Abstract
The Saccharomyces cerevisiae MUM2 gene is essential for meiotic, but not mitotic, DNA replication and thus sporulation. Genetic interactions between MUM2 and a component of the origin recognition complex and polymerase α-primase suggest that MUM2 influences the function of the DNA replication machinery. Early meiotic gene expression is induced to a much greater extent in mum2 cells than in meiotic cells treated with the DNA synthesis inhibitor hydroxyurea. This result indicates that the mum2 meiotic arrest is downstream of the arrest induced by hydroxyurea and suggests that DNA synthesis is initiated in the mutant. Genetic analyses indicate that the recombination that occurs in mum2 mutants is dependent on the normal recombination machinery and on synaptonemal complex components and therefore is not a consequence of lesions created by incompletely replicated DNA. Both meiotic ectopic and allelic recombination are similarly reduced in the mum2 mutant, and the levels are consistent with the levels of meiosis-specific DSBs that are generated. Cytological analyses of mum2 mutants show that chromosome pairing and synapsis occur, although at reduced levels compared to wild type. Given the near-wild-type levels of meiotic gene expression, pairing, and synapsis, we suggest that the reduction in DNA replication is directly responsible for the reduced level of DSBs and meiotic recombination.
MEIOSIS is a specialized cell cycle that enables diploid organisms to reproduce sexually by generating haploid gametes through two successive divisions. At meiosis I, also known as the reductional division, homologous chromosomes disjoin from each other. At meiosis II, as at mitosis, sister chromatids separate and move to opposite poles. The meiotic divisions, in turn, are tightly linked to gamete differentiation. Fusion of gametes at fertilization restores the diploid chromosome number and initiates zygotic development. High fidelity of meiotic chromosome segregation is essential for the propagation of all sexually reproducing organisms.
The replication of chromosomes is the first detectable cytological event in meiosis. The coordinated synthesis of genomic DNA requires multiple levels of regulation and a large number of gene products. The origin recognition complex (ORC) and the multi-subunit minichromosome maintenance (MCM) complex, among others, are essential in initiating DNA replication (reviewed in Dutta and Bell 1997). Primase and the replicative polymerases, themselves multi-protein complexes, are the machines that carry out DNA synthesis (reviewed in Baker and Bell 1998). Genetic analyses in the yeast Saccharomyces cerevisiae indicate that the replicative machinery used to synthesize DNA in vegetative cells is also required for the duplication of chromosomes in meiosis (Buddet al. 1989). In contrast, the regulatory controls for these two replication modes appear distinct (Schild and Byers 1978; Hollingsworth and Sclafani 1993; Diricket al. 1998; Stuart and Wittenberg 1998).
During prophase, the duplicated homologous chromosomes align, synapse, and recombine. Alignment refers to the presynaptic association of the homologous chromosomes (Loidlet al. 1994; Weiner and Kleckner 1994), while chromosome synapsis is the intimate association of homologous chromosomes in the context of the synaptonemal complex (SC). The SC is a meiosis-specific structure that is elaborated along the lengths of the chromosomes (reviewed in von Wettsteinet al. 1984). Several genes that encode structural proteins of the SC have been identified; genetic analyses indicate that the proper assembly of the SC is essential for chromosome segregation (reviewed in Kupiecet al. 1997).
All meiotic recombination in yeast (and presumably in other organisms as well) that has been studied to date is initiated by double strand breaks (DSBs; Sunet al. 1989; Caoet al. 1990; Cervanteset al. 2000). DSBs occur before the SC is fully formed, while mature recombinants do not occur until chromosome synapsis is complete (Padmoreet al. 1991). In most organisms reciprocal exchange is essential for proper chromosome segregation at the reductional division (Hawley 1987).
Studies in S. cerevisiae have been instrumental in elucidating the requirement for chromosome pairing, synapsis, and genetic recombination in ensuring proper chromosome segregation at the meiosis I division (reviewed in Kupiecet al. 1997). However, very little attention has been devoted to the role of DNA replication in meiosis. Recent studies indicate that DNA replication is important for meiotic recombination (Bordeet al. 2000). Further, SPO11 (Klapholzet al. 1985) and REC8 (Kleinet al. 1999), two yeast genes originally identified for their roles in recombination and sister chromatid cohesion, respectively, have subsequently been shown to play distinct roles in meiotic DNA replication (Chaet al. 2000). However, the connection between DNA replication and homologous chromosome interaction remains unknown.
