mei-P22 Encodes a Chromosome-Associated Protein Required for the Initiation of Meiotic Recombination in Drosophila melanogaster

Corresponding author: Kim S. McKim, Rutgers University, 190 Frelinghuysen Rd., Piscataway, NJ 08854., mckim{at}rci.rutgers.edu (E-mail)

Communicating editor: S. HENIKOFF

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Double-strand breaks (DSB) initiate meiotic recombination in a variety of organisms. Here we present genetic evidence that the mei-P22 gene is required for the induction of DSBs during meiotic prophase in Drosophila females. Strong mei-P22 mutations eliminate meiotic crossing over and suppress the sterility of DSB repair-defective mutants. Interestingly, crossing over in mei-P22 mutants can be restored to almost 50% of wild-type by X irradiation. In addition, an antibody-based assay was used to demonstrate that DSBs are not formed in mei-P22 mutants. This array of phenotypes is identical to that of mei-W68 mutants; mei-W68 encodes the Drosophila Spo11 homolog that is proposed to be an enzyme required for DSB formation. Consistent with a direct role in DSB formation, mei-P22 encodes a basic 35.7-kD protein, which, when examined by immunofluorescence, localizes to foci on meiotic chromosomes. MEI-P22 foci appear transiently in early meiotic prophase, which is when meiotic recombination is believed to initiate. By using an antibody to C(3)G as a marker for synaptonemal complex (SC) formation, we observed that SC is present before MEI-P22 associates with the chromosomes, thus providing direct evidence that the development of SC precedes the initiation of meiotic recombination. Similarly, we found that MEI-P22 foci did not appear in a c(3)G mutant in which SC does not form, suggesting that DSB formation is dependent on SC formation in Drosophila. We propose that MEI-P22 interacts with meiosis-specific chromosome proteins to facilitate DSB creation by MEI-W68.

MEIOTIC crossovers mature into chiasmata and thereby direct the segregation of homologs at the first meiotic division. It is now commonly accepted that, in many organisms, meiotic recombination is initiated with a double-strand break (DSB), which is then repaired by using the homolog as a template. This process results in either a noncrossover (simple gene conversion) or a crossover (SZOSTAK et al. 1983 ; LICHTEN 2001 ). In support of this model, DSBs (reviewed in GOLDMAN and LICHTEN 1996 ) and double Holliday junctions (SCHWACHA and KLECKNER 1995 ) have been detected during meiotic prophase in Saccharomyces cerevisiae. While it is not known how DSB sites are chosen, several genetic and epigenetic factors are known to have important roles. For example, at least 10 known gene products are required for DSB formation in S. cerevisiae meiosis, and 7 of these are presumed to be meiosis specific (reviewed in ROEDER 1997 ). Some of these proteins could have a role in choosing the DSB site, but one of these proteins, Spo11, is believed to be the enzyme that generates the DSB break. After the DSB has been made, a combination of meiosis-specific and general DSB repair proteins are required for the production of crossovers or gene conversions.

Few homologs of these DSB-inducing genes have been identified in higher eukaryotes. The significant exception to this observation is Spo11, for which homologs have been identified in several species. Strong mutations in the Drosophila Spo11 homolog, mei-W68, eliminate all meiotic recombination, both simple gene conversion and crossing over, from which it has been inferred that the wild-type allele is responsible for making meiosis-specific DSBs (MCKIM et al. 1998 ). In yeast and mice, DSBs are needed not only for meiotic recombination but also for formation of the synaptonemal complex (SC), a proteinaceous structure that forms between aligned homologs in meiotic prophase (reviewed in ROEDER 1997 ; LICHTEN 2001 ). In Drosophila and Caenorhabditis elegans, however, SC formation is normal in strong mutants of their Spo11 homologs, implying that in some organisms meiotic recombination is not required for synapsis (DERNBURG et al. 1998 ; MCKIM et al. 1998 ). Further studies and direct observations of DSB formation in higher eukaryotes have technical limitations due to the rarity of DSBs at any given site. Therefore, experiments to determine the time course of key events in a wild-type meiosis such as SC formation, in DSB formation, and in DSB repair have not been possible.

mei-P22 mutants have an identical phenotype to mei-W68 in that they eliminate meiotic recombination but form normal SC (MCKIM et al. 1998 ). Here we present genetic evidence that mei-P22 is required for the initiation of DSBs. Consistent with this function, we have found that MEI-P22 localizes to foci on meiotic chromosomes, suggesting that MEI-P22 may have a close relationship with DSB formation. Observing the expression pattern of MEI-P22 provides an opportunity to study the regulation of DSB formation in a higher eukaryote. We have found that the binding of MEI-P22 to chromosomes is restricted to a brief period of meiotic prophase and occurs after SC formation. In support of the conclusion that DSB formation is dependent on SC formation, we have found that MEI-P22 localization requires the SC component C(3)G.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Genetic techniques:
All fly crosses to measure X chromosome nondisjunction or crossing over were raised at 25°. The original mei-P22 mutation, mei-P22¹, was recovered in a screen for meiotic mutants caused by insertion of a w⁺ P element (P{lacW}; SEKELSKY et al. 1999 ). X chromosome nondisjunction was assayed by crossing females to C(1;Y)1, v f B/0; C(4)RM, ey^R ci males. From this cross, progeny resulting from normal X chromosome disjunction were B/+ females and wild-type males. Nondisjunction of the X chromosome resulted in Bar males and wild-type females. Crossing over on the second chromosome was assayed by generating females homozygous for a mei-P22 mutant and heterozygous for an al dp b pr cn chromosome. Crossing over on the third chromosome was assayed by recombining a mei-P22 mutation onto a th st cu e ca chromosome and then backcrossing to make the mei-P22 mutation homozygous. To assay for radiation-induced crossing over, 2- to 5-day-old virgin females of the genotype mei-P22¹⁰³/mei-P22¹⁰³ th st cu e ca were exposed to 4000 rad of X rays at a dose of 114 rad/min.

Screen for new mei-P22 alleles:
Males were fed 25 mM ethyl methanesulfonate in 1% sucrose for 24 hr and then transferred to yeasted bottles for 1 day to recover. The cross was conducted as follows:

The F₂ cross was brothers to sisters to avoid the requirement for virgins. Normal progeny from this cross were yellow females and wild-type males, whereas the presence of a mutant on the +* chromosome was indicated by yellow males and wild-type females. If these were observed, the cross was set up again for confirmation and a stock made by crossing the white-eyed males and females (y w/y+Y; +*/TM3 x y w/y w; +*/TM3). We could select against the mei-P22¹ chromosome because it is associated with a P-element insertion carrying the mini-white marker gene.

Molecular analysis of mei-P22 expression and mutations:
Genomic sequences flanking the P-element insertion site in mei-P22¹ were isolated and sequenced. An EcoRI fragment containing 1.3 kb of DNA flanking the 3' end of the P{lacW} insertion was isolated by plasmid rescue (ASHBURNER 1989 ). The P-element was inserted between two open reading frames (ORFs) and could have affected either or both transcripts. The nondisjunction phenotype was rescued by constructs that contained only one of the ORFs. Using the rescue fragment as a probe to screen a Drosophila -DASHII genomic library (FINELLI et al. 1994 ), we identified three overlapping clones encompassing the mei-P22 locus. Subclones from one of these were sequenced (GenBank accession no. AF199369). Our sequence differs from that of the genome project by 13 amino acids, perhaps because of different backgrounds (MYERS et al. 2000 and Fig 3). For in situ hybridization, digoxygenin-labeled RNA probes were made from a linearized mei-P22 clone (pNH4) using the Boehringer Mannheim (Indianapolis) RNA labeling kit and hybridized using a procedure modified from the protocol for embryos (TAUTZ and PFEIFLE 1989 ).

