Monomer–dimer control of the ColE1 Pcer promoter

Heather M. Chatwina,1 and David K. Summers1

Department of Genetics, University of Cambridge, Downing Site, Cambridge CB2 3EH, UK1

Author for correspondence: David K. Summers. Tel: +44 1223 333991. Fax: +44 1223 333992. e-mail: dks11{at}cam.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
XerCD-mediated recombination at cer converts multimers of plasmid ColE1 to monomers, maximizing the number of independently segregating molecules and minimizing the frequency of plasmid loss. In addition to XerCD, recombination requires the accessory factors ArgR and PepA. The promoter Pcer, located centrally within cer, is also required for stable plasmid maintenance. Pcer is active in plasmid multimers and directs transcription of a short RNA, Rcd, which appears to inhibit cell division. It has been proposed that Rcd is part of a checkpoint which ensures that multimer resolution is complete before the cell divides. This study has shown that ArgR does not act as a transcriptional repressor of Pcer in plasmid monomers. Pcer is unusual in that the -35 and -10 hexamers are separated by only 15 bp and this study has demonstrated that increasing this to a more conventional spacing results in elevated activity. An increase to 17 bp resulted in a 10- to 20-fold increase in activity, while smaller effects were seen when the spacer was increased to 16 bp or 18 bp. These observations are consistent with the hypothesis that Pcer activation involves realignment of the -35 and -10 sequences within a recombinational synaptic complex. This predicts that a 17 bp spacer promoter derivative should be down-regulated by plasmid multimerization, and this is confirmed experimentally.

Keywords: dimer resolution, cer–Xer recombination system, cell-cycle checkpoint, Rcd, plasmid stability

Abbreviations: GalK, galactokinase

a Present address: Laboratory of Receptor Signalling, The Babraham Institute, Babraham Hall, Cambridge CB2 4AT, UK.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Stable maintenance of bacterial plasmids requires efficient replication control and distribution to daughter cells. High-copy-number plasmids are distributed randomly at cell division and the probability of forming a plasmid-free daughter (p0) is given by p0=2(1–n), where n is the plasmid copy number of the dividing cell (Summers, 1991 ). Any factor which reduces the plasmid copy number increases the frequency of plasmid loss. Multimer formation by homologous recombination is an important cause of multicopy plasmid instability as multimer-containing cells contain fewer independent plasmids. Furthermore, dimers replicate at twice the rate of monomers and accumulate rapidly in the descendants of the cell in which they first appeared. This generates a sub-population of dimer-only cells from which plasmid-free cells arise at high frequency (Summers et al., 1993 ). The multicopy plasmid ColE1 counters the effects of multimerization through the activity of the cer–Xer recombination system (Summers & Sherratt, 1984 ). Recombination between cer sites in a plasmid multimer regenerates monomers and restores the normal copy number. Recombination at cer requires four chromosome-encoded proteins: ArgR (Stirling et al., 1988a , b ), PepA (Stirling et al., 1989 ), XerC (Colloms et al., 1990 ) and XerD (Blakely et al., 1993 ).

cer-mediated dimer resolution is necessary, but not sufficient, for the stable maintenance of ColE1 and related plasmids. A mutation which inactivates the promoter, Pcer, within the cer site reduces plasmid stability despite having no detectable effect on plasmid dimer resolution (Summers & Sherratt, 1988 ; Patient & Summers, 1993 ). Pcer directs the synthesis of an approximately 75 nt transcript, Rcd, whose predicted secondary structure is reminiscent of small antisense RNAs involved in the control of plasmid replication and transposition (Sharpe et al., 1999 ). The level of Rcd is elevated in multimer-containing cells and it has been proposed that the transcript is part of a cell-cycle checkpoint which blocks division until multimer resolution is complete (Patient & Summers, 1993 ). An unusual feature of Rcd action is that growth and division appear to be stopped in a co-ordinated manner since cells recovered from solid medium after Rcd overexpression show cell-cycle arrest just before septation but do not filament (Patient & Summers, 1993 ). In broth culture septation is blocked by Rcd induction and moderate filamentation is apparent, with cells increasing to approximately three times their normal length (Rowe & Summers, 1999 ).

One of the most intriguing aspects of the Rcd checkpoint hypothesis is the activation of Pcer by plasmid multimerization. How might this be achieved? The separation between the -10 and -35 hexamers in Pcer is only 15 bp, compared to the consensus spacing for an Escherichia coli promoter of 17±1 bp (Hawley & McClure, 1983 ; Harley & Reynolds, 1987 ). We have speculated that distortion of the promoter in a cer–cer synaptic complex might align the -10 and -35 regions for recognition by RNA polymerase (Summers, 1998 ). We refer to this as the ‘local twisting hypothesis’, although the distortion may conceivably include elements of both twist and writhe in the strict topological sense. Since recombination (and therefore synaptic complex formation) is limited to cer sites on the same molecule, this would ensure that Rcd was only made in multimer-containing cells. Support for the idea that Pcer can be activated by realignment of its -10 and -35 sequences came from the analysis of , an up-promoter derivative of Pcer in which the -35 hexamer is replaced by a new sequence at a more conventional spacing from the -10 (Patient & Summers, 1993 ; Fig. 1). This derivative promoter is active even in the absence of multimers, suggesting that the short spacer is crucial to proper monomer–dimer control at Pcer.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 1. Sequences of Pcer and two mutant promoters. The -35 and -10 hexamers are underlined and the ArgR binding site in Pcer is shown by a box. The insertion site used to generate promoters with increased spacer length is indicated by a solid arrow. The nucleotide deleted to generate the 14 bp spacer mutant is indicated by a broken arrow. The putative alternative -35 hexamers for Pcer{Delta}T and the up-promoter mutant (Patient & Summers, 1993 ) are shown by broken boxes. The ThaI site used for subcloning the promoter fragments is indicated, as are the transcription start sites (*).

