Replication fork blockage by transcription factor-DNA complexes in Esc
http://www.100md.com
《核酸研究医学期刊》
School of Medical Sciences, Institute of Medical Sciences, University of Aberdeen Foresterhill, Aberdeen AB25 2ZD, UK 1 Department of Biochemistry, University of Oxford South Parks Road, Oxford OX1 3QU, UK
*To whom correspondence should be addressed. Tel: +44 0 1224 555183; Fax: +44 0 1224 555844; Email: p.mcglynn@abdn.ac.uk
ABSTRACT
All organisms require mechanisms that resuscitate replication forks when they break down, reflecting the complex intracellular environments within which DNA replication occurs. Here we show that as few as three lac repressor-operator complexes block Escherichia coli replication forks in vitro regardless of the topological state of the DNA. Blockage with tandem repressor-operator complexes was also observed in vivo, demonstrating that replisomes have a limited ability to translocate through high affinity protein–DNA complexes. However, cells could tolerate tandem repressor-bound operators within the chromosome that were sufficient to block all forks in vitro. This discrepancy between in vitro and in vivo observations was at least partly explained by the ability of RecA, RecBCD and RecG to abrogate the effects of repressor-operator complexes on cell viability. However, neither RuvABC nor RecF were needed for normal cell growth in the face of such complexes. Holliday junction resolution by RuvABC and facilitated loading of RecA by RecF were not therefore critical for tolerance of protein–DNA blocks. We conclude that there is a trade-off between efficient genome duplication and other aspects of DNA metabolism such as transcriptional control, and that recombination enzymes, either directly or indirectly, provide the means to tolerate such conflicts.
INTRODUCTION
Genome duplication must occur concurrently with many other essential metabolic processes. Chemical damage to the DNA template caused, in part, by the byproducts of normal metabolism is known to block replication (1), but the consequences of duplicating template DNA unavoidably bound by a wide-range of proteins is largely unknown. Both prokaryotic and eukaryotic replication forks can be paused by specialized protein–DNA complexes that have evolved to act as preprogrammed blocks to replication (2,3). However, the majority of proteins associated with chromosomal DNA function in DNA packaging and transcriptional control and their impact on replication fork movement is unclear, although large tandem arrays of transcription factor-operator complexes do provide potent blocks to Escherichia coli chromosomal duplication (4).
Regardless of the source of replication blockage, any block must be removed or bypassed for replication to be restarted. Recombination enzymes play pivotal roles in the repair and restart of blocked replication forks both in prokaryotes and eukaryotes, facilitating bypass of the original block and/or the formation of intermediates upon which the replication machinery can be reloaded (5–7). Indeed, the enhanced recombination noted at both preprogrammed replicative blocks (8–10) and at transcription factor–DNA complexes (11,12) point to the potential problems of duplicating protein-bound DNA.
Studies with E.coli have suggested that unwinding of the nascent DNA at blocked replication forks to form four-stranded Holliday junctions plays an important role in the repair of damaged forks (13). Such junctions may be targeted by RecBCD helicase/nuclease, allowing degradation of the spooled out DNA end and resetting of the fork for subsequent replisome reloading by PriA (14). However, it could also be cleaved by the Holliday junction-specific helicase/endonuclease RuvABC. This cleavage would result in chromosome breakage, release of a free double-stranded DNA (dsDNA) end and a requirement for RecA-catalysed strand exchange with the sister chromosome to form a D-loop onto which the replication apparatus can be reloaded (15,16). The situations in which these processing events might take place are unclear at present, and evidence has been presented that cleavage of reversed forks by RuvABC may be a pathological event occurring only in the absence of RecBCD (7). One problem is that how recombination enzymes act at blocked or otherwise damaged replication forks may depend on the context. Different types of replication block will result in different DNA structures at the block, dictating how the blocked fork will be processed by recombination enzymes, all of which display distinct polarities of action on DNA. Indeed, RecA in conjunction with RecFOR may stabilize replication forks blocked by ultraviolet (UV) light-induced DNA damage to inhibit subsequent processing, rather than to promote processing by catalysing strand exchange (17).
In this study, we have analysed the movement of replication forks through tandem lac repressor-operator complexes and have demonstrated that such high affinity protein–DNA complexes present additive barriers to fork movement in vitro. Lac repressor–operator complexes also inhibited fork movement in vivo. Blockage of fork movement by such barriers lead to a requirement for wild-type recA, recB and recG genes to maintain cell division. However, there was no such requirement for ruvABC, recF, rep or uvrD. We conclude that RecA, RecBCD and RecG act to facilitate cell proliferation in the face of protein–DNA complexes that hinder replisomes, but that these tolerance mechanism(s) do not rely on Holliday junction cleavage by RuvABC or the loading of RecA by RecFOR. Moreover, Rep and UvrD, proposed to act as accessory replicative helicases during chromosomal duplication, do not appear to underpin replication through protein–DNA roadblocks.
MATERIALS AND METHODS
DNAs and proteins
DNA PolIII*, ?, DnaB, DnaC, DnaG, HU and SSB were purified as described (18,19). LacI was purified as indicated (20). DnaA was purified with an N-terminal hexahistidine tag on a nickel NTA superflow column (Qiagen). DNA gyrase was supplied by A. Maxwell (John Innes Centre, Norwich, UK).
Replication plasmids were based on pBROTB535-I (21). pPM308 was constructed from pBROTB535-I by replacement of terB site 1 with EcoRV and XbaI sites. pPM437, pD506 and pIK02 were constructed by ligation of a Klenow-treated XhoI–SalI fragment from a derivative of pFX92 carrying 3, 6 and 22 tandem lacO repeats, respectively (22) into the EcoRV site of pPM308. pPM374, pPM378, pPM393 and pPM379 were constructed by cloning an apramycin resistance cassette (23) linked to 0, 3, 6 or 22 tandem lacO sites between the EcoRI and XbaI sites of pPM308. pPM306 was constructed by amplification of the lacI gene from E.coli MG1655 and cloning into pBAD/HisB (Invitrogen) as an NcoI–HindIII fragment.
In vitro replication assays
Standard oriC DNA replication reactions (15 μl) using DNA gyrase were performed as described (18) except incubation was at 37°C for 10 min. Electrophoretic analysis of replication products under denaturing conditions was performed in 0.7% alkaline agarose gels as described (18) using a 5'-labelled HindIII digest as a size marker. Fork blockage was quantified in standard replication assays by measuring the amount of replication product above the blocked 4 kb leading strand using reactions lacking LacI as the baseline.
Determination of replication fork movement along template DNA relaxed by cleavage with a restriction enzyme was based on a previous pulse-chase approach (24). All reaction components except the restriction enzyme and dCTP (LacI at 100 nM tetramers final concentration as indicated, the standard concentrations of DnaB, DnaC, SSB, ?, PolIII*, HU and DnaG plus DnaA at 300 nM) were assembled on ice in reaction volumes of 14 μl and incubated at 37°C for 3 min. A total of 1 μl of a mixture of SmaI (64 U/μl, Promega) and dCTP (0.3 pmol; 200 kBq/pmol) was then added to each reaction and incubation continued for 2 min at 37°C before reactions were stopped by addition of 1 μl of 0.5 M EDTA. This concentration of SmaI was sufficient to cleave all the template plasmid within 20 s. Reactions were then analysed under denaturing conditions as described above. The degree of fork blockage using this approach was estimated by comparing the 2 kb leading strand peak heights and setting the blockage as 1 for 22 lacO sites—blockage was taken to be complete at lacO22 based on the absence of leading strand product above 4 kb in the presence of LacI (Figure 3B, lane 9).
Strain construction
To integrate lac operators into argEC, the same 22 tandem lac operators used to construct oriC replication plasmids described above and a fragment containing 34 tandem lac operators were each cloned downstream of an apramycin resistance cassette (23) and then subcloned between the HindIII and EcoRV sites of a plasmid carrying argE and the 5' end of argC (25). The apramycin resistance gene was also cloned in the same manner but with no associated lacO sites. The apramycin cassette ± lacO together with flanking argEC sequences was then amplified and integrated into DY330 as described (26). These loci were then moved by P1 vir transduction from DY330 into PM203 (MG1655 ara714). The resultant strains were BP38 (lacO0), PM222 (lacO22) and BP41 (lacO34) (see Table 1). Successful integration was confirmed by arginine auxotrophy, Southern blotting, diagnostic PCR and DNA sequencing of PCR products. Derivatives of PM222 and BP41 with mutations in recombination and replication genes were made by P1 vir transduction (see Table 1).
Table 1 Strains used in this study
Growth experiments
Strains were transformed with pPM306 and selected on LB ampicillin plates overnight at 37°C. Single colonies were grown in 10 ml Luria–Bertani (LB) ampicillin in 50 ml conical flasks at 37°C with shaking to mid log phase and then inoculated into fresh 10 ml LB ampicillin, together with glucose, arabinose and isopropyl-?-D-thiogalactopyranoside (IPTG) as indicated, to an A650 of 0.01. Growth was continued at 37°C and samples removed at 0, 1, 2, 3 and 4 h after subculturing and viable cell counts performed on LB ampicillin IPTG plates at 37°C overnight. Concentrations of ampicillin, apramycin, glucose, arabinose and IPTG were 100 μg/ml, 50 μg/ml, 0.1%, 0.02% and 1 mM, respectively. Comparison of the number of cell divisions was performed using the increase in the number of viable cells between 0 and 4 h in the above growth experiments. Growth experiments were performed between two and eight times for each strain.
Electrophoretic analysis of in vivo replication intermediates
Strains were grown as described above for the growth experiments except that the culture volumes were increased to 40 ml in 250 ml conical flasks. Upon addition of arabinose plus or minus IPTG, growth was continued at 37°C with shaking until the A650 reached 0.4 (3 h after subculturing). Genomic DNA plugs were made and processed as described (25) except that digestion was with PvuII. Electrophoresis was performed through 0.8% agarose gels in 1x TBE for 6 h at 3.5 V/cm, the gels were Southern blotted and then probed using the complete apramycin resistance gene labelled by random priming with dCTP. Gene ruler 1 kb DNA ladder (Fermentas) was used as a marker.
