DDBJ/EMBL/GenBank accession no. AF006380
ABSTRACT
G-DNA, a polymorphic family of four-stranded DNA structures, has been proposed to play roles in a variety of biological processes including telomere function, meiotic recombination and gene regulation. Here we report the purification and cloning of TGP1, a G-DNA specific binding protein from Tetrahymena thermophila. TGP1 was purified by three-column chromatographies, including a G-DNA affinity column. Two major proteins (~80 and ~40 kDa) were present in the most highly purified column fraction. Renaturation experiments showed that the ~80 kDa protein contains TGP1 activity. Biochemical characterization showed that TGP1 is a G-DNA specific binding protein with a preference for parallel G-DNAs. The TGP1/DNA complex has a dissociation constant (Kd) of ~2.2 × 10-8 M and TGP1 can form supershift in gel mobility shift assays. The cDNA coding TGP1 was cloned and sequenced based upon an internal peptide sequence obtained from the ~80 kDa protein. Sequence analyses showed that TGP1 is a basic protein with a pI of 10.58, and contains two extensively hydrophilic and basic domains. Homology searches revealed that TGP1 is a novel protein sharing weak similarities with a number of proteins.
G-quartets are unusual nucleic acid structures first described about three decades ago (reviewed in 1). In a G-quartet, four guanine molecules lie in a plane with each guanine forming G:G (instead of Watson-Crick A:T or G:C) hydrogen bonds with two adjacent guanines (1-3). DNA containing such structures is known as quadruplex, tetraplex or G-DNA (3). Many guanine-rich sequences of biological significance have been found to be capable of forming G-quartet structures in vitro under physiological conditions. Such G-rich sequences include most telomeres (4-6), immunoglobulin switch regions (7), a few gene promoters (8-10), fragile X repeats (11) and the dimerization domain in the human immunodeficiency virus (HIV) genome (12,13).
The ability to form G-DNA by these important sequences implies that G-DNA may be biologically relevant. However, direct evidence for the existence of G-DNA in vivo is lacking. An alternative approach to study G-DNA is to identify and investigate proteins that interact with this structure. A number of proteins have been found to be able to bind to G-DNA. These proteins include a hepatocyte chromatin protein QUAD (14), transcription factor MyoD (15), chick topoisomerase II (16), two yeast proteins G4p1 and G4p2 (17,18) and the yeast KEM1 protein (19). In contrast to all other G-DNA binding proteins, KEM1 not only binds to G-DNA, but also acts as a nuclease to cut the single-stranded DNA 5' to the G-quartet domain (19,20). The G-DNA specific nuclease activity of KEM1, together with the studies in which meiotic recombination was found to be severely affected in KEM1 deletion mutants (21,22), suggests that KEM1 and the G-DNA structure play roles in meiotic pairing (20). In addition to G-DNA binding proteins, two proteins have been found to facilitate the formation of G-DNA structure: the Oxytricha telomere binding protein [beta] subunit, which acts as a molecular chaperone (23,24), and the multifunctional yeast telomere protein RAP1 (25). Since both proteins are telomere-related, their G-DNA promotion activities suggest that G-DNA may have a role in telomere function.
The biological relevance of G-DNA was further suggested by several other studies showing that G-DNA can inhibit the activities of a number of proteins. One of these G-DNA-inhibited proteins is telomerase, the enzyme that synthesizes telomeres (26). It was found that folding of telomere DNA (specifically the primer) into G-DNA inhibited the activity of telomerase in vitro, suggesting that formation of G-DNA in the primer could down-regulate telomere elongation in vivo (27). Another in vitro study showed that an antiparallel G-DNA aptamer can inhibit thrombin activity (28). G-DNA was also shown to inhibit the activity of macrophage scavenger receptors, the glycoproteins which may function in the deposition of lipoprotein cholesterol (29). More interestingly, an oligonucleotide in G-DNA form was found to be a potent inhibitor of HIV-1 integrase (30). These inhibitory activities of G-DNA suggest its potential use as a pharmaceutical agent.