 
In this paper we describe an investigation of factors which influence the activity of Pcer. The results lend support to the idea that realignment of the -10 and -35 regions is indeed crucial to Pcer activation, and the mechanism by which this may be achieved in plasmid multimers is discussed.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacteria and plasmids.
Most strains used in this work were derivatives of Escherichia coli DS941 (AB1157 recF lacIq lacZ {Delta}M15). DS941 argR and DS941 xerC were deficient in cer-mediated recombination. The relevant gene had been inactivated by a transposon insertion (Stirling et al., 1988a , b ; Blakely et al., 1993 ). JC8679 was a recBC sbcA derivative of AB1157 (Bachmann, 1972 ). Plasmids used in this study included pKS490 (Summers & Sherratt, 1988 ), pUC8 (Vieira & Messing, 1982 ) and pKO1 (McKenney et al., 1981 ). Construction of the PcergalK transcriptional fusions based on pKO1 is described in the results. M13mp19 (Messing, 1983 ; Norrander et al., 1983 ; Yanisch-Perron et al., 1985 ) and M13mp19cer were prepared using TG1 (Sambrook et al., 1989 ) as a host strain. M13mp19cer was constructed by inserting the cer site (as a 290 bp EcoRI–HindIII fragment excised from pKS490) between the EcoRI and HindIII sites of the M13mp19 multiple-cloning site.

Media and bacterial transformation.
For the routine growth of bacteria L-broth (Kennedy, 1971 ) and Oxoid iso-sensitest agar were used (37 °C). Where appropriate, 50 µg ampicillin ml-1 (Sigma) was added to the media. Plasmid transformation of calcium chloride-treated cells was by the method of Cohen et al. (1972) . TG1 was maintained on minimal M9 agar (Sambrook et al., 1989 ). For transformation TG1 cells were grown in 2YT broth (Sambrook et al., 1989 ) and transfected cells were spread in soft agar overlays on L-agar plates. Enzymes were used according to the manufacturers’ recommendations. All enzymes were obtained from Promega except ThaI, which was obtained from Gibco-BRL.

Site-directed mutagenesis.
Single-stranded M13mp19cer phage DNA was prepared by phenol extraction of a polyethylene glycol phage precipitate (Sambrook et al., 1989 ) and used as a template for mutagenesis. Site-directed mutagenesis reactions were performed using the Sculptor in vitro mutagenesis system (Amersham). Oligonucleotides for mutagenesis were synthesized by the Protein and Nucleic Acid Chemistry Facility, Department of Biochemistry, University of Cambridge. Oligonucleotides used to introduce extra bases into the promoter (Fig. 1) were degenerate within the 1–4 base insert region. The oligonucleotide to generate a 14 bp intervening spacing (i.e. to remove 1 base) was synthesized with a T missing from the wild-type sequence (Fig. 1). Plaques were picked at random and sequenced to identify mutants. DNA sequences were determined using the Sequenase DNA sequencing kit (version 2.0, US Biochemicals) and an automated DNA sequencer (model 377; Applied Biosystems). Approximately 75% of the plaques obtained by this method were mutant.

Sequencing of Pcer derivative constructs.
Plasmid templates for sequencing were isolated using a Qiagen plasmid mini kit (Qiagen tip-20). DNA sequences were determined using the ABI PRISM Dye Terminator cycle sequencing ready reaction kit with AmpliTaq DNA Polymerase FS (Perkin Elmer), and a model 377 DNA sequencer.

Assay of galactokinase activity.
Transformant colonies were inoculated into L-broth containing 50 µg ampicillin ml-1 and incubated in a 37 °C shaking water bath overnight. They were then subcultured into fresh medium and incubated at 37 °C until an OD650 0·2–0·3 was reached. GalK activity was assayed by the method of McKenney et al. (1981) . Enzyme activity is expressed as the mean±SD of a minimum of six independent assays.

Preparation of dimer plasmids and recombination analysis.
Plasmid DNA was prepared from JC8679 by alkaline lysis (Sambrook et al., 1989 ). DNA was electrophoresed on a 0·85% (w/v) agarose gel and dimer bands were cut out from the gel and eluted into TE buffer (pH 8·0) using the Geneclean kit (Bio 101). DS941 xerC was then transformed with dimer plasmid DNA and transformant colonies containing dimers were identified after preparation of plasmid DNA by standard techniques. Analysis of cer-mediated recombination was as described by Summers & Sherratt (1984 , 1988 ).

Assay of host cell growth.
A single colony of the host E. coli strain harbouring the test plasmid was picked from a selective plate and inoculated into 10 ml L-broth containing 100 µg ampicillin ml-1 and incubated overnight at 30 °C in a shaking water bath. Cultures were diluted in L-broth to OD600 0·005–0·01; a change in density of the culture at 37 °C was followed by measurement of OD600. To prevent the accumulation of plasmid-free cells, antibiotic selection was maintained throughout. The overnight incubation was carried out at 30 °C to reduce Rcd growth inhibition of cells carrying the +CA Pcer derivative. If the overnight incubation was at 37 °C, cells carrying the +CA Pcer derivative became effectively non-viable.

Oligonucleotide probe preparation.
For Northern blot and dot blot analysis oligonucleotide probes were used. Probes were end-labelled using T4 polynucleotide kinase and [{gamma}-32P]ATP (Amersham). The phosphorylation reaction contained Promega T4 polynucleotide kinase buffer, 200 ng probe, and 10 units polynucleotide kinase in 20 µl. After 30 min incubation at 37 °C, the mixture was heated to 90 °C to inactivate the kinase. The labelled probe was then purified using a S200-HR spin column (Pharmacia). To probe for Rcd an oligonucleotide complementary to a region just downstream of Pcer was used: 5'-GGTAAAAATGGCAACAAACC-3'. To probe for ß-lactamase, an oligonucleotide probe with the sequence 5'-CGGGATAATACCGCGCC-3' was used.