RESULTS
Movement of replisomes in vitro is inhibited by LacI–lacO complexes
To assess the ability of the replication apparatus of E.coli to duplicate protein-bound DNA, the impact of lac repressor-operator complexes on replisome movement was analysed in vitro. DnaA-directed bidirectional replication of oriC-containing plasmids resulted in generation of 0.5 kb lagging strands plus 3 kb leading strands formed by convergence of the two forks in a region opposite oriC (Figure 1A, lane 1) (18). The sizes of leading and lagging strands generated on a plasmid lacking lacO sequences was not altered upon addition of lac repressor (LacI) (Figure 1A, lanes 1–4). Non-sequence-specific binding of LacI was not therefore a major impediment to replisome movement. However, addition of LacI to plasmid encoding 22 lacO sites located 2 and 4 kb from oriC lead to accumulation of leading strands of 2 and 4 kb in size (Figure 1A, lanes 5–8). Addition of the inducer IPTG to reactions with LacI on template DNA containing lacO22 precluded appearance of 2 and 4 kb leading strand products (Figure 1B). These data demonstrate that LacI–lacO complexes inhibit replisome movement, allowing convergence of the two forks at lacO22.
Figure 1 lac repressor–operator complexes block replisome movement in vitro. (A) Replication of plasmid templates with no lac operator sequences (pPM308, lanes 1–4) and with 22 tandem lacO sequences (pIK02, lanes 5–8) in the presence of 0, 5, 25 and 100 nM LacI tetramers. Products of replication were analysed on a denaturing agarose gel, with DNA size markers shown in kb. (B) Production of distinct 2 and 4 kb leading strand products on template pIK02 (lacO22) bound by LacI is abrogated by addition of IPTG. LacI (100 nM tetramers) and IPTG (1 mM) were present as indicated.
To determine the efficacy of replisome blockage by LacI–lacO complexes a series of plasmids were constructed bearing 0, 3, 6 and 22 lacO sites 2 and 4 kb away from oriC. Replication of each template in the presence of LacI resulted in accumulation of the 2 and 4 kb leading strands expected if forks stopped at the LacI–lacO complexes (Figure 2A). At 100 nM LacI tetramers, 24, 43 and 90% of replisomes were blocked at 3, 6 and 22 lacO sites, respectively (Figure 2B). The lacO3 and lacO6 data suggest that the effects of LacI–lacO complexes on replisome movement are additive with one lacO site blocking 5–10% of forks when bound by LacI.
Figure 2 Additive blockage of replisome movement by repressor–operator complexes. (A) Replication of plasmid templates bearing 0, 3, 6 and 22 lacO sites (pPM374, 378, 393 and 379) in the presence of 0, 5, 25 and 100 nM LacI tetramers was monitored by denaturing agarose gel electrophoresis. DNA size markers are shown in kb. (B) Replisome blockage as a function of LacI concentration. Open circles, lacO0; filled circles, lacO3; open squares, lacO6; filled triangles, lacO22.
Inhibition of replication by repressor–operator complexes occurs on relaxed as well as supercoiled template
The above data suggest that the energy input required to disrupt lac repressor–operator interactions presents a barrier to replication. However, the binding of repressor to tandem operators may have induced topological alterations within the plasmid template leading to replication fork stalling, possibly by inhibiting access of DNA gyrase to the template DNA ahead of the replisome. To determine whether altered topology was the primary determinant in replication blockage, the effect of lac repressor–operator complexes on fork movement along a topologically unconstrained DNA template was analysed.
Omission of a topoisomerase from replication reactions still allows initiation of replication at oriC in negatively supercoiled plasmid template (24). However, only one of the two forks initiates DNA synthesis and the fork that does initiate stalls after about 1 kb of DNA synthesis due to accumulation of positive torsional strain (27). This inhibition of fork movement can be relieved by addition of a restriction enzyme (SmaI) that cleaves near oriC thus relieving any topological constraint by linearization of the template (24) . Inclusion of radiolabelled dCTP at the time of SmaI addition allows fork progression to be monitored.
Figure 3 Replication blockage occurs on linearized as well as supercoiled template DNA. (A) In the absence of a topoisomerase, replication can initiate at oriC but is inhibited after about 1 kb of synthesis (i). Addition of SmaI results in cleavage of the DNA near to oriC (ii and iii) and allows one of the two forks to progress around the now-linearized template. Unimpeded replication would generate leading strands of 6 kb for both forks (v and vii) whereas 4 and 2 kb leading strands would result from blockage of the forks at lacO (iv and vi). (B) Replication of pPM308 (lacO0), pPM437 (lacO3), pD506 (lacO6) and pIK02 (lacO22) in the absence of a topoisomerase but with LacI and SmaI as indicated, monitored by denaturing agarose gel electrophoresis. DNA size markers are shown in kb. The position of early replication intermediate (ERI) (37), which accumulated in the absence of SmaI, is shown using pPM308 as template (lane 1).
To determine whether replication forks were blocked by repressor–operator complexes on linear template DNA, plasmids containing 0, 3, 6 or 22 lacO sites were employed. Forks moving rightwards would generate 4 and 6 kb leading strands upon addition of SmaI, depending on whether blockage occurred at lacO. Similarly, forks moving leftward would generate 2 or 6 kb leading strands .
2 and 4 kb leading strands were formed in the presence of LacI on template with 3, 6 or 22 lacO sites but were absent when LacI was absent, or when there were no lacO sites within the template (Figure 3B). Thus replication fork movement was inhibited by repressor–operator complexes on linearized template DNA. The reduction (lacO6) or absence (lacO22) of 6 kb leading strand products supported this conclusion (Figure 3B, lanes 7 and 9). The degree of blockage was 30 and 70% for 3 and 6 lacO sites, respectively, similar to the inhibition of replication estimated on covalently closed supercoiled template (Figure 2). We conclude that blockage of replication forks by repressor–operator complexes occurs to a similar extent regardless of the topological state of the DNA template.
LacI–lacO complexes present barriers to DNA replication in vivo
Two hundred and forty tandem lac operators bound by repressor on the E.coli chromosome provide a complete block to replication and cell proliferation in vivo (22). The in vitro data in Figures 1–3 suggest that far fewer repressor–operator complexes are needed for effective replication blockage. We therefore analysed the ability of replisomes to traverse 22 and 34 lacO sites integrated within the chromosome of E.coli (Figure 4A). To ensure that the lacO sites were saturated with repressor a plasmid (pPM306) encoding lacI under the control of an arabinose-inducible promoter was introduced into the strains. Upon addition of arabinose to the growth medium this plasmid generated an increase in levels of repressor from approximately 300 tetramers per cell to 10–20 000 tetramers per cell, with significant overexpression beginning 1 h after addition of arabinose (data not shown).
Figure 4 lac repressor–operator complexes present blocks to DNA replication in vivo. (A) Insertion of tandem lac operators within the chromosomal arg locus, linked to an apramycin resistance gene. The direction of replication fork movement is indicated with a dashed arrow. (B) Southern blot of a 1D agarose gel of PvuII digests from strains bearing 0 (BP38), 22 (PM222) and 34 (BP41) lacO repeats together with pPM306, a plasmid bearing an arabinose-inducible lacI gene, both with and without 1 mM IPTG. DNA size markers are shown in kb. The entire apramycin resistance gene was used to probe the blot. The positions of replication intermediates are marked with arrows. (C) Growth curves of strains BP38 (i), PM222 (ii) and BP41 (iii) containing pPM306. Open circles, medium supplemented with glucose (lacI expression repressed); filled circles, medium supplemented with arabinose (lacI expression induced). c.f.u.: colony forming units.
In the absence of lacO sites, no slowly migrating DNA species indicative of replication intermediates could be detected as judged by Southern blot analysis of chromosomal DNA digests (Figure 4B, lanes 1 and 2). Replication intermediates did accumulate at 22 and 34 lacO sites, with more intermediates accumulating at lacO34 as compared with lacO22 (Figure 4B, compare lanes 3 and 5). These slowly migrating species were absent when cells were grown in medium supplemented with IPTG (Figure 4B, lanes 4 and 6). The requirement for lacO sites to observe these slowly migrating species, and the effect of IPTG, suggest that these DNA species were replication intermediates caused by blockage of replisomes by repressor–operator complexes. Similar slowly migrating DNA species have been observed upon blockage of replication by Tus–ter complexes (28,29). However, growth of lacO22 and lacO34 strains did not display any major defect in the presence of elevated lac repressor expression (Figure 4C) although there was a small decrease in growth of the lacO34 strain in arabinose as compared with glucose (Figures 4C, 6A and B).
Given the blockage of DNA replication by tandem lac repressor–operator complexes observed in vitro (Figures 1 and 2) it was surprising there was little effect on growth upon induction of lac repressor expression. A trivial explanation for this discrepancy was that neither the lacO22 nor the lacO34 loci were saturated with repressor, even at the high concentrations of repressor achieved upon induction of overexpression. Levels of replication intermediates detectable by gel electrophoresis at the lacO22 and lacO34 loci were compared therefore in the presence of native and overexpressed levels of repressor. Replication intermediates detected at lacO34 increased from 1.1% (±0) as a proportion of the total DNA signal with wild-type levels of repressor to 5.5% (±1.5) upon induction of repressor overexpression. This increase demonstrates that tandem lacO sites were accessible to enhanced levels of repressor. In contrast, the levels of replication intermediates detected at lacO22 were similar in the presence of wild-type or overexpressed levels of repressor (0.7% and 1.2% , respectively). The lack of enhanced blockage at the lacO22 locus, in spite of the accessibility of tandem lacO sites to enhanced levels of repressor, indicates that lacO22 was saturated at wild-type repressor levels. Concerning lacO34, these data cannot unambiguously establish whether the locus was saturated with repressor but, given the 30- to 60-fold enhancement of repressor concentration obtained upon induction of overexpression, it is likely that lacO34 was also saturated in the presence of elevated levels of repressor.
