|Home | About | Journals | Submit | Contact Us | Français|
The human Ena/Vasp-like (EVL) protein is considered to be a bifunctional protein, involved in both actin remodeling and homologous recombination. In the present study, we found that human EVL forms heat-stable multimers of circular single-stranded DNA (ssDNA) molecules in the presence of a type I topoisomerase in vitro. An electron microscopic analysis revealed that the heat-stable ssDNA multimers formed by EVL and topoisomerase were ssDNA catemers. The ssDNA catenation did not occur when either EVL or topoisomerase was omitted from the reaction mixture. A deletion analysis revealed that the ssDNA catenation completely depended on the annealing activity of EVL. Human EVL was captured from a human cell extract by TOPO IIIα-conjugated beads, and the interaction between EVL and TOPO IIIα was confirmed by a surface plasmon resonance analysis. Purified TOPO IIIα catalyzed the ssDNA catenation with EVL as efficiently as the Escherichia coli topoisomerase I. Since the ssDNA cutting and rejoining reactions, which are the sub-steps of ssDNA catenation, may be an essential process in homologous recombination, EVL and TOPO IIIα may function in the processing of DNA intermediates formed during homologous recombination.
Human Ena/Vasp-like (EVL) is a member of the ENA/VASP family, which is involved in actin-remodeling processes (1). We previously reported (2) that EVL may also function in homologous recombination, because it directly binds to RAD51 and RAD51B, which are essential proteins for meiotic homologous recombination and mitotic recombinational repair of DNA double-strand breaks (3–6). Biochemical studies revealed that EVL forms tetramer-based multimers, and actually stimulates the RAD51-mediated recombinase reactions, such as homologous pairing and strand exchange, in vitro (2,7). In addition to the RAD51-stimulating activity, EVL also possesses ssDNA annealing activity (2), which is considered to be important for the homologous-recombination processes. Therefore, EVL may have dual functions in cytoplasmic actin remodeling and nuclear homologous recombination.
Topoisomerases promote DNA strand cutting and rejoining, and are known to be important in homologous recombination. Escherichia coli RecA, a bacterial homolog of RAD51, forms homologous joint molecules between circular ssDNA and closed circular dsDNA by its recombinase activity (8). Escherichia coli topoisomerase I (Topo I) reportedly converts the homologous joint molecules formed by RecA into hemicatemers (8). Escherichia coli topoisomerase III efficiently catenates closed circular dsDNAs in the presence of RecQ helicase (9,10), which is suggested to function in homologous recombination. In humans, TOPO IIIα forms a complex with BLM and BLAP75 (11,12), and the complex is reportedly involved in the dissolution of the Holliday junction (13–18), which is a DNA intermediate formed in the late stage of homologous recombination. These facts suggest that the DNA cutting and rejoining activities of topoisomerases play important roles in homologous recombination.
In the present study, we unexpectedly found that EVL, with either E. coli Topo I or human TOPO IIIα, catalyzed ssDNA catenation. The omission of either EVL or topoisomerase quenched the ssDNA-catenating reaction, indicating that both proteins are essential for this reaction. A deletion analysis revealed that the EVL C-terminal domain, which possesses the annealing activity, is responsible for the ssDNA catenation. We also found that EVL physically interacted with human TOPO IIIα in a human cell extract and in vitro. These new findings suggest that EVL may function with a type I topoisomerase, such as TOPO IIIα, by catalyzing ssDNA cutting and re-joining reactions in the conversion of DNA intermediates during homologous recombination.
Human EVL, EVL(1–221) and EVL(222–418) were prepared by the methods described earlier (2,7). In these methods, human EVL, EVL(1–221) and EVL(222–418) were expressed as His6-tagged proteins, and the His6 tag was removed by a thrombin treatment during the purification procedure. Human RPA was produced in E. coli cells and was prepared according to the published protocol (19).
