During early embryogenesis in Caenorhabditis elegans, the ATL-1–CHK-1 (ataxia telangiectasia mutated and Rad3 related–Chk1) checkpoint controls the timing of cell division in the future germ line, or P lineage, of the animal. Activation of the CHK-1 pathway by its canonical stimulus DNA damage is actively suppressed in early embryos so that P lineage cell divisions may occur on schedule. We recently found that the rad-2 mutation alleviates this checkpoint silent DNA damage response and, by doing so, causes damage-dependent delays in early embryonic cell cycle progression and subsequent lethality. In this study, we report that mutations in the smk-1 gene cause the rad-2 phenotype. SMK-1 is a regulatory subunit of the PPH-4.1 (protein phosphatase 4) protein phosphatase, and we show that SMK-1 recruits PPH-4.1 to replicating chromatin, where it silences the CHK-1 response to DNA damage. These results identify the SMK-1–PPH-4.1 complex as a critical regulator of the CHK-1 pathway in a developmentally relevant context.

Introduction

In somatic cells, DNA damage or stalled DNA replication can activate the S-phase checkpoint, resulting in delayed cell cycle progression to allow the damage to be repaired (for reviews see Bartek et al., 2004; Sancar et al., 2004). S-phase checkpoint signaling is mediated by ataxia telangiectasia mutated and Rad3 related (ATR) and Chk1 protein kinases. Replication forks that stall at sites of DNA damage activate ATR, which then phosphorylates and activates Chk1. Finally, cell cycle progression is delayed by activated Chk1 through the modulation of core cell cycle regulators, such as the Cdc25 protein phosphatase.

In contrast to somatic cells, early embryonic cell cycles typically lack a checkpoint response to DNA damage (for review see O'Farrell et al., 2004). In both Xenopus laevis and Drosophila melanogaster, this is because an insufficient number of nuclei are present in early embryos, and, thus, an insufficiently robust checkpoint signal is generated to thwart the mitosis-promoting activity of maternally supplied and abundant Cdk1–cyclin B complexes. In both flies and frogs, it is only later in embryogenesis that the checkpoint signal produced by replication stress is strong enough to neutralize Cdk1–cyclin B, and this is caused by the accumulation of nuclei (Dasso and Newport, 1990; Sibon et al., 1997, 1999; Su et al., 1999; Yu et al., 2000; Conn et al., 2004; Crest et al., 2007). In Caenorhabditis elegans, the situation is quite different. The ATR–Chk1 pathway is present and active from the first division onwards in worms, and it plays an important role in controlling the timing of cell division during the early cycles (Brauchle et al., 2003). Checkpoint function is restricted to the P lineage, or future germ line, in C. elegans embryos, and its activation by as of yet undetermined developmental cues results in the delayed division of P cells relative to their sisters. This asynchrony in cell division is critical for embryonic and germ line development, as reducing the delay through inactivation of the ATR–Chk1 pathway results in germ line developmental failure and sterility, whereas extending the delay through hyperactivation of the ATR–Chk1 pathway results in patterning defects and embryonic lethality (Encalada et al., 2000; Brauchle et al., 2003; Kalogeropoulos et al., 2004; Holway et al., 2006).

Although C. elegans differs from Xenopus and Drosophila in that the ATR–Chk1 pathway controls the pace of the early embryonic cycles, what is common between them is that like frog and fly embryos, the checkpoint is nonresponsive to DNA damage in early nematode embryos. This is not the result of insufficient signal strength but rather of the presence of an active silencing mechanism that suppresses the checkpoint response to DNA damage but allows the checkpoint to respond to developmental cues (Holway et al., 2006). This silencing mechanism has presumably evolved to prevent unscheduled checkpoint activation, which would cause extended delays in cell division and, ultimately, embryonic lethality. Our laboratory identified this checkpoint silencing mechanism, and, to date, we have isolated three genes that are required for silencing: the gei-17 SUMO E3 ligase, the polh-1 translesion synthesis DNA polymerase, and the mutationally defined but uncloned rad-2 gene (Holway et al., 2006). Previous work has shown that gei-17 and polh-1 silence the checkpoint through their ability to promote the rapid replication of damaged DNA (Holway et al., 2006), whereas the role of rad-2 in silencing was as of yet unknown.

The rad-2 mutation was isolated 25 yr ago in a screen for mutations causing embryonic sensitivity to DNA-damaging agents (Hartman and Herman, 1982). Follow-up phenotypic analysis of rad-2 showed that mutant animals were competent for excision repair and that the period of DNA damage sensitivity was restricted to early embryogenesis (Hartman, 1984; Hartman et al., 1989; Jones and Hartman, 1996). More recently, we have shown that rad-2 is a component of the silencing pathway that suppresses chk-1 activation by DNA damage in early embryos (Holway et al., 2006). This conclusion was based largely on effects of the rad-2 mutation on the timing of cell division in early embryos exposed to DNA-damaging agents. Wild-type embryos did not delay the cell cycle after exposure to either methyl methanesulphonate (MMS) or UV-C or UV light, whereas rad-2 mutant embryos showed a substantial delay. Importantly, the damage-induced delay in rad-2 embryos was reversed upon the RNAi-mediated depletion of chk-1. These genetic experiments indicated that rad-2 antagonizes the chk-1 pathway during the early embryonic DNA damage response and prompted us to further explore rad-2 function in checkpoint silencing.

In this study, we report the cloning of rad-2 and show that the rad-2 phenotype is caused by mutations in the smk-1 gene. smk-1 is an evolutionally conserved regulatory subunit of protein phosphatase 4 (PP4; or pph-4.1 in C. elegans) and has recently been shown to control lifespan in the worm (Wolff et al., 2006). We report that the roles of smk-1 in checkpoint silencing and longevity are distinct, and we show that the function of SMK-1 in silencing is to recruit PPH-4.1 to replicating chromatin so that it may antagonize checkpoint signaling during a DNA damage response. These results link PP4 to negative regulation of the ATR–Chk1 checkpoint, provide a targeting function for the SMK-1 regulatory subunit, and illustrate how during development primordial inputs into the ATR–Chk1 pathway such as DNA damage may be bypassed so that the checkpoint can respond exclusively to developmentally programmed inputs.

