Progressive modulation of endothelial phenotype during in vitro blood vessel formation.

"Sprouting" vascular endothelial cells were used as an in vitro model system to study the progressive morphologic and biosynthetic changes associated with the formation of tubular structures. In vitro, sprouting endothelial cells formed spontaneously without the addition of any exogenous factors from cultures of cloned endothelium exhibiting a polygonal/cobblestone phenotype. These phenotypically variant endothelial cells differentiated to form associated cell networks or nodules which gradually reorganized into tubular structures. Concomitant with these morphologic changes, the biosynthesis of extracellular matrix proteins was modulated, as determined by Northern blot analysis, metabolic labeling, and immunocytochemistry. The initial sprouting phase was characterized by the induction of type I collagen synthesis and the appearance of fibronectin containing the ED-A domain, in comparison to their absence in cloned cultures displaying a stable polygonal/cobblestone phenotype. The organizational stage, where the sprouting endothelial cells assembled into tubular structures, was additionally characterized by the expression of type IV collagen. These studies demonstrate that the progression from polygonal/cobblestone to sprouting cultures, and subsequent tubular organization, involves major alterations in extracellular matrix protein expression. This developmental phenomenon, although not completely analogous to blood vessel formation in vivo, nevertheless may be helpful in understanding the role of matrix macromolecules in the angiogenic process.

Cyclic biaxial strain of pulmonary artery endothelial cells causes an increase in cell layer-associated fibronectin.

Bovine pulmonary artery endothelial (PAE) cells were cultured on an artificial compliant substrate (Mitrathane) and were strained biaxially at a frequency of 1/s for 2, 4, 6, 7, or 24 h. Total protein synthesis, determined by estimating the incorporation of radiolabeled precursors into nondialyzable protein, was increased in cultures that had been biaxially strained for 6, 7, or 24 h, with differences more apparent in the cell layer fraction than in the medium fraction. Medium and cell layer-associated fibronectin were quantitated by enzyme-linked immunosorbent assay and by densitometric analysis of the autoradiograms of electrophoresed protein. Fibronectin levels in the medium of biaxially strained cells were initially depressed in comparison to nonstrained controls but, with time, began to approach control values. Cell layer-associated fibronectin of biaxially strained cultures was significantly elevated at 24 h, whereas DNA synthesis was not altered. Immunohistochemical localization of fibronectin and factor VIII-von Willebrand antigen revealed a more intense staining pattern in strained cultures. Distribution of stress fibers containing fibrous actin was visualized by staining with rhodamine-phalloidin and was altered in strained cultures. These observations indicate that cells respond to cyclic biaxial strain by selectively enhancing structural components associated with cell adhesion.

A system to reproduce and quantify the biomechanical environment of the cell.

An in vitro system that permits application of a uniform biomechanical stimulus to a population of cells with great precision has been developed. The device is designed to subject living cells to reproducible and quantifiable biaxial strains from 0 to 10% at rates from quasi-static to 1 s-1 and frequencies from 0 to 5 Hz. Equations for determining the strain in the substrate upon which the cells are grown, based on easily measured parameters, are derived and validated experimentally. The mechanical properties of the substrate are determined, and it is demonstrated that cells can easily be cultured in the apparatus. By use of the system, cloned bovine pulmonary artery endothelial cell clones are subjected to 5% biaxial strains applied at a peak strain rate of 0.5 s-1 and a frequency of 1 Hz for 7 h with cell viability greater than 84% and cell detachment less than 8%. We demonstrate that cells must be attached to the substrate for them to be stretched and that cell strain and substrate strain are not equal. With the use of fluorescently labeled beads as cell surface markers to measure the actual strain produced in the cells as a result of the deformation of the substrate, cell elongation was found to be approximately 60% of the strain in the substrate. This constant appeared to be affected by both in vitro cell age and morphology.

Effects of biaxial deformation on pulmonary artery endothelial cells.

An apparatus has been designed to subject vascular cells grown on a compliant substrate in vitro to uniform, quantifiable levels of biaxial deformation. The system described can be controlled with respect to strain level, rate, and frequency to mimic the pulsatile force to which vascular cells are exposed in vivo under both physiologic and pathologic conditions. In the experiments presented here, bovine pulmonary artery endothelial cells were grown on a substrate of segmented polyurethane urea (Mitrathane). Cell growth and morphology on this substrate were compared with those of cells grown on standard tissue culture polystyrene with no difference noted between the two substrates. Primary cultures of pulmonary artery endothelial cells were seeded onto Mitrathane, which was then subjected to cyclic biaxial deformation-producing strains of 0.78%, 1.76%, 4.9%, or 12.5% at a frequency of 1 sec-1 and a duty cycle of 0.5 sec-1 for 7 h. Cells subjected to deformations generating strains of either 4.9% or 12.5% secreted significantly less fibronectin than nondeformed cells. Similar results were obtained in experiments using cloned pulmonary artery endothelial cells on Mitrathane subjected to the 4.9% strain; however, total protein synthesis was increased. Cell viability and DNA synthesis were not affected by cyclic biaxial deformation in these experiments.

Studies on cell proliferation and tracer localization in the kidneys of guinea pigs with experimental autoimmune anti-tubular basement membrane nephritis.

The mitotic activity in kidneys of guinea pigs with experimental autoimmune anti-tubular basement membrane (TBM) nephritis was investigated using autoradiographic techniques to determine the uptake of [3H]thymidine by actively dividing cells. It was observed in these animals that cells of proximal tubules, distal tubules, cortical and medullary interstitium, medullary collecting ducts, and loops of Henle took up significantly greater amounts of [3H]thymidine when compared with normal animals. In addition, the behaviour of horseradish peroxidase (HRP) and goat anti-HRP IgG in extraglomerular sites in the kidneys of these animals was studied. Contrary to what was expected, these tracers appeared to be less concentrated in the tubules and interstitium of animals with anti-TBM disease, with tracer movement restricted in areas of disrupted TBM. The significance of these observations is discussed.

An ACTH-induced renal glomerular lesion in the mouse: immunofluorescence microscopy.

Mice were injected with adrenocorticotropic hormone (ACTH) and the glomerular lesion that was induced was studied by light and immunofluorescence microscopy. By light microscopy, kidneys from ACTH-treated mice showed typical ACTH-induced glomerular lesions. Immunofluorescence of kidneys from ACTH-treated mice revealed intense staining for IgG and IgM in the extraglomerular mesangium (EGM), in Bowman's space, and in the ascending thick limb of Henle near the macula densa. Staining for immunoglobulins was unchanged after treatment with acid buffer. Immunoreactivity for complement (C3) was confined largely to the EGM and Bowman's space. Staining for albumin was almost exclusively in Bowman's space or the peripheral glomerular tuft in discrete aggregates. The above patterns of IgG, IgM, C3 and albumin were seen in control mice, although much less frequently. The results show that in mice, treatment with ACTH results in the increased accumulation of plasma proteins in the juxtaglomerular apparatus (JGA). This effect may reflect a role for the JGA in the normal clearing of plasma proteins from the glomerulus and/or directly from the blood.