Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Dendritic cells are highly efficient antigen-presenting cells found in lymphoid organs and a variety of other tissues, including the respiratory tract (1, 2). After capturing antigen in peripheral locations, these cells migrate via lymphatics to the T-cell-dependent areas of regional lymph nodes, where they encounter naive lymphocytes and initiate a primary immune response (3). Mature dendritic cells constitutively express high levels of major histocompatibility complex class II (Ia) molecules on their surface and are thought to be especially important in the activation of CD4⁺ helper T cells (4, 5).

Immunocytochemical studies of humans and rodents have revealed an abundance of Ia-immunoreactive (Ia⁺) dendritic cells in the mucosa of the conducting airways (6- 8). These cells have long, branching processes that extend in the plane of the epithelial basement membrane, forming a network optimally situated to sample inhaled antigen. Because of their repertoire of costimulatory molecules needed to activate naive T cells, airway dendritic cells are considered the most potent antigen-presenting cells in the respiratory tract (4, 5).

Recent evidence indicates that airway dendritic cells constitute a highly reactive population, whose numbers and Ia content change rapidly under experimental conditions. For example, dendritic cells become more numerous in response to inhaled antigen (9, 10) and rapidly decrease in number after treatment with glucocorticoids (11, 12). Nelson and coworkers (12) reported that the number of Ia⁺ dendritic cells in the rat tracheal mucosa decreases by nearly 75% within 24 h of systemic treatment with dexamethasone. However, the mechanism of this decrease is uncertain. One potential mechanism is apoptosis, a form of cell death characterized by distinct biochemical and morphologic changes (13). Glucocorticoids have been shown to induce apoptosis in murine thymocytes in vivo (14, 15) and may trigger similar changes in airway dendritic cells. Alternatively, glucocorticoids may alter the rate at which dendritic cells enter or leave the airway mucosa. One way of distinguishing between apoptosis and changes in cell influx or efflux is to determine whether the dendritic cells become fragmented (suggestive of apoptosis) or become rounder (suggestive of cell migration). Another approach is to use the terminal deoxynucleotidyl transferase (TdT)- mediated dUTP-biotin nick end-labeling (TUNEL) method to detect evidence of apoptosis (16).

In the present study, we examined formalin-fixed tracheal whole mounts, which provided an unprecedented way to visualize the three-dimensional morphology, arrangement, and number of dendritic cells in the airway mucosa. We had three objectives: (1) to define the changes in dendritic cell morphology induced by glucocorticoid treatment, with a focus on evidence of apoptosis versus cell migration; (2) to correlate such changes with the time-dependent reduction in the number of dendritic cells following treatment; and (3) to determine the contribution of apoptosis to the steroid-induced reduction in dendritic cell number.

Rats were given injections of the potent glucocorticoid dexamethasone, and tracheal dendritic cells were examined 4 to 72 h later. The Ia⁺ dendritic cells were revealed by immunoperoxidase staining using the OX-6 (anti-Ia) monoclonal antibody (17). Evidence of dendritic cell apoptosis was sought in histologic sections of trachea stained by the TUNEL method. Some of our findings have been reported in abstract form (18).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

Male, pathogen-free Wistar rats were purchased from Charles River Breeding Laboratories (Hollister, CA) and housed under barrier conditions in autoclaved microisolator cages. At the time of the experiments, the rats were approximately 12 wk old and had a mean body weight of 250 g. All of the experimental procedures used in this study were approved by the Committee on Animal Research of the University of California, San Francisco.

