The antiproliferative effect of interferon and the mitogenic activity of growth factors are independent cell cycle events

Experimental

Cell Research 161 (1985) 297-306

The Antiproliferative Effect of Interferon and the Mitogenic Activity of Growth Factors Are Independent Cell Cycle Events Studies

with

Vascular

Smooth

Muscle

Cells and Endothelial

ANTHON du P. HEYNS,’ AMIRAM ELDOR,‘, NURIT KAISER,3 RAFAEL FRIDMAN2

Cells

* ISRAEL VLODAVSKY and AMOS PANET3,4

‘Department of Hematology, 2Department of Radiation and Clinical Oncology, University Hospital, ‘Department of Endocrinology and 4Department of Virology, University-Hadassah Medical School, Jerusalem, Israel 91120

,*

Hadassah the Hebrew

We studied the antagonistic effects of interferon (IFN) and growth factors in GO/Glarrested normal bovine aortic smooth muscle cells @MC) which were stimulated by serum, or purified platelet derived growth factor (PDGF), supplemented with plasma-derived serum (PDS). The growth response, measured as [3H]thymidine incorporation into DNA, was dependent on the concentration of the mitogen. Human IFNa, recombinant human lFNa2, or a crude bovine-IFN preparation prepared from virus-infected bovine aortic endothelial cells, inhibited SMC growth induced by either serum or PDGF with PDS. The extent of IFN inhibition was inversely related to the concentration of the mitogenic stimulus. We also investigated whether IFN inhibited the early events in Gl phase, stimulated by the competence factor PDGF, or the progression of the cell into the S phase induced by PDS. The results indicated that IFN inhibited these two stages of the Gl phase independently. In addition, we investigated the antiproliferative effect of IFN on bovine aortic endothelial cells (BAEC), which do not respond to PDGF but to the mitogenic activity of fibroblast growth factor (FGF). IFN inhibited the mitogenic activity of FGF in a dose-dependent manner. The results indicate that the anti-proliferative activity of IFN and the mitogenic effects of different growth factors are independent. @ 1985 Academic RUSS, IX.

Animal cells require growth factors to sustain proliferation in tissue culture. Serum provides a variety of such regulatory factors. Platelet-derived growth factor (PDGF), one of the most important of the serum factors, is unique in that it is not normally present in plasma, but located within the a-granules of platelets. PDGF becomes available for smooth muscle cell (SMC) proliferation in vivo only after its release from adherent platelets in areas of exposed subendothelium [l]. Components present in plasma-derived serum (PDS), are necessary for optimal growth response to PDGF 12, 31. Some of these factors belong to the family of polypeptide hormones termed somatomedins. The role of somatomedins, including insulin, in regulating the growth of cells in culture has been studied extensively [4,5]. It has been suggested that they do not control the same events in the cell cycle as does PDGF. According to this view, PDGF initiates cell replication by * To whom offprint request should be addressed. Copyright @ 1985 by Academic Press, Inc. All rights of reproduction in any form resewed 0014-4827/85 $03.00

