Characterization of Functional Domains in Human Claspin

Özdemirhan Serçin and Michael G. Kemp*

Department of Biochemistry and Biophysics; University of North Carolina School of Medicine; Chapel Hill, NC USA

Key words: Claspin, DNA replication, checkpoint, DNA damage, Cdc45, DNA polymerase, Rad17

Abbreviations: RFC, replication factor C; Pol ε, DNA polymerase epsilon; ssDNA, single-stranded DNA; RPA, replication protein A; ATRIP, ATR-interacting protein; ATR, ataxia telangiectasia-mutated and Rad3-related; 9-1-1, Rad9-Hus1-Rad1; Chk1, checkpoint kinase 1; Mrc1, mediator of replication checkpoint 1; RFID, replication fork-interacting domain; BPI and BPII, basic patch I and II; CBD, Chk1-binding domain; GST, glutathione S-transferase; aa, amino acids; MCM, minichromosome maintenance

Claspin is a mediator of the ATR-dependent DNA replication checkpoint in human cells and also promotes DNA replication fork progression and stability. Though Claspin has been shown to bind DNA and co-immunoprecipitate with other replication fork-associated proteins, the specific protein-protein and protein-DNA interactions that are important for Claspin function are not known. We therefore purified several domains of human Claspin and then tested for direct interactions of these fragments with several replication fork-associated proteins and with DNA. Our data show that the N terminus of Claspin binds to the replicative helicase co-factor Cdc45, the Timeless protein and a branched, replication fork-like DNA structure. In contrast, the C terminus of Claspin associates with DNA polymerase epsilon and Rad17- Replication Factor C (RFC). We conclude that multiple protein-DNA and protein-protein interactions may be important for Claspin function during DNA replication and DNA replication checkpoint signaling.

Introduction homolog of Xenopus Claspin functions similarly in DNA damage and replication checkpoint signaling.10,11 Unlike many other DNA damage and replication checkpoint signaling pathways proteins involved in DNA damage signaling responses, Claspin respond to DNA damage and replication stress during S phase appears to be specifically required for the phosphorylation of by delaying cell cycle progression, stabilizing stalled replication Chk1 but not other DNA damage response substrates in response forks and promoting DNA repair.1,2 This response is conserved to genotoxic stress.12,13 Consistent with these data using human throughout eukaryotes and is initiated by specific proteins that cells and Xenopus egg extracts, in vitro reconstitution assays with sense the presence of single-stranded DNA (ssDNA) and primer- purified proteins recently showed that Claspin directly mediates template junctions at stalled replication forks. Through an inter- Chk1 phosphorylation by ATR.14 action of the ssDNA-binding protein Replication Protein A The human Claspin protein shows a limited degree of sequence (RPA) with the ATR-interacting protein ATRIP, the DNA dam- similarity with the Mrc1 (Mediator of replication checkpoint) age response kinase ATR is recruited to the ssDNA at these sites.3 proteins in budding and fission yeast. Yeast strains lacking Mrc1 Similarly, the Rad17-Replication Factor C (RFC) complex recog- fail to show normal DNA replication checkpoint responses nizes primer-template junctions and loads the 9-1-1 (Rad9-Hus1- to DNA damage and replication stress during S phase.15,16 Rad1) complex.4 An interaction of the C-terminal tail of Rad9 Independent of its replication checkpoint activity, however, sev- with the ATR-activating protein TopBP15,6 then allows ATR to eral studies have shown that Mrc1 also co-localizes with the rep- become fully activated to phosphorylate and activate a number of lication fork apparatus, where it appears to maintain replication proteins involved in DNA damage checkpoints and DNA repair,7 fork stability and promote replication fork progression.17-19 For including the DNA damage signal transducing kinase Chk1. example, under conditions of replication stress in a Saccharomyces Using Xenopus egg extracts, a Chk1-interacting protein cerevisiae strain lacking Mrc1, the replicative MCM (minichro- termed Claspin was identified and shown to be required for the mosome maintenance) helicase co-factor Cdc45 becomes physi- phosphorylation and activation of Chk1 in response to agents cally uncoupled from the sites of DNA synthesis,18 and the leading that stall DNA polymerases and activate ATR in the extract.8,9 strand polymerase Pol ε (DNA polymerase epsilon) becomes Subsequent studies in human cells found that the human destabilized at stalled replication forks.20 Though much of this

*Correspondence to: Michael Kemp; Email: [email protected] Submitted: 03/07/11; Accepted: 03/22/11 DOI: 10.4161/cc.10.10.15562

