Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Epithelial surfaces of respiratory tracts are coated by mucus that plays a critical role in the hydration and protection of airway epithelium (1). Mucus is a viscoelastic gel, formed by composite secretions of submucosal gland cells and goblet cells in the surface airway epithelium. Mucins are high molecular-weight glycoproteins that constitute the major component of mucus secretions in the airway and other secretory epithelia such as the gastrointestinal and reproductive tracts. To date, nine mucin genes (MUC1, MUC2, MUC3, MUC4, MUC5B, MUC5AC, MUC6, MUC7, and MUC8) have been identified (2). Characteristic structural features of these mucin genes include tandem repeats that vary in length and sequence resulting in the polymorphisms at the DNA and protein level (11, 12). The repeat units of the different mucins exhibit little similarity to each other either in sequence or number of amino acids in the repeats. Several studies have demonstrated that multiple mucin and apomucin genes are normally expressed in secretory tissues (13). In the lung, seven mucin genes (i.e., MUC1, MUC2, MUC3, MUC4, MUC5B, MUC5AC, and MUC8) have been demonstrated to be expressed at the protein and/or mRNA level (3, 10). However, little is known about the heterogeneity of mucin gene expression in the various cellular compartments of the lung.

Mucus overproduction is observed in several diseases such as asthma, chronic bronchitis, cancer, and cystic fibrosis (CF) (16). CF is an autosomal, recessive, inherited disorder that affects children and young adults, resulting from dysfunction of an epithelial chloride channel (cystic fibrosis transmembrane conductance regulator [CFTR]) (23). The clinical manifestations of airway disease in CF include chronic bacterial infections and inflammation, as well as mucus overproduction, which leads to respiratory failure and death (18). The hypersecretory disease state in the CF lung is characterized by hyperplasia and metaplasia of goblet cells in the surface airway epithelium (18, 24) as well as hypertrophy and hyperplasia of submucosal gland regions in the airways (25). Although it is generally accepted that hypersecretion in the CF lung is secondary to airway infection and not primary to CFTR dysfunction, it has been difficult to establish this fact conclusively. Recent studies in a human bronchial xenograft model have demonstrated that in the absence of infection, CF and non-CF airways have an equivalent percentage of goblet cells in the surface airway epithelium (26). Such findings support the hypothesis that goblet cell hyperplasia and metaplasia in the CF airway are secondary to bacterial infection. Knowledge of the normal and disease cellular patterns of mucin expression is needed in order to fully understand the airway abnormalities in the CF hypersecretory condition. Such information will aid in elucidating the pathophysiologic mechanisms involved in mucus hypersecretion.

The two primary sites of mucin production in the lung are goblet cells of the surface airway epithelium and mucous cells of the submucosal glands. Submucosal glands, which are present in the cartilaginous airways, are composed of an interconnecting network of tubules and ducts that secrete a proteinaceous fluid into the airways. Two predominant types of tubules in submucosal glands are composed of either serous or mucous acinar and ductular cell types (27). Serous cells of submucosal glands secrete a number of bactericidal proteins such as lysozyme and lactoferrin (28), whereas mucous cells primarily secrete mucin (1). Serous cells express high levels of CFTR protein, which in part serves to secrete fluid and electrolytes important in glandular secretion and hydration of mucins (29). Submucosal glands are believed to be important in the pathophysiology of CF airways disease.

Numerous studies have demonstrated alterations in the biochemical properties of CF airway mucins. These have included decreased sialylation, increased sulfation, and increased fucosylation of carbohydrate chains on secreted mucins (26, 30). These differences may result from defective function of intracellular CFTR, which acidifies and alters the ionic content of glycoprotein-processing compartments such as the Golgi (33). Additionally, amino acid analysis of purified mucin from asthmatic and CF patients has revealed increased levels of serine and threonine content in CF mucins (34). These findings suggest that alterations in the proportion of different mucin gene products expressed in the airways may occur in CF. Given that alteration in the biochemical properties of airway mucins accompany the CF phenotype, we hypothesized that the type and/or cellular distribution of mucin expression in the airway may also be affected in CF airways disease. To this end, we sought to determine whether mucin gene expression may be altered in CF airways.

