Spontaneous telomere to telomere fusions occur in unperturbed fission yeast cells
Nucleic Acids Research
Spontaneous telomere to telomere fusions occur in unperturbed fission yeast cells
Hugo Almeida 0
Miguel Godinho Ferreira 0
0 Instituto Gulbenkian de Cieˆ ncia , Rua da Quinta Grande 6, 2780-156 Oeiras , Portugal
Telomeres protect eukaryotic chromosomes from illegitimate end-to-end fusions. When this function fails, dicentric chromosomes are formed, triggering breakage-fusion-bridge cycles and genome instability. How efficient is this protection mechanism in normal cells is not fully understood. We created a positive selection assay aimed at capturing chromosome-end fusions in Schizosaccharomyces pombe. We placed telomere sequences with a head to head arrangement in an intron of a selectable marker contained on a plasmid. By linearizing the plasmid between the telomere sequences, we generated a stable mini-chromosome that fails to express the reporter gene. Whenever the ends of the mini-chromosome join, the marker gene is reconstituted and fusions are captured by direct selection. Using telomerase mutants, we recovered several fusion events that lacked telomere sequences. The end-joining reaction involved specific homologous subtelomeric sequences capable of forming hairpins, suggestive of ssDNA stabilization prior to fusing. These events occurred via microhomology-mediated end-joining (MMEJ)/ single-strand annealing (SSA) repair and also required MRN/Ctp1. Strikingly, we were able to capture spontaneous telomere-to-telomere fusions in unperturbed cells. Similar to disruption of the telomere regulator Taz1/TRF2, end-joining reactions occurred via non-homologous end-joining (NHEJ) repair. Thus, telomeres undergo fusions prior to becoming critically short, possibly through transient deprotection. These dysfunction events induce chromosome instability and may underlie early tumourigenesis.
The ends of eukaryotic chromosomes prevent DNA
double-strand break (DSB) repair through a specialized
protective structure called the telomere. Telomeres are
composed of G-rich DNA repeats bound by a protein
complex known as shelterin (
). In the absence of
shelterin, telomeres undergo end-to-end fusions that give
rise to genomic instability (
). Several studies suggest
that genomic instability initiated by telomere dysfunction
may underlie carcinogenesis (6).
The functions of telomere protection have been
dissected in several organisms from yeast to humans. One
major function is to compensate for the incomplete
replication of the ends of chromosomes, known as the ‘end
replication problem’ (
). This is generally achieved by
telomerase, a reverse transcriptase that adds new telomere
repeats to the ends of chromosomes (
). In the fission
yeast Schizosaccharomyces pombe, telomere elongation by
telomerase is regulated by Taz1, orthologue of both
human TRF1 and TRF2, also involved in
chromosomeend protection (
). In absence of Taz1,
telomereto-telomere fusions occur via non-homologous
endjoining (NHEJ) repair, a mode of DSB repair up-regulated
in the G1 phase of the cell cycle (
disruption of either human TRF2 or budding yeast Rap1
results in telomere-to-telomere fusions perpetrated by
NHEJ repair (
). During DNA replication, the
replication fork unwinds the telomere and exposes
chromosome ends. While this permits the engagement of
telomerase, it also causes momentarily deprotection that
triggers DNA damage checkpoints (
Although loss of the telomere repeat factor Taz1 triggers
NHEJ-mediated fusions, erosion of telomeric DNA leads
to a different mode of DNA repair. In the absence of
telomerase, Stn1, Ten1 or Pot1—components that protect the
30-overhangs from degradation (
undergo fusions through a DNA repair process that
requires exposure of complementary ssDNA regions on
both DNA ends (
). Depending on the size of the
homology region involved, this process has been termed
single-strand annealing (SSA) or
microhomologymediated end-joining (MMEJ) repair (
). In fission
yeast, these fusions normally occur 7–13 kb internally
after complete disappearance of telomere sequences (
In mouse cells, a microhomology-dependent mechanism of
DNA repair termed A-NHEJ also mediates fusions after
removal of Pot1a/b-Tpp1 (
). Similar results were
obtained for fusions observed in telomerase-deficient
budding yeast and mice models (
fusions were also observed in telomerase-positive human
cell lines (23). Most involve critically short telomeres,
suggesting that telomeres fuse after eroding past a minimum
length required for protection. In contrast to these studies,
wild-type length telomeres can engage end fusions.
Gottschling and colleagues used a positive selection assay
in budding yeast based on an induced DSB and observed
extremely rare telomere-to-DSB fusions in WT cells,
highlighting that normal telomeres are protected from engaging
in NHEJ repair with other DNA breaks (
Here we investigate the nature of telomere protection
in fission yeast using a novel quantitative telomere
fusion assay. We find that telomerase mutants show
microhomology-dependent fusions possibly mediated by
transient DNA hairpin structures. In contrast to
previous observations, we show that wild-type telomeres
are subjected to spontaneous fusions via NHEJ, which do
not require critical telomere shortening. Thus, functional
telomeres can be involved in spontaneous telomere
fusions, which may trigger genomic instability preceding
MATERIALS AND METHODS
Mini-chromosome and plasmid construction
The his3+ gene was cloned into pBluescriptII KS(+) using
a blunt-ended NotI/SmaI digestion. A synthetic 100 mer
oligonucleotide containing subtelomeric sequences in a
head-to-head orientation, derived from the pNSU70
) separated by an ApaI restriction fragment
was cloned into the XbaI restriction site of his3+. The
resulting his3+ fragment was then removed from
pBluescriptII using PstI and SacI and cloned into the
PstI and SacI sites of pREP3X. The SacI site was
subsequently destroyed by digestion and blunt-end
ligation to generate pREP-his-telo. Subsequently, an ApaI
restriction fragment containing opposing telomere
sequences separated by an Escherichia coli kanR
resistance gene was derived from the pEN53 plasmid (
cloned into the ApaI site, thus spacing the subtelomeric
sequences of pREP3X-his-telo to give pFT2. Linearization
of the plasmid was achieved by removing the
SacIenclosed kanR sequence in between the telomere
The his3+ gene was cloned into pREP3X using PstI and
SacI sites to generate pREP-his3+.
pREP42- taz1+ was used in the construction of the taz1+
o/e strain. The taz1+ORF was cloned in a pTOPO vector
(Invitrogen). pTOPO was then digested with BamHI and
SalI and the resulting taz1+fragment was cloned into the
BamHI and SalI sites of pREP42 vector.
