The p23 molecular chaperone promotes functional telomerase complexes through DNA dissociation

Abstract

Telomeres are the composite of short DNA element tandem arrays and heterotypic protein components that protect and maintain chromosomal termini. As proper telomere maintenance requires a multitude of DNA extension events, it is important to understand the factors that modulate telomerase DNA association. Here, we show that the endogenous levels of the yeast p23 molecular chaperone Sba1p are required for telomere length maintenance and that Sba1p can modulate telomerase DNA binding and extension activities in vitro. Notably, telomere occupancy by telomerase and the extension rate of a shortened telomere fluctuated with changing Sba1 protein levels in vivo. In addition, we found that Sba1p displayed a cell cycle-dependent telomere interaction that paralleled telomerase binding; telomere association by Sba1p depended on its inherent chaperone activity. Taken together, our results support a model in which Sba1p modulates telomerase DNA binding activity for optimal function in vitro and in vivo.

Telomerase maintains genomic integrity, in part, by preserving chromosome length after DNA replication (1–3). Because conventional DNA polymerases require priming events to initiate synthesis, the extreme terminus of each lagging strand cannot be completed, which is commonly referred to as the end replication problem (4). In the absence of a compensatory process, this limitation would lead to chromosome erosion with each round of DNA replication. However, almost all eukaryotes circumvent this problem by adding simple DNA sequence motifs to each terminus that buffers against the loss. Typically the DNA motifs are added by the specialized ribonucleoprotein complex telomerase that is composed of a reverse transcriptase protein and an associated template RNA; in yeast, EST2 and TLC1 encode these factors. Depending on the organism, telomerase increases chromosome ends between a few hundred to a few thousand nucleotides to create, in part, a telomere. Perhaps unexpectedly, it was realized that the length of each telomere is not added at once but rather telomerase may append 6–8 nt per binding event (5). Although it had been argued that telomerase activity might not need to be iterative to maintain telomere length, recent studies in yeast indicate that telomeres can be extended over 100 nt per cell cycle (6, 7). Unfortunately, yeast telomerase does not add multiple repeats in vitro but rather extends a DNA substrate a single repeat and remains bound in a stalled state (5). Thus, the question is raised about how a bound telomerase complex is removed to allow the redundant binding and addition cycles necessary to support proper telomere maintenance.

Prior studies indicated that the p23 and Hsp90 molecular chaperones are required for reconstitution of human telomerase activity in vitro (8, 9); unlike other telomerase cofactors the molecular chaperones seemingly remain associated during reverse transcription (9). A recent study indicates that the Hsp90 interaction might be necessary to promote telomerase DNA binding activity (10); the functional consequence for the p23 interaction during DNA extension has not been revealed. However, studies have demonstrated that human p23 destabilizes DNA bound transcription factors in vitro and in vivo (11). Hence, we investigated whether the yeast p23 ortholog Sba1p serves to modulate DNA binding by telomerase.

Results

Sba1p Affects Telomerase DNA Extension Activity in Vitro.

We determined whether yeast telomerase was p23-dependent, as is the human enzyme in vitro (8, 9). We prepared telomerase-enriched cell fractions from WT and sba1Δ yeast by resolving whole-cell extracts sequentially over DEAE and MonoQ anion exchange resins as described in supporting information (SI) Fig. 7 and SI Text. Using WT and sba1Δ extracts we found that yeast telomerase, like human, relies on a p23 homolog for proper function in vitro; the sba1Δ extract activity was ≈20% relative to WT (Fig. 1, lanes 2 and 3). Importantly, we found that purified Sba1p was sufficient to affect function. Surprisingly, supplementation with full-length Sba1p showed a biphasic response; at first DNA extension levels increased to near WT but then declined (Fig. 1, lanes 4–9). Perhaps of note, optimal recovery of telomerase function was achieved with an Sba1p concentration (800 nM) comparable to physiological levels (1 μM; ref. 12).

Fig. 1.

