been described in detail.1'2 Lambs delivered ... - Europe PMC

2 downloads 0 Views 7MB Size Report
tion of osmiophilic inclusion bodies in type II alveolar epithelial cells, ... Presented in part at Symposium on Altered Pulmonary Structure in Relation to. Function ...
THE ULTRASTRUCTURE OF THE LUNGS OF LAMBS THE RELATION OF OSMIOPHILIC INCLUSIONS AND ALVEOLAR LINING LAYER TO FETAL MATURATION AND EXPERIMENTALLY PRODUCED RESPIRATORY DISTRESS YUTAKA KIKKAWA, M.D.*; ETSURO K. MOTOYAMA, M.D., AND CHALEs D. COOK, M.D. From the Department of Pathology, Albert Einstein College of Medicine, New York, N.Y., and the Department of Pediatrics, Yale University School of Medicine, New Haven, Conn.

In previous reports, experimental respiratory distress in lambs has been described in detail.1'2 Lambs delivered prematurely developed respiratory distress and loss of the normal surface activity of their lung extracts; in slightly more mature lambs severe prenatal asphyxia was required before similar respiratory difficulty appeared. The lungs of full term animals were found to be resistant and did not exhibit features consistent with this type of respiratory distress. Since there is a strong association between prematurity and the respiratory distress syndrome in newborn human infants3'4 and a preand perinatal stress such as asphyxia has been considered an additional important predisposing factor,5'6 the experimental model in lambs was considered similar in a number of respects to the human disorder.2 This report presents studies with electron microscopy of the morphologic development of the lungs in normal fetal and newborn lambs and in those with experimentally induced respiratory distress, atelectasis and hyaline membranes. The time of appearance during fetal maturation of osmiophilic inclusion bodies in type II alveolar epithelial cells, the osmiophilic alveolar lining layer, and the normal surface activity of lung extracts were investigated together with the interrelations of these facets of lung development. MATERIAL AND METHODS Lung development and the experimental production of respiratory distress had been studied previously in I23 mixed breed lambs.1'2 The lungs from 69 of these were mined by electron microscopy. The lambs were divided into 3 groups. Supported by Grants No. AM-2967-C7, HD-OOO3-0I, HD-oo248-o5 and HD-oo989-ox from the National Institutes of Health, United States Public Health Service. Presented in part at Symposium on Altered Pulmonary Structure in Relation to Function, at the Sixty-second Annual Meeting of the American Association of Pathologists and Bacteriologists, Philadelphia, Pa., March 7, i965. Accepted for publication June iI, I965. * Fellow of New York Heart Association.

877

878

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

Group I. Twenty-seven fetal lambs with estimated gestational ages ranging from 55 days to full term (I4 7 to iSo days). Group II. Twenty-four lambs of gestational age above I25 days; these had breathed spontaneously for 30 minutes to 4 hours. Older animals breathed spontaneously, while younger animals required assistance of a respirator (an additional 6 spontaneously born newborn lambs were included in this group). Group III. Twelve animals of more than 130 days gestation; these were subjected to severe prenatal asphyxia for IY2 to 4 hours and were breathed for an additional i to 5X2 hours with or without the assistance of a respirator. Specimens for electron microscopy were obtained from the lower lobes immediately after the animals were sacrificed and were fixed in cold Veronal buffered i per cent osmium tetroxide for i5 minutes. The tissues were then degassed in a partial vacuum for a few minutes and fixation was continued for an additional 75 minutes at atmospheric pressure. Specimens were then dehydrated 5 minutes each in 95 and ioo per cent ethanol and placed in propylene oxide for 20 minutes at room temperature. Tissues were then infiltrated overnight with half and half propylene oxide and Epon mixture, transferred to ioo per cent Epon mixture and after I2 hours polymerized at 600 C. Sections were made with a Porter-Blum microtome, stained with uranyl acetate, lead citrate or both and were examined with an RCA-3C electron microscope using an accelerating voltage of ioo kv. The number of osmiophilic inclusions were counted in at least ioo type II alveolar epithelial cells (giant alveolar cells, granular pneumonocytes, special cells or alveolar macrophages) from each specimen and the average number of inclusion bodies per cell was compared in the different gestational age groups. The number of type II cells containing inclusion bodies per Ir,ooo or more epithelial, septal and endothelial cells was determined and expressed as per cent. Tissues for light microscopy were obtained from several lobes and were fixed in buffered io per cent formalin. Paraffin sections of these were stained with hematoxylin and eosin. Five to 7 gm of lung tissue from the upper and lower lobes were used for the measurement of surface tension using saline extracts with a modified Wilhelmy balance as described elsewhere.7

RESULTS

Development of Fetal Lung By light microscopy the lungs of fetal lambs at I20 days gestation appeared indistinguishable from those at full term. The terminal air sac was well developed and the majority of the epithelial cells were flattened. By electron microscopy, however, cuboidal epithelial cells began to show cytoplasmic extension as early as 95 days (Fig. I). Two types of epithelial cells became distinguishable at about i io days. Type I cells had larger nuclei and a small amount of cytoplasm with well defined cytoplasmic extension. Type II cells exhibited a large amount of cytoplasm with very few organelles; Golgi zones were regularly observed and the remainder of cytoplasm consisted of scattered free ribosomes in a moderately dense background (Fig. 2). Type I and II cells were joined by i or several terminal bars (Fig. 3). At the same time capillaries became incorporated within the alveolar wall. The alveolar septum at this stage was still thick.

