JOURNAL OF CELLULAR PHYSIOLOGY 152507-519 (1992)

Transforming Growth Factor-pl Inhibitory Effect of Platelet-Derived Growth Factor-Induced Signal Transduction on Human Bone Marrow Fibroblasts: Possible Involvement of Protein Phosphatases MICHAELA FONTENAY, MARIJKE BRYCKAERT, AN[) GERARD TOEELEM* lnserm U 348, Hbpita/ Larihoisi+re, 7.50 70 f'aris, France

Transforming growth factor-pl (TGF-p1) is a potent growth inhibitor for many cell types. On fibroblasts, TGF-PI has been shown to inhibit human platelet-derived growth factor ( P U G - i n d u c e d mitogenicity. The mechanism implicated in this growth inhibition is unknown. In this work, we show on human bone marrow fibroblasts that TGF-P1, which inhibited PDGF-BB mitogenicity, was able to block PUGF-BB-induced early events such as polyphosphoinositide (Ptdlns 4,5P, Ptdlns 4-P, and Ptdlns) breakdown and Ins 1,4,5-P, formation. No significant modification by TGF-p1 of PDCF-BB binding (n, = 200,000 vs. nz = 135,000 sites per cell with TGF-PI; Kd, = Kd, = 0.5 x 10 'M) and of internalization kinetic5 was observed. In addition, TGF-P1 was shown to inhibit PDGF-BB receptor autophosphorylation either in intact cells or in partially isolated membranes and to partially inhibit PDGF-K tyrosine kinase activity. Since a dephosphorylation mechanism through protcin phosphatases could be implicated, we used okadaic acid, a potent inhibitor of type 1 and 2A serineithreonine phosphatases and showed that okadaic acid restored PDCF-receptor autophosphorylation on tyrosine residues. Based on these data, we suggest that an alternative regulatory mechanism of PDGF tyrosine phosphorylation seems to involve serineithreonine c 1992 WiIey- L i s i . Inc. phosphatase activation.

Transforming growth factor-pl (TGF-p1) is a n homodimeric 25 kD polypeptide secreted by most mammalian cells in a latent form (reviewed in Sporn et al., 1986; Massague, 1990). Five types of TGF-P receptor have been identified and two ubiquitous types (type I and 11) seem to be involved in mediating TGF-P biological activities (Cheifetz e t al., 1987, 1990; Laiho et al., 1990aj. TGF-p1 is described as a biphasic multifunctional factor that displays a reversible growth inhibitory activity in fibroblasts, epithelial, and endothelial cells in vitro (Moses et al., 1985; Roberts et al., 19851,but it has a proliferative effect for many mesenchymal cells (Sporn e t al., 1987). TGF-pl was shown to exhibit a mitogenic activity on human fibroblasts through the induction of platelet-derived growth factor (PDGF) receptor a subunits (Battegay e t al., 1990; Ishikawa et al., 1990). Moreover, TGF-p1 can modulate the cellular response to other growth factors, either stimulating or inhibiting their mitogenicity. TGF-p1 has been found to inhibit EGF mitogenicity, although a n increase of the number of epidermal growth factor (EGF)-receptors (Assoian et al., 1984; Assoian, 1985; Baskin et al., 1991) has been observed in response to TGF-P1, which also enhanced the effect of EGF on polyphosphoinositide and Ca+' mobilisation (Muldoon e t al., 1988a). In addition, TGF-p1 inhibited platelet-derived growth factor ", 1992

WILEY-LISS, INC.

(PDGF)-induced human bone marrow fibroblast proliferation (Bryckaert et al., 1988).Moreover, high concentrations of TGF-Pl decreased the mitogenic response to PDGF-AA but not to PDGF-AB or -BB on NIH 3T3 fibroblasts (Gronwald et al., 1989). On human fibroblasts, the three isoforms (AA, AB, and BB) of PDGF were shown to bind two (Y and p types of receptor (Hart e t al., 1988; Heldin et al., 1988). The a type receptor is able to bind the three isoforms: whereas the p type only binds PDGF-BB with high affinity. Specific ligand binding activates the protein tyrosine kinase of both receptors leading to receptor autophosphorylation (Ek et al., 1982; Frackelton et al., 1984; Escobedo et al., 1988; Claesson-Welsch e t al., 1989; Hammacher et al., 1989). Signal transduction mediated by PDGF-R involves protein kinase C pathway through the modification of polyphosphoinositide metabolism (Habenicht et al., 1981). TGF-pl-induced transduction mechanism is presently unknown. TGF-P1 inhibits proliferation of hamster lung fibroblasts but does not alter polyphosphoinositide breakdown, activation of protein kinase C

Received October 7, 1991; accepted March 4,1992.