www.landesbioscience.com Cell Cycle 1599 initial work focused on exogenous agents that induce replication have indicated that several replication fork-associated proteins stress, it has since been discovered that repetitive DNA sequence interact with Claspin and Mrc1, including DNA polymerase elements and alternative DNA structures are unstable and sensi- epsilon (Pol ε), Cdc45, Rad17-RFC and the Timeless-Tipin tive to breakage during DNA replication in yeast strains lacking complex.18,20,25-28,39-41 To determine whether any of these inter- Mrc1.21-24 Observations that Mrc1 immunoprecipitates with sev- actions are conserved in human Claspin, we immunoprecipi- eral DNA replication fork-associated proteins, including Cdc45 tated endogenous Claspin from HeLa nuclear extract using an and Pol ε,18,20,25-27 led to the suggestion that Mrc1 physically cou- antibody raised against an internal segment of human Claspin ples the activities of the replicative helicases and polymerases dur- (amino acids 725–775). As shown in Figure 1A, we observed ing DNA replication to prevent genome destabilization. Studies the presence of the catalytic subunit of Pol ε, the DNA dam- using Xenopus egg extracts have also isolated protein complexes age/ replication checkpoint protein Rad17 and the MCM helicase containing Claspin with these and other replication-fork asso- co-factor Cdc45 in the Claspin immunoprecipitates, but not in ciated factors,28 thus indicating that the role of Mrc1 in DNA the control IgG immunoprecipitates. We note that at an associa- replication may be conserved through evolution. tion of Claspin with the human homologs of Pol ε and Cdc45 has Recent studies in human cells have similarly suggested that not been previously reported, though it was expected based on human Claspin is required for normal rates of replication fork the evolutionary conservation of these proteins. Rad17 was previ- progression29,30 and to prevent genome rearrangements that may ously shown to co-immunoprecipitate with human Claspin41 and arise at stalled replication forks.31 How Claspin associates with thus served as a positive control in this experiment. Furthermore, replication forks and regulates replication fork progression is we did not observe significant amounts of the replicative clamp currently unclear. Though Claspin and its fission yeast homolog protein PCNA in complex with Claspin, suggesting that the Mrc1 both preferentially bind to replication fork-like DNA struc- isolated Claspin immunocomplexes were specific and unique. tures in vitro,32,33 Claspin does not stably associate with chroma- Though the replication fork-associating Timeless-Tipin complex tin in the absence of Cdc45.34,35 Since Cdc45 is required for the has been shown to co-immunoprecipitate with Claspin,35,39,42,43 replicative MCM helicase to unwind DNA during DNA repli- its absence in these particular immunocomplexes indicates that cation,36-38 these results suggest that the association of Claspin the interaction of Claspin with Pol ε, Rad17-RFC and Cdc45 with replication forks may be dependent on the replication fork occur independent of Timeless-Tipin. We therefore conclude that structures that are generated by the MCM helicase and/or on the Claspin-Cdc45/Pol ε/Rad17 complexes are specific and inde- direct protein-protein interactions with Cdc45. Furthermore, pendent of other proteins that are also present at replication forks. the Timeless-Tipin complex (budding yeast Tof1-Csm3 and fis- To confirm the interaction of Pol ε and Cdc45 with human sion yeast Swi1-Swi3) is an additional protein factor that asso- Claspin, we added recombinant Flag-tagged Claspin purified ciates and moves with replication forks in a Cdc45-dependent from baculovirus-infected insect cells to HeLa nuclear extract manner.18,35 The removal of Timeless-Tipin from Xenopus egg and then retrieved the Flag-Claspin protein complexes with anti- extracts and its deletion in yeast also prevents the association Flag agarose. As shown in Figure 1B, both Pol ε and Cdc45 of Claspin and Mrc1 with replicating chromatin.26,35,39 Taken were present in the Flag-Claspin immunoprecipitate. We did together, these results suggest that several protein-protein and not detect the 34 kDa RPA2 subunit of Replication Protein A protein-DNA interactions may be important for Claspin recruit- (RPA) in the immunoprecipitate, however, suggesting that the ment and function at replication forks during DNA replication complexes that Claspin forms with Pol ε and Cdc45 are specific. and in replication checkpoint signaling. Purification of Claspin and Claspin fragments. We next To identify the specific protein-protein and protein-DNA wanted to determine which, if any, of the Claspin co-immunopre- interactions that may be important for Claspin function, we cipitating proteins directly bind to Claspin. We therefore cloned, purified human Claspin, several of its functional domains and expressed and purified full-length and several smaller fragments a number of replication fork-associated proteins that have been of human Claspin in either E. coli or baculovirus-infected insect shown to co-immunoprecipitate with Claspin. We then charac- cells. A schematic of human Claspin and the various fragments terized the binding of the Claspin fragments to DNA polymerase that we purified is shown in Figure 2A. The fragments were epsilon, Rad17-RFC, Cdc45, Timeless and branched, replication designed, in part, on recent analyses of the Xenopus homolog fork-like DNA. Our results show that the Claspin N terminus of Claspin and its functional interactions with DNA, chromatin binds to Cdc45, Timeless and DNA, and that the C terminus and other proteins.28 In the schematic in Figure 2A, we note a binds to DNA polymerase epsilon and Rad17-RFC. These results region termed the replication fork-interaction domain (RFID), suggest that the binding of different regions of Claspin to specific which was shown to be sufficient for Claspin to associate with proteins and to DNA may be important for its function in DNA chromatin-containing stalled replication forks. The RFID con- replication and replication checkpoint signaling. tains two smaller regions rich in basic amino acids, which were purified and are termed basic patch I (BP1; aa 273–492) and basic Results patch II (BPII; aa 492–622). The C-terminal half of Claspin (aa 679–1,332) contains a smaller, minimal region (M; 837–1,206) Claspin interacts with several replication fork-associated recently demonstrated to be sufficient to mediate the phosphory- proteins in nuclear extracts. Co-immunoprecipitation stud- lation Chk1 by ATR in a reconstituted kinase assay with puri- ies of Xenopus Claspin and budding and fission yeast Mrc1 fied proteins.14 This fragment contains both a smaller region

