A Morphometric Examination of Type II Alveolar Epithelial Cells in Normal and IsolatedPerfused Dog Lungs1 DAVID 0. DEFOUW AND PETER B. BERENDSEN Department of Anatomy, New Jersey Medical School, 100 Bergen Street, Newark, New Jersey 07103
ABSTRACT The volume densities of type I1 alveolar cell cytoplasmic organelles and alveolar surface densities were estimated by established stereologic procedures. The morphometric measurements were obtained from normal dog lungs (in situ) and isolated dog lungs perfused for 30-minute, l-hour, and 2-hour periods. The type I1 cell lamellar body volume densities and the alveolar surface densities progressively decreased as the times of perfusion were increased. The volume densities of the granular and agranular endoplasmic reticulum progressively increased during the periods of perfusion. These morphometric parameters from lungs in situ and isolated lungs suggest that changes occur in pulmonary surfactant synthesis and activity during perfusion. It is further postulated that progressive increases in the rates of surfactant removal and/or inactivation during perfusion may contribute to spontaneous edema in lungs isolated for periods exceeding two hours. The morphologic and physiologic integrity of isolated perfused lung preparations, widely used as models of lungs in vivo, in situ requires further evaluation. Autoradiographic and biochemical studies suggest that the lamellar bodies of type I1 alveolar epithelial cells contain pulmonary surfactant (Askin and Kuhn, '71; Chevalier and Collet, '72; Page-Roberts, '72). The recent analysis of isolated type I1 cells has provided more direct evidence that type I1 cells synthesize surface active material (Kikkawa et al., '75). The rate of surfactant synthesis is reportedly very rapid, as calculated from studies using radioactive precursors of dipalmitoyl lecithin (Tierney et al., '67). This suggests that the surface tension properties of the lungs depend on active metabolic processes within the type I1 cells t o insure sufficient quantities of surface active material. It was recently postulated that the demand for rapid synthesis and secretion of surfactant, necessary in species with rapid respiratory rates, was met by an increased number of type I1 cells and by an increased rate of synthesis and secretion by the individual type I1 cells of these species (Massaro and Massaro, '75). These observations led t o the present study in which ultrastructural changes in the type AM. J. ANAT., 150: 139-148.
I1 cells of dog lungs, under varying conditions of pulmonary vascular perfusion, were examined. Specifically, the volume densities of the type I1 cell cytoplasmic organelles and alveolar surface densities were determined by established stereological procedures in normal and in isolated perfused dog lungs. MATERIALS AND METHODS
Mongrel dogs of either sex, weighing approximately 20 kg, were anesthetized with pentobarbital (30 mg/kg, i.v.1, heparinized (25,000) IU),artificially ventilated with 5% CO, in humidified air on a Harvard respirator (12 strokedmin a t 400 ml/stroke) and given the antihistamine, diphenylhydramine hydrocholoride (35 mg). A femoral artery was cannulated and 0.5-1.5 liters of blood was withdrawn while a comparable volume of 5% bovine serum albumin in 0.15 M NaHCO, was infused through a femoral vein. The withdrawn blood was centrifuged and the supernaAccepted April 4, '77. This investigation was supported in part by Public Health Service Grants HL-19571 and HL-12879.
'
139
140
DAVID 0. DEFOUW AND PETER B. BERENDSEN
tant plasma was used as perfusate. Papaverine (180 mg) was added to the perfusate in hopes of producing a uniformly vasodilated lung. The dogs were killed with an overdose of pentobarbital (2 g), the thorax opened, and the heart and lungs were removed. The lungs were isolated from the heart and the pulmonary trunk connected by 3J8-inch tubing to a gravity-feed, upper reservoir containing the perfusate. The outflow connection was via 3l.8inch Tygon tubing inserted through the mitral valve and sewn in place by a pursestring suture. The outflow connection led to a lower reservoir from which the perfusate waB roller-pumped back to the upper reservoir. The flow rate was maintained at 10 ml/sec and the perfusate temperature was controlled (37°C) by a heater coil and thermostat in the upper reservoir. The lungs were placed on a platform which was suspended from a Statham load cell connected to a Honeywell Visicorder to monitor changes in lung weight. Arterial inflow and venous outflow pressures were also monitored by Statham gauges connected to the recorder. The lungs were continuously ventilated throughout the periods of isolation and perfusion. These preparations are similar to those used for studies of endothelial reflection and permeability coefficients (Per1 et al., '75). The isolated perfused lung preparations were divided into three groups of 30 minutes, 1 hour, and 2 hours of perfusion. Those lungs that demonstrated a weight gain greater than 2-3%were not considered stable preparations and were excluded from the sample. A second group of animals which was heparinized, injected with diphenylhydramine hydrochloride, and killed with an overdose of pentobarbital, served as the control preparations. In these preparations, the thorax was opened, the thoracic contents removed, and the lungs isolated from the heart. The trachea was cannulated and instilled a t a pressure of 20 cm H 2 0 with cold 2% glutaraldehyde in 0.15 M NaHCO, buffer a t pH 7.4. The lungs from the three experimental groups were fixed, like those of the control group, for electron microscopy. After the initial tracheal instillation of glutaraldehyde, the lungs remained in cold fixative for 24 to 48 hours and were then cut into 1 2 longitudinal slices. Two tissue blocks were taken from each slice according to a stratified sampling procedure (Weibel, '73). The primary sample tis-
