sign in sign up
<<
Home , Elizabeth Blackburn, Telomerase, Telomere, Transferase

COMMENTARY

Telomeres and : the path from maize, and to and aging

Elizabeth H Blackburn, Carol W Greider & Jack W Szostak

The problem research report that described the frequent join- the 1970s, Joe Gall had been delving into the Scientific discoveries are each individual and ing of broken ends, “no case was processes by which some organisms produce

http://www.nature.com/naturemedicine occur by their own unique path. However, there found of the attachment of a piece of one chro- extra copies of the for ribosomal RNA are key ingredients that set the stage for them. mosome to the end of another [intact chromo- (rRNA). This occurs, for example, early in Many of these ingredients were important in some]”1. In 1938, Muller named the natural ends development, when large amounts of the discovery of telomerase: talking with sci- of ‘’2. But neither Muller synthesis must occur rapidly. Joe had discovered entists from different fields, paying attention to nor McClintock had the tools to understand the that the genes encoding rDNA are amplified unusual findings and taking the risks of doing molecular nature of these chromosome ends. on circular DNA that are present in crazy experiments. We will describe how a com- The question of the molecular nature of the very high numbers in the developing oocyte in bination of these ingredients and productive chromosome end only became meaningful in frogs. He then found that the same thing hap- collaborations led us to postulate and discover 1953, when the structure of DNA was described3. pened in a very different organism—the cili- telomerase. By the 1960s, had discovered ated protozoan Tetrahymena thermophila—but The earliest functional description of telo- DNA and its mechanism had been this time, the rDNA was amplified into linear meres was by geneticist Hermann Muller when determined4. This understanding posed yet DNA molecules. Tetrahymena contained large he used X-rays to fragment chromosomes. another question about DNA ends—how was numbers of nearly identical, relatively short Nature Publishing Group Group 200 6 Nature Publishing Muller, working with fruit flies, and Barbara their complete replication ensured? Because . This was the material that © McClintock, working with maize, converged on DNA polymerase could only extend a preformed Liz decided to use to analyze natural ends of the same conclusion around the same time: that primer, it could not copy the very end of a linear chromosomes. the natural ends of chromosomes are different DNA; this became known as the DNA end-rep- There was no road map for how to do this. from those created at the site of a chromosomal lication problem5. By the early 1970s, studies of But Joe Gall had already shown that a fraction of break. The natural ends were somehow protected DNA had shown that the molecules were circular when extracted from from the frequent rearrangements that occur at the answers to the DNA end-replication problem cells, a property reminiscent of phage lambda, broken ends. As McClintock wrote in a 1931 differed between one virus and another6. How, the linear DNA of which circularizes to replicate. then, was the DNA at the very end of eukaryotic Because Liz had grown to believe that nature uses Elizabeth H. Blackburn is at the Department chromosomes arranged? In 1975, Liz Blackburn elegant and universal solutions, she thought that of and Biophysics, University of arrived at Yale in Joe Gall’s lab to do postdoctoral the lambda ends might be a possible model for California, , 600 16th Street, research, having recently completed her gradu- the molecular nature of the Tetrahymena termini. MC 2200, San Francisco, California 94158-2517, ate work in Fred Sanger’s group in Cambridge, Thus, she decided to use the DNA ‘repair’ USA. Carol W. Greider is Professor England, where DNA sequencing was being reaction of DNA polymerase, which had been and Director, Department of Molecular invented. Liz wanted to apply her knowledge successfully used by Ray Wu and colleagues to and , Johns Hopkins University School from the Sanger lab to understanding the molec- sequence the cohesive ends of the lambda phage of Medicine, 603 PCTB, 725 North Wolfe Street, ular nature of chromosome termini. family genomic DNAs7. Baltimore, Maryland 21205, USA. Jack W. Szostak This was a fortunate choice, because the is Professor of Genetics at Harvard Medical School, Telomeres go molecular: mysterious DNA molecular ends of the rDNA turned out to Department of and Center for termini have discontinuities (the significance of which Computational and Integrative Biology, Simches One of the most daunting aspects of addressing remains mysterious to this day) within the Research Center, 85 Cambridge Street, Boston, the question of the DNA at chromosomal ends telomeric repeat tract DNA that allowed DNA Massachusetts 02114, USA. was the enormous length of the chromosomal polymerase to label them readily in vitro using e-mails: [email protected], of . DNA methods radiolabeled triphosphate substrates. Liz was [email protected] or [email protected]. had not yet been invented, so to be able to study then able to piece together the DNA sequence harvard.edu. the ends, short chromosomes were needed. In of the telomeres by combining a variety of