Media and genetic methods
We used standard recipes for Yeast Extract with
supplements (YES) and Edinburgh minimal medium
(EMM) as described in (
). Methods of transformation,
sporulation and tetrad dissection are described in (
Strains used in this study are described in Supplementary
Table S1. Lines derived from genetic crosses are indicated
in the parental column with both parental strains
indicated. All deletions were performed using the
procedures described in (
To create strain MGF1898, pREP42-taz1+ was digested
with KpnI and transformed for integration. Insertion at
the taz1+ locus was confirmed by PCR.
Primers used for amplification and sequencing of captured
fusions were as follows: 495:GAACTTCAGCCTTATCG
CTG; 496: CCACGGAAATAACCGAACCA; 613: GGG
TAATAATTGATATGAGGGC; 614: CCACGGAAAT
AACCGAACCA. Primers used for confirmation of
insertion of pREP42-taz1+ fragment at the taz1+ locus were
as follows: 178: TTCGCCTCGACATCATCTGC; 375:
CCTCAGTGGCAAATCCTAAC; 711: TCTTTTACAG
Serial dilution assays
Cultures were grown at 32 C to logarithmic phase and
re-suspended to a cell density of 1 107 cells/ml.
Tenfold serial dilutions were performed in EMM media, and
5 ml of each dilution were spotted onto EMM plates with
the described supplements and incubated at 32 C up to 7
Telomere fusion assays
Cultures were grown at 32 C to logarithmic phase in
EMM media containing Leucine and plated on solid
EMM media containing Leucine ± Histidine. Plates were
incubated at 32 C up to 7 days and colonies were counted.
For time course experiments, cultures were diluted at each
time point to 5 105 cells/ml and incubated the
appropriate times in EMM media containing Leucine ± Histidine.
Cells were lysed in 20% TCA and the resulting protein
extracts were resolved by SDS-PAGE, transferred to
PVDF membranes (GE Healthcare RPN203D) and
probed with primary a-Taz1 antibody (a gift from Julia
P. Cooper). After incubation with HRP-conjugated
secondary antibodies (GE Healthcare NA934), bands were
visualized using ECLPlus (GE Healthcare RPN2132) in
a STORM scanner (Amersham). Quantification was
performed using ImageJ software.
Genomic DNA was obtained from exponentially growing
cells in YES or EMM media with supplements, using the
phenolic extraction method described (
) and digested
with the appropriate restriction enzymes. Southern blot
analysis was performed as described (
). Briefly, DNA
was separated in 0.8% agarose gels and transferred by
capillarity to genomic blotting membranes (Bio-Rad,
#162-0196). The DNA was then cross-linked using UV
radiation and the membranes were hybridized using
Church–Gilbert solution at 65 C. An overnight
incubation was performed with a telomere repeat probe or a
rad4+ genomic probe labelled with 32P using the Prime-it
II random primer labeling kit (Stratagene). The membrane
was washed for 30 min at 65 C with a 1 SSC 1% SDS
solution and exposed to a PhosphoImager screen
(Amersham) for 1–3 days depending on signal strength.
The PhosphoImager screen was scanned with a STORM
scanner (Amersham). Telomere length was calculated by
normalizing molecular weight with telomere signal
intensity, as described in (
). Number of pFT2 copies per cell
was calculated by dividing signal intensity from ars1
fragment present in pFT2 with the signal intensity from
the endogenous ars1. This number was multiplied by a
factor of 0.3 to account for the measured fraction of
cells in a population under selection for LEU2 harbouring
pFT2 (See Supplementary Figure S2).
Mini-chromosome end fusions were amplified by PCR.
Forward primers (495: GAACTTCAGCCTTATCGC
TG; 613:GGGTAATAATTGATATGAGGGC) were
used for promoter proximal sequencing and the
reverse primers (496:CCACGGAAATAACCGAACCA;
614:CCACGGAAATAACCGAACCA) for terminator
Direct assay to capture telomere fusions
To devise a scheme to detect telomere fusions, we resorted
to the ability of introns to bear gene-unrelated sequences,
such as telomeres. We cloned the his3+ gene in a fission
yeast plasmid carrying a LEU2 marker, allowing for
double selection in media lacking both histidine and
leucine. Subsequently, we introduced two telomeres of
258 bp in a head to head arrangement in a unique site of
the second intron of his3+ thus creating pFT2 (Figure 1A).
These telomeres were flanked by 80 bp of subtelomeric
sequences on each side. The telomere sequences did not
impair his3+ expression and allowed for growth in media
lacking both leucine and histidine (pFT2-Cir, Figure 1B).
We then linearized pFT2 between the telomere sequences
and transformed WT cells, thus creating a
minichromosome in which the his3+ gene was split between
the two extremities of the plasmid (Figure 1A).
