Histamine-releasing factor in severe asthma and rhinovirus-associated asthma exacerbation

Abstract

Background:

Histamine-releasing factor (HRF) is implicated in allergic diseases. We previously showed its pathogenic role in murine models of asthma.

Objective:

To present data analysis from three separate human samples (sera samples from asthmatic patients, nasal washings from rhinovirus (RV) infected individuals and sera samples from patients with RV-induced asthma exacerbation) and one mouse sample in order to investigate correlates of HRF function in asthma and virus-induced asthma exacerbations.

Methods:

Total IgE and HRF-reactive IgE/IgGs as well as HRF in sera from patients with mild/moderate (MA) or severe asthma (SA) and healthy controls (HC) were quantified by enzyme-linked immunosorbent assay (ELISA). HRF secretion in culture media from RV-infected BEAS-2B human bronchial epithelial cells and in nasal washings from experimentally RV-infected subjects was analyzed by western blotting. HRF-reactive IgE/IgG levels in longitudinal serum samples from patients with asthma exacerbations were also quantified.

Results:

HRF-reactive IgE and total IgE were higher in SA than in HC, whereas HRF-reactive IgGs (and IgG1) were lower in asthmatics vs. HC. In comparison to HRF-reactive IgE^low asthmatics, HRF-reactive IgE^high asthmatics had tendency to release more tryptase and prostaglandin D₂ upon anti-IgE stimulation of bronchoalveolar lavage (BAL) cells. RV infection induced HRF secretion from BEAS-2B cells, and intranasal RV infection of human subjects induced increased HRF secretion in nasal washes. Asthmatic patients had higher levels of HRF-reactive IgE at the time of asthma exacerbations associated with RV infection, compared with those after the resolution. This phenomenon was not seen in asthma exacerbations without viral infections.

Conclusion:

HRF-reactive IgE is higher in severe asthmatics. RV infection induces HRF secretion from respiratory epithelial cells both in vitro and in vivo. These results suggest the role of HRF in asthma severity and RV-induced asthma exacerbation.

Clinical Implications

HRF and HRF-reactive IgE likely play a pathogenic role in severe asthma and RV-induced asthma exacerbation. Thus, inhibition of HRF functions might ameliorate severe asthma and RV-induced exacerbations.

Capsule Summary

In keeping with previous murine airway inflammation results, HRF-IgE interactions likely play a significant role in the pathogenesis of severe asthma and RV infection-induced asthma exacerbation. Therefore, these interactions can be a preventive/therapeutic target.

Graphical Abstract

Introduction

Histamine-releasing factor (HRF), also known as translationally controlled tumor protein, fortilin, p23, p21, and Q23, is a multifunctional protein composed of 172 amino acid residues that are 96% identical between human and mouse.¹ The extracellular function of HRF is a cytokine-like activity that activates mast cells and basophils in an IgE-dependent manner,² while intracellularly it is required for cell cycle progression, proliferation, survival, malignant transformation, and DNA repair.¹ Secretion of HRF was found in bodily lavage fluids during allergic reactions, implying its role in allergic diseases including asthma.³ We previously identified a subset of IgE and IgG molecules as HRF-binding proteins ⁴: Mapping of the immunoglobulin (Ig) Fab-binding sites within the HRF molecule led to the discovery of HRF sequence-based inhibitors, N19 (the N-terminal 19 residues) and H3 (the third helical region) peptides, which blocked HRF-IgE interactions without affecting HRF’s intracellular functions. Administration of these inhibitors drastically reduced type 2 inflammation, mucus secretion, and airway hyperresponsiveness in murine models of atopic asthma.⁴ We also solved crystal structures of mouse and human HRF molecules, including a human HRF homodimer linked by a disulfide bond of the C-terminal Cys172 from two monomers.⁵ This structure supports our model of HRF dimer-mediated activation of mast cells via HRF-reactive IgE-bound FcεRI (high-affinity IgE receptors).^4,5 We recently found that HRF is secreted spontaneously from human bronchial epithelial cells and HRF proteins secreted as monomers are converted to disulphide-linked dimers and oligomers.⁶ Interestingly, the perennial aeroallergen house dust mites (HDM), cell deaths, and cell-death byproducts (ATP and adenosine) induced HRF secretion from human bronchial epithelial cells. Moreover, HDM-induced HRF secretion was dramatically enhanced by various cytokines including asthma-related epithelial-derived (IL-25, IL-33, TSLP), Th2 (IL-4, IL-5, IL-13), and proinflammatory (IL-1β, IL-6, TNF) cytokines, although the cytokines themselves had modest effects on HRF secretion. Thus, we proposed that HRF is a novel alarmin. HDM-mediated HRF secretion required the toll-like receptors, TLR2 and TLR1.⁶

