Modulation of Fibronectin Production of Bovine Bronchial Epithelial Cells by Transforming Growth Factor-S Debra J. Romberger, Joe D. Beckmann, Lorene Claassen, Ronald F. Ertl, and Stephen I. Rennard Department of Internal Medicine, University of Nebraska Medical Center, Omaha, Nebraska
Regulation of airway repair after injury is poorly understood but is thought to be important in the development of airway diseases such as chronic bronchitis and asthma. There is evidence that fibronectin (Fn), an extracellular matrix glycoprotein, has a role in repair processes. In addition, transforming growth factor{3 (TGF-{3) is also likely involved in wound healing and is known to influence extracellular matrix constituents in other cell systems. We postulated that TGF-{3 may effect airway repair by modulating Fn production from airway epithelial cells. To examine this hypothesis, we studied the effect of TGF-{31 on Fn production by bovine bronchial epithelial cells in culture. Fn, released into the media of cultures exposed to TGF-{3t, increased in a dose- and time-responsive fashion. Fn in the cell layer also increased in response to TGF-{31. De novo protein synthesis was demonstrated by an increase in [3sS]methionine incorporation into Fn immunoprecipitated from media of TGF-{3-treated cultures. TGF-{31 also induced an increase in expression of Fn mRNA from cultured bronchial epithelial cells, suggesting that TGF-{3 modulates Fn production of these cells, at least in part, through modulation of Fn gene expression. These data support a role for TGF-{3 in airway repair through modulation of Fn production by airway epithelium.
Fibrosis of the airways is an important feature of persistent asthma and chronic bronchitis (1, 2). Both syndromes are characterized by repeated episodes of airway inflammation, and the subsequent repair that occurs after each injury is likely to contribute to the development of such anatomic changes as fibrosis. Although the mechanisms involved with injury and repair are not well understood, it is thought that migration, attachment, proliferation, and differentiation of airway epithelial cells at a site of injury playa role in subsequent repair (3-5). Fibronectin (Fn), an extracellular matrix (ECM) glycoprotein, has suggested roles in such cellular processes (6). In addition, Fn has been documented to be present during wound repair of epidermal injuries in vivo (7-10). Therefore, it is likely that Fn may be involved in the repair of injured airways. Bronchial epithelial cells are known to produce Fn (11, 12). However, the mechanisms controlling the regulation of Fn production by bronchial epithelial cells have not been investigated. The regulation of Fn production in fibroblasts, however, has been studied, and several growth factors and
(Received in original form February 12, 1991 and in final form February 11, 1992) Address correspondence to: Dr. Debra 1. Romberger, Pulmonary and Critical Care Medicine Section, Department of Internal Medicine, 600 South 42nd St., Omaha, NE 68198-2465. Abbreviations: extracellular matrix, ECM; epithelial lining fluid, ELF; enzyme-linked immunosorbent assay, ELISA; fibronectin, Fn; Eagle's minimal essential medium, MEM; phosphate-buffered saline, PBS; sodium dodecyl sulfate, SDS; transforming growth factor-S, TGF-~. Am. J. Respir. Cell Mol. BioI. Vol. 7. pp. 149-155, 1992
hormones have been identified that stimulate Fn expression (13). Transforming growth factor-S (TGF-{3) is known to influence Fn production in fibroblasts and other cell types and is also thought to play a role in wound healing in vivo (14, 15). We postulated that TGF-{3 may be involved in airway repair by regulating the production of Fn from bronchial epithelial cells. To investigate our hypothesis, we examined the effect of exogenous TGF-{31 on cultured bovine bronchial epithelial cells. We determined that TGF-{31 increases the release of Fn into the media of bronchial epithelial cell cultures, increases the cell-associated Fn, increases de novo synthesis ofFn, and increases Fn mRNA in bovine bronchial epithelial cells.