NATURE MEDICINE VOLUME 12 | NUMBER 10 | OCTOBER 2006 1133 COMMENTARY

in vitro labeling and other analytical techniques, ciliates16,17. The question then became: how mid, resulting in a linear replicating with and she came up with a very un-lambda-like was the telomere repeat added? stable DNA ends. The experiment was technically surprise. The rDNA end In considering telomere formation in simple to perform, and the prediction—that lin- sequence and the structure at the termini of Tetrahymena, Liz wrote in 1982: “...the sequences ear plasmid molecules would be seen instead of these molecules were complex and unlike common to the macronuclear DNA termini the usual circles—would be trivial to confirm. any previously described. At each end of the must be acquired by these subchromosomal Armed with a purified Tetrahymena telomeric Tetrahymena rDNA molecules there were segments during their formation. Two types of DNA fragment supplied by Liz, Jack generated around 50 tandem repeats of the hexanucleo- routes can be envisaged: Telomeric sequences a few nanograms of a linearized yeast plasmid tide unit CCCCAA—TTGGGG on the comple- are transposed or recombined onto the devel- capped with Tetrahymena telomeres, and intro- mentary strand—with the latter (G-rich) strand oping macronuclear DNA termini, or the simple, duced the DNA into yeast cells. He obtained a bearing the 3ʹ OH end at each end of the linear repeating telomeric sequences are synthesized dozen or so transformants that were analyzed DNA. The C-rich CCCCAA-repeat strand had de novo onto these termini by specific synthetic by Southern blotting. The result was immedi- single-stranded discontinuities in at least a por- machinery”13. This idea for de novo telomere ately clear: over half of the colonies maintained tion of the repeat array. Oddly, too, the number addition was further supported by ongoing the introduced DNA in linear form. This result of tandem repeats per end was heterogeneous experiments in yeast. was confirmed by more detailed DNA analysis19. among the population of purified molecules, This result showed that telomeres could function ranging from an estimated minimum of 20 to Telomere function transcends kingdom across phylogenetic kingdoms, implying remark- up to about 70 (ref. 8). Late in 1977, Liz gave boundaries able functional conservation. a talk describing the unusual molecular fea- By 1980 we understood the molecular struc- tures of the rDNA ends at the University of ture of Tetrahymena telomeres, but the con- Yeast telomeres reveal conservation of California, San Francisco to Herbert Boyer’s nection between this structure and the special telomere structure group in the Biochemistry Department. In the properties of telomeres remained obscure. These linear then provided the

http://www.nature.com/naturemedicine discussion after her presentation, a member of Whether this surprising structure was unique ideal vector for cloning yeast telomeres. Jack the Boyer lab asked whether the heterogeneity to Tetrahymena and its relatives or was removed one Tetrahymena telomere, generat- in the number of CCCCAA repeats in the DNA more broadly conserved was also unknown. ing a linear DNA fragment that could not be population might arise by addition of repeats The answers to these questions began to emerge maintained in yeast. He then went fishing for to the chromosome ends. Liz was intrigued but from an unlikely collaboration between Jack functional yeast telomeres by joining random at the time could see no known way this could Szostak and Liz. Jack had recently completed pieces of yeast genomic DNA onto this linear occur. Looking back, this conversation seems his graduate and postdoctoral work with Ray DNA fragment; only when the missing telo- prescient, as later Liz would find that, in fact, Wu at Cornell, where he had begun to study mere was replaced with a yeast telomere could this addition does occur. recombination in yeast. In 1979, he had set up the DNA survive in yeast as a linear plasmid. his own lab at what was then the Sidney Farber Three of the expected linear plasmids were New telomeres are added to fragmented Cancer Institute in Boston, and was investigat- recovered, and Southern blots of genomic Tetrahymena chromosomes ing the highly recombinogenic nature of DNA yeast DNA showed that the linear plasmids In collaboration with Meng-Chao Yao in Joe ends in the budding yeast Saccharomyces cere- indeed carried a functional yeast telomere Nature Publishing Group Group 200 6 Nature Publishing Gall’s lab, Liz showed that the same termi- visiae. At that time, plasmid vectors for yeast (Fig. 1). With yeast telomeres in hand, it was © nal, heterogeneous array of CCCCAA repeats transformation were all maintained as circu- now possible to study the structure of telo- occurred at the ends of the other chromosomal lar DNA molecules. Furthermore, lineariza- meres of a eukaryotic chromosome capable of DNA molecules of the nucleus (the tion of these plasmids by restriction proper mitotic and meiotic segregation. macronucleus) in Tetrahymena, but that these digestion led to DNA ends that were extremely Closer examination of the initial linear plas- sequences were not present in the precursor reactive inside yeast cells: if the DNA ends were mids revealed that the Tetrahymena telomeres DNAs in the nucleus (the micronu- homologous to yeast DNA, recombination had become longer and more heterogeneous in cleus) from which the somatic nucleus is gener- would result in integration of the plasmid into length during their maintenance in yeast. Janis ated9,10. Soon after, the telomeric sequences of the chromosome18; otherwise, the DNA ter- Shampay in Liz’s lab sequenced the subcloned the linear rDNA minichromosomes from the mini would be degraded, ligated or otherwise yeast telomeres and the Tetrahymena telomeres slime molds Physarum11 and Dictyostelium12 rearranged. Liz and Jack’s collaboration began maintained in yeast and found that yeast-spe- were also determined. Liz continued her work with an intense conversation in the summer cific TG1–3 repeats had been added onto the on ciliate DNA termini when she set up her of 1980 at a New Hampshire school, the site Tetrahymena TTGGGG repeats20. Liz labeled own lab in the Molecular Biology Department of that year’s Gordon Research Conference the yeast telomeres in vitro and found that they of the University of California, Berkeley in on Nucleic Acids. After hearing Liz’s descrip- also had single-stranded discontinuities in the 1978. There she found that new telomere tion of the remarkable molecular biology of C-rich DNA strand, as seen in Tetrahymena sequences were added to the ends of the lin- Tetrahymena telomeres, Jack asked Liz about telomeres. Janice then showed that yeast chro- ear rDNA minichromosomes by an unknown testing whether Tetrahymena telomeres might mosomal telomeres also consisted of a variable 13 mechanism . These subchromosomal macro- function in yeast. The idea was so far-fetched number of terminal TG1–3 repeats. Together nuclear DNAs are generated by developmen- that it seemed outlandish: to test whether the these results indicated that these telomeres tally controlled DNA fragmentation of the telomere replication mechanisms are conserved were very similar in structure to the archetypal germline nuclear chromosomes. Surprisingly, between such evolutionarily distant species. telomeres of Tetrahymena. The length hetero- there was no invariant DNA sequence to which Liz and Jack reasoned that if Tetrahymena geneity of telomeric fragments reflected vari- the telomeric repeats were joined13–15. At telomeres retained their ability to stabilize DNA able amounts of DNA at the very ends of the around the same time, David Prescott’s group ends when transferred into yeast cells, they might plasmid or chromosome. These results led Jack obtained similar results in a different group of be used to cap the ends of a linearized yeast plas- and Liz to propose the existence of a terminal

