- Email: [email protected]
Hyper-recombinogenic RecA Protein from Pseudomonas aeruginosa with Enhanced Activity of its Primary DNA Binding Site
doi:10.1016/S0022-2836(03)00242-0
J. Mol. Biol. (2003) 328, 1–7
COMMUNICATION
Hyper-recombinogenic RecA Protein from Pseudomonas aeruginosa with Enhanced Activity of its Primary DNA Binding Site Dmitry M. Baitin, Eugene N. Zaitsev and Vladislav A. Lanzov* Molecular Genetics Laboratory Division of Molecular and Radiation Biophysics B. P. Konstantinov Petersburg Nuclear Physics Institute Russian Academy of Sciences Gatchina, St. Petersburg 188350, Russian Federation
According to one prominent model, each protomer in the activated nucleoprotein filament of homologous recombinase RecA possesses two DNA-binding sites. The primary site binds (1) single-stranded DNA (ssDNA) to form presynaptic complex and (2) the newly formed doublestranded (ds) DNA whereas the secondary site binds (1) dsDNA of a partner to initiate strand exchange and (2) the displaced ssDNA following the strand exchange. RecA protein from Pseudomonas aeruginosa (RecAPa) promotes in Escherichia coli hyper-recombination in an SOS-independent manner. Earlier we revealed that RecAPa rapidly displaces E. coli SSB protein (SSB-Ec) from ssDNA to form presynaptic complex. Here we show that this property (1) is based on increased affinity of ssDNA for the RecAPa primary DNA binding site while the affinity for the secondary site remains similar to that for E. coli RecA, (2) is not specific for SSB-Ec but is also observed for SSB protein from P. aeruginosa that, in turn, predicts a possibility of enhanced recombination repair in this pathogenic bacterium. q 2003 Elsevier Science Ltd. All rights reserved
*Corresponding author
Keywords: P. aeruginosa; SSB protein; RecA protein; recombinogenic activity; DNA binding sites
Homologous genetic recombination in bacteria is catalysed by a filamentous DNA-transferase RecA which promotes two central steps of the process, the homologous DNA pairing and strand transfer.1 In the presence of ATP (one molecule per each protein subunit), RecA is polymerised on singlestranded DNA (ssDNA) to form a ternary complex RecA< ATP< ssDNA that can serve either the presynaptic helical filament for recombination or the allosteric co-protease for SOS regulon derepression.2,3 In recombination, the filament interacts with naked homologous double-stranded DNA (dsDNA) partner to make “the switch of Present address: E. N. Zaitsev, NICHD, NIH, 49 Convent Dr., Rm 6B71, Bethesda, MD 20892, USA. Abbreviations used: RecAEc and RecAPa, RecA proteins from E. coli and P. aeruginosa, respectively; ssDNA, single-stranded DNA; dsDNA, double-stranded DNA; SSB protein, single-stranded DNA-binding protein; SSB-Ec and SSB-Pa, SSB protein from E. coli and P. aeruginosa, respectively. E-mail address of the corresponding author: [email protected]
pairing” and to form one newly born heteroduplex DNA and one displaced strand. As originally proposed by Howard-Flanders et al.4 and then proven by Zlotnik et al.5 and Mazin & Kowalczykowski,6,7 each RecA molecule features, at least, two distinct DNA binding sites, the primary and the secondary. The former with its strong affinity for ssDNA is necessary in order to form the presynaptic filament. Several lines of evidence indicate that after the switch of pairing this site also binds the newly formed heteroduplex DNA.8 – 11 Thus the primary site should possess both the ssDNA and the dsDNA binding activity. Relatively weaker in its DNA binding ability,7 the secondary site binds both homologous and non-homologous dsDNA.12,13 When the presynaptic complex has been formed, this site serves in searching for sequence homology between ssDNA from this complex and the dsDNA of the recombining partner. This site also displays the affinity for ssDNA,5 because after the switch of pairing the displaced DNA strand temporarily stays in this site, before being removed by singlestrand DNA-binding (SSB) protein, to stabilise the
0022-2836/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved
2
Primary DNA Binding Site of P. aeruginosa RecA Protein
Table 1. Comparison of basic parameters of poly(dT)-dependent ATP hydrolysis RecAEc Parameters S05 a (mM) kcat a (min21) nH a Site sizeb (nucleotides)
RecAPa
10 mM MgCl2
1 mM MgCl2
10 mM MgCl2
1 mM MgCl2
76.0 ^ 9.5 29.7 ^ 4.9 13.0 ^ 1.2 3.0 ^ 0.2
142.0 ^ 12.0 21.0 ^ 2.6 11.5 ^ 0.6 5.6 ^ 0.6
71.5 ^ 6.9 33.4 ^ 3.5 12.9 ^ 1.4 3.1 ^ 0.3
77.5 ^ 6.3 31.7 ^ 3.9 12.4 ^ 0.8 3.1 ^ 0.5
a The data were calculated from the dependence of the rate of ATP hydrolysis (mM min21) on ATP concentration (mM). The ATP hydrolysis was monitored using a spectrophotometric assay that couples the production of ADP to the oxidation of NADH. The reactions were carried out at 37 8C in TD buffer (25 mM Tris –HCl (pH 7.5), 1 mM DTT) containing 1 mM RecA, 10 mM poly(dT), ATP regenerating system (phosphoenolpyruvate at a final concentration of 5 mM and 30 units ml21 pyruvate kinase), 0.56 mM NADH, 30 units ml21 lactate dehydrogenase, MgCl2 as indicated, and increasing concentration of ATP to obtain a maximum rate of hydrolysis. The data were averaged from two repeats. b The data were averaged from three repeats. One set of experiments is presented in Figure 1.
heteroduplex DNA formation and thus prevent the reversal of DNA strand exchange.7 Comparative to Escherichia coli RecA (RecAEc) protein, RecA from the opportunistic Gram-negative pathogen Pseudomonas aeruginosa (RecAPa) causes hyper-recombination in E. coli cells.14 Relatively infrequent recombination exchanges promoted by RecAEc in conjugational recombination15,16 (V.A.L., I. Bakhlanova & J. Clark, unpublished results) become 6.5-fold more frequent when RecAPa replaces RecAEc and, what is most important, this increase is not associated with an activation of SOS functions but directly results from properties of RecAPa.17 In search for differences in recombinogenic activities of RecAEc and RecAPa we found several distinguishing characteristics of the latter,14,17,18 one of which seems to be most important: an abnormally high affinity for ssDNA and, as a consequence, a quick displacement of SSB protein from ssDNA followed by rapid formation of presynaptic complex that involves new sites in recombination (for example, ssDNA gaps which exist during replication of donor or recipient DNA19). Here we continue the analysis of distinguishing characteristics of RecAPa in an attempt to answer two questions. What intrinsic property of RecAPa does predetermine the quick presynaptic complex formation? Is the improved ability to displace SSB protein from E. coli (SSB-Ec) a specific or a general characteristic of RecAPa? Basal characteristics of ssDNA-dependent ATP hydrolysis catalysed by presynaptic complexes formed with RecAEc and RecAPa proteins ATP is an obligatory component of the presynaptic complex RecA< ATP< ssDNA, which converts RecA molecules into a high recombination affinity state. Although ATP hydrolysis is not necessary for initiation of recombination20 and rather serves for its elongation or completion,21 it can be used to characterise the basic parameters of the presynaptic complex.
