Overlapping Roles of the Spindle Assembly and DNA Damage Checkpoints in the Cell-Cycle Response to Altered Chromosomes in Saccharomyces cerevisiae

Corresponding author: Jasper Rine, 401 Barker Hall, University of California, Berkeley, CA 94720., jrine{at}uclink4.berkeley.edu (E-mail)

Communicating editor: L. PILLUS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The MAD2-dependent spindle checkpoint blocks anaphase until all chromosomes have achieved successful bipolar attachment to the mitotic spindle. The DNA damage and DNA replication checkpoints block anaphase in response to DNA lesions that may include single-stranded DNA and stalled replication forks. Many of the same conditions that activate the DNA damage and DNA replication checkpoints also activated the spindle checkpoint. The mad2 mutation partially relieved the arrest responses of cells to mutations affecting the replication proteins Mcm3p and Pol1p. Thus a previously unrecognized aspect of spindle checkpoint function may be to protect cells from defects in DNA replication. Furthermore, in cells lacking either the DNA damage or the DNA replication checkpoints, the spindle checkpoint contributed to the arrest responses of cells to the DNA-damaging agent methyl methanesulfonate, the replication inhibitor hydroxyurea, and mutations affecting Mcm2p and Orc2p. Thus the spindle checkpoint was sensitive to a wider range of chromosomal perturbations than previously recognized. Finally, the DNA replication checkpoint did not contribute to the arrests of cells in response to mutations affecting ORC, Mcm proteins, or DNA polymerase . Thus the specificity of this checkpoint may be more limited than previously recognized.

THE ability to accurately transmit genetic material to daughter cells is essential to all life. Eukaryotic organisms have evolved mechanisms called checkpoints that increase the fidelity of genetic transmission. Checkpoints enhance fidelity by delaying cell-cycle progression in cells with defects in chromosomes or in the machinery that segregates chromosomes. Cancer cells display reduced fidelity of genetic transmission and frequently have mutations in checkpoint genes (LI and BENEZRA 1996 ; CAHILL et al. 1998 ; LENGAUER et al. 1998 ). Thus checkpoint failure contributes to cancer. Since the defects that checkpoints respond to are likely to be defects that destabilize the genome, identification of those defects should increase our understanding of genetic instability and also our understanding of how checkpoints prevent cancer.

A variety of conditions that disrupt chromosomes and/or chromosome segregation cause Saccharomyces cerevisiae cells to arrest prior to anaphase through one or more of three different checkpoints. These checkpoints differ in the types of agents that elicit their response and also in the genes that are required for their function. A checkpoint termed the DNA damage checkpoint arrests cells that have been treated with DNA-damaging agents. This checkpoint requires RAD9, RAD17, RAD24, RAD53, DDC1, DDC2, MEC1, and MEC3 for full function (reviewed by FOIANI et al. 2000 ). A second checkpoint, variously called the replication, S-phase, or S-M checkpoint, arrests cells in which replication has been blocked by deoxyribonucleotide depletion. This checkpoint overlaps with the DNA damage checkpoint in its requirement for RAD53, MEC1, and DDC2, but unlike the DNA damage checkpoint, does not require RAD9, RAD17, RAD24, MEC3, or DDC1 (reviewed by LOWNDES and MURGUIA 2000 ). A third checkpoint, termed the spindle assembly checkpoint, arrests cells in which replicated chromatids fail to achieve bipolar spindle attachment. This checkpoint requires MAD1, MAD2, MAD3, BUB1, BUB3, NDC10, and MPS1 for full function (reviewed by HOYT 2001 ).

In this study, we quantify the roles of these three checkpoints in the preanaphase arrests that occur in cells that have lost the function of various essential replication proteins. Initially, in an attempt to gain insight into the function of the eukaryotic DNA replication initiator, the origin recognition complex (ORC), we quantified the roles of these three checkpoints in the preanaphase arrests of cells that had lost ORC function. We were surprised to find that, although the DNA damage and spindle assembly checkpoints contributed to the arrests of orc cells, the DNA replication checkpoint did not. To determine whether this pattern of checkpoint responses was unique to orc mutants, we conducted similar analyses of cells harboring conditional mutations affecting Mcm proteins, DNA Pol, and DNA Pol. We found that the spindle assembly and DNA damage checkpoints jointly mediated arrest responses to a variety of replication mutations and that the spindle assembly checkpoint was capable of mediating arrest responses to a DNA-damaging agent and a DNA replication inhibitor.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains:
mcm2-1, mcm3-1, cdc2-1, and pol1-17 alleles from other strain backgrounds were backcrossed to W303 a minimum of four times. The orc2-1 strain was constructed by gene replacement of ORC2 in W303. The resulting collection of replication mutants was crossed to rad9, rad24, mad2, and mec1 checkpoint mutant strains isogenic to W303 to create the replication mutant strains used in this study (Table 1). The mad1::KanMX allele, derived from the Research Genetics (Birmingham, AL) MATa deletion collection, was backcrossed once to W303 before crossing to a W303-isogenic rad9 rad24 strain to create JRY7309–JRY7320. Standard genetic procedures were as described (ROSE et al. 1989 ).

Growth, synchronization, methyl methanesulfonate treatment, hydroxyurea treatment, and 4',6-diamidino-2-phenylindole staining:
YPD (rich medium) was used in all experiments. Temperature-sensitive replication mutants were maintained at 23°. The restrictive temperature used in all temperature-shift experiments was 37°. Strains lacking replication mutations were grown at 25–28°. -Factor was used at 2.5–5 µg/ml to synchronize MATa cells in G1. -Factor-arrested cells were washed twice in prewarmed YPD before being released into prewarmed YPD containing 5 µg/ml Pronase (Calbiochem 53702) protease. Methyl methanesulfonate (MMS, M-4016; Sigma, St. Louis) was added to YPD medium at 0.033%, except as noted. Hydroxyurea (HU) was added to YPD medium at 200 mM. 4',6-Diamidino-2-pheynylindole (DAPI) staining of fixed cells was as described (ROSE et al. 1989 ), and cell-cycle arrest or progression was determined by calculating the percentage of large-budded uninucleate cells by fluorescence microscopy of >200 cells. All cultures were coded during scoring so that the scorer was blind to the genotype of the culture being scored.

