Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The pulmonary response to inhaled particulates is now being reexamined in the light of recent epidemiologic evidence that links episodes of high levels of airborne small particulates with increases in morbidity and mortality (1- 3); at particular risk are individuals with preexisting lung disease (4). This may indicate that the lung is "primed" by earlier events so that exposure to a normally harmless dose of fine particulate matter is sufficient to trigger events such as cytokine release or cytotoxicity, which may lead to pathologic changes in the lung and other organs. There is now some experimental evidence to support this hypothesis; for example, when rats with pulmonary inflammation or chronic bronchitis were exposed to concentrated ambient air particles, 20 to 40% died, whereas others showed bronchoconstriction (5).

We have previously postulated that particle deposition in the lung at a time of epithelial injury could allow particles or cytokines, normally restricted to the airspaces, to gain access to the lung interstitium and beyond (6). From in vivo and in vitro studies it is also known that there is greater particle uptake across the pulmonary epithelium if it is damaged by cigarette smoke or ozone (7, 8). Using the system of intratracheal instillation, we have shown that silica deposition in the lung when the alveolar epithelium is injured results in greater translocation of particles to the pulmonary interstitium, increased fibrosis, and greater transport of silica to regional lymph nodes (9). This illustrates the importance of epithelial integrity in preventing particle access to tissue and demonstrates a possible mechanism of inducing significant pulmonary effects when a relatively low dose of particles is deposited in injured lungs. This was followed up in a more relevant inhalation study in which rats were coexposed to 0.8 ppm ozone and urban dust for 4 h (10). In this case, the injury induced by ozone was potentiated by the dust exposure, which alone caused no visible or measurable changes in the lung. However, the dust did induce the production of macrophage inflammatory protein-2 (MIP-2) and endothelin-1 by lavaged alveolar macrophages, an effect that was significantly exacerbated by the coexposure to dust plus ozone (11). There was little change in lavaged protein and inflammatory cell numbers in that study, but these values represent an average whole-lung response that might underestimate localized lesions. Ozone is known to induce damage specifically at the alveolar duct regions while sparing the distal lung (10, 12, 13). If ozone-induced injury is potentiated by dust exposure, effects at this anatomic location may be more severe but not adequately measured by whole-lung studies.

In the present study, we wish to confirm that coexposure to ozone and urban dust increases lung injury specifically around the alveolar ducts, and to quantitate injury and inflammation at that site using morphometric methods that allow comparison of different regions in the same lung and among the various exposure groups. By using lung tissue fixed by vascular perfusion, the localized nature of injury and subsequent inflammation at the epithelial surface is not disturbed and can be quantitated in situ in both alveolar and interstitial tissue compartments.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

The animals used were specific pathogen-free male Fischer 344 rats weighing 200 to 250 g. Ozone was produced from pure oxygen in an arc generator, and the concentration was monitored at nose-only inhalation ports (10, 11). The dust used was prepared from baghouse filters collected from the air purification system of the Environmental Health Centre (EHC), Ottawa. The sample EHC-93 was sieved to produce respirable range particles that have been analyzed as detailed elsewhere (10, 14). The dust was dispersed into the airstream and directed into a single flow past the nose-only exposure system. Particle measurements were made as described by Vincent and colleagues (10); the target concentration of EHC-93 was 50 mg/m³ with an actual measured level of 57 mg/m³. Although the individual particle size was < 1 µm diameter, the mass median aerodynamic diameter of the suspension was 4.15 µm with a geometric standard deviation of 1.93. Groups of four rats were exposed for 4 h as follows: (1) to clean air, (2) to ozone at 0.8 ppm, (3) to 50 mg/m³ of EHC-93, and (4) simultaneously to ozone and EHC-93. At the end of the exposure period, rats were returned to cages in clean air for 32 h. Each animal then received tritiated thymidine (³HT) at 1 mCi/kg intraperitoneally and was killed 90 min later by barbiturate overdose.

The chest wall was opened, the lungs were allowed to collapse and then were fixed by vascular perfusion via the pulmonary artery. A solution of 2% glutaraldehyde in cacodylate buffer at pH 7.4 was infused at a constant rate of 2.5 ml/min for 20 min, and the fluid was allowed to run out through a cut in the left atrium. After this time, the lungs were slightly expanded by fluid crossing from vasculature to airways and alveoli. The lungs were tied off, removed, and further fixed by immersion in the glutaraldehyde for 1 h. The lung was then sliced and washed in buffer, and most slices were processed for embedding in glycol methacrylate. The remaining slices were chopped smaller and postfixed in osmic acid before being processed for electron microscopy (15).

