The structure of a molecular motor

Abstract

Two proteins, actin and myosin, together form a molec- ular motor that produces the work performed by mus- cle tissues, as well as many of the movements seen in all eukaryotic cells. IJntil recently, high resolution struc- tures of these two proteins have not been available. One obstacle in determining these structures has been the dif- ficulty of obtaining crystals suitable for X-ray analysis. In the case of actin, crystallization was hindered by the ten- dency of the protein ~to polymerize. In the case of myosin subfragment 1 (Sl), the head of the myosin molecule, crystallization was prevented by heterogeneity produced by both proteolysis and the presence of a number of dif- ferent myosin isoforms. Now, however, after a scientific odyssey of Homeric iscale involving more than ten years effort on each protein, these problems have been solved by Rayment, Holden a.nd coworkers [ 1 I, working on myo- sin Sl, and by Holmes, Kabsch and coworkers, working on actin [2]. The tendency of actin to polymerize is inhibited by DNase 1, which forms a tight complex with actin. This was ex- ploited by Kabsch, H:olmes and coworkers, who crystal lized a complex of actin with ,DNase 1 [2]. In the case of myosin Sl, crystallization was achieved by a com- bination of innovative protein modification and careful biochemistry [l]. Isoform heterogeneity is exhibited by the essential myosin light chain, one of the Sl small sub- units. In many previous attempts to crystallize Sl by other investigators, these isoforms were separated, a process that requires removal of a second subunit, the regulatory light chain. Rayment and coworkers, however, retained both subunits, along with their heterogeneity. From the crystal structure of Sl, it is clear that retention of both sub- units was essential to produce a well-defined, crystallizable molecule. However, the crystallization required another crucial step. The solubility of the protein was changed by dimethyla- tion of essentially all the lysine residues; the modified, but not the unmodified, protein was capable of being crys- tallized. Although lysine dimethylation alters some of the kinetic constants of the actinS interaction, the modified Sl retains activity [3]. Dimethylation of the lysines of an- other protein, lysozyme, was shown to have little effect upon its structure [4]. It is noteworthy that this protein modification may be of general utility for inducing other recalcitrant proteins to crystallize. The structures of crystallized actin and myosin Sl have been determined to 2.8A resolution [ 1,2]. These structures alone answer many questions and provide fascinating insight into the functions of actin and myo- sin. However, research has proceeded beyond this step at a dizzying speed. The structures of these two proteins have now been fitted into macromolecular complexes to