|Home | About | Journals | Submit | Contact Us | Français|
Previously, we found that mast cell tryptases and carboxypeptidase A3 (CPA3) are differentially expressed in the airway epithelium in asthmatic subjects. We also found that asthmatic subjects can be divided into 2 subgroups (“TH2 high” and “TH2 low” asthma) based on epithelial cell gene signatures for the activity of TH2 cytokines.
We sought to characterize intraepithelial mast cells (IEMCs) in asthma.
We performed gene expression profiling in epithelial brushings and stereology-based quantification of mast cell numbers in endobronchial biopsy specimens from healthy control and asthmatic subjects before and after treatment with inhaled corticosteroids (ICSs). We also performed gene expression and protein quantification studies in cultured airway epithelial cells and mast cells.
By means of unsupervised clustering, mast cell gene expression in the airway epithelium related closely to the expression of IL-13 signature genes. The levels of expression of mast cell genes correlate positively with lung function improvements with ICSs. IEMC density was 2-fold higher than normal in subjects with TH2-high asthma compared with that seen in subjects with TH2-low asthma or healthy control subjects (P = .015 for both comparisons), and these cells were characterized by expression of tryptases and CPA3 but not chymase. IL-13 induced expression of stem cell factor in cultured airway epithelial cells, and mast cells exposed to conditioned media from IL-13–activated epithelial cells showed downregulation of chymase but no change in tryptase or CPA3 expression.
IEMC numbers are increased in subjects with TH2-high asthma, have an unusual protease phenotype (tryptase and CPA3 high and chymase low), and predict responsiveness to ICSs. IL-13–stimulated production of stem cell factor by epithelial cells potentially explains mast cell accumulation in TH2-high asthmatic epithelium.
Mast cells play key roles in host defense, homeostasis, tissue repair, and mechanisms of allergic inflammation.1 Two major human mast cell phenotypes have been described2: (1) mast cell–tryptase (MC-T), or mucosal mast cells found in the lung and gastrointestinal tract, and (2) mast cell–tryptase/chymase (MC-TC), or connective tissue mast cells found predominantly in the skin, peritoneum, and joints. MC-T cells predominantly express high levels of tryptases but not chymase or carboxypeptidases, whereas MC-TC cells express all 3 proteases.3 In lung tissues MC-T cells usually predominate, but smooth muscle, submucosal glands, blood vessels, and pleura are enriched in MC-TC cells.4,5
In recent genome-wide expression studies of cells obtained by means of epithelial brushing during research bronchoscopy in asthmatic and control subjects, we found that 2 mast cell genes (tryptase and carboxypeptidase A3 [CPA3]) are among the top 10 most differentially expressed genes in asthmatic subjects. We also found that 3 other genes (periostin, chloride channel accessory 1 [CLCA1], and serpin B2) among the top 10 most differentially expressed genes are regulated by IL-13.6 We subsequently proposed these 3 genes as a signature for IL-13–induced activation of airway epithelial cells.7 Surprisingly, when we classified asthmatic subjects based on expression of IL-13–inducible genes in epithelial brushings, we identified 2 evenly sized and distinct subgroups: one that had an IL-13 signature and another that did not.7 Our approach was unsupervised hierarchical clustering based on the microarray expression levels of periostin, CLCA1, and serpin B2. These 2 subgroups differed significantly in gene expression for IL-5 and IL-13 in homogenates of bronchial biopsy specimens, and we therefore named the 2 groups “TH2 high” and “TH2 low.”
Our findings for increases in expression of mast cell proteases in airway epithelial brushings in an unbiased gene-profiling study in asthmatic subjects suggest that mast cells infiltrate the airway epithelium in asthmatic subjects. Increases in intraepithelial mast cell (IEMC) numbers in asthmatic subjects have been shown in some prior studies8,9 but not in others,10,11 and all of these studies have been limited by relatively small numbers of subjects, imprecise methods of quantification, and confounding by inhaled corticosteroid (ICS) use. Therefore uncertainty remains about whether IEMC numbers are increased in asthmatic subjects, and little is known about the phenotype of this mast cell subset in vivo. In addition, it is unknown whether TH2 status influences mast cell accumulation in the airway epithelium. Here we set out to use rigorous stereology-based methods in a relatively large number of steroid-naive asthmatic subjects to determine whether the numbers of IEMCs are increased in asthmatic subjects and to determine the influence of TH2 status on these numbers. We also used PCR-based gene profiling and immunohistology to determine the expression of tryptases, CPA3, and chymase in IEMCs in asthmatic subjects.
