Organized microtubule arrays in γ-tubulin-depleted Drosophila spermatocytes

Organized microtubule arrays in γ-tubulin-depleted Drosophila spermatocytes Paula Sampaio1, Elena Rebollo2, Hanne Varmark2, Claudio E. Sunkel1,3 and Cayetano González2 1 Instituto de Biologia Molecular e Celular, Universidade do Porto, Rua do Campo Alegre 823, 4150 Porto, Portugal. 2 European Molecular Biology Laboratory, Cell Biology and Biophysics Programme, Meyerhofstrasse 1, 69117 Heidelberg, Germany. 3 Instituto de Ciências Biomédicas Abel Salazar, Largo do Prof. Abel Salazar N°2, 4000 Porto, Portugal. Correspondence: Dr. Cayetano González E-mail: [email protected] Abstract To assess the role of γ-tubulin in spindle assembly in vivo, we have followed meiosis progression by immunofluorescence and time-lapse video microscopy in -yTub23CPI mutant spermatocytes. We have found that centrosomes associate with large numbers of astral microtubules even though γ-tubulin is severely depleted; bipolar meiotic spindles are never assembled; and later in meiosis, the microtubules get organized into a conical structure that is never observed in wild-type cells. Several lines of evidence suggest that these cones may be related to wild-type central spindles. First, they are assembled midway through meiosis and elongate during anaphase. Second, they are constricted during late meiosis, giving rise to a pointed end similar to those that form in each half of the wild-type spindle midzone. Third, Klp3A and Polo, two markers of the wild-type central spindle are also found around the pointed end of the mutant cones. Finally, ectopic cytokinesis furrows are often formed at the distal end of the cone. Our results suggest that microtubule polymerization or stabilization from the centrosome may be possible in a γ-tubulinindependent manner in Drosophila spermatocytes. However, γ-tubulin seems to be essential for spindle assembly in these cells. Finally, our results show that at least part of the central spindle and constriction-ring assembly machinery can operate on microtubule bundles that are not organized as bipolar spindles. Results and Discussion Two γ-tubulin isoforms have been identified in Drosophila, γTub37C and γTub23C. γTub37C is largely restricted to the female germ line and early stages of embryogenesis [1–3]. The γTub23C isoform, on the other hand, is expressed in a variety of tissues in both sexes and is the only isoform present in testes [2, 4, 5]. Genetic, biochemical, and cytological data suggest that the -γTub23CPI mutant allele leads to a very severe loss of function of the γTub23C gene and thus to a severe reduction of the levels of γ-tubulin in mutant testes ([4]; also see the Supplementary material available with this article online). At the onset of meiosis, the two asters segregate normally in -γTub23CPI mutant spermatocytes, but soon afterwards, they collapse back together again (asterisks in Figure 1e). As meiosis proceeds, the normal figures found in control cells (Figure 1b,c) are never observed in -γTub23CPI spermatocytes. Instead, the microtubules form cone-shaped structures (Figure 1f,g), one per cell, with a pointed end (arrowheads) and the two asters fused or very close to each other at the base (asterisk). Such figures had not been reported before in other meiotic mutants in Drosophila [6, 7]. -γTub23CPI early spermatid cysts contain only 16 cells, and each spermatid has several nuclei of different sizes associated with a single large nebenkern, revealing the failure of the two meiotic divisions (Figure 1h). Around 75% of these cysts contain small anuclear and anastral cell fragments that carry a small nebenkern (arrows in Figure 1h), suggesting that some highly asymmetric cytokinesis takes place in these cells. To further investigate the organization of microtubules in -γTub23CPI meiotic spermatocytes, we stained fixed preparations with antibodies against α-tubulin. At prophase, a pair of welldefined astral arrays can be observed both in wild-type and -γTub23CPI mutant spermatocytes (Figure 2a,d). By the time of prometaphase/metaphase, a considerable number of microtubules are present in -γTub23CPI mutant spermatocytes (Figure 2e), but they only form a disordered mesh that bears no resemblance to the bipolar spindles found at this stage in control spermatocytes (Figure 2b). As meiosis proceeds, the microtubules found in the mutant spermatocytes are organized into cones (Figure 2f). With very rare exceptions, the chromosomes and centrosomes (not shown) are located at the base of the cone. The abundance of microtubules in -γTub23CPI spermatocytes is quite remarkable given the very low γ-tubulin levels observed in these cells. If the leaky function provided by this mutant allele is able to sustain the observed levels of microtubule polymerization, γ-tubulin must be present in vast excess in