- Email: [email protected]
SGS1 is required for telomere elongation in the absence of telomerase
Brief Communication 125
SGS1 is required for telomere elongation in the absence of telomerase Pei-Hsiu Huang*, Fiona E. Pryde*†, Darren Lester*, Rachelle L. Maddison‡, Rhona H. Borts*‡, Ian D. Hickson† and Edward J. Louis*‡ In S. cerevisiae, mutations in genes that encode telomerase components, such as the genes EST1, EST2, EST3, and TLC1, result in the loss of telomerase activity in vivo [1–3]. Two telomeraseindependent mechanisms can overcome the resulting senescence. Type I survival is characterized by amplification of the subtelomeric Yⴕ elements with a short telomere repeat tract at the terminus [4]. Type II survivors arise through the abrupt addition of long tracts of telomere repeats [4–6]. Both mechanisms are dependent on RAD52 [4, 5] and on either RAD50 or RAD51 [6, 7]. We show here that the telomere elongation pathway in yeast (type II) is dependent on SGS1, the yeast homolog of the gene products of Werner’s (WRN) [8] and Bloom’s (BLM) [9] syndromes. Survival in the absence of SGS1 and EST2 is dependent upon RAD52 and RAD51 but not RAD50. We propose that the RecQ family helicases are required for processing a DNA structure specific to eroding telomeres. Addresses: *Department of Biochemistry, University of Oxford, Oxford OX1 3QU, United Kingdom. †Imperial Cancer Research Fund Laboratories, Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DS, United Kingdom. ‡Department of Genetics, University of Leicester, University Road, Leicester LE1 7RH, United Kingdom. Correspondence: Edward J. Louis E-mail: [email protected] Received: 12 October 2000 Revised: 28 November 2000 Accepted: 6 December 2000 Published: 23 January 2001 Current Biology 2001, 11:125–129 0960-9822/01/$ – see front matter 2001 Elsevier Science Ltd. All rights reserved.
Results and discussion SGS1 is required for telomere elongation in the absence of EST2
In a previous study [10], we assessed the role of Sgs1p in the maintenance of telomere structure and found no obvious defects in sgs1 strains. However, we found a general increase in recombination in sgs1 strains in a number of assays, including those for telomere and Y⬘ recombina-
tion [10]. As both telomerase-independent pathways for telomere maintenance depend on recombination, we tested whether the increased recombination of sgs1 strains affected survival in the absence of telomerase. Diploids heterozygous for both SGS1 and the catalytic subunit of telomerase, EST2, were sporulated. Isogenic haploid spores that differed only in their EST2 and SGS1 genotypes were isolated. Sequential restreaking for individual colonies on YPD agar was used to propagate cells from each spore clone. Each round of subculturing represents approximately 25 cell generations. The est2 strains exhibited senescence on continuous culturing, as expected, but all gave rise to surviving subclones after about 75 generations. Southern analysis, for which we used a Y⬘ TG1–3 probe, revealed that 3 of 20 independent survivors in est2 strains displayed a hybridization pattern characteristic of type II survivors, while the remaining 17 were of type I (Figure 1). DNA from type I survivors showed both a hybridization pattern with strong Y⬘ signals arising from tandem arrays of the long and short Y⬘ forms and a small terminal telomeric fragment of around 1 kb. In contrast, type II survivors displayed hybridizing fragments of variable lengths. These fragments represent long tracts of TG1–3 at different chromosome ends. In 40 out of 40 independent isolates, the sgs1 est2 strains gave rise only to type I survivors (Figure 1). This is significantly different from the 3 out of 20 type II survivors seen in est2 strains (p ⫽ 0.044, Fisher’s exact test). These type I survivors exhibited highly variable intensities of the unit-sized fragments, corresponding to amplified Y⬘s (Figure 1). This appears to be qualitatively different from the variability seen in est2 cells [4], where both fragments always appear to increase in intensity. In sgs1 est2 one of the two fragments was often reduced in copy number or was lost entirely (see lanes 6 and 7 in Figure 1). Two interpretations of the lack of type II survivors in sgs1 est2 strains are possible. The first interpretation is that Sgs1p is required for type II survival. The second is that the frequency of type I survival is elevated in sgs1 mutants, and this heightened frequency obscures any type II events that might arise. Y⬘ recombination increases in sgs1 mutants [10], and this could enrich for type I survivors. Given that type II survivors tend to grow better than type I survivors and that this results in type I cultures switching to type II [5], we tested this second explanation by propagating survivors for a further 300 generations. Continuous
126 Current Biology Vol 11 No 2
subculturing of 12 of the type I survivors from est2 strains resulted in a switch to type II in eight cases (Figure 2a). In contrast, 24 type I survivors from sgs1 est2 strains remained type I after 300 generations of continuous propagation (Figure 2b). This is significantly different from the 75% switches to type II seen in the est2 survivors (p ⫽ 0.00071, Fisher’s exact test). We conclude, therefore, that SGS1 is required for type II survivors to arise.