Dexamethasone Treatment

For the whole mounts, a total of 16 rats were treated with dexamethasone (10 mg/kg given intraperitoneally once daily; Sigma Chemical Company, St. Louis, MO). This dose has been shown to reduce profoundly the number of dendritic cells in the rat tracheal mucosa (11). At intervals of 4, 8, 24, and 72 h after treatment (four rats per time point), the rats were anesthetized with sodium pentobarbital (50 mg/ kg, intraperitoneally; Abbott Laboratories, North Chicago, IL) and perfused via the thoracic aorta with 1% paraformaldehyde in 0.01 M phosphate-buffered 0.9% NaCl (PBS; pH 7.4) for 2 min at a pressure of 120 mm Hg (19). An additional six untreated rats (controls) were studied along with the others. The tracheas were incised along the ventral midline and immersed in 4% paraformaldehyde overnight at 4°C. For the histological sections, a total of eight rats was treated with dexamethasone (10 mg/kg, intraperitoneally) and perfused with fixative at 4 h and 8 h after the treatment (four rats per time point). An additional four rats received no treatment (controls). The tracheas were immersed intact in 4% paraformaldehyde overnight at 4°C.

Staining of Dendritic Cells in Whole Mounts

Airway dendritic cells were stained using the OX-6 mouse monoclonal antibody (PharMingen, San Diego, CA), which is directed against the Ia surface determinant (17). After fixation, the tracheas were washed for 6 h in PBS containing 0.3% Triton X-100 (PBS/Triton; Sigma), pinned flat onto small slabs of Sylgard rubber (Dow Corning Corporation, Midland, MI), and then processed for immunocytochemistry using a modification of the indirect immunoperoxidase techniques described by Holt and colleagues (20) and Baluk and coworkers (21). All steps were carried out at room temperature and all solutions contained 0.01% thimerosal (Sigma) to inhibit bacterial growth. The tracheas were incubated for 2 h in PBS/Triton containing 5% normal goat serum (NGS; Jackson ImmunoResearch Laboratories, West Grove, PA) and then for 36 h in PBS/ Triton containing 1% NGS and the OX-6 antibody diluted 1:1,000. After a 6-h wash in PBS/Triton, the tracheas were incubated for 24 h in PBS/Triton containing peroxidase-conjugated goat antimouse immunoglobulin G (IgG) (Jackson) diluted 1:200, washed again, and reacted with hydrogen peroxide and 3,3'-diaminobenzidine (DAB) to reveal the peroxidase label. Finally, the stained tracheas were dehydrated in ethanol, cleared in toluene, and mounted mucosal surface up in Permount (Fisher Scientific Company, Fair Lawn, NJ) (19). The resulting whole mounts were examined with a Zeiss Axiophot microscope (Carl Zeiss, Thornwood, NY) equipped with differential interference contrast optics or with an Edge R400 microscope (Edge Scientific Instrument Corporation, Los Angeles, CA), which uses oblique specimen illumination to produce three-dimensional images.

Staining of Dendritic Cells in Histologic Sections

To avoid the harsh effects of paraffin processing on OX-6 antigenic sites, we embedded the tracheas in polyethylene glycol 900 (PEG; Aldrich Chemical Company, St. Louis, MO), a water-soluble wax with a melting point of 37°C (22). After fixation, the tracheas were washed for 2 h in PBS and placed in a 1:1 solution of PEG and PBS for 1 h at room temperature. The tracheas were then passed through three changes of molten 100% PEG for 1 h each at 37°C and allowed to cool in fresh PEG to form solid tissue blocks. Transverse histologic sections (10 µm) were cut in a cryostat at 10°C, floated on a water bath to dissolve the PEG, and collected on Superfrost Plus-coated slides (VWR Scientific Products, West Chester, PA). All subsequent steps were carried out at room temperature. After air-drying overnight, the sections were incubated for 20 min in 0.01 M Tris-HCl (pH 8.0) containing 20 µg/ml proteinase K (Sigma), rinsed in 0.02 M Tris-buffered 0.9% NaCl (TBS; pH 7.6), incubated for 5 min in methanol containing 3% hydrogen peroxide, and rinsed again in TBS. The sections were incubated for 30 min in TBS containing 5% NGS and then for 1 h in TBS containing 1% NGS and the OX-6 antibody diluted 1:1,000. Following a TBS rinse, the sections were incubated for a second time in TBS containing 5% NGS, then incubated for 1 h in TBS containing peroxidase-conjugated goat antimouse IgG diluted 1:200, rinsed in TBS, and reacted with hydrogen peroxide and DAB. Finally, the sections were counterstained with 0.3% methyl green, dehydrated, cleared, and mounted in Permount.

