DNA strand exchange activity of rice recombinase OsDmc1 monitored by fluorescence resonance energy transfer and the role of ATP hydrolysis Chittela Rajanikant 1 , Manoj Kumbhakar 2 , Haridas Pal 2 , Basuthkar J. Rao 3 and Jayashree K. Sainis 1 1 Molecular Biology Division, Bhabha Atomic Research Center, Mumbai, India 2 Radiation Chemistry and Chemical Dynamics Division, Bhabha Atomic Research Center, Mumbai, India 3 Department of Biological Sciences, Tata Institute of Fundamental Research, Mumbai, India Homologous recombination is a fundamental process by which two DNA molecules physically interact with each other. This process is important for repairing the double strand breaks (DSBs) induced during mitosis, meiosis and other stages where chromosomal break- ages ensue. There are several sequential biochemical Keywords Dmc1; FRET; renaturation; rice; strand exchange Correspondence J. K. Sainis, Molecular Biology Division, Bhabha Atomic Research Centre, Mumbai 400 085, India Fax: +91 22 25505326 Tel : +91 22 25595079 E-mail: jksainis@magnum.barc.ernet.in (Received 18 October 2005, revised 2 February 2006, accepted 8 February 2006) doi:10.1111/j.1742-4658.2006.05170.x Rad51 and disrupted meiotic cDNA1 (Dmc1) are the two eukaryotic DNA recombinases that participate in homology search and strand exchange reactions during homologous recombination mediated DNA repair. Rad51 expresses in both mitotic and meiotic tissues whereas Dmc1 is confined to meiosis. DNA binding and pairing activities of Oryza sativa disrupted mei- otic cDNA1 (OsDmc1) from rice have been reported earlier. In the present study, DNA renaturation and strand exchange activities of OsDmc1 have been studied, in real time and without the steps of deproteinization, using fluorescence resonance energy transfer (FRET). The extent as well as the rate of renaturation is the highest in conditions that contain ATP, but sig- nificantly less when ATP is replaced by slowly hydrolysable analogues of ATP, namely adenosine 5¢-(b,c-imido) triphosphate (AMP-PNP) or adeno- sine 5¢-O-(3-thio triphosphate) (ATP-c-S), where the former was substan- tially poorer than the latter in facilitating the renaturation function. FRET assay results also revealed OsDmc1 protein concentration dependent strand exchange function, where the activity was the fastest in the presence of ATP, whereas in the absence of a nucleotide cofactor it was several fold ( 15-fold) slower. Interestingly, strand exchange, in reactions where ATP was replaced with AMP-PNP or ATP-c-S, was somewhat slower than that of even minus nucleotide cofactor control. Notwithstanding the slow rates, the reactions with no nucleotide cofactor or with ATP-analogues did reach the same steady state level as seen in ATP reaction. FRET changes were unaffected by the steps of deproteinization following OsDmc1 reaction, suggesting that the assay results reflected stable events involving exchanges of homologous DNA strands. All these results, put together, suggest that OsDmc1 catalyses homologous renaturation as well as strand exchange events where ATP hydrolysis seems to critically decide the rates of the reac- tion system. These studies open up new facets of a plant recombinase func- tion in relation to the role of ATP hydrolysis. Abbreviations AMP-PNP, adenosine 5¢-(b,c-imido) triphosphate; ATP-c-S, adenosine 5¢-O-(3-thio triphosphate); Dmc1, disrupted meiotic cDNA1; DS, double stranded; DSBs, double strand breaks; FRET, fluorescence resonance energy transfer; OsDmc1, Oryza sativa disrupted meiotic cDNA1; SS, single stranded; RecA, DNA recombinase A; RPA, replication protein A. FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS 1497 reactions that prepare the DNA molecules for repair via homologous recombination. These sequential reac- tions involve identification of DSBs, processing the broken DNA at the damage site by resecting the ends, facilitating homology search ⁄ synapses, strand exchange between the paired chromosomes, followed by resolu- tion of the Holliday junction. Specific proteins required for each of these steps have been identified mainly from Escherichia coli and now also from yeast and mammalian systems. The proteins implicated in homology search and pairing activities are called strand exchange proteins or DNA recombinases. DNA recombinase A (RecA) from E. coli has been extensively investigated both at the biochemical and molecular level [1]. These studies revealed that RecA protein binds to single strand overhangs generated as a result of processing at the DSB site. In this presynaptic complex, RecA protein coats single stranded DNA (ssDNA) as a helical filament (three nucleotides per protein monomer) resulting in stretching of DNA by 1.5 times its original length. This conformational change in DNA is hypothesized to facilitate homology search in the double stranded DNA (dsDNA). Once the homologous region is found, RecA protein medi- ates strand exchange with its duplex partner. This pro- cess is known to require ATP. E. coli RecA is thus shown to be a remarkable protein which alone can cat- alyze the DNA interactions necessary for establishment of homologous contacts, eventually leading to homo- logous recombination mediated repair. This feature stimulated the search for homologues of RecA in euk- aryotes using classical biochemical techniques, which did not yield much success. Introduction of molecular techniques led to identification of two RecA homo- logues, namely Rad51 and disrupted meiotic cDNA1 (Dmc1) first in yeast and later in mammals and plants. Rad51 was found to express during mitosis as well as meiosis, whereas Dmc1 was confined mainly to meiosis [2]. Rad51 homologues in yeast, mouse, and humans have