Figure 1
SGS1 is in the RAD50-dependent pathway leading to type II survival
It is known that type I and type II survivors arise via a RAD52-dependent recombination mechanism [4–7]. We could not recover survivors in sgs1 rad52 est2 strains despite culturing large numbers of cells. In contrast, all cultures of est2 and sgs1 est2 strains gave rise to survivors. This excludes the possibility of the RAD52-independent recombination, previously seen in sgs1 strains [10], having any role in survival in sgs1 est2 strains. The RAD52-dependence of telomere maintenance in the absence of telomerase can be genetically divided into two pathways defined by RAD50 and RAD51 [7]. It has recently been shown [6] that rad51 tlc1 survivors are type II, while those of rad50 tlc1 strains are type I. We tested the role of SGS1 in these two genetic pathways. The rad51 sgs1 est2 strains rapidly underwent senescence, and no survivors were recovered (Figure 3a). These results further support the conclusion that SGS1 is required for type II survival. However, type I (and only type I) survivors were recovered in sgs1 rad50 est2 strains (Figure 3a). We conclude, therefore, that SGS1 is in the same pathway as RAD50 but defines a RAD51-independent branch of the telomere maintenance pathway found in cells lacking telomerase. The subtelomeric composition of Yⴕs influences survival
The results presented above were obtained with strain YP1, which has numerous Y⬘s, some of which exist in tandem arrays [11]. In YP1 the Y⬘s are separated from the adjacent subtelomeric region by short tracts of TG1–3 [12]. In contrast, in strain Y55 there are fewer Y⬘s and no tandem arrays [11], and no internal tracts of TG1–3 have been identified [12]. A prediction based on this genomic organization is that the majority of survivors in Y55 est2 strains would be type II and that, therefore, SGS1 would be required for survival of these strains. Indeed, only one in ten early survivors of Y55 est2 strains appeared to be type I. More significantly, the sgs1 est2 Y55 strains exhibited rapid senescence, and survivors were rare (Figure 3b). These findings further support the role of Sgs1p in type II survival. Does Sgs1p process an eroding telomere–specific structure?
The apparent absence of a role for Sgs1p in normal telomere maintenance [10] and its seemingly paradoxical requirement for the type II pathway may be due to a differ-
Southern analysis of early survivors was performed with a Y⬘ TG1–3containing probe on XhoI restriction digest fragments as previously described [10]. Examples of type I (lanes 4 and 5) and type II (lane 3) survivors from est2 strains are shown along with examples of type I survivors (lanes 6 and 7) from sgs1 est2 strains. Lane 2 contains a senescing strain. The markers (M) are DNA digested with BstEII, with sizes indicated in kilobase pairs (kb). Type I survivors are typified by a 1 kb terminal restriction fragment (TRF, lower arrows). There are two fragments (upper arrows) representing tandem arrays of short (Y⬘-S) and long (Y⬘-L) Y⬘ elements. These can vary in intensity in different survivors, but both generally increase in copy number [4] in est2 survivors. The type I survivors from the sgs1 est2 strains show qualitatively more variable tandem fragment intensities, as seen in lanes 6 and 7. The type II survivor is typified by several terminal restriction fragments ranging in size from 1 kb to many kb, each representing a long telomere tract addition to an individual chromosome end.
ence in telomere structure between telomerase-maintained chromosome ends and eroding ends in telomerase-deficient cells. With normal yeast telomeres, there is a highly regulated cycle of addition and removal of telomere repeats that maintains a narrow telomere size range. This process involves a number of proteins, including the Rad50p/Xrs2p/Mre11p complex, the Ku70p/Ku80p complex, Rap1p, the Rif1p/Rif2p complex, and the telomerase complex (Est2p, TLC1, and accessory factors Est1p and Cdc13p) [13]. These highly regulated nucleoprotein struc-
Brief Communication 127
Figure 2
Southern analysis of long-term cultures of type I survivors. (a) Twelve independent long-term cultures (300 generations) of est2 survivors that were type I when they first arose (lanes 3–14). Eight have switched to type II from type I (lanes 5–6, 8–10, and 12–14). Experimental conditions are as in Figure 1. Lane 2 is a senescing culture. Arrows indicate the long- and short-unit Y⬘ fragments as well as the 1.0 kb
terminal fragment seen in type I survivors. The intermediate-sized terminal restriction fragments of type II survivors that result from the addition of long TG1–3 tracts at different individual ends are clearly visible. (b) Twelve long-term cultures (300 generations) of sgs1 est2 survivors that retain a type I pattern.
Figure 3 Survival in strains lacking Sgs1p where type I survival is precluded. (a) Examples of longterm cultures (approximately 100 generations) of sgs1 rad50 est2 and sgs1 rad51 est2 strains show survivors in the case of rad50 but a lack of survivors in the case of rad51, which cannot give rise to type I survivors. (b) Long-term cultures of wild-type, sgs1, est2, and sgs1 est2 Y55 strains show a lack of survivors in an sgs1 est2 strain, while survivors have arisen in an est2 strain. The majority of sgs1 est2 Y55 strains (more than 75%) fail to give rise to survivors as type I survivors are rare in Y55.
