JOURNAL OF BACTERIOLOGY, Oct. 2003, p. 6025–6031 0021-9193/03/$08.00⫹0 DOI: 10.1128/JB.185.20.6025–6031.2003 Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Vol. 185, No. 20
DpiA Binding to the Replication Origin of Escherichia coli Plasmids and Chromosomes Destabilizes Plasmid Inheritance and Induces the Bacterial SOS Response Christine Miller,1 Hanne Ingmer,1,2† Line Elnif Thomsen,2 Kirsten Skarstad,3 and Stanley N. Cohen1,4* Departments of Genetics1 and Medicine,4 Stanford University, Stanford, California 94305-5120; Department of Veterinary Microbiology, Royal Veterinary and Agricultural University, Stigboejlen 4, Frederiksberg C, DK-1870, Denmark2; and Department of Cell Biology, Institute for Cancer Research, Montebello, 0310 Oslo, Norway3 Received 23 May 2003/Accepted 23 July 2003
The dpiA and dpiB genes of Escherichia coli, which are orthologs of genes that regulate citrate uptake and utilization in Klebsiella pneumoniae, comprise a two-component signal transduction system that can modulate the replication of and destabilize the inheritance of pSC101 and certain other plasmids. Here we show that perturbed replication and inheritance result from binding of the effector protein DpiA to AⴙT-rich replication origin sequences that resemble those in the K. pneumoniae promoter region targeted by the DpiA ortholog, CitB. Consistent with its ability to bind to AⴙT-rich origin sequences, overproduction of DpiA induced the SOS response in E. coli, suggesting that chromosomal DNA replication is affected. Bacteria that overexpressed DpiA showed an increased amount of DNA per cell and increased cell size—both also characteristic of the SOS response. Concurrent overexpression of the DNA replication initiation protein, DnaA, or the DNA helicase, DnaB—both of which act at AⴙT-rich replication origin sequences in the E. coli chromosome and DpiAtargeted plasmids—reversed SOS induction as well as plasmid destabilization by DpiA. Our finding that physical and functional interactions between DpiA and sites of replication initiation modulate DNA replication and plasmid inheritance suggests a mechanism by which environmental stimuli transmitted by these gene products can regulate chromosomal and plasmid dynamics. Previous work has identified an Escherichia coli gene, dpiA, whose overexpression in E. coli destabilizes the inheritance of pSC101 and certain other iteron-containing plasmids (19). The DpiA protein, which can function as the effector component of a two-component signal transduction system, and DpiB, which is encoded by an adjacent gene and is the cognate histidine kinase component, show 47 and 39% identity, respectively, to the CitB and CitA proteins, which modulate citrate metabolism in Klebsiella pneumoniae (6). However, while DpiA can regulate transcription from citrate lyase operon homologs in E. coli (19) and DpiB can function as a citrate receptor (21), their roles in citrate metabolism have not been conserved in E. coli, which, unlike K. pneumoniae, cannot use citrate as a carbon and energy source for aerobic growth. The ability of DpiA and DpiB to function as parts of a canonical two-component system and the identification of target promoters for DpiA (19) raised the prospect that destabilization of plasmid inheritance by overproduction of DpiA may result from transcriptional regulation of a gene or genes that control plasmid DNA replication or partitioning. The speculation that DpiA might control a gene involved in plasmid maintenance was supported by the fact that mutations introduced into DpiA at consensus phosphorylation sites affected
the ability of DpiA to both regulate transcription from targeted promoters and destabilize pSC101 inheritance (19). However, here we show that DpiA instead affects plasmid stability and replication by interacting directly and highly specifically with plasmid replication origin sequences that resemble sequences in DpiA target promoters. We further provide evidence that DpiA interaction with replication origin sequences has pleiotropic effects that include induction of the SOS response. MATERIALS AND METHODS Bacterial strains, plasmids, and growth conditions. Bacterial strains and plasmids are listed in Table 1. Plasmid pHI1510 was constructed as described previously for pHI1511 (19) except that a PstI-EcoRV fragment in the citC gene was introduced into StuI-PstI-cleaved pHI1496 (19). Unless otherwise stated, bacteria were grown in Luria broth (LB) medium containing ampicillin (20 g/ml; U.S. Biochemical Corp. [USB]), kanamycin (30 g/ml; USB), or chloramphenicol (20 g/ml; Sigma). For flow cytometry, cells were grown in AB minimal medium (9) supplemented with 10 g of thiamine/ml, 0.2% glucose, and 0.5% Casamino Acids. For glutathione S-transferase (GST) fusion protein production, cells were grown in 2⫻ yeast extract-tryptone (Amersham Pharmacia) and induced with 1 mM isopropyl--D-thiogalactopyranoside. Transformation assays and segregation rates. Cells containing the pSC101 derivatives were transformed using standard procedures (10) with plasmids expressing the various host replication proteins and plated on LB plates containing antibiotics selective for either the incoming or both types of plasmids. Plasmid stability was determined as described by Meacock and Cohen (28) and Tucker et al. (48). DNA preparations and manipulations. Plasmid DNA was isolated by using either Triton X (12) or Qiagen minipreps. DNA digestion by restriction endonucleases was performed according to protocols obtained from suppliers (New England BioLabs and Invitrogen). The gel-purified DNA fragments (purified from agarose gels with a Qiagen gel extraction kit) used were the origin region of pSC101 (4) (the 310-bp EcoRI-SpeI piece of pCM328 [48]), the origin region of P1 (the 230-bp BamHI-HindIII piece of pALA630 [1]), and the origin region
* Corresponding author. Mailing address: Department of Genetics, Stanford University, 300 Pasteur Dr., Stanford, CA 94305-5120. Phone: (650) 723-5315. Fax: (650) 725-1536. E-mail:
[email protected]. † Present address: Department of Veterinary Microbiology, Royal Veterinary and Agricultural University, Stigboejlen 4, Frederiksberg C, DK-1870, Denmark. 6025
6026
MILLER ET AL.