Possible biological functions for G-DNA are diverse, but remain to be definitively demonstrated. Further studies on G-DNA and its related proteins are needed to elucidate the exact role(s) of these structures in vivo. The ciliated protozoan, Tetrahymena thermophila, has proven to be a useful system for identification of G-DNA binding proteins. This organism has two nuclei, one of which contains a fragmented genome with thousands of telomeres (31). A relatively abundant G-DNA binding activity has been identified in T.thermophila (32). The protein, TGP1 (Tetrahymena G-DNA binding protein 1), binds to an intermolecular G-DNA form of d(TTGGGG)4 under physiological conditions, and competition experiments showed that G-DNAs competed strongly for TGP1 binding, while non-G-DNA and G-RNA oligonucleotides did not (32). In this paper, we report the purification, molecular cloning and further biochemical characterization of TGP1. TGP1 was shown to be an 83 kDa protein with binding activity specific for G-DNA. Complete cDNA sequence of TGP1 was obtained and shown to encode a novel protein with two extensively hydrophilic and basic putative DNA binding domains and weak similarities to a number of other proteins. In light of these data, possible functions for TGP1 are discussed.
Table 1.
Oligonucleotides (oligos) (Table 1) were synthesized on an ABI 394 DNA/RNA synthesizer (ISU DNA facility), and purified as previously described (32). Briefly, oligos were separated on a 20% denaturing polyacrylamide gel (7 M urea), then desired oligo products were excised and purified by C-18 chromatography (Waters). 5'-32P-labeling of oligos using T4 polynucleotide kinase was carried out according to a standard protocol (33). Radiolabeled oligos were purified by G-25 spin columns (5 prime -> 3 prime).
Fifty nanograms of 32P-labeled oligo Y were boiled for 3 min in G-DNA formation buffer (50 mM KCl, 10 mM MgCl2, 10 mM Tris-HCl, pH 8.0, 5% glycerol), cooled and incubated at room temperature for at least 30 min to make G-DNA. Mobility shift experiments were performed as previously described with minor modifications (32). About 2.5 ng of labeled Y(G4) were mixed with desired amounts of protein extract or TGP1 fractions in binding reaction buffer (10 mM Tris-HCl, pH 7.5, 6% glycerol). One hundred fold (~250 ng) non-specific competitor poly(dI·dC) (Pharmacia) was added to each binding reaction. The total volume of each reaction was 20 µl. After incubation on ice for 20 min, the reaction mixtures were loaded onto a 6% polyacrylamide gel. Electrophoresis was carried out in 0.6× TBE at room temperature. The gel was then vacuum-dried, and exposed to X-ray film or a PhosphorImager screen (Molecular Dynamics).
For competition assays, unlabeled competitors were made as follows. Single-stranded DNA (Y) was made by boiling in ddH2O for 3 min to prevent G-quartet structure formation, and chilling on ice for 30 min. Double-stranded DNAs (Y-cY, Tet4-cTet4) were formed by annealing of single-stranded oligos with their complementary strands at 95°C for 5 min, 65°C for 10 min and 37°C for 10 min in buffer containing 50 mM NaCl and 10 mM Tris-HCl, pH 7.5. G-DNAs were formed according to the methods described by Sen and Gilbert (34). In mobility shift assays, unlabeled competitors were diluted as indicated and added to the binding buffer before the addition of TGP1 and labeled Y(G4) probe. About 50 ng of TGP1 and saturating amounts (2.5 ng) of labeled probe Y(G4) were used in each binding reaction. Twenty units of RNasin ribonuclease inhibitor (Promega) were included in reactions where RNA oligo (rTet4) was used. Binding reactions and mobility shift experiments were performed as described above.
SDS-PAGE was performed according to the standard protocol (33). After electrophoresis, protein bands were visualized by silver staining following the manufacturer's protocol (BioRad).
Cross-linking of G-DNA to TGP1 was performed in situ by exposing a wet 8% mobility shift gel on ice to 254 nm UV light with the gel no more than 5 cm from the UV source. The gel was then exposed to film to reveal the position of the TGP1/G-DNA binding complex. The gel piece containing the complex was excised, denatured in SDS sample buffer for 5 min, and polymerized into the stacking gel of a 10% SDS-polyacrylamide gel. After electrophoresis, the gel was dried and exposed to X-ray film. Molecular weight markers were included to identify the positions of cross-linked bands.
Tetrahymena cell culture and total protein extract preparation were performed as previously described (32). Briefly, T.thermophila cells (strain C3V) were grown vegetatively to mid-log phase (2.5 × 105 cells/ml) and harvested. Cells were washed twice with 10 mM Tris-Cl, pH 7.5, resuspended in 5 volumes of TMG buffer (10 mM Tris-Cl, pH 7.5, 1 mM MgCl2, 10% glycerol and 10 mM [beta]-mercaptoethanol) with protease inhibitors (0.01 mM Leupeptin, 0.01 mM Pepstatin and 0.1 mM Pefabloc, all from Boehringer-Mannheim), and lysed by addition of one tenth volume of 2% NP-40 (Sigma). The cell lysate was centrifuged at 100 000 g for 70 min. The supernatant (S100) was aliquoted, immediately frozen in liquid nitrogen and stored at -70°C.