View larger version (23K): In this window In a new window Download PPT slide   Figure 1. Crossing over is increased and nondisjunction decreased by irradiating mei-P22103 females. The frequency of crossing over between st and ca (control distance = 50.9) is shown with shaded (irradiated) or solid (not irradiated) bars. Nondisjunction frequencies are shown with shaded (irradiated) or solid (not irradiated) lines. The high nondisjunction frequency for the irradiated 10- to 12-day brood may not be significant due to a small sample size.

View larger version (47K): In this window In a new window Download PPT slide   Figure 2. mei-P22 is epistatic to spnB mutants. (A) spnBBU mutants are sterile after the first 2–3 days of mating (blue). This effect is suppressed by mutations that reduce DSBs such as mei-W681 and mei-P22N1. (B) Unrepaired DSBs are present in a spnBBU mutant but not a mei-P22N1 spnBBU double mutant. In each image the oocyte was recognized due to the enrichment of the ORB protein in the cytoplasm (green). Region 3 and stage 2 oocytes are in a relatively late stage of meiotic prophase (see also Fig 4). In wild type the -H2AX foci are usually absent because the DSBs have been repaired. In a spnB mutant, however, the -H2AX antibody stains numerous foci because the DSBs are not repaired (D. SHERIZEN and K. MCKIM, unpublished results). In the mei-P22N1 spnBBU double mutant no foci were visible, suggesting that the DSBs were never formed.

View larger version (24K): In this window In a new window Download PPT slide   Figure 3. Sequence of MEI-P22 including the changes in the EMS mutants. The sequence shown corresponds to our sequence from the -DASHII library. In the published genome sequence, there is an insertion (arrowhead) encoding 13 amino acids (EDSSCAT RRRSLS; MYERS et al. 2000 ). The positions of EMS-induced mutations are shown in boldface type, the predicted BP-NLS is underlined, and the region predicted to form an -helix is shown in italic.

To sequence our new alleles, some of which were on chromosomes with extraneous lethal mutations, genomic DNA was made from females homozygous for a mei-P22 allele or that were heterozygous to the original mei-P22¹ allele. PCR was used to amplify the entire mei-P22 gene. The primers flanked the site of the P-element insertion in mei-P22¹; therefore, in heterozygotes only the DNA from the EMS allele was amplified. Amplified fragments were cloned using the Perfectly Blunt Cloning system (Novagen) and prepared for sequencing by alkaline lysis minipreps and polyethylene glycol precipitation. In some cases, the PCR products were directly sequenced following PCR and isolation from an agarose gel. Sequencing was performed by the University of Medicine and Dentistry of New Jersey sequencing facility and analyzed using the Wisconsin Package Version 9.1 (Genetics Computer Group, Madison, WI). Sequences from different mutant DNAs were compared to identify the nucleotide changes.

Construction and analysis of the mei-P22 epitope-tag fusion gene:
The vector containing the triple hemagglutinin (HA) epitope was constructed by cloning double-stranded oligonucleotides into pBluescript. To make an amino-terminal fusion of mei-P22 to the epitope tag, the entire coding region, including 1 kb after the stop codon, was amplified by PCR. A ClaI site at the amino terminus (introduced in the PCR reaction) and a SacII site were used to clone the fragment into the pBluescript vector containing the epitope tag. The hsp83 promoter was inserted as a 900-bp KpnI/SalI fragment into the epitope tag/mei-P22 construct cut with KpnI and XhoI. The whole construct was then transferred to the transformation vector pCaSper 4 using the KpnI and SacII sites. We chose to use the hsp83 promoter because in previous experience it has reliably driven gene expression in the germline. When the hsp83 promoter was used to drive expression of mei-218, we found that every transgenic rescued the mei-218 mutant phenotype and produced detectable amounts of protein in the germline when examined by immunofluorescence and Western blot (MANHEIM et al. 2002 ).

With an antibody to the HA tag, we detected approximately equal amounts of MEI-P22^3XHA in each of the hsp83::mei-P22^3XHA transgenics by Western blot of ovarian protein (data not shown). However, in many of these lines, the protein was not detectable in the germline by immunofluorescence and did not rescue the mei-P22 mutant phenotype. It appeared that certain transgenic lines expressed MEI-P22^3XHA in the germline although for unknown reasons it was not visible in the nucleus. An interesting explanation for this finding is based on the observation that the hsp83::mei-218 transgenic lines differed in the timing of the earliest protein expression. Some lines expressed MEI-218 in early region 2a, whereas in others the protein was not observed until region 2b. Of the transgenics where MEI-P22 was observed, staining was limited to region 2a. Unlike mei-218, the timing of MEI-P22 expression appears to be critical for its function: If not present early in the development of the 16-cell cysts, it may be unable to aggregate at potential DSB sites into complexes that are visible by fluorescence. These considerations suggest that the transgenic lines that failed to produce foci expressed MEI-P22 too late in pachytene.

Cytology:
For immunolocalization experiments, virgin females were aged for 16 hr at room temperature, dissected, and fixed using the "Buffer A" protocol (BELMONT et al. 1989 ). Young females were used to optimize for early germarium stages without the complication of the much larger stage 14 oocytes. All the experiments were performed without heat shock because the basal level of hsp83 expression is similar to the pattern of mei-P22 in the germarium. To detect the HA epitope-tagged MEI-P22, the primary antibody used was either the monoclonal mouse anti-HA 12CA5 (1:20) or the monoclonal rat anti-HA "high affinity" 3F10 (1:100; Roche Biochemicals). The guinea pig anti-C(3)G antibody (PAGE and HAWLEY 2001 ) was used at 1:500, and a combination of two Orb antibodies (4H8 and 6H4; LANTZ et al. 1994 ) was used at 1:150. Secondary antibodies were FITC-labeled goat anti-guinea pig (Vector Laboratories, Burlingame, CA) used at 1:300, Cy5-labeled goat anti-mouse (Amersham-Pharmacia) used at 1:40, and Cy3-labeled goat anti-rat (Amersham-Pharmacia or Jackson ImmunoResearch) used at 1:250. Chromosomes were stained with 4',6-diamidino-2-phenylindole (0.2 µM) for 10 min or Hoechst (0.1 µl/ml of a 10 mg/ml solution) for 5 min. For immunolocalization of DSBs, the anti-phospho-H2AX (Ser139) rabbit polyclonal antibody (Upstate Biotechnology, Lake Placid, NY; ROGAKOU et al. 1999 ) was used at 1:100 and the secondary Cy3-labeled goat anti-rabbit (Amersham) was used at 1:300.

For BrdU staining, ovaries were incubated with 10 mM BrdU (Sigma, St. Louis) in Grace's insect cell culture medium (Life Technologies) for 1 hr. The ovaries were then fixed as described above and washed 2 x 15 min in PBS + 0.6% Triton-X100 and 2 x 15 min in DnaseI buffer (Life Technologies) and then incubated in 25 units DnaseI (in 0.5 ml DnaseI buffer) at 37° for 30 min. Monoclonal mouse anti-BrdU (Becton-Dickinson, San Jose, CA) was used at 1:20.