Consequently, the linear mini-chromosome pFT2 could
be maintained through selection using LEU2 expression
but was now unable to grow on media lacking histidine
(pFT2-Lin, Figure 1B). We reasoned that any event that
would disrupt telomere protection leading to plasmid end
joining would restore the initial circular configuration of
pFT2 and allow for his3+ expression. We also asserted
that intron length would not be limiting in our assay by
observing the unimpaired growth of strains harbouring
circular pFT2 carrying a 1.3 kbp Kanr cassete in the
his3+ intron (data not shown). As long as the
his3+coding sequence is not eroded, this system identifies any
type of end joining to occur between the ends of the
mini-chromosome, as introns do not rely on a coding
We next gathered evidence that the mini-chromosome
was propagated linearly in cells and that telomeres were
functional. We produced genomic DNA from fission yeast
cells and digested it with either EcoRV that would release
the his3+ promoter proximal telomere (T Prom) or XhoI
that cuts near the his3+ terminator, generating a fragment
containing the terminator proximal telomere (T Term;
Supplementary Figure S1). We next performed Southern
blotting using a telomere probe that revealed the telomere
sequences present in the cell’s chromosomes and in the
newly generated mini-chromosome (Figure 1C). Both
chromosome ends showed a distribution of sizes typical
of telomeres, suggesting that the ends of the
minichromosome were free and were being used as a substrate
To test the mini-chromosome stability and to confirm
that it was independent of the remaining chromosomes,
we analysed the ability of these cells to lose pFT2 in either
the linear or the circular form. Cells were grown in media
containing leucine and histidine for several generations
and then tested for the ability to retain pFT2. Fission
yeast plasmids lack centromeric sequences and are thus
missegregated and lost when not under selection (
We confirmed that pFT2 either in the circular or linear
configuration were lost at a similar rate as other plasmids
(Supplementary Figure S2). Thus, our mini-chromosome
was maintained independently of the remaining genome,
showing that its telomeres were able to recruit telomerase
and protect its ends.
We wanted to know whether we could use our assay to
analyse chromosome-end fusions. As a proof of principle,
we crossed an early generation telomerase mutant with the
strain harbouring the pFT2 mini-chromosome and
allowed the progeny to undergo telomere erosion over
several days. On each time point, we collected samples
for Southern blotting and plated cells on media containing
and lacking histidine to measure the ability to fuse the
ends of the mini-chromosome. Using a telomere probe,
we could observe the slow erosion of telomeres in trt1
mutants over several days (Figure 1D). Throughout this
experiment, control WT did not generate detectable
colonies in histidine-depleted media (Figure 1E).
However, by day 6, coincident with the lowest signal for
the mini-chromosome telomeres, trt1 cells showed a
steadily increase in his3+-expressing cells. Over the
course of the experiment, productive mini-chromosome
fusions increased to represent 5–8% of all the cells
harbouring the pFT2 plasmid. Given that several
mini-chromosome fusions are likely to occur in a way
that fail to reconstitute his3+ expression, we anticipate
that we are under-estimating the whole population of
Taz1 levels are limiting to control telomere length
We wanted to define the impact for the cell to harbour
more telomeres than the ones it usually carries. Haploid
fission yeast contains three chromosomes and spends most
of its cell cycle in S/G2 phases. Thus, cells carry a
maximum of 12 telomeres at any given time. Introducing
several other telomeres could alter telomere dynamics. To
investigate the impact of extra telomeres per cell, we first
quantified how many mini-chromosomes would a cell
carry on average. We used Southern blot analysis to
quantify the ars1 fragments, which are present as a
single copy both in the genome and in the
minichromosome (Supplementary Figure S3). Given that not
all cells in the population carry a mini-chromosome due to
natural plasmid loss, even when under selection, we
calculated that the mini-chromosomes would be in a
range of 5–7 copies per cell, thus increasing the overall
number of telomeres.
To assess the impact of extra copies of telomeres, we
looked at telomere length in cells carrying the
mini-chromosome and compared to control cells. Using
an ApaI restriction enzyme that is unable to discriminate
the source of the telomeres analysed, we observed that the
presence of the mini-chromosome gives rise to a 3- to
5-fold increase in overall telomere length (Figure 2A,
lane 2). We then used an EcoRI digest to discriminate
between the chromosome telomeres and those harboured
in the mini-chromosome. Southern blots verified that the
chromosome telomeres were elongated in cells carrying the
mini-chromosome (Figure 2B, lane 2). The
minichromosome was the cause for the changes in telomere
length, since telomere length decreased in cells that lost
the linear plasmid (Figure 2B, lane 3). The same increase
was observed in a strain carrying a circular configuration
of the mini-chromosome (Supplementary Figure S4B).
Thus, as a consequence of increasing the number of
telomeres in the cell, the overall telomere length is increased,
suggesting the titration of a limiting factor that controls
We aimed at having a plasmid fusion assay that mimics
WT cells as close as possible. Even though there is no
indication that small variation in length, such as the one
observed, results in deficiencies in telomere protection, we
set up to identify the possible regulator of telomere length
that was limiting. Using plasmids carrying the nmt1+
promoter, we independently over-expressed Taz1, Rap1,
Pot1 and Ccq1 in cells carrying the mini-chromosome
(data not shown). Only Taz1 over-expression resulted in
an almost complete reduction of telomere length to
wildtype length of all telomeres. The same result was observed
when we integrated a Taz1 over-expression cassette in the
genome of a strain, to which we called taz1+ o/e
(Figure 2B). This strain was used in all remaining
experiments. By western blotting, we estimated that Taz1 was
over-expressed about 2-fold when compared with WT
levels (Figure 2C). Thus, levels of the telomere binding
Taz1 are limiting in fission yeast and, when telomere
number increases, this results in net telomere elongation.
Accordingly, budding yeast’s telomere-binding factor
Rap1 was also limiting upon increase in telomere
trt1" mini-chromosome fusions are MRN dependent and
occur via SSA/MMEJ repair
Capturing telomere fusions residing on a plasmid allows
us to study not only their abundance but also the
mechanism whereby they arise. Previous studies revealed that
chromosome-end fusions in trt1 in fission yeast were a
consequence of the rad16+ (ScRAD1 and mammalian
XPF)-dependent SSA/MMEJ pathway (18). To test the
genetic requirements for mini-chromosome fusions, we
produced trt1 mutants in which we deleted specific key
factors regulating different DNA repair pathways. We
allowed trt1 , trt1 rad16 and trt1 lig4 mutants
carrying the mini-chromosome to undergo telomere
erosion and quantified the number of colonies expressing
histidine as a measure of end-joining events. Both trt1
and trt1 lig4 mutants produced substantial
minichromosome fusions (Figure 3A). This was consistent
with previous observations (
), in which Lig4-dependent
NHEJ repair was dispensable for generating trt1
survivors in fission yeast. In contrast, rad16+ was required
for trt1 mini-chromosome end fusions (Figure 3A).
