Fine-Tuning the Chromosome Ends the Last Base of Human Telomeres

Extra View Fine-Tuning the Chromosome Ends The Last Base of Human Telomeres

Agnel J. Sfeir ABSTRACT Jerry W. Shay Telomeres protect chromosomes from degradation and loss of vital sequence, block end-end fusion, and allow the cell to distinguish between broken ends and chromosome Woodring E. Wright* ends. Mammalian telomeres end in single-stranded (TTAGGG)-rich 3'-overhangs. that are tucked back into the preceding double stranded region to form a T-loop. The end structure Department of Cell Biology; University of Texas Southwestern Medical Center; Dallas, Texas USA of mammalian telomeres has just started to be elucidated and through this extra views we highlight one aspect of that structure. We have recently identified the terminal nucleotides *Correspondence to: Woodring E. Wright; Department of Cell Biology; UT of both the C-rich and G-rich telomere strands in human cells and showed that ~80% of Southwestern Medical Center; 5323 Harry Hines Boulevard; Dallas, Texas 75390- 9039 USA; Tel.: 214.648.2933; Email: [email protected] the C-rich strands terminate precisely in ATC-5', while the last base of the G-strand is less precise. This finding has important implications for the processing events that act on the Received 09/07/05; Accepted 09/08/05 telomere ends post-replication. While the mechanism behind this phenotype is yet to be Previously published online as a Cell Cycle E-publication: unraveled, we discuss potential models that could explain the last base specificity. http://www.landesbioscience.com/journals/cc/abstract.php?id=2161

KEY WORDS Mammalian telomeres consist of many kilobases of 5'-TTAGGG/3'-AATCCC DNA repeats that are bound by specialized proteins forming a unique end-structure.1 The G-rich telomere, end-replication problem, STELA, pot1, T DISTRIBUTE strand (TTAGGG) extends beyond the double stranded region to form a single stranded terminal nucleotide 3'overhang that is believed to be important for telomere structure and proper functioning.2-5 Every time the cell divides, telomeric repeats are lost due to the “end-replication problem” ACKNOWLEDGEMENTS that was first predicted by James Watson6 and Alexei Olovnikov.7 According to the semi- This work was supported by Department of conservative model of DNA replication, synthesizing the ends of the telomere generates Defense BC031037 to A.J.S. and NIH AG01228 two structurally distinct ends. Leading strand synthesis is continuous and copies the to W.E.W.. W.E.W. is an Ellison Medical C-rich strand (3'-AATCCC. -5')DO to the NO very end to initially generate a blunt ended DNA. Foundation Senior Scholar. Alternatively, the replication machinery could fall off before it reaches the end generating a 5'C-rich overhang.8 Lagging strand synthesis of the G-rich strand is discontinuous and carried out by small Okazaki fragments. A 3'overhang will result upon the removal of the RNA primer used to generate the last fragment (Fig. 1). This overhang could be exactly the size of the RNA primer if it were positioned at the end of the telomere or the overhang could be much longer if the placement of the final Okazaki fragment primer were random. Ultrastructural studies at the EM level have shown that the 3' G-rich overhang does not always exist as a free extension in cells, but is often tucked back into the preceding double stranded region to form a lariat like structure the “t-loop”.3 Furthermore, T- loops can be present on both ends of a chromosome, suggesting that the leading DNA strands (inititially being blunt-ended or possessing a 5' C-rich extension) has been processed in order to generate the 3' G-overhang that is required for t-loop formation. Whether overhangs of lagging strand synthesis are processed further upon the removal of the last RNA primer is yet to be determined. Overhangs constitute an important structure of the ends that are needed for proper telomere function, yet information about their generation remains scarce. Most of the information we know comes from model organisms (yeast and ciliates), in which genetic and structural studies are more easily undertaken. Ciliates have short and very abundant telomeres that facilitate analysis. Studies by Price et al have shown that the overhangs in Euplotes are generated with precise terminal nucleotides at both ends (GGTTTTGG-3' at the G-rich strand and AAAACCC-5' at the C-rich strand) and the length of the overhang ©2005 LANDES BIOSCIENCEis always 14nt.9 Extensive studies characterizing mechanisms of overhang processing have also been done in Tetrahymena. Instead of ending in GGGTTG-3' as would be expected if the terminus is generated by dissociation of telomerase during the translocation step, most Tetrahymena telomeres end in TGGGGT-3'. Tetrahymena C-rich strands end with CAACCC-5' or CCAACC-5'. This suggests that overhang generation is mediated by two separate processing steps; one cleaves the G strand and the other resects the C strand, and both steps are distinctively terminated at a specific base.10,11 www.landesbioscience.com Cell Cycle 1467

