Allele specific synthetic lethality between priC and dnaAts alleles at the permissive temperature of 30 degrees C in E. coli K-12.

BACKGROUND: DnaA is an essential protein in the regulation and initiation of DNA replication in many bacteria. It forms a protein-DNA complex at oriC to which DnaC loads DnaB. DNA replication forks initiated at oriC by DnaA can collapse on route to the terminus for a variety of reasons. PriA, PriB, PriC, DnaT, Rep and DnaC form multiple pathways to restart repaired replication forks. DnaC809 and dnaC809,820 are suppressors of priA2::kan mutant phenotypes. The former requires PriC and Rep while the latter is independent of them. RnhA339::cat mutations allow DnaA-independent initiation of DNA replication.
RESULTS: It is shown herein that a priC303::kan mutation is synthetically lethal with either a dnaA46 or dnaA508 temperature sensitive mutation at the permissive temperature of 30 degrees C. The priC-dnaA lethality is specific for the dnaA allele. The priC303::kan mutant was viable when placed in combination with either dnaA5, dnaA167, dnaA204 or dnaA602. The priC-dnaA508 and priC-dnaA46 lethality could be suppressed by rnhA339::cat. The priC-dnaA508 lethality could be suppressed by a dnaC809,820 mutation, but not dnaC809. Neither of the dnaC mutations could suppress the priC-dnaA46 lethality.
CONCLUSIONS: A hitherto unknown function for either DnaA in replication restart or PriC in initiation of DNA replication that occurs in certain dnaA temperature sensitive mutant strains at the permissive temperature of 30 degrees C has been documented. Models considering roles for PriC during initiation of DNA replication and roles for DnaA in replication restart were tested and found not to decisively explain the data. Other roles of dnaA in transcription and nucleoid structure are additionally considered.

Background

The loading of the DnaB replicative helicase at the E. coli origin of DNA replication (oriC) is a highly regulated and is thought to be a key step in the assembly of the replisome. DnaB makes important contacts with the τ-subunit of DNA Polymerase III Holoenzyme and DNA primase [1]. DnaB loading at oriC during initiation of DNA replication is a sequence specific, cell cycle regulated event dependent on the DnaA and DnaC proteins (reviewed in [2-4]). In vitro, in the presence of several other accessory proteins (i.e. RNA polymerase, DNA gyrase, HU protein), multiple DnaA proteins bind to four asymmetric 9 bp DnaA binding sites in the 225 bp oriC region allowing formation of a protein/DNA complex [5-7]. This in turn causes melting of a nearby AT rich sequence. A complex of DnaB6-DnaC6 then interacts with the DnaA/oriC complex to load DnaB at the AT rich sequence.

It is thought that DnaA may have other roles in the cells in addition to initiation. These additional roles stem from the fact that there are many DnaA binding sites in the chromosome outside the oriC region [8,9] and that DnaA binding to these asymmetric 9 bp sites, can bend the DNA [10]. It can be hypothesized based on large number of potential DnaA binding sites in the chromosome and the ability of DnaA to bend DNA that it may influence the structure of nucleoid. It has been shown that if a DnaA binding site falls within a promoter region that mutations in dnaA can affect the level of transcription from that promoter [11-14]. Thus, mutations in dnaA may have global effects in gene expression and nucleoid structure as well as affecting initiation of DNA replication at oriC.

In E. coli, dnaA is an essential gene. Several different dnaA temperature sensitive mutant alleles have been isolated. Many of these share the property that they are double mutants (dnaA5, dnaA46, dnaA508 and dnaA602 – Table 1 and [15]). Several of the double mutants share a mutation: a change at codon 184 that replaces an alanine with a valine. Additionally, a dnaA850::Tn10 mutant has been isolated. This is only viable in strains that have an alternate, oriC-independent method of initiation of DNA replication [16].

Table 1

dnaA alleles used in this work and some phenotypes

Allele	Amino Acid Change	Domain Affecteda	Ability to replicate P+ plasmids at 30°Cb	Ability to grow in priC303::kan at 30°C
dnaA46	A184VH252Y	IIIIII	No	No
dnaA508	P28LT80I	II	No	No
dnaA5	A184VG426S	IIIIV	Yes	Yes
dnaA167	V157E	III	Yes	Yes
dnaA204	I389N	IV	No	Yes
dnaA602	A184VA347V	IIIlinker	Yes	Yes

Domains are defined according to [25, 35, 36].

