The protection of genetic information is essential for cells and org-
anisms, because the accumulation of chromosomal aberrations leads
to genomic instability. The evolution of linear chromosomes presents a
complex puzzle, because chromosome ends need to be protected from
enzymatic attack to avoid the loss of genetic information. In addition,
all cells have developed mechanisms to detect DNA lesions, and nat-
ural chromosome ends need to be hidden from this machinery. The
solution in most eukaryotes lies in nucleoprotein complexes known as
telomeres, which consist of G-rich DNA repeats covered by specialized
binding proteins. The actual terminus of a telomere is not blunt-ended
but consists of a single-stranded 3ʹ protrusion of the G-rich strand (or
G strand), known as a G tail or G overhang. These overhangs have been
observed in humans, mice, ciliates, yeast, trypanosomes and plants,
demonstrating that they are evolutionarily conserved and an essential
feature of telomeres.
Loss of telomere function has various consequences in many model
organisms, such as loss of the telomere G overhang, resection of the
C-rich strand (or C strand), increased levels of recombination at chro-
mosome ends, altered gene-expression patterns, fusion of chromo-
somes, instability of the genome, growth arrest and cell death. Most
of our knowledge about telomere structure and function is derived
from studies in a diverse range of organisms, such as the ciliates
Euplotes crassus (also known as Moneuplotes crassus), Tetrahymena
thermophila and Oxytricha nova (also known as Sterkiella nova), and the
yeasts Saccharomyces cerevisiae, Schizosaccharomyces pombe and Kluy-
ver omyces lactis. In human and mouse cells, the only direct conse-
quences of telomere dysfunction that have been identified so far are
degradation of the G strand and/or chromosome fusion. Whether
similar phenotypes to those observed in the ciliates and/or yeasts also
occur in mammalian cells with impaired telomere function remains
to be seen.
Studies in human and mouse cells suggest that the G-rich single-
stranded telomere overhang can invade homologous double-stranded
telo meric tracts, resulting in a large lasso-like structure, known as
a telomeric loop (t-loop)1 (Fig. 1). This provides an elegant and appealing
mechanism by which chromosome ends could be protected. However,
at present, it is not clear whether t-loops are present at all chromosome
ends, whether they are required for chromosome protection, or whether,
instead, they have a role in regulating other features of telomeres (for
example, access for the telomere-specific reverse transcriptase, known
as telomerase, and therefore telomere length). Telomeres in cells from
humans, mice, ciliates, trypanosomes and plants, as well as yeast engi-
neered to have long telomeres, have been shown to have t-loops2.
Terminal loops are an attractive model for a specialized configuration
of chromosome ends; however, it is clear from the ciliate O. nova that
other equally efficient structures have evolved. O. nova chromosome
ends are tightly bound by a complex of two proteins of 56 and 41 kDa,
efficiently protecting both the single-stranded telomere overhang and
the double-stranded telomeric DNA from modifying enzymes3–6. This
suggests that t-loops have evolved as only one of several means of chro-
mosome end protection.
Mammalian telomeres are associated with the shelterin complex,
a complex of interdependent telomeric core proteins consisting of
telomeric-repeat-binding factor 1 (TRF1), TRF2, TRF1-interacting
protein 2 (TIN2), the transcriptional repressor/activator protein RAP1,
protection of telomeres 1 (POT1) and the POT1- and TIN2-organizing
protein TPP1 (ref. 7; Fig. 1). TRF1 was originally reported to be involved
mainly in the control of telomere length, and TRF2 was mainly impli-
cated in chromosome end protection, by preventing end-to-end
fusions8,9. Now, however, these distinctions are less clear-cut, because
TRF2 has been shown to have a role in telomere length regulation10, and
the targeted deletion of Trf1 in mice leads to early embryonic lethality.
This lethality is probably due to telomere deprotection, because con-
comitant deletion of the DNA-damage sensor p53 extended the life of
the embryos11. Such overlapping phenotypes can be explained by the
nature of the shelterin complex, which is destabilized by the removal of
In addition to shelterin, mammalian telomeres interact with a
number of other factors that can influence chromosome end integ-
rity and dynamics, such as tankyrase 1 and tankyrase 2, poly(ADP-
ribose) polymerase (PARP), meiotic recombination 11 homologue
(MRE11), the RecQ-like helicases WRN (Werner’s syndrome protein)
and BLM (Bloom’s syndrome protein), Ku70, Ku86, DNA-dependent
protein kinase (DNA-PK; also known as PRKDC), ataxia-telangiectasia
mutated (ATM), ATM and Rad3-related (ATR), excision repair cross-
complementing 1 (ERCC1), RNA-polymerase σ 70 factor (XPF) and
the DNA-repair protein RAD51D7 (Fig. 1). This plethora of factors,
many of which are involved in DNA recombination and repair, not
only demonstrates the flexibility and dynamic nature of the complex
but also presents a paradox, because telomeres have long been defined
as structures that are protected against becoming substrates for DNA
repair or recombination. However, it is becoming increasingly clear that
the repair and recombination machineries are an important component
of telomere replication, protection and stability. One challenge for the
telomere field is to address how these machineries contribute to these
different classes of DNA end.
Replication and protection of telomeres
Ramiro E. Verdun1 & Jan Karlseder1
During the evolution of linear genomes, it became essential to protect the natural chromosome ends to
prevent triggering of the DNA-damage repair machinery and enzymatic attack. Telomeres — tightly regulated
complexes consisting of repetitive G-rich DNA and specialized proteins — accomplish this task. Telomeres
not only conceal linear chromosome ends from detection and inappropriate repair but also provide a buffer
to counteract replication-associated shortening. Lessons from many model organisms have taught us about
the complications of maintaining these specialized structures. Here, we discuss how telomeres interact and
cooperate with the DNA replication and DNA-damage repair machineries.
1The Salk Institute for Biological Studies, 10010 North Torrey Pines Road, La Jolla, California 92037-1099, USA.
NATURE|Vol 447|21 June 2007|doi:10.1038/nature05976
The end-replication problem and ageing
In 1972, James Watson wrote, “While 5ʹ to 3ʹ oriented growth should
proceed smoothly to the end of its template, I see no simple way for
3ʹ to 5ʹ growth to reach the 3ʹ end of its template”12. Thus, he correctly
predicted that the lagging strand of linear chromosomes copied by the
semi-conservative replication machinery would not be fully replicated12.
In 1973, A. M. Olovnikov proposed the ‘marginotomy theory of ageing’,
suggesting that ‘telogenes’ located at opposite ends of DNA molecules
carry no genetic information and fulfil a buffer function. He stated that
these telogenes are stochastically shortened during each mitotic cycle,
providing a mechanism for ageing13.
Observations made by L. Hayflick in 1961 suggested that human cells
derived from embryonic tissues can only divide about 50 times, and this
became known as the Hayflick limit14. Since then, the assumption that
the Hayflick limit is determined by the initial length of the telomeres
and the rate of telomere shortening, as laid out in the mathematical
approach of A. M. Olovnikov13, has been proved experimentally15,16.