1134 VOLUME 12 | NUMBER 10 | OCTOBER 2006 NATURE MEDICINE COMMENTARY

Ori Artificial chromosomes were subsequently used tionated extract containing radiolabeled to very long DNA fragments in yeast, and dGTP. When the products were fractionated by they became an essential tool early in the analy- size on a sequencing gel, a ladder of repeats with sis of the human . a six-base periodicity extended up the gel26. We Marker now know that this experiment worked because Some unusual goings-on at the ends the telomeric substrate was pres- The addition of yeast sequences directly onto ent in a very high concentration compared with Linear plasmid the Tetrahymena telomeres on the linear plas- that in the earlier experiments using restriction Add Tetrahymena telomeres mid in yeast, the de novo telomere addition in fragments. Transform yeast Tetrahymena and the length heterogeneity of telomeres indicated that an unusual process Yeast telomeres in Tetrahymena extracts: might be involved in telomere maintenance. proof comes full circle Maintain linear In addition, the curious fact that telomeres This first hint of an activity that generated a six- plasmid in yeast grew longer when trypanosomes were grown base repeating pattern was very exciting. Carol in culture also indicated something unusual and Liz then set out to determine whether this was going on at the ends24. By 1983, there repeating pattern was due to a new enzyme or Remove right telomeres were two competing models for the molecular to an established DNA polymerase that might Add yeast genomic DNA mechanism that generated the telomere length be using the input oligonucleotide and copy- Yeast telomere stabilizes linear plasmid heterogeneity and maintained the telomere ing endogenous telomere repeats in the cell length. One model proposed that there was an extract. Carol used different approaches to rule unknown enzyme that synthesized the ends20. out endogenous DNA as a template. First, she The second model proposed that the addition showed that the elongation activity was specific Tetrahymena TTGGGG repeat sequence http://www.nature.com/naturemedicine of telomere repeats occurred through a recom- to telomeric sequences, as an oligonucleotide 24,25 Yeast TG1–3 repeat sequence bination-mediated process . Recombination with an unrelated sequence was not extended was a well established process and the mecha- in the reaction. The key experiment, however, Figure 1 Yeast sequences are added to nism proposed for telomere repeat addition came when Carol and Liz decided to do the Tetrahymena telomeres in vivo. The cross- seemed plausible. To distinguish between these inverse of the Szostak and Blackburn cross- kingdom experiment that showed telomere mechanisms, direct experimental evidence was kingdom telomere-function experiment19: this function is conserved, and it allowed the cloning needed. time in Tetrahymena cell extracts. They used a of yeast telomeres shown in the diagram. A circular yeast plasmid with an synthetic oligonucleotide consisting of the yeast (Ori, black) and a selectable marker (green) was The discovery of telomerase telomeric sequence repeat. This 24-base oligo- linearized and Tetrahymena telomeres (blue) The biochemical evidence for telomerase came had the sequence TGGG at its 3ʹ end, were ligated onto the ends. When this plasmid from a series of experiments carried out in Liz but was otherwise not related in sequence to the was transformed into and grown in yeast, the Blackburn’s lab by Carol Greider. Carol joined Tetrahymena repeats. The experiment showed plasmid remained linear but yeast telomere Liz’s lab as a graduate student in 1984, and she that not only was this yeast telomere efficiently Nature Publishing Group Group 200 6 Nature Publishing sequence (red) was added to the end of the set out to study how telomeres replicated and to elongated in the extracts, but the pattern of the © Tetrahymena telomere repeats. This plasmid was extracted from yeast, and the right end was investigate what caused the sequence additions repeats on the gel was shifted up by one base removed and replaced by fragments of yeast DNA. seen in Tetrahymena, yeast and trypanosomes. (Fig. 2a). This was the exciting breakthrough A yeast telomere (red) captured by this method After preliminary experiments done by Liz and that finally convinced Carol and Liz that the represented the first cloned telomere. her graduate student Peter Challoner, Carol activity was indeed a unique telomere-synthe- initiated biochemical experiments to look for sizing enzyme. The one-base shift in the band- –like enzyme that would add repeat evidence for a telomere-addition enzyme. Carol ing pattern was due to the correct synthesis of units to telomeric DNA as a way of compensat- began by adding DNA restriction fragments the TTGGGG sequences by the Tetrahymena ing for the erosion caused by incomplete termi- and radiolabeled to Tetrahymena enzyme (Fig. 2b). As the Tetrahymena substrate 20 nal replication . cell extracts to determine whether a telomere (TTGGGG)3 has 4 G’s at the end, and the yeast The availability of cloned telomeric frag- sequence on one end of the restriction frag- oligonucleotide had only 3 G’s, it had to first be ments allowed Andrew Murray, then a graduate ment would become preferentially labeled. extended by an extra G before the TTGGGG student in Jack’s lab, to begin construction of These experiments initially showed promising repeat pattern could be continued. This resulted the first artificial chromosomes. Andrew com- results. However, closer examination showed in an upward shift in the entire banding pat- bined , replication origins, genetic that the labeling was due to known DNA poly- tern—compelling evidence that the activity was markers and telomeres on a single yeast plasmid. merase activities. Carol therefore tried a variety a specific telomere terminal transferase activ- Surprisingly, this and subsequent work showed of different substrates and reaction conditions ity26. In a later paper, the name of the enzyme that this full complement of known chromo- to determine whether there was an activity that was shortened to ‘telomerase’27. somal elements was insufficient for proper would act only on telomere sequences. mitotic segregation, and that a minimum length The breakthrough came in December of 1984, Telomerase: how is the sequence of DNA was also required. This initial work on when she used a synthetic DNA oligonucleotide specified? artificial chromosomes21 set the stage for later as a substrate in the reaction. This synthetic oli- Having a newly discovered telomere-synthesiz- studies of highly engineered chromosomes gonucleotide represented a different molecular ing activity in hand, the next most important as a path toward understanding the cellular substrate that could also be added in higher con- question was how the telomeric sequence is mechanisms that ensure accurate chromosomal centration than a restriction fragment. She added specified. One model was that there might be 22,23 inheritance, in both and . the DNA oligonucleotide (TTGGGG)4 to unfrac- a nucleic acid component to specify telomere

