Grcrwth Factors, 1992, Vol. 6, pp. 265-275 Reprints available directly from the publisher Photocopyingpermitted by license only
0 1992Harwood Academic Publishers GmbH
Printed in the United Kingdom
Transforming Growth Factor Beta Stimulates Mitogenically Mouse NIH3T3 Fibroblasts and Those Cells Transformed by the EJ-H-ras Oncogene OMAR BENZAKOUR, ABDERRAHIM MERZAK, YOLANDE DOOGHE, MARTINE PIRONIN, DAVID LAWRENCE and PHILIPPE VIGIER* Unit6 1443 CNRS, Institut Curie-Biologie, Bat. 110, Centre Universitaire, 91405 Orsay Cidex, France
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
(Received August 20 1991, Accepted November 25 1991)
TGF-j3l stimulates thymidine incorporation and the growth rate of mouse NIH3T3 fibroblasts and of those cells transformed by the EJ-H-ras oncogene (TR15 cells), in the presence and the absence of serum. Thymidine incorporation, in serum-deprived cells, is stimulated to a higher degree by 0.1-1 n g / d of TGF-/3 in NIH3T3 than in TR15 cells, which have a 10-fold higher basal level of incorporation. In both cell types TGF-D is as active, or more active than other mitogens (TGF-a, PDGF-AB, bFGF) at the same concentration. The growth rate of NIH3T3 cells, in low serum or serum-free (S-) medium, is stimulated by only 10 picograms/ml of TGF-fl, and that of TR15 cells, in Smedium, by only 1 picogram/ml. In contrast, TGF-fl inhibits mitogenically unestablished mouse embryo fibroblasts and these fibroblasts immortalized spontaneously and able to grow in S- medium. It also inhibits the anchorageindependent growth of TR15 cells. NIH3T3 and TR15 cells respond, similarly, to TGF-/3 activated by acification of their culture medium. The kinetics of thymidine incorporation and of activation of the c-myc proto-oncogene, observed already after 1 hr, in treated NIH3T3 and TR15 cells, suggests a direct mitogenic stimulation. The level of activated c-myc RNA is 2-fold higher at 2 hr, and subsequently decreases relatively less in the TR15 cells. KEYWORDS transforming growth factor beta (TGF-P), NIH3T3 cells, EJ-H-ras oncogene, stimulation of growth
lished, but non transformed rat NRK-49F cells, in synergy with the epidermal growth factor (EGF) TGF-/.?is a bifunctional regulator of cell growth, or TGF-a (Anzano et al., 1983), and that of mouse active on a great variety of cells, and able to AKR-28 cells without the necessity of exogenous either stimulate or inhibit their multiplication, EGF (Tucker et al., 1983). However, it was subdepending on the cell type, the conditions of cul- sequently found to inhibit the growth of anchture (with or without anchorage), and the context ored NRK-49F cells and its stimulation by EGF set by other factors present. It also has effects (Roberts et al., 1985). Moreover, following transunrelated to proliferation, notably promotion or formation of these cells by the K-ras oncogene of inhibition of cell differentiation, stimulation of Ki-MSV, TGF-P inhibits both their anchored and extracellular matrix formation and suppression A1 growth, and EGF reverses only partially the of production of immunoglobulin by B cells (for inhibition of the latter growth (Jullien et al., reviews, see Barnard et al., 1990; Massaguk, 1990; 1988). TGF-P also inhibits the anchored growth of Roberts and Sporn, 1990). As regards fibroblastic primary rat embryo fibroblasts and the stimucells, TGF-P was first shown to stimulate the lation of their A1 growth by PDGF; and it simianchorage-independent (A11 growth of estab- larly inhibits the A1 growth of established NIH3T3 fibroblasts and its stimulation by PDGF (Anzano et al., 1986). Inhibition of growth of both anchored and unanchored cells was further “Corresponding author.
INTRODUCTION
265
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
266
BENZAKOUR el al.
observed with cancer cell lines of various origins, and appears to be the most frequent response to TGF-P. Yet, it may be modulated by accompanying growth factors (Roberts et al., 1985). The study of the mouse AKR-2B cells which are stimulated mitogenically by TGF-P, with or without anchorage, has also shown that the G1 period of the cell cycle induced by this factor is prolonged by about 12 hr, when compared to G1 in other growth factor-stimulated cells (Shipley et al., 1985). This delayed response appears to be due to the induction by TGF-/3 of the c-sis mRNA and the PDGF-B chain encoded by this gene, which is presumably the true mitogen, acting autocrinally. Hence, TGF-P may act as an indirect mitogen (Leof et al., 1986). Keeping in mind the above findings, and since viral and cellular oncogenes seem able to alter the response of fibroblastic cells to TGF-P, we decided to reexamine the action of this factor on mouse NIH3T3 fibroblasts, and to extend this study to these same cells transformed by the Hras oncogene, which is closely related to the K-ras oncogene (Barbacid, 1987). Since all previous studies with TGF-P had been carried out in medium containing serum, we also decided to study its action in defined serum-free medium containing only adjuvant growth factors (insulin and transferrin). We have observed that the growth of attached non transformed and transformed NIH3T3 cells is stimulated, with and without serum, by very low concentrations of TGF-PI, which appears to be more active on the transformed cells. In contrast, TGF-fl inhibits the A1 growth of the transformed cells, in semisolid medium, with or without serum, and also the growth of attached unestablished mouse embryo fibroblasts and of these same fibroblasts after spontaneous immortalization and acquisition of the capacity to grow without serum.
MATERIALS AND METHODS
Cells and Media Mouse NIH3T3 fibroblasts were kindly provided by Dr G. M. Cooper (Sydney Farber Cancer Institute, Boston, MA, USA) and by Dr J. Ghysdael, in our Institute, and have been called, respectively NIH3T3-1 and NIH3T3-2 cells. A clonal line of NIH3T3-1 cells transformed by the H-ras onco-
gene, cloned by Shih and Weinberg (1982) from the human EJ bladder carcinoma cell line, was obtained by transfecting a monolayer culture with the original molecular clone of these authors (pEJ6.6) and explanting a single transformed focus, with a cloning cylinder. Transfection was carried out by the calcium phosphate precipitation method of Graham and van der Eb (19731, as modified by Copeland and Cooper (1979). The transformed clone (TR15) was highly tumorigenic in athymic mice, able from the start to grow in serum-free (S-) medium (see below), and also, after a few passages, to form A1 colonies in semisolid S- medium. Twenty other clones, isolated in the same way, behaved similarly (Pironin et al., in preparation). Tertiary cultures of mouse embryo fibroblasts (MEF, P3) were obtained by subculturing secondary cultures derived from cultures of 2-week-old embryos from a syngeneic line of mice (NC-Z), bred in our Institute and characterized by a very low incidence of spontaneous tumors. An immortalized cell line established spontaneously, after repeated passages of these fibroblasts, was also studied, at its 18th passage. It was then able to grow actively in S- medium, and, consequently, called MEF/S-. But it formed no A1 colonies, in semi-solid medium containing serum, and was only weakly tumorigenic in athymic mice. The NIH3T3 cells were cultivated in DMEM supplemented with 5-10% newborn calf serum (NCS), and MEF, P3 in DMEM plus 10% fetal calf serum (FCS); whereas the TR15 and MEF/S- ceIls were passaged in S- medium consisting of equal volumes of DMEM and HamF-12 medium (DHM). Two other S- media were also used to study the action of TGF-P on cell growth: (i) DHM supplemented with insulin (5 pglml), transferrin (5 pg/ml) and sodium selenite (5 ng/ml), in the form of Premix (Collaborative Research, Waltham, MA, USA) which was called DHM+; (ii) McClure's medium (MCM), containing the same components as DHM', plus other additives favoring cell growth in the absence of serum (McClure, 1983). This medium was formerly shown to allow the growth of NIH3T3 cells transformed by EJ-H-ras and other oncogenes (Zhan and Goldfarb, 1986). Growth Factors Highly purified TGF-Pl and TGF-a were pur-
TGF-p STIMULATES MITOGENICALLY NIH3T3 FIBROBLASTS
chased from R. and D. Systems Inc. (Minneapolis, MN, USA). Tissue grade mouse EGF and purified PDGF (AB heterodimer) from human platelets were obtained from Sigma (St Louis, MO, USA), and purified basic FGF (bFGF) kindly supplied by the laboratory of Dr D. Barritault (Universite de Crkteil, France). TGF-fl, TGF-a and PDGF were dissolved and diluted in 4mM HCl plus 1 mg/ml bovine serum albumin (BSA), and EGF and bFGF in DMEM plus 1 mg/ml BSA.
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Bioassays
267
2x104 trypsinized cells in semi-solid S' or Smedium (DHM plus 10% NCS or DHM') in the 16 mm diameter wells of 24-well Linbro Plates, as previously described (Li Wen Xin et al., 1987); except that Sea Plaque agarose (0.6% in the underlayer and 0.4% in the cell containing overlayer) was substituted for agar. TGF-fl, or other growth factors, were added on top of the overlayer, in 0.1 ml of diluent, and colonies of over 30 cells (estimated by their size under the microscope) were scored after 14 days at 37 "C, in replicate wells, following cristal violet staining.
Growth of Attached Cells
DNA Synthesis
The growth of attached cells was appreciated by their growth rate, in S' and S- media, and their cloning efficiency, in S- medium, for TR15 cells. In the first assay, lo4 to 5x104 trypsinized cells were plated in 3.5cm Falcon dishes, in DMEM plus 10% NCS (or FCS, for MEF,P3), and this medium was replaced, one day later, by DMEM plus 2% CS (or 5% FCS for MEF, P3), or by DHM' or MCM, for growth experiments in S- medium. In the latter case, the cultures were rinsed twice, for 1 hr, with DMEM, before addition of the Smedium; and the dishes were also precoated with polylysine and fibronectin (Zhan and Goldfarb, 19861, to prevent cell detachment, in studies with NIH3T3 cells and MEF, P3. This treatment was not necessary for TR15 cells and MEF/S-. TGF-P1 was added at the time of medium renewal, one day after plating, and the S' or S- medium, with or without the factor, was renewed at 2-3 days interval. The cells were trypsinized and counted with a Coulter counter, in replicate cultures, 6-10 days after the first medium renewal. The cloning efficiency of the TR15 cells in S- medium was assayed by plating 2x103cells in 3.5 cm dishes, in DMEM plus 10% NCS, and switching the cultures one day later to DHM', with or without TGF-PI, as described above. The medium was renewed as in the growth assays, and attached colonies of over 100 cells scored with a microscope 11 days after plating, following fixation with 2% glutaraldehyde and staining with cristal violet. Visible (macroscopical) colonies were also counted.