J. BACTERIOL. TABLE 1. Strains and plasmids
Strain or plasmid
Genotypea
Source or reference
Strains PM191 BR5171 WJW61 BL21DE3 SC1088 SC1148
recA, C600 derivative MC4100 Smr araD139 ⌬[lacIPOZYA] U169 strA thi L(cI ind sfiA::lacZ) W3110 with lacIq, ⌬M15/gal490::152 [lacZ::N rex⫹, cI857 (cro-bioA)⌬] F⫺ OmpT hsdSC rB⫺ mB⫺ galS lacZ4075::mini-Tet ⌬(Tcs) ⫽ 3.300 lacI22 relA1 spoT1 thi-1
28 17 37 45 5 35
Plasmids pCM328 pALA630 pAC12 pNU121 pHI1429 pHI1447 pGEX6P1 pHI1735 pHI1737 pHI1739 pDLC8 pACYC184 pDnaB pDnaG pDnaA pHI1510 pHI1511
pSC101 par copy up derivative, Apr P1 ori cloned onto pUC oriC plasmid, Spr Apr cI Tetr under Pr control, pBR322-like pNU121 Apr wt dpiA pNU121, Apr, wt dpiA up regulated GST fusion vector pGEX6P1, with DpiA12D-L from pHI1539 (19) pGEX6P1, with DpiA wt from pHI1447 (19) pGEX6P1, with DpiA57D-E from pHI1712 (19) pSC101 Kmr derivative P15A, Cmr Tcr ⫽ pMJRDnaB, Cmr ⫽ pPL184 DnaG, Cmr Aps pACYC184, DnaA pHI1496 (19) vector with part of citC pHI1496 with citC-dpiB operator region cloned
48 1 43 19 19 19 Amersham Pharmacia This paper This paper This paper 13 11 39 44 46 This paper 19
a
Reference numbers are given in parentheses. wt, wild-type.
of oriC (the 290-bp SmaI-HindII fragment of pAC12 [43]). The fragments were end labeled with Klenow (40), gel purified, and eluted from the gel (27). Protein purification and gel shift assay. DNA fragments containing mutant and wild-type DpiA proteins were cloned into pGEX6P1, resulting in plasmids pHI1735 (DpiA12D-L), pHI1737 (wild-type DpiA), and pHI1739 (DpiA57D-E). GST fusion proteins were made using the Amersham Pharmacia GST fusion system and purified according to the manufacturer: the fusion proteins were bound to glutathione Sepharose 4B (Amersham Pharmacia), and the DpiA fragment was removed from the matrix by treatment with Precission protease (Amersham Pharmacia). The proteins were analyzed on 10% Criterion XT (Bio-Rad) protein gels to assess purity and concentration (also determined by Western blot analysis [19] and Bio-Rad protein assay). Gel shifts were done according to the method of Kustu (22, 34), using buffers and conditions that favored the phosphorylated form of the proteins. Calf thymus DNA was added to a concentration of 50 g/ml to increase specificity. Protein and DNA fragments were incubated for 20 min at room temperature before being loaded onto gels. The specific activity was calculated from gel shifts as the number of micrograms of protein needed to shift 90% of a 0.6-ng sample of the pSC101 origin fragment away from the original banding position, measured by a Molecular Dynamics Storm PhosphorImager or PDI densitometer. Footprint analysis. We modified the DNase I footprinting method for identifying protein binding sites on DNA (15) as follows: DNA-protein binding was carried out as described above; CaCl2 and MgCl2 were added to final concentrations of 2.5 and 5 mM, respectively, and 4 U of DNase I (Amersham) was added. After incubation for 1 min at room temperature, an equal volume of stop buffer was added (3). Mixtures were treated and run on 6% sequencing gels according to the method of Bao and Cohen (3). The DNA ladder showing the sequence of the region footprinted was prepared using a chain-terminating reaction (Sequenase 7-deaza-dGTP sequencing kit; USB) and the oligonucleotide primer complementary to the 5⬘ end of the origin fragment of pSC101. -Galactosidase assay. The SOS response was measured by lacZ fusions to SOS-responsive genes for the amount of -galactosidase activity. Cells were grown in LB at 37°C to mid-log phase. Assays were performed as recommended by Sambrook and Russell (40). Flow cytometry. Exponentially growing cells (optical density at 450 nm ⫽ 0.15) were treated with 150 g of rifampin (Fluka)/ml and 10 g of cephalexin (Eli Lilly)/ml for four to five generations to complete ongoing rounds of replication. Rifampin inhibits transcription, which in turn inhibits initiation of DNA repli-
cation, whereas cephalexin inhibits cell division (7, 42). In the presence of these drugs, cells end up with an integral number of chromosomes (42), which represents the number of origins at the time of drug treatment. In a culture of cells with synchronous initiation, the integral number of chromosomes is 2, 4, or 8, but with asynchronous initiation, 3, 5, 6, or 7 chromosomes appear. Cells treated with rifampin and cephalexin or exponentially growing cells were collected and resuspended in Tris-EDTA buffer and then treated and analyzed as described previously (47). The average cell mass, determined as the average fluorescein isothiocyanate fluorescence per cell, was calculated by taking the average of the fluorescein isothiocyanate fluorescence intensity distribution. The average DNA content per cell, determined as the average Hoechst fluorescence per cell, was calculated by taking the average of the Hoechst fluorescence intensity distribution.
