Charles E. Hart

Recent evidence has demonstrated that there is more than one form of platelet-derived growth factor (PDGF) receptor and that these receptors differ in their specificity for the multiple isoforms of PDGF. We present evidence that high affinity binding of PDGF requires association of two different receptor subunits: an alpha-subunit that can bind either a B- or an A-chain of PDGF, and a beta-subunit that can bind only a B-chain. The alpha- and beta-subunits appear to be similar in size but can be distinguished by binding specificity and by an antireceptor monoclonal antibody, PR7212, which recognizes only the beta-subunit. In the absence of PDGF, these subunits either exist separately or form rapidly reversible complexes. In the presence of PDGF, receptor subunits of appropriate specificity interact with a PDGF molecule to form a high affinity complex. Both the absolute and relative numbers of these two PDGF receptor subunits vary on different cell types and correspond to differences in the mitogenic sensitivity of cells to the different PDGF isoforms.

Previous studies involving platelet-derived growth factor (PDGF) have been based on the premise that a single cell-surface receptor binds all three isoforms of PDGF (AA, BB, and AB). It is now shown that two populations of PDGF receptor exist and can be distinguished by their ligand binding specificity. The B receptor binds only the BB dimer, whereas the A/B receptor binds AA, BB, and AB dimers. Human dermal fibroblasts appear to express seven times as much B receptor as A/B receptor. The B receptor is responsible for most PDGF receptor phosphorylation.

A lambda gt11 cDNA library containing DNA inserts prepared from human liver mRNA has been screened with an antibody to human factor X, a plasma protein participating in the middle phase of the blood coagulation cascade. Ten positive clones were isolated from 2 X 10(6) phage and plaque purified. The cDNA in the phage containing the largest insert has been sequenced and shown to code for human factor X. This cDNA insert contained 1137 base pairs coding for a portion of the light chain of the molecule, a connecting region, the heavy chain, a stop codon, a short 3' noncoding region, and a poly(A) tail. The sequence of A-T-T-A-A-A, which functions as a potential recognition site for polyadenylylation or processing, was present in the 3' end of the coding sequence and preceded the stop codon of TGA by 1 base pair and the poly(A) tail by 14 base pairs. The amino acid sequence deduced from the cDNA indicated that factor X is synthesized as a single-chain polypeptide containing the light and heavy chains connected by an Arg-Lys-Arg tripeptide. The single-chain molecule is then converted to the light and heavy chains by cleavage of two (or more) internal peptide bonds. In plasma, these two chains are linked together by a disulfide bond. The DNA sequence coding for the active site of human factor X showed a high degree of identity with prothrombin and factor IX, two other vitamin K-dependent serine proteases that participate in blood coagulation. These data along with the protein sequence data previously published for the light chain of human factor X establish the complete amino acid sequence for the mature protein present in plasma.

Smooth muscle cells (SMC) in rat carotid artery leave the quiescent state and proliferate after balloon catheter injury, but the signals for mitogenesis are not known. In this study, the possibility that cells within damaged arteries produce a growth factor that could act locally to stimulate SMC replication and repair was examined. We found that the genes for PDGF-A and -B (ligand) and PDGF receptor (alpha and beta subunits) were expressed in normal and injured carotid arteries and were independently regulated during repair of carotid injury. Two phases of PDGF ligand and receptor gene expression were observed: (a) In the early stage, a large decrease in PDGF beta-receptor mRNA levels preceded 10- to 12-fold increases in PDGF-A transcript abundance in the first 6 h after wounding. No change in PDGF alpha-receptor or PDGF-B gene expression was found at these times. (b) In the chronic phase, 2 wk after injury, neointimal tissue had lower levels of PDGF alpha-receptor mRNA (threefold) and higher levels of PDGF beta-receptor mRNA (three- to fivefold) than did restored media. Moreover, in situ hybridization studies identified a subpopulation of neointimal SMC localized at or near the luminal surface with a different pattern of gene expression than the underlying carotid SMC. Luminal SMC were strongly positive for PDGF-A and PDGF beta-receptor transcripts, while showing little or no hybridization for PDGF-B or PDGF alpha-receptor. Immunohistochemical studies showed strongly positive staining for PDGF-A in SMC along the luminal surface. These data show that changes in PDGF ligand and receptor expression occur at specific times and locations in injured carotid artery and suggest that these changes may play a role in regulating arterial wound repair.

Osteopontin is an Arg-Gly-Asp-containing acidic phosphoprotein recently shown to be upregulated in vascular smooth muscle during rat arterial neointima formation and in human atherosclerotic plaques. Functional studies showed that osteopontin promoted adhesion of both cultured aortic endothelial cells and aortic smooth muscle cells. Adhesion of vascular cells to osteopontin was dose dependent and half maximal when solutions containing 7 and 30 nmol/L osteopontin were used to coat wells for endothelial and smooth muscle cells, respectively. Smooth muscle cells adherent to osteopontin were spread after 60 minutes, whereas endothelial cells remained round, although flattened, at this time point but were spread at 90 minutes. Cell spreading on osteopontin was accompanied by the formation of focal adhesion plaques. A newly developed anti-osteopontin antibody completely inhibited adhesion of both cell types to osteopontin but not to fibronectin or vitronectin. In addition, the peptide GRGDSP blocked adhesion to osteopontin, suggesting that integrins mediate Arg-Gly-Asp-dependent adhesion. Indeed, an antibody against the alpha v beta 3 integrin neutralized adhesion of both endothelium and smooth muscle cells to osteopontin by approximately 50%, demonstrating that alpha v beta 3 is one osteopontin receptor on vascular cells. Osteopontin also promoted the migration of smooth muscle cells in a Boyden-type chamber, with half-maximal effects observed at 77 nmol/L osteopontin. Checkerboard analysis demonstrated that this stimulus was chemotactic in nature. Our findings suggest that osteopontin may be functionally important as an adhesive and chemotactic molecule for vascular cells, particularly when levels of osteopontin are dramatically increased, as is the case after arterial angioplasty and in atherosclerotic plaques.