Search

In mammalian cells the centrosome or diplosome is defined by the two parental centrioles observed in electron microscopy and by the pericentriolar material immunostained with several antibodies directed against various centrosomal proteins (gamma-tubulin, pericentrin, centrin and centractin). Partial destabilization of the microtubule cytoskeleton by microtubule-disassembling substances induced a splitting and a slow migration of the two diplosome units to opposite nuclear sides during most of the interphase in several mammalian cell lines. These units relocated close together following drug removal, while microtubule stabilization by nM taxol concentrations inhibited this process. Cytochalasin slowed down diplosome splitting but did not affect its relocation after colcemid washing. These results account for the apparently opposite effects induced by microtubule poisons on centriole separation. Moreover, they provide new information concerning the centrosome cycle and stability. First, the centrosome is formed by two units, distinguished only by the number of attached stable microtubules, but not by pericentrin, gamma-tubulin, centrin and centractin and their potency to nucleate microtubules. Second, the centrosomal units are independent during most of the interphase. Third, according to the cell type, these centrosomal units are localized in close proximity because they are either linked or maintained close together by the normal dynamics of the microtubule cytoskeleton. Finally, the relocalization of the centrosomal units with their centrioles in cells possessing one or two centrosomes suggests that their relative position results from the overall tensional forces involving at least partially the microtubule arrays nucleated by each of these entities.

Cells of eukaryotic organisms exhibit microtubules with various functions during the different developmental stages. The identification of multiple forms of alpha- and beta-tubulins had raised the question of their possible physiological roles. In the myxomycete Physarum polycephalum a complex polymorphism for alpha- and beta-tubulins has been correlated with a specific developmental expression pattern. Here, we have investigated the potential heterogeneity of gamma-tubulin in this organism. A single gene, with 3 introns and 4 exons, and a single mRNA coding for gamma-tubulin were detected. They coded for a polypeptide of 454 amino acids, with a predicted molecular mass of 50,674, which presented 64–76% identity with other gamma-tubulins. However, immunological studies identified two gamma-tubulin polypeptides, both present in the two developmental stages of the organism, uninucleate amoebae and multinucleate plasmodia. The two gamma-tubulins, called gamma s- and gamma f-tubulin for slow and fast electrophoretic mobility, exhibited apparent molecular masses of 52,000 and 50,000, respectively. They were recognized by two antibodies (R70 and JH46) raised against two distinct conserved sequences of gamma-tubulins. They were present both in the preparations of amoebal centrosomes possessing two centrioles and in the preparations of plasmodial nuclear metaphases devoid of structurally distinct polar structures. These two gamma-tubulins exhibited different sedimentation properties as shown by ultracentrifugation and sedimentation in sucrose gradients. Moreover, gamma s-tubulin was tightly bound to microtubule organizing centers (MTOCs) while gamma f-tubulin was loosely associated with these structures. This first demonstration of the presence of two gamma-tubulins with distinct properties in the same MTOC suggests a more complex physiological role than previously assumed.

Six novel docetaxel analogues that possess a N-(7-nitrobenz-2-oxa-1,3-diazo-4-yl)amido-6-caproyl chain in position 7 or 3′ (11 and 16a), a N-(7-nitrobenz-2-oxa-1,3-diazo-4-yl)amido-3-propanoyl group at 3′ (16b) and a 5′-biotinylamido-6-caproyl chain in position 7, 10 or 3′, respectively, have been synthesized. These compounds exhibit activity against microtubule disassembly similar to that of docetaxel but show discrepant activities on living cells. Although addition of microtubules to 11, 16a and b enhance their fluorescence, no shift of the emission maxima was observed. The fluorescent docetaxel derivatives show a specific labeling of microtubules in living cells, demonstrating that the microtubule cytoskeleton constitutes their main subcellular localization.

