The diversity of cytoplasmic microtubules

Microtubules (MTs) were first identified by electron microscopy of flagella and cilia. Cytoplasmic MTs, however, were seen only after glutaraldehyde was introduced as a fixative (Sabatini et al., 1963). Most early studies of MTs were descriptive, listing where they were seen in different cells. Because all MTs looked alike, it was assumed that all were the same. Their functions were unknown, but because of their cellular locations it was assumed that MTs might be involved in cell motility or cell structure. Articles by Pickett-Heaps and Northcote (Pickett-Heaps and Northcote, 1966a; Pickett-Heaps and Northcote, 1966b), Tilney (Tilney, 1968), and Behnke and Forer (Behnke and Forer, 1967), in early volumes of Journal of Cell Science, were among the first to do more than just describe the presence of MTs. They started us towards our present ideas about MTs and raised questions that remain unanswered today.

Pickett-Heaps and Northcote deduced a potential function of specific preprophase MTs in wheat mitotic cells (Pickett-Heaps and Northcote, 1966a; Pickett-Heaps and Northcote, 1966b). Searching for cytoplasmic ultrastructural changes indicating that the cells were preparing to enter division, they saw nothing unusual at the ends of the cells where the spindle poles were going to form (Pickett-Heaps and Northcote, 1966a). In the plane of the future cell plate, however, they observed a ring of more than 150 MTs, three to four MTs deep, encircling the cells just under the cell wall. This `preprophase band' of MTs disappeared as the spindle MTs began to appear and, in these symmetrically dividing cells, seemed to predict the cell plate position. To test this, Pickett-Heaps and Northcote studied the three successive asymmetric divisions of wheat stomatal complex cells, in which two of these divisions are perpendicular to the first and the positions of all future cell plates can be predicted from the positions of organelles in the adjacent cells (Pickett-Heaps and Northcote, 1966b). In every pre-prophase cell in the stomatal complex, they found a band of MTs near the cell wall where the cell plate subsequently forms. Because changed positions of the cell plates were reflected by changed positions of the preprophase bands, there could be little doubt that the pre-prophase band indicated the future cleavage plane.

The preprophase band has been studied a great deal since this first description – a Google search yields over 25,000 hits. We now know that actin filaments are present in the preprophase band (e.g. Mineyuki and Palevitz, 1990; Cleary et al., 1992), and that actin and myosin cause it to narrow after it forms initially (e.g. Eleftheriou and Palevitz, 1992; Granger and Cyr, 2001; Li et al., 2006), presumably by interacting with the MTs. Kinases also are involved (e.g. Katsuta and Shibaoka, 1992; Nogami and Mineyuki, 1999), but what causes the band to form at this position and what exactly it signifies still are not known.

Shortly after these articles on the preprophase band, Tilney (Tilney, 1968) described experiments on MTs in Actinosphaerium axopodia that were among the first to suggest that MTs function in both structural roles and force generation. Earlier, Tilney and Porter (Tilney and Porter, 1965) had shown that the thin, 250- to 300-μm long axopodia of Actinosphaerium contain an array of two intertwined 12-sided coils of evenly spaced MTs that are separated by `flocculent' material. Each axopodial MT is straight and extends lengthwise along the axopod; some extend from end to end of the axopodial `rod', but others end before the tip as the width narrows. Tilney (Tilney, 1968) assumed that all MTs have similar properties and thus he treated Actinosphaerium with colchicine, an agent known for its effects on spindle structure. Because colchicine caused depolymerization of axopodial MTs and retraction of the axopodia, and because both axopodia and MTs reformed after colchicine was removed (see Fig. 1), Tilney concluded that MTs are involved in both axopod maintenance and axopod formation, and thus that MTs have structural and force-producing roles (Tilney, 1968).

