MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Teng, S.-C.
Right arrow Articles by Zakian, V. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Teng, S.-C.
Right arrow Articles by Zakian, V. A.

Molecular and Cellular Biology, December 1999, p. 8083-8093, Vol. 19, No. 12
0270-7306/99/$04.00+0
Copyright © 1999, American Society for Microbiology. All rights reserved.

Telomere-Telomere Recombination Is an Efficient Bypass Pathway for Telomere Maintenance in Saccharomyces cerevisiae

Shu-Chun Teng and Virginia A. Zakian*

Department of Molecular Biology, Princeton University, Princeton, New Jersey 08544

Received 24 May 1999/Returned for modification 29 June 1999/Accepted 25 August 1999


    ABSTRACT
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Many Saccharomyces telomeres bear one or more copies of the repetitive Y' element followed by ~350 bp of telomerase-generated C1-3A/TG1-3 repeats. Although most cells lacking a gene required for the telomerase pathway die after 50 to 100 cell divisions, survivors arise spontaneously in such cultures. These survivors have one of two distinct patterns of telomeric DNA (V. Lundblad and E. H. Blackburn, Cell 73:347-360, 1993). The more common of the two patterns, seen in type I survivors, is tandem amplification of Y' followed by very short tracts of C1-3A/TG1-3 DNA. By determining the structure of singly tagged telomeres, chromosomes in type II survivors were shown to end in very long and heterogeneous-length tracts of C1-3A/TG1-3 DNA, with some telomeres having 12 kb or more of C1-3A/TG1-3 repeats. Maintenance of these long telomeres required the continuous presence of Rad52p. Whereas type I survivors often converted to the type II structure of telomeric DNA, the type II pattern was maintained for at least 250 cell divisions. However, during outgrowth, the structure of type II telomeres was dynamic, displaying gradual shortening as well as other structural changes that could be explained by continuous gene conversion events with other telomeres. Although most type II survivors had a growth rate similar to that of telomerase-proficient cells, their telomeres slowly returned to wild-type lengths when telomerase was reintroduced. The very long and heterogeneous-length telomeres characteristic of type II survivors in Saccharomyces are reminiscent of the telomeres in immortal human cell lines and tumors that maintain telomeric DNA in the absence of telomerase.


    INTRODUCTION
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Telomeres, the protein-DNA structures found at the natural ends of eukaryotic chromosomes, are required to protect chromosomes from degradation and end-to-end fusion and to facilitate their complete replication. In most organisms, telomeric DNA consists of a short, tandemly repeated sequence that has clusters of G residues in the strand that runs 5' to 3' toward the chromosome end. For example, Saccharomyces chromosomes end in ca. 350 ± 75 bp of C1-3A/TG1-3 DNA (see Fig. 1A). In addition, many eukaryotes have middle repetitive DNA elements or telomere-associated (TA) sequences immediately internal to the simple repeats. In S. cerevisiae, there are two such sequences, X and Y'. X is a heterogeneous sequence found at virtually all telomeres (25). Y' is found in one to four tandem copies, immediately internal to the C1-3A/TG1-3 repeats, on about two-thirds of yeast telomeres (10, 48). There are two classes of Y' elements, Y'-short and Y'-long, with the 5.2-kb Y'-short differing from the 6.7-kb Y'-long by a 1.5-kb internal deletion (25). When Y' is tandemly repeated, a given array consists of all Y'-long or all Y'-short elements (24). In wild-type cells, Y' sequences can be lost or duplicated by mitotic recombination between sister chromatids or different chromosome ends (24).

In most eukaryotes, including yeast, telomere replication is carried out by a special reverse transcriptase, telomerase, that uses a small C-rich stretch in its RNA component as a template for the extension of the G-rich strand (reviewed in reference 36). The genes encoding the RNA (TLC1) (44) and protein catalytic subunit (EST2) (12, 22) of the Saccharomyces telomerase have been identified. Several additional genes, including EST1, which encodes a telomerase RNA-associated protein (20), and CDC13, which encodes a protein that binds telomeres in vivo (4), are also required for telomerase replication in vivo (19, 21, 35). When any of the yeast genes that are essential for the telomerase pathway are deleted, the telomere length gradually shortens, chromosome loss increases, and most cells die (see, for example, reference 28).

Telomerase is not the only mechanism that can maintain telomeric DNA. In Drosophila, transposition of telomere-specific retrotransposons is the major pathway for telomere maintenance (2). Both telomerase and transposition contribute to telomere maintenance in the green alga Chlorella (17). Telomere-telomere recombination is thought to be the sole mechanism for maintaining the repeats at chromosome ends in some insects, such as the mosquito Anopheles (40) and the dipteran Chironomus (23).

Even in organisms that normally rely on telomerase, telomerase-independent mechanisms of telomere maintenance exist. Although most cells in S. cerevisiae (27), Schizosaccharomyces pombe (34), and Kluyveromyces lactis (29) that lack the gene for a telomerase component die, survivors arise relatively frequently in all three organisms. In both S. cerevisiae and K. lactis, generation of survivors requires RAD52-dependent recombination. In S. cerevisiae (discussed in more detail below), the survivors that have been characterized in detail have very short telomeric C1-3A/TG1-3 tracts but long tandem arrays of Y' DNA. In contrast, in K. lactis, survivors have long tracts of telomeric repeats (29). S. pombe can escape the telomerase requirement in two ways, by amplification of its TA repeats, presumably by recombination, or by loss of both TA and telomeric DNA followed by end-to-end fusions to generate circular chromosomes (33). Since some human cell lines (7) and tumors (8) that lack telomerase have very long telomeres, telomerase bypass pathways exist in mammals as well.

The generation of survivors in the absence of telomerase has been studied most extensively in est1Delta strains of Saccharomyces (27). In that pioneering study, the authors described two types of telomerase-independent survivors based on the pattern of restriction fragments produced after digestion with XhoI. Type I survivors had tandem duplication of the subtelomeric Y' element, whereas type II survivors were suggested to arise by rearrangement and/or tandem duplication of the distal portion of Y'. Similar type I and type II survivors were observed in tlc1, est2, est3, and est4 strains, but the structure of DNA in these strains has not been characterized in detail (19).

We reinvestigated the structure of telomeric DNA in type II survivors arising in a tlc1 strain. We found that cDNA-mediated recombination of Y' elements to chromosome ends occurred but its frequency was too low to support telomere maintenance in the absence of telomerase. Likewise, type II survivors did not arise as a result of chromosome circularization. Rather, type II survivors had very long terminal tracts of C1-3A/TG1-3 DNA, with some telomeres being as much as 12 kb longer than telomeres in wild-type cells. This pattern is similar to the exceptionally long telomeres in human tumors (8) or cultured cells (7) that lack telomerase. The maintenance of these elongated telomeres required Rad52p, but reintroduction of telomerase resulted in the slow loss of telomeric DNA until all telomeres returned to wild-type lengths.


    MATERIALS AND METHODS
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Plasmids, yeast strains, yeast transformations, and genetic manipulation. All the yeast operations were performed by standard methods (39). Yeast strains used in this study were derivatives of YPH501 (MATa/MATalpha ura3-52/ura3-52 lys2-801 amber/lys2-801 amber ade2-101 ochre/ade2-101 ochre trp1Delta 63/trp1Delta 63 his3Delta 200/his3Delta 200 leu2-Delta 1/leu2-Delta 1) (43). YPH501 tlc1::LEU2/TLC1 was constructed by transforming XhoI-digested pBlue61::LEU2 (kindly provided by D. Gottschling) (44) into YPH501 and selecting Leu+ transformants. To complement a tlc1 strain, a CEN plasmid containing TLC1 was made. The full-length TLC1 gene plus 1 kb of 5'-flanking and 0.5 kb of 3'-flanking sequences was cloned into pRS317, a vector having LYS2 as a selectable marker (43). pRS317TLC1 was transformed into YPH501 tlc1::LEU2/TLC1 and sporulated. YPH tlc1::LEU2 segregants carrying pRS317TLC1 were selected by growth on complete medium lacking leucine and lysine.