"To whom reprint requestsicorrespondence should be addressed.

508

FONTENAY ET AL

(PKC), elevation of intracellular pH (Chambard and Pouyssegur, 1988), or intracellular calcium mobilization (Muldoon et al., 1988b3. Moreover, several studies have shown that TGF-P1 rapidly inhibits c-myc transcription (Braun et al., 1990; Pietenpol et al., 1990) and suppresses phosphorylation of the retinoblastoma gene product (Laiho et al., 1990b). The role of serine/ threonine phosphatase activation in the inhibition of keratinocyte growth by TGF-P1 has recently been suggested (Gruppuso et al., 1991). In the present study, we have shown on human bone marrow fibroblasts that (I)the inhibition by TGF-P1 of PDGF mitogenicity was related to a block of PDGFinduced polyphosphoinositide turnover and of PDGFreceptor autophosphorylation and tyrosine kinase activity, (2) and the inhibitory effect of TGF-P1 on PDGF-R autophosphorylation was reversed by okadaic acid, a potent inhibitor of type 1 and 2A protein phosphatases.

MATERIALS AND METHODS Materials Recombinant PDGF-BB (c-sis) was purchased from Amersham Laboratories (Buckinghamshire, UK) and recombinant PDGF-AA from Upstate Biotechnology, Inc. (Lake Placid, NY). Recombinant human TGF-P1 and PDGF-AB, purified over 95% from human platelets, were from British Bio-Technology (Oxford, UK). PR 7212 anti-PDGF-R type p antibody was purchased from Genzyme Corporation (Boston, MA) and 4G10 antiphosphotyrosine antibody was from Upstate Biotechnology. Okadaic acid was from Scientific Marketing Associates (Barnet, UK).

Isolation and culture of h u m a n bone m a r r o w fibroblasts Normal human bone marrow samples were provided by hip surgery performed on patients without bone marrow pathology. As previously described (CastroMalaspina et al., 19811, fibroblasts were isolated and cultured in a-modification of Eagle's medium (a-MEM) supplemented with 20% fetal calf serum (FCS), 2 mM glutamine, 50 Uiml penicillin, 50 pg/ml streptomycin, and 2.5 pgiml amphotericin B (Labsystems, Les Ulis,

Abbreviations

a-MEM

a-modification of Eagle's medium bovine serum albumin FCS fetal calf serum Ins 1,4,6-P, inositol 1,4,6-trisphosphate OA okadaic acid PA phosphatidic acid PAGE polyacrylamide gel electrophoresis PBS phosphate-buffered saline PDGF platelet-derived growth factor PDGF-R platelet-derived growth factor receptor PPase serine/threonine phosphatase PtdIns phosphatidylinositol PtdIns 4-P phosphatidylinositol 4-phosphate PtdIns 45-P, phosphatidylinositol 4,5-bisphosphate PTPase phosphotyrosine phosphatase SDS sodium dodecyl sulfate TGF-Pl transforming growth factor-pl BSA

