Ty1 retrotransposons in Saccharomyces cerevisiae are activated by telomere erosion. Ty1-dependent reverse transcription of mRNA from subtelomeric Y′ repeats generates chimeric Y′-Ty1 cDNA. Here, we show that Y′-Ty1 cDNA is incorporated at eroding telomeres in the absence of telomerase. Telomeric incorporation of Y′-Ty1 cDNA promotes genome rearrangements.
TELOMERASE, a ribonucleoprotein complex consisting of a reverse transcriptase enzyme and an RNA template, extends telomeric DNA repeats at the ends of linear chromosomes. In the absence of telomerase, Saccharomyces cerevisiae strains undergo replicative senescence as telomeric DNA shortens over 50–100 generations. A small fraction of cells escape senescence, and these “survivors” have altered telomere structures that are maintained by homologous recombination (Lundblad and Blackburn 1993; Teng and Zakian 1999). Mobility of the Ty1 long terminal repeat (LTR) retrotransposon in S. cerevisiae is progressively induced by erosion of telomeres in the absence of telomerase, and mobility remains variably induced in survivors (Scholes et al. 2003). One consequence of the induction of Ty1 reverse transcriptase activity is the mobilization of subtelomeric Y′ elements in trans (Maxwell et al. 2004). Chimeric Y′-Ty1 cDNA molecules are incorporated into the genome at frequencies as high as 7 × 10−6 in telomerase-negative survivors (Maxwell et al. 2004). Y′-Ty1 cDNA is synthesized by reverse transcription initiating at the 3′ poly(A) tail of Y′ mRNA, using the end of either the Ty1 LTR or, less frequently, internal regions of Ty1 cDNA in either orientation as a primer (Maxwell et al. 2004). Other chimeric retrosequences consisting of cDNA derived from a variety of cellular mRNAs fused to Ty1 cDNA have also been detected, and they are incorporated into the genome at high frequencies in the absence of telomerase (Derr et al. 1991; Schacherer et al. 2004; Maxwell and Curcio 2007). Here, we determine whether incorporation of Y′-Ty1 cDNA extends telomeres by recombining with subtelomeric Y′ elements, as previously proposed (Maxwell et al. 2004), and whether the presence of Y′-Ty1 cDNA at chromosome ends affects the stability of eroding telomeres.
To detect incorporation of Y′-Ty1 retrosequences into the genome, we used a strain harboring a single chromosomal Y′ element marked in the 3′ untranslated region with the retrotranscript indicator gene his3AI (Maxwell et al. 2004). Splicing of the AI intron from the Y′his3AI transcript, followed by reverse transcription, results in the formation of Y′HIS3 cDNA. Incorporation of Y′HIS3 cDNA into the genome allows cells to become His+ prototrophs. The majority of Y′HIS3 events in tlc1Δ mutants, which lack the telomerase RNA template, have Ty1 sequences 3′ of the oligo(A) tract of the Y′HIS3 cDNA (Maxwell et al. 2004) (Figure 1A).
As an initial test of the hypothesis that Y′HIS3-Ty1 cDNA is incorporated into the genome by recombination with subtelomeric Y′ elements at eroding telomeres, we determined if Y′HIS3 retrosequence formation is dependent on the homologous recombination proteins Rad51p or Rad52p. Following segregation of a TLC1 plasmid in tlc1Δ Y′his3AI derivatives of strain BY4742 (Maxwell et al. 2004) with or without a deletion of RAD52 or RAD51, the frequency of His+ prototroph formation was measured. Rad52p is required to form telomerase-negative survivors (Teng and Zakian 1999), so we measured His+ frequencies after subculturing once in the absence of telomerase (∼25 generations; Table 1). The median His+ frequency of tlc1Δ rad52Δ isolates was at least 10-fold lower and significantly different (P = 0.001) from the median His+ frequency of tlc1Δ isolates (Table 1). Rad51p is not required for recovery from senescence (Teng et al. 2000; Chen et al. 2001), so we assayed His+ frequencies after subculturing twice (∼50 generations) when cell populations were senescent. The median His+ frequency of tcl1Δ rad51Δ populations was 16-fold lower and significantly different (0.005 < P <0.01) from that of tlc1Δ populations (Table 1). There was no difference in the population doubling times that could account for the difference in Y′HIS3 retrosequence formation between tlc1Δ and tlc1Δ rad51Δ strains (data not shown). Furthermore, there was no substantial difference in the quantity of unmarked Y′-Ty1 cDNA detected using a competitive PCR assay (Maxwell et al. 2004). The tlc1Δ and tlc1Δ rad51Δ strains had 4.8 × 10−3 (±0.76 × 10−3) copies/genome and 6.9 × 10−3 (±1.7 × 10−3) copies/genome, respectively. Using PCR analysis, we determined that sequences from the first 50 bp of Y′ are almost always detected 5′ of the HIS3 marker in His+ isolates (data not shown). This is also consistent with incorporation of Y′HIS3-Ty1 cDNA through homologous recombination rather than integration, since integrated Y′HIS3-Ty1 cDNA would lack sequences from untranscribed regions of Y′. We conclude from these data that incorporation of Y′HIS3 cDNA occurs predominantly through homologous recombination.
