Exposure of single-stranded telomeric DNA causes G2/M cell cycle arrest in Saccharomyces cerevisiae.

In Saccharomyces cerevisiae, Cdc13p is a single-stranded TG(1-3) DNA binding protein that protects telomeres and maintains telomere length. A mutant allele of CDC13, cdc13-1, causes accumulation of single-stranded TG(1-3) DNA near telomeres along with a G(2)/M cell cycle arrest at non-permissive temperatures. We report here that when the single-stranded TG(1-3) DNA is masked by its binding proteins, such as S. cerevisiae Gbp2p or Schizosaccharomyces pombe Tcg1, the growth arrest phenotype of cdc13-1 is rescued. Mutations on Gbp2p that disrupt its binding to the single-stranded TG(1-3) DNA render the protein unable to complement the defects of cdc13-1. These results indicate that the presence of a single-stranded TG(1-3) tail in cdc13-1 cells serves as the signal for the cell cycle checkpoint. Moreover, the binding activity of Gbp2p to single-stranded TG(1-3) DNA appears to be associated with its ability to restore the telomere-lengthening phenotype in cdc13-1 cells. These results indicate that Gbp2p is involved in modulating telomere length.

In Saccharomyces cerevisiae, Cdc13p is a singlestranded TG 1-3 DNA binding protein that protects telomeres and maintains telomere length. A mutant allele of CDC13, cdc13-1, causes accumulation of singlestranded TG 1-3 DNA near telomeres along with a G 2 /M cell cycle arrest at non-permissive temperatures. We report here that when the single-stranded TG 1-3 DNA is masked by its binding proteins, such as S. cerevisiae Gbp2p or Schizosaccharomyces pombe Tcg1, the growth arrest phenotype of cdc13-1 is rescued. Mutations on Gbp2p that disrupt its binding to the single-stranded TG 1-3 DNA render the protein unable to complement the defects of cdc13-1. These results indicate that the presence of a single-stranded TG 1-3 tail in cdc13-1 cells serves as the signal for the cell cycle checkpoint. Moreover, the binding activity of Gbp2p to single-stranded TG 1-3 DNA appears to be associated with its ability to restore the telomere-lengthening phenotype in cdc13-1 cells. These results indicate that Gbp2p is involved in modulating telomere length.
Telomeres are the structure at the ends of eukaryotic linear chromosomes (1,2). In most organisms, the telomeric DNA is composed of short, tandem repeated sequences with a strand rich in guanine residues (G-strand) running 5Ј to 3Ј toward the end of telomere. For example, the telomeric sequences in the brewers yeast Saccharomyces cerevisiae are ϳ250 -300 base pair-long TG 1-3 /C 1-3 A repeats. Sequences of ciliate telomeres reveal that the G-strand extends beyond the duplex region, creating a short single-stranded 3Ј-overhang (3)(4)(5). In yeast, longer G-strand DNA with varying lengths, presumably an intermediate during telomere replication, is detected during late S phase (6,7). Telomeres are essential for the maintenance of chromosome integrity. Telomeres protect chromosomes from degradation by nucleases, facilitate complete replication of chromosomes, and differentiate linear chromosome ends from broken ends (1,2).
Several single-stranded telomeric DNA binding proteins, including Oxytricha ␣ and ␤ subunits (8,9), Cdc13p (10, 11), Gbp2p (12,13), hnRNPs (14 -16), and Pot1 (17), have been identified in vitro. Among these proteins, Oxytricha ␣ and ␤ subunits have been well characterized because of the abun-dance of telomeres in this organism. The ␣ subunit binds to the G 4 T 4 G 4 T 4 single-stranded end of the telomere, and the ␤ subunit is required for making the terminus-specific binding (8,18,19). Oxytricha ␣and ␤-like-binding proteins represent a novel type of DNA-binding protein because their binding is extremely salt-resistant. The binding of other single-stranded telomeric binding proteins to telomeric DNA is relatively salt-sensitive. A protein Pot1 that shares partial sequence homology with the Oxytricha ␣ subunit has been identified from human and Schizosaccharomyces pombe (17). Mutation in S. pombe POT1 causes loss of telomeric DNA and circulation of chromosomes. In S. cerevisiae, Cdc13p binds specifically to single-stranded TG 1-3 DNA in vitro and affects telomere function in vivo (10,11,20,21). Interestingly, even though Cdc13p does not share sequence similarity with the ciliate proteins or Pot1, the binding regions of Cdc13p and ciliate proteins that interact with singlestranded telomeric DNA are conserved (22,23).
