Receptors for Transforming Growth Factor-S (lGF-{j) on Rat Lung Fibroblasts Have Higher Affinity for IDF-,81 than for IDF-{j2 Valerie G. Kalter and Arnold R. Brody Laboratory of Pulmonary Pathobiology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina

Most cell types have receptors for transforming growth factor-S (TGF-{j) and respond similarly to TGF-{jl and TGF-{j2' We have demonstrated the presence of a single class of high-affinity receptors (1'\.110,000 sites/cell) for TGF-{jl (K; = 23 pM) and TGF-{j2 (K, = 41 pM) on early-passage rat lung fibroblasts (RLF). Incubation with unlabeled TGF-{jl and TGF-{j2 resulted in concentration-dependent inhibition of binding of 15 pM (12SI]TGF-{j1 (EDso, 20 and 28 pM, respectively) and (1 2SI]TGF-{j2 (ED so, 36 and 56 pM, respectively). TGF-{j receptors affinity-cross-linked with 100 pM (12SI] TGF-{j I or (1 2SI]TGF-{j2 were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis and exhibited labeled protein bands of 68, 88, and 286 kD. Densitometric analysis of the resulting autoradiograms showed that the different molecular weight TGF-{j binding proteins exhibited separate affinities for the two forms of TGF-{j. Both TGF-{jl and TGF-{j2 altered the morphology and cytoskeleton of RLF in a similar manner, but TGF-{jl was more potent than TGF-{j2 in the inhibition of RLF growth and colony formation, with 50 % inhibition by 0.12 pM TGF-{j. and 4.4 pM TGF-{j2' Different affinities for the TGF-{js may indicate selectivity among the receptor subtypes with regard to the biologic responsiveness of RLF to TGF-{js. We believe this to be the first demonstration of biologically responsive TGF-{j receptors with different affinities for TGF-{j. and TGF-{j2 on cells derived from normal, nonimmortal RLF. In establishing the basic mechanisms of pulmonary fibrosis, it will be essential to understand the biology and biochemistry of the receptors that may control cell division and production of extracellular matrix components by fibroblasts.

Transforming growth factor-S (TGF-{j) is a 25-kD protein produced by most normal and transformed cells and was originally characterized by its ability to reversibly induce anchorage-independent growth of nontransformed cells (1). Now known to exist in at least two distinct but structurally related forms, TGF-{jl and TGF-{j2 (2, 3), the TGF-l3s are multifunctional agents capable of either stimulating or inhibiting proliferation of a multitude of cell types (4, 5). Importantly, in terms of understanding interstitial pulmonary fibrosis, TGF-{j induces profound effectson the synthesis and secretion of many components of the extracellular matrix (Receivedin originalform October25, 1990 and in revisedform December 4, 1990) Addresscorrespondence to: Arnold R. Brody, Ph. D., Laboratory of Pulmonary Pathobiology, National Institute of Environmental Health Sciences, P.O. Box 122331D2-02, Research Triangle Park, NC 27709.

Abbreviations: alveolar macrophages, AM; bovine serum albumin, BSA; colony-forming efficiency, CFE; Dulbecco's modified Eagle's medium, DMEM; disuccinirnidyl suberate, DSS; dithiothreitol, DTT; extracellular matrix, ECM; epidermal growth factor, EGF; fetal bovine serum, FBS; N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid, Hepes; insulin, INS; polyacrylamide gel electrophoresis, PAGE; phosphate-buffered saline, PBS; platelet-derived growth factor, PDGF; phenylmethylsulfonyl fluoride, PMSF; rat lung fibroblasts, RLF; sodium dodecyl sulfate, SDS; transforming growth factor-d. TGF-p. Am. J. Respir. Cell Mol. BioI. \bl. 4. pp. 397-407, 1991

(ECM) and on cell differentiation (6-8). Much of the pulmonary fibrogenic response to inhaled particles consists of increased synthesis and deposition of collagen, fibronectin, and other components of the ECM in the tissues surrounding the alveolar spaces (9-11). The excess production of ECM proteins by fibroblasts, usually accompanied by fibroblast proliferation (12, 13), is largely responsible for thickening of the pulmonary interstitium, which decreases respiratory gas exchange. Modulation of cell growth and stimulation of ECM production by TGF-{j in several types of normal and transformed fibroblastic cells and cell lines have been reported (8, 14-16). These important biologic properties have been shown recently in a human embryonic lung fibroblast cell line (17) and in fibroblasts obtained from normal adult human lungs (18). Almost all cell types tested have been shown to possess specific high-affinity receptors for TGF-{j (19-24), with a range of binding affinity constants (Kd) of 1 to 100 pM and variable numbers of receptors (1,000to 80,000 per cell) (24). Three distinct types of TGF-{j glycoprotein receptor molecules with different molecular weight ranges have been identified using affinity-labeling techniques (2, 19, 21, 25). These have been designated as type I (Mr , 50,000 to 80,000), type II (Mr , 85,000 to 140,000), and type ill (Mr , 250,000 to 330,000) (19, 25, 26). The type III receptor is a disulfide-linked membrane proteoglycan complex of 565 to

398

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

615 kD (27). The pattern of TGF-13 binding to different cell strains and established cell lines is somewhat variable; some cells lack the type III receptor protein but remain responsive to TGF-13 (25). We have established a reliable procedure for obtaining normal primary rat lung fibroblasts and maintaining earlypassage rat lung fibroblasts (RLF) in culture (28). Using this cell culture system, we present here in detail the binding kinetics and receptor subunit structures for both [J25I]TGF-/31 and [I25I]TGF-132 under identical conditions. In these experiments, we demonstrate that the RLF TGF-13 receptor exhibits different affinities for TGF-/31 and TGF-132 in both homologous and heterologous ligand-receptor studies of the two TGF-l3s, and that the biologic potency of TGF-131 is greater than that of TGF-/32 in growth-inhibition studies. We believe this to be the first demonstration of biologically responsive TGF-13 receptors with different affinities for TGF-/31 and TGF-132 on normal, nonimmortal fibroblastic cells derived from adult rat lung.

Materials and Methods Growth Factors and Reagents 125I-Iabeled porcine TGF-131 (sp act, 66 to 144 p.Ci/p.g) and TGF-132 (sp act, 45 to 165 p.Ci/p.g), unlabeled pure and crude porcine TGF-/31 and TGF-/32, and pure human platelet-derived growth factor (PDGF) were purchased from R&D Systems (Minneapolis, MN). Mouse epidermal growth factor (EGF) was from Collaborative Research (Lexington, MA). The following reagents were obtained from Sigma Chemical Co. (St. Louis, MO): bovine serum albumin (BSA), N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (Hepes), deoxyribonuclease I, soybean trypsin inhibitor, Triton X-100, dithiothreitol (DTT), antipain, aprotinin, bestatin, leupeptin, pepstatin A, benzamidine hydrochloride, phenylmethylsulfonyl fluoride (PMSF), bovine insulin (INS), and Brilliant Blue G colloidal concentrate. Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), trypsin (1:250, lOx concentrate in normal saline), and L-glutamine were supplied by GIBCO (Grand Island, NY). Collagenase (284 U/mg) was purchased from Worthington Biochemicals (Freehold, NJ). Sodium dodecyl sulfate (SDS), electrophoresis purity grade premixed acrylamide/bisacrylamide, and bromophenol blue were from BioRad Laboratories (Richmond, CA). Rainbow protein molecular weight markers were purchased from Amersham (Amersham, UK), and disuccinimidyl suberate (DSS) was from Pierce Chemical Co. (Rockford, IL). Seprabuff (Laemmli running buffer) was obtained from Enprotech (Hyde Park, MA), and Tris (ultrapure) was supplied by Bethesda Research Laboratories (Gaithersburg, MD). Rhodamine-conjugated phalloidin was purchased from Molecular Probes (Eugene, OR). AR-XOmat film and Dupont Lightning Plus Screens were obtained from Eastman Kodak Co. (Rochester, NY). Giemsa stain, glycerol, formaldehyde, and Hematall (lsoton IT) were from Fisher Scientific (Orangeburg, NY). All other reagents used were either reagent grade or electrophoresis purity grade chemicals. Plastic tissue culture flasks and dishes were supplied by Costar (Cambridge, MA) and Falcon (Becton Dickinson Labware, Lincoln Park, NJ). Lab-Tek tis-

sue culture chamber slides were purchased from Nunc (Naperville, IL). RLF Cultures