sue blocks were then postfixed in 1%osmium tetroxide, dehydrated in graded ethanols and propylene oxide, and embedded in Epon. From the primary sample of 24 blocks, 10 blocks were selected randomly and thin sections (6090 nm) from each block were mounted on 200mesh copper grids and stained with uranyl acetate and lead citrate. Ten electron photomicrographs a t X 21,600 were obtained randomly from the upper left hand corners of the mesh squares of each copper grid. Pointcounting volumetry, using a test system of 84 test lines and 168 test points, was used to estimate volume densities of the type I1 cytoplasmic organelles. In addition, line-intercept counts were recorded to estimate the pulmonary alveolar surface densities. The stratified sampling procedure provided 100 micrographs per lung for the stereologic analyses. This sample size was sufficient to yield an expected relative error in estimating the stereologic parameters of less than 5% (Weibel, '73). The statistical significance of the differences between the means of the lungs in situ and the isolated perfused lung samples was evaluated by a two-tailed t-test (Freund, '67). RESULTS
Volume densities of cytoplasmic organelles are defined as estimates of the percentages of type I1 cell cytoplasmic volume occupied by the respective organelles. Table 1 presents the volume densities of lamellar bodies, mitochondria, and granular and agranular endoplasmic reticulum from type I1 cells of the control and isolated perfused lungs. The lamellar body volume densities progressively decreased as the time of isolation continued from 30 minutes to 2 hours. The difference between the volume density in situ and the values after 1 hour and after 2 hours of isolation were statistically significant (P < 0.05). The suggestion that continuous depletion of lamellar body contents occurs during the periods of perfusion in isolated lungs is consistent with these findings. The mitochondria1 volume densities were initially unchanged but after 2 hours of isolated perfusion a slight increase in volume density was noted. The granular endoplasmic reticulum volume densities demonstrated a slight but continuous rate of increase as the periods of isolation were increased. Likewise, the combined volume densities of the agranular endoplasmic reticulum and Golgi complex-
141
TYPE I1 CELLS IN NORMAL AND ISOLATED DOG LUNGS TABLE 1
Volume densities of WDe IIcell cvtoDlasmk oreanelles from the control and isolated Derfused lunes
VLB '
VM '
33.120.05 22.6k0.08 14.1k0.04 11.520.08
9.420.01 8 . 3 20.01 9.420.02 13.6k0.05
Preparation
Control in situ 1/2-hr isolation 1-hrisolation 2-hr isolation
Vnm'
7.620.03 10.1~0.01 12.250.01 13.320.05
VSER
2.950.01 5.750.02 7.220.02 9.350.07
' Values represent mean & one standard error. 'Lamellar body volume density. Mitochondria1 volume density. 'Granular endoplasmic reticulum volume density. Agranular endoplasmic reticulum (including Golgi complex) volume density.
es were progressively increased during the periods of isolation. The differences between the volume densities in situ and the volume densities after two hours of isolated perfusion was statistically significant (P < 0.05). The progressive increases in the volume densities suggest continuously rising rates of synthetic activity within these cytoplasmic organelles. The progressive changes in volume densities of the cytoplasmic organelles as the periods of perfusion were increased are illustrated in figures 1-4.Each photomicrograph of a single field was chosen to represent the large number of fields that were randomly sampled for the stereologic analyses. The morphologic features of each micrograph illustrate the particular changes that occurred during the varying periods of perfusion as determined by statistical analysis of the stereologic data. Alveolar surface densities are defined as estimates of alveolar surface area present within each milliliter of lung volume. Table 2 presents the alveolar surface densities in the control and isolated perfused lungs. A progressive decrease in alveolar surface density is revealed as the periods of isolation proceed from 30 minutes to 2 hours. The differences between the surface density in situ and that after two hours of isolation is statistically significant (P < 0.05) suggesting progressively increasing rates of removal or inactivation of surfactant. TABLE 2 Alveolar surface densities in the control and isolated perfused lungs' ~
~~
PreDaration
Surface densities (cm2/ml lune)
Control in situ 1/2-hourisolation 1-hour isolation 2-hour isolation 1
Values represent mean
%
566.2%31.6 536.75 31.3 464.7k36.7 345.0266.8 one standard error.