NATURE MEDICINE VOLUME 12 | NUMBER 10 | OCTOBER 2006 1135 COMMENTARY

(CCCCAA) Yeast 4 pBR repeats. Indeed, treatment of the extract with telomerase or had some other role in telomere RNase abolished the activity. A complex series maintenance. (TTGGGG)4 No oligo a of controls eventually showed that this loss of In characterizing the Tetrahymena telomerase activity was not an artifact: the enzyme con- RNA mutants, Liz found that one mutant had tained an essential RNA component27. However, a phenotype that paralleled the EST1 mutant finding the specific RNA was trickier. Carol set of yeast29. This template somehow out to purify telomerase and sequence the blocked all DNA addition onto telomeric ends that copurified with the enzyme activity. Several in vivo. Instead, the telomeres shortened progres- years of cold-room work led to highly purified sively, and the cells grew for about 20–25 more fractions, but the RNA remained elusive. After cell divisions, and then they senesced (stopped moving to an independent position at the Cold dividing)29. This result helped support the idea Spring Harbor Laboratory, Carol was able to use that the EST1 mutant in yeast might indeed be partial sequence information from copurifying a telomerase mutant, as the phenotype was so RNAs to clone the RNA component of telomer- similar to that of this Tetrahymena telomerase ase. The appearance on a sequencing gel of the mutant. Together, these experiments established sequence CAACCCCAA in one clone imme- that a functional telomerase is necessary for diately heralded success. Carol then set out to the indefinite replicative capacity of yeast and test whether this RNA was required for enzyme Tetrahymena. activity. She accumulated evidence that not only Biochemical identification of the yeast telom- did the RNA copurify with telomerase, but oli- erase activity came from the work of postdoc- gonucleotides that blocked the CAACCCCAA toral fellows Marita Cohn and John Prescott proposed template region also blocked enzyme in Liz’s lab31,32. An initially puzzling result was

, 405–413 (1985). 28 http://www.nature.com/naturemedicine activity . The presence of nine nucleotides in that this in vitro activity did not require EST1 43 the template region then implied a mechanism (refs. 31,33). However, subsequent biochemical Cell for how telomerase might synthesize tandem evidence showed that the Est1 protein is indeed

Input: TTGGGG repeats. The model, presented in a component of the yeast telomerase complex, 1989, is still the accepted mechanism for telom- although it is not the catalytic component34. pBRÐ erase action28. The EST2 , which was later identified Carol sent the cloned RNA gene back to Liz by Vicki Lundblad35, turned out to encode (TTGGGG) 4 Input Yeast at Berkeley. Guo-Liang Yu and John Bradley, the catalytic subunit. EST2 was shown to be graduate students in Liz’s lab, made homologous to a ciliate telomerase component in the template. Guo-Liang overexpressed the purified and identified by and (CCCCAA) Ð 4 mutant telomerase RNA in Tetrahymena cells Tom Cech, leading the three of them to show