These results demonstrate not only that the SSA/MMEJ
pathway is required for joining the ends of the
mini-chromosome, but also that the repair mechanisms
triggered on chromosomes are the ones responsible for
repairing the ends present in the mini-chromosome, thus
validating our assay.
In parallel, we investigated the role of MRN (Rad32/
Rad50/Nbs1)/Ctp1 (MRX/ScSAE2 or mammalian MRN/
CtIP) in generating mini-chromosome end fusions in
trt1 mutants. For this purpose, we established mutants
of each component of this complex and conjugated them
in a trt1 background. Absence of MRN/Ctp1 also failed
to produce trt1 survivors capable of expressing histidine
(Figure 3B). Thus, MRN and its nuclease partner Ctp1 are
required to process mini-chromosome ends for the SSA/
MMEJ repair reaction.
In contrast to TRF2 and taz1 mutants,
chromosomeend fusions in trt1 mutants occurred only after total
telomere erosion. We set out to investigate the sequences
present at the junction of the mini-chromosome. We
devised PCR reactions with primers on each side of the
intron and verified that all of the trt1
histidineproducing cells possessed an intact his3+ intron (n = 45),
denoting that they occurred between the cloned
subtelomeric sequences of the mini-chromosome. We
sequenced trt1 and trt1 lig4 fusion junctions and,
to our surprise, >90% involved a set of three
pentanucleotide G-rich sequences present within the
85 bp subtelomeric region (Figure 3C and D). Consistent
to what was previously reported (
), all but one of
the trt1 fusions involved telomere sequences. The
pentanucleotide sequences were arranged in a direct
repeat spanning 30 bp and were present at both ends of
the mini-chromosome (Figure 3C, Supplementary Figure
S5), two promoter proximal (depicted as ABC and DEF)
and one terminator proximal (GHI).
Not all pentanucleotides were equally involved in
fusions, and there was a clear preference for those with
C/I and F/I pairs (73.3% and 70.5% in trt1 and trt1
lig4 , respectively, Figure 3D). We wondered why fusions
comprising the pentanucleotide present in C, F and I were
dominant. We realized that the remaining other two
pentanucleotides formed an inverted repeat spaced by
14 bp that could be engaged in a hairpin structure
(Figure 3E). Conceivably, as telomeres erode, ssDNA is
generated at subtelomeric regions of the
mini-chromosome creating the opportunity for secondary DNA
structures such as hairpins. This may momentarily
stabilize the sequences involved in the hairpin, thus exposing the
adjacent pentanucleotide for SSA/MMEJ repair. This
model is consistent with the requirement of MRN/Ctp1
in trt1 fusions, since the complex would be engaged in
processing the hairpin as part of the DNA repair reaction
Telomere–telomere fusions occur in unperturbed WT cells via NHEJ repair
As we generated our telomere fusion assay, we wondered
about the rare colonies appearing in unperturbed WT cells
grown on media lacking histidine (Figure 1B). These
escapers originated from previous cells that carried the
linear mini-chromosome and did not express his3+,
suggesting that they were the outcome of unsuspected fusions
taking place in unperturbed cells. We measured the
frequency of these events. In taz1+ o/e cells, the frequency
of colonies expressing his3+ was 1.4 10 4 ± 0.5 10 4
(SEM, n = 3; Figure 4A). In contrast, cells with longer
telomeres in which Taz1 was limiting had a higher
reversion frequency of 3.4 10 4 ± 0.9 10 4 (SEM, n = 3),
suggesting that insufficient Taz1 levels to maintain
normal telomere length also resulted in reduced protection
from end fusions.
Chromosome-end fusions in unperturbed cells
expressing telomerase were previously reported in transformed
human cell lines (
). These were the result of completely
eroded telomeres fusing to a shorter telomere involving
microhomologies at the junction. To identify the
mechanism behind our mini-chromosome fusions, we first
investigated the genetic requirements underlying these
events. As previously, we deleted the genes involved in
cells containing the mini-chromosome and quantified the
ability of expressing the his3+ reporter gene. In contrast to
the observed mini-chromosome fusions in trt1 survivors,
rad16+-dependent SSA/MMEJ repair is dispensable in
taz1+ o/e (Figure 4A). However, mutations in key
elements of the NHEJ pathway, such as pku70 and
lig4 , reduced the number of his3+-expressing colonies
about 10-fold (Figure 4A). These results indicate that
end-fusions occurring in unperturbed cells are processed
by a distinct pathway to the ones caused by gradual
telomere erosion, suggesting that they are fundamentally
different uncapping events.
We next asked whether MRN/Ctp1 was required for
NHEJ-mediated fusions of mini-chromosome ends. In
contrast to budding yeast, MRN is dispensable for
NHEJ repair in S. pombe as measured on plasmid-based
). Surprisingly, mutants in all subunits of MRN,
including Ctp1, were deficient in joining the two ends of
the mini-chromosome by NHEJ repair (Figure 4B). Our
data suggests that, in contrast to NHEJ measured in
naked plasmid ends, MRN/Ctp1 is required to process
mini-chromosome ends before undergoing end-joining
The requirement of NHEJ repair for chromosome
fusions in unperturbed cells suggested that a different
event occurred at the ends of our mini-chromosome. To
attempt at identifying the incident that originated the
end-fusion, we sequenced the junction at his3+
mini-chromosomes. Almost all fusion events involved
two telomeres on each side of the junction (89.3%,
n = 28; Figure 4C and D and Supplementary Figure S6).
The remaining involved one telomere and a subtelomere
or two subtelomeres (7.1% and 3.6%, respectively). In all
cases, we did not observe clear microhomologies at the
junction, even though fusions often involve guanidine
bases, as expected from G-rich telomere sequences (data
not shown). Thus, in contrast to the previous studies in
immortalized humans cell lines (
), our data suggest that
uncapping events occur at telomere sequences that result
in end-fusions via NHEJ repair.