Most of the images were collected using a Leica TCS SP2 confocal microcope or a Zeiss Axioplan II imaging microscope equipped with a Cooke Sensicam CCD camera. All images were collected using a x63 or x100 objective. For analysis by deconvolution, the images were collected with a DeltaVision restoration microscopy system (Applied Precision) equipped with a Nikon x60 N.A.1.4 oil immersion objective. The restoration and modeling was performed with softWoRx software (Applied Precision) on an Octane Workstation (Silicon Graphics).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The effect of mei-P22 on chromosome disjunction and crossing over:
Previous studies found that the mei-P22¹ mutation caused a reduction in the frequency of gene conversion, which is evidence that mei-P22 has a role in initiating meiotic recombination (MCKIM et al. 1998 ). This mutation was not ideal for further studies because it is not a null allele (see below). We isolated four additional mei-P22 mutations (MATERIALS AND METHODS) and characterized their effects on X chromosome nondisjunction (Table 1) and crossing over (Table 2). Two, mei-P22²⁰⁵ and mei-P22²⁰⁶, were weaker alleles and had relatively low X chromosome nondisjunction. Crossing over in mei-P22²⁰⁶ homozygotes was decreased to 50% of wild type in the cu-e and e-ca regions. In the st-cu region, however, crossing over was increased to 156% of wild type. The st-cu region includes the centromere and in wild type exhibits crossover suppression relative to the genome average. Therefore, these results show that the crossover reductions in mei-P22²⁰⁶ were accompanied by a change in the distribution of residual events along the chromosome. In Drosophila, many, although not all, recombination-defective mutants also change the distribution of crossing over relative to wild type ("precondition mutants"; BAKER and HALL 1976 ).

The other two mutations, mei-P22¹⁰³ and mei-P22^N1, caused higher levels of X chromosome nondisjunction and drastically reduced the frequency of crossing over. mei-P22¹⁰³ may be a null allele because it reduced crossing over to extremely low levels (0.3% of wild type) similar to the mei-W68 mutant phenotype. mei-P22^N1 had a low level of crossing over on the third chromosome (13% of wild type), suggesting that it reduced but did not eliminate the initiation of recombination. Crossing over was more severely affected on the second chromosome in mei-P22^N1 females (4.2% of wild type), suggesting that in this mutant the recombination frequency was sensitive to the chromosome or genetic background. The mei-P22^N1 second chromosome crossover frequency was similar to our observations with the original allele mei-P22¹, suggesting that this mutation is also not a null allele.

mei-P22 mutants lack DSBs:
The initial step in meiotic recombination is the formation of DSBs. To examine whether mei-P22 is required before or after establishment of DSBs, we performed two experiments. The first experiment was to determine if an exogenous source of DSBs, X rays, could rescue a strong mei-P22 mutant phenotype. The second experiment was to determine if a mei-P22 mutation could suppress the phenotype of a mutant defective in the repair of DSBs.

Previous experiments have shown that X-irradiation partially rescues the meiotic recombination defect in mutants unable to generate DSBs, including spo11 homolog mutants of S. cerevisiae (THORNE and BYERS 1993 ), C. cinereus (CELERIN et al. 2000 ), C. elegans (DERNBURG et al. 1998 ), and Drosophila (R. BHAGAT and K. MCKIM, unpublished results). Conversely, irradiation has no effect on mei-218 mutants that act after DSB formation or in c(3)G mutants (ROBERTS 1969 ; R. BHAGAT and K. MCKIM, unpublished results). We exposed mei-P22¹⁰³ mutant females to 4000 rad of X-irradiation, crossed them with males to score either crossing over or X chromosome nondisjunction, and then transferred them to new vials every 3 days. Crossing over was induced by as much as 120-fold compared to unirradiated controls, almost half of wild-type levels (Table 2 and Fig 1), demonstrating that exogenously induced DSBs could compensate for the mei-P22 defect. Nondisjunction was also decreased, as was expected if the induction of crossing over resulted in chiasmata that could direct the segregation of homologous chromosomes. The most significant increases in crossing over and decreases in X chromosome nondisjunction were observed in the offspring derived from oocytes fertilized 4–6 days after radiation treatment. In these offspring, for example, X chromosome nondisjunction was reduced to 7.9% in contrast to 34.6% in unirradiated mutant females. These oocytes were in pachytene when their mothers were exposed to X rays (KING 1970 ; SPRADLING et al. 1997 ), suggesting that there is a limited time frame in oogenesis during which DSBs efficiently generate crossovers.

The st-cu region of the third chromosome was more sensitive to radiation-induced crossing over than were other regions. In the 4- to 6-day brood, crossing over in the st-cu interval occurred at 194% of wild type, whereas in the more distal cu-ca region crossing over occurred at 31% of wild type. The st-cu region was also significant because crossing over was induced by radiation at relatively high levels through all broods. In contrast, an increased frequency of crossing over in the cu-e and e-ca regions occurred only in the 4- to 6- and 7- to 9-day broods (data not shown). From the markers used, we cannot determine if these interval-specific effects of the st-cu region are a result of heterochromatic crossovers or if X-ray-induced crossovers exhibit the same changes in distribution observed with endogenously induced crossovers in mei-P22 hypomorphs.

As a second test of the relationship between mei-P22 function and DSB formation, we constructed a double mutant with spnB^BU. spnB encodes a meiosis-specific Rad51 homolog and is required for meiotic DSB repair. spnB mutants are sterile because defects in meiotic DSB repair cause the oocyte to develop abnormally (GHABRIAL et al. 1998 ). Mutants that eliminate DSBs, such as mei-W68, suppress the oogenesis defects and partially restore fertility to spnB (GHABRIAL and SCHUPBACH 1999 ; R. PATEL and K. MCKIM, unpublished results). In females homozygous for both mei-P22^N1 and spnB^BU, fertility was increased more than threefold compared to those for spnB^BU alone (Fig 2). While the spnB^BU mutant females are completely sterile by the time they have been laying eggs for 3 days, the mei-P22^N1 spnB^BU double mutants continued to produce progeny as they aged. This increase in fertility was similar to that observed with mei-W68¹; spnB^BU homozygotes and is consistent with a failure to create DSBs in mei-P22 mutants. The fact that the fertility in the double mutants was not restored to mei-P22 (or mei-W68) single-mutant levels indicates that spnB might have additional roles during oogenesis.