The procedure for the isolation of fibroblasts from adult rat lungs by enzymatic dissociation has been described (28, 29). Isolated fibroblasts were grown to confluence in DMEM + 10% FBS in T-150 flasks (Costar). They were then trypsinized, counted, and cryopreserved at 1 x 1Q6 cells/ml in DMEM containing 10% dimethyl sulfoxide. For binding assays, fibroblasts were thawed, grown to confluence, trypsinized, and seeded at 4 to 10 X 1~ cells/well in Costar 24well plastic dishes. All experiments were carried out with second-passage RLF isolated from a single animal. Ongoing experiments with second-passage cells from three other animals have yielded similar results. Cell counts were determined with a Coulter counter or a hemocytometer. Cells were allowed to reach confluence in DMEM + 10% FBS, usually after 2 to 3 d, and were then used for radioligand binding studies. By allowing the cells to reach confluence, the nonspecific binding was reduced from 30 % to 5 % of the input radioactivity. Fibroblast cultures were photographed with a Polaroid camera mounted on a Leitz Labovert phase-contrast microscope. Binding Assays The binding assay for confluent fibroblast monolayers was a modification of the procedures of Massague (21) and Wakefield and colleagues (24). Confluent monolayer cultures were washed twice with DMEM containing 25 mM Hepes and 0.1% BSA (dissociation buffer) and incubated in this medium for 2 h at 37 0 C to allow dissociation and/or internalization of endogenous bound TGF-/3. The monolayers were then washed twice with ice-cold binding buffer (128 mM NaCI, 5 mM KCI, 5 mM MgS04 , 1.2 mM csci, 50 mM Hepes [pH 7.5], 10 mg/ml BSA) and allowed to equilibrate in this buffer for at least 30 min at 4 0 C. Duplicate or triplicate wells were incubated with 0.5 ml of binding buffer containing 0 to 400 pM unlabeled TGF-13 and 15 to 20 pM [J25I]TGF-13 for competition and time-course experiments. For saturation binding analysis (30), fibroblasts were grown as described above, followed by the dissociation and equilibration procedures. Triplicate cultures were then incubated with 0.5 ml of binding buffer containing 1 to 500 pM [J25I]TGF-/31 or ['25I]TGF-/32 in the absence or presence of a 50-fold excess of unlabeled TGF-/3. Incubation with binding media proceeded for 4 to 6 h at 4 0 C with constant shaking at 150 rpm on a rotary shaker platform. After equilibration, aliquots of the binding media were saved for the determination of free radioactivity, and the binding reaction was terminated by rapidly washing the cultures 5 times with ice-cold binding buffer. After washing, the cells were incubated with 0.5 ml of solubilization buffer (1% [vol/vol] Triton X-100, 10% [vol/vol] glycerol, 25 mM Hepes, 10 mg/ml BSA) for 40 min at 4 0 C with constant shaking. The solubilized extracts were removed and counted in a Packard Auto Gamma 500C gamma counter. Specific binding of [J25I]TGF-/3 was determined by subtracting the amount of radioactivity bound in the presence of unlabeled TGF-13 from the radioactivity bound in wells incubated with (125I]TGF-13 alone. Addi-

Kalter and Brody: Rat Lung Fibroblast TGF-13 Receptors

tional wells were washed and incubated without labeled TGF-I3, then trypsinized and processed for the determination of the number of cells per well. Results are expressed in femtomoles (fmol) ['2SI]TGF-13 bound/l X 106 cells, except where indicated otherwise. Usually there were 120,000 to 200,000 cells/well. For downregulation experiments, some cultures were incubated with 400 pM TGF-131 or TGF-132 for 16 h at 37 0 C, then washed extensively with dissociation buffer at 37 0 C for 2 h. Saturation binding assays were performed as previously described. Affinity Labeling with DSS The method for the affinity labeling of TGF-13 receptors has been described (21). Briefly, cell monolayers were incubated with 100 pM ['2SI]TGF-131 or (12SI]TGF-132 plus 0 to 4,000 pM unlabeled TGF-13 and washed as above. The medium was changed to 1 ml/well binding buffer without BSA, and cells were incubated with 27 mM DSS for 15 min at 4 0 C with agitation. Cells were detached in ice-cold 0.25 M sucrose, 10 mM Tris, 1 mM EDTA (pH 7.4),0.3 mM PMSF; scraped; pelleted at 12,000 x g for 2 min; and solubilized in 60-I.d cell solubilization buffer (125 mM NaCI, 10 mM Tris, 1 mM EDTA [pH 7.0], 1% Triton X-lOO) containing eight protease inhibitors (21) for 40 min at 4 0 C. Alternatively, cells were solubilized in the culture dishes. After solubilization, debris was removed by centrifugation at 12,000 x g for 15 min. The supernatants were mixed with an equal volume of 2 x concentrated electrophoresis sample buffer (31) containing 100 mM DTT, heated at 100 0 C for 2 min, and frozen until they were subjected to electrophoresis. Gel Electrophoresis and Autoradiography Discontinuous gel electrophoresis on 7.5 % SDS-polyacrylamide gels was performed according to the procedures of Laemmli (31) with either a Bio-Rad Protean II or Hoeffer SE-600 vertical slab gel electrophoresis unit. The gels were fixed, stained in Brilliant Blue G, destained, and dried under vacuum. Dried gels were exposed to Kodak AR-XOmat film using Dupont Lightning Plus Screens and stored at -70 0 C for 1 to 5 wk. The films were developed, and the molecular weights of the resulting bands were determined from calibration curves of simultaneously electrophoresed molecular weight markers for each gel. The intensity of the labeled bands was quantified with a Joyce Loebel densitometer or a Bio-Rad Model 620 Video Densitometer. Effects of TGF-{3s on the Cytoskeleton of RLF For the localization of actin microfilaments, RLF were grown on glass coverslips for 72 h in the presence or absence of 1 ng/ml (40 pM) TGF-l3t or TGF-132 in DMEM + 10% FBS. The coverslips were washed in phosphate-buffered saline (PBS), fixed for 15 min in 3.7% paraformaldehyde, washed, and the cell membranes were permeabilized by a l-min exposure to 0.5 % Triton X-I00 in PBS. After washing, the cells were stained with rhodamine-phalloidin for 20 min at 37 0 C and then washed extensively in PBS. The coverslips were mounted with Gelvatol and sealed with nail polish. Labeled cells were examined with a rhodamine filter on a Vanox fluorescence photomicroscope equipped for epifluorescence, and photographed with Kodak TMax film.

399

Cell Proliferation and Colony-forming Efficiency (CFE) Assays For cell proliferation and CFE assays, RLF were thawed and grown to confluency as above, trypsinized, counted, and seeded at 250 cells/60-mm dish (Falcon) in 4 ml of DMEM + 10% FBS. After allowingthe cells to attach for 1 h at 370 C, 1 ml of test medium (DMEM + 10% FBS with or without added growth factors) was added to the dishes, and the cells were maintained for up to 8 d in this medium. The media were not changed, and the growth factors were added only once as indicated. CFE dishes were washed twice with 70% ethanol, fixed in Giemsa stain for 15 min at room temperature, rinsed with distilled water, and air-dried. The %CFE was defined as the number of colonies larger than 12 cells/250 multiplied by 100. Control cultures contained 100 to 120 colonies/dish after 7 to 8 d in culture, with a CFE of 40 to 47%. The number, size, and appearance of the colonies were also recorded. For proliferation assays, cells were washed twice with PBS containing 1 mM EDTA and trypsinized, and l-ml aliquots were mixed with 19 ml of Hematall for the determination of cell number in a Coulter counter. Results were expressed as the mean ± SEM. Statistical Analysis Competition and saturation binding assays were analyzed with the LIGAND program (32). Statistical comparisons were made with the SYSTATprogram (33), using t tests or ANOVA when appropriate.

