Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Overproduction and hypersecretion of airway mucus is a common feature of several chronic human respiratory diseases, including asthma, chronic bronchitis and rhinitis, and cystic fibrosis. Individuals with these diseases often have excessive numbers of mucous (goblet) secretory cells in their airways and elevated numbers of inflammatory cells within the airway mucosa and lumens, often associated with recurrent bouts of bacterial infections. An increase in the number of mucous cells in airway epithelium that normally contains these cells is termed mucous-cell (or secretory-cell) hyperplasia. However, when mucous secretory cells appear in regions of the respiratory tract that are normally devoid of these cells, this change in epithelial phenotype is referred to as mucous-cell metaplasia (MCM). The cellular and molecular mechanisms involved in the development and persistence of these hypersecretory airway diseases are unknown; however, the chronic presence of inflammatory cells and their secreted products may contribute to the hyperplastic or metaplastic changes in the mucous secretory apparatus in these individuals.

Several animal models of human hypersecretory airway diseases have been developed. MCM/hyperplasia can be induced in rodent airways by exposure to commonly encountered toxicants such as tobacco smoke (1), ozone (4), and bacterial endotoxin (7, 8), or by the instillation of neutrophil elastase (9). We have previously shown that, in rats, MCM induced by ozone (6, 12, 13) or endotoxin (8, 14) is preceded by acute, transient, neutrophilic inflammation and epithelial cell proliferation. In addition, we have reported that when the inflammatory response is reduced by depleting rats of circulating neutrophils prior to exposure (15), or by administration of an anti-inflammatory topical steroid during exposure (16), the toxicant-induced MCM is significantly attenuated. This suggests that neutrophils or their secreted products are important elements in the development or expression of a mucous-cell phenotype during toxicant-induced MCM.

Infiltrating neutrophils secrete soluble products such as cytokines (tumor necrosis factor [TNF]-, interleukin [IL]-1, and IL-6) (17, 18) and proteases (i.e., elastase and cathepsin G) (19). TNF-, IL-1, and IL-6 have all been reported to induce airway mucin hypersecretion (20). Neutrophil elastase has been shown to induce MCM in rodent pulmonary airways (9, 23). Recently, TNF- has been shown to induce the expression of a major airway mucin gene (MUC-2) in a human airway epithelial cell line (20). These data suggest that neutrophils or their soluble products may directly affect mucin gene expression as well as mucin synthesis and secretion.

To study the pathogenesis of toxicant-induced MCM we have developed an in vitro culture system to examine the regulation of airway mucous-cell differentiation and mucin gene expression in microdissected rat nasal airway tissues in the absence of cellular inflammation or other systemic effects found in vivo. In the present study we report the maintenance of two morphologically distinct nasal airway epithelia in vitro for up to 2 wk. We cultured microdissected proximal nasal septa, which are lined by a respiratory epithelium (RE) with numerous mucous (goblet) cells, to determine whether the culture conditions were sufficient to maintain this mucociliary phenotype in vitro. We also cultured microdissected maxilloturbinates which are lined by nasal transitional epithelium (NTE) to determine the effect of this culture system on this normally nonsecretory epithelium. These tissues were selected because (1) they have morphologically distinct epithelial cell populations with different phenotypic responses to toxicant exposure, and (2) they represent common sites of toxicant-induced epithelial injury and repair in rodent nasal airways. Light microscopy, electron microscopy, and morphometric analyses were used to assess the effects of in vitro culture on the type and number of epithelial cells present and the quantity of stored intraepithelial mucosubstances. We employed a quantitative reverse transcription-polymerase chain reaction (RT-PCR) assay to measure steady-state levels of rMuc-5AC messenger RNA (mRNA) in these nasal tissues to determine the effect of our in vitro culture system on this airway mucin gene expression.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals, Necropsies, and Tissue Culture

Animals. Twenty-four male specific pathogen-free F344/N Hsd (Harlan Sprague-Dawley, Indianapolis, IN) rats, 12 wk of age, were used in this study. Each rat was randomly assigned to one of four experimental groups (n = 6 per group). The rats were housed in pairs with free access to food and water at the University Research Containment Facility (Michigan State University).