Finally, we employed a cytological assay of DSB creation by using an antibody to the phosphorylated form of human H2AX (-H2AX), which detects a histone modification that occurs at DSBs in meiotic cells of male mice (MAHADEVAIAH et al. 2001 ). We have shown that this antibody detects DSB-dependent foci during meiotic prophase of Drosophila females (D. SHERIZEN, J. K. JANG and K. MCKIM, unpublished results). In mei-P22^N1 and mei-P22¹⁰³ mutants, these foci were not observed (data not shown). This result is most striking, however, when a spnB double mutant is examined because the foci of -H2AX staining persist longer in DSB repair-defective mutants than in wild type (D. SHERIZEN and K. MCKIM, unpublished results). An example of this is shown in Fig 2B; all oocytes from spnB^BU females accumulate -H2AX foci late in meiotic prophase. In a mei-P22^N1 spnB^BU double mutant, however, these foci were not observed. It is likely that -H2AX foci are absent in the double mutant because DSBs are not created in a mei-P22 mutant.

mei-P22 encodes a small, novel, basic protein:
Genomic sequences flanking the P-element insertion site in mei-P22¹ were used to clone the gene (MATERIALS AND METHODS). The assignment of the mei-P22 coding region was confirmed by rescuing mutants with transgenes and the sequencing of mutations. The original P-element mutation inserted 52 bp upstream of the ATG. For each of the mei-P22 EMS alleles, we identified a single nucleotide sequence change within the coding region (Fig 3). The two severe EMS alleles mei-P22^N1 and mei-P22¹⁰³ were found to be stop codons while the hypomorphs mei-P22²⁰⁵ and mei-P22²⁰⁶ were missense mutations. Surprisingly, some genetic tests showed mei-P22^N1 to be a weaker allele than mei-P22¹⁰³ (e.g., Table 2). Hypomorphic alleles that are stop codons have been observed in the meiotic recombination genes mei-9 and mei-217 (MANHEIM et al. 2002 ; J. SEKELSKY, personal communication). In this case, it is possible that in some genetic backgrounds there may be a low frequency of translational readthrough of the mei-P22^N1 stop codon. Genomic sequence analysis and RT-PCR (data not shown) showed that mei-P22 encodes a 954-bp ORF with no introns. MEI-P22 is predicted to be 318 amino acids with a potential bipartite nuclear localization sequence (NLS) between amino acids 145 and 162, has a predicted isoelectric point of 10.23, and has no homologs in the sequence databases (Fig 3).

We used the hsp83 promoter to express the transgene because its expression pattern in the germarium is similar to that of mei-P22 (see below), and we experienced difficulties achieving transgenic rescue of the mutant phenotype with the endogenous promoter, presumably due to position effects. Fused to the 5' end of the mei-P22 coding region was an epitope tag encoding three hemagglutinin (3XHA) peptide sequences to facilitate immunolocalization of the protein (see below). Several hsp83::mei-P22^3XHA transgenic lines that rescued the mei-P22 mutant phenotype were isolated as well as some that did not (Table 1 and MATERIALS AND METHODS).

Analysis of mei-P22 expression: overview of Drosophila oogenesis:
Our analysis of mei-P22 RNA and protein expression was performed using whole mounted ovaries. Not only did this approach maintain the three-dimensional structure of the meiotic cells, but it also allowed us to determine the time course of MEI-P22 expression since cells in the ovary are arranged in order of developmental age. An overview of Drosophila oocyte development is described below, and a schematic of the germarium and summary of our results are shown in Fig 4A.

View larger version (86K): In this window In a new window Download PPT slide   Figure 4. Expression of mei-P22 in the germline. (A) Schematic diagram of the Drosophila germarium showing the stages of development, SC formation (green), and MEI-P22 foci dynamics (red). Not all 16 cells in each cyst are shown. In region 1, four rounds of mitotic divisions create the 16-cell cysts. Cytoplasmic ORB staining (blue) is initially equal in all 16 cells and then localizes to the oocyte late in region 2a. The first SC forms in the 2 cells with four intercellular connections or ring canals, the pro-oocytes. Later, more cells form SC, with the 2 three-ring canal cells forming extensive amounts, although distinguishable from the 2 four-ring canal cells by comparatively lighter staining with C(3)G. In region 2b cysts, the cells flatten out. Finally, all cells but the oocyte exit the meiotic program and lose their SC. On the basis of cyst position and ORB, C(3)G, and MEI-P22 staining, we propose four stages of meiotic prophase in region 2a cysts (labeled I–IV). In temporal order they are: (I) cysts with no SC, ORB, or MEI-P22; (II) cysts that stain with C(3)G but lack cytoplasmic ORB and nuclear MEI-P22 staining; (III) cysts with cytoplasmic ORB and MEI-P22 in the nucleus of the 2 cells with C(3)G, the pro-oocytes; (IV) cysts with C(3)G staining in additional cells, such as those with three- or two-ring canals; MEI-P22 persists in these cells and may be absent from the 2 four-ring canal cells; and (V) cysts with SC, which is beginning to be restricted to the oocyte, but in which MEI-P22 is absent. (B) mei-P22 is expressed during meiosis in the Drosophila ovary. In situ hybridization using an antisense RNA probe for the mei-P22 coding region is shown. The figure is labeled to show how the transcript is enriched in the germarium. Stage 2 cysts are the first stage of the vitellarium. The staining on the edge of some vitellarium cysts is nonspecific. (C) MEI-P22 (red) appears early in the germarium. The anterior of the germarium, and hence the earlier stages of development, is at the top in B–D and is shown by an arrow. In this volume projection of a confocal stack of images, C(3)G (green) appears prior to the earliest appearance of ORB (blue) and MEI-P22 (asterisk). Late region 2a and region 2b cysts have no MEI-P22 foci. (D) Volume projection of confocal images showing MEI-P22 (red) closely associated with C(3)G (green). A white outline shows the approximate shape of four region 2a cysts, two with MEI-P22 staining. The logically inferred developmental order of the cysts is indicated by a number. In cyst 1, MEI-P22 appears in the two nuclei with the strongest C(3)G staining (pro-oocytes) but also in nuclei with weaker staining (nurse cells). In cyst 2, MEI-P22 is present at low levels in the pro-oocytes and is most abundant in cells with punctate C(3)G staining. In later stage cysts (3 and 4) the oocytes still have C(3)G staining but no MEI-P22.

Drosophila females have two ovaries, each composed of 10–15 ovarioles containing chains of developing oocytes. At the anterior end of each ovariole is the germarium, where four rounds of incomplete mitotic divisions produce a 16-cell cyst with intercellular junctions termed ring canals. Although each cyst will eventually contain just one oocyte, several cells enter meiosis. Specifically, the SC first develops in the two cells with four ring canals (the pro-oocytes) and later SC develops to variable extents in the other cells with fewer ring canals. Eventually, all cells but the oocyte exit the meiotic program and SC is maintained in only one cell, the oocyte (CARPENTER 1975A , CARPENTER 1979 ).

The cysts move down the germarium as they mature but their absolute position does not equate to a specific stage in meiotic prophase (CARPENTER 1975A ). More significantly, the cysts are usually arrayed in order of developmental age such that a cyst in a posterior position is usually at a later stage of meiotic prophase than that of a cyst in a more anterior position. There can be exceptions to this arrangement, but relative cyst position is a useful tool to identify and compare oocytes in different stages of meiosis (Fig 4A). The germarium is divided into four regions on the basis of the morphology of the 16-cell cysts, which we distinguished using an antibody to the ORB protein (LANTZ et al. 1994 ). ORB is a cytoplasmic protein that first appears in early region 2a and becomes enriched in the oocyte by region 2b. In region 2a of the germarium, meiotic prophase begins and recombination is initiated (CARPENTER 1979 ).

mei-P22 RNA is expressed in the germarium:
We were able to detect mei-P22 mRNA in the Drosophila ovary using in situ hybridization with an antisense RNA probe. mei-P22 RNA was specifically observed in the postmitotic region of the germarium, in regions 2 and 3 where meiotic prophase occurs (Fig 4B). No staining was observed in later stages of the vitellarium. In contrast, we were unable to detect the transcript by Northern blot of ovary total RNA or by screening two ovarian cDNA libraries, suggesting that the transcript is rare (data not shown).