Three steady-state kinetic and one structural quantitative characteristics of ATP hydrolysis catalysed by RecAPa and RecAEc are summarised in Table 1. The former include values of S05 (the substrate concentration for the half-maximal observed rate of hydrolysis for enzyme – substrate systems that do not display classic Michaelis –Menten behaviour), kcat (the rate constant for the rate-determining step of a transition from a low to a highaffinity state), and nH (the Hill coefficient that serves as a measure of the degree of cooperativity for binding ATP to RecA during filament formation). As can be seen, these characteristics of ATP hydrolysis catalysed by either RecAPa or RecAEc are similar at 10 mM Mg2þ, which is optimal and 2.5-fold higher than the upper limits of free magnesium concentration in E. coli. (In fact, the total intracellular pool of this cation in E. coli stands at about 100 mM.22 Most of this is bound up in ribosomes, and the concentration of free magnesium available for RecA to use is thought to be from 1 mM to 4 mM.23,24) Moreover, RecAPa protein appeared to be similarly active at both optimal or suboptimal (1 mM) for E. coli amounts of magnesium whereas two parameters, S05 and kcat , demonstrated a reduced RecAEc ATPase activity at the suboptimal magnesium concentration. This finding has no obvious structural basis because two amino acid residues, Asp144 and Thr73, participating in magnesium binding25 are identical in both RecAPa and RecAEc. Besides, this finding does not seem to be important for realisation of a main RecA function in DNA strand transfer reaction because neither RecAEc nor RecAPa is able to produce joint molecules at a magnesium concentration lower than 4 mM (data not shown). As has been shown earlier,26 RecAEc protein possesses a very high cooperativity for binding ATP during the protein polymerisation on ssDNA. The cooperative binding is manifested by sigmoid curves presenting dependence of the rate of ATP hydrolysis on ATP concentration. This dependence was converted into a Hill plot, the slope of which, denoted nH ; gave the measure of cooperativity. Like RecAEc protein, RecAPa has a large nH value
3
Primary DNA Binding Site of P. aeruginosa RecA Protein
Figure 1. RecA/DNA binding isotherms monitored by ATP hydrolysis. Each point on the curves obtained represents an individual sample. The spectrophotometric assay was performed at 37 8C in TMD buffer (25 mM Tris – HCl (pH 7.5), 10 mM MgCl2, 1 mM dithiothreitol) containing 3 mM poly(dT), 1 mM ATP with its regenerating system, 0.56 mM NADH, 30 units ml21 lactate dehydrogenase, RecA as indicated, and either 10 mM MgCl2 (open symbols) or 1 mM MgCl2 (filled symbols). The curves with RecAEc are designated by triangles and those with RecAPa by circles. The preparation of RecAPa and RecAEc proteins as well as determination of their molar concentrations were as described earlier.14
(about 12) that suggests the functional unit for both proteins to be approximately 36 bases or, at least, two turns of RecA spiral filament.27 The structural parameter presented in Table 1 is a stoichiometry of RecA – ssDNA interaction or the number of nucleotide residues of ssDNA per RecA protomer (the site size). Using the ssDNAdependent ATPase to monitor binding between RecAEc and ssDNA, Zlotnick et al.5 have found that each RecA protomer has two non-equivalent DNA binding sites distinguished by their effects on ATPase activation. In fact, the first ssDNA binds RecA to form a presynaptic filament with a stoichiometry of about three bases per protomer and this is sufficient to fully activate the ATPase. Then, this filament can bind to the second ssDNA that results in final stoichiometry of about seven bases per protomer, but no additional ATPase activity is observed. Figure 1 shows typical examples of the RecA/ DNA binding isotherms for both RecAEc and RecAPa monitored by poly(dT)-dependent ATP hydrolysis. Extrapolation of linear portions of the isotherms (as shown by broken lines) demonstrates how estimations of RecA/ssDNA stoichiometry were done. As shown in Table 1, RecAPa and RecAEc interact with poly(dT) with identical stoichiometry. The only and expected exclusion was the RecAEc/poly(dT) stoichiometry at nonphysiological magnesium concentrations when RecAEc probably interacts with two DNA binding sites. In order to find distinguishing characteristics of RecAPa protein, the isotherms were measured for
a stronger substrate (dATP)28 and a weaker matrix for RecA polymerisation [poly(dA)]28 at optimal (10 mM) amount of co-factor MgCl2. The RecA/ poly(dA) stoichiometry in the presence of ATP was 5.6(^ 0.4) for RecAEc and 3.5(^ 0.3) for RecAPa, whereas in the presence of dATP this parameter was 2.9(^ 0.2) and 3.2(^ 0.4), respectively. In summary, poly(dA) at 10 mM MgCl2, like poly(dT) at 1 mM MgCl2, reduced the affinity of the RecAEc primary site for ssDNA. As expected, dATP improved the RecAEc primary site affinity for poly(dA). In turn, the RecAPa protein demonstrated the high affinity of its primary site for both strong poly(dT) and weak poly(dA) matrices at both optimal and suboptimal magnesium concentrations. The data suggest that the high affinity of the RecAPa primary site for ssDNA is the characteristic that distinguishes the RecAPa protein from RecAEc.
The primary DNA binding site of RecAPa has enhanced affinity for ssDNA In order to understand the molecular basis of the higher RecAPa affinity for ssDNA, it is necessary to have a quantitative estimation of how the primary and secondary sites of RecAPa and RecAEc proteins interact with ssDNA. In previous studies29,30 etheno-modified polynucleotides have been used successfully to show that RecA binding to etheno-DNA (eDNA) results in a proportional increase in fluorescence upon complex formation. Thus the fluorescent method is a direct way to compare the ssDNA affinity of different RecA structures. To examine the relative ssDNA binding affinities of the primary and the secondary sites in RecAPa and RecAEc filaments, the following findings made previously with RecAEc were used. (1) The primary and the secondary sites in the RecA nucleoprotein filament bind equal amount of ssDNA.7 Thus, being completely filled in, these sites are expected to increase equally the fluorescence intensity. (2) The primary site, which binds ssDNA with higher affinity, should be saturated preferentially in the presence of a limiting amount of ssDNA whereas a subsequent titration with additional ssDNA will result in binding to the secondary site.5,7 In other words, the excess of RecA relative to ssDNA results in binding to the primary site whereas the ssDNA in excess over RecA results in binding to both sites. (3) The stability (i.e. affinity) of ternary complexes RecA< ATP< ssDNA is decreasing with increasing NaCl concentration30 and thus the decrease can be used as a measure of the complex instability. (4) The ternary complex with ATPgS (the nearly nonhydrolysable ATP analogue) is extremely stable and cannot be dissociated by even 2.5 M NaCl.30 However, it remains unclear whether both sites or only the primary one, which is responsible for
4
Primary DNA Binding Site of P. aeruginosa RecA Protein
Figure 2. Sensitivity of the presynaptic complexes formed by RecAEc or RecAPa, ATP or ATPgS, and calf thymus eDNA to increasing concentration of NaCl that was monitored by change of the eDNA fluorescence: quenching of fluorescence resulted from the decay of the complexes accompanied by the transfer of eDNA from a bound to free state. Complexes with RecAPa and RecAEc are shown by filled and open symbols, respectively. Complexes with ATPgS or ATP are designated in the Figure. The value of 100% RecA< eDNA means a pure fluorescence of eDNA bound in the complex (see the text for details). A, The presynaptic complex RecA
ATPase activity5 and ATP binding,31 become extremely stable to NaCl in the presence of ATPgS. Taking into account all these findings, the dissociation of three types of RecAPa and RecAEc nucleoprotein filaments by means of increasing concentrations of NaCl was monitored by the eDNA fluorescence quenching that accompanies this dissociation. The first type of filaments, RecAPa< ATPgS< eDNA and RecAEc< ATPgS< eDNA, contained eDNA only in the primary sites of their protomers. The filaments were formed under conditions of excess RecA (2 mM) over eDNA (2 mM) in the presence of ATPgS. Because the RecAPa to ssDNA stoichiometry is one protomer to three bases, the ratio 1:1 between RecA and eDNA correctly provided the required excess of RecA protein. The results are displayed in Figure 2A, where 100% of RecA< eDNA complex means the fluorescence intensity (measured in arbitrary units) corrected for the linear increase in apparent fluorescence that occurs upon addition of RecA protein to the reaction mixture just before addition of eDNA. As seen, both RecAEc (as expected) and RecAPa filaments were stable in the presence of NaCl concentration as high as 1.4 M. The data indicate that in the presence of ATPgS the RecAPa protein, like RecAEc, forms a very stable complex with ssDNA through its primary site. Quite another situation was found in the second type of filaments, RecAPa< ATP< eDNA and RecAEc< ATP< eDNA, formed under the same
conditions of excess RecA (2 mM) over eDNA (2 mM) but in the presence of ATP and its regenerating system (Figure 2(A)). Judging by the quenching of fluorescence, the increase of NaCl concentration in the reaction mixture resulted in dissociation of both RecAPa and RecAEc complexes. Because in both cases the dissociation increases linearly with increase of NaCl concentration, it was easy to calculate that the presynaptic complex formed by RecAPa appeared to be twofold more stable than that of RecAEc. In other words, affinity of the RecAPa primary site for ssDNA is twofold greater than that of RecAEc. As was firstly supposed by Story et al.27 and then directly or indirectly supported by several studies,32,33 the mobile loop L2 of RecAEc structure contained the primary DNA binding site. Moreover, a 20 amino acid residue peptide including this L2 loop has been shown to form the homologous DNA pairing domain of RecA that also mediates the allosteric regulation of DNA binding and ATP hydrolysis.31,34 As a matter of fact, these peptides are identical in both the RecAEc and RecAPa structures35 and thus cannot be responsible for the strong difference observed in their affinities for ssDNA. It seems reasonable to suggest that the primary site is not limited only by the L2 loop but includes some other amino acids from the DNAbinding surface mapped by photochemical crosslinking between RecAEc protein and a 55-base ssDNA.36 The surface is formed by four known amino acid residues (Tyr65,36 Tyr103,36 Tyr264,36
5
Primary DNA Binding Site of P. aeruginosa RecA Protein
Figure 3. Time-course for SSB-Ec and SSB-Pa protein displacement by RecAPa: compared rates of poly(dT)dependent ATP hydrolysis. The spectrophotometric assay was performed at 37 8C in TMD buffer (25 mM Tris – HCl (pH 7.5), 10 mM MgCl2, 1 mM dithiothreitol) containing 3 mM poly(dT), 2 mM ATP with its regenerating system, 0.56 mM NADH, 30 units ml21 lactate dehydrogenase, 0.75 mM RecA, and 10 mM MgCl2. The SSB protein (0.6 mM) was added after preforming the mixture (zero on the time-scale). The SSB-Ec protein was from USB (USA). The SSB-Pa protein was kindly supplied to us by Dr C. Urbanke and collaborators.37 The SSB-Pa concentrations were determined spectrophotometrically using an absorption coefficient at 280 nm of 93,000 M21 cm21.
