Remodeling of Nucleoprotein Complexes Is Independent of the Nucleotide State of a Mutant AAA+ Protein

Abstract

DnaA protein, a member of the AAA+ (ATPase associated with various cellular activities) family, initiates DNA synthesis at the chromosomal origin of replication (oriC) and regulates the transcription of several genes, including its own. The assembly of DnaA complexes at chromosomal recognition sequences is affected by the tight binding of ATP or ADP by DnaA. DnaA with a point mutation in its membrane-binding amphipathic helix, DnaA(L366K), previously described for its ability to support growth in cells with altered phospholipid content, has biochemical characteristics similar to those of the wild-type protein. Yet DnaA(L366K) fails to initiate in vitro or in vivo replication from oriC. We found here, through in vitro dimethyl sulfate footprinting and gel mobility shift assays, that DnaA(L366K) in either nucleotide state was unable to assemble into productive prereplication complexes. In contrast, at the dnaA promoter, both the ATP and the ADP form of DnaA(L366K) generated active nucleoprotein complexes that efficiently repressed transcription in a manner similar to wild-type ATP-DnaA. Thus, it appears that unlike wild-type DnaA protein DnaA(L366K) can adopt architectures that are independent of its bound nucleotide, and instead the locus determines the functionality of the higher order DnaA(L366K)-DNA complexes.

Introduction

In eukaryotes and prokaryotes protein-DNA complexes regulate a variety of cellular functions. Some nucleoprotein structures include proteins that contain an AAA+ (ATPase associated with various cellular activities) domain, such as polymerase clamp loader proteins (1, 2), transcriptional regulators (3), eukaryotic replication initiator protein Cdc6 (4), and the bacterial homologue DnaA (4–6).

The Escherichia coli chromosome possesses ∼300 recognition sites for DnaA protein (7). The unique 245-base pair origin of replication (oriC)² harbors tandem repeats of five 9-mers, termed R boxes, that largely adhere to the consensus sequence TGTGNA(T/A)AA (8, 9). Although DnaA protein tightly associates with ATP and ADP (K_D of 0.03 and 0.1 μm, respectively) (10) and both forms have high affinity for sites R1, R2, and R4 (11, 12), only ATP-DnaA is active for initiating new rounds of replication in vitro (10, 13) or in vivo (14).

Several low affinity DnaA binding sites, known as I sites and tau sites, flank the center of oriC and differ from the R box consensus by 3–4 bases (11, 12, 15). Among them, I2, I3, and tau sites discriminate between ATP-DnaA and ADP-DnaA with a preference for the ATP form (12). Throughout most of the bacterial cell cycle, DnaA protein remains bound to the high affinity sites R1, R2, and R4, forming a structure similar to the eukaryotic origin recognition complex (16–19). At the time of initiation, these high affinity sites facilitate loading of additional DnaA to the low affinity sites, generating a functional prereplication complex (pre-RC) (19).

In addition to oriC, consensus sequences for DnaA binding are located throughout the E. coli genome, including within promoter regions of several genes, such as rpoH (20), mioC (21), nrdAB (22), and dnaA itself (23, 24), which contains two promoters, dnap1 and dnap2 (23–25). The dnaA promoter region contains DnaA recognition motifs, boxes a, b, and c (consensus sequence of AGATCT), that like sites I2 and I3 in oriC are specific for the ATP form of DnaA (25). In contrast, DnaA boxes 1 and 2, located between dnap1 and dnap2, have a 9-mer consensus sequence similar to that for R boxes in oriC. Nucleotide-dependent assembly of DnaA within the dnaA promoter region regulates dnaA expression; ATP-DnaA represses transcription of the dnaA gene from both promoters to differing degrees, whereas ADP-DnaA has no such effect (25).

