Tubulin-nucleotide interactions. Effects of removal of exchangeable guanine nucleotide on protein conformation and microtubule assembly.

The removal of tightly bound GDP from the exchangeable nucleotide-binding site of tubulin has been performed with alkaline phosphatase under conditions which essentially retain the assembly properties of the protein. When microtubule protein is treated with alkaline phosphatase, nucleotide is selectively removed from tubulin dimer rather than from MAP (microtubule-associated protein)-containing oligomeric species. Tubulin devoid of E-site (the exchangeable nucleotide-binding site of the tubulin dimer) nucleotide shows enhanced proteolytic susceptibility of the beta-subunit to thermolysin and decreased protein stability, consistent with nucleotide removal causing changes in protein tertiary structure. Pyrophosphate ion (3 mM) is able to promote formation of normal microtubules in the complete absence of GTP by incubation at 37 degrees C either with nucleotide-depleted microtubule protein or with nucleotide-depleted tubulin dimer to which MAPs have been added. The resulting microtubules contain up to 80% of tubulin lacking E-site nucleotide. In addition to its effects on nucleation, pyrophosphate competes weakly with GDP bound at the E-site. It is deduced that binding of pyrophosphate at a vacant E-site can promote microtubule assembly. The minimum structural requirement for ligands to induce tubulin assembly apparently involves charge neutralization at the E-site by bidentate ligation, which stabilizes protein domains in a favourable orientation for promoting the supramolecular protein-protein interactions involved in microtubule formation.     

Microtubule-associated proteins from Antarctic fishes

Microtubules and presumptive microtubule-associated proteins (MAPs) were isolated from the brain tissues of four Antarctic fishes (Notothenia gibberifrons, N. coriiceps neglecta, Chaenocephalus aceratus, and a Chionodraco sp.) by means of a taxol-dependent, microtubule-affinity procedure (cf. Vallee: Journal of Cell Biology 92:435-442, 1982). MAPs from these fishes were similar to each other in electrophoretic pattern. Prominent in each preparation were proteins in the molecular weight ranges 410,000-430,000, 220,000-280,000, 140,000-155,000, 85,000-95,000, 40,000-45,000, and 32,000-34,000. The surfaces of MAP-rich microtubules were decorated by numerous filamentous projections. Exposure to elevated ionic strength released the MAPs from the microtubules and also removed the filamentous projections. Addition of fish MAPs to subcritical concentrations of fish tubulins at 0-5 degrees C induced the assembly of microtubules. Both the rate and the extent of this assembly increased with increasing concentrations of the MAPs. Sedimentation revealed that approximately six proteins, with apparent molecular weights between 60,000 and 300,000, became incorporated into the microtubule polymer. Bovine MAPs promoted microtubule formation by fish tubulin at 2-5 degrees C, and proteins corresponding to MAPs 1 and 2 co-sedimented with the polymer. MAPs from C. aceratus also enhanced the polymerization of bovine tubulin at 33 degrees C, but the microtubules depolymerized at 0 degrees C. We conclude that MAPs are part of the microtubules of Antarctic fishes, that these proteins promote microtubule assembly in much the same way as mammalian MAPs, and that they do not possess special capacities to promote microtubule assembly at low temperatures or to prevent cold-induced microtubule depolymerization.     

Nucleotide binding and phosphorylation in microtubule assembly in vitro.

Two non-hydrolyzable analogs of GTP, guanylyl-β,γ-methylene diphosphonate and guanylyl imidodiphosphate, have been found to induce rapid and efficient microtubule assembly in vitro by binding at the exchangeable site (E-site) on tubulin. Characterization of microtubule polymerization by several criteria, including polymerization kinetics, nucleotide binding to depolymerized and polymerized microtubules, and microtubule stability, reveals strong similarities between microtubule assembly induced by GTP and non-hydrolyzable GTP analogs. Nucleoside triphosphates which bind weakly or not at all to tubulin, such as ATP, UTP and CTP, are shown to induce microtubule assembly by means of a nucleoside diphosphate kinase (NDP-kinase, EC 2.7.4.6.) activity which is not intrinsic to tubulin. The NDP-kinase mediates microtubule polymerization by phosphorylating tubulin-bound GDP in situ at the E-site. Although hydrolysis of exchangeably bound GTP occurs, it is found to be uncoupled from the polymerization reaction. The non-exchangeable nucleotide binding site on tubulin (N-site) is not directly involved in microtubule assembly in vitro. The N-site is shown to contain almost exclusively GTP which is not hydrolyzed during microtubule assembly. A scheme is presented in which GTP acts as an allosteric effector at the E-site during microtubule assembly in vitro.     

