Microtubules self-repair in living cells

Abstract

Microtubule self-repair has been studied both in vitro and in vivo as an underlying mechanism of microtubule stability. The turnover of tubulin dimers along the microtubule network has challenged the pre-existing dogma that only growing ends are dynamic. However, although there is clear evidence of tubulin incorporation into the shaft of polymerized microtubules in vitro, the possibility of such events taking place in living cells remains uncertain. In this study, we investigated this possibility by microinjecting purified tubulin dimers labeled with a red fluorophore into the cytoplasm of cells expressing GFP-tubulin. We observed the appearance of red dots along pre-existing green microtubule network within minutes. We found that the fluorescence intensities of these red dots were inversely correlated with the green signal, suggesting that the red dimers were incorporated into the microtubules and replaced the pre-existing green dimers. We then characterized the size and spatial frequency of these incorporations as a function of injected tubulin concentration and post-injection delay. The saturation of these measurements contradicted the hypothesis of nonspecific adsorption along microtubules and suggested that the injected dimers incorporated into a finite number of damaged sites. By our low estimate, within a few minutes of the injections, free dimers incorporated into major repair sites every 70 micrometers of microtubules. Finally, we mapped the location of these sites in micropatterned cells and found that they were more concentrated in regions where the actin filament network was less dense and where microtubules exhibited greater lateral fluctuations. These results provide evidences that microtubules do self-repair in living cells, and they provide a quantitative characterization of the temporal and spatial dynamics of this process in PtK2 cells.