Acidic phospholipids present in a fluid bilayer stimulate the release of bound nucleotide from DnaA in vitro and in the presence of oriC exchange DnaA-bound ADP for ATP (26, 27). Depletion of the acidic phospholipids phosphatidylglycerol and cardiolipin in the E. coli membrane through controlled repression of phosphatidylglycerophosphate synthase (product of pgsA) results in arrested growth but not a loss of cell viability (28). However, growth can be restored by overexpressing DnaA proteins that contain certain mutations within the DNA-binding or membrane-binding domains, including DnaA(L366K) (29). Although DnaA(L366K) as the sole form of DnaA is unable to initiate DNA synthesis in vivo (29) or in vitro (30), purified DnaA(L366K) is similar to wild-type DnaA for nucleotide binding affinities, ATP hydrolysis, and specificity for acidic phospholipids promoting nucleotide exchange (30). The stoichiometry of DnaA bound to oriC is still debatable, varying from as low as five DnaA molecules per origin as detected by electrophoretic mobility shift assays (31) and gel filtration (32), to 15 when measured by filter retention (10), to even 20–30 as suggested by electron microscopy (8, 34). Filter retention data (30) suggest that DnaA and DnaA(L366K) bind oriC comparably.

Using in vitro dimethyl sulfate (DMS) footprinting and electrophoretic mobility shift assays, we observed that DnaA(L366K) in either nucleotide form was poor at forming a proper pre-RC due to its inability to bind low affinity I2, I3, and R5M sites. Interaction at these sites could be restored partially in the presence of limiting quantities of wild-type DnaA, suggesting formation of functionally active hetero-oligomeric pre-RCs. These observations led us to consider whether the leucine to lysine substitution abolishes the discrimination that ATP-DnaA has for low affinity sites in forming nucleoprotein complexes. Therefore, higher order structures formed by DnaA(L366K) were examined at a different locus, the dnaA promoter region, which also has sites specific for ATP-DnaA. We observed that DnaA(L366K) binding to the dnaA promoter formed structures that repressed transcription independently of the nucleotide state of the protein. Based on these results from oriC and the dnaA promoter, we propose that DnaA(L366K) generates higher order structures regardless of its bound nucleotide, with active complexes at some loci and inactive complexes at other loci.

EXPERIMENTAL PROCEDURES

Reagents and Strains

DNA-modifying enzymes and DNA polymerase were obtained from New England Biolabs; RNA polymerase was from Sigma; and plasmid DNA preparation kits, PCR purification kits, nucleotide removal kits, and nickel-nitrilotriacetic acid-agarose matrix were purchased from Qiagen. Sequa Gel solution to prepare sequencing gels was obtained from National Diagnostics. [γ-³²P]ATP (3,000 Ci/mmol for labeling of footprinting primers and 6,000 Ci/mmol for labeling of electrophoretic mobility shift assay probes) and [α-³²P]UTP (3,000 Ci/mmol) were from PerkinElmer Life Sciences. Bacterial growth medium components were from Difco. Oligonucleotides used in this study were obtained from Integrated DNA Technologies. Strains EH3827 (CM1565 zia::pKN500 ΔdnaA mad-1) (35) and M182 (F⁻, Δ(codB-lacI)3, galK16, galE15(GalS), λ⁻, e14-?, relA1, rpsL150(str), spoT1) (36) were used for in vivo dnaA transcription assays. Plasmids pZL606 and pZL607 are as described previously (29).

Expression and Purification of Recombinant Proteins

Recombinant wild-type DnaA and DnaA(L366K) proteins were expressed and purified as described previously (30, 37). Protein concentrations were determined according to Bradford (38). In vitro DNA replication and ATP binding activities of DnaA protein were determined as described elsewhere (30, 37, 39).

DNA Footprinting

DnaA footprinting was performed as described previously (12, 15, 40). Briefly, in 50-μl reactions, E. coli oriC DNA (0.75 μg of pBSoriC) was incubated (7 min at 38 °C) with DnaA protein (as indicated) in 40 mm HEPES-KOH, pH 7.6, 8 mm MgCl₂, 30% glycerol, 320 μg/ml BSA, and 5 mm ATP or ADP. Following treatment with DMS (5 μl of a 1.4% solution; 5 min at 38 °C), samples were cleaved with piperidine (100 μl of a 10% solution) at 95 °C for 30 min, and DNA fragments were purified using Micro Bio-Spin 6 chromatography columns (Bio-Rad). Relevant fragments were examined by primer extension with primer RS4 (5′-TATACAGATCGTGCGATC-3′) that hybridizes to bases 272–290 at 52–53 °C to analyze the upper strand and primer SR4 (5′-GGATCATTAACTGTGAATG-3′) that hybridizes to bases 124–142 at 56–57 °C to analyze the lower strand. Both primers (150 pmol) used in the primer extension reactions were previously radiolabeled with 100 of μCi [γ-³²P]ATP and T4 polynucleotide kinase at 37 °C for 45 min. Primer extension reactions were stopped by adding quenching solution (4 m ammonium acetate and 20 mm EDTA), and DNA was precipitated by mixing with 95% ethanol and collected by centrifugation (15,000 × g for 5 min at 4 °C). Samples were air-dried and dissolved in gel loading buffer. Extension products were resolved on 6% acrylamide sequencing gels, and radioactive bands were detected with a GE Healthcare Storm 840 imager. Images were analyzed using Bio-Rad Quantity One software.