Results

The rad-2 mutation negatively affects CHK-1 activation during the DNA damage response

To gain cytological and biochemical evidence that rad-2 antagonizes chk-1 during a DNA damage response, we examined the phosphorylation status of CHK-1 in wild-type and rad-2 embryos exposed to MMS. To do this, we used an antibody that recognizes the Ser345-phosphorylated (CHK-1–S345-P) and activated form of the enzyme and examined early embryos by immunofluorescence microscopy (Fig. 1, A–L). Wild-type (N2) embryos displayed a punctate staining pattern with this antibody that was specific for the P lineage in both two-cell (Fig. 1, A–C) and four-cell (Fig. 2 F) embryos, and this signal was largely reduced in chk-1 RNAi embryos (Fig. S1). Exposure of N2 embryos to MMS did not substantially alter the CHK-1–S345-P signal intensity (Fig. 1, D–F), which is consistent with the checkpoint being silenced in wild-type embryos (Holway et al., 2006). In contrast to wild type, however, rad-2 embryos showed a noticeable increase in CHK-1–S345-P signal intensity after exposure to MMS (Fig. 1, G–L). To confirm these cytological observations biochemically, we prepared whole embryo extracts for the purpose of detecting activated CHK-1 by immunoblotting. As shown in Fig. 1 M, activated CHK-1 was not readily detected in control or MMS-exposed N2 embryos. In contrast, slightly more activated CHK-1 was observed in rad-2 embryos, and this was substantially increased upon MMS exposure. To ensure equal loading, we also probed the blots for total CHK-1 and PCN-1, the worm orthologue of proliferating cell nuclear antigen, and found that equivalent amounts of these factors were present in all extracts. Image densitometry of the blot in Fig. 1 M revealed that approximately threefold more activated CHK-1 was present in rad-2 embryos relative to wild type after exposure to MMS (Fig. 1 N). Based on the data in Fig. 1, we conclude that DNA damage activates CHK-1 to a greater extent in rad-2 embryos relative to wild type.

The data in Fig. 1 (A–L) show that activated CHK-1 localizes to punctate cytoplasmic structures in P cells that are reminiscent of P granules. To determine directly whether these structures are indeed P granules, we performed colabeling experiments using antibodies against activated CHK-1 and the P granule component PGL-1 (Kawasaki et al., 1998). As shown in Fig. 2 (A–D), the activated CHK-1 and PGL-1 signals overlapped, and, from this, we conclude that activated CHK-1 resides in P granules. To determine whether P granule residency was controlled by rad-2, we also stained early rad-2 embryos with these antibodies and found that activated CHK-1 still resides in P granules despite the loss of rad-2 function (Fig. 2, E–H). We conclude that activated CHK-1 localizes to cytoplasmic P granules in a rad-2–independent manner. The mechanism by which activated CHK-1 accumulates in P granules and the importance of this for CHK-1's ability to control the cell cycle is not yet known and is currently under investigation.

The rad-2 mutation primarily affects early embryos during the DNA damage response

Having found that rad-2 negatively regulates chk-1 during the DNA damage response in early embryos, we next asked whether rad-2 function was restricted to early embryogenesis or whether it was required throughout the embryonic period. Earlier studies had shown that plating rad-2 embryos on media containing MMS did not prevent hatching, whereas exposing adults to MMS prevented the hatching of their progeny (Hartman and Herman, 1982; Hartman, 1985). This suggested that very early embryogenesis represented the period of DNA damage sensitivity in rad-2 mutants; therefore, we sought a more direct test of this hypothesis. To do this, we collected early embryos from gravid adults by bleaching and plated these embryos. Next, we UV irradiated the embryos and determined survival as a function of both dose and time of administration of the UV light (Fig. 3 A). Early rad-2 embryos (i.e., those irradiated immediately after plating) were more sensitive to UV light than early wild-type embryos at all doses of UV that were tested. Interestingly, there was little difference in the UV light sensitivities of rad-2 relative to wild type if the UV light was administered ≥4 h after plating (Fig. 3 A). From this, we conclude that early but not late embryos require rad-2 to survive DNA damage.

In C. elegans, there are two sources of rapidly proliferating cells: the early embryo and the adult hermaphrodite gonad (for review see Lambie, 2002). We have previously shown that the chk-1 pathway responds to DNA damage in the gonad but is silenced in the early embryo (Holway et al., 2006). Therefore, it was of interest to determine whether rad-2 function was restricted to early embryos or whether it was also required in the germ line to survive DNA damage. To do this, we UV light irradiated hermaphrodites to damage the germ cells and mated them to untreated males harboring a GFP–ribonucleotide reductase (RNR) transgene (Zhong et al., 2003). We then asked whether viable cross progeny could be produced from the UV-irradiated germ cells. We performed the mating step because we required a source of undamaged sperm so that all effects on the survival of progeny would be through DNA damage inflicted specifically in the mitotic zone of the hermaphrodite gonad. As shown in Fig. 3 B, the cross progeny from this experiment were viable, but the self progeny were not. This result indicates that mitotically dividing germ cells in the hermaphrodite gonad do not require rad-2 function to survive DNA damage. The fact that the self progeny in this experiment were sensitive to DNA damage likely reflects the inability of early rad-2 embryos to survive the damaged DNA supplied by the UV-irradiated sperm.

To pursue these observations further, we next asked whether the rad-2 mutation hyperactivates the ATR–Chk1 pathway in the gonad, as it does in early embryos. Previous work has shown that mitotically dividing germ cells in the distal tip of the gonad arrest in an atl-1–dependent manner after exposure to UV light (Holway et al., 2006). This arrest is reflected by a reduction in the number of nuclei at the distal tip (or mitotic zone) and an increase in their size. Therefore, we compared cell cycle arrest in wild-type versus rad-2 gonads after exposure to UV light (Fig. 3 C). If the loss of rad-2 function hyperstimulates the ATR–Chk1 pathway in germ cells, we would expect a more pronounced reduction in the number of mitotic nuclei at the distal tip in rad-2 relative to wild-type gonads. We observed that UV light caused a reduction of 17.4 mitotic zone nuclei on average in wild-type animals and a reduction of 10.1 nuclei in rad-2 gonads. These data show that the loss of rad-2 function in distal tip germ cells does not reduce the number of UV light–exposed mitotic zone nuclei beyond what is observed in wild type and, in fact, that rad-2 gonads are modestly more refractory to atl-1–dependent cell cycle arrest than are wild-type gonads. We conclude that the stimulatory effect of the rad-2 mutation on the ATR–Chk1 pathway is specific for the early embryonic cell cycle.