Isolation of RNA.
E. coli cells harbouring plasmids were cultured in L-broth containing 100 µg ampicillin ml-1, overnight in a shaking water bath at 30 °C. The cells were then diluted into fresh medium to an OD650 0·01–0·02 and incubated at 37 °C until an OD650 0·3 was reached. The cells were harvested and resuspended in 200 µl of solution 1 (0·3 M sucrose, 0·01 M sodium acetate, pH 4·8). They were lysed by the addition of 200 µl of solution 2 (2% SDS, 0·01 M sodium acetate, pH 4·8) preheated to 65 °C. RNA was isolated by three extractions with acid phenol (Bioline) at 65 °C. The initial phenol extraction was followed by rapid freezing in liquid nitrogen. The RNA was then extracted with phenol/chloroform at room temperature, followed by extraction with chloroform/isoamyl alcohol. Then 10% (v/v) 3 M sodium acetate was added and the RNA was precipitated by adding 2 vols of ethanol (at -20 °C, 90 min). The RNA pellet was washed with 70% ethanol and dissolved in water. DNA was removed by treatment with DNase I (Sigma); the RNA was extracted with phenol/chloroform, re-precipitated and resuspended in water. RNA concentrations were determined by measurement of A260.

Northern blot analysis.
Twenty micrograms of total RNA was electrophoresed on polyacrylamide gels containing 7 M urea (Sambrook et al., 1989 ). RNA was blotted onto Genescreen Plus nylon membrane (DuPont) with a Bio-Rad electroblot apparatus (5 h at 75 V). The transfer buffer used was 0·5x TBE (0·045 M Tris/borate, 0·001 M EDTA, pH 8·0). Pre-hybridization was at 37 °C for 1 h in a solution containing 6x standard saline citrate (SSC), 0·5% SDS, 7·5x Denhardts solution and sonicated salmon sperm DNA (200 µg ml-1). Hybridization was at 37 °C for 24 h in pre-hybridization solution containing the labelled probe. Hybridized membranes were then washed for 4x10 min at 37 °C in 6x SSC containing 0·1% SDS, followed by 2x10 min at 40 °C in 6x SSC containing 0·1% SDS.

Dot blot analysis.
Serial dilutions of RNA samples were prepared and dot blotted onto Genescreen Plus nylon membrane (DuPont) by the method of Sambrook et al. (1989) . Filters were probed for Rcd as described above. Filters were probed for ß-lactamase in the same pre-hybridization solution but the pre-hybridization and hybridization steps were at 40 °C and the washes were at 40 °C and 43 °C.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Site-directed mutagenesis of Pcer
It has been suggested that the distortion (bending or twisting) of Pcer in response to plasmid multimer formation may realign the -10 and -35 sequences and activate the promoter (Summers, 1998 ). Experimentally an alternative way to alter the relative positions of these promoter elements is to add or subtract bases in the spacer region. We therefore constructed a series of Pcer derivatives with altered spacer length, using single-stranded M13mp19cer phage DNA as a template for site-directed mutagenesis. When inserting bases we used degenerate oligonucleotides so that a variety of insert sequences were obtained. This was necessary to disentangle the contributions of spacer sequence and spacer length to promoter activity. The site of insertion was central in the spacer, two bases before the 3'-end of the binding site for ArgR (the ARG box), which overlaps the -35 hexamer and spacer of Pcer (Fig. 1). Forty-one mutants with insertions in the spacer region were isolated (Table 1): four with a 16 bp spacer (+1); ten with a 17 bp spacer (+2); 12 with an 18 bp spacer (+3); and 15 with a 19 bp spacer (+4). Mutants which have an insertion beginning with CA reconstitute the ArgR binding site, and there was at least one such mutant in each of the groups where two or more bases were added. A promoter mutant with a 14 bp spacer (-1) was also generated by removing a T from the wild-type sequence just 3' of the end of the ArgR binding site (Fig. 1).


View this table:
[in this window]
[in a new window]
 
Table 1. Promoter activity of Pcer and derivatives with altered spacer length

 
Dimer resolution assays
A plasmid dimer resolution assay was used to determine to what extent the addition of bases to the Pcer spacer affected recombination at cer. Pcer activity is not required for plasmid dimer-resolution (Summers & Sherratt, 1988 ), but the insertion of a non-integral number of helical turns between the binding-sites in cer for ArgR and XerCD has been shown to inhibit recombination, presumably by interfering with the formation of nucleoprotein recombination complexes (Guhathakurta et al., 1996 ). The +C, +CA, +CAT and +CATT cer site-derivatives were transferred from phage M13mp19 into pUC8 as EcoRI–HindIII fragments. Dimers of pKS490 (which contained wild-type cer) and plasmids containing the +C, +CA, +CAT, and +CATT cer derivatives were transformed into DS941 and DS941 xerC (the latter acting as a recombination-deficient control). Plasmid DNA from the transformant colonies was analysed on an agarose gel for the presence of monomers. As expected, there was no breakdown of plasmid dimers in DS941 xerC (data not shown). In DS941, dimers of pKS490 were resolved efficiently to monomers. Dimers of plasmids containing the +C and +CA insertions showed partial resolution to monomers, but no resolution was observed for plasmids containing the +CAT or +CATT insertions (Fig. 2a). Analysis of dimer resolution with another 12 plasmids containing cer sites with 1–4 bp insertions confirmed the general pattern of these results. Insertions of 1 or 2 bp reduced but did not completely abolish dimer resolution, while insertions of 3 or 4 bp abolished recombination completely (data not shown). Among the sites with 1 or 2 bp insertions the efficiency of dimer resolution varied with the insert sequence. The pattern for the 1 bp insertions was (with best recombination first) G>C>A>T, and CA>CG>AA>GC for the 2 bp insertions. Note that the CA insertion regenerates the consensus sequence for the ARG box.



