Introduction

Top Abstract Introduction Materials and methods Results Discussion References

It has been shown that the interactions between the epithelium and mesenchyme play crucial roles in each sequential stage of organ development, including organogenesis, growth, morphogenesis, and cytodifferentiation (1). These interactions have been extensively investigated on a variety of organs, for example, the skin (6), kidneys (9, 10), lungs (11), and digestive system (12, 13). In addition, there is a growing body of evidence that interactions between epithelial cells and mesenchymal fibroblasts are strongly involved in morphogenesis (14). Therefore, fibroblasts are capable not only of producing components of the extracellular matrix (e.g., collagen, fibronectin, and laminin), but also of secreting a number of growth factors (e.g., basic fibroblast growth factor, epidermal growth factor [EGF] and keratinocyte growth factor), which could modulate proliferation and differentiation of the epithelial cells (18). On the other hand, it has been shown that epithelial cells can regulate the proliferation of fibroblasts (26, 27). To investigate the interactions between the epithelium and mesenchyme, numerous in vitro three-dimensional culture systems have been developed (28), with the results indicating that co-culturing with fibroblasts induces successful reconstitution of the epithelium. For example, the reconstruction of the epithelium of the urinary bladder (28), differentiation of cryptlike gut epithelial cells (29), development of early airway gland (30), and tubular morphogenesis of Mardin-Darby canine kidney epithelial cells (31) have all been reported. However, little is known about the role of the interaction between epithelial and mesenchymal cells during development of the tracheal epithelium.

Recently, we have reported on the establishment of a unique culture system using a human amnion membrane (32). Using this system, ciliated epithelial cells were differentiated and their frequency was almost equivalent to that seen in vivo. However, it should be noted that the epithelial cells formed a simple cuboidal epithelium and that the other types of epithelial cells, such as goblet and basal cells, were not observed. In this study, to investigate the role of epithelial-mesenchymal interaction during the development of the trachea, we used the same system to co-culture tracheal epithelial cells with tracheal fibroblasts. The results indicate that co-culturing of the two types of cells leads not only to formation of a pseudostratified columnar epithelium, but also to differentiation of goblet-like and basal cells. Moreover, the frequency of each type of cell was almost identical to that seen in situ. Thus, we have developed an in vitro reconstitution of tracheal epithelia.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Isolation of the Human Amnion

The human amnion membrane was prepared as described previously (32). In brief, the amnion was peeled away from the chorion of a normal term placenta obtained immediately after delivery, and immersed in 0.25 M NH₄OH. It was then rinsed twice with phosphate-buffered saline (PBS) containing a mixture of antibiotics: 2 mg/ml aminobenzyl penicillin, 400 µg/ml minomycin, and 200 µg/ml amphotericin B (Sigma, St. Louis, MO). The epithelial layer of each amnion was scraped off and placed within the tissue-holding device of the two-compartment apparatus under aseptic conditions. This tissue-holding device was composed of two concentric polycarbonate rings, each having an outside diameter of 30 mm and an inside diameter of 14 mm. With its epithelial side facing the upper ring, the amnion was stretched over the bottom of the upper ring and placed between the two rings. These devices were washed with PBS containing the above mixture of antibiotics to remove debris. They were stored in Dulbecco's modified Eagle's medium (DMEM)/F12 (50/50) (GIBCO-BRL, Grand Island, NY) containing 100 U/ml penicillin, 100 µg/ml streptomycin (BioWhittaker, Walkersville, MD), 2.5 µg/ml of amphotericin B, and 50 µg/ml gentamycin (Wako Pure Chemicals, Ltd., Osaka, Japan) at 4°C until use.

