beta-Interferon alters the pattern of proteins secreted from quiescent and platelet-derived growth factor-treated BALB/c-3T3 cells

Feedback inhibitors in normal and tumor tissues

Negative feedback represents the principal mechanism for regulating growth in biological systems. Over the past 20 years, our understanding of the role played by inhibitory factors governing this process has advanced considerably. This is particularly well illustrated in the field of experimental hematology with the recognition of hemopoietic progenitor cell proliferation inhibitors, an expanding group of unrelated peptides that act to limit proliferation in hemopoietic precursor cells. The characterization and subsequent production of these molecules by chemical synthesis or recombinant DNA technology has enabled investigators to explore their role in normal hemopoiesis and define a potential role in clinical medicine. A number of inhibitory factors, including macrophage inflammatory protein-1 alpha (MIP-1 alpha) and the tetrapeptide AcSDKP appear to share a relative specificity to hemopoietic progenitor cell subsets. Others, such as interferon and tumor necrosis factor, have a more complex action and their hemopoietic effects are likely to be indirect and nonspecific. In addition to the role of inhibitors in normal steady state, it has become increasingly evident that loss of sensitivity to the normal feedback inhibitory signals may be of central importance in carcinogenesis and tumor promotion. This presumably represents a developmental strategy that allows the neoplastic cell to maintain a growth advantage over its normal cell counterpart. The underlying mechanisms that terminate in inhibitor-resistance are yet to be elucidated, but in some instances they may be associated with aberrant tumor suppressor gene function.

Interferon-beta inhibits progression of relapsing-remitting experimental autoimmune encephalomyelitis

The results of two phase III clinical trials have recently shown that interferon-beta (IFNbeta) is effective in the treatment of relapsing-remitting multiple sclerosis (RRMS). Treatment with IFNbeta results in a significant decrease in the rate of clinical relapse and a marked delay in progression to disability compared to placebo-treated control patients. In the present study, we demonstrate similar therapeutic effects after treating (SWR X SJL)F1 mice with IFNbeta at the onset of clinical signs of experimental autoimmune encephalomyelitis (EAE), a disease animal model widely used in MS studies. EAE was actively induced by immunization of (SWR X SJL)F1 mice with the immunodominant encephalitogenic peptide 139-151 of myelin proteolipid protein (PLP). In blinded testing, mice treated with IFNbeta at EAE onset showed a delay in progression to clinical disability as determined by marked improvement with time in mean clinical score, significant delay in onset of relapse, and significant decrease in exacerbation frequency compared to placebo-treated control mice. The therapeutic effect of IFNbeta was accompanied by a significant inhibition of delayed-type hypersensitivity (DTH) but not proliferation in response to the priming PLP 139-151. In addition, IFNbeta treatment resulted in an overall decrease in severity of both inflammation and demyelination in the central nervous system. These results mimic in an autoimmune animal model the effectiveness of IFNbeta treatment observed in MS. Moreover, our study suggests that anti-viral properties of IFNbeta are not essential for producing therapeutic effects in autoimmune demyelinating disease, and that the efficacy of IFNbeta in the treatment of MS may be due to inhibition of autoreactivity.

The existence of a growth-specific DNA binding factor for the promoter region of mouse ST2 gene

A comparison of the 5'-flanking regions of human and mouse ST2 genes revealed the presence of two highly conserved DNA sequences. The promoter activity assay with a luciferase gene as a reporter showed that the deletion of the upstream conserved region diminished the transcriptional activity in growing BALB/c-3T3 cells. By electrophoretic mobility-shift analysis, the presence of a factor that binds to the positive regulatory region of the mouse ST2 gene was found in growing but not in quiescent BALB/c-3T3 cells. These results suggest the functional importance of this conserved region and the requirement of a binding factor for the expression of the ST2 gene.

Different factors bind to the regulatory region of the Na+,K(+)-ATPase alpha 1-subunit gene during the cell cycle

Three factors that bind to the positive regulatory region (ARE) of the Na+,K(+)-ATPase alpha 1-subunit gene were shown to be present in growing BALB/c-3T3 cells as shown by the gel retardation assay pattern in which three specific complexes (C1, C2 and C3) were identified. The complexes are similar to those observed in MDCK cell nuclear extracts in which linker substitution mutations in the competitor gave parallel specific effects in both cells. During the process of the cell growth cycle, the relative mobility of C3 was altered, and the amount of C1 decreased in the G0 state. All three complexes (C1, C2 and C3) disappeared and other specific complexes with higher mobilities were alternatively observed at 6 h after serum stimulation and thereafter. The expression of the mRNA for the alpha 1-subunit gene was repressed at G0 and gradually increased after serum stimulation. These results suggest that different sets of factors are responsible for the transcription of the gene at different stages of the cell cycle.