Nov., I965

ALVEOLAR LINING LAYER

879

Osmiophilic inclusion bodies first appeared within type II cells in i animal as early as II3 days gestation and were regularly manifest at about 12i days. At the latter time the number of free ribosomes increased and a few endoplasmic reticulum channels appeared. Multivesicular bodies and Golgi zones also increased in number but mitochondria were scarce (Fig. 3). At I30 days free ribosomes, granular endoplasmic reticulum and inclusion bodies were all increased in number (Fig. 4). At I35 days, the ribosomes and inclusion bodies filled up most of the cytoplasm (Fig. 5). At all stages of development mitochondria were rare. The inclusion bodies were not found to be laminated; they measured between O. I5 and 2 /u in maximum size and were usually bound by a unit membrane (Fig. 6b). The periphery of the inclusion was generally dense and its center electron-lucent, although occasionally a dense central core was present (Fig. 3). The smallest inclusion bodies tended to occur deep in the cytoplasm (Fig. 6a). At the periphery of the cell they were of average diameter, as though they had increased in size as they moved toward the cell surface (Figs. 6b and 6c). The unit membrane apparently joined the epithelial cell membrane and excreted the inclusion content into the alveolar space (Fig. 6d). When this content was exposed to the alveolar space, fine laminations, measuring 30 A in width and separated by clear spaces 25 A wide, were seen within the otherwise amorphous electron-dense material (Fig. 7). The laminations were arranged in parallel lines and were quite different from those described by others 8-10 in so-called "laminated inclusion bodies." In the latter the laminations were larger, irregular in width, and arranged in more random fashion, making the measurement of spacing difficult. In some areas electron-dense homogeneous material similar to inclusion body content was in continuity with multilaminated membranes (myelin figure) which measured 8o to IOO A in width (Fig. 8). In several areas the myelin figure was further divided into 2 or 3 finer laminations measuring the same as those in the discharging inclusion bodies (Fig. 9). After I2I days gestation a dense osmiophilic layer measuring ioo to I50 A thick appeared on the epithelial cell membrane (Fig. io). This was best seen when the epithelial surface was compared with that of the endothelial cell (Fig. ii). Osmiophilic layers were only found in lungs containing inclusion bodies and they became more prevalent with the progress of maturation until I35 days. There appeared to be a direct relation between the number of interruptions in the alveolar lining and the number of myelin figures in an alveolus. Type II cells containing inclusion bodies increased in number from about 3 per cent of the total parenchymal cells at I2 i days to 9 per cent

Vol. 47, No. 5

KIKKAWA, MOTOYAMA AND COOK

88o

at full term (Text-fig. I). The average number of inclusion bodies per type II cell also increased with maturation from about 4 at I2 Idays to 8 to at full term (Text-fig. 2); thus the average number of inclusion io

bodies at I 2

i

lung parenchymal cells ranged from approximately 12 days to 8o at term. When the per cent of type II cells with incluper

ioo

.e

I

0

i

i i i i i i i i i i i

10

0

0 0

i i i i i i i I i i i i

TYPE E CELLS

CONTAINING

I~~~

6

INCLUSIONS

0

00 0

0

i

4

0 xx

(PER CENT)

*

I:

X * x

2 i tX

0

I

X

I

S 150 140 130 120 110 GESTATIONAL AGE (DAYS) TEXT-FIG. i. The number of type II alveolar cells containing osmiophilic inclusion bodies (expressed as per cent of total lung parenchymal cells) versus gestional age. 0, fetuses; 0, lambs which breathed; X, animals with respiratory distress. Lambs in the section marked S were born spontaneously. The interrupted vertical line at I26 days indicates the gestational age after which normal surface activity of lung extracts is usually found. Normal surface activity was found in all lungs which contained more than 4 per cent type II cells with inclusion bodies. Lungs of lambs with respiratory distress apparently have a decreased number of inclusion-containing type II cells and normal surface activity Is lost.

