The aim of the present study was to verify whether the modifications of the extracellular matrix, described in varicose veins, are also present in cultures of smooth muscle cells from human varicose veins. The accumulation of collagen type III and fibronectin was determined by immunofluorescence in cultures of smooth muscle cells at passage 2–3 during the proliferation phase. After 5 days of culture, the immunostaining of both collagen type III and fibronectin was weaker in cells from varicose than in those of control veins while the expression of collagen type III and fibronectin messenger ribonucleic acids was not significantly different. Collagen type I and III synthesis were quantified by tritiated proline incorporation in control and varicose cell layers at postconfluence. Collagen type I deposition was similar in both types of cell layers while collagen type III was decreased in cell layers from varicose veins. Matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) were also quantified by enzyme immunoassays in supernatants from smooth muscle cell cultures at postconfluence. No significant difference was observed in the synthesis of any of the MMPs (–1, –2 and –9) or their inhibitors (–1 and –2) tested. These data illustrate that smooth muscle cells cultured from varicose veins deposit less collagen type III and fibronectin than control cells despite comparable levels of mRNAs for these proteins suggesting dysregulation of posttranslational steps in the synthesis of both proteins by smooth muscle cells from varicose veins.

Porcine coronary arteries with regenerated endothelium exhibit impaired endothelium-dependent relax-ations. Experiments were designed to analyze the structural and functional changes occurring in regenerated endothelial cells. Primary cultures from regenerated endothelium contained giant endothelial cells, with an increased number of cells with diameter 14.5 m, a reduced ability to proliferate, and signs of apoptosis. The uptake of fluorescent acetylated LDL was increased 2-fold in cultures from regenerated endothelium. The increased uptake of acetylated LDL was confirmed ex vivo in injured coronary arteries. In cultures from regenerated endothelium, cGMP production was decreased under basal conditions and during stimulation with serotonin, bradykinin, and A23187. Thus, during regeneration, there is accelerated senescence of endothelial cells accompanied by increased incorporation of modified LDL and reduction of NO production without decrease in endothelial NO synthase expression. These alterations help to explain the altered endothelium-dependent responses 28 days after balloon injury. (Circ Res. 2000;86:854-861.)

1 Experiments were designed to investigate whether the pertussis toxin-dependent endothelial dysfunction following balloon injury is due to a reduced expression or an insu cient function of Gproteins. 2 Endothelium-dependent responses of porcine coronary arteries were examined in vitro by use of conventional organ chambers. Morphological analysis was performed by isolating and culturing the endothelial cells from these arteries. The expression of Gi-proteins in regenerated endothelial cells was measured by Western blots and immunolabelling. The function of G-proteins was assessed by measuring the GTPase activity of cultured endothelial cells. 3 Eight days following denudation, endothelial regrowth was con®rmed by histological examination and by demonstrating the presence of endothelium-dependent relaxations to bradykinin and 5-hydroxytryptamine (5-HT). In primary culture, the regenerated endothelial cells displayed a`cobblestone' pattern as seen with native endothelial cells. 4 Twenty eight days after denudation, the endothelium-dependent relaxations induced by 5-HT were impaired, but those to bradykinin were maintained. However, the latter were reduced when endotheliumdependent hyperpolarization was prevented. 5 Twenty eight days after denudation, multinucleated giant cells were present in the regenerated but not in the native cultured endothelial cell populations. These regenerated endothelial cells incorporated less tritiated thymidine than native endothelial cells. 6 The intensities of the bands on the immunoblot of the regenerated endothelial cells, when several antibodies against Gia1/a2/a3 were used, were the same as those obtained in native endothelial cells. The immunolabelling with the same antibodies was similar between the giant cells and the regenerated endothelial cells of normal size. The hydrolysis of GTP was lower in regenerated than in native endothelial cell membranes. 7 In conclusion, endothelium-dependent relaxations mediated by Gi-proteins are impaired in balloon denuded coronary arteries. This dysfunction following regeneration cannot be explained by a reduced expression of Gi proteins but rather re¯ects an abnormal function of the G-proteins in the regenerated endothelium.