Published ahead of print on October 9, 2003, doi:10.1164/rccm.200303-371OC
© 2003 American Thoracic Society Human Alveolar Wall Fibroblasts Directly Link Epithelial Type 2 Cells to Capillary EndotheliumiCAPTUR4E Center, McDonald Research Laboratories and Department of Pathology and Laboratory Medicine, University of British Columbia, St. Paul's Hospital, Providence Health Care, Vancouver, British Columbia, Canada Correspondence and requests for reprints should be addressed to David C. Walker, iCAPTUR4E Center, McDonald Research Laboratories, Room 166 Burrard Building, St. Paul's Hospital, 1081 Burrard Street, Vancouver, BC, V6Z 1Y6 Canada. E-mail: dwalker{at}mrl.ubc.ca
Alveolar wall fibroblasts directly link type 2 (T2) pneumocytes to capillary endothelium through apertures in their respective basal laminae in rabbit lung. These fibroblasts provide a bridge from the capillary to the airway lumen along which leukocytes may migrate without disrupting extracellular matrix. Normal human lungs were examined by transmission electron microscopy and serial section 3D reconstruction. We found contacts between fibroblasts and T2 pneumocytes and between fibroblasts and type 1 pneumocytes that occur at holes in the epithelial basal lamina. The same fibroblast also made contact with pericytes and endothelial cells through similar apertures. A survey of 41 T2 pneumocytes revealed that 54% of T2 pneumocytes had at least one gap in their basal lamina. A morphometric analysis showed these gaps occupied approximately 5.58 ± 1.51% (mean ± SE) of the area underneath T2 pneumocytes. We conclude that a population of single fibroblasts link T2 pneumocytes to adjacent capillary endothelial cells in alveolar walls of human lung. We propose that fibroblasts are organized to maintain communication between epithelium and mesenchyme and to provide directional information to migrating leukocytes.
Key Words: leukocyte migration • transmission electron microscopy • 3D reconstruction • heterotypic cellular contacts Previous studies in rabbits from our laboratory showed that a population of fibroblasts link alveolar wall capillary endothelium and type 2 (T2) pneumocytes through apertures in their basal laminae (1). These myofibroblast-like cells contact endothelial cells by means of cytoplasmic extensions that protrude through apertures in the endothelial basal lamina. At the epithelial basal lamina, T2 pneumocyte cytoplasmic extensions protrude through similar apertures to contact and often invaginate these fibroblasts (1). These connecting fibroblasts provide migratory leukocytes with portals to enter and exit the interstitium and a cellular substrate on which to migrate through the interstitium with minimal damage to normal lung architecture. Leukocytes migrating from the pulmonary capillaries into alveolar airspace during an inflammatory response face several important obstacles (2–8). They must first cross the endothelium and its basal lamina, migrate through the interstitium of the alveolar wall, and then pass through the epithelium and its basal lamina to enter the alveolar lumen (8). In rabbits with streptococcal pneumonia, migrating neutrophils use the apertures in the endothelial and epithelial basal laminae to enter and exit the alveolar wall interstitium (1, 9). They also travel through the interstitium on the surface of single fibroblasts that guide them from the endothelial basal lamina aperture to the epithelial basal lamina aperture and cross the epithelium adjacent to T2 pneumocytes to enter the alveolar lumen (4, 10, 11). The purpose of this study was to determine whether or not there are fibroblasts in the alveolar walls of human lung that link the endothelium to T2 pneumocytes through apertures in their respective basal laminae as there are in rabbit lung. To accomplish this we used computer-assisted transmission electron microscopic serial section 3D reconstruction of alveolar wall fibroblasts. Observations made in this study demonstrate that human alveolar wall fibroblasts do link endothelial cells to T2 pneumocytes through basal lamina apertures. Furthermore, there is more diversity in the complexity of the intercellular contacts between the fibroblasts and T2 pneumocytes than observed previously in rabbit lungs.
Specimens Specimens were acquired from lungs entered into St. Paul's Hospital Pulmonary Research Laboratory Lung Registry that has been described elsewhere (12, 13). Approximately 1-cm3 pieces of fresh tissue were taken from macroscopically normal lung parenchyma well away from the tumor and regions of emphysema. All tissue was processed for transmission electron microscopy as before (1, 6).
