Inhibitory Effects of Basic or Neutral Phospholipid on Acidic Phospholipid-mediated Dissociation of Adenine Nucleotide Bound to DnaA Protein, the Initiator of Chromosomal DNA Replication*

DnaA protein activity, the initiator of chromosomal DNA replication in bacteria, is regulated by acidic phospholipids such as phosphatidylglycerol (PG) or cardiolipin (CL) via facilitation of the exchange reaction of bound adenine nucleotide. Total lipid isolated from exponentially growing Staphylococcus aureus cells facilitated the release of ATP bound to S. aureus DnaA protein, whereas that from stationary phase cells was inert. Fractionation of total lipid from stationary phase cells revealed that the basic phospholipid, lysylphosphatidyl-glycerol (LPG), inhibited PG- or CL-facilitated release of ATP from DnaA protein. There was an increase in LPG concentration during the stationary phase. A fraction of the total lipid from stationary phase cells of an integrational deletion mprF mutant, in which LPG was lost, facilitated the release of ATP from DnaA protein. A zwitterionic phospholipid, phosphatidylethanolamine, also inhibited PG-facilitated ATP release. These results indicate that interaction of DnaA protein with acidic phospholipids might be regulated by changes in the phospholipid composition of the cell membrane at different growth stages. In addition, the mprF mutant exhibited an increased (cid:3) M [ (cid:2) - 32 P]ATP (44,000 cpm/pmol), and bound nucleotides were then determined. The results indicated that DnaA protein binds to ATP in the presence of PG with a higher K d value (1.5 (cid:3) M ; data not shown) than in the absence of PG (1 n M ; Fig. 1). The stoichiometry of bound ATP to DnaA protein in the presence of PG was 0.1, which was equivalent to that without PG. The results suggest that PG-induced dissociation of nucleotide from S. aureus DnaA protein is not the result of irreversible denaturation of the protein, but rather of a decreased in affinity for the nucleotides.

Chromosomal DNA replication in bacteria is regulated at the initiation step (1,2), where the activity and quantity of the initiator DnaA protein is critically controlled (3)(4)(5). DNA replication in Escherichia coli is initiated by the binding of DnaA protein to the DnaA boxes in the oriC, the origin of chromosomal DNA replication, followed by melting of the duplex at the three AT-rich 13-mers that locate close to the DnaA boxes. Loading of DnaB helicase in complex with DnaC protein leads to priming by DnaG primase, and DNA is synthesized by the DNA polymerase III holoenzyme (3,6).
Based on biochemical and molecular genetic research in E. coli, DnaA protein activity is proposed to be regulated by its binding of adenine nucleotides. DnaA protein has a high affinity for ATP and ADP, and the ATP binding form of DnaA protein (ATP-DnaA) initiates DNA replication in vitro, whereas the ADP binding form (ADP-DnaA) does not (7). In E. coli cells, the majority of DnaA protein is present as ADP-DnaA, and ATP-DnaA increases at the time of initiation (8,9). The ratio of ATP-DnaA to ADP-DnaA is regulated by at least four elements, including de novo synthesis of DnaA protein accompanied by formation of ATP-DnaA, an intrinsic ATP hydrolysis activity that generates ADP-DnaA from ATP-DnaA, a stimulation factor for the ATP hydrolysis named RIDA 1 (for regulatory inactivation of DnaA), and an exchange reaction of bound adenine nucleotide. RIDA requires Hda protein and a ␤-clamp of the DNA polymerase III holoenzyme as essential components and stimulates DnaA ATPase coupling with progressive DNA replication by the DNA polymerase III holoenzyme (8,10). Thus, RIDA suppresses the initiation of DNA replication (5).
Acidic phospholipid is a well characterized factor that facilitates the exchange reaction of ADP to ATP bound to DnaA protein (11)(12)(13)(14). Acidic phospholipids in a fluid membrane facilitate dissociation of ADP bound to DnaA protein in vitro, and the resultant free form of DnaA protein is reactivated through binding to ATP, which is present at high concentrations under physiological conditions. In the conditional pgsA deletion mutant that is defective for the synthesis of acidic phospholipids, phosphatidylglycerol (PG) and cardiolipin (CL), initiation of replication at oriC does not occur (15). Moreover, the growth defect of the pgsA mutant is restored by expression of DnaA proteins possessing certain mutations (16). E. coli DnaA protein is located at the cell membrane (17,18). In addition, there is direct evidence of an exchange reaction of ADP to ATP bound to DnaA protein in the cells (9). These findings strongly suggest that reactivation of DnaA protein by acidic phospholipids is indispensable for the initiation of DNA replication at oriC. DnaA protein of other bacteria was purified (19 -22), and its interaction with acidic phospholipids has been investigated (23). How the reactivation of DnaA protein by phospholipids is coordinated with cell cycle progression remains uncertain.
