Mandelkow CO CellBiol95

Microtubules and microtubule-associated proteins Eckhard Mandelkow and Eva-Maria Mandelkow

Max Planck Uni t for Structural Molecular Biology, Hamburg, Germany

Microtubule research is becoming increasingly diverse, reflecting the many isoforms and modifications of tubulin and the many proteins with which microtubules interact. Recent advances are particularly visible in four areas: microtubule motor proteins (their structures, stepping modes, and forces); microtubule nucleation (the roles of centrosomes and y-tubulin); tubulin folding (mediated by cytoplasmic chaperones); and the expanding list of microtubule-associated proteins, knowledge of their phosphorylation states,

and information on their effects on microtubule dynamics.

Current Opinion in Cell Biology 1995, 7:72-81

Introduction

Microtubules represent one of the fiber systems of the eukaryotic cytoskeleton. They are essential for a wide variety of cellular functions, notably cell motility, transport, cell shape and polarity, and mitosis. Microtubules consist of a core cylinder built from heterodimers of et- and ~-tubulin monomers. Three main classes of proteins interact with tubulin. The 'inicrotubule-associated proteins', or MAPs, are sometimes called more specifically 'structural' MAPs because they bind to, stabilize and promote the assembly ofmicrotubules, and because they can be copurified with tubulin through several cycles of microtubule assembly and disassembly. Representatives are MAPla and lb, MAP2a, 2b, and 2c, MAP4, tau protein, 205 kDa MAP, and isoforms of these proteins that are often generated by alternative splicing.

Another broad class comprises the motor proteins, so called because they generate movement along microtubules using the chemical energy of ATP hydrolysis. Representatives are kinesin (a motor moving towards the distal, or plus, end of microtubules), dynein (a rearward motor, also involved in the bending of cilia and flagella), and their many relatives. The motors can bind tightly to microtubules, for example in the absence ofnucleotides, but they do not copurify through cycles of assembly and disassembly.

A third and more heterogeneous class includes proteins that are not normally called MAPs but are often found associated with microtubules and inay even copurify with them. Examples are glycolytic enzymes (e.g. GAPDH and aldolase), kinases (e.g. protein kinase A, GSK-3 and c-mos), proteins involved in biosynthesis (elongation factor EF-lCt and even entire ribosomes), proteins linking up to membrane receptors (dynamin and G proteins), and ribonucleoproteins carrying mKNA. These interactions are not always well defined, but they

point to the cytoskeleton as a (transient) anchor for many cytoplasmic proteins.

The fields of microtubules, MAPs, and motors have ex- panded so nmch that they have become the subject of separate reviews. For example, microtubule structure, assembly and regulation have been reviewed recently [1-4], as have MAPs [5,6] and motors [7-9]. General surveys of microtubules and related proteins are given in [10,11].

This review covers selected articles concerning microtubules, MAPs and motor proteins published in the last eighteen months.

Structural aspects of microtubules revealed by motors

The information on microtubule structure we have today is derived mostly from electron microscopy combined with image reconstruction, X-ray scattering, and video microscopy; information on tubulin structure is indi- rect and derived mostly from biochemical experiments (binding sites ofnucleotides or drugs, cross-linking studies, and so on). Novel insights on the microtubule lattice have recently come from studies of the interaction between microtubules and kinesin. The head domain of kinesin binds to [3-tubulin with a stoichiometry of one kinesin head per tubulin heterodimer, generating an axial periodicity of 8 nm (the height of the dimer) and a 'B'-lattice in which adjacent protofilaments are staggered by about 0.9 nm [12°-14"]. The dimer polarity is such that ~-tubulin points to the plus (fast-growing) end of the microtubule and [3-tubulin points towards the minus (slow-growing) end (see Fig. 1). This implies that the plus end has a crown ofet-tubulin subunits and the minus end terminates with [3-tubulin [14°]. This polarity en- ables the ~-subunits to interact with T-tubulin, which is a key component ofmicrotubule-organizing centers (reviewed in [3]). As the exchangeable GTP-binding site is

Abbreviations GAPDH--glyceraldehyde-3-phosphate dehydrodgenase; GSK-3--glycogen synthase kinase-3; MAP~microtubule-associated protein.

72 © Current Biology Ltd ISSN 0955-0674

Microtubules and microtubule-associated proteins Mandelkow and Mandelkow 73

on ~-tubulin and the microtubule-bound GTP is at the plus end [15°], another consequence of this arrangement of the microtubule is that the 'GTP-cap' sits on the plus end in an inverted fashion so that the last ~-tubulin-GTP is buried by a terminal layer of et-tubuhn (see Fig. 1).

Kinesin moves along the microtubule in 8 nm intervals and produces forces in the pN range [16°-20"]. The direction of movement is parallel to the tracks of protofilaments [21], and the step size matches the separation of the ~-tubulin 'stepping stones'. As kinesin works as a dimer in vivo and shows alternating head catalysis (using each head of the dimer alternately to hydrolyze ATP) [22°], one could integrate the structural and kinetic data by assuming that a kinesin dimer would 'walk' or 'hop' along one or two adjacent protofilaments. (Note that if the center of the kinesin molecule advances by 8 nm, this may imply 16 nm steps for each head.) It also seems that kinesin prefers rigid stepping stones over wobbly ones: microtubules are fairly stiff structures, with persistence lengths (a measure of their straightness) in the mm range [23-25], but it is possible to make them even stiffer using slowly hydrolyzable GTP analogues such as GMP-PCP, in which case the movement of kinesin becomes 30% faster [26].

The microtubule-motor interaction is surprisingly ver- satile: there are variants ofkinesin such as Drosophila ncd or Saccharomyces cerevisiae Kar3 that can walk in the op- posite direction (towards the minus end), and this property is inherent to the head domains [27°,28,29]. The motor can twist its neck in order to attach to microtubules in the proper direction [30], and may even cause rotation of microtubules instead o f translation, as is sug- gested by the presence ofkinesin-like proteins on the C2 central microtubule in flagella in Chlamydomonas [31]. Finally, CENP-E, a kinesin-like protein that migrates from kinetochores to the midzone of mitotic spindles, cross-links the antiparallel microtubules and presumably controls their rigidity or ghding during anaphase [32°].

Tubulin structure and interactions

Drugs that target tubulin Tubulin has many applications as a target of drugs (e.g. anti-cancer drugs) and therefore is a workhorse for certain drug screening tests. The different modes of ac- tion and binding sites on tubulin of various drugs con- tinue to be explored (see reviews [33,34]). Colchicine and vinblastine are examples of drugs that destabilize microtubules and thus disrupt mitosis; however, their effect can be felt even at drug concentrations too low to cause net microtubule disassembly. This can be attributed to a pronounced effect of the drugs on microtubule dynamic instability parameters [35-37], which are closely linked to GTP binding and hydrolysis (see below). The binding site for colchicine has been mapped near the amino terminus of ~-tubulin [38]. By contrast, taxol stabihzes microtubules, thereby also disrupting mitosis. It was shown using photoactivatable taxol derivatives that taxol binds predominantly to the amino-terminal

region of ~-tubulin [39,40]. The stoichiometry of taxol binding is one molecule per tubulin dimer; dimers sta- bilized by taxol are capable of assembling with bound GDP, without the usual need for GTP hydrolysis [41].