Isolation of Epithelial Cells from Guinea Pig Tracheas

Tracheal epithelial cells were isolated from female Hartley strain guinea pigs (Charles River, Kanagawa, Japan) by mild digestion with 1 mg/ml of pronase (Sigma) as described previously (32). The cells were collected by flushing the inside of the tracheal lumen with the culture medium, DMEM/F12 containing 5% fetal calf serum (FCS; Summit, Fort Collins, CO), hereafter referred to as DMEM/F12-FCS, and the antibiotics mentioned previously, and were washed several times with PBS. The cells were resuspended at a density of 4 × 10⁶ cells/ml in DMEM/F12-FCS.

Isolation of Fibroblasts from the Denuded Guinea Pig Tracheas and Skin

The five tracheas, from which epithelia had been denuded by the enzymatic digestion described above, were minced to pieces of around 1 to 2 mm³. The minced pieces were transferred to a centrifuge tube (15 ml; Iwaki Glass Co., Ltd., Chiba, Japan) and washed in PBS by vigorous mixing with a vortex mixer. After setting the tube aside for several minutes, the supernatant was removed by suction. The precipitated pieces were again resuspended in PBS. The remaining epithelial cells on the minced trachea were released into the supernatant during the washings, which were repeated until the supernatant was clear. The pieces were then cultured in Eagle's minimum essential medium (MEM; Nissui, Tokyo, Japan) containing 10% FCS (hereafter referred to as MEM-FCS) on a six-well culture plate (Iwaki) and the MEM-FCS medium was replaced each 2 to 3 d of the first 2 wk of culturing. It was observed that fibroblasts migrated from each piece and grew to subconfluence on each well for the first 2 wk. The wells were washed with PBS to eliminate the minced pieces. The remaining pieces were removed by gentle pipetting and/or were mechanically eliminated. The growing fibroblasts were harvested from the wells using trypsin-ethylenediaminetetraacetic acid (EDTA) (GIBCO-BRL) after two additional washings of each well with PBS. The recovered cells were washed twice with PBS and then resuspended in MEM-FCS. They were divided equally into five portions and each portion was transferred to a dish of 90 mm diameter (Iwaki). The MEM-FCS was replaced every 2 to 3 d. Non-spindle-shaped cells were mechanically removed by a Pasteur pipette under a microscope. The fibroblasts of each dish showed confluence until Day 4 or 5. The confluent fibroblasts were harvested with trypsin-EDTA after washing with PBS. The recovered cells per each dish were divided equally into four aliquots, and each was transferred to a new dish (90-mm diameter). Contamination of smooth-muscle cells was determined by immunostaining with mouse monoclonal antibodies (mAb) to human -smooth-muscle actin (Sigma) after the fifth passage.

After the fifth to seventh passage, the fibroblasts were collected from the confluent dishes using trypsin-EDTA. For each experiment, the recovered cells were washed twice with PBS and once with DMEM/F12-FCS just before use. In some experiments, the confluent dishes were washed with PBS and then incubated in 10 ml DMEM/ F12-FCS. The conditioned medium was collected after 48 h incubation and kept at 20°C until use.

In some experiments, fibroblasts from abdominal skin were used. The abdominal skin of guinea pigs was removed after shaving. The fatty tissue was eliminated from the skin, which was then minced into small pieces (1 to 2 mm³). The pieces were then washed and cultured in the same manner as for the minced trachea, above. The growing fibroblasts were collected from the dishes and passaged 3 to 4 times. The skin fibroblasts were collected and subjected to co-culturing as described previously.

Three-Dimensional Co-Culturing of Tracheal Epithelial Cells and Fibroblasts

Epithelial cells were cultured according to a previously described method (32). Briefly, 250 µl of epithelial cell suspension (4 × 10⁶ cells/ml) was added to the epithelial side of the amnion membrane. Each chamber was placed in a six-well culture plate containing 5 ml DMEM/F12-FCS and maintained at 37°C in an incubator with 5% CO₂. Three days later, the epithelial side of the membrane was washed with PBS to eliminate nonattached cells. Thereafter, the DMEM/F12-FCS was replaced every 2 d. Three weeks later, the chamber was turned over. Two hundred microliters of fibroblast suspension in DMEM/F12-FCS were added to the serosal side of the membrane at a density of 5 × 10⁵ cells per chamber. The plates were incubated for 24 h. The epithelial cells were maintained by immersion feeding during the incubation. The nonadherent fibroblasts were eliminated by washing with PBS. The chambers were returned to the original orientation after 24 h incubation, after which culturing was continued for an additional 10 d. The epithelial cells were maintained by immersion feeding during the first week of culturing, and thereafter were maintained using air-liquid interface feeding.

