Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Defining the mechanisms by which reactive oxidant gases such as ozone (O₃) damage the respiratory system is hindered by a number of factors, including a lack of homogeneity in the formation of cytotoxic products (1, 2) and the pattern of acute cellular injury (3). The latter is highly focal, site-selective, and variable by species and duration of exposure (3). Because dosimetry models indicate substantial variation in the rates of O₃ uptake relative to longitudinal position within the airway tree (15), this variability in the cellular response could be attributable to variations in absorbed dose and the subsequent formation of bioactive reaction products. Because O₃ likely does not diffuse through the lung surface lining layer (SLL) without reacting, the formation of SLL-derived products is presumably a critical component in exposure-induced toxicity (1, 2). The majority of the studies that have defined the pattern of epithelial injury and changes in surface chemistry associated with short-term exposure have relied on assessment of total protein, lactate dehydrogenase (LDH), and recovered cells in bronchoalveolar lavage (BAL) fluid (BALF) (18, 19), or on histopathologic approaches combined with specific sampling methodologies to define site specificity within the airway tree (20). Reliance on information from BAL is limited by the inability to discriminate pathophysiologic foci and the requirement of relatively pronounced epithelial damage before changes can be detected. Two of the major drawbacks of histopathologic approaches are that they are extremely time-consuming and they restrict the perspective that could be obtained if the entire target surface of the tracheobronchial airways were evaluated. Even when combined with sampling by airway dissection, these methods are two-dimensional and therefore rely on very small subsamples of specific regions. These methods also limit characterization of the injury pattern at sites where airflow may be modified, such as at branch points, where there are marked changes in airway diameter and flow pattern (from laminar to nonlaminar), and where there is a decrease in bulk flow at the terminal bronchiole/alveolar duct junctional region. Local dose to the epithelium might be expected to vary at these anatomic sites.

Complete appreciation of the extent and pattern of epithelial injury using a conventional two-dimensional histopathologic approach requires computer-aided three-dimensional reconstruction strategies associated with serial sections of large blocks of defined airway specimens. This approach has been used to define the extent of local injury in one of these regions, the centriacinar region, but has been inadequate for providing a three-dimensional perspective of the distribution of injured epithelium even of that one region (10, 21). The current study was designed to develop and validate a method of mapping acute cell injury and to compare its sensitivity for detecting injury with measurement of total protein and LDH in BALF.

The method is based on the use of specific nuclear DNA-binding fluorochromes to label the nuclei of cells with compromised plasma membrane integrity combined with airway microdissection to allow comparison of: (1) different airway levels of the same pathway, (2) both branches at a bifurcation, (3) bifurcation points and the airway walls between them, (4) proximal bifurcation points with distal bifurcation points, and (5) the terminal bronchiole/alveolar duct junction with more proximal airways. The goal was a rapid, relatively reliable assay that could provide quantitative data on cell injury in a reproducible fashion. To control ventilatory parameters believed to affect regional gas distribution, we used an isolated perfused lung preparation in these initial studies. Our study shows that the method can be used to quantify the distribution of acute cell injury; that O₃ exposure produces a heterogeneous pattern of injury throughout the conducting airways, with the greatest injury occurring in small side branches from the main airway path just downstream of a bifurcation; and that the method displays far greater sensitivity for detecting cell injury than does the measurement of total protein or LDH activity in BALF.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Exposure Conditions