All purification steps were conducted at 4°C unless otherwise indicated. Tetrahymena S100 protein extract (100 ml, ~10 mg/ml) was quickly thawed and filtered through a 0.4 µm filter (Costar). The filtrate was then loaded onto a SP-Sepharose column (4.91 cm2 × 14.5 cm, Pharmacia), which had been equilibrated and packed with at least 2 column volumes of chromatography buffer [CB buffer: 10 mM Tris-HCl, pH 7.7, 1 mM EDTA, 0.01% NP-40, 10% glycerol and 1 mM dithiothreitol (DTT)], at a flow rate of 0.8 ml/min. After the extract had passed through the column, 1.5 column volumes of CB buffer were passed through the column at the same flow rate (0.8 ml/min). The column was then washed with 2 column volumes of CB buffer containing 0.45 M NaCl at a flow rate of 2 ml/min to remove weakly bound proteins. Four column volumes of CB buffer containing 0.8 M NaCl were then applied to elute TGP1 activity. About sixty 3 ml fractions were collected and assayed for TGP1 activity in mobility shift experiments.
Fractions from the SP-Sepharose column containing TGP1 activity were pooled and dialyzed against CB buffer overnight at 4°C. The fractions were then loaded at a flow rate of 0.5 ml/min onto a DE52 column (3.21 cm2 × 4 cm, Whatman), which had been equilibrated with CB buffer. The column was then washed with 1 column volume of CB buffer at a rate of 0.5 ml/min. A linear NaCl gradient (0-0.8 M in CB buffer, 2 column volumes) was applied to elute TGP1 activity at a flow rate of 0.4 ml/min. About thirty 2 ml fractions were collected and assayed in mobility shift experiments.
A G-DNA affinity column (~2 ml) was constructed according to a published procedure (20) with minor modifications. Briefly, 200 µg of 3'-biotinylated oligo Y (Midland) were mixed with 1 mg of oligo Y (5-fold) in the presence of 10 mM MgCl2 and 100 mM KCl. The mixture was boiled for 3 min and then cooled to room temperature for 30 min to allow G-DNA formation. The biotinylated Y(G4) was then mixed with 2 ml of 50% avidin-agarose (Pierce), which had been equilibrated with CB buffer containing 100 mM KCl. After incubation at room temperature for 1 h, the mixture was poured and packed into a 5 ml disposable column (BioRad). Fractions from the DE52 column containing TGP1 activity were pooled and dialyzed against CB buffer overnight at 4°C. The fractions were then loaded at a flow rate of 0.1 ml/min onto the G-DNA affinity column, which had been equilibrated with CB buffer containing 0.1 M KCl. After the sample passed through, the column was washed with 1 column volume of 0.1 M KCl/CB buffer, and proteins were eluted with a linear KCl gradient (0.1-2.0 M) in CB buffer. About thirty 0.2 ml fractions were collected and tested for TGP1 activity in mobility shift experiments.
Renaturation experiments were performed using a procedure similar to that described by Hager and Burgess (35). Briefly, partially purified TGP1 (~10 µg) was displayed on a 10% SDS-PAGE, and the gel was then stained with Coomassie blue. Gel slices containing different protein bands were excised, rinsed with ddH2O, put into different tubes, and crushed into small pieces. One milliliter of elution buffer (50 mM Tris-HCl, pH 7.5, 0.1% SDS, 0.1 mg/ml BSA, 1 mM DTT, 0.2 mM EDTA and 2.5% glycerol) was then added to each tube. Proteins were eluted from the gel into buffer with agitation for at least 2 h at room temperature. The mixtures were centrifuged for 2 min at 10 000 r.p.m. Supernatants were transferred to fresh tubes, mixed with 4 volumes of 100% acetone (-20°C), stored at -20°C for >2 h and centrifuged for 10 min at 10 000 r.p.m. Precipitates were rinsed with 0.5 ml of 100% methanol (-20°C) twice, air dried and dissolved in 5 ml of 6 M guanidine-HCl in dilution buffer (50 mM Tris-HCl, pH 7.6, 20% glycerol, 0.1 mg/ml BSA, 0.15 M NaCl, 1 mM DTT and 0.1 mM EDTA) for 20 min at room temperature. Solutions were 50-fold diluted with dilution buffer, and proteins were allowed to renature overnight at room temperature. Renatured proteins were tested for TGP1 activity in mobility shift experiments.