and Phe20332) and one unknown located between positions 178 and 183. Because these three Tyr residues and Phe203 are identical in both RecAEc and RecAPa, the difference discussed can be determined by any of three following amino acid changes: L178I, A179T and L182I between RecAEc and RecAPa located in the 178 – 183 region. It is interesting that other known substitutions in the two latter positions, A179V37 and L182E,38 have been described earlier as those related to regulation of SOS functions, and consequently the hyper-recombination too (for references, see Lanzov19). The third type of filaments, RecAPa< ATPgS< eDNA and RecAEc< ATPgS< eDNA, contained eDNA in both the primary and the secondary sites. These filaments were formed under conditions of excess eDNA (8 mM) over RecA (1 mM) in the presence of ATPgS. In order to calculate the corrected fluorescence generated only by binding RecA to eDNA, we should subtract from the measured fluorescence intensity not only that originating from RecA (as we did in previous experiments) but also that generated by the excess of unbound eDNA molecules. Two following findings should be taken into account to calculate the portion of eDNA bound to RecA through the first and the second sites. (1) The binding of calf thymus eDNA used in the study to RecAEc or RecAPa proteins resulted in increase of the specific fluorescence by a factor of 3.25 (data not shown).
(2) This increase has been found to be practically identical from binding to either the primary or the secondary site.5 As shown in Figure 2B, RecAPa and RecAEc filaments of the third type produce NaCl-dependent dissociation in a similar manner. The dissociation was more active than that observed from the primary site (Figure 2A), achieved plateau at 800 mM NaCl, and was limited by only a half of the bound eDNA. The latter makes obvious that the observed dissociation proceeds only at the expense of the portion of eDNA involved in the filaments through secondary sites. The data indicate: (1) as expected, the RecA secondary sites are much weaker than the primary ones in their affinity for ssDNA; (2) the affinity of both RecAPa and RecAEc secondary sites for ssDNA is similar; (3) the binding of ssDNA to the secondary site in the presence of ATPgS results in formation of the nucleoprotein complex sensitive to increasing concentrations of NaCl. The latter finding is in good accordance with an observation that after binding of a labelled ssDNA to the primary and the secondary sites in the presence of ATPgS the former differs from the latter by its resistance to increasing concentration of the challenger, unlabelled ssDNA.7 Like SSB-Ec, SSB-Pa cannot compete with RecAPa in binding to ssDNA The SSB-Ec protein easily displaces RecAEc from poly(dT). This observation is part of a more general characteristic of RecAEc filament formed on linear ssDNA: the 50 end-dependent dissociation of RecA protomers from the DNA.39 This dissociation requires ATP hydrolysis, and in the presence of SSB protein is largely irreversible because the dissociated RecA molecules are replaced by SSB. The described process reflects a fine balance between the affinities of these proteins for ssDNA that provides a relatively low recombinogenic activity of RecAEc in vivo. The SSB-Ec practically cannot displace RecAPa from poly(dT).18 The question arises; which situation is realised in the case of competition for ssDNA between RecAPa and its native SSB protein, SSB-Pa?40 Figure 3 shows that both SSB-Pa and SSB-Ec are very similar in their ability to displace RecAEc from the presynaptic ternary complex RecAEc< ATP< poly(dT), but they are identically ineffective to displace RecAPa under the same conditions. The same similarity between SSB-Ec and SSB-Pa activities was found in experiments with displacement of these proteins by RecAPa from a circular ssDNA of bacteriophage M13 as was monitored by growth of ATPase activity of the presynaptic complex formed (data not shown). The data suggest that relationships between RecAPa and SSB-Ec, when these proteins compete for ssDNA, is not specific because the RecAPa also displaces from ssDNA its native partner, SSB-Pa. In turn, this observation suggests that RecAPa
6
Primary DNA Binding Site of P. aeruginosa RecA Protein
protein could provide to its pathogenic host a significantly more active mechanism of recombination repair than that acting in E. coli by means of RecAEc.
