- Email: [email protected]
Drosophila oogenesis: Versatile spn doctors
Dispatch
R55
Drosophila oogenesis: Versatile spn doctors Jason Morris and Ruth Lehmann
Recent work on Drosophila oogenesis has uncovered connections between cell-cycle checkpoints and pattern formation. Genes of the spindle class, which encode double-strand break repair enzymes and RNA helicases, affect oocyte polarity and the decision whether to differentiate as an oocyte or a nurse cell. Address: Developmental Genetics Program, Skirball Institute, Howard Hughes Medical Institute, Department of Cell Biology, NYU School of Medicine, 540 First Avenue, New York, New York 10016, USA. Current Biology 1999, 9:R55–R58 http://biomednet.com/elecref/09609822009R0055 © Elsevier Science Ltd ISSN 0960-9822
Meiotic and mitotic ‘checkpoints’ control the timing of cell-cycle events by ensuring that earlier steps in the cell cycle are completed before later steps are initiated. Given the importance of regulated cell division in development, one might expect to find links between developmental events and cell-cycle checkpoints that would ensure the two processes are coordinated. Indeed, in many species a correlation has been observed between specific stages of the meiotic cell cycle and cellular events required for oocyte differentiation and maturation. The connection has now been put on a molecular footing, with the discovery that Drosophila ‘spindle’ genes are required for meiosis and have roles in establishing the embryonic axes and restricting oocyte fate to one cell in a germ-line cyst [1]. Oocyte determination and axis establishment
The main body axes of the Drosophila embryo are determined prior to fertilization by the localization of three mRNAs — bicoid, oskar and gurken — to discrete regions within the oocyte (Figure 1). These patterns of mRNA localization are established by a series of intercellular signaling events, involving communication between the oocyte and the surrounding somatic follicle cells [2]. These patterning events all occur while the oocyte is arrested in meiotic prophase I [3,4]. Before explaining the new work on spindle genes it would be useful to give a general outline of oogenesis in Drosophila. Drosophila oogenesis begins when a germ-line cystoblast undergoes four rounds of mitotic division to form an oogenic cyst of 16 germ cells [4]. These mitotic divisions occur with incomplete cytokinesis, so that each cell within the cyst is connected to one, two, three or four of its sister cells (Figure 1a). Initially, several of the germ cells enter meiosis, as shown by the presence of the characteristic synaptonemal complexes and recombination nodules [3]. Only one of these cells will maintain meiotic arrest and
develop as an oocyte; the other 15 cells undergo DNA endoreduplication and develop as nurse cells, which define the anterior end of the oocyte. The oocyte moves to a posterior position, relative to the nurse cells, and its chromosomes condense to form a compact ‘karyosome’ (Figure 1b). Each oogenic cyst is surrounded by somatic follicle cells (Figure 1b). Once the cell that will be the oocyte has been determined and properly positioned within the cyst, a complex set of intercellular signals are passed between the oocyte and the follicle cells, which lead to the establishment of the embryonic axes. The key early oocyte signal is a member of the transforming growth factor α (TGFα) family encoded by the gurken gene, the follicle-cell receptor for which is the Drosophila homolog of the epidermal growth factor (EGF) receptor. The position of the nucleus plays a crucial role in determining the polarity of the oocyte, and gurken RNA is localized close to the nucleus at this stage. The nucleus eventually moves from its posterior position to define the anterodorsal position, taking with it the associated gurken RNA; gurken signaling from this position is the initial trigger that determines dorsoventral polarity (Figure 1c). Gurken signaling is regulated at every step, including gurken RNA synthesis, localization and translation, and activation of the Gurken protein product [2] (Figure 1d). Although oocyte determination and positioning, as well as the establishment of polarity, occur when the oocyte is arrested in meiotic prophase, it has been unclear whether and how these different processes are coordinated. That the processes are coordinated is strongly implied by the identification of a rather large number of genes — more than 15 — that affect aspects of both meiosis and oocyte differentiation. Just how the coordination is