The model for the polymerization of ATP-actin proposed last year states that ATP hydrolysis occurs on the F-actin subsequent to the addition of the actin subunit. As a consequence of the fact that hydrolysis on F-actin is generally slower than the addition reaction, there exists a cap of ATP-actin subunits, even at steady state, and the length of the cap will vary with the G-actin concentration above the critical concentration. The model proposes that the hydrolysis of ATP occurs preferentially, it not exclusively, at the interface between the ATP cap and the ADP core, i.e. on an ATP-actin subunit adjacent to a more interior ADP-actin subunit. This vectorial or zipper hydrolysis predicts taht the initial rates of ATP hydrolysis and elongation as a function of G-actin concentration should be the same near the critical concentration but, as the G-actin concentration increases, the rate of ATP hydrolysis will become constant while the rate of elongation will continue to increase. This prediction has been verified for the polymerization of Mg-actin with an hydrolysis rate constant for the preferred site of 18 s/-1 and an additional very slow random hydrolysis in the ATP cap with a rate constant of 0.001 s/-1. With Ca-actin, however, ATP hydrolysis occurs randomly in the ATP cap and is always slower than the rate of elongation so that a long ATP cap builds up in proportion to the G-actin concentration. G-actin binds Ca-2+ with a K-D of 5 nM (304 orders of magnitude tighter than previously thought) and Mg-2+ with a K-D of 0.5 MuM. Ca-actin elongates at the same rate as Mg-actin but nucleates much more slowly. A new dimeric G-actin monomer-binding protein and a hexameric, ATP-sensitive F-actin crosslinking protein have been discovered in Acanthamoeba castellanii.
