Introduction

Topoisomerases play a crucial role in cellular metabolism (for a review see Duguet, 1997), and variations in topoisomerase expression levels result in pleiotropic phenotypes in bacteria and yeast (Kim and Wang, 1989; Wallis et al., 1989; Drlica, 1990). DNA topoisomerases are essential for the replication of DNA molecules (reviewed in Wang, 1996), for chromosome condensation (Adachi et al., 1991; Downes et al., 1994; Castano et al., 1996) and for the proper segregation of chromosomes at mitosis and meiosis (DiNardo et al., 1984; Holm et al., 1985, 1989; Rose et al., 1990). Another fundamental role for topoisomerases involves the relaxation of positive and negative superhelical domains of DNA that are generated by transcription (Liu and Wang, 1987; Giaever and Wang, 1988; Wu et al., 1988; Tsao et al., 1989). Additionally, the regulation of gene expression for many functions is dependent on the degree of supercoiling found in intracellular DNA (Gangloff et al., 1994a; Wang and Droge, 1996). Recently, a role for topoisomerases I and II in checkpoint control was proposed (Downes et al., 1994; Castano et al., 1996).

In Saccharomyces cerevisiae, one mitochondrial and three nuclear topoisomerase activities have been described. Among the nuclear activities, DNA topoisomerases I and II (encoded by the TOP1 and TOP2 genes, respectively) are capable of relaxing both negatively and positively supercoiled DNA molecules (Goto and Wang, 1984; Thrash et al., 1984), while DNA topoisomerase III (encoded by the TOP3 gene) weakly relaxes only negatively supercoiled DNA (Kim and Wang, 1992). Little is known about the recently identified mitochondrial activity (Ezekiel et al., 1994). The genes encoding the nuclear activities (TOP1, TOP2 and TOP3) have been cloned (Goto and Wang, 1984, 1985; Thrash et al., 1985; Wallis et al., 1989) and studied extensively (for reviews see Champoux, 1994; Gangloff et al., 1994a; Hsieh et al., 1994; Lima and Mondragon, 1994; Watt and Hickson, 1994; Roca, 1995; Wang, 1996).

The TOP3 gene is unique in that it encodes a type I-5' topoisomerase in yeast. It is homologous to the Escherichia coli topA and topB genes but not to the S.cerevisiae TOP1 (I-3') or TOP2 (II) genes. The relationships originally proposed on the basis of sequence comparison have been confirmed biochemically (Kim and Wang, 1992). Purified Top3 protein exhibits many of the properties of TopA and TopB, including relaxation of only negatively supercoiled DNA and a 5'-covalent phosphotyrosine ester linkage between the conserved tyrosine in the active site and the DNA. The enzyme is also proficient in preferentially binding single-stranded DNA and in decatenating single strands (Kim and Wang, 1992). Recently, homologs in man and mouse have been identified, and the gene has been shown to be essential during early mouse embryogenesis (Hanai et al., 1996; Fritz et al., 1997; Li and Wang, 1998; Seki et al., 1998).

The top3 mutants originally were isolated because they stimulate recombination between repeated sequences (Wallis et al., 1989; Arthur, 1991; Gangloff et al., 1996). This hyper-recombinogenic phenotype is not restricted to direct repeats, since the absence of Top3 also elevates recombination between homologous ectopic genes (Bailis et al., 1992). In addition to the recombination phenotype, top3 mutants exhibit a cell cycle aberration characterized by an accumulation of cells containing an undivided nucleus in the neck of the bud. This cell cycle delay translates into slow growth that is believed to result from a defect in single-stranded DNA decatenation in an alternative pathway for DNA replication termination (Gangloff et al., 1994a). Among other mitotic phenotypes, top3 mutants are affected in the transcriptional control of several genes (Arthur, 1991).

A search for suppressors of the top3 slow growth phenotype led to the isolation of mutations in the SGS1 gene (Gangloff et al., 1994b). The Sgs1 protein is a member of the helicase family characterized by the E.coli RecQ protein (Nakayama et al., 1985; Umezu et al., 1990). Over the past few years, Sgs1 homologs have been identified in many organisms, including yeast and man (Puranam and Blackshear, 1994; Ellis et al., 1995; Puranam et al., 1995; Yu et al., 1996; Stewart et al., 1997; Yan et al., 1998). The absence of Sgs1 suppresses all the mitotic phenotypes caused by the lack of Top3 (Gangloff et al., 1994b), which led to the hypothesis that Top3 is required to act on substrates created by Sgs1. This model is supported by the finding that Sgs1 interacts physically with Top3 in a two-hybrid system, suggesting that Top3 and Sgs1 may work together as part of a complex (Gangloff et al., 1994b).

Although Top3 is involved in every aspect of DNA metabolism (Wallis et al., 1989; Arthur, 1991; Bailis et al., 1992; Gangloff et al., 1994a, 1996; Kim et al., 1995), its mitotic function is dispensable. In contrast, Top3 function is essential during meiosis, and homozygous mutant diploids are unable to form asci. In this report, we investigate the metabolic alteration responsible for the inability of top3 strains to sporulate. We show that the top3 mutants fail to complete reductional division at a step following initiation of recombination and formation of recombinant molecules. Additionally, overexpression of the E.coli topA gene restores sporulation, which suggests that a type I-5' topoisomerase activity is essential for processing recombination intermediates generated during meiosis.