All previous studies of the interaction of E. coli DnaA protein with acidic phospholipids used chemically synthesized lipids or those isolated from eukaryotic cells. Little attention has been paid to alterations of lipid composition caused by physiological changes in the growth condition. In the present study, we focused on a Gram-positive bacterium, Staphylococcus aureus, in which lipid composition is altered from the exponential phase to the stationary growth phase (24). The amino acid sequence of domains 3 and 4 of S. aureus DnaA protein has 49% identity with that of E. coli DnaA protein (25). This region is responsible for ATP binding, ATP hydrolysis, membrane interaction, and DNA binding properties of the E. coli DnaA protein (26). Therefore, the properties of DnaA protein for ATP binding and interaction with acidic phospholipids are expected to be conserved between the two bacterial species. Here we describe that total lipid isolated from exponentially growing S. aureus cells releases ATP from ATP-DnaA, whereas total lipid isolated from stationary phase cells does not. Moreover, in the stationary phase, the activities of the acidic phospholipids, PG and CL, are inhibited by a basic phospholipid, lysylphosphatidylglycerol (LPG).

EXPERIMENTAL PROCEDURES
Reagents-[␣- 32  Bacteria and Plasmids-S. aureus strain RN4220 (27) was kindly provided by Dr. Kenichi Hiramatsu (Juntendo University). S. aureus strain CK1001 was constructed as described below. S. aureus strain Cowan I (28) and E. coli strain KA450 (⌬oriC1071::Tn10, dnaA17(Am), rnhA199(Am)) (9) were from our laboratory stock. Shuttle vectors between E. coli and S. aureus, pHY300PLK and pND50 (29), were from Takara (Japan) and Dr. Matsuhisa Inoue (Kitasato University), respectively. S. aureus DnaA-overproducing plasmid was constructed as follows. The coding region of the S. aureus dnaA gene (1.4 kbp) (25) was amplified by polymerase chain reaction (PCR) using primers 5Ј-gggaattccatatgtcggaaaaaaagaaatttggg-3Ј and 5Ј-ccggaattcttatacatttcttatttctttttc-3Ј and chromosomal DNA extracted from S. aureus Cowan I as template. EcoRI and NdeI restriction sites or the EcoRI site were added to the 5Ј or 3Ј ends of the coding region, respectively. The EcoRI fragment of the PCR product was subcloned into the EcoRI site of M13mp19. The NdeI-EcoRI fragment of the resultant plasmid containing the S. aureus dnaA coding region was replaced with the E. coli dnaA coding region of E. coli DnaA overproducer pMZ001, which contains an arabinose promoter (30). The resultant plasmid, named pSAd-naA002, was used for overproduction of S. aureus DnaA protein. pHYm-prF was obtained by insertion of the DNA fragment, which was amplified by PCR using primers 5Ј-tgaaacgagtatttgccacttga-3Ј and 5Јtccaagcgcttcaggcataa-3Ј and chromosomal DNA extracted from S. aureus RN4220 as template, into the SmaI site of pHY300PLK in the same orientation as the tetracycline resistance gene.
Bacterial Culture Conditions-E. coli and S. aureus cells were grown in Luria Bertani (LB) medium (1% tryptone (Difco), 0.5% yeast extract (Difco), and 1% NaCl). Thymine, up to 50 g/ml, ampicillin, up to 50 g/ml, tetracycline, up to 5 g/ml, or chroramphenicol, up to 12.5 g/ml, was added if required. For S. aureus lipid preparations, a 0.5-ml aliquot of overnight culture was diluted into 100 ml of LB medium in a 225-ml round tube (Falcon), capped tightly, and incubated at 37°C. Cells were harvested at the exponential phase (OD 660 of 0.3) or at the stationary phase (20 h incubation with an OD 660 of ϳ2.5).