Affinity purified TGP1 fractions were pooled, dialyzed against CB buffer and concentrated using Centricon concentrators (Amicon). TGP1 protein concentrations were estimated by SDS-PAGE and Coomassie blue staining. Concentrated TGP1 was then serially-diluted 2-fold and used in mobility shift experiments. Bound and unbound Y(G4) probe was quantitated by a PhosphorImager (Molecular Dynamics). The Kd for TGP1 G-DNA binding reaction was estimated from the plot of percentage of bound Y(G4) versus protein concentration.
Purified TGP1 from affinity column was displayed on a 10% SDS-PAGE and electrotransferred to PVDF membrane according to the manufacturer's protocol (BioRad). The 80 kDa band was excised and N-terminal sequencing was attempted (ISU Protein Facility), but failed due to N-terminal blockage, thus internal peptide sequencing was then performed. About 5 µg of TGP1 (80 kDa) was eluted from 10% SDS-PAGE gel and acetone precipitated (see the Renaturation section for details). The protein was then subjected to CNBr digestion according to the protocol described by Smith (36). After digestion, peptides were resolved on a 16% SDS-PAGE, and transferred to PVDF membrane. One of the peptides (~35 kDa) yielded a clear peptide sequence: ??GRQSAEG?VG?hYIEV?rFgQYi (`?' indicates that no amino acid had been assigned, while letters in lower case indicate tentative assignments).
Total RNA was isolated from mid-log phase T.thermophila cells using TRIzol reagent according the manufacturer's protocol (Gibco-BRL). First strand cDNA synthesis from total RNA was done by reverse transcription using the M-MLV reverse transcriptase (Gibco-BRL). Based on the internal peptide sequence and with consideration of Tetrahymena genetic codon usage (37), two partially degenerate primers were designed: primer 1 [5'-GGTAGA(C/T)A(A/G)(T/A)(C/G)(T/C)GC(T/C)GAAGG] corresponding to the peptide sequence GRQSAEG, and primer 2 [5'-AATCTGCA(A/G)ACTTC(A/G)AT(A/G)TA] corresponding to the antisense sequence of peptide YIEV?rF. PCR was performed using the first strand cDNAs and primer 1 and primer 2. The PCR product (~55 bp) was cloned into TA cloning vector (Invitrogen), and sequenced (ISU DNA Facility).
Both 5' and 3' rapid amplification of cDNA ends (RACE) were performed according to the recommended protocols (Gibco-BRL). For 3' RACE, a TGP1 gene specific primer (primer 3: 5'-GCTGAAGGAAAAGTCGGTGGTCAC) was designed based on the sequence of the 55 bp PCR product. First strand cDNA synthesis was done using the poly(dT)-anchor oligo (Gibco-BRL). Subsequent PCR was performed over the first strand cDNAs using the anchor and primer 3. PCR products were cloned and sequenced. For 5' RACE, the first strand cDNA synthesis was done using a gene specific primer (primer 4: 5'-TTGACCGAATTTCTTAACTTCG), which was based on the DNA sequence obtained in the 3' RACE. After that, the first strand cDNAs were tailed with poly(dC), and purified. PCR was performed over the tailed cDNAs using a poly(dG)-anchor primer and a nested gene specific primer (primer 5: 5'-TAGTGACCACCGACTTTTCC) as primers. PCR products were cloned and sequenced.
GCG (Genetics Computer Group, Inc.) software was used for sequence analyses. BLAST (38) and FASTA (39) database searches were performed at the National Center for Biotechnology Information (NCBI). Secondary structure analysis was done at the European Molecular Biology Laboratory (EMBL).
