Cultured endothelial cells produce multiple growth factors for connective tissue cells

Cultured bovine aortic endothelial cells (BAEC) secrete into their medium a growth-promoting factor that stimulates many connective tissue cells in culture. We now report that this growth-promoting activity is due to at least two different proteins which are biochemically separable and immunologically distinct. Cation exchange chromatography (Carboxymethyl-Sephadex) of concentrated BAEC-conditioned medium yields two major peaks of growth-promoting activity which adsorb at pH 8 and elute with a salt gradient. One of these peaks contains as well a protein that inhibits the binding of radioiodinated platelet-derived growth factor (PDGF) to its receptor on target cells. The PDGF-like mitogen is purified approx. 25-fold by this chromatographic step. A second peak of mitogenic activity exhibits no binding to the PDGF receptor. Both the PDGF-like mitogenic activity and the PDGF-distinct mitogenic activity are highly cationic, stable to boiling, sensitive to beta-mercaptoethanol, and between 30 and 50 kD in molecular weight. Complementary studies with human umbilical vein endothelial cells in culture were performed. These human cells also produce both growth-promoting activity and a protein that binds to the PDGF receptor. The latter activity is greatly inhibited by a specific antiserum to human PDGF, whereas the growth-promoting activity of the conditioned medium is minimally affected. The degree of inhibition of the two activities is, however, quantitatively consistent: 3.5 ng of PDGF-like activity in the radioreceptor assay is inhibited, while 5 ng of PDGF-like activity in the DNA synthesis assay is inhibited. The data from the two species are consistent with the proposal that cultured endothelial cells produce at least two distinct mitogens, one of which is biochemically and immunologically related to PDGF.     

Reversion of autocrine transformation by a dominant negative platelet-derived growth factor mutant.

A non-receptor-binding mutant of the platelet-derived growth factor (PDGF) A chain, PDGF-0, was generated by exchanging 7 amino acids in the sequence. The mutant chains formed dimers that were similar to wild-type PDGF-AA with regard to stability and rate of processing to the mature 30-kDa secreted forms. Moreover, the mutant chains formed disulfide-bonded heterodimers with the PDGF B chain in NIH 3T3 cells heterodimer underwent the same processing and secretion as PDGF-AB. Transfection of c-sis-expressing 3T3 cells with PDGF-0 significantly inhibited the transformed phenotype of these cells, as determined by the following criteria. (i) Compared with PDGF-0-negative clones, PDGF-0-producing clones showed a reverted morphology. (ii) Clones producing PDGF-0 grew more slowly than PDGF-0-negative clones, with a fivefold difference in cell number after 14 days in culture. (iii) The expression of PDGF-0 completely inhibited the ability of the c-sis-expressing 3T3 cells to form colonies in soft agar; this inhibition was overcome by the addition of recombinant PDGF-BB to the culture medium, showing that the lack of colony formation of these cells was not due to a general unresponsiveness to PDGF. The specific expression of a PDGF-0/PDGF wild-type heterodimer in COS cells revealed that the affinity of the mutant heterodimer for the PDGF alpha receptor was decreased by approximately 50-fold compared with that of PDGF-AA. Thus, we show that a non-receptor-binding PDGF A-chain mutant neutralizes in a trans-dominant manner the autocrine transforming potential of the c-sis/PDGF B chain by forming low-affinity heterodimers with wild-type PDGF chains. This method of specifically antagonizing the effect of PDGF may be useful in investigations of the role of PDGF in normal and pathological conditions.     

A significant part of macrophage-derived growth factor consists of at least two forms of PDGF

The macrophage has been suggested to be responsible for the connective tissue cell proliferation that accompanies most chronic inflammatory responses. One of the secretory products of activated macrophages is MDGF, a growth factor (or factors) for fibroblasts, 3T3 cells, smooth muscle, and vascular endothelium. This report demonstrates that a significant portion of the mitogenic activity for 3T3 cells secreted by cultured human alveolar and peritoneal macrophages is due to a molecule (or molecules) similar to platelet-derived growth factor (PDGF). Two size classes (approximately 37,000-39,000 and 12,000-17,000 daltons) of mitogenically active PDGF-like molecules are detected by two criteria--antigenic similarity with PDGF and ability to compete with 125I-PDGF for high-affinity binding to the PDGF receptor. The presence of mRNA for the B chain of PDGF is demonstrated by Northern analysis, and de novo synthesis of these molecules by activated macrophages is shown by immunoprecipitation of 35S-labeled proteins with anti-PDGF IgG.     