Full-length Sba1p modulates telomerase DNA extension activity in vitro. Telomerase-dependent extension of an immobilized seven-base single-stranded 3′ overhang DNA substrate was examined. The DNA extension activities of WT, sba1Δ, or sba1Δ MonoQ fractions supplemented with varying levels of Sba1p or Sba1pΔ50 (0.1, 0.2, 0.4, 0.8, 1.6, or 3.2 μM) were tested as indicated. To serve as a loading control a polynucleotide kinase end-labeled 27-base oligonucleotide was added before the precipitation of all of the telomerase extension products. The position of +1 was determined by using terminal transferase-labeled DNA and α-ddATP (lanes 1 and 10).

We wanted to substantiate that the activity changes were caused by Sba1p and not through its association with Hsp90. Work on human p23 indicated that the extreme carboxyl terminus is required for in vitro chaperone activity (13), whereas Sba1p's amino terminus mediates Hsp90 interactions (14). Thus, we purified two Sba1 carboxyl-terminal truncations (Δ50 or Δ84). As expected, we found that the Sba1p carboxyl terminus was necessary for inherent chaperone activity but dispensable for Hsp82p (yeast Hsp90 homolog) interactions (see SI Fig. 8 and SI Text). In contrast to the full-length Sba1p, titration of either truncation protein had no apparent effect (Fig. 1, lanes 12–17, and data not shown). Taken together, the data indicate that Sba1p relies on its inherent chaperone activity to modulate telomerase DNA extension activity in vitro.

Sba1p Promotes Telomerase DNA Substrate Exchange in Vitro.

Although there are a number of possibilities for the biphasic response to Sba1p levels, we suggest two models: (i) Sba1p modulates telomerase's conformation, which initially promotes a function (e.g., nucleotide processivity) or (ii) Sba1p releases bound, inactive telomerase from DNA, thereby providing an opportunity to rebind in an active state (i.e., telomerase recycling); however, high Sba1p levels interfere with rebinding. A fundamental difference between the two models is that the first does not involve telomerase release from the DNA. To differentiate between the two models we performed a primer challenge experiment in which a second, longer substrate was added in increasing amounts after telomerase binding to a shorter telomeric DNA. The two paradigms will be distinguished depending on whether the second substrate is lengthened (i.e., extension of the longer substrate can only occur if Sba1p promotes telomerase release from the first primer).

As expected no extension of the longer (15-base) substrate was observed in the absence of Sba1p; however, upon addition of full-length Sba1p the longer DNA was extended (Fig. 2A). Furthermore, we observed increasing extension levels of the 15-base substrate as higher amounts of this DNA were added in the presence of Sba1p (Fig. 2A, lanes 13–19), which is consistent with an active telomerase turnover. Thus, the productive dissociation of telomerase from DNA depends on the level of a releasing factor (Sba1p) and/or availability of a suitable substrate (naïve unextended DNA).

Fig. 2.

Sba1p promotes telomerase DNA substrate exchange in vitro. (A) The ability of Sba1p or Sba1pΔ50 to promote telomerase transfer between DNA substrates was determined by challenging a bound 7-base substrate with increasing levels of a 15-base 3′ overhang telomeric DNA substrate. The position of +1 for the 7-base and 15-base substrates was determined by using terminal transferase end-labeled DNAs (lanes 1 and 2, respectively). As positive controls, extension reactions were performed by using sba1Δ MonoQ extract on the 7-base substrate, 15-base substrate, or the mixture of the 7- and 15-base substrates (2 pmol) (lanes 3, 4, and 5, respectively). The ability of either Sba1pΔ50 or Sba1p to promote telomerase transfer was examined by adding the Sba1 proteins (400 nM final) to prebound 7-base DNA (2 pmol, gray bar) followed by varying amounts of the 15-base telomeric substrate (2, 4, 6, 8, 10, 15, or 20 pmol). (B) Association of Sba1p and Hsp82p with DNA-bound telomerase was determined. WT MonoQ telomerase extract was incubated with immobilized DNA terminating with a G-rich single stranded seven-base 3′ overhang (G) in the absence or presence of RNase A or with DNA terminating with a C-rich sequence (C) as indicated. The presence of Sba1p or Hsp82p in the DNA bound (B) or free supernatant (F) fraction was detected by immunoblotting; as a control the total chaperones levels in the MonoQ extracts were determined (T).