P1 Endonuclease Digestion

P1 endonuclease digestions were performed as described previously (40). Briefly, supercoiled oriC plasmid (60 fmol; pOC170) in 20 μl of 40 mm HEPES-KOH, pH 7.6, 8 mm MgCl₂, 30% glycerol, 320 μg/ml BSA, and 5 mm ATP or ADP was incubated with DnaA or DnaA(L366K) (as indicated) for 5 min at 38 °C. P1 endonuclease (1.2 units) was added for 10 s, and then the endonuclease reaction was stopped by the addition of stop solution (2% SDS and 50 mm EDTA). Samples were heated (65 °C) for 5 min, and DNA species were resolved by electrophoresis through a 1% agarose gel and stained with ethidium bromide. Quantitation of the linear form was done using Bio-Rad Quantity One software.

Electrophoretic Mobility Shift Assay

Complementary single-stranded oligonucleotides (5′-CAGTCATTGGTTATACACAGACTTCCTGACAGAGTTATCCACAGTAGATCGCA-3′ and 5′-TGCGATCTACTGTGGATAACTCTGTCAGGAAGTCTGTGTATAACCAATGACTG-3′) were resuspended to a concentration of 10 μg/μl in deionized water. DNA strands were annealed by incubating at 90 °C for 10 min followed by a 50 °C incubation for 10 min and then cooled to room temperature. Annealed oligonucleotide (100 μg) was applied to a 10% non-denaturing polyacrylamide gel in Tris borate-EDTA buffer (45 mm Tris borate and 1 mm EDTA). The double-stranded fragment was localized in the gel by UV shadowing, excised, and incubated in 0.3 m sodium acetate, pH 4.5 overnight at 37 °C. Fragments were extracted with 24:1:1 phenol:chloroform:isoamyl alcohol and precipitated with 95% ethanol. The double-stranded fragments were radiolabeled as described above.

Reactions contained 20 mm HEPES-KOH, pH 8.0, 2.5 mm magnesium acetate, 1 mm EDTA, 4 mm DTT, 100 μm ATP, 5 mg/ml BSA, 0.2% Triton X-100, 5% glycerol, and 4 nm radiolabeled fragment. Either 1 nm wild-type DnaA or DnaA(L366K) was added, and the mixture was incubated for 8 min at 38 °C. Additional wild-type DnaA or DnaA(L366K) (2, 5, 10, 20, 40, 80, or 160 nm) was added, and incubation was continued for an additional 8 min. Resulting complexes were resolved by non-denaturing polyacrylamide gel (7%) electrophoresis in Tris borate-EDTA buffer. Dried gels were scanned on a Bio-Rad Molecular Image FX phosphorimaging system.

DNase I Protection Assay

A DNA fragment of ∼200 base pairs containing dnaAp1 and dnaAp2 was amplified by PCR using forward primer 5′-AATTTTCCAATGCCGCGTAAATCGTGC-3′ and reverse primer 5′-GAACCGCTGTCTGCGGTTATATGC-3′. Two consecutive sets of PCRs were carried out using Taq DNA polymerase at an annealing temperature of 60 °C. In each set, one of the two primers used for amplification was radiolabeled with [γ-³²P]ATP and T4 polynucleotide kinase. The ATP and ADP forms of the wild-type and mutant proteins were generated as described earlier (25). The indicated amounts of protein were incubated (10 min at 37 °C) with the radiolabeled PCR product (20 ng) in binding buffer (25 mm HEPES, pH 7.6, 100 mm potassium acetate, 5 mm magnesium acetate, 5 mm calcium acetate, 4 mm DTT, 0.2% Triton X-100, 0.5 mg/ml BSA, and 3 mm ATP or ADP). DNA digestion reactions were initiated by the addition of Dnase I (20 pg), samples were incubated for 4 min at 37 °C, and then the reaction was terminated with the addition of stop buffer (20 μl; 1% SDS, 200 mm NaCl, 20 mm EDTA, pH 8.0, and 0.1 mg/ml glycogen). Protein was removed by phenol-chloroform extraction, and the DNA was precipitated with ethanol. Samples were air-dried, dissolved in gel loading buffer, and applied to 8% acrylamide sequencing gels. Radioactive bands were detected with a GE Healthcare Storm 840 imager, and images were analyzed by using Bio-Rad Quantity One software.

