Rat Lung Alveolar Type I Epithelial Cell Injury and Response to Hyperoxia James B. Harris, Ling-Yi Chang, and James D. Crapo Departments of Medicine and Pathology, Duke University Medical Center, Durham, North Carolina

Hyperoxia has been shown to cause extensive lung injury, which involves all components of the alveolar septum, although the type I epithelium has generally been reported to be resistant to significant injury. Electron microscopic morphometry was performed to define changes in volumes of subcellular components of alveolar epithelial cells in rats exposed to 85% O2 for 0, 7, and 14 d. Because of their large size, type I cells in control animals actually contain a greater volume of most of the organelles involved in cell metabolism than do type II cells. Hyperoxic exposure causes a dramatic change in the subcellular composition of the average type I cell, suggesting significant injury and/or response. Injury was suggested by the finding that lysosomes plus peroxisomes increased 1,250% after 7 d in hyperoxia and remained elevated by 200 % after 14 d of exposure. Volumes of mitochondria, rough endoplasmic reticulum, smooth endoplasmic reticulum, and Golgi apparatus increased by 100%, 51%, 91%, and 500%, respectively, after hyperoxia. Qualitative analysis showed an altered, ruffled air border with focal areas of cytoplasmic translucency (suggesting injury) and focal areas of subcellular hypertrophy. Exposure to hyperoxia was associated with more organelles being found in peripheral or attenuated portions of type I alveolar cells. Since the increase in type I organelles exceeds the volume of these organelles in its progenitor, the type II cell, it is likely that hyperoxia causes hypertrophy of the type I alveolar epithelium itself, independent of simple type II cell differentiation. Because of the large size and wide distribution of the type I cell, dramatic shifts in cell substructure caused by hyperoxia are more difficult to detect and require quantitative analysis to fully ascertain the extent of cell alterations.

The lung is a primary site of injury and is the organ most severely damaged by hyperoxia (1). Exposure to elevated, but sublethal, levels of O2 can lead to the development of tolerance in rats (2-4). When exposed to 100% O 2 , almost all adult rats die by 72 h of exposure. When exposed to 85 % O2 , significant lung injury occurs; however, the rats survive, and after 5 to 7 d of exposure can show tolerance to 100% O2 • Exposure to high concentrations of O2 causes a variety of dramatic effects to the alveolar septum, of which the pulmonary capillary endothelium has been identified as a major site of cell injury (5). Hyperoxia has been postulated to increase oxygen radical production at subcellular sites such as mitochondria, endoplasmic reticulum, nuclear membrane, cytoplasmic membrane, peroxisomes, and lysosomes (6-8). Oxygen radical production at specific intracellular sites is likely to be important in initiating pulmonary oxygen toxicity (6, 8, 9). With exposure to high enough levels of 01, cell destruction occurs. This process has been well documented for alveolar capillary endothelial cells and is postu(Received in original form March 16, 1990 and in revised form July 26, 1990) Address correspondence to: James D. Crapo, M.D., Division of Allergy, Critical Care and Respiratory Medicine, Duke University Medical Center, Durham, NC 27710. Am. J. Respir. Cell Mol. BioI. Vol. 4. pp. 115-125, 1991

lated to lead to loss of the alveolar capillary bed and death. The sensitivity of alveolar epithelial cells to hyperoxic injury is more controversial. Pathologic findings have been detected in studies of the rat, but, as a whole, these prior investigations have described subtle or no changes in the type I alveolar epithelium (4, 10-12). In contrast, results from studies of monkeys show evidence of extensive epithelial cell injury (13). Although the alveolar capillary endothelium in the monkey is also a major site of cellular injury, the alveolar epithelial lining in the monkey is almost completely destroyed after 4 d of exposure to 100% O2 • The dramatic discrepancy between the studies of rats and primates could be a reflection of species difference or a lack of recognition of the extent of epithelial cell injury in the rat. Functional changes in the alveolar epithelium have been found in animals exposed to hyperoxia. In addition to causing an alteration in permeability of the capillary endothelium, exposure to high levels of O2 causes a defect in the permeability of the alveolar epithelium (14-21). This damage commonly precedes the appearance of pulmonary edema, and can contribute to respiratory failure (22, 23). The structural pattern of the epithelial changes in rat lungs in response to hyperoxia has not been well defined. Questions of how extensive it is and whether the epithelium is injured and remodeled or replaced have not been addressed. Accurate knowledge of structural changes in the lung is necessary to gain new insight into cell microanatomy