Staining of Apoptotic Bodies in Histologic Sections

To stain the nucleosomal DNA fragments in apoptotic bodies, we used a modification of the TUNEL method described by Gavrieli and colleagues (16). All reagents were supplied by the TdT-FragEL DNA Fragmentation Detection Kit (Oncogene Research Products, Cambridge, MA) and applied according to the kit protocol. In brief, the PEG-embedded tracheas were cut to obtain paired adjacent sectionsone section of each pair was processed for OX-6 staining as described above, the other section was processed as a free-floating, 30-µm section for TUNEL staining.¹ After dissolving the PEG with a TBS rinse, the free-floating sections were incubated for 20 min in Tris-HCl containing 20 µg/ml proteinase K and for 5 min in methanol containing 3% hydrogen peroxide. Subsequently, the sections were incubated for 1.5 h at 37°C in a mixture of TdT enzyme and biotinylated nucleotides to label the exposed 3'-OH ends of DNA fragments generated during apoptosis, then the sections were incubated for 30 min at room temperature in a solution of streptavidin-peroxidase conjugate and reacted with hydrogen peroxide and DAB. Negative controls were obtained by omitting the TdT enzyme. The sections were transferred to Superfrost Plus-coated slides, air-dried for 2 h, and counterstained with methyl green before mounting in Permount.

Quantification of Dendritic Cells and Apoptotic Bodies

Dendritic cells were quantified in the tracheal whole mounts by counting all Ia⁺ cells having one or more cytoplasmic processes in 12 consecutive regions along the right side of each trachea, at sites located over cartilage rings 3-14. Each region measured 0.06 mm². The number of Ia⁺ dendritic cells was expressed per square millimeter of mucosa. Apoptotic bodies were quantified in the histologic sections by counting all TUNEL⁺ bodies in three widely separated regions around the mucosal perimeter of each section, at sites located near the middle and ends of the cartilage ring. Each region measured 0.014 mm². A total of 10 sections per trachea was examined. The number of TUNEL⁺ bodies was expressed per square millimeter of mucosa.

Statistical Analysis

Values are expressed as means ± SEM. Differences between group means were assessed by one-way analysis of variance (ANOVA) and Scheffé's test. Differences were considered significant when P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Morphology of Dendritic Cells in Whole Mounts

Normal tracheal mucosa. The Ia⁺ dendritic cells were more abundant in the mucosa between the cartilage rings than in the mucosa overlying the rings, which was thinner and provided the best vantage point for examining the detailed morphology of individual cells (Figure 1A). By focusing through successive depths of the whole mounts, we were able to observe the full extent of the dendritic cell network in three dimensions. We observed Ia⁺ cells scattered throughout the mucosa, but most cells were located in a thin layer just beneath the epithelium. Although the Ia⁺ cells formed a tightly spaced array, actual contact between dendritic cells was infrequently seen. In addition to these mucosal dendritic cells, numerous Ia⁺ cells were observed alongside blood vessels in the underlying adventitia (not shown).

View larger version (161K): [in this window] [in a new window]   Figure 1.   Micrographs showing Ia+ dendritic cells in tracheal whole mounts from normal, pathogen-free rats. Immunoperoxidase staining with the OX-6 (anti-Ia) antibody. Cells were photographed using differential interference contrast optics. Whole mounts viewed from the luminal surface. (A) Ia+ cells in the mucosa overlying a cartilage ring. Each cell has three to five branched cytoplasmic processes. (B) Ia+ cell with a bilobed nucleus (arrow) and processes that surround basal cells (asterisk) of the tracheal epithelium. (C) Ia+ cell with an immunoreactive granule (arrow) in one of its processes. Scale bar = 20 µm.

