Stimulation of DNA Strand Exchange by the Human TBPIP/Hop2-Mnd1 Complex*

In Saccharomyces cerevisiae, the Hop2 protein forms a complex with the Mnd1 protein and is required for the alignment of homologous chromosomes during meiosis, probably through extensive homology matching between them. The Rad51 and Dmc1 proteins, the eukaryotic RecA orthologs, promote strand exchange and may function in the extensive matching of homology within paired DNA molecules. In the present study, we purified the human TBPIP/Hop2-Mnd1 complex and found that it significantly stimulates the Dmc1- and Rad51-mediated strand exchange. The human Hop2-Mnd1 complex preferentially binds to a three-stranded DNA branch, which mimics the strand-exchange intermediate. These findings are consistent with genetic data, which showed that the Hop2 and Mnd1 proteins are required for homology matching between homologous chromosomes. Therefore, the human TBPIP/Hop2-Mnd1 complex may ensure proper pairing between homologous chromosomes through its stimulation of strand exchange during meiosis.

In Saccharomyces cerevisiae, the Hop2 protein forms a complex with the Mnd1 protein and is required for the alignment of homologous chromosomes during meiosis, probably through extensive homology matching between them. The Rad51 and Dmc1 proteins, the eukaryotic RecA orthologs, promote strand exchange and may function in the extensive matching of homology within paired DNA molecules. In the present study, we purified the human TBPIP/ Hop2-Mnd1 complex and found that it significantly stimulates the Dmc1-and Rad51-mediated strand exchange. The human Hop2-Mnd1 complex preferentially binds to a three-stranded DNA branch, which mimics the strand-exchange intermediate. These findings are consistent with genetic data, which showed that the Hop2 and Mnd1 proteins are required for homology matching between homologous chromosomes. Therefore, the human TBPIP/ Hop2-Mnd1 complex may ensure proper pairing between homologous chromosomes through its stimulation of strand exchange during meiosis.
In meiosis, a high level of homologous recombination occurs only between homologous chromosomes but not between sister chromatids. This meiotic homologous recombination is initiated by the formation of a double strand break (DSB), 3 which is introduced by the SPO11 protein (1)(2)(3). On the other hand, in mitosis, homologous recombination functions to repair DSBs, which are introduced by DNA damaging agents, such as ionizing radiation, DNA cross-linking reagents, oxidative stress, and replication errors. This mitotic homologous recombination repair mainly occurs between sister chromatids or between the abundant intra-and inter-chromosomal homologous repeat sequences.
After the DSB formation, a single-stranded DNA (ssDNA) tail derived from a DSB site invades the homologous double-stranded DNA (dsDNA). This strand-invasion step, called homologous pairing, primes heteroduplex formation, in which new Watson-Crick base pairs are formed between the invading strand and its complementary strand of parental dsDNA. Then, the heteroduplex region is expanded by the subsequent strand-exchange step. This strand-exchange step may be important for the extensive matching of homology between paired chromosomes to ensure that homologous recombination occurs between the proper partners. The Escherichia coli RecA protein is known to catalyze the homologous pairing and strand-exchange steps (4 -7), and two RecA homologues, the Rad51 and Dmc1 proteins, have been identified in eukaryotes (8 -11). The Rad51 and Dmc1 proteins have been shown to catalyze strand exchange in vitro, but their strandexchange activities are low as compared with that of RecA (12)(13)(14).