116

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

and to accurately interpret biochemical studies carried out to define the processes that are important in both injury and adaptation to hyperoxic exposures. In addition, the changes occurring in the alveolar epithelium in response to hyperoxic stress cannot be easily detected qualitatively. No ultrastructural characterizations have been made that include a quantitative evaluationof the subcellular structures of the alveolar type I epithelium, nor of the changes occurring with hyperoxic insult. This work was carried out to define how the alveolar epithelium structurally responds to hyperoxic stress and to further elucidate and clarify the pathogenesis of pulmonary oxygen toxicity and tolerance. We chose to use 85 % O2 because it is a level that causes extensive injury without death to the rat, in addition to inducing tolerance to hyperoxia. Further, this level of O2 exposure has been widely used in studies of hyperoxia.

Materials and Methods Animals and Exposure Twelve, male CD strain specific pathogen-free, SpragueDawley rats, weighing 300 to 350 g each, were used. Four animals were exposed to 85 % O2 continuously for 7 d and four animals were exposed to 85 % O2 continuously for 14 d using polystyrene cages and followingprocedures previously described (3). The O2 concentration varied less than 2 % and the CO2 concentration was maintained below 0.5 % by providing sevento eight complete gas changes per hour. Four animals kept in air under identical exposure conditions were used as controls. During the exposures, food and water were provided ad libitum and the animals were kept in a 12-h-on, 12-h-off light cycle during the exposure period. Tissue Preparation After the exposures, each rat was anesthetized with an intraperitoneal injection of pentobarbital. Then in rapid succession the trachea was cannulated, the abdomen was opened, both hemidiaphragms were punctured to allow collapse of the lungs, and fixative was infused into the trachea at a constant pressure of 20 em of fixative above the midpoint of the chest wall. The fixative was 2 % glutaraldehyde in 0.1% CaCl 2 , and 0.085 M cacodylate buffer at pH 7.4 and had a total osmolality of 350 (24). After a minimum of 30 min of fixation inside the chest wall, the lungs were dissected out and the total lung volume was determined by volume displacement (24-26). Tissue samples for electron microscopy were randomly selected from four peripheral sites in the left lung of each animal. Before randomly selecting these sites, the lung was divided into four regions with one sampling being taken from each of the following regions: upper caudal, lower caudal, upper rostal, and lower rostal. These tissue blocks were washed in cacodylate buffer, postfixed with 2 % osmium tetroxide, then dehydrated in a graded series of alcohol, transferred to propylene oxide, and embedded in Epon. Thin sections were cut using a diamond knife, placed on 200mesh uncoated grids, and then stained with uranyl acetate and lead citrate for electron microscopic studies.