The DNA fragment encoding human TOPO IIIα was isolated from a human cDNA pool (purchased from Clontech Laboratories, Mountain View, CA, USA) by the polymerase chain reaction. The TOPO IIIα DNA fragment was cloned in the NdeI site of the pET15b vector (Novagen, Darmstadt, Germany). In this construct, the His6 tag sequence was fused to the N terminus of the protein. Human TOPO IIIα was produced in the E. coli BL21(DE3) codon plus-RP strain (Stratagene, La Jolla, CA, USA) and was purified by the following procedure. The cells producing His6-tagged TOPO IIIα were resuspended in a 50mM Tris–HCl buffer (pH 7.5), containing 1M NaCl, 5mM 2-mercaptoethanol and 10% glycerol, and were disrupted by sonication. The cell debris was removed by centrifugation for 20min at 30000g, and the lysate was mixed gently by the batch method with Ni-NTA agarose beads (3ml, Qiagen, Hilden, Germany) at 4°C for 1h. The His6-tagged TOPO IIIα-bound beads were washed with 40ml of 20mM potassium phosphate buffer (pH 7.4), containing 500mM NaCl, 5mM 2-mercaptoethanol, 40mM imidazole and 10% glycerol, and then were washed again with 40ml of 20mM potassium phosphate buffer (pH 7.4), containing 500mM NaCl, 5mM 2-mercaptoethanol, 30mM imidazole and 10% glycerol. The beads were then packed into an Econo-column (Bio-Rad Laboratories, Hercules, CA, USA) and were washed with 90ml of 20mM potassium phosphate buffer (pH 7.4), containing 500mM NaCl, 5mM 2-mercaptoethanol, 30mM imidazole and 10% glycerol. His6-tagged TOPO IIIα was eluted with a linear gradient of 30–300mM imidazole (13-column volumes), in 20mM potassium phosphate (pH 7.4), 100mM NaCl, 5mM 2-mercaptoethanol and 10% glycerol. Peak fractions containing His6-tagged TOPO IIIα were diluted 5-fold with a 20mM potassium phosphate buffer (pH 7.4), containing 5mM 2-mercaptoethanol and 10% glycerol and were mixed gently by the batch method with Hydroxyapatite resin (5ml, Bio-Rad Laboratories) at 4°C for 1h. The unbound fraction was then dialyzed against a 20mM potassium phosphate buffer (pH 7.4), containing 100mM NaCl, 5mM 2-mercaptoethanol and 10% glycerol. After the dialysis, the sample was loaded on an SP Sepharose column (1ml, GE Healthcare Biosciences, Uppsala, Sweden), which was equilibrated with 20ml of 20mM potassium phosphate buffer (pH 7.4), containing 100mM NaCl, 5mM 2-mercaptoethanol and 10% glycerol. The resin was washed with 20ml of 20mM potassium phosphate buffer (pH 7.4), containing 225mM NaCl, 5mM 2-mercaptoethanol and 10% glycerol, and the His6-tagged TOPO IIIα was then eluted with a linear gradient of 225–600mM NaCl (24-column volumes). The purified His6-tagged TOPO IIIα was dialyzed against 20mM HEPES–NaOH buffer (pH 7.5), containing 100mM NaCl, 5mM 2-mercaptoethanol and 20% glycerol, and was stored at −80°C. The concentration of the purified His6-tagged TOPO IIIα was determined by the Bradford method (20), using bovine serum albumin as the standard.
The ϕX174 circular ssDNA (20µM) was mixed with EVL in 10µl of a standard reaction solution, containing 36mM HEPES–NaOH (pH 7.5), 1mM dithiothreitol, 4mM 2-mercaptoethanol, 80mM NaCl, 1mM MgCl2, 24% glycerol and 0.1mg/ml bovine serum albumin. The reaction mixtures were incubated at 37°C for 15min, and were then analyzed by 0.8% agarose gel electrophoresis in 1× TAE buffer (40mM Tris–acetate and 1mM EDTA) at 3.3V/cm for 2h. The bands were visualized by ethidium bromide staining.
The ϕX174 circular ssDNA (20µM) was incubated with EVL and Topo I (New England Biolabs, Ipswich, MA, USA), in a reaction buffer containing 24mM HEPES–NaOH (pH 7.5), 1mM MgCl2, 1mM Tris–HCl (pH 7.5), 1.1mM dithiothreitol, 1mM 2-mercaptoethanol, 20mM NaCl, 5mM KCl, 3.5mM ammonium sulfate, 0.01mM EDTA, 11% glycerol and 0.1mg/ml bovine serum albumin, at 37°C for 1h. For the experiments with TOPO IIIα, the reactions were conducted in a buffer containing 30mM HEPES–NaOH (pH 7.5), 1mM MgCl2, 1mM dithiothreitol, 2.5mM 2-mercaptoethanol, 70mM NaCl, 14% glycerol and 0.1mg/ml bovine serum albumin. The samples were then deproteinized by a treatment with 0.2% SDS and 1.3mg/ml proteinase K at 37°C for 15min. The products were then incubated at 100°C for 5min and were chilled on ice for 5min. The products were separated by 1% agarose gel electrophoresis, and the bands were visualized by SYBR Gold (Invitrogen, Carlsbad, CA, USA) staining.