Acute exacerbation is the major cause of asthma morbidity, mortality, and health-care costs. Among respiratory viruses, human rhinovirus (RV) is the most important risk factor for the development of atopic asthma.^7,8 RV infections are also associated with the majority of asthma exacerbations in pediatric patients and with loss of symptom control among adult asthmatics.^9,10 The risk for bronchoconstriction with RV infection is associated with allergic sensitization (mostly with HDM) and type 2 airway inflammation.^7,8 Clinical evidence supports the pathogenic role for IgE and mast cells in asthma and RV-induced asthma exacerbation: Mast cells are increased in the airway epithelium¹¹ and within the smooth muscle layer¹² in atopic asthma. Increased degranulated mast cells are found in the mucous glands from fatal asthma.¹³ Unlike the Th2-low group, the Th2-high group of asthmatics highly expressing Th2 cell (IL4, IL5, IL13, POSTN, CLCA1, SERPINB2) and mast cell (TPSB2, TPSAB1, CPA3) genes in the airway epithelium¹⁴ show the recruitment of mast cells to the airway epithelium from the submucosa.¹⁵ Mast cells are also recruited to the bronchial epithelium following RV infection.¹⁶ RV can replicate in and activate mast cells.¹⁷ Importantly, the humanized anti-IgE mAb omalizumab is indicated for moderate-to-severe allergic asthma^18,19 and can prevent or ameliorate RV-induced asthma exacerbation.^20,21

In light of the background literature, we identified three populations (at Pittsburgh, Boston, and Virginia) where samples of blood, BAL cells, and mucus could be analyzed to study the role of the interaction between HRF and HRF-reactive IgE in asthma and RV-induced asthma exacerbation. In this study, we have sought clinical evidence for the involvement of HRF in asthma and asthma exacerbations.

Methods

Study design

Collections of samples were identified in three separate cohorts. Each cohort’s samples provide insight about the following questions: 1) Do HRF-reactive IgE levels differ between asthmatics and healthy controls? 2) Can RV infection induce HRF secretion from respiratory epithelial cells? 3) During an exacerbation are levels of HRF and HRF-reactive IgE and IgG affected by viral infection with RV? The third question is answered twice by two separate cohorts (Boston and Virginia cohorts). For these objectives, we quantified HRF-related factors, including HRF-reactive IgE and IgG and HRF levels in blood samples and anti-IgE-mediated activation of BAL cells collected from patients with asthma and asthma exacerbation. To complement these analyses, we investigated whether RV infection induces HRF secretion from cultured human bronchial epithelial cells and in nasal mucosae from those who went through experimental RV infection. We also sought to find clinical correlates consistent with HRF’s role in asthma exacerbations due to RV infection or non-viral causes.