The his3AI-5', URA3, and his3AI-3' were amplified by PCR with, respectively, his3AI-5',5' (GGACTAGTGCTGCAGCTTTAAATATCG) and his3AI-5',3' (CCCGCTCGAGATGGTCCTCTAGTACACTC), URA3,5' (CCCGCTCGAGCTTTTCAATTCAATTCATC) and URA3,3' (CTCCCCGCGGGTAATAACTGATATAAT), and his3AI-3',5' (CTCCCCGCGGTGTCACTACATAAGAAC) and his3AI-3',3' (TGCTCTAGATGGTCCTCTAGTACTCTC) as primers (underlined segments indicate restriction sites) and pTyhis3AI (45) as a template. To make pSL300his3AI-URA3-his3AI, SpeI-XhoI-digested his3AI-5', XhoI-SacII-digested URA3, and SacII-XbaI-digested his3AI-3' PCR-amplified fragments were cloned sequentially into the multiple-cloning sites of pSL300 (6). The fragment for tagging the 3' untranslated region of the Y' elements with his3AI-URA3-his3AI was amplified by PCR with 50-bp Y' sequences that spanned the stop codon of the Y' ORF2 at the ends of the primers and pSL300his3AI-URA3-his3AI as a template. The resulting Y'-his3AI-URA3-his3AI-Y' PCR-amplified fragment was transformed into the YPH tlc1::LEU2 strain carrying pRS317TLC1. Y'-his3AI-URA3-his3AI-tagged strains were selected on medium lacking uracil. Cells that had lost the URA3 gene by popout recombination were selected on 5-fluoroorotic acid (5-FOA) (3). Tagging of individual telomeres by his3AI was confirmed by Southern blot analysis with Y' and his3AI probes as described below.

Formation of survivors. To lose the pRS317TLC1 plasmid, cells were grown on yeast extract-peptone-dextrose (YEPD) plates overnight and then replica plated to alpha -aminoadipate plates to identify Lys- cells (11). Single colonies were restreaked on alpha -aminoadipate plates. Colonies from alpha -aminoadipate plates were then streaked on YEPD plates for single-colony purification. This procedure was repeated five times on YEPD plates to allow cellular senescence to occur and survivors to appear. The plates were incubated at 30°C for 3 days. Survivors first appeared after four restreaks on YEPD plates. Alternatively, survivors were obtained by inoculating single colonies from the alpha -aminoadipate plates into 10 ml of YEPD medium, growing these to stationary phase by incubation at 30°C for 3 days, and then diluting the cultures 1:10,000 into fresh YEPD medium. This procedure was repeated three or four times, and then the cells were plated on YEPD plates to identify survivor colonies. Most survivors obtained by the liquid growth method were type II survivors due to their faster growth compared to type I cells.

To determine if maintenance of survivors required Rad52p, YPH500 (43) was mated to Y0025 (12) (from R. Weinberg), a strain in which the RAD52 gene was replaced with HIS3. One copy of TCL1 was replaced with TRP1, and a centromere plasmid containing URA3 and RAD52 was introduced by transformation. This strain was sporulated, and tlc1:TRP1 rad52:HIS3 spores containing the RAD52 plasmid were identified. Suppressors were isolated as described above, and then cells that lost the RAD52 plasmid were identified by their ability to grow on FOA medium.

DNA preparation, enzyme digestion, Southern blot analysis, and gel electrophoresis. Genomic DNA preparation and Southern blot analysis were performed as previously described (30). S1 nuclease and mung bean nuclease treatments were performed as specified by the manufacturer (New England BioLabs). Two-dimensional gel electrophoresis (5) and alkaline denaturing gel electrophoresis (41) were performed as described previously. For the Bal31 exonuclease digestion experiment, 70 µg of genomic DNA from wild-type or type II survivors was digested with 3 U of Bal31 (New England BioLabs) in a 100-µl final volume. A 14-µl volume of digested DNAs was removed every 10 min, subjected to phenol-chloroform extraction and ethanol precipitation, and digested with XhoI. The following probes were used for Southern hybridization: a 270-bp C1-3A fragment, a 1.5-kb SphI-SalI fragment from the 5' end of Y', a 4.2-kb SalI-XhoI fragment from the middle region of Y', a 341-bp XhoI-KpnI fragment from the 3' end of Y', a 586-bp NdeI-NsiI fragment of HIS3, a 1-kb 5'-EcoRI fragment of PIF1, and a 350-bp PCR fragment of TLC1. Probes were randomly labeled with the RTS-Rad prime system (Life Technologies).

For pulsed-field gel electrophoresis (PFGE), yeast chromosomal DNA blocks were prepared by mixing equal volumes of yeast cells from stationary-phase cultures with 1% low-melting agarose (FMC BioProducts) as described previously (39). PFGE was performed with the contour-clamped homogeneous electric field-dynamically regulated CHEF-DR III system (Bio-Rad). Chromosomes were separated on a 1% agarose gel in 0.5× Tris-borate-EDTA (TBE) buffer at 14°C for 30 h at 6.0 V/cm (200 V) with a 120° included angle and a 60- to 120-s linear switch time ramp.

Inverse PCR, cloning and sequencing. Portions (3 µg) of genomic DNA from wild-type and two independently isolated type II survivors with his3AI-tagged telomeres were digested with XhoI. The XhoI-digested fragments were made blunt ended by using Klenow enzyme in the presence of all four nucleotides. Half of each reaction mixture was subjected to Southern blot analysis to determine the size of the his3AI-tagged XhoI-digested telomere fragments. The other half of the reaction mixture was separated on a 0.6% gel. DNAs in the correct size range to contain the his3AI-tagged XhoI-digested telomere fragments were gel purified and ligated at 14°C overnight in a total volume of 200 µl. PCR was carried out with 4 µl of the ligation mix. The PCR conditions were 30 to 40 cycles of 30 s of 94°C denaturation, 1 min of 69°C annealing, and 3 min of 72°C extension. The primers were P1 (5'-TAGCGACCAGCCGGAATGCTTGG-3') and P2 (5'-ACGATGTTCCCTCCACCAAAGGTG-3') facing opposite to each other in his3AI (see Fig. 6). A further PCR amplification step was performed with nested primers P3 (5'-AGCGCTCGTCATGGAACGCAAAC-3') and P4 (5'-CGAGAGTAGAGGTAGATGTGAGAG-3') facing opposite to each other in the Y' element (see Fig. 6). The PCR-generated products were cloned into the pCRIITOPO vector (Invitrogene) and transformed into Escherichia coli STAB2 cells (GIBCO-BRL). Sequencing was performed with cycle-sequencing kits (Epicentre Technologies) with primers P3 and P4.


    RESULTS
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Identification of survivors in a tlc1 strain. Although most est1 cells die, telomerase-independent survivors appear after ~50 to 100 generations (27). This previous study identified two types of survivors in an est1 strain that are distinguishable by their pattern of telomeric XhoI fragments. There is a single site for XhoI in Y' (Fig. 1A). The majority (63%) of est1 survivors (called type I survivors in this paper) have three major XhoI fragments that hybridize to the 3' Y' probe (Fig. 1A), which detects the distal portion of Y'. The sizes of these bands are ~1.3, 6.7, and 5.2 kb. The ~1.3-kb fragment is the terminal fragment from Y' telomeres and consists mainly of Y' DNA with a very short stretch of C1-3A/TG1-3 DNA. The strong hybridization at 6.7 and 5.2 kb is due to tandemly repeated Y' long and Y' short elements, respectively (Fig. 1A). In contrast, XhoI digestion of DNA from type II survivors yields many differently sized XhoI fragments that hybridize to both C1-3A/TG1-3 and 3' Y' probes but not to probes from other regions of Y' (27). The pattern of telomeric XhoI fragments varies among independent type II survivors. The authors concluded that telomeres of type II survivors sustain substantial Y' deletions and rearrangements, possibly containing tandem duplications of the distal segment of Y' (27).


