Molecular design of inhibitors of in vitro oriC DNA replication based on the potential to block the ATP binding of DnaA protein.

DnaA protein, the initiation factor for chromosomal DNA replication in Escherichia coli, is activated by binding to ATP. We earlier reported that 3-acetoxy-2,2'-bi-1H-indol inhibited the ATP binding to DnaA protein (Sasaki, S., Mizushima, T., Hashimoto, T., Maeda, M., and Sekimizu, K. (1994) Bioorg. Med. Chem. Lett. 4, 1771-1774). In the present study, derivatives of 3-acetoxy-2,2'-bi-1H-indol with different lengths of aliphatic chains at the 3-O position were synthesized, and their potential to inhibit the ATP binding to DnaA protein was examined. Elongation of the aliphatic chain resulted in inhibition of the ATP binding to DnaA protein at lower concentrations. Among the derivatives, 3-[N-(11-carboxyundecyl)]carbamoylmethoxy-2,2'-bi-1H-indol (structure 7 (3-CUCM-BI)) exhibited the most potent inhibition with an IC50 value of 7 microM. The mode of the inhibition was competitive. We further demonstrated that structure 7 (3-CUCM-BI) inhibited DNA replication of the oriC plasmid in a system reconstituted from purified proteins. This inhibition was specific for the initiation of DNA replication rather than for the elongation. The inhibition was overcome by preincubation of DnaA protein with ATP. Furthermore, structure 7 (3-CUCM-BI) showed little inhibition on DNA synthesis in the ABC primosome system. We propose that structure 7 (3-CUCM-BI) functions in the in vitro oriC DNA replication by inhibiting the ATP binding to DnaA protein.

Replication of chromosomal DNA in Escherichia coli is regulated at the step of initiation. DnaA protein is the initiation factor for chromosomal DNA replication (1)(2)(3); thus, DnaA protein has been considered to play an important role in regulating DNA replication. DnaA protein has a high affinity for ATP (K d ϭ 0.03 M) and for ADP (K d ϭ 0.1 M) (4). In the oriC DNA replication system reconstituted from purified proteins, the ATP binding form of DnaA protein is active in DNA replication, whereas the ADP binding form is inactive (4). These results suggest that the ATP binding to DnaA protein activates the protein; however, the possibilities that the ADP binding to the protein inhibits the activity of DnaA protein in the initiation of oriC DNA replication and that the ATP binding to the protein is not essential for the process would need to be excluded. Studies on DnaAcos protein, which loses the affinity for ATP and ADP but is active in the initiation of oriC DNA replication in vitro, imply this notion (5,6). To better understand the requirement of the ATP binding for the initiation of oriC DNA replication, development of specific inhibitors for the ATP binding to DnaA protein and examination of their effects on oriC DNA replication in vitro are important. The availability of such inhibitors would be good tools to study the biological relevance of the ATP binding to DnaA protein.
We reported that 3-acetoxy-2,2Ј-bi-1H-indol inhibited the ATP binding of DnaA protein (7). This indol is the first known synthetic organic compound to inhibit the ATP binding of DnaA protein. However, concentration of the drug necessary for inhibition is relatively high, which makes it difficult to examine effects on DNA replication, in vitro or in vivo. In the present work, we attempted to decrease the IC 50 value of the drug for inhibition of the ATP binding of DnaA protein by introducing hydrophobic residues and carboxyl residue to the 3-acetoxy-2,2Ј-bi-1H-indol. We obtained evidence that the most potent inhibitory compound, 7 (3-CUCM-BI), 1 inhibited oriC DNA replication in a system reconstituted from purified enzymes through specific inhibition of the ATP binding activity of DnaA protein.

EXPERIMENTAL PROCEDURES
Materials-DnaA protein was purified by the method described elsewhere (8), except that a newly constructed overproducer was used. 2 Specific activity of the protein was 0.7 ϫ 10 6 units/mg. Purity of the fraction used exceeded 90%, as determined by SDS-polyacrylamide gel electrophoresis.
