JOURNAL OF CELLULAR PHYSIOLOGY 144:473-484 (1990)

Transforming Growth Factor Type p1 Modulates the Effects of Basic Fibroblast Growth Factor on Growth and Phenotypic Expression of Rat Astroblasts In Vitro GERARD LABOURDETTE,* THIERRY JANET, PASCAL LAENG, FREDERIC PERRAUD, D A V I D LAWRENCE, AND BRIGITTE PETTMANN Centre de Neurochimie du Centre National de la Recherche Scientifique and lnstitut National de /a 5ant6 et de la Recherche Medicale U44, 67084 Strasbourg Cedex (C.L, JJ., P. L., F.P., B.P.), and In5titut Curie-Biologie, Centre Universitaire, 9 1405 Orsay Cedex, France (D.L.1 In a search of the growth factors possibly involved in brain ontogenesis we have examined the effects of transforming growth factor p l (TGF-P1) on the growth and phenotypic expression of rat astroblasts in primary culture. Along TGF-p1 elicited only a slight negative effect on the growth of these cells. However, this factor was found to modulate the mitogenic effects of other growth factors. O n quiescent cells it potentiates the mitogenic effect of basic fibroblast growth factor (bFGF) but not that of other growth factors, namely, epidermal growth factor (EGF), platelet-derived growth factor (PDGF), and thrombin. TGF-P1 did not modulate significantly the stimulatory effect of these growth factors on the activity of the enzyme glutamine synthetase (CS); but kinetic studies showed that TGF-PI delays the stimulation of GS activity. D N A synthesis monitored by the incorporation of ['L'l]iododeoxyuridine ('251-dUrd) was maximum after 24-30 h of treatment with bFGF. With bFGF plus TGF-PI the maximum was shifted to 30-36 h. This shift is compatible with the idea that TCF-P1 induces responsiveness in some cells which are otherwise unresponsive to the mitogenic action of bFGF, and that this induction requires some time. This hypothesis is sustained by the observation that in cells treated for only 12 h with bFGF, the treatment with TGF-PI for the same 12 h or for longer time did not stimulate significantly the cell growth. Stimulation occurred only when the bFGF treatment was continued after 12 h. Potentiation of the mitogenic effect of bFGF and shift of the maximum Iz51-dUrd incorporation towards 24 h was seen with cells pretreated with TGF-P1. This potentiation effect decreased with increasing time between the two treatments. The potentiation effect of TGF-Pl i s not mediated by an induction of new bFGF membrane receptors as seen by binding studies.

TGFP belongs to a family of sequence-related proteins involved in the regulation of growth and development, including Mullerian inhibiting substance, inhibins, activins, and the decapentaplegic gene transcript of Drosophila (Cate et al., 1986; Vale et al., 1986; Massague, 1987; Padgett et al., 1987). Blood platelets are the major source of TGFP found in normal serum (Childs et al., 1982). However, TGFB is a highly ubiquitous protein synthesized by most tissues and cell types normal and neoplastic (Roberts and Sporn, 1985) and is often secreted in a latent form (Lawrence et al., 1984, 1985; Pircher et al., 1986). TGFp was originally identified by its capacity, in the presence of other growth factors, to induce certain non-transformed rodent fibroblasts to grow without anchorage in semisolid media, a characteristic of transformed cells (Roberts et al., 1981). Subsequent studies have shown that TGFp elicits complex effects on the proliferation of nor-

Q

1990 WILEY-LISS, INC

mal and transformed cells (Massague, 1985; Roberts et al., 1985; Goustin et al., 1986; Moses et al., 1987) depending on the particular cell type and on the presence of other growth factors. Received February 20, 1990; accepted May 17, 1990. "To whom reprint requestdcorrespondence should be addressed. Frederic Perraud's present address is Transgene S.A., 11, rue de Molsheim, 67082 Strasbourg Cedex, France. Abbreviations used: bFGF, basic fibroblast growth factor; BrdUrd, bromodeoxyuridine; db CAMP, N6,O2'-dibutylyladenosine-3': 5'cyclic-monophosphate; EDTA, ethylenediamine tetracetic acid; EGF, epidermal growth factor; G F M , glial fibrillary acidic protein; GS, glutamine synthetase; "'I-dUrd, 5-L'251110do-2'-deoxyuridine; PBS, phosphate-buffered saline: 0.15 M NaC1,0.05 M sodium phosphate, pH 7.2; PDGF, platelet-derived growth factor; TGF-P, transforming growth factor type P.

474

LABOURDETTE ET AL.

12 10

8

6 4

2

TABLE 1. Effect of bFGF and TGF-pl on the percentage of aroliferating (BrdUrd-uositive) astroblasts (GFAP-oositive cells)' Treatment Control bFGF (20 ngiml) TGF-pl (1 ngiml) bFGF + TGF-P1

BrdUrd-positive cells (70) Growth phase2 Stationary phase3 10 t 9 22 t 15 80 -c 18 57 k 11 22 t 11 8t6 76 2 12 80 2 11

'The cells were treated with the growth factors alone for 16 h and then wlth BrdUrd in addition, for 24 h. Results are expressed as percentage of BrdUrdpositive cells among the GFAP-positive cells, which represent more than 90% of the cells in these cultures. 'Cells treated at day 6 after seeding. %ells treated a t day 24 after needing.