The yeast MUM2 gene was identified in a screen for meiotic mutants and is required specifically for meiotic DNA replication (Engebrechtet al. 1998). The Mum2 protein is not similar to other proteins; however, the C terminus is predicted to form a coiled coil. Coiled-coil domains are found in diverse classes of proteins including molecular motors and SC components. Synthetic interactions with a component of ORC and polymerase α-primase suggest that MUM2 influences the functioning of the replication machinery. Genetic analyses of mum2 mutants in meiosis indicate that perturbation of DNA replication modestly affects chromosome pairing and synapsis, while meiotic recombination is greatly impaired. Further, these studies reveal a direct correlation between the level of meiotic DSBs and DNA replication and suggest that DNA replication is an essential prerequisite to DSB formation.
MATERIALS AND METHODS
Yeast strains and media: Routine growth and manipulation of S. cerevisiae strains were performed as described (Roseet al. 1990). Yeast strains were transformed using the lithium acetate procedure (Itoet al. 1983). All integrative transformants were verified by Southern blot analysis or polymerase chain reaction (PCR). Gene disruptions were performed by one-step gene replacement; orc2 and pol1 alleles were introduced by two-step gene replacement (Rothstein 1991) using plasmids JR1267 (orc2-1) and YIp5-pol1-17 (Budd and Campbell 1987). Yeast strains used in this study are listed in Table 1.
Diploid strains harboring the arg4-bgl and arg4-nsp alleles at different locations in the genome were created from haploids Y1244 and Y1245 after targeted integration of the appropriate plasmids as described (Goldman and Lichten 1996). The arg4 alleles were inserted in the context of a pBR322-based plasmid that contains a 1.2-kb URA3 HindIII fragment inserted at the HindIII site of pBR322 and a 3.3-kb arg4 PstI fragment inserted between the BamHI and SalI sites of pBR322 (Wu and Lichten 1995). To construct homozygous arg4 insertions at HIS4 and URA3, Y1319 was transformed with pME1141 (MUM2 CEN4 TRP1; Engebrechtet al. 1998) and induced to sporulate and resultant tetrads dissected. Haploid segregants containing pME1141 and harboring both arg4 alleles were mated. Mitotic segregants that had lost the plasmid were isolated, thus generating isogenic MUM2 and mum2 strains homozygous for the arg4 alleles. sae2Δ derivatives were made by transforming pME1210 (see below) into Y1244 and Y1245 derivatives. orc and pol temperature-sensitive alleles were introduced into these same strains.
hop1::LEU2, rec104::LEU2, red1::LEU2, and spo11::TRP1 derivatives of JB128 (Bharagavaet al. 1992) were made by transforming Y898 (JB128 MUM2/mum2::LYS2) with pNH37-2 (Hollingsworth and Byers 1989), pNH131-10, pNH119-8 (Hollingsworth and Johnson 1993), and pGB324 (Girouxet al. 1989), respectively. Transformants were induced to sporulate and the resulting tetrads were dissected to generate homozygous deletions.
Plasmids: SAE2 sequences were isolated by PCR amplification using primers P68, 5′-TCC AAG CTT TTG CAC GTC-3′, and P69, 5′-GTT CCC GTG GTA GAA ATG-3′. The PCR product was digested with SacII and SalI and the resulting 1.9-kb product was inserted into the SacII and XhoI sites of Bluescript SK+ (Stratagene, La Jolla, CA) to generate pME1214. The 1.1-kb HindIII-HpaI URA3 fragment from Yep352 (Hillet al. 1986) was inserted into the HindIII-EcoRV sites of pME1214 to generate the sae2Δ::URA3 deletion allele in pME1220. Plasmid ME1220 was digested with SacI and HpaI to allow substitution of the SAE2 locus in yeast. Sequences near THR4 (YCR47c), which contain a meiotic hotspot (Wu and Lichten 1994), were isolated by PCR amplification using primers P66, 5′-GAC TAC ATG GAC ACT GAG-3′, and P67, 5′-CTA GAG TTC AAC TGC AAG G-3′. The PCR product was digested with HindIII and the resulting 0.9-kb product was inserted in the HindIII site of Bluescript SK+ to generate pME1210. The 618-bp DraI-EcoRI SPO11 fragment from pGB320 (Girouxet al. 1989) was inserted into the EcoRV and EcoRI sites of Bluescript SK+ to generate pME1845.