been characterized at both the genetic and bio- chemical levels [3]. The deletion of Rad51 resulted in embryonic lethality in mice [4]. Using transposon induced mutagenesis, Bishop et al. [5] identified DMC1 gene in Saccharomyces cereveciae, which had sequence similarities with recA and RAD51. ScDMC1 mutants showed defective reciprocal recombination, accumula- tion of double strand breaks and failure to form syn- aptonemal complexes, eventually leading to arrest of the cells in meiotic prophase. DMC1 gene was also detected in mouse and DMC1 – ⁄ – mice were viable but sterile, showing reduced reproductive organ sizes and asynapsis or random segregation of chromosomes [6,7]. In the case of plants, the DMC1 knockout in Arabidopsis resulted in sterile plants and showed asynapsis in meiosis [8]. As a sequel to identification of recA homologues in eukaryotes, attempts were made to clone and overex- press these genes for biochemical characterization. The biochemical properties of Dmc1 proteins from yeast, Coprinus and human systems have emerged recently. Dmc1 proteins from S. cerevisiae and Coprinus cenere- us were shown to catalyze strand assimilation of radio- labeled oligonucleotides into homologous duplex DNA in a reaction promoted by ATP and ATP analogues [9,10]. Coprinus cenereus Dmc1 protein interacts homo- typically and mediates a homology dependent strand exchange reaction [11]. Using fluorescence resonance energy transfer (FRET), hDmc1 was shown to catalyze the strand exchange and strand assimilation in a homology dependent manner [12]. Recent studies on hDmc1 showed that it was able to mediate the strand exchange reaction at least up to several kilobase pairs (5.4 kb) in vitro in cooperation with a heterotrimeric protein, namely replication protein A (RPA) [13]. Though eukaryotic RecA homologues have been well characterized from yeast and animal model sys- tems, the information from plants is still in its infancy. Previously, homologues of DMC1 and RAD51 have been reported from Lilium longiflorum [14,15], Arabid- opsis thaliana [16,17] and Oryza sativa [18–21]. It is important to gather biochemical information about the proteins participating in homologous recombination in plant systems, which are constantly under genetic improvement programs through conventional or induced mutation breeding or through new transgenic manipulations. We have initiated studies on the bio- chemical characterization of DNA recombinases from rice. As gametophytic tissue is limited in amount and also the expression of Oryza sativa disrupted meiotic cDNA1 (OsDmc1) protein is temporal, we have cloned and overexpressed OsDmc1 protein in E. coli [21]. The overexpressed OsDmc1 protein was purified under denaturing conditions to homogeneity and renatured to its native state. Renatured protein exhibited all the important functional hallmarks of a recombinase, as described below, thereby showing it has attained its native structure. This purified protein showed single and double stranded DNA binding activity. Binding to single stranded DNA stimulated a significant level of DNA dependent ATPase activity. OsDmc1 also showed renaturation of complementary strands as well as assimilation of single strands into homologous supercoiled duplexes leading to D-loop formation [22]. In the present study, we extend our efforts further and show that OsDmc1 mediates renaturation as well as strand exchange activities whose rates are critically DNA strand exchange activity of OsDmc1 by FRET C. Rajanikant et al. 1498 FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS dependent on the state of ATP hydrolysis. The results, based on FRET assays conducted in real time, report that stable changes are associated with homologous strands in the native state of the reaction and involve no protein removal steps. Results Strand exchange activity of several recombinases have been investigated in vitro, typically using agarose gel assays with radiolabeled oligonucleotides as substrates [1,9,10,12]. Though commonly used because of its sim- plicity, this assay has drawbacks because it involves removal of DNA bound proteins and hence does not score the reaction at its equilibrium state. An assay using FRET, involving fluorophore labeled oligo- nucleotides as substrates, is not only sensitive but also scores the reaction without perturbing its equilibrium state, as it involves no sample deproteinization and can be carried out in real time. Therefore, in the pre- sent study we have used FRET for measuring recombi- nation activities of OsDmc1 in vitro. Design of the assay Two complementary oligonucleotides (Phi-W and Phi- C, Fig. 1A), were labeled with fluorescein and rhodam- ine at their 5¢ and 3¢ ends, respectively. The assay is based on nonradiative fluorescence resonance energy transfer from fluorescein (donor) to rhodamine (accep- tor). FRET, being highly distance dependent between donor and acceptor dyes, will therefore unveil the sta- tus of physical union and separation of complementary strands during renaturation and strand exchange, respectively, and thereby assesses the recombinase activity of OsDmc1, as has been shown for other rec- ombinase enzymes previously [12,23]. Fluorescein and rhodamine are a good FRET pair, where energy trans- fer from fluorescein results in loss of its emission inten- sity at 522 nm following excitation at 490 nm. Thus, renaturation between Phi-W and Phi-C strands can be assayed through FRET as the decrease in fluorescein emission intensity at 522 nm (Fig. 1A). Conversely, strand exchange leads to separation of these duplexed strands resulting in an increase in the fluorescein emis- sion intensity (Fig. 1B). In all the experiments des- cribed