In Vitro Transcription

Single round runoff transcription assays were performed as described elsewhere (25) with minor modifications. Plasmid pLSK4 (10 ng), which contains the dnaA promoter region, was used as template DNA in reaction buffer containing 25 mm HEPES, pH 7.6, 100 mm potassium acetate, 5 mm magnesium acetate, 4 mm DTT, 0.2% Triton X-100, 0.5 mg/ml BSA, and 3 mm ATP or ADP. The indicated amounts of wild-type or mutant protein were added, and the mixture was incubated for 5 min at 37 °C. RNA polymerase (66 nm) was added, and samples were incubated at the same temperature for an additional 5 min. In vitro transcription was initiated by the addition of [α-³²P]UTP (7.5 μCi; 3,000 Ci/mmol), unlabeled UTP (60 μm), ATP (200 μm), CTP (400 μm), and GTP (400 μm), and heparin (200 μg/ml). After 5 min, transcription was terminated by the addition of stop buffer (20 μl). Samples were purified and resolved (6% acrylamide sequencing gels), and bands were analyzed as described for DNase I footprint assays.

In Vivo β-Galactosidase Reporter Assay

E. coli EH3827 (35) or M182 (36) was lysogenized using a λ phage containing a reporter gene comprising the dnaA promoter driving expression of β-galactosidase (23). The lysogens were transformed with plasmids containing an arabinose-inducible dnaA or dnaA(L366K) gene. Empty vector (pBAD) was used as a control. Cells were grown exponentially in LB medium in the presence of arabinose (0.2%) for 10 generations, and the amount of β-galactosidase activity was measured using a Miller assay (41).

The levels of DnaA and DnaA(L366K) overexpression in each strain were determined by immunoblotting whole-cell lysates prepared from uninduced and induced cells with anti-DnaA antiserum. The intensities of resulting immunoreactive bands were determined using NIH ImageJ software. The amount of sample loaded in each lane was normalized to the equivalent of 1.0 ml of culture at an A_{600 nm} of 0.5. To confirm that equivalent amounts of cell lysates were compared, the membranes of the anti-DnaA immunoblots were stripped and reprobed with anti-OmpA protein antiserum.

RESULTS

DnaA(L366K) Binds oriC High Affinity Sites in Manner Similar to DnaA

DMS footprinting was performed to examine the interaction of DnaA and DnaA(L366K) with the high affinity oriC sites R1, R2, and R4. DnaA binding enhanced the methylation of the guanosine residue present at the fourth position (G4) of the high affinity R boxes and suppressed modification of the guanosine at the second position (G2 if present) (Fig. 1). To minimize the binding of wild-type DnaA to the low affinity sites, the ADP form of the proteins was used because its known that ADP-DnaA protein primarily engages the high affinity sites (12). Purified DnaA or DnaA(L366K) was incubated with ADP to generate the ADP form of each protein, which was then mixed with oriC-containing plasmids in the presence of ADP and assessed for DNA binding by DMS footprinting. The ∼3,600 base pairs of the plasmids harboring oriC served as competitor DNA, preventing nonspecific binding of DnaA protein to oriC. Depending on the amount of protein present, the intensity of the G4 band in box R2 increased ∼2.5-fold (with 20 nm protein) to 4.5-fold (with 160 nm protein) for both types of DnaA protein (Fig. 1). Binding of ADP-DnaA(L366K) was also similar to wild-type ADP-DnaA at boxes R1 and R4 as indicated by the ∼1.7–2.3-fold (20 and 160 nm protein, respectively) increases in band intensities for box R1 and 2.0–2.6-fold increases for box R4 (Fig. 1). Thus, DnaA(L366K) seems able to form origin recognition complex-like structures involving the high affinity sites similar to wild-type DnaA.

