Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The adverse health effects of ozone are well documented (1). In clinical studies, respiratory inflammation and respiratory symptoms occur with exercise in an atmosphere of ozone as low as 0.08 ppm (2) and ozone exposure has been shown to exacerbate the frequency and severity of asthma attacks (3). Epidemiologic studies have shown strong correlations between seasonal ozone levels and the incidence of nasal mucosal atrophy, elevated nasal polymorphonuclear leukocyte levels, and asthma symptoms in children living in polluted urban environments where ozone levels episodically exceed 0.3 ppm (4, 5). However, the cellular and molecular mechanisms involved in ozone toxicity to pulmonary tissues are not entirely understood.

Epithelial cells can be considered the first line of defense as well as the first cellular target of respiratory damage affected by ozone. As reviewed by Leikauf and colleagues (6), ozone toxicity in respiratory tissue can be detected in different chronological stages. The first, involving interaction of ozone with the epithelial lining fluid and cell membrane components, results in the formation of secondary ozonolysis products which, along with ozone itself, may convey the oxidant stress to the epithelial cells and the subepithelial tissues. We have, therefore, turned to primary cultures of nasal epithelia in order to study the earliest effects of ozone.

Repeated exposure to 0.5 ppm ozone induces mucous cell metaplasia in the nasal transitional epithelium of rats (7) in which the acute response includes a transient increase in the epithelial labeling index, which peaks at 20- 24 h after exposure, and a concomitant return of cell density to control levels (8). Ozone-induced airway hyperresponsiveness (hypersensitive airway constriction) in guinea pigs is maximal at 2 to 5 h after exposure and occurs concomitantly with epithelial detachment and histologically demonstrable disruption (9). The epithelial cell layer is an important barrier between airborne pollutants and the tissue interstitium through the presentation of antioxidant scavengers (10) and as a barrier to ionic flux (11). Immediately following 3 h of exposure to ozone, cell morphology is altered through changes in actin polymerization, resulting in increased pericellular permeability (11). Similar effects on actin polymerization have been observed in endothelial cells exposed to hydrogen peroxide (12); however, the mechanism by which ozone alters epithelial cell morphology is not known.

Maintenance of a differentiated, nonproliferative state in epithelial cells as well as the barrier function of the epithelium depends upon the composition of the extracellular matrix (ECM) and the presence of matrix ligands on the cells (13). Thus, changes in composition of the epithelial ECM could be an early stage in the pathologic processes initiated by exposure of respiratory tissues to ozone. In the present study, we examined the effect of ozone exposure on mRNA levels for five different ECM components in primary cultures of primate nasal epithelial cells (PNE). Tenascin-C, a molecule associated with detachment and migration of cells in several different systems (16), displayed the greatest change in mRNA levels in response to ozone. Further investigation showed that exposure to ozone increased the expression and organization of tenascin-C in cultured PNE as determined by Northern and Western analyses and immunohistochemistry.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Cell Culture

Primate nasal tissue was obtained from Macaca nemestrina monkeys. Animals were free of respiratory disease and were killed for other experimental purposes under the guidelines of the Animal Care Committee at the University of Washington. Immediately after necropsy, specimens were placed in 1:1 Dulbecco's Modified Eagle's Medium: Ham's F-12 medium supplemented with 10% NuSerum, 2 mM L-glutamine, 10 mM Hepes, 100 U/ml penicillin, 100 µg/ml streptomycin sulfate, and 2.5 µg/ml fungisone (designated as "E medium"). Once in the laboratory, tissue specimens were washed with fresh E-medium and incubated overnight at 4°C in 0.4% dispase in phosphate-buffered saline (PBS) without calcium or magnesium. The epithelium was physically separated from the connective tissue as intact sheets by gentle scraping with a scalpel. The resulting cells were pelleted by centrifugation and agitated for 5 min in trypsin, to dissociate the cells. The cell suspension was seeded at 2 × 10⁶ cells per well onto collagen-coated tissue culture semipermeable inserts (Costar, Cambridge, MA) and, after 24 h incubation under medium, maintained at the interface between air and tissue-culture medium. This procedure results in the development of a differentiated stratified cuboidal epithelium (17). In our hands, cells cultured in such a fashion may begin to senesce after 12 to 14 d. In an effort to standardize our exposure protocols, all exposures were carried out at 6-8 d, when the cultures consisted of confluent bilayers prior to the onset of senescence. Cells cultured in our laboratory under these conditions included ciliated cells, as was determined by scanning electron microscopy (Figure 1). The entire ciliated epithelium from maxilloturbinate and ethmoturbinate regions of the superior, medial, and inferior turbinates was used in the preparations where the dispase treatment allows removal of all of the epithelial cells residing above the basement membrane. This was confirmed by examination of cryosections of the digested tissue (data not shown). Representative intact portions from dissected turbinates were collected for tenascin-C immunostaining without reference to their specific original location within the nasal structure.

