Properties of the kinetochore in vitro. I. Microtubule nucleation and tubulin binding

We have isolated chromosomes from Chinese hamster ovary cells arrested in mitosis with vinblastine and examined the interactions of their kinetochores with purified tubulin in vitro. The kinetochores nucleate microtubule (MT) growth with complex kinetics. After an initial lag phase, MTs are continuously nucleated with both plus and minus ends distally localized. This mixed polarity seems inconsistent with the formation of an ordered, homopolar kinetochore fiber in vivo. As isolated from vinblastine-arrested cells, kinetochores contain no bound tubulin. The kinetochores of chromosomes isolated from colcemid-arrested cells or of chromosomes incubated with tubulin in vitro are brightly stained after anti-tubulin immunofluorescence. This bound tubulin is probably not in the form of MTs. It is localized to the corona region by immunoelectron microscopy, where it may play a role in MT nucleation in vitro.     

Microtubule assembly affected by the presence of denatured tubulin

The effects of denatured tubulin on microtubule assembly from active phosphocellulose-tubulin have been studied. The presence of denatured tubulin resulted in an inhibition of the assembly and in the increase of the critical concentration to trigger the assembly. Inhibition of both the rate and extent of microtubule assembly was dependent on denatured tubulin concentration. This perturbation of microtubule assembly by denatured tubulin is likely to be specific as non-microtubule proteins did not significantly affect the assembly.     

Microtubule assembly dynamics: new insights at the nanoscale

Although the dynamic self-assembly behavior of microtubule ends has been well characterized at the spatial resolution of light microscopy (~200 nm), the single-molecule events that lead to these dynamics are less clear. Recently, a number of in vitro studies used novel approaches combining laser tweezers, microfabricated chambers, and high-resolution tracking of microtubule-bound beads to characterize mechanochemical aspects of MT dynamics at nanometer scale resolution. In addition, computational modeling is providing a framework for integrating these experimental results into physically plausible models of molecular scale microtubule dynamics. These nanoscale studies are providing new fundamental insights about microtubule assembly, and will be important for advancing our understanding of how microtubule dynamic instability is regulated in vivo via microtubule-associated proteins, therapeutic agents, and mechanical forces.     

Microtubule dynamics in nerve cells: analysis using microinjection of biotinylated tubulin into PC12 cells

To study microtubule (MT) dynamics in nerve cells, we microinjected biotin-labeled tubulin into the cell body of chemically fused and differentiated PC12 cells and performed the immunofluorescence or immunogold procedure using an anti-biotin antibody followed by secondary antibodies coupled to fluorescent dye or colloidal gold. Incorporation of labeled subunits into the cytoskeleton of neurites was observed within minutes after microinjection. Serial electron microscopic reconstruction revealed that existing MTs in PC12 neurites incorporated labeled subunits mainly at their distal ends and the elongation rate of labeled segments was estimated to be less than 0.3 micron/min. Overall organization of MTs in the nerve cells was different from that in undifferentiated cells such as fibroblasts. Namely, we have not identified any MT-organizing centers from which labeled MTs are emanating in the cell bodies of the injected cells. Stereo electron microscopy revealed that some fully labeled segments seemed to start in the close vicinity of electron dense material within the neurites. This suggests new nucleation off some structures in the neurites. We have also studied the overall pattern of the incorporation of labeled subunits which extended progressively from the proximal part of the neurites toward their tips. To characterize the mechanism of tubulin incorporation, we have measured mean density of gold labeling per unit length of labeled segments at different parts of the neurites. The results indicate access of free tubulin subunits into the neurites and local incorporation into the neurite cytoskeleton. Our results lead to the conclusion that MTs are not static polymers but dynamic structures that continue to elongate even within the differentiated nerve cell processes.     

