Telomere maintenance by recombination in human cells

letter

© 2000 Nature America Inc. • http://genetics.nature.com

Telomere maintenance by recombination in human cells


Melissa A. Dunham, Axel A. Neumann, Clare L. Fasching & Roger R. Reddel Telomeres of eukaryotic chromosomes contain many tandem repeats of a G-rich sequence (for example, TTAGGG in vertebrates1). In most normal human cells, telomeres shorten with each cell division, and it is proposed that this limits the number of times these cells can replicate2. Telomeres may be maintained in germline cells, and in many immortalized cells and cancers, by the telomerase holoenzyme3 (first discovered in the ciliate Tetrahymena4), which uses an RNA subunit as template for synthesis of telomeric DNA by the reverse transcriptase catalytic subunit5. Some immortalized human cell lines and some tumours maintain their telomeres in the absence of any detectable telomerase activity by a mechanism referred to as alternative lengthening of telomeres6,7 (ALT). Here we show that DNA sequences are copied from telomere to telomere in an immortalized human ALT cell line, indicating that ALT occurs by means of homologous recombination and copy switching.

Non-telomerase mechanisms of telomere maintenance have been found in several eukaryotes. Drosophila melanogaster and related dipterans use retrotransposons for telomere maintenance8. Homologous recombination is used for telomere maintenance in the mosquito, Anopheles gambiae9. Telomerase-negative mutant yeast cells undergo telomere shortening over many cell generations followed by cell death10–14. Rare survivors maintain their telomeres by a mechanism dependent on the gene RAD52, the product of which is involved in recombination. Type I survivors have amplifi- a cation of Y´ subtelomeric eleKpnI ments, and type II survivors have long and heterogeneous telomeres11,14 similar to those seen in human ALT cell lines and tumours6,7. Telomerasenull mouse embryonic stem cells that survived long-term had amplified DNA of unknown origin in the subtelomeric region, possibly due to recombination15.

The importance of understanding the ALT mechanism(s) is underscored by proposals to treat cancer with telomerase inhibitors. Tumours containing ALT cells are likely to be drugresistant, and telomerase-positive tumours will be placed under strong selection pressure to activate ALT. ALT cell lines and tumours have nuclear bodies containing PML protein, telomeric DNA, specific telomere-binding proteins and proteins involved in recombination, such as RAD51 and RAD52 (ref. 16). As each telomere contains the same sequence tandemly repeated, we, and others, have proposed that telomeres in ALT cells may use other telomeres17,18, or themselves19 following the formation of a telomeric loop structure20, as a copy template. We assayed for inter-telomeric templating by placing a plasmid tag within the telomeres of ALT cells (Fig. 1). If the plasmid DNA tag is located within the telomere, inter-telomeric recombination that occurs proximal (centromeric) to the tag may result in copying of the tag to another telomere (Fig. 1a). In contrast, if the tag is immediately proximal to the telomeric DNA, recombination can only occur distal to the tag, and therefore the tag will not be copied (Fig. 1b). We targeted the telomeres of ALT cells (GM847) with the Tel plasmid (Fig. 1a) and determined whether telomeric integration had occurred. The G418-selected clones GM847/Tel-1 and Tel-2 contained three and two telomeric tags, respectively, at popula-

b XbaI KpnI

NotI

XbaI

NotI

Fig. 1 Strategy for detecting intertelomeric copying of DNA. a, Intertelomeric recombination proximal to an integrated telomere targeting plasmid, Tel, will result in copying of the plasmid DNA ‘tag’ sequences to another telomere. b, If Subtel plasmid DNA is located immediately proximal to telomeric DNA, inter-telomeric recombination can only occur distal to the plasmid and the tag will not be copied onto the other chromosome. For each plasmid, the locations of KpnI, XbaI and NotI restriction sites and the neomycin resistance gene (neo) are shown.

Cancer Research Unit, Children’s Medical Research Institute, Westmead, Sydney, Australia. Correspondence should be addressed to R.R.R. (e-mail: [email protected]). nature genetics • volume 26 • december 2000

447

letter


a


Fig. 2 Detection of plasmid tag DNA at the telomere by FISH. Full or partial metaphase spreads are shown of GM847/Tel-1 (a) and GM847/Tel-2 (b), which are clones of the ALT cell line (GM847) transfected with the telomere-targeting plasmid, Tel. Both clones are shown at PD 23 (early) and PD 63 (late). c, Partial metaphase spreads of GM847/Subtel-1 at PD 24 (early) and PD 64 (late); GM847/Subtel-2 at PD 24 (early) and PD 64 (late); HT1080/Tel-1 at PD 23 (early) and PD 63 (late); and HT1080/Tel-2 at PD 23 (early) and PD 63 (late). The tagged telomeres are indicated (arrowheads) and are described in Table 1.

