The mechanism of microtubule assembly in vitro

The mechanism of microtubule polymerization and depolymerization has been studied with protein purified from extracts of porcine brain. Under polymerizing conditions characteristic microtubules composed of parallel protofilaments are observed in the electron microscope. Under depolymerizing conditions three forms are observed: double rings of outside diameter 49 nm, spirals, and 7‐nm globular subunits. Under the same conditions two boundaries are observed in the analytical ultracentrifuge at 6S and 36S, whether depolymerization is accomplished by cooling to 0°C, by addition of 1 mM CaCl 2 at 25°C, or by removal of GTP. On polymerization all of the 36S and most of the 6S is converted to a fast‐sedimenting form which the electron microscope reveals to be microtubules. The depolymerization mixture may be fractionated by gel chromatography into two fractions, one consisting solely of 6S and the other mostly 36S. Neither fraction regenerates the original equilibrium mixture. The 36S form may be reversibly dissociated into 6S subunits by addition of NaCl. From these and other considerations we have postulated that microtubule protein is composed of two different types of tubulin, both of which participate in polymerization. Studies are reported showing that colchicine does not dissociate microtubule rings but blocks polymerization by interfering with their proper lateral association into a protofilament array within microtubules. The role of GTP in polymerization is also discussed. Electron micrographic evidence is presented suggesting the conversion of protofilaments directly into rings and spirals, and a pathway for microtubule assembly is proposed.