We have investigated the effects of the microtubule poison rhazinilam on microtubule assembly in vivo and in vitro. In mammalian cells, rhazinilam mimics the effects of taxol and leads to microtubule bundles, multiple asters, and microtubule cold stability. In vitro, rhazinilam protected preassembled microtubules from cold-induced disassembly, but not from calcium ion-induced disassembly. Moreover, both at 0 degrees C and at 37 degrees C, rhazinilam induced the formation of anomalous tubulin assemblies (spirals). This process was prevented by maytansine and vinblastine, but not by colchicine. Preferential saturable and stoichiometric binding of radioactive rhazinilam to tubulin in spirals was observed with a dissociation constant of 5 microM. This binding was abolished in the presence of vinblastine and maytansine. In contrast, specific binding of radioactive rhazinilam to tubulin assembled in microtubules was undetectable. These results demonstrate that rhazinilam alters microtubule stability differently than taxol, and that the overall similar effects of rhazinilam and taxol on the cellular cytoskeleton are the consequence of two distinct mechanisms of action at the molecular level.

Animal cells undergoing cytokinesis form an inter-cellular bridge containing two bundles of microtubules interdigitated at their plus ends, which constitute the midbody. Polyclonal antibodies raised against three specific amino acid sequences of gamma-tubulin (EEFATEGGDRKDV, NIIQGEADPTDVHKSL and EYHAATRPDYISWGTQEQ) specifically stained the centrosome in interphase, the spindle poles in all stages of mitosis, and the extremities of the midbody in mammalian cells (Potorous, human, Chinese hamster, mouse). This staining was prevented by the corresponding peptides, by Xenopus gamma-tubulin, but was not modified by purified alpha beta-tubulin heterodimer. An identical staining was obtained with affinity-purified antibodies against the carboxyl-terminal amino acid sequence of human gamma-tubulin. No gamma-tubulin could be detected in the interzone during anaphase and early telophase. Material containing gamma-tubulin first appeared in the two daughter cells on each side of the division plane in late telophase, and accumulated transiently at the minus ends of the two microtubule bundles constituting the midbody for one hour after metaphase. Micro-injection of gamma-tubulin antibodies into anaphase cells prevented the subsequent formation of the microtubule bundles between the two daughter cells. In contrast with previous views, these observations suggest that the microtubules constituting the midbody may be nucleated on special microtubule organizing centres, active during late telophase only, and assembled on each side of the dividing plane between the daughter cells.

Physarum possesses two different microtubule cytoskeletons. In amoebae, cytoplasmic and mitotic microtubules are nucleated by a typical centrosome. In contrast, it has been reported that plasmodia have an intranuclear spindle organizing centre (SPOC) devoid of centrioles. We present genetic evidence suggesting that the SPOC located in the centrosome is very similar to the intranuclear plasmodial SPOC. The immunostaining properties of a new monoclonal antibody against Physarum centrosome has been used to compare these different MTOCs. Moreover, a dense plasmodial microtubule network was present in interphase plasmodia and absent in plasmodia undergoing mitosis. MTOCs responsible for the nucleation of the cytoplasmic microtubule network and intranuclear SPOCs were located in two different compartments of the plasmodium.

The functional significance of distinct gamma-tubulins in several unrelated eukaryotes remains an enigma due to the difficulties to investigate this question experimentally. Using specific nucleotidic and immunological probes, we have demonstrated that the two divergent Drosophila gamma-tubulins, gamma-tub23C and gamma-tub37CD, are expressed in cultured cells. Gamma-tub37CD is constantly detected at the centrosome and absent in the mitotic spindle, while gamma-tub23C is extensively recruited to the centrosome during mitosis and relocalizes in the mitotic spindle. The two gamma-tubulins exhibit distinct biochemical properties. Gamma-tub23C is present in the soluble gamma-tubulin small complexes (10S) and gamma-tubulin big complexes (35S) and is loosely associated to the cytoskeleton. In contrast, gamma-tub37CD is undetectable in the soluble fraction and exhibits a tight binding to the centrosome. Syncytial embryos also contain the two gamma-tubulin isotypes, which are differentially recruited at the centrosome. Gamma-tub23C is present in the 10S soluble complexes only, while y-tub37CD is contained in the two soluble complexes and is recruited at the centrosome where it exhibits an heterogeneous binding. These results demonstrated an heterogeneity of the two Drosophila gamma-tubulin isotypes both in the cytoskeletal and the soluble fractions. They suggest the direct implication of the 35S complex in the centrosomal recruitment of gamma-tubulin and a conditional functional redundancy between the two gamma-tubulins.

