Keywords: DnaA box, DnaA protein,oriC region, orisome (protein–oriC complex)

Abbreviations: AAA+, ATPase associated with a variety of cellular activities;amp^r, ampicillin resistance; DTT, dithiothreitol; EMSA, electrophoretic mobility-shift assay; HTH, helix–turn–helix; ORB, origin recognition boxes; ORC, origin recognition complex;oriC, origin of chromosomal replication; RU, resonance units; SPR, surface plasmon resonance;tsr^r, thiostrepton resistance

DNA manipulations

E. coli XL1-Blue (endA1,gyrA46,hsdR17,lac,recA1,supE44,thi; F'lac:lacI^q, Δ(lacZ)M15, Tn10,proA⁺,proB⁺) served as the host for the plasmids. WM1785 (=W3110) and itspolA derivative WM1838 (polA,fadA::Tnl0) were used as host strains in theori assay [18]. Transformation ofS. coelicolor has been described previously [19].S. coelicolor transformants were selected for resistance to 10 μg of thiostrepton per ml. The DNA fragments and plasmids were purified using kits according to the manufacturer's protocols (Qiagen). DNA fragments for gel retardation were 5′-end-labelled using [γ-³²P]ATP and T4 polynucleotide kinase. Enzymes were supplied by Roche, Fermentas MBI and Gibco BRL. Isotopes were obtained from MP Biomedicals (Irvine, CA, U.S.A.). The oligonucleotides used for PCR or gel retardation were from Bionovo (Bioresearch Laboratory & Biochemicals, Legnica, Poland).

Plasmid and minichromosomes

TheoriC regions of the analysed organisms were cloned into pBR322 plasmid. TheoriC regions ofM. tuberculosis andS. coelicolor were amplified by PCR using a pair of primers tailed by a BamHI motif and were cloned into pBR322 vector linearized by the same restriction enzyme. The resultant constructs were named pBR322_MtoriC and pBR322_StoriC respectively. The remainder constructs, pBR322_HporiC and pBR322_EcoriC (originally named pOC170) containing theoriC region fromH. pylori andE. coli respectively were obtained as described previously [20,21].

For minichromosome replication studies inS. coelicolor, the PCR-amplifiedoriC regions were ligated with a DNA fragment carrying thetsr^r (thiostrepton resistance) gene and were then introduced intoS. coelicolor protoplasts. Thiostrepton-resistant colonies were examined for the presence of minichromosomes (closed circular plasmid DNA) as described previously [22].

DnaA purification

Recombinant DnaA proteins ofE. coli,H. pylori,M. tuberculosis andS. coelicolor were purified according to procedures described previously by Krause et al. [23], Zawilak et al. [24], Zawilak et al. [25] and Majka et al. [26] respectively. The purified DnaA proteins were more than 95% homogeneous, as judged by SDS/PAGE analysis. ATP-bound DnaA proteins were used in allin vitro experiments.

Gel-retardation assay

For binding assays,³²P-labelled DNA was incubated with DnaA protein in the presence of a non-specific competitor [poly-(dI-dC)·(dI-dC); Roche] at room temperature (20 °C) for 30 min in 0.5× Marians' binding buffer [1× Marians' buffer: 20 mM Hepes/KOH, pH 7.6, 5 mM magnesium acetate, 1 mM EDTA, 4 mM DTT (dithiothreitol), 0.2% Triton X-100, 3 mM ATP and 50 μg/ml BSA] [27]. The bound complexes were separated by electrophoresis in 4 or 6% polyacrylamide gels [0.25× TBE (89 mM Tris/89 mM borate/1 mM EDTA) at 4 V/cm, 4 °C]. Gels were dried and analysed by a Typhoon 8600 Variable Mode Imager and ImageQuant software. The apparent equilibrium dissociation constantK_d was determined using the method described by Carey [28]. A reaction mixture contained a fixed amount of DNA and various concentrations of DnaA protein. The DNA concentration chosen was much lower than the protein concentration required for half-maximal binding, so the protein concentration at half-maximal binding is very close toK_d (app). TheK_d (app) was deduced from a curve [unbound DNA (%) against DnaA (nM)], based on the equationK_d=[S]·[P]·[SP]⁻¹, where [S] is DNA concentration, [P] is protein concentration and [SP] is DNA–protein complex concentration. When [S]«K_d, then [P]_free≈[P]_total, soK_d=[P]_total·[S]·[PS]⁻¹.

SPR (surface plasmon resonance)

The BIAcore system (Biosensor AB) was used to study the interaction between protein and DNA. The system allows analysis of molecular interactions in real time by monitoring changes of the light angle inducing SPR. The SPR signal is directly proportional to the mass changes at the sensor surface and is expressed in RU (resonance units). For most proteins and DNA, 1000 RU corresponds to a surface concentration of approx. 1 ng/mm² (BIA-evaluation software handbook, 1996, BIAcore AB). The biotinylated double-stranded oligonucleotides were immobilized on a streptavidin-coated SA chip of the BIAcore apparatus. Approx. 50 RU of DNA were immobilized. DNA loosely attached to the surface of the chip was removed with a pulse of 0.05% SDS. In order to exclude the effects of mass transport on the kinetics of the protein–DNA interactions, the measurements were performed at various protein concentrations (0.39–100 nM) and at a continuous flow rate (100 μl·min⁻¹). The measurements were performed in KAC buffer (25 mM Hepes, 10 mM magnesium acetate, 100 potassium acetate, 5 mM EDTA and 0.005% BIAcore surfactant P20) in the presence of poly(dI-dC)·(dI-dC) as a competitor. At the end of each cycle, bound DnaA protein was removed by washing with 0.05% SDS. The kinetic constants for binding of DnaA protein to the double-stranded DNA were determined from the association and dissociation curves of the sensorgrams using the BIAevaluation version 3 program, as described previously [29].