below, reactions were performed where the decrease (renaturation) or increase (strand exchange) of fluorescein emission intensity at 522 nm is scored as a function of reaction time. The assays are carried out in real time, where reproducibly similar trends were observed in duplicate sets performed together as a set on a given day. However, the initial rates of a reaction encompassing the first few seconds of the time course, performed on different days, showed variation, per- haps arising from minor differences in reaction condi- tions. We believe that our readouts being entirely fluorescence based were very sensitive to reaction con- ditions, thereby leading to high confidence levels for comparison within the sets, rather than across the sets. Because the results within the set were highly reprodu- cible, the differences in reaction rates were significant, based on which, we describe below the contrasting effects of ATP (hydrolyzing versus nonhydrolyzing conditions) in renaturation and strand exchange reac- tion kinetics. Renaturation activity of OsDmc1 Renaturation was measured at a fixed concentration of protein using Phi-C⁄ Phi-W strands tagged with fluores- cence labels (Fig. 1A) in the presence of ATP, as a func- tion of time (Fig. 2A). OsDmc1 was presynapsed with Phi-C oligonucleotide, followed by the addition of com- plementary strands (Phi-W). As expected, strand anneal- ing led to a time dependant drop in fluorescein emission Fig. 1. Schematic representation of the renaturation (A) and strand exchange (B) reactions mediated by OsDmc1 protein. Phi-C and Phi-W represent rhodamine and fluorescein carrying strands, respectively, at their 3¢ and 5¢ ends. The assays were based on nonradiative energy transfer from fluorescein to rhodamine when fluorescein was excited at 490 nm, followed by measurement of emission intensity at 522 nm, due to the two dyes being in close proximity. Renaturation activity was measured as the decrease in fluorescein emission intensity and strand exchange was monitored as the increase in fluorescein emission intensity at 522 nm. C. Rajanikant et al. DNA strand exchange activity of OsDmc1 by FRET FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS 1499 intensity (Fig. 2). The rate of renaturation catalyzed by OsDmc1 was rather high; within 200 s the reaction reached steady state level (Fig. 2A, line 3). Under the same assay conditions, a control with no protein revealed much slower spontaneous renaturation reaction (Fig. 2A, line 2). Even after several minutes of incubation sponta- neous renaturation did not seem to have reached its steady state level. The rate of OsDmc1 catalyzed renatur- ation was much faster at higher concentration of protein (more than 1.25 lm; data not shown). The protein to nucleotide ratio (1 : 10–1 : 20) used in these reactions is close to the reported optimum of ScDmc1, where each protein molecule was observed to bind about 10–40 nucleotides for its maximum activity [9]. As expected, another control where only the presynaptic complex (con- taining Phi-W strand) was incubated in the absence of Phi-C strand, showed no loss in fluorescence intensity as a function of time (Fig. 2A, line 1), thereby revealing that the rapid loss of fluorescence signal catalyzed by OsDmc1 (Fig. 2A, line 3) reflects genuine FRET change associated with strand renaturation. Homology dependence of renaturation activity In the following experiment, we examined the homol- ogy dependence of the renaturation reaction. FRET efficiency was assessed as a function of different doses of unlabeled competitor strand (homologous or heterologous) premixed with labeled Phi-C oligonucleo- tide, followed by renaturation for 5 min. FRET effi- ciency of a reaction was estimated by subtracting the fluorescence emission intensity measured after renatur- ation (at 5 min) from the starting value (0 min). There- fore, the measured efficiencies stem from direct readouts of fluorescence emission. It was surmised that the renaturation as measured by FRET is competed specifically by the homologous unlabeled strand (it cannot accept FRET because it lacks the label) in a dose dependent manner, while the heterologous unlabeled competitor will have no significant effect on the same. FRET efficiency decreased specifically in the presence of unlabeled homologous competitor strand (Phi-C) as a function of its concentration (Fig. 2B, his- togram 2 and 3). On the other hand, no significant change in FRET efficiency was observed when unlabe- led heterologous (M13C) strand was added to the renaturation mixture (Fig. 2B, histograms 4 and 5). In these sets, FRET efficiency was highly comparable to that where no competitor was present (Fig. 2B, histo- gram 1). This experiment revealed that OsDmc1 cata- lyzed renaturation as measured by FRET change was homology dependent. Effect of ATP and its hydrolysis on renaturation activity We also examined the effect of ATP on renaturation catalyzed by OsDmc1. Interestingly enough, in the Fig. 2. (A) Time course of renaturation reaction as monitored by decrease in fluorescein emission intensity at 522 nm expressed as arbitrary units normalized to one. (1) Reaction containing Phi-W oligonucleotide with 1.25 l M of OsDmc1. (2) Reaction containing Phi-W oligonucleotide and Phi-C oligonucleotide without OsDmc1. (3) Reaction containing Phi-W oligonucleotide and Phi-C oligonucleo- tide with 1.25 l M of OsDmc1. (B) Homology dependent renatura- tion reaction mediated by OsDmc1. Renaturation reaction of OsDmc1 (1) without competitor; (2) with 27.5 l M unlabeled of Phi- C along with 27.5 l M of rhodamine labeled Phi-C; (3) with 55.0 lM unlabeled of Phi-C along with 27.5 lM of rhodamine labeled Phi-C; (4) with 