wild-type spermatocytes. Alternatively, microtubule polymerization and stabilization at the centrosome may not be completely dependent upon γ-tubulin in these cells. This interpretation is also supported by previous work carried out in Drosophila, Caenorhabditis elegans, and Schizosaccharomyces pombe [3, 4, 8–10]. However, despite the presence of a significant number of microtubules, bipolar spindles are never assembled in -γTub23CPI mutant spermatocytes. This is a surprising result given that spindle self-assembly through a centrosome-independent pathway has been reported in several experimental models, including Drosophila neuroblasts and spermatocytes, in vitro [11] and in vivo [12–16]. We do not understand the reason for this failure to organize the meiotic spindle. A simple explanation could be the presence of impaired, but partially functional, centrosomes that may still function as the major MTOCs (microtubule organizing centers) in these cells, thus overriding the organizing activity of the acentrosomal spindle assembly pathway. This hypothesis is not supported by the absence of organized spindles in -γTub23CPI asl double-mutant spermatocytes (data not shown). However, some centrosomal function may still remain in the double mutant. In this regard, it would be very interesting to investigate the phenotype of cnn -γTub23CPI double-mutant spermatocytes. An alternative interpretation could be that the dynamic properties of the microtubules present in -γTub23CPI cells are such that they no longer serve as good substrates for the organizing activities of the molecular motors involved in the acentrosomal spindle assembly pathway [17]. To get a better understanding of the process of assembly of the cones found in -γTub23CPI spermatocytes, we followed meiosis in these cells by time-lapse microscopy. At late prophase, the two mutant asters that had previously migrated apart start to approach each other, and the remains of the nuclear envelope become deformed (Figure 3f). Similar results have been observed in C. elegans embryos in which, following γTub inactivation by RNAi, the two asters that had initially segregated collapse back together again [10]. As meiosis proceeds in the mutant cells (Figure 3g), the condensed bivalents (black arrow) are clearly visible, and a large number of microtubules is revealed by their association with phase-dark membranes (pdm in Figure 3g). At this stage, most microtubules are sorted into two populations that emanate from each aster. Their distal ends can be seen moving along the cell membrane, becoming clustered at one side of the cell. Coinciding with the separation of the homolog univalents (Figure 3h, arrows), and thus at a stage similar to anaphase, the microtubule array found in the mutant starts to elongate and continues to do so until the new nuclear envelopes are formed (Figure 3i). Finally, this microtubule network is disassembled at the end of meiosis I (Figure 3j). Thus, the time of organization and elongation of the cone found in -γTub23CPI spermatocytes corresponds with the timing of organization and elongation of the wild-type central spindle. Given the similarities in the dynamics of assembly and elongation of mutant cones and wildtype central spindles, we decided to investigate the presence of common molecular components. To this end, we studied the localization of the wild-type central spindle markers Polo and Klp3A [18, 19] in the mutant cones. Immediately before nuclear envelope breakdown, the V-shaped centrosomes characteristic of these cells [20] are strongly labeled by a functional GFP-Polo fusion [21] that, at this stage, also very distinctively labels the nucleolus. Two pairs of centrosomes can be seen at nearly opposite sides of the nucleus, both in wild-type and in -γTub23CPI mutant spermatocytes (arrows in Figure 4a,a',e,e'). However, as meiosis proceeds in the mutant, the centrosomes get closer to each other and the spindle fails to become organized (Figure 4f,f'). When the microtubule cone is formed, the GFP-Polo fusion strongly accumulates at the pointed end (arrowhead in Figure 4g,g'), just as it does in the wild-type spindle midzone (Figure 4c,c'). It also labels the centrosomes in both wild-type and mutant cells. During cytokinesis, GFP-Polo colocalizes with the constriction ring in the wild-type (arrowheads in Figure 4d,d'). In those instances in which highly asymmetric cytokinesis take place in the mutant cells, the GFP-Polo fusion is localized near the furrow (Figure 4h,h'). Similar results were obtained with Klp3A. In wild-type spermatocytes, the kinesin-like protein Klp3A accumulates in the central spindle midzone (Figure 4i) [19]. In - γTub23CPI mutant spermatocytes, Klp3A is found at the pointed end of the cone (Figure 4j). Therefore, at least part of the machinery involved in the assembly of the constriction