Genetic studies with Saccharomyces cerevisiae, Schizosaccharomyces pombe, and Arabidopsis thaliana have identified the HOP2, meu13 ϩ , and AHP2 genes, respectively, as being essential for meiotic homologous recombination (15)(16)(17). The HOP2, meu13 ϩ , and AHP2 genes are orthologs, and the human and mouse TBPIP proteins have been identified as mammalian orthologs of Hop2 (18,19). The hop2 deletion mutant in S. cerevisiae fails to sporulate, due to a uniform arrest at the pachytene stage of meiosis I (15). The TBPIP/Hop2 knock-out mice also display meiotic cell cycle arrest, due to the failure of DSB repair (20). Interestingly, in the hop2 deletion mutant cells, most of the chromosomes are engaged in synapsis with nonhomologous partners but not with homologous chromosomes (15), suggesting that the Hop2 protein functions to align homologous chromosomes during meiosis. The S. cerevisiae Hop2 protein reportedly forms a complex with the Mnd1 protein, which has been identified as a multicopy suppressor of a temperature-sensitive hop2 mutant allele (21). The S. cerevisiae MND1 gene and its S. pombe ortholog, mcp7 ϩ , are also required for meiotic recombination (21)(22)(23). The mnd1-null mutant exhibits a strikingly similar phenotype to that of the hop2-null mutant (21,22). These results suggest that the Hop2 and Mnd1 proteins function as a complex to promote meiotic chromosome pairing. Biochemical analyses also revealed that the yeast and mouse Hop2 (TBPIP/Hop2) and Mnd1 proteins form a complex that can stimulate the strand-invasion step promoted by their cognate Dmc1 proteins (24,25).
In the present study, we found that the purified human TBPIP/Hop2-Mnd1 complex (hTBPIP/Hop2-hMnd1) significantly stimulates the Dmc1-or Rad51-mediated strand exchange, which may be required for the extensive matching of homology within homologous chromosomes. Therefore, hTBPIP/Hop2-hMnd1 may function in the strand-exchange step, which occurs just after homologous pairing, to ensure that the correct partner is involved in meiotic homologous recombination.

Overexpression and Purification of the Human TBPIP/Hop2-Mnd1
Complex-The human TBPIP/Hop2 gene was previously cloned as a homologue of the mouse TBPIP gene (18,19). The human Mnd1 gene (GenBank TM accession number NM_032117) was cloned from a human testis cDNA pool (Clontech). The human TBPIP/Hop2 and Mnd1 genes were ligated into the NdeI-BamHI sites of the pET15b and pET11a vectors, respectively. Then, the pET11a vector containing the Mnd1 gene was digested with BglII and BamHI, and the resulting Mnd1-containing fragment was ligated into the BamHI site of the TBPIP/Hop2containing pET15b vector. In this construct, the human Mnd1 protein (hMnd1) and the N-terminally His 6 -tagged human TBPIP/Hop2 protein (hTBPIP/Hop2) were coexpressed in the E. coli strain BL21-CodonPlus(DE3)-RIL (Stratagene). The cells expressing both hMnd1 and hTBPIP/Hop2 were grown at 37°C in 10 liters of LB medium, containing 100 g/ml ampicillin and 34 g/ml chloramphenicol. At the logarithmic phase of growth (A 600 ϭ 0.6), both hTBPIP/Hop2 and hMnd1 were produced in the presence of 500 M isopropyl 1-thio-␤-Dgalactopyranoside (IPTG). Then, the cells were harvested after a 4 h incubation at 37°C and were disrupted by sonication in buffer A (50 mM Tris-HCl buffer (pH 7.5) containing 0.5 M NaCl, 10% glycerol, 5 mM 2-mercaptoethanol, and protease inhibitors (Complete EDTA-free; Roche Diagnostics)) on ice. The cell lysate was centrifuged at 27,700 ϫ g for 20 min, and then the proteins in the supernatants were precipitated