From the methacrylate-embedded tissue, sections from at least three blocks were cut at 1.5 µm for histologic examination and quantitation of the inflammatory response. Because cell injury and inflammation occur predominantly at the bronchiolar-alveolar (BR-ALV) duct junction, the numbers of polymorphonuclear leukocytes (PMN) and macrophages (MAC) identified near this point were counted on each section. At high power using a ×100 oil immersion lens, one edge of a counting grid in the microscope eyepiece was lined up at the junction point, and this field plus the next three moving distally were evaluated. Each field measured approximately 75 × 75 µm. The number of PMN and MAC in air spaces and in the lung tissue were recorded separately for each duct area per slide, three slides per rat; if necessary, extra slides were cut so that at least 24 duct areas were counted per animal. In addition, similar counts were made on 12 subpleural fields per animal using the same slides. The PMN were readily identified by their small size and lobed nuclei. Alveolar MAC could be recognized in the alveolar spaces as large irregular cells at the epithelial surface, whereas interstitial MAC were identified in the connective tissue as larger cells with rounded nuclei. Because the vasculature was almost totally free of cells, there was no confusion with monocytes in capillaries, for example. Sections were also cut at 1-µm thickness for autoradiography (15). On three slides per rat, the overall percentage of labeled lung parenchymal cells was determined. Using the foregoing grid-counting system, the percentages of cells labeled in four fields at the BR-ALV duct region and at the lung periphery were also quantitated. Samples of lung tissue, particularly from these central areas, were also sectioned and stained for electron microscopy.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Autoradiography

Counts of ³HT-labeled cells in the total lung parenchyma after inhalation of dust were low and not different from the air-only group (Figure 1). There was a significant increase in labeling in the lung after ozone exposure when 0.8% of all cells were in DNA synthesis. This value further increased to 1.6% of cells when rats were exposed to both ozone and dust (Figure 1). In the latter two groups, labeling was predominately observed near the BR-ALV duct junction where epithelial cells were the almost exclusive cellular sites of label (Figure 2). Cells in four alveolar fields at this specific anatomic location were counted, and it was found that the percentage of labeled cells was considerably higher in the ozone and ozone plus dust groups, as compared with peripheral tissue from the same rats, and from air- or dust-only groups (Figure 1). The highest level, at over 4% of cells labeled, was found in the BR-ALV duct region of the coexposed group. Labeling was virtually confined to epithelial cells; no MAC or PMN were labeled.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Percentages of 3HT-labeled parenchymal cells in different regions of lung in the four exposure groups. Comparing the same zones, *P < 0.05 versus air or dust or distal lung counts; #P < 0.05 versus ozone alone.

View larger version (155K): [in this window] [in a new window]   Figure 2.   Autoradiograph of BR-ALV duct junctional area after coexposure to dust and ozone. Many labeled cells are seen, mainly epithelial cells of the bronchiole (B) or alveoli (arrows). Original magnification ×950.

Morphology

Lungs of rats exposed to air only showed normal morphology with no evidence of cell injury or inflammation and only an occasional red blood cell or PMN where a capillary had not been completely cleared by fixative. Animals exposed to dust only showed no change in overall lung structure or in alveolar cell integrity but did show a few PMN in the lung tissue compartment near alveolar duct regions. Ozone-induced epithelial cell injury occurred at the BR-ALV duct junction where the tissue appeared more cellular, and some PMNs could be identified (Figure 3). Electron microscopy confirmed that there was necrosis of some type 1 alveolar epithelial cells after ozone, particularly at the alveolar duct region as previously reported (15). Cell injury was more severe and readily detected in the ozone plus dust group. At the BR-ALV duct junction, the tissue appeared more cellular by light microscopy, and many PMN were seen in the interstitium, with a few also in the air spaces (Figure 4). By electron microscopy, type 1 cell injury was more extensive at this location, and cell debris with edema fluid was seen in the alveoli (Figures 5 and 6). In tissue adjacent to the necrotic type 1 cells, PMN were present in the interstitium and occasionally a neighboring type 2 cell was seen in mitosis (Figure 6). Similarly, at the terminal end of the bronchioles, evidence of injury and repair was seen with mitotic figures in some epithelial cells and underlying PMN in the connective tissue (Figure 7). In all groups, tissue at the lung periphery did not appear to be affected; there was no evidence of injury and no increase in cellularity at this site when compared with air-only control lungs.

View larger version (134K): [in this window] [in a new window]   Figure 3.   Methacrylate section of lung, 33 h after ozone exposure. Some alveolar edema is seen, the tissue at the duct junctional area is more cellular than normal, and several PMN can be recognized in the connective tissue (arrows). Original magnification ×800.