ring may also be used to organize the pointed end of the cone. Therefore, the cone-shaped microtubule arrays found in -γTub23CPI mutant spermatocytes bear some striking similarities to wild-type central spindles (schematically shown in Figure 4k,l). First, they are assembled midway through meiosis and elongate soon afterwards, just as wild-type central spindles do during anaphase B. Second, they are constricted during late meiosis, giving rise to a pointed end similar to those that form in each half of the wild-type central spindle during telophase. Third, Klp3A and Polo, two of the proteins localized in the wild-type central spindle midzone that are required for the recruitment of several components of the contractile ring [18, 19, 22, 23], are also found around the distal pointed end of the mutant cones. Finally, highly asymmetric cytokinesis can occur within -γTub23CPI mutant cells at the distal end of the cone. In summary, we have found that the depletion of γ-tubulin caused by the -γTub23CPI mutant allele does not prevent microtubule polymerization or stabilization, abolishes spindle assembly, and allows for the activity of the mechanisms that organize central spindles. Acknowledgements We are grateful to Mike Goldberg for providing antibodies and to the members of our laboratories for helpful discussions. P.S. is supported by a fellowship from the PRAXIS XXI program of Portugal. E.R. is a Marie Curie postdoctoral research fellow. Work in our laboratories is supported by the Training, Mobility, and Research program of the European Union and grants from the Ramón Areces Foundation to C.G. and the Fundação para a Ciência e Tecnologia of Portugal to C.E.S. References 1. Tavosanis G, Llamazares S, Goulielmos G, Gonzalez C: Essential role for gamma-tubulin in the acentriolar female meiotic spindle of Drosophila. EMBO J 1997, 16: 1809-1819. 2. Wilson PG, Zheng Y, Oakley CE, Oakley BR, Borisy GG, Fuller MT: Differential expression of two γ-tubulin isoforms during gametogenesis and development in Drosophila. Dev Biol 1997, 184: 207-221. 3. Llamazares S, Tavosanis G, Gonzalez C: Cytological characterisation of the mutant phenotypes produced during early embryogenesis by null and loss-of-function alleles of the γTub37C gene in Drosophila. J Cell Sci 1999, 112: 659-667. 4. Sunkel CE, Gomes R, Sampaio P, Perdigão J, González C: γ-tubulin is required for the structure and function of the microtubule organizing centre in Drosophila neuroblasts. EMBO J 1995, 14: 28-36. 5. Riparbelli MG, Whitfield WG, Dallai R, Callaini G: Assembly of the zygotic centrosome in the fertilized Drosophila egg. Mech Dev 1997, 65: 135-44. 6. Fuller MT: Spermatogenesis. In Development of Drosophila. Edited by Martinez-Arias A, Bate M. New York: Cold Spring Harbor Laboratory Press; 1993: 71-147. 7. Hackstein JH: Spermatogenesis in Drosophila. A genetic approach to cellular and subcellular differentiation. Eur J Cell Biol 1991, 56: 151-169. 8. Horio T, Uzawa S, Jung MK, Oakley BR, Tanakka K, Yanagida M: The fission yeast γtubulin is essential for mitosis and is localized at microtubule organizing centers. J Cell Sci 1991, 99: 693-700. 9. Vardy L, Toda T: The fission yeast γ-tubulin complex is required in G(1) phase and is a component of the spindle assembly checkpoint. EMBO J 2000, 19: 6098-6111. 10. Strome S, Powers J, Dunn M, Reese K, Malone CJ, White J, et al.: Spindle dynamics and the role of γ-tubulin in early Caenorhabditis elegans embryos. Mol Biol Cell 2001, 12: 1751-1764. 11. Heald R, Tournebize R, Blank T, Sandaltzopoulos R, Becker P, Hyman A, et al.: Selforganization of microtubules into bipolar spindles around artificial chromosomes in Xenopus egg extracts. Nature 1996, 382: 420-425. 12. Khodjakov A, Cole RW, Oakley BR, Rieder CL: Centrosome-independent mitotic spindle formation in vertebrates. Curr Biol 2000, 10: 59-67. 13. Bonaccorsi S, Giansanti MG, Gatti M: Spindle self-organization and cytokinesis during male meiosis in asterless mutants of Drosophila melanogaster. J Cell Biol 1998, 142: 751-761. 14. Megraw TL, Kao L, Kaufman TC: Zygotic development without functional mitotic centrosomes. Curr Biol 2001, 11: 116-120. 15. Giansanti MG, Gatti M, Bonaccorsi S: The role of centrosomes and astral microtubules during asymmetric division of Drosophila neuroblasts. Development 2001, 128: 11371145. 16. Raff JW: Centrosomes: central no more? Curr Biol 2001, 11: R159-R161. 17. Hyman AA, Karsenti E: Morphogenetic properties of microtubules and mitotic spindle assembly. Cell 1996, 84: 401-410. 18. Herrmann S, Amorim I, Sunkel CE: The POLO kinase is required at multiple stages during spermatogenesis in Drosophila melanogaster. Chromosoma 1998, 107: 440451. 19. Williams BC, Riedy MF, Williams EV, Gatti M, Goldberg ML: The Drosophila kinesin-like protein Klp3A is a midbody component required for central spindle assembly and initiation of cytokinesis. J Cell Biol 1995, 129: 709–723. 