Telomere fusions occur between half-sized telomeres
Even though mini-chromosome fusions exhibited
substantial telomere sequences in a head to head arrangement, it
was not clear whether the end-joining event resulted from
one of them being too short to sustain telomere
protection. To investigate the telomere length distribution in
fusions, we organized telomeres in a length histogram
(Figure 5A). There was a 2-fold size difference between
the median telomere size at T Prom and T Term (T
Prom 232 bp and T Term 111 bp; interquartile range
(IQR) 92–291 bp and 29–147 bp, respectively). This was
consistent with the average telomere lengths exhibited by
the mini-chromosome, as analysed by Southern blot (T
Prom 415 bp and T Term 184 bp, Figure 2A and B).
Interestingly, a 2-fold difference between the size of T
Term and T prom was also observed for the strain with
deficient Taz1 levels (Figure 2B, lane 2). Thus, even
though the methods used provide measures of different
accuracy, our data suggest that most fusions occur
between telomeres of about half the normal length
observed in the linear configuration.
The difference in length between the promoter and
terminator telomeres could arise due to a telomerase
preference for one telomere over the other. This preference may
reflect transcription of both the T Prom and endogenous
telomeres, either through his3+ or TERRA transcription,
which does not occur at the T Term telomere. To
discriminate whether one telomere was more frequently engaged
by telomerase over the other, we compared the
degenerated telomere sequences of fission yeast present
at fusion junctions. As in most yeast species, S. pombe
telomerase is ‘faulty’ providing other nucleotides in
between the consensus (GGTTAC)n. This property
allowed us to recognize the original telomere sequence
and identify how much erosion telomeres suffered and
which new sequences were added (Figure 5B). We
suggest that telomerase has a preference for T Prom
since it had added repeats to most telomeres engaged in
fusions (26/28). In contrast, only about half of the T Term
telomeres have new sequences added (16/28). As a result,
telomere shortening is more evident on terminator than on
promoter proximal telomeres (Figure 5B). Median length
of original telomere was half as long on the terminator
proximal telomere (T Prom 101 bp and T Term 68 bp;
IQR 75–149 bp and 18–103 bp, respectively). These
differences could be explained by different affinity for
telomerase, such that the T Term telomere is elongated only half
of the times as the T Prom and thus erodes further
acquiring a smaller steady state length.
Why are telomeres engaging in end-joining reactions?
We reasoned that if one telomere would be critically
short, it would engage in fusions with other telomeres
independently of their size. Such was observed in budding
yeast for fusions between a DSB and telomeres (
However, this was not the case for our assay. By
plotting telomere length of T Prom against T Term, we
observed a positive correlation of sizes such that the
longer the telomere was on one side, the longer it was
on the other (R2 = 0.62, n = 28; Figure 5C). Thus, the
fusions observed cannot be explained by critical telomere
shortening at one end of the mini-chromosome. Instead,
telomeres appear to become stochastically deprotected,
thereby undergoing NHEJ repair with neighbour
Telomere erosion leading to critically short telomeres has
been widely implicated in cancer (
). Studies using
primary human cell lines have shown that telomeres
engage in end fusions upon continuous erosion (23).
Here, we show that unperturbed cells with seemingly
functional telomeres can also engage in chromosome-end
fusions. Telomere-to-telomere fusions are an undervalued
phenomenon owing to the difficulties in capturing and
sequencing these events. This partially stems from the
lack of quantitative positive detection assays (
One such assay using an induced DSB in budding yeast
provided an extremely low frequency of telomere-to-DSB
fusion events [8.4 10 8 (24)]. Since we found far more
frequent rate of telomere-to-telomere fusions (1.4 10 4),
it is tempting to speculate that spontaneous chromosome
fusions may be a relevant trigger of genome instability and
carcinogenesis. Consistent with this idea, recent evidence
show telomere-to-telomere fusions in human breast
). The disparity in frequencies between our
result and the one previously reported may simply reflect
differences in the experimental setup. Unlike our assay,
the telomere-to-DSB fusion assay relies on coinciding a
DSB with transient telomere deprotection. Since G1
phase is predominant in budding yeast asynchronous
cultures, a DSB generated during this stage may not be
available in late S phase when telomeres are replicated and
possibly deprotected. In contrast, our assay relies
exclusively on transient telomere deprotection and these events
occur synchronously by their own nature. Alternatively,
the higher frequency of telomere fusions measured in our
assay may simply reflect the artificial nature of the
reporter construct. Our mini-chromosome possesses
considerably short subtelomeric regions (ca. 80 bp) in
comparison with chromosomal subtelomeres that encompass
10 kbp. Moreover, telomeres were introduced within an
intron of a transcribed gene, a feature that may affect
The frequency of telomere fusions measured in our
assay is likely to under-represent the total amount of
those present in WT and trt1 mutants. Our system is
unable to detect unproductive mini-chromosome-end
fusions, i.e. all those that fail to regenerate his3+
expression. These could involve (i) the destruction of the coding
sequence or intronic splice sites of the reporter gene; (ii)
fusions between mini-chromosome and endogenous
chromosome ends; (iii) fusions between
minichromosomes that do not pair opposite ends or even (iv)
the generation of linear trt1 survivors that continuously
undergo homologous recombination (HR) at the
Telomere-to-telomere fusions require the NHEJ
pathway similar to what was found in taz1 , rap1 and
TRF2F/- mutants (
). They also depend on the
MRN complex, as was observed in TRF2F/- and taz1
). Because in addition to MRN, fusions
require the Ctp1 nuclease that is not present during G1
), these events are likely to occur later in the cell cycle,
perhaps during DNA replication when telomeres are
exposed. In fission yeast, taz1 telomeres are subject to
HR during S/G2 phase of the cell cycle (
dysfunctional telomeres can undergo NHEJ-mediated
fusions in S/G2 when HR is compromised (
anticipate that upon DNA replication, telomere protection may
occasionally be faulty. In late S phase, telomeres unfold to
accommodate the DNA replication machinery causing
momentary deprotection (
). This is corroborated by
DNA damage responses being initiated every S phase as
the replication fork reaches chromosome ends (
Taz1 levels may also be limiting during this period to
protect the newly duplicated telomeres. In support of
this hypothesis, we observed that a strain with insufficient
Taz1 undergoes telomere fusions at a higher frequency.