Electron Microscopic Observation

Amnion membranes with cultured cells were removed from the tissue-holding devices and washed twice with PBS. Each membrane was fixed in 2% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.2) and postfixed in 1% osmium tetroxide for 1 h in sodium phosphate buffer (0.1 M, pH 7.2). The membranes were then dehydrated in a graded series of ethanol (50 to 100%) followed by propylene oxide, and were embedded in Epon 812 (Abbot, North Chicago, IL) for transmission electron microscopy. Ultrathin sections of 70 to 80 nm, cut with a diamond knife by Ultrotome III 8800 (LKB-Produkter AB, Bromma, Sweden), were stained with uranyl acetate and lead citrate and examined with an H-7000 electron microscope (Hitachi, Ltd., Tokyo, Japan).

To compare the subpopulation between in situ guinea pig tracheas and the cultured cells, approximately 200 tracheal epithelial cells from each of five normal guinea pigs (i.e., a total of more than 1,000 epithelial cells) and over 100 cultured epithelial cells from five different batches (i.e., more than 500 cultured cells) were observed with a transmission electron microscope. In in situ tracheas, the ciliated cells had electron-lucent cytoplasms and many cilial structures. The goblet cells, which were distinguished by their electron-lucent secretory granules, appeared relatively dark, and the basal cells situated over the basement membrane had electron-dense cytoplasms but no granules.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Purity of Tracheal Fibroblasts

After the fifth passage, the purity of fibroblasts was morphologically and immunohistologically determined. Fibroblasts formed a confluent monolayer during each passage, but no formation of multiple layers or piling up of cells was observed. Based on the observation of over 200 cells, more than 95% of the cells showed a typical fibroblast shape (i.e., a peculiar spindle shape) but endothelial and epithelial cells were not observed. In addition, these spindle-shaped cells were not stained with mAb to -smooth-muscle actin (95%). The observations indicate that the predominant cells in the preparation were fibroblasts. Moreover, the majority of cells from skin were also shown to be fibroblasts, based on microscopic observation.

Co-Culturing of Tracheal Epithelial Cells with Fibroblasts on the Amnion Membrane

It is known that, in situ, the tracheal epithelium in guinea pigs exhibits a pseudostratified columnar epithelium on a basal membrane, that the epithelial cells are squamously placed on the basement membrane, and that fibroblasts are located in the mesenchymal matrix. Thus, it is inferred that the three-dimensional location of the epithelial cells, the basement membrane, and the fibroblasts would play a role in organogenesis of the trachea. Therefore, the following conditions were designed for the three-dimensional culturing. The epithelial cells were first cultured on the epithelial side of the amnion membrane on Day 1; the feeding was shifted to air-liquid interface feeding on Day 7; the culturing of fibroblasts on the serosal side of the membrane was started on Day 21; and the morphologic examination was carried out on Day 31.