Reagents. Ham's F12 nutrient mixture with L-glutamine and without sodium bicarbonate was purchased from Life Technologies (Grand Island, NY). Hydrocortisone was obtained from Collaborative Biomedical Products (Bedford, MA), and bovine pituitaries were obtained from Pel-Freez Biologicals (Rogers, AZ). Retinol, human epidermal growth factor (EGF), human apotransferrin, insulin, penicillin/streptomycin, gentamicin, L-cystine, sodium bicarbonate, and N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid buffer were all purchased from Sigma Chemical Co. (St. Louis, MO).

Microdissection and culture conditions. Rats were deeply anesthetized with sodium pentobarbital (60 mg/kg) and exsanguinated by severing the abdominal aorta and renal arteries. Immediately after death, the head of each rat was removed from the carcass. The lower jaw and skin were removed and the head was split in half along the medial longitudinal suture. The proximal nasal septum (between the anterior surface of the incisor teeth and the incisive papilla) and both maxilloturbinates were removed by careful blunt dissection using fine ophthalmic surgical instruments and an Olympus SZH10 stereo zoom microscope (Figure 1). The proximal septum (a rectangular piece of tissue 6 to 8 mm on a side) and individual maxilloturbinates (approximately 8 to 10 mm long and 2 to 3 mm wide) were placed in tissue culture-treated Transwell (Costar, Pleasanton, CA) mesh inserts (12-well plates with clear 0.4 µm pore-size inserts) and grown in culture for periods up to 14 d at the air-liquid interface. Proximal septa and maxilloturbinates were placed in the center of the mesh inserts. Medium was added to the lower (400 µl) and upper chambers (80 µl) to provide a thin film of fluid over the tissue. Culture medium (supplemented Ham's F12) was a modification of the culture media used to culture hamster (24) and rat (25, 26) tracheal epithelium. Nasal tissue explants were cultured for up to 14 d in serum-free medium (F12 supplemented with 0.1 µg/ml hydrocortisone, 5 µg/ml transferrin, 10 µg/ml insulin, 25 ng/ml EGF, 50 µg/ml bovine pituitary extract, 28 ng/ml retinol, 25 µg/ml gentamicin, 100 U/ml penicillin, and 100 µg/ml streptomycin) with complete replacement after 1 d in culture (DIC) and every other day thereafter.

View larger version (29K): [in this window] [in a new window]   Figure 1.   Tissue sample sites. (a) Exposed septum of nasal airway. Shaded area indicates the portion of the septum dissected for culture. Line 1 indicates the transverse section taken for morphometric analysis. (b) Exposed lateral wall of nasal airway. Shaded area indicates the maxilloturbinate microdissected for culture. Line 2 indicates the transverse section taken for morphometric analysis. (c) Illustration of the ciliated RE lining the proximal nasal septum. Arrows indicate mucous cells. (d) Illustration of the nonciliated transitional epithelium lining the maxilloturbinate. N = naris; B = brain; HP = hard palate; NP = nasopharynx.

Tissue fixation and processing. Proximal septa and maxilloturbinates were removed after 0, 3, 7, or 14 DIC and fixed for a minimum of 48 h in zinc-formalin for light microscopic analyses or 2% glutaraldehyde in 0.1 M sodium cacodylate buffer with 0.05% CaCl₂ for transmission electron microscopy. Maxilloturbinates were decalcified before being embedded in either 13% formic acid (for light microscopic analysis) or 10% ethylenediaminetetraacetic acid (EDTA) (for transmission electron microscopy).

