| Nucleic Acids Research | Pages |
Taz1p and Teb1p, two telobox proteins in Schizosaccharomyces pombe, recognize different telomere-related DNA sequences
Introduction
Materials And Methods
Whole cell extract preparation
Expression of full-length telobox proteins
Expression of hybrid MalE-telobox proteins
Band shift assay and DNA probes
Results
Taz1p specifically binds S.pombe telomeric DNA repeats
Binding of Taz1p and TRF1 to G-rich telomeric single-stranded DNA
A second S.pombe protein binds to telomeric repeats
Teb1p binds to vertebrate-type telomeric repeats
Discussion
Acknowledgements
References
Taz1p and Teb1p, two telobox proteins in Schizosaccharomyces pombe, recognize different telomere-related DNA sequences
Received September 9, 1999; Revised and Accepted October 28, 1999
ABSTRACT Band shift assays were used to study proteins from the fission yeast that bind double-stranded telomeric repeat sequences. We also examine general DNA binding properties of the telobox domain, which characterizes telomere-binding proteins from a range of species. We demonstrate that Taz1p has a high affinity for the fission yeast telomeric repeat, consistent with genetic results implicating this protein in telomere maintenance. A second Schizosaccharomyces pombe telobox protein, Teb1p, is shown to bind with high affinity to the vertebrate repeat and with low affinity to the fission yeast telomeric DNA. When tested on G-rich single-stranded telomeric DNA, all these proteins bind with very low affinity, much like the human telomere-binding protein TRF1. Recombinant proteins containing just the telobox domains reproduce the specificity of binding demonstrated for the corresponding full-length proteins, indicating that the telobox domain is indeed responsible for specific DNA recognition. The presence of possible Teb1p-binding sites upstream of many genes suggests a role for this protein as a general transcription factor. Finally, band shift experiments with whole cell extracts from wild-type and taz1- strains suggest that in addition to Taz1p, S.pombe has another major telomere-binding activity.
INTRODUCTION
In eukaryotic organisms the extremities of the chromosomes, or telomeres, are essential to maintain genome integrity by enabling complete DNA replication and by protecting chromosome ends from double-strand break repair (for a review see 1). Telomeres also participate in various aspects of functional organization of the nucleus: they interact with each other and with other substructures like the nuclear scaffold and nuclear envelope, drive polarized chromosome movements during karyogamy and meiotic prophase I, and exert position effects (telomeric position effects) on transcription, the timing of DNA replication, transposition and recombination (reviewed in 2,3). In yeast, telomeres are considered to be heterochromatin-like regions that serve as molecular sinks for factors involved in chromatin-mediated repression and DNA repair (reviewed in 4-7). In many organisms, telomeric DNA is composed of tandemly repeated sequences with a G-rich strand oriented 5[prime]->3[prime] towards the end of the chromosome (for a review see 1). The repeated motif can be regular, like TTAGGG in vertebrates and several flagellate and fungal species, or irregular, like the (TG1-3)n of Saccharomyces cerevisiae. In Schizosaccharomyces pombe, the repeat appears to be unique and several consensus sequences have been proposed, such as T1-2 ACA0-1C0-1G1-6 (1) or G0-3GGTTACA (8), the most frequent motif being GGTTACA (9). We will refer hereafter to GnTTACA as the S.pombe telomeric repeat.
At least two domains can be described within telomeric chromatin. One reflects the binding of specific proteins to the single-stranded 3[prime]-overhang. This domain constitutes the extremity of the chromosome and is essential for chromosome capping and telomerase regulation. The second corresponds to the double-stranded telomeric repeats which are organized, at least in part, in a non-nucleosomal manner. These two domains interact with specific telomeric factors which play a critical role in telomere maintenance, telomere function and telomerase activity (reviewed in 8-10). The structural and functional relationship between these two domains is largely unknown.
The multifunctional repressor-activator protein 1 (Rap1p) in S.cerevisiae was the first protein identified as a telomeric factor binding to the duplex part of the telomeric DNA (for a review see 11). In mice and men, TRF1 and TRF2 are duplex telomeric DNA-binding factors that localize immunologically to the ends of metaphase chromosomes and have essential telomeric functions in cultured cells (12-18). The telomere-associated protein in Schizosaccharomyces pombe, Taz1p, found in a one-hybrid screen using S.pombe telomeric DNA as a target, is required for telomeric functions in interphase cells and for meiosis during sporulation (19-22). Interestingly, Taz1p, TRF1 and TRF2 share a conserved Myb-related sequence involved in specific telomeric DNA recognition, called the telobox (Fig. 1) (13; for a review on telobox proteins see 8). This suggests that Taz1p may bind directly to the duplex part of S.pombe telomeric DNA.
Figure 1. Schematic structure of telobox proteins Taz1p, Teb1p (S.pombe) and human TRF1 as well as fusion proteins between Mal-E and the telobox portion of the corresponding proteins E-Taz and E-Teb. Telobox motifs are shaded and amino acid positions are indicated above.