Coherent singlet-triplet transitions at the initial step of tubulin assembly into microtubules

Quantum chemistry calculations [DFT-B3LYP QM/MM method, 6-31G** basis set, + ab initio molecular dynamics] were used to study the action of Mg2+ on tubulin properties. It was shown that the hydration of the guanosine triphosphate-tubulin forms a protein zone structure, which includes a electron-occupied zone and a conductivity zone. The binding of Mg2+ to guanosine triphosphate-tubulin results in the unpairing of electrons in the occupied zone (triplet state formation) followed by their transition to the conductivity zone in which the inversion of spin occurs (singlet state formation). The formation of triplet state is the initial step in the subsequent protein dynamics in the picosecond range of time. The dynamics shows up as a coherent oscillating transition of tubulin between the triplet and singlet states, which is evidence of a simultaneous adjustment between nuclear and electron configurations of the protein (ab initio molecular dynamics calculations). The barrier between the triplet and singlet states does not exceed 0.60 kcal x mol(-1). The barrier overcome is considered as electron tunneling through the Fermi surface, which separates the occupied and conductivity zones. Zone formation occurs in the presence of the shell of biological water surrounding the protein.     

Microtubule initiation at kinetochores and centrosomes in lysed mitotic cells. Inhibition of site-specific nucleation by tubulin antibody

A lysed cell system was developed to determine whether tubulin antibody can block the nucleation of exogenous tubulin at kinetochores and centrosomes. Mitotic PtK2 cells were pretreated with colcemid to remove all endogenous microtubules and were lysed with Triton X-100 in PIPES-EGTA-Mg++ buffer. This procedure left centrosomes, chromosomes, and kinetochores intact as determined by electron microscopy of thin-sectioned cells. Exposure of the lysed cells to phorphocellulose-purified tubulin dimers at 37 degrees C in the presence of 1 mM GTP resulted in site-specific nucleation of microtubules at centrosomes and kinetochores. Treatment of the lysed cell preparations with tubulin antibody before subsequent exposure to the exogenous tubulin resulted in almost complete blockage of microtubule nucleation, especially at kinetochores. Pretreatment of the lysed cell preparations with control antibody or buffer without antibody had no effect on the ability of centrosomes and kinetochores to initiate microtubule assembly. The implications of these results with respect to the molecular composition of centrosomes and kinetochores are discussed.     

Taxol stabilization of microtubules in vitro: dynamics of tubulin addition and loss at opposite microtubule ends

We have investigated the effects of taxol on steady-state tubulin flux and on the apparent molecular rate constants for tubulin addition and loss at the two ends of bovine brain microtubules in vitro. These microtubules, which consist of a mixture of 70% tubulin and 30% microtubule-associated proteins (MAPs), undergo a net addition of tubulin at one end of each microtubule (A end) and a precisely balanced net loss of tubulin at the opposite end (D end) at steady state in vitro. They do not exhibit to a detectable extent the "dynamic instability" behavior described recently for MAP-free microtubules, which would be evident as an increase in the mean microtubule length and a decrease in the number of microtubules in the suspensions [Mitchison, T., & Kirschner, M. (1984) Nature (London) 312, 237-242]. We used a double-label procedure in which microtubules were labeled with tritium and carbon-14 at A ends and carbon-14 at D ends to distinguish the two ends, combined with a microtubule collection procedure that permitted rapid and accurate analysis of retention of the two labels in the microtubules. We found that taxol slowed the flux of tubulin in a concentration-dependent manner, with 50% inhibition occurring between 5 and 7 microM drug. The effects of taxol on the apparent molecular rate constants for tubulin addition and loss at the two microtubule ends were determined by dilution analysis at an intermediate taxol concentration. The results indicated that taxol decreased the magnitudes of the dissociation rate constants at the two ends to similar extents, while exerting little effect on the association rate constants.(ABSTRACT TRUNCATED AT 250 WORDS)     

Microtubule dynamics at the G2/M transition: abrupt breakdown of cytoplasmic microtubules at nuclear envelope breakdown and implications for spindle morphogenesis