Purification of S. aureus DnaA Protein-E. coli strain KA450 harboring pSAdnaA002 was grown in 3 liters of LB medium containing thymine and ampicillin at 30°C. When the OD 660 value reached 0.5, 60 ml of 50% L-(ϩ)-arabinose was added (final 1%), and the culture was further incubated for 3 h. Cells were harvested by centrifugation at 4°C, resuspended in buffer CЈ (50 mM HEPES-KOH (pH 7.6), 1 mM EDTA, 2 mM dithiothreitol (DTT), 20% glycerol) containing 0.25 M KCl (1 ml/1 g of cell paste), frozen in liquid nitrogen, and stored at Ϫ80°C. After thawing the frozen cell suspension on ice, lysozyme and spermidine were added up to 0.3 mg/ml and 20 mM, respectively. The suspension was mixed by inversion, placed on ice for 30 min, frozen in liquid nitrogen, and thawed on ice. After brief sonication, the extract was collected by centrifugation in a Beckman 80Ti rotor at 35,000 rpm for 20 min at 4°C. Ammonium sulfate was slowly added to the supernatant up to 0.2 g/ml with stirring. After additional stirring for 3 h, precipitates were collected by centrifugation in an 80Ti rotor at 30,000 rpm for 20 min at 4°C, dissolved in buffer CЈ containing 10 mM magnesium acetate, and dialyzed for 12 h against 1.6 liters of buffer CЈ containing 10 mM magnesium acetate with replacement of the buffer once. After centrifugation in an 80Ti rotor at 30,000 rpm for 20 min, the supernatant was applied to a MonoS HR5/5 column (Amersham Biosciences fast protein liquid chromatography) equilibrated with buffer CЈ containing 10 mM magnesium acetate. DnaA protein was eluted with a linear gradient of 0 to 1 M KCl in buffer CЈ containing 10 mM magnesium acetate. The DnaA protein fraction that was active for ATP binding was used as purified S. aureus DnaA fraction. This fraction had a single band on SDS-polyacrylamide gel (10.5%) electrophoresis stained with Coomassie Brilliant Blue R250. Protein concentration was determined by the Lowry method (31) using bovine serum albumin as a standard.
Purification of E. coli DnaA Protein-E. coli DnaA protein was purified as described previously (32), except for using an E. coli DnaA overproducer, pBAD-dnaA, 2 that was kindly provided from Dr. T. Ogawa (Nagoya University, Nagoya, Japan).
Isolation of S. aureus Lipid-Exponentially growing cells and stationary phase cells were prepared as described above. Total lipid was extracted by a modified Bligh-Dyer method as described previously (11), dried with a rotary evaporator, and suspended in water by sonication (Branson Sonifier 450). The S. aureus phospholipids were purified from total lipid obtained during either the exponentially growing phase or the stationary phase as indicated. LPG was purified by thin layer chromatography (TLC) developed with chloroform/methanol/acetic acid (65:25:10, v/v/v) using a silica gel plate (Silicagel 60, type PK6F, Whatman). PG and CL were further subjected to a second TLC with chloroform, methanol, and 29% ammonium (65:35:5, v/v/v). Re-extraction of the lipid from the thin layer plate was performed as follows. One of the TLC plates was sprayed with 100 mg/ml CuSO 4 containing 8% phosphoric acid and heated to detect lipids (33). The region containing PG, CL, and LPG was scraped off the other thin layer plates, and lipids were extracted by stirring for 30 min at 4°C in chloroform/methanol (2:1, v/v). Methanol and distilled water were added to the samples up to a chloroform/methanol/water ratio of 1:1:0.9 (v/v/v) followed by vigorous mixing. After centrifugation at 3000 rpm for 3 min, the lower phase was collected, dried with a rotary evaporator, dissolved in a small amount of chloroform/methanol (2:1, v/v), and stored at Ϫ20°C. All procedures, except for evaporation, were performed at 4°C. Each purified phospholipid fraction had a single spot by TLC (Silicagel 60, Merck) with three different solvent systems; chloroform/methanol/water (65:25:4, v/v/v), chloroform/methanol/29% ammonium (65:35:5, v/v/v), and chloroform/ methanol/acetic acid (65:25: Quantification of Phospholipids-The amount of phosphorus was determined (34), and the concentration of phospholipids was calculated assuming an average molecular weight of 750 for total lipid, 750 for PG, 850 for LPG, 1400 for CL, and 750 for PE. Radioactivity retained on the filter was measured in a liquid scintillation counter. ATP binding assay for E. coli DnaA protein was performed as described previously (7).
Exposure of ATP-DnaA to Lipids-For S. aureus DnaA protein, the [␣-32 P]ATP-DnaA complex (0.2-0.5 pmol) was formed as described above in 50 l of standard reaction buffer and magnesium acetate was added to 10 mM. Lipid suspension and excess cold ATP (final 1 mM) were added, and the mixture was incubated at 4°C for 14 min or as indicated. Reactions were also performed at 37°C for indicated time periods (Fig. 4). Samples were filtered as described above and radioactivity retained on the membrane was measured. When basic or zwitterionic lipids were mixed with PG, each lipid was mixed in organic solvent, dried, and suspended in 250 l of water (PG, 1.2 mg/ml). Lipid suspensions were sufficiently sonicated before use. An ATP releasing assay for E. coli DnaA protein was performed under conditions for E. coli DnaA protein, as described previously (11). ATP binding and ATP releasing assays for both S. aureus DnaA protein and E. coli DnaA protein were performed under conditions for S. aureus DnaA protein, and releasing reactions were performed on ice or at 37°C for various time periods (Fig. 4).