Synthesis of Inhibitors-All the inhibitors used in this study were synthesized starting with indigo (1) via the key intermediate 2 (3-Ac-BI) (7), as shown in Fig. 1.
General-1 H-NMR spectra were taken at 500 or 270 MHz. Chemical shifts are reported in ppm downfield from tetramethylsilane. IR spectra were taken on a JASCO IR Report-100 spectrometer, and mass spectra were obtained with a JEOL JMS DX-300 or D-300 mass spectrometer. Column chromatography was done using silica gel BW200 (150 -350 mesh, Fuji Division).

3-Carboxymethoxy-2,2Ј-bi-1H-indol Potassium Salt (4 (3-KCM-BI)
)-A mixture of 3 (3-MCM-BI) (0.495 g, 1.54 mmol) and 1 N aqueous solution of NaOH (3.09 ml) in ethanol (104 ml) was stirred for 20 min at room temperature, and then the solvent was evaporated to produce a pale green powder (4 (3-KCM-BI), 0.657 g, quantitative). This salt was used for the next reaction, without further purification. mp 200 -208°C.  (5.5 ml) at Ϫ5°C, and the mixture was stirred for 3 h at the same temperature and overnight at room temperature. The reaction mixture was filtered, and the filtrate was diluted with AcOEt/benzene (1:1). The organic solvents were washed with brine and then dried over Na 2 SO 4 .
The solvent was evaporated, and the residue was chromatographed on a silica gel column (acetone:hexane ϭ 1:7) to afford the ethyl ester of 5 (3-CMCM-BI) (92.7 mg, 0.236 mmol, 83%) as a pale green powder. This was used for the next reaction without further purification.
A mixture of the above product (10.5 mg, 0.0268 mmol) and 1 N aqueous NaOH (47 l) in ethanol (1.9 ml) was stirred for 40 min at room temperature, and then the solvent was evaporated. The residue was chromatographed on a silica gel column (AcOE) to afford 5 (4.8 mg, 0.0132 mmol, 49%) as a pale yellow powder. mp 140 -145°C. 1  As the drugs were dissolved in Me 2 SO, the ATP binding reaction to DnaA protein and DNA replication reaction was performed in the presence of 1% Me 2 SO (including control experiment without drug), which did not affect the reactions.
ATP Binding Assay of DnaA Protein-Inhibitory effects by various compounds on the ATP binding activity of DnaA protein were determined by filter binding assay (4). DnaA protein (1 pmol) was incubated with a compound at 0°C for 15 min in 40 l of buffer G (50 mM HEPES/KOH, pH 8.0, 0.5 mM magnesium acetate, 0.3 mM EDTA, 5 mM dithiothreitol, 10 mM ammonium sulfate, 17% (v/v) glycerol, and 0.005% Triton X-100). Then, [␣-32 P]ATP (final concentration, 1 M) was added, the preparation was incubated at 0°C for 15 min, and samples were passed through nitrocellulose membranes. The radioactivity remaining on the filters was counted in a liquid scintillation counter.
Assay of Dissociation of the ATP⅐DnaA Complex-Influence of compounds on the release of ATP from the ATP⅐DnaA complex was examined as described (4). DnaA protein (1 pmol) and [␣-32 P]ATP (final concentration, 1 M) were incubated at 0°C for 15 min in 40 l of buffer G, and a synthetic organic compound was added. Incubation was continued at 37°C for various periods. Samples were passed through nitrocellulose filters, and radioactivity remaining on the filter was counted in a liquid scintillation counter.
oriC Replication System in Vitro Reconstituted from Purified Proteins-Reaction mixtures (9)  Samples were passed through Whatman GF/C glass-fiber filters. The amount of radioactivity on the filter was measured in a liquid scintillation counter, and the amount of DNA synthesized (pmol nucleotides) was calculated.
Replication of Single-stranded DNA by ABC Primosome-The DNA replication reaction by ABC primosome was done as described (10). M13 A-site single-stranded DNA (220 pmol as nucleotide) was added to the reaction mixture (25 l), and samples were incubated at 30°C for 10 min. The reaction was terminated by chilling on ice and adding 10% trichloroacetic acid. Samples were passed through glass fiber filters. The amount of radioactivity on the filter was measured in a liquid scintillation counter, and the amount of DNA synthesized (pmol nucleotides) was calculated.

