ABSTRACT
Gel retardation assays using a probe containing the repeat region of a
Schizosaccharomyces pombe
chromosomal telomere identified four specific DNA-protein complexes in
S.pombe
total protein extracts (I, I
'
, IIa and IIb). The proteins responsible for these complexes bound to the
telomeric repeat region irrespective of whether or not the repeats were in
close proximity to the end of a DNA molecule, and none of them bound strongly to single-stranded DNA. The protein responsible for complex I (TeRF I) was separated from the activity responsible for complexes IIa and IIb (TeRF II) using
heparin-Sepharose chromatography. Both factors were efficiently cross-competed by an oligonucleotide containing the 18 bp sequence 5
'
-GGTTACAGGTTACAGGTT-3', which corresponds to two complete telomeric repeat units. Mutation of the T
residues at positions 4 and 11 in the oligonucleotide dramatically reduced
binding by TeRF II, but had no affect on binding by TeRF I. The protein
responsible for complex I' did not bind strongly to either the wild-type or mutant oligonucleotide.
Telomeres are the physical ends of eukaryotic chromosomes and they have a number
of important roles within the cell. These roles include preventing chromosome
fusion, protecting the chromosomes from exonucleolytic attack and facilitating
the complete replication of chromosomes (
1
,
2
). In most eukaryotes, with the notable exception of
Drosophila
, telomeres are made up of multiple copies of a short (5-8 bp) repeat unit (
3
,
4
). This repeat is usually rich in T and G residues in the DNA strand which runs 5' -> 3' towards the chromosome end. In ciliated protozoa this strand overlaps its partner strand at the very end of the
chromosome to form a short 3' extension (
5
,
6
). In the budding yeast,
Saccharomyces cerevisiae
, telomeres gain long single-stranded extensions during S phase of the cell cycle (
7
). The functions of telomeres are thought to be achieved via interactions
between specific proteins and the telomeric repeat sequences, and by the
formation of specific DNA structures at the chromosome ends. Both the single-stranded 3' extension and the double-stranded repeat region are the targets for specific DNA
binding proteins in a range of different organisms (
8
). In the ciliated protozoan
Oxytricha nova
, a dimeric protein containing [alpha]- and [beta]-subunits interacts with the single-stranded 3' extension (
9
-
11
). In
Euplotes
, a monomeric protein with homology to the [alpha]-subunit of the
Oxytricha
protein, appears to have a similar role (
12
). The budding yeast,
S.cerevisiae
, has been used extensively as a model system to investigate telomere function.
In yeast, telomeres consist of an average of 300 bp of a variable TG rich
sequence, abbreviated as TG
1-3
(
3
). Several different telomere binding proteins have been described. The best
characterised of these is the multi-functional protein Rap1p (
13
). This protein binds to both the double-stranded repeat region and the single-stranded extension of yeast telomeres, and these interactions appear
to be important for telomere function (
8
,
14
-
19
). Reduction in the amount of functional Rap1p in the yeast cell, by growing a
rap1
ts
strain at a semi-permissive temperature, resulted in a gradual reduction in telomere length. This could be reversed by returning the
cells to a permissive temperature (
19
). Similarly, overexpression of Rap1p caused an increase in telomere length and an increase in the rate of
chromosome loss and mitotic recombination (
8
). Immunofluorescence experiments have shown that the majority of Rap1p in yeast
cells is associated with the telomeres and that this association requires the
products of the
SIR3
and
SIR4
genes (
20
,
21
).