1600 Cell Cycle Volume 10 Issue 10 required for binding of Xenopus Claspin to Chk1 (Chk1-binding domain; CBD) and a novel Mrc1-like domain (MRC1) identified bioinformatically and shown to be required for mediating phosphorylation of Chk1 by ATR in vitro.14 As shown in Figure 2B, we obtained relatively high quality and quantity of the GST- tagged smaller fragments of Claspin (expressed in E. coli) and Flag-tagged forms of larger regions of Claspin (expressed in Sf21 cells). These purified Claspin proteins were then used in direct protein-protein interaction assays to test for direct binding of Claspin to other replication fork-associated proteins. Claspin directly interacts with DNA polymerase epsilon, Rad17-RFC and Timeless. To examine the direct interaction of Claspin with other replication fork-associated proteins, we initially chose to purify Pol ε, Rad17-RFC and Timeless. Pol ε is a DNA polymerase involved in leading-strand replication44 and DNA replication checkpoint control. Both the Xenopus and yeast homologs of Claspin have been reported to co-immunoprecipitate with Pol ε.20,28 The human Pol ε holoenzyme was purified from insect cells co-infected with baculoviruses expressing the four subunits of Pol ε (p261, p59, p17 and p12) using a hexa- Figure 1. Claspin co-immunoprecipitates with replication fork- histidine tag on the p12 subunit. As shown in Figure 3A (left), we associated proteins in cell extracts. (A) HeLa nuclear extract (200 μg) was immunoprecipitated (IP) with either rabbit IgG or anti-Claspin obtained a high yield of Pol ε at a high level of purify. antibody. Immunoprecipitates (50%, except for the Claspin immunob- The Rad17-RFC complex facilitates the binding of the 9-1-1 lot, which was 10%) were fractionated by SDS-PAGE and analyzed by checkpoint clamp to primer-template junctions at sites of DNA immunoblotting (IB) with antibodies against the indicated proteins. damage and replication stress.1,4 Phosphorylation of Rad17 was Input lane represents 1% of the starting nuclear extract used for the im- recently shown to be important for Claspin recruitment and Chk1 munoprecipitation. The SDS gel was also stained with Coomassie blue 41 to visualize the immunoprecipitated Claspin. (B) HeLa nuclear extract activation in response to replication stress. Human Rad17-RFC (400 μg) was supplemented (+) or not (-) with recombinant full-length was purified from insect cells infected with baculoviruses express- Flag-Claspin (500 ng) and then incubated with anti-Flag agarose. Im- ing Flag-Rad17 and the RFC subunits (p36, p37, p38 and p40),4,45 munoprecipitated (Flag IP) proteins were analyzed by SDS-PAGE and using anti-Flag agarose. We obtained high quality and a high immunoblotting (IB) with antibodies against the indicated proteins. yield of the Rad17-RFC complex (Fig. 3A, middle). Input represents 1% of the starting nuclear extract used for the immunoprecipitation. The immunoprecipitated Flag-Claspin was visualized Several reports using budding and fission yeast, Xenopus by staining the nitrocellulose membrane with Ponceau S. and human cells have shown that the Timeless protein and its eukaryotic homologs (XlTim1, ScTof1 and SpSwi1) co-immunoprecipitate with Claspin and Mrc1.26,35,39,42,43,46 The Timeless We then incubated the Rad17-RFC complex with the GST- protein plays numerous roles in the cell, including in DNA repli- Claspin fragments to test for a direct interaction. As shown in cation, DNA damage checkpoints, circadian rhythms and chro- Figure 3C, Rad17 bound only to the C-terminal 679C fragment mosome cohesion.40,47 The fission yeast homolog of Timeless, of Claspin but not to any of the other fragments. We conclude Swi1, was initially reported to function as a component of a that like Pol ε, Rad17-RFC and Claspin directly interact with replication fork protection complex that stabilizes replication one another through the C-terminal half of Claspin. forks that stall at sites of DNA damage or replication stress.48,49 Using an identical approach, we incubated Flag-tagged In Xenopus egg extracts depleted of Timeless, Claspin is unable Timeless with the GST-Claspin fragments and retrieved the to associate with replicating chromatin.35 We therefore puri- complexes with glutathione resin. As shown in Figure 3D, a fied Timeless from baculovirus-infected insect cells using anti- small amount of Timeless reproducibly bound to the BPII frag- Flag agarose (Fig. 3A, right) to examine its interactions with ment (aa 492–622) of Claspin but not to any of the other frag- Claspin. ments. Importantly, deletion of this segment from the Xenopus To test for direct interactions of Pol ε, Rad17-RFC and homolog of Claspin interferes with its ability to stably associate Timeless with Claspin, we mixed each of the purified proteins with chromatin.28 (Fig. 3A) with the panel of GST-tagged Claspin fragments shown Our data showing that the Claspin C terminus directly binds in Figure 2 and then used glutathione resin to isolate the com- to Pol ε and Rad17-RFC, and that the BPII fragment interacts plexes. As shown in Figure 3B, Pol ε was specifically retrieved with Timeless indicates that specific protein-protein contacts with the C-terminal fragment of Claspin (679C). Since nearly are made between Claspin and these replication fork-associated 20% of the Pol ε bound to this Claspin fragment under these proteins. conditions, these results demonstrate that Pol ε is a bona fide Claspin directly interacts with Cdc45. We then examined Claspin-interacting protein, and that the C-terminal half of whether Claspin forms a direct protein-protein interaction Claspin is sufficient for the interaction. with Cdc45. The Cdc45 protein is an accessory factor for the

