Extracellular Glutathione Suppresses Human Lung Fibroblast Proliferation Andre M. Cantin, Pierre Larivee, and Raymond O. Begin Unite de Recherche Pulmonaire, Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, Quebec, Canada
Alveolar epithelial lining fluid glutathione (GSH) is markedly decreased in patients with idiopathic pulmonary fibrosis (IPF). Because patients with IPF have exaggerated numbers of fibroblasts in their lower respiratory tract, we hypothesized that GSH can suppress lung fibroblast proliferation. To verify this hypothesis, we examined the ability of GSH to suppress human lung fibroblast (ATCC; HFL-l) proliferation in vitro in the presence of either IPF bronchoalveolar lavage fluid (BAL) or calf serum (CS). Both CS at a concentration of 10% and IPF BAL markedly increased fibroblast proliferation when compared to cells grown without CS or IPF BAL (10% CS = 93 ± 4%, P < 0.001; IPF BAL = 47 ± 4%, P < 0.001). In the presence of physiologic concentrations of GSH (0 to 500 t-tM), both CS- and IPF BAL-mediated fibroblast proliferation were markedly reduced, with 500 t-tM GSH inducing complete inhibition. Interestingly, glutathione disulfide (GSSH) and S-methylglutathione did not suppress proliferation, whereas various sultbydryl-containing molecules (cysteine, N-acetylcysteine, 2-mercaptoethanol, and low concentrations of dithiothreitol) induced an inhibition of fibroblast proliferation similar to that observed with GSH. Most of the suppressive effect of GSH was mediated at the cell level since incubation of fibroblasts with 500 t-tM GSH for 1 h completely blocked the ability of the cells to subsequently proliferate in the presence of untreated 10% CS. Treatment ofCS with 500 t-tM GSH for 1 h followed by removal ofGSH by molecular sieve chromatography had no detectable effect on fibroblast proliferation. Catalase inhibited the GSHmediated suppression of fibroblast proliferation, suggesting that oxidation of the sultbydryl group may produce H20 rrelated oxidants, which could inhibit fibroblast proliferation. We conclude that (1) GSH in the extracellular milieu can suppress lung fibroblast proliferation in response to mitogens and (2) suppression is related to the presence of the sultbydryl group of GSH. It is possible that a lower-respiratorytract GSH deficiency as observed in IPF may create an environment favoring excessive fibroblast proliferation.
Glutathione (t-v-glutamyl-t-cysteinylglycine) (GSH) is a sultbydryl-containing tripeptide present in most mammalian cells. In addition to a variety of metabolic functions, GSH plays an essential role in protecting cells against toxic oxidants and xenobiotics (1). Although GSH is exported from many cell types, the extracellular concentration of this peptide is most often low (1). However, the alveolar epithelial surface of the lung is different in this regard since high GSH levels (i.e., 429 ± 34 t-tM) are found in normal alveolar epitheliallining fluid (ELF) (2). Furthermore, ELF GSH levels are increased in normal smokers, but markedly decreased in patients with the fibrotic lung disease idiopathic pulmonary fibrosis (IPF) (2, 3). Key Words: glutathione, lung fibrosis, fibroblast, idiopathicpulmonary fibrosis fibrosis
(Received in original form December 7, 1989 and in revised form February 13, 1990) Address correspondence to: Andre Cantin, M.D., CHUS, 3001, 12e Ave Nord, Sherbrooke (QC) JlH 5N4 Abbreviations: bronchoalveolar lavage fluid, BAL; calf serum, CS; epithelial lining fluid, ELF; glutathione, OSH; glutathione disulphide, GSSG; idiopathic pulmonary fibrosis, IPF; platelet-derived growth factor, PDGF. Am. J. Respir. Cell Mol. BioI. Vol. 3. pp. 79-85, 1990
Both normal smokers and IPF patients have an alveolitis characterized by increased numbers of neutrophils and macrophages that release large amounts of oxidants within their lower respiratory tracts (4, 5), yet interstitial fibrosis is minimal in the lungs of smokers, and extensive in patients with IPF (6, 7). Whether GSH can modulate the fibrotic process in these alveolitides is unknown. Current concepts of the pathogenesis of IPF suggest that fibroblast proliferation is a major component of the interstitial changes leading to pulmonary fibrosis (8). Fibroblast proliferation is regulated by a variety of molecules derived mainly from serum and inflammatory cells. The fibroblast growth-promoting molecule accounting for most of the mitogenic activity in serum is platelet-derived growth factor (PDGF) (9). This growth factor has also been shown to be released in exaggerated amounts by alveolar macrophages of patients with IPF (10). Interestingly, PDGF is a disulfide dimer of two homologous polypeptides termed A and B (11). When PDGF is exposed to a reducing agent, it irreversibly loses its biologic activity (12). Furthermore, the extracellular portion of the PDGF receptor and other growth-factor receptors have immunoglobulin-like domains with regularly spaced cysteine residues (13, 14). These cysteine residues likely form disulfide bonds that are essential to the structural
80
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 3 i990
conformation of the receptors (14). Reduction of these extracellular disulfide bonds could conceivably alter the conformation of the growth-factor receptors and their ability to bind specific growth factors. Finally, it is known that GSH can undergo autoxidation to produce hydrogen peroxide, and that extracellular GSH autoxidation likely occurs in vivo (1). Because the alveolar surface is rich in both oxygen and GSH, cell growth inhibitory concentations of H2 0 2 and/or H20 2related oxidants could conceivably be produced. In the context of the above, we hypothesized that extracellular GSH at concentrations found in normal alveolar epitheliallining fluid may be capable of suppressing lung fibroblast proliferation.
Effects of GSH, GSSG, and S-methylglutathione on Fibroblast Proliferation To determine the effects of GSH on fibroblast proliferation, cells were incubated with either 10% CS or IPF BAL, as described above, in the presence of 0 to 500 ~M GSH (Sigma Chemical Co., St. Louis, MO) for 1 h, followed by incubation in basal medium for 72 h, at 37° C, 10% CO 2 • The cells were then counted in an electronic particle counter as described above. To determine whether the observed effects were due to the sulfhydryl group of GSH, identical experiments were performed using 0 to 500 ~M of either glutathione disulphide (GSSG) or S-methylglutathione.
Materials and Methods
Evaluation of the Effects of GSH on the Fibroblasts To determine whether GSH could act directly on the fibroblast to prevent it from subsequently proliferating in response to mitogens, the following experiments were performed. Cells were incubated in basal medium or HBSS (i.e., without mitogens) with 0 to 500 ~M GSH for 1 h, at 37° C, 10% CO2 • The cells were washed 3 times with PBS, and the medium was then replaced by DMEM with 10% CS. The cells were incubated for 72 h at 37° C, 10% CO2 , before being counted in an electronic particle counter.
Study Population Six patients (mean age = 61 ± 4 yr; 5 men, 1 woman) with biopsy-proven IPF were enrolled in this study. Four were nonsmokers and two were ex-smokers who had stopped smoking for more than 1 yr. All underwent bronchoalveolar lavage at the time of diagnosis and none were treated at the time of lavage. Bronchoalveolar Lavage Fluid Preparation Bronchoalveolar lavage was performed by placing the tip of the bronchoscope in a wedge position in either the middle lobe or lingula, followed by the sequential infusion and aspiration of 50 ml sterile PBS solution, as previously described (15). The cells were separated from the lavage fluid (BAL) by centrifugation at 500 x g for 5 min. The BAL albumin content was determined by nephelometric immunoassay (16). The BAL was subsequently dialyzed against distilled water for 18 h, lyophilized, and reconstituted in DMEM (GIBCO Laboratories, Grand Island, NY), containing 1 mg/ml transferrin (both from Collaborative Research, Waltham, MA) to yield a human albumin concentration of 500 ~g/ml.
Fibroblast Proliferation Assay Human fetal lung diploid fibroblasts (HFL-l strain, ATCC CCL 153; American Type Culture Collection, Rockville, MD) were maintained in DMEM plus 10% calf serum (CS) (GIBCO) and used before the twentieth population doubling. To determine fibroblast proliferation, 3.5 x 104 cells/emwere plated in DMEM plus 0.4% CS and cultured in this medium for 72 h at 37° C, 10% CO 2 • The medium was then changed to DMEM with 1 mg/ml albumin and om mg/ml transferrin (referred to subsequently as basal medium) for a further 24 h. After this incubation period, the cells were washed 3 times with PBS and incubated in basal medium and either 10% CS or IPF BAL at an albumin concentration of 500 ~g/ml, for 72 h at 3]0 C, 10% CO2 , then harvested by exposure to 0.25 % trypsin (GIBCO), and counted in an electronic particle counter (Model ZM; Coulter Electronics, Inc., Hialeah, FL). The number of fibroblasts were expressed as the percent increase over the number of fibroblasts incubated in basal medium (control). The absolute number of fibroblasts in the control samples (i.e., basal medium) ranged from 0.77 x lOS to 1.15 x 105 cells/em'. A 100% increase represents a twofold increase in cell number.
