Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The distal conducting airway epithelium is a primary site of injury after exposure to many air pollutants, oxidant gases, metabolically activated toxicants, and particles. In the mouse and other species, nonciliated bronchiolar (Clara) cells are the predominant epithelial cell type present in the distal airways (1). Clara cells are also the principal site of xenobiotic metabolism by the cytochrome P450 monooxygenase system within the lung (2, 3). The abundance of Clara cells in a primary target region for air pollutants, coupled with their ability to metabolize chemicals to toxic intermediates, results in the Clara cell being uniquely susceptible to injury. Despite the well-recognized site- and cell-selective toxicity to Clara cells, the intracellular changes associated with toxicity are undefined.

The liver, because of its location and complement of metabolic enzymes, is another frequent site of chemically induced injury, and it has been extensively studied. Hepatocyte toxicity can be caused by several forms of cell injury, including hypoxia, hyperthermia, chemical toxicity, and oxidative stress (4), all of which result in plasma membrane blebbing. In fact, blebbing is widely used as a morphologic indicator of general hepatocyte stress in toxicologic tests in vitro (10). Bioactivated toxicants cause characteristic changes in liver cell morphology in vivo, including formation of large, clear cytoplasmic vacuoles; dilation of smooth endoplasmic reticulum (SER); loss of ribosomes from rough endoplasmic reticulum (RER); aggregation of microfilaments; and protrusion of organelle-depleted blebs extending into the space of Disse (11). Similar changes also occur during toxicity in isolated hepatocytes and in the hypoxic isolated perfused liver (4, 12). Hepatocyte bleb formation precedes the onset of mitochondrial permeability which, in turn, precedes loss of cell membrane integrity (7). Blebs are characterized as lacking organelles but containing free ribosomes and amorphous material. Although these acute ultrastructural changes have been documented in liver cells exposed to P450-activated cytotoxicants, they have not been defined for the metabolically active cells of the lung, including Clara cells; nor have they been correlated directly to a defined marker for epithelial injury such as membrane permeability.

To study acute Clara cell injury, we chose to use the metabolically activated Clara cell toxicant naphthalene. Previous studies have shown that naphthalene causes severe dose- and site-selective injury to distal bronchiolar Clara cells in the mouse (13). Clara cells within the most distal conducting airways, the terminal bronchioles, are selectively injured. Toxicity is due to the metabolism of naphthalene by cytochrome P450 monooxygenases to a toxic intermediate that causes distal airway Clara cell swelling, vacuolization, and exfoliation into the lumen of the airways 24 h after injury is initiated (14). Susceptibility correlates with the presence of CYP2F2 within Clara cells (15, 17). For our studies of the early phases of acute injury (1 to 6 h) we chose a naphthalene dose for which both the long-term injury and repair responses and the injury extent at 24 h have been characterized previously (13, 16, 18).

The purpose of our current study was to define the temporal pattern of intracellular changes in the 6 h immediately after naphthalene exposure to identify critical events involved in the initiation of cytotoxicity. To accomplish this goal we adapted the use of cell-permeant and -impermeant nuclear binding fluorochromes, conventionally used on isolated cells, for use on live whole-lung preparations from naphthalene-treated mice (E. Postlethwait and colleagues, unpublished observations; and [19] and [20]). Samples were also examined using high-resolution light, electron, and epifluorescent microscopy. These methods allowed direct correlation of the characteristic features of Clara cell necrosis with cell permeability as indicated by fluorescence on the same samples. This study addresses the following questions: (1) what is the short-term (1- to 6-h) spatial and temporal pattern of cell permeability after naphthalene injury to bronchiolar epithelium in vivo? (2) What is the temporal pattern of key organellar changes in injured Clara cells? (3) How does membrane permeability relate to the temporal pattern of intracellular changes?

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Naphthalene was purchased from Aldrich Chemical (Milwaukee, WI). Corn oil (Mazola) was manufactured by Best Foods/CPC International (Englewood Cliffs, NJ). Glutaraldehyde, paraformaldehyde, lead citrate, toluidine blue, azure II, and Araldite 502 resin were obtained from Electron Microscopy Sciences (Fort Washington, PA). JB-4 Embedding Kit (methacrylate resin), osmium tetroxide, and uranyl acetate were purchased from Polysciences, Inc. (Warrington, PA). The fluorochromes ethidium homodimer-1 and Yo-Pro-1, as well as nonfluorescing mounting media (ELF: Enzyme Linked Fluorescence Kit), were obtained from Molecular Probes (Eugene, OR).