www.landesbioscience.com Cell Cycle 1601 associated with Cdc45 but not Cdc25C. The 851C fragment (aa 851–1,332) did not bind either protein. The Cdc45 interaction was reproducible and significant, indicating that Cdc45 is a bona fide Claspin-interacting protein. Claspin binds to replication fork-like DNA. Though protein-protein interactions likely play an important role in the stable association of Claspin with DNA replication forks or sites of DNA damage and replication stress, the direct interaction of Claspin with DNA may also be functionally relevant. Indeed, Claspin and Mrc1 both show a higher affinity for replication fork-like DNA structures in vitro than for either ssDNA or double-stranded DNA.32,33 However, the precise domains in human Claspin that are important for binding to DNA remain unclear. We therefore used an in vitro pull-down assay to examine the binding of Claspin fragments to a branched, replication fork-like DNA structure immobilized on streptavidin-coupled magnetic beads. We initially tested full-length Claspin and Claspin fragments under low stringency conditions with a binding buffer composed of 50 mM NaCl. As shown in Figure 5A, full-length Claspin (FL) efficiently associated with the immobilized DNA. However, two C-terminal fragments of Claspin (679C and M) completely failed to bind to DNA. To identify the specific Claspin domain(s) that bind the branched DNA, we then tested the N-terminal GST-fragments (224N, BPI and BPII). Though all three fragments associated with the DNA (Fig. 5A), the BPII region showed the strongest binding, since a higher percentage of this fragment than the Figure 2. Schematic and purification of recombinant Claspin frag- others was retained on the DNA beads. To further analyze the ments. (A) Schematic of Claspin fragments cloned and purified in either E. coli or insect cells. Abbreviations for the names are provided DNA binding domains of Claspin under more stringent condi- to the right of the schematic. The numbers indicate the amino acid tions, we modified the binding buffer conditions with a higher (aa) positions. Several important functional domains are presented salt concentration (100 mM). As shown in Figure 5B, only in the schematic, including the replication fork-interacting domain the full-length and BPII fragment of Claspin were retained on (RFID, amino acids 273–622), basic patch I (BPI, aa 273–492), basic patch the immobilized DNA under these conditions. Based on these II (BPII, aa 492–622), Chk1-binding domain (CBD, aa 899–999) and an Mrc1-like domain (MRC1, aa 1,045–1,203). The purified Claspin proteins results, we conclude that the BPII region of Claspin plays an and fragments included: full-length (FL), aa 851–1,332 (851C), aa 1–851 important role in the binding of Claspin to branched, replica- (851N), aa 679–1,332 (679C), aa 837–1,206 (M; minimal checkpoint tion fork-like DNA structures. domain), aa 1–224 (224N), aa 273–492 (BPI) and aa 492–622 (BPII). (B) The indicated recombinant Claspin fragments were purified with either a GST epitope from E. coli or with a Flag-epitope from baculovirus-infect- Discussion ed insect cells. An aliquot of each of the purified proteins was analyzed by SDS-PAGE and Coomassie blue staining. The numbers to the left In this report, we showed that human Claspin co-immunopre- of the Coomassie-stained gel indicates the location of the molecular cipitates with Pol ε and the replicative helicase co-factor Cdc45 weight markers (kDa). in human cells (Fig. 1). Similar approaches have previously shown that such complexes exist in yeast and in Xenopus egg replicative MCM helicase and is essential for MCM helicase activ- extracts.18,20,25,27,28,51 However, a major limitation of the co-immu- ity in vitro36,37 and for unwinding of DNA at replication origins.38 noprecipitation approach is that it does not provide information Cdc45 is also a target of the ATR- and Chk1-mediated intra- about the specific protein-protein interactions that are impor- S-phase DNA damage checkpoint.50 In Xenopus egg extracts tant for protein complex formation. These details are necessary depleted of Cdc45, Claspin is unable to associate with chroma- to investigate how specific protein-protein interactions regulate tin.34,35 We therefore cloned and expressed a GST-tagged form of enzyme activities. This information is particularly important for Cdc45 and purified it with glutathione resin (Fig. 4A). We chose understanding the mechanisms of multi-protein enzymatic pro- to use GST-tagged Cdc25C (Fig. 4B) as a negative control for cesses, such as DNA replication and DNA replication checkpoint these binding experiments. We then incubated Flag-tagged forms signaling. of Claspin (full-length, 851N and 851C; Fig. 2) with the GST- To gain a better understanding of the potential mechanisms Cdc45 and GST-Cdc25C proteins and used glutathione resin of Claspin function, we purified all of the relevant replication to isolate the protein complexes. As shown in Figure 4C, the fork-associated, Claspin co-immunoprecipitating proteins in full-length Claspin and 851N fragment (aa 1–851) specifically recombinant form and examined direct binding to Claspin. We