Although the above electrophoretic analysis of replication intermediates did indicate saturation of lacO22 with repressor, this analysis also suggested that replication fork blockage in vivo (1% for lacO22) was much less efficient than that seen at lacO22 in vitro (100%). However, such direct comparisons are inappropriate since the ratios of replication intermediates to linear DNA fragments as observed by 1D and 2D gel techniques may not correlate directly with the actual levels of replication fork blockage—nicking or branch migration of replication intermediates to form linear DNA structures during their isolation likely result in underestimates of fork blockage (4).
Enzymes of recombination facilitate tolerance of repressor–operator complexes in vivo
If the lack of any major growth defects in cells harbouring 22 and 34 lacO sites bound by repressor could not be explained by sub-saturating levels of repressor, what other factors could explain this discrepancy? Fork movement through protein–DNA complexes may be facilitated by accessory factors, such as Rep helicase in vivo (30), factors which were absent from the in vitro replication reactions. Alternatively, direct or indirect processing of blocked replication forks, possibly by recombination enzymes (5–7), may have ameliorated the effects of protein–DNA blocks to replication in vivo. Therefore we examined the genetic requirements for growth of strains bearing either 22 or 34 lacO sites by analysing growth in the absence or presence of elevated levels of lac repressor.
Deletion of rep had no major effect on growth with either 22 or 34 lacO sites (Figures 5 and 6). Rep did not therefore have a critical role in promotion of fork movement through these repressor–operator complexes. We also tested whether UvrD, implicated in a fork clearing role (31), aided cell growth in the face of repressor–operator complexes. However, a uvrD mutation had no effect on the ability of cells to grow in the face of lacO34 (Figure 6A and B).
Figure 5 Recombination enzymes are not required in vivo for tolerance of 22 lacO sites bound by repressor. (A) Growth of strains harbouring 22 lacO sites and the indicated mutations, plus the lacI plasmid pPM306. Open circles: cells grown in glucose (no lac repressor overexpression); filled circles: cells grown in arabinose (elevated lac repressor). lacO22 strains were PM222 (otherwise wild-type), BP13 (recA), BP22 (recB), BP60 (recF), BP33 (recG), BP19 (ruvABC), BP16 (rep). Individual growth curves are shown but each curve was performed between two and eight times with all curves yielding similar results. C.f.u.: colony forming units. (B) Mean number of cell divisions between 0 and 4 h of growth for strains bearing 22 lacO sites, determined from growth curves as represented in (A). Open bars: cells grown in glucose; filled bars: cells grown in arabinose. Strains are as described in (A).
Figure 6 RecA, RecB and RecG are needed in vivo to tolerate 34 lacO sites bound by repressor. (A) Growth of strains harbouring 34 lacO sites and the indicated mutations, plus the lacI plasmid pPM306. Open circles: cells grown in glucose (no lac repressor overexpression); filled circles: cells grown in arabinose (elevated lac repressor); filled triangles: cells grown in arabinose plus IPTG. Strains were BP41 (otherwise wild-type), BP43 (recA), BP45 (recB), BP44 (recF), BP54 (recG), BP52 (ruvABC), BP47 (rep), BP55 (uvrD). Each curve was performed between two and six times with very similar results. (B) Mean number of cell divisions between 0 and 4 h of growth for strains bearing 34 lacO sites, determined from growth curves as represented in (A). Open bars: cells grown in glucose; filled bars: cells grown in arabinose. Strains are as described in (A). (C) Southern blot of a 1D agarose gel of PvuII digests of chromosomal DNA from strains bearing lacO34, all containing pPM306, grown in the presence of arabinose with and without 1 mM IPTG. Strains were BP41 (otherwise wild-type), BP43 (recA), BP45 (recB) and BP54 (recG). The entire apramycin resistance cassette was used to generate the radiolabelled probe. Note also that the amounts of chromosomal DNA detected by the probe varied between strains and likely reflected variation between strains in the number of chromosomes per cell. However, the amount of replication intermediate as a proportion of the total DNA signal did not vary greatly (5 to 8%).
Analysis of recA, recB, recF, recG and ruvABC recombination mutations also did not reveal any significant impact on growth in the lacO22 background (Figure 5). Recombination enzymes appeared therefore not to be required for tolerance of this repressor–operator block. However, deletion of recA, recB or recG had major impacts on growth in the lacO34 background upon repressor overexpression (Figure 6A and B). These defects in cell growth became apparent between 1 and 2 h after addition of arabinose (Figure 6A) and correlated with the rise of lac repressor levels as detected by western blotting (data not shown). These effects were abolished if IPTG was present in the growth medium in addition to arabinose (Figure 6A), demonstrating that the observed inhibition of growth was dependent on formation of lac repressor–operator complexes. In contrast, ruvABC or recF mutations had only minor effects on growth rate upon repressor overexpression in the lacO34 background (Figure 6A and B). Thus fork blockage at this artificial lac repressor-lacO34 array presented a significant impediment to normal cell proliferation that required the activities of RecA, RecB and RecG to tolerate, but that did not require RuvABC, RecF, Rep or UvrD.
The accumulation of replication intermediates at the lacO34 locus was analysed in recA, recB and recG strains upon induction of lac repressor overexpression. Slowly migrating DNA species were detected in all three mutant backgrounds but were absent when IPTG was present in the growth medium (Figure 6C). However, deletion of recA, recB or recG did not lead to a large increase in accumulation of replication intermediates as a proportion of the total DNA signal (Figure 6C, compare lanes 1, 3, 5 and 7) implying that, as expected, these gene products acted downstream of the initial blockage event.
DISCUSSION
Collisions between the replisome and protein–DNA complexes that have evolved to terminate replication lead to enhanced genome instability near the site of termination (10,29,32). The consequences of unavoidable encounters between the replisome and other protein–DNA complexes remain obscure. The data presented here suggest that collisions between the DNA replication apparatus and transcription factors may present barriers to replisome movement. The inhibition of replication observed in vitro occurred on both supercoiled (Figures 1 and 2) and linearized (Figure 3) template DNA indicating that blockage of forks was a direct consequence of collision between the replication apparatus and the repressor–operator complexes. Therefore displacement of transcription factor–DNA complexes by the replisome appears to present an energetic barrier to fork movement. The use of multiple repressor–operator complexes to hinder replication in vitro (Figure 1A) is clearly a situation never encountered in vivo. However, the additive nature of the blockage shown by 3 and 6 lacO sites bound by repressor (Figure 2) suggest that even single repressor–operator complexes will pose barriers to fork movement. Indeed, although we have failed to detect reproducible replication blockage with a single repressor–operator complex in vitro (data not shown), one lac repressor–operator complex does act as a recombination hotspot in vivo (11), a signature of replication fork problems.
Cells were capable of tolerating tandem protein–DNA complexes that were sufficient to block all replication in vitro (Figure 4), demonstrating that duplication of protein-bound DNA is enhanced in vivo, either directly or indirectly. Tolerance of 22 lacO sites within the chromosome did not depend on any of seven recombination enzyme genes tested (Figure 5A). Other mechanisms, such as enhancement of replicative helicase function by chromosomal topology (28), might therefore underpin this tolerance. However, mutations in recA, recB and recG did have major impacts on viable cell growth in the presence of 34 lacO sites (Figure 6A and B). Thus the strand exchange protein RecA, the helicase/exonuclease RecBCD and the branch migration helicase RecG all function to facilitate replication through these protein–DNA complexes, either directly or indirectly. In contrast, Rep helicase, previously implicated in removal of proteins ahead of replication forks (30,33), was not essential for this tolerance (Figure 6). The possible fork clearing role ascribed to UvrD (31) was also not required for tolerance of this block (Figure 6).
These data show that recombination enzymes somehow promote fork movement along protein-bound template, correlating with the enhanced recombination noted at a lac repressor–operator complex in vivo (11). However, given that any homologous recombination reaction between sister chromosomes will occur upstream of the protein–DNA array, it is difficult to envisage how such reactions would promote bypass of the block. Moreover, whilst recombination generates DNA structures onto which PriA can reload the replication machinery (34), any reassembled replisomes would still have to confront the protein–DNA barrier that caused the original fork to stall. However, given the stochastic nature of replication blockage seen in vitro (Figure 2) reinitiation of replication by recombination might provide another chance for the fork to translocate through the protein–DNA array. Such a mechanism would promote replication through barriers other than those presented by direct chemical damage to the template DNA. It would also explain how replication forks can eventually move through multiple Tus–ter complexes, given enough time (4).
If recombination via RecA-catalysed strand exchange is involved in tolerance of repressor–operator arrays then Holliday junctions would be expected to be formed. However, RuvABC was not critical for normal growth in the face of the lacO34 protein–DNA array (Figure 6) indicating that Holliday junction resolution by RuvABC-directed cleavage was not essential. The lack of any requirement for RuvABC in a lacO34 background also suggests that cleavage of Holliday junctions formed by unwinding of stalled replication forks (13) was not essential for tolerance of the lac arrays.
One proposed mechanism of replication fork processing that invokes RecA catalysis without a need to resolve Holliday junctions posits that RecA facilitates stabilization of reversed replication forks formed as a result of blockage by UV-induced DNA damage, thus protecting the forks from nucleolytic attack (17). However, this stabilization also requires RecF (35). The lack of any effect of a recF mutation on cell viability in the face of repressor–operator arrays (Figure 6) indicates that similar stabilization is unlikely to occur at forks blocked by protein–DNA complexes. Such differences likely reflect the diversity of fork repair mechanisms needed to face the different sources of replicative blocks that challenge all cells.