To identify proteins responsible for TGP1 activity, UV cross-linking experiments were performed. For these experiments, Tet4(G4) probe and a saturating amount of S100 extract were used. The results are shown in Figure 1. There were no detectable bands in samples which were not exposed to UV-light and which did not contain both the protein extract and probe. In contrast, a band of ~95 kDa was observed in UV-exposed samples containing both the protein extract and DNA probe. These results suggested that the 95 kDa band was the cross-linked complex between TGP1 and DNA probe. Depending on how many Tet4 monomers cross-linked to TGP1, the molecular weight (MW) of the protein can range from 61 to 87 kDa (the Tet4 monomer has a MW ~8.4 kDa). As UV exposure time increased, a minor cross-linked product ~45 kDa was observed. This 45 kDa band could have resulted from a degradation product of the 95 kDa band, or may represent a different protein. Subsequent experiments (affinity purification and renaturation) suggested that this band represents a different G-DNA binding protein.
A number of chromatography columns were tested on analytical scales (1-10 ml) to determine their usefulness in purifying TGP1. Three columns (SP-Sepharose, DE52 and G-DNA affinity) were chosen for the purification. The first column used, SP-Sepharose, is a cation-exchange column (Fig. 2A). As shown in the mobility shift assay, TGP1 was eluted when the salt concentration reached 0.8 M NaCl. This column resulted in removal of >90% of non-TGP1 proteins (data not shown). The high salt concentration (0.8 M NaCl) needed to elute TGP1 activity suggested that TGP1 is a basic protein, or may contain basic domains. After the SP-Sepharose column, the TGP1 fractions were applied to an anion-exchange (DE52) column (Fig. 2B). TGP1 activity was eluted from the column at NaCl concentrations of 0.15-0.5 M.
To confirm that the 80 kDa protein was TGP1, renaturation of proteins from SDS-PAGE was performed. The renatured proteins were used in mobility shift experiments to assay for TGP1 activity. The results (Fig. 4) showed that the protein renatured from the 80 kDa band had TGP1 activity, while no shifted bands were observed for two control proteins (50 and 30 kDa). These results demonstrated that the 80 kDa band contains the TGP1 protein. The size of TGP1 (80 kDa) is consistent with that predicted by the UV cross-linking experiment. Interestingly, a strong band that shifted to a position lower than the TGP1/G-DNA complex was observed for the renatured 40 kDa protein. To test whether there are any interactions between the 80 and 40 kDa proteins, we mixed both renatured proteins and tested them in the mobility shift assay. No additional shifted bands were observed, suggesting no obvious interactions between these two proteins. However, this result does not rule out the possibility of interactions between these two proteins, because the proteins used in the assay were likely to have been only partially renatured.
The DNA binding specificity of TGP1 was studied by competition assays using purified TGP1. Oligonucleotides used in the assays are listed in Table 1. The results are shown in Figure 5. The strongest competitors for TGP1 G-DNA binding are GL(G4), Tet(G4) and Y(G4). These three oligos are in the intermolecular parallel G-DNA (G4) form, suggesting that TGP1 has high binding affinity for such a structure. Non-G-DNAs, including single-stranded DNA (Y) and double-stranded DNAs (Y-cY, Tet4-cTet4), did not compete for TGP1 binding under our assay conditions. Antiparallel G-DNAs (G4' and G'2) showed intermediate competition between those of parallel G-DNAs and non-G-DNAs. G-RNA rTet4(G4) was also tested for competition, and showed much weaker competitive activity than its DNA counterpart Tet4(G4). Hence, our competition results demonstrate that TGP1 is a G-quartet structure-specific binding protein that prefers G-DNAs in the parallel form.
Purified TGP1 was also used to determine the Kd of the TGP1/G-DNA complex. The result of the mobility shift assay using serially-diluted (2-fold) TGP1 is shown in Figure 6A. A supershift band was observed when high concentrations of TGP1 were used. With TGP1 at low concentration, the supershift band disappeared. This supershift could be a dimer of TGP1 binding to a G-DNA substrate. However, because the TGP1 fraction used in the assay contains other proteins (mainly the 40 kDa protein), it is possible that other protein components may be involved in forming such a supershift band. Bands in the mobility shift gel were quantitated and the data were plotted in the graph shown in Figure 6B. The estimated Kd for TGP1/G-DNA binding is ~2.2 × 10-8 M, under our assay conditions [2.5 ng Y(G4) probe, 10 mM Tris-HCl, pH 7.5, 4°C]. This Kd value is in a range similar to those of other G-DNA binding proteins such as G4p1 (17) and G4p2 (18).
To clone the cDNA coding the TGP1 protein, peptide sequencing was performed. No sequence data were obtained from the N-terminal sequencing directly from TGP1, possibly due to N-terminal blockage (data not shown). However, a 35 kDa peptide obtained from CNBr digestion of TGP1 yielded an amino acid sequence (??GRQSAEG?VG?hYIEV?rFgQYi).