MEI-P22 protein is detected during early meiotic prophase:
To gain further insights into the function of mei-P22 and its relationship to DSB formation, we examined the localization of the protein. Since two attempts at raising antibodies to the endogenous protein failed, we used the rescuing transgenes (described above) containing the mei-P22 coding region fused at the amino terminus to three copies of the HA epitope tag. For most immunolocalization experiments we used the transformant line P{hsp83::mei-P22^3XHA}9 because it provided the best rescue of the mei-P22 mutant defects in disjunction (Table 1) and crossing over (Table 2). The efficient rescue was possible because the hsp83 promoter drives expression without heat shock throughout regions 2 and 3 of the germarium (DING et al. 1993 ; MANHEIM et al. 2002 ).

When Drosophila ovaries carrying the P{hsp83::mei-P22^3XHA}9 transgene were stained with an anti-HA antibody, MEI-P22 fusion protein was detected in region 2a of the germarium (Fig 4 and Fig 5). This staining pattern was observed whether the transgene was in a wild-type or mei-P22 mutant background. On the basis of a comparison to ORB staining, MEI-P22 staining was restricted to region 2a (i.e., early pachytene) and usually disappeared by the end of region 2a (Table 3), suggesting that relative to other events in the oocyte, MEI-P22 staining is short lived. In addition, MEI-P22 staining appeared as numerous foci, and because it always overlapped with the DNA stain, it appears to be a nuclear protein (Fig 5A). Similar patterns were observed with other transgenic lines that rescued the mei-P22 mutant phenotype, such as P{hsp83::mei-P22^3XHA}4 and P{hsp83::mei-P22^3XHA}X1. No staining was observed in transgenics that failed to rescue the mutant phenotype such as P{hsp83::mei-P22^3XHA}2 (data not shown).

View larger version (39K): In this window In a new window Download PPT slide   Figure 5. MEI-P22 is localized to the nuclei of wild type but not of c(3)G mutants. (A) Wild-type germarium with MEI-P22 (red) appearing in the nuclei (DNA in blue) of the earliest cyst in region 2a with cytoplasmic ORB (green) staining. (B) MEI-P22 is absent in c(3)G mutant oocytes. Volume projection of a c(3)G mutant germarium with ORB to show these are region 2a cysts where MEI-P22 would normally be. The few nuclei with red staining are follicle cells. In A and B, an arrow indicates the anterior end of the germarium. (C) MEI-P22 in the mitotically dividing somatic follicle cells. The focal plane of this image is the outside edge of an ovarian stage 3 cyst. This cyst is located in the vitellarium, the region after the germarium. BrdU incorporation experiments showed that this expression pattern did not correspond to S-phase (data not shown). Within the nucleus of these cells, MEI-P22 colocalizes with the DNA and has a fibrous appearance, suggesting that it is binding to the chromosomes, although there are no discrete foci. Bars, 8 µm.

Evidence for post-transcriptional regulation of mei-P22:
The MEI-P22 expression pattern may be the result of post-transcriptional regulation. The mRNAs of mei-P22 (Fig 4) and hsp83 (DING et al. 1993 ) are expressed throughout regions 2 and 3 of the germarium. Furthermore, when the hsp83 promoter was used to drive expression of mei-218, protein was observed in regions 2 and 3 of the germarium (MANHEIM et al. 2002 ). Thus, we suspect that the mRNA from the P{hsp83::mei-P22^3XHA}9 transgene is expressed throughout regions 2 and 3. In contrast, MEI-P22 was detected in a more restricted pattern and was never observed in region 3 cells.

In addition to expression in the germarium, using the hsp83 promoter resulted in expression of mei-P22 protein in a subset of the somatic follicle cells (Fig 5C). Because the follicle cells are asynchronously dividing, this result suggests that MEI-P22 levels or nuclear entry are regulated at a post-transcriptional level by the cell cycle. Therefore, the mechanism that causes MEI-P22 to be present for only a limited portion of meiotic prophase could also be functioning in mitotic cells. In contrast to this similarity in regulation, the pattern of nuclear staining in the meiotic (germline) and mitotic (follicle) cells differed in appearance. Unlike the discrete foci observed in the germline meiotic cells, mitotic cell nuclear localization was uniformly spread along the chromosomes. This difference suggests that there must be meiosis-specific factors that restrict the binding of MEI-P22 to a small number of chromosomal sites during meiotic prophase.

MEI-P22 appears after SC formation begins:
To accurately relate MEI-P22 localization to meiotic events in the oocyte, we compared the localization of MEI-P22 to the development of the SC. To detect the presence of the SC, we used an antibody to the c(3)G protein, which is proposed to be a component of the transverse elements (PAGE and HAWLEY 2001 ). Prior to ORB localization, SC formation is the earliest indicator of the oocyte. C(3)G antibody staining was observed in at least the two pro-oocytes in each region 2a cyst. In addition, C(3)G stained transiently in additional cells of some region 2a cysts but was usually limited to the oocyte by region 3 (Fig 4).

In 50% of the germaria, there was one early region 2a cyst with C(3)G staining but no MEI-P22 foci or ORB staining (Fig 4C). These cysts are probably a normal stage in ovary development but were not observed in all germaria because they persist for a short period of time. These cells also stained lighter for C(3)G, suggesting that they are at the zygotene-pachytene transition. As noted by CARPENTER 1975A , zygotene is a relatively brief stage of prophase in Drosophila. In the remaining germaria, the earliest MEI-P22 was observed simultaneously with the appearance of ORB and C(3)G staining (Fig 4D and Fig 5A). MEI-P22 was never observed in a cyst that did not have C(3)G staining in the oocyte. On the basis of these data, we conclude that SC forms first and is rapidly followed by MEI-P22 accumulation.

There is probably an indirect relationship between C(3)G and MEI-P22. That is, although the MEI-P22 localization occurred after and was dependent on C(3)G, these proteins may not directly interact. Three-dimensional reconstructions of deconvolved image stacks revealed that MEI-P22 usually occupied a space that was separate from SC, although the foci always overlapped the DNA staining (Fig 6). These observations suggest that SC development alters the chromosomes to allow MEI-P22 binding.

View larger version (37K): In this window In a new window Download PPT slide   Figure 6. MEI-P22 is nuclear and co-localizes with the DNA but not C(3)G. (A) Image following deconvolution and projection of the stack with MEI-P22 (red) and C(3)G (green). The frequent presence of red and infrequent presence of yellow staining indicate that most of the MEI-P22 staining does not overlap with C(3)G. Bar, 1 µm. (B) A similar type of image from a different nucleus showing MEI-P22 (red) is always associated with the DNA (blue). (C and D) A model of the nucleus shown in A with the two images rotated 90° relative to each other to show that the MEI-P22 foci are adjacent to the SC. As is typical for wild type, 10 foci are visible (see text).

MEI-P22 foci transiently associate with pachytene chromosomes:
Compared to the other cells in each 16-cell cyst, C(3)G staining is the strongest and forms the most threadlike structures in the two pro-oocytes (PAGE and HAWLEY 2001 ). Abundant MEI-P22 foci were observed in the nuclei of these cells but usually in only two to four cysts per germarium (Table 3). Some pachytene pro-oocytes in late region 2a and all in region 2b had no MEI-P22 staining, however, suggesting that the appearance and disappearance of MEI-P22 foci in the pro-oocytes is rapid. For example, MEI-P22 foci were usually observed in the pro-oocytes at the earliest stages of pachytene, whereas oocytes that appeared to be in later stages of pachytene, such as those in more posterior located cysts of region 2a and 2b, lacked MEI-P22 foci (Fig 4D). The nurse cells in region 2a cysts usually contained MEI-P22 foci as well. These MEI-P22 positive nurse cells often had punctate C(3)G staining, suggesting that they had partially entered the meiotic program. In fact, in late region 2a cysts MEI-P22 foci commonly were still present in the nurse cells even though they had disappeared from the pro-oocytes. Furthermore, the foci were more abundant in some nurse cells than in the pro-oocytes. These observations of the nurse cells suggest that MEI-P22 accumulated for a longer time in those cells where SC did not fully develop (Fig 4D).