Morphometric Analyses

Epithelial cell numeric densities. Only the RE and NTE overlaying the proximal septa and maxilloturbinates, respectively, were analyzed. One-micron-thick, toluidine blue- stained, plastic sections were analyzed at a magnification of ×1000. Total epithelial cell numeric densities were determined by counting the number of epithelial cell nuclear profiles present in the surface epithelium of the proximal septa and maxilloturbinates and dividing by the length of the underlying basal lamina (12, 27). The length of the basal lamina present in each tissue section was calculated from the contour length of the digitized image of the basal lamina on a Power Macintosh 7100/66 computer using the public-domain NIH Image program (written by Wayne Rasband, U.S. National Institutes of Health; available from the Internet by anonymous ftp from ). The data are presented as the mean number of epithelial cells per millimeter of basal lamina ± the standard deviation (SD) (n = 3 per group). In addition, the numeric densities of each of the cell types present in the RE and NTE were determined by assigning each nuclear profile to a cell type category based on the following morphologic criteria. The RE covering the proximal septa contained basal cells (small ovoid to polygonal cells with little darkly staining cytoplasm; touching the basal lamina but not reaching the luminal surface), ciliated cells (tall cuboidal or columnar cells with very pale blue cytoplasm containing numerous cilia on their apical surface), secretory cells (tall cuboidal or columnar cells with cytoplasmic staining intermediate between basal and ciliated cells containing few to many intracellular secretory granules), and indeterminate (other) cells that could not be morphologically assigned to the previous groups. The NTE covering the maxilloturbinates contained basal cells, ciliated cells, and other cells (described previously) plus nonciliated, cuboidal cells (predominant cell type with intermediately staining cytoplasm; cells reaching the luminal surface had microvilli). Cell type-specific numeric densities were determined by dividing the total number of cells of each type by the total basal lamina length of that sample. The data are presented as the mean number of each cell type per millimeter of basal lamina ± SD (n = 3 per group).

Epithelial cell volume densities. The size and abundance of respiratory and transitional epithelium were analyzed by procedures discussed by Hyde and colleagues (28, 29) using the morphometric data acquisition software Stereology Toolbox v1.1 (Morphometrix, Davis, CA). All measurements were made using a ×100 objective and 1-µm-thick plastic sections. The proportion of the epithelium composed of basal cells, ciliated cells, mucous cells, and indeterminate (other) cells was estimated by point counting using a 100-point cycloid grid. The mass, as measured by volume (µm³) of epithelial cells per unit area (µm²) of basement membrane (V/S), was estimated from point and intercept counts using the equation V/S = (2.07 × Po)/ (Ibl × 2), where Po represents the points counted for each object of interest and Ibl represents the number of intercepts of the basal lamina with the cycloid arcs. Thickness of the epithelium was estimated from point and intercept counts from all of the epithelial cell types. The data are presented as the mean mass ± SD (n = 3 per group).

Stored intraepithelial mucosubstances. The volume density of stored intraepithelial mucosubstances (IM) in the surface epithelium overlying the proximal septa and maxilloturbinates was determined using image analysis and standard morphometric techniques (30, 31). Briefly, the area of AB/PAS-stained mucosubstances in the surface epithelium was estimated from the length of the circumscribed perimeter of the stained material using the NIH Image program described previously. The length of the basal lamina was calculated from the contour length of the digitized image of the basal lamina. The quantity of stored mucosubstances per unit area of basal lamina was determined as described by Harkema and associates (31) and expressed as the mean volume (nl) of IM/mm² of basal lamina ± SD (n = 6 per group).

Analysis of rMuc-5AC mRNA Levels

Steady-state levels of rMuc-5AC mRNA were determined in proximal septa and maxilloturbinates after 0, 3, 7, and 14 DIC using quantitative RT-PCR. This assay employs a recombinant competitor RNA (rcRNA), used as an internal standard (IS), that is reverse-transcribed and amplified in the same tubes as the target sequence (i.e., rMuc-5AC). Several publications have described protocols that use an IS that contains the same sequences recognized by the target-gene amplification primers but has a different-sized intervening sequence and therefore yields a different-sized RT-PCR product (32, 33). The concentration of target mRNA can be estimated by adding increasing amounts of IS to the RT-PCR mixtures that contain a constant amount of sample RNA. This approach provides an absolute experimental readout (e.g., molecules of target gene mRNA per unit sample) as opposed to a radioisotope "signal" or "fold" readout. The primary advantages of quantitative RT-PCR are that (1) it does not require normalization against a housekeeping gene, (2) it facilitates comparison of data generated in different experiments within a laboratory as well as comparison among different laboratories, and (3) it corrects for variability during both the RT and PCR steps. The design and use of the IS (described below) are based on the work of Vanden Heuvel and coworkers (33).