TGP1 cDNA was cloned based on a PCR strategy. Two partially degenerate primers were designed based on the internal peptide sequence of TGP1. Using these two primers, a ~55 bp fragment was amplified from the first strand cDNAs by reverse transcriptase PCR (RT-PCR). The DNA sequence between the two primers encodes a peptide KVGGH which is consistent with the peptide sequence obtained from direct peptide sequencing. Based on this DNA sequence, the full length sequence of TGP1 cDNA was obtained using 5' and 3' RACE.
Translation of the longest open reading frame (ORF) in the TGP1 full length cDNA sequence produces a 725 amino acid long protein (the TGP1 cDNA and protein sequences had been deposited into GenBank with accession no. AF006380). The sequence of this predicted protein is considered correct and complete for TGP1 for the following reasons: (i) this predicted protein sequence contains the peptide sequence obtained from the direct peptide sequencing; (ii) the predicted molecular weight for TGP1 from the sequence is 83.2 kDa, which is consistent with the SDS-PAGE result (~80 kDa); (iii) the amino acid composition of this predicted TGP1 protein matches the composition data obtained from acid hydrolysis of TGP1 (data not shown).
Analyses of the predicted TGP1 protein sequence showed that TGP1 has an unbalanced amino acid composition: it is rich in asparagine (13.9%), lysine (9.4%), glutamine (7.2%) and arginine (7.0%), whereas it has only one tryptophan, two cystines and five methionines. The richness of basic amino acids (lysine, arginine and histidine) results in a predicted pI of 10.58, and likely relates to the DNA binding activity of TGP1. Hydrophobicity plot of TGP1 showed that TGP1 contains two extensively hydrophilic and basic regions. One of the two regions is from residue 476 to residue 603. In this 128 amino acid region, there are 109 hydrophilic residues, and among them, 40 asparagines (33%, twice the average percentage in the whole TGP1 sequence), 17 glutamines (14%, twice the average), and more strikingly, 33 basic residues (compared with only five acidic residues). Another hydrophilic and basic region is from residue 251 to 285. The hydrophilic and basic properties of these regions suggest that they may serve as the DNA binding sites (domains) for TGP1. However, none of the known DNA binding motifs were found in the TGP1 sequence (including the hydrophilic and basic regions).
Comparison of the TGP1 protein sequence with available databases using BLAST and FASTA programs showed that TGP1 does not share significant homology or similarity with any other proteins, including many known G-DNA binding proteins, yet is weakly similar to a number of proteins. An entry with one of the best scores (P = 0.02) in the BLAST search is the yeast NUF1 protein, which contains coiled-coil structures and is probably a component of the yeast nuclear skeleton (40). The GAP alignment (gap creation penalty: 7; extension penalty: 2) between TGP1 and NUF1 over the whole sequence showed a 22% identity. In addition to NUF1, several other proteins with coiled-coil structures, including the tropomyosin [alpha] chain of smooth muscle, paramyosin and myosin regulatory light chain from different species, were returned in the FASTA search. However, a secondary structure prediction program (the PHD program in EMBL) did not predict similar structures for TGP1. Of note, the BLAST search also returned two yeast G-quartet binding proteins, G4p1 and G4p2. TGP1 has a region (152-268 amino acids) similar (19% identity, 40% similarity) to a region in G4p1 (46-162 amino acids) (Fig. 7A). TGP1 also has several short sequences similar to sequences in G4p2, and one of the short sequences is within the longer extensively hydrophilic region in TGP1 (Fig. 7B). However, alignments of TGP1 with either G4p1 or G4p2 over the whole sequence resulted in much lower (probably random) similarity.
G-DNAs are novel DNA structures whose proposed biological roles include participation in telomere function, recombination and gene regulation. These putative roles are still unproven and need further exploration. One approach is to identify and study proteins that interact with G-DNA. In this paper, we described the purification, biochemical characterization and molecular cloning of TGP1, the first identified Tetrahymena G-DNA binding protein (32).