Structural and functional domains of tubulin

The molecular aspects of the microtubule system is a research area that has developed very rapidly during the past decade. Research on the assembly mechanisms and chemistry of tubulin and the molecular biology of microtubules have advanced our understanding of microtubule formation and its regulation. The emerging view of tubulin is of a macromolecule containing spatially discrete sequences that constitute functionally different domains with respect to self-association, interactions with microtubule associated proteins (MAPs) and specific ligands. Recent studies point to the role of the carboxyl-terminal moiety of tubulin subunits in regulating its assembly into microtubules. These investigations combined with further studies on the spatial relationships between tubulin domains should provide new insights into the detailed structural basis of microtubule assembly.     

Assembly of pure tubulin in the absence of free GTP: effect of magnesium, glycerol, ATP, and the nonhydrolyzable GTP analogues

We describe in vitro microtubule assembly that exhibits, in bulk solution, behavior consistent with the GTP cap model of dynamic instability. Microtubules assembled from pure tubulin in the absence of free nucleotides could undergo one cycle of assembly, but could not sustain an assembly plateau. After the initial peak of assembly was reached and bound E-site GTP hydrolyzed to GDP, the microtubules gradually disassembled. We studied buffer conditions that maximized this disassembly while still allowing robust assembly to take place. While both glycerol and glutamate increased the rate of initial assembly and then slowed disassembly, magnesium promoted initial assembly and, surprisingly, enhanced disassembly. After cooling, a second cycle of assembly was unsuccessful unless GTP or the hydrolyzable GTP analogue GMPCPOP was readded. The nonhydrolyzable GTP analogues GMPPNP and GMPPCP could not support the second assembly cycle in the absence of E-site GTP. Analysis using HPLC found no evidence that GMPPNP, GMPPCP, or ATP could bind to free tubulin, and these nucleotides did not compete with GTP for the E-site. We have, however, demonstrated that the nonhydrolyzable GTP analogues and ATP do have an important effect on microtubule assembly. GMPPNP, GMPPCP, and ATP could each enhance the rate of assembly and stabilize the plateau of assembled microtubules against disassembly, while not binding appreciably to free tubulin. We conclude that these nucleotides, as well as GTP itself, enhance assembly by binding to a site on microtubules that is not present on free, unpolymerized tubulin. We estimate the affinity (KD) of the polymeric site for nucleotide triphosphates to be approximately 10(-4)M.     

Towards an understanding of microtubule function and cell organization: an overview.

Microtubules exhibit dynamic instability, converting abruptly between assembly and disassembly with continued growth dependent on the presence of a tubulin-GTP cap at the plus end of the organelle. Tubulin, the main structural protein of microtubules, is a heterodimer composed of related polypeptides termed alpha-tubulin and beta-tubulin. Most eukaryotic cells possess several isoforms of the alpha- and beta-tubulins, as well as gamma-tubulin, an isoform restricted to the centrosome. The isoforms of tubulin arise either as the products of different genes or by posttranslational processes and their synthesis is subject to regulation. Tubulin isoforms coassemble with one another and isoform composition does not appear to determine whether a microtubule is able to carry out one particular activity or another. However, the posttranslational modification of polymerized tubulin may provide chemical signals which designate microtubules for a certain function. Microtubules interact with proteins called microtubule-associated proteins (MAPs) and they can be divided into two groups. The structural MAPs stimulate tubulin assembly, enhance microtubule stability, and influence the spatial distribution of microtubules within cells. The dynamic MAPs take advantage of microtubule polarity and organization to vectorially translocate cellular components. The interactions between microtubules and MAPs contribute to the structural-functional integration that characterizes eukaryotic cells.     