The ssDNA-annealing assay was performed as described earlier (2). Briefly, the ssDNA oligonucleotide 49-mer (0.2µM) was incubated with the indicated amounts of EVL or the EVL fragments at 30°C for 5min, in 9µl of reaction buffer, containing 28mM HEPES–NaOH (pH 7.5), 50mM NaCl, 2mM 2-mercaptoethanol, 12% glycerol, 0.1mM MgCl2, 1mM DTT and 0.1mg/ml bovine serum albumin. The reactions were initiated by the addition of 0.2µM antisense 32P-labeled 49-mer oligonucleotide. At the times indicated, the reactions were quenched with an excess of the unlabeled 49-mer oligonucleotide. The DNA substrates and products were deproteinized by a treatment with 0.2% SDS and 1.5mg/ml proteinase K at 30°C for 10min. The products were fractionated by 10% PAGE in 0.5× TBE. The gels were dried, exposed to an imaging plate and visualized using an FLA-7000 imaging analyzer (Fujifilm, Tokyo, Japan).
The ssDNA catemers formed by EVL and Topo I were extracted by phenol/chloroform and precipitated with ethanol. The ssDNA catemers were then coated with RecA in the absence of ATP and were stained on a copper-plated carbon grid with 2% uranyl acetate. The samples were visualized by tungsten rotary shadowing, and were observed with a JEOL JEM 2000FX electron microscope (JEOL, Akishima, Tokyo, Japan).
Purified human TOPO IIIα was covalently conjugated to Affi-Gel 10 beads (100µl, Bio-Rad), according to the manufacturer's instructions. To block the remaining active ester sites, ethanolamine (pH 8.0) was added to a final concentration of 100mM, and the resin was incubated at 4°C overnight. The unbound proteins were removed by washing the Affi-Gel 10-TOPO IIIα beads three times with 500µl of binding buffer, which contained 20mM HEPES–NaOH (pH 7.5), 300mM NaCl, 5mM 2-mercaptoethanol, 20% glycerol and 0.05% Triton X-100. After washing the resin, the Affi-Gel 10-protein matrices were adjusted to 50% slurries, and were stored at 4°C. The control beads were made by the same method, except that the TOPO IIIα was replaced by the TOPO IIIα storage buffer. For the EVL-binding assay, the TOPO IIIα-beads were incubated with an MCF7 whole-cell extract (1.8mg of protein), and the beads were washed three times with 200µl of washing buffer, containing 50mM Tris–HCl (pH 7.5), 100mM NaCl, 5mM EDTA, 0.5% NP-40, 1mM phenylmethylsulfonyl fluoride and protease inhibitor cocktail (Nacalai Tesque, Kyoto, Japan). The proteins that copelleted with the TOPO IIIα beads were fractionated by 7.5% SDS–PAGE. The EVL protein was detected with the EVL-specific rabbit polyclonal antibodies.
The surface plasmon resonance (SPR) signals were measured with a Biacore X100 instrument (GE Healthcare Biosciences, Uppsala, Sweden). Flow cells were maintained at 25°C during the measurement, and the instrument was operated at the mid-flow rate (~30µl/min). Purified EVL was conjugated to the activated surface of the CM5 sensor chip (GE Healthcare Biosciences, Uppsala, Sweden), using the standard amine coupling conditions recommended by the manufacturer. The level of the conjugated EVL protein was 5600 resonance units. The SPR signals of the flow cell containing a sensor chip without the proteins were subtracted from those of the SPR signals of the flow cell containing the EVL-conjugated sensor chip. The running buffer was 20mM HEPES–NaOH (pH 7.5), 200mM NaCl, 2.5% glycerol, 1mM dithiothreitol and 0.1% Tween-20. For the binding assay, 0.1µM TOPO IIIα, RPA, Topo I or bovine serum albumin was injected for 2min.