RESULTS Binding of DpiA to origin DNA sequences. DpiA can regulate transcription either positively or negatively (19). Our initial experiments were designed to identify genes whose possible regulation by DpiA affects plasmid stability. Using DNA microarray methods (24), we globally searched for E. coli genes whose expression is differentially altered by upregulation of wild-type DpiA. Of the many genes showing at least twofold alteration in expression during overexpression of DpiA, some regulate known sugar pathways (glycerol pathway genes were induced, and maltose pathway genes were repressed) or are known heat shock or chaperonin proteins (including GroES and GroEL, Hsp70 and Hsp90, and DnaK). These last genes are all controlled through the global regulator H-NS (2), which our earlier experiments had shown to be regulated by DpiA (H. Ingmer, data not shown). Although expression of certain other open reading frames was also perturbed by DpiA overexpression, we observed no change in any gene known to
VOL. 185, 2003
DpiA BINDS ORIGINS, DESTABILIZES PLASMIDS, INDUCES SOS
6027
FIG. 1. DNA sequences and gel shifts with DpiA protein. (A) The sequence of the repeat regions of the cit operator-promoter from K. pneumoniae and E. coli as well as the A⫹T-rich repeats from the origins of the plasmids pSC101 and P1 and the E. coli chromosome, oriC, are shown. (B) Binding of DpiA protein and the mutant proteins to the origin regions of pSC101 and P1 are shown. Lanes 1 to 5 contain 0.6 ng of pSC101 origin region/lane, and lanes 6 to 10 contain 0.6 ng of P1 origin region/lane. Five micrograms of protein was added to each reaction mixture: lanes 1 and 6, no protein added; lanes 2 and 7, 5 l of control GST extract; lanes 3 and 8, 5 g of DpiA12D-L protein; lanes 4 and 9, 5 g of wild-type DpiA protein; lanes 5 and 10, 5 g of DpiA57D-E protein. (C) Binding of DpiA protein and the mutant proteins to pSC101 origin fragment. All lanes have 0.6 ng of the pSC101 origin fragment. The amounts of protein vary to show the binding specificities of various DpiA proteins relative to that of the wild-type protein. Lane 1, no protein; lane 2, 5 l of GST control extract; lanes 3 to 5, 10, 20, and 30 g, respectively, of DpiA12D-L protein; lanes 6 to 8, 7.5, 12.5, and 20 g of wild-type DpiA protein; lanes 9 to 12, 2.5, 5, 12.5, and 17.5 g of DpiA57D-E protein. (D) Binding of DpiA protein to pSC101 origin fragment and oriC. Lanes 1 and 2, 0.2 ng of pSC101 origin fragment; lanes 3 to 5, 0.2 ng of oriC fragment; lanes 1 and 3 have no protein added, lane 2 has 5 g, lanes 4 has 10 g, and lane 5 has 20 g of wild-type DpiA protein added. Lane 4 shows 50% less labeled DNA, and lane 5 shows 90% less labeled DNA, than lane 3 by PhosphorImager analysis; however, no shifted band is visible.