In the present study, we have identified the cells in the human bronchial airway responsible for MUC5B and MUC7 secretion and determined whether any alterations in the expression patterns of these mucins occur in CF as compared with non-CF cells. To achieve these goals, we performed in situ hybridization and immunofluorescence detection of MUC5B and MUC7 mRNA and protein expression, respectively. Our data demonstrate that MUC5B expression was confined to mucous cells of airway submucosal glands, whereas MUC7 was expressed mainly in a subpopulation of submucosal gland serous cells. There were no differences in MUC5B and MUC7 expression observed between CF and non-CF bronchial tissues.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Bronchial Tissue Samples

Six CF and ten non-CF bronchial samples were obtained from lung transplant tissue. Typically, samples were obtained from airways ranging from the distal trachea to third-generation bronchus. The CF genotypes analyzed in this study included F508/F508 (two); F508/unknown (three); and F508/G551D (one). The unknown CF alleles are defined as the absence of the following 26 common mutations: G85E, R117H, 621 + 1G > T, R334W, R347C, R347P, R347H, S492F, Q493R, 1609delCA, I507, F508, 1717-1G > A, G542X, S549I, S549N, S549R(A > C), G551D, R553X, R506T, R1162X, S1251N, W1282X, R1283M, R1283K, and N1303K. This panel of mutations has detected 87% of 301 CF chromosomes analyzed from the University of Pennsylvania Molecular Diagnostic Laboratory (Philadelphia, PA). Bronchial samples were excised and immediately frozen in optimal cutting temperature (OCT) embedding media (Miles, Elkhardt, IN). Cryosections (6 µm) were analyzed for mRNA and protein localization studies.

In Situ Hybridization

In situ analysis for mRNA was performed as described previously (35). Briefly, frozen CF and non-CF bronchial sections were thawed and fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 5 min. Following fixation, sections were treated with 0.25% acetic anhydride in 0.1 M triethanolamine, pH 8.0, for 10 min at room temperature. Sections were then dehydrated in graded concentrations of ethanol, and air-dried. Hybridization was performed using alkaline phosphatase-labeled oligonucleotides. The specificity of each probe has been described previously (35, 36). Antisense oligonucleotide probes for MUC5B (5'-GCGGGTCCAGGTGGTGCCCAGGGAGGAGGA-3') and MUC7 (5'-GTGGAGCTGGTGTAGTTGCAGAAGGTGTGG-3') were used to detect mRNA expression (6, 9). Pretreatment of sections with ribonuclease A (200 µg/ml) at 37°C for 2 h prior to hybridization with antisense probes was used to control for specificity of hybridization. Additional controls included sections left untreated with oligonucleotide probes. Sections were probed with antisense MUC5B (20 fmol/µl) and MUC7 (5 fmol/ µl) probes at 37°C overnight in hybridization buffer containing 50% formamide, 1× Denhardt's, 5× standard saline citrate (SSC), and salmon sperm DNA (400 µg/ml). Following hybridization, slides were washed in 0.1× SSC for 30 min at 37°C (four changes) followed by a 1-h wash at room temperature in the same buffer. Sections were developed using nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate (Sigma, St. Louis, MO) in 100 mM Tris (pH 9.0), 100 mM NaCl, and 50 mM MgCl₂ for 5 h at 37°C, then mounted and analyzed by brightfield microscopy.

Immunofluorescence Detection of MUC5B and MUC7 in Human Bronchus

MUC5B and MUC7 proteins were detected using monoclonal antibodies PANH2 and PANH3, respectively. The specificities of these antibodies have been described previously (35, 36). PANH2 antibody was raised against partially deglycosylated MG1 and recognizes MUC5B, and PANH3 was raised against a 21-residue synthetic peptide derived from the N-terminus sequence of MUC7 (9). For detection of MUC5B protein, sections were thawed, fixed in 95% ethanol, and blocked using PBS containing 20% goat serum (GS) for 30 min at room temperature. Sections were then incubated overnight with PANH2 antibody (hybridoma supernatant at a 1:1 dilution) in PBS/1.5% GS. Following staining, the sections were washed in three changes of PBS/1.5% GS and incubated in antimouse fluorescein isothiocyanate (FITC)-conjugated antibody (7.5 µg/ ml) for 30 min at room temperature. Sections were then washed three times in PBS/1.5% GS for 10 min each and mounted in Citifluor antifadent (UKC Chem Lab, Canterbury, UK) and then examined using a Leica microscope (Leitz DMR, Wetzlar, Germany). For detection of MUC7 protein expression the procedure was the same as for MUC5B except the slides were fixed in methanol (20°C) for 10 min, followed by air-drying.

