Robin Holliday

Summary

The replication of linear chromosome DNA by DNA polymerase leads to the loss of terminal sequences, in the absence of a special mechanism to maintain ends or telomeres. This mechanism is known to consist of short terminal repeats and the enzyme telomerase, which contains RNA complementary to the DNA repeats. There is evidence that telomeric DNA continually decreases in size in the absence of telomerase, and this is followed by cellular senescence. immortalisation of somatic cells is accompanied, at least in some cases, by acquisition of telomerase activity. The cloning of DNA coding for the RNA component of telomerase has opened up some new experimental approaches, including the study of telomerases with mutant RNA(’j2). The telomere theory of cellular senescence appears to provide a molecular basis for the ‘Hayflick limit’ to human fibroblast growth. However the telomeres and behaviour of primary mouse cells are anomoloud3), and many immortalised human cell lines lack normal telomerase activity(4). These exceptions are not easily accommodated in the telomere theory.

Telomeres and cellular ageing The replication of linear chromosome DNA presents a prob- lem, which was first pointed out by Watsod5) and Olovinkov(6). In the absence of a special mechanism to maintain the ends on telomeres, chromosomes will become progressively shorter. It was subsequently discovered that the special mechanism depends on short DNA repeats at chromosome termini and on the enzyme telomerase (reviewed in refs 7 and 8). It was proposed that the loss of terminal DNA, in the absence of the special mechanism, could be a cause of ageing of somatic cells or, more specifi- cally, an explanation of the ‘Hayflick limit’ to human fibro- blast proliferation@). In Hayflick’s classical study, it was shown that culture lifespan was a function of population doublings rather than chronological time(g).

The first important experiments on ageing were, how- ever, carried out with the yeast Saccharomyces cerevisiae, populations of which do not age. The isolation of mutants defective in telomerase maintenance showed that indefi- nite growth was no longer possible(10). Instead, the popula- tions initially grew, but then became senescent and died out. Soon after, it was shown that the telomeric DNA of human diploid fibroblasts progressively shortened during their in vitro lifespan(’’). It was also shown that telomere length was decreased in human chromosomes from donors of increasing age(l2!l3). In addition, the length of telomeric DNA in primary human skin fibroblasts could be

used to predict the proliferative potential of each culture(14). The telomerase hypothesis of cellular senescence was fur- ther strengthened by the demonstration that immortal cell lines maintained telomeric DNA(15). This strongly suggested that immortalisation of a normal somatic cell might depend on the acquisition of telomerase activity. A number of virus-transformed immortalised human cell lines have been shown to have telomerase activity (reviewed in ref. 4). Finally, a large sample of tumour-derived cell lines were shown to have telomerase activity(16).

Thus all the evidence seemed to favour the general hypothesis, namely, that somatic cells with finite prolifera- tion have lost telomerase activity, and that the progressive loss of telomeres provides an explanation of the Hayflick limit. Immortal germ line cells and transformed cells have telomerase activity.

The telomere theory of ageing encounters problems From the beginning there were some problems with the hypothesis. Primary mouse fibroblasts (species Mus mus- culus) grow for only 10-20 population doublings (PDs) before senescence or ‘crisis’ sets in(17,18). Subsequently, these cultures usually produce immortalised, but initially untransformed derivatives. Why then should M. musculus chromosomes have telomeres 5-20 times longer than humans(lg)? It has long been known that the lifespan of

clones and populations of human fibroblasts is very vari- able, and also that sister cells can have very different growth potential(20121). If senescence depends on telomere short- ening, one would surely expect the daughters of one cell to have very similar telomere lengths. It is also known that the lifespan of human fibroblasts can be strongly influenced by experimental treatments (reviewed in ref. 22). Agents as diverse as hydrocortisone, the peptide carnosine, ionising radiation and antisense oligonucleotides to the tumour sup- pressor genes Rb and p53, can all significantly increase in human fibroblast lifespan. It can be reduced by a single treatment of the young cells with the DNA methylating agent 5-azacytidine, incubation at 40°C, or by treatment with the aminoglycoside antibiotic paramomycin, which reduces translational fidelity. Why should such a wide range of treat- ments affect telomere shortening? Just conceivably, they could influence the numbers of repeats lost per cell gener- ation, or alter the phenotypic effect of telomere loss. In any event, additional assumptions have to be made which demand further experimentation.