128 Current Biology Vol 11 No 2
tures may not allow access to the machinery necessary for recombination-based telomere maintenance in the absence of telomerase. It has been reported that Rif1p and Rif2p, which are inhibitors of telomerase, repress the telomere-telomere recombination that gives rise to type II survivors [6]. Hence, in the absence of both Rif1p/2p and TLC1, only type II survivors are found [6]. One suggestion [6] is that extra-chromosomal telomere repeat–containing circles could recombine with the critically short telomeres (i.e., those with loss of Rif1p/2p association, which then allows access to a Rad50p complex) to generate the type II survivors. This recombination is RAD50-dependent but RAD51-independent, as we see in sgs1 strains. We propose (Figure 4) that type II survivors arise from inter- or intra-telomere interactions involving a 3⬘ overhang of TG1–3 repeats. In this model Rad50p is required for the invasion of the 3⬘ overhang directly or for the generation of the 3⬘ overhang prior to invasion. Sgs1p is then required either for the formation of the D-loop, which allows DNA synthesis to be primed from the invading end (Figure 4, step 2), and/or for migration of the D-loop structure, which enables DNA synthesis to proceed to the end of the chromosome. This process is analogous to break-induced replication (BIR), one proposed mechanism for repair of double-stranded DNA (dsDNA) breaks [14]. However, in at least one system that requires extensive replication up to the end of the chromosome, BIR does not require Sgs1p [15]. If the type II mechanism is BIR, then Sgs1p must be an additional component specifically required at telomeres. One possibility is that Sgs1p is required for disruption of G-quadruplex DNA that can form in the G-rich DNA of telomeres, as has been demonstrated in vitro [16]. Sgs1p and other members of the RecQ family exhibit several enzymatic properties that are consistent with this proposed role. Eukaryotic RecQ helicases associate with the polymerase accessory factor, PCNA [17], and with replication protein A [18]. WRN stimulates the initiation of DNA synthesis by Pol␦ [19]. Sgs1p and its homologs promote migration of 3- and 4-stranded DNA structures [20, 21] and can initiate unwinding from “bubbles” within DNA (I. D. H., unpublished data). This allows for the ability to initiate D-loop formation/migration. An intriguing possibility for this mechanism is the T-loop structure seen at telomeres in human and mouse cells [22]. Although a similar structure has not yet been reported for yeast telomeres, some researchers have proposed loop and fold-back structures for yeast telomeres to explain the repression of nearby genes [23, 24] and the rapid deletion of long telomeres [25]. A T-loop structure may not be possible in normal telomeres, but in eroding telomeres the end may be free to form the necessary invasion. This invasion may only occur when a critical length is reached
Figure 4
Model for the role of Sgs1p in telomerase-independent telomere maintenance. (a) The substrate for recombination-dependent telomere replication is a duplex with a 3⬘ tail that invades homologous telomere sequences, and this results in the formation of a D-loop. (b) This substrate could originate from a sister chromatid or another chromosome end (path 1) or from the T-loop structure known to form at human telomeres [22] (path 2). (c) The 3⬘-OH terminus of the invading strand primes the DNA synthesis that continues until the end of the templating chromosome. (d) Upon completion, DNA strands are resolved, and the second strand is synthesized. If second-strand synthesis follows shortly after or is coincident with the first-strand synthesis, then continuous replication around the T-loop would be possible and would allow for the abrupt elongation of the telomere tract beyond available template lengths.
in which the Rif1p/2p complex is no longer associated and the Rad50p complex has accessibility [6]. Replication primed from this structure could directly result in the abrupt elongation [5] seen in type II survivors. Alternatively, it could generate the extra-chromosomal circles that have been proposed as the template for a rollingcircle replication mechanism of telomere elongation [6]. The Sgs1p and Rad50p-dependent type II survival mechanism in yeast could be analogous to the process whereby elongated telomeres arise in some mammalian cell lines lacking telomerase activity, the so-called ALT pathway.
Brief Communication 129
Circumstantial evidence for a role of WRN in telomere maintenance comes from the observation that ectopic expression of telomerase can overcome the premature replicative senescence characteristic of primary cultures of Werner’s cells [26]. The Werner’s protein also has a potential role in dissociating extra-chromosomal telomere repeat complexes found in telomerase-negative immortalized cells [27]. In addition, hRAD50 is found constitutively associated with telomeres and is associated with TRF2, which is involved in T-loop formation [28]. This places the appropriate proteins at the right location for a mechanism similar to type II survival to be occurring.
14.
15.
16. 17.
18.
Materials and methods The YP1 and Y55 strains of yeast, their subtelomeric composition and use in SGS1 studies are described elsewhere [10–12]. Yeast culture and telomere length analysis were performed as previously described [2–5, 10]. Disruption of SGS1, RAD50, RAD51, and RAD52 was with LEU2 or LYS2 as previously described [10]. Disruption of EST2 with KANMX was by one-step PCR-mediated transplacement using short, flanking homology to EST2 [29].
Acknowledgements This work was supported by The Wellcome Trust and The Imperial Cancer Research Fund. The authors would like to thank members of the I. D. H., R. H. B., and E. J. L. laboratories for critical reading and Jim Haber for sharing information prior to publication.
19. 20. 21. 22. 23. 24.