GTP binding to tubulin Microtubules contain a non-exchangeable GTP on ct-tubulin and an exchangeable one on ~-tubulin (nucleotides on the latter site can be exchanged with nucleotides from the solution). Extending earlier work, two recent studies showed that the exchangeable GTP site is in the amino-terminal domain of [~-tubulin; an additional ATP-binding site (not involved in microtubule assembly) was found on et-tubulin [42,43]. That microtubule stability is closely linked to GTP binding and hydrolysis has now been demonstrated directly by Davis et al. [44"], who made several mutant tubulins in which altered GTPase activity is paralleled by altered dynamic instability. Similarly, Reijo et al. [45 °] made systematic mutations of clusters of charged amino acids for alanine in yeast [~-tubulin and deduced that three regions are critical for ~-tubulin function: near the carboxyl terminus, around residue 330, and around residue 150 (near the glycine cluster that is probably involved in GTP binding).

Modifications of tubulin An enigmatic feature of tubulin is its heterogeneity. Not only are et- and [~-tubulin each encoded by up to six different genes, but the protein can also be modified in several ways: phosphorylation, acetylation, detyrosi- nation, and glutamylation. The complexity of neuronal tubulins is particularly great and appears to increase with differentiation. In other cells, such as avian erythrocytes, the degree o f heterogeneity is conspicuously low [46]. One function of neuronal tubulin modification may be the selective stabihzation of certain microtubules [47]. One of the modifying enzymes, the tubulin tyrosine ligase, has now been sequenced and characterized in detail [48]. This enzyme restores a carboxy-terminal tyrosine to et-tubulin after it has been cleaved by a tubulin carboxypeptidase. However, et-tubulin can also lose two carboxy-terminal residues (glutamic acid and tyrosine). This A2-tubulin can no longer be retyrosi- nated by the ligase and is enriched in very stable neuronal microtubules [49]. Another sign of neuronal differentiation is the increasing length ofpoly-glutamic acid chains attached to glutamic acid residues near the carboxyl terminus of 0t- and ~-tubulin; this makes the already very acidic carboxyl termini even more negatively charged [50]. The functions of these post-translational modifications have been ditlqcult to ascertain, but Gurland and Gundersen [51] found that long-lived stable microtubules (known as 'Glu-microtubules' because they have lost the carboxy-terminal tyrosine from ct-tubulin) selectively break down when phosphatases are inhibited by okadaic acid. One interpretation of this observation is that stabilization factors (caps) that are normally associated with stable microtubules can be lost as a result of unchecked phosphorylation.

74 Cytoskeleton

(D 1995 Current Opinion in Cell Biology

cz-tubulin O [3-tubulin-GTP 0 J3-tubulin-GDP

Fig. 1. Several features of the microtubule structure deduced from labeling microtubules with the kinesin motor domain (for details, see [14"]). The plus (fast-growing, distal) end is up, the minus end (slow-growing, proximal) is down. The minus end is normally buried in the microtubule-organizing structure containing y-tubulin (not shown here). An ct[B-tubulin heterodimer is shown near the top (height 2 x 4 = 8 nm). The plus end of a microtubule terminates with a crown of (~-tubulin, the minus end with a crown of ifl-tubulin. [3-Tubulin carrying GTP occurs in solution and at the plus end [15"], but not in the interior of the microtubule. Note that the [3- tubulin-GTP cap is covered by the terminal layer of ct-tubulin. The protofilaments are shifted axially by 0.9 nm, corresponding to the 'B4attice' which results in a discontinuity in the microtubule wall (moved to the back and thus not shown in this diagram). In the fully saturated microtubule lattice every tubulin heterodimer binds one kinesin head, thus emphasizing the dimer lattice of microtubules. Native kinesin is a dimer of two heavy chains, linked via a flexible neck and a coiled-coil stalk (simplified here as a V-shape), and two light chains. The diagram shows three hypothetical modes of at- taching the kinesin dimer to the microtubule: (a) on two adjacent protofilaments side by side; (b) on one protofilament (kinesin heads displaced axially by one tubulin dimer, or 8 nm); (c) on two adjacent protofilaments (heads staggered by one tubulin dimer). Each of these modes would be compatible with the observed structure of the microtubule decorated with kinesin.

Regulation of tubulin synthesis The level of tubulin synthesis is regulated via the soluble tubulin pool, whose size is in turn coupled to the degree of microtubule assembly. In the case of ~-tubulin this autoregulation involves the four amino-terminal residues (MR_El, in the single letter code for amino acids) [52]. In an extension of their earlier work, Cleveland and co- workers [52] showed that 0t-tubulin follows a different pathway (even though both subunits are downregulated by degrading their mRNAs).

Prokaryote homologs of tubulin? Tubulin is thought to be restricted to eukaryotic cells, but there is a curious homology of the conserved glycine-rich cluster (GGGTGSG) to the bacterial FtsZ protein from Escherichia coll. Like tubulin, this protein is also involved in cell division and binds GTP, and Bramhill and Thompson [53] have now shown that it even polymerizes in a GTP-dependent fashion, suggesting that the similarity is more than accidental, even though the overall homology is very low.

Centrosomes and chaperones

Components of the centrosome For microtubules to operate properly they must be nucleated at the right time and in the right place. This normally happens at the centrosome, which consist of a pair of centrioles surrounded by periocentriolar material; the latter is where microtubule nucleation actu- ary occurs. The stages and components of centrosome assembly are beginning to be understood. One of the components is 7-tubulin, which exists in the cytoplasm as part of a multiprotein complex ('gammasome') with several other proteins. 7-Tubulin is recruited to the centrosome for nucleation of microtubules at their minus ends; nucleation is blocked when 7-tubulin is depleted or inactivated. Other components of the centrosome pre- cursor include ct- and ~-tubulin, actin, and the molecular chaperone hsp70 [54"-57"]. Additional proteins are required to assemble the pericentriolar material, such as pericentrin, an elongated 220kDa protein with a coiled-coil stalk and two globular ends, identified using antisera from patients with scleroderma [58"]. This general domain structure is shared with other proteins found in microtubule-organizing structures, such as the SP-H antigen/NuMa [59], and variants of tektin, an (~-helical protein first found in flagellar axonemes and recently in centrosomes [60]. The microtubule-nucleating function of 7-tubulin in centrosomes appears to be highly conserved as human 7-tubulin functions even when introduced into yeast [61].

Chaperones Before microtubules can be nucleated, the tubulin must first be folded; this is achieved by another oligomeric complex, the cytoplasmic chaperone complex TCP-1,


also known as TriC. This 800 kDa structure contains several homologs of the heat shock protein Hsp60 (termed CCT0t, -~, and so on) arranged in the two-layered ring structure also found in bacterial and mitochon- drial chaperones; this complex assists in the folding of relatively few proteins, among them the three tubulins and actin [62"-64"]. TCP-1 subunits are the cytosolic equivalents of the bacterial chaperone GroEL and mito- chondrial Hsp60. The folding takes place in several steps such that the initial association occurs with ADP bound to the complex, followed by a later release involving ATP hydrolysis [65"]. Yeast have equivalent chaperones (BIN2, BIN3 and so on); even though they are related, they appear to serve somewhat different functions and do not replace one another [66"].

Location of assembly and transport A related and more specialized problem is the nucleation of microtubules in nerve cells, whose long processes are filled with microtubules that serve as tracks for intracellular transport. Are the microtubules assembled in the axon or in the cell body? The latter appears to be the case: nucleation sites containing ~,-tubulin are found only in the cell body, and microtubules are then transported into the axons or dendrites (see below). Superimposed on this transport process is a length redistribution which allows microtubules to change their length, presumably by dynamic instability, in the axon. The plus ends of the transported microtubules are oriented distally, and this end is where incorporation of new subunits can occur [67].

