|Home | About | Journals | Submit | Contact Us | Français|
A critical event in eukaryotic DNA replication involves association of minichromosome maintenance (MCM2–7) proteins with origins, to form prereplicative complexes (pre-RCs) that are competent for initiation. The ability of mutants defective in MCM2–7 function to complete meiosis had suggested that pre-RC components could be irrelevant to premeiotic S phase. We show here that MCM2–7 proteins bind to chromatin in fission yeast cells preparing for meiosis and during premeiotic S phase in a manner suggesting they in fact are required for DNA replication in the meiotic cycle. This is confirmed by analysis of a degron mcm4 mutant, which cannot carry out premeiotic DNA replication. Later in meiosis, Mcm4 chromatin association is blocked between meiotic nuclear divisions, presumably accounting for the absence of a second round of DNA replication. Together, these results emphasize similarity between replication mechanisms in mitotic and meiotic cell cycles.
In meiosis, the normal alternation of S phase and chromosome disjunction seen in the mitotic cell cycle is altered so that a single round of premeiotic DNA replication is followed by two consecutive nuclear divisions, thus achieving a reduction in ploidy. Strict control to ensure a single round of DNA replication in premeiotic S phase is important to produce haploid cells containing a single complete copy of the genome, but the mechanism of this control has not been subject to as much analysis as in the vegetative cell cycle.
Analysis of mutants affecting replication enzymes such as DNA polymerases suggests that the basic mechanisms of S phase in the meiotic and mitotic cell cycles are similar (Schild and Byers, 1978 ; Johnston et al., 1982 ; Budd et al., 1989 ; Forsburg and Hodson, 2000 ). Furthermore, analysis of replication origin usage in Saccharomyces cerevisiae is consistent with basic conservation in the initiation mechanism (Collins and Newlon, 1994 ). In general, the same origins are used in both premeiotic and vegetative S phases, perhaps reflecting the fact that similar events at the origin recognition complex (ORC), which is bound to initiation sites, are occurring in both types of cell cycle. In the mitotic cell cycle, this involves association of six MCM2–7 proteins at ORC during late mitosis/early G1 in a process dependent on Cdc6/Cdc18 and Cdt1 (reviewed by Kearsey and Labib, 1998 ; Kelly and Brown, 2000 ; Maiorano et al., 2000 ; Nishitani et al., 2000 ). This process of prereplicative complex (pre-RC) formation confers replicative competence on the origin, allowing Cdc7 and cyclin-dependent kinase (CDK)-activated initiation of DNA synthesis during the subsequent S phase. During DNA replication, MCM2–7 proteins are thought to provide helicase activity for the elongation step of DNA replication (reviewed in Labib and Diffley, 2001 ). These proteins dissociate from chromatin, probably during replication termination, and cannot rebind because this step is inhibited by CDK activity and other mechanisms involving pre-RC components (Dahmann et al., 1995 ; Tanaka et al., 1997 ; Labib et al., 1999 ; Wohlschlegel et al., 2000 ; Tada et al., 2001 ). Thus, reinitiation is dependent on CDK inactivation in mitosis, limiting DNA replication to a single round per cell cycle.
In spite of these general similarities, some observations suggest that premeiotic S phase is not identical to vegetative DNA replication, and these differences could be related to specialized meiotic processes. Premeiotic S phase is universally longer than the S phase in cycling cells of the same organism (Holm, 1977 ; Cha et al., 2000 ), and this may reflect the activity of proteins needed for reductional chromosome segregation, such as Rec8, whose correct function is intimately associated with DNA replication (Cha et al., 2000 ; Watanabe et al., 2001 ). In S. cerevisiae, deletion of replication origins can delay double-strand break (DSB) appearance, perhaps because of coupling between replication and recombination (Borde et al., 2000 ) and preventing S phase activation also blocks DSB formation (Smith et al., 2001 ). There is also evidence that pre-RC formation and the mechanism of initiation and elongation in premeiotic S phase may be significantly different. Specifically, mcm2, mcm4, and cdc18 mutants of Schizosaccharomyces pombe are not arrested in premeiotic S phase or the subsequent nuclear divisions under conditions that block vegetative DNA replication (Forsburg and Hodson, 2000 ). Also, in budding yeast, although CDK activity is needed for activation of premeiotic S phase (Iino et al., 1995 ; Dirick et al., 1998 ; Stuart and Wittenberg, 1998 ), Cdc7 may not be required (Schild and Byers, 1978 ; Hollingsworth and Sclafani, 1993 ), which could reflect different regulatory controls over DNA replication.