Materials and Methods Preparation of Bovine Bronchial Epithelial Cells Bovine bronchial epithelial cells were obtained by a modification of the method ofWu and Smith (16). Briefly, the lungs of freshly slaughtered cattle were obtained from a local slaughterhouse. Bronchi were dissected from the lungs, cut into large pieces, and trimmed of connective tissue. Bronchi were incubated overnight at 4 0 C in calcium-free Eagle's minimal essential medium (MEM) (GIBCO, Grand Island, NY), which contained 0.1% bacterial protease (type XIV; Sigma Chemical Co., St. Louis, MO). The following day, the bronchial lumens were washed with MEM containing 10% fetal calf serum (Biofluids, Rockville, MD) to detach the bronchial epithelial cells. The bronchial epithelial cells were then washed once with MEM with 10% fetal calf serum, filtered through 100-p,m
150
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
nitex mesh (Tetko, Elmsford, NY), and resuspended in LHC-8 media (see below). The cells were plated at 5.0 X l O' cells/plate on 35-mm tissue culture plates (Becton Dickinson, Lincoln Park, NJ) in 2 ml media/plate, and incubated at 37 0 C in 5% CO 2/95% air. Media and Supplements LHC-8 media (17) contains LHC basal media (Biofluids), 5 j.tg/ml bovine insulin (Sigma), 5 ng/ml epidermal growth factor (Sigma), 10 j.tg/ml bovine transferrin (Sigma), 10 nM 3,3~5 triiodothyronine (Biofluids), bovine pituitary extract (50 j.tg protein/ml; Pel Freeze, Rogers, AK), 0.2 j.tM hydrocortisone (Biofluids), 5 j.tM phosphoethanolamine/ethanolamine (Sigma), trace elements, 0.11 mM calcium (Fisher, Springfield, NJ), 50 U/ml penicillin and streptomycin (GIBCO), and 2 ug/ml fungizone (GIBCO). Bovine Fn Enzyme-linked Immunosorbent Assay (ELISA) An indirect inhibition ELISA for bovine Fn was used to quantitate Fn from the cultures of the bovine bronchial epithelial cells. Bovine Fn was purified from bovine serum using the technique of Engvall and Ruoslahti (18). Bovine Fn was diluted in Voller's buffer (0.5 j.tg/ml) and used to coat 96-well, flatbottomed plates (Immulon 2, Flat Bottom Plates; Dynatech Laboratories, Chantilly, VA) with 200 j.tllwell. These plates were then stored overnight at 4 0 C in a humidified chamber. Samples were diluted 1:1 to 1:27 in 96-well, roundbottomed polystyrene plates (Linbro/Titertech, McLean, VA) with phosphate-buffered saline (PBS)-Tween (0.08% Tween). The samples were incubated overnight at 4 0 C in a humidified chamber with a 1:400 dilution of rabbit antibovine Fn antiserum. The following day, the flat-bottomed plates were washed 3 times with PBS-Tween, and the samples were transferred to the flat-bottomed plates. After a 30min incubation at room temperature, the flat-bottomed plates were again washed 3 times with PBS-Tween. Two hundred microliters/well of a 1:1,000dilution of goat anti-rabbit IgG (ICN ImmunoBiologicals, Lisle, IL) conjugated with horseradish peroxidase were added to the plates and allowed to incubate for 90 min at room temperature. The plates were again washed with PBS-Tween 3 times, and 200 j.tl/well of o-phenylenediamine substrate (10 mg dissolved in 1 ml methanol/100 ml H20) with 10 j.tl of 30% H202 were added. The enzyme reaction was stopped with 28 j.tl/well of 4 N H2S04 , The absorbance was then measured at 492 nm using an ELISA reader (Bio-Rad Model 2550 EIA Reader; Bio-Rad Laboratories, Richmond, CA). Data were analyzed using the algorithm of Rodbard (19) by a Macintosh computer, which was interfaced to the ELISA reader. Western blot of bovine bronchial epithelial cell culture supernatants utilizing the same antibodies as are used in the ELISA demonstrated a single band compatible with Fn. Increasing quantities of Fn by ELISA correlate to increasing intensity of the Fn band (data not shown). Dose Response and Time Course of TGF-IJ Stimulation of Bronchial Epithelial Cell Fn Release To examine the dose response oflGF-J3 on bronchial epithelial cell Fn release, bovine bronchial epithelial cells obtained
in the manner described above were plated on 35-mm tissue culture plates at 0.5 X 106 cells/plate in LHC-8 media. The cells were maintained in culture, and the media were changed every 2 days. When plates were nearly confluent (after about 1 wk), the media were removed and replaced with fresh LHC-8 media containing 0, 0.5, 5.0, 50, and 500 pM TGF-J31 (R&D Systems, Minneapolis, MN). All experiments were done in triplicate. After 48 h, the media were harvested and stored at -20 0 C until Fn was assayed by ELISA. After removing the media for Fn assay, the cell layers were rinsed with isotonic buffered saline without azide (American Scientific Products, McGaw Park, IL) and then exposed to 0.05% trypsin (GIBCO) to detach the cells from the plate, and the cells were counted by Coulter Counter. This allowed Fn to be expressed in a quantity per cell per time. The time course of lGF-J31 stimulation of bronchial epithelial cell Fn release was examined by culturing cells as described above and then exposing triplicate dishes to fresh media containing 500 pM TGF-J31 for 6, 12, 24, and 48 h. Effect of TGF-J3 on Released and Cell-associated Fn Because Fn is located in the ECM, it is possible that an increase in Fn in the media in response to TGF-J31 could represent only an increase in the release of Fn from a preformed pool of Fn within the cell layer. Toexamine the effectof TGF131 on both cell-associated Fn and Fn released into the media of cell cultures, bronchial epithelial cell cultures were exposed to 0 or 100 pM TGF-J31 for 48 h. The media were harvested, and cell layers were scraped into 0.05 M Na3P04 with 2 mM phenylmethylsulfonyl fluoride (Sigma) and 1% Triton X-100 (Sigma) and sonicated using a microtip for 2 min at room temperature. Both media and cell layers were then assayed for Fn by ELISA.