directly affect DNA replication. Additionally, an E. coli gene expression library containing chromosomal gene inserts under the control of a promoter whose constitutive expression is independent of DpiA did not yield any clones that reversed the instability of pSC101 caused by excess DpiA. Given the negative findings summarized above, we considered the possibility that direct interaction of DpiA protein with plasmid DNA can mediate the effect on plasmid stability we had observed during DpiA overexpression. The experiments we subsequently designed are based on the rationale that replication origins of chromosomes and plasmids commonly contain A⫹T-rich regions that resemble the operator region binding site for CitB, the K. pneumoniae homolog of DpiA (29) (Fig. 1A). The A⫹T-rich sequences at replication origin regions are sites where the coordinated actions of the DnaA replication initiation protein and the DnaB helicase unwind duplex DNA, enabling DNA replication to begin (8). Previously we had found that inheritance of pSC101 and F is destabilized by DpiA, whereas the inheritance of plasmid P1 is not (19). Comparison of sequences at the replication origins of these plasmids indicates that only the A⫹T-rich region of destabilized plasmids resembles the cit operator region (Fig. 1A). To learn whether DpiA binds preferentially to the replication origins of plasmids that it destabilizes, we performed gel electrophoretic mobility shift assays. We produced GST fusion proteins containing DpiA or the DpiA mutants characterized previously (19) and overexpressed them, purified them, and tested them for their ability to bind to DNA fragments con-
taining origin regions. Whereas the DpiA-GST fusion protein, which was not active in vivo, showed no origin region binding, purification of DpiA without the GST tag (see Materials and Methods) produced a protein that interacted specifically with DNA fragments containing the replication origin of pSC101 but not the origin of P1 (Fig. 1B). Previous work has shown that mutations that affect the predicted phosphorylation conformation at the two highly conserved aspartic acid residues of response regulators (36) also influence the regulatory properties of DpiA and its ability to alter inheritance of pSC101 (19). The mutated protein DpiA12D-L abolishes the instability normally seen with DpiA overproduction; however, it retains its ability to induce citC expression (19). Mutation at the corresponding position of other response regulators interferes with phosphorylation (36). The mutated protein DpiA57D-E increases plasmid instability and increases citC expression (19); mutation at the corresponding position on several other response regulators results in a constitutively active conformation (22). Further analysis showed that the wild-type DpiA protein and the mutant proteins DpiA12D-L and DpiA57D-E all shifted the origin region fragment of pSC101 (Fig. 1B). However, the three proteins possessed different affinities of binding. More DpiA12D-L mutant protein than wild-type protein was required to alter the migration of an equal amount of the pSC101 origin fragment in gel shift assays, while less of the DpiA57D-E mutant protein was needed (Fig. 1C). These results parallel the effects of phosphorylation on binding of the DpiA homolog, CitB, to its operator sequences in K. pneu-
6028
MILLER ET AL.
J. BACTERIOL. TABLE 2. Induction of the SOS response and reversal Plasmid added or control
None pACYC184 pDnaAd pDnaBd pDnaGd pHI1511 pHI1510 Mitomycin Ce Backgroundf
SOS response measured in Miller unitsa pHI1429b in WJW61 (N::lacZ fusion)
pHI1447c in BR5171 (sfiA::lacZ fusion)
410 409 56 20 350 92 408 329 6
1,898 2,253 722 677 2,114 1,280 30
a All experiments were repeated at least three times, and the numbers are averages. b Produces two times the normal amount of DpiA (19). c Produces 10 times the normal amount of DpiA (19) d Each produces 10-fold the normal concentration of DnaA, DnaB, or DnaG as measured by Western blotting (30). e Mitomycin C was added to a concentration of 100 g/ml for 2 h before samples were taken as a control for the induction of the SOS response. f Background for each strain with no plasmids.
FIG. 2. DNA footprint of DpiA on the A⫹T-rich region of the pSC101 origin. The sequence of the pSC101 origin is known (49), and the sequencing tracts are shown for orientation. The sequence between the known binding sites for DnaA and IHF is shown. The italicized region is the two 13-mers, and the bold region is the area protected by the DpiA protein in the footprint. The footprint and protein purification of DpiA was performed as described in Materials and Methods. The amount of wild-type DpiA added is shown in micrograms; all lanes were treated with DNase I for 1 min.
moniae DNA (9). DpiA is a small (24.5-kDa) protein, but small shifts in the position of the DNA were not seen even with small amounts of protein. Additionally, there was an abrupt change from very little DNA shifted to a large fraction of the DNA shifted with a small increase in the amount of protein. The occurrence of multiple shifted bands (Fig. 1B) suggests that complexes containing different amounts of protein bound to the DNA were formed. From experiments similar to those shown in Fig. 1C, we calculated specific activity in terms of the protein required to shift 90% of the DNA; if the specific activity for wild-type DpiA was assigned a value of 1, the specific activity for DpiA12D-L was 1.6, while the specific activity for DpiA57D-E was 0.5. No shift of the P1 plasmid origin region fragment occurred with any of the DpiA proteins examined under similar conditions (Fig. 1B). To identify the specific sequence in the pSC101 origin that DpiA binds to, we did DNase I footprint assays. Using binding conditions similar to those used for the gel shift assays, we observed that the A⫹T-rich region of the pSC101 origin was protected by wild-type DpiA protein (Fig. 2). The A⫹T-rich region between the DnaA binding site and the IHF binding site includes the two 13-mers (23). Within this region is a 30-bp stretch that is 90% A⫹T, and it was this 30-bp stretch that was protected by DpiA. Effects of DpiA on the chromosomal origin of replication. The A⫹T-rich sequences at the pSC101 plasmid replication