activity on cells grown in monolayer culture has been described with bone cells from fetal bovine calvariae C bFGF EGF T h r PDGF c A M P (Globus et al., 1988) and with corneal endothelial cells (Plouet and Gospodarowicz, 1989). In addition to its effect on cell growth, TGFp also elicits various effects on cell differentiation (or maturation)-positive on bronchial epithelial cells (Masui et al., 1986) and on chondrocytes (Seyedin et al., 1986)but negative on adipocytes (Ignotz and Massague, 19851, myocytes (Massague et al., 1986), B and T lymphocytes (Kehrl et al., 1986a,b),and some other cell types (Massaguh et al., 1987). Various properties and effects of TGFP, such as its presence in embryos and its stimulation of angiogenesis, suggest that this factor is involved in the control of embryonic development (Proper et al., 1982; Heine et al., 1987). A role for TGFp in tissue repair is also likely, given its high concentration in blood platelets (Assoian et al., 1983) and its acceleration of wound healing in rats (Mustoe et al., 1987). 0 C bFGF EGF T h r PDGF c A M P Since we are interested in the role of growth factors Fig. 1. Effect of growth factors and db cAMP on '"I-dUrd incorpo- on astroglial cells during brain development and durration (A) and specific glutamine synthetase activity (B) in rat astro- ing the process of reactive gliosis after tissue damage, blasts in primary culture. The cells were grown in 35 mm dishes for 20 a possible role of TGFP in these phenomena warranted days in the presence of 10%fetal calf serum and then in a chemically investigation. In the present work we studied the effect defined medium for 4 days. They were treated with the agents as indicated at the final concentrations of bFGF, 20 ngiml; EGF, 10 of TGFp, alone or in the presence of other growth facng/ml; thrombin, 1 unitiml; PDGF, 1 unitiml, and db CAMP, 1 mM tors, on the proliferation and on some aspects of the (open blocks) and in the same conditions, but in the presence of TGF- phenotypic expression of rat astroblasts in primary culp l 0.5 ng/ml (hatched blocks). For determination of cell proliferation (A) 0.2 pCiiml of lZ51-dUrdwere added after 24 h of treatment and left ture. Those aspects of phenotypic expression examined for 24 h. Radioactivity incorporated was then determined. For GS were cell morphology and activity of glutamine synspecific activity (B) the cells were treated for 4 days and harvested. thetase (GS) the latter being a marker of astroglial Results of a representative experiment are shown; they represent the maturation (Caldani et al., 1982; Fages et al., 1988). mean of triplicate dishes. MATERIALS AND METHODS Materials In fact, TGFp inhibits the division of most cell types, Tissue culture Petri dishes were from Falcon (Becparticularly cells of epithelial origin which are respon- ton Dickinson, USA). Basal culture medium MD 70511 sive to EGF (Tucker et al., 1984; Sporn and Roberts, and fetal calf serum were purchased from Flow 1985). It stimulates the growth of cells of mesenchymal Laboratories. Fatty acid-free bovine serum albumin origin, such as fibroblasts, in soft agar (Moses et al., (A-6003), epidermal growth factor (E-1257) and insu1987), but also in monolayer culture (Shipley et al., lin (1-5500)were from Sigma. Highly purified thrombin (3,600 unitslmg) was prepared and kindly provided 1985). On cells grown in monolayer, TGF-Pl can inhibit the by J.M. Freyssinet from INSERM U311 a t the Centre mitogenic effect of bFGF. Thus, it inhibits the mitoge- de Transfusion Sanguine in Strasbourg. Acidic and nic effect of FGFs on bovine aortic endothelial cells basic fibroblast growth factors were purified using (Frater-Schroder et al., 1986) on vascular and capillary heparin-Ultrogel affinity chromatography as described endothelial cells (Baird and Durkin, 1986) and on cor- previously (Pettmann et al., 1985) or purchased from R nea and fetal heart endothelial cells (Muller et al., & D Systems, USA. lz5I Na was from CEA (Gif1987) but it does not affect the potent mitogenic activ- sur-Yvette, France). 1251-dUrd was from NEN, Enity of FGF for adrenocortical cells (Hotta and Baird, gland. TGF-p1 was prepared from human platelets 1986). A potentiation by TGF-61 of bFGF mitogenic according to the method of Pircher et al. (1986) or 0

l n r N

475

TGFB MODULATION OF bPGF EFFECTS ON RAT ASTROBLASTS

0

10

20

30

bFGF ( n g / m l )

0

0.5

1

1.5

2

TGF-O ( n g / n ; l )

Fig. 2. Dose response for the proliferation rate of rat astroblasts determined by incorporation of 1251dUrd. Basic FGF alone (A, open circles), TGF-Pl alone (B, open squares), bFGF with a constant dose of 0.5 ngiml TGF-p1 (A, black circles), and TGF-Pl with a constant dose of 20 ngiml bFGF (B, black squares). The cells were grown and treated as described in Figure 1.

purchased from R & D Systems. Anti-BrdUrd monoclonal antibody was from Becton Dickinson. AntiGFAP antibody was from Dakopatts, Denmark, and other antibodies were from Biosys, France.

Rat astroblast culture Primary cultures of astroblasts were prepared as described by Booher and Sensenbrenner (1972)with some modifications: brain hemispheres from newborn rats were dissociated by passage through a 2 mm diameter needle in a small volume of nutrient medium composed of Waymouth's basal medium MD 705/1 enriched with sodium pyruvate (110 mgiL), 10%fetal calf serum, and antibiotics. Thirty dishes (35 mm diameter) or 15 dishes (60 mm diameter) were seeded with cells originating from one brain. Cultures were incubated at 37°C in a 5 % CO, humidified atmosphere. Culture medium was changed after 5 days in vitro and twice a week thereafter. Cell treatments For the experiments shown rat astroblasts were grown in the presence of fetal calf serum for 20 days. Then they were rinsed with basal medium and changed to a serum-free chemically defined medium composed of the same enriched basal medium Waymouth MD 705/1 supplemented with bovine serum albumin (0.5 mg/ml) and insulin (5 pgiml). The cells were maintained for 4 days in the defined medium before treat-

ment. Other details will be given in the text or in the legends.

BrdUrd incorporation and immunocytochemical detection of BrdUrd incorporated and of GFAP The cells were incubated for 24 h at 37°C with BrdUrd at 10 pM in the culture medium. Then, for fixation, cultures were rinsed with PBS and incubated a t room temperature for 10 min first with 4% paraformaldehyde in PBS, then with 20% methanol, 3% hydrogen peroxide in PBS. For detection of GFAP, cultures were rinsed three times with PBS for 5 min, incubated overnight at 4°C with PBS containing 0.5% rabbit anti-GFAP antiserum and 10% normal sheep serum, rinsed three times for 5 min with 0.1% Tween 20 in PBS and once with PBS alone, incubated for 2 h in the dark with PBS containing 1%goat anti-rabbit peroxidase-conjugated y-globulins, rinsed three times with 0.1% Tween 20 in PBS and once with PBS alone, incubated for 20 rnin in the dark with a solution of 5 mg 3,3'-diaminobenzidine and 10 p1 hydrogen peroxide in 10 ml PBS and rinsed with PBS. For detection of BrdUrd, cultures were rinsed for 5 min with 1%Triton X-100 in PBS and with PBS alone, incubated for 30 min with 2 N HC1, rinsed with 0.02 N NaOH, twice with PBS, for 5 min with 0.1 M sodium borate buffer, pH 8.5, then twice with PBS alone and incubated overnight at 4°C with PBS containing 2%