De Novo Production of Fn in Response to TGF-,B De novo production of Fn was investigated by studying the incorporation of [3sS]methionine into newly synthesized Fn and by identifying newly synthesized Fn with immune precipitation, sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis, and autoradiography. To accomplish this, nearly confluent bronchial epithelial cell cultures were exposed to media containing either 0 or 250 pM TGF-J31 for 48 h. The media were then aspirated and replaced with methionine-free Dulbecco's modified Eagle's medium (GIBCO) containing 20 j.tCi/ml of pSS]methionine (10 mCi/ml; Amersham, Arlington Heights, IL) for 24 h. Media and cell layers were then harvested. One-milliliter samples of the media were precipitated by the addition of 100 j.tl of 100% (wt/vol) trichloroacetic acid (Fisher Scientific, Fair Lawn, NJ). One-milliliter samples of the media were immunoprecipitated by adding 2 j.tl of rabbit anti-bovine Fn antiserum for 2 h at 4 0 C, followed by overnight incubation with 8 j.tl of goat anti-rabbit IgG (Miles Scientific, Naperville, IL). The resulting precipitations were layered over a sucrose cushion (0.5 ml of 0.5 M sucrose and 1.0ml of 1.0M sucrose) and centrifuged (2,000 rpm for 10 min). The pellets;were washed 2 times with PBS-Tween (0.08% Tween) and then dissolved in gel sample buffer. One-milliliter samples of media and cell layer extract were also incubated with gelatinSepharose 4B (Pharmacia, Uppsala, Sweden) overnight to
Romberger, Beckmann, Claassen et al.: Fn Modulation of Bovine Bronchial Epithelial Cells by TGF-{3
recover Fn from the samples. The gelatin-Sepharose was washed 3 times with PBS-Tween and then placed in gel sample buffer. After heat denaturation and reduction with 2-mercaptoethanol, samples were applied to a 7.5 % polyacrylamide gel containing 0.2 % SDS for electrophoresis. The resulting gels were dried, and autoradiography was then performed by standard methods (20).
151
7.0 6.0
..
I Q
5.0
~
)C
c .....
;.c: u .....
•= c.
4.0
.cD -c
3.0
OU
Effect of TGF-{3 on Fn mRNA Levels from Bronchial Epithelial Cells To begin to examine the mechanisms by which TGF-{31 increases Fn production from bronchial epithelial cells, we investigated whether TGF-{3 affected the levels of Fn mRNA from bronchial epithelial cells. Cells were obtained and cultured as described. When the cells were nearly confluent, the media were changed to contain LHC-8 alone or LHC-8 with 2.5, 25, or 250 pM TGF-{31. Cultures with each of the above conditions were maintained for 6, 12, 24, and 48 h. Cell conditions were studied in triplicate. The media were harvested, stored at -20° C and later assayed for Fn by ELISA. The RNA was extracted using the method of Chomczynski and Sacchi (21), utilizing guanidine thiocyanate extraction. Ten micrograms of RNA from each condition were electrophoresed on an 0.8 % agarose/formaldehyde gel, and Northern blot transfer to nitrocellulose was performed (22). A human Fn cDNA probe (gift of F. Baralle, Oxford, UK) was oligolabeled with [3zP]dCTP (3,000 Cil mmol; Amersham) for 6 h (23). The blot was then hybridized with the probe for 18 h at 42 ° C. The blot was washed once in 2 X SSPE/O.1 % SDS at room temperature, twice in O.1x SSPE/O.1 % SDS at room temperature, and once in 0.1X SSPE/O.1 % SDS at 42° C SSPE is 0.15 M NaCI, 0.01 M NaH zP0 4 , 0.001 M EDTA [pH 7.4]). Autoradiography was then performed. Labeling and hybridization with the mouse {35 tubulin cDNA (from Don Cleveland, Baltimore, MD) was also performed as above (24).
nx
Statistical Methods Significance was determined by using Student's unpaired t test.
u..--
2.0
1.0
o
5.0
0.5
50.0
500.0
TGF-8 dose, pM
Figure 1. Dose response of TGF-{3 on bronchial epithelial cell Fn release. Fn was released into the media of bovine bronchial epithelial cells in culture in response to 0, 0.5, 5.0, 50, and 500 pM TGF-{31. The concentrations of TGF-{31 used are represented on the x-axis. Fn was assayed by ELISA and expressed as ng/cell/h, represented on the y-axis. Cell numbers were obtained after the media were harvested by trypsinizing the cells and counting by Coulter Counter. Values are means of triplicate plates, with error bars representing standard deviation.
ure 2). This increase in Fn persisted at 48 h. It would appear that the mechanisms controlling the influence of TGF-{31 on Fn release from bronchial epithelial cells required several hours before becoming maximally effective. This is suggestive of de novo protein biosynthesis with or without gene activation rather than stimulated release of a preformed pool of Fn. Effect of TGF-{3 on Released and Cell-associated Fn Because Fn is a constituent of the ECM, it is possible that the increase in Fn with TGF-131 found in the media of the cultures might represent only a change in distribution of Fn.
7.0
Results Dose Response and Time Course of TGF-{3 Stimulation of Bronchial Epithelial Cell Fn Release The amount of Fn in the supernatants of bronchial epithelial cell cultures increased as the cultures were exposed to increasing concentrations of TGF-{31 (Figure 1). There was a 3-fold increase in Fn in cultures exposed to 50 pM TGF-{31' and a 4-fold increase in response to 500 pM TGF-{31. The half-maximal response of Fn release was at a concentration of approximately 10 pM TGF-{31. Further experiments demonstrated (data not shown) that a dose of 250 pM TGF{31 produced nearly the same response as 500 pM TGF-{31. These data demonstrate that TGF-{31 stimulated the release of Fn from bronchial epithelial cells in culture in a doseresponsive manner. There is only a minimal increase in Fn release at 6 and 12 h. By 24 h, however, the amount of Fn released into the supernatant of bronchial epithelial cell cultures exposed to 500 pM TGF-{31 is significantly increased (P < 0.007) (Fig-
6.0
. I Q
M
.-
c-.c:
u:::
c·
o~
"0
5.0
4.0 3.0
~..s
2.0 1.0
6
12
24
48
Time, h
Figure 2. Time course ofTGF-{3 stimulation of bronchial epithelial cell Fn release. Fn released into the media of bovine bronchial epithelial cell cultures at 0, 6, 12, 24, and 48 h after exposure to 500 pM TGF-{31 was assayed by ELISA and expressed as ng/cell/h. Fn is represented on the y axis, and the time (in hours) of TGF-{3 exposure is represented on the x axis. Values are means of triplicate plates, with error bars representing standard deviation.
