Specific Binding of Single-stranded Telomeric DNA by Cdc13p ofSaccharomyces cerevisiae *

Cdc13p is a single strand telomere-binding protein of Saccharomyces cerevisiae; its telomere-binding region is within amino acids 451–693, Cdc13(451–693)p. In this study, we used purified Cdc13p and Cdc13(451–693)p to characterize their telomere binding activity. We found that the binding specificity of single-stranded TG1–3 DNA by these two proteins is similar. However, the affinity of Cdc13(451–693)p to DNA was slightly lower than that of Cdc13p. The binding of telomeric DNA by these two proteins was disrupted at NaCl concentrations higher than 0.3m, indicating that electrostatic interaction contributed significantly to the binding process. Because both proteins bound to strand TG1–3 DNA positioned at the 3′ end, the 5′ end, or in the middle of the oligonucleotide substrates, our results indicated that the location of TG1–3 in single-stranded DNA does not appear to be important for Cdc13p binding. Moreover, using DNase I footprint analysis, the structure of the telomeric DNA complexes of Cdc13p and Cdc13(451–693)p was analyzed. The DNase I footprints of these two proteins to three different telomeric DNA substrates were virtually identical, indicating that the telomere contact region of Cdc13p is within Cdc13(451–693)p. Together, the binding properties of Cdc13p and its binding domain support the theory that the specific binding of Cdc13p to telomeres is an important feature of telomeres that regulate telomerase access and/or differentiate natural telomeres from broken ends.

Single-stranded guanosine-rich DNA tail is a common structural feature in most of the eukaryotic telomeres (1)(2)(3)(4)(5)(6)(7). For example, ciliated protozoa telomeres are extended to form a 12-16-base single-stranded G-tail (1,3). In Saccharomyces cerevisiae, transient single-stranded TG 1-3 tails with lengths larger than 30 bases are detected late in the S phase (4). This presence of single-stranded G-tails was postulated as an intermediate for telomere replication (8). Single-stranded G-tails could form a DNA quadriplex in vitro known as the G-quartet (9,10), although it remains to be determined whether such a structure indeed exists in cells.
Protein factors that bind to the single-stranded telomeric DNA have been identified in several organisms (11)(12)(13)(14)(15)(16)(17)(18)(19). Among these protein factors, Oxytricha telomere-binding protein has been well characterized. It is heterodimeric and is composed of an ␣ and ␤ subunit (20 -22). The ␣ subunit is a single-stranded DNA-binding protein that binds to the G 4 T 4 single-stranded end of a telomere. Although the ␤ subunit is not directly involved in binding, it is required for making the terminusspecific binding. Cdc13p is a single-stranded TG 1-3 -binding protein that interacts with telomeres in S. cerevisiae (13,14,23). However, although Cdc13p in yeast is the functional equivalent of Oxytricha ␣and ␤-binding proteins, it shares no sequence similarity with Oxytricha telomere-binding proteins (13,14,23).
The binding of Cdc13p to telomeric DNA is essential for its function in telomeres (24), and it appears to have multiple functions in cells. For example, Cdc13p is involved in cell cycle control since a temperature-sensitive allele of CDC13, cdc13-1, causes cell cycles to arrest in the G2/M phase at non-permissive temperatures (25). The binding of Cdc13p to telomeres might cause yeast cells to differentiate whether the ends of linear DNA are telomeres or broken ends (26). In addition, Cdc13p appears to be a key factor in telomere replication. It interacts with Est1p that is associated with telomerase RNA (27)(28)(29)(30) to recruit telomerase to telomeres for replication. This was evidenced in part by the presence of a mutant allele of CDC13, cdc13 est , which causes a gradual loss of the telomere (13). Moreover, Cdc13p could interact with the catalytic subunit of DNA polymerase ␣, suggesting that it might be involved in a C-strand synthesis of telomeres (28). 1 It might also prevent end-to-end fusion of chromosomes and protect chromosome from degradation by nucleases.