Human subjects

Pittsburgh participants: Subjects 18–65 years were recruited through the Severe Asthma Research Program (SARP2) University of Pittsburgh site (NIH grant HL69174), approved by the Institutional Review Board. All subjects gave informed consent. The 50 serum samples (healthy controls and asthma participants) from the Pittsburgh cohort were collected upon the participant’s enrollment into SARP2 over a 5 year period (March 2007- April 2012). Healthy controls (HC) had normal lung function, no history of chronic respiratory diseases, with or without atopy. Severe asthmatics (SA) required high-dose inhaled corticosteroid and/or oral corticosteroids for >50% of the year and met ≥2 of seven minor criteria per the American Thoracic Society (ATS) definition.²² Participants completed blood draw, exhaled FeNO measurement, and spirometry per ATS guidelines as previously described.^11,23 Participants underwent bronchoscopy where BAL cells and fluids were obtained and processed according to prior studies and the SARP manual of procedures.^11,24 In terms of sample selection, this was a convenience cohort and participant samples were selected from the parent cohort based on sample availability. Over this time frame, there were a total of 188 participants (healthy controls and asthma patients) enrolled at the Pittsburgh site; of those, 118 participants completed a bronchoscopy; following bronchoscopy 6 participants withdrew (1 healthy, 2 mild/moderate [MA] and 3 severe asthmatics). Of these 112 eligible participants, samples were selected from those who had undergone evaluation for mast cell markers in the airway lumen. Of the available 29 healthy controls. 26 mild/moderate asthmatics and 36 severe asthmatics, the total 50 samples were chosen based on those with available volume of serum for HRF analysis (18 healthy controls, 16 mild/moderate asthmatics and 16 severe asthmatics)(Table I). From the initial Pittsburgh cohort of 32 asthma participants, 21 of these asthma participants had enough available BAL cells at the time of bronchoscopy for anti-IgE stimulation experiments (Fig. E1 in this article’s Online Repository at www.jacionline.org). BAL cells from 21 asthma participants were suspended in X-Vivo media (Lonza) with or without stimulation with polyclonal rabbit anti-human IgE antibody (Dako) for 20 minutes before collection. BAL cell supernatant measurements for tryptase and PGD2 were determined by ELISA: Tryptase was measured by Dr. L.B. Schwartz’s laboratory (VCU, VA) and PGD2 measured by Elisatech (Denver, CO) as previously described.^11,25 The BAL cell pellet was collected, and quantitative reverse transcription-polymerase chain reaction was performed for tryptase and FceRIa mRNAs (indexed to GAPDH). For each participant, the ELISA measurements were indexed to baseline BAL cell tryptase mRNA to normalize for varying mast cell numbers in the BAL cell population.

Table I.

Baseline demographic characteristics of Pittsburgh participants

Demographic	Healthy controls (n=18)	Mild/Moderate asthma (n=16)	Severe asthma (n=16)	Overall p value
Age, yr*	27 (24–35)	28 (22–34)	38 (33–49)	0.004, ¥ɸ
Sex, male/female	9/9	5/11	4/12	0.28
Race, AA/white/other	2/12/4	2/10/4	0/13/3	0.65
BMI, kg/m2*	23 (22–26)	26 (24–33)	32 (27–37)	0.0005, ¥ɸ
Blood Eosinophils,/uL*	100 (100–200)	200 (100–3000	300 (100–500)	0.013, ¥Δ
Exhaled NO, ppb*	21 (13–46)	28 (15–40)	33 (30–46)	0.18
Baseline FEV1%p*	96 (92–105)	93 (83–107)	59 (51–73)	<0.0001, ¥ɸ
Use of ICS, n (%)	n/a	10 (63)	16 (100) [12 on OCS]	0.007

^*Data presented as medians (25th–75th percentiles), analysed by Wilcoxon/ANOVA; Categorical variable analysed by Chi-square. Significant intergroup comparisons as follows:

^¥:SA vs. HC

^ɸ:SA vs. MA

^:Δ:MA vs. HC; Abbreviations: AA, African American; BMI, body mass index; ICS= inhaled corticosteroids, NO= nitric oxide, OCS= oral corticosteroids