DNA synthesis was assayed by measuring the incorporation of 3H-thymidine into the TCAinsoluble material of confluent cultures, in the wells of Linbro plates. These cultures were obtained by plating lo4 TR15 cells or 5x104 NIH3T3 cells in DMEM plus 5% NCS, and allowing them to grow for 2 days. They were then rinsed twice with PBS and switched to SDMEM containing TGF-fl, or other growth factors (0.5 ml/well). After 6-24 hr, the cultures were labeled for 1 hr with 3H-thymidine (NEN, France, 80 Ci/mMol, 1pCi/ml), then washed and extracted with 5% TCA, and lysed overnight, at room temperature, with 0.5N NaOH. TCAinsoluble radioactivity was subsequently measured by scintillation counting, in Optiphase I1 scintillation fluid (Pharmacia).
Anchorage-Independant (AI) Growth A1 growth of TR15 was assayed by seeding lo3 or
TGF-/3Assays TGF-P activity in the media of NIH3T3 and TR15 cells was assayed by measuring the inhibition by these media of DNA synthesis (i.e., of incorporation of 3H-thymidine)in mink lung CC164 cells, as described by van Zoelen et al. (1986), by the stimulation of the formation of A1 colonies by rat NRK-49F fibroblasts, in the presence of EGF, as described by Jullien et al. (1988), and by the inhibition of binding of lz5ITGF-fl to its receptors on these fibroblasts (Massaguk and Like, 1985).
Acidification of Cell-Conditioned Media Cell-conditioned media were obtained by incubating confluent cultures 2 days in S- medium (DHM'). The crude media were centrifuged, to eliminate cells and debris, and acidified with HC1
268
BENZAKOUR et al.
to pH 2.0, then reneutralized with NaOH. Alternatively, they were concentrated under acidic conditions, by dialysis against 1 M acetic acid, followed by lyophilization. The lyophilized residue was dissolved and diluted in 4 mM HCl plus 1 mg/ml BSA and sterilized with 6oCo gamma-rays.
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Chromatography (FPLC) Reverse phase chromatography (FPLC) of the cell-conditioned TR15 medium, concentrated 20fold under acidic conditions, was carried out as described earlier (Pircher et al., 1984). The fractions collected from the acetonitrile gradient were lyophilized, redissolved in 4 mM HC, plus 1 mg/ml BSA, and sterilized with 6oCogammarays. Cytoplasmic RNA: Isolation and Slot Blot Analysis Cytoplasmic RNA was isolated as described by Dooley et al. (1989). For each experimental point, 10 p g of RNA were applied onto a nitrocellulose filter, using a minifold apparatus (Biorad), as described by Maniatis et al. (1989). The RNA amounts tested were initially measured by absorbance at 260 nm, and later more accurately by quantitating the 28s rRNA in each sample by migration on an agarose-formaldehyde gel. The slot blot filter was prehybridized and hybridized with the c-myc probe as described by Maniatis et al. (1989). The c-myc probe was an Alul/HaeIII fragment of 0.6 Kb, representing the third exon of c-myc and labelled by nick translation (BRL kit) at a specific activity of 10' cpmlpg.
capacity to grow in S- medium (MEF/S-: see Materials and Methods). 3H-Thymidine Incorporation Serum-deprived cultures of NIH3T3 on TR15 cells were treated for 16-24 hr with TGF-P1, or other growth factors (for comparison), then labeled for 1 hr with 3H-thymidine. As seen from Table 1, thymidine incorporation into cellular DNA in the absence of additives was 10-fold higher in TR15 cells than in NIH3T3 cells, presumably due to the fact that the former, but not the latter, could grow in the S- medium. The low basal level of incorporation in NIH3T3 cells was increased over 10-fold by TGF-PI, at 0.1 or lng/ml, and also by TGF-a and bFGF at the same concentrations, but only 2-fold by PDGF, at 1 ng/ml. TGF-fl also increased the high basal level of incorporation in TR15 cells (over 2-fold for 1 ng and 1.5-fold for 0.1 ng/ml), whereas no significant enhancement was seen with the other factors. Similar results were obtained in repeated experiments, which also showed that the incorporation observed with 1 ng/ml of TGF-PI was not increased by addition of any of the other factors, at the same concentration. We further studied the kinetics of thymidine incorporation in serum-deprived cells treated with 1 ng/ml of TGF-fl. As is seen from Fig. 1, in both cell types, incorporation was increased already after 12 hr in S- medium containing TGFPl. Thus, stimulation of thymidine incorporation by TGF-PI in NIH-3T3 and TR15 cells appears to TABLE 1 Incorporation of 'H-thymidine in Serum-deprived NIH3T3 and TR15 Cells: Action of TGF-fl and Other Growth Factors ~~
Additives
RESULTS
The capacity of TGF-Dl to stimulate mitogenically NIH3T3 and TR15 cells was first studied by assaying its action on the incorporation of 'H-thymidine into DNA of serum-deprived cells. Subsequently, its capacity to stimulate the growth of these cells was studied. For comparison, we also studied the action of TGF-PI on the growth of mouse embryo fibroblasts, at their third passage (MEF, P3) and following their spontaneous establishment and acquisition of the
None TGF-PI
1 .o
TGF-CI
0.1 1 .o 0.1
PDGF-AB bFGF
~
-
Concentration 'H-TdR incorporation (cprn/well (ng/ml) x l o-')
1.0 0.1 1 .o
0.1
NIH3T3 cells
TR15 cells
0.5 10.0 7.0 6.0 5.5
4.5 11.0 7.0 5.5 5.O 4.5
1 .o 0.7 10.5 8.0
5.0 5.0 6.0
(6
NIH3T3 cells, seeded a t 5xlO'/well, or TR15 cells, a t 10'/well 16 mm wells of Linbro plates), were switched arter 2 days to S DMEM, and labeled, 24 hr later for 1 hr with 'H-TdR. TCA-insoluble radioactivity was measured as described in Materials and Methods.
TGF-P STIMULATES MITOGENICALLY NIH3T3 FJBROBLASTS
269
3.5cm dishes, in DMEM plus 2% NCS (2A, 2B). Growth stimulation was a function of the concen15- A tration of TGF-PI, and already observed with r 0.01 ng/ml (Fig. 3A). In contrast, under the same conditions, no stimulation of growth was Q) observed with TR15 cells which grew much fas\ ter than the nontransformed parental cells (Fig. E 2C). TR15 cells, at 5x104/dish, also grew actively in S- medium (DHM') and were not stimulated either by TGF-fl (Fig. 2C). However, when they 6 12 24 6 12 were seeded at 104/dish, TGF-P1 stimulated Hours markedly their growth rate, at concentrations FIGURE 1. Kinetics of 3H-thymidine incorporation in serum above 0.01 ng/ml. A plateau of maximal stirnudeprived anchored NIH3T3 (A) and TR15 (B) cells, treated ng/ml Of TGF-fl, in s' with TGF-m and PDGF-AB. The factors were added. at lation was reached at 1 ng/ml, to replicate cultures switched to S' medium, as in medium, and 0.3 ng/ml, in-S- medium, and the Table 1, and the cultures labeled 6, 12 and 24 hr later. 0, cell number attained in the presence of TGF-fl controls; 0,TGF-PI; PDGF. was higher in S- then in S' medium, although it was 3-fold lower without TGF-P1 (Fig. 3 0 . The pursuit of this study confirmed that follow usual kinetics. In this experiment, with both cell types, no increase of incorporation at growth stimulation of TR15 in S- medium was 12 hr was seen after addition of 1 ng/ml of dose-dependent with a maximum stimulation PDGF-AB, and only a low increase at 24 hr, in corresponding to a 10-fold increase of the cell NIH3T3 cells. The low increase also observed, at number, relative to controls, at about 0.3 ng/ml, and a significant stimulation at 0.001 ng/ml (Fig. this time, in TR15 cells was not significant. 3D). A dose-dependent stimulation of growth in S- medium (DHM') was also observed with both Growth of Attached Cells lines of NIH3T3 cells, seeded in dishes coated As seen from Fig. 2, TGF-P, at 2ng/ml, stimu- with polylysine and fibronectin. In these conlated markedly the growth rate of attached ditions, the cells of both lines remained attached, NIH3T3 cells obtained from two different labora- but divided only once or twice in the absence of tories (NIH3T3-1 and -2, see Materials and TGF-PI, over a 10-day period, whereas their final Methods), and seeded at 5x104 cells/dish, in number was increased 6- to 12-fold, relative to
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
-
.,
Oeo3
FIGURE 2. Effect of TGF-fl on the anchored growth of NIH3T3 and TR15 cells, and unestablished (MEF, P3) and established (MEF/S-) mouse embryo fibroblasts. Cells were plated at 5x104/dish in 3.5 cm dishes, in DMEM+10% NCS (or 10% FCS, for MEF, P3) and switched one day later to DMEM+2% NCS (or 5% FCS, for MEF, P3), or to S- DHM', containing 2 ng/ml of TGF-PI. The medium was renewed as indicated by arrows and the cells counted, in duplicate cultures, 6 or 8 days after plating. Full symbols: cultures+ TGF-fl; empty symbols: controls. A, B: NIH3T3-1 and NIH3T3-2 cells, in DMEM+2% NCS; C: TR15 cells in DMEM+Z% NCS (0, .), or in DHM' (A, A); D MEF,P3 in DMEM+5% FCS (0, .), and MEF/S- in DMEM+Z% NCS (A, A), or in D H M (V, V).
-
Days
270
BENZAKOUR et al.
