Restitution of an epithelial layer after environmental or biological damage is important to maintain the normal function of the respiratory tract. We have investigated the role of transforming growth factor (TGF)-β isoforms in the repair of layers of 16HBE 14o− bronchial epithelial-derived cells after damage by multiple scoring. ELISA showed that both latent TGF-β1 and TGF-β2 were converted to their active forms 2 h after wounding. Time-lapse microscopy showed that the addition of TGF-β1, but not TGF-β2, progressively increased the rate of migration of damaged monolayers at concentrations down to 250 pg/ml. This increase was blocked by addition of a neutralizing TGF-β1 antibody. Phase-contrast microscopy and inhibition of proliferation with mitomycin C showed that proliferation was not required for migration. These results demonstrate that conversion of latent to active TGF-β1 and TGF-β2 during in vitro epithelial wound repair occurs quickly and that TGF-β1 speeds epithelial repair. A faster repair may be advantageous in preventing access of environmental agents to the internal milieu of the lung although the production of active TGF-β molecules may augment subepithelial fibrosis.
- transforming growth factor-β
- 16HBE 14o− bronchial epithelial cell
- cell migration
- wound healing
- primary bronchial epithelial cells
transforming growth factor (TGF)-β is a multifunctional cytokine with significant regulatory effects on extracellular matrix production from mesenchymal cells in the lung (8, 12). Produced in an inactive form tethered to a latency-associated peptide (LAP) by covalent bonding (17), its signaling through TGF-β receptors requires exposure of the active site of the ligand either through conformational change or through cleavage of the LAP (17). The three isoforms secreted by mammalian cells have been studied extensively in connection with wound healing in skin where, in contrast to TGF-β3, TGF-β1 and TGF-β2 provide a quicker wound repair but induce more scarring (34).
The bronchial epithelium forms a barrier between the external and internal milieu of the respiratory tract and consists of ciliated and secretory columnar cells attached to a basal cell layer forming a pseudostratified epithelium. The intact epithelium prevents passive movement of environmental agents by the paracellular route, which is blocked by tight junctions (2). Damage to the epithelium involving loss of columnar epithelial cells can impact on the effectiveness of this barrier, allowing antigen access to the underlying bronchial tissue, as has been established in animal models (13). The restitution processes involved in the repair of damage to the bronchial epithelium are progressive and follow a number of characteristic and identifiable steps. After damage to the epithelial layer, epithelial cells abutting the wound edge dedifferentiate into a migratory phenotype to quickly cover the area of damage (15). The formation of a fibrin-fibrinogen gel rich in leukocytes at this point protects the underlying tissue until cell-cell contact has been reestablished. Finally, redifferentiation provides an intact fully functioning epithelium (16). Restitution to an intact undifferentiated epithelial barrier after damage to the airway can take as little as 1 h in the guinea pig (14), whereas the time taken in the human airway has not been determined.
Damage to the epithelium of the conducting airways has long been recognized to be a characteristic feature of asthma (27), with the shed epithelial cells, predominately ciliated columnar cells (24), found in bronchoalveolar lavage fluid and the morphological appearance of a single monolayer of epithelial cells seen in samples from bronchoscopy (24). Thus the proportion of epithelial cells in a repairing, dedifferentiated phenotype would be increased in asthmatic subjects. There is also an increase in the “activation” of epithelial cells after damage associated with increased release of mediators, including interleukin-5, interleukin-8, and granulocyte-macrophage colony-stimulating factor, which may affect the immunological responses in the surrounding tissue (5,32). The increase in subbasement membrane thickening resulting from the deposition of collagens I and III by myofibroblasts in the lamina reticularis (31) is well established and implicates TGF-β as a potential mediator in the pathogenesis of asthma.
Given the increased proportion of repairing epithelial cells in asthma, this study investigated the changes in TGF-β1 and TGF-β2 isoforms produced during the wound repair process and their effect on the speed of movement of the monolayer.
