Report

Est1 and Cdc13 as Comediators of Telomerase Access

See allHide authors and affiliations

Science  01 Oct 1999:
Vol. 286, Issue 5437, pp. 117-120
DOI: 10.1126/science.286.5437.117

Abstract

Cdc13 and Est1 are single-strand telomeric DNA binding proteins that contribute to telomere replication in the yeastSaccharomyces cerevisiae. Here it is shown that fusion of Cdc13 to the telomerase-associated Est1 protein results in greatly elongated telomeres. Fusion proteins consisting of mutant versions of Cdc13 or Est1 confer similar telomere elongation, indicating that close physical proximity can bypass telomerase-defective mutations in either protein. Fusing Cdc13 directly to the catalytic core of telomerase allows stable telomere maintenance in the absence of Est1, consistent with a role for Est1 in mediating telomerase access. Telomere length homeostasis therefore is maintained in part by restricting access of telomerase to chromosome termini, but this limiting situation can be overcome by directly tethering telomerase to the telomere.

In most species, telomeres are composed of G-rich repetitive sequences that are elongated by telomerase (1). Several factors govern the balance between sequence addition and loss to maintain telomeres at a stable length, including positive and negative regulation of telomerase access to the chromosome terminus (2–4). In S. cerevisiae, five genes are required for the telomerase pathway (4–7). TLC1 andEST2 encode the RNA and reverse transcriptase subunits of telomerase, respectively, and as expected for subunits that are essential for catalysis, telomerase activity is absent in extracts from strains defective in EST2 orTLC1(7–9). In contrast, mutations inEST1, EST3, and CDC13 do not eliminate enzyme activity in vitro (9, 10), despite the fact that strains carrying mutations in any of these three genes have the same severe telomere replication defect as est2-Δ ortlc1-Δ strains (6, 10).

Both Cdc13 and Est1 bind single-strand telomeric DNA (4,11, 12), although they make separate contributions to telomere replication and stability. Est1 is required solely for the telomerase pathway (11), whereas Cdc13 has an essential function at the telomere, presumably in protecting the end of the chromosome (13), as well as a role in telomere replication (4). This latter activity was revealed by a telomerase-defective allele of Cdc13, calledcdc13est [originally named est4(6)], leading to the proposal that Cdc13, like Est1, mediates telomerase access (4). The two proteins also display different biochemical properties. Est1, but not Cdc13, requires a free single-strand 3′ terminus for DNA binding and binds telomeric DNA with a 500-fold reduced affinity compared with Cdc13 (4, 11). In addition, Est1 is associated with telomerase, whereas Cdc13 does not exhibit a detectable interaction with the enzyme (14).

These results suggest that telomerase is recruited to the telomere due to a direct (but weak) protein-protein interaction between Cdc13 and the enzyme, and the telomere shortening in thecdc13est mutant strain is due to a further reduction in this interaction. This model predicts that increasing the association between Cdc13 and telomerase would increase telomere length. To test this, we examined the consequences of fusing Cdc13 to the telomerase-associated protein Est1 (15). Introduction of the gene encoding this Cdc13-Est1 fusion, present on a single-copy plasmid and expressed by the CDC13 promoter, into aCDC13 + strain resulted in substantial telomere elongation (Fig. 1A, lanes 2 and 3). This effect was dependent on functional telomerase, because telomeres were not elongated when the fusion was introduced into anest2-Δ strain (Fig. 1A, lanes 14 to 16). Telomere elongation was even more pronounced in a cdc13-Δ strain, in which telomere length increased by ∼800 base pairs after ∼100 generations of growth (Fig. 1A, lanes 4 and 5); continued propagation resulted in even further telomere lengthening (16).

Figure 1

The Cdc13-Est1 fusion confers telomere elongation. (A) Telomere Southern (DNA) blots were performed as in (6). The bracket indicates a heterogeneous telomeric band that represents about two-thirds of the telomeres in this strain background. Cultures for lanes 1 to 13 were grown for ∼100 generations before DNA preparation. Lanes 1, 13, and 14,CDC13 + EST + control strain; lanes 2 to 5, 10, 15, and 16, pVL1091 (expressing the Cdc13-Est1 fusion); lanes 6 to 9 and 11, pVL1092 (Cdc13est-Est1); lane 12, pVL1098 (Cdc13est-Est1-47); the relevant genotypes of the strains are indicated. Plasmids were introduced into strains deleted forCDC13 for this and subsequent figures by first transforming into a cdc13-Δ/pVL438 (CDC13 + URA3 +) strain followed by subsequent eviction of pVL438 by plating on media containing 5-fluoro-orotic acid. Molecular sizes are indicated on the left (in kilobase pairs). (B)cdc13-Δ strains containing pVL648 (CDC13 +), pVL1091, or pVL762 (cdc13ts ) (13) were grown at 23°C, and equivalent numbers of cells, as serial 10-fold dilutions, were plated at 23° and 36°C. (C) Growth after ∼50 generations of est1-Δ, est2−Δ , or est3-Δ strains, with single-copy plasmids bearing the CDC13-EST1 fusion gene or the appropriate WT ESTgene (each under their native promoter). (D) Immunoprecipitation from extracts prepared from strains expressing proteins with a triple hemagglutinin epitope (HA3) introduced at the NH2-terminus: HA3Est1 (pVL1106), HA3Cdc13 (pVL841), or HA3Cdc13-Est1 (pVL1102), followed by detection of the telomerase RNA (TLC1) levels in the extract (E) and immunoprecipitates (P) by Northern (RNA) blotting (27).