achieved is beginning to be revealed by analyses of these genes and their products. Mutations in these genes were first identified because they cause a ‘ventralized’ phenotype similar to that of gurken mutants [1,5]. The egg shell of spindle mutant oocytes lacks dorsal structures and ventral pattern elements encroach into the dorsal half of the embryo. The genes defined by these mutations include those of the spindle class (spn-A, spn-B, spn-C, spn-D and spn-E), okra, aubergine and vasa (in what follows these will be collectively referred to as ‘spindle genes’). In spindle mutants, the translation and localization of gurken mRNA is defective, which probably accounts for their ventralized phenotypes [1,5–9]. But in contrast to the phenotype caused by strong gurken alleles, which is consistent and fully penetrant, the ventralized phenotype of even the
R56
Current Biology, Vol 9 No 2
Figure 1
(a) Restriction of oocyte fate
(b) Oocyte positioning
(c) Polarity establishment
(d) Oocyte maturation Anterior
Dorsal appendage Ventral
Anterior
Dorsal
Pro-oocytes Chorion Oocyte nucleus
Nurse cell Follicle cell
bicoid RNA gurken RNA
Oocyte nucleus
Oocyte nucleus
Oocyte Posterior
Current Biology
oskar RNA Posterior follicle
Posterior
Stages in Drosophila oogenesis. (a) The cystoblast divides to give 16 interconnected cells; two cells (blue and orange) arrest in prophase I, form synaptonemal complexes and initiate recombination. Only one of these cells (blue) will remain arrested and develop as an oocyte; the other fifteen cells will re-enter the cell cycle and develop as nurse cells. (b) Early in oogenesis, the oocyte is at the posterior end of the egg chamber and the oocyte chromosomes condense into the karyosome. Nurse cell nuclei become polyploid; gurken RNA and protein are enriched in the oocyte and Gurken signals to the follicle cells closest to the oocyte, at the posterior end of the cyst. Signaling from the
follicle cells back to the oocyte leads to the establishment of the anteroposterior axis via localization of bicoid (yellow) and oskar (pink) mRNA to the anterior and posterior poles of the oocyte, respectively. (c) The oocyte nucleus and gurken RNA (green) move from the posterior of the oocyte to an anterodorsal position during midoogenesis. Gurken signals to the overlying follicle cells to determine the position of the future dorsal side of the egg. (d) In the mature egg, the oocyte nucleus enters and arrests in metaphase I, and the nuclear membrane breaks down. The nurse cells and the follicle cells have degenerated by this stage.
strongest lack-of-function spindle alleles is variable. The spindle genes appear to be partially redundant — flies with mutations in two different spindle genes have a more severe phenotype than those with a mutation that completely abolishes the function of any one spindle gene [5].
gurken, they must also affect oocyte fate restriction via other, as yet unidentified effectors.
Gurken signaling cannot, however, be the only target of the spindle genes, because mutations in this class of genes disrupt not only oocyte axis formation, but also the oocyte cell-fate decision. Instead of compacting into a tiny, spherical karyosome, the oocyte chromosomes in spindle mutants have a diffuse, thread-like appearance. In some cases, the oocyte fails to move to the posterior end of the germ-line cyst [1,5,8,9]. Moreover, in the early cysts of some multiple spindle mutants, two oocytes are occasionally observed in a single cyst; this is most impressively seen in spn-C spn-B double mutants [5]. But gurken mutants do not exhibit these phenotypes, so while the spindle genes do regulate oocyte pattern formation through
A link between cell-cycle control and pattern formation
Several of the spindle genes have now been characterized at the molecular level [1,5,6,10], and the analysis has proven revealing and surprising. Several of the encoded spindle proteins are predicted to act in the cytoplasm to regulate gurken, whereas others are predicted to localize to the nucleus where they act in meiosis (Figure 2). The okra gene turns out to be identical to the previously cloned gene DmRad54, which encodes a homolog of the yeast protein Rad54. Rad54 is a DNA helicase with a wellestablished role in double-strand break repair [1,11]; in a similar vein, spn-B encodes a protein that shows significant sequence similarity to yeast Dmc1, a meiosis-specific double-strand break repair enzyme [12,13] (Figure 2). The spn-D gene has not yet been cloned, but its mutant phenotype is indistinguishable from that of spn-B.