Design and synthesis of rat rMuc-5AC IS. The IS primers were synthesized by Southwest Scientific Research (Albuquerque, NM) with the following sequences: forward IS primer (5'-T7 promoter-rat rMuc-5AC forward-human glutathione-S-transferase [GST] forward-3') = 5'-TAATACGAC-TCACTATAGG-CATCATTCCTGTAGCAGTAGTGAGG-AGGCCATGGTTTGCAGGAA-3'; reverse IS primer (5'-poly d[T]₁₈-rat rMuc-5AC reverse-human GST reverse-3') = 5'-TTT-TTTTTTTTTTTTTTT-GGTACCCAGGTCTACACCTACTCCG-GTTGGGCTCAAATA- TA-CGGTGG-3'. The PCR steps for synthesizing the rMuc-5AC IS were conducted in a final volume of 50 µl containing PCR buffer (16.6 mM [NH₄]₂SO₄, 50 mM -mercaptoethanol, 6.8 mM EDTA, 67 mM Tris [pH 8.8], 0.1 mg/ml bovine serum albumin, 3 mM MgCl₂, 0.1 mM each deoxyribonucleotide [dNTP], 0.6 pmol each of the forward and reverse IS primers, 100 ng human genomic DNA, and 2.5 units Taq DNA Polymerase [AmpliTaq; Perkin-Elmer, Norwalk, CT]). PCR mixtures were incubated at 85°C before adding Taq, then heated to 94°C for 4 min and cycled 10 times through a "touchdown" (a 15-s denaturation step at 94°C, a 30-s annealing step starting at 65°C and decreasing by 1°C per cycle, and a 30-s elongation step at 72°C), followed by 25 cycles using a 15-s denaturation step at 94°C, a 30-s annealing step at 55°C, and a 30-s elongation step at 72°C. An additional 5-min extension at 72°C was included at the end of the last cycle. PCR-amplified products were purified using the Wizard PCR Prep DNA purification system (Promega, Madison, WI). After electrophoresis on a 3% Nusieve 3:1 agarose gel (FMC Bioproducts, Rockland, ME), a single faint band was observed at the anticipated size of approximately 300 base pairs (bp). The first-round PCR products were diluted 1:100 and multiple tubes were amplified, gel-purified, and pooled.

Isolation of total RNA from nasal tissues. Proximal septa and maxilloturbinates were removed from the Transwells and immediately homogenized in 0.5 ml of Tri-Reagent (Molecular Research Center, Cincinnati, OH) to stabilize the sample. The homogenates were snap-frozen in liquid nitrogen and stored at 80°C. Total cellular RNA was isolated according to the protocol provided with the Tri-Reagent. To avoid DNA contamination, samples were incubated with a DNase solution (100 units rRNasin [Promega], 30 mM dithiothreitol [Life Technologies], and 10 units DNase [Boehringer Mannheim, Indianapolis, IN] in 5× transcription buffer [Promega]). After sequential extraction with phenol/chloroform/isoamyl alcohol (25:24:1) and chloroform/isoamyl alcohol (24:1), the RNA was precipitated with 1/3 volume 10 M ammonium acetate and 1.5 volumes isopropanol. The pellets were washed with 75% ethanol, air-dried, and resuspended in nuclease-free water containing rRNasin (40 units/100 µl). Total RNA was quantitated by measuring absorbance at 260 nm. RNA integrity was assessed by running 1 µg total RNA on a 1% SeaKem LE agarose gel (FMC Bioproducts).

Quantitative RT-PCR. RT-PCR was performed as outlined by Gilliland and colleagues (34, 35), except that known amounts of IS rcRNA were reverse-transcribed into complementary DNA (cDNA) in a volume of 20 µl containing PCR buffer plus 5 mM MgCl₂, 1 mM each dNTP, 10 units rRNasin, 125 ng oligo(dT)_12-18 (Becton Dickinson, Bedford, MD), 100 ng total RNA from proximal septa or maxilloturbinates, and 40 units MMLV reverse transcriptase (Promega). For each RNA sample, eight aliquots were made, to which a range of IS rcRNA molecules were added. First, a wide range-finding experiment was performed on pooled RNA samples from proximal septa and maxilloturbinates isolated after 0, 3, 7, and 14 DIC. Ten-fold serial dilutions of the internal standard were added to each tissue RNA sample to yield a range of 10⁹ to 10² IS molecules per reaction tube. The RNA samples were incubated at 42°C for 15 min, followed by 95°C for 5 min. A PCR master-mix consisting of PCR buffer, 4 mM MgCl₂, 6 pmol each of rat rMuc-5AC forward (5'-CATCATT-CC-TGTAGCAGTAGTGAGG-3') and reverse (5'-GGTACCCAGGTCTACACCTACTCCG-3') primers, and 1.25 units Taq DNA polymerase were added to the cDNA samples, for a final volume of 50 µl. Taq polymerase was added after samples had been heated to 85°C for 5 min. Immediately, samples were heated to 95°C for 4 min and then cycled (28 times for proximal septa, 36 times for maxilloturbinates) at 94°C for 30 s, 56°C for 30 s, and 72°C for 30 s, after which an additional final extension step at 72°C for 10 min was included.

