ABSTRACT
Ku is a heterodimer of polypeptides of approximately 70 and 80 kDa (Ku70 and
Ku80, respectively) that binds to DNA ends. Mammalian cells lacking Ku are
defective in DNA double-strand break (DSB) repair and in site- specific V(D)J recombination. Here, we describe the identification and characterisation of
YKU80
, the gene for the
Saccharomyces cerevisiae
Ku80 homologue. Significantly, we find that
YKU80
disruption enhances the radiosensitivity of
rad52
mutant strains, suggesting that
YKU80
functions in a DNA DSB repair pathway that does not rely on homologous
recombination. Indeed, through using an
in vivo
plasmid rejoining assay, we find that
YKU80
plays an essential role in illegitimate recombination events that result in the
accurate repair of restriction enzyme generated DSBs. Interestingly, in the absence of
YKU80
function, residual repair operates through an error-prone pathway that results in recombination between short direct repeat
elements. This resembles closely a predominant DSB repair pathway in vertebrates. Together, our data suggest that multiple, evolutionarily conserved mechanisms for DSB repair exist in eukaryotes.
Furthermore, they imply that Ku binds to DSBs
in vivo
and promotes repair both by enhancing accurate DNA end joining and by suppressing alternative error-prone repair pathways. Finally, we report that
yku80
mutant yeasts display dramatic telomeric shortening, suggesting that, in
addition to recognising DNA damage, Ku also binds to naturally occurring
chromosomal ends. These findings raise the possibility that Ku protects
chromosomal termini from nucleolytic attack and functions as part of a
telomeric length sensing system.
DNA repair is of the utmost importance in maintaining the integrity and stability of the genome. Without an ability to mediate DNA repair
effectively, cells have a greatly elevated risk of acquiring mutations and, in
multicellular organisms, this can lead ultimately to tumourigenesis. Perhaps the most dangerous type of DNA damage is the double strand break (DSB), which is generated following exposure to ionizing radiation or to radio-mimetic chemicals. Consequently, eukaryotic cells have evolved several systems to recognise and repair this form of DNA damage. Work on the budding
yeast,
Saccharomyces cerevisiae
, has established that genes comprising the
RAD52
epistasis group play important roles in DSB repair (
1
,
2
). Investigations into the mechanisms of DSB repair
in vivo
, together with structural and functional characterisation of
RAD52
epistasis group gene products, has shown that these genes direct repair through
homologous recombination, a process in which the damaged DNA molecule retrieves
genetic information by pairing with an undamaged partner. Mammalian homologues
of the yeast
RAD51
and
RAD52
genes have been identified (
3
-
5
), suggesting that the homologous recombination apparatus is highly conserved throughout the eukaryotic kingdom. Consistent with this, mammalian cells can repair DNA DSBs by
homologous recombination, but this takes place much less efficiently than in yeast (
6
).
Although
RAD52
-dependent homologous recombination is the predominant mechanism for DSB
repair in
S.cerevisiae
, recent investigations have revealed the existence of alternative DSB repair pathways in this organism. For example, in
rad52
mutant backgrounds, HO-endonuclease generated DSBs can be repaired by a single strand annealing
pathway that results in homologous recombination between directly repeated
elements that flank the DSB and the concurrent loss of intervening sequences (
7
,
8
). In addition, studies in
rad52
mutant strains have identified mechanisms of DSB repair that require little or no sequence homology between the
recombining DNA segments. Consequently, these mechanisms are referred to as non-homologous or illegitimate DSB repair pathways. Situations where
illegitimate DSB repair has been observed in
S.cerevisiae
include the
RAD52
independent non-homologous integration of plasmids into the yeast genome (
9
), and the recircularisation of linearised plasmid DNA (
10
-
12
). In these situations, repair occurs through either of two pathways. The first
of these is illegitimate end-joining, a process whereby two DNA ends are ligated together without the
loss of nucleotide sequences. The other, which we term illegitimate
recombination, results in the loss of variable amounts (normally one to several
hundred bp) of terminal DNA through the joining of the two DNA ends via short
(usually 1-7 bp) direct repeats. In contrast to the situation in yeast, where
homologous recombination normally predominates, DNA DSBs are usually repaired
by illegitimate recombination pathways in vertebrate systems (
13
). Nevertheless, as in yeast, the two principal pathways for illegitimate DSB repair in vertebrates are direct end joining and illegitimate recombination between
short direct repeat sequences. These observations therefore suggest that mechanisms of illegitimate DSB repair are highly conserved
throughout evolution and raise the possibility that the identification and
characterisation of components of
S.cerevisiae
illegitimate recombination pathways will provide valuable insights into DSB
repair in mammalian systems.