References 1. Lendvay TS, Morris DK, Sah J, Balasubramanian B, Lundblad V: Senescence mutants of Saccharomyces cerevisiae with a defect in telomere replication identify three additional EST genes. Genetics 1996, 144:1399-1412. 2. Lundblad V, Szostak JW: A mutant with a defect in telomere elongation leads to senescence in yeast. Cell 1989, 57:633643. 3. Singer MS, Gottschling DE: TLC1: template RNA component of Saccharomyces cerevisiae telomerase. Science 1994, 266:404-409. 4. Lundblad V, Blackburn EH: An alternative pathway for yeast telomere maintenance rescues est1- senescence. Cell 1993, 73:347-360. 5. Teng SC, Zakian VA: Telomere-telomere recombination is an efficient bypass pathway for telomere maintenance in Saccharomyces cerevisiae. Mol Cell Biol 1999, 19:8083-8093. 6. Teng S-C, Chang J, McCowan B, Zakian VA: Telomeraseindependent lengthening of yeast telomeres occurs by an abrupt Rad50p-dependent, Rif-inhibited recombinational process. Molecular Cell 2000, 6:947-952. 7. Le S, Moore JK, Haber JE, Greider CW: RAD50 and RAD51 define two pathways that collaborate to maintain telomeres in the absence of telomerase. Genetics 1999, 152:143-152. 8. Yu C-E, Oshima J, Fu Y-H, Wijsman EM, Hisama F, Alisch R, Mathews S, Nakura J, Miki T, Ouais S, Martin GM, Mulligan J, Schellenberg GD: Positional cloning of the Werner’s syndrome gene. Science 1996, 272:258-262. 9. Ellis NA, Groden J, Ye T-Z, Straughen J, Lennon DJ, Ciocci S, et al.: The Bloom’s syndrome gene product is homologous to RecQ helicases. Cell 1995, 83:655-666. 10. Watt PM, Hickson ID, Borts RH, Louis EJ: SGS1, a homologue of the Bloom’s and Werner’s syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae. Genetics 1996, 144:935-945. 11. Louis EJ, Haber JE: The subtelomeric Yⴕ repeat family in Saccharomyces cerevisiae: an experimental system for repeated sequence evolution. Genetics 1990, 124:533-545. 12. Louis EJ, Naumova ES, Lee A, Naumov G, Haber JE: The chromosome end in yeast: its mosaic nature and influence on recombinational dynamics. Genetics 1994, 136:789-802. 13. Cooper JP: Telomere transitions in yeast: the end of the
25. 26.
27.
28. 29.
chromosome as we know it. Curr Opin Genet Dev 2000, 10:169177. Malkova A, Ivanov EL, Haber JE: Double-strand break repair in the absence of RAD51 in yeast: a possible role for breakinduced DNA replication. Proc Natl Acad Sci USA 1996, 93:71317136. Signon L, Malkova A, Naylor ML, Klein H, and Haber JE: Genetic requirements for RAD51,RAD54-independent breakinduced replication repair of a chromosomal double-strand break. Mol Cell Biol 2001, in press. Karow JK, Wu L, Hickson ID: RecQ family helicases: roles in cancer and aging. Curr Opin Genet Dev 2000, 10:32-38. Lebel M, Spillare EA, Harris CC, Leder P: The Werner syndrome gene product co-purifies with the DNA replication complex and interacts with PCNA and topoisomerase I. J Biol Chem 1999, 274:37795-37799. Brosh RM Jr, Li JL, Kenny MK, Karow JK, Cooper MP, Kureekattil RP, et al.: Replication protein A physically interacts with the Bloom’s syndrome protein and stimulates its helicase activity. J Biol Chem 2000, 275:23500-23508. Kamath-Loeb AS, Johansson E, Burgers PM, Loeb LA: Functional interaction between the Werner syndrome protein and DNA polymerase delta. Proc Natl Acad Sci USA 2000, 97:4603-4608. Bennett RJ, Keck JL, Wang JC: Binding specificity determines polarity of DNA unwinding by the Sgs1 protein of S. cerevisiae. J Mol Biol 1999, 289:235-248. Karow JK, Constantinou A, Li JL, West SC, Hickson ID: The Bloom’s syndrome gene product promotes branch migration of holliday junctions. Proc Natl Acad Sci USA 2000, 97:6504-6508. Griffith JD, Comeau L, Rosenfield S, Stansel RM, Bianchi A, Moss H, et al.: Mammalian telomeres end in a large duplex loop. Cell 1999, 97:503-514. Grunstein M: Yeast heterochromatin: regulation of its assembly and inheritance by histones. Cell 1998, 93:325328. Pryde FE, Louis EJ: Limitations of silencing at native yeast telomeres. EMBO J 1999, 18:2538-2550. Li BB, Lustig AJ: A novel mechanism for telomere size control in Saccharomyces cerevisiae. Gene Develop 1996, 10:13101326. Wyllie FS, Jones CJ, Skinner JW, Haughton MF, Wallis C, WynfordThomas D, et al.: Telomerase prevents the accelerated cell ageing of Werner syndrome fibroblasts. Nat Genet 2000, 24:16-17. Ohsugi I, Tokutake Y, Suzuki N, Ide T, Sugimoto M, Furuichi Y: Telomere repeat DNA forms a large non-covalent complex with unique cohesive properties which is dissociated by Werner syndrome DNA helicase in the presence of replication protein A. Nucleic Acids Res 2000, 28:3642-3648. Zhu XD, Kuster B, Mann M, Petrini JH, Lange T: Cell-cycleregulated association of RAD50/MRE11/NBS1 with TRF2 and human telomeres. Nat Genet 2000, 25:347-352. Wach A, Brachat A, Pohlmann R, Philippsen P: New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 1994, 10:1793-1808.