Although it had been suggested that human p23 remains associated with active telomerase (9), based on the presented data an interaction between Sba1p and DNA-bound telomerase seems unlikely. In an attempt to resolve this point, we tested whether Sba1p remained bound to telomerase on the DNA and found a very low level of Sba1p associated with the G-rich substrate (Fig. 2B); the majority of Sba1p was found in the unbound, free fraction. In contrast to Sba1p, we observed a higher proportion of Hsp82p associated with the G-rich substrate (Fig. 2B). Hence, DNA binding by telomerase is largely incompatible with Sba1p interactions, whereas Hsp82p remains bound and likely serves a distinct function. The low level of Sba1p detected in the bound fraction may represent a transient intermediate that occurs before dissociation.

Sba1p Modulates Telomerase DNA Binding Activity in Vitro.

Our initial attempts to visualize RNA-dependent DNA binding with DEAE telomerase extracts were unsuccessful; the DEAE extracts contained significant binding activities, yet it was not appreciably altered upon RNase A inclusion (data not shown). However, after resolution of the DEAE extracts over MonoQ resin the protein fraction displayed an RNase A-sensitive binding activity that was selective for G-rich ssDNA by either fluorescence anisotropy or EMSA (Fig. 3A, SI Fig. 9A, and SI Text); a rise in the anisotropy value is indicative of complex formation as the increased mass results in a slower tumbling of the fluorescein-labeled telomeric primer. Notably, no binding activity was apparent when using extracts prepared from est2Δ or tlc1Δ yeast (Fig. 3A). Taken together, the two DNA binding assays in conjunction with MonoQ telomerase extracts provide a means to evaluate telomerase DNA binding activity.

Fig. 3.

Telomerase DNA binding is Sba1p-dependent in vitro. (A) Fluorescence anisotropy was used to detect DNA binding to a fluorescein-labeled telomeric oligonucleotide. The used MonoQ fractions were from WT (empty bars), sba1Δ (filled bars), est2Δ (striped bar), or tlc1Δ (hatched bar) yeast. The WT and sba1Δ reactions contain equivalent levels of TLC1 RNA and total protein; the tlc1Δ and est2Δ fractions contained equivalent total protein. (B) To measure the binding affinity between telomerase and a telomeric DNA WT (○) or sba1Δ (●) fractions were titrated into a reaction containing a fluorescein-labeled oligonucleotide (12.5 nM). (C) Purified recombinant full-length Sba1p or carboxyl-terminal truncations (Δ50 or Δ84) were titrated (0.1, 0.2, 0.4, 0.8, 1.6, or 3.2 μM) into anisotropy reactions containing sba1Δ MonoQ extract; for comparison, the binding activities of unsupplemented telomerase fractions with equivalent TLC1 and protein levels from WT (open bar) and sba1Δ (filled bar) yeast are shown.

Sba1p Chaperone Activity Is Required to Alter Telomerase DNA Binding.

Using the anisotropy assay we determined the telomerase DNA binding affinity in MonoQ extracts prepared from WT or sba1Δ yeast. Perhaps surprisingly, titration of WT or sba1Δ telomerase extracts with a fixed level of oligonucleotide revealed a enhanced binding affinity in the sba1Δ extracts (Fig. 3B). Based on the observed data telomerase displayed an ≈5.0-fold affinity increase (K_d 4.94 vs. 1.02 nM) in DNA binding in the absence of Sba1p. Given the decreased extension activity in the sba1Δ extract one might have expected the DNA binding activity to decrease. However, the observed increase in binding supports a model in which Sba1p destabilizes the telomerase/DNA structure.

Using the Sba1 protein derivatives we tested whether this chaperone was sufficient to influence DNA binding by telomerase and which Sba1p activities were required. We found that full-length Sba1p was necessary to reduce telomerase DNA binding activity, as neither carboxyl-terminal mutant had an effect (Fig. 3C). To affect telomerase DNA binding Sba1p alters the off-rate between telomerase and DNA as described in SI Fig. 9B and SI Text. Hence, Sba1p dissociates telomerase from DNA by using its inherent chaperone activity.

Sba1p Modulates Telomere Occupancy by Telomerase in Vivo.