1602 Cell Cycle Volume 10 Issue 10 Figure 3. Claspin directly interacts with Pol ε, Rad17-RFC and Timeless. (A) Purified Pol ε (left), Rad17-RFC (middle part) and Timeless (right) were purified as described in the Materials and Methods section and analyzed by SDS-PAGE and Coomassie blue staining. Arrows indicate each protein or subunit. The asterisk represents impurities in the purified protein. The numbers to the left of the Coomassie-stained gel indicate the location of the molecular weight markers (kDa). The indicated GST-Claspin fragments shown in Figure 2 were mixed with either (B) Pol ε, (C) Rad17-RFC or (D) Timeless, retrieved with glutathione resin and analyzed by SDS-PAGE and immunoblotting with antibodies against the indicated proteins. The GST-Claspin fragments were visualized by staining the membrane with Ponceau S. Input represents 10% of the protein tested in the binding assays. The relative amount of each protein that co-precipitated with each Claspin fragment was quantified and normalized relative to the input signal. Error bars represent the standard deviation of the mean from two (Timeless) or three independent experiments (Pol ε, Rad17-RFC). further characterized the specific regions of Claspin that are involved in these protein-protein interactions and for binding to branched, replication fork-like DNA. A summary of these find- ings are provided in Figure 6. Our data show that the N terminus of Claspin, which contains a replication fork-interacting domain (RFID) previously shown to be sufficient for Claspin association with chromatin in Xenopus egg extracts,28 is also involved in directly binding to the replicative MCM helicase co-factor Cdc45 (Fig. 4) and the replication fork protection complex protein Timeless (Fig. 3). Because Claspin fails to associate with replicating chromatin in Xenopus egg extracts lacking either Cdc45,34 or Timeless,35 these results suggest that the interactions of Cdc45 and Timeless with the Claspin N terminus may be important for stably recruiting Claspin to newly generated replication forks. Alternatively, Cdc45-dependent MCM helicase activity may gen- erate unwound DNA structures that allow Claspin to directly bind to DNA. Consistent with this latter hypothesis, both Claspin and Mrc1 preferentially bind branched DNA structures through N-terminal domains.32,33 Though there are multiple DNA binding elements within this region of Claspin, the BPII domain has the greatest affinity for DNA (Fig. 5). In yeast strains lacking Mrc1 and exposed to the nucleotide- depleting agent hydroxyurea, Cdc45 becomes physically uncoupled from the sites of DNA replication,18 suggesting that Mrc1 may control the activity of the replicative DNA helicase. Thus, the direct interaction that we observed between Claspin and Cdc45 (Fig. 4) may regulate Cdc45-dependent MCM helicase activity. Clearly, additional work is necessary to understand how the Cdc45-Claspin, Timeless-Claspin and DNA-Claspin interactions affect the functions of these proteins in DNA replication and DNA replication checkpoint signaling. Whereas the N terminus of Claspin appears to have functions relating to its association with DNA replication forks, the C terminus is primarily thought to be involved in ATR-mediated DNA damage and replication checkpoint signaling.14,28 Importantly, a Interestingly, the C-terminal half of Claspin also mediates region in the C-terminal part of Claspin was recently shown to its interactions with two additional proteins—DNA polymerase be sufficient to directly mediate Chk1 phosphorylation by ATR epsilon and Rad17-RFC. The Rad17-RFC complex has a well- in a defined system in vitro.14 This minimal checkpoint domain recognized role in loading the alternative DNA damage 9-1-1 (Fig. 6) contains both a Chk1-binding domain (CBD) and a clamp onto primer-template junctions at sites of DNA damage region of similarity to yeast Mrc1. and replication stress in order to facilitate activation of ATR.1,4

www.landesbioscience.com Cell Cycle 1603 In conclusion, our results suggest a model where Claspin function in DNA replication and DNA damage as well as replication checkpoints may be influenced by multiple protein-protein and protein- DNA interactions. An understanding of the domains important for these interactions will therefore facilitate future work aimed at understanding how these various protein-protein and DNA-protein interactions influence DNA replication, DNA damage checkpoints and genomic stability.

Materials and Methods

Protein purification. Proteins were purified from either bacterial or insect expression systems. Monolayer Sf21 insect cells were cultured in Grace’s insect medium with 10% fetal bovine serum (FBS). The DNA polymerase epsilon holoenzyme was expressed in insect cells after infection with baculoviruses encoding the p12 (His-tagged), p17, p59 and p261 subunits (gift from J. Hurwitz) and then purified with Ni-NTA agarose using standard pro- cedures (Qiagen). The Rad17-RFC complex was purified from insect cells infected with baculoviruses encoding the Rad17 (Flag-tagged), p37, p40, p37 and p38 (His-tagged) subunits using anti-Flag agarose (Sigma).4,45 Flag-Timeless and Flag-Claspin con- structs were also expressed in baculovirus-infected Figure 4. Claspin directly interacts with Cdc45. (A) GST-Cdc45 and (B) GST-Cdc25C insect cells and purified with anti-Flag agarose, as pre- were purified from E. coli with glutathione resin and analyzed by SDS-PAGE and Coo- viously described in reference 32 and 52. The GST- massie blue staining. The arrows indicate the full-length purified proteins.T he num- Claspin fragments shown in Figure 2 were cloned bers to the left of the Coomassie-stained gel indicates the location of the molecular into pGEX-4T1 (GE Life Sciences), expressed in weight markers (kDa). The purified GST-Cdc45 had a number of degradation prod- ucts, as visualized in the Coomassie-stained gel and verified by immunoblotting with DH5α and purified with Glutathione Sepharose 4B an anti-Cdc45 antibody. (C) Flag-tagged Claspin fragments shown in Figure 2 were (Amersham Biosciences). GST‑Cdc25C (Addgene mixed with either GST-Cdc45 or GST-Cdc25C, incubated with glutathione resin and Plasmid 10969)53 was purified in a similar man- bound proteins analyzed by SDS-PAGE and immunoblotting with anti-Flag antibody ner. The Cdc45 cDNA was cloned into pGEX-4T1, and Ponceau S staining of the membrane. (D) The relative amount of each protein expressed in DH5α and purified with Glutathione that co-precipitated with each GST-tagged protein was quantified and normalized relative to the input signal (10% of material used in the binding assay). Error bars rep- Sepharose 4B resin. The quality of each protein resent the standard deviation of the mean from three independent experiments. was analyzed by SDS-PAGE and Coomassie blue staining. Detailed protocols regarding the cloning, expression and purification of specific proteins and Importantly, a previous report suggested that the phosphoryla- protein fragments are available upon request. tion of Rad17 by checkpoint kinases was required for Claspin Protein-protein interaction studies. Target and bait proteins recruitment and function at sites of DNA replication stress.41 (5 pmol) were mixed in 50 μl reactions containing Tris-buffered Though our data confirm that Rad17-Claspin complexes exist saline (50 mM Tris pH 7.5, 150 mM NaCl) and then incubated in human cells (Fig. 1), our data now show that the interaction at room-temperature for 1 h. Reactions were then diluted to 1 ml involves direct contacts between the two proteins (Fig. 3). with Buffer A (50 mM Tris pH 7.5, 150 mM NaCl, 0.1% NP-40, We further found that the same C-terminal region of Claspin 0.1% Tween-20), supplemented with Glutathione Sepharose (679C) is also involved in binding to Pol ε. This is in agreement 4B resin, and then incubated for 2 h at 4°C. After centrifugation with a recent publication in fission yeast, where it was reported and three washes with Buffer A, proteins were eluted with 1x SDS- that the C terminus of Mrc1 was sufficient to co-immunopre- PAGE sample buffer, fractionated by SDS-PAGE and transferred cipitate with the C terminus of Pol ε.20 It should be noted that to nitrocellulose. Membranes were stained with Ponceau S and this interaction affected yeast growth rate, independent of Mrc1 then analyzed by immunoblotting with the indicated antibodies. checkpoint functions. It will therefore be important to determine Protein-DNA interaction studies. A branched DNA struc- how this interaction affects cell growth or DNA replication rate ture was generated by mixing Oligo1 (5'-Biotin-GAC GCT GCC in human cells. GAA TTC TGG CTT GCT AGG ACA TCT TTG CCC ACG