Influence of Monoclonal Antibodies on Microtubule Assembly

The influence on microtubule assembly in vitro of monoclonal antibodies against microtubule-associated proteins (MAPs) was studied. Light scattering was used for measuring net polymer formation and electron microscopy for determining the influence of antibodies on microtubule morphology. Control experiments showed that nonimmune mouse IgG had no effect on either the assembly or appearance of microtubules. The same was true for monoclonal antibodies against MAP1. At low levels, antibodies against MAP2 caused the aggregation of microtubules into bundles, an effect that did not occur with antibodies against any other MAP type studied. At increasing concentrations, anti-MAP2 progressively inhibited tubulin polymerization, producing irregular, shortened filaments. Anti-MAP5 produced a striking fragmentation of microtubules into very short pieces that were otherwise morphologically identical to control microtubules. The different effects of these antibodies show the potential of monoclonal antibodies for investigating MAP function and form an important adjunct to cellular microinjection experiments.     

Separation of active tubulin and microtubule-associated proteins by ultracentrifugation and isolation of a component causing the formation of microtubule bundles

A new method for separating microtubule-associated proteins (MAPs) and tubulin, appropriate for relatively large-scale preparations, was developed. Most of the active tubulin was separated from the MAPs by centrifugation after selective polymerization of the tubulin was induced with 1.6 M 2-(N-morpholino)ethanesulfonate (Mes) and GTP. The MAPs-enriched supernatant was concentrated and subsequently clarified by prolonged centrifugation. The supernatant (total soluble MAPs) contained almost no tubulin, most of the nucleosidediphosphate kinase activity of the microtubule protein, good activity in promoting microtubule assembly in 0.1 M Mes, and proteins with the electrophoretic mobility of MAP-1, MAP-2, and tau factor. The pellet, inactive in supporting microtubule assembly, contained denatured tubulin, most of the ATPase activity of the microtubule protein, and significant amounts of protein with the electrophoretic mobility of MAP-2. Insoluble material at this and all previous stages, including the preparation of the microtubule protein, could be heat extracted to yield soluble protein active in promoting microtubule assembly and containing MAP-2 as a major constituent. The total soluble MAPs were further purified by DEAE-cellulose chromatography into bound and unbound components, both of which induced microtubule assembly. The bound component (DEAE-MAPs) contained proteins with the electrophoretic mobility of MAP-1, MAP-2, and tau factor. The polymerization reaction induced by the unbound component (flow-through MAPs) produced very high turbidity readings. This was caused by the formation of bundles of microtubules. Although the flow-through MAPs contained significantly more ATPase, tubulin-independent GTPase, and, especially, nucleosidediphosphate kinase activity than the DEAE-MAPs, preparation of a MAPs fraction without these enzymes required heat treatment.     

Stable expression of heterologous microtubule-associated proteins (MAPs) in Chinese hamster ovary cells: evidence for differing roles of MAPs in microtubule organization

To study the effects of microtubule-associated proteins (MAPs) on in vivo microtubule assembly, cDNAs containing the complete coding sequences of a Drosophila 205-kD heat stable MAP, human MAP 4, and human tau were stably transfected into CHO cells. Constitutive expression of the transfected genes was low in most cases and had no obvious effects on the viability of the transfected cell lines. High levels of expression, as judged by Western blots, immunofluorescence, and Northern blots, could be induced by treating cells with sodium butyrate. High levels of MAPs were maintained for at least 24-48 h after removal of the sodium butyrate. Immunofluorescence analysis indicated that all three MAPs bound to cellular microtubules, but only the transfected tau caused a rearrangement of microtubules into bundles. Despite high levels of expression of these exogenous MAPs and the bundling of microtubules in cells expressing tau, transfected cells had normal levels of assembled and unassembled tubulin. With the exception of the tau-induced bundles, microtubules in transfected cells showed the same sensitivity as control cells to microtubule depolymerization by Colcemid. Further, all three MAPs were ineffective in reversing the taxol-dependent phenotype of a CHO mutant cell line. The absence of a quantitative effect of any of these heterologous proteins on the assembly of tubulin suggests that these MAPs may have different roles in vivo from those inferred previously from in vitro experiments.     