with 0.24 g/ml ammonium sulfate (40% saturation). The precipitate was removed by centrifugation at 27,700 ϫ g for 20 min, and the proteins in the supernatant were further precipitated by the addition of 0.13 g/ml ammonium sulfate (60% saturation). The precipitate was dissolved in 30 ml of buffer A and was gently mixed with 4 ml of Ni-NTA-agarose beads (Qiagen) for 1 h at 4°C. The Ni-NTA-agarose beads bound to the His 6tagged hTBPIP/Hop2 protein were packed into an Econo-column (Bio-Rad) and were washed with 20-column volumes of buffer A containing 7 mM imidazole, at a flow rate of about 0.3 ml/min. The His 6 -tagged hTBPIP/Hop2 protein complexed with the hMnd1 protein was eluted in a 15-column volume linear gradient from 7 to 400 mM imidazole in buffer A. The His 6 tag was uncoupled from the hTBPIP/Hop2 portion by a digestion with 1 unit of thrombin protease (Amersham Biosciences) per mg of the hTBPIP/Hop2-hMnd1 proteins. After the thrombin treatment, the hTBPIP/Hop2-hMnd1 proteins were immediately dialyzed against buffer B (20 mM Tris-HCl buffer (pH 7.5) containing 0.2 M NaCl, 5 mM 2-mercaptoethanol, 0.25 mM EDTA, and 10% glycerol) for more than 12 h at 4°C. The hTBPIP/Hop2-hMnd1 proteins were loaded onto a 6-ml heparin-Sepharose (Amersham Biosciences) column previously equilibrated with buffer B. The column was washed with 10-column volumes of buffer B, and the proteins were eluted with a 10-column volume linear gradient from 0.2 to 1.2 M NaCl in buffer B. The peak fractions of the hTBPIP/Hop2-hMnd1 proteins were dialyzed against buffer C (20 mM Tris-HCl buffer (pH 7.5) containing 0.2 M NaCl, 5 mM 2-mercaptoethanol, and 0.25 mM EDTA). The hTBPIP/Hop2-hMnd1 proteins eluted from the heparin-Sepharose column were loaded onto a Hiload 26/60 Superdex 200 column (Amersham Biosciences) previously equilibrated with buffer C. The purified hTBPIP/Hop2-hMnd1 proteins were eluted from the Superdex 200 column with buffer C, at a flow rate of about 0.8 ml/min. Protein concentrations were determined using a Bio-Rad protein assay kit with bovine serum albumin as the standard.
The Human Replication Protein A (RPA), Dmc1, and Rad51 Proteins-The human replication protein A (RPA) was purified as described previously (26). The human Dmc1 and Rad51 proteins (hDmc1 and hRad51, respectively) were expressed in the E. coli strain BL21-CodonPlus(DE3)-RIL (Stratagene) and were purified to homogeneity as described previously (27)(28)(29). The His 6 tags of these proteins were removed by thrombin protease digestion during the purification process. Protein concentrations were determined using a Bio-Rad protein assay kit, with bovine serum albumin as the standard.
DNA Substrates-In the D-loop formation assay, alkaline treatment of the cells harboring the plasmid DNA was avoided, to prevent the dsDNA substrates from undergoing irreversible denaturation. Instead, the cells were gently lysed using Sarkosyl, as described previously (29). The pGsat4 DNA was created by inserting a 198-base pair fragment of the human ␣-satellite sequence into the pGEM-T easy vector (Promega) (30). For the ssDNA substrate used in the D-loop assay, the following high performance liquid chromatography (HPLC)-purified oligonucleotide was purchased from Roche Diagnostics. SAT-1 (50-mer, 5Ј-ATT TCA TGC TAG ACA GAA GAA TTC TCA GTA ACT TCT TTG TGC TGT GTG TA-3Ј). The 5Ј ends of the oligonucleotides were labeled with T4 polynucleotide kinase (New England Biolabs) in the presence of [␥-32 P]ATP at 37°C for 30 min. DNA concentrations are expressed in moles of nucleotides.