Fine-Tuning the Chromosome Ends

THE LAST BASE OF HUMAN TELOMERES The considerable length (5–15kb) of human telomeres makes it hard to analyze the end structure of these repetitive sequences. Recently developed assays to measure the overhang lengths in human cells show that overhang length control is less stringent than model organisms, ranging from ~35 -400 nucleotides.5,12-15 We developed assays that enabled us to determine the end-nucleotides of the telomere. For C-strand terminal nucleotide identification, we performed six different reactions ligating oligonucleotides called C-telorettes to the last base of the C-strand. Six C-telorettes were designed, each ending in a specific permutation of the AATCCC repeat. The telomeric overhang was used as a template that guides Figure 1. Initial products of DNA replication. As the replication machinery is progressing through the telomere, the annealing of the C-telorettes in close it will generate two structurally different ends. The products of lagging strand synthesis initially have proximity to the C-strand end such that 3'G-rich overhangs while products of leading strand synthesis are initially either blunt ended telomeres or those telorettes that anneal next to the 5' telomeres possessing 5'C-rich overhang. Processing events could then modify these structures. end in register would be ligated to the C-strand. Following ligation, DNA was diluted and PCR amplification was performed to determine the fraction of chromosomes that ligated to a given C-telorette, thereby determining their terminal nucleotide. A similar assay was applied for G-strand terminal nucleotide identification, except that a long C-rich template was first annealed to the telomeric overhang to create a 5' overhang that drives the subsequent annealing and ligation of the G-telorettes to the G-strand (Fig. 2). We showed that the C-strand is very tightly regulated such that 80% of the chromosomes end in ATC-5'. The same last base preference was observed for leading and lagging strands, suggesting that the overhang processing step that specifies the end nucleotides is common to both strands of replication. This is in agreement with model organisms, that show a very precise terminal nucleotide of the C-strand, Figure 2. Strategy for end-nucleotide determination. Individual C and G-rich telorettes, representing all six permutations of the telomeric repeats (AATCCC/TTAGGG), are ligated separately to the terminal and suggests the existence of regulated nucleotide of the C-rich and the G-rich telomeric strands. The fraction of the telomeres that ligate to a given processing machinery that acts on the telorette is determined by PCR amplification reactions using a forward primer that is chromosome specific telomere ends post-replication. Unlike (located within subtelomere) and a reverse primer corresponding to the unique sequence part of each the G-terminus of model organisms that telorette oligonucleotide. was specific, the last G-base of humans was less precise. Nevertheless we found a bias towards three nucleotides occur. This is further supported by the fact that in the presence of (TAG-3', TTA-3' and GTT-3'). Since the C-strand ending in 3'- telomerase a shift in terminal nucleotide identity was seen so that CCAATC-5' is the template for leading strand replication, synthe- almost 50% of human chromosomes ended in TAG-3'. This sequence sizing the G-strand all the way to the very end generates a sequence matches the last base of the RNA template region of telomerase, and ending in GGTTAG-3'. If the replication machinery fell off one or is the pause site at which telomerase is predicted to dissociate from two nucleotides before the end, the G-strands would terminate in the telomere.16 It would be the sequence present if no further GGTTA-3' and GGTT-3'. These three ends accounted for 70% of processing of the G-terminus occurred thereafter (Fig. 3). the ends, suggesting that specific cleavage of the G-strand did not

1468 Cell Cycle 2005; Vol. 4 Issue 11 Fine-Tuning the Chromosome Ends

5'-3' exonuclease starts resecting the 5' end of the C-rich strand. In parallel, telomere binding proteins (such as TRF1 and TRF2) are loaded onto the preceding double-stranded DNA. The last protein complex loaded close to the telomere end could create a barrier that blocks further nuclease resection. That protein should bind to the double stranded DNA with great specificity to dictate the identity of the last base (Fig. 4A). An alternative model predicts that single stranded DNA binding proteins such as Pot-1, RPA or hnRNPA are being loaded on a 3'-G-rich overhang that is created by partial exonuclease resection of the 5'-end of the C-rich strand. When a certain threshold protein level is achieved, only the ATC 5'-ends are protected from further resection by the 5'-3' exonuclease (Fig. 4B). The docking space for telomeric single stranded DNA binding proteins might be created by helicase unwinding of the telomere end instead of partial exonuclease resection. Once sufficient proteins are loaded, they could recruit an endonuclease that would specifically cleave at the AATC^CC boundary (Fig. 4C).