Because the telobox sequence appears to be a signature motif for a large class of proteins that bind the duplex part of telomeric DNA, we were able to use it as a probe to screen for new telomeric proteins. Accordingly, one uncharacterized fission yeast open reading frame, SpX (8), named here Teb1p for telobox protein 1, was found to contain two telobox sequences (Fig. 1), suggesting that it is an authentic telomeric protein. In addition, in vitro DNA-binding studies have revealed telomeric candidates in fission yeast, namely the TeRFI and II activities (23), and a telomeric DNA-binding activity detected in fission yeast nuclear extracts (N.S.Vassetzky, E.Gilson and S.M.Gasser, unpublished results). In this study we analyze the in vitro DNA-binding specificity of Taz1p and Teb1p and identify a third telomeric DNA-binding activity in crude protein extracts from S.pombe which is distinct from Taz1p and Teb1p.
MATERIALS AND METHODS
Whole cell extract preparation
To prepare the S.pombe whole cell extracts (WCE) we used taz1+ strain 732 (h- ura4-D18) and taz1- strain 1A (h+ taz1::ura4+ ura4-D18 ade6-M210 leu1-32) or J28 (h- taz1::ura4+ ura4-D18), all kindly provided by Dr Julia Cooper. The S.pombe cells were cultivated in 50 ml YES medium to mid-log phase (~5 × 106 cells/ml), centrifuged and washed twice with water. The pellet was resuspended in 20 µl breakage buffer (50 mM KCl, 50 mM Tris-HCl, pH 7.5, 25% glycerol, 2 mM DTT, 0.2 mM PMSF, 0.1% Triton X-100, 25 µg/ml leupeptin, 0.5% v/v aprotinin, 25 µg/ml pepstatin, 25 µg/ml antipain, and 0.3 µg/ml benzamidine) on ice. The cells were broken by vortexing with glass beads for 5 min; the liquid was recovered and vortexing with beads in breakage buffer was repeated three times. Insoluble material was removed from the pooled liquid by centrifugation (13 000 r.p.m., 5 min, 4°C) and protein was estimated using the Bio-Rad protein assay kit (typically ~5 mg/ml).
Expression of full-length telobox proteins
The DNA fragment of the teb1+ cDNA was PCR amplified from a S.pombe cDNA library for the two-hybrid system (Clontech) (using primers gcaggtctagaTCTAAAAGGGAGGTAGCTCAAGTTCCAGG and ctaagatctagaCTATCCCCGGTTGTCCCACGGTATATCCTCGG) and cloned into the pGEM-T vector (Promega). Three independent clones of pGEMT-Teb1 with the cDNA under control of the T7 RNA polymerase promoter were selected. The primers ORF1-A (5[prime]-CGAGGAATTCAACATGGCGGAGGATGTTTCCTCAGC-3[prime]) and Pb (13) were used in RT-PCR in order to recover the hTRF1 cDNA. The reaction was performed on HeLa cell poly(A)+ mRNA, using the TitanTM One Tube RT-PCR System (Boehringer Mannheim). RT-PCR products were then gel purified, cloned into pGEM-T (Promega) and sequenced with an ABI model sequencer. The resulting plasmid was named pGEMT-TRF1. The pBS-taz1 vector carrying the taz1 coding sequence under control of the T7 promoter was kindly provided by Dr Julia Cooper.
Radiolabeled telobox proteins were synthesized using the TnT reticulocyte lysate system (Promega), [14C]leucine and pGEMT-Teb1, pGEMT-TRF1 and pBS-Taz1 plasmid DNA as templates. After checking the expected apparent molecular weight of the translated protein by SDS-PAGE (data not shown), cold proteins were synthesized and the corresponding reticulocyte lysates were used for DNA-protein interactions.
Expression of hybrid MalE-telobox proteins
A DNA fragment carrying the two telobox motifs of teb1+ was PCR amplified from the S.pombe cDNA library for the two-hybrid system (Clontech) using primers gcaggtctagaTCTAAAAGGGAGGTAGCTCAAGTTCCAGG and ctggctctagaCGTAGAACTATCTTTCGGGGTAGCATC. A DNA fragment encoding the telobox of Taz1p was obtained by PCR from pBS-taz1 using the primers acgttgaattcGTGTCCATTGAAAGATCTGCTGCTCGTTCGGG and gcaggtctagaTAAAAGTGGCGGAGTTGCCTCTATGTAAGG. Both telobox-containing DNA fragments were cloned into the XbaI (teb1) or XbaI + EcoRI (taz1) sites of pMal-c2 (Biolabs). At least three independent clones of pE-Teb and pE-Taz were selected. The structures of the corresponding hybrid proteins are presented in Figure 1.
