Erythrocyte membrane fractions contain free barbed filament ends despite sufficient concentrations of retained capper(s) to prevent barbed end growth.

Many cellular functions depend on rapid cytoskeletal rearrangements localized to specific cytoplasmic domains. Tight regulation of the submembranous microfilament network is accomplished in large part in erythrocytes and granulocytes by actin binding proteins that cap the fast-growing barbed filament ends. Study of this dynamic system is necessarily hampered by the confounding perturbations of cell lysis and dilution. In this paper, we characterize the functional properties of the membrane-associated spectrin-actin complex from human erythrocytes as it exists after hypotonic lysis. Purified spectrin-actin "seeds" extracted from erythrocyte membranes effectively nucleated actin elongation from their barbed ends. However, polymerization from spectrin-actin complexes associated with the membrane fraction prematurely slowed despite the presence of G-actin in great excess of the critical monomer concentration. The addition of cytochalasin B decreased (rather than augmented) the slowing of elongation attributable to the membrane fraction, indicating that capping of barbed filament ends (not monomer sequestration) was the major mechanism underlying this effect. The paradoxical implication of our findings is that, despite the presence of excess capper(s) in the membrane fraction, the membrane-associated spectrin-actin seeds were not capped until after dilution into physiological ionic strength buffer containing monomeric actin. Furthermore, by comparing the degrees of contamination of the extracted and membrane-associated spectrin-actin preparations, it appeared that recognized capping proteins (including gelsolin and capping protein beta2) were not the predominant cappers found in the membrane pellet after hypotonic lysis. We hypothesize that the barbed ends of membrane-associated spectrin-actin complexes, while not excluding actin monomers, may be selectively inaccessible to certain cappers (perhaps simply as the result of steric hindrance). Growth from such complexes in vivo could be limited by the availability of polymerization-competent G-actin.