SfiI probing

Plasmid pGEM®-T Easy-SfiIScoriCSfiI (150 ng) was incubated with different amounts of DnaA protein in binding buffer (10 mM Tris/HCl, pH 7.5, 10 mM MgCl₂, 50 mM NaCl and 1 mM DTT) at room temperature for 30 min. Then, SfiI digestion was carried out at 37 °C for 20 min according to the manufacturer's protocol (Amersham Biosciences). The reactions were stopped by incubation at 65 °C for 20 min. The SfiI cleavage products were separated on a 1% agarose gel. After staining with Sybr Green I, the gel was analysed by a Typhoon 8600 Variable Mode Imager using ImageQuant software.

Different DnaA proteins exhibit various affinities toward single DnaA boxes

The consensus sequence for theE. coli DnaA box is 5′-TTATNCACA-3′. Binding studies revealed that theE. coli DnaA protein exhibits the highest affinity towards the 5′-TTATCCACA-3′ sequence (EcDnaA box) that is named the ‘perfect’ DnaA box [9]. In high G+C content organisms, such asMycobacterium andStreptomyces, the third position is replaced with G or C: 5′-TT(G/C)TCCACA-3′ to give theMtScDnaA box [30–32]. Comparison of DnaA boxes from theH. pylori oriC region allowed us to determine the consensus sequence: 5′-TTCTTCACA-3′ (HpDnaA box) [33]. Since the consensus sequences for DnaA boxes differ slightly among the organisms analysed, it was interesting to compare the binding of the DnaA proteins to distinct single DnaA boxes (HpDnaA,EcDnaA andMtScDnaA). To evaluate the interactions of the DnaA proteins with individual DnaA boxes, SPR and/or a gel-retardation assay was applied. Oligonucleotides containing a single DnaA box,HpDnaA,EcDnaA orMtScDnaA box (Table 1), were end-labelled with biotin or [γ-³²P]ATP for SPR or a gel-retardation assay respectively and were then incubated with increasing amounts of a given DnaA protein. Representative binding experiments are shown inFigure 2. TheK_d (app) values (Figure 2) were calculated from the sensorgrams and/or by quantification of gel-retardation assays as described previously in detail [29,34]. Our results show that the affinities of the analysed proteins for a single DnaA box vary significantly. TheE. coli DnaA protein exhibited the highest affinity toward its own (EcDnaA) box. This protein, in addition to its own DnaA box, also bound theMtScDnaA box. As was shown previously, theS. coelicolor DnaA protein prefers theE. coli DnaA box over its own DnaA box [34]. TheH. pylori DnaA bound only its ownHpDnaA box, although with weak affinity. What is more, no other analysed DnaA protein bound theHpDnaA box. In contrast with the other proteins, theM. tuberculosis DnaA binds neither its cognateMtScDnaA box nor other single DnaA boxes (Figure 2).

Spatial arrangement of two DnaA boxes is crucial for theH. pylori andS. coelicolor DnaA proteins, but not forE. coli orM. tuberculosis proteins

Our previous studies demonstrated that theS. coelicolor DnaA protein exhibits preference for DNA containing two DnaA boxes in reverse orientation separated by 3 bp; such a box arrangement has been found in the promoter region of thednaA gene as well as in theoriC regions of differentStreptomyces species, includingS. coelicolor, e.g. the fifth and sixth DnaA boxes (Figure 1) [30,35,36]. The structure of theH. pylori oriC suggests thatH. pylori DnaA may prefer the same orientation of two DnaA boxes (2 bp between them,Hp boxes-wt;Figure 1). Thus it seems interesting to examine the influence of orientation and distance between two DnaA boxes on the binding affinity of the analysed DnaA proteins. For this purpose, we synthesized a set of six oligonucleotides bearing two boxes in reverse (Scboxes-wt,Scboxes-wt+10 bp,Hpboxes-rv) or direct (Scboxes-fw,Hpboxes-wt,Hpboxes-wt+10 bp) orientation (seeFigure 3). None of theH. pylori oligonucleotides was bound by theS. coelicolor DnaA (and vice versa; results not shown). Incubation of theM. tuberculosis orE. coli DnaA protein with each of the six oligonucleotides caused the appearance of single or two nucleoprotein complexes respectively; both proteins exhibited higher affinity toward theS. coelicolor boxes than toward theH. pylori DnaA boxes (results not shown). Neither the relative orientation nor the spacing between the two DnaA boxes significantly affected the affinity of theE. coli andM. tuberculosis DnaA for the DNA. However, if theE. coli DnaA binds as dimer, the orientation and distance of DnaA boxes does affect the binding affinity [11]. In contrast, the arrangement of boxes was important for theH. pylori DnaA protein, which exhibited the highest affinity toward the wild-type arrangement of DnaA boxes (Hpboxes-wt). The twoH. pylori DnaA boxes in reverse orientation (Hpboxes-rv) were also bound by theH. pylori DnaA, albeit with lower affinity than the wild-type boxes. Addition of 10 bp (Hpboxes-wt+10 bp) significantly reduced the interaction betweenH. pylori DnaA and two boxes (Figure 3). TheS. coelicolor DnaA bound all three oligonucleotides; however, the boxes in direct orientation (Scboxes-fw) were bound less efficiently. Separation of the DnaA boxes (Scboxes-wt+10 bp) did not change the affinity of the protein (mainly single nucleoprotein complex was formed).