We recently developed a direct fluorescence ratio assay (Zhai, Y., and G.G. Borisy. 1994. J. Cell Sci. 107:881-890) to quantify microtubule (MT) polymer in order to determine if net MT depolymerization occurred upon anaphase onset as the spindle was disassembled. Our results showed no net decrease in polymer, indicating that the disassembly of kinetochore MTs was balanced by assembly of midbody and astral MTs. Thus, the mitosis-interphase transition occurs by a redistribution of tubulin among different classes of MTs at essentially constant polymer level. We now examine the reverse process, the interphase-mitosis transition. Specifically, we quantitated both the level of MT polymer and the dynamics of MTs during the G2/M transition using the fluorescence ratio assay and a fluorescence photoactivation approach, respectively. Prophase cells before nuclear envelope breakdown (NEB) had high levels of MT polymer (62%) similar to that previously reported for random interphase populations (68%). However, prophase cells just after NEB had significantly reduced levels (23%) which recovered as MT attachments to chromosomes were made (prometaphase, 47%; metaphase, 56%). The abrupt reorganization of MTs at NEB was corroborated by anti- tubulin immunofluorescence staining using a variety of fixation protocols. Sensitivity to nocodazole also increased at NEB. Photoactivation analyses of MT dynamics showed a similar abrupt change at NEB, basal rates of MT turnover (pre-NEB) increased post-NEB and then became slower later in mitosis. Our results indicate that the interphase-mitosis (G2/M) transition of the MT array does not occur by a simple redistribution of tubulin at constant polymer level as the mitosis-interphase (M/G1) transition. Rather, an abrupt decrease in MT polymer level and increase in MT dynamics occurs tightly correlated with NEB. A subsequent increase in MT polymer level and decrease in MT dynamics occurs correlated with chromosome attachment. These results carry implications for understanding spindle morphogenesis. They indicate that changes in MT dynamics may cause the steady-state MT polymer level in mitotic cells to be lower than in interphase. We propose that tension exerted on the kMTs may lead to their lengthening and thereby lead to an increase in the MT polymer level as chromosomes attach to the spindle.     

Microtubule organization during maturation of Xenopus oocytes: assembly and rotation of the meiotic spindles.

Assembly of the meiotic spindles during progesterone-induced maturation of Xenopus oocytes was examined by confocal fluorescence microscopy using anti-tubulin antibodies and by time-lapse confocal microscopy of living oocytes microinjected with fluorescent tubulin. Assembly of a transient microtubule array from a disk-shaped MTOC was observed soon after germinal vesicle breakdown. This MTOC-TMA complex rapidly migrated toward the animal pole, in association with the condensing meiotic chromosomes. Four common stages were observed during the assembly of both M1 and M2 spindles: (1) formation of a compact aggregate of microtubules and chromosomes; (2) reorganization of this aggregate resulting in formation of a short bipolar spindle; (3) an anaphase-B-like elongation of the prometaphase spindle, transversely oriented with respect to the oocyte A-V axis; and (4) rotation of the spindle into alignment with the oocyte axis. The rate of spindle elongation observed in M1 (0.7 microns min-1) was slower than that observed in M2 (1.8 microns min-1). Examination of spindles by immunofluorescence with antitubulin revealed numerous interdigitating microtubules, suggesting that prometaphase elongation of meiotic spindles in Xenopus oocytes results from active sliding of antiparallel microtubules. A substantial number of maturing oocytes formed monopolar microtubule asters during M1, nucleated by hollow spherical MTOCs. These monasters were subsequently observed to develop into bipolar M1 spindles and proceed through meiosis. The results presented define a complex pathway for assembly and rotation of the meiotic spindles during maturation of Xenopus oocytes.     

Stimulation of the CLIP-170--dependent capture of membrane organelles by microtubules through fine tuning of microtubule assembly dynamics

Cytoplasmic microtubules (MTs) continuously grow and shorten at their free plus ends, a behavior that allows them to capture membrane organelles destined for MT minus end-directed transport. In Xenopus melanophores, the capture of pigment granules (melanosomes) involves the +TIP CLIP-170, which is enriched at growing MT plus ends. Here we used Xenopus melanophores to test whether signals that stimulate minus end MT transport also enhance CLIP-170-dependent binding of melanosomes to MT tips. We found that these signals significantly (>twofold) increased the number of growing MT plus ends and their density at the cell periphery, thereby enhancing the likelihood of interaction with dispersed melanosomes. Computational simulations showed that local and global increases in the density of CLIP-170-decorated MT plus ends could reduce the half-time of melanosome aggregation by ~50%. We conclude that pigment granule aggregation signals in melanophores stimulate MT minus end-directed transport by the increasing number of growing MT plus ends decorated with CLIP-170 and redistributing these ends to more efficiently capture melanosomes throughout the cytoplasm.     