Animals

Adult, 8- to 10-wk old, male, viral antibody-free Swiss Webster mice (CFW; Charles River Laboratories, Wilmington, MA) were used. Mice were housed in a high-efficiency particle air-filtered cage rack in Association for Accreditation and Assessment of Laboratory Animal Care (AAALAC)-approved facilities on a 12/12 light/dark cycle with food and water ad libitum for at least 5 d before use.

Experimental Protocol

All animals were treated at the same time of day, between 8:00 and 10:00 A.M., with either 200 mg/kg naphthalene or a corresponding volume of carrier (corn oil). Animals were killed with an overdose of pentobarbital at 1, 2, 3, and 6 h after treatment (PT). The trachea was cannulated and lungs were inflated to capacity in situ with 37°C ethidium homodimer-1 (5 µM) in Ham's F12 Nutrient Mixture (Life Technologies, Grand Island, NY) for 10 min to label the nuclei of any membrane-permeable cells fluorescent red. Lungs were then lavaged three times with 37°C F12 media to remove any unincorporated ethidium and were fixed at 30 cm of pressure with 330 mOsm Karnovsky's fixative (1% glutaraldehyde/0.5% paraformaldehyde in cacodylate buffer, pH 7.4) for 1 h. Lungs were stored in fixative in the dark until used. Each experiment contained one naphthalene-treated and one carrier-treated animal at each time point: 1, 2, 3, and 6 h PT. The experiment was repeated three times.

Mapping the Distribution of Permeable Cells

The right middle lobe from the ethidium-lavaged fixed lung was microdissected by cutting down the airway lumen to expose the long axis of the conducting airway tree including the most distal six to eight conducting airway generations and the terminal bronchioles. The microdissected lung was counterstained with the nuclear binding fluorochrome Yo-Pro-1 (4 µM) for 20 min to label the remaining nuclei with green fluorescence. Samples were imaged using a scanning laser confocal microscope (Bio-Rad, Hercules, CA), a 20× long working distance water-immersion objective, and a dual filter set. Using the dual filter set (excitation/emission 514/ 527 and 540 /600), two fluorescent wavelengths were captured. The laser scanned through the entire three-dimensional sample, producing a series of images (for both wavelengths) at sequential focal planes that were "stacked" one on top of the other to generate two separate pseudocolor (red and green) images of the entire depth of the sample. The two stacked images were then merged into one image to generate the image of red (permeable) cell nuclei in the context of the rest of the cell nuclei (green, impermeable). Each airway map was a composite of 8 to 10 fields collected in this manner. To determine whether injury varied by position within each airway circumference, both halves of the sliced dissected airways were imaged for three additional mice at 6 h PT. There was no difference in abundance or distribution of permeable cells from side to side within the same airway (data not shown).

High-Resolution Fluorescence Microscopy

Ethidium-stained tissue was embedded in methacrylate resin and sectioned at 1 µm. Sections were coverslipped with a water-soluble, nonfluorescing mounting media. Using an Olympus BH-2 epifluorescent microscope and a wide-band ultraviolet (UV) fluorescence excitation/emission filter set, tissue sections containing terminal bronchioles were photographed using slide film. Due to two phase shifts in the wide-band UV excitation/emission filter set, the tissue autofluoresces blue and the ethidium-positive nuclei fluoresce white. Slides were scanned into computer images using a slide scanner. Coverslips were removed, and the same sections were stained with toluidine blue. Staining with toluidine blue quenched the ethidium fluorescence. Images of the same fields as the fluorescent slides were captured using an Olympus Provis computerized microscope in brightfield mode. Brightfield and darkfield images were aligned and adjusted to the same magnification in Adobe Photoshop v4.0 (Adobe Systems, San Jose, CA).

