Heterogeneity of tight junction morphology in extrapulmonary and intrapulmonary airways of the rat.код для вставкиСкачать
THE ANATOMICAL RECORD 198193-208 (1980) Heterogeneity of Tight Junction Morphology in Extrapulmonary and Intrapulmonary Airways of the Rat EVELINE E. SCHNEEBERGER Department of Pathology, Massachusetts General Hospital and Haruard Medical School, Boston, Massachusetts 02114 ABSTRACT In the present study morphology of tight junctions was related to the various cell types lining extrapulmonary and intrapulmonary airways of the rat. Freeze fracture replicas were prepared from extrapulmonary airway epithelium derived from the cartilagenous and membranous sides of upper, middle, and lower thirds of the trachea. Intrapulmonary airway epithelium was obtained from airways < 1 mm in diameter. Tight junction fibrils on the P fracture face were organized into three types of patterns. Type 1: parallel, sparsely interconnected lumenal fibrils with large ablumenal fibril loops. Type 2: richly interconnected lumenal fibrils with large ablumenal fibril loops. Type 3: narrow network of interconnected fibrils. On the E fracture face complementary grooves were organized in a similar pattern. Ciliated cells on both sides and all levels of the trachea were associated with type 1 junctions. In intrapulmonary airways, however, the junctional pattern of ciliated cells changed to type 2. Brush cells at all levels of the airways were bounded by type 2 and occasionally by type 1 junctions. Secretory cell junctions displayed the following patterns: Mucous cells were bounded solely by type 3, serous cells by either types 2 or 3, and Clara cells predominantly by type 2. Cells tentatively identified as intermediate cells displayed all three junctional patterns. The number of parallel fibrils comprising tight junctions was higher in extrapulmonary as compared to intrapulmonary airways. No difference was seen in the various locations sampled in the trachea. Gap junctions were observed between secretory cells of extrapulmonary but not intrapulmonary airways. These observations are discussed in relation to current physiologic data. Recent evidence suggests that cells comprising the tracheal epithelium elaborate a watery periciliary fluid (sol)which is present beneath the mucus (gel)layer lining the trachea (Nadel et al., '79). One possible function for the low viscosity sol layer is to permit the cilia to beat freely, thereby propelling the overlying mucus layer towards the mouth. Secretion of this watery fluid phase is a consequence of active ion transport into the tracheal lumen (Olver et al., '75, Mangos, '76, Al-Bazzaz and Al-Aquati, '77), and therefore, depends on the presence of a permeability barrier across which an osmotic gradient can be established. Tight junctions between cells, by virtue of their continuous network of strands, provide such a barrier (Staehelin, '74). Correlative studies suggest that the number of horizontally parallel strands comprising the junction determines, in part, the tightness of the intercellular seal that 0003-276X/80/1982-0193$02.900 1980 ALAN R. LISS, INC. is formed (Claude and Goodenough, '73, Claude, '78). On the other hand, the extent to which strands are interconnected with one another may affect the degree to which the junction can be stretched, either as a consequence of an increase in cell size, or of pressure applied to the cells from within the lumen (Hull and Staehelin, '76). A variety of epithelial cells line the mammalian extrapulmonary and intrapulmonary airways; the relative number of each cell type, however, varies depending on the species, and the level of the respiratory tract from which the tissue is taken (Jeffrey and Reid, '75, Breeze and Wheeldon, '77). In the rat these types include mucous, serous, intermediate, basal, brush, ciliated, and Clara cells; the first Received March 10. 19.30: accepted April 2, 1980. 194 E.E. SCHNEEBERGER being sparse in number in the healthy rat and the last being confined to the intrapulmonary airways. While the appearance of freeze-fractured intercellular j unctions has been described in the guinea pig (Inoue and Hogg, '77), the rat (Marin et al., '79), and mouse (Inoue and Richardson, '79), no data are available relating the morphology of tight junctions to either the adjacent cell type or the level of the airway from which the junction is obtained. The present study on the morphology of epithelial tight junctions in extrapulmonary and intrapulmonary airways of normal rats was designed to answer the following questions: 1. What is the morphology of the tight junctions which form between several of the combinations of the epithelial cell types enumerated above? 2. How does the morphology of the junction differ in the extrapulmonary as compared to the intrapulmonary airways? 