Results

The absence of sporulation in top3 mutants is not due to loss of chromosome III

Diploid cells homozygous for the top3 deletion fail to sporulate (Wallis et al., 1989; Arthur, 1991; Gangloff et al., 1994a). Light microscopy analysis of 2000 diploid cells homozygous for top3 indicated that, even 7 days after transfer to sporulation medium, no visible asci could be detected. We previously have observed that diploid cells homozygous for top3, at a much higher frequency than wild-type cells, become capable of mating with haploid cells and form colonies on selective medium, an event due to chromosome III loss (Arthur, 1991). In yeast, each chromosome III carries a different MAT allele at the mating type locus, and the products of both are required for entry into meiosis. A high frequency of chromosome loss may therefore lead to top3's inability to sporulate. To test this hypothesis, we measured the frequency at which chromosome III is lost in a population of cells ready to enter meiosis. We used a strain (D5-4DW1193-2C; Table I) heterozygous for LEU2, a proximal gene on chromosome III. At the t = 0 of meiosis induction, diploid cells were plated out onto rich medium. After 3 days of vegetative growth, the colonies were replicated onto synthetic medium lacking leucine. In those conditions, cells that have lost the LEU2-bearing chromosome III prior to the first cell division on the plate become unable to form colonies. This analysis was performed on several thousand colonies and revealed that <1/3000 diploid cells had lost the LEU2 marked chromosome III. This indicates that the total frequency of chromosome III loss in top3 mutants is <1/1500 cells, and thus cannot account for the inability of top3 mutants to sporulate.

The absence of TOP3 prevents the cells from undergoing reductional division

To determine at which stage in meiosis top3 cells are blocked, we next examined the fate of the nucleus of top3 cells during meiosis. Meiosis was induced, and samples of wild-type and mutant cells were taken at different times, washed, fixed and incubated with 4',6-diamidino-2-phenylindole (DAPI). Analysis by fluorescence microscopy revealed that after 12 h, 5–10% of wild-type cells have already completed both reductional and equational division, and formed four haploid staining bodies. However, in a sample of 2000 top3 cells, reductional division was never observed. Instead, after 2 days of incubation in sporulation medium, nuclear DAPI-stained material fragmentation is observed with subsequent loss of the intense fluorescent staining of the DNA (Figure 1).

Double-strand breaks (DSBs) are present in top3 mutants

In S.cerevisiae, double-strand breaks (DSBs) initiate meiotic recombination, a process necessary to achieve proper chromosome segregation (for reviews see Kleckner, 1996; Roeder, 1997). To investigate whether top3 homozygous diploids enter meiosis, the occurrence of DSBs at time t = 0, t = 8 and t = 24 h in sporulation medium was analyzed (de Massy and Nicolas, 1993; de Massy et al., 1994). Interestingly, DSBs at both the CYS3 and the ARG4 loci were detected. The levels of breaks as well as their kinetics of appearance and processing are comparable with those observed in wild-type controls (Figure 2). Global analysis of DSB formation along the 340 kb of chromosome III indicated that the pattern of DSBs formed on this chromosome is similar in the top3 mutants and the wild-type controls, suggesting that the results observed at the ARG4 and CYS3 loci reflect a general trend in the genome (data not shown).

Recombinant molecules are formed, but generate a lethal substrate

To analyze the progression of recombination after the introduction of a DSB, we examined the formation of recombinant molecules by using arg4 heteroalleles that differ from each other by their restriction pattern (Figure 3), The presence of recombinant molecules does not, however, imply that Holliday junctions are resolved actively. In the arg4-EcoRV/arg4-BglII diploid cells, the great majority of the convertants at the ARG4 locus involve only the EcoRV restriction site which lies close to the upstream region of the gene where the DSBs occur. After a double digestion with EcoRV and BglII, and in the absence of a pre-treatment with DNA cross-linking agents, passive migration of the junction releases linear DNA fragments. Using this experimental design, the two parental fragments were detected. However, two additional signals corresponding to arg4 recombinant molecules were also detected by Southern analysis in both wild-type and top3 mutants (see 1 and 7 kb fragments, Figure 3). Together, these results and those presented in the previous section indicate that top3 cells are progressing through meiosis and that the Top3 function is not required for the formation of either DSBs or recombinant molecules.

In yeast cells, meiotic levels of recombination can be induced before commitment to reductional division in a return to growth (RTG) experiment (Sherman and Roman, 1963). In top3 mutants, although DSBs and recombinant molecules are observed, no stimulation of recombination could be measured through the formation of Arg⁺ prototrophs. At t = 0, we recovered four and 20 Arg⁺ clones for 10⁶ cells in wild-type and top3 mutants, respectively. After 24 h in the sporulation medium, 17 000 Arg⁺ recombinants were recovered for 10⁶ wild-type cells, compared with only 25 for top3 mutants. We interpret this result as a failure of top3 cells to resume vegetative growth following meiotic recombination initiation. This is not due to a general deficiency in DSB repair since mitotic top3 mutants switch mating type in the presence of the HO endonuclease (S.Gangloff, unpublished result).

Bypassing recombination between homologs restores sporulation

To determine whether the top3 arrest in meiosis I is related to recombination, we examined whether cells that do not undergo meiotic recombination are capable of sporulating in the absence of Top3. The spo11 mutation abolishes meiotic DSB formation (Klapholz et al., 1985; Bergerat et al., 1997; Keeney et al., 1997) and, thus, recombination. Since spo11 mutants do not sporulate efficiently and produce mostly inviable spores due to non-disjunction, we tested our hypothesis in strains homozygous for both the spo11 and the spo13 mutations. spo13 mutants bypass meiosis I and produce asci containing two diploid spores (dyads) that are viable even in the absence of recombination (Malone and Esposito, 1981; Klapholz et al., 1985). Isogenic diploid strains were constructed in which the spo11, spo13 and top3 mutations are homozygous (Table I). These diploid cells were analyzed for their ability to undergo meiosis and form spores. The results summarized in Table II reveal that TOP3 is not required to form viable spores in the absence of recombination: while the top3 spo13 cells form no asci, the triple mutant top3 spo11 spo13 sporulates and forms dyads with a viability similar to that of spo11 spo13 controls. We conclude that the meiotic arrest of top3 spo13 cells depends on the activity of Spo11 and, therefore, on meiotic recombination. This result suggests that resolution of entangled homologs after recombination is probably responsible for the sporulation defect observed in top3 mutants. This hypothesis is also in agreement with the observation that nuclear division occurs within the same time frame in the presence or absence of TOP3 when recombination is abolished. As expected, two DAPI-staining bodies could be observed in <24 h following meiotic induction in the top3 spo11 spo13 triple mutant background, while no nuclear division was detected in >2000 cells after 2 days (Figure 4) or even after 6 days of incubation (data not shown) when the SPO11 controlled initiation of recombination is functional. Furthermore, the timing of dyad formation is independent of TOP3, which also suggests that the absence of this topoisomerase does not greatly affect pre-meiotic DNA synthesis.