Double-labeling immunofluorescence detection of lysozyme and MUC7 utilized sequential staining with anti-lysozyme antibody (BioGenex Laboratories, San Ramon, CA) followed by staining with PANH3 antibody. For lysozyme staining, sections were fixed in 95% ethanol for 10 min at room temperature followed by blocking in PBS/ 20% GS for 30 min. Sections were then incubated sequentially in anti-lysozyme for 12 h at 4°C (1:250 dilution), followed by detection with antirabbit Texas Red-conjugated antibody (7.5 µg/ml). Sections were then post-fixed in methanol at 20°C for 10 min, followed by air-drying prior to staining for MUC7. The sections were then blocked with PBS/20% GS and incubated sequentially with PANH3 (1:1 dilution) in PBS/1.5% GS for 16 h at 4°C, followed by antimouse FITC-labeled antibody (7.5 µg/ml).

Morphometric Analysis of MUC7 and MUC5B Protein Expression

Morphometric analysis of MUC7 protein expression in submucosal glands was performed at low magnification on a Leica DMR fluorescent microscope. Glandular structures of ten non-CF and seven CF bronchial samples were quantified for MUC5B and MUC7 expression in three independent sections from each bronchial sample. For CF samples, two of the six bronchial samples were from different regions of the same lung. Criteria for mucous cell identification included flat lucent cellular features, characteristic for Nomarski optics, and nuclei situated at the base of the cell. The percentage of mucous tubules positive for MUC5B protein was quantified for each submucosal gland in a given section. Individual submucosal glands were identified as spatially distinct groups of serous and mucous tubules within the submucosa that were separated by an elastic lamina and connective tissue.

Serous tubules were identified by their distinctive morphology under Nomarski optics, which included characteristic dense cytoplasm and centrally situated nuclei. Serous demilunes were often seen at the distal ends of mucous tubules. Lysozyme was used as a marker for serous cells in studies quantifying MUC7 expression. In sections where MUC7 expression was seen, the number of MUC7-expressing serous tubules were quantified for each gland within a given section using double immunofluorescence with lysozyme and MUC7 antibodies. On average, more than nine glands were quantitated from each bronchial sample for the percentage of mucous and serous tubules that expressed MUC5B and MUC7 protein, respectively.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Detection of MUC7 and MUC5B RNA in Non-CF Human Bronchus

The distribution of MUC7 and MUC5B gene expression was examined in non-CF bronchial samples by in situ hybridization using alkaline phosphatase-labeled oligonucleotide probes. The hybridization specificity was demonstrated by comparing antisense-probed sections with tissue sections pretreated with RNase prior to hybridization. Hybridization of non-CF bronchial samples with MUC7 antisense probes demonstrated high levels of expression in a subpopulation of serous tubules within submucosal glands (Figures 1A and 1E). An absence of detectable MUC7 mRNA in sections pretreated with RNase and hybridized to antisense probes confirmed the specificity of hybridization (Figures 1B and 1F). No MUC7 expression was seen in the surface airway epithelium. In contrast to MUC7, in situ hybridization of non-CF bronchial samples with antisense probes to MUC5B demonstrated staining in all the mucous tubules of the submucosal glands (Figures 1C and 1G). Hybridization specificity was demonstrated by the absence of signal with an antisense probe following RNase pretreatment (Figures 1D and 1H). Similar to MUC7, no expression of MUC5B was seen in the surface airway epithelium.

View larger version (114K): [in this window] [in a new window]   Figure 1.   Nonradioactive in situ detection of MUC5B and MUC7 mRNA in submucosal glands of non-CF human bronchus. Antisense alkaline phosphatase-labeled oligonucleotides to MUC7 (A and E) and MUC5B (C and G) were hybridized to non-CF bronchial tissue section and reactivity was developed by alkaline phosphatase staining. Arrows indicate positive expression of MUC5B confined to mucous tubules and MUC7 confined to subpopulation of serous tubules in the submucosal glands. No staining was observed in sections pretreated with RNase followed by hybridization with antisense probes to MUC7 (B and F) and MUC5B (D and H). Photomicrographs represent Nomarski (top row) and brightfield (bottom row) optics of the same field. Panels A-D represent ×200 fields; panels E-H represent ×400 fields.

Detection of MUC7 and MUC5B Protein in CF and Non-CF Human Bronchus

The localization of MUC7 protein was studied in human bronchial tissue samples from non-CF (n = 10) individuals and CF (n = 6) individuals using a previously characterized antibody (PANH3) specific to the N-terminal peptide region of MUC7. Results from immunocytochemical studies confirmed the expression pattern of MUC7 determined by in situ hybridization within a subpopulation of serous tubules of submucosal glands (Figures 2A and 2D). Only bronchial samples that demonstrated mRNA expression of MUC7 also showed reactivity to PANH3. Staining was not observed in sections probed with secondary antibody alone (Figures 2C and 2F). Additionally, no alterations in the intensity, frequency, or cellular patterns of MUC7 expression were observed in CF as compared with non-CF human bronchial samples (Figures 2B and 2E).