1234 5 6 789 by microinjecting the genes into Tetrahymena that EST2 encodes the catalytic protein com- Reprinted from Greider, C.W. & Blackburn, E.H. C.W. Reprinted from Greider, using a new vector system he had devised. He ponent of telomerase—the telomerase reverse Nature Publishing Group Group 200 6 Nature Publishing found that the telomeres in these cells now incor- transcriptase (TERT)36. © b porated the altered telomere repeat sequence. GGTTGGGGTTGGGGttGgggttGgggttGggg Primer Tetrahymena This showed definitively that telomerase uses Telomeres in cancer and cell renewal oligonucleotide the CAACCCCAA template sequence to specify The discovery of telomerase in Tetrahymena TGGGTGTGTGTGGGgttGgggttGgggttGggg the addition of telomeric repeats and therefore and yeast was pure curiosity-driven research, Primer yeast established its mode of with no obvious medical impact. That soon oligonucleotide action in vivo as well as in vitro29. changed. Although the sequence of the human Figure 2 Tetrahymena sequences are added to telomere took longer to identify, there was early yeast telomeres in vitro. (a) The autoradiogram Evidence for telomerase in yeast evidence that it had a similar structure37–39 . In shows the addition of Tetrahymena telomere repeats onto either a Tetrahymena telomere In parallel with the biochemical effort to identify 1988, the sequence of the human telomere was sequence primer (lane 5) or a yeast telomere the components of telomerase in Tetrahymena, identified as simple repeats of TTAGGG, very sequence primer (lane 6). In both cases, a 6-base Vicki Lundblad in Jack’s lab initiated a genetic similar to the Tetrahymena repeats38,39. This is added, but with the yeast screen in yeast for mutants defective in telomere and the identification of human telomerase40 primer the pattern is shifted up by one nucleotide maintenance30. The most interesting mutant sparked research in two areas: cellular senes- (arrows). Lanes 1–4 show the size of the input had the phenotype predicted for one defective cence and cancer. Alexei Olovnikov had pre- primers, and lanes 7–9 show that with unrelated sequences are not substrates for in telomere maintenance—a continuous short- dicted that the end-replication problem would addition. (b) Diagram representing the sequence ening of the telomeric DNA. Because the telo- result in telomere shortening that might be the of the Tetrahymena primer (top, blue) and the meres in this strain became shorter and shorter cause of the limited potential of yeast primer (bottom, blue). The sequence added over time, the gene was named EST1 for ‘ever primary fibroblasts in culture—cellular senes- onto the primer is dictated by the sequence at the shorter telomeres.’ Most gratifyingly, this strain cence41. Carol, in collaboration with Calvin 3ʹ end of the primer. Thus, the banding pattern is also showed a delayed phenotype, in Harley, then found that telomere shortening shifted up by one nucleotide for the yeast primer which many generations of normal growth were does indeed occur in human cells in culture42. because a G must be added to complete a set of GGGG before the TT sequence follows. This extra followed by an increase in chromosome loss and This is due to a lack of telomerase expression G sets the phase for all the other repeats that are a significant loss of growth potential, as predicted and a consequent inability to maintain normal added, shifting the entire pattern for every repeat from the gradual loss of telomeres. It was not telomeres. If telomerase is expressed in these that follows (arrows). known at the time whether EST1 was related to cells, telomeres do not shorten and the cells do

1136 VOLUME 12 | NUMBER 10 | OCTOBER 2006 NATURE MEDICINE COMMENTARY

not show senescence43. Thus, short telomeres How much telomerase is enough? loss of tissue renewal capacity58. This implies that can limit the ability of cells to divide, indicat- The continuing biochemical studies of telomer- the remarkable ability of telomerase to recognize ing that telomerase inhibition might limit the ase throughout the 1990s led to an entirely new and efficiently elongate telomeres is strictly lim- growth of cancer cells42. discovery: short telomeres have a role in genetic ited in the cell. Experiments aimed at under- This idea of a role for telomeres in cancer was . The structure of the human telomer- standing the mechanisms that limit telomere pursued independently by two independent ase RNA component49 contains a motif that is elongation and at understanding why telomer- groups. Nick Hastie and both present in an unrelated class of small RNAs, the ase activity is so tightly regulated in mammals showed that telomeres were shorter in cancer snoRNAs50. This motif binds to a protein termed promise more exciting discoveries in telomere cells than in normal tissue44,45. This shorten- that functions in ribosomal RNA modi- research in the coming years. ing is likely to be due to the large number of cell fication. Alterations in dyskerin lead to altered divisions that cancer cells undergo without suf- rRNA51, but they also result in a reduction in The power of curiosity-driven research ficient telomerase. For cancer cells to overcome human telomerase RNA concentration and, The stories of the functional analysis of the the limit to cell division imposed by very short most importantly, short telomeres52,53. Dyskerin telomere, the discovery of telomerase and the telomeres, telomerase must be present to allow mutations underlie the X-linked human genetic initially unseen implications of both serve to telomere maintenance. The subsequent demon- disease , which is charac- illustrate several points. First, there is great stration that telomerase is active in many human terized by abnormal skin and by value in talking to people working in different tumor cells but is not detectable in most normal failure. This initial association of dyskerin with fields. As the specialization of grows— tissues further strengthened the idea that telom- telomerase paved the way for the demonstration perhaps inevitably, as a result of the increase in erase inhibition might be an effective way to kill that the autosomal dominant form of dyskerato- knowledge—it becomes more likely that sci- cancer cells46. sis congenita is caused by mutations in human entists will remain cocooned within their own These initial findings stimulated the cancer telomerase RNA54. research specialty. But big problems such as community to study telomerase activity in a Remarkably, mutations in both the RNA and cancer, infectious disease and aging will never