Qualitative Observations Digital images were captured at x10,000 magnification for all cells recognized as T2 pneumocytes. A total of 41 T2 cells were assessed to estimate the frequency of apertures seen in basal lamina beneath T2 pneumocytes. The numbers of T2 cells assessed per patient ranged from 5 to 11. We counted the number of gaps beneath all 41 T2 pneumocytes, and the frequency of apertures was calculated by dividing the number of gaps in the epithelial basal lamina by the number of T2 pneumocytes. We have assumed that the presence of a gap in the basal lamina represents an aperture. The linear length of the basal lamina beneath the T2 pneumocytes as well as the length on the gaps in the basal lamina was also measured at this time using AMT Advantage Software (Version 5.3.0; Advanced Microscopy Techniques Corp., Danvers, MA). To estimate the relative surface area of basal lamina and apertures, we used a line intercept technique with cycloids (14). The relative surface area of apertures was calculated by dividing the number of intercepts that fell on a basal lamina gap by the total number of intercepts and multiplying by 100. Color segmentation with Image Pro Plus (Media Cybernetics, Silver Spring, MD). allowed us to directly measure the relative surface area of apertures beneath the reconstructed T2 pneumocyte using Figure 1D .
3D Reconstruction A total of 86 serial 0.1-µm thick sections were cut from "normal" tissue and picked up on formvar-coated slot grids. The area of interest was recorded at x3,550 magnification, and each transmission electron microscopic negative was developed according to standard darkroom procedures and scanned at 600 dpi on a transilluminating flatbed scanner (Epson 1680). Adobe Photoshop 6.0 (Adobe Systems Inc., San Jose, CA) was used to trace T2 pneumocytes, type 1 (T1) pneumocytes, fibroblasts, basal laminae, and endothelial cells each in separate grayscales. After the establishment of registry, the program "T3D" (Research Systems, Inc., Boulder, CO) was used for 3D reconstruction as we have done previously for leukocytes in diapedesis (15).
3D Reconstruction Generally, cuboidal T2 pneumocytes were located in depressions created by the reticulum of capillaries of the alveolar wall (Figure 1). Careful inspection of the epithelial surface reveals that several, usually three, T1 pneumocyte borders converge at the perimeter of the T2 pneumocyte (Figure 1). T2 pneumocytes are easily identified topographically by their numerous short microvilli (Figure 1). In more favorable views the T1 pneumocytes may be seen to extend up the lateral sides of the T2 pneumocyte, allowing adequate surface area for the tight junctions. The serial section reconstruction in this study comes from an area like that pictured in Figure 1. In the 3D reconstruction (Figures 2A–2F) , it is clear that the low cuboidal T2 pneumocyte is surrounded by the adjacent squamous T1 pneumocytes (Figures 2A, 2B, and 2E). The cell body of the fibroblast immediately below the T2 pneumocyte has squamous cytoplasmic extensions that partially surround the adjacent capillaries (Figures 2A, 2E, and 2F). The T2 pneumocyte is surrounded by capillary segments that apparently converge at an intersection behind it (Figure 2C). The epithelial basal lamina is present under both types of pneumocytes, however, apertures were observed either under or immediately adjacent to the T2 pneumocyte (Figures 2C–2F). Of the 14 apertures observed in the epithelial basal lamina, 13 were located between only the T2 pneumocyte and the fibroblast; two of these 13 were located at the margin of the T2 pneumocyte (Figure 2D). The fourteenth aperture was found between the fibroblast and an adjacent T1 pneumocyte. In addition, one aperture was observed in the capillary basal lamina along the intersection of the thick and thin walls of that capillary (Figures 2C and 2D), and another aperture was found in the pericyte basal lamina (not shown in the reconstruction because of the orientation). All 16 of these apertures permit direct contacts between the fibroblast and neighboring epithelial cell, endothelial cells, or a pericyte. Therefore, this 3D reconstruction demonstrates that this fibroblast simultaneously is in direct contact with alveolar epithelial cells, the endothelium of the adjacent capillaries, and their pericytes through apertures in the basal laminae of both the epithelium and endothelium.