Identification of the product of the murine ST2 gene

The murine ST2 gene is expressed in growth-stimulated BALB/c-3T3 cells. This gene encodes a protein that is similar to the extracellular portions of the interleukin-1 receptors (types 1 and 2). In this study, we prepared a polyclonal antibody against the recombinant ST2 protein produced in Escherichia coli. This antibody detected recombinant ST2 protein in the culture fluid of COS7 cells transfected with a mammalian expression vector (pEF-BOS) carrying ST2 cDNA. Using this antibody, we could detect the ST2 protein in the culture fluid of growth-stimulated BALB/c-3T3 cells, and in the medium of continuously growing cells, but not in that of growth-arrested cells. ST2 proteins produced in COS7 cells and BALB/c-3T3 cells were N-glycosylated as predicted from nine putative N-glycosylation sites in its deduced amino-acid sequence.

Murine ST2 gene is a member of the primary response gene family induced by growth factors

The murine ST2 gene, which encodes a protein remarkably similar to the extracellular portion of murine interleukin 1 receptor types 1 and 2, is expressed in growth-stimulated BALB/c-3T3 cells in the presence of 50 micrograms/ml of cycloheximide. The treatment with 1,000 U/ml of purified native murine beta-interferon superinduced, rather than suppressed, the ST2 mRNA expression as in the cases of c-myc and JE mRNAs. These results suggested that the murine ST2 gene belongs to the family of primary response genes induced by growth factors. Furthermore, a longer ST2-related mRNA was found in BALB/c-3T3 cells that were stimulated to proliferate in the presence of cycloheximide.

Cytotoxic and cytostatic effects of antitumor agents induced at the plasma membrane level

A variety of antitumor agents inhibit cell proliferation by interacting with the plasma membrane. They act as growth factor antagonists, growth factor receptor blockers, interfere with mitogenic signal transduction or exert direct cytotoxic effects. The P-glycoprotein encoded by the MDR1 gene represents a transmembrane protein which catalyzes the efflux of various antitumor agents. This membrane protein is the target of compounds acting as Multi-Drug Resistance (MDR)-modulators. Finally, several established antitumor agents which are considered to represent DNA-targeted drugs, including anthracyclines, platinum complexes and alkylating agents, cause a variety of membrane lesions. Their contribution to the antitumor activity of these drugs is discussed.

Posttranscriptional regulation of c-myc proto-oncogene expression and growth inhibition by recombinant human interferon-beta ser17 in a human colon carcinoma cell line

Recombinant human interferon-beta ser17 (IFN-beta ser17), a cytokine that exhibits both antiviral and antiproliferative activity against a wide variety of cell types, causes a time- and dose-dependent inhibition of monolayer growth and of the expression of the c-myc proto-oncogene in DLD-1 Clone A human colon-carcinoma cells. The suppression of c-myc expression mediated by IFN-beta ser17 is due to a posttranscriptional destabilization of c-myc mRNA rather than to an inhibition of c-myc mRNA transcription. There is evidence suggesting that the selective reduction in the half-life of c-myc mRNA in IFN-beta ser17-treated cells occurs through an increase in the activity of the 2',5'-oligoadenylate synthetase/RNase L [2',5'-oligo (A) synthetase] pathway in DLD-1 Clone A cells. Cotreatment of these cells with IFN-beta ser17 and the anticancer agent N-methylformamide leads to the partial abrogation of 2',5'-oligo (A) synthetase activity and the stabilization of c-myc mRNA. These findings suggest that there is a correlation between the IFN-beta ser17-mediated suppression of c-myc expression and the induction of 2',5'-oligo (A) synthetase activity in DLD-1 clone A cells.