Several experiments indicated that both Cdc13 and Est1 retained function in the context of the fusion. First, the essential function ofCDC13 was fully complemented by the fusion protein (Fig. 1B). The Cdc13-Est1 fusion also complemented the senescence phenotype of an est1-Δ strain (Fig. 1C). This was not due to a general bypass of the telomerase pathway, because this same fusion did not complement est2-Δ or est3-Δ strains (Fig. 1C). The complementation behavior of the fusion protein indicates that Est1 and Cdc13 normally function in temporal and physical proximity in their respective roles in telomere replication. Telomerase also has been shown to coimmunoprecipitate with Est1 but not with Cdc13 (Fig. 1D) (14). The Cdc13-Est1 fusion protein coimmunoprecipitated the RNA subunit of the telomerase complex (Fig. 1D) and enzyme activity (17), indicating that telomerase is associated with the fusion protein.

These results suggest that the proposed recruitment function of Cdc13 can be enhanced by fusing it to a telomerase component, and predict that the telomerase-defective cdc13est mutation would be bypassed in a fusion. Consistent with this prediction, a fusion containing the mutant Cdc13est protein behaved indistinguishably from the wild-type (WT) fusion: Telomere elongation occurred to the same degree in both CDC13 + andcdc13-Δ strains (Fig. 1A, compare lanes 6 to 9 and 2 to 5), and no senescence was observed when the Cdc13est-Est1 fusion was introduced into a cdc13-Δ strain (17). Similar results were observed in a reciprocal experiment with a mutant allele of EST1(est1-47); this mutation disrupts telomere replication (Fig. 1C), although the mutant Est1 protein still physically associates with telomerase (18), suggesting a defect in the same telomerase-accessing function that is altered by thecdc13est allele. This defect was bypassed when the Est1-47 protein was fused to Cdc13. In fact, the double-mutant fusion protein (Cdc13est fused to Est1-47) conferred the same degree of telomere elongation as the WT fusion protein, even in a strain deleted for both est1-Δ and cdc13-Δ (Fig. 1A, lane 12); moreover, this strain did not exhibit senescence (17). The ability of the double-mutant fusion to complement a cdc13-Δ est1-Δ strain indicates that the fusion is acting as a dimeric molecule that bridges telomerase and the telomere.

One alternative interpretation of our data is that telomere elongation is due to perturbation of chromatin structure, rather than to increased access of telomerase to the telomere. In particular, telomere lengthening could be a secondary consequence of altered Cdc13 function, because mutations in CDC13 have been identified that increase telomere length (19, 20). However, these recessive alleles of CDC13 have a set of genetic and biochemical features that distinguish them from the gain-of-function properties of the Cdc13-Est1 fusion (17, 20). In addition, fusion of several unrelated protein sequences, or an inactive telomerase subunit (see below), to the COOH-terminus of either the WT Cdc13 protein or the mutant Cdc13est protein does not increase telomere length (17, 21). The most direct argument against this alternative interpretation is the result of an experiment in which we examined the behavior of a fusion in which only the high-affinity DNA binding domain of Cdc13 (DBDCdc13) was fused to Est1 (15). This experiment was based on our previous demonstration that DBDCdc13 can be expressed as a stable, functional subdomain (22) and therefore could be used as an alternative means of directing Est1 to the telomere with high efficiency, while leaving the full-length Cdc13 protein intact. As predicted, the Est1-DBDCdc13 fusion conferred extensive telomere lengthening in a CDC13 + strain (Fig. 2A) and bypassed senescence of acdc13est strain (Fig. 2B). Furthermore, fusion of the defective Est1-47 protein to DBDCdc13 bypassed bothest1-Δ and cdc13est mutations (Fig. 2B), even though the est1-47 allele fails to complement either mutation (Fig. 1C) (17). Expression of either Est1 or DBDCdc13 had no effect on telomere length or viability in CDC13 + orcdc13est strains, and the Est1-DBDCdc13 fusion failed to rescue the inviability of acdc13-Δ strain (Fig. 2A) (17). Thus, the telomere lengthening properties of these fusions are likely to be a consequence of delivery of telomerase to the telomere, rather than a perturbation of Cdc13 function.