Dispatch
In addition to causing gurken-dependent phenotypes, mutations in okra, spn-B or spn-D affect meiotic recombination. The loss of function of any of these genes causes an increased rate of chromosomal non-disjunction and a decreased frequency of recombination [1]. The okra gene is also implicated in mitotic DNA repair on the basis of the survival rates of okra mutant larvae exposed to the mutagen methyl methane sulfonate [1]; the survival rates of spn-B and spn-D mutant animals were unaffected by methyl methane sulfonate. As spn-B and spn-D mutants are phenotypically so similar, these two genes might have redundant functions in mitotic DNA repair, or they might act specifically in meiosis, as does the yeast protein Dmc1. Spn-E, vasa and aubergine belong to the class of spindle genes that might act in the cytoplasm rather than the nucleus. The Spn-E and Vasa proteins show significant sequence similarity to members of the RNA-dependent ATPase family defined by the motif DExH, which suggests that Spn-E and Vasa somehow interact with gurken mRNA in the cytoplasm and control either its localization or translation [6,8–10]. The oocyte karyosomes in spn-E and vasa mutants are thread-like and diffuse, however, as in mutants for the spindle genes implicated in doublestrand break repair. For both classes of spindle gene, therefore, mutations cause nuclear and cytoplasmic phenotypes; this might be indicative of feedback communication between the nuclear and cytoplasmic pathways [1], though direct roles for both sets of gene products in the nucleus cannot be ruled out. Double-strand break repair as a developmental checkpoint
In yeast, the presence of unrepaired double-strand breaks activates a meiotic checkpoint [12]. For example, dmc1 mutant cells, which are defective in double-strand break repair, arrest at meiotic prophase [13]. And mutations in genes required for initiation of recombination, such as SPO11, suppress the mutant dmc1 phenotype [12–14]. These observations suggest a model that may explain how double-strand break repair enzymes might regulate cell fate in Drosophila. In the cysts of spn-C spn-B double mutants, as in the wild type, several cells initiate meiosis; in the mutant cysts, however, more than one of these cells remain arrested in meiotic prophase I and are diverted from the nurse cell fate. This suggests that double-strand breaks have to be repaired for cells to follow the nurse cell fate. Double-strand break repair seems also to be required for the oocyte to proceed to the ‘karyosome’ stage. Support for this view comes from the observation that the oocyte nucleus in the mutants resembles the wild-type nucleus at an earlier stage (cited in [5]). This would be consistent with the mutants cells suffering a dmc1-like meiotic arrest [13]. Observations on a particular Drosophila cyclinE mutant further refine the model: cysts of such mutants contain one or two cells in addition to the oocyte
R57
Figure 2
Prophase Initiation of recombination, which requires: mei-W68 (SP011 homolog) Prophase arrest Double-strand break repair, which requires: Dmrad51 okra (RAD54 homolog) spn-B spn-D Metaphase arrest Crossover resolution, which requires: mei-9 (RAD1 homolog) Current Biology
During Drosophila oogenesis, the oocyte nucleus arrests initially at prophase, and recombination initiation (top) and double-strand break repair (middle) take place at this stage. The nucleus arrests again, almost three days later, in metaphase I, and crossovers are resolved by this stage (bottom). A number of genes (red) that play a part in doublestrand break repair have recently been implicated in a process of feedback signaling to the cytoplasm that regulates oocyte fate restriction and gurken localization (see text for details).
which, probably because of a defect in S-phase completion, do not undergo the DNA endoreduplication that is usually seen in nurse cells [15]. These cells are nevertheless not oocytes, as they fail to accumulate oocyte-specific markers [5]. Endoreplication may be required after the doublestrand break repair checkpoint has been passed to maintain nurse cell fate. Moreover, the double-strand break repair checkpoint may also have to be passed for activation of cytoplasmic factors such as Vasa and Spn-E, which are required for gurken RNA localization and translation. The discovery that genes identified by mutations that affect pattern formation may have a primary role in meiotic progression raises many questions and possibilities for further study. In order to determine the regulatory steps that synchronize progression through the cell cycle and development of the oocyte, it will be important to see which of the genes that affect Drosophila meiosis also affect pattern formation [16,17]. Useful insights might come from exploiting mei-W68, the recently identified Drosophila homolog of the yeast SPO11 gene [18]: mei-W68 mutants are recombination deficient [19] and should