In contrast to unperturbed cells, chromosome fusions in
the absence of telomerase were mediated by SSA/MMEJ
repair. Choice of DSB repair may simply reflect telomere
length and sequence. Although shelterin-protected longer
telomeres prevent DNA-end 50-resection required for
SSA/MMEJ repair (
), ssDNA generated at critically
short telomere inhibits NHEJ repair (
microhomologies are largely absent from telomere
sequences involved in end-joining reactions [e.g. 50-(TTA
GGG)n/(CCCTAA)n-30]. Previous work in pot1 and
trt1 mutants showed extensive degradation of
chromosome ends up to 13 kb (
). In contrast, our
minichromosome ends in trt1 mutants fuse at specific
pentanucleotide microhomologies contiguous to telomeres
but not within the juxtaposed intronic region of his3+,
suggesting that intron sequences may be refractory to
end-joining reaction. We observed that microhomologies
were arranged in inverted repeats that, when resected,
could fold back and engage in DNA hairpin formation.
Fusions in trt1 mutants require the MRN complex and
Ctp1. Thus, hairpins may serve as stabilizing structures for
eroding DNA ends, which are subsequently processed by
MRN/Ctp1 and captured in end-fusion reactions (
MRN was similarly required for chromosome fusions
involving critically short telomeres in Arabidopsis and
human cells (
), and telomerase RNA mutants in
budding yeast (
). Alternatively, hairpins formed
during DNA replication can lead to end-joining reactions
). Recent evidence shows that DNA replication
plays a role in chromosome-end fusions in
Caenorhabditis elegans (61). However, we were unable to
detect fusions involving pentanucleotide repeats in WT
cells, suggesting they require telomere resection prior to
fusion. Future studies will reveal the role of these naturally
occurring subtelomeric DNA microhomologies in
Our mini-chromosome allowed us to assess end-to-end
fusions in several distinct genetic backgrounds.
Furthermore, we established that there are two distinct
pathways that generate chromosome-end fusions, which
are determined by the mechanism of uncapping, and
that both are likely to be of importance in the
development of genome instability and cancer.
Supplementary Data are available at NAR Online:
Supplementary Table 1 and Supplementary Figures 1–6.
We thank members of our laboratory for helpful
discussions. We are indebted to Julie Cooper for support at the
initial stages of this work and for kindly providing the
a-Taz1 antibody. We are grateful to Kurt Runge, Karel
Riha and Kazunori Tomita for critically reading our
manuscript. Miguel Godinho Ferreira is an HHMI
international early career scientist.
Portuguese Fundac¸ a˜ o para a Ci eˆncia e Tecnologia
Association for International Cancer Research [Ref:
06-396]. Funding for open access charge: Portuguese
Funda c¸a˜ o para a Ci eˆncia e Tecnologia [PTDC/
Conflict of interest statement. None declared.
1. Stewart , J.A. , Chaiken , M.F. , Wang , F. and Price , C.M. ( 2012 ) Maintaining the end: roles of telomere proteins in end-protection, telomere replication and length regulation . Mutat. Res. , 730 , 12 - 19 .
2. Cooper , J.P. , Nimmo , E.R. , Allshire , R.C. and Cech , T.R. ( 1997 ) Regulation of telomere length and function by a Myb-domain protein in fission yeast . Nature , 385 , 744 - 747 .
3. Hackett , J.A. , Feldser , D.M. and Greider , C.W. ( 2001 ) Telomere dysfunction increases mutation rate and genomic instability . Cell , 106 , 275 - 286 .
4. Sfeir ,A. and de Lange , T. ( 2012 ) Removal of shelterin reveals the telomere end-protection problem . Science , 336 , 593 - 597 .
5. de Lange, T. ( 2005 ) Shelterin: the protein complex that shapes and safeguards human telomeres . Genes Dev. , 19 , 2100 - 2110 .
6. Martı´ nez,P. and Blasco , M.A. ( 2010 ) Role of shelterin in cancer and aging . Aging Cell , 9 , 653 - 666 .
7. Lingner , J. , Cooper , J.P. and Cech , T.R. ( 1995 ) Telomerase and DNA end replication: no longer a lagging strand problem? Science , 269 , 1533 - 1534 .
8. Blackburn , E.H. and Greider , C.W. ( 1985 ) Identification of a specific telomere terminal transferase activity in Tetrahymena extracts . Cell , 43 , 405 - 413 .
9. Blackburn , E.H. , Greider , C.W. , Henderson , E. , Lee , M.S. , Shampay , J. and Shippen-Lentz , D. ( 1989 ) Recognition and elongation of telomeres by telomerase . Genome , 31 , 553 - 560 .
10. Nakamura , T.M. , Morin , G.B. , Chapman , K. , Weinrich , S.L. , Andrews , W. , Lingner , J. , Harley , C.B. and Cech , T.R. ( 1997 ) Telomerase catalytic subunit homologs from fission yeast and human . Science , 277 , 955 - 959 .
11. Ferreira , M.G. and Cooper , J.P. ( 2001 ) The fission yeast Taz1 protein protects chromosomes from Ku-dependent end-to-end fusions . Mol. Cell , 7 , 55 - 63 .
12. Miller , K.M. , Ferreira , M.G. and Cooper , J.P. ( 2005 ) Taz1, Rap1 and Rif1 act both interdependently and independently to maintain telomeres . EMBO J ., 24 , 3128 - 3135 .
13. Smogorzewska , A. , Karlseder , J. , Holtgreve-Grez , H. , Jauch ,A. and de Lange , T. ( 2002 ) DNA ligase IV-dependent NHEJ of deprotected mammalian telomeres in G1 and G2 . Curr. Biol ., 12 , 1635 - 1644 .