For optimal control of the inspired dose rate, and to facilitate direct quantitative comparisons based solely on inspired O₃ concentrations, we used isolated rat lung preparations as our exposure model. Adult male Sprague- Dawley rats (body weight: 390 to 573 g) were deeply anesthetized with pentobarbital sodium (65 mg/kg) and the lungs surgically resected as described previously (24, 25). Briefly, the trachea was cannulated and animals were placed on positive pressure-support ventilation while the chest cavity was opened and the pulmonary artery cannulated. The pulmonary vascular bed was flushed free of erythrocytes and the lungs were placed in an artificial thorax equipped to provide subatmospheric ventilation. The lungs were perfused with recirculated Krebs bicarbonate buffer containing 4.5% albumin and 6 mM glucose (pH 7.4; 37°C). O₃ was generated from 100% O₂ via a silent arc electrode and bled into a biased flow of 95% air/5% CO₂ that flowed past the tracheal cannula in excess of peak inspiratory flow to ensure a constant inhalation concentration. O₃ concentrations were monitored using either a Thermo-Environmental Model 49 UV Photometric Analyzer (Thermo-Environmental, Franklin, MA) or a Dasibi 1003 AH (Dasibi Environmental Corp., Glendale, CA). Isolated perfused lungs were exposed to O₃ under standard conditions (24). The lungs were stabilized with filtered air (FA), then exposed for 20 to 90 min to one of the following target concentrations: FA (0.0 ppm) or 0.25, 0.5, or 1.0 ppm O₃. Mean ventilation parameters were: tidal volume = 2.4 ml, and ventilation frequency = 40 breaths/min.

In separate studies, the effects of O₃ exposure on LDH release and on LDH catalytic activity were examined. Isolated rat lungs were exposed to 0.25 to 2.0 ppm O₃ for 20 to 60 min, followed immediately by BAL to assess LDH activity and protein content (see BIOCHEMICAL ANALYSIS OF BALF, below). Additionally, cell-free BALF was obtained from naive lungs and 300 U/ml LDH (EC 1.1.1.27; from rabbit muscles; Sigma Chemical Co., St. Louis, MO) added. The mixture was exposed in vitro under well-mixed steady-state conditions (1) to 0.32 ppm O₃ in air for 20 min. At selected times, samples were withdrawn and the LDH activity was evaluated.

Cell Permeability

After exposure, ethidium homodimer-1 (EthD-1) (12 M; Molecular Probes, Eugene, OR) in Waymouth's medium was infused by tracheal cannula for 15 min at 37°C. The EthD-1 was then removed by a single gentle lavage with the same medium, and the lungs were fixed via infusion through the tracheal cannula with glutaraldehyde/paraformaldehyde (1.1%/0.9%) in cacodylate buffer (pH 7.4, 330 mOsm) at 30 cm fluid pressure. After 10 min of fixation, the trachea was ligated and the lungs were removed and placed in the same fixative. The analysis of EthD-1-permeable cells was conducted on the right middle lobe. The lobe was separated into two pieces: the proximal two-thirds (Figure 1, upper left) and the distal one-third (Figure 1, lower left). Each piece was glued to a coverslip by its costal surface using Nexaband S/C veterinary adhesive (Veterinary Products Laboratories, Phoenix, AZ). Beginning at the lobar bronchus (Site A in Figure 1, upper left), the axial pathway in the proximal piece was exposed by microdissection and its side branches were opened as previously described (13, 20). The analysis was performed on the costal half of these airways. The axial pathway at the second intrapulmonary branch point (Site B in Figure 1, upper left) and its accompanying side branch (Site C in Figure 1, upper left) (airway generations #5-7) were compared with a more distal part of the axial path (Site D in Figure 1, upper left and right) and the associated side branch (Site E in Figure 1, upper left and right) (airway generations #10-12). In the distal portion of the lung (Figure 1, lower left and right), the terminal bronchiole in the axial pathway and those of the associated side branches were visually compared with the airway generations immediately proximal to them. The microdissected lungs were imaged using laser scanning confocal microscopy (Bio-Rad MRC-600 mounted on an Olympus BH-2) with water-immersion objectives. The nuclei of cells impermeant to EthD-1 were identified by labeling all nuclei in microdissected fixed lungs with YOPRO-1 (2-4 M; Molecular Probes). YOPRO-1 labeled all nuclei within the airways as well as the cytoplasm but at a much lower level of intensity. EthD-1 (EX 518; EM 605 nm)-positive cells were imaged using the GHS channel, and YOPRO-1 (EX 491; EM 509 nm)-positive cells were imaged using the BHS channel. Position within the dissected airway tree was identified with a ×4 objective, and the confocal Z-series images for a specific epithelial site were collected using a ×20 objective. Images within a Z-series were 10 µm apart and had a depth of focus of 10 µm.