Until a few years ago, very little was known about the components of the
mammalian illegitimate DSB repair pathways. Recently, however, it was shown
that the nuclear protein Ku is defective in the radiosensitive Hamster cell
line
xrs-6
, which is impaired both in DNA DSB repair and in V(D)J recombination (
14
-
16
). This finding established Ku as an important component of the mammalian DSB
repair machinery. Ku is a heterodimer of two tightly-associated subunits of approximately 70 and 80 kDa (Ku70 and Ku80,
respectively) and serves as the DNA binding component of the DNA dependent
protein kinase (DNA-PK). Cells defective in the other DNA-PK component, the DNA-PK catalytic subunit (DNA-PK
cs
), are also deficient in DSB repair and in V(D)J recombination, implicating the DNA-PK holoenzyme in these processes (
17
-
19
). Since Ku binds specifically to DNA DSBs
in vitro
, it is likely the DNA-PK
cs
/Ku complex directly senses DNA damage and is an integral component of the
illegitimate DSB repair apparatus.
Feldmann and Winnacker (
20
) identified a heterodimeric DNA end-binding factor in
S.cerevisiae
and showed that the gene for the 70 kDa subunit of this factor encodes a
polypeptide with homology to Ku70. Given the importance of Ku in mammalian
systems, we and others have tested for an involvement of this putative yeast
Ku70 homologue, which we refer to as Yku70p, in DSB repair (
21
-
23
). Consistent with the dominance of
RAD52
-dependent homologous recombination in yeast, disruption of
YKU70
does not result in marked hypersensitisation to ionizing radiation. However, when the homologous recombination system is rendered inactive by mutations in
RAD52
,
yku70
mutations significantly enhance the radiosensitivity of yeast strains. These
results reveal that YKu70p is involved in the
RAD52
-independent repair of ionising radiation-induced DNA damage and are consistent with the existence of a Ku-dependent illegitimate DSB repair apparatus in
S.cerevisiae
.
Since mammalian Ku70 is complexed with Ku80 and in light of the fact that Yku70p
co-purifies with a polypeptide of ~80 kDa (
20
), we have searched for an
S.cerevisiae
Ku80 homologue. In this manuscript we describe the identification of a yeast gene, termed
YKU80
, that encodes a protein that is related to mammalian Ku80. Consistent with this
gene encoding a
bona fide
Ku80 homologue, disruption of
YKU80
hypersensitises yeast
rad52
mutant strains to ionising radiation and to the radio-mimetic drug methyl methanesulphonate (MMS). Moreover, through using an
in vivo
plasmid repair assay, we demonstrate that Yku80p plays an important role in illegitimate DNA end joining and suppresses error-prone illegitimate recombination. Finally, we demonstrate that
inactivation of
YKU80
leads to telomeric shortening. These results reveal that the Ku-dependent DSB repair pathway is highly conserved from yeast to man and
indicate that Yku80p plays important roles in maintaining chromosomal integrity
both under normal growth conditions and in response to exogenous mutagenic
agents.
Protein sequence alignments were performed using the PILEUP package (Genetics Computer Group, Wisconsin). The resulting pileup was edited in GCG using the LINEUP facility and was further edited in SEP-APP. The edited pileup was then used in the Boxshade program which identifies conserved and semi-conserved amino acids.
Yeast strains are given in Table
1
and yeast media were as described by Sherman
et al.