CDC13 is an essential gene that encodes a 924-amino acid protein with a molecular mass of 104,895 Da (25). The Cdc13p fragment ranging from amino acids 451 to 693, Cdc13(451-693)p, contains the telomere-binding region of Cdc13p; it is sufficient to bind single-stranded telomeric DNA in vitro and interacts with telomeres in vivo (24,31). However, a sequence comparison among Cdc13(451-693)p and known DNA-or RNA-binding proteins did not provide any information on Cdc13(451-693)p responsible for binding to telomeres (24). Thus, Cdc13(451-693)p contains a novel motif for telomere binding. To understand how Cdc13p interacts with telomeres, we used purified Cdc13p and Cdc13(451-693)p to analyze their binding properties. The results of both EMSA 2 and DNase I footprint analysis revealed the specific binding of telomeric DNA by these two proteins.

MATERIALS AND METHODS
Expression and Purification of 6xHis-tagged Cdc13p and Cdc13(451-693)p-A baculovirus system was used to purify Cdc13p. An insect cell line sf21 was used as the host for virus propagation and protein purification. Escherichia coli DH5␣ was used as a host for plasmid construc-* This work was supported by NSC Grants 88-2314-010-070, 89-2314-B-010-008, and in part by Grant 89-B-FA22-2-4 (Program for Promoting Academic Excellence of Universities). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
To purify 6xHis-tagged Cdc13p, ϳ5 ϫ 10 7 sf21 cells were infected with recombinant virus for 4 days. Cells were washed with phosphatebuffered saline and then lysed by the addition of Nonidet P-40 lysis buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 200 M phenylmethylsulfonyl fluoride). The suspensions were incubated on ice for 40 min and then sonicated. Total cell extracts were then collected by centrifugation. Ni-NTA-agarose (Qiagen) was used to purify the 6xHistagged Cdc13p. Batch purification protocol was used according to the manufacturer's recommendations. The bound 6xHis-tagged Cdc13p was eluted by buffer containing 50 mM NaH 2 PO 4 , pH 8.0, 250 mM imidazole, 20% glycerol. Purified protein was aliquoted and frozen by a dry ice-ethanol bath. The yield of Cdc13p was ϳ0.5 mg/10 7 cells.
To purify 6xHis-tagged Cdc13(451-693)p, a 1-liter culture of E. coli harboring pET6H-CDC13(451-693) was grown at 25°C to A 600 to 0.5 and induced with the addition of 1 mM isopropyl-1-thio-␤-D-galactopyranoside. The cells were grown at 25°C for another 4 h before harvesting by centrifugation. Cells were resuspended in 10 ml of sonication buffer (50 mM NaH 2 PO 4 , pH 7.8, 300 mM NaCl, 5 mM ␤-mercapto ethanol, 1 ϫ protease inhibitors (Calbiochem)) and sonicated to release the cell contents. The sonicated cells were centrifuged at 13,000 g for 15 min at 4°C to obtain total cell free extracts. 0.5 ml of Ni-NTA-agarose (Qiagen) 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, 250 mM imidazole, 20% glycerol. Purified protein was aliquoted and frozen by the dry ice-ethanol bath. The yield of Cdc13(451-693)p was ϳ4.2 mg from 1 liter of E. coli culture.
Electrophoretic Mobility Shift Assay (EMSA)-Oligonucleotides (Table I) were 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 (12), Cdc13p or Cdc13(451-693)p in Buffer A (50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 50 mM NaCl, 1 mM dithiothreitol, 1 g of heat-denatured poly[dI-dC]) was mixed with 5 nM 32 P-labeled TG15 DNA with a total volume of 15 l. The reaction mixture was incubated at room temperature for 10 min. Then, the mixtures were loaded directly onto an 8% nondenaturing polyacrylamide gel, which was prerun 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, autoradiographed, and the amounts of oligonucleotides bound to the proteins were quantified with a PhosphorImager (Molecular Dynamics).