3. How does the location of a cell on the cartilagenous as opposed to the membranous side of the trachea affect the morphology of its tight junction? MATERIALS AND METHODS Animals The extrapulmonary and intrapulmonary airways of 36 Wistar-Furth rats (Microbiological Associates, Bethesda, Md.) of both sexes, weighing between 70 and 150 grams, were examined. The cages containing the rats were maintained on glass enclosed shelves supplied with filtered air. To ensure that tissue was obtained from animals free of respiratory infections, lungs from randomly selected animals were periodically checked by light microscopy to ascertain that there was no lymphocytic cuffing of the airway epithelium. Tissue preparation The animals were anesthetized by the intraperitoneal injection of sodium pentobarbital (Diabutal, Diamond Laboratories, Inc., Des Moines, Iowa), 5 mgllOO gm body weight. After exposing the trachea and lungs, and allowing the lungs to collapse, formaldehydeglutaraldehyde (FG) fixative diluted 1:2 with 0.1 M cacodylate buffer, pH 7.3 (Karnovsky, '65) was gently instilled into the airways by means of a 3-ml syringe and a 30-gauge needle inserted just cephalad to the larynx. Care was taken not to over distend the lungs with fixative. The trachea, lungs, and heart were removed en bloc. The trachea was then divided into three equal parts: 1. upper, 2. middle, and 3. lower trachea. These tracheal rings together with slices of lung, 1-mm thick, were fixed in FG fixative for an additional 20-45 minutes a t 4"C, and washed overnight in 0.15 M cacodylate buffer, pH 7.3. Light and electron microscopy Tracheal rings, 1-mm thick, and small cubes (1mm3)of lung were postfixed in 1% OsO, with 15 mg/ml of potassium ferrocyanide (Karnovsky, '71) for 1 hour a t 4°C. The tissue was stained en bloc with 1.5% uranyl acetate in 0.05 M maleate buffer, pH 6.2, dehydrated in graded ethanols, infiltrated, and embedded in Epon. Sections 1-pmthick from the three levels of the trachea and the lung were stained with toluidine blue and examined by light microscopy. For electron microscopy, thin sections were cut with a diamond knife (BalzersUnion, Balzers, Lichtenstein) on an LKB Ultrotome I11 (LKB, Bromma, Sweden) picked up on carbon-coated grids, stained with lead citrate (Venable and Coggeshall, '69), and examined in a Philips EM 300 electron microscope. Freeze fracture Small pieces of fixed tissue were taken from lung containing airways less than 1.0 mm in diameter (intrapulmonary airways) and from the cartilagenous or membranous portion of the trachea, (extrapulmonary airway) and were infiltrated with glycerol at concentrations increasing from 10 to 30% in 0.1 M cacodylate buffer, pH 7.3 for 2 hours at 4°C. The tissue was then rapidly frozen in liquid nitrogen cooled to -210°C under vacuum, and fractured in a double replica device at -150°C, using a Balzers high-vacuum freeze-etch unit. The carbon-platinum replicas were washed in 5.2% sodium hypochlorite (Chlorox Co., Oakland, California) followed by distilled water, picked up on Formvar-coated grids, and examined in the electron microscope. Measurements In each electron micrograph of the freeze fracture replica, the cell on which the junction was localized was identified by morphological criteria which are described in Results. All measurements were made on elecron micrographs at a final magnification of x 50,000. A grid of 0.25 pm spacings was placed on the micrographs so that the grid lines were perpendicular to the longitudinal axis of the tight junction network. The number of fibrils intersected by each grid line was counted and the total length of junction examined was measured. Care was taken to count only those por- TIGHT JUNCTIONS IN AIRWAYS OF THE RAT tions of the junctions which appeared complete in the fracture plane of the replica. The mean, mode, and range of the fibril counts were determined. RESULTS Light microscopy Examination of sections 1 pm thick, stained with toludine blue, revealed neither lymphocytic cuffing beneath the tracheal epithelium nor an inflammatory cell infiltrate in the epithelium of extrapulmonary and intrapulmonary airways of randomly selected animals. Ciliated cells, staining weakily with toluidine blue, stood in sharp contrast to the other more darkly staining epithelial cell types. I t was possible to distinguish mucous from serous epithelial cells by virtue of the fact that the secretory granules of the former were larger and stained less intensely than those of the latter cell. Mucous cells were sparse in number and rarely observed caudal to the midportion of the trachea. Brush and intermediate cells could not be identified with certainty at the light microscope level. In the intrapulmonary airways Clara cells were easily distinguished by their bulbous apices which projected into the lumen of the airway above the level of the adjacent ciliated cells. The latter cell type was more numerous in the intrapulmonary than in the extrapulmonary airways. 