View larger version (101K): [in this window] [in a new window]   Figure 1.   Comparison of airway microdissection in right middle lobe of rat lung with projected series from laser scanning confocal microscope. In the axial pathway of the proximal two-thirds of the lobe (upper left), two sites in the axial path (Sites B and D) were quantified. Both of these axial path sites (B and D) were compared with epithelial cells in the adjacent minor daughter path (Site C and E, respectively) (upper right). In the distal one-third of the lobe (lower left), the most distal three to five generations of bronchioles and terminal bronchoalveolar duct junctions (the area marked in a rectangle in lower left), were evaluated by confocal microscopy (lower right). The two images on the right are higher magnification of the boxes outlined in black in the left images. The inset diagrams in the left images indicate the portion of the lobe shown in that image.

Morphometric Assessment

The number of cells with EthD-1-positive nuclei per unit volume (mm³) of epithelium (nucleus and cytoplasm of YOPRO-1-positive cells) was estimated using an optical "disector" (26, 27). Each Z-series was separated into individual files in two folders (directories) of EthD-1- and YOPRO-1-positive cells using NIH Image.* EthD-1-positive nuclei were imported into Stereology Toolbox. Using an unbiased optical "disector," EthD-1-positive nuclei were counted in serial sections excluding those that intersected the top section and two sides of the counting frame. For a small number of EthD-1-positive nuclei, signal from one nucleus bordered that of a neighboring nucleus. When this occurred, the two could be discerned as double by shape and size and were counted as two. YOPRO-1-positive cells were imported into the Stereology Toolbox software and the total volume of cells was estimated using point counts on individual sections and an associated volume with each point. Stacks of serial sections (spaced 10 µm apart and with a depth of focus of 10 µm) varied from four to 40 images depending on the orientation of the airway surface in the dissected whole mounts. Sections were separated from the image stack and every other section was used for counting.

Expressed in a formula, the number of EthD-1-positive nuclei (N_ETH) per volume of YOPRO-1-positive cells (V) is:

where P_EA is the point hits on epithelial cells that have YOPRO-1-positive nuclei totaled from all sections in the stack, and V/P is the volume per point (2.738 × 10⁷ mm³) in the epithelium per Z- series section thickness.

Histopathology

To assess the nature of the cells labeled with EthD-1, areas of proximal airways identified as Sites B and C in Figure 1 were evaluated by light microscopy. Those regions were removed from the whole mounts of lungs exposed to 1.0 ppm O₃ by slicing perpendicular to the long axis of the axial airway after laser scanning was completed. The slices were dehydrated in ethanol and embedded in glycolmethacrylate (Immuno-bed; Polysciences, Warrington, PA). The blocks were sectioned at 1 µm on a Sorvall JB-4 microtome. After unstained sections were evaluated by confocal microscopy, the same sections were stained with Toluidine blue and evaluated by transmission light microscopy using an Olympus Provis microscope.

Biochemical Analysis of BALF

For LDH activity and total protein analysis, lungs were lavaged immediately after exposure through the tracheal cannula using 8 ml of warmed (37°C) phosphate-buffered saline (pH 7.0, 310 mOsmol) that was gently instilled and withdrawn three times to yield BALF. Cells were removed by centrifugation (4,000 × g for 10 min, 4°C) and flash-frozen for later analysis. LDH was analyzed by following the oxidation of reduced nicotinamide adenine dinucleotide at optical density₃₄₀ (28) and protein by the Lowry Method (29).

