© 1992 Oxford University Press
Human Molecular Genetics, Vol. I, No. 1
3-6
HMG MINI-REVIEW Beginning or end? Telomere structure, genetics and biology David Kipling and Howard J.Cooke MRC Human Genetics Unit, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK Received January 3, 1992; Revised and Accepted February 4, 1992
ABSTRACT
MAMMALIAN TELOMERE STRUCTURE AND BEHAVIOUR The structure of mammalian telomeres It was the discovery of short repeated sequences at the ends of ciliate chromosomes that led to a search for similar sequences at the ends of human chromosomes. Allshire et ai, (6) demonstrated that the TTGGGGn terminal repeats from Tetrahymena would cross-hybridize to human telomeres. Subsequently both TTAGGGn and TTTAGGGn repeats (the latter being the terminal repeats of Arabidopsis) were also shown to cross-hybridize to human telomeres (7,8). These, and other, telomeric sequences can function in yeast, and a popular method of cloning human telomeres uses a modified yeast artificial chromosome vector bearing only a single functional telomere; the second telomere is provided by the repeats at the end of the human insert DNA, to which yeast terminal repeats are added (9, 10, 11, 12, 13). Subcloning and sequencing of the resulting artificial chromosomes demonstrated directly that human chromosomes terminate in TTAGGGn repeat arrays (9, 11). It seems likely from in situ hybridization data that this sequence is present at the end of all vertebrate chromosomes, some nonhuman ones also having large internal tracts of TTAGGGn, commonly in a pericentric or centric location (14, 15). Small regions of TTAGGGn have been cloned from interstitial sites in the human genome and a modified PCR protocol has also been used to clone TTAGGGn-adjacent human sequences (16, 17). The subtelomeric regions of human chromosomes (of which sequence data for the first few kilobases of a number of chromosomes is available) contain a complex mixture of repetitive sequences. A number of different repeats have been described, often present as short tandem repeats with differing periodicities, with 29, 37, 46 and 61bp repeats having been described (16,
18, 19, 20). Some of these repeats are very G+C rich; one 29bp repeat is over 85% G+C and behaves as a hypervariable minisatellite (18, 19, 20). These repeats are present in variable copy numbers at some telomeres and are entirely absent from others. The particular subset of telomeres detected by these repeats using in situ hybridization is polymorphic in the population (18, 19), possibly reflecting subtelomeric translocation between nonhomologous chromosomes, one specific example of which has been described recently (21). An additional family of subtelomeric repeated sequences, located at least 20kb from many chromosome ends, has also been described (22). An additional level of polymorphism has been found at mouse telomeres which are some 5 — 10 times larger than those of humans, enabling individual telomeres to be visualized as discrete terminal restriction fragments using pulsed-field gel electrophoresis (23, 24). The length of these fragments is polymorphic within a population of inbred mice, and can be accounted for by a high rate of generation of new size alleles, although the mechanism underlying this variation is not clear. One major finding of these studies is that only TTAGGGn is present at all human telomeres, implying that other repeats are not absolutely required for telomere function (18, 19, 20). In addition, the subtelomeric repeats only detect fragments in primate species on 'zoo' blots, in marked contrast to the presence of TTAGGGn at all vertebrate telomeres so far analysed (19). Further evidence in support of the hypothesis that the TTAGGGn repeats alone are sufficient for telomere function comes from a case of a thalassaemia with a terminal deletion of chromosome 16 (25). Sequence analysis shows that TTAGGGn repeats have been added directly at the breakpoint site, and in a subsequent study Morin has shown that human telomerase can recognize an oligonucleotide corresponding to this a thalassaemia breakpoint and utilize it as a substrate for TTAGGGn addition in vitro (26). Telomere-mediated chromosome breakage A major handicap in the analysis of mammalian telomere behaviour has been the lack of a method for altering telomere structure in a mammalian cell. In yeast the structure of endogenous chromosome ends can be altered by homologous recombination, or a linear minichromosome can be introduced terminating in any sequence desired, and its behaviour in mitosis and meiosis then followed (27, 28). Recent work by Farr et al. (29) suggests a similar approach is feasible in mammalian cells. A linear DNA molecule carrying a selectable marker and terminating in TTAGGGn repeats was introduced into a Chinese hamster cell line in culture and allowed
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
The word telomere derives from the Greek word telos meaning 'end', roughly translating as 'the thing at the end' when the end is that of a chromosome. Telomeres belie their apparent simplicity of structure by being involved in a wide range of diverse biological phenomena. Much of our understanding of telomere behaviour comes from studies in lower eukaryotes such as ciliates and yeast, the subject of many recent reviews (1, 2, 3, 4, 5). Here we concentrate on the mammalian telomere, recent progress in its study, and how recent evidence for an involvement of telomeres in the regulation of gene expression and DNA replication in yeast points to new aspects of mammalian telomere function yet to be explored.
4 Human Molecular Genetics, Vol. 1, No. 1 to integrate at random into the hamster genome. In 6 out of 27 transformants examined the integrated DNA is located at a telomere as defined by in situ hybridization and exonuclease sensitivity. The terminal restriction fragments are heterogeneous in size and larger than the introduced DNA; it seems likely that integration at an interstitial site has been accompanied by chromosome breakage, with the introduced TTAGGGn repeats conferring telomere function at the break site (perhaps by acting to prime TTAGGGn repeat addition by telomerase (30)).