Overexpression of TopA suppresses the sporulation deficiency of top3 mutants

Wallis et al. (1989) have shown that expression in yeast of the E.coli topA gene, encoding a functionally related type I-5' topoisomerase, is capable of complementing the slow growth phenotype of top3 null mutants. Thus, we tested whether expression of topA (YEptopA-PGAL1) would also complement the sporulation defect of top3 mutant strains. Diploid cells homozygous for the top3 deletion and containing the topA plasmid were tested for spore formation. The diploid strains were pre-grown on galactose medium to induce the expression of topA and then replica-plated to sporulation medium containing 0.1% galactose. Under these conditions, three- and four-spored asci accounted for 20% of the cells present after 3 days of induction at 30°C compared with 50% for wild-type diploid controls transformed by YEptopA-PGAL1 plasmid (Table III). When four-spored asci from the top3/top3 YEptopA-PGAL1 diploids were dissected on either YPD or YPGal, the spore viability was poor. From a total of 71 tetrads dissected, only 110 viable colonies (39%) were recovered, while 89% viability was observed for isogenic wild-type strains containing the topA plasmid. Of the 71 top3/top3 YEptopA-PGAL1 tetrads, only five yielded four viable spores. Among tetrads where only three spores form colonies, some spores were found to grow on medium lacking the two amino acids corresponding to the auxotrophic markers used to disrupt the two copies of the TOP3 gene. This implies that the two chromosomes bearing the top3 deletion did not disjoin and segregated to the same cell. These observations indicate that overexpression of a type I-5' topoisomerase activity can restore sporulation in cells lacking Top3. However, the efficiency of the process is poor, and chromosome non-disjunction may be important in spore inviability. This result suggests that the essential enzymatic activity necessary for performing meiosis is shared with type I-5' enzymes, but absent from the type I-3' or II counterparts.

The absence of SGS1 allows some sporulation in top3 mutants

The SGS1 gene encodes a protein containing a DNA helicase domain conserved from bacteria to man (Rothstein and Gangloff, 1995). Mutations in either BLM or WRN, two human genes encoding proteins structurally and functionally homologous to Sgs1, lead to genome instability and a high incidence of cancer (Ellis et al., 1995; Yu et al., 1996). We have shown previously that Sgs1 interacts with Top3 in a two-hybrid system (Gangloff et al., 1994b). Interestingly, TOP3 and SGS1 also interact genetically, since the absence of SGS1 suppresses the slow growth, the altered cell cycle distribution and the hyper-recombination phenotype of top3 mutants (Gangloff et al., 1994b). Because of the potential for Sgs1 and Top3 to act together in a complex (Gangloff et al., 1994b), we analyzed the possibility that unprocessed structures introduced by Sgs1 in the absence of Top3 are responsible for the sporulation defect. As shown in Table II, top3 sgs1 double mutants sporulate very poorly, but produce viable spores. Out of 18 top3 sgs1 tetrads dissected, the overall spore viability was 67%, with five, six, three and four tetrads giving rise to four, three, two and one viable spore, respectively. We found that, unlike top3 strains, sgs1 mutant strains complete the meiotic process, but with a long delay in the appearance of four-spored asci and a very low yield (Figure 5). Out of 33 sgs1 tetrads dissected, the overall spore viability was 74%, with 14, eight, seven and four tetrads giving rise to four, three, two and one viable spore, respectively, indicating that sgs1 is epistatic to top3 with respect to spore viability.

We next examined whether the 48 h delay in the meiotic process is related, similarly to the arrest of top3 mutants, to the recombination process. We therefore constructed sgs1 spo11 spo13 triple mutant diploid cells and monitored both nuclear division and spore (dyad) formation. In the spo11 spo13 background, the kinetics of spore formation and the timing of equational division are independent of the presence of SGS1, which indicates that Sgs1 is also involved in the meiotic recombination process (Figure 6).

Discussion

The absence of the type I-5' topoisomerase Top3 results in a pleiotropic mitotic phenotype (Wallis et al., 1989; Gangloff et al., 1994a,b, 1996). Vegetative cells are viable but meiosis does not proceed. This indicates that Top3 plays either a specific and essential role in meiosis or that another protein can partially substitute for Top3 during the mitotic but not the meiotic cycle. The experiments described in this study were designed to address the meiotic role of Top3. Since the Sgs1 helicase interacts with Top3, and since sgs1 is epistatic to top3 with respect to mitotic top3 phenotypes, we have extended this study to the sgs1 and sgs1 top3 mutants.

Aberrant gene expression is not a likely cause for the sporulation defect

TOP3 is known to control the expression of several genes (Arthur, 1991). IME1 and/or IME2, which belong to the cascade of regulatory genes involved in the control of entry into meiosis (Sia and Mitchell, 1995), could depend on TOP3 for their expression. This is not the case since overexpression of these genes was not found to alleviate the sporulation defect of top3 cells (unpublished results). Although it cannot be formally excluded that TOP3 controls the expression of other meiosis-specific genes, altered gene expression is not a probable candidate for the sporulation defect since spo11 spo13 top3 triple mutants do sporulate similarly to spo11 spo13 strains.

Recombination is responsible for the top3 meiotic defect

Analysis of top3 cells at the cellular level indicates that very few, if any, cells are capable of completing the first meiotic division (MI). Since this is the part of meiosis where recombination takes place, several aspects of this process were analyzed. Initiation, as evidenced by formation of DSBs, occurs at a similar frequency in both normal and top3 cells. Additionally, comparable amounts of recombinant molecules could be detected in wild-type and mutants, suggesting that TOP3 is not involved in the formation of these recombinant molecules (Figure 3).