Gamma-interferon-induced nerve growth factor receptors in colorectal carcinoma cell lines

Gamma-interferon (gamma IFN) was found to induce expression of the 150,000 M(r) cell surface and the 35,000 M(r) chromatin receptors for nerve growth factor (NGF) in the SW1116 colorectal carcinoma cell line that does not express NGF receptors. In the SW707 colorectal carcinoma cell line that expresses a low level of NGF receptors, gamma IFN stimulated expression of the cell surface and the nuclear receptors. Induction of NGF receptors in SW1116 cells resulted in internalization and nuclear translocation of 125I-NGF. When NGF bound to the chromatin, ribosomal RNA synthesis was inhibited. Two-dimensional gel electrophoresis of [35S]methionine-labeled chromatin proteins indicated significant changes in chromatin protein composition in cells treated and not-treated with gamma IFN. gamma IFN effectively stimulated the expression of NGF receptors in two colorectal carcinoma cell lines, but inhibited the expression in melanoma and breast carcinoma cells. It is suggested that gamma IFN, by modulating the expression of NGF receptors may affect the NGF-dependent growth of some tumor cell lines.

Interferons: Pleiotropic cellular modulators

Interferons play a key role in host response as pleiotropic modulators of cell function. As induced proteins, interferons contrast with other physiologic regulators such as glucocorticoids which are produced relatively continuously. Antitumor effects have been suggested to be principally the result of two mechanisms: a direct effect on the functional capacity or antigenic composition of tumor cells or an indirect effect on modulation of immunological effector cell populations with tumor specificities. Over the past decade, interferons have been established as therapeutically useful molecules for malignant and viral diseases.

Interferon-induced modulation of epidermal growth factor-stimulated growth of a human breast tumor cell line

We have determined that interferon-alpha (IFN-alpha) and IFN-gamma inhibit the growth of a human breast tumor cell line, S4, in vitro. Cells were more sensitive to the antiproliferative effects of low-dose IFN-gamma than IFN-alpha. As the growth of the S4 cell line is enhanced by epidermal growth factor (EGF), we examined the effect of IFN on EGF-dependent growth of S4 cells. Cells plated in 2.5% serum alone failed to grow. EGF stimulated these cells to grow more than twofold. IFN substantially attenuated the EGF-stimulated growth of S4 cells. Binding of EGF to its receptor was unaffected by pretreatment of cells with IFN-alpha. However, a 24-h exposure of cells to IFN-gamma significantly increased the number of EGF receptors on S4 cells. Internalization of the EGF receptor was unaffected by IFN treatment. Binding remained elevated through 4 days of IFN-gamma exposure. Scatchard analysis of receptor binding data revealed that IFN-gamma increased the number of binding sites without changing the affinity of the receptor for its ligand. These results demonstrate that IFN inhibits EGF-stimulated growth of a breast tumor cell line and suggest that the antiproliferative effect of IFN may be due, in part, to its interaction with growth factor-initiated pathways.

Dissociation of proto-oncogene induction from growth response in normal human fibroblasts

Proto-oncogenes are cellular homologues of viral oncogenes that are known to be associated with regulation of growth and differentiation. The c-myc and c-fos proto-oncogenes have been extensively studied by using established cell lines but to a lesser extent by using normal cells. Using physiologic growth modulators, we have shown that mitotic stimulation of normal human dermal fibroblasts is associated with induction of c-myc and c-fos, but that growth inhibition of these cells is not necessarily accompanied by their down-regulation. When treated with both serum and interferon-alpha during quiescence, fibroblasts were delayed in their progress into the S-phase of the cell cycle as compared to cells treated with serum alone and displayed substantial growth inhibition as measured by cell number at the end of 1 week. However, this growth inhibition was not preceded by down-regulation or delay in induction of c-myc and c-fos mRNA. The above studies suggest that in normal fibroblasts growth inhibition is not necessarily dependent on down-regulation of transcription of either c-myc or c-fos and that interferon may act to inhibit cell growth either through a post-transcriptional effect on cellular proto-oncogenes necessary for cell proliferation or through induction of other, as yet unrecognized gene(s) associated with growth arrest.