Replication forks blocked at 240 tet operators bound by tet repressor have no requirement for RecA or RecB to restart replication upon induced dissociation of the repressor–operator complexes (4). This lack of requirement appears to conflict with the need for both RecA and RecB to allow cell proliferation in the face of lacO34 bound by lac repressor (Figure 6). However, recombination may not be needed for resumption of replication once a block has been removed by addition of inducer. Stable association of the replisome with the template DNA at the site of blockage (4) would obviate the need for processing by recombination enzymes. In contrast, continued proliferation in the face of lacO34–repressor complexes may require active mechanisms to facilitate replication through these complexes. What these mechanisms are, and how they rely on RecA, RecBCD and RecG, remain to be elucidated. However, our data demonstrate that transcription factor–operator complexes can act to halt DNA replication, and that recombination enzymes are needed subsequently to endure such blocks. Indeed, occasional blockage of replisomes at transcription factor–DNA complexes in vivo may contribute to the decreased viability observed in recombination mutants (16,36). Our findings also imply that such collisions may be a factor in the coevolution of replication with other DNA metabolic processes.
ACKNOWLEDGEMENTS
The authors would like to thank Ken Marians, Bob Lloyd and Divijendra Sirigiri for plasmids, strains and proteins, Tom Meddows for advice on agarose gels and Tony Maxwell for DNA gyrase. This work was supported by the MRC and the Lister Institute of Preventive Medicine (P.M.) and by the Wellcome Trust (D.J.S.). Funding to pay the Open Access publication charges for this article was provided by the MRC.
REFERENCES
Lindahl, T. (1996) The Croonian Lecture, 1996: endogenous damage to DNA Philos. Trans. R. Soc. Lond. B. Biol. Sci, . 351, 1529–1538 .
Rothstein, R., Michel, B., Gangloff, S. (2000) Replication fork pausing and recombination or ‘gimme a break’ Genes Dev, . 14, 1–10 .
Mulcair, M.D., Schaeffer, P.M., Oakley, A.J., Cross, H.F., Neylon, C., Hill, T.M., Dixon, N.E. (2006) A molecular mousetrap determines polarity of termination of DNA replication in E.coli Cell, 125, 1309–1319 .
Possoz, C., Filipe, S.R., Grainge, I., Sherratt, D.J. (2006) Tracking of controlled Escherichia coli replication fork stalling and restart at repressor-bound DNA in vivo EMBO J, . 25, 2596–2604 .
Cox, M.M., Goodman, M.F., Kreuzer, K.N., Sherratt, D.J., Sandler, S.J., Marians, K.J. (2000) The importance of repairing stalled replication forks Nature, 404, 37–41 .
McGlynn, P. and Lloyd, R.G. (2002) Recombinational repair and restart of damaged replication forks Nature Rev. Mol. Cell. Biol, . 3, 859–870 .
Michel, B., Grompone, G., Flores, M.J., Bidnenko, V. (2004) Multiple pathways process stalled replication forks Proc. Natl Acad. Sci. USA, 101, 12783–12788 .
Kobayashi, T., Heck, D.J., Nomura, M., Horiuchi, T. (1998) Expansion and contraction of ribosomal DNA repeats in Saccharomyces cerevisiae: requirement of replication fork blocking (Fob1) protein and the role of RNA polymerase I Genes Dev, . 12, 3821–3830 .
Bierne, H., Ehrlich, S.D., Michel, B. (1991) The replication termination signal terB of the Escherichia coli chromosome is a deletion hot spot EMBO J, . 10, 2699–2705 .
Lambert, S., Watson, A., Sheedy, D.M., Martin, B., Carr, A.M. (2005) Gross chromosomal rearrangements and elevated recombination at an inducible site-specific replication fork barrier Cell, 121, 689–702 .
Vilette, D., Uzest, M., Ehrlich, S.D., Michel, B. (1992) DNA transcription and repressor binding affect deletion formation in Escherichia coli plasmids EMBO J, . 11, 3629–3634 .
Fan, Q., Xu, F., Petes, T.D. (1995) Meiosis-specific double-strand DNA breaks at the HIS4 recombination hot spot in the yeast Saccharomyces cerevisiae: control in cis and trans Mol. Cell. Biol, . 15, 1679–1688 .
Seigneur, M., Bidnenko, V., Ehrlich, S.D., Michel, B. (1998) RuvAB acts at arrested replication forks Cell, 95, 419–430 .
Flores, M.J., Ehrlich, S.D., Michel, B. (2002) Primosome assembly requirement for replication restart in the Escherichia coli holDG10 replication mutant Mol. Microbiol, . 44, 783–792 .
McGlynn, P. and Lloyd, R.G. (2000) Modulation of RNA polymerase by (p)ppGpp reveals a RecG-dependent mechanism for replication fork progression Cell, 101, 35–45 .
Gregg, A.V., McGlynn, P., Jaktaji, R.P., Lloyd, R.G. (2002) Direct rescue of stalled DNA replication forks via the combined action of PriA and RecG helicase activities Mol. Cell, 9, 241–251 .
Courcelle, J., Donaldson, J.R., Chow, K.H., Courcelle, C.T. (2003) DNA damage-induced replication fork regression and processing in Escherichia coli Science, 299, 1064–1067 .
Hiasa, H. and Marians, K.J. (1994) Primase couples leading- and lagging-strand DNA synthesis from oriC J. Biol. Chem, . 269, 6058–6063 .
Marians, K.J. (1995) X174-type primosomal proteins: purification and assay Meth. Enzymol, . 262, 507–521 .
Rosenberg, J.M., Khallai, O.B., Kopka, M.L., Dickerson, R.E., Riggs, A.D. (1977) Lac repressor purification without inactivation of DNA binding activity Nucleic Acids Res, . 4, 567–572 .
Hiasa, H. and Marians, K.J. (1994) Tus prevents overreplication of oriC plasmid DNA J. Biol. Chem, . 269, 26959–26968 .
Lau, I.F., Filipe, S.R., Soballe, B., Okstad, O.A., Barre, F.X., Sherratt, D.J. (2003) Spatial and temporal organization of replicating Escherichia coli chromosomes Mol. Microbiol, . 49, 731–743 .
Wilkinson, C.J., Hughes-Thomas, Z.A., Martin, C.J., Bohm, I., Mironenko, T., Deacon, M., Wheatcroft, M., Wirtz, G., Staunton, J., Leadlay, P.F. (2002) Increasing the efficiency of heterologous promoters in actinomycetes J. Mol. Microbiol. Biotechnol, . 4, 417–426 .
Marians, K.J., Hiasa, H., Kim, D.R., McHenry, C.S. (1998) Role of the core DNA polymerase III subunits at the replication fork. is the only subunit required for processive replication J. Biol. Chem, . 273, 2452–2457 .
Meddows, T.R., Savory, A.P., Lloyd, R.G. (2004) RecG helicase promotes DNA double-strand break repair Mol. Microbiol, . 52, 119–132 .
Yu, D., Ellis, H.M., Lee, E.C., Jenkins, N.A., Copeland, N.G., Court, D.L. (2000) An efficient recombination system for chromosome engineering in Escherichia coli Proc. Natl Acad. Sci. USA, 97, 5978–5983 .
Smelkova, N. and Marians, K.J. (2001) Timely release of both replication forks from oriC requires modulation of origin topology J. Biol. Chem, . 276, 39186–39191 .
Valjavec-Gratian, M., Henderson, T.A., Hill, T.M. (2005) Tus-mediated arrest of DNA replication in Escherichia coli is modulated by DNA supercoiling Mol. Microbiol, . 58, 758–773 .
Bidnenko, V., Ehrlich, S.D., Michel, B. (2002) Replication fork collapse at replication terminator sequences EMBO J, . 21, 3898–3907 .
Lane, H.E. and Denhardt, D.T. (1975) The rep mutation IV. Slower movement of replication forks in Escherichia coli rep strains J. Mol. Biol, . 97, 99–112 .
Flores, M.J., Sanchez, N., Michel, B. (2005) A fork-clearing role for UvrD Mol. Microbiol, . 57, 1664–1675 .
Horiuchi, T., Fujimura, Y., Nishitani, H., Kobayashi, T., Hidaka, M. (1994) The DNA replication fork blocked at the Ter site may be an entrance for the RecBCD enzyme into duplex DNA J. Bacteriol, . 176, 4656–4663 .
Yancey-Wrona, J.E. and Matson, S.W. (1992) Bound Lac repressor protein differentially inhibits the unwinding reactions catalyzed by DNA helicases Nucleic Acids Res, . 20, 6713–6721 .
Xu, L. and Marians, K.J. (2003) PriA mediates DNA replication pathway choice at recombination intermediates Mol. Cell, 11, 817–826 .
Courcelle, J., Carswell-Crumpton, C., Hanawalt, P.C. (1997) recF and recR are required for the resumption of replication at DNA replication forks in Escherichia coli Proc. Natl Acad. Sci. USA, 94, 3714–3719 .
Capaldo, F.N., Ramsey, G., Barbour, S.D. (1974) Analysis of the growth of recombination-deficient strains of Escherichia coli K-12 J. Bacteriol, . 118, 242–249 .
Hiasa, H. and Marians, K.J. (1994) Topoisomerase IV can support oriC DNA replication in vitro J. Biol. Chem, . 269, 16371–16375 .
Bachmann, B.J. (1996) Derivations and genotypes of some mutant derivatives of Escherichia coli K-12 In Neidhardt, F.C., Curtiss, R., III, Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., Reznikoff, W.S., Riley, M., Schaechter, M., Umbarger, H.E. (Eds.). Escherichia coli and Salmonella cellular and molecular biology,, 2nd edn Washington, DC ASM Press pp. 2460–2488 .
Guzman, L.M., Belin, D., Carson, M.J., Beckwith, J. (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter J. Bacteriol, . 177, 4121–4130 .