298 Heyns et al.

making the cells competent to progression growth factors. Somatomedins and other growth factors present in PDS, which lacks PDGF, allow the competent cells to progress through GO/G1 and enter the S phase of the cell cycle [4, 51. Endothelial cells line the inner surface of blood vessels and form a protective barrier between the circulating blood and the SMC which constitute the media of the vessel wall [6]. While proliferation of SMC is mediated by PDGF, endothelial cells respond to the mitogenic activities of growth factors, such as fibroblast growth factor (FGF) and endothelial cell growth factor (ECGF) [6, 71. Little is known about how the growth of cells in the vessel wall, induced by various growth factors, is modulated (reviewed by Schwartz et al. [6]), the blood vessel wall may be damaged [8] and exposed to circulating interferons (IFN) induced by viruses as well as other agents [9]. Since IFNs are known to inhibit cell growth, they may act as mitogen antagonists. Indeed, it has been suggested that IFN and growth factors are two families of hormones with antagonistic action [lo]. Studies on the antiproliferative activity of IFN were performed almost exclusively with transformed cell lines [lo]. Since transformed cells may produce autonomous growth factors, or are not regulated like normal cells by exogenous mitogens, they may not constitute a good model for these studies. In the present study we investigated the effect of IFN on the proliferation of cultured vascular SMC and endothelial cells. These two cells maintain the integrity of the vessel wall and are involved in the pathogenesis of atherosclerosis [I]. We also tested the hypothesis that PDGF and IFN are hormones with competitive antagonistic action. METHODS Materials Dulbecco’s modified Eagle medium, H-16 (DMEM), penicillin and streptomycin were obtained from Gibco Grand Island, N.Y., newborn calf serum was supplied by Bio-Lab Laboratories, Ierusalem; human leukocyte IFN-a (5~10~ U/mg protein) was a gift from Dr T. Bino, the biological Institute, Ness Ziona, and was prepared as described before [ll]. Recombinant human IFNaz (10’ U/m& produced in E. co/i and purified to apparent homogeneity, was a gift from Dr D. Novick, the Weizmann Institute, Rehovot. IFN units were defined by protection of bovine MDBK cells from vascular stomatitis vitus’(VSV) infection [20]. The two HuIFNa preparations contained no mitogenic activity when tested on mouse NIHi3T3 cultures arrested at the GO/G1 phase. Newcastle disease virus (NDV) was grown in fertile eggs; VSV was propagated and titrated in mouse L cells. Purified porcine PDGF was supplied by Speywood Laboratories, Nottingham, England. The preparation had a specific activity of 0.94 Ulug protein. A unit of activity was defined as the amount of PDGF required to stimulate [“Hlthymidine incorporation by confluent 3T3 cells equal to half that in medium containing 10% fetal calf serum (FCS). A highly purified PDGF [12] preparation (240 U/ug protein), a kind gift from Dr L. Witte, Columbia University, New York, was used in some experiments. SMC viability was not diminished after treatment with IFN, or PDGF as determined by trypan blue exclusion and neutral red uptake [13]. Fibroblast growth factor (FGF) was purified as described previously from bovine brains [14]. Brain FGF yielded a single band on PAGE at pH 4.5 and on an isoelectric focusing column (pH range 3.4 to 11.0). [Methyl-3H]thymidine, 5000 mCi/mmol, was supplied by Nuclear Research Center, Beersheba, Israel. All other reagents were of analytical grade and supplied by commercial sources. Plasma-derived serum (PDS) was prepared from platelet-poor human plasma as described previously [15, 161. PDS (5 %), in the absence of exogenous PDGF failed to stimulate the proliferation of SMC cultures. Exp Cell

Res 161 (1985)

Antagonistic

effects of interferon

and growth factors

299

Vascular Smooth Muscle Cells After removal of the intima by scraping with a surgical blade, the media of the bovine aortic arch was dissected with a scalpel and fragments were cultured as described [ 171. Cultures, propagated in DMEM medium containing 10 % calf serum and antibiotics were composed of small, spindle-shaped and overlapping cells, which by EM exhibited numerous myofilament bundles and vesicles near the cell surface. Staining for factor VIII antigen was negative. Stock cultures were passaged every 5-7 days and the experiments conducted with subcultures of passages 2-8.

Vascular Endothelial

Cells

Cloned populations of bovine aortic endothelial cells were obtained from the adult aortic arch, as previously described [18, 191. Cells were cultured in DMEM supplemented with 10% calf serum and antibiotics (50 U/ml penicillin and 50 &ml streptomycin). Once a week the cells were dissociated with 0.05 % trypsin and 0.02 % Versene in phosphate-buffered saline and subcultured at a split ratio of 1: 20. FGF (100 &ml) was added at alternate days during the phase of active growth, and confluent cultures (1 100 cells/mm2) were maintained without further addition of FGF. The cells formed a closely apposed non-overlapping and contact-inhibited cell monolayer. The presence of factor VIII antigen, production of prostacyclin, secretion of an underlying basal lamina and expression of a nonthrombogenic apical surface have been constant features of all the subcultures of vascular endothelial cells. Both the vascular endothelial and smooth muscle cell cultures were maintained in 10% CO2 humidified incubators at 37°C.