View larger version (33K):
[in this window]
[in a new window]
  		FIG. 1.   (A). Telomeric and subtelomeric structure of S. cerevisiae. C1-3A/TG1-3 DNA is shown in black. The Y'-long element is between a ~50- to 100-bp internal stretch of C1-3A/TG1-3 DNA (26, 46) and a 300- to 400-bp terminal C1-3A/TG1-3 tract. The two open reading frames in Y' are shown as striped boxes, and their direction of transcription is indicated by the arrow. The position of the 0.9-kb his3AI insert is shown as a triangle. The deleted region in Y'-short is indicated by brackets (25). The 3' end of Y' is shown as an open rectangle. The solid lines indicate the restriction fragments used as probes. Restriction sites used for Southern blot analysis are as follows: P, SphI; S, SalI; E, EcoRI; X, XhoI; K, KpnI. (B). Telomeres of tlc1 survivors show two distinct patterns. Genomic DNA from the wild type, senescing tlc1 cells, two independent type I tlc1 survivors, and two independent type II tlc1 survivors was digested with XhoI, fractionated through 1% agarose, and analyzed by Southern blotting with a C1-3A/TG1-3 probe. Size markers are in kilobases.

To further understand telomerase-independent mechanisms for telomere maintenance in S. cerevisiae, a strain lacking TLC1, the gene encoding the RNA component of telomerase was created. This strain contained a LYS2 CEN plasmid harboring the wild-type TLC1 gene to complement the chromosomal tlc1 deletion. Cells that lost the TLC1 plasmid were identified and then restreaked multiple times to obtain survivors. Although most tlc1 cells died, faster-growing survivor cells appeared after ca. 100 to 125 generations (data not shown). Genomic DNA from 24 independent survivors, as well as from wild-type and early-passage tlc1 strains, was isolated, digested with XhoI, and analyzed by Southern blotting with a C1-3A/TG1-3 probe. Most (21 of 24) survivors were type I (two examples are shown in Fig. 1B, lanes type I), and three were type II (lanes type II). Additional type II survivors were obtained in independent experiments.

The growth characteristics of nine type I and nine type II survivors were analyzed by restreaking each survivor 10 times on YEPD plates. The growth rate of most type II survivors and wild-type colonies was similar, whereas type I survivors grew more slowly and their growth rate fluctuated in different restreaks (see, for example, Fig. 2), with senescing cells reappearing at different times during outgrowth. Type I and type II survivors from an est1 strain had similar growth properties (27). We conclude that the survivors obtained in a tlc1 strain were indistinguishable from those obtained in an est1 strain.

Survivors require the continuous presence of Rad52p. Survivors are not obtained in either rad52 est1 strain (27) or rad52 tlc1 strains (18). To determine if Rad52p is required to maintain survivors in cells lacking telomerase, we isolated eight independent type I and eight independent type II survivors in a rad52Delta tlc1 strain carrying a RAD52 URA3 plasmid. After survivors were generated, cells that lost the RAD52 plasmid were identified by their ability to grow on plates containing 5-FOA. Type I survivors grew very poorly or not at all, even on the first restreak after loss of the RAD52 plasmid (Fig. 2). Although type II tlc1 rad52 survivors divided more times than type I tlc1 rad52 survivors did, they formed heterogeneously sized colonies similar to those seen in senescent tlc1 cells and stopped growing altogether after one to three restreaks (Fig. 2). Thus, RAD52 function is needed continuously to maintain both type I and type II survivors.


View larger version (108K):
[in this window]
[in a new window]
  		FIG. 2.   The maintenance of type I and type II survivors requires Rad52p. Eight independent type I and type II survivors were isolated in a rad52 tlc1 strain carrying a plasmid bearing the RAD52 gene. After survivors were characterized, the RAD52 plasmid was lost and cells were streaked onto rich medium. Shown here are cells from one type I and one type II survivor with (left) or without (right) the RAD52 plasmid. The plate is the first restreak after plasmid loss. Some type II survivors grew on the first restreak but not on subsequent streaks. This figure also demonstrates that type II survivors grew faster than type I survivors (compare streaks of +pRAD52 for type I and type II).

Telomeres of tlc1 survivors are not maintained through cDNA-mediated recombination. Drosophila telomeres consist of retrotransposons (1, 17). Y' elements share several structural features with retrotransposons, such as having two overlapping open reading frames that are in different frames but oriented in the same direction (14, 25). Moreover, many integrated copies of retrotransposons have truncated 5' ends due to premature termination of reverse transcription (14), a situation that could explain the proposed tandem duplication of just the 3' end of Y' (27). In yeast, a chromosome without a telomere is seen as a double-strand break (42) and retrotransposons are able to repair chromosomal breaks (31, 45). The RAD52 dependence for generating survivors would be explained if Y' cDNAs were added to chromosome ends by homologous recombination. These considerations led us to test if type II survivors are generated by RNA-mediated transposition of Y' DNA.

To test this possibility, we first tagged the 3' end of Y' elements with the his3AI marker (13) (Fig. 3). The his3AI gene is designed to detect reverse transcription-mediated events in yeast. In his3AI, the HIS3 ORF is interrupted by a 104-bp artificial intron (AI) that is oriented opposite to the direction of HIS3 transcription. Transcription of the his3AI gene results in a nonspliceable RNA. The his3AI gene was inserted within Y' such that transcription from the Y' promoter generates a transcript containing antisense HIS3 sequences interrupted by the AI intron in a spliceable orientation. Reverse transcription of the spliced RNA, followed by either recombination or transposition of the HIS3 cDNA, will generate His+ colonies.


View larger version (45K):
[in this window]
[in a new window]
  		FIG. 3.   The his3AI gene can be used to detect reverse transcriptase-mediated addition of Y' to chromosome ends. The his3AI gene was inserted within the 3' end of Y' (Fig. 1A). The his3AI gene is transcribed from the Y' promoter to generate an antisense HIS3 transcript having the AI intron in a spliceable orientation. If this transcript is spliced and reverse transcribed, it will generate a cDNA containing an uninterrupted HIS3 gene flanked on either side by Y' sequences. This Y' DNA can direct the recombination of the cDNA to a chromosome end. HIS3 gene expression is detected by growth on plates lacking histidine. The deleted region in Y'-short is denoted by brackets. Restriction sites used for Southern blot analysis are indicated: S, SalI; X, XhoI. ds, double-stranded.

Both wild-type and tlc1 strains that contained TLC1 on a plasmid were transformed with a construct having the his3AI marker inserted into the middle of the 0.9-kb 3' end of Y' DNA at the downstream boundary of ORF2 (Fig. 3). Transformants were screened by Southern blotting to obtain strains that had a single his3AI-tagged telomere. Since different telomeres have zero to four copies of Y', the his3AI marker could insert within either an internal or a terminal Y' element. We recovered eight wild-type strains, of which three had the his3AI gene inserted within an internal Y' and five had it inserted within a terminal Y'. We recovered 14 tlc1 strains containing the TLC1 plasmid, comprising 4 tagged at an internal Y' element and 10 tagged at a terminal Y' element.

We selected for loss of the TLC1 plasmid in tlc1 cells containing a his3AI-tagged telomere by using the liquid assay described in Materials and Methods. The liquid culture scheme favored the isolation of type II survivors because of their growth advantage compared to type I survivors. Indeed, Southern blot analysis revealed that most survivors (24 of 24 examined) had the XhoI restriction pattern diagnostic for type II survivors (data not shown). The cultures were then plated onto YEPD plates to determine the total cell number and onto complete plates lacking histidine to determine the fraction of His+ cells. If type II survivors are generated via a cDNA intermediate, the majority of them should have a His+ phenotype. Three tlc1 His+ colonies were recovered from a total of 4.2 × 109 postsenescent tlc1 cells. No His+ colonies were identified in 6.6 × 109 wild-type cells. The three His+ colonies from the tlc1 survivors contained a 0.8-kb XhoI-NdeI fragment that hybridized to the HIS3 probe but not to a probe for AI, as expected if the His+ phenotype resulted from cDNA-mediated movement of HIS3 to a chromosome end. In addition, these strains contained one or more copies of a 0.9-kb XhoI-NdeI fragment that hybridized to both the HIS3 and AI probes, as expected for the original his3AI locus. We conclude that cDNA-mediated movement of a his3AI-tagged Y' occurred but at too low a rate to explain the formation of type II survivors. A similarly small number of His+ cells was found among type I survivors.