12. Takahashi, M., Kubista, M. & Norden, B. (1989). Binding stoichiometry and structure of RecA– DNA complexes studied by flow linear dichroism and fluorescence spectroscopy. Evidence for multiple heterogeneous DNA co-ordination. J. Mol. Biol. 205, 137– 147. 13. Muller, B., Koller, T. & Stasiak, A. (1990). Characterization of the DNA binding activity of stable RecA – DNA complexes. Interaction between the two DNA binding sites within RecA helical filaments. J. Mol. Biol. 212, 97 – 112. 14. Namsaraev, E. A., Baitin, D. M., Bakhlanova, I. V., Alexseyev, A. A., Ogawa, H. & Lanzov, V. A. (1998). Biochemical basis of hyper-recombinogenic activity of Pseudomonas aeruginosa RecA protein in Escherichia coli cells. Mol. Microbiol. 27, 727– 738. 15. Wood, T. H. & Walmsley, R. H. (1969). Conjugation in Escherichia coli K-12 and its modification by irradiation. Biophys. J. 9, 391– 420. 16. Bresler, S. E., Krivonogov, S. V. & Lanzov, V. A. (1978). Scale of the genetic map and genetic control of recombination after conjugation in Escherichia coli K-12. Mol. Gen. Genet. 166, 337– 346. 17. Bakhlanova, I. V., Ogawa, T. & Lanzov, V. A. (2001). Recombinogenic activity of chimeric recA genes (Pseudomonas aeruginosa/Escherichia coli): a search for RecA protein regions responsible for this activity. Genetics, 159, 7 – 15. 18. Chervyakova, D., Kagansky, A., Petukhov, M. & Lanzov, V. (2001). [L29M] substitution in the interface of subunit –subunit interactions enhances Escherichia coli RecA protein properties important for its recombinogenic activity. J. Mol. Biol. 314, 923– 935. 19. Lanzov, V. A. (2002). Hyper-recombination in Escherichia coli with and without SOS response. In Recent Research Development in DNA Repair and Mutagenesis (Ruiz-Rubio, M., Alejandre-Duran, E. & Roldan-Arjona, T., eds), pp. 21 – 38, Research Signpost, Kerala, India. 20. Kowalczykowski, S. C. & Krupp, R. A. (1995). DNAstrand exchange promoted by RecA protein in the absence of ATP: implications for the mechanism of energy transduction in protein-promoted nucleic acid transactions. Proc. Natl Acad. Sci. USA, 92, 3478– 3482. 21. Cox, M. M. (1994). Why does RecA protein hydrolyze ATP? Trends Biochem. Sci. 19, 217– 222. 22. Moncany, M. L. J. & Kellenberger, E. (1981). High magnesium content of Escherichia coli B. Experimentia, 37, 846– 847. 23. Lusk, J. E., Williams, J. P. & Kennedy, E. P. (1968). Magnesium and the growth of Escherichia coli. J. Biol. Chem. 243, 2618– 2624. 24. Alatossava, T., Ju¨tte, H., Kuhn, A. & Kellenberger, E. (1985). Manipulation of intracellular magnesium content in polymyxin B nonapeptide-sentisized Escherichia coli by ionophore A23187. J. Bacteriol. 162, 413– 419. 25. Story, R. M. & Steitz, T. A. (1992). Structure of the RecA protein – ADP complex. Nature, 355, 374– 376. 26. Mikawa, T., Masui, R. & Kuramitsu, S. (1998). RecA protein has extremely high cooperativity for substrate in its ATPase activity. J. Biochem. 123, 450– 457. 27. Story, R. M., Weber, I. T. & Steitz, T. A. (1992). The structure of the E. coli recA protein monomer and polymer. Nature, 355, 318– 324. 28. Menetski, J. P. & Kowalczykowsky, S. C. (1989). Enhancement of Escherichia coli RecA protein enzymatic function by dATP. Biochemistry, 28, 5871– 5881.
Acknowledgements We are grateful to Drs J. Genschel, L. Litz, C. Urbanke and R. Hilgenfeld for kindly supplying us with the SSB-Pa protein used in the study and Dr A. Alexseev for providing etheno-modified DNA. We thank also Dr L. Firsov and unknown reviewers for valuable comments and Dr L. Bresler for English correction. This work was supported by a Fogarty International Research Collaboration Award (grant 1 R03 TWO1319-01A1) and Russian Foundation for Basic Research (grant 02-04-48332).
References 1. Kowalczykowski, S. C. (2000). Initiation of genetic recombination and recombination-dependent replication. Trends Biochem. Sci. 35, 156– 165. 2. Kowalczykowski, S. C., Dixon, D. A., Eggleston, A. K., Lauder, S. D. & Rehrauer, W. M. (1994). Biochemistry of homologous recombination in Escherichia coli. Microbiol. Rev. 58, 401– 465. 3. Roca, A. I. & Cox, M. M. (1997). RecA protein: structure, function, and role in recombinational DNA repair. Prog. Nucl. Acid Res. Mol. Biol. 56, 129– 223. 4. Howard-Flanders, P., West, S. C., Rusche, J. R. & Egelman, E. H. (1984). Molecular mechanisms of general genetic recombination: the DNA-binding sites of RecA protein. Cold Spring Harbor Symp. Quant. Biol. 49, 571– 580. 5. Zlotnick, A., Mitchell, R. S., Steed, R. K. & Brenner, S. L. (1993). Analysis of two distinct single-stranded DNA binding sites on the RecA nucleoprotein filament. J. Biol. Chem. 268, 22525– 22530. 6. Mazin, A. V. & Kowalczykowski, S. C. (1996). The specificity of the secondary DNA binding site of RecA protein defines its role in DNA strand exchange. Proc. Natl Acad. Sci. USA, 93, 10673– 10678. 7. Mazin, A. V. & Kowalczykowski, S. C. (1998). The function of the secondary DNA-binding site of RecA protein during DNA strand exchange. EMBO J. 17, 1161 –1168. 8. Chow, S. A., Honigberg, S. M., Bainton, R. J. & Radding, C. M. (1986). Patterns of nuclease protection during strand exchange: RecA protein forms heteroduplex DNA by binding to strands of like polarity. J. Biol. Chem. 261, 6961– 6971. 9. Pugh, B. F. & Cox, M. M. (1990). RecA protein binding to the heteroduplex product of DNA strand exchange. J. Biol. Chem. 262, 1337– 1343. 10. Rosselli, W. & Stasiak, A. (1990). Energetics of RecAmediated recombination reaction: without ATP hydrolysis RecA can mediate polar strand exchange but is unable to recycle. J. Mol. Biol. 216, 335– 352. 11. Ulsperger, C. J. & Cox, M. M. (1995). Quantitative RecA protein binding to the hybrid duplex product of DNA strand exchange. Biochemistry, 34, 10859 –10866.
Primary DNA Binding Site of P. aeruginosa RecA Protein
7
29. McEntee, K., Weinstock, G. M. & Lehman, I. R. (1981). Binding of the recA protein of Escherichia coli to single- and double-stranded DNA. J. Biol. Chem. 256, 8835– 8844. 30. Menetski, J. P. & Kowalczykowski, S. C. (1985). Interaction of recA protein with single-stranded DNA. Quantitative aspects of binding affinity modulation by nucleotide cofactors. J. Mol. Biol. 181, 281– 295. 31. Voloshin, O. N., Wang, L. & Camerini-Otero, R. D. (2000). The homologous pairing domain of RecA also mediates the allosteric regulation of DNA binding and ATP hydrolysis: a remarkable concentration of functional residues. J. Mol. Biol. 303, 709– 720. 32. Malkov, V. A. & Camerini-Otero, R. D. (1995). Photocross-links between single-stranded DNA and Escherichia coli RecA protein map to loops L1 (amino acid residues 157– 164) and L2 (amino acid residues 195–209). J. Biol. Chem. 270, 30230– 30233. 33. Maraboeuf, F., Voloshin, O., Camerini-Otero, R. D. & Takahashi, M. (1995). The central aromatic residue in loop L2 of RecA interacts with DNA. Quenching of the fluorescence of a tryptophan reporter inserted in L2 upon binding to DNA. J. Biol. Chem. 270, 30927– 30932. 34. Voloshin, O. N., Wang, L. & Camerini-Otero, R. D. (1996). Homologous DNA pairing promoted by a 20-amino acid peptide derived from RecA. Science, 272, 868– 872.
35. Roca, A. I. & Cox, M. M. (1990). The RecA protein: structure and function. Crit. Rev. Biochem. Mol. Biol. 25, 415–456. 36. Morimatsu, K. & Horii, T. (1995). DNA-binding surface of RecA protein. Photochemical cross-linking of the first DNA binding site on RecA filament. Eur. J. Biochem. 234, 695–705. 37. Wang, W.-B. & Tessman, E. S. (1996). Location of functional regions of the Escherichia coli RecA protein by DNA sequence analysis of RecA protease-constitutive mutants. J. Bacteriol. 168, 901– 910. 38. Liu, S-K., Eisen, J. A., Hanawalt, P. C. & Tessman, I. (1993). recA mutations that reduce the constitutive coprotease activity of the RecA1202 (Prtc) protein: possible involvement of interfilament association in proteolytic and recombination activities. J. Bacteriol. 175, 6518– 6529. 39. Shan, Q., Bork, J. M., Webb, B. L., Inman, R. B. & Cox, M. M. (1997). RecA protein filaments: end-dependent dissociation from ssDNA and stabilization by RecO and RecR proteins. J. Mol. Biol. 265, 519– 540. 40. Genschel, J., Litz, L., Thole, H., Roemling, U. & Urbanke, C. (1996). Isolation, sequencing and overproduction of the single-stranded DNA binding protein from Pseudomonas aeruginosa PAO. Gene, 182, 137 –143.