Transmission Electron Microscopy

Samples for electron microscopy were taken from the same lungs that were used for the fluorescence-permeability studies. Lung slices from the left lobe were postfixed in 1% osmium tetroxide in Zetterquist's buffer, processed by large block methodology, and embedded in Araldite 502 epoxy resin (21). Specimens were sectioned at 1 µm and stained with methylene blue/azure II and imaged on an Olympus Provis microscope. Selected areas of the large blocks were sectioned at 70 nm with a Sorvall 5000 ultramicrotome, stained with uranyl acetate and lead citrate, and examined using a Zeiss EM-10 electron microscope at 60 kV.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Permeability and Morphology

Low-magnification composite "maps" (Figure 1) of the distribution of permeable cells (red) in the context of all the cell nuclei (green) were generated in distal bronchioles using a laser scanning confocal microscope on fixed, microdissected tissue specimens. These composite micrographs mapped at least the most distal six airway generations in the axial path from the right middle lobe of mice treated with either carrier or naphthalene. The lungs of carrier-treated control mice had little to no ethidium-positive (red) cells in the airway epithelium at all time points after carrier treatment (Figure 1A). Likewise, the lungs from naphthalene-treated mice had no ethidium-positive cells at 1 h PT and very few positive cells (two or fewer per bronchiole) at 2 h PT (Figures 1B and 1C). The few permeable cells present were located in terminal bronchioles one airway-branching generation removed from the axial path and were not present in the terminal bronchioles located at the end of the axial path (Figure 1C). By 3 h PT more permeable cells were present, and these cells were located in the terminal bronchioles and at distal airway bifurcations contiguous to the terminal bronchioles (Figure 1D). All terminal bronchioles examined had some permeable cells. At 6 h PT (Figure 1E), many more permeable (red) cells were present, and these were predominantly located within terminal bronchioles. Permeable cells were also present in the lung parenchyma immediately distal to the terminal bronchiole.

View larger version (133K): [in this window] [in a new window]   Figure 1.   Laser scanning confocal composite micrograph of the distribution of ethidium homodimer-1-positive permeable cells (red, arrows) and impermeant cells labeled after fixation with Yo-Pro-1 (green) in the distal bronchioles of the right middle lobe from mice treated with either carrier (A; 6 h PT) or 200 mg/kg naphthalene 1, 2, 3, or 6 h previously (B-E, respectively). Bar = 500 µm.

Paired fluorescent and brightfield images of the same fields, from 1-µm sections of mouse lungs that had been infused with ethidium, were used to compare cell permeability directly with morphology (Figure 2). Terminal bronchioles from carrier-treated mice (Figure 2A) contained no permeable epithelial cells (permeable cells have white fluorescent nuclei; see Figure 2G). The epithelium of the mouse terminal bronchiole contained both Clara and ciliated cells (Figure 2B). Cells were not vacuolated or swollen. The bronchiolar epithelium from mice 1 h PT was relatively unchanged compared with controls (Figures 2C and 2D) and there were no permeable cells (Figure 2C). However, some of the Clara cells were swollen and slightly vacuolated (Figure 2D). By 2 h PT, the terminal bronchiolar epithelium contained many vacuolated and swollen Clara cells (Figures 2E and 2F). Ciliated cells extended under the adjacent vacuolated Clara cells. Although the Clara cells were clearly vacuolated, they were not permeable (Figures 2E and 2F). By 3 h PT, as illustrated in Figures 2G and 2H, apical cytoplasmic blebs were a prominent structural feature of the vacuolated Clara cells. Cytoplasmic blebs were not identified in permeable Clara cells (see fluorescent nucleus in Figure 2G). Ciliated cells were squamated (Figure 2H). Cell debris was found rarely in the lumen of the airway (not shown). At 6 h after naphthalene treatment, the terminal bronchiolar epithelium contained many permeable Clara cells with condensed nuclei (Figures 2I and 2J). No Clara cells with apical blebs were observed. Cell debris was frequently seen in the airway lumen.

View larger version (109K): [in this window] [in a new window]   Figure 2.   Paired epifluorescent and brightfield images of terminal bronchiolar epithelium from mice injected with either carrier (A and B; 6 h PT) or naphthalene at 1 (C and D), 2 (E and F), 3 (G and H), or 6 (I and J) h PT. Epifluorescent images (left column) are of ethidium homodimer-1-positive permeable cell nuclei (white arrowheads) through a wide-band UV excitation and emission filter set that also transmits the normal fluorescence of fixed tissue (blue). Brightfield images (right column) are the same section after additional staining with toluidine blue to show cellular detail, including Clara cells (CC), ciliated cells (Ci), and cells with apical membrane blebs (*). Bar = 50 µm.