Dexamethasone-treated tracheal mucosa. The Ia⁺ dendritic cells displayed several conspicuous morphologic abnormalities after dexamethasone treatment. At 4 h and 8 h after treatment, the Ia immunoreactivity of many dendritic cells was internalized into cytoplasmic vacuoles (Figure 2B). In addition, small Ia⁺ bodies, which appeared to be fragments of dendritic cells, were seen scattered in the interstitium (Figure 2B).

View larger version (128K): [in this window] [in a new window]   Figure 2.   (A-D) Micrographs showing Ia+ dendritic cells in tracheal whole mounts from rats that received no treatment (A) or were treated with dexamethasone (10 mg/kg/d, intraperitoneally; B-D). (A) No treatment. Ia immunoreactivity is abundant on the surface of most dendritic cells. The dendritic cell processes are broad, distinct, and highly branched. (B) 4 h after dexamethasone. Ia immunoreactivity is visible in the cytoplasm of many dendritic cells. Some cells appear to be fragmenting. (C) 24 h after dexamethasone. The dendritic cell processes contain prominent granules, some of which appear isolated from identifiable cells (arrows). The dendritic cell body has a swollen appearance. (D) 72 h after dexamethasone. The dendritic cell processes are narrower and the surface Ia immunoreactivity is reduced in comparison with the untreated controls (A). (E-H) Micrographs showing Ia+ dendritic cells (E, F) and TUNEL+ bodies (G, H) in histologic sections of trachea from rats that received no treatment (E, G) or were treated with dexamethasone (F, H). (E) No treatment. Dendritic cells are abundant in the lamina propria. Some dendritic cells are visible in the epithelium (arrows). (F  ) 4 h after dexamethasone. Fewer dendritic cells are visible in the lamina propria. (G) No treatment. Only one TUNEL+ body (arrow) is visible in this region of the lamina propria. (H) 4 h after dexamethasone. Numerous TUNEL+ bodies are scattered in the lamina propria. Scale bar = 20 µm.

Detection of Apoptosis in Histologic Sections

The Ia⁺ dendritic cells were numerous in the lamina propria of the normal tracheal mucosa, but TUNEL⁺ bodies were rare (Figures 2E and 2G). By contrast, at 4 h and 8 h after dexamethasone treatment, the Ia⁺ cells were noticeably less abundant, but many more TUNEL⁺ bodies were seen scattered in the lamina propria (Figures 2F and 2H). Many of the TUNEL⁺ bodies closely resembled the small Ia⁺ bodies seen in the whole mounts after dexamethasone treatment (Figures 2B and 2H).

Dendritic Cell Number

In the untreated rats, the number of Ia⁺ dendritic cells in the tracheal mucosa overlying the cartilage rings was 1,405 ± 140 cells/mm² mucosa (n = 6; Figure 3). The corresponding number of Ia⁺ cells in the dexamethasone-treated rats was reduced by 24% at 4 h, 56% at 8 h, 68% at 24 h, and 77% at 72 h (Figure 3). The reductions after 8 h were statistically significant (P < 0.001, one-way ANOVA).

View larger version (9K): [in this window] [in a new window]   Figure 3.   Graph showing the number of Ia+ dendritic cells in the rat tracheal mucosa at different times after the onset of dexamethasone treatment (10 mg/kg/d, intraperitoneally). Values are means ± SEM; n = four to six tracheas per group. *Significantly different from the untreated controls (P < 0.001, one-way ANOVA).

Apoptotic Body Number

In the untreated rats, the number of TUNEL⁺ bodies in the tracheal mucosa overlying the cartilage rings was 174 ± 47 bodies/mm² mucosa (n = 4; Figure 4). The corresponding number of TUNEL⁺ bodies in the dexamethasone-treated rats was increased 12.1-fold at 4 h and 5.4-fold at 8 h (Figure 4). The increase at 4 h was statistically significant (P < 0.002, one-way ANOVA).