Analysis of growth: Strains harboring conditional alleles with and without MUM2 were grown overnight to saturation at 25°. Serial dilutions were spotted onto rich medium and incubated at the indicated temperature for 3 days. Viability was monitored by plating cells in triplicate at 25° after the indicated times at different temperatures; doubling times were calculated as described (Engebrechtet al. 1998).
Flow cytometric analysis: Cultures in log phase were split and propagated at 25°, 33°, and 36° for the indicated times and prepared for flow cytometric analysis as described (Engebrechtet al. 1998). At least 10,000 cells were analyzed per sample.
RNA isolation and analysis: Strains were cultured in YPAcetate, introduced into sporulation medium, and assessed by 4′6-diamidino-2-phenylindole (DAPI) staining as described (Engebrechtet al. 1998). Hydroxyurea (HU, 200 mm; Aldrich Chemicals, Milwaukee) was added after 2.5 hr in sporulation medium to a portion of the culture. RNA was prepared and Northern blot analysis was performed as described (Roseet al. 1995). Radioactive RNA probes were synthesized by in vitro transcription (Krieg and Melton 1987) of plasmids ME1845 (SPO11, see above) digested with HindIII using T3 polymerase (Roche Biochemicals); NH50-1 (HOP1; obtained from N. Hollingsworth) digested with BamHI; and ME817 (PYK1; Roseet al. 1995) digested with XbaI using T7 polymerase (Roche Biochemicals). Quantification was done on a Molecular Dynamics (Sunnyvale, CA) 445 SI phosphoimager using Image QuaNT software.
DSBs: DSB formation was monitored by extracting DNA from cells induced to undergo meiosis (Engebrechtet al. 1998). The resulting DNA was digested with BglII and separated on a 0.7% agarose gel. The DNA was transferred to Gene Screen (NEN) and hybridized with sequences at THR4 (Wu and Lichten 1994). The 0.9-kb HindIII fragment from pME1210 was purified and radiolabeled by random priming (New England Biolabs, Beverly, MA). Quantification was done on a Molecular Dynamics 445 SI phosphoimager using ImageQuaNT software.
Yeast strains
Recombination frequencies: Yeast strains were grown to saturation in yeast extract, peptone, adenine, dextrose (YPAD), and aliquots of triplicate cultures were plated onto YPAD medium and synthetic media lacking the appropriate amino acids to determine the mitotic prototroph frequencies.
Meiotic (return-to-growth) recombination frequencies were determined by taking 0.8 ml from each saturated culture, washing once with dH2O, resuspending the cells in 4 ml of 2% potassium acetate, and incubating them at 30° with aeration. Aliquots of the sporulating cultures were removed after 15 hr (this time corresponds to pachytene in this strain background under these sporulation conditions) for plating onto YPAD and synthetic media lacking the appropriate amino acids. The percentage sporulation was monitored after 2 days in all cultures to ensure that meiosis had progressed as expected.
Cytology: Chromosome spreads and fluorescence in situ hybridization (FISH) were performed as described (Loidl et al. 1994, 1998). Synaptonemal complex formation was analyzed by staining spreads with antibodies directed against Hop1 (Smith and Roeder 1997) and Zip1 (Symet al. 1993).
RESULTS
Genetic interactions between mum2 and components of the DNA replication machinery: MUM2 was identified in a genetic screen for mutants defective in meiosis and is required specifically for meiotic DNA replication (Engebrechtet al. 1998). MUM2 is expressed in vegetative cells (data not shown); however, no impairment of growth is observed in the absence of MUM2, suggesting that MUM2 does not play an important role in mitotic DNA replication (Engebrechtet al. 1998). We tested whether a role could be uncovered by synthetic interactions with known components of the DNA replication machinery.
