![]() |
|
My research interests pertain to the cytoskeleton, its morphology and regulation. The cytoskeleton is a highly adaptive scaffold of proteins that respond to both internal and external stimuli. The cytoskeleton thus functions as both a structure and a signaling complex. My work in the Taylor lab has centered on the structural analysis of macromolecular assemblies of cytoskeletal proteins on lipid monolayers. In this manner we can create 2D paracrystalline arrays of F-actin cross-linked by actin-bundling proteins such as α-actinin and villin. With these arrays we have pushed the image analysis envelope beyond simple spatial averaging techniques into the realm of single-particle methods with 3D volume alignment and correspondence analysis.
α-Actinin is a modular protein belonging to the spectrin superfamliy that cross-links and bundles actin filaments in both muscle and non-muscle cells. α-Actinin cross-links F-actin to form large (> 1μm) rafts on the lipid monolayer. In order to improve the stochastic signal-to-noise ratio of the cross-link motifs, we first aligned all motifs to the common actin filament and followed with classification of the left- and right-side cross-links independently. The resulting left and right class averages could then be pasted together down the central, aligned actin filament to recreate a class average for every motif. With the improved detail observed in these averages we could perform quantitative measurements of cross-link length and build comparative molecular models to show the observed variation in the length and the angle of the α-actinin cross-links. Unique to this study are the numbers of α-actinin molecules bound to successive crossovers on the same actin filament. These monofilament-bound α-actinins may resemble the form of α-actinin involved in actin-membrane attachments in focal adhesions. These results suggest that α-actinin is not simply a rigid spacer between actin filaments, but rather a flexible cross-linking, scaffolding, and anchoring protein. We suggest these properties of apha-actinin may contribute to tension sensing in actin bundles.
The polar actin bundles of the microvillus are tightly cross-linked by two proteins, villin and fimbrin. Villin is an F-actin nucleating, cross-linking, severing, and capping protein within the gelsolin superfamliy. Villin is unique in this group by its ability to cross-link filaments, which is due to an additional small headpiece domain. Villin shares high sequence homology to gelsolin, which has no known cross-linking ability. In order to define the villin cross-linking structure we have used electron tomography of 2-D rafts of F-actin cross-linked with villin on a lipid monolayer to generate 3-D volumes of F-actin: villin cross-links. These rafts are > or = to 1μm across and consist of polar arrays of F-actin spaced ~126 angstroms apart with villin cross-links occurring approximately once per actin crossover. More than 6,000 paired F-actin crossover repeats with villin protein bound between them were selected as single particles, aligned, and classified by correspondence analysis to produce class averages. Docking of the homologous gelsolin domain structures plus the villin head piece structure into the average density reveals the invariant localization on the actin N-terminus which is quite distinct from that of other actin-binding proteins, such as cofilin, profillin, DNase I, or gelsolin domains G1 and G2. This is the first glimpse of the entire structure of villin in a cross-linking role. Up until now it has been assumed that villin interacts with F-actin in a similar fashion to its close homolog, gelsolin. This study shows this assumption to be wrong and instead lends concrete evidence to the notion that there can be different modes of interaction with actin among highly homologous actin-binding proteins.