View larger version (131K): [in this window] [in a new window]   Figure 4.   Methacrylate section of lung 33 h after coexposure to ozone and dust. Some PMN and MAC are seen in air spaces where edema fluid is also present. Many PMN are evident in the connective tissue (arrows). Original magnification ×800.

View larger version (138K): [in this window] [in a new window]   Figure 5.   EM of alveolar wall after coexposure to ozone and dust shows some edema fluid and an alveolar macrophage containing a few dust particles (arrows). Ccapillary; EDedema. Original magnification ×7,500.

View larger version (176K): [in this window] [in a new window]   Figure 6.   EM of alveolar wall at BR-ALV duct region after coexposure to ozone and dust shows type 1 epithelial cell necrosis (arrows) with cell debris in the alveolar space. In the interstitium, PMN (P) are present and a type 2 cell is seen in mitosis (M). Original magnification ×5,500.

View larger version (171K): [in this window] [in a new window]   Figure 7.   EM of an area near the end of a bronchiole (BR) after coexposure to ozone and dust shows an epithelial cell in mitosis (M). In the connective tissue, PMN (arrows) are seen. Original magnification ×4,500.

Quantitation of Inflammation

To quantitate the inflammatory-cell response in air spaces and tissue at the alveolar duct junction, it was not necessary to lavage the lung or to expand the alveoli or airways with fluid but simply to fix tissue by vascular perfusion. Although all lungs were fixed using the same procedure, there was some variation in the degree of alveolar inflation within any particular lung unrelated to the treatment group. For that reason, we counted a minimum of 24 duct regions per animal to give a more valid, though still "semiquantitative," analysis.

The total numbers of PMN in four fields around the BR-ALV duct junction were counted (Figure 8). In the alveolar spaces, very few if any PMN were seen after air alone, and though the numbers were small there was an increase seen after dust or ozone, but particularly after coexposure to ozone plus dust. The changes were more dramatic when cells in the tissue compartment were counted. Even exposure to dust alone significantly increased PMN numbers in lung tissue. Ozone alone also resulted in an influx of cells to the tissue, whereas the largest increase in PMN followed the combined exposure (Figure 8). In each group, no PMN were found in peripheral lung air spaces, whereas only an occasional PMN was seen in the peripheral tissue.

View larger version (19K): [in this window] [in a new window]   Figure 8.   The total number of PMN in four microscopic fields closest to BR-ALV duct junction; cells in air spaces and in the tissue compartment were counted separately in the four exposure groups. Comparing counts in the same zone, *P < 0.05 versus air only; #P < 0.05 versus ozone or dust alone.

Macrophages were also counted and the numbers showed a similar trend, although values were lower (Figure 9). Most MAC were found in the interstitium near the BR-ALV duct junctions after ozone plus dust exposure. Numbers of cells in peripheral regions were low and equal to those of the air-only group. Even if the distinction between air space and tissue compartment is not made, it is clear from the results shown in Figures 8 and 9 that the total number of extravascular PMN and MAC (added values for air spaces and tissue) is highest in the ozone and dust exposure groups. In addition, the total number of each cell type after dust or ozone alone is still significantly greater than that of the air-only group.

View larger version (20K): [in this window] [in a new window]   Figure 9.   The total number of MAC in four microscopic fields at the BR-ALV duct junction. Comparing counts in the same zone, *P < 0.05 versus air; #P < 0.05 versus ozone or dust alone.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The dust EHC-93 was prepared from filtered air in Ottawa, Ontario, and its composition has been analyzed (10, 14). About 20% of the mass of EHC-93 has been estimated to belong to a particulate matter < 2.5 µm in diameter (PM_2.5) fraction as deduced from the elemental analysis and concentrations, the sulphate content, and water solubility (10). The dry dust was dispersed in a venturi to simulate ambient air pollution and analysis of particle size at inhalation ports supported the relevance to ambient air particulates and predictable deposition within the pulmonary region (10). Although the concentrations of ozone and particles in the present study appear high in comparison with ambient air conditions, the relevance to human exposure is better appreciated when doses and deposition efficiencies are scaled and compared within the lungs of rats and humans. As previously discussed (15), the dose of ozone at 0.8 ppm to rats has been estimated as equivalent to about 0.2 ppm in the central acinus, a level of ozone that can be reached in urban air. Similarly, after a detailed analysis, Vincent and colleagues (10) concluded that a particle level of 5 mg/m³ EHC-93 will produce an internal dose that differs by less than one order of magnitude from plausible human exposure. We chose to use the higher level of 50 mg/m³ dust, as it appeared to produce no changes in the lung when delivered alone but potentiated ozone- induced injury to a greater degree than did the 5 mg/m³ dose on coexposure (10, 11). This higher level was then more likely to give measureable effects for the present study while following the same pathway of injury.