Although detection of intragenic recombinant molecules does not require that the Holliday junctions are resolved, it is still possible that Top3 may affect the recombination process itself. It has been shown previously that meiotic recombinant molecules can be detected by genomic blot analysis in mutants such as rad51, rad52, rad55 and rad57 that are deficient in homologous recombination (Borts et al., 1986; Shinohara et al., 1992; Schwacha and Kleckner, 1997). However, in these studies, the level of recombinant molecules was much lower than in control cells, which is clearly not the case for top3.

Cells mutant for TOP3 appear to proceed into an irreversible phase leading to cell death (Figure 1). In addition, RTG analysis did not indicate any increase in the recovery of Arg⁺ recombinants with top3, suggesting a viability loss of the meiotic cells following DSBs. Finally, we used spo11, an early meiotic mutation that prevents DSB formation (Cao et al., 1990), to correlate the top3 defect directly with recombination. Viable diploid spores are obtained if spo11 is coupled to spo13, a mutation that leads to a bypass of the reductional division. While spo13 top3 cells are blocked before MI, the spo11 spo13 top3 cells do sporulate, with kinetics of nuclear division similar to the kinetics in spo11 spo13 cells. Furthermore, viable diploid spores were recovered, indicating that the meiotic phenotype of top3 mutants is due to recombination.

Does the top3 defect trigger a checkpoint response?

It is still not clear whether the absence of MI division in top3 meiotic cells relates to a checkpoint activity. Indeed, after a transient delay in prophase, the chromatin appears to be fragmented and the cells die. It is possible, therefore, that some kind of DNA degradation process is activated after a time of arrest with unresolved recombination structures. We tested a possible effect of rad17 and rad24, mutations known to allow MI to occur in mutants such as dmc1 or rad51, affected in intermediate recombination steps (Lydall and Weinert, 1995; Lydall et al., 1996). These mutations had no effect on top3 cells (unpublished results), in agreement with the proposal that the arrested cells do not contain single-stranded DNA that would activate this checkpoint control. It will be of interest, therefore, to determine if red1 or mek1 mutation allows MI division to take place in top3 cells. These genes have been proposed to monitor the completion of recombination, and to arrest cells until resolution of the recombination structures (Xu et al., 1997).

The sgs1 mutant also displays a meiotic defect related to recombination

The relationship between top3 and sgs1 prompted us to question if sgs1 diploids display a meiotic defect related to the processing of recombination events.

We found that sgs1 decreases sporulation efficiency by approximately two-thirds, reduces spore viability (Table II) and exhibits a slower progression through MI (Figure 5). However, the rate of intragenic recombination among viable spores studied at different loci is unaffected (data not shown) and confirms previously reported results (Watt et al., 1995, 1996). The MI division is also delayed by 48 h, indicating that the cells are held up transiently in prophase. This delay was likewise seen in a spo13 sgs1 double mutant, but not in spo11 spo13 sgs1 cells, indicating that the effect is due to recombination (Figure 6). It is conceivable that a recombination checkpoint is activated in sgs1 cells, leading to the arrest of the majority of the meiotic cells, while delaying progression through meiosis of the unarrested cells.

Top3 and DNA replication

Although Top3 is dispensable for vegetative growth, its function is required for the completion of MI. Previous work has shown that mitotic and meiotic DNA replication initiate at the same origins and terminate non-specifically (Collins and Newlon, 1994). The absence of Top3 in mitosis leads to the accumulation of cells with an undivided nucleus in the neck of an elongated bud (Gangloff et al., 1994b). The associated cell cycle delay at the G₂–M transition is consistent with an impediment of chromosome segregation, suggesting that Top3 is involved directly in the separation of replicated chromatids or, alternatively, in the resolution of topological structures resulting from either recombination during replication or collapsed replication forks (Young et al., 1984; Kawasaki et al., 1994; Zou and Rothstein, 1997). Because no noticeable difference in the timing of diploid spore formation could be detected when meiotic levels of recombination are abolished, we do not believe it likely that Top3 plays a meiotic-specific role in the DNA synthesis process.

Topoisomerase activities and recombination

Of the three nuclear DNA topoisomerases characterized in yeast, the activity of Top2 is essential in mitosis and meiosis, while that of Top1 is dispensable in both (Holm et al., 1985, 1989; Rose et al., 1990; Wang, 1996). On the other hand, Top3 is non-essential during vegetative growth but cells mutant for the Top3 topoisomerase do not achieve the first meiotic division. This result indicates that Top3 performs a function in meiosis that cannot be replaced by Top1 or Top2. Different models can account for this observation.

Top3 may exhibit a substrate specificity or may perform a specific biochemical activity. During the early steps of the meiotic process, DNA replication and recombination take place (reviewed in Kleckner, 1996; Roeder, 1997). The observation that MI does not take place in the absence of Top3 when recombination is proficient suggests that physically connected homologs cannot be disentangled efficiently. It has been shown that meiotic recombination intermediates evolve to form double Holliday junctions that need to be resolved to complete the meiotic process (Schwacha and Kleckner, 1994). Failure to resolve the junctions prevents the cells from executing a correct anaphase I, and it has been postulated that a type I topoisomerase could serve as a backup mechanism to process persisting double Holliday junctions or derived hemicatenane structures in which the DNA duplexes are intertwined at a single point (Schwacha and Kleckner, 1994; Gilbertson and Stahl, 1996; Duguet, 1997). Top3 could fulfill such a function. Sequence analysis and biochemical studies have shown that Top3 resembles its bacterial type I-5' homologs TopA or TopB, in that it binds preferentially single-stranded DNA and is proficient in decatenating DNA molecules (Wallis et al., 1989; Kim and Wang, 1992). In the context of a double Holliday junction, unwound single-stranded DNA is available for Top3 to bind to and to untwine. Prokaryotic and eukaryotic enzymes are all capable of relaxing negatively supercoiled DNA, but only type I-5' enzymes exhibit an extremely strong preference for single-stranded DNA (for review see Wang, 1996). When bound to ssDNA, type I-5' enzymes become very potent decatenases. Support for a specific topoisomerase type I-5' activity in the resolution of recombination intermediates derives from the observation that ectopic expression of the E.coli topA gene allows top3 mutants to form viable meiotic products. An alternative to the previous models relies on interactions with proteins capable of specifically targeting the topoisomerase to its substrate. The Top3 topoisomerase has been shown to associate physically with the Sgs1 helicase (Gangloff et al., 1994b). It is therefore conceivable that Sgs1 or another protein could deliver the activity of Top3 to a region of the genome that is not accessible to Top1 or Top2.