Edited by J. Karn (Received 3 December 2002; received in revised form 7 February 2003; accepted 10 February 2003)
J. Mol. Biol. (2003) 328, 1–7
COMMUNICATION
Hyper-recombinogenic RecA Protein from Pseudomonas aeruginosa with Enhanced Activity of its Primary DNA Binding Site Dmitry M. Baitin, Eugene N. Zaitsev and Vladislav A. Lanzov* Molecular Genetics Laboratory Division of Molecular and Radiation Biophysics B. P. Konstantinov Petersburg Nuclear Physics Institute Russian Academy of Sciences Gatchina, St. Petersburg 188350, Russian Federation
According to one prominent model, each protomer in the activated nucleoprotein filament of homologous recombinase RecA possesses two DNA-binding sites. The primary site binds (1) single-stranded DNA (ssDNA) to form presynaptic complex and (2) the newly formed doublestranded (ds) DNA whereas the secondary site binds (1) dsDNA of a partner to initiate strand exchange and (2) the displaced ssDNA following the strand exchange. RecA protein from Pseudomonas aeruginosa (RecAPa) promotes in Escherichia coli hyper-recombination in an SOS-independent manner. Earlier we revealed that RecAPa rapidly displaces E. coli SSB protein (SSB-Ec) from ssDNA to form presynaptic complex. Here we show that this property (1) is based on increased affinity of ssDNA for the RecAPa primary DNA binding site while the affinity for the secondary site remains similar to that for E. coli RecA, (2) is not specific for SSB-Ec but is also observed for SSB protein from P. aeruginosa that, in turn, predicts a possibility of enhanced recombination repair in this pathogenic bacterium. q 2003 Elsevier Science Ltd. All rights reserved
*Corresponding author
Keywords: P. aeruginosa; SSB protein; RecA protein; recombinogenic activity; DNA binding sites
Homologous genetic recombination in bacteria is catalysed by a filamentous DNA-transferase RecA which promotes two central steps of the process, the homologous DNA pairing and strand transfer.1 In the presence of ATP (one molecule per each protein subunit), RecA is polymerised on singlestranded DNA (ssDNA) to form a ternary complex RecA< ATP< ssDNA that can serve either the presynaptic helical filament for recombination or the allosteric co-protease for SOS regulon derepression.2,3 In recombination, the filament interacts with naked homologous double-stranded DNA (dsDNA) partner to make “the switch of Present address: E. N. Zaitsev, NICHD, NIH, 49 Convent Dr., Rm 6B71, Bethesda, MD 20892, USA. Abbreviations used: RecAEc and RecAPa, RecA proteins from E. coli and P. aeruginosa, respectively; ssDNA, single-stranded DNA; dsDNA, double-stranded DNA; SSB protein, single-stranded DNA-binding protein; SSB-Ec and SSB-Pa, SSB protein from E. coli and P. aeruginosa, respectively. E-mail address of the corresponding author: [email protected]
pairing” and to form one newly born heteroduplex DNA and one displaced strand. As originally proposed by Howard-Flanders et al.4 and then proven by Zlotnik et al.5 and Mazin & Kowalczykowski,6,7 each RecA molecule features, at least, two distinct DNA binding sites, the primary and the secondary. The former with its strong affinity for ssDNA is necessary in order to form the presynaptic filament. Several lines of evidence indicate that after the switch of pairing this site also binds the newly formed heteroduplex DNA.8 – 11 Thus the primary site should possess both the ssDNA and the dsDNA binding activity. Relatively weaker in its DNA binding ability,7 the secondary site binds both homologous and non-homologous dsDNA.12,13 When the presynaptic complex has been formed, this site serves in searching for sequence homology between ssDNA from this complex and the dsDNA of the recombining partner. This site also displays the affinity for ssDNA,5 because after the switch of pairing the displaced DNA strand temporarily stays in this site, before being removed by singlestrand DNA-binding (SSB) protein, to stabilise the
0022-2836/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved
2
Primary DNA Binding Site of P. aeruginosa RecA Protein
Table 1. Comparison of basic parameters of poly(dT)-dependent ATP hydrolysis RecAEc Parameters S05 a (mM) kcat a (min21) nH a Site sizeb (nucleotides)
RecAPa
10 mM MgCl2
1 mM MgCl2
10 mM MgCl2
1 mM MgCl2
76.0 ^ 9.5 29.7 ^ 4.9 13.0 ^ 1.2 3.0 ^ 0.2
142.0 ^ 12.0 21.0 ^ 2.6 11.5 ^ 0.6 5.6 ^ 0.6
71.5 ^ 6.9 33.4 ^ 3.5 12.9 ^ 1.4 3.1 ^ 0.3
77.5 ^ 6.3 31.7 ^ 3.9 12.4 ^ 0.8 3.1 ^ 0.5
a The data were calculated from the dependence of the rate of ATP hydrolysis (mM min21) on ATP concentration (mM). The ATP hydrolysis was monitored using a spectrophotometric assay that couples the production of ADP to the oxidation of NADH. The reactions were carried out at 37 8C in TD buffer (25 mM Tris –HCl (pH 7.5), 1 mM DTT) containing 1 mM RecA, 10 mM poly(dT), ATP regenerating system (phosphoenolpyruvate at a final concentration of 5 mM and 30 units ml21 pyruvate kinase), 0.56 mM NADH, 30 units ml21 lactate dehydrogenase, MgCl2 as indicated, and increasing concentration of ATP to obtain a maximum rate of hydrolysis. The data were averaged from two repeats. b The data were averaged from three repeats. One set of experiments is presented in Figure 1.
heteroduplex DNA formation and thus prevent the reversal of DNA strand exchange.7 Comparative to Escherichia coli RecA (RecAEc) protein, RecA from the opportunistic Gram-negative pathogen Pseudomonas aeruginosa (RecAPa) causes hyper-recombination in E. coli cells.14 Relatively infrequent recombination exchanges promoted by RecAEc in conjugational recombination15,16 (V.A.L., I. Bakhlanova & J. Clark, unpublished results) become 6.5-fold more frequent when RecAPa replaces RecAEc and, what is most important, this increase is not associated with an activation of SOS functions but directly results from properties of RecAPa.17 In search for differences in recombinogenic activities of RecAEc and RecAPa we found several distinguishing characteristics of the latter,14,17,18 one of which seems to be most important: an abnormally high affinity for ssDNA and, as a consequence, a quick displacement of SSB protein from ssDNA followed by rapid formation of presynaptic complex that involves new sites in recombination (for example, ssDNA gaps which exist during replication of donor or recipient DNA19). Here we continue the analysis of distinguishing characteristics of RecAPa in an attempt to answer two questions. What intrinsic property of RecAPa does predetermine the quick presynaptic complex formation? Is the improved ability to displace SSB protein from E. coli (SSB-Ec) a specific or a general characteristic of RecAPa? Basal characteristics of ssDNA-dependent ATP hydrolysis catalysed by presynaptic complexes formed with RecAEc and RecAPa proteins ATP is an obligatory component of the presynaptic complex RecA< ATP< ssDNA, which converts RecA molecules into a high recombination affinity state. Although ATP hydrolysis is not necessary for initiation of recombination20 and rather serves for its elongation or completion,21 it can be used to characterise the basic parameters of the presynaptic complex.