View larger version (213K): [in this window] [in a new window]   Figure 1.   Low-magnification scanning electron micrograph of an air-exposed control culture showing a portion of the culture surface. Cells grown at an air-liquid interface form a confluent mat of cellsmost with a microvillous surface but some cells, as shown here, are differentiating as ciliated cells. Bar = 20 µm.

Air Chemistry and Exposure Conditions

Cell cultures were exposed in an environmental exposure system which permitted simultaneous exposure of cells to either filtered air alone or filtered air with 0.3 ppm ozone (18). Pairs of six well cell culture plates containing epithelial cells were simultaneously exposed for 3 h in identical 0.5-cubic-foot exposure chambers above a water bath maintained at 37°C. The gaseous pollutant/air mixture was added to the chamber at 40 liters/min and exited via exhaust ports adjusted to maintain pressure at 1 atm within each chamber. The ozone concentration was monitored in the exposure chamber above the cell cultures. Humidity was maintained at 95% and continuously monitored with a dew point hygrometer (EG & C Model 880; Waltham, MA). Both chambers received 5% CO₂ as monitored with a medical gas analyzer (Beckman Model LB-2; Irvine, CA). The system permitted treatment and control chambers to be randomized. Ozone was produced by ultraviolet irradiation of clean compressed air (OREC Model 03V1-0; Ozone Research and Equip. Co., Phoenix, AZ) and monitored with an ultraviolet photometric analyzer (Model 1003; Dasibi, Glendale, CA). The ozone concentration and duration of exposure were selected to approximate the episodically high levels which are detected in urban environments (1) while sample times were chosen to replicate the observations from other studies of acute ozone toxicity in vivo and in vitro (6, 8, 9, 19).

Northern Blot Analysis

Total RNA was isolated from cultured epithelial cells using guanidine isothiocyanate solubilization followed by phenol extraction at pH 4 and subsequent precipitation in isopropanol (20). Purified samples were fractionated on 1.0 or 0.8% formaldehyde-agarose gels, transferred to nylon blotting membranes (Zeta Probe; Bio-Rad Laboratories, Richmond, CA), and hybridized to a primate-specific cDNA probe for tenascin-C which was generated in our laboratory by polymerase chain reaction cloning and which hybridized to a region of tenascin-C common to all splice variants (21). In addition, blots were hybridized to cDNA probes to the following human matrix molecules: human versican, clone 7 (24); human biglycan (25), clone p16; human collagen type I, clone HF677 (26); and human perlecan, clone HS-1 (27). These were generous gifts from, respectively, Erkki Ruoslahti, La Jolla Cancer Research Foundation, La Jolla, CA; Larry Fisher, National Institute of Dental Research, Bethesda, MD; Francisco Ramirez, Mt. Sinai School of Medicine, New York, NY; and Renato Iozzo, Thomas Jefferson University, Philadelphia, PA. The prehybridizing and hybridizing solutions contained 50% formamide, 5× Denhardt's solution, 6× standard sodium phosphate and EDTA buffer (SSPE), 0.5% sodium dodecyl sulfate (SDS), and 100 µg/ml salmon testes DNA. cDNA probes were labeled with [³²P]dCTP by random priming and added to the hybridizing solutions at a final concentration of of 2 × 10⁶ cpm/ml. Hybridizations were performed at 42°C for 15 to 20 h. All blots were washed twice for 10 min in 2× SSPE and 0.1% SDS and twice for 30 min in 2× SSPE and 0.1% SDS at 42°C. Blots hybridized to cDNA for tenascin-C were then washed twice for 15 min in 0.3× SSPE and 0.1% SDS at 42°C, and twice for 15 min in 0.3× SSPE and 0.1% SDS at 55°C. All other blots were washed twice in 0.3× SSPE and 0.1% SDS for 30 min at 65°C. Exposures were done on Kodak XAR-2 film at 70°C. For quantitation of mRNA levels, autoradiograms were analyzed using densitometric scanning and normalized to the amount of 28S mitochondrial RNA as revealed by ethidium bromide staining. Each blot was probed one time for each specific mRNA. Normalized data from different experiments, using cells from different animals, were averaged to obtain the average increase in mRNA present in ozone-treated cells over controls.