http://www.nature.com/naturemedicine variety of tumor types, which led to an explo- the protein components of telomerase result in be solved purely by studying one aspect of one sion of publications on telomerase (Fig. 3). progressive telomere shortening over many gen- biological system; such targeted work is neces- Subsequent work using cultured human cells erations in , which limits tissue renewal sary but not sufficient. We can see this clearly and a telomerase knockout mouse model con- capacity55–57. This limited tissue renewal mani- in, for example, recent advances in understand- firmed that inhibition of telomerase expression fests clinically as dyskeratosis congenita, aplastic ing the biology of aging, in which the integra- can limit division and tumor pro- anemia or other syndromes depending on the tion of studies in species from yeast through duction47. However, loss of telomere function tissue that is most affected in the individual. The humans has been critical for the progress that in some situations may lead to chromosome striking thing about the autosomal dominant has been made. Beyond this, the emerging dis- rearrangements that can fuel tumor progres- inheritance pattern is that it implies that one cipline of is helping to inte- sion48. The details of the pathways that deter- functional copy of telomerase is not sufficient for grate entirely new and unfamiliar techniques, mine whether telomerase inhibition will be telomere maintenance. Indeed Carol’s lab dem- such as the mathematical analysis of network effective in fighting particular types of cancer onstrated this directly in mice by showing that architecture, into our methodological arma- are still being worked out. This is a very com- having half the normal amount of telomerase mentarium. Nature Publishing Group Group 200 6 Nature Publishing plex and rich area of current research. results in progressive telomere shortening and A second important lesson is the value of high- © risk, high-payoff experiments. When we were 7,000 first discussing the idea of putting Tetrahymena telomeres into yeast, the experiment seemed 6,000 unlikely to work. After all, these organisms are in distantly related phyla and their cell biology

5,000 is remarkably diverse. However, a large body of experimental work showed that, in all eukaryotes that had been studied, the DNA ends located at 4,000 chromosomal termini were unusual. Given that there had to be some mechanism for telomere 3,000 replication and stabilization, it seemed possible that this mechanism might be conserved across eukaryotes. Likewise, initiating experiments 2,000 aimed at finding telomerase in Tetrahymena seemed very risky at the time. The success of 1,000 these experiments also illustrates the value of work on ‘nonstandard’ organisms—we will miss

0 a lot if we focus exclusively on a few well worked-

0 2 3 4 5 6 7 8 9 0 1 2 3 4 5 out model systems. We can learn so much from 1985* 1986 1987 1988 1989 199 1991 199 199 199 199 199 199 199 199 200 200 200 200 200 200 the diversity of life. Biology sometimes reveals its Figure 3 Cumulative citations for telomerase in Medline.The number of citations for “telomerase” is general principles through that which appears shown from every year since its discovery. While there were many important fundamental papers on to be arcane and even bizarre. But in evolution, telomerase published between 1985 and 1996, it was after this time that the medical relevance of function is frequently fundamentally conserved telomerase became established in the scientific community. The number of cancer publications on at the mechanistic and molecular levels, and telomerase then skyrocketed. *The first paper, in 1985, did not use the name “telomerase”; this term was not used until 1987. what changes is the extent and setting in which