We obtained subtelomeric tags by transfecting cells with a plasmid (Subtel; Fig. 1b) that causes chromosomal truncation and seeding of a new telomere in both telomerase-positive and ALT cell lines21,22. As predicted by the recombination/copy switching model (Fig. 1), the b number of tagged telomeres in GM847/Subtel clones did not vary with increasing PD. At both early and late PD, the same two telomeres were tagged in GM847/Subtel-1, and only one chromosome was tagged in GM847/Subtel-2 (Fig. 2 and Table 1). FISH analysis of several GM847 clones using a panel of subtelomeric probes23 did not detect subtelomeric translocation events, further demonstrating that the recombination events specifically involve the c telomeres. To test whether the effect was specific for ALT, the Tel plasmid was transfected into telomerase-positive fibrosarcoma HT1080 cells. HT1080/Tel1 and HT1080/Tel-2 showed no increase in the number of tagged telomeres from early to later PD level (Fig. 2 and Table 1). We confirmed the FISH data by Southern-blot analysis of integrated plasmid DNA: the GM847/Tel-1 and Tel-2 lines showed an increase in the number of hybridizing bands from early to later PD level (Fig. 3). This is most tion doubling (PD) 23, as detected by fluorescence in situ obvious for clone GM847/Tel-2, in which PD 63 cells have 7 hybridization (FISH). By PD 63 the number of tagged telomeres bands not present at PD 23. In contrast, the GM847/Subtel-1 and within the cell population had increased to 10 in both clones Subtel-2 lines and the HT1080/Tel-1 and Tel-2 lines contained (Fig. 2 and Table 1); the maximum number of detectable tags per fewer bands, and the number of bands did not change (Fig. 3). metaphase was 5. These data are consistent with telomere length The bands present in the GM847/Tel clones differed in relative being dynamic in ALT cells, with telomere shortening deleting intensity from early to late PD, suggesting loss of plasmid tags some of the tags and telomere lengthening by means of telomere- from some telomeres and enrichment of the population with telomere recombination resulting in copying of the tag on to pre- cells containing other tags. To demonstrate that the tag from a single telomere was copied viously untagged telomeres. into other telomeres, we transferred a single tagged chromoTable 1 • Telomere-tagged chromosomes in GM847 and HT1080 clones some from GM847/Tel-1 cells Cell line Earlya PD Latea PD into the mouse A9 cell line using microcell-mediated chromoGM847/Tel-1 10pter, 13pter, 16pter 5pter, 10pter, 13pter, 16pter, 17qter, 18pter, 20pter, 21pter, Cpter, Cqter some transfer (MMCT) folGM847/Tel-2 16pter, small marker 1pter, 1qter, 2pter, 7pter, 3pter, 16pter, lowed by G418 selection, small marker, Bqter, Cpter, marker creating the hybrid cell line GM847/Subtel-1 1pter, small marker 1pter, small marker CMC3c2 (Fig. 4a). We detected GM847/Subtel-2 13qter 13qter HT1080/Tel-1 2pter 2pter four bands hybridizing to the HT1080/Tel-2 15pter 15pter plasmid core by Southern analyaEarly and late PD levels are specified in Fig. 2. sis, and each one corresponded 448

nature genetics • volume 26 • december 2000


-1 0/ Te l T1 H

C

08

08 T1

H

0/ Te M lC 3c 2 G 2 M 84 H T1 7/T 08 el-1 0

-2

-1

te l 7/ Su b

te l G M 84

b

24 79 24 79 23 73 23 68 23 23 21

91 23 63 23 63

23.1

G M 84

7/ Su b

-2

-1

7/ Te l

7/ Te l

G M 84

7

G M 84

G M 84

PD

a

letter

23.1 9.4

9.4 6.5 6.5 4.3 4.3 2.3 2.0 2.3 2.0


Fig. 3 Southern-blot analysis of GM847 and HT1080 lines. Genomic DNA was isolated from the cells at the indicated PD. CMC3c2 is a mouse A9 cell line containing a single tagged human chromosome microcelled from GM847/Tel-1. The size of HindIII-digested λ DNA markers is shown in kilobases.