R58
Current Biology, Vol 9 No 2
therefore not accumulate double-strand breaks or require double-strand break repair enzymes (Figure 2). One would then predict that in mei-W68 okra or mei-W68 spn-B double mutants, oogenesis would proceed beyond the okra/spn-B arrest point. These double mutants could also be used to test whether karyosome morphology and the correct regulation of gurken RNA localization and translation are a direct consequence of successfully passing the double-strand break repair checkpoint. Another useful genetic tool might be the Drosophila mei-9 gene, a homolog of yeast RAD1 that has been suggested to be required for resolution of Holliday junctions after double-strand break repair [20]. One would predict that mei-9 mutations would not affect oocyte fate and polarity, as the checkpoint is passed before Holliday junctions are resolved. In this case, okra and spn-B should be epistatic to mei-9 (Figure 2). In general, genes originally identified in screens for defects in either meiosis or oocyte patterning might turn out to play important roles in both processes. A convergence of cell-cycle and patterning research will be necessary to understand fully the coordination of cellcycle control and developmental events in oogenesis. References 1. Ghabrial A, Ray RP, Schüpbach T: okra and spindle-B encode components of the RAD52 DNA repair pathway and affect meiosis and patterning in Drosophila oogenesis. Genes Dev 1998, 12:2711-2723. 2. Ray RP, Schüpbach T: Intercellular signaling and the polarization of body axes during Drosophila oogenesis. Genes Dev 1996, 10:1711-1723. 3. Carpenter AT: Electron microscopy of meiosis in Drosophila melanogaster females. I. Structure, arrangement, and temporal change of the synaptonemal complex in wild-type. Chromosoma 1975, 51:157-182. 4. Spradling AC: Developmental genetics of oogenesis. In The Development of Drosophila melanogaster. Edited by Bate M, Arias AM. New York: Cold Spring Harbor Laboratory Press; 1993:1-70. 5. Gonzalez-Reyes A, Elliot H, St Johnston D: Oocyte determination and the origin of polarity in Drosophila: the role of the spindle genes. Development 1997, 124:4927-4937. 6. Gillespie DE, Berg CA: homeless is required for RNA localization in Drosophila oogenesis and encodes a new member of the DE-H family of RNA-dependent ATPases. Genes Dev 1995, 9:2495-2508. 7. Wilson JE, Connell JE, Macdonald PM: aubergine enhances oskar translation in the Drosophila ovary. Development 1996, 122:1631-1639. 8. Styhler S, Nakamura A, Swan A, Suter B, Lasko P: vasa is required for Gurken accumulation in the oocyte, and is involved in oocyte differentiation and germline cyst development. Development 1998, 125:1569-1578. 9. Tomancak P, Guichet A, Zavorszky P, Ephrussi A: Oocyte polarity depends on regulation of gurken by Vasa. Development 1998, 125:1723-1732. 10. Lasko PF, Ashburner M: The product of the Drosophila gene vasa is very similar to eukaryotic initiation factor 4A. Nature 1988, 335:611-617. 11. Koolstra R, Vrecken K, Zonneveld JB, de Jong A, Eeken JC, Osgood CJ, Buerstedde JM, Lohman PH, Pastink A: The Drosophila melanogaster RAD54 homolog, DmRad54, is involved in repair of radiation damage and recombination. Mol Cell Biol 1997, 17:6097-6104. 12. Roeder GS: Meiotic chromosomes: it takes two to tango. Genes Dev 1997, 11:2600-2621. 13. Bishop D, Park D, Xu L, Kleckner N: DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex formation, and cell cycle progression. Cell 1992, 69:439-456.
14. McKee AH, Kleckner N: Mutations in Saccharomyces cerevisiae that block meiotic prophase chromosome metabolism and confer cell cycle arrest at pachytene identify two new meiosis-specific genes SAE1 and SAE3. Genetics 1997, 146:817-834. 15. Lilly MA, Spradling AC: The Drosophila endocycle is controlled by cyclin E and lacks a checkpoint ensuring S-phase completion. Genes Dev 1996, 10:2514-2526. 16. Baker BS, Carpenter ATC, Esposito MS, Esposito RE, Sandler L: The genetics of meiosis. Annu Rev Genet 1976, 10:53-134. 17. Sekelsky JJ, Burtis KC, Hawley RS: Damage control: the pleiotropy of DNA repair genes in Drosophila melanogaster. Genetics 1998, 148:1587-1598. 18. McKim KS, Hayashi-Hagihara A: mei-W68 in Drosophila melanogaster encodes a Spo11 homolog: evidence that the mechanism for initiating meiotic recombination is conserved. Genes Dev 1998, 12:2932-2942. 19. McKim KS, Green-Marroquin BL, Sekelsky JJ, Chin G, Steinberg C, Khodosh R, Hawley RS: Meiotic synapsis in the absence of recombination. Science 1990, 279:876-878. 20. Sekelsky JJ, McKim KS, Chin GM, Hawley RS: The Drosophila meiotic recombination gene mei-9 encodes a homologue of the yeast excision repair protein Rad1. Genetics 1995, 141:619-627.