Data from [34]

The loading of DnaB by DnaC in E. coli can occur away from oriC at a repaired replication fork. This is governed by the Replication Restart Proteins (RRPs): PriA, PriB, PriC, DnaT and Rep. The genes coding for these products form multiple pathways to identify the proper substrate and then help DnaC load DnaB (Figure 1 and [17]). Since replication restart is thought to be an essential process, the poor viability (versus complete inviability) of priA mutants suggested the availability of alternate pathways. The PriA-independent pathway depends on PriC and Rep [17]. Two types of priA suppressor mutations have been found and both map to the dnaC gene. The first typified by dnaC809 (E176G) and is dependent on the genes in the PriA-independent pathway of replication restart, rep and priC [17]. The second type of priA suppressor has an additional mutation (K178N) in dnaC relative to dnaC809 and makes this protein's suppression of priA mutant phenotypes independent of priC and rep. This dnaC allele is called dnaC809,820 [17,18]. The multiplicity of replication restart pathways may be a general property in Bacteria since Bacillus subtilis also has a similar arrangement of PriA-dependent and PriA-independent pathways [19,20].

Figure 1

This diagram compares the ways in which the replicative helicase can be loaded either from oriC or a repaired replication fork in E. coli. Left side of the diagram indicates the starting substrate to which the replisome is to be loaded. The horizontal arrows indicate the way in which the proteins may interact to load the replisome. The dotted lines represent suppressor pathways.

As mentioned above, DnaA-dependent initiation of DNA replication at oriC and replication restart share several properties. The most important of these is that both strive to make protein-DNA complexes that are recognized by the DnaC protein so that DnaB can be loaded. Previous work by Kogoma and colleagues showed that oriC-DnaA-independent initiation of DNA replication could take place in an rnhA mutant strain [16,21]. This type of initiation of DNA replication was termed Constitutive Stable DNA Replication (cSDR) and is dependent on both recombination and replication restart functions (reviewed in [22] and Sandler, submitted).

To begin testing the roles of priB and priC in cSDR (reviewed in [22]) required the construction of dnaAtsrnhA priB and dnaAtsrnhA priC triple mutant strains. However when trying to construct these strains, we found that priC was required for growth in two different dnaAts strains, dnaA46 and dnaA508, at the permissive temperature of 30°C. When priC303::kan was tested with other dnaAts alleles, the synthetic lethality was found to be allele specific. Two different mutations were found to suppress the priC-dnaA lethality. One was rnhA339::cat, a non-allele specific suppressor of dnaA mutants. The other was dnaC809,820, a suppressor of priC and priA mutations. The first type suppressed both the dnaA508-priC and dnaA46-priC lethality while the latter only suppressed the dnaA508-priC lethality. These studies suggest that priC may have a role in initiation of DNA replication at oriC with certain dnaA alleles and or that dnaA may have an additional role in the cell important for replication restart.

Results

PriC, but not priB, is required for growth in a dnaA508 mutant at the permissive temperature of 30°C

We began this study by asking if priB and priC are required for cSDR. During this study it was found that certain dnaAts could be introduced into a strain containing a priB mutation, but not into a strain containing a priC mutation. The standard P1 transductional cross used for the introduction of the dnaAts mutant alleles is shown in Figure 2. Table 2 shows that the co-transduction frequency between tnaA300::Tn10 and dnaA508 in a wild type strain is 92% (49/53). Using a priB mutant strain as the recipient, the co-transduction frequency between tnaA300::Tn10 and dnaA508 was approximately the same as when the wild type strain was used as the recipient (data not shown). Surprisingly, when a priC mutant was used as a recipient, the co-transduction frequency was 0/72 or 0% (Table 2). This suggested that the priC303::kan and dnaA508 mutation may be synthetically lethal at the permissive temperature of 30°C. It is also formally possible that priC303::kan suppressed the temperature sensitive nature of dnaA508. These two possibilities are tested below.

Figure 2

The tnaA- dnaA region of the E. coli chromosome is diagramed on the lower line. The upper line is symbolic of the DNA introduced by the P1 transduction in the standard cross described in this paper where a tnaA300::Tn10 dnaAts donor is introduced to a dnaA+ recipient. Potential crossover events between the two markers are shown.