SGS1 is required for telomere elongation in the absence of telomerase Pei-Hsiu Huang*, Fiona E. Pryde*†, Darren Lester*, Rachelle L. Maddison‡, Rhona H. Borts*‡, Ian D. Hickson† and Edward J. Louis*‡ In S. cerevisiae, mutations in genes that encode telomerase components, such as the genes EST1, EST2, EST3, and TLC1, result in the loss of telomerase activity in vivo [1–3]. Two telomeraseindependent mechanisms can overcome the resulting senescence. Type I survival is characterized by amplification of the subtelomeric Yⴕ elements with a short telomere repeat tract at the terminus [4]. Type II survivors arise through the abrupt addition of long tracts of telomere repeats [4–6]. Both mechanisms are dependent on RAD52 [4, 5] and on either RAD50 or RAD51 [6, 7]. We show here that the telomere elongation pathway in yeast (type II) is dependent on SGS1, the yeast homolog of the gene products of Werner’s (WRN) [8] and Bloom’s (BLM) [9] syndromes. Survival in the absence of SGS1 and EST2 is dependent upon RAD52 and RAD51 but not RAD50. We propose that the RecQ family helicases are required for processing a DNA structure specific to eroding telomeres. Addresses: *Department of Biochemistry, University of Oxford, Oxford OX1 3QU, United Kingdom. †Imperial Cancer Research Fund Laboratories, Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DS, United Kingdom. ‡Department of Genetics, University of Leicester, University Road, Leicester LE1 7RH, United Kingdom. Correspondence: Edward J. Louis E-mail: [email protected] Received: 12 October 2000 Revised: 28 November 2000 Accepted: 6 December 2000 Published: 23 January 2001 Current Biology 2001, 11:125–129 0960-9822/01/$ – see front matter 2001 Elsevier Science Ltd. All rights reserved.
Results and discussion SGS1 is required for telomere elongation in the absence of EST2
In a previous study [10], we assessed the role of Sgs1p in the maintenance of telomere structure and found no obvious defects in sgs1 strains. However, we found a general increase in recombination in sgs1 strains in a number of assays, including those for telomere and Y⬘ recombina-
tion [10]. As both telomerase-independent pathways for telomere maintenance depend on recombination, we tested whether the increased recombination of sgs1 strains affected survival in the absence of telomerase. Diploids heterozygous for both SGS1 and the catalytic subunit of telomerase, EST2, were sporulated. Isogenic haploid spores that differed only in their EST2 and SGS1 genotypes were isolated. Sequential restreaking for individual colonies on YPD agar was used to propagate cells from each spore clone. Each round of subculturing represents approximately 25 cell generations. The est2 strains exhibited senescence on continuous culturing, as expected, but all gave rise to surviving subclones after about 75 generations. Southern analysis, for which we used a Y⬘ TG1–3 probe, revealed that 3 of 20 independent survivors in est2 strains displayed a hybridization pattern characteristic of type II survivors, while the remaining 17 were of type I (Figure 1). DNA from type I survivors showed both a hybridization pattern with strong Y⬘ signals arising from tandem arrays of the long and short Y⬘ forms and a small terminal telomeric fragment of around 1 kb. In contrast, type II survivors displayed hybridizing fragments of variable lengths. These fragments represent long tracts of TG1–3 at different chromosome ends. In 40 out of 40 independent isolates, the sgs1 est2 strains gave rise only to type I survivors (Figure 1). This is significantly different from the 3 out of 20 type II survivors seen in est2 strains (p ⫽ 0.044, Fisher’s exact test). These type I survivors exhibited highly variable intensities of the unit-sized fragments, corresponding to amplified Y⬘s (Figure 1). This appears to be qualitatively different from the variability seen in est2 cells [4], where both fragments always appear to increase in intensity. In sgs1 est2 one of the two fragments was often reduced in copy number or was lost entirely (see lanes 6 and 7 in Figure 1). Two interpretations of the lack of type II survivors in sgs1 est2 strains are possible. The first interpretation is that Sgs1p is required for type II survival. The second is that the frequency of type I survival is elevated in sgs1 mutants, and this heightened frequency obscures any type II events that might arise. Y⬘ recombination increases in sgs1 mutants [10], and this could enrich for type I survivors. Given that type II survivors tend to grow better than type I survivors and that this results in type I cultures switching to type II [5], we tested this second explanation by propagating survivors for a further 300 generations. Continuous
126 Current Biology Vol 11 No 2
subculturing of 12 of the type I survivors from est2 strains resulted in a switch to type II in eight cases (Figure 2a). In contrast, 24 type I survivors from sgs1 est2 strains remained type I after 300 generations of continuous propagation (Figure 2b). This is significantly different from the 75% switches to type II seen in the est2 survivors (p ⫽ 0.00071, Fisher’s exact test). We conclude, therefore, that SGS1 is required for type II survivors to arise.