Moolenaar, G.F., Moorman, C., Goosen, N. (2000) Role of the Escherichia coli nucleotide excision repair proteins in DNA replication J. Bacteriol, . 182, 5706–5714 .(Bryony T. I. Payne, Ingeborg C. van Knip)
*To whom correspondence should be addressed. Tel: +44 0 1224 555183; Fax: +44 0 1224 555844; Email: p.mcglynn@abdn.ac.uk
ABSTRACT
All organisms require mechanisms that resuscitate replication forks when they break down, reflecting the complex intracellular environments within which DNA replication occurs. Here we show that as few as three lac repressor-operator complexes block Escherichia coli replication forks in vitro regardless of the topological state of the DNA. Blockage with tandem repressor-operator complexes was also observed in vivo, demonstrating that replisomes have a limited ability to translocate through high affinity protein–DNA complexes. However, cells could tolerate tandem repressor-bound operators within the chromosome that were sufficient to block all forks in vitro. This discrepancy between in vitro and in vivo observations was at least partly explained by the ability of RecA, RecBCD and RecG to abrogate the effects of repressor-operator complexes on cell viability. However, neither RuvABC nor RecF were needed for normal cell growth in the face of such complexes. Holliday junction resolution by RuvABC and facilitated loading of RecA by RecF were not therefore critical for tolerance of protein–DNA blocks. We conclude that there is a trade-off between efficient genome duplication and other aspects of DNA metabolism such as transcriptional control, and that recombination enzymes, either directly or indirectly, provide the means to tolerate such conflicts.
INTRODUCTION
Genome duplication must occur concurrently with many other essential metabolic processes. Chemical damage to the DNA template caused, in part, by the byproducts of normal metabolism is known to block replication (1), but the consequences of duplicating template DNA unavoidably bound by a wide-range of proteins is largely unknown. Both prokaryotic and eukaryotic replication forks can be paused by specialized protein–DNA complexes that have evolved to act as preprogrammed blocks to replication (2,3). However, the majority of proteins associated with chromosomal DNA function in DNA packaging and transcriptional control and their impact on replication fork movement is unclear, although large tandem arrays of transcription factor-operator complexes do provide potent blocks to Escherichia coli chromosomal duplication (4).
Regardless of the source of replication blockage, any block must be removed or bypassed for replication to be restarted. Recombination enzymes play pivotal roles in the repair and restart of blocked replication forks both in prokaryotes and eukaryotes, facilitating bypass of the original block and/or the formation of intermediates upon which the replication machinery can be reloaded (5–7). Indeed, the enhanced recombination noted at both preprogrammed replicative blocks (8–10) and at transcription factor–DNA complexes (11,12) point to the potential problems of duplicating protein-bound DNA.
Studies with E.coli have suggested that unwinding of the nascent DNA at blocked replication forks to form four-stranded Holliday junctions plays an important role in the repair of damaged forks (13). Such junctions may be targeted by RecBCD helicase/nuclease, allowing degradation of the spooled out DNA end and resetting of the fork for subsequent replisome reloading by PriA (14). However, it could also be cleaved by the Holliday junction-specific helicase/endonuclease RuvABC. This cleavage would result in chromosome breakage, release of a free double-stranded DNA (dsDNA) end and a requirement for RecA-catalysed strand exchange with the sister chromosome to form a D-loop onto which the replication apparatus can be reloaded (15,16). The situations in which these processing events might take place are unclear at present, and evidence has been presented that cleavage of reversed forks by RuvABC may be a pathological event occurring only in the absence of RecBCD (7). One problem is that how recombination enzymes act at blocked or otherwise damaged replication forks may depend on the context. Different types of replication block will result in different DNA structures at the block, dictating how the blocked fork will be processed by recombination enzymes, all of which display distinct polarities of action on DNA. Indeed, RecA in conjunction with RecFOR may stabilize replication forks blocked by ultraviolet (UV) light-induced DNA damage to inhibit subsequent processing, rather than to promote processing by catalysing strand exchange (17).
In this study, we have analysed the movement of replication forks through tandem lac repressor-operator complexes and have demonstrated that such high affinity protein–DNA complexes present additive barriers to fork movement in vitro. Lac repressor–operator complexes also inhibited fork movement in vivo. Blockage of fork movement by such barriers lead to a requirement for wild-type recA, recB and recG genes to maintain cell division. However, there was no such requirement for ruvABC, recF, rep or uvrD. We conclude that RecA, RecBCD and RecG act to facilitate cell proliferation in the face of protein–DNA complexes that hinder replisomes, but that these tolerance mechanism(s) do not rely on Holliday junction cleavage by RuvABC or the loading of RecA by RecFOR. Moreover, Rep and UvrD, proposed to act as accessory replicative helicases during chromosomal duplication, do not appear to underpin replication through protein–DNA roadblocks.
MATERIALS AND METHODS
DNAs and proteins
DNA PolIII*, ?, DnaB, DnaC, DnaG, HU and SSB were purified as described (18,19). LacI was purified as indicated (20). DnaA was purified with an N-terminal hexahistidine tag on a nickel NTA superflow column (Qiagen). DNA gyrase was supplied by A. Maxwell (John Innes Centre, Norwich, UK).
Replication plasmids were based on pBROTB535-I (21). pPM308 was constructed from pBROTB535-I by replacement of terB site 1 with EcoRV and XbaI sites. pPM437, pD506 and pIK02 were constructed by ligation of a Klenow-treated XhoI–SalI fragment from a derivative of pFX92 carrying 3, 6 and 22 tandem lacO repeats, respectively (22) into the EcoRV site of pPM308. pPM374, pPM378, pPM393 and pPM379 were constructed by cloning an apramycin resistance cassette (23) linked to 0, 3, 6 or 22 tandem lacO sites between the EcoRI and XbaI sites of pPM308. pPM306 was constructed by amplification of the lacI gene from E.coli MG1655 and cloning into pBAD/HisB (Invitrogen) as an NcoI–HindIII fragment.
In vitro replication assays
Standard oriC DNA replication reactions (15 μl) using DNA gyrase were performed as described (18) except incubation was at 37°C for 10 min. Electrophoretic analysis of replication products under denaturing conditions was performed in 0.7% alkaline agarose gels as described (18) using a 5'-labelled HindIII digest as a size marker. Fork blockage was quantified in standard replication assays by measuring the amount of replication product above the blocked 4 kb leading strand using reactions lacking LacI as the baseline.
Determination of replication fork movement along template DNA relaxed by cleavage with a restriction enzyme was based on a previous pulse-chase approach (24). All reaction components except the restriction enzyme and dCTP (LacI at 100 nM tetramers final concentration as indicated, the standard concentrations of DnaB, DnaC, SSB, ?, PolIII*, HU and DnaG plus DnaA at 300 nM) were assembled on ice in reaction volumes of 14 μl and incubated at 37°C for 3 min. A total of 1 μl of a mixture of SmaI (64 U/μl, Promega) and dCTP (0.3 pmol; 200 kBq/pmol) was then added to each reaction and incubation continued for 2 min at 37°C before reactions were stopped by addition of 1 μl of 0.5 M EDTA. This concentration of SmaI was sufficient to cleave all the template plasmid within 20 s. Reactions were then analysed under denaturing conditions as described above. The degree of fork blockage using this approach was estimated by comparing the 2 kb leading strand peak heights and setting the blockage as 1 for 22 lacO sites—blockage was taken to be complete at lacO22 based on the absence of leading strand product above 4 kb in the presence of LacI (Figure 3B, lane 9).
Strain construction
To integrate lac operators into argEC, the same 22 tandem lac operators used to construct oriC replication plasmids described above and a fragment containing 34 tandem lac operators were each cloned downstream of an apramycin resistance cassette (23) and then subcloned between the HindIII and EcoRV sites of a plasmid carrying argE and the 5' end of argC (25). The apramycin resistance gene was also cloned in the same manner but with no associated lacO sites. The apramycin cassette ± lacO together with flanking argEC sequences was then amplified and integrated into DY330 as described (26). These loci were then moved by P1 vir transduction from DY330 into PM203 (MG1655 ara714). The resultant strains were BP38 (lacO0), PM222 (lacO22) and BP41 (lacO34) (see Table 1). Successful integration was confirmed by arginine auxotrophy, Southern blotting, diagnostic PCR and DNA sequencing of PCR products. Derivatives of PM222 and BP41 with mutations in recombination and replication genes were made by P1 vir transduction (see Table 1).
Table 1 Strains used in this study
Growth experiments
Strains were transformed with pPM306 and selected on LB ampicillin plates overnight at 37°C. Single colonies were grown in 10 ml Luria–Bertani (LB) ampicillin in 50 ml conical flasks at 37°C with shaking to mid log phase and then inoculated into fresh 10 ml LB ampicillin, together with glucose, arabinose and isopropyl-?-D-thiogalactopyranoside (IPTG) as indicated, to an A650 of 0.01. Growth was continued at 37°C and samples removed at 0, 1, 2, 3 and 4 h after subculturing and viable cell counts performed on LB ampicillin IPTG plates at 37°C overnight. Concentrations of ampicillin, apramycin, glucose, arabinose and IPTG were 100 μg/ml, 50 μg/ml, 0.1%, 0.02% and 1 mM, respectively. Comparison of the number of cell divisions was performed using the increase in the number of viable cells between 0 and 4 h in the above growth experiments. Growth experiments were performed between two and eight times for each strain.
Electrophoretic analysis of in vivo replication intermediates
Strains were grown as described above for the growth experiments except that the culture volumes were increased to 40 ml in 250 ml conical flasks. Upon addition of arabinose plus or minus IPTG, growth was continued at 37°C with shaking until the A650 reached 0.4 (3 h after subculturing). Genomic DNA plugs were made and processed as described (25) except that digestion was with PvuII. Electrophoresis was performed through 0.8% agarose gels in 1x TBE for 6 h at 3.5 V/cm, the gels were Southern blotted and then probed using the complete apramycin resistance gene labelled by random priming with dCTP. Gene ruler 1 kb DNA ladder (Fermentas) was used as a marker.