195 Secretory cells. The two secretory cell types present in the trachea (mucous and serous epithelial cells) (Fig. l a ) are distinguished from one another by the morphology of their secretory vacuoles: Those of the serous epithelial cells tend to be smaller (up to 0.6 pm in diameter) and contain a homogeneous, moderately electron dense matrix; those of mucous cells are larger (up to 0.8 pm in diameter), and contain a rounded dense core within a weakly electron dense matrix. The accumulation of such vacuoles within mucous cells results in the outward bulging of the apex and sides of the cell, a change which is less apt to occur in the serous cells of the normal rat. This feature together with the larger size of the secretory vacuoles, and the presence of mucus cells primarily in the proximal half of the trachea, are useful in identifying these cells in freeze fracture replicas. Both these secretory cells have on their lumenal surface a sparse array of short, blunt microvilli, measuring 0.3-0.6 pm in length and 0.08-0.1 pm in width. Their basolateral membranes, particularly where they are in contact with basal cells, form interdigitations with adjacent cells. A large amount ot rough endoplasmic reticulum is present on the ablumenal side of the cell and moderate numbers of mitochondria are present throughout the cytoplasm. While the function of Clara cells remains to be established, it is thought to be secretory in nature (Kuhn et al., '74). In contrast to the cells of the trachea, Clara cells have few secretory vacuoles, moderate numbers of mitochondria, and abundant smooth endoplasmic reticulum. Their cell borders are smooth and show few interdigitations. The most distinctive morphological feature of the Clara cell which is helpful in its identification of freeze fracture is a bulbous apex which projects into the airway lumen and has few surface microvilli (Fig. 5a). Transmission electron microscopy While some of the ultrastructural features of epithelial cells lining the rat extrapulmonary and intrapulmonary airways have recently been described (Kuhn et al., '74, Jeffrey and Reid, '75, Marin et al., '79), little attention has been focused on those ultrastructural features which have been correlated with fluid secretion in other epithelia (Oschman and BerIntermediate cells usually have no distincridge, '71, DiBona and Mills, '79). These in- tive cytoplasmic features. However, occasionalclude, for example, cell shape, intercellular ly either a small secretory vacuole or a 'fibrinojunctions, surface microvilli, and basolateral granular' accumulation indicative of ciliogenmembrane infolds. Since the present study esis may be observed, suggesting that these deals primarily with the freeze fracture ap- cells may be precursors of either secretory or pearance of intercellular junctions as it per- ciliated epithelial cells. The surface microvilli, tains to the specific cell types lining the basolateral membranes, and numbers of mitoextrapulmonary and intrapulmonary airways, chondria are similar to those described for a brief overview of the cellular morphology is secretory cells (Fig. la). given. Comments will be made with regard to Brush cells. These distinctive cells are few in those structural features thought to play a role in electrolyte and water transport, as well as number, but are distributed throughout the those found helpful in identifying specific cell length and circumference of the extrapulmonary and intrapulmonary airways. They have a types in freeze fracture replicas. 196 E.E. SCHNEEBERGER Fig. 1.a. Epithelium obtained from the upper third of the rat trachea. A serous cell (SC). ciliated cell (CC). basal cell (BC), intermediate cell (IC).and a mucous cell (MCI are indicated. Mag. x 9,560. b. Portion of a brush cell present in the upper third of the trachea. Bundles of microfilaments (asterisk)extend upward into the surface microvilli. The tight junctions are indicated (arrow heads). Mag. x 15.120. c,d,e.f. Tight junctions between ciliated and intermediate cells (c). mucous and mucus cells (d), ciliated and mucous cells (e),serous and serous cells (f). Areas of close membrane apposition are indicated by short lines. Mag. x 99,000 TIGHT JUNCTIONS I N AIRWAYS OF THE RAT characteristic array of straight, stiff-appearing surface microvilli, measuring approximately 2 pm in length and 0.2 pm in width, which are the most useful feature in their identification in freeze fracture replicas (Fig. lb). Dense bundles of microfilaments extend from the underlying cytoplasm into each microvillus. A moderate number of mitochondria are present, and the basolateral cell membranes show frequent areas of interdigitation with adjacent cells. Ciliated cells. The cells increase in number towards the intrapulmonary airways, and are readily identified by their numerous apical cilia which contain the characteristic array of nine outer doublet and two inner microtubules. Surrounding each cilium are numerous microvilli measuring up to 1.1 pm in length and 0.08O.lpm in width. The basolateral membranes are relatively straight. While the number of mitochondria is similar to that of the cell types described above, the electron density of the cytoplasm is consistently less (Fig. la). Junctional complexes. All the above cell types are joined by junctional complexes which, by transmission electron microscopy, are similar in appearance, but vary in the width of the tight