152
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
A
8
3000
~ ~ :J
.ti,
200 kD -
2000
c: c:
U QI c: ~
1000
116 kD
D U.
97kD
a Media t-)
Media (.j
Cell layer (-) Cell
--
layer (+)
Figure 3. Effectof TGF-13 on released and cell-associated Fn. Fn assayed in the media and cell layer of bovine bronchial epithelial cell cultures with (hatched bars) and without (solid bars) exposure to 100pM TGF-131 for 48 h. The celllayer wasextracted by scraping the cells from the plates into 0.05 M Na3PO. with 2 mM phenylmethylsulfonyl fluoride and 1% Triton, followed by sonication. Fn wasassayed byELISA and expressed in ng/culture (y-axis), Values are mean ± SD of six plates.
In particular, TGF-I3, might cause a release of Fn from a previously formed pool without increasing the amount of Fn present in the culture. In order to examine if TGF-I3, caused only a release of Fn from a previously formed pool, Fn levels in both the supernatant and cell layer of bronchial epithelial cell cultures after exposure to TGF-I3, were assayed. TGF-I3, increased the amount of Fn in the cell layer of bronchial epithelial cell cultures as well as the amount of Fn released into the media of the cultures (Figure 3) . The amount of cell layer-associated Fn is smaller than the amount of Fn found in the media of the cultures both with and without exposure to TGF-I3,. However, the amount of Fn associated with the cell layer increased by approximately 3-fold when cultures were exposed to 100 pM TGF-I3,. This demonstrates that TGF-131 causes both an increase of Fn released into the culture media as well as an increase in the amount of Fn present in the cell layer. An increase in Fn in both the cell layer and media suggests that TGF-I3, increases total production of Fn, rather than causing release from a preformed pool.
De Novo Production of Fn in Response to TGF-,6 De novo Fn production was evaluated by examining radiolabeled amino acid incorporation in cultures exposed to TGF13, . From the media of bronchial epithelial cells cultured in the presence of [35S] methionine, Fn was immunoprecipitated with anti-bovine Fn antiserum as demonstrated by an appropriate sized band on the autoradiograph (Figure 4). This band increased in density in cultures exposed to 250 pM TGF-I3, when compared with control cultures, suggesting that more Fn was produced in cultures treated with TGF-I3,. Similarly, there was an increase in Fn in the cell layer of cultures exposed to TGF-131 evaluated by radiolabeled amino acid incorporation. An autoradiograph of cell layer extracts showed an increase in the band of the appropriate size for Fn with exposure of cultures to TGF-,61 (Figure 5) .
66kD -
Figure 4. Effect of TGF-13 on radio1abeled Fn in media of bovine bronchial epithelial cells in culture. Bovine bronchial epithelial cells in culture were exposed to either 0 or 250 pM TGF-I3 , for 48 h. Media were then changed to include 20 ~Ci/rnl of [35S]methionine for 24 h. One-milliliteraliquots of harvested media were imrnunoprecipitated by anti-bovine Fn antibody before SDS-acrylamidegel electrophoresis, followed byautoradiography. (A) Radiolabeled Fn imrnunoprecipitated from the media of cultures not exposedto TGF-,6. (B)Radiolabeled Fn imrnunoprecipitated fromthe mediaof cultures exposed to 250pM TGF-I3. The molecularweight standards appear to the left.
Fn was not the only protein in the cell layer isolated by gelSepharose binding that was increased by TGF-,6,. Interestingly, several bands on the autoradiograph also decreased in intensity, suggesting that TGF-I3, increased the amount of specific proteins such as Fn rather than nonspecifically increasing protein synthesis. Effect of TGF-13 on Fn mRNA Levels from Bronchial Epithelial Cells
Dose responseofTGF-13 on FnmRNA expression. Twentyfour hours of TGF-131exposure increased the mRNA levels for Fn in bronchial epithelial cells (Figure 6) in a doseresponsive fashion. The constitutive expression of Fn mRNA is demonstrated in the control cultures: an appropriate 8.6-kb mRNA band hybridizes with a 32P-Iabeled human cDNA Fn probe. With exposure to increasing concentrations of TGF-I3" there is increased expression of Fn mRNA. Cultures exposed to the highest concentration of TGF-I3" 250 pM, demonstrated the most intense Fn mRNA band . This suggests that TGF-I3, increases the production of Fn from bronchial epithelial cells at least in part by regulating Fn mRNA levels. TIme course ofTGF-,6 on FnmRNA expression. The effect of TGF-I3, on the Fn mRNA levels is present minimally after 6 h of expo sure of the cultures to TGF-131 (Figure 7). The peak effect is observed at 12 to 24 h of exposure and appears to be declining slightly by 48 h . This timing is consistent with the observed timing of the increase in Fn released
Romberger, Beckmann, Claassen et al. : Fn Modulation of Bovine Bronchial Epithelial Cells by TGF-i3
153
- TGF·/3 + TGF-/3
FN mRNA Fn _ 200kD
116 kD 97kD
-
-
o
-
2.5
25 pM
250
Regulation of airway repair after injury is poorly understood but is thought to be important in the development of airway diseases such as chronic bronchitis and asthma. There is evidence that fibronectin (Fn), an extracellular matrix glycoprotein, has a role in repair processes. In addition, transforming growth factor{3 (TGF-{3) is also likely involved in wound healing and is known to influence extracellular matrix constituents in other cell systems. We postulated that TGF-{3 may effect airway repair by modulating Fn production from airway epithelial cells. To examine this hypothesis, we studied the effect of TGF-{31 on Fn production by bovine bronchial epithelial cells in culture. Fn, released into the media of cultures exposed to TGF-{3t, increased in a dose- and time-responsive fashion. Fn in the cell layer also increased in response to TGF-{31. De novo protein synthesis was demonstrated by an increase in [3sS]methionine incorporation into Fn immunoprecipitated from media of TGF-{3-treated cultures. TGF-{31 also induced an increase in expression of Fn mRNA from cultured bronchial epithelial cells, suggesting that TGF-{3 modulates Fn production of these cells, at least in part, through modulation of Fn gene expression. These data support a role for TGF-{3 in airway repair through modulation of Fn production by airway epithelium.