Evaluation of the Effect of GSH on Serum Fibroblast Growth-promoting Activity The effect of GSH on the mitogenic activity of serum for fibroblasts was evaluated by incubating 10% CS in DMEM with 0 to 500 ~M GSH for 1 hat 23° C. The GSH was then removed from the serum by chromatography of the medium on a PD-lO column (Pharmacia Fine Chemicals, Piscataway, NJ) which was eluted with DMEM. The ability of the GSHtreated serum-containing medium to stimulate fibroblast proliferation was then evaluated as previously described. Effect of Sulfhydryl Molecules Other Than GSH on Fibroblast Proliferation To determine whether sulfhydryl molecules other than GSH can suppress fibroblast proliferation, the following experiments were performed. Fibroblasts were incubated with 0 to 500 ~M of each of the following sulfhydryl molecules: cysteine, N-acetylcysteine, 2-inercaptoethanol, and dithiothreitol (all from Sigma) in basal medium for 1 hat 37° C, 10% CO2 • The cells were then washed twice with basal medium and incubated in basal medium containing 10% CS for 72 h, at 3]0 C, 10% CO 2 , The cells were then counted in an electronic particle counter. Fibroblast Viability Fibroblast viability was estimated by evaluating cell morphology under phase-contrast microscopy, by trypan blue dye exclusion (17), and by assay of chromium release. For chromium release, cells were prepared as described for the fibroblast proliferation assay, then labeled with 5 ~Ci "Cr/well in 24-well culture plates for 12 h, at 37° C, 10% CO2 in basal medium. The cells were washed 3 times with PBS and incubated in basal medium with 0 to 500 ~M GSH for 1 h. The medium was then removed, and the cells were incubated in DMEM with 10% CS for 7 h, at 37° C, 10% CO2 • The amount of slCr released in the supernatant was
Cantin, Larivee, and Begin: Glutathione Suppression of Fibroblast Growth
80
... Q)
.0
E==:l0
60
1::'::
_I:: l/lO tlSU
40
:o~
...o~ 20
.0
;::
0
100
0
200
GSH
300
400
500
(~M)
Figure 1. Glutathione (GSH)-mediated inhibition of fibroblast proliferation in response to molecules in bronchoalveolar lavage fluid (BAL) of patients with idiopathic pulmonary fibrosis. BAL was dialyzed extensively against water, lyophilized, and reconstituted in sufficient DMEM to yield an albumin concentration of 500 JLg/rnl. Human lung fibroblasts (HFL-l) were incubated in the reconstituted BAL in the presence of 0, 100, and 500 JLM GSH for 3 d, at 37° C, 10% CO 2 • After 3 d, the cells were counted in an electronic particle counter. Results are expressed as the percent increase in cell number over control fibroblasts incubated in DMEM basal medium (see text for details).
measured, and the cytotoxicity index (CI) was calculated as follows: CI =
(Sample dpm) - (background dpm) (Triton-X 1% dpm) - (background dpm)
Background dpm was the amount of 51Cr released by cells in DMEM-lO% CS, and Triton-X 1% dpm was the amount of 51Cr released by cells incubated in 1% Triton-X. 150
81
(3H]thymidine Incorporation Assays To determine whether GSH affected HFL-l fibroblast proliferation only at an early time (i.e., nonconfluence) or at all times (i.e., including cells at confluence), pH]thymidine uptake in the presence or absence of GSH was measured in cells at low density and in confluent cells. Low-density cells were prepared as described above for the fibroblast proliferation assay. After 72 h in DMEM plus 0.4 % CS, the media was changed to 500 JLI per well of either basal medium or DMEM plus 10% CS, 0 to 500 JLM GSH, and 2 JLCi/ml pH]thymidine (70 to 80 Ci/mmol; New England Nuclear, Boston, MA). After a 16-h incubation, the cells were harvested by washing twice with 5 % TCA, followed by solubilization of TCA-insolublematerial with 0.25 M NaOH (1 ml), and counting in 5 ml Aquasol (New England Nuclear). Identical assays were performed on HFL-l cells that had reached confluence after 4 d in DMEM, 10% CS, at 37° C, 10% CO 2 , Biochemical Assays Hydrogen peroxide produced during the incubation of GSH with the fibroblasts was measured by adding 500 JLI DMEM basal medium (GIBCO) containing 500 JLM GSH, 0.25 JLM phenol red, and 10 Vlml horseradish peroxidase to each well of HFL-l fibroblasts (68 X 103 ± 2 X 103 cells/well) for 1 h (18). At the end of the incubation period, 10 JLI of 1 N NaOH were added to each well and the absorbance was read in a spectrophotometer (Model DU-7; Beckman Instruments, Palo Alto, CA) at a wavelength of 610 nm. GSH and GSSG were measured after a l-h incubation in HBSS with the fibroblasts by following spectrophotometrically the enzymatic reduction of 5,5'-dithiobis(2-nitrobenzoic acid) (2). Role of HzOz in GSH Fibroblast Suppression To determine whether H20 2 could play a role in the suppression of fibroblast proliferation by GSH, 100 Vlml catalase (Sigma) was added to HFL-l cells in the presence of 0
A
200
B S-methyl glutathione
a;
120
GSSG
a;
.Q
E= ;:,0
I::~ _I::
.Q
E=
90
;:,0 I::~ _I:: I/) 0
1/)0
",u
:o~ 0_
160
60
"' u
:o~ 0_
~
~
0;:
0;:
30
100
200
300
400
glutathione (11M)
500
120 80
GSH
40
100
200
300
400
500
glutathione (11M)
Figure 2. Effects of (A) GSH (closed circles) and glutathione disulphide (GSSG) (open circles) and (B) GSH (closed circles) and S-methylglutathione (open squares) on serum-driven fibroblast proliferation. Human lung fibroblasts were incubated with 10% calf serum (CS) and concentrations varying from 0 to 500 JLM of either GSH, GSSG, or S-methylglutathione for 1 hat 37° C, 10% CO 2 • The cells were subsequently washed 3 times in PBS, incubated in DMEM basal medium for 3 d at 37° C, 10% CO 2 and counted. Each data point represents the mean ± SEM of triplicates, and each experiment was performed at least 4 times. Results are expressed as the percent increase in fibroblast number over control fibroblasts incubated in basal medium alone.
82
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 3 1990
250
...
~
200
.,""
E::::::-
Eg 150 _I::
"
1 ~
Media ,~
........... '§-------Cf------D""
~,
~
i
1110 (\J(,)
:co~ ~ 100 ....c ;;::
Cells
50
GSH Effect on Cells and on Media Mitogens
O'---.L...---.L...---.L...--_.L...-_----'
o
100
200
300
400
34 t-tM) (2) induced complete supression of fibroblast proliferation (fibroblast number = a ± 4 % above control at 500 t-tM GSH). Similar results were obtained in six different experiments. In contrast, GSH in which the sulfhydryl group had been either oxidized (i.e., GSSG, Figure 2A) or methylated (i.e., S-methylglutathione, Figure 2B) had no suppressive effect on serum-driven fibroblast proliferation.
500
GSH (IlM) Figure 3. Effects of GSH on either the human lung fibroblasts (closed circles) or the 10% CS medium (open circles). Cells were
incubatedin basal mediumwithout serum or BALin the presence of 0 to 500 I-lM GSH for 1 h and subsequently incubated in untreated 10% CS for 3 d at 37° C, 10% CO2 • Results are represented by the closed circles. Alternatively, 0 to 500 I-lM GSH was addedto 10% CS for 1 hat 23° C, and removed from the 10% CS by PD-lO chromatography. The GSH-treated 10% CS mediumwas thenaddedto the fibroblasts, whichwere incubated for 3 d at 37° C, 10% CO2 • Results are representedby the open circles. Each data point represents the mean ± SEM of triplicates.
to 500 t-tM GSH for 1 h and the fibroblast proliferation assay was performed as previously described. Statistical Analysis All results are expressed as the mean, and the standard error is used as an index of dispersion. With the exception of the IPF BAL experiments, which were performed once for each patient, each experiment was performed in triplicate and repeated at least 4 times. Comparisons between mean values were performed using the two-tailed Student's t test (19).
