Dynamics of tubulin and calmodulin in the mammalian mitotic spindle.

Microtubules (MTs) have attracted interest and experimental attention in part because they are involved in numerous cellular processes and in part because they are intriguingly dynamic. Many studies with light and electron microscopes have shown that MTs can change their location and extent of polymerization, depending on the physiological state of the cell. Efforts to study this dynamism and the cellular factors that control it have, however, approached the problem either indirectly or with low time resolution. For example, one can look a t the changes in the birefringent retardation of the mitotic spindle as a result of anaphase or of some experimental treatment. Such observations are, however, indirect assays of tubulin assembly, because birefringence depends on the orientation and bunching of spindle MTs in addition to the amount of polymer and of other birefringent components present. One can directly study the effects of a given cellular perturbation on MT assembly by fixation and subsequent immunofluorescence, but such investigations are necessarily of low time resolution. MT assembly can be studied quickly and directly in vitro, but the systems of buffers and the proteins used for chemical analysis may not be an accurate reflection of the state within the cell. To study the assembly of MTs in their natural context, one wants a way to look directly a t the polymerization behavior of tubulin in cells. Fluorescence microscopy of d ye-tagged molecules, or fluorescent analogue cytochemistry, is a powerful way to follow the polymerization dynamics of specific proteins in vivo.' The requirements of the method are well defined: 1) a fluorophore-conjugated protein that is indistinguishable from the native macromolecule by all available functional tests, 2) a way to introduce the labeled protein into appropriate cells, 3) sufficient fluorescence signal to permit localization of the labeled protein a t adequate space resolution, and 4) a device for following the fluorophore that can quantify the observed fluroescence. This approach to the study of tubulin was first used by Keith, Feramisco, and Shelanski2 who recognized the suitability of dichlorotriazinyl aminofluorescein (DTAF) as a fluorophore to label tubulin. They showed that DTAF tubulin will form cytoplasmic fibers that resembled MTs in their distribution and response to colchicine. Wadsworth and Sloboda' confirmed these findings in a study of DTAF tubulin microinjected into sea urchin eggs, noting particularly the speed with which the fluoresecent tubulin analogue would incorporate into spindle-like structures. Our laboratory has been investigating tubulin dynamics of living cells in collabora-