Telomere-mediated chromosome breakage will be a powerful mapping tool, and the ability of cloned telomeres to function when reintroduced into mammalian cells is a significant step in the development of a mammalian artificial chromosome. Regulation of telomere length It has been known for some years that telomeres in human germline cells (eg. sperm) are longer than those in somatic tissue such as blood (6, 11, 31, 32, 33, 34). One proposed explanation for this is the absence of telomere repeat addition (ie. absence of telomerase activity) in somatic cells (31). If so, incomplete end replication would be expected to result in the progressive loss of terminal repeats as somatic cells undergo successive rounds of division. This is indeed what appears to happen in vivo for humans, with both blood (34) and skin (35) cells showing shorter telomeres with increasing donor age, and telomere loss may contribute to the chromosome aberrations typically seen in senescent cells. Senescence and the measurement of cellular time is an intriguingly complex subject (36, 37, 38) and it will be interesting to see to what extent telomere shortening has a causal role. The large telomeres possessed by both young and old mice (23, 24) would seem to preclude a simple relationship between telomere loss and ageing, but more elaborate schemes cannot be ruled out. The long telomeric repeat arrays of mice (23, 24) raised the question of how cells regulate telomere length. This is not as
FUTURE DIRECTIONS FOR MAMMALIAN TELOMERE RESEARCH Organization in the interphase nucleus The mitotic nucleus in interphase has a somewhat undistinguished appearance under the light or electron microscope, but there is considerable evidence suggesting that it is far from lacking in large-scale order (41). We are, for example, familiar with the concept of ribosome assembly being localized to specific regions of the nucleus (the nucleoli) and it seems that RNA splicing may be compartmentalized in an analogous fashion (42, 43,44). There are, of course, situations where discrete chromosomes are visible in mitotic interphase, such as the polytene chromosomes of Drosophila salivary gland nuclei. In this case the centromeres are seen clustered at one end of the nucleus, with the telomeres attached to the nuclear envelope at the other (45, 46, 47, 48), possibly reflecting chromosome orientation as it is at the end of mitosis. There is a large body of indirect evidence to suggest that there are attachments between telomeres, and between telomeres and the nuclear envelope (41, 49). However, much of the data does not directly concern mitotic interphase nuclei but instead uses
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
A human X chromosome with a terminal deletion of Xq. In situ hybridisation with an X specific or satellite probe and MIC2 sequences from the telomere construct used to produce the terminal deletion. Both the normal p-arm copy and the introduced telomeric q-arm copy of the sequences are labelled.
straightforward as might first be thought. A simple balance between a constant rate of terminal repeat addition (eg. by telomerase) and loss (for instance due to incomplete replication) is unstable over a long period of time; any alteration to either rate would result in steady growth or shrinkage that could continue unabated. It is necessary therefore that changes in telomere length must somehow feed back to the process of loss and/or addition and alter them accordingly. What might the cell be measuring as its criterion for 'telomere length'? One possibility would be the total TTAGGGn in the nucleus, perhaps by titration of some sequence-specific DNA binding protein. However, for an individual chromosome this mechanism would not provide stable length regulation. In addition, it would have to be exquisitely sensitive to enable control in species such as the Chinese hamster where there are large internal blocks of TTAGGGn and where loss of all the telomeric TTAGGGn would result in only a small percentage change in the total TTAGGGn level (15). An alternative possibility is suggested by results from Saccharomyces cerevisiae, where it has been shown that telomere length is remarkably stable in the face of changes in the amount of telomeric DNA in the nucleus. Introduction of a massive excess of yeast telomeres or telomeric DNA on replicating circular or linear plasmids has only a mild effect on average telomere length. For example, even a 25-fold excess of telomeres produces only a 50% increase in average telomere length (39). This leads to the interesting possibility that the cell is measuring the length of each individual telomere separately and regulating the number of terminal repeats added to each one accordingly. How might this be achieved? One possibility is that the recognition and binding of telomerase to a chromosome end is enhanced by an interaction with an accessory factors) bound at a fixed internal site on the chromosome. Telomeres with shorter terminal repeat arrays would have their free ends closer to this fixed 'reference point', and might be more favoured substrates for telomerase action. Support for this model comes from the recent isolation of a yeast protein which binds the junction between the subterminal X sequence and the TGj_3 repeats of a cloned yeast telomere (40) and is a candidate for such a reference point factor.
Human Molecular Genetics, Vol. 1, No. 1 5
Telomere behaviour in meiosis The pairing of homologous chromosomes is an essential prerequisite for meiotic recombination. However, although the cytological aspects of pairing in meiotic prophase have been extensively studied, the underlying molecular mechanisms are poorly understood. Pairing in human cells usually starts at sites very close to the telomeres, although with careful observation the contacts are seen to be slightly internal to the very tips of the chromosomes (56, 57). The three-dimensional organization of meiotic chromosomes can be reconstructed from electron micrographs of serial thin sections cut through meiotic nuclei. Applying this elegant if laborious technique to mammalian cells has shown that the telomeres are bound to the nuclear envelope by so-called attachment plaques (56). These sites of attachment are scattered around the nuclear envelope in the early stages of meiosis, but later on become clustered at one end in what is termed the bouquet formation. One consequence of this could be that the ends of the chromosomes are helped to align; sequences near telomeres are restricted to a search through a two-dimensional plane for homologous sequences with which to pair, whereas internal chromosomal regions are free to explore three-dimensional space. Therefore the preferential initiation of pairing near telomeres may be due to such pairings being favoured kinetically. Such initial alignments of chromosomes are often incorrect, and it has been proposed that this may contribute to the generation of new alleles in the germline (58). This would be consistent with the high level of variability of mammalian telomeres.
The influence of telomeres on transcription and DNA replication It is becoming evident from work in budding yeast that telomeres can influence both gene expression and DNA replication. Locating a reporter gene near a yeast telomere can result in its transcription being repressed. This repression is reversible, the gene switching between repressed and active states which are themselves inherited mitotically in a semi-stable fashion (59). The presence of the gene near TG]_3 repeats (the yeast telomeric sequence) is not by itself sufficient to cause repression, which occurs only if the repeat sequences are located at the end of a
linear molecule. The mechanism of telomeric repression may be similar to the transcriptional repression of the silent mating-type loci in yeast, as a number of mutations relieve repression at both sites (60). There is evidence for a distinct chromatin structure at the silent mating-type loci (61, 62, 63) and the chromatin at yeast telomeres may be in similar heterochromatin-like form. Heterochromatic silencing of a locus (such as expression of a gene or accessibility of a rra/is-acting factor to its binding site) within a defined distance of a telomere could also provide a possible mechanism for regulating telomere length. One hallmark of heterochromatic regions of mammalian chromosomes is that they generally replicate late in S phase. Recent work in yeast by Ferguson et al. (64) suggests that one determinant of replication timing is proximity to a telomere. A yeast origin of replication was identified which is activated late in S phase. The timing of firing of this origin appears not to be determined by its primary sequence but by its location near a telomere, as other origins moved near a telomere become late firing, and conversely this late origin fires early if moved to a nontelomeric location. As with transcriptional repression, proximity to TG]_3 repeats is insufficient to cause late initiation unless the repeats terminate a linear molecule. It is tempting to speculate that the heterochromatin-like structure at yeast telomeres responsible for transcriptional repression also results in delayed firing of adjacent origins of replication, and it will be interesting to determine if late replication can be relieved by the same mutations which relieve telomeric transcriptional repression. Although human telomeres replicate throughout the duration of S phase (65) it is not known to what extent they influence the timing of replication of adjacent genomic regions. This epigenetic control of gene expression in yeast may be relevant with respect to understanding metazoan phenomena such as position-effect variegation in Drosophila (66) and the regulation of telomeric genes in mammalian genomes.