Terminal but not internal Y' elements are altered during generation of type II survivors. We used strains having a single his3AI-tagged telomere to determine the structure of telomeric DNA in type II survivors. Twelve independent tlc1 strains (strains 1 to 12), each with a single his3AI-tagged Y' element, were generated. Of the 12 strains, 3 (strains 1, 2, and 6) had his3AI at an internal Y'. Strain 1 was marked at an internal Y'-short element, and strains 2 and 6 were marked at an internal Y'-long element. The nine other tlc1 strains had his3AI inserted within a terminal Y' element. Each of the 12 strains was diluted into liquid YEPD medium to generate survivors. Two different survivors (a and b) from each of the 12 strains were examined in detail.

Genomic DNA was prepared from each of the 24 survivors, digested with XhoI, and examined by Southern blotting with probes for telomeric C1-3A/TG1-3 DNA (Fig. 4A), the 3' end of Y' (Fig. 4B), the 5' and middle parts of Y' (data not shown), and his3AI (Fig. 4C). The telomeric and 3' Y' probes detected multiple irregularly sized bands that ranged in size from ~1.5 to 10 kb in each of the 24 strains. Most XhoI fragments hybridized to both the telomeric and 3' Y' probe (Fig. 4A and B) but not to the 5' or middle regions of Y' (data not shown). Like type I survivors, some (but not all [see also Fig. 1]) of the type II survivors had elevated levels of tandem Y'-long and Y'-short elements compared to the starting wild-type strain (Fig. 4B). The hybridization pattern reported here is identical to that described previously for type II survivors (27). Thus, 24 of 24 recovered survivors were type II survivors. Given the growth advantage of type II over type I survivors and the method used to generate survivors, the predominance of type II survivors was expected.


View larger version (74K):
[in this window]
[in a new window]
  		FIG. 4.   Southern blot analysis of telomeric DNA in independent type II survivors. (A to C) Two tlc1 survivors (named a and b) from each of 12 independent tlc1 strains (strains 1 to 12) with a single his3AI-tagged Y' element were collected. Genomic DNA from each was digested with XhoI, fractionated through a 1% agarose gel, and analyzed by Southern blotting. The filter was hybridized sequentially with a C1-3A/TG1-3 (A), 3' Y' (B), or his3AI (C) probe. Most fragments smaller than 5.2 kb that hybridized to the C1-3A/TG1-3 probe (A) also hybridized to the 3' Y' probe. Fragments in panel A that did not hybridize to the Y' probe are marked with asterisks. Size markers are in kilobases. (D) Interchromosomal recombination in telomerase-minus S. cerevisiae. Chromosomes from two sets of type II survivors (strains 3 a, 3 b, 10 a, and 10 b) and their parent strains (strains 3 and 10) were separated by PFGE. Size markers (in kilobases) are shown on the left. Chromosome numbers are shown on the right (9). (E). Inter-Y' recombination in telomerase-minus S. cerevisiae. Genomic DNAs from the same sets of samples in panel D were digested with SalI. After transfer to a nylon filter, the filters in panels D and E were hybridized with the his3AI probe.

To determine the fate of individual telomeres during survivor formation, the same blot was hybridized with a his3AI probe (Fig. 4C). The sizes of the his3AI-hybridizing sequences in survivors obtained from tlc1 strains with his3AI inserted into an internal Y' (strains 1, 2, and 6) were unchanged compared to the starting tlc1 strain. These bands had the size expected for insertion into Y'-short (strain 1, a and b) or the size expected for insertion into Y'-long (strains 2 and 6, a and b). These data suggest that during formation of type II survivors, internal Y' elements were not subjected to major rearrangements.

The pattern of his3AI hybridization in the 18 survivors obtained from tlc1 strains that had his3AI inserted at a terminal Y' was diverse. Of the 18 survivors, 12 still had a single his3AI-tagged telomere, 1 had no tagged telomere (strain 11 a), and 5 had two or more his3AI tags (strains 9 a, 10 a and b, and 12 a and b). In no case did the two survivors obtained from the same starting tlc1 strain have the same structure. The only consistent feature was that the tagged Y' in the survivor was almost always larger than the tagged Y' fragment in the parental tlc1 strain (the position of the parent band is marked terminal Y'-his3AI in Fig. 4C). These data indicate that the structure of terminal Y' elements is usually altered during the formation of a type II survivor.

To obtain a better understanding of the kinds of changes that can occur during formation of type II survivors, we used PFGE to determine the chromosomal location of the his3AI tag in four independent type II survivors as well as in their parent strains (Fig. 4D). In strain 3, there was a single copy of his3AI inserted within a terminal Y' element on either chromosome VII or XV (these chromosomes comigrate in PFGE [9]). In this case, the telomere lengthening that accompanied the transition to a type II survivor did not involve movement of his3AI to a different chromosome, since the positions of his3AI in the two survivors were the same as in the starting strain (Fig. 4D). Strain 10 also had a single copy of his3AI within a terminal Y' element, this time on chromosome VI. Both strain 10 survivors retained his3AI on chromosome VI but in addition had his3AI on other chromosomes (two new copies of his3AI in survivor 10 a, on chromosomes VII and IV or XII; one new copy in survivor 10 b, on chromosome IV or XII). Digestion with SalI can be used to determine if his3AI was inserted into Y'-short or Y'-long (Fig. 1A). In strain 3 and its two survivors, his3AI was embedded in Y'-long (Fig. 4E). However, during generation of survivors 10 a and 10 b, the his3AI tag moved not only to a new chromosome but also from a Y'-short to a Y'-long element (Fig. 4E). These results indicate that both intra- and interchromosomal events occur during generation of type II survivors.

The his3AI sequences are still near the physical end of the chromosome in type II survivors. Since the his3AI sequences were invariably on larger XhoI fragments in type II survivors than in the starting strain (Fig. 4C), it is possible that they were no longer at the ends of linear chromosomes. For example, chromosome ends without a telomere could fuse to form a circular chromosome, as seen in telomerase-minus S. pombe (33). If the his3AI marker were at the physical end of a chromosome in type II survivors, it would be sensitive to digestion by the exonuclease Bal 31. To address this possibility, genomic DNA was prepared from a wild-type strain with his3AI inserted within a terminal Y' (Fig. 5A, left) and from four independent type II survivors (analysis of survivor 3 b is shown in Fig. 5A, right). DNA was digested with Bal 31 with samples removed at 10-min intervals. The DNA was then digested with XhoI, subjected to electrophoresis, and analyzed by hybridization to a his3AI probe. Since the his3AI hybridizing sequences shortened at the same rate in both wild-type and type II survivor 3 b DNA, his3AI sequences were near a free end in the 3 b survivor. However, it took longer to degrade the his3AI sequences in the 3 b survivor DNA than in its parent strain, suggesting that in the type II survivor, the his3AI sequences were further from the physical end of the chromosome, a result consistent with the larger size of this fragment. Hybridization of the same blot with the internal PIF1 probe demonstrated that nontelomeric sequences were not Bal 31 sensitive (Fig. 5A). Similar results were obtained with three other type II survivors (data not shown). These data argue that the alterations in his3AI-bearing restriction fragments that accompanied the generation of type II survivors do not alter the telomere-proximal location of the tag.