Statistical Analyses

The morphometric and quantitative mRNA data were evaluated for the potential effects of time in culture using a one-way analysis of variance. Significant differences were determined by using a post-hoc multiple comparison procedure (Tukey test) to identify the source of the variance. The mRNA data were assessed after a natural logarithmic transformation. Statistical analyses were performed using the SigmaStat software program (Jandel Scientific, San Rafael, CA). The data are expressed as the group mean ± SD.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Culture Conditions and Tissue Morphology

The proximal nasal septa and the maxilloturbinates are richly supplied with blood in vivo, with numerous capillaries and large-capacitance blood vessels (swell bodies) underlying the surface epithelium. The blood vessels and capillaries were clearly visible under the dissecting microscope when the tissues were first placed in culture. However, with increasing time in culture the tissues became more pale and the blood vessels were no longer visible. There were numerous ciliated cells in the RE lining the surface of the proximal nasal septum. The NTE lining the maxilloturbinates had few ciliated cells, most of which were found in the distal region. The cilia continued to beat throughout the 14 DIC and maintained a continuous flow of material (i.e., media, cellular secretions) across the apical surface of the tissues exposed to air.

Epithelial Morphology

The epithelium at the air-liquid interface remained healthy and retained the normal morphology (Figures 2 and 3). The majority of cells within the lamina propria died during the 2 wk in culture. Numerous apoptotic cells with highly fragmented nuclei were evident. After 14 DIC, the normal morphology of the RE and NTE in contact with the microporous membrane was lost and replaced by a poorly differentiated squamous epithelium, 1 to 2 cell layers thick.

View larger version (125K): [in this window] [in a new window]   Figure 2.   Light micrographs of nasal septum placed in culture for 0 (a), 3 (b), 7 (c), or 14 (d) d. One-micron-thick epon/araldite sections were stained with toluidine blue. Arrowheads indicate basal lamina between epithelium and lamina propria; arrows indicate mucous (goblet) cells; e = epithelium; lp = lamina propria containing blood vessels, submucosal glands, and ducts.

View larger version (118K): [in this window] [in a new window]   Figure 3.   Light micrographs of transitional epithelium of maxilloturbinates placed in culture for 0 (a), 3 (b), 7 (c), or 14 (d) d. One-micron-thick epon/araldite sections were stained with toluidine blue. Arrowheads indicate basal lamina between epithelium and lamina propria; e = epithelium; lp = lamina propria containing blood vessels, submucosal glands, and ducts; tb = turbinate bone.

In contrast to the epithelium in contact with the Transwell inserts, no evidence of squamous metaplasia was observed in RE or NTE cultured at the air-liquid interface. Only a few minor ultrastructural changes were noted in the various epithelial cell populations in either the NTE or the RE during time in culture. Compared with the RE, in vivo, there was a progressive reduction in the height of ciliated and secretory cells with time in culture (Figure 2). By 14 DIC, these cells had changed from tall columnar to short columnar or tall cuboidal. No other cytoplasmic or nuclear changes were morphologically evident, with the exception of a conspicuous reduction of secretory products in the mucous cells by 14 DIC (Figure 4). Goblet cells still contained electron-lucent granules in the apical cytoplasm (although there were fewer granules in the RE cultured for 14 d than in the control RE).

View larger version (159K): [in this window] [in a new window]   Figure 4.   Transmission electron photomicrographs of nasal septum placed in culture for 0 (a) or 14 (b) d. Arrowheads indicate basal lamina between epithelium and lamina propria; M = mucous (goblet) cell; Ci = ciliated cell; B = basal cell.