To address whether Sba1p is able to alter DNA association in vivo we used the ChIP assay. We determined the relative telomerase occupancy at a single telomere of chromosome 15 (Chr XV) or monitored the average occupancy at a population of telomeres by using an oligonucleotide set specific to a subtelomeric Y′ element found at 11 chromosomal termini. To assess the Sba1p effect we performed ChIP assays using yeast strains expressing Myc-tagged Est2p with either endogenous levels of Sba1p (WT) or overexpressed Sba1p (WT+Sba1p), or in the absence of Sba1p (sba1Δ). Importantly, we found that telomerase association decreased to ≈30% of WT levels upon overexpression of Sba1p but was elevated ≈2-fold in the absence of Sba1p at either a single telomere or at the population (Fig. 4A). Hence, Sba1p cellular levels affect telomerase occupancy at telomeres.

Fig. 4.

In vivo levels of Sba1p affect telomere occupancy by telomerase. (A) Telomere association by telomerase was gauged by using the ChIP assay. The relative levels of telomerase-telomere binding were determined in the WT, WT overexpressing Sba1p (WT + Sba1p), and sba1Δ yeast. Telomerase residency at a population of telomeres was determined by using oligonucleotides selected for a subtelomeric Y′ element found at 11 chromosomal termini (Y′ elements) or at a single telomere using primers specific for a subtelomeric region of chromosome XV (Chr. XV); all values were normalized to the signal from an internal nontelomeric DNA. (B) The ability of the carboxyl-terminal Sba1p deletions to associate with telomeric DNA was addressed by using the ChIP assay and asynchronous sba1Δ yeast expressing the indicated proteins. All data represent average values (mean ± SD) from three independent assays.

In an attempt to understand the Sba1p cellular role with telomerase we examined whether the Sba1p telomere association was cell cycle-dependent and also determined the domain requirements for the interaction. Using α-factor synchronized yeast we found that Sba1p was telomere-associated throughout the cell cycle as shown in SI Fig. 10 and SI Text. Sba1p's ability to interact at telomeres depended on its inherent chaperone activity because neither Δ50 nor Δ84 showed an apparent interaction (Fig. 4B). Thus, an Sba1p telomere association depends on intrinsic chaperone activity and not on an Hsp82p interaction.

Sba1p Levels Affect Telomere Length.

Our in vitro telomerase DNA binding and extension results suggest that there is an optimal Sba1p level for telomerase actions. We tested whether changes to telomere length occur after disruption of the SBA1 gene or upon Sba1p overexpression. Although the telomere effect was mild in the first passage after SBA1 disruption, we observed a further decline during serial passages (Fig. 5A); no further shortening was observed after five passages (data not shown). Importantly, no change in the migration of a control, nontelomeric DNA was observed, indicating that the decline in telomere length was specific (Fig. 5A) (15). It might have been difficult to predict that SBA1 has a role in telomere maintenance (16); however, the phenotype likely was not revealed because it only becomes pronounced after multiple passages.

Fig. 5.

Telomere length is Sba1p-dependent. (A) Genomic DNA was isolated from WT or yeast passaged one, three, or five times after SBA1 disruption. The asterisk indicates the approximate position of the Y′ subtelomeric fragment. As a migration control, the position of a nontelomeric DNA (control) was visualized by using a probe specific to an internal fragment of chromosome IV (15). (B) The effect of Sba1p overexpression on telomere length was examined by Southern blot analysis in WT yeast constitutively expressing Sba1p (GPD promoter) for the indicated passages. The asterisk indicates the approximate position of the Y′ subtelomeric fragment.

We also investigated the effect of Sba1p overexpression and observed a decrease in telomere length (Fig. 5B). Thus, a correlation between our in vitro and in vivo data is apparent in that telomerase activity and telomere length depend on the relative Sba1p levels. We suggest that telomere telomerase dynamics require an optimal rate; too slow and shortening results as the necessary transitions do not efficiently occur and if the kinetics are too fast then telomerase cannot productively engage the telomere.

Sba1p Is Necessary for Proper Telomere Extension Rates.