rad-2 corresponds to mutations in the smk-1 gene

To pursue these observations further, we next sought to identify the gene encoding rad-2. Previous genetic analysis of rad-2 had mapped the position of the gene to 1.09 ± 0.46 cM on chromosome V (Hartman and Herman, 1982). Using a combination of bulk segregation analysis, three-factor crosses, and single nucleotide polymorphism (SNP) mapping, we were able to refine this position to the interval between 1.38 and 1.88 cM. To identify rad-2, we performed an RNAi screen across this interval using the soaking method. We initially searched for genes that would render embryos sensitive to UV light after depletion by RNAi. This resulted in the identification of smk-1 at position 1.49 cM as a candidate gene encoding rad-2. To pursue this further, we performed more detailed analysis of the smk-1 RNAi phenotype. Two different regions of the gene, the central region and the 3′ end, were targeted for RNAi knockdown. RNAi against the central region (RNAi#1) resulted in a low level of embryonic lethality, and this was greatly increased when embryos were exposed to MMS (Table I). Therefore, RNAi#1 phenocopies rad-2. RNAi against the 3′ end (RNAi#2) resulted in high embryonic lethality even in the absence of MMS. When either RNAi#1 or #2 were combined with the rad-2 mutation, embryonic lethality was higher than that observed in any single case alone (Table I). These results show that smk-1 is an essential gene and that RNAi#1 represents a hypomorphic condition. These results are also consistent with the idea that rad-2 represents a hypomorphic allele of the smk-1 gene.

A hallmark of the rad-2 phenotype is that these embryos show a checkpoint-dependent delay in cell cycle progression in response to DNA damage. This is in contrast to wild-type embryos, which silence their checkpoint responses during a DNA damage response. If rad-2 represents a hypomorphic allele of smk-1, smk-1 RNAi#1 should phenocopy rad-2 for checkpoint silencing. To address this, we timed cell cycle progression in early embryos as described previously (Holway et al., 2006). In both rad-2 and smk-1 RNA#1 embryos, the first cell cycle occurred normally in the absence of DNA damage but was substantially delayed after exposure to MMS (Fig. 4 A). Importantly, in both cases, this MMS-induced delay was reversed after chk-1 RNAi. These results show that smk-1 RNAi#1 phenocopies the checkpoint silencing defect of rad-2. To determine whether a wild-type copy of the smk-1 gene could rescue the rad-2 phenotype, we made an smk-1–GFP fusion transgene (Fig. 4 B) driven by the pie-1 promoter and introduced the gene into rad-2 animals by particle bombardment to produce the rad-2 (pie-1smk-1–GFP) strain. Transformants were selected by virtue of GFP signals and were tested for sensitivity to DNA-damaging agents. Introduction of wild-type smk-1 coding sequences into rad-2 animals increased resistance to both MMS and UV light (Fig. 4 C). Furthermore, when the timing of cell division was examined in early embryos, we observed that rad-2 (pie-1smk-1–GFP) embryos did not delay the cell cycle to the same extent as rad-2 mutants after exposure to MMS (Fig. 4 A). From this, we conclude that smk-1 RNAi#1 phenocopies the DNA damage response phenotypes of rad-2 and that introduction of an smk-1–GFP transgene into rad-2 mutants partially suppresses these phenotypes.

As further evidence that rad-2 represents an allele of smk-1, we sought to link rad-2 to a previously identified phenotype of smk-1, longevity. The smk-1 gene was first identified in C. elegans as a regulator of lifespan (Wolff et al., 2006). RNAi against smk-1 reduces both the lifespan of wild-type animals and the extended lifespan of daf-2 mutant animals. Therefore, we performed longevity assays on rad-2 animals and rad-2 animals exposed to daf-2 RNAi and compared these lifespans with N2 and N2 daf-2 RNAi animals. As shown in Table II, in both cases, the N2 animals lived longer than rad-2 animals. Thus, like smk-1 RNAi, the rad-2 allele reduces the lifespan of otherwise wild-type animals, and it reduces the extended lifespan that results from the depletion of daf-2.

To determine the molecular basis of the rad-2 mutation, we sequenced the smk-1 gene in the rad-2 strain. smk-1 encodes an evolutionally conserved regulatory subunit of PP4 (Wolff et al., 2006). Homologues of rad-2 include human PP4R3, yeast PSY2, and Drosophila falafel (Spradling et al., 1999; Wu et al., 2004; Gingras et al., 2005). Three differences were found in the smk-1 gene from rad-2 relative to wild-type strains (E497G, D580G, and D703G; Fig. 5 A). Of particular interest is the mutation occurring at position 703, as this aspartic acid residue is absolutely conserved from yeast to humans (Fig. 5 B) and is found within a highly conserved subdomain of the SMK-1 protein, conserved region 3 (Wolff et al., 2006). Collectively, our data show that smk-1 RNAi phenocopies the rad-2 allele for both DNA damage response and lifespan phenotypes, that a smk-1 transgene can partially suppress the rad-2 phenotype, and that the smk-1 gene from the rad-2 strain contains mutations, including an amino acid substitution at an evolutionally conserved position. We conclude that the rad-2 phenotype is caused by mutations in the smk-1 gene.

SMK-1 is a PPH-4.1–binding partner, and PPH-4.1 controls the early embryonic DNA damage response

Recent work has demonstrated that smk-1 functions in lifespan regulation by controlling transcriptional activity of the daf-16 forkhead box O (FOXO) transcription factor (Wolff et al., 2006). Thus, it was possible that the effects of smk-1 on checkpoint silencing were through the regulation of daf-16. If so, we would expect that daf-16 mutant embryos would be sensitive to DNA-damaging agents, but this was not the case (Fig. 4 C). These results show that although rad-2 is an allele of smk-1, the role of smk-1 in checkpoint silencing is distinct from its role in daf-16–mediated longevity.

In other organisms, smk-1 orthologues form complexes with PP4 (Gingras et al., 2005). To see whether SMK-1 did the same, we performed coimmunoprecipitation experiments using proteins expressed by in vitro transcription/translation in rabbit reticulocyte lysate. Lysates expressing PPH-4.1, the C. elegans homologue of PP4, were mixed with lysates expressing epitope-tagged SMK-1. The mixtures were then immunoprecipitated with an antibody that recognizes the tag on SMK-1, and, as shown in Fig. 5 C, PPH-4.1 was found in these immune complexes. PPH-4.1 was not found in the immune complexes when epitope-tagged SMK-1 was omitted from the reaction or when nonspecific antibody was used in the coimmunoprecipitation, demonstrating specificity. We conclude that SMK-1 interacts with PPH-4.1. We next asked whether the D703G mutation in the rad-2 allele of smk-1, which lies in conserved region 3 of the protein, influenced interaction between SMK-1 and PPH-4.1. As shown in Fig. 5 C, PPH-4.1 did not efficiently coimmunoprecipitate with a mutant form of SMK-1 containing the D703G substitution (SMK-1 D703G). These data suggest that at least in part, the rad-2 phenotype is caused by a compromised interaction between SMK-1 and PPH-4.1.