Thus, among the DnaA proteins analysed, the orientation of the boxes and distance between them have influence on theH. pylori andS. coelicolor DnaA binding.

TheoriC regions are optimally adjusted to their cognate DnaA proteins

Interactions between wholeoriC regions and DnaA proteins were analysed by a series of gel-retardation assays. For our studies, fouroriC regions fromH. pylori,E. coli,M. tuberculosis andS. coelicolor were amplified by PCR using the templates and pairs of primers listed inTable 1. The number of DnaA boxes within theoriC regions analysed varies from five (H. pylori andE. coli) to 19 (S. coelicolor) (Figure 1). The labelledoriC fragments were incubated with increasing amounts of purified DnaA proteins, and then nucleoprotein complexes were analysed in a 4% native polyacrylamide gel. Formation of nucleoprotein complexes was studied in homologous (oriC and DnaA from the same organism) and heterologous (oriC and DnaA from two different organisms) systems. It is worthwhile to mention that DnaA proteins bind with the same affinity to linear and supercoiledoriC DNA ([21], and results not shown). DnaA binding tooriC is determined by the DNA sequence rather than by the topology. Therefore we used for our assay a linear form of DNA that allowed us to follow in an easy manner a nucleoprotein complex formation.

In homologous, as well as heterologous, systems, the nucleoprotein complexes observed were formed in a manner that is dependent on the protein concentration (Figure 4). In homologous systems, nucleoprotein complexes readily appeared at the lowest protein concentrations (Figure 4). In most heterologous systems (but not forM. tuberculosis DnaA), nucleoprotein complexes were detectable only at elevated protein concentrations. TheH. pylori andS. coelicolor DnaA proteins and theiroriC regions exhibited particularly low reciprocal affinity in heterologous system; diffuse nucleoprotein complexes were detectable only at the highest protein concentrations (Figure 4). In contrast,M. tuberculosis DnaA bound with a high avidity not only its ownoriC region, but also ‘foreign’oriC region fromE. coli andS. coelicolor and, to some extent,H. pylori oriC (Figure 4).

Interestingly, interactions of theE. coli DnaA protein with each of the analysedoriC regions led to the formation of multiple discrete nucleoprotein complexes which were visible as a ladder of retarded bands (Figure 4). As theE. coli DnaA protein concentration increased, the complexity of the band pattern increased until a critical point was reached at which the ladder pattern disappeared and was replaced by diffuse highly retarded bands, indicative of large, but not very stable, complexes (Figure 4). The number of discrete retarded bands seemed to be proportional to the complexity of theoriC region analysed: the highest number of nucleoprotein complexes could be observed for theS. coelicolor oriC region that contains 19 DnaA boxes. The formation of discrete nucleoprotein complexes can be explained by monomeric binding of theE. coli DnaA to multiple DnaA boxes; at lower concentrations, theE. coli DnaA bound each of the DnaA boxes individually (discrete bands were visible on the electrophoretic mobility-shift assay gel;Figure 4), starting from high-affinity DnaA boxes (e.g. DnaA boxes R1 and R4; [21,37]).

H. pylori DnaA protein also bound DnaA boxes sequentially, but preferentially those from its ownoriC region.H. pylori DnaA exhibited the highest affinity towards two closely spaced DnaA boxes that were arranged ‘head–2 bp–tail’; such an arrangement is repeated within theH. pylori oriC region (Figure 1). TheH. pylori DnaA protein bound randomly to the repeated pair of the DnaA boxes (2-3 and4-5); each pair of DnaA boxes is probably bound by a dimer of the DnaA protein. Higher protein concentrations are required to observe binding to the DnaA box1 that is not bound as an individual DnaA box [33].

In contrast withE. coli andH. pylori, incubation of theM. tuberculosis andS. coelicolor DnaA proteins with theoriC regions led to the formation of high-molecular-mass complexes already at a low protein concentration.S. coelicolor DnaA first bound the ‘head–3 bp–head’ DnaA boxes5 and6, and then ‘weaker’ DnaA boxes [26]; the discrete complexes were only visible at the lowest protein concentration (Figure 4). ForM. tuberculosis DnaA, discrete nucleoprotein complexes were not observed, even at a low protein concentration (Figure 4). The formation of higher-molecular-mass complexes, which were visible as a diffuse band(s) already at a low protein concentration, was probably caused by a property of theM. tuberculosis DnaA protein; it does not bind a single DnaA box and, for efficient binding, presumably requires at least two DnaA boxes, which are bound in a co-operative manner.

In order to elucidate further DnaA protein binding specificity with regard to the origin region, a series of competition gel-retardation assays were performed. All four labelledoriC fragments were incubated separately with each of the analysed DnaA proteins (Figure 5). Competition gel-retardation assays revealed that the DnaA proteins fromH. pylori andS. coelicolor exhibited the highest affinity towards their ownoriC regions (Figure 5). These experiments corroborated an earlier observation that theoriC regions ofH. pylori andS. coelicolor were poorly recognized by heterologous DnaA proteins. TheM. tuberculosis DnaA protein preferred theoriC region ofS. coelicolor over its cognate origin:S. coelicolor oriC was completely bound at the lowest concentration of this protein (Figure 5).