Expression of the proteins in Escherichia coli cells was induced as described elsewhere (13). Correct synthesis was checked by loading the cells directly on SDS-PAGE gels and staining with Coomassie brilliant blue or anti-MalE antibodies (western blotting). Alternatively, the cells from 2 ml of the induced culture were lysed in 200 µl 20 mM Tris-HCl, pH 7.4, 1 mM EDTA, 400 mM NaCl, 0.4 mg/ml lysozyme, 0.2 mM PMSF, 25 µg/ml leupeptin, 0.5% (v/v) aprotinin, 25 µg/ml pepstatin, 25 µg/ml antipain, and 0.3 µg/ml benzamidine on ice for 20 min. After incubation with 0.1 mg/ml sodium deoxycholate at 37°C for 3 min and sonication, the liquid was cleared by centrifugation and used for DNA-protein interactions.
Band shift assay and DNA probes
Unless specifically indicated, 10 µg protein extract was incubated on ice for 5 min in 10 ml of 50-100 mM Tris-HCl, pH 7.5, 50 mM KCl, 5% glycerol, and 0.2 mM PMSF in the presence of non-specific competitor [0.1 mg/ml poly(dI·dC) (Boehringer Mannheim) and 0.1 mg/ml denatured E.coli DNA] and specific competitors. Then, end-labeled DNA probe (0.5 nM) was added and incubation continued for 15 min at room temperature. DNA-protein complexes were separated in 1% agarose gels containing 0.5× TBE for 2-3 h, dried on a DE81 filter in a vacuum heater, and exposed overnight with intensifying screens.
Plasmid pDW232T carries ~33 telomeric repeats of S.pombe (24). We used the 0.3 kb HindIII-EcoRI fragment as the S.pombe telomeric probe (SpTel, Table 1). In order to distinguish telomeric repeat-specific binding from binding to subtelomeric and polylinker sequences present in the S.pombe telomeric probe, we inserted the subtelomeric fragment aagcttGCATGCCTGCAGGTCCCACCCGTCAGCCGAGCCGTAAGGCGAGGCTgaattc into the HindIII and EcoRI sites of pBluescript II SK+ and used the 0.5 kb PvuII fragment of the resulting pSpStel plasmid as probe (SpStel, Table 1).
Table 1. Table 1. DNA fragments used in the band shift assays
| DNA fragment | Relevant sequence | Length (ds/ss) |
| SpTel | HindIII-//-CCCACCCGTCAGCCGAGCCGTAAGGCGAGGCTGC(G1-6TTACA)33-//-EcoRI | 337 nt (ds) |
| SpStel | PvuII-//-CCCACCCGTCAGCCGAGCCGTAAGGCGAGGCTGC-//-PvuII | 488 nt (ds) |
| HuTel | HindIII-//-(TTAGGG)10-//-EcoRI | 80 nt (ds) |
| ScTel | HindIII-//-(TG1-3)100-//-EcoRI | 300 nt (ds) |
| SpTelG | AGCTTACAAGGTTACGTGGTTACACGGTTACAGGTTACAGGTTACAGGGGGG | 52 nt (ss) |
| SpTelC | AATTCCCCCCTGTAACCTGTAACCTGTAACCGTGTAACCACGTAACCTTGTA | 52 nt (ss) |
| HTelG | (TTAGGG)9 | 54 nt (ss) |
| HTelC | (CCCTAA)9 | 54 nt (ss) |
| HIS3 | CCGGTTCGTAAGGCCGACCAGCGA | 24 nt (ss) |
To probe for vertebrate and yeast telomeric repeats we used the 0.06 kb HindIII-EcoRI fragment of pHuTel (25) carrying 10 TTAGGG repeats (HuTel, Table 1) and the 0.3 kb HindIII-EcoRI fragment of pTel270 (25) carrying 100 TG1-3 repeats (ScTel, Table 1), respectively.
RESULTS
Taz1p specifically binds S.pombe telomeric DNA repeats
In order to test whether Taz1p binds to telomeric DNA specifically, the in vitro translated protein was used in a band shift assay with a radiolabeled DNA probe (SpTel, Table 1) containing 33 S.pombe telomeric DNA repeats [(GnTTACA)33], next to a 34 bp fragment of S.pombe telomere-associated sequences. The reticulocyte lysate containing Taz1p gives rise to a discrete complex (Fig. 2A, lane b), while the same type of lysate devoid of Taz1p does not exhibit any DNA-binding activity under the same conditions (data not shown). Competitive binding experiments at low protein:probe ratios were performed to assess the specificity of Taz1p binding to SpTel DNA. A 20-fold excess (10 nM) of cold SpTel is enough to fully displace the radiolabeled probe (Fig. 2A, lanes c-e) while 50 nM DNA fragment containing the subtelomeric sequence without the telomeric repeat (SpStel, Table 1) did not (Fig. 2A, lanes f-h). Therefore, the binding of Taz1p appears to be highly specific for the S.pombe GnTTACA repeat. Increasing concentrations of a cold human telomeric DNA (HuTel, Table 1) was able to compete for Taz1p binding, but much less efficiently than SpTel (Fig. 2A, compare lanes c-e and i-k). This may be due to either a reduced number of telomere repeats in the HuTel probe (10 repeats) as compared to the SpTel probe (33 repeats), or it may indicate that Taz1p binds S.pombe sequence with higher affinity than it binds the human sequence. It is worth noting that Taz1p binding to S.pombe repeats is not at all affected by up to a 100-fold molar excess (50 nM) of S.cerevisiae telomeric DNA (ScTel, Table 1; Fig. 2A, lanes l-n). This indicates that Taz1p is not a general TG-rich DNA-binding protein, but one with selective affinity for fission yeast telomeric repeats.