junction band, and the prevalence of desmosomes. The junctional complexes consist of an apical tight junction which varies in width from 0.12-0.52 pm, and displays varying numbers of membrane contacts in which the extracellular space is apparently obliterated (Figs. lc,d,e,f). Distal to the tight junction an intermediate junction is often present. Desmosomes are infrequent in the lumenal junctional complexes, except for those adjacent to brush cells where several may be observed in a single plane of section. Freeze fracture Technical considerations. Fixation of airway epithelium from 20-45 minutes did not affect the length of tight junctional areas exposed by the freeze fracturingprocess. However, shorter fixation times resulted in the tight junction fibrils being particulate and caused them to partition onto both P and E fracture faces. Similar observations have been made in fetal lamb lungs (Schneeberger e t al., '78). Conversely, a longer fixation times gave rise to continuous fibrils on the P face and empty, complementary grooves on the E face. These observations support the findings of Van Deurs and Luft ('79), and emphasize the importance of examining complementary replicas of the same junction before drawing 197 any conclusions as to its integrity. Approximately 108 double replicas were examined and these yielded a total of 116 freeze-fractured junctional areas which could be catergorized as to the cell types with which they were associated (Table 1). Classification of tight junctions. In both the extra (trachea)pulmonary and intrapulmonary airways (bronchioles)essentially three geometrical patterns of tight junction networks are observed (Fig. 2). The first (type 1)consists of sparsely interconnected, but fairly closely spaced parallel fibrils (P face) along the lumenal portion of the junction. This arrangement of junctional elements forms oblong, interfibril compartments, the long axes of which are parallel to the lumenal surface of the cell. On the ablumenal side of the junction, the fibrils form large, irregular loops with the ends of some of the fibrils not connected to the junctional network (Fig. 2). Mirror-image replicas show complementary grooves containing a variable number of junctional particles (E face) (Figs. 3a,b). The second (type 2) consists of a richly interconnected network of fibrils (Pface) on the lumenal side of the junction and irregular large loops of fibrils on the ablumenal side (Figs. 2,3c).As in the type 1junctions, some of the ablumenal fibrils end blindly, and mirrorimage replicas show complementary grooves containing a variable number of junctional particles (E face). The third (type 3) consists of a narrow band of richly interconnected fibrils (P face) and complementary grooves (E face) with no large ablumenal loops (Figs. 2,4,a,b). In both types 2 and 3 junctions the interfibril compartments are smaller and more irregular than in type 1,and they do not show the parallel alignment with the lumenal surface. Distribution of types of tight junctions i n pulmonary airways. Ciliated cells on both the cartilagenous and membranous side, at all levels of the trachea are bound by type 1 tight junctions- that is, junctions with poorly connected parallel lumenal fibrils and large ablumenal loops (Fig. 3a,b, Table 1).While the mean number of fibrils is slightly less on the cartilagenous than on the membranous side of the trachea, the difference is not statistically significant (Table 2). The mode and range of fibril numbers of ciliated cells in both locations of the trachea are also similar. In the intrapulmonary portions of the airways, at the level of the bronchioles, however, the geometrical pattern of the tight junctions between ciliated E.E. SCHNEEBERGER 198 1 2 Fig, 2. 'Racings of the three junctional patterns observed. Type 1, present on tracheal ciliated cells. consists of parallel, sparsely interconnected lumenal fibrils and large ablumenal loops. Q p e 2, present on serous, intermediate, and brush cells in the trachea and on ciliated, Clara, and brush cells in bronchioles, consists of a richly interconnected network of lumenal fibrils and large ablumenal loops. Q p e 3, present on tracheal mucous and serous cells, consists of a narrow interconnected network of fibrils. Fig.S.a,b.c. Complementary double replica (a,b)of a tight junction between ciliated cells in the mid-portion of the trachea. The junction consists on the P face of parallel, sparsely interconnected,lumenal fibrils which form large hops on the ablumenal side (b), and complementary grooves on the E. face (a). Some of the ablumenal fibrilsigrooves end blindly (small arrow). Mag. x 50.000. c. Freeze fracture replica of a tight junction between ciliated cells in a bronchiole. In contrast to the tight junctions in the trachea, the lumenal fibrils form a richly interconnected network. The ablumenal fibrils form large loops, and some end blindly (arrows).Mag. x 50,000. TIGHT JUNCTIONS IN AIRWAYS OF THE RAT Figure 3 199 200 E.E. SCHNEEBERGER cells changes to that of type 2 (Fig. 3c). Rarely a type 3 pattern may