View larger version (35K):
[in this window]
[in a new window]
 
Fig. 2. (a) Assay of cer-mediated dimer resolution. Dimers of pKS490 (containing wild-type cer) or derivatives with modified Pcer were transformed into DS941. Plasmid DNA from transformant colonies was analysed on a 0·85% agarose gel. The positions of supercoiled monomers and dimers (1m SC, 2m SC) and relaxed monomers and dimers (1m OC, 2m OC) are indicated. Lane 1, monomer marker; lane 2, dimer marker; lane 3, wild-type pKS490 (w-t); lane 4, +C (+1); lane 5, +CA (+2); lane 6, +CAT (+3); lane 7, +CATT (+4). (b) Plasmid DNA from DS941 (lanes 1 and 2) and JC8679 (lanes 3 and 4), containing pKS490 (w-t) or the +CA promoter derivative (+CA). Supercoiled monomers (1m SC), dimers (2m SC), trimers (3m SC) and tetramers (4m SC) are indicated.

 
Transcriptional activity of Pcer derivatives
Pcer and its mutant derivatives were cloned into plasmid pKO1, upstream of a promoter-less galactokinase gene (McKenney et al., 1981 ). This was achieved by excising promoters from the pUC8 derivatives used for the recombination assay on a 157 bp EcoRI–ThaI fragment and inserting them between the EcoRI and SmaI sites of pKO1. The ThaI site lies immediately downstream of the Pcer transcription start (Fig. 1). Transcriptional activity of the PcergalK fusions was assayed in DS941 and the results are shown in Table 1 and Fig. 3. The wild-type promoter had an activity of 80±16 GalK units. Pcer derivatives with a 16 bp spacer showed a small (1·6- to 2·0-fold) increase relative to the wild-type promoter, while derivatives with a 17 bp spacer showed a much more substantial (14- to 21-fold) increase. The majority of derivatives with an 18 bp spacer showed a 4·8- to 9·3-fold increase relative to the wild-type promoter: an exception was the TTT insertion, which gave only a 2·1-fold increase. Derivatives with a 19 bp spacer showed a three- to fivefold decrease in promoter activity relative to the wild-type. The mutant with a 14 bp spacer ({Delta}T) showed a 1·9-fold increase. The data suggest that the optimum spacer length for Pcer activity is 17 bp and that the length of the spacer is a more important determinant of promoter activity than its sequence.



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 3. Promoter activity of Pcer (in terms of GalK activity) and a range of promoter derivatives with altered spacer regions (mutants). Assays were carried out in DS941 (black bars) and DS941argR (hatched bars). The sequence inserted into the spacer is indicated below each pair of results. Error bars extend an equal distance above and below the blocks and indicate SD for six independent assays.

 
One complication in interpreting these data is that the cer ARG box overlaps the Pcer spacer (Fig. 1) and in some cases the insertion alters the ARG box sequence. It is therefore possible that changes in ArgR binding, as well as changes in spacer length, underpin the observed changes in promoter activity. To test this the activities of Pcer and 17 derivative promoters were reassayed in DS941 argR. The activity of these promoters was not substantially different in wild-type and argR backgrounds (Fig. 3), implying that spacer length rather than ArgR binding was the key determinant of promoter activity. This result also excludes the simple hypothesis that transcription from Pcer in plasmid monomers is repressed by ArgR alone.

An isolated Pcer promoter does not show monomer–dimer regulation
We next investigated the monomer–dimer control of Pcer and its derivatives. A key question was whether monomer–dimer regulation is exhibited by isolated promoters or whether the rest of the cer site plays an essential part. It is possible to formulate a plausible hypothesis in which isolated promoters would respond to plasmid dimerization. ArgR and PepA are involved in interactions between pairs of cer sites (Guhathakurta & Summers, 1995 ) and in the model of Hodgman et al. (1998) sites are brought together when PepA forms a bridge between ArgR hexamers bound to their respective ARG boxes. The ARG box overlaps Pcer so that the PepA/ArgR bridge provides a means of communication between promoters in a dimer: this might be necessary, or even sufficient, for their activation.

To investigate monomer–dimer control of isolated promoters we used the PcergalK fusion plasmids whose construction was described in the previous section. These plasmids contain the cer ARG box and Pcer, but the cer sequences downstream of the Rcd transcription start (including the XerCD binding site) have been deleted. Transcription from Pcer and its derivatives was compared in strains containing monomers or dimers of the PcergalK constructs. Plasmids used in the assay contained either wild-type Pcer, the {Delta}T deletion or the +C, +CA, +CAT and +CATT insertions. These four insertion derivatives were chosen because they reconstitute the ARG box. Dimers of all the plasmids showed a small (13–39%) increase in transcription compared to the monomers (Table 2). The fact that this increase was small and that similar increases were seen for the wild-type promoter and its up-regulated derivatives suggests that the effect is unlikely to represent a biologically significant activation of Pcer. It seems more likely to result from differences in the galK dosage in monomers and dimers. This interpretation was supported by the observation that even in an argR mutant strain (where a PepA/ArgR bridge would be impossible) the wild-type PcergalK fusion plasmid showed an increased GalK activity of 15% between monomers and dimers (55±10 and 63±13 GalK units, respectively). We noticed that the monomer plasmids all showed slightly lower GalK activity in the xerC strain than in the wild-type (compare Tables 1 and 2). This may be due to the reduced viability of xerC cells because XerC is required for the resolution of chromosome dimers formed during replication.


