Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The airway epithelium is a target of inflammatory and physical stimuli in obstructive diseases such as asthma and bronchopulmonary dysplasia. The epithelium provides both a physical barrier to the external environment and regulates several key metabolic functions of airways, including fluid and ion transport to the airway lumen, mucociliary clearance, and airway diameter. Injury to the epithelium is a common finding in pathologic studies of patients with even mild asthma, and is associated with worsening of clinical symptoms (1, 2). Repair of a damaged epithelium, then, may be a necessary part of restoring the airway to its normal state after an exacerbation of asthma. Repair generally involves several steps, including migration and spreading of epithelial cells at the margin of the injury into the damaged region and proliferation of new epithelial cells (3). Each step can be modulated actively by growth factors secreted by constitutive cells within the airway, such as fibroblasts (6), or depressed by mediators secreted by inflammatory cells that have migrated into the airways (7, 8).

Epidermal growth factor (EGF) has long been recognized as a mitogen for several types of epithelium, including human keratinocytes and airways epithelium of several species (9). Epidermal growth factor has been suggested to accelerate wound healing over 2 wk in the central airways of sheep following smoke inhalation injury (13), though it is not clear how much of this effect is due to proliferation of new epithelium versus migration and spreading of epithelial cells into the injury site. Other growth factors, such as calcitonin gene-related peptide and the tachykinin substance P have been demonstrated to be both mitogenic and chemotactic for airways epithelium (14), and thus have been termed "mitoattractants." However, it is not clear that EGF is also a mitoattractant for airways epithelium. Using a model of keratinocyte culture, Barrandon and Green have suggested that EGF cannot stimulate keratinocyte proliferation unless the cells are free to migrate outward from the rim of a growing colony of cells (11). The effect of EGF on airway epithelial cell migration and spreading, however, is not clear.

We postulated that EGF would stimulate migration and spreading of airway epithelial cells in culture, and would accelerate wound closure in vitro partly through its effects on cell migration. We studied cell migration in culture using two techniques: chemotaxis through a porous, matrix-coated membrane in a 48-well chemotaxis chamber; and wound closure in a cell monolayer by video microscopy. Our data demonstrate that EGF is a chemotactic factor for airway epithelial cells in culture, and enhances wound closure in epithelial cell monolayers partly through its effect on cell migration. These effects are seen regardless of underlying matrix protein support and are not due to cell proliferation at the wound leading edge. These data suggest that EGF may have a role in early epithelium repair via stimulation of cell migration.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Isolation and Culture of Guinea-Pig Tracheal Epithelial Cells (GPTEC)

Institutional approval for the use of guinea pigs in the study was given by The University of Chicago Animal Care and Use Committee. Twenty-five male Hartley guinea pigs were anesthetized with 100 mg/kg pentobarbital and killed by placement into a 100% CO₂ chamber for 6 to 10 min. Under sterile conditions, the mid-cervical trachea was dissected free and placed into 0.1% protease solution (type 25) in Hank's balanced salt solution for 2 h at 37°C. Tracheal segments were then transferred to sterile plates containing Ham's F12 medium with 5% fetal calf serum (FCS) and opened. Epithelial cells were dislodged using a micro-spatula, triturated through a small-bore pipette tip to break up clumped cells, collected, and centrifuged at 850 × g for 11 min, then washed twice. Cell pellets were resuspended in Ham's F12 medium containing 10 µg/ml insulin, 5 ng/ml EGF, 0.5 µg/ml transferrin, 0.4 µg/ml hydrocortisone, 2 ng/ml triiodothyronine, 100 µg/ml penicillin, 25 µg/ml amphotericin, and 5% FCS (medium A) (15, 16). For chemotaxis and wound-repair experiments, cells were plated at a density of 3-5 × 10⁵ cells/well in 35-mm-diameter collagen-coated culture wells and incubated at 37°C in 5% CO₂ atmosphere for 7-10 days. High-density plating was done for wound-repair experiments to assure generation of confluent monolayers. For proliferation experiments, cells were plated at a density of 2.5 × 10⁴ cells/ well in similar culture wells and incubated under similar conditions for 2 days. Low-density plating was done for proliferation experiments to prevent deleterious effects of edge-contact inhibition on proliferation. Epithelial cell origin was confirmed by typical morphology in these studies and by staining to an anti-cytokeratin antibody and electron microscopy in past studies (14). Cells from more than one animal were never mixed in any way in a given experiment. In the following, "N" refers to an individual animal.