Here we have investigated the role of MCM2–7 proteins as fission yeast cells exit the mitotic cell cycle and enter meiosis. By a combination of chromatin association assays and genetic analysis using novel degron alleles, we provide clear evidence that MCM2–7 proteins have an essential role in meiotic as well as vegetative S phases.
Strains used are shown in Table Table1.1. Media and growth conditions and standard genetic methods were as described by Moreno et al. (1991) . Diploid pat1 strains were made by protoplast fusion. P-factor arrest was carried out as described in Stern and Nurse (1997) using a P-factor concentration of 1.5 μg/ml in minimal medium supplemented with leucine. Thiamine at 5 μg/ml was used to repress the nmt1 promoter. Nitrogen starvation was carried out using EMM medium lacking NH4Cl.
Mcm2+ and mcm6+ genes, expressed from their own promoters, were tagged to express C-terminally-fused GFP5, as described earlier for mcm4+ (Kearsey et al., 2000 ). We constructed a GFP-containing integration vector, pSMRG2+, containing GFP5 and the kanMX6 selectable marker (EMBL accession no. AJ306910). This involved replacing the NgoM IV fragment containing the ura4+ gene in pSMUG (Kearsey et al., 2000 ) with a kanMX6 fragment, which was amplified using the primers 5′-atttagccggctgtttagcttgcctcgtccc-3′ and 5′-aattgccggcgagctcgtttaaactggatgg-3′ from pFA6a-kanMX6 (Bahler et al., 1998 ). Also the linker region upstream of GFP was enlarged by inserting the sequence 5′-ctcgagggtagatctggtgcccggggtggtgctggtgccggagccggtgctggtgctgaagctt-3′ between the unique XhoI and HindIII sites. The C-terminal encoding region of the mcm2+ gene was amplified using primers 5′- acgactcgagacactacaattccttttaat-3′ and 5′- ccaccccgggcaataagatatttagcaaatgttc-3′ and cloned into the XhoI and SmaI sites of pSMRG2+. Homologous integration into the mcm2+ gene was directed by linearization with NheI. For mcm6+, a similar procedure was used, using the primers 5′-gaacggggcccgcaagagcaaactgtgtag-3′ and 5′- cttgccccgggcgttcggaacatcgccattgc-3′ and cloning into the ApaI and SmaI sites. The plasmid was linearized using XhoI to direct integration into the mcm6+ locus. Constructs were verified by sequencing. The Mcm2-GFP and Mcm6-GFP strains have a normal growth rate and DNA content by flow cytometry, indicating that the tagged proteins are functional.
For tagging the mcm4+ gene in the background of the cyclin B shut-off strain (nmt1(41X)-cdc13, cdc13Δ cig1Δ cig2Δ; Fisher and Nurse, 1995 ) we replaced the ura4+ marker in pSMUG+mcm4-GFP (Kearsey et al., 2000 ) with an NgoM IV fragment containing the kanMX6 (kanr) marker (see above), to give pSMRG+mcm4-GFP. This was linearized with HpaI to direct integration at the mcm4+ locus, thus generating strain P886. Because this strain is thiamine sensitive and kanMX6 cannot be selected for on minimal (EMM) medium, we devised a modified medium (KsnoT, Kanamycin-selective, no thiamine: bacto-peptone 10 g/l, 3% glucose, 2% agar, 250 mg/l adenine, 250 mg/l uracil, 250 mg/l leucine, 75 μg/ml geneticin), which allows kanamycin selection in a thiamine-deficient medium. pSMRG+mcm4-GFP was also used to tag mcm4+ in pat1ts strains.