In vitro
experiments suggest that Rap1p may bind once per 18 bp of telomeric DNA,
resulting in up to 25-30 binding sites per telomere (
22
). The presence of Rap1p at telomeres is particularly intriguing because this
protein is also a transcription factor, involved both in activating and
silencing transcription at a range of loci (
23
,
24
,
13
). It interacts with the UAS of many yeast housekeeping genes, including genes
encoding ribosomal proteins, components of the translational machinery and
glycolytic enzymes (
25
-
32
). Rap1p also interacts with the silencers at
HML
and
HMR
and plays a role in repressing the inactive mating type genes (
30
,
33
-
36
). Recently a
RAP1
gene was cloned from the closely related yeast
Kluyveromyces lactis
(
37
). The protein product of this gene binds a similar DNA sequence to budding
yeast Rap1p and contains a conserved region which in the budding yeast gene
encodes a domain of Rap1p involved in telomere function (
37
). It is not yet known if the
K.lactis
protein interacts with
K.lactis
telomeres. A second budding yeast gene,
TBF1
, encodes an essential TTAGGG repeat binding factor, with a molecular weight of
63 kDa (
38
). This protein binds to two TTAGGG sequences proximal to the TG
1-3
repeat sequences, but its role in telomere function is unclear. Recently, three
further genes were isolated from a yeast gene library based on their ability to
make protein products which specifically interact with TG
1-3
DNA
in vitro
(
39
). However, the role, if any, that these proteins play at telomeres
in vivo
is unclear.
Telomere binding factors have also been identified in higher eukaryotes.
Xenopus laevis
contains a well characterised activity called
Xenopus
telomere end factor (XTEF) which interacts with two repeats of the sequence
TTAGGG in the single-stranded telomeric extension (
40
). Single strand TTAGGG binding factors have also been purified from mouse liver
extracts (
41
). Activities have been identified in human cells which interact with both the
double-stranded and single-stranded forms of the TTAGGG human telomeric repeat (
42
,
43
). The single-stranded binding factors also interact with RNA and are components of
hnRNPs (
42
). The human double-stranded telomeric repeat binding factor (TRF) requires six copies of
TTAGGG to form an efficient substrate and was initially estimated to be ~50 kDa in size (
43
). The gene encoding TRF has been isolated from a human cDNA library (
44
). It encodes a protein of 439 amino acids, with a predicted molecular weight of
50.3 kDa. The protein has a region at the N-terminus rich in aspartic and glutamic acid residues and a region close to
the C-terminus containing an myb-like DNA binding motif (
44
).
Fission yeast,
Schizosaccharomyces pombe
, is distantly related to both budding yeasts and higher eukaryotes. A
comparison between telomere organisation and function in fission yeast and
other eukaryotic organisms may therefore provide important insights into the
evolution and roles of these key structures. Fission yeast has a genome of
about the same size as budding yeast, but a haploid chromosome number of only
three. The chromosomal telomeres have been implicated in the specific movements
chromosomes make as haploid nuclei fuse during karyogamy, and the movements the
fused nucleus makes within the cell prior to meiosis (
45
). These movements appear to be mediated via attachment of the telomeres to the
spindle pole body (
45
). Four
S.pombe
chromosomal telomeres have been cloned and sequenced (
46
). They are ~300 bp in length and are made up of a repeat unit of consensus C
1-8
G
0-1
T
0-2
GTA
1-3
(
46
,
47
). Although this consensus accurately describes the sequences at fission yeast
telomeres, it suggests that the repeat unit is very variable, and it relates to
the sequence of the A and C rich DNA strand, rather than the T and G rich
strand. To facilitate comparison with the repeats in other organisms, we have
used the simpler consensus 5'-TTACAG
1-8
-3', which describes the majority of telomeric repeats in
S.pombe.
Although cloned
S.pombe
telomeres have been available for a number of years, no direct telomere binding
proteins have been characterised. The only protein so far shown to be localised
to
S.pombe
telomeres is the chromodomain protein Swi6p. This protein associates with both
centromeres and telomeres in
S.pombe
and is required for proper centromere function (
48
). In this paper we describe the identification and initial characterisation of
fission yeast protein factors which interact with the double-stranded repeat region of the telomeric DNA.
The haploid
S.pombe
wild-type strains
975h
+
and
975h
-
were used throughout. They were routinely grown using YEPD medium on plates and
in liquid culture (
49
).