27.5 l M unlabeled of M13C along with 27.5 lM of rhodam- ine labeled Phi-C; and (5) with 55.0 l M unlabeled of M13C along with 27.5 l M of rhodamine labeled Phi-C. Samples 2, 3 and 4, 5 represent renaturation reactions in the presence of homologous and heterologous unlabeled competitors, respectively. In each reac- tion, FRET efficiency was measured by the decrease in fluorescein emission intensity at 522 nm following 5 min of renaturation reac- tion. The FRET efficiencies are plotted: without any competitor (his- togram 1); with the competitor (histograms 2–5). DNA strand exchange activity of OsDmc1 by FRET C. Rajanikant et al. 1500 FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS absence of ATP, OsDmc1 exhibited hardly any rena- turation activity (Fig. 3, line 1). It is interesting to note that under these conditions, presence of OsDmc1 seems to attenuate even spontaneous strand anneal- ing, as evidenced by negligible annealing observed in this experiment compared to that observed in the absence of any protein (Fig. 2A, line 2). However, when the reaction mixture contained slowly hydroly- sable forms of ATP, namely adenosine 5¢-(b,c-imido) triphosphate (AMP-PNP) or adenosine 5¢-O-(3-thio triphosphate) (ATP-c-S), renaturation was augmented (Fig. 3, line 2 and 3), but still fell short of that observed in spontaneous annealing (Fig. 2A, line 2). Moreover, the enhancement in renaturation activity was most dramatic when ATP was added, where not only the overall rate but also the total extent of rena- turation by OsDmc1 was stimulated several fold (Fig. 3, line 4). However, it is intriguing to note that the initial rate of renaturation in the presence of ATP-c-S was similar to that of ATP, but the reaction rate suddenly plummeted after about 50 s of the reac- tion. There is no simple explanation for this observa- tion. These results demonstrated that in contrast to ScDmc1, where renaturation was not dependent on ATP [9], OsDmc1 protein requires not only the pres- ence of ATP, but also perhaps its hydrolysis, for mediating the maximum renaturation activity (see below). Strand exchange activity of OsDmc1 Protein concentration dependence OsDmc1 was shown to have the ability to mediate homology dependent D-loop formation activity [22]. In the present study, we extended our analyses further, and the strand exchange property of OsDmc1 was monitored by pairing an unlabeled Phi-C single strand with duplex oligonucleotide formed by complementary annealing of FRET dye labeled strands (Phi-W and Phi-C; Fig. 1B). Interestingly enough, the three stran- ded pairing reaction exhibited OsDmc1 concentration dependent increase in fluorescein emission intensity at 522 nm as a function of reaction time (Fig. 4A). This was in stark contrast to the renaturation reaction, where the emission intensity at the same wavelength had decreased as a function of time (Figs 2 and 3). The observed increase in fluorescein emission intensity is consistent with strand exchange as detected by FRET. It should be noted that no emission increase (indicative of lack of strand exchange) was detected in the absence of OsDmc1 protein (Fig. 4A, line 1). Inter- estingly, the rate as well as the steady state levels of strand exchange increased as a function of OsDmc1 protein concentration (Fig. 4A, lines 2 and 3). How- ever, at concentrations higher than 2.5 lm OsDmc1, there was only an increase in the rate, but no further increase in the steady state levels of the reaction (Fig. 4A, lines 3–6). At high enough concentrations of protein, presumably reflecting a binding saturation, both rate as well as the steady state level of the reac- tion reached a maximum (Fig. 4A, lines 5 and 6); at this condition the ratio of protein to ssDNA nucleo- tides approached a value close to 1 : 3, implying a binding stoichiometry similar to that reported for other recombinase enzymes [1,9,12,23]. FRET changes versus deproteinization of strand exchange products In order to establish that the observed FRET changes are related to strand exchange, as depicted in Fig. 1B, and not due simply to the transient pairing events dur- ing homology search, we performed the following experiment. We inferred that because strand exchange related changes, unlike that of transient pairing events, are more stable to the steps of protein removal, the observed FRET changes, if related to strand exchange, must be stable even after deproteinization treatment. Strand exchange was performed in the presence and absence of OsDmc1 for 15 min, where expected FRET changes were observed specifically in the protein con- taining reaction (Fig. 4B, lines 2 and 3). At this point, Fig. 3. Renaturation activity mediated by OsDmc1 protein in the presence of ATP and slowly hydrolysable ATP analogues as monit- ored by the decrease in fluorescein emission intensity at 522 nm expressed as arbitrary units normalized to one. Reaction contained fluorescein labeled Phi-W and rhodamine labeled Phi-C (1) with 1.25 l M of OsDmc1 in absence of ATP; (2) with 1.25 lM of OsDmc1 in presence of 2.0 m M of AMP-PNP; (3) with 1.25 lM of OsDmc1 in presence of 2.0 mM ATP-c-S; (4) with 1.25 lM of OsDmc1 in presence of 2.0 m M of ATP. C. Rajanikant et al. DNA strand exchange activity of OsDmc1 by FRET FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS 1501 one of the sets was treated with deproteinization steps (shown by arrow in Fig. 4B, line 3) and the other reac- tion set continued with OsDmc1 action (Fig. 4B, line 2). The fluorescence emission changes remained stable to deproteinization treatment. This was evidenced by fluorescence values that were similar in deproteinized and nondeproteinized samples, suggesting that the steady state