Microtubule-associated proteins

The majority of studies have traditionally been done with brain MAPs, partly because brain tissue was the most abundant source of microtubules, and partly because the MAPs provide selective markers for brain development and cell type. The best known MAPs are the heat-stable proteins MAP2 and tau; they are localized (mainly) to the dendritic and axonal compartment, respectively, and they share homologous repeats in the carboxy-terminal region that contribute to microtubule binding (reviewed by Schoenfeld and Obar [51).

Tau Tau is thought to stabilize microtubules in axons and thus provide the basis for axonal transport. Consistent with this idea, it was found that when tau is transfected into culture cells microtubules become more stable and the degree of assembly increases [68,69]. This effect can be observed even more dramatically in insect Sf0 ceils transfected with tau using the baculovirus system: these normally round cells develop 'neurite-like' processes. In this system, tau also protects microtubules against drug- induced disassembly [70",71"]. The simple picture does not always hold, however, as transfection oftau into Chi- nese hamster ovary cells did not result in an increase in microtubule stability or mass [72"]. Even more surprising

is the fact that tau-null transgenic mice are viable; in this case, the stabilization ofmicrotubules may be achieved by other MAPs [73"].

Functions of tau

Tau has several distinct effects on microtubules in vitro: it binds to them, promotes their nucleation and elongation, protects against disassembly, and induces them to form parallel arrays (bundles). These functions can be traced back to different domains of tau. For example, efficient assembly of microtubules requires not only the internal repeat domain, but also regions on either side, acting as 'jaws'; the flanking regions in turn tend to promote the formation of bundles [74-76]. Bundles of microtubules may be important not only in neurite outgrowth, but also in other functions such as in the erythrocyte marginal band, where tau is one of the MAPs responsible for cohesion [77]. Although the con- tribution of the repeat domain to microtubule binding is generally weak, there are certain 'hotspots' (e.g. the region between repeats 1 and 2) which could regulate the affinity, especially during the developmentally regulated alternative splicing of tau [78].

Distribution Another characteristic feature of tau is its heterogeneous distribution, both within cells and between different cell populations. Whereas six tau isoforms are found in central nervous tissue (in humans), 'big' tau (distinguished mainly by the presence of an additional exon, 4a) occurs in peripheral nerves. Within neurons, tau is mainly axonal (in contrast to the somatodendritic MAP2), but there is also a nuclear variant of tau. This intracellular sorting may take place at the level of the transcripts, which come in sizes between 2.5 and 8 kilobases and may contain targeting information. Microtubules may play a role in distributing tau, because mRNA is transported along them [79-82].

AIzheimer's disease Apart from its cell biological role, tau has generated considerable interest because of its relationship with Alzheimer's disease. Tau is the main component of the paired helical filaments that make up the neurofibril- lary tangles in the brains of patients with Alzheimer's disease. In these tangles, tau is not only abnormally aggregated, but also modified in several other ways fie- viewed in [83]). The most conspicuous modification is excess phosphorylation, frequently at Set-Pro or Thr- Pro motifs (indicating the influence of proline-directed kinases). There are other sites (such as Ser262) that strongly influence the affinity of tau for microtubules [84"] and are phosphorylated in the tau protein found in Alzheimer's disease [85]; phosphorylation might explain why this tau does not bind to microtubules and ceases to stabilize them [86]. This effect contrasts with that re- sulting from 'fetal' phosphorylation of tau, which shares some of the phosphorylation sites of Alzheimer's disease,

76 Cytoskeleton

but which still binds well to microtubules [87]. Several phosphorylation sites in fetal and adult rat tau have now been determined directly by mass spectrometry and se- quencing [88]. From a practical point of view, most studies of phosphorylation rely on phosphorylation-dependent antibodies, which can also be used with cell models (epitopes are summarized in [83]). All of the sites found so far can be dephosphorylated by phosphatases PP-2A and PP-2B (calcineurin), both of which are abundant in brain [89]. Other modifications include ubiquitination at several lysines [90] and glycation [91].

Structure How the normal and pathological properties of tau are related to its structure remains enigmatic. In solution, tau lacks secondary structure or even a defined shape and is best described as a random coil; the lack of secondary structure is observed even in paired helical filaments, suggesting that these may not assemble by way of interacting ~-strands, in contrast to other fibers found in pathological states [92]. Nevertheless, tau tends to associate into antiparallel dimers, and in this form it is prone to assembly into paired helical filaments [93].

MAP2, MAPla, MAPlb and MAP4 MAP2 MAP2 is a relative o f tau in that it has a homologous repeat domain near the carboxyl terminus but has a much larger extension towards the amino terminus. The mid- dle part can be spliced out to give MAP2c, a 'juvenile' isoform [94]. Whereas tau can have either three or four repeats, MAP2 used to be found with three repeats only; however, this gap has now been closed by the discovery of a four-repeat MAP2c. Unlike the neuronal MAP2, this isoform occurs in gila cells [94]. The question of how MAPs bind to microtubules, and how they interfere with motor proteins, continues to-be a matter of study. Wallis et al. [95] showed that MAP2 binds in a posi- tively cooperative manner, leading to the clustering of MAP2 on the microtubule surface (which makes it distinct from tau [76]). Echoing an earlier debate, Hagiwara et al. [96] reported that MAP2 and motor proteins com- pete for the same binding sites (in the carboxy-terminal domain of tubulin), whereas Marya et al. [97] found distinct sites and no competition. A compromise solution was offered by Lopez and Sheetz [98], who found distinct sites but steric hindrance, as might be expected with molecules as large as MAP2. Whatever the explanation of the results, in practice there seems to be no doubt that motors can work their way along microtubules quite ef- ficiently, even when MAPs are present.

MAPla and M A P l b

MAPla and MAPlb are two other high molecular weight neuronal MAPs; they do not belong to the same class as MAP2 and tau because they have different microtubule interaction motifs. MAPla and MAPlb share partial sequence homology, and both are syn-

thesized as a polyprotein coupled to a light chain (LC; MAPlb-LC1, 34 kDa, and MAPla-LC2, 28 kDa, respectively [99,100]). As MAPlb binds to microtubules through a series of basic repeats of the type KKE, the same mode of binding had also been assumed for MAPla. This is not the case, however, as Cravchik et al. [101] have discovered that MAPla has a novel binding motif which is unique, as it is acidic rather than basic. They proposed that MAPla has a bipolar helical structure, with positive and negative charges separated on the two sides. The third light chain, LC3 (18kDa), is encoded separately and has now been cloned [102]. It is a basic protein, abundant in neurons, that binds to both MAPla and MAPlb and could modulate their interaction with microtubules.

MAP4

In contrast to the neuronal proteins discussed above, MAP4 has a ubiquitous distribution. It resembles MAP2 and tau in having a homologous repeat domain that interacts with microtubules. Contrary to the view that MAPs are freely diffusible in cells, Chapin and Bulinski [103] have observed a surprising extent of microhetero- geneity of MAP4-microtubule interactions in cell cul- tures, suggesting the selective stabilization of subsets of microtubules within cells. A related MAP4-1ike protein has been purified from Xenopus eggs [104]. This protein becomes phosphorylated during mitosis and thus is a candidate for MAP4-dependent regulation of microtubule dynamics.

Plant MAPs Compared to the proteins mentioned so far, the study of plant microtubules has been lagging behind, partly because of the difficulty in preparing sufficient quantities of protein. Recent work shows that plant microtubules are quite comparable with those of animals; for example, plant tubulin can be assembled reversibly using taxol, hy- brid systems of maize tubulin assembled with neuronal MAP2 show tight binding (dissociation constant in the micromolar range), and some plant MAPs appear to be related to tau [105-107].