Three steady-state kinetic and one structural quantitative characteristics of ATP hydrolysis catalysed by RecAPa and RecAEc are summarised in Table 1. The former include values of S05 (the substrate concentration for the half-maximal observed rate of hydrolysis for enzyme – substrate systems that do not display classic Michaelis –Menten behaviour), kcat (the rate constant for the rate-determining step of a transition from a low to a highaffinity state), and nH (the Hill coefficient that serves as a measure of the degree of cooperativity for binding ATP to RecA during filament formation). As can be seen, these characteristics of ATP hydrolysis catalysed by either RecAPa or RecAEc are similar at 10 mM Mg2þ, which is optimal and 2.5-fold higher than the upper limits of free magnesium concentration in E. coli. (In fact, the total intracellular pool of this cation in E. coli stands at about 100 mM.22 Most of this is bound up in ribosomes, and the concentration of free magnesium available for RecA to use is thought to be from 1 mM to 4 mM.23,24) Moreover, RecAPa protein appeared to be similarly active at both optimal or suboptimal (1 mM) for E. coli amounts of magnesium whereas two parameters, S05 and kcat , demonstrated a reduced RecAEc ATPase activity at the suboptimal magnesium concentration. This finding has no obvious structural basis because two amino acid residues, Asp144 and Thr73, participating in magnesium binding25 are identical in both RecAPa and RecAEc. Besides, this finding does not seem to be important for realisation of a main RecA function in DNA strand transfer reaction because neither RecAEc nor RecAPa is able to produce joint molecules at a magnesium concentration lower than 4 mM (data not shown). As has been shown earlier,26 RecAEc protein possesses a very high cooperativity for binding ATP during the protein polymerisation on ssDNA. The cooperative binding is manifested by sigmoid curves presenting dependence of the rate of ATP hydrolysis on ATP concentration. This dependence was converted into a Hill plot, the slope of which, denoted nH ; gave the measure of cooperativity. Like RecAEc protein, RecAPa has a large nH value
3
Primary DNA Binding Site of P. aeruginosa RecA Protein
Figure 1. RecA/DNA binding isotherms monitored by ATP hydrolysis. Each point on the curves obtained represents an individual sample. The spectrophotometric assay was performed at 37 8C in TMD buffer (25 mM Tris – HCl (pH 7.5), 10 mM MgCl2, 1 mM dithiothreitol) containing 3 mM poly(dT), 1 mM ATP with its regenerating system, 0.56 mM NADH, 30 units ml21 lactate dehydrogenase, RecA as indicated, and either 10 mM MgCl2 (open symbols) or 1 mM MgCl2 (filled symbols). The curves with RecAEc are designated by triangles and those with RecAPa by circles. The preparation of RecAPa and RecAEc proteins as well as determination of their molar concentrations were as described earlier.14
(about 12) that suggests the functional unit for both proteins to be approximately 36 bases or, at least, two turns of RecA spiral filament.27 The structural parameter presented in Table 1 is a stoichiometry of RecA – ssDNA interaction or the number of nucleotide residues of ssDNA per RecA protomer (the site size). Using the ssDNAdependent ATPase to monitor binding between RecAEc and ssDNA, Zlotnick et al.5 have found that each RecA protomer has two non-equivalent DNA binding sites distinguished by their effects on ATPase activation. In fact, the first ssDNA binds RecA to form a presynaptic filament with a stoichiometry of about three bases per protomer and this is sufficient to fully activate the ATPase. Then, this filament can bind to the second ssDNA that results in final stoichiometry of about seven bases per protomer, but no additional ATPase activity is observed. Figure 1 shows typical examples of the RecA/ DNA binding isotherms for both RecAEc and RecAPa monitored by poly(dT)-dependent ATP hydrolysis. Extrapolation of linear portions of the isotherms (as shown by broken lines) demonstrates how estimations of RecA/ssDNA stoichiometry were done. As shown in Table 1, RecAPa and RecAEc interact with poly(dT) with identical stoichiometry. The only and expected exclusion was the RecAEc/poly(dT) stoichiometry at nonphysiological magnesium concentrations when RecAEc probably interacts with two DNA binding sites. In order to find distinguishing characteristics of RecAPa protein, the isotherms were measured for
a stronger substrate (dATP)28 and a weaker matrix for RecA polymerisation [poly(dA)]28 at optimal (10 mM) amount of co-factor MgCl2. The RecA/ poly(dA) stoichiometry in the presence of ATP was 5.6(^ 0.4) for RecAEc and 3.5(^ 0.3) for RecAPa, whereas in the presence of dATP this parameter was 2.9(^ 0.2) and 3.2(^ 0.4), respectively. In summary, poly(dA) at 10 mM MgCl2, like poly(dT) at 1 mM MgCl2, reduced the affinity of the RecAEc primary site for ssDNA. As expected, dATP improved the RecAEc primary site affinity for poly(dA). In turn, the RecAPa protein demonstrated the high affinity of its primary site for both strong poly(dT) and weak poly(dA) matrices at both optimal and suboptimal magnesium concentrations. The data suggest that the high affinity of the RecAPa primary site for ssDNA is the characteristic that distinguishes the RecAPa protein from RecAEc.
The primary DNA binding site of RecAPa has enhanced affinity for ssDNA In order to understand the molecular basis of the higher RecAPa affinity for ssDNA, it is necessary to have a quantitative estimation of how the primary and secondary sites of RecAPa and RecAEc proteins interact with ssDNA. In previous studies29,30 etheno-modified polynucleotides have been used successfully to show that RecA binding to etheno-DNA (eDNA) results in a proportional increase in fluorescence upon complex formation. Thus the fluorescent method is a direct way to compare the ssDNA affinity of different RecA structures. To examine the relative ssDNA binding affinities of the primary and the secondary sites in RecAPa and RecAEc filaments, the following findings made previously with RecAEc were used. (1) The primary and the secondary sites in the RecA nucleoprotein filament bind equal amount of ssDNA.7 Thus, being completely filled in, these sites are expected to increase equally the fluorescence intensity. (2) The primary site, which binds ssDNA with higher affinity, should be saturated preferentially in the presence of a limiting amount of ssDNA whereas a subsequent titration with additional ssDNA will result in binding to the secondary site.5,7 In other words, the excess of RecA relative to ssDNA results in binding to the primary site whereas the ssDNA in excess over RecA results in binding to both sites. (3) The stability (i.e. affinity) of ternary complexes RecA< ATP< ssDNA is decreasing with increasing NaCl concentration30 and thus the decrease can be used as a measure of the complex instability. (4) The ternary complex with ATPgS (the nearly nonhydrolysable ATP analogue) is extremely stable and cannot be dissociated by even 2.5 M NaCl.30 However, it remains unclear whether both sites or only the primary one, which is responsible for
4
Primary DNA Binding Site of P. aeruginosa RecA Protein
Figure 2. Sensitivity of the presynaptic complexes formed by RecAEc or RecAPa, ATP or ATPgS, and calf thymus eDNA to increasing concentration of NaCl that was monitored by change of the eDNA fluorescence: quenching of fluorescence resulted from the decay of the complexes accompanied by the transfer of eDNA from a bound to free state. Complexes with RecAPa and RecAEc are shown by filled and open symbols, respectively. Complexes with ATPgS or ATP are designated in the Figure. The value of 100% RecA< eDNA means a pure fluorescence of eDNA bound in the complex (see the text for details). A, The presynaptic complex RecA
ATPase activity5 and ATP binding,31 become extremely stable to NaCl in the presence of ATPgS. Taking into account all these findings, the dissociation of three types of RecAPa and RecAEc nucleoprotein filaments by means of increasing concentrations of NaCl was monitored by the eDNA fluorescence quenching that accompanies this dissociation. The first type of filaments, RecAPa< ATPgS< eDNA and RecAEc< ATPgS< eDNA, contained eDNA only in the primary sites of their protomers. The filaments were formed under conditions of excess RecA (2 mM) over eDNA (2 mM) in the presence of ATPgS. Because the RecAPa to ssDNA stoichiometry is one protomer to three bases, the ratio 1:1 between RecA and eDNA correctly provided the required excess of RecA protein. The results are displayed in Figure 2A, where 100% of RecA< eDNA complex means the fluorescence intensity (measured in arbitrary units) corrected for the linear increase in apparent fluorescence that occurs upon addition of RecA protein to the reaction mixture just before addition of eDNA. As seen, both RecAEc (as expected) and RecAPa filaments were stable in the presence of NaCl concentration as high as 1.4 M. The data indicate that in the presence of ATPgS the RecAPa protein, like RecAEc, forms a very stable complex with ssDNA through its primary site. Quite another situation was found in the second type of filaments, RecAPa< ATP< eDNA and RecAEc< ATP< eDNA, formed under the same
conditions of excess RecA (2 mM) over eDNA (2 mM) but in the presence of ATP and its regenerating system (Figure 2(A)). Judging by the quenching of fluorescence, the increase of NaCl concentration in the reaction mixture resulted in dissociation of both RecAPa and RecAEc complexes. Because in both cases the dissociation increases linearly with increase of NaCl concentration, it was easy to calculate that the presynaptic complex formed by RecAPa appeared to be twofold more stable than that of RecAEc. In other words, affinity of the RecAPa primary site for ssDNA is twofold greater than that of RecAEc. As was firstly supposed by Story et al.27 and then directly or indirectly supported by several studies,32,33 the mobile loop L2 of RecAEc structure contained the primary DNA binding site. Moreover, a 20 amino acid residue peptide including this L2 loop has been shown to form the homologous DNA pairing domain of RecA that also mediates the allosteric regulation of DNA binding and ATP hydrolysis.31,34 As a matter of fact, these peptides are identical in both the RecAEc and RecAPa structures35 and thus cannot be responsible for the strong difference observed in their affinities for ssDNA. It seems reasonable to suggest that the primary site is not limited only by the L2 loop but includes some other amino acids from the DNAbinding surface mapped by photochemical crosslinking between RecAEc protein and a 55-base ssDNA.36 The surface is formed by four known amino acid residues (Tyr65,36 Tyr103,36 Tyr264,36
5
Primary DNA Binding Site of P. aeruginosa RecA Protein
Figure 3. Time-course for SSB-Ec and SSB-Pa protein displacement by RecAPa: compared rates of poly(dT)dependent ATP hydrolysis. The spectrophotometric assay was performed at 37 8C in TMD buffer (25 mM Tris – HCl (pH 7.5), 10 mM MgCl2, 1 mM dithiothreitol) containing 3 mM poly(dT), 2 mM ATP with its regenerating system, 0.56 mM NADH, 30 units ml21 lactate dehydrogenase, 0.75 mM RecA, and 10 mM MgCl2. The SSB protein (0.6 mM) was added after preforming the mixture (zero on the time-scale). The SSB-Ec protein was from USB (USA). The SSB-Pa protein was kindly supplied to us by Dr C. Urbanke and collaborators.37 The SSB-Pa concentrations were determined spectrophotometrically using an absorption coefficient at 280 nm of 93,000 M21 cm21.
