Rate constants for actin polymerization in ATP determined using cross-linked actin trimers as nuclei.

Chemically cross-linked actin trimer, prepared from F-actin covalently cross-linked by N-N'-p-phenylenebismaleimide, was used to nucleate the assembly of actin polymers under a variety of ionic conditions at 25 degrees C. Gel-filtered G-actin (5% labeled with N-pyrenyl iodoacetamide) was prepared in buffers containing 5 mM Tris X HCl, 0.2 mM ATP, 0.05 mM MgCl2 or 0.1 mM CaCl2, 0.2 mM dithiothreitol, and 0.01% sodium azide, pH 7.5. Polymerization was initiated by adding salt to a solution of actin monomer (2-25 microM) mixed with actin trimer (0-24 nM). Pseudo-first-order polymerization kinetics were observed in all cases following a lag phase of 30-75 s. The length of the lag phase was independent of trimer concentration, monomer concentration, or preincubation of monomer or trimer in polymerizing buffer. A simple model is presented which explains all of the observed features of the lag phase. Lower limits for the sums of the elongation rate constants at the two filament ends, calculated assuming each trimer acts as a seed, were in the range 1.4-5.2 microM-1 s-1, and lower limits for the polymer dissociation rate constants (calculated from the elongation rate constants and the critical actin concentration) were 0.4-1.9 s-1, depending on ionic conditions. The observed rate constants were independent of actin concentration or trimer concentration under a given ionic condition. These polymerization rate constants are in good agreement with the values others have obtained by measuring the growth of actin filaments by electron microscopy.