Other proteins associated with microtubules As mentioned in the Introduction, not all associated proteins can be classified as structural or motor MAPs; conversely, as microtubules are 'sticky', not all associa- tions may be meaningful. As phosphorylation regulates many intracellular interactions, it is intriguing that microtubules associate with a number ofkinases (such as the mitogen-activated protein kinase cdk5, and GSK-3; see [83]) that can cause the 'pathological' phosphorylation of tau protein. The kinases often bind to microtubules via MAPs. An interesting case is that of protein kinase A associated with MAP2: phosphorylation of the regulatory subunit of protein kinase A by the cell cycle kinase cdc2 releases the catalytic subunit and presumably allows it to phosphorylate other targets [108]. Dynamin is a GTPase involved in endocytosis; it has two microtubule-bind-


ing sites near the amino terminus and a proline-rich domain (resembling Src homology region 3 domains) near the carboxyl terminus whose affinity is regulated by phosphorylation [109].

Several proteins involved in signal transduction bind tightly to tubulin or microtubules. Recent examples include neurofibromin (a neuronal GTPase-activating pro-

tein) [110], the ~-subunits of several G proteins (which can take up GTP from tubulin) [111], and gephyrin (a protein linked to the glycine receptor) [112]. Associa- tions with microtubules are not restricted to proteins; the binding of mRNAs was mentioned above [80], and recently Surridge and Burns [113] reported that the mid- die part of MAP2 has a high affinity for phospholipids, a property not found in tau or MAP2c.

Of special interest in this context are proteins that are capable ofmicrotubule severing, which could contribute to their rapid reorganization, for example during mitosis. Examples include katanin, a heterodimeric protein (60 and 81kDa) that severs microtubules in an ATP- dependent fashion [114"], and elongation factor EF-I~, a GTP-binding protein involved in protein synthesis [115°]. On the other hand, a protein like EF-10t has also been found to be able to cause the formation of bundles of microtubules [116 °] and has been found in the pro-centrosomal protein complex (see above; [54°]). It will be interesting to see how these seemingly con- tradictory activities are related to one another. Finally, flagellar microtubules can be excised near their base, and it turns out that this is achieved by contractile fibers consisting of centrin, a relative of the calcium-binding protein calmodulin ([117"]; see also Salisbury, this issue, pp 39-45).

Microtubule dynamics

potential for bidirectional gliding during anaphase [123]. Plant ceils have the capacity for reorienting their entire microtubule network in response to extracellular factors and gravity [124]. Some of these features can be repro- duced in vitro, where microtubules can be induced to display elaborate waves of assembly and orientation in time and space [125"].

Conclusions

In contrast to the actin-based cytoskeleton, where the structures of the main players (actin, myosin and their complex) have been solved by X-ray crystallography [126,127], the microtubule system does not yet offer such insights. A direct structural analysis may be possible by electron crystallography, because zinc induces the formation of tubulin sheets that diffract to near- atomic resolution [128°]. For the time being, we have to learn by analogy. For example, the structure of the actin-binding domains of gelsolin [129] illustrates that repetitive sequences need not reflect repetitive binding; the interactions may in fact be based on the variable parts, whereas the repeats could simply provide a frame- work for protein folding (this argument might apply to models of rnicrotubule-MAP interactions). Tubulin is a GTP-binding protein that shares weak homology with signal-transducing G proteins. The structure of the ct- subunit of transducin has become available [130]; part of it folds in a manner similar to the small G protein 1Las/p21 (reviewed in [131]). Finally, if one is interested in an example of an assembly of 0r- and [~-subunits, with nucleotides bound in a non-exchangeable and exchangeable fashion, then the recent solution of the F1-ATPase structure may be worth studying [132]. Even though these proteins have different functions from tubulin, they provide intriguing food for thought.

One of the most fascinating aspects of microtubules is their rapid turnover and potential for reorganization, unexpected for a cyto-'skeletal' polymer. The half-lives of most microtubules are in the range of minutes, and the predominant mechanism of redistribution appears to be dynamic instability (reviewed by Cassimeris, [118]), perhaps in combination with the severing factors just mentioned [114",115°]. Dynamic instability can also be observed in plant cells [119]. On the other hand, many cellular microtubules are intrinsically stable [120], suggesting that the observed dynamics are tightly regulated by factors controlling catastrophes and rescues, including kinases such as the cell cycle dependent kinase cdc2. Independent of these mechanisms is a steady flux (or treadmill) of subunits, such that labeled subunits migrate slowly from the plus to the minus end of the microtubule [121]. In spite of this activity, and contrary to intuition, the assembled microtubule mass does not change nmch during mitosis, indicating that reorganization does not require net disassembly [122]. A reconstruction of an entire spindle microtubule assembly shows that the nearest-neighbor interactions are mainly between antiparallel microtubules, supporting the

Acknowledgements

We thank Alexander Marx for help in compiling the references, and Stefan Sack for the preparation of Fig. 1.

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as: * of special interest ** of outstanding interest

1. Guenette S, Solomon F: Microtubule assembly: pathways, dynamics, and regulators. Curr Opin Struct Bio1199 3, 3:198-20].

2. Mandelkow E, Mandelkow E-M: Microtubule structure. Curr Opin Struct Biol 1994, 4:171-179.

3. Joshi HC: Microtubule organizing centers and y-tubulin. Curt Opin Cell Biol 1994, 6:55-62.

4. Hirokawa N: Microtubule organization and dynamics dependent on microtubule-associated proteins. Curt Opin Cell Biol 1994, 6:74-81.

5. Schoenfeld TA, Obar RA: Diverse distribution and function of fibrous microtubule-associated proteins in the nervous system. Int Rev Cytol 1994, 151:67-137.

78 Cytoskeleton

6. Lee G: Non-motor microtubule-associated proteins. Curt Opin Cell Biol 1993, 5:88-94.

7. Goldstein LSB: With apologies to Scheherazade: tails of 1001 kinesin motors. Annu Rev Genet 1993, 27:319-351.

8. Walker R, Sbeetz MP: Cytoplasmic microtubule-associated motors. Annu Rev Biochem 1993, 62:429-451.

9. Skoufias DA, Scholey JM: Cytoplasmic microtubule-based motor proteins. Curr Opin Cell Biol 1993, 5:95-] 04.

10. Kreis 1-, Vale R (Eds): Guidebook to the cytoskeletal and motor proteins. Oxford: Oxford University Press; 1993.

11. Hyams JS, Lloyd CW (Eds): Microtubules. New York: Wiley-Liss; 1993.

12. Song Y-H, Mandelkow E: Recombinant kinesin motor domain • binds to 13-tubulin and decorates microtubules with a B-surface

lattice. Proc Nat/ Acad Sci USA 1993, 90:1671-1675. Decoration of microtubules with recombinant kinesin head domains shows that microtubules have a B-type lattice of dimers. The stoichiometry is one kinesin head per tubulin dimer, and chemical cross-linking con- nects kinesin to [[]-tubulin.

13. Harrison B, Marchese-Ragona S, Gilbert S, Cheng N, Steven • A, Johnson KA: Decoration of the microtubule surface by one

kinesin head per tubulin heterodimer. Nature 1993, 362:73-7,5. Describes microtubules decorated with kinesin head domains and im- aged by cryo-electron microscopy.

14. Song Y-H, Mandelkow E: Anatomy of flagellar microtubules: • polarity, seam, junctions, and lattice. J Cell Biol 1995,

128:81-94. Analysis of flagellar outer doublet and central pair microtubules by decoration with kinesin heads, showing the lattices (B-type in all cases) and the polarity of dimers (u.-tubulin at the plus end).