Immunohistochemistry

Intact nasal tissue and air- and ozone-exposed cultures were embedded in Tissue-Tek O.C.T. compound (Miles, Elkhart, IN) for cryosectioning. Ten-micrometer-thick sections were stained with hematoxylin and eosin or with methyl green and F9A5 anti-tenascin-C antibodies, a generous gift from William Carter (Fred Hutchinson Cancer Research Center, Seattle, WA) (28). Bound F9A5 was detected using a biotin-linked secondary antibody in conjunction with the avidin-peroxidase method (Vector, Burlingame, CA) resulting in the precipitation of a brown product of diaminobenzoate (DAB). Cultures were prepared for en face immunostaining by permeabilization for 1 min in 0.05% Triton-X-100 detergent in PBS, followed by fixation in 4.0% paraformaldehyde in PBS. Nuclei were stained with 4, 6 diamidino-2-phenylindole (DAPI) (Sigma, St. Louis, MO). Tenascin-C was visualized with a system utilizing a biotin-linked secondary antibody bound to streptavidin-Texas Red (Zymed, San Francisco, CA).

Western Blotting

Epithelial cell layers were scraped into 3-[cyclohexylamino]-1 propane sulfonic acid buffer at pH 8 containing the protease inhibitors N-ethylmaleimide, aprotinin, and phenylmethylsulfonyl fluoride. Samples were concentrated on 50K molecular weight cutoff Centricon membranes (Amicon, Beverly, MA), applied to 7.5% SDS-polyacrylamide gel electrophoresis (29) and electrophoretically transferred to nitrocellulose membranes (Schleicher and Schuell, Keene, NH) using a Mini Trans-Blot Cell (Bio-Rad, Hercules, CA). Tenascin-C was detected with antibody F9A5 and enhanced chemiluminescence (Western-Light Chemiluminescent Detection System with CSPD substrate; Tropix, Bedford, MA).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Ozone Selectively Increases Tenascin-C mRNA Levels

Cell layers were harvested for Northern analysis at 24 h after a 3-h exposure to 0.3 ppm ozone or air and examined for differences in the levels of mRNA for several matrix molecules. Tenascin-C mRNA transcripts were increased significantly, although considerable individual variation was observed (Figures 2a and 2b). No significant difference was found between air- and ozone-treated cells for other ECM molecules such as biglycan, versican, perlecan, or collagen type I mRNA.

View larger version (24K): [in this window] [in a new window]   Figure 2.   Northern blot analysis of PNE after exposure to 0.3 ppm ozone for 3 h. Cell layers were harvested at 24 h after exposure to ozone. Total RNA was hybridized to cDNA probes specific for tenascin, biglycan, versican, perlecan, or collagen type I. Representative data are presented in (a), where increased tenascin-C message in response to ozone was detected, whereas no differences were observed for biglycan, versican, perlecan, or collagen type I. The results of several different normalized experiments are presented in (b), where the amount of message in ozone-treated cultures is expressed as percent increase over controls.

Increased Levels of Tenascin-C are Detected by Western Blotting in PNE After Ozone Exposure

PNE cell layers were subjected to Western blotting with anti-tenascin-C antibody at 24 h after exposure to ozone or air. In three separate experiments, more tenascin-C was detected in the ozone-treated cell layers than in the air-treated controls (Figure 3).

View larger version (20K): [in this window] [in a new window]   Figure 3.   Equal numbers of air- and ozone-treated cultures of PNE were harvested for Western analysis at 24 h after exposure, hybridized to anti-tenascin-C antibody, F9A5, and visualized by reaction with a chemiluminescent substrate. A representative Western blot is presented in (a). The results of three different normalized experiments are presented in (b), where the amount of tenascin-C in ozone-treated cultures is expressed as percent of control.

Tenascin-C Immunostaining is Present in the Normal Nasal Mucosal Subepithelial Space

Although the presence of tenascin-C in the bronchial subepithelial space has been demonstrated (30), it has not previously been found in nasal tissue. To determine the location of tenascin in uninjured nasal epithelial tissue in vivo, cryosections of intact nasal turbinate were labeled with anti-tenascin-C antibody and visualized with the DAB reaction method. Tenascin-C staining was restricted to the subepithelial spaces of the airway surfaces and secretory ducts, with the most prominent immunoreactivity occurring in the ductal regions. The mucosal subepithelial space contained regions which were alternately positive or negative for tenascin-C (Figure 4c). To test the possibility that cells isolated for experimental treatments contained tenascin-C from the subepithelial space, mucosal epithelia released by dispase treatment were immunostained with anti-tenascin-C antibody. No tenascin-C was detected (Figure 4a), suggesting that tenascin-C detected in cell cultures is due to synthesis in vitro.