origin resemble those of the chromosomal origin, oriC (Fig. 1A). While DpiA interacted with the oriC A⫹T-rich replication origin, as assayed by gel shift, we found that shifting of the oriC fragment required more than four times as much DpiA protein in order to produce a shift comparable to that seen for the pSC101 origin fragment. Additionally, the DpiA-mediated oriC shift, which was demonstrated by loss of the DNA band observed in the absence of protein (Fig. 1D), resulted in a diffuse smear rather than a clearly shifted band or retention in the wells. Since we had previously observed that DpiA overexpression is associated with inhibition of plasmid DNA replication (19), we investigated whether the chromosomal replication was also affected by production of excess DpiA. In these experiments, perturbation of replication was assessed using the SOS response, which is induced by DNA damage or inhibition of replication (50). We measured SOS induction using lacZ gene fusions to the SOS-controlled promoters of the E. coli sfiA gene (17) or the bacteriophage N gene (37). Twofold overexpression of DpiA (from pHI1429 [19]) and 10-fold overexpression (from pHI1447 [19]) both were found to induce the SOS response (Table 2). We used flow cytometry to observe changes in the cell size and DNA content per cell in E. coli SC1088 when DpiA was overexpressed. Twofold overexpression of DpiA from pHI1429 did not affect the cell growth rate in minimal medium containing CAA and glucose, compared with growth of cells containing the vector plasmid, pNU21 (both had doubling times of 38 min). However, as seen in Fig. 3, this level of DpiA overexpression resulted in an increase in cells having more than the normal complement of chromosomes. A 40% increase in DNA and origin content per cell was seen during DpiA overexpression compared with SC1088 containing the vector plasmid. No abnormality in chromosome number was observed, indicating that all origins within a cell replicated synchronously. A 30% increase in cell size was also observed (Fig. 3), consistent with earlier evidence that induction of the SOS response generates cells having increased size secondary to inhibition of cell divi-
VOL. 185, 2003
DpiA BINDS ORIGINS, DESTABILIZES PLASMIDS, INDUCES SOS
6029
TABLE 3. Stability of plasmids and reversal Plasmid(s) in PM191
pDLC8 pDLC8, pDLC8, pDLC8, pDLC8, pDLC8, pDLC8,
pHI1447a pHI1447a, pDnaAb pHI1447a, pDnaBb pHI1447a, pDnaGb pHI1510 (control), pHI1447d pHI1511 (citC-dpiB operator), pHI1447d
% of cells retaining pDLC8c
100 20 60 80 20 28 86
a
Produces 10 times the normal amount of DpiA (19). Each produce 10-fold the normal concentration of DnaA, DnaB, or DnaG as measured by Western analysis (30). c Loss measured over 40 generations of unselected growth. All experiments were repeated at least three times and the numbers are averages. d Segregations in strain SC1148. b
FIG. 3. Overexpression of DpiA increases the average DNA content and size of cells. Wild-type strain 1088 (in panels A and B and top row of table) and 1088pHI1429, twofold overproducer of DpiA, (in panels C and D and second row of table) were grown in CAA plus glucose. Flow cytometry data of exponential growing cells (A and C) and cells treated with rifampin and cefalexin (B and D) are shown. The average amount of DNA/cell (calculated from panels A and C) and the average mass/cell are relative numbers. The average origin number/cell is calculated from panels B and D.
sion (16). During the SOS response, the larger cells in the population contain a greater-than-normal amount of DNA, albeit a lower-than-normal DNA/mass ratio (32). Notwithstanding these usual concomitants of the SOS response, during DpiA-induced SOS, the DNA/mass ratio was observed to be slightly higher than normal. This possibly suggests slight overinitiation, or stimulation of initiation. Alternatively, this may reflect aberrant chromosome segregation in which some cells have greater-than-normal amounts of DNA while other cells lack DNA. Reversal of effects of overexpression of DpiA on plasmid stability and SOS response. The A⫹T-rich sequences at the replication origins of both the E. coli chromosome and the pSC101 plasmid are the site of DnaA-facilitated attachment of the DnaC-DnaB complex and of DnaB-mediated unwinding of the two strands of duplex DNA (23). We hypothesized that binding of overexpressed DpiA to this region of DNA may compete with binding of the host replication proteins DnaA and DnaB. Consequently, we tested the ability of overexpression of these host replication proteins to compensate for excess DpiA by restoring plasmid stability. Our results (Table 3) showed that a 10-fold excess of either DnaA or DnaB reversed the pSC101 plasmid instability normally observed when the DpiA protein is overexpressed at 10-fold its normal level (19, 30). Excess DpiA also has been observed to lower the copy number of pSC101 plasmids (19). However, in cells containing an excess of both DpiA and either DnaA or DnaB, the copy number of pSC101 returned to approximately normal (data not
shown). Consistent with these findings, the same conditions that reversed plasmid instability also suppressed DpiA-mediated induction of the SOS response; as seen in Table 2, concomitant overexpression of either the DnaA or DnaB protein decreased SOS induction by 68 and 70%, respectively. During the course of these experiments, we found that introduction of a multicopy plasmid (pHI1511 [19]) carrying the citC-dpiB operator region of E. coli reduced the instability of pSC101 plasmids observed in cells overexpressing DpiA (Table 3), whereas introduction of a control plasmid (pHI1510) had no effect. This finding was consistent with the ability of DpiA to bind to A⫹T-rich sequences present in both its target operator region and the replication origin of plasmids whose inheritance it destabilizes. Increasing cellular binding sites for DpiA by the introduction of plasmid pHI1511 also reduced induction of the SOS response mediated by DpiA overexpression (Table 2). DISCUSSION We initially anticipated that the observed destabilization of pSC101 plasmid inheritance by overexpression of the DpiA two-component system effector protein would result from altered expression of a cellular protein governing replication or partitioning of the plasmid. However, we discovered instead that DpiA destabilizes pSC101 by binding directly to the plasmid replication origin. In contrast, DpiA did not interact detectably with the replication origin of the P1 plasmid, whose inheritance it does not affect. The in vitro affinity of mutant DpiA proteins, as measured by their binding to the origin of pSC101 in gel shift assays, correlated with their predicted phosphorylation conformation and the ability in vivo to destabilize pSC101 inheritance. The mutant DpiA57D-E protein predicted to be in the constitutively active conformation (19, 22) maximally destabilized pSC101 plasmid inheritance in vivo (19) and was more effective than the wild-type DpiA protein in shifting the origin region fragment of pSC101. In contrast, the mutant DpiA12D-L protein that was predicted to be unable to undergo phosphorylation (19, 36) did not destabilize pSC101 inheritance and was less effective than wild-type protein in shifting the pSC101 origin region DNA fragment. Footprint analysis indicates that the binding of DpiA to the origin region of pSC101 occurs at the 13-mers, the A⫹T-rich
6030
MILLER ET AL.
region that melts to allow DnaB (the helicase) access to begin unwinding the helix. The A⫹T-rich region of the pSC101 origin contains sequences similar to those in the operator region of the dpi and cit operons (8, 19, 29). In contrast, the cit operon and pSC101 binding sites for DpiA differ from the sequence of the A⫹T-rich region of the replication origin of P1, a plasmid that is not destabilized by excess DpiA protein and whose origin is not detectably bound by DpiA. The A⫹T-rich regions of the pSC101 replication origin and the E. coli chromosomal replication origin, oriC, share overall sequence similarity, and thus, it is not surprising that DpiA also interacts with the A⫹T-rich sequence of oriC. The origin region A⫹T-rich sequences are also the sites of DNA binding of the host replication proteins DnaA and DnaB, and our results suggest that interaction of DpiA with these sites competes with the binding of both of these essential replication proteins to their targets. Consistent with this conclusion, the SOS response, which can be induced by single-strand DNA (41), produced by either DNA damage or inhibition of replication (50), is induced by DpiA overexpression. An excess of the DnaA or DnaB protein compensates for and reverses both the plasmid instability associated with DpiA protein binding to the plasmid origin and the induction of the SOS response associated with DpiA binding to the E. coli chromosome replication origin. These parallel effects suggest that a similar mechanism (i.e., binding of DpiA to origin region sequences) may account for both phenotypes. They further suggest that the DnaA and DnaB proteins normally are in equilibrium with DpiA for their common binding sites. While DnaA and DnaB have been shown to mediate partitioning of pSC101 as well as plasmid DNA replication (18, 30, 31), the roles of these proteins in the separate functions of what has been termed a replication/partitioning complex are distinct (30). The ability of DpiA overexpression to interfere with both processes, taken together with the ability of DnaA or DnaB overproduction to reverse the detrimental effects of excess DpiA, suggests that pSC101 plasmid partitioning, as well as plasmid DNA replication, is dependent on binding of the DnaA and DnaB proteins to the replication origin. In addition to their well-established functions in the initiation of chromosomal DNA replication, DnaA and DnaB proteins may have a role in chromosome segregation—just as they do in plasmid segregation (30). DnaA is a membrane-bound protein (33), and earlier work from our lab has shown that the membrane-binding domain of DnaA is specifically required for proper segregation of pSC101 (30). If this domain is also important for chromosomal segregation as previously speculated (33), disruption of normal chromosomal segregation by excessive DpiA binding to the A⫹T-rich regions of oriC may indicate that the A⫹T-rich region is the region of the chromosome that binds to the cell membrane at segregation. For Bacillus subtilis, the A⫹T-rich region located between the dnaA and dnaN genes must be positioned near the chromosome origin to ensure accurate segregation of the origin at division (20). A number of other proteins have also recently been found to bind to the E. coli chromosomal origin, i.e., SeqA (26), ArcA (25), CspD (51), and the yccV product (14). Rare earlier instances of bacterial cells using the regulators of two-component systems to control DNA replication have been observed. In Caulobacter crescentus, a two-component
J. BACTERIOL.
system transcriptional regulator, CtrA, can modulate the expression of genes required for replication and interact with DnaA binding sites in the origin region to block replication (38). It also was found that in vitro the ArcA two-component regulator protein binds to DnaA binding sites in the chromosome origin of E. coli, and it was speculated that ArcA may be induced during stationary phase to inhibit replication (25). Collectively, these earlier findings and the data reported here suggest that stimuli sensed by two-component signal transduction systems may enable cells to couple DNA replication to changes in the extracellular or intracellular environment. ACKNOWLEDGMENTS These studies were supported by NIH grants AI08619 and GM 26355 to S.N.C. We thank Chris Hackett for performing the experiments with the overexpression library, Jon Bernstein for help with the microarray experiments, Kirsti Solberg Landsverk and Anne Wahl for performing flow cytometry experiments, and Marcia Seyler for editorial help on the manuscript. We also thank Stuart Austin, Judith Zyskind, Tove Atlung, Michael Yarmolinsky, and Jon Kaguni for strains and plasmids. REFERENCES 1. Abeles, A. L., L. D. Reaves, B. Youngren-Grimes, and S. J. Austin. 1995. Control of P1 plasmid replication by iterons. Mol. Microbiol. 18:903–912. 2. Atlung, T., and H. Ingmer. 1997. H-NS: a modulator of environmentally regulated gene expression. Mol. Microbiol. 24:7–17. 3. Bao, K., and S. N. Cohen. 2003. Recruitment of terminal protein to the ends of Streptomyces linear plasmids and chromosomes by a novel telomerebinding protein essential for linear DNA replication. Genes Dev. 17:774– 785. 4. Bastia, D., C. Vocke, J. Germino, and J. Gray. 1985. DNA-protein interaction at the replication origins of plasmid chromosomes. Basic Life Sci. 30:397–414. 5. Biek, D. P., and S. N. Cohen. 1989. Involvement of integration host factor (IHF) in maintenance of plasmid pSC101 in Escherichia coli: mutations in the topA gene allow pSC101 replication in the absence of IHF. J. Bacteriol. 171:2066–2074. 6. Bott, M., M. Meyer, and P. Dimroth. 1995. Regulation of anaerobic citrate metabolism in Klebsiella pneumoniae. Mol. Microbiol. 18:533–546. 7. Boye, E., and A. Lobner-Olesen. 1991. Bacterial growth control studied by flow cytometry. Res. Microbiol. 142:131–135. 8. Bramhill, D., and A. Kornberg. 1988. Duplex opening by dnaA protein at novel sequences in initiation of replication at the origin of the E. coli chromosome. Cell 52:743–755. 9. Clark, D. J., and O. Maaloe. 1967. DNA replication and the division cycle in Escherichia coli. J. Mol. Biol. 23:99–112. 10. Cohen, S. N., A. C. Chang, and L. Hsu. 1972. Nonchromosomal antibiotic resistance in bacteria: genetic transformation of Escherichia coli by R-factor DNA. Proc. Natl. Acad. Sci. USA 69:2110–2114. 11. Cohen, S. N., and A. C. Change. 1977. Revised interpretation of the origin of the pSC101 plasmid. J. Bacteriol. 132:734–737. 12. Cohen, S. N., and C. A. Miller. 1970. Non-chromosomal antibiotic resistance in bacteria. 3. Isolation of the discrete transfer unit of the R-factor R1. Proc. Natl. Acad. Sci. USA 67:510–516. 13. Conley, D. L., and S. N. Cohen. 1995. Isolation and characterization of plasmid mutations that enable partitioning of pSC101 replicons lacking the partition (par) locus. J. Bacteriol. 177:1086–1089. 14. d’Alencon, E., A. Taghbalout, C. Bristow, R. Kern, R. Aflalo, and M. Kohiyama. 2003. Isolation of a new hemimethylated DNA binding protein which regulates dnaA gene expression. J. Bacteriol. 185:2967–2971. 15. Galas, D. J., and A. Schmitz. 1978. DNAse footprinting: a simple method for the detection of protein-DNA binding specificity. Nucleic Acids Res. 5:3157– 3170. 16. George, J., M. Castellazzi, and G. Buttin. 1975. Prophage induction and cell division in E. coli. III. Mutations sfiA and sfiB restore division in tif and lon strains and permit the expression of mutator properties of tif. Mol. Gen. Genet. 140:309–332. 17. Huisman, O., R. D’Ari, and S. Gottesman. 1984. Cell-division control in Escherichia coli: specific induction of the SOS function SfiA protein is sufficient to block septation. Proc. Natl. Acad. Sci. USA 81:4490–4494. 18. Ingmer, H., and S. N. Cohen. 1993. Excess intracellular concentration of the pSC101 RepA protein interferes with both plasmid DNA replication and partitioning. J. Bacteriol. 175:7834–7841.
VOL. 185, 2003
DpiA BINDS ORIGINS, DESTABILIZES PLASMIDS, INDUCES SOS