It is well established that critically short telomeres cease to function as
protective units and cause the cell to die or to arrest permanently.
Telomeres are now known to have many more roles than simply
buffering against DNA loss; however, the initial concept of replication-
associated telomere shortening was correct. At present, we refer to the
inability of conventional DNA polymerases to replicate linear molecules
fully as the ‘end-replication problem’. This is caused by the deletion of
the RNA primer of the most distal Okazaki fragment and results in the
loss of about five bases of terminal genetic material per population
doubling15,17–19. However, the sequence loss that is predicted to occur
as a result of the end-replication problem is considerably less than that
which has been observed in primary human cells, which lose about
100–200 bases of TTAGGG repeats per cell division20–22. Consequently,
replication-associated terminal sequence loss is caused by a combina-
tion of the end-replication problem and the processing that must occur
to create the G overhang on the telomeres generated by leading- and
When telomeres become critically short, they are detected by the cell-
ular DNA-damage repair machinery23. As demonstrated in S. cerevisiae,
chromosomes that lose a telomere are often eliminated, despite check-
point and DNA-damage repair machineries24. In human cells, p53- and
RB1 (retinoblastoma 1)-dependent pathways are responsible for moni-
toring telomere function, whereas p53 seems to be the main sensor in
mouse cells25. The minimal functional telomere length, and whether this
length varies among cell types, has not been clearly defined. But even in
senescent human cells, telomeric double-stranded repeats are readily
detectable, suggesting that several kilobases of TTAGGG repeats are
required at all times. Similarly, when the telomere-protection factor TRF2
is overexpressed in telomerase-negative primary human fibroblasts,
telomere-shortening rates almost double, and cells enter senescence
with considerably shorter telomeres than control populations, indica-
ting that telomere structure, not telomere length, is the main determinant
of functional telomeres26.
Analysis of signalling from experimentally induced dysfunctional
mammalian telomeres and from chromosome ends in senescent cells
suggests that the same machinery, the intracellular DNA-damage-
monitoring system, recognizes both. Telomeres that are stripped of the
protective shelterin complex by expression of a dominant form of TRF2
become associated with factors involved in DNA-damage responses
such as the p53-binding protein 53BP1, the histone protein γ-H2AX
(phosphorylated H2AX) (see page 951), RAD17, ATM and MRE11, and
are visualized as TIF, or telomere-dysfunction-induced foci27. Inhibition
of the phosphatidylinositol-3-OH-kinase-like kinases ATM and ATR
reduces TIF formation, confirming that dysfunctional telomeres are
detected by the ATM–p53 pathway27. Cells carrying senescent telomeres
trigger a cellular response remarkably similar to that elicited by double-
strand DNA breaks28; in these cells, several DNA-damage-response fac-
tors congregate at the eroded telomeres similarly to TIF. These findings
underscore the importance of the DNA-damage repair machinery for
telomere function, emphasizing the fact that this machinery carries out
several essential tasks.
Replicative senescence can be viewed as a mechanism to limit the
potential number of population doublings a cell can undergo, hypothet-
ically rendering it a powerful tumour-suppressor mechanism29. Each
time a cell divides, telomeres shorten as a result of the end-replication
problem and end processing. After telomeres have become critically
short, they are detected by the DNA-damage repair machinery, and
the cell dies or enters senescence. At present, senescence in human
cells is regarded as an irreversibly arrested state, effectively inhibiting
the generation of immortal cells and therefore cancer formation. As a
result, major tumour-suppressive mechanisms need to be deactivated
before a cell can overcome this block to immortality. Cells that continue
to divide past their normal replicative limit lose all remaining protec-
tive telomeric DNA and enter a stage termed crisis. Crisis is marked
by massive genomic instability and cell death. Eventually, transformed
clones emerge, and although the activation of telomerase is not essential
for the acquisition of a transformed phenotype30, most cells that suc-
cessfully exit from crisis have upregulated its activity. This observation
emphasizes the dual role of telomerase in the immortalization process.
On the one hand, reactivation of the enzyme in cells with critically
short telomeres allows genomically unstable and immortal clones to be
established, which is a major step towards cancer. On the other hand,
switching on telomerase in cells that have not reached crisis prevents
telomere-mediated genomic instability, which is a hallmark of can-
cer cells31. Consequently, it could be argued that telomerase fulfils a
tumour-suppressive role before it contributes to the establishment of
immortality (Fig. 2).
Figure 1 | The mammalian telomeric
complex. The fluorescence image shows the
location of a telomere within a chromosome.
Mammalian telomeres consist of TTAGGG
repeats with a single-stranded 3ʹ overhang of
the G-rich strand. Specific protein complexes
bind to the double- and single-stranded
telomeric DNA. The components of the
shelterin complex are shown in bold text.
The single-stranded overhang can invade
the double-stranded portion of the telomere,
forming protective loops — such as t-loops
with displacement loops (D-loops) — at the
invasion site. The telomerase complex (which
contains the telomerase RNA template and the
reverse transcriptase TERT) interacts with
the overhang and is regulated by shelterin
and other telomeric proteins7. Other factors
that can interact with telomeres are listed.
Bidirectional arrows indicate interactions.
NATURE|Vol 447|21 June 2007
Approximately 10% of human tumours rely on a telomerase-
independent method to maintain their telomeres. Known as ALT (alter-
native lengthening of telomeres), it is based on recombination between
telomeres32 (Box 1). Tumours resulting from Wrn–/– mouse cells without
telomerase activity readily engage the ALT pathway, and although the
exact molecular mechanism is not understood, preliminary findings point
to aberrant homologous recombination as the underlying cause33,34.
Despite the finding that expression of the catalytic subunit of
telomerase readily immortalizes primary cells without seeming to
cause genomic instability16,35, the in vivo evidence for telomere involve-
ment in human ageing is limited to correlations. For example, the mean
length of telomeric restriction fragments from DNA isolated from sperm
cells is considerably longer than comparable fragments isolated from
replicating cells in vivo36, and it has recently been demonstrated that
senescent cells account for up to 15% of the cell population in the skin
of aged baboons37. Many such correlations have been documented, and
although they show that telomeres shorten with age, it is unclear whether
telomere shortening causes ageing in vivo.
Little is known about the role of telomeres during the ageing pro cess
of differentiated cells and in organisms that do not contain mitotic cells.