Cell Cycle Studies A system was used in which SMC were arrested in GO/G1 of the cell cycle by depriving the cultures of serum [16]. Cells from confluent stock cultures were detached from the plates by treatment with 0.05 % trypsin, 0.02 % Versene solution and resuspended in DMEM containing 0.2 % calf serum. SMC (4-6x 10’ cells in 0.5 ml) were seeded in ldmm multiwell plates. Cells became quiescent after 48 h. GO/Gl-arrested cells received a mitogenic stimulus consisting of either calf serum or purified PDGF supplemented with PDS (5 % V/V). The arrest of SMC at GO/G1 phase of the cell cycle was evident from the incorporation of [3H]thymidine (5% of that in growing cells). Moreover, after stimulation with serum, DNA synthesis started at 12 h and reached a maximum at 18 h. IFN was added as indicated in the captions to the figures. 13H]thymidine, 1.0 @/well, was always added with the serum, or with the PDS. Sixteen hours after addition of [‘Hlthymidine, DNA synthesis was assayed by measuring the incorporation of [‘Hlthymidine into trichloroacetic acid (JCA)-insoluble material as previously reported [17]. The final volume of medium in the well was 0.6 ml, the balance was made up with DMEM supplemented with 0.2 % calf serum. Concentrations of all additives are expressed per ml. Results were the means of triplicate wells. Standard deviations were less than 10%. DNA synthesis in the control stimulated cultures (100% [3H]thymidine incorporation) was at the range of 100000-200000 cpm/ld cells, in these experiments. BAEC (2X ld cells) were plated into each l&run tissue culture well in DMEM containing 10% calf serum. Under these conditions, cell division was non-existent or very slow [14]. Cells were stimulated by adding FGF directly to the culture medium at day 2 after seeding. The cells were allowed to proliferate in the presence of FGF for 3 days. r3H]Thymidine (1 @i/well) was added at day 5 for a 2-h pulse-labelling and the cultures tested for the amount of DNA incorporated radioactivity.

IFN Induction

in Endothelial

Cells

As reported previously [20], confluent BAEC monolayers in 16mm wells were infected with Newcastle Disease Virus (NDV, 1000 hemagglutination/ml) and the cultures incubated for 24 h. The supematants were removed, subjected to acid treatment (pH 2.5) with HCl for 1 h, and neutralized with NaOH. The acid treatment did not affect the bovine IFN (boIFN) titers, but inactivated the virus [20]. This was evident from the lack of cytopathic effect after exposure of bovine endothelial cells to the boIFN-containing supematants. IFN activity in the supematants was titrated with a microassay which measures the extent of protection from the cytopathic effect of VSV in bovine aortic endothelial cells [20, 221. The boIFN activity was compared with that of human IFNa, and the boIFN titers were expressed as ,laboratory units [20]. The antiviral substance produced by the bovine endothelial Exp Ceil Res 161 (1985)

300 Heyns et al.

;lrn B d g

75

E

25

Fig. 1. DNA synthesis in synchronow SMC cultures stimulated by (A) serum; (II) purified PDGF. Incorporation of [3H]thymidine into DNA was measured 16 h after the addition of the mitogen to WGlarrested SMC.

2 7 L 50 2 % 9

I:--0

a5 PDGF

10 iunilr

15 /ml

20 1

cells was characterized by standard treatments (nucleases, trypsin, thermal inactivation, high speed centrifugation) according to Lockart [21]. Results indicated that the antiviral effect in the supematant conforms to that of IFN [20].