be observed (Table 1). Moreover, the mean number of fibrils is slightly reduced (0.02 < P < 0.05) and the lower limit of the range of fibril numbers is smaller than in the trachea (Table 2). In both extrapulmonary and intrapulmonary airways the tight junctions of brush cells are most frequently of the type 2 variety (Table 1, Fig. 4d). Unlike the ciliated cells, the mean number of fibrils is not significantly reduced (Table 2) in the intrapulmonary airways and in none of the brush cells examined was a type 3 pattern observed. Brush cells were observed in replicas from both the membranous and carti- TABLE 1. Frequency distribution Level in airway Trachea3 Bronchioles ~~ ~ lagenous side of the trachea; however, only those from the cartilagenous side yielded fracture planes through the tight junction areas (Table 2). Among the secretory cells (mucous, serous, Clara cells) of the extrapulmonary and intrapulmonary airways, essentially two patterns of tight junctions are present: type 2 and type 3 (Table 1).Except for some serous cells in the membraneous portion of the trachea, the type 1 pattern is not observed. Serous cells at all levels of the trachea may have either a type 2 or type 3 pattern with the latter predominating (Table 1,Fig. 4b). Mucous cells areconsistently bounded by type 3 junctions having relatively of the three types of tight junction patterns in extrapulmonary and intrapulmonary airways Percent of tight junction type2 5Pe2 5Pe3 Cell type' 5pe 1 Nonsecretory: ciliated, brush cells (29) Secretory: serous, mucous cells (25) Nonciliated: intermedate cells (15) 90 4 7 104 21 40 Nonsecretory: ciliated, brush cells (34) Secretary: Clara cells (11) Nonciliated: intermediate cells (3) 6 - a5 73 1005 - 75 53 9 27 - ~ 'The information in this table is derived from the same cells a s are listed in Table 2. The numbers in parentheses indicate the number of tight junctions examined in each cell lype. "@pe 1 junction: parallel poorly interconnected lumenal fibrils and ablumenal loops. Type 2 junction: interconnected luinenal fibril network and ablumenal loops. Q p e 3 junction: interconnected network of fibrils, no ablumcnal loops. 'The cells examined are derived from both the membranous and cartilagenous portions of the trachea a t levels 1, 2. 3. 'The junctions of three out of the four brush cells examined were of the type 2 pattern. All the junctions of the ciliated cells were of type 1. 'Too few junctional areas in nonciliated cells were exposed, in the replicas examined, t o state with assurance whether any of the other types of junctions were present. TABLE 2. Tight junctions in extrapulmonary and intrapulmonary airways of the rat ~ Fibril number Area in Airway Cell type2 Mean f S.E. Membranous portion of trachea a t levels 1, 2. 3 Ciliated cells (11) Serous cells ( 12) Nonciliated cells (lo)* Mucous cells (3) 7.56 6.88 6.07 4.80 f 0.24 f 0.29 Cartilagenous portion of trachea a t levels 1, 2, 3 Ciliated cells (14) Serous cells (9) Nonciliated cells (5) Brush cells (4)3 Bronchioles Cliliated cells (31) Clara cells (11) Nonciliated cells (3) Brush cells 13) 6.50 0.11 5.96 f 0.23 6.00 0.44 6.43 f 0.44 Mode Range Length of junction measured (pm) i 0.35 6 6.7 6 4,5 5-12 4-10 3-9 3-8 7.30 f 0.13 6.28 0.25 7.10 f 0.26 6.88 f 0.30 7 6 6 6.7 5-10 4-9 5-11 5-9 22.8 9.7 6 5 4 56 3-12 2-11 4-10 3-11 52.1 20.5 4.0 5.6 f 0.18 * * * 16.3 1.7 13.6 4.8 7.4 4.0 'The absence of d a t a from a particular cell type in either the membranous or cartilagenous portion of the trachea indicates t h a t no tight junction of that cell type was present in the replicas examined. However. all five cell types were present The number in parenthesis indicates the number of junctions measured in each cell type. 'Nonciliated cells are those in whirh the fracture plane passed solely through the lateral membrane. They could, therefore. represent either serous. niucous. or intermediate cells. 'llrush cells were also present in the replicas from the inenibranous portion of the trachea. hut unlike those of the cartilagenous portion of thc trachea. none nf the fracture planes passed through the Light junctinn region. TIGHT JUNCTIONS IN AIRWAYS OF THE RAT few fibrils and no ablumenal loops (Table 2, Fig. 4a). Clara cells, on the other hand, are bounded primarily by type 2 junctions (Fig. 5c,d). As in the case of ciliated cells in extrapulmonary and intrapulmonary locations, the mean number of fibrils comprising the tight junctions of serous cells is somewhat greater than Clara cells (0.02 < p < 0.05). Cells designated as nonciliated include those in which the fracture plane passes along the lateral cell membrane without entering the underlying cytoplasm. This makes it impossible to determine whether these cells are serous, intermediate, or possibly mucous cells which have recently discharged their secretory vacuoles. In the trachea all three types of junctional patterns are observed in these nonciliated cells; however, types 2 and 3 predominate (Fig. 4c, Table 1). Too few nonciliated cells were available in the replicas from intrapulmonary