NATURE MEDICINE VOLUME 12 | NUMBER 10 | OCTOBER 2006 1137 COMMENTARY

the various molecular processes are played out. 12. Emery, H.S. & Weiner, A.M. An irregular satellite 34. Nugent, C.I. & Lundblad, V. The telomerase reverse sequence is found at the termini of the linear extra- transcriptase: components and regulation. Genes Dev. Thus, what may at first look like an exception chromosomal rDNA in Dictyostelium discoideum. Cell 12, 1073–1085 (1998). in biology often turns out to be the manifesta- 26, 411–419 (1981). 35. Lendvay, T.S., Morris, D.K., Sah, J., Balasubramanian, B. tion of a molecular process that is much more 13. Blackburn, E.H. et al. DNA termini in ciliate mac- & Lundblad, V. Senescence mutants of Saccharomyces ronuclei. Cold Spring Harb. Symp. Quant. Biol. 47, cerevisiae with a defect in telomere replication identify fundamental. Such was the case with the ciliated 1195–1207 (1983). three additional EST genes. Genetics 144, 1399–1412 protozoa and their telomeres and telomerase. 14. Pan, W.C. & Blackburn, E.H. Single extrachromosomal (1996). Even more broadly, the quiet beginnings of ribosomal RNA gene copies are synthesized during 36. Lingner, J. et al. Reverse transcriptase motifs in the amplification of the rDNA in Tetrahymena. Cell 23, catalytic subunit of telomerase. Science 276, 561–567 telomerase research emphasize the importance 459–466 (1981). (1997). of basic, curiosity-driven research. At the time 15. Katzen, A.L., Cann, G.M. & Blackburn, E.H. Sequence- 37. Cooke, H.J., Brown, W.R. & Rappold, G.A. Hypervariable telomeric sequences from the human sex chromosomes that it is conducted, such research has no appar- specific fragmentation of macronuclear DNA in a holot- richous ciliate. Cell 24, 313–320 (1981). are pseudoautosomal. Nature 317, 687–692 (1985). ent practical applications. Our understanding 16. Klobutcher, L.A., Swanton, M.T., Donini, P. & Prescott, 38. Moyzis, R.K. et al. A highly conserved repetitive DNA of the way the world works is fragmentary and D.M. All gene-sized DNA molecules in four species of sequence (TTAGGG)n, present at the telomeres of human chromosomes. Proc. Natl. Acad. Sci. USA 85, incomplete, which means that progress does not hypotrichs have the same terminal sequence and an unusual 3ʹ terminus. Proc. Natl. Acad. Sci. USA 78, 6622–6626 (1988). occur in a simple, direct and linear manner. It is 3015–3019 (1981). 39. Allshire, R.C. et al. Telomeric repeats from T. thermoph- important to connect the unconnected, to make 17. Boswell, R.E., Klobutcher, L.A. & Prescott, D.M. ila cross hybridize with human telomeres. Nature 332, Inverted terminal repeats are added to genes during 656–659 (1988). leaps and to take risks, and to have fun talking macronuclear development in Oxytricha nova. Proc. 40. Morin, G.B. The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG and playing with ideas that might at first seem Natl. Acad. Sci. USA 79, 3255–3259 (1982). repeats. Cell 59, 521–529 (1989). 18. Orr-Weaver, T.L., Szostak, J.W. & Rothstein, R.J. Yeast outlandish. Fundamental advances are often 41. Olovnikov, A.M. A theory of marginotomy. J. Theor. Biol. transformation: a model system for the study of recom- 41, 181–190 (1973). stimulated by unexplained observations made bination. Proc. Natl. Acad. Sci. USA 78, 6354–6358 42. Harley, C.B., Futcher, A.B. & Greider, C.W. Telomeres (1981). in the course of applied work, and the resulting shorten during of human fibroblasts. Nature 19. Szostak, J.W. & Blackburn, E.H. Cloning yeast telo- growth in understanding directs our attention 345, 458–460 (1990). meres on linear plasmid vectors. Cell 29, 245–255 to new avenues of research. This cycle of interac- 43. Bodnar, A.G. et al. Extension of life-span by introduc- (1982). tion of telomerase into normal human cells. Science 20. Shampay, J., Szostak, J.W. & Blackburn, E.H. DNA http://www.nature.com/naturemedicine tion between fundamental and applied research 279, 349–352 (1998). is autocatalytic and contributes greatly to the sequences of telomeres maintained in yeast. Nature 44. Hastie, N.D. et al. Telomere reduction in human 310, 154–157 (1984). colorectal carcinoma and with ageing. Nature 346, explosive increase in scientific knowledge. If the 21. Murray, A.W. & Szostak, J.W. Construction of arti- 866–868 (1990). cycle is broken, through a failure to recognize ficial chromosomes in yeast. Nature 305, 189–193 45. de Lange, T. et al. Structure and variability of human the importance of basic science, the continu- (1983). chromosome ends. Mol. Cell. Biol. 10, 518–527 22. Murray, A.W., Schultes, N.P. & Szostak, J.W. (1990). ation of progress in more applied domains of Chromosome length controls mitotic chromosome seg- 46. Kim, N.W. et al. Specific association of human telom- science, medicine and engineering will surely be regation in yeast. Cell 45, 529–536 (1986). erase activity with immortal cells and cancer. Science limited. 23. Dawson, D.S., Murray, A.W. & Szostak, J.W. An alterna- 266, 2011–2014 (1994). tive pathway for meiotic chromosome segregation in 47. de Lange, T. & Jacks, T. For better or worse? Telomerase yeast. Science 234, 713–717 (1986). inhibition and cancer. Cell 98, 273–275 (1999). 1. McClintock, B. Cytological observations of deficiencies 24. Bernards, A., Michels, P.A., Lincke, C.R. & Borst, P. 48. Hackett, J.A. & Greider, C.W. Balancing instability: involving known genes, translocations and an inversion Growth of chromosomal ends in multiplying trypano- dual roles for telomerase and telomere dysfunction in in Zea mays. Missouri Agr. Exp. Sta. Res. Bull. 163, somes. Nature 303, 592–597 (1983). tumorigenesis. 21, 619–626 (2002). 1–48 (1931). 25. Walmsley, R.W., Chan, C.S., Tye, B.-K. & Petes, T.D. 49. Chen, J.L., Blasco, M.A. & Greider, C.W. Secondary 2. Muller, H.J. The remaking of chromosomes. Collecting Unusual DNA sequences associated with the ends of structure of telomerase RNA. Cell 100,