Figure 2. DNA binding affinity of in vitro translated Taz1p in a band shift assay with a S.pombe telomeric probe. (A) Full-length Taz1p translation by the reticulocyte system. (a) Probe alone; (b) probe + protein; titration with telomeric DNA specific for (c-e) S.pombe, (i-k) vertebrates, and (l-n) S.cerevisiae, as well as (f-h) subtelomeric S.pombe DNA. (B) Taz1 telobox-Mal-E hybrid translation in E.coli. (a) Probe alone; (b) probe + protein; titration with S.pombe (c-f) telomeric and (g-j) subtelomeric DNA. (C) Full-length Taz1p translation by the reticulocyte system. (a) Probe alone; (b) probe + protein; titration with double-stranded (c-e) S.pombe telomeric DNA and single-stranded (f-k) S.pombe and (l-q) vertebrate telomeric DNA, as well as (r-t) S.pombe HIS3 gene fragment. The arrow indicates the unbound probe position; competitor amounts are indicated above.
In order to determine whether the telobox domain is responsible for the Taz1p DNA-binding specificity, we expressed the C-terminal part of Taz1p (comprising the telobox, between positions 533 and 663) in E.coli cells as a fusion protein with MalE (Fig. 1, E-Taz hybrid). Band shift assays reveal a SpTel-binding activity in the bacterial extracts of cells induced for the expression of E-Taz (Fig. 2B), while no activity is detected in extracts from non-induced cells (data not shown). The interaction is specific for S.pombe GnTTACA repeats, since it can be competed by SpTel but not by SpStel DNA (Fig. 2B, compare lanes c-f and g-j). In conclusion, the telobox motif is sufficient to confer specific interactions with GnTTACA DNA repeats.
Binding of Taz1p and TRF1 to G-rich telomeric single-stranded DNA
Since telomeric DNA at chromosome ends is composed of a duplex part and a 3[prime] G-rich overhang, we evaluated whether the Taz1p-probe complex can be competed by single-stranded G-rich oligonucleotide. Indeed, the interaction between Taz1p and SpTel can be efficiently competed by 200 µM oligonucleotide corresponding to the G-rich strand of S.pombe or human telomeric repeats (HTelG and SpTelG, Table 1; Fig. 2C, lanes f-h and l-n). This interaction was specific for the G-rich oligonucleotides since in the same range of concentration, the corresponding C-strand telomeric oligonucleotide (HTelC and SpTelC, Table 1) and a non-telomeric oligonucleotide (HIS3, Table 1) did not compete for the interaction of Taz1p with SpTel (Fig. 2C, lanes i-k and o-t). Surprisingly, an increased concentration of the C-rich or HIS3 oligonucleotides enhanced the intensity of the Taz1p-DNA complex. This phenomenon not being due to a spurious oligonucleotide-binding activity in the reticulocyte lysate (data not shown), we conclude that a very large excess of these oligonucleotides actually improves Taz1p binding. Since unrelated oligonucleotides have similar effects, this improved binding appears to be mainly non-sequence-specific. Whether this effect results from the decreased void volume of the reaction, from the removal of a competitor or inhibitor of the binding reaction, and/or from complex allosteric effects between the single-stranded competitor and the DNA-binding complex requires further studies.
Because the ability of other telobox proteins to bind G-rich single-stranded DNA is unknown, human TRF1 was tested in a similar assay. In vitro translated TRF1 demonstrated a clear preference for binding to HTelG, as opposed to HTelC: competition for the double-stranded human telomeric repeat was complete at 20 µM HTelG and only partial at 200 µM HTelC (Fig. 3, lanes f-h). In this case, we did not observe any increase in double-stranded complex formation when HTelC was added, whereas a partial competition was detected at 200 µM (Fig. 3, lane k). This further suggests that the non-specific effect of nucleotides on Taz1p binding is a sequence-independent concentration effect. Overall, we conclude that Taz1p and TRF1 bind the G-rich strand of telomeric DNA with a much lower affinity than the corresponding duplex telomeric DNA.
Figure 3. Band shift assay using the vertebrate telomeric probe and TRF1 translated in vitro by the reticulocyte system. (a) Probe alone; (b) probe + protein; titration with (c-e) double-stranded vertebrate telomeric DNA and (f-k) single-stranded vertebrate DNA. The arrow indicates the unbound probe position; competitor amounts are indicated above.