POT-1 AND LAST BASE SPECIFICITY Pot-1 (protection of telomeres) is a telomeric Figure 3. Telomere end-nucleotides. More than 80% of human telomeres end in ATC-5' at the single stranded DNA binding protein that binds C-rich strand. The G-rich strand end is less precise with preference for TAG-3' (~40%), and lesser to the 3'-overhang directly via two OB folds at its preference for TTA-3' and GTT-3'. (insert: the pattern observed for the G-terminal nucleotide is consistent with the replication complex generating a blunt end (leaving a GGTTAG-3' end) or N terminus and indirectly by being part of the dissociating one or two nucleotides prior to the terminus (GGGTTA-3' or AGGGTT-3' ends).) six-member telomeric complex (TRF1, TRF2, Tin2, PIP/PTOP/TINT (PPT1), hRAP1 and Pot1).24-26 In vitro binding assays together with HOW ARE THE ENDS SPECIFIED? the solved crystal structure of its OB folds indicate that Pot1 binds Determining the terminal nucleotide of the telomeres provides us very specifically to the 5'-TTAGGGTTAG-3' sequence.27-29 Given with a tool to better understand the nature of the resecting nuclease(s) this great specificity with which it binds the telomere, Pot1 constitutes and the regulatory proteins that generate the overhangs. The speci- a good candidate for specifying the precise overhang end structure. ficity observed for the C-strand terminal nucleotide could stem from For that purpose, C-strand last base determination studies were done a sequence specific nuclease that preferentially clips following a on cells with altered Pot-1.30 When Pot-1 levels are diminished by particular nucleotide. There have been no reports of an exonuclease shRNA knockdowns, the C-strand terminal nucleotide is randomized possessing base-specificity. If it is the nuclease by itself that defines such that the clear preference for ATC-5' is lost, suggesting that the precision of the terminal nucleotide, then such nuclease would Pot-1 is involved in processing of the telomere end post-replication most likely be an endo-nuclease that preferentially cleaves the C-strand to determine the identity of its last base.30 Is Pot-1 determining the at the AATC^CC position. Possible end-processing nucleases last base of the C-strand by the direct binding of its OB fold to the include already identified endonucelases like FEN-117 and Dna218,19 overhang or does it interact with other telomeric proteins to help or exonucleases such as Artemis,20 Werner,21,22 and MRE11.23 Such create the proper ends? If the latter is the case, then what is the iden- nucleases are involved in normal cell functions like VDJ recombina- tity of the associated protein(s)? Furthermore, does Pot-1 interact tion (Artemis), NonHomologous End Joining and Homologous with the telomere end nuclease, such that one can now identify the Recombination (MRE11), leading strand replication (Werner) and long sought after nuclease? The process of end-structure formation Okazaki fragment processing upon DNA synthesis (FEN-1 and is important for our understanding of telomere replication, telomerase Dna2). Some of these nucleases (Werner, Artemis and MRE11) have recruitment, t-loop function and telomere shortening. Answering been linked to telomere stability. However, their direct involvement these questions about the nuclease and mechanisms determining in telomeric overhang generation post-replication has not been end-structure will provide much needed information for exploiting determined. Alternatively, the end-nuclease could be an unidentified our knowledge of telomere biology for therapeutic purposes. telomere specific nuclease that is yet to be characterized. The precision of the C-strand last base could also result from telomere binding proteins that direct the cleavage of a given nuclease to the specified terminal nucleotide, and many different models are possible. One model predicts that after replication is complete, a www.landesbioscience.com Cell Cycle 1469

Fine-Tuning the Chromosome Ends

A B

Figure 4. Potential processing mechanisms. At least three different processing mechanisms can be envisioned. Telomeric Proteins binding to the double stranded DNA after replication could create the boundary for exonuclease resection (A).Alternatively, telomeric binding proteins that bind to the single-stranded overhang could create the boundary blocking further exonuclease resection beyond ATC-5' (B) or they could recruit an endonuclease that would specifically cleave at the precise location ATC^CC (C). Helicases with or without interactions with the telomere-specific single stranded binding proteins, could unwind the C-strand prior to endonuclease cleaage.