A mutation allele of CDC13, cdc13-1, causes yeast cells to arrest at the G 2 /M phase of the cell cycle and accumulate single-stranded TG 1-3 DNA at telomeres (20). Various types of DNA damages, including x-ray, UV, and chemical mutagens, induce DNA damage-dependent cell cycle arrest (24 -28). Presumably, these damages on DNA were processed by cellular nucleases to generate single-stranded DNA, which could then serve as the signal for DNA damage-dependent cell cycle arrest (20,29). Here, we show that S. cerevisiae Gbp2p and S. pombe Tcg1 rescue the growth arrest of cdc13-1 cells and that their single-stranded TG 1-3 DNA binding activity is required for their complementation. Our results suggest that the appearance of single-stranded telomeric DNA in cdc13-1 cells is the signal that induces the DNA damage-dependent checkpoint. Moreover, we also provide evidence that Gbp2p is involved in modulating telomere length.
Cloning and Sequencing of S. pombe TCG1-To clone S. pombe genes that complement the cdc13-1 temperature-sensitive phenotype, an S. pombe cDNA library (32) was transformed into W13a. Transformants were plated, incubated at 25°C for 16 h, and then at 30°C for 4 days. A total of 72 colonies grew at 30°C from ϳ120,000 transformants. DNAs from these 72 colonies were prepared and transformed into Escherichia coli XL1-Blue. Plasmid DNAs were prepared and transformed back into W13a to retest their temperature sensitivity at 30°C. Only one clone grew at 30°C after the retransformation of purified plasmids. A 1.4-kbp HindIII-HindIII DNA insert within this complementing plasmid, pDB-TCG1, was subcloned into HindIII-digested pVZ1 to generate pVZ-TCG1. The sequence of this DNA was determined by the exonuclease III deletion method and sequenced using Sequenase (Upstate Biotechnology). To construct pGST-TCG1, the TCG1 fragment was first PCR amplified with primers TCG5 (5Ј-CGAATTCATGATGTCCGCTGAGGAAAC-3Ј) and TCG3 (5Ј-CGTC-GACTTAGGCAGTTATAGCACTAG-3Ј). A PCR-amplified ϳ1.0-kbp DNA fragment was digested with EcoRI and SalI and ligated with the EcoRI-and XhoI-digested pGEX-4T-1 to generate pGST-TCG1.
Site-specific Mutagenesis-PCR-based site-specific mutagenesis was applied to generate mutations on the RNA recognition motifs (RRMs) of GBP2. Mutations on residues Arg-161 and Arg-259 of Gbp2p were changed to Ala on plasmid pMAL-GBP2 to generate pMAL-GBP2 R161A and pMAL-GBP2 R259A , respectively. The mutation pMAL-GBP2 R161A, R259A was generated by conducting a second mutagenesis on plasmid pMAL-GBP2 R259A by changing Arg-161 to Ala. A similar approach was applied to generate mutations on pGAL-GBP2 R161A , pGAL-GBP2 R259A , and pGAL-GBP2 R161A, R259A . All the mutations were confirmed by DNA sequencing.
Protein Purification-To purify glutathione S-transferase (GST) 2fused Tcg1, a 1000-ml culture of E. coli harboring pGST-TCG1 was grown at 30°C to an A 600 of 0.8 and induced with the addition of 1 mM IPTG. The cells were grown at 30°C for another 3 h before being harvested by centrifugation. Cells were resuspended in 20 ml of sonication buffer (50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 0.5% Triton X-100, 1 mM DTT, 50 mM NaOAc, 20% glycerol, and 1ϫ protease inhibitors (Calbiochem)) and sonicated to release the cell contents. The sonicated cells were centrifuged at 13,000 ϫ g for 15 min at 4°C to get rid of soluble cell fractions. The pellet was then extracted with sonication buffer with 0.25% Tween 20 and 0.1 mM EGTA to obtain GST-Tcg1 fusion protein. Purified protein was dialyzed against storage buffer (50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 0.5% Triton X-100, 1 mM DTT, 50 mM NaOAc, 50% glycerol), aliquoted, and frozen by dry ice-ethanol bath.