RESULTS Dose Response of SMC to the Mitogenic

Activity

of Serum and PDGF

Cultures of SMC were arrested at the GO/G1 phase of the cell cycle by exposure to 0.2% calf serum for 48 h [16]. The mitogenic activities of serum and purified PDGF were compared by measuring the incorporation of r3H]thymidine into DNA (S phase), 16 h after addition of the mitogen. A typical dose-response curve to serum is shown in fig. 1A. SMC proliferation was dependent on the serum concentration. In repeated experiments, the addition of 2.5% serum produced 50% of the maximum stimulating effect. Quiescent SMC do not proliferate significantly if stimulated only by purified PDGF, but require somatomedins and other plasma factors for the orderly progression of the cell cycle into the S phase [2, 31. Initial studies were carried out to determine the effect of various concentrations of PDS on the growth-promoting activity of purified PDGF. The optimal PDS concentration was 2.5-5 %, to stimulate proliferation in the presence of PDGF (2 U/ml). A typical dose-response curve of quiescent SMC, stimulated with various concentrations of PDGF in the presence of 2.5 % PDS, is shown in fig. 1 B. At a concentration of 0.45 U/ml of PDGF the incorporation of [3H]thymidine into DNA was 50% of the maximum. This was equivalent to the response observed with 2.5 % serum (compare fig. 1 A, B). The dose-response curves were linear at low concentrations of the mitogens. SMC Responses to hulFN Human IFNs, particularly those of the a-type, act very effectively as antiviral agents on cells from a variety of species. This may not always be paralleled by an effect on cell growth. In fact, it has been reported that some bovine cell lines were sensitive to the antiviral, but not the antiproliferative effect on huIFNa [231. The protective effect of huIFNa on bovine SMC infected with VSV was tested as described previously [20, 221. Preincubation with ten units of huIFNa protected 50% of the SMC against the cytopathic effect of the virus. Subsequently we Erp CellRes

161 (1985)

Antagonistic 100 t .; f z. 75

*

effects of interferon

and growth factors

301

IFN lulmtl ,,.--

_ ---.-..

0

I ;-- 50 8 I -25 h 0

,’0,” -0 1250 ; #’ ’ e- -.-•- - - - .~*ooo i ,a” ,’ ’ 0 j/. ’ 2 L 6 8 10 Serum Concentration t % 1

Fig. 2. Effect of huIFNa on SMC stimulated to proliferate by serum. huIFNa (0, 1250 U/ml; Cl,2000 U/ml) and serum were added together and DNA synthesis was measured 16 h later (see Methods for details). Fig. 3. Effect of crude boIFN on the proliferation of SMC. Supematant containing boIFN, 100 units/ml (0), and serum were added together and [‘Hlthymidine incorporation into DNA measured 16 h later.

examined the antiproliferative effect of huIFNa on bovine SMC. Quiescent SMC were stimulated to proliferate synchronously by adding increasing concentrations of serum in the presence of 1250 or 2000 U/ml huIFNa (fig. 2). The huIFNaevoked inhibition of SMC entry to the S phase of the cell cycle was concentration-dependent. The anti-proliferative activity of huIFNa was not associated with the cytotoxic effect, since the number of viable cells determined by both trypan blue dye exclusion and neutral red dye uptake did not decrease after incubation with up to 10000 units of IFN. Moreover, the inhibition of SMC proliferation was reversible upon withdrawal of the IFN (data not shown). The antiproliferative effect of a crude preparation of boIFN was compared with that of huIFNa. Cultured bovine vascular endothelial cells were infected with NDV, and the medium was collected and inactivated by acid pH. This supematant contained 1000 laboratory U/ml of boIFN activity [20]. Quiescent bovine SMC were stimulated by increasing concentrations of calf serum, with and without the addition of 100 laboratory units/ml of boIFN. The results depicted in fig. 3 demonstrate that this supematant containing boIFN effectively inhibits SMC growth in a pattern similar to that of purified huIFNa (compare figs 2,3). It should be noted that the crude boIFN preparation, like huIFNa, had no cytotoxic effect on SMC. To understand the relationships between growth factors and IFN, the antiproliferative activity on SMC stimulated with purified PDGF was studied. In the first experiment (fig. 4A), quiescent SMC were stimulated with suboptimal concentrations (0.4 or 0.5 U/ml) of PDGF (see fig. 1B), supplemented with 2.5 % PDS. Various concentrations of huIFN were added with the mitogens, and [3H]thymidine incorporation into DNA was measured 16 h later. The results depicted in fig. 4A show that huIFNa-inhibition of SMC proliferation is concentration-dependent. The half-maximal effective concentration (EC& of IFN is about 125 U/ml for cells stimulated by 0.4 U/ml. PDGF, and 150 U/ml for cells 20-858342

Erp Cell Res 161 (1985)