(
24
). MMS sensitivity was measured as described by Milne and Weaver (
25
). Briefly, yeast colonies were suspended into dH
2
O, and diluted six times by 10-fold serial dilution. Aliquots (15 [mu]l ) of each dilution were then spotted in duplicate onto YPED plates
with and without MMS (0.005%), and were incubated at 30oC for 3-4 days. Temperature sensitivity was determined by spotting 15 [mu]l aliquots of serially diluted cultures in duplicate onto non-selective media which were then incubated at 30oC or 37oC for 3-4 days.
Table 1
yku80-3
GAATGTACCGTCAACCAAAGATTAGCAGTTG
yku80-4
GGCAGCAGCGTGTGTTGAGATTAGGAACCGC
His3-3
GATTGTCTGCGAGGCAAGAATGAT
Ura3-1
GGCGGATAATGCCTTTAGCGGC
Ura3-2
GGAGAATATACTAAGGGTACTG
Trp1
GTGCACTTGCCTGCAGGCC
All yeast-
E.coli
shuttle vectors used have centromeric and ARS sequences for stable maintenance
in yeast, an auxotrophic yeast selectable marker, an OriC for high copy number
propagation in
E.coli
and a [beta]-lactamase gene for ampicillin selection in
E.coli
. p70FL (
21
) consists of
YKU70
and flanking regions cloned into pRS413 (Stratagene). p80FL was generated as
follows: a 2.5 kb region of yeast genomic DNA containing the full length
YKU80
gene and flanking regions was amplified by PCR using oligonucleotide primers yku80-3 and yku80-4 by employing the Bio-X-act proof reading DNA polymerase (Bioline). After first
cloning into pGEM-T (Promega), this fragment was cloned into the yeast-
E.coli
shuttle plasmid pRS414 (Stratagene) to generate p80FL. The p80FL insert was
sequenced fully by automated sequencing (Applied Biosystems) to verify that no
PCR or cloning errors had taken place. pBTM116 (Fig.
3
A) contains a
TRP1
selectable marker. pRS413 and pRS414 (Stratagene) contain
HIS3
and
TRP1
selectable markers, respectively.
The
YKU80
disruption construct was generated by PCR cloning of a 2.1 kb fragment of the
YKU80
gene into pGEM-T, using primers yku80-1 and yku80-2. This fragment was then subcloned into pBluescript-SK+ using the unique restriction sites,
Xho
I and
Bam
HI, and the ORF was disrupted by inserting a
URA3
marker into
Eco
RI/
Sph
I sites, unique within the resulting vector. The disruption fragment was then
excised using
Bam
HI and
Xho
I and was used to transform the appropriate strains. Transformants were selected
on minimal media lacking uracil, and the presence of a disrupted
YKU80
gene was verified by the use of
yku80
and
URA3
primers in PCR studies. Two
RAD52
disruption constructs were provided by D. Weaver and have
TRP1
and
URA3
selection, respectively. Strains disrupted in
RAD52
were checked by PCR using
RAD52
/
TRP1
or
RAD52
/
URA3
primers, as appropriate.
A yeast colony was inoculated into 10 ml YPED and was grown overnight at 30oC. The culture was then diluted in dH
2
O to an OD
600nm
value equivalent to 1 * 10
7
cells/ml and 1 ml aliquots were irradiated using a
137
Cs source at a dose of 0.18 kRad/min. Irradiated samples and an unirradiated
control were than serially diluted in dH
2
O and were plated in duplicate on YPED followed by incubation at 30oC for 3-4 days.
Plasmid DNA (5 [mu]g) was digested with restriction enzyme to completion as determined by gel
electrophoresis followed by ethidium bromide staining and Southern blotting,
and the restriction enzyme was inactivated by treatment at 65oC for 20 min. This linearised DNA was then used to transform yeast by the lithium
acetate method as described by Schiestl and Gietz (
26
)-parallel transformations were performed with an equivalent amount of uncut plasmid to enable normalisation for minor differences in
transformation efficiencies between strains and between experiments. Serial
dilutions were plated and colonies arising on selective media after 3-4 days were counted.