Results Time Course of (12SI]TGF-{3 Binding to RLF The time course of (12SI]TGF-131 binding was determined at 4 0 C (Figure lA). All experiments were conducted at 4 0 C in order to avoid internalization and/or degradation of the labeled TGF-13 (34, 35). About 10 to 15% of the ['2SI]TGF-l3t added per well (15 to 20 pM (125I]TGF-l3t, or 50,000 to 100,000 cpm/well) was specifically bound and 2 to 3 % was nonspecifically bound to the cells. The level of specific binding was 75 to 80 % of the total binding at each time point assessed. Saturation and maximal binding were achieved at 4 to 5 h of incubation, and all subsequent experiments were performed with a 4- to 5-h incubation period. Similar results were obtained for the time course of ['25I]TGF-132 binding to RLF at 4 0 C (Figure lB). The maximal amount of ['2SI]TGF-132 specifically bound at saturation (6 to 8 fmol/ 106 cells) was always less than that of ['2SI]TGF-131 (12 to 16 fmol/l()6 cells), with all experiments carried out under the same conditions. Specificity of [125I]TGF-{3 Binding to RLF The specificity of (125I]TGF-13 binding to RLF was determined by incubation of RLF with 15 pM (125I]TGF-13 and 0 to 100 ng/ml unlabeled TGF-I3I' TGF-132' EGF, PDGF, or INS. As illustrated in Figure 2, only TGF-l3s were able to inhibit the specific binding of (125I]TGF-l3t or (12SI]TGF-132 in a concentration-dependent manner. The homologous form of TGF-13 competed more effectively than the heterologous form. For (125I]TGF-131 the EDso was 20 pM for TGF-l3t and 28 pM for TGF-132' while for (12SI]TGF-132 the ED so was 36

400

~

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

..

~

20

Gi

III

~

16

S

12

~

16

S

12

Iii CII '0

'0

'0

::3 0

::3 0

e::

e::

ID

c;

ID

8

N

on

8

~

IL C)

IL C)

~

B

0

0 III

Iii CII '0

20

'7

4

on

4

~

~

2

4

3 Time

5

6

7

8

2

(Hr)

Time

pM for TGF-J32 and 56 pM for TGF-J31. Statistical analysis of the combined homologous and heterologous displacement curves with the LIGAND program (32) revealed a significant difference in the affinity of the RLF TGF-J3 receptors for lGF-J31 and lGF-J32 (P < 0.05) and also showed that RLF possess only one set of binding sites for labeled lGF-J3s. 120

r--......-'P"""I"'......,~

-.-......... .....,.-.....................rr--.-.....................,

100

80

"c::s 0

m

60

'#. 40

20

0 .01

.1

3

10

100

[Growth Factor] (ng/ml)

Figure 2. Specificity of [J251]TGF-{3\ and (1 251]TGF-{32 binding to RLF cultures. Cells were incubated for 4 h at 4 0 C with 20 pM (1 251]TGF-{31 or [J25I]TGF-{32 and various amounts of unlabeled TGF-(3), TGF-{32' platelet-derived growth factor (PDGF), epidermal growth factor (EGF), or insulin (INS). Closed circles: [ 125I]TGF-{31 versus TGF-{3\; open circles: (1 251]TGF-{3\ versus TGF-{32; open triangles: (125I]TGF-{32 versus TGF-{32; closed triangles: (125I]TGF-iS2 versus TGF-iS\; closed squares: PDGF; open squares: EGF; crossmarks: INS. Only TGF-{3s inhibited the specific binding of [I25I]TGF-,8. or (1 251]TGF-iS2 in a concentrationdependent manner. Data points are the mean ± SD of triplicate determinations. Two-way ANOVA revealedthat RLF TGF-{3 receptors have different affinities for the two TGF-{3s (F-ratio, 14.079; P < 0.002). A representative experiment of three repetitions is shown here.

4

5 (Hr)

6

7

8

Figure 1. Time course of binding of (1 251]TGF-{3 . (A) and (1 251] TGF-{32 (B) to rat lung fibroblasts (RLF) at 4 0 C. Triplicate wells were incubated with 20 pM (1 251]TGF-{3 in the absence (open circles) or presence of2 nM unlabeled TGF-{3 (triangles). Specific (1 251]TGF-{3 bound (closed circles) was determined as the difference between total binding (open circles) and nonspecific binding (triangles). At saturation (4 to 5 h), RLF cultures specifically bound 12 to 16 fmol (1 251] TGF-{3\I1Q6 cells (A) or 6 to 8 fmol [J251]TGF-,82/1Q6 cells (B). Data points are the mean ± SD.

Nonlinear curve-fitting revealed no significant improvement with a two-site model (P> 0.05). Two-way ANOVA of the combined homologous and heterologous competition data from several repetitions of this experiment with the SYSTAT program (33) showed that RLF lGF-J3 receptors have different affinities for the two TGF-J3s (F-ratio = 14.079; P < 0.002). Affinity and Number of TGF-,8 Receptors on RLF Estimation of the receptor binding affinity (Kd) and number of binding sites (Ro) for [t25I]TGF-J31 and [t25I]TGF-J32 were obtained by equilibrium binding of the labeled TGF-J3s to RLF cultures at 4 0 C. Analysis of the binding isotherms indicated that saturation ofTGF-J3 receptors was achieved at 150 to 300 pM for both (I25I]TGF-J3. (Figure 3A) and (I25I]lGF132 (Figure 3B). Scatchard analysis (30) of the binding curves revealed the presence of a single class of high-affinity binding sites for lGF-J3\ and lGF-J32 (Figure 3C). The apparent K; values obtained from the slopes of the Scatchard plots shown in Figure 3 are 17.9 pM for (125I]TGF-131 and 44.2 pM for [t25I]lGF-J32' The number of binding sites per cell determined from the x-axis intercepts in the Scatchard plots in Figure 3 are 10,200 sites/cell for (125I]TGF-131 and 8,500 sites/cell for [I25I]TGF-J32' Although there was some variation in the K, and R, values obtained in different experiments (Table 1), there was always at least a 2-fold difference in affinity for the two lGF-J3s when saturation analyses were performed on the same day with the same batch of cells, and this difference was significant (P < 0.05) when the results were analyzed by Student's t test with the SYSTAT program (33). The number of binding sites for TGF-J32 was usually within 10% of that for lGF-J3\. Replotting the data according to the method of Klotz (36) confirmed that the number of binding sites had been correctly estimated by Scatchard analysis and that we had achieved saturation in the equilibrium binding assays. The K, values obtained by saturation analysis were in good agreement with the ED 50 values from the competition assays (Figure 2). Combined analysis of the data for all saturation binding experiments with the LIGAND program (32) indicated that no significant improvement in the single-binding site model could be obtained with the use of nonlinear curve-fitting for a two-site model for either [t25I]TGF-J3\ or [I25I]TGF-J32 (P > 0.05).

Kalter and Brody: Rat Lung Fibroblast TGF-{j Receptors

A

30

B

30

~

401

0.4 ...--................

~

~ 25

~ 25 ,..

CD 0

CD 0

,..

:20

~0

'0

0.3

20 Gl Gl

~

~

"t:: 15

"e 15

~O.2

CD

CD

:::I 0

at

:::I

:::I 0

CD

:;

c

--..-.--.,..............."""'T'""---.----,

10

c-.

o

10

~

0.1

u.

u, C)

C)

'":" 5

'":" 5

.n

,;) ~

~

0.0

100

200

300

100

[ 125 1_TGF Bl] (pM)

200

[125 1_TGF B2] (pM)

300

'--"""--"""-----'-----L..--o.----I----:_ _...J.....),..-"----J

o

4

8

12

16

20

125 1_TGF B Bound (fmoles/l0 6 Cells)

Figure 3. Equilibrium binding of (125I]TGF-J31 (A) and (125I]TGF-J32 (B) to RLFcultures. Triplicate cultures were incubated with 1 to 500 pM (125I]TGF-J3 alone (open circles) or in thepresence of a 50-fold excess of unlabeled TGF-J3 (triangles) for 4 to 5 h at 4 C. The saturation binding isotherm for specific (125I]TGF-J3 bound (closed circles) was determined as the difference between total (open circles) and nonspecific (triangles) binding of labeled TGF-J3. Representative experiments for [t~I]TGF-J31 (n = 6) and (125I]TGF-J32 (n = 4) saturation binding are shown here. (See Table 2.) Saturation of TGF-J3 receptors was reached at 150 to 300 pM for both (125I]TGF-J31 (A) and (125I]TGF-J32 (B). (C) Scatchard analysis ofthe specific binding of (125I]TGF-J31 (closed circles) and (125I]TGF-J32 (open circles) to RLFcultures. The line is the least-squares fit to the points. 0