France). Only first and second passages were used for experiments. Fibroblasts were characterized by immunofluorescent staining and electron microscopy, a s previously described (Bryckaert et al., 1988). Thymidine incorporation a s s a y The mitogenic activity of PDGF-AA, -AB, and -BB was assayed by incorporation of 1methyl-3H]thymidine (r3H1TdR).Cells were seeded in 12-well plates (Costar, Cambridge, MA) in a-MEM + 20% FCS. Three days later, subconfluent cells were placed for 48 h under serum-deprived conditions in Ham's F12 medium (JBio, Les Ulis, France) + 0.5%bovine serum albumin (BSA, Sigma, St Louis). Increasing concentrations of PDGF-AA, -AB, or -BB (1-50 ngiml) and ['HITdR (1 pCilwell) were then added simultaneously in the presence or in the absence of TGF-Pl(O.1, 1.0, or 10 ng/ml). Cells were rinsed twice in 20 mM phosphate-buffered saline (PBS) (pH 7.40) containing 0.9 mM CaC1, 2H,O and 0.5 mM MgCls 6H,O. After precipitation with 10% ice-cold trichloracetic acid, cells were washed once in phosphate buffer (pH 7.40) and solubilized at room temperature for 30 min with 0.2 M NaOH. The radioactive lysates were counted in a p-matic Kontron counter. Polyphosphoinositide metabolism Bone marrow fibroblasts were plated into 6-well dishes at lo5 cellsiwell in a-MEM t 20% FCS. Two days later, cells were placed for 36 h in serum-free Ham's F12 medium containing 0.5% BSA and then incubated overnight with ["P]orthophosphate (50 pCi/well) in phosphate free-DMEM. Cells were washed three times with 20 mM Hepes buffer + 0.1% BSA, pH 7.40. PDGFAA, -AB or -BB (10 ngiml) in the presence or absence of 10 ng/ml of TGF-p1 was added to a final volume of 1ml. The reaction was stopped at different times by rapid removal of supernatants and by addition of 1 ml of ice-cold CH,OH. Cells were scraped off the dishes and transferred into glass vials. Phospholipids were extracted with acidified solvants as previously described (Chu et al., 1986) and then separated by thin layer chromatography on silica plates (Merck, Darmstadt, Deutschland) with the following migration solvent: CHC1,ICH3C0CH3/CH30H/CH3C00H/H20 (40:15:13: 12:7, v/v/v/vlv). 132Plphosphatidylinositol 4,5-bisphosphate (r3'Pl PtdIns 4,5-P2), ~32Plphosphatidylinositol 4-monophosphate (["PIPtdIns 4-P), [32P]phosphatidylinositol ([32P]PtdIns), and ["P]phosphatidic acid ([""PJPA) were located by autoradiography and quantified by scraping the spots and measurement in a fi counter. Determination of Inositol 1,4,54risphosphate (Ins 1,4,5-P,) formation Fibroblasts were cultured in 6-well dishes with a-MEM T 20%,FCS until a density of 2 x lo5 cellsiml was obtained and them incubated in a serum-free medium for 48 h at 37°C. Next, cells were placed in serumfree medium containing 20 mM LiCl and then stimulated with 10 ngiml PDGF-AA, -AB, and -BB or the equivalent volume of medium in the absence or presence of TGF-Pl (10 ngiml) for 2 min at 37°C. In some experiments, fibroblasts were pretreated with 10 nM

TCFp INHIBITS PDGF-R TRANSDUCTION VIA PHOSPHATASES

okadaic acid for 30 min at 37°C and then incubated with growth factors in fresh medium, At the end of incubation, cells were placed on ice, supernatants were discarded, and ice-cold 20% perchloric acid was added for 20 min. Cells were then scraped off the dishes and transferred into vials and the resulting precipitates were sedimented by centrifugation for 15 min at 2,000 g and 4°C. Supernatants were titrated at pH 7.50 with 10 N KOH and then with 1.5 M KOH + 60 mM Hepes. KC10, precipitates were eliminated by centrifugation and supernatants precleared by anion exchange chromatography to eliminate most of other inositolphosphate isoforms (Amprep, Amersham, Buckinghamshire, UK). Ins 1,4,5-P3 was measured using the Ins 1,4,5-P3 radioimmunological assay kit (Amershamj by competition with D-myo L‘Hlinositol 1,4,5-trisphosphate tracer. Radioactivity was measured in a p counter and results were expressed in % relative to the control without growth factor.

