Assembly of tubulin from cultured cells and comparison with the neurotubulin model

Spontaneous microtubule assembly can be obtained in extracts from a variety of cultured cell lines by including glycerol in the assembly buffer. An analysis of the effects of cultured cell extracts on brain tubulin (neurotubulin) assembly has shown that the extracts contain initiation inhibitors whose effects are diminished by glycerol. By using glycerol during the assembly step, cultured cell tubulin can be purified by assembly-dissassembly procedures. The amount of glycerol necessary for significant spontaneous assembly varies from 1–6 M among the different cell lines and is dependent upon their content of inhibitor. Comparison of the assembly products obtained from NA, C₆ and CHO cells at increasing glycerol concentrations shows that glycerol enhances the purification of tubulin and a polypeptide of molecular weight 49,000 daltons in all three systems. These preparations contain a number of other polypeptides, including a group with gel electrophoretic mobilities characteristic of tau-factor, but lack the high molecular weight microtubule-associated proteins (MAPs) which are present in neurotubulin preparations. Phosphocellulose chromatography of NA tubulin removes several proteins from the tubulin and results in a loss of polymerizability. Among three proteins which are completely removed from the inactive tubulin, the most prominent is the 49K protein. This observation and the co-purification of the 49K protein with tubulin through two assembly-disassembly cycles suggest that it is a true MAP. The difference in MAP proteins between brain and tissue culture cells is parallelled by an absence of ring structures in NA tubulin preparations. NA tubulin, however, does form rings when brain MAPs are added. The early steps of NA tubulin assembly differ from those of neurotubulin; no rings are involved, and the first assembly intermediates are straight protofilament bundles. The differences between MAPs from cultured cells and brain and the absence of ring formation in NA tubulin preparations suggest that the assembly model based on neurotubulin is not completely general. A comparison of extracts from CHO cells grown with and without dibutyryl cAMP revealed no differences between the behavior of these extracts in spontaneous tubulin assembly or in mixture experiments with brain tubulin.