Actin Filament Density Research Articles

Tumor Necrosis Factor Enhances Endothelial Cell Susceptibility to Oxygen Toxicity: Role of Glutathione

The effect of tumor necrosis factor-alpha (TNF) on hyperoxia-induced endothelial injury in vitro was investigated. TNF caused a time- and dose-dependent reduction in the number of viable pulmonary artery endothelial cells. The TNF-mediated endothelial cytotoxicity was more pronounced under hyperoxia (95% O2 and 5% CO2) than under normoxia (95% air and 5% CO2). Pretreatment of endothelial cells with TNF (0.01 micrograms/ml or 240 U/ml) for 18 h at normoxia reduced the intracellular concentration of total glutathione (GSH), whereas the concentration of oxidized GSH was increased. These TNF-treated endothelial cells were more susceptible to hyperoxia- or hydrogen peroxide-mediated cytotoxicity. TNF also induced changes in endothelial morphology and in the distribution and density of actin filaments. Exogenous GSH or L-2-oxothiazolidine-4-carboxylate, which enhanced endothelial GSH concentrations, partially protected endothelial cells against TNF-mediated cytotoxicity, morphologic changes, and actin filament redistribution, especially under the hyperoxic condition. These results suggest an important role of GSH in modulating endothelial response to TNF.

Control of actin polymerization in live and permeabilized fibroblasts.

We have investigated the spatial control of actin polymerization in fibroblasts using rhodamine-labeled muscle actin in; (a) microinjection experiments to follow actin dynamics in intact cells, and (b) incubation with permeabilized cells to study incorporation sites. Rhodamine-actin was microinjected into NIH-3T3 cells which were then fixed and stained with fluorescein-phalloidin to visualize total actin filaments. The incorporation of newly polymerized actin was assayed using rhodamine/fluorescein ratio-imaging. The results indicated initial incorporation of the injected actin near the tip and subsequent transport towards the base of lamellipodia at rates greater than 4.5 microns/min. Furthermore, both fluorescein- and rhodamine-intensity profiles across lamellipodia revealed a decreasing density of actin filaments from tip to base. From this observation and the presence of centripetal flux of polymerized actin we infer that the actin cytoskeleton partially disassembles before it reaches the base of the lamellipodium. In permeabilized cells we found that, in agreement with the injection studies, rhodamine-actin incorporated predominantly in a narrow strip of less than 1-microns wide, located at the tip of lamellipodia. The critical concentration for the rhodamine-actin incorporation (0.15 microM) and its inhibition by CapZ, a barbed-end capping protein, indicated that the nucleation sites for actin polymerization most likely consist of free barbed ends of actin filaments. Because any potential monomer-sequestering system is bypassed by addition of exogenous rhodamine-actin to the permeabilized cells, these observations indicate that the localization of actin incorporation in intact cells is determined, at least in part, by the presence of specific elongation and/or nucleation sites at the tips of lamellipodia and not solely by localized desequestration of subunits. We propose that the availability of the incorporation sites at the tips of lamellipodia is because of capping activities which preferentially inhibit barbed-end incorporation elsewhere in the cell, but leave barbed ends at the tips of lamellipodia free to add subunits.