airways to say with assurance whether types other than 2 are present. Distribution of gap junctions and rectilinear arrays. While the primary purpose of this study is an analysis of the morphology of epithelial tight junctions in the airways of the rat, observations are also presented with regard to the presence and location of gap junctions in the freeze fracture replicas examined. The gap junctions observed are present in replicas derived from the extrapulmonary airways and are associated with serous, mucus, and possibly intermediate cells (Fig. 5b). They are usually present towards the basal portion of the cell, although an occasional gap junction is observed on the lumenal half of the lateral cell membrane. They vary in maximum diameter from 0.30 to 0.91 pm, and more than one gap junction per cell may be observed. The location of gap junctions near the base of the secretory cells suggests that some of these junctions may serve to connect secretory to basal cells. Although they were not present in any of the ciliated cells examined, rare gap junctions have been reported to be present near the lumenal half of ciliated cells in the rat (Marinet al., '79). Insufficient numbers of Clara cells were examined to make a definitive statement as to the presence of gap junctions in these cells; however, none were observed in the replicas obtained of these cells. Rectilinear arrays of intramembranous particules, such as those described in the cells of the guinea pig trachea (Inoue and Hogg, '77), were seen in the lateral membranes of only a single Clara cell. 20 1 freeze fracture replicas prepared for the present study, the epithelial tight junctions at the point of transition between respiratory bronchioles and alveolar ducts were also examined. I t is in this region that Macklin ('55) has suggested that pulmonary sumps are located beneath the epithelium. Furthermore, it has been postulated that these sumps may be anatomical sites into which excess alveolar fluid can drain (Macklin, '55) or from which alveolar flooding occurs in a retrograde fashion (Staub, '79). I t is, therefore, of interest to know what the structure of the tight junctions is in this region of the pulmonary epithelium. Only a single such tight junction (Fig. 6) was apparent in the freeze fracture replicas examined, and its structure resembles that of the junctions observed between type 1 pneumocytes (Schneeberger and Karnovsky, '76) rather than that of junctions present between epithelial cells lining the respiratory bronchioles. DISCUSSION The present study shows that in the epithelium lining extrapulmonary and intrapulmonary airways, the morphology of tight junctions varies depending on the cell type involved and its level in the tracheobronchial tree, but not on its location on the membranous or cartilagenous side of the trachea. The junctions are organized into three geometrical patterns. w p e 1, consisting of the P fracture face of poorly interconnected parallel fibrils and large ablumenal loops, is primarily limited to ciliated cells and occasionally to some brush cells at all levels and on all sides of the extrapulmonary airway (trachea).Q p e 2, consisting on the P fracture face of an interconnected network of lumenal fibrils and large ablumenal loops, is the predominant pattern for ciliated, Clara, and brush cells in intrapulmonary airways (bronchioles), but is also seen on some serous and nonciliated (intermediate ?) cells in the trachea. Finally, type 3, consisting of a narrow band of interconnected fibrils on the P face with no ablumenal loops, is present primarily on mucous and serous cells in the trachea. The factors governing the morphological organization of tight junctions are at present unclear. I t is of interest however, that the greatest uniformity of tight junction pattern is observed among ciliated cells which are the most common cell type in the airways (Breeze and Wheeldon, '77) and are, therefore, the most likely to form tight junctions with each other. A quantitative assessment of the impermeTight junctions of transition between respir- ability of tight junctions to solutes is obtained atory bronchioles and alveolar ducts. In the by the measurement of the electrical resistance 202 E.E. SCHNEEBERGER Figure 4 TIGHT JUNCTIONS I N AIRWAYS OF THE RAT 203 Fig, 4. a. Tight junction (P face) on the surface of a mucous cell. The junction consists of a relatively narrow network of interconnected fibrils. Rarely a fibril ends blindly on the ablumenal side. Mag. x 62.000. b. Tight junction (Pface) on the surface of a serous cell. The junction consists of a network of interconnected fibrils with some slightly larger ablumenal compartments. Mag x 64.000. c. Tight junction (E face)from a presumptive intermediate cell. The junction consists of lumenal parallel grooves which show relatively few interconnections and large ablumenal loops. Mag. x 63,000. d. Tight junction (mostly E face) of a brush cell. The junction consists of an interconnected network of lumenal grooves with some larger ablumenal loops. The closely spaced surface microvilli are seen on the left. Mag. x 42.000. 