To purify maltose-binding protein (MBP)-fused Gbp2p, 100 ml culture of E. coli harboring pMBP-GBP2, was grown at 30°C to an A 600 of 0.5 and induced with the addition of 1 mM IPTG. The cells were grown at 30°C for another 16 h before being harvested by centrifugation. Cells were resuspended in 10 ml of sonication buffer (50 mM NaH 2 PO 4 , pH 7.8, 1 mM EDTA, 0.5% Triton X-100, 1 mM DTT, 5% glycerol, 1 mM phenylmethylsulfonyl fluoride, and 1ϫ protease inhibitors (Calbiochem)) and sonicated to release the cell contents. The sonicated cells were centrifuged at 13,000 ϫ g for 15 min at 4°C to obtain total cell-free extracts. 0.5 ml of amylose-agarose (New England Biolabs) was added to the total cell-free extracts and incubated at 4°C for 1 h. The resin was washed and eluted with 2 ml of buffer containing 50 mM NaH 2 PO 4 , pH 8.0, 100 mM maltose, and 20% glycerol. Purified protein was aliquoted and frozen by a dry ice-ethanol bath. Gbp2p with different point mutations were purified using similar procedures.
Electrophoretic Mobility Shift Assay (EMSA)-Oligonucleotide TG43 (5Ј-GTGGTGGGTGGGTGTGTGTGGGTGTGGTGGGTGTGTGGGTG-TG-3Ј) was labeled with [␥-32 P]ATP (3000 mCi/mM, PerkinElmer Life Sciences) using T4 polynucleotide kinase (New England Biolabs) and subsequently purified from a 10% sequencing gel after electrophoresis. To perform the assays, Gbp2p, its mutants, or E. coli extracts harboring Tcg1 in Buffer A (50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 50 mM NaCl, 1 mM DTT, and 1 g of heat-denatured poly(dI-dC)) was mixed with 5 nM of 32 P-labeled TG43 DNA in a total volume of 15 l. The reaction mixture was incubated at room temperature for 10 min and then loaded directly onto an 8% non-denaturing polyacrylamide gel already pre-run at 125 V for 10 min. Electrophoresis was carried out in TBE (89 mM Tris borate, 2 mM EDTA) at 125 V for 105 min. The gels were dried and autoradiographed.
Single-stranded Telomere Tail and Telomere Length Determinations-Yeast cells CDC13/pSH380 or cdc13-1 cells harboring plasmid pSH380, pGAL-GBP2, pGAL-GBP2 R161A , pGAL-GBP2 R259A , or pGAL-GBP2 R161A, R259A were grown in YC-Leu in the presence of 2% glucose at 25°C for 2 days. Cells were then washed with water, resuspended in YC-Leu medium with 3% galactose, and continued to grow at 30°C for another 6 h. DNA from these cells was prepared, digested with XhoI, and separated on 1% agarose gels. Single-stranded telomere tails were determined by nondenaturing Southern hybridization as described previously (6). To determine the telomere length, DNA fragments were transferred to a Hybond N ϩ paper (Amersham Pharmacia Biotech) for hybridization using a random primed C 1-3 A/TG 1-3 DNA probe (12).

RESULTS AND DISCUSSION
In S. cerevisiae, a cdc13-1 mutation causes yeast cells to arrest at the G 2 /M phase and accumulate single-stranded TG 1-3 DNA at telomeres at non-permissive temperatures (20). Analogous to what is caused by various types of DNA damage, the accumu- 2 The abbreviations used are: GST, glutathione S-transferase; IPTG, isopropyl-1-thio-␤-D-galactopyranoside; DTT, dithiothreitol; MBP, maltosebinding protein; RNP, ribonucleoprotein; RRM, RNA recognition motif. Purified GST (0.6 g) and GST-Tcg1 (1 g) were analyzed on a 10% SDS-polyacrylamide gel, and the Coomassie Blue-stained gel is shown. Arrows indicate the positions of purified proteins (left). Purified GST or GST-Tcg1 (200 nM) was mixed with 5 nM 32 P-labeled TG43 in 15-l reaction mixtures. The reactions were incubated at room temperature for 10 min and then subjected to gel shift assay. Lane 1 has no extracts. An autoradiogram is shown on the right. Position of the protein-DNA complex is indicated by an arrow. d, TCG1 complements cdc13-1 mutation. Yeast 4053-5-2a (CDC13), W13a (cdc13-1), or the W13a-carrying plasmid pTCG1or pDB20 (vector) was spotted in 10-fold serial dilutions on YC plates and grown at 25°C (left), 30°C (middle), or 33°C (right) until colonies formed. lated single-stranded telomeric DNA may serve as the signal for cell cycle arrest in cdc13-1 cells (20,29). In this scenario, it is conceivable that single-stranded telomeric binding proteins in excess amount could mask the single-stranded DNA present in cdc13-1 cells at the non-permissive temperature and prevent the G 2 /M arrest. To test this possibility, genes encoding singlestranded TG 1-3 DNA binding proteins were expressed in cdc13-1 cells to see if they rescue the temperature-sensitive phenotype (Fig. 1). It has been reported that S. cerevisiae Gbp2p binds specifically to single-stranded TG 1-3 DNA in vitro (12,13). As shown in Fig. 1a, cdc13-1 cells could not grow at 30°C, whereas the same cells expressing GBP2 grew to colonies. Expression of GBP2 cannot complement the cdc13-1 phenotype at 37°C and was not sufficient to complement the cdc13⌬ mutation (data not shown). Another single-stranded TG 1-3 DNA-binding protein, the product of S. pombe TCG1 (thirteen complementation gene, GenBank TM accession number AY007252), also complements cdc13-1. Similar to Gbp2p, TCG1 encodes a protein with two RRM motifs that binds single-stranded TG 1-3 DNA (Fig. 1, b and  c). Plasmid harboring TCG1 rescued the growth defect of cdc13-1 at 30°C but could not complement cdc13-1 at 33°C (Fig. 1d). It should be noted that the above complementation is specific to single-stranded telomeric DNA-binding proteins because other RRM-containing proteins that do not bind single-stranded telomeric DNA, including Ssb1p, Pub1p, or Hrb1p, did not complement the growth defect of cdc13-1 at 30°C (Fig. 1a, and data not shown).