View larger version (104K): [in this window] [in a new window]   Figure 2.   Immunofluorescent localization of MUC7 protein in human bronchus. Methanol-fixed cryosections (6 µm) of non-CF (A and D) and CF (B and E) human bronchus were stained with PANH3 antibody followed by FITC-labeled antimouse antibody. Expression of MUC7 is observed in a subpopulation of serous tubules of the submucosal glands. No expression was seen in tissues incubated with secondary antibody alone (C and F ). Photomicrographs represent Nomarski (top row) and fluorescent (bottom row) photomicrographs. Panels A-C represent ×200 fields; panels D-F represent ×400 fields. Serous tubules (s); mucous tubules (m).

Localization of MUC5B protein using the PANH2 antibody confirmed the glandular mucous cell-specific expression pattern determined by mRNA localization studies. In contrast to MUC7 expression, MUC5B was expressed in all mucous cells of submucosal glands in non-CF bronchial samples (Figures 3A and 3D). No staining was observed in sections probed with secondary antibody alone (Figures 3C and 3F). CF samples demonstrated an identical pattern of glandular MUC5B expression as seen in non-CF samples (Figures 3B and 3E). MUC7 expression in serous cells was confirmed by co-localization studies with anti-lysozyme and anti-MUC7 antibodies (Figure 4). The example shown in Figure 4 demonstrates both positive and negative MUC7-expressing serous cells, which are positive for lysozyme (Figures 4A and 4B). In summary, results of MUC5B and MUC7 protein expression confirmed findings from the mRNA localization studies.

View larger version (107K): [in this window] [in a new window]   Figure 3.   Immunofluorescent localization of MUC5B protein in human bronchus. Ethanol-fixed cryosections (6 µm) of non-CF (A and D) and CF (B and E) human bronchus were stained with PANH2 antibody followed by FITC-labeled antimouse antibody. Expression of MUC5B is observed in all mucous tubules of the submucosal glands. No expression was seen in tissues incubated with secondary antibody alone (C and F ). Photomicrographs represent Nomarski (top row) and fluorescent (bottom row) photomicrographs. Panels A-C represent ×200 fields; panels D-F represent ×400 fields. Serous tubules (s); mucous tubules (m).

View larger version (59K): [in this window] [in a new window]   Figure 4.   Co-localization of MUC7 and lysozyme in non-CF human bronchus using double-labeling immunofluorescence. Ethanol-fixed human bronchial sections (6 µM) were incubated with anti-lysozyme antibody followed by Texas Red-labeled antirabbit antibody. The section was then refixed in methanol and incubated with PANH3 followed by FITC-labeled antimouse antibody. Photomicrographs represent Nomarski (panel A), Texas Red channel lysozyme expression (panel B), FITC channel MUC7 expression (panel C), and combined Texas Red/FITC channels (panel D). MUC7 protein is observed in a subpopulation of lysozyme-expressing serous tubules. The arrowheads mark serous cells that express both MUC7 and lysozyme, whereas the arrows indicate the serous tubules that express lysozyme but not MUC7.

Morphometric Analysis of MUC7 and MUC5B Expression

Morphometric analysis was performed to evaluate heterogeneity of MUC7 and MUC5B protein expression in submucosal glands of bronchial tissues (Table 1). The percentage of MUC7- and MUC5B-expressing tubules was quantified from at least nine independent glands from three sections of each bronchial sample. Quantification of serous cells expressing MUC7 was performed using double labeling of the same section with MUC7 and lysozyme antibodies. Of ten non-CF samples analyzed, only two expressed MUC7 protein. The average percentage of serous tubules expressing MUC7 in each submucosal gland for the two samples was 27%. The heterogeneity of MUC7 expression between glands in a given tissue ranged from 0 to 93%. Of six CF samples analyzed, one bronchial sample expressed MUC7 protein. The percentage of serous tubules positive for MUC7 in this sample averaged 32% with a range of 0 to 85%. Interestingly, heterogeneity of MUC7 expression was observed in two independent bronchial samples from the same patient; only one of the two demonstrated MUC7 protein expression. This suggests that heterogeneity in MUC7 expression is observed among different patient samples and also between different airways in a single lung. In contrast, MUC5B was uniformly expressed in all submucosal gland mucous tubules of all the bronchial samples analyzed (both CF and non-CF).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In the present study, expression of MUC5B and MUC7 mRNA was evaluated in non-CF human bronchus using oligonucleotide probes to MUC5B and MUC7 tandem repeats. The use of these repeat probes afforded increased sensitivity of detection over other traditional methods of in situ mRNA localization. The use of specific monoclonal antibodies to MUC5B and MUC7 proteins confirmed cellular expression patterns determined by mRNA localization and facilitated more definitive characterization of cellular phenotypes. The results indicate that MUC7 mRNA and protein are localized in a subpopulation of serous cells. This observation was confirmed by co-localization studies of MUC7 and lysozyme, as a serous protein marker (38). In contrast, MUC5B mRNA and protein were expressed specifically in all mucous tubules of the submucosal glands. These results indicate that MUC5B and MUC7 are produced by mutually exclusive cellular compartments of submucosal glands. The cellular partitioning of mucin gene expression within serous and mucous cells of the submucosal glands represents differences in the differentiated state of these two cell types. The function of these two mucins in airways remains to be investigated.