View larger version (10K): [in this window] [in a new window]   Figure 4.   Graph showing the number of TUNEL+ bodies in the rat tracheal mucosa at different times after the onset of dexamethasone treatment (10 mg/kg, intraperitoneally). Values are means ± SEM; n = four tracheas per group. *Significantly different from the untreated controls (P < 0.002, one-way ANOVA).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results of this study show that glucocorticoid treatment causes conspicuous morphologic changes in dendritic cells of the rat tracheal mucosa, and that these changes coincide with a reduction in the number of dendritic cells and an increase in the number of TUNEL⁺ (apoptotic) bodies. Examination of tracheal whole mounts from untreated rats revealed a dense network of Ia⁺ dendritic cells with highly branched cytoplasmic processes. After dexamethasone treatment, four distinct morphologic abnormalities were observed in these cells: (1) appearance of large Ia⁺ granules in cytoplasmic processes, (2) narrowing of cytoplasmic processes, (3) loss of Ia immunoreactivity from the cell surface, and (4) fragmentation of cells into small Ia⁺ bodies, which was evident 4 h after treatment and coincided with the appearance of TUNEL⁺ bodies indicative of apoptosis. The number of dendritic cells decreased 56% in 8 h and 77% in 72 h.

Our use of tracheal whole mounts enabled us to view the intricate morphology of Ia⁺ dendritic cells in greater detail than has been attainable using conventional sectioning methods. The intact nature of the whole mounts made it possible to see features not previously described, such as the intimate association of dendritic cells and epithelial basal cells. This association is reminiscent of the interaction between Langerhans cells and keratinocytes of the skin (7). Another feature not readily appreciated in histologic sections is the multilobed nucleus, which was obvious in many of the larger dendritic cells and resembled the highly indented nucleus seen in dendritic cells isolated from the peripheral lymph (23). Particularly intriguing are the large Ia⁺ granules that we observed in some of the dendritic cell processes. Although the function of these granules is unknown, the fact that they became more numerous after dexamethasone treatment suggests their involvement in the response to steroids. Holt and Thomas (24) have presented evidence that steroids inhibit the uptake and/or processing of exogenous antigen by airway dendritic cells. Perhaps the Ia⁺ granules reflect functional abnormalities in the endosomal compartments of these cells.

The number of Ia⁺ dendritic cells we counted in the tracheal mucosa of untreated rats (1,405 ± 140 cells/mm² mucosa) is approximately twice as large as that reported by Schon-Hegrad and coworkers (600-800 cells/mm² epithelium) (8). However, these investigators examined histologic sections that were cut through the epithelium parallel to the basement membrane, which excluded the dendritic cells located deeper in the lamina propria. By contrast, our count reflects the total number of Ia⁺ cells throughout the entire mucosa overlying the cartilage rings.²

The rapid decline we observed in the number of Ia⁺ dendritic cells after dexamethasone treatment is consistent with the findings of others (11, 12). By measuring the rate of Ia⁺ cell decline after X-ray or high-dose dexamethasone, Holt and coworkers (11) estimated that the half-life of dendritic cells in the rat trachea is about 2 d. We used a similar approach, but included Ia⁺ cell counts taken at earlier time points after the onset of treatment. We found that the number of Ia⁺ cells decreased rapidly during the first 8 h after dexamethasone treatment (56% reduction) and decreased more slowly thereafter. Based on these data, our estimate for the half-life of dendritic cells is approximately 8 h.

