Vitronectin at sites of cell-substrate contact in cultures of rat myotubes.

Affinity-purified antibodies to the serum glycoprotein, vitronectin, were used to study sites of cell-substrate contact in cultures of rat myotubes and fibroblasts. Cells were removed from the substrate by treatment with saponin, leaving fragments of plasma membrane attached to the glass coverslip. When stained for vitronectin by indirect immunofluorescence, large areas of the substrate were brightly labeled. The focal contacts of fibroblasts and the broad adhesion plaques of myotubes appeared black, however, indicating that the antibodies had failed to react with those areas. Contact sites within the adhesion plaque remained unlabeled after saponin-treated samples were extracted with Triton X-100, or after intact cultures were sheared with a stream of fixative. These procedures expose extracellular macromolecules at the cell-substrate interface, which can then be labeled with concanavalin A. In contrast, when samples were sheared and then sonicated to remove all the cellular material from the coverslip, the entire substrate labeled extensively and almost uniformly with anti-vitronectin. Extracellular molecules associated with substrate contacts were also studied after freeze-fracture, using a technique we term "post-release fracture labeling." Platinum replicas of the external membrane were removed from the glass with hydrofluoric acid to expose the extracellular material. Anti-vitronectin, bound to the replicas and visualized by a second antibody conjugated to colloidal gold, labeled the broad areas of close myotube-substrate attachment and the nearby glass equally well. Our results are consistent with the hypothesis that vitronectin is present at all sites of cell-substrate contact, but that its antigenic sites are obscured by material deposited by both myotube and fibroblast cells.