Orc2 is a subunit of the origin recognition complex, which is required for initiation of DNA synthesis (Bellet al. 1993; Rosset al. 1993). Strains harboring temperature-sensitive alleles of orc2 die when shifted to the restrictive temperature due to a failure to initiate DNA synthesis (Bellet al. 1993; Rosset al. 1993; Looet al. 1995). The MCM complex is also required for the initiation of DNA replication, although, unlike ORC, these proteins are transiently associated with origins of DNA replication and move with the replication fork (Aparicioet al. 1997). The ORC and MCM multi-protein complexes (as well as a number of other proteins, reviewed in Dutta and Bell 1997) recruit polymerases to carry out DNA synthesis. DNA polymerase α-primase (pol α) synthesizes the RNA primers and begins leading and lagging strand DNA synthesis and polymerase ε and δ are required for bulk DNA replication (review in Sugino 1995). POL1, POL2, and POL3 encode components of pol α, ε, and δ, respectively.
To investigate the relationship between MUM2 and the DNA replication machinery, we examined the effect of deleting MUM2 in strains harboring temperature-sensitive alleles of genes required for initiation and synthesis of DNA. Deletion of MUM2 in orc2-1 strains improved growth (Figure 1A). The doubling time at 30° of the orc2 mutant was 162 ± 20 min, while the doubling time of the orc2 mum2 double mutant was 105 ± 12. However, deletion of MUM2 in a mcm5 temperature-sensitive strain (cdc46-1) had no effect (data not shown). In sharp contrast to the results with orc2-1, deletion of MUM2 in pol1-17 (Buddet al. 1989) strains lowered the restrictive temperature for growth (Figure 1A). This was confirmed by plating assays in which viability was more severely perturbed in the pol1-17 mum2 double mutant compared to pol1-17 at both 33° and 36° (Figure 1B). No effect on the growth of pol2-18 and pol3-14 mutants was observed in the absence of MUM2 (data not shown).
—mum2 displays genetic interactions with ORC2 and POL1. (A) Serial dilutions of saturated cultures were spotted on rich medium and incubated at the indicated temperature for 3 days. Strains Y1243 (pol1-17), Y1287 (pol1-17 mum2), Y1679 (orc2-1), and Y2032 (orc2-1 mum2) were used. Viability (B) and DNA content (C) of pol1 (shaded histogram) and pol1 mum2 (clear histogram with thick line) strains after shift to the indicated temperatures.
To determine if deletion of MUM2 influenced DNA replication in the mutant backgrounds, DNA content of yeast cells incubated at various temperatures was analyzed by flow cytometry. Log phase cultures have approximately equal numbers of cells in G1 and G2, as represented by the peaks of 2C and 4C content. At 33°, the restrictive temperature for pol1-17 mum2 strains, a greater majority of the pol1-17 mum2 cells appear to be in S phase as compared to the pol1-17 strain (Figure 1C). At 36°, both cultures showed predominantly G1 DNA content of the cells, consistent with previous analyses (Buddet al. 1989). No differences were observed in the DNA content of the orc2 and orc2 mum2 mutants (data not shown). These results implicate MUM2, directly or indirectly, in DNA replication.
Meiotic early gene expression is induced in mum2 mutants to a greater extent than in wild-type cells treated with hydroxyurea: mum2 mutants fail to complete meiotic DNA replication and consequently arrest prior to the first meiotic division (Engebrechtet al. 1998). As perturbation of DNA replication has been shown to inhibit early meiotic gene expression (Lamb and Mitchell 2001), we examined the expression level of SPO11 and HOP1, prototypical meiotic early genes (Mitchell 1994), in mum2 mutants. Northern blot analysis indicates that the steady-state level of SPO11 and HOP1 in meiosis in mum2 mutants is induced similarly to wild type (Figure 2).