Glucocorticoid dexamethasone reversibly complements EJ-RAS oncogene to transform mouse embryo BALB-3T3 cells

EJ-A is a Balb-3T3 transfectant cell line that bears a small number of EJ-ras oncogene copies/cell, has low EJ-ras expression, and resembles the parental cell line in displaying a non-transformed phenotype. The glucocorticoid hormone dexamethasone reversibly induces transformation traits in EJ-A cells, namely: 1) morphological transformation; 2) increased growth rate and saturation density; 3) reduced G1 length; and 4) independence of the FGF requirement to initiate DNA synthesis. Western blot analysis revealed that dexamethasone does not increase the p21ras protein intracellular level. beta-IFN, added to the culture medium, does not suppress the dexamethasone-induced growth stimulation and morphological transformation. Therefore, glucocorticoid hormones can complement low EJ-ras expression to transform Balb-3T3 cells, by a mechanism that is likely to be independent of p21ras increase and beta-IFN decrease.

Interferon inhibition of bombesin-stimulated mitogenesis in Swiss 3T3 cells occurs without blocking c-fos and c-myc expression

We have investigated the effect of murine interferon-beta on the growth of quiescent Swiss 3T3 cells stimulated by the amphibian tetradecapeptide mitogen bombesin. We found that IFN inhibited mitogenesis in bombesin-stimulated cells in a dose-dependent manner. In addition, bombesin, in common with platelet-derived growth factor (PDGF), was able to reverse the inhibitory effects of IFN in cells stimulated by epidermal growth factor (EGF) and insulin, which elicit a mitogenic response via a distinct signaling pathway. The observation that IFN was as effective in inhibiting DNA synthesis when added 48 h before or as late as 3 h after the mitogen, indicated that a block in one of the early regulatory signals was not essential for the anti-growth effects. Indeed, IFN did not inhibit the increased expression of c-fos and c-myc mRNA induced by bombesin in quiescent Swiss 3T3 cells. The combined data indicate that events occurring late in G1 are more likely targets for IFN action in Swiss 3T3 cells.

Growth factor modulation of melanoma growth stimulatory activity mRNA expression in human malignant melanoma cells correlates with cell growth

This report demonstrates that the expression of melanoma growth stimulatory activity (MGSA) mRNA can be modulated in a positive fashion in the Hs294T human melanoma cell line by PDGF and MGSA. There is close correlation between MGSA expression and the pattern of cell growth in Hs294T cells.

Interferon induces cell fusion and the formation of multinuclear cells in a culture of Ehrlich ascites tumor cells

The increase in number of Ehrlich ascites tumor (EAT) cells was diminished significantly when the cell culture was treated with 1,000 IU/ml of recombinant mouse alpha or beta interferon (IFN). Microscopical observation revealed that almost all the cells showed bi- or multinuclear morphology 3 to 5 days after IFN treatment. Furthermore, a videorecording showed that each multinuclear cell arose by fusion after mitotic division of one parental cell.

Effect of interferon on secretion of proteins by various murine cell lines

The interferons (IFNs) have been shown to be antagonistic to the growth stimulatory effects of mitogens on cultured cells. A report of the interactions of IFN-beta and platelet-derived growth factor on BALB/c-3T3 mouse cells established that IFN itself induced the secretion of a limited number of proteins from this cell line. The present work was undertaken to determine if other murine cell lines treated with homologous IFN-beta also secreted new or additional protein(s) in response to this agent and if this response correlated with other phenotypic properties of the cells. The cell lines examined included L929 cells and two derivatives of this line (GM347 and WDIFN), CAK-TK-, Swiss-3T3, and BALB/c-3T3. Each line was exposed to [35S]methionine in the absence and in the presence of IFN-beta, the supernatant fluids collected, and the radioactive, secreted proteins examined by fluorography after electrophoresis through SDS-containing polyacrylamide gels. Two cell lines (GM347 and Swiss-3T3) did not appear to secrete new or additional proteins after IFN treatment. However, four lines (L929, WDIFN, CAK-TK-, and BALB/c-3T3) did secrete new or additional proteins in response to IFN. Thus IFN-induced secretion of protein appeared to be a common but not universal phenomenon. In addition, although the number and apparent size(s) of the IFN-induced, secreted proteins were different in these various lines, one protein (Mr = 89-90,000) appeared to be secreted by each of them. In this respect it was unique. Moreover the IFN-induced secretion of protein did not appear to correlate with the antiviral or antiproliferative effects of IFN.