View larger version (130K): [in this window] [in a new window]   Figure 4.   Immunostaining for tenascin in PNE in culture and in intact tissue. Cryosections of dispase released nasal epithelium (a, b) and intact nasal turbinate (c, d), were labeled with anti-tenascin-C antibody-mediated DAB product (brown) (a, c) or hematoxylin and eosin (b, d). Narrow arrows indicate the abluminal side of the epithelial cell layer. Tenascin-C (wide arrows) was detected in the basement membrane of intact epithelial cells but was absent from dispase-released cells. The tissue in (a, c) was not counterstained but, to increase its visibility, the color balance of the photograph was altered in favor of magenta by Adobe Photoshop software (Adobe Systems, Mountain View, CA). PNE were cultured on permeable membranes at an air-medium interface (e, f  ) and exposed to air (e) or 0.3 ppm ozone ( f  ) for 3 h and harvested for cryosectioning 7 h after exposure. Brown DAB reaction product linked to anti-tenascin-C antibody was light and diffuse in control cultures but intense on the lateral and basal borders of the apical layer cells and intermittent between the basal cells and the cell support in ozone-treated cultures. The cultures were counterstained with methyl green. Bar: (a, b, c, d = 25 µm; e, f = 10 µm).

Tenascin-C is Deposited by PNE Exposed to Ozone

Air- and ozone-treated PNE cultures were sectioned for immunostaining with anti-tenascin-C antibody, and representative micrographs are shown in Figures 4e and 4f. Air-treated cultures consisted of two to three layers of epithelial cells which were adherent to the synthetic culture substrate and showed very little tenascin-C immunostaining. In contrast, cultures exposed to ozone had a markedly different morphology and exhibited regions of prominent tenascin-C immunoreactivity. Although the morphology of the uppermost or apical cell layer was similar to that of the air-treated controls, the lateral and basal borders of the apical layer cells in ozone-exposed cultures were delineated by tenascin-C immunostaining. Moreover, regions of ozone-treated cultures were detached from the underlying support, with cells in the basal and intermediate layers in such regions appearing disrupted, as if by lytic or degradative processes. Punctate deposits of tenascin-C staining were detected subjacent to the basal layer in these regions and between the cell layer and the culture support.

En face immunofluorescent staining for tenascin in ozone-treated cultures revealed regions of tenascin-C deposition in areas which were negative for nuclear staining, indicating that tenascin-C was present on the Teflon membranes in areas devoid of cells (Figures 5b and 5d). Moreover, the ozone-treated cultures contained many areas devoid of cells, whereas there were only a few such areas in the control cultures and these did not stain brightly for tenascin-C (Figure 5a). The denuded areas were negative when stained with only the secondary antibody (Figure 5c). The presence or absence of tenascin-C in areas of the culture where cells remain may not be determined by this method.

View larger version (88K): [in this window] [in a new window]   Figure 5.   PNE were cultured on permeable membranes at an air- medium interface, exposed to air or 0.3 ppm ozone for 3 h, and prepared for en face immunofluorescent staining with DAPI (blue, nuclei) and Texas red linked to anti-tenascin-C antibody, F9A5 (extracellular tenascin-C). Red fluoresence showed tenascin-C on the denuded portions of the membrane only in cultures exposed to ozone. Air (a), ozone (b, c, d). Secondary antibody only (c). Bar = 100 µm.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

We found that exposure to ozone increases the expression of tenascin-C, but not perlecan, versican, collagen type I, or biglycan, in cultured PNE. In response to ozone, extracellular tenascin-C outlined the basal and lateral aspects of apical layer cells and was also deposited between the cell layer and the membrane on which cells are grown. En face immunostaining also showed tenascin-C on denuded areas of the matrix in ozone-treated cultures.

