852) and then as polymers created from a defined protein component (see “The discovery of tubulin” JCB 169:552). It took a series of studies in the 1980s to emphasize that dynamics were essential for microtubule action in the cell.
First, in vitro polymerization was needed as proof that there was no additional magic ingredient needed for microtubule formation. That achievement was forthcoming once EGTA was added to the mix to get rid of inhibitory calcium (Weisenberg, 1972). But this didn't exactly bring hordes rushing into the field of in vitro microtubule polymerization. Summers and Kirschner (1979) did find that microtubule growth was polar, occurring more readily at one end of the polymer than another. But it would be another 12 years after the initial in vitro method was reported before detailed in vitro studies led to the landmark theory of dynamic instability (Mitchison and Kirschner, 1984). The phenomenon was later observed in vivo in real time (Cassimeris et al., 1988).
The discovery of dynamic instability, in which microtubules grow persistently but suffer stochastic switches to catastrophic shrinkage, was only possible because of the observations of the dynamics of individual microtubules. Initially, says Tim Mitchison of Harvard Medical School, “the microtubule mafia were totally surprised and didn't really believe it.” If it was true, however, then these individual behaviors must be reshaping bulk microtubule populations in vivo. Just a month after Mitchison and Kirschner's paper, there were two reports indicating how important turnover and dynamic instability might be for the cell. Salmon et al. (1984) found that the half-time for spindle microtubule turnover (based on fluorescence recovery) is only ∼19 s, and Saxton et al. (1984) reported that microtubules turn over in minutes during interphase but seconds during mitosis.
One way that the turnover could be put to work is via microtubule flux—the poleward movement of microtubule subunits resulting from depolymerization at the pole balanced by polymerization at kinetochores. Bleaching experiments showed no sign of flux (Salmon et al., 1984), but it was possible that the rapid turnover of nonkinetochore microtubules was obscuring the flux of the less dynamic kinetochore microtubules. A direct demonstration of flux came when Mitchison (1989) reversed the contrast by making a photoactivatable, fluorescent derivative of tubulin. Encouraged by chemist David Trenton, “it was a fairly simple extension to think of turning fluorescence on instead of off,” he says.
The attached chemical group became fluorescent only when illuminated by light of a particular frequency. Mitchison used this light to mark a bar of fluorescence on an otherwise nonfluorescent spindle. Initially the bar faded as nonkinetochore microtubules turned over, but the remaining fluorescence moved steadily poleward. A flux-based force is still thought by many to contribute significantly to the poleward movement of chromosomes during anaphase (Rogers et al., 2004).