R55
Drosophila oogenesis: Versatile spn doctors Jason Morris and Ruth Lehmann
Recent work on Drosophila oogenesis has uncovered connections between cell-cycle checkpoints and pattern formation. Genes of the spindle class, which encode double-strand break repair enzymes and RNA helicases, affect oocyte polarity and the decision whether to differentiate as an oocyte or a nurse cell. Address: Developmental Genetics Program, Skirball Institute, Howard Hughes Medical Institute, Department of Cell Biology, NYU School of Medicine, 540 First Avenue, New York, New York 10016, USA. Current Biology 1999, 9:R55–R58 http://biomednet.com/elecref/09609822009R0055 © Elsevier Science Ltd ISSN 0960-9822
Meiotic and mitotic ‘checkpoints’ control the timing of cell-cycle events by ensuring that earlier steps in the cell cycle are completed before later steps are initiated. Given the importance of regulated cell division in development, one might expect to find links between developmental events and cell-cycle checkpoints that would ensure the two processes are coordinated. Indeed, in many species a correlation has been observed between specific stages of the meiotic cell cycle and cellular events required for oocyte differentiation and maturation. The connection has now been put on a molecular footing, with the discovery that Drosophila ‘spindle’ genes are required for meiosis and have roles in establishing the embryonic axes and restricting oocyte fate to one cell in a germ-line cyst [1]. Oocyte determination and axis establishment
The main body axes of the Drosophila embryo are determined prior to fertilization by the localization of three mRNAs — bicoid, oskar and gurken — to discrete regions within the oocyte (Figure 1). These patterns of mRNA localization are established by a series of intercellular signaling events, involving communication between the oocyte and the surrounding somatic follicle cells [2]. These patterning events all occur while the oocyte is arrested in meiotic prophase I [3,4]. Before explaining the new work on spindle genes it would be useful to give a general outline of oogenesis in Drosophila. Drosophila oogenesis begins when a germ-line cystoblast undergoes four rounds of mitotic division to form an oogenic cyst of 16 germ cells [4]. These mitotic divisions occur with incomplete cytokinesis, so that each cell within the cyst is connected to one, two, three or four of its sister cells (Figure 1a). Initially, several of the germ cells enter meiosis, as shown by the presence of the characteristic synaptonemal complexes and recombination nodules [3]. Only one of these cells will maintain meiotic arrest and
develop as an oocyte; the other 15 cells undergo DNA endoreduplication and develop as nurse cells, which define the anterior end of the oocyte. The oocyte moves to a posterior position, relative to the nurse cells, and its chromosomes condense to form a compact ‘karyosome’ (Figure 1b). Each oogenic cyst is surrounded by somatic follicle cells (Figure 1b). Once the cell that will be the oocyte has been determined and properly positioned within the cyst, a complex set of intercellular signals are passed between the oocyte and the follicle cells, which lead to the establishment of the embryonic axes. The key early oocyte signal is a member of the transforming growth factor α (TGFα) family encoded by the gurken gene, the follicle-cell receptor for which is the Drosophila homolog of the epidermal growth factor (EGF) receptor. The position of the nucleus plays a crucial role in determining the polarity of the oocyte, and gurken RNA is localized close to the nucleus at this stage. The nucleus eventually moves from its posterior position to define the anterodorsal position, taking with it the associated gurken RNA; gurken signaling from this position is the initial trigger that determines dorsoventral polarity (Figure 1c). Gurken signaling is regulated at every step, including gurken RNA synthesis, localization and translation, and activation of the Gurken protein product [2] (Figure 1d). Although oocyte determination and positioning, as well as the establishment of polarity, occur when the oocyte is arrested in meiotic prophase, it has been unclear whether and how these different processes are coordinated. That the processes are coordinated is strongly implied by the identification of a rather large number of genes — more than 15 — that affect aspects of both meiosis and oocyte differentiation. Just how the coordination is achieved is beginning to be revealed by