The force for poleward chromosome motion in Haemanthus cells acts along the length of the chromosome during metaphase but only at the kinetochore during anaphase

The force for poleward chromosome motion during mitosis is thought to act, in all higher organisms, exclusively through the kinetochore. We have used time-lapse. video-enhanced, differential interference contrast light microscopy to determine the behavior of kinetochore-free "acentric" chromosome fragments and "monocentric" chromosomes containing one kinetochore, created at various stages of mitosis in living higher plant (Haemanthus) cells by laser microsurgery. Acentric fragments and monocentric chromosomes generated during spindle formation and metaphase both moved towards the closest spindle pole at a rate (approximately 1.0 microm/min) similar to the poleward motion of anaphase chromosomes. This poleward transport of chromosome fragments ceased near the onset of anaphase and was replaced. near midanaphase, by another force that now transported the fragments to the spindle equator at 1.5-2.0 microm/min. These fragments then remained near the spindle midzone until phragmoplast development, at which time they were again transported randomly poleward but now at approximately 3 microm/min. This behavior of acentric chromosome fragments on anastral plant spindles differs from that reported for the astral spindles of vertebrate cells, and demonstrates that in forming plant spindles, a force for poleward chromosome motion is generated independent of the kinetochore. The data further suggest that the three stages of non- kinetochore chromosome transport we observed are all mediated by the spindle microtubules. Finally, our findings reveal that there are fundamental differences between the transport properties of forming mitotic spindles in plants and vertebrates.     

Microtubule inhibitors alter the secretion of beta-glucuronidase by human blood platelets: involvement of microtubules in release reaction II

The effects of the antimicrotubular drugs colchicine and vinblastine on the blood platelet release reaction were studied by measuring release of ¹⁴C-5-hydroxytryptamine (¹⁴C-5-HT, release I) and β-glucuronidase (release II) from gel-filtered human platelets. β-glucuronidase release induced by thrombin was significantly inhibited by colchicine (0.01-1 mM) or vinblastine (0.05–0.1 mM). Release of ¹⁴C-5-HT, however, was unaffected at low concentrations of colchicine and only slightly inhibited at higher concentrations. Inhibition of β-glucuronidase release depended on colchicine or vinblastine concentrations and decreased with longer time intervals (1′, 5′, 20′) after thrombin stimulation. Levels of the cytoplasmic enzyme, lactic acid dehydrogenase, in supernatants of colchicine treated platelets were not significantly different from controls. Colchicine also inhibited β-glucuronidse release, but not ¹⁴C-5-HT release, induced by trypsin and sodium arachidonate. Binding of ¹⁴C-colchicine by platelets was measured and it was found that platelet aggregation and release of 5-HT induced by adenosine diphosphate, epinephrine and collagen proceeded without any alteration in colchicine binding. However, significant increases in the rate and degree of colchicine binding were observed when platelets were stimulated by thrombin, trypsin and arachidonic acid which induced aggregation, release of both 5-HT and β-glucuronidase. The results suggest that an alteration in platelet microtubules is correlated with the physiologic response resulting in release II and that the cellular mechanisms effecting release I and II by platelets differ qualitatively in that the microtubules may facilitate release II.     