and Phe20332) and one unknown located between positions 178 and 183. Because these three Tyr residues and Phe203 are identical in both RecAEc and RecAPa, the difference discussed can be determined by any of three following amino acid changes: L178I, A179T and L182I between RecAEc and RecAPa located in the 178 – 183 region. It is interesting that other known substitutions in the two latter positions, A179V37 and L182E,38 have been described earlier as those related to regulation of SOS functions, and consequently the hyper-recombination too (for references, see Lanzov19). The third type of filaments, RecAPa< ATPgS< eDNA and RecAEc< ATPgS< eDNA, contained eDNA in both the primary and the secondary sites. These filaments were formed under conditions of excess eDNA (8 mM) over RecA (1 mM) in the presence of ATPgS. In order to calculate the corrected fluorescence generated only by binding RecA to eDNA, we should subtract from the measured fluorescence intensity not only that originating from RecA (as we did in previous experiments) but also that generated by the excess of unbound eDNA molecules. Two following findings should be taken into account to calculate the portion of eDNA bound to RecA through the first and the second sites. (1) The binding of calf thymus eDNA used in the study to RecAEc or RecAPa proteins resulted in increase of the specific fluorescence by a factor of 3.25 (data not shown).
(2) This increase has been found to be practically identical from binding to either the primary or the secondary site.5 As shown in Figure 2B, RecAPa and RecAEc filaments of the third type produce NaCl-dependent dissociation in a similar manner. The dissociation was more active than that observed from the primary site (Figure 2A), achieved plateau at 800 mM NaCl, and was limited by only a half of the bound eDNA. The latter makes obvious that the observed dissociation proceeds only at the expense of the portion of eDNA involved in the filaments through secondary sites. The data indicate: (1) as expected, the RecA secondary sites are much weaker than the primary ones in their affinity for ssDNA; (2) the affinity of both RecAPa and RecAEc secondary sites for ssDNA is similar; (3) the binding of ssDNA to the secondary site in the presence of ATPgS results in formation of the nucleoprotein complex sensitive to increasing concentrations of NaCl. The latter finding is in good accordance with an observation that after binding of a labelled ssDNA to the primary and the secondary sites in the presence of ATPgS the former differs from the latter by its resistance to increasing concentration of the challenger, unlabelled ssDNA.7 Like SSB-Ec, SSB-Pa cannot compete with RecAPa in binding to ssDNA The SSB-Ec protein easily displaces RecAEc from poly(dT). This observation is part of a more general characteristic of RecAEc filament formed on linear ssDNA: the 50 end-dependent dissociation of RecA protomers from the DNA.39 This dissociation requires ATP hydrolysis, and in the presence of SSB protein is largely irreversible because the dissociated RecA molecules are replaced by SSB. The described process reflects a fine balance between the affinities of these proteins for ssDNA that provides a relatively low recombinogenic activity of RecAEc in vivo. The SSB-Ec practically cannot displace RecAPa from poly(dT).18 The question arises; which situation is realised in the case of competition for ssDNA between RecAPa and its native SSB protein, SSB-Pa?40 Figure 3 shows that both SSB-Pa and SSB-Ec are very similar in their ability to displace RecAEc from the presynaptic ternary complex RecAEc< ATP< poly(dT), but they are identically ineffective to displace RecAPa under the same conditions. The same similarity between SSB-Ec and SSB-Pa activities was found in experiments with displacement of these proteins by RecAPa from a circular ssDNA of bacteriophage M13 as was monitored by growth of ATPase activity of the presynaptic complex formed (data not shown). The data suggest that relationships between RecAPa and SSB-Ec, when these proteins compete for ssDNA, is not specific because the RecAPa also displaces from ssDNA its native partner, SSB-Pa. In turn, this observation suggests that RecAPa
6
Primary DNA Binding Site of P. aeruginosa RecA Protein
protein could provide to its pathogenic host a significantly more active mechanism of recombination repair than that acting in E. coli by means of RecAEc.