FIGURE 1.

DnaA(L366K) forms an origin recognition complex by binding to oriC sites R1, R2, and R4. Top, ADP-DnaA or ADP-DnaA(L366K) at 20 and 160 nm was incubated with oriC DNA in the presence of 5 mm ADP, and modifications to DMS footprint patterns caused by bound protein were assessed. Sites R1, R5, I2, I3, R2, and R4 and the guanines at positions 2 and 4 within each site are marked. Primer extension was carried out using primer SR4 and primer RS4. Note that I3 lacks a guanine at position 2. Bottom, band intensities at positions G2 and G4 are plotted relative to the intensity of the corresponding band in the no protein (0 nm) control lane. Error bars denote S.D. from three independent experiments. Each bar on the graph (left to right) at the G2 and G4 positions of a recognition site corresponds to the gel electrophoresis lane (top; from left to right) for that site.

L366K Mutation Impairs Binding of DnaA to oriC Low Affinity Sites

Interactions of the ATP forms of DnaA and DnaA(L366K) with the high affinity boxes and low affinity binding sites (I1, I2, I3, and R5M) were also examined by DMS footprinting. When present at 160 nm, both wild-type and mutant DnaA caused an approximately 2.2-fold increase in the intensity of the G4 band of R4 box and a 4.5-fold increase for the R2 site (Fig. 2B). Even at the lower DnaA concentration of 20 nm, a 2.0–2.5-fold increase in band intensities due to wild-type and mutant proteins was attained at these sites (Fig. 2B). A lower but still significant increase was observed at box R1 with either protein even at their higher concentration (Fig. 2A). Combined, these data support that ATP-DnaA(L366K) behaves like wild-type DnaA with respect to stable binding to high affinity R boxes and thus in forming an origin recognition complex.

FIGURE 2.

ATP-DnaA(L366K) is feeble at binding low affinity sites but can form a pre-RC when augmented with low amount of DnaA. A and B, purified DnaA(WT) and DnaA(L366K) proteins (20 and 160 nm) were incubated with oriC in the presence of ATP, and modifications to DMS footprint patterns caused by bound protein were assessed. Primer extension was carried out using primer SR4 (A) or primer RS4 (B). C and D, a mixture at a molar ratio of 1:9 of DnaA(WT) and DnaA(L366K), respectively, was incubated (20 and 160 nm total protein) with oriC in the presence of ATP, and modifications to DMS footprint patterns were assessed. Primer extension was carried out using primer SR4 (C) or primer RS4 (D). Band intensities are plotted as in Fig. 1. The left panel corresponds to gels A and B. The right panel corresponds to gels C and D. Error bars denote S.D. from three independent experiments.

In contrast, ATP-DnaA(L366K) differed from ATP-DnaA for binding to low affinity DnaA binding sites I2, I3, and R5M. Although wild-type DnaA produced a 2–4-fold increase in G4 band intensities at these sites, little difference was observed with the mutant protein even at a concentration of 160 nm (Fig. 2, A and B). The binding of ATP-DnaA to I2, I3, and R5M is a necessary step in pre-RC formation and a prerequisite for melting the duplex 13-mer AT-rich region of oriC (12); on wild-type oriC, filling of the low affinity sites with ATP-DnaA is coincident with origin unwinding (40), and if loading to low affinity sites is blocked, origin unwinding is prevented (12, 40). Likewise, ADP and ATP forms of DnaA(L366K), each unable to engage the low affinity sites (Fig. 2), were as poor as or even more so than ADP-DnaA at promoting strand opening of supercoiled oriC DNA as assessed by P1 endonuclease digestion (40) (Fig. 3). Hence, the mutant protein in either its ATP or ADP form like ADP-DnaA cannot form a pre-RC and thus is unable to initiate replication.

FIGURE 3.

Neither ATP- or ADP-DnaA(L366K) can unwind oriC. Top, supercoiled oriC plasmid was incubated with 0, 10, 20, 40, 80, 100, or 200 nm DnaA or DnaA(L366K) in the presence of ATP (top panel) or ADP (bottom panel). After P1 endonuclease digestion, DNA species were resolved by electrophoresis. The supercoiled (sc), nicked, and linear forms are marked. Bottom, quantitation of the linear form generated by P1 digestion in the presence of DnaA or DnaA(L366K).