It has been demonstrated that DnaA proteins mediated DNA bending oforiC regions [23,35,38]. The DnaA protein ofBacillus subtilis forms stable loops by interaction of protein molecules bound to the DnaA box groups separated by the long spacer,dnaA gene (∼1400 bp). Interestingly, theE. coli DnaA is also able to loop out theB. subtilis oriC region [23]. TheoriC region fromS. coelicolor also contains two clusters of DnaA boxes; however, they are separated by a short spacer (120 bp;Figure 1). In order to check whether theS. coelicolor oriC is exclusively looped out by its cognate DnaA, we examined the ability of DnaA proteins fromE. coli andM. tuberculosis to form the loop structure. For this purpose, the SfiI probing assay was applied. The interaction with two SfiI sitesin cis involves the formation of a DNA loop between the sites, presumably as a result of the tetrameric protein binding simultaneously to both sites [39]. For SfiI probing, theoriC region ofS. coelicolor was PCR-amplified with a pair of primers (ScoriSfiIbend_fw andScoribend_rv) in such a way that two SfiI cleavage sites flank theoriC. The PCR product was cloned into a T-vector (pGEM®-T Easy), which does not contain any SfiI sites. The resulting plasmid pGEM®-T Easy-SfiIScoriCSfiI was incubated with different amounts of theS. coelicolor,E. coli orM. tuberculosis DnaA protein and then subjected to the SfiI enzyme cleavage. Under the conditions used, the SfiI digestion was inefficient, but was enhanced significantly (∼3.0 times;Figure 6) by theS. coelicolor DnaA at a protein/DnaA box ratio of 1:1. In contrast, whenE. coli orM. tuberculosis DnaA was incubated with pGEM®-T Easy-SfiIScoriCSfiI, no distinct enhancement of SfiI cleavage was observed (Figure 6). Thus SfiI probing demonstrated that theS. coelicolor oriC is exclusively bent by its own DnaA protein.

Host specificity oforiC regions

In order to investigate the host specificity of the analysedoriC regions, we performed a set of heterologous transformations; the ability of the origins to promote replication was examined inE. coli andS. coelicolor (Table 2). For assay inE. coli, the PCR-amplifiedoriC regions were cloned into the pBR322 plasmid. The resultant constructs, named pBR322_HporiC, pBR322_EcoriC, pBR322_MtoriC and pBR322_ScoriC, were assayed fororiC-dependent initiation of replicationin vivo using the frequently usedpolA system [40]. ColE1-type plasmids (such as pBR322) require DNA polymerase I for their replication. Thus only the constructs containing a functionaloriC region are able to replicate (and confer pBR322-encoded ampicillin resistance,amp^r) in the absence of DNA polymerase I (inE. coli polA⁻ strain). Similar to pBR322 itself (a negative control), none of the ‘foreign’oriC regions gave ampicillin-resistant transformants inpolA deficient strain; theamp^r-positive colonies did not appear even after prolonged incubation (of a few days). The pBR322_EcoriC was the only construct in whichoriC was active (positive control). For a replication assay inS. coelicolor, minichromosomes (or shuttle vectors, containing, in addition to theoriC region andtsr^r, pMB1 replicon andamp^r to propagate inE. coli; results not shown) were constructed, and their ability to replicatein vivo was tested as described previously [22]. In contrast with pScoriC minichromosome (positive control), pEcoriC minichromosome was found not to replicate inS. coelicolor (Table 2). Interestingly, despite the fact that ScDnaA protein binds theoriC region ofM. tuberculosis with a high affinity, theM. tuberculosis oriC was not able to initiate minichromosome replication inS. coelicolor. Replication ability of theH. pylori oriC region inS. coelicolor was not tested, since this region exhibits particularly low affinity for theScDnaA protein.

The DnaA proteins analysed differ in binding affinity

The DnaA proteins analysed are able to bind a single DnaA box, with the exception of theM. tuberculosis DnaA protein. However, they exhibit different affinity towards single DnaA boxes; theE. coli andH. pylori DnaA proteins bind their cognate DnaA boxes with the highest and lowest affinity respectively. In contrast with theE. coli andH. pylori DnaA proteins, theS. coelicolor DnaA protein reveals higher affinity for theE. coli DnaA box than for its own DnaA box. Recently, analysis of the crystal structure of theE. coli DnaA binding domain complexed with R1 DnaA box revealed that 18 amino acid residues have direct contact with DNA (seeFigure 7): 14 residues interact with DNA backbone phosphate groups, four residues are involved in the base-specific interactions (two of them, Asp⁴³⁴ and Arg³⁹⁹ interact with both bases and phosphor groups) [16]. Sequence comparison of theE. coli binding domain with the corresponding protein fragments formH. pylori,M. tuberculosis andS. coelicolor showed principal differences that are summarized inFigure 7. First, interesting differences between theE. coli andM. tuberculosis orH. pylori DnaA binding domains concern residues involved directly in the base recognition: His⁴³⁹ and Pro⁴²³ both make van der Waals contact with the C-5 methyl group of T4 and T9* (*, opposite strand) respectively (Figure 7). His⁴³⁹ is replaced in theM. tuberculosis DnaA (and other mycobacterial DnaA proteins as well) by the non-closely related amino acid, tyrosine (Figure 7). InE. coli, His⁴³⁹ belongs to the A-signature motif that has been shown by extensive mutational analyses to define both the affinity and specificity of DnaA box binding. Thus it is tempting to conclude that the substitution (H439Y) may cause a decrease in theM. tuberculosis DnaA affinity. In theH. pylori DNA-binding domain, Pro⁴²³ (a well-conserved residue in other DnaAs) is replaced by leucine; this substitution probably causes reduction of protein affinity. It should be noted that the binding domain of theS. coelicolor DnaA protein does not contain any substitutions in amino acids that are directly involved in base recognition (Figure 7). Thus it explains why theS. coelicolor DnaA protein, in contrast with that ofH. pylori andM. tuberculosis, binds theE. coli R1 DnaA box with higher affinity than its own DnaA box. In contrast with T*, C* at the third position ofMtScDnaA box is probably unable to form a water-mediated hydrogen bond with the highly conserved Arg³⁹⁹ (present inS. coelicolor, but replaced by a closely related residue, lysine, inM. tuberculosis) [16]. Thus the third position of the DnaA box may also contribute to the high-affinity binding. Interesting differences between theE. coli and the analysed proteins also concern residues that are directly involved in the interactions with DNA backbone phosphate groups (Figure 7), e.g. the triad PTL (Pro-Thr-Leu) present in theH. pylori binding domain probably contributes to the lower specificity of the protein; all three residues show significant divergence from the consensus, and threonine is the most divergent residue of all [24].