476

LABOURDETTE ET AL. bFGF

I C

c- fos

TGFO

- 1 1 min

bFGF+TGFO

-

1 1 min 30 90 4h 30 W 4h 30 90 min

i 4h

-

-28

s

-18s

bFGF

C

TGFB

bFGF+TGFB

- - -

I min 30

11 mln

904h30904h

1 1 min

3090

1 4h

428s

c-myc-

-18s

Fig. 3. Kinetics of c-fos and c-myc mRNA expression in rat astroblasts grown for 20 days. The cells were treated with bFGF (20 ng/ml), TGF-61 (0.5 ng/ml), or both together and harvested at the times indicated. Five micrograms of RNA were loaded per lane, electrophoresed, transferred to a nylon membrane and hybridized with v-fos and c-myc probes.

cell pellets. The tubes were vortexed and the cell suspensions were sonicated three times for 10 s at 24 W with an MSE Ultrasonic disintegrator Mk2. The glutamine synthetase assay was performed as described by Miller et al. (1978) with slight modifications. Final concentrations in the reaction mixture were 40 mM imidazole-HC1, pH 7, 30 mM glutamine, 0.4 mM sodium ADP, 0.5 mM MnCl,, 20 mM sodium arsenate, and 65 mM hydroxylamine-HC1. A small volume (50250 ~ 1of) the sonicated sample, containing 100 kg proteins, was included in a total volume of 500 k1 and the mixture was incubated at 37°C for 20 min. The reaction was stopped by addition of 1ml of 220 mM Fe(NO,), in 2.5% trichloracetic acid (Iqbal and Ottaway, 1980). The reaction was clarified by centrifugation at 4,000g for 5 min, degassed under vacuum, and the y-glutamylhydroxamate produced was determined by measuring the optical density at 500 nm. Proteins were determined by the method of Lowry et al. (1951) using bovine serum albumin as standard.

RNA extraction and analysis of gene expression Total cytoplasmic RNA was isolated by the LiClUrea technique (Auffray and Rougeon, 1980). The RNAs were electrophoresed in agarose gels in the presence of formaldehyde, and transferred onto Hybond N membranes. Hybridization and washing conditions were performed according to the manual provided by the manufacturer (Amersham). Blots were exposed to X-ray films (X-AR5) for 3 days. C-myc and v-fos probes were 32P-labeled by random primer method, to a specific activity of 2.10' cpmikg.

Iodination of bFGF Ten micrograms of bFGF were labelled with 1251by using chloramine T. Free iodine was separated by gel filtration on a 10 ml Trisacryl GF 05 column in 0.9% NaCl containing 0.1% BSA. The specific activity was anti-BrdUrd monoclonal antibody and 10% normal about 5.6 x lo4 cpmhg (900 cpmifmolej. Analysis of the sheep serum. The next day these cultures were rinsed iodinated bFGF by SDS-PAGE revealed by autoradifor 5 min with three changes of 0.1% Tween 20 in PBS ography a single band migrating a t the position of the and once with PBS alone, incubated for 1 h with 5% unlabelled FGF. Biological activity was preserved. goat anti-mouse antiserum in PBS, rinsed as before and incubated for 2 h in the dark with 1%mouse perBinding of lZ5I-bFGFto astroblasts oxidase-antiperoxidase complex in PBS. Cultures were The cells, previously grown in chemically defined then rinsed three times for 5 min with 0.1% Tween 20 medium for 4 days, were changed to a similar medium in PBS and once with 0.9 % NaCl and incubated for 10 but without sodium bicarbonate and supplemented min in the dark with 0.9 % NaCl containing 0.04% with 20 mM Hepes, pH 7.2. '"I-bFGF was added to the NiC12, 0.018% 4-chloro-1-naphtol, and 0.1% hydrogen medium and the cells were incubated for 2 h a t 4°C peroxide. under orbital shaking. Binding was terminated by removing the medium and washing three times with cold I2%dUrdincorporation 0.15 M NaCl and then three times with 2 M NaC1, 50 1251-dUrd was added at a final concentration of 0.2 mM Tris-HC1, p H 7.5. '251-bFGF bound to high affinity PCilrnl. At the end of the incorporation period the sites was extracted with 2 M NaC1, 20 mM sodium dishes were rinsed three times with 0.9% NaCl at room acetate pH 4 (Moscatelli, 1987). temperature. Cells were scraped off and sedimented. Radioactivity of the pellet was counted with a gamma RESULTS counter. Modulation, by TGF-f31,of the effects of other agents on the growth and GS activity of Glutamine synthetase assay quiescent astroblasts Cultures were rinsed three times with cold 0.9% After 20 days of culture in the presence of fetal calf NaCl and put on ice. Cells were harvested by scraping, sedimented by low speed centrifugation, and kept fro- serum, rat astroblasts are mainly quiescent, showing zen at -20°C. A small volume (200 to 500 ~ 1of) 10 mM only a low residual proliferation rate (Perraud e t al., imidazole-HC1,0.5 mM EDTA, pH 7, was added to the 1988). Treatment of the cells with TGF-P1 alone re-

477

TGFp MODULATION OF bFGF EFFECTS ON RAT ASTROBLASTS

I

m

II

A

"Enn

I 1 I I

2000 -

I I

"

O

K

""",

\

I

I

1

I I

I I

1

1

I

\

I

1

I

-

8

-

*

cc

0

0

1

2

3

1

2

3

D a y s o'f t r e a t m e n t Z

.-

I

4 2

c

t

500

I I

y'i I

I

I

400

\

'

I

I

300

\

200

III

........ *....

100

... & .._._ -

0

1

2

3

0

1

2

3

0

0

Days of treatment

Fig. 4. Time course effects of bFGF (20 ngiml, circles), TGF-61 (0.5 ng/ml, triangles), and their combination (squares) on growth and GS specific activity of rat astroblasts in the stationary phase of growth. The cells were grown and treated as described in Figure 1. In A, B, and C the cells were treated once at time 0 (which is 24 days after seeding) and the culture medium was not changed thereafter. In D, E,

and F the culture medium was changed daily and the cells were treated at the same time. Results are expressed as percent of control. Controls are untreated cells at the day considered. A representative experiment is shown among three which gave similar results. Given values are the mean of triplicate dishes. Errors, which were within the range of 208, are not shown for clarity.

sulted in a slight inhibition of this proliferation (Fig. 1A). Several growth factors have been shown to elicit a mitogenic effect on rat astroblasts in culture, the FGFs (Pettmann et al., 1985), thrombin (Perraud et al., 1987), PDGF (Heldin et al., 1979; Besnard et al., 19871, and EGF (Westermark, 1976). These factors were tested in conjunction with TGF-P1 (hatched blocks), and alone (open blocks) for control (Fig. I). The effect on growth of the potent mitogen bFGF was strongly potentiated by TGF-PI. In contrast, there was no such effect with thrombin, PDGF, and EGF. The activity of the enzyme GS was determined in astroblasts in the same conditions but after treatment for 5 days (Fig. 1B). TGF-P1 (hatched blocks) did not alter significantly GS activity with any of the treatments. Even the effect of db CAMP,tested for comparison, was not modified. It can be observed that GS activity was stimulated by all the mitogenic growth

factors and that EGF was the most potent of them. It appears that the effects on proliferation and on GS activity are not quantitatively related.