Increase in the Inhibitory Activity of 3-Acetoxy-2,2Ј-bi-1Hindol for the ATP Binding of DnaA Protein by Introduction of
Alkyl Chains-We earlier reported that 2 (3-Ac-BI) inhibited the ATP binding of DnaA protein (7). We have now modified this compound to increase the inhibitory activity. Acidic phospholipids, cardiolipin and phosphatidylglycerol, inhibit the ATP binding of DnaA protein (11)(12)(13)(14). Fatty acid moieties are essential for inhibition as preincubation of phospholipids with phospholipase A 2 diminishes the inhibition (11). We thus assumed that introduction of hydrophobic residues to the bisindol skeleton would increase the inhibitory activity of the drug to the ATP binding to DnaA protein. We chose the 3-O site of bis-indol as the target for modification. Simple exchange of the acetyl group of 2 (3-Ac-BI) with methoxycarbonylmethyl (3 (3-MCM-BI)) did not increase the inhibitory activity of the drug to the ATP binding to DnaA protein; the IC 50 value for the inhibition of the ATP binding of DnaA protein increased about 10 times by this modification (data not shown).
As negative charges of lipid molecules are essential to inhibit the ATP binding to DnaA protein (13,14), we hydrolyzed the methyl ester residue of 3 (3-MCM-BI) and examined the inhibitory effect on the ATP binding of DnaA protein. The compound with the carboxyl residue (4 (3-KCM-BI)) was 10 times more potent than 3 (3-MCM-BI) (data not shown). Next, we planned to introduce a carboxyalkyl chain onto the structure of 4 (3-KCM-BI). Three amino acid derivatives having alkyl chains of different lengths between the amino and the carboxy residues were condensed with 4 (3-KCM-BI) to produce the inhibitors (5 (3-CMCM-BI), 6 (3-CPCM-BI), and 7 (3-CUCM-BI)) ( Fig. 1). When the longer alkyl chain was introduced, a more potent inhibition of the ATP binding to DnaA protein was obtained (Fig. 2) To study the role of the alkyl chain in the inhibition, we examined the influence of a mixture of 4 (3-KCM-BI) and 12-aminododecanoic acid methyl ester hydrochloride (8) on the ATP binding to DnaA protein. As shown in Fig. 3, the inhibitory activity of the mixture on the ATP binding is less than that of 7 (3-CUCM-BI) and similar to that of 4 (3-KCM-BI). 8 (12aminododecanoic acid methyl ester hydrochloride) showed little inhibitory effect (Fig. 3). These results suggest that the alkyl chain and bis-indol skeleton cooperate in inhibiting the ATP binding to DnaA protein.
To study the mode of the inhibition of the ATP binding of DnaA protein by 7 (3-CUCM-BI), we examined the titration of ATP on the ATP binding to DnaA protein in the presence and absence of the drug. Double reciprocal plot analysis revealed that the lines in the presence and absence of 7 (3-CUCM-BI) crossed on the Y axis despite that the lines were distinguishable (Fig. 4). The apparent K d values of the ATP binding to DnaA protein in the presence and absence of the drug were 4 and 0.2 M, respectively. Therefore, the inhibitory mode of 7 (3-CUCM-BI) on the ATP binding of DnaA protein is competitive.
