postmortem human lungs may be adequately fixed for electron microscopy. An electron microscopy and morphometry study of lungs of patients dying of adult respiratory distress syndrome revealed that the alveolar epithelium was transformed to a vastly cuboidal epithelium irrespective of the inspired oxygen concentration, which in several cases did not exceed 40 percent. This constant change in alveolar epithelium, resulting in a consistent overall thickening of the epithelial barrier by a factor of 4: 5 in all cases, was interpreted to represent a characteristic reaction pattern of alveolar epithelium to various kinds of injury. Arguments are presented to indicate that hypertrophy of cuboidal type 2 epithelial cells is a necessary step in the repair process, since type 1 cells are unable to divide by mitosis. ACKNOWLEDGMENT: We gratefully acknowledge the generous assistance received from the staH of the Department of Pathology, in particular from Prof. B. Roos and Prof. H. Cottier; skillful technical assistance was received from Miss H. Claassen, Mr. Karl Bahl, and Miss G. Reber.
REFERENCES 1 Alley MR, Manktelow BW: Alveolar epithelialization in ovine pneumonia. J Patholl03:219-224, 1971 2 Carrington CB, Green TJ: Granular pneumocytes in early repair of diffuse alveolar injury. Arch Intern Med 126: 464465, 1970 3 Evans MJ, Cahral LJ, Stephens RJ, et al: Renewal of alveolar epithelium in the rat following exposure to N02. Am J Pathol 70:175-198,1973 4 Gould VE, Tosco R, Wheelis RF, et al: Oxygen pneu-
9 10 11
monitis in man. Ultrastructural observations on the development of alveolar lesions. Lab Invest 26: 499-508, 1972 Kapanci Y, Tosco R, Eggermann J, et al: Oxygen pneumonitis inman. Light- and electronmicroscopic morphometric studies. Chest 62:162-169, 1972 Kapanci Y, Weibel ER, Kaplan HP, et al: Pathogenesis and reversibility of the pulmonary lesions of oxygen toxicity in monkeys. 2. Ultrastructural and morphometric studies. Lab Invest 20: 101-118, 1969 KauHman SL, Burri PH, Weibel ER: The postnatal growth of the rat lung. Part 2: Autoradiography. Submitted to Anat Rec Kistler GS, Caldwell PRB, Weibel ER: Electronmicroscopic and morphometric study of rat lungs exposed to 97S oxygen at 258 Torr (27,000 feet). Tech Rep US-Air Force, AMRL-TR-66-103, 1966 Nash G, Blennerhassett JB, Pontoppidan H: Pulmonary lesions associated with oxygen therapy and artificial ventilation. N Engl J Med 276:368-374, 1967 Omar AR: The characteristic cells of the lung and their reaction to injury. 1,2. Vet Bull 34:371 and 431, 1964 Schwinger G, Weibel ER, Kaplan HP: Pulmonary pathology of oxygen toxicity. Part 3: Electronmicroscopic and morphomebic study of dog and monkey lungs exposed to 98S 02 at 258 torr for seven months and followed by one month recovery in room air. Interim Scienti6.c Report USAir Force, AF 61 (052)-941, 1967 Weibel ER: Morphological basis of alveolar-capillary gas exchange. Physiol Rev 53:419-495, 1973 Weibel ER: The mystery of "non-nucleated plates" in the alveolar epithelium of the lung explained. Acta Anat 78:425-443, 1971
A Note on Differentiation and Divisibility of Alveolar Epithelial Cells Ewald R. Weibel, M.D.O
of the most intriguing recent findings in the O nebiology of lung cells is that 1 cells of alveolar
type ep1thelium are apparently unable to undergo mitotic division.s-" Consequently, damage to the epithelial lining of alveoli must be repaired by proliferation of type 2 cells, with subsequent trans differentiation into squamous type 1 cells,2-4 a topic that has become a central theme at the 1973 Aspen Conference. It is, however, generally assumed that the type 1 cell of alveolar epithelium is less highly differentiated than the type 2 cell. This judgment is based on the fact that the type 1 cell, which is sometimes also called the "small" alveolar cell, contains a thin shell of cytoplasm surrounding the nucleus, and that this cytoplasm is extraordinarily poor in organellae. In contrast, the type 2 cell appears "large" on section; its abundant cytoplasm is rich in organellae and contains a highly specific granule in the form of the osmiophilic lamellar bodies, which are . the secretory granules for the phospholipid component of pulmonary surfactant. In this sense, cytoplasmic differentiation of type 2 cells greatly exceeds that of type 1 cells, and one would expect the cell of greater diHerentiation to be a less likely candidate for cell division than the poorly diHerentiated type 1 cell. Accordingly, the o Anatomisches