The DNA substrates used in the competitive DNA binding assay were identical to those in the synthetic Holliday junction (I) designed by Iwasaki et al. (31). The DNA substrates containing the three-stranded DNA branch, the Y-form DNA, and the single-stranded oligonucleotide were produced with three 49-mer single-stranded oligonucleotides, 1, 2, and 3, with the sequences 5Ј-ATCGA TGTCT CTAGA CAGCT GCTCA GGATT GATCT GTAAT GGCCT GGGA-3Ј, 5Ј-GTCCC AGGCC ATTAC AGATC AATCC TGAGC ATGTT TACCA AGCGC ATTG-3Ј, and 5Ј-TGATC ACTTG CTAGC GTCGC AATCC TGAGC AGCTG TCTAG AGACA TCGA-3Ј, respectively. The three-stranded DNA branch substrate contained a 12-bp mobile region at the center, and the Y-form DNA substrate was composed of a 30-bp duplex region and two single-stranded tails (18 and 19 bp). All of the oligonucleotides were purified by HPLC, and the DNA concentrations are expressed in moles of nucleotides.
Competitive DNA Binding Assays with the Three-stranded DNA Branch, the Y-form DNA, and the Single-stranded Oligonucleotides-The 32 P-labeled oligonucleotide 1 (0.16 M) was mixed with the unlabeled oligonucleotides 2 (0.16 M) and 3 (0.16 M), and the 32 P-labeled DNA substrates containing the three-stranded DNA branch, the Y-form DNA, and the single-stranded oligonucleotide were produced by annealing. The DNA sample used in this assay contained the threestranded DNA branch, the Y-form DNA, and the single-stranded oligonucleotide at 1:2.2:2.9 ratios. The indicated amounts of hTBPIP/Hop2-hMnd1 were incubated with the DNA substrates (total 64 nM) at 37°C for 10 min in 10 l of 20 mM Tris-HCl buffer (pH 8.0), containing 1 mM ATP, 1 mM MgCl 2 , 100 g/ml bovine serum albumin, 2 mM creatine phosphate, and 75 g/ml creatine kinase. The products were analyzed by 10% polyacrylamide gel electrophoresis in 1 ϫ TBE buffer (90 mM Tris borate and 2 mM EDTA) and were visualized by autoradiography of the dried gel.
Analytical Ultracentrifugation-Sedimentation equilibrium experiments were performed in a Beckman ProteomeLab XL-1 instrument. The hTBPIP/Hop2-hMnd1 proteins (0.5 mg/ml) were spun in a Beckman An-50Ti rotor with a 6-sector centerpiece. Equilibrium distributions were analyzed after 20 h of centrifugation at 9,000 rpm and 20°C. For the molecular weight analysis, a partial specific volume of 0.675 cm 3 /g and a solution density of 1 g/cm 3 were used.
The ATPase Assay-The ATPase activities of the proteins were analyzed by the release of 32   phosphate, and 30 g/ml creatine kinase), at 37°C for 3 min, and the supercoiled pGsat4 (3,218 bp) (181 M) DNA, which had been incubated with hTBPIP/Hop2-hMnd1 (0.2 M) at 37°C for 3 min, was added to the reaction mixture. After incubations at 37°C for the indicated times, 10-l aliquots of the reaction mixture were withdrawn, and the reactions were terminated by the addition of 0.5% SDS and 883 g/ml proteinase K (Roche Diagnostics), followed by an incubation at 37°C for 15 min. After 6-fold loading dye was added, and the products were separated by 1% agarose gel electrophoresis in 0.5 ϫ TBE buffer (45 mM Tris borate and 1 mM EDTA) at 3.3 V/cm for 2 h. The bands were visualized by autoradiography of the dried gel.