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
controls, in the presence of TGF-P1, at 0.3-1 ng/ml; i.e., to the same degree as the number of TR15 cells. However, the lowest TGF-PI concentration active on NIH3T3 cells was TO-fold higher (0.01 ng/ml) than that active on TR cells (Fig. 3B). TGF-P1 also stimulated, in a dose-dependent manner, the cloning efficiency of TR15 cells seeded in low number (2x103/dish), in nontreated dishes, in S' medium, and switched the next day to S- medium. The stimulation of formation of attached colonies was maximum at 0.1 ng/ml, and half the maximum increase was
observed at 0.01 ng/ml, the lowest concentration tested. Large colonies visible without a microscope, were also scored in the treated, but not in the control cultures (Fig. 4A). In contrast to its enhancing action on the growth of NIH3T3 and TR15 cells, TGF-P1 at 2 ng/ml, inhibited markedly (2-fold reduction of the final cell number) the growth of attached MEF,P3 in DMEM plus 5% FCS, and also inhibited significantly the growth of established MEF/S-, in DMEM plus 2% NCS, or in DHM' (Fig. 2D). The non-established MEF, P3 failed to grow in S- medium, even in MCM (see Materials
FIGURE 3. Effect of TGF-PI on the growth of anchored NIH3T3 and TR15 cells with and without serum. Same protocol as in Fig. 2, except that NIH3T3 cells switched to SDHM' were plated in dishes coated with polylysine and fibronectin, and TR cells were seeded at IO'/dish. A: NIH3T3-1 cells in DMEM+Z% NCS at 6 days; B: NIH3T3-1 cells (0) and NIH3T3-2 cells (A) in DHM' at 11 days; C: TRl5 cells in DMEM+Z% NCS (B) or DHM' ( 0 )at 8 days; D TR15 cells in DHM', at 11 days. N/No: final cell No/No of plated cells. Dotted line: N/No in untreated controls.
FIGURE 4. Effect of TGF-P, on the formation of attached colonies (cloning efficiency) and anchorageindependent (AI) colonies by TR15 cells. A: Cloning efficiency of 2x10' cells, plated in DMEM+I% NCS, in 3.5 cm dishes, and switched one day later to S- DHM'. Colonies of over 100 cells ( 0 ) and macroscopical colonies (W were scored at 11 days. B, C: Formation of A1 colonies by 10' TR15 cells in 16 mm wells of Linbro plates, in S- DHM' (B) or DHM+10% NCS (C). Colonies were scored at 14 days.
ov,
.1
I
1
1
0
I
L
.01 .1
1
10
0
1
I
1
.01 .1
ng/ml TGF-fll
1
10
TGF-/3 STIMULATES MITOGENICALLY NIH3T3 FIBROBLASTS
27 I
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
and Methods) and in dishes coated with polylysine and fibronectin, and were neither stimulated nor inhibited by TGF-fl, in these conditions.
inhibition of DNA synthesis in mink CC164 cells (Pironin et al., in preparation). The conditioned media were either acidified and reneutralized, or concentrated 5-fold, by dialysis against 1 M acetic acid, followed by lyophilization (see Growth of Unanchored TR15 Cells Materials and Methods), and assayed on replicate As is seen from Fig. 4 (B, C), TGF-fl also cultures. inhibited the formation of A1 colonies by TR15 As is seen from Table 2, the untreated crude cells, in semi-solid S' medium (DMEM plus 10% NIH3T3 medium did not increase thymidine NCS) or S- medium (DHM'). However, inhibition incorporation in NIH3T3 cells and the crude of colony formation was observed with TR15 medium displayed only a low enhancing 0.1 ng/ml of TGF-/?l in S- medium, against activity; whereas the acidified/neutralized media 1 ng/ml in S' medium. This suggests that serum and these media concentrated by dialysis against contains some factorb) counteracting the inhibi- acetic acid increased markedly incorporation in tory action of TGF-fl. This factor is unlikely to the target cells. No increase of the high level of be PDGF which did not revert the inhibitory basal incorporation in TR15 cells was observed action of TGF-fl, at a concentration of 1 ng/ml after addition of the crude or the acidified and reneutralized media; but a significant increase (data not shown). (over 2-fold) was seen after addition of the concentrated acidified media. Action on NIH3T3 and TR15 Cells of TGF-P When tested on unanchored TR15 cells, in Activated in Cell Conditioned Medium semi-solid DHM', the crude media stimulated Since TGF-P is present in a latent form, activat- significantly (over 2-fold) A1 colony formation, able by acid treatments, in the culture medium of due to the presence of cell-secreted factors able to nontransformed and transformed avian and stimulate this growth (Pironin et al., in preprodent fibroblastic cells (Lawrence et al., 1984; aration). Following acidification/neutralization, Pircher et al., 1984), the acidified medium of or concentration with acetic acid, the TR medium these cells, containing activated TGF-fl, should simultaneously lost its enhancing activity and stimulate mitogenically NIH3T3 and TR15 cells became inhibitory, with a maximum effect for and inhibit A1 growth of the latter cells. The experiments were carried out with the S- medium of confluent cultures of NIH3T3 and TR15 cells which contain no active TGF-P, detectable by TABLE 2 Action of Untreated and Acidified S- Conditioned Media on 'H-thymidine Incorporation in NIH3T3 and TR15 Cells and A1 Growth of TR15 Cells, in S- Medium Additives
'H-TdR incorporation (cpm/ wellxlo") NIH3T3 cells TRl5 cells
None TR med, crude id.AN id.,conc. 5xAC NIH3T3.med,crude id.AN id,conc. 5xAC
2.5 3.5(1.4) 12.5(5) 25(10) 2.5(1.0) 6(2.4) 41(16)
25 25(1.0) 25(1.O) 58Q.3) 25(1.O) 25(1.O) 60(2.4)
Percentage colony formation by TR15 cells 0.18 0.40(2.2) 0.16(0.9) 0.03(0.17) 0.45Q.5) 0280.6) 0.06(0.33)
Assays as for TGF-PI. The untreated (crude) or acidified/neutralized (AN) media, or these media concentrated under acidic conditions (AC). by dialysis against 1 M acetic acid and lyophilization, were added at 1 : l O (v/v) to replicate cultures of serum-deprived cells, for 'H-TdR incorporation, or to unanchored cells, in semi-solid medium, for A1 growth of TR15 cells. In parentheses, cpm or A1 colbnies in treated cultures: cpm or A1 colonies in controls.
FIGURE 5. Stimulation of 'H-thymidine incorporation in serum-deprived NIH3T3 cells by the fractions obtained by reverse phase chromatography (FPLC) of the S- medium of TR15 cells. One ml of TR medium, concentrated 20-fold under acidic conditions, was injected onto a reverse phase column, type Pro-RP 5/2 (Pharmacia). After 8 m n at 25% acetonitrile/O.l% TFA, a gradient of acetonitrile to 45% was applied over 32 mn. The fractions were assayed as in Table 1, and the active fractions of the second peak reassayed for TGF/3 activity (cf. Materials and Methods).
212
BENZAK.OUR er al.
lating thymidine incorporation in serumdeprived NIH3T3 cells eluted in the same fractions of the acetonitrile gradient as TGF-P activity, detected by inhibition of DNA synthesis in CC164 cells, or by the capacity to enhance A1 colony formation of NRK-49F cells, in the presence of EGF, or to compete for TGF-P receptors on the latter cells (see Materials and Methods). Figure 5 shows the elution profile of the mitogenic activity assayed on NIH3T3 cells. The additional early peak of activity presumably corresponds to PDGF activity which was detected in the TR medium (Pironin et al., in preparation).
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Mechanism of Action: Stimulation of c-myc Transcription The kinetics of thymidine incorporation in serum-deprived NIH3T3 and TR15 cells treated with TGF-fl suggested that it might stimulate their division by a direct mechanism. We, therefore, studied, in these cells, the kinetics of activation of the c-myc proto-oncogene, which is upregulated in early response to growth factors with a direct mitogenic action on quiescent cells (reviewed in Cole, 1986). As seen from Fig. 6, c-myc was up-regulated, in both cell lines, iH 2H 3H already 1 hr after addition of TGF-P1 (2 ng/ml). FIGURE 6. Induction of c-myc RNA by TGF-B. (a) NIH3T3 The level of c-myc RNA increased markedly at and TR15 cells, serum starved for 24 hr, were stimulated by 2 hr and was then 2-fold higher in TR15 cells than TGF-Pl (2 ng/ml) for the indicated times before preparation of cytoplasmic RNA, slot blot analysis, and hybridization with a in NIH3T3 cells, as indicated by densitometric c-myc probe (see Materials and Methods). (b) Densitometric scanning analysis. At 12 hr, it had declined 30scanning values of the slot blot of treated cells expressed as the fold in NIH3T3 cells, but only 5-fold in TR15 percentage of c-myc RNA at 2hr in the TR15 cells. No induction was observed in the untreated controls, which are cells. not presented.