DISCUSSION
It has been reported that lamellar bodies occupied similar volume densities in both small (mouse) and large (dog) species (Massaro and Massaro, '75). Mitochondrial volume densities, however, were greater in the smaller species (Massaro et al., '75). This was attributed to their more rapid respiratory rates. Previously, Faridy et al. ('66) had reported that rapid respiratory rates induce a more rapid surfactant turnover which suggests a need for greater surfactant synthesis by the type I1 cells. Therefore, the increased mitochondrial volume density was attributed to the higher energy production in type I1 cells to accommodate the suggested increases in surfactant synthesis and secretion (Massaro et al., '75). McClenahan and Urtnowski ('67) also reported a loss of functional surfactant by rapid ventilation with large tidal volumes in isolated dog lungs. The present comparisons of lungs in situ with 30-minute, 1-hour and 2hour periods of isolation and perfusion describe a progressive decrease in lamellar body volume density with increased periods of isolation. Concomitant increases in both granular and agranular (including the Golgi complexes) endoplasmic reticulum provide morphologic evidence that surfactant synthesis progressively increases as the period of isolation and perfusion is increased. Note that after two hours of isolation the combined volume densities of the agranular endoplasmic reticulum and Golgi complex have tripled, the granular endoplasmic reticulum volume density has doubled, while the lamellar body volume density has decreased to one third of its value in the lungs in situ. Previously, Chevalier and Collet ('72) described the endoplasmic reticulum and Golgi complex as the sites of surfactant synthesis in type I1 cells. The mitochondrial volume densities estimated in this study are similar to those previ-
142
DAVID 0.DEFOUW AND PETER B. BERENDSEN
ously estimated in the dog lung (Massaro et al., '75). This suggests that appropriate levels of energy production were provided to meet the demands of increased surfactant synthesis. An alternate explanation for the changes in organelle volume densities during the varying periods of perfusion must also be considered. Substantial increases in the volume densities of granular and agranular endoplasmic reticulum after two hours of perfusion could result from an overall decrease in cytoplasmic volumes. That is, lamellar body volume densities were significantly reduced after the 2hour perfusion period. Thus, as the contents of the lamellar bodies are lost, the granular and agranular endoplasmic reticulum would appear to occupy a larger relative proportion of the cytoplasmic volume. Therefore, additional biochemical evidence is required before a progressive increase in surfactant production during the varying periods of perfusion can be substantiated. If it is assumed that lamellar bodies contain only material destined to be secreted onto the alveolar surface, then the finding of diminished lamellar body volume densities in the isolated lungs suggests a rapid rate of surfactant secretion. However, the progressive decrease in the alveolar surface densities with increased periods of isolation suggests inactivation or rapid removal of surfactant from the alveolar surface in the isolated perfused lungs. Previous reports of decreases in alveolar surface area proportional to decreases in lung compliance (Tierney et al., '67)have also suggested that diminution of alveolar surfactant leads to diminished alveolar surface areas. The mechanisms responsible for surfactant inactivation or removal in the present isolated-perfused lung preparations remains uncertain. The respiratory rates and volumes approximated normal canine lung values; therefore, a removal of surfactant via excessive ventilation is not suggested as a contributing factor. Finley et al. ('64) reported that two hours of pulmonary arterial occlusion were necessary before minimal surface tension values appeared slightly elevated. Weber and Visscher ('69) reported normal metabolic activity in plasma-perfused isolated dog lungs. Thus, the presence and utilization of phospholipid, protein, and polysaccharide precursors of surfactant by type I1 cells would also appear sufficient in the present investiga-
tion. However, the concentrations of these constituents in the perfusates were not recorded. Gil and Weibel ('69) described a duplex lining layer of the alveoli after perfusionfixation of dog lungs. The thickness of this layer was reported to be approximately 50 nm. Thus, it added only 3%to the total thickness of the air-blood barrier. The lining layer, which represented a n ultrastructural visualization of pulmonary surfactant, was not preserved after instillation of fixative into the airways. Therefore, the means of fixation used in the present study precluded a morphometric assessment of changes in surfactant layer thickness during the periods of isolation. Currently, stereologic analyses of perfusion-fixed lungs have been initiated t o examine changes in surfactant layer thickness as a further means of evaluating the progressive loss of surfactant as the periods of perfusion are increased. Nicolaysen ('71) attributed spontaneous edema in isolated lungs perfused for periods of two to six hours to gradual increases in capillary permeability. However, ultrastructural evidence of changes in pulmonary capillary endothelial cells was not presented. More recently filtration coefficients were found to increase and the reflection coefficients for sodium were found to decrease as isolated perfused