Multiple Genetic Pathways for Restarting DNA Replication Forks in Escherichia coli K-12

Corresponding author: Steven J. Sandler, 203 Morrill Science Center IVN, University of Massachusetts, Amherst, MA 01003., sandler{at}microbio.umass.edu (E-mail)

Communicating editor: P. L. FOSTER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In Escherichia coli, the primosome assembly proteins, PriA, PriB, PriC, DnaT, DnaC, DnaB, and DnaG, are thought to help to restart DNA replication forks at recombinational intermediates. Redundant functions between priB and priC and synthetic lethality between priA2::kan and rep3 mutations raise the possibility that there may be multiple pathways for restarting replication forks in vivo. Herein, it is shown that priA2::kan causes synthetic lethality when placed in combination with either rep::kan or priC303:kan. These determinations were made using a nonselective P1 transduction-based viability assay. Two different priA2::kan suppressors (both dnaC alleles) were tested for their ability to rescue the priA-priC and priA-rep double mutant lethality. Only dnaC809,820 (and not dnaC809) could rescue the lethality in each case. Additionally, it was shown that the absence of the 3'-5' helicase activity of both PriA and Rep is not the critical missing function that causes the synthetic lethality in the rep-priA double mutant. One model proposes that replication restart at recombinational intermediates occurs by both PriA-dependent and PriA-independent pathways. The PriA-dependent pathways require at least priA and priB or priC, and the PriA-independent pathway requires at least priC and rep. It is further hypothesized that the dnaC809 suppression of priA2::kan requires priC and rep, whereas dnaC809,820 suppression of priA2::kan does not.

THE primosome assembly proteins, PriA, PriB, DnaT, PriC, DnaC, DnaB, and DnaG, were defined in an in vitro DNA replication system based on the life cycle of the X174 ssDNA phage (reviewed in KORNBERG and BAKER 1992 ; MARIANS 1992 ). These proteins load sequentially onto the X174 chromosome to create a multiprotein complex that synthesizes an RNA oligonucleotide that primes DNA synthesis by Pol III holoenzyme. Briefly, PriA binds to the primosome assembly site (PAS) on the X174 chromosome. Then PriB, DnaT, and PriC bind sequentially to the PriA-DNA complex. PriB may stabilize PriA at PAS and facilitate the binding of DnaT (LIU et al. 1996 ). PriC is only partially required for the full assembly reaction. Omitting PriC lowers priming by about three- to fourfold (NG and MARIANS 1996A , NG and MARIANS 1996B ). In an ATP-dependent reaction, DnaC then loads DnaB into the complex. DnaC is not retained in the complex. This PriABC-DnaBT complex is competent to translocate around the chromosome. DnaG can then interact transiently with the protein complex to synthesize RNA primers. This system was thought to model RNA priming of Okazaki fragment DNA synthesis on the lagging strand at a replication fork.

Recent studies on the roles of the primosome assembly proteins in Escherichia coli have raised the possibility that their role in E. coli may be different, or in addition to, their roles predicted by the X174 system. This was first indicated when it was found that priA2::kan mutants were unexpectedly viable, but where UV^S, Rec^-, they could not plate Mu phage and had high basal levels of SOS expression (LEE and KORNBERG 1991 ; NURSE et al. 1991 ; MASAI et al. 1994 ; SANDLER et al. 1996 ; JONES and NAKAI 1997 ). The UV^S and Rec^- phenotypes suggested a role for the primosome assembly proteins in recombination and DNA repair. This idea was supported by the finding that in vitro PriA can initiate assembly of a replisome, competent for DNA synthesis, on a recombinational intermediate (LIU and MARIANS 1999 ; LIU et al. 1999 ).

The second result not predicted by the X174 system was seen when mutations in priB and priC (priBC) were first created and studied (SANDLER et al. 1999 ). It was expected that priB and priC should have different roles in the cell, yet mutations in these genes should lead to the same phenotype as priA2::kan mutants. This was not seen. Instead priB and priC had functional redundancy and their combined absence led to much poorer growth than with priA2::kan alone. This suggested that if assembly of replication forks at recombinational intermediates was important for viability, there may be multiple pathways by which this occurs, and these pathways may be dependent on priB and priC.