Gbp2p contains two RRM motifs. The RRM motif, an ϳ90amino acid module, is used for RNA binding in many RNAbinding proteins (33). The signature of the RRM motif is the two consensus sequences, RNP1 and RNP2, located about 30 amino acids apart in the RRM domain (Fig. 2a). These two conserved regions are involved in making contacts with RNA. In U1 small nuclear RNP, the Arg residue in RNP1 forms a salt bridge with the phosphate of U1 RNA, whereas residues in RNP2 form hydrogen bonds with the RNA (34). We changed both Arg-161 and Arg-259, the conserved Arg residues within the two RRM motifs of Gbp2p, to Ala (Fig. 2a), and tested the effect on the singlestranded TG 1-3 binding activity in a gel mobility shift assay using purified mutant proteins (Fig. 2, b and c). The results indicate that the R259A mutation greatly reduced the binding of Gbp2p to single-stranded TG 1-3 DNA, whereas the R161A mutation only partially inactivated Gbp2p. Mutation on a non-  6) were expressed as MBP fusion proteins in E. coli and purified using amyloseagarose resins. 2 g each of the purified protein was analyzed on a 10% SDS-polyacrylamide gel, and the Coomassie blue-stained gel is shown. Positions of the alterations on Gbp2p are indicated above the lanes. c, the single-stranded TG 1-3 DNA binding activity of Gbp2p and its mutants. ϳ5 nM each of 32 P-labeled TG43 was mixed with several concentrations of the purified MBP, wild-type Gbp2p, Gbp2p R161A , Gbp2p R259A , or Gbp2p R161A, R259A , and the gel shift assay was then carried out. Proteins used in each set of experiments were 300, 100, and 33 nM. An autoradiogram is shown here. Fig. 1, yeast W13a (cdc13-1) carrying plasmids pSH380, pGAL-GBP2, pGAL-GBP2 R161A , pGAL-GBP2 R259A , or pGAL-GBP2 R161A, R259A was spotted in 10-fold serial dilutions on YC-Leu plates with the addition of 2% glucose (left) or 3% galactose (right) and grown at 25°C (top) or 30°C (bottom) until colonies formed (a). Yeast strains 4053-5-2a (CDC13) carrying pSH380 (vector) or W13a (cdc13-1) carrying plasmids pSH380, pGAL-GBP2, pGAL-GBP2 R161A , pGAL-GBP2 R259A , or pGAL-GBP2 R161A, R259A were grown in the presence of 2% glucose at 25°C for 3 days. Cells were then washed, resuspended in YC-Leu medium with 3% galactose and continued to grow at 30°C for another 6 h. Total yeast DNA was then isolated from these cells, digested with XhoI, run in a 1% agarose gel, and analyzed by Southern blotting under denatured (b) or native (c) conditions. The blots were hybridized with a 32 P-labeled C 1-3 A probe. Arrow indicates the YЈ-bearing telomeres. conserved residue, Arg-127, did not affect the binding activity (data not shown). These results indicate that Arg-161 and Arg-259 are important for binding to a single-stranded TG 1-3 DNA.

FIG. 3. The single-stranded TG 1-3 binding activity of Gbp2p is required to complement the cdc13-1 mutant. As in
Each Gbp2p binding mutant was tested for efficiency in complementing the growth defect of cdc13-1 cells. The experiments were conducted in the presence of glucose or galactose, because GBP2 expression was driven by the GAL1 promoter. As shown in Fig 3a, whereas wild-type GBP2 complemented cdc13-1 even under the non-inducing condition (ϪLeu ϩ Glc, 30°C) and gbp2 R161A complemented only under the inducing conduction (ϪLeu ϩ Gal, 30°C), neither gbp2 R259A nor gbp2 R161A, R259A was able to complement cdc13-1 at 30°C. These results suggested that complementation depends on the singlestranded telomeric DNA binding activity.