To construct the degron mcm4 strain, a plasmid was constructed containing the mcm4+/cdc21+ promoter, expressing an N-terminal ubiquitin-degron-HA cassette fused to the mcm4+/cdc21 gene. The degron was amplified as an ApaI-XhoI fragment using primers 5′-atagggcccctgcttatctttcttcttcc-3′ and 5′-atactcgaggcttgccctcctaaaaatgc-3′, with plasmid pPW66R (Dohman et al., 1994 ) as template. The mcm4+ promoter was amplified as a KpnI-ApaI fragment using the primers: 5′-ataggtaccccgcatttgatggtttcgcc-3′ and 5′-atagggccccgtggtgggtgtagaaagac-3′. Both fragments were cloned into KpnI and XhoI cleaved pSMUG2+ (ura4+-containining integration vector, identical to pSMRG2+, except containing the ura4+ gene instead of the kanMX6 marker; EMBL accession no. AJ306911) to give pSMUG2+degron. The 5′ region of the mcm4+ reading frame was amplified as a XhoI-BglII fragment using the primers 5′-atactcgaggtcctctagtcagcaaagtg-3′ and 5′-ataagatcttcaatttgtcaatgtcaccag-3′. This fragment was cloned into the XhoI-BglII region of pSMUG2+degron to give pSMUG2+degron+mcm4. The final construct was verified by sequencing. This plasmid was cleaved with SpeI to direct integration into the mcm4+ locus and thus tag the endogenous gene with the degron. The same strategy was used to construct the mcm4ts-td allele as the mcm4ts mutation (cdc21-M68) causes a Leu to Pro substitution at position 238, i.e., is not in the N-terminal region of the protein (S. Montgomery and S.E. Kearsey, manuscript in preparation).
Chromatin binding of GFP-tagged proteins was analyzed using a modified version of the protocol described in Kearsey et al. (2000) . Instead of ZM buffer, cells were resuspended in ZM2 buffer (15 mM potassium hydrogen phthalate, 15 mM Na2HPO4, pH 7.0, 90 mM NH4Cl, 1.2 M sorbitol, 10 mM dithiothreitol) and zymolyase 20-T was added to 2 mg/ml. Cells were washed twice in ZM buffer, once in EB2 (20 mM PIPES-KOH, pH 6.8, 0.4 M sorbitol, 1 mM EDTA, 0.5 mM spermidine-HCl, 1.5 mM spermine-HCl, 150 mM KAc, 1/1000 volume protease inhibitor cocktail; P-8215, Sigma, St. Louis, MO), and cells were extracted in EB2 containing 1% (wt/vol) Triton X-100 for 5 min at 20°C. Cells were fixed with methanol/acetone and analyzed by fluorescence microscopy as previously described (Kearsey et al., 2000 ). At least 100 cells were counted for each data point, and error bars show the range of two experiments. For flow cytometry, methanol/acetone fixed cells were rehydrated in 10 mM EDTA, pH 8.0, 0.1 mg/ml RNase A, 2 μg/ml propidium iodide or 1 μM sytox green, and incubated at 37°C for 2 h. Cells were analyzed using a Coulter Epics XL-MCL (Fullerton, CA).
Protein extracts were made by TCA extraction and analyzed by Western blotting as described previously (Grallert et al., 2000 ). Mcm4 was detected using a mouse mAb KL2.2, which was generated against full-length, bacterially expressed Mcm4 (Maiorano et al., 1996 ). The recognized epitope is in the N-terminal 302 amino acids (unpublished observations). α-Tubulin was detected using Sigma T5168 at a dilution of 1/10,000.