204 E.E. SCHNEEBERGER Figure 5 TIGHT JUNCTIONS I N AIRWAYS OF THE RAT 205 Fig. 5.a. Clara cell (Cl)and ciliated cells (CC)from a respiratory bronchiole. The Clara cell is characterized by it bulbous apex which often contains a moderate number of secretory vacuoles. Mag. x 4,600. b. Two gap junctions [G)present on the basolateral membrane of a secretory cell in the midportion of the trachea. Mag. x 67,000. c. Freeze fracture replica (Pand E faces) of the apical portion of a Clara cell in a bronchiole. The apical, bulbous portion of the cell extends into the lumen (L)above the level of the tight junction. A portion of the junction on the right hand side (within rectangle) is enlarged in Figure 5d. Mag. x 16,400. d. Enlargement of a portion of the Clara cell tight junction shown in Figure 5c. The P face (P)is on the left and the E face (El is on the right. Mag. x 44.000. 206 E.E. SCHNEEBERGER Fig. 6. Tight junction between epithelial cells at the bronchoalveolar junction. A type I (I)cell is on the right and a cuboidal epithelial cell (C) lining the bronchiole is on the left. The tight junction forms on interconnected network of fibrils (arrow) which resembles the tight junctions usually observed between type I cells. The air space (AS) and capillary Iumen (CL) are shown. Mag. x 51.300. across the epithelium (Fromter and Diamond, '72), and a rough correlation appears to exist between the electrical resistance and the number of sealing fibirls comprising the junction (Claude and Goodenough, '73, Claude, '78). In airway epithelium of several species an electrical resistance of about 300 Q cm2 has been measured (Nadel et al., '79), putting it in the category of a moderately tight epithelium (Schultz, '77). While it is recognized that junctional permeability may be governed by factors other than fibril numbers (MartinezPaloma and Erlij, '75, Dziegielewska et al., '79, Hainau et al., '79), the present morphological study suggest that the airway epithelium of the rat may be expected to be moderately tight as well. A consistent reduction in the mean number of fibrils comprising all tight junctions was observed in the intrapulmonary as compared to the extrapulmonary airway epithelium, suggesting the possibility that the latter may be less permeable than the former. Similarly, the number of fibrils constituting the junctions of secretory cells is consistently less than of ciliated cells. Recent studies have shown that tracheal epithelium actively transports C1- towards the lumen and Na' towards the submucosa. While the Na-K-ATPase has been localized to the basolateral membranes of tracheal epithelial cells (Widdicombe et al., '79), the anatomic location from which C1- is secreted has not been established. Furthermore, it is not known whether C1- secretion is a property of all cells comprising the tracheal epithelium. The present study shows that most of the cells lining the extrapulmonary and intrapulmonary airways are endowed with some surface microvilli of varying dimensions, and their basolateral membranes display a moderate degree of interdigitation with adjacent cells thereby forming tortuous intercellular spaces; these are all features that have been observed in epithelia involved in isotonic fluid secretion (DiBona 207 TIGHT JUNCTIONS IN AIRWAYS O F THE RAT and Mills, '79). Of the tight junctions associated with the various cells lining the extrapulmonary and intrapulmonary airways, those associated with the mucous cells have the fewest fibrils and might be expected to provide the least resistance to the passage of small water soluble molecules. However, in the normal rat these cells are few in number and are limited to the proximal portions of the trachea. Since the secretion of the watery periciliary fluid appears to occur throughout the airways, it is likely that C1- transport is not limited to a single cell type. In their studies of the tight junction morphology in the alimentary tract of Xenopus laevis, Hull and Staehelin ('76) suggested that tight junctions with parallel, poorly interconnected networks of fibrils are flexible and can be stretched more readily than those with richly interconnected networks. Interestingly in Xenopus laevis the flexible type of junctions are associated with mucus cells and the junctions rich in interconnections are seen on ciliated and absorptive cells. The reverse appears to be the case for cells in the mammalian airways. Not only are the poorly interconnected tight junctions limited largely to ciliated cells in the trachea, but the fibrils of these junctions become more richly interconnected in the distal airways, where physiological evidence suggests that the airways are more distensible (Martin and Proctor, '58, Burnard et al., '65). Furthermore, it might be expected that the membranous part of the trachea is more likely to be subjected to stretching than the cartilagenous side, and yet the tight junctions show similar, poorly interconnected fibrils in both locations. I t remains to be established to what extent the