Determination of GPTEC Proliferation

Cell proliferation was assessed in response to EGF to ensure that these cells responded in a manner similar to that described for other epithelial cell types (9), and to ascertain a concentration range for subsequent experiments. Proliferation was measured by the incorporation of the thymidine analog bromodeoxyuridine (BrdU) into the DNA of proliferating cells (17). Subconfluent GPTEC grown in collagen-coated wells for 2 days were incubated with 3 ml of medium B (medium A minus FCS and EGF) for 24 h to make the cells quiescent. Wells were then filled with 3 ml of medium B containing 10 mM BrdU and 1-50 ng/ml EGF. Control wells received BrdU only. After 24 h, cells were harvested and cytospin preparations were prepared and fixed in 70% ethanol. The percentage of cells incorporating BrdU into their nuclei was assessed using an avidin-biotin-peroxidase complex technique (14) and a specific mouse anti-BrdU antibody (Dako, Inc., Carpinteria, CA). Cells were counterstained with hematoxylin. BrdU-positive cells were counted as a percentage of all cells for each slide (at least 200 cells counted).

Determination of GPTEC Chemotaxis

We have described this method in detail previously (15, 16). Cells were grown until confluent in medium A, and then made quiescent by incubation for 24 h with medium B. Cells were harvested, triturated through a small-bore pipette tip to break up clumped cells, washed twice, and then resuspended in bland F12 medium at a concentration of 10⁶ cells/ml. Cell viability was  85% as assessed by trypan blue exclusion. Epithelial cell chemotaxis experiments were done in a 48-well blindwell chamber (Neuroprobe, Bethesda, MD). In five experiments, 25 µl of medium B containing 1-50 ng/ml EGF was placed into each of the bottom wells. Additional wells received bland F12 medium alone as a control. The wells were covered with an 8-µm-pore polycarbonate filter (Nucleopore, Pleasanton, CA) coated with 0.1% gelatin. The upper plate was then placed above the filter and 50,000 epithelial cells in 50 µl of medium were placed into each of the top wells. The chamber was incubated at 37°C in 5% CO₂ atmosphere for 6 h. The filter was then removed and adherent cells on the top surface of the filter were removed by careful scraping. The filter was stained with modified Wright stain (Leukostat; Fisher, Hasca, IL), and migrated epithelial cells on the underside of the filter were counted using a light microscope at ×400 total magnification.

Repair of GPTEC by Video Microscopy

We have previously published details of this method (18). Briefly, cells were grown until confluent in medium A. The cells were then made quiescent by incubation overnight with medium B. The next day, cells were washed and placed in 3 ml Ham's F12 medium ± EGF. A small wound was made in the confluent monolayer with a rubber stylet; this removed the epithelial cells without disturbing the underlying collagen matrix. Wound closure was measured serially for 24 h. Microscope images were photographed using a Sony Iris CCD camera (Sony, Inc., Rolling Meadows, IL) on a Nikon Diaphot inverted stage microscope. Video images were digitized using a Power Macintosh 6100 AV computer (Apple Computer, Inc., Cupertino, CA) and Video Monitor software (Apple Computer). Analysis of perimeter length and area of the remaining wound in each image was performed using NIH Image software (Wayne Rasband, National Institutes of Health, Bethesda, MD, available on the Internet at http://rsb.info.nih.gov/nih-image).