changes were stable (Fig. 4B, lines 2 and 3). Under the conditions of the assay, deproteinizing agents SDS, EDTA and proteinase K had no effect on the emission intensity of fluorescein as evidenced by the set where OsDmc1 was omitted in the sample (Fig. 4B, line 1). In order to confirm that the increase in the fluorescence signal is specific to the emission maximum of fluorescein, the donor dye in the FRET pair, we compared the emission spectra of the strand exchange reaction mixture with that where OsDmc1 protein was not added (no strand exchange control). Emission scan revealed that enhancement of emission at 522 nm was specific and related to strand exchange reaction conditions (Fig. 5). The results indicated that the FRET changes, observed in real time, reflected changes consistent with strand exchange reaction. Effect of homology dependence and ATP hydrolysis on strand exchange activity Strand exchange was monitored as a function of reac- tion time, at a fixed concentration of protein, by varying nucleotide cofactor conditions. The reaction was the fastest when it contained ATP, where within about 100–150 s the reaction essentially reached completion (Fig. 6, line 2). On the other hand, when ATP was omit- ted (no nucleotide cofactor condition) the same reaction took about 1700 s, a drop in the reaction rate by  15-fold (Fig. 6, line 3). Most interestingly, the pres- ence of slowly hydrolyzing nucleotide cofactors (AMP- PNP or ATP-c-S) appears to further slow down the reaction as compared to that where no nucleotide cofac- tor was added (Fig. 6, compare lines 4 and 5 with 3). As Fig. 5. Fluorescence emission spectra of fluorescein in strand exchange reaction mediated by OsDmc1 protein (12.5 l M). Fluor- escein was excited at 490 nm and emission was monitored from 500 to 550 nm. Spectrum obtained in absence of OsDmc1 protein (1), and in presence of OsDmc1 protein (2). Fluorescein emission intensity is expressed as arbitrary units. Fig. 4. (A) Time course and OsDmc1 protein concentration depend- ence of strand exchange reaction monitored by increase in fluo- rescein emission intensity at 522 nm in arbitrary units. (1) Without OsDmc1 protein; (2) 1.25 l M; (3) 2.5 lM; (4) 5.0 lM; (5) 10.0 lM; and (6) 12.5 l M of OsDmc1 protein. (B) Effect of deproteinization on strand exchange reaction mediated by OsDmc1. Fluorescence was monitored at 522 nm in arbitrary units. Reaction mixture (1) without OsDmc1 deproteinized after 15 min; (2) with 5.0 l M OsDmc1 protein without deproteinization; and (3) with 5.0 lM OsDmc1 protein with deproteinization after 15 min as indicated in figure with arrow. DNA strand exchange activity of OsDmc1 by FRET C. Rajanikant et al. 1502 FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS expected, when OsDmc1 was omitted or homologous single strand oligonucleotide was replaced with non- complementary sequence M13C as heterologous con- trol, no increase in fluorescence intensity was observed, re-establishing the veracity of the assay to an ongoing homologous strand exchange activity dependent on OsDmc1 function (Fig. 6, lines 1 and 6). The results indicated that the strand exchange activity catalyzed by OsDmc1 is homology dependent and is facilitated by ATP and its hydrolysis at kinetic level. Discussion We have been studying the biochemistry of OsDmc1 where we have shown earlier [22] that the rice enzyme exhibits many hallmarks typical of recombinases. In the present study, we extend our analyses further and show that the rice recombinase exhibits strand exchange function and that its rates are critically ATP hydrolysis dependent. We have used a fluorimetric assay to monitor the time course of renaturation and strand exchange activ- ities of OsDmc1. The renaturation activity was found to be stimulated in the presence of ATP and was satur- ated at 1.25 lm of OsDmc1. In the competition assay, renaturation activity was observed to be dependent on the presence of homologous sequence partner in the reaction mixture (Fig. 2). The calculated protein : nuc- leotide ratio is 1 : 10–1 : 20, which compares well with the reported values (1 : 10–1 : 40) for ScDmc1 [9]. In contrast to ScDmc1, renaturation activity by OsDmc1 was more efficient in the presence of ATP, though the slowly hydrolysable analogues of ATP partially promoted the renaturation activity (Fig. 3). This prop- erty appears to be more similar to RecA and Rad51, which were shown to promote efficient renaturation even in the absence of ATP hydrolysis [23]. However, ATP-c-S was more efficient in promoting renaturation when compared to AMP-PNP, consistent with the property that the former is a little more hydrolysable than the latter. Interestingly, it appears that protein binding to ssDNA in the absence of any nucleotide cofactor results in attenuation of spontaneous anneal- ing (compare line 1, Fig. 3 with line 2, Fig. 2A); this attenuation is undone only in the presence of nucleo- tide cofactors where the effect by non- or slowly hydrolysable analogues is much weaker compared to ATP (lines 2 and 3 versus 4, Fig. 3). This result seems to suggest that the protein binding mode ⁄ dynamics are significantly different in these diverse conditions. OsDmc1 was found to promote strand exchange in a protein concentration dependent manner. The rate of strand exchange was highest in ATP and lowest in conditions that had either no nucleotide cofactor or had slowly hydrolysable analogues of ATP. The rate enhancement by the presence of ATP was in the order of about 15-fold (Fig. 6), strongly suggesting that the hydrolysis of ATP somehow overcomes the rate limit- ing barriers in the reaction pathway. The FRET chan- ges observed were indeed related to stable changes