Institut der U niversitat Bern, Bern, Switzer-
CHEST 65: 4, APRIL, 1974 SUPPLEMENT
observation that type 1 cells cannot divide, but are derived from freely dividing type 2 cells appears as a paradox. This point of view does not regard, however, another type of diHerentiation which we choose to call cCtopologic differentiation." It is the purpose of this appendix to demonstrate that type 1 cells show a very high level of topologic diHerentiation, and that this may inhibit their faculty to divide by mitosis. The alveolar epithelium is a simple mosaiclike epithelium composed of squamous (type 1) and cuboidal (type 2) cells." In this sort of epithelium all cells have three faces: an apical face in contact with the external space; a basal face attached to the basement membrane; and a lateral face in contact with the neighboring cell. The intercellular clefts are closed by tight junctions or terminal bars, which mark the boundary between apical and lateral face. Figures 1 (a) and (b) show a semischematic representation of this feature, with respect to type 1 and 2 cells. In 1971 we showed that both cell types of alveolar epithelium, but in particular, type 1 cells, may deviate from this general scheme in a unique and characteristic way. It can be observed that a type 1 cell may form a OWe are disregarding in this context the rarely occurring third cell type, the brush cell,8 which is also cuboidal.
16TH ASPEN LUNG CONFERENCE 19S
FIGURE 1. Schematic diagram of type 2 cell ( Q. ), simple type 1 cell (b), and type 1 cells with two (c) and four ( d) apical cytoplasmic plates. Heavy contour indicates apical plasma membrane bounded by terminal bars (dots). Cell's division plane is marked by vertical line; in ( 0 ) there are several possibilities for its course.
cytoplasmic flap with its apical face bounded by a terminal bar ring on one side of the interalveolar septum, extending across the septum with a cytoplasmic stem and forming a second cytoplasmic flap on the other side, which is again bounded by a closed terminal bar ring (Fig Ie). The consequence is that one nucleus will serve two apical cytoplasmic plates. This observation explains the mysterious postulate of KollikerG that alveolar epithelium consisted of "nonnucleated plates/" We 7 also showed that the diHerentiation of type I cells could be even more complex, since we found direct evidence of the occurrence of a minimum of three apical plates; furthermore, that cytoplasmic stems could extend across the septum also in regions where no nucleus was observed, suggesting a peripheral "branching" of cytoplasmic plates. Recent observations have, indeed, shown that multiple apical plates belonging to one and the same cell are possible. Figure Id shows a diagram of a section of a type I cell from the lung of an Etruscan shrew, a small mammal. In this case, a minimum of four apical plates are directly traceable on sections, but it is most likely that many more exist. It is noteworthy that branching of cytoplasmic flaps at a certain distance from
20S 18TH ASPEN LUNG CONFERENCE
the nucleus is directly demonstrated in this instance. Indirect evidence suggests that, in the rat and shrew at least, each type 1 cell must have a minimum of ten apical plates emanating from one nucleus! One can, indeed, calculate from the alveolar surface area and the number of epithelial cells present in young rat lungs that each type 1 cell must cover an area of 16,000 ,.,..2 With an estimated diameter of apical plates (terminal bar rings) of 30 to 50 p., this area would accommodate about ten plates, on the average, but it is possible that there are even more. This would first demonstrate why epithelial nuclei are so