Genomic DNA and plasmid DNA from
S.cerevisiae
was isolated as described previously (
27
). Briefly, cells were grown to saturation, then were harvested, washed, and suspended in lysis buffer (2% Triton X-100, 1% SDS, 100 mM NaCl, 10 mM Tris-HCl, 1 mM EDTA). Next, glass beads and phenol/chloroform/isoamyl alcohol (50:49:1) was added followed by 3 min of high speed vortexing. TE
buffer was added and the organic and aqueous phases were separated by
centrifugation. The aqueous phase was transferred to a fresh tube and DNA was
precipitated with 100% ethanol. After dissolving the DNA in TE buffer, RNA was
removed by treating with RNase A at 37oC. For telomere analysis, 1 [mu]g genomic DNA was digested overnight with 10 U
Xho
I (Boehringer Mannheim). The digested DNA was then run on a 0.8% agarose gel and
was transferred to nitrocellulose as described previously (
27
). The DNA used for hybridisation to telomere sequences was a poly(GT)
20
oligonucleotide that was
32
P end-labelled by polynucleotide kinase (Boehringer Mannheim). Telomere blots
were hybridised in Church buffer (7% SDS, 1% BSA, 0.25 M Na
2
HPO
4
, 1 mM EDTA) overnight at 60oC and were washed in 0.2* SSC, 0.1% SDS at room temperature for 20 min before exposing to X-ray film.
Initial attempts to identify a
YKU80
homologue in the
S.cerevisiae
genome were unsuccessful, despite the use of several database search
facilities. During the course of these investigations, however, we identified a
C.elegans
open reading frame (ORF) capable of encoding a protein of predicted molecular mass of 82 kDa that is ~20% identical in sequence to human Ku80 (Fig.
1
). Although the sequence alignment between the two proteins is weak, it spans
the whole length of the polypeptides, suggesting that the
C.elegans
ORF indeed encodes a Ku80 homologue. Interestingly, we noted that a central
portion of the
C.elegans
sequence, spanning amino acid residues 386 to 482, is ~36% identical to human Ku80, implying that this region corresponds to a
particularly important domain of the protein. Given this, we reasoned that a
putative yeast Ku80 homologue would also display highest similarity in this
region. Indeed, when the central conserved portion of human Ku80 was used in
database searches using the blastp facility, we identified an
S.cerevisiae
open reading frame (ORF) from 26387-28326 in the reverse complement of accession number Z49702 as a significant hit. This sequence resides on Chromosome XIII. Subsequent analyses revealed that this ORF has the potential to encode a polypeptide of 646
amino acid residues that is ~15% identical in sequence to the mammalian and
C.elegans
Ku80 proteins (Fig.
1
). Although the homology is not striking, it spans the entire length of the
yeast sequence and, encouragingly, the regions that are most conserved tend to
correspond to regions that are most similar between the human and
C.elegans
sequences. We therefore tentatively concluded that the yeast ORF encodes a Ku80
homologue and named the corresponding gene
YKU80
.
As an approach to address whether
YKU80
indeed encodes a functional yeast Ku80 homologue, we inactivated this gene in
the haploid yeast strains W3031B and GY1 by one-step gene disruption (
28
). This involved the PCR cloning of a 2.0 kb genomic fragment spanning the
YKU80
locus and the disruption of the ORF at amino acid residue 74 by the
introduction of a
URA3
cassette. Since disruption of
YKU70
results in an inability to grow at 37oC (
20
,
21
,
23
), we initially tested
YKU80
disrupted strains for temperature sensitivity. As shown in Figure
2
A, like
yku70
mutant yeasts, strains disrupted for
YKU80
are unable to form colonies on solid media when incubated at 37oC. Consistent with the temperature-sensitive growth defect being attributable directly to mutations in
yku80
, this phenotype is observed in several independently derived
yku80
mutants. Moreover, the growth defect at 37oC is complemented by plasmid p80FL that contains the full length wild-type
KU80
gene, but is not complemented by the parental plasmid pRS414. Although we have
so far been unable to determine the precise reason for the growth arrest of
yku80
mutant strains at 37oC, FACS analyses reveal that most cells arrest in the G2 phase of the cell
cycle, suggesting an inability to enter into mitosis.