Ultrastructure

In carrier-treated control animals (Figure 3), both Clara and ciliated cells were present (Figure 3A). Clara cells contained SER in the apex of the cell and RER next to the nucleus. There were many mitochondria in both circular and columnar profiles distributed throughout the Clara cell cytoplasm (Figure 3A). The apex of the cell also contained circular, electron-dense, membrane-bound secretory granules. Higher magnification of the apex of a Clara cell revealed mitochondria with few cristae and abundant SER in a relatively electron-dense cytoplasm (Figure 3B). Secretory granules were segregated from the plasmalemma by a zone of intermediate filaments.

View larger version (125K): [in this window] [in a new window]   View larger version (181K): [in this window] [in a new window]   Figure 3.   Ultrastructure of bronchiolar epithelium in carrier-treated (corn oil) control mice. (A) The Clara cell, flanked on either side by ciliated cells, contains SER in the apex of the cell and RER next to the nucleus (N). There are many mitochondria in both circular and columnar profiles distributed throughout the Clara cell cytoplasm. The apex of the cell also contains circular, electron-dense, membrane-bound secretory granules (G). (B) Higher magnification of the apex of a Clara cell reveals mitochondria (M) with few cristae and abundant SER (arrowheads) in a relatively electron-dense cytoplasm. Organelles are held from the plasmalemma by a zone of intermediate filaments (arrows). L = airway lumen. (A) Bar = 2 µm; (B) bar = 1 µm.

At 1 h PT (Figure 4), Clara cells had increased cytoplasmic spaces near the nucleus. Abundant mitochondria, in predominantly circular profiles, were distributed both apically and basally throughout the cell cytoplasm (Figure 4A). The apex of the cell also contained circular, electron-dense, membrane-bound secretory granules. Higher magnification of the apex of a Clara cell showed a loss of electron density in the cytoplasm and large cytoplasmic spaces (Figure 4B). Mitochondria were round with few cristae. The SER was swollen in focal areas. Secretory granules were separated from the plasmalemma by a zone of intermediate filaments.

View larger version (123K): [in this window] [in a new window]   View larger version (174K): [in this window] [in a new window]   Figure 4.   Ultrastructure of bronchiolar epithelium from mice 1 h PT with naphthalene. (A) A Clara cell with increased cytoplasmic spaces (*) near the nucleus (N) and abundant mitochondria distributed both apically and basally throughout the cell cytoplasm. The apex of the cell also contains circular, electron-dense, membrane-bound secretory granules (G). (B) Higher magnification of the apex of a Clara cell with loss of electron density in the cytoplasm and many cytoplasmic spaces (*). Mitochondria (M) are round with few cristae. The SER is beginning to swell in focal areas (arrowhead). Secretory granules (G) are held from the plasmalemma by a zone of intermediate filaments (arrows). L = airway lumen. (A) Bar = 2 µm; (B) bar = 1 µm.

As shown in Figure 5, terminal bronchiolar Clara cells had increased consolidation of cytoplasmic spaces predominantly above the nucleus in the apex of the cell 2 h PT. Fewer mitochondria were found below the nucleus. The apex of the cell contained few circular, electron-dense, membrane-bound secretory granules. Higher magnification of the apex of a Clara cell at 2 h PT revealed a continuing loss of electron density in the cytoplasm and increased cytoplasmic spaces, many of which contained free ribosomes (Figure 5B). The SER was markedly swollen. The apex of the cell contained numerous unorganized filaments.

View larger version (140K): [in this window] [in a new window]   View larger version (214K): [in this window] [in a new window]   Figure 5.   Ultrastructure of bronchiolar epithelium from mice 2 h PT with naphthalene. (A) Clara cell with increased consolidation of cytoplasmic spaces (*) predominantly above the nucleus (N) in the apex of the cell. Fewer mitochondria are found below the nucleus. The apex of the cell contains a few circular, electron-dense, membrane-bound secretory granules (G). (B) Higher magnification of the apex of a Clara cell with loss of electron density in the cytoplasm and cytoplasmic spaces which contain free ribosomes (open arrow). Swollen SER (arrowheads) is abundant. The apex of the cell contains numerous unorganized filaments (arrows) as well as a layer of filaments near the plasmalemma (arrows). L = airway lumen. (A) Bar = 2 µm; (B) bar = 1 µm.