Cloning the RNA components of telomerases Recently, several important studies of telomeres and telo- merases have The gene coding for the pro- tein component of telomerase has not been cloned, but sev- eral laboratories have cloned the DNA coding for the RNA component of the enzyme. This opens the way for a number of new experimental techniques, including the phenotypic effects of mutating the sequences. McEachern and Black- burn('), using the yeast Kluyveromyces lactis, find that some mutations bring about immediate runaway telomere elongation, whereas others cause elongation after a long latent period of growth. They propose a self-regulatory model for the control of normal telomeric length, which can be deranged by a change in telomeric DNA sequence. Feng et a/.(2) have cloned the RNA component of human telo- merase (hTR). This comprises 1 1 nucleotides (5'CUAACC- CUAAC) complementary to the human telomere sequence (TTAGGG)". Cell lines with mutations in the template com- ponent of the enzyme generated the predicted mutant telomerase activity. Antisense expression constructs were introduced into HeLa cells. It is reported that 41 cultures expressing antisense to hTR initially grew normally, but at 23-26 PDs after transfection, 33 cultures underwent crisis, with about 1% of cells eventually resuming growth. A decline in telomere length during the period of growth was not documented, but the average telomere length in the cul- tures expressing antisense was about 25% shorter than in controls.

lmmortalised cells without telomeres Although the first studies on immortalised human cell lines confirmed the existence of telomerase, a more recent inves-

tigation of human cells immortalised by oncogenic viruses shows that many lack this enzyme activity. Bryan et a/.(4) examined, using the sensitive PCR-based assay, 35 cell lines immortalised in vitro and found that 15 had no detectable telomerase. The telomerase-negative cell lines had long heterogenous telomeres, much longer than immortalised lines with telomerase. They suggest three possible explanations: (1) the telomerase may be altered or mutant, and does not use the oligonucleotide substrate in the PCR assay; (2) the same mutant is dysregulated in the cell, resulting in abnormal telomere lengthening; or (3) that an alternative mechanism for maintaining telomeres may exist, such as a recombination mechanism, which is known to generate additional telomeric sequences in yeast(23).

Problems with two mouse species The very long telomeres of Mus musculis have already been mentioned, but another species, Mus spretus, has telo- meres roughly the same size as those in human cells. Prowse and Greideh3) studied primary M. spretus fibrob- lasts and report that the length of telomeric DNA shortens over about 60 PDs (comparable to the lifespan of human cells), but subsequently stabilises. Oddly, no discernible cri- sis was seen, although this is a standard feature of M. mus- culus primary cultures. Fibroblast lifespan is stongly corre- lated with the longevity of the donor species(24), so why should a short-lived rodent such as M. spretus have such long-lived cells, if that is indeed the case? In the same study, telomerase was assayed in somatic tissues of M. spretus. None was detected in the brain, but it was present in liver, with weaker activity in spleen and kidney. Testis, as expected, was also positive. In contrast, several normal human somatic tissues have been reported to have no telomerase(16). However, another recent study detected telomerase in human bone marrow tissue and peripheral blood leukocytes(25).

Steady states and unbalanced growth The telomere theory of cell senescence and immottalisation is so beguiling, that there is a danger of accepting it as truth. What we do know is that cell populations that can grow indefi- nitely are essentially in a steady state, albeit with variability between cells and continual selection of the fastest growing cells. In such immortalised populations, most phenotypic fea- tures will remain constant. The length of telomeres is one example; another is the level of DNA methylation(l*). In con- trast, diploid somatic cells with finite proliferative potential can be said to be in a state of unbalanced growth. Something is changing during their period of normal proliferation, which has no outward phenotypic effect, but eventually gives rise to senescence. This 'something' could be the loss of telomeric DNA, or it could be one or more other cumulative molecular changes, an example being the progressive loss of DNA

methylation(18B26). It is not difficult to envisage the interaction, or coupling, of different molecular changes, which may make it difficult to distinguish cause and effect. Is telomere loss suf- ficient to drive the cell towards senescence, or do some other cumulative events at the protein or DNA level also lead to the loss of telomere maintenance? The quest for the complete pic- ture may not be endless, but much remains to be discovered.

Acknowledgements I thank Roger Reddel and Tracy Bryan for helpful comments on the manuscript.