SUMMARY AND PROSPECTS The last few years have seen a dramatic increase in our knowledge of mammalian telomere structure, stemming in part from new techniques developed for cloning these DNA sequences. However, various lines of evidence suggest that of all the telomeric and subtelomeric sequences described, only the terminal TTAGGGn repeats are essential for telomere function. One recurring theme from studies of these regions of the mammalian genome is a high rate of structural polymorphism and genetic instability. The underlying reasons remain a mystery, but may reflect the involvement of mammalian telomeres in the initiation of meiotic chromosome pairing that precedes recombination. A significant new tool for the analysis of mammalian telomere structure is- the ability to functionally reintroduce cloned telomeres into a mammalian cell. Coupled with the established technologies of targeted recombination and transgenesis in the mouse we may be close to manipulating mammalian telomere structure in vivo. There are many aspects of telomere biology of which we as yet have only the most fragmentary understanding, such as the relationship between telomeres and the higher order structure of the nucleus. Meanwhile, work in other species continues to uncover new and unexpected aspects of telomere behaviour. We can therefore be confident that the end is by no means in sight for the investigation of the mammalian telomere.
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
polytene or meiotic chromosomes, or instances where telomeretelomere interactions are so strong that they are maintained in squash preparations of prophase nuclei (50). Telomere-telomere interactions, together with attachments to the nuclear envelope, would be expected to contribute to chromosome organization in the nucleus, with implications for processes such as gene expression (51). Telomere distribution in the mammalian mitotic interphase nucleus may be resolved by a combination of in situ hybridization and optical sectioning techniques, as have been successfully applied to trypanosomes and plants (52, 53). Indeed, such techniques have been used recently to study the distribution of telomeres in murine neuronal cells in culture, where they were found not to be associated with the nuclear envelope but distributed throughout the volume of the interphase nucleus (54). In another study the inactive human X chromosome has been shown to form itself into a loop via telomere-telomere association near the nuclear envelope, in contrast to a linear orientation for the active X (55).
6 Human Molecular Genetics, Vol. 1, No. 1 ACKNOWLEDGEMENTS We would like to thank many colleagues in the MRC Unit for stimulating discusssions and comments on the manuscript. D.K. is funded by a grant from the UK MedicaJ Research Council's Human Genome Mapping Project.
38. 39. 40. 41. 42.
REFERENCES
43.
1. 2. 3. 4. 5. 6. 7.
11. 12. 13. 14. 15.
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
8. 9. 10.
Zakian.V.A. (1989) Ann. Rev. Genet. 23, 579-604. Greider,C.W. (1990) BioEssays 12, 363-369. Zakian.V.A., Runge.K. and Wang,S-S. (1990) Trends Genet. 6, 12-16. Blackburn.E.H. (1991) Nature 350, 569-573. Greider.C.W. (1991) Curr. Opinion Cell Biol. 3, 444-451. Allshire.R.C, Gosden.J.R., Cross,S.H., Cranston.G., Rout.D., Sugawara.N., SzostakJ.W., Fantes.P.A. and Hastie.N.D. (1988) Nanire 332, 656-659. Moyzis.R.K., BuckinghamJ.M., Cram.L.S., Dani.M., Deaven,L.L., Jones.M.D., MeyneJ., Ratliff.R.L. and WuJ-R. (1988) Proc. Nail. Acad. Sd. USA 85, 6622-6626. Richards.E.J. and Ausubel.F.M. (1988) Cell 53, 127-136. Brown.W.R.A. (1989) Nature 338, 774-776. Cheng,J-F., Smith.C.L. and Cantor.C.R. (1989) Nucleic Acids Res. 17, 6109-6127. Cross,S.H., Allshire.R.C., McKay,SJ., McGill.N.I. and Cooke.HJ. (1989) Nature 338,771 -774. RJethman.H.C, Moyzis.R.K., MeyneJ., Burke.D.T. and01son,M.V. (1989) Proc. Natl. Acad. Sd. USA 86, 6240-6244. Bates.G.P., MacDonald.M.E., Baxendale.S., Sedlacek.Z., Youngman,S., Romano.D., Whaley.W.L., Allitto,B.A., Poustka,A., Gusella.J.F. and Lehrach.H. (1990) Am. J. Hum. Genet. 