302 Heyns et al.

i”‘-

Fig. 4. Effect of huIFNa on quiescent SMC stimulated for growth with PDGF and PDS. PDGF, PDS and huIFNa were added at the same time and [%]thymidine incorporation into DNA measured 16 h later. (A) huIFNa ;“,zputitiedPDGF(0 05U/ml 0, 0.4 U/ml). (B) Recombinan; huIFNaj (X); huIFNa (Cl) and highly purified PDGF

‘gg

0

500 tow IFN

I units

,mt

15ca I

2mw

0

500 IFN

Iwo I unitslmt

1500

2000 )

exposed to 0.5 U/ml PDGF. Concentrations of huIFNa higher than about 500 U/ml did not have a greater antiproliferative effect on cells stimulated with PDGF. The degree of inhibition by IFN is inversely related to the intensity of the mitogenic stimulus. It may be argued that contaminating proteins in the partially purified preparations of huIFNa and PDGF influence the apparent antagonistic effects of these two hormones. To rule out this possibility, the activity of pure huIFNa,, produced in recombinant E. c&i, was investigated in SMC stimulated with highly purified PDGF (fig. 4B). It is evident that huIFNa, (10’ units/mg protein) and huIFNa (sp. act. 5x lo6 units/mg protein) were similarly effective in inhibiting the mitogenic stimulus of a highly purified PDGF preparation. From the literature it is not clear whether PDGF and IFN act on the same event in the cell cycle [lo, 271. The following experiments were designed to clarify this question. GO/G1 SMC cultures were stimulated by PDGF, 0.4 U/ml. The interactions between huIFN and PDGF and their effect on cell proliferation were studied in three types of experiment (fig. 5). In the first experiment (fig. 5A) huIFNa at concentrations ranging from 20 to 2 000 U/ml was added together with PDGF and PDS. There was a dose-dependent inhibition of [3H]thymidine incorporation into DNA as measured 16 h later. In the second experiment (fig. 5 B) PDGF was added and allowed to remain in contact with the SMC for 2 h to induce competence. PDGF was then removed by washing, and by brief exposure to 2-p mercaptoethanol (28 mM) as described previously by Pledger et al. [2]. The culture medium was replaced by DMEM containing PDS with the appropriate concentration of huIFNa and [3H]thymidine. The incorporation of [3H]thymidine into DNA was measured 16 h later. As may be seen in fig. 5 curve B, the inhibition by IFN is similar to that observed when huIFN and PDGF are added simultaneously (fig. 5, curve A). In a third experiment (fig. 5, curve C), the time interval between transient exposure of SMC to PDGF and the introduction of huIFNa and PDS was prolonged to 8 h. Quiescent SMC were stimulated by 0.4 U/ml PDGF and 2 h later the exogenous PDGF was washed away and the cells were briefly exposed to 2-mercaptoethanol (28 mM). Following this treatment, the medium was replaced with DMEM containing 0.2% calf serum. After 8 h PDS (2.5 %), the appropriate concentration of huIFNa, and r3H]thymidine were added to the Erp Cell Res 161 (1985)

Antagonistic 3

30

-100 .-8 3 i

effects

5

6 +

and growth

factors

'15, 5.

4

.,;; .-------a-.-.-.-

-----.A .-.-.___._.

500 IFN

1000 15cQ I units/ml 1

q c

i E P c; ,=

303

IFN (Ultil

r6 ,,,'

B 2 .z

of interferon

m,' ,'

,a

,'

,*O

_' ,'

2*-

25 F‘-Ol 2000

0

10

20 30 FGF ,ng,m,,

40

50

Fig. 5. Interaction between huIFNa and PDGF on events in the cell cycle. Quiescent SMC were stimulated for growth by PDGF, 0.4 U/ml. Curve (A): huIFNa was added with the PDGF and PDS at zero time, and [‘Hlthymidine incorporation measured 16 h later (0). Curves B, CZ Two hours after addition of PDGF to the cultures it was removed. The remaining PDGF was inactivated by washing the cultures once with medium containing 28 mM 2-/?-mercaptoethanol, and once with DMEM; this treatment had no apparent toxic effect on the cultures [2]. In curve B PDS and huIFNa were added after removal of PDGF, and DNA synthesis measured 16 h later (0). In curve C PD.3 and huIFNa were added 8 h after removal of PDGF, and DNA synthesis measured 16 h later (0). Fig. 6. Effect of huIFNa on BAEC stimulated to proliferate by FGF. Two ~10~ endothelial cells were seeded in 1 ml DMEM containing 10% calf serum into each of 16 mm tissue culture wells. Various amounts of FGF were added on 2 days after seeding, without (0) or together with 100 U (0), or 1000 U (Cl) of huIFNa. Five days after seeding [‘H]thymidine was added (1 @i/well), and its incorporation into acid-insoluble material was measured after 2 h pulse.