14. Pardo , B. and Marcand , S. ( 2005 ) Rap1 prevents telomere fusions by nonhomologous end joining . EMBO J ., 24 , 3117 - 3127 .
15. Teixeira , M.T. , Arneric , M. , Sperisen , P. and Lingner , J. ( 2004 ) Telomere length homeostasis is achieved via a switch between telomerase-extendible and-nonextendible states . Cell , 117 , 323 - 335 .
16. Verdun , R.E. and Karlseder , J. ( 2006 ) The DNA damage machinery and homologous recombination pathway act consecutively to protect human telomeres . Cell , 127 , 709 - 720 .
17. Martı´ n,V., Du , L.L. , Rozenzhak , S. and Russell , P. ( 2007 ) Protection of telomeres by a conserved Stn1-Ten1 complex . Proc.Natl. Acad. Sci. , 104 , 14038 - 14043 .
18. Wang , X. and Baumann , P. ( 2008 ) Chromosome fusions following telomere loss are mediated by single-strand annealing . Mol. Cell , 31 , 463 - 473 .
19. Decottignies , A. ( 2007 ) Microhomology-mediated end joining in fission yeast is repressed by Pku70 and relies on genes involved in homologous recombination . Genetics , 176 , 1403 - 1415 .
20. Nakamura , T.M. , Cooper , J.P. and Cech , T.R. ( 1998 ) Two modes of survival of fission yeast without telomerase . Science , 282 , 493 - 496 .
21. Rai , R. , Zheng, H. , He , H. , Luo , Y. , Multani , A. , Carpenter , P.B. and Chang , S. ( 2010 ) The function of classical and alternative non-homologous end-joining pathways in the fusion of dysfunctional telomeres . EMBO J ., 29 , 2598 - 2610 .
22. Hemann , M.T. , Strong , M.A. , Hao , L.Y. and Greider , C.W. ( 2001 ) The shortest telomere, not average telomere length, is critical for cell viability and chromosome stability . Cell , 107 , 67 - 77 .
23. Capper , R. , Britt-Compton , B. , Tankimanova , M. , Rowson , J. , Letsolo , B. , Man , S. , Haughton , M. and Baird , D.M. ( 2007 ) The nature of telomere fusion and a definition of the critical telomere length in human cells . Genes Dev. , 21 , 2495 - 2508 .
24. DuBois , M.L. , Haimberger , Z.W. , McIntosh , M.W. and Gottschling , D.E. ( 2002 ) A quantitative assay for telomere protection in Saccharomyces cerevisiae . Genetics , 161 , 995 - 1013 .
25. Sugawara , N. ( 1989 ) Thesis . Harvard University, Cambridge, MA.
26. Nimmo , E.R. , Cranston , G. and Allshire , R.C. ( 1994 ) Telomere-associated chromosome breakage in fission yeast results in variegated expression of adjacent genes . EMBO J ., 13 , 3801 .
27. Moreno , S. , Klar , A. and Nurse , P. ( 1991 ) Molecular genetic analysis of fission yeast Schizosaccharomyces pombe . Methods Enzymol ., 194 , 795 .
28. Ba¨ hler,J., Wu , C. , Longtine , M. , Shah , N. , McKenzie , A. III, Steever , A.B. , Wach , A. , Philippsen , P. and Pringle , J.R. ( 1998 ) Heterologous modules for efficient and versatile PCR-based gene targeting in Schizosaccharomyces pombe . Yeast , 14 , 943 - 951 .
29. Sato , M. , Dhut , S. and Toda , T. ( 2005 ) New drug-resistant cassettes for gene disruption and epitope tagging inSchizosaccharomyces pombe . Yeast , 22 , 583 - 591 .
30. Kimura , M. , Stone , R.C. , Hunt , S.C. , Skurnick , J. , Lu , X. , Cao , X. , Harley , C.B. and Aviv , A. ( 2010 ) Measurement of telomere length by the Southern blot analysis of terminal restriction fragment lengths . Nature Protocols , 9 , 1596 - 1607 .
31. Beach , D. and Nurse , P. ( 1981 ) High-frequency transformation of the fission yeast Schizosaccharomyces pombe . Nature , 290 , 140 - 142 .
32. Wiley, E.A. and Zakian , V.A. ( 1995 ) Extra telomeres, but not internal tracts of telomeric DNA, reduce transcriptional repression at Saccharomyces telomeres . Genetics , 139 , 67 - 79 .
33. Runge , K.W. and Zakian , V.A. ( 1989 ) Introduction of extra telomeric DNA sequences into Saccharomyces cerevisiae results in telomere elongation . Mol. Cell. Biol. , 9 , 1488 - 1497 .
34. Baumann , P. and Cech , T.R. ( 2000 ) Protection of telomeres by the Ku protein in fission yeast . Mol. Biol. Cell , 11 , 3265 - 3275 .
35. Lobachev , K.S. , Gordenin , D.A. and Resnick , M.A. ( 2002 ) The Mre11 complex is required for repair of hairpin-capped double-strand breaks and prevention of chromosome rearrangements . Cell , 108 , 183 - 193 .
36. Lengsfeld , B.M. , Rattray , A.J. , Bhaskara , V. , Ghirlando , R. and Paull,T.T. ( 2007 ) Sae2 Is an endonuclease that processes hairpin DNA cooperatively with the Mre11/Rad50/Xrs2 complex . Mol. Cell , 28 , 638 - 651 .
37. Manolis , K.G. , Nimmo , E.R. , Hartsuiker , E. , Carr , A.M. , Jeggo , P.A. and Allshire , R.C. ( 2001 ) Novel functional requirements for non-homologous DNA end joining in Schizosaccharomyces pombe . EMBO J ., 20 , 210 - 221 .
38. Artandi , S.E. , Chang , S. , Lee , S.L. , Alson , S. , Gottlieb , G.J. , Chin , L. and DePinho , R.A. ( 2000 ) Telomere dysfunction promotes non-reciprocal translocations and epithelial cancers in mice . Nature , 406 , 641 - 644 .