Although it has been suggested that telomere elongation extends the
lifespan of the nematode Caenorhabditis elegans38 (in which cells do
not undergo mitosis after development is complete), this is probably a
secondary effect of hrp-1 overexpression in these animals, because clonal
wild-type nematode strains with varying telomere length did not show
any differences in organismal ageing and lifespan39. Consequently, and
in line with Watson’s and Olovnikov’s concepts of the end-replication
problem, it is unlikely that telomeres have a major role in the ageing of
The end-replication problem correctly predicts that linear DNA mol-
ecules shorten during every replicative cycle. Consequently, in the
absence of a mechanism to maintain the absolute ends, chromosomes
eventually lose the protective cap provided by their ends, resulting
in the loss of genomic integrity. To counteract replication-associated
telomere shortening, telomerase evolved. Telomerase is a specialized
reverse transcriptase complex and can add G-rich telomeric repeats to
the absolute ends of chromosomes using its own internal RNA template,
effectively stabilizing telomere length. In S. cerevisiae, telomeres switch
between extendable and non-extendable states40. In human cells, the lim-
ited amount of telomerase contributes to telomere length homeostasis,
because increased amounts of telomerase change telomere length settings
to a different equilibrium41. In S. cerevisiae, deletion of the gene encoding
the telomerase RNA template (TLC1) or the catalytic subunit (EST2)
leads to gradual telomere shortening and growth arrest17–19. Similarly,
mice lacking the gene encoding either the telomerase RNA (Terc) or the
reverse transcriptase domain (Tert) gradually lose their telomeres over
several generations, resulting in degeneration of highly proliferative cell
populations and sterility42–44. Expression of human TERT in fibroblasts
causes telomere elongation and renders the cells immortal, effectively
avoiding telomere-shortening-dependent replicative senescence16.
No origin of replication has been detected in telomeres, rendering the
closest origin, resident in the subtelomeric region of the chromosome,
the starting point for the replication of chromosome ends. Passage of the
replication fork through the telomere is thought to generate a blunt-
ended leading-strand product and a lagging-strand product with a short
3ʹ G overhang (Fig. 3). In the presence of telomerase, and during new
telomere synthesis, the actions of the conventional replication machinery
and telomerase are closely coordinated. Inhibition of C-strand synthesis
in the ciliate E. crassus by using aphidicolin, a specific inhibitor of DNA
polymerase-α and DNA polymerase-δ, leads to a general lengthening of
the G strand, thereby showing that C- and G-strand synthesis is coordi-
nated45. Similarly, addition of new telomeres by telomerase in S. cerevisiae
requires not only extension of the 3ʹ G-rich end by telomerase but also
fill-in synthesis of the C strand by DNA primase, DNA polymerase-α and
DNA polymerase-δ. G-strand polymerization by telomerase is inhibited
in S. cerevisiae if DNA polymerase-α and DNA polymerase-δ are inactive,
suggesting that telomerase needs to interact with the lagging-strand syn-
thesis machinery to be active46. An excellent candidate for regulating the
coordination between telomerase and the conventional DNA polymerases
is the Cdc13 complex, which attracts telomerase to chromosome ends
in S. cerevisiae. Cdc13 binds to single-stranded G-rich telomeric
DNA and then recruits telomerase47,48. Several studies have suggested
that, subsequently, lagging-strand synthesis fills in the C strand, and then
inhibits telomerase in a Cdc13-dependent manner49,50.
Little is known about coordinated C- and G-strand synthesis in mam-
malian cells, but activation of a temperature-sensitive allele encoding
DNA polymerase-α causes elongation of the G tail and of the telomere
overall, suggesting that coordination of telomerase with the replication
machinery is a common feature in all organisms51.
The G-rich and repetitive nature of telomeric DNA complicates rep-
lication as well, because it potentially allows the formation of second-
ary structures, such as G quartets52. Consequently, telomeric proteins
support the progressing replication fork, allowing efficient telomere
syn thesis. In S. pombe, Taz1, the homologue of TRF1 and TRF2, is
required for telomere replication. Without Taz1, replication forks stall
at telomeric sequences, regardless of whether the repeats are located at
the ends or in the interior of chromosomes53. This suggests that the Taz1-
dependent telomere-replication phenotype is due to characteristics of
the telomeric sequence itself and not to its position on the chromosome.
In human cells, the RecQ-like helicase WRN contributes to efficient
telomere replication. Overexpression of a helicase-defective WRN allele
causes the occasional loss of telomeres generated by the lagging-strand
machinery, and telomerase expression compensates for this loss of
telomeric sequence, implicating telomere maintenance in the pathology
of Werner’s syndrome54. Accordingly, targeted deletion of Wrn in mice
leads to phenotypes that resemble the human Werner’s syndrome only
when telomerase is also deleted55. In summary, it is becoming increas-
ingly clear that telomere replication and telomerase-dependent telomere
elongation are highly coordinated processes and that telomeric proteins
have essential roles in the regulation of these processes.
(immortal, stable genome,
Loss of p53 and RB1
telomeres 1–4 kb)
Crisis (>80 PD,
telomeres critically short,
Figure 2 | Telomere shortening, senescence and cancer. Primary cells
divide exponentially, and telomeres shorten from ~15 kilobases (kb) until
they reach a critical length, 4–6 kb. Irreversible cell-cycle arrest then occurs
(blue). Activation of telomerase before senescence allows cells to divide
indefinitely and maintain a stable genome (green). If, instead, the p53
and RB1 pathways are suppressed, cells continue dividing (orange) until
end protection is completely lost, resulting in telomeric crisis, cell death
and massive genomic instability (dark pink). If telomerase is activated
before erosion is complete, this rescues the genome from instability by
re-establishing telomere maintenance (light pink). Activation of telomerase
after the accumulation of mutations results in an unstable genome, allowing
clones that carry multiple mutations to escape cell death (that is, to become
immortal). Such cells are predisposed to oncogenic transformation
(brown). PD, population doublings.
NATURE|Vol 447|21 June 2007
The generation of telomere overhangs
After telomere replication is completed, the newly generated lagging-
strand telomere carries a short 3ʹ overhang, resulting from the removal
of the most distal RNA primer used for Okazaki fragment synthesis. It
is not clear whether this distal primer is placed at the absolute terminus
of the chromosome or a few bases from the end, therefore allowing an
overhang that could be longer than the length of the RNA fragment. By
contrast, leading-strand synthesis is expected to continue until it reaches
the end of the template, resulting in blunt-ended products (Fig. 3). In
ciliate, yeast and human cells, overhangs can be detected at both ends of
the chromosomes, suggesting that there are regulated mechanisms for
G-tail generation56–60. Currently, no candidates for overhang-generating
nucleases have been identified in any organism, but the field is looking
to ciliates for clues, because much of the work on telomere replication
was pioneered in these organisms.
T. thermophila maintains G overhangs with defined sequence and
length on both chromosome ends56. Because both the G strand and the
C strand are processed accurately in the absence of telomerase, it has
been suggested that C-strand resection works in collaboration with
G-strand cleavage to generate a functional telomere end. These steps
potentially include more than one nuclease, and when the telomeric
sequence is artificially changed, the terminal nucleotides were not
altered, suggesting that the nuclease activities do not show sequence
preference and that additional factors, such as proteins that bind to the
single-stranded overhang, regulate specificity61.