RESULTS
Movement of replisomes in vitro is inhibited by LacI–lacO complexes
To assess the ability of the replication apparatus of E.coli to duplicate protein-bound DNA, the impact of lac repressor-operator complexes on replisome movement was analysed in vitro. DnaA-directed bidirectional replication of oriC-containing plasmids resulted in generation of 0.5 kb lagging strands plus 3 kb leading strands formed by convergence of the two forks in a region opposite oriC (Figure 1A, lane 1) (18). The sizes of leading and lagging strands generated on a plasmid lacking lacO sequences was not altered upon addition of lac repressor (LacI) (Figure 1A, lanes 1–4). Non-sequence-specific binding of LacI was not therefore a major impediment to replisome movement. However, addition of LacI to plasmid encoding 22 lacO sites located 2 and 4 kb from oriC lead to accumulation of leading strands of 2 and 4 kb in size (Figure 1A, lanes 5–8). Addition of the inducer IPTG to reactions with LacI on template DNA containing lacO22 precluded appearance of 2 and 4 kb leading strand products (Figure 1B). These data demonstrate that LacI–lacO complexes inhibit replisome movement, allowing convergence of the two forks at lacO22.
Figure 1 lac repressor–operator complexes block replisome movement in vitro. (A) Replication of plasmid templates with no lac operator sequences (pPM308, lanes 1–4) and with 22 tandem lacO sequences (pIK02, lanes 5–8) in the presence of 0, 5, 25 and 100 nM LacI tetramers. Products of replication were analysed on a denaturing agarose gel, with DNA size markers shown in kb. (B) Production of distinct 2 and 4 kb leading strand products on template pIK02 (lacO22) bound by LacI is abrogated by addition of IPTG. LacI (100 nM tetramers) and IPTG (1 mM) were present as indicated.
To determine the efficacy of replisome blockage by LacI–lacO complexes a series of plasmids were constructed bearing 0, 3, 6 and 22 lacO sites 2 and 4 kb away from oriC. Replication of each template in the presence of LacI resulted in accumulation of the 2 and 4 kb leading strands expected if forks stopped at the LacI–lacO complexes (Figure 2A). At 100 nM LacI tetramers, 24, 43 and 90% of replisomes were blocked at 3, 6 and 22 lacO sites, respectively (Figure 2B). The lacO3 and lacO6 data suggest that the effects of LacI–lacO complexes on replisome movement are additive with one lacO site blocking 5–10% of forks when bound by LacI.
Figure 2 Additive blockage of replisome movement by repressor–operator complexes. (A) Replication of plasmid templates bearing 0, 3, 6 and 22 lacO sites (pPM374, 378, 393 and 379) in the presence of 0, 5, 25 and 100 nM LacI tetramers was monitored by denaturing agarose gel electrophoresis. DNA size markers are shown in kb. (B) Replisome blockage as a function of LacI concentration. Open circles, lacO0; filled circles, lacO3; open squares, lacO6; filled triangles, lacO22.
Inhibition of replication by repressor–operator complexes occurs on relaxed as well as supercoiled template
The above data suggest that the energy input required to disrupt lac repressor–operator interactions presents a barrier to replication. However, the binding of repressor to tandem operators may have induced topological alterations within the plasmid template leading to replication fork stalling, possibly by inhibiting access of DNA gyrase to the template DNA ahead of the replisome. To determine whether altered topology was the primary determinant in replication blockage, the effect of lac repressor–operator complexes on fork movement along a topologically unconstrained DNA template was analysed.
Omission of a topoisomerase from replication reactions still allows initiation of replication at oriC in negatively supercoiled plasmid template (24). However, only one of the two forks initiates DNA synthesis and the fork that does initiate stalls after about 1 kb of DNA synthesis due to accumulation of positive torsional strain (27). This inhibition of fork movement can be relieved by addition of a restriction enzyme (SmaI) that cleaves near oriC thus relieving any topological constraint by linearization of the template (24) . Inclusion of radiolabelled dCTP at the time of SmaI addition allows fork progression to be monitored.
Figure 3 Replication blockage occurs on linearized as well as supercoiled template DNA. (A) In the absence of a topoisomerase, replication can initiate at oriC but is inhibited after about 1 kb of synthesis (i). Addition of SmaI results in cleavage of the DNA near to oriC (ii and iii) and allows one of the two forks to progress around the now-linearized template. Unimpeded replication would generate leading strands of 6 kb for both forks (v and vii) whereas 4 and 2 kb leading strands would result from blockage of the forks at lacO (iv and vi). (B) Replication of pPM308 (lacO0), pPM437 (lacO3), pD506 (lacO6) and pIK02 (lacO22) in the absence of a topoisomerase but with LacI and SmaI as indicated, monitored by denaturing agarose gel electrophoresis. DNA size markers are shown in kb. The position of early replication intermediate (ERI) (37), which accumulated in the absence of SmaI, is shown using pPM308 as template (lane 1).
To determine whether replication forks were blocked by repressor–operator complexes on linear template DNA, plasmids containing 0, 3, 6 or 22 lacO sites were employed. Forks moving rightwards would generate 4 and 6 kb leading strands upon addition of SmaI, depending on whether blockage occurred at lacO. Similarly, forks moving leftward would generate 2 or 6 kb leading strands .
2 and 4 kb leading strands were formed in the presence of LacI on template with 3, 6 or 22 lacO sites but were absent when LacI was absent, or when there were no lacO sites within the template (Figure 3B). Thus replication fork movement was inhibited by repressor–operator complexes on linearized template DNA. The reduction (lacO6) or absence (lacO22) of 6 kb leading strand products supported this conclusion (Figure 3B, lanes 7 and 9). The degree of blockage was 30 and 70% for 3 and 6 lacO sites, respectively, similar to the inhibition of replication estimated on covalently closed supercoiled template (Figure 2). We conclude that blockage of replication forks by repressor–operator complexes occurs to a similar extent regardless of the topological state of the DNA template.
LacI–lacO complexes present barriers to DNA replication in vivo
Two hundred and forty tandem lac operators bound by repressor on the E.coli chromosome provide a complete block to replication and cell proliferation in vivo (22). The in vitro data in Figures 1–3 suggest that far fewer repressor–operator complexes are needed for effective replication blockage. We therefore analysed the ability of replisomes to traverse 22 and 34 lacO sites integrated within the chromosome of E.coli (Figure 4A). To ensure that the lacO sites were saturated with repressor a plasmid (pPM306) encoding lacI under the control of an arabinose-inducible promoter was introduced into the strains. Upon addition of arabinose to the growth medium this plasmid generated an increase in levels of repressor from approximately 300 tetramers per cell to 10–20 000 tetramers per cell, with significant overexpression beginning 1 h after addition of arabinose (data not shown).
Figure 4 lac repressor–operator complexes present blocks to DNA replication in vivo. (A) Insertion of tandem lac operators within the chromosomal arg locus, linked to an apramycin resistance gene. The direction of replication fork movement is indicated with a dashed arrow. (B) Southern blot of a 1D agarose gel of PvuII digests from strains bearing 0 (BP38), 22 (PM222) and 34 (BP41) lacO repeats together with pPM306, a plasmid bearing an arabinose-inducible lacI gene, both with and without 1 mM IPTG. DNA size markers are shown in kb. The entire apramycin resistance gene was used to probe the blot. The positions of replication intermediates are marked with arrows. (C) Growth curves of strains BP38 (i), PM222 (ii) and BP41 (iii) containing pPM306. Open circles, medium supplemented with glucose (lacI expression repressed); filled circles, medium supplemented with arabinose (lacI expression induced). c.f.u.: colony forming units.
In the absence of lacO sites, no slowly migrating DNA species indicative of replication intermediates could be detected as judged by Southern blot analysis of chromosomal DNA digests (Figure 4B, lanes 1 and 2). Replication intermediates did accumulate at 22 and 34 lacO sites, with more intermediates accumulating at lacO34 as compared with lacO22 (Figure 4B, compare lanes 3 and 5). These slowly migrating species were absent when cells were grown in medium supplemented with IPTG (Figure 4B, lanes 4 and 6). The requirement for lacO sites to observe these slowly migrating species, and the effect of IPTG, suggest that these DNA species were replication intermediates caused by blockage of replisomes by repressor–operator complexes. Similar slowly migrating DNA species have been observed upon blockage of replication by Tus–ter complexes (28,29). However, growth of lacO22 and lacO34 strains did not display any major defect in the presence of elevated lac repressor expression (Figure 4C) although there was a small decrease in growth of the lacO34 strain in arabinose as compared with glucose (Figures 4C, 6A and B).
Given the blockage of DNA replication by tandem lac repressor–operator complexes observed in vitro (Figures 1 and 2) it was surprising there was little effect on growth upon induction of lac repressor expression. A trivial explanation for this discrepancy was that neither the lacO22 nor the lacO34 loci were saturated with repressor, even at the high concentrations of repressor achieved upon induction of overexpression. Levels of replication intermediates detectable by gel electrophoresis at the lacO22 and lacO34 loci were compared therefore in the presence of native and overexpressed levels of repressor. Replication intermediates detected at lacO34 increased from 1.1% (±0) as a proportion of the total DNA signal with wild-type levels of repressor to 5.5% (±1.5) upon induction of repressor overexpression. This increase demonstrates that tandem lacO sites were accessible to enhanced levels of repressor. In contrast, the levels of replication intermediates detected at lacO22 were similar in the presence of wild-type or overexpressed levels of repressor (0.7% and 1.2% , respectively). The lack of enhanced blockage at the lacO22 locus, in spite of the accessibility of tandem lacO sites to enhanced levels of repressor, indicates that lacO22 was saturated at wild-type repressor levels. Concerning lacO34, these data cannot unambiguously establish whether the locus was saturated with repressor but, given the 30- to 60-fold enhancement of repressor concentration obtained upon induction of overexpression, it is likely that lacO34 was also saturated in the presence of elevated levels of repressor.
