consider regulation of IL-18 signaling and differentiation of naïve eosinophils to CD274-expressing pathogenic eosinophils for therapeutic trials in asthmatic patients. In support of these presented data, and to establish the synergy of IL-18 and IL-5-induced eosinophilia in allergic asthma, we generated IL-5-IL-18 double knockout mice (IL-18^−/−/IL-5^−/−) that completely depleted both naïve and pathogenic blood and tissue eosinophilia and significantly reduced airway hyperactivity following the induction of A. fumigatus -induced experimental asthma. In short, the presented data support the therapeutic effectiveness of CD274 and IL-18 neutralization for eosinophilic asthma. Taken together, the current study advances our knowledge of eosinophil biology and the existence of an eosinophil subset that is critical for promoting mucus, mucus cell hyperplasia, and airway resistance. We present for the first time strong novel data that establishes the significance of induced IL-18 and its responsive CD274⁺ eosinophils in asthma pathogenesis and provide evidence for new targeted therapeutic and diagnostic interventions. First, we show allergen-challenged WT, IL-5^−/−, or IL-18^−/− mice develop CD274⁺ eosinophilic inflammation and airway obstruction in the lungs. Second, we show that rIL-18 treatment induces perivascular and peribronchial eosinophilia and increases the severity of asthma pathogenesis in CD2-IL-5 transgenic mice. Third, we show that rIL-18 treatment caused the development of lung eosinophilia-associated asthma pathogenesis, including airway obstruction, even in eosinophil-deficient ΔdblGATA mice. Fourth, we show that IL-5^−/−/IL-18^−/− double gene-deficient mice are protected from induction of CD274⁺ eosinophilic lung inflammation and airway hyperactivity, Lastly, we show that both IL-18 and CD274 neutralization protect mice from asthma pathogenesis. Based on our previous report of CD274⁺ eosinophil subsets and their presence in human allergic patients^22,80,81 and current findings, we propose anti-CD274 and anti-IL-18 clinical immunotherapy trial (with proper control placebo) to improve quality of life for asthma patients. The anti-CD274 or anti-IL-18 immunotherapy will restrict only pathogenic eosinophils and will not disturb the IL-5-responsive eosinophil-induced innate immunity in patients suffering from allergic asthma or other diseases.

Abbreviations:

ΔdblGATA

GATA1 deficiency

A. fumigatus

Aspergillus fumigatus

BALF

Bronchoalveolar lavage fluid

EPX

Eosinophil peroxidase

IL

Interleukin

IL-18^−/−

IL-18 knockout

IL-5^−/−

IL:-5 knockout

i.n.

intranasal

MBP

Major Basic Protein

MCh

Methacholine

PAS

Periodic Acid Schiff

PBS

Phosphate-buffered saline

RI

Airway resistance

Tg

Transgenic mice

WT

wild type