degree of fibril interconnection plays a role either in the permeability properties or distensibility of tight junctions. ACKNOWLEDGMENTS The expert technical help of Mrs. Joanne McCormack and Mssrs. Bruce Kaynor and William Flaherty and the critical review of the manuscript by Dr. R.D. Lynch are gratefully acknowledged. This study was supported by USPHS grant HL-25822. LITERATURE CITED Al-Bazzaz. F.. and Q. Al-Aqwati (1977) Characteristics of ion transport in canine tracheal epithelia. Fed. Proc., .76:479 (Abstract).. Breeze. R.G., and E.R. Wheeldon (19771The cells of the Dulmonary airways. Am. Rev. Resp. Dis., Zl6:705-777. Burnard, E.D., P. Gratton-Smith, C.G. Picton-Warlow, and A. Grauaug (1965)Aust. Pediat. J., 1:12-38. Claude, P., and D. Goodenough (1973) Fracture faces of zonulae occludentes from "tight" and "leaky" epithelium. J. Cell Biol.. 58:390-400. Claude, P. (1978)Morphological factors influencing transepithelial permeability: A model for the resistance of the ronida occludens. J. Membr. Biol.. 39219-232. DiBona, D.R., and J.W. Mills (1979) Distribution of Na'pump sites in transporting epithelia. Fed. Proc., 38:134143. Dziegielewska. K.M.. C.AN. Evans, D.H. Malinowska, K. Mollgard, J.H. Reynolds, M.L. Reynolds, and N.H. Saunders (1979) Studies of the development of brain barrier systems to lipid insoluble molecules in fetal sheep. J. Physiol. 292:207-231. Fromter, E.. and J. Diamond (1972) Route of passive ion permeation in epithelia. Nat. New Hiol., 235:9-13. Hainau, B.. C.J. Bentzel, U. IIegel. M. Fromm. and E.L. Benedetti (1979)Cytoplasmic regulation of tight junction permeability in a leaky epithelium: A morphologic electrophysiolgic study. Ultrastruct. Res., 69:145-146 (Abstract). Hull. B.. and A. Staehelin (1976) Functional significance of the variations in the geometrical organization of tight junction networks. J. Cell Biol.. 68:688-704. Inoue, S., and J.C. Hogg (1977)Freeze-etch study of the tracheal epithelium of normal guinea pigs with particular reference to intercellular junctions. J. Ultrastruc. Hes., 61:89-99. Inoue. S.,and J.G. Richardson (1979) A correlated thin section and freeze fracture study of mouse tracheal epithelium before and after ether anesthesia. Lab. Invest., 40:583-586. Jeffrey, P.K.. and L. Reid 11975) New observations of rat airway epithelium: A quantitative and electron microscopic study. J. Anat., 120:295-320. Karnovsky, M.J. (1965)A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy. J. Cell Biol., 27:137 (Abstract).. Karnovsky, M.J. 11971)Use of ferrocyanide reduced osmium tetroxide in electron microscopy. 11t h Annual Meeting of Am. Soc. Cell Biol. New Orleans. LA. p. 146 (Abstract). Kuhn, C. 111. L.A. Callaway. and F.B. Askin (1974) The formation of granules in the bronchiolar Clara cells of the rat. J. Ultrastruct. Res., 49:387-400. Macklin. C.C. (19551 Pulmonary sumps, dust accuniulations, alveolar fluid and lymph vessels. Acta. Anat., 23:l-33. Mangos, d.A. (1976) Fluxes of electrolytes and water in rat trachea. Cystic Fibrosis Club Abstracts, 17th Annual Meeting, St. Louis, MO. p. 11 (Abstract). Marin, M.L.. R.P. Lane. R.E. Gordon, and E. Drummond (1979) Ultrastructure of rat tracheal epithelium. Lunr, 156i223-236. Martin, H.B.. and D.F. Proctor (1958) Pressure-volume measurements on dog bronchi. J. Appl. Physiol., 1.7: 337-343. Martinez-Paloma, A,, and D. Erlij (1975) Structure of tight junctions in epithelia with different permeability. Proc. Nat. Acad. Sci. (USA), 724487-4491. Nadel, J.A., B. Davis, and R.J. Phipps (1979)Control of mucus secretion and ion transport in airways. Ann. Rev. Phvsiol., 42:369-381. Olver, R.E.. B. Davis, M.G. Marin, and J.A. Nadel (1975) Active transport of Na* and C1- across the canine tracheal epithelial in uitro. Am. Rev. Resp. Dis.. 122:811-815. Oschman, J.L.. and M.J. Berridge (1971) The structural basis of fluid secretion. Fed. Proc.. .70:49-56. Schneebereer. - . E.E.. and M.J. Karnovskv (19761 Substructure of intercellular junctions in freeze-fractured alveolar~~ 208 E.E. SCHNEEBERGER capillary membranes of mouse lung. Circul. Res., .38;404411. Schneeberger, E.E., D.V. Waiters, and H.E. Olver 11978)Development of intercellular junctions in the pulmonary epithelium of the foetal lamb. J. Cell Sci..3 2 3 - 3 2 4 . Schultz. S.G. (1977)The role of paracellular pathways in isotonic fluid transport. Yale J. Biol. Med.. 50:99-113. St'aehelin A. (1974) Structure and function of intercellular junctions. Int. Rev. Cytol., .39191-283. Staub, N.C. (1979) Pathways for fluid and solute fluxes in pulmonary edema. In: Pulmonary Edema, A.P. Fishman and E.M. Renkin eds. American Physiological Society, Bethesda. Md. p. 113-124. Van Deurs, B., and J.H. Luft (1979) Effects of glutaraldehvde fixation on the structure of tight junctions, J . Ultrastruct. Res. 68:160-172. Venable, J.H., and R. Coggeshall (1969) A simplified lead citrate stain for use in electron microscopy. J. Cell Biol.. 25;407. Widdicombe, J.H.. C.B. Bashaum. and J.Y. Yee (1979)Localization of Na pumps in the tracheal epithelium of the dog. J. Cell Biol.. 82:380-390.