The Strand-exchange Assay-The hDmc1 (15 M) or hRad51 (10 M) protein was incubated at 37°C for 5 min with 30 M X174 circular ssDNA in the presence of the indicated amounts of hTBPIP/ Hop2-hMnd1, in 10 l of 20 mM Tris-HCl buffer (pH 8.0), containing 2 mM ATP, 1 mM MgCl 2 , 100 g/ml bovine serum albumin, 2 mM creatine phosphate, and 75 g/ml creatine kinase. After this incubation, 2 M RPA and 0.2 M KCl were added to the reaction mixture, which was incubated at 37°C for 5 min. Then, the reactions were initiated by the addition of 30 M X174 linear dsDNA and were continued for 1 h. To eliminate the salt effect from the storage buffer for the proteins, in the protein titration experiments, the salt concentrations (NaCl and KCl) of the reaction mixtures were adjusted to be the same in all reactions. For the time course experiments, the concentrations of hTBPIP/Hop2-hMnd1 were 0.5 M for hDmc1 and 2 M for hRad51. The reactions were stopped by the addition of 0.5% SDS and 1.82 mg/ml proteinase K (Roche Diagnostics) and were further incubated at 37°C for 30 min. After 6-fold loading dye was added, the deproteinized reaction products were separated by 1% agarose gel electrophoresis in 1 ϫ TAE buffer at 3.3 V/cm for 3 h. The products were visualized by ethidium bromide, which was added to the gel and the running buffer (0.5 g/ml).
The hTBPIP/Hop2 and hMnd1 proteins co-purified through all of the purification steps in about 1:1 stoichiometry, and no fraction containing hTBPIP/Hop2 or hMnd1 alone was found in any chromatographic step, indicating that these proteins formed a stable complex. Purified hTBPIP/Hop2-hMnd1 bound to both ssDNA and dsDNA (data not shown). A competitive DNA binding experiment with circular ssDNA and linearized dsDNA showed that hTBPIP/Hop2-hMnd1 preferentially bound to dsDNA (Fig. 1B). This DNA binding property of hTBPIP/Hop2-hMnd1 is the same as that of the yeast (S. cerevisiae) Hop2-Mnd1 complex containing the C-terminally His 6 -tagged Mnd1 protein (24). However, the mouse TBPIP/Hop2-Mnd1 complex containing the C-terminally His 6 -tagged TBPIP/Hop2 protein reportedly binds to ssDNA and dsDNA without any preference (25), suggesting that the DNA binding properties of the human and mouse TBPIP/ Hop2-Mnd1 complexes may be somewhat different.
hTBPIP/Hop2-hMnd1 Stimulates Homologous Pairing Mediated by hDmc1-Next, we tested whether hTBPIP/Hop2-hMnd1 stimulates the homologous pairing promoted by hDmc1, like the yeast and mouse orthologs (24,25). The D-loop formation assay was employed to test the Dmc1-mediated homologous pairing ( Fig. 2A). In this assay, hDmc1 formed D-loops within 5 min, and the D-loops were dissociated by the subsequent strand exchange promoted by hDmc1 (Fig. 2, B and D). hTBPIP/Hop2-hMnd1 itself did not promote D-loop formation (Fig.  2C, lane 3, and D) but significantly stimulated the Dmc1-mediated D-loop formation (Fig. 2, B-D). In the experiments presented here, 0.2 M hTBPIP/Hop2-hMnd1 was used to detect stimulation of the Dmc1mediated D-loop formation in the presence of 6.5 M hDmc1. Therefore, the stimulation of the Dmc1-mediated homologous pairing by hTBPIP/Hop2-hMnd1 may be catalytic. Interestingly, hTBPIP/Hop2-hMnd1 significantly enhanced the D-loop yield, but it did not inhibit the D-loop dissociation (Fig. 2, C and D). This observation suggests that the hTBPIP/Hop2-hMnd1 complex may stimulate strand exchange, in addition to homologous pairing.
hTBPIP/Hop2-hMnd1 Stimulates the Strand-exchange Reaction Mediated by hDmc1 and hRad51-We next performed a strand-exchange assay to test whether hTBPIP/Hop2-hMnd1 stimulates strand exchange, which occurs just after homologous pairing. The X174 phage circular ssDNA (5,386 bases) and the linearized X174 dsDNA (5,386 base pairs) were used as DNA substrates (Fig. 3A). The strandexchange reactions were conducted in the presence of human RPA (2 M). In this assay, two reaction products were detected: a joint molecule between ssDNA and dsDNA (JM) and a nicked circular DNA (NC). The JM is a reaction intermediate containing an incomplete strand-ex-change product (Fig. 3A, left to center). On the other hand, the NC is a complete strand-exchange product, whose formation is mediated by the extensive strand-exchange activity of Dmc1 or Rad51 (Fig. 3A, center to  right). The strand exchange must be promoted over a span of five thousand base pairs, to form the NC product with the DNA substrates used in this assay.