Subsequent experiments have indicated that, in protrusions emanating from various cells, MTs exert an outward force that is resisted by actin and myosin (e.g. Solomon and Magedantz, 1981; Joshi et al., 1985; Madreperla and Adler, 1989; Ahmad et al., 2000). Other experiments (Mitchison et al., 2005) showed that, as spindles shrink after treatment with a MT-depolymerizing drug, spindle MTs become bent by forces from a spindle `matrix'. Because Actinosphaerium MTs also bend during colchicine treatment (Tilney, 1968), similar `matrix forces' might act in the axopods, and might involve actin and myosin, both of which interact with MTs in a variety of systems (e.g. Waterman-Storer and Salmon, 1997; Waterman-Storer et al., 2000; Yvon et al., 2001; Mandato and Bement, 2003; Rodriguez et al., 2003; Weber et al., 2004; Fabian et al., 2007). Indeed, in some longitudinal sections, the material between the axopodial MTs looks similar to arrowhead-labeled actin (Tilney and Porter, 1965).

How the double-coiled MT arrangement is generated is still not known. Tilney (Tilney, 1968) speculated that the cytoplasm might have some epigenetic `memory', as it does in other systems (e.g. Beisson and Sonneborn, 1965) [discussed in Frankel (Frankel, 1989)], but, like the preprophase band, what causes this remarkable structure to form, and what the pattern signifies, remain enigmas.

When cytoplasmic MTs were being described, we knew nothing of their molecular components; as I have already mentioned, the prevailing assumption was that MTs are all the same. To see whether this was in fact the case, Behnke and I (Behnke and Forer, 1967) subjected MTs from crane-fly spermatids, rat sperm and rat tracheal cilia to various treatments, both in tissue and outside the cell. When whole cells were treated, cytoplasmic MTs responded differently from all axoneme MTs. Moreover, in the axoneme, the A-tubules responded differently from the B-tubules, and both were different from the two central MTs. There were similar differences when MTs were treated outside the cell, and when sections of cells were treated with pepsin, and there were differences along the lengths of the MTs (Behnke and Forer, 1967). We concluded that, despite their similar appearances, not all MTs are the same – that there can be `intrinsic physical and/or chemical differences among the tubules themselves'.

There have been giant strides in our understanding of MT chemistry since then. We now know that all MTs have the same basic substructure of α-tubulin–β-tubulin heterodimers; that they exhibit `dynamic instability' (e.g. Nogales and Wang, 2006); that motor proteins transport cargo along MTs and occasionally double as enzymes that depolymerize them (e.g. Helenius et al., 2006; Moores and Milligan, 2006; Howard and Hyman, 2007); that MTs can be severed by enzymes such as katanin (e.g. Baas et al., 2005); and that bacteria, for a long time thought to have no cytoskeleton, contain relatives of tubulin (e.g. FtsZ) (e.g. Gitai, 2007). But we also know that there are chemically unique tubulin subunits and different subunit isotypes, which confirms the earlier conclusion that there are chemical and physical differences between MTs. For example, MTs formed from different β-tubulin isotypes differ in their solubility (e.g. Verdier-Pinard et al., 2003) and their function in axonemes (e.g. Raff et al., 1997). The different α-tubulin isotypes and post-translational modifications confer different properties (e.g. Matsuyama et al., 2002), different positions within cytoplasmic MT arrays (e.g. Venkei et al., 2006) and specific localizations in cells (e.g. Walss-Bass et al., 2001).

The articles on MTs by Pickett-Heaps and Northcote (Pickett-Heaps and Northcote, 1966a; Pickett-Heaps and Northcote, 1966b), Behnke and Forer (Behnke and Forer, 1967) and Tilney (Tilney, 1968) published in the early issues of Journal of Cell Science were among the first to do more than just describe MTs; they provided experimental insights into various aspects of MTs and their behavior. Crucially, they gave impetus to further experimentation. We have come a long way since 1968, but we are still grappling with some questions raised then. For example, how do the different morphological arrangements of MTs arise, and what do these signify? And how do the different cytoskeletal systems – MTs, actin, intermediate filaments – interact? We have come to recognize that there is crosstalk between the different systems; we now need to understand the language and its grammar.