MATERIALS AND METHODS
The 16HBE 14o− bronchial epithelial cell line was a kind gift from Dr. D. Gruenert (Cardiovascular Research Institute, University of California; see Ref. 9). Stock cultures were routinely maintained in MEM with Earle's salts (Life Technologies, Paisley, UK); supplemented with 10% FCS (Life Technologies), 20 mMl-glutamine, 10 U/ml penicillin G, 10 μg/ml streptomycin sulfate, and 2.5 μg/ml amphotericin B (Life Technologies); and incubated at 37°C in a 5% CO2 incubator. Experimental cultures were grown free from antibiotics for a minimum of 24 h before the experimental protocol. Epithelial cells were used with a maximum range of 20 passages.
To induce mechanical damage, 16HBE 14o− cells were grown to confluence in serum-containing medium on 35-mm petri dishes (Greiner) and were damaged by scraping off a line of cells in a cross-hatch pattern using a P200 Gilson pipette tip. The medium and damaged cells were removed, the remaining cells were washed one time in fresh medium, and 2 ml of fresh serum-containing medium were added. For damage experiments in the absence of serum, the 16HBE 14o−monolayers were serum starved for 24 h before mechanical damage and the addition of fresh serum-free media for the duration of the experiment.
Primary cell culture.
Primary epithelial cells were obtained from tissue after lung resection. Mucosal strips were dissected from bronchial samples, and excess matrix material was removed. Strips of mucosa (1–2 mm) were placed on 24-well plates with the epithelial surface downward. Epithelial cultures were maintained in LHC-9 medium (Life Technologies; see Ref. 21) with 2% Ultroser G serum replacement (Life Technologies). All fibroblast-contaminated cultures were discarded.
mRNA was extracted using Trizol (Life Technologies) following the manufacturer's instructions and was stored at −20°C in diethyl pyrocarbonate-treated water until use. mRNA (1 μg) was reverse transcribed into cDNA with avian myeloblastosis virus reverse transcriptase (Promega, Southampton, UK).
PCR for TGF-β isoforms was performed as described previously (32). The TGF-β1, TGF-β2, and TGF-β3 primers were generated from the published sequences as follows (4): TGF-β1, GCC CTG GAC ACC AAC TAT TGC (5′), GCT GCA CTT GCA GGA GCG CAC (3′); TGF-β2, GTC TTG GAT GCG GCC TAT TGC (5′), GCT GCA TTT GCA AGA CTT TAC (3′); TGF-β3, GCT TTG GAC ACC AAT TAC TGC (5′), GCT ACA TTT ACA AGA CTT CAC (3′). TGF-β1 and TGF-β2 were primed with Taq polymerase (Promega) at an annealing temperature of 56°C, 1 mM MgCl2 for 30 cycles, whereas TGF-β3 resulting from the lower expression was primed for 40 cycles; expected length for all isoforms was 336 bp. The gene product for adenosyl phosphoribosyl transferase (APRT) was used to control for total mRNA [APRT sequence GCT GCG TGC TCA TCC GAA AG (5′) and AGG GCG TCT TTC TGA ATC TC (3′) at the same annealing temperature and number of cycles, with the expected length of 246 bp].
PCR products were visualized by gel electrophoresis with an ethidium bromide-stained 2% agarose gel and were observed and photographed on an ultraviolet transilluminator with a Polaroid camera.
After damage to epithelial monolayers by the cross-hatch method (see above) and subsequent culture for 5 min and 2, 6, 24, 48, and 72 h, the supernatants were removed and stored at −20°C. TGF-β1 and TGF-β2 were analyzed using a sandwich ELISA system (Promega) with monoclonal capture and polyclonal detection antibodies and visualization with tetramethylbenzidine according to the manufacturer's instructions. Cross-reactivity between all three isoforms was quantified, by the manufacturer, as <5% for both TGF-β1 and TGF-β2 antibodies. After the reaction was stopped with 1 M phosphoric acid, the plate was read on an ELISA plate reader at 450 nm, and all values were corrected using a standard curve.