46, 762-775. MeyneJ., Ratliff,R.L. and Moyzis.R.K. (1989) Proc. Natl. Acad. Sd. USA 86, 7049-7053. MeyneJ., Baker,R.J., Hobart.H.H., Hsu.T.C, Ryder,O.A., Ward,O.G., WileyJ.E., Wurster-HUl.D.H., Yates.T.L. and Moyzis.R.K. (1990) Chromosoma 99, 3 — 10. Weber.B., CoUins.C, Robtxns.C, Magenis.R.E., Delaney.A.D., Gray.J.W. and Hayden.M.R. (1990) Nudeic Adds Res. 18, 3353-3361. Wells.R.A., Germino.G.G., Krishna.S., Buckle.V.J. and Reeders.S.T. (1990) Genomics 8, 699-704 Brown,W.R.A., MacKinnon.PJ., Villasante,A., Spurr.N., Buckle.V.J. and Dobson,M.J. (1990) Cell 63, 119-132. Cross.S., UndseyJ., FantesJ., McKay.S., McGill.N. and Cooke,H. (1990) Nucleic Adds Res. 18, 6649-6657. ChengJ-F., Smith.C.L. and Cantor.C.R. (1991) Nucleic Acids Res. 19, 149-154. Wilkie.A.O.M., Higgs.D.R., Rack.K.A., Buckle.V.J., Spurr.N.K., HschelGhodsian.N., Ceccherini.I., Brown.W.R.A. and Harris.P.C. (1991) Cell 64, 595-606. Rouyer.F., de la Chapelle.A., Andersson.M. and WeissenbachJ. (1990) EMBOJ. 9, 505-514. Kipling.D., and Cooke.H.J. (1990) Nature 347, 400-402. StarlingJ.A., MauleJ., Hastie.N.D. and Allshire.R.C. (1990) Nucleic Adds Res. 18, 6881-6888. Wilkie.A.O.M., LambJ., Harris.P.C., Rnney.R.D. and Higgs,D.R. (1990) Nature 346, 868-871. Monn.G.B. (1991) Nature 353, 454-456. Vollrath,D., Davies.R.W., Connelly.C. and Hieter,P. (1988) Proc. Natl. Acad. Sd. USA 85, 6027-6031. Pavan,W.J., Hieter,P. and Reeves,R.H. (1990) Proc. Natl. Acad. Sd. USA 87, 1300-1304. Farr,C., Fantes^l., Goodfellow.P. and Ccoke,H. (1991) Proc. Natl. Acad. Sci. USA 88, 7006-7010. Morin,G.B. (1989) Cell 59, 521-529. Cooke.H.J. and Smith.B.A. (1986) Cold Spring Harbor Symp. Quant. Biol. 51, 213-219. Allshire.R.C., Dempster.M. and Hastie.N.D. (1989) Nudeic Adds Res. 17, 4611-4627. de Lange.T., Shiue.L., Myers.R.M., Con.D.R., Naytor.S.L., Killery^.M. and Varmus.H.E. (1990) Mol. Cell. Biol. 10, 518-527. Hastie.N.D., Dempster,M., Dunlop.M.G., Thompson.A.M., Green,D.K. and Allshire.R.C. (1990) Nature 346, 866-868. LindseyJ., McGUI.N.I., Undsey.L.A., Green.D.K. and Cooke.HJ. (1991) Mutation Res. 256, 4 5 - 4 8 . Goldstein,S. (1990) Science 249, 1129-1133. Groves.A.K., Bogler.O., Jat.P.S. and NoWe.M. (1991) Curr. Opinion Cell Biol. 3, 224-229.
McCormkk.A. and CampisiJ. (1991) Curr. Opinion Cell Biol. 3, 230-234. Runge.K.W. and Zakian.V.A. (1989) Mol. Cell. Biol. 9, 1488-1497. Liu.Z. and Tye,B-K. (1991) Genes Dev. 5, 49-59. Hilliker.A.J. and Appels.R. (1989) Exp. Cell Res. 185, 287-318. Carmo-Fonseca.M., Pepperkok.R., Sproat.B.S., Ansorge.W., Swanson.M.S. and Lamond.A.I. (1991) EMBO J. 10: 1863-1873. Carmo-Fonseca.M., Tollervey.D., Pepperkok.R , Barabino.S M.L., Merdes.A., Brunner.C, Zamore.P.D., Green,M.R., Hurt.E. and Lamond.A.I. (1991) EMBO J. 10, 195-206. Zamore.P.D. and Green.M.R. (1991) EMBO J. 10, 207-214. Agard.D.A. and Sedat.J.W. (1983) Nature 302, 676-681. Mathog.D., Hochstrasser.M., Gruenbaum,Y., Saumweber.H. and Sedat.J. (1984) Nature 308, 414-421. Hochstrasser.M., Mathog.D., Gruenbaum,Y., Saumweber.H. and SedatJ.W.(1986)./. Cell Biol. 102, 112-123. Hiraoka.Y., Agard.D.A. and SedatJ.W. (1990) J. Cell Biol. I l l , 2815-2828. Blackbum.E.H. and Szostak J.W. (1984) Ann, Rev. Biochem. 53, 163-194. Wagenaar.E.B. (1969) Chromosoma 26, 410-426. Jackson.D.A. (1991) BioEssays 13, 1-10. Chung,H-M.M., Shea.C, Fields.S., Taub.R N. and Van der Ploeg.L.H.T. (1990) EMBOJ. 9, 2611-2619. Rawlins,D.J., Highett.M.I. and Shaw.P.J. (1991) Chromosoma 100, 424-431. Billia.F. and De Boni, U. (1991) J. Cell Sd. 100, 219-226. Walker,C.L., Cargile.C.B., Floy.K.M., Delannoy.M. and Migeon.B.R. (1991) Proc. Natl. Acad. Sci. USA 88, 6191-6195. Rasmussen.S.W. and Holm.P.B. (1980) Hereditas 93, 187-216. Laurie.D.A. and Hulten,M.A. (1985) Ann. Hum. Genet. 49, 203-214. Chandley.A.C. (1989) J. Med. Genetics 26, 546-552. Gottschling.D.E., Aparicio.O.M., Billington.B.L. and Zakian.V.A. (1990) Ce//63, 751-762. Apancio.O.M., BUlington.B.L. and Gottschling.D.E. (1991) Cell 66, 1279-1287. Nasmyth.K.A. (1982) Cell 30, 567-578. Strathern.J.N., Klar.A.J.S., Hicks.J.B., Abraham.J.A., Ivy.J.M., Nasmyth.K.A. and McGill.C. (1982) Cell 31, 183-192. Kostriken.R., StrathernJ.N., Klar,AJ.S , HicksJ.B. and Heffron.F. (1983) Ce//35, 167-174. Ferguson.B.M., Brewer.BJ., Reynolds.A.E. and Fangman.W.L. (1991) Cell 65, 507-515. Ten Hagen.K.G., Gilbert.D.M., Willard.H.F. and Cohen.S.N. (1990) Mol. Cell. Biol. 10, 6348-6355. Henikoff.S. (1990) Trends Genet. 6, 422-426.