Viability in MMS:
The number of colony-forming units (CFU) per microliter in cultures of wild-type, mad2, rad9 rad24, and mad2 rad9 rad24 cells was determined both before and at various times after the addition of 0.033% MMS by plating cells on non-MMS-containing medium and counting colonies after 3 days of growth at 25°. At each time point after MMS addition, the viability of each culture was expressed as the relation (CFU per microliter at that time point)/(CFU per microliter before MMS addition). To determine the significance of the effect of mad2 on the viability of the rad9 rad24 strains, the viabilities of both the rad9 rad24 strains and the mad2 rad9 rad24 strains were normalized to the mean viability of the rad9 rad24 strains at that time point. This normalization permitted compiling the viability of these strains at all time points, which was then expressed as (viability of strains X)/(viability of rad9 rad24 strains) ± 95% confidence limits.

Growth rate:
Six (wild-type, mad2) or seven (rad9 rad24, mad2 rad9 rad24) log-phase cultures of strains of the indicated genotypes were diluted to OD₆₀₀ 0.06 and grown for 3.5–5 hr at 25°. The ratios of the final ODs to the initial ODs were used to compute the doubling time of each culture, which was then normalized to the mean doubling time of the wild-type cultures. The means of these normalized doubling times ±SD are shown.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

To test the roles of the DNA damage, DNA replication, and spindle assembly checkpoints in the arrest responses of cells with replication defects, budding yeast strains harboring temperature-sensitive mutations in genes encoding one of five different replication proteins were studied. Three of these mutations affect proteins involved in replication initiation: orc2-1 (affecting ORC subunit 2); mcm2-1 [affecting minichromosome maintenance (MCM) protein 2]; and mcm3-1 (affecting MCM protein 3) (YAN et al. 1991 ; FOSS et al. 1993 ). These proteins are components of the preinitiation complex, which assembles at replication origins, rendering them competent to initiate DNA replication. Intriguingly, mutations in each of these proteins cause cells to arrest prior to anaphase with a genome that is either fully replicated or nearly so (GIBSON et al. 1990 ; YAN et al. 1991 ; BELL et al. 1993 ; PFLUMM and BOTCHAN 2001 ). Since it was not clear why mutations affecting the preinitiation complex should cause cells to arrest prior to anaphase, knowledge of which checkpoints, if any, were responsible for this phenotype was a first step toward understanding it.

The other two mutations affect proteins involved in replication elongation. The cdc2-1 mutation affects DNA polymerase , the major replicative DNA polymerase (BOULET et al. 1989 ), and causes cells to arrest in mid-S-phase (BUDD and CAMPBELL 1993 ; P. GARBER and J. RINE, unpublished results). The pol1-17 mutation affects DNA polymerase (BUDD and CAMPBELL 1987 ), responsible for priming DNA synthesis, and also causes cells to arrest in mid-S-phase (BUDD et al. 1989 ). Whereas it was unclear whether unreplicated DNA would be present during the late S/G2 arrests of the initiation mutants, the mid-S arrest of these mutants ensured that significantly underreplicated chromosomes would be present for potential detection by the various checkpoints.

In a current model of the DNA-responsive checkpoint pathways in S. cerevisiae, MEC1 is essential to the checkpoint responses to both DNA damage and stalled replication. In contrast, RAD9 and RAD24 are essential only to the checkpoint response to DNA damage and are not required for the response to stalled replication (Fig 6A). In preliminary experiments with the replication mutants, we found that combined rad9 and rad24 mutations relieved the cell-cycle arrests of all of the mutants other than pol1-17 to the same degree as did the mec1 mutation. Therefore the RAD9- and RAD24-independent, MEC1-dependent replication checkpoint pathway did not make a significant contribution to the arrests of these mutants. Since use of the mec1 mutation prevents distinguishing between the contributions of the DNA replication and DNA damage checkpoints, we used the combined rad9 and rad24 mutations and left MEC1 intact in the majority of these experiments.

View larger version (55K): In this window In a new window Download PPT slide   Figure 1. MAD2-mediated arrest responses to a variety of replication mutations. Cultures of strains of the indicated genotypes were incubated at the restrictive temperature of 37° for 3.5 hr and then harvested, fixed, and stained with DAPI. The percentage of large-budded uninucleate cells in each culture was then determined by fluorescence microscopic examination of >200 cells. Each bar represents the mean of 3–11 such trials, and the error bars represent the 95% confidence limits of that mean. Excepting mcm3-1 rad9 rad24, each genotype was represented by a minimum of two, and usually three or more, independently derived strains. "No treatment" refers to control strains lacking replication mutations but containing the indicated checkpoint mutations.

View larger version (55K): In this window In a new window Download PPT slide   Figure 2. The spindle-checkpoint-arrested cells lacking the DNA damage or DNA replication checkpoints in response to MMS and HU. The percentage of large-budded uninucleate cells in each culture was determined as in Fig 1. (a) Cultures of strains with the indicated checkpoint mutations were treated with 0.033% MMS for 3.5 hr prior to fixation. From 7 to 11 cultures of each genotype were tested, and the mean and 95% confidence limits of that mean are shown. (b) Experiments were performed as in a. Three strains of each genotype were tested either two (mad1) or three (wt, rad9 rad24, mad1 rad9 rad24) times each, and the mean and 95% confidence limits are shown. (c) Three cultures each of a mad2 strain and a rad9 rad24 strain were incubated with either 0.033% MMS (shaded) or 0.008% MMS (stippled) for 3.5 hr. Means and 95% confidence limits are shown. (d) Three cultures each of wild-type, mec1, and mec1 mad2 strains were incubated with 200 mM hydroxyurea for 3.5 hr prior to analysis. Means ±95% confidence limits are shown.