1604 Cell Cycle Volume 10 Issue 10 Figure 6. Schematic of Claspin functional domains. A summary of the functional regions of Claspin are indicated, based on data described in the text. BPI and BPII indicate basic patches I and II. RFID, replication fork-interacting domain. CKB, Chk1-binding domain. MRC1, Mrc1-like domain.

Antibodies. Rat monoclonal anti-Cdc45 antibody was a gift from Cyrus Vaziri.50 Mouse anti-Rad17 antibody was described previously in reference 54. Anti-RPA34 antibody (NA-18) was Figure 5. The Claspin N terminus, but not the C terminus, binds to rep- purchased from Calbiochem. A monoclonal antibody recogniz- lication fork-like DNA in vitro. Each of indicated the GST-Claspin frag- ing the catalytic subunit (p261) of DNA polymerase epsilon ments (20, 40 and 100 nM) was incubated with a biotinylated, branched (ab104) was purchased from Abcam. Antibodies against Claspin DNA substrate (20 nM) immobilized on magnetic beads, as described in the Materials and Methods section. Binding reactions were performed (H-300; sc-48771) and PCNA (PC10; sc-56) were from Santa with either (A) 50 mM NaCl or (B) 100 mM NaCl in the binding and wash Cruz Biotechnology. The guinea pig anti-Timeless antibody was buffer. Proteins bound to the DNA beads were analyzed by SDS-PAGE a gift from Anthony Gotter.42 The Claspin antibody used for and immunoblotting. Input lanes contain 0.5 pmol of the indicated immunoprecipitation was from Bethyl (A300-266). protein. Signals were quantified using the input signal for normalization Immunoprecipitation. HeLa nuclear extracts were pre- and then graphed. Data represent the mean and standard deviation from three independent experiments. pared as previously described in reference 55. Nuclear extracts (200 μg; 100 μl) were mixed with 20 μg of control rabbit IgG or anti-Claspin antibody (Bethyl A300-266) prebound to rProtein TTG ACC CG-3') and Oligo2 (5'-GCG ATA GTC TCT AGA A/G+ agarose (Santa Cruz) for 4 h. Reactions were then diluted CAG CAT GTG GTA GCA AGC CAG AAT TCG GCA GCG to 1 ml with Buffer A and incubated for an additional 1 h at 4°C. TC-3') in an annealing buffer (50 mM Tris pH 7.4, 50 mM NaCl). The resin was washed with Buffer A three times and the proteins After heating at 95°C for 5 min and slow cooling, the biotinyl- were eluted with 1x SDS-PAGE loading buffer for analysis by ated, branched DNA structure was immobilized on Dynabeads SDS-PAGE and immunoblotting. Immunoprecipitation of Flag- M-280 Streptavidin (Invitrogen) according to the manufacturer’s Claspin with anti-Flag agarose was performed in a similar man- recommendations. Pull-down assays examining the association ner after adding recombinant Flag-tagged Claspin (500 ng) to of the Claspin fragments with the immobilized DNA were per- HeLa nuclear extract (400 μg). formed as previously described in reference 43. Binding reactions contained the indicated concentrations of proteins and 20 nM Acknowledgements DNA in a 50 μl reaction buffer (20 mM Tris pH 7.4, 50 or 100 We thank Laura Lindsey-Boltz for her comments on the manu- mM NaCl, 0.02% NP-40 and 8% glycerol). Associated proteins script. This work was supported by National Institutes of Health were analyzed by SDS-PAGE and immunoblotting. grant GM32833 to Aziz Sancar.