lobes, similarly handled, gained weight, i.e., became edematous (Chinard, Ritter, Chowdhury, personal communication). The present results suggest that the progressive decreases in alveolar surface density may be interpreted as progressive elevations in surface tension as surfactant is inactivated or removed. Progressive elevations in surface tension would thus contribute to the onset of edema. Subjective observations failed to detect obvious ultrastructural evidence of pulmonary edema in the 30-minute, l-hour or 2hour periods of isolation. In general, attempts to extend the periods of perfused isolation were regularly met by the development of alveolar flooding with edema fluid. A morphometric examination of the alveolar septa during each of the isolation periods is needed to detect progressive changes in the septa1 compartments that may arise as the periods of isolation are increased. Progressive changes in the metabolic activity of type I1 cells in the isolated-perfused lungs must be considered in morphologic and
TYPE I1 CELLS IN NORMAL AND ISOLATED DOG LUNGS
physiologic evaluations of isolated systems. The present morphometric results from isolated-perfused dog lungs indicate that the production and secretion of surface active materials by the alveolar type I1 cells occur a t fluctuating rates. These various rates of synthesis and secretion are also associated with apparent alterations in the rate of surfactant inactivation or removal. Evaluations of the correlation between surfactant secretion and its inactivation are necessary to further define the role of surfactant in the maintenance of alveolocapillary membrane integrity. ACKNOWLEDGMENTS
The authors are grateful to Doctor F. P. Chinard and the late Doctor W. Perl for their suggestions and assistance, and to Doctor P. Chowdhury for his technical assistance. LITERATURE CITED Askin, F. B., and C. Kuhn 1971 The cellular origin of pulmonary surfactant. Lab. Invest., 25: 260-268. Chevalier, G., and A. J. Collet 1972 In vivo incorporation of ~holine-~H l e,~ c i n e - ~ H and , g a l a ~ t o s e - ~in H alveolar type I1 pneumocytes in relation to surfactant synthesis. A quantitative radioautographic study in mouse by electron microscopy. Anat. Rec., 174: 289-310. Faridy, E. E., S. Permutt and R. L. Riley 1966 Effect of ventilation on surface forces in excised dog lungs. J. Appl. Physiol., 21: 1453-1462. Finley, T. N., W. H. Tooley, E. W. Swenson, R. E. Gardner and J. A. Clements 1964 Pulmonary surface tension
143
i n experimental atelectasis. Am. Rev. Res. Dis., 89: 373~~. -378. Freund, J. E. 1967 Modern Elementary Statistics. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Gil, J.,and E. R. Weibel 1969 Improvements in demonstration of the lining layer of lung alveoli by electron microscopy. Resp. Physiol., 8: 13-36. Kikkawa, Y., K. Yoneda, F. Smith, B. Packard and K. Suzuki 1975 The type I1 epithelial cells of the lung. 11. Chemical composition and phospholipid synthesis. Lab. Invest., 32: 295-302. Massaro, G. D., D. B. Gail and D. Massaro 1975 Lung oxygen consumption and mitochondria of alveolar epithelial and endothelial cells. J. Appl. Physiol., 38: 588-592. Massaro, G. D., and D. Massaro 1975 Stereologic evaluation of granular pneumocyte lamellar bodies in different Exp. Biol. Med., 149: 333-335. species. Proc. SOC. McClenahan, J. B., and A. Urtnowski 1967 Effect of ventilation on surfactant, and its turnover rate. J. Appl. Physiol., 23: 215-220. Nicolaysen, G. 1971 Perfusate qualities and spontaneous edema formation in a n isolated perfused lung preparation. Acta Physiol. Scand., 83: 563-570. Page-Roberts, B. A. 1972 Preparation and partial characterization of a lamellar body fraction from r a t lung. Biochem. Biophys. Acta, 260: 334-338. Perl, W. P.. P. Chowdhury and F. P. Chinard 1975 Reflection coefficients of dog endothelium to small hydrophilic solutes. Am. J. Physiol., 228: 797-809. Tierney, D.F.,J. A. Clements and H. J. Trahan 1967 Rates of replacement of lecithins and alveolar instability in rat lungs. Am. J. Physiol., 213: 671-676. Weber, K. C., and M. B. Visscher 1969 Metabolism of the isolated canine lung. Am. J. Path., 21 7: 1044-1052. Weibel, E. R. 1973 Stereologic techniques for rnicroscopic morphometry. In: Principle and Techniques of Electron Microscopy. Vol. 3. M. A. Hyatt, ed. Von Nostrand Reinhold, New York, pp. 237-296. ~
PLATE 1 EXPLANATION OF FIGURES
1 A type I1 cell from a dog lung in situ. The cytoplasm is characterized by prominent lamellar bodies, mitochondria and endoplasmic reticulum. This appearance suggests a metabolically active cell. X 21,600.
2 A type I1 cell from an isolated lung perfused for 30 minutes. The mitochondria and endoplasmic reticulum are obvious while the lamellar bodies are now less prominent. Note the lamellar body at the cell's free surface that appears to be secreting its contents. X 21,600.
144
TYPE II CELLS IN NORMAL AND ISOLATED DOG LUNGS David 0. DeFouw and Peter B. Berendsen
PLATE 1
145
PLATE 2 EXPLANATION OF FIGURES
3 A type I1 cell from a n isolated lung after a 1-hour period of perfusion. Note the prominent endoplasmic reticulum and cluster of mitochondria in the perinuclear cytoplasm. The lamellar body volume densities are now obviously diminished. The integrities of t h e endothelial membrane and interstitial space appear intact. X 21,600. 4
146
A type I1 cell from a lung isolated and perfused for a 2-hour period. The paucity of lamellar bodies is obvious. The endoplasmic reticulum and mitochondria remain prominent, suggesting continuous synthetic activity within the type I1 cells. X 21,600.