The progressive telomere shortening in the sba1Δ strain indicates that Sba1p is required for telomere maintenance, which can involve a number of activities, including extension or protection of the chromosomal termini. In an attempt to directly address whether Sba1p is required for telomere lengthening in vivo, we exploited an established system to determine the extension rates of a shortened telomere in WT and sba1Δ yeast. In brief, the left arm terminus of chromosome VII contains an internal telomere tract that is flanked by Flp1p-recognition target (FRT) sites, the tract can be specifically removed upon the transient Flp1p expression and changes in length can be monitored during subsequent growth by Southern blot analysis (17, 18). Using this system we determined the telomere extension rates in WT and sba1Δ yeast and found that the sba1Δ rate was 50% of WT after 3 generations and remained at 65% after 6 and 10 generations of growth (Fig. 6). Hence, Sba1p loss correlates not only with shortened telomeres but also with a slower telomere extension rate. Taken together, our results indicate that Sba1p has a role in telomerase-mediated telomere extension by modulating its DNA binding activity.

Fig. 6.

Disruption of SBA1 alters telomere extension rates in vivo. The elongation rates of an Flp-FRT-directed shortened telomere were monitored in WT and sba1Δ yeast (17, 18). Briefly, genomic DNA was isolated from cells growing exponentially in raffinose media (P), DNA was recovered after the FRT-encompassed telomere tract was removed by shifting the cells to galactose media (0), yeast were switched to glucose media, and genomic DNA samples were recovered after 3, 6, and 10 generations as indicated. Representative Southern blot results are shown (A), and the quantified data from five independent experiments are presented (B). In A, the asterisk indicates the position of the telomeric restriction fragment of the left arm of chromosome VII, and the arrow marks the residual uninduced nontelomeric URA3–ADH4.

Discussion

Numerous studies have added to our knowledge of telomerase events, particularly recruitment to a telomere and extension of the DNA by reverse transcription (1–3). In addition to these actions, effective telomere maintenance requires telomerase release during the cell cycle to efficiently maintain chromosome length and genomic integrity (19, 20). It has been established that during G₁ of the cell cycle telomerase is telomere-associated, presumably to foster telomere capping; at late G₁/early S phase it transiently dissociates and in mid- to late-S phase it rebinds to permit rapid telomere extension (3). Given the intrinsic stable association telomerase displays with telomeric DNA in vitro, the involvement of an auxiliary factor(s) to release telomerase from a telomere is likely. Our data support a role for the Sba1p molecular chaperone in promoting telomerase dissociation.

We suggest a model of telomere extension in which the process is cyclical and comprised of three distinct phases. The first two phases, assembly and extension, involve recruitment of the telomerase enzyme (minimally, Est2p and TLC1 RNA), alignment of the RNA substrate, and extension of a chromosome terminus by the reverse transcriptase. If, by chance, telomerase engages the DNA in a nonproductive manner, dissociation by Sba1p would permit an opportunity to rebind until an active state is reached. After extension of the DNA the third phase (disassembly) would occur, which entails disruption of the telomere structure by minimally dissociating the core components Est2p and TLC1. Once the core complex has been removed, the DNA can be lengthened further by the same telomerase (processive extension) or a different telomerase complex. Although our data do not distinguish between these two modes of lengthening, the work does demonstrate that proper telomere maintenance likely relies on balanced dynamics in which the stability of a telomerase/DNA complex is modulated to permit efficient reverse transcription without impeding subsequent telomere extensions.

We showed that Sba1p could modulate telomerase activity both positively and negatively. Although there are several plausible models to account for these effects, we suggest that Sba1p minimally contributes to telomerase activity by dissociating stalled complexes that have yet to initiate DNA extension. Hence, the Sba1p-dependent increase in extension activity in the sba1Δ fraction results from telomerase recycling that permits “resampling” of the telomeric substrate until a suitable telomerase/DNA interaction is formed that is competent for extension. Presumably, any resampling in the WT telomerase fraction occurs rapidly upon initial DNA contacts so that telomerase effectively binds the DNA in an extension-competent state. In contrast, as the levels of Sba1p become excessive telomerase activity declines because its interaction with the telomeric DNA becomes too transient to allow proper engagement and extension. The biphasic response to Sba1p levels may be applicable to numerous proteins involved in genome maintenance, as a prior study indicated that Sba1p loss or overexpression enhances chromosome instability (21).