Migratory and proliferative effect of platelet-derived growth factor in rabbit retinal endothelial cells: Evidence of an autocrine pathway of platelet-derived growth factor

Angiogenesis is a crucial event in the progression of diabetic retinopathy. Migration and proliferation of endothelial cells (EC) are important steps in angiogenesis and are caused by angiogenic factors such as basic fibroblast growth factor (bFGF). In this work, capillary EC were isolated from rabbit retinal tissues and rabbit retinal EC (RREC) were found to secrete a migration factor for RREC in conditioned medium (CM). The activity was inhibited by an anti-platelet-derived growth factor (PDGF) antibody, but not by an anti-bFGF antibody. We also found that RREC showed a migratory response to PDGF. The response was induced by PDGF-BB and PDGF-AB dose dependently, but not by PDGF-AA, indicating that it was mediated by PDGF-β receptor-dependent pathways, and that the PDGF-like factor was PDGF-BB or -AB. In addition, PDGF-BB induced the proliferation of RREC as well as bFGF. These data indicate that RREC have an autocrine pathway of PDGF by the secretion of and the response to PDGF. PDGF may play significant parts in angiogenesis in the progression of diabetic retinopathy. © 1994 Wiley-Liss, Inc.     

Ultrastructural localization of platelet-derived growth factor and related factors in normal and transformed cells

Visualization of platelet-derived growth factor (PDGF) and PDGF-like growth factors in cultured cells has been achieved by cryo-ultramicrotomy in combination with immunogold labeling. Immunogold staining of cryosections requires a mild chemical fixation in order to ensure preservation of antigenicity and ultrastructural details. Therefore the effect of several chemical fixatives on the antigenic properties of PDGF and PDGF-like growth factors was studied by indirect immunofluorescence using a polyclonal anti-PDGF antiserum. These studies demonstrated that formaldehyde has no effect on antigenicity, in contrast to glutaraldehyde or acrolein. For this reason formaldehyde was used as the only fixative for the visualization of PDGF in cryosections. PDGF was visualized in cryosections of normal human fibroblasts, preincubated with PDGF under various conditions. Preincubation at 4 degrees C with PDGF resulted in partial internalization of the growth factor. During subsequent warming of the cells to 37 degrees C PDGF was translocated to the nucleus. PDGF was also detected in the cytoplasm of tumor cells producing endogenous PDGF-like growth factors (neuroblastoma and simian sarcoma virus-transformed cells) but in these cases no significant amounts of these growth factors were present in the nucleus or at the extracellular surface of these cells. These results will be discussed in view of the intracellular routing of PDGF in normal responsive cells and of PDGF-like growth factors in factor-producing cells.     

Identification of a PDGF-like mitoattractant produced by NIH/3T3 cells after transformation with SV40

It has previously been shown that fibroblastic cells transformed by SV40 exhibit a reduced requirement for PDGF for growth. In addition, NIH/3T3 cells lose both their chemotactic response to PDGF and specific cell surface binding of PDGF after transformation with SV40. We have now examined whether the SV40 transformed NIH/3T3 cells are producing a factor which acts similarly to PDGF. Our studies indicate that NIH/3T3 cells transformed with SV 40 produce a factor which shares many biological properties with PDGF. We were unable to detect this activity in conditioned media from nontransformed NIH/3T3 cells. The SV40/NIH/3T3 derived factor appears to possess both chemotactic and mitogenic activity for connective tissue cells but not endothelial or epithelial cells. Furthermore, in preliminary studies, this activity competes with 125I-PDGF for binding to smooth muscle cells. The biochemical properties of the SV40/NIH/3T3 derived factor are different from those of PDGF. The SV40 activity appears to reside in a heat labile acidic protein (pI less than 7.0) of MW less than 30,000 whereas PDGF is a heat stable basic protein (pI9.8) of 30,000 MW. Production of this factor may play a role in the decreased serum requirement for cell replication exhibited by SV40-transformed NIH/3T3 cells by supplying the cells with their own PDGF-like growth factor.     