Nature Publishing Group Group 200 6 Nature Publishing Net 13, 181–198 (1938). yeast chromosomes. Nature 310, 157–160 (1984). 503–514 (2000). 3. Watson, J.D. & Crick, F.H. Molecular structure of © 26. Greider, C.W. & Blackburn, E.H. Identification of 50. Mitchell, J.R., Cheng, J. & Collins, K. A box H/ACA small nucleic acids; a structure for deoxyribose nucleic acid. a specific telomere terminal transferase activity in nucleolar RNA-like domain at the human telomerase Nature 171, 737–738 (1953). Tetrahymena extracts. Cell 43, 405–413 (1985). RNA 3ʹ end. Mol. Cell. Biol. 19, 567–576 (1999). 4. Kornberg, A. DNA Synthesis (W.H. Freeman, San 27. Greider, C.W. & Blackburn, E.H. The telomere termi- 51. Ruggero, D. et al. Dyskeratosis congenita and cancer in Francisco, 1974). nal transferase of Tetrahymena is a ribonucleoprotein mice deficient in ribosomal RNA modification. Science 5. Watson, J.D. Origin of concatameric T7 DNA. Nat. New enzyme with two kinds of primer specificity. Cell 51, 299, 259–262 (2003). Biol. 239, 197–201 (1972). 887–898 (1987). 52. Mitchell, J.R., Wood, E. & Collins, K. A telomerase com- 6. Blackburn, E.H. & Szostak, J.W. The molecular 28. Greider, C.W. & Blackburn, E.H. A telomeric sequence ponent is defective in the human disease dyskeratosis structure of centromeres and telomeres. Annu. Rev. in the RNA of Tetrahymena telomerase required for congenita. Nature 402, 551–555 (1999). Biochem. 53, 163–194 (1984). telomere repeat synthesis. Nature 337, 331–337 53. Chen, J.L. & Greider, C.W. Telomerase RNA structure 7. Wu, R. & Taylor, E. Nucleotide sequence analysis of and function: implications for dyskeratosis congenita. (1989). DNA. II. Complete nucleotide sequence of the cohesive Trends Biochem. Sci. 29, 183–192 (2004). 29. Yu, G.-L., Bradley, J.D., Attardi, L.D. & Blackburn, E.H. ends of bacteriophage lambda DNA. J. Mol. Biol. 57, 54. Vulliamy, T. et al. The RNA component of telomerase In vivo alteration of telomere sequences and senes- 491–511 (1971). is mutated in autosomal dominant dyskeratosis con- cence caused by mutated Tetrahymena telomerase 8. Blackburn, E.H. & Gall, J.G. A tandemly repeated genita. Nature 413, 432–435 (2001). RNAs. Nature 344, 126–132 (1990). sequence at the termini of the extrachromosomal ribo- 55. Vulliamy, T. et al. Disease anticipation is associated 30. Lundblad, V. & Szostak, J.W. A mutant with a defect in somal RNA genes in Tetrahymena. J. Mol. Biol. 120, with progressive telomere shortening in families with 33–53 (1978). telomere elongation leads to senescence in yeast. Cell dyskeratosis congenita due to mutations in TERC. Nat. 9. Yao, M.C., Blackburn, E. & Gall, J.G. Amplification of 57, 633–643 (1989). Genet. 36, 447–449 (2004). the rRNA genes in Tetrahymena. Cold Spring Harb. 31. Cohn, M. & Blackburn, E.H. Telomerase in yeast. 56. Yamaguchi, H. et al. Mutations in TERT, the gene for Symp. Quant. Biol. 43, 1293–1296 (1979). Science 269, 396–400 (1995). telomerase reverse transcriptase, in . 10. Yao, M.C., Blackburn, E. & Gall, J. Tandemly repeated 32. Prescott, J. & Blackburn, E.H. Telomerase RNA muta- N. Engl. J. Med. 352, 1413–1424 (2005). C-C-C-C-A-A hexanucleotide of Tetrahymena rDNA is tions in alter telomerase 57. Armanios, M. et al. Haploinsufficiency of telomerase present elsewhere in the genome and may be related action and reveal nonprocessivity in vivo and in vitro. reverse transcriptase leads to anticipation in autosomal to the alteration of the somatic genome. J. Cell Biol. Genes Dev. 11, 528–540 (1997). dominant dyskeratosis congenita. Proc. Natl. Acad. Sci. 90, 515–520 (1981). 33. Lingner, J., Cech, T.R., Hughes, T.R. & Lundblad, V. USA 102, 15960–15964 (2005). 11. Johnson, E.M. A family of sequences and Three Ever Shorter Telomere (EST) genes are dispens- 58. Hao, L.Y. et al. Short telomeres, even in the presence specific single-strand gaps at the termini of the Physarum able for in vitro yeast telomerase activity. Proc. Natl. of telomerase, limit tissue renewal capacity. Cell 123, rDNA palindrome. Cell 22, 875–886 (1980). Acad. Sci. USA 94, 11190–11195 (1997). 1121–1131 (2005).

1138 VOLUME 12 | NUMBER 10 | OCTOBER 2006 NATURE MEDICINE