Analysis of the interactions between the DnaA proteins and two DnaA boxes demonstrated that, besides the sequences of DnaA boxes, their relative spacing and orientation affect binding of the proteins, particularly those ofH. pylori andS. coelicolor. TheH. pylori DnaA protein exhibits the highest affinity for the two DnaA boxes arranged ‘head-to-tail’ (2 bp between); such arrangement is repeated twice within theH. pylori oriC region (Figures 1 and3). TheS. coelicolor DnaA protein prefers a ‘head-to-head’ (3 bp between) arrangement that is present not only within itsoriC region, but also in the promoter region of thednaA gene (its expression is autoregulated) [35,36,41]. Insertion of exactly 10 bp (B-DNA has 10 bp per turn of helix) between the twoS. coelicolor DnaA boxes practically did not change the affinity of theS. coelicolor DnaA protein, while such a separation between the twoH. pylori DnaA boxes nearly completely abolished binding of theH. pylori DnaA protein (Figure 3). Thus in contrast with theH. pylori DnaA, theS. coelicolor DnaA protein could contact widely spaced (to some extent) DnaA boxes. Interestingly,H. pylori andS. coelicolor DnaA proteins possess the shortest (66 amino acids) and the longest (253 amino acids) domain II respectively. Domain II does not contain any relevant secondary structural motif and is considered as a flexible linker that connects the N-terminal domain involved in dimerization (domain I) with the C-terminal part of the DnaA protein responsible for ATPase activity (domain III) and DNA binding (domain IV) [9]. The presence of the long flexible domain II allows theS. coelicolor DnaA protein to bind separated DnaA boxes, while the efficient binding of theH. pylori DnaA is limited to the closely spaced DnaA boxes, presumably because of the presence of the short domain II.

The relative orientation and spacing of paired DnaA boxes do not significantly affect the binding ofE. coli andM. tuberculosis DnaA proteins (domain II ofE. coli andM. tuberculosis DnaA proteins consist of 77 amino acids and 83 amino acids respectively;Figure 1). In the case ofE. coli, most DnaA boxes in the fragments analysed, i.e. inoriC, are strong boxes, and therefore it is not surprising that DnaA binds individually to them with no apparent co-operativity. However, detailed SPR analysis demonstrated that also forE. coli DnaA, the spatial arrangement of the DnaA boxes has an effect on the binding affinity of this protein [9].

The architecture of the nucleoprotein complexes: theoriC regions are optimally adjusted to their cognate DnaA proteins

The number and positions of the DnaA boxes determine the architecture and composition of the initiator complexes. Successive binding to sites with different affinities generates a particular order of assembly. The arrangement of DnaA boxes generates protein–DNA and protein–protein interactions that are unique to a given organism. Binding of the DnaA proteins fromE. coli andH. pylori to their cognateoriC regions proceeds sequentially. However, theE. coli DnaA protein exhibits much higher affinity for its individual DnaA boxes than theH. pylori DnaA protein does for its own DnaA boxes (Figure 2). In contrast with theH. pylori oriC region, the DnaA boxes from theE. coli oriC are separated by considerably long spacers (seeFigure 1). The lower affinity of theH. pylori DnaA protein towards DnaA boxes is presumably compensated by the ‘compact’ arrangement of the DnaA boxes (Figure 1). In the case ofE. coli DnaA, binding starts at box number R4, and then other boxes are bound with no apparent co-operativity [21]. However, in the last step of initiation of replication, the DnaA box R1 serves as an anchor for DnaA molecules that are engaged in the unwinding reaction [9,11,12,42].