Traces of brain microtubule-associated proteins affect dynamic properties of microtubules

A method is described for measuring the quantities of stable and dynamic microtubules in a population in vitro. The method exploits the tendency of dynamic microtubules to depolymerize rapidly after being sheared. Stable microtubules, such as those protected by microtubule-associated proteins (MAPs), are broken to a smaller size by shearing, but do not depolymerize into subunits. The usual difficulty with this procedure is that the tubulin released from the dynamic microtubules rapidly repolymerizes before the end point of depolymerization can be measured. This has been overcome by including a small quantity of tubulin-colchicine complex in the mixture to block the repolymerization. For a total of 24 microM tubulin in a polymerization mixture, 10 microM of the sample polymerized originally under the conditions used. When 1.05 microM tubulin-colchicine complex was added at the time of shearing, the dynamic microtubules depolymerized, but the tubulin was released was unable to repolymerize and a small fraction of stable microtubules that resisted shear-induced depolymerization could then be detected. When traces of MAPs (0.23-2.8% by mass) were included in the tubulin mixture, the fraction of stable microtubules increased from 5% in the absence of added MAPs to 41% in the presence of 2.8% MAPs. All the MAPs in the mixture were found in the stable fraction and this stable fraction forms early during microtubule assembly. Calculations on the extent of enrichment of MAPs in the stable fraction indicated that as little as 4% MAPs in a microtubule protected it from shear-induced disassembly. The results suggest that low levels of MAPs may distribute nonrandomly in the microtubule population.     

Dependency of microtubule-associated proteins (MAPS) for tubulin stability and assembly; use of estramustine phosphate in the study of microtubules

Summary Microtubule-associated proteins (MAPS) were separated from tubulin with several different methods. The ability of the isolated MAPs to reinduce assembly of phosphocellulose purified tubulin differed markedly between the different methods. MAPs isolated by addition of 0.35 M NaCl to taxol-stabilized microtubules stimulated tubulin assembly most effectively, while addition of 0.6M NaCl produced MAPs with a substantially lower ability to stimulate tubulin assembly. The second best preparation was achieved with phosphocellulose chromatographic separation of MAPs with 0.6 M NaCl elution.The addition of estramustine phosphate to microtubules reconstituted of MAPS prepared by 0.35 M NaCl or phosphocellulose chromatography, induced less disassembly than for microtubules assembled from unseparated proteins, and was almost without effect on microtubules reconstituted from MAPs prepared by taxol and 0.6 M NaCl. Estramustine phosphate binds to the tubulin binding part of the MAPs, and the results do therefore indicate that the MAPs are altered by the separation methods. Since the MAPs are regarded as highly stable molecules, one probable alteration could be aggregation of the MAPs, as also indicated by the results. The purified tubulin itself seemed not to be affected by the phosphocellulose purification, since the microtubule proteins were unchanged by the low buffer strenght used during the cromatography. However, the assembly competence after a prolonged incubation of the microtubule proteins at 4° C was dependent on intact bindings between the tubulin and MAPs.Abbreviations Pipes 1,4-Piperazinediethanesulfonic acid - EDTA Ethylenedinitrilo Tetraacetic Acid - MAPs Microtubule-Associated Proteins - SDS-PAGE SDS-Polyacrylamide Gel Electrophoresis     

Generation of microtubule stability subclasses by microtubule- associated proteins: implications for the microtubule "dynamic instability" model

We have developed a method to distinguish microtubule associated protein (MAP)-containing regions from MAP-free regions within a microtubule, or within microtubule sub-populations. In this method, we measure the MAP-dependent stabilization of microtubule regions to dilution-induced disassembly of the polymer. The appropriate microtubule regions are identified by assembly in the presence of [3H]GTP, and assayed by filter trapping and quantitation of microtubule regions that contain label. We find that MAPs bind very rapidly to polymer binding sites and that they do not exchange from these sites measurably once bound. Also, very low concentrations of MAPs yield measurable stabilization of local microtubule regions. Unlike the stable tubule only polypeptide (STOP) proteins, MAPs do not exhibit any sliding behavior under our assay conditions. These results predict the presence of different stability subclasses of microtubules when MAPs are present in less than saturating amounts. The data can readily account for the observed "dynamic instability" of microtubules through unequal MAP distributions. Further, we report that MAP dependent stabilization is quantitatively reversed by MAP phosphorylation, but that calmodulin, in large excess, has no specific influence on MAP protein activity when MAPs are on microtubules.     