Stimulation of Dissociation of the ATP⅐DnaA Complex by Drugs-We examined the influence of organic compounds we synthesized on stability of the ATP⅐DnaA complex. DnaA protein was preincubated with [␣-32 P]ATP, and each drug with a concentration 5 times exceeding the IC 50 (1 mM, 4 (3-KCM-BI); 280 M, 5 (3-CMCM-BI); 200 M, 6 (3-CPCM-BI); 20 M, 7 (3-CUCM-BI)) was added, followed by incubation at 37°C. The ATP⅐DnaA protein complex was stable in the absence of the drug; more than 90% of the complex remained in an intact form after incubation at 37°C for 5 min under these conditions (data not shown). The exchange rate of ATP bound to DnaA protein is low because even in the presence of excess amounts of nonradiolabeled ATP, the radiolabeled ATP⅐DnaA complex is stable as described previously (4). All the drugs almost completely inhibited the ATP binding of DnaA protein when added to DnaA protein prior to ATP (data not shown). This result means that the re-association of ATP to free DnaA protein does not occur under these conditions. All the drugs tested increased the dissociation rate of ATP bound to DnaA protein (Fig. 5). The stimulatory activity of drug for dissociation of the ATP⅐DnaA complex was higher when the longer alkyl chains were introduced into the 3-O site of 2 (3-AC-BI) (Fig. 5). Thus, 4 (3-KCM-BI), as well as 2 (3-AC-BI), stimulated little dissociation of the ATP⅐DnaA complex, whereas 7 (3-CUCM-BI) strongly enhanced the reaction (Fig. 5). The results suggest that the hydrophobic residues introduced contributed to not only inhibition of the ATP binding to DnaA protein but also to stimulation of the dissociation of the ATP⅐DnaA complex.
We also examined the influence of temperature on the stimulation by 7 (3-CUCM-BI) for the dissociation of the ATP⅐DnaA protein complex. Stimulation of the dissociation of the ATP⅐DnaA protein complex by 7 (3-CUCM-BI) required a high temperature; no dissociation was observed at 0°C, even in the presence of 7 (3-CUCM-BI) (Fig. 6). The result suggests that a higher order structure of DnaA protein is altered at a high temperature, which is probably essential for stimulation of the release of ATP from the complex by the drug. The activation energy for the dissociation of ATP⅐DnaA complex in the pres- ence of 20 M 7 (3-CUCM-BI) calculated from the Arrhenius plot was 120 kcal/mol, a value somewhat higher than the activation energy for the dissociation of the ATP⅐DnaA complex in the presence of acidic phospholipid (30 kcal/mol) (7).

Influence of 7 (3-CUCM-BI) on the oriC Plasmid DNA Replication in Vitro in a System
Reconstituted from Purified Proteins-Next, we examined the influence of 7 (3-CUCM-BI) on oriC DNA replication in vitro, an event dependent on the function of DnaA protein and requiring the ATP binding to DnaA protein (4). DnaA protein was preincubated with a drug and functioned in the oriC replication system reconstituted from purified proteins. As shown in Fig. 7, the drug inhibited in vitro DNA replication in a dose-dependent manner. The concentration of the drug required for inhibition of oriC DNA replication (IC 50 ϭ 60 M) was higher than that for inhibition of the ATP binding to DnaA protein (IC 50 ϭ 7 M) (Fig. 2).
The reaction in the oriC DNA replication can be separated into two stages: formation of the prepriming complex that requires a high temperature, 38°C, and the following priming and elongation stages that proceed even at a low temperature, 16°C (9). We examined which step was sensitive to 7 (3-CUCM-BI) using the staged reaction. Reaction mixtures depleting primase and DNA polymerase III holoenzyme were incubated at 38°C for 2 min, followed by priming and elongation reactions with radiolabeled deoxyribonucleoside triphosphates at 16°C. When the drug was added after the prepriming complex formation, a much higher concentration of the drug was necessary to inhibit DNA synthesis (Fig. 8), thereby suggesting that the action of the drug is specific for the prepriming stage rather than for the following steps of priming and elongation.