The third result not predicted by the X174 system was found when dnaC809, a suppressor of priA2::kan (SANDLER et al. 1996 ), did not fully suppress the absence of priBC. From dnaC809's role in priA2::kan suppression, it was hypothesized that dnaC809 created a bypass pathway that allowed the loading of DnaB at the proper recombinational intermediate in the absence of PriA. This was predicted based on data that PriA, PriB, DnaT, and PriC loaded onto the X174 chromosome in a linear fashion (the loading of each protein was dependent on the loading of the former; LIU et al. 1996 ; NG and MARIANS 1996A , NG and MARIANS 1996B ) and so the mutant DnaC protein should bypass the need for the entire PriABC-DnaT complex and load DnaB. Additional work showed that a second mutation in dnaC809, called dnaC820, creating the doubly mutated dnaC809,820 gene (SANDLER et al. 1999 ) could increase the suppression of priBC significantly so that the priBC dnaC809,820 strain behaved almost like the wild type. To explain this, it was hypothesized that dnaC809, selected in a priA2::kan priB⁺ priC⁺ strain, was partially priB and/or priC dependent and that the dnaC820 mutation relieved that dependency. dnaC809,820 suppressed the phenotypes of priA2::kan mutants as well as dnaC809 (SANDLER et al. 1999 ). Note that contrary to the in vivo situation, DnaC809 will suppress the absence of all four proteins in an in vitro reaction (LIU et al. 1999 ).

A current general model to explain the role of recombination in DNA replication suggests that replication forks that begin at oriC stall or collapse at some frequency while they transverse the chromosome. These forks are then repaired by recombinational processes. It has been proposed that this type of replication fork repair is nonmutagenic and part of a larger cellular housekeeping program that integrates recombinational repair of the collapsed forks, chromosome partitioning, and cell division (COX et al. 2000 ). CPR (for coordinated processing of damaged replication forks) has been suggested as an acronym for this combination of processes that resuscitates these broken replication forks (SANDLER and MARIANS 2000 ) and allows subsequent chromosome partitioning and cell division.

Since multiple mechanisms can cause the demise of the forks (stalling, arrest, breakage, or collapse), the way in which the DNA strands and attendant proteins at these forks may be associated could be different. Thus, the cell may need multiple mechanisms to deal with these different DNA-protein complexes. PriA has been shown to bind both SSB-coated ssDNA with a PAS (ALLEN and KORNBERG 1993 ; NG and MARIANS 1996A ) and naked D-loop structures in vitro (MCGLYNN et al. 1997 ; LIU and MARIANS 1999 ; NURSE et al. 1999 ). It is possible that these two may only be a subset of possible protein-DNA structures PriA may bind in vivo. Likewise, there may be other proteins giving the cell additional flexibility in recognizing multiple substrates and priming the restart process. Models for the recombinational repair of these "collapsed" forks are currently best understood in terms of postreplication repair (KUZMINOV 1999 ) and double-strand break repair (STAHL 1994 ). It is reasonable to assume however, that there may be recombination or an recA-independent mechanism to repair replication forks as well since recA mutants are viable. For this article, I refer to the many different possible DNA-protein structures that could be produced as a consequence of replication fork demise and repair collectively as "recombinational intermediates," knowing that other, yet to be identified structures may also exist.

If the process of restarting DNA replication is essential for growth, then why are priA mutants not dead or at least as poorly viable as priBC mutants? As mentioned above, one possible explanation is that there are PriA-independent pathways for restarting replication forks. If these pathways exist, then what proteins participate in them? One way to answer these questions is to find mutations that in combination with priA2::kan are lethal. This experimental approach supposes that there are two pathways, that each protein is in one of these pathways, and that the process is essential for viability. A possibility for at least one gene in the priA-independent pathway was rep. This was suggested by the preliminary result of Seigneur et al. that the rep3 and priA2::kan mutations are synthetically lethal (SEIGNEUR et al. 1998 ).

Rep is a 3'-5' helicase (LOHMAN and BJORNSON 1996 ). Although nonessential for DNA replication, one study proposed that its role in the cell is to optimize the speed and stability of the elongating replication fork (LANE and DENHARDT 1974 , LANE and DENHARDT 1975 ; COLASANTI and DENHARDT 1987 ). More recently, Michel and colleagues have suggested that the lack of the Rep helicase function causes replication forks to arrest more frequently (MICHEL et al. 1997 ). Additionally, it has been shown that rep mutations are lethal in combination with recB mutations (UZEST et al. 1995 ). Thus it is thought that the combination of rep and recB mutations leads to increased numbers of double-strand DNA breaks (DSBs) that are not repaired by recombination and hence the cells do not yield progeny (MICHEL et al. 1997 ). More recently, it was found that the recB-rep synthetic lethality can be rescued by additionally removing ruvAB (SEIGNEUR et al. 1998 ).

The experiments reported herein confirm that rep and priA cause synthetic lethality. One model to explain this posits that these two proteins function in different pathways for replication restart. dnaC809,820, but not dnaC809, can suppress the absence of priA and rep. It had been suggested previously that dnaC809 suppression of priA2::kan mutant phenotypes may be priB and/or priC dependent (SANDLER et al. 1999 ). Thus these mutations were tested for their synthetic lethality with priA2::kan. Only priC mutations created synthetic lethality. This suggested that rep and priC may be important for a PriA-independent pathway of restarting replication forks. Furthermore, evidence is offered that supports the idea that dnaC809 suppression of priA2::kan mutations requires priC and rep, whereas suppression of priA2::kan by dnaC809,820 is independent of these genes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Bacterial strains:
Bacterial strains:All bacterial strains used in this work are derivatives of E. coli K-12 and are described in Table 1. Due to priA1::kan strain's sensitivity to rich medium (MASAI and KOGOMA 1994), cultures for all experiments described, unless otherwise indicated, were grown in 56/2 minimal medium (WILLETTS et al. 1969 ) at 37°. The protocol for P1 transduction has been described elsewhere (WILLETTS et al. 1969 ).