Dose response effects of bFGF and TGF-P1 on astroblast growth The quiescent astroblasts were treated for 24 h by combinations of bFGF and TGF-PI. 1251-dUrdincorporation proceeded from 24 h to 48 h. Basic FGF was tested alone (Fig. 2A, dashed line) and in the presence of 0.5 ngiml TGF-P1 (continuous line). The potentiation of bFGF mitogenic effect was observed at the lowest bFGF levels. The two curves are nearly proportional. These data indicate that the cells which become responsive to bFGF following TGF-P1 treatment manifest the same sensitivity to bFGF as the spontaneously responsive cells. TGF-Pl tested alone (Fig. ZB, open squares) inhibits, in a dose-dependent manner, the

478

LABOURDETTE ET AL.

Fig. 5. Incorporation of BrdUrd into nuclei of rat astroblasts not treated A) and treated with bFGF (20 ng/ml, B). The cells were grown and treated a s described in Figure 1. At day 24 after seeding cells were treated with bFGF. After 16 h BrdUrd was added and left until

40 h, following which the dishes were rinsed, the cells fixed (see Methods), and BrdUrd revealed by iinmunocytochemical reaction. The gliofilaments were stained by immunocytochemical reaction with antiGFAP. Bar: 50 ym.

slight increase in DNA content and a decrease in GS specific activity were seen after the daily treatment (Fig. 4E,F). Basic FGF (continuous lines) stimulated transiently the proliferation and progressively the GS specific activity. The renewed treatment extended the duration of Induction of c-fos and c-myc the stimulation of proliferation (Fig. 4D), leading to a The expression of c-fos and c-myc mRNAs was exam- greater increase in DNA content (Fig. 4E). When both ined as a function of time after treatment with bFGF factors were used together (dashed lines), there was a alone, TGF-Pl alone, and with both factors together. strong, but still transient, potentiation of the mitogenic We observed a transient expression of c-fos after bFGF effect of bFGF (Fig. 4A,D). GS specific activity was still treatment, with a maximum around 30 min (Fig. 3). enhanced, but this effect was delayed for 1 or even 2 TGF-P1 alone elicited a weak expression. The combi- days with the daily treatment (Fig. 4F). nation of the two factors induced a much stronger expression than bFGF alone or than the addition of the Cell growth determined by incorporation signals induced by bFGF and TGF-61 individually. Exof BrdUrd pression of c-myc was also transient after bFGF treatment, but with a maximum a t 90 min. TGF-P1 alone Cells were treated with the growth factors on day 6 induced a weak expression at 4 h, visible after a longer or day 24 for 40 h. BrdUrd was added from 16 h to 40 h after the beginning of treatment. GFAP- and BrdUrdtime of exposure (not shown). containing cells were revealed by immunocytochemical Time course effects on quiescent cells methods (Fig. 5). Almost all the cells were characterQuiescent 24-day-old astroblasts were treated either istic GFAP-positive astrocytes. Results are shown in once (Fig 4A-C) or daily, with medium changes (Fig. Table 1. In the stationary phase and without any 4D-F), with bFGF, TGF-P1, or both. 1251-dUrdincorpo- growth factor treatment only 10% of the GFAP-positive ration, DNA content, and GS activity were determined cells proliferate. This figure increased to 57 % upon every day for 3 days. A single treatment with TGF-(31 treatment with bFGF and to 80%with bFGF and TGFP alone (dotted lines) did not affect significantly the together. In the growth phase with bFGF only, about growth and GS activity after one treatment, though a 80% of the cells were proliferating. Such a high persmall residual proliferation of the astroblasts. With a constant dose of bFGF (20 ngiml) (filled squares), a maximum potentiation was seen at 0.5 ngiml TGF-P1. Some decrease was observed at higher concentration of TGF-P1.

Fig. 6 . Morphological appearance in phase-contrast of rat astroblasts after 25 days in culture: control (A); cells treated for 24 h with and bFGF plus TGF-pl bFGF, 20 ng/ml (B); TGF-p1, 0.5 ngiml (0; (D). Bar = 50 ym.

Fig. 7. Morphological appearance in phase contrast of rat astroblasts after 8 days in culture. Other specifications as for Figure 6 .

480

LABOURDETTE ET AL.

2000 -

2000

A

-.-

B 0

S v)

?

-

;\

.!

1500 -

1500

( I

0

.-m

0

! ! ,

;

!

I

I

!

1000

1000 -

E 500

500 -

e

!

.9

! ! 8 !

C

E

I '

.

!

I

!

'!

0

0 C

z

-In N

r

Y

A

0-1 12 18 2 4 3 0 3 6 4 2 4 8

I

I

I

I

1

Time (hours)

1

I

1

I

I

Symbol

Treatments

0 -12h

I

After12h

bFGF TGF-I31 bFGF TGF-I31

1 1

Treatments

Symbol

0 -12h

bFGF TGF-I31

1

After12h bFGF TGF-I31

----y--Fig 8 Kinetics of stimulation of DNA synthesis in rat astroblasts in the stationary phase by bFGF (20 ngiml) and TGF-pl(0 5 ng/ml) The cells were grown for 20 days in the presence of 10% fetal calf serum, and then in the chemically defined medium Treatments began on day 24 (time 0). Culture medium was changed at times 0 and 12 h In A all the cells were treated with bFGF at least from 0 to 12 h In B all the cells were treated with bFGF plus TGF-P1 at least from 0 to 12 h. At the times indicated 0 2 pCi of lZ5I-dUrdwas added per dish (35 mm

diameter, 2 ml culture medium). The radioactivity incorporated after 1 h incubation was counted. Results are means of triplicate dishes. Errors are not shown but were in the range of 15%. For clarity the results of control cells and of cells treated with TGF-P1 alone are not shown. They were almost constant around 60 cpm. A representative experiment is shown. Note that in all the conditions presented the cells were treated with bFGF during the first 12 h of the experiment.