Vinflunine, a newly synthesized derivative, possesses marked in vivo antitumor properties and, like other alkaloids, inhibits in vitro tubulin assembly at microM concentrations. However, in contrast to other vinca alkaloids, vinflunine exhibits relatively low in vitro cytotoxic potency. The aim of this report was to investigate whether the action(s) of vinflunine on the microtubule cytoskeleton could account for its cytotoxicity or if its cellular action requires another molecular target. Four vinca alkaloids used in cancer therapy and vinflunine were studied using PtK2 cells. Their activities on the most dynamic microtubules were investigated in mitosis and in interphase by evaluating the disturbance of the metaphase plate and the splitting of the diplosome, respectively. No correlation was observed between the cellular accumulation of these compounds and either their cytotoxicity or their action(s) on the microtubule cytoskeleton. In contrast, cytotoxicity, mitotic disturbance and diplosome splitting were observed in the nM range for vinblastine, vincristine, vindesine and vinorelbine, although these events occurred at 10 times higher concentrations in the case of vinflunine. Hence, dynamic modifications of both the mitotic and interphasic microtubule cytoskeleton are compatible with in vitro cytotoxicity of vinflunine, raising questions about the conventional biochemical screening of these vinca alkaloids.

The variations of thymidine kinase or ATP:thymidine 5'-phosphotransferase (EC 2.7.1.21) during the cell cycle of Physarum polycephalum plasmodia have been studied at two extreme physiological temperatures: 22 degrees C and 32 degrees C. At 22 degrees C the enzyme activity increases near mitosis and stays constant during late S and G2 phases, exhibiting the typical pattern of a 'step enzyme'. But at 32 degrees C thymidine kinase activity goes through a maximum 1 h 30 min after mitosis and decreases during the subsequent phases as expected for a 'peak enzyme'. The rate of enzyme degradation and/or inactivation, measured in the presence of metabolic poisons (cycloheximide or dinitrophenol), appears to follow a simple exponential function with a half-life of approximately 3 h and 1 h at 22 degrees C and 32 degrees C respectively. The effect of growth temperature on the decrease of thymidine kinase activity can account entirely for the differences in the pattern of enzyme activity at the two extreme temperatures. Tentative calculations indicate that the rate of enzyme synthesis is nearly constant during the cell cycle except near mitosis, where it is temporarily increased. The results suggest the existence of a regulatory mechanism able to modulate the rate of synthesis of thymidine kinase during the cell cycle.

Physarum synchronous plasmodia were submitted to temperature shifts during the cell cycle and the onset of mitosis was followed at both temperatures. After 22 to 31 or 32° shifts, delays in mitosis onset, dependent upon protein synthesis, were observed at 32° and found to increase as the time separating the shift from the control mitosis decreases. The modification of a general metabolic process or the inactivation of a catalytic heat sensitive substance cannot account for such a result. The proposed model postulates a substance acting in a stoichiometric way, which can occur under three structural forms: two active forms synthesized at low and high temperatures respectively and an inactive one which comes from the transformation of the low temperature active form placed at high temperature. The constant delays observed after some shifts (29 to 32°) suggest that this substance is acting through a polymeric structure which would be necessary for the mitotic process and the initiation of the following DNA synthesis.

Temperature shifts from 22 to 32 °C perturb one of the systems responsible for mitosis triggering in the plasmodia of Physarum (Myxomycetes). In order to determine if the same regulatory mechanism could also be involved in some other cell cycle events, the effects of temperature shifts on the peak of thymidine kinase (EC 2.7.1.21, ATP : thymidine 5′-phosphotransferase) synthesis have been studied. At 22 °C, the increase in thymidine kinase (tdk) activity begins shortly before mitosis and is thus always associated with the end of the G2 phase, the mitosis and the beginning of the S phase. The consequences of temperature shifts depend upon their position in the cell cycle. In all cases, a peak of tdk occurs concomitantly with the 32 °C mitosis. But, when the temperature shift is applied 90-15 min before the control metaphase at 22 °C, another peak of tdk is observed at 32 °C in absence of mitosis, but at the same time as the control mitosis at 22 °C. These results indicate that the increase in the synthesis of tdk is controlled by the heat-sensitive regulatory system which plays a role in the onset of mitosis and S phase. We further suggest that the increase in the synthesis of tdk and the triggering of mitosis are both controlled by the amount of a heat-sensitive effector. But the former takes place when the amount of the effector reaches a critical value lower than the value necessary to trigger mitosis.