This study ultimately identified TGP1 as an 83 kDa protein. UV cross-linking experiments suggested that a protein of mw of ~61-87 kDa was responsible for the TGP1 activity. During a three-column purification process, an ~80 kDa protein was found in the most highly purified fractions. Renaturation experiments demonstrated that the ~80 kDa protein contained TGP1 activity. Furthermore, the cloned TGP1 cDNA encodes a protein with a predicted mw of 83 kDa. In addition to the 83 kDa TGP1 protein, we identified an ~40 kDa protein which copurified with TGP1 during the three-column purification. Interestingly, renaturation experiments showed that this 40 kDa protein possessed a G-DNA binding activity distinct from TGP1. Since the N-terminal peptide sequence of the 40 kDa protein was not found in the TGP1 sequence (data not shown), the 40 kDa protein is unlikely to be a degradation product of TGP1, but rather represents an additional G-DNA binding protein.
The binding specificity of TGP1 was evaluated using the purified protein. The results of these experiments are consistent with a previous study (32). The subtle DNA binding specificity of TGP1 distinguishes it from most known G-quartet binding proteins. First, TGP1 showed very low affinity for non-G-quartet nucleic acids (i.e., single- or double-stranded DNAs), whereas some G-DNA binding proteins are able to bind to these structures. Such proteins include the yeast telomere protein RAP1 which binds to the double-stranded yeast telomeric DNA with much higher affinity (25), and a rat hepatocyte protein qTBP42 which can bind to single-stranded DNAs (41). Second, TGP1 has a weaker binding affinity for G-quartet structures formed by RNA, while some G-quartet binding proteins [e.g., a mouse cytoplasmic exoribonuclease mXRN1p (42), two yeast protein G4p1 and G4p2 (17,18)] have high affinity for G-RNA. Third, TGP1 has a higher affinity for parallel-stranded G-DNA than for the antiparallel form. In contrast, most known G-DNA binding proteins do not distinguish between these two G-DNA forms. The differences in binding specificity suggest that G-DNA/RNA binding proteins can be divided into several subfamilies, each with specificity for different G-quartet forms. TGP1 could therefore belong to a distinct parallel G-DNA-binding protein subfamily.
Sequence analyses showed that none of the known DNA binding motifs were found in TGP1, including a recently identified Myb-like domain characteristic of some telomere binding proteins (43,44). This result is not surprising in light of the fact that TGP1 binds specifically to G-DNA, and has almost no affinity for double- or single-stranded DNAs. It is reasonable to predict that novel DNA binding motifs would exist to account for the specific G-DNA binding activity. Consistent with this hypothesis, two extensively hydrophilic and basic regions, which have no similarity to any known DNA/RNA binding motifs, were identified in TGP1. We propose that the hydrophilic and basic regions comprise novel G-DNA binding domains. We further suggest that these novel domains could be shared by other G-DNA binding proteins to some extent. The weak similarities between sequences of G4p2 and the longer extensively hydrophilic region within TGP1 lend support to this hypothesis. Since no G-DNA specific domains have been identified in any G-DNA binding proteins thus far, further characterization of these putative G-DNA binding domains might provide valuable information on mechanism(s) by which proteins can specifically recognize the G-DNA structure.
Homology/similarity searches revealed that TGP1 is a novel protein with very limited similarity to a number of proteins. The weak similarity between TGP1 and NUF1, the yeast putative nuclear skeleton protein, suggests that TGP1 may be a component of the nuclear skeleton (matrix) in Tetrahymena nuclei. Nuclear matrix, the insoluble non-chromatin scaffold structure of nucleus, has been thought to be involved in many nuclear events such as chromosome organization and gene regulation, through associations with chromosomes (45). The relative abundance of TGP1 [~1.6 × 106 molecules/cell, (32)] is consistent with a role for TGP1 as a structural protein related to the nuclear matrix. One possibility is that TGP1 may bind to telomeres, where G-DNA structures could form (46), linking the telomeres to the nuclear matrix, thus helping to organize the chromosomes. This speculation is supported by data from other species that link telomere proteins to the nuclear matrix. For example, the yeast telomere protein RAP1 is associated with the nuclear matrix (47), and interacts with the SIR4 protein (48), which is weakly similar to the human nuclear matrix proteins lamin A and C. In humans, the telomere protein TRF was found to be a component of the nuclear matrix, and the TRF/telomeric DNA complexes are associated with the nuclear matrix network (49).