cultures and the incorporation of radioactivity into DNA was measured 16 h later. The SMC, transiently stimulated with PDGF, remained competent to progress to orderly DNA synthesis, even though the PDS was added 8 h later. As may be seen in fig. 5, curve C, huIFNa added to SMC made competent 8 h earlier by PDGF, induced its antiproliferative effect. The inhibition was of the same magnitude as that observed in the experiments in which PDGF and IFN were added simultaneously or within 2 h of each other. In experiments B and C, PDGF was removed before the addition of IFN. Nevertheless, we found in a control experiment that even after extensive washings in medium containing 2-mercaptoethanol, some residual PDGF mitogenic activity was left in the tissue culture dishes (data not shown). This observation is in agreement with the report of Dicker & Rozengurt on the difficulties of inactivation of PDGF in tissue culture [39]. However, since the antiproliferative activity was very similar whether IFN was added together with PDGF or up to 8 h after its removal, the two proteins do not appear to compete on the same functional site. BAEC Responses to huIFN

Vascular endothelial cells do not respond for active proliferation they require FGF in These cells present therefore a unique model proliferative activity of IFN. The induction

to the mitogenic stimulus of PDGF; addition to some serum factors [6]. to study the specificity of the antiof endothelial cell proliferation by Exp

Cell Res

161 (1985)

304 Heyns et al. FGF was tested in the absence and presence of IFN (fig. 6). For this purpose FGF and huIFNa were added together to low density cultures of bovine aortic endothelial cells maintained in the presence of 10% calf serum. Three days later cell proliferation was estimated by pulse-labelling for 2 h with [3H]thymidine. Control low density (2x lo3 cells/well) endothelial cell cultures maintained in the absence of FGF exhibited a very low rate of thymidine incorporation (fig. 6). Under these conditions there was no change in the number of cells (data not shown). As demonstrated in fig. 6, an almost complete inhibition of the FGF mitogenic activity was obtained in the presence of 1000 U/ml IFN. A partial inhibition (30%) was observed when the BAEC were treated with 100 U/ml of IFN. These results further indicated that the anti-proliferative effect of IFN is not related to a specific growth factor. DISCUSSION The results of the present study clearly demonstrate that purified huIFNo, and crude boIFN, inhibit bovine SMC and BAEC proliferation induced by serum and purified growth factors. This effect of IFN on synchronized cell transition into S phase was determined by [3H]thymidine incorporation into DNA. Long-term exposure of subconfluent BAEC to IFN resulted in the inhibition of cell division, measured by cell counts [22]. Similar results were also obtained with SMC (data not shown). However, the antiproliferative effect of IFN was more pronounced when the synchronized cell transition from GO/G1 into S phase was measured. The precise mode of action of IFN on cell proliferation is not clear. IFN markedly extends all phases of the cell cycle [24], but Gl and G2 are the most prolonged [25, 261. However, these studies were performed almost exclusively with transformed cell lines, and it has been pointed out that interpretation of these results may be complicated by the autonomous production of transformed growth factors [23]. We have studied the antiproliferative activity of IFN in a system composed of normal vascular SMC and BAEC, stimulated from quiescence with growth factors. In these cells, huIFN had both an antiviral and an anti-proliferative effect. This is contrary to the findings of Taylor-Papadimitriou et al. [lo, 231, who reported that transformed bovine cell lines were resistant to the antiproliferative effect of huIFNa. This discrepancy may be ascribed to the differences in sensitivity of normal and transformed cells to huIFNa. In the present study the GO/G1 phase of the SMC cycle was investigated by adding PDGF and PDS at different intervals. We were particularly interested to test the hypothesis that IFN and growth factors are two families of hormones with antagonistic action [27, 281. The results of this study confirm that quiescent SMC, like 3T3 cells, need only to be transiently exposed to exogenous PDGF in order to become competent to go through the Gl phase, when the necessary progression factors are added [2]. During this stage of competence several genes including c-myc and c-fos are transiently induced by PDGF [29, 301. The antiExp Cell