The preference of theS. coelicolor andM. tuberculosis DnaA proteins in binding concertedly several separate DnaA boxes distinguishes these complexes from those ofE. coli andH. pylori. InS. coelicolor, there is only one DnaA box with the consensus sequence (box6) and none inM. tuberculosis. Consequently, only a few of the 19 DnaA boxes from theoriC region exhibit specific binding of the DnaA protein, while none of the 13 DnaA boxes from theM. tuberculosis oriC region is recognized individually by its own DnaA protein. At least two ‘weak’ DnaA boxes are required for specific efficient binding [25,30,34]. Thus it implies co-operativity for binding to origins with low-affinity DnaA boxes. Presumably due to the high GC pressure exerted during the course ofMycobacterium andStreptomyces evolution, theoriC regions ofS. coelicolor andM. tuberculosis have been changed; the presence of ‘weak’ GC-rich DnaA boxes within theS. coelicolor andM. tuberculosis oriC regions is compensated by their abundance (seeFigure 1). TheS. coelicolor nucleoprotein complex is more intricate than other analysed initiation complexes [35]. TheS. coelicolor DnaA protein, owing to the presence of long domain II, is able to bind widely spaced DnaA boxes and to bend the short spacer that links two clusters of DnaA boxes. The SfiI probing demonstrated that theS. coelicolor oriC is exclusively bent only by its own DnaA protein. The other proteins analysed containing shorter domain II are not able to bend theS. coelicolor oriC region.

Our experiments demonstrated that theoriC regions, particularly those ofH. pylori andS. coelicolor, are optimally adjusted to their cognate DnaA proteins. During evolution, the structure of theoriC regions changed, particularly the sequence and the length of the spacers that link DnaA boxes. Alterations in spacer lengths caused changes in the spatial arrangement of the DnaA boxes. The sequence of DnaA boxes also evolved differently for low- and high-GC-rich organisms such asH. pylori andS. coelicolor respectively. The weak interactions between a single DnaA box and DnaA protein are presumably compensated by the arrangement of DnaA boxes and/or by an abundance of DnaA boxes, e.g. a pair of closely spacedH. pylori DnaA boxes are bound by its cognate DnaA protein with the highest affinity. Thus, during the evolution of a given organism, the two key elements of initiation of replication, DnaA protein andoriC region, were tuned to optimal interaction. Ourin vivo functional analysis of replication origins indicates the host specificity oforiC regions. The results corroborate earlier observations; the analysis of the hybridB. subtilis/E. coli replication origin suggested that the species specificity resides in the DnaA box part of the origin, probably in the spatial arrangement of DnaA boxes [43].

The results of the present study indicate that the primary functions of multiple DnaA boxes are to determine the positioning and order of assembly of the DnaA molecules. Gradual transition from the sequence-specific binding of the DnaA protein to binding through co-operative protein–protein interactions seems to be a common conserved strategy to generate oligomeric initiator complexes that are bound to multiple sites within the chromosomal, plasmid [44] and viral origins [45–48].