View larger version (36K): In this window In a new window Download PPT slide   Figure 3. The spindle-checkpoint-arrested cells in the first cell cycle after inactivation of Mcm2p. Cultures of () mcm2-1 (JRY7208), () mcm2-1 mad2 (JRY7214), () mcm2-1 rad24 (JRY7270), and (•) mcm2-1 mad2 rad24 (JRY7271) cells were arrested in G1 with -factor and then shifted to 37° for 45 min prior to release into -factor-free 37° medium. At the indicated time points, aliquots were removed and fixed for DAPI staining. At each time point, >200 cells of each strain were scored as unbudded (a), small-budded (b), large-budded uninucleate (c), large budded with dividing nucleus (not shown), or large-budded binucleate (d).

View larger version (36K): In this window In a new window Download PPT slide   Figure 4. The spindle-checkpoint-arrested cells in the first cell cycle after MMS exposure. Cultures of () wild-type (JRY2334), () mad2 (JRY7201), () rad9 rad24 (JRY7196), and (•) mad2 rad9 rad24 (JRY7204) cells were arrested in G1 with -factor and then released into -factor-free medium containing 0.033% MMS. At the indicated time points, aliquots were removed and fixed for DAPI staining. At each time point, >200 cells of each strain were scored as unbudded (a), small-budded (b), large-budded uninucleate (c), large-budded with dividing nucleus (not shown), or large-budded binucleate (d).

View larger version (35K): In this window In a new window Download PPT slide   Figure 5. The spindle-checkpoint-arrested cells in the first cell cycle after HU treatment. Cultures of () wild-type (JRY2334), () mec1 (JRY7274), and () mec1 mad2 (JRY7275) cells were arrested in -factor and then released into -factor-free medium containing 200 mM HU. At the indicated time points aliquots were removed and fixed for DAPI staining. At each time point, >200 cells of each strain were scored as unbudded (a), small-budded (b), large-budded uninucleate (c), large-budded with dividing nucleus (not shown), or large-budded binucleate (d).

View larger version (46K): In this window In a new window Download PPT slide   Figure 6. (a) A model of checkpoint pathways involved in responding to stalled replication and DNA damage. The MEC1 pathway is the cell's most sensitive DNA-responsive checkpoint. The response to DNA-damaging agents such as MMS and ionizing radiation requires RAD9, the RAD24 epistasis group, MEC1, and additional downstream protein kinases. In contrast, the response to the replication inhibitor hydroxyurea requires MEC1, but does not require RAD9 or the RAD24 epistasis group. The response to mitotic spindle disruptors requires MAD2 and occurs independently of MEC1. Both the MEC1-dependent checkpoint and the MAD2-dependent checkpoint block anaphase by stabilizing the anaphase inhibitor Pds1p. RAD9 and the RAD24 group activate this pathway in response to DNA damage and DNA structures resulting from stalled or aberrant DNA replication. (b) Enhancements to the model derived from results presented here. The MEC1-dependent pathway's requirement for RAD9 and the RAD24 group was abrogated only in pol mutants and hydroxyurea-treated cells. The hydroxyurea effect may be mediated by the impact of lowered deoxynucleotide pools on Pol. The MAD2-dependent spindle checkpoint was less sensitive than the MEC1-dependent pathway to most chromosome perturbations, but was equally or more sensitive than the MEC1-dependent pathway to certain aberrant chromosome structures such as those found in pol1-17 and mcm3-1 mutants.

The spindle checkpoint protein Mad2p arrested cells in response to replication mutations:
Activation of the DNA damage or the spindle checkpoint causes budding yeast cells to arrest with a large bud and an undivided nucleus. This type of arrest can be quantified by fluorescence microscopic determination of the percentage of large-budded uninucleate cells in a population. After creating yeast strains that harbored the replication mutations as well as mutations in the DNA damage checkpoint (rad9 rad24), the spindle checkpoint (mad2), or both, we incubated cultures of these strains at the restrictive temperature and then determined the percentage of large-budded uninucleate cells in each.

In strains in which both the DNA damage and DNA replication checkpoints were intact, mad2 significantly reduced the arrests of mcm3-1 and pol1-17 strains (Fig 1; checkpoint+ vs. mad2). Thus, the full arrest response to these mutations required Mad2p. The effect of the mad2 mutation on mcm2-1, cdc2-1, and orc2-1 strains was highly variable; thus whether the arrest responses to these mutations required Mad2p was not resolved by this experiment. However, Mad2p did contribute to the residual arrests of mcm2-1 and orc2-1 strains lacking the DNA damage checkpoint (Fig 1; compare the rad9 rad24 double mutant to the rad9 rad24 mad2 triple mutant). Furthermore, Mad2p, in conjunction with Rad9p and Rad24p, was required for the full arrest response to the cdc2-1 mutation (Fig 1; compare cdc2-1 with cdc2-1 mad2 rad9 rad24). Thus, Mad2p, and therefore presumably also the spindle checkpoint, detectably responded to the altered chromosomes generated in each of the replication mutants studied. These included mutations affecting both the initiation and elongation of replication and mutations causing cells to arrest either in mid-S-phase or in late S-G2.The combined mad2rad9rad24 mutations eliminated the accumulation of large-budded uninucleate cells in mcm2-1, mcm3-1, orc2-1, and cdc2-1 cultures held at the restrictive temperature (Fig 1; mad2 rad9 rad24 vs. no treatment). Thus, in the absence of these checkpoints, none of these replication mutations created a block to anaphase progression. Moreover, although MEC1 was intact in these strains, they failed to arrest. Therefore, inactivation of these replication proteins failed to detectably activate the MEC1-dependent, RAD9 RAD24-independent replication checkpoint.

The spindle checkpoint arrested cells in response to methyl methanesulfonate and hydroxyurea:
These results with replication mutants suggested that the spindle checkpoint may be sensitive to a wide range of DNA perturbations. To explore this range further, the ability of Mad2p to arrest cells treated with the DNA-damaging agent MMS was tested. No effect of mad2 was observed on the arrests of strains with an intact DNA damage checkpoint (Fig 2A), consistent with previous reports that MAD2 is not required for normal DNA damage responses (HOYT et al. 1991 ; HARDWICK et al. 1999 ). However, rad9 rad24 mutations only partially relieved the MMS-induced arrest. MMS-treated rad9 rad24 strains accumulated 45% large-budded uninucleate cells compared to the 10–14% observed in the untreated strains (Fig 2A, rad9 rad24 vs. no treatment). The mad2 mutation significantly reduced this residual accumulation to 26% (Fig 2A, rad9 rad24 vs. mad2 rad9 rad24). Thus, Mad2p was able to arrest a portion of MMS-treated cells and did so when the DNA damage checkpoint was not present.