Fibrosis of the airways is an important feature of persistent asthma and chronic bronchitis (1, 2). Both syndromes are characterized by repeated episodes of airway inflammation, and the subsequent repair that occurs after each injury is likely to contribute to the development of such anatomic changes as fibrosis. Although the mechanisms involved with injury and repair are not well understood, it is thought that migration, attachment, proliferation, and differentiation of airway epithelial cells at a site of injury playa role in subsequent repair (3-5). Fibronectin (Fn), an extracellular matrix (ECM) glycoprotein, has suggested roles in such cellular processes (6). In addition, Fn has been documented to be present during wound repair of epidermal injuries in vivo (7-10). Therefore, it is likely that Fn may be involved in the repair of injured airways. Bronchial epithelial cells are known to produce Fn (11, 12). However, the mechanisms controlling the regulation of Fn production by bronchial epithelial cells have not been investigated. The regulation of Fn production in fibroblasts, however, has been studied, and several growth factors and
(Received in original form February 12, 1991 and in final form February 11, 1992) Address correspondence to: Dr. Debra 1. Romberger, Pulmonary and Critical Care Medicine Section, Department of Internal Medicine, 600 South 42nd St., Omaha, NE 68198-2465. Abbreviations: extracellular matrix, ECM; epithelial lining fluid, ELF; enzyme-linked immunosorbent assay, ELISA; fibronectin, Fn; Eagle's minimal essential medium, MEM; phosphate-buffered saline, PBS; sodium dodecyl sulfate, SDS; transforming growth factor-S, TGF-~. Am. J. Respir. Cell Mol. BioI. Vol. 7. pp. 149-155, 1992
hormones have been identified that stimulate Fn expression (13). Transforming growth factor-S (TGF-{3) is known to influence Fn production in fibroblasts and other cell types and is also thought to play a role in wound healing in vivo (14, 15). We postulated that TGF-{3 may be involved in airway repair by regulating the production of Fn from bronchial epithelial cells. To investigate our hypothesis, we examined the effect of exogenous TGF-{31 on cultured bovine bronchial epithelial cells. We determined that TGF-{31 increases the release of Fn into the media of bronchial epithelial cell cultures, increases the cell-associated Fn, increases de novo synthesis ofFn, and increases Fn mRNA in bovine bronchial epithelial cells.
Materials and Methods Preparation of Bovine Bronchial Epithelial Cells Bovine bronchial epithelial cells were obtained by a modification of the method ofWu and Smith (16). Briefly, the lungs of freshly slaughtered cattle were obtained from a local slaughterhouse. Bronchi were dissected from the lungs, cut into large pieces, and trimmed of connective tissue. Bronchi were incubated overnight at 4 0 C in calcium-free Eagle's minimal essential medium (MEM) (GIBCO, Grand Island, NY), which contained 0.1% bacterial protease (type XIV; Sigma Chemical Co., St. Louis, MO). The following day, the bronchial lumens were washed with MEM containing 10% fetal calf serum (Biofluids, Rockville, MD) to detach the bronchial epithelial cells. The bronchial epithelial cells were then washed once with MEM with 10% fetal calf serum, filtered through 100-p,m
150
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
nitex mesh (Tetko, Elmsford, NY), and resuspended in LHC-8 media (see below). The cells were plated at 5.0 X l O' cells/plate on 35-mm tissue culture plates (Becton Dickinson, Lincoln Park, NJ) in 2 ml media/plate, and incubated at 37 0 C in 5% CO 2/95% air. Media and Supplements LHC-8 media (17) contains LHC basal media (Biofluids), 5 j.tg/ml bovine insulin (Sigma), 5 ng/ml epidermal growth factor (Sigma), 10 j.tg/ml bovine transferrin (Sigma), 10 nM 3,3~5 triiodothyronine (Biofluids), bovine pituitary extract (50 j.tg protein/ml; Pel Freeze, Rogers, AK), 0.2 j.tM hydrocortisone (Biofluids), 5 j.tM phosphoethanolamine/ethanolamine (Sigma), trace elements, 0.11 mM calcium (Fisher, Springfield, NJ), 50 U/ml penicillin and streptomycin (GIBCO), and 2 ug/ml fungizone (GIBCO). Bovine Fn Enzyme-linked Immunosorbent Assay (ELISA) An indirect inhibition ELISA for bovine Fn was used to quantitate Fn from the cultures of the bovine bronchial epithelial cells. Bovine Fn was purified from bovine serum using the technique of Engvall and Ruoslahti (18). Bovine Fn was diluted in Voller's buffer (0.5 j.tg/ml) and used to coat 96-well, flatbottomed plates (Immulon 2, Flat Bottom Plates; Dynatech Laboratories, Chantilly, VA) with 200 j.tllwell. These plates were then stored overnight at 4 0 C in a humidified chamber. Samples were diluted 1:1 to 1:27 in 96-well, roundbottomed polystyrene plates (Linbro/Titertech, McLean, VA) with phosphate-buffered saline (PBS)-Tween (0.08% Tween). The samples were incubated overnight at 4 0 C in a humidified chamber with a 1:400 dilution of rabbit antibovine Fn antiserum. The following day, the flat-bottomed plates were washed 3 times with PBS-Tween, and the samples were transferred to the flat-bottomed plates. After a 30min incubation at room temperature, the flat-bottomed plates were again washed 3 times with PBS-Tween. Two hundred microliters/well of a 1:1,000dilution of goat anti-rabbit IgG (ICN ImmunoBiologicals, Lisle, IL) conjugated with horseradish peroxidase were added to the plates and allowed to incubate for 90 min at room temperature. The plates were again washed with PBS-Tween 3 times, and 200 j.tl/well of o-phenylenediamine substrate (10 mg dissolved in 1 ml methanol/100 ml H20) with 10 j.tl of 30% H202 were added. The enzyme reaction was stopped with 28 j.tl/well of 4 N H2S04 , The absorbance was then measured at 492 nm using an ELISA reader (Bio-Rad Model 2550 EIA Reader; Bio-Rad Laboratories, Richmond, CA). Data were analyzed using the algorithm of Rodbard (19) by a Macintosh computer, which was interfaced to the ELISA reader. Western blot of bovine bronchial epithelial cell culture supernatants utilizing the same antibodies as are used in the ELISA demonstrated a single band compatible with Fn. Increasing quantities of Fn by ELISA correlate to increasing intensity of the Fn band (data not shown). Dose Response and Time Course of TGF-IJ Stimulation of Bronchial Epithelial Cell Fn Release To examine the dose response oflGF-J3 on bronchial epithelial cell Fn release, bovine bronchial epithelial cells obtained
in the manner described above were plated on 35-mm tissue culture plates at 0.5 X 106 cells/plate in LHC-8 media. The cells were maintained in culture, and the media were changed every 2 days. When plates were nearly confluent (after about 1 wk), the media were removed and replaced with fresh LHC-8 media containing 0, 0.5, 5.0, 50, and 500 pM TGF-J31 (R&D Systems, Minneapolis, MN). All experiments were done in triplicate. After 48 h, the media were harvested and stored at -20 0 C until Fn was assayed by ELISA. After removing the media for Fn assay, the cell layers were rinsed with isotonic buffered saline without azide (American Scientific Products, McGaw Park, IL) and then exposed to 0.05% trypsin (GIBCO) to detach the cells from the plate, and the cells were counted by Coulter Counter. This allowed Fn to be expressed in a quantity per cell per time. The time course of lGF-J31 stimulation of bronchial epithelial cell Fn release was examined by culturing cells as described above and then exposing triplicate dishes to fresh media containing 500 pM TGF-J31 for 6, 12, 24, and 48 h. Effect of TGF-J3 on Released and Cell-associated Fn Because Fn is located in the ECM, it is possible that an increase in Fn in the media in response to TGF-J31 could represent only an increase in the release of Fn from a preformed pool of Fn within the cell layer. Toexamine the effectof TGF131 on both cell-associated Fn and Fn released into the media of cell cultures, bronchial epithelial cell cultures were exposed to 0 or 100 pM TGF-J31 for 48 h. The media were harvested, and cell layers were scraped into 0.05 M Na3P04 with 2 mM phenylmethylsulfonyl fluoride (Sigma) and 1% Triton X-100 (Sigma) and sonicated using a microtip for 2 min at room temperature. Both media and cell layers were then assayed for Fn by ELISA.