We studied biological samples from the Airway Tissue Bank at the University of California, San Francisco (UCSF). These samples had been collected during research bronchoscopy in healthy and asthmatic volunteers. Bronchoscopy included collection of epithelial brushings and bronchial biopsy specimens by using methods previously described.6,12–14 For this study, we reviewed mast cell gene expression data from our previously generated epithelial microarray and quantitative RT-PCR (qPCR) data6 and then generated new data by using samples from the tissue bank.
We studied 42 steroid-naive subjects with mild-to-moderate asthma and 36 healthy nonasthmatic control subjects (Table I). Thirty-two of the asthmatic subjects were also enrolled in a double-blind randomized controlled clinical trial of inhaled fluticasone (500 μg twice daily) or matched placebo (ClinicalTrials.gov Identifier: NCT00187499).
Microarray analyses on epithelial brushings had been performed as described previously.7 These data are available in MIAME-compliant format at Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/, accession no. GSE4302). Real-time RT-PCR (qPCR) analyses for tryptases, CPA3, and chymase were performed with RNA from epithelial brushings by using methods previously described6 and with primers and probes listed in Table E1 (available in this article’s Online Repository at www.jacionline.org). Summary data for the tryptase and CPA3 PCR results presented here were published previously as validation of gene expression data obtained by using microarrays.6
Four to 6 bronchial biopsy specimens had been embedded in paraffin with isector molds and regular paraffin molds, as previously described.13 Three serial sections, 3 μm thick, were cut, and tryptase immunostaining was used to identify mast cells (see below). We used the Computer Assisted Stereology Toolbox (C.A.S.T.) grid system (Olympus, Albertslund, Denmark) to measure the numeric density of mast cells using the physical dissector method (full details are available in the Methods section of this article’s Online Repository at www.jacionline.org).14,15
The antibody reagents used were as follows: (1) tryptase, AA1, mouse monoclonal anti-human mast cell tryptase (Thermo Scientific, Fremont, Calif); (2) chymase, CC1 mouse monoclonal anti-human mast cell chymase (ABD Serotec, Oxford, United Kingdom); (3) CPA3, rabbit anti-human mast cell CPA3 (Sigma, St Louis, Mo); and (4) basophils, BB1 mouse mAb against basogranulin (a generous gift from Dr Andrew Walls). Three-micrometer sections of formalin-fixed and paraffin-embedded lung sections from a patient who died from asthma were used to test and optimize the mast cell protease antibodies.
Normal human bronchial epithelial cells (Clonetics, San Diego, Calif) were cultured in Grey media and seeded onto 12-well transwell inserts, as described in the Methods section of this article’s Online Repository. Cells were then grown at the air-liquid interface with the addition of cytokines to the basal media for 4 days: control (Grey media alone), IL-13 (10 ng/mL), TNF-α (10 ng/mL), or IL-1β (10 ng/mL). On day 4, cells and media were collected for analysis (full details are available in the Methods section of this article’s Online Repository).
Mast cells were derived from umbilical cord blood mononuclear cells (as described in the Methods section of this article’s Online Repository). The cells were exposed to conditioned media (CM) from human bronchial epithelial cells that had been activated with IL-13 (10 ng/mL on days 0 and 2) or control (no cytokine). Mast cells were cultured with the epithelial cell CM for 2 days and processed for immunostaining and qPCR.16
Values are presented as means ± SDs or medians (interquartile ranges) unless otherwise specified. Correlation was performed with the Pearson correlation. Nonparametric 2-group comparisons were made with the Wilcoxon signed-rank test, and multiple group comparisons were made with Kruskal-Wallis 1-way ANOVA. A 2-tailed P value of less than .05 was taken as statistically significant. All statistical analyses were performed with STATA SE 10.1 software (StataCorp, College Station, Tex). Unsupervised hierarchical clustering of gene expression data with Pearson correlation as a distance metric was performed with the R package hclust (see the Methods section of this article’s Online Repository).