In vegetative cells, the DNA synthesis inhibitor HU inhibits ribonucleotide reductase, resulting in stalled replication fork movement via a regulatory circuit that monitors nucleotide levels (Santocanale and Diffley 1998). The precise effect of HU on meiotic cells has not been delineated; however, by analogy with the mitotic block, HU-treated cells would presumably initiate meiotic DNA synthesis but arrest during fork elongation. Because the HU-induced block in meiotic DNA replication has been found to inhibit early meiotic gene expression (Lamb and Mitchell 2001), we reasoned that mum2 mutants either bypass the HU block or arrest downstream of this block. To address this question, the level of SPO11 and HOP1 mRNA present in MUM2 and mum2 cells exposed to HU during meiosis was also examined. Treatment of wild-type and mum2 cells with HU at a concentration that is known to elicit meiotic arrest (200 mm; Stuart and Wittenberg 1998) similarly inhibited the induction of the SPO11 and HOP1 genes (Figure 2). Analysis of DNA replication by cell sorting indicates that DNA synthesis does not occur in the presence of this concentration of HU (data not shown). We conclude that, in contrast to cells exposed to HU, early gene expression is not affected in the absence of MUM2 despite the block in DNA replication. In addition, the results suggest that the mum2 arrest occurs downstream of the arrest induced by HU after meiotic DNA synthesis is initiated.
The recombination observed in the mum2 mutant is dependent on the meiotic recombination machinery and components of the SC: To determine whether the recombination observed in mum2 mutants when transferred to sporulation medium is dependent on the meiotic recombination machinery or is the consequence of incompletely replicated DNA, double mutant analysis was undertaken by combining mum2 with various mutants alleles of meiosis-specific genes required for recombination and chromosome synapsis. SPO11 encodes a topoisomerase-like transesterase that is most likely responsible for catalyzing DSB formation, the initiating lesion in meiotic recombination (Bergeratet al. 1997; Keeneyet al. 1997). Accordingly, meiotic recombination is completely eliminated in spo11 mutants (Klapholzet al. 1985). REC104 also plays an essential role in the initiation of meiotic recombination, as DSB formation does not occur in rec104 mutants (Bullardet al. 1996). The induction of meiotic ectopic recombination between ura3 heteroalleles on chromosomes III and IV, which occurs at a high level in mum2 mutants (Engebrechtet al. 1998), was monitored. Deletion of either SPO11 or REC104 in mum2 strains eliminated the induction of meiotic recombination (Table 2). These findings indicate that the recombination observed in mum2 mutants is most likely initiated by the meiotic recombination machinery and is not the result of incompletely replicated DNA. Examination of other heteroallelic pairs indicates that all recombination observed in mum2 mutants is dependent on SPO11 (data not shown).
—Early meiotic gene expression in the mum2 mutant. (A) Meiotic progression of MUM2 cultures used for Northern analysis. The percentage of cells having completed the meiosis I division (▵) and both the meiosis I and II divisions (▪) are shown. (B) Northern blot analysis of HOP1, SPO11, and PYK1 (as loading control) in MUM2, MUM2 + HU (200 mm), mum2, and mum2 + HU (200 mm). The numbers below the panels represent the hours in sporulation medium. (C) Quantification of the relative expression of HOP1 in MUM2 (•) and mum2 (▪) and of SPO11 in MUM2 (○) and mum2 (□).
mum2 recombination is dependent on the meiotic recombination machinery and SC components
HOP1 and RED1 encode components of meiotic chromosomes essential for the formation of the SC (Hollingsworthet al. 1990; Rockmill and Roeder 1990; Smith and Roeder 1997). Mutations in both genes reduce, but do not eliminate, meiotic recombination, suggesting that SC morphogenesis is required to obtain full levels of meiotic recombination. To determine if the recombination observed in mum2 mutants is dependent on SC components, HOP1 and RED1 were deleted in the mum2 mutant and recombination was measured in return-to-growth experiments. As shown in Table 2, deletion of RED1 greatly reduced the recombination observed in the mum2 mutant; the double mutants had a recombination level below that of the mum2 or red1 single mutants. These results suggest that MUM2 functions in a pathway different from RED1 and both pathways are required to obtain full meiotic levels of recombination. Deletion of HOP1 resulted in a drastic reduction in the levels of recombination in this ectopic assay and introduction of the MUM2 deletion completely eliminated the induction of meiotic recombination (Table 2).