View this table:
[in this window]
[in a new window]
 
Table 2. Activity of PcergalK fusions in monomer and dimer plasmids (host DS941 xerC)

 
Analysis of Rcd expression from Pcer
The experiments described so far involved promoters which were excised from their normal sequence environment and fused to a promoter-less galK. However, within the cer site Pcer is likely to be affected by the assembly of a nucleoprotein complex which contains XerC, XerD, PepA and ArgR (Hodgman et al., 1998 ; Sträter et al., 1999 ). To obtain a more realistic picture of Pcer regulation it was necessary to investigate its activity when it was part of an intact cer site. Northern blot analysis was used to detect Rcd produced from pKS490 (wild-type cer) or plasmid derivatives containing cer sites with mutated Pcer. To ensure that the structures of nucleoprotein complexes which assemble at the mutant sites were not radically different from the wild-type, the analysis was restricted to cells containing plasmids with +C and +CA insertions in Pcer. These remained capable of dimer resolution, albeit at a reduced rate (Fig. 2a), implying that functional recombination complexes still formed at these sites. As expected, very little Rcd was detected in the pKS490-containing strain since the plasmid is predominantly monomeric. There was a slight increase for the strain containing the +C mutant promoter and a very large increase for the +CA mutant (Fig. 4). Transcription from the +CA promoter was higher than from , the up-promoter mutant of Pcer (Fig. 1; Patient & Summers, 1993 ). The Northern blot data are consistent with the results of the GalK fusion experiment, where the increases for the +C and +CA mutants were twofold and 20-fold, respectively.



View larger version (61K):
[in this window]
[in a new window]
 
Fig. 4. Northern blot analysis of Rcd expression from Pcer and its derivatives. Total RNA was prepared from strain DS941 carrying the appropriate plasmids. Equal amounts of total RNA were subjected to electrophoresis and probed for Rcd. Lane1, DS941 pKS490 (wt); lane 2, DS941 pKS496 (); lane 3, +C derivative (+C); lane 4, +CA derivative (+CA). The Rcd transcript is indicated by an arrow.

 
Mechanism of Pcer activation and the local twisting hypothesis
We have seen that in isolation Pcer is insensitive to plasmid dimerization. It seems likely that the assembly of a nucleoprotein complex at cer is required to activate the promoter. We have developed this idea in the local twisting hypothesis (Fig. 5), which proposes that Pcer is inactive in the nucleoprotein complex formed in a plasmid monomer (the single-site complex) but is activated when it finds itself in the synaptic complex formed between sites in a dimer. We envisage that in the single-site complex the -35 and -10 sequences are incorrectly aligned for RNA polymerase recognition (Fig. 5a). In the synaptic complex, distortion of the promoter brings these sequences into the correct alignment for transcription to occur. For simplicity the model invokes only twisting of the promoter, with no change in writhe of the DNA. Since the -35 and -10 sequences are separated by 15 bp, twisting must introduce a relative rotation of approximately 69°, which is equivalent to an insertion of 2 bp. This model is supported by both our galK fusion and Northern blot data which demonstrate that promoter derivatives with a 2 bp insertion are extremely active in plasmid monomers. On its own this might be regarded as a trivial result since it is well known that the optimum -35/-10 separation is 17 bp. However, our model makes the important prediction for the regulation of a promoter derivative with a 17 bp spacer (e.g. the +CA derivative). Twisting in the synaptic complex should move the -35 and -10 out of alignment and down-regulate the promoter (Fig. 5b).



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 5. Local twisting hypothesis for Pcer activation. (a) Wild-type Pcer. In the single-site complex (which predominates in strain DS941) Pcer is inactive because the -35 and -10 boxes are misaligned. Plasmid multimerization (stimulated in strain JC8679) leads to synaptic complex formation. This twists the promoter, aligns the -10 and -35 boxes and stimulates transcription. (b) +CA Pcer derivative. In the single-site complex (in strain DS941) the -10 and -35 boxes are already aligned and Pcer is active. Plasmid multimerization (stimulated in strain JC8679) leads to synaptic complex formation. Associated twisting of the promoter misaligns the -10 and -35 boxes and down-regulates the promoter.

 
To test this prediction we compared Rcd production from wild-type Pcer and its +CA derivative in strains DS941 and JC8679. In DS941 plasmid monomers predominate, while in JC8679 plasmid multimerization is stimulated (Fig. 2b). Comparison of Northern blots of RNA from the two strains revealed that less Rcd was produced from the +CA promoter in JC8679 than DS941 (data not shown). However, such comparisons are at best semi-quantitative and may be influenced by changes in plasmid copy-number and the efficiency of RNA isolation from the two strains. To exclude these variables, quantitative dot blot analysis was carried out. Samples of total RNA prepared from DS941 or JC8679 containing either pKS490, the +C plasmid or the +CA plasmid were spotted onto duplicate filters. One was probed for Rcd and the other for ß-lactamase mRNA. The plasmid-borne ß-lactamase gene is transcribed constitutively and can be used to correct for variations in plasmid copy-number and RNA extraction efficiency. Autoradiography was carried out using film pre-flashed according to the manufacturer’s instructions. The autoradiographs were analysed using the densitometry package NIH Imaging 1.52, and the ratios of Rcd to ß-lactamase mRNA were calculated for each combination of plasmid and host (Table 3). The data show that moving pKS490 from DS941 to JC8679 resulted in a 60% increase in Rcd expression, consistent with the higher level of plasmid multimerization in the latter. The +C plasmid showed a similar level of Rcd expression in both hosts. However, moving the +CA plasmid from DS941 to JC8679 resulted in a 2·6-fold decrease in Rcd expression.