In human cells, in the absence of telomerase, the leading-strand daugh-
ter telomeres carry longer overhangs than the telomeres synthesized by
the lagging-strand machinery62. However, when the catalytic subunit of
telomerase is introduced, similar lengths are found at both leading and
lagging daughter telomeres. Moreover, both daughter telomeres have
conserved terminal nucleotides at their 5ʹ ends. Small-interfering-RNA-
dependent knockdown of expression of the shelterin component POT1
in human tumour cells randomizes the last nucleotides of the 5ʹ telomere
end, suggesting that this single-stranded TTAGGG-binding protein is
involved in regulation of terminal specificity63. This process differs for
the 3ʹ end, where the terminal residues seem to be much more variable
in the absence of telomerase. Overexpression or suppression of indi-
vidual components of the telomere-binding protein complex or of the
multisubunit telomerase complex might already disturb the equilibrium
at telomeres, limiting the conclusions that can be drawn about the in vivo
situation from such experiments.
Eventually, detailed knowledge about overhang length and base
specificity of leading- and lagging-strand telomeres in the presence
and absence of telomerase will provide insight into the coordination
of telomerase activity at telomeres. There are several possible ways in
which this coordination might occur. Both daughter telomeres are
nucleolytically recessed in the 5ʹ to 3ʹ direction after replication, but the
efficiency varies at each strand, potentially rendering the leading-strand
telomere with a longer overhang62. The presence of telomerase compli-
cates processing, because telomerase could take advantage of the short
overhang at the lagging strand before nuclease action, specifically elon-
gating telomeres replicated by the lagging-strand machinery. However,
if telomerase acts after overhangs are generated on both strands, more
than one fill-in step by the lagging-strand machinery and more than one
resection step might be required to generate functional chromosome
ends. The field of telomere research is in agreement that well-regulated
overhang generation is an essential step for telomere function and thus
for chromosome protection; however, at this stage, only the surface of
this complex problem has been scratched.
Telomeres and the DNA-damage repair machinery
The protective features of telomeres are lost when chromosome ends
become uncapped, through the mechanisms described here. These sub-
sequently dysfunctional telomeres are then subject to DNA repair by
non-homologous end joining (NHEJ) or homologous recombination.
In accordance with a requirement for the NHEJ pathway (which
depends on Ku, DNA-PKcs (the catalytic subunit of DNA-PK), and DNA
ligase IV and its cofactor XRCC4) for the processing of dysfunctional
telomeres, DNA ligase IV is required for end-to-end fusion of criti-
cally short or dysfunctional telomeres. DNA ligase IV fuses telomeres
in S. pombe lacking taz1 (a homologue of the mammalian TRF genes)64,
in S. cerevisiae with mutated TEL1 (which encodes a protein kinase) or
MEC1 (which encodes a signal transducer)65, and in mouse cells that
lack TRF2 and therefore contain uncapped chromosome ends66. Ku con-
tributes to telomere protection in S. cerevisiae, S. pombe and mammalian
A subset of immortalized cells do not show telomerase upregulation
and use a recombination-based pathway known as alternative
lengthening of telomeres (ALT) to maintain chromosome ends. One of
the main characteristics of ‘ALT cells’ is the presence of promyelocytic
leukaemia (PML) bodies, which are subnuclear structures that
contain telomeric DNA, telomeric proteins and factors involved
in DNA recombination and repair. ALT cells have heterogeneous
telomere lengths, ranging from critically short telomeres to telomeres
of up to 100 kb, as well as telomeric DNA circles of 1–60 kb94. It has
been suggested that the homologous-recombination machinery is
responsible for the amplification of the telomeric sequences in
ALT cells, because a selection marker introduced into telomeres of ALT
and non-ALT cells spreads throughout telomeres only in ALT cells33.
ALT cells show an increased rate of sister chromatid exchange95,96,
suggesting that the homologous-recombination pathway is involved.
Overexpression of a mutant TRF2 allele in non-ALT cells generates
t-loop-sized DNA circles that depend on the ERCC1–XRCC3 complex81.
This phenotype resembles the ALT-associated DNA circles and again
suggests a possible role for homologous recombination in the ALT
The molecular mechanism of the ALT pathway is far from
understood but resembles break-induced replication (BIR). BIR is a
gene-conversion mechanism that is induced only when one DNA end
invades a homologous sequence and initiates DNA replication with the
homologous sequence as template. Proteins such as RAD50, RAD52
and MRE11 are involved in BIR97. It is feasible that invasion of the single-
stranded telomere overhang into double-stranded TTAGGG repeats
is a BIR-like situation. Consequently, it is possible that the telomere
maintenance mechanism used by ALT cells is similar to that proposed
for BIR. The figure compares telomere structure and regulation in non-
ALT cells and ALT cells.
In the absence of telomerase, a subset of yeast cells elongate their
telomeres through amplification of telomeric and subtelomeric repeat
sequences. This recombination maintenance pathway depends on
RAD52 (ref. 98), suggesting that homologous recombination can
elongate telomeres in a telomerase-independent manner.
Box 1 | ALT, homologous recombination and BIR at telomeres
Elongation No elongation
Normal cellsALT cells
NATURE|Vol 447|21 June 2007
cells. In all of these organisms, Ku associates with telomeric DNA, regu-
lates telomerase and inhibits inappropriate recombination and repair
events at telomeres67,68.
In S. cerevisiae, the single-stranded telomeric DNA-binding factor
Cdc13, which also has a prominent role in telomerase recruitment, is
central to chromosome end protection69. A protective complex of Cdc13,
Stn1 and Ten1 protects the telomere from excessive degradation of the
C strand47,70,71, resulting in long G tails, which trigger Rad9-dependent
Loss of the S. pombe double-stranded telomere-binding protein
Taz1 results in uncontrolled elongation of single- and double-stranded
telomeric tracts74, as well as loss of viability and telomere fusion in yeast
arrested in the G1 phase of the cell cycle64. Chromosome end fusions
are also a consequence of loss of Pot1, a single-stranded telomere-
binding protein, which is a homologue of the α-subunit of the O. nova
end-binding factor75. S. pombe cells lacking Pot1 rapidly lose telomeric
and subtelomeric DNA and become inviable, although a subset of cells
survive and have circularized chromosomes. Human POT1 regulates
telomere length in a telomerase-dependent manner, relaying information
from shelterin to the telomerase complex75–77. Mouse cells have two
POT1 proteins, POT1A and POT1B, suggesting recent expansion of the
telomeric complex in rodents78,79. POT1A is sufficient to repress DNA-
damage signalling at telomeres in the absence of POT1B. However, lack
of POT1B leads to a substantial increase in the length of the G tail in a
telomerase-independent manner, a phenotype that could not be rescued
by overexpression of POT1A, establishing that the two proteins have
distinct roles in telomere protection. In addition to POT1, the mam-
malian factor TRF2 has a substantial role in protecting telomeres from
NHEJ and homologous recombination. Telomeres in mouse and human
cells that lack TRF2 lose the G tail, are detected as damage sites, and are
substrates for DNA-ligase-IV-dependent fusion9,66,80. However, telomeres
in cells that lack both DNA ligase IV and TRF2 do not show overhang
loss, although they are still recognized as DNA damage, indicating that
the loss of the G tail is a consequence of NHEJ and is not required for the
DNA-damage pathways at telomeres80. In addition, overexpression of
a human TRF2 allele encoding a protein that lacks the amino termi-
nus but retains the DNA-binding region causes the rapid shortening
of telomeres and the generation of extrachromosomal telomeric DNA
circles81. These circles can readily be detected in cells in which the ALT
pathway is engaged (Box 1) and depend on the hom ologous-recombi-
nation protein XRCC3, which is involved in the resolution of Holliday
junctions (junctions between four DNA strands), suggesting that TRF2
protects telomeres not only from NHEJ but also from homologous-
recombination-based deletion of large stretches of DNA.