Construction of S. aureus mprF Mutant-The S. aureus mprF mu-tant was constructed by integrational disruption through a singlecrossover event (35) as follows. An 804-bp DNA fragment corresponding to the internal region of the mprF gene was amplified by PCR using 5Ј-gcaatcacattgtatcgggag-3Ј and 5Ј-cgggatccggtacaaaatagtacgcaa-3Ј primers and chromosomal DNA extracted from RN4220 as template, then cloned into the SmaI site of the plasmid, pCK20. pCK20, containing both the E. coli pUC19-derived origin and the chloramphenicolresistant gene, was constructed by self-ligation of the AflII-AvaII fragment of pND50 (29) to delete the S. aureus pUB110-derived origin. The resultant plasmid (pCK20-mprF) was electroporated into S. aureus RN4220, and homologous recombinants were selected on LB-agar plates containing 12.5 g/ml chloramphenicol and named CK1001. Southern blot hybridization was performed to ensure a single integration of pCK20-mprF into the chromosomal mprF locus. Flow Cytometric Analysis-Flow cytometry analysis of S. aureus cells was performed as described previously (36) with slight modifications. The exponentially growing cells at 37°C with an OD 660 of 0.2 were treated with 100 g/ml rifampicin and 10 g/ml cephalexin for 4 h. The samples were harvested by centrifugation, washed, and resuspended in 0.1 ml of wash buffer. Fixing solution (1 ml) was added to the suspensions and mixed by vortex. The fixed cells were treated with 800 g/ml RNase A at 50°C for 2 h and sonicated for 30 s to disrupt conjugation of the cells. After washing with TE buffer, samples were stained with a combination of ethidium bromide and mithramycin. Cells were analyzed using a Becton Dickinson FACSCalibur.

Effects of Total Lipid Isolated from S. aureus Cells on the
Dissociation of ATP-DnaA-S. aureus DnaA protein was overexpressed in E. coli and purified to near homogeneity. Purified protein had a single band of ϳ50 kDa judged by SDS-polyacrylamide gel electrophoresis with Coomassie Brilliant Blue R-250 staining. S. aureus DnaA protein has a high affinity for ATP and ADP with K d values of 1 and 5 nM, respectively (Fig. 1). One molecule of DnaA protein bound to 0.1-0.3 molecules of ATP (Fig. 1), depending on the different preparations.
Lipids from S. aureus were isolated during the exponential phase and stationary growth phase to compare their ability to facilitate the release of ATP from ATP-DnaA. Total lipid isolated from exponentially growing cells facilitated the release of ATP from DnaA protein (Fig. 2). The activity was dependent on the lipid dose and exposure time (Fig. 2). On the other hand, total lipid from stationary phase cells did not have this activity (Fig. 2). Unlike S. aureus, there was little difference in the action of the E. coli lipids between the two growth phases as described before (11).
PG Facilitates the Dissociation of ATP-DnaA-To determine the mechanism underlying the alterable interaction of DnaA protein with lipids, we examined which lipid was responsible for facilitating the dissociation of ATP-DnaA. Total lipid isolated from exponentially growing cells was fractionated by two-dimensional TLC and separated into eight fractions (Fig.  3A). In S. aureus cells, PG, CL, and LPG are the major phospholipids that form ϳ98% of the total phospholipids (37). Staining with Dittmer-Lester reagents revealed that fractions 4, 5, and 7 contained lipid phosphates (data not shown). Staining with ninhydrin reagent revealed that fraction 7 also contained amino groups (data not shown), indicating that fraction 7 contained LPG. The R F value of the lipid in fraction 5 on onedimensional TLC was much the same as that of PG isolated from egg yolk (data not shown). Mass spectrometric analyses indicated that the molecular mass of the most abundant lipid in fraction 5 was 736, which corresponded to PG. Lipid in fraction 4 was determined to be CL by comparing the R F value from one-dimensional TLC with CL from bovine heart. Mass spectrometric analyses supported this conclusion (data not shown).
Lipid in each fraction was isolated from the thin layer silica gel plate, suspended in water, and subjected to assay for the release of ATP from ATP-DnaA. Fraction 5 (PG) had the most potent activity (Fig. 3C). Fraction 8, which contained lipids from the entire region except for fractions 1-7, had weak, but significant activity (Fig. 3C). It remains uncertain which lipid(s) is responsible for the activity in fraction 8. There was no activity in fraction 4 (CL), possibly caused by the small amount of CL in the exponentially growing phase. CL purified from stationary phase cells had releasing activity as shown below (Fig. 5).
