C.A. Sukow, D. Hanein, J. Condeelis*, P. Matsudaira**, and D.J. DeRosier
Rosenstiel Basic Medical Sciences Research Center, Brandeis University, Waltham, MA 02254; *Dept. of Anatomy and Structural Biology, Albert Einstein College of Medicine, Bronx, NY 10461; **Whitehead Institute for Biomedical Research, Massachusetts Institute of Technology, Cambridge, MA 02142
Actin, by its interaction with different actin-binding proteins, can form a variety of specialized structures. Examples of such structures include the lamellipodia and filopodia of Dictyostelium, the microvilli of the cells of the intestinal brush border, and the stereocilia of the hair cells of the inner ear. Some of these actin bundling proteins also have enzymatic activity. The 3D nature of actin bundles limits the structural information that can be gleaned from them. Bundles generally lack crystalline order, and the polymorphism in the filament organization makes it nearly impossible to combine different images to obtain a 3D map of the structure. The approach taken in this work is to use 2D arrays of actin ("rafts"), crossbridged by bundling proteins (Taylor and Taylor, 1992, 1994) as a model for the 3D bundles. The restriction of filaments to 2D simplifies the interpretation of the images, permitting visualization of individual cross-bridges. This data can be combined with complementary information, such as 3D images of actin filaments decorated with the actin binding protein, and crystallographic structures of bundling proteins. The bundling proteins initially being studied are diverse: EF-1 alpha, fimbrin, and aldolase. EF-1 alpha bundles actin in Dictyostelium lamelli-podia/filopodia and also functions in protein synthesis; fimbrin bundles actin in microvilli and stereocilia; aldolase, an important enzyme in the metabolic process, has been shown to bundle actin in vitro, in a way that affects its metabolic function. We are currently in the process of standardizing the tools for characterization of these 2D crossbridged actin arrays. For each bundling protein, appropriate conditions for the formation of crossbridged actin arrays are determined by image analysis, correlated with biochemical data. The biochemistry is performed on specimens taken directly from an electron microscope grid. The polarity of the actin arrays has also been examined by image analysis, and a biochemical approach using S1 decoration is being tried. These tools will enable us to proceed to our goal of analyzing the interactions of actin and its bundling proteins.