analyses of these genes and their products. Mutations in these genes were first identified because they cause a ‘ventralized’ phenotype similar to that of gurken mutants [1,5]. The egg shell of spindle mutant oocytes lacks dorsal structures and ventral pattern elements encroach into the dorsal half of the embryo. The genes defined by these mutations include those of the spindle class (spn-A, spn-B, spn-C, spn-D and spn-E), okra, aubergine and vasa (in what follows these will be collectively referred to as ‘spindle genes’). In spindle mutants, the translation and localization of gurken mRNA is defective, which probably accounts for their ventralized phenotypes [1,5–9]. But in contrast to the phenotype caused by strong gurken alleles, which is consistent and fully penetrant, the ventralized phenotype of even the
R56
Current Biology, Vol 9 No 2
Figure 1
(a) Restriction of oocyte fate
(b) Oocyte positioning
(c) Polarity establishment
(d) Oocyte maturation Anterior
Dorsal appendage Ventral
Anterior
Dorsal
Pro-oocytes Chorion Oocyte nucleus
Nurse cell Follicle cell
bicoid RNA gurken RNA
Oocyte nucleus
Oocyte nucleus
Oocyte Posterior
Current Biology
oskar RNA Posterior follicle
Posterior
Stages in Drosophila oogenesis. (a) The cystoblast divides to give 16 interconnected cells; two cells (blue and orange) arrest in prophase I, form synaptonemal complexes and initiate recombination. Only one of these cells (blue) will remain arrested and develop as an oocyte; the other fifteen cells will re-enter the cell cycle and develop as nurse cells. (b) Early in oogenesis, the oocyte is at the posterior end of the egg chamber and the oocyte chromosomes condense into the karyosome. Nurse cell nuclei become polyploid; gurken RNA and protein are enriched in the oocyte and Gurken signals to the follicle cells closest to the oocyte, at the posterior end of the cyst. Signaling from the
follicle cells back to the oocyte leads to the establishment of the anteroposterior axis via localization of bicoid (yellow) and oskar (pink) mRNA to the anterior and posterior poles of the oocyte, respectively. (c) The oocyte nucleus and gurken RNA (green) move from the posterior of the oocyte to an anterodorsal position during midoogenesis. Gurken signals to the overlying follicle cells to determine the position of the future dorsal side of the egg. (d) In the mature egg, the oocyte nucleus enters and arrests in metaphase I, and the nuclear membrane breaks down. The nurse cells and the follicle cells have degenerated by this stage.
strongest lack-of-function spindle alleles is variable. The spindle genes appear to be partially redundant — flies with mutations in two different spindle genes have a more severe phenotype than those with a mutation that completely abolishes the function of any one spindle gene [5].
gurken, they must also affect oocyte fate restriction via other, as yet unidentified effectors.
Gurken signaling cannot, however, be the only target of the spindle genes, because mutations in this class of genes disrupt not only oocyte axis formation, but also the oocyte cell-fate decision. Instead of compacting into a tiny, spherical karyosome, the oocyte chromosomes in spindle mutants have a diffuse, thread-like appearance. In some cases, the oocyte fails to move to the posterior end of the germ-line cyst [1,5,8,9]. Moreover, in the early cysts of some multiple spindle mutants, two oocytes are occasionally observed in a single cyst; this is most impressively seen in spn-C spn-B double mutants [5]. But gurken mutants do not exhibit these phenotypes, so while the spindle genes do regulate oocyte pattern formation through
A link between cell-cycle control and pattern formation
Several of the spindle genes have now been characterized at the molecular level [1,5,6,10], and the analysis has proven revealing and surprising. Several of the encoded spindle proteins are predicted to act in the cytoplasm to regulate gurken, whereas others are predicted to localize to the nucleus where they act in meiosis (Figure 2). The okra gene turns out to be identical to the previously cloned gene DmRad54, which encodes a homolog of the yeast protein Rad54. Rad54 is a DNA helicase with a wellestablished role in double-strand break repair [1,11]; in a similar vein, spn-B encodes a protein that shows significant sequence similarity to yeast Dmc1, a meiosis-specific double-strand break repair enzyme [12,13] (Figure 2). The spn-D gene has not yet been cloned, but its mutant phenotype is indistinguishable from that of spn-B.