Regulation of production of a platelet-derived growth factor-like protein by cultured bovine aortic endothelial cells

Platelet-derived growth factor (PDGF) is a potent mitogen for cultured cells of mesenchymal origin. Known sources of PDGF or PDGF-like protein are blood platelets, several transformed cell lines, and cultured endothelial cells (EC). We have examined the regulation of production of a PDGF-like protein in cultures of bovine aortic EC using a specific radioreceptor assay for PDGF. EC constitutively secreted PDGF-like protein into serum-containing or serum-free medium. The rate of production of PDGF-like protein was constant for at least 3 weeks and was not due to release of an internal store, since cell lysis by repeated freeze/thaw cycles did not relase significant amounts of the protein. Synthesis of PDGF-like protein was sensitive to changes in the pH of the media and was maximal at pH 8.5. Production of PDGF-like protein was independent of EC growth rate: rapidly dividing cells and confluent, quiescent cells produced equal amounts per cell. However, sparse, quiescent EC produced more PDGF-like protein per cell than did confluent, quiescent cells. Several phorbol esters stimulated production of PDGF-like protein. At a concentration of 10^?6 M, a twofold stimulation was observed upon addition of the tumor promoter 12-O-tetradecanoylphorbol 13-acetate (TPA) and nearly a fourfold stimulation upon addition of the nonpromoting analog, methyl TPA. Incubation of EC with endotoxin (10 μ/ml) resulted in a twofold stimulation of PDGF-like protein production. In all experiments with endotoxin and phorbol esters, an increase in the production of PDGF-like protein was accompanied by morphological changes in the EC cultures. The cells appeared elongated and fibroblastic and exhibited low viability. A mathematical model was developed in which PDGF-like protein production was shown to consist of two separate components—production at a constant rate by healthy cells and a large burst of synthesis and secretion by dying cells. These results suggest that injurious agents may be capable of stimulating production of a growth factor by the endothelium.     

Cultured endothelial cells do not respond to a platelet-derived growth-factor-like protein in an autocrine manner

Cultured endothelial cells produce a growth factor similar or identical to platelet-derived growth factor (PDGF). Endothelial cells are able to proliferate in plasma-supplemented medium, while most nontransformed cells require serum-supplemented medium. Since PDGF is a major serum mitogen, we have tested the possibility that endothelial cells interact with and respond to the autologously produced PDGF-like (PDGF-c) protein. We have found that bovine aortic and rat heart endothelial cells express little or no cell surface PDGF receptors as determined by binding of pure 125I-PDGF. Treating these cells under acidic conditions, which release receptor-bound PDGF in control cells without affecting receptor function, did not reveal a population of cryptic receptors. In addition, when rat heart endothelial cells were grown in the presence of an antibody to PDGF, proliferation was unimpaired, though no detectable free PDGF was present in the medium. An equivalent amount of antibody completely blocked the mitogenic response of human fibroblasts that had been preincubated for 1 h at 37 degrees C with an equivalent dose of PDGF. Thus, endothelial cells do not respond mitogenically in a manner that would be expected from the interaction of autologously produced PDGF with its cell surface receptor. Endothelial cells were detergent-solubilized and immobilized on nitrocellulose in an attempt to detect the presence of intracellular PDGF receptors. Specific binding of 125I-PDGF to adsorbed, solubilized bovine aortic or rat heart endothelial cells was undetectable, though significant binding to adsorbed, solubilized fibroblasts, used as a positive control, was observed. We conclude that endothelial cells do not have detectable intracellular PDGF receptors.     

A homologue of platelet-derived growth factor produced by rat alveolar macrophages

Rat alveolar macrophages secrete a growth factor that renders rat lung fibroblasts competent to initiate DNA synthesis in vitro in the presence of platelet-poor plasma. This biological activity resembles that of platelet-derived growth factor (PDGF). After separation from putative associated binding proteins by chromatography under acidic conditions, the macrophage-derived factor exhibited a relative molecular weight similar to that of highly purified human PDGF. The factor bound to a monospecific antibody to human PDGF and thus could be quantitated in an enzyme immunoassay for PDGF. It competed with radiolabeled human PDGF for receptor sites for PDGF on rat lung fibroblasts, and binding to these receptor sites could be specifically inhibited by anti-PDGF. These data strongly support the view that the factor derived from rat alveolar macrophages is homologous to human PDGF and is similar to human macrophage-derived PDGF-like growth factor. Furthermore, we have demonstrated that the lung contains both an effector cell (pulmonary macrophage) and a potential target cell (interstitial fibroblast) for this cytokine. Therefore the rat appears to be an appropriate animal model in which to study macrophage-derived PDGF-like growth factors as mediators of proliferation in pulmonary fibrogenesis.     