To dissociate the effect of GSH on the fibroblast from its effect on mitogens in media, cells were preincubated for 1 h with GSH in the absence of media mitogens, and subsequently incubated in 10% CS for 3 d at 37° C, 10% CO 2 • Preincubation of cells with GSH in basal medium blocked their capacity to subsequently proliferate in the presence of 10% CS (Figure 3). Cells not treated with GSH responded to 10% CS within 3 d by increasing their number to 191 ± 5 % above control, while preincubation of the cells with 500 t-tM GSH decreased their number to 16 ± 3 %. Similarly.preincubation of cells with 500 t-tM GSH in HBSS reduced the proliferative response to 10% CS from 130 ± 8% to 15 ± 5% (P 0.05 [Figure 4]). Effects of GSH on Rapidly Dividing Versus Confluent Fibroblasts pH]thymidine incorporation was measured to determine the effect of GSH on nonconfluent versus confluent fibroblasts. The addition of 10% CS to nonconfluent fibroblasts increased pH]thymidine uptake (10% CS = 12,860 ± 467 dpm/well versus basal medium = 1,705 ± 35 dpm/well; P < 0.001). In the presence of 400 J.tM GSH, the ability of 10% CS to induce PH]thymidine uptake by nonconfluent cells markedly reduced (10% CS and 400 J.tM GSH = 4,769 ± 322 dpm/well; P < 0.01 compared to 10% CS alone). In contrast, GSH, up to concentrations of 1 mM, had no detectable effect on pH]thymidine uptake by confluent fibroblasts (10% CS = 13,594 ± 862 versus 10% CS and 1 mM GSH = 12,935 ± 653 dpm/well; P > 0.20). Role of GSH-derived Oxidants Because GSH can undergo autoxidation to form H202, the role of GSH-derived oxidants in suppressing fibroblast
200 GSH + catalase
....Q)
160
.c
E~
~o
c:= _c
120
00
nsu
.c~
0_
80
~
;;:::
40
100
200
300
400
500
glutathione ().1M) Figure 5. Evaluation of the influence of catalase on the GSHmediated suppression of fibroblast proliferation. Cells were incubated for 1 h, at 37° C, 10% CO2 in basal medium supplemented with 0 to 500 1tM GSH alone (closed circles) or in the presenceof 100 U/ml catalase (open circles). The cells were then washed and incubated in 10% CS for 3 d at 37° C, 10% CO2 • The number of fibroblasts was determined at the end of the incubation period.
83
proliferation was investigated. The amount of H20 2 that was detectable in the culture medium after l-h incubation of the cells with 500 J.tM GSH was 6.5 ± 2.1 J,tM. In addition, 100 U/ml catalase added to the media containing 500 J.tM GSH completely inhibited the suppressive effect of GSH on fibroblast proliferation (Figure 5). GSH Uptake by Fibroblasts Very little, if any, GSH was taken up by the fibroblasts during the l-h incubation with 500 J.tM GSH in HBSS, since at the end of the incubation period 511 ± 14 J.tM GSH was detected in the fibroblast supernatant. When incubated in HBSS, most of the total GSH remained in the reduced state throughout the incubation period (GSH = 492 ± 15 J.tM at 1 h); however, small amounts of GSSG were detectable in the fibroblast supernatant (GSSG = 9.3 ± 0.4 J.tM at 1 h). Evaluation of Fibroblast Viability In all experiments, the morphology of the fibroblasts as evaluated by phase-contrast microscopy- was similar in cells exposed and not exposed to GSH. No cell detachment from the culture dish was observed at concentrations of GSH up to 500 J.tM. More than 95 % of the cells incubated in the presence or absence of GSH excluded the trypan blue dye. The "Cr cytotoxicity index (CI) was not increased in the presence of up to 500 J.tM GSH (CI = 0 ± 2 %, P > 0.20 compared to 0 J.tM GSH).
Discussion An increase in the number of interstitial fibroblasts is thought to be a key feature of pulmonary interstitial fibrosis (20). In addition, alveolar inflammatory cells in patients with IPF release high levels of molecules capable of inducing fibroblast proliferation (10, 21), consistent with the concept that excessive fibroblast proliferation within the lung interstitium plays an important role in the pathogenesis of IPF. Patients with IPF also have a relative deficiency in lower respiratory tract extracellular GSH (3). The present study demonstrates that extracellular GSH, at concentrations similar to those found in normal ELF, inhibits fibroblast proliferation and suggests that the low GSH levels in IPF ELF (i.e., 97 ± 18 J.tM) (3) may facilitate fibroblast proliferation in response to growth factors within the alveolar structures. Although human diploid cells export GSH to the extracellular milieu, little GSH is detectable in plasma because it is rapidly metabolized by the enzyme ')'-glutamyl transpeptidase (')'-GT) in the basolateral membrane of the renal proximal tubule and luminal brush border (22). In contrast, ELF GSH is high in normal nonsmokers and markedly increased in normal smokers (2). Interestingly, smokers, despite the presence of an alveolitis characterized by the release of large amounts of toxic oxidants from lung inflammatory cells, develop little interstitial fibrosis (6). Whether the high GSH level in normal smokers limits fibroblast proliferation in vivo is unknown. The mechanisms by which GSH inhibits fibroblast proliferation seem dependent upon the presence of the sulfhydryl group. GSSG and S-methylglutathione had no effect on fibroblast proliferation. Furthermore, the sulfhydryl-containing molecules N-acetylcysteine, cysteine, 2-mercaptoethanol and
84
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 3 1990
low concentrations of dithiothreitol had an inhibitory effect on fibroblast proliferation similar to that of GSH. Many potential mechanisms exist by which thiol groups may suppress fibroblast proliferation. First, the major fibroblast mitogen in serum in PDGF, a disulfide polypeptide that completely loses its biologic activity upon reduction (9-12). Interestingly, exaggerated amounts of PDGF are synthesized by alveolar macrophages in IPF, likely resulting in increased extracellular levels of this fibroblast growth-factor within IPF alveolar structures (10). However, suppression ofPDGF activity alone cannot account for all of the inhibitory effect of GSH on fibroblast proliferation observed in the present study since treatment of the cells in the absence of mitogens was sufficient to block a subsequent proliferative response to incubation with untreated serum. A second mechanism by which thiols may inhibit fibroblast proliferation would be through direct interaction between GSH and cell surface molecules. Among the cell surface molecules likely susceptible to thiol exposure are members of what has been called the immunoglobulin superfamily, a group of diverse molecules thought to be derived from a common evolutionary pathway (23). These molecules are characterized by extracellular regions rich in regularly spaced cysteine groups likely forming disulfide bonds essential to their structure. Among these molecules are numerous growthfactor receptors, cell adhesion molecules, and surface adhesion molecules such as fibronectin. The structure of the PDGF receptor, as recently proposed by Williams (14), would be typical of this family of molecules since its extracellular portion likely is composed of five immunoglobulin-like domains, four of which have a structure determined by disulfide bonds. Similar extracellular domains with disulfide bonds have been proposed for other growth-factor receptors, including the interleukin 1 receptor (14). It is therefore conceivable that extracellular thiol may alter the extracellular configuration of critical growth-factor receptors to prevent mitogen binding and subsequent fibroblast proliferation. However, the effects of DTT on fibroblast proliferation observed in the present study argue against this possibility. Although low concentrations of DTT suppressed fibroblast proliferation, higher concentrations had no effect on fibroblast proliferation. Because DTT is a most effective reducing agent, it is unlikely that reduction of cell surface molecules can explain the inhibition of fibroblast proliferation by thiolcontaining molecules. The biphasic effect of DTT is similar to observations by Trotta and colleagues (24) in which low DTT concentrations were found to suppress glutaminase activity of carbamyl phosphate synthetase whereas high DTT concentrations did not affect enzyme activity. The likely explanation for this paradoxical effect was felt to be that high, but not low, concentrations of DTT reacted with H 202 molecules formed from the autoxidation of thiol groups before the H 202 could inactivate the enzyme. Consistent with the concept that H 202 or H202-derived oxidants may mediate suppression of fibroblast proliferation by GSH is the observation in the present study that catalase completely inhibited the effects of GSH on fibroblast proliferation. Interestingly, although autoxidation likely mediated fibroblast growth inhibition, most of the GSH (96 ± 1%) remained in the reduced state as it is in normal alveolar epithe-
liallining fluid (2). These results suggest that GSH at the alveolar surface may simultaneously play a role in fibroblast growth regulation as observed in this study and in providing antioxidant protection of alveolar cells as previously described (2). GSH may also act indirectly on the fibroblast to suppress its proliferation by inducing the release of growth-inhibitory molecules from the fibroblast itself. An example of this would be the enhanced synthesis of prostaglandin E 2 (PG~), a known suppressant of lung fibroblast proliferation (25). GSH is a cofactor in the synthesis of PGE 2 (26). Although this mechanism cannot be excluded, it would seem unlikely, since intact GSH is not usually taken up by cells (27) and, in the present study, all added GSH was recovered in the extracellular milieu after the incubation period. Finally, a nonspecific cytotoxic effect of GSH must be considered. The concentrations ofGSH used in these studies are well within the physiologic range found in normal ELF (2). At these concentrations, we found no morphologic evidence of cytotoxicity and no increase in 5\Cr release, and all cells excluded trypan blue dye. Furthermore, the cells were still capable of proliferating in the presence of GSH, albeit at a reduced rate. However, we cannot exclude the possibility that GSH induced a sublethal nonspecific toxic effect. Although the mechanisms of GSH-mediated inhibition of fibroblast proliferation are not entirely clear, several lines of evidence suggest that GSH deficiency may be associated with enhanced fibrosis. First, IPF patients have marked interstitial fibrosis and are deficient in alveolar epithelial lining fluid GSH (3). Second, an excessive oxidant burden rapidly depletes cellular stores ofGSH (28-30), and several investigators have reported experimental and clinical data demonstrating that an excessive lower respiratory tract oxidant burden leads to interstitial pulmonary fibrosis (31, 32). Third, histologic studies of liver tissue have demonstrated that areas of liver fibrosis are associated with increased '}'-GT (33), an enzyme that can break down GSH (22). Finally, depletion of lung GSH by buthionine sulfoximine treatment after ozone exposure markedly increased pulmonary interstitial fibrosis in mice (34). Because GSH depletion was induced after ozone exposure, it is possible that the mechanisms of enhanced fibrosis were not only related to the antioxidant properties of GSH but also to its suppressive effects on fibroblast proliferation during the tissue repair phase. In summary, GSH, at concentrations measured in the extracellular fluid of the normal lower respiratory tract, can suppress fibroblast proliferation. The sulfhydryl group of GSH is essential to inhibit fibroblast proliferation, and most of the observed suppressive effect seems to result from the interaction between the rapidly dividing fibroblast and lowgrade GSH autoxidation. These results suggest that the decreased extracellular GSH levels in the lower respiratory tract of patients with IPF may create an environment that facilitates lung fibroblast proliferation in response to mitogens within the local milieu. Acknowledgments: The writers thank Ginette Bilodeau and Denis Bisson for technical support and Marguerite Cloutier for help in preparing this manuscript. This work was supported by a grant from the Medical Research Council (Canada) and Association Pulmonaire du Quebec. Dr. Cantin is a scholar of the Fonds de la Recherche en Sante du Quebec.