Our results are consistent with those observed by Nielsen and colleagues (35, 36) where expression of MUC5B was observed in all the mucous cells of the sublingual, submandibular, and labial glands, whereas MUC7 labeled only the serous cells. Such differential expression of mucin genes has also been observed in the intestine, where MUC2 expression is limited to goblet cells and MUC3 is expressed by both goblet cells and absorptive cells (36, 39). Similarly, MUC5 protein was expressed in the surface mucous cells of the cardia, funds, and antrum. In contrast, MUC6 protein expression was limited to the mucous neck cells of the fundus, antral-type glands of the antrum and cardia, and Brunners gland of the duodenum (40). Previous studies that used Northern blot analysis with oligonucleotide probes to the tandem repeats of MUC5B failed to detect significant levels of MUC5B mRNA in the human bronchus, pancreas, and intestine (19, 41). This discrepancy with our present data may be due to a low abundance of submucosal glands represented in tissue samples analyzed by Northern blot analysis. Furthermore, the assays used to determine expression in this study have greater sensitivity at the single-cell level than Northern blot analyses. Comparisons between studies that use mRNA as the sole endpoint for expression must be interpreted with caution. For example, discrepancies in mucin mRNA and protein expression have been observed previously in serous cells that demonstrate expression of MUC2 and MUC3 mRNA, but have not been shown to produce the corresponding protein (42). In contrast to these studies, we observed a complete concordance between both protein and mRNA expression patterns of MUC5B and MUC7.

Changes in mucin gene expression patterns in response to disease are a well-documented phenomenon in both cancer and infection (19). Studies in rat trachea using MUC2 probes demonstrated elevated levels of MUC2 in response to infection and irritation by microorganisms (24). The levels of MUC2 were elevated due to increases in mucin gene transcription and may represent increases in MUC2-expressing cellular compartments caused by mucus cell hyperplasia and/or metaplasia. Similar phenomena also occur in goblet-cell hyperplasia seen in the CF airways. However, it is unclear whether changes in the types of mucin expressed, rather than abundance, also accompany the disease phenotype in the CF airway. Although numerous studies have documented alterations in the biochemical properties of secreted mucins in the CF airway, none has yet determined whether these alterations reflect changes in the cellular phenotypes of mucin-secreting cells (i.e., through alterations in the mucin gene expression patterns). Our results indicate no significant differences in the intensity or cellular patterns of MUC5B and MUC7 expression between CF and non-CF bronchial tissues. However, because of the infrequent nature of MUC7 gene expression, larger numbers of samples will be necessary to definitively address whether phenotypic changes occur in CF airways.

The identification of cell-specific mucin gene expression patterns has advanced our understanding about the serous- and mucous-cell phenotypes in the lung. Furthermore, the identification of MUC5B and MUC7 expression in submucosal glands suggests that these mucins may have roles functionally distinct from mucin produced by surface airway epithelium. The heterogeneity of MUC7 expression seen between different glands of the same tissue sample suggests the possibility that not all glands are composed of phenotypically identical cell types. The fact that glands in the same section demonstrated a range of 0 to 93% MUC7-positive serous cells can be explained by two potential mechanisms: (1) local environmental influences may affect mucin gene expression in serous cells, and/or (2) different glands in the airway may be derived from different cell types that have varying capacities for differentiation into MUC7-expressing serous-cell phenotypes. Additionally, variation in MUC7 expression could be a function of the level of the airway, which could not be systematically controlled in this study. Despite the mechanism for heterogeneity of MUC7 expression patterns in airway submucosal glands, these studies conclusively demonstrate that not all glands in a given airway are phenotypically identical.