DISCUSSION the acidified concentrate. Acidification/ neutralization of the NIH3T3 medium reduced its enhancing activity without suppressing it, but the acidified concentrate was highly inhibitory. Hence, the acidified media mimicked the actions of TGF-PI on NIH3T3 and TR15 cells. In line with this finding, DNA synthesis in mink CC164 cells was strongly inhibited by the acidified concentrates, somewhat less by the non concentrated acidified/neutralized TR medium, and much less by the non concentrated acidified/neutralized NIH3T3 medium (data not shown). Moreover, following reverse phase chromatography (FPLC) of the TR medium, concentrated 20-fold with acetic acid, the activity stimu-
Our results show that TGF-P is an effective mitogen for NIH3T3 fibroblasts and these cells transformed by EJ-H-ras, in S' and S- medium. This activity is demonstrated by its capacity to stimulate DNA synthesis, in serum-deprived attached cells, and the growth rate of these cells, in S' medium and S- medium containing insulin, transferrin and sodium selenite. In serumdeprived NIH3T3 cells, TGF-P1, at 0.1 ng/ml, increases markedly thymidine incorporation and is as active as TGF-a or bFGF, and much more active than PDGF, at the same concentration. It also increases, at only 0.01 ng/ml, the growth rate of the attached cells, in low serum medium,
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
TGF-B STIMULATES MITOGENICALLY NIH3T3 FIBROBLASTS
or in S- medium, in dishes coated with polylysine and fibronectin; and a marked increase is observed at 1-2 ng/ml, the highest concentrations tested. It should be noted that the two sublines of NIH3T3 cells used in our experiments were able to divide once or twice in S- medium, in the precoated dishes, contrary to the earlier report of Zhan and Goldfarb (1986). In any case, TGF-PI appears to cooperate with serum growth factors and factors in the S- medium to stimulate their growth. TGF-fl also stimulates DNA synthesis of attached serum-deprived TR15 cells, and their growth in S' and S- medium. DNA synthesis in confluent cultures is stimulated to a much lower degree than in NIH3T3 cells, presumably due to the fact that it is 10-fold higher in the untreated serum-deprived TR cells. This high basal level appears to be due to the secretion by the TR cells of autocrinally active mitogenic factors related to TGF-a, PDGF, and FGF and, apparently, other unrelated factors (Pironin et al., in preparation). Therefore, TGF-P presumably acts in cooperation with these factors. On the other hand, it appears to be more active on serum-deprived TR cells than exogenous TGF-a, PDGF or bFGF, which do not increase the basal level of thymidine incorporation, at the same concentration. In attached cultures of TR15 cells plated at high density (5x104/3.5cm dish), the growth rate, in low serum medium or S- medium, is not increased by TGF-P, at 2ng/ml, whereas it is increased significantly in cultures plated at lower density, at concentrations comparable to those active on NIH3T3 cells. Moreover, in the less dense cultures, in S- medium, the lowest active concentration of TGF-PI is 10-fold lower than the lowest active concentration for NIH3T3 cells, i.e. only 1 picogram/ml. These findings suggest that, in the dense cultures, TGF-fl, at the concentrations tested, does not enhance the already high mitogenic activity of the autocrine factors, whereas it is able to enhance it, in the less dense cultures, in coooperation with these factors. The capacity of TR cells to respond to a very low concentration of TGF-P1 is presumably due to this cooperation. However, TR cells may also be more sensitive to TGF-P than NIH3T3 cells. Along this line, we have further checked that the responsiveness of the TR15 cells to TGF-PI, in Smedium, is as high, or higher than that of mink lung CC164 cells, assayed by inhibition of DNA
273
synthesis (van Zoelen et al., 1986). Thus, the stimulation of growth of TR15 cells, in Smedium, appears to constitute a new sensitive assay for TGF-P. This assay may be used to detect activated TGF-P in acidified cell-conditioned medium, as shown by our experiments with the acidified NIH3T3 and TR media. However, other factors in these media might modulate the TGF-/3 activity. It should be recalled here that only the acidified media contained active TGF-P. Thus, the autocrine activity of the untreated TR cell medium must be due to other factors. Regarding the mechanism of the mitogenic stimulation of NIH3T3 and TR cells by TGF-PI, the question arises as to whether it initiates directly the cell cycle, as do other bona fide growth factors, or acts indirectly by the intermediary of activated PDGF, as observed for murine AKR-2B fibroblasts by Leof et al. (1986). As recalled in the introduction, these authors observed a 12 hr prolongation of the G1 period of the cell cycle, following stimulation by TGF-P, which presumably corresponds to the time required for activation of the c-sis gene and "indirect" autocrine stimulation by the PDGF encoded by this gene. However, this indirect, mechanism is ruled out in the case of NIH3T3 and TR15 cells, since normal kinetics of thymidine incorporation are observed in these cells, stimulated by TGF-P, together with an early upregulation of the c-myc proto-oncogene. This is, to our knowledge, the first observation showing that TGF-fl may act as a direct mitogen. As regards the higher level of transcription and of persistance of c-myc RNA in TR cells, relative to NIH3T3 cells, it is in line with the responsiveness of the former cells to a lower concentration of TGF-fl, and might, therefore, also be due to the cooperative action of the autocrine factors released by the TR cells. Since, in addition, the growth of anchored fibroblastic cells is generally inhibited by TGF-P (see Introduction), the apparently direct mitogenic response of NIH3T3 and TR cells to TGF-fl seems exceptional and remains to be explained. One of the events involved in the acquisition of the responsiveness may be immortalization. However, the fact that immortalized NC-Z embryo fibroblasts are inhibited mitogenically by TGF-P, like their early passage ancestors, shows it is insufficient. In contrast to its mitogenic action on anchored
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
214
BENZAKOUR et al.
NIH3T3 and TR cells, TGF-PI inhibits the A1 growth of the latter cells, in S' and S- media and induces a cell response similar to that generally observed with other transformed cells, including other fibroblastic cells transformed by rus oncogenes (see Introduction). As concerns the inhibition of A1 growth observed in S' medium, it may be noted that TGF-fl was previously shown to inhibit the A1 growth of nontransformed NIH3T3 cells in this medium (Anzano et al., 1986). Hence, the loss of anchorage appears to reverse the action of TGF-p on both nontransformed and transformed NIH3T3 cells. This reversion constitutes a new example of the bifunctional activity of this TGF, depending on the conditions of culture and environmental factors (Roberts et al., 1985; Sporn and Roberts, 1988; Roberts and Sporn, 19901, and only hypothetical models may be proposed to explain it. One possibility is that, in the absence of anchorage, genes required for the mitogenic response to TGF-/3 may cease to be expressed, allowing reversion to antiproliferative responsiveness. The unanchored cells may also cease to produce some factodd which synergizes with TGF-P to induce the mitogenic response. Further studies are required to test these, and other possible models of action of TGF-P, and to precise the mechanism of its direct mitogenic action. ACKNOWLEDGMENTS This work was s u p p o r t e d by grants from la Ligue Nationale Frangaise contre le Cancer and from t h e INSERM (Contract no 884015).W e thank Drs G. Goubin and G. Calothy, i n our Institute, for kindly supplying respectively, the pEJ6.6 plasmid and the c-myc probe. W e are also indebted t o Mrs A. Besnard for her skillful secretarial assistance.
REFERENCES Anzano, M. A., Roberts, A. B., Smith, J. M., Sporn, M. B. and De Larco, J. E. (1983) Sarcoma growth factor from conditioned medium of virally transformed cells is composed of both type a and type p transforming growth factors. Proc. Natl. Acad. Sci. U S A 80, 6264-6268. Anzano, M., Roberts, A. 8. and Sporn, M. B. (1986) Anchorage-independent growth of primary rat embryo cells is induced by platelet-derived growth factor and inhibited by type-beta transforming growth factor. ]. Cell. Physiol. 126, 312-318. Barbacid, M. A. (1987) ras genes. Ann. Rev. Biochem. 56, 779-827.
Barnard, J. A,, Lyons, R. M. and Moses, H. L. (1990) The cell biology of transforming growth factor p. Biochem. Biophys. Acta 1032,7947. Cole, M. D. (1986) The myc oncogene: its role in transformation and differentiation. Ann. Rev. Genet. 20,361-384. Copeland, N. G. and Cooper, G. M. (1979) Transfection by exogenous and endogenous murine retrovirus DNAs. Cell 16,347-356. Dooley, S., Ehrhart, E., Radtke, J., Unteregger, G. and Blin, N. A. (1989) A procedure for simultaneous isolation of undegraded RNA, DNA and nuclear proteins suitable for interaction analysis in homologous systems. Methods in Mol. Cell. Biol. 1,95-105 Graham, F. L. and van der Eb, A. J. (1973) A new technique for the assay of human adenovirus 5 DNA. Virology 52, 4-67. Jullien, P., Berg, T. M., De Lannoy, C. and Lawrence, D. A. (1988) Bifunctional activity of transforming growth factor type /3 on the growth of NRK-49F cells, normal and transformed by Kirsten murine sarcoma virus. ]. Cell. Physiol. 136,175-181. Lawrence, D. A., Pircher, R., Kryceve-Martinerie, C. and Jullien, P. (1984) Normal embryo fibroblasts release transforming growth factors in a latent form. ]. Cell. Physiol. 121, 184-188. Leof, E. B., Proper, J. A., Goustin, A. S., Shipley, G. D., DiCorleto, P. E. and Moses, H. L. (1986) Induction of c-sis mRNA and activity similar to platelet-derived growth factor by transforming growth factor p: a proposed model for indirect mitogenesis involving autocrine activity. PTOC. Natl. Acad. Sci. U S A 83,2453-2457. Li Wen Xin, Jullien, P., Lawrence, D. A,, Pironin, M. and Vigier, P. (1987). Chemically and virally transformed cells able to grow without anchorage in serum-free medium: evidence for an autocrine growth factor. ]. Cell Physiol. 131, 175-1 83. Maniatis, T., Fritsch, E. F. and Sambrook, J. (1989) Molecular Cloning: A Lnboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Massaguk, J. (1990) The transforming growth factor-p family, Ann. Rev. Cell. Biol. 6, 587-641. Massagu6, J. and Like, B. (1985) Cellular receptors for type-P transforming growth factor: ligand binding and affinity labeling in human and rodent cell lines. ]. Biol. Chem. 260, 2636-2645. McClure, D. B. (1983) Anchorage-independent colony formation of SV40 transformed BALB/c-3T3 cells in serum-free medium: role of cell- and serum-derived factors. Cell 32, 999-1006. Pircher, R., Lawrence, D. A. and Jullien, P. (1984) Latent /% Transforming Growth Factor in non-transformed and Kirsten sarcoma virus-transformed Normal Rat Kidney cells, clone 49F. Cancer Res. 44,5538-5543. Roberts, A. B., Anzano, M. A,, Wakefield, L. M., Roche, N. S., Stem, D. F. and Sporn, M. B. (1985) Type p transforming growth factor. A bifunctional regulator of cellular growth. Proc. Natl. Acad. Sci. U S A 82,119-123. Roberts, A. 8. and Sporn, M. B. (1990) The transforming growth factor-beta. In Handbook of Experimental Pharmacology. 1. Peptide Growth Factors and Their Receptors, (M. B. Sporn and A. B. Roberts, eds), Springer-Verlag, Berlin, pp. 419-472. Shih, C. and Weinberg, R. A. (1982) Isolation of transforming sequence from a human bladder carcinoma cell line. Cell 29, 161-169. Shipley, G. D., Tucker, R. F. and Moses, H. L. (1985) Type p transforming growth factor/growth inhibitor stimulates entry of monolayer cultures of AKR-2B cells into S phase afte; a prolong& prereplicative interval. Proc. Natl. Acad. Sci. USA 82,4147-4151.
TGF-/3 STIMULATES MITOGENICALLY NEI3T3 FIBROBLASTS
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Sporn, M. 8. and Roberts, A. B. (1988) Peptide growth factors are multi-functional Nature 332,217-219. Tucker, R. F., Volkenant, M. E., Branum, E. i. and Moses, H. L. (1983) Comparison of intra- and extracellular transforming growth factors from nontransformed and chemicallly transformed mouse embryo cells. Cancer Res. 43, 1581-1 586. Van Zoelen, E. J. J., Van Oostwaard, T. M. J. and De Laat,
215
S. W. (1986) PDGF-like growth factor induces EGFpotentiated phenotypic transformation of normal rat kidney cells in the absense of TGF-P. Biochem. Biophys. Res. Commun. 141,1229-1235. Zhan, X. and Goldfarb, M. (1986) Growth factor requirements of oncogene-transformed NIH3T3 and BALB/c-3T3 cells cultured in defined media. Mol. Cell. Biol. 6,3541-3544.