centage is the reason for the lack of a clear potentiation effect of TGF-P1 in these conditions.

treated for 12 h (not shown) and that of the cells treated continuously (filled circles) with bFGF. Probably in both cases one cycle of cell division was induced with a maximum incorporation around 24-30 h after the beginning of the treatment. A combination treatment (bFGF plus TGF-P1) for only 12 h (open squares) resulted in a weak potentiation of the bFGF effect at 30 h. However, the continuous combination treatment (bFGF plus TGF-PI, filled squares) was much more efficient, and the strongest potentiation was then seen after 36 h. When cells were treated with bFGF and TGF-P1 together for 12 h (Fig. 8B) and then with bFGF alone (filled triangles), a strong potentiation was elicited and was even stronger than with the continuous combination treatment. However, if TGF-P1 was continued after 12 h instead of bFGF, the effect was much weaker (open triangles). Thus, a prolonged treatment with TGF-P1 seems to be deleterious. Other experiments were performed to examine the effects of pretreatment of the cells with TGF-P1. In these experiments the cells were treated continuously with bFGF (Fig. 9). As seen before, treatment with TGF-P1 for 12 h potentiated the effect of bFGF (Fig. 9B, open circles) and delayed the maximum incorporation. When cells were only pretreated with TGF-P1

Effects on cell morphology The morphological aspect of the cells was observed after 24 h of treatment. In the stationary phase (Fig. 6) no difference was visible between TGF-P1 treated cells (Fig. 6C) and control cells (Fig. 6A). The effect of bFGF alone was drastic (Fig. 6B) and TGF-P1 reversed transiently this strong effect (Fig. 6D). In the growth phase (Fig. 7 ) ,TGF-P1 induced a slight morphological effect with an elongation of some cells (Fig. 7C). TGF-P1 together with bFGF did not reverse the effect of bFGF but induced a spectacular morphological effect. Nearly all the cells in the dish were arranged in parallel arrays (Fig. 7D). Kinetic aspects This study was undertaken in order to understand better the interrelationship between the effects of bFGF and TGF-P1 on quiescent cells. Proliferation rate was determined every sixth hour by a 1h pulse of lZ5IdUrd. In a first set of experiments (Fig. 8A) all the cells were treated with bFGF at least for 12 h. NO great difference was seen between the growth of the cells

481

TGFp MODULATION OF bFCF EFFECTS ON RAT ASTROBLASTS 2000

2000

B

A c

.-

s!

5

1500

1500

1000

1000

500

500

0

v

c 0 .+

z

g 0 0 E

-

E

3

m

@ r I I

a

n

12

2 18 24 3 0 3 6 4 2 48

18

Time (hours) Symbol

-

1

.-*.

+....*..', ----

- -

TGF-O1Treatment -36h--24 h---12 h I I I I I I I I I I 1 + 1 I

+

30

24

36

42

Time (hours)

I

Fig. 9. Kinetics of stimulation of DNA synthesis in rat astroblasts in the stationary phase. Same specifications as for Figure 8. In these experiments all the cells were treated with bFGF continuously starting a t time 0. In A cells were treated with TGF-Pl from 0 to 12 h (0)

I

0h I I

1

+

I I

-

bFGF Treatment 12 h I I 1 I I I

Oh

I

, I

+ +

I

+

I

+

I

+

or with TGF-61 from 12 h before time 0 to time 0 (A). In panel B (another experiment) are shown cells treated with only bFGF (*I and similar cells pretreated with TGF-pl for 12 h periods starting 12 h (A),24 h (m), or 36 h (0) before the beginning of bFGF treatment.

from 12 h before bFGF treatment until time 0 (filled squares) there was a potentiation of the bFGF effect and a shift towards 24 h. Another experiment is shown in Figure 9B. Pretreatments with TGF-PI for 12 h periods all resulted in potentiation of the bFGF effect and in a shift of the maximum incorporation towards 24 h. The effect decreased with increasing time between the treatment with bFGF and the pretreatment with TGF-p1. It seems that after a pretreatment with TGF-61 astroblasts respond more readily to bFGF.

Binding - of lz5IbFGFto astroblasts The potentiation of the mitogenic effect of bFGF by TGF-p1 could be mediated by a n increase in the number of bFGF specific membrane receptors. To test this hypothesis the cells were treated with TGF-P1 for various periods of time, from 3 h to 48 h, and specific binding of 1251bFGFwas determined. A typical Scatchard analysis is shown in Figure 10, with cells treated or untreated with TGF-P1 for 12 h. No significant differences were seen at any time between untreated and TGF-P1 treated cells. The number of specific binding sites per cell is about 12,000,

0.00

'

0

I

1 0

30

20

bFGF bound (fmole/lO

cells)

Fig. 10. Scatchard analysis of the binding of ""I-bFGF to astroblasts. The cells were grown for 20 days in the serum containing medium and then in serum free defined medium containing 0.1 %J BSA for 4 days (01. Some cells were treated with 0.5 ngiml TGF-PI for 1 2 h before the binding assay + ). All the cells were then rinsed once with a cold serum free medium containing 0.1% BSA, Hepes but without sodium bicarbonate and incubated at 4°C in the same medium for 2 h, with various amounts of lZ51-bFGF.Cells were washed three times with 2 M NaCI, 20 mM Tris-HCI, pH 7.5. Bound *"I-bFGF was extracted with 2 M NaCl, in 20 mM sodium acetate, pH 4.0.

