We have studied the contribution of the C-terminus of actin to filament stability by chemical modification and limited proteolysis. Formation of mixed disulfides of the penultimate C-terminal cysteine residue 374 with various low-molecular-mass thiols resulted in filament destabilization, as reflected by an increase in critical concentration and steady-state ATPase activity. These effects were fully reversed by the addition of phalloidin. Both the destabilization by glutathionylation and the reversal of it by phalloidin exhibited a high degree of cooperativity ; half-maximal destabilization required the modification of four out of five actin subunits, and half-maximal restabilization by phalloidin was already reached when only one out of 20 actin subunits was complexed. C-terminal truncation by limited trypsinolysis of filamentous actin resulted in a similar destabilization of the polymer, as shown by a 2-3-fold increase in the steady-state ATPase activity. This effect was likewise cooperative and could be reversed by phalloidin. Since truncation of the C-terminus of actin has an effect on stability similar to that of chemical modification with bulky substituents, the possibility can be excluded that, in the latter case, destabilization was caused by steric hindrance. Rather, it seems that the highly conserved C-terminal part of actin plays an active role in establishing a tight contact between neighbouring subunits.Under physiological conditions, actin exists as a long, apparently double-stranded helical array of globular monomers (reviewed in [l-31). Recently, atomic models of the filament were presented based on X-ray diffraction [4, 51and electron microscopy [6, 71. From these models, it was proposed that actin units in the filament exhibit strong longitudinal interactions but only tenuous diagonal interactions. It was shown that the C-terminus is probably localized at the bottom of the outer domain, not far from both contact sites [7]. The C-terminal residues of actin show a high degree of flexibility [8, 91. This may explain why, prior to crystallization of the DNAse complex, the three C-terminal residues, Lys373, Cys374 and Phe375, had to be removed by limited proteolysis [lo]. It may be that the C-terminal part of the molecule has a function in the assembly process. This is suggested by experiments with reporter groups attached to Cys374, which show that the environment of the C-terminus is changed during polymerization [8, 9, 11 -131.During our studies on the chemistry of the thiol groups of actin, we prepared some mixed disulfides of the protein with small thiol compounds [14-181. We observed that modification of Cys374, for example by formation of a mixed disulfide with glutathione, could greatly influence fila- Enzyme. Trypsin (EC 3.4.21.4). ment stability [15]. Based on the enhanced steady-state ATPase accompanying filament destabilization, the modification could be used to screen potentially filament-stabilizing agents and study their mechanism of action [16]. In a different approach, we succeeded ...