References 24. Ye JZ, Hockemeyer D, Krutchinsky AN, Loayza D, Hooper SM, Chait BT, de Lange T. 1. Smogorzewska A, de Lange T. Regulation of telomerase by telomeric proteins. Annu Rev POT1-interacting protein PIP1: A telomere length regulator that recruits POT1 to the Biochem 2004; 73:177-208. TIN2/TRF1 complex. Genes Dev 2004; 18:1649-54. 2. Greider CW. Telomeres do D-loop-T-loop. Cell 1999; 97:419-22. 25. Liu D, Safari A, O’Connor MS, Chan DW, Laegeler A, Qin J, Songyang Z. PTOP inter- acts with POT1 and regulates its localization to telomeres. Nat Cell Biol 2004; 6:673-80. 3. Griffith JD, Comeau L, Rosenfield S, Stansel RM, Bianchi A, Moss H, de Lange T. Mammalian telomeres end in a large duplex loop. Cell 1999; 97:503-14. 26. Loayza D, De Lange T. POT1 as a terminal transducer of TRF1 telomere length control. Nature 2003; 423:1013-8. 4. Wellinger RJ, Ethier K, Labrecque P, Zakian VA. Evidence for a new step in telomere main- tenance. Cell 1996; 85:423-33. 27. Kelleher C, Kurth I, Lingner J. Human protection of telomeres 1 (POT1) is a negative regulator of telomerase activity in vitro. Mol Cell Biol 2005; 25:808-18. 5. Wright WE, Tesmer VM, Huffman KE, Levene SD, Shay JW. Normal human chromosomes have long G-rich telomeric overhangs at one end. Genes Dev 1997; 11:2801-9. 28. Lei M, Podell ER, Cech TR. Structure of human POT1 bound to telomeric single-stranded DNA provides a model for chromosome end-protection. Nat Struct Mol Biol 2004; 6. Watson JD. Origin of concatemeric T7 DNA. Nat New Biol 1972; 239:197-201. 11:1223-9. 7. Olovnikov AM. A theory of marginotomy. The incomplete copying of template margin in 29. Loayza D, Parsons H, Donigian J, Hoke K, de Lange T. DNA binding features of human enzymic synthesis of polynucleotides and biological significance of the phenomenon. J Theor Biol 1973; 41:181-90. POT1: a nonamer 5’-TAGGGTTAG-3’ minimal binding site, sequence specificity, and internal binding to multimeric sites. J Biol Chem 2004; 279:13241-8. 8. Cimino-Reale G, Pascale E, Battiloro E, Starace G, Verna R, D’Ambrosio E. The length of telomeric G-rich strand 3’-overhang measured by oligonucleotide ligation assay. Nucleic 30. Hockemeyer D, Sfeir AJ, Shay JW, Wright WE, de Lange T. POT1 protects telomeres from Acids Res 2001; 29:E35. a transient DNA damage response and determines how human chromosomes end. Embo J 2005; 24:2667-78. 9. Fan X, Price CM. Coordinate regulation of G- and C strand length during new telomere synthesis. Mol Biol Cell 1997; 8:2145-55. 10. Jacob NK, Kirk KE, Price CM. Generation of telomeric G strand overhangs involves both G and C strand cleavage. Mol Cell 2003; 11:1021-32. 11. Jacob NK, Skopp R, Price CM. G-overhang dynamics at Tetrahymena telomeres. Embo J 2001; 20:4299-308. 12. Huffman KE, Levene SD, Tesmer VM, Shay JW, Wright WE. Telomere shortening is pro- portional to the size of the G-rich telomeric 3’-overhang. J Biol Chem 2000; 275:19719-22. 13. Keys B, Serra V, Saretzki G, Von Zglinicki T. Telomere shortening in human fibroblasts is not dependent on the size of the telomeric-3’-overhang. Aging Cell 2004; 3:103-9. 14. Makarov VL, Hirose Y, Langmore JP. Long G tails at both ends of human chromosomes suggest a C strand degradation mechanism for telomere shortening. Cell 1997; 88:657-66. 15. Stewart SA, Ben-Porath I, Carey VJ, O’Connor BF, Hahn WC, Weinberg RA. Erosion of the telomeric single-strand overhang at replicative senescence. Nat Genet 2003; 33:492-6. 16. Morin GB. The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG repeats. Cell 1989; 59:521-9. 17. Liu Y, Kao HI, Bambara RA. Flap endonuclease 1: A central component of DNA metab- olism. Annu Rev Biochem 2004; 73:589-615. 18. Kao HI, Campbell JL, Bambara RA. Dna2p helicase/nuclease is a tracking protein, like FEN1, for flap cleavage during Okazaki fragment maturation. J Biol Chem 2004; 279:50840-9. 19. Tomita K, Kibe T, Kang HY, Seo YS, Uritani M, Ushimaru T, Ueno M. Fission yeast Dna2 is required for generation of the telomeric single-strand overhang. Mol Cell Biol 2004; 24:9557-67. 20. Rooney S, Alt FW, Lombard D, Whitlow S, Eckersdorff M, Fleming J, Fugmann S, Ferguson DO, Schatz DG, Sekiguchi J. Defective DNA repair and increased genomic instability in Artemis-deficient murine cells. J Exp Med 2003; 197:553-65. 21. Bai Y, Murnane JP. Telomere instability in a human tumor cell line expressing NBS1 with mutations at sites phosphorylated by ATM. Mol Cancer Res 2003; 1:1058-69. 22. Crabbe L, Verdun RE, Haggblom CI, Karlseder J. Defective telomere lagging strand synthesis in cells lacking WRN helicase activity. Science 2004; 306:1951-3. 23. D’Amours D, Jackson SP. The Mre11 complex: At the crossroads of dna repair and check- point signalling. Nat Rev Mol Cell Biol 2002; 3:317-27.

1470 Cell Cycle 2005; Vol. 4 Issue 11