482

LABOURDETTE ET AL

DISCUSSION TGF-f31 alone elicits some effects on rat astroblasts After one treatment for 24 h or 48 h TGF-P1 inhibits the growth of rat astroblasts in a dose-dependent manner. However, after daily or prolonged treatment it induces a slight but significant increase in proliferation and some decrease of GS activity. In low density cultures TGF-P1 induces a morphological modification of some cells. These results document that TGF-p1 alone is able to induce several effects on rat astroblasts, but they are of limited extent. A slight induction of c-fos mRNA was detected. It has been reported previously (Robertson et al., 1988) that TGFp is active on rat astroblasts, in which it induces the translocation of protein kinase C and stimulates phosphoinositol metabolism. The primary effect of TGF-P1 on astroblasts, like on most cells, should be a n inhibition of proliferation, later stimulation could be indirect, related for instance to the secretion and deposition of adhesion proteins, such as fibronectin or collagen, as has been shown for other cell types like fibroblasts, myoblasts, or osteoblasts (Wrana et al., 1988; Fine and Goldstein, 1987; Ignotz and Massague, 1986) and to the expression of the receptors for these proteins (Ignotz and Massague, 1987). It is generally admitted that extracellular matrix components are able to regulate the cell morphology or even cell proliferation (Goetschy et al., 1987). I n t e r p l a y between TGF-P1 and the mitogenic activity of b F G F TGF-P1 was found to potentiate only the mitogenic activity of bFGF (and of aFGF, results not shown), which is the most potent mitogen for astroblasts. I n cells taken in the stationary phase of growth, in which only 10% of the cells proliferate, bFGF alone triggers the proliferation of 57% of the cells, whereas treatment with TGF-P1 and bFGF together triggered 80% of the cells to proliferate. This last percentage is identical to that found in the cultures during their growth phase. Such a n observation shows that rat astroblasts, which have spent 10 to 15 days in the stationary phase of growth, retain the capacity to proliferate after a suitable stimulation. The cells (23%),which are induced to proliferate by the two factors together, have been rendered sensitive to the mitogenic effect of bFGF by TGF-P1. The same conclusion can be drawn from the observation that pretreatment of the cells with TGF-P1 alone is still able to potentiate the mitogenic activity of bFGF. An induction of the synthesis or of the availability of FGF membrane receptors, under the effect of TGF-P1, could be an explanation. However, TGF-P1 treatment did not alter the mean number of bFGF receptors per cell (Fig. 10). When the cells are treated with bFGF, the proliferation induced lasts for only about 48 h. If TGF-P1 is added at that time (48 h), bFGF being still present, no more proliferation is induced (result not shown). Thus, cells which can be made responsive by a n earlier treatment with TGF-p1 do not respond later. A possibility is that these cells, which do not proliferate under the action of bFGF alone, are, however, down-regulated. Basic FGF would bind to these cells, activate the specific

membrane receptors, and be internalized, but without triggering proliferation. The modulation of the mitogenic activity of bFGF on astroblasts by TGF-P1 was found to depend also on the growth phase of the cells. On quiescent cells, the effect of bFGF was potentiated, whereas for cells in the growth phase it was not. In this latter case, the counting of proliferating cells indicated t h a t with bFGF alone 80% of the cells were growing and it seems t h a t this percentage cannot be exceeded. Kinetic aspects and mechanisms of action The strongest potentiation of bFGF mitogenic activity was achieved when the cells were treated with TGF01 for 12 h starting either 12 h before bFGF treatment or at the same time as bFGF treatment. In the latter case bFGF treatment must be prolonged after 12 h for maximal effect. The other main observation is that the pretreatment with TGF-P1 shifts forward the maximal mitogenic effect by about 6 h compared to treatment starting a t time 0. Thus cells treated with TGF-P1 respond more readily to bFGF. This result indicates that TGF-P 1 pretreatment renders some cells, which normally would not respond to bFGF, responsive to bFGF, though this process requires some time. Under our experimental conditions, i n which TGFP l potentiates the effect of bFGF, a n early effect of bFGF, the expression of c-fos mRNA, is modulated by TGF-P1. The expression of c-fos is considered as a key event in the transduction of some external signals leading to growth and differentiation (Adamson, 1987; Ruther et al., 1987). TGF-P1 alone stimulates very slightly the expression of c-fos. A combined TGF-P1bFGF treatment enhances strongly this expression. The superinduction of c-fos by TGF-P1 shows that the effect induced by TGF-P1 is manifested very early. TGF-P1 could also concern other steps of the transduction process since a long treatment with TGF-P1 is required for a maximum potentiation. It is known also that c-fos induction is a necessary but not sufficient event in the triggering of cell proliferation. Binding studies have shown that TGF-P1 does not induce a n increase in the mean number of bFGF specific binding sites on the cells. This observation suggests that TGF-P1 action takes place at a n intermediate transduction step inside the cell, facilitating the mitogenic effect of bFGF. Such a lack of effect on the number of bFGF specific membrane receptors has also been shown with endothelial cells (Plouet and Gospodarowicz, 1989). TGF-p1, bFGF, and phenotypic expression During the stationary phase of growth, TGF-P1 delays the increase of GS activity induced by bFGF. The delay could be related to the induction of proliferation, preventing transiently the expression of GS. Numerous examples have been reported of TGF-pl, alone or in combination with another growth factor, affecting the phenotypic expression of cells or their differentiation. For instance, TGFP alone stimulates the c-sis oncogene expression in fibroblasts and the FGF expression in corneal endothelial cells (Plouet and GOSpodarowicz, 1989), but it inhibits tumor necrosis factor a (TNF-a) production by human monocytes (Espevik et al., 1987). Cooperation between FGF and TGFP has

TGFp MODULATION OF bFGF EFFECTS ON RAT ASTROBLASTS

been shown to affect cell maturation or differentiation. TGFp enhances the effects of FGF to induce normal levels of muscle actin in animal pole explants (Kimelman and Kirschner, 1987). From this work it has been suggested that FGF and TGFp are the natural inducers of mesoderm in vertebrate development. Thus, the association of TGF-Pl with bFGF could be involved in the development of specific organs or cell types. The observations that the FGF-induced increase in GS activity is delayed by TGF-p1 and that the morphological change induced by bFGF is also modulated by TGF-PI suggest that the latter factor is acting on most if not all the cells as regards the phenotypic parameter examined in the culture. Only a subpopulation of these cells respond by proliferating. Two different mechanisms are possibly involved in the triggering of these two effects. Our finding that TGF-P1 potentiates the effect of bFGF on proliferation and on cell morphology and modulates GS activity of rat astroblasts suggests that associations of TGF-p1 and FGF could play a role in the development of glial cells and also in the phenomenon of reactive gliosis. Support for this hypothesis is also provided by the presence of both FGF and TGF(3in the central nervous system (Pettmann et al., 1986; Moses et al., 1981). The complexity of the effects of TGF-p1, associated with bFGF, on rat astroblasts will stimulate further investigations on the mechanisms of action of these two factors.