Statistics

The differences between O₃ concentrations for each site were compared, after log transformation of the data, by repeated measures analysis of variance (SAS; SAS Institute Inc., Cary, NC) with one grouping factor (O₃ concentration) and one within factor (airway site). The post hoc analysis was a polynomial decomposition comparing each site with the terminal bronchioles and comparing all adjacent sites. Statistical significance was P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

LDH and Protein in BALF

As we have previously observed (24, 25), the O₃ exposure conditions employed in this study did not produce overt pulmonary damage as assessed by lung weight or the appearance of pulmonary edema. Figure 2 shows the correlations between O₃ exposure and the appearance of LDH activity and total protein in BALF. Neither LDH nor protein levels differed from FA controls until exposure concentrations exceeded 1 ppm O₃. To determine whether LDH is inactivated during exposure, LDH-supplemented BALF was exposed in vitro and resulted in notable losses of LDH activity (Figure 3). After 20-min exposure to 0.3 ppm O₃, we observed at least a 75% decline in the measurable activity.

View larger version (28K): [in this window] [in a new window]   Figure 2.   Isolated perfused rat lungs were exposed to O3 under steady-state conditions. Exposure conditions encompassed combinations of inspired concentrations ([O3] inspired = 0 to 2.0 ppm) and exposure times (20 to 60 min). Tidal volumes and ventilation frequencies were similar for all exposures. Immediately after exposure, lungs were lavaged and the cell-free BALF was assayed for LDH activity (expressed as Wroblewski units) and protein content. Because minute ventilations were similar, the data are plotted against the exposure magnitude and expressed as the product of the inspired concentration and the exposure time (i.e., ppm × min). Data are presented as means ± 1 standard deviation with n = 3 observations. Data are plotted against the product of O3 concentration and exposure time (ppm × min). LDH and protein values did not differ from air control except for the 2-ppm exposures (* P < 0.05).

View larger version (12K): [in this window] [in a new window]   Figure 3.   A total of 300 U/ml LDH was added to cell-free BALF obtained from naive rat lungs and the solutions were exposed in small glass vessels to 0.32 ppm O3 in air (flow = 150 ml/min) under well-mixed, steady-state conditions (1). At various time points, samples were removed and assayed for LDH activity. The data are presented as means ± 1 standard deviation of three observations. O3 exposure induced a rapid loss of enzyme activity.

EthD-1-Positive Cells in the Axial Pathway

Once the axial pathway of the right middle lobe of each rat was exposed by microdissection (Figure 1, upper and lower left), the distribution of EthD-1-positive cells was evaluated on the costal half of this pathway beginning with the most proximal airway, the lobar bronchus (Site A in Figure 1, upper left). The axial path was defined as the series of airways that deviated least from the long axis of the preceding airway at each point of airway branching. This allowed us to compare the most proximal airways with the succeeding distal airways and the terminal bronchioles. Two of these axial sites (Sites B and D in Figure 1, upper left) were quantified. For each of five animals in each exposure group, the data for each site were expressed on a per-animal basis. The greatest abundance of EthD-1-positive cells in the airways making up the axial path were in the most proximal segments (Sites A and B in Figure 1, upper left) in the lungs exposed to 1 ppm O₃ (Figure 4, left image). Compared with FA-exposed lungs, there was no obvious qualitative increase in the abundance of EthD-1- positive cells in airways more distal than the fifth to seventh generations, represented by Site B in Figure 1. There was an elevation in the abundance of EthD-1-positive cells in some of the lungs exposed to 0.5 ppm at this and more proximal positions, but this was not consistent for all animals evaluated. Morphometric evaluation found that the number of EthD-1-positive cells in the proximal airways of the axial path was more than double in animals exposed to 0.5 ppm compared with those exposed to 0.25 ppm, and that the abundance of EthD-1-positive cells in lungs exposed to 1.0 ppm was twice that at 0.5 ppm (Figure 5a). The overall concentration-injury response was statistically significant (P < 0.0001). In the epithelium of all of the more distal branches in the axial path, EthD-1-positive cells were only slightly more abundant in some animals compared with FA control lungs. In the terminal bronchioles (Figure 6), which were the most distal extensions of the axial pathway (Figure 1, lower left and right), there was a slight increase in the abundance of EthD-1-positive cells compared with FA control lungs. Morphometric evaluation (Figure 5a) indicated an inhalation concentration- related increase in the abundance of EthD-1-positive cells in terminal bronchioles. When the three regions quantified in the axial pathway were compared, the regional analysis found that there was more injury in the most proximal positions (Sites B and D in Figure 1, upper left) than in the terminal bronchioles (P < 0.05).