Although Mad2p is not known to have a function outside of its role in the spindle checkpoint, we considered the possibility that the Mad2p-dependent arrest responses in our experiments reflected a spindle checkpoint-independent function of Mad2p. To test this possibility, we determined whether inactivation of a different component of the spindle checkpoint, Mad1p, would also relieve cell-cycle arrest responses to DNA damage. Similarly to mad2, the mad1 mutation significantly reduced the arrest response of rad9 rad24 cells to MMS (Fig 2B). Thus, two different spindle checkpoint genes each promoted cell-cycle arrest in response to DNA damage. The simplest interpretation of this finding is that the spindle checkpoint itself promotes cell-cycle arrest in response to DNA damage.

The cell-cycle arrest defect of rad9 rad24 cells relative to mad2 cells treated with MMS indicated that the spindle checkpoint was less efficient at mediating this response than was the DNA damage checkpoint. One explanation for this difference could be that the DNA damage checkpoint recognizes most MMS-induced lesions whereas the spindle checkpoint recognizes only a subset of them. For example, only a subset of the MMS-induced lesions might interfere with centromere function and hence activate the spindle checkpoint. If this model were correct, then it should be possible to reduce the concentration of MMS to a level at which most or all cells experience a lesion that activates the damage checkpoint, while only a subset of cells experience a lesion that activates the spindle checkpoint. Therefore, reduced MMS concentrations were evaluated for their effects on the arrests of mad2 and rad9 rad24 strains. A fourfold reduction in MMS concentration had no detectable effect on the arrest of the mad2 strains, presumably reflecting the ability of the DNA damage checkpoint to respond to low levels of MMS-induced damage. In contrast, the lower MMS concentration reduced the arrest of the rad9 rad24 strains from 45 to 25% (Fig 2C). These data suggested that only a subset of MMS-induced lesions could activate the spindle checkpoint.

To explore further the spindle checkpoint's ability to respond to DNA perturbations, the ability of mad2 to relieve the arrest response to an agent that stalls DNA replication was tested. HU stalls DNA replication by inhibiting ribonucleotide reductase, thereby depleting cells of deoxyribonucleotides. This condition activates the replication checkpoint, which is independent of RAD9 and RAD24 but requires MEC1. Therefore we tested the ability of mad2 to relieve arrest in hydroxyurea-treated mec1 cells, which lack the replication checkpoint. After 3.5 hr in 200 mM HU, 80% of wild-type cells or cells lacking the DNA damage checkpoint arrested (Fig 2D). The mec1 cells arrested less well, yet still exhibited more cell-cycle arrest than has been reported by others (Fig 2D, wild type vs. mec1; WEINERT et al. 1994 ; DESANY et al. 1998 ). At later time points mec1 relieved a larger portion of the HU-induced arrest (Fig 5 and data not shown), possibly explaining this discrepancy. mad2 significantly reduced the residual arrest observed in HU-treated mec1 cells (Fig 2D, mec1 vs. mec1 mad2). Since HU treatment stalls replication efficiently and creates only small amounts of DNA damage, this result indicated that incompletely replicated chromosomes per se may activate the spindle checkpoint.

The checkpoints acted in the first cell cycle following damage:
In the preceding experiments, cell-cycle arrest was quantified in cultures that were growing asynchronously at the time of insult. When cells fail to accumulate at the arrest point in such an experiment, they may do so either by passing through the arrest point (a checkpoint defect) or by failing to ever arrive at the arrest point (a viability defect). Experiments on synchronized cell populations allowed us to distinguish between these possibilities. This question was relevant to the mechanism of spindle-checkpoint-mediated arrest. Mitosis in the presence of damaged or partially replicated chromosomes can lead to aneuploidy and chromosome breakage. Since aneuploidy and small linear chromosomes can activate the spindle checkpoint in S. cerevisiae (WELLS and MURRAY 1996 ), it was possible that the spindle checkpoint responses to MMS, HU, or replication mutations required a prior mitosis in the presence of these insults. To test these ideas, we synchronized cells in G1 with -factor prior to the time of insult and then quantified the number of cells both arriving at and passing through the arrest point.

In the first synchronized-cell experiment, the mcm2-1 mutation was used to activate the checkpoints. RAD9 was left intact in the strains used in this experiment since rad9 in combination with other mutations caused low viability that prevented synchronization. The mcm2-1, mcm2-1 mad2, mcm2-1 rad24, and mcm2-1 mad2 rad24 strains were synchronized in G1 at the permissive temperature and then released from the G1 block into restrictive-temperature medium. Under these conditions, bud emergence and bud growth occurred with similar kinetics in the mcm2-1 and mcm2-1 mad2 strains, but were slightly slower and/or less synchronous in the mcm2-1 rad24 and mcm2-1 rad24 mad2 strains (Fig 3, a and b). By 100 min postrelease, 80% of the mcm2-1 cells with both checkpoints intact accumulated at the large-budded uninucleate stage (Fig 3C, mcm2-1). By contrast, only 12% of mcm2-1 cells lacking both the DNA damage and spindle checkpoints accumulated at this stage (Fig 3C, mcm2-1 mad2 rad24). Subsequent increases in the proportions of binucleate and unbudded cells in this strain demonstrated that these cells were passing through the arrest point rather than failing to arrive at it (Fig 3D, mcm2-1 mad2 rad24; Fig 3A, mcm2-1 mad2 rad24). Thus, the preanaphase arrest of mcm2-1 cells was due solely to the responses of the DNA damage and spindle assembly checkpoints to the mcm2-1 mutation.