Human Molecular Genetics, Vol. I, No. 1
3-6
HMG MINI-REVIEW Beginning or end? Telomere structure, genetics and biology David Kipling and Howard J.Cooke MRC Human Genetics Unit, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK Received January 3, 1992; Revised and Accepted February 4, 1992
ABSTRACT
MAMMALIAN TELOMERE STRUCTURE AND BEHAVIOUR The structure of mammalian telomeres It was the discovery of short repeated sequences at the ends of ciliate chromosomes that led to a search for similar sequences at the ends of human chromosomes. Allshire et ai, (6) demonstrated that the TTGGGGn terminal repeats from Tetrahymena would cross-hybridize to human telomeres. Subsequently both TTAGGGn and TTTAGGGn repeats (the latter being the terminal repeats of Arabidopsis) were also shown to cross-hybridize to human telomeres (7,8). These, and other, telomeric sequences can function in yeast, and a popular method of cloning human telomeres uses a modified yeast artificial chromosome vector bearing only a single functional telomere; the second telomere is provided by the repeats at the end of the human insert DNA, to which yeast terminal repeats are added (9, 10, 11, 12, 13). Subcloning and sequencing of the resulting artificial chromosomes demonstrated directly that human chromosomes terminate in TTAGGGn repeat arrays (9, 11). It seems likely from in situ hybridization data that this sequence is present at the end of all vertebrate chromosomes, some nonhuman ones also having large internal tracts of TTAGGGn, commonly in a pericentric or centric location (14, 15). Small regions of TTAGGGn have been cloned from interstitial sites in the human genome and a modified PCR protocol has also been used to clone TTAGGGn-adjacent human sequences (16, 17). The subtelomeric regions of human chromosomes (of which sequence data for the first few kilobases of a number of chromosomes is available) contain a complex mixture of repetitive sequences. A number of different repeats have been described, often present as short tandem repeats with differing periodicities, with 29, 37, 46 and 61bp repeats having been described (16,
18, 19, 20). Some of these repeats are very G+C rich; one 29bp repeat is over 85% G+C and behaves as a hypervariable minisatellite (18, 19, 20). These repeats are present in variable copy numbers at some telomeres and are entirely absent from others. The particular subset of telomeres detected by these repeats using in situ hybridization is polymorphic in the population (18, 19), possibly reflecting subtelomeric translocation between nonhomologous chromosomes, one specific example of which has been described recently (21). An additional family of subtelomeric repeated sequences, located at least 20kb from many chromosome ends, has also been described (22). An additional level of polymorphism has been found at mouse telomeres which are some 5 — 10 times larger than those of humans, enabling individual telomeres to be visualized as discrete terminal restriction fragments using pulsed-field gel electrophoresis (23, 24). The length of these fragments is polymorphic within a population of inbred mice, and can be accounted for by a high rate of generation of new size alleles, although the mechanism underlying this variation is not clear. One major finding of these studies is that only TTAGGGn is present at all human telomeres, implying that other repeats are not absolutely required for telomere function (18, 19, 20). In addition, the subtelomeric repeats only detect fragments in primate species on 'zoo' blots, in marked contrast to the presence of TTAGGGn at all vertebrate telomeres so far analysed (19). Further evidence in support of the hypothesis that the TTAGGGn repeats alone are sufficient for telomere function comes from a case of a thalassaemia with a terminal deletion of chromosome 16 (25). Sequence analysis shows that TTAGGGn repeats have been added directly at the breakpoint site, and in a subsequent study Morin has shown that human telomerase can recognize an oligonucleotide corresponding to this a thalassaemia breakpoint and utilize it as a substrate for TTAGGGn addition in vitro (26). Telomere-mediated chromosome breakage A major handicap in the analysis of mammalian telomere behaviour has been the lack of a method for altering telomere structure in a mammalian cell. In yeast the structure of endogenous chromosome ends can be altered by homologous recombination, or a linear minichromosome can be introduced terminating in any sequence desired, and its behaviour in mitosis and meiosis then followed (27, 28). Recent work by Farr et al. (29) suggests a similar approach is feasible in mammalian cells. A linear DNA molecule carrying a selectable marker and terminating in TTAGGGn repeats was introduced into a Chinese hamster cell line in culture and allowed
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
The word telomere derives from the Greek word telos meaning 'end', roughly translating as 'the thing at the end' when the end is that of a chromosome. Telomeres belie their apparent simplicity of structure by being involved in a wide range of diverse biological phenomena. Much of our understanding of telomere behaviour comes from studies in lower eukaryotes such as ciliates and yeast, the subject of many recent reviews (1, 2, 3, 4, 5). Here we concentrate on the mammalian telomere, recent progress in its study, and how recent evidence for an involvement of telomeres in the regulation of gene expression and DNA replication in yeast points to new aspects of mammalian telomere function yet to be explored.
4 Human Molecular Genetics, Vol. 1, No. 1 to integrate at random into the hamster genome. In 6 out of 27 transformants examined the integrated DNA is located at a telomere as defined by in situ hybridization and exonuclease sensitivity. The terminal restriction fragments are heterogeneous in size and larger than the introduced DNA; it seems likely that integration at an interstitial site has been accompanied by chromosome breakage, with the introduced TTAGGGn repeats conferring telomere function at the break site (perhaps by acting to prime TTAGGGn repeat addition by telomerase (30)).