Since mammalian cell lines defective in Ku80 display heightened sensitivity
towards agents that generate DNA strand breaks (
14
,
15
), we tested the sensitivity of
yku80
mutant strains towards ionising radiation and towards the radiomimetic drug
MMS. As measured by the ability of yeast cells to form colonies on solid media
after treatment with ionising radiation or MMS, we were unable to detect
statistically significant differences in rates of cell killing between
yku80
mutant strains and their parents (Fig.
2
B and data not shown). However, in the course of these studies, we noted that
MMS impairs the growth rate of
yku80
mutant strains considerably more than that of parental yeasts and this is
evident both in reduced colony size and survival on solid media, and in
increased cell doubling times in liquid cultures. This suggests that mutations
in
YKU80
impair the ability of
S.cerevisiae
to cope with MMS-induced cellular damage.
Yeasts employ homologous recombination to repair DSBs under most circumstances.
Consistent with this, loss of Yku70p activity significantly hypersensitises
yeast strains to ionising radiation only when the homologous recombination
apparatus is debilitated. To see whether this is also the case for the putative
yeast Ku80 homologue, we assessed the effect of disrupting
YKU80
in
rad52
mutant backgrounds. Notably, as shown in Figure
2
B, mutations in
YKU80
enhance substantially the radiosensitivity of
rad52
mutant strains and this hypersensitisation is observed at all radiation doses tested. Similarly, the MMS sensitivity of
rad52
mutant strains is elevated significantly when
YKU80
is inactivated. By contrast, inactivation of
YKU80
does not result in detectable increases in sensitivity towards agents such as
ultraviolet light, which do not generate DNA DSBs as the principal lethal
lesion (data not shown). Together, these results reveal that
YKU80
is involved in the repair of ionising radiation-induced or MMS- induced DNA strand breaks by a
RAD52
-independent pathway, and are consistent with
YKU80
encoding a functional homologue of mammalian Ku80.
The phenotypes of
yku80
mutant strains described above are very similar to those reported for strains
disrupted in
YKU70
(
20
-
23
). Although this is consistent with Yku70p and Yku80p functioning in the same
pathway, an alternative explanation is that
YKU70
and
YKU80
define two distinct
RAD52
-independent DNA DSB repair systems. To distinguish between these
possibilities, we determined whether mutations in
YKU70
and
YKU80
synergistically affect the sensitivity of
rad52
mutant yeasts towards DNA damaging agents. Importantly,
rad52
mutant strains defective in both
YKU70
and
YKU80
are not more sensitive towards ionising radiation or towards MMS than are
rad52
strains mutant for
YKU70
or
YKU80
alone (Fig.
2
B and data not shown). These data therefore imply that
YKU70
and
YKU80
operate in the same DNA repair pathway, and are consistent with a model in
which Yku70p and Yku80p function in a heterodimeric complex, as is the case for
their mammalian counterparts.
Previous studies have revealed that mammalian cells impaired in Ku function are
deficient in the rejoining of radiation-induced DSBs of chromosomal DNA (
29
,
30
). To determine whether
S.cerevisiae
Ku also functions in DSB repair, we utilised a transformation-based plasmid rejoining assay that measures the ability of yeast cells to
repair restriction enzyme generated DSBs
in vivo
. In this assay, a
S.cerevisiae
strain is transformed with a yeast-
E.coli
shuttle plasmid that has been linearised by treatment with a restriction
enzyme. To normalise for differences in transformation efficiency between
strains and between repeats of the same experiment, a supercoiled version of
the same plasmid is transformed into the yeast strain, in parallel. Since the
plasmid must be recircularised in order to be propagated, the number of
transformants obtained with the linear plasmid normalised to the number
obtained with the supercoiled plasmid provides a quantitation of the ability of
the yeast strain to mediate repair of the restriction enzyme-generated DSB. Figure
3
A is a representation of the plasmid, pBTM116, which is used in these experiments-to prevent the DSBs from becoming repaired by homologous recombination
with the yeast genome, the sites for restriction enzyme cleavage of pBTM116 are
within regions that are not homologous to chromosomal sequences.