Figure 1
SGS1 is in the RAD50-dependent pathway leading to type II survival
It is known that type I and type II survivors arise via a RAD52-dependent recombination mechanism [4–7]. We could not recover survivors in sgs1 rad52 est2 strains despite culturing large numbers of cells. In contrast, all cultures of est2 and sgs1 est2 strains gave rise to survivors. This excludes the possibility of the RAD52-independent recombination, previously seen in sgs1 strains [10], having any role in survival in sgs1 est2 strains. The RAD52-dependence of telomere maintenance in the absence of telomerase can be genetically divided into two pathways defined by RAD50 and RAD51 [7]. It has recently been shown [6] that rad51 tlc1 survivors are type II, while those of rad50 tlc1 strains are type I. We tested the role of SGS1 in these two genetic pathways. The rad51 sgs1 est2 strains rapidly underwent senescence, and no survivors were recovered (Figure 3a). These results further support the conclusion that SGS1 is required for type II survival. However, type I (and only type I) survivors were recovered in sgs1 rad50 est2 strains (Figure 3a). We conclude, therefore, that SGS1 is in the same pathway as RAD50 but defines a RAD51-independent branch of the telomere maintenance pathway found in cells lacking telomerase. The subtelomeric composition of Yⴕs influences survival
The results presented above were obtained with strain YP1, which has numerous Y⬘s, some of which exist in tandem arrays [11]. In YP1 the Y⬘s are separated from the adjacent subtelomeric region by short tracts of TG1–3 [12]. In contrast, in strain Y55 there are fewer Y⬘s and no tandem arrays [11], and no internal tracts of TG1–3 have been identified [12]. A prediction based on this genomic organization is that the majority of survivors in Y55 est2 strains would be type II and that, therefore, SGS1 would be required for survival of these strains. Indeed, only one in ten early survivors of Y55 est2 strains appeared to be type I. More significantly, the sgs1 est2 Y55 strains exhibited rapid senescence, and survivors were rare (Figure 3b). These findings further support the role of Sgs1p in type II survival. Does Sgs1p process an eroding telomere–specific structure?
The apparent absence of a role for Sgs1p in normal telomere maintenance [10] and its seemingly paradoxical requirement for the type II pathway may be due to a differ-
Southern analysis of early survivors was performed with a Y⬘ TG1–3containing probe on XhoI restriction digest fragments as previously described [10]. Examples of type I (lanes 4 and 5) and type II (lane 3) survivors from est2 strains are shown along with examples of type I survivors (lanes 6 and 7) from sgs1 est2 strains. Lane 2 contains a senescing strain. The markers (M) are DNA digested with BstEII, with sizes indicated in kilobase pairs (kb). Type I survivors are typified by a 1 kb terminal restriction fragment (TRF, lower arrows). There are two fragments (upper arrows) representing tandem arrays of short (Y⬘-S) and long (Y⬘-L) Y⬘ elements. These can vary in intensity in different survivors, but both generally increase in copy number [4] in est2 survivors. The type I survivors from the sgs1 est2 strains show qualitatively more variable tandem fragment intensities, as seen in lanes 6 and 7. The type II survivor is typified by several terminal restriction fragments ranging in size from 1 kb to many kb, each representing a long telomere tract addition to an individual chromosome end.
ence in telomere structure between telomerase-maintained chromosome ends and eroding ends in telomerase-deficient cells. With normal yeast telomeres, there is a highly regulated cycle of addition and removal of telomere repeats that maintains a narrow telomere size range. This process involves a number of proteins, including the Rad50p/Xrs2p/Mre11p complex, the Ku70p/Ku80p complex, Rap1p, the Rif1p/Rif2p complex, and the telomerase complex (Est2p, TLC1, and accessory factors Est1p and Cdc13p) [13]. These highly regulated nucleoprotein struc-
Brief Communication 127
Figure 2
Southern analysis of long-term cultures of type I survivors. (a) Twelve independent long-term cultures (300 generations) of est2 survivors that were type I when they first arose (lanes 3–14). Eight have switched to type II from type I (lanes 5–6, 8–10, and 12–14). Experimental conditions are as in Figure 1. Lane 2 is a senescing culture. Arrows indicate the long- and short-unit Y⬘ fragments as well as the 1.0 kb
terminal fragment seen in type I survivors. The intermediate-sized terminal restriction fragments of type II survivors that result from the addition of long TG1–3 tracts at different individual ends are clearly visible. (b) Twelve long-term cultures (300 generations) of sgs1 est2 survivors that retain a type I pattern.
Figure 3 Survival in strains lacking Sgs1p where type I survival is precluded. (a) Examples of longterm cultures (approximately 100 generations) of sgs1 rad50 est2 and sgs1 rad51 est2 strains show survivors in the case of rad50 but a lack of survivors in the case of rad51, which cannot give rise to type I survivors. (b) Long-term cultures of wild-type, sgs1, est2, and sgs1 est2 Y55 strains show a lack of survivors in an sgs1 est2 strain, while survivors have arisen in an est2 strain. The majority of sgs1 est2 Y55 strains (more than 75%) fail to give rise to survivors as type I survivors are rare in Y55.