To pursue these observations further, we assessed DNA damage response phenotypes for embryos depleted of pph-4.1 by RNAi. Unlike daf-16 mutants, embryos depleted of pph-4.1 were very sensitive to both UV light and MMS (Fig. 4 C). Furthermore, pph-4.1–depleted embryos displayed a DNA damage–dependent delay in progression through the first cell cycle in a manner similar to rad-2 embryos (Fig. 4 A). Based on these data, we conclude that the rad-2 phenotype is caused by an inability of SMK-1 to control PPH-4.1 function during the DNA damage response.

SMK-1 is a chromosomal protein that recruits PPH-4.1 to replicating chromatin

To learn more about how smk-1 performs its checkpoint silencing function, we used the rad-2 (pie-1smk-1–GFP) strain to localize SMK-1–GFP in early embryos (Fig. 6, A–L). The fusion protein was nuclear throughout all stages of the cell cycle. At prophase, SMK-1–GFP colocalized with condensed chromosomes, indicating that SMK-1 is a chromosomal protein (Fig. 6, D–I). To make certain that these localization patterns were not an artifact of the exogenous pie-1 promoter used in our construct, we repeated this analysis with a strain driving SMK-1–GFP off the endogenous smk-1 promoter and obtained identical results (unpublished data). To see whether chromosomal occupancy of SMK-1 was dependent on DNA replication, we treated rad-2 (pie-1smk-1–GFP) animals with RNAi against the replication initiation factor cdt-1. As shown in Fig. 6 (M–O), the chromosomal localization of SMK-1–GFP was abolished in cdt-1 RNAi embryos. This was not the case for embryos expressing a histone H2B-GFP fusion protein, which localized to condensed chromatin regardless of the depletion of cdt-1 (Fig. 6, S–X). We also asked whether abrogation of the ATR pathway influenced the chromosomal localization of SMK-1 and found that SMK-1 localization was not perturbed by atl-1 RNAi (Fig. 6, P–R). The effectiveness of the atl-1 RNAi in this experiment was ascertained by the high level of embryonic lethality that resulted, which is a known consequence of atl-1 RNAi (Aoki et al., 2000). From this experiment, we conclude that SMK-1 is recruited to chromatin in a replication-dependent and checkpoint-independent manner.

The results obtained thus far indicate that SMK-1 and PPH-4.1 form a complex, that both proteins confer DNA damage resistance to early embryos, and that SMK-1 is recruited to chromatin in early embryos in a manner dependent on DNA replication. To pursue the chromatin-binding properties of SMK-1 further, we developed a chromatin-binding assay for early embryos (Fig. 7 A) based on previously published procedures (Polanowska et al., 2004). Large quantities of early embryos were isolated from adults and sonicated to produce an embryo extract. The extract was centrifuged to produce two fractions: a supernatant (A) and the chromatin-containing pellet (B). The pellet fraction was then treated with micrococcal nuclease to degrade the DNA and to release the DNA-bound chromatin proteins. This reaction was then centrifuged again to produce a supernatant (C) and pellet (D) fractions. Proteins that were originally in the first pellet fraction (B) but were found in the second supernatant fraction (C) after micrococcal nuclease treatment were defined as chromatin proteins and identified by immunoblotting. As shown in Fig. 7 A, the known chromatin protein PCN-1 was found in fractions B and C but not in fraction D as expected. In contrast, the nonchromatin protein tubulin was found exclusively in fraction A, verifying that this procedure can identify chromatin proteins. We also examined the behavior of SMK-1–GFP and PPH-4.1 under these fractionation conditions. As expected, based on the localization data in Fig. 6, SMK-1–GFP was found in the chromatin protein–containing fraction C. PPH-4.1 was also found in fraction C, and some was observed in fraction D. It may be that a subset of PPH-4.1 associates with a nonchromosomal, easily sedimenting structure such as the centrosome (Sumiyoshi et al., 2002).

We next used this assay to monitor the chromatin association of SMK-1–GFP and PPH-4.1 under different conditions. As shown in Fig. 7 B, SMK-1–GFP was found in the chromatin protein–containing C fraction in both control and MMS- exposed embryos (lanes 2 and 3). RNAi-mediated depletion of gei-17, another checkpoint silencing gene, had no effect on the chromatin binding of SMK-1–GFP (Fig. 7 B, lanes 4 and 5), whereas RNAi against smk-1 itself did prevent the recovery of SMK-1–GFP in the chromatin fraction (Fig. 7 B, lane 6) as expected. PCN-1 was used as a control for these experiments and was found in the chromatin fraction under all conditions. To pursue these observations further, we extended this analysis to PPH-4.1. The PPH-4.1 protein was found in the chromatin fraction of both control and MMS-exposed wild-type embryos (Fig. 7 C, lanes 1 and 2). Importantly, the amount of PPH-4.1 that associated with chromatin in rad-2 embryos was noticeably reduced relative to wild-type embryos (Fig. 7 C, lanes 3 and 4). The overall level of PPH-4.1 in rad-2 versus wild-type extracts was only modestly reduced. To confirm these data using an alternative method, we immunostained MMS-exposed wild-type and rad-2 embryos with antiserum directed against PPH-4.1. As shown in Fig. S2 , the PPH-4.1 signal was nuclear in wild-type embryos but not in rad-2 embryos. Based on these results, we conclude that SMK-1 functions to recruit the PPH-4.1 phosphatase to chromatin and that a failure to do so, such as in rad-2 embryos, leads to hyperactivation of the chk-1 response to DNA damage and subsequent embryonic lethality.