Tenascin-C is a member of the tenascin gene family which is implicated in numerous developmental processes (31). Tenascin-C causes increased cell rounding and reduced substrate attachment in vitro (16). These cellular changes facilitate migration, proliferation, and apoptosis, all of which occur in tissues where tenascin-C is transiently upregulated in vivo: in wound healing, tumor invasion, and developmental histogenesis (14, 32). Immunohistochemical data reported here demonstrate that 7 h after ozone exposure, tenascin-C immunostaining is abundant in ozone-treated cultures in the same regions where cell attachment to the subcellular matrix is greatly reduced. This is consistent not only with the findings of Cheek and associates (35), who found focal areas of epithelial degradation in cultured primary type II alveolar epithelial cells after ozone exposure, but also with our previous report that exposure of cultured PNE to ozone resulted in a consistent reduction in cell number to approximately 90% of control (19). We have found similar reductions in cell number during this current set of experiments. Although the apical cell layer was exposed directly to ozone and showed the most intense tenascin-C staining, morphologic effects were more evident in the basal cell layer. While the reason for this difference is not apparent, it may be due to the formation of reactive oxidative intermediates, secretion of cytokines (6, 36, 37), or increased susceptibility of basal layer cells. In any case, it is plausible that selective secretion of tenascin-C by ozone-treated epithelial cells occurs concomitantly with morphologic changes resulting in cell detachment and cell loss.

Ligands for tenascin-C include fibronectin and perlecan in the ECM, and syndecan and the integrins ₅₆, ₂₁, _v6, and ₉₁ at the cell surface. The latter two integrins, as well as tenascin-C, are present in lung tissue (30). The integrin _v6 binds either tenascin-C or fibronectin; is tightly linked to morphogenetic events, tumorigenesis, and epithelial repair (38); and is upregulated in injured human lung tissue (30). In vivo exposure of respiratory epithelium to ozone (0.1 to 2.0 ppm) causes increased epithelial synthesis of fibronectin and collagen type I mRNA (39), along with epithelial damage and a variety of inflammatory responses (19, 40). Elevated fibronectin secretion also occurs in immortalized epithelial cells in response to ozone (44), and we have recently found that ozone exposure causes the fibronectin receptor, integrin ₅₁, to be enriched on the apical surface of primary respiratory epithelial cell cultures (A. Jacob Jabbour, Department of Environmental Health, University of Washington, Seattle, WA; personal communication). Thus, the selective increase in tenascin-C expression we observed in vitro is consistent not only with an increase in fibronectin, a potential ligand for tenascin-C, and collagen type I, but is also in keeping with changes in the expression and distribution patterns of cell surface ligands for both tenascin-C and fibronectin.

Epithelial cell injury in vivo, regardless of its cause, results in a series of increasingly severe consequences, ranging from loss of tight junction integrity to basement membrane denudation, depending on the nature and duration of the insult. Recovery involves reversible epithelial phenotype change which allows the epithelium to regenerate and restore its pseudostratified structure. Recolonization of denuded areas requires migration and proliferation of participating cells, which occurs as early as 2 to 6 h after injury (45). Both of these processes require reorganization of the extracellular matrix and, in particular, upregulation of tenascin-C (16, 34).

The synthesis of tenascin-C by epithelial cells in vitro is regulated by growth factors, including epidermal growth factor (EGF) and transforming growth factor  (TGF) (46, 47), and is positively correlated with cellular proliferation, migration, and/or regulation of differentiation (32). The synthesis of tenascin-C by transformed bronchial cells in culture is increased by exposure to the inflammatory cytokines interferon-, tumor necrosis factor , and TGF (48, 49); and, in addition, an integrin receptor for tenascin-C, _v6, is upregulated by EGF and TGF (50). Whether these specific cytokines are involved in regulation of tenascin-C expression as a result of ozone exposure in the primary cultures utilized in this study awaits further examination. Secretion of cytokines by epithelial cells as well as changes in the expression of other surface receptors and matrix components, such as integrins, proteoglycans, and fibronectin, may contribute to a multifaceted alteration in the epithelial cell environment in response to ozone, allowing the cells to change shape and proliferate or detach from the culture.

Tenascin-C has been suggested to be a marker of inflammation (51). It has also been shown to be inhibited by corticosteroids in several different systems (47, 52) and its downregulation has been proposed as an explanation for the inhibition of wound healing by glucocorticoids (53) because tenascin-C can promote fibroblast migration, which is necessary for wound healing. Since ozone has been shown to cause respiratory epithelial changes usually associated with inflammation, including cell detachment and morphologic reorganization, and tenascin-C has been associated with these same processes, ozone may promote apoptosis and epithelial detachment in respiratory tissue by upregulating tenascin-C synthesis. Upregulation of tenascin-C expression in response to ozone may play an early and important role in the disruptive effects of this pollutant.