De Novo Production of Fn in Response to TGF-,B De novo production of Fn was investigated by studying the incorporation of [3sS]methionine into newly synthesized Fn and by identifying newly synthesized Fn with immune precipitation, sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis, and autoradiography. To accomplish this, nearly confluent bronchial epithelial cell cultures were exposed to media containing either 0 or 250 pM TGF-J31 for 48 h. The media were then aspirated and replaced with methionine-free Dulbecco's modified Eagle's medium (GIBCO) containing 20 j.tCi/ml of pSS]methionine (10 mCi/ml; Amersham, Arlington Heights, IL) for 24 h. Media and cell layers were then harvested. One-milliliter samples of the media were precipitated by the addition of 100 j.tl of 100% (wt/vol) trichloroacetic acid (Fisher Scientific, Fair Lawn, NJ). One-milliliter samples of the media were immunoprecipitated by adding 2 j.tl of rabbit anti-bovine Fn antiserum for 2 h at 4 0 C, followed by overnight incubation with 8 j.tl of goat anti-rabbit IgG (Miles Scientific, Naperville, IL). The resulting precipitations were layered over a sucrose cushion (0.5 ml of 0.5 M sucrose and 1.0ml of 1.0M sucrose) and centrifuged (2,000 rpm for 10 min). The pellets;were washed 2 times with PBS-Tween (0.08% Tween) and then dissolved in gel sample buffer. One-milliliter samples of media and cell layer extract were also incubated with gelatinSepharose 4B (Pharmacia, Uppsala, Sweden) overnight to
Romberger, Beckmann, Claassen et al.: Fn Modulation of Bovine Bronchial Epithelial Cells by TGF-{3
recover Fn from the samples. The gelatin-Sepharose was washed 3 times with PBS-Tween and then placed in gel sample buffer. After heat denaturation and reduction with 2-mercaptoethanol, samples were applied to a 7.5 % polyacrylamide gel containing 0.2 % SDS for electrophoresis. The resulting gels were dried, and autoradiography was then performed by standard methods (20).
151
7.0 6.0
..
I Q
5.0
~
)C
c .....
;.c: u .....
•= c.
4.0
.cD -c
3.0
OU
Effect of TGF-{3 on Fn mRNA Levels from Bronchial Epithelial Cells To begin to examine the mechanisms by which TGF-{31 increases Fn production from bronchial epithelial cells, we investigated whether TGF-{3 affected the levels of Fn mRNA from bronchial epithelial cells. Cells were obtained and cultured as described. When the cells were nearly confluent, the media were changed to contain LHC-8 alone or LHC-8 with 2.5, 25, or 250 pM TGF-{31. Cultures with each of the above conditions were maintained for 6, 12, 24, and 48 h. Cell conditions were studied in triplicate. The media were harvested, stored at -20° C and later assayed for Fn by ELISA. The RNA was extracted using the method of Chomczynski and Sacchi (21), utilizing guanidine thiocyanate extraction. Ten micrograms of RNA from each condition were electrophoresed on an 0.8 % agarose/formaldehyde gel, and Northern blot transfer to nitrocellulose was performed (22). A human Fn cDNA probe (gift of F. Baralle, Oxford, UK) was oligolabeled with [3zP]dCTP (3,000 Cil mmol; Amersham) for 6 h (23). The blot was then hybridized with the probe for 18 h at 42 ° C. The blot was washed once in 2 X SSPE/O.1 % SDS at room temperature, twice in O.1x SSPE/O.1 % SDS at room temperature, and once in 0.1X SSPE/O.1 % SDS at 42° C SSPE is 0.15 M NaCI, 0.01 M NaH zP0 4 , 0.001 M EDTA [pH 7.4]). Autoradiography was then performed. Labeling and hybridization with the mouse {35 tubulin cDNA (from Don Cleveland, Baltimore, MD) was also performed as above (24).
nx
Statistical Methods Significance was determined by using Student's unpaired t test.
u..--
2.0
1.0
o
5.0
0.5
50.0
500.0
TGF-8 dose, pM
Figure 1. Dose response of TGF-{3 on bronchial epithelial cell Fn release. Fn was released into the media of bovine bronchial epithelial cells in culture in response to 0, 0.5, 5.0, 50, and 500 pM TGF-{31. The concentrations of TGF-{31 used are represented on the x-axis. Fn was assayed by ELISA and expressed as ng/cell/h, represented on the y-axis. Cell numbers were obtained after the media were harvested by trypsinizing the cells and counting by Coulter Counter. Values are means of triplicate plates, with error bars representing standard deviation.
ure 2). This increase in Fn persisted at 48 h. It would appear that the mechanisms controlling the influence of TGF-{31 on Fn release from bronchial epithelial cells required several hours before becoming maximally effective. This is suggestive of de novo protein biosynthesis with or without gene activation rather than stimulated release of a preformed pool of Fn. Effect of TGF-{3 on Released and Cell-associated Fn Because Fn is a constituent of the ECM, it is possible that the increase in Fn with TGF-131 found in the media of the cultures might represent only a change in distribution of Fn.
7.0
Results Dose Response and Time Course of TGF-{3 Stimulation of Bronchial Epithelial Cell Fn Release The amount of Fn in the supernatants of bronchial epithelial cell cultures increased as the cultures were exposed to increasing concentrations of TGF-{31 (Figure 1). There was a 3-fold increase in Fn in cultures exposed to 50 pM TGF-{31' and a 4-fold increase in response to 500 pM TGF-{31. The half-maximal response of Fn release was at a concentration of approximately 10 pM TGF-{31. Further experiments demonstrated (data not shown) that a dose of 250 pM TGF{31 produced nearly the same response as 500 pM TGF-{31. These data demonstrate that TGF-{31 stimulated the release of Fn from bronchial epithelial cells in culture in a doseresponsive manner. There is only a minimal increase in Fn release at 6 and 12 h. By 24 h, however, the amount of Fn released into the supernatant of bronchial epithelial cell cultures exposed to 500 pM TGF-{31 is significantly increased (P < 0.007) (Fig-
6.0
. I Q
M
.-
c-.c:
u:::
c·
o~
"0
5.0
4.0 3.0
~..s
2.0 1.0
6
12
24
48
Time, h
Figure 2. Time course ofTGF-{3 stimulation of bronchial epithelial cell Fn release. Fn released into the media of bovine bronchial epithelial cell cultures at 0, 6, 12, 24, and 48 h after exposure to 500 pM TGF-{31 was assayed by ELISA and expressed as ng/cell/h. Fn is represented on the y axis, and the time (in hours) of TGF-{3 exposure is represented on the x axis. Values are means of triplicate plates, with error bars representing standard deviation.