The ELISA was used to measure active TGF-β and total TGF-β (after acid activation) in the supernatant following the manufacturer's instructions. The concentration of latent TGF-β was calculated from the active and total concentrations. Where the ELISA reading for the active molecule exceeded that of the total, the latent concentration was recorded as zero.
Time-lapse microscopy methods.
The 16HBE 14o− cells were grown to confluence in 35-mm petri dishes (Greiner) and were damaged with a single scrape from a pipette tip. All culture medium and damaged cells were removed, and the wounded monolayer was washed one time in medium and changed to fresh serum-containing medium containing 100 mM HEPES. The cells were incubated at 37°C in a 5% CO2 incubator for 1 h before time-lapse analysis. For cultures involving treatment with TGF-β1 or TGF-β2 (R&D Systems, Abingdon, UK), 0.25, 2.5, or 25 ng/ml were added to the serum-containing medium immediately before time-lapse analysis.
Where serum starvation was required, the epithelial monolayer was incubated for 24 h in MEM with Earle's salts and 20 mMl-glutamine without the presence of serum. For neutralization of TGF-β1, confluent monolayers were serum starved for 24 h before damage and time lapsed for 3 h in the presence of 0.25 ng/ml TGF-β1. Monoclonal anti-TGF-β1 (1 μg/ml; R&D Systems) was added, and the time-lapse recording continued for a further 3 h. For blockade of proliferation, monolayers were serum starved for 24 h before damage and recorded by time lapse for 6 h in the presence of 1 μg/ml mitomycin C (Sigma, Poole, UK). According to the manufacturers, the TGF-β1 antibody showed <1% cross-reactivity with the other two isoforms.
Time-lapse images were taken at 1-h intervals on a Leica DM IRB phase-contrast inverted microscope (Leica; Milton Keynes) maintained at 36 ± 1°C with a thermocouple-linked heater unit and thermometer. The images were collected with a cooled charge-coupled device array camera (Digital Pixel, Brighton, UK) connected to a computer running IPLAB version 2.41 (Scanalytics) and Digital Pixel capture software (version 1) over a 6-h time period.
All images were converted into TIFF format and analyzed using Scion Image PC (Scion, Frederick, MD). The area of damage within a set 0.7 × 0.47-mm box was measured at each time, and the difference in area covered between time points was calculated, thus providing a measure of the area covered by the repairing epithelium. Because the length of the wound edge was the same in all experiments (∼0.8 mm), the area covered during migration was used as an index of the linear rate of migration (mm2/h).
To measure the movement of individual cells within the epithelial sheet, the position of 20 wound edge cells was tracked through each frame. This was performed for control monolayers and 0.25 ng/ml TGF-β1-treated monolayers, and the distance travelled by each cell between each frame was calculated (mm).
Statistical analysis was performed using the SPSS statistical program version 10. Two independent sample t-tests were used to compare ELISA data between damaged and undamaged groups and time-lapse data between treated and untreated groups. Statistical significance was highlighted on individual data points.
Epithelium and epithelial damage.
The 16HBE 14o− cell line, an SV40 transformed bronchial epithelial cell line, grew in a typical epithelial fashion forming a fully confluent monolayer with a “cobblestone” appearance (9). After confluence in 35-mm petri dishes, mechanical damage in a cross-hatch pattern generated islands of epithelial cells, with gaps between each island of 5–25 cell widths (Fig.1). Cells from each island subsequently migrated in all directions to cover the area of damage giving a high proportion of cells in a migratory phase. Repair of small “wounds,” five cell widths, commonly occurred within 6 h of damage, with cell-cell contact being reestablished and confluence resumed ∼24 h after damage (Fig. 1). At this stage, the lines of damage could still be seen where cell-cell borders overlapped (Fig. 1); however, over the course of the next 24 h, the monolayer returned to a morphology indistinguishable from undamaged cultures. Few mitotic figures were seen in the islands of cells as they migrated to confluence. With the exception of time-lapse analysis, this model of epithelial repair was used for all subsequent experiments.