As shown in Fig. 3, B and C, hTBPIP/Hop2-hMnd1 significantly enhanced the NC yield in the Dmc1-mediated strand-exchange reaction, indicating that this complex stimulates the strand exchange promoted by hDmc1. Robust enhancement of the NC yield was detected in the presence of 0.5 or 1 M hTBPIP/Hop2-hMnd1 with 15 M hDmc1 (Fig. 3B, lanes 5 and 6, and C). At this optimal hTBPIP/Hop2-hMnd1 concentration, the decrease in the JM yield was concomitant with the increase in the NC yield, suggesting that hTBPIP/Hop2-hMnd1 stimulates the conversion from JM to NC in the Dmc1-mediated strandexchange reaction (Fig. 3, B and C). In addition, hTBPIP/Hop2-hMnd1 also stimulated the strand-exchange promoted by hRad51 (Fig. 3, D and  E). In both the Dmc1-mediated and Rad51-mediated reactions, NC formation was inhibited in the presence of an excess amount of hTBPIP/ Hop2-hMnd1 (Fig. 3, B-E). Time course experiments showed that the NC yield by the Dmc1-and Rad51-mediated reactions was significantly stimulated by hTBPIP/Hop2-hMnd1 at the early reaction times (10 -30 min) (Fig. 3, F and G).
Since the Dmc1-and Rad51-mediated strand exchange is an ATPdependent reaction, we tested whether hTBPIP/Hop2-hMnd1 enhances the ATP hydrolyzing abilities of hDmc1 and hRad51. As shown in Fig. 4, the ATP hydrolyzing abilities of Dmc1 and Rad51 were observed in the presence of ssDNA or dsDNA. However, hTBPIP/ Hop2-hMnd1 did not enhance the DNA-dependent ATP hydrolyzing abilities of hDmc1 and hRad51 (Fig. 4).
hTBPIP/Hop2-hMnd1 Preferentially Binds to a Three-stranded DNA Branch-We next tested the DNA binding specificity of hTB-PIP/Hop2-hMnd1. A competitive DNA binding assay with a 49-mer three-stranded DNA branch, which mimics an intermediate for strand exchange, a 49-mer Y-form DNA (a two-stranded DNA branch), and a 49-mer ssDNA, was performed. Intriguingly, hTBPIP/ Hop2-hMnd1 preferentially bound to the three-stranded DNA branch rather than the Y-form DNA (Fig. 5, A and B). hTBPIP/Hop2-hMnd1 also preferred to bind the Y-form DNA rather than the canonical dsDNA (data not shown), indicating that it has a binding preference for the branched DNA structure. These findings support the idea that hTBPIP/Hop2-hMnd1 preferentially binds to the threestranded DNA branch to stimulate the JM-NC conversion during the strand-exchange process promoted by the Dmc1 and Rad51 proteins.  3, 4, 7, and 8). Lanes 1 and 5 indicate the negative control experiments without proteins. The samples were separated by thin layer chromatography, and were visualized with the BAS2500 image analyzer. B, the reactions were conducted using the same methods as shown in A, except hRad51 was used instead of hDmc1. The reaction mixtures contained 6.5 M hRad51 (lanes 2, 4, 6, and 8) and 1.3 M hTBPIP/Hop2- Mnd1 (lanes 3, 4, 7, and 8). Lanes 1 and 5 indicate the negative control experiments without proteins.