Although the above electrophoretic analysis of replication intermediates did indicate saturation of lacO22 with repressor, this analysis also suggested that replication fork blockage in vivo (1% for lacO22) was much less efficient than that seen at lacO22 in vitro (100%). However, such direct comparisons are inappropriate since the ratios of replication intermediates to linear DNA fragments as observed by 1D and 2D gel techniques may not correlate directly with the actual levels of replication fork blockage—nicking or branch migration of replication intermediates to form linear DNA structures during their isolation likely result in underestimates of fork blockage (4).
Enzymes of recombination facilitate tolerance of repressor–operator complexes in vivo
If the lack of any major growth defects in cells harbouring 22 and 34 lacO sites bound by repressor could not be explained by sub-saturating levels of repressor, what other factors could explain this discrepancy? Fork movement through protein–DNA complexes may be facilitated by accessory factors, such as Rep helicase in vivo (30), factors which were absent from the in vitro replication reactions. Alternatively, direct or indirect processing of blocked replication forks, possibly by recombination enzymes (5–7), may have ameliorated the effects of protein–DNA blocks to replication in vivo. Therefore we examined the genetic requirements for growth of strains bearing either 22 or 34 lacO sites by analysing growth in the absence or presence of elevated levels of lac repressor.
Deletion of rep had no major effect on growth with either 22 or 34 lacO sites (Figures 5 and 6). Rep did not therefore have a critical role in promotion of fork movement through these repressor–operator complexes. We also tested whether UvrD, implicated in a fork clearing role (31), aided cell growth in the face of repressor–operator complexes. However, a uvrD mutation had no effect on the ability of cells to grow in the face of lacO34 (Figure 6A and B).
Figure 5 Recombination enzymes are not required in vivo for tolerance of 22 lacO sites bound by repressor. (A) Growth of strains harbouring 22 lacO sites and the indicated mutations, plus the lacI plasmid pPM306. Open circles: cells grown in glucose (no lac repressor overexpression); filled circles: cells grown in arabinose (elevated lac repressor). lacO22 strains were PM222 (otherwise wild-type), BP13 (recA), BP22 (recB), BP60 (recF), BP33 (recG), BP19 (ruvABC), BP16 (rep). Individual growth curves are shown but each curve was performed between two and eight times with all curves yielding similar results. C.f.u.: colony forming units. (B) Mean number of cell divisions between 0 and 4 h of growth for strains bearing 22 lacO sites, determined from growth curves as represented in (A). Open bars: cells grown in glucose; filled bars: cells grown in arabinose. Strains are as described in (A).
Figure 6 RecA, RecB and RecG are needed in vivo to tolerate 34 lacO sites bound by repressor. (A) Growth of strains harbouring 34 lacO sites and the indicated mutations, plus the lacI plasmid pPM306. Open circles: cells grown in glucose (no lac repressor overexpression); filled circles: cells grown in arabinose (elevated lac repressor); filled triangles: cells grown in arabinose plus IPTG. Strains were BP41 (otherwise wild-type), BP43 (recA), BP45 (recB), BP44 (recF), BP54 (recG), BP52 (ruvABC), BP47 (rep), BP55 (uvrD). Each curve was performed between two and six times with very similar results. (B) Mean number of cell divisions between 0 and 4 h of growth for strains bearing 34 lacO sites, determined from growth curves as represented in (A). Open bars: cells grown in glucose; filled bars: cells grown in arabinose. Strains are as described in (A). (C) Southern blot of a 1D agarose gel of PvuII digests of chromosomal DNA from strains bearing lacO34, all containing pPM306, grown in the presence of arabinose with and without 1 mM IPTG. Strains were BP41 (otherwise wild-type), BP43 (recA), BP45 (recB) and BP54 (recG). The entire apramycin resistance cassette was used to generate the radiolabelled probe. Note also that the amounts of chromosomal DNA detected by the probe varied between strains and likely reflected variation between strains in the number of chromosomes per cell. However, the amount of replication intermediate as a proportion of the total DNA signal did not vary greatly (5 to 8%).
Analysis of recA, recB, recF, recG and ruvABC recombination mutations also did not reveal any significant impact on growth in the lacO22 background (Figure 5). Recombination enzymes appeared therefore not to be required for tolerance of this repressor–operator block. However, deletion of recA, recB or recG had major impacts on growth in the lacO34 background upon repressor overexpression (Figure 6A and B). These defects in cell growth became apparent between 1 and 2 h after addition of arabinose (Figure 6A) and correlated with the rise of lac repressor levels as detected by western blotting (data not shown). These effects were abolished if IPTG was present in the growth medium in addition to arabinose (Figure 6A), demonstrating that the observed inhibition of growth was dependent on formation of lac repressor–operator complexes. In contrast, ruvABC or recF mutations had only minor effects on growth rate upon repressor overexpression in the lacO34 background (Figure 6A and B). Thus fork blockage at this artificial lac repressor-lacO34 array presented a significant impediment to normal cell proliferation that required the activities of RecA, RecB and RecG to tolerate, but that did not require RuvABC, RecF, Rep or UvrD.
The accumulation of replication intermediates at the lacO34 locus was analysed in recA, recB and recG strains upon induction of lac repressor overexpression. Slowly migrating DNA species were detected in all three mutant backgrounds but were absent when IPTG was present in the growth medium (Figure 6C). However, deletion of recA, recB or recG did not lead to a large increase in accumulation of replication intermediates as a proportion of the total DNA signal (Figure 6C, compare lanes 1, 3, 5 and 7) implying that, as expected, these gene products acted downstream of the initial blockage event.
DISCUSSION
Collisions between the replisome and protein–DNA complexes that have evolved to terminate replication lead to enhanced genome instability near the site of termination (10,29,32). The consequences of unavoidable encounters between the replisome and other protein–DNA complexes remain obscure. The data presented here suggest that collisions between the DNA replication apparatus and transcription factors may present barriers to replisome movement. The inhibition of replication observed in vitro occurred on both supercoiled (Figures 1 and 2) and linearized (Figure 3) template DNA indicating that blockage of forks was a direct consequence of collision between the replication apparatus and the repressor–operator complexes. Therefore displacement of transcription factor–DNA complexes by the replisome appears to present an energetic barrier to fork movement. The use of multiple repressor–operator complexes to hinder replication in vitro (Figure 1A) is clearly a situation never encountered in vivo. However, the additive nature of the blockage shown by 3 and 6 lacO sites bound by repressor (Figure 2) suggest that even single repressor–operator complexes will pose barriers to fork movement. Indeed, although we have failed to detect reproducible replication blockage with a single repressor–operator complex in vitro (data not shown), one lac repressor–operator complex does act as a recombination hotspot in vivo (11), a signature of replication fork problems.
Cells were capable of tolerating tandem protein–DNA complexes that were sufficient to block all replication in vitro (Figure 4), demonstrating that duplication of protein-bound DNA is enhanced in vivo, either directly or indirectly. Tolerance of 22 lacO sites within the chromosome did not depend on any of seven recombination enzyme genes tested (Figure 5A). Other mechanisms, such as enhancement of replicative helicase function by chromosomal topology (28), might therefore underpin this tolerance. However, mutations in recA, recB and recG did have major impacts on viable cell growth in the presence of 34 lacO sites (Figure 6A and B). Thus the strand exchange protein RecA, the helicase/exonuclease RecBCD and the branch migration helicase RecG all function to facilitate replication through these protein–DNA complexes, either directly or indirectly. In contrast, Rep helicase, previously implicated in removal of proteins ahead of replication forks (30,33), was not essential for this tolerance (Figure 6). The possible fork clearing role ascribed to UvrD (31) was also not required for tolerance of this block (Figure 6).
These data show that recombination enzymes somehow promote fork movement along protein-bound template, correlating with the enhanced recombination noted at a lac repressor–operator complex in vivo (11). However, given that any homologous recombination reaction between sister chromosomes will occur upstream of the protein–DNA array, it is difficult to envisage how such reactions would promote bypass of the block. Moreover, whilst recombination generates DNA structures onto which PriA can reload the replication machinery (34), any reassembled replisomes would still have to confront the protein–DNA barrier that caused the original fork to stall. However, given the stochastic nature of replication blockage seen in vitro (Figure 2) reinitiation of replication by recombination might provide another chance for the fork to translocate through the protein–DNA array. Such a mechanism would promote replication through barriers other than those presented by direct chemical damage to the template DNA. It would also explain how replication forks can eventually move through multiple Tus–ter complexes, given enough time (4).
If recombination via RecA-catalysed strand exchange is involved in tolerance of repressor–operator arrays then Holliday junctions would be expected to be formed. However, RuvABC was not critical for normal growth in the face of the lacO34 protein–DNA array (Figure 6) indicating that Holliday junction resolution by RuvABC-directed cleavage was not essential. The lack of any requirement for RuvABC in a lacO34 background also suggests that cleavage of Holliday junctions formed by unwinding of stalled replication forks (13) was not essential for tolerance of the lac arrays.
One proposed mechanism of replication fork processing that invokes RecA catalysis without a need to resolve Holliday junctions posits that RecA facilitates stabilization of reversed replication forks formed as a result of blockage by UV-induced DNA damage, thus protecting the forks from nucleolytic attack (17). However, this stabilization also requires RecF (35). The lack of any effect of a recF mutation on cell viability in the face of repressor–operator arrays (Figure 6) indicates that similar stabilization is unlikely to occur at forks blocked by protein–DNA complexes. Such differences likely reflect the diversity of fork repair mechanisms needed to face the different sources of replicative blocks that challenge all cells.