TGF-β expression profiles of 16HBE 14o− and primary epithelial cells.
To confirm the similarity in the TGF-β profile between the 16HBE 14o− cell line and primary epithelial cells from lung resection, RT-PCR for the three isoforms of TGF-β was performed (Fig.2). Primary epithelial cells gave strong PCR product bands for TGF-β1 and TGF-β2 that were broadly equivalent to each other and showed a lower level of expression for TGF-β3 (Fig. 2 A). A similar expression profile was observed in 16HBE 14o− cells (Fig. 2 B). The 16HBE 14o− cell line was used for all subsequent experiments.
Production of active TGF-β isoforms during epithelial repair.
Supernatants were taken from epithelial monolayers damaged by the cross-hatch method at 5 min and 2, 6, 24, 48, and 72 h after addition of serum-containing medium. Although the overall amount of the latent form of TGF-β1 declined over 72 h, a peak of latent TGF-β1 release was observed in the damaged cultures at 2 h postwounding. This induction of latent TGF-β1 was mirrored by a similar significant (P = 0.03) increase in the active form of the molecule at the same time point compared with undamaged control cultures, with the total percentage of TGF-β1 in the active form peaking at 26% in damaged cultures (Fig.3 A).
Similarly, a peak of both active and latent TGF-β2 was found in supernatants from damaged cultures 2 h after wounding. The level of active TGF-β2 thereafter paralleled undamaged control cultures; however, the concentration of latent TGF-β2 in damaged cultures increased significantly (P = 0.004) over undamaged control cultures between 6 and 72 h, when repair of the wound had occurred (Fig. 3 B). Compared with TGF-β1, the percentage of TGF-β2 in its active form peaked higher (38% active), with the concentration of active TGF-β2 at this time point being similar (0.31 ng/ml TGF-β1, 0.35 ng/ml TGF-β2; Fig. 3 B). The concentration of total TGF-β1 and TGF-β2 in the serum-containing medium was measured at 1 ng/ml TGF-β1 and 150 pg/ml TGF-β2, with 90–95% in the latent form.
Cultures damaged in the absence of serum showed a contrasting pattern to those in the presence of serum (Fig.4). Where serum was absent, there was no detectable TGF-β1 in either the active or latent form in undamaged cultures, and there was a small increase in active TGF-β1 in damaged cultures 24 h after damage (Fig. 4 A). Similarly, although in the absence of serum the peak in active TGF-β2 2 h after damage was missing, the presence of TGF-β2 at 24–48 h demonstrated that the cells were capable of its production (Fig.4 B).
Migration of epithelial cells after exogenous addition of active TGF-β isoforms.
To measure the effect of TGF-β1 and TGF-β2 on the rate of cell migration, 16HBE 14o− cells were tracked by time-lapse microscopy after a single scrape wound to the confluent monolayer. Images taken every hour were measured to determine the area covered by the cells per hour. Because the wound edge in all samples was of equal length (0.8 mm), the value of the area covered per unit time was directly proportional to the mean migration rate per hour. Thus a positive migratory rate indicates closure of the wound.
Untreated monolayers in serum-containing medium migrated at an average speed of 0.02 mm2/h, which decreased slightly over the 6-h period of migration studied (Fig.5 A-1). In contrast, addition of both 0.25 and 2.5 ng/ml TGF-β1 to the medium induced an increase in the rate of migration, with the addition of 0.25 ng/ml TGF-β1 reaching significance by 6 h (P = 0.03). The rate of migration reached a plateau at 0.0275 and 0.026 mm2/h, respectively (Fig. 5, A-1 and A-2). A similar pattern was found after serum starvation of the monolayer for 24 h before TGF-β1 treatment (results not shown).