The starting plasmid for the telomere sub-clones was pNSU28 which contains 0.9 kb of
S.pombe
telomeric DNA in pUC19 (
46
). The 0.9 kb telomeric fragment was isolated from pNSU28 by digestion with
Eco
RI. The isolated
Eco
RI fragment was then cut with
Rsa
I and a 425 bp fragment containing the telomere repeat region plus telomere associated sequence was isolated. This fragment was ligated
into the
Sma
I site of plasmid pSP56 to generate plasmid pAJ25 (
50
). pAJ25 was cut with
Eco
RI and
Bam
HI to release the cloned 425 bp fragment. This was then digested with
Hae
III to release a fragment containing 140 bp of telomeric repeat sequence plus
telomere associated sequence. This was end-filled using Klenow polymerase and dNTPs to convert the
Bam
HI end to a blunt end. It was then cloned into the
Hin
cII site of pSP56 such that the extreme telomeric sequences were at the 5' end of the polylinker and the telomere associated sequences were at the
3' end. This generated plasmid pAJ34. The telomeric sub-fragment used in most retardation assays was isolated by digestion
of pAJ34 with
Bam
HI and
Pst
I.
Schizosaccharomyces pombe
cultures were grown to mid-log phase in 50 ml YEPD medium. Cells were harvested and washed twice with 1 ml 25 mM NaPO
4
pH 7.5, then resuspended in 300 [mu]l ice-cold 25 mM NaPO
4
pH 7.5, containing 1 mM PMSF. The cells were broken with glass beads by
vortexing for 2 min then centrifuged briefly in a microfuge. The supernatant
was collected and recentrifuged for 10 min at 4oC. The supernatant was again collected and the protein concentration
determined using the Bradford assay (
51
). Typically 50 ml of cells yielded ~1 mg protein, of which 5 [mu]g was used in each gel retardation assay. Retardation assays were
performed at room temperature in a total reaction volume of 20 [mu]l. The labelled DNA fragment for use in the assay was generated by end-labelling ~120 ng of isolated DNA fragment using [[gamma]-
32
P]ATP (>185 TBq/mmol; Amersham International plc) and T4 polynucleotide kinase
(Life Technologies, Inc.). Protein extract (5 [mu]g) was incubated with 2 [mu]g poly(dI:dC) and 2 ng of the labelled DNA fragment in a binding buffer
containing 5% glycerol, 1 mM EDTA, 10 mM [beta]-mercaptoethanol, 25 mM Tris-HCl pH 7.5, 25 mM NaCl and 20 mM KCl. After incubation for 30
min at room temperature, DNA-protein complexes were separated by electrophoresis at 180V for ~90 min using 16 cm-long 5% polyacrylamide gels containing 0.5* TBE.
Unlabelled competitor DNAs consisted of either gel isolated DNA fragments or
annealed oligonucleotide pairs. The standard telomere competitor fragment was
obtained by digestion of pAJ34 with
Bam
HI and
Pst
I and isolation of the 140 bp telomere fragment. A competitor fragment with
telomeric repeats away from the ends was obtained by digestion of pAJ34 with
Sph
I and
Pvu
II and isolation of a 810 bp fragment. A control fragment for these experiments
was isolated by digestion of plasmid pSP65 with the same enzymes and isolation
of a 670 bp fragment.
The telomeric repeat oligonucleotides consisted of the sequences:
AC1 5'-GATCTCAGCT
GGTTACAGGTTACAGGTT
G-3'
AC2 5'-GATCC
AACCTGTAACCTGTAACC
AGCTGA-3'
AC3 5'-
GGTTACAGGGGGG
TT-3'
AC4 5'-AA
CCCCCCTGTAACC
-3'
AC5 5'-
TTACAGGTTACAGG
-3'
AC6 5'-
CCTGTAACCTGTAA
-3'
AC7 5'-GATCTCAGCT
GGTGACAGGTGACAGGTT
G-3'
AC8 5'-GATCC
AACCTGTCACCTGTCACC
AGCTGA-3'
In each case the telomeric sequences are shown in bold type.