Next, we asked whether the inhibition by 7 (3-CUCM-BI) of oriC DNA replication was caused by inhibitory effects of the drug on the ATP binding to DnaA protein. If such was the case, preincubation of DnaA protein with ATP should diminish the inhibitory action of the drug on oriC DNA replication. On the contrary, if the target of the drug was not specific for the ATP binding of DnaA protein, even after preincubation of DnaA protein with ATP, the drug should have an inhibitory effect on DNA replication. The inhibitory effect of the drug was greatly decreased by preincubation of DnaA protein with ATP (Fig. 9). This means that the drug minimally inhibits oriC DNA replication when DnaA protein is complexed with ATP. The drug may inhibit oriC DNA replication through a specific inhibition FIG. 8. Influence of 7 (3-CUCM-BI) on pre-priming complex formation or priming and elongation step of DNA replication reaction. oriC DNA replication was staged as described previously (9). DnaA protein (1 pmol) was preincubated with various concentrations of 7 (3-CUCM-BI) (E) or without the drug (q) at 4°C for 15 min; then the reaction mixture (except primase and DNA polymerase III) was added, and pre-priming reaction was run at 38°C for 2 min. Various concentrations of 7 (3-CUCM-BI) were added (q) followed by incubation with dNTP, primase, and DNA polymerase III at 16°C for 10 min.  (15). It was also reported that DnaAcos protein, which has no affinity for ATP or ADP, was active in the system (5,6). Thus, if 7 (3-CUCM-BI) inhibited DNA synthesis in the oriC system through inhibition of the ATP binding of DnaA protein, it should have little effect on DNA synthesis in the ABC primosome system. We noted a significant amount of DNA synthesis in the ABC primosome system even after DnaA protein had been preincubated with 1 mM 7 (3-CUCM-BI) (Fig. 10), which completely inhibited the DNA replication in the oriC DNA replication system (Fig. 9). The results suggest that action of DnaA protein in the ABC primosome system is much less sensitive to the drug than that in the oriC system. DISCUSSION In this study, we attempted to enhance the inhibitory activity of 2 (3-Ac-BI) on the ATP binding to DnaA protein. Based on previous observations that hydrophobic domains and negative charges of acidic phospholipids are important for inhibition of the ATP binding to the DnaA protein (11)(12)(13)(14), we introduced alkyl chains and the carboxyl residue to the 3-O site of 2 (3-Ac-BI) and were able to increase the inhibitory effect of 2 (3-Ac-BI) on the ATP binding of DnaA protein.
The inhibitory activity of the mixture of 4 (3-KCM-BI) and 8 (12-aminododecanoic acid methyl ester hydrochloride) for the DnaA-ATP binding is less than that of 7 (3-CUCM-BI) and similar to that of 4 (3-KCM-BI) (Fig. 3). From an analogy of the structure-function relationship of inhibitors of protein kinases (16), we consider that the bis-indol structure may bind to the ATP binding site of the DnaA protein and that the alkyl chain of the inhibitor binds to a hydrophobic region near the ATP binding site of DnaA protein resulting in increase in affinity of the drug for DnaA protein. This assumption is supported by findings that a number of hydrophobic amino acids cluster near the putative ATP binding domain of DnaA protein (17). Hydropathy analysis of DnaA protein showed that this region is the most hydrophobic part of the protein (data not shown).
We found that 7 (3-CUCM-BI) potently inhibited the oriC DNA replication reaction reconstituted from purified proteins. The concentration of the drug necessary for inhibition of DNA synthesis was higher when the drug was added after formation of the prepriming complex (Fig. 8). Preincubation of DnaA protein with ATP before the incubation with the drug greatly diminished the inhibitory effect of the drug on oriC DNA replication (Fig. 9). DNA synthesis in the ABC primosome system, which does not require the ATP binding to DnaA protein, was less sensitive to the drug than that in oriC replication system (Fig. 10). These results suggest that 7 (3-CUCM-BI) inhibits oriC DNA replication by inhibiting the ATP binding to DnaA protein. This observation is the first evidence that a specific inhibitor for the ATP binding to DnaA protein inhibits oriC DNA replication, thereby suggesting that the ATP binding to DnaA protein is essential for activity of DnaA protein in the initiation of oriC DNA replication.
The concentration of 7 (3-CUCM-BI) required for inhibition of oriC DNA replication was higher than that for ATP binding to DnaA protein (Figs. 2 and 7). One explanation for this discrepancy is that a certain factor in the oriC DNA replication system decreases the affinity of the drug for DnaA protein.
Another is that high concentration of ATP in the oriC replication system (2 mM) leads to reconstruction of the ATP binding form of DnaA protein after the dissociation of DnaA-7 (3-CUCM-BI) complex.