Furthermore, 55% of the mcm2-1 mad2 cells accumulated at the arrest point, indicating that the DNA damage checkpoint arrested this proportion of cells in the first cycle following inactivation of Mcm2p. Similarly, 45% of the mcm2-1 rad24 cells accumulated at the arrest point, indicating that the spindle checkpoint arrested this proportion of cells in the first cycle following inactivation of Mcm2p. Thus, both DNA damage checkpoint- and spindle checkpoint-mediated arrest occurred in the first cycle after inactivation of Mcm2p.

To assess checkpoint responses in the first cell cycle after exogenously induced DNA damage, the experiment was repeated using MMS rather than mcm2-1 to activate the checkpoints. Wild-type, rad9 rad24, mad2, and mad2 rad9 rad24 cells were synchronized in G1 and then released into medium containing MMS. In this experiment, bud emergence and growth occurred with similar kinetics in all the strains (Fig 4, a and b). In cells with both checkpoints intact (WT in Fig 4) and in cells with just the DNA damage checkpoint intact (mad2), >80% of the population arrested (Fig 4C). In cells with only the spindle checkpoint intact (rad9 rad24), 55% of the population arrested (Fig 4C). However, in cells with neither checkpoint intact (mad2 rad9 rad24), only 14% of the population arrested (Fig 4C). Cell-cycle progression in the triple mutant cells was confirmed by subsequent increases in unbudded cells, small-budded cells, and large-budded binucleate cells (Fig 4, a, b, and d). Thus both DNA damage checkpoint activation and spindle checkpoint activation occurred in the first cycle after treatment with MMS.

To explore further the role of MAD2 in the arrest of mec1 cells in HU, we monitored the cell-cycle progression of wild-type, mec1, and mec1 mad2 cells released from the -factor block into HU-containing medium. In this experiment, all three strains behaved similarly up until 120 min after release into HU-containing medium, accumulating from 50 to 70% large-budded uninucleate cells. Thus, the initial cell-cycle arrest response to HU occurred as efficiently in the mec1 mad2 strain as in the mec1 strain (Fig 5C). However, by 150 min, the percentage of large-budded uninucleate cells in the wild-type strain continued to increase, whereas this percentage remained constant in the mec1 strain, and in the mec1 mad2 strain it declined. The rise in the percentage of binucleate cells during this time indicated that the loss of uninucleate cells was due to nuclear division (Fig 5D). Thus the spindle checkpoint was able to block nuclear division in a portion of HU-treated mec1 cells. Since the initial arrest response to HU occurred as efficiently in the mec1 mad2 strain as it did in the mec1 strain, these results suggested a role for MAD2 in maintaining the anaphase block in HU-treated cells rather than in establishing the block.

The spindle checkpoint contributed to the growth rate and DNA damage resistance of rad9 rad24 mutants:
Cell-cycle delay in the presence of DNA damage often preserves cell viability, and indeed many checkpoint mutants have been identified by their sensitivity to DNA-damaging agents. Therefore we tested whether the mad2 mutation affected survival of MMS treatment. In tests for growth on solid MMS-containing medium, the mad2 mutation did not significantly affect the survival of either wild-type cells or rad9 rad24 cells (data not shown). Thus spindle checkpoint function did not enhance survival of cells during chronic exposure to MMS. However, to test whether spindle checkpoint function could rescue cells from acute exposure to MMS, we treated wild-type, mad2, rad9 rad24, and mad2 rad9 rad24 cells in liquid culture with MMS, removing cells at various times to test viability. This analysis revealed that the viability of the mad2 rad9 rad24 strains (39 ± 12% relative to rad9 rad24 viability) was significantly lower than that of the rad9 rad24 strains (100 ± 12%) at all times after MMS treatment. Thus, spindle checkpoint function did contribute to the viability of rad9 rad24 cells during short-term exposure to MMS.

Relative to wild-type yeast strains, the mad2 rad9 rad24 strains formed small colonies. To quantify this effect, we determined the doubling times of wild-type, mad2, rad9 rad24, and mad2 rad9 rad24 strains. The growth of the mad2 strain was indistinguishable from wild type (99 ± 3.6% of wild-type doubling time). The rad9 rad24 strain grew more slowly (107 ± 3.3% of wild-type doubling time), and the mad2 mutation enhanced this defect (117 ± 3.4% of wild-type doubling time). Since the cell-cycle data presented above indicated that the checkpoints that these genes control respond to aberrant replication, the slower doubling times of these strains likely reflected a requirement for checkpoint response to aberrant replication events during normal cell divisions.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We investigated the relative contributions of the DNA damage, DNA replication, and spindle assembly checkpoints to the preanaphase arrest responses that occur in S. cerevisiae cells exposed to a variety of chromosome-perturbing conditions. These conditions included mutations affecting the origin recognition complex, Mcm proteins, DNA polymerase , and DNA polymerase , as well as nucleotide depletion and exposure to a DNA-damaging agent. Several findings arose from this work. First, the spindle checkpoint was able to contribute to the arrest responses to all of the conditions tested. Second, spindle checkpoint function was essential for cells to achieve a full arrest response to mutations affecting Mcm3p and Pol1p. Third, the RAD9-independent, MEC1-dependent replication checkpoint made a detectable contribution to the arrests of pol1-17 mutants and HU-treated cells, but not to any of the other conditions tested.

Identification of the checkpoints that become activated under a certain condition should offer insight into the molecular defects associated with that condition. In the case of the spindle checkpoint, the activating molecular defect has been well characterized. A large body of evidence indicates that kinetochores not under tension from the mitotic spindle cause this checkpoint to become activated (RIEDER et al. 1995 ; CHEN et al. 1996 ; LI and NICKLAS 1997 ). Some of this evidence derives from studies of S. cerevisiae, in which unreplicated chromosomes and chromosomes with defects in sister chromatid cohesion both activate the spindle checkpoint (CASTANO et al. 1996 ; SKIBBENS et al. 1999 ; HANNAH et al. 2001 ; MAYER et al. 2001 ; STERN and MURRAY 2001 ). Similarly, incomplete DNA replication in Drosophila embryos results in a BUB1-dependent mitotic arrest (GARNER et al. 2001 ). Since some of our experimental conditions caused cells to arrest in early S-phase (the DNA polymerase mutations, hydroxyurea treatment, and MMS treatment), mono-oriented kinetochores on unreplicated centromeres were likely sources of spindle checkpoint activation in these cells. However, the orc and mcm mutations caused cells to arrest in late S-phase or G2. Since unreplicated centromeres were less likely to be present in these cells, spindle checkpoint activation in these mutants may signal defects in sister chromatid cohesion or may reflect heretofore unsuspected roles of ORC and MCM proteins in promoting the bipolar attachment of chromosomes to the mitotic spindle.