15. Mitchison T): Localization of an exchangeable GTP-binding site • at the plus end of microtubules. Science 1993, 261:1044-1047. A direct demonstration that the exchangeable GTP on 13-tubulin is near the plus end of microtubules.

16. Kuo SC, Sheetz MP: Force of single kinesin molecules meas- • ured with optical tweezers. Science 1993, 260:232-234. This series of papers [16*-20 "] provides examples of how the microme- chanics of molecular motors can be analyzed by laser tweezers, video microscopy, and related techniques. The kinesin step sizes are related to the tubulin dimer repeat, and the forces are in the picoNewton range per kinesin head.

17. Svoboda K, Schmidt CF, Schnapp BJ, Block SM: Direct obser- • vation of kinesin stepping by optical trapping interferomelry.

Nature 1993, 365:721-727. Demonstration by optical trapping that kinesin moves along microtubules with step sizes of 8 nm. See also [16"] annotation.

18. Svoboda K, Block SM: Force and velocity measured for single • kinesin molecules. Cell 1994, 77:773-784. Force-velocity relationships of kinesin coupled to silica beads and moving along microtubules argue that ATP hydrolysis and kinesin displace- ment are only loosely coupled. See also [16*] annotation.

19. Hunt AJ, Gittes F, Howard J: The force exerted by a single • kinesin molecule against a viscous load. Biophys J 1994,

67:766-781. The speed of microtubules moving on a kinesin-coated surface was measured in solvents of different viscosities. The results are compared to different 'ratchet' or 'power stroke' models. See also [16 ° ] annotation.

20. Hall K, Cole DG, Yeh Y, Scholey JM, Baskin RJ: Force- * velocity relationships in kinesin-driven motility. Nature 1993,

364:457-459. A centrifuge microscope is used to measure force-velocity relationships for kinesin, yielding force values in the range of 0.1 pN per kinesin head. See also [16 •] annotation.

21. Ray S, Meyhofer E, Milligan RA, Howard J: Kinesin follows the microtubules protofilament axis. J Cell Biol 1993, 121:1083-1093.

22. Hackney DD: Evidence for alternating head catalysis by kinesin • during microtubule-stimulated ATP hydrolysis. Proc Nat/Acad

Sci USA 1994, 91:6865~869.

A kinetic analysis of the interactions between microtubules, kinesin, and ATP hydrolysis.

23. Dye RB, Fink SP, Williams RC: Taxol-induced flexibility of microtubules and its reversal by MAP2 and tau. J Biol Chem 1993, 268:6847-6850.

24. Gittes F, Mickey B, Nettleton J, Howard J: Flexural rigidity of microtubules and actin filaments measured from thermal fluctuations in shape. J Cell Biol 1993, 120:923-934.

25. Venier P, Maggs AC, Carlier MF, Pantaloni D: Analysis of microtubule rigidity using hydrodynamic flow and thermal fluctuations. J Biol Chem 1994, 269:13353-13360.

26. Vale RD, Coppin CM, Malik F, Kull FJ, Milligan RA: Tubulin GTP hydrolysis influences the structure, mechanical properties, and kinesin driven transport of microtubules. J Biol Chem 1994, 269:23769-23775.

27. Stewart RJ, Thaler JP, Goldstein LSB: Direction of microtubule • movement is an intrinsic property of the motor domains of

kinesin heavy chain and Drosophila ncd protein. Proc Natl Acad Sci USA 1993, 90:5209-5213.

The paper shows that the directionality of the movement is encoded in the head domain of a motor, independently of its arrangement in the protein.

28. Endow S, Kang S, Satterwhite L, Rose M, Skeen V, Salmon ED: Yeast Kar3 is a minus-end microtubule motor protein that destabilizes microtubules preferentially at the minus ends. EMBO J 1994, 13:2708-2713.

29. Middleton K, Carbon J: Kar3-encoded kinesin is a minus-end- directed motor that functions with centromere-binding proteins (cbf3) on an in-vitro yeast kinetochore. Proc Natl Acad Sci USA 1994, 91:7212-7216.

30. Hunt AJ, Howard J: Kinesin swivels to permit microtubule movement in any direction. Proc Natl Acad Sci USA 1994, 90:11653-11657.

31. Bernstein M, Beech PL, Katz SG, Rosenbaum JL: A new kinesin- like protein (kip1) localized to a single microtubule of the Chlamydomonas flagellum. J Cell Biol 1994, 125:1313-1326.

32. Liao H, Li G, Yen TJ: Mitotic regulation of microtubule cross- • linking activity of CENP-E kinetochore protein. Science 1994,

265:394-398. An example of how a kinesin-like head domain can cooperate with another microtubule-binding domain on the same protein under the control of phosphorylation.

33. Sackett DL: Podophyllotoxin, steganacin and combretastatin: natural products thai bind at the colchicine site of lubulin. Pharmacol Ther 1993, 59:163-228.

34. Hamel E: Interactions of tubulin with small ligands. In Mi- crotubule proteins. Edited by Avila J. Boca Raton: CRC Press; 1990:89-191.

35. Toso RJ, Jordan MA, Farrell KW, Matsumoto B, Wilson L: Ki- netic stabilization of microtubule dynamic instability in vitro by vinblastine. Biochemistry 1993, 32:1285-1293.

36. Vandecandelaere A, Martin SR, Schilstra MJ, Bayley PM: Effects of the tubulin-colchicine complex on microtubule dynamic instability. Biochemistry 1994, 33:2792-2801.

37. Perez-Ramirez B, Shearwin KE, Timasheff SN: The colchicine- induced GTPase activity of tubulin: state of the product activa- tion by microtubule-promoting cosolvents. Biochemistry 1994, 33:6253-6261.

38. Uppuluri S, Knipling L, Sackett DL, Wolff J: Localization of the colchicine-binding site of tubulin. Proc Nail Acad Sci USA 1993, 90:11598-11602.

39. Combeau C, Commercon A, Mioskowski C, Rousseau B, Aubert F, Goeldner M: Predominant labeling of ~-tubulin over ct-tubulin from porcine brain by a photoactivatable taxoid derivative. Biochemistry 1994, 33:6676-6683.

40. Rao S, Krauss N, Heerding J, Swindell C, Ringel I, Orr G, Horwitz S: 3'-(p-azidobenzamido)taxol photolabels the N- terminal 31 amino-acids of ~-tubulin. J Biol Chem 1994, 269:3132-3134.

Microtubules and microtubule-associated proteins M a n d e l k o w and M a n d e l k o w 79

41. Diaz JF, Andreu JM: Assembly of purified GDP tubulin into microtubules induced by taxol and taxotere: reversibility, ligand stoichiometry, and competition. Biochemistry 1993, 32:2747-2755.

42. Shivanna BE), Mejillano MR, Williams TD, Himes RH: Ex- changeable GTP binding-site of [3-tubulin: identification of cysleine-12 as the major site of cross-linking by direct pholoaffinity-labeling. J Bio/ Chem 1993, 268:127-132.

43. Jayaram B, Haley BE: Identification of peptides within the base binding domains of the GTP-specific and ATP-specific binding- sites of tubulin. J Biol Chem 1994, 269:3233-3242.

44. Davis A, Sage CR, Dougherty CA, Farrell KW: Microtubule • dynamics modulated by guanosine triphosphate hydrolysis ac-

tivity of ~-tubulin. Science 1994, 264:839-842. This paper and [45"] provide examples of how yeast genetics can be used to answer questions about tubulin structure, functional domains, and GTPase activity.