152
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
A
8
3000
~ ~ :J
.ti,
200 kD -
2000
c: c:
U QI c: ~
1000
116 kD
D U.
97kD
a Media t-)
Media (.j
Cell layer (-) Cell
--
layer (+)
Figure 3. Effectof TGF-13 on released and cell-associated Fn. Fn assayed in the media and cell layer of bovine bronchial epithelial cell cultures with (hatched bars) and without (solid bars) exposure to 100pM TGF-131 for 48 h. The celllayer wasextracted by scraping the cells from the plates into 0.05 M Na3PO. with 2 mM phenylmethylsulfonyl fluoride and 1% Triton, followed by sonication. Fn wasassayed byELISA and expressed in ng/culture (y-axis), Values are mean ± SD of six plates.
In particular, TGF-I3, might cause a release of Fn from a previously formed pool without increasing the amount of Fn present in the culture. In order to examine if TGF-I3, caused only a release of Fn from a previously formed pool, Fn levels in both the supernatant and cell layer of bronchial epithelial cell cultures after exposure to TGF-I3, were assayed. TGF-I3, increased the amount of Fn in the cell layer of bronchial epithelial cell cultures as well as the amount of Fn released into the media of the cultures (Figure 3) . The amount of cell layer-associated Fn is smaller than the amount of Fn found in the media of the cultures both with and without exposure to TGF-I3,. However, the amount of Fn associated with the cell layer increased by approximately 3-fold when cultures were exposed to 100 pM TGF-I3,. This demonstrates that TGF-131 causes both an increase of Fn released into the culture media as well as an increase in the amount of Fn present in the cell layer. An increase in Fn in both the cell layer and media suggests that TGF-I3, increases total production of Fn, rather than causing release from a preformed pool.
De Novo Production of Fn in Response to TGF-,6 De novo Fn production was evaluated by examining radiolabeled amino acid incorporation in cultures exposed to TGF13, . From the media of bronchial epithelial cells cultured in the presence of [35S] methionine, Fn was immunoprecipitated with anti-bovine Fn antiserum as demonstrated by an appropriate sized band on the autoradiograph (Figure 4). This band increased in density in cultures exposed to 250 pM TGF-I3, when compared with control cultures, suggesting that more Fn was produced in cultures treated with TGF-I3,. Similarly, there was an increase in Fn in the cell layer of cultures exposed to TGF-131 evaluated by radiolabeled amino acid incorporation. An autoradiograph of cell layer extracts showed an increase in the band of the appropriate size for Fn with exposure of cultures to TGF-,61 (Figure 5) .
66kD -
Figure 4. Effect of TGF-13 on radio1abeled Fn in media of bovine bronchial epithelial cells in culture. Bovine bronchial epithelial cells in culture were exposed to either 0 or 250 pM TGF-I3 , for 48 h. Media were then changed to include 20 ~Ci/rnl of [35S]methionine for 24 h. One-milliliteraliquots of harvested media were imrnunoprecipitated by anti-bovine Fn antibody before SDS-acrylamidegel electrophoresis, followed byautoradiography. (A) Radiolabeled Fn imrnunoprecipitated from the media of cultures not exposedto TGF-,6. (B)Radiolabeled Fn imrnunoprecipitated fromthe mediaof cultures exposed to 250pM TGF-I3. The molecularweight standards appear to the left.
Fn was not the only protein in the cell layer isolated by gelSepharose binding that was increased by TGF-,6,. Interestingly, several bands on the autoradiograph also decreased in intensity, suggesting that TGF-I3, increased the amount of specific proteins such as Fn rather than nonspecifically increasing protein synthesis. Effect of TGF-13 on Fn mRNA Levels from Bronchial Epithelial Cells
Dose responseofTGF-13 on FnmRNA expression. Twentyfour hours of TGF-131exposure increased the mRNA levels for Fn in bronchial epithelial cells (Figure 6) in a doseresponsive fashion. The constitutive expression of Fn mRNA is demonstrated in the control cultures: an appropriate 8.6-kb mRNA band hybridizes with a 32P-Iabeled human cDNA Fn probe. With exposure to increasing concentrations of TGF-I3" there is increased expression of Fn mRNA. Cultures exposed to the highest concentration of TGF-I3" 250 pM, demonstrated the most intense Fn mRNA band . This suggests that TGF-I3, increases the production of Fn from bronchial epithelial cells at least in part by regulating Fn mRNA levels. TIme course ofTGF-,6 on FnmRNA expression. The effect of TGF-I3, on the Fn mRNA levels is present minimally after 6 h of expo sure of the cultures to TGF-131 (Figure 7). The peak effect is observed at 12 to 24 h of exposure and appears to be declining slightly by 48 h . This timing is consistent with the observed timing of the increase in Fn released
Romberger, Beckmann, Claassen et al. : Fn Modulation of Bovine Bronchial Epithelial Cells by TGF-i3
153
- TGF·/3 + TGF-/3
FN mRNA Fn _ 200kD
116 kD 97kD
-
-
o
-
2.5
25 pM
250