Cantin, Larivee, and Begin: Glutathione Suppression of Fibroblast Growth
References 1. Meister, A., and M. E. Anderson. 1983. Glutathione. Annu. Rev. Biochem. 52:711-760. 2. Cantin, A. M., S. L. North, R. C. Hubbard, and R. G. Crystal. 1987. Normal Alveolar epithelial lining fluid contains high levels of glutathione. J. Appl. Physiol. 63:152-157. 3. Cantin, A. M., R. C. Hubbard, and R. G. Crystal. 1989. Glutathione deficiency in the epithelial lining fluid of the lower respiratory tract in idiopathic pulmonary fibrosis. Am. Rev. Respir. Dis. 139:370-372. 4. Hoidal,1. R., R. B. Fox, P. A. Lemarbre, R. Perri, and 1. E. Repine. 1981. Altered oxidative metabolism responses in vitro of alveolar macrophages from asymptomatic smokers. Am. Rev. Respir. Dis. 123:85-89. 5. Cantin, A. M., S. L. North, G. A. Fells, R. C. Hubbard, and R. G. Crystal. 1989. Oxidant-mediated epithelial cell injury in idiopathic pulmonary fibrosis. J. Clin. Invest. 79:1665-1673. 6. Cosio, M. G., K. A. Hale, and D. E. Niewoehner. 1980. Morphologic and morphometric effects of prolonged cigarette smoking on the small airways. Am. Rev. Respir: Dis. 122:265-271. 7. Crystal, R. G., 1. D. Fulmer, W. C. Roberts, M. L. Moss, B. R. Line, and H. Y. Reynolds. 1976. Idiopathic pulmonary fibrosis: clinical, histologic, radiologic, physiologic, scintigraphic, cytological and biochemical aspects. Ann. Intern. Med. 85:769-788. 8. Crystal, R. G., P. B. Bitterman, S. I. Rennard, A. 1. Hance, and B. A. Keogh. 1984. Interstitial lung diseases of unknown cause: disorders characterized by chronic inflammation of the lower respiratory tract. N. Engl. J. Med. 310:154-166. 9. Deuel, T. E, 1. S. Huang, R. T. Proffitt, 1. V. Baenziger, D. Chang, and B. B. Kennedy. 1981. Human platelet-derived growth factor: purification and resolution into two active protein fractions. J. Bioi. Chem. 256:88968899. 10. Martinet, Y., W. N. Rom, G. R. Grotendorst, G. R. Martin, andR. G. Crystal. 1987. Exaggerated spontaneous release of platelet-derived growth factor by alveolar macrophages from patients with idiopathic pulmonary fibrosis. N. Engl. J. Med. 317:202-209. 11. Betsholtz, C., A. Johnsson, C. H. Heldin et al. 1986. cDNA sequence and chromosomal localization of human platelet-derived growth factor Achain and its expression in tumor cell lines. Nature 320:695-699. 12. Ross, R. and A. Vogel. 1978. The platelet-derived growth factor. Cell 14:203-210. 13. Yarden, Y., 1. A. Escobedo, W. 1. Kuang etal. 1986. Structure of the receptor for platelet-derived growth factor helps define a family of closely related growth factor receptors. Nature 323:226-232. 14. Williams, L. T. 1989. Signal transduction by the platelet-derived growth factor receptor. Science 243:1564-1570. 15. Cantin, A., R. Begin, M. Rola-Pleszczynski, and R. Boileau. Heterogeneity of bronchoalveolar lavage cellularity in stage 11I pulmonary sarcoidosis. Chest 83:485-486. 16. Killingsworth, L. M., and J. Savory. 1972. Manual nephelometric methods for imrnunochemical determination of immunoglobulins IgG, IgA, and IgM in human serum. Clin. Chem. 18:335-339. 17. Freshney, R. I. Measurement of cytotoxicity and viability. 1987. In Culture of Animal Cells: A Manual of Basic Technique. R. I. Freshney, editor.
85
Alan R. Liss, New York: 245-256. 18. Pick, R., and Y. Keisari. 1980. A simple colorimetric method for the measurement of hydrogen peroxide produced by cells in culture. J. Immunol. Methods 38:161-170. 19. Snedecor, G. w., and W. G. Cochrane. 1980. Statistical Methods, 7th ed. Iowa State University Press. Ames, IA. 20. Brody, A. R., and 1. E. Craighead. 1976. Interstitial association of cells lining air spaces in human pulmonary fibrosis. Virchows Arch. [A] 372: 39-49. 21. Bitterman, P. B., S. Adelberg, and R. G. Crystal. 1983. Mechanismsofpulmonary fibrosis: spontaneous release of the alveolar macrophage-derived growth factor in the interstitial lung disorders. J. Clin. Invest. 72:18011813. 22. Abbott, W. A., R. 1. Bridges, and A. Meister. 1984. Extracellular metabolism of glutathione accounts for its disappearance from the basolateral circulation of the kidney. 1. Bioi. Chem. 259:15393-15400. 23. Edelman, G. M. 1987. CAMs and IGs: cell adhesion and the evolutionary origins of immunity. Immunol. Rev. 100:11-45. 24. Trotta, P. P., L. M. Pinkus, and A. Meister. 1974. Inhibition by dithiothreitol of the utilization of glutamine by carbamyl phosphate synthetase. J. Bioi. Chem. 249:1915-1921. 25. Bitterman, P. B., M. D. Wewers, S. I. Rennard, S. Adelberg, and R. G. Crystal. 1986. Modulation of alveolar macrophage-driven fibroblast proliferation by alternative macrophage mediators. J. Clin. Invest. 77:700708. 26. Metz, S. A. 1981. Anti-inflammatory agents as inhibitors of prostaglandin synthesis in man. Med. Clin. North Am. 65:7I'J-757. 27. Meister, A. 1985. Methods for the selective modification of glutathione metabolism and study of glutathione transport. Methods Enzymot. 113: 571-585. 28. Schraufstiitter, I. V., D. B. Hinshaw, P. A. Hyslop, R. G. Spragg, and C. G. Cochrane. 1985. Glutathione cycle activity and pyridine nucleotide levels in oxidant induced injury of cells. J. Clin. Invest. 76:1131-1139. 29. Dunbar, 1. R., A. 1. Delucia, and L. R. Bryant. 1984. Glutathione status of isolated rabbit lungs: effects of nitrofurantoin and paraquat perfusion with normoxic and hyperoxic ventilation. Biochem. Pharmacol. 33:13431348. 30. Delucia, A. J., M. G. Mustafa, M. Z. Hussain, and C. E. Cross. 1975. Ozone interaction with rodent lung. III. Oxidation of reduced glutathione and formation of mixed disulfides between protein and nonprotein sulfhydryls. J. Clin. Invest. 55:794-802. 31. Johnson, K. 1., and P. A. Ward. 1982. Acute and progressive lung injury after contact with phorbol myristate acetate. Am. J. Pathol. 107:29-35. 32. Pratt, P. C. 1974. Pathology of pulmonary oxygen toxicity. Am. Rev. Respir: Dis. 110(Suppl):51-57. 33. Yamauchi, M., K. Kimura, H. Kawase et at. 1984. Hepatic gammaglutamyl transferase and hepatic fibrosis in patients with alcoholic liver disease. Enzyme 32:110-115. 34. Sun, J. D., J. A. Pickrell, J. R. Harkema, S. J. McLaughlin, E E Hahn, and R. E Henderson. 1988. Effects ofbuthionine sulfoximine on the development of ozone-induced pulmonary fibrosis. Exp. Mol. Pathol. 49: 254-266.