39. Shay , J.W. and Wright , W.E. ( 2011 ) Role of telomeres and telomerase in cancer . Semin. Cancer Biol ., 21 , 349 - 353 .
40. Letsolo , B.T. , Rowson , J. and Baird , D.M. ( 2010 ) Fusion of short telomeres in human cells is characterized by extensive deletion and microhomology, and can result in complex rearrangements . Nucleic Acids Res ., 38 , 1841 - 1852 .
41. Lin , T.T. , Letsolo , B.T. , Jones , R.E. , Rowson , J. , Pratt , G. , Hewamana , S. , Fegan , C. , Pepper , C. and Baird , D.M. ( 2010 ) Telomere dysfunction and fusion during the progression of chronic lymphocytic leukemia: evidence for a telomere crisis . Blood , 116 , 1899 - 1907 .
42. Tanaka , H. , Abe , S. , Huda , N. , Tu , L. , Beam , M.J. , Grimes , B. and Gilley , D. ( 2012 ) Telomere fusions in early human breast carcinoma . Proc. Natl Acad. Sci. , 109 , 14098 - 14103 .
43. McEachern , M.J. , Iyer , S. , Fulton ,T.B. and Blackburn , E.H. ( 2000 ) Telomere fusions caused by mutating the terminal region of telomeric DNA . Proc. Natl. Acad. Sci. , 97 , 11409 - 11414 .
44. Mieczkowski , P.A. , Mieczkowska , J.O. , Dominska , M. and Petes , T.D. ( 2003 ) Genetic regulation of telomere-telomere fusions in the yeast Saccharomyces cerevisae . Proc. Natl. Acad. Sci. , 100 , 10854 - 10859 .
45. Heacock , M. , Spangler , E. , Riha , K. , Puizina , J. and Shippen , D.E. ( 2004 ) Molecular analysis of telomere fusions in Arabidopsis: multiple pathways for chromosome end-joining . EMBO J ., 23 , 2304 - 2313 .
46. Lowden , M.R. , Meier , B. , Lee , T.W.s. , Hall , J. and Ahmed , S. ( 2008 ) End joining at caenorhabditis elegans telomeres . Genetics , 180 , 741 - 754 .
47. Bae , N.S. and Baumann , P. ( 2007 ) A RAP1/TRF2 Complex inhibits nonhomologous end-joining at human telomeric DNA ends . Mol. Cell , 26 , 323 - 334 .
48. Muraki , K. , Nyhan , K. , Han , L. and Murnane , J.P. ( 2012 ) Mechanisms of telomere loss and their consequences for chromosome instability . Front. Oncol. , 2 , 135 .
49. Dimitrova ,N. and de Lange , T. ( 2009 ) Cell cycle-dependent role of MRN at dysfunctional telomeres: ATM signaling-dependent induction of nonhomologous end joining (NHEJ) in G1 and resection-mediated inhibition of NHEJ in G2 . Mol. Cell . Biol., 29 , 5552 - 5563 .
50. Reis , C.C. , Batista , S. and Ferreira , M.G. ( 2012 ) The fission yeast MRN complex tethers dysfunctional telomeres for NHEJ repair . EMBO J. , 1 - 11 .
51. Limbo , O. , Chahwan , C. , Yamada ,Y., de Bruin , R.A.M. , Wittenberg , C. and Russell , P. ( 2007 ) Ctp1 is a cell-cycle-regulated protein that functions with Mre11 complex to control double-strand break repair by homologous recombination . Mol. Cell , 28 , 134 - 146 .
52. Rog , O. , Miller , K.M. , Ferreira , M.G. and Cooper , J.P. ( 2009 ) Sumoylation of RecQ helicase controls the fate of dysfunctional telomeres . Mol. Cell , 33 , 559 - 569 .
53. Verdun , R.E. and Karlseder , J. ( 2007 ) Replication and protection of telomeres . Nature , 447 , 924 - 931 .
54. Moser , B.A. , Subramanian , L. , Chang , Y.T. , Noguchi , C. , Noguchi , E. and Nakamura , T.M. ( 2009 ) Differential arrival of leading and lagging strand DNA polymerases at fission yeast telomeres . EMBO J ., 28 , 810 - 820 .
55. Pitt , C.W. and Cooper , J.P. ( 2010 ) Pot1 inactivation leads to rampant telomere resection and loss in one cell cycle . Nucleic Acids Res ., 38 , 6968 - 6975 .
56. Helmink , B.A. , Tubbs , A.T. , Dorsett , Y. , Bednarski , J.J. , Walker , L.M. , Feng , Z. , Sharma , G.G. , McKinnon , P.J. , Zhang,J., Bassing , C.H. et al. ( 2010 ) H2AX prevents CtIP-mediated DNA end resection and aberrant repair in G1-phase lymphocytes . Nature , 469 , 245 - 249 .
57. Tankimanova , M. , Capper , R. , Letsolo , B.T. , Rowson , J. , Jones , R.E. , Britt-Compton , B. , Taylor ,A.M. and Baird , D.M. ( 2012 ) Mre11 modulates the fidelity of fusion between short telomeres in human cells . Nucleic Acids Res ., 40 , 2518 - 2526 .
58. Chan , S.W.L. and Blackburn , E.H. ( 2003 ) Telomerase and ATM/ Tel1p protect telomeres from nonhomologous end joining . Mol. Cell , 11 , 1379 - 1387 .
59. Murnane , J.P. ( 2012 ) Telomere dysfunction and chromosome instability . Mutat. Res. , 730 , 28 - 36 .
60. Haber , J.E. and Debatisse , M. ( 2006 ) Gene amplification: yeast takes a turn . Cell , 125 , 1237 - 1240 .
61. Lowden , M.R. , Flibotte , S. , Moerman , D.G. and Ahmed , S. ( 2011 ) DNA synthesis generates terminal duplications that seal end-to-end chromosome fusions . Science , 332 , 468 - 471 .