These were annealed in pairs to generate double-stranded oligo- nucleotides for use as competitor DNAs.
A control oligonucleotide used in some experiments consisted of the sequence:
AS1 5'-GATCCTAAATATAAAAA-3'
Heparin-Sepharose (Pharmacia) was treated according to the manufacturer's
instructions. A column was prepared at 4oC with a 2.5 ml bed volume of heparin-Sepharose in Z buffer (10 mM Tris pH 7.5, 1 mM EDTA, 10% glycerol,
100 mM KCl, 50 mM [beta]-mercaptoethanol, 1 mM PMSF). Total protein extract (10-12 ml; 50 mg protein) was applied to the column using a BioRad
EconoSystem. Proteins were eluted using increasing concentrations of KCl ranging from 100 mM to 1 M in Z buffer. Fractions of 0.5 ml were collected automatically and dialysed overnight at 4oC against a large volume of Z buffer. Five [mu]l of each dialysed fraction was used in gel retardation assays.
pAJ34 was digested with
Bgl
II and
Sma
I and the telomere DNA fragment was isolated. This was radioactively labelled at
the
Bgl
II end by end filling using Klenow polymerase in the presence of [[alpha]-
32
P]dCTP (Amersham). The end-labelled fragment was incubated in standard binding buffer with 10 [mu]l of a fraction eluted from heparin-Sepharose at 400 mM and 3 [mu]g poly(dI:dC) in a total volume of 50 [mu]l. Binding was allowed to proceed for 30 min on ice
before addition of diluted DNase I at room temperature. One [mu]l of a 1:100 dilution of Promega DNase I was allowed to digest the DNA for 30 s and the reaction terminated by the addition of 100 [mu]l stop buffer (50 mM Tris pH 8, 2% SDS, 10 mM EDTA pH 8, 0.4 mg/ml proteinase K, 100 [mu]g/ml glycogen). The reaction was then incubated at 37oC for 30 min and 70oC for a further 2 min. It was then extracted once with
phenol-chloroform and dried in a vacuum concentrator. The dried samples were
resuspended in sequencing gel loading buffer and subjected to electrophoresis
on a 8% denaturing polyacrylamide gel. Approximately 1200 c.p.m. was loaded per
lane. The marker was the product of the Maxam and Gilbert A+G reaction on the
same DNA.
In order to identify factors which interact with the double-stranded regions of fission yeast telomeres, we isolated a 140 bp DNA
fragment from a cloned
S.pombe
chromosomal telomere (
46
). This fragment consisted of 102 bp of telomeric repeat sequence and 38 bp of telomere associated sequence. The 102 bp of repeat sequence contained a total of 13 repeats, the majority of which
conformed to the consensus 5'-TTACAG
1-8
-3' (Fig.
1
A). The isolated telomeric DNA was radioactively labelled and tested in gel
retardation assays with a fission yeast total protein extract (Fig.
1
B). Three DNA-protein complexes were detected; a strong complex (I) and two fainter,
lower mobility complexes (IIa and IIb) (lanes 2 and 5). The complexes were
cross-competed by the addition of an excess of the unlabelled telomere fragment
(lanes 6 and 7) but not by equivalent amounts of an unlabelled, non-specific DNA fragment, of approximately the same size (lanes 3 and 4).
This suggested that the complexes were the result of specific DNA-protein interactions. The telomeric DNA fragment used in these initial
retardation assays contained telomeric repeat sequence, a short region of
telomere associated sequence and a small amount of plasmid polylinker sequence.
Shorter fragments which contained only the telomeric repeat sequence generated
the same pattern of complexes as the original probe fragment (data not shown),
indicating that complex formation did not require the telomere associated
sequence or the polylinker.