Boston participants: Subjects aged 6–17 years with physician-diagnosed asthma were enrolled at the time of acute asthma exacerbation and followed longitudinally. Subjects were consecutively enrolled based on availability of study personnel to approach for informed consent. The diagnosis of asthma exacerbation was based on clinical features and response to treatment.²⁶ Exclusion criteria were as follows: radiographic pneumonia, other pulmonary or structural airway diseases, prematurity, or immunodeficiency. At enrollment, the following data were collected: validated respiratory disease questionnaire data,²⁷ physical exam, peak expiratory flow, nasal swabs, serum IgE levels. Subjects returned for an outpatient visit after 6 weeks (or when they returned to symptomatic baseline) and the following data were collected: questionnaire data, pulmonary function testing, and serum IgE levels. The Boston Children’s Hospital IRB approved this study (Study ID# P00000084). Written informed consent was obtained prior to participation. Blood samples were collected at the time of enrollment and the second visit. PCR-based detection of rhinovirus/enterovirus from nasal swabs was performed by Viracor-IBT (Lee’s Summit, MO) as part of the respiratory viral panel which detects the following: human metapneumovirus, rhinovirus, influenza A, influenza A subtype H1 and H3, influenza B, respiratory syncytial virus A and B, parainfluenza virus 1 – 3, and adenovirus. Samples for this study were chosen consecutively based on meeting study inclusion criteria as well as availability of sufficient quantity of serum to perform the analysis. Serum samples from 10 RV-infected asthmatics and 10 non-virally infected asthmatics were provided for this study (Table 2).

Table 2.

Baseline demographic characteristics of Boston participants

Demographic	RV-infected asthmatics (n=10)	Non-virally infected asthmatics (n=10)	p values
Age, yr	9.8 ± 3.2	9.5 ± 3.1	0.85
Sex (% male)	70	60	0.661
Race, AA/white/other	5/4/1	4/4/2	ns
HDM IgE (kU/L)	13.67 ± 21.85	11.84 ± 18.67	0.842
HDM exposed (% total)	62.5	60	0.532

Virginia participants: All subjects enrolled were adults (ages 19 to 39 years). The asthmatics had physician-diagnosed mild asthma treated intermittently with inhaled bronchodilators. Subjects eligible for participation based on screening were enrolled in a clinical trial which included 11 asthmatics and 13 non-allergic, non-asthmatic healthy controls (NCT02111772). The atopic status of the asthmatics was consistently high (total IgE values ranging from 726–2128 IU/ml with ≥ 4 positive skin tests to aeroallergens, including HDM). RV-16 virus inoculation was performed in the evening on day 0, after which each subject was isolated in a hotel room for the first 4 days of the infection. Procedures and assessments were done at visits scheduled before rhinovirus inoculation; daily in the morning in the hotel; and during clinic visits scheduled on days 7, 14, and 21. Institutional review board approval and informed consent for screening were obtained at the University of Virginia and Virginia Commonwealth University, and also for the enrollment and evaluation of subjects who participated in this trial at the University of Virginia. The original study was described.²⁸ Nasal washes from 6 mild asthmatics and 6 healthy controls were provided for this study (Table 3).

Table 3.

Baseline demographic characteristics of Virginia participants

Demographic	Mild asthmatics (n=6)	Healthy controls (n=6)	p values
Mean age, yr (range)	23.8 (19–38)	22.5 (20–27)	0.48
Sex, male/female	83	83	ns
Race, AA/white/other	1/5/0	0/6/0	ns
Total serum IgE; IU/ml (intraquartile range)†	1456 (749–2239)	17 (8–22)	< 0.001
# positive skin tests (range)	4–9	0

^†Total IgE results are median values with intra-quartile ranges in parentheses.