To determine the relevance of MCM2–7 proteins for premeiotic DNA replication, we examined the chromatin association of these proteins as cells exit vegetative growth and prepare for mating and meiosis. In fission yeast, meiosis is induced by nutrient deprivation, which causes haploid cells to express mating pheromones, arrest in G1, and conjugate to form a diploid zygote. Usually, meiosis then commences immediately, followed by sporulation to form a four-spored ascus. To follow pre-RC formation during this process we used an in situ detergent-washing procedure to examine the chromatin binding of GFP-tagged MCM2–7 proteins in single fission yeast cells. MCM proteins that are bound to chromatin are resistant to detergent extraction and remain nuclear, whereas unbound nucleoplasmic protein is washed away. Using this method, we have previously shown that chromatin association of Mcm4 (Cdc21) is restricted to the interval from mid-anaphase to S phase and shows an expected dependence on ORC and Cdc18 (Kearsey et al., 2000 ). To study the effect of the P-factor mating pheromone on Mcm4 chromatin binding without the need for simultaneous nutrient deprivation, we used a genetic background that allows pheromone-induced G1 arrest of nonstarved cells. Deleting the cyr1+ gene lowers intracellular cAMP levels, thus activating genes needed for the mating pheromone reponse (Maeda et al., 1990 ), and deleting the sxa2+ gene reduces P-factor proteolysis (Imai and Yamamoto, 1992 ; Ladds et al., 1996 ). Thus, by introducing Mcm4-GFP into a cyr1Δ sxa2Δ genetic background, Mcm4 chromatin binding could be monitored during P-factor arrest of exponentially growing cells (Figure (Figure1A).1A). Before addition of P-factor, Mcm4 was only chromatin associated in binucleate (late M/G1/S phase) cells (Figure (Figure1,1, B and C) as in a wild-type strain. In contrast, 3.5 h after addition of P-factor, 1C cells were prominent by flow cytometry and a high proportion of uninucleate cells were positive for Mcm4 chromatin binding, implying that pre-RC assembly had occurred in G1-arrested cells (Figure (Figure1,1, B–D). Chromatin association of Mcm4 appeared to be stable at least up to 7 h, but it was not possible to investigate this for longer periods because pheromone-arrested cells start to enter S phase after about 8 h of arrest (Davey and Nielsen, 1994 ; Imai and Yamamoto, 1994 ). We also observed that wild-type cells arrested in G1 phase by a brief period (4–7 h) of nitrogen starvation alone showed chromatin-associated Mcm4, although this association was not stable on longer periods of arrest (>12 h, see below). These results show that cells arrested in a state competent for mating and meiosis have chromatin-bound Mcm4, implying this could be relevant to execution of DNA replication in the meiotic cycle.
Because previous studies have shown that inhibition of Cdc2 activity is responsible for cell cycle arrest in G1 by mating pheromone (Stern and Nurse, 1997 , 1998 ), we examined the effect of directly inhibiting Cdc2 activity on Mcm4 chromatin association. This was examined in a strain containing a thiamine-repressible cdc13+ gene in the background of cyclin B gene deletions (cig1Δ, cig2Δ, and cdc13Δ), which arrests mainly in G1 after addition of thiamine to the medium (Figure (Figure2,2, A and D; Fisher and Nurse, 1995 ). Mcm4 was shown to be chromatin associated in G1-arrested cells (Figure (Figure2,2, B and C), implying that inhibition of Cdc2 activity alone is sufficient to explain Mcm4 chromatin association seen in cells arrested by mating pheromone.