It is now well established that the natural chromosome ends need to
be protected from inappropriate repair, so it seems paradoxical that sev-
eral proteins involved in the detection or repair of DNA lesions local-
ize to telomeres. Such localizing proteins include the MRX/N complex
(Mre11–Rad50–Xrn2 in yeast and MRE11–RAD50–NBS1 in mammals)
and the protein kinases Tel1 and Mec1 in S. cerevisiae or their mammalian
homologues, ATM and ATR, all of which have crucial roles in DNA-
damage signalling and telomeric integrity. Deletion of MRE11 or RAD50
in S. cerevisiae leads to gradual telomere shortening, which does not
increase when EST2 is also deleted, placing the MRX complex in the same
pathway as telomerase82, probably recruiting the enzyme to telomeres83.
The finding that MRX localizes to telomeres in late S phase suggests that
this complex is required to prepare the telomere for telomerase-dependent
replication84. Co-deletion of MRN components and rad3 (the hom ologue
of MEC1 and ATR) in S. pombe results in telomere loss and chromosome
circularization85, again placing the complex in a pathway that provides
protection against telomere loss. Similarly, Tel1 and Mec1 can be found
at S. cerevisiae telomeres86, and deletion of the genes encoding both
molecules causes telomere loss, a phenotype that can be overcome by
attracting active telomerase to the chromosome ends87. At the same time,
Tel1 protects telomeres against NHEJ-dependent fusion, because the fre-
quency of telomeres fused to an inducible double-strand break increases
sharply in a strain that lacks both Tel1 and Tlc1 (ref. 88). In humans, cells
derived from patients with cancer-prone syndromes such as the Nijmegen
breakage syndrome or ataxia-telangiectasia (which carry mutations in
NBS1 and ATM, respectively) show accelerated telomere shortening and
chromosome fusions89,90. Similarly to yeast, the MRN complex and ATM
have been detected at telomeres91,92, suggesting dual roles in DNA-damage
signalling and telomere maintenance for these factors.
All of these observations indicate that functional telomeres require
interaction with DNA-damage repair proteins, suggesting that the DNA-
damage repair machinery is involved in replication of telomeres, protec-
tion of functional chromosome ends, and detection of, and signalling
from, dysfunctional ones.
Similarities between the proteins responsible for the detection and
repair of DNA lesions and those found at functional and dysfunctional
telomeres suggest that, for a cell, the difference between a DNA break
and a telomere is less pronounced than previously assumed. The find-
ing that functional telomeres are detected by the DNA-damage repair
machinery in every cell cycle, and the presence of the homologous-
recombination machinery at telomeres in G2 phase, suggests that this
pathway is involved in telomere end processing92,93 (Box 2).
Formation of a functional
5′ to 3′ nucleolytic resection
Figure 3 | End replication and processing. There are 3ʹ G overhangs at both
ends of the chromosome, and these are thought to be generated by 5ʹ to 3ʹ
nucleolytic activity. Semi-conservative replication of telomeres generates a
blunt-ended leading-strand product and a lagging-strand product with
a short overhang. Nucleolytic digestion (pink) in the 5ʹ to 3ʹ direction
then generates G overhangs, which allow the formation of a functional
telomere structure (not shown). The short overhang generated by lagging-
strand synthesis could be sufficient for a functional telomere, so it has
been proposed that only the leading-strand product undergoes nucleolytic
NATURE|Vol 447|21 June 2007
The repair of a double-strand break (DSB)
by the homologous-recombination repair
pathway involves well-characterized
molecular steps and proteins (see figure,
part a). After a DSB is detected, the MRN
complex (MRE11–NBS1–RAD50) is one of the
first repair factors to be recruited, and this
is followed by MRN-dependent activation
of the protein kinases ATM and ATR99.
Activation of these protein kinases results
in signals that lead to the recruitment of
processing factors, which generate 3ʹ single-
stranded overhangs. The exposed overhangs
are coated by replication protein A, which
protects the DNA against degradation and
inhibits the formation of secondary structures
(not shown). Next, the single-stranded DNA
invades homologous duplex DNA sequences,
forming a displacement loop (D-loop) by
homologous pairing and strand exchange, a
process catalysed by RAD51 and stimulated
by RAD52, RAD54 and RAD55–RAD57. Using
the homologous sequence as a template, the
invading strand primes DNA synthesis,
generating a Holliday junction. The genetic
information is restored after the Holliday
junction is cleaved by RAD51C–XRCC3-
dependent resolvase activity100.
Detection of ends, generation of 3ʹ overhangs
and invasion of homologous sequences are
essential steps for DSB repair, and these steps
seem strikingly similar to the predicted order
of post-replicative telomere processing and of
t-loop formation (see figure, part b). Recently,
we suggested that functional human telomeres
in primary fibroblasts interact with proteins of
the DNA-damage repair machinery92. MRE11,
NBS1 and activated ATM localize to telomeres
from late S phase until G2 phase, suggesting
that telomeres are detected as DNA damage.
The damage signal is localized to chromosome
ends, because neither stabilization of p53
nor phosphorylation of CHK2 (checkpoint 2
homologue; also known as CHEK2) can be
observed in the nucleus, suggesting that the
signal is well controlled, and does not lead to
cell-cycle arrest. All of these results promote
the hypothesis that telomeres are not always
hidden from the DNA-damage repair machinery
and suggest that telomeres require a close
relationship with the DNA-damage-response
pathways for function, fuelling the idea that
the processing of telomeres is similar to the
processing of DNA breaks.
In a manner analogous to that of DSB
processing, telomeres recruit RAD51, RAD52
and XRCC3 before mitosis, potentially resulting
in a search for homologous DNA sequences,
followed by strand invasion. However, invasion
of another chromosome by a telomere overhang
can lead to a deleterious phenotype. TRF2 is a
good candidate for involvement in avoidance
of inter-telomere invasion, because this protein
can form t-loops in vitro by keeping the telomere
end and the duplex DNA of the same telomere
in proximity. This model is supported by the
finding that in an in vitro assay with telomeric
substrates, not only is the homologous-
recombination machinery required for efficient
invasion but also the telomeric protein TRF2
(ref. 93). Consequently, this TRF2 activity,
together with the homologous-recombination
machinery, potentially facilitates the formation
of the D-loop that forms the core of the t-loop
structure and provides a substrate for POT1.