Initially, we analysed the repair of linearised plasmid molecules bearing
cohesive 5' overhanging termini that were generated by the restriction endonuclease,
Eco
RI. As shown in Figure
3
B, this type of DSB is repaired with high efficiency in wild-type yeast strains in that transformant yields with
Eco
RI linearised pBTM116 are over 70% of the values obtained with supercoiled
plasmid. Consistent with
RAD52
-dependent homologous recombination mechanisms playing only a minor role in DSB repair under these
experimental circumstances, transformant yields with
Eco
RI linearised pBTM116 are decreased <2-fold in
rad52
mutant strains (Fig.
3
B). In marked contrast, strains mutated in
YKU80
show a dramatic 40-100-fold decrease in transformant recoveries with linearised DNA, and
these effects are observed in both the presence or absence of functional
RAD52
(Fig.
3
B). Demonstrating that this phenotype is not specific to
Eco
RI-generated DNA ends, very similar results are obtained with 5' overhanging ends created by
Xho
I
cleavage and with 3' overhanging ends generated by the enzyme
Pst
I (data not shown). Furthermore, similar dramatic effects are observed with several independently derived
yku80
mutant strains and with various other plasmids. As a further verification that the
deficiency in plasmid repair in
yku80
mutant strains is a direct consequence of mutations in
YKU80
, this defect is complemented fully by p80FL, which directs the expression of
full-length wild-type Yku80p, but is not complemented by the parental plasmid pRS414 (Fig.
3
C). Since transformation efficiencies with supercoiled plasmid DNA are unaffected by mutations in
YKU80
(see Fig.
3
legend), these data imply that
YKU80
is required for the efficient repair of cohesive DNA termini by a
RAD52
-independent mechanism. Mutations in
YKU70
and
YKU80
have very similar effects in the plasmid repair assay (Fig.
3
A and B). Moreover, strains mutant in both
YKU70
and
YKU80
are no more impaired in plasmid repair than strains mutated for either
YKU70
or
YKU80
alone (Fig.
3
D). These data therefore support a model in which
YKU70
and
YKU80
function in the same DNA DSB repair pathway.
To analyse the type(s) of DNA repair that requires Yku80p, repaired plasmids were recovered from
rad52
mutant strains that either possessed or lacked functional
YKU80
. These plasmids were shuttled into
E.coli
and were then analysed by restriction enzyme digestion and by DNA sequencing. Strikingly, of over 100 plasmids recovered from
YKU80
strains, all had been repaired by the simple religation of the cohesive DNA
termini. In stark contrast, every plasmid that was recovered from
yku80
mutant strains had suffered deletion of terminal sequences and had been joined
at sites corresponding to short direct repeat elements of 2-16 bp (Fig.
4
). In some cases, plasmid repair products retrieved from
yku80
mutant strains also possessed small insertions at the sites of joining.
Together, these results reveal that Yku80p plays a crucial role in an
illegitimate DSB repair pathway that is capable of rejoining cohesive DNA ends
with very high efficiency and fidelity. Furthermore, they show that, in the
absence of functional
YKU80
, cohesive DNA ends cannot be repaired by simple end-joining. Instead, they are repaired by a relatively inefficient error-prone illegitimate recombination pathway that results in the
deletion of terminal DNA sequences and the joining of the two DNA ends via
short direct repeat motifs.
Recently, Porter
et al.
(
31
) have reported that inactivation of
YKU70
leads to telomeric shortening. We therefore wished to determine whether
telomeric length is altered as a consequence of
YKU80
disruption. To address this question, genomic DNA was isolated from newly
sporulated
yku80
mutant strains, then this DNA was digested with the restriction enzyme
Xho
I and was subjected to Southern blot-hybridisation analysis using the radiolabelled oligonucleotide poly(dG-dT)
20
that hybridises to the telomeric repeat elements (consensus C
1-3
A; Fig.
6
A).
