Most myosins have a positively charged loop 2 with a cluster of lysine residues that bind to the negatively charged N-terminal segment of actin. However, the net charge of loop 2 of very fast Chara myosin is zero and there is no lysine cluster in it. In contrast, Chara myosin has a highly positively charged loop 3. To elucidate the role of these unique surface loops of Chara myosin in its high velocity and high actin-activated ATPase activity, we have undertaken mutational analysis using recombinant Chara myosin motor domain. It was found that net positive charge in loop 3 affected Vmax and Kapp of actin activated ATPase activity, while it affected the velocity only slightly. The net positive charge in loop 2 affected Kapp and the velocity, although it did not affect Vmax. Our results suggested that Chara myosin has evolved to have highly positively charged loop 3 for its high ATPase activity and have less positively charged loop 2 for its high velocity. Since high positive charge in loop 3 and low positive charge in loop 2 seem to be one of the reasons for Chara myosin's high velocity, we manipulated charge contents in loops 2 and 3 of Dictyostelium myosin (class II). Removing positive charge from loop 2 and adding positive charge to loop 3 of Dictyostelium myosin made its velocity higher than that of the wild type, suggesting that the charge strategy in loops 2 and 3 is widely applicable.actin ͉ ATPase ͉ motility ͉ cytoplasmic streaming ͉ molecular engineering C ytoplasmic streaming in characean algal cells is extremely fast, and this streaming is brought about by the movement of myosin-coated organelles along actin filament bundles fixed inside the cell (1). Myosin purified from Chara corallina could translocate actin filaments in the in vitro motility assay at a velocity comparable to that of the cytoplasmic streaming (approximately 50 m s Ϫ1 ) (2, 3). This velocity is about 10 times faster than that of the fast skeletal muscle myosin and the Chara myosin is the fastest motor protein known so far. We have cloned cDNA of the Chara myosin heavy chain (4) and succeeded in expressing functional motor domain (5, 6). The velocity of the expressed motor domain measured by in vitro motility assay was comparable to that of the native Chara myosin if we consider the difference in the lever arm length. Relevant steps in actomyosin ATPase cycle to the sliding velocity are ADP release from actomyosin and ATP reassociation resulting in actin-myosin dissociation. Kinetic analyses of Chara myosin motor domain revealed that both the ADP release and the ATP-induced dissociation from actin were very fast (6). Time spent in the strongly bound state with actin, which was estimated from these rates, was less than 1 ms. This very short strongly bound state, together with a large step size (7), are probably the reason for the very high velocity of Chara myosin. Actin-activated ATPase activity of expressed Chara motor domain was approximately 500 s Ϫ1 head Ϫ1