ATP-DnaA(L366K) Forms Functional Hetero-oligomeric Complexes with ATP-DnaA

Previously it has been observed that supplementing DnaA(L366K) with a limiting amount of DnaA results in efficient in vitro replication from oriC (30). To gain insight on the molecular mechanism of how DnaA(L366K) participates in initiating replication, a mixture of wild-type and mutant DnaA proteins in a molar ratio of 1:9 was incubated with DNA in the presence of ATP and evaluated for oriC binding. Whereas marginal binding was observed at sites R5M, I1, I2, and I3 with either 20 nm wild-type ATP-DnaA or 160 nm ATP-DnaA(L366K) alone (Fig. 2, A and B), the extent of binding with 160 nm total protein (16 nm and 144 nm wild-type and DnaA(L366K), respectively) resulted in a 1.6–2.5-fold increase in band intensities at these low affinity sites (Fig. 2, C and D), indicating that DnaA(L366K) in a mixed oligomer with wild-type DnaA can efficiently generate functional pre-RCs identical to those produced solely by wild-type DnaA.

DnaA Cannot Assist Loading of DnaA(L366K) to Low Affinity Sites

DnaA is unable to bind low affinity sites directly (19). Rather, DnaA must be donated to the weak site by DnaA at a proximal higher affinity site. It is possible that DnaA(L366K) is inactive in this donation but could bind to weak sites if assisted by DnaA. To investigate this, we used gel shift assays to examine binding of DnaA to a DNA fragment containing two strong sites (R4 and R2) flanking a low affinity DnaA site (21).³ Binding of DnaA to this fragment resulted in the formation of three complexes with different electrophoretic mobilities (Fig. 4). Control assays (data not shown) in which one or more of the DnaA binding sites was converted to a non-binding form indicated that R4 occupation corresponds to complex 1, R4 and R2 binding corresponds to complex 2, and occupation of all three sites results in complex 3 (Fig. 4). Both DnaA and DnaA(L366K) were proficient at forming complexes 1 and 2 (Fig. 4, A and B), consistent with DnaA(L366K) binding to strong R boxes. When increasing concentrations of ATP-DnaA were added to a fragment that was prebound with ATP-DnaA(L366K) (Fig. 4C), complex 3 was formed at levels only slightly lower than those seen with DnaA alone (Fig. 4A), indicating that DnaA(L366K) at high affinity sites was able to donate DnaA to weak sites. However, even if assisted by DnaA, DnaA(L366K) was unable to engage the weak site (Fig. 4D), indicating that the mutant DnaA is deficient in some aspect of weak site occupation.

FIGURE 4.

DnaA(L366K) cannot bind low affinity sites but when bound to a strong site can assist loading of wild-type DnaA to low affinity site. A–D, a radiolabeled DNA fragment containing R4 and R2 flanking a weak site was incubated with DnaA(WT) or DnaA(L366K) (1 nm; top bars) and then with increasing concentrations (2–160 nm) of the indicated protein. DnaA-DNA nucleoprotein complexes were separated by non-denaturing gel electrophoresis. The unbound probe and complexes containing one, two, or three molecules of DnaA bound to the fragment are indicated.

Complex Formation by Either Nucleotide Form of DnaA(L366K) Is Not Unique to oriC

The above findings suggest the L366K mutation impairs the discrimination between ATP-DnaA and ADP-DnaA in the assembly of nucleoprotein complexes involving DnaA binding sites. To test whether this holds true for loci other than oriC, we used an established method (25) to examined the architecture of complexes produced by DnaA(L366K) at the dnaA promoter where it is known that the ATP and ADP forms of DnaA differentially interact. Binding of DnaA to specific DNA sequences located between the dnaAp1 and dnaAp2 promoter elements generates sites hypersensitive to DNase I on both the lower and upper strands of promoter region DNA (25).

Here, DnaA and DnaA(L366K) in their ATP or ADP forms were incubated with a DNA fragment containing the dnaA promoter, and resulting complexes were assessed using a DNase I protection assay as described previously (25). Changes in band intensities were compared with the intensity of the respective bands generated by DNase I digestion in the absence of DnaA protein. Both ATP-DnaA(L366K) and ADP-DnaA(L366K) conferred protection to DnaA box 1 at 50 nm. Similarly, protection at DnaA box 2 was achieved by either nucleotide form of DnaA(L366K) at concentrations slightly higher than 100 nm (Fig. 5, A and B). These results suggest that both ATP-DnaA(L366K) and ADP-DnaA(L366K) like either nucleotide form of wild-type DnaA (Ref. 24 and supplemental Fig. 1) protect DnaA box 1 and DnaA box 2.