PG purified from Fraction 5 facilitated the release of ATP bound to DnaA protein in a dose-dependent manner (data not shown) and the recovery of the activity from total lipid in fraction 5 was ϳ45%. PG also facilitated the dissociation of ADP-DnaA (data not shown). These results suggest that PG was mainly responsible for facilitating the release of the ATP bound to DnaA protein in the total lipid of exponentially growing S. aureus cells, and confirmed the importance of PG in bacterial cell membranes for interaction with DnaA protein.
There is a possibility that the PG-induced release of the nucleotide from S. aureus The initial amount of ATP bound to DnaA protein before exposure to the lipid fraction is defined as 100%. Blank indicates a negative control experiment using a sample extracted from silica gel containing no lipid whose area was equal to that of fraction 1. The vertical bars represent differences between duplicate experiments.
M [␣-32 P]ATP (44,000 cpm/pmol), and bound nucleotides were then determined. The results indicated that DnaA protein binds to ATP in the presence of PG with a higher K d value (1.5 M; data not shown) than in the absence of PG (1 nM; Fig. 1). The stoichiometry of bound ATP to DnaA protein in the presence of PG was 0.1, which was equivalent to that without PG. The results suggest that PG-induced dissociation of the nucleotide from S. aureus DnaA protein is not the result of irreversible denaturation of the protein, but rather of a decreased in affinity for the nucleotides.
Effects of Temperature on the Dissociation of ATP-DnaA-Dissociation of the nucleotide form of E. coli DnaA protein by acidic phospholipid membrane required a temperature above the phase transition point (11,14). On the other hand, the nucleotide dissociation from S. aureus DnaA protein proceeded on ice (Fig. 2). To examine whether the difference in these results was the result of a difference in the assay conditions, dissociation of ATP from S. aureus DnaA was compared with that from E. coli DnaA under the same assay conditions. When the dissociation of E. coli DnaA-ATP was performed under the conditions for S. aureus DnaA used in Fig. 2, dissociation of ATP from E. coli DnaA protein by PG did not occur on ice, but was observed at 37°C (Fig. 4, C and D), consistent with the results under the conditions for E. coli DnaA protein (11,14). On the other hand, dissociation of ATP from S. aureus DnaA protein by PG took shorter time periods at 37°C than on ice (Fig. 4, A and B). Therefore, facilitation of the dissociation of S. aureus DnaA-ATP by acidic membranes differs from the case of E. coli DnaA protein because the dissociation occurred on ice. The dissociation rate of the DnaA proteins increased at 37°C compared with those on ice (Fig. 4, B and D). Therefore, membrane fluidity might have promoted the dissociation of ATP bound to either species of DnaA protein.
LPG Inhibits the PG-mediated Dissociation of ATP-DnaA-Although total lipid extracted from S. aureus cells during the stationary phase have too little activity to facilitate the dissociation of ATP-DnaA (Fig. 2), the fraction did contain PG. The amount of PG was ϳ70% of that during the exponential phase (Fig. 3B). In addition, the fraction also contained an increased amount of CL, another acidic phospholipid (Fig. 3B). When PG and CL were purified from stationary phase lipids and assayed for the dissociation, they facilitated the dissociation of ATP-DnaA (Fig. 5). Their specific activities were much the same as PG purified from exponentially growing cells (data not shown). Thus, the loss of activity in stationary phase lipids was not the result of a decrease in content nor of a decrease in the activity of the acidic phospholipids, PG and CL, but might be the result of the presence of a lipid that inhibits the actions of PG and CL and that exists in the stationary phase.
The identity of the lipid that inhibits the activities of PG and CL to facilitate the dissociation of ATP-DnaA was determined by fractionation of the stationary phase lipids by one-dimensional TLC using three different solvent systems. Lipid in each fraction was extracted from silica gel plates and mixed with PG, and the resultant lipid mixture was exposed to ATP-DnaA. In each TLC condition, there was a lipid fraction with an inhibitory effect on PG-mediated dissociation of ATP-DnaA (fraction 2, fraction 3, and fraction 1 of Fig. 6, A, B, and C, respectively). Each lipid in these fractions stained positive for both the Dittmer-Lester reagent, which detects phospholipids, and the ninhydrin reagent, which detects amino groups (data not shown). LPG is the major phospholipid containing an amino group in S. aureus and its content in the stationary phase increased 3.6-fold over that in the exponential phase (Fig. 3, A and B). Thus, we decided to examine whether LPG inhibited the PG-mediated dissociation of ATP-DnaA.
