We have studied the correlation between the actomyosin organization and microvillar position in an epithelial cell line derived from the proximal pig kidney tubule (LLC-PK1). When grown on glass, these cells are approximately 5-6 microns in height and develop numerous microvilli that project from the dorsal membrane. A fairly homogeneous distribution of microvilli was achieved by synchronization of the cell cycle. These microvilli are of the brush border type, as defined by their content of villin and their anchorage in a myosin-rich terminal web-like structure. When LLC-PK1 cells were injected with two monoclonal antibodies against pig brain nonmuscle myosin, in concentrations yielding a 1:1 ratio of antibody to myosin, neither microvillar number nor length was affected. However, when we examined the cells by scanning electron microscopy 1-3 h after microinjection, we found that one of the antibodies (a-PBM 4) had a profound effect on microvillar position: more than 50% were seen tilted or lying prone on the plasma membrane. The microvilli of cells injected with the other antibody (a-PBM 9) were not significantly different from those of cells injected with control antibodies. This difference correlates with in vitro properties of the antibodies: a-PBM 4 decreases the actin-activated Mg(2+)-ATPase of pig brain nonmuscle myosin quite substantially, while a-PBM 9 affects it only moderately. These differential effects are probably a consequence of the different epitope location as determined for both antibodies, not of differences in antibody affinity. Our data are compatible with the hypothesis that a-PBM 4 also interferes with the actomyosin interaction in situ, thus decreasing the effective cross-linking of microvillar rootlets by myosin filaments in the terminal web. On the basis of this model, we suggest that myosin filaments are essential for the upright position of brush-border type microvilli.

We have used polyclonal and monoclonal antibodies raised against calf thymus profilin to localize the corresponding protein in translocating, spreading, and stationary rat fibroblasts. Immunofluorescence of whole cells and immunogold labeling on ventral membranes of lysis-squirted cells showed that profilin was markedly enriched in the highly dynamic lamellipodia or pseudopodial lobes. Within these regions, a significant fraction was colocalized with dynamic actin filaments organized in actin ribs, cortical filaments, or stress fiber-like bundles, and little profilin was found in membrane areas appearing free of actin. In contrast, stress fibers of stationary cells as well as actin arcs and ring-like bundles of spreading and migrating cells showed very little label. These results are discussed in context with the proposed role of profilin in regional membrane dynamics typical for fibroblasts and are compared to previous data (Hartwig et al.: J. Cell Biol. 109:1571-1579, 1989) on profilin distribution in platelets and granulocytes.

The microfilament system is thought to provide motor elements needed for plasma membrane dynamics. This article focuses on two protein components that may play key roles in this process: (1) Profilin, a G-actin binding protein which is considered as the source of actin subunits necessary for rapid changes in the amount of actin filaments. Our data demonstrate that profilin is synthesized even in terminally differentiated blood cells of a high dynamic potential. In addition, we show that plasma membrane-associated profilin in fibroblasts is unevenly distributed and is concentrated in areas that are highly motile. (2) The filament-forming myosin which is the classical motor protein in the microfilament system. We show that interfering with myosin filaments by microinjecting antibodies causes brush border-type microvilli on epithelial cells to loose their upright position. This result, together with our previous observations on the effects of anti-myosin injection into fibroblastic and epithelial cells (loss of stress fibers and cellular contact sites, increase in locomotory activity, delay of cytokinesis), suggests that bipolar myosin filaments are needed to maintain a specific cortical tension which is lost upon antibody binding.