Discussion

In this study, we have shown that mutations in the smk-1 gene cause the rad-2 phenotype. We have also shown that although the rad-2 mutation has a strong effect on early embryonic DNA damage resistance, it does not affect damage resistance in proliferating cells of the germ line. Consistent with a role for smk-1 in early embryos but not the germ line is published data showing that an SMK-1–GFP fusion protein expressed off the endogenous smk-1 promoter is abundant in early embryos as well as other tissues of the worm but is not readily observed in the germ line (Wolff et al., 2006). Therefore, it may be that the embryonic specificity of the checkpoint silencing pathway is achieved through preferential expression of the SMK-1–PPH-4.1 complex in embryos relative to germ cells. The lack of a rad-2 phenotype in germ cells must be interpreted with caution, however, given the hypomorphic nature of the rad-2 allele.

The work presented here has uncovered a role for SMK-1 in silencing DNA damage–based CHK-1 activation in early embryos. In C. elegans, SMK-1 also functions in the insulin-mediated control of longevity (Wolff et al., 2006). In longevity, SMK-1 modulates the activity of the DAF-16 transcription factor through an unknown mechanism to regulate the expression of DAF-16 target genes. We have shown here that DAF-16 is not required for checkpoint silencing, and, thus, it appears that the roles for SMK-1 in aging and checkpoint silencing are distinct. DAF-16 is a member of the FOXO superfamily of transcriptional regulators, and, therefore, it is possible that SMK-1 functions with a FOXO transcription factor that is distinct from DAF-16 in the checkpoint silencing pathway. We do not favor this hypothesis, however, as it is generally true that early embryonic cell cycle control is driven by maternally supplied regulators and not via zygotic transcription. Although the roles of smk-1 in longevity and checkpoint silencing can be unlinked in the embryo, we note that chk-1, the smk-1 target for silencing, has been shown to reduce lifespan in the worm by acting in postmitotic cells (Olsen et al., 2006). Therefore, it may be that smk-1 antagonizes the chk-1 effect on lifespan, and experiments are in progress to test this hypothesis.

SMK-1 is an evolutionally conserved regulatory subunit of the PP4 phosphatase. Links between the PP4 complex and DNA damage response have been uncovered before, although not in the context of regulation of the ATR–Chk1 pathway as has been reported here. In Drosophila, loss of the SMK-1 orthologue falafel causes sensitivity to the DNA-damaging agent cisplatin (Gingras et al., 2005). In yeast, the SMK-1 orthologue Psy2 and the PP4 orthologue Pph3 have been shown to control the phosphorylation status of the histone variant H2AX after DNA damage (Keogh et al., 2006). In this case, dephosphorylation of H2AX by Pph3 is required for attenuation of the checkpoint response to double-strand breaks. This is somewhat similar to the results reported here, in which SMK-1 and PPH-4.1 negatively regulate the ATR–Chk1 pathway after DNA damage; however, the mechanism in C. elegans is clearly different, as worms do not have H2AX. Although these previous reports clearly linked SMK-1 orthologues to the DNA damage response (Gingras et al., 2005; Keogh et al., 2006), they did not explain the role of SMK-1 in this process. We report here that SMK-1 is a chromosomal protein and that its recruitment to chromatin is dependent on ongoing DNA replication. Furthermore, we show that SMK-1 is required to recruit PPH-4.1 to chromatin, the site of CHK-1 activation during a DNA damage response. Collectively, these data supply a function for SMK-1 during the DNA damage response (the targeting of PPH-4.1 to chromatin) and suggest that the SMK-1–PPH-4.1 complex may be a general regulator of the ATR–Chk1 pathway in metazoan cells.

Although our data clearly identify SMK-1–PPH-4.1 as an important negative regulator of the checkpoint response to DNA damage in early nematode embryos, we do not at present know the critical target for this phosphatase complex in attenuating the checkpoint response. Chk1 is known to be regulated directly by protein phosphatase 2A (Leung-Pineda et al., 2006), by PPM1D, a type 2C phosphatase (Lu et al., 2005), and by Dis2, a type I phosphatase in fission yeast (den Elzen and O'Connell, 2004). Preliminary results from our laboratory have shown that PPH-4.1 and CHK-1 form a complex (unpublished data), and, thus, it may be that PP4-type phosphatases are also capable of the direct regulation of Chk1. Regulation of Chk1 is likely to be complex in any given cell type, and is likely to involve multiple phosphatases controlling Chk1 under different circumstances and in different subcellular locations. Our data show that a site for regulation of the ATR–Chk1 pathway by PP4 is chromatin, and this is consistent with the embryo's requirement that the Chk1 pathway be rapidly inactivated so as to prevent potentially lethal delays in cell cycle progression. To completely understand how the checkpoint is silenced in early embryos, it will be necessary to identify the SMK-1–PPH-4.1 target and to determine how this target is accessed by SMK-1–PPH-4.1 on replicating DNA.

Materials And Methods

C. elegans strains

The wild-type N2 Bristol strain was used in all control experiments (Brenner, 1974). SP488 (rad-2[mn156]), CF1038 (daf-16[mu86]), and AZ212 (unc-119(ed3) ruIs32[unc-119(+) pie-1–GFP–H2B]) strains were provided by T. Stiernagle (Caenorhabditis Genetics Center, University of Minnesota, Minneapolis, MN). An RNR-GFP strain (maIs103[unc-36(+), rnr–GFP]) was provided by E. Kipreos (University of Georgia, Athens, GA; Zhong et al., 2003).

Genetic mapping of the rad-2 locus

Previous genetic mapping had determined the chromosome location of rad-2 to be V:1.09 ± 0.461 cM (Hartman and Herman, 1982). To confirm and extend these initial mapping data, the TH37 strain containing dpy-11 and unc-23 markers located at V:0.00 cM and V:1.88 cM, respectively, was used in three-factor cross mapping of rad-2. TH37 hermaphrodites (dpy-11/rad-2+/unc-23−) were mated with rad-2 males (dpy-11+/rad-2/unc-23), and cross progeny were isolated. These worms were allowed to self-fertilize, and recombinants representing the dpy-11/unc-23+ and dpy-11+/unc-23− genotype were identified. Once it has been determined that recombinant worms were homozygous for each marker, the status of the rad-2 gene was determined for each recombinant. 27 dpy-11/unc-23+ and 21 dpy-11+/unc-23− homozygous recombinants were screened for MMS sensitivity. The fact that dpy-11/rad-2− and rad-2/unc-23− recombinants were isolated confirms that rad-2 is to the right of dpy-11 and to the left of unc-23, or between 0.00 and 1.88 cM. In a total of 48 recombination events, 30 events occurred between dpy-11 and rad-2, and 18 events occurred between rad-2 and unc-23. These numbers translated to map ratios of 0.625 and 0.375 for dpy-11 and unc-23, respectively. Therefore, the three-factor cross indicated that rad-2 lies closer to unc-23 at V:1.175 with a 95% confidence interval of ±0.273 cM.