At 3 h PT (Figure 6), many Clara cells had apical membrane blebbing and heavy consolidation of cytoplasmic spaces (Figure 6A). A few swollen mitochondria were located in the apex of the cell. Clara cells contained several circular, electron-dense, membrane-bound secretory granules, some on the lateral margins of the cell. Figures 6B and 6C are higher magnifications of the apex of two different Clara cells at the junction between the main cell body and an apical membrane bleb. The bleb was packed with swollen SER and was walled off from the body of the cell by a zone of cytoplasmic filaments. The main body of the cell contained large cytoplasmic spaces, mitochondria with increased granulation in the matrix, and secretory granules. Some blebs contained mitochondria in addition to the swollen endoplasmic reticulum that was uniformly found in all blebs. The blebs that also contained mitochondria generally had larger clear cytoplasmic spaces than did the blebs that did not contain mitochondria (compare Figures 6B and 6C).

View larger version (132K): [in this window] [in a new window]   View larger version (158K): [in this window] [in a new window]   View larger version (206K): [in this window] [in a new window]   Figure 6.   Ultrastructure of bronchiolar epithelium from mice 3 h PT with naphthalene. (A) Clara cells with consolidation of cytoplasmic spaces (*) and pronounced apical membrane blebbing (B). A few swollen mitochondria are now found in the apex of the main cell body as well as in the bleb. The cell contains several circular, electron-dense, membrane-bound secretory granules (G). An adjacent cell has increased consolidation of cytoplasmic spaces into large vacuoles (*). (B) Higher magnification of the apex of a Clara cell at the junction of the main cell body with an apical membrane bleb. The bleb is packed with very swollen SER (arrowheads), contains mitochondria (M) with granular matrices, and is walled off from the body of the cell by a zone of cytoplasmic filaments (arrows). L = airway lumen. (C) Another Clara cell apex with less swollen SER (arrowheads), but lacking large organelles such as mitochondria, is also walled off from the main cell body by a zone of intermediate filaments (arrows). The main body of the cell contains large cytoplasmic spaces, mitochondria (M) with increased granulation in the matrix, and secretory granules (G). Bars: (A), 2 µm; (B) and (C), 1 µm.

At 6 h PT, Clara cells had widespread loss of cytoplasmic density (Figure 7) and pronounced swelling of the SER. The majority of the mitochondria, whether swollen or with granular matrices, were clustered above the nucleus (Figures 7A and 7B). Secretory granules were present. Figure 7C is a higher magnification of the apex of the same Clara cell shown in Figure 7B, with very swollen SER and abundant mitochondria at various stages of degeneration. Compared with 3 h PT (Figure 6), more mitochondria had heavy granulation of their matrices and were quite swollen. Cell debris, including apparently intact cells with nuclei as well as small pieces of membrane-bound cytoplasm that lacked nuclei but contained swollen SER, were frequently noted in the lumen of the airway. Cell debris without an associated membrane was rarely found in the airway lumen.