The pattern of complexes detected in the gel retardation assays could be
explained in several ways: first, multiple binding by a single protein could
account for the multiple complexes detected; secondly, different forms of a
single protein might generate the complexes; thirdly, the lower mobility
complexes may be ternary complexes resulting from protein-protein interactions between proteins present in the extract and a single
DNA binding protein; finally, the different complexes might result from DNA
binding by two or more proteins with overlapping specificities. Experiments in
which increasing amounts of total protein extract were added to retardation
assays containing the radioactively labelled telomere fragment demonstrated
that as the ratio of protein:DNA increased, the amounts of all three complexes
increased in parallel. Complexes IIa and IIb were not preferentially formed at high protein
concentrations (data not shown). This suggested that multiple binding by the
complex I protein was not responsible for the formation of complexes IIa and
IIb. To distinguish between the other possibilities the total protein extract
was fractionated by passing it through heparin-Sepharose and eluting bound
proteins using a gradient of KCl. Fractions were tested in gel retardation
assays using the telomere fragment as a probe. The proteins responsible for
complexes I, IIa and IIb eluted between 350 and 500 mM KCl (Fig.
2
). Complexes IIa and IIb were detected in fractions ranging from ~350 to 400 mM, although in these fractions a small amount of a faster
migrating complex was also detected (lanes 2-5). This faster migrating complex was probably the result of the presence
of small amounts of the complex I protein. These fractions generated more of
complex IIb than complex IIa, suggesting that the proteins responsible for
these two complexes did not copurify during the fractionation procedure. Later
fractions, eluting between ~430 and 500 mM, gave rise to a strong complex I, but produced only
extremely faint complexes IIa and IIb (lanes 6-12). These fractions also generated a new high mobility complex (NC),
which was probably formed by a degradation product of the protein responsible
for complex I. These results confirmed that complexes IIa and IIb were not
formed as a result of multiple binding by the complex I protein because some of
the fractions (lanes 8 and 9) contained relatively large amounts of the complex
I protein but did not produce strong complexes IIa and IIb. It is also unlikely
that the three complexes were generated by differently modified forms of a
single protein, unless that modification led to significantly different
properties on heparin-Sepharose fractionation. The most likely explanation for
these results is that complex I resulted from binding by one factor [telomere
repeat factor I (TeRF I)] and that complexes IIa and IIb resulted from binding
by different forms of a second factor (telomere repeat factor II (TeRF II)].
The protein responsible for Complex I' was not clearly present in any of the fractions collected, although the
fractions in lanes 2-5 do contain a complex below complex I, which might have been complex I'. Complex I' probably resulted from binding by a third protein which
also recognised sequences present in the telomere fragment (see later).
In order to localise the positions within the telomeric sequence at which DNA-protein interactions occurred, DNase I footprinting was performed. A
partially purified protein fraction which generated complexes I, IIa and IIb in
gel retardation assays (see Fig.
2
, lane 5) and the telomeric DNA fragment labelled at the 3' end of the AC rich strand were used (Fig.
3
). The positions on the gel of the individual repeats can be determined from the
positions of the AA*G repeated pattern in lane 2 (Maxam and Gilbert A+G
reaction), corresponding to the conserved TTAC of each repeat on the opposite
strand. We reasoned that if a regular defined footprint was obtained it would
be evidence to suggest that only one protein interacted directly with the
telomeric DNA. The footprint produced contained various regions of both
protection and hypersensitivity (compare lanes 1 and 3 control ladders with
lane 4 containing the protein fraction). The most clearly protected bands are
indicated by arrows on the figure. These show no regular pattern in relation to
the telomeric repeats. Five clear hypersensitivities are also indicated. None
of these are within the TTAC region of the repeat, they all occur within the 3' end of the repeat unit. The pattern of footprint obtained might have
been the result of a single DNA binding protein interacting with different
repeats in different ways, but is perhaps more likely to have resulted from the
presence of different proteins with overlapping DNA binding specificities.