Replication forks blocked at 240 tet operators bound by tet repressor have no requirement for RecA or RecB to restart replication upon induced dissociation of the repressor–operator complexes (4). This lack of requirement appears to conflict with the need for both RecA and RecB to allow cell proliferation in the face of lacO34 bound by lac repressor (Figure 6). However, recombination may not be needed for resumption of replication once a block has been removed by addition of inducer. Stable association of the replisome with the template DNA at the site of blockage (4) would obviate the need for processing by recombination enzymes. In contrast, continued proliferation in the face of lacO34–repressor complexes may require active mechanisms to facilitate replication through these complexes. What these mechanisms are, and how they rely on RecA, RecBCD and RecG, remain to be elucidated. However, our data demonstrate that transcription factor–operator complexes can act to halt DNA replication, and that recombination enzymes are needed subsequently to endure such blocks. Indeed, occasional blockage of replisomes at transcription factor–DNA complexes in vivo may contribute to the decreased viability observed in recombination mutants (16,36). Our findings also imply that such collisions may be a factor in the coevolution of replication with other DNA metabolic processes.
ACKNOWLEDGEMENTS
The authors would like to thank Ken Marians, Bob Lloyd and Divijendra Sirigiri for plasmids, strains and proteins, Tom Meddows for advice on agarose gels and Tony Maxwell for DNA gyrase. This work was supported by the MRC and the Lister Institute of Preventive Medicine (P.M.) and by the Wellcome Trust (D.J.S.). Funding to pay the Open Access publication charges for this article was provided by the MRC.
REFERENCES
Lindahl, T. (1996) The Croonian Lecture, 1996: endogenous damage to DNA Philos. Trans. R. Soc. Lond. B. Biol. Sci, . 351, 1529–1538 .
Rothstein, R., Michel, B., Gangloff, S. (2000) Replication fork pausing and recombination or ‘gimme a break’ Genes Dev, . 14, 1–10 .
Mulcair, M.D., Schaeffer, P.M., Oakley, A.J., Cross, H.F., Neylon, C., Hill, T.M., Dixon, N.E. (2006) A molecular mousetrap determines polarity of termination of DNA replication in E.coli Cell, 125, 1309–1319 .
Possoz, C., Filipe, S.R., Grainge, I., Sherratt, D.J. (2006) Tracking of controlled Escherichia coli replication fork stalling and restart at repressor-bound DNA in vivo EMBO J, . 25, 2596–2604 .
Cox, M.M., Goodman, M.F., Kreuzer, K.N., Sherratt, D.J., Sandler, S.J., Marians, K.J. (2000) The importance of repairing stalled replication forks Nature, 404, 37–41 .
McGlynn, P. and Lloyd, R.G. (2002) Recombinational repair and restart of damaged replication forks Nature Rev. Mol. Cell. Biol, . 3, 859–870 .
Michel, B., Grompone, G., Flores, M.J., Bidnenko, V. (2004) Multiple pathways process stalled replication forks Proc. Natl Acad. Sci. USA, 101, 12783–12788 .
Kobayashi, T., Heck, D.J., Nomura, M., Horiuchi, T. (1998) Expansion and contraction of ribosomal DNA repeats in Saccharomyces cerevisiae: requirement of replication fork blocking (Fob1) protein and the role of RNA polymerase I Genes Dev, . 12, 3821–3830 .
Bierne, H., Ehrlich, S.D., Michel, B. (1991) The replication termination signal terB of the Escherichia coli chromosome is a deletion hot spot EMBO J, . 10, 2699–2705 .
Lambert, S., Watson, A., Sheedy, D.M., Martin, B., Carr, A.M. (2005) Gross chromosomal rearrangements and elevated recombination at an inducible site-specific replication fork barrier Cell, 121, 689–702 .
Vilette, D., Uzest, M., Ehrlich, S.D., Michel, B. (1992) DNA transcription and repressor binding affect deletion formation in Escherichia coli plasmids EMBO J, . 11, 3629–3634 .
Fan, Q., Xu, F., Petes, T.D. (1995) Meiosis-specific double-strand DNA breaks at the HIS4 recombination hot spot in the yeast Saccharomyces cerevisiae: control in cis and trans Mol. Cell. Biol, . 15, 1679–1688 .
Seigneur, M., Bidnenko, V., Ehrlich, S.D., Michel, B. (1998) RuvAB acts at arrested replication forks Cell, 95, 419–430 .
Flores, M.J., Ehrlich, S.D., Michel, B. (2002) Primosome assembly requirement for replication restart in the Escherichia coli holDG10 replication mutant Mol. Microbiol, . 44, 783–792 .
McGlynn, P. and Lloyd, R.G. (2000) Modulation of RNA polymerase by (p)ppGpp reveals a RecG-dependent mechanism for replication fork progression Cell, 101, 35–45 .
Gregg, A.V., McGlynn, P., Jaktaji, R.P., Lloyd, R.G. (2002) Direct rescue of stalled DNA replication forks via the combined action of PriA and RecG helicase activities Mol. Cell, 9, 241–251 .
Courcelle, J., Donaldson, J.R., Chow, K.H., Courcelle, C.T. (2003) DNA damage-induced replication fork regression and processing in Escherichia coli Science, 299, 1064–1067 .
Hiasa, H. and Marians, K.J. (1994) Primase couples leading- and lagging-strand DNA synthesis from oriC J. Biol. Chem, . 269, 6058–6063 .
Marians, K.J. (1995) X174-type primosomal proteins: purification and assay Meth. Enzymol, . 262, 507–521 .
Rosenberg, J.M., Khallai, O.B., Kopka, M.L., Dickerson, R.E., Riggs, A.D. (1977) Lac repressor purification without inactivation of DNA binding activity Nucleic Acids Res, . 4, 567–572 .
Hiasa, H. and Marians, K.J. (1994) Tus prevents overreplication of oriC plasmid DNA J. Biol. Chem, . 269, 26959–26968 .
Lau, I.F., Filipe, S.R., Soballe, B., Okstad, O.A., Barre, F.X., Sherratt, D.J. (2003) Spatial and temporal organization of replicating Escherichia coli chromosomes Mol. Microbiol, . 49, 731–743 .
Wilkinson, C.J., Hughes-Thomas, Z.A., Martin, C.J., Bohm, I., Mironenko, T., Deacon, M., Wheatcroft, M., Wirtz, G., Staunton, J., Leadlay, P.F. (2002) Increasing the efficiency of heterologous promoters in actinomycetes J. Mol. Microbiol. Biotechnol, . 4, 417–426 .
Marians, K.J., Hiasa, H., Kim, D.R., McHenry, C.S. (1998) Role of the core DNA polymerase III subunits at the replication fork. is the only subunit required for processive replication J. Biol. Chem, . 273, 2452–2457 .
Meddows, T.R., Savory, A.P., Lloyd, R.G. (2004) RecG helicase promotes DNA double-strand break repair Mol. Microbiol, . 52, 119–132 .
Yu, D., Ellis, H.M., Lee, E.C., Jenkins, N.A., Copeland, N.G., Court, D.L. (2000) An efficient recombination system for chromosome engineering in Escherichia coli Proc. Natl Acad. Sci. USA, 97, 5978–5983 .
Smelkova, N. and Marians, K.J. (2001) Timely release of both replication forks from oriC requires modulation of origin topology J. Biol. Chem, . 276, 39186–39191 .
Valjavec-Gratian, M., Henderson, T.A., Hill, T.M. (2005) Tus-mediated arrest of DNA replication in Escherichia coli is modulated by DNA supercoiling Mol. Microbiol, . 58, 758–773 .
Bidnenko, V., Ehrlich, S.D., Michel, B. (2002) Replication fork collapse at replication terminator sequences EMBO J, . 21, 3898–3907 .
Lane, H.E. and Denhardt, D.T. (1975) The rep mutation IV. Slower movement of replication forks in Escherichia coli rep strains J. Mol. Biol, . 97, 99–112 .
Flores, M.J., Sanchez, N., Michel, B. (2005) A fork-clearing role for UvrD Mol. Microbiol, . 57, 1664–1675 .
Horiuchi, T., Fujimura, Y., Nishitani, H., Kobayashi, T., Hidaka, M. (1994) The DNA replication fork blocked at the Ter site may be an entrance for the RecBCD enzyme into duplex DNA J. Bacteriol, . 176, 4656–4663 .
Yancey-Wrona, J.E. and Matson, S.W. (1992) Bound Lac repressor protein differentially inhibits the unwinding reactions catalyzed by DNA helicases Nucleic Acids Res, . 20, 6713–6721 .
Xu, L. and Marians, K.J. (2003) PriA mediates DNA replication pathway choice at recombination intermediates Mol. Cell, 11, 817–826 .
Courcelle, J., Carswell-Crumpton, C., Hanawalt, P.C. (1997) recF and recR are required for the resumption of replication at DNA replication forks in Escherichia coli Proc. Natl Acad. Sci. USA, 94, 3714–3719 .
Capaldo, F.N., Ramsey, G., Barbour, S.D. (1974) Analysis of the growth of recombination-deficient strains of Escherichia coli K-12 J. Bacteriol, . 118, 242–249 .
Hiasa, H. and Marians, K.J. (1994) Topoisomerase IV can support oriC DNA replication in vitro J. Biol. Chem, . 269, 16371–16375 .
Bachmann, B.J. (1996) Derivations and genotypes of some mutant derivatives of Escherichia coli K-12 In Neidhardt, F.C., Curtiss, R., III, Ingraham, J.L., Lin, E.C.C., Low, K.B., Magasanik, B., Reznikoff, W.S., Riley, M., Schaechter, M., Umbarger, H.E. (Eds.). Escherichia coli and Salmonella cellular and molecular biology,, 2nd edn Washington, DC ASM Press pp. 2460–2488 .
Guzman, L.M., Belin, D., Carson, M.J., Beckwith, J. (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter J. Bacteriol, . 177, 4121–4130 .
Moolenaar, G.F., Moorman, C., Goosen, N. (2000) Role of the Escherichia coli nucleotide excision repair proteins in DNA replication J. Bacteriol, . 182, 5706–5714 .(Bryony T. I. Payne, Ingeborg C. van Knip)