View larger version (40K):
[in this window]
[in a new window]
  		FIG. 5.   The his3AI gene remains near a free chromosome end and embedded within Y' DNA in type II survivors. (A) Genomic DNA from the wild-type strain and a type II survivor (survivor 3 b in Fig. 4C) was digested with Bal 31 exonuclease for increasing lengths of time. Samples were removed at 10-min intervals, subjected to phenol-chloroform extraction and ethanol precipitation, and digested with XhoI. Digested DNAs were fractionated in a 0.7% agarose gel and analyzed by Southern hybridization with, sequentially, a his3AI probe (top) and an internal probe (the PIF1 gene) (bottom). (B) Genomic DNA from two sets of type II survivors (strains 3 b, 3 a, 10 a, and 10 b) and their parent strains (strains 3 and 10) were digested with XhoI (X) and EcoRI (E) separately. Digested DNA was fractionated in a 0.7% agarose gel and analyzed by Southern hybridization with a his3AI probe. Size markers are in kilobases.

The his3AI sequences are still embedded in Y' DNA in type II survivors and have a local structure very similar to that of the starting strain. The Bal 31 data suggest that the his3AI sequences in a type II survivor were on a terminal XhoI fragment but were further from the chromosome end than in the parental tlc1 strain. Restriction enzyme analysis was used to demonstrate that the DNA surrounding the his3A gene in different type II survivors had the characteristics expected for Y' DNA. Genomic DNA from each of four survivors (strains 3 a, 3 b, 10 a, and 10 b in Fig. 4C) as well as DNA from their parent strains (strains 3 and 10) was digested with five different restriction enzymes, SalI, HindIII, EcoRI, BglII, and KpnI, all of which have recognition sites within Y' and/or his3AI DNA (see Fig. 1A for the locations of the sites). Digested DNA was analyzed by Southern blotting with a his3AI probe. The data obtained with EcoRI are shown in Fig. 5B. The XhoI fragments that hybridized to the his3AI probe were larger in each type II survivor than the 2.1-kb XhoI fragment produced in the parent tlc1 strains (Fig. 5B). For example, in survivor 3 b, a 3.5-kb XhoI fragment hybridized to his3AI, whereas in survivor 3 a, a 2.4-kb XhoI fragment hybridized to his3AI (Fig. 5B). There is an EcoRI site in the Y' element 2.6 kb internal to the XhoI site (Fig. 1A). If his3AI were still embedded in the same place in Y' in survivor strains, his3AI should hybridize to a 6.1-kb EcoRI fragment in survivor 3 b (3.5 + 2.6 kb) and a 5-kb EcoRI fragment in survivor 3 a (2.4 + 2.6 kb), exactly the pattern seen (Fig. 5B). Likewise, in survivors 10 a and b, the EcoRI fragment(s) that hybridized to his3AI were 2.6 kb larger than the his3AI-hybridizing XhoI fragments. Since this result was obtained with five of five restriction enzymes, the his3AI fragments in type II survivors behaved as if they were embedded within Y' DNA. Thus, we found no evidence for rearrangement or deletion of Y' elements on the tagged telomeres of type II survivors.

One explanation for the structure of telomeres in type II survivors is that their sequence is similar to that of the starting strain but their structure is altered in a manner that reduces their mobility in agarose gels. For example, type II survivors might have telomeres with very long single-stranded TG1-3 tails, as seen in cdc13-1 cells at semipermissive temperatures (15), and these single-stranded tails might form secondary structures that reduced fragment mobility. However, the mobility of his3AI-tagged telomeres did not change after treatment with the single-strand-specific S1 or mung bean nucleases (data not shown). Moreover, analysis by both two-dimensional (5) and alkaline denaturing (41) gel electrophoresis revealed no difference between telomeres of type II survivors and wild-type cells (data not shown). Thus, there was no evidence for any change in the structure of Y' (Fig. 5B) or in the physical structure of the chromosome end. We conclude that the slower migration of the terminal XhoI fragments from his3AI-tagged telomeres in type II survivors (Fig. 4C) is probably due to the addition of DNA distal to the tagged Y' element.

C1-3A/TG1-3 DNA is distal to Y' on his3AI-tagged telomeres in type II survivors. An inverse PCR strategy (37) (Fig. 6) was used to obtain the sequence of the DNA to either side of his3AI in two type II survivors, 3 a and b, as well as in the parental tlc1 strain (Fig. 6). Genomic DNA from each of the three strains was digested with XhoI, and the restriction fragments were rendered blunt ended by treatment with the Klenow fragment of DNA polymerase I. The his3AI XhoI fragments from the parent (2.1 kb), survivor 3 a (2.4 kb), and survivor 3 b (3.5 kb) strains were gel purified and treated with ligase. Because the DNA was dilute, ligation generated intramolecular circles. The circularized XhoI fragments were subjected to PCR amplification with primers P1 and P2, which are 395 bp apart and facing in opposite directions within the his3AI gene (Fig. 6). For each strain, a PCR product of the expected size was obtained but only in ligase-treated DNA (data not shown). Since there were other ligase-independent PCR products in the first PCR amplification, we did an additional PCR amplification on the products of the first reaction by using nested primers P3 and P4. P3 annealed at the very 3' end of Y', just upstream of the telomeric C1-3A/TG1-3 telomeric tract. P4 annealed just downstream of the XhoI site in Y' and was oriented toward the XhoI site (Fig. 6). Again, PCR fragments of the appropriate size were obtained. These PCR products hybridized to a C1-3A/TG1-3 probe (data not shown). Because the PCR products were from a population of DNA molecules, the exact sequence of telomeric DNA varied from molecule to molecule. Although we could not obtain a precise telomeric sequence from PCR-amplified DNA, when the pool of molecules was sequenced with the P3 primer, the products of all three strains consisted of only T and G residues. When the P4 primer was used, 45 bp of Y' sequence was followed by DNA consisting of only A and C residues (data not shown).


View larger version (28K):
[in this window]
[in a new window]
  		FIG. 6.   Inverse PCR strategy to determine the sequence of the DNA flanking the his3AI gene in type II survivors. Wild-type and tlc1 strains with TLC1 on a plasmid and containing a single his3AI tag were identified and then allowed to lose the TLC1 plasmid. The black box represents the telomeric C1-3A/TG1-3 tract present before survivors were formed. The DNA added to the his3AI gene during the formation of type II survivors is indicated by the question mark. Genomic DNA from the wild type and type II survivors was digested with XhoI and made blunt ended by treatment with Klenow enzyme in the presence of all four nucleotides. The his3AI-tagged XhoI-digested telomere fragments were gel purified and intramolecularly ligated. Ligated telomeric fragments were subjected to PCR amplification with primers P1 and P2. The P1-P2 PCR products were then subjected to nested PCR amplification with primers P3 and P4. Nested PCR products were further analyzed by cloning and sequencing.

To determine the precise sequence of an individual telomere, we cloned the products from the second PCR from survivor 3 a and from its parent tlc1 strain prior to its losing the TLC1 plasmid. Whereas telomeres from the parent strain were easily recovered, type II survivor telomeres were clonable only in STAB2 E. coli (GIBCO-BRL), a strain used to stabilize long tracts of repetitive DNA. We obtained an insert of 0.6 kb, the appropriate size for the telomeric his3AI fragment from survivor 3 a. Using the P3 and P4 primers, we sequenced ~100 to 200 bp from each end of the insert and found that it consisted solely of C2-3A(CA)1-6/(TG)1-6TG2-3 DNA. Although we did not obtain a precise sequence for the rest of the tract, it had the appropriate pattern to be C1-3A/TG1-3 DNA. Thus, a combination of restriction digestion (Fig. 5B) and sequence analysis indicates that the structure of the his3AI-tagged telomere in survivor 3 a was unchanged from the same telomere in the parent strain except that its telomeric C1-3A/TG1-3 tract was much longer.