Because the gel retardation assays and footprinting experiments suggested
multiple interactions between proteins in the extract and the telomeric sequences, we synthesised a series of double-stranded oligonucleotides and used these to examine the DNA sequence requirements for production of the different complexes. Initially we
tested two oligonucleotides designated AC1/2 and AC3/4. AC1/2 contained the
sequence 5'-GG TTACAGGTTACAGG TT-3'. This corresponds to two complete repeats with two
base pairs of flanking sequence at each end. AC3/4 contained the sequence 5'-GG TTACAGGGGGG TT-3'. This is a particular variant of a single repeat in
which six G/C pairs are present, again with two base pairs of flanking sequence
at each end. These oligonucleotides were added as unlabelled competitor DNAs to gel retardation
reactions containing the 140 bp telomere fragment and an
S.pombe
total protein extract (Fig.
4
). The double repeat oligonucleotide (AC1/2, lanes 2 and 3) efficiently cross-competed complexes I, IIa and IIb, suggesting that this oligonucleotide
contained the recognition sequences for both TeRF I and TeRF II. Interestingly
it did not cross-compete formation of complex I', suggesting that this complex must be formed by a third factor
with different sequence requirements to TeRF I and TeRF II.
Telomeres in several different organisms have been shown to end in a single-stranded 3' extension, usually only a few repeats in length (
5
,
6
). In budding yeast, telomeres gain long single-stranded extensions during the S phase of the cell cycle (
7
). Single-stranded extensions are the targets for several telomere binding proteins,
including the budding yeast protein Rap1p, which promotes the formation of G
tetrad structures (
15
). Although it is not known if such extensions are ever found at fission yeast
telomeres, we have tested whether any of the telomere binding factors we have
identified will bind to single-stranded DNA, using a range of single-stranded oligonucleotides as competitors. Because binding by budding
yeast Rap1p to single-stranded DNA is less efficient than to double-stranded DNA, we used the single-stranded oligonucleotides at higher concentrations than the
double-stranded oligonucleotides (Fig.
5
). Three single-stranded oligonucleotides were tested; AC1 corresponds to the TG rich
strand of the AC1/2 double repeat oligonucleotide, AC2 corresponds to the AC
rich strand, and AS1 is a control oligonucleotide. None of the oligonucleotides
cross competed the formation of any of the complexes (Fig.
5
, lanes 3 and 4, 6 and 7, 9 and 10), even when used at high concentrations. This
indicated that neither of the individual strands of the double repeat is a
strong substrate for either TeRF I or TeRF II.
Schizosaccharomyces pombe
telomeres, like those in other eukaryotes, are made up of many copies of a
short repeat unit (
46
,
47
). We have now shown that proteins are present in
S.pombe
which can interact specifically with these telomeric repeats. Four specific DNA-protein complexes were detected by gel retardation assays, using a telomere fragment and a total protein extract. These
complexes were produced as a result of binding by at least three different factors (see below). All four complexes were produced irrespective of
the proximity of the repeats to the end of the DNA fragment, indicating that
the factors we have identified are not end-specific. Fractionation of the total protein extract using heparin-Sepharose separated the factor responsible for complex I away from
the factor responsible for complexes IIa and IIb. The partially purified
factors were termed TeRF I and TeRF II respectively. Complexes IIa and IIb were
probably produced by differently modified forms of TeRF II because IIa and IIb
were generated by the same column fractions, and their DNA binding
specificities appeared identical in cross-competition experiments. The existence of differently modified forms could
be a property of certain telomere binding proteins because the human telomere
repeat binding factor TRF also produced multiple complexes in gel retardation
assays when isolated from HeLa cells (
44
). The factor responsible for complex I' was not unambiguously identified in any of the fractions tested.