While initially surprising, the presence of MEI-P22 in cells fated to become nurse cells can be explained in the context of germline cyst development. While some of the nurse cell staining could be due to abnormally high expression levels from the transgene, there is also good evidence based on the appearance of recombination nodules (CARPENTER 1994 ) and the detection of DSBs with antibodies (D. SHERIZEN and K. MCKIM, unpublished results) that the nurse cell progenitors experience double-strand breaks. Given that SC develops in several cells of the 16-cell cyst, but then rapidly degrades, and that c(3)G is required for MEI-P22 localization, one possibility is that MEI-P22 accumulates only in regions of the chromosomes where SC previously developed. Cells with punctate C(3)G staining may have persistent MEI-P22 foci because the completion of SC formation (pachytene) and/or DSB repair in the pro-oocytes may hasten the removal of MEI-P22 from the chromosomes. This conclusion is supported by our findings, described below, that delays in DSB repair alter the dynamics of the MEI-P22 foci.

The relationship of MEI-P22 localization to double-strand-break repair:
To determine if DSB repair plays a role in removing MEI-P22 from meiotic chromosomes, we stained for foci in mutants where DSBs either do not form or are not correctly repaired. We constructed hsp83::mei-P22^3XHA mei-W68⁴⁵⁷² females to observe MEI-P22 localization when DSBs do not occur (meiotic recombination is eliminated in a mei-W68⁴⁵⁷² mutant) and hsp83::mei-P22^3XHA ; spnB^BU females to observe MEI-P22 when there is a defect in DSB repair (GHABRIAL et al. 1998 ). In both types of females, we found that MEI-P22 foci appeared at the normal time in region 2a, but there were two important differences from the wild-type dynamics of the foci. First, in both mei-W68 and spnB mutants, the MEI-P22 foci usually persisted into region 2b cysts (Fig 7). Second, based on double staining with C(3)G in these mutants, there were more cysts per germarium with MEI-P22 foci in the two pro-oocytes. In wild type, approximately four cysts with MEI-P22 staining were in the pro-oocytes, whereas in these mutants approximately six cysts with MEI-P22 staining were in the two pro-oocytes (Table 3). Both mutants and wild type had a similar number of cysts in each germaria and the increase in the mutant germaria could be attributed to an increased number of consecutive cysts with pro-oocyte MEI-P22 staining. One explanation for these results is that DSB repair may increase the rate at which MEI-P22 is removed from the chromosomes. Even in the absence of DSB repair, however, MEI-P22 foci were not observed later in prophase (region 3). Therefore, other factors must have a role in regulating MEI-P22 chromosome binding.

View larger version (94K): In this window In a new window Download PPT slide   Figure 7. Expression of MEI-P22 (shown in red) in meiotic mutants. MEI-P22 is present in mei-W68 (A and B) and spnB (C) mutants. (A) A volume projection of a germarium showing normal appearance of MEI-P22 (red) in region 2a along with C(3)G (green) and ORB (blue) but unusual persistence into region 2b (indicated by an asterisk). (B and C) Single confocal sections showing that MEI-P22 is associated with the C(3)G of region 2b cysts.

If DSB repair increases the rate at which MEI-P22 is removed from chromosomes, the relatively high abundance of MEI-P22 foci in pro-nurse cells can be explained. In the cells that never complete SC formation (pro-nurse cells), the rate of DSB repair may be lower than that in the cells where SC forms completely (pro-oocytes). The low rate of DSB repair in the pro-nurse cells could occur because they only partially enter the meiotic program. Consistent with this hypothesis, in the mutants without DSB repair the MEI-P22 foci appeared and then disappeared with similar dynamics in both the pro-oocytes and the pro-nurse cells.

The frequency of MEI-P22 foci:
We counted the number of MEI-P22 foci in nuclei with the strongest C(3)G staining, the two cells with four ring canals. We focused on these cells for two reasons: first, they are the only cells in each cyst that reliably achieve full synapsis of the homologs (pachytene) and second, one of these cells will become the oocyte for which genetic data on recombination frequencies are available for comparison. In wild type, we counted an average of 8.7 foci per cell whereas, on the basis of limited data, we estimate 15 per DSB per nucleus [on the basis of the estimate of one crossover every five gene conversions at the rosy locus (HILLIKER and CHOVNICK 1981 ) that the crossing over at the rosy locus is less frequent (1/2) than the genome average and that there is an average of 1.2 crossovers per chromosome arm]. It is likely that the lower number of MEI-P22 foci was due to their dynamic and transient nature in the pro-oocytes.

MEI-P22 localization on the chromosomes is c(3)G dependent:
In c(3)G mutants, SC does not develop (SMITH and KING 1968 ; RASMUSSEN 1975 ) and meiotic recombination is drastically reduced (CARLSON 1972 ; HALL 1972 ). To investigate the possibility that C(3)G is required to recruit proteins required for DSB formation to meiotic chromosomes, we constructed hsp83::mei-P22^3XHA; c(3)G⁶⁸ females. In these females, we did not detect MEI-P22 foci in the germline (Fig 5B, n = 13), providing a strong correlation between the presence of the foci and meiotic recombination. It is possible that MEI-P22 stability or localization to meiotic chromosomes requires the SC.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

What do the MEI-P22 foci represent?
Several lines of evidence support the conclusion that mei-P22 is required for an early step in recombination during Drosophila meiosis. First, both gene conversion and crossing over are eliminated in strong mei-P22 mutants (MCKIM et al. 1998 and this article). Second, mei-P22 mutants suppress the sterility phenotype of DSB repair-defective mutants such as spnB^BU. Third, crossing over is restored by an exogenous source of double-strand breaks, implying that mei-P22 mutants lack them. Finally, the staining by an antibody that detects histone modifications at DSB sites, -H2AX, is eliminated in mei-P22 mutants. From the similarity of these phenotypes to those observed in mei-W68 mutants, we hypothesize that mei-P22 is required for DSB formation.

Our most significant and novel finding is that MEI-P22 is found as chromosome-associated foci early in meiotic prophase. In our results, the strong correlation between the presence of MEI-P22 foci and meiotic recombination suggests that these foci are relevant to DSB formation. The MEI-P22 foci are present at a narrower window of time during pachytene than is the pattern of RNA expression, making it likely that the appearance and disappearance of the foci are unrelated to the promoter used and instead are a product of post-transcriptional regulation. During this portion of pachytene, the early recombination nodules appear on the chromosomes (CARPENTER 1975B ). Early recombination nodules are the structures proposed to be associated with early stages of meiotic recombination and are observed at the same time, early pachytene (region 2a of the germarium), as the MEI-P22 foci are observed. This correlation suggests that there may be a relationship between the MEI-P22 foci and DSB formation.