The lesions recognized by the DNA damage and DNA replication checkpoints have not been determined as precisely as that of the spindle checkpoint. It has been proposed that single-stranded DNA (ssDNA) is the lesion recognized by the DNA damage checkpoint. Correlations between the presence of ssDNA and DNA damage checkpoint activation support this model (LYDALL and WEINERT 1995 ; LEE et al. 1998 ). However, since ssDNA is an expected intermediate in the processing of most types of DNA damage, this correlation is compatible with other models as well. Therefore, we interpret the DNA damage checkpoint activation in our experiments to signal the presence of some DNA lesion that may be an intermediate in DNA damage processing that may or may not be ssDNA.

The molecular defect responsible for activation of the MEC1-dependent, RAD9-independent DNA replication checkpoint is widely thought to be ongoing or stalled replication forks (reviewed by LOWNDES and MURGUIA 2000 ). However, our results, in conjunction with data from other labs, suggest that this view may need refinement. Specifically, the preanaphase arrest responses to a variety of conditions in which replication is stalled or ongoing require RAD9; thus, these conditions all fail to significantly activate the RAD9-independent replication checkpoint. For example, the preanaphase arrest responses to the cdc2-1 and cdc2-2 mutations, which inactivate DNA Pol, require RAD9 (WEINERT and HARTWELL 1993 ; P. GARBER and J. RINE, unpublished results). Similarly, although two-dimensional gel analyses indicate that arrested orc2-1, orc5-1, and mcm2-1 mutant cells contain replication forks, these mutants require RAD9 for preanaphase arrest (LIANG et al. 1995 ; LEI et al. 1997 ). Furthermore, yeast cells harboring an origin-deficient artificial chromosome require RAD9 to stably maintain the chromosome (VAN BRABANT et al. 2001 ). Thus, neither the stalled forks in orc and mcm mutants nor the ongoing replication forks of the artificial chromosome significantly activate the RAD9-independent, MEC1-dependent replication checkpoint.

In our experiments and in those of others, the replication checkpoint contributed to the arrests of deoxyribonucleotide-depleted cells and cells lacking Pol DNA polymerase activity, but not to the other conditions tested (WEINERT and HARTWELL 1993 ; WEINERT et al. 1994 ; P. GARBER and J. RINE, unpublished results; Fig 2 and Fig 3C). Since stalled replication structures are likely to be present in orc, mcm, and pol mutants as well as in pol mutants and HU-treated cells, we propose that the replication checkpoint, rather than recognizing stalled replication structures per se, recognizes a DNA lesion that is specific to cells lacking Pol DNA polymerase activity and cells lacking deoxyribonucleotides. When considering what lesion might be common to these two conditions, we note that pol mutants and HU-treated cells are each compromised for Pol DNA polymerase activity. Thus, one possibility is that the replication checkpoint recognizes an intermediate in DNA replication that persists only when Pol's DNA polymerase is not active. Studies in Xenopus laevis extracts indicate that RNA primers activate the replication checkpoint. Hence, we suggest that the unextended RNA primers that might accumulate in pol mutants activate the replication checkpoint in S. cerevisiae. A model that takes into account this more limited role of the replication checkpoint in responding to replication problems is presented in Fig 6B.

Although the spindle checkpoint was capable of arresting cells in response to all of the chromosome-perturbing conditions we employed, the responses to mcm2-1, orc2-1, HU, and MMS did not require spindle checkpoint function as long as the MEC1-dependent checkpoints were intact. Thus the MEC1-dependent checkpoints responded readily to these conditions and mediated a maximal arrest response to them whether or not the spindle checkpoint was present. By contrast, the MEC1-dependent checkpoints responded less readily to the mcm3-1 and pol1-17 mutations, and under these conditions the spindle checkpoint contributed to arrest even when the MEC1-dependent checkpoints were intact. Thus chromosome perturbations vary in the degree to which they activate each of these checkpoints. On one end of the spectrum lie chromosome perturbations such as the presence of multiple linear minichromosomes, which activate the spindle checkpoint without apparently activating the DNA damage checkpoint (WELLS and MURRAY 1996 ). On the other end of the spectrum lie chromosome perturbations such as low levels of MMS and the cdc13-1 mutation, which activate the DNA damage checkpoint while causing little or no spindle checkpoint activation (HARDWICK et al. 1999 ; P. GARBER and J. RINE, unpublished results; Fig 2). The mcm3-1 and pol1-17 mutations fall between these two extremes. The ability to thus classify the checkpoint responses to a particular mutation may lend insight into the underlying molecular phenotype of the mutants.

The majority of cancers display a chromosome instability phenotype (LENGAUER et al. 1998 ). While some cancer cells are defective for spindle checkpoint function and/or the expression of spindle checkpoint genes (LI and BENEZRA 1996 ; CAHILL et al. 1998 ), the molecular basis for chromosome instability in most cancers remains unknown (LENGAUER et al. 1998 ). Therefore, discovery of the lesions responsible for chromosome instability remains a critical step in understanding tumorigenesis. The ability of a variety of replication perturbations to activate the spindle checkpoint suggests that the proper attachment of chromosomes to the mitotic spindle can be disrupted by defects in replication. If so, then replication defects could contribute directly to the segregation defects underlying much of the chromosome instability in cancer cells. Furthermore, the response of the spindle checkpoint to the same types of perturbations as the DNA damage checkpoint suggests that mutations affecting these two pathways would synergistically contribute to genome instability. Hence, tumor cells with spindle checkpoint mutations may be more sensitive to chemotherapeutic agents that damage DNA than are cells with functional spindle checkpoints.