128 Current Biology Vol 11 No 2
tures may not allow access to the machinery necessary for recombination-based telomere maintenance in the absence of telomerase. It has been reported that Rif1p and Rif2p, which are inhibitors of telomerase, repress the telomere-telomere recombination that gives rise to type II survivors [6]. Hence, in the absence of both Rif1p/2p and TLC1, only type II survivors are found [6]. One suggestion [6] is that extra-chromosomal telomere repeat–containing circles could recombine with the critically short telomeres (i.e., those with loss of Rif1p/2p association, which then allows access to a Rad50p complex) to generate the type II survivors. This recombination is RAD50-dependent but RAD51-independent, as we see in sgs1 strains. We propose (Figure 4) that type II survivors arise from inter- or intra-telomere interactions involving a 3⬘ overhang of TG1–3 repeats. In this model Rad50p is required for the invasion of the 3⬘ overhang directly or for the generation of the 3⬘ overhang prior to invasion. Sgs1p is then required either for the formation of the D-loop, which allows DNA synthesis to be primed from the invading end (Figure 4, step 2), and/or for migration of the D-loop structure, which enables DNA synthesis to proceed to the end of the chromosome. This process is analogous to break-induced replication (BIR), one proposed mechanism for repair of double-stranded DNA (dsDNA) breaks [14]. However, in at least one system that requires extensive replication up to the end of the chromosome, BIR does not require Sgs1p [15]. If the type II mechanism is BIR, then Sgs1p must be an additional component specifically required at telomeres. One possibility is that Sgs1p is required for disruption of G-quadruplex DNA that can form in the G-rich DNA of telomeres, as has been demonstrated in vitro [16]. Sgs1p and other members of the RecQ family exhibit several enzymatic properties that are consistent with this proposed role. Eukaryotic RecQ helicases associate with the polymerase accessory factor, PCNA [17], and with replication protein A [18]. WRN stimulates the initiation of DNA synthesis by Pol␦ [19]. Sgs1p and its homologs promote migration of 3- and 4-stranded DNA structures [20, 21] and can initiate unwinding from “bubbles” within DNA (I. D. H., unpublished data). This allows for the ability to initiate D-loop formation/migration. An intriguing possibility for this mechanism is the T-loop structure seen at telomeres in human and mouse cells [22]. Although a similar structure has not yet been reported for yeast telomeres, some researchers have proposed loop and fold-back structures for yeast telomeres to explain the repression of nearby genes [23, 24] and the rapid deletion of long telomeres [25]. A T-loop structure may not be possible in normal telomeres, but in eroding telomeres the end may be free to form the necessary invasion. This invasion may only occur when a critical length is reached
Figure 4
Model for the role of Sgs1p in telomerase-independent telomere maintenance. (a) The substrate for recombination-dependent telomere replication is a duplex with a 3⬘ tail that invades homologous telomere sequences, and this results in the formation of a D-loop. (b) This substrate could originate from a sister chromatid or another chromosome end (path 1) or from the T-loop structure known to form at human telomeres [22] (path 2). (c) The 3⬘-OH terminus of the invading strand primes the DNA synthesis that continues until the end of the templating chromosome. (d) Upon completion, DNA strands are resolved, and the second strand is synthesized. If second-strand synthesis follows shortly after or is coincident with the first-strand synthesis, then continuous replication around the T-loop would be possible and would allow for the abrupt elongation of the telomere tract beyond available template lengths.
in which the Rif1p/2p complex is no longer associated and the Rad50p complex has accessibility [6]. Replication primed from this structure could directly result in the abrupt elongation [5] seen in type II survivors. Alternatively, it could generate the extra-chromosomal circles that have been proposed as the template for a rollingcircle replication mechanism of telomere elongation [6]. The Sgs1p and Rad50p-dependent type II survival mechanism in yeast could be analogous to the process whereby elongated telomeres arise in some mammalian cell lines lacking telomerase activity, the so-called ALT pathway.
Brief Communication 129
Circumstantial evidence for a role of WRN in telomere maintenance comes from the observation that ectopic expression of telomerase can overcome the premature replicative senescence characteristic of primary cultures of Werner’s cells [26]. The Werner’s protein also has a potential role in dissociating extra-chromosomal telomere repeat complexes found in telomerase-negative immortalized cells [27]. In addition, hRAD50 is found constitutively associated with telomeres and is associated with TRF2, which is involved in T-loop formation [28]. This places the appropriate proteins at the right location for a mechanism similar to type II survival to be occurring.
14.
15.
16. 17.
18.
Materials and methods The YP1 and Y55 strains of yeast, their subtelomeric composition and use in SGS1 studies are described elsewhere [10–12]. Yeast culture and telomere length analysis were performed as previously described [2–5, 10]. Disruption of SGS1, RAD50, RAD51, and RAD52 was with LEU2 or LYS2 as previously described [10]. Disruption of EST2 with KANMX was by one-step PCR-mediated transplacement using short, flanking homology to EST2 [29].
Acknowledgements This work was supported by The Wellcome Trust and The Imperial Cancer Research Fund. The authors would like to thank members of the I. D. H., R. H. B., and E. J. L. laboratories for critical reading and Jim Haber for sharing information prior to publication.
19. 20. 21. 22. 23. 24.