Several additional observations support a functional significance between the MEI-P22 foci and DSB sites. For instance, the MEI-P22 foci are influenced by mutations in other meiotic recombination genes. The MEI-P22 foci are absent in a c(3)G mutant in which the SC does not form, and they persist in the oocyte for a longer time in mutants that do not induce or repair DSBs. Also, MEI-P22 foci form only in the germline and their numbers are similar to the predicted number of recombination events in the genome. Finally, there was a perfect correlation between the transgenes that rescued the mei-P22 mutant phenotype and the presence of the foci. In combination with the genetic results summarized above, and on the basis of these cytological observations, we suggest that the MEI-P22 foci represent the sites along the chromosomes where DSBs will form.

Similar examples in other systems of proteins required for DSB formation to be found localized to meiotic chromosomes are limited. In S. cerevisiae, the Mre11/Rad50/Xrs2 complex has been observed, but is visible only in a rad50S mutant where DSBs accumulate because they are not resected and repaired (USUI et al. 1998 ). This complex was also observed in male mouse meiosis but there was no correlation with DSB sites (EIJPE et al. 2000 ). More relevant to our findings is the Rec102 protein. This S. cerevisiae meiosis-specific protein is required for DSB formation and has been observed as foci on meiotic chromosomes by immunofluorescence (KEE and KEENEY 2002 ).

MEI-P22 foci and DSB formation are dependent on chromosome structure:
If we assume that MEI-P22 foci respond to the same factors that regulate meiotic recombination, our observations provide a view of the regulatory mechanisms that ensure that only a small number of DSBs are generated in each cell. We propose the following sequence of events for regulating MEI-P22 and DSB formation in Drosophila female meiosis (Fig 4A). First, the homologs align and form SC. Second, SC-dependent changes in chromatin structure provide the conditions that promote MEI-P22 accumulation at discrete chromosome sites. If MEI-P22 is not expressed early enough, it may not be able to gain access to the chromosomes. Third, MEI-P22, and likely other proteins, promote the DSB activity of the Spo11 homolog MEI-W68. The observation that MEI-P22 foci are observed in the absence of DSB activity (i.e., a mei-W68 mutant) demonstrates that MEI-P22 is not recruited to chromosomes in response to a breakage event. Finally, the MEI-P22 foci begin to disappear, first from the pachytene cells and later from the other cells that form partial SC.

The dependence of MEI-P22 localization on C(3)G is consistent with genetic studies in Drosophila females where meiotic recombination does not occur in the absence of SC (CARLSON 1972 ; HALL 1972 ; PAGE and HAWLEY 2001 ) and SC can form normally between homologs in the absence of recombination (MCKIM et al. 1998 ). C(3)G and SC formation are not sufficient, however, for MEI-P22 localization. In later stages of pachytene, regions 2b and 3, C(3)G is still assembled between the homologs but MEI-P22 foci are never present. The disappearance of MEI-P22 is not due to transcriptional regulation and demonstrates that the SC has an essential but not sufficient role in regulating MEI-P22 accumulation. In fact, the completion of synapsis may have a negative influence on the foci (see below).

Limiting the time when MEI-P22 binds to chromosomes may regulate DSB formation:
The brief appearance of MEI-P22 in region 2a cells demonstrates that its expression is tightly controlled. Our evidence suggests that the rapid removal of MEI-P22 foci may be related to DSB repair activities, which for a number of reasons could be more efficient in the oocyte. For example, complete SC formation might stimulate DSB repair. Alternatively, and akin to SC formation in the 16-cell cyst, DSB repair proteins might be present in the greatest concentration in the two prospective oocytes. The ultimate inhibition or removal of MEI-P22 foci, however, occurs even in the absence of DSBs and therefore is unrelated to DSB repair. Our observations are consistent with two nonexclusive processes that may have a role in the final removal and/or inhibition of MEI-P22 chromosome binding. First, the nuclear localization and chromosome binding of MEI-P22 in the germline may be regulated by a mechanism involving cell-cycle controls that also function in mitotic cells. This has an obvious evolutionary attraction, given that meiosis is thought to have evolved from a mitotic cell cycle. Second, not only is the SC required for the formation of MEI-P22 foci but also its maturation and/or complete synapsis may contribute to their disappearance. This inhibition of foci formation may correspond to changes in SC morphology previously noted by CARPENTER 1975A . As the cells reach pachytene, the SC progressively shortens and thickens. The corresponding structural changes to the SC could modify the chromosomes such that MEI-P22 can no longer bind and therefore prevent further DSB formation.

There are two reasons to regulate DSB formation within the cell. First, the timing of DSB formation is important for its meiotic function. The analysis of X-ray-induced crossing over shows that for DSBs to effectively induce crossovers, they must occur during pachytene. Second, once a sufficient number of DSBs have been induced, the activity of generating breaks must then be attenuated. Failure to stop the induction of breaks could lead to excessive chromosomal breakage and cell lethality. It is reasonable to conclude from the foci dynamics that MEI-P22 is under the control of a system regulating DSB formation. The ability to directly observe MEI-P22 localizing to potential DSB sites will facilitate further studies on the mechanisms that regulate DSB formation.

	ACKNOWLEDGMENTS

We are grateful to Rajal Bhagat and Dalia Sherizen for help with the screen for mei-P22 alleles; to Hugo Dooner, R. Scott Hawley, Ken Irvine, Elizabeth Manheim, Ruth Steward, and Anne Villeneuve for insightful discussions and comments on the manuscript; to Eric Lam in whose laboratory the deconvolution microscopy was performed; to Jay Tischfield for use of an X-irradiator; to Brian Calvi and Mary Lilly for advice on BrdU labeling; and to Li Nguyen for technical assistance. We also thank Scott Page and Scott Hawley for providing the C(3)G antibody and data prior to publication. The initial molecular characterization and sequencing of the mei-P22 locus was carried out in the laboratory of R. Scott Hawley. Some of the stocks used in this study were received from the Bloomington Stock Center. The ORB antibodies were obtained from the Developmental Studies Hybridoma Bank at the University of Iowa, developed under the auspices of the National Institute of Child Health and Human Development. This work was supported by grant MCB-0077705 from the National Science Foundation.

Manuscript received March 7, 2002; Accepted for publication June 17, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ASHBURNER, M., 1989 Drosophila: A Laboratory Handbook. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

BAKER, B. S., and J. C. HALL, 1976 Meiotic mutants: genetic control of meiotic recombination and chromosome segregation, pp. 351–434 in The Genetics and Biology of Drosophila, edited by M. ASHBURNER and E. NOVITSKI. Academic Press, New York.