Cuboidal cells of the transitional epithelium that line the maxilloturbinates also retained their ultrastuctural appearance throughout 14 DIC with no apparent alteration in size or shape (Figure 5). Nonciliated cuboidal cells had microvilli covering the luminal surface and numerous mitochondria similar to control tissue after 14 DIC. There was an increase in the thickness of the transitional epithelium at 14 DIC, compared with in vivo controls, due to an increase in the number of basal cells present.

View larger version (129K): [in this window] [in a new window]   Figure 5.   Transmission electron photomicrographs of NTE placed in culture for 0 (a) or 14 (b) d. Arrowheads indicate basal lamina between epithelium and lamina propria; C = cuboidal cell; B = basal cell.

Epithelial Morphometry

We used image analysis and standard morphometric techniques to evaluate the effects of in vitro culture on RE and NTE morphology. Volumetric data was used to determine changes in epithelial thickness and the mass (volume density) of each cell type. Epithelial-cell numeric densities were determined to evaluate culture-induced changes in epithelial-cell populations.

Proximal septum. As previously mentioned, there was a decrease in the average thickness of the RE overlying the proximal septum with increasing time in culture (Figure 6a). Compared with microdissected proximal septa that were not placed in culture (0 DIC; control), there was a 14% decrease in the average thickness after 7 DIC and a 33% decrease in thickness after 14 DIC (1.5-fold thinner than control septa). Table 1 summarizes the volumetric changes in the individual cell types within the RE (i.e., mucous, ciliated, basal, and other). The reduction in epithelial thickness was associated with a decrease in the volume density (mass) of mucous secretory cells after 7 (45% decrease) and 14 (47% decrease) DIC. There were no changes in the volume density of ciliated or basal cells during the 14 DIC.

View larger version (26K): [in this window] [in a new window]   Figure 6.   Numeric cell density (open bars) and epithelial thickness (closed circles) of RE (a) and transitional epithelium (b) after 0, 3, 7, and 14 DIC. Error bars denote SD. *Significantly different from 0 DIC, P < 0.05.

Maxilloturbinates. There was no change in the thickness of the NTE and no change in the volume densities of the constituent cells within the NTE (cuboidal, ciliated, basal, and other) after 3, 7, or 14 DIC, compared with microdissected (control) maxilloturbinates that were not placed in culture (Figure 6b and Table 3, respectively). Although there was no change in NTE thickness, there was a 35% increase in the number of NTE cells/mm basal lamina in maxilloturbinates that were in culture for 14 d (Figure 6b). The increase was due to an approximately twofold increase in the number of basal cells present in the epithelium and a slight increase in the number of cuboidal cells present in the NTE (Table 4).

IM

There were no measurable AB/PAS-stainable IM in the NTE of control maxilloturbinates (0 DIC) or in the NTE of maxilloturbinates after 3, 7, or 14 DIC. In contrast, the RE covering the proximal septum contained numerous mucous cells with copious amounts of stored IM (Figure 7a). The amounts of stored mucosubstances within the RE decreased with increasing time in culture (Figures 7 and 8). There was a 44% decrease in stored mucosubstances after 3 DIC. There was an additional 37% decrease in stored mucosubstances between 3 and 7 DIC (2.9-fold less than 0 DIC, controls), and a further 13% decrease between 7 and 14 DIC (3.3-fold less than 0-d controls).

View larger version (144K): [in this window] [in a new window]   Figure 7.   Light micrographs of stored mucosubstances in nasal septum placed in culture for 0 (a), 3 (b), 7 (c), or 14 (d) d. One-micron-thick epon/araldite sections were stained with AB/PAS to detect acidic and neutral mucosubstances. Arrows indicate mucous (goblet) cells; e = epithelium; lp = lamina propria containing blood vessels, submucosal glands, and ducts.