Table 2

P1 crosses using dnaAts strains as the donors and isogenic priC303::kan and dnaC mutants as the recipients

Donor		Recipient Strain Number	Recipient Genotype		Ts/total
Strain Number	dnaA		priC	dnaC
JC12390	508	JC13509	+	+	49/53
AL0454	46	JC13509	+	+	60/69
JC12390	508	SS145	303	+	0/72
AL0454	46	SS145	303	+	0/72
SS1750	5	SS145	303	+	7/12
SS1751	167	SS145	303	+	11/11
WM433	204	SS145	303	+	4/5
SS1752	602	SS145	303	+	10/12
JC12390	508	SS1099	303	809	0/63
JC12390	508	SS1100	303	809,820	41/48
AL0454	46	SS1099	303	809	0/32
AL0454	46	SS1100	303	809,820	0/24

aThese are the results of either single or multiple transductions. The numbers show the number of TetR transductants that were also sensitive to growth at 42°C. If the transductions yielded temperature sensitive transductants then only a few transductants from one transduction is reported. The recipient cells were grown at 37°C in Luria broth. They were then treated with donor lysate made on the tnaA300::Tn10 dnaAts strain indicated. The transductions were then spread on Luria plates containing 10 μgm/ml tetracycline and incubated 24–48 hours at 30°C. Transductants were then picked, purified and tested for growth at 42°C by replica plating.

Since it is known that the absence of other gene products (i.e., rnhA [23] and trxA [24]) can suppress the essentiality of dnaA, it is possible that the absence of priC might also suppress the temperature sensitivity of the dnaAts allele. If so, then one should be able to detect the presence of the dnaA mutation on the chromosome of the 42°C resistant transductants. To test this possibility, TetR transductants selected at 30°C were screened for a 42°C R phenotype. These were then further screened for the presence of a restriction site polymorphism (a Eco NI site) created by the dnaA508 mutation. To do this, the dnaA region from the 42°C R TetR transductants was amplified by standard Polymerase Chain Reaction methods using the primers, prSJS480 and prSJS481 (Table 3). The amplified DNA was then restricted with Eco NI. Examination of eight independent 42°CR TetR transductants, constructed using JC12390 as the donor, revealed no case in which a restriction pattern was consistent with the presence of the temperature sensitive allele (data not shown). From these results, it is concluded that the priC303::kan mutation does not suppress the absence of dnaA508 and is synthetically lethal with it.

Table 3

List of Oligonucleotide Primers

Name	5' to 3' oligonucleotide sequence	Position
prSJS283	ATATTGAGTGTTGTCAGC	Upstream of priC
prSJS284	TCCTCCAGCAGCACAATC	Downstream of priC
prSJS480	CCGCGGTCCCGATCGTTTTG	dnaA upstream primer
prSJS481	GCAGGGCGTTGAAGGTGTGG	dnaA downstream primer

PriC303::kan is synthetically lethal with dnaA46 and dnaA508, but not with dnaA5, dnaA167, dnaA204 or dnaA602

It is possible that either the dnaA508-priC303::kan synthetic lethality at the permissive temperature of 30°C is allele specific or occurs with all dnaAts alleles. To distinguish between these two possibilities, several other dnaAts alleles were tested. Selection of a diverse collection of mutations to test was aided by an already large repertoire of characterized dnaAts alleles [15] and the recent elucidation of the crystal structure of DnaA from Aquifex aeolicus [25]. This allowed the selection of several temperature sensitive dnaA alleles that had different amino acids substitutions in different parts of the protein (Table 1). Hence, it was attempted to introduce dnaA5, dnaA46, dnaA167, dnaA204 and dnaA602 into the priC303::kan strain (SS145) using the selectable marker tnaA300::Tn10 as before. Table 2 shows that the synthetic lethality only occurred additionally with dnaA46, but not with dnaA5, dnaA167, dnaA204 or dnaA602. It is concluded that the synthetic lethality between priC303::kan and dnaA508 and dnaA46 at 30°C is allele specific.

The priC303::kan dnaAts synthetic lethality is solely due to the absence of only priC and the presence of the dnaAts mutation

Since both the dnaA and priC genes are in operons, it is possible that the synthetic lethality seen above is due to not just the mutation in priC or dnaA, but also due to that mutation and or polar effects on downstream genes within their respective operons. While the potential polar effects of a priC303::kan insertion mutation are easily envisioned, the potential polar effects of a dnaA missense mutation are less obvious. This is tested here as well because, as introduced above and discussed below, dnaA mutations can have effects on the level of transcription of promoters in which there are DnaA binding sites. Since dnaA binds to its own promoter and autoregulates its own expression [12,26], it is possible that dnaA mutation may effect transcription from its own promoter and subsequently effect dnaN and recF expression.

