Cell-free plasma-derived serum (PDS) is deficient in the platelet-derived growth factor and will not support the growth of 3T3 cells, fibroblasts, or smooth muscle cells. However, when PDS-containing medium is preincubated with endothelial cells, the medium becomes modified so that it will support growth. The activity produced by the endothelial cells results from a polypeptide of 10,000 to 30,000 daltons which has several features that differ from those of the platelet-derived growth factor, including heat instability and lack of adsorption to CM Sephadex .Growth of nontransformed cell strains in culture requires one or more exogenous growth factors .

The current state of our knowledge of the control of endothelial growth and the role of endothelial injury in the pathogenesis of atherosclerosis can be summarized as follows: 1. Endothelial cells can be grown in plasma-derived serum in the absence of exogenous growth factors. This is quite different from the growth requirements of most other nontransformed cells. These factors may, however, prolong replicative life span and increase the ability of endothelium to grow at sparse density. The relevance of these phenomena to the control of endothelial growth in vivo is unclear. There is no evidence that exogenous growth factors are required for wound edge regeneration. In view of the relative lack of growth factor requirements, it is intriguing to consider the possibility that the critical control factor for endothelial cell growth is cell contact. 2. Endothelial cell regeneration may be dependent on endothelial cell motility. The nature of this relationship may be important in controlling the ability of the endothelium to regenerate itself under different flow conditions around lesions or in different parts of the vessel tree and in determining the ability of the endothelium to respond to changes in the connective tissue overlying lesions. 3. Endothelial cells in vivo are able to regenerate small areas of denudation extremely rapidly. This process may be sufficiently rapid to permit the endothelium to replace dying cells as they are being lost, resulting in desquamation without denudation. 4. We have little evidence for endothelial denudation either spontaneously or in response to atherosclerosis risk factors until after lesion formation has begun. This does not rule out the possibility that small, repeated, transient episodes of denudation occur and play a role in the initiation of atherosclerotic lesions. It is important, however, to begin considering the role of nondenuding injuries in atherosclerosis. 5. The fact that thrombosis occurs in atherosclerosis implies an eventual breakdown of endothelial integrity. The mechanism of that breakdown remains unknown. 6. Finally, there is the question of interactions between smooth muscle cells and endothelial cells at the level of growth control. This includes the evidence that there is a critical amount of endothelium that must be lost before lesion formation is stimulated and the recent evidence that endothelial cells produce substances able to regulate growth of smooth muscle cells.

Recognition that the sis gene codes for a protein homologous with at least one of the two chains of plateletderived growth factor has made it possible to directly assess transcriptional expression of platelet-derived growth factor both in cultured cells and in tissue obtained in vivo. We have found that a 3.7-kilobase RNA homologous to the sis gene is expressed at moderate levels in cultured human and bovine endothelial cells, at low levels in in vivo endothelium from human umbilical vein, and at very low levels in bovine aortic endothelium in vivo. This RNA migrates at the same rate as the previously reported sis band in the HUT 102 human T-cell lymphoma line. This band is not found in RNA extracted from freshly obtained bovine aortic media or from human foreskin fibroblasts or cultured fetal human aortic smooth muscle cells. Our in vitro results suggest that the sis gene is responsible for at least part of the platelet-derived growth factor-like mitogenic activity secreted by cultured endothelial cells and indicate that the sis gene is readily activated in endothelial cells during the transition from in vivo conditions to in vitro growth as a monolayer on plastic. Expression of the sis gene by endothelium in vivo raises the possibility that platelet-derived growth factor has a role in the development of the vascular system in the young animal and in the maintenance of the normal vascular system in the adult.Although the production of peptide growth factors by tissues in vivo has long been suspected to be of biological importance (1), the in vivo detection of growth factor production has been limited by the sensitivity of available assay systems. Most known growth factors may be present in amounts too low to be directly detected in vivo by current techniques. Cultured endothelial cells secrete a significant amount of mitogenic activity as assayed on fibroblasts and smooth muscle cells (SMC) (2). Most of this mitogenic activity has been attributed to a growth factor or factors unique and specific to endothelial cells, which is biochemically distinct from platelet-derived growth factor (PDGF) and several other characterized growth factors (3). Approximately 25% of the mitogenic activity in endothelial cell conditioned media, though, has been specifically attributed to a PDGF-like protein (4,5), as determined by 125I-labeled PDGF radioreceptor competition assay and by inhibition of the competitor activity by antiserum against human PDGF.The discovery that the peptide sequence of the simian sarcoma virus transforming protein (p28515) predicted from the nucleotide sequence of the simian sarcoma virus has considerable homology to the known sequence for at least one of the two chains of PDGF (6, 7) and more recent data that the coding sequence of the human c-sis gene predicts a peptide highly homologous to this same chain of PDGF (8) MATERIALS AND METHODSCells. Human umbilical vein endothelial (HUVE) cells for growth in culture were obtained from fresh intact umbilical cords. The umbilical vein was filled wit...

Although the basic fibroblast growth factor (bFGF) gene lacks a traditional consensus signal peptide domain indicative for secretion, many cell types have receptors for bFGF. Since endothelium is a rich source of cell-associated bFGF, we asked under what conditions could bFGF be released or secreted from confluent cultures of bovine aortic endothelial (BAE) cells. The level of bFGF in BAE cell lysates was compared with the level of heparin-releasable bFGF in intact BAE cell monolayers, intact cells with exposed extracellular matrix (nonlytic matrices), and extracellular matrices prepared by cell lysis (lytic matrices). Less than 10% of total cell-associated bFGF was released from intact cell monolayers and nonlytic matrices. In contrast, the levels of bFGF released from lytic matrices depended upon the conditions used to prepare the matrices. Cell lysis at neutral pH generated matrices that released the highest bFGF levels (approximately 50% of total cell-associated bFGF). These matrices were heavily contaminated by histones, indicating the cellular release and adsorption of intracellular proteins to the matrix. Matrices prepared by BAE cell exposure to basic pH (100 mM NH4OH) contained low bFGF content and minor histone contamination. These latter matrices were chosen to study bFGF sequestration, under physiological conditions, into the extracellular matrix of confluent BAE cell cultures. Incubation with endotoxin, an agent acutely toxic to BAE cells, resulted in cellular release and adsorption of endogenous bFGF to cells and matrices, accompanied by histone deposition in the matrices. These results suggested that one mechanism for bFGF release from BAE cell monolayers was passive release induced by severe cell injury and/or cell lysis with secondary adsorption to the matrix.