Characterization of a high-molecular-weight protein immunoprecipitated by platelet-derived growth factor antisera

We previously demonstrated that a high-molecular-weight glycoprotein could be immunoprecipitated from metabolically labeled U-2 OS cells with platelet-derived growth factor (PDGF) antiserum and that it appears to be derived from a different precursor than is the 30 kD PDGF-like mitogen produced by these cells. These findings were unexpected, since the molecular weight of this glycoprotein is too large to be encoded by the PDGF structural genes. From experiments with metabolically labeled U-2 OS human osteosarcoma, fibroblasts, and NRK cells, we report here that a 185 kD protein immunoprecipitated with PDGF antiserum has the following characteristics. 1) It is a PDGF binding protein that is unrelated to alpha 2-macroglobulin. 2) It is phosphorylated in response to PDGF stimulation. 3) It is immunoprecipitated by phosphotyrosine antibodies. 4) It is not a substrate of epidermal growth factor-induced tyrosine kinase activity. These studies indicate that high-molecular-weight proteins immunoprecipitated by antiserum to PDGF represent a complex between PDGF and a binding protein capable of being phosphorylated by a PDGF-induced tyrosine kinase. These characteristics are identical to those of the PDGF receptor.     

Analysis of a transformed cell line using antisense c-fos RNA

Simian sarcoma virus (SSV)-infected NIH-3T3 cells (SSV-NIH-3T3), express a homologue of platelet-derived growth factor, (PDGF) a powerful inducer of the c-fos gene. We have used these cells to test the hypothesis that autocrine stimulation by PDGF-like molecules leads to c-fos expression which is functional in the transformed phenotype. We have transfected SSV-NIH-3T3 cells with a c-fos antisense-RNA expression vector, pSVsof, or control plasmids. pSVsof-transfected cells exhibit markedly decreased c-fos mRNA and protein levels, restored density-dependent growth arrest and reduced (three of five clones) tumorigenicity compared to control lines. The results confirm that c-fos cooperates in the transformed phenotype of SSV-NIH-3T3 cells.     

Persistence of the mitogenic response to platelet-derived growth factor (competence) does not reflect a long-term interaction between the growth factor and the target cell

Quiescent BALB/c-3T3 cells exposed briefly to platelet-derived growth factor (PDGF) become "competent" to replicate their DNA even if PDGF is removed from cell culture medium prior to the onset of DNA synthesis. We have suggested that persistence of the PDGF-induced competent state reflects a rapidly induced and relatively stable biochemical change within the target cells. Others suggest that the phenomenon reflects a long-term association between PDGF and its target cells or perhaps between PDGF and the cell culture dish. This controversy has been addressed (a) by examining the effect of anti-PDGF antibodies on PDGF- induced competence and (b) by studying the chemical fate of 125I- labeled PDGF. Anti-PDGF antibodies inactive both soluble and surface- bound PDGF. However, if quiescent 3T3 cells are exposed to PDGF for as little as 30 min, subsequent addition of these antibodies to the culture medium does not prevent the mitogenic response. Under conditions where the PDGF-induced competent state decays stochastically with a t1/2 of 18-20 h, cell-associated 125I-PDGF decays with a t1/2 of approximately 50 min. These data do not support the concept that persistence of the PDGF-induced competent state reflects a long-term association between PDGF and the target cells or between PDGF and the culture dish.     

