Published ahead of print on August 17, 2006, doi:10.1164/rccm.200603-310OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200603-310OC
Characterizing Mucous Cell Remodeling in Cystic FibrosisRelationship to NeutrophilsDivision of Pulmonary and Critical Care Medicine, Department of Medicine, Cardiovascular Research Institute, University of California, San Francisco, San Francisco, California Correspondence and requests for reprints should be addressed to John V. Fahy, M.D., University of California, San Francisco, 505 Parnassus Avenue, Box 0111, San Francisco, CA 94143. E-mail: john.fahy{at}ucsf.edu
Rationale: Relatively few studies have characterized mucous cells or mucins in detail in cystic fibrosis (CF), and the relationship between mucous cell abnormalities and neutrophilic inflammation is uncertain. Objectives: To characterize mucous cell phenotypes and mucin profiles in CF and to determine if neutrophils accumulate around goblet cells in the epithelium and gland acini in the submucosa. Methods: Bronchial biopsies were collected from 7 subjects with CF and 15 control subjects, and the morphology of mucous cells was measured. Immunostains for gel-forming mucins and neutrophil elastase were quantified. Measurements and Main Results: Goblet cell size was increased in CF (p = 0.004), but the number of goblet cells was normal. The volume of submucosal glands was fourfold higher than normal (p = 0.031), but the proportion of mucous and serous cells in CF glands was normal. The patterns of expression of gel-forming mucins in epithelial and submucosal compartments in CF were similar to normal. Although neutrophil elastase immunostaining was intense in the epithelium in CF, neutrophils were largely absent around gland acini in the submucosa. Conclusion: The most prominent pathologic feature in the CF airway is an increase in submucosal gland volume, but serous cell transdifferentiation to mucous cells does not occur, nor are gland acini inflamed with neutrophils. The mechanism for increased submucosal gland volume in CF deserves further study.
Key Words: cystic fibrosis • MUC5AC • MUC5B • neutrophil elastase • submucosal glands The pathologic hallmarks of airway disease in cystic fibrosis (CF) include inflammation, especially neutrophilic inflammation, and airway remodeling, especially remodeling of mucous cells. Neutrophilic airway inflammation is one of the earliest findings in CF, occurring in infants before the onset of obvious infection (1). Neutrophil products such as proteases are important, both as growth factors for mucous cells and as secretagogues for mucins. For example, neutrophil elastase stimulates mucin gene expression and mucin secretion in airway epithelial cells in vitro and in vivo (2–6) and induces airway mucous metaplasia in hamsters and mice (7, 8). These findings suggest that neutrophils may be found in close proximity to mucous cells in the epithelium or submucosa in CF, but little information exists about the relationship between neutrophilic inflammation and mucous cells in the epithelium and glands. Remodeling of mucous cells in the surface epithelium and in submucosal glands is characteristic of CF, but surprisingly few studies have examined mucous cell phenotypes in CF in detail or have characterized the expression patterns of gel-forming mucins in the surface epithelium and submucosa. Morphologic studies of lungs from infants show that submucosal glands are normal in size and number (9, 10), regardless of the presence of airway infection (10). However, autopsy studies in older patients with CF show that submucosal glands are usually prominent (11), with a preponderance of mucous acini within the glands (12). Thus, patients with CF are not born with mucous cell abnormalities, but clear changes occur by the time of death. Data on the intervening years are limited, because very few studies have used bronchoscopy to investigate the airway phenotype of patients living with CF. Gel-forming mucins, such as MUC2, MUC5AC, and MUC5B, are secreted by airway mucous cells, and these mucins are subject to regulation by inflammatory stimuli (13, 14). Changes in mucin gene expression are therefore likely to occur in CF, but most studies have been in the upper airway and have yielded conflicting results. One study shows decreased MUC5AC expression in nasal epithelial cells from patients with CF (15), whereas another shows increased MUC2 expression (16). Recently, mucin proteins have been quantified in airway secretions in CF. Surprisingly, it was found that MUC5B and MUC5AC concentrations are lower than normal (17). To our knowledge, there has been no study quantifying the expression of gel-forming mucins in tissues from the lower airway in CF. In this study, we examined neutrophilic inflammation and mucous cell remodeling in CF by performing detailed analyses of neutrophils, mucous cells, and mucins in tissue sections from biopsies obtained during research bronchoscopy. We used stereology to quantify the size and number of mucous cells, and their mucin expression patterns. We also examined the spatial distribution of neutrophils around goblet cells in the surface epithelium and gland acini in the submucosa. Some of these data have been previously presented in abstract form (18).