100

sion bodies were related to the presence or absence of normal surface activity of the lung extracts, it was apparent that all lungs of lambs above I26 days gestation contained more than 4 per cent type II cells and had normal surface activity (Text-fig. 3). Influence of Air Breathing None of the 24 non-asphyxiated lambs at I30 or more days gestation which had breathed for 30 minutes to 4 hours developed respiratory distress. The lungs in these animals were slightly more distended when

ALVEOLAR LINING LAYER

Nov., z965q

88I

compared to those of fetuses of comparable age. Electron microscopy revealed the presence of the same osmiophilic lining layer seen in fetuses but with fewer interruptions. Inclusion bodies were here also discharged into the alveoli but the number of intra-alveolar myelin figures also appeared reduced when compared to fetuses. There was no apparent .

A

I0

8

NUMBER OF 6

INCLUSIONS PER

4

TYPE I CELL 2

0 -

130 S 140 120 130 GESTATIONAL AGE (DAYS) TEXT-FIG. 2. The average number of osmiophilic inclusion bodies per type II alveolar 110

cells versus gestational age. *, fetuses; 0, lambs which breathed; X, animals with respiratory distress. The lambs in the section marked S were born spontaneously. The interrupted vertical line at 126 days indicates the gestational age after which normal surface activity of lung extracts is usually found.

difference in either the number of type II cells containing inclusion bodies or the total number of inclusion bodies in the lungs of lambs which breathed as compared to fetuses of similar gestational ages. All 6 lambs at gestational age between 125 and 129 days that breathed for I to 4 hours developed respiratory distress with a severe disturbance in blood-gas exchange and abnormal surface activity of the lung extract.2 The lungs were grossly liver-like, showed diffuse atelectasis and contained hyaline membranes. Electron microscopy showed that the type I and II cells were frequently severely damaged; their cell membranes were ruptured and most of the cytoplasmic organelles, including inclusion bodies, were lost. In better preserved cells inclusion bodies were fewer in number and the cytoplasm contained many clear

Vol. 47, No. 5

KIKKAWA, MOTOYAMA AND COOK

882

vacuoles about the size of the inclusion bodies. Even the better preserved inclusion bodies appeared to have lost electron density as compared to normal (Fig. 12). Thus, both the number of type II cells containing inclusion bodies and the total number of inclusions were decreased as compared to animals of comparable gestational age (Text-

25 0

0~~~~~~

*?*

20 LOWEST 15

*

*

.;

*F 0

0

i

0~~~~~~~~~~~~~0:

I~~~~~~~~~~~~~~~!

SURFACE

i

0

i

0

TENSION 10

O

;

(DYNES/CM)

i.

5

Os

@0 O-g

00

-

0

v 0000

0@

0 0

i0

0

I

*

3

Ia

I

I,I

I

100 110 120 130 140 150 S GESTATIONAL AGE (DAYS) TEXT-FIG. 3. The minimum surface tension of lung extracts versus gestational age. The lambs in the section marked S were born spontaneously. Connected points represent values in twins. The horizontal line indicates the dividing point between normal (< I5 dynes per cm) and abnormal (> I5 dynes per cm) surface activity of lung extracts.1"'7 The interrupted line at 126 days shows the age after which normal surface activity is usually found. (Reprinted from Pediatrics' with permission of the editor.)

60

70

80

90

figs. and 2). The dense osmiophilic lining layer seen in normal lungs was either absent or scanty in the area of atelectasis and the capillary endothelium contained increased numbers of pinocytotic vacuoles (Fig. I3). Hyaline membranes consisted mostly of electron-dense granular material intermixed with cellular debris. In some areas, however, bundles of fibrils were observed (Fig. 14). In some of these bundles, a 230 A periodicity could be demonstrated (Fig. I 5). i

Effect of Prenatal Asphyxia In 4 of the 6 lambs between I30 and I36 days gestation and in I of 6 animals between I37 and I42 days, respiratory distress, similar to that seen in the younger group, was observed. This was apparently the result of immaturity plus severe prenatal asphyxia. The more mature lambs did not develop respiratory distress following severe prenatal asphyxia