View this table:
[in this window]
[in a new window]
 
Table 3. Effect of plasmid multimerization on Rcd expression from Pcer and mutant promoters

 
As an independent (albeit less direct) measure of Rcd production we compared the growth rates of DS941 and JC8679 carrying pKS490 and the +CA insertion. In DS941 the +CA mutation increased the generation time more than fivefold, compared with the same strain containing pKS490 (164±43 min and 32±43 min, respectively; mean±SD, n>=4). The high SD in the generation time of DS941 carrying the +CA plasmid reflected the difficulty in culturing this very unfit strain. However, in JC8679 the +CA insertion had little effect on growth (generation time 51±9 min), causing an increase in generation time of only 25% when compared to pKS490 (42±3 min). Assuming that increased Rcd production is the cause of the increased generation time, these data are consistent with the results of the Northern blot analysis. Theoretically it is possible that the results of the growth rate experiment were due to a difference between DS941 and JC8679 which is unrelated to plasmid multimerization. However, this is excluded by the observation that, as expected if the +CA promoter is down-regulated within the synaptic complex, the growth of JC8679pepA is severely inhibited by the +CA promoter plasmid (data not shown).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A very clear result of this study is that Pcer activity is dependent upon spacer length. The +2 derivatives (spacer length 17 bp) had activities 14- to 21-fold higher than the wild-type promoter, and approximately three times stronger than induced Plac or Pgal (McKenney et al., 1981 ). More modest increases were observed with +1 and +3 derivatives, while derivatives with a 19 bp spacer showed a three- to fivefold decrease in activity. These results are consistent with many previous observations that maximum promoter activity and open complex formation by RNA polymerase occurs with a spacer length of 17 bp and that lengthening or shortening the spacer leads to a decrease in promoter activity (Berman & Landy, 1979 ; Ackerson & Gralla, 1982 ; Mandecki & Reznikoff, 1982 ; Stefano & Gralla, 1982 ; Aoyama et al., 1983 ; Mandecki et al., 1985 ). A less consistent observation was that the Pcer derivative with a 14 bp spacer showed a 1·9-fold increase in promoter activity relative to wild-type Pcer, and was similar to Pcer promoter derivatives with a 16 bp spacer. It is conceivable that this promoter uses alternative -10 or -35 sequences with a more conventional spacing. Inspection of the promoter sequence suggests that RNA polymerase may recognize the normal Pcer -10 hexamer and a weaker TGGTTG -35 hexamer 17 bp upstream (Fig. 1).

Although the activity of Pcer derivatives was determined primarily by spacer length there was some variation within each group, reflecting the fact that the sequence of the spacer region also plays a role. The interaction between RNA polymerase and the spacer is largely non-specific, but if the sequence is intrinsically bent, or has the necessary flexibility to bend in the appropriate orientation, this may assist promoter activation (Collis et al., 1989 ; Lozinski et al., 1991 ). Equally, bends in the spacer that mis-orientate the DNA and disrupt protein–DNA interactions can decrease promoter activity. The Pcer mutant promoter with an 18 bp spacer created by insertion of TTT shows considerably lower promoter activity than the other promoter mutants with an 18 bp spacer. The insertion of TTT at the insertion site generates a run of five Ts. Such a T-tract is a rigid structure and will distort the DNA (Koo et al., 1986 ; Hagerman, 1990 ). Presumably this T-tract bending misplaces the protein–DNA interfaces, thereby reducing promoter activation.

In isolation, our observation that a +2 bp derivative of Pcer shows high, constitutive activity was unremarkable. It did, however, suggest a mechanism by which the wild-type promoter might be activated: instead of inserting extra bases, the effect of a non-optimal spacer length might be counteracted by twisting or bending the promoter. Inserting 2 bp into the spacer is equivalent to a 69° relative rotation of the -10 and -35 sequences; something which could perhaps be achieved by appropriate protein–DNA interactions. A precedent for promoter activation by protein-mediated twisting is provided by the mer operon of Tn501, which encodes the mercuric ion detoxification system. The promoter of the mer operon has a 19 bp spacer region which is essential for induction of the operon in the presence of mercuric ions and for repression in their absence. Induction is controlled by binding of MerR at the mer operator sequence, which is a 7 bp perfect dyad sequence between the -35 and -10 hexamers (Parkhill & Brown, 1990 ). The MerR protein binds to this sequence in the presence or absence of mercuric ions. In the absence of mercuric ions MerR acts as a repressor, but in the presence of mercuric ions the MerR protein is allosterically modified. This modified form of MerR causes underwinding of the DNA at the operator site, which realigns the -35 and -10 hexamers into the orientation found in a promoter with an 18 bp spacer (O’Halloran et al., 1989 ; Ansari et al., 1992 ).

The MerR system is relatively simple, with a single protein activating the promoter by inducing a change in its structure. By analogy it seemed possible that Pcer activation in a plasmid dimer might involve only the two promoters and ArgR hexamers bound to their ARG boxes (the ‘isolated promoter’ model). Alternatively it might require the formation of full nucleoprotein complexes at cer. A previous observation that dimerization of a plasmid containing a PcergalK fusion resulted in increased GalK activity (H. Withers & D. Summers, unpublished data) seemed to support the isolated promoter model but the more detailed study reported here suggests that this is not the case. The small increase in GalK activity observed for fusion plasmid dimers in this work appears likely to reflect a copy-number effect (i.e. dimers of the fusion plasmid had slightly more than half the copy number of monomers) and in any case was not ArgR-dependent. A slight increase in galK gene dosage for dimers is not unexpected because the copy-number control of ColE1-like plasmids exhibits exponential kinetics (Summers, 1998 ).

The lack of support for the isolated promoter model implies that monomer–dimer regulation of Pcer is likely to require the full nucleoprotein complexes which form at cer during dimer resolution. We have published a structural hypothesis for cer–Xer recombination (Hodgman et al., 1998 ) which proposes that nucleoprotein complexes form initially at individual cer sites in plasmid monomers. These single-site complexes are able to pair via an ArgR/PepA bridge which can form between two sites in a dimer or in separate monomers. However, only in the dimer is the paired structure sufficiently long-lived to isomerize into a recombination-competent synaptic complex. Some of the basic concepts of the Hodgman model were included in a model proposed by Sträter et al. (1999) . However, the latter does not discuss structures for single-site complexes but envisages that the synaptic complex forms de novo on a plasmid dimer. In either case, the synaptic complex is the only nucleoprotein assembly unique to a plasmid dimer so we presume that it is within this that Pcer activation must take place.