According to Nelson and coworkers (12), the steroid-induced decline in dendritic cells probably reflects the normal emigration of the cells to regional lymph nodes without a corresponding influx of precursor cells to offset the loss. These investigators concluded that steroids act by inhibiting the recruitment of precursor cells from the circulation, perhaps by altering the binding of incoming cells to the vascular endothelium. If this explanation is correct, the rapid loss of Ia⁺ cells observed after dexamethasone treatment indicates that dendritic cells in the rat trachea have a very fast turnover under conditions of steady-state antigenic exposure. The net flux of dendritic cells through the tracheal mucosa, as reflected by the turnover, is thought to depend on the prevailing antigenic load and the need for continuous immune surveillance (11). Indeed, inhalation of bacterial antigen can substantially increase the turnover of dendritic cells in the rat trachea (9, 10). However, factors other than emigration of dendritic cells could account for some of the reduction in Ia⁺ cells seen after dexamethasone treatment.

One factor that appears to contribute to this rapid loss of cells is apoptosis. Apoptotic cell death is morphologically characterized by the fragmentation of dying cells into small apoptotic bodies, which are then rapidly removed by phagocytes (13). Endonuclease cleavage of the genomic DNA into nucleosome-length fragments usually accompanies this process and is considered the biochemical hallmark of apoptosis (25). Because the small Ia⁺ bodies observed in the whole mounts at 4 h and 8 h after dexamethasone treatment were suggestive of apoptosis, we used the TUNEL staining method to detect nucleosomal DNA fragments in transverse histologic sections of the tracheal mucosa (16). Examination of paired adjacent sectionsone stained with OX-6, the other with TUNELrevealed a close correspondence between the locations of Ia⁺ cells and TUNEL⁺ bodies, indicating that dendritic cells are the likely source of these apoptotic nuclear fragments. Moreover, the number of TUNEL⁺ bodies in the histologic sections was greatest 4 h after dexamethasone treatment, which was when Ia⁺ cell fragmentation in the whole mounts was most pronounced. These findings lead us to believe that many tracheal dendritic cells undergo apoptosis in response to steroids, although the attendant cellular fragmentation makes it difficult to determine the exact proportion of dendritic cells so affected.

Steroid-induced apoptosis is a well-documented phenomenon in lymphocytes (26), and Kaempgen and coworkers (27) reported that mouse splenic dendritic cells undergo apoptosis when incubated with dexamethasone in vitro. The mechanism by which steroids induce apoptosis is uncertain but may be related to the suppression of critical cytokines required for cell survival. For example, Moser and coworkers (28) showed that the addition of granulocyte-macrophage colony stimulating factor prevented the steroid-induced loss of mouse splenic dendritic cells in culture. The locus of steroid action in this regard may be the transcription factor known as nuclear factor-kappa B, inactivation of which has been linked to the induction of apoptosis in certain cell types (29). A similar mechanism might explain the reduced numbers of dendritic cells seen in the bronchial mucosa of atopic asthmatic patients after prolonged treatment with inhaled glucocorticoids (30).

Another factor that could contribute to the reduction in Ia⁺ cells is the progressive loss of Ia immunoreactivity from the cell surface. The amount of Ia antigen may have been reduced sufficiently to render some dendritic cells undetectable by OX-6 immunoperoxidase staining. Reduced staining intensity might also explain the apparent narrowing of dendritic cell processes. Similar findings were reported by Holt and coworkers (11), who showed that the relative intensity of OX-6 immunoperoxidase staining in the rat trachea decreased by an average of 55% during the first 2 d after dexamethasone treatment. These investigators further showed that the intensity of Ia staining decreased faster than the number of dendritic cells. Likewise, we observed internalization of OX-6 staining only 4 h after dexamethasone treatmentbefore there was a significant reduction in dendritic cell number. This suggests that downregulation of Ia surface antigen is a very early event in the response to steroids and may be a requisite step for subsequent changes in the dendritic cell population.

We conclude that dexamethasone treatment reduces the number of Ia⁺ dendritic cells in the rat tracheal mucosa through multiple mechanisms. The rapid decline in Ia⁺ cells that occurs during the first 8 h probably results largely from the death of dendritic cells by apoptosis. The slower decline in Ia⁺ cells that occurs 8 to 72 h after treatment may reflect the emigration of dendritic cells to the regional lymph nodes and the progressive loss of surface Ia immunoreactivity.