Electron Microscopy The four sample sites from each rat were studied by taking two independent series of random electron photomicrographs using a Zeiss lO-Ctransmission electron microscope. One series was for assessment of cell characteristics and the other was for assessment of subcellular characteristics. For the first series of micrographs, tissue was photographed in the upper left corner in each of 15 consecutive grid spaces from each section. These photographs were printed at 8,500x on 11- X 14-inch photographic paper containing an overlay pattern of 112 lines 2 ern long, as previously described (27). The end of each line was assumed to be a point, yielding a total of 224 points per micrograph. The distribution of points that fell on each of the various tissue components was tallied, and the number of times any line intercepted an alveolar or capillary surface was counted as previously described (lO). Since the type II cell represents the progenitor cell to the type I cell (28, 29) and is important in reparative processes after epithelial damage (13, 30-32), there is a point where the cells in transition from a type II cell to a type I cell possess characteristics of both cell types and are difficultto classify as either cell type. We attempted at one point in this study to identify these transitional cells as a separate group but found that it created even greater problems in cell classification. In actuality, the cell profiles in the alveolar epithelium that could not be classified as definitely either type I or type II were quite rare. Using straightforward criteria for identifying type II and type I cells, we were able to clearly categorize > 98 % of all alveolar epithelial cells. If the cell profile had any lamellar bodies and/or distinct microvilli, it was identified as a type II cell. If no lamellar bodies were present and there were numerous pinocytotic vesicles, it was regarded as type I, even if it approached a cuboidal shape. Cuboidal cells rich in Golgi, endoplasmic reticulum, and mitochondria and that lacked lamellar bodies, microvilli, and pinocytotic vesicles were very rare. A cell having these characteristics would have been classified as a type I cell. For the second series of micrographs, the tissue was photographed following an unbiased, systematic, nonoverlapping pattern in consecutive grid spaces from each section until 45 photographs were taken per site. These photographs were printed at 45,000 X on 11- X 14-inch photographic paper containing the same overlay pattern described above. The distribution of points that fell on each of 12 categories of subcellular structures for both type I and type II alveolar epithelium was counted. Identifications of nuclei, mitochondria, rough endoplasmic reticulum, smooth endoplasmic reticulum, free polyribosomes, Golgi apparatus, lamellar bodies (generally spherical membrane-bounded organelles with a lamellar appearance that are unique to the type II pneumocyte), and cytosol were made. Pinocytotic vesicles and small vesicular structures that look like pinocytotic vesicles were classified as vesicles. Multivesicular bodies were identified as membranous sacs containing numerous vesicles. Organelles bound by a single membrane and containing granular electron-opaque material were classified as lysosomes or peroxisomes (see Figures 5b and 6a). "Other" con-

Harris, Chang, and Crapo: Hyperoxic Changes in Type I Alveolar Epithelium

sisted of microtubules, fibers, and rare, nonidentifiable structures. Although intermediate fibers and microtubules are present in normal alveolar tissue (Figure 4c), they were not added as a separate category, and any points landing on fibers or microtubules were counted as "other." We did not include the ultrastructural category of composite body in the type II organelles. This structure has attributes of both lamellar bodies and multivesicular bodies and is presumably a combination of the two, present specifically in the type II cell (33, 34). Although we did recognize this structure, we did not include it as a separate category of cell substructure since it cannot always be identified accurately on a random two-dimensional section. If vesicles were present in an organelle profile, it was identified as a multivesicular body; otherwise, it was classified as a lamellar body. The category of vesicles included not only classic pinocytotic vesicles, but also small vesicular structures within the cell. It therefore could include Golgi vesicles that were sectioned on a plane such that the Golgi apparatus was not detected. In addition, the type II cell, which normally has very few pinocytotic vesicles, apparently acquires them after hyperoxic insult. This may be a consequence of an enhanced rate of differentiation of type II cells into type I cells. Morphometric Analyses The procedures for carrying out morphometric analyses in our laboratory were adapted from those described by Weibel (24), Underwood (35), and Weibel and Bolender (27) and have been reported in detail (11). The previously determined value of 0.81 (l0, 11)was used for the proportion of total lung volume in the rat that is alveolar region, defined as including alveoli and alveolar ducts. Two different determinations of surface area were made for both type I and type II pneumocytes, as previously described (36-38). Since the air surface side of alveolar type I epithelium has more surface irregularities than does the basement membrane on which it lies, we made surface area determinations for both air surface and basement membrane surface. For alveolar type II epithelium, since microvilli increase the air surface of the cell and the basement membrane is irregular and not parallel to the air surface, we made surface area determinations for both the entire air surface, including all microvilli and the interface created by drawing a smooth line across the surface of the cell and that went through the base of microvilli. For both cell types, the second described method was assumed to give the best estimate of the portion of "smooth" alveolar surface covered. The numeric density of lung cells (Ny) was determined using the relationship N; = NJJ5 (27), where NA is the frequency of occurrence of nuclear profiles per unit area on a random sectioning plane, and 15 is the mean caliper diameter of the nucleus. The frequency of occurrence of nuclear profiles per unit area of random sectioning plane (N A ) was determined directly on the electron microscope. Counting was performed only on grid squares covered entirely by the section and that contained no preparation artifacts that could hinder the accuracy of cell identification. The number of nuclear profiles of each class of cells found in 9 to 27 (aver-