View larger version (142K): [in this window] [in a new window]   View larger version (159K): [in this window] [in a new window]   View larger version (173K): [in this window] [in a new window]   Figure 7.   Ultrastructure of bronchiolar epithelium from mice 6 h PT with naphthalene. (A) Clara cell with widespread loss of cytoplasmic density and pronounced swelling of the SER (arrowheads). Severely swollen mitochondria (M) and increasing numbers of mitochondria with granular matrices are distributed throughout the cell but predominantly clustered above the nucleus (N). (B) Another Clara cell with loss of cytoplasmic density, small amounts of swollen SER (arrowheads), and numerous mitochondria (M) with granular matrices. (C) Higher magnification of the Clara cell shown in (A) with very swollen SER (arrowheads) and abundant mitochondria (M) at various stages of degeneration. Some mitochondria have increased granulation of their matrices and other mitochondria are quite swollen. L = airway lumen. Bars: (A) and (B), 2 µm; (C), 1 µm.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The purpose of our current study was to define the spatial and temporal pattern of intracellular events preceding the loss of membrane integrity in bronchiolar epithelium during acute, chemically induced cytotoxicity. We focused on cells occupying a specific site within the airway tree (terminal bronchioles) for which the longer-term sequelae of naphthalene injury (Clara cell necrosis and exfoliation) are well-established (13, 16). There is a close spatial relationship between the abundance of permeable cells within terminal bronchioles observed in our current study and the pattern of toxicity reported previously at 24 h after exposure to naphthalene (13, 16); terminal bronchioles contain the most extensive concentrations of permeable cells, and injured cells are Clara cells. We have also identified a temporal sequence of events involved in the early phases of cytotoxicity. The earliest phase includes focal swelling of the Clara cell SER that occurs simultaneously with early rounding and loss of Clara cell profile in some cells. This is indicative of the onset of cell swelling. In the second phase, most of the Clara cells are swollen and some have consolidation of cytoplasmic spaces visible by light microscopy as vacuoles. Cytoskeletal rearrangement begins as unorganized filaments are relocated in the apex of injured cells and lost from the plasmalemma. Very few cells (two or fewer per bronchiole) are permeable. The third phase of cell injury includes pronounced changes in Clara cell ultrastructure that result in formation of apical membrane blebs that are segregated from the main body of the cell by a zone of organized filaments. All blebs are packed with swollen endoplasmic reticulum and some contain swollen mitochondria. All the injured Clara cells contain mitochondria with granular matrices. All terminal bronchioles contain some permeable cells. However, only the few Clara cells that lack blebs are permeable. In the fourth and final phase many cells have lost membrane integrity, and these are present in great abundance in terminal bronchioles. Clara cell blebs have been lost and many Clara cells contain condensed nuclei, vacuoles, and rarifaction of the cytoplasm, as well as mitochondria in various stages of degeneration. The mitochondria tend to cluster above the nucleus, and many are granular and very swollen. Some ciliated cells have begun to squamate at this phase. These findings suggest that the sequence of major events during acute Clara cell toxicity in vivo is as follows: cytoplasmic vacuolization, SER swelling, rearrangement of cytoskeletal filaments, formation of a cytoplasmic apical bleb, and cell membrane permeability. Changes in mitochondria are apparent following the initiation of SER dilation and progress into the period when plasma integrity is lost.

The cytoplasmic blebs formed in Clara cells are quite similar to blebs formed in hepatocytes after a variety of toxic stresses. They resemble the blebs found in hepatocytes exposed both in vivo and in vitro to the P450-activated toxicants dacarbazine and menadione (11, 12, 22). Both Clara cells and hepatocytes form blebs that contain dilated SER and are separated from the main cell body by a zone of intermediate filaments. However, blebs in hepatocytes do not contain organelles other than endoplasmic reticulum, whereas some of the blebs in Clara cells also contain mitochondria (4). The inclusion of mitochondria in Clara cell blebs may be unique to Clara cells or may reflect a greater relative injury to the Clara cells as compared with hepatocytes. The blebs that contain mitochondria are found in Clara cells that have the most dilated SER. These cells may represent the most severely injured cells within the spectrum of Clara cell injury. Bleb formation in hepatocytes in culture is an early sign of toxicity that precedes mitochondrial membrane depolarization, disintegration of lysosomes, and loss of cell membrane integrity (7). Our findings in Clara cells are in agreement with this pattern of ultrastructural changes preceding loss of cell membrane integrity in hepatocytes. There have been other studies of acute injury to the lung in vivo, but these have focused primarily on injury by oxidant pollutants such as ozone and NO₂, and on later time points (such as 24 h PT). Although it is not known whether nonbioactivated cytotoxicants produce this type of blebbing, other studies with P450-activated compounds suggest that Clara cell blebbing may be a uniform response to bioactivated toxicants (23, 24). On the basis of the data presented in these previous studies it is difficult to determine whether the blebs they described are passive deformations of the Clara cell apex or are part of an active process such as we have identified in response to acute injury. Neither the sequence of cytotoxic events (before and after bleb formation) nor the airway location of the blebbing cells are identified in these earlier studies. Our work suggests that, for the bioactivated cytotoxicant naphthalene, the process of bleb formation is an active process and that bleb formation is restricted to a single phase of the cytotoxic response.