In 15 experiments, repair of a wound in a confluent monolayer was followed for 24 h after creation of the wound and addition of 2.5-15 ng/ml EGF in medium B or medium B alone as a control (one concentration per well). In these experiments, three wound sizes were created (n = 5 for each) to compare the effect of EGF on closure of differently sized monolayer wounds.

In four additional experiments, repair was followed for 24 h after wound creation in cells grown on different matrix support. In these experiments, 35-mm-diameter cell culture wells were coated either 50 µg/ml with collagen, as above, or 50 µg/ml laminin, or 50 µg/ml fibronectin. Plates were dried overnight and washed twice with phosphate-buffered saline. Cells were then plated at a density of 3- 5 × 10⁵ cells/well and grown to confluence in medium A. Monolayers were made quiescent in medium B for 24 h prior to starting the experiment. Wounds of ~ 0.8 mm² were created and imaged, after which monolayers were treated with either 15 ng/ml EGF in medium B or medium B alone as a control, and imaged over 24 h.

Materials

Insulin, penicillin, hydrocortisone, transferrin, bovine serum albumin, triiodothyronine, Ham's F12 medium, trypsin, EGF, EDTA, and type 25 protease from Bacillus polymyxia were obtained from Sigma Chemical Co. (St. Louis, MO). FCS was obtained from Hyclone (Logan, UT) and was heat-denatured prior to use. Mouse anti-BrdU antibody was obtained from Dako, Inc.

Data Analysis

BrdU labeling is expressed as mean ± SEM. Chemotaxis is expressed as the number of cells per 10 high-power fields (hpf). Wound closure is expressed either as area (µm²), length of the wound perimeter (µm), or percent control area or length. In preliminary video microscopy experiments using cell monolayers, intra-observer variability was < 2% and inter-observer variability was < 4% for all measurements. The time required for 50% closure of a wound was calculated by linear interpolation between the two closest data points and expressed in minutes. When the 50% closure time exceeded 24 h, the maximum time (24 h) was used. The time required for closure of the first 100,000 µm² of wound area (100K time) was also calculated by linear interpolation and expressed in minutes. Comparisons between multiple groups were made by repeated-measures analysis of variance (ANOVA); when significant differences were found, further comparisons were made by Fisher's least significant difference test. Comparisons between two groups were made by paired Student's t test. Bonferroni's correction was made as appropriate for multiple comparisons. Differences were considered significant when P < 0.05.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Proliferation of Airway Epithelial Cells after Stimulation with EGF

As expected, treatment with EGF stimulated GPTEC proliferation in a concentration-dependent manner as shown in Figure 1. In five experiments, BrdU labeling after stimulation with 10 ng/ml EGF was 12.3 ± 3.5% versus 2.6 ± 0.6% for control (P < 0.01). The EGF concentration range of 5 to 15 ng/ml was different than control (P = 0.03 by repeated-measures ANOVA). These data demonstrate that GPTEC, much like other epithelial cell types, proliferated in response to treatment with EGF; and suggested an appropriate concentration range of EGF for subsequent migration and wound-repair experiments.

View larger version (14K): [in this window] [in a new window]   Figure 1.   Proliferation of guinea-pig tracheal epithelial cells after stimulation with EGF, as measured by labeling with BrdU. Stimulation with EGF for 24 h elicited concentration-dependent proliferation, as expected. n = 5 experiments.

Chemotaxis of Airway Epithelial Cells after Stimulation with EGF

Treatment with EGF stimulated GPTEC migration in a concentration-dependent manner as shown in Figure 2. In five experiments, cell migration after stimulation with 10 ng/ml EGF was 42.1 ± 9.7 cells versus 7.5 ± 0.4 cells per 10 hpf for control (P < 0.03) (Figure 2). The EGF concentration range of 5 to 50 ng/ml was different than control (P = 0.001 by repeated-measures ANOVA).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Migration of guinea-pig tracheal epithelial cells after stimulation with EGF. Stimulation with EGF for 6 h elicited concentration-dependent migration. n = 5 experiments.