ACKNOWLEDGMENTS The technical assistance of M.F. Knoetgen was appreciated. We are grateful to C. Thomassin-Orphanides for typing and to J.M. Freyssinet for providing highly purified human thrombin. This work was supported in part by a grant (6451) from the “Association pour la Recherche sur le Cancer”. LITERATURE CITED Adamson E.D. (1987) Oncogenes in development. Development. 99: 449-471. Assoian R.K., Komoriya A., Meyers C.A., Miller D.M., and Sporn M.B. (1983)Transforming growth factor-p in human platelets. Identification of a major storage site, purification, and characterization. J. Biol. Chem., 258:7155-7160. Auffray C., and Rougeon F. (1980) Purification of mouse immunoglobulin heavy-chain messenger RNAs from total myeloma tumor RNA. Eur. J. Biochem., 107r303-309. Baird A., and Durkin T. (1986) Inhibition of endothelial cell proliferation by type p-transforming growth factor : interactions with acidic and basic fibroblast growth factors. Biochem. Biophys. Res. Commun., 138:476-482. Besnard F., Perraud F., Sensenbrenner M., and Labourdette G. (1987) Platelet-derived growth factor is a mitogen for glial but not for neuronal rat brain cells in vitro. Neurosci. Lett., 73:287-292. Booher J., and Sensenbrenner M. (1972) Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flasks culture. Neurobiology, 2~97-105. Caldani, M., Roland B., Fages C., and Tardy M. (1982) Glutamine synthetase activity during mouse brain development. Experientia, 38:1199-1202. Cate, R.L., Mattaliano R.J., Hession C., Tizard R., Farber N.M., Cheung A,, Ninfa E.G., Frey A.Z., Gash A.Z., Chow D.J., Fisher E.P., Bertonis R.A., Torres G., Wallner B.P., Ramachandran K.L., Ragin R.C., Manganaro T.F., MacLaughlin D.T., and Donahoe P.K. 119861 Isolation of the bovine and human genes for Mullerian inhibiting substance and expression of the human gene in animal cells. Cell, 45:685-698.

483

Childs C.J., Proper J.A., Tucker R.F., and Moses H.L. (1982) Serum contains a platelet-derived transforming growth factor. Proc. Natl. Acad. Sci. USA, 79:5312-5316. Espevik T., Figari IS., Shalaby M.R., Lackides G.A., Lewis G.D., Shepard H.M., and Palladino M.A. (1987)Inhibition of cytokine production by cyclospain A and transforming growth factor p. J. Exp. Med., 166:571-576. Fagees C., Rolland B.. Dias Costa M.F., Khelil M., Dupre G., CampagGoni A.T., and Tardy M. (1988) Messenger RNA-coding for g1;tamine synthetase in cerebral hemispheres and astroglial cultures from mouse brain : a developmental study. Neurochem. Int., 12: 307-313. Fine A., and Goldstein R.H. (1987) The effect of transforming growth factor-p on cell proliferation and collagen formation by lung fibroblasts. J. Biol. Chem., 262:3897-3902. Frater-Schroder M., Muller G., Bircheier W., and Bohlen P. (1986) Transforming growth factor-beta inhibits endothelial cell proliferation. Biochem. Biophys. Res. Commun.. 137t295-302. Globus R., Buckendahl P., and Gospodarowicz D. (1988) Regulation of osteoblast proliferation by fibroblast growth factor and transforming growth factor B. Endocrinology, 123r98-105. Geotschy J.F., Ulrich G., Aunis D., and Ciesielski-Treska J. (1987) Fibronectin and collagens modulate the proliferation and morphology of astroglial cells in culture. Int. J . Devl. Neurosci., 5:63-70. Goustin A.S., Leof E.B., Shipley G.B., and Moses H.L. (1986) Growth factors and cancer. Cancer Res., 46:1015-1029. Heine, U.I., Munoz E.F., Flanders K.C., Ellingsworth L.R., Peterlam H.Y., Thompson N.L., Roberts A.B., and Sporn M.B. (1987) Role of transforming growth factor-p in the development of the mouse embryo. J. Cell. Riol., 105:2861-2876. Heldin C.H., Westermark B., and Wasteson A. (1979) Platelet-derived growth factor : purification and partial characterization. Proc. Natl. Acad. Sci. USA, 76~3722-3726. Hotta M., and Baird A. (1986) Differential effects of transforming growth factor type p on the growth and function of adrenocortical cells in vitro. Proc. Natl. Acad. Sci. USA, 83r7795-7799. Ignotz R.A., and Massague J. (1985) Type beta transforming growth factor controls the adipogenic differentiation of 3T3 fibroblasts. Proc. Natl. Acad. Sci. USA, 82t8530-8534. Ignotz R.A., and Massague J. (1986) Transforming growth factor+ stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J. Biol. Chem., 261r43374345. Ignotz R.A., and Massague J. (1987) Cell adhesion protein receptors as targets for transforming growth factor-p action. Cell, 51r189197. Iqbal K., and Ottaway J.H. (1980) Improved sensitivity of glutamine synthetase assay and of acyl hydroxamates. Neurochem. Res., 5: 805-808. Kehrl J.H., Roberts A.B., Wakefield L.M., Jakowlew S., Sporn M.B., and Fauci A S . (1986a) Transforming growth factor-beta is a n important immunomodulatory protein for human B lymphocytes. J. Immunol., 137t3855-3860. Kehrl H.H., Wakefield L.M., Roberts A.B., Jakowlew S.,Alvarez-Mon M., Derynck R., Sporn M.B., and Fauci A S . (1986b) Production of transforming growth factor-beta by human T lymphocytes and its potential role in the regulation of T cell growth. J. Exp. Med., 163: 1037-1050. Kimelman D., and Kirschner M. (1987) Synergistic induction of mesoderm by FGF and TGFp and the identification of a n mRNA coding for FGF in the early Xenopus embryo. Cell, 5:869-877. Lawrence D.A., Pircher R., Kryceve-Martinerie C., and Jullien P. (1984) Normal embryo fibroblasts release transforming growth factors in a latent form. J. Cell. Physiol., 121:184-188. Lawrence D.A., Pircher R., and Jullien P. (1985) Conversion of a high molecular weight latent TGF-p from chicken embryo fibroblasts into a low molecular weight active TGF-p under acidic conditions. Biochem. Biophys. Res. Commun., 133t1026-1034. Lowry O.H., Rosebrough N.J., Farr A.L., and Randall R.J. (1951) Protein measurement with the fohn phenol reagent. J. Biol. Chem., 193,265-275, Massague J. (1985) The transforming growth factors. Trends Biochem. Sci., 10:237-240. Massague J., Cheifetz S., Endo T., and Natal-Ginard B. (1986) Type p transforming growth factor is an inhibitor of myogenic differentiation. Proc. Natl. Acad. Sci. USA, 833206-8210. Massague J. (1987) The TGF-p family of growth and differentiation factors. Cell, 49t437-438. Massague J.,Cheifetz S., Ignotz R.A., and Boyd F.T. (1987) Multiple type-p transforming growth factors and their receptors. J. Cell. Physiol. [Suppl.], 543-47.