Nov., z965y

ALVEOLAR LINING LAYER

883

and there was no apparent damage to the type II cells. The osmiophilic lining layer was present as often as in control animals. One gained the impression, however, that some inclusion bodies were less electron-dense as compared to controls. All the lambs which developed respiratory distress showed pulmonary atelectasis and hyaline membrane formation. Electron microscopic observations were similar to those described in the previous group of animals. DIscusSION Osmiophilic Inclusion Bodies and Surfactant Inclusion bodies were first described by Sj6strand and Sjostrand11 in 1938 and were demonstrated electron microscopically by Schlipkoter 12 and Low13 in 1954. Karrer14 found the inclusions to be present in type II alveolar epithelial cells and Macklin 15 suggested that inclusions might be secreted into the alveolar space. In the mammalian species studied to date the inclusion bodies have appeared concentrically laminated,8 made up of stacks of membranes,8 or have been septate16 in appearance. In the study of the development of the mouse lung, Buckingham and Avery17 found that normal surface activity of lung extracts was first detectable at I8 days gestation (i.e., I day before term). Since there was evidence by electron microscopy that inclusion bodies first appeared at the same gestational age,10 they, as well as others,18 have suggested that inclusion bodies and surfactant might be closely related. In the present study a relation between the number of inclusion bodies and the time of appearance of normal surface activity was demonstrated. The bodies first appeared consistently at about I2I days gestation (corresponding to approximately 27 weeks gestation in the human) and steadily increased in number with maturation. Normal surface activity of the lung extracts was found at about I26 days. The fact that the inclusion bodies were found within deeply invaginated sacs of the cell membrane (Fig. 6d) as reported by others, 8,19 and that no osmiophilic lining layer or myelin figures were observed within an alveolus before the appearance of inclusion bodies in type II cells, may be interpreted as evidence of the excretory function of these cells. Inclusion bodies have been thought, on the basis of histochemical analysis, to contain primarily lipid11"14 or phospholipid.9 Surface active substances in the mammalian lungs are apparently also phospholipid 20 or lipoprotein.21 The present study provides convincing additional evidence concerning the phospholipid nature of both the inclusion body content and the osmiophilic alveolar lining layer. This is indicated by the presence of fine laminations with 2 5 A spacing within the otherwise

884

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

homogeneous content of inclusions; the laminations are similar to those demonstrated by Stoeckenius 22 in phospholipid preparations. Further support is the persistence of myelin figures (Figs. 8 and 9), which have been shown to contain phospholipid,23 in an electron-dense material identical to both the content of the inclusions and to that of the surface lining. The presence of myelin figures within the alveoli appears to be a demonstration in situ of what Mendenhall and Sun24 have shown in preparations of pig lung washings. Following the atelectasis caused by C02 poisoning2l or homotransplantation of the lung,"' inclusion bodies have been shown to lose their electron density. In the present study in lambs with atelectasis and hyaline membranes, inclusion bodies also lost electron density and many intracytoplasmic vacuoles appeared. A decrease in the number of inclusion bodies was associated with cellular necrosis and the apparent transition of the inclusion to a vacuole as electron-dense material was lost. Some evidence presented by Schaefer, Avery and Bensch25 has suggested that such a vacuole may still contain altered inclusion content. The loss of electron density, therefore, can be due to either a quantitative or a qualitative alteration, but this cannot be determined without tracer studies. Consistent with these experimental observations are the observations of Campiche,9 who reported a decrease in the number of inclusions in the lungs of infants dying with the respiratory distress syndrome. He suggested a relationship between this decrease and the development of the disease. The correlation between the number of inclusion bodies and the time of appearance of normal surface activity, the demonstration of the excretory function of type II cells, the probable phospholipid or lipoprotein nature of inclusion bodies and the morphologic alteration of inclusion bodies with such stress as severe prenatal asphyxia, coincident with the loss of normal surface activity, all appear to justify the hypothesis that the inclusion body is a precursor of the pulmonary surfactant.

Surfactant and the Alveolar Lining Layer Tyler and Pangborn,26 as the result of electron microscopy studies, described a laminated membrane on the epithelial surface of the tertiary bronchi and atria of chicken lung. Because of the affinity of the membrane for osmium tetroxide, they suggested that the membrane contained a phospholipid or a lipoprotein. In mammals, however, a similar osmiophilic layer has not been reported, although a preparation of pig lung washings was shown to have laminated membranes with 95 A spacing.24 The present studies are the first to demonstrate the osmiophilic surface layer in the mammalian lung in situ.