A second S.pombe protein binds to telomeric repeats
In order to identify the Taz1p-dependent binding activity in S.pombe extracts, we compared the pattern of complexes formed on incubation of SpTel with WCE prepared from either taz1+ or taz1- cells. With increasing amounts of the taz1+ WCE, four complexes (designated C1-C4; Fig. 4A, lanes b-e) were observed. One of these complexes (C2) was clearly missing from the taz1- WCE (compare lanes b-e with f-i in Fig. 4A, Fig. 4B with D and lane c with d in Fig. 4C). Increasing concentrations of the unlabeled competitor SpTel resulted in loss of the C2 complex prior to loss of the C1 complex (Fig. 4B, lanes a-c). Both C1 and C2 appeared to be specific for the GnTTACA repeats, since the competition of C1 and C2 by SpStel is inefficient when compared to that by SpTel (Fig. 4B, lanes d-f). These data indicate that the C1 and C2 complexes are highly specific for S.pombe telomeric repeats, and the absence of C2 in a taz1- WCE strongly suggests that C2 contains Taz1p. Interestingly, the complex formed with the Taz1p protein translated in vitro migrates differently than C2 (Fig. 4C, compare lanes b and c), suggesting that the protein may be modified post-translationally in intact cells. Since the lower migrating complexes C3 and C4 are present in the taz1- WCE and since their relative amount is augmented with increasing extract concentration, they are likely to represent multiple binding to the probe of the protein forming the C1 complex.
Figure 4. Band shift assay using S.pombe telomeric DNA and whole cell extracts (WCE). (A) (a) Probe alone; increasing amounts of (c-e) taz1+ WCE and (f-i) taz1- WCE: 2 (b,f), 4 (c,g), 8 (d,h) or 16 µg (e,i). (B) taz1+ WCE titrated with S.pombe (a-c) telomeric and (d-f) subtelomeric DNA. (C) (a) Probe alone; (b) in vitro translated Taz1p; (c) taz1+ WCE; (d) taz1- WCE. (D) taz1- WCE. (a) Probe alone; (b) probe + extract; titration with telomeric DNA specific for (c-e) S.pombe, (i-k) vertebrates, and (l-n) S.cerevisiae as well as (f-h) subtelomeric S.pombe DNA. The arrow indicates the unbound probe position; C1-C4 complexes (see text) are marked on the right; competitor amounts are indicated above.
The C1 complex present in taz1- WCE is competed much more efficiently by SpTel than by SpStel, while no competition is observed with HuTel or with ScTel (Fig. 4C). We conclude that formation of the C1 complex is highly specific for S.pombe telomeric repeats and results from a specific interaction of telomeric repeats with an S.pombe protein distinct from Taz1p.
Teb1p binds to vertebrate-type telomeric repeats
An obvious candidate for the second telomeric DNA-binding protein is the previously described S.pombe ORF spX, which contains two telobox motifs in its N-terminal domain (8). The corresponding gene is renamed here teb1 for telobox protein 1. In Figure 5, we show that in vitro translated Teb1p binds the SpTel probe, although this interaction is more efficiently competed by unlabeled HuTel than by SpTel, SpStel, or ScTel (Fig. 5A). The high affinity of Teb1p for human telomeric repeats was confirmed using HuTel as probe in the band shift assay (Fig. 5B). Reminiscent of the Taz1p binding properties, Teb1p binding can be competed by a large excess of S.pombe or human G-rich telomeric oligonucleotides (Fig. 5B, lanes f-h and l-n, respectively), but not by the corresponding C-strands (Fig. 5B, lanes i-k and o-q, respectively). Furthermore, using a hybrid protein expressed in E.coli, composed of the teb1 telobox N-terminal domain and MalE (E-Teb, Fig. 1), the specificity for S.pombe telomeric repeats was confirmed (Fig. 6A). Interestingly, specific competition by the G-rich telomeric DNA was also observed with E-Teb (Fig. 6B), strongly suggesting that the ability to bind G-rich DNA is characteristic of the telobox.
Figure 5. DNA binding affinity of full-length Teb1p in a band shift assay. (A) Teb1ptranslated in vitro by the reticulocyte system and S.pombe telomeric probe. (a) Probe alone; (b) probe + protein; titration with telomeric DNA specific for (c-e) S.pombe, (i-k) vertebrates, and (l-n) S.cerevisiae, as well as (f-h) subtelomeric S.pombe DNA. (B and C) Full-length Teb1p translation (B) and taz1- WCE (C) with vertebrate telomeric probe. (a) Probe alone; (b) probe + protein; titration with (c-e) double-stranded vertebrate telomeric DNA and single-stranded telomeric DNA specific for (f-k) S.pombe and (l-q) vertebrates. The arrow indicates the unbound probe position; competitor amounts are indicated above.
Figure 6. Analyses of the DNA binding affinity of in vitro translated Teb1p in a band shift assay with vertebrate telomeric probe. (A) (a) Probe alone; (b) probe + protein; titration with double-stranded telomeric DNA specific for (c-e) S.pombe, (i-k) vertebrates, and (l-n) S.cerevisiae, as well as (f-h) subtelomeric S.pombe DNA. (B) (a) Probe alone; (b) probe + protein; titration with (c-e) double-stranded S.pombe telomeric DNA and single-stranded telomeric DNA specific for (f-k) S.pombe and (l-q) vertebrates. The arrow indicates the unbound probe position; competitor amounts are indicated above the gel.