It was first tested if the synthetic lethality between the dnaA508 and priC303::kan was dependent solely on the priC gene. This was necessary to determine because priC303::kan is an insertion mutation and could be polar on the downstream gene, ybaM. This was tested by cloning priC into a plasmid (pTH1, see Methods) and seeing if the priC plasmid could complement the synthetically lethal phenotype. pTH1, containing just the priC promoter and gene, in the priC303::kan mutant strain (SS145), allowed the dnaA508 allele to be introduced into that strain at the wild type co-transduction frequency of 90% (data not shown).

It was then tested if the synthetic lethality was due to polar effects of dnaA508 or dnaA46 on downstream dnaN-recF expression. This was tested in a similar way. A plasmid, pAB3 [27], that expresses the dnaA gene in trans was introduced into the priC303::kan mutant strain (SS145). This strain was then used as a recipient in a cross with either ALO450 (dnaA46 tnaA300::Tn10) or JC12390 (dnaA508 tnaA300::Tn10). TetR transductants were selected at 30°C. In each case, several transductants were selected and screened for the presence of the dnaAts mutation by backcrosses to JC13509. The temperature sensitive phenotype associated with dnaA508 and dnaA46 was detected in each case (data not shown).

It is concluded that the synthetic lethality seen between priC303::kan and dnaA508 or dnaA46 is due to solely the absence of priC and the presence of the dnaAts mutation.

RnhA mutations suppress the priC-dnaA synthetic lethality

It has been shown that rnhA mutations are non-allele specific suppressors of both dnaAts and dnaA insertion mutations [16]. The mechanism of suppression is thought to be the stabilization of R-loops on the chromosome [22]. To determine if the priC-dnaA synthetic lethality is suppressed by a mutation in rnhA, it was attempted to introduce dnaA508 and dnaA46 into a priC303::kan rnhA339::cat (SS1531) double mutant strain. It was found that this dnaAts rnhA priC triple mutant combination was viable in each case (see SS1543 and SS3032 in Table 4). This suggested that the cause of the priC-dnaA lethality was a defect in the mutant dnaA protein's ability to initiation of DNA replication and that priC has some role in initiation of DNA replication in the dnaA508 and dnaA46 mutant strains.

Table 4

Strain List

Strain	priC	dnaA	dnaC	Relevant Genotype	Source or Derivation
ALO454	+	46	+	tnaA300::Tn10	Kirsten Skarstad
AQ12251	+	+	+	rnhA339::cat	T. Kogoma
CAG18442	+	+	+	thr-34::Tn10	[38]
CM740	+	5	+		Kirsten Skarstad
CM2556	+	167	+		Kirsten Skarstad
CM2733	+	602	+		Kirsten Skarstad
JC12390	+	508	+	tnaA300::Tn10	Lab stock
JC13509a	+	+	+		[18]
JC19008	+	+	809	priA2::kan	[28]
JC19165	303	+	+		[18]
JC19257	+	+	809,820	priA2::kan	[18]
SS145	303	+	+		JC19165 -> JC13509b
SS1091	+	+	809		JC19008 -> SS1213c
SS1092	+	+	809,820		JC19257 -> SS1213c
SS1099	303	+	809		JC19165 -> SS1091b
SS1100	303	+	809,820		JC19165 -> SS1092b
SS1201	+	508	+	tnaA300::Tn10	JC12390 -> JC13509e
SS1213	+	+	+	thr-34::Tn10	CAG18442 -> JC13509d
SS1504	303	508	809,820	tnaA300::Tn10	JC12390 -> SS1100e
SS1531	303	+	+	rnhA339::cat	AQ12251 -> SS145
SS1543	303	508	+	rnhA339::cat	JC12390 -> SS1531f
SS1738	+	+	+	tnaA300::Tn10	Lab Stock
SS1750	+	5	+	tnaA300::Tn10	SS1738 -> CM740e
SS1751	+	167	+	tnaA300::Tn10	SS1738 -> CM2556e
SS1752	+	602	+	tnaA300::Tn10	SS1738 -> CM2733e
SS1793	303	167	+	tnaA300::Tn10	SS1751 -> SS145e
SS1796	303	602	+	tnaA300::Tn10	SS1752 -> SS145e
SS1797	303	5	+	tnaA300::Tn10	SS1750 -> SS145e
SS1798	303	204	+	tnaA300::Tn10	WM433 -> SS145e
SS3032	303	46	+	rnhA339::cat	ALO454 -> SS1531f
WM433	+	204	+	tnaA300::Tn10	Mark Sutton

