Sutoh, T., Shimada, N., Sasaki, H., Asukagawa, and R. Ohkura
Department of Life Sciences, Graduate School of Arts and Sciences, University of Tokyo, Komaba, Tokyo 153, Japan
X-ray crystallography of the motor domain of Dictyostelium myosin II showed the similarity of the nucleotide binding sites of myosin and the Ras protein. In both proteins, three loop regions (the P-loop, the switch I, and the switch II) are essential components for nucleotide hydrolysis and for structural changes accompanying the hydrolysis. We carried out alanine-scanning mutagenesis on the switch I and switch II loops of myosin to understand how each residue in the regions is involved in the ATP hydrolysis and the structural changes induced by the hydrolysis. The target sequence of the switch I is 233NNNSSRFG240. The sequence is highly conserved among myosin families. Together with the P-loop, this stretch of the sequence form the surface of the ATP binding pocket. Multicopy vector was constructed for introducing mutant myosin II heavy chain genes into the myosin-null Dictyostelium cells so that the transformed cells express myosin II carrying each mutation. After transformation, we examined the following phenotypes of the transformants: 1) growth in suspension culture, 2) cell morphology on a plastic plate, 3) rate of clearing of bacterial cells on a plate, and 4) development after starvation. From these observation, we could identify three types of transformants. One type of the cells (N235A, S236A, F239A) had the phenotypes very similar to those of the wild type cells. The phenotypes of the second type of the cells (R238A) were very similar to those of the myosin-null cells. The last type of the cells (N233A, S237A) showed the phenotypes worse than those of the myosin-null cells. Myosins purified from the first type of the cells had rather high actin-activated ATPase activities and could drive sliding motion of F-actin, though at slower rate than the wild type myosin. Myosins from the second and the third types of the cells, however, showed low basal and actin-activated ATPase activities and could not drive the sliding motion. Given the fact that the N233A and S237A cells showed the dominant-negative phenotypes, it seems that the N233A and S237A mutations resulted in weaker binding of MgATP, and then slower ATP hydrolysis. On the other hand, it seems that the R238A mutation blocks the hydrolysis step directly. The target sequence of the switch II is 454DISGFE459. The stretch covers the bottom of the ATPase pocket. Among them, G457 is assumed to play a pivotal role as a g-phosphate sensor around which the upper and the lower 50K domains rotate during the transition from M/ATP to M/ADP/Pi. As above, we first examined phenotypes of the transformants. The I455 and S456 cells showed the phenotypes very similar to those of the wild type cells. On the other hand, the D454A and G457A cells showed the phenotypes similar to or worse than those of the myosin-null cells. Purified myosins from the former cells had rather high actin-activated ATPase activities and could drive the sliding motion of F-actin as expected from the fact that I455 and S456 are not conserved among the myosin superfamily. However those purified from the latter types of the cells had low basal and actin-activated ATPase activities and could not drive the sliding motion. Implication of these results will be discussed.