associated with DNA in the strand exchange reaction as they were stable to the steps of protein removal (Fig. 4B). OsDmc1 mediated strand transfer was not observed when homologous ssDNA oligonucleotide was replaced with a heterologous sequence oligonucleo- tide in the assay mixture. From this result we conclu- ded that the OsDmc1 mediated strand exchange reaction is homology dependent and is not related to any contaminating helicase activity spuriously associ- ated with the purified OsDmc1 preparation. It is rele- vant to point out that OsDmc1 protein is somewhat distinct as compared to either human or yeast Dmc1 proteins: while human protein requires ATP to promote pairing and strand exchange [12] the same is not true with either OsDmc1 (this study) or yeast Dmc1 [9]. However, unlike yeast Dmc1 which does not require nucleotide cofactor for renaturation activity, OsDmc1 requires ATP hydrolysis for efficient renatur- ation reaction. Sequence comparison between OsDmc1 and that of human and yeast proteins reveals no signi- Fig. 6. Strand exchange reaction mediated by OsDmc1 in presence of ATP or slowly hydrolysable ATP analogues monitored by increase in fluorescein emission intensity at 522 nm in arbitrary units. Reactions contained OsDmc1 (5.0 l M) and ATP or slowly hydrolysable ATP analogues (2.0 m M). (1) Without OsDmc1 protein in presence of ATP; (2) with OsDmc1 protein with ATP; (3) with OsDmc1 protein without ATP; (4) with OsDmc1 protein with ATP- c-S; (5) with OsDmc1 protein with AMP-PNP; and (6) with OsDmc1 protein and ATP in presence of heterologous sequence M13C oligo- nucleotide. C. Rajanikant et al. DNA strand exchange activity of OsDmc1 by FRET FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS 1503 ficant changes either in ATPase or in DNA binding domains of OsDmc1 [22]. The only discernible change in OsDmc1 is an insertion of seven amino acid residues at the N-terminal region, a change also observed in Arabidopsis protein [22]. We do not know the struc- tural consequence of such an insertion because there is no structure available for this region due to its highly flexible nature [24]. If the Dmc1 oligomeric protein uses this flexible region for regulating ring to helix transformation steps, the dynamics of the same are likely to be different in OsDmc1 protein system compared to that of human and yeast proteins. Our results on renaturation and strand exchange activities of OsDmc1 showed interesting differences in protein requirement for optimal activity. As renatura- tion is an inherent property of complementary strands, it may not require complete coating of ssDNA with OsDmc1. Therefore, a protein to nucleotide ratio of as low as 1 : 20 promoted the complete renaturation. In contrast, the strand exchange reaction required one monomer of OsDmc1 for two to three nucleotides, suggesting that strand exchange probably needs ssDNA filament saturated with OsDmc1 whereas for renaturation partially coated ssDNA was sufficient for optimal activity. These results are in agreement with the results reported for ScDmc1 [9]. Passy et al. [25] and Masson et al. [26] have demon- strated that human Dmc1 forms octameric ring like structures on ssDNA. Recent atomic force microscopy studies have shown that ScDmc1 forms 90% octameric ring-like structures as well as 10% helical filaments upon binding to ssDNA. The helical forms were hypo- thesized to represent the active form responsible for recombination reactions [27]. The presence of ATP was found to result in the formation of helical fila- ments on ssDNA, whereas in the absence of ATP there was a preponderance of octameric rings [13]. Recent studies showed that Ca 2+ enhances the strand exchange activity of human Dmc1 protein by increas- ing the affinity of Dmc1 protein to ATP, mediated by a conformational change [28]. In our strand exchange assays, the presence of ATP significantly enhanced the reaction rate of strand exchanges as compared to that of no nucleotide control or that containing slowly hydrolysable analogues of ATP. Put together, these observations suggest that perhaps the ATP hydrolysis function of the protein may be related to some critical steps in the conversion of inactive forms of protein to that of active forms. It is interesting and intriguing to note that in the absence of any nucleotide cofactor, OsDmc1 leads to a slow process of strand exchange essentially going to completion (line 3, Fig. 6), but under the same conditions, the renaturation reaction is strongly suppressed (line 1, Fig. 3) even as compared to that of spontaneous annealing (line 2, Fig. 2A). The results are best rationalized by invoking better binding of protein to ssDNA compared to dsDNA under these conditions, thereby leading to suppression of ssDNA renaturation, as two interacting protein–ssDNA com- plexes are unlikely to pair. Conversely, the protein– ssDNA complex might facilitate ATP independent, kinetically slow steps of pairing and exchange with rel- atively naked duplex DNA. Fine structural studies are required on OsDmc1 to explain the structure–function relationship of the recombinase from rice. It will fur- ther enhance our understanding of the homologous recombination and DNA repair in plant systems. It is relevant to point out that the real time assay employed here enabled us to distinguish the kinetic differences in ATP containing reaction sets versus those containing no nucleotide control or slowly hydrolysable analogues of ATP. The assays that simply measure the steady state levels of products formed after several minutes of the reaction will fail to detect these important changes, which sometimes contribute to conflicting results on strand exchange versus nucleotide cofactor