If Pcer is activated by distortion within the synaptic complex, the +CA derivative is likely to be subjected to similar forces, but they may have different consequences for promoter regulation. For the +CA derivative the -10 and -35 sequences in an isolated promoter are positioned appropriately for recognition by RNA polymerase, and the local twisting model predicts that in the synaptic complex they should be moved out of alignment, deactivating the promoter (Fig. 5). To test this hypothesis we needed to generate a high level of synaptic complexes in the host cells. One possible approach would be to use recombination-deficient mutants of the XerC to trap the synaptic complex. However, there may be subtle differences in structure between a ‘frozen’ synaptic complex and its recombination-proficient counterpart; in any case we have no idea whether Pcer is activated before or after the XerC-mediated strand exchange. Given our poor understanding of the system, we decided it was more appropriate to increase the frequency of synaptic complexes by stimulating recombination. We used the sbcA strain JC8679, in which plasmid multimers are generated continuously by homologous recombination, thus providing a constant supply of substrate for Xer–cer recombination. The demonstration by both Northern blot analysis and growth rate experiments that the +CA promoter derivative is less active in JC8679 than in DS941 (where the plasmids are mostly monomeric) is consistent with the predictions of the local twisting hypothesis. We would not expect complete de-activation of the +CA promoter since only about 50% of the plasmids are multimeric at any time. Furthermore, the synaptic complex modelled by Hodgman et al. (1998) is asymmetric, with only one of the recombining sites subject to distortion during synaptic complex formation (the other site retaining the structure seen in the single-site complex). This would mean that only one wild-type promoter in a synaptic complex would fire and only one +CA derivative promoter would be down-regulated.

Finally it is interesting to contrast our analysis of Pcer with a study of proU promoter regulation. The proU promoter responds to changes in extracellular osmolarity and, like Pcer, has a suboptimal (16 bp) spacer (Jordi et al., 1995 ). Jordi et al. (1995) predicted that if the proU promoter is activated by twisting, increasing the spacer length from 16 to 18 bp would reverse the effect of increased osmolarity on its activity. However, a proU promoter derivative with an 18 bp spacer showed essentially normal osmoregulation. Jordi et al. (1995) discounted the hypothesis that changes in DNA twist activate the proU promoter and proposed instead that changes in promoter flexibility, facilitating writhe, were responsible. In contrast to the proU results, we have found that a Pcer derivative with an 18 bp spacer does indeed show a reversed response to plasmid multimerization, consistent with promoter activation involving twist. Monomer–dimer control of Pcer therefore appears to represent a distinct and novel mechanism of transcriptional control.


   ACKNOWLEDGEMENTS
 
This work was supported by a Project Grant from the UK Medical Research Council. We thank Marion Martin for her technical assistance and acknowledge Helen Withers’ trailblazing work with the galK fusion constructs. We also thank our friends and colleagues in Cambridge and elsewhere for their encouragement, ideas and suggestions.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Ackerson, J. W. & Gralla, J. D. (1982). In vivo expression of lac promoter variants with altered -10, -35, and spacer sequences. Cold Spring Harbor Symp Quant Biol 47, 473-476.

Ansari, A. Z., Chael, M. L. & O’Halloran, T. V. (1992). Allosteric underwinding of DNA is a critical step in positive control of transcription by Hg-MerR. Nature 355, 87-89.[Medline]

Aoyama, T., Takanami, M., Ohtsuka, E., Taniyama, Y, Marumoto, R., Sato, H. & Ikehara, M. (1983). Essential structure of E. coli promoter: effect of spacer length between the two consensus sequences on promoter function. Nucleic Acids Res 17, 5855-5864.[Medline]

Bachmann, B. J. (1972). Pedigrees of some mutant strains of Escherichia coli K-12. Bacteriol Rev 36, 525-557.[Medline]

Berman, M. L. & Landy, A (1979). Promoter mutations in the transfer RNA gene tyrT of Escherichia coli. Proc Natl Acad Sci USA 76, 4303-4307.[Abstract]

Blakely, G., May, G., McCulloch, R., Arciszewska, L. K., Burke, M., Lovett, S. T. & Sherratt, D. J. (1993). Two related recombinases are required for site-specific recombination at dif and cer in E. coli K12. Cell 75, 351-361.[Medline]

Cohen, S. N., Chang, A. C. Y. & Hsu, L. (1972). Non-chromosomal antibiotic resistance in bacteria: genetic transformation of Escherichia coli by R-factor DNA. Proc Natl Acad Sci USA 69, 2110-2114.[Abstract]

Collis, C. M., Molloy, P. L., Both, G. W. & Drew, H. R. (1989). Influence of the sequence-dependent flexure of DNA on transcription in E. coli. Nucleic Acids Res 17, 9447-9468.[Abstract]

Colloms, S. D., Sykora, P., Szatmari, G. & Sherratt, D. J. (1990). Recombination at ColE1 cer requires the Escherichia coli xerC gene product, a member of the lambda integrase family of site-specific recombinases. J Bacteriol 172, 6973-6980.[Medline]

Guhathakurta, A. & Summers, D. K. (1995). Involvement of ArgR and PepA in the pairing of ColE1 dimer resolution sites. Microbiology 141, 1163-1171.[Abstract]

Guhathakurta, A., Viney, I. & Summers, D. K. (1996). Accessory proteins impose site selectivity during ColE1 dimer resolution. Mol Microbiol 20, 613-620.[Medline]

Hagerman, P. J. (1990). Sequence-directed curvature of DNA. Annu Rev Biochem 59, 755-781.[Medline]

Harley, C. B. & Reynolds, R. P. (1987). Analysis of E. coli promoter sequences. Nucleic Acids Res 15, 2343-2361.[Abstract]

Hawley, D. K. & McClure, W. R. (1983). Compilation and analysis of Escherichia coli promoter DNA sequences. Nucleic Acids Res 11, 2237-2255.[Abstract]

Hodgman, T. C., Griffiths, H. & Summers, D. K. (1998). Nucleoprotein architecture and ColE1 dimer resolution: a hypothesis. Mol Microbiol 29, 545-558.[Medline]