Telomere-mediated chromosome breakage will be a powerful mapping tool, and the ability of cloned telomeres to function when reintroduced into mammalian cells is a significant step in the development of a mammalian artificial chromosome. Regulation of telomere length It has been known for some years that telomeres in human germline cells (eg. sperm) are longer than those in somatic tissue such as blood (6, 11, 31, 32, 33, 34). One proposed explanation for this is the absence of telomere repeat addition (ie. absence of telomerase activity) in somatic cells (31). If so, incomplete end replication would be expected to result in the progressive loss of terminal repeats as somatic cells undergo successive rounds of division. This is indeed what appears to happen in vivo for humans, with both blood (34) and skin (35) cells showing shorter telomeres with increasing donor age, and telomere loss may contribute to the chromosome aberrations typically seen in senescent cells. Senescence and the measurement of cellular time is an intriguingly complex subject (36, 37, 38) and it will be interesting to see to what extent telomere shortening has a causal role. The large telomeres possessed by both young and old mice (23, 24) would seem to preclude a simple relationship between telomere loss and ageing, but more elaborate schemes cannot be ruled out. The long telomeric repeat arrays of mice (23, 24) raised the question of how cells regulate telomere length. This is not as
FUTURE DIRECTIONS FOR MAMMALIAN TELOMERE RESEARCH Organization in the interphase nucleus The mitotic nucleus in interphase has a somewhat undistinguished appearance under the light or electron microscope, but there is considerable evidence suggesting that it is far from lacking in large-scale order (41). We are, for example, familiar with the concept of ribosome assembly being localized to specific regions of the nucleus (the nucleoli) and it seems that RNA splicing may be compartmentalized in an analogous fashion (42, 43,44). There are, of course, situations where discrete chromosomes are visible in mitotic interphase, such as the polytene chromosomes of Drosophila salivary gland nuclei. In this case the centromeres are seen clustered at one end of the nucleus, with the telomeres attached to the nuclear envelope at the other (45, 46, 47, 48), possibly reflecting chromosome orientation as it is at the end of mitosis. There is a large body of indirect evidence to suggest that there are attachments between telomeres, and between telomeres and the nuclear envelope (41, 49). However, much of the data does not directly concern mitotic interphase nuclei but instead uses
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
A human X chromosome with a terminal deletion of Xq. In situ hybridisation with an X specific or satellite probe and MIC2 sequences from the telomere construct used to produce the terminal deletion. Both the normal p-arm copy and the introduced telomeric q-arm copy of the sequences are labelled.
straightforward as might first be thought. A simple balance between a constant rate of terminal repeat addition (eg. by telomerase) and loss (for instance due to incomplete replication) is unstable over a long period of time; any alteration to either rate would result in steady growth or shrinkage that could continue unabated. It is necessary therefore that changes in telomere length must somehow feed back to the process of loss and/or addition and alter them accordingly. What might the cell be measuring as its criterion for 'telomere length'? One possibility would be the total TTAGGGn in the nucleus, perhaps by titration of some sequence-specific DNA binding protein. However, for an individual chromosome this mechanism would not provide stable length regulation. In addition, it would have to be exquisitely sensitive to enable control in species such as the Chinese hamster where there are large internal blocks of TTAGGGn and where loss of all the telomeric TTAGGGn would result in only a small percentage change in the total TTAGGGn level (15). An alternative possibility is suggested by results from Saccharomyces cerevisiae, where it has been shown that telomere length is remarkably stable in the face of changes in the amount of telomeric DNA in the nucleus. Introduction of a massive excess of yeast telomeres or telomeric DNA on replicating circular or linear plasmids has only a mild effect on average telomere length. For example, even a 25-fold excess of telomeres produces only a 50% increase in average telomere length (39). This leads to the interesting possibility that the cell is measuring the length of each individual telomere separately and regulating the number of terminal repeats added to each one accordingly. How might this be achieved? One possibility is that the recognition and binding of telomerase to a chromosome end is enhanced by an interaction with an accessory factors) bound at a fixed internal site on the chromosome. Telomeres with shorter terminal repeat arrays would have their free ends closer to this fixed 'reference point', and might be more favoured substrates for telomerase action. Support for this model comes from the recent isolation of a yeast protein which binds the junction between the subterminal X sequence and the TGj_3 repeats of a cloned yeast telomere (40) and is a candidate for such a reference point factor.
Human Molecular Genetics, Vol. 1, No. 1 5
Telomere behaviour in meiosis The pairing of homologous chromosomes is an essential prerequisite for meiotic recombination. However, although the cytological aspects of pairing in meiotic prophase have been extensively studied, the underlying molecular mechanisms are poorly understood. Pairing in human cells usually starts at sites very close to the telomeres, although with careful observation the contacts are seen to be slightly internal to the very tips of the chromosomes (56, 57). The three-dimensional organization of meiotic chromosomes can be reconstructed from electron micrographs of serial thin sections cut through meiotic nuclei. Applying this elegant if laborious technique to mammalian cells has shown that the telomeres are bound to the nuclear envelope by so-called attachment plaques (56). These sites of attachment are scattered around the nuclear envelope in the early stages of meiosis, but later on become clustered at one end in what is termed the bouquet formation. One consequence of this could be that the ends of the chromosomes are helped to align; sequences near telomeres are restricted to a search through a two-dimensional plane for homologous sequences with which to pair, whereas internal chromosomal regions are free to explore three-dimensional space. Therefore the preferential initiation of pairing near telomeres may be due to such pairings being favoured kinetically. Such initial alignments of chromosomes are often incorrect, and it has been proposed that this may contribute to the generation of new alleles in the germline (58). This would be consistent with the high level of variability of mammalian telomeres.
The influence of telomeres on transcription and DNA replication It is becoming evident from work in budding yeast that telomeres can influence both gene expression and DNA replication. Locating a reporter gene near a yeast telomere can result in its transcription being repressed. This repression is reversible, the gene switching between repressed and active states which are themselves inherited mitotically in a semi-stable fashion (59). The presence of the gene near TG]_3 repeats (the yeast telomeric sequence) is not by itself sufficient to cause repression, which occurs only if the repeat sequences are located at the end of a
linear molecule. The mechanism of telomeric repression may be similar to the transcriptional repression of the silent mating-type loci in yeast, as a number of mutations relieve repression at both sites (60). There is evidence for a distinct chromatin structure at the silent mating-type loci (61, 62, 63) and the chromatin at yeast telomeres may be in similar heterochromatin-like form. Heterochromatic silencing of a locus (such as expression of a gene or accessibility of a rra/is-acting factor to its binding site) within a defined distance of a telomere could also provide a possible mechanism for regulating telomere length. One hallmark of heterochromatic regions of mammalian chromosomes is that they generally replicate late in S phase. Recent work in yeast by Ferguson et al. (64) suggests that one determinant of replication timing is proximity to a telomere. A yeast origin of replication was identified which is activated late in S phase. The timing of firing of this origin appears not to be determined by its primary sequence but by its location near a telomere, as other origins moved near a telomere become late firing, and conversely this late origin fires early if moved to a nontelomeric location. As with transcriptional repression, proximity to TG]_3 repeats is insufficient to cause late initiation unless the repeats terminate a linear molecule. It is tempting to speculate that the heterochromatin-like structure at yeast telomeres responsible for transcriptional repression also results in delayed firing of adjacent origins of replication, and it will be interesting to determine if late replication can be relieved by the same mutations which relieve telomeric transcriptional repression. Although human telomeres replicate throughout the duration of S phase (65) it is not known to what extent they influence the timing of replication of adjacent genomic regions. This epigenetic control of gene expression in yeast may be relevant with respect to understanding metazoan phenomena such as position-effect variegation in Drosophila (66) and the regulation of telomeric genes in mammalian genomes.