Xho
I cleaves within the sub-telomeric Y' region that is found in many
S.cerevisiae
telomeres (`Y'-type telomeres') and, in wild-type strains, generates a predominant terminal chromosomal fragment of ~1.3 kb that hybridises to poly(G-T)
20
(
32
; Fig.
6
A and B). In addition, several larger poly(G-T)
20
hybridising fragments are observed, which correspond to telomeric ends from the
subset of telomeres (`X-type telomeres') that lack Y' regions (Fig.
6
B). Significantly, yeasts disrupted for
YKU80
or
YKU70
contain dramatically shortened telomeres (Fig.
6
B). Indeed, we estimate that the predominant telomeric product in Figure
6
B becomes reduced in size by ~300 bp, corresponding to a decrease of ~65-70% in the length of the region bearing the terminal repetitive
elements. Consistent with this, the hybridisation intensity of poly(GT)
20
to the predominant telomeric fragment is reduced ~5-fold in
yku80
mutant strains. In addition, a decrease is observed in both the hybridisation
intensity and the length of several of the larger DNA fragments that hybridise
to poly(GT)
20
, indicating that shortening also occurs at X-type telomeres. Similar reductions in telomeric length are observed in
numerous independently-derived
yku80
mutant strains (Fig.
6
B). Furthermore, the telomeric shortening phenotype is complemented fully when
yku80
mutant yeasts are transformed with the plasmid p80FL, which contains the full-length, wild-type
YKU80
gene (Fig.
6
C). Together, these data indicate that inactivation of Ku function in yeast
leads to a dramatic decrease in telomeric length.
Over recent years, several
S.cerevisiae
genes have been identified that regulate telomere length. In many cases, the inactivation of such genes
leads to progressive telomere shortening, and the full expression of the mutant phenotype generally requires sub-culturing for over 100 generations (
33
,
34
,
35
). To see whether this is the case for strains mutant in
YKU80
, we measured telomeric lengths of newly-sporulated wild-type and
yku80
mutant strains after they had been subcultured for 25, 50, 75 or 100
generations (Fig.
6
D). Significantly, these studies revealed that, upon inactivation of
YKU80
, the reduction in size and hybridisation intensity of the predominant Y' band telomeric ends is complete after 25 cell division cycles. Similarly, the reduction in length and hybridisation intensity of X-type telomeric ends is also manifest fully after just 25
generations. These results indicate that, when
YKU80
is inactivated, telomeric attrition occurs relatively rapidly until telomeres have been reduced to a certain size and that, after such a size has been
reached, a new equilibrium is established such that telomere ends do not
shorten further. Having analysed the kinetics of telomere shortening after
YKU80
inactivation, we next assessed the rate of telomere lengthening when
yku80
mutant strains are complemented by the wild-type
YKU80
gene. Notably, full restoration of telomeric length to wild-type levels is attained after just 25 generations and, after this, no further telomeric lengthening is evident (Fig.
6
D). These data reveal that changes in yeast Ku activity lead to relatively rapid alterations in telomeric length.
In this paper, we have described the identification of sequences in the genomes
of
C.elegans
and
S.cerevisiae
which encode for polypeptides that are related to mammalian Ku80. Furthermore, we have shown that the yeast gene is involved in illegitimate recombination and, therefore, appears to correspond to a
bona fide
Ku80 homologue. Our data are in line with those published very recently by
Milne
et al.
(
36
). Together with recent work showing that the Ku70 homologues of
S.cerevisiae
and
Drosophila melanogaster
(
37
) are involved in DNA DSB repair, these data imply that the Ku-dependent illegitimate DSB repair apparatus is conserved throughout the
eukaryotic kingdom. Consistent with homologous recombination playing a dominant
role in DSB repair in yeasts, disruption
YKU80
only leads to significantly elevated radiosensitivity in genetic backgrounds where homologous recombination
is rendered inoperative. Nevertheless, through using a plasmid repair assay, we show that
YKU80
plays a crucial role in illegitimate recombination reactions that result in the
accurate rejoining of cohesive DNA termini. Furthermore, we are able to
complement the plasmid repair defects of
yku80
mutant strains with a plasmid bearing wild-type
YKU80
. The establishment of a simple yet highly sensitive and reproducible complementation assay for Yku80p should
greatly facilitate mutational studies directed at defining functionally important domains of this protein. In this regard, it will be of particular interest to examine the effects of mutating the regions of Ku80 which we have
identified as being particularly highly conserved from yeast to man.