Intercellular Contacts Around the perimeter of the T2 pneumocyte, a cytoplasmic extension of the T1 pneumocyte extended both laterally up the side of the T2 pneumocyte and basally underneath it. Tight junctions were observed in the lateral space between the two epithelial cells, whereas interdigitations between the two epithelial cells were confined to the basal region of contact (Figures 3A, 3B and Figure 4 , 7). Two apertures in the epithelial basal lamina existed at the basal region on opposite sides of the T2 cell that permitted simultaneous contact between the fibroblast, T1, and the T2 pneumocytes (Figures 3B and 4, 1). The contact was between a T1 pneumocyte cytoplasmic extension that reaches basally beneath the T2 pneumocyte, T2 cell finger-like projections, and the surface of a squamous fibroblast extension. Both of these apertures were located adjacent to tricellular corners in the epithelium, where two different T1 pneumocytes and the T2 pneumocyte meet (Figure 3B). The aperture at the left side of the reconstruction extended from Section 41 to 45, whereas the adjacent tricellular corner extended from Section 59 to 63. At the right of the reconstruction, the aperture extended from Section 58 to 66 and its adjacent tricellular corner from Section 61 to 70. The limits of the tricellular corner were defined by the presence of two profiles of T1 pneumocytes separated by intercellular space adjacent to or under the T2 pneumocyte (Figure 3B). The most frequent type of cell contact observed in this reconstruction was that between the T2 pneumocyte and the fibroblast. In 4 of the 14 epithelial contacts observed, short T2 pneumocyte extensions projected through the epithelial basal lamina aperture into the interstitium and interdigitated with fibroblast extensions that extended up into the epithelium through the same aperture (Figures 3C and 4, 5). In three contacts, numerous T2 pneumocyte extensions contacted the flat surface of a fibroblast extension (Figure 4, 6). Twice, a T2 pneumocyte extension protruded through the epithelial basal lamina aperture and invaginated the fibroblast (Figure 4, 2), whereas one fibroblast extension protruded through an aperture and invaginated the T2 pneumocyte (Figure 4, 3). One contact consisted of only the apposition of the flat surfaces of both the T2 pneumocyte and fibroblast (Figure 4, 4). Figures 3D and 4 (9) show a type of contact in which a T1 pneumocyte cytoplasmic extension projects through an aperture in the epithelial basal lamina to invaginate the surface of the subtending fibroblast. This type of contact always occurred with a T1 pneumocyte immediately adjacent to a T2 pneumocyte. A total of three such contacts were observed in the sections used for this reconstruction, but two of these contacts occurred in association with another T2 pneumocyte and were thus not reconstructed. We estimated that the minimum diameter of this type of aperture is 300 nm because the basal lamina was discontinuous in at least three serial sections. This same fibroblast also made contact with elements of the adjacent capillaries. One fibroblast cytoplasmic extension was in contact with an endothelial cell through a small aperture in the endothelial basal lamina (Figure 4, 10) at the junction of the thin and thick walls of the capillary. Another contact was observed between a pericyte and the fibroblast through an aperture in the pericyte basal lamina (Figure 4, 8). This occurred at the thick wall of the capillary.
Aperture Size and Frequency After calculating the frequency of apertures seen in the T2 pneumocyte basal lamina using the number of gaps and the number of T2 pneumocytes found, we estimate the frequency to be 0.54 ± 0.14. We interpreted this to mean that approximately 5 out of 10 T2 pneumocytes observed should have at least one aperture in its basal lamina. We have counted these parameters from five patients and a total of nine blocks.
Aperture Areas
This study shows that single human alveolar wall fibroblasts connect the endothelium to epithelial T2 pneumocytes across the interstitium. These connections occur through apertures in the basal laminae of both endothelial and epithelial cells. Observations from the serial sections revealed that the single fibroblast was also in simultaneous contact with capillary pericytes and T1 pneumocytes through similar apertures in their basal laminae. In addition, many different types of intercellular contacts were observed between the fibroblast and T2 pneumocyte. These ranged from convergent contacts, where three cells contacted each other through a basal lamina aperture, to simple interdigitating contacts, where single cell extensions protruded through apertures in the basal lamina to invaginate the opposing cell, to more complex interdigitations between many microvilli of the two cells (see summary diagram in Figure 4). The fibroblasts in the human alveolar wall appear to be organized much like they are in the alveolar walls of rabbit lung. These fibroblasts link the entrance to the interstitium from the capillary to an exit from the interstitium at the base of the T2 pneumocyte through apertures in the respective basal laminae (1, 6). In both human and rabbit lung, the tight junctions between T1 and T2 pneumocytes were located on the lateral surfaces of the T2 pneumocytes (6). There are additional features of human fibroblast organization not observed in rabbit. In human lung, interdigitations between T1 and T2 pneumocytes were confined to the basal surface around the perimeter of the T2 pneumocytes, whereas in rabbit lung