To further narrow the region of the rad-2 gene, we performed SNP mapping. dpy-11/rad-2 worms were mated to the Hawaiian CB4856 strain, and five recombinants that were dpy-11 and wild type for rad-2 (based on MMS sensitivity) were isolated. SNPs located between 1.14 and 1.46 cM were PCR amplified from recombinant worm lysates, and the origin of DNA at each locus was determined by either snip-SNP analysis or sequencing of the SNP. This analysis revealed that among the recombinants, rad-2 DNA could be found at positions 1.14, 1.27, and 1.38 cM but not at 1.46 cM. This analysis positioned the rad-2 locus to the right of 1.38 cM and, in combination with the three-factor crosses, defined the interval between 1.38 and 1.88 cM as the location of the rad-2 gene.

RNAi

smk-1 and pph-4.1 RNAi by soaking method was performed as described previously (Maeda et al., 2001). smk-1, gei-17, daf-2, cdt-1, and atl-1 RNAi by feeding method was performed as described previously (Timmons and Fire, 1998), and chk-1 RNAi was performed as described previously (Holway et al., 2006).

UV sensitivity assays

To examine whether rad-2 function was restricted to early embryogenesis or whether it was required throughout the embryonic period, UV sensitivity assay was performed using embryos prepared by bleaching N2 and rad-2 gravid hermaphrodites on the basis of published protocols (Edgar and McGhee, 1988). About 50 early embryos were plated on fresh plates and exposed to UV light at 0, 10, and 25 J/m2 at the indicated times in Fig. 3 A. 24 h after UV irradiation, the unhatched eggs were counted. Embryonic lethality was determined by dividing the number of eggs remaining after 24 h by the total number plated.

To examine whether rad-2 function was restricted to early embryos or whether it was also required in the germ line to survive DNA damage, UV light–irradiated hermaphrodites were crossed with untreated male worms harboring an RNR-GFP transgene, and the UV light sensitivity of progeny was examined. To do this, 10 L4-stage N2 and rad-2 hermaphrodites were exposed to 100 J/m2 UV light followed by plating eight males harboring an RNR-GFP transgene. 48 h after transferring males, all worms were removed from the plate, and GFP and non-GFP embryos were counted. After 24 h, the embryos were scored for survival to determine embryonic lethality.

Cloning of the rad-2 gene and lifespan analysis

Using a combination of bulk segregation analysis, three-factor crosses, and SNP mapping, the position of the rad-2 gene was refined to the genetic interval between 1.38 and 1.88 cM. To clone the rad-2 gene, UV-sensitive genes across this genetic interval were initially identified by UV sensitivity assay after depletion by soaking RNAi and were analyzed further by MMS sensitivity assay and timing of cell division in living embryos, which were performed as described previously (Holway et al., 2006). For longevity assay of rad-2, lifespan and statistical analyses were performed as described previously (Wolff et al., 2006).

Rescue of rad-2 mutant and genomic DNA sequencing

To construct an smk-1–GFP fusion transgene, a full-length cDNA of the smk-1 gene was cloned into a pie-1–GFP germline expression vector (Reese et al., 2000). The smk-1–GFP fusion transgene was introduced into the rad-2 mutant by microparticle bombardment to generate the rad-2 (pie-1–smk-1–GFP) strain. Using this transgenic strain, rescue of the rad-2 mutant was assessed by restoring normal embryonic viability and timing of cell division in living embryos in response to UV light and MMS exposures. Additionally, we also monitored the behavior of SMK-1–GFP expressed under the control of an endogenous smk-1 promoter in early embryos of a transgenic strain, smk-1p–smk-1–GFP, which were generated previously (Wolff et al., 2006). For genomic DNA sequencing, genomic DNA corresponding to the smk-1 gene was isolated from the rad-2 mutant and cloned into pCR2.1-TOPO vector (Invitrogen). Mutations in the genomic DNA were identified by DNA sequencing performed by Agencourt Bioscience Corp.

Antibodies, whole embryo extracts, and immunoblotting

C. elegans proliferating cell nuclear antigen orthologue PCN-1 antibody was generated by immunizing rabbits with the peptide DIDSEHLGIPDQDYAVVCE (Bethyl Laboratories). C. elegans PP4 orthologue PPH-4.1 antibody was a gift from M. Yamamoto (University of Tokyo, Tokyo, Japan; Sumiyoshi et al., 2002). Antibodies against phospho-Chk1 Ser345 (Cell Signaling Technology), Chk1 (Santa Cruz Biotechnology, Inc.), GFP (Abcam), OIC1D4 (Developmental Studies Hybridoma Bank), α-tubulin (Sigma-Aldrich), and c-myc (Santa Cruz Biotechnology, Inc.) were purchased. To prepare whole embryo extracts, embryos were obtained by bleaching gravid hermaphrodites and were suspended in twice the pellet volume of homogenization buffer (50 mM Hepes, pH 7.6, 200 mM KCl, 1 mM EDTA, 1 mM EGTA, 0.2% Triton X-100, 5% glycerol, and protease inhibitors). The embryo suspension was sonicated and clarified by centrifugation at 12,000 g at 4°C for 30 min. The pellet was resuspended in half the pellet volume of homogenization buffer. Immunoblotting was performed by the standard procedures with HRP-conjugated mouse or rabbit secondary antibodies (GE Healthcare), and protein bands were detected by enhanced chemiluminescence (Pierce Chemical Co.).

Chromatin protein fractions and chromatin-binding assay

To prepare embryo extract fractions containing chromatin proteins, large quantities of embryos were obtained by bleaching gravid hermaphrodites and were suspended in twice the pellet volume of homogenization buffer. The embryo suspension was sonicated briefly on ice until the mixture had lost its viscosity. The sonicated embryo mixture was clarified by centrifugation at 8,000 g at 4°C for 5 min, and the pellet was resuspended in half the pellet volume of homogenization buffer. The chromatin proteins were extracted from the pellet by adding micrococcal nuclease (Roche Applied Science) followed by centrifugation at 15,000 g for 20 min at 4°C. The supernatant fraction containing chromatin proteins was identified by immunoblotting with antibodies against α-tubulin and PCN-1, which are nonchromatin and chromatin proteins, respectively.