JC13509 has the following genotype: sulB103 lacMS286 Φ 80dIIlacBK1 argE3 his-4 thi-1 xyl-5 mtl-1 SmR T6R. The lacMS286 Φ 80dIIlacBK1 code for two partial copies of the lac operon [39, 40].

bSelect for kanamycin resistance on minimal media.

cSelect for threonine prototrophy on minimal media. Screen for the presence of the dnaC mutation by the method described in [18].

dSelect for tetracycline resistance on rich media and screen for threonine auxotrophy.

eSelect for tetracycline resistance on rich media at 30°C and then screen for temperature sensitive growth at 42°C.

fSelect for tetracycline resistance on rich media at 30°C and then screen for restriction site polymorphism and or backcross to test the presence of the temperature sensitive allele.

DnaC809,820, but not dnaC809, suppresses the absence of priC in the dnaA508 mutant at 30°C

The above experiment suggested that PriC has a role in initiation of DNA replication in certain dnaA mutants. If so, then suppressors of priC's role in replication restart should not suppress the priC-dnaA synthetic lethality. Two types of replication restart suppressor mutations are known and were tested [18,28]. DnaC809 suppresses the phenotypes of priA2::kan and dnaT822 [28,29]. In vitro, DnaC809 can suppress the absence of all RRPs on several different substrates [30]. In vivo however, priA2::kan suppression requires priC and rep [17]. DnaC809 can be additionally mutated to make the priA suppression both priC and rep independent [17]. This additionally mutated dnaC allele is called dnaC809,820 [18]. To test the above hypothesis, priC303::kan dnaC809 (SS1099) and priC303::kan dnaC809,820 (SS1100) strains were constructed (Table 4) and used as recipients in crosses with the donor P1 from either JC12390 (dnaA508) or AL0450 (dnaA46). Table 2 shows that when dnaC809,820 was used as the recipient and JC12390 as the donor that 41/48 or 83% of the TetR transductants were also temperature sensitive (they inherited the dnaA508 allele). However, when only dnaC809 was used as the recipient, 0/63 TetR transductants inherited the temperature sensitive phenotype. It is concluded that dnaC809,820 can suppress the absence of priC in the dnaA508 mutant and dnaC809 cannot. The dnaA46 allele was additionally tested and was not suppressed by either dnaC809 or dnaC809,820 (Table 3). From this it is concluded that dnaC809,820 is able to suppress the absence of some dnaAts allele. In contradiction to the suggestion of the above rnhA experiment, this result suggests that dnaA may have some role in replication restart necessary in a priC mutant.

Discussion

This paper shows that priC, a gene involved in both the PriA-dependent and PriA-independent pathways for replication restart, is also required for cell viability in two of six dnaAts mutants at the permissive growth temperature of 30°C. These results were surprising on at least two accounts. The first is that in vitro systems for either replication restart or initiation of DNA replication at oriC posed no requirement for the DnaA or PriC protein respectively. The second is that priC has no known role in vivo in initiation of DNA replication (the only reported role is in replication restart [17]) and that dnaA has no known role in replication restart.

One way to answer the question of why the mutations are synthetically lethal is to see what types of mutations may suppress the lethality. RnhA339::cat, a non-allele specific suppressor of dnaA mutants role in initiation of DNA replication, could suppress the priC-dnaA synthetic lethality for both dnaAts mutant alleles. Such suppression is strong evidence that priC and dnaA may both be missing a function needed during initiation of DNA replication. It was further observed, however, that dnaC809,820 (but not dnaC809) could suppress the absence of priC in the dnaA508 mutant. Neither dnaC809 nor dnaC809,820 could suppress the dnaA46-priC synthetic lethality. DnaC809,820 is a PriC-independent suppressor of several genes required in replication restart. This suppression implicates dnaA in replication restart. Thus the inferences from the two types of suppressors seem to contradict one another.