View larger version (9K): [in this window] [in a new window]   Figure 8.   Volume density of AB/PAS-positive mucosubstances in the RE lining nasal septum placed in culture for 0, 3, 7, or 14 d. Error bars denote SD. *Significantly different from 0-DIC controls, P < 0.05.

rMuc-5AC mRNA Levels

Steady-state levels of rMuc-5AC mRNA present in proximal nasal septa and maxilloturbinates cultured for 0, 3, 7, and 14 d were determined using a quantitative RT-PCR assay. Basal levels (0 DIC) of rMuc-5AC mRNA were 66-fold greater in the proximal nasal septa lined with secretory RE compared with maxilloturbinates that are lined by nonsecretory NTE (Figure 9). There was a 35-fold decrease in septal rMuc-5AC mRNA during the first 3 d of culture (Figure 9a). After 14 DIC the steady-state level of rMuc-5AC mRNA in the proximal nasal septa was 13% of basal levels (222 ± 39 and 1713 ± 465 aM, respectively).

View larger version (16K): [in this window] [in a new window]   Figure 9.   Amount of rMuc-5AC mRNA expressed in RE (a) and transitional epithelium (b) after 0, 3, 7, and 14 DIC. Error bars denote SD. *Significantly different from 0 DIC, P < 0.05.

Although maxilloturbinates had no measurable stored IM, they did express low levels of rMuc-5AC mRNA (Figure 9b). Similar to the proximal nasal septa, there was a large (23-fold) decrease in steady-state levels of mucin mRNA after 3 DIC. Four days later (7 DIC), the amount of rMuc-5AC mRNA per microgram of RNA was two times greater than control (0 DIC) maxilloturbinates. However, after a total of 14 DIC, the amount of rMuc-5AC mRNA per microgram of RNA had decreased again to 30% of basal (0 DIC, control) levels.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study we have demonstrated that microdissected nasal airway tissues can be maintained in culture, at an air-liquid interface, for up to 14 d with minimal effects on the morphology of the surface epithelial cells. Our goal was to develop a culture system that could be used to investigate mechanisms of toxicant-induced epithelial injury and repair, mucous-cell differentiation, and mucin gene expression in the absence of inflammatory cells or systemic influences that exist in vivo. We selected the proximal nasal septum, which is covered by an RE with numerous mucous (goblet) secretory cells, to determine whether we could maintain a mucous-cell phenotype using this culture system. We also cultured microdissected maxilloturbinates, which are covered by a sparsely ciliated, cuboidal NTE, to determine the effect of this culture system on this normally nonsecretory epithelium. These tissues represent common sites of toxicant-induced epithelial injury and repair in rodent nasal airways and have distinctly different phenotypic responses to exposure to the same toxicant. For example, repeated ozone exposure (0.5 ppm ozone, 8 h/d for 5 d to 13 wk) induces MCM in the NTE lining maxilloturbinates but has no effect on the RE lining the proximal nasal septum (4, 13).

After 2 wk in culture the RE covering the proximal nasal septum was thinner, had less stored IM, and had fewer mucous and ciliated cells. The decrease in the amount of stored IM reflected both a decrease in the total number of mucous secretory cells present in the epithelium and a decrease in the amount of stored material present in each mucous cell. It did not reflect a change in the type of secretory product present in the cells. Similar to the RE covering the proximal septum in vivo, or immediately after being removed by microdissection, the secretory cells present in the RE after 3, 7, or 14 DIC continued to have large electron-lucent secretory granules within their cytoplasm which contained both acidic (AB-positive) and neutral (PAS-positive) mucosubstances. The decrease in total surface epithelial cells was due to a loss of both ciliated and mucous cells. It is interesting to note that the numeric densities of ciliated and mucous cells within the RE covering the proximal nasal septum were similar before culture, and the relative abundance of ciliated cells compared with mucous cells was retained throughout 2 wk in culture.

In contrast to the proximal nasal septum, there was no change in the thickness of the NTE covering the maxilloturbinates during the 2 wk in culture, but there was an increase in the number of surface epithelial cells due to a proliferation of basal cells (i.e., basal cell hyperplasia). The NTE is unique because even though it normally has few to no secretory cells, it can rapidly undergo MCM after exposure to ozone (6, 27, 30).

The similarities of our nasal explant culture method to others in the literature include using air-liquid interface to maintain a fully differentiated airway epithelial morphology. Human nasal mucosal explants have been cultured for up to 48 h on a gelatin sponge-supported histoculture (36, 37) or on plastic plates on a rocking platform for up to 6 d (38). Human nasal epithelial cell isolates have also been cultured successfully at an air-liquid interface (39). To create this interface, explants (or cells) are placed on a porous-bottomed insert suspended in a tissue culture well. Medium is added to the basolateral aspect of the explants (i.e., in the bottom reservoir) so that the mucosal face of the explants/cells on the insert is covered by a thin layer of fluid. Levels of aerobic respiration in tissue grown at an air-liquid interface are higher than when immersed (40) and this biphasic system provides an environment similar to that found in the airway lumen in vivo (41).