LPG was purified and further characterized based on the dissociation of ATP-DnaA. LPG alone had little effect on the dissociation of the S. aureus ATP-DnaA (Fig. 5). When increasing amounts of LPG were mixed with PG and the mixture was exposed to ATP-DnaA, the amounts of ATP-DnaA increased (Fig. 7). With an LPG to PG ratio of 0.5:1 (by mol), LPG masked the action of PG on the dissociation of ATP-DnaA (Fig. 7). Because LPG content during the stationary phase is approximately half that of PG and CL (Fig. 3B), it is reasonable to assume that the loss of activity in the stationary phase lipid is caused by the inhibitory effect of LPG. The LPG content in exponentially growing cells was less than one fifth that of PG and CL (Fig. 3A). Therefore, the relative ratio of LPG to PG and CL might determine the activity of the total lipid for facilitating the release of ATP from ATP-DnaA.
Inhibitory Effect of LPG on the PG-mediated Release of ATP Bound to E. coli DnaA Protein-We then examined whether the 1% Triton X-100. S. aureus DnaA protein complexed with [␣-32 P]ATP was exposed to each lipid, and bound ATP to DnaA protein was measured as described in the legend for Fig. 2. inhibitory effect of LPG on the PG-mediated dissociation of the ATP form of S. aureus DnaA protein was also observed with E. coli DnaA protein. An increase in LPG while PG levels remained constant prevented PG-mediated dissociation of the ATP form of E. coli DnaA protein (Fig. 8A). Moreover, LPG also inhibited the CL-mediated release of ATP from E. coli DnaA (Fig. 8B), where twice the amount of LPG (in mol) was required than that of PG. LPG alone did not affect the dissociation of the ATP form of E. coli DnaA protein (data not shown). Thus, LPG had an effect not only on S. aureus DnaA protein but also on E. coli DnaA protein.
Whereas the previous paper reported that CL was more effective than PG (dipalmitoyl-) for promoting the dissociation of the bound nucleotide from E. coli DnaA protein (11), this was not observed in the present study (Fig. 8). The relatively high efficiency of PG in the present study might be caused by the high content of unsaturated fatty acids in egg yolk PG (Fig. 8). Unsat- acetic acid (65:25:10, v/v/v). The lower panels show lipid staining patterns by CuSO 4 /phosphate. Each plate was divided into 10 -12 fractions as shown. The arrowheads represent a spot of lipid that was positive for staining by both the ninhydrin reagent and the Dittmer-Lester reagent. Lipid in each fraction was re-extracted from silica gel plates at 4°C, mixed with PG (300 g, egg yolk), dried to remove the solvent, and suspended in 250 l of water. The mixed lipid suspension (6 l) was assayed for dissociation of the ATP form of S. aureus DnaA protein (4 pmol) as described in the legend for Fig. 2. C (control) represents an assay with only PG. The vertical bars represent differences between duplicate experiments.  (12,14).
Inhibitory Effect of PE on the PG-mediated Dissociation of ATP-DnaA-E. coli membranes do not contain basic phospholipids such as LPG, but have a zwitterionic phospholipid, PE (ϳ75% of total phospholipids). We next examined whether PE had an inhibitory effect on PG activity to facilitate the dissociation of ATP-DnaA. PE inhibited PG-mediated release of ATP from DnaA protein in a dose-dependent manner regardless of the source of DnaA protein, although the amount needed for inhibition was larger than that for LPG (Fig. 9). Therefore, not only LPG but also PE has an inhibitory effect on the PGmediated dissociation of ATP-DnaA.
Effect of Total Lipid Isolated from a Mutant Deficient in LPG on the Dissociation of ATP-DnaA-To examine the role of LPG in the regulation of DnaA protein function in vivo, an integrational deletion mutant of the mprF gene, which is essential for LPG synthesis in S. aureus (38), was constructed by a method described previously (35). A DNA fragment coding an internal MprF protein region was cloned into an E. coli plasmid containing the chloramphenicol-resistant gene, and the resultant plasmid was integrated into the mprF locus of RN4220 cells by homologous recombination. The integration of the plasmid at a single mprF gene locus on the genome was confirmed by Southern blot hybridization using the chloramphenicol-resistant gene region as a probe (data not shown). The cell membrane of the constructed mprF mutant cells lost LPG (Fig. 10A), consistent with a report by Peschel et al. (38). The loss of LPG was complemented by a plasmid harboring the mprF gene (data not shown). Total lipid isolated from stationary phase mprF mutant cells facilitated the dissociation of ATP-DnaA (Fig. 10B), which was not the case of the wild-type strain (Fig. 2). This result was consistent with biochemical evidence that purified LPG in- hibited the PG-mediated dissociation of ATP-DnaA (Fig. 7). Total lipid isolated from exponentially growing mprF mutant cells was more efficient than that from wild-type cells in facilitating the dissociation of ATP-DnaA (Figs. 2 and 10B). This result suggests that LPG also functions in exponentially growing cells.