Nov., z96,5

ALVEOLAR LINING LAYER

885

The presence of an osmiophilic lining layer is directly related to the number of inclusion bodies and the surface activity of the lung extract. This layer first appears at about I2I days gestation, concomitant with the first appearance of inclusion bodies. During maturation the number of inclusion bodies increases. The osmiophilic layer is more consistently present with maturation but appears to reach a plateau at about I35 days. Normal surface activity of the lung extract is detectable with the surface balance for the first time at about 126 days. In both the premature and the prenatally asphyxiated animals which develop respiratory distress and associated loss of normal surface activity, this layer is either scanty or absent in the areas of atelectasis. On the basis of these observations, it seems probable that such an osmiophilic layer represents surfactant and is derived from inclusion bodies. If one accepts the hypothesis that the alveolar lining layer is a continuous film of phospholipid or lipoprotein whose function is to stabilize alveoli during breathing,27-29 one must assume that its apparent discontinuity in the lungs of mature fetal and newborn lambs is an artifact. This possibility is supported by the observation in mature lambs that there are more interruptions in the osmiophilic lining layer within a given alveolus when there are more myelin figures, suggesting that the latter may have been formed from defective portions of the osmiophilic layer. In immature animals, the interruptions in the lining layer probably more closely represent the situation in vivo since the inclusion bodies, and therefore the amount of surfactant, are both insufficient. In newborn lambs which breathed spontaneously, a greater amount of osmiophilic lining layer and fewer myelin figures than in fetal animals of the same gestational age were observed. Since the method of tissue preparation was identical in both groups of animals, some additional mechanism must have been involved in the formation of the lining layer. If one speculates that the formation of the continuous lining layer is accelerated by the onset of breathing, myelin figures in the fetal lungs may represent the freely floating surfactant contributing to the formation of the lining layer. The relation between the osmiophilic layer described here, the fluorescent layer described by Hackney, Rounds and Schoen,"0 and Bolande and Klaus,31 the mucopolysaccharide layer observed by Chase32 and Groniowski and Biczyskowa 38 still remains to be explored. It is probable that the osmiophilic layer and the fluorescent layer are closely related because of the similar correlation of both layers with surface activity. The mucopolysaccharide layer probably represents an additional lining coat, since it is not known to be osmiophilic or to possess surface activity. The comparison of our findings with others reveals several important differences. In lambs inclusion bodies are usually septate in appearance

886

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

or contain a large component of electron-dense, homogeneous substance, unlike other mammalian species studied by the same technique.16'34 The 25 A spacing of laminations in the inclusions and in the myelin figures in the lamb is identical to the spacing of the multiple laminations in the chicken lung but differs from the 95/2 A spacing noted in pig lung washings. The dense osmiophilic surface layer seen in the lungs of lambs is different from the multilaminated surface layer in the chicken lung. These differences suggest that the composition of inclusion bodies and surfactant may be different in various animal species. If this were so it would explain some of the difficulties in the identification of an osmiophilic lining layer in other animals. In some animals, there may be fewer available double bonds with resulting absence of osmiophilia.

Cytologic Immaturity and Respiratory Distress Since there was a close association between prematurity and the development of respiratory distress in these experimental animals, prematurity is considered the most important factor in the production of this condition with perinatal stress such as asphyxia a probable contributory factor.2 By light microscopy it was not possible to differentiate the lungs of a I2 i-day fetus from those of the full term lamb. The surface activity of lung extracts also showed no change from I26 days to term. In the present study, however, using electron microscopy, in lambs between I25 and I35 days gestation, the number of ribosomes and inclusions in type II cells were fewer than in the fully mature cell (Fig. 4 and Text-figs. i and 2). These findings clearly demonstrate the cytologic immaturity of type II cells until the fetus reaches at least I35 days gestation, and probably explain why the premature lamb fetus, younger than I35 days, is apt to develop respiratory distress when subjected to asphyxia. Hyaline Membranes Gitlin and Craig,35 utilizing a fluorescent antibody technique, have shown that hyaline membranes are composed of fibrin. Van Breeman, Neustein and Bruns,36 using electron microscopy, demonstrated bundles of fibrin within the mesh of granular material; the fibrils showed a suggestion of periodicity. We have demonstrated fibrils with 230 A periodicity. This corresponds to the findings of Hall 37 who used a pure fibrin preparation. This periodicity, as shown by Still and Boult,38 using pure fibrin preparation, does not always appear to be demonstrable, especially in thin sections. In both studies, however, the published photomicrographs consist entirely of the bundles of fibrils with or without periodicity.

Nov., 1965

ALVEOLAR LINING LAYER

887

In tissue preparations the presence of bundles of fibrils suggest the presence of fibrin, but for definitive identification one must find periodicity of 220 to 250 A. In our studies bundles of fibrin, with or without periodicity, were rare and the bulk of the hyaline membrane consisted of a granular material, which could not be further characterized. It is probable that the granular material contains various stages of fibrinogen polymers and serum proteins. This supposition does not contradict with a diffuse fluorescence shown by Gitlin and Craig in hyaline membranes, because fluorescein labeled anti-fibrin antibody is known to react with any intermediary polymers of fibrinogen. The relative amount of these different proteins cannot be determined by electron microscopy, since most serum proteins are granular without other recognizable structural features.39 SUMMARY AND CONCLUSIONS The lungs in 69 fetal and newborn lambs were studied by the electron microscope. Osmiophilic inclusion bodies first appeared at about 12I days gestation and their total number increased with maturation. Normal surface activity of lung extracts was detectable a few days following the appearance of inclusion bodies. The excretory nature of the type II alveolar epithelial cell, the phospholipid nature of the inclusion body content, the decrease in the number of inclusion bodies and their loss of density associated with respiratory distress and with the loss of normal surface activity of lung extracts provide strong evidence that inclusion bodies are the source of pulmonary surfactant. A dense osmiophilic alveolar lining layer has been described in mammals for the first time. Since the presence of such a layer is well correlated with surface tension values and the number of inclusions, it is suggested that this layer consists of surface active substances. The difficulty in detecting such an osmiophilic layer in other mammals is thought to be, at least in part, due to species differences in the composition of materials constituting the pulmonary surfactant. Cytologic immaturity exists until lamb fetuses reach I35 days gestation. This is considered to be the basis for the susceptibility of immature lambs to respiratory distress. Fibrin with 230 A periodicity has been shown to be a component, although a small one, of hyaline membranes in lambs. In view of the fact that mature fibrin appearing as bundles of fibrils is rarely found, it is suggested that the bulk of the hyaline membrane is not a mature fibrin, but probably consists of polymers of fibrinogen and serum protein. REFERENCES I. ORZALESI, M. M.; MOTOYAMA, E. K.; JACOBSON, H. N.; KIKKAWA, Y.; REYNOLDS, E. 0. R., and COOK, C. D. The development of the lungs of lambs. Pediatrics, I965, 35, 373-381.