Box 2 | Telomere maintenance and double-strand-break repair
Recognition of DNA ends, and signalling
Replication of telomere ends
Generation of 3′ overhang
Loading of homologous-
recombination proteins and TRF2
Recognition and signalling
Holliday junction resolution
Generation of 3′ overhang
Loading of homologous-
5′ to 3′ exonuclease
NATURE|Vol 447|21 June 2007
The field of telomere biology has progressed considerably from the
simplistic view that telomeres function only as non-coding buffer zones
at the ends of linear chromosomes. We now view telomeres as highly
specialized and regulated complexes in which length and structure
determine integrity and function. Despite the open questions about
the requirements of t-loops for end protection or telomere length regu-
lation, it has been accepted that the G-rich 3ʹ single-stranded overhang
is required for telomere function.
However, the cell not only needs to control the formation of the
3ʹ overhang (and therefore the adoption of a functional telomere struc-
ture) but also needs to monitor telomere length, as demonstrated by the
finding that critically short telomeres lose their protective function.
A similar bipolar relationship is observed between telomeres and
DNA-damage-response pathways. On the one hand, the intracellular
DNA-damage repair machinery is required to detect dysfunctional
telomeres, which are consequently processed like any other double-
strand break. On the other hand, the DNA-damage repair machinery
is required for telomere replication and telomere protection, and it
also seems to be essential for the formation of a functional telomere
Taken together, all of these observations demonstrate that Watson and
Olovnikov were correct when they suggested a problem in the replica-
tion of terminal DNA. The telomere, ageing and cancer fields have since
managed to advance our understanding of the problem considerably,
although it has not yet been solved.
1. Griffith, J. D. et al. Mammalian telomeres end in a large duplex loop. Cell 97, 503–514
de Lange, T. T-loops and the origin of telomeres. Nature Rev. Mol. Cell Biol. 5, 323–329
Gottschling, D. E. & Zakian, V. A. Telomere proteins: specific recognition and protection of
the natural termini of Oxytricha macronuclear DNA. Cell 47, 195–205 (1986).
Price, C. M. & Cech, T. R. Telomeric DNA–protein interactions of Oxytricha macronuclear
DNA. Genes Dev. 1, 783–793 (1987).
Gray, J. T., Celander, D. W., Price, C. M. & Cech, T. R. Cloning and expression of genes for the
Oxytricha telomere-binding protein: specific subunit interactions in the telomeric complex.
Cell 67, 807–814 (1991).
Horvath, M. P., Schweiker, V. L., Bevilacqua, J. M., Ruggles, J. A. & Schultz, S. C. Crystal
structure of the Oxytricha nova telomere end binding protein complexed with single strand
DNA. Cell 95, 963–974 (1998).
de Lange, T. Shelterin: the protein complex that shapes and safeguards human telomeres.
Genes Dev. 19, 2100–2110 (2005).
van Steensel, B. & de Lange, T. Control of telomere length by the human telomeric protein
TRF1. Nature 385, 740–743 (1997).
van Steensel, B., Smogorzewska, A. & de Lange, T. TRF2 protects human telomeres from
end-to-end fusions. Cell 92, 401–413 (1998).
10. Smogorzewska, A. et al. Control of human telomere length by TRF1 and TRF2. Mol. Cell. Biol.
20, 1659–1668 (2000).
11. Karlseder, J. et al. Targeted deletion reveals an essential function for the telomere length
regulator Trf1. Mol. Cell. Biol. 23, 6533–6541 (2003).
12. Watson, J. D. Origin of concatemeric T7 DNA. Nature New Biol. 239, 197–201 (1972).
13. Olovnikov, A. M. A theory of marginotomy. J. Theor. Biol. 41, 181-190 (1973).
14. Hayflick, L. & Moorhead, P. S. The serial cultivation of human diploid cell strains. Exp. Cell
Res. 25, 585–621 (1961).
15. Lundblad, V. & Szostak, J. W. A mutant with a defect in telomere elongation leads to
senescence in yeast. Cell 57, 633–643 (1989).
16. Bodnar, A. G. et al. Extension of life-span by introduction of telomerase into normal human
cells. Science 279, 349–352 (1998).
17. Singer, M. S. & Gottschling, D. E. TLC1: template RNA component of Saccharomyces
cerevisiae telomerase. Science 266, 404–409 (1994).
18. 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).
19. Lingner, J. et al. Reverse transcriptase motifs in the catalytic subunit of telomerase. Science
276, 561–567 (1997).
20. Harley, C. B., Futcher, A. B. & Greider, C. W. Telomeres shorten during ageing of human
fibroblasts. Nature 345, 458–460 (1990).
21. Counter, C. M. et al. Telomere shortening associated with chromosome instability is
arrested in immortal cells which express telomerase activity. EMBO J. 11, 1921–1929
22. Shay, J. W. & Wright, W. E. Hayflick, his limit, and cellular ageing. Nature Rev. Mol. Cell Biol.
1, 72–76 (2000).
23. de Lange, T. Protection of mammalian telomeres. Oncogene 21, 532–540 (2002).
24. Sandell, L. L. & Zakian, V. A. Loss of a yeast telomere: arrest, recovery, and chromosome
loss. Cell 75, 729–739 (1993).
25. Smogorzewska, A. & de Lange, T. Different telomere damage signaling pathways in human
and mouse cells. EMBO J. 21, 4338–4348 (2002).
26. Karlseder, J., Smogorzewska, A. & de Lange, T. Senescence induced by altered telomere
state, not telomere loss. Science 295, 2446–2449 (2002).
27. Takai, H., Smogorzewska, A. & de Lange, T. DNA damage foci at dysfunctional telomeres.
Curr. Biol. 13, 1549–1556 (2003).
28. d’Adda di Fagagna, F. et al. A DNA damage checkpoint response in telomere-initiated
senescence. Nature 426, 194–198 (2003).
29. de Lange, T. & Jacks, T. For better or worse? Telomerase inhibition and cancer. Cell 98,
30. Seger, Y. R. et al. Transformation of normal human cells in the absence of telomerase
activation. Cancer Cell 2, 401–413 (2002).
31. Artandi, S. E. & DePinho, R. A. A critical role for telomeres in suppressing and facilitating
carcinogenesis. Curr. Opin. Genet. Dev. 10, 39–46 (2000).
32. Reddel, R. R. & Bryan, T. M. Alternative lengthening of telomeres: dangerous road less
travelled. Lancet 361, 1840–1841 (2003).
33. Dunham, M. A., Neumann, A. A., Fasching, C. L. & Reddel, R. R. Telomere maintenance by
recombination in human cells. Nature Genet. 26, 447–450 (2000).