Assays for IgE, IgG, and IgG1 in human sera

Total IgE, IgGs, and IgG1 in serum were analyzed by commercial ELISA kits (eBioscience). HRF-reactive IgE and HRF-reactive IgG were analyzed by in-house ELISAs, as previously described.²⁹ Briefly, ELISA wells were coated with recombinant human HRF-His₆ overnight. After washings, the wells were blocked with ImmunoBlock (DS Pharma Biomedical, Japan, Cat# CTKN001) for 2 h. The wells were washed, and incubated with biotin anti-human IgE or anti-human IgG (or IgG1) for 1 h. Then the wells were washed and incubated with streptavidin-β-Gal conjugate, followed washings and incubation with 0.2 mM 4-MU-Gal for 1 or 2 h. Fluorescence was measured at excitation of 365 nm and emission of 445 nm. Biotin anti-human IgE, IgG, and IgG1 antibodies were purchased from BD Biosciences.

Measurements of serum HRF

Nunc immunosorp2 plates (clear, flat, high binding) were coated with anti-TPT1 antibody (clone 2C4, Abnova) at 10 μg/ml. Serum samples diluted 1:5 in 1% BSA in PBS were incubated in the plates. Standards were 100 ng/mL and serial 1:2 dilutions of human HRF-His. After overnight incubation and washes with 1% BSA in PBS, HRF bound to immobilized antibody was detected with 300 ng/ml anti-TCTP (clone D10F2 from Cell signaling) and then followed by HRP-conjugated anti-Rabbit IgG (Cell Signaling). Color development was done with TMB substrate (BD Bioscience).

Culture of BEAS-2B human bronchial epithelial cells

BEAS-2B cells were cultured in DMEM supplemented with 10% FCS, 100 μg/ml penicillin and 100 U/ml streptomycin at 37°C in a humidified incubator with 5% CO₂. Cells were incubated overnight at 5 × 10⁵ cells/well in a 6-well plate or 1.25 × 10⁵ cells/well in a 24-well plate. The culture medium was replaced with fresh medium containing 0.1% FCS, and cells were cultured for an additional 4 h for starvation. Then, cells were stimulated with HDM allergen or RV-1B at a multiplicity of infection of 0.1 or combinations of both agents. Culture supernatants were analyzed by SDS-PAGE and followed by anti-TPT1 antibody (clone 2C4, Abnova). Horseradish peroxidase-linked anti-mouse IgG Ab was used as secondary antibody.

Mouse experiments

Mice were inoculated in the foot pad with 5 × 10⁴ PFU MCMV (Smith strain generated in cultured fibroblasts),³⁰ and were serially bled at days 1, 7, 14, 21, and 28. HRF-reactive IgE, IgG1, and IgG2b in sera were measured by ELISA.²⁹ Animal experiments were approved by the Animal Care and Use Committee of the La Jolla Institute for Immunology.

Statistical analysis

Data are presented as mean ± SEM in all figure parts in which error bars are shown. As normal distribution was not found by Shapiro-Wilk W test, Mann-Whitney U test was used to compare HRF-reactive IgE and IgG levels and ratios of HRF-reactive IgE to HRF-reactive IgG between two groups in the Pittsburgh cohort and significant differences were also confirmed by Kruskal-Wallis test. Numbers of BAL cells available to anti-IgE stimulation varied drastically unlike serum samples. Thus, 21 asthmatics with BAL cells enough for anti-IgE stimulation were divided into two groups, HRF-reactive IgE^high and HRF-reactive IgE^low groups, according to the median split of HRF-reactive IgE levels. The two groups were compared by Student’s t-test. In the Boston cohort, Wilcoxon Matched Pairs test was used to compare HRF-reactive IgE/IgG levels within RV⁺ or virus-negative groups. Comparison between RV⁺ vs. virus-negative groups was made using Mann-Whitney U test. HRF-reactive IgE and IgG levels in mouse experiments were analyzed as described above. p < 0.05 was used as a significant difference.

