ABSTRACT
Checkpoint controls exist in eukaryotic cells to ensure that cells do not enter mitosis in the presence of DNA damage or unreplicated chromosomes. In Schizosaccharomyces pombe many of the checkpoint genes analysed to date are required for both the DNA damage and the replication checkpoints, an exception being chk1. We report here on the characterization of nine new methylmethane sulphonate (MMS)-sensitive S.pombe mutants, one of which is defective in the DNA damage checkpoint but not the replication checkpoint. We have cloned and sequenced the corresponding gene. The predicted protein is most similar to the Saccharomyces cerevisiae Rad9 protein, having 46% similarity and 26% identity. The S.pombe protein, which we have named Rhp9 (
Multiple checkpoints exist in cells to prevent mitosis from occurring in the presence of DNA damage or incomplete DNA replication (1 -6 ). DNA damage (radiation) checkpoints act both at G1/S and G2/M to prevent entry of cells into S phase and mitosis respectively (e.g. 2,7), while the S/M checkpoint prevents mitosis in the presence of unreplicated DNA (see for example 8 ). In Saccharomyces cerevisiae G2 arrest in the presence of DNA damage requires the RAD9, RAD17, RAD24, MEC1/ESR1, RAD53 and MEC3 genes (9 ,10 ). In addition, the RAD9, RAD24 and RAD53 genes are also required for S phase delay following irradiation of cells in G1 (7 ,11 ,12 ). Two of the genes, MEC1 and RAD53, are also involved in a separate checkpoint which is required for the prevention of entry into mitosis in the presence of incomplete DNA replication (10 ,11 ).
In Schizosaccharomyces pombe the DNA damage checkpoint requires the functions of at least seven genes, namely rad1, rad3, rad9, rad17, rad26, hus1 and chk1 (3 ,5 ,6 ,13 ). All except chk1 are also required for the S/M checkpoint, where entry into mitosis is dependent on completion of DNA replication (6 ,13 ).
A number of the S.cerevisiae and S.pombe checkpoint genes are related in sequence, e.g. RAD17 and rad1 (14 ), RAD24 and rad17 (15 ) and MEC1 and rad3 (16 ). The proteins encoded by these genes have been proposed to act as a guardian complex (15 ), however, the precise role of the gene products in the checkpoint process remains obscure. In addition to their checkpoint roles, the S.cerevisiae RAD17 and RAD24 gene products have been proposed to act in damage processing in cdc13-arrested cells, with the RAD9 gene product having a regulatory role (14 ).
Here we report on the isolation of nine methylmethane sulphonate (MMS)-sensitive S.pombe mutants, six of which have an altered response to the DNA replication inhibitor hydroxyurea. In particular, we describe the further characterization of one new mutant, rhp9-64, which is defective in the DNA damage checkpoint. We have cloned and sequenced the rhp9 gene and show that the gene product has 26% identity and 46% similarity to the S.cerevisiae Rad9 protein. The Rhp9 protein contains a domain present in the human and mouse BRCA1 proteins and 53BP1, a p53 binding protein. The rhp9 null allele displays a slightly increased response to both UV and ionizing radiation compared with the original allele, but is less sensitive to both types of radiation than a checkpoint rad null allele. The rhp9 gene has a DNA damage checkpoint role but is not required for the S/M checkpoint. Epistasis analysis indicates that rhp9 functions in a process involving the checkpoint rad genes and, specifically, the chk1 gene. Double mutant analysis also indicates that rhp9 is likely to be required for an additional process independent of the arrest phenotype, possibly suggesting a damage processing role related to that reported for S.cerevisiae RAD9.
The S.pombe plasmid pUR19 and the genomic library used in this study have been described elsewhere (17 ). The S.pombe rad3-containing plasmid (pSUB41) was a gift from S.Subramani (18 ). Schizosaccharomyces pombe strains used are shown in Table 1 . Procedures and media used for routine growth and maintenance of S.pombe strains and general S.pombe genetic techniques were as described in our previous work (19 ). General molecular biology techniques were as described by Sambrook et al. (20 ). MMS was used at a concentration of 50 [mu]l/l (21 ).