Subjects Seven persons with CF and 15 healthy control subjects were enrolled. Subjects with CF were 23 to 37 years old, met Cystic Fibrosis Foundation criteria for the diagnosis of CF, had compatible lung disease, and sweat chloride testing results ranging from 82 to 120 mM/L. Five subjects were homozygous for  F508, one subject was F508/3905insT, and one subject was 1717-1/unidentified. Five of the seven subjects with CF had sputum that was culture positive for Pseudomonas aeruginosa at the time of the study. Control subjects were healthy volunteers, aged 22 to 44, with PC20 methacholine greater than 16 mg/ml and FEV1 greater than 80% predicted. Subjects were excluded for a CF exacerbation or respiratory infection within the previous 6 wk, smoking history (> 10 pack-years lifetime or any cigarette smoking in the last year), history of hemoptysis requiring intensive care unit admission, or other CF complication that significantly increased the risk of bronchoscopy. All subjects provided written, informed consent and the University of California–San Francisco Committee on Human Research approved the study. Biopsy samples from these same subjects had previously been used for analysis of airway smooth muscle morphology (19, 20).
Protocol
Stereology
Statistics
Subjects The clinical characteristics of the study subjects are shown in Table 1. Subjects with CF had moderate to severe airflow obstruction, with FEV1 ranging from 45 to 71% predicted. Five of the seven subjects with CF (71%) were allergic to at least one aeroallergen, a fraction that was not different from healthy subjects (12 of the 15 healthy subjects, 80%). In addition, the mean number of positive skin tests in the healthy subgroup (2.86) and CF subgroup (2.42) was not significantly different (p = 0.697).
Epithelium and Goblet Cells The epithelium was higher than normal in CF (Table 2). There was a trend for an increase in epithelial mucin stores in CF (as referenced to the surface area of basal lamina), and this was a consequence of an increase in the average size of goblet cells, rather than in the number of goblet cells (Table 2).
Submucosal Glands The volume fraction of submucosal glands normalized to the surface area of basal lamina was fourfold greater in subjects with CF than in healthy subjects (Figure 1). The volume of serous cells as a fraction of overall gland volume was similar in CF and healthy subjects (Figure 2, Table 2); likewise, the volume of mucous cells as a fraction of overall gland volume was no different in CF and healthy subjects (Figure 2, Table 2). Thus, the increase in gland volume in CF is the result of increases in both serous and mucous cell compartments, with no change in the overall proportions of these gland cell types.
Mucin Glycoproteins In the surface epithelium, MUC5AC was the most prominent gel-forming mucin in both healthy and CF subjects, and we found no difference in MUC5AC between groups (Figures 3 and 4). In the submucosa, MUC5B immunostaining was very prominent in glands with little or no immunostaining for MUC2 or MUC5AC. The volume of MUC5B immunostain was higher than normal in the CF submucosa, but when normalized to the volume of glands, there was no difference between groups (Figure 4).
Neutrophil Elastase Neutrophil elastase immunostaining was most intense in the epithelium of CF biopsies and, when quantified, the volume of neutrophil elastase staining in the epithelium was fourfold higher in subjects with CF than in healthy subjects (Figure 5). In contrast, neutrophil elastase immunostaining was minimal in the region of submucosal glands in CF (Figure 5). When quantified, the volume of neutrophil elastase staining in the submucosa was similar in subjects with CF and healthy subjects (Figure 5). In hematoxylin and eosin–stained sections from the same subjects, we found polymorphonuclear cells (PMNs) in the epithelial layer in CF and only rarely in the epithelial layer of healthy subjects (Figure E2). We did not find PMNs in the submucosal gland regions in these hematoxylin and eosin sections, consistent with the immunostaining results.