The uniform lack of blebs at 6 h PT suggests that Clara cell blebs may also be shed. Loss of small amounts of apical cytoplasm has been reported previously in studies of Clara cell apocrine secretion of granules in mice (25), and we found similar membrane-bound structures in the airway lumen 6 h PT. Hepatocytes are thought to shed blebs as a result of cytotoxicity. Cytoplasmic fragments that may represent bleb shedding can be induced by reoxygenation of liver following hypoxia (26). This suggests that bleb shedding from injured hepatocytes may be a protective mechanism. Clara cell blebbing may be a similar mechanism to protect the cell by isolating and removing damaged SER. SER is a likely site for high concentrations of reactive naphthalene intermediates and their associated damage because the SER is the intracellular region that contains abundant cytochrome P450 monooxygenases, which metabolize naphthalene (15, 17). However, blebbing is clearly not sufficient to protect the injured cells, because the majority of Clara cells in terminal bronchioles will become membrane-permeable in 6 h and will exfoliate by 24 h PT (13, 16). The only cells permeable to ethidium were those Clara cells that lacked blebs. One possibility is that membrane permeability in cells lacking blebs is due to bleb rupture. In this case, blebs would not be protective. Both bleb shedding and bleb rupture are supported by the observation of pieces of cytoplasm bound by cell membrane, as well as debris without an accompanying membrane, in the airway lumen 6 h PT. Despite the potentially protective mechanism of cell blebbing, there may be some other unidentified change in the cell that occurs earlier than 3 h PT that results in inevitable Clara cell injury and exfoliation, such as adduct formation with a molecule that is critical for cell function. We have previously observed this during naphthalene injury and repair in an ex vivo model where bronchioles were removed from naphthalene-treated mice 2 h PT and then underwent the cytotoxic changes in vitro (27). This suggests that the injury process was already initiated within the Clara cells at 2 h PT and could not be due to further accumulation of naphthalene or its metabolites. Further studies are needed to define the mechanism of bleb shedding, or rupture, and to determine whether cells with blebs can be "rescued" before loss of membrane integrity (i.e., does bleb formation indicate that the injury process is irreversible?).

Research on bleb formation in hepatocytes has suggested that blebs are caused by one or more of these various mechanisms: transformation of preexisting microvilli, changes in the cortical cytoskeleton, and disturbances in both thiol and calcium homeostasis within the injured cell (8, 12, 22, 26, 28). It is likely that disturbances in cytoskeletal, thiol, and calcium homeostasis are involved in the Clara cell response to P450-activated toxicants. Glutathione conjugation is a major element of naphthalene metabolite detoxification, and a drop in glutathione levels is likely to precede bleb formation and mitochondrial damage. The cytoskeleton of Clara cells has pronounced changes in response to naphthalene toxicity, as indicated by the filaments present at 2 h PT, and organized into blebs at 3 h PT. Calcium homeostasis has not been assessed in individual injured Clara cells in vivo. However, the calcium content of entire lungs (micrograms per gram of dry weight) from mice treated with high doses of another P450-activated Clara cell toxicant, trichloroethylene, was significantly increased 24 h PT (29). Further work is needed to determine the nature and extent of changes in thiol and calcium homeostasis in response to Clara cell injury. Additional studies are also needed to determine whether the Clara cell ultrastructural changes reported in this study are specific to naphthalene metabolites or to P450-activated compounds in general, or are a uniform response to any toxic stress.

In this study we determined that the early (1- to 6-h) spatial and temporal pattern of cell permeability after naphthalene injury to bronchiolar epithelium in vivo involved terminal bronchioles exclusively; and that cell permeability was a late event, occurring in some cells as early as 3 h PT and in many cells by 6 h PT. We used three-dimensional approaches to define sites of increased cell permeability because of the focal and site-specific nature of bioactivated toxicant injury in the lung. Membrane permeability was then related to the temporal pattern of intracellular changes that preceded it, which included SER swelling, rearrangement of cytoskeletal filaments, formation of cytoplasmic apical blebs on injured terminal bronchiolar Clara cells, and swollen mitochondria with granular matrices. These studies are the first steps toward defining the intracellular mechanisms involved in Clara cell susceptibility to metabolically activated cytotoxicants. Identification of the mechanisms of toxicity, which is the long-term goal of these studies, is a requisite step before interventions for lung injury caused by exposure to environmental pollutants can be designed.