Dispatch
In addition to causing gurken-dependent phenotypes, mutations in okra, spn-B or spn-D affect meiotic recombination. The loss of function of any of these genes causes an increased rate of chromosomal non-disjunction and a decreased frequency of recombination [1]. The okra gene is also implicated in mitotic DNA repair on the basis of the survival rates of okra mutant larvae exposed to the mutagen methyl methane sulfonate [1]; the survival rates of spn-B and spn-D mutant animals were unaffected by methyl methane sulfonate. As spn-B and spn-D mutants are phenotypically so similar, these two genes might have redundant functions in mitotic DNA repair, or they might act specifically in meiosis, as does the yeast protein Dmc1. Spn-E, vasa and aubergine belong to the class of spindle genes that might act in the cytoplasm rather than the nucleus. The Spn-E and Vasa proteins show significant sequence similarity to members of the RNA-dependent ATPase family defined by the motif DExH, which suggests that Spn-E and Vasa somehow interact with gurken mRNA in the cytoplasm and control either its localization or translation [6,8–10]. The oocyte karyosomes in spn-E and vasa mutants are thread-like and diffuse, however, as in mutants for the spindle genes implicated in doublestrand break repair. For both classes of spindle gene, therefore, mutations cause nuclear and cytoplasmic phenotypes; this might be indicative of feedback communication between the nuclear and cytoplasmic pathways [1], though direct roles for both sets of gene products in the nucleus cannot be ruled out. Double-strand break repair as a developmental checkpoint
In yeast, the presence of unrepaired double-strand breaks activates a meiotic checkpoint [12]. For example, dmc1 mutant cells, which are defective in double-strand break repair, arrest at meiotic prophase [13]. And mutations in genes required for initiation of recombination, such as SPO11, suppress the mutant dmc1 phenotype [12–14]. These observations suggest a model that may explain how double-strand break repair enzymes might regulate cell fate in Drosophila. In the cysts of spn-C spn-B double mutants, as in the wild type, several cells initiate meiosis; in the mutant cysts, however, more than one of these cells remain arrested in meiotic prophase I and are diverted from the nurse cell fate. This suggests that double-strand breaks have to be repaired for cells to follow the nurse cell fate. Double-strand break repair seems also to be required for the oocyte to proceed to the ‘karyosome’ stage. Support for this view comes from the observation that the oocyte nucleus in the mutants resembles the wild-type nucleus at an earlier stage (cited in [5]). This would be consistent with the mutants cells suffering a dmc1-like meiotic arrest [13]. Observations on a particular Drosophila cyclinE mutant further refine the model: cysts of such mutants contain one or two cells in addition to the oocyte
R57
Figure 2
Prophase Initiation of recombination, which requires: mei-W68 (SP011 homolog) Prophase arrest Double-strand break repair, which requires: Dmrad51 okra (RAD54 homolog) spn-B spn-D Metaphase arrest Crossover resolution, which requires: mei-9 (RAD1 homolog) Current Biology
During Drosophila oogenesis, the oocyte nucleus arrests initially at prophase, and recombination initiation (top) and double-strand break repair (middle) take place at this stage. The nucleus arrests again, almost three days later, in metaphase I, and crossovers are resolved by this stage (bottom). A number of genes (red) that play a part in doublestrand break repair have recently been implicated in a process of feedback signaling to the cytoplasm that regulates oocyte fate restriction and gurken localization (see text for details).
which, probably because of a defect in S-phase completion, do not undergo the DNA endoreduplication that is usually seen in nurse cells [15]. These cells are nevertheless not oocytes, as they fail to accumulate oocyte-specific markers [5]. Endoreplication may be required after the doublestrand break repair checkpoint has been passed to maintain nurse cell fate. Moreover, the double-strand break repair checkpoint may also have to be passed for activation of cytoplasmic factors such as Vasa and Spn-E, which are required for gurken RNA localization and translation. The discovery that genes identified by mutations that affect pattern formation may have a primary role in meiotic progression raises many questions and possibilities for further study. In order to determine the regulatory steps that synchronize progression through the cell cycle and development of the oocyte, it will be important to see which of the genes that affect Drosophila meiosis also affect pattern formation [16,17]. Useful insights might come from exploiting mei-W68, the recently identified Drosophila homolog of the yeast SPO11 gene [18]: mei-W68 mutants are recombination deficient [19] and should