Top3 and Top2

Previous work that focused on the role of Top2 in the meiotic process indicates that top2 mutants fail to undergo both reductional division and the mitosis-like equational division (Holm et al., 1985; Rose et al., 1990). The reductional defect is caused by recombination events, suggesting that Top2 is required to resolve intertwined sister chromatids distal to the site of recombination (Rose et al., 1990).

Two hypotheses can explain the requirement for Top3 in MI. Schwacha and Kleckner (1995) echoed earlier suggestions that topoisomerases are involved in the resolution of Holliday junctions, and it is therefore possible that Top2 and Top3 are involved in this process. During vegetative growth, a limited number of recombination events involving homologous chromosomes occur (Fabre et al., 1984; Kadyk and Hartwell, 1992), which largely limits the role of Top2 to disentangling replicated chromatids (Holm et al., 1985) and that of Top3 to participating in replication termination (Gangloff et al., 1994b). In meiosis, however, connections between homologous chromosomes are stimulated greatly (for a review see Petes et al., 1991). Top2 activity may, therefore, be required, not only to decatenate replicated chromatids to allow recombining chromosomes to segregate, but also directly to resolve recombination intermediates. If Top2 is limiting, it is possible that Top3's function may become essential for resolving recombination intermediates in MI. In the second alternative, Top2 is not involved primarily in the resolution of recombinant structures and Top3 may represent a major activity capable of resolving recombination intermediates involving homologous chromosomes, a structure that is rare in vegetative cells.

Top3 and Sgs1

We have shown previously that the Sgs1 helicase interacts genetically and physically with the Top3 topoisomerase (Gangloff et al., 1994b). Epistasis analysis has suggested that the Sgs1 helicase creates the structures that Top3 acts on. We have proposed, by analogy with the bacterial TopB activity with which Top3 shares many features (Kim and Wang, 1992), that Top3 and Sgs1 may play an important role in an alternative pathway of DNA replication termination. The absence of Sgs1 fully suppresses all the mitotic phenotypes of top3 mutants, and meiosis is the first instance where the activities of the complex can be uncoupled. The tetrad yield in the top3 sgs1 double mutant is extremely low (1–2%), compared with that found in sgs1 mutants (30%), although sgs1 viability is retained. This suggests that, in meiosis, Top3 can act on substrates that are not created by the Sgs1 helicase. This view is in agreement with the idea that Top3 and Sgs1 may work together in replication termination (both in mitosis and in meiosis), and with the hypothesis involving Top3 alone in the resolution of some meiotic double Holliday junctions.

Human genes and perspectives

Finally, it is interesting to note that TOP3 and SGS1 are present in humans. Mutations in BLM or WRN, the SGS1 homologs, are responsible for cancer-prone disorders characterized by genomic instability, a phenotype shared with sgs1 mutants (Ellis et al., 1995; Rothstein and Gangloff, 1995; Watt et al., 1995; Yu et al., 1996). Two human genes homologous to TOP3 have also been identified (Hanai et al., 1996; Kawasaki et al., 1997). Disruption of TOP3 in mouse revealed that this gene is essential during early embryogenesis (Li and Wang, 1998). It is therefore tantalizing to speculate that mammalian Top3 may both be involved in the resolution of recombination intermediates present during early development, and play a more general role in replication termination in association with Blm or Wrn proteins.

Materials and methods

Strains and genetic manipulation methods

Standard yeast genetic methods were employed for analysis of strains and crosses (Sherman et al., 1986). All strains used in this study are isogenic to W303 (Thomas and Rothstein, 1989), except for ORD2130, 2184 and 2186, and are listed in Table I. The rad17::LEU2 and rad24::TRP1 mutants were kindly provided by T.Weinert, the YEPtopA-PGAL1 plasmid is a gift from J.C.Wang, and the spo11 and spo13 strains were a kind gift from R.E.Esposito.

Monitoring nuclear divisions

Meiotic cultures were stained with DAPI and examined by fluorescence microscopy for one, two (MI) and four (MII) staining bodies. In meiotic time course experiments, 1 ml aliquots from the sporulation medium were taken at 2 h intervals for the first 12 h, then every day for the next 6 days. The cells were resuspended in 0.5 ml of methanol:glacial acetic acid (3:1 v/v) and fixed for 1–24 h at 4°C. Following fixation, cells were washed and resuspended in distilled water, and stored at 4°C until the time of examination. Then 10⁸ cells were resuspended in 50 l of distilled water, and 1 ml of ethanol and 1 l of the DAPI (10 mg/ml) solution were added. After 5 min of incubation at room temperature, the cells were washed several times with distilled water and resuspended at 10⁸ cells/ml for microscopy (Sherman et al., 1986).

Commitment to meiosis

Cells were grown to saturation overnight in YPD (Sherman et al., 1986), diluted into SPS (yeast extract 0.5%, bactopeptone 1%, yeast nitrogen base 0.17%, potassium acetate 1%, ammonium sulfate 0.5%, potassium phthalate 50 mM pH 5.0) and allowed to grow until they reached the stationary phase. At this time, the cells were washed with distilled water and resuspended at a concentration of 10⁷ cells/ml in sporulation medium (1% potassium acetate) and incubated at 28°C with vigorous agitation. The proper meiotic characteristics of the strains were analyzed systematically by dissection of the tetrads (or dyads) followed by a test for viability and segregation of the available markers.

Double-strand break analyses

For DSB detection, the rad50S mutation commonly is used because it prevents the metabolism of the broken ends and thus leads to DSB accumulation (Cao et al., 1990). We could not use this mutation because top3 rad50S double mutants are extremely sick. ARG4 and CYS3 DSBs were analyzed using the standard techniques previously described (de Massy and Nicolas, 1993). Recombinant molecules present at the ARG4 locus were analyzed quantitatively on a Molecular Dynamics PhosphorImager with the ImageQuant v1.1 software. Distribution of meiotic DSBs on yeast chromosome III was established using the procedure described by Baudat and Nicolas (1997).