As illustrated in Figures 1a and 2a, tracheal epithelial cells exhibited a confluent monolayer on the amnion membrane, and some ciliated cells were observed after 31 d of culturing (Figure 1a). A pseudostratified region was observed in some parts of the membrane (Figure 2a). Although the shape of the epithelium was still cuboidal rather than columnar, the appearance of ciliated cells was clearly observed. Moreover, these cells formed tight junctions with each other. Dramatic morphologic changes of the epithelial cells were observed when the epithelial cells were co-cultured with fibroblasts during the last 10 d (Figures 1b and 2b). Surprisingly, the epithelial cells seemed to form a pseudostratified columnar epithelium and they were squamously placed at the top of the layer (Figure 1b). The electron-dense basal cells, which were very small and flattened, appeared to be closely attached to the amnion membrane. Comparing Figures 2a and 2b, it can be seen that the ciliated epithelial cells became more columnar than those without co-culture. Interestingly, some goblet-like cells with electron-dense cytoplasms were observed and stained positively with periodic acid-Schiff (data not shown). In addition, the intracellular localization of nuclei and other organelles (e.g., the endoplasmic reticulum and secretory granules) was similar to that seen in vivo. Furthermore, tight junctions among the epithelial cells were observed at the top of the epithelial layer. These observations indicate that the three-dimensional co-culturing system resulted in formation of a pseudostratified columnar epithelium. Moreover, the results suggest that interaction between epithelial cells and fibroblasts would be involved in the organogenesis of the tracheal epithelium.

View larger version (105K): [in this window] [in a new window]   Figure 1.   Semithin section of tracheal epithelial cells cultured with or without fibroblasts. (a) Tracheal epithelial cells were cultured on the amnion membrane for 31 d without fibroblasts. At 7 d after the beginning of the culturing, the epithelial cells were maintained under an air-liquid interface feeding. (b) Tracheal epithelial cells were cultured with tracheal fibroblasts for the last 10 d of a 31-d culturing. The procedure is described in MATERIALS AND METHODS. (c) Tracheal epithelial cells were cultured with skin fibroblasts in the same manner as tracheal fibroblasts. Bar = 50 µm.

View larger version (93K): [in this window] [in a new window]   Figure 2.   Electron microscopic photograph. High magnification of cultured epithelial cells. Tracheal epithelial cells cultured without (a) and with (b) tracheal fibroblasts. Bar = 10 µm.

Dramatic changes in fibroblast layers were frequently observed during the co-culturing. When the fibroblasts were cultured on the serosal side of the amnion membrane without epithelial cells, they usually formed a confluent monolayer for 10 d. However, a multilayer of fibroblasts was not consistently formed under this culturing condition (data not shown). The frequency of formation of the multilayers was dramatically increased by co-culturing with epithelial cells (Figure 1b). These results suggest that epithelial cells modulate the proliferation of fibroblasts.

Fibroblasts are thought to represent a group of functionally heterogeneous cells (33). We compared epithelium that had co-cultured with fibroblasts from the trachea and the skin. The co-culture with skin fibroblasts resulted in a pseudostratified columnar epithelium, as did that with tracheal fibroblasts (Figure 1c). However, the co-culturing with skin fibroblasts did not consistently lead to formation of the same epithelium as that from tracheal fibroblasts; the former showed partially enlarged intercellular spaces like edema in their basal portions (data not shown). Although the data are not shown, the epithelium consisted of ciliated, goblet-like, basal and unknown cells. These results indicate that the skin fibroblasts have the same ability to form a pseudostratified columnar epithelium and to induce differentiation of epithelial cells as do the tracheal fibroblasts.

Comparison of Subcellular Population among Natural and Cultured Epithelial Cells

In the trachea of the guinea pig, we recognized at least four types of epithelial cells: ciliated, goblet, basal, and intermediate cells (34). According to our observations, ciliated, goblet, basal and unidentified cells represented 39.5, 27.7, 23.9, and 8.9% of total epithelial cells, respectively (32) (Table 1). In the same way, we classified the epithelial cells that had been co-cultured with fibroblasts into four types. These were composed of 46.6% ciliated, 15.4% goblet-like, 26% basal, and 12% unidentified cells (Table 1). This composition was very close to that seen in in situ epithelia of the guinea pig trachea. These findings are in contrast to the previous report (32) for epithelial cells cultured alone for the same period, consisting of 37.2% ciliated and 62.8% unknown cells.