The proposed function for TGP1 in the nuclear matrix is speculative and does not exclude other possible telomere-related functions for TGP1. For example, TGP1 may stablize telomere-telomere interactions that had been observed in many species (50). Such interactions have been suggested to have a role in chromosome separation during mitosis (51,52). A mutation in Tetrahymena telomeric DNA repeat (GGGGTT changed to GGGGTTTT) caused delayed and abnormal separation between the sister chromatids during mitosis (52). The mutant chromatids became elongated up to twice the normal length when they finally separated, indicating stronger physical association exists between the mutant chromatids, especially at the telomeres (52). However, little is known about how telomeres interact with each other. One of the possible mechanisms is that G-rich strands or 3' G-overhangs (53-58) of telomeres from different chromosomes can stack together by forming a G-DNA structure (46). If that is the case, protein components such as TGP1 could bind to the G-DNA, and stabilize the interaction between sister chromatids. Any alterations in the telomere sequence may change the G-DNA/protein structure, and thus cause problems in chromosome separation.
Preferred TGP1 binding to parallel G-DNA suggests that TGP1 may also be involved in recombination, since parallel G-DNA structures have long been thought to be involved in recombination processes (7). It was proposed that four G-rich DNA strands from sister chromosomes can bind together to form parallel G-DNA stabilized structures (7). Protein factors that interact with such structures could therefore have roles in recombination. The yeast G-DNA specific nuclease KEM1, which recognizes the parallel G-DNA structure and cuts the single-stranded DNA 5' to the G-quartet domain in vitro, was proposed to function in such a way in meiotic pairing (20). However, in contrast to KEM1, no nuclease activity was observed for TGP1 (data not shown).
All hypotheses regarding the biological functions for TGP1 and G-DNA need to be rigorously tested by further studies such as immunolocalization and gene disruption, and these experiments are in progress. These studies will further define the function(s) of TGP1, and should also contribute to the understanding of biological role(s) of G-DNA in general.
The authors want to thank Scott Schaus for critical reading of the manuscript. This work was funded partly by a research grant from the American Cancer Society to E.H. (NP906). Q.L. was supported in part by fellowships from the MCDB program. This is Journal Paper No. J-17788 of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA. Project No. 3064, and supported by Hatch Act and State of Iowa Funds.
Nucleic Acids Research
Pages
Introduction
Materials And Methods
Oligonucleotide synthesis, purification and labeling
Mobility shift experiments and competition assays
SDS-PAGE and silver staining
UV cross-linking
Tetrahymena cell culture and extract preparation
Purification of TGP1
Protein renaturation
Kd determination of TGP1/G-DNA binding complex
Peptide sequencing
Molecular cloning of TGP1 cDNA
Sequence analyses and database searches
Results
Identification of TGP1 by UV cross-linking
Purification of TGP1
Renatured ~80 kDa protein contains TGP1 activity
TGP1 binds specifically to G-DNA
TGP1/G-DNA complex has a Kd of 2.2 × 10-8 M
Molecular cloning of TGP1 cDNA
TGP1 is a novel protein with unusual sequence features
Discussion
Acknowledgements
References
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 24 Mar 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Etzioni, A. Yafe, S. Khateb, P. Weisman-Shomer, E. Bengal, and M. Fry Homodimeric MyoD Preferentially Binds Tetraplex Structures of Regulatory Sequences of Muscle-specific Genes J. Biol. Chem., July 22, 2005; 280(29): 26805 - 26812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Yafe, S. Etzioni, P. Weisman-Shomer, and M. Fry Formation and properties of hairpin and tetraplex structures of guanine-rich regulatory sequences of muscle-specific genes Nucleic Acids Res., May 20, 2005; 33(9): 2887 - 2900. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Weisman-Shomer, E. Cohen, and M. Fry Distinct domains in the CArG-box binding factor A destabilize tetraplex forms of the fragile X expanded sequence d(CGG)n Nucleic Acids Res., September 1, 2002; 30(17): 3672 - 3681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Lu and E. Henderson Two Tetrahymena G-DNA-binding proteins, TGP1 and TGP3, share novel motifs and may play a role in micronuclear division Nucleic Acids Res., August 1, 2000; 28(15): 2993 - 3001. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kozik, E. M. Bradbury, and A. O. Zalensky Identification and Characterization of a Bovine Sperm Protein That Binds Specifically to Single-Stranded Telomeric Deoxyribonucleic Acid Biol Reprod, February 1, 2000; 62(2): 340 - 346. [Abstract] [Full Text]
|
|
|
|
|
|
M. I. N. Cano, J. J. Blake, E. H. Blackburn, and N. Agabian A Trypanosoma brucei Protein Complex That Binds G-overhangs and Co-purifies with Telomerase Activity J. Biol. Chem., January 4, 2002; 277(2): 896 - 906. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||