Microtubule plus end‐tracking proteins play critical roles in directional growth of hyphae by regulating the dynamics of cytoplasmic microtubules in Aspergillus nidulans

Cytoplasmic microtubules (MTs) serve as a rate‐limiting factor for hyphal tip growth in the filamentous fungus Aspergillus nidulans. We hypothesized that this function depended on the MT plus end‐tracking proteins (+TIPs) including the EB1 family protein EBA that decorated the MT plus ends undergoing polymerization. The ebAΔ mutation reduced colony growth and the mutant hyphae appeared in an undulating pattern instead of exhibiting unidirectional growth in the control. These phenotypes were enhanced by a mutation in another +TIP gene clipA. EBA was required for plus end‐tracking of CLIPA, the Kinesin‐7 motor KipA, and the XMAP215 homologue AlpA. In addition, cytoplasmic dynein also depended on EBA to track on most polymerizing MT plus ends, but not for its conspicuous appearance at the MT ends near the hyphal apex. The loss of EBA reduced the number of cytoplasmic MTs and prolonged dwelling times for MTs after reaching the hyphal apex. Finally, we found that colonies were formed in the absence of EBA, CLIPA, and NUDA together, suggesting that they were dispensable for fundamental functions of MTs. This study provided a comprehensive delineation of the relationship among different +TIPs and their contributions to MT dynamics and unidirectional hyphal expansion in filamentous fungi.     

Microtubule assembly is required for the formation of the pronuclei,nuclear lamin acquisition,and DNA synthesis during mouse,but not sea urchin,fertillization

Microtubule assembly is required for the formation of the male and female pronuclei during mouse, but not sea urchin, fertilization. In mouse oocytes, 50 μM colcemid prevents the decondensation of the maternal meiotic chromosomes and of the incorporated sperm nucleus during in vitro fertilization. Nuclear lamins do not associate with either of the parental chromatin sets although peripherin, the PI nuclear peripheral antigen, appears on both. DN A synthesis docs not occur in these fertilized, colcemid-arrested oocytes. This effect is limited to the first hours after ovulation, since colcemid added 4–6 hours later no longer prevents pronuclear development, lamin acquisition, or DNA synthesis. Neither microtubule stabilization with 10 μM taxol nor microfilament inhibition with 10 μM cytochalasin D or 2.2 μg/ml lalrunculin A prevent these pronuclear events; these drugs will inhibit the apposition of the pronuclei at the egg center. In sea urchin eggs, colcemid or griseofulvin treatment doe? not result in the same effect and the male pronucleus forms with the attendant accumulation of the nuclear lamins. The differences in the requirement for microtubule assembly during pronucleus formation may be related to the cell cycle: In mice the sperm enters a meiotic cytoplasm, whereas in sea urchin eggs it enters an interphase cytoplasm. Refertilization of mitotic sea urchin eggs was performed to test the possibility that this phenomenon is related to whether the sperm enters a meiotic/mitotic cytoplasm or one at interphase; during refertilization at first mitosis, the incorporated sperm nucleus is unable to decondense and acquire lamins. These results indicate a requirement for microtubule assembly for the progression from meiosis to first interphase during mouse fertilization and suggest that the cytoskeleton is required for changes in nuclear architecture necessary during fertilization and the cell cycle.     

Stoichiometry of estramustine phosphate binding to MAP2 measured by the disassembly of chick brain MAP2:tubulin microtubules

The concentration of estramustine phosphate required to inhibit the assembly or to induce the disassembly of chick brain MAP2:tubulin microtubules is markedly dependent upon the microtubule protein concentration. Analysis of this relationship shows that estramustine phosphate and tubulin compete for common MAP2 sites, that MAP2 can bind 5-6 moles.mole-1 estramustine phosphate, and that the Kd of these sites is congruent to 20 microM estramustine phosphate. It is proposed that two molecules of estramustine phosphate interact with each of the three tubulin-binding sites of MAP2 and inhibit the MAP2:tubulin interaction by neutralising two highly conserved basic residues.     

Preferential incorporation of tubulin into the junctional region of ciliary outer doublet microtubules: a model for treadmilling by lattice dislocation