45. Reijo RA, Cooper EM, Beagle G J, Huffaker TC: Systematic mu- • tational analysis of the yeast [3-tubulin gene. Mol Biol Cell

1994, 5:2943. A systematic mutation analysis of [~-tubulin (converting charged residues to alanine) to reveal regions important for the functions of microtubules in yeast.

46. Rudiger M, Weber K: Characterization of the post-translational modifications in tubulin from the marginal band of avian erythrocytes. Eur J Biochem 1993, 218:107-116.

47. Lu Q, Luduena RF: In vitro analysis of microtubule assembly of isotypically pure tubulin dimers: intrinsic differences in the assembly properties of c~-[~-ii, c~-13-iii, and (~-[3-iv tubulin dimers in the absence of microtubule-associated protein. J Biol Chem 1994, 269:2041-2047.

48. Ersfeld K, WehJand J, Plessmann U, Dodemont H, Gerke V, Weber K: Characterization of the tubulin tyrosine ligase. J Cell Biol 1993, 120:725-732.

49. Paturle-Lafanechere L, Manier M, Trigault N, Pirollet F, Mazar- gull H, Job D: Accumulation of delta-2-tubulin, a major tubulin variant that cannot be tyrosinated in neuronal tissues and in stable microtubule assemblies. J Cell 5ci 1994, 107:1529-1543.

50. Audebert S, Koulakoff A, Berwald-Netter Y, Gros F, Denoulet P, Edde B: Developmental regulation of polyglutamylated c~- tubulin and ~-tubulin in mouse-brain neurons. J Cell 5ci 1994, 107:2313-2322.

51. Gurland G, Gundersen GG: Protein phosphatase inhibitors induce the selective breakdown of stable microtubules in fibroblasts and epithelial-cells. Proc Natl Acad 5ci USA 1993, 90:8827-8831

52. Bachurski C, Theodorakis N, Coulson RM, Cleveland DW: An amino-terminal tetrapeptide specifies cotranslational degrada- tion of [3-tubulin but not ct-tubulin messenger RNAs. Mol Cell Biol 1994, 14:4076~.086.

53. Bramhill D, Thompson CM: GTP-dependent polymerization of Escherichia coil FtsZ protein to form tubules. Proc Natl Acad Sci USA ]994, 91:5813-5817.

54. Marchesi VT, Ngo N: In vitro assembly of multiprotein com- . plexes containing cc-tubulin, [3-tubulin, and y-tubulin, heat-

shock protein hsp70, and elongation factor-lct. Proc Nat/Acad Sci USA 1993, 90:3028-3032.

Protein compexes that represent centrosome precursors are isolated from Chinese hamster ovary cells. They contain cytoskeletal proteins (a.-, [3-, y-tubulin and actin) as well as elongation factors and chaperones.

55. Stearns T, Kirschner M: In vitro reconstitution of centrosome • assembly and function: the central role of y-tubulin. Cell 1994,

76:623-637. A 25 S complex containing y-tubulin in Xenopus eggs can combine with the sperm centriole to form the centrosome; this assembly requires ATP but not microtubules. The complex confers the microtubule-nucleating activity upon the centrosome.

56. Felix MA, Antony C, Wright M, Maro B: Centrosome assembly • in vitro: role of y-tubulin recruitment in Xenopus sperm aster

formation. J Biol Chem 1994, 124:19-31

This study distinguishes between the nucleation of microtubules on other microtubules (template elongation) and on the pericentriolar material (as- tral nucleation) in Xenopus eggs, and describes the role of y-tubulin in the nucleation activity.

57. Ahmad FJ, Joshi HC, Centonze VE, Baas PW: Inhibition of mi- • crotubule nucleation at the neuronal centrosome compromises

axon growth. Neuron 1994, 12:271-280. Growth of axons depends on a supply of microtubules which are nucleated on the neuronal centrosomes and require the presence of ",f-tubulin.

58. Doxsey SJ, Stein P, Evans L, Calarco PD, Kirschner M: Peri- • centrin, a highly conserved centrosome protein involved in

microtubule organization. Cell 1994, 76:639-650. Pericentrin is a newly discovered 220 kDa protein involved in organizing the pericentriolar material and other structures from which microtubules can be nucleated.

59. Maekawa T, Kuriyama R: Primary structure and microtubule- interacting domain of the SP-H antigen: a mitotic MAP located at the spindle pole and characterized as a homologous protein to NuMa. J Cell Sci 1993, 105:589-600.

60. Steffen W, Fajer EA, Linck RW: Centrosomal components immunologically related to tektins from ciliary and flagellar microtubules. J Cell 5ci 1994, 107:2095-2105.

61. Horio T, Oakley BR: Human y-tubulin functions in fission yeast. J Biol Chem 1994, 126:1465-1473.

62. Melki R, Vainberg IE, Chow RL, Cowan NJ: Chaperonin-medi- • ated folding of vertebrate actin-related protein and y-tubulin.

J Cell Biol 1993, 122:1301-1310. The actin homologue centractin and ~-tubulin, two centrosomal proteins, are folded through the cytoplasmic chaperof~in complex TCP-1 ; the folding reaction requires ATP.

63. Sternlicht H, Farr GW, Sternlicht ML, Driscoll JK, Willison K, • Yaffe MB: The t-complex polypeptide-1 complex is a chaper-

onin for tubulin and actin in vivo. Proc Nat/ Acad Sci USA 1993, 90:9422-9426.

A pulse-labelling analysis of actin and tubulin biogenesis in Chinese hamster ovary cells showing that thay must associate with the 900 kDa TCP-1 complex in order to fold properly and associate into functional tubulin heterodimers.

64. Gao YI, Vainberg IE, Chow RL, Cowan NJ: Two cofactors and • cytoplasmic chaperonin are required for the folding of cc-tubu-

lin and [[3-tubulin. Mol Cell Biol 1993, 13:2478-2485. Analysis of the interactions between the TCP-1 complex, actin, tubulin, and other cofactors involved in the folding reaction.

65. Melki R, Cowan NJ: Facilitated folding of actins and tubulins • occurs via a nucleotide-dependent interaction between cyto-

plasmic chaperonin and distinctive folding intermediates. Mol Cell Biol 1994, 14:2895-2904.

This paper describes the sequence of steps in the folding of actin and tubulin by the cytoplasmic chaperonin complex. The initial interaction requires the chaperonin to be in the ADP-bound state, whereas release requires ATP, which thus acts as a switching device.

66. Chen XY, Sullivan DS, Huffaker TC: Two yeast genes with simi- • larlty to TCP-1 are required for microtubule and actin function

in vivo, Proc Nat/ Acad 5ci USA 1994, 91:9111-9115. The folding of actin and tubulin in yeast requires chaperonin complexes similar to the TCP-1 complex.

67. Yu WQ, Baas PW: Changes in microtubule number and length during axon differentiation. J Neurosci 1994, 14:2818-2829.

68. Lo MMS, Fieles AW, Norris TE, Dargis PG, Caputo CB, Scott CW, Lee VMY, Goedert M: Human tau isoforms confer distinct morphological and functional properties to stably transfected fibroblasts. Mol Brain Res 1993, 20:209-220.

69. Degarcini EM, Delaluna S, Dominguez JE, Avila J: Overexpres- sion of tau protein in COS-1 cells results in the stabilization of centrosome-independent microtubules and extension of cytoplasmic processes. Mol Cell Biochem 1994, 130:187-196.

70. Kosik KS, McConlogue L: Microtubule-associated protein func- • tion: lessons from expression in Spodoptera frugiperda cells.