effects. Nevertheless, the kinetic assay described here is a sim- ple and generally applicable one that will be used in our future studies to understand the mechanistic details of OsDmc1 function compared to ATP hydro- lysis rates as well as OsDmc1 changes in the presence of its functional interactors. Experimental procedures Materials Oligonucleotides (55-mers) for strand exchange assay were synthesized by Metabion (Martinsreid, Germany) with the following sequences: PhiC: 5¢-CGATACGCTCAAAGTCA AAATAATCAGCGTGACATTCAGAAGGGTAATAAG AACG-3¢;, PhiW: 5¢-CGTTCTTATTACCCTTCTGAA TGTCACGCTGATTATTTTGACTTTGAGCGTATCG-3¢ and M13C: 5¢-CTACAACGCCTGTAGCATTCCACAGA CAGCCCTCATAGTTAGCGTAACGAGATCG-3¢. Phi-C and Phi-W were complementary to each other. Phi-C was labeled with rhodamine at the 3¢ end and Phi-W was labe- led with fluorescein at the 5¢ end. M13C was used as the heterologous strand in the strand exchange assay. cDNA for OsDmc1 protein was obtained from rice anthers by RT-PCR and was cloned previously in pET28a, intro- duced into E. coli BL21(DE3) expression cells. Protein was overexpressed by 1.0 mm isopropyl thio-b-d-galactoside [21]. The OsDmc1 was purified and stored at )20 °C as described previously [22]. ATP, AMP-PNP and ATP-c-S were pur- chased form Sigma Chemical Company (St Louis, MO, DNA strand exchange activity of OsDmc1 by FRET C. Rajanikant et al. 1504 FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS USA). Duplex oligonucleotide was prepared by mixing equal amounts of rhodamine labeled Phi-C and fluorescein labeled Phi-W followed by thermal denaturation at 92 °C for 10 min. The mixture was subjected to a slow renaturation step by bringing the temperature to 25 °C over a period of 2 h [29]. Renaturation assay The renaturation activity was carried out according to Gupta et al. [23] with the following modifications. A reaction mixture (100 lL) containing 20 mm HEPES (pH 7.9), 2 mm ATP, 10.0 mm MgCl 2 , 3.0% (w ⁄ v) glycerol, 1.0 mm dithiothreitol and 1.25 lm of OsDmc1 was preincu- bated at 37 °C for 5 min with Phi-C oligonucleotide (27.5 lm of nucleotides) labeled with rhodamine at the 3¢ end. Complementary oligonucleotide Phi-W (27.5 lm of nucleotides) labeled with fluorescein at the 5¢ end was added, and the decrease in fluorescein emission intensity as a result of FRET was measured at 522 nm after excitation at 490 nm using a F-4010 Hitachi fluorescence spectrophotometer (Hitachi Ltd, Tokyo, Japan). The decrease in fluorescence emission intensity of fluorescein at 522 nm after excitation at 490 nm was measured at 20 s intervals for 15 min. ATP was omitted or replaced with 2.0 mm AMP-PNP or ATP-c-S in some assays as mentioned in figure legends. To show that the observed renaturation activity was homology dependent, we carried out the following compet- itive FRET assay. In the standard renaturation assay, oligonucleotide Phi-W (27.5 lm of nucleotides) labeled with fluorescein at the 5¢ end was preincubated with 1.25 lm of OsDmc1 at 37 °C for 5 min. Phi-C oligonucleotide (27.5 lm of nucleotides) labeled with rhodamine at the 3¢ end was premixed with either unlabelled Phi-C as homolog- ous competitor (0, 27.5, 55.0 lm of nucleotides) or unla- belled M13C as heterologous competitor (27.5, 55.0 lm), followed by its addition to the reaction mixture. The extent of homologous versus heterologous competition in FRET was monitored by the decrease in the fluorescein emission intensity at 522 nm after 5 min. FRET efficiencies were cal- culated by subtracting the fluorescence value obtained at 5 min from that of 0 min. Therefore the measured efficien- cies stem from direct readouts of fluorescence emission. Strand exchange assay Strand exchange activity was monitored essentially accord- ing to the procedure mentioned by Gupta et al. [12]. A reaction mixture (100 lL) containing 20 mm HEPES (pH 7.9), 2 mm ATP, 10.0 mm MgCl 2 , 3.0% (w ⁄ v) gly- cerol, 1.0 mm dithiothreitol and different concentration of OsDmc1 was preincubated with unlabeled Phi-C oligo- nucleotide (27.5 l m of nucleotides) for 5 min at 37 °C. Duplex 55-mer made from fluorescein labeled Phi-W and rhodamine labeled Phi-C was added (27.5 lm of nucleo- tides) as homologous duplex. Increase in fluorescence emission intensity of fluorescein at 522 nm after excitation at 490 nm was measured at 20 s intervals for the first 15 min and at 1 min intervals for the next 15 min due to the loss of FRET as a result of strand exchange. Oligonu- cleotide M13C was used as heterologous control in strand exchange assay. ATP was omitted or replaced with 2.0 mm AMP-PNP or ATP-c-S in some assays as men- tioned in figure legends. Effect of deproteinazation on strand exchange activity To check whether the strand exchange activity of OsDmc1 is due to complete strand exchange or is due to transient pairing events leading to partial local separation of strands, unlabeled Phi-C oligonucleotide (27.5 lm of nucleotides) was preincubated with 5.0 lm of OsDmc1 protein for 5 min at 37 °C followed by the addition duplex 55-mer made from fluorescein labeled Phi-W and rhodamine labe- led Phi-C. Increase in the fluorescence was measured at 522 nm for 15 min. The same reaction mixture was subse- quently deproteinized with 20 mm EDTA, 1.0% SDS and 100 lgÆmL )1 of freshly prepared proteinase K. Fluorescence emission was measured for a further 15 min. Acknowledgements We would like to acknowledge Dr S.K. Apte and Dr A.S. Bhagwat, Molecular Biology Division, BARC, Mumbai, India for critical reviewing of this manu- script. References 1 Kowalczykowski SC, Dixon DA, Eggleston AK, Lauder SD & Rehrauer WM (1994) Biochemistry of homologous recombination E. coli. Microbiol Rev 58, 401–465. 