Even in the presence of colchicine or Taxol(R), sea urchin embryonic cilia undergo substantial steady-state turnover, with a rate of tubulin incorporation approaching half that seen in full regeneration [Stephens: Mol Biol Cell 8:2187-2198, 1997]. Preliminary experiments suggest that tubulin incorporates differentially into the most stable portion of the outer doublet, the junctional protofilaments [Stephens: Cell Struct Funct 24:413-418, 1999]. To explore this possibility further, embryos of the sea urchin Tripneustes gratilla, a ciliary length inducible system [Stephens: J Exp Zool 269:106-115, 1994a], were pulse labeled with (3)H leucine during steady-state turnover or induced elongation, followed by regeneration in the presence of unlabeled leucine. Cilia were isolated by hypertonic shock and fractionated into detergent-soluble membrane plus matrix, thermally-solubilized microtubule walls, and insoluble 9-fold symmetric remnants of A-B junctional protofilaments plus associated architectural elements. The fractions were resolved by SDS-PAGE and the specific activity of alpha-tubulin was determined. In cilia undergoing turnover or elongation during an isotope pulse, the specific activity of tubulin in the junctional region approximated that of precursor membrane plus matrix tubulin but surpassed that of the tubule wall by a factor of approximately 1.5. In cilia regenerated during an isotope chase, the specific activity of junctional tubulin exceeded that of both the membrane plus matrix and the tubule wall by a similar factor. These data indicate that tubulin is preferentially incorporated into junctional protofilaments during steady-state turnover, induced elongation and regeneration. A model for directional incorporation based on surface lattice discontinuities in the outer doublet is proposed.     

Membrane phospholipid dynamics during cytokinesis: regulation of actin filament assembly by redistribution of membrane surface phospholipid

To study molecular motion and function of membrane phospholipids, we have developed various probes which bind specifically to certain phospholipids. Using a novel peptide probe, RoO9-0198, which binds specifically to phosphatidylethanolamine (PE) in biological membranes, we have analyzed the cell surface movement of PE in dividing CHO cells. We found that PE was exposed on the cell surface specifically at the cleavage furrow during the late telophase of cytokinesis. PE was exposed on the cell surface only during the late telophase and no alteration in the distribution of the plasma membranebound peptide was observed during the cytokinesis, suggesting that the surface exposure of PE reflects the enhanced transbilayer movement of PE at the cleavage furrow. Furthermore, cell surface immobilization of PE induced by adding of the cyclic peptide coupled with streptavidin to prometaphase cells effectively blocked the cytokinesis at late telophase. The peptide-streptavidin complex bound specifically to cleavage furrow and inhibited both actin filament disassembly at cleavage furrow and subsequent plasma membrane fusion. Binding of the peptide complex to interphase cells also induced immediate disassembly of stress fibers followed by assembly of cortical actin filaments to the local area of plasma membrane where the peptide complex bound. The cytoskeletal reorganizations caused by the peptide complex were fully reversible; removal of the surface-bound peptide complex by incubating with PE-containing liposome caused gradual disassembly of the cortical actin filaments and subsequent formation of stress fibers. These observations suggest that the redistribution of plasma membrane phospholipids act as a regulator of actin cytoskeleton organization and may play a crucial role in mediating a coordinate movement between plasma membrane and actin cytoskeleton to achieve successful cell division.     

F actin assembly modulated by villin: Ca++-dependent nucleation and capping of the barbed end

We have studied the mechanism of Ca++-dependent restriction of actin filament length by villin, one of the major actin-associated proteins of intestinal microvilli microfilament bundles. Villin acts, even at a ratio of 1 to 1000 with respect to actin, very efficiently as a Ca++-dependent nucleation factor on actin assembly. This gives rise to unidirectional assembly, with the morphologically defined "barbed" end of the resulting filament being capped. Consequently, at steady state treadmilling of actin monomers through the filament is inhibited. Increase of the villin-to-actin ratio enhances the number of nucleated filaments necessarily shorter in length. This results finally in nonsedimentable F actin and a low molecular weight complex of one villin and three monomeric actins, which itself is a potent nucleator. Thus restriction of actin assembly by villin is not due to a direct inhibition of assembly but arises as the consequence of strongly enhanced nucleation followed by unidirectional elongation at the pointed end of the nucleated filaments. In addition, in the presence of Ca++-villin, but not the villin-actin complex, seems able to "break" or "sever" preformed F actin filaments. Thus a variety of cellular phenomena-nucleation, unidirectional assembly, filament end capping, nonpolymerizable actin and F actin bundles-can be observed in vitro in a two-protein component system modulated by the concentration of free Ca++.