The apparent binding constant of Cdc13p and Cdc13(451-693)p to telomeric DNA was determined using EMSA and quantified by a Phos-phorImager. The DNA substrates used were at 5 nM in all experiments. Values presented in Table II were determined from interpolation on a Hill plot. Each value was the average of 2-3 experiments.
DNase I Footprint-Telomeric DNA strands were 5Ј-labeled and purified as described above. The DNA was mixed with Cdc13p or Cdc13(451-693)p in 40 l of Buffer B (40 mM Tris-HCl, pH 8.0, 10 mM MgSO 4 , 1 mM CaCl 2 ) and incubated at 25°C for 10 min. 1 unit of DNase I was added and incubated at 37°C for another 10 min. The reaction was stopped with 10 l of 250 mM EGTA. The DNA was then precipitated by adding 1 l of 10 mg/ml oyster glycogen and 150 l of ethanol. The precipitant was collected by centrifugation, dried, and analyzed by electrophoresis using a 12% polyacrylamide sequencing gel.

Isolation of Recombinant Cdc13p and Its Specific Binding to
Single-stranded TG 1-3 DNA-The baculovirus expression system was used to purify Cdc13p. In this study, Cdc13p with six histidines tagged at the N terminus was expressed in insect sf21 cells, and the protein was purified to homogeneity using a Ni-NTA-agarose column (Fig. 1, lane 2). The purified protein has an apparent mass of 105 kDa, which is in agreement with the predicted size of Cdc13p. Western blotting analysis further confirmed that it is Cdc13p (data not shown).
The binding specificity of Cdc13p to telomeric DNA was determined using EMSA analysis. Purified protein was mixed with 32 P-labeled single-stranded TG 1-3 with various amounts of unlabeled nucleic acid competitors before subjecting to the EMSA analysis. As shown in Fig. 2, unlabeled TG15 competed efficiently with 32 P-labeled TG15. The binding was reduced by ϳ50% when the competitor was presented at equal concentrations ( Fig. 2A, lanes 3-6, and Fig. 2B). Vertebrate (T 2 AG 3 ) telomeric DNA also competed for Cdc13p binding ( Fig. 2A,  lanes 7-9), although a 10-fold molar excess of (T 2 AG 3 ) 3 was needed to obtain the same level of competition as that from TG15 (Fig. 2B). In contrast, Oxytricha (T 4 G 4 ) and Tetrahymena (T 2 G 4 ) telomeric DNA did not compete for the binding of TG15 to Cdc13p (Fig. 2A, lanes 7-12). Total yeast RNA, singlestranded C 1-3 A DNA, or duplex TG 1-3 /C 1-3 A DNA did not compete for Cdc13p binding (Fig. 3A, lanes 16 -18) (data not shown). Interestingly, Cdc13p formed two complexes with TG15, although the nature of these multiple complexes is unclear. Nevertheless, these results indicated that Cdc13p bound specifically to single-stranded TG 1-3 telomeric DNA.
Binding of Cdc13p and Cdc13(451-693)p to Long Telomeric DNA-To determine whether Cdc13p and Cdc13(451-693)p could bind to long single-stranded TG 1-3 telomeric DNA, Cdc13p and Cdc13(451-693)p were mixed with telomeric DNA substrates with different lengths, and the complexes were analyzed. The results shown in Fig. 4 demonstrated that both Cdc13p and Cdc13(451-693)p were capable of forming complexes with long telomeric DNA. Multiple protein-DNA com-plexes were apparent in long DNA substrates, and the patterns of the complexes suggested that more than one protein could bind to a single DNA molecule. We also determined the binding affinity of Cdc13p and Cdc13(451-693)p to these DNA substrates. An apparent binding constant was determined from the Hill plot. As shown in Table II, Cdc13(451-693)p appears to bind to telomeric DNA with an affinity similar to that of Cdc13p. Cdc13p required telomeric sequences longer than 13 bases for proper binding, whereas Cdc13(451-693)p required 15 bases for proper binding.