As described earlier (2), EVL bound to ϕX174 circular ssDNA (5386 bases) and formed large complexes (Figure 1A). We found that ssDNA multimers were formed within this EVL–ssDNA complex, after the EVL protein was removed from the complex by an SDS and proteinase K treatment (Figure 1B). Since we previously reported that EVL promotes the annealing of complementary ssDNAs (2), we suspected that the ssDNA multimers may be annealed products between short, complementary sequences within the ϕX174 circular ssDNA. As anticipated, the ssDNA multimers were dissociated into monomers, when the samples were incubated at 100°C for 5min (Figure 1B, lane 7). In addition, the formation of the ssDNA multimers by EVL was completely suppressed by an ssDNA-binding protein, RPA (Figure 1C), which significantly inhibited the ssDNA annealing by EVL (Figure 1D). Therefore, we concluded that EVL promotes annealing between short, complementary sequences within the ϕX174 circular ssDNA and forms ssDNA multimers.
The ssDNA multimers formed by EVL may be annealed products, because the multimers were dissociated by heating (Figure 1B). Interestingly, we found that the ssDNA multimers formed by EVL in the presence of a type I topoisomerase, E. coli Topo I, were not resolved after incubation at 100°C for 5min (Figure 2A, lane 6). These heat-stable ssDNA multimers may be ssDNA catemers. To assess whether ssDNA catemers were formed, we visualized the heat-stable ssDNA multimers by electron microscopy. To do so, the ssDNA multimers were purified, and were then coated with RecA to visualize the ssDNA. As anticipated, circular ssDNA catemers, containing two or three ssDNA molecules, were observed (Figure 2B). We counted a total of 102 molecules from the DNA samples and found that about 72.5, 18.6, 6.9 and 2% of ssDNA molecules were single circles, two ssDNA catemers, three ssDNA catemers and four or five ssDNA catemers, respectively. Therefore, we concluded that the heat-stable ssDNA multimers are ssDNA catemers.
The ssDNA-catenating reaction did not occur when either EVL or Topo I was omitted from the reaction mixture (Figure 3A, lanes 2 and 7), indicating that both EVL and Topo I are essential for the ssDNA-catenating reaction. In addition, the formation of EVL-dependent DNA catemers was not detected when circular dsDNA was used as a substrate (Figure 3B). Protein titration experiments revealed that a very small amount of Topo I (0.04nM) was sufficient to form the ssDNA catemers (Figure 3A, lane 3). This Topo I concentration (0.04nM) was far below the amount required for inducing a topological change in supercoiled DNA, because 0.4nM of Topo I is not sufficient to relax supercoiled DNA (Figure 3B, lanes 3 and 5). Therefore, these results suggested that Topo I catalytically functions in the EVL-dependent ssDNA-catenating reaction.
We next tested whether the ssDNA annealing by EVL plays an essential role in the ssDNA-catenating reaction with Topo I. To do so, we prepared two EVL fragments, EVL(1–221) and EVL(222–418), which contained amino acid residues 1–221 and 222–418, respectively (7). As shown in Figure 4A and B, EVL(222–418) promoted annealing to a similar extent as full-length EVL, but EVL(1–221) did not. We then tested whether EVL(1–221) and EVL(222–418) support the ssDNA-catenating reaction. As anticipated, EVL(222–418), which possesses the annealing activity, promoted the ssDNA-catenating reaction, with indistinguishable efficiency from the full-length EVL (Figure 4C, lane 7). In contrast, EVL(1–221) did not promote the ssDNA-catenating reaction (Figure 4C, lane 6). These results indicated that the EVL annealing activity is essential for the ssDNA-catenating reaction with Topo I.
We then determined whether human topoisomerase can also promote the EVL-dependent ssDNA-catenating reaction. A eukaryotic type I topoisomerase, human TOPO IIIα, reportedly functions in homologous recombination (11–18). Therefore, we purified human TOPO IIIα as a recombinant protein (Figure 5A). The purified TOPO IIIα was chemically conjugated to Affi-Gel 10, and pull-down assays were performed with MCF7 cell extracts. As shown in Figure 5B (lane 3), the endogenous EVL protein in the MCF7 cells was clearly detected in the TOPO IIIα-bound fraction. A SPR analysis also revealed that purified TOPO IIIα efficiently bound to EVL (Figure 5C). We next performed the EVL-dependent ssDNA-catenating assay with TOPO IIIα. As shown in Figure 5D, TOPO IIIα significantly stimulated the ssDNA-catenating reaction in the presence of EVL. Like E. coli Topo I (Figure 3A), the ssDNA catemers were formed with a very small amount of TOPO IIIα (0.01nM, Figure 5D, lane 4), suggesting its catalytic function in the reaction. The ssDNA catemers were not formed when either EVL or TOPO IIIα was omitted from the reaction mixture (Figure 5D, lanes 2 and 8). In addition, we found that the ssDNA-catenating reaction mediated by EVL and TOPO IIIα was significantly inhibited by RPA (Figure 6A), which suppresses the EVL-dependent ssDNA annealing (Figure 1C and D). The EVL-TOPO IIIα interaction was still observed in the presence of RPA (Figure 6B). These results suggested that RPA inhibits the ssDNA-catenating reaction by inhibiting the EVL-mediated ssDNA annealing.