We investigated the effect of mutant or wild-type Gbp2p on the cdc13-1-associated telomere-lengthening phenotype (35). In S. cerevisiae, middle repetitive sequences known as YЈ elements are found in the subtelomeric regions of most chromosomes. As shown in Fig. 3b, in CDC13/pSH380 cells a XhoI digest produces a ϳ1.3 kbp fragment from the ends of YЈbearing chromosomes that contains ϳ950 bp of YЈ and the terminal tract of ϳ350 bp of TG 1-3 /C 1-3 A DNA. However, the length of YЈ-bearing telomeres in cdc13-1 cells appeared to be longer and more heterogeneous. Expression of the wild-type GBP2 or gbp2 R161A , but not gbp2 R259A or gbp2 R161A, R259A , suppressed telomere lengthening in cdc13-1 cells. Expressing Gbp2p in wild-type cells or cells with gbp2⌬ or gbp2⌬ cdc13-1 mutations did not affect the telomere length (data not shown). The effect of expressing Gbp2p on the accumulation of singlestranded telomeric DNA in cdc13-1 cells at 30°C was also evaluated. The cdc13-1 cells, but not the wild-type cells, accumulate abnormally high levels of single-stranded telomeric DNA at 30°C (Fig. 3c). By taking the single-stranded TG 1-3 DNA level in cdc13-1 cells at 30°C as 100%, the singlestranded TG 1-3 DNA level in wild-type CDC13 cells, cdc13-1 cells overexpressing wild-type GBP2, R161A, R259A, or R161A/ R259A mutants is 26%, 45%, 50%, 102%, and 91%, respectively. Thus, expressing the wild-type Gbp2p or R161A mutant moderately decreased the levels of single-stranded telomeric DNA, whereas R259A or R161A/R259A mutants of Gbp2p did not have this effect. These results demonstrate that the singlestranded TG 1-3 DNA binding activity of Gbp2p is required for suppressing the telomere lengthening of cdc13-1 cells and affecting the accumulation of single-stranded telomeric tails.
The presence of single-stranded DNA has been speculated as being the signal for the DNA-damage checkpoint (29) because it is present at the replication fork when replication is incomplete. Moreover, single-stranded DNA is a common intermediate in many repair pathways, including double-strand break repair, excision repair, and recombination repair (36 -38). Here we have shown that expressing two single-stranded TG 1-3 binding proteins, S. cerevisiae Gbp2p and S. pombe Tcg1, complemented the defects of cdc13-1 cells. In addition, the singlestranded TG 1-3 DNA binding activity of Gbp2p is required for this complementation. Because the cell cycle defect in cdc13-1 is dependent on the DNA-damage checkpoint, our results strongly suggest that the presence of single-stranded TG 1-3 DNA in cdc13-1 at a non-permissive temperature sends the signal for the DNA damage checkpoint. In wild-type cells, the binding of Cdc13p to the telomeric tail may convey a message to sensor proteins of the DNA-damage checkpoint, thus marking telomeres different from broken DNA ends. In cdc13-1 cells, masking the arrest signal with other single-stranded TG 1-3 DNA binding activity is necessary at non-permissive temperatures to circumvent the chromosome-surveillance mechanism and prevent cell cycle arrest.
A group of RRM-containing proteins that bind to singlestranded telomeric DNA has been identified in several organisms. For examples, human hnRNP A1, A2/B1, D, and E proteins and mouse hnRNP A2/B1 bind specifically to G-strand telomeric DNA (14 -16). In S. cerevisiae, Gbp2p and Nsr1p each have two RRMs and bind specifically to single-stranded TG 1-3 DNA in vitro (12,13) These RRM-containing proteins were shown to affect different telomere functions. For instance, mouse cells deficient in hnRNP A1 have telomeres shorter than those of cells expressing a normal level of hnRNP A1, suggesting a role for hnRNP A1 in telomere maintenance (39). Therefore, these RRMs containing telomeric DNA-binding proteins might function both in RNA metabolism and at telomeres. However, how these proteins affect telomeres remains unclear (16). It is conceivable that these proteins exert their influence through interaction with telomere-associated proteins. For example, the association between hnRNP A1 or D proteins and the telomerase might affect telomere maintenance (39,40).