To examine more directly whether Mcm4 associated with chromatin in G1-arrested cells is relevant to meiosis, we determined whether this and other MCM2–7 proteins bind to chromatin during premeiotic S phase. It is difficult to examine this process in a wild-type diploid, because entry to meiosis on nutrient starvation is rather asynchronous. We therefore made use of a temperature-sensitive pat1 allele (pat1ts), which encodes a defective negative regulator of meiosis, and even haploid strains containing this allele can be induced to enter meiosis by shifting to the restrictive temperature (Iino and Yamamoto, 1985 ; Nurse, 1985 ; McLeod and Beach, 1986 ). Pat1 inactivation leads to a meiosis that is very similar to that induced physiologically by nutrient deprivation and has been generally used for analyzing meiotic mechanisms. A pat1ts strain was arrested in G1 by nitrogen starvation for 16 h, after which meiosis was induced by shifting to the restrictive temperature and refeeding (Figure (Figure3A).3A). Initially Mcm4 was not chromatin bound, although binding increased after 2 h, and peaked slightly in advance of the time of premeiotic S phase (Figure (Figure3,3, B–D). As premeiotic S phase finished, Mcm4 chromatin association was lost. Similar results were obtained using a diploid pat1 strain shifted to 34°C (Figure (Figure4,4, A and D), these conditions being compatible with a viable meiosis (Bähler et al., 1991 ), although the timing of premeiotic S phase was a little advanced compared with the haploid strain (Figure (Figure4C).4C). Thus, the timing of Mcm4 association with chromatin suggests that MCM proteins function in premeiotic DNA replication as in a normal S phase.
Following later stages of the diploid meiosis showed that Mcm4 remained nuclear during meiosis I and II, but Mcm4 chromatin association was not seen at any stage between these nuclear divisions (Figure (Figure4,4, A, B, and D, 5–6 h). Thus, a block to MCM chromatin association and pre-RC formation could account for absence of DNA replication between meiosis I and II. The resistance of spores to zymolyase digestion made it difficult to examine Mcm4 chromatin binding after meiosis II, but a high proportion of cells showed absence of Mcm4 chromatin binding (Figure (Figure4A,4A, 6 h; 4D, 5–6 h) before obvious spore formation. Thus, MCM chromatin association may have to be re-established after spore germination.
To test if the correlation of Mcm4 chromatin association with premeiotic DNA replication reflects a direct involvement with DNA synthesis, we examined how blocking S phase affected MCM2–7 chromatin binding. In the vegetative cell cycle, arresting the elongation step of DNA synthesis with hydroxyurea (HU) prevents the displacement of Mcm4 (Kearsey et al., 2000 ). To extend the meiotic analysis to other MCM2–7 proteins, we also tagged Mcm2 and Mcm6 with GFP. In vegetative cells, these proteins are similar to Mcm4 in terms of cell cycle changes in chromatin binding. Pat1ts strains were arrested in G1 and induced to undergo premeiotic S phase after refeeding and shifting to 37°C as before, except that HU was added to half the culture (Figure (Figure5A).5A). In the presence of HU, most cells showed chromatin binding of Mcm2, Mcm4, and Mcm6 after 5.5 h, whereas control cultures without HU had completed premeiotic S phase and chromatin binding of these proteins was not detected (Figure (Figure5,5, B and D). Similar results were obtained using a diploid pat1ts mcm4-GFP strain. When the same experiment was carried out with an Mcm2-GFP strain containing a temperature-sensitive mcm4 mutation (mcm4ts-td, see below), Mcm2 chromatin association was blocked (Figure (Figure5,5, C and D). Thus, there is mutual dependency of MCM2–7 proteins for chromatin association, as has been shown in the vegetative cell cycle (Pasion and Forsburg, 1999 ; Labib et al. 2001 ). Overall, these results suggest that premeiotic S phase involves chromatin association of MCM2–7 proteins and that displacement of these proteins from chromatin requires completion of DNA replication as in the mitotic cell cycle.