117

age = 18) grid spaces in each of the 4 grids per animal was counted. The area of any grid space on any given grid varied less than 2 %. However, the variation of grid space area between grids was significant. Therefore, we photographed one grid space from each grid, enlarged the photograph of that grid space to 2,000x, and determined the grid space area by digitization. The mean caliper diameters of rat pulmonary cell nuclei have been previously defined for both control animals and animals exposed to 85% O2 (38-41). The oxygen exposure for both 7 and 14 d was found to increase the 15 for type II cells by 21% while increasing the value for type I cell nuclei by only 7 %. The values used in this study (given in micrometers) were 7.97 (control) and 8.50 (02 exposed) for the alveolar type I cell and 6.92 (control) and 8.36 (02 exposed) for the alveolar type II cell. The volume densities that each subcellular component occupied per cell of the type I and of type II epithelium within each treated group were determined by dividing the number of points falling on each subcellular structure by the total number of points falling on the particular tissue type within the specific treated group. The absolute volumes that each subcellular structure occupied per cell were calculated by multiplying the volume density of each cell substructure by the average cell volume corresponding to the particular tissue type and treated group. Statistical Analyses Duncan's multiple comparison test (42) was used to evaluate both significant changes occurring between treated groups and the control group, and between the treatment groups themselves. All tests were two-sided, and P < 0.05 was considered to be significant.

Results Morphometric Assessment of Mean Epithelial Cell Characteristics The total volumes, total numbers, average volumes, and average surface areas of both alveolar type I and type II cells are given in Table 1, in addition to final body weights and fixed lung volumes. The results for the 12 animals included in this study are similar to those previously reported for similar animals and exposures (10). The normal rat lung contains 70 to 80 million alveolar type I cells and their average volume (2,240 1Lm3) and surface area characteristics do not change substantially after hyperoxic exposure, which is consistent with previous reports that have concluded that these cells are not highly sensitive to hyperoxic injury (4, 10-12). The alveolar type II epithelial cell is substantially smaller (about 500 1Lm3) and undergoes substantial hyperplasia and hypertrophy during hyperoxic exposure as previously reported (10, 11). Morphometric Assessment of Subcellular Structure in Normal Epithelial Cells The distribution of major subcellular constituents within the normal type I and type II alveolar epithelium is shown in the first column of Table 2. The absolute volumes per cell of these subcellular compartments are shown in the first column

118

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 41991

TABLE I

Mean cellular characteristics of the alveolar epithelium* Control

Characteristic

n Body weight (g) Fixed lung volume (ern') Alveolar type I cells Total volume (cmt/lung) Total number (X 106 /lung) Average volume (JLm 3/cell) Average surface area (JLm 2/cell) Air border Basement member Alveolar type II cells Total volume (cmvlung) Total number (X 106/lung) Average volume (JLm 3/cell) Average surface area (JLm 2/cell) Total air border Air border (without microvilli)

* All data are mean ± SEM. < 0.05 compared to control. * < 0.05 for comparison to 7 d of 85% t p P

Seven Days of 85 % O2

Fourteen Days of 85 % 0,

4 360.0 ± 5.0 10.40 ± 0.70

293.8 ± 1lAt 11.28 ± 0.88

4 291.2 ± 9.2t 11.52 ± 0.25

0.161 ± 0.013 79 ± 13 2,200 ± 39

0.200 ± 0.034 73 ± 5 2,740 ± 400

0.146 ± 0.024 65 ± 5 2,220 ± 230

7,070 ± 1,230 6,170 ± 1,080

7,340 ± 313 6,260 ± 300

7,130 ± 660 6,340 ± 690

0.057 ± 0.014 113 ± 8 500 ± 105

0.226 ± 0.062t 228 ± 15t 975 ± 231

0.128 ± 0.Q25 173 ± 13U 750 ± 140

4

231 ± 58 166 ± 75

334 ± 53 229 ± 66

319 ± 73 225 ± 42

0,.