FIGURE 5.

ATP-DnaA(L366K) and ADP-DnaA(L366K) confer the same DNase I protection to dnaA promoter as ATP-DnaA. A radiolabeled DNA fragment containing the dnaA promoter region was incubated with ATP-DnaA(L366K) or ADP-DnaA(L366K). Wedges above the gel lanes represent final concentrations of 50, 100, 200, and 300 nm for each protein. DNase I footprints were obtained for the lower strand (A) and upper strand (B). DnaA boxes 1 and 2 along with ATP-DnaA boxes a, b, and c are indicated. Promoter region elements −10 and −35 sites within dnaAp1 and dnaAp2 are also shown. Changes in intensities of representative bands (*) within ATP-DnaA boxes a, b, and c are plotted relative to the intensities of the corresponding bands in the no protein (0 nm) control lane. Note that limited electrophoretic resolution prevented accurate measurement of representative band intensities for lower strand box a. Error bars denote S.D. from three independent experiments. Each bar on the graph (left to right) corresponds to the gel electrophoresis lane (top; from left to right) for indicated amount of ATP-DnaA(L366K) and ADP-DnaA(L366K) protein.

Binding of both nucleotide forms of DnaA(L366K) was analyzed at the ATP-specific DnaA boxes a, b, and c. As seen previously (25), we found that ADP-DnaA protein at any concentration used caused no to little changes at ATP-DnaA boxes a, b, and c (supplemental Fig. 1). The slight changes in DNase I sensitivity offered by ADP-DnaA at ≥200 nm to the lower strand of box b and the upper strand of boxes a and c are in close agreement with earlier studies (25). On the other hand, ATP-DnaA (supplemental Fig. 1) and more interestingly both the nucleotide forms of DnaA(L366K) at concentrations between 100 and 300 nm yielded ∼1.5–4.0-fold changes in DNase I sensitivity at each ATP-DnaA box (Fig. 5). Thus, these results suggest that DnaA(L366K) unlike wild-type DnaA generates similar nucleoprotein complexes at the dnaA promoter regardless of whether the mutant form of the protein is in its ATP or ADP form.

Both Nucleotide Forms of DnaA(L366K) Repress Transcription from dnaA Promoter in Vitro

At oriC, ATP-DnaA(L366K) and ADP-DnaA(L366K) behaved like ADP-DnaA with respect to structures formed (Figs. 1–4), corresponding to its feeble initiation activity (30). In contrast, at the dnaA promoter, DnaA(L366K) in both its ATP and ADP forms not only produced structures like ATP-DnaA (Fig. 5), but both forms also repressed transcription like ATP-DnaA (Fig. 6). In vitro runoff transcription from template DNA by RNA polymerase was performed and monitored for the generation of transcripts from dnaAp1 (295 bp) and dnaAp2 (212 bp) (25). At 50 nm, ATP-DnaA showed 40% repression from dnaAp1 and 60% repression from dnaAp2 (Fig. 6A). Whereas almost complete repression (90%) was found at the higher concentration (300 nm) of ATP-DnaA, little if any repression was observed at 50 nm ADP-DnaA, and only up to 50% repression was seen at 300 nm from either dnaAp1 or dnaAp2 (Fig. 6A). However, at 50 nm, both ATP-DnaA(L366K) and ADP-DnaA(L366K) were only slightly weaker than ATP-DnaA for repressing transcription from dnaAp1, and at 300 nm, they were essentially identical to ATP-DnaA, causing ∼80–85% repression (Fig. 6B). Similarly, at dnaAp2, both nucleotide forms of DnaA(L366K) at 300 nm yielded 85% repression, indistinguishable from that caused by ATP-DnaA. Thus, these results are in agreement with those from the DNase I footprint studies: DnaA(L366K) in either the ATP or ADP form acts like ATP-DnaA at the dnaA promoter region.

FIGURE 6.