To investigate whether chromatin association of MCM2–7 proteins in premeiotic S phase reflects a functional requirement for these proteins, we analyzed the effect on an mcm4 mutation on meiosis. Because available temperature-sensitive alleles of mcm2–7 genes are rather leaky and do not block vegetative S phase efficiently (e.g., Forsburg and Nurse, 1994 ; Takahashi et al. 1994 ), we explored the use of temperature-sensitive degron (td) alleles in S. pombe, because these have been useful for clarifying MCM2–7 function in S. cerevisiae (Labib et al., 2000 ). We constructed a degron fusion of Mcm4 (mcm4td) where the N-terminus of Mcm4 is fused to the DHFR degron (Dohman et al., 1994 ), expressed from the native mcm4+ promoter. In S. cerevisiae this degron is stable at 25°C, but at 37°C the degron is ubiquitylated by Ubr1, probably because of an increase in the accessibility of its N-terminal arginine, leading to its rapid proteolysis (Dohman et al., 1994 ; Lévy et al., 1999 ). In S. pombe, the mcm4td strain grew normally and levels of the degron-Mcm4 protein were similar to Mcm4 levels at 25°C, but at 37°C cells were elongated, implying that the degron confers temperature sensitivity in this organism. Flow cytometry did not show a tight block to DNA replication; however, and Mcm4 levels were not dramatically reduced, implying that Mcm4 degradation is too inefficient to block DNA replication at initiation. To obtain a mutant with a tighter phenotype, we modified a temperature-sensitive mcm4 allele (mcm4ts/cdc21-M68; Nasmyth and Nurse, 1981 ) by fusion to the degron, to give a mcm4ts-td strain. This mutant arrested with predominantly 1C DNA after shifting exponentially growing cells to the restrictive temperature in contrast to the mcm4ts mutant, which showed a leakier phenotype (Figure (Figure6A).6A). Western blotting showed that levels of the Mcm4 protein were similar in the mcm4ts-td and mcm4ts strains at 25°C and in the mcm4ts strain at 37°C, but a rapid reduction in protein levels in the mcm4ts-td strain was seen at 37°C (Figure (Figure6B).6B). To determine if the mcm4ts-td allele affected premeiotic DNA replication, we constructed a double mutant of mcm4ts-td with pat1ts, and G1-arrested cells were refed and shifted to 37°C. Control mcm4+ pat1ts cells carried out premeiotic S phase around 3 h, whereas mcm4ts-td pat1ts cells did not replicate their DNA, although a minor fraction of cells exhibited partial replication (Figure (Figure6C).6C). This effect on premeiotic S phase was more severe than that seen with a mcm4ts (cdc21-M68) pat1ts mutant, which showed more extensive replication as previously reported (Forsburg and Hodson, 2000 ; Murakami and Nurse, 2001 ). Meiotic nuclear divisions were also reduced in the degron mcm4 strain, consistent with a block in DNA replication (Figure (Figure6D;6D; Murakami and Nurse, 1999 ). Western analysis showed that levels of Mcm4 were significantly reduced in the mcm4ts-td mutant compared with mcm4+ strain after the nitrogen starvation step, and there was a further reduction in protein levels only 30 min after the shift to 37°C (Figure (Figure6E).6E). Protein levels were also lower than in a mcm4ts strain, where we could detect Mcm4 throughout premeiotic S phase. We have also shown that a degron mcm6 mutation blocks premeiotic S phase. Thus, these results, taken together with analysis of MCM chromatin binding, indicate that execution of premeiotic S phase requires the participation of MCM2–7 proteins.
In this article we have investigated MCM2–7 chromatin binding and, by inference, pre-RC formation, during G1 arrest of the mitotic cycle and entry into meiosis. In summary, S. pombe cells arrested in G1 by mating pheromone have chromatin-bound Mcm4. This pre-RC formation is relevant to premeiotic S phase, because Mcm4 binds to chromatin around the time of premeiotic S phase and S phase completion is necessary to allow displacement of Mcm2, Mcm4, and Mcm6. Analysis of a degron mutant shows that Mcm4 is essential for premeiotic S phase and given that in the vegetative cell cycle MCM2–7 proteins interact (Adachi et al., 1997 ; Pasion and Forsburg, 1999 ) and are required for licensing and the elongation steps of replication (Labib et al., 2000 ; Prokhorova and Blow, 2000 ; Tye and Sawyer, 2000 ), it is likely that all MCM2–7 proteins are required for premeiotic DNA replication.