Repair of GPTEC Monolayers by EGF as Assessed by Video Microscopy

Creation of small wounds within confluent GPTEC monolayers by mechanical abrasion was achieved in every experiment, and wound closure could be followed readily over 24 h (Figure 3). Initial monolayer wounds were generally ovoid in appearance, with smooth edges. No subsequent lifting or sloughing of edge margins was noted in any experiment. Wounds of three different sizes were created, as noted in Table 1. The coefficient of variation for the initial area and perimeter length of monolayer wounds within an individual experiment was  20% in each experiment. Larger injuries, as expected, closed more slowly. After 24 h, the remaining wound size in group A controls was 1 ± 1% of time 0, whereas in group B controls, wound size was 57 ± 15% of time 0 and in group C controls, wound size was 80 ± 12% of time 0 (Figure 4).

View larger version (87K): [in this window] [in a new window]   Figure 3.   Representative video micrographs of GPTEC monolayer wound closure from group B (~ 0.68 mm2 starting area). Monolayer images were digitized over 0-24 h. Remaining wound area in µm2 is shown for each image. Wound margins are somewhat irregular but are easily visualized and traced on the digitized image. Original magnification: ×40 for all images.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Wound closure in control GPTEC monolayers in each size group. Left: remaining wound area over time. Right: remaining perimeter length over time. Results are normalized to original (time 0) wound area or perimeter length for each experiment. Smaller wounds closed more quickly than larger wounds, as expected. n = 5 experiments in each group. Groups A, B, and C reference original wound area as defined in Table .

For each group of starting-injury size, treatment with EGF accelerated wound closure (Figure 4). A clear concentration-response effect to EGF was noted in each group. Differences in the remaining wound area 12 h after EGF or sham treatment were evaluated, as this time point represented the best range of response across each group. At 12 h, treatment with EGF decreased wound area in each group compared with control (Table 2). The 50% closure time was different in control monolayer wounds across the three size groups, as expected (P < 0.0001 by ANOVA). The 100K time in control monolayer wounds across the size groups was similar, also as expected (P = 0.55 by ANOVA). In each group, the time required for 50% closure and the 100K time (Table 3) were decreased significantly after treatment with EGF.

Perimeter length of the wound also was significantly accelerated over 24 h by treatment with EGF (Figure 5). In control monolayers, perimeter length changed less rapidly than wound area. This was due to projections of new cells into the wound which created a more irregular margin during the closure process, so that while area decreased, perimeter decreased less or even increased at early time points. Monolayer wounds in groups A and B treated with 15 ng/ml EGF had a substantial decrease in perimeter compared with control monolayer wounds 12 h after wound creation (Table ), and perimeter length in EGF-treated monolayer wounds started to decrease before such changes were noted in control monolayers (Figure 5).

View larger version (20K): [in this window] [in a new window]   Figure 5.   Closure of monolayer wounds in GPTEC monolayers after treatment with EGF. Results are normalized to original (time 0) wound area for each experiment. There is substantial acceleration of wound closure after treatment with EGF. n = 5 experiments in each group. Error bars for some curves are omitted for clarity; SEM for all curves are similar. Groups A, B, and C reference original wound area as defined in Table .

Wound Repair in Epithelial Monolayers Grown on Different Extracellular Matrix Protein Substrates

Epithelial cells grew to confluence equally well when grown on either collagen, fibronectin, or laminin substrate. Initial wound area in these experiments was 796,290 ± 28,640 µm² and initial perimeter length was 3,910 ± 80 µm, both of which were comparable to group B monolayers in initial experiments. There was no difference between the three substrate groups in wound closure in control monolayers or monolayers treated with EGF, as demonstrated by wound area (Figure 6). The time required for 50% closure and the 100K time were also similar between the three groups (data not shown).