Type II survivors have very long tracts consisting only of C1-3A/TG1-3 DNA. The cloning and sequencing results confirmed that the his3AI marker in survivor strain 3 a was still embedded at the same site within Y'DNA as in the starting strain and that it had ~600 bp of C1-3A/TG1-3 DNA distal to Y'. This result suggests that the structure of telomeres in type II survivors is similar to that in wild-type cells except that the telomeres in type II survivors are longer and much more heterogeneous. If telomeres of type II survivors consist solely of C1-3A/TG1-3 DNA, they should lack recognition sites for most restriction enzymes, even enzymes that cut frequently in yeast DNA. To test this possibility, genomic DNAs from a wild-type strain (Fig. 7, lane wt), a type I survivor (lane I), and 10 independent type II survivors (lanes II) were digested with a mixture of AluI, HaeIII, HinfI, and MspI. Each of these enzymes recognizes a different 4-bp sequence, and together they are expected to reduce yeast DNA to, on average, 96 bp. There are many sites for these enzymes within Y', including sites 358 bp downstream of the 5' end of Y' and 42 bp upstream of the 3' end of Y'. The DNA was analyzed by Southern blotting with a C1-3A/TG1-3 probe. If many telomeres in type II survivors have long C1-3A/TG1-3 tracts, there will be many large fragments that hybridize to a telomeric probe after digestion with the four enzymes.


View larger version (79K):
[in this window]
[in a new window]
  		FIG. 7.   Newly elongated telomeric fragments in type II survivors are C1-3A/TG1-3 tracts. Genomic DNAs from a wild-type strain, a type I survivor, and 10 type II survivors were digested with a combination of AluI, HaeIII, HinfI, and MspI, which cut at the AGCT, GGCC, GANTC, and CCGG sequences, respectively; fractionated through 1.3% agarose; and transferred to a nylon filter. The filter was hybridized sequentially to a C1-3A/TG1-3 probe (top) and a probe for a nontelomeric fragment (taken from a region of TLC1 that was not deleted; serves as a control for DNA loading). Size markers (kilobases) are shown at the left. An asterisk marks the position of critically short telomeres.

Since the wild-type strain used for this study had telomeres of ~375 ± 75 bp, digestion with the four enzymes is expected to generate C1-3A/TG1-3 hybridizing fragments of ~375 + 42 bp from Y'-bearing telomeres, fragments of ~0.5 kb from tandem Y' elements, and fragments of up to 1.1 kb from X telomeres. Consistent with this expectation, digestion of DNA from wild-type cells with the four enzymes released C1-3A/TG1-3 fragments that were mostly smaller than 1 kb (Fig. 7). DNA from the type I survivor yielded very short fragments, which hybridized to a telomeric probe. In contrast, most fragments containing C1-3A/TG1-3 DNA in type II survivors were large, ranging up to 12 kb. The fact that these large fragments hybridized intensely to the C1-3A/TG1-3 probe provided additional evidence that they consisted solely of C1-3A/TG1-3 DNA. We conclude that most chromosome ends in type II survivors bear very long and variable-length tracts of C1-3A/TG1-3 DNA. However, many of the type II survivors also had a subset of telomeres that were as short as those in type I survivors (asterisk in Fig. 7).

Type II survivors are stable over time, but their telomeres continuously shorten. To test whether the telomere structure in type I and type II survivors was stable, three independent type I survivors and 20 independent type II survivors were restreaked 10 times on YEPD plates. DNA was prepared from different restreaks, digested with XhoI, and analyzed by Southern blotting with a C1-3A/TG1-3 probe (Fig. 8A). Two of the type I survivors maintained a type I telomeric pattern throughout the restreaking period, while the third type I survivor converted to a type II telomeric pattern between the first and fourth restreaks (Fig. 8A, left panel). Concomitant with this switch, the growth rate of this survivor improved and became similar to that of other type II survivors (data not shown). Thus, type I survivors can convert to a type II pattern during outgrowth.


View larger version (90K):
[in this window]
[in a new window]
  		FIG. 8.   Telomeric changes in type I and type II survivors during outgrowth and after introduction of telomerase. (A) Stability of telomeric structure in type I and type II survivors. Two type I and two type II survivors were restreaked 10 times on YEPD plates. Genomic DNA from the 1st, 4th, 7th, and 10th restreaks was digested with XhoI. Southern blot analysis was performed as in Fig. 1B. (B) Telomeres of type II survivors are subject to continuous shortening and gene conversion. The blot shown on the right of panel A was stripped and rehybridized with a his3AI probe. The his3AI marker was lost between the fourth and seventh restreaks in the second type II survivor. (C and D) Telomere length reverts to wild type in type I and type II survivors after reintroduction of telomerase. Plasmid pRS317TLC1 was transformed into both type I and type II survivors. Transformants were restreaked multiple times. Genomic DNA from the 1st and 4th restreaks from a type I survivor and from the 1st, 4th, 7th, 10th, 13th, 16th, and 19th restreaks from a type II survivor was digested with XhoI (C) or with a combination of AluI, HaeIII, HinfI, and MspI (D); fractionated through a 1% agarose gel; and analyzed by Southern blotting with a C1-3A/TG1-3 probe. DNAs from the wild type and from the two survivors before introduction of the pRS317TLC1 plasmid were also analyzed. Similar results were obtained with three additional type II survivors. Size markers are in kilobases. The C1-3A/TG1-3 hybridizing fragments in panel C that are >0.5 kb and that are absent in DNA from the type I survivor after telomerase was reintroduced are from X telomeres. Not all type I survivors lacked X telomeres after reintroduction of TLC1 (data not shown).

In contrast, all 20 type II survivors retained the XhoI pattern of variable-length telomeric fragments characteristic of type II survivors for ~250 cell divisions (~25 cell divisions per restreak) (see Fig. 8A, right panel, for two examples). Although the general pattern of telomeric XhoI fragments did not change during subculturing and the average telomere length remained high, individual C1-3A/TG1-3-hybridizing XhoI fragments in type II survivors appeared to shorten slowly over time (Fig. 8A). This shortening was especially apparent when these blots were reprobed with his3AI (Fig. 8B), which detects a single telomere. The rate of telomere shortening in type II survivors was ~3 bp/cell division. The his3AI marker at the internal Y' of the second type II survivor was lost between the fourth and seventh restreaks. Thus, even though the pattern of variable and very long telomeres characteristic of type II survivors was maintained for at least 250 cell divisions, individual type II telomeres continuously shorten and, as inferred from the loss of the his3AI tag during outgrowth, engage continuously in gene conversion events with other telomeres.

Normal telomere length regulation is restored in type II survivors when TLC1 is reintroduced. Since most cells that lack telomerase do not form survivors, generation of survivors might require a second event that activates a telomerase-independent telomere maintenance pathway. If this model were true, reintroduction of telomerase might not be sufficient to restore a wild-type pattern of telomere structure. To determine if the telomere changes that occur in type I and II survivors were reversible, we introduced a plasmid-borne TLC1 gene into individual type I and type II survivors. Transformants were restreaked multiple times, and DNA was prepared for Southern analysis after variable number of cell divisions in the presence of telomerase. DNA was digested with both XhoI (Fig. 8C) and multiple 4-bp cutters (Fig. 8D). Telomeres of type I survivors reverted to wild-type lengths soon after reintroduction of TLC1, whereas it took many generations for type II telomeres to return to wild-type lengths (~475 cell divisions). Similar results were obtained with three other type II survivors. These data suggest that generation of type II survivors does not require mutations of genes involved in telomere length regulation.