to a band present in the GM847/Tel-1 cells (Fig. 3). The tagged chromosome was then transferred by MMCT from CMC3c2 into GM847 cells. At PD 24 the tag was already detected in the telomere of a different chromosome (Fig. 4b). These data demonstrate for the first time that, in ALT cells, telomeric DNA is copied to other telomeres. The assay was not designed to detect intratelomeric copying, and it is possible that this occurs in addition to intertelomeric copying. It is also possible that the extrachromosomal DNA derived from telomeres that is present in ALT cells may be used16. The proteins involved in this process need to be identified, but it is highly likely that they are proteins involved in DNA recombination and replication rather than proteins involved in only telomere maintenance. Recent data regarding telomerase-negative yeast type II survivors indicate that the process may depend on RAD50 (ref. 24). The existence of such a mechanism was predicted before the discovery of telomerase25. ALT supports the long-term growth of immortalized cells in culture and of tumours in vivo6,7, so it is interesting to speculate why cells use an additional enzyme, telomerase, for telomere maintenance. One possibility might be that telomerase is more efficient for healing of broken chromosomes. We did not detect telomeretelomere recombination in the telomerase-positive HT1080 cell line. This is consistent with the finding that inhibition of telomerase in HT1080 cells results in telomere shortening and the onset of a senescence-like state26, indicating that these cells do not possess a telomere maintenance mechanism other than telomerase. In view of the abundance of telomeric sequences that could potentially be used as a template for telomere elongation by a recombinational mechanism in normal cells and telomerase-positive immortalized cells, it will be important to determine how ALT is normally repressed in such cells27.

Methods Plasmids. We used pSXneo-1.6-T2AG3 (ref. 28); the NotI-linearized DNA is referred to as Subtel (Fig. 1b). This plasmid contains two small multiple cloning sites (each containing a KpnI site) at the right-hand end, and in the middle of the 1.6 kb of TTAGGG repeats. The KpnI site at the right-hand end was destroyed by partial KpnI digestion, polishing with Klenow enzyme and religating. The resulting telomere targeting plasmid is referred to as Tel. Cells and cell culture. We obtained and cultured ouabain-resistant GM847 SV40-immortalized human skin fibroblasts (ALT) and HT1080 fibrosarcoma cells (telomerase-positive) as described27. Transfected GM847 and HT1080 cells were maintained in 300 and 700 µg/ml G418, respectively. We obtained A9 mouse fibrosarcoma cells from the American Type Culture nature genetics • volume 26 • december 2000

Fig. 4 Transfer of a single tagged telomere into GM847 cells. A single telomeretagged chromosome from GM847/Tel-1 cells was transferred into mouse A9 cells, creating the CMC3c2 hybrid line, and then transferred from this line into GM847 cells. a, A partial metaphase from the CMC3c2 microcell hybrid donor, which contains a telomere-tagged human chromosome (FITC, green arrow) from GM847/Tel-1. The mouse chromosomes have been hybridized with biotinylated mouse Cot-1 DNA (FITC). b, A partial metaphase from GM847 cells containing the tagged donor chromosome (green arrow) transferred from CMC3c2, and a chromosome that has a newly acquired telomere tag (white arrow).