34. Varley, H., Pickett, H. A., Foxon, J. L., Reddel, R. R. & Royle, N. J. Molecular characterization
of inter-telomere and intra-telomere mutations in human ALT cells. Nature Genet. 30,
35. Morales, C. P. et al. Absence of cancer-associated changes in human fibroblasts
immortalized with telomerase. Nature Genet. 21, 115–118 (1999).
36. Cooke, H. J. & Smith, B. A. Variability at the telomeres of the human X/Y pseudoautosomal
region. Cold Spring Harb. Symp. Quant. Biol. 51, 213–219 (1986).
37. Herbig, U., Ferreira, M., Condel, L., Carey, D. & Sedivy, J. M. Cellular senescence in aging
primates. Science 311, 1257 (2006).
38. Joeng, K. S., Song, E. J., Lee, K. J. & Lee, J. Long lifespan in worms with long telomeric DNA.
Nature Genet. 36, 607–611 (2004).
39. Raices, M., Maruyama, H., Dillin, A. & Karlseder, J. Uncoupling of longevity and telomere
length in C. elegans. PLoS Genet. 1, e30 (2005).
40. Teixeira, M. T., Arneric, M., Sperisen, P. & Lingner, J. Telomere length homeostasis is
achieved via a switch between telomerase-extendible and -nonextendible states. Cell 117,
41. Cristofari, G. & Lingner, J. Telomere length homeostasis requires that telomerase levels are
limiting. EMBO J. 25, 565–574 (2006).
42. Blasco, M. A. et al. Telomere shortening and tumor formation by mouse cells lacking
telomerase RNA. Cell 91, 25–34 (1997).
43. Lee, H. W. et al. Essential role of mouse telomerase in highly proliferative organs. Nature
392, 569–574 (1998).
44. Liu, Y. et al. The telomerase reverse transcriptase is limiting and necessary for telomerase
function in vivo. Curr. Biol. 10, 1459–1462 (2000).
45. Fan, X. & Price, C. M. Coordinate regulation of G- and C strand length during new telomere
synthesis. Mol. Biol. Cell 8, 2145–2155 (1997).
46. Diede, S. J. & Gottschling, D. E. Telomerase-mediated telomere addition in vivo requires
DNA primase and DNA polymerases α and δ. Cell 99, 723–733 (1999).
47. Pennock, E., Buckley, K. & Lundblad, V. Cdc13 delivers separate complexes to the telomere
for end protection and replication. Cell 104, 387–≠96 (2001).
48. Bianchi, A., Negrini, S. & Shore, D. Delivery of yeast telomerase to a DNA break depends on
the recruitment functions of Cdc13 and Est1. Mol. Cell 16, 139–146 (2004).
49. Chandra, A., Hughes, T. R., Nugent, C. I. & Lundblad, V. Cdc13 both positively and
negatively regulates telomere replication. Genes Dev. 15, 404–414 (2001).
50. Grossi, S., Puglisi, A., Dmitriev, P. V., Lopes, M. & Shore, D. Pol12, the B subunit of DNA
polymerase α, functions in both telomere capping and length regulation. Genes Dev. 18,
51. Nakamura, M., Nabetani, A., Mizuno, T., Hanaoka, F. & Ishikawa, F. Alterations of DNA and
chromatin structures at telomeres and genetic instability in mouse cells defective in DNA
polymerase α. Mol. Cell. Biol. 25, 11073–11088 (2005).
52. Rhodes, D. & Giraldo, R. Telomere structure and function. Curr. Opin. Struct. Biol. 5, 311–322
53. Miller, K. M., Rog, O. & Cooper, J. P. Semi-conservative DNA replication through telomeres
requires Taz1. Nature 440, 824–828 (2006).
54. Crabbe, L., Verdun, R. E., Haggblom, C. I. & Karlseder, J. Defective telomere lagging strand
synthesis in cells lacking WRN helicase activity. Science 306, 1951–1953 (2004).
55. Chang, S. et al. Essential role of limiting telomeres in the pathogenesis of Werner
syndrome. Nature Genet. 36, 877–882 (2004).
56. Jacob, N. K., Skopp, R. & Price, C. M. G-overhang dynamics at Tetrahymena telomeres.
EMBO J. 20, 4299–4308 (2001).
57. Wellinger, R. J., Ethier, K., Labrecque, P. & Zakian, V. A. Evidence for a new step in telomere
maintenance. Cell 85, 423–433 (1996).
58. Dionne, I. & Wellinger, R. J. Cell cycle-regulated generation of single-stranded G-rich DNA
in the absence of telomerase. Proc. Natl Acad. Sci. USA 93, 13902–13907 (1996).
59. Makarov, V. L., Hirose, Y. & Langmore, J. P. Long G tails at both ends of human
chromosomes suggest a C strand degradation mechanism for telomere shortening. Cell
88, 657–666 (1997).
60. Chai, W., Du, Q., Shay, J. W. & Wright, W. E. Human telomeres have different overhang
sizes at leading versus lagging strands. Mol. Cell 21, 427–435 (2006).
61. Jacob, N. K., Kirk, K. E. & Price, C. M. Generation of telomeric G strand overhangs involves
both G and C strand cleavage. Mol. Cell 11, 1021–1032 (2003).
62. Sfeir, A. J., Chai, W., Shay, J. W. & Wright, W. E. Telomere-end processing the terminal
nucleotides of human chromosomes. Mol. Cell 18, 131–138 (2005).
63. Hockemeyer, D., Sfeir, A. J., Shay, J. W., Wright, W. E. & de Lange, T. POT1 protects
telomeres from a transient DNA damage response and determines how human
chromosomes end. EMBO J. 24, 2667–2678 (2005).
64. Ferreira, M. G. & Cooper, J. P. The fission yeast Taz1 protein protects chromosomes from
Ku-dependent end-to-end fusions. Mol. Cell 7, 55–63 (2001).
65. Mieczkowski, P. A., Mieczkowska, J. O., Dominska, M. & Petes, T. D. Genetic regulation of
telomere–telomere fusions in the yeast Saccharomyces cerevisae. Proc. Natl Acad. Sci. USA
100, 10854–10859 (2003).
66. Smogorzewska, A., Karlseder, J., Holtgreve-Grez, H., Jauch, A. & de Lange, T. DNA ligase
IV-dependent NHEJ of deprotected mammalian telomeres in G1 and G2. Curr. Biol. 12, 1635
NATURE|Vol 447|21 June 2007
67. Fisher, T. S., Taggart, A. K. & Zakian, V. A. Cell cycle-dependent regulation of yeast Download full-text
telomerase by Ku. Nature Struct. Mol. Biol. 11, 1198–1205 (2004).
68. Fisher, T. S. & Zakian, V. A. Ku: a multifunctional protein involved in telomere maintenance.
DNA Repair (Amst.) 4, 1215–1226 (2005).