These results show pheromone arrest in S. pombe is similar to that in S. cerevisiae, where α-factor also arrests cells in G1 with pre-RCs assembled and MCM2–7 proteins bound to chromatin (Diffley et al., 1994 ; Donovan et al., 1997 ). The mechanisms of these cell cycle arrests are distinct in detail, although in both cases pheromone blocks CDK activation needed for S phase entry (Peter et al., 1993 ; Peter and Herskowitz, 1994 ; Stern and Nurse, 1998 ), whereas expression of Cdc18/Cdc6 needed for pre-RC formation is not prevented (Zwerschke et al., 1994 ; Stern and Nurse, 1997 ). Budding and fission yeast cells have different fates after diploid formation, which is relevant to the function of chromatin-associated MCM2–7 proteins. Because S. cerevisiae haploid cells are constitutively competent for conjugation, MCM2–7 proteins in mating cells would generally function in a vegetative S phase. On the other hand, because S. pombe cells only fuse on nutrient limitation and zygotes progress directly to meiosis, MCM2–7 proteins assembled onto chromatin in mating competent cells are likely to function in premeiotic S phase.
It is likely that a previous report suggesting that MCM2–7 proteins are not needed for premeiotic S phase (Forsburg and Hodson, 2000 ) reflects the leaky nature of the original conditional alleles compared with more efficient inactivation of Mcm4 function in our degron allele While this article was in preparation, Murakami and Nurse (2001) reported that using a higher restrictive temperature than that used in the initial study does in fact prevent completion of premeiotic S phase, using mcm2 and mcm4 mutants. One factor that could be relevant to why premeiotic DNA replication is not blocked by mcm mutations under conditions that prevent completion of vegetative S phase is the nitrogen starvation step used for G1 synchronization before meiotic entry. Nitrogen-starved cells have increased levels of Rum1 compared with vegetative cells (Maekawa et al., 1998 ) and thus depressed CDK activity, and CDK levels are further reduced by enhanced proteolysis of cyclin B (Yamaguchi et al., 1997 ; Kitamura et al., 1998 ; Kominami et al., 1998 ). Both these changes could suppress mutations affecting Cdc18 or MCM2–7 function by promoting pre-RC formation, based on analysis of mutants with reduced CDK levels (Jallepalli and Kelly, 1996 ; Grallert et al., 2000 ). For instance, enhanced Rum1 levels suppresses a cdc18 mutation via inhibition of Cdc18 proteolysis (Jallepalli and Kelly, 1996 ) and deletion of the cig2+ B-type cyclin suppresses cdc18, mcm2, and mcm4 mutations (Grallert et al., 2000 ). Another possibility is that the difference between premeiotic and vegetative DNA replication might be quantitative in that less MCM2–7 function is required for premeiotic S phase. It seems less likely that there is an alternative meiotic replication pathway capable of compensating for loss of MCM2–7 function, given the effective replication arrest seen in the degron mcm4 mutant.
Mcm4 levels are maintained through meiosis (Forsburg and Hodson, 2000 ), but chromatin association of Mcm4 does not occur between meiosis I and II even though the protein remains nuclear (Figure (Figure4),4), presumably accounting for the absence of a second round of DNA replication. Inactivation of components required for pre-RC formation would provide an obvious mechanism to block MCM chromatin association, and it is of interest that in Xenopus incomplete inactivation of Cdc2 after meiosis I is required for preventing DNA replication (Iwabuchi et al., 2000 ; Nakajo et al., 2000 ). If a similar situation applies to fission yeast, Cdc2-mediated destabilization of Cdc18 (Baum et al., 1998 ) could constitute one mechanism to ensure that only a single round of DNA replication occurs in meiosis.
The authors thank Paul Nurse's group for strains and plasmids and for communicating results before publication; Sue Cotterill, Tim Humphrey, and Karim Labib for discussion and comments on the manuscript; and Ben Martynoga and Ashley Spearing for technical help. This work was supported by the EU TMR program (contract ERB-MRX-CT970125) and the Cancer Research Campaign (grant SP1897/0301).
Article published online ahead of print. Mol. Biol. Cell 10.1091/mbc.01–11-0537. Article and publication date are at www.molbiolcell.org/cgi/10.1091/mbc.01–11-0537.