PerspectiveCell Biology

Persistence Pays

See allHide authors and affiliations

Science  13 Jun 2003:
Vol. 300, Issue 5626, pp. 1675-1677
DOI: 10.1126/science.1086055

Microtubules are polar cytoskeletal filaments that are crucial for intracellular motility, mitosis, and cellular organization. A key feature of microtubules is their ability to undergo rapid turnover, which results from the addition and loss of subunits at filament ends. Two types of dynamic turnover characterize microtubule arrays: dynamic instability, in which individual “plus” ends switch stochastically between phases of growth and shortening, and treadmilling, in which the rate of subunit addition at the plus end equals the rate of subunit loss at the opposite “minus” end (13). Microtubules of the interphase cortical array of plant cells are known to turn over more rapidly (4) than interphase microtubules in animal cells (5). In a research article on page 1715 of this issue, Shaw et al. (6) now provide a molecular explanation for this difference in dynamic behavior. By in vivo imaging of cells from the model plant Arabidopsis that express green fluorescent protein-tagged tubulin, the authors show that the faster turnover in the plant array results from a hybrid treadmilling mechanism. This mechanism is characterized by loss of subunits at minus ends and dynamic instability, with a bias toward subunit addition, at the plus ends. Microtubules in the cortical array grow and shorten more slowly (7) than microtubules in animal cells, but spend a much greater fraction of time in a dynamic state. Thus, like the fabled tortoise, faster turnover results from slow but persistent activity.

In a typical interphase animal cell, microtubules radiate from a perinuclear microtubule organizing center, the centrosome, composed of a pair of centrioles and associated γ-tubulin-containing ring complexes that enable microtubules to nucleate (see the figure, A). Once nucleated, microtubules grow in a highly persistent manner toward the cell periphery, with the more dynamic plus end leading (8). Near the cell edge, microtubule behavior changes and individual plus ends display dynamic instability (2). Most individual growth and shortening excursions are short in distance and duration, resulting in a dynamic population of microtubule plus ends that continually probe cellular space in the peripheral region.

Plants in motion.

(A) Microtubule arrays in animal cells with centrosomes and in plant cells without centrosomes. Centrosomal microtubule arrays are associated with nucleating structures in the pericentriolar material; a subset of microtubules is associated with the centriolar subdistal appendages (not shown). In noncentrosomal arrays, microtubules are nucleated at dispersed sites that may contain nucleating material. (B) After a microtubule is released from the nucleating site, it can (i) treadmill by balanced loss and gain of subunits from opposite ends of the filament, (ii) become capped at the minus end by complexes that bind to the end and/or sides of the filament and undergo dynamic instability at the plus end, or (iii) be transported by molecular motors or by linkage to motile elements of the cytoskeleton. Transported microtubules may be capped at one or both ends.

The contribution of the minus end to microtubule turnover has been more difficult to determine because the centrosomal region is highly congested, hence it is difficult to see individual microtubule minus ends. Nonetheless, microtubule release from the centrosome can be observed, and after release, minus ends move away from the centrosome by motor-driven microtubule transport, loss of subunits from the liberated minus end, or a combination of the two (9). Thus, a picture emerges that microtubules in animal cells show dynamic instability at the plus end, while a subset of microtubules disassemble completely after release from the centrosome.

The new work by Shaw et al. on the plant cortical microtubule cytoskeleton provides fresh insights into microtubule nucleation and minus-end behavior (6). The authors report the first observations of microtubule nucleation in the plant cell cortex. Nucleation occurs at dispersed sites; a single site can give rise to one, or occasionally several, microtubules that grow out in all directions. Most strikingly, once nucleated, these cortical microtubules do not remain associated with the nucleating structure but are released, as evidenced by subunit loss from the minus ends (see the figure, A). Thus, the cortical nucleating sites can overcome the unfavorable reaction to form a microtubule, but soon after formation, the microtubule is liberated. It will be of considerable interest to determine the molecular composition of the nucleating sites in Arabidopsis and the mechanism of microtubule release.

The new observations of microtubule behavior in plant cells, which lack a conventional centrosome, are remarkably similar to previous observations of animal cells treated to remove the centrosome. These cell fragments, or cytoplasts, contain fewer (and more disorganized) microtubules than the parental cells from which they were derived. In cytoplasts prepared from fibroblasts, microtubules with two free ends are observed, and these microtubules treadmill (10). Both in the plant cortex and the cytoplast, the apparent motion of the microtubule results from gain and loss of subunits at opposite ends; the lattice of the microtubule is stationary, presumably because of linkage to other elements of the cytoskeleton (see the figure, B). Interestingly, microtubule behavior in cytoplasts is cell type specific—microtubules in cytoplasts derived from epithelial cells have a static minus end and a plus end that shows dynamic instability (10). Minus-end capping complexes are likely to regulate microtubule behavior in these cells (see the figure, B).

In contrast to the tethered microtubules observed in the plant cortex, animal cells use motor-driven microtubule transport to establish and maintain noncentrosomal microtubule arrays in diverse cell types (see the figure, B) and to rearrange the microtubule array during cell locomotion (11, 12). This behavior is particularly striking in nerve axons, where microtubules are transported long distances by rapid but highly infrequent episodes of motility (11). In this case, the behavior is more like the proverbial hare, with bursts of activity followed by stasis or even reversal. The absence of cytoplasmic dynein from the Arabidopsis genome may partially explain the lack of microtubule transport, as well as the inability of microtubules in higher plant cells to organize their minus ends into tightly focused arrays.

Although the plus end of the microtubule has been the focus of much study, it is clear that a complete understanding of microtubule behavior will require an appreciation of what happens at the minus end as well. It will be particularly important to determine the mechanism by which microtubules are nucleated in vivo, and to learn which aspects of this process are conserved in diverse cells. A better understanding of microtubule nucleation may also shed light on the mechanism of release and on how cells regulate its frequency. For example, recent observations of motile fibroblasts have shown that microtubule release from the centrosome is more frequent than previously estimated and that this release contributes to cell motility (13). Clearly, much remains to be learned about how plant and animal cells generate microtubule arrays with distinct dynamic properties by regulating microtubule nucleation, release, and dynamic turnover.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
View Abstract

Stay Connected to Science

Navigate This Article