Other laboratories have had success maintaining differentiated rat tracheobronchial epithelium in vitro (42, 43). This histotypic method requires that airway epithelial cells be dissociated from the basement membrane and then plated on collagen inserts at an air-liquid interface and allowed to re-differentiate. Although this technique requires time (approximately 8 d) to reestablish a fully differentiated pseudostratified epithelium, this method is excellent for airway epithelial cells from large-diameter pulmonary airways. Also, having a pseudostratified layer of cells on the mesh inserts allows the introduction of additives to the basolateral side of the cells, mimicking systemic influx in vivo. Our explant method does not allow for this because there is bone or cartilage between the epithelial surface of interest and the bottom reservoir.

In contrast to methods involving cell dissociation, the method described in the present paper is quick, gentle, and maintains normal cell-cell contacts within the surface epithelium. This results in a population of fully differentiated cells with normal epithelial organization and structure that is immediately available for experimental manipulation and that can be maintained for up to 14 DIC. This method is very similar to and based upon the one used by Van Winkle and colleagues (44) to maintain distal bronchiolar epithelial cells for up to 7 DIC.

Although the overproduction and hypersecretion of airway mucins is a common pathologic feature of several chronic human respiratory diseases (e.g., asthma, chronic bronchitis and rhinitis, cystic fibrosis), the cellular and molecular mechanisms involved in the development and persistence of these hypersecretory airway diseases are unknown. This culture system will be a useful tool to investigate mechanisms of toxicant- or allergen-induced MCM and mucin gene regulation. We have previously reported that intratracheal instillation of bacterial endotoxin rapidly induces MCM, with increased amounts of stored mucosubstances, within the RE lining rat pulmonary airways in vivo (7, 8). We have recently extended these in vivo experiments by demonstrating that endotoxin can induce MCM with increased amounts of stored mucosubstances in a morphologically similar RE lining rat distal nasal septa in vitro (45). The endotoxin-induced increase in stored mucosubstances was dose-dependent, and occurred within 72 h of endotoxin exposure. This culture system can therefore not only maintain normal, fully differentiated, airway epithelium in vitro for up to 2 wk but also support characteristic toxicant-induced epithelial responses normally observed in vivo (i.e., MCM).

In the present study we found that in the RE the steady-state levels of rMuc-5AC mRNA roughly correlated with the amount of histochemically stainable mucosubstances present in the surface epithelium. The rMuc-5AC mRNA levels, the number of mucous cells, and the amount of stored IM all decreased with time in culture. Decreases in differentiated cell function have also been reported in the culture of isolated murine intestinal villus and crypt cells (46), isolated rat pulmonary Clara and alveolar type 2 cells (47), and microdissected murine pulmonary bronchioles (44). In contrast to the RE, the NTE had no identifiable mucous cells and no stored mucosubstances, although it did have measurable steady-state levels of rMuc-5AC. This apparent discrepancy may reflect the sensitivity of the method we used to quantitate rMuc-5AC mRNA (i.e., detecting rMuc-5AC mRNA in a small number of mucous cells present in the tissue but not observed in tissue sections). Alternatively, it may indicate that rMuc-5AC apomucin expression is post-transcriptionally regulated.

In conclusion, RE and NTE were maintained in vitro for up to 14 d using explant cultures of rat nasal septa and maxilloturbinates. These cultured nasal epithelia retained histochemical markers of differentiated cell types (acidic and neutral mucosubstances), preserved cell viability and structural integrity at the air-liquid interface, and maintained levels of mucin-specific rMuc-5AC mRNA. The results of this study indicate that this culture system will be a useful tool to investigate mechanisms of toxicant-induced epithelial injury and repair, MCM, as mucin gene regulation. In the future, this approach will permit the use of microdissected human airway tissues to help bridge the gap between our understanding of the pathogenesis of toxicant-induced MCM in rodents and the events that take place in normal and diseased human airways.