Meiotic allelic and ectopic recombination are similarly impaired in mum2 mutants: Examination of the ura3 heteroalleles arranged in an ectopic configuration (see above) and HIS4 and LEU2 heteroalleles arranged in an allelic configuration suggested that allelic recombination was more impaired than ectopic recombination in mum2 mutants (Engebrechtet al. 1998). To investigate this issue further, we introduced the arg4 alleles described by Goldman and Lichten (1996) at multiple locations in the genome. As shown in Table 3, induction of arginine prototrophy is severely decreased in the mum2 mutant, whether the alleles are in an ectopic or an allelic configuration (20- to 100-fold decrease). The same decrease is observed when the arg4 alleles are placed at the same location as the ura3 alleles (URA3 → HIS4; Table 3). To determine if the level of recombination in the mum2 mutant is influenced by the insertion being hemizygous (i.e., arg4 alleles) or homozygous (i.e., ura3 alleles), strains harboring homozygous arg4 alleles were induced in sporulation medium and assayed for recombination. As shown in Table 3, no differences were observed whether the arg4 alleles were hemizygous or homozygous. Therefore, the level of induction of recombination between the ura3 alleles in mum2 mutants cannot be accounted for by the alleles’ ectopic configuration or homozygous state (see discussion).
DSB formation in mum2 mutants occurs at reduced levels: Because mum2 mutants fail to sporulate, recombination is measured in return-to-growth experiments, which reflects both meiotic and mitotic functions within the cell. To examine the initiation of meiotic recombination, formation of DSBs was monitored physically. In wild type, DSBs are transient; therefore, to determine the total number of breaks generated during meiosis, we also analyzed DSBs in MUM2 and mum2 strains harboring the sae2/com1 mutation. sae2 mutants fail to resect DSBs, and consequently any DSBs formed are stable (McKee and Kleckner 1997; Prinzet al. 1997). These experiments indicate that mum2 mutants make detectable, albeit reduced, levels of DSBs (Figure 3). The level of DSBs observed in the mum2 mutant correlates well with the level of heteroallelic recombination observed in return-to-growth experiments (>10-fold reduction).
Allelic and ectopic meiotic recombination are uniformily reduced in mum2 mutants
We also monitored DSBs in MUM2 sae2 and mum2 sae2 cells induced for meiosis in the presence of HU. HU inhibits the formation of DSBs (Bordeet al. 2000) and mature meiotic recombinants (data not shown, Silva-Lopezet al. 1975; Simchenet al. 1976). No DSBs were detectable in the presence of HU in either culture (Figure 3). As HU also inhibits the residual DSBs in mum2 mutants, it is unlikely that mum2 uncouples DNA replication and events during prophase. Finally, this result is consistent with our analysis of meiotic early gene expression, in that the HU arrest is epistatic to the mum2 arrest, and suggests a correlation between levels of DSBs and DNA replication.
Chromosome pairing and SC formation in the mum2 mutant: To assay homolog pairing, meiotic chromosomes were surface spread and painted with composite probes for chromosome I and IV. Only nuclei showing a compact appearance using DAPI were scored (Loidlet al. 1994). Homologs were classified as paired if they were so close together that their FISH signals had fused into a single spot or if their signals were touching each other. In the wild-type cells harvested after 9 hr in sporulation medium, 85% of the homologous chromosome regions were paired. In the mum2 mutant at the same time point, 40% of the homologous chromosome regions were paired (Table 4A). After 10 hr in sporulation medium, homologous pairing in the mutant approached 64% of the level seen in wild type. Thus there is a considerable capacity to pair in the mutant.
To examine the effect of the mum2 mutation on chromosome synapsis, meiotic chromosomes were surface spread and stained with antibodies directed against Hop1, an axial element component (Hollingsworthet al. 1990), and Zip1, a central region component of the SC (Symet al. 1993). While Hop1 staining was similar in the wild-type and mutant strains, Zip1 staining in mum2 was much more diffuse and mesh-like, although significant staining was present (Table 4B; Figure 4). These results are consistent with the genetic epistasis studies indicating that SC components are present in the mum2 mutant, even though full synapsis is reduced compared to wild type.