12. Takahashi, M., Kubista, M. & Norden, B. (1989). Binding stoichiometry and structure of RecA– DNA complexes studied by flow linear dichroism and fluorescence spectroscopy. Evidence for multiple heterogeneous DNA co-ordination. J. Mol. Biol. 205, 137– 147. 13. Muller, B., Koller, T. & Stasiak, A. (1990). Characterization of the DNA binding activity of stable RecA – DNA complexes. Interaction between the two DNA binding sites within RecA helical filaments. J. Mol. Biol. 212, 97 – 112. 14. Namsaraev, E. A., Baitin, D. M., Bakhlanova, I. V., Alexseyev, A. A., Ogawa, H. & Lanzov, V. A. (1998). Biochemical basis of hyper-recombinogenic activity of Pseudomonas aeruginosa RecA protein in Escherichia coli cells. Mol. Microbiol. 27, 727– 738. 15. Wood, T. H. & Walmsley, R. H. (1969). Conjugation in Escherichia coli K-12 and its modification by irradiation. Biophys. J. 9, 391– 420. 16. Bresler, S. E., Krivonogov, S. V. & Lanzov, V. A. (1978). Scale of the genetic map and genetic control of recombination after conjugation in Escherichia coli K-12. Mol. Gen. Genet. 166, 337– 346. 17. Bakhlanova, I. V., Ogawa, T. & Lanzov, V. A. (2001). Recombinogenic activity of chimeric recA genes (Pseudomonas aeruginosa/Escherichia coli): a search for RecA protein regions responsible for this activity. Genetics, 159, 7 – 15. 18. Chervyakova, D., Kagansky, A., Petukhov, M. & Lanzov, V. (2001). [L29M] substitution in the interface of subunit –subunit interactions enhances Escherichia coli RecA protein properties important for its recombinogenic activity. J. Mol. Biol. 314, 923– 935. 19. Lanzov, V. A. (2002). Hyper-recombination in Escherichia coli with and without SOS response. In Recent Research Development in DNA Repair and Mutagenesis (Ruiz-Rubio, M., Alejandre-Duran, E. & Roldan-Arjona, T., eds), pp. 21 – 38, Research Signpost, Kerala, India. 20. Kowalczykowski, S. C. & Krupp, R. A. (1995). DNAstrand exchange promoted by RecA protein in the absence of ATP: implications for the mechanism of energy transduction in protein-promoted nucleic acid transactions. Proc. Natl Acad. Sci. USA, 92, 3478– 3482. 21. Cox, M. M. (1994). Why does RecA protein hydrolyze ATP? Trends Biochem. Sci. 19, 217– 222. 22. Moncany, M. L. J. & Kellenberger, E. (1981). High magnesium content of Escherichia coli B. Experimentia, 37, 846– 847. 23. Lusk, J. E., Williams, J. P. & Kennedy, E. P. (1968). Magnesium and the growth of Escherichia coli. J. Biol. Chem. 243, 2618– 2624. 24. Alatossava, T., Ju¨tte, H., Kuhn, A. & Kellenberger, E. (1985). Manipulation of intracellular magnesium content in polymyxin B nonapeptide-sentisized Escherichia coli by ionophore A23187. J. Bacteriol. 162, 413– 419. 25. Story, R. M. & Steitz, T. A. (1992). Structure of the RecA protein – ADP complex. Nature, 355, 374– 376. 26. Mikawa, T., Masui, R. & Kuramitsu, S. (1998). RecA protein has extremely high cooperativity for substrate in its ATPase activity. J. Biochem. 123, 450– 457. 27. Story, R. M., Weber, I. T. & Steitz, T. A. (1992). The structure of the E. coli recA protein monomer and polymer. Nature, 355, 318– 324. 28. Menetski, J. P. & Kowalczykowsky, S. C. (1989). Enhancement of Escherichia coli RecA protein enzymatic function by dATP. Biochemistry, 28, 5871– 5881.
Acknowledgements We are grateful to Drs J. Genschel, L. Litz, C. Urbanke and R. Hilgenfeld for kindly supplying us with the SSB-Pa protein used in the study and Dr A. Alexseev for providing etheno-modified DNA. We thank also Dr L. Firsov and unknown reviewers for valuable comments and Dr L. Bresler for English correction. This work was supported by a Fogarty International Research Collaboration Award (grant 1 R03 TWO1319-01A1) and Russian Foundation for Basic Research (grant 02-04-48332).
References 1. Kowalczykowski, S. C. (2000). Initiation of genetic recombination and recombination-dependent replication. Trends Biochem. Sci. 35, 156– 165. 2. Kowalczykowski, S. C., Dixon, D. A., Eggleston, A. K., Lauder, S. D. & Rehrauer, W. M. (1994). Biochemistry of homologous recombination in Escherichia coli. Microbiol. Rev. 58, 401– 465. 3. Roca, A. I. & Cox, M. M. (1997). RecA protein: structure, function, and role in recombinational DNA repair. Prog. Nucl. Acid Res. Mol. Biol. 56, 129– 223. 4. Howard-Flanders, P., West, S. C., Rusche, J. R. & Egelman, E. H. (1984). Molecular mechanisms of general genetic recombination: the DNA-binding sites of RecA protein. Cold Spring Harbor Symp. Quant. Biol. 49, 571– 580. 5. Zlotnick, A., Mitchell, R. S., Steed, R. K. & Brenner, S. L. (1993). Analysis of two distinct single-stranded DNA binding sites on the RecA nucleoprotein filament. J. Biol. Chem. 268, 22525– 22530. 6. Mazin, A. V. & Kowalczykowski, S. C. (1996). The specificity of the secondary DNA binding site of RecA protein defines its role in DNA strand exchange. Proc. Natl Acad. Sci. USA, 93, 10673– 10678. 7. Mazin, A. V. & Kowalczykowski, S. C. (1998). The function of the secondary DNA-binding site of RecA protein during DNA strand exchange. EMBO J. 17, 1161 –1168. 8. Chow, S. A., Honigberg, S. M., Bainton, R. J. & Radding, C. M. (1986). Patterns of nuclease protection during strand exchange: RecA protein forms heteroduplex DNA by binding to strands of like polarity. J. Biol. Chem. 261, 6961– 6971. 9. Pugh, B. F. & Cox, M. M. (1990). RecA protein binding to the heteroduplex product of DNA strand exchange. J. Biol. Chem. 262, 1337– 1343. 10. Rosselli, W. & Stasiak, A. (1990). Energetics of RecAmediated recombination reaction: without ATP hydrolysis RecA can mediate polar strand exchange but is unable to recycle. J. Mol. Biol. 216, 335– 352. 11. Ulsperger, C. J. & Cox, M. M. (1995). Quantitative RecA protein binding to the hybrid duplex product of DNA strand exchange. Biochemistry, 34, 10859 –10866.
Primary DNA Binding Site of P. aeruginosa RecA Protein
7
29. McEntee, K., Weinstock, G. M. & Lehman, I. R. (1981). Binding of the recA protein of Escherichia coli to single- and double-stranded DNA. J. Biol. Chem. 256, 8835– 8844. 30. Menetski, J. P. & Kowalczykowski, S. C. (1985). Interaction of recA protein with single-stranded DNA. Quantitative aspects of binding affinity modulation by nucleotide cofactors. J. Mol. Biol. 181, 281– 295. 31. Voloshin, O. N., Wang, L. & Camerini-Otero, R. D. (2000). The homologous pairing domain of RecA also mediates the allosteric regulation of DNA binding and ATP hydrolysis: a remarkable concentration of functional residues. J. Mol. Biol. 303, 709– 720. 32. Malkov, V. A. & Camerini-Otero, R. D. (1995). Photocross-links between single-stranded DNA and Escherichia coli RecA protein map to loops L1 (amino acid residues 157– 164) and L2 (amino acid residues 195–209). J. Biol. Chem. 270, 30230– 30233. 33. Maraboeuf, F., Voloshin, O., Camerini-Otero, R. D. & Takahashi, M. (1995). The central aromatic residue in loop L2 of RecA interacts with DNA. Quenching of the fluorescence of a tryptophan reporter inserted in L2 upon binding to DNA. J. Biol. Chem. 270, 30927– 30932. 34. Voloshin, O. N., Wang, L. & Camerini-Otero, R. D. (1996). Homologous DNA pairing promoted by a 20-amino acid peptide derived from RecA. Science, 272, 868– 872.
35. Roca, A. I. & Cox, M. M. (1990). The RecA protein: structure and function. Crit. Rev. Biochem. Mol. Biol. 25, 415–456. 36. Morimatsu, K. & Horii, T. (1995). DNA-binding surface of RecA protein. Photochemical cross-linking of the first DNA binding site on RecA filament. Eur. J. Biochem. 234, 695–705. 37. Wang, W.-B. & Tessman, E. S. (1996). Location of functional regions of the Escherichia coli RecA protein by DNA sequence analysis of RecA protease-constitutive mutants. J. Bacteriol. 168, 901– 910. 38. Liu, S-K., Eisen, J. A., Hanawalt, P. C. & Tessman, I. (1993). recA mutations that reduce the constitutive coprotease activity of the RecA1202 (Prtc) protein: possible involvement of interfilament association in proteolytic and recombination activities. J. Bacteriol. 175, 6518– 6529. 39. Shan, Q., Bork, J. M., Webb, B. L., Inman, R. B. & Cox, M. M. (1997). RecA protein filaments: end-dependent dissociation from ssDNA and stabilization by RecO and RecR proteins. J. Mol. Biol. 265, 519– 540. 40. Genschel, J., Litz, L., Thole, H., Roemling, U. & Urbanke, C. (1996). Isolation, sequencing and overproduction of the single-stranded DNA binding protein from Pseudomonas aeruginosa PAO. Gene, 182, 137 –143.
Edited by J. Karn (Received 3 December 2002; received in revised form 7 February 2003; accepted 10 February 2003)