View larger version (19K): [in this window] [in a new window]   Figure 6.   Perimeter length in GPTEC monolayers after treatment with EGF. Results are normalized to original (time 0) perimeter length of the wound for each experiment. There is substantial reduction in the perimeter after treatment with EGF. n = 5 experiments in each group. Error bars for some curves are omitted for clarity; SEM for all curves are similar. Groups A, B, and C reference original wound area as defined in Table .

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The objective of this study was to determine whether EGF can stimulate the migration of airway epithelial cells that are required in the repair of a damaged airway epithelium. A blindwell chemotaxis chamber was used to measure both chemotaxis and chemokinesis. Our data demonstrate that EGF stimulated GPTEC migration as measured in the chemotaxis chamber assay (Figure 2). Migration stimulated by EGF was equivalent to that measured after treatment with insulin, a peptide previously demonstrated to be a potent chemotaxin for airway epithelial cells in several species (15, 16, 19) and for other epithelial cell types (20). Treatment with EGF also accelerated closure of a wound in an epithelial cell monolayer (Figures 4-6 and Tables 123). This began rapidly and was substantial over the first 24 h. The maximal response to EGF in each experimental assay occurred at similar concentrations (~ 10 to 15 ng/ml). Thus, EGF elicits both migration and proliferation (Figure 1) of airway epithelial cells and thus properly is a mitoattractant. These data suggest that the trophic effects of EGF on the airways epithelium are not limited to proliferation alone, and each effect may be useful in the repair of an epithelial injury.

Epithelial repair in both control and EGF-stimulated conditions was not dependent on the composition of the underlying protein matrix, in that cells grown on collagen, fibronectin, or laminin responded equally well after injury (Figure 7). These results are similar to previous studies that have examined the ability of airway epithelial cells to adhere to and migrate through extracellular matrix. Spurzem and colleagues (21) have demonstrated that cultured bovine bronchial epithelial cells migrated equally well in response to fibronectin, laminin, and collagen IV in blindwell chambers. In a similar study, Rickard and associates (22) demonstrated that both freshly isolated and cultured bovine bronchial epithelial cells adhered equally well to fibronectin, laminin, and collagen IV. Epithelial cells make fibronectin in response to inflammatory mediators (23, 24), and it has been suggested that both fibronectin and exudated plasma proteins provide a provisional matrix for repair to occur after injury (25). Our data suggest that epithelial cells may utilize any of the matrix proteins found in existing basement membrane as a substrate for repair. Our results should be compared with those obtained by Garat and coworkers (28), using techniques similar to those in our study, in which wound repair in rat alveolar epithelial cells was accelerated in the presence of soluble or insoluble fibronectin compared with either type I or type IV collagen.

View larger version (19K): [in this window] [in a new window]   Figure 7.   Closure of monolayer wound in GPTEC monolayers after treatment with EGF: effect of underlying matrix protein. (A) Control, untreated monolayers. (B) Monolayers treated with 15 ng/ml EGF at time 0. Results are normalized to original (time 0) wound area for each experiment. Original wound area was ~ 1 mm2 for each experiment. Growing cells on fibronectin, laminin, or collagen did not change the rate of wound closure. n = 4 experiments in each group. Error bars for some curves are omitted for clarity; SEM for all curves are similar.

Much of what is known about epithelium repair in airways comes from studies in which the epithelium is injured and repair followed histologically over time. Early events in the repair process include cell spreading, migration, and proliferation (3). Adjacent stem cells initiate proliferation, and other cells migrate into the denuded area. These cells then flatten and spread over the site of injury. In a model of human nasal epithelium in primary explant culture, closure of a small wound (10,000 to 30,000 µm²) over 8 h involved both cell spreading and migration. In contrast, closure of larger epithelial wounds required both migration and proliferation of epithelial cells (29). In a model that evaluated wound closure in cultured type II rat alveolar epithelial cells with time-lapse video microscopy, EGF at higher concentrations enhanced wound closure (30). In a similar study, EGF was a chemoattractant for type II rat alveolar epithelial cells in blindwell chemotaxis chambers at concentrations similar to those used in the present study (31). Our data extend these observations and make clear that EGF can stimulate migration and wound repair in both major types of lung epithelia.