Site-directed mutagenesis, in vitro transcription and translation, and coimmunoprecipitation

The full-length cDNAs of smk-1 and pph-4.1 genes were cloned into pCS2+MT vector containing myc epitope tags and pSP72 vector, respectively. The myc-tagged SMK-1 mutant displacing an aspartic acid residue at position 703 to a glycine (myc-SMK-1 [D703G]) was generated by the Quik-Change II Site-Directed Mutagenesis kit (Stratagene). The myc–SMK-1, myc–SMK-1 (D703G), and untagged PPH-4.1 were transcribed and translated (TNT reaction) in the presence of [35S]methionine according to the manufacturer's instructions (Promega). For coimmunoprecipitation, 10 μl TNT reactions were mixed in 400 μl of binding buffer (20 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5 mM MgCl2, 10% glycerol, and 0.1% NP-40) and incubated with 0.5 μg of anti–mouse myc antibody at 4°C for 4 h. A mouse IgG (Santa Cruz Biotechnology, Inc.) was used as a nonspecific antibody for demonstrating specificity. After an overnight incubation with 20 μl of protein A/G PLUS-Agarose (Santa Cruz Biotechnology, Inc.) at 4°C, immunoprecipitated beads were washed three times with binding buffer. The protein bound to the beads was eluted by boiling in 30 μl of 2× Laemmli sample buffer. The samples were run on SDS-polyacrylamide gels and detected by autoradiography.

DNA staining, immunostaining, and fluorescence microscopy

Embryos and gonads were dissected form adult hermaphrodites and were fixed and stained by Hoechst 33258 as described previously (Holway et al., 2005, 2006). The images of nuclei in the gonad (Fig. 3 C) were captured on camera (2.1.1; Diagnostic Instruments) and processed using SPOT Advanced version 3.2.4 software (Diagnostic Instruments). UPlanAPO 40× NA 1.40 oil objective lenses were used. The nuclei in the mitotic zone of the gonad were then counted as described previously (Holway et al., 2006). For immunostaining, the fixed embryos were incubated with antibodies against phospho-Chk1 (Ser345), OIC1D4, and PPH-4.1 overnight at 4°C followed by a 2-h incubation with FITC- or rhodamine-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories). All confocal images (Figs. 1, A–J; 2, A–H; 6, A–X; S1, A–F; and S2, A–I) were obtained by a confocal system (LSM510 META; Carl Zeiss MicroImaging, Inc.) attached to a laser-scanning microscope (Axiovert 100M; Carl Zeiss MicroImaging, Inc.). Plan-Neofluar 40× NA 1.30 oil objective lenses (Carl Zeiss MicroImaging, Inc.) were used. All microscopic experiments were performed at room temperature.

Online supplemental material

Fig. S1 shows that the activated CHK-1 signal is largely reduced in chk-1 RNAi embryos. Fig. S2 shows that the nuclear localization of PPH-4.1 is abolished in MMS-exposed rad-2 and pph-4.1 RNAi embryos. Table S1 shows the embryonic lethality in Fig. 3 A.

Acknowledgments

We thank Masayuki Yamamoto for providing a PPH-4.1 antibody, Alain Viel for help with a bombardment transformation, the Caenorhabditis Genetics Center and E. Kipreos for providing worm strains, and Craig Hunter and his laboratory members for advice.

Support for this work was provided by the National Institute of General Medical Sciences (grant R01GM67735 to W.M. Michael).