The results support the observations that in vitro reconstitution of tracheal epithelia is achievable using our culture system, and that epithelial-mesenchymal interactions likely play a critical role(s) in the appearance of goblet and basal cells.

Factor-Dependent Differentiation of Epithelial Cells

In the next series of experiments, we examined whether the effects of co-culturing with tracheal fibroblast are mediated by a soluble factor or factors. For this purpose, the fibroblasts were incubated in DMEM/F12-FCS for 48 h. The conditioned medium was added to the epithelial culture at the original concentration instead of co-culturing with fibroblasts. As illustrated in Figure 3, the addition of the conditioned medium from the fibroblasts led to essentially the same results as the co-culturing. In both this figure and Figure 2b, it can be seen that the epithelial cells formed a pseudostratified columnar epithelium, that some electron-dense basal-like cells were closely attached to the amnion membrane, and that some goblet-like cells were observed. These results suggest that differentiation of epithelial cells during organization of the tracheal epithelium is mediated by a soluble factor or factors derived from fibroblasts.

View larger version (95K): [in this window] [in a new window]   Figure 3.   Tracheal epithelial cells cultured with conditioned medium from tracheal fibroblasts. (a) Semithin section of tracheal epithelial cells. Bar = 50 µm. (b) High magnification of cultured epithelial cells. Bar = 10 µm. Tracheal epithelial cells were cultured on the amnion membrane for 21 d without fibroblasts, and then with conditioned medium from fibroblasts for the last 10 d of the 31-d culture period.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

In the present studies, we showed that three-dimensional co-culturing of epithelial cells with fibroblasts induced differentiation of goblet cells, appearance of basal cells, and pseudostratified epithelia. The cell population and morphology of in vitro epithelia were very similar in the in situ tracheal epithelia of guinea pigs, indicating that the present culture system has the important advantage of differentiating three types of epithelial cells simultaneously within the same culture well. Thus, this is the first report of a three-dimensional culturing system to achieve in vitro reconstitution of the tracheal epithelium without addition of any exogenous factors. Moreover, the results suggest that epithelial-mesenchymal interactions play a crucial role in the organogenesis, morphogenesis, and cytodifferentiation of the tracheal epithelial cells.

In our previous study (32), we showed that both the change from immersion feeding to air-liquid interface feeding of epithelial culture and the cultivation of epithelial cells on the amnion membrane were important for the appearance of ciliated cells. Thus, the cultivation of epithelial cells on a proper extracellular matrix under conditions of air-liquid interface feeding appeared to be a sufficient stimulus for the differentiation to ciliated cells, but not for that to goblet-like or basal cells. The present results using our co-culture system suggest that differentiation of the latter two types of cells and formation of a pseudostratified epithelium require interaction between epithelial cells and fibroblasts in addition to that between epithelial cells and matrix.

One of our present aims was to elucidate the role or roles of mesenchymal cells in the differentiation of epithelial cells during reconstruction of trachea at an adult stage. For this purpose, we also examined whether the smooth-muscle cells exerted the same effect as the tracheal fibroblasts and whether the fibroblasts from different organs had the same capability as the tracheal fibroblasts. Because it was difficult to obtain a sufficient number of smooth-muscle cells from the trachea, we used smoothmuscle cells from the aorta. The smooth muscle led to neither further differentiation of epithelial cells nor formation of pseudostratified layers (data not shown). It would thus appear that fibroblasts, but not smooth-muscle cells, play a crucial role in differentiation of epithelial cells during tracheal reconstitution. The results shown in Figure 1c indicate that skin fibroblasts share the same capability as tracheal fibroblasts in regard to the formation of pseudostratified epithelia and the induction of differentiation of epithelial cells. A number of sophisticated studies on the role of fibroblasts in organogenesis have been reported (29, 31). These studies have also indicated that fibroblasts play a critical role in differentiation of epithelial cells, although they used fibroblasts taken from different organs. Together these results suggest that fibroblasts are generally involved in organogenesis and/or differentiation of epithelial cells.