DISCUSSION
MUM2 was identified in a genetic screen for mutants defective in meiosis and is specifically required for meiotic DNA replication (Engebrechtet al. 1998). mum2 mutants display no growth defect as measured by doubling times (Engebrechtet al. 1998); however, Mum2 is present in cycling cells, as both the MUM2 transcript and its protein product are detectable (data not shown). Thus, MUM2 is essential for meiotic DNA replication, while playing a still-undefined role in mitotic DNA synthesis. In this study, we show genetic interactions between MUM2 and components of the DNA replication machinery in vegetative cells. Interestingly, deletion of MUM2 has opposite effects on strains harboring temperature-sensitive alleles of ORC2 or POL1, improving growth of the former, but lowering the restrictive temperature of the latter. These effects may reflect a requirement for MUM2 in facilitating the coupling of DNA synthesis with its initiation steps. Given the complex interactions and multitude of proteins required to couple initiation and DNA synthesis (Dutta and Bell 1997), a specific requirement for MUM2 in this process is difficult to infer from such genetic interactions. Nonetheless, these results do suggest that MUM2 influences the functioning of the replication machinery. The differential requirements for MUM2 in mitotic vs. meiotic DNA replication emphasize that these two replication modes are distinct and provide a unique tool for investigating the requirement for DNA replication in meiotic chromosome behavior. Recent studies in fission yeast also indicate that the process of replication may be fundamentally different in mitotic and meiotic cells, as some of the components essential for initiation of mitotic DNA replication are not required for meiotic S phase (Forsburg and Hodson 2000).
—DSBs occur at reduced levels in mum2 mutants. (A) Southern blot anlysis of DSB at the THR4 hotspot (Wu and Lichten 1994). DNA was extracted from strains Y1541 (MUM2), Y1542 (mum2), Y1233 (MUM2 sae2), and Y1234 (mum2 sae2) and the latter two in the presence of HU. The numbers on the bottom refer to the hours after transfer to sporulation medium. In this experiment, ∼50% of the cells had completed meiosis I by 20 hr. P indicates the parental band and the bracket indicates diagnostic DSBs. (B) Quantification of the percentage of DSB bands in comparision to the parental fragment in MUM2 (solid bars), mum2 (black hatched bars), MUM2 sae2 (open bars), and mum2 sae2 (white hatched bars).
Comparison of the meiotic phenotypes of yeast cells deleted for MUM2 vs. cells treated with HU suggests that at least some meiotic replication is initiated in mum2 mutants. The effect on early meiotic gene expression, DSB formation, and recombination is more severe in HU-treated cells than in mum2 mutants. These results indicate that the mum2 arrest is downstream of the HU arrest. Inhibition of meiotic DNA replication by HU activates a Mec1p-dependent checkpoint (Diricket al. 1998; Stuart and Wittenberg 1998) and also prevents the induction of early genes (Figure 2; Lamb and Mitchell 2001). It is not known whether it is inhibition of DNA replication itself or the activation of the replication checkpoint that is responsible for preventing the induction of early genes in HU. While this meiotic DNA replication checkpoint is functional in mum2 mutants, it is not activated by the mum2 replication block. In fact, meiotic prophase proceeds as demonstrated by cytological analyses, which indicate that chromosome pairing and synapsis occur, although at reduced levels in the mum2 mutant. Thus, bulk DNA replication is not required for these prophase events. Analysis of the Coprinus spo22 mutant also indicates that synapsis does not depend on bulk meiotic DNA replication (Pukkilaet al. 1995; Merinoet al. 2000).
SPO11 and HOP1 represent members of a class of genes that are expressed in prophase of meiosis I (Atchesonet al. 1987; Hollingsworthet al. 1990; Mitchell 1994). The induction of SPO11 and HOP1 mRNA in mum2 mutants most likely reflects a similar induction of the whole class of early meiotic genes, which includes genes required for meiotic recombination and chromosome synapsis (Mitchell 1994). Thus the phenotype of mum2 mutants in meiosis is probably not a consequence of a defect in early gene expression.