888

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

2. REYNOLDS, E. 0. R.; JACOBSON, H. N.; MOTOYAMA, E. K.; KIKYAwA, Y.; CWG, J. M.; ORzALEsI, M. M., and COOK, C. D. The effect of immaturity and prenatal asphyxia on the lungs and pulmonary function of newborn lambs: The experimental production of respiratory distress. Pediatrics, I965, 35,

382-392. 3. SILVERMAN, W. A., and SILVERMAN, R. H. "Incidence" of hyaline membrane in premature infants. Lancet, I958, -2, 588. 4. SIVANESAN, S. Neonatal pulmonary pathology in Singapore. J. Pediat., 1961,

59, 6oo-6o6. 5. COHEN, M. M.; WEINTRAUB, D. H., and LILIENFELD, A. M. The relationship of pulmonary hyaline membrane to certain factors in pregnancy and delivery.

Pediatrics, I960, 26, 42-50.

6. JAMES, L. S. Physiology of respiration in newborn infants and in the respiratory distress syndrome. Pediatrics, 1959, 24, IO69-IIOI. 7. REYNOLDS, E. 0. R.; ORZALESI, M. M.; MOTOYAMA, E. K.; CRAIG, J. M., and COOK, C. D. Surface properties of saline extracts from lungs of newborn infants. Acta Paed. Scand. (In press) 8. BALIS, J. U., and CONEN, P. E. The role of alveolar inclusion bodies in the developing lung. Lab. Invest., I964, I3, 1215-I229. 9. CAMPICHE, M. A. Les inclusions lamellaires des cellules alveolaires dans le poumon du raton. Relations entre l'ultrastructure et la fixation. J. Ulstrastruct. Res., i960, 3, 302-312. IO. WOODSIDE, G. L., and DALTON, A. J. The ultrastructure of lung tissue from newborn and embryo mice. J. Ultrastruct. Res., I958, 2, 28-54. I I. SJ6STRAND, F., and SJ6STRAND, T. tYber die granulierte Alveolarzelle und ihre Funktion. Z. Mikr-Anat. Forsch., 1938, 44, 370-41I. I2. SCHLIPKOTER, H. W. Elektronenoptische Untersuchungen ultradiinner Lungenschnitte. Deutsch. Med. Wschr., 1954, 79, I658-I659. I3. Low, F. N. The electron microscopy of sectioned lung tissue after varied duration of fixation in buffered osmium tetroxide. Anat. Rec., 1954, I20, 82785i. I4. KARRER, H. E. The ultrastructure of mouse lung. General architecture of capillary and alveolar walls. J. Biophys. & Biochem. Cytol., i956, 2, 241-252. 15. MACKLIN, C. C. The pulmonary alveolar mucoid film and the pneumonocytes. Lancet, I954, I, 1099-II04. I6. AMIRANA, M. T.; ROHMAN, M.; OKA, M.; KiKKAWA, Y.; GUEFT, B., and STATE, D. Functional and pathologic changes in the reimplanted lung. Surg. Forum, 1964, 15, 177-I79. I7. BUCKINGHAM, S., and AVERY, M. E. Time of appearance of lung surfactant in the foetal mouse. Nature (London), T962, 193, 688-689. I8. KLAUS, M.; REiss, 0. K.; TOOLEY, W. H.; PIEL, C., and CLEMENTS, J. A. Alveolar epithelial cell mitochondria as source of the surface-active lung lining. Science, I962, I37, 750-751. I9. BENSCH, K.; SCHAEFER, K., and AVERY, M. E. Granular pneumocytes: electron microscopic evidence of their exocrinic function. Science, I964, I45, I3I8-I3I9. 20. KLAUS, M. H.; CLEMENTS, J. A., and HAVEL, R. J. Composition of surfaceactive material isolated from beef lung. Proc. Nat. Acad. Sci. USA, 1961, 47, I858-I859. 2I. PATTLE, R. E., and THOMAS, L. C. Lipoprotein composition of the film lining the lung. Nature (London), I96I, I89, 844.