TYPE I1 CELLS IN NORMAL AND ISOLATED DOG LUNGS David 0. DeFouw and Peter B. Berendsen
PLATE 2
147
ABSTRACT The volume densities of type I1 alveolar cell cytoplasmic organelles and alveolar surface densities were estimated by established stereologic procedures. The morphometric measurements were obtained from normal dog lungs (in situ) and isolated dog lungs perfused for 30-minute, l-hour, and 2-hour periods. The type I1 cell lamellar body volume densities and the alveolar surface densities progressively decreased as the times of perfusion were increased. The volume densities of the granular and agranular endoplasmic reticulum progressively increased during the periods of perfusion. These morphometric parameters from lungs in situ and isolated lungs suggest that changes occur in pulmonary surfactant synthesis and activity during perfusion. It is further postulated that progressive increases in the rates of surfactant removal and/or inactivation during perfusion may contribute to spontaneous edema in lungs isolated for periods exceeding two hours. The morphologic and physiologic integrity of isolated perfused lung preparations, widely used as models of lungs in vivo, in situ requires further evaluation. Autoradiographic and biochemical studies suggest that the lamellar bodies of type I1 alveolar epithelial cells contain pulmonary surfactant (Askin and Kuhn, '71; Chevalier and Collet, '72; Page-Roberts, '72). The recent analysis of isolated type I1 cells has provided more direct evidence that type I1 cells synthesize surface active material (Kikkawa et al., '75). The rate of surfactant synthesis is reportedly very rapid, as calculated from studies using radioactive precursors of dipalmitoyl lecithin (Tierney et al., '67). This suggests that the surface tension properties of the lungs depend on active metabolic processes within the type I1 cells t o insure sufficient quantities of surface active material. It was recently postulated that the demand for rapid synthesis and secretion of surfactant, necessary in species with rapid respiratory rates, was met by an increased number of type I1 cells and by an increased rate of synthesis and secretion by the individual type I1 cells of these species (Massaro and Massaro, '75). These observations led t o the present study in which ultrastructural changes in the type AM. J. ANAT., 150: 139-148.
I1 cells of dog lungs, under varying conditions of pulmonary vascular perfusion, were examined. Specifically, the volume densities of the type I1 cell cytoplasmic organelles and alveolar surface densities were determined by established stereological procedures in normal and in isolated perfused dog lungs. MATERIALS AND METHODS
Mongrel dogs of either sex, weighing approximately 20 kg, were anesthetized with pentobarbital (30 mg/kg, i.v.1, heparinized (25,000) IU),artificially ventilated with 5% CO, in humidified air on a Harvard respirator (12 strokedmin a t 400 ml/stroke) and given the antihistamine, diphenylhydramine hydrocholoride (35 mg). A femoral artery was cannulated and 0.5-1.5 liters of blood was withdrawn while a comparable volume of 5% bovine serum albumin in 0.15 M NaHCO, was infused through a femoral vein. The withdrawn blood was centrifuged and the supernaAccepted April 4, '77. This investigation was supported in part by Public Health Service Grants HL-19571 and HL-12879.
'
139
140
DAVID 0. DEFOUW AND PETER B. BERENDSEN
tant plasma was used as perfusate. Papaverine (180 mg) was added to the perfusate in hopes of producing a uniformly vasodilated lung. The dogs were killed with an overdose of pentobarbital (2 g), the thorax opened, and the heart and lungs were removed. The lungs were isolated from the heart and the pulmonary trunk connected by 3J8-inch tubing to a gravity-feed, upper reservoir containing the perfusate. The outflow connection was via 3l.8inch Tygon tubing inserted through the mitral valve and sewn in place by a pursestring suture. The outflow connection led to a lower reservoir from which the perfusate waB roller-pumped back to the upper reservoir. The flow rate was maintained at 10 ml/sec and the perfusate temperature was controlled (37°C) by a heater coil and thermostat in the upper reservoir. The lungs were placed on a platform which was suspended from a Statham load cell connected to a Honeywell Visicorder to monitor changes in lung weight. Arterial inflow and venous outflow pressures were also monitored by Statham gauges connected to the recorder. The lungs were continuously ventilated throughout the periods of isolation and perfusion. These preparations are similar to those used for studies of endothelial reflection and permeability coefficients (Per1 et al., '75). The isolated perfused lung preparations were divided into three groups of 30 minutes, 1 hour, and 2 hours of perfusion. Those lungs that demonstrated a weight gain greater than 2-3%were not considered stable preparations and were excluded from the sample. A second group of animals which was heparinized, injected with diphenylhydramine hydrochloride, and killed with an overdose of pentobarbital, served as the control preparations. In these preparations, the thorax was opened, the thoracic contents removed, and the lungs isolated from the heart. The trachea was cannulated and instilled a t a pressure of 20 cm H 2 0 with cold 2% glutaraldehyde in 0.15 M NaHCO, buffer a t pH 7.4. The lungs from the three experimental groups were fixed, like those of the control group, for electron microscopy. After the initial tracheal instillation of glutaraldehyde, the lungs remained in cold fixative for 24 to 48 hours and were then cut into 1 2 longitudinal slices. Two tissue blocks were taken from each slice according to a stratified sampling procedure (Weibel, '73). The primary sample tis-