Affinity Labeling of (125I]1GF-J3 Receptor Proteins

(1 251]TGF-J31 and (1 251]TGF-{j2 were cross-linked to RLF TGF-J3 receptor proteins with DSS (21), and the labeled proteins were displayed by SDS polyacrylamide gel electrophoresis (PAGE) and detected by autoradiography (Figure 4). (1 251]TGF-J3\ specifically labeled three major protein bands with approximate molecular weights of 68, 88, and 286 kD, similar to the previously identified TGF-{j type I, type II, and type ill receptor species (19, 21, 25). The labeling of the three receptor species was inhibited in a concentration-dependent manner by the presence of increasing amounts of unlabeled TGF-{jlo but was not inhibited by excess PDGF, EGF, or INS (Figure 4A). Solubilized extracts from cells incubated with 100 pM (1 251] TGF-{j I without cross-linking to DSS showed no specific labeling of any protein bands (Figure 4A, lane 8). In the absence of any competitor, 100 pM

TABLE 1 Summary of Scatchard analysis of binding data* s, (PM)t Ligand Treatment Ro (sites/cell)t [125I]TGF-~1

Control DR

P25I]TGF-~2

Control DR

23.2 ± 3.5 46.6 ± 6.9* 40.7 ± 7.8* 40.6 ± 0.8

10,550 2,950 9,100 2,600

± ± ± ±

1,400 800:1: 1,600 5OO§

n

6 3 4 2

* Rat lung fibroblast cultures were incubated with increasing amounts of or P25I]TGF-~2 as described inMATERIALS AND METHODS. The results were subjected to Scatchard analysis (30) with curve-fitting performed with the LIGAND program ofMunson and Rodbard (32) and statistical analysis with the SYSTAT program (33). The values in the table represent the mean ± SEM for the K.:t and number of binding sites for untreated (control) and downregulated (DR) rat lung fibroblast cultures. n = number ofexperiments. t Values are mean ± SEM. :I: P < 0.05, significantly different from TGF-~I control (Student's t test). § P < 0.05, significantly different from TGF-~2 control (Student's t test). P2SI]TGF-~1

(1 251] TGF-{j I labeled the 286-kD species about 7 to 8 times as intensely as the 68- and 88-kD species of receptors (Figure 4A). As with labeled TGF-{jlo 100 pM (1 251]TGF-{j2 labeled three receptor species with molecular weights similar to those labeled by (1 251]TGF-131(Figure 4B). While the intensity of labeling of the 286-kD receptor by (1 251]TGF-J32 was similar to that of (1 251]TGF-J31o this type III receptor species was labeled about 18 to 20 times as intensely as the 68- and 88-kD subunits (Figure 4B). The labeling with (125I]TGF-132 was inhibited by the presence of increasing amounts of unlabeled TGF-{j2' but was not decreased by PDGF, EGF, or INS (Figure 4B). The labeling of (125I]TGF-131 was reduced by TGF-{j2 as was the labeling of (125I]TGF-{j2 by TGF-I3\ (Figure 4C), but not to the same extent as with the homologous ligands (Figures 4A and 4B). The competition of unlabeled TGF-{js with (1 251]TGF-J3labeled receptor subunits in the autoradiograms from Figure 4 was analyzed by densitometric scanning. The amount of labeling in each lane was expressed as the percentage of the peak area labeled in the control cell extracts, i.e., 100 pM (1 251]TGF-{jl or (1 251]TGF-J32 labeling with no competitors. Unlabeled TGF-J3\ inhibited the labeling intensity of [1251] TGF-131 receptors by one-half with the following concentrations: type I, 150 pM; type II, 130 pM; and type ill, 120 pM (Table 2). Unlabeled TGF-132 inhibited one-half the intensity of labeling of (1 25I]TGF-132 receptors with these concentrations: type I, 480 pM; type II, 800 pM; and type ill, 160 pM (Table 2). TGF-J32 and TGF-131 competed much less successfully for each other's labeled receptor proteins, except that TGF-J31 and TGF-132 competed similarly for the (1251]TGF-{j2-labeled type III receptor proteins (Table 2). The affinity of (1 251]TGF-131 and (1 251]TGF-{j2 for the type III receptor was approximately the same as previously demon-

402

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL.

4 1991

+100pM 1251-TGFP

A

I,-----~A 1

2

3 4

5 6 7

1

,

8

9 10 11

B

46-

Addition

C)

evil?

it)

-? ~ ~ ~

Addition

0

it)

-? ~

J? ~

C)