In addition to recycling bound but nonfunctional enzymes, the chaperone-mediated dissociation may be important to promote telomerase release after DNA extension; removal would be necessary to reposition existing or bind new telomerase. In addition to Sba1p, we speculate that other proteins may aid in telomerase release after telomere extension in vivo. It has recently been shown that the Pif1 DNA helicase can dissociate telomerase from telomeric DNA by disrupting the base-pairing between telomeric DNA and TLC1 RNA (22). Hence, it is conceivable that Sba1p and Pif1p work together to disengage Est2p/DNA and TLC1/DNA interactions in vivo. Nonetheless, it seems likely that proper telomere maintenance requires both an efficient means to turnover telomerase after DNA extension and a way to remove/reposition bound but nonfunctional telomerase complexes. In brief, we suggest that telomere protein associations are driven forward by high-affinity interactions between the less abundant telomere-binding proteins such as telomerase, Cdc13p, and Est1p and that these complexes are disengaged through lower-affinity interactions between key components (e.g., telomerase) and the abundant molecular chaperones. Thus, the molecular chaperones help create a dynamic telomere environment without interfering with the functional activities required to maintain chromosome stability and length.

In addition to telomerase, our work is applicable to our understanding of the p23 molecular chaperone and possibly our appreciation of the modular nature of the eukaryotic molecular chaperone network (23, 24). In general, molecular chaperones mediate fundamental cellular events from nascent polypeptide folding to regulation of “native” proteins. Based on intracellular expression levels the primary eukaryotic molecular chaperones are members of the Hsp90 and Hsp70 families. In addition to these chaperones, a plethora of Hsp90 and/or Hsp70 cofactors have been revealed. By and large, studies have addressed the impact of the associated components on the chaperone and ATPase activities of Hsp70 or Hsp90; yet, many of the constituents, including p23, Hsp104, HiP, Hdj2, Cdc37, and large immunophilins, have inherent molecular chaperone activities (23, 24). For instance, Sba1p-mediated dissociation of telomerase from telomeric DNA relies on Sba1p's chaperone activity rather than on Hsp82p. Nonetheless, there are possible points in which the p23–Hsp90 interactions might be necessary. For example, both chaperones are known telomerase-binding proteins that may promote the assembly of the telomerase RNA substrate and reverse transcriptase (8). As both chaperones contribute to the same endpoint, formation of the assembled ribonucleoprotein enzyme, a need for chaperone interactions may be necessary. It is also possible that an interaction between p23 and Hsp90 might be necessary to rapidly cycle telomerase on and off the DNA because p23 appears to have a role in DNA release and Hsp90 promotes binding (10). The differential use of p23 and Hsp90 with telomerase is perhaps analogous to their roles with intracellular hormone receptors. For example, both p23 and Hsp90 contribute to the hormone binding activity of the estrogen receptor in a manner that depends on the p23–Hsp90 interaction (25). Yet, p23 is able to dissociate a receptor from its response element independent of Hsp90 and Hsp90 appears to support DNA binding by a receptor (11, 26). Hence, in a manner analogous to hormone receptors, molecular chaperones might have multifaceted effects on telomerase depending on the composition or context of the chaperone complex.

Materials and Methods

Yeast Strains.

The Saccharomyces cerevisiae strains used in this work were YPH499 (MATa), EST2-MYC13x (MATa; est2::EST2–13xmyc-KAN), and LEV336 (MATa; adh4:: FRT-URA3-TEL270-TEL270-FRT-TEL). The EST2-MYC13x strain was a kind gift from E. H. Blackburn (University of California, San Francisco, CA). The LEV336 strain was a kind gift of E. Gilson (Ecole Normale Supérieure de Lyon, Lyon, France) and S. Marcand (Commissariat à l'Energie Atomique, Fontenay-aux-Roses, France). YBF100, YOT100 and YOT336 strains are SBA1 disruptions derived from YPH499, EST2-MYC13x, or LEV336, respectively, using standard homologous recombination procedures (27).

Telomerase Extract Preparation.

The telomerase extracts were prepared by using a modification of the established DEAE protocol (5) by sequentially resolving whole-cell extracts over DEAE and MonoQ anion exchange resins as described in SI Text.

Protein Purification.

The full-length Sba1p and carboxyl-terminal truncations (Δ50 or Δ84) were expressed in Rosetta (Novagen, San Diego, CA) bacteria using pET expression constructs (pET-SBA1, pET-sba1Δ50, or pET-sba1Δ84). The proteins were isolated in a manner analogous to the purification of human p23 (28).