sue blocks were then postfixed in 1%osmium tetroxide, dehydrated in graded ethanols and propylene oxide, and embedded in Epon. From the primary sample of 24 blocks, 10 blocks were selected randomly and thin sections (6090 nm) from each block were mounted on 200mesh copper grids and stained with uranyl acetate and lead citrate. Ten electron photomicrographs a t X 21,600 were obtained randomly from the upper left hand corners of the mesh squares of each copper grid. Pointcounting volumetry, using a test system of 84 test lines and 168 test points, was used to estimate volume densities of the type I1 cytoplasmic organelles. In addition, line-intercept counts were recorded to estimate the pulmonary alveolar surface densities. The stratified sampling procedure provided 100 micrographs per lung for the stereologic analyses. This sample size was sufficient to yield an expected relative error in estimating the stereologic parameters of less than 5% (Weibel, '73). The statistical significance of the differences between the means of the lungs in situ and the isolated perfused lung samples was evaluated by a two-tailed t-test (Freund, '67). RESULTS
Volume densities of cytoplasmic organelles are defined as estimates of the percentages of type I1 cell cytoplasmic volume occupied by the respective organelles. Table 1 presents the volume densities of lamellar bodies, mitochondria, and granular and agranular endoplasmic reticulum from type I1 cells of the control and isolated perfused lungs. The lamellar body volume densities progressively decreased as the time of isolation continued from 30 minutes to 2 hours. The difference between the volume density in situ and the values after 1 hour and after 2 hours of isolation were statistically significant (P < 0.05). The suggestion that continuous depletion of lamellar body contents occurs during the periods of perfusion in isolated lungs is consistent with these findings. The mitochondria1 volume densities were initially unchanged but after 2 hours of isolated perfusion a slight increase in volume density was noted. The granular endoplasmic reticulum volume densities demonstrated a slight but continuous rate of increase as the periods of isolation were increased. Likewise, the combined volume densities of the agranular endoplasmic reticulum and Golgi complex-
141
TYPE I1 CELLS IN NORMAL AND ISOLATED DOG LUNGS TABLE 1
Volume densities of WDe IIcell cvtoDlasmk oreanelles from the control and isolated Derfused lunes
VLB '
VM '
33.120.05 22.6k0.08 14.1k0.04 11.520.08
9.420.01 8 . 3 20.01 9.420.02 13.6k0.05
Preparation
Control in situ 1/2-hr isolation 1-hrisolation 2-hr isolation
Vnm'
7.620.03 10.1~0.01 12.250.01 13.320.05
VSER
2.950.01 5.750.02 7.220.02 9.350.07
' Values represent mean & one standard error. 'Lamellar body volume density. Mitochondria1 volume density. 'Granular endoplasmic reticulum volume density. Agranular endoplasmic reticulum (including Golgi complex) volume density.
es were progressively increased during the periods of isolation. The differences between the volume densities in situ and the volume densities after two hours of isolated perfusion was statistically significant (P < 0.05). The progressive increases in the volume densities suggest continuously rising rates of synthetic activity within these cytoplasmic organelles. The progressive changes in volume densities of the cytoplasmic organelles as the periods of perfusion were increased are illustrated in figures 1-4.Each photomicrograph of a single field was chosen to represent the large number of fields that were randomly sampled for the stereologic analyses. The morphologic features of each micrograph illustrate the particular changes that occurred during the varying periods of perfusion as determined by statistical analysis of the stereologic data. Alveolar surface densities are defined as estimates of alveolar surface area present within each milliliter of lung volume. Table 2 presents the alveolar surface densities in the control and isolated perfused lungs. A progressive decrease in alveolar surface density is revealed as the periods of isolation proceed from 30 minutes to 2 hours. The differences between the surface density in situ and that after two hours of isolation is statistically significant (P < 0.05) suggesting progressively increasing rates of removal or inactivation of surfactant. TABLE 2 Alveolar surface densities in the control and isolated perfused lungs' ~
~~
PreDaration
Surface densities (cm2/ml lune)
Control in situ 1/2-hourisolation 1-hour isolation 2-hour isolation 1
Values represent mean
%
566.2%31.6 536.75 31.3 464.7k36.7 345.0266.8 one standard error.