Cell Motil Cytoskeleton 1994, 28:195-198. A review of how insect Sf9 cells can be used as a model system to generate 'neurite-like' cell processes by transfecting them with microtubule- associated proteins.

80 Cy toske le ton

71. Baas PW, Pienkowski TP, Cimbalnik KA, Toyama K, Bakalis • S, Ahmad FJ, Kosik KS: Tau confers drug stability but not

cold stability to microtubules in living cells. J Cell Sci 1994, 107:135-143.

Expression of tau in insect Sf9 cells leads to the formation of 'neurite-like' cell processes based on assembled microtubules. Probing their resistance to depolymerizing drugs or low temperature shows that tau makes these microtubules stable against drugs but not against cold.

72. Barlow S, Gonzalez-Garay ML, West R, Olmsted JB, Cabral F: • Stable expression of heterologous microtubule-associated pro-

teins (MAPs) in Chinese hamster ovary cells: evidence for dif- fering roles of MAPs in microtubule organization. J Ce// Bio/ 1994, 126:1017-1029.

COS cells were transfected with Drosophila 250 kDa MAP, human MAP4 and human tau. The exogenous MAPs had no major effect on the levels of assembled microtubules or their drug sensitivity, suggesting that MAPs may have roles distinct from those in microtubule stabilization.

73. Harada A, Oguchi K, Okabe S, Kuno J, Terada S, Ohshima T, • Sato-Yoshitake R, Takei Y, Noda T, Hirokawa N: Altered mi-

crotubule organization in smaU-caliber axons of mice lacking tau protein. Nature 1994, 369:488-49].

The first tau-null mouse appears to survive reasonably well, suggesting that neurons have a back-up system for tau.

74. Brandt R, Lee G: Functional organization of microtubule-associated protein tau: identification of regions which affect microtubule growth, nucleation, and bundle formation in vitro. J Biol Chem 1993, 268:3414-3419.

75. Brandt R, Lee G: Orientation, assembly, and stability of microtubule bundles induced by a fragment of tau protein. Cell Motil Cytoskeleton 1994, 28:143-154.

76. Gustke N, Trinczek B, Biemat J, Mandelkow E-M, Mandelkow E: Domains of tau protein and interactions with microtubules. Biochemistry ] 994, 33:95 ] 1-9522.

77. Sanchez I, Cohen WD: Assembly and bundling of marginal band microtubule protein: role of tau. Cell Motil Cytoskeleton 1994, 29:57-71.

78. Goode BL, Feinstein SC: Identification of a novel microtubule binding and assembly domain in the developmentally regulated inter-repeat region of tau. J Cell Biol 1994, 124:769-782.

79. Litman P, Barg J, Rindzoonski L, Ginzburg I: Subcellular localization of tau messenger RNA in differentiating neuronal cell culture: implications for neuronal polarity. Neuron 1993, 10:627~38.

80. Bassell GJ, Singer RH, Kosik KS: Association of poly(A) messenger-RNA with microtubules in cultured neurons. Neu- ron 1994, 12:571-582.

81. Georgieff I, Liem R, Couchie D, Mavilia C, Nunez J, Shelanski M: Expression of high molecular weight tau in the central and peripheral nervous systems. J Cell Sci 1993, 105:729-737.

82. Wang Y, Loomis PA, Zinkowski RP, Binder LI: A novel tau transcript in cultured human neuroblastoma cells expressing nuclear tau. J Cell Biol 1993, 121:257-267.

83. Mandelkow E-M, Mandetkow E: Tau as a marker for Alzheimer's disease. Trends Biochem Sci 1993, 18:480-483.

84, Biernat J, Gustke N, Drewes G, Mandelkow E-M, Mandelkow E: • Phosphorylation of serine 262 strongly reduces the binding

of tau protein to microtubules: distinction between PHF- like immunoreactivity and microtubule binding. Neuron 1993, 11:153-]63.

This paper describes a critical residue in the repeat region of tau protein whose phosphorylation detaches tau from microtubules and destabilizes them.

85. Hasegawa M, Morishima-Kawashima M, Takio K, Suzuki M, Titani K, Ihara Y: Protein sequence and mass spectrometric analyses of tau in the Alzheimer's disease brain. J Biol Chem 1992, 26:1 7047-1 7054.

86. Yoshida H, Ihara Y: Tau in paired helical filaments is function- ally distinct from fetal tau: assembly incompetence of paired helical filament tau. J Neurochem 1993, 61:1183-1186.

87. Bramblett GT, Goedert M, Jakes R, Merrick SE, Trojanowski JQ, Lee VMY: Abnormal tau phosphorylation at Ser(396) in Alzheimer's disease recapitulates development and contributes to reduced microtubule binding. Neuron 1993, 10:1089-1099.

88. Morishima-Kawashima M, Hasegawa M, Takio K, Suzuki M, Titani K, Ihara Y: Ubiquitin is conjugated with amino-termi- nally processed tau in paired helical filaments. Neuron 1993, 10:1151-1160.

89. Yan SD, Chen X, Schmidt A, Brett J, Godman G, Zou Y, Scott CW, Caputo C, Frappier T, Smith MA et al.: Glycated tau protein in Alzheimer-disease: a mechanism for induction of oxidant stress. Proc Natl Acad Sci USA 1994, 91:7787-7791.

90. Watanabe A, Hasegawa M, Suzuki M, Takio K, Morishima- Kawashima M, Titani K, Arai T, Kosik KS, Ihara Y: In-vivo phosphorylation sites in fetal and adult rat tau. J Biol Chem 1993, 268:257] 2-25717.

91. Drewes G, Mandelkow E-M, Baumann K, Goris J, Mer- levede W, Mandelkow E: Dephosphorylation of tau protein and Alzheimer paired helical filaments by calcineurin and phosphatase-2A. FEBS Lett 1993, 336:425-432.

92. Schweers O, Sch(3nbrunn-Hanebeck E, Marx A, Mandelkow E: Structural studies of tau protein and Alzheimer paired helical filaments show no evidence for [~ structure. ] Biol Chem 1994, 269:24290-24297.

93. wi l le H, Drewes G, Biernat J, Mandelkow E-M, Mandelkow E: Alzheimer-like paired helical filaments and antiparallel dimers formed from microtubule-associated protein tau in vitro. J Cell Biol 1992, 118:573-584.

94. Doll T, Meichsner M, Riederer BM, Honegger P, Matus A: An isoform of mlcrotubule-associated proteln-2 (MAP2) containing 4 repeats of the tubulin-binding motif. J Cell Sci 1993, 106:633-640.

95. Wallis K, Azhar S, Rho M, Lewis S, Cowan N, Murphy DB: The mechanism of equilibrium binding of microtubule-associated protein-2 to microtubules: binding is a multi-phasic process and exhibits positive cooperativity. J Biol Chem 1993, 268:15158-15167.

96. Hagiwara H, Yorifuji H, Sato-Yoshitake R, Hirokawa N: Com- petition between motor molecules (kinesin and cytoplasmic dynein) and fibrous microtubule-associated proteins in binding to mlcrotubules. J Biol Chem 1994, 269:3581-3589.

97. Marya PK, Syed Z, Fraylich PE, Eagles PAM: Kinesin and tau bind to distinct sties on microlubules. J Cell Sci 1994, 107:339-344.

98. Lopez LA, Sheetz MP: Steric inhibition of cytoplasmic dynein and kinesin motility by MAP2. Cell Motil Cytoskeleton 1993, 24:1-16.