Results

Patients with severe asthma have higher serum levels of HRF-reactive IgE than healthy controls and asthmatics have lower serum levels of HRF-reactive IgG than healthy controls

We quantified HRF and HRF-reactive Igs in sera from severe asthmatics (SA), mild/moderate asthmatics (MA), and healthy controls (HC) of the University of Pittsburgh cohort (Table I). Results indicate that total IgE and HRF-reactive IgE were higher in SA than in HC (Fig. 1A,B), whereas HRF levels of all serum samples were less than 6.25 ng/ml, the detection limit. While MA patients showed numerically intermediate levels of total IgE and HRF-reactive IgE between SA and HC, there was not a statistically significant difference between MA vs. SA (Fig. 1A,B). In contrast, HRF-reactive IgGs (total IgG and IgG1) were lower in asthmatics than in HC (Fig. 1C and data not shown), consistent with the negative regulatory effects of IgG on airway inflammation³¹ and sequential class switch recombination from IgM through IgG to IgE.³² Accordingly, ratios of HRF-reactive IgE to HRF-reactive IgG were higher in asthmatics vs. HC (Fig. 1D). HRF-reactive IgE or HRF-reactive IgGs were not correlated with clinical factors including atopy status, eosinophil levels, lung function (FEV₁), systemic or inhaled use of corticosteroids, and health care utilization (data not shown). In respiratory epithelial cells mRNA levels from selected chemokines and cytokines (Eotaxin-2/CCL24; Eotaxin-3/CCL26; IP10/CXCL10, MIG/CXCL9) did not correlate with HRF IgE nor HRF IgG (Table E1 in this article’s Online Repository at www.jacionline.org). Next, in this cohort, 21 asthmatics with anti-IgE stimulation data were divided into two groups, HRF-reactive IgE^high and HRF-reactive IgE^low groups, according to the median split of HRF-reactive IgE levels. These two groups showed no significant differences in mRNA levels of tryptase and FcεRIα in BAL cells (data not shown). Immunostaining of the BAL cell cytospins for tryptase did identify tryptase positive cells in both HRF-reactive IgE^high and IgE^low asthmatics; there was no difference in the percentage of tryptase positive cells (median of 3.59 [IQR 1.2–4.99] for HRF-reactive IgE^high vs. 1.99 [IQR 1.35–3.69] for HRF-reactive IgE^low (p=0.53) (Fig. E2 in this article’s Online Repository at www.jacionline.org). However, in comparison to HRF-reactive IgE^low asthmatics (n=11), HRF-reactive IgE^high asthmatics (n=10) had tendency to release more tryptase (1.40 fold, p=0.0882) and more prostaglandin D₂ (PGD₂) (1.81 fold, p=0.0445) upon anti-IgE stimulation of BAL cells. These results suggest that mast cells in the BAL cells of these patients are more easily activated by anti-IgE antibody. Therefore, HRF-reactive IgE likely contributes to or reflects increased asthma severity.

Fig. 1.

Increased HRF-reactive IgE in severe asthmatics and reduced HRF-reactive IgG in asthmatics. (A-D) Total IgE and HRF-reactive IgE and IgG in sera were quantified by ELISA. (D) Ratio of HRF-reactive IgE to HRF-reactive IgG. HC, healthy control (n=18); MA, mild/moderate asthma (n=16); SA, severe asthma (n=16). P values by Mann-Whitney U test (A-D).

RV infection induces HRF secretion from cultured human bronchial epithelial cells and from nasal washes in experimentally RV-infected individuals

As shown previously, BEAS-2B cells secreted HRF spontaneously, and HDM increased its secretion.⁶ RV infection of BEAS-2B cells also induced the secretion of HRF (Fig. 2A), which formed oligomers by disulfide bonds. Importantly, HRF oligomers induced by HDM plus RV or HDM stimulation followed by RV infection were much more abundant than those induced by each alone (Fig. 2A). HDM plus RV also induced enhanced levels of inflammatory cytokines and chemokines including IL-6, IL-8, CXCL1, CXCL2, and CCL5 (data not shown), as shown previously.³³ Given the enhanced secretion of HRF by RV, particularly by HDM plus RV, we tested whether HRF secretion is influenced by experimental RV infection in human subjects (Table 3). Amounts of HRF oligomers secreted into nasal mucosa in the subjects with mild asthma exhibited their peak amounts at day 4 post-infection and mostly returned to baseline levels by day 21 postinfection, while some healthy participants reacted more slowly to RV infection (Fig. 2B). These results suggest that RV infection induces HRF secretion from respiratory epithelial cells in vitro and in vivo.