References 1 McEachern, M.J. and Blackburn, E.H. (1995). Runaway telomere elongation caused by telomerase gene mutations. Nature 376,403-408. 2 Feng, J. Funk, W.D., Wang, S.S., Weinrich, S.L., Avilion, A.A., Chiu, C.P., Adams, R.R., Chang, E., Allsop, R.C., Yu, J., Le, S., West, M.D., Harley, C.B., Andrews, W.H., Greider, C.W. and Villeponteau, B. (1995). The RNA component of human telomerase. Science 269,1236-1241. 3 Prowse, K.R. and Greider, C.W. (1 995). Developmental and tissue-specific regulation of mouse telomerase and telomere length. Proc. Nafl Acad. Sci. USA 92,4818-4822. 4 Bryan, T.M., Englezou, A., Gupta, J., Bucchetti, S. and Reddel, R.R. (1995). Telomere elongation in immortal human cells without detectable telomerase activity. EM60 J. 14,4240-4248. 5 Watson, J.D. (1972). Origin of concatameric T7 DNA. Nature New Biol. 239, 197-201. 6 Olovnikov, A.M. (1 973). A theory of marginotomy. J. Theoret. Biol. 41, 181 -1 90. 7 Blackburn, E.H. (1 991). Structure and function of telomeres. Nature 350, 569- 573. 8 Blackburn, E.H. (1992). Telomerases. Ann. Rev. Biochem. 61,113-129. 9 Hayflick, L. (1965). The limited in vitrolife time of human diploid cell strains. Exp. Cell Res. 37,614-636. 10 Lundblad, V. and Szostak, J.W. (1989). A mutant with a defect in telomere elongation leads to senescence in yeast. Cel/57,633-643.

11 Harley, C.B., Futcher, A.B. and Greider, C.W. (1990). Telomeres shorten during ageing of human fibroblasts. Nature 345, 458-460. 12 Hastie, N.D., Dempster, M., Dunlop, M.G., Thompson, A.M., Green, D.K. and Allshire, R.C. (1990). Telomere reduction in human colorectal carcinoma and with ageing. Nature346,866-868. 13 Lindsey, J., McGill, N.I., Lindsey, L.A.,Green, D.K.andCooke, H.J. (1991). In vivo loss of telomeric repeats with age in humans. Mutat. Res. 256,45-48. 14 Allsop, R.C. eta/. (1992). Telomere length predicts replicative capacity of human fibroblasts. Proc. NatlAcad. Sci. USA 89, 101 14-101 18. 15 Counter, C.M. et a/. (1992). Telomere shortening associated with chromosomal instability is arrested in immortal cells which express telomerase activity. EM60J. 11, 1921-1929. 16 Kim, N.W. eta/. (1994). Specific association of human telomerase activity with immortal cells and cancer. Science266,2011-2015. 17 Todaro, G.J. and Green, H. (1963). Quantitative studies of the growth of mouse embryo cells in culture and their development into established lines. J. Cell BiOi. 17,299-313. 18 Wilson, V.C. and Jones, P.A. (1983). DNA methylation decreases in aging but not in immortal cells. Science220, 1055-1057. 19 Kipling, D. and Cooke, H.J. (1990). Hypervariable ultra-long telomeres in mice. Narure347,400-402. 20 Holliday, R., Huschtscha, L.I., Tarrant, G.M. and Kirkwood, T.B.L. (1977). Testing the commitment theory of cellular ageing. Science 198, 366-372. 21 Smith, J.R. and Whitney, R.G. (1980). lntraclonal variation in proliferative potential of human diploid fibroblasts- stochastic mechanism for cellular aging. Science 207,8244. 22 Holliday, R. (1995). Understanding Ageing. Cambridge University Press. 23 Wang, S.S. and Zokian, V.A. (1 990). Telomere-telomere recombination provides an express pathway for telomere acquisition. Nature345, 456-458. 24 Rohme, D. (1981). Evidence for a relationship between longevity of mammalian species and lifespan of normal fibroblasts in vitro and erythrocytes in vivo. Proc. NatlAcad. Sci. USA 78,5009-501 3. 25 Broccoli, D., Young, J.W. and de Lange, T. (1995). Telomerase activity in normal and haematopoietic cells. Proc. Natl Acad. Sci. USA 92, 9082-9086. 26 Fairweather, S., Fox, M. and Margison, G.P. (1987). The in vitro lifespan of MRC-5 cells is shortened by 5-azacytidine induced demethylation. Exp. Cell Res. 168, 153-159.

Robin Holliday is at the CSIRO Division of Biomolecular Engineering, PO Box 184, North Ryde, NSW 2113, Sydney, Australia.