DISCUSSION
It has been reported that lamellar bodies occupied similar volume densities in both small (mouse) and large (dog) species (Massaro and Massaro, '75). Mitochondrial volume densities, however, were greater in the smaller species (Massaro et al., '75). This was attributed to their more rapid respiratory rates. Previously, Faridy et al. ('66) had reported that rapid respiratory rates induce a more rapid surfactant turnover which suggests a need for greater surfactant synthesis by the type I1 cells. Therefore, the increased mitochondrial volume density was attributed to the higher energy production in type I1 cells to accommodate the suggested increases in surfactant synthesis and secretion (Massaro et al., '75). McClenahan and Urtnowski ('67) also reported a loss of functional surfactant by rapid ventilation with large tidal volumes in isolated dog lungs. The present comparisons of lungs in situ with 30-minute, 1-hour and 2hour periods of isolation and perfusion describe a progressive decrease in lamellar body volume density with increased periods of isolation. Concomitant increases in both granular and agranular (including the Golgi complexes) endoplasmic reticulum provide morphologic evidence that surfactant synthesis progressively increases as the period of isolation and perfusion is increased. Note that after two hours of isolation the combined volume densities of the agranular endoplasmic reticulum and Golgi complex have tripled, the granular endoplasmic reticulum volume density has doubled, while the lamellar body volume density has decreased to one third of its value in the lungs in situ. Previously, Chevalier and Collet ('72) described the endoplasmic reticulum and Golgi complex as the sites of surfactant synthesis in type I1 cells. The mitochondrial volume densities estimated in this study are similar to those previ-
142
DAVID 0.DEFOUW AND PETER B. BERENDSEN
ously estimated in the dog lung (Massaro et al., '75). This suggests that appropriate levels of energy production were provided to meet the demands of increased surfactant synthesis. An alternate explanation for the changes in organelle volume densities during the varying periods of perfusion must also be considered. Substantial increases in the volume densities of granular and agranular endoplasmic reticulum after two hours of perfusion could result from an overall decrease in cytoplasmic volumes. That is, lamellar body volume densities were significantly reduced after the 2hour perfusion period. Thus, as the contents of the lamellar bodies are lost, the granular and agranular endoplasmic reticulum would appear to occupy a larger relative proportion of the cytoplasmic volume. Therefore, additional biochemical evidence is required before a progressive increase in surfactant production during the varying periods of perfusion can be substantiated. If it is assumed that lamellar bodies contain only material destined to be secreted onto the alveolar surface, then the finding of diminished lamellar body volume densities in the isolated lungs suggests a rapid rate of surfactant secretion. However, the progressive decrease in the alveolar surface densities with increased periods of isolation suggests inactivation or rapid removal of surfactant from the alveolar surface in the isolated perfused lungs. Previous reports of decreases in alveolar surface area proportional to decreases in lung compliance (Tierney et al., '67)have also suggested that diminution of alveolar surfactant leads to diminished alveolar surface areas. The mechanisms responsible for surfactant inactivation or removal in the present isolated-perfused lung preparations remains uncertain. The respiratory rates and volumes approximated normal canine lung values; therefore, a removal of surfactant via excessive ventilation is not suggested as a contributing factor. Finley et al. ('64) reported that two hours of pulmonary arterial occlusion were necessary before minimal surface tension values appeared slightly elevated. Weber and Visscher ('69) reported normal metabolic activity in plasma-perfused isolated dog lungs. Thus, the presence and utilization of phospholipid, protein, and polysaccharide precursors of surfactant by type I1 cells would also appear sufficient in the present investiga-
tion. However, the concentrations of these constituents in the perfusates were not recorded. Gil and Weibel ('69) described a duplex lining layer of the alveoli after perfusionfixation of dog lungs. The thickness of this layer was reported to be approximately 50 nm. Thus, it added only 3%to the total thickness of the air-blood barrier. The lining layer, which represented a n ultrastructural visualization of pulmonary surfactant, was not preserved after instillation of fixative into the airways. Therefore, the means of fixation used in the present study precluded a morphometric assessment of changes in surfactant layer thickness during the periods of isolation. Currently, stereologic analyses of perfusion-fixed lungs have been initiated t o examine changes in surfactant layer thickness as a further means of evaluating the progressive loss of surfactant as the periods of perfusion are increased. Nicolaysen ('71) attributed spontaneous edema in isolated lungs perfused for periods of two to six hours to gradual increases in capillary permeability. However, ultrastructural evidence of changes in pulmonary capillary endothelial cells was not presented. More recently filtration coefficients were found to increase and the reflection coefficients for sodium were found to decrease as isolated perfused lobes, similarly handled, gained weight, i.e., became