Res 161 (1985)

Antagonistic

effects of interferon

and growth factors

305

proliferative effects of IFN on SMC are apparently divorced from the activities of PDGF, since DNA synthesis is inhibited, regardless of whether huIFN is added at the same time as PDGF, or as late as 8 h after removal of PDGF from the SMC culture medium. Moreover, the inhibitory effect of IFN on SMC and BAEC cultures could not be completely overcome by increasing the concentration of the mitogenic stimulus. It should also be emphasized that the bovine vascular endothelial cells, as opposed to the SMC, were continuously exposed to 10% calf serum prior and during their incubation with IFN. Hence, it is inconceivable that the inhibitory effect of IFN on BAEC is due to an antagonistic competition with the serum growth-promoting factors. We interpret these findings to indicate that IFN and FGF, PDGF or other serum growth factors are not competing hormones. Rather, these substances appear to act by unrelated mechanisms on different metabolic pathways and cellular events. It has recently been reported that IFN inhibits the c-myc mRNA expression in certain cell lines [31, 321. If our interpretation, that the growth factors and IFN act independently is correct, then the down-regulation of c-myc is a consequence rather than the cause of the antiproliferative activity of IFN. Does IFN participate in the control of the growth of vessel wall cells? We have recently shown that cultured vascular endothelial cells, of bovine and human origin, produce significant amounts of a- and /3-IFN when infected with viruses [20]. In the present study a crude supernatant, containing boIFN, inhibited the growth of SMC. In vivo, this IFN may protect against excessive SMC growth in response to endothelial damage caused by viruses and other agents. Since IFNa is relatively stable and circulates [35, 361, it seems plausible that IFNa may act as a circulating molecule which regulates SMC proliferation. Animal studies have shown that indeed IFN may have an anti-atherogenic effect in vivo. In rabbits fed a cholesterol-rich diet, concomitant treatment with IFN or IFN-inducers, reduced the extent of intimal surface lesions and intimal lipid deposits [37, 381. We have recently found that IFN enhances the capacity of bovine vascular endothelial cells to produce prostacyclin [22]. Prostaglandins, on their part, are known to inhibit the proliferation of SMC in vitro [33]. Thus, IFN may be involved directly and indirectly in the control of SMC proliferation. It may be. however, that the IFN-enhanced production of prostacyclin in vivo affects the control of SMC proliferation by suppressing platelet aggregation and the release of PDGF [34]. Platelet aggregation which lead to the release of PDGF is considered a major event in vascular SMC proliferation in vivo and atherosclerosis [I]. Based on these considerations and the animal studies [37, 381 it is proposed that under certain circumstances IFN may play an important role in the biological regulation of SMC proliferation. This project was supported by grants from the South African Medical Research Council and the Israel Council for Research and Development (A. du P. H.); the A. Jurzykowski Foundation and the Israel Academy of Sciences (A. E.); the National Institutes of Health (CA30289) (I. V.); the National Exp Cell Res 161 (1985)

306 Heyns et al. Council for Research and Development, Israel, and the D.K.F.Z., Germany, and the Sir Zelman Cowen University research fund, Australia (A. P.). A. du P. H. is a research associate from MRC and University of the Orange Free State Blood Platelet Research Institute, Bloemfontein, South Africa.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