	ACKNOWLEDGMENTS

We thank Peter Burgers, Judith Campbell, Andrew Murray, Rodney Rothstein, Bik Tye, and Ted Weinert for yeast strains; Paul Kaufman and Judith Sharp for helpful discussions and protocols; Christopher Beh, Orna Cohen-Fix, Alexa Franco, Lena Hwang, Paul Kaufman, Jim Keck, Ann Kirchmaier, Judith Sharp, and Jeremy Thorner for comments on this manuscript; and anonymous reviewers for suggestions that improved both the substance and presentation of the manuscript. This work was supported by a grant from the National Institutes of Health (GM-31105) and by core support from a National Institute of Environmental Health Sciences Mutagenesis Center grant (P30 ES01896).

Manuscript received July 31, 2001; Accepted for publication March 4, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BELL, S. B., R. KOBAYASHI, and B. STILLMAN, 1993  Yeast origin recognition complex functions in transcription silencing and DNA replication. Science 262:1844-1849[Abstract/Free Full Text].

BOULET, A., M. SIMON, G. FAYE, G. A. BAUER, and P. M. J. BURGERS, 1989  Structure and function of the Saccharomyces cerevisiae CDC2 gene encoding the large subunit of DNA polymerase III.. EMBO J. 8:1849-1855[Medline].

BUDD, M. and J. L. CAMPBELL, 1987  Temperature-sensitive mutations in the yeast DNA polymerase I gene. Proc. Natl. Acad. Sci. USA 84:2838-2842[Abstract/Free Full Text].

BUDD, M. E. and J. L. CAMPBELL, 1993  DNA polymerases and are required for chromosomal replication in Saccharomyces cerevisiae. Mol. Cell. Biol. 13:496-505[Abstract/Free Full Text].

BUDD, M. E., K. D. WITTRUP, J. E. BAILEY, and J. L. CAMPBELL, 1989  DNA polymerase I is required for premeiotic DNA replication and sporulation but not for X-ray repair in Saccharomyces cerevisiae. Mol. Cell. Biol. 9:365-376[Abstract/Free Full Text].

CAHILL, D. P., C. LENGAUER, J. YU, G. J. RIGGINS, and J. K. WILLSON et al., 1998  Mutations of mitotic checkpoint genes in human cancers. Nature 392:300-303[Medline].

CASTANO, R. B., P. M. BRZOSKA, B. U. SADOFF, H. CHEN, and M. F. CHRISTMAN, 1996  Mitotic chromosome condensation in the rDNA requires TRF4 and DNA topoisomerase I in Saccharomyces cerevisiae.. Genes Dev. 10:2564-2576[Abstract/Free Full Text].

CHEN, R.-H., J. C. WATERS, E. D. SALMON, and A. W. MURRAY, 1996  Association of spindle assembly checkpoint component XMAD2 with unattached kinetochores. Science 274:242-246[Abstract/Free Full Text].

DESANY, B. A., A. A. ALCASABAS, J. B. BACHANT, and S. J. ELLEDGE, 1998  Recovery from DNA replicational stress is the essential function of the S-phase checkpoint pathway. Genes Dev. 12:2956-2970[Abstract/Free Full Text].

FOIANI, M., A. PELLICIOLI, M. LOPES, C. LUCCA, and M. FERRARI et al., 2000  DNA damage checkpoints and DNA replication controls in Saccharomyces cerevisiae.. Mutat. Res. 451:187-196[Medline].

FOSS, M., F. J. MCNALLY, P. LAURENSON, and J. RINE, 1993  Origin recognition complex (ORC) in transcriptional silencing and DNA replication in S. cerevisiae.. Science 262:1838-1844[Abstract/Free Full Text].

GARNER, M., S. VAN KREEVELD, and T. T. SU, 2001  mei-41 and bub1 block mitosis at two distinct steps in response to incomplete DNA replication in Drosophila embryos. Curr. Biol. 11:1595-1599[Medline].

GIBSON, S. I., R. T. SUROSKY, and B. TYE, 1990  The phenotype of the minichromosome maintenance mutant mcm3 is characteristic of mutants defective in DNA replication. Mol. Cell. Biol. 10:5707-5720[Abstract/Free Full Text].

HANNAH, J. S., E. S. KROLL, V. LUNDBLAD, and F. A. SPENCER, 2001  Saccharomyces cerevisiae CTF18 and CTF4 are required for sister chromatid cohesion. Mol. Cell. Biol. 21:3144-3158[Abstract/Free Full Text].

HARDWICK, K. G., R. LI, C. MISTROT, R-H. CHEN, and P. DANN et al., 1999  Lesions in many different spindle components activate the spindle checkpoint in budding yeast. Genetics 152:509-518[Abstract/Free Full Text].

HOYT, M. A., 2001  A new view of the spindle checkpoint. J. Cell Biol. 154:909-912[Abstract/Free Full Text].

HOYT, M. A., L. TOTIS, and B. T. ROBERTS, 1991  S. cerevisiae genes required for cell-cycle arrest in response to loss of microtubule function. Cell 66:507-517[Medline].

LEE, S. E., J. K. MOORE, A. HOLMES, K. UMEZU, and R. D. KOLODNER et al., 1998  Saccharomyces Ku70, Mre11/Rad50, and RPA proteins regulate adaptation to G2/M arrest after DNA damage. Cell 94:399-409[Medline].

LEI, M., Y. KAWASAKI, M. R. YOUNG, M. KIHARA, and A. SUGINO et al., 1997  Mcm2 is a target of regulation by Cdc7-Dbf4 during the initiation of DNA synthesis. Genes Dev. 11:3365-3374[Abstract/Free Full Text].

LENGAUER, C., K. W. KINZLER, and B. VOGELSTEIN, 1998  Genetic instabilities in human cancers. Nature 396:643-649[Medline].

LI, X. and R. B. NICKLAS, 1997  Tension-sensitive kinetochore phosphorylation and the chromosome distribution checkpoint in praying mantid spermatocytes. J. Cell Sci. 110:537-545[Abstract].