• 1.Kornberg A., Baker T. A. DNA Replication. New York: W. H. Freeman & Co.; 1992. [Google Scholar]
• 2.Baker T. A., Bell S. P. Polymerases and the replisome: machines within machines. Cell. 1998;92:295–305. doi: 10.1016/s0092-8674(00)80923-x. [DOI] [PubMed] [Google Scholar]
• 3.Matsunaga F., Forterre P., Ishino Y., Myllykallio H. In vivo interactions of archaeal Cdc6/Orc1 and minichromosome maintenance proteins with the replication origin. Proc. Natl. Acad. Sci. U.S.A. 2001;98:11152–11157. doi: 10.1073/pnas.191387498. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 4.Robinson N. P., Dionne I., Lundgren M., Marsh V. L., Bernander R., Bell S. D. Identification of two origins of replication in the single chromosome of the archaeon Sulfolobus solfataricus. Cell. 2004;116:25–38. doi: 10.1016/s0092-8674(03)01034-1. [DOI] [PubMed] [Google Scholar]
• 5.Bell S. P. The origin recognition complex: from simple origins to complex functions. Genes Dev. 2002;16:659–672. doi: 10.1101/gad.969602. [DOI] [PubMed] [Google Scholar]
• 6.Bell S. P., Dutta A. DNA replication in eukaryotic cells. Annu. Rev. Biochem. 2002;71:333–374. doi: 10.1146/annurev.biochem.71.110601.135425. [DOI] [PubMed] [Google Scholar]
• 7.Kelman L. M., Kelman Z. Archaea: an archetype for replication initiation studies? Mol. Microbiol. 2003;48:605–615. doi: 10.1046/j.1365-2958.2003.03369.x. [DOI] [PubMed] [Google Scholar]
• 8.Giraldo R. Common domains in the initiators of DNA replication in Bacteria, Archaea and Eukarya: combined structural, functional and phylogenetic perspectives. FEMS Microbiol. Rev. 2003;26:533–554. doi: 10.1111/j.1574-6976.2003.tb00629.x. [DOI] [PubMed] [Google Scholar]
• 9.Messer W. The bacterial replication initiator DnaA: DnaA and oriC, the bacterial mode to initiate DNA replication. FEMS Microbiol. Rev. 2002;26:355–374. doi: 10.1111/j.1574-6976.2002.tb00620.x. [DOI] [PubMed] [Google Scholar]
• 10.Erzberger J. P., Pirruccello M. M., Berger J. M. The structure of bacterial DnaA: implications for general mechanisms underlying DNA replication initiation. EMBO J. 2002;21:4763–4773. doi: 10.1093/emboj/cdf496. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 11.Messer W., Blaesing F., Jakimowicz D., Krause M., Majka J., Nardmann J., Schaper S., Seitz H., Speck C., Weigel C., et al. Bacterial replication initiator DnaA: rules for DnaA binding and roles of DnaA in origin unwinding and helicase loading. Biochimie. 2001;83:5–12. doi: 10.1016/s0300-9084(00)01216-5. [DOI] [PubMed] [Google Scholar]
• 12.Speck C., Messer W. Mechanism of origin unwinding: sequential binding of DnaA to double- and single-stranded DNA. EMBO J. 2001;20:1469–1476. doi: 10.1093/emboj/20.6.1469. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 13.Messer W. Blaesing, F. Majka, J. Nardmann, J. Schaper, S. Schmidt, A. Seitz, H. Speck, C. Tungler, D. Wegrzyn G., et al. Functional domains of DnaA proteins. Biochimie. 1999;81:819–825. doi: 10.1016/s0300-9084(99)00215-1. [DOI] [PubMed] [Google Scholar]
• 14.Weigel C., Schmidt A., Seitz H., Tungler D., Welzeck M., Messer W. The N-terminus promotes oligomerization of the Escherichia coli initiator protein DnaA. Mol. Microbiol. 1999;34:53–66. doi: 10.1046/j.1365-2958.1999.01568.x. [DOI] [PubMed] [Google Scholar]
• 15.Simmons L. A., Felczak M., Kaguni J. M. DnaA protein of Escherichia coli: oligomerization at the E. coli. chromosomal origin is required for initiation and involves specific N-terminal amino acids. Mol. Microbiol. 2004;49:849–858. doi: 10.1046/j.1365-2958.2003.03603.x. [DOI] [PubMed] [Google Scholar]
• 16.Fujikawa N., Kurumizaka H., Nureki O., Terada T., Shirouzu M., Katayama T., Yokoyama S. Structural basis of replication origin recognition by the DnaA protein. Nucleic Acids Res. 2003;31:2077–2086. doi: 10.1093/nar/gkg309. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 17.Mackiewicz P., Zakrzewska-Czerwińska J., Zawilak A., Dudek M. R., Cebrat S. Where does bacterial replication start? Rules for predicting the oriC region. Nucleic Acids Res. 2004;32:3781–3791. doi: 10.1093/nar/gkh699. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Woelker B., Messer W. The structure of the initiation complex at the replication origin, oriC, of Escherichia coli. Nucleic Acids Res. 1993;21:5025–5033. doi: 10.1093/nar/21.22.5025. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 19.Hopwood D. A., Bibb M. J., Chater K. F., Kieser T., Bruton C. J., Kieser H. M., Lydiate D. J., Smith C. P., Ward J. M., Schrempf H. Norwich: The John Innes Foundation; 1985. Genetic Manipulation of Streptomyces: a Laboratory Manual. [Google Scholar]
• 20.Zawilak A. Ph.D. Thesis. Ludwik Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław; 2003. Characterization of the key elements of initiation of Helicobacter pylori chromosome replication. [Google Scholar]
• 21.Weigel C., Schmidt A., Ruckert B., Lurz R., Messer W. DnaA protein binding to individual DnaA boxes in the Escherichia coli replication origin, oriC. EMBO J. 1997;16:6574–6583. doi: 10.1093/emboj/16.21.6574. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 22.Zakrzewska-Czerwińska J., Majka J., Schrempf H. Minimal requirements of the Streptomyces lividans 66 oriC region and its transcriptional and translational activities. J. Bacteriol. 1995;177:4765–4771. doi: 10.1128/jb.177.16.4765-4771.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 23.Krause M., Rückert B., Lurz R., Messer W. Complexes at the replication origin of Bacillus subtilis with homologous and heterologous DnaA protein. J. Mol. Biol. 1997;274:365–380. doi: 10.1006/jmbi.1997.1404. [DOI] [PubMed] [Google Scholar]
• 24.Zawilak A., Durrant M. C., Jakimowicz P., Backert S., Zakrzewska-Czerwińska J. DNA binding specificity of the replication initiator protein, DnaA from Helicobacter pylori. J. Mol. Biol. 2003;334:933–947. doi: 10.1016/j.jmb.2003.10.018. [DOI] [PubMed] [Google Scholar]