Alveolar epithelial lining fluid glutathione (GSH) is markedly decreased in patients with idiopathic pulmonary fibrosis (IPF). Because patients with IPF have exaggerated numbers of fibroblasts in their lower respiratory tract, we hypothesized that GSH can suppress lung fibroblast proliferation. To verify this hypothesis, we examined the ability of GSH to suppress human lung fibroblast (ATCC; HFL-l) proliferation in vitro in the presence of either IPF bronchoalveolar lavage fluid (BAL) or calf serum (CS). Both CS at a concentration of 10% and IPF BAL markedly increased fibroblast proliferation when compared to cells grown without CS or IPF BAL (10% CS = 93 ± 4%, P < 0.001; IPF BAL = 47 ± 4%, P < 0.001). In the presence of physiologic concentrations of GSH (0 to 500 t-tM), both CS- and IPF BAL-mediated fibroblast proliferation were markedly reduced, with 500 t-tM GSH inducing complete inhibition. Interestingly, glutathione disulfide (GSSH) and S-methylglutathione did not suppress proliferation, whereas various sultbydryl-containing molecules (cysteine, N-acetylcysteine, 2-mercaptoethanol, and low concentrations of dithiothreitol) induced an inhibition of fibroblast proliferation similar to that observed with GSH. Most of the suppressive effect of GSH was mediated at the cell level since incubation of fibroblasts with 500 t-tM GSH for 1 h completely blocked the ability of the cells to subsequently proliferate in the presence of untreated 10% CS. Treatment ofCS with 500 t-tM GSH for 1 h followed by removal ofGSH by molecular sieve chromatography had no detectable effect on fibroblast proliferation. Catalase inhibited the GSHmediated suppression of fibroblast proliferation, suggesting that oxidation of the sultbydryl group may produce H20 rrelated oxidants, which could inhibit fibroblast proliferation. We conclude that (1) GSH in the extracellular milieu can suppress lung fibroblast proliferation in response to mitogens and (2) suppression is related to the presence of the sultbydryl group of GSH. It is possible that a lower-respiratorytract GSH deficiency as observed in IPF may create an environment favoring excessive fibroblast proliferation.
Glutathione (t-v-glutamyl-t-cysteinylglycine) (GSH) is a sultbydryl-containing tripeptide present in most mammalian cells. In addition to a variety of metabolic functions, GSH plays an essential role in protecting cells against toxic oxidants and xenobiotics (1). Although GSH is exported from many cell types, the extracellular concentration of this peptide is most often low (1). However, the alveolar epithelial surface of the lung is different in this regard since high GSH levels (i.e., 429 ± 34 t-tM) are found in normal alveolar epitheliallining fluid (ELF) (2). Furthermore, ELF GSH levels are increased in normal smokers, but markedly decreased in patients with the fibrotic lung disease idiopathic pulmonary fibrosis (IPF) (2, 3). Key Words: glutathione, lung fibrosis, fibroblast, idiopathicpulmonary fibrosis fibrosis
(Received in original form December 7, 1989 and in revised form February 13, 1990) Address correspondence to: Andre Cantin, M.D., CHUS, 3001, 12e Ave Nord, Sherbrooke (QC) JlH 5N4 Abbreviations: bronchoalveolar lavage fluid, BAL; calf serum, CS; epithelial lining fluid, ELF; glutathione, OSH; glutathione disulphide, GSSG; idiopathic pulmonary fibrosis, IPF; platelet-derived growth factor, PDGF. Am. J. Respir. Cell Mol. BioI. Vol. 3. pp. 79-85, 1990
Both normal smokers and IPF patients have an alveolitis characterized by increased numbers of neutrophils and macrophages that release large amounts of oxidants within their lower respiratory tracts (4, 5), yet interstitial fibrosis is minimal in the lungs of smokers, and extensive in patients with IPF (6, 7). Whether GSH can modulate the fibrotic process in these alveolitides is unknown. Current concepts of the pathogenesis of IPF suggest that fibroblast proliferation is a major component of the interstitial changes leading to pulmonary fibrosis (8). Fibroblast proliferation is regulated by a variety of molecules derived mainly from serum and inflammatory cells. The fibroblast growth-promoting molecule accounting for most of the mitogenic activity in serum is platelet-derived growth factor (PDGF) (9). This growth factor has also been shown to be released in exaggerated amounts by alveolar macrophages of patients with IPF (10). Interestingly, PDGF is a disulfide dimer of two homologous polypeptides termed A and B (11). When PDGF is exposed to a reducing agent, it irreversibly loses its biologic activity (12). Furthermore, the extracellular portion of the PDGF receptor and other growth-factor receptors have immunoglobulin-like domains with regularly spaced cysteine residues (13, 14). These cysteine residues likely form disulfide bonds that are essential to the structural
80
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 3 i990
conformation of the receptors (14). Reduction of these extracellular disulfide bonds could conceivably alter the conformation of the growth-factor receptors and their ability to bind specific growth factors. Finally, it is known that GSH can undergo autoxidation to produce hydrogen peroxide, and that extracellular GSH autoxidation likely occurs in vivo (1). Because the alveolar surface is rich in both oxygen and GSH, cell growth inhibitory concentations of H2 0 2 and/or H20 2related oxidants could conceivably be produced. In the context of the above, we hypothesized that extracellular GSH at concentrations found in normal alveolar epitheliallining fluid may be capable of suppressing lung fibroblast proliferation.
Effects of GSH, GSSG, and S-methylglutathione on Fibroblast Proliferation To determine the effects of GSH on fibroblast proliferation, cells were incubated with either 10% CS or IPF BAL, as described above, in the presence of 0 to 500 ~M GSH (Sigma Chemical Co., St. Louis, MO) for 1 h, followed by incubation in basal medium for 72 h, at 37° C, 10% CO 2 • The cells were then counted in an electronic particle counter as described above. To determine whether the observed effects were due to the sulfhydryl group of GSH, identical experiments were performed using 0 to 500 ~M of either glutathione disulphide (GSSG) or S-methylglutathione.
Materials and Methods
Evaluation of the Effects of GSH on the Fibroblasts To determine whether GSH could act directly on the fibroblast to prevent it from subsequently proliferating in response to mitogens, the following experiments were performed. Cells were incubated in basal medium or HBSS (i.e., without mitogens) with 0 to 500 ~M GSH for 1 h, at 37° C, 10% CO2 • The cells were washed 3 times with PBS, and the medium was then replaced by DMEM with 10% CS. The cells were incubated for 72 h at 37° C, 10% CO2 , before being counted in an electronic particle counter.
Study Population Six patients (mean age = 61 ± 4 yr; 5 men, 1 woman) with biopsy-proven IPF were enrolled in this study. Four were nonsmokers and two were ex-smokers who had stopped smoking for more than 1 yr. All underwent bronchoalveolar lavage at the time of diagnosis and none were treated at the time of lavage. Bronchoalveolar Lavage Fluid Preparation Bronchoalveolar lavage was performed by placing the tip of the bronchoscope in a wedge position in either the middle lobe or lingula, followed by the sequential infusion and aspiration of 50 ml sterile PBS solution, as previously described (15). The cells were separated from the lavage fluid (BAL) by centrifugation at 500 x g for 5 min. The BAL albumin content was determined by nephelometric immunoassay (16). The BAL was subsequently dialyzed against distilled water for 18 h, lyophilized, and reconstituted in DMEM (GIBCO Laboratories, Grand Island, NY), containing 1 mg/ml transferrin (both from Collaborative Research, Waltham, MA) to yield a human albumin concentration of 500 ~g/ml.
Fibroblast Proliferation Assay Human fetal lung diploid fibroblasts (HFL-l strain, ATCC CCL 153; American Type Culture Collection, Rockville, MD) were maintained in DMEM plus 10% calf serum (CS) (GIBCO) and used before the twentieth population doubling. To determine fibroblast proliferation, 3.5 x 104 cells/emwere plated in DMEM plus 0.4% CS and cultured in this medium for 72 h at 37° C, 10% CO 2 • The medium was then changed to DMEM with 1 mg/ml albumin and om mg/ml transferrin (referred to subsequently as basal medium) for a further 24 h. After this incubation period, the cells were washed 3 times with PBS and incubated in basal medium and either 10% CS or IPF BAL at an albumin concentration of 500 ~g/ml, for 72 h at 3]0 C, 10% CO2 , then harvested by exposure to 0.25 % trypsin (GIBCO), and counted in an electronic particle counter (Model ZM; Coulter Electronics, Inc., Hialeah, FL). The number of fibroblasts were expressed as the percent increase over the number of fibroblasts incubated in basal medium (control). The absolute number of fibroblasts in the control samples (i.e., basal medium) ranged from 0.77 x lOS to 1.15 x 105 cells/em'. A 100% increase represents a twofold increase in cell number.