LI, Y. and R. BENEZRA, 1996  Identification of a human mitotic checkpoint gene: hsMad2. Science 274:246-248[Abstract/Free Full Text].

LIANG, C., M. WEINREICH, and B. STILLMAN, 1995  ORC and Cdc6p interact and determine the frequency of initiation of DNA replication in the genome. Cell 81:667-676[Medline].

LOWNDES, N. F. and J. R. MURGUIA, 2000  Sensing and responding to DNA damage. Curr. Opin. Genet. Dev. 10:17-25[Medline].

LYDALL, D. and T. WEINERT, 1995  Yeast checkpoint genes in DNA damage processing: implications for repair and arrest. Science 270:1488-1491[Abstract/Free Full Text].

MAYER, M. L., S. P. GYGI, R. AEBERSOLD, and P. HIETER, 2001  Identification of RFC(Ctf18p, Ctf8p, Dcc1p): an alternative RFC complex required for sister chromatid cohesion in S. cerevisiae.. Mol. Cell 7:959-970[Medline].

PFLUMM, M. F. and M. R. BOTCHAN, 2001  Orc mutants arrest in metaphase with abnormally condensed chromosomes. Development 128:1697-1707[Abstract].

RIEDER, C. L., R. W. COLE, A. KHODJAKOV, and G. SLUDER, 1995  The checkpoint delaying anaphase in response to chromosome monoorientation is mediated by an inhibitory signal produced by unattached kinetochores. J. Cell Biol. 130:941-948[Abstract/Free Full Text].

ROSE, M. D., F. WINSTON and P. HIETER, 1989 Laboratory Course Manual for Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SKIBBENS, R. V., L. B. CORSON, D. KOSHLAND, and P. HIETER, 1999  Ctf7p is essential for sister chromatid cohesion and links mitotic chromosome structure to the DNA replication machinery. Genes Dev. 13:307-319[Abstract/Free Full Text].

STERN, B. M. and A. W. MURRAY, 2001  Lack of tension at kinetochores activates the spindle checkpoint in budding yeast. Curr. Biol. 11:1462-1467[Medline].

VAN BRABANT, A. J., C. D. BUCHANAN, E. CHARBONNEAU, W. L. FANGMAN, and B. J. BREWER, 2001  An origin-deficient yeast artificial chromosome triggers a cell-cycle checkpoint. Mol. Cell 7:705-713[Medline].

WEINERT, T. A. and L. H. HARTWELL, 1993  Cell cycle arrest of cdc mutants and specificity of the RAD9 checkpoint. Genetics 134:63-80[Abstract].

WEINERT, T. A., G. L. KISER, and L. H. HARTWELL, 1994  Mitotic checkpoint genes in budding yeast and the dependence of mitosis on DNA replication and repair. Genes Dev. 8:652-665[Abstract/Free Full Text].

WELLS, A. E. and A. W. MURRAY, 1996  Aberrantly segregating centromeres activate the spindle assembly checkpoint in budding yeast. J. Cell Biol. 33:75-84.

YAN, H., S. I. GIBSON, and B. TYE, 1991  Mcm2 and Mcm3, two proteins important for ARS activity, are related in structure and function. Genes Dev. 5:944-957[Abstract/Free Full Text].

This article has been cited by other articles:

				A. M. Lyndaker, T. Goldfarb, and E. Alani Mutants Defective in Rad1-Rad10-Slx4 Exhibit a Unique Pattern of Viability During Mating-Type Switching in Saccharomyces cerevisiae Genetics, August 1, 2008; 179(4): 1807 - 1821. [Abstract] [Full Text] [PDF]

				E. D. Strome, X. Wu, M. Kimmel, and S. E. Plon Heterozygous Screen in Saccharomyces cerevisiae Identifies Dosage-Sensitive Genes That Affect Chromosome Stability Genetics, March 1, 2008; 178(3): 1193 - 1207. [Abstract] [Full Text] [PDF]

				F. Dotiwala, J. Haase, A. Arbel-Eden, K. Bloom, and J. E. Haber The yeast DNA damage checkpoint proteins control a cytoplasmic response to DNA damage PNAS, July 3, 2007; 104(27): 11358 - 11363. [Abstract] [Full Text] [PDF]

				C. Clemenson and M.-C. Marsolier-Kergoat The Spindle Assembly Checkpoint Regulates the Phosphorylation State of a Subset of DNA Checkpoint Proteins in Saccharomyces cerevisiae Mol. Cell. Biol., December 15, 2006; 26(24): 9149 - 9161. [Abstract] [Full Text] [PDF]

				D. Blake, B. Luke, P. Kanellis, P. Jorgensen, T. Goh, S. Penfold, B.-J. Breitkreutz, D. Durocher, M. Peter, and M. Tyers The F-Box Protein Dia2 Overcomes Replication Impedance to Promote Genome Stability in Saccharomyces cerevisiae Genetics, December 1, 2006; 174(4): 1709 - 1727. [Abstract] [Full Text] [PDF]

				D. G. Gibson, S. P. Bell, and O. M. Aparicio Cell cycle execution point analysis of ORC function and characterization of the checkpoint response to ORC inactivation in Saccharomyces cerevisiae Genes Cells, June 1, 2006; 11(6): 557 - 573. [Abstract] [Full Text] [PDF]

				N. Dhillon, M. Oki, S. J. Szyjka, O. M. Aparicio, and R. T. Kamakaka H2A.Z Functions To Regulate Progression through the Cell Cycle Mol. Cell. Biol., January 15, 2006; 26(2): 489 - 501. [Abstract] [Full Text] [PDF]

				J. A. Daniel, B. E. Keyes, Y. P. Y. Ng, C. O. Freeman, and D. J. Burke Diverse Functions of Spindle Assembly Checkpoint Genes in Saccharomyces cerevisiae Genetics, January 1, 2006; 172(1): 53 - 65. [Abstract] [Full Text] [PDF]

V. Archambault, A. E. Ikui, B. J. Drapkin, and F. R. Cross Disruption of Mechanisms That Pre