Collection and cultured them in DME medium plus 10% fetal bovine serum (FBS), with 600 µg/ml G418 following microcell fusion. Telomere targeting. We plated 1×106 cells in a 10-cm dish 18 h before transfection. For transfections, linearized Tel or Subtel (10 µg) and Lipofectamine (100 µg; Life Technologies) were added to the cells in a total volume of 8 ml serum-free DME medium. Five hours after transfection, the medium was supplemented with 20% FBS. Two days after transfection cells were trypsinized and 1×105 cells plated out for selection. After 18 h, we added G418 at the appropriate concentration. Two weeks later when controls had died, we collected individual foci arising after transfection by trypsinization using plastic cloning cylinders and subcultured them to give rise to clonal cell lines. Clones containing telomeric and subtelomeric integrants were identified initially by FISH using the plasmid core as the probe, and also by showing that Bal31 digestion reduced the size of the terminal restriction fragment detected by probing a Southern blot with the plasmid core sequence. The first two clones identified for each category of integrant were cultured for a minimum of 60 PD, and analysed in detail. For both GM847/Tel and GM847/Subtel, the first two clones analysed had tags in the desired location, and for HT1080/Tel, we examined four clones to find two with telomeric integration; ‘correct’ tagging thus occurred in 2 of 2, 2 of 2 and 2 of 4 clones, respectively. In the case of GM847/Tel, we examined an additional eight clones by FISH only, and found that each of these had telomeric tags. No other type of integration event was detected. By this criterion, telomeric tagging occurred in 10 of 10 GM847 clones. FISH. We labelled plasmid pSXneo (equivalent to the plasmid DNA core of Tel and Subtel) with bio-16-dUTP using the Biotin-Nick Translation Mix (Boehringer) according to the manufacturer’s instructions. Approximately 30 ng/µl of pSXneo probe was hybridized onto separately denatured chromosome preparations for 16–18 h in a humidified chamber at 37 ºC. We detected the hybridized pSXneo probe with fluorescein-conjugated avidin DCS (Vector Laboratories) followed by biotinylated anti-avidin antibody (Vector Laboratories). Chromosomes were counterstained with propidium iodide (PI, 120 ng/ml final concentration; Sigma) and diamidino-phenylindole-dihydrochloride (DAPI, 0.6 µg/ml final concentration; Sigma) for chromosome identification, and slides were evaluated on a Leica DMLB epifluorescence microscope with appropriate filter sets. We analysed between 50 and 100 metaphases for each clone, and scored chromosomes as positive if a signal doublet could be detected on both sister chromatids of the respective chromosome. FITC, PI and DAPI images were captured separately with a cooled CCD (SPOT2, Diagnostic Instruments) camera, merged using SPOT2 software, and further processed using Adobe Photoshop Version 5.5 software. Microcell-mediated chromosome transfer. We performed microcell-mediated chromosome transfer29 using GM847/Tel-1 as the donor and A9 as the recipient to yield the CMC3c2 hybrid line, and then CMC3c2 as the donor

449

letter


with GM847 cells as the recipient. Briefly, we grew the donor cells in tissue culture flasks for 48 h in colcemid (20 ng/ml; Sigma), treated them with cytochalasin B (10 µg/ml; Sigma) and pelleted them in a Beckman J2-21 centrifuge. We then fractionated the pellet through a series of Nucleopore filters (Whatman) and overlaid the resulting filtrate onto the recipient cells. We treated the recipient cells with 50% polyethylene glycol 1300-1600 (Sigma), washed them and left them to recover. After 24 h we plated the recipient cells onto 10-cm dishes and selected for G418 resistance. We isolated colonies and analysed the clones by FISH to determine positive microcell hybrids. Southern-blot analysis. We electrophoresed genomic DNA (3.0 µg) that had been digested with XbaI through a 0.8% agarose gel in TAE buffer for 16 h at 30 V. The gel was Southern blotted onto Hybond N+ (Amersham) membrane

1. 2.

3. 4.


5. 6.

7.

8. 9.

10. 11. 12.

13.

14.

Blackburn, E.H. Structure and function of telomeres. Nature 350, 569–573 (1991). Olovnikov, A.M. A theory of marginotomy. The incomplete copying of template margin in enzymic synthesis of polynucleotides and biological significance of the phenomenon. J. Theor. Biol. 41, 181–190 (1973). Shay, J.W. & Bacchetti, S. A survey of telomerase activity in human cancer. Eur. J. Cancer 33A, 787–791 (1997). Greider, C.W. & Blackburn, E.H. Identification of a specific telomere terminal transferase activity in Tetrahymena extracts. Cell 43, 405–413 (1985). Colgin, L.M. & Reddel, R.R. Telomere maintenance mechanisms and cellular immortalization. Curr. Opin. Genet. Dev. 9, 97–103 (1999). Bryan, T.M., Englezou, A., Gupta, J., Bacchetti, S. & Reddel, R.R. Telomere elongation in immortal human cells without detectable telomerase activity. EMBO J. 14, 4240–4248 (1995). Bryan, T.M., Englezou, A., Dalla-Pozza, L., Dunham, M.A. & Reddel, R.R. Evidence for an alternative mechanism for maintaining telomere length in human tumors and tumor-derived cell lines. Nature Med. 3, 1271–1274 (1997). Mason, J.M. & Biessmann, H. The unusual telomeres of Drosophila. Trends Genet. 11, 58–62 (1995). Roth, C.W., Kobeski, F., Walter, M.F. & Biessmann, H. Chromosome end elongation by recombination in the mosquito Anopheles gambiae. Mol. Cell. Biol. 17, 5176–5183 (1997). Lundblad, V. & Szostak, J.W. A mutant with a defect in telomere elongation leads to senescence in yeast. Cell 57, 633–643 (1989). Lundblad, V. & Blackburn, E.H. An alternative pathway for yeast telomere maintenance rescues est1– senescence. Cell 73, 347–360 (1993). McEachern, M.J. & Blackburn, E.H. Cap-prevented recombination between terminal telomeric repeat arrays (telomere CPR) maintains telomeres in Kluyveromyces lactis lacking telomerase. Genes Dev. 10, 1822–1834 (1996). Lendvay, T.S., Morris, D.K., Sah, J., Balasubramanian, B. & Lundblad, V. Senescence mutants of Saccharomyces cerevisiae with a defect in telomere replication identify three additional EST genes. Genetics 144, 1399–1412 (1996). Teng, S.-C. & Zakian, V.A. Telomere-telomere recombination is an efficient bypass pathway for telomere maintenance in Saccharomyces cerevisiae. Mol. Cell. Biol. 19, 8083–8093 (1999).