Ross, R & Glomset, J A, New Eng j med 295 (1976) 369, 420. Pledger, W J, Stiles, C D, Antoniades, H L Scher, C D, Proc natl acad sci US 74 (1977) 4481. Vogel, A, Raines, E, Kariya, B, Rivest, M J J & Ross, R, Proc natl acad sci US 75 (1978) 2810. Stiles, C D, Capone, G T, Scher, C D, Antoniades, H N, Van Wyk, J J & Pledger, W J, Proc natl acad sci US 76 (1979) 1279. Scher, C D, Shepard, R C, Antoniades, H N Kc Stiles, C D, Biochim biophys acta 560 (1979) 217. Schwartz, S M M, Gajdusek, C M & Selden III, S C, Arteriosclerosis 1 (1981) 107. Maciang, T, Cerundolo, J, Ilsley, S, Keller, P R & Forand, R, Proc natl acad sci US 76 (1979) 5674. McKay, D G & Margaretten, W, Arch intern med 12 (1967) 125. Jameson, P, The interferon system, Texas reports in biology & medicine (ed S Baron, F Dianzani & G J Stanton) vol. 41, p. 133. University of Texas Medical Branch at Galveston (1982). Taylor-Papadimitriou, J, Interferons (ed D C Burke & A G Morris) vol. 109. Cambridge University Press, London (1983). Camel, K & Hirvonen, S, J gen virol 39 (1978) 541. Raines, E W&Ross, R, J biol them 257 (1982) 5154. Borden, E C & Leohardt, P H, J lab clin med 89 (1977) 1036. Gospodarowicz, D, Moran, J & Birdwell, C R, Proc natl acad sci US 73 (1976) 4120. Rutherford, R B & Ross, R, J cell bio169 (1976) 1%. Ross, R, Nist, C, Kariya, B, Rivest, M J, Raines, E & Callis, J, J cell physiol 97 (1980) 497. Kaiser, N, Vlodavsky, I, Tur-Sinai, A, Fuks, Z & Cerasi, E, Endocrinology 113 (1983) 228. Gospodarowicz, D, Moran, J & Brain, D, J cell physiol91 (1977) 377. Vlodavsky, I, Johnson, L K, Greenberg, G Bc Gospodarowicz, D, J cell bio183 (1978) 468. Einhom, S, Eldor, A, Vlodavsky, I, Fuks, Z & Panet, A, J cell physiol 122 (1985) 200. Lockart, R Z, Interferons and interferon inducers (ed N B Finster) vol. 11. Elsevier, Amsterdam, New York (1973). Eldor, A, Fridman, R, Vlodavsky, I, Hy-Am, E, Fuks, Z & Panet, A, J clin invest 73 (1984) 251. Taylor-Papadimitriou, J, Shearer, M, Balkwill, F R 62 Fantes, K H, J interferon res 2 (1982) 479. Collyn d’Hooghe, M, Brouty-Boye, D, Malaise, E P & Gresser, I, Exp cell res 105 (1977) 73. Killander, D, Lindahl, P, Lundin, L, Leary, P & Gresser, I, Eur j immuno16 (1976) 56. Balkwill, F & Taylor-Papadimitriou, J, Nature 274 (1978) 798. Inglot, A D, Oleszak, E & Kisielow, B, Arch viro163 (1980) 291. Oleszak, E & Inglot, A D, J interferon res 1 (1980) 37. Kelly, K, Co&ran, B H, Stiles, D & Leder, P, Cell 35 (1983) 603. Greenberg, M E & Zitf, E B, Nature 311 (1984) 433. Jonak, G J & Knight, E, Proc natl acad sci US 81 (1984) 1747. Einat, M, Resnitzky, D & Kimchi, A, Nature 313 (1985) 597. Hutter, J J, Gwebu, E T, Panganomola, R U, Milo, G E & Comwell, D E, Science 197 (1977) 289. Smith, D L, Willis, A L, Ham, W, Reid, D, Mahmud, I & Nakamura, G, Prostaglandins and leukotrienes (ed J M Bailey) p. 217. Raven Press, New York (1984). Abstr. Cesario, T, Vaziri, N, Slater, L & Tillis, J, Infect immun 24 (1979) 851. Levin, S & Hahn, T, Clin exp immuno146 (1981) 475. Kuo, P T, Wilson, A C, Goldstein, R C & Schaub, R G, J Am co11cardiol3 (1984) 129. Wilson, A C, Schoub, R G, Goldstein, R C, Conn, H L & Kuo, P T, Clin res 32 (1984) 217. Dicker, P & Rosengurt, E, J cell physiol 109 (1981) 99.

Received May 3, 1985

Exp CellRes

16I (1985)

F’rinted

in Sweden