Evaluation of the Effect of GSH on Serum Fibroblast Growth-promoting Activity The effect of GSH on the mitogenic activity of serum for fibroblasts was evaluated by incubating 10% CS in DMEM with 0 to 500 ~M GSH for 1 hat 23° C. The GSH was then removed from the serum by chromatography of the medium on a PD-lO column (Pharmacia Fine Chemicals, Piscataway, NJ) which was eluted with DMEM. The ability of the GSHtreated serum-containing medium to stimulate fibroblast proliferation was then evaluated as previously described. Effect of Sulfhydryl Molecules Other Than GSH on Fibroblast Proliferation To determine whether sulfhydryl molecules other than GSH can suppress fibroblast proliferation, the following experiments were performed. Fibroblasts were incubated with 0 to 500 ~M of each of the following sulfhydryl molecules: cysteine, N-acetylcysteine, 2-inercaptoethanol, and dithiothreitol (all from Sigma) in basal medium for 1 hat 37° C, 10% CO2 • The cells were then washed twice with basal medium and incubated in basal medium containing 10% CS for 72 h, at 3]0 C, 10% CO 2 , The cells were then counted in an electronic particle counter. Fibroblast Viability Fibroblast viability was estimated by evaluating cell morphology under phase-contrast microscopy, by trypan blue dye exclusion (17), and by assay of chromium release. For chromium release, cells were prepared as described for the fibroblast proliferation assay, then labeled with 5 ~Ci "Cr/well in 24-well culture plates for 12 h, at 37° C, 10% CO2 in basal medium. The cells were washed 3 times with PBS and incubated in basal medium with 0 to 500 ~M GSH for 1 h. The medium was then removed, and the cells were incubated in DMEM with 10% CS for 7 h, at 37° C, 10% CO2 • The amount of slCr released in the supernatant was
Cantin, Larivee, and Begin: Glutathione Suppression of Fibroblast Growth
80
... Q)
.0
E==:l0
60
1::'::
_I:: l/lO tlSU
40
:o~
...o~ 20
.0
;::
0
100
0
200
GSH
300
400
500
(~M)
Figure 1. Glutathione (GSH)-mediated inhibition of fibroblast proliferation in response to molecules in bronchoalveolar lavage fluid (BAL) of patients with idiopathic pulmonary fibrosis. BAL was dialyzed extensively against water, lyophilized, and reconstituted in sufficient DMEM to yield an albumin concentration of 500 JLg/rnl. Human lung fibroblasts (HFL-l) were incubated in the reconstituted BAL in the presence of 0, 100, and 500 JLM GSH for 3 d, at 37° C, 10% CO 2 • After 3 d, the cells were counted in an electronic particle counter. Results are expressed as the percent increase in cell number over control fibroblasts incubated in DMEM basal medium (see text for details).
measured, and the cytotoxicity index (CI) was calculated as follows: CI =
(Sample dpm) - (background dpm) (Triton-X 1% dpm) - (background dpm)
Background dpm was the amount of 51Cr released by cells in DMEM-lO% CS, and Triton-X 1% dpm was the amount of 51Cr released by cells incubated in 1% Triton-X. 150
81
(3H]thymidine Incorporation Assays To determine whether GSH affected HFL-l fibroblast proliferation only at an early time (i.e., nonconfluence) or at all times (i.e., including cells at confluence), pH]thymidine uptake in the presence or absence of GSH was measured in cells at low density and in confluent cells. Low-density cells were prepared as described above for the fibroblast proliferation assay. After 72 h in DMEM plus 0.4 % CS, the media was changed to 500 JLI per well of either basal medium or DMEM plus 10% CS, 0 to 500 JLM GSH, and 2 JLCi/ml pH]thymidine (70 to 80 Ci/mmol; New England Nuclear, Boston, MA). After a 16-h incubation, the cells were harvested by washing twice with 5 % TCA, followed by solubilization of TCA-insolublematerial with 0.25 M NaOH (1 ml), and counting in 5 ml Aquasol (New England Nuclear). Identical assays were performed on HFL-l cells that had reached confluence after 4 d in DMEM, 10% CS, at 37° C, 10% CO 2 , Biochemical Assays Hydrogen peroxide produced during the incubation of GSH with the fibroblasts was measured by adding 500 JLI DMEM basal medium (GIBCO) containing 500 JLM GSH, 0.25 JLM phenol red, and 10 Vlml horseradish peroxidase to each well of HFL-l fibroblasts (68 X 103 ± 2 X 103 cells/well) for 1 h (18). At the end of the incubation period, 10 JLI of 1 N NaOH were added to each well and the absorbance was read in a spectrophotometer (Model DU-7; Beckman Instruments, Palo Alto, CA) at a wavelength of 610 nm. GSH and GSSG were measured after a l-h incubation in HBSS with the fibroblasts by following spectrophotometrically the enzymatic reduction of 5,5'-dithiobis(2-nitrobenzoic acid) (2). Role of HzOz in GSH Fibroblast Suppression To determine whether H20 2 could play a role in the suppression of fibroblast proliferation by GSH, 100 Vlml catalase (Sigma) was added to HFL-l cells in the presence of 0
A
200
B S-methyl glutathione
a;
120
GSSG
a;
.Q
E= ;:,0
I::~ _I::
.Q
E=
90
;:,0 I::~ _I:: I/) 0
1/)0
",u
:o~ 0_
160
60
"' u
:o~ 0_
~
~
0;:
0;:
30
100
200
300
400
glutathione (11M)
500
120 80
GSH
40
100
200
300
400
500
glutathione (11M)
Figure 2. Effects of (A) GSH (closed circles) and glutathione disulphide (GSSG) (open circles) and (B) GSH (closed circles) and S-methylglutathione (open squares) on serum-driven fibroblast proliferation. Human lung fibroblasts were incubated with 10% calf serum (CS) and concentrations varying from 0 to 500 JLM of either GSH, GSSG, or S-methylglutathione for 1 hat 37° C, 10% CO 2 • The cells were subsequently washed 3 times in PBS, incubated in DMEM basal medium for 3 d at 37° C, 10% CO 2 and counted. Each data point represents the mean ± SEM of triplicates, and each experiment was performed at least 4 times. Results are expressed as the percent increase in fibroblast number over control fibroblasts incubated in basal medium alone.
82
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 3 1990
250
...
~
200
.,""
E::::::-
Eg 150 _I::
"
1 ~
Media ,~
........... '§-------Cf------D""
~,
~
i
1110 (\J(,)
:co~ ~ 100 ....c ;;::
Cells
50
GSH Effect on Cells and on Media Mitogens
O'---.L...---.L...---.L...--_.L...-_----'
o
100
200
300
400
34 t-tM) (2) induced complete supression of fibroblast proliferation (fibroblast number = a ± 4 % above control at 500 t-tM GSH). Similar results were obtained in six different experiments. In contrast, GSH in which the sulfhydryl group had been either oxidized (i.e., GSSG, Figure 2A) or methylated (i.e., S-methylglutathione, Figure 2B) had no suppressive effect on serum-driven fibroblast proliferation.
500
GSH (IlM) Figure 3. Effects of GSH on either the human lung fibroblasts (closed circles) or the 10% CS medium (open circles). Cells were
incubatedin basal mediumwithout serum or BALin the presence of 0 to 500 I-lM GSH for 1 h and subsequently incubated in untreated 10% CS for 3 d at 37° C, 10% CO2 • Results are represented by the closed circles. Alternatively, 0 to 500 I-lM GSH was addedto 10% CS for 1 hat 23° C, and removed from the 10% CS by PD-lO chromatography. The GSH-treated 10% CS mediumwas thenaddedto the fibroblasts, whichwere incubated for 3 d at 37° C, 10% CO2 • Results are representedby the open circles. Each data point represents the mean ± SEM of triplicates.
to 500 t-tM GSH for 1 h and the fibroblast proliferation assay was performed as previously described. Statistical Analysis All results are expressed as the mean, and the standard error is used as an index of dispersion. With the exception of the IPF BAL experiments, which were performed once for each patient, each experiment was performed in triplicate and repeated at least 4 times. Comparisons between mean values were performed using the two-tailed Student's t test (19).