Our studies were designed to evaluate cell migration and spreading without the confounding effects of proliferation. Airway epithelial cells in our experiments closed a small wound readily, such that these wounds, ~ 0.35 mm² on creation, were about 50% closed after 24 h (Figure 4). Larger wounds (~ 0.7 and 3.2 mm², respectively) closed less under control conditions, though the initial rate of closure, shown as the 100K time, was similar for controls in each group (Table ). Therefore, the effect of EGF in terms of stimulating migration and wound closure was similar regardless of initial wound area.

One important consideration in these data is whether the wound closure noted in the first 24 h could be explained by processes other than cell migration or spreading. The time required for mammalian airway epithelial cells to traverse cell cycle is generally 22 to 28 h (32). In a previous study, we demonstrated via flow cytometry that GPTEC require  24 h to replicate (16). Thus the wound closure noted in our studies, especially within the first 12 to 18 h, when new daughter cells generated in response to a mitogenic signal delivered at time 0 would not be available, almost certainly was not due to proliferation. The video microscopy assay permitted an examination of the relative contribution of migration and spreading in two dimensions during early injury repair. Our data confirm that EGF elicited migration that could be measured both in a single dimension (a chemotaxis chamber) and in two dimensions (wound monolayer).

There are limitations to the chemotaxis experiments. Migration through a gelatin-covered filter may not approximate migration along a basement membrane in vivo. However, the utility of this assay has been demonstrated and it allows multiple experiments on cells derived from a single animal (15, 16, 18). A second limitation in these studies is that the GPTEC used are relatively de-differentiated, a problem common to many experiments utilizing cultured airway epithelial cells. It is not clear that more differentiated airway epithelial cells, such as ciliated or secretory columnar cells, would migrate differently than cultured cells, or whether differentiated cells would migrate at all.

We chose to study cell migration using video microscopy which is complementary to traditional studies of chemotaxis during early wound closure. This method, however, also has limitations. Wound closure has been shown to involve both migration and cell spreading, and distinguishing between the two on a digitized image may be difficult. Morphologic measurements of spreading during wound repair by imaging techniques have to take into consideration several technical factors. First, the irregular margins of a wound make measurements of spreading inward from the leading edge difficult. Second, there is non-uniformity of spreading, even if cells are evenly distributed on a regular wound edge. Third, the effects of migration may confound measurements of spreading. Fourth, there is difficulty in demonstrating without a marker that might in itself alter repair the location of the original, time 0 wound margin from which measurements of spreading might be made. Finally, migrating cells themselves may have a flatter appearance even if not "spreading"; without frequent, dynamic measurements of a single cell, it may be difficult to know if a cell is flattened and migrating or simply flattening in one place. For these reasons it is difficult to estimate the relative contributions of migration and spreading in the initial closure of a monolayer wound. The effects of EGF, then, may be on either migration or spreading alone, or on both in some proportion.

Another limitation relates to the simple matrix on which our cells were grown (collagen), which is different from the more complex basement membrane in which a variety of matrix proteins, such as fibronectin, laminin, type IV collagen, proteoglycans, entactin, and nidogen, are found (33). Interactions with multiple matrix proteins may have a different modulating effect on wound repair when compared with collagen alone. Finally, while the video microscopic method takes into account more factors that are involved with wound repair in vivo, it is still an in vitro approximation. Cells present within (e.g., sensory nerves) or immediately beneath (e.g., fibroblasts) the epithelial layer may modulate the repair process of a damaged epithelium (15, 16, 33). Such effects are not present in our in vitro assay.

In summary, we conclude that EGF is a chemoattractant and repair accelerant for guinea-pig airway epithelial cells in primary culture. Migration is stimulated by nanogram concentrations of EGF. Thus the trophic effects of EGF on airways epithelium are several, and each may play a significant role in early airway epithelial cell repair.