    DISCUSSION
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

Previous studies demonstrated that yeast cells lacking telomerase can maintain telomeric DNA by Rad52p-dependent recombination (27). Type I survivors arise by tandem duplication of the subtelomeric Y' element. Type II survivors were proposed to have deleted Y' elements but were not characterized in detail (27). To characterize the structure of telomeres in individual type II survivors, we tagged the distal portion of Y' on individual telomeres with the his3AI marker. We showed that cDNA-mediated movement of Y' to chromosome ends is not a general mechanism for generating type II survivors. We found no evidence for deletion or rearrangement of Y' elements during the formation of type II survivors (Fig. 3 and 5B). Although the his3AI tags remained near a chromosome end (Fig. 5A), the terminal XhoI fragments were invariably larger (Fig. 4C and 5A) than in the parental strain. DNA sequencing revealed that the DNA distal of the his3AI tag in one type II survivor consisted solely of C1-3A/TG1-3 DNA (Fig. 6). Moreover, 10 of 10 type II survivors had heterogeneous-length C1-3A/TG1-3 telomeric fragments that were much longer than the C1-3A/TG1-3 tracts in wild-type cells (Fig. 7). The simplest interpretation of these data is that telomeres in type II survivors differ from wild-type telomeres by having longer and irregularly extended C1-3A/TG1-3 tracts. Since type II survivors required the continuous presence of Rad52p (Fig. 2), the long terminal tracts of C1-3A/TG1-3 DNA on type II telomeres are probably maintained by nonreciprocal recombination between two telomeres (38, 46, 47).

In contrast to type I survivors, which often reverted to a type II pattern of telomeric DNA, the type II pattern of telomeres was persistent, being maintained for at least 250 cell generations (Fig. 8A). Despite the stability of the general pattern of telomeric structure in type II survivors, individual telomeres were not static but, rather, steadily shortened (Fig. 8A and B). Moreover, the his3AI tags on these telomeres could be lost (Fig. 4C and 7B), duplicated (Fig. 4C to E), or transferred to a different chromosome (Fig. 4D). Since most type II survivors had a growth rate similar to that of wild-type cells, telomere-telomere recombination must be an efficient mechanism for maintaining telomeric DNA. Nonetheless, after reintroduction of telomerase, type II telomeres slowly returned to wild-type lengths (Fig. 8D), suggesting that expression of telomerase in some way suppressed telomere-telomere recombination.

Saccharomyces telomeric C1-3A/TG1-3 repeats are associated in vivo with at least eight distinct proteins (4, 16). We propose that as telomeres become critically short, the binding of one or more of these proteins is impaired, an event that could expose the single-stranded 3' TG1-3 tail, freeing it to invade another telomere. Consistent with this possibility, a tlc1 strain that lacks Rif2p, an in vivo telomere binding protein (4), generates exclusively type II survivors, suggesting that Rif2p normally inhibits telomere-telomere recombination (45a). According to this model, telomere-telomere recombination does not occur once telomerase is reintroduced, because telomerase lengthens critically short telomeres, eliminating the ends that would otherwise initiate recombination. Telomere-telomere recombination in K. lactis is also proposed to result from loss of telomere binding proteins (29).

Since all telomeres in a senescing strain are short, a single telomere-telomere recombination event will at most double the size of the recombining telomere. For a telomere to become as long as the typical type II telomere, it must either undergo multiple recombination events with a short telomere or invade a very long telomere. If, as we propose, only critically short telomeres lose telomere binding proteins and initiate telomere-telomere recombination, a short telomere cannot become a long telomere by recombination with another short telomere, because after a single recombination event, it will no longer be critically short. We propose that the rate-limiting step in generating a type II survivor is the creation of one or more telomeres that are long enough to serve as efficient, one-step donors of telomeric DNA. This long telomere might be generated by a series of successive recombination events in which a critically short telomere invades a telomere that was itself lengthened by recombination with another short telomere. Alternatively, formation of the first long telomere might involve a rare replication event, such as repeated replication slippage during gene conversion. If generation of a single long telomere were rare, it would explain why type II survivors are less common than type I. Even after type II telomeres are generated, telomere-telomere recombination must be relatively rare, since individual type II telomeres continuously shortened (Fig. 8B), presumably until they became sufficiently short to initiate recombination.

The pattern of telomere structure in Saccharomyces type II survivors is similar to that described in telomerase-minus K. lactis cells (29). However, since K. lactis survivors go through repeated rounds of telomere elongation, telomere shortening, and senescence (29), their growth characteristics are more similar to those of type I than type II Saccharomyces survivors, perhaps because telomere-telomere recombination is even rarer in K. lactis than in Saccharomyces. The structure of telomeres in type II survivors in Saccharomyces was also similar to that described in human cells that maintain their telomeres by the telomerase-independent alternative lengthening of telomeres (ALT) pathway (7, 8). These cells also have very long and heterogeneous-length telomeres. Moreover, individual telomeres in cells maintaining telomeres by the ALT pathway are seen to both gradually shorten and undergo rapid, one-step elongation (32).

In summary, previous work showed that linear chromosomes in Saccharomyces could be maintained either by telomerase or by recombination-driven amplification of subtelomeric Y' DNA (27). Here we show that there is yet a third mechanism that can maintain the ends of yeast chromosomes, telomere-telomere recombination. Although this pathway appears to be as efficient as telomerase in maintaining linear chromosomes, its occurrence is suppressed in telomerase-proficient cells.


    ACKNOWLEDGMENTS

We thank A. Ivessa, B. McCowan, E. Monson, J. Stavenhagen, A. Taggart, L. Vega, and other members of the Zakian laboratory for their helpful comments on the manuscript.

This work was supported by U.S. Army Breast Cancer grant DAMD17-97-1-7242 to S.-C.T. and NIH grant GM26938 to V.A.Z.


    FOOTNOTES

* Corresponding author. Mailing address: Department of Molecular Biology, Princeton University, Princeton, NJ 08544. Phone: (609) 258-6770. Fax: (609) 258-1701. E-mail: vzakian{at}molbio.princeton.edu.