Short-term inhalation of EHC-93 particles alone up to a dose level of 50 mg/m³ has been reported as causing no effects on the lung (10, 11). We also have found that a 4-h exposure produced no overt cell necrosis even when tissues are examined by election microscopy, and we found no evidence of DNA synthesis associated with repair. However, when the number of PMN in the periductal regions was counted, there was an increase in these cells, particularly in the lung interstitium, which implies some damage and the generation of a chemotactic factor in response to particle deposition. In another study, when lavaged MAC were collected 1 day after a 4-h inhalation of EHC-93, these cells produced more MIP-2 and endothelin-1 than did MAC after exposure to air (11). This indicates that macrophage-derived or possibly even epithelial cell-derived chemotactic factors are produced in response to the dust and may induce inflammatory cell migration into the tissue. In the earlier study using the same exposure conditions, no change in inflammatory cell numbers was found when these cells were counted in bronchoalveolar lavage fluid (BALF) after dust inhalation (11). This represents a "whole-lung" finding, whereas in the present study in which the lungs were fixed in situ by vascular perfusion, we observed a small increase in PMN and MAC in air spaces near the alveolar duct region. A much greater increase in the number of inflammatory cells in the connective tissue at this anatomic site was demonstrated, suggesting that the dust does induce a localized inflammatory response at the site of deposition.

Exposure to ozone alone at the level used in the present study is known to induce epithelial lesions after a single exposure (10, 15). As we demonstrate here, this was confirmed by increased DNA synthesis in bronchiolar epithelial and type 2 alveolar epithelial cells about 1.5 d after exposure. This response is predominantly found around the junctions of the alveolar duct and illustrates the reparative response after the epithelial cell necrosis produced by ozone exposure. This injury was also accompanied by a highly localized inflammatory response, again with more PMN and MAC being found in the pulmonary interstitium at that anatomic site than in the peripheral air spaces.

Coexposure of rat lung to urban dust and ozone clearly potentiated the lesions at the BR-ALV duct junctions, the specific site of ozone-induced injury. Areas of type 1 alveolar epithelial necrosis were found, and highly localized pockets of alveolar edema were observed. Cell regeneration was also underway as increased DNA synthesis and mitotic figures were seen in bronchiolar and type 2 alveolar epithelial cells. Increased cell injury and repair in the whole lung after such a coexposure has been reported (10), but we demonstrate here that these events are much more pronounced at the duct junctional region where cell labeling is at least three times higher than in total lung counts (Figure 1). In addition, the number of interstitial PMN and MAC are also significantly increased in the coexposure group. In this respect, the assessment of lung inflammation by whole-lung BALF cell counts underestimates the cellular events at the localized center of the acinus and does not measure tissue inflammation. The increased inflammatory response seen in the pulmonary interstitium may have significant consequences. It has been noted that alveolar macrophages in response to dust and ozone coexposure produce much more MIP-2 and endothelin-1 (11), so it is likely that the interstitial MAC population will react similarly. In addition, at sites of epithelial necrosis, such proteins generated in the alveoli can readily cross to the interstitium (16). At this site the increased accumulation of inflammatory cells suggests that the release of PMN products such as free radicals (17) and/or the activation and release of macrophage-derived cytokines (18, 19) could contribute to any subsequent pathologic changes in the pulmonary interstitium or even more distal sites.

The results demonstrate a cytotoxic interaction on coexposure to ozone and airborne particles. Although the mechanism for this interaction is not known, it is possible that ozone alone, although inducing necrosis of some epithelial cells, may also cause a nonlethal injury to other cells at the same anatomic location. Further or concomitant exposure of these injured cells to the otherwise minor effects of EHC dust may cause the injury to progress to cell death. Because particles preferentially deposit in the BR-ALV duct region (20), the increased cell necrosis predominates in the same general anatomic location as the ozone-induced injury, and the distal lung is not affected. As a consequence, an influx of inflammatory cells is also concentrated at the centriacinar regions. It may then follow that any lung with preexisting epithelial injury at the time of particle inhalation is more susceptible to further damage. There is some evidence for this from experimental studies in which particle deposition during a phase of epithelial damage induced by virus or chemicals results in greater injury, more particle penetration of the tissue, and increased pathologic changes (6, 9, 21). In one study, exposure of rats with bronchitis to ambient air particles resulted in increased mortality (5). The present study demonstrates how coexposure to ozone and to urban dust at reasonable dose levels can result in increased cell injury in the lung and can induce a localized inflammatory effect that may not be appreciated when the lung response is studied by BAL or by whole-tissue analysis.