The use of a radiolabeled HuTel probe with taz1- WCE revealed a complex (Fig. 5C) with an apparent mobility similar to that obtained with in vitro translated Teb1p (data not shown). Furthermore, the competition pattern of this complex is indistinguishable from that obtained with Teb1p (compare Fig. 5B and C). These results strongly suggest that the HuTel complex detected in the taz1- WCE corresponds to one involving Teb1p. The higher affinity demonstrated for human, rather than for S.pombe, telomeric repeats by the in vitro synthesized Teb1p, E.coli-synthesized E-Teb1p, and natural yeast Teb1p suggests that the second S.pombe telomeric DNA-binding protein identified in the taz1- extract is unlikely to be Teb1p (see above).
DISCUSSION
This work provides the first direct demonstration that Taz1p binds the S.pombe telomeric DNA (GnTTACA)n in a sequence-specific manner. This specificity is fully consistent with activities monitored by the one-hybrid assay, the presence of a telobox motif, and the telomere-associated phenotypes that correlate with loss of taz1 (19-22). Although the significance is still unclear, we also note that Taz1p exhibits a much higher affinity for vertebrate-type repeats (TTAGGG)n than for S.cerevisiae repeats (TG1-3)n.
In an earlier study, Duffy and Chambers (23) detected four DNA-protein complexes specific for telomeric repeats in S.pombe cell extracts. They proposed that these correspond to two or three different proteins. Indeed, in crude protein extracts prepared from cells lacking Taz1p, we detected a major S.pombe telomeric DNA-binding activity. In particular, this activity exhibits a much higher affinity for S.pombe than for vertebrate-type telomeric DNA. We expected that this binding was due to the product of an S.pombe ORF previously identified by the presence of two teloboxes at its N-terminus (8). However, the in vitro translated Teb1p, as well as its N-terminal domain produced in bacteria as a fusion protein with MalE, have much higher affinity for the vertebrate-type repeat (TTAGGG) than for the S.pombe telomeric repeat (GnTTACA). This difference in specificity makes it unlikely that the telomere-binding factor in taz1- cells corresponds to Teb1p. Alternatively, this activity could be due to another telobox-containing ORF recently sequenced by the S.pombe genome project (EMBL accession no. AL021839.1). This sequence exhibits a high overall similarity to the S.cerevisiae telobox protein Tbf1p, which itself recognizes the TTAGGG repeats that are found in subtelomeric regions, not far from the authentic S.cerevisiae telomeric repeats (13,26,27). Using an in vitro translated cDNA corresponding to this protein, we have failed, however, to observe any DNA-binding activity (unpublished observations). This protein may have no DNA-binding activity on its own, or its activity might be affected by the absence of additional S.pombe factors, abnormal in vitro synthesis, and/or a lack of post-translational processing. Whether the non-Taz1p activity binding to S.pombe repeats corresponds to Tbf1p, to another as yet unidentified telobox protein, or to a protein belonging to another family of telomere-binding factors remains to be determined.
It is worth noting that none of the five characterized telobox-containing proteins (human TRF1 and TRF2, fission yeast Taz1p and Teb1p, and budding yeast Tbf1p) show specific interaction with S.cerevisiae telomeric DNA, which is recognized by the non-telobox protein Rap1p. In fact, both vertebrate-type and S.pombe telomeric DNA contain multiple GnTTA motifs (8,9) which are related to the core sequence of the TRF1 telobox binding site (GGGTTA) (28). Therefore, teloboxes might be general GnTTA-binding domains, which explains why they might recognize telomeric DNA of S.pombe and vertebrates, but not of S.cerevisiae. Nevertheless, the telobox proteins can vary in their relative affinity for GnTTA variants. For instance, Taz1p appears to exhibit a preference for S.pombe repeats as compared to vertebrate-type repeats, while Teb1p displays the reverse specificity. This may be due to variations in the telobox sequence, to different spatial arrangements of teloboxes, or to the presence of additional DNA-binding domains.
The function of Teb1p is unknown, but its binding properties suggest that it may bind interspersed TTAGGG repeats in the S.pombe genome. Therefore, we made a BLAST search for the sequence (TTAGGG)3 in the available S.pombe genomic sequences. Strikingly, 23 out of the 24 occurrences lie in intergenic regions (Table 2). They include the previously described conserved sequence present in the intergenic spacer sequences or in the 5[prime] upstream region of all histone genes, named the AACCCT box (29). Others are located in the vicinity of the transcriptional start site of various genes, including heat shock factors, translation factors, splicing factors, and GTP-binding proteins. One notes that all these sequences are conserved (Table 2) and are highly similar to the vertebrate telomeric repeat TTAGGG (Fig. 7). In agreement with the DNA-binding properties of Teb1p presented in this work, we recently showed that Teb1p specifically binds to the AACCCT box DNA (unpublished observation). These data strongly suggest that vertebrate-type telomeric repeats participate in the expression of S.pombe gene families involved in essential cellular functions, including transcription, translation and chromatin structure.