Contrary to the increase in migratory rate seen for TGF-β1, there was no increase or decrease in the rate of migration observed with either concentration of TGF-β2 compared with untreated control cultures (Fig. 5, B-1 and B-2).
To exclude the possibility that the increased rate of migration of the epithelial layer after TGF-β1 treatment was the result of cell spreading only, the distance moved was calculated for 20 cells at the wound edge of the control and 0.25 ng/ml TGF-β1-treated monolayers (Fig. 5 C). This showed the same pattern of increased cell migration for the epithelial sheets after TGF-β1 treatment (Fig.5 A-1) and reached statistical significance after 4 h (P < 0.001).
Anti-TGF-β1 blocks the increased migration induced by TGF-β.
To further support the role of TGF-β1 in the enhanced migratory rate, a monoclonal antibody to the active form of TGF-β1 was used after 24 h of serum starvation to remove the exogenous active molecule. After a single scrape damage to the serum-starved 16HBE 14o− cell monolayers, addition of 0.25 ng/ml TGF-β1 significantly enhanced the rate of migration at 2 and 3 h (P = 0.05 and 0.038, respectively) compared with control cultures without additional TGF-β1. Addition of a monoclonal neutralizing antibody to the active form of TGF-β1 3 h after the initial stimulus reduced the rate of migration such that it was not significantly different from control cultures (Fig.6).
Relationship between cell migration and proliferation.
To establish that the enhanced rate of migration was not as a result of increased cell proliferation, time-lapse microscopy was performed in the presence of 1 μg/ml mitomycin C in cultures starved of serum for 24 h. This showed that, despite the inhibition of proliferation induced by mitomycin C, there was an increase in the rate of migration above the level of controls for the first 4 h of the 6-h time-lapse experiment (Fig. 7).
Repair of the bronchial epithelium after damage is an essential feature of respiratory tract function. Through time-lapse analysis and mechanical wounding of epithelial monolayers, this report highlights the importance of the isoforms of TGF-β for the repair of the bronchial epithelium in vitro and shows that damage to the bronchial epithelium can induce the activation of latent TGF-β.
Two models of epithelial repair were developed using the 16HBE 14o− bronchial epithelial cell line. The cross-hatch damage model was devised to generate a high proportion of cells in a migratory phase, thus allowing the study of cells and cell products during migration and later stages of the repair cycle (20). In comparison, the time-lapse model generated detailed measurements of the speed of migration over the initial 6-h time period after damage. Thus the time-lapse system showed that an enhanced migratory rate after TGF-β1 stimulation of the wounded monolayer involved a progressive increase of speed over the first 4 h studied. The SV40-transformed 16HBE 14o− cell line has been well characterized and has a number of the features of primary human bronchial epithelial cells (9). Indeed, RT-PCR confirmed that both primary cells and 16HBE 14o−cells expressed mRNA for all three TGF-β isoforms, and the 16HBE 14o− cell line was used for all further experiments.
Few differences have previously been observed between TGF-β1 and TGF-β2 in terms of their expression in epithelial repair. Skin wound models show that TGF-β1 and TGF-β2 show a similar pattern in terms of kinetics of expression (1). In the present study, damage to a bronchial epithelial monolayer demonstrated a divergence between TGF-β1 and TGF-β2 in terms of TGF-β release and function in cell culture supernatants. In damaged cultures only, and in contrast to the pattern for TGF-β1, latent TGF-β2 was detected in progressively increasing concentrations at the later stages of the time course, with this increase starting at a time point where resolution of the wound had occurred. This is consistent with previous published results using a similar model (37). Whether this reflects an increased production of TGF-β2 at this stage or a decreased breakdown and clearance of the molecule is unclear. However, time-lapse measurements of the wound edge after addition of TGF-β2 to the culture showed no detectable enhancement of migration, in contrast to increased migration with additional TGF-β1. Therefore, TGF-β2 appears to play no role in cell migration during wound repair of bronchial epithelial cells.