The effects of hydroxyurea have been investigated on three events of the cell cycle, S-phase, mitosis, and the cyclic synthesis of thymidine kinase, in the synchronous plasmodium of the myxomycete Physarum. DNA synthesis was slowed down with limited action on other macromolecular syntheses and any increase of thymidine kinase that had already been triggered was indistinguishable from that of the control. When DNA synthesis was inhibited, the onset of the following cyclic increase of thymidine kinase synthesis occurred at the same time as in the control, but mitosis was delayed in a very early prophase stage. The arrest of thymidine kinase synthesis occurred after completion of the delayed mitosis. All these effects were suppressed when the action of hydroxyurea was prevented by the addition, to the medium, of the four deoxyribonucleosides. These observations show that (1). The blockage of S-phase does not prevent the nuclei from entering a very early prophase stage but does prevent them from proceeding through metaphase. (2) The transient blockage of DNA synthesis does not perturb the normal timing of the triggering of thymidine kinase synthesis. (3) The signal which triggers the arrest of thymidine kinase synthesis is postmitotic but does not require extensive DNA synthesis. The effect of hydroxyurea is not limited to an inhibition of S-phase. The blockage of DNA replication also led to the dissociation of the normal coordination between two other events of the cell cycle, mitosis and thymidine kinase synthesis. This observation could have strong implications in cell synchronization with chemical agents.

When synchronous plasmodia of the myxomycete Physarum polycephalum were submitted to temperature shifts from 22 degrees C to 32 degrees C, the highest physiological temperature, protein synthesis was increased during at least 10 h. Moreover during 2 h, four proteins (69, 74, 82 and 105 kDa) showed a transient increase of their synthesis, independently of the period of the temperature shift during the cell cycle. The stability of these proteins and the susceptibility of their synthesis to actinomycin suggested that they corresponded to four different proteins. Temperature shifts from 22 degrees C or 29 degrees C to 37 degrees C, a non-physiological temperature, demonstrated that the 69-kDa, 74-kDa, 82-kDa and 105-kDa proteins were identical to the four heat-shock proteins which could be detected in Physarum. Although the physiological significance of these heat-shock proteins remained unclear, comparison between the extent of their synthesis and the length of the mitotic delays induced by various temperature shifts ruled out a direct relationship between mitotic delays and synthesis of the 74-kDa, 82-kDa and 105-kDa proteins.

It has been claimed repeatedly that gamma-tubulin is exclusively localized at the spindle poles in mitotic animal cells, where it plays a role in microtubule nucleation. In addition to this localization, we have observed a gamma-tubulin-specific staining of the mitotic spindle in several animal cells (human, kangaroo rat, mouse, Chinese hamster, Xenopus and Drosophila) using five polyclonal antibodies raised against unique gamma-tubulin sequences and four different fixation protocols. In HeLa and PtK2 cells, gamma-tubulin was detected in the mitotic spindle from late prometaphase to telophase. In contrast, in other cell types, it was detected in metaphase only. In all cases we failed to detect gamma-tubulin in the short aster microtubules at the spindle poles. Electron microscopic observation revealed that at least part of the gamma-tubulin localized on the surface of spindle microtubules with a preferential distribution along kinetochore microtubules. In HeLa cells, the amount of antigenic gamma-tubulin was fairly constant in the spindle poles during mitosis from prometaphase to telophase. In contrast, gamma-tubulin appeared in the mitotic spindles in prometaphase. The amount of gamma-tubulin decreased in telophase, where it relocalized in the interzone. In metaphase cells about 15–25% of the total fluorescence was localized at the spindle poles, while 75–85% of the fluorescence was distributed over the rest of the spindle. These results suggest that the localization and timing of gamma-tubulin during the cell cycle is highly regulated and that is physiological role could be more complex and diverse than initially assumed.