View larger version (138K): [in this window] [in a new window]   Figure 4.   Confocal microscopic comparison of labeled nuclei from epithelium in adjacent airway branches from the proximal bronchi of the right middle lung lobe from a rat lung exposed to 1 ppm O3 for 90 min. These two cell populations are from branches in the axial path (Site B in Figure 1, upper left) and the associated minor daughter side branch (Site C in Figure 1, upper left) from the most proximal branch point evaluated. Nuclei of EthD-1- positive cells (red signal) were more abundant in epithelium from minor daughter side branches.

View larger version (13K): [in this window] [in a new window]   Figure 5.   Morphometric comparison of the abundance of nuclei of EthD-1-permeable epithelial cells in the airways of the axial pathway (a) (Sites B and D and terminal bronchiole [TB]) and two minor daughter side branches (b) (Sites C and E) of rat lungs exposed to O3 for 90 min. The sites are as illustrated in Figure 1. Abundance of EthD-1-positive cells increased with increased O3 concentration (P = 0.0001). Abundance of EthD-1-positive cells in proximal (B and C) and intermediate (D and E) sites in the axial pathway and side branches was greater than in the TBS (P < 0.05 for all comparisons). Abundance of EthD-1-positive cells was significantly greater in the minor daughter (E) than in the adjacent axial pathway (D) at the intermediate site (P = 0.004). Values are means ± 1 standard deviation; n = 5 for all points.

View larger version (90K): [in this window] [in a new window]   Figure 6.   Confocal microscopic comparison of the distribution of nuclei of EthD-1-permeable cells (red) within the epithelium of the terminal bronchioles in rat lungs exposed to FA or O3 at 1 ppm for 90 min.

EthD-1-Positive Cells in Side Branches

In the process of exposing the airway branches in the axial path by microdissection, the side branches, termed "minor daughters" because they had greater deviation from the long axis of the parent branch and had the smaller diameter of the two branches, were also opened (Sites C and E in Figure 1). Comparison of these branches at most sites indicated that regardless of position (proximal or distal) within the airway tree, they had higher densities of EthD-1-positive cells than did their accompanying sites in the axial path (Figure 4). Evaluation of the sites of bifurcation themselves did not reveal a consistent increase in the abundance of EthD-1-positive cells at any O₃ concentration. The majority of the EthD-1-positive cells in the minor daughter airways were located slightly downstream from these points. In lungs exposed to different concentrations of O₃, there appeared to be a concentration-dependent increase in the abundance of EthD-1-positive cells in minor daughter airways (Figure 7). In more distal airway branches, there was no apparent difference in the density of EthD-1-positive cells in lungs exposed to FA or to the two lowest concentrations of O₃. However, in lungs exposed to 1.0 ppm there was a marked increase in the density of positive cells. Morphometric evaluation of the density of EthD-1- positive cells in two of these minor daughter branches, one proximal (Site C in Figure 1, upper left) and one more distal (Site E in Figure 1, upper left) showed concentration-dependent increases in both (Figure 5b). Both of these airways had significantly increased densities of EthD-1-positive cells compared with their corresponding branches located in the axial path (Site B versus Site C, P < 0.07; Site D versus Site E, P < 0.004; Figure 5). Comparison of minor daughter branches by region and with axial path airways indicated that there were concentration- and region-dependent differences in density of EthD-1-positive cells, with the highest density in the proximal minor daughter airway (Site C) at 1.0 ppm O₃ and the least abundance in terminal bronchioles at 0.25 ppm O₃. Minor daughter airways at both sites (C and E) had more EthD-1-positive cells than did the terminal bronchioles (P < 0.05).

View larger version (170K): [in this window] [in a new window]   Figure 7.   Confocal microscopic comparison of nuclei from proximal minor daughter airway (Site C in Figure 1) from rat lungs exposed to FA or to O3 at 0.25, 0.5, or 1.0 ppm for 90 min. Nuclei of EthD-1-positive cells (red) increased in abundance with increasing inhaled concentration.