R58
Current Biology, Vol 9 No 2
therefore not accumulate double-strand breaks or require double-strand break repair enzymes (Figure 2). One would then predict that in mei-W68 okra or mei-W68 spn-B double mutants, oogenesis would proceed beyond the okra/spn-B arrest point. These double mutants could also be used to test whether karyosome morphology and the correct regulation of gurken RNA localization and translation are a direct consequence of successfully passing the double-strand break repair checkpoint. Another useful genetic tool might be the Drosophila mei-9 gene, a homolog of yeast RAD1 that has been suggested to be required for resolution of Holliday junctions after double-strand break repair [20]. One would predict that mei-9 mutations would not affect oocyte fate and polarity, as the checkpoint is passed before Holliday junctions are resolved. In this case, okra and spn-B should be epistatic to mei-9 (Figure 2). In general, genes originally identified in screens for defects in either meiosis or oocyte patterning might turn out to play important roles in both processes. A convergence of cell-cycle and patterning research will be necessary to understand fully the coordination of cellcycle control and developmental events in oogenesis. References 1. Ghabrial A, Ray RP, Schüpbach T: okra and spindle-B encode components of the RAD52 DNA repair pathway and affect meiosis and patterning in Drosophila oogenesis. Genes Dev 1998, 12:2711-2723. 2. Ray RP, Schüpbach T: Intercellular signaling and the polarization of body axes during Drosophila oogenesis. Genes Dev 1996, 10:1711-1723. 3. Carpenter AT: Electron microscopy of meiosis in Drosophila melanogaster females. I. Structure, arrangement, and temporal change of the synaptonemal complex in wild-type. Chromosoma 1975, 51:157-182. 4. Spradling AC: Developmental genetics of oogenesis. In The Development of Drosophila melanogaster. Edited by Bate M, Arias AM. New York: Cold Spring Harbor Laboratory Press; 1993:1-70. 5. Gonzalez-Reyes A, Elliot H, St Johnston D: Oocyte determination and the origin of polarity in Drosophila: the role of the spindle genes. Development 1997, 124:4927-4937. 6. Gillespie DE, Berg CA: homeless is required for RNA localization in Drosophila oogenesis and encodes a new member of the DE-H family of RNA-dependent ATPases. Genes Dev 1995, 9:2495-2508. 7. Wilson JE, Connell JE, Macdonald PM: aubergine enhances oskar translation in the Drosophila ovary. Development 1996, 122:1631-1639. 8. Styhler S, Nakamura A, Swan A, Suter B, Lasko P: vasa is required for Gurken accumulation in the oocyte, and is involved in oocyte differentiation and germline cyst development. Development 1998, 125:1569-1578. 9. Tomancak P, Guichet A, Zavorszky P, Ephrussi A: Oocyte polarity depends on regulation of gurken by Vasa. Development 1998, 125:1723-1732. 10. Lasko PF, Ashburner M: The product of the Drosophila gene vasa is very similar to eukaryotic initiation factor 4A. Nature 1988, 335:611-617. 11. Koolstra R, Vrecken K, Zonneveld JB, de Jong A, Eeken JC, Osgood CJ, Buerstedde JM, Lohman PH, Pastink A: The Drosophila melanogaster RAD54 homolog, DmRad54, is involved in repair of radiation damage and recombination. Mol Cell Biol 1997, 17:6097-6104. 12. Roeder GS: Meiotic chromosomes: it takes two to tango. Genes Dev 1997, 11:2600-2621. 13. Bishop D, Park D, Xu L, Kleckner N: DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex formation, and cell cycle progression. Cell 1992, 69:439-456.
14. McKee AH, Kleckner N: Mutations in Saccharomyces cerevisiae that block meiotic prophase chromosome metabolism and confer cell cycle arrest at pachytene identify two new meiosis-specific genes SAE1 and SAE3. Genetics 1997, 146:817-834. 15. Lilly MA, Spradling AC: The Drosophila endocycle is controlled by cyclin E and lacks a checkpoint ensuring S-phase completion. Genes Dev 1996, 10:2514-2526. 16. Baker BS, Carpenter ATC, Esposito MS, Esposito RE, Sandler L: The genetics of meiosis. Annu Rev Genet 1976, 10:53-134. 17. Sekelsky JJ, Burtis KC, Hawley RS: Damage control: the pleiotropy of DNA repair genes in Drosophila melanogaster. Genetics 1998, 148:1587-1598. 18. McKim KS, Hayashi-Hagihara A: mei-W68 in Drosophila melanogaster encodes a Spo11 homolog: evidence that the mechanism for initiating meiotic recombination is conserved. Genes Dev 1998, 12:2932-2942. 19. McKim KS, Green-Marroquin BL, Sekelsky JJ, Chin G, Steinberg C, Khodosh R, Hawley RS: Meiotic synapsis in the absence of recombination. Science 1990, 279:876-878. 20. Sekelsky JJ, McKim KS, Chin GM, Hawley RS: The Drosophila meiotic recombination gene mei-9 encodes a homologue of the yeast excision repair protein Rad1. Genetics 1995, 141:619-627.