Nov., I965

ALVEOLAR LINING LAYER

889

22. STOECKENIUS, W. Structure of the plasma membrane. An electron-microscopic study. Circulation, I962, 26, io66-io69. 23. REVEL, J. P.; ITO, S., and FAWCETT, D. W. Electron micrographs of myelin figures of phospholipide simulating intracellular membranes. J. Biophys. & Biochem. Cytol., I958, 4, 495-498. 24. MENDENHALL, R. M., and SUN, C. N. Surface lining of lung alveoli as a structure. Nature (London), I964, 201, 713-7I4. 25. SCHAEFER, K. E.; AVERY, M. E., and BENSCH, K. Time course of changes in surface tension and morphology of alveolar epithelial cells in C02-induced

hyaline membrane disease. J. Clin. Invest., I964, 43, 2080-2093. 26. TYLER, W. S., and PANGBORN, J. Laminated membrane surface and osmiophilic inclusions in avian lung epithelium. J. Cell. Biol., I964, 20, 157-I64. 27. BROWN, E. S.; JOHNSON, R. P., and CLEMENTS, J. A. Pulmonary surface tension. J. Appi. Physiol., I959, 14, 717-720. 28. PATTLE, R. E. Properties, function and origin of the alveolar lining layer. Nature (London), I955, I75, II25-II26. 29. CLEMENTS, J. A.; BROWN, E. S., and JOHNSON, R. P. Pulmonary surface tension and the mucus lining of the lungs: Some theoretical considerations. J. Appl. Physiol., I958, I2, 262-268. 30. HACKNEY, J. D.; ROUNDS, D. E., and SCHOEN, A. W. Observation of a lipid lining in mammalian lung. (Abstract) Fed. Proc., I963, 22, 339. 3I. BOLANDE, R. P., and KLAUS, M. H. The morphologic demonstration of an alveolar lining layer and its relationship to pulmonary surfactant. Amer. J. Path., I964, 45, 449-463. 32. CHASE, W. H. The surface membrane of pulmonary alveolar walls. Exp. Cell Res., I959, i8, 15-28. 33. GRONIOWSKI, J., and BICZYSKOWA, W. Structure of the alveolar lining film of the lungs. Nature (London), I964, 204, 745-747. 34. KIKKAWA, Y., and MOTOYAMA, E. K. Unpublished observations. 35. GITLIN, D., and CRAIG, J. M. The nature of the hyaline membrane in asphyxia of the newborn. Pediatrics, I956, I7, 64-7I. 36. VAN BREEMEN, V. L.; NEUSTEIN, H. B., and BRUNS, P. D. Pulmonary hyaline membranes studied with the electron microscope. Amer. J. Path., 1957, 33,

769-789. 37. HALL, C. E. Electron microscopy of the fibrinogen molecule and the fibrin clot. Lab. Invest., I963, I2, 998-IOOI. 38. STILL, W. J. S., and BOULT, E. H. Electron microscopic appearance of fibrin in thin sections. Nature (London), I957, 179, 868-869. 39. HOGLUND, S. Macroglobulin from Alpha 2 and Gamma i Fractions of Normal Human Serum. Third European Regional Conference of Electron Microscopy. Publishing House of Czechoslovak Academy of Sciences, Czechoslovakia, I964, Vol. B., pp. 55-56. We are indebted to Drs. A. Angrist, J. C. Craig, B. Gueft, H. N. Jacobson, M. M. Orzalesi and E.O.R. Reynolds for helpful assistance and valuable advice.

[ Illustrations follow ]

890

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

LEGENDS FOR FIGURES Unless otherwise stated, micrographs were prepared from sections stained with uranyl acetate. FIG. I. A portion of cuboidal epithelium with early cytoplasmic extension (arrow) in the lung of a 95-day-old fetal lamb. Ribosomes and granular endoplasmic reticulum are conspicuous. At this stage type I and II cells cannot be differentiated on the basis of the cytologic features. Normal surface activity of lung extract absent. X IO,OOO. FIG. 2. Type II epithelium in a I io-day fetus. The cell is cuboidal and its cytoplasm is mostly devoid of organelles. Occasional granular endoplasmic reticulum (GE) and free ribosomes can be seen. The Golgi zone (G) is prominent. A mitochondrion (M) appears inactive. The cytoplasm consists of granular material of low density. Type II cell is bound by basement membrane (BM). Mitochondria (m) of a septal cell are also shown. Normal surface activity absent. X IO,OOO.