Histopathology of Whole Mounts

To assess the nature of the cells labeled with EthD-1, portions of the conducting airway from the most proximal airway generation quantified were assessed by high-resolution light microscopy. Sections of tissue removed from the airways of the lungs exposed to 1.0 ppm O₃ showed EthD-1-positive cells in 1-µm sections when evaluated by laser confocal scanning microscopy (Figure 8a). Subsequent staining of the same sections with Toluidine blue indicated that the EthD-1-positive nuclei belonged to epithelial cells (Figure 8b). High-magnification evaluation of these cells indicated that only the cells in the epithelial lining that contained pyknotic nuclei and disrupted cytoplasm were positive for EthD-1, whether they appeared to be in the process of exfoliation (Figure 8c) or still attached to the basal lamina (Figure 8d).

View larger version (56K): [in this window] [in a new window]   Figure 8.   Analysis of the same site on methacrylate sections using the confocal laser microscope (a) and with transmission microscopy after staining with Toluidine blue (b, c, and d). The nuclei of EthD-1-positive cells (red in a) are associated with pyknotic nuclei of cells attached to the basal lamina (d) or partially or completely exfoliated from the epithelium (c).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The current study presents an approach for defining cellular injury across broad regions of the lung. Our strategy was to combine and adapt well-established methods for isolating lungs for controlled ventilation and perfusion (24, 25) and defining local airway microenvironments by microdissection (20, 30) with morphometric approaches (26, 27, 31). We also modified methods used in vitro to characterize cellular injury on the basis of the loss of plasma membrane integrity. This approach utilizes dyes routinely employed for identifying the nuclei of permeable cells (32). To establish that the nuclei identified within intact lungs by the impermeant dye EthD-1 represented injured cells, we embedded the same specimens for high-resolution epifluorescence and histopathology (Figure 8). The labeled cells exhibited the histopathologic characteristics of cell necrosis. Using laser confocal microscopy, we have demonstrated that mapping acute cellular injury throughout the distal respiratory system not only is feasible but also provides spatial information not available through other strategies. In addition, the application of three- dimensional morphometric techniques (26, 27, 31) to the image stacks produced by a confocal laser scanning microscope provides quantitative information regarding the distribution of cell injury within the airway tree and allows statistical comparisons between cell populations occupying closely contiguous microenvironments.

Results from this study indicate that this method is more sensitive in detecting cell injury than are injury markers (LDH activity and total protein) routinely tested in BALF. For example, exposure to 0.25 ppm O₃ produced localized regions of epithelial damage that were readily observable by the fluorochrome labeling technique. We observed injured cells during even relatively brief exposures (0.25 to 0.5 ppm for 20 min). However, the other markers of injury were not altered at these concentrations. Changes in these markers were not detectable in BALF until the exposure concentration exceeded 1.0 ppm. This lack of sensitivity with BALF may result from the combined effects of lavage-induced dilution, O₃ exposure-induced inactivation, or modification of marker proteins and variable perfusion and protein delivery to injury sites. Our observations also emphasize the focal nature of the injury response to O₃ even in a contiguous airway path. There were pronounced differences between proximal and distal sites in the axial path of airflow, and airway side branches downstream of bifurcation points had more injury than did the comparable sites in the axial pathway. When ventilation and perfusion were carefully controlled, the injury pattern appeared to be dependent on inhaled concentration.

It should be noted that this methodology and interpretation of the resulting data may be confounded by a number of factors. Other stresses, such as respiratory infections, may alter epithelial permeability and increase the background of EthD-1-positive cells. Further, we observed that if lungs were ventilated with filtered air for a 30 to 45-min period after exposure cessation, the number of EthD-1-positive cells declined due to shedding of damaged epithelial cells. Consequently, the time at which samples are obtained relative to the exposure protocol is critical. Nonetheless, this new approach to assessing acute lung injury enables the addressing of questions that currently remain unanswered because the method can be combined with specimen preparations applicable to a wide range of biochemical and physiologic analyses (34, 36).