TGF-  ImmunostainingTGF- is an important ligand for the epithelial growth factor receptor (EGFR), and EGFR signaling has been implicated in the mechanism of mucin gene up-regulation and epithelial tubulogenesis (25, 26). Thus, we examined the CF tissue sections for evidence of TGF- expression. TGF- staining, as detected by the MF9 antibody, was evident in epithelial cells in a perinuclear distribution (Figure 6), and the volume of immunostain was higher in the subjects with CF but not significantly so (Table 2). In the submucosa, TGF- immunostain was very prominent in glands (Figure 6), and the volume of TGF- was significantly greater than normal, reflecting the higher gland volume in CF (Table 2).
Epithelial Tubulogenesis An increase in gland volume may result from new gland growth through a process of epithelial tubulogenesis, whereby glands form as a result of invaginations or buds of surface epithelial cells. We examined biopsy tissues from the subjects with CF for evidence of these buds or invaginations. Although not frequent, we did find structures in the CF sections close to the epithelium that could be new gland forming (Figure 7). This finding needs to be interpreted cautiously not only because of our limited sample size but also because there are no markers that distinguish new glands from old. A related finding in the CF biopsy sections was that the basal lamina zone was 20% thinner than normal (Figure 8), a characteristic that may facilitate epithelial tubulogenesis.
Neutrophil Elastase and Mucous Cell Remodeling We considered it possible that neutrophilic inflammation in the surface epithelium is mechanistically linked to remodeling of goblet cells and submucosal gland cells. When using data from all subjects, we found a direct relationship between the volume of neutrophil elastase immunostaining in the epithelial layer and the volume of glands in the submucosa (r = 0.53, p = 0.01). This correlation did not reach statistical significance when using the data for the subjects with CF alone because of the limited number of subjects.
Although there have been post mortem studies of the airway pathology of CF (10–12, 27), we believe ours is the first study to quantify mucous cells, mucins, and inflammatory changes in CF using tissue obtained by bronchoscopy and methods of design-based stereology. We found that the volume fraction of submucosal glands was fourfold higher than normal in CF, a finding that is much more striking than the changes in goblet cells, which were more subtle. These subtleties included a modest increase in the volume of stored mucin in the epithelium, which was the result of increased goblet cell size rather than increase in goblet cell number. Mucin stores in goblet cells may be relatively low in CF because of ongoing degranulation signals. Nevertheless, we were surprised that we did not find more evidence of goblet cell hyperplasia in the large airways in CF. Using immunohistochemistry, we found that MUC5AC was the predominant mucin expressed by goblet cells, and MUC5B was the predominant mucin in gland mucous cells. This pattern was not different in subjects with CF compared with healthy subjects. Thus, although the large inflammatory stimuli in CF may up-regulate mucin gene expression in mucous cells in the epithelium and submucosa, they do not alter basic patterns of expression of gel-forming mucins. The increased mucin stores in the submucosal glands represent a reservoir of mucins that could contribute to mechanisms of mucin hypersecretion in CF. Although mechanisms of mucous cell degranulation are distinct from those of intracellular mucin accumulation, it is reasonable to assume that increases in mucin stores in the submucosa lead to increases in mucins in secretions. Neutrophilic airway inflammation is a known pathologic feature of CF, but our goal was to examine neutrophils in relation to goblet cells in the epithelium and gland cells in the submucosa. We found that neutrophil elastase immunostaining was largely absent around glands in the submucosa, so that neutrophil-mediated degranulation of gland mucous cells is very unlikely. Other mucin secretagogues must regulate gland mucin secretion in CF. In contrast, in the epithelium, we found intense neutrophil elastase immunostaining. It is unclear what causes neutrophils to accumulate in the epithelium. Recent data suggest that there are epithelial cell changes related to CFTR deficiency that cause neutrophil accumulation in the epithelial layer (28, 29). For example, epithelial cells in CF have an augmented production of interleukin 8 (30–32) and increased intercellular adhesion molecule-1 expression (28, 29). Whatever the mechanism of neutrophil accumulation in the epithelial layer, the consequences could include degranulation of goblet cells, or stimulation of goblet cell hypertrophy. The increased volume of glands in the submucosa was the most striking aspect of the pathologic phenotype that we observed in the subjects with CF. Our data showing an increase in the volume fraction of glands in the submucosa could result from the numbers of gland units being increased, or because individual glands become hypertrophied. Our methods do not allow us to distinguish between these possibilities. We are able to state that the gland volume increases with proportionate increases in the serous and mucous acini. Published data do not shed light on whether gland enlargement in CF results from growth of new submucosal glands, the enlargement of existing glands, or both. There are data to support the concept of new gland growth in adult airways. For example, in health, submucosal glands are confined to large cartilaginous airways. However, in autopsy studies of CF lungs, submucosal glands are found in more distal, noncartilaginous airways (33). In addition, using animal models, it is possible to induce gland growth in the airways (34, (35), suggesting that new gland growth is possible in adult lungs. Any such development of glands from surface epithelial cells would require epithelial progenitor cells and recapitulation of the epithelial tubulogenesis that occurs during lung development. During fetal lung development, buds form from the surface epithelium followed by lateral morphogenesis into the mesenchyme (36–38), a sequence that can be reproduced in cell culture. Normal human tracheobronchial epithelial cells cocultured with human fetal lung fibroblasts penetrate collagen matrices (39), form tubular structures, and undergo dichotomous branching (40). Polarized noninvasive epithelial cells may transform into nonpolar invasive mesenchymal cells ("epithelial mesenchymal transition"), break through the basement membrane, and invade surrounding extracellular matrix (41).
Our study design did not address the mechanism of gland enlargement in CF, but we hypothesize that neutrophil-directed epithelial cell activation could lead to epithelial mesenchymal transition, epithelial tubulogenesis, and new glands. Our findings for the epithelial-restricted nature of neutrophilic inflammation in CF and the positive correlation between epithelial neutrophils and gland volume in the submucosa provide some support for this hypothesis as does the finding of thinning of the basal lamina zone in CF. This latter result could occur because of the actions of neutrophil proteases and thinning of the basal lamina, which could, in turn, facilitate tubulogenesis. Interactions between epithelial cells and extracellular matrix have been particularly well studied in the kidney. For example, locally produced growth factors, such as EGFR ligands, have roles in nephrogenesis and in renal regeneration after acute tubular injury (42, 43). Both the number and extent of tyrosine phosphorylation of EGF receptors increase in the developing kidney during late fetal development, coincident with the timing of tubulogenesis and glomerulogenesis (26, 43). TGF-
The authors thank the following staff at UCSF for assistance with this study: Peggy Cadbury and Hofer Wong for recruiting subjects and for assistance with bronchoscopy, and Roderick Carter and Ronald Ferrando for assistance in tissue processing and analysis. The authors also thank Prescott Woodruff and Anh Innes for editing the manuscript and Ingemar Carlstedt for the generous contribution of the monoclonal antibodies LUM2-3 and LUM5B-2.
Supported by the UCSF Cystic Fibrosis Foundation Research Development Program, the Diamond Family Trust, and by NIH grant RO1 HL66563 (J.V.F.). S.R.H. has received fellowship training support from the NIH (5T32HL007185) and was a recipient of a fellowship training award from the Cystic Fibrosis Foundation (HAYS02) and NIH grant K23 HL077660. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200603-310OC on August 17, 2006 Conflict of Interest Statement: Neither author has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form March 1, 2006; accepted in final form August 15, 2006
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