    REFERENCES
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References

1. Biessman, H., J. M. Mason, K. Ferry, M. d'Hulst, K. Valgeirsdottir, K. L. Traverse, and M. L. Pardue. 1990. Addition of telomere-associated HeT DNA sequences "heals" broken chromosome ends in Drosophila. Cell 61:663-673[Medline].
2. Biessman, H., and J. M. Mason. 1997. Telomere maintenance without telomerase. Chromosoma 106:63-69[Medline].
3. Boeke, J. D., F. LaCroute, and G. R. Fink. 1984. A positive selection for mutants lacking orotidine-5'-phosphate decarboxylase activity in yeast: 5-fluoro-orotic acid resistance. Mol. Gen. Genet. 197:345-346[Medline].
4. Bourns, B. D., M. K. Alexander, A. M. Smith, and V. A. Zakian. 1998. Sir proteins, Rif proteins, and Cdc13p bind Saccharomyces telomeres in vivo. Mol. Cell. Biol. 18:5600-5608[Abstract/Free Full Text].
5. Brewer, B. J., and W. L. Fangman. 1987. The localization of replication origins on ARS plasmids in S. cerevisiae. Cell 51:463-471[Medline].
6. Brosius, J. 1992. Compilation of superlinker vectors. Methods Enzymol. 216:469-483[Medline].
7. Bryan, T. M., A. Anglezou, J. Gupta, S. Bacchetti, and R. R. Reddel. 1995. Telomere elongation in immortal human cells without detectable telomerase activity. EMBO J. 14:4340-4348.
8. Bryan, T. M., A. Englezou, L. Dalla-Pozza, M. A. Dunham, and R. R. Reddel. 1997. Evidence for an alternative mechanism for maintaining telomere length in human tumors and tumor-derived cell lines. Nat. Med. 3:1271-1274[Medline].
9. Carle, G. F., and M. V. Olson. 1985. An electrophoretic karyotype for yeast. Proc. Natl. Acad. Sci. USA 82:3756-3760[Abstract/Free Full Text].
10. Chan, C. S. M., and B.-K. Tye. 1983. Organization of DNA sequences and replication origins at yeast telomeres. Cell 33:563-573[Medline].
11. Chattoo, B. B., F. Sherman, D. A. Atubalis, T. A. Fiellstedt, D. Mehnert, and M. Ogur. 1979. Selection of lys2 mutants of the yeast S. cerevisiae by the utilization of alpha-aminoadipate. Genetics 93:51-65[Abstract/Free Full Text].
12. Counter, C. M., M. Meyerson, E. N. Eaton, and R. A. Weinberg. 1997. The catalytic subunit of yeast telomerase. Proc. Natl. Acad. Sci. USA 94:9202-9207[Abstract/Free Full Text].
13. Curcio, M. J., and D. J. Garfinkel. 1991. Single-step selection for Ty1 element retrotransposition. Proc. Natl. Acad. Sci. USA 88:936-940[Abstract/Free Full Text].
14. Eickbush, T. H. 1992. Transposing without ends: the non-LTR retrotransposable elements. New Biol. 4:430-440[Medline].
15. Garvik, B., M. Carson, and L. Hartwell. 1995. Single-stranded DNA arising at telomeres in cdc13 mutants may constitute a specific signal for the RAD9 checkpoint. Mol. Cell. Biol. 15:6128-6138[Abstract].
16. Gravel, S., M. Larrivee, P. Labrecque, and R. J. Wellinger. 1998. Yeast Ku as a regulator of chromosomal DNA end structure. Science 280:741-744[Abstract/Free Full Text].
17. Higashiyama, T., Y. Noutoshi, M. Fujie, and T. Yamada. 1997. Zepp, a LINE-like retrotransposon accumulated in the Chlorella telomeric region. EMBO J. 16:3715-3723[Medline].
18. Le, S., J. Moore, J. Haber, and C. Greider. 1999. RAD50 and RAD51 define two pathways that collaborate to maintain telomeres in the absence of telomerase. Genetics 152:143-152[Abstract/Free Full Text].
19. Lendvay, T. S., D. K. Morris, J. Sah, B. Balasubramanian, and V. Lundblad. 1996. Senescence mutants of Saccharomyces cerevisiae with a defect in telomere replication identify three additional EST genes. Genetics 144:1399-1412[Abstract].
20. Lin, J.-J., and V. A. Zakian. 1995. An in vitro assay for Saccharomyces telomerase requires EST1. Cell 81:1127-1135[Medline].
21. Lingner, J., T. R. Cech, T. R. Hughes, and V. Lundblad. 1997. Three Ever Shorter Telomere (EST) genes are dispensable for in vitro yeast telomerase activity. Proc. Natl. Acad. Sci. USA 94:11190-11195[Abstract/Free Full Text].
22. Lingner, J., T. R. Hughes, A. Shevchenko, M. Mann, V. Lundblad, and T. R. Cech. 1997. Reverse transcriptase motifs in the catalytic subunit of telomerase. Science 276:561-567[Abstract/Free Full Text].
23. Lopez, C. C., L. Nielsen, and J. E. Edstrom. 1996. Terminal long tandem repeats in chromosomes from Chironomus pallidivittatus. Mol. Cell. Biol. 16:3285-3290[Abstract].
24. Louis, E. J., and J. E. Haber. 1990. Mitotic recombination among subtelomeric Y' repeats in Saccharomyces cerevisiae. Genetics 124:547-559[Abstract].
25. Louis, E. J., and J. E. Haber. 1992. The structure and evolution of subtelomeric Y' repeats in Saccharomyces cerevisiae. Genetics 131:559-574[Abstract].
26. Louis, E. J., E. S. Naumova, A. Lee, G. Naumov, and J. E. Haber. 1994. The chromosome end in yeast: its mosaic nature and influence on recombinational dynamics. Genetics 136:789-802[Abstract].
27. Lundblad, V., and E. H. Blackburn. 1993. An alternative pathway for yeast telomere maintenance rescues est- senescence. Cell 73:347-360[Medline].
28. Lundblad, V., and J. W. Szostak. 1989. A mutant with a defect in telomere elongation leads to senescence in yeast. Cell 57:633-643[Medline].
29. McEachern, M. J., and E. H. Blackburn. 1996. Cap-prevented recombination between terminal telomeric repeat arrays (telomere CPR) maintains telomeres in Kluyveromyces lactis lacking telomerase. Genes Dev. 10:1822-1834[Abstract/Free Full Text].
30. Monson, E. K., D. de Bruin, and V. A. Zakian. 1997. The yeast Cac1 protein is required for the stable inheritance of transcriptionally repressed chromatin at telomeres. Proc. Natl. Acad. Sci. USA 94:13081-13086[Abstract/Free Full Text].
31. Moore, J. K., and J. E. Haber. 1996. Capture of retrotransposon DNA at the sites of chromosomal double-strand breaks. Nature 383:644-646[Medline].
32. Murnane, J. P., L. Sabatier, B. A. Marder, and W. F. Morgan. 1994. Telomere dynamics in an immortal human cell line. EMBO J. 13:4953-4962[Medline].
33. Nakamura, T. M., J. P. Cooper, and T. R. Cech. 1998. Two modes of survival of fission yeast without telomerase. Science 282:493-496[Abstract/Free Full Text].
34. Nakamura, T. M., G. B. Morin, K. B. Chapman, S. L. Weinrich, W. H. Andrews, J. Lingner, C. B. Harley, and T. R. Cech. 1997. Telomerase catalytic subunit homologs from fission yeast and human. Science 277:955-959[Abstract/Free Full Text].
35. Nugent, C. I., T. R. Hughes, N. F. Lue, and V. Lundblad. 1996. Cdc13p: a single-strand telomeric DNA-binding protein with a dual role in yeast telomere maintenance. Science 274:249-252[Abstract/Free Full Text].
36. Nugent, C. I., and V. Lundblad. 1998. The telomerase reverse transcriptase: components and regulation. Genes Dev. 12:1073-1085[Free Full Text].
37. Ochman, H., A. S. Gerber, and D. L. Hartl. 1988. Genetic applications of an inverse polymerase chain reaction. Genetics 120:621-623[Abstract/Free Full Text].
38. Pluta, A. F., and V. A. Zakian. 1989. Recombination occurs during telomere formation in yeast. Nature 337:429-433[Medline].
39. Rose, M. D., F. Winston, and P. Hieter. 1990. Methods in yeast genetics: a laboratory course manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y
40. Roth, C. W., F. Kobeski, M. F. Walter, and H. Biessmann. 1997. Chromosome end elongation by recombination in the mosquito Anopheles gambiae. Mol. Cell. Biol. 17:5176-5183[Abstract].
41. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y
42. Sandell, L. L., and V. A. Zakian. 1993. Loss of a yeast telomere: arrest, recovery and chromosome loss. Cell 75:729-739[Medline].
43. Sikorski, R. S., and P. Hieter. 1989. A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122:19-27[Abstract/Free Full Text].
44. Singer, M. S., and D. E. Gottschling. 1994. TLC1, the template RNA component of the Saccharomyces cerevisiae telomerase. Science 266:404-409[Abstract/Free Full Text].
45. Teng, S. C., B. Kim, and A. Gabriel. 1996. Retrotransposon reverse-transcriptase-mediated repair of chromosomal breaks. Nature 383:641-644[Medline].
45a. Teng, S.-C., B. McCowan, and V. Zakian. Unpublished data.
46. Walmsley, R. M., C. S. M. Chan, B.-K. Tye, and T. D. Petes. 1984. Unusual DNA sequences associated with the ends of yeast chromosomes. Nature 310:157-160[Medline].
47. Wang, S.-S., and V. A. Zakian. 1990. Telomere-telomere recombination provides an express pathway for telomere acquisition. Nature 345:456-458[Medline].
48. Zakian, V. A., and H. M. Blanton. 1988. Distribution of telomere-associated sequences on natural chromosomes of Saccharomyces cerevisiae. Mol. Cell. Biol. 8:2257-2260[Abstract/Free Full Text].

Molecular and Cellular Biology, December 1999, p. 8083-8093, Vol. 19, No. 12
0270-7306/99/$04.00+0
Copyright © 1999, American Society for Microbiology. All rights reserved.


This article has been cited by other articles: