Of 750 temperature‐sensitive mutants of Gram‐positive Staphylococcus aureus, one was complemented by the dnaA gene. This mutant had a single base transition in the dnaA gene causing the amino‐acid substitution mutation, Ala40Thr. Phage transduction experiments showed that this temperature‐sensitive phenotype was linked with a drug‐resistant marker inserted near the dnaA gene, suggesting the dnaA mutation is responsible for the phenotype. Flow cytometric analysis revealed that the dnaA mutant was unable to initiate DNA replication at a restrictive temperature and exhibited asynchrony in the replication initiation at a permissive temperature. This is the first report of a temperature‐sensitive dnaA mutant in S. aureus, and the results show that DnaA is required for the initiation of chromosomal replication and for the regulation of synchrony in the bacterial cells.
Chromosomal DNA replication in bacteria is regulated at the initiation step. Control of the activity and the amount of initiator DnaA protein play an indispensable role in this regard (Skarstad & Boye, 1994; Messer, 2002). In Escherichia coli, chromosome replication is initiated by oligomeric binding of DnaA protein to the DnaA boxes in the origin of chromosomal replication, oriC. This leads to melting of duplex DNA at AT‐rich regions within oriC. DnaA protein next loads DnaB helicases from DnaB–DnaC complexes, resulting in the priming and initiation of DNA synthesis.
Biochemical and molecular genetic studies in E. coli reveal that the DnaA protein is regulated by adenine nucleotide binding. DnaA is a member of the AAA+family of ATPases (Erzberger , 2002), and the ATP‐bound form of DnaA is active for replication initiation, whereas the ADP form is inactive (Sekimizu , 1987). In E. coli cells, the ADP form is abundant, and the ATP form increases coincident with the replication initiation (Kurokawa , 1999). The ratio of the ADP and ATP forms of DnaA in vivo appears to be determined by several factors, including the protein synthesis‐dependent generation of new ATP‐bound DnaA, the hydrolysis of bound ATP to ADP by the intrinsic ATPase activity of DnaA (Nishida , 2002), the level of regulatory inactivation of DnaA protein (RIDA) (Katayama , 1998), and the exchange of bound adenine nucleotides (Kurokawa , 1999). RIDA requires both Hda protein and the β‐clamp of the DNA polymerase III holoenzymes, and it stimulates hydrolysis of ATP bound to DnaA concomitant with DNA synthesis (Katayama , 1998; Kato & Katayama, 2001). Furthermore, exchange of ADP for ATP is stimulated by fluid membranes containing acidic phospholipids (Sekimizu & Kornberg, 1988; Crooke, 2001; Fujimitsu & Katayama, 2004) as well as by the presence of specific DNA regions containing DnaA boxes (Sekimizu & Kornberg, 1988; Crooke, 2001; Fujimitsu & Katayama, 2004).
Gram‐positive Bacillales including Staphylococcus aureus and Bacillus subtilis appear to have a different mechanism for regulating the replication initiation than Gram‐negative E. coli. First, DnaA in E. coli is not interchangeable with DnaA of either S. aureus or B. subtilis even though DnaA is conserved among these bacteria (Skarstad & Boye, 1994; Katayama , 1997; Krause , 1997). Second, E. coli Hda protein and sequestration of oriC via Dam methyltransferase and SeqA protein (Lu , 1994) do not exist in Bacillales. Instead, chromosome replication is negatively regulated by YabA in B. subtilis (Noirot‐Gros , 2002). Third, the copy number of the oriC plasmid is very low in B. subtilis but moderate in E. coli (Moriya , 1999). Finally, for the initiation, B. subtilis and S. aureus require the dnaD gene and two helicase loader genes, dnaB (different from E. coli dnaB) and dnaI, which do not have homologs in E. coli (Moriya , 1999; Li , 2004). Generally, understanding of the mechanisms regulating the initiation of chromosomal replication in Gram‐positive bacteria, including S. aureus, is far behind that in Gram‐negative E. coli.
Studies of temperature‐sensitive (TS) mutants have greatly contributed to the elucidation of the functions of genes that are responsible for essential processes like DNA replication. Therefore, we investigated the role of DnaA in the regulation of chromosome replication initiation by isolating and characterizing a TS mutant of dnaA in S. aureus.
Materials and methods
Bacteria and plasmids
Bacterial strains and synthetic primers (Proligo, Japan) were listed in Tables 1 and S1 (supplementary data), respectively. Bacterial cells were cultured in Luria–Bertani (LB) medium (1% bactotryptone, 0.5% yeast extract, and 1% NaCl) with 50 μg mL−1 thymine, 50 μg mL−1 ampicillin, and 5 μg mL−1 tetracycline, if required. Chlorpromazine was from Shionogi. Transformation of Staphylococcus aureus with plasmids was as described (Inoue , 2001). TS mutants of S. aureus were derived from RN4220 (Novick , 1993) treated with ethylmethansulfate (Sigma) (Inoue , 2001). To construct the thyA strain NI8, the central coding region of the thyA gene was amplified by polymerase chain reaction (PCR) using Pfu polymerase (Stratagene) and primers thyAL and thyAR, and then inserted between the EcoRI and BamHI sites of a kanamycin‐resistant pSF151 suicide vector (Kaito , 2005). This, in turn, was integrated into RN4220, resulting in NI8. For inserting an erythromycin‐resistant marker near the dnaA gene, the noncoding chromosomal region between SA0009 and SA0010 was amplified by PCR using primers 09L and 09R and then inserted into the EcoRI and BamHI sites of a pMutinT3 suicide vector (Kaito , 2005). This was then integrated into TS8822, resulting in NM1000. Phage transduction using phage 80alpha (Novick, 1991) was performed using NM1000 as a donor; NM1001 and NM1002 (dnaA8822) were constructed as a temperature resistant and TS transductant of RN4220, respectively. Transduction was also performed on NI8 (thyA), that resulted in NM1003 and NM1004 (dnaA8822). Escherichia coli JM109 was used as a host for shuttle vectors between E. coli and S. aureus, pND50 (Inoue , 2001) and pHY300PLK (Takara Bio). Escherichia coli KA450 [ΔoriC1071::Tn10, rnhA199(Am), dnaA17(Am)] (Katayama, 1994) was for construction of the pHYdnaA plasmid, which is a pHY300PLK‐derived plasmid containing the 2023 bp dnaA region at the SmaI site in the same orientation as the tetracycline‐resistance gene. The dnaA fragment was amplified using primers 1.5L and 4.5R that were constructed on N315 genome sequences.
|RN4220||dnaA+thyA+||Novick . (1993)|
|RN4220||dnaA+thyA+||Novick . (1993)|
The region containing the dnaA gene was amplified by PCR using primers 1L and 5R. The fragments were then used as a template for sequence reactions using a PRISM Bigdye terminator kit (Applied Biosystems, Japan) and several primers. The nucleotide sequence of the RN4220 dnaA gene was registered in DDBJ under accession no. AB201418.
Measurement of DNA or protein synthesis in Staphylococcus aureus
DNA and protein synthesis was as described (Li , 2004) with slight modifications. The thyA mutant was used as the parent strain, and 50 μg mL−1 [Methyl‐3H]thymine (47.5 Ci mmol−1, Moravek) replaced 0.1 μM [3H]thymidine for monitoring DNA synthesis. [35S]methionine (1000 Ci mmol−1) was from Amersham Biosciences.
Flow cytometric analysis was as described (Skarstad , 1995; Li , 2004) with slight modifications. Exponentially growing cells (OD600 nm=0.15) were treated at 30°C for 1 h with 50 μg mL−1 rifampicin and 20 μg mL−1 cephalexin or were incubated at 43°C for 1 h without rifampicin but with cephalexin. Cells under each condition were harvested, washed, and fixed in 1 mL of 70% ethanol. The fixed cells were treated with 20 μg mL−1 RNaseA at 50°C for 30 min, sonicated, washed, stained in staining buffer (10 mM Tris‐HCl [pH 7.4] and 10 mM MgCl2) with 1 μM SYTOX Green (Molecular Probes), and analyzed by FACSCaliber (Becton Dickinson). The sonication separated cells into single cells, whose conditions were determined by a combination of microscopic analysis and flow cytometric analysis, in which greater than 90% of stationary‐phase cells or dnaD1726 replication initiation mutant cells (Li , 2004) cultured at a restrictive temperature for 2 h showed the chromosome number of one that is the smallest number by our hands.
Exponentially growing cells at 30°C (OD600 nm=0.3) were treated with the same volume 20% trichloroacetic acid, centrifuged at 16 000 g, and washed twice with acetone. The cells were treated with 0.2 mg mL−1 lysostaphin at 37°C for 30 min in RIPA buffer (50 mM Tris‐HCl pH 8.0, 150 mM NaCl, 0.1% sodium dodecyl sulfate, 0.5% sodium deoxycholate and 1% Triton X‐100). Total proteins extracted from 108 cells or purified S. aureus DnaA (Ichihashi , 2003) were subjected to sodium dodecyl sulfate polyacrylamide gel (10.5%) electrophoresis and transferred onto polyvinylidine difluoride membrane (Millipore). The membrane was blocked with 5% skim milk, incubated with anti‐DnaA IgY and further with goat antichicken IgY antibody conjugated with alkaline phosphatase (Promega), and developed using nitrobluetertazorium and 5‐bromo‐4‐chloro‐3‐indolylphosphate according to the manufacturer's instruction (Roche). The band intensity was determined using the NIH image that was available at http://rsb.info.nih.gov/nih‐image/. Anti‐DnaA IgY was prepared from eggs of chicken immunized by S. aureus DnaA. Purified DnaA was measured using the Lowry method with bovine serum albumin (Ichihashi , 2003).
Complementation of TS cell growth by the dnaA gene
To identify the essential gene for Staphylococcus aureus cell growth, about 750 TS mutants were collected from ethylmethansulfate‐treated RN4220 cells. These mutants grew on LB medium at 30°C but not at 43°C. Genes that complemented the temperature sensitivity of these mutants were screened by introducing pND50 shuttle vector library containing the 17 PCR‐amplified putative replication genes including dnaA. A plasmid containing an operon of the dnaA and dnaN genes was found to complement the temperature sensitivity of the TS8822 strain (data not shown). To determine which gene is responsible for the complementation, we examined the complementation ability of the pHYdnaA plasmid, which contained only the dnaA gene. We found that pHYdnaA complemented the temperature sensitivity of TS8822, whereas vector plasmid pHY300PLK did not (Table 2).
|Transformants × 103 (43°C/30°C)|
|Transformants × 103 (43°C/30°C)|
Competent cells of TS8822 (22 μL) were electroporated with 100 ng of pHYdnaA or vector pHY300PLK, followed by incubation at 43 or 30°C on Luria–Bertani plates containing 5 μg mL−1 tetracycline. The number of transformants was counted. Data are representative of more than five independent experiments.
Determination of the mutation site in the dnaA gene
Next, we searched for the mutation site in the chromosomal dnaA gene of TS8822. We found that the mutant has a single transition from G : C to A : T (G118A) in the open reading frame, resulting in the amino‐acid substitution mutation, Ala40Thr, in the DnaA protein. We named this mutation dnaA8822. The amino‐acid sequence of the DnaA protein in the RN4220 strain is identical to that in the Cowan I (Katayama , 1997) (Genbank accession no. D89066) and N315 strains.
To determine whether the dnaA8822 mutation is responsible for the temperature sensitivity of TS8822, we performed phage transduction experiments using phage 80alpha. As a selection marker, the erythromycin‐resistance gene was inserted into a noncoding region between SA0009 and SA0010 near the dnaA gene of TS8822. Of 43 erythromycin‐resistant transductants of RN4220, seven gained a TS phenotype, while the remaining 36 transductants remained temperature resistant. Repeated experiments showed that the erythromycin resistance and temperature sensitivity were cotransduced with a frequency of about 16% (Table 3). A plasmid harboring the dnaA gene again complemented the temperature sensitivity of the transductants (data not shown). These results suggest that the dnaA8822 mutation is responsible for the TS phenotype of TS8822.
|TS and CPZs||TR and CPZs||TS and CPZr||TR and CPZr|
|TS and CPZs||TR and CPZs||TS and CPZr||TR and CPZr|
Transductants using phage 80alpha were first selected by erythromycin resistance (Ermr) because the resistance gene was inserted near dnaA. Selected transductants were then tested for both temperature‐sensitive (TS)/temperature‐resistant (TR) cell growth at 43°C and for chlorpromazine (0.2 mM) sensitivity (CPZs)/resistance (CPZr). The number of transductants for each condition is indicated. Data are representative of two independent experiments.
TS DNA synthesis in the dnaA8822 mutant cells
Studies in E. coli and B. subtilis have revealed that the dnaA gene is essential for chromosome replication initiation (Messer, 2002). We therefore asked whether DNA synthesis in the S. aureus dnaA8822 mutant cells is stopped at a restrictive temperature. For this purpose, the dnaA8822 mutation was transduced into the thyA background NI8 using phage 80alpha, and DNA synthesis was determined by continuous labeling of DNA with [3H]thymine. When exponentially growing cells were shifted from 30 to 43°C, DNA synthesis in the mutant NM1004 cells decreased, while it continued in the parent NM1003 cells (Fig. 1). The impairment of DNA synthesis at the restrictive temperature was suppressed by a plasmid containing the wild‐type dnaA gene (data not shown). In contrast to DNA synthesis, protein synthesis, as monitored by the incorporation of [35S]methionine, continued in all of the strains even after the temperature increase (Fig. 1), which was consistent with swollen cells for the mutant at a restrictive temperature (Fig. S1, supplementary data). These results suggest that S. aureus dnaA8822 impairs DNA synthesis but not protein synthesis.
Flow cytometric analysis
To determine whether the restrictive temperature causes DNA synthesis in the S. aureus dnaA8822 mutant to stop at initiation, we examined the chromosomal content in cells by flow cytometry (Skarstad , 1995). We found a single broad peak in DNA histograms of exponentially growing cells at 30°C in both the dnaA8822 mutant (Fig. 2a) and the parent strain (Fig. 2b). This indicates that the cells were in different stages of the replication cycle. Next, we treated the exponentially growing cells for 1 h at 30°C with rifampicin, which inhibits the initiation step of DNA replication but allows ongoing DNA synthesis, and cephalexin, which inhibits cell division. Under these conditions, the chromosome number reaches the oriC number at the time of drug addition because ongoing DNA replication completes without initiation of the next round of replication and without cell division. The parent NM1001 cells, a temperature‐resistant transductant of RN4220, displayed a small 2N and a major 4N peak, corresponding to oriC numbers of 2 and 4, respectively (Fig. 2d) (Skarstad , 1995). In contrast, the mutant NM1002 cells, a TS transductant of RN4220, had a 3N peak (oriC number of 3) in addition to the 2N and 4N peaks (Fig. 2c). This result suggests that the dnaA8822 mutant loses the ability to ensure synchronous regulation of chromosome replication initiation even at a permissive temperature.
When parent NM1001 cells were cultured in the presence of cephalexin and switched from 30 to 43°C, the histogram was more widely distributed and was shifted to the right (Fig. 2f). This indicates accumulation of chromosomal DNA as a result of continuous DNA synthesis without cell division. On the other hand, mutant NM1002 cells displayed 2N, 3N, and 4N peaks under these conditions (Fig. 2e). These results suggest that NM1002 cells complete ongoing DNA synthesis but cannot initiate chromosomal replication at the restrictive temperature. Consistent with these findings, NM1002 cells cultured at 43°C for 1 h without drugs had major 1N and minor 2N peaks in an 8 : 1 ratio (Fig. 2g), suggesting that, at the nonpermissive temperature, NM1002 cells can carry out cell division after completion of chromosomal replication.
Biochemical and genetic studies have suggested that the initiation of DNA replication is regulated by the interaction of DnaA with membrane phospholipids (Crooke, 2001). In E. coli, activation of DnaA by acidic phospholipids in vitro is blocked by membrane‐interacting reagents (Sekimizu & Kornberg, 1988), and dnaA mutants are sensitive to these reagents (Shinpuku , 1995; Mizushima , 1996). In addition, dissociation of the adenine nucleotide bound to S. aureus DnaA is stimulated by liposomes containing acidic phospholipids in vitro, and this, in turn, is inhibited by cationic phospholipids (Ichihashi , 2003). Therefore, we examined whether the S. aureus dnaA8822 mutant is sensitive to membrane‐interacting reagents. We found that NM1002 cells were more sensitive than parent NM1001 cells to chlorpromazine, a phenothiazine derivative that changes membrane fluidity (Tanji , 1992) (Fig. 3). Furthermore, in transduction analyses, chlorpromazine sensitivity was cotransduced with temperature sensitivity (Table 3). These results suggest that the chlorpromazine sensitivity of the NM1002 cells is caused by the dnaA8822 mutation.
Decreased DnaA amount in the dnaA8822 mutant cells
A kind of E. coli DnaA TS mutant cells contains a lower concentration of DnaA (Torheim , 2000), and decrease in cellular DnaA amount is associated with asynchrony of chromosome replication initiation (Ogura , 2001). So, DnaA amount in the dnaA8822 mutant was examined by Western blotting (Fig. 4). Exponentially growing cells at 30°C of the parent strain contained 700±100 molecules of DnaA per cell. On the other hand, amount of DnaA in the dnaA8822 mutant was reduced to 40±10 molecules per cell: c. 6% of the parent strain.
In the current study, of 750 TS mutants of Staphylococcus aureus, we identified one whose TS growth was complemented by dnaA. This mutant contained a single amino‐acid substitution. Phage 80alpha transduction experiments showed that the TS phenotype was linked to the drug‐resistance marker that was inserted in the region flanking the dnaA gene. This represents the first report of a dnaA TS mutant in S. aureus. Moreover, this shows that the product of this gene is essential for S. aureus cell growth, which is consistent with previous findings based on antisense RNA expression (Ji , 2001). DNA synthesis by the dnaA mutant was impaired by switching to the restrictive temperature, a response that was suppressed by addition of a plasmid containing the dnaA gene. Finally, flow cytometric analysis revealed that, at the restrictive temperature, DNA replication in these mutant cells was stopped at the initiation step. Collectively, these results suggest that DnaA is essential for the initiation of chromosome replication in S. aureus.
The oriC number in bacterial cells is known to be 2n, where n is a whole number of not less than 0 (Skarstad , 1995). This is due to synchronous initiation at multiple oriC sites. In E. coli, mutations in several genes, including dnaA, are known to cause a loss in synchrony (Skarstad & Boye, 1994). The dnaA mutant of S. aureus cultured at 30°C displayed an odd number of chromosomes, suggesting that it lost the synchronous regulation of replication initiation. Thus, the dnaA gene appears to be involved in the regulation of synchronous chromosomal replication in S. aureus.
The dnaA8822 mutation results in a single amino‐acid substitution mutation, Ala40Thr. This amino acid is predicted to be located in a β‐sheet structure in domain I, which is essential for oligomerization of DnaA, loading of DnaB replicative helicases, and interaction with the novel DnaA initiator‐associating factor, DiaA (Ishida , 2004). A Leu38Ser substitution in E. coli DnaA, which corresponds to Ala40 in S. aureus DnaA, failed to complement the TS phenotype of the dnaA46 mutant, suggesting that Leu38 in E. coli DnaA is essential for the ability of DnaA to initiate replication (Mima , 1999). Furthermore, the finding that the Ala40 mutation leads to a TS phenotype suggests that this residue is important for DnaA activity in all bacteria.
The S. aureus dnaA8822 mutant had higher sensitivity to chlorpromazine, a phenothiazine derivative that changes membrane fluidity. This agrees with previous findings that E. coli dnaA mutants are sensitive to membrane‐interacting drugs (Shinpuku , 1995; Mizushima , 1996). Like these previous reports, our studies show that mutations in the N‐terminal region of DnaA (dnaA508 and dnaA167 for E. coli and dnaA8822 for S. aureus) cause higher sensitivity to phenothiazine derivatives. In vitro, liposomes containing acidic phospholipids facilitate the dissociation of adenine nucleotides from DnaA from E. coli (Sekimizu & Kornberg, 1988), S. aureus (Ichihashi , 2003), and Mycobacterium (Yamamoto , 2002). Despite the fact that the membrane of S. aureus contains lysylphosphatidylglycerol rather than phosphatidylethanolamine, we speculate that, in all bacterial species, DnaA is regulated by membrane phospholipids or that DnaA is involved in membrane homeostasis (Wegrzyn , 1999).
The amount of DnaA in S. aureus cells was determined to be 700±100 molecules per cell, which is similar to that in an E. coli cell of 1000 (Skarstad & Boye, 1994). Cellular amount of DnaA8822 was reduced to 6% of the parent strain. In B. subtilis, synchrony regulation of DNA replication initiation is disturbed if the cellular DnaA amount is reduced to 20% (Ogura , 2001). Thus, the decreased DnaA concentration in dnaA8822 mutant cells might be one reason for the defect of synchrony regulation, that is, control of cellular amount of DnaA is necessary for the synchrony regulation of DNA replication initiation in S. aureus. Moreover, the decreased concentration of DnaA8822 protein could have an additive effect on the temperature sensitivity and chlorpromazine sensitivity of the mutant, in combination with a membrane‐mediated DnaA regulation.
Temperature‐sensitive mutants facilitate the identification of the molecular function of genes. Suppressor mutations of the dnaA mutant isolated in this study are expected to be useful for identifying proteins that regulate the initiation DNA replication. In addition, TS mutants are useful for assaying the specificity of new drugs. Thus, the S. aureus dnaA mutant is expected to aid in the development of drugs that target DnaA as novel means of treating multidrug‐resistant S. aureus.
We thank Drs N. Ogasawara and J. J. Ferretti for kindly providing pMutinT3 and pSF151, respectively. We also thank M. Miyatani, H. Komaki and K. Saito for their technical assistance. This work was supported in part by Grants‐in‐Aid for Scientific Research from JSPS, by the Industrial Technology Research Grant Program in '04 from NEDO of Japan, and by grants from Kyorin Pharmaceutical Co., Ltd. and Genome Pharmaceuticals Co., Ltd. N.I. was the recipient of a predoctoral fellowship from JSPS.