To dissociate the effect of GSH on the fibroblast from its effect on mitogens in media, cells were preincubated for 1 h with GSH in the absence of media mitogens, and subsequently incubated in 10% CS for 3 d at 37° C, 10% CO 2 • Preincubation of cells with GSH in basal medium blocked their capacity to subsequently proliferate in the presence of 10% CS (Figure 3). Cells not treated with GSH responded to 10% CS within 3 d by increasing their number to 191 ± 5 % above control, while preincubation of the cells with 500 t-tM GSH decreased their number to 16 ± 3 %. Similarly.preincubation of cells with 500 t-tM GSH in HBSS reduced the proliferative response to 10% CS from 130 ± 8% to 15 ± 5% (P 0.05 [Figure 4]). Effects of GSH on Rapidly Dividing Versus Confluent Fibroblasts pH]thymidine incorporation was measured to determine the effect of GSH on nonconfluent versus confluent fibroblasts. The addition of 10% CS to nonconfluent fibroblasts increased pH]thymidine uptake (10% CS = 12,860 ± 467 dpm/well versus basal medium = 1,705 ± 35 dpm/well; P < 0.001). In the presence of 400 J.tM GSH, the ability of 10% CS to induce PH]thymidine uptake by nonconfluent cells markedly reduced (10% CS and 400 J.tM GSH = 4,769 ± 322 dpm/well; P < 0.01 compared to 10% CS alone). In contrast, GSH, up to concentrations of 1 mM, had no detectable effect on pH]thymidine uptake by confluent fibroblasts (10% CS = 13,594 ± 862 versus 10% CS and 1 mM GSH = 12,935 ± 653 dpm/well; P > 0.20). Role of GSH-derived Oxidants Because GSH can undergo autoxidation to form H202, the role of GSH-derived oxidants in suppressing fibroblast
200 GSH + catalase
....Q)
160
.c
E~
~o
c:= _c
120
00
nsu
.c~
0_
80
~
;;:::
40
100
200
300
400
500
glutathione ().1M) Figure 5. Evaluation of the influence of catalase on the GSHmediated suppression of fibroblast proliferation. Cells were incubated for 1 h, at 37° C, 10% CO2 in basal medium supplemented with 0 to 500 1tM GSH alone (closed circles) or in the presenceof 100 U/ml catalase (open circles). The cells were then washed and incubated in 10% CS for 3 d at 37° C, 10% CO2 • The number of fibroblasts was determined at the end of the incubation period.
83
proliferation was investigated. The amount of H20 2 that was detectable in the culture medium after l-h incubation of the cells with 500 J.tM GSH was 6.5 ± 2.1 J,tM. In addition, 100 U/ml catalase added to the media containing 500 J.tM GSH completely inhibited the suppressive effect of GSH on fibroblast proliferation (Figure 5). GSH Uptake by Fibroblasts Very little, if any, GSH was taken up by the fibroblasts during the l-h incubation with 500 J.tM GSH in HBSS, since at the end of the incubation period 511 ± 14 J.tM GSH was detected in the fibroblast supernatant. When incubated in HBSS, most of the total GSH remained in the reduced state throughout the incubation period (GSH = 492 ± 15 J.tM at 1 h); however, small amounts of GSSG were detectable in the fibroblast supernatant (GSSG = 9.3 ± 0.4 J.tM at 1 h). Evaluation of Fibroblast Viability In all experiments, the morphology of the fibroblasts as evaluated by phase-contrast microscopy- was similar in cells exposed and not exposed to GSH. No cell detachment from the culture dish was observed at concentrations of GSH up to 500 J.tM. More than 95 % of the cells incubated in the presence or absence of GSH excluded the trypan blue dye. The "Cr cytotoxicity index (CI) was not increased in the presence of up to 500 J.tM GSH (CI = 0 ± 2 %, P > 0.20 compared to 0 J.tM GSH).
Discussion An increase in the number of interstitial fibroblasts is thought to be a key feature of pulmonary interstitial fibrosis (20). In addition, alveolar inflammatory cells in patients with IPF release high levels of molecules capable of inducing fibroblast proliferation (10, 21), consistent with the concept that excessive fibroblast proliferation within the lung interstitium plays an important role in the pathogenesis of IPF. Patients with IPF also have a relative deficiency in lower respiratory tract extracellular GSH (3). The present study demonstrates that extracellular GSH, at concentrations similar to those found in normal ELF, inhibits fibroblast proliferation and suggests that the low GSH levels in IPF ELF (i.e., 97 ± 18 J.tM) (3) may facilitate fibroblast proliferation in response to growth factors within the alveolar structures. Although human diploid cells export GSH to the extracellular milieu, little GSH is detectable in plasma because it is rapidly metabolized by the enzyme ')'-glutamyl transpeptidase (')'-GT) in the basolateral membrane of the renal proximal tubule and luminal brush border (22). In contrast, ELF GSH is high in normal nonsmokers and markedly increased in normal smokers (2). Interestingly, smokers, despite the presence of an alveolitis characterized by the release of large amounts of toxic oxidants from lung inflammatory cells, develop little interstitial fibrosis (6). Whether the high GSH level in normal smokers limits fibroblast proliferation in vivo is unknown. The mechanisms by which GSH inhibits fibroblast proliferation seem dependent upon the presence of the sulfhydryl group. GSSG and S-methylglutathione had no effect on fibroblast proliferation. Furthermore, the sulfhydryl-containing molecules N-acetylcysteine, cysteine, 2-mercaptoethanol and
84
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 3 1990
low concentrations of dithiothreitol had an inhibitory effect on fibroblast proliferation similar to that of GSH. Many potential mechanisms exist by which thiol groups may suppress fibroblast proliferation. First, the major fibroblast mitogen in serum in PDGF, a disulfide polypeptide that completely loses its biologic activity upon reduction (9-12). Interestingly, exaggerated amounts of PDGF are synthesized by alveolar macrophages in IPF, likely resulting in increased extracellular levels of this fibroblast growth-factor within IPF alveolar structures (10). However, suppression ofPDGF activity alone cannot account for all of the inhibitory effect of GSH on fibroblast proliferation observed in the present study since treatment of the cells in the absence of mitogens was sufficient to block a subsequent proliferative response to incubation with untreated serum. A second mechanism by which thiols may inhibit fibroblast proliferation would be through direct interaction between GSH and cell surface molecules. Among the cell surface molecules likely susceptible to thiol exposure are members of what has been called the immunoglobulin superfamily, a group of diverse molecules thought to be derived from a common evolutionary pathway (23). These molecules are characterized by extracellular regions rich in regularly spaced cysteine groups likely forming disulfide bonds essential to their structure. Among these molecules are numerous growthfactor receptors, cell adhesion molecules, and surface adhesion molecules such as fibronectin. The structure of the PDGF receptor, as recently proposed by Williams (14), would be typical of this family of molecules since its extracellular portion likely is composed of five immunoglobulin-like domains, four of which have a structure determined by disulfide bonds. Similar extracellular domains with disulfide bonds have been proposed for other growth-factor receptors, including the interleukin 1 receptor (14). It is therefore conceivable that extracellular thiol may alter the extracellular configuration of critical growth-factor receptors to prevent mitogen binding and subsequent fibroblast proliferation. However, the effects of DTT on fibroblast proliferation observed in the present study argue against this possibility. Although low concentrations of DTT suppressed fibroblast proliferation, higher concentrations had no effect on fibroblast proliferation. Because DTT is a most effective reducing agent, it is unlikely that reduction of cell surface molecules can explain the inhibition of fibroblast proliferation by thiolcontaining molecules. The biphasic effect of DTT is similar to observations by Trotta and colleagues (24) in which low DTT concentrations were found to suppress glutaminase activity of carbamyl phosphate synthetase whereas high DTT concentrations did not affect enzyme activity. The likely explanation for this paradoxical effect was felt to be that high, but not low, concentrations of DTT reacted with H 202 molecules formed from the autoxidation of thiol groups before the H 202 could inactivate the enzyme. Consistent with the concept that H 202 or H202-derived oxidants may mediate suppression of fibroblast proliferation by GSH is the observation in the present study that catalase completely inhibited the effects of GSH on fibroblast proliferation. Interestingly, although autoxidation likely mediated fibroblast growth inhibition, most of the GSH (96 ± 1%) remained in the reduced state as it is in normal alveolar epithe-
liallining fluid (2). These results suggest that GSH at the alveolar surface may simultaneously play a role in fibroblast growth regulation as observed in this study and in providing antioxidant protection of alveolar cells as previously described (2). GSH may also act indirectly on the fibroblast to suppress its proliferation by inducing the release of growth-inhibitory molecules from the fibroblast itself. An example of this would be the enhanced synthesis of prostaglandin E 2 (PG~), a known suppressant of lung fibroblast proliferation (25). GSH is a cofactor in the synthesis of PGE 2 (26). Although this mechanism cannot be excluded, it would seem unlikely, since intact GSH is not usually taken up by cells (27) and, in the present study, all added GSH was recovered in the extracellular milieu after the incubation period. Finally, a nonspecific cytotoxic effect of GSH must be considered. The concentrations ofGSH used in these studies are well within the physiologic range found in normal ELF (2). At these concentrations, we found no morphologic evidence of cytotoxicity and no increase in 5\Cr release, and all cells excluded trypan blue dye. Furthermore, the cells were still capable of proliferating in the presence of GSH, albeit at a reduced rate. However, we cannot exclude the possibility that GSH induced a sublethal nonspecific toxic effect. Although the mechanisms of GSH-mediated inhibition of fibroblast proliferation are not entirely clear, several lines of evidence suggest that GSH deficiency may be associated with enhanced fibrosis. First, IPF patients have marked interstitial fibrosis and are deficient in alveolar epithelial lining fluid GSH (3). Second, an excessive oxidant burden rapidly depletes cellular stores ofGSH (28-30), and several investigators have reported experimental and clinical data demonstrating that an excessive lower respiratory tract oxidant burden leads to interstitial pulmonary fibrosis (31, 32). Third, histologic studies of liver tissue have demonstrated that areas of liver fibrosis are associated with increased '}'-GT (33), an enzyme that can break down GSH (22). Finally, depletion of lung GSH by buthionine sulfoximine treatment after ozone exposure markedly increased pulmonary interstitial fibrosis in mice (34). Because GSH depletion was induced after ozone exposure, it is possible that the mechanisms of enhanced fibrosis were not only related to the antioxidant properties of GSH but also to its suppressive effects on fibroblast proliferation during the tissue repair phase. In summary, GSH, at concentrations measured in the extracellular fluid of the normal lower respiratory tract, can suppress fibroblast proliferation. The sulfhydryl group of GSH is essential to inhibit fibroblast proliferation, and most of the observed suppressive effect seems to result from the interaction between the rapidly dividing fibroblast and lowgrade GSH autoxidation. These results suggest that the decreased extracellular GSH levels in the lower respiratory tract of patients with IPF may create an environment that facilitates lung fibroblast proliferation in response to mitogens within the local milieu. Acknowledgments: The writers thank Ginette Bilodeau and Denis Bisson for technical support and Marguerite Cloutier for help in preparing this manuscript. This work was supported by a grant from the Medical Research Council (Canada) and Association Pulmonaire du Quebec. Dr. Cantin is a scholar of the Fonds de la Recherche en Sante du Quebec.