509

Lysates were then scraped, collected for sonication, and centrifuged for 10 min at 4°C and 800 g. Supernatants were ultracentrifuged at 100,000 g for 1 h a t 4°C. Pellets were resuspended in 50-100 pl lysis buffer and Bradford method (1976) was used for protein quantitation (IgG1 used as standard, Sigma). The autophosphorylation assay was performed as following: equal amounts of membrane were incubated a t either 0°C or 37°C with 10 ng PDGF-BB in the presence or absence of TGF-P1 (10 ng) under a total volume of 50 p1 reaction buffer (20 mM Hepes pH 7.40,60 mM NaCl,O.l% BSA, 3 mM MnCl,, and 0.05%Triton x 100).After incubation with growth factors, 10 pM Ty 32PJATP (1OpCi) was added for either 10 rnin a t 0°C or 5 rnin a t 37°C. Samples were analyzed on 7% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Immunoprecipitation Fibroblasts, in 10 cm dishes, were preincubated for PDGF binding assay and internalization 10 min at 37°C with 25 ng TGF-P1 and then for 5 min PDGF-AA was iodinated using a modified chlora- with 50 ng of PDGF-BB in a total volume of 1.5 ml mine T method (Hunter and Greenwood, 1962) with reaction buffer (50 mM Hepes pH 7.40, 145 mM NaCl, specific activity of 150,000 cpming. [ ‘‘‘IIPDGF-BB was 5.4 mM KC1,l.S mM CaCl,, 2 mM MnCl,, 20 pM ZnCl,, obtained from Amersham Laboratories and its specific 1 mM Na,VO,). Cells were washed twice with ice-cold activity was 75,000 cpming. PBS pH 7.40 and then lysed with 1ml of lysis buffer (20 Binding of [lZ5I]PDGF-AAand -BB to human bone mM Hepes pH 7.40,125 mM NaCl, 1% Triton x100,20 marrow fibroblasts was measured as previously de- pM ZnCl,, 50 mM NaF, 10 mM pyrophosphate, 1.5 mM scribed (Heldin et al., 1981). Confluent cells in 12-well Na,VO,, 10 mM EDTA, 2 mM EGTA plus a mixture of plates were rinsed once with a binding buffer (20 mM protease inhibitors (1mM PMSF, 10 pgiml Aprotinin, 5 PBS containing 0.9 mM CaC1, 2H,O, 0.5 mM MgC1, pgiml Leupeptin, 20 pgiml Benzamidine, 10 pgiml soy6H,O and 0.1% BSA, pH 7.40). Binding of [12511PDGF- bean trypsin inhibitor). After 30 rnin of incubation on AA and -BB was determined by incubating cell mono- ice, cell extracts were centrifuged a t 11,000 g for 15 layers for 3 h at 4°C in binding buffer with 1 ng of min. Aliquots were taken for protein quantitation. In radiolabeled ligands and increasing unlabeled ligand some experiments before adding growth factors, cells concentrations. Fibroblasts were then washed five were pretreated with 100 nM okadaic acid for 30 rnin a t times with 0.5 ml ice-cold binding buffer and cell-asso- 37°C. Five plipoint of antiphosphotyrosine 4G10 MoAb ciated radioactivity was determined in a gamma (Upstate Biotechnology) were incubated with 60 pl of a counter. Nonspecific binding was evaluated by addition 50%(viv) Protein A-Sepharose suspension for 1h a t 4°C of a 200-fold excess of unlabeled PDGF. In some experi- on rotator. The preparation was washed 4 times in ments, cells were preincubated with 10 ngiml of wash buffer (20 mM Hepes pH 7.40, 125 mM NaCI, 0.1% Triton x 100,20 p M ZnCl,, 1rnM Na,VO,, 10 mM TGF-P1 for 10 minutes a t 37°C. PDGF receptor internalization was studied a s follow- EDTA, 2 mM EGTA, 50 mM NaF, 10 mM pyrophosing: TlZ5I1PDGF-BB(100,000 cpm per well) was incu- phate) and a 50% (viv) antibody-Proteine A-sepharose bated for 3 h a t 4°C in the presence or absence of 10 suspension was made. For immunoprecipitation reacngiml of TGF-p1 in 0.5 ml binding buffer, then cells tions, 50 pl of this suspension was added to equal were washed and incubated a t 37°C to allow internal- amounts of protein for a 3-h incubation at 4°C on rotaization, in the presence or absence of TGF-Pl(10 ngiml) tor. added in fresh medium. At different times (0,15,30,60, Alternatively, cells were lysed in RIPA buffer (Tris and 120 rnin), surface bound L1251]PDGF was deter- 50 mM pH 8.0, NaCll25 mM, 0.5%NP-40,0,5% sodium mined by treating cells with ice-cold 0.5 M acetic acid deoxycholate and 0.5 mgiml BSA globulin-free). Four and 150 mM NaCl for 5 rnin (Betsholtz et al., 1984). pl of anti-PDGF-R type B antibody PR7212 (Hart et al., Internalized [1251]PDGFwas measured in cell lysates 1987)were added to each lysate for 3 h a t 4°C on rotator following 30 min at 37°C. The cells were lysed in 20 mM and then a 50% suspension of protein A-sepharose couHepes, 1%Triton x 100,10%glycerol (viv) pH 7.40 with pled to a rabbit antimouse antibody (IgG 7S, Nordic) for 0.1 mgiml of BSA. 1h at 4°C. Samples were washed 4 times in wash buffer and analyzed by 7% SDS-PAGE. Proteins were transMembrane preparation and ferred with a transblot apparatus (Hoefer Scientific Inautophosphorylation assay struments, San Francisco). For immunoblotting, nitroFor membrane preparation, cells were washed twice cellulose membranes were saturated in 10 mM Tris pH in 17 mM Hepes pH 7.0, 160 mM KC1, 50 pM EGTA, 7.40, 0.17 M NaC1, 5% dried milk and incubated with and lysed in 10 mM Hepes pH 7.0, 10 mM KCI, 50 II.M 4G10 antiphosphotyrosine antibody (2 Fgiml). After EGTA, 0.5 mM Dithiotreitol, 10 pg/ml Leupeptin, 50 several washes, phosphoproteins were revealed using pgiml Aprotinin, 100 pgiml soybean trypsin inhibitor. 1”51-proteinA (0.1 pCiiml) (Amersham).