few, compared to the large number of cytoplasmic areas with terminal bars demonstrable on silver-impregnated specimens; we must therefore conclude that Kolliker's! observation was well founded. Secondly, this finding establishes an extraordinary complexity for type 1 cells: to my knowledge, no other epithelial cell has had multiple apical areas. The complexity might be compared, at least in principle, to that of neurons. We assumed that type 1 cells possess an unusually high level of differentiation of a very special kind. For cells of simple surface epithelia the apical plate is a region of special differentiation, since it provides the contact surface with the external space; the membrane belonging to this region is clearly distinguished from the basolateral membranes by the junctional complex. The number of such apical plates formed can certainly be considered a measure of the level of differentiation for these cells. We have chosen to call the nature of this diHerentiation topologic for the following reason: If one subjects the cell to a topologic deformation (involving no tearing or cutting) until it is spherical, the apical plates will appear as spotlike surface regions, well separated by basolateral membrane parts (Fig 2). A "normal" squamous cell contains one such region, but a type 1 cell contains multiple spots. To "transform" the cell shown in Figure 2b into the one in Fig 2a (or viceversa), cutting of the surface is necessary. The number of apical plates present is therefore a topologic property of the cell. What influence does this topologic differentiation have on the divisibility of type 1 cells? Here we are left to speculation. The purpose of mitotic division of the cells of a simple epithelial lining is to increase the number of lining cells, or, more specifically, the number of apical faces of the cell membrane. This would mean that the apical plate of the dividing cell must be split into two plates. Schematically, the division plane
2. Topologic analogue to type 1 cell with single ( Q.) and multiple (b) apical plates. Division plane is easily conceived in (Q.), but not in (b). FIGURE
CHEST 65: 4, APRIL, 1974 SUPPLEMENT
should extend vertically from the apical to the basolateral cell surface, a process which seems easily possible in the cuboidal type 2 cell or in the simple squamous type 1 cell (Fig la, Ib, 2a). However, in the topologically more complex type 1 cell (Fig Ic, Id, 2b), it would be very difficult to conceive how a nuclear division could be followed by a cytoplasmic division, which adequately divides the apical and basoalteral cell surfaces. This is rendered even more difficult by the fact that the cytoplasmic flaps of type 1 cells extend over long distances from the nucleus. In contrast, the type 2 cell is predominantly a simple cuboidal cell, although some cells with two or even three apical faces have also been observed, but never with the same level of complexity as the type 1 cells." For the majority of these cells there is, therefore, no difficulty in adequately dividing, whereby the division plane passes across the apical and the basolateral face." Their cytoplasmic diHerentiation does not seem to be an obstacle, since it can be shown that the specific lamellar bodies are retained in mitotic cells and distributed among the two daughter cells. S In conclusion, we have demonstrated that type 2 cells have a high level of cytoplasmic but a low level of topologic diHerentiation, whereas the type 1 cells combine a low level of cytoplasmic with an exceptionally high level of topologic diHerentiation. Considering the process of mitotic cell division, it seems plausible that
type 1 cells should have great difficulty in proper division of its elements of the surface membrane. This could explain the fact that type 1 cells cannot divide, but that damaged type 1 cells are replaced by division and subsequent transdiHerentiation of type 2 cells. If this hypothesis can be proved, it would imply that the level of topologic rather than that of cytoplasmic diHerentiation is the major obstacle to mitotic divisibility.