www.landesbioscience.com Cell Cycle 1605 References 21. Freudenreich CH, Lahiri M. Structure-forming CAG/ 39. Errico A, Costanzo V, Hunt T. Tipin is required for CTG repeat sequences are sensitive to breakage in the stalled replication forks to resume DNA replication 1. Sancar A, Lindsey-Boltz LA, Unsal-Kacmaz K, Linn absence of Mrc1 checkpoint function and S-phase after removal of aphidicolin in xenopus egg extracts. S. Molecular mechanisms of mammalian DNA repair checkpoint signaling: Implications for trinucleotide Proc Natl Acad Sci USA 2007. and the DNA damage checkpoints. Annu Rev Biochem repeat expansion diseases. Cell Cycle 2004; 3:1370-4. 40. Gotter AL. A timeless debate: Resolving TIM’s noncir- 2004; 73:39-85. 22. Razidlo DF, Lahue RS. Mrc1, Tof1 and Csm3 inhibit cadian roles with possible clock function. Neuroreport 2. Ciccia A, Elledge SJ. The DNA damage response: CAG.CTG repeat instability by at least two mecha- 2006; 17:1229-33. Making it safe to play with knives. Mol Cell 2010; nisms. DNA Repair (Amst) 2008; 7:633-40. 41. Wang X, Zou L, Lu T, Bao S, Hurov KE, Hittelman 40:179-204. 23. Voineagu I, Narayanan V, Lobachev KS, Mirkin WN, et al. Rad17 phosphorylation is required for 3. Zou L, Elledge SJ. Sensing DNA damage through SM. Replication stalling at unstable inverted repeats: claspin recruitment and Chk1 activation in response to ATRIP recognition of RPA-ssDNA complexes. Science Interplay between DNA hairpins and fork stabilizing replication stress. Mol Cell 2006; 23:331-41. 2003; 300:1542-8. proteins. Proc Natl Acad Sci USA 2008; 105:9936-41. 42. Gotter AL, Suppa C, Emanuel BS. Mammalian 4. Bermudez VP, Lindsey-Boltz LA, Cesare AJ, Maniwa 24. Voineagu I, Surka CF, Shishkin AA, Krasilnikova MM, TIMELESS and tipin are evolutionarily conserved Y, Griffith JD, Hurwitz J, et al. Loading of the Mirkin SM. Replisome stalling and stabilization at replication fork-associated factors. J Mol Biol 2007; human 9-1-1 checkpoint complex onto DNA by the CGG repeats, which are responsible for chromosomal 366:36-52. checkpoint clamp loader hRad17-replication factor fragility. Nat Struct Mol Biol 2009; 16:226-8. 43. Kemp MG, Akan Z, Yilmaz S, Grillo M, Smith-Roe C complex in vitro. Proc Natl Acad Sci USA 2003; 25. Gambus A, Jones RC, Sanchez-Diaz A, Kanemaki M, SL, Kang TH, et al. Tipin-RPA interaction mediates 100:1633-8. van Deursen F, Edmondson RD, et al. GINS maintains Chk1 phosphorylation by ATR in response to geno- 5. Kumagai A, Lee J, Yoo HY, Dunphy WG. TopBP1 acti- association of Cdc45 with MCM in replisome progres- toxic stress. J Biol Chem 2010. vates the ATR-ATRIP complex. Cell 2006; 124:943-55. sion complexes at eukaryotic DNA replication forks. 44. Pursell ZF, Isoz I, Lundstrom EB, Johansson E, 6. Delacroix S, Wagner JM, Kobayashi M, Yamamoto Nat Cell Biol 2006; 8:358-66. Kunkel TA. Yeast DNA polymerase epsilon participates K, Karnitz LM. The Rad9-Hus1-Rad1 (9-1-1) clamp 26. Bando M, Katou Y, Komata M, Tanaka H, Itoh T, in leading-strand DNA replication. Science 2007; activates checkpoint signaling via TopBP1. Genes Dev Sutani T, et al. Csm3, Tof1 and Mrc1 form a hetero- 317:127-30. 2007; 21:1472-7. trimeric mediator complex that associates with DNA 45. Lindsey-Boltz LA, Bermudez VP, Hurwitz J, Sancar A. 7. Cimprich KA, Cortez D. ATR: An essential regulator replication forks. J Biol Chem 2009; 284:34355-65. Purification and characterization of human DNA dam- of genome integrity. Nat Rev Mol Cell Biol 2008; 27. Nedelcheva MN, Roguev A, Dolapchiev LB, age checkpoint rad complexes. Proc Natl Acad Sci USA 9:616-27. Shevchenko A, Taskov HB, Shevchenko A, et al. 2001; 98:11236-41. 8. Kumagai A, Dunphy WG. Claspin, a novel protein Uncoupling of unwinding from DNA synthesis implies 46. Tanaka T, Yokoyama M, Matsumoto S, Fukatsu R, You Z, required for the activation of Chk1 during a DNA rep- regulation of MCM helicase by Tof1/Mrc1/Csm3 Masai H. Fission yeast Swi1-Swi3 complex facilitates DNA lication checkpoint response in xenopus egg extracts. checkpoint complex. J Mol Biol 2005; 347:509-21. binding of Mrc1. J Biol Chem 2010; 285:39609-22. Mol Cell 2000; 6:839-49. 28. Lee J, Gold DA, Shevchenko A, Shevchenko A, 47. Unsal-Kacmaz K, Mullen TE, Kaufmann WK, Sancar 9. Kumagai A, Kim SM, Dunphy WG. Claspin and the Dunphy WG. Roles of replication fork-interacting A. Coupling of human circadian and cell cycles by the activated form of ATR-ATRIP collaborate in the activa- and Chk1-activating domains from claspin in a DNA timeless protein. Mol Cell Biol 2005; 25:3109-16. tion of Chk1. J Biol Chem 2004; 279:49599-608. replication checkpoint response. Mol Biol Cell 2005; 48. Noguchi E, Noguchi C, Du LL, Russell P. Swi1 pre- 10. Chini CC, Chen J. Human claspin is required for 16:5269-82. vents replication fork collapse and controls checkpoint replication checkpoint control. J Biol Chem 2003; 29. Petermann E, Helleday T, Caldecott KW. Claspin kinase Cds1. Mol Cell Biol 2003; 23:7861-74. 