• 25.Zawilak A., Kois A., Konopa G., Smulczyk-Krawczyszyn A., Zakrzewska-Czerwińska J. Mycobacterium tuberculosis DnaA initiator protein: purification and DNA binding requirements. Biochem. J. 2004;382:247–252. doi: 10.1042/BJ20040338. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 26.Majka J., Messer W., Schrempf H., Zakrzewska-Czerwińska J. Purification and characterization of the Streptomyces lividans initiator protein DnaA. J. Bacteriol. 1997;179:2426–2432. doi: 10.1128/jb.179.7.2426-2432.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 27.Parada C. A., Marinas K. J. Mechanism of DNA A protein-dependent pBR322 DNA replication: DNA A protein-mediated trans-strand loading of the DNA B protein at the origin of pBR322 DNA. J. Biol. Chem. 1991;266:18895–18906. [PubMed] [Google Scholar]
• 28.Carey J. Gel retardation. Methods Enzymol. 1991;208:103–117. doi: 10.1016/0076-6879(91)08010-f. [DOI] [PubMed] [Google Scholar]
• 29.Majka J., Jakimowicz D., Messer W., Schrempf H., Lisowski M., Zakrzewska-Czerwińska J. Interactions of the Streptomyces lividans initiator protein DnaA with its target. Eur. J. Biochem. 1999;260:325–335. doi: 10.1046/j.1432-1327.1999.00168.x. [DOI] [PubMed] [Google Scholar]
• 30.Jakimowicz D., Majka J., Messer W., Speck C., Fernandez M., Martin M. C., Sanchez J., Schauwecker F., Keller U., Schrempf H., Zakrzewska-Czerwińska J. Structural elements of the Streptomyces oriC region and their interactions with the DnaA protein. Microbiology. 1998;144:1281–1290. doi: 10.1099/00221287-144-5-1281. [DOI] [PubMed] [Google Scholar]
• 31.Qin M. H., Madiraju M. V., Rajagopalan M. Characterization of the functional replication origin of Mycobacterium tuberculosis. Gene. 1999;233:121–130. doi: 10.1016/s0378-1119(99)00148-1. [DOI] [PubMed] [Google Scholar]
• 32.Rajagopalan M., Qin M.-H., Nash D. R., Madiraju M. V. V. S. Mycobacterium smegmatis dnaA region and autonomous replication activity. J. Bacteriol. 1995;177:6527–6535. doi: 10.1128/jb.177.22.6527-6535.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 33.Zawilak A., Cebrat S., Mackiewicz P., Krol-Hulewicz A., Jakimowicz D., Messer W., Gosciniak G., Zakrzewska-Czerwińska J. Identification of a putative chromosomal replication origin from Helicobacter pylori and its interaction with the initiator protein DnaA. Nucleic Acids Res. 2001;29:2251–2259. doi: 10.1093/nar/29.11.2251. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 34.Majka J., Zakrzewska-Czerwińska J., Messer W. Sequence recognition, cooperative interaction, and dimerization of the initiator protein DnaA of Streptomyces. J. Biol. Chem. 2001;276:6243–6252. doi: 10.1074/jbc.M007876200. [DOI] [PubMed] [Google Scholar]
• 35.Jakimowicz D., Majka J., Konopa G., Wegrzyn G., Messer W., Schrempf H., Zakrzewska-Czerwińska J. Architecture of the Streptomyces lividans DnaA protein–replication origin complexes. J. Mol. Biol. 2000;298:351–364. doi: 10.1006/jmbi.2000.3686. [DOI] [PubMed] [Google Scholar]
• 36.Zakrzewska-Czerwińska J., Nardmann J., Schrempf H. Inducible transcription of the dnaA gene from Streptomyces lividans 66. Mol. Gen. Genet. 1994;242:440–447. doi: 10.1007/BF00281794. [DOI] [PubMed] [Google Scholar]
• 37.Margulies C., Kaguni J. M. Ordered and sequential binding of DnaA protein to oriC, the chromosomal origin of Escherichia coli. J. Biol. Chem. 1996;271:17035–17040. doi: 10.1074/jbc.271.29.17035. [DOI] [PubMed] [Google Scholar]
• 38.Schaper S., Messer W. Interaction of the initiator protein DnaA of Escherichia coli with its DNA target. J. Biol. Chem. 1995;270:17622–17626. doi: 10.1074/jbc.270.29.17622. [DOI] [PubMed] [Google Scholar]
• 39.Wentzell L. M., Halford S. E. DNA looping by the SfiI restriction endonuclease. J. Mol. Biol. 1998;281:433–444. doi: 10.1006/jmbi.1998.1967. [DOI] [PubMed] [Google Scholar]
• 40.Langer U., Richter S., Roth A., Weigel C., Messer W. A comprehensive set of DnaA-box mutations in the replication origin, oriC, of Escherichia coli. Mol. Microbiol. 1996;21:301–311. doi: 10.1046/j.1365-2958.1996.6481362.x. [DOI] [PubMed] [Google Scholar]
• 41.Jakimowicz D., Majka J., Lis B., Konopa G., Wegrzyn G., Messer W., Schrempf H., Zakrzewska-Czerwińska J. Structure and regulation of the dnaA promoter region in three Streptomyces species. Mol. Gen. Genet. 2000;262:1093–1102. doi: 10.1007/pl00008652. [DOI] [PubMed] [Google Scholar]
• 42.McGarry K. C., Ryan V. T., Grimwade J. E., Leonard A. C. Two discriminatory binding sites in the Escherichia coli replication origin are required for DNA strand opening by initiator DnaA–ATP. Proc. Natl. Acad. Sci. U.S.A. 2004;101:2811–2816. doi: 10.1073/pnas.0400340101. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 43.Seitz H., Welzeck M., Messer W. A hybrid bacterial replication origin. EMBO Rep. 2001;2:1003–1006. doi: 10.1093/embo-reports/kve225. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 44.Brendler T. G., Abeles A. L., Reaves L. D., Austin S. J. The iteron bases and spacers of the P1 replication origin contain information that specifies the formation of a complex structure involved in initiation. Mol. Microbiol. 1997;23:559–567. doi: 10.1046/j.1365-2958.1997.d01-1869.x. [DOI] [PubMed] [Google Scholar]
• 45.Tjian R. The binding site on SV40 DNA for a T antigen-related protein. Cell. 1978;13:165–179. doi: 10.1016/0092-8674(78)90147-2. [DOI] [PubMed] [Google Scholar]
• 46.Chen G., Stenlund A. Characterization of the DNA-binding domain of the bovine papillomavirus replication initiator E1. J. Virol. 1998;72:2567–2576. doi: 10.1128/jvi.72.4.2567-2576.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 47.Chen G., Stenlund A. The E1 initiator recognizes multiple overlapping sites in the papillomavirus origin of DNA replication. J. Virol. 2001;75:292–302. doi: 10.1128/JVI.75.1.292-302.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 48.Chen G., Stenlund A. Sequential and ordered assembly of E1 initiator complexes on the papillomavirus origin of DNA replication generates progressive structural changes related to melting. Mol. Cell. Biol. 2002;22:7712–7720. doi: 10.1128/MCB.22.21.7712-7720.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]