Production of TGF-β in most cell types is in the latent form, and activation is necessary for production of the active form and for cell signaling (17). In vitro, this can be performed by acidification and then neutralization before ELISA measurement; however, in vivo, a number of putative activation mechanisms have been identified, including plasmin and the urokinase plasminogen activator system, the extracellular matrix molecule thrombospondin-1, αVβ6-integrin (10, 26, 35), and most recently a matrix metalloproteinase (MMP)-9/CD44 complex (36).
Despite the differences in isoform synthesis and function detected by ELISA, one similarity was evident. Damage to the bronchial epithelial monolayer allowed the detection of active TGF-β1 in the culture supernatants at a peak of 2 h after wounding, and a similar profile was seen for TGF-β2. Furthermore, TGF-β2 was only observed in its active form during the early stages of the repair process. This transient accumulation was only seen in cultures in which serum-containing medium was used, indicating that latent TGF-β in the serum was the substrate for activation. This was confirmed after damage to the monolayer in the absence of serum, where the pattern of release of both TGF-β1 and TGF-β1 was different from that in the presence of serum. In particular, the peak in active TGF-β1 and TGF-β2 demonstrated in the presence of serum was absent after serum removal. The use of serum-containing media in these systems reflects the in vivo situation where immunocytochemistry has shown that the majority of TGF-β present in the bronchial airway is found in the subepithelial compartment (29), probably complexed with fibrillary latent TGF-β binding proteins (11). Therefore, damage to a bronchial epithelial monolayer enhances the ability of cells to convert latent TGF-β1 and TGF-β2 into their active forms. This has recently been shown graphically in a coculture system where damage to guinea pig tracheal epithelial cells induced the loss of matrix-bound TGF-β1 on the underlying human amniotic basement membrane (25). The specific identity of these activation molecules has not been elucidated fully and will likely differ between isoforms and between organs. αVβ6-Integrin, which is upregulated at the wound edge of keratinocyte monolayers (18), has been shown to activate TGF-β1 in vitro, binding through an RGD sequence on the TGF-β1 LAP, a sequence not present on TGF-β2 LAP (26). Similarly, both MMP-9 and CD44 have an increased expression during repair of the bronchial epithelium (7, 22), with the complex of molecules at the cell surface showing greater activation of TGF-β2 than TGF-β1.
The consequences of this concentration of active TGF-β1 at the wound edge were investigated using time-lapse microscopy. In line with previous studies using dispersed cultures of mink lung epithelial cells and cultures of primary rabbit tracheal cells, TGF-β1 on a wounded human epithelial monolayer induced an increased migratory speed (6, 38). However, the present study showed that the concentration of TGF-β1 sufficient to establish an increased speed could be reduced to 250 pg/ml. Furthermore, TGF-β2 at any concentration showed no change in migration different from untreated control cultures.
By the use of time lapse, we could observe that, in TGF-β1-stimulated wounded monolayers, migration speed progressively increased up to 4 h, thereafter plateauing at an enhanced level. Measurement of the distance traveled during this time indicated that although some cell spreading occurred, the increase in the migratory speed of the epithelial layer was predominately the result of cell migration. Moreover, the total distance traveled by the monolayer (240 μm) was >10-fold greater than the average cell diameter of the 16HBE 14o− cell line (∼20 μm). The importance of continued TGF-β1 stimulation during this enhanced but stable migration was demonstrated by the neutralization of TGF-β1 in the system. After initial stimulation with an optimal concentration of TGF-β1 (3 h), an anti-TGF-β1 antibody reduced the rate of migration to the level of cultures without addition of TGF-β1. Therefore, TGF-β1 but not TGF-β2 can induce an increased migratory speed for the wounded epithelial monolayer; furthermore, this increased speed was dependent on the continued presence of active TGF-β1 in the system. This is, to our knowledge, the first demonstration of a differential migration response to TGF-β isoforms in bronchial epithelial cells. However, an enhanced migration to TGF-β1 has previously been shown in bovine aortic smooth muscle, and an enhanced migration to TGF-β2 has been shown in bovine aortic endothelial cells (23). This differential response has been linked to the relative expression of the type III TGF-β receptor (33). Furthermore, the X-ray crystal structure of the two isoforms reveals differences in a region of the molecule implicated in receptor binding (19).