510

FONTENAY ET AL

In another set of experiments, quiescent fibroblasts in 10 cm dishes were incubated for 1 h in phosphatefree DMEM and then prelabeled with 0.2 mCi/ml [32P]orthophosphatefor 4 h at 37°C. After growth factor addition, in the absence or presence of 100 nM OA, lysates prepared as previously were immunoprecipitated with PR7212 antibody. Samples were washed 3 times in lysis buffer and then twice in wash buffer and analyzed on 7% SDS-PAGE. All the experiments were exposed for autoradiography with Kodak X-OMAT AR films.

1000

4

800

-

600

-

400

-

200

-

Statistical analysis The statistical analysis performed on 13H]TdRincorporation, polyphosphoinositide turnover, and Ins 1,4,5-P3formation results was the ANOVA test for repeated measures.

o! 0

'

I

.

I

.

20

10

I

I

.

40

30

41

.

I

50

PDGF concentration (nghl) i

RESULTS Inhibition of PDGF-induced mitogenic activity by TGF-PI Human bone marrow fibroblasts were first tested for their response to the different isoforms (AA, AB, BB) of PDGF in the presence or absence of different concentrations of TGF-Pl. [Methyl-3H1thymidine uptake increased in a dose-dependent manner with PDGF (0 to 50 ngiml). The stimulation reached a plateau a t 830% t 87 and 436%)+- 58 of the control value with 15 ngiml of PDGF-BB and -AB, respectively. PDGF-AA induced no significant mitogenic activity. Increasing concentrations of TGF-Pl(O.1to 10 ngiml) added simultaneously decreased PDGF-BB and PDGF-AB (10 ngiml) -induced DNA synthesis (Fig. 1B). A significant inhibition of PDGF induced-mitogenic activity from 794% * 106 to 456% 2 66 of the control value for PDGF-BB (p < 0.05) and from 447% i72 to 295% ? 58 of the control value for PDGF-AB (p < 0.05) was observed with 10 ngiml TGF-P1. PDGF-AA 110 ng/ml) stimulation, which varied according to the cell passage used (125%--234%)was slightly reduced by 1 ngiml of TGF-PI.

BB

aoo 3

El

POGFAA POGFAB

0 0 I-

0

z c P

iL

600-

a

BE 8 z

a

U

: z

2ooL 400-

1.o 0.1 TGF-Di (ng/rnl)

0

10

Fig. 1. Effcct of TGF-pl on PDGF-AA, -AB, and -BB mitogenicity. Quiescent cells were incubated for 48 h a t 37°C (A) with increasing concentrations (1to 50 ngimlJ of PDGF-AA 1 A,"'.A 1, -AR (D----c)) or -BB (.--a) or (B) with 10 ngiml of each isoform in the absence or presence of TGF-pl (0.1-10 ngjml) and 1 (*Ciiml 1"HJthymidinewas added simultaneously. Results are expressed relativc to a control without growth factor a s 100% and are means SEM of five separate expcriments in duplicates. "'P .: 0.05.