Demonstration of an antibody against platelet-derived growth factor

An antibody was raised in a rabbit against platelet-derived growth factor (PDGF), purified from human platelets. It recognized specifically ¹²⁵I-labelled PDGF, as demonstrated by gel electrophoretic analysis of precipitated radioactivity. Addition of immunoglobulins from the immune serum to cell cultures together with PDGF inhibited the multiplication-stimulating activity of the growth factor. Further, multiplication-stimulating activity in crude platelet lysate was retained upon passage over a column of Sepharose-bound anti-PDGF; it was recovered after elution of the column with acid. A radioimmunoassay for PDGF was developed, permitting the determination of 5 ng PDGF/ml. Platelet factor 4 and β-thromboglobulin, both being proteins of the platelet α-granule, did not cross-react in the radioimmunoassay for PDGF. The radioimmunoassay was utilized to examine the immunological relationship between PDGF and some other growth factors. The anionic growth factors previously demonstrated in pletelets were found to be antigenically distinct from cationic PDGF. Similar results were obtained with epidermal growth factor (EGF) or fibroblast growth factor (FGF). In contrast, osteosarcomaderived growth factor, a PDGF-like growth factor previously known as a product of cultured human osteosarcoma cells, showed some cross-reactivity in the radioimmunoassay for PDGF. It was concluded that osteosarcoma-derived growth factor (ODGF) has both physical-chemical and immunological properties in common with PDGF.     

Altered growth factor sensitivity in EL2 rat fibroblasts: influence of this biological characteristic on cell growth

Extensive evidence indicate that platelet-derived growth factor (PDGF) and epidermal growth factor (EGF) play a key role in the stimulation of the 3T3 fibroblast replication: in this connection, PDGF and EGF act as a competence and a progression factor, respectively. We have previously demonstrated that EGF alone leads density-arrested EL2 rat fibroblasts to synthesize DNA and proliferate in serum-free cultures. Here, we have analyzed the role of EGF in the control of EL2 cell proliferation. Our data show a dose-related effect of EGF on DNA synthesis and cell growth, with maximal stimulation for both parameters at 20 ng/ml. On the other hand, autocrine production of PDGF or PDGF-like substances by EL2 cells is seemingly excluded by experiments with anti-PDGF serum or medium conditioned by EL2 fibroblasts. EGF binding studies show that EL2 cells possess high affinity EGF receptors, at a density level 3 to 4-fold higher than other fibroblastic lines. In addition, EL2 cells show a normal down-regulation of EGF receptors, following exposure to EGF, but PDGF, fibroblast growth factor (FGF), transforming growth factor beta (TGF beta) and bombesin have not decreased the affinity of EGF receptor for its ligand. Moreover, in EL2 cells, the EGF is able to induce the synthesis of putative intracellular regulatory proteins that govern the PDGF-induced competence in 3T3 cells. Our data indicate that EGF in EL2 cells may act as both a competence and a progression factor, via induction of the mechanisms, regulated in other cell lines by cooperation between different growth factors.(ABSTRACT TRUNCATED AT 250 WORDS)     

Transmodulation of epidermal growth factor binding by platelet-derived growth factor and 12-O-tetradecanoylphorbol-13-acetate is not sodium-dependent in Balb/c/3T3 cells

The addition of platelet-derived growth factor (PDGF) to many types of cells causes a rapid decrease in high affinity binding of 125I-epidermal growth factor (EGF), a process which has been termed transmodulation. Treatment with the tumor promoter, 12-O-tetradecanoylphorbol-13-acetate (TPA) also results in the transmodulation of the EGF receptor in many cell types. PDGF can transmodulate EGF binding through a mechanism that is not dependent on protein kinase C activity. A recent report (Wattenberg, E. V., McNeil, P. L., Fujiki, H., and Rosner, M. R. (1989) J. Biol. Chem. 264, 213-219) described the requirement for a sodium ion influx in the down-modulation of the EGF receptor stimulated by a non-TPA-type tumor promoter, palytoxin, in Swiss 3T3 cells. We tested for a similar sodium requirement in Balb/c/3T3 and Swiss 3T3 cells stimulated by PDGF or TPA in Balb cells treated with TPA for prolonged periods to down-regulate protein kinase C activity. Our results clearly show that the PDGF- and TPA-stimulated transmodulation of the EGF receptor does not require external sodium nor is the process affected by amiloride. In each of these experiments, the loss of 125I-EGF binding occurred to a similar extent and at a similar rate in the presence or absence of sodium. Intracellular pH also did not appear to have a role in the response. The sodium ionophore, monensin, was previously shown to bring about the down-modulation of 125I-EGF binding in Swiss cells. However, our results indicate that monensin-induced transmodulation of the EGF receptor occurs with or without external sodium, suggesting that the loss of binding is not the result of a sodium ion influx. These findings demonstrate that an increase in intracellular sodium does not cause nor is it required for PDGF- or TPA-stimulated EGF receptor transmodulation.