TGF-β1 is recognized for its ability to inhibit cell proliferation in most cell types (17); however, there was neither a reduction nor enhancement of proliferation in the current study. This does not reflect a lack of responsiveness to TGF-β1, which has been suggested to contribute to carcinogenesis (30), because application of TGF-β1 enhanced cell migration. Cell proliferation has been shown to assist in the closure of chemically wounded primary epithelial monolayers 48 h after damage at a site 160–400 μm distal to the wound edge. The use of time-lapse analysis of wounded monolayers after inhibition of proliferation with mitomycin C in this study demonstrated that proliferation of the monolayer was not responsible for the increased migration observed. In contrast, the enhanced migration after inhibition of proliferation is consistent with previous studies where induction of G1 arrest in mink lung epithelial cells increased the migratory response to TGF-β1 (38). In addition, by phase-contrast microscopy, few cells could be observed in the process of division during the time course (Fig. 1). Therefore, active TGF-β1 at a similar concentration to that produced in damaged cultures could enhance and thereafter sustain the rate of migration of a wounded monolayer.
Bronchial epithelial repair is essential to normal airway function for maintenance of the barrier between the external and internal milieu of the lung after damage induced by inhaled pathogens, pollutants, or allergen. It has been established that asthmatic airways contain more evidence of epithelial damage than airways from nonasthmatic patients (3), as indicated by the presence of creola bodies in induced sputa and bronchial lavage samples (27). Therefore, by inference, asthmatic airways contain a greater degree of epithelial repair. It is clear from this study that bioactive TGF-β1 can be released from its latent form after damage to an epithelial cell monolayer and upregulation of activating molecules. A recent study has shown that chemical and mechanical damage of 16HBE 14o−monolayers can induce myofibroblast proliferation (37), partly through growth factors. The current study reinforces and augments this by providing evidence that mechanical damage and repair of the bronchial epithelial monolayer can induce the synthesis and activation of growth factors, in this case the TGF-β isoforms, and that activation increases the speed of repair of the monolayer. This corresponds with skin wound repair where blockade of the TGF-β isoforms TGF-β1 and TGF-β2 or injection of TGF-β3 in the wound site prolongs the wound repair but induces the production of matrix and alignment of fibers similar to normal skin (34). The release of TGF-β after bronchial damage initiated by inhalation of allergen (14) could explain the detection of increased TGF-β in the bronchoalveolar lavage fluid of asthmatic patients, both basally and 24 h after allergen challenge (28).
In summary, this report highlights the divergence in function between TGF-β1 and TGF-β2 in terms of bronchial epithelial repair. Activation mechanisms that facilitate the conversion of latent to active forms of TGF-β may be enhanced at an early stage in the repair process. However, although active forms of both TGF-β2 and TGF-β1 are produced during wound repair, only TGF-β1 increases the speed of epithelial repair.
This work was supported by Medical Research Council Program Grant G8604034, Rhone-Poulenc Rorer, United Kingdom, and Hope Charity.
Address for reprint requests and other correspondence: W. J. Howat, Div. of Respiratory, Cell, and Molecular Biology, MP810, Level D, Centre Block, Southampton General Hospital, SO16 6YD United Kingdom (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society