While incompletely replicated DNA is probably present in mum2 mutants, the recombination that occurs is dependent on the meiotic recombination machinery. As SPO11 has been shown to have separable roles in the control of meiotic S-phase length and in recombination (Chaet al. 2000), the loss of recombination in mum2 mutants deleted for SPO11 may be a consequence of altered S-phase length, absence of DSBs, or both. We favor the hypothesis that deletion of SPO11 eliminates recombination by preventing DSB formation in mum2 mutants, because deletion of REC104, another meiosis-specific gene essential for the initiation of meiotic DSBs (Bullardet al. 1996), also eliminates the recombination that occurs in mum2 mutants. Taken together with the finding that the pattern of DSBs is similar in MUM2 and mum2 cells, these results indicate that the recombination observed in mum2 mutants is initiated by meiosis-specific DSBs.
Chromosome pairing and synapsis in mum2 mutants
The recombination observed in the mum2 mutant is also dependent on SC components. Deleting both RED1 and HOP1 in mum2 mutants results in levels of recombination lower than those observed in any of the single mutants, indicating that these gene products function in different pathways, both of which are required for meiotic levels of recombination. In the case of RED1, there is clearly a synergistic effect of deletion of MUM2, indicating that these pathways cooperate to ensure meiotic levels of recombination. The low levels of meiotic recombination observed in the hop1 mutant in our assay make it difficult to distinguish between a synergistic or additive relationship between these gene products.
MUM2 was identified in a screen for meiotic mutants proficient for the induction of ectopic recombination (Engebrechtet al. 1998). High levels of meiotic recombination were observed in return-to-growth experiments with the ectopic heteroallele pair used, but not allelic recombination (Table 2 and 3; Engebrechtet al. 1998). In this study, we show that the reduction of recombination in return-to-growth assays in the mum2 mutant is not influenced by the location of the particular heteroallele pair being measured or by whether the introduced alleles are hemizygous or homozygous (Table 3). Further, the differences are not due to strain backgrounds (J. Engebrecht, unpublished results). There are three possible explanations for these findings. First, the level of recombination between the ura3 heteroalleles is low and hence deletion of MUM2 does not have much effect. Second, different heteroalleles are repaired differently in mum2 mutants. Third, insertion of different pairs of heteroalleles at a particular locus alters an important sequence or chromatin structure such that recombination is affected differently in the mum2 mutant.
Meiosis-specific DSBs are greatly reduced in mum2 mutants, although not eliminated as they are in the presence of HU. The failure to form DSBs in HU-treated cells is probably due to the absence of Spo11p (as well as other meiosis-specific gene products required for DSB formation). Because Spo11 is present and functional in mum2 mutants (Figure 2, Table 2), the reduction in DSBs may be a consequence of the defect in DNA replication. The direct coupling of DSB formation with DNA synthesis has been suggested on the basis of the observation that a delay in replication results in a similar delay in DSB formation (Bordeet al. 2000). In addition, the finding that SPO11 plays a role in the regulation of S phase in addition to its role in DSB formation (Chaet al. 2000) suggests that DNA replication and DSB formation are interrelated events. Determining the molecular requirement for DNA replication in DSB formation will be facilitated by analysis of Mum2p and elucidation of its unique role in meiotic DNA replication.
—SC formation in mum2 mutants. Surface spread chromosomes from either wild type (MUM2/MUM2) or mutant (mum2/mum2) were labeled with antibodies directed against (A) Hop1 (Smith and Roeder 1997) or (B) Zip1 (Symet al. 1993).
Acknowledgments
We are grateful to Jim Haber, Nancy Hollingsworth, Janet Leatherwood, and Carol Newlon for strains, plasmids, and thoughtful discussions. Chun Zhou and the FACS lab at University Hospital provided excellent technical assistance. We also thank Nancy Hollingsworth for comments on this manuscript. This work was supported by grants from the National Science Foundation (MCB-9870780) and in part by research grant no. 1-FY99-573 from the March of Dimes Birth Defects Foundation to J.E., from the National Institutes of Health (GM-28220) to R.S., and from the Austrian Science Fund (grant no. S8202) to J.L.
Footnotes
-
Communicating editor: M. Lichten
- Received July 27, 2000.
- Accepted December 7, 2000.
- Copyright © 2001 by the Genetics Society of America