450

and hybridized to pSXneo plasmid radiolabelled with [α-32P] by random priming. The washed membrane was exposed to Kodak MS X-ray film. Acknowledgements

We thank T. Haaf for the panel of subtelomeric CEPH mega-YACs, and T. de Lange for plasmid pSXneo-1.6-T2AG3. Supported by a project grant from the National Health and Medical Research Council of Australia, the Carcinogenesis Fellowship of the New South Wales Cancer Council, and a scholarship of the Judith Hyam Memorial Trust Fund for Cancer Research.

Received 10 August; accepted 16 October 2000.

15. Niida, H. et al. Telomere maintenance in telomerase-deficient mouse embryonic stem cells: Characterization of an amplified telomeric DNA. Mol. Cell. Biol. 20, 4115–4127 (2000). 16. Yeager, T.R. et al. Telomerase-negative immortalized human cells contain a novel type of promyelocytic leukemia (PML) body. Cancer Res. 59, 4175–4179 (1999). 17. Reddel, R.R., Bryan, T.M. & Murnane, J.P. Immortalized cells with no detectable telomerase activity. A review. Biochemistry (Mosc.) 62, 1254–1262 (1997). 18. Kass-Eisler, A. & Greider, C.W. Recombination in telomere-length maintenance. Trends Biochem. Sci. 25, 200–204 (2000). 19. Reddel, R.R. The role of senescence and immortalization in carcinogenesis. Carcinogenesis 21, 477–484 (2000). 20. Griffith, J.D. et al. Mammalian telomeres end in a large duplex loop. Cell 97, 503–514 (1999). 21. Murnane, J.P. & Yu, L.-C. Acquisition of telomere repeat sequences by transfected DNA integrated at the site of a chromosome break. Mol. Cell. Biol. 13, 977–983 (1993). 22. Sprung, C.N., Sabatier, L. & Murnane, J.P. Telomere dynamics in a human cancer cell line. Exp. Cell Res. 247, 29–37 (1999). 23. Stout, K. et al. Somatic pairing between subtelomeric chromosome regions: implications for human genetic disease? Chromosome Res. 7, 323–329 (1999). 24. Teng, S.-C., Chang, J., McCowan, B. & Zakian, V.A. Telomerase independent lengthening of Saccharomyces telomeres occurs by an abrupt Rad50p dependent, Rif inhibited recombinational process. Mol. Cell 6, 947–952 (2000). 25. Walmsley, R.W., Chan, C.S.M., Tye, B.-K. & Petes, T.D. Unusual DNA sequences associated with the ends of yeast chromosomes. Nature 310, 157–160 (1984). 26. Colgin, L.M. et al. The hTERTα splice variant is a dominant negative inhibitor of telomerase activity. Neoplasia (in press). 27. Perrem, K. et al. Repression of an alternative mechanism for lengthening of telomeres in somatic cell hybrids. Oncogene 18, 3383–3390 (1999). 28. Hanish, J.P., Yanowitz, J.L. & de Lange, T. Stringent sequence requirements for the formation of human telomeres. Proc. Natl Acad. Sci. USA 91, 8861–8865 (1994). 29. Saxon, P.J., Srivatsan, E.S., Leipzig, G.V., Sameshima, J.H. & Stanbridge, E.J. Selective transfer of individual human chromosomes to recipient cells. Mol. Cell. Biol. 5, 140–146 (1985).

nature genetics • volume 26 • december 2000