Jordi, B. J. A. M., Owen-Hughes, T. A., Hulton, C. S. J. & Higgins, C. F. (1995). DNA twist, flexibility and transcription of the osmoregulated proU promoter of Salmonella typhimurium. EMBO J 14, 5690-5700.[Abstract]

Kennedy, C. K. (1971). Induction of colicin production by high temperature or inhibition of protein synthesis. J Bacteriol 108, 10-19.[Medline]

Koo, H.-S., Wu, H.-M. & Crothers, D. M. (1986). DNA bending at adenine thymine tracts. Nature 320, 501-506.[Medline]

Lozinski, T., Adrych-Rozek, K., Markiewicz, W. T. & Wierzchowski, K. L. (1991). Effect of DNA bending in various regions of a consensus-like Escherichia coli promoter on its strength in vivo and structure of the open complex in vitro. Nucleic Acids Res 19, 2947-2953.[Abstract]

McKenney, K., Shimatake, H., Court, D., Schmeisser, U., Brady, C. & Rosenberg, M. (1981). A system to study promoter and terminator signals recognized by Escherichia coli RNA polymerase. In Gene Amplification and Analysis , pp. 383-415. Edited by J. G. Chirikjian & T. S. Papa. Amsterdam:Elsevier.

Mandecki, W. & Reznikoff, W. S. (1982). A lac promoter with a changed distance between -10 and -35 regions. Nucleic Acids Res 10, 903-912.[Abstract]

Mandecki, W., Goldman, R. A., Powell, B. S. & Caruthers, M. H. (1985). lac up-promoter mutants with increased homology to the consensus promoter sequence. J Bacteriol 164, 1353-1355.[Medline]

Messing, J. (1983). New M13 vectors for cloning. Methods Enzymol 101, 20-78.[Medline]

Norrander, J., Kempe, T. & Messing, J. (1983). Construction of improved M13 vectors using oligodeoxynucleotide-directed mutagenesis. Gene 26, 101-106.[Medline]

O’Halloran, T. V., Frantz, B., Shin, M. K., Ralston, D. M. & Wright, J. G. (1989). The MerR heavy metal receptor mediates positive activation in a topologically novel transcription complex. Cell 56, 119-129.[Medline]

Parkhill, J. & Brown, N. L. (1990). Site-specific insertion and deletion mutants in the mer promoter-operator region of Tn501; the nineteen base-pair spacer is essential for normal induction of the promoter by MerR. Nucleic Acids Res 18, 5157-5162.[Abstract]

Patient, M. E. & Summers, D. K. (1993). ColE1 multimer formation triggers inhibition of Escherichia coli cell division. Mol Microbiol 9, 1089-1095.[Medline]

Rowe, C. D. & Summers, D. K. (1999). The quiescent-cell expression system for protein synthesis in Escherichia coli. Appl Environ Microbiol 65, 2710-2715.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sharpe, M. E., Chatwin, H. M., Macpherson, C., Withers, H. L. & Summers, D. K. (1999). Analysis of the ColE1 stability determinant Rcd. Microbiology 145, 2135-2144.[Abstract]

Stefano, J. E. & Gralla, J. D. (1982). Spacer mutations in the lac ps promoter. Proc Natl Acad Sci USA 79, 1069-1072.[Abstract]

Stirling, C. J., Stewart, G. & Sherratt, D. J. (1988a). Multicopy plasmid stability in Escherichia coli requires host-encoded functions that lead to plasmid site-specific recombination. Mol Gen Genet 214, 80-84.[Medline]

Stirling, C. J., Szatmari, G., Stewart, G., Smith, M. C. M. & Sherratt, D. J. (1988b). The arginine repressor is essential for plasmid stabilizing site-specific recombination at the ColE1 cer locus. EMBO J 7, 4389-4395.[Abstract]

Stirling, C. J., Colloms, S. D., Collins, J. F., Szatmari, G. & Sherratt, D. J. (1989). xerB, an Escherichia coli gene required for plasmid ColE1 site-specific recombination, is identical to pepA, encoding aminopeptidase A, a protein with substantial similarity to bovine lens leucine aminopeptidase. EMBO J 8, 1623-1627.[Abstract]

Sträter, N., Sherratt, D. J. & Colloms, S. D. (1999). X-ray structure of aminopeptidase A from Escherichia coli and a model for the nucleoprotein complex in Xer site-specific recombination. EMBO J 18, 4513-4522.[Abstract/Free Full Text]

Summers, D. K. (1991). The kinetics of plasmid loss. Trends Biotechnol 9, 273-278.[Medline]

Summers, D. K. (1998). Timing, self-control and a sense of direction are the secrets of multicopy plasmid stability. Mol Microbiol 29, 1137-1145.[Medline]

Summers, D. K. & Sherratt, D. J. (1984). Multimerization of high copy number plasmids causes instability: ColE1 encodes a determinant essential for plasmid monomerization and stability. Cell 36, 1097-1103.[Medline]

Summers, D. K. & Sherratt, D. J. (1988). Resolution of ColE1 dimers requires a DNA sequence implicated in the three-dimensional organization of the cer site. EMBO J 7, 851-858.[Abstract]

Summers, D. K., Beton, C. W. H. & Withers, H. L. (1993). Multicopy plasmid instability: the dimer catastrophe hypothesis. Mol Microbiol 8, 1031-1038.[Medline]

Vieira, J. & Messing, J. (1982). The pUC plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19, 259-268.[Medline]

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103-119.[Medline]

Received 17 May 2001; revised 19 July 2001; accepted 26 July 2001.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Chatwin, H. M.
Articles by Summers, D. K.
Articles citing this Article
PubMed
PubMed Citation
Articles by Chatwin, H. M.
Articles by Summers, D. K.
Agricola
Articles by Chatwin, H. M.
Articles by Summers, D. K.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2001 Society for General Microbiology.