edematous (Chinard, Ritter, Chowdhury, personal communication). The present results suggest that the progressive decreases in alveolar surface density may be interpreted as progressive elevations in surface tension as surfactant is inactivated or removed. Progressive elevations in surface tension would thus contribute to the onset of edema. Subjective observations failed to detect obvious ultrastructural evidence of pulmonary edema in the 30-minute, l-hour or 2hour periods of isolation. In general, attempts to extend the periods of perfused isolation were regularly met by the development of alveolar flooding with edema fluid. A morphometric examination of the alveolar septa during each of the isolation periods is needed to detect progressive changes in the septa1 compartments that may arise as the periods of isolation are increased. Progressive changes in the metabolic activity of type I1 cells in the isolated-perfused lungs must be considered in morphologic and
TYPE I1 CELLS IN NORMAL AND ISOLATED DOG LUNGS
physiologic evaluations of isolated systems. The present morphometric results from isolated-perfused dog lungs indicate that the production and secretion of surface active materials by the alveolar type I1 cells occur a t fluctuating rates. These various rates of synthesis and secretion are also associated with apparent alterations in the rate of surfactant inactivation or removal. Evaluations of the correlation between surfactant secretion and its inactivation are necessary to further define the role of surfactant in the maintenance of alveolocapillary membrane integrity. ACKNOWLEDGMENTS
The authors are grateful to Doctor F. P. Chinard and the late Doctor W. Perl for their suggestions and assistance, and to Doctor P. Chowdhury for his technical assistance. LITERATURE CITED Askin, F. B., and C. Kuhn 1971 The cellular origin of pulmonary surfactant. Lab. Invest., 25: 260-268. Chevalier, G., and A. J. Collet 1972 In vivo incorporation of ~holine-~H l e,~ c i n e - ~ H and , g a l a ~ t o s e - ~in H alveolar type I1 pneumocytes in relation to surfactant synthesis. A quantitative radioautographic study in mouse by electron microscopy. Anat. Rec., 174: 289-310. Faridy, E. E., S. Permutt and R. L. Riley 1966 Effect of ventilation on surface forces in excised dog lungs. J. Appl. Physiol., 21: 1453-1462. Finley, T. N., W. H. Tooley, E. W. Swenson, R. E. Gardner and J. A. Clements 1964 Pulmonary surface tension
143
i n experimental atelectasis. Am. Rev. Res. Dis., 89: 373~~. -378. Freund, J. E. 1967 Modern Elementary Statistics. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Gil, J.,and E. R. Weibel 1969 Improvements in demonstration of the lining layer of lung alveoli by electron microscopy. Resp. Physiol., 8: 13-36. Kikkawa, Y., K. Yoneda, F. Smith, B. Packard and K. Suzuki 1975 The type I1 epithelial cells of the lung. 11. Chemical composition and phospholipid synthesis. Lab. Invest., 32: 295-302. Massaro, G. D., D. B. Gail and D. Massaro 1975 Lung oxygen consumption and mitochondria of alveolar epithelial and endothelial cells. J. Appl. Physiol., 38: 588-592. Massaro, G. D., and D. Massaro 1975 Stereologic evaluation of granular pneumocyte lamellar bodies in different Exp. Biol. Med., 149: 333-335. species. Proc. SOC. McClenahan, J. B., and A. Urtnowski 1967 Effect of ventilation on surfactant, and its turnover rate. J. Appl. Physiol., 23: 215-220. Nicolaysen, G. 1971 Perfusate qualities and spontaneous edema formation in a n isolated perfused lung preparation. Acta Physiol. Scand., 83: 563-570. Page-Roberts, B. A. 1972 Preparation and partial characterization of a lamellar body fraction from r a t lung. Biochem. Biophys. Acta, 260: 334-338. Perl, W. P.. P. Chowdhury and F. P. Chinard 1975 Reflection coefficients of dog endothelium to small hydrophilic solutes. Am. J. Physiol., 228: 797-809. Tierney, D.F.,J. A. Clements and H. J. Trahan 1967 Rates of replacement of lecithins and alveolar instability in rat lungs. Am. J. Physiol., 213: 671-676. Weber, K. C., and M. B. Visscher 1969 Metabolism of the isolated canine lung. Am. J. Path., 21 7: 1044-1052. Weibel, E. R. 1973 Stereologic techniques for rnicroscopic morphometry. In: Principle and Techniques of Electron Microscopy. Vol. 3. M. A. Hyatt, ed. Von Nostrand Reinhold, New York, pp. 237-296. ~
PLATE 1 EXPLANATION OF FIGURES
1 A type I1 cell from a dog lung in situ. The cytoplasm is characterized by prominent lamellar bodies, mitochondria and endoplasmic reticulum. This appearance suggests a metabolically active cell. X 21,600.
2 A type I1 cell from an isolated lung perfused for 30 minutes. The mitochondria and endoplasmic reticulum are obvious while the lamellar bodies are now less prominent. Note the lamellar body at the cell's free surface that appears to be secreting its contents. X 21,600.
144
TYPE II CELLS IN NORMAL AND ISOLATED DOG LUNGS David 0. DeFouw and Peter B. Berendsen
PLATE 1
145
PLATE 2 EXPLANATION OF FIGURES
3 A type I1 cell from a n isolated lung after a 1-hour period of perfusion. Note the prominent endoplasmic reticulum and cluster of mitochondria in the perinuclear cytoplasm. The lamellar body volume densities are now obviously diminished. The integrities of t h e endothelial membrane and interstitial space appear intact. X 21,600. 4
146
A type I1 cell from a lung isolated and perfused for a 2-hour period. The paucity of lamellar bodies is obvious. The endoplasmic reticulum and mitochondria remain prominent, suggesting continuous synthetic activity within the type I1 cells. X 21,600.
TYPE I1 CELLS IN NORMAL AND ISOLATED DOG LUNGS David 0. DeFouw and Peter B. Berendsen
PLATE 2
147