99. Hammarback JA, Obar RA, Hughes SM, Vallee RB: MAPlb is encoded as a polyprotein that is processed to form a complex N-terminal microtubule-binding domain. Neuron 1991, 7:129-139.

100. Langkopf A, Hammarback J, Muller R, Vallee R, Garner C: Microtubule-associated proteins 1A and LC2: two proteins encoded in one messenger RNA. J Biol Chem 1992, 267:16561 16566.

101. Cravchik A, Reddy D, Matus A: Identification of a novel microtubule-binding domain in microtubule-associated protein-la (MAPla). J Cell Sci 1994, 107:661-672.

102. Mann SS, Hammarback JA: Molecular characterization of light chain-3, a microtubule-binding subunit of MAPla and MAPlb. J Biol Chem 1994, 269:11492-11497.

103. Chapin SJ, Bulinski JC: Cellular microtubules heterogeneous in their content of microtubule-associated protein-4 (MAP4). Cell Motil Cytoskeleton 1994, 27:133-149.

104. Faruki S, Karsenti E: Purification of microtubule proteins from Xenopus egg extracts: identification of a 230k MAP4-1ike protein. Cell Motil Cytoskeleton 1994, 28:108-118.

105. Bokros C, Hugdahl J, Hanesworth V, Murthy J, Morejohn L: Characterization of the reversible taxol-induced polymeriza-

Mic ro tubu les and mic ro tubu le -assoc ia ted pro te ins M a n d e l k o w and M a n d e l k o w 81

tion of plant tubulin into microtubules. Biochemistry 1993, 32:3437-3447.

106. Hugdahl J, Bokrok C, Hanesworth V, Aalund G, Morejohn LC: Unique functional characteristics of the polymerization and MAP binding regulatory domains of plant tubulin. Plant Cell 1993, 5:1063-1080.

107. Vantard M, Peter C, Fellous A, Schellenbaum P, Lambert A: Characterization of a 100 kDal heat-stable microtubule- associated protein from higher plants. Eur J Biochem 1994, 220:847-853.

108. Keryer G, Luo ZJ, Cavadore JC, Erlichman J, Bornens M: Phos- phorylation of the regulatory subunit of type-ii-~ cAMP-dependent protein kinase by cyclin B/p34(cdc2) kinase impairs its binding to microtubule-associated protein-2. Proc Nat/ Acad Sci USA 1993, 90:5418-5422.

109. Herskovits JS, Shpetner HS, Burgess CC, Vallee RB: Micro- tubules and src homology-3 domains stimulate the dynamin GTPase via its C-terminal domain. Proc Natl Acad Sci USA 1993, 90:11468-11472.

110. Bollag G, McCormick F, Clark R: Characterization of full-length neurofibromin: tubulin inhibits ras GAP activity. EMBO J 1993, 12:1923-1927.

111. Roychowdhury S, Rasenick MM: Tubulin45 protein association stabilizes GTP-binding and activates GTPase: cytoskeletal par- ticipation in neuronal signal transduction. Biochemistry 1994, 33:9800-9805.

112. Kirsch J, Betz H: Widespread expression of gephyrin, a putative glycine receptor tubulin linker protein, in rat brain. Brain Res 1993, 621:301-310.

113. Surridge CD, Burns RG: The difference in the binding of phosphatidylino~itol distinguishes MAP2 from MAP2c and tau. Biochemistry 1994, 33:8051-8057.

114. McNally FJ, Vale RD: Identification of katanin, an ATPase * thai severs and disassembles stable microtubules. Cell 1993,

75:419-429. Katanin is a heterodimeric protein complex that is capable of cutting microtubules in the presence of ATP.

115. Shiina N, Gotoh Y, Kubomura N, Iwamatsu A, Nishida E: • Microlubule severing by elongation factor 1-co. Science 1994,

266:282-285. A new role - - severing of microtubules - - is discovered for the elongation factor-lu.. See also [116"].

116. Durso NA, Cyr RJ: A calmodulin-sensitive interaction between * microtubules and a higher plant homolog of elongation factor

let. Plant Cell 1994, 6:893-905. In this paper, elongation factor-l~ is shown to stabilize and bundle microtubules in a calmodulin-dependent fashion. See also [11S•].

117. Sanders MA, Salisbury JL: Centrin plays an essential role in mi- • crotubule severing during flagellar excision in Chlamydomonas

reinhardtii. J Biol Chem 1994, 124:795-805. Severing of flagella in Chlamydomonas is achieved by cutting microtubules using contractile fibers that contain centrin, a calmodulin homo- Iogue.

118. Cassimeris L: Regulation of microtubule dynamic instability. Cell Motil Cytoskeleton 1993, 26:275-281.

119. Hush JM, Wadsworth P, Callaham DA, Hepler PK: Quantifica- tion of microtubule dynamics in living plant cells using fluo- rescence redistribution after photobleaching. J Cell Sci 1994, 107:775-784.

120. Lieuvin A, Labbe J-C, Dome M, Job D: Intrinsic microtubule stability in interphase cells. J Cell Biol 1994, 124:985-996.

121. Sawin KE, Mitchison TJ: Microtubule flux in mitosis is independent of chromosomes, centrosomes, and antiparallel microtubules. Mol Biol Cell 1994, 5:217-226.

122. Zhai Y, Borisy GG: Quantitative determination of the proportion of microtubule polymer present during the mitosis-interphase transition. J Cell Sci 1994, 107:881-890.

123. Mastronarde DN, McDonald KL, Ding R, Mclntosh JR: In- terpolar spindle microtubules in PtK cells. J Cell Biol 1993, 123:1475-1489.

124. Yuan M, Shaw PI, Warn RM, Lloyd CW: Dynamic reorientation of cortical microtubules, from transverse to longitudinal, in living plant-cells. Proc Natl Acad Sci USA 1994, 91:6050-6053.

125. Tabony J: Morphological bifurcations involving reaction-diffu- * sion processes during microtubule formation. Science 1994,

264:245-248. An example of how tubulin assembly alone can lead to spatial and tem- poral order, independently of other cellular factors.

126. Rayment I, Holden HM, Whittaker M, Yohn CB, Lorenz M, Holmes KC, Milligan RA: Structure of the actin-myosin complex and its implications for muscle contraction. Science 1993, 261:58-65.

127. Schroder RR, Manstein DJ, Jahn W, Holden H, Rayment I, Holmes KC, Spudich IA: Three-dimensional atomic model of F-actin decorated with Dictyostelfum myosin-S1. Nature 1993, 364:171-174.

128. Wolf SG, Mosser G, Downing KH: Tubulin conformation • in zinc-induced sheets and macrotubes. J Struct Biol 1993,

111:190-199. A high-resolution (4,~) electron-diffraction analysis of tubulin sheets, re- vealing the presence of secondary structure elements such as [8-sheets.

129. McLaughlin PJ, Gooch JT, Mannherz HG, Weeds AG: Structure of gelsolin segment-l-actin complex and the mechanism of filament severing. Nature 1993, 364:685-692.

130. Noel J, Hamm H, Sigler P: The 2.2A crystal structure of transducin-ct complexed with GTP-y-S. Nature 1993, 366:654-663.

131. Wittinghofer A: The structure of transducin Gct: more to view than just ras. Cell 1994, 76:201-204.

132. Abrahams JP, Leslie A, Lutter R, Walker J: Structure at 2.8,~ resolution of F1-ATPase from bovine heart mitochondria. Nature 1994, 370:621-628.

E Mandelkow and E-M Mandelkow, Max Planck Unit for Struc- tural Molecular Biology, Notkestrasse 85, D-22603, Hamburg 52, Germany.

Documents

Mandelkow CO CellBiol95