References

References
Aoki, H., S. Sato, T. Takanami, T. Ishihara, I. Katsura, H. Takahashi, and A. Higashitani.
2000
. Characterization of Ce-atl-1, an ATM-like gene from Caenorhabditis elegans.
Mol. Gen. Genet.
264
:
119
–126.
Bartek, J., C. Lukas, and J. Lukas.
2004
. Checking on DNA damage in S phase.
Nat. Rev. Mol. Cell Biol.
5
:
792
–804.
Brauchle, M., K. Baumer, and P. Gonczy.
2003
. Differential activation of the DNA replication checkpoint contributes to asynchrony of cell division in C. elegans embryos.
Curr. Biol.
13
:
819
–827.
Brenner, S.
1974
. The genetics of Caenorhabditis elegans.
Genetics.
77
:
71
–94.
Conn, C.W., A.L. Lewellyn, and J.L. Maller.
2004
. The DNA damage checkpoint in embryonic cell cycles is dependent on the DNA-to-cytoplasmic ratio.
Dev. Cell.
7
:
275
–281.
Crest, J., N. Oxnard, J.Y. Ji, and G. Schubiger.
2007
. Onset of the DNA replication checkpoint in the early Drosophila embryo.
Genetics.
175
:
567
–584.
Dasso, M., and J.W. Newport.
1990
. Completion of DNA replication is monitored by a feedback system that controls the initiation of mitosis in vitro: studies in Xenopus.
Cell.
61
:
811
–823.
den Elzen, N.R., and M.J. O'Connell.
2004
. Recovery from DNA damage checkpoint arrest by PP1-mediated inhibition of Chk1.
EMBO J.
23
:
908
–918.
Edgar, L.G., and J.D. McGhee.
1988
. DNA synthesis and the control of embryonic gene expression in C.
elegans. Cell.
53
:
589
–599.
Encalada, S.E., P.R. Martin, J.B. Phillips, R. Lyczak, D.R. Hamill, K.A. Swan, and B. Bowerman.
2000
. DNA replication defects delay cell division and disrupt cell polarity in early Caenorhabditis elegans embryos.
Dev. Biol.
228
:
225
–238.
Gingras, A.C., M. Caballero, M. Zarske, A. Sanchez, T.R. Hazbun, S. Fields, N. Sonenberg, E. Hafen, B. Raught, and R. Aebersold.
2005
. A novel, evolutionarily conserved protein phosphatase complex involved in cisplatin sensitivity.
Mol. Cell. Proteomics.
4
:
1725
–1740.
Hartman, P.S.
1984
. UV irradiation of wild type and radiation-sensitive mutants of the nematode Caenorhabditis elegans: fertilities, survival, and parental effects.
Photochem. Photobiol.
39
:
169
–175.
Hartman, P.S.
1985
. Epistatic interactions of radiation-sensitive (rad) mutants of Caenorhabditis elegans.
Genetics.
109
:
81
–93.
Hartman, P.S., and R.K. Herman.
1982
. Radiation-sensitive mutants of Caenorhabditis elegans.
Genetics.
102
:
159
–178.
Hartman, P.S., J. Hevelone, V. Dwarakanath, and D.L. Mitchell.
1989
. Excision repair of UV radiation-induced DNA damage in Caenorhabditis elegans.
Genetics.
122
:
379
–385.
Holway, A.H., C. Hung, and W.M. Michael.
2005
. Systematic, RNA-interference-mediated identification of mus-101 modifier genes in Caenorhabditis elegans.
Genetics.
169
:
1451
–1460.
Holway, A.H., S.H. Kim, A. La Volpe, and W.M. Michael.
2006
. Checkpoint silencing during the DNA damage response in Caenorhabditis elegans embryos.
J. Cell Biol.
172
:
999
–1008.
Jones, C.A., and P.S. Hartman.
1996
. Replication in UV-irradiated Caenorhabditis elegans embryos.
Photochem. Photobiol.
63
:
187
–192.
Kalogeropoulos, N., C. Christoforou, A.J. Green, S. Gill, and N.R. Ashcroft.
2004
. chk-1 is an essential gene and is required for an S-M checkpoint during early embryogenesis.
Cell Cycle.
3
:
1196
–1200.
Kawasaki, I., Y.H. Shim, J. Kirchner, J. Kaminker, W.B. Wood, and S. Strome.
1998
. PGL-1, a predicted RNA-binding component of germ granules, is essential for fertility in C. elegans.
Cell.
94
:
635
–645.
Keogh, M.C., J.A. Kim, M. Downey, J. Fillingham, D. Chowdhury, J.C. Harrison, M. Onishi, N. Datta, S. Galicia, A. Emili, et al.
2006
. A phosphatase complex that dephosphorylates γH2AX regulates DNA damage checkpoint recovery.
Nature.
439
:
497
–501.
Lambie, E.J.
2002
. Cell proliferation and growth in C. elegans.
Bioessays.
24
:
38
–53.
Leung-Pineda, V., C.E. Rvan, and H. Piwnica-Worms.
2006
. Phosphorylation of Chk1 is antagonized by a Chk1-regulated protein phosphatase 2A circuit.
Mol. Cell. Biol.
26
:
7529
–7538.
Lu, X., B. Nannenga, and L.A. Donehower.
2005
. PPM1D dephosphorylates Chk1 and p53 and abrogates cell cycle checkpoints.
Genes Dev.
19
:
1162
–1174.
Maeda, I., Y. Kohara, M. Yamamoto, and A. Sugimoto.
2001
. Large-scale analysis of gene function in Caenorhabditis elegans by high-throughput RNAi.
Curr. Biol.
11
:
171
–176.
O'Farrell, P.H., J. Stumpff, and T.T. Su.
2004
. Embryonic cleavage cycles: how is a mouse like a fly?
Curr. Biol.
14
:
R35
–R45.
Olsen, A., M.C. Vantipalli, and G.J. Lithgow.
2006
. Checkpoint proteins control survival of the postmitotic cells in Caenorhabditis elegans.
Science.
312
:
1381
–1385.
Polanowska, J., J.S. Martin, R. Fisher, T. Scopa, I. Rae, and S.J. Boulton.
2004
. Tandem immunoaffinity purification of protein complexes from Caenorhabditis elegans.
Biotechniques.
36
:
778
–780, 782.
Reese, K.J., M.A. Dunn, J.A. Waddle, and G. Seydoux.
2000
. Asymmetric segregation of PIE-1 in C. elegans is mediated by two complementary mechanisms that act through separate PIE-1 protein domains.
Mol. Cell.
6
:
445
–455.
Sancar, A., L.A. Lindsey-Boltz, K. Unsal-Kacmaz, and S. Linn.
2004
. Molecular mechanisms of mammalian DNA repair and the DNA damage checkpoints.
Annu. Rev. Biochem.
73
:
39
–85.
Sibon, O.C., V.A. Stevenson, and W.E. Theurkauf.
1997
. DNA-replication checkpoint control at the Drosophila midblastula transition.
Nature.
388
:
93
–97.
Sibon, O.C., A. Laurencon, R. Hawley, and W.E. Theurkauf.
1999
. The Drosophila ATM homologue Mei-41 has an essential checkpoint function at the midblastula transition.
Curr. Biol.
9
:
302
–312.
Spradling, A.C., D. Stern, A. Beaton, E.J. Rhem, T. Laverty, N. Mozden, S. Misra, and G.M. Rubin.
1999
. The Berkeley Drosophila Genome Project gene disruption project: single P-element insertions mutating 25% of vital Drosophila genes.
Genetics.
153
:
135
–177.
Su, T.T., S.D. Campbell, and P.H. O'Farrell.
1999
. Drosophila grapes/CHK1 mutants are defective in cyclin proteolysis and coordination of mitotic events.
Curr. Biol.
9
:
919
–922.
Sumiyoshi, E., A. Sugimoto, and M. Yamamoto.
2002
. Protein phosphatase 4 is required for centrosome maturation in mitosis and sperm meiosis in C. elegans.
J. Cell Sci.
115
:
1403
–1410.
Timmons, L., and A. Fire.
1998
. Specific interference by ingested dsRNA.
Nature.
395
:
854
.
Wolff, S., H. Ma, D. Burch, G.A. Maciel, T. Hunter, and A. Dillin.
2006
. SMK-1, an essential regulator of DAF-16-mediated longevity.
Cell.
124
:
1039
–1053.
Wu, H.I., J.A. Brown, M.J. Dorie, L. Lazzeroni, and J.M. Brown.
2004
. Genome-wide identification of genes conferring resistance to the anticancer agents cisplatin, oxaliplatin, and mitomycin C.
Cancer Res.
64
:
3940
–3948.
Yu, K.R., R.B. Saint, and W. Sullivan.
2000
. The Grapes checkpoint coordinates nuclear envelope breakdown and chromosome condensation.
Nat. Cell Biol.
2
:
609
–615.
Zhong, W., H. Feng, F.E. Santiago, and E.T. Kipreos.
2003
. CUL-4 ubiquitin ligase maintains genome stability by restraining DNA-replication licensing.
Nature.
423
:
885
–889.

Abbreviations used in this paper: ATR, ataxia telangiectasia mutated and Rad3 related; FOXO, forkhead box O; MMS, methyl methanesulphonate; PP4, protein phosphatase 4; RNR, ribonucleotide reductase; SNP, single nucleotide polymorphism.