Intriguingly, we also found that EVL bound to E. coli Topo I, with reduced affinity as compared to human TOPO IIIα in vitro (Figure 5C). Although this EVL-Topo I combination does not exist in the natural context, it suggested that EVL binding to the Topo I region, which may be evolutionarily conserved with TOPO IIIα, may enhance the ssDNA-catenating reaction in vitro.
In the present study, we found that human EVL and a type I topoisomerase promote the catenation of ssDNA molecules. It has been reported that RecA, a bacterial recombinase, promotes hemicatenation between ssDNA and dsDNA in the presence of Topo I (8); however, no report for the ssDNA-catenating reaction has been published thus far. We also confirmed that both prokaryotic and eukaryotic type I topoisomerases are functional for the ssDNA catenation with EVL.
Type I topoisomerase promotes cutting and rejoining reactions on a strand within dsDNA. In the present study, we showed that the annealing activity of EVL plays an essential role in the ssDNA-catenating reaction with a type I topoisomerase. Therefore, we have proposed an annealing-mediated model for the ssDNA-catenating reaction (Figure 7). In this model, EVL first binds to ssDNA and forms large complexes, each containing multiple ssDNA molecules (Step 1). Since human EVL contains a tetramerization domain at its C-terminus (21), the formation of multiple ssDNA complexes may be mediated through its tetramerization activity. The short complementary sequences of the ssDNA may be annealed within the EVL-multiple ssDNA complexes (Step 2). A type I topoisomerase, which may be recruited on the ssDNA annealed sites through binding to EVL, then cuts an ssDNA strand within the annealed region (Step 3). This nicked site rotates and rejoins (Step 4).
The functional significance of the ssDNA catenation by EVL and topoisomerases remains to be elucidated; however, it may be involved in homologous recombination because EVL directly interacts with the eukaryotic RecA homologs, RAD51 and RAD51B (2). Our previous analyses suggested that EVL may function as a RAD51 mediator, which stimulates the RAD51-mediated homologous-pairing and strand-exchange reactions (2,7). The ssDNA-catenating activity, comprising the ssDNA cutting and rejoining reactions, may be utilized to process the DNA intermediates formed by the homologous-pairing and strand-exchange reactions by RAD51. In the homologous-recombination process, it has been proposed that a DNA intermediate containing a D-loop, in which the ssDNA has invaded a homologous region of the intact dsDNA, is first formed by the RAD51-mediated homologous pairing. After this homologous-pairing step, the invading strand primes repair synthesis of the DNA strands, and the D-loop structure is converted to a four-way junction, the Holliday-junction intermediate, which moves along the DNA to expand the newly paired heteroduplex region. The ssDNA cutting and rejoining activities of EVL and topoisomerase may be involved in the conversion process from the D-loop to Holliday junction intermediates.
Alternatively, the ssDNA cutting and rejoining activities may function in the resolution of the Holliday junction. The Holliday-junction intermediate must be resolved into two parental DNA molecules by the ssDNA-nicking activity in the late stage of the homologous-recombination pathway. Interestingly, human TOPO IIIα, which was found to promote the ssDNA-catenating reaction with EVL, is reportedly involved in Holliday junction dissolution (13–16). In addition, EVL is known to bind RAD51B, which preferentially binds to the Holliday junction (2,22). Therefore, EVL may also be involved in the late stage of the HRR pathway, together with TOPO IIIα. Further analyses are required to clarify these issues.
Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), and the Japanese Society for the Promotion of Science (JSPS) of Japan. Funding for open access charge: Waseda University.
Conflict of interest statement. None declared.
H. K. is a research fellow in the Waseda Research Institute for Science and Engineering.