278:30057-62. promotes normal replication fork rates in human cells. 49. Noguchi E, Noguchi C, McDonald WH, Yates JR, 3rd, 11. Lin SY, Li K, Stewart GS, Elledge SJ. Human claspin Mol Biol Cell 2008; 19:2373-8. Russell P. Swi1 and Swi3 are components of a replica- works with BRCA1 to both positively and negatively 30. Scorah J, McGowan CH. Claspin and Chk1 regulate tion fork protection complex in fission yeast. Mol Cell regulate cell proliferation. Proc Natl Acad Sci USA replication fork stability by different mechanisms. Cell Biol 2004; 24:8342-55. 2004; 101:6484-9. Cycle 2009; 8:1036-43. 50. Liu P, Barkley LR, Day T, Bi X, Slater DM, Alexandrow 12. Liu S, Bekker-Jensen S, Mailand N, Lukas C, Bartek 31. Focarelli ML, Soza S, Mannini L, Paulis M, MG, et al. The Chk1-mediated S-phase check- J, Lukas J. Claspin operates downstream of TopBP1 Montecucco A, Musio A. Claspin inhibition leads to point targets initiation factor Cdc45 via a Cdc25A/ to direct ATR signaling towards Chk1 activation. Mol fragile site expression. Genes Chromosomes Cancer Cdk2-independent mechanism. J Biol Chem 2006; Cell Biol 2006; 26:6056-64. 2009; 48:1083-90. 281:30631-44. 13. Lupardus PJ, Cimprich KA. Phosphorylation of xeno- 32. Sar F, Lindsey-Boltz LA, Subramanian D, Croteau 51. Calzada A, Hodgson B, Kanemaki M, Bueno A, Labib pus Rad1 and Hus1 defines a readout for ATR acti- DL, Hutsell SQ, Griffith JD, et al. Human claspin is K. Molecular anatomy and regulation of a stable repli- vation that is independent of claspin and the Rad9 a ring-shaped DNA-binding protein with high affin- some at a paused eukaryotic DNA replication fork. carboxy terminus. Mol Biol Cell 2006; 17:1559-69. ity to branched DNA structures. J Biol Chem 2004; Genes Dev 2005; 19:1905-19. 14. Lindsey-Boltz LA, Sercin O, Choi JH, Sancar A. 279:39289-95. 52. Unsal-Kacmaz K, Chastain PD, Qu PP, Minoo P, Reconstitution of human claspin-mediated phosphory- 33. Zhao H, Russell P. DNA binding domain in the repli- Cordeiro-Stone M, Sancar A, et al. The human Tim/ lation of Chk1 by the ATR (ataxia telangiectasia- cation checkpoint protein Mrc1 of Schizosaccharomyces Tipin complex coordinates an intra-S checkpoint mutated and rad3-related) checkpoint kinase. J Biol pombe. J Biol Chem 2004; 279:53023-7. response to UV that slows replication fork displace- Chem 2009; 284:33107-14. 34. Lee J, Kumagai A, Dunphy WG. Claspin, a Chk1- ment. Mol Cell Biol 2007; 27:3131-42. 15. Alcasabas AA, Osborn AJ, Bachant J, Hu F, Werler regulatory protein, monitors DNA replication on 53. Savitsky PA, Finkel T. Redox regulation of Cdc25C. J PJ, Bousset K, et al. Mrc1 transduces signals of DNA chromatin independently of RPA, ATR and Rad17. Biol Chem 2002; 277:20535-40. replication stress to activate Rad53. Nat Cell Biol 2001; Mol Cell 2003; 11:329-40. 54. Chang MS, Sasaki H, Campbell MS, Kraeft SK, 3:958-65. 35. Tanaka H, Kubota Y, Tsujimura T, Kumano M, Masai Sutherland R, Yang CY, et al. HRad17 colocalizes with 16. Tanaka K, Russell P. Mrc1 channels the DNA replica- H, Takisawa H. Replisome progression complex links NHP2L1 in the nucleolus and redistributes after UV tion arrest signal to checkpoint kinase Cds1. Nat Cell DNA replication to sister chromatid cohesion in xeno- irradiation. J Biol Chem 1999; 274:36544-9. Biol 2001; 3:966-72. pus egg extracts. Genes Cells 2009; 14:949-63. 55. Dignam JD, Lebovitz RM, Roeder RG. Accurate tran- 17. Tourriere H, Versini G, Cordon-Preciado V, Alabert 36. Ilves I, Petojevic T, Pesavento JJ, Botchan MR. scription initiation by RNA polymerase II in a soluble C, Pasero P. Mrc1 and Tof1 promote replication fork Activation of the MCM2-7 helicase by association with extract from isolated mammalian nuclei. Nucleic Acids progression and recovery independently of Rad53. Mol Cdc45 and GINS proteins. Mol Cell 2010; 37:247-58. Res 1983; 11:1475-89. Cell 2005; 19:699-706. 37. Moyer SE, Lewis PW, Botchan MR. Isolation of the 18. Katou Y, Kanoh Y, Bando M, Noguchi H, Tanaka H, Cdc45/Mcm2-7/GINS (CMG) complex, a candidate Ashikari T, et al. S-phase checkpoint proteins Tof1 for the eukaryotic DNA replication fork helicase. Proc and Mrc1 form a stable replication-pausing complex. Natl Acad Sci USA 2006; 103:10236-41. Nature 2003; 424:1078-83. 38. Pacek M, Walter JC. A requirement for MCM7 and 19. Szyjka SJ, Viggiani CJ, Aparicio OM. Mrc1 is required Cdc45 in chromosome unwinding during eukaryotic for normal progression of replication forks throughout DNA replication. EMBO J 2004; 23:3667-76. chromatin in S. cerevisiae. Mol Cell 2005; 19:691-7. 20. Lou H, Komata M, Katou Y, Guan Z, Reis CC, Budd M, et al. Mrc1 and DNA polymerase epsilon function together in linking DNA replication and the S phase checkpoint. Mol Cell 2008; 32:106-17.

1606 Cell Cycle Volume 10 Issue 10