A second prediction of our hypothesis is that Y′HIS3-Ty1 retrosequences are located centromere distal to X elements and in the same orientation as native Y′ elements. X elements are subtelomeric repeats found centromere proximal to Y′ elements or telomeric DNA at all S. cerevisiae telomeres (Zakian 1996) (Figure 1B). Genomic DNA from 29 His+ survivors harboring Y′HIS3-Ty1 cDNA and 2 His− survivors was digested with NsiI or PvuII, neither of which have cleavage sites in Y′ or HIS3 DNA. Diluted DNA was ligated and used as a template for inverse PCR with an X primer and a primer to the HIS3 splice junction (Figure 1B). Products were obtained from 22 of the 29 samples prepared from survivors harboring Y′HIS3-Ty1 cDNA, but no specific products were obtained from His− survivors. The 22 products obtained each contained Ty1 sequences ligated to sequences adjacent to X DNA (Figure 1, B and C). Therefore, Y′HIS3-Ty1 cDNA was centromere distal to X and in the orientation expected if Y′HIS3-Ty1 cDNA molecules recombined with subtelomeric Y′ elements in at least 76% of His+ strains. Twelve inverse PCR products contained unique subtelomeric sequences, which allowed us to identify eight different telomeres that acted as recipients in recombination with Y′HIS3-Ty1 cDNA (Figure 1C and data not shown). The absence of a product for seven samples could be due to the presence of long Y′ arrays between X and Y′HIS3-Ty1 retrosequences, which would reduce the efficiency of intramolecular ligation, or to the incorporation of Y′HIS3-Ty1 cDNA at nontelomeric sites. Regardless, our results indicate that Y′HIS3-Ty1 cDNA is frequently incorporated at telomeres.
The incorporation of Y′-Ty1 cDNA at telomeres could compensate for the loss of DNA from chromosome termini in telomerase mutants (Maxwell et al. 2004), thereby contributing to the formation of stable alternative telomere structures. On the other hand, the introduction of Ty1 sequences at chromosome termini could destabilize telomeres if telomeric Ty1 sequences recombine with Ty1 elements at other genomic locations. To determine if telomeres harboring Y′-Ty1 retrosequences are stable, we examined isolates of telomerase-negative survivors that harbored or lacked Y′HIS3-Ty1 cDNA for evidence of chromosomal rearrangements using pulsed-field gel electrophoresis. Only 5% of His− isolates lacking Y′HIS3-Ty1 cDNA had a missing or new chromosome band relative to the original His− survivors from which they were derived. By comparison, 76% of the His+ isolates had new or missing chromosome bands compared to the original His− survivors (Figure 2 and Table 2). Of the His+ isolates that had one or more new chromosome bands, 60% had HIS3 sequences on a new chromosome band (Figure 2 and Table 2), consistent with the involvement of the Y′HIS3-Ty1 cDNA in the rearrangement. The strong correlation between incorporation of Y′HIS3-Ty1 cDNA and the presence of chromosomal rearrangements supports the hypothesis that incorporation of Y′-Ty1 cDNA destabilizes telomeres. The observations strengthen the argument that Y′HIS3-Ty1 retrosequences destabilize telomeres and promote genome rearrangements.
In summary, this work indicates that Ty1's role in mobilizing Y′ elements can result in the incorporation of Ty1 sequences at a specific genomic site, telomeres, at which Ty1 sequences are not normally found. Telomeres that terminate with Ty1 sequences are likely unstable and undergo secondary recombination events that produce chromosomal rearrangements. On the basis of our previous characterization of chimeric retrosequences at chromosomal breakpoint junctions (Maxwell and Curcio 2007), the secondary events are likely to involve recombination between Ty1 cDNA at telomeres and Ty1 elements at internal positions on other chromosomes to produce translocations. Potentially, translocations formed by Y′-Ty1 cDNA retrosequence incorporation could form dicentric chromosomes, which would lead to the generation of additional rearrangements through the chromosome breakage–fusion–bridge cycle (Bailey and Murnane 2006). This phenomenon could explain the existence of multiple rearrangements in some strains harboring Y′HIS3 retrosequences (Figure 2).
The frequency of Y′-Ty1 retrosequence formation in telomerase-negative survivors is high enough to compensate for the loss of telomeric DNA in the absence of telomerase (Maxwell et al. 2004). However, the results presented here are not consistent with the idea that incorporation of Y′-Ty1 cDNA contributes to telomere maintenance. Instead, incorporation of Y′-Ty1 retrosequences at telomeres, similar to the incorporation of single-copy gene retrosequences at chromosomal breakpoints (Maxwell and Curcio 2007), promotes restructuring of the genome during telomere crisis. Our work also supports the hypothesis that retrotransposition of mammalian L1 elements to telomeres triggers the formation of chromosome rearrangements (Morrish et al. 2007).
We thank the Wadsworth Center Molecular Genetics Core for DNA sequencing and J. Moran for comments on the manuscript. This work was supported by National Institutes of Health grant GM52072.
Communicating editor: D. Voytas
- Received March 12, 2008.
- Accepted May 12, 2008.
- Copyright © 2008 by the Genetics Society of America