Return to growth experiment

Commitment of the wild-type (ORD2130), TOP3. top3 (ORD2186) and top3/top3 (ORD2184) strains to meiotic recombination is followed throughout sporulation. After transfer to sporulation medium, aliquots were taken at different times (Sherman and Roman, 1963). The cells were plated after appropriate dilutions on YPD plates and on plates lacking arginine, allowing determination of the frequency of Arg⁺ prototroph recombinants per viable cell.

Acknowledgements

Many thanks to Eric Coïc, Christine Soustelle and Julie Smith for critically reading the manuscript. This work was supported by an NIH grant (GM55628) to R.R and by the Centre National de la Recherche Scientifique, the Institut Curie and the Commissariat à l'Energie Atomique to F.F. S.G. was supported by fellowships from La Fondation pour la Recherche Médicale, l'Association pour la Recherche sur le Cancer and La Ligue Nationale contre le Cancer.

References

Adachi Y, Luke M and Laemmli UK (1991) Chromosome assembly in vitro: topoisomerase II is required for condensation. Cell, 64, 137–148. | PubMed | ISI | ChemPort |

Arthur L (1991) Characterization of a novel eukaryotic topoisomerase (TOP3) in Saccharomyces cerevisiae that affects recombination and gene expression. PhD thesis, Columbia University, New York, NY.

Bailis AM, Arthur L and Rothstein R (1992) Genome rearrangement in top3 mutants of Saccharomyces cerevisiae requires a functional RAD1 excision repair gene. Mol Cell Biol, 12, 4988–4993. | PubMed | ChemPort |

Baudat F and Nicolas A (1997) Clustering of meiotic double-strand breaks on yeast chromosome III. Proc Natl Acad Sci USA, 94, 5213–5218. | Article | PubMed | ChemPort |

Bergerat A, de Massy B, Gadelle D, Varoutas PC, Nicolas A and Forterre P (1997) An atypical topoisomerase II from Archaea with implications for meiotic recombination. Nature, 386, 414–417. | Article | PubMed | ISI | ChemPort |

Borts RH, Lichten M and Haber JE (1986) Analysis of meiosis-defective mutations in yeast by physical monitoring of recombination. Genetics, 113, 551–567. | PubMed | ChemPort |

Cao L, Alani E and Kleckner N (1990) A pathway for generation and processing of double-strand breaks during meiotic recombination in S.cerevisiae. Cell, 61, 1089–1101. | PubMed | ISI | ChemPort |

Castano IB, Brzoska PM, Sadoff BU, Chen HY and Christman MF (1996) Mitotic chromosome condensation in the rDNA requires TRF4 and DNA topoisomerase I in Saccharomyces cerevisiae. Genes Dev, 10, 2564–2576. | PubMed | ISI | ChemPort |

Champoux JJ (1994) Mechanism of catalysis by eukaryotic DNA topoisomerase I. Adv Pharmacol, 29, 71–82.

Collins I and Newlon CS (1994) Chromosomal DNA replication initiates at the same origins in meiosis and mitosis. Mol Cell Biol, 14, 3524–3534. | PubMed | ISI | ChemPort |

de Massy B and Nicolas A (1993) The control in cis of the position and the amount of the ARG4 meiotic double-strand break of Saccharomyces cerevisiae. EMBO J, 12, 1459–1466. | PubMed | ChemPort |

de Massy B, Baudat F and Nicolas A (1994) Initiation of recombination in Saccharomyces cerevisiae haploid meiosis. Proc Natl Acad Sci USA, 91, 11929–11933. | PubMed | ChemPort |

DiNardo S, Voelkel K and Sternglanz R (1984) DNA topoisomerase II mutant of Saccharomyces cerevisiae: topoisomerase II is required for segregation of daughter molecules at the termination of DNA replication. Proc Natl Acad Sci USA, 81, 2616–2620. | PubMed | ChemPort |

Downes CS, Clarke DJ, Mullinger AM, Gimenez-Abian JF, Creighton AM and Johnson RT (1994) A topoisomerase II-dependent G₂ cycle checkpoint in mammalian cells. Nature, 372, 467–470. | Article | PubMed | ISI | ChemPort |

Drlica K (1990) Bacterial topoisomerases and the control of DNA supercoiling. Trends Genet, 6, 433–437. | Article | PubMed | ChemPort |

Duguet M (1997) When helicase and topoisomerase meet! J Cell Sci, 110, 1345–1350. | PubMed | ChemPort |

Ellis NA, Groden J, Ye T-Z, Straughen J, Lennon DJ, Ciocci S, Proytcheva M and German J (1995) The Bloom's syndrome gene product is homologous to RecQ helicases. Cell, 83, 655–666. | PubMed | ISI | ChemPort |

Ezekiel UR, Towler EM, Wallis JW and Zassenhaus HP (1994) Evidence for a nucleotide-dependent topoisomerase activity from yeast mitochondria. Curr Genet, 27, 31–37. | PubMed | ChemPort |

Fabre F, Boulet A and Roman H (1984) Gene conversion at different points in the mitotic cycle of Saccharomyces cerevisiae. Mol Gen Genet, 195, 139–143. | PubMed | ChemPort |

Fritz E, Elsea SH, Patel PI and Meyn MS (1997) Overexpression of a truncated human topoisomerase III partially corrects multiple aspects of the ataxia-telangiectasia phenotype. Proc Natl Acad Sci USA, 94, 4538–4542. | Article | PubMed | ChemPort |

Gangloff S, Lieber MR and Rothstein R (1994a) Transcription, topoisomerases and recombination. Experientia, 50, 261–269. | PubMed | ChemPort |

Gangloff S, McDonald JP, Bendixen C, Arthur L and Rothstein R (1994b) The yeast type I topoisomerase Top3 interacts with Sgs1, a DNA helicase homolog: a potential eukaryotic reverse gyrase. Mol Cell Biol, 14, 8391–8398. | PubMed | ISI | ChemPort |