We previously provided summary data for gene expression for tryptase and CPA3 in epithelial brushings from 42 steroid-naive subjects with asthma and 28 healthy control subjects.6 These summary data showed that tryptase and CPA3 expression was higher than normal in asthmatic subjects. Here we show that each of these 2 genes is overexpressed in some, but not all, of the 42 asthmatic subjects (Fig 1). Interestingly, using microarray- or qPCR-based gene expression profiling, we found no difference in chymase expression in airway epithelial brushings in asthmatic subjects (Fig 1). Mast cells expressing both tryptase and CPA3 traditionally are of the MC-TC phenotype and also coexpress chymase. This was not the case here, suggesting a unique pattern of protease gene expression by IEMCs in asthmatic subjects. Notably, the ratio of CPA3 to tryptase was similar in the asthmatic and healthy subgroups (0.46 vs 0.39, P = .13), indicating that the IEMC phenotype of increased CPA3 expression is not asthma specific but rather indicative of an intraepithelial cell phenotype, of which there is an accumulation in asthmatic subjects. We did observe a trend for an increased CPA3/tryptase ratio among subjects with TH2-high asthma, suggesting that CPA3 expression is especially increased in IEMCs from subjects with TH2-high asthma (see Fig E1 in this article’s Online Repository at www.jacionline.org).
To gain insight into potential mechanisms of mast cell accumulation within the airway epithelium, we examined the relationships between expression levels of the 2 mast cell genes and other genes in the microarray expression data from epithelial brushings among the 42 asthmatic and 28 healthy subjects. Specifically, we performed unsupervised hierarchical clustering analysis to classify genes based on overall gene expression without regard to subject category labels. The hierarchical tree yielded 40 gene clusters, the largest of which contained 40 genes. This cluster includes several mast cell protease genes: tryptases TPSAB1 and TPSB2, CPA3, and the mast cell surface receptor c-kit (KIT), which is a transmembrane receptor for the mast cell growth stimulant and attractant stem cell factor (SCF; otherwise known as kit ligand; Fig 2). Notably, in the same “mast cell cluster” we found periostin, serpin B2, and CLCA1, the 3 genes we have previously proposed as an epithelial gene signature for TH2-high asthma. Tryptase gene expression was found to strongly correlate with each of these TH2 signature genes (see Fig E2 in this article’s Online Repository at www.jacionline.org). Although SCF was not specifically found in this cluster, we did find significantly higher epithelial SCF gene expression among subjects with TH2-high asthma compared with that seen in subjects with TH2-low asthma and healthy control subjects (see Fig E3 in this article’s Online Repository at www.jacionline.org). These findings suggest that upregulation of mast cell gene expression in the airway epithelium in asthmatic subjects occurs specifically in the TH2-high subgroup.
A subset of the 42 asthmatic subjects (n = 32) had been enrolled in a clinical trial in which they had been randomized to either inhaled fluticasone, 500 μg twice daily, or inhaled placebo for 8 weeks after bronchoscopy. Here we examined the relationship between baseline gene expression levels for tryptase and CPA3 in epithelial brushings and fluticasone-induced improvement in lung function. In fluticasone-treated asthmatic subjects we found significant positive correlations between baseline expression of tryptase in epithelial brushings and the change in FEV1 at 4 weeks (r = 0.53, P =.03) and between baseline expression of CPA3 and the change in FEV1 at 4 weeks (r = 0.48, P < .05; Fig 3). This analysis shows that asthmatic subjects with higher baseline expression levels for mast cell genes in their epithelium have greater improvements in lung function, suggesting that some of the clinical improvement from ICSs could be a result of decreased activation of IEMCs.
We examined a subset of samples from the UCSF Airway Tissue Bank with adequate intact epithelium (26 asthmatic 23 healthy control subjects) and quantified IEMC density using design-based stereology. We found that the number of IEMCs was higher than normal in asthmatic subjects (Fig 4, A); when stratified by TH2 subgroup, the number of IEMCs was significantly higher in the subjects with TH2-high asthma than the subjects with TH2-low asthma, and the TH2-low subgroup did not differ significantly from normal values (Fig 4, B).
Basophils express tryptase, albeit at lower expression levels than mast cells, and basophils are increased in sputum from asthmatic subjects17 and increase after allergen challenge.18 To confirm that the tryptase gene expression in epithelial brushings in asthmatic subjects derives from mast cells and not basophils, we immunostained serial sections of airway biopsy specimens for basophils by using a basophil-specific antibody (BB1).19 We found that basophil staining was sparse or absent in the airway epithelium and that tryptase-positive cells within the epithelium did not costain with the basophil antibody (Fig 5, A and B). The positive control slides (fatal asthma) showed that the BB1 antibody worked to show basophil staining (see Fig E4 in this article’s Online Repository at www.jacionline.org). We conclude that tryptase-positive cells in the airway epithelium in asthmatic subjects are mast cells and not basophils.