Increase in DNA Amount per Cell in the mprF Deletion Mutant-Although Peschel et al. (38) reported that the growth of the mprF transposon-insertion mutant was apparently normal, we examined whether the mprF integrational deletion mutant had a disorder in the cell cycle regulation based on the in vitro results described above. Exponentially growing cells were treated with rifampicin and cephalexin, which inhibit an initiation step of DNA replication but allow ongoing DNA replication to complete with an inhibition of cell division, and were analyzed by flow cytometry to determine the number of replication origins in each cell (36). Doubling time of the mprF mutant at 37°C was much the same as wild-type strain. Wildtype cells contained two or four copies of origin with a ratio of 5 to 3 (Fig. 11A). The mprF mutant cells contained two, four, or eight copies of origin with a ratio of 2 to 5 to 2 (Fig. 11B). Therefore, the number of replication origins per cell of the mprF mutant strain increased ϳ1.6-fold over that of wild-type strain. This phenotype of increased origin numbers per cell was complemented by the wild-type mprF gene (Fig. 11C). Severe asynchronous initiation of DNA replication, which was judged by the presence of an abnormal number of origin (3, 5, 6, or 7), was not observed in the wild-type strain or in the mprF mutant strain. When the amount of DNA of exponentially growing and randomly replicated cells was determined by flow cytometry, where cells were not treated with drugs, there were higher amounts of DNA per cell in the mprF strain than in the wildtype strain, as determined by a shift of the peak position in the DNA histograms (data not shown). In addition, this increased amount of DNA per cell was complemented by the wild-type mprF gene (data not shown). DISCUSSION Acidic phospholipids in fluid membrane dissociate inactive ADP-DnaA, and the resultant free DnaA protein becomes the active ATP-bound form under physiological concentrations of ATP. Thus, acidic phospholipids rejuvenate DnaA protein by facilitating the exchange reaction of adenine nucleotide bound to DnaA protein. Although this process seems to be regulated in response to cell cycle progression or growth phase transition, the nature of the rejuvenation remains uncertain. In the present study, total lipid isolated from exponentially growing, but not stationary phase, S. aureus cells, a Gram-positive bacterium, facilitated the release of adenine nucleotide bound to DnaA protein. During the stationary phase, the PG content decreased to ϳ70%, whereas LPG content increased 3.6-fold (Fig. 3, A and B). LPG strongly inhibited the actions of PG and CL, suggesting that basic phospholipids negatively regulate activation of the DnaA protein mediated by acidic phospholipids. In addition, total lipid during the stationary phase isolated from the mprF deletion mutant, which lost LPG, recovered the activity for facilitating the dissociation of ATP-DnaA. These biochemical findings suggest that DnaA activity is regulated by a change in the phospholipid composition of the cell membrane under physiological conditions.
The acidic phospholipid-dependent release of bound nucleotide from S. aureus DnaA protein is considered to be equivalent to that from E. coli DnaA protein for the following reasons. 1) Acidic phospholipids, not neutral phospholipids, support the release of bound nucleotide from S. aureus DnaA protein (Fig.  5) (11). 2) A similar molar ratio of LPG and PG is required for the inhibitory effect of LPG on both E. coli and S. aureus DnaA proteins (Figs. 8 and 7, respectively).
3) The release of bound nucleotide by PG from S. aureus DnaA protein might not be caused by an irreversible denaturation of the protein as in the case of E. coli DnaA (11,13). S. aureus DnaA protein binds to ATP with a K d value of 1.5 M in the presence of PG (0.83 mM), whereas the K d value in the absence of PG was 1 nM (Fig. 1). This result also indicates that PG might change a part of the S. aureus DnaA protein structure, resulting in facilitated release of the bound nucleotide.