Ii)

~~~ ~

~~ '-----.v,,----'

TGFP j (ng/mll

(lOOng/ml)

(lOOng/ml)

TGFP2 (ng/mll

TGFP I (ng/mll

Figure 4. Affinity labeling of (125I]TGF-13 receptor proteins by cross-linking with disuccinimidyl suberate (DSS). Near-confluent cell monolayers were incubated with 100 pM (125I]TGF-I3, or 100 pM ['25I]TGF-132 alone or with various concentrations of unlabeled TGF-l3h TGF-132, EGF, PDGF, or INS as indicated, and then reacted with DSS. The cells were solubilized with Triton X-100 and electrophoresed on 7.5 % sodium dodecyl sulfate gels in the presence of dithiothreitol. Three-week autoradiograms from the fixed, dried gels are shown. The positions of simultaneously electrophoresed molecular weight markers are indicated by numbers at the left. The major species labeled are denoted by arrows. (A) Affinity labeling with [125I]TGF-131 (lane 1) and competition with 2.5 to 100 ng/ml (100 to 4,000 pM) TGF-I3, (lanes 2 through 7), 100 ng/ml PDGF (lane 9), EGF (lane 10), or INS (lane 11). Solubilized extracts from cells incubated with ['251]_ TGF-131 not reacted with DSS (lane 8). (B) Affinity labeling with ['25I]TGF-132 (lanes 1 and 8), and competition with 2.5 to 100 ng/ml TGF-132 (lanes 2 through 7), 100 ng/ml PDGF (lane 9), EGF (lane 10), or INS (lane 11). (C) Affinity labeling with [' 251] TGF-13 I (lane 1), and competition with TGF-132 (lanes 2 through 6); affinity labeling with ['25I]TGF-132 (lane 7), and competition with TGF-I3, (lanes 8 through 12). The experiment was repeated 3 times, with similar results.

strated (21). However, the apparent affinity of ['251]lGF-,82 for type I and type IT receptors was 3 to 6 times lower than that of ['251]lGF-,81 (Table 2). The competition ED50 values obtained with these experiments were an order of magnitude higher than shown in Figure 2 because the earlier experiments utilized only 15 to 20 pM oflabeled lGF-,8s, whereas 100 pM of labeled lGF-,8s were used in the affinity-labeling experiments to enable visualizationof the labeled proteins by autoradiography. TABLE 2

Inhibition of affinity labeling of TGF-f3 receptors by unlabeled TGF-f31 and TGF-f32 TGF-131 [I 25I]TGF-l31 [I 251]TGF-l32

Type I Type II Type III Type I Type II Type III

(PM)

150* 130 120 1,000 1,000 240

TGF-132

(PM)

2,320 3,360 4,000 480 800 160

"TGF-13 receptors were affinity-labeled with 100 pM [1251]TGF-.B1 or [l25I]TGF-132 and reacted with disuccinimidyl suberate. The values in thetable indicate the competition ED so for each type of TGF-13 receptor when labeled in thepresence of increasing concentrations of unlabeled TGF-131 or TGF-132' These are the densitometric scanning results of the autoradiograms in Figure 4. The area under the peaks corresponding to the labeled bands was expressed as thepercentage of thearea under thepeaks for labeled bands in the absence ofany competitors. The data for type I and type II receptors were determined from 3-wk exposures, and thedata for type III receptors were from 8-d exposures of theautoradiograms.

Effect of Preincubation with TGF-131 or TGF-132 on the Number of RLF TGF-13 Receptors Although it has been shown that acute downregulation (DR) of lGF-,8 receptors does not occur in some cell lines (35), this has not been studied in early-passage RLF in culture. Incubation of RLF cultures for 16 h at 37° C with 400 pM lGF,81 resulted in a 72 % lowering of receptor number and an increase in the apparent K, (decreased affinity) when compared to control cultures assessed for ['251] lGF-,8, binding (P < 0.05) (Figure SA). This experiment was repeated 3 times with similar results (Table 1). Incubation with 400 pM lGF-,82 also resulted in a 70% loss in receptor number (P < 0.05), but with no change in the apparent K, for ['251]lGF,82 receptors (Figure 5B and Table 1). In addition, incubation with 400 pM TGF-,82 induced a 65% loss in [I25I]TGF-,8, binding capacity (not shown). This indicates that prebinding with unlabeled lGF-,82 blocked the subsequent binding of TGF-,8, to its receptors on the RLF. Although the cultures were washed extensivelyand ligands allowed to dissociate at 37° C before saturatio..n binding was performed, a significant portion of the decrease in cell-associated lGF-,8 is probably due to residual receptor occupancy and/or incomplete dissociation, as previously described (35). In another experiment, when lGF-,82-DR cultures were washed extensively and allowed to recover for 24 h rather than the 2-h period in the absence of exogenous lGF-,82, there was an increase in ['251]lGF-132 binding capacity to 64 % of control cultures (not shown).

Kalter and Brody: Rat Lung Fibroblast lGF-13 Receptors

403

Figure 5. Chronic "downregulation" of (125I]TGF-{j receptors by excess TGF-{j. (A) Near-confluB A 0.20 0.20 ent cultures were incubated with 400 pM unlabeled TGF-{jl for 0.16 0.16 16 h as described in MATERIALS •• AND METHODS. Binding isoir. therms for control (1 251] TGF-{jI ~ 0.12 0.12 ::::I bound were determined as the difo CD ference between total binding 0.08 0.08 with 1 to 200 pM (1 251] TGF-{j1 alone and nonspecific binding with (125I]TGF-{j1 and a 50-fold excess 0.04 0.04 of TGF-{jl' Data points are the means of triplicate determina0.00 .........--'-.......,;;:,""'--.......... --'o 2 tions. (A) Scatchard analysis of 4 6 8 10 12 14 16 2 4 6 8 10 12 14 16 the specific saturation binding 1251_TGF B1 Bound (fmoleSI10 6Cells) 1251_TGF B2 Bound (fmOleSl10 6 Cells) curves for control (closed circles) and downregulated cultures (open circles). (B) Chronic downregulation of (125I]TGF-{j2 receptors by excess TGF-{j2. Scatchard analysis ofthe specific saturation binding curves for control (closed circles) and downregulated cultures (open circles). Representative experiments for (125I]TGF-{j1 (n = 3) and (125I]TGF-{j2 (n = 2) are shown (see Table 1). .L......o--L.~...L_~

Biologic Responses of RLF Cultures to lGF-l3. and TGF-132 Morphology. Cells exposed to lGF-.B have altered morphologies and also have been shown to undergo reorganization of the actin cytoskeleton (25, 37). We examined the effects of incubation with lGF-.BI and lGF-132 on lowdensity RLF cultures. RLF grown in the presence of 0.5 ng/ml lGF-131 assumed a more flattened, spread-out appearance with fewer processes providing cell-to-cell contacts compared to control cultures. Similar results were obtained with lGF-132 (not shown). Incubation of RLF grown on glass coverslips with 1.0 nglml lGF-132 resulted in an apparent increase in the intensity of the specific labeling of actin microfilaments with rhodamine-phalloidin and thickening of the bundles of actin in contrast to the less intense fluorescence seen in control cells. The actin filaments in RLF incubated with lGF-131 were also brighter than the controls. These effects are not shown since they are similar to those observed in several cell lines and have been previously reported (25). CFE and inhibition of cell proliferation. Both lGF-131 and lGF-132 inhibited the formation of RLF colonies in a concentration-dependent manner (Figure 6A). The maximal decrease in colony number was 40% with 10 ng/ml lGF-I3. and 20 to 25% with 10 ng/ml TGF-132. In addition, TGF-131 clearly reduced the size of the colonies. Both TGF-.BI and TGF-.B2 significantly decreased the total number of cells per dish (Figure 6B). AT 10 nglml, TGF-131 reduced the number of cells to only 22 % of control while lGF-132 caused a reduction to ""30% of control. TGF-.BI was more potent than lGF-132 at the lower concentrations, with an ED 50 of 0.003 nglml (0.12 pM) and 0.11 ng/ml (4.4 pM) for lGF-.B2' It has been shown that lGF-13 can antagonize the stimulatory effects of other growth factors such as PDGF and EGF in monolayer cultures (38). RLF incubated with 1.0 nglml PDGF form very large colonies (not shown). Incubation of the cells to test CFE with both lGF-131 and PDGF resulted

in both a decrease in the size of the colonies and an inhibition of the fibroblast proliferative response to PDGF (Figure 7). Figure 7 illustrates the ability of TGF-131 to completely antagonize PDGF with a molar ratio of PDGF:lGF-131 of 10:1.