TABLE 2

Effects of hyperoxia on the volume density of cell substructures* Organelles

Alveolar type I epithelium Nucleus Cytosols Vesicles' Mitochondria Rough ER Polyribosomes Smooth ER Golgi apparatus Multivesicular bodies Lysosomes/peroxisomes Other** Alveolar type II epithelium Nucleus Cytosol! Lamellar bodies Vesicles' Mitochondria Rough ER Polyribosomes Smooth ER Golgi apparatus Multivesicular bodies Lysosomes/peroxisomes Other**

Control

8.57 74.60 8.04 2.36 3.39 1.47 1.19 0.10 0.05 0.01 0.22

± ± ± ± ± ± ± ± ± ± ±

Seven Days of 85 % 0,

Fourteen Days of 85% 0,

0.68 1.01 1.22 0.27 0.20 0.16 0.25 0.07 0.02 0.01 0.06

12.02 67.66 7.48 4.04 4.38 1.36 1.94 0.50 0.22 0.10 0.29

± ± ± ± ± ± ± ± ± ± ±

3.78 3.32 0.55 0.40t 0.59 0.10 0.20 O.13t 0.03t O.04t 0.Q7

10.91 67.69 7.54 5.09 4.89 1.51 1.64 0.36 0.12 0.03 0.21

± ± ± ± ± ± ± ± ± ± ±

0.94 1.80 0.77 0.25U 0.76 0.14 0.25 0.12 0.03* 0.01 0.Q7

23.23 ± 2.41 41.83 ± 1.94 11.55±1.oo 0.17 ± 0.09 8.07 ± 0.43 8.69 ± 0.64 2.44 ± 0.13 1.41 ± 0.29 1.54 ± 0.16 0.92 ± 0.10 0.02 ± 0.02 0.13 ± 0.03

20.02 43.20 12.05 0.98 7.58 8.74 2.75 2.05 1.54 0.86 0.04 0.19

± ± ± ± ± ± ± ± ± ± ± ±

1.23 1.33 1.20 0.13t 0.53 0.59 0.14 0.37 0.18 0.13 0.01 0.06

28.60 38.98 7.10 0.98 10.65 7.18 2.25 1.73 1.72 0.66 0.04 0.12

± ± ± ± ± ± ± ± ± ± ± ±

0.67U 1.00 0.59u 0.14t 0.82U 0.54 0.37 0.39 0.20 0.Q7 0.02 0.04

Abbreviation: ER = endoplasmic reticulum. * All data are expressed as a percentage of total cell volume (mean ± SEM; n = 4). t P < 0.05 compared to control. P < 0.05 for comparison to 7 d of 85% 0,. § Cytosol includes all components of the cytoplasm exclusive of membrane-bound structures plus endoplasmic reticulum and ribosomes. , Vesicles include pinocytotic vesicles and small vesicular structures within the cytoplasm. ** Other includes microtubules, fibrils, occasional portions of tissue hidden by artifact of preparation, and rare, nonidentifiable structures.

*

Harris, Chang, and Crapo: Hyperoxic Changes in Type I Alveolar Epithelium

119

TABLE 3

Effects of hyperoxia on the mean volume of cell substructures* Organelles

Alveolar type I epithelium Nucleus Cytosol! Vesicles' Mitochondria Rough ER Polyribosomes Smooth ER Golgi apparatus Multivesicular bodies Lysosomes/peroxisomes Other** Alveolar type II epithelium Nucleus Cytosol! Lamellar bodies Vesicles' Mitochondria Rough ER Polyribosomes Smooth ER Goigi apparatus Multivesicular bodies Lysosomes/peroxisomes Other**

Control

Fourteen Days of 85 % O2

15.00 22.14 26.72 5.88 4.47 3.58 5.53 1.52 0.50 0.26 1.28

329.89 1,856.59 205.33 110.78 120.12 87.37 53.23 13.83 5.94 2.81 8.09

± 103.80 ± 91.03t ± 14.96 ± 1O.85t ± 16.13t ± 2.67 ± 5.53t ± 3.64t ± 0.81t ± 1.13t ± 1.96

242.04 1,501.21 167.32 112.87 108.46 33.47 36.43 7.89 2.61 0.67 4.72

116.16 ± 12.05 209.14 ± 9.71 57.76 ± 5.00 0.86 ± 0.44 40.33 ± 2.16 43.46 ± 3.22 12.18 ± 0.66 7.05 ± 1.43 7.82 ± 0.79 4.63 ± 0.48 0.08 ± 0.08 0.67 ± 0.17