Telomeric DNA Binding Properties of Cdc13p and Cdc13(451-693)p-The telomeric DNA sequences of S. cerevisiae are combinations of TG, TGG, and TGGG repeats. To test if Cdc13p and Cdc13(451-693)p have a sequence preference, the binding affinities of these two proteins for (TG) 12 , (TGG) 8 , or (TGGG) 6 were determined. As shown in Table II, the apparent binding constants of Cdc13p to these three substrates were similar to other telomeric DNA substrates. This result sug-
Oxytricha telomere-binding proteins were shown to bind to T 4 G 4 telomeric DNA at a high concentration of salt (22, 32, 33). We then investigated whether the affinity of Cdc13p or Cdc13(451-693)p for TG15 may depend on NaCl concentrations. As shown in Fig. 5, both proteins dissociated from TG15 at a NaCl concentration higher than 0.3 M. However, Cdc13(451-693)p appeared to tolerate NaCl better than Cdc13p.
To evaluate the stability of the protein-DNA complex, the dissociation rate of protein-DNA complex was measured. Cdc13p or Cdc13(451-693)p was first bound to 32 P-labeled TG15, and then an excess amount of unlabeled TG15 was added to prevent a re-association of protein to labeled DNA. The dissociation rate was estimated as the time required for half of the protein-DNA complex to dissociate. As shown in Fig.  6, both Cdc13p and Cdc13(451-693)p had dissociation rates ϳ30 min, suggesting that these two proteins bound to telomeric DNA with similar stability.
Cdc13p Does Not Require a 3Ј end for Binding-To determine if single-stranded TG 1-3 has to locate at the 3Ј end for Cdc13p binding, we synthesized oligonucleotides, 5Ј-Tel, Int-Tel, and 3Ј-Tel (Table I), of identical size harboring telomeric sequences at various locations. They were tested for Cdc13p binding by EMSA. Because all three substrates bound to Cdc13p or Cdc13(451-693)p to the same extent (Fig. 7), the location of TG 1-3 in a single-stranded DNA does not appear significant for Cdc13p binding.
In Fig. 7, the three oligonucleotides migrated differently on a polyacrylamide gel even though they have the same size and were heat-denatured and quick-cooled on ice before the binding assay. This result suggested an intrinsic position-dependent structure of these oligonucleotides. Also, although Int-Tel DNA gave the slowest mobility on the polyacrylamide gel, it had a slightly faster mobility upon binding to Cdc13(451-693)p. This migration behavior suggested that Cdc13(451-693)p bound to  5. Both Cdc13p and Cdc13(451-693)p prefer low salt for binding. 5 nM 32 P-labeled TG15 were incubated with 50 nM of Cdc13p or Cdc13 (451-693)p and 0, 0.05, 0.1, 0.15, 0.2, 0.3, or 0.5 M NaCl at room temperature for 10 min. 1 M unlabeled TG15 was added to the reaction mixtures, and gel shift assay then was performed. The binding was quantified by a PhosphorImager, and the activity without NaCl was taken as 100%.
FIG. 6. Slow dissociation of Cdc13p and Cdc13(451-693)p from single-stranded TG 1-3 DNA. 15 nM 32 P-labeled TG15 were incubated with 50 nM Cdc13p or Cdc13(451-693)p at room temperature for 10 min. 1 M unlabeled TG15 was then added (t ϭ 0). Aliquots of the mixtures were withdrawn at indicated time points and were loaded onto a running gel. After electrophoresis, the amount of 32 P-labeled TG15 that remained bound was quantified by a PhosphorImager, and the binding at t ϭ 0 was taken as 100%.  Figs. 8 and 9, respectively, and the results are summarized in Fig. 10. Both Cdc13p and Cdc13(451-693)p selectively protected telomeric sequences from DNase I digestion. The footprints of these two proteins with various DNA were nearly identical, suggesting that amino acid 451-693 of Cdc13p was the region that made contact with telomeres. It is interesting to note that the binding of Cdc13p produced several DNase I hypersensitive sites right next to the protection area, although the nature of these hypersensitive sites is unclear.