DnaA(L366K) in either nucleotide form represses in vitro transcription from dnaA promoter similar to ATP-DnaA. The indicated amounts of ATP and ADP forms of DnaA(WT) (A) and DnaA(L366K) (B) were incubated with the template DNA, and RNA synthesis was initiated by the addition of RNA polymerase and nucleotide triphosphates. The resulting transcripts from dnaAp1 and dnaAp2 are indicated. Effects of the ATP and ADP forms of DnaA(WT) and DnaA(L366K) on transcript production (percent expression) are relative to that obtained with no protein (bottom). Values represent the mean ± S.D. from three independent set of experiments.

DnaA(L366K) Represses Transcription from dnaA Promoter in Vivo

Because DnaA(L366K) was functional in repressing transcription from the DnaA promoter in vitro, we next examined whether it also acts as a repressor in vivo. Two different strains of exponentially growing cells containing a reporter gene comprising the dnaA promoter driving β-galactosidase expression were induced for plasmid-based DnaA or DnaA(L366K) overexpression and assayed for β-galactosidase activity (41). In both strains, approximately 10-fold increased expression of DnaA or DnaA(L366K) resulted in dnaA promoter activity repressed by ∼50% (Table 1), consistent with the in vitro findings at 50 nm DnaA and the 2–3-fold repression in vivo seen by others (23). It did not escape our notice that a higher degree of repression was achievable in vitro with 300 nm DnaA (Fig. 6) than was seen in vivo (23) (Table 1). However, it is not uncommon for in vivo effects to be more subtle with multiple factors often contributing to the control of biological processes. Thus, it appears that although DnaA(L366K) in either nucleotide form is not proficient in building the higher order structure that assists in the initiation of DNA replication it is capable of forming the functional nucleoprotein complex associated with another DnaA activity, autorepression.

TABLE 1.

DnaA(L366K) is competent at repressing the dnaA promoter in vivo

E. coli strain	Induced protein expressed	Amount of overexpression relative to uninduced	β-Galactosidase activity
			Miller units
EH3827/pBAD	None	1.1	250
EH3827pZL606	DnaA	12	130
EH3827pZL607	DnaA(L366K)	8.9	125
M182/pBAD	None	1.0	207
M182/pZL606	DnaA	12	100
M182/pZL607	DnaA(L366K)	9.7	80

DISCUSSION

We have established that there is a lack of discrimination between the ATP and ADP forms of DnaA(L366K) in the formation of higher order nucleoprotein structures. Surprisingly, the activity of such complexes varied depending on the locus in question: at oriC, DnaA(L366K) in either nucleotide form behaved like inactive ADP-DnaA, whereas at the dnaA promoter, it behaved like active ATP-DnaA.

Assembly of productive wild-type DnaA complexes at oriC and the dnaA promoter requires both the ATP form of the protein (12, 25) and cooperative binding of individual protomers to multiple cognate recognition sequences (19, 31, 42–45). A defective DnaA protomer, such as DnaA(L366K), if unable to bind to low affinity DnaA sites would logically be expected to negatively affect generation of a functional complex. However, this result was observed only at oriC (30). In contrast, at the dnaA promoter, both nucleotide forms of DnaA(L366K) were competent to bind all cognate recognition sequences and form a complex that repressed transcription. Thus, our studies with DnaA(L366K) suggest that the nucleotide sequence of an individual locus also is a determinant of functional DnaA complex formation. In support of this, we have recently found that the positioning, orientation, and number of DnaA boxes modulate the functional nature of the complexes obtained.³

High resolution structural knowledge of the different nucleotide forms of DnaA and DnaA(L366K) either alone or in a co-crystal with DNA of different loci would go far in solving the mystery of why DnaA (L366K) and wild-type DnaA behave differently. The L366K mutation is within an amphipathic helix that lies between nucleotide-binding domain III and DNA-binding domain IV (5, 13, 33, 46) and thus may cause some sort of an action-at-a-distance effect on the activity of DnaA protein. Among several possibilities are (i) conformational changes in the nucleotide binding motif that remove specificity for the nucleotide, (ii) alterations in the helix-turn-helix motif responsible for sequence-specific DNA binding, (iii) alterations in protomer-protomer interactions, and (iv) affects on the assembly of the nucleoprotein complex on a potential scaffold, such as a cytoskeletal element or the plasma membrane.