195.08 921.02 117.47 9.51 73.88 85.16 26.81 19.93 15.02 8.41 0.36 1.89

± 12.02t ± 12.95t ± 11.71t ± 1.31t ± 5.20t ± 5.78t ± 1.39t ± 3.61t ± 1.79t ± 1.28t ± 0.13 ± 0.56

219.64 292.57 53.32 7.34 79.91 53.90 16.87 13.00 12.94 4.93 0.31 0.86

188.59 1,640.98 176.86 51.90 76.64 32.24 26.16 2.20 1.20 0.21 4.71

± ± ± ± ± ± ± ± ± ± ±

Seven Days of 85 % O2

± ± ± ± ± ± ± ± ± ± ±

20.94 39.80* 17.01 5.44t 16.78 3.21 5.61 2.66 0.62* 0.23 1.60

± 5.00t ± 7.48U ± 4.40* ± 1.03t ± 6.12t ± 4.06* ± 2.74* ± 2.90 ± 1.50t ± 0.55* ± 0.13 ± 0.30

Abbreviation: ER = endoplasmic reticulum. * All data are average organelle volumes per cell (I'm') (mean ± SEM; n = 4). t P < 0.05 compared to control. P < 0.05 for comparison to 7 d of 85% O2 • § Cytosol includes all components of the cytoplasm exclusive of membrane-bound structures plus endoplasmic reticulum and ribosomes. , Vesicles include pinocytotic vesicles and small vesicular structures within the cytoplasm. ** Other includes microtubules, fibrils, occasional portions of tissue hidden by artifact of preparation, and rare, nonidentifiable structures.

*

of Table 3. When a comparison is made of the total organelle volumes per cell between the type I and type II epithelium, there were statistically significant differences in every category. Some of the major differences are graphically illustrated in Figure 1. While alveolar type II cells are generally thought of as highly metabolically active compared to type I cells, this concept does not take into account the massive total volume of the alveolar type I cell. The concept of relative metabolic activity when assessed by ultrastructure characteristics is usually based on the presence of high volume densities of cell substructure that contribute to cell metabolism. Table 1 shows that type II cells have a much higher volume density of mitochondria, rough and smooth endoplasmic reticulum, free polyribosomes, and Golgi apparatus than do type I cells. Despite this, the normal type I cell actually contains 29 % more total mitochondria, 72 % more total rough endoplasmic reticulum, 165% more total free polyribosomes, and 270% more total smooth endoplasmic reticulum (Figure 1). When looking at absolute volumes per cell, the type II cell, as expected, has more of the components that relate to secretory function: Golgi apparatus, lamellar bodies, and multivesicular bodies (Figure 1). Since a single type II cell covers only 1/37th of the alveolar surface covered by a type I cell and contains 78 % less

volume, its subcellular organelles are more tightly packed. If the metabolic activity of these cells is considered relative to their alveolar surface area, the type II cell would have a dramatically higher relative metabolic activity as indicated by its greater volume/alveolar surface density ratio of mitochondria, rough endoplasmic reticulum, free polyribosomes, smooth endoplasmic reticulum, and Golgi apparatus (Figure 2). Morphometric Assessment of Cell Substructure in Hyperoxic-exposed Epithelial Cells The effects of hyperoxia on organelle volume densities and absolute volumes are shown in Tables 2 and 3, respectively. The comparison of change in cellular volume of lysosomes and peroxisomes among the two cell types is graphically represented in Figure 3. Further, the percent change of six selected organelles in type I pneumocytes (mitochondria, rough endoplasmic reticulum, ribosomes, smooth endoplasmic reticulum, Golgi apparatus, and multivesicular bodies) in response to hyperoxic exposure is shown in Figure 4. Hyperoxic exposure caused a dramatic change in the subcellular composition of both type I and type II pneumocytes. With minor exceptions, there were increases in volume densities in every category of alveolar epithelial substructure at some point following hyperoxic exposure (Table 2). Changes

120

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

a.

4

7 •

o

T

Type II Type I

3

·p

Rat lung alveolar type I epithelial cell injury and response to hyperoxia.

Hyperoxia has been shown to cause extensive lung injury, which involves all components of the alveolar septum, although the type I epithelium has gene...
1MB Sizes 0 Downloads 0 Views