Table 1
107 cells (sp.011) were washed once in water and once in 0.1 M sodium citrate, pH 5.5, and then resuspended in 1 ml 0.1 M sodium citrate. An aliquot of 110 [mu]l 2 mg/ml N-methyl-N'-nitro-N-nitrosoguanine (MNNG) was added and the cells were incubated at 30oC for 30 min. The cells were then diluted and washed twice in 0.1 M sodium citrate and twice in YES medium. Mutagenized cells were resuspended in 5 ml YES plus 1 ml glycerol and incubated at 30oC for 20 h.
Cultures of (50 ml) exponentially growing cells were grown overnight to a cell density of ~5 * 106/ml. The cultures were split into three aliquots of 10 ml and treated with either 400 Gy irradiation or 20 mM hydroxyurea (HU), with the third aliquot acting as control. Samples of 100 [mu]l were taken at 30 min intervals into 1 ml methanol. Cells were then scored for per cent septa and `cut' phenotype following staining with DAPI and calcofluor (22 ). Synchronous cultures were prepared on 7.5-30% lactose gradients as described by Carr and Murray (23 ). Cells were treated with a range of doses of ionizing radiation using a Gammacell 1000 137Cs source (12 Gy/min) as described by Al-Khodairy et al. (24 ).
In order to identify new recombination or checkpoint mutants we initiated a search for novel S.pombe mutants sensitive to the alkylating agent MMS. Sp.011 S.pombe cells (rad+ mms+) were mutagenized using MNNG to produce a killing rate of 98%. Approximately 10 000 independent colonies were either picked or replica plated to YES plates containing MMS (50 [mu]l/l). Following this selection, 120 colonies were rescreened on MMS-containing medium, resulting in nine MMS-sensitive mutants. These were back-crossed at least three times with the rad+ mms+ strain sp.011 or sp.012.
Previous screens for S.pombe checkpoint mutants have resulted in the isolation of a disproportionately large number of alleles of rad3 (see for example 6 ). To eliminate any rad3 alleles from our screen, all the MMS-sensitive mutants were transformed with pSUB41, a rad3-containing plasmid (18 ). Of the nine mutants, the radiation sensitivity of one mutant (mms70) was complemented fully by pSUB41 (data not shown), indicating that this mutant is likely to be an allele of rad3.
An S.pombe genomic library (17 ) was used to transform the mms64 strain, sp.353, to uracil prototrophy. A total of 30 000 ura+ transformants were pooled and subjected to three rounds of increasing doses of UV irradiation as previously described (19 ). Individual colonies were tested for co-instability of the rad+ and ura+ phenotypes. Plasmid pGa7 (Fig. 3 A) was isolated from a radiation-resistant colony and on retransformation into sp.353 was shown to complement the radiation-sensitive phenotype of mms64 (Fig. 2 c). Analysis of a stable integrant of pGA7 in mms64 indicated that the plasmid integrated at the mms64 locus.
A 2 kb region of the rhp9 coding sequence (from the 5' EcoRV site to the 3' HindIII site within the rhp9 coding sequence) was replaced by the S.pombe ura4 gene (33 ). A disruption fragment of 3.6 kb was excised by digestion with EcoRI (see Fig. 3 A) and transformed into the haploid strain sp.012. ura+ transformants were checked by Southern blotting (data not shown) for homologous replacement of the wild-type rhp9 gene by the disruption construct. Comparison of Figures 2 a and 5 a indicates that the rhp9-d null allele displays a somewhat increased level of sensitivity to UV radiation over that of the original rhp9-64 allele. The PstI-BglII fragment of pGa7 when cloned into pAL19 was able to complement the radiation sensitivity of the rhp9-d null allele to approaching wild-type levels (data not shown).
To determine whether rhp9 acts in a process involving other radiation checkpoint gene products or in any of the previously characterized repair pathways, epistasis analysis was carried out. Analysis of double mutants with the rad2-d and rad13-d null alleles indicates that rhp9 functions in a process other than that involving rad2 (the S.pombe equivalent of human FEN1; 34 ,35 ) and other than an excision repair pathway (36 ; Fig. 5 a and b).