DISCUSSION
Cdc13p is an essential sequence-specific DNA-binding protein involved in a wide range of telomere functions including telomere length maintenance (13,27,28), telomere position effect (14), and cell cycle regulation (25). Therefore, it is important to elucidate the interaction of this protein with telomeric DNA. In this study, we purified Cdc13p in native form and its telomere-binding domain, Cdc13(451-693)p. Using these purified proteins, the telomeric DNA binding properties were characterized. Both Cdc13p and Cdc13(451-693)p bound to telomeric DNA with similar specificity and stability. Furthermore, DNase I footprints showed virtually identical binding of telomeric DNA by both Cdc13p and Cdc13(451-693)p, indicating that amino acid 451-693 is the region within Cdc13p that contacts telomere DNA.
A dissociation constant of ϳ10 Ϫ7 M is corresponding to ϳ10 kcal/mol of binding energy. Binding energies of this magnitude are certainly common among macromolecule-macromolecule interactions. We proposed that the binding of Cdc13p to telomeric DNA was the result of at least two factors, specific bond formation to nucleic acid bases and nonspecific electrostatic interaction. The binding to yeast telomeric DNA was not re-duced by competition with telomeric DNA from other species, suggesting that the binding is sequence-specific, and specific interaction between Cdc13p and the bases of nucleic acids should be critical for telomere binding. Moreover, because the binding is sensitive to ionic strength, electrostatic interaction must contribute significantly to the binding. This property is significantly different from that of Oxytricha telomere-binding protein, which used a series of aromatic amino acid residues to interact with the extended bases of single-stranded T 4 G 4 DNA (34). In Oxytricha, the phosphodiester-sugar backbone that contributes to electrostatic interaction was largely solvent-exposed. It could be worthwhile in future work to gain more precise information on this protein-DNA interaction.
In Fig. 7, EMSA analysis implicated a structure alteration of telomeric DNA induced by Cdc13(451-693)p. Similarly, DNase I footprint analysis also supports the presence of protein-induced alteration on DNA (Figs. 8 and 9). DNase I hypersensitive sites are positions of DNA that are easily digested by DNase I. These positions usually represent DNA with aberrant structure. For example, DNA bends on double-stranded DNA usually cause hypersensitive sites by DNase I digestion. Although we do not know the exact nature of the structural change in the single-stranded telomeric DNA induced by Cdc13p, it appeared that Cdc13p indeed caused structural alteration in telomeric DNA. Such alteration caused by Cdc13p may be an important feature of telomeres. Protein-induced DNA distortion at specific sites is considered to be an important mechanism for promoting the multiprotein interactions involved in regulation of gene activity (35,36), initiation of replication (37)(38)(39), site-specific recombination (40,41), or recognition of DNA damage (42). Cdc13p was shown to interact with Stn1p (43), the catalytic subunit of polymerase ␣, and Est1p (28). Moreover, Cdc13p was proposed to mediate telomerase access to telomeres (27). Thus, the alteration of telomeric DNA by Cdc13p might facilitate the formation of a multiprotein complex on telomeres. Alternatively, the single-stranded telomeric tail was shown to loop back to the double-stranded region of telomeres to form a "t-loop" structure in mammalian telomeres (44). In yeast, telomeres are folded back to form a looped structure (45,46). By cooperating with the doublestranded telomeric DNA bends induced by Rap1p (47,48), the alteration of single-stranded telomeric DNA by Cdc13p might facilitate the formation of such a structure in yeast.