We thank J. Murray for yeast strain GY1, D. Weaver for
RAD52
-disrupted yeast strains, A. Bannister for pEG202fos, H. Feldmann and E. L. Winnacker for Yku70p-deficient yeast strains, and S. Fields for pBTM116. Thanks also to members of the SPJ
laboratory for their advice and critical comments on this work and especially
to Graeme Smith and David Gell for help with the protein sequence alignment.
SJB is supported by a studentship from the Cancer Research Campaign. The
research described in this manuscript was made possible by a grant from the Kay Kendall Leukaemia Fund and by grants SP2143/0101, SP2143/0201 and SP2143/0401 from the
Cancer Research Campaign.
*To whom correspondence should be addressed. Tel: +44 1223 334 102; Fax: +44
1223 334 089; Email: spj13@mole.bio.cam.ac.uk
Strain
Genotype
Source
W303-1A
Mat
[alpha]
ade2 his3 leu2 trp1 ura3 can1-100
ref. 20
W303-1B
Mat
a
ade2 his3 leu2 trp1 ura3 can1-100
ref. 20
GY1
Mat
[alpha]
ade2 leu2 ura3 trp1 CAN1
J. Murray
yku70[alpha]
W303-1A yku70::URA3
ref. 20
yku70a
W303-1B yku70::LEU2
ref. 20
yku80[alpha]
W303-1A yku80::URA3
This study
yku80a
W303-1B yku80::URA3
This study
GW
W303-1B
*
GY1
This study
GW80
yku80
[alpha]
*
yku80a
This study
DWY85
Mat
[alpha]
ho::LYS2 leu2::hisG rad52::URA3
D. Weaver
DWY86
Mat
a
l ho::LYS2 leu2::hisG rad52::URA3
D. Weaver
DWY91
Mat
[alpha]
arg4 .RV leu2-3-112 cyhR trp1-289
D. Weaver
rad52::URA3
DWY176
Mat
a
his3 cyhS ade2 ura3-52 trp1-289
D. Weaver
rad52::URA3
rh7a
Mat
a
arg4.RV leu2.3-112 trp1 ade2 his3
This study
yku70::LEU2 rad52::URA3
rh8b
Mat
a
ade2 his3 trp1 ura3 can1-100
This study
yku70::LEU2 rad52::URA3
rh10b
Mat
[alpha]
ade2 his3 trp1 ura3 can1-100
This study
yku70::LEU2 rad52::URA3
rh12a
Mat
a
ade2 trp1 his3 yku70::LEU2
This study
rad52::URA3
rh16b
Mat
[alpha]
ade2 trp1 his3 yku70::LEU2
This study
rad52::URA3
r80-1[alpha]
Mat
[alpha]
ade2 trp1 his3 yku80::URA3
This study
rad52::TRP1
r80-1a
Mat
a
ade2 trp1 his3 yku80::URA3
This study
rad52::TRP1
ku-1
Mat
a
ade2 trp1 his3 yku80::URA3
This study
yku70::LEU2
rku-1
Mat
[alpha]
ade2 his3 yku70::LEU2
This study
yku80::URA3 rad52::TRP1
yku70-3
GAGATTTCTATGCTCGAGGAGAACTTC
yku70-5
GGGACCCACAAAGGATTCTCAGGAAGTGG
rad52-1
TATTGGGAATAAATGCCAATGCCAGTTC
rad52-2
TAATGATCTATTGTTTTTCCGAGTTGCC
yku80-1
GGAAATGCTCGAGTATGAGACCTTGAACCAG
yku80-2
CAGCGGATCCCCGGATGTAGTTGTTCG
REFERENCES
Return