What function(s) in dnaAts strains are missing for initiation of DNA replication that make the strain dependent on priC at 30°C? A structure-function analysis of DnaA would help to answer this question. Briefly, based on alignments of DnaA proteins, the X-ray crystal structure of the DnaA protein from Aquifex aeolicus and much research on the genetics and biochemistry of DnaA, the DnaA protein can be divided into four domains with four proposed functions: Domain I) DnaB recruitment, Domain II) Linker region, Domain III) ATP binding and Domain IV) DNA binding [25,35,36]. Table 1 shows that the six mutations tested substitute amino acids spread throughout DnaA. The two mutants that show a requirement for priC have mutations in Domains I (DnaB recruitment) and III (ATP hydrolysis). However, several of the mutations not requiring priC also affect Domain III. An interesting aspect to dnaA genetics is that many temperature sensitive mutants have two mutations (Table 1 and [2]). The dnaA5, dnaA46 and dnaA602 all have mutations in Domain III near the ATP binding region. Their second mutations cause amino acid replacements in other domains. Unfortunately, the positions of the second changes yield no clues about what might make dnaA508 and dnaA46 mutants require priC for growth at 30°C and why the other four dnaA mutants do not.

What might PriC be doing to help initiation of DNA replication in a dnaAts strain? One idea is that the dnaAts protein is defective in its ability to create a region of ssDNA at oriC. Since the RRPs are also thought to help create regions of ssDNA (away from oriC) so that DnaC binds and loads DnaB, it is possible that PriC may help the DnaA mutants in this endeavor. Another type of model that is formally possible is that PriC may somehow stabilize the dnaAts protein. This seems unlikely, however, given that dnaC809,820 can rescue the synthetic lethality of priC303::kan and dnaA508. Other models may also be possible.

One needs to consider if DnaA may be involved in replication restart. In considering this, one needs to remember that DnaA has the ability to bind DNA at specific sites and bend it. It has been shown that different dnaAts allele can differentially influence the rate of initiation of transcription in some promoters with DnaA binding sites (see below). This in turn can influence replication restart in two ways. First changes in the level of gene expression of a single gene (or groups of genes) may indirectly influence the replication restart process. Second, the ability of DnaA to bind to many sites on the chromosome may influence structure of the nucleoid and the sites at which replication restart may occur.

There are many examples where several phenotypes had been tested systematically for several dnaA alleles (Table 1 and [31-33]). The only other system that seems to have some similarity to the data here is one in which the ability to replicate λ P+ plasmids was investigated [27,34]. Table 1 shows that dnaA46, dnaA204 and dnaA508 all fail to replicate these plasmids while dnaA5, dnaA167 and dna602 can. The model proposed to explain this phenomenon suggests that DnaA is required to activate transcription at the λ Ppromoter and that the dnaA46, dnaA204 and dnaA508 mutations decrease this ability[27]. With the exception of dnaA204, the inability to replicate these plasmids mirrors the ability of the dnaAts mutants to grow in the absence of priC.

The results presented in this paper do not allow one to definitively know whether the synthetic lethality studied here is due to a failure in initiation of replication at oriC or is it due to a failure in replication restart based on the study of the suppressors. Since dnaA mutations can affect more than just initiation of DNA replication, it is tempting to speculate that some other dnaA function: transcription of a particular gene or set of genes or the shape of nucleoid in the priC mutant may contribute or be the cause of the synthetic lethality. Understanding the molecular mechanism underlying the dnaAts-priC synthetic lethality may require appreciation of these other aspects of dnaA biology.

Conclusions

A hitherto unknown function for either DnaA in replication restart or PriC in initiation of DNA replication that occurs in certain dnaA temperature sensitive strains at the permissive temperature of 30°C has been documented. Models considering roles for PriC during initiation of DNA replication and roles for DnaA in replication restart were tested and found not to decisively explain the data. Other roles of dnaA in transcription and nucleoid structure are additionally considered.

Methods

Bacterial strains

All bacterial strains used in this work are derivatives of E. coli K-12 and are described in Table 4. The protocol for P1 transduction has been described elsewhere [37]. All P1 transduction were selected on 2% agar plates containing either minimal or rich media and either tetracycline 10 μg/ml or kanamycin 50 μg/ml final concentration. All transductants were first purified on the same type of media on which they were selected. Tests for temperature sensitivity were then done by replica plating patches of the purified transductants at 30°C and 42°C on solid rich media without any antibiotics. Growth was scored by either the presence or absence of a patch after 24 hours.

Cloning of the priC gene

Wildtype chromosome DNA was used as the template in a standard PCR reaction using prSJS283 and prSJS284 (Table 3) as the priming oligonulceotides. The amplified PCR fragment (that includes the putative promoter) was purified by gel electrophoresis and cloned into the pCR 2.1 using the TOPO-TA cloning system from Invitrogen. The priC containing plasmid was called pTH1.

Authors contributions

TH carried out the initial part of the molecular genetic studies. These were completed by SJS. SJS conceived of the study and wrote the manuscript.