Fig. 2.

HRF secretion is induced by RV in vitro and in vivo. (A) BEAS-2B cells were stimulated with HDM (100 μg/ml) or RV-1B (0.1 MOI) for 88 h. HDM>RV, cells stimulated with HDM for 72 h, then incubated with RV for another 16 h; HDM+RV, cells stimulated with HDM for 72 h and with HDM plus RV for the last 16 h. Secreted HRF was analyzed by SDS-PAGE (10% polyacrylamide gel) under nonreducing (upper panel) or reducing (lower panel) conditions and western blotting. 1*, HRF monomer; 2*, dimer; HMW-HRFs, high molecular weight HRFs (Mr > 148 kD); med, medium; US, unstimulated. (B) Mild asthmatics (19–39 years in age) and normal subjects (20–27 years) were inoculated with RV-16 on day 0. Nasal washes were analyzed by SDS-PAGE (4–12% gradient Tris-glycine SDS gel) under nonreducing conditions and western blotting for HRF. Note that the use of a gradient gel allowed a better resolution in the high molecular weight range. n*, HRF n-mer.

HRF-reactive IgE levels are higher during RV infection-associated asthma exacerbation, but not non-viral exacerbation, in pediatric patients

In light of RV infection-induced HRF secretion in respiratory epithelial cells, we tested if there would be any changes in HRF-reactive IgE levels in asthma exacerbation. Interestingly, we found that all pediatric patients with asthma tested at Boston Children’s Hospital (Table 2) had higher levels of HRF-reactive IgE at the time of asthma exacerbation associated with RV infection, compared with those after the resolution (Fig. 3A,C). In contrast, this phenomenon was not observed with non-virally induced exacerbations (Fig. 3B). Moreover, HRF-reactive IgE levels at the exacerbation were more than 3 times higher in RV infection-associated exacerbations than in non-virally associated exacerbations (Fig. 3C). HRF-reactive IgG levels did not change significantly between the exacerbation and after-resolution in RV infection-associated and non-virally associated exacerbations (Fig. 3D). HRF levels of these serum samples were less than 6.25 ng/ml. These data suggest that RV infection induces HRF secretion from respiratory epithelial cells and that HRF-reactive IgE, together with HRF, whose concentrations would be locally high, may exert exacerbating effects in RV-induced asthmatic patients. In contrast, HRF does not seem to significantly contribute to non-virally induced asthma exacerbations. In these cases, exposure to high levels of allergen or other stimulants may be more important.

Fig. 3.

Increased HRF-reactive IgE during RV-induced asthma exacerbations. Sera from pediatric asthma patients during exacerbation (Exac) with RV infection (A, n=10) or non-viral causes (B, n=10) and after resolution (Resol)³⁴ were subjected to ELISA to quantify HRF-reactive IgE. C,D Comparisons of HRF-reactive IgE (C) and HRF-reactive IgG (D) within RV+ or virus- groups were made using Wilcoxon Matched Pairs test. Comparison between RV+ vs. virus- samples was made by Mann-Whitney U test. Data are presented as mean ± SEM.

HRF-reactive IgE levels are higher during mouse cytomegalovirus (MCMV) infection in mice

Given the increased serum HRF-reactive IgE in RV-infected individuals, we tested whether another virus infection has the similar effect in mice. Low serum levels of HRF (mostly <1 ng/ml) were also observed in mice infected with MCMV. However, significant induction of not only HRF-reactive IgE but also HRF-reactive IgG1/IgG2b was observed (Fig. E3 in this article’s Online Repository at www.jacionline.org), consistent with our recent data that cell deaths cause HRF release from epithelial cells.⁶