the interdigitations were confined to the lateral surface of the T2 pneumocytes. In contrast to rabbit, we observed two three-way contacts between the fibroblast, T1, and T2 pneumocytes and each of these were adjacent to a tricellular corner in the epithelium between the T2 and two T1 pneumocytes. Finally, we showed that human pulmonary fibroblasts may contact T1 pneumocytes. It was of particular interest to discover that the two points at which the three cells, T1 and T2 pneumocytes and fibroblast, converged were both at an aperture adjacent to tricellular corners in the alveolar epithelium. We have previously shown that the tight junctions are discontinuous at tricellular corners in the epithelium and endothelium (11, 15–17) and that these corners may provide favored avenues for leukocyte emigration (18). The observation of two T1/T2/fibroblast contacts located adjacent to epithelial tricellular corners is consistent with there being a favored avenue of leukocyte migration (19). The fibroblast we reconstructed was distinguished from pericytes by its prominent rough endoplasmic reticulum, cytoplasmic extensions, and lack of a basal lamina. At least two populations of fibroblasts have been described in mammalian lungs. A population of fibroblasts is oriented parallel to the alveolar wall and is intimately associated with the fibrillar elements of the extracellular matrix (20, 21). Another population of fibroblasts, termed myofibroblasts, are oriented perpendicular to the alveolar epithelium and is described as having bundles of microfilaments believed to have contractile properties (20, 21). It is these myofibroblast-like cells that we describe as linking the epithelial T2 pneumocytes to the endothelium in rabbit lung and here in human lung (1). Epithelial–interstitial cell contacts have previously been demonstrated, and their numbers have been shown to vary with lung development, sex, and repair processes in both rat and mouse lung (22, 23). These observations confirm in animal models the existence of constitutive links between T2 and interstitial cells that we demonstrate here in human lung. In contrast, Brody and Craighead and Brody and coworkers have previously demonstrated T2/fibroblast contacts in fibrotic but not in normal human lungs (24, 25). These contradictory reports of epithelial/interstitial contacts in "normal" human lung may be explained in several ways. For example, our normal lung samples came from the lungs of chronic smokers undergoing a resection for a malignancy and thus may not truly be normal. On the other hand, ultrastructural sampling of point phenomena such as basal lamina apertures is limited by the inherently small size of thin sections. Forty percent of T2 pneumocyte observed in a single thin section revealed no basal lamina apertures. These differences may be reconciled by sampling from nonsmokers. Our results in lung and similar observations in gut and conducting airways (26–30) support the proposition that a population of fibroblasts in the lamina propria of various epithelia may link the epithelium to the endothelium of the adjacent elements of the vasculature through apertures in the basal laminae. It is unlikely that fibroblast contacts with both endothelium and epithelium described here are mechanical (1). There are distinctly separate adherens-like junctions between the fibroblasts and epithelial basal lamina that coexist with the fibroblast/epithelial contacts in lung tissue (1). Distinct mechanical contacts also occur between fibroblasts and the basal laminae under T1 pneumocytes, where the thick and thin walls of capillaries meet (1). Because we did not observe any microfilaments associated with the fibroblast/epithelial/endothelial contacts that are the focus of this study, we suggest that they do not perform a mechanical function. Our previous observations in rabbit, and now human lung, suggest an important role for alveolar fibroblasts in directing leukocyte migration (1). Several studies suggest that leukocytes are attracted to and indeed crawl along fibroblasts during an inflammatory response (26–33). In fact, all leukocytes observed within the interstitium in the present study were in close juxtaposition to fibroblast cell extensions. In summary, by performing a 3D reconstruction of human alveolar wall fibroblasts using transmission electron microscopy, we determined that there is a population of fibroblasts that directly connect the endothelium to epithelial T2 pneumocytes through apertures in the basal laminae. This effectively creates a continuous cellular substrate from the capillary to the alveolar lumen along which leukocytes may migrate. This model we have developed for fibroblast organization in human lung leads us to speculate that disruption of this pathway for leukocytes in diseased lung, such as emphysema, could lead to architectural changes characteristic of that disease.
The authors thank Dr. James C. Hogg for reading this manuscript. They also thank Mr. Stuart Green and Mr. Dean English for their assistance with the photography and imaging in this manuscript.
Supported by a National Institutes of Health Grant HL66569. Conflict of Interest Statement: F.E.S. has no declared conflict of interest; F.S.F.C. has no declared conflict of interest; D.C.W. has no declared conflict of interest. Received in original form March 13, 2003; accepted in final form October 8, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
and interleukin-4 differentially regulate ICAM-1 and VCAM-1 expression on human lung fibroblasts. Eur Respir J 1999;14:759–766.