Figure 7. TTAGGG-related sequences in the S.pombe genome (see text for explanation)
Table 2. Table 2. ACCCT boxes and their flanking genes
| Sequence | D | Gene name | Function | Locus |
| TTCAGGGTTAGGGTTTCTG | 111 | tdf1 | TFIID | SPAC29E6 |
| TACAGGGTTAGGGTTTGAG | 46 | snu5 | U5 snRNA | SPBC19F8 |
| AGCAGGGTTAGGGTTTTAA | 1631/454 | SPBC30B4.03c/snu14 | ?/U14 snRNA | SPBC30B4 |
| ATCAGGGTTAGGGTTTTGT | 104 | spi1 | GTP-binding | SPBC1289 |
| CCCAGGGTTAGGGTTGTCC | 2946 | SPAC27E2.03c | GTP-binding | SPAC27E2 |
| TTTA-GGTTAGGGTTAATG | 66 | rpl3 | Ribosomal protein | SPBC24E9 |
| GTGTGGGTTAGGGTTTGCC | 435/1582 | SPBC6B1.01c/SPBC6B1.02 | Helicase/? | SPBC6B1 |
| TCGAGGGTTAGGGTACGGG | 484 | SPAC4A8.03c/isp6 | ?/Ser protease | SPAC4A8 |
| CACAGGGTTAGGGTTTCAG | 71 | BAA23591 | EF-2 | D83976 |
| GTAAGGGTTAGGGTTGATT | 193/1746 | tif51/SPAC26H5.11 | eIF5/? | SPAC26H5 |
| AGCAGGGTTAGGGTTTTGA | 166/250 | H3.1/H4.1 | Histones | SPHIS41 |
| ATCAGGGTTAGGGTTGTGA | 235/81 | H4.2/H3.2 | Histones | SPBC8D2 |
| TAAAGGGTTAGGGTTGTGA | 128/92 | H4.3/H3.3 | Histones | SPHIS33 |
| ACCAGGGTTAGGGTTGTGA | 167/370 | H2A.1/H2B.1 | Histones | SPCC622 |
| TACAGGGTTAGGGTTTTGA | 124/428 | H2A.2/H2B.2 | Histiones | SPAC19G12 |
| TCGAGGGTTAGGGTTGGTG | 265/942 | SPAC13G7.02c/ SPAC13G7.03 | hsp70 family/? | SPAC13G7 |
| GCGAGGGTTAGGGTTTTGG | 260 | SPCC1739.13 | hsp70 family | SPCC1739 |
| ACCAGGGTTAGGGTTTTGG | 309 | hsp90 | hsp90 | YSPHSP90X |
| TTCAGGGTTAGGGTTATCG | 445 | SPAC4H3.11c | ? | SPAC4H3 |
| AACGAGGTTAGGGTTATGG | 1568 | SPBC24C6.09c | ? | SPBC24C6 |
| TTTAGG-TTAGGGTTAATG | 414/295 | SPB24E9.11c/SPBC24E9.12 | ?/? | c839 |
| AAACAGGTTAGGGTTATAA | 458/1276 | SPAC10F6.14c/ SPAC10F6.15 | ?/? | SPAC10F6 |
| TTCCTGGTTAGGGTTAAAG | 1987 | SPCC1183.11 | ? | SPCC1183 |
Another sequence motif of S.pombe, previously identified in the promoter of several ribosomal protein genes as the Homol E-box (30,31), is also reminiscent of vertebrate-type telomeric repeats (Fig. 7). Homol E-box sequences are proximal activation sequences and protein binding sites (30). Since Teb1p binds with high affinity to a (TTAGGG)10 sequence, it might also bind to the AACCCT or to Homol E-boxes, raising the interesting possibility that Teb1p is a general transcriptional regulator. In addition, Teb1p may bind in vivo to S.pombe telomeric regions, since it also has a weak but detectable affinity for the S.pombe telomeric repeat.
We observed an unusual affinity of the tested telobox proteins (Taz1p, Teb1p and TRF1) for the G-rich strand of telomeric DNA. This affinity is revealed by displacement of the binding of these proteins to the radiolabeled double-stranded probe by a large excess of single-stranded oligonucleotides. The affinity of these proteins for the single-stranded G-rich DNA is very low as compared to that for the double-stranded telomeric DNA, the competition starting at micromolar concentrations of the oligonucleotides, i.e. over 100-fold concentration of the corresponding double-stranded DNA required for a similar level of competition (Figs 2C, 3, 5B and 6B). Importantly, this binding to the G-rich strand is specific: the same range of concentrations of the corresponding C-strand oligonucleotide failed to significantly compete binding of the probe. This property, reminiscent of Rap1p (25,32), may be a clue to the natural function of these proteins considering the presence of a single-stranded G-tail at chromosome ends. Furthermore, TRF2, another telobox protein, may be implicated in the structure of the G-rich overhang of human chromosomes (17,33). It is tempting to speculate that the capacity of telomeric proteins to bind both the duplex part of telomeric DNA and the G-rich tail plays an important role in the association of single-stranded with double-stranded telomeric DNA in the so-called t-loop. Indeed, Rap1p was shown to stimulate a similar single- to double-stranded association of yeast telomeric DNA in vitro (25).