69. Lin, J. J. & Zakian, V. A. The Saccharomyces CDC13 protein is a single-strand TG1–3 telomeric
DNA-binding protein in vitro that affects telomere behavior in vivo. Proc. Natl Acad. Sci. USA
93, 13760–13765 (1996).
70. Grandin, N., Reed, S. I. & Charbonneau, M. Stn1, a new Saccharomyces cerevisiae protein, is
implicated in telomere size regulation in association with Cdc13. Genes Dev. 11, 512–527
71. Grandin, N., Damon, C. & Charbonneau, M. Ten1 functions in telomere end protection
and length regulation in association with Stn1 and Cdc13. EMBO J. 20, 1173–1183
72. Garvik, B., Carson, M. & Hartwell, L. Single-stranded DNA arising at telomeres in cdc13
mutants may constitute a specific signal for the RAD9 checkpoint. Mol. Cell. Biol. 15,
6128–6138 (1995); erratum 16, 457 (1996).
73. Booth, C., Griffith, E., Brady, G. & Lydall, D. Quantitative amplification of single-stranded
DNA (QAOS) demonstrates that cdc13-1 mutants generate ssDNA in a telomere to
centromere direction. Nucleic Acids Res. 29, 4414–4422 (2001).
74. Cooper, J. P., Nimmo, E. R., Allshire, R. C. & Cech, T. R. Regulation of telomere length and
function by a Myb-domain protein in fission yeast. Nature 385, 744–747 (1997).
75. Baumann, P. & Cech, T. R. Pot1, the putative telomere end-binding protein in fission yeast
and humans. Science 292, 1171–1175 (2001).
76. Loayza, D. & de Lange, T. POT1 as a terminal transducer of TRF1 telomere length control.
Nature 423, 1013–1018 (2003).
77. Colgin, L. M., Baran, K., Baumann, P., Cech, T. R. & Reddel, R. R. Human POT1 facilitates
telomere elongation by telomerase. Curr. Biol. 13, 942–946 (2003).
78. Hockemeyer, D., Daniels, J. P., Takai, H. & de Lange, T. Recent expansion of the telomeric
complex in rodents: two distinct POT1 proteins protect mouse telomeres. Cell 126, 63–77
79. Wu, L. et al. Pot1 deficiency initiates DNA damage checkpoint activation and aberrant
homologous recombination at telomeres. Cell 126, 49–62 (2006).
80. Celli, G. B. & de Lange, T. DNA processing is not required for ATM-mediated telomere
damage response after TRF2 deletion. Nature Cell Biol. 7, 712–718 (2005).
81. Wang, R. C., Smogorzewska, A. & de Lange, T. Homologous recombination generates
T-loop-sized deletions at human telomeres. Cell 119, 355–368 (2004).
82. Nugent, C. I. et al. Telomere maintenance is dependent on activities required for end repair
of double-strand breaks. Curr. Biol. 8, 657–660 (1998).
83. Tsukamoto, Y., Taggart, A. K. & Zakian, V. A. The role of the Mre11–Rad50–Xrs2 complex
in telomerase-mediated lengthening of Saccharomyces cerevisiae telomeres. Curr. Biol. 11,
84. Takata, H., Tanaka, Y. & Matsuura, A. Late S phase-specific recruitment of Mre11 complex
triggers hierarchical assembly of telomere replication proteins in Saccharomyces cerevisiae.
Mol. Cell 17, 573–583 (2005).
85. Nakamura, T. M., Moser, B. A. & Russell, P. Telomere binding of checkpoint sensor and
DNA repair proteins contributes to maintenance of functional fission yeast telomeres.
Genetics 161, 1437–1452 (2002).
86. Takata, H., Kanoh, Y., Gunge, N., Shirahige, K. & Matsuura, A. Reciprocal association of
the budding yeast ATM-related proteins Tel1 and Mec1 with telomeres in vivo. Mol. Cell 14,
87. Chan, S. W., Chang, J., Prescott, J. & Blackburn, E. H. Altering telomere structure allows
telomerase to act in yeast lacking ATM kinases. Curr. Biol. 11, 1240–1250 (2001).
88. Chan, S. W. & Blackburn, E. H. Telomerase and ATM/Tel1p protect telomeres from
nonhomologous end joining. Mol. Cell 11, 1379–1387 (2003).
89. Vaziri, H. et al. ATM-dependent telomere loss in aging human diploid fibroblasts and DNA
damage lead to the post-translational activation of p53 protein involving poly(ADP-ribose)
polymerase. EMBO J. 16, 6018–6033 (1997).
90. Ranganathan, V. et al. Rescue of a telomere length defect of Nijmegen breakage syndrome
cells requires NBS and telomerase catalytic subunit. Curr. Biol. 11, 962–966 (2001).
91. Zhu, X. D., Kuster, B., Mann, M., Petrini, J. H. & de Lange, T. Cell-cycle-regulated association
of RAD50/MRE11/NBS1 with TRF2 and human telomeres. Nature Genet. 25, 347–352
92. Verdun, R. E., Crabbe, L., Haggblom, C. & Karlseder, J. Functional human telomeres are
recognized as DNA damage in G2 of the cell cycle. Mol. Cell 20, 551–561 (2005).
93. Verdun, R. E. & Karlseder, J. The DNA damage machinery and homologous recombination
pathway act consecutively to protect human telomeres. Cell 127, 709–720 (2006).
94. Cesare, A. J. & Griffith, J. D. Telomeric DNA in ALT cells is characterized by free telomeric
circles and heterogeneous t-loops. Mol. Cell. Biol. 24, 9948–9957 (2004).
95. Bailey, S. M., Brenneman, M. A. & Goodwin, E. H. Frequent recombination in telomeric
DNA may extend the proliferative life of telomerase-negative cells. Nucleic Acids Res. 32,
96. Bechter, O. E., Shay, J. W. & Wright, W. E. The frequency of homologous recombination in
human ALT cells. Cell Cycle 3, 547–549 (2004).
97. McEachern, M. J. & Haber, J. E. Break-induced replication and recombinational telomere
elongation in yeast. Annu. Rev. Biochem. 75, 111–135 (2006).
98. Lundblad, V. Telomere maintenance without telomerase. Oncogene 21, 522–531 (2002).
99. Carson, C. T. et al. The Mre11 complex is required for ATM activation and the G2/M
checkpoint. EMBO J. 22, 6610–6620 (2003).
100. Haber, J. E. Partners and pathways repairing a double-strand break. Trends Genet. 16,
Acknowledgements We are indebted to V. Lundblad for constructive comments on
the manuscript, the National Institutes of Health for funding (J.K.) and the Leukemia
and Lymphoma Society for a long-term postdoctoral fellowship (R.E.V.).
Author Information Reprints and permissions information is available at
npg.nature.com/reprintsandpermissions. The authors declare no competing financial
interests. Correspondence should be addressed to J.K. (email@example.com).
NATURE|Vol 447|21 June 2007