A microtubule assembly/disassembly model based on drug effects and depolymerization kinetics after rapid dilution.

One of two models for microtubule assembly/disassembly involves a condensation equilibrium with microtubule formation occurring only when the microtubule protein concentration exceeds a critical or threshold level, and the state of polymerization is subject to any change in the level of assembly-competent tubulin dimers. The second has been called a “treadmilling” process with assembly and disassembly taking place predominantly, and possibly even exclusively, at opposite ends of each tubule. We present evidence here for a combined model which incorporates the essential features of each model. By rapidly diluting assembled microtubules to a level below the critical concentration, we have obtained a rate constant of 1 min-’ for the first order loss of polymerized tubulin. This value exceeds the rate constant for treadmilling in the steady state by about lOOO-fold, suggesting that dimer release from the assembly end is relatively rapid. Podophyllotoxin, a poison of microtubule assembly at 50 pM does not inhibit the disassembly rate at steady state of assembly or upon dilution, but 5 to 10 PM tubulin/podophyllotoxin does. This is consistent with the capping of the assembly end by the drug* tubulin complex, but the residual 0.1 h-’ rate of disassembly observed in the presence of the complex agrees well with earlier values of 0.07 h-’ (Margolis, R. L., and Wilson, L. (1978) Cell 13,1-S) for the treadmilling process. Thus, the data are consistent with the observed polarity of drug action of microtubules. These and other findings lend credence to a unified or combined “equilibrium-treadmilling” model.