Ca2+– and Calmodulin-Dependent Phosphorylation of Microtubule-Associated Protein 2 and t Factor, and Inhibition of Microtubule Assembly

Microtubule-associated proteins (MAPs) were phosphorylated by a Ca2+- and calmodulin-dependent protein kinase from rat brain cytosol. The maximal amount of phosphate incorporated into MAPs was 25 nmol of phosphate/mg protein. A Ka value of the enzyme for calmodulin was 57.0 nM, with MAPs as substrates. Among MAPs, MAP2 and tau factor were phosphorylated in a Ca2+- and calmodulin-dependent manner. The phosphorylation of MAPs led to an inhibition of microtubule assembly in accordance with its degree. This reaction was dependent on addition of the enzyme, Ca2+, and calmodulin, and had a greater effect on the initial rate of microtubule assembly rather than on the final extent. The critical tubulin concentration for microtubule assembly was unchanged by the MAPs phosphorylation. Therefore assembly and disassembly of brain microtubule are regulated by the Ca2+- and calmodulin-dependent protein kinase that requires only a nanomolar concentration of calmodulin for activation.     

On the rigidity of the cytoskeleton: are MAPs crosslinkers or spacers of microtubules?

Microtubules are fibers of the cytoskeleton involved in mitosis, intracellular transport, motility and other functions. They contain microtubule-associated proteins (MAPs) bound to their surface which stabilize microtubules and promote their assembly. There has been a debate on additional functions of MAPs, e.g. whether MAPs crosslink microtubules and thus increase their rigidity, or whether they act as spacers between them. We have studied the packing of microtubules in the presence of MAPs by solution X-ray scattering using synchrotron radiation. Microtubules free in solution produce a scattering pattern typical of an isolated hollow cylinder, whereas tightly packed microtubules generate a pattern dominated by interparticle interference. The interference patterns are interpreted in terms of the Hosemann paracrystal concept, adapted for arrays of parallel fibers with hexagonal arrangement in the plane perpendicular to the fiber axes (Briki et al., 1998). Microtubules without MAPs can rapidly and efficiently be compressed by centrifugation, as judged by the transition from a "free microtubule" to a "packed microtubule" X-ray scattering pattern. MAPs make the microtubule array highly resistant to packing, even at high centrifugal forces. This emphasizes the role of MAPs as spacers of microtubules rather than crosslinkers. A possible function is to keep the microtubule tracks free for the approach of motor proteins carrying vesicle or organelle cargoes along microtubules.     

Effects of microtubule-associated proteins on network formation by neurofilament-induced polymerization of tubulin

It has been revealed that neurofilaments stimulate polymerization of tubulin and thereby cause gelation. Addition of a very small amount of MAPs to the reaction mixture of tubulin and neurofilaments resulted in promotion of gelation. This could not be ascribed to MAP-induced cross-linking between microtubules and neurofilaments because further increases in the MAP concentration (still substoichiometric amount) resulted in total suppression of gelation. It is concluded that MAPs promote microtubule assembly independently of neurofilaments, and lower the concentration of tubulin available for neurofilament-induced polymerization, then preventing network formation.     

Biochemical dissection of the role of the one-kilodalton carboxyl-terminal moiety of tubulin in its assembly into microtubules

The 4-kDa C-terminal domain of both tubulin subunits plays a major role in the regulation of microtubule assembly [Serrano et al. (1984) Biochemistry 23, 4675]. Controlled proteolysis of tubulin with subtilisin produces the selective cleavage of this 4-kDa moiety from alpha- and beta-tubulin with a concomitant enhancement of the assembly. Here we show that gradual removal of the last six to eight amino acid residues of the C-terminal region of alpha and beta subunits by an exopeptidase, carboxypeptidase Y, produces a modified protein (C-tubulin) without relieving the modulatory effect of the C-terminal domain and the usual need of MAPs for microtubule assembly. Actually, treatment with this proteolytic enzyme did not change tubulin assembly as promoted by either MAP-2, taxol, MgCl2, dimethyl sulfoxide, or glycerol. The critical concentration for the assembly of C-tubulin remained the same as that for the unmodified tubulin control. Microtubule-associated proteins MAP-2 and tau incorporated into C-tubulin polymers. Clearly, pure C-tubulin did not assemble in the absence of MAPs or without addition of assembly-promoting compounds. However, proteolysis with the exopeptidase induced changes in tubulin conformation as assessed by biophysical methods and double-limited proteolysis. The cleavage with subtilisin after carboxypeptidase digestion did not result in enhancement of the assembly to the levels observed after the treatment of native tubulin with subtilisin. Interestingly, Ca2+ ions affected neither C-tubulin assembly nor depolymerized microtubules assembled from C-tubulin.(ABSTRACT TRUNCATED AT 250 WORDS)     