The similarity in the binding properties between Taz1p and TRF1 substantiates the proposal that Taz1p is a functional analog of TRF1 (19). It appears likely that S.pombe has another telomeric DNA-binding protein, since we detect an additional band shift activity in taz1- cells. This situation is similar to that in vertebrates where two telobox proteins, TRF1 and TRF2, are involved in telomere function. Further studies will determine which, if any, of the three fission yeast telobox proteins actually binds fission yeast telomeres in vivo.
ACKNOWLEDGEMENTS
We thank J. Cooper for kindly providing S.pombe strains and the pBS-Taz plasmid, J. Berman for plasmid pDW232T, D. Rhodes for pTel270, and J.M. Clément for anti-MalE antibodies. The work in E. Gilson's laboratory was supported by the Ligue Nationale contre le Cancer and the CNRS program Génome. The work in S. Gasser's laboratory was supported by the Swiss National Science Foundation and the Human Frontiers Science Program. Nikita Vassetzky was supported by Russian Foundation for Basic Research (96-04-49561) and Russian Human Genome Program (5-95/99-3).
REFERENCES
*To whom correspondence should be addressed. Tel: +33 4 72 72 84 53; Fax: +33 4 72 72 80 80; Email: eric.gilson{at}ens-lyon.fr
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. W. Pitt, L. P. Valente, D. Rhodes, and T. Simonsson Identification and Characterization of an Essential Telomeric Repeat Binding Factor in Fission Yeast J. Biol. Chem., February 1, 2008; 283(5): 2693 - 2701. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Tomaska, S. Willcox, J. Slezakova, J. Nosek, and J. D. Griffith Taz1 Binding to a Fission Yeast Model Telomere: FORMATION OF TELOMERIC LOOPS AND HIGHER ORDER STRUCTURES J. Biol. Chem., December 3, 2004; 279(49): 50764 - 50772. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Yang, S. K. Kim, and W. T. Kim Perturbation of NgTRF1 Expression Induces Apoptosis-Like Cell Death in Tobacco BY-2 Cells and Implicates NgTRF1 in the Control of Telomere Length and Stability PLANT CELL, December 1, 2004; 16(12): 3370 - 3385. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Sadaie, T. Naito, and F. Ishikawa Stable inheritance of telomere chromatin structure and function in the absence of telomeric repeats Genes & Dev., September 15, 2003; 17(18): 2271 - 2282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. W. Yang, D. H. Kim, J. J. Lee, Y. J. Chun, J.-H. Lee, Y. J. Kim, I. K. Chung, and W. T. Kim Expression of the Telomeric Repeat Binding Factor Gene NgTRF1 Is Closely Coordinated with the Cell Division Program in Tobacco BY-2 Suspension Culture Cells J. Biol. Chem., June 6, 2003; 278(24): 21395 - 21407. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Nakanishi, T. Kawai, F. Kumaki, S. Hiroi, M. Mukai, E. Ikeda, C. E. Koering, and E. Gilson Expression of mRNAs for Telomeric Repeat Binding Factor (TRF)-1 and TRF2 in Atypical Adenomatous Hyperplasia and Adenocarcinoma of the Lung Clin. Cancer Res., March 1, 2003; 9(3): 1105 - 1111. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Ancelin, M. Brunori, S. Bauwens, C.-E. Koering, C. Brun, M. Ricoul, J.-P. Pommier, L. Sabatier, and E. Gilson Targeting Assay To Study the cis Functions of Human Telomeric Proteins: Evidence for Inhibition of Telomerase by TRF1 and for Activation of Telomere Degradation by TRF2 Mol. Cell. Biol., May 15, 2002; 22(10): 3474 - 3487. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-L. Mergny, J.-F. Riou, P. Mailliet, M.-P. Teulade-Fichou, and E. Gilson Natural and pharmacological regulation of telomerase Nucleic Acids Res., February 15, 2002; 30(4): 839 - 865. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E. Koering, G. Fourel, E. Binet-Brasselet, T. Laroche, F. Klein, and E. Gilson Identification of high affinity Tbf1p-binding sites within the budding yeast genome Nucleic Acids Res., July 1, 2000; 28(13): 2519 - 2526. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Y. Yu, S. E. Kim, J. H. Kim, J. H. Ko, M. H. Cho, and I. K. Chung Sequence-specific DNA Recognition by the Myb-like Domain of Plant Telomeric Protein RTBP1 J. Biol. Chem., July 28, 2000; 275(31): 24208 - 24214. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||