0 1992Harwood Academic Publishers GmbH
Printed in the United Kingdom
Transforming Growth Factor Beta Stimulates Mitogenically Mouse NIH3T3 Fibroblasts and Those Cells Transformed by the EJ-H-ras Oncogene OMAR BENZAKOUR, ABDERRAHIM MERZAK, YOLANDE DOOGHE, MARTINE PIRONIN, DAVID LAWRENCE and PHILIPPE VIGIER* Unit6 1443 CNRS, Institut Curie-Biologie, Bat. 110, Centre Universitaire, 91405 Orsay Cidex, France
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
(Received August 20 1991, Accepted November 25 1991)
TGF-j3l stimulates thymidine incorporation and the growth rate of mouse NIH3T3 fibroblasts and of those cells transformed by the EJ-H-ras oncogene (TR15 cells), in the presence and the absence of serum. Thymidine incorporation, in serum-deprived cells, is stimulated to a higher degree by 0.1-1 n g / d of TGF-/3 in NIH3T3 than in TR15 cells, which have a 10-fold higher basal level of incorporation. In both cell types TGF-D is as active, or more active than other mitogens (TGF-a, PDGF-AB, bFGF) at the same concentration. The growth rate of NIH3T3 cells, in low serum or serum-free (S-) medium, is stimulated by only 10 picograms/ml of TGF-fl, and that of TR15 cells, in Smedium, by only 1 picogram/ml. In contrast, TGF-fl inhibits mitogenically unestablished mouse embryo fibroblasts and these fibroblasts immortalized spontaneously and able to grow in S- medium. It also inhibits the anchorageindependent growth of TR15 cells. NIH3T3 and TR15 cells respond, similarly, to TGF-/3 activated by acification of their culture medium. The kinetics of thymidine incorporation and of activation of the c-myc proto-oncogene, observed already after 1 hr, in treated NIH3T3 and TR15 cells, suggests a direct mitogenic stimulation. The level of activated c-myc RNA is 2-fold higher at 2 hr, and subsequently decreases relatively less in the TR15 cells. KEYWORDS transforming growth factor beta (TGF-P), NIH3T3 cells, EJ-H-ras oncogene, stimulation of growth
lished, but non transformed rat NRK-49F cells, in synergy with the epidermal growth factor (EGF) TGF-/.?is a bifunctional regulator of cell growth, or TGF-a (Anzano et al., 1983), and that of mouse active on a great variety of cells, and able to AKR-28 cells without the necessity of exogenous either stimulate or inhibit their multiplication, EGF (Tucker et al., 1983). However, it was subdepending on the cell type, the conditions of cul- sequently found to inhibit the growth of anchture (with or without anchorage), and the context ored NRK-49F cells and its stimulation by EGF set by other factors present. It also has effects (Roberts et al., 1985). Moreover, following transunrelated to proliferation, notably promotion or formation of these cells by the K-ras oncogene of inhibition of cell differentiation, stimulation of Ki-MSV, TGF-P inhibits both their anchored and extracellular matrix formation and suppression A1 growth, and EGF reverses only partially the of production of immunoglobulin by B cells (for inhibition of the latter growth (Jullien et al., reviews, see Barnard et al., 1990; Massaguk, 1990; 1988). TGF-P also inhibits the anchored growth of Roberts and Sporn, 1990). As regards fibroblastic primary rat embryo fibroblasts and the stimucells, TGF-P was first shown to stimulate the lation of their A1 growth by PDGF; and it simianchorage-independent (A11 growth of estab- larly inhibits the A1 growth of established NIH3T3 fibroblasts and its stimulation by PDGF (Anzano et al., 1986). Inhibition of growth of both anchored and unanchored cells was further “Corresponding author.
INTRODUCTION
265
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
266
BENZAKOUR el al.
observed with cancer cell lines of various origins, and appears to be the most frequent response to TGF-P. Yet, it may be modulated by accompanying growth factors (Roberts et al., 1985). The study of the mouse AKR-2B cells which are stimulated mitogenically by TGF-P, with or without anchorage, has also shown that the G1 period of the cell cycle induced by this factor is prolonged by about 12 hr, when compared to G1 in other growth factor-stimulated cells (Shipley et al., 1985). This delayed response appears to be due to the induction by TGF-/3 of the c-sis mRNA and the PDGF-B chain encoded by this gene, which is presumably the true mitogen, acting autocrinally. Hence, TGF-P may act as an indirect mitogen (Leof et al., 1986). Keeping in mind the above findings, and since viral and cellular oncogenes seem able to alter the response of fibroblastic cells to TGF-P, we decided to reexamine the action of this factor on mouse NIH3T3 fibroblasts, and to extend this study to these same cells transformed by the Hras oncogene, which is closely related to the K-ras oncogene (Barbacid, 1987). Since all previous studies with TGF-P had been carried out in medium containing serum, we also decided to study its action in defined serum-free medium containing only adjuvant growth factors (insulin and transferrin). We have observed that the growth of attached non transformed and transformed NIH3T3 cells is stimulated, with and without serum, by very low concentrations of TGF-PI, which appears to be more active on the transformed cells. In contrast, TGF-fl inhibits the A1 growth of the transformed cells, in semisolid medium, with or without serum, and also the growth of attached unestablished mouse embryo fibroblasts and of these same fibroblasts after spontaneous immortalization and acquisition of the capacity to grow without serum.
MATERIALS AND METHODS
Cells and Media Mouse NIH3T3 fibroblasts were kindly provided by Dr G. M. Cooper (Sydney Farber Cancer Institute, Boston, MA, USA) and by Dr J. Ghysdael, in our Institute, and have been called, respectively NIH3T3-1 and NIH3T3-2 cells. A clonal line of NIH3T3-1 cells transformed by the H-ras onco-
gene, cloned by Shih and Weinberg (1982) from the human EJ bladder carcinoma cell line, was obtained by transfecting a monolayer culture with the original molecular clone of these authors (pEJ6.6) and explanting a single transformed focus, with a cloning cylinder. Transfection was carried out by the calcium phosphate precipitation method of Graham and van der Eb (19731, as modified by Copeland and Cooper (1979). The transformed clone (TR15) was highly tumorigenic in athymic mice, able from the start to grow in serum-free (S-) medium (see below), and also, after a few passages, to form A1 colonies in semisolid S- medium. Twenty other clones, isolated in the same way, behaved similarly (Pironin et al., in preparation). Tertiary cultures of mouse embryo fibroblasts (MEF, P3) were obtained by subculturing secondary cultures derived from cultures of 2-week-old embryos from a syngeneic line of mice (NC-Z), bred in our Institute and characterized by a very low incidence of spontaneous tumors. An immortalized cell line established spontaneously, after repeated passages of these fibroblasts, was also studied, at its 18th passage. It was then able to grow actively in S- medium, and, consequently, called MEF/S-. But it formed no A1 colonies, in semi-solid medium containing serum, and was only weakly tumorigenic in athymic mice. The NIH3T3 cells were cultivated in DMEM supplemented with 5-10% newborn calf serum (NCS), and MEF, P3 in DMEM plus 10% fetal calf serum (FCS); whereas the TR15 and MEF/S- ceIls were passaged in S- medium consisting of equal volumes of DMEM and HamF-12 medium (DHM). Two other S- media were also used to study the action of TGF-P on cell growth: (i) DHM supplemented with insulin (5 pglml), transferrin (5 pg/ml) and sodium selenite (5 ng/ml), in the form of Premix (Collaborative Research, Waltham, MA, USA) which was called DHM+; (ii) McClure's medium (MCM), containing the same components as DHM', plus other additives favoring cell growth in the absence of serum (McClure, 1983). This medium was formerly shown to allow the growth of NIH3T3 cells transformed by EJ-H-ras and other oncogenes (Zhan and Goldfarb, 1986). Growth Factors Highly purified TGF-Pl and TGF-a were pur-
TGF-p STIMULATES MITOGENICALLY NIH3T3 FIBROBLASTS
chased from R. and D. Systems Inc. (Minneapolis, MN, USA). Tissue grade mouse EGF and purified PDGF (AB heterodimer) from human platelets were obtained from Sigma (St Louis, MO, USA), and purified basic FGF (bFGF) kindly supplied by the laboratory of Dr D. Barritault (Universite de Crkteil, France). TGF-fl, TGF-a and PDGF were dissolved and diluted in 4mM HCl plus 1 mg/ml bovine serum albumin (BSA), and EGF and bFGF in DMEM plus 1 mg/ml BSA.
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Bioassays
267
2x104 trypsinized cells in semi-solid S' or Smedium (DHM plus 10% NCS or DHM') in the 16 mm diameter wells of 24-well Linbro Plates, as previously described (Li Wen Xin et al., 1987); except that Sea Plaque agarose (0.6% in the underlayer and 0.4% in the cell containing overlayer) was substituted for agar. TGF-fl, or other growth factors, were added on top of the overlayer, in 0.1 ml of diluent, and colonies of over 30 cells (estimated by their size under the microscope) were scored after 14 days at 37 "C, in replicate wells, following cristal violet staining.
Growth of Attached Cells
DNA Synthesis
The growth of attached cells was appreciated by their growth rate, in S' and S- media, and their cloning efficiency, in S- medium, for TR15 cells. In the first assay, lo4 to 5x104 trypsinized cells were plated in 3.5cm Falcon dishes, in DMEM plus 10% NCS (or FCS, for MEF,P3), and this medium was replaced, one day later, by DMEM plus 2% CS (or 5% FCS for MEF, P3), or by DHM' or MCM, for growth experiments in S- medium. In the latter case, the cultures were rinsed twice, for 1 hr, with DMEM, before addition of the Smedium; and the dishes were also precoated with polylysine and fibronectin (Zhan and Goldfarb, 19861, to prevent cell detachment, in studies with NIH3T3 cells and MEF, P3. This treatment was not necessary for TR15 cells and MEF/S-. TGF-P1 was added at the time of medium renewal, one day after plating, and the S' or S- medium, with or without the factor, was renewed at 2-3 days interval. The cells were trypsinized and counted with a Coulter counter, in replicate cultures, 6-10 days after the first medium renewal. The cloning efficiency of the TR15 cells in S- medium was assayed by plating 2x103cells in 3.5 cm dishes, in DMEM plus 10% NCS, and switching the cultures one day later to DHM', with or without TGF-PI, as described above. The medium was renewed as in the growth assays, and attached colonies of over 100 cells scored with a microscope 11 days after plating, following fixation with 2% glutaraldehyde and staining with cristal violet. Visible (macroscopical) colonies were also counted.