19. Ingmer, H., C. A. Miller, and S. N. Cohen. 1998. Destabilized inheritance of pSC101 and other Escherichia coli plasmids by DpiA, a novel two-component system regulator. Mol. Microbiol. 29:49–59. 20. Kadoya, R., A. K. Hassan, Y. Kasahara, N. Ogasawara, and S. Moriya. 2002. Two separate DNA sequences within oriC participate in accurate chromosome segregation in Bacillus subtilis. Mol. Microbiol. 45:73–87. 21. Kaspar, S., and M. Bott. 2002. The sensor kinase CitA (DpiB) of Escherichia coli functions as a high-affinity citrate receptor. Arch. Microbiol. 177:313– 321. 22. Klose, K. E., D. S. Weiss, and S. Kustu. 1993. Glutamate at the site of phosphorylation of nitrogen-regulatory protein NTRC mimics aspartyl-phosphate and activates the protein. J. Mol. Biol. 232:67–78. 23. Kornberg, A., and T. A. Baker. 1992. DNA replication, 2nd ed. W. H. Freeman, New York, N.Y. 24. Lee, K., J. A. Bernstein, and S. N. Cohen. 2002. RNase G complementation of rne null mutation identifies functional interrelationships with RNase E in Escherichia coli. Mol. Microbiol. 43:1445–1456. 25. Lee, Y. S., J. S. Han, Y. Jeon, and D. S. Hwang. 2001. The arc two-component signal transduction system inhibits in vitro Escherichia coli chromosomal initiation. J. Biol. Chem. 276:9917–9923. 26. Lu, M., J. L. Campbell, E. Boye, and N. Kleckner. 1994. SeqA: a negative modulator of replication initiation in E. coli. Cell 77:413–426. 27. McDowall, K. J., S. Lin-Chao, and S. N. Cohen. 1994. A⫹U content rather than a particular nucleotide order determines the specificity of RNase E cleavage. J. Biol. Chem. 269:10790–10796. 28. Meacock, P. A., and S. N. Cohen. 1980. Partitioning of bacterial plasmids during cell division: a cis-acting locus that accomplishes stable plasmid inheritance. Cell 20:529–542. 29. Meyer, M., P. Dimroth, and M. Bott. 1997. In vitro binding of the response regulator CitB and of its carboxy-terminal domain to A ⫹ T-rich DNA target sequences in the control region of the divergent citC and citS operons of Klebsiella pneumoniae. J. Mol. Biol. 269:719–731. 30. Miller, C., and S. N. Cohen. 1999. Separate roles of Escherichia coli replication proteins in synthesis and partitioning of pSC101 plasmid DNA. J. Bacteriol. 181:7552–7557. 31. Miller, C. A., H. Ingmer, and S. N. Cohen. 1995. Boundaries of the pSC101 minimal replicon are conditional. J. Bacteriol. 177:4865–4871. 32. Morigen, A. Lobner-Olesen, and K. Skarstad. Titration of the Escherichia coli DnaA protein to excess datA sites causes destabilization of replication forks, delayed replication initiation and delayed cell division. Mol. Microbiol., in press. 33. Newman, G., and E. Crooke. 2000. DnaA, the initiator of Escherichia coli chromosomal replication, is located at the cell membrane. J. Bacteriol. 182:2604–2610. 34. North, A. K., and S. Kustu. 1997. Mutant forms of the enhancer-binding protein NtrC can activate transcription from solution. J. Mol. Biol. 267:17– 36. 35. Pardee, A. B., F. Jacob, and J. Monod. 1959. The genetic control and
36. 37. 38.
39.
40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.
6031
cytoplasmic expression of “inducibility” in the synthesis of B-galactosidase by E. coli. J. Mol. Biol. 1:165–178. Parkinson, J. S., and E. C. Kofoid. 1992. Communication modules in bacterial signaling proteins. Annu. Rev. Genet. 26:71–112. Patterson, T. A., N. Costantino, S. Dasgupta, and D. L. Court. 1993. Improved bacterial hosts for regulated expression of genes from lambda pL plasmid vectors. Gene 132:83–87. Quon, K. C., B. Yang, I. J. Domian, L. Shapiro, and G. T. Marczynski. 1998. Negative control of bacterial DNA replication by a cell cycle regulatory protein that binds at the chromosome origin. Proc. Natl. Acad. Sci. USA 95:120–125. Ruiz-Echevarria, M. J., G. Gimenez-Gallego, R. Sabariegos-Jareno, and R. Diaz-Orejas. 1995. Kid, a small protein of the parD stability system of plasmid R1, is an inhibitor of DNA replication acting at the initiation of DNA synthesis. J. Mol. Biol. 247:568–577. Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. Sassanfar, M., and J. W. Roberts. 1990. Nature of the SOS-inducing signal in Escherichia coli. The involvement of DNA replication. J. Mol. Biol. 212: 79–96. Skarstad, K., E. Boye, and H. B. Steen. 1986. Timing of initiation of chromosome replication in individual Escherichia coli cells. EMBO J. 5:1711– 1717. Smith, D. W., A. M. Garland, G. Herman, R. E. Enns, T. A. Baker, and J. W. Zyskind. 1985. Importance of state of methylation of oriC GATC sites in initiation of DNA replication in Escherichia coli. EMBO J. 4:1319–1326. Stamford, N. P., P. E. Lilley, and N. E. Dixon. 1992. Enriched sources of Escherichia coli replication proteins. The dnaG primase is a zinc metalloprotein. Biochim. Biophys. Acta 1132:17–25. Studier, F. W., and B. A. Moffatt. 1986. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189:113–130. Sutton, M. D., and J. M. Kaguni. 1995. Novel alleles of the Escherichia coli dnaA gene are defective in replication of pSC101 but not of oriC. J. Bacteriol. 177:6657–6665. Torheim, N. K., E. Boye, A. Lobner-Olesen, T. Stokke, and K. Skarstad. 2000. The Escherichia coli SeqA protein destabilizes mutant DnaA204 protein. Mol. Microbiol. 37:629–638. Tucker, W. T., C. A. Miller, and S. N. Cohen. 1984. Structural and functional analysis of the par region of the pSC101 plasmid. Cell 38:191–201. Vocke, C., and D. Bastia. 1983. Primary structure of the essential replicon of the plasmid pSC101. Proc. Natl. Acad. Sci. USA 80:6557–6561. Walker, G. C. 1984. Mutagenesis and inducible responses to deoxyribonucleic acid damage in Escherichia coli. Microbiol. Rev. 48:60–93. Yamanaka, K., W. Zheng, E. Crooke, Y. H. Wang, and M. Inouye. 2001. CspD, a novel DNA replication inhibitor induced during the stationary phase in Escherichia coli. Mol. Microbiol. 39:1572–1584.