*

Polyphosphoinositide turnover Ins 1,4,5-P, formation We have hypothesized that the inhibitory effect of TGF-P1 should be related to a n inhibition of PDGFIn order to confirm these results, Ins 1,4,5-P, formainduced early events. First, changes in the amount of tion was quantified using a radioimmunoassay (Amer[32P1-prelabeled phospholipids were quantified at different times after PDGF-AA or -BB addition. With 10 ngiml of PDGF-BB, a rapid and transient decrease of Fig. 2. Effects of PDGF-AA and -BB on polyphosphoinositide metab["PIPtdIns 4,5-P2,["PIPtdIns 4-P, and ["PIPtdIns lev- olism. Cells were plated a t a density of 2.5 x lo5 cel1s:ml for 24 h and els was observed and reached 82%- * 4, 88% -t 5 and then serum-depleted for 48 h. Fibroblasts were prelabeled overnight 50 pCi:ml I"'PP1orthophosphatc in phosphate-free DMEM. At dif89% % 5 of the initial control value, respectively, after 1 with ferent times; rJ2P1labeled fihroblasts were stimulated with 10 nginil min. In addition, PA synthesis (116% t 5 of the control PDGF-AA ( A - A ) or -BB (*-*I a t 37°C. Results a r e expressed as the value) was observed after 30 min of stimulation. In ratio of the percentage of radioactivity of a given phospholipid in the contrast, no modification of polyphosphoinositide me- presence of growth factor to its percentage of radioactivity i n unstimulatcd fibroblasts. Results are means SEM of five experiments. tabolism was detectable in the presence of 10 ngiml of Fig. 3. Effects of TGF-Bl on PDGF-BE-induced polyphosphoinositidc PDGF-AA (Fig. 2). Added in the presence of 10 ngiml breakdown. Fibroblasts were labeled as prcviously described. At the PDGF-BB, TGF-Pl (10 ng/ml) was found to inhibit times indicated, cells were stimulated with 10 ngiml PDGF-RB in the absence (0-0) or presence (O----n) of 10 ng/ml TGF-P1 a t 37°C. RePDGF-BB-induced phosphoinositide breakdown: ["PI sults are expressed as the ratio of the percentage of radioactivity of a PtdIns 43-P, (94.5 ? 5.0; p < 0.051, ["PIPtdIns 4-P given phospholipid in the presence of growth factor to its percentage of (103.5 ? 2.3; p < 0.05), ["PIPtdIns (101.3 i 3.3; p < radioactivity i n unstimulated fibroblasts. Results are means :SEM 0.05) (Fig. 3). of five experiments. +

51 1

TGFB INHIBITS PDGF-R TRANSDUCTION VIA PHOSPHATASES

j

PIP2

c

8 70

I

0

20

10

30

-

l110

z

0

8oL 70 0

0

,

1

I

.

10

20

.

20

10

30

130

PA

0

' " I

m

' 0

30

20

10

30

Time (min) Fig. 2

-

N

80

Y

-

8

70 I

0

10

'

1

1

20

30

.

70

I

I

0

10

I

.

20

30

0

10

20

30

L CI

c

. a 0 0

n

105

n I

n N

n

95

Y

0

10

20

30

Time (min) Fig. 3.

512

FONTENAY ET AL

o

l

.

"0

I I : I-

, 20

0

.

,

*

I

.

,

40 60 PDGF (nglml)

.

80

I

I

100

400. McoMAol

TGF-Oi

+TGF-R

-

+

-

+

-

+

-

+

Fig. 4. Dose-response study of Ins 1.4,5-P, formation. (A) Quiescent cells were stimulated for 2 minutes with increasing concentrations of PDGF-AA ( A . . . . . A1, -AB (o---o)or -BB ( 0 -mi at 37°C or (B) with each isoform of PDGF (10 ng:ml) without or with 10 ngiml of TGY-pl a t 37°C. Ins 1,4,5-P, formation was evaluated using an Ins 1,4,5-P, radioimmunological assay kit. Results as percentage of the control without growth factor, are means i- SEM of three separate experiments. :?*I)

Transforming growth factor-beta 1 inhibitory effect of platelet-derived growth factor-induced signal transduction on human bone marrow fibroblasts: possible involvement of protein phosphatases.

Transforming growth factor-beta 1 (TGF-beta 1) is a potent growth inhibitor for many cell types. On fibroblasts, TGF-beta 1 has been shown to inhibit ...
1MB Sizes 0 Downloads 0 Views