Cantin, Larivee, and Begin: Glutathione Suppression of Fibroblast Growth
References 1. Meister, A., and M. E. Anderson. 1983. Glutathione. Annu. Rev. Biochem. 52:711-760. 2. Cantin, A. M., S. L. North, R. C. Hubbard, and R. G. Crystal. 1987. Normal Alveolar epithelial lining fluid contains high levels of glutathione. J. Appl. Physiol. 63:152-157. 3. Cantin, A. M., R. C. Hubbard, and R. G. Crystal. 1989. Glutathione deficiency in the epithelial lining fluid of the lower respiratory tract in idiopathic pulmonary fibrosis. Am. Rev. Respir. Dis. 139:370-372. 4. Hoidal,1. R., R. B. Fox, P. A. Lemarbre, R. Perri, and 1. E. Repine. 1981. Altered oxidative metabolism responses in vitro of alveolar macrophages from asymptomatic smokers. Am. Rev. Respir. Dis. 123:85-89. 5. Cantin, A. M., S. L. North, G. A. Fells, R. C. Hubbard, and R. G. Crystal. 1989. Oxidant-mediated epithelial cell injury in idiopathic pulmonary fibrosis. J. Clin. Invest. 79:1665-1673. 6. Cosio, M. G., K. A. Hale, and D. E. Niewoehner. 1980. Morphologic and morphometric effects of prolonged cigarette smoking on the small airways. Am. Rev. Respir: Dis. 122:265-271. 7. Crystal, R. G., 1. D. Fulmer, W. C. Roberts, M. L. Moss, B. R. Line, and H. Y. Reynolds. 1976. Idiopathic pulmonary fibrosis: clinical, histologic, radiologic, physiologic, scintigraphic, cytological and biochemical aspects. Ann. Intern. Med. 85:769-788. 8. Crystal, R. G., P. B. Bitterman, S. I. Rennard, A. 1. Hance, and B. A. Keogh. 1984. Interstitial lung diseases of unknown cause: disorders characterized by chronic inflammation of the lower respiratory tract. N. Engl. J. Med. 310:154-166. 9. Deuel, T. E, 1. S. Huang, R. T. Proffitt, 1. V. Baenziger, D. Chang, and B. B. Kennedy. 1981. Human platelet-derived growth factor: purification and resolution into two active protein fractions. J. Bioi. Chem. 256:88968899. 10. Martinet, Y., W. N. Rom, G. R. Grotendorst, G. R. Martin, andR. G. Crystal. 1987. Exaggerated spontaneous release of platelet-derived growth factor by alveolar macrophages from patients with idiopathic pulmonary fibrosis. N. Engl. J. Med. 317:202-209. 11. Betsholtz, C., A. Johnsson, C. H. Heldin et al. 1986. cDNA sequence and chromosomal localization of human platelet-derived growth factor Achain and its expression in tumor cell lines. Nature 320:695-699. 12. Ross, R. and A. Vogel. 1978. The platelet-derived growth factor. Cell 14:203-210. 13. Yarden, Y., 1. A. Escobedo, W. 1. Kuang etal. 1986. Structure of the receptor for platelet-derived growth factor helps define a family of closely related growth factor receptors. Nature 323:226-232. 14. Williams, L. T. 1989. Signal transduction by the platelet-derived growth factor receptor. Science 243:1564-1570. 15. Cantin, A., R. Begin, M. Rola-Pleszczynski, and R. Boileau. Heterogeneity of bronchoalveolar lavage cellularity in stage 11I pulmonary sarcoidosis. Chest 83:485-486. 16. Killingsworth, L. M., and J. Savory. 1972. Manual nephelometric methods for imrnunochemical determination of immunoglobulins IgG, IgA, and IgM in human serum. Clin. Chem. 18:335-339. 17. Freshney, R. I. Measurement of cytotoxicity and viability. 1987. In Culture of Animal Cells: A Manual of Basic Technique. R. I. Freshney, editor.
85
Alan R. Liss, New York: 245-256. 18. Pick, R., and Y. Keisari. 1980. A simple colorimetric method for the measurement of hydrogen peroxide produced by cells in culture. J. Immunol. Methods 38:161-170. 19. Snedecor, G. w., and W. G. Cochrane. 1980. Statistical Methods, 7th ed. Iowa State University Press. Ames, IA. 20. Brody, A. R., and 1. E. Craighead. 1976. Interstitial association of cells lining air spaces in human pulmonary fibrosis. Virchows Arch. [A] 372: 39-49. 21. Bitterman, P. B., S. Adelberg, and R. G. Crystal. 1983. Mechanismsofpulmonary fibrosis: spontaneous release of the alveolar macrophage-derived growth factor in the interstitial lung disorders. J. Clin. Invest. 72:18011813. 22. Abbott, W. A., R. 1. Bridges, and A. Meister. 1984. Extracellular metabolism of glutathione accounts for its disappearance from the basolateral circulation of the kidney. 1. Bioi. Chem. 259:15393-15400. 23. Edelman, G. M. 1987. CAMs and IGs: cell adhesion and the evolutionary origins of immunity. Immunol. Rev. 100:11-45. 24. Trotta, P. P., L. M. Pinkus, and A. Meister. 1974. Inhibition by dithiothreitol of the utilization of glutamine by carbamyl phosphate synthetase. J. Bioi. Chem. 249:1915-1921. 25. Bitterman, P. B., M. D. Wewers, S. I. Rennard, S. Adelberg, and R. G. Crystal. 1986. Modulation of alveolar macrophage-driven fibroblast proliferation by alternative macrophage mediators. J. Clin. Invest. 77:700708. 26. Metz, S. A. 1981. Anti-inflammatory agents as inhibitors of prostaglandin synthesis in man. Med. Clin. North Am. 65:7I'J-757. 27. Meister, A. 1985. Methods for the selective modification of glutathione metabolism and study of glutathione transport. Methods Enzymot. 113: 571-585. 28. Schraufstiitter, I. V., D. B. Hinshaw, P. A. Hyslop, R. G. Spragg, and C. G. Cochrane. 1985. Glutathione cycle activity and pyridine nucleotide levels in oxidant induced injury of cells. J. Clin. Invest. 76:1131-1139. 29. Dunbar, 1. R., A. 1. Delucia, and L. R. Bryant. 1984. Glutathione status of isolated rabbit lungs: effects of nitrofurantoin and paraquat perfusion with normoxic and hyperoxic ventilation. Biochem. Pharmacol. 33:13431348. 30. Delucia, A. J., M. G. Mustafa, M. Z. Hussain, and C. E. Cross. 1975. Ozone interaction with rodent lung. III. Oxidation of reduced glutathione and formation of mixed disulfides between protein and nonprotein sulfhydryls. J. Clin. Invest. 55:794-802. 31. Johnson, K. 1., and P. A. Ward. 1982. Acute and progressive lung injury after contact with phorbol myristate acetate. Am. J. Pathol. 107:29-35. 32. Pratt, P. C. 1974. Pathology of pulmonary oxygen toxicity. Am. Rev. Respir: Dis. 110(Suppl):51-57. 33. Yamauchi, M., K. Kimura, H. Kawase et at. 1984. Hepatic gammaglutamyl transferase and hepatic fibrosis in patients with alcoholic liver disease. Enzyme 32:110-115. 34. Sun, J. D., J. A. Pickrell, J. R. Harkema, S. J. McLaughlin, E E Hahn, and R. E Henderson. 1988. Effects ofbuthionine sulfoximine on the development of ozone-induced pulmonary fibrosis. Exp. Mol. Pathol. 49: 254-266.