Our observations indicate that interaction between fibroblasts and epithelial cells is likely mediated by a soluble factor or factors, because the conditioned medium of the fibroblasts was able to reconstitute tracheal epithelia as well as the fibroblasts themselves (Figure 3). In vitro reconstitution of human tracheal epithelia was reported by Yamaya and colleagues (35). In their system, cultivation of epithelial cells on Vitrogen gel and Ultroser G serum was required for reconstitution. Although co-culturing with fibroblasts was not required in their system, the use of FCS rather than Ultroser G serum failed to induce reconstitution. Their results suggest that this specific serum contains some growth and/or differentiation factors which lead to the appearance of those types of cells. In addition, Halttunen and colleagues reported an in vitro three-dimensional cell culture model in which intestinal cryptlike T84 epithelial cells organized and differentiated into a distinct phenotype when supplemented by addition of fibroblasts (29). The fibroblast-induced organization and differentiation were induced by transforming growth factor- (TGF-). Montesano and associates indicated that a soluble factor or factors (e.g., hepatocyte growth factor [HGF]) from mesenchymal cells was involved in the morphogenesis of canine kidney tubular epithelia (14, 31). Nogawa and Ito reported that acidic fibroblast growth factor is involved in differentiation of mouse lung epithelia during embryogenesis (36). Also, EGF has been reported to be involved in differentiation of airway epithelia (37). Furthermore, it was shown that the cultivation of epithelial cells on collagen type I gel resulted in differentiation to mucociliated cells in the presence of EGF and retinoic acid, and that addition of retinoic acid was essential for differentiation to mucociliated cells (40). In contrast, our culture system does not need an artificial chemical (i.e., retinoic acid) to observe differentiation to mucociliated cells. Thus, our system has an adantage for analyzing factors involved in adult tracheal epithelium. In preliminary experiments, we also tested the roles of HGF, EGF, and TGF- in differentiation of our tracheal epithelial cells. However, the results indicated that none of the tested growth factors induced the mucociliary differentiation of epithelial cells (data not shown). Further studies on the identification of soluble factors from the fibroblast and for the expression of its receptor on epithelial cells are currently under way in our laboratories.

It is well known that contact inhibition is observed in fibroblast culturing. Indeed, contact inhibition is usually observed whether the fibroblasts are cultured on the serosal side of the amnion membrane or in plastic dishes. Our findings indicate that the co-culturing with epithelial cells would lead to deregulation of contact inhibition of the normal fibroblasts. A similar abnormal growth of fibroblasts is observed in chronic inflammatory lesions, for example, interstitial pneumonitis. Many soluble factors that include cytokines produced from epithelial cells and infiltrating inflammatory cells are known as mitogens for fibroblasts: interleukin-6 (43), TGF- (44), platelet-derived growth factor (45), granulocyte macrophage colony-stimulating factor (46), and endothelin-1 (47). However, it is rare to observe the formation of fibrosis or overgrowth of fibroblasts even in an acute inflammatory trachea. Because the epithelium in the respiratory tract is exposed to infectious microorganisms, chemical irritants, and endotoxin, it is assumed that repair of damaged epithelial cells would always occur in vivo. Thus, it is speculated that the growth of fibroblasts could be negatively regulated under in vivo circumstances, whereas this regulation might not occur in the co-culture system. Thus, our system could be used as an in vitro system which would reflect the course of repair in chronically damaged epithelia. The system would also be a useful model for investigation of the course of recovery and remodeling of epithelia following acute and chronic damage.

In conclusion, we have established a three-dimensional co-culture system using epithelial cells and fibroblasts of guinea pig tracheas and denuded human amnion, and leading to in vitro reconstitution of the epithelium of the guinea pig. These findings suggest that epithelial-mesenchymal interactions play a crucial role in the morphogenesis and differentiation of the epithelium of the trachea.