Figure 2

The DNA binding domain of Cdc13 is sufficient to deliver telomerase to the telomere. (A) Telomere length after ∼100 generations of growth. Lanes 1 and 2, est1-Δ/pVL1091; lanes 3 and 4,est1-Δ/pVL1120 (expressing the Est1- DBDCdc13 fusion); lane 5,EST1 + control. (B) Growth after ∼50 generations of an est1-Δcdc3est strain with plasmids expressing the Est1 protein (pVL499), the Est1-DBDCdc13 fusion (pVL1120), or the Est1-47-DBDCdc13 fusion (pVL1121). Strains expressing the Est1-DBDCdc13 and the Est1-47-DBDCdc13 fusions have been propagated a further 75 generations with no signs of senescence (17).

We next fused Cdc13 directly to Est2, the catalytic subunit of telomerase (8). The Cdc13-Est2 fusion (15) resulted in telomere lengthening to levels comparable to that of the Cdc13-Est1 fusion (Fig. 3A, lanes 3 and 4). The fusion complemented cdc13-Δ andest2-Δ null mutations, and telomere elongation occurred to the same degree in est2-Δ and EST2 +strains (17). A Cdc13-Est2D670Afusion, containing an Asp to Ala mutation at position 670 in the active site of Est2 (8), did not confer extensive telomere elongation but instead maintained telomere length at WT levels in acdc13-Δ EST2 + strain (Fig. 3A, lanes 5 and 6) (23), showing that telomere elongation is only observed when a catalytically active version of telomerase is tethered to the telomere. Strikingly, the Cdc13-Est2 fusion allowed cell growth in the complete absence of Est1 function, because anest1-Δ strain carrying this fusion was viable for more than 250 generations (Fig. 3B) (24). Long-term propagation in the absence of Est1 was not due to a previously described alternative pathway that can maintain telomeres in the absence of telomerase function (6, 25): Telomeres in an est1-Δ strain carrying the Cdc13-Est2 fusion were stably maintained at a length slightly below that of WT telomere length (Fig. 3C), with none of the striking changes in telomere structure that characterize the alternative pathway (6, 25). The ability of the Cdc13-Est2 fusion to maintain an est1-Δ strain required tethering of a functional telomerase, because anest1-Δ strain carrying the Cdc13-Est2D670Afusion exhibited senescence (Fig. 3B). This supports the hypothesis that a critical function of the Est1 protein is to mediate access of telomerase to the telomere. Notably, neither the Cdc13-Est1 fusion nor the Cdc13-Est2 fusion bypassed the requirement for Est3 (Fig. 1C) (17), showing that Est1 and Est3 perform functionally distinct roles in telomere replication.

Figure 3

Fusing Cdc13 to the catalytic subunit of telomerase bypasses the requirement for Est1. (A) Telomere length, after ∼75 generations of growth. Lanes 1 and 2,cdc13-Δ/pVL1091; lanes 3 and 4,cdc13-Δ/pVL1107 (expressing the Cdc13-Est2 fusion); lanes 5 and 6, cdc13-Δ/pVL1111 (Cdc13-Est2D670); lane 7, CDC13 + control strain. (B) Growth of an est1−Δ cdc13-Δ strain with either pVL1091 or pVL1107 for ∼100 (4×) to ∼150 (6×) generations after eviction of the CDC13 plasmid. Theest1-Δ cdc13-Δ/pVL1111 (Cdc13-Est2D670) strain (constructed by dissection of aest1-Δ/EST1 + cdc13-Δ/CDC13 + diploid strain with pVL1111) is shown after ∼25 and ∼50 generations of growth. (C) Telomere length, after ∼150 generations of growth, of a CDC13 + EST + control strain (lane 1) or an est1−Δcdc13-Δ/pVL1107 strain (lanes 2 and 3). This strain has been propagated for an additional ∼125 generations with no signs of senescence or changes in telomere length (17).

Our results are consistent with a model in which Cdc13 mediates telomerase access by a direct interaction with the enzyme (Fig. 4). Furthermore, these data indicate that Est1 is a comediator of this “accessing” function, potentially as a direct binding partner of Cdc13, although we cannot rule out the possibility of additional intervening protein or proteins. These experiments may have also uncovered an additional role for Est1 in telomerase function, as the Cdc13-Est2 fusion was not capable of promoting extensive telomere elongation in the absence of Est1 (Fig. 3C) (26). Because Est1 is a terminus-specific DNA binding protein (11), we speculate that this second role may be to promote accessibility of the 3′ terminus to the active site of telomerase.

Figure 4

Model for Cdc13 and Est1 as positive regulators of telomerase function. Cdc13 is proposed to bind the single-stranded overhang present at the ends of chromosomes and to mediate telomerase access by a direct but weak protein interaction with a component of the telomerase holoenzyme, possibly Est1. Telomerase is shown as a multisubunit RNA-containing complex that may include additional proteins such as Est3 (14). In addition to the positive regulation described here, telomeres are also subject to negative length regulation in both yeast and human cells, which has been proposed to be mediated by cis-inhibition of telomerase through the action of duplex telomere DNA binding proteins (3). Whether Est1 and Cdc13 are the direct recipients of such negative regulators is an intriguing question.

  • * To whom correspondence should be addressed: E-mail: lundblad{at}bcm.tmc.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article