DNA synthesis was assayed by measuring the incorporation of 3H-thymidine into the TCAinsoluble material of confluent cultures, in the wells of Linbro plates. These cultures were obtained by plating lo4 TR15 cells or 5x104 NIH3T3 cells in DMEM plus 5% NCS, and allowing them to grow for 2 days. They were then rinsed twice with PBS and switched to SDMEM containing TGF-fl, or other growth factors (0.5 ml/well). After 6-24 hr, the cultures were labeled for 1 hr with 3H-thymidine (NEN, France, 80 Ci/mMol, 1pCi/ml), then washed and extracted with 5% TCA, and lysed overnight, at room temperature, with 0.5N NaOH. TCAinsoluble radioactivity was subsequently measured by scintillation counting, in Optiphase I1 scintillation fluid (Pharmacia).
Anchorage-Independant (AI) Growth A1 growth of TR15 was assayed by seeding lo3 or
TGF-/3Assays TGF-P activity in the media of NIH3T3 and TR15 cells was assayed by measuring the inhibition by these media of DNA synthesis (i.e., of incorporation of 3H-thymidine)in mink lung CC164 cells, as described by van Zoelen et al. (1986), by the stimulation of the formation of A1 colonies by rat NRK-49F fibroblasts, in the presence of EGF, as described by Jullien et al. (1988), and by the inhibition of binding of lz5ITGF-fl to its receptors on these fibroblasts (Massaguk and Like, 1985).
Acidification of Cell-Conditioned Media Cell-conditioned media were obtained by incubating confluent cultures 2 days in S- medium (DHM'). The crude media were centrifuged, to eliminate cells and debris, and acidified with HC1
268
BENZAKOUR et al.
to pH 2.0, then reneutralized with NaOH. Alternatively, they were concentrated under acidic conditions, by dialysis against 1 M acetic acid, followed by lyophilization. The lyophilized residue was dissolved and diluted in 4 mM HCl plus 1 mg/ml BSA and sterilized with 6oCo gamma-rays.
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Chromatography (FPLC) Reverse phase chromatography (FPLC) of the cell-conditioned TR15 medium, concentrated 20fold under acidic conditions, was carried out as described earlier (Pircher et al., 1984). The fractions collected from the acetonitrile gradient were lyophilized, redissolved in 4 mM HC, plus 1 mg/ml BSA, and sterilized with 6oCogammarays. Cytoplasmic RNA: Isolation and Slot Blot Analysis Cytoplasmic RNA was isolated as described by Dooley et al. (1989). For each experimental point, 10 p g of RNA were applied onto a nitrocellulose filter, using a minifold apparatus (Biorad), as described by Maniatis et al. (1989). The RNA amounts tested were initially measured by absorbance at 260 nm, and later more accurately by quantitating the 28s rRNA in each sample by migration on an agarose-formaldehyde gel. The slot blot filter was prehybridized and hybridized with the c-myc probe as described by Maniatis et al. (1989). The c-myc probe was an Alul/HaeIII fragment of 0.6 Kb, representing the third exon of c-myc and labelled by nick translation (BRL kit) at a specific activity of 10' cpmlpg.
capacity to grow in S- medium (MEF/S-: see Materials and Methods). 3H-Thymidine Incorporation Serum-deprived cultures of NIH3T3 on TR15 cells were treated for 16-24 hr with TGF-P1, or other growth factors (for comparison), then labeled for 1 hr with 3H-thymidine. As seen from Table 1, thymidine incorporation into cellular DNA in the absence of additives was 10-fold higher in TR15 cells than in NIH3T3 cells, presumably due to the fact that the former, but not the latter, could grow in the S- medium. The low basal level of incorporation in NIH3T3 cells was increased over 10-fold by TGF-PI, at 0.1 or lng/ml, and also by TGF-a and bFGF at the same concentrations, but only 2-fold by PDGF, at 1 ng/ml. TGF-fl also increased the high basal level of incorporation in TR15 cells (over 2-fold for 1 ng and 1.5-fold for 0.1 ng/ml), whereas no significant enhancement was seen with the other factors. Similar results were obtained in repeated experiments, which also showed that the incorporation observed with 1 ng/ml of TGF-PI was not increased by addition of any of the other factors, at the same concentration. We further studied the kinetics of thymidine incorporation in serum-deprived cells treated with 1 ng/ml of TGF-fl. As is seen from Fig. 1, in both cell types, incorporation was increased already after 12 hr in S- medium containing TGFPl. Thus, stimulation of thymidine incorporation by TGF-PI in NIH-3T3 and TR15 cells appears to TABLE 1 Incorporation of 'H-thymidine in Serum-deprived NIH3T3 and TR15 Cells: Action of TGF-fl and Other Growth Factors ~~
Additives
RESULTS
The capacity of TGF-Dl to stimulate mitogenically NIH3T3 and TR15 cells was first studied by assaying its action on the incorporation of 'H-thymidine into DNA of serum-deprived cells. Subsequently, its capacity to stimulate the growth of these cells was studied. For comparison, we also studied the action of TGF-PI on the growth of mouse embryo fibroblasts, at their third passage (MEF, P3) and following their spontaneous establishment and acquisition of the
None TGF-PI
1 .o
TGF-CI
0.1 1 .o 0.1
PDGF-AB bFGF
~
-
Concentration 'H-TdR incorporation (cprn/well (ng/ml) x l o-')
1.0 0.1 1 .o
0.1
NIH3T3 cells
TR15 cells
0.5 10.0 7.0 6.0 5.5
4.5 11.0 7.0 5.5 5.O 4.5
1 .o 0.7 10.5 8.0
5.0 5.0 6.0
(6
NIH3T3 cells, seeded a t 5xlO'/well, or TR15 cells, a t 10'/well 16 mm wells of Linbro plates), were switched arter 2 days to S DMEM, and labeled, 24 hr later for 1 hr with 'H-TdR. TCA-insoluble radioactivity was measured as described in Materials and Methods.
TGF-P STIMULATES MITOGENICALLY NIH3T3 FJBROBLASTS
269
3.5cm dishes, in DMEM plus 2% NCS (2A, 2B). Growth stimulation was a function of the concen15- A tration of TGF-PI, and already observed with r 0.01 ng/ml (Fig. 3A). In contrast, under the same conditions, no stimulation of growth was Q) observed with TR15 cells which grew much fas\ ter than the nontransformed parental cells (Fig. E 2C). TR15 cells, at 5x104/dish, also grew actively in S- medium (DHM') and were not stimulated either by TGF-fl (Fig. 2C). However, when they 6 12 24 6 12 were seeded at 104/dish, TGF-P1 stimulated Hours markedly their growth rate, at concentrations FIGURE 1. Kinetics of 3H-thymidine incorporation in serum above 0.01 ng/ml. A plateau of maximal stirnudeprived anchored NIH3T3 (A) and TR15 (B) cells, treated ng/ml Of TGF-fl, in s' with TGF-m and PDGF-AB. The factors were added. at lation was reached at 1 ng/ml, to replicate cultures switched to S' medium, as in medium, and 0.3 ng/ml, in-S- medium, and the Table 1, and the cultures labeled 6, 12 and 24 hr later. 0, cell number attained in the presence of TGF-fl controls; 0,TGF-PI; PDGF. was higher in S- then in S' medium, although it was 3-fold lower without TGF-P1 (Fig. 3 0 . The pursuit of this study confirmed that follow usual kinetics. In this experiment, with both cell types, no increase of incorporation at growth stimulation of TR15 in S- medium was 12 hr was seen after addition of 1 ng/ml of dose-dependent with a maximum stimulation PDGF-AB, and only a low increase at 24 hr, in corresponding to a 10-fold increase of the cell NIH3T3 cells. The low increase also observed, at number, relative to controls, at about 0.3 ng/ml, and a significant stimulation at 0.001 ng/ml (Fig. this time, in TR15 cells was not significant. 3D). A dose-dependent stimulation of growth in S- medium (DHM') was also observed with both Growth of Attached Cells lines of NIH3T3 cells, seeded in dishes coated As seen from Fig. 2, TGF-P, at 2ng/ml, stimu- with polylysine and fibronectin. In these conlated markedly the growth rate of attached ditions, the cells of both lines remained attached, NIH3T3 cells obtained from two different labora- but divided only once or twice in the absence of tories (NIH3T3-1 and -2, see Materials and TGF-PI, over a 10-day period, whereas their final Methods), and seeded at 5x104 cells/dish, in number was increased 6- to 12-fold, relative to
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
-
.,
Oeo3
FIGURE 2. Effect of TGF-fl on the anchored growth of NIH3T3 and TR15 cells, and unestablished (MEF, P3) and established (MEF/S-) mouse embryo fibroblasts. Cells were plated at 5x104/dish in 3.5 cm dishes, in DMEM+10% NCS (or 10% FCS, for MEF, P3) and switched one day later to DMEM+2% NCS (or 5% FCS, for MEF, P3), or to S- DHM', containing 2 ng/ml of TGF-PI. The medium was renewed as indicated by arrows and the cells counted, in duplicate cultures, 6 or 8 days after plating. Full symbols: cultures+ TGF-fl; empty symbols: controls. A, B: NIH3T3-1 and NIH3T3-2 cells, in DMEM+2% NCS; C: TR15 cells in DMEM+Z% NCS (0, .), or in DHM' (A, A); D MEF,P3 in DMEM+5% FCS (0, .), and MEF/S- in DMEM+Z% NCS (A, A), or in D H M (V, V).
-
Days
270
BENZAKOUR et al.