trations may have produced significant ultrastructural alterations, especially in the mitochondria. I have been evaluating the structural alterations or damage which can be produced in rat liver cells if the tonicity of the fixing solutions is altered (Acta histochem 35: 1-17, 1970). It seems reasonable to assume that similar damage could be induced in pulmonary cells also. If the buDer used with formaldehyde was isotonic or slightly hypertonic, best results were achieved. However, when dilute buffer (0.05-0.1 M) was used with 4 percent formaldehyde or glutaraldehyde preceding the osmium tetroxide postfixation, marked differences, including dilated endoplasmic reticulum, swollen mitochondria and even ruptured mitochondrial membranes were observed in the specimens. This was true even if the tonicity of the fixative brought the osmolarity of the entire solution to isotonic or hypertonic levels. This observation diHers from the usual custom of carefully controlling the tonicity of the buffer and fixative together, ie keeping it nearly isotonic. I found that during the initial fixation, the combined tonicity of buffer and fixative was less critical than the concentration of the buller alone. Postfixation osmium tetroxide could not reverse the damage induced by the earlier use of a dilute buffer. Dr. Weibel: These are critical considerations, about which we have been very concerned also. We measure the osmolarity of each batch of fixative prepared, by the method of freezing point depression, and routinely use an osmolarity of 350 mosm/L. Another point which should be made is that the buffer employed in the fixative is important; we have found, for example, that glutaraldehyde buffered with collidine may produce holes in the tissue sections.
Dr. Stanford: Using the Cowdry terminology, are type 1 alveolar epithelial cells fixed postmitotic cells, or can they, under certain conditions, revert to the status of a dividing cell, similar to hepatocytes? Dr. Weibel: I will refer the question to Dr. Evans, who has done a great deal of work in this area recently. Dr. Evans: We have seen occasional cells with the morphologic features of type 1 cells 'labeled with tritiated thymidine label in young animals, but as yet not in adult animals, even under pathologic conditions. I would like to ask Dr. Weibel a question: What was the regional distribution of the lesions you described in your patient's lungs? Dr. Weibel: The lesions were multifocal, but we could not find any particular relationship to the pulmonary acinus or lobule. Dr. MitcheU: Is cellular proliferation in the pulmonary interstitium characteristic of oxygen toxicity? Dr. Weibel: Not necessarily; it may.bepresent in that and a wide variety of other pathologic processes, as Dr. Stanford mentioned in the introductory remarks this morning, and as I have shown in this series of patients who did not receive high-concentration or longtenn oxygen. Dr. Winter: What was the immediate cause of death in your patients? Dr. Weibel: Acute respiratory failure with impaired gaseous diffusion and consequent systemic hypoxemia. Dr. Paegle: I am concerned that during your presentation and those of several preceding speakers, who performed ultrastructural morphometric studies, relatively little was said about the precautions taken to exclude the possibility that variations in buffer and fixative concen-
CHEST 65: 4, APRIL, 1974 SUPPLEMENT
1 Bachofen M, Weibel ER: Basic pattern of tissue repair in
human lungs following unspecific injury. Chest (this issue) 2 Carrington CB, Green TJ: Granular pneumocytes in early repair of diffuse alveolar injury. Arch Intern Moo 126:464465, 1970 3 Evans MJ, Cabral LJ, Stephens RJ, et al: Renewal of alveolar epithelium in the rat following exposure to N02. Am J Pathol 70:175-198, 1973 4 KauHman SL, Burri PH, Weibel ER: The postnatal growth of the rat lung. Part 2: Autoradiography. (Submitted to Anat Rec) 5 Kolliker A: Zur Kenntnis des Bans der Lunge des Menschen. Verhandl Physiol Moo Ges (Wiirzburg) 16:1,
6 Meyrick B, Reid L: The alveolar brush cell in rat lung-a third pneumocyte. J Ultrastruct Res 23:71-80, 1968 7 Weibel ER: The mystery of "non-nucleated plates" in the alveolar epithelium of the lung explained. Acta Anat (Basel) 78:425-443, 1971
18TH ASPEN LUNG CONFERENCE 215