We immunostained serial sections of airway mucosal tissue biopsy specimens from asthmatic subjects for tryptase, CPA3, and chymase. We consistently found that IEMCs expressed tryptase and CPA3 but not chymase (Fig 5). We observed chymase-positive staining in submucosal mast cells and mast cells localized to smooth muscle cells, but we were unable to detect chymase staining in IEMCs. These data confirm the results of gene expression by means of microarray and qPCR from epithelial brushings and suggest a unique pattern of protease expression among IEMCs.
As noted above, we identified a mast cell gene cluster that included several mast cell protease genes and c-kit, an important cell-surface receptor on mast cells. The major ligand for c-kit is SCF, which is an important growth factor for mast cells and is expressed by airway epithelial cells.20,21 Given our finding for an increase in IEMCs in subjects with TH2-high asthma, we hypothesized that IL-13 stimulates SCF production by epithelial cells and therefore promotes intraepithelial accumulation of mast cells. Using an in vitro model of primary bronchial epithelial cells grown at an air-liquid interface and stimulated with IL-13, we found that IL-13 induced a significant increase in SCF gene expression (Fig 6, A). In addition, we found that IL-13 induced secretion of SCF protein into the basal medium underneath the epithelial cells (Fig 6, B), a finding that provides a mechanism for how epithelial secretion of SCF could promote migration of mast cells from the submucosa to the epithelium.
Epithelial cells can promote and support the growth of mast cells in vitro.21 Given the unique protease phenotype of IEMCs (chymase low and CPA3/tryptase high) and their accumulation in subjects with TH2-high asthma, we hypothesized that CM from epithelial cells treated with IL-13 would promote changes in IEMC gene expression consistent with this phenotype. We cultured cord blood–derived mast cells in the presence of CM from epithelial cells treated with IL-13–supplemented or control media and analyzed gene and protein expression (Fig 7). We found that CM from IL-13–treated epithelial cells significantly downregulated chymase gene expression by cord blood–derived cultured mast cells, whereas there was no significant change in expression of tryptase or CPA3. (Of note, there were no significant changes in gene expression with IL-13 alone [data not shown].) Mast cell chymase is packaged within granules, and therefore the effect of IL-13 we observe here might be to cause degradation of granules with chymase and formation of new granules lacking chymase. This is consistent with our immunostaining of IEMCs in vivo and suggests a direct role for the epithelium under the influence of TH2 cytokines to directly modify the phenotype of IEMCs after their recruitment into the epithelium.
We found that the numbers of mast cells in the airway epithelium were increased in some, but not all, of the 26 asthmatic subjects we studied. By classifying the asthmatic subjects as having TH2-high or TH2-low asthma based on expression of IL-13–responsive genes in their epithelial brushings, we discovered that IEMC numbers in the TH2-high subgroup were significantly higher than those in the TH2-low subgroup. In fact, the numbers of IEMCs in the TH2-low subgroup were similar to those in the healthy nonasthmatic subjects. We also found that mast cell genes and TH2 signature genes clustered together in unsupervised hierarchical clustering analyses of the microarray data from epithelial brushings from asthmatic and healthy subjects, lending further support to a link between IEMCs and TH2 inflammation. Thus our findings here add increases in IEMC numbers to the list of pathologies characteristic of TH2-high asthma. Other characteristics we have reported include airway and systemic eosinophilia, sub-epithelial fibrosis, and abnormal expression of gel-forming mucins.7
The levels of expression of tryptase and carboxypeptidase in epithelial brushings in the asthma subgroup correlated positively with lung function improvement after treatment with ICSs. This finding is consistent with the TH2 subtype of asthma being corticosteroid sensitive,7 and it suggests that some of the clinical improvement from ICSs could be a result of decreased activation of IEMCs.
c-kit, a cell-surface receptor on mast cells, was also in this mast cell cluster. SCF is a key ligand for c-kit, and levels of SCF correlate with asthma disease severity.22 We found SCF expression to be increased among subjects with TH2-high asthma and also found SCF expression to be increased at both the gene transcript and protein levels in cultured airway epithelial cells stimulated with IL-13. Interestingly, SCF protein was secreted basally by the epithelial cells, raising the possibility that basal secretion of this mast cell chemokine and growth factor by epithelial cells in response to IL-13 could be a mechanism for mast cell migration from the subepithelial matrix into the intraepithelial space.