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
controls, in the presence of TGF-P1, at 0.3-1 ng/ml; i.e., to the same degree as the number of TR15 cells. However, the lowest TGF-PI concentration active on NIH3T3 cells was TO-fold higher (0.01 ng/ml) than that active on TR cells (Fig. 3B). TGF-P1 also stimulated, in a dose-dependent manner, the cloning efficiency of TR15 cells seeded in low number (2x103/dish), in nontreated dishes, in S' medium, and switched the next day to S- medium. The stimulation of formation of attached colonies was maximum at 0.1 ng/ml, and half the maximum increase was
observed at 0.01 ng/ml, the lowest concentration tested. Large colonies visible without a microscope, were also scored in the treated, but not in the control cultures (Fig. 4A). In contrast to its enhancing action on the growth of NIH3T3 and TR15 cells, TGF-P1 at 2 ng/ml, inhibited markedly (2-fold reduction of the final cell number) the growth of attached MEF,P3 in DMEM plus 5% FCS, and also inhibited significantly the growth of established MEF/S-, in DMEM plus 2% NCS, or in DHM' (Fig. 2D). The non-established MEF, P3 failed to grow in S- medium, even in MCM (see Materials
FIGURE 3. Effect of TGF-PI on the growth of anchored NIH3T3 and TR15 cells with and without serum. Same protocol as in Fig. 2, except that NIH3T3 cells switched to SDHM' were plated in dishes coated with polylysine and fibronectin, and TR cells were seeded at IO'/dish. A: NIH3T3-1 cells in DMEM+Z% NCS at 6 days; B: NIH3T3-1 cells (0) and NIH3T3-2 cells (A) in DHM' at 11 days; C: TRl5 cells in DMEM+Z% NCS (B) or DHM' ( 0 )at 8 days; D TR15 cells in DHM', at 11 days. N/No: final cell No/No of plated cells. Dotted line: N/No in untreated controls.
FIGURE 4. Effect of TGF-P, on the formation of attached colonies (cloning efficiency) and anchorageindependent (AI) colonies by TR15 cells. A: Cloning efficiency of 2x10' cells, plated in DMEM+I% NCS, in 3.5 cm dishes, and switched one day later to S- DHM'. Colonies of over 100 cells ( 0 ) and macroscopical colonies (W were scored at 11 days. B, C: Formation of A1 colonies by 10' TR15 cells in 16 mm wells of Linbro plates, in S- DHM' (B) or DHM+10% NCS (C). Colonies were scored at 14 days.
ov,
.1
I
1
1
0
I
L
.01 .1
1
10
0
1
I
1
.01 .1
ng/ml TGF-fll
1
10
TGF-/3 STIMULATES MITOGENICALLY NIH3T3 FIBROBLASTS
27 I
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
and Methods) and in dishes coated with polylysine and fibronectin, and were neither stimulated nor inhibited by TGF-fl, in these conditions.
inhibition of DNA synthesis in mink CC164 cells (Pironin et al., in preparation). The conditioned media were either acidified and reneutralized, or concentrated 5-fold, by dialysis against 1 M acetic acid, followed by lyophilization (see Growth of Unanchored TR15 Cells Materials and Methods), and assayed on replicate As is seen from Fig. 4 (B, C), TGF-fl also cultures. inhibited the formation of A1 colonies by TR15 As is seen from Table 2, the untreated crude cells, in semi-solid S' medium (DMEM plus 10% NIH3T3 medium did not increase thymidine NCS) or S- medium (DHM'). However, inhibition incorporation in NIH3T3 cells and the crude of colony formation was observed with TR15 medium displayed only a low enhancing 0.1 ng/ml of TGF-/?l in S- medium, against activity; whereas the acidified/neutralized media 1 ng/ml in S' medium. This suggests that serum and these media concentrated by dialysis against contains some factorb) counteracting the inhibi- acetic acid increased markedly incorporation in tory action of TGF-fl. This factor is unlikely to the target cells. No increase of the high level of be PDGF which did not revert the inhibitory basal incorporation in TR15 cells was observed action of TGF-fl, at a concentration of 1 ng/ml after addition of the crude or the acidified and reneutralized media; but a significant increase (data not shown). (over 2-fold) was seen after addition of the concentrated acidified media. Action on NIH3T3 and TR15 Cells of TGF-P When tested on unanchored TR15 cells, in Activated in Cell Conditioned Medium semi-solid DHM', the crude media stimulated Since TGF-P is present in a latent form, activat- significantly (over 2-fold) A1 colony formation, able by acid treatments, in the culture medium of due to the presence of cell-secreted factors able to nontransformed and transformed avian and stimulate this growth (Pironin et al., in preprodent fibroblastic cells (Lawrence et al., 1984; aration). Following acidification/neutralization, Pircher et al., 1984), the acidified medium of or concentration with acetic acid, the TR medium these cells, containing activated TGF-fl, should simultaneously lost its enhancing activity and stimulate mitogenically NIH3T3 and TR15 cells became inhibitory, with a maximum effect for and inhibit A1 growth of the latter cells. The experiments were carried out with the S- medium of confluent cultures of NIH3T3 and TR15 cells which contain no active TGF-P, detectable by TABLE 2 Action of Untreated and Acidified S- Conditioned Media on 'H-thymidine Incorporation in NIH3T3 and TR15 Cells and A1 Growth of TR15 Cells, in S- Medium Additives
'H-TdR incorporation (cpm/ wellxlo") NIH3T3 cells TRl5 cells
None TR med, crude id.AN id.,conc. 5xAC NIH3T3.med,crude id.AN id,conc. 5xAC
2.5 3.5(1.4) 12.5(5) 25(10) 2.5(1.0) 6(2.4) 41(16)
25 25(1.0) 25(1.O) 58Q.3) 25(1.O) 25(1.O) 60(2.4)
Percentage colony formation by TR15 cells 0.18 0.40(2.2) 0.16(0.9) 0.03(0.17) 0.45Q.5) 0280.6) 0.06(0.33)
Assays as for TGF-PI. The untreated (crude) or acidified/neutralized (AN) media, or these media concentrated under acidic conditions (AC). by dialysis against 1 M acetic acid and lyophilization, were added at 1 : l O (v/v) to replicate cultures of serum-deprived cells, for 'H-TdR incorporation, or to unanchored cells, in semi-solid medium, for A1 growth of TR15 cells. In parentheses, cpm or A1 colbnies in treated cultures: cpm or A1 colonies in controls.
FIGURE 5. Stimulation of 'H-thymidine incorporation in serum-deprived NIH3T3 cells by the fractions obtained by reverse phase chromatography (FPLC) of the S- medium of TR15 cells. One ml of TR medium, concentrated 20-fold under acidic conditions, was injected onto a reverse phase column, type Pro-RP 5/2 (Pharmacia). After 8 m n at 25% acetonitrile/O.l% TFA, a gradient of acetonitrile to 45% was applied over 32 mn. The fractions were assayed as in Table 1, and the active fractions of the second peak reassayed for TGF/3 activity (cf. Materials and Methods).
212
BENZAK.OUR er al.
lating thymidine incorporation in serumdeprived NIH3T3 cells eluted in the same fractions of the acetonitrile gradient as TGF-P activity, detected by inhibition of DNA synthesis in CC164 cells, or by the capacity to enhance A1 colony formation of NRK-49F cells, in the presence of EGF, or to compete for TGF-P receptors on the latter cells (see Materials and Methods). Figure 5 shows the elution profile of the mitogenic activity assayed on NIH3T3 cells. The additional early peak of activity presumably corresponds to PDGF activity which was detected in the TR medium (Pironin et al., in preparation).
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Mechanism of Action: Stimulation of c-myc Transcription The kinetics of thymidine incorporation in serum-deprived NIH3T3 and TR15 cells treated with TGF-fl suggested that it might stimulate their division by a direct mechanism. We, therefore, studied, in these cells, the kinetics of activation of the c-myc proto-oncogene, which is upregulated in early response to growth factors with a direct mitogenic action on quiescent cells (reviewed in Cole, 1986). As seen from Fig. 6, c-myc was up-regulated, in both cell lines, iH 2H 3H already 1 hr after addition of TGF-P1 (2 ng/ml). FIGURE 6. Induction of c-myc RNA by TGF-B. (a) NIH3T3 The level of c-myc RNA increased markedly at and TR15 cells, serum starved for 24 hr, were stimulated by 2 hr and was then 2-fold higher in TR15 cells than TGF-Pl (2 ng/ml) for the indicated times before preparation of cytoplasmic RNA, slot blot analysis, and hybridization with a in NIH3T3 cells, as indicated by densitometric c-myc probe (see Materials and Methods). (b) Densitometric scanning analysis. At 12 hr, it had declined 30scanning values of the slot blot of treated cells expressed as the fold in NIH3T3 cells, but only 5-fold in TR15 percentage of c-myc RNA at 2hr in the TR15 cells. No induction was observed in the untreated controls, which are cells. not presented.