20. Tates AD: Cytodifferentiation during spermatogenesis in Drosophila melanogaster. PhD thesis, Department of Radiation Genetics, Transitorium voor Geneeskunde, Leiden, Netherlands. 1971. 21. Moutinho-Santos T, Sampaio P, Amorim I, Costa M, Sunkel CE: In vivo localisation of the mitotic POLO kinase shows a highly dynamic association with the mitotic apparatus during early embryogenesis in Drosophila. Biol Cell 1999, 91: 585-596. 22. Giansanti MG, Bonaccorsi S, Williams B, Williams EV, Santolamazza C, Goldberg ML, et al.: Cooperative interactions between the central spindle and the contractile ring during Drosophila cytokinesis. Genes Dev 1998, 12: 396-410. 23. Carmena M, Riparbelli MG, Minestrini G, Tavares AM, Adams R, Callaini G, et al.: Drosophila polo kinase is required for cytokinesis. J Cell Biol 1998, 143: 659-671. Figure 1 Phase-contrast images of unfixed (a–d) wild-type and (e–h) γTub23CPI mutant testis. (a) By late prophase, the phase-dark membranes that associate with the microtubules are well organized into asters (indicated by asterisks) located at opposite sides within wild-type spermatocytes. (e) Aster segregation also occurs in γTub23C$^{PI}$ mutant spermatocytes, but is not maintained, resulting in deformed nuclei with two indentations occupied by the two asters as they progressively approach each other. (b,c) The meiotic figures found in wild-type primary spermatocytes are bipolar, with two asters at each side of the spindle. The spindle microtubules are associated with phase-dark membranes (pdm). Different stages of meiosis from anaphase to telophase can be observed in this cyst. After telophase, the spindle is constricted in the middle (arrowheads), and one daughter nucleus (nu) is formed at each pole. (f,g) None of these figures are found in γTub23CPI mutant spermatocytes, in which the only organized arrays are cone shaped, with a pointed end (arrowheads) and a single aster (asterisk) at the base. One or more nuclei can often be seen associated with the aster. (d) Successful completion of the two meiotic divisions in the wild-type generates cysts of 64 early spermatids, each containing a phase-light nucleus (nu) and a phase-dark mitochondrial derivative (nebenkern; nb). (h) Early spermatid cysts in γTub23CPI males contain only 16 cells that have a large nebenkern associated with a few nuclei (nu) of different sizes. Several small cellular fragments containing only a small nebenkern (arrows) are also present in these mutant cysts. The scale bar represents 10 μm, except for in (c) and (g). Figure 2 Images of testis squashes showing microtubules (green) and DNA (blue) in (a–c) wild-type and (d–f) γTub23CPI mutant testis. (a) Well-organized asters can be observed during prophase in wild-type and (d) γTub23CPI mutant spermatocytes. (b) Bipolar spindles in the wild-type at prometaphase. (c) Telophase spindles in the wild-type showing the remains of the central spindle cleaved in two halves by the cytokinesis furrow (arrowheads). (e) A disorganized meshwork of microtubules surrounds the condensed chromosomes at this stage in the mutant. (f) The only organized microtubule arrays found in the mutant are cone shaped with a pointed end (arrowheads) and a broad base. A few nuclei (nu) of different sizes are always located at the base. The scale bar represents 10 μm. Figure 3 Time-lapse series of (a–e) wild-type and (f–j) γTub23CPI spermatocytes during meiosis I. (a) At prometaphase, in the wild-type, the two asters (asterisks) are diametrically opposed to each other near the outline of the remains of the nuclear envelope. (f) At this stage, the mutant asters are getting closer to each other and the nucleus is being deformed. (b) By late prometaphase, the bivalents (double arrow) are fully condensed, sitting on the middle of the spindle. The spindle is associated with numerous phase-dark membranes (pdm). (g) Microtubule polymerization and organization is very intense at this stage in the mutant. Essentially, the microtubules are being sorted into two populations that emanate from each aster toward the cell membrane. They get clustered through a circular movement (white long arrows) on one side of the cell, which accumulates most of the phase-dark membranes. (c) As the homolog chromosomes (arrows) segregate during anaphase, the wild-type central spindle starts to elongate and so does the (h) microtubule bundle in the mutant. (d,i) The elongation of both the wild-type spindle and the mutant cone continues until the new nuclei (nu) are formed. (e) During cytokinesis, the wild-type spindle and the associated membranes are disassembled, and the cytokinesis furrow (arrowheads) cleaves the cell into two secondary spermatocytes. (j) In the mutant, at the end of meiosis, the cone is also disassembled.