Fluorescence Anisotropy Assay.

The anisotropy reactions were performed in TMG30 (20 mM Tris, pH 8.0/1.1 mM MgCl₂/0.1 mM EDTA/1.5 mM DTT/10% glycerol/0.1% Triton X-100/30 mM NaOAc, pH 7.0) supplemented with a fluorescein-labeled telomere oligonucleotide (fl-GTGTGGTGTGTGGG). The reactions were incubated 5 min at 25°C before determining the anisotropy values with an Ultra Evolution plate reader (Tecan, Zurich, Switzerland). RNase sensitivity of the binding was determined by preincubating aliquots of the MonoQ extracts with 10 μg of RNase A for 10 min at room temperature. The apparent dissociation constants (K_d) were determined by fitting a curve according to the equation: (((m1 + m2 + m0) − ((m1 + m2 + m0)∧ 2–4*m2*m0)∧0.5)/(2*m2))*m3, where m0 = (total TLC1 RNA); m1 = K_d; m2 = (total oligonucleotide); m3 = the maximal change in anisotropy, which is set for each experiment (29).

ChIP Assay.

ChIP analysis was performed essentially as described (30). To determine the relative telomere occupancy by telomerase transformants of the EST2-MYC13x (pRS425-GPD or pRS425-GPD-SBA1) or YOT100 (pRS425-GPD) were grown to an OD₅₉₅ = 1.0 before formaldehyde addition. We performed the cell-cycle experiments as described in SI Text. The relative level of select DNA was determined by quantitative PCR as per manufacturer's instructions (Bio-Rad, Hercules, CA) using primers specific for a Y′ element found within 11 subtelomeric regions or primers that amplify a subtelomeric region of chromosome XV (ACCACAGCGAACCACGATCCA and GGTGAGTATGGCATGTGGTGG). The fold enrichments were determined by normalizing the ratio of the specific/nonspecific signal [e.g., α-Myc ChIP threshold cycle (C_t values)/normal mouse IgG ChIP C_t value] at the telomeric DNA targets to the ratio of the specific/nonspecific signal produced by amplifying a region of the genome not targeted by Sba1p (YJL052W; TGCTGCTAAGGCTGTCGGTA and CAACGGCATCTTCGGTGTAA) as determined by using ChIP-on-CHIP (B.C.F., unpublished work).

Telomere Length Analysis.

Genomic DNA was prepared from WT (YPH499) and sba1Δ (YBF100) yeast after the indicated passages. The DNA was digested with XhoI, resolved, and transferred to Immobilon-Ny⁺ membrane (Millipore, Billerica, MA). The telomeric DNA was identified after hybridization in Ekono buffer (ISC BioExpress Inc., Kaysville, UT) supplemented with a radiolabeled telomeric oligonucleotide [(TGTGGGT)₄]. As a loading and migration control we used an established probe that recognizes a 1,621-bp fragment of chromosome IV (15). After appropriate washes using SSC/SDS buffers the telomeric and control DNA was visualized with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Telomere Extension Assay.

The telomere elongation assay was performed as described (17, 18). Briefly, WT (LEV336) or sbaΔ (YOT336) yeast growing exponentially in the raffinose (2%) media were harvested by centrifugation, and the cells were resuspended in a galactose (2%) media; before the switch to galactose an aliquot was removed (“preflip” sample). After a 3-h incubation at 30°C, a sample was removed for genomic DNA extraction (0 generation). The remaining culture was clarified by a brief centrifugation. The cells were resuspended in glucose containing media and propagated by dilution into fresh media every two to four doublings as determined by cell counting with a hemocytometer. Samples for genomic DNA isolation were collected after 3, 6, and 10 doublings. The genomic DNAs were digested with StuI, the products were resolved, and the DNA was transferred to Immobilon-Ny⁺ membrane. To determine the average size of each telomeric fragment we followed the established protocol by generating a standard curve from the migration distances of the complete set of coresolved molecular weight markers (Invitrogen, Carlsbad, CA) (17, 18). The presented data represent average values (mean ± SD) from five independent assays.

Abbreviation

FRT

Flp1p-recognition target.