DISCUSSION the acidified concentrate. Acidification/ neutralization of the NIH3T3 medium reduced its enhancing activity without suppressing it, but the acidified concentrate was highly inhibitory. Hence, the acidified media mimicked the actions of TGF-PI on NIH3T3 and TR15 cells. In line with this finding, DNA synthesis in mink CC164 cells was strongly inhibited by the acidified concentrates, somewhat less by the non concentrated acidified/neutralized TR medium, and much less by the non concentrated acidified/neutralized NIH3T3 medium (data not shown). Moreover, following reverse phase chromatography (FPLC) of the TR medium, concentrated 20-fold with acetic acid, the activity stimu-
Our results show that TGF-P is an effective mitogen for NIH3T3 fibroblasts and these cells transformed by EJ-H-ras, in S' and S- medium. This activity is demonstrated by its capacity to stimulate DNA synthesis, in serum-deprived attached cells, and the growth rate of these cells, in S' medium and S- medium containing insulin, transferrin and sodium selenite. In serumdeprived NIH3T3 cells, TGF-P1, at 0.1 ng/ml, increases markedly thymidine incorporation and is as active as TGF-a or bFGF, and much more active than PDGF, at the same concentration. It also increases, at only 0.01 ng/ml, the growth rate of the attached cells, in low serum medium,
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
TGF-B STIMULATES MITOGENICALLY NIH3T3 FIBROBLASTS
or in S- medium, in dishes coated with polylysine and fibronectin; and a marked increase is observed at 1-2 ng/ml, the highest concentrations tested. It should be noted that the two sublines of NIH3T3 cells used in our experiments were able to divide once or twice in S- medium, in the precoated dishes, contrary to the earlier report of Zhan and Goldfarb (1986). In any case, TGF-PI appears to cooperate with serum growth factors and factors in the S- medium to stimulate their growth. TGF-fl also stimulates DNA synthesis of attached serum-deprived TR15 cells, and their growth in S' and S- medium. DNA synthesis in confluent cultures is stimulated to a much lower degree than in NIH3T3 cells, presumably due to the fact that it is 10-fold higher in the untreated serum-deprived TR cells. This high basal level appears to be due to the secretion by the TR cells of autocrinally active mitogenic factors related to TGF-a, PDGF, and FGF and, apparently, other unrelated factors (Pironin et al., in preparation). Therefore, TGF-P presumably acts in cooperation with these factors. On the other hand, it appears to be more active on serum-deprived TR cells than exogenous TGF-a, PDGF or bFGF, which do not increase the basal level of thymidine incorporation, at the same concentration. In attached cultures of TR15 cells plated at high density (5x104/3.5cm dish), the growth rate, in low serum medium or S- medium, is not increased by TGF-P, at 2ng/ml, whereas it is increased significantly in cultures plated at lower density, at concentrations comparable to those active on NIH3T3 cells. Moreover, in the less dense cultures, in S- medium, the lowest active concentration of TGF-PI is 10-fold lower than the lowest active concentration for NIH3T3 cells, i.e. only 1 picogram/ml. These findings suggest that, in the dense cultures, TGF-fl, at the concentrations tested, does not enhance the already high mitogenic activity of the autocrine factors, whereas it is able to enhance it, in the less dense cultures, in coooperation with these factors. The capacity of TR cells to respond to a very low concentration of TGF-P1 is presumably due to this cooperation. However, TR cells may also be more sensitive to TGF-P than NIH3T3 cells. Along this line, we have further checked that the responsiveness of the TR15 cells to TGF-PI, in Smedium, is as high, or higher than that of mink lung CC164 cells, assayed by inhibition of DNA
273
synthesis (van Zoelen et al., 1986). Thus, the stimulation of growth of TR15 cells, in Smedium, appears to constitute a new sensitive assay for TGF-P. This assay may be used to detect activated TGF-P in acidified cell-conditioned medium, as shown by our experiments with the acidified NIH3T3 and TR media. However, other factors in these media might modulate the TGF-/3 activity. It should be recalled here that only the acidified media contained active TGF-P. Thus, the autocrine activity of the untreated TR cell medium must be due to other factors. Regarding the mechanism of the mitogenic stimulation of NIH3T3 and TR cells by TGF-PI, the question arises as to whether it initiates directly the cell cycle, as do other bona fide growth factors, or acts indirectly by the intermediary of activated PDGF, as observed for murine AKR-2B fibroblasts by Leof et al. (1986). As recalled in the introduction, these authors observed a 12 hr prolongation of the G1 period of the cell cycle, following stimulation by TGF-P, which presumably corresponds to the time required for activation of the c-sis gene and "indirect" autocrine stimulation by the PDGF encoded by this gene. However, this indirect, mechanism is ruled out in the case of NIH3T3 and TR15 cells, since normal kinetics of thymidine incorporation are observed in these cells, stimulated by TGF-P, together with an early upregulation of the c-myc proto-oncogene. This is, to our knowledge, the first observation showing that TGF-fl may act as a direct mitogen. As regards the higher level of transcription and of persistance of c-myc RNA in TR cells, relative to NIH3T3 cells, it is in line with the responsiveness of the former cells to a lower concentration of TGF-fl, and might, therefore, also be due to the cooperative action of the autocrine factors released by the TR cells. Since, in addition, the growth of anchored fibroblastic cells is generally inhibited by TGF-P (see Introduction), the apparently direct mitogenic response of NIH3T3 and TR cells to TGF-fl seems exceptional and remains to be explained. One of the events involved in the acquisition of the responsiveness may be immortalization. However, the fact that immortalized NC-Z embryo fibroblasts are inhibited mitogenically by TGF-P, like their early passage ancestors, shows it is insufficient. In contrast to its mitogenic action on anchored
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
214
BENZAKOUR et al.
NIH3T3 and TR cells, TGF-PI inhibits the A1 growth of the latter cells, in S' and S- media and induces a cell response similar to that generally observed with other transformed cells, including other fibroblastic cells transformed by rus oncogenes (see Introduction). As concerns the inhibition of A1 growth observed in S' medium, it may be noted that TGF-fl was previously shown to inhibit the A1 growth of nontransformed NIH3T3 cells in this medium (Anzano et al., 1986). Hence, the loss of anchorage appears to reverse the action of TGF-p on both nontransformed and transformed NIH3T3 cells. This reversion constitutes a new example of the bifunctional activity of this TGF, depending on the conditions of culture and environmental factors (Roberts et al., 1985; Sporn and Roberts, 1988; Roberts and Sporn, 19901, and only hypothetical models may be proposed to explain it. One possibility is that, in the absence of anchorage, genes required for the mitogenic response to TGF-/3 may cease to be expressed, allowing reversion to antiproliferative responsiveness. The unanchored cells may also cease to produce some factodd which synergizes with TGF-P to induce the mitogenic response. Further studies are required to test these, and other possible models of action of TGF-P, and to precise the mechanism of its direct mitogenic action. ACKNOWLEDGMENTS This work was s u p p o r t e d by grants from la Ligue Nationale Frangaise contre le Cancer and from t h e INSERM (Contract no 884015).W e thank Drs G. Goubin and G. Calothy, i n our Institute, for kindly supplying respectively, the pEJ6.6 plasmid and the c-myc probe. W e are also indebted t o Mrs A. Besnard for her skillful secretarial assistance.
REFERENCES Anzano, M. A., Roberts, A. B., Smith, J. M., Sporn, M. B. and De Larco, J. E. (1983) Sarcoma growth factor from conditioned medium of virally transformed cells is composed of both type a and type p transforming growth factors. Proc. Natl. Acad. Sci. U S A 80, 6264-6268. Anzano, M., Roberts, A. 8. and Sporn, M. B. (1986) Anchorage-independent growth of primary rat embryo cells is induced by platelet-derived growth factor and inhibited by type-beta transforming growth factor. ]. Cell. Physiol. 126, 312-318. Barbacid, M. A. (1987) ras genes. Ann. Rev. Biochem. 56, 779-827.
Barnard, J. A,, Lyons, R. M. and Moses, H. L. (1990) The cell biology of transforming growth factor p. Biochem. Biophys. Acta 1032,7947. Cole, M. D. (1986) The myc oncogene: its role in transformation and differentiation. Ann. Rev. Genet. 20,361-384. Copeland, N. G. and Cooper, G. M. (1979) Transfection by exogenous and endogenous murine retrovirus DNAs. Cell 16,347-356. Dooley, S., Ehrhart, E., Radtke, J., Unteregger, G. and Blin, N. A. (1989) A procedure for simultaneous isolation of undegraded RNA, DNA and nuclear proteins suitable for interaction analysis in homologous systems. Methods in Mol. Cell. Biol. 1,95-105 Graham, F. L. and van der Eb, A. J. (1973) A new technique for the assay of human adenovirus 5 DNA. Virology 52, 4-67. Jullien, P., Berg, T. M., De Lannoy, C. and Lawrence, D. A. (1988) Bifunctional activity of transforming growth factor type /3 on the growth of NRK-49F cells, normal and transformed by Kirsten murine sarcoma virus. ]. Cell. Physiol. 136,175-181. Lawrence, D. A., Pircher, R., Kryceve-Martinerie, C. and Jullien, P. (1984) Normal embryo fibroblasts release transforming growth factors in a latent form. ]. Cell. Physiol. 121, 184-188. Leof, E. B., Proper, J. A., Goustin, A. S., Shipley, G. D., DiCorleto, P. E. and Moses, H. L. (1986) Induction of c-sis mRNA and activity similar to platelet-derived growth factor by transforming growth factor p: a proposed model for indirect mitogenesis involving autocrine activity. PTOC. Natl. Acad. Sci. U S A 83,2453-2457. Li Wen Xin, Jullien, P., Lawrence, D. A,, Pironin, M. and Vigier, P. (1987). Chemically and virally transformed cells able to grow without anchorage in serum-free medium: evidence for an autocrine growth factor. ]. Cell Physiol. 131, 175-1 83. Maniatis, T., Fritsch, E. F. and Sambrook, J. (1989) Molecular Cloning: A Lnboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Massaguk, J. (1990) The transforming growth factor-p family, Ann. Rev. Cell. Biol. 6, 587-641. Massagu6, J. and Like, B. (1985) Cellular receptors for type-P transforming growth factor: ligand binding and affinity labeling in human and rodent cell lines. ]. Biol. Chem. 260, 2636-2645. McClure, D. B. (1983) Anchorage-independent colony formation of SV40 transformed BALB/c-3T3 cells in serum-free medium: role of cell- and serum-derived factors. Cell 32, 999-1006. Pircher, R., Lawrence, D. A. and Jullien, P. (1984) Latent /% Transforming Growth Factor in non-transformed and Kirsten sarcoma virus-transformed Normal Rat Kidney cells, clone 49F. Cancer Res. 44,5538-5543. Roberts, A. B., Anzano, M. A,, Wakefield, L. M., Roche, N. S., Stem, D. F. and Sporn, M. B. (1985) Type p transforming growth factor. A bifunctional regulator of cellular growth. Proc. Natl. Acad. Sci. U S A 82,119-123. Roberts, A. 8. and Sporn, M. B. (1990) The transforming growth factor-beta. In Handbook of Experimental Pharmacology. 1. Peptide Growth Factors and Their Receptors, (M. B. Sporn and A. B. Roberts, eds), Springer-Verlag, Berlin, pp. 419-472. Shih, C. and Weinberg, R. A. (1982) Isolation of transforming sequence from a human bladder carcinoma cell line. Cell 29, 161-169. Shipley, G. D., Tucker, R. F. and Moses, H. L. (1985) Type p transforming growth factor/growth inhibitor stimulates entry of monolayer cultures of AKR-2B cells into S phase afte; a prolong& prereplicative interval. Proc. Natl. Acad. Sci. USA 82,4147-4151.
TGF-/3 STIMULATES MITOGENICALLY NEI3T3 FIBROBLASTS
Growth Factors Downloaded from informahealthcare.com by McMaster University on 02/12/15 For personal use only.
Sporn, M. 8. and Roberts, A. B. (1988) Peptide growth factors are multi-functional Nature 332,217-219. Tucker, R. F., Volkenant, M. E., Branum, E. i. and Moses, H. L. (1983) Comparison of intra- and extracellular transforming growth factors from nontransformed and chemicallly transformed mouse embryo cells. Cancer Res. 43, 1581-1 586. Van Zoelen, E. J. J., Van Oostwaard, T. M. J. and De Laat,
215
S. W. (1986) PDGF-like growth factor induces EGFpotentiated phenotypic transformation of normal rat kidney cells in the absense of TGF-P. Biochem. Biophys. Res. Commun. 141,1229-1235. Zhan, X. and Goldfarb, M. (1986) Growth factor requirements of oncogene-transformed NIH3T3 and BALB/c-3T3 cells cultured in defined media. Mol. Cell. Biol. 6,3541-3544.