484

LABOURDETTE ET AL.

Masui T., Wakeield L.M., Lechner J.F., LaBeck M.A., Sporn M.B., and Harris C.C. (1986) Type beta transforming growth factor is the primary differentiation-inducing serum factor for normal human bronchial epithelial cells. Proc. Natl. Acad. Sci. USA, 83: 2438-2442. Miller R.E., Hackenberg R., and Gershman H. (1978) Regulation of glutamine synthetase in cultured 3T3-U cells by insulin, hydrocortisone and dibutyryl cyclic AMP. Proc. Natl. Acad. Sci. USA, 75t1418-1422. Moscatelli D. (1987) High and low affinity binding sites for basic fibroblast growth factor on cultured cells : absence of a role for low affinity binding in the stimulation of plasminogen activator production by bovine capillary endothelial cells. J. Cell. Physiol., 131: 123-130. Moses H.L., Branum E.B., Proper J.A., and Robinson R.A. (1981) Transforming growth factor production by chemically transformed cells. Cancer Res., 41:2842-2848. Moses H.L., Coffey R.J., Leof E.B., Lyons R.M., and Keski-Oja J. 11987) Transforming growth factor p regulation of cell proliferation. J. Cell. Physiol. LSuppl.1, 5tl-7. Muller G., Behrens J., Nussbaumer H., Bohlen P., and Birchmeier W. (1987) Inhibitory action of transforming growth factor B on endothelial cells. Proc. Natl. Acad. Sci. USA, 84:5600-5604. Mustoe T.A., Pierce G.F., Thomason A., Gramates P., Sporn M.B., and Deuel T.F. (1987) Accelerated healing of incisional wounds in rats induced by transforming growth factor-p. Science, 237r1333-1336. Padgett R.W., St Johnston D.R., and Gelbart W.M. (1987) A transcript from a Drosophila pattern gene predicts a protein homologous to the transforming growth factor-p family. Nature, 325:81-84. Perraud F., Besnard F., Sensenbrenner M., and Labourdette G. (1987) Thrombin is a potent mitogen for rat astroblasts but not for oligodendroblasts and neuroblasts in primary culture. Int. J. Dev. Neurosci., 5r181-188. Perraud F., Besnard F., Sensenbrenner M., and Labourdette G. (1988) Effect of acidic and basic fibroblast growth factors (aFGF and bFGF) on the proliferation and glutamine synthetase expression of rat astroblasts in culture. Glia, 1:124-131. Pettmann B., Weibel M., Sensenbrenner M., and Labourdette G. (1985) Purification of two astroelial .>mowth factors from bovine brain FEBS Lett., 189 102-108 Pettmann B., Labourdette G., Weibel M , and Sensenbrenner M. (1986) The brain fibroblast growth factor (FGF) IS localized in neurons. Neurosci. Lett., 68t175-180. Pircher R., Jullien D., and Lawrence D.A. (1986) p-Transforming growth factor is stored in human blood platelets as a latent high molecular weight complex. Biochem. Biophys. Res. Commun., 136: 30-37. Plouet J., and Gospodarowicz D. (1989) Transforming growth factor Y

p-1 positively modulates the bioactivity of fibroblast growth factor on corneal endothelial cells. J. Cell. Physiol., 141t392-399. Proper J.A., Bjornson C.L., and Moses H.L. (1982) Mouse embryos contain polypeptide growth factorb) capable of inducing a reversible neoplastic phenotype in nontransformed cells in culture. J . Cell. Physiol., 110:169-174. Roberts A.B., Anzano M.A., Lamb L.C., Smith J.M., and Sporn M.B. (1981) New class of transforming growth factors potentiated by epidermal growth factor : isolation from non-neoplastic tissues. Proc. Natl. Acad. Sci. USA, 78x5339-5343. Roberts A.B., Anzano M.A., Wakefield L.M., Roche N.S., Stern D.F., and Sporn M.B. (1985) Type p transforming growth factor : a bifunctional regulator of cell growth. Roc. Natl. Acad. Sci. USA, 82t119-123. Roberts A.B., and Sporn M.B. (1985) Transforming growth factors. Cancer Surveys, 4583-705. Robertson P.L., Markovac J.,Datta S.C., and Goldstein G.W. (1988) Transforming growth factor beta stimulates phosphoinositol metabolism and translocation of protein kinase C in cultured astrocytes. Neurosci. Lett., 93t107-113. Ruther U., Garber C., Komitowski D., Muller R., and Wagner E.F. (1987) Deregulated c-fos expression interferes with normal bone development in transgenic mice. Nature (London), 324:412-416. Seyedin S.M., Thompson A.Y., Bentz H., Rosen D.M., McPherson J.M., Conti A., Siege1 N.R., Galluppi G.R., and Piez K.A. (1986) Cartilage- inducing factor-A. Apparent identity to transforming growth factor-beta. 5. Biol. Chem., 261:5693-5695. Shipley G.D., Tucker R.F., and Moses H.L. (1985) Type ptransforming growth factorigrowth inhibitor stimulates entry of monolayer cultures of AKR-2B cells into S phase after a prolonged prereplicative interval. Proc. Natl. Acad. Sci. USA, 82:4147-4151. Sporn M.B., and Roberts A.B. (1985) Autocrine growth factors and cancer. Nature (Lond.), 313t745-747. Tucker R.F., Shipley G.D., Moses H.L., and Holley R.W. (1984) Growth inhibitor from BSC-1 cells closely related to platelet type p transforming growth factor. Science, 226;705-707. Vale W., Rivier J., Vaughan J., McClintock R., Corrigan A., WOOW., Karr D., and Spiess J. (1986) Purification and characterization of an FSH releasing protein from porcine ovarian follicular fluid. Nature, 321 t776-779. Westermark B. (1976) Density dependent proliferation of human glia cells stimulated by epidermal growth factor. Biochem. Biophys. Res. Commun., 69:304-309. Wrana J.L., Maeno, M., Hawrylyshyn K.-L., Yao C. Domenicucci, and Sodek J . (1988) Differential effects of transforming growth factor-p on the synthesis of ertracellular matrix proteins by normal fetal rat calvarial bone cell populations. J. Cell. Biol. 106:915-924.