We found IEMCs to have upregulation of tryptase and CPA3 but no upregulation of chymase. This pattern of protease expression is not consistent with the patterns observed for MC-T or MC-TC cells. MC-T cells predominantly express high levels of tryptase, but not chymase or carboxypeptidases, whereas MC-TC cells express all 3 proteases.3 To our knowledge, our data for a tryptase-high and CPA3-high subset of mast cells is the first evidence that IEMCs in subjects with mild-to-moderate asthma have a unique phenotype with a different program of gene expression from mast cells in other lung microenvironments. It has previously been shown that mast cells grown in coculture with epithelial cells downregulate chymase.21 We have shown here that CM from IL-13–stimulated epithelial cells downregulate chymase while maintaining CPA3 and tryptase expression in cord blood–derived cultured mast cells. These findings raise the possibility that the TH2-activated epithelium alters the phenotype of mast cells once they accumulate in this location. Our data contrast with those for IEMCs in mouse intestine, which produce chymase that is important for worm expulsion.23 In addition, increased numbers of chymase-producing mast cells have been reported in the small airways of patients with severe asthma,24 suggesting that mast cell phenotypes in the airway vary by disease severity.
The expression of CPA3 in IEMCs suggests a potentially important role for this protease in asthma. Mast cell carboxypeptidase cleaves C-terminal aromatic and some aliphatic amino acids, and several potential natural substrates have been identified, including angiotensins, apolipoproteins, endothelin, and venoms.25–28 Murine mast cell CPA3 can convert angiotensin I to angiotensin II by removing C-terminal amino acids.29 Angiotensin II is proposed to play a major role in airway bronchoconstriction and enhanced airway hyperresponsiveness in asthmatic subjects.30 It is therefore possible that CPA3 plays a role in asthma through production of angiotensin II. Our finding of CPA3 as a major product of IEMCs should prompt further investigation into the potential biologic role of carboxypeptidases in asthma.
The mechanism for how mast cells accumulate in the epithelium is incompletely understood, although there are clues. For example, mast cells adhere avidly to tracheal epithelium through interactions between epithelial cells and cell-surface proteins on mast cells.31 Human lung mast cells express several integrins and immunoglobulin superfamily adhesion receptors, including CD18, very late antigen 4, and αEβ7.32–34 Normal epithelial cells express important counterligands for these receptors, including intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1, and E-cadherin, respectively.35,36
Mast cell products have several important paracrine effects on airway epithelial cell function. First, tryptases can stimulate production of IL-8 and ICAM-1 by epithelial cells.37 ICAM-1 is the primary receptor for the major group rhinoviruses, an important cause of asthma exacerbations. Second, numerous mast cell products can act as mucin secretagogues, including prostaglandin D2, chymase, and leukotriene C4.38–40 Finally, mast cells can produce TH2 cytokines, including IL-4 and IL-13,41 which activate specific programs of gene expression by epithelial cells, leading to mucus hypersecretion, eosinophilic inflammation, and airway hyperresponsiveness.42,43
In summary, we conclude that increases in IEMC numbers occurs in subjects with TH2-high asthma but not in subjects with TH2-low asthma and that IEMCs have an unusual phenotype, showing upregulation of tryptases and CPA3 but no upregulation of chymase. These findings highlight an important pathologic abnormality in asthma that has implications for mechanisms of epithelial cell activation.
Supported by the National Institutes of Health National Research Service Award F32HL093999-01 (R.H.D.), A1077439-025 and R01 HL080414 (J.V.F.), R01 HL095372 (P.G.W.), and P01 HL024136 (G.H.C.).
Disclosure of potential conflict of interest: G. H. Caughey receives research support from the National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute. P. G. Woodruff receives research support from Genentech. J. V. Fahy has consultant arrangements with Amira, Cytokinetics, Abbott, and GlaxoSmithKline and receives research support from Genentech and the NIH. The rest of the authors have declared that they have no conflict of interest.