Discussion The existence of multiple forms of lGF-13 and their receptors appears to be well established (19,21,22,39,40). However, to our knowledge, there are no previous studies that compare the biologic potencies of lGF-131 and TGF-132 with their receptor-binding characteristics on a finite cell line derived from primary RLF in culture. A wide variety of cell types has been assayed for their (125I]TGF-131 binding patterns with affinity-labeling techniques (19, 21, 25, 26) and equilibrium binding (Scatchard analysis) (21, 23, 24), but little data are available for (125I]lGF-132 binding. In binding studies of monolayers of NRK and A549 cells, Segarini and associates (22, 25) found both cell types to contain only one-fourth to one-third as many binding sites for [125I]TGF-132 (cartilageinducing factor B [CIF-B]) as for (125I]TGF-I3I; both cell types had different K, values for TGF-131 and CIF-B, with limited cross reactivity in competition and labeling studies. In contrast, we have found that RLF contain approximately the same number of binding sites per cell for both forms of lGF-I3, but with a lower affinity for lGF-132' These RLF receptor characteristics were further demonstrated in both competition studies and downregulation experiments in which either co-incubation or preincubation with the unlabeled form of one of the lGF-l3s inhibited (125I]TGF-13 binding by the other form. In addition, affinity-labeling studies showed that the intensity of labeling of (125I]TGF-131 was more effectively inhibited by TGF-.B I than by lGF-I3z, and vice versa, with the ED50 for [I25I]lGF-I3I-labeled components in the range of a 1.5- to 4-fold molar excess for unlabeled lGF-131 and a 20- to 40-fold molar excess for unlabeled lGF-132' In addition, the ED 50 for [I25I]lGF-132-labeled components was in the range of a 1.6- to 8-fold molar excess

404

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

-..

-

'0

350000 100

c

Q en

0

0

80

~

+I

~

C

aa G)

s:

en

enCD

C 0 '0

~

-..

40

A

E

::::s

Z

0 .001

.01

.1

10

28000

.001

+I

24000

e as

~

II>

20000

.c en

16000

is ~

12000

0

8000

II>

4000

::::s

z

.01

.1

10

100

Figure 7. TGF-.81 inhibits PDGF-stimulated growth of RLF in culture. RLF were grown for 8 din 60-mm dishes (see Figure 7B). Closed circles: TGF-.81; open circles: PDGF; closed triangles: PDGF + 0.1 ng/ml TGF-.81; open triangles: PDGF + 1.0 ng/ml TGF-.81. Data points are mean ± SEM of triplicate determinations.

en

E

50000

[Growth Factor] (ng/ml)

i" LLI

.c

100000

100

[TGF B] (ng/ml)

0

150000

E

::::s

-..

a;

'0 Ci .0

20

.c

CD

200000

0

0

Z

en

is ]!

0

CD

250000

!.

60

is

300000

a

.0001

B .001

.01

.1

10

100

[TGFB] (ng/ml)

Figure 6. Effects of TGF-,81 and TGF-,82 on RLF colony formation and cell growth. The effect of TGF-,8s on the anchoragedependent growth of RLF was determined by seeding the cells at 250 per 60-mm dish and growing them in Dulbecco's modified Eagle's medium + 10% fetal bovine serum for 8 d with varying concentrations of TGF-.81 and TGF-.82. (A) Inhibition of RLF colony formation in the presence of TGF-.81 (closed circles) or TGF-.82 (open circles). Data points are mean ± SD of five determinations. (B) Inhibition of cell growth in the presence of TGF-.81 (closed circles) or TGF-.82 (open circles). Data points are mean ± SEM of triplicate determinations. Representative experiments of three repetitions are shown.

for unlabeled TGF-132 and a 2.4- to lO-fold molar excess for unlabeled TGF-131. The only exception was the nearly equal competition of TGF-131 and TGF-132 for type ITI receptors labeled by (1 2SI]TGF-132(Table 2). Our results are somewhat different from those of Cheifetz and colleagues (19), who showed that TGF-131 and TGF-132 competed equally for (12SI]TGF-132 binding in several cell lines. This discrepancy may be due to variation among different cell types (19) or may indicate the presence of two groups of receptor glyco-

proteins with similar molecular weights, with one group selective for lGF-131 and the other group selectivefor TGF132 (22). The lGF-13 receptor has been shown to be a complex combination of molecules (25-27). We noted that the 1'\.165and 1'\.185-kD receptor proteins were poorly labeled by (12SI]TGF-132, but the 1'\.1280-kD species was labeled with equal intensity by (12SI]TGF-131 and (1lSl]TGF-l3h similar to the other (12SI]TGF-132 cross-linking studies (2, 19, 22). The differences in receptor affinities for TGF-131 and TGF-132 we have observed by saturation analysis, competition, and affinity-labeling studies are in fairly good agreement with each other. The functions of these differentreceptor proteins remain largely undefined, although it is clear that the different receptor species maintain different selectivities for the various forms of TGF-13 (19, 39). While Scatchard analysis and competition experiments revealed only a 2- to 3-fold difference in binding affinities (Kd ) for TGF-I3\ and TGF-132, this represents an average difference since. the affinitylabeling competition studies show that the type III receptor recognizes lGF-131 and TGF-132 about equally well; type I has 3 times the affinity for TGF-131 as TGF-132; and type IT has 6 times the affinity for TGF-131 as TGF-132. While it is possible that there are different receptor proteins with similar molecular weights that have different affinities for TGF-131 and TGF-132 (22), the differentstructures of the two TGF-l3s may also result in differential levels of cross-linking to the receptor, thus affecting the intensity of labeling achieved. The former explanation is supported by our densitometric results obtained from affinity-labeling competition studies that show that both TGF-131 and TGF-132 compete more efficiently for their own labeled receptors than in heterologous

Kalter and Brody: Rat Lung Fibroblast TGF-/3 Receptors

competition (Table 2). On the other hand, the reduced intensity of labeling of type I and type II receptors by (125I]TGF{32 implies that there may be considerably fewer numbers of the receptor subtypes that may be capable of distinguishing the two TGF-/3 analogs. The majority oflabeled receptors on RLF are in the type III group, which has a similar affinity for TGF-/3 I and TGF-/32 (Table 2 and references 2, 19, and 22). However, the relative abundance of the three types of receptor proteins varies with cell type, as determined by affinity-cross-linking studies (2, 22, 25). Ligand-induced loss of receptors (i.e., "downregulation") has been observed for other growth factors such as EGF and PDGF (41, 42). This loss of receptors may be accompanied by decreased responsiveness to the respective ligand, followed by eventual recovery to normal receptor levels. An acute 50 to 70 % decrease in the number of TGF-/3 receptors has been observed by some investigators (20, 24) but not by others (35). In these experiments, several types of cell lines were exposed to saturating concentrations of TGF-/3 for less than 6 h at 37° C, with varying results. Our experiments suggest that a chronic, long-term exposure (16 h) to saturating levels of TGF-/3 may induce up to a 70 % loss in receptor binding capacity, which is reversible, as measured by Scatchard analysis. However, a significant portion of the decrease in cell-associated TGF-/3 could be due to incomplete dissociation and/or residual receptor occupancy despite the washing and incubation steps (35). Whether or not this effect is due to actual downregulation or receptor occupancy will have to be established by further experimentation. Also, in the biologic responses measured here and by others, the loss of 70 % of the cell surface receptors for TGF-/3 would not be expected to result in a decreased responsiveness at least to lGF-/3J, since only 10 to 20% of the receptors need be occupied for near-maximal biologic responses (Figure 7A and reference 24). In addition, because the rate at which the different receptor subtypes are downregulated is not known, we have assumed that they are all downregulated at equal rates. Interestingly, prebinding with saturating concentrations of TGF-/32 blocked the subsequent binding of labeled TGF-/31 to RLF TGF-/3 receptors. The ability of TGF-/31 and TGF-/32 to "downregulate" or occupy each other's receptors suggests that they share binding sites, rather than having separate sets of receptors. Alternatively, they may share some binding sites or classes of receptor protein species, with each ligand having, in addition, receptor species more selective for TGF-/3 I or TGF-/32 (22). It is also possible that TGF-/3 receptors exist as single sites within each group (i.e., types I, II, and III) but in altered conformations, with different conformations having higher specificity for one analog of lGF-/3. TGF-/3 J and TGF-/32 have been shown to be equipotent in several cell systems (2, 3, 19, 21,43). Recently, some functional differences in biologic responses have emerged for the two forms of TGF-/3. TGF-/31 was more effective at inhibiting DNA synthesis in endothelial cells (44) and in the inhibition of hematopoietic progenitor cell proliferation (45), while lGF-/32, but not TGF-/3h is involved in amphibian mesoderm induction (46). We have demonstrated that lGF-/31 is 10 to 30 times as potent as TGF-/32' as measured by the inhibition of the anchorage-dependent growth of RLF in cul-

405