Gangloff S, Zou H and Rothstein R (1996) Gene conversion plays the major role in controlling the stability of large tandem repeats in yeast. EMBO J, 15, 1715–1725. | PubMed | ISI | ChemPort |

Giaever GN and Wang JC (1988) Supercoiling of intracellular DNA can occur in eukaryotic cells. Cell, 55, 849–856. | PubMed | ChemPort |

Gilbertson LA and Stahl FW (1996) A test of the double-strand break repair model for meiotic recombination in Saccharomyces cerevisiae. Genetics, 144, 27–41. | PubMed | ISI | ChemPort |

Goto T and Wang JC (1984) Yeast DNA topoisomerase II is encoded by a single-copy, essential gene. Cell, 36, 1073–1080. | PubMed | ChemPort |

Goto T and Wang JC (1985) Cloning of yeast TOP1, the gene encoding DNA topoisomerase I and construction of mutants defective in both DNA topoisomerase I and DNA topoisomerase II. Proc Natl Acad Sci USA, 82, 7178–7182. | PubMed | ChemPort |

Hanai R, Caron PR and Wang JC (1996) Human TOP3: a single-copy gene encoding DNA topoisomerase III. Proc Natl Acad Sci USA, 93, 3653–3657. | Article | PubMed | ChemPort |

Holm C, Goto T, Wang JC and Botstein D (1985) DNA topoisomerase II is required at the time of mitosis in yeast. Cell, 41, 553–563. | PubMed | ChemPort |

Holm C, Stearns T and Botstein D (1989) DNA topoisomerase II must act at mitosis to prevent nondisjunction and chromosome breakage. Mol Cell Biol, 9, 159–168. | PubMed | ChemPort |

Hsieh T, Lee MP and Brown SD (1994) Structure of eukaryotictype I DNA topoisomerase. Adv Pharmacol, 29, 191–200.

Kadyk LC and Hartwell LH (1992) Sister chromatids are preferred over homologs as substrates for recombinational repair in Saccharomyces cerevisiae. Genetics, 132, 387–402. | PubMed | ISI | ChemPort |

Kawasaki I, Bae YS, Eki T, Kim Y and Ikeda H (1994) Homologous recombination of monkey alpha-satellite repeats in an in vitro simian virus 40 replication system: possible association of recombination with DNA replication. Mol Cell Biol, 14, 4173–4182. | PubMed | ChemPort |

Kawasaki K, Minoshima S, Nakato E, Shibuya K, Shintani A, Schmeits JL, Wang J and Shimizu N (1997) One-megabase sequence analysis of the human immunoglobulin gene locus. Genome Res, 7, 250–261. | PubMed | ISI | ChemPort |

Keeney S, Giroux CN and Kleckner N (1997) Meiosis-specific DNA double-strand breaks are catalyzed by Spo11, a member of a widely conserved protein family. Cell, 88, 375–384. | PubMed | ISI | ChemPort |

Kim RA and Wang JC (1989) A subthreshold level of DNA topoisomerases leads to the excision of yeast rDNA as extrachromosomal rings. Cell, 57, 975–985. | PubMed | ISI | ChemPort |

Kim RA and Wang JC (1992) Identification of the yeast TOP3 gene product as a single strand-specific DNA topoisomerase. J Biol Chem, 267, 17178–17185. | PubMed | ISI | ChemPort |

Kim RA, Caron PR and Wang JC (1995) Effects of yeast DNA topoisomerase III on telomere structure. Proc Natl Acad Sci USA, 92, 2667–2671. | PubMed | ChemPort |

Klapholz S, Waddell CS and Esposito RE (1985) The role of the SPO11 gene in meiotic recombination in yeast. Genetics, 110, 187–216. | PubMed | ISI | ChemPort |

Kleckner N (1996) Meiosis: how could it work? Proc Natl Acad Sci USA, 93, 8167–8174. | Article | PubMed | ChemPort |

Li W and Wang JC (1998) Mammalian DNA topoisomerase III is essential in early embryogenesis. Proc Natl Acad Sci USA, 95, 1010–1013. | Article | PubMed | ChemPort |

Lima CD and Mondragon A (1994) Mechanism of type II DNA topoisomerases: a tale of two gates. Structure, 2, 559–560. | PubMed | ChemPort |

Liu LF and Wang JC (1987) Supercoiling of the DNA template during transcription. Proc Natl Acad Sci USA, 84, 7024–7027. | PubMed | ChemPort |

Lydall D and Weinert T (1995) Yeast checkpoint genes in DNA damage processing: implications for repair and arrest. Science, 270, 1488–1491. | PubMed | ISI | ChemPort |

Lydall D, Nikolsky Y, Bishop DK and Weinert T (1996) A meiotic recombination checkpoint controlled by mitotic checkpoint genes. Nature, 383, 840–843. | Article | PubMed | ISI | ChemPort |

Malone RE and Esposito RE (1981) Recombinationless meiosis in Saccharomyces cerevisiae. Mol Cell Biol, 1, 891–901. | PubMed | ISI | ChemPort |

Nakayama K, Irino N and Nakayama H (1985) The recQ gene of Escherichia coli K12: molecular cloning and isolation of insertion mutants. Mol Gen Genet, 200, 266–271. | PubMed | ISI | ChemPort |

Petes TD, Malone RE and Symington LS (1991) Recombination in yeast. In Broach,J.R., Pringle,J.R. and Jones,E.W. (eds), The Molecular and Cellular Biology of the Yeast Saccharomyces: Genome Dynamics, Protein Synthesis and Energetics. 1, 407–521.