Tumor necrosis factor stimulates proliferation of human osteosarcoma cells and accumulation of c-myc messenger RNA

The objective of this study was to establish whether human recombinant tumor necrosis factor (TNF) can significantly stimulate the proliferation of some tumor cells. Treatment with TNF had little or no effect on the growth of human tumor cells and murine NIH/3T3 cells cultured in medium with high serum concentration. Two tumor lines, SK-MEL-109 melanoma and HOS osteosarcoma cells, were adapted to grow in medium supplemented with 0.5% serum. The growth of these SK-MEL-109 cells was inhibited by TNF, but that of the HOS cells was greatly stimulated by TNF in a dose-dependent way. Treatment with 10 ng/ml of TNF resulted in a two-fold increase in the rate of cell division. This effect of TNF was also shown by measuring DNA and protein synthesis. The continuous presence of TNF was not required for its mitogenic activity on HOS cells cultured with 0.5% serum, since treatment for only one day with TNF resulted in prolonged growth stimulation. The failure of TNF to promote division of cells cultured in medium with 10% serum may possibly be explained by the presence of saturating amounts of growth factors in serum. Interferons abolished the mitogenic activity of TNF on HOS cells. Furthermore, TNF did not show synergism with insulin or epidermal growth factor in stimulating growth of these cells. The level of c-myc mRNA was increased five-fold after 30 minutes of treatment with TNF. This shows that TNF is a growth factor for HOS cells and that it induces accumulation of c-myc mRNA.

Recombinant interferon-gamma inhibits the mitogenic effect of platelet-derived growth factor at a level distal to the growth factor receptor

Highly purified preparations of recombinant human interferons (rIFNs)-alpha A, -beta, and -gamma all inhibited platelet-derived growth factor (PDGF)-induced DNA synthesis in normal human dermal fibroblasts, as monitored by incorporation of [3H]-thymidine into trichloroacetic acid (TCA)-insoluble material. rIFN-gamma was the most potent, since it blocked the PDGF response by 50% at about 10 U/ml or 0.3 ng/ml, whereas with rIFN-alpha A and rIFN-beta 4000 U/ml and 600 U/ml, respectively (10 ng/ml in both cases), were required to achieve the same effect. There was a close parallelism between the ability of these rIFNs to inhibit PDGF mitogenic activity and their capacity to inhibit cell proliferation in serum-containing medium. None of the rIFNs inhibited specific binding of 125I-PDGF to fibroblasts, and none interfered with receptor internalization. The mechanism of action of rIFN-gamma was analyzed further. rIFN-gamma did not inhibit uptake of [3H]-thymidine into these cells. However, it shifted if the time point of initiation of DNA synthesis from about 14 h after stimulation with PDGF to about 18 to 21 h and decreased significantly the rate of the DNA synthesis. rIFN-gamma could be added up to 6 h following stimulation with PDGF with no loss of its inhibitory effect. rIFN-gamma also blocked the mitogenic activity of epidermal growth factor and basic fibroblast growth factor. Taken together these results implicate that rIFN-gamma exerts its antimitogenic effect by inhibiting a process that occurs late in the PDGF signaling pathway and onto which the activity pathways of other mitogens converge. In view of the important role PDGF may play in wound-healing and in the pathogenesis of the proliferative lesions of arteriosclerosis, these data point to a possible role IFN-gamma may play as a regulator of these processes in vivo.

A 63 kDa protein is secreted from BALB/c-3T3 cells entering the G1 phase from the G0 state

A 63 kDa protein is detectable in the culture fluid of mouse BALB/c-3T3 cells traversing from the G0 state to the G1 phase, whereas it is undetectable in the culture fluid of quiescent or growing BALB/c-3T3 cells. Secretion of the protein is maximal at 10 h after serum addition. G0-specific ts mutant cells (rat tsJT60) also secrete the 63 kDa protein only when the quiescent cells are stimulated by serum addition at permissive temperature. These facts indicate that the 63 kDa protein is secreted only from cells traversing from the G0 state to the G1 phase.