SUMMARY AND PROSPECTS The last few years have seen a dramatic increase in our knowledge of mammalian telomere structure, stemming in part from new techniques developed for cloning these DNA sequences. However, various lines of evidence suggest that of all the telomeric and subtelomeric sequences described, only the terminal TTAGGGn repeats are essential for telomere function. One recurring theme from studies of these regions of the mammalian genome is a high rate of structural polymorphism and genetic instability. The underlying reasons remain a mystery, but may reflect the involvement of mammalian telomeres in the initiation of meiotic chromosome pairing that precedes recombination. A significant new tool for the analysis of mammalian telomere structure is- the ability to functionally reintroduce cloned telomeres into a mammalian cell. Coupled with the established technologies of targeted recombination and transgenesis in the mouse we may be close to manipulating mammalian telomere structure in vivo. There are many aspects of telomere biology of which we as yet have only the most fragmentary understanding, such as the relationship between telomeres and the higher order structure of the nucleus. Meanwhile, work in other species continues to uncover new and unexpected aspects of telomere behaviour. We can therefore be confident that the end is by no means in sight for the investigation of the mammalian telomere.
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
polytene or meiotic chromosomes, or instances where telomeretelomere interactions are so strong that they are maintained in squash preparations of prophase nuclei (50). Telomere-telomere interactions, together with attachments to the nuclear envelope, would be expected to contribute to chromosome organization in the nucleus, with implications for processes such as gene expression (51). Telomere distribution in the mammalian mitotic interphase nucleus may be resolved by a combination of in situ hybridization and optical sectioning techniques, as have been successfully applied to trypanosomes and plants (52, 53). Indeed, such techniques have been used recently to study the distribution of telomeres in murine neuronal cells in culture, where they were found not to be associated with the nuclear envelope but distributed throughout the volume of the interphase nucleus (54). In another study the inactive human X chromosome has been shown to form itself into a loop via telomere-telomere association near the nuclear envelope, in contrast to a linear orientation for the active X (55).
6 Human Molecular Genetics, Vol. 1, No. 1 ACKNOWLEDGEMENTS We would like to thank many colleagues in the MRC Unit for stimulating discusssions and comments on the manuscript. D.K. is funded by a grant from the UK MedicaJ Research Council's Human Genome Mapping Project.
38. 39. 40. 41. 42.
REFERENCES
43.
1. 2. 3. 4. 5. 6. 7.
11. 12. 13. 14. 15.
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.
Downloaded from http://hmg.oxfordjournals.org/ at University of Guelph on May 28, 2012
8. 9. 10.
Zakian.V.A. (1989) Ann. Rev. Genet. 23, 579-604. Greider,C.W. (1990) BioEssays 12, 363-369. Zakian.V.A., Runge.K. and Wang,S-S. (1990) Trends Genet. 6, 12-16. Blackburn.E.H. (1991) Nature 350, 569-573. Greider.C.W. (1991) Curr. Opinion Cell Biol. 3, 444-451. Allshire.R.C, Gosden.J.R., Cross,S.H., Cranston.G., Rout.D., Sugawara.N., SzostakJ.W., Fantes.P.A. and Hastie.N.D. (1988) Nanire 332, 656-659. Moyzis.R.K., BuckinghamJ.M., Cram.L.S., Dani.M., Deaven,L.L., Jones.M.D., MeyneJ., Ratliff.R.L. and WuJ-R. (1988) Proc. Nail. Acad. Sd. USA 85, 6622-6626. Richards.E.J. and Ausubel.F.M. (1988) Cell 53, 127-136. Brown.W.R.A. (1989) Nature 338, 774-776. Cheng,J-F., Smith.C.L. and Cantor.C.R. (1989) Nucleic Acids Res. 17, 6109-6127. Cross,S.H., Allshire.R.C., McKay,SJ., McGill.N.I. and Cooke.HJ. (1989) Nature 338,771 -774. RJethman.H.C, Moyzis.R.K., MeyneJ., Burke.D.T. and01son,M.V. (1989) Proc. Natl. Acad. Sd. USA 86, 6240-6244. Bates.G.P., MacDonald.M.E., Baxendale.S., Sedlacek.Z., Youngman,S., Romano.D., Whaley.W.L., Allitto,B.A., Poustka,A., Gusella.J.F. and Lehrach.H. (1990) Am. J. Hum. Genet. 