Puranam KL and Blackshear PJ (1994) Cloning and characterization of RECQL, a potential human homologue of the Escherichia coli DNA helicase RecQ. J Biol Chem, 269, 29838–29845. | PubMed | ISI | ChemPort |

Puranam KL, Kennington E, Sait SN, Shows TB, Rochelle JM, Seldin MF and Blackshear PJ (1995) Chromosomal localization of the gene encoding the human DNA helicase RECQL and its mouse homologue. Genomics, 26, 595–598. | Article | PubMed | ChemPort |

Roca J (1995) The mechanisms of DNA topoisomerases. Trends Biochem Sci, 20, 156–160. | Article | PubMed | ChemPort |

Roeder GS (1997) Meiotic chromosomes: it takes two to tango. Genes Dev, 11, 2600–2621. | PubMed | ISI | ChemPort |

Rose D, Thomas W and Holm C (1990) Segregation of recombined chromosomes in meiosis I requires DNA topoisomerase II. Cell, 60, 1009–1017. | PubMed | ChemPort |

Rothstein R and Gangloff S (1995) Hyper-recombination and Bloom's syndrome: microbes again provide clues about cancer. Genome Res, 5, 421–426. | PubMed | ISI | ChemPort |

Schwacha A and Kleckner N (1994) Identification of joint molecules that form frequently between homologs but rarely between sister chromatids during yeast meiosis. Cell, 76, 51–63. | PubMed | ISI | ChemPort |

Schwacha A and Kleckner N (1995) Identification of double Holliday junctions as intermediates in meiotic recombination. Cell, 83, 783–791. | PubMed | ISI | ChemPort |

Schwacha A and Kleckner N (1997) Interhomolog bias during meiotic recombination: meiotic functions promote a highly differentiated interhomolog-only pathway. Cell, 90, 1123–1135. | PubMed | ISI | ChemPort |

Seki T, Seki M, Katada T and Enomoto T (1998) Isolation of a cDNA encoding mouse DNA topoisomerase III which is highly expressed at the mRNA level in the testis. Biochim Biophys Acta, 1396, 127–131. | Article | PubMed | ChemPort |

Sherman F and Roman H (1963) Evidence for two types of allelic recombination in yeast. Genetics, 48, 255–261. | PubMed | ChemPort |

Sherman F, Fink GR and Hicks JB (1986) Methods in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Shinohara A, Ogawa H and Ogawa T (1992) Rad51 protein involved in repair and recombination in S.cerevisiae is a RecA-like protein. Cell, 69, 457–470. | PubMed | ISI | ChemPort |

Sia RA and Mitchell AP (1995) Stimulation of later functions of the yeast meiotic protein kinase Ime2p by the IDS2 gene product. Mol Cell Biol, 15, 5279–5287. | PubMed | ChemPort |

Stewart E, Chapman CR, Al-Khodairy F, Carr AM and Enoch T (1997) rqh1⁺, a fission yeast gene related to the Bloom's and Werner's syndrome genes, is required for reversible S phase arrest. EMBO J, 16, 2682–2692. | Article | PubMed | ChemPort |

Thomas BJ and Rothstein R (1989) Elevated recombination rates in transcriptionally active DNA. Cell, 56, 619–630. | PubMed | ISI | ChemPort |

Thrash C, Voelkel K, DiNardo S and Sternglanz R (1984) Identification of Saccharomyces cerevisiae mutants deficient in DNA topoisomerase I activity. J Biol Chem, 259, 1375–1377. | PubMed | ChemPort |

Thrash C, Bankier AT, Barrell BG and Sternglanz R (1985) Cloning, characterization and sequence of the yeast DNA topoisomerase I gene. Proc Natl Acad Sci USA, 82, 4374–4378. | PubMed | ChemPort |

Tsao Y-P, Wu H-Y and Liu LF (1989) Transcription-driven supercoiling of DNA: direct biochemical evidence from in vitro studies. Cell, 56, 111–118. | PubMed | ChemPort |

Umezu K, Nakayama K and Nakayama H (1990) Escherichia coli RecQ protein is a DNA helicase. Proc Natl Acad Sci USA, 87, 5363–5367. | PubMed | ChemPort |

Wallis JW, Chrebet G, Brodsky G, Rolfe M and Rothstein R (1989) A hyper-recombination mutation in S.cerevisiae identifies a novel eukaryotic topoisomerase. Cell, 58, 409–419. | PubMed | ISI | ChemPort |

Wang JC (1996) DNA topoisomerases. Annu Rev Biochem, 65, 635–692. | Article | PubMed | ISI | ChemPort |

Wang Z and Droge P (1996) Differential control of transcription-induced and overall DNA supercoiling by eukaryotic topoisomerases in vitro. EMBO J, 15, 581–589. | PubMed | ChemPort |

Watt PM and Hickson ID (1994) Structure and function of type II DNA topoisomerases. Biochem J, 303, 681–695. | PubMed | ChemPort |

Watt PM, Louis EJ, Borts RH and Hickson ID (1995) Sgs1: a eukaryotic homolog of E.coli RecQ that interacts with topoisomerase II in vivo and is required for faithful chromosome segregation. Cell, 81, 253–260. | PubMed | ISI | ChemPort |

Watt PM, Hickson ID, Borts RH and Louis EJ (1996) SGS1, a homologue of the Bloom's and Werner's syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae. Genetics, 144, 935–945. | PubMed | ISI | ChemPort |

Wu H, Shyy S, Wang J and Liu L (1988) Transcription generates positively and negatively supercoiled domains in the template. Cell, 53, 433–440. | PubMed | ChemPort |

Xu L, Weiner BM and Kleckner N (1997) Meiotic cells monitor the status of the interhomolog recombination complex. Genes Dev, 11, 106–118. | PubMed | ChemPort |

Yan H, Chen CY, Kobayashi R and Newport J (1998) Replication focus-forming activity 1 and the Werner syndrome gene product. Nature Genet, 19, 375–378. | Article | PubMed | ISI | ChemPort |

Young CS, Cachianes G, Munz P and Silverstein S (1984) Replication and recombination in adenovirus-infected cells are temporally and functionally related. J Virol, 51, 571–577. | PubMed | ChemPort |

Yu CE et al. (1996) Positional cloning of the Werner's syndrome gene. Science, 272, 258–262. | PubMed | ISI | ChemPort |

Zou H and Rothstein R (1997) Holliday junctions accumulate in replication mutants via a RecA homolog-independent mechanism. Cell, 90, 87–96. | PubMed | ISI | ChemPort |