For the acidic phospholipid-mediated dissociation of adenine nucleotide bound to E. coli DnaA protein, electrostatic interactions followed by insertion of a part of the DnaA protein into the membrane are required (39,40). We previously reported the requirement of a cluster formation, a phase separation, of acidic phospholipid for the process on the mixed membranes using chemically synthesized lipids possessing different acyl groups (41). Therefore, the inhibitory effect of LPG on the interaction of acidic phospholipids with DnaA protein might be the result of (i) a decrease in negative electricity in the membrane, and/or (ii) reduction of the cluster domains of acidic phospholipids. For the inhibition of PG-mediated dissociation, a smaller amount of LPG, a basic phospholipid, was required than that of PE, a zwitterionic phospholipid. In addition, 2 times the amount of LPG (in mol) was required to inhibit the action of diacidic CL than that of monoacidic PG. These results support the notion that the inhibition of PG-mediated dissociation by LPG might be caused by a decrease in negative electricity on the membrane surface. Further studies are required to examine whether LPG disturbs the cluster formation of acidic phospholipids on the mixed membranes or not.
Membrane fluidity seems to be required for reactivation of the E. coli DnaA protein by acidic phospholipids because the process is dependent on both the presence of unsaturated fatty FIG. 11. Flow cytometric determination of the number of replication origins per cell. Exponentially growing cells at OD 660 of 0.2 were treated with rifampicin and cephalexin for 4 h. In this condition, the initiation of DNA replication and septum formation were inhibited but the elongation stage was completed. DNA histograms were obtained by flow cytometry. A, RN4220/pHY300PLK. B, CK1001/pHY300PLK. C, CK1001/pHYmprF. acids in lipid molecules and incubation temperatures above the phase transition point of the membrane (12,14). In the present study, the reaction of the acidic phospholipid-mediated dissociation of the ATP from S. aureus DnaA protein was performed at a low temperature. This result suggests that membrane fluidity is not required for the acidic phospholipid-mediated dissociation of ATP from S. aureus DnaA protein, and this biochemical characteristic of S. aureus DnaA protein is distinct from that of E. coli DnaA. On the other hand, the dissociation rate of ATP from S. aureus DnaA protein by PG was increased at 37°C (Fig. 4B). Therefore, membrane fluidity is considered to have a promotive role in the dissociation of the nucleotides from S. aureus DnaA protein as described in E. coli DnaA protein.
Peschel et al. (38) reported that the mprF gene is necessary for the pathogenicity of S. aureus and for synthesis of LPG. In the transposon insertion mprF mutant cells, LPG content in the cell membrane decreased below detectable levels, whereas the growth of the mutant was reported to be much the same as the wild-type strain. Thus, LPG seems to be dispensable for growth of S. aureus cells. In the mutant cells, it is possible that a change in lipid composition occurs to compensate for the loss of LPG and the interaction between acidic phospholipid and DnaA protein. Actually, in the pgsA3 mutant of E. coli, in which the PG and CL synthesis pathway is defective, the phosphatidic acid content increases (42). Thus, we examined the effect of changes in lipid composition in the mprF mutant. An alteration in the lipid composition other than LPG was observed in both the exponentially growing and stationary phases in the mprF mutant (data not shown). Still, total lipid extracted from the mutant in the stationary phase had the facilitating activity (Fig. 10B). The result supports the notion that LPG is responsible for inhibiting PG-facilitated nucleotide release and excludes the possibility that the loss of LPG might be suppressed by an increase in other basic lipids that could inhibit PG-facilitated nucleotide release. Regulatory mechanisms for the DnaA protein activity other than interactions with phospholipids might act to control the initiation of DNA replication in the mprF mutant.
The mprF integrational deletion mutant, which lost LPG, had more origin copy numbers per cell than did the wild-type strain (Fig. 11). This alteration could be caused by a mutation defective in the initiation of DNA replication, chromosomal partition, or cell division. Thus, the results suggest that the mprF gene is involved in the regulation of cell cycle event(s). The molecular mechanism, based on the present finding, might be the loss of LPG-mediated regulation of the DnaA activity. This is consistent with a report describing an alteration in the copy number of chromosomal DNA in a temperature-sensitive strain of E. coli with the dnaA 167 mutation (43). The involvement of the disordered regulation of DnaA activity in the altered origin copy number in the mprF mutant is currently under further investigation. Another possibility is that the loss of LPG might affect membrane binding of origin and/or terminus regions. Membrane binding of these regions was reported in B. subtilis (44), whereas roles of the membrane-chromosome association in the regulation of DNA replication have not been ruled out. Various cell cycle-related proteins other than DnaA interact with cell membrane or phospholipids (44 -47). LPG might affect these interactions, thereby altering their activities.
LPG is present not only in Staphylococci, including S. aureus, but also in other Gram-positive bacteria like Bacillus subtilis (48), Enterococcus faecalis (24), Mycoplasma laidlawii (49), Vagococcus fluvialis (50), and the Gram-negative bacterium, Pseudomonas aeruginosa (51). In addition to LPG, other basic lipids are present in bacterial membranes (52). These basic lipids might have a role in activity regulation of DnaA protein in each type of bacteria.