46, 762-775. MeyneJ., Ratliff,R.L. and Moyzis.R.K. (1989) Proc. Natl. Acad. Sd. USA 86, 7049-7053. MeyneJ., Baker,R.J., Hobart.H.H., Hsu.T.C, Ryder,O.A., Ward,O.G., WileyJ.E., Wurster-HUl.D.H., Yates.T.L. and Moyzis.R.K. (1990) Chromosoma 99, 3 — 10. Weber.B., CoUins.C, Robtxns.C, Magenis.R.E., Delaney.A.D., Gray.J.W. and Hayden.M.R. (1990) Nudeic Adds Res. 18, 3353-3361. Wells.R.A., Germino.G.G., Krishna.S., Buckle.V.J. and Reeders.S.T. (1990) Genomics 8, 699-704 Brown,W.R.A., MacKinnon.PJ., Villasante,A., Spurr.N., Buckle.V.J. and Dobson,M.J. (1990) Cell 63, 119-132. Cross.S., UndseyJ., FantesJ., McKay.S., McGill.N. and Cooke,H. (1990) Nucleic Adds Res. 18, 6649-6657. ChengJ-F., Smith.C.L. and Cantor.C.R. (1991) Nucleic Acids Res. 19, 149-154. Wilkie.A.O.M., Higgs.D.R., Rack.K.A., Buckle.V.J., Spurr.N.K., HschelGhodsian.N., Ceccherini.I., Brown.W.R.A. and Harris.P.C. (1991) Cell 64, 595-606. Rouyer.F., de la Chapelle.A., Andersson.M. and WeissenbachJ. (1990) EMBOJ. 9, 505-514. Kipling.D., and Cooke.H.J. (1990) Nature 347, 400-402. StarlingJ.A., MauleJ., Hastie.N.D. and Allshire.R.C. (1990) Nucleic Adds Res. 18, 6881-6888. Wilkie.A.O.M., LambJ., Harris.P.C., Rnney.R.D. and Higgs,D.R. (1990) Nature 346, 868-871. Monn.G.B. (1991) Nature 353, 454-456. Vollrath,D., Davies.R.W., Connelly.C. and Hieter,P. (1988) Proc. Natl. Acad. Sd. USA 85, 6027-6031. Pavan,W.J., Hieter,P. and Reeves,R.H. (1990) Proc. Natl. Acad. Sd. USA 87, 1300-1304. Farr,C., Fantes^l., Goodfellow.P. and Ccoke,H. (1991) Proc. Natl. Acad. Sci. USA 88, 7006-7010. Morin,G.B. (1989) Cell 59, 521-529. Cooke.H.J. and Smith.B.A. (1986) Cold Spring Harbor Symp. Quant. Biol. 51, 213-219. Allshire.R.C., Dempster.M. and Hastie.N.D. (1989) Nudeic Adds Res. 17, 4611-4627. de Lange.T., Shiue.L., Myers.R.M., Con.D.R., Naytor.S.L., Killery^.M. and Varmus.H.E. (1990) Mol. Cell. Biol. 10, 518-527. Hastie.N.D., Dempster,M., Dunlop.M.G., Thompson.A.M., Green,D.K. and Allshire.R.C. (1990) Nature 346, 866-868. LindseyJ., McGUI.N.I., Undsey.L.A., Green.D.K. and Cooke.HJ. (1991) Mutation Res. 256, 4 5 - 4 8 . Goldstein,S. (1990) Science 249, 1129-1133. Groves.A.K., Bogler.O., Jat.P.S. and NoWe.M. (1991) Curr. Opinion Cell Biol. 3, 224-229.
McCormkk.A. and CampisiJ. (1991) Curr. Opinion Cell Biol. 3, 230-234. Runge.K.W. and Zakian.V.A. (1989) Mol. Cell. Biol. 9, 1488-1497. Liu.Z. and Tye,B-K. (1991) Genes Dev. 5, 49-59. Hilliker.A.J. and Appels.R. (1989) Exp. Cell Res. 185, 287-318. Carmo-Fonseca.M., Pepperkok.R., Sproat.B.S., Ansorge.W., Swanson.M.S. and Lamond.A.I. (1991) EMBO J. 10: 1863-1873. Carmo-Fonseca.M., Tollervey.D., Pepperkok.R , Barabino.S M.L., Merdes.A., Brunner.C, Zamore.P.D., Green,M.R., Hurt.E. and Lamond.A.I. (1991) EMBO J. 10, 195-206. Zamore.P.D. and Green.M.R. (1991) EMBO J. 10, 207-214. Agard.D.A. and Sedat.J.W. (1983) Nature 302, 676-681. Mathog.D., Hochstrasser.M., Gruenbaum,Y., Saumweber.H. and Sedat.J. (1984) Nature 308, 414-421. Hochstrasser.M., Mathog.D., Gruenbaum,Y., Saumweber.H. and SedatJ.W.(1986)./. Cell Biol. 102, 112-123. Hiraoka.Y., Agard.D.A. and SedatJ.W. (1990) J. Cell Biol. I l l , 2815-2828. Blackbum.E.H. and Szostak J.W. (1984) Ann, Rev. Biochem. 53, 163-194. Wagenaar.E.B. (1969) Chromosoma 26, 410-426. Jackson.D.A. (1991) BioEssays 13, 1-10. Chung,H-M.M., Shea.C, Fields.S., Taub.R N. and Van der Ploeg.L.H.T. (1990) EMBOJ. 9, 2611-2619. Rawlins,D.J., Highett.M.I. and Shaw.P.J. (1991) Chromosoma 100, 424-431. Billia.F. and De Boni, U. (1991) J. Cell Sd. 100, 219-226. Walker,C.L., Cargile.C.B., Floy.K.M., Delannoy.M. and Migeon.B.R. (1991) Proc. Natl. Acad. Sci. USA 88, 6191-6195. Rasmussen.S.W. and Holm.P.B. (1980) Hereditas 93, 187-216. Laurie.D.A. and Hulten,M.A. (1985) Ann. Hum. Genet. 49, 203-214. Chandley.A.C. (1989) J. Med. Genetics 26, 546-552. Gottschling.D.E., Aparicio.O.M., Billington.B.L. and Zakian.V.A. (1990) Ce//63, 751-762. Apancio.O.M., BUlington.B.L. and Gottschling.D.E. (1991) Cell 66, 1279-1287. Nasmyth.K.A. (1982) Cell 30, 567-578. Strathern.J.N., Klar.A.J.S., Hicks.J.B., Abraham.J.A., Ivy.J.M., Nasmyth.K.A. and McGill.C. (1982) Cell 31, 183-192. Kostriken.R., StrathernJ.N., Klar,AJ.S , HicksJ.B. and Heffron.F. (1983) Ce//35, 167-174. Ferguson.B.M., Brewer.BJ., Reynolds.A.E. and Fangman.W.L. (1991) Cell 65, 507-515. Ten Hagen.K.G., Gilbert.D.M., Willard.H.F. and Cohen.S.N. (1990) Mol. Cell. Biol. 10, 6348-6355. Henikoff.S. (1990) Trends Genet. 6, 422-426.