Insulin-like growth factor II increases cytoplasmic free calcium in competent Balb/c 3T3 cells treated with epidermal growth factor

To determine the role of calcium in the action of insulin-like growth factor II (IGF-II), we have examined the effect of multiplication stimulating activity, the rat IGF-II, on cytoplasmic-free calcium concentration, [Ca2+]c, in aequorin-loaded Balb/c 3T3 cells. IGF-II does not cause any change in [Ca2+]c in quiescent cells. By contrast, IGF-II induces changes in [Ca2+]c in platelet-derived growth factor(PDGF) - pretreated competent cells: when competent cells are incubated with epidermal growth factor (EGF) for 10 min, subsequent IGF-II induces an immediate increase in [Ca2+]c. Without EGF treatment, IGF-II does not cause any increase in [Ca2+]c. The priming action of EGF is time dependent, requiring approximately 10 min for the maximum effect. The IGF-II-mediated increase in [Ca2+]c is totally dependent on extracellular calcium and is blocked by lanthanum. When DNA synthesis in PDGF-treated competent cells is assessed by measuring [3H]thymidine incorporation, IGF-II by itself has only a small effect. Likewise, a brief treatment with EGF results in only a small increase in [3H]thymidine incorporation. By contrast, in competent cells briefly treated with EGF, IGF-II causes a marked stimulation of [3H]thymidine incorporation. These results indicate that IGF-II increases [Ca2+]c in competent Balb/c 3T3 cells treated with EGF by stimulating calcium influx and that IGF-II-stimulated calcium influx may be related causally to its action on cell proliferation.

Preferential inhibition of cytokine-stimulated bone resorption by recombinant interferon gamma

It is likely that immune cells in the bone marrow produce factors which are involved in the local control of bone remodeling. Immune cell products such as interleukin-1 and the tumor necrosis factors are potent stimulators of bone resorption in vitro. In this paper, we have studied the effects of recombinant murine interferon-gamma on bone resorption stimulated by these agents and the systemic calcium-regulating hormones 1,25(OH)2 vitamin D3 and parathyroid hormone. We found that interferon-gamma completely abolished bone resorption stimulated by the cytokines interleukin-1, tumor necrosis factor alpha and tumor necrosis factor beta. In contrast, parathyroid hormone- and 1,25(OH)2 vitamin D3-stimulated bone resorption were not significantly affected by the addition of interferon-gamma under the same conditions. Parathyroid hormone-stimulated bone resorption was inhibited slightly when larger concentrations of interferon-gamma were used for more prolonged periods. The inhibitory effects on cytokine-stimulated bone resorption occurred at interferon concentrations of 100 U/ml (half-maximal) to 300 U/ml (complete inhibition). This relatively selective inhibition of cytokine-stimulated bone resorption by an immune cell product may have physiological significance in the local control of trabecular bone volume and bone remodeling.

Interferon inhibits the establishment of competence in Go/S-phase transition

Addition of mouse interferon-alpha/beta (IFN) to confluent, quiescent BALB/c 3T3 (clone A31) mouse fibroblasts resulted in a block or delay in serum-induced activation of the cell cycle. It was necessary to add IFN within 6 hours after serum stimulation to inhibit nuclear labeling with [3H]thymidine. This is consistent with the time required for platelet-derived growth factor (PDGF) to induce cells to become competent to respond to additional growth factors present in platelet-poor plasma. Simultaneous addition of IFN with PDGF inhibited the PDGF-induced synthesis of a 29-kilodalton and a 35-kilodalton protein that normally occurs within 1 hour after PDGF addition. IFN also suppressed the general increase in protein synthesis that occurs by the fifth hour after PDGF addition. These results show that IFN antagonizes the action of PDGF, thereby interfering with the activation of Go cells for G1 traverse and S-phase entry.

Platelet-derived growth factor and double-stranded ribonucleic acids stimulate expression of the same genes in 3T3 cells

Platelet-derived growth factor (PDGF) stimulates expression of a "competence" gene family in Balb/c-3T3 cells. The competence family contains the c-myc and c-fos genes together with several functionally uncharacterized genes (JE, KC, and r-fos) that have been isolated as cDNA clones. We show that double-stranded ribonucleic acid is a potent inducer of the competence gene family. Infection with vesicular stomatitis virus also induces expression of this gene family. Conversely, PDGF stimulates expression of genes hitherto characterized as responsive to double-stranded ribonucleic acids, including the beta-fibroblast interferon and (2'-5')-oligoadenylate synthetase genes. These PDGF-inducible genes could conceivably function in a feedback loop to control 3T3 cell growth. Some of the genes, such as c-fos and c-myc, are induced quickly by PDGF and may initiate a round of cell division. Others, such as beta-fibroblast interferon and (2'-5')-oligoadenylate synthetase, are induced more slowly and may function as feedback inhibitors of the growth response to PDGF.