We have probed the DNA recognition specificities of the different factors using
a series of unlabelled oligonucleotides as competitor DNAs in gel retardation
assays. These experiments demonstrated that the factors responsible for the
different complexes had similar, but distinct, recognition sequences. An
oligonucleotide (AC3/4) containing one complete telomeric repeat failed to
compete any of the complexes, suggesting that a single repeat is not a strong
binding site for TeRF I, TeRF II or the complex I' factor. An oligonucleotide (AC1/2) containing two complete telomere
repeat units cross-competed formation of complexes I, IIa and IIb, but not complex I'. AC1/2 therefore contained binding sites for both TeRF I and TeRF
II but not for the protein which formed complex I'. The presence of binding sites for TeRF I and TeRF II in an
oligonucleotide containing just two complete repeat units, suggests that these
factors may not be analogous to the human telomere binding factor TRF, which
requires six repeat units for strong binding (
43
). The oligonucleotide AC1/2 contained additional telomeric sequences, flanking
the two complete repeats. These sequences comprised two G/C bp at the 5' end and two T/A bp at the 3' end. When these extra bases were absent (AC5/6) the ability of
the oligonucleotide to compete complex formation was lost. When an
oligonucleotide identical to AC1/2, but containing mutations of the second T in
each repeat (AC7/8), was used as a competitor DNA, complex I was still
efficiently competed, but competition of complexes IIa and IIb was much
reduced. The mutations in this oligonucleotide have therefore dramatically
reduced the strength of binding of TeRF II, but had little effect on TeRF I
binding. The T/A bp which was mutated in each repeat within this
oligonucleotide is highly conserved and is present in all telomeric repeat
units in
S.pombe.
The lack of effect of these mutations on TeRF I binding may indicate that the
blocks of G/C bp within the oligonucleotide, and the spacing between these
blocks, are the critical factors for strong binding by this protein. The
difference in DNA recognition sequence requirements for formation of the
complexes confirmed that complexes IIa and IIb were produced by a different
factor to complex I.
Individual single-stranded oligonucleotides from AC1/2 failed to cross-compete formation of any of the complexes, even when present at
relatively high levels. This suggested that none of the factors interacts
efficiently with single-stranded DNA. Our experiments were designed to identify strong single-stranded binding activities. It remains possible that one or more of our factors
possesses a weak single-stranded binding activity that requires the presence of multiple copies of
a recognition sequence for complex formation
in vitro
, analogous to the interaction between budding yeast Rap1p and single-stranded DNA (
15
).
Budding yeast Rap1p is currently the best characterised telomere binding factor
in any yeast species. We have now identified two new telomere binding factors
in fission yeast. Is either of these factors a good candidate to be a homologue
of Rap1p? Rap1p is an abundant protein which interacts with a consensus
sequence generated in telomeric DNA by juxtaposition of particular variants of
the TG
(1-3)
repeat unit. Both TeRF I and TeRF II have recognition sequences formed by two
copies of the telomeric repeat unit. However, TeRF I appears to be an abundant
DNA binding activity, whilst TeRF II is less abundant. Because Rap1p in budding
yeast also binds to the promoters of many glycolytic and ribosomal protein
genes (
25
-
32
) we have performed preliminary experiments to test whether TeRF I or TeRF II
binds to a ribosomal protein gene promoter from fission yeast. Cross
competition experiments suggested that neither factor interacted with the
single ribosomal protein gene promoter (from the
K5
ribosomal protein gene) that we tested (data not shown).
Telomere binding factors play a role in telomere function and also appear to be
important in controlling the length of telomeres. In budding yeast, changes in
the level of Rap1p have been shown to result in changes to the lengths of
telomeres (
8
,
19
). In addition, it has been proposed recently that factors which bind to the telomeric repeats of
Kluyveromyces lactis
play a role in negatively modulating the activity of telomerase (
52
). As changes in telomerase activity and in the lengths of telomeres have been
implicated in carcinogenesis and cellular ageing, characterisation of telomere
binding proteins and an understanding of their roles, may be of key importance
in understanding these processes (
53
-
55
). If the interactions which we have identified
in vitro
prove to be important
in vivo
,
S.pombe
may be a useful model organism in this regard.
We thank Neal Sugawara for the generous gift of telomere subclones and Jack Szostak for valuable information. We also thank Ian Graham and Liz Packham for critical reading of the manuscript. This work was
supported by the UK Medical Research Council.
REFERENCES
Return