References 1. Lendvay TS, Morris DK, Sah J, Balasubramanian B, Lundblad V: Senescence mutants of Saccharomyces cerevisiae with a defect in telomere replication identify three additional EST genes. Genetics 1996, 144:1399-1412. 2. Lundblad V, Szostak JW: A mutant with a defect in telomere elongation leads to senescence in yeast. Cell 1989, 57:633643. 3. Singer MS, Gottschling DE: TLC1: template RNA component of Saccharomyces cerevisiae telomerase. Science 1994, 266:404-409. 4. Lundblad V, Blackburn EH: An alternative pathway for yeast telomere maintenance rescues est1- senescence. Cell 1993, 73:347-360. 5. Teng SC, Zakian VA: Telomere-telomere recombination is an efficient bypass pathway for telomere maintenance in Saccharomyces cerevisiae. Mol Cell Biol 1999, 19:8083-8093. 6. Teng S-C, Chang J, McCowan B, Zakian VA: Telomeraseindependent lengthening of yeast telomeres occurs by an abrupt Rad50p-dependent, Rif-inhibited recombinational process. Molecular Cell 2000, 6:947-952. 7. Le S, Moore JK, Haber JE, Greider CW: RAD50 and RAD51 define two pathways that collaborate to maintain telomeres in the absence of telomerase. Genetics 1999, 152:143-152. 8. Yu C-E, Oshima J, Fu Y-H, Wijsman EM, Hisama F, Alisch R, Mathews S, Nakura J, Miki T, Ouais S, Martin GM, Mulligan J, Schellenberg GD: Positional cloning of the Werner’s syndrome gene. Science 1996, 272:258-262. 9. Ellis NA, Groden J, Ye T-Z, Straughen J, Lennon DJ, Ciocci S, et al.: The Bloom’s syndrome gene product is homologous to RecQ helicases. Cell 1995, 83:655-666. 10. Watt PM, Hickson ID, Borts RH, Louis EJ: SGS1, a homologue of the Bloom’s and Werner’s syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae. Genetics 1996, 144:935-945. 11. Louis EJ, Haber JE: The subtelomeric Yⴕ repeat family in Saccharomyces cerevisiae: an experimental system for repeated sequence evolution. Genetics 1990, 124:533-545. 12. Louis EJ, Naumova ES, Lee A, Naumov G, Haber JE: The chromosome end in yeast: its mosaic nature and influence on recombinational dynamics. Genetics 1994, 136:789-802. 13. Cooper JP: Telomere transitions in yeast: the end of the
25. 26.
27.
28. 29.
chromosome as we know it. Curr Opin Genet Dev 2000, 10:169177. Malkova A, Ivanov EL, Haber JE: Double-strand break repair in the absence of RAD51 in yeast: a possible role for breakinduced DNA replication. Proc Natl Acad Sci USA 1996, 93:71317136. Signon L, Malkova A, Naylor ML, Klein H, and Haber JE: Genetic requirements for RAD51,RAD54-independent breakinduced replication repair of a chromosomal double-strand break. Mol Cell Biol 2001, in press. Karow JK, Wu L, Hickson ID: RecQ family helicases: roles in cancer and aging. Curr Opin Genet Dev 2000, 10:32-38. Lebel M, Spillare EA, Harris CC, Leder P: The Werner syndrome gene product co-purifies with the DNA replication complex and interacts with PCNA and topoisomerase I. J Biol Chem 1999, 274:37795-37799. Brosh RM Jr, Li JL, Kenny MK, Karow JK, Cooper MP, Kureekattil RP, et al.: Replication protein A physically interacts with the Bloom’s syndrome protein and stimulates its helicase activity. J Biol Chem 2000, 275:23500-23508. Kamath-Loeb AS, Johansson E, Burgers PM, Loeb LA: Functional interaction between the Werner syndrome protein and DNA polymerase delta. Proc Natl Acad Sci USA 2000, 97:4603-4608. Bennett RJ, Keck JL, Wang JC: Binding specificity determines polarity of DNA unwinding by the Sgs1 protein of S. cerevisiae. J Mol Biol 1999, 289:235-248. Karow JK, Constantinou A, Li JL, West SC, Hickson ID: The Bloom’s syndrome gene product promotes branch migration of holliday junctions. Proc Natl Acad Sci USA 2000, 97:6504-6508. Griffith JD, Comeau L, Rosenfield S, Stansel RM, Bianchi A, Moss H, et al.: Mammalian telomeres end in a large duplex loop. Cell 1999, 97:503-514. Grunstein M: Yeast heterochromatin: regulation of its assembly and inheritance by histones. Cell 1998, 93:325328. Pryde FE, Louis EJ: Limitations of silencing at native yeast telomeres. EMBO J 1999, 18:2538-2550. Li BB, Lustig AJ: A novel mechanism for telomere size control in Saccharomyces cerevisiae. Gene Develop 1996, 10:13101326. Wyllie FS, Jones CJ, Skinner JW, Haughton MF, Wallis C, WynfordThomas D, et al.: Telomerase prevents the accelerated cell ageing of Werner syndrome fibroblasts. Nat Genet 2000, 24:16-17. Ohsugi I, Tokutake Y, Suzuki N, Ide T, Sugimoto M, Furuichi Y: Telomere repeat DNA forms a large non-covalent complex with unique cohesive properties which is dissociated by Werner syndrome DNA helicase in the presence of replication protein A. Nucleic Acids Res 2000, 28:3642-3648. Zhu XD, Kuster B, Mann M, Petrini JH, Lange T: Cell-cycleregulated association of RAD50/MRE11/NBS1 with TRF2 and human telomeres. Nat Genet 2000, 25:347-352. Wach A, Brachat A, Pohlmann R, Philippsen P: New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 1994, 10:1793-1808.