Having established a role for Rhp9 in the damage checkpoint, double mutants were created with several of the checkpoint rad mutants known to be defective in both the S phase and damage checkpoints. rhp9-d double mutants with either rad3-d (3 ) or rad26-d (6 ) showed no increase in radiation sensitivity over that of the single rad3-d or rad26-d mutants (Fig. 5 c and d), indicating that rhp9 functions in a process involving the checkpoint rad genes. Double mutants with other checkpoint rad mutants, namely rad9-d and rad17-d, (15 ,26 ) also showed no increase in sensitivity over that of the most sensitive single mutant (data not shown). Another mutant generally classified with the checkpoint rad mutants is hus1 (27 ). Interestingly, an rhp9-d hus1-d double mutant was more sensitive than the most sensitive single mutant, hus1-d (5 ,27 ; Fig. 5 e), suggesting that hus1-d retains a function which is defective in the rad3-d, rad9-d, rad17-d and rad26-d mutants.
To further define the process involving rhp9, two additional rhp9-d double mutants were created. A double mutant with chk1-d (which is defective in the damage checkpoint but not the S phase checkpoint; 6 ) demonstrated no increase in radiation sensitivity over that of rhp9-d (Fig. 5 f). The rad26-T12 mutant has previously been characterized as having no obvious defect in the mitotic checkpoint but displays radiation and HU sensitivity (6 ). This has been interpreted as a defect in a DNA damage tolerance process (6 ,37 ). A rad26-T12 rhp9-d double mutant is more sensitive to radiation than the most sensitive single mutant, rhp9-d (data not shown), suggesting that rhp9 is not required for the damage tolerance process.
These results suggest that in addition to their requirement in the S phase and damage checkpoints, the checkpoint rad genes rad3, rad9, rad17 and rad26 may also be involved in an additional process which is independent of hus1 but which requires rhp9. To determine whether chk1 is required for this process, a further double mutant was created, chk1-d hus1-d. The double mutant displayed no increase in radiation sensitivity over that of the hus1-d mutant, unlike the rhp9-d hus1-d double mutant, indicating that the additional process is independent of chk1.
The chk1 gene has been shown to be required to prevent mitosis following cell cycle arrest in cdc10 and cdc17 mutants as well as following exposure to DNA damaging agents (38 ). To investigate whether rhp9 is required in a similar manner, double mutants were created with cdc10-V50 (V.Simanis, personal communication), cdc17-K42 and cdc22-M45 (39 ). An rhp9 cdc22 double mutant displayed a `cdc' phenotype after 4 h incubation at 36oC (Fig. 6 ), consistent with the HU phenotype shown in Figure 4 . In contrast, rhp9 cdc10 and rhp9 cdc17 double mutants displayed a `cut' phenotype. These results are consistent with those obtained with chk1 (38 ).
We have isolated a number of new MMS-sensitive mutants, some of which are also altered in response to the DNA replication inhibitor HU. Of these mutants, two, which show no growth on HU, are likely to be defective in the DNA replication checkpoint. Three further mutants, which form microcolonies on HU, are likely to be defective in some other aspect of S phase arrest. We have cloned two of the corresponding wild-type genes by complementation of the radiation-sensitive phenotypes. Sequence analysis indicates that the gene which fully complements the radiation sensitivity of mms8 (data not shown) is the S.pombe rad4/cut5 gene (40 ,41 ).
The mms64 mutant phenotype is complemented fully by a novel gene which we have called rhp9. The rhp9 gene product, which is identical (apart from a couple of amino acid differences) to a recently identified sequence (Crb2) in the GenBank database (accession no. D86478) is most homologous to the S.cerevisiae Rad9 protein and contains motifs resembling the BRCT domains in human and mouse BRCA1 proteins. Human BRCA1 was identified as a breast cancer susceptibility gene, with mutations in BRCA1 accounting for 45% of families with high incidence of breast cancer and for 80-90% of families with both breast and ovarian cancer (42 ). The biological role of the BRCT domains is unknown, however, two missense mutations in the C-terminal 160 amino acids of BRCA1 (which include the BRCT domains) are associated with breast cancer (43 ,44 ). Additionally, the C-terminal 270 amino acids of 53BP1 are sufficient for binding p53 (45 ).
The S.cerevisiae RAD9 gene has been shown to be required for DNA damage-specific checkpoints in both G1 and G2 (1 ,7 ,12 ,30 ) and more recently roles for RAD9 in DNA damage processing (14 ) and damage-dependent transcription (46 ) have also been demonstrated. Data presented here indicate that rhp9 is also required for the damage checkpoints in both G1 and G2 (Figs 4 and 6 ). Epistasis analysis indicates that rhp9 functions in one or more processes requiring the rad3, rad9, rad17 and rad26 genes, with one of these processes likely to be the damage checkpoint.
The S.pombe checkpoint rad mutants rad1, rad3, rad9, rad7, rad26 and hus1 have all been shown to be defective in both the replication and DNA damage checkpoints (3 ,5 ,6 ). Double mutants of rhp9-d with rad3, rad9, rad17 and rad26 null alleles display no increase in radiation sensitivity over that of the most sensitive single mutant, indicating that the genes are required for one or more common processes (Fig. 7 ). The increased sensitivity of an rhp9 hus1 double mutant over that of hus1 suggests that hus1 is not required for all of the functions which are defective in other checkpoint rad mutants such as rad3, rad17 and rad26. This additional function may be reflected in the response of the rhp9 null allele to HU (Fig. 4 ). In response to HU, synchronous cultures of rhp9-d do not undergo septation for at least 5 h following exposure to HU, indicating an intact replication checkpoint. After 5 h, the expected length of replication delay induced by HU exposure (47 ), rhp9-d cells undergo septation, unlike chk1 cells, which do not. This may reflect an inability of rhp9-d cells to properly regulate the processing of DNA structures which occur as a result of the HU block. This potential defect may then lead to loss of the arrest signal after a particular time of delay. Such an rhp9-dependent DNA processing function may be analogous to that carried out by RAD9 in S.cerevisiae (14 ).
We thank Tony Carr for discussions and help with the damage checkpoint analysis. J.W. and N.W. acknowledge receipt of MRC studentships and J.W. thanks the Biochemical Society for a travel grant. S.W. acknowledges receipt of a BBSRC studentship. F.W. thanks the Royal Society and Wellcome Trust for travel grants and the CRC for project grant SP2212/0102.
*To whom correspondence should be addressed. Tel: +44 1273 678257; Fax: +44 1273 678433; Email f.z.watts@sussex.ac.uk
Strain
Genotype
sp.011
ade6-704, leu1-32, ura4-D18, h-
sp.012
ade6-704, leu1-32, ura4-D18, h+
sp.096
rad9::ura4, ade6-704, leu1-32, h-
sp.218
rad2::ura4, ade6-704, leu1-32, h-
sp.222
rad13::ura4, ade6-704, leu1-32, h-
sp.310
rad17::ura4, ade6-704, leu1-32, h-
sp.304
cdc10-V50, ura4-D18, h+
sp.374
chk1::ura4, h+
sp.378
rad1::ura4, leu1-32, ade6-704, h-
sp.379
rad3::ura4, leu1-32, ade6-704, h-
sp.380
hus1::leu1, h+
sp.381
rad26::ura4, ade6-704, leu1-32, h+
sp.382
cdc22.M45, ade6-704, ura4-D18, leu1-32, h+
sp.383
rad26-T12, h+
sp.417
hus1::leu1, h-
Strains created during this study
sp.346
mms1, ade6-704, leu1-32, ura4-D18, h+
sp.347
mms8, ade6-704, leu1-32, ura4-D18, h+
sp.348
mms35, ade6-704, leu1-32, ura4-D18, h-
sp.349
mms38, ade6-704, leu1-32, ura4-D18, h-
sp.350
mms52, ade6-704, leu1-32, ura4-D18, h+
sp.351
mms57, ade6-704, leu1-32, ura4-D18, h-
sp.352
mms60, ade6-704, leu1-32, ura4-D18, h-
sp.353
rhp9-64, ade6-704, leu1-32, ura4-D18, h-
sp.354
mms70, ade6-704, leu1-32, ura4-D18, h-
sp.391
rhp9::ura4, ade6-704, leu1-32, h+
sp.392
rhp9::ura4, rad13::ura4, ade6-704, leu1-32, h+
sp.394
rhp9::ura4, hus1::leu1, h-
sp.395
rhp9::ura4, chk1::ura4, h-
sp.397
rhp9::ura4, rad2::ura4, ade6-704, leu1-32, h+
sp.398
rhp9::ura4, rad26.T12, h+
sp.408
rhp9::ura4, rad26::ura4, ade6-704, leu1-32, h+
sp.409
rhp9::ura4, rad3::ura4, ade6-704, leu1-32, h+
sp.411
rhp9::ura4, cdc10-V50, h-
sp.412
rhp9::ura4, cdc17, h+
sp.413
rhp9::ura4, cdc22-M45, h-
sp.419
chk1::ura4, hus1::leu1, h+
REFERENCES