Discussion

In our previous study, HRF inhibitors, GST-N19 and GST-H3, that blocked HRF-IgE interactions, reduced type 2 inflammation, mucus secretion, and airway hyperresponsiveness in murine models of atopic asthma.⁴ Our clinical data in the present study support for the role of the interaction between HRF and HRF-reactive IgE in severe asthma and RV-induced asthma exacerbation: 1) serum levels of HRF-reactive IgE were higher in patients with severe asthma than in healthy controls and patients with mild/moderate asthma had intermediate levels. 2) RV infection induced HRF secretion in nasal mucosa. 3) HRF-reactive IgE levels were higher during RV-induced asthma exacerbation than after resolution of the acute symptoms. In contrast, this phenomenon was not observed in non-viral exacerbation. 4) HRF-reactive IgE levels were higher during RV-induced asthma exacerbation than during non-viral asthma exacerbation. In vitro experiments potentially connect the above clinical data: RV infection together with HDM enhances HRF secretion in BEAS-2B cells, suggesting that RV infection will strongly enhance HRF secretion from HDM-exposed asthmatics. Since a large majority of asthmatics are sensitized with HDM,³⁴ a subset of HDM-specific IgE antibodies recognize HRF (data not shown), and these IgE antibodies will promote mast cell activation, eventually leading to more severe disease. Although the negative regulatory effects of IgG were reported on airway inflammation,³¹ we do not know the function of HRF-reactive IgG. HRF-reactive IgG might have both activating and inhibitory functions, as Fcγ receptors have both activating and inhibitory receptors, unlike FcεRI that are exclusively activating. Comparison in function between HRF-reactive IgE and HRF-reactive IgG will be potentially an important issue in the future. Alternatively, reduced HRF-reactive IgG in asthmatics might simply reflect the sequential class switching from IgM through IgG to IgE.

Potential limitations of this study are due to small sample size or potential bias in selection of samples. However, smaller sample sizes for BAL cell data (Pittsburgh cohort) were simply due to limited BAL cell numbers available for anti-IgE stimulation. Our clinical data are based on three cohorts, two based on young adults and one on children. At this point, we do not know if the data in Pittsburgh adult participants can be extrapolated to pediatric patients or if the data in Boston pediatric participants can be replicated in adult asthmatics. There are no data on the level of HRF or HRF-reactive IgE/IgG in different ages, genders, races, and other demographic traits. There could also be differences in the role or extent of HRF involvement in variant asthma phenotypes or endotypes. Our previous animal experiments suggest that this may be the case, as only IgE/mast cell-dependent animal models of asthma can be suppressed by HRF inhibitors.⁴

Collectively, our observations suggest that HRF-reactive IgE antibodies contribute to increased severity of asthma and that, together with allergens (such as HDM) and HRF oligomers secreted by respiratory epithelial cells in response to RV, HRF-reactive IgE may activate mast cells during RV-induced asthma exacerbation. Both severe asthma and RV-induced asthma exacerbation can be prevented or ameliorated with the humanized anti-IgE monoclonal antibody omalizumab.^20,35 Omalizumab binds to free IgE, lowers free IgE levels, and downregulates FcεRI on mast cells and basophils.³⁶ In addition to this main mechanism, omalizumab at high concentrations can disrupt IgE-FcεRI complexes. In contrast, HRF inhibitors likely act by blocking HRF-IgE interactions. Therefore, it will be interesting to compare the effects of these two types of anti-mast cell agents side-by-side in animal models of asthma and RV-induced asthma exacerbation.

Abbreviations used

ATS

American Thoracic Society

BAL

bronchoalveolar lavage

BEAS-2B cell

adenovirus-12 SV40 hybrid virus transformed bronchial epithelial cell

BEC

bronchial epithelial cell

ELISA

enzyme-linked immunosorbent assay

HC

healthy controls

HDM

house dust mites

HRF

histamine-releasing factor

Ig

Immunoglobulin

MA

mild/moderate asthma

MCMV

mouse cytomegalovirus

RV

rhinovirus

SA

severe asthma

SARP

Severe Asthma Research Program