ture, although very high concentrations of TGF-/32 inhibited cell growth in a manner similar to that of TGF-/31 (Figure 6). The ability of TGF-/3 to cause a maximum biologic response at low levels of receptor occupancy suggests the presence of "spare" receptors (24). This has also been demonstrated for other growth factors such as insulin and EGF (47, 48). Moreover, we showed that the growth stimulatory activity of the fibroblast proliferative agent PDGF was effectively antagonized by much lower concentrations of TGF-/31 (Figure 7). The antagonism of the effects of mitogenic growth factors such as PDGF, EGF, and insulin by TGF-/3 has been established (4, 23, 38,40,49). While the growth-promoting factors act via a tyrosine kinase receptor, the blocking action of TGF-/3 does not occur at these other growth factor receptors but at some distal site (49, 50). The effects of TGF-/3 on cells may be due not only to TGF-/3 itself but to its effects on activation of other growth factors or their receptors present in the cellular environment (40). The receptor binding studies shown here for TGF-/3s in the early-passage RLF have yielded results similar to those for embryonic and adult human lung fibroblastic cell lines (19, 24), where the K, varied from 13 to 40 pM and the It> from 7,000 to 19,000 sites/cell. In addition, we have shown in preliminary studies that RLF in vitro possess high-affinity receptors that specifically bind a TGF-/3-like product produced by alveolar macrophages (AM) in vitro (51). The further identification, characterization, and regulation of this AM-derived TGF-/3currently are being investigated. Preliminary results indicate that freshly isolated rat AM constitutively express a high level of mRNA for TGF-/31 (Kalter and Brody, unpublished observations) as has been observed previously in human AM (52). Activation of rat AM in vitro with carbonyl iron spheres resulted in an elevation of the amount ofTGF-/3-like activity secreted by AM (51). We have also shown that carbonyl iron spheres or chrysotile asbestos fibers induce increased levels of PDGF secretion by lung macrophages (53). Because RLF also have specific receptors that recognize macrophage-derived PDGF (28, 29, 53), the interactions of TGF-/3 and PDGF (c.f. Figure 7), and their respective roles in the pulmonary fibrogenic response should be explored and manipulated in cell culture systems, as well as in animal models of lung fibrosis (e.g., references 9-11, 13, and 54). The participation of TGF-/3 in normal and pathobiologic processes is beginning to receive considerable attention. TGF-/3 appears to be a significant contributor in wound healing (40). In addition, TGF-/3 was shown to be involved in both in vivo and in vitro models of hepatic fibrogenesis in rats (55). Recently, Khalil and colleagues (56) showed that TGF-/3 was increased in the lungs of animals treated with the fibrogenic agent bleomycin. During lung fibrogenesis, TGF-/3 was demonstrated intracellularly in macrophages and extracellularly in the collagenous matrix. The investigators postulated that TGF-/3 from macrophages mediates the bleomycin-induced lung fibrosis. We agree, at least in part, with this hypothesis but suggest that other growth factors such as PDGF also may be influential in mediating fibrotic pulmonary lesions (28, 29, 53). The role of growth factors in pulmonary fibrosis is just beginning to be explored. As referenced above (52, 56), TGF-/3 is produced by lung mac-

406

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

rophages and is increased when the cells are stimulated by inorganic particles (51). It is not known which form ofTGF-J3 predominates in macrophage secretions. If TGF-J31 is the predominant type, then the data presented here suggest that lung fibroblasts would be extremely sensitive to its biologic signal. On the other hand, if both types of lGF-J3 are secreted by AM, the differential response of RLF to the two TGF-J3s may depend upon the role of secreted TGF-J3binding proteins in the regulation of the concentration of active TGF-J3. One such potential binding protein, other than the cell-surface lGF-J3 receptors, is exrmacroglobulin, which has been found to form complexes with TGF-J3 (57). Both TGF-J31 and lGF-J32 bind preferentially to the protease-activated, or "fast," form of ex2-macroglobulin (58). In studies with (1 2SI]-labeled TGF-J31 and TGF-J32, Danielpour and Sporn (58) showed that purified CX2macroglobulin could inhibit the binding of both TGF-J3s to receptors on A549 human carcinoma cells. However, lower concentrations of purified cx2-macroglobulin were required for the inhibition of TGF-J32 binding than for TGF-J31 binding to these cells, indicating that TGF-J32 has a higher affinity than TGF-J31 for this serum binding protein. In addition, cx2-macroglobulin preferentially inhibited the inhibitory effect of TGF-J32 on the growth of mink lung cells (58). Recently, LaMarre and co-workers (59) have demonstrated that the antiproliferative effect of TGF-J3s on rat hepatocytes can be counteracted by the presence of ex2-macroglobulin or serum. The reversal of this inhibitory effect of TGF-J3 was much more pronounced with TGF-J32 than with TGF-J3 I. The investigators have hypothesized a potential role for the local production of this binding protein in the regulation of TGF-J32 activity during liver injury or inflammation (59). Because TGF-J3 has also been shown to be involved in hepatic fibrogenesis (55), differential regulation of the response to different forms of TGF-J3 may be important in the disease process. In previous studies, we have shown that ex-macroglobulin secreted by rat AM in culture is an important binding protein for AM-derived PDGF (29) and that the biologic potency of PDGF can be regulated by this ex-macroglobulin (60). Experiments are in progress to determine whether the biologic activity of AM-derived TGF-J3 may also be regulated by rat ex-macroglobulin. We intend to elucidate the role of both exogenous and AM -derived TGF-J3s and their putative binding proteins in ECM production and growth of lung fibroblasts. In addition, we must understand the responses of these lung fibroblasts to other growth factors and how they might be regulated by TGF-J3.

ADDED IN PROOF: Our results suggesting the presence of different TGF-J3 receptor subtypes that recognize TGF-J32 with higher affinity than TGF-J3 I in RLF are supported by the recent findings of Cheifetz et al. (J. Bioi. Chern. 265: 20533-20538, 1990). Using homologous competition binding assays with both labeled and unlabeled TGF-J32, they found that some cell types demonstrated subsets of TGF-J3 type I and type II receptors with higher affinity for TGF-J32 than for either TGF-J31 or the recently purified TGF-J33. Thus, cell type-specific differences in the number of high-affinity receptors for the different TGF-J3s may be associated with cell type-specific differences in the biologic responsiveness to the TGF-J3 isoforms.

Acknowledgments: We thank Ms. Annette Badgett, Ms. Lynn Moore, Dr. Alice T. Robertson, and Dr. Ronald W. Steigerwalt for valuable advice and assistance.

References 1. Delarco, J. E., and G. J. Todaro. 1978. Growth factors from murine sarcoma virus-transformed cells. Proc. Natl. Acad. Sci. USA 75:400 1-4005. 2. Cheifetz, S., A. Bassols, K. Stanley, M. Ohta, J. Greenberger, andJ. Massague, 1988. Heterodirneric transforming growth factor {3. Biological properties and interaction with three types of cell surface receptors. J. Bioi. Chem. 263:10783-10789. 3. Seyedin, S. M., P. R. Segarini, D. M. Rosen, A. Y. Thompson, H. Bentz, and J. Graycar. 1987. Cartilage-inducing factor-B is a unique protein structurally and functionally related to transforming growth factor-beta. J. Bioi. Chem. 262:1946-1949. 4. Roberts, A. B., M. A. Anzano, L. M. Wakefield, N. S. Roche, D. F. Stern, and M. B. Sporn. 1985. Type {3 transforming growth factor: a bifunctional regulator of cellular growth. Proc. Natl. Acad. Sci. USA 82: 119-123. 5. Roberts, A. B., and M. B. Sporn. 1988. Transforming growth factor {3. Adv. Cancer Res. 51:107-145. 6. Ignotz, R. A., T. Endo, and J. Massague, 1987. Regulation of fibronectin and type I collagen mRNA levels by transforming growth factor-d. J. Bioi. Chem. 262:6443-6446. 7. Rizzino, A. 1988. Transforming growth factor-S: multiple effects on cell differentiation and extracellular matrices. Dev. BioI. 130:411-422. 8. Varga, J., J. Rosenbloom, and S. A. Jimenez. 1987. Transforming growth factor {3 (TGF{3) causes a persistent increase in steady-state amounts of type I and type ill collagen and fibronectin mRNAs in normal human dermal fibroblasts. Biochem. J. 247:597-604. 9. Cutroneo, K. R., and K. M. Sterling, Jr. 1989. The biochemical and molecular bases of bleomycin-induced pulmonary fibrosis. In Focus on Pulmonary Pharmacology and Toxicology. M. A. Hollinger, editor. CRC Press, Boca Raton, FL. 1-22. 10. Snider, G. L. 1986. Interstitial pulmonary fibrosis. Chest 89: 115S-121S. 11. Chang, L.-Y., L. H. Overby, A. R. Brody, and J. D. Crapo. 1988. Progressive lung cell reactions and extracellular matrix production after a brief exposure to asbestos. Am. J. Pathol. 131: 156-170. 12. Goldstein, R. H., and A. Fine. 1986. Fibrotic reactions in the lung: the activation of the lung fibroblast. Exp. Lung Res. 11:245-261. 