Analysis of the microtubule-binding domain of MAP-2

We examined the microtubule-binding domain of the microtubule- associated protein (MAP), MAP-2, using rabbit antibodies that specifically bind to the microtubule-binding region ("stub") and the projection portion ("arm") of MAP-2. We found that (a) microtubules decorated with arm antibody look similar to those labeled with whole unfractionated MAP antibody, though microtubules are not labeled with stub antibody; (b) incubation of depolymerized microtubule protein with stub antibody prior to assembly partially inhibits the rate of microtubule elongation, presumably because MAPs that are complexed with antibody cannot bind to microtubules and stabilize elongating polymers; (c) the rate of appearance and amounts of 36- and 40-kD microtubule- binding peptides produced by digestion with chymotrypsin are distinct for MAPs associated with microtubules vs. MAPs free in solution. The enhanced stability of the 40-kD peptide when associated with microtubules suggests that this domain of the protein is closely associated with, or partially buried in, the microtubule surface; (d) MAP-2 is a slender, elongate molecule as determined by unidirectional platinum shadowing (90 +/- 30 nm), which is in approximate agreement with previous observations. Stub antibody labels MAP-2 in the terminal one-quarter of the extended protein, indicating an intrinsic asymmetry in the molecule.     

Visualization of detyrosination along single microtubules reveals novel mechanisms of assembly during cytoskeletal duplication in trypanosomes

We have been able to use immunogold labeling with monoclonal antibodies specific for tyrosinated alpha-tubulin to define new microtubule assembly within the T. brucei pellicular cytoskeleton. Using this approach, we have been able to visualize and define the detyrosination gradient along single microtubules in vivo. New microtubules are seen to invade the cytoskeletal array early in the cell cycle between old microtubules. In post-mitotic cells, a unique form of microtubule assembly occurs, with very short microtubules being intercalated in the array. We propose that these are nucleated by lateral interaction with the MAPs on existing adjacent microtubules. This construction pattern suggests a templated morphogenesis of microtubule arrays with semi-conservative distribution to the daughter cells.     

Dynamics of microtubules bundled by microtubule associated protein 2C (MAP2C)

MAP2C is a microtubule-associated protein abundant in immature nerve cells. We isolated a cDNA clone encoding whole mouse MAP2C of 467 amino acid residues. In fibroblasts transiently transfected with cDNA of MAP2C, interphase microtubule networks were reorganized into microtubule bundles. To reveal the dynamic properties of microtubule bundles, we analyzed the incorporation sites of exogenously introduced tubulin by microinjection of biotin-labeled tubulin and the turnover rate of microtubule bundles by photoactivation of caged fluorescein- labeled tubulin. The injected biotin-labeled tubulin was rapidly incorporated into distal ends of preexisting microtubule bundles, suggesting a concentration of the available ends of microtubules at this region. Although homogenous staining of microtubule bundles with antibiotin antibody was observed 2 h after injection, the photoactivation study indicated that turnover of microtubule bundles was extremely suppressed and < 10% of tubulin molecules would be exchanged within 1 h. Multiple photoactivation experiments provided evidence that neither catastrophic disassembly at the distal ends of bundles nor concerted disassembly due to treadmilling at the proximal ends could explain the observed rapid incorporation of exogenously introduced tubulin molecules. We conclude that microtubules bundled by MAP2C molecules are very stable while the abrupt increase of free tubulin molecules by microinjection results in rapid assembly from the distal ends within the bundles as well as free nucleation of small microtubules which are progressively associated laterally with preexisting microtubule bundles. This is the first detailed study of the function of MAPs on the dynamics of microtubules in vivo.     

Mechanism of microtubule stabilization by doublecortin

Neurons undertake an amazing journey from the center of the developing mammalian brain to the outer layers of the cerebral cortex. Doublecortin, a component of the microtubule cytoskeleton, is essential in postmitotic neurons and was identified because its mutation disrupts human brain development. Doublecortin stabilizes microtubules and stimulates their polymerization but has no homology with other MAPs. We used electron microscopy to characterize microtubule binding by doublecortin and visualize its binding site. Doublecortin binds selectively to 13 protofilament microtubules, its in vivo substrate, and also causes preferential assembly of 13 protofilament microtubules. This specificity was explained when we found that doublecortin binds between the protofilaments from which microtubules are built, a previously uncharacterized binding site that is ideal for microtubule stabilization. These data reveal the structural basis for doublecortin's binding selectivity and provide insight into its role in maintaining microtubule architecture in maturing neurons.