BELMONT, A. S., M. B. BRAUNFELD, J. W. SEDAT, and D. A. AGARD, 1989  Large-scale chromatin structural domains within mitotic and interphase chromosomes in vivo and in vitro. Chromosoma 98:129-143.[Medline]

CARLSON, P. S., 1972  The effects of inversions and the c(3)G mutation on intragenic recombination in Drosophila. Genet. Res. 19:129-132.[Medline]

CARPENTER, A. T. C., 1975a  Electron microscopy of meiosis in Drosophila melanogaster females. I. Structure, arrangement, and temporal change of the synaptonemal complex in wild-type. Chromosoma 51:157-182.[Medline]

CARPENTER, A. T. C., 1975b  Electron microscopy of meiosis in Drosophila melanogaster females. II. The recombination nodule—A recombination-associated structure at pachytene? Proc. Natl. Acad. Sci. USA 72:3186-3189.[Abstract/Free Full Text]

CARPENTER, A. T. C., 1979  Synaptonemal complex and recombination nodules in wild-type Drosophila melanogaster females. Genetics 92:511-541.[Abstract/Free Full Text]

CARPENTER, A. T. C., 1994  egalitarian and the choice of cell fates in Drosophila melanogaster oogenesis. Ciba Found. Symp. 182:223-254.[Medline]

CELERIN, M., S. T. MERINO, J. E. STONE, A. M. MENZIE, and M. E. ZOLAN, 2000  Multiple roles of Spo11 in meiotic chromosome behavior. EMBO J. 19:2739-2750.[Medline]

DERNBURG, A. F., K. MCDONALD, G. MOULDER, R. BARSTEAD, and M. DRESSER et al., 1998  Meiotic recombination in C. elegans initiates by a conserved mechanism and is dispensable for homologous chromosome synapsis. Cell 94:387-398.[Medline]

DING, D., S. M. PARKHURST, S. R. HALSELL, and H. D. LIPSHITZ, 1993  Dynamic Hsp83 RNA localization during Drosophila oogenesis and embryogenesis. Mol. Cell. Biol. 13:3773-3781.[Abstract/Free Full Text]

EIJPE, M., H. OFFENBERG, W. GOEDECKE, and C. HEYTING, 2000  Localisation of RAD50 and MRE11 in spermatocyte nuclei of mouse and rat. Chromosoma 109:123-132.[Medline]

FINELLI, A. L., C. A. BOSSIE, T. XIE, and R. W. PADGETT, 1994  Mutational analysis of the Drosophila tolloid gene, a human BMP-1 homolog. Development 120:861-870.[Abstract]

GHABRIAL, A. and T. SCHUPBACH, 1999  Activation of a meiotic checkpoint regulates translation of Gurken during Drosophila oogenesis. Nat. Cell Biol. 1:354-357.[Medline]

GHABRIAL, A., R. P. RAY, and T. SCHUPBACH, 1998  okra and spindle-B encode components of the RAD52 DNA repair pathway and affect meiosis and patterning in Drosophila oogenesis. Genes Dev. 12:2711-2723.[Abstract/Free Full Text]

GOLDMAN, A. S. H. and M. LICHTEN, 1996  The efficiency of meiotic recombination between dispersed sequences in Saccharomyces cerevisiae depends upon their chromosomal location. Genetics 144:43-55.[Abstract]

HALL, J. C., 1972  Chromosome segregation influenced by two alleles of the meiotic mutant c(3)G in Drosophila melanogaster.. Genetics 71:367-400.

HILLIKER, A. J. and A. CHOVNICK, 1981  Further observations on intragenic recombination in Drosophila melanogaster.. Genet. Res. 38:281-296.[Medline]

KEE, K. and S. KEENEY, 2002  Functional interactions between SPO11 and REC102 during initiation of meiotic recombination in Saccharomyces cerevisiae.. Genetics 160:111-122.[Abstract/Free Full Text]

KING, R. C., 1970 Ovarian Development in Drosophila melanogaster. Academic Press, New York.

LANTZ, V., J. S. CHANG, J. I. HORABIN, D. BOPP, and P. SCHEDL, 1994  The Drosophila ORB RNA-binding protein is required for the formation of the egg chamber and establishment of polarity. Genes Dev. 8:598-613.[Abstract/Free Full Text]

LICHTEN, M., 2001  Meiotic recombination: breaking the genome to save it. Curr. Biol. 11:R253-R256.[Medline]

MAHADEVAIAH, S. K., J. M. TURNER, F. BAUDAT, E. P. ROGAKOU, and P. DE BOER et al., 2001  Recombinational DNA double-strand breaks in mice precede synapsis. Nat. Genet. 27:271-276.[Medline]

MANHEIM, E. A., J. K. JANG, D. DOMINIC, and K. S. MCKIM, 2002  Cytoplasmic localization and evolutionary conservation of MEI-218, a protein required for meiotic crossing over in Drosophila. Mol. Biol. Cell 13:84-95.[Abstract/Free Full Text]

MCKIM, K. S., B. L. GREEN-MARROQUIN, J. J. SEKELSKY, G. CHIN, and C. STEINBERG et al., 1998  Meiotic synapsis in the absence of recombination. Science 279:876-878.[Abstract/Free Full Text]

MYERS, E. W., G. G. SUTTON, A. L. DELCHER, I. M. DEW, and D. P. FASULO et al., 2000  A whole-genome assembly of Drosophila. Science 287:2196-2204.[Abstract/Free Full Text]

PAGE, S. L. and R. S. HAWLEY, 2001  c(3)G encodes a Drosophila synaptonemal complex protein. Genes Dev. 15:3130-3143.[Abstract/Free Full Text]

RASMUSSEN, S. W., 1975  Ultrastructural studies of meiosis in males and females of the c(3)G mutant in Drosophila melanogaster meigen. C.R. Trac. Lab. Carlsberg 40:163-173.

ROBERTS, P. A., 1969  Some components of X-ray-induced crossing over in females of Drosophila melanogaster.. Genetics 63:387-404.[Free Full Text]

ROEDER, G. S., 1997  Meiotic chromosomes: it takes two to tango. Genes Dev. 11:2600-2621.[Free Full Text]

ROGAKOU, E. P., C. BOON, C. REDON, and W. M. BONNER, 1999  Megabase chromatin domains involved in DNA double-strand breaks in vivo. J. Cell Biol. 146:905-916.[Abstract/Free Full Text]

SCHWACHA, A. and N. KLECKNER, 1995  Identification of double Holliday junctions as intermediates in meiotic recombination. Cell 83:783-791.[Medline]

SEKELSKY, J. J., K. S. MCKIM, L. MESSINA, R. L. FRENCH, and W. D. HURLEY et al., 1999  Identification of novel Drosophila meiotic genes recovered in a P-element screen. Genetics 152:529-542.[Abstract/Free Full Text]

SMITH, P. A. and R. C. KING, 1968  Genetic control of synaptinemal complexes in Drosophila melanogaster.. Genetics 60:335-351.[Free Full Text]

SPRADLING, A. C., M. DE CUEVAS, D. DRUMMOND-BARBOSA, L. KEYES, and M. LILLY et al., 1997  The Drosophila germarium: stem cells, germ line cysts, and oocytes. Cold Spring Harbor Symp. Quant. Biol. 62:25-34.[Abstract/Free Full Text]

SZOSTAK, J. W., T. L. ORR-WEAVER, R. J. ROTHSTEIN, and F. W. STAHL, 1983  The double-strand-break repair model for recombination. Cell 33:25-35.[Medline]

TAUTZ, D. and C. PFEIFLE, 1989  A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback.. Chromosoma 98:81-85.[Medline]

THORNE, L. W. and B. BYERS, 1993  Stage-specific effects of X irradiation on yeast meiosis. Genetics 134:29-42.[Abstract]

USUI, T., T. OHTA, H. OSHIUMI, J. TOMIZAWA, and H. OGAWA et al., 1998  Complex formation and functional versatility of Mre11 of budding yeast in recombination. Cell 95:705-716.[Medline]

This article has been cited by other articles:

H. Murakami and S. Keeney Regulating the formation of DNA double-strand breaks in meiosis Genes & Dev., February 1, 2008; 22(3): 286 - 292. [Full Text]