13. Reiser, K. M., and J. A. Last. 1986. Early cellular events in pulmonary fibrosis. Exp. Lung Res. 10:331-335. 14. Hill, D. J., S. F. Elstrow, I. Swenne, and R. D. G. Milner. 1986. Bifunctional action of transforming growth factor-S on DNA synthesis in early passage human fetal fibroblasts. J. Cell. Physiol. 128:322-328. 15. Sorrentino, B., and S. Bandyopadhyay. 1989. TGF{3 inhibits GoIS-phase transition in primary fibroblasts. Loss of response to the antigrowth effect of TGF{3 is observed after immortalization. Oncogene 4:569-574. 16. Wrana, J. L., J. Sodek, R. L. Ber, and C. G. Bellows. 1986. The effects of platelet-derived transforming growth factor {3 on normal human diploid gingival fibroblasts. Eur. J. Biochem. 159:69-76. 17. Fine, A., and R. H. Goldstein. 1987. The effect of transforming growth factors on cell proliferation and collagen formation by lung fibroblasts. J. Bioi. Chem. 262:3897-3902. 18. Raghu, G., S. Masta, D. Meyers, and A. S. Narayanan. 1989. Collagen synthesis by normal and fibrotic human lung fibroblasts and the effect of transforming growth factor-d. Am. Rev. Respir. Dis. 140:95-100. 19. Cheifetz, S., J. A. Weatherbee, M. L.-S. Tsang, et al. 1987. The transforming growth factor {3 system, a complex pattern of cross-reactive ligands and receptors. Cell 48:409-415. 20. Frolik, C. A., L. M. Wakefield, D. M. Smith, and M. B. Sporn. 1984. Characterization of a membrane receptor for transforming growth factor{3 in normal rat kidney fibroblasts. J. Bioi. Chem. 259:10995-11000. 21. Massague, J. 1987. Identification of receptors for type-S transforming growth factor. Methods Enzymol. 146:174-195. 22. Segarini, P. R., A. B. Roberts, S. M. Rosen, and S. M. Seyedin. 1987. Membrane binding characteristics of two forms of transforming growth factor-S. J. Bioi. Chem. 262: 14655-14662. 23. Tucker, R. F., E. L. Branum, G. D. Shipley, R. J. Ryan, andH. L. Moses. 1984. Specific binding to cultured cells of 12sI-labeled type {3 transforming growth factor from human platelets. Proc. Natl. Acad. Sci. USA 81: 6757-6761. 24. Wakefield, L. M., D. M. Smith, T. Masui, C. C. Harris, andM. B. Sporn. 1987. Distribution and modulation of the cellular receptor for transforming growth factor-beta. J. Cell Bioi. 105:965-975. 25. Segarini, P. R., D. M. Rosen, and S. M. Seyedin. 1989. Binding of transforming growth factor {3 to cell surface proteins varies with cell type. Mol. Endocrinol. 3:261-272. 26. Cheifetz, S., B. Like, and J. Massague. 1986. Cellular distribution of type I and type II receptors for transforming growth factor-S. J. Bioi. Chem. 261 :9972-9978. 27. Cheifetz, S., J. L. Andres, and J. Massague. 1988. The transforming

Kalter and Brody: Rat Lung Fibroblast

28. 29.

30. 31. 32. 33. 34. 35. 36. 37. 38.

39. 40. 41. 42.

43. 44.

45.

roF-~

407

Receptors

growth factor-S receptor type III is a membrane proteoglycan. Domain structure of the receptor. J. Bioi. Chern. 263: 16984-16991. Kumar, R. K., R. A. Bennett, and A. R. Brody. 1988. A homologue of platelet-derived growth factor produced by rat alveolar macrophages. FASEB J. 2:2272-2277. Bonner, J. C., M. Hoffman, andA. R. Brody. 1989. Alpha-macroglobulin secreted by alveolar macrophages serves as a binding protein for a macrophage-derived homologue of platelet-derived growth factor. Arn. J. Respir. Cell Mol. Bioi. 1:171-179. Scatchard, G. 1949. The attractions of proteins for small molecules and ions. Ann. NY Acad. Sci. 51:600-672. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. Munson, P. J., and D. Rodbard. 1980. LIGAND: a versatile computerized approach for characterization of ligand-binding systems. Anal. Biochem. 107:220-239. Wilkinson, L. 1987. SYSTAT: The System for Statistics. SYSTAT, Inc., Evanston, IL. Massague, J., and B. Kelly. 1986. Internalization oftransforming growth factor-S and its receptor in BALB/C 3T3 fibroblasts. J. Cell. Physiol. 128:216-222. Massague, J., and B. Like. 1985. Cellular receptors for type {3 transforming growth factor. Ligand binding and affinity labeling in human and rodent cell lines. J. Bioi. Chern. 260:2636-2645. Klotz, I. M. 1982. Numbers of receptor sites from Scatchard graphs: facts and fantasies. Science 217: 1247-1249. Fanger, B. 0., L. M. Wakefield, and M. B. Sporn. 1986. Structure and properties of the cellular receptor for transforming growth factor type {3. Biochemistry 25:3083-3091. Anzano, M. A., A. B. Roberts, and M. B. Sporn. 1986. Anchorageindependent growth of primary rat embryo cells is induced by plateletderived growth factor and inhibited by type-beta transforming growth factor. J. Cell. Physiol. 126:312-318. Massague, J., S. Cheifetz, R. A. Ignotz, and F. T. Boyd. 1987. Multiple type-d transforming growth factors and their receptors. J. Cell. Physiol. Suppl. 5:43-47. Sporn, M. B., A. B. Roberts, L. M. Wakefield, and B. de Crombrugghe. 1987. Some recent advances in the chemistry and biology of transforming growth factor-beta. J. Cell BioI. 105: 1039-1045. Bowen-Pope, D. F., and R. Ross. 1982. Platelet-derived growth factor. II. Specific binding to cultured cells. J. Biol. Chern. 257:5161-5171. Heldin, C.-H., A. Wasteson, and B. Westermark. 1982. Interaction of platelet-derived growth factor with its fibroblast receptor: demonstration of ligand degradation and receptor modulation. J. Biol. Chern. 257: 4216-4221. Ignotz, R, and J. Massague. 1986. Transforming growth factor-beta stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J. Bioi. Chern. 261:4337-4345. Jennings, J. C., S. Mohan, T. A. Linkhart, R. Widstrom, and D. J. Baylink. 1988. Comparison of the biological actions ofTGF beta-l and TGF beta-2: differential activity in endothelial cells. J. Cell. Physiol. 137: 167-172. Ohta, M., J. S. Greenberger, P. Anklesaria, A. Bassols, and J. Massague,

46. 47. 48. 49. 50.

51. 52. 53.

54.

55. 56. 57. 58. 59.

60.

1987. Two forms of transforming growth factor-S distinguished by multipotential haematopoietic progenitor cells. Nature 329:539-541. Rosa, F., A. B. Roberts, D. Danielpour, L. L. Dart, M. B. Sporn, and I. B. Dawid. 1988. Mesoderm induction in amphibians: the role of TGF.82-like factors. Science 239:783-785. Aharonov, A., R M. Pruss, and H. R. Herschman, 1978. Epidermal growth factor. Relationship between receptor regulation and mitogenesis in 3T3 cells. J. Bioi. Chern. 253:3970-3977. Kono, T., and F. W. Barham. 1971. The relationship between the insulinbinding capacity of fat cells and the cellular response to insulin. J. Bioi. Chern. 246:6210-6216. Like, B., and J. Massague. 1986. The antiproliferative effect of type beta transforming growth factor occurs at a level distal from receptors for growth-activating factors. J. Bioi. Chern. 261:13426-13429. Jaffer, F., C. Saunders, P. Shultz, D. Throckmorton, E. Weinshell, and H. E. Abond. 1989. Regulation of mesangial cell growth by polypeptide mitogens. Inhibitory role of transforming growth factor beta. Arn. J. Pathol. 135:261-269. Kalter, V. G., and A. R. Brody. 1989. Secretion of TGF{3 by rat alveolar macrophages and characterization of receptors for TGF{3 on rat lung fibroblasts. Cytokine 1:76. (Abstr. no. 24) Assoian, R. K., B. E. Fleurdelys, H. C. Stevenson et al. 1987. Expression and secretion of type {3 transforming growth factor by activated human macrophages. Proc. Nati. Acad. Sci. USA 84:6020-6024. Bauman, M. D., A. M. Jetten, J. C. Bonner, R. K. Kumar, R. A. Bennett, and A. R. Brody. 1990. Secretion of a platelet-derived growth factor homologue by rat alveolar macrophages exposed to particulates in vitro. Eur. J. Cell BioI. 51:327-334. Warheit, D. 8., L.-Y. Chang, L. H. Hill, G. E. R. Hook, J. D. Crapo, and A. R.' Brody. 1986. Pulmonary macrophage accumulation and asbestos-induced lesions at the site of fiber deposition. Arn. Rev. Respir. Dis. 134: 128-133. Czaja, M. J., F. R. Weiner, K. C. Flanders et al. 1989. In vitro and in vivo association of transforming growth factor {3 with hepatic fibrosis. J. Cell Bioi. 108:2477-2482. Khalil, N., O. Bereznay, M. Sporn, and A. H. Greenberg. 1989. Macrophage production of transforming growth factors and fibroblast collagen synthesis in chronic pulmonary inflammation. J. Exp. Med. 170:727-737. Huang, S. S., P. O'Grady, and J. S. Huang. 1988. Human transforming growth factor {3-cxTmacroglobulin complex is a latent form of transforming growth factor {3. J. Bioi. Chern. 263:1525-1541. Danielpour, D., and M. B. Sporn. 1990. Differential inhibition of transforming growth factor .81 and .82 activity by cx2-macroglobulin. J. Biol. Chern. 265:6973-6977. LaMarre, J., G. K. Wollenberg, J. Gauldie, and M. A. Hayes. 1990. CX2Macroglobulin and serum preferentially counteract the mitoinhibitory effect of transforming growth factor-{32 in rat hepatocytes. Lab. Invest. 62:545-551. Bonner, J. C., A. Badgett, A. R. Osornio-Vargas, M. Hoffman, andA. R Brody. 1990. PDGF-stimulated fibroblast proliferation is enhanced synergistically by receptor-recognized cxrmacroglobulin. J. Cell. Physiol. 145:1-8.