2 Masson JY & West SC (2001) The Rad51 and Dmc1 recombinases: a non-identical twin relationship. Trends Biochem Sci 26, 131–136. 3 Sung P, Krejci L, Van Komen S & Sehorn MG (2003) Rad51 recombinase and recombination mediators. J Biol Chem 278, 42729–42732. 4 Tsuzuki T, Fujii Y, Sakumi K, Tominaga Y, Nakao K, Sekiguchi M, Matsushiro A, Yoshimura Y & Morita T (1996) Targeted disruption of the Rad51 gene leads to lethality in embryonic mice. Proc Natl Acad Sci USA 93, 6236–6240. 5 Bishop DK, Park D, Xu L & Kleckner N (1992) DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex for- mation, and cell cycle progression. Cell 69, 439–456. C. Rajanikant et al. DNA strand exchange activity of OsDmc1 by FRET FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS 1505 6 Yoshida K, Kondoh G, Matsuda Y, Habu T, Nishi- mune Y & Morita T (1998) The mouse RecA-like gene Dmc1 is required for homologous chromosome synopsis during meiosis. Mol Cell 1, 707–718. 7 Pittman DL, Cobb J, Schimenti KJ, Wilson LA, Cooper DM, Brignull E, Handel MA & Schimenti JC (1998) Meiotic prophase arrest with failure of chromosome synapsis in mice deficient for Dmc1, a germline-specific RecA homolog. Mol Cell 1, 697–705. 8 Couteau F, Belzile F, Horlow C, Grandjean O, Vezon D & Doutriaux MP (1999) Random chromosome segregation without meiotic arrest in both male and female meiocytes of a dmc1 mutant of Arabidopsis. Plant Cell 11, 1623–1634. 9 Hong EL, Shinohara A & Bishop DK (2001) Saccharo- myces cerevisiae Dmc1 protein promotes renaturation of single-strand DNA (ssDNA) and assimilation of ssDNA into homologous super-coiled duplex DNA. J Biol Chem 276, 41906–41912. 10 Nara T, Hamada F, Namekawa S & Sakaguchi K (2001) Strand exchange reaction in vitro and DNA-dependent ATPase activity of recombinant LIM15 ⁄ DMC1 and RAD51 proteins from Coprinus cinereus. Biochem Biophys Res Commun 285, 92–97. 11 Nara T, Yamamoto T & Sakaguchi K (2000) Characteri- zation of interaction of C- and N-terminal domains in LIM15 ⁄ DMC1 and RAD51 from a basidiomycetes, Coprinus cinereus. Biochem Biophys Res Commun 275, 97–102. 12 Gupta RC, Golub E, Bi B & Radding CM (2001) The synaptic activity of HsDmc1, a human recombination protein specific to meiosis. Proc Natl Acad Sci USA 98, 8433–8439. 13 Sehorn MG, Sigurdsson S, Bussen W, Unger VM & Sung P (2004) Human meiotic recombinase Dmc1 pro- motes ATP-dependent homologous DNA strand exchange. Nature 429, 433–437. 14 Kobayashi T, Hotta Y & Tabata S (1993) Isolation and characterization of a yeast gene that is homologous with a meiosis specific cDNA from a plant. Mol Gen Genet 237, 225–232. 15 Kobayashi T, Kobayashi E, Sato S, Hotta Y, Miyajima N, Tanaka A & Tabata S (1994) Characterization of cDNAs induced in meiotic prophase in lily microsporo- cytes. DNA Res 1, 15–26. 16 Sato S, Hotta Y & Tabata S (1995) Structural analysis of a recA-like gene in the genome of Arabidopsis thali- ana. DNA Res 2, 89–93. 17 Doutriaux MP, Couteau F, Bergounioux C & White C (1998) Isolation and characterisation of the RAD51 and DMC1 homologues from Arabidopsis thaliana. Mol Gen Genet 257, 283–291. 18 Shimazu JC, Matsukura M, Senda R, Ishikawa S, Akada T, Harada T & Niizeki M (2001) Characteriza- tion of a DMC1 homologue, RiLIM15, in meiotic pani- cles, mitotic cultured cells and mature leaves of rice (Oryza sativa L.). Theor Appl Genet 102, 1159–1163. 19 Ding Z, Wang T, Chong K & Bai S (2001) Isolation and characterization of OsDMC1, the rice homologue of the yeast DMC1 gene essential for meiosis. Sex Plant Reprod 13, 285–288. 20 Kathiresan A, Khush GS & Bennett J (2002) Two rice DMC1 genes are differentially expressed during meiosis and during haploid and diploid mitosis. Sex Plant Reprod 14, 257–267. 21 Metkar SS, Sainis JK & Mahajan SK (2004) Cloning and characterization of the DMC1 genes in Oryza sativa. Curr Sci 87, 353–357. 22 Kant CR, Rao BJ & Sainis JK (2005) DNA binding and pairing activity of OsDmc1, a recombinase from rice. Plant Mol Biol 57, 1–11. 23 Gupta RC, Bazemore LR, Golub EI & Radding CM (1997) Activities of human recombination protein Rad51. Proc Natl Acad Sci USA 94, 463–468. 24 Kinebuchi T, Kagawa W, Enomoto R, Tanaka K, Miyagawa K, Shibata T, Kurumizaka H & Yokoyama S (2004) Structural basis for octameric ring formation and DNA interaction of the human homologous-pairing protein Dmc1. Mol Cell 14, 363–374. 25 Passy SI, Yu X, Li Z, Radding CM, Masson JY, West SC & Egelman EH (1999) Human Dmc1 protein binds DNA as octameric ring. Proc Natl Acad Sci USA 96, 10684–10688. 26 Masson JY, Davies AA, Hajibagheri N, Van Dyck E, Benson FE, Stasiak AZ, Stasiak A & West SC (1999) The meiosis-specific recombinase hDmc1 forms ring structures and interacts with hRad51. EMBO J 18, 6552–6560. 27 Chang Lo YH, Lee MH, Leng CH, Hu SM, Chang CS & Wang TF (2005) Molecular Visualization of the Yeast Dmc1 Protein Ring and Dmc1-sDNA Nucleoprotein Complex. Biochemistry 44, 6052–6058. 28 Bugreev DV, Golub EI, Stasiak AZ, Stasiak A & Mazin AV (2005) Activation of human meiosis-specific recom- binase Dmc1 by Ca 2+ . J Biol Chem 280, 26886–26895. 29 Rao BJ, Chiu SK & Radding CM (1993) Homologous recognition and triplex formation promoted by RecA protein between duplex oligonucleotides and single- stranded DNA. J Mol Biol 229, 328–343. DNA strand exchange activity of OsDmc1 by FRET C. Rajanikant et al. 1506 FEBS Journal 273 (2006) 1497–1506 ª 2006 The Authors Journal compilation ª 2006 FEBS . DNA strand exchange activity of rice recombinase OsDmc1 monitored by fluorescence resonance energy transfer and the role of ATP hydrolysis Chittela. lM of OsDmc1 in presence of 2.0 mM ATP- c-S; (4) with 1.25 lM of OsDmc1 in presence of 2.0 m M of ATP. C. Rajanikant et al. DNA strand exchange activity of