Nov.. 1965

ALVEOLAR LINING LAYER

89I

1

892

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

FIG. 3. Type II cell in a I2i-day lamb fetus. There is no cytoplasmic extension. The cell is cuboidal and joined to an attenuated portion of a type I cell by terminal bars (TB) on either side. Multivesicular bodies (MVB) are present. Free ribosomes and the granular endoplasmic reticulum are both seen occasionally. An inclusion body (IB) appears near the cell membrane. Microvilli (MV) are observed for the first time. Normal surface activity absent. x I I ,000.

FIG. 4. Type II cell in a Q3o-day fetus. Free ribosomes and the granular endoplasmic reticulum are increased. Inclusion bodies (IB) with electron-dense, homogeneous areas and electron-lucent centers are numerous at this stage. Normal surface activity of lung extract is present. Uranyl acetate, lead citrate stain. X 9,ooo.

Nov., I965

ALVEOLAR LINING LAYER

893

3

4

894

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. 5

FIG. 5. Type II cell in a I35-day fetal lung. Much of the cytoplasm is occupied by clumped ribosomes and inclusion bodies, the latter often showing septate appearance at this stage. Normal surface activity present. X IO,OOO. FIG. 6. Type II cell in a mature fetal lung. a. A small inclusion appears deep in the cytoplasm (IB). Nearby there are a few others with lighter density (arrows). The granular endoplasmic reticulum is dilated. X 35,000. b. An inclusion body is bound by a unit membrane (arrow). The epithelial cell membrane also has a unit membrane of the same thickness. X 6o,ooo. c. An inclusion body approaches the epithelial surface. X 50,000. d. The content of an inclusion body is discharged into an alveolus. A portion of the content still remains deep in a pocket (arrow). X 50,000.

Nov., I965

ALVEOLAR LINING LAYER

895

5

6b

6c

6d

896

KIKKAWA, MOTOYAMA AND COOK

Vol. 47, No. g

FIG. 7. Mature fetal lung. Another inclusion is in the processs of discharge. Fine laminations (arrow) can be seen. X 75,000. Inset: Fine laminations are photographically magnified. Each line is about 3o A thick and separated by a clear space 25 A wide. X I70,000. FIG. 8. Mature fetal lamb. A multilaminated myelin figure is seen in an alveolus. Each membrane measures about 8o to ioo A in thickness. Laminated membranes are in continuity with a dense homogenous material (arrow). Inclusion bodies (IB) also appear within the cytoplasm. X 48,ooo.

897

ALVEOLAR LINING LAYER

Nov., I965

__

k

:3i.:

7

*~ ~ ~ ~ *XS.is;.,', _|_{'8k6;

-.^a x-*MF

mm!

Ll¢.;

t-

> ; ]d-

A I. .

7'. s.

8

KIKKAWA, MOTOYAMA AND COOK

898

Vol. 47, No. 5

FIG. 9. The continuity of the homogeneous dense material and the laminated membranes is clearly seen (arrow). Some membranes are divided into 3 lines measuring about 30 A wide (FL). The clear spacing measures 25 A. X I35,000. FIG. io. Mature fetal lung. A low power view of an osmiophilic alveolar lining layer (OLL) covering the attenuated portion of type I cell. A red cell (RBC) is seen within a capillary. Normal surface activity of lung extract is present. X 4,200.

Nov., I965

ALVEOLAR LINING LAYER

899

9

10

KIKKAWA, MOTOYAMA AND COOK

goo

Vol. 47, No. .s

FIG. ii. A higher power view of an osmiophilic alveolar lining layer (OLL) on an epithelial cell membrane. When the endothelial cell membrane is compared with surface lining layer, the difference is quite obvious. Epithelial cytoplasm (EP), basement membrane (BM) and endothelial cytoplasm (EN) are clearly separated. A red cell appears in a capillary. Normal surface activity present. X 65,ooo.

FIG. I2. A portion of a type II cell from a lamb at I35 days gestation with respiratory distress. The central electron-lucent area is wider than normal. Some inclusions (IBi) resemble vacuoles, while others (lB2) show a small amount of the electron-dense content. Normal surface activity of lung extract is lost. X 50,000.

Nov., Ir965

ALVEOLAR LINING LAYER

90I

11

12

902

13

14

KIKKAWA, MOTOYAMA AND COOK

VOI. 47, No. 5

Nov., 1965

ALVEOLAR LINING LAYER

903

15

FIG. I3. The osmiophilic lining layer is absent in a lamb of I35 days gestation with respiratory distress. The epithelial cytoplasm (EP) appears degenerated. Capillary endothelial cytoplasm (EN) contains many vesicles. Normal surface activity absent. X 65,ooo. FIG. I4. Bundles of fibrils within an alveolus are mixed with cellular debris. X 36,ooo. FIG. I5. A higher power view of these fibrils shows a periodicity of 230 A. X I 00,000.