Method for gene replacement:
A plasmid called pET-3c (K230R) was the source of the priA300 allele (ZAVITZ and MARIANS 1992 ). To replace this allele with the wild type on the chromosome, priA300 was first transferred to pKO3, a gene replacement vector (LINK et al. 1997 ). To do this, pET-3c (K230R) and pKO3 were restricted with BamHI and SalI. The mixtures of DNA fragments were separated by gel electrophoresis and the appropriate DNA fragments were isolated, mixed, treated with T4 DNA ligase, and used to transform competent cells. One pKO3 derivative containing priA300 was identified and called pSJS1299. This was then used to replace priA300 for priA⁺ on the chromosome by the method of LINK et al. 1997 . The presence of the priA300 allele was confirmed by detecting a BsiWI restriction endonuclease site associated with this mutation.

PCR analysis of transductants:
The PCR primers used for analysis of the priA locus (LEE et al. 1990 ; NURSE et al. 1990 ) are prSJS361 GGCATGACGGTTAAAGCTGG and prSJS362 GCGTAGCGCCTGCAACGCGG. Using a standard reaction mixture, thermocycling conditions used were initially 3 min denaturing at 94°. This was followed by 30 rounds of denaturing at 94° for 1 min, followed by hybridizing at 47° for 1 min, followed by extending the primer at 71° for 2.5 min. Finally a 10-min extension at 71° was used to finish the reaction. PCR products were analyzed by agarose gel electrophoresis. PriA⁺ and priA2::kan strains yield 0.45- and 1.8-kb DNA fragments, respectively.

The primers used for analysis of the priC locus (ZAVITZ et al. 1991 ) are prSJS283 (ATATTGAGTGTTGTCAGC) and prSJS284 (TCCTCCAGCAGCACAATC). The conditions used for the PCR reaction were identical to the ones above.

Testing temperature sensitivity of recBC strains:
The plating efficiencies of strains SS422 (rep::kan recBC^ts) and SS423 (priC303::kan recBC^ts) at 30° and 42° were determined by growing the strains in rich media to log phase. Aliquots of the cultures were serially diluted into 56/2 buffer, plated on Luria plates, placed at either 30° or 42°, and incubated for 24 hr. Colony-forming units per milliliter of culture were then determined and compared.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Seigneur et al. reported that the priA2::kan and rep3 mutations are synthetically lethal because they could not transduce a rep3 (sulA⁺ sulB⁺) mutant with priA2::kan by selecting directly for kanamycin resistance (SEIGNEUR et al. 1998 ; B. MICHEL, personal communication). To test the result of Seigneur et al. more critically, several changes were made to their experiment. First, a sulA mutation was incorporated into the recipient strain to suppress the filamentation and improve the viability of any priA2::kan transductant that might form (NURSE et al. 1991 ). Second, the btuB3191::Tn10 metB1 mutations were added to the recipient strain so that Met⁺ could be selected, rather than kanamycin resistance. This allows the nearby priA2::kan mutation to be recombined into the strain in a nonselective fashion. Thus the effect of the priA2::kan mutation on the strain's viability may be assessed independently of its ability to grow on media containing kanamycin. Whether priA2::kan was introduced into the strain was then determined by PCR analysis of the Met⁺ transductants. Last, a well-defined rep::kan (COLASANTI and DENHARDT 1987 ) mutation was tested instead of rep3.

Fig 1 shows a diagram of the cross used in this experiment. The donor used in many of the crosses reported here was JC19008 (priA2::kan metB⁺ btuB⁺). The recipient was JC19250 (priA⁺ metB1 btuB3191::Tn10) or an isogenic strain that varied at either the rep, priC, or dnaC locus. As stated above, Met⁺ was the selected phenotype and the condition of the priA locus was determined by PCR analysis. An example of the fragments produced by a PCR analysis of priA⁺ (0.5 kb) and priA2::kan (2.0 kb) alleles is shown in Fig 2A. The bands in lane 3 that are <2.0 kb in size are spurious and not seen in every PCR reaction. They may be breakdown products of the 2.0-kb band.

View larger version (10K): [in this window] [in a new window] [Download PPT slide]   Figure 1. This is a diagrammatic representation of the possible crossovers when a P1 lysate grown on JC19008 (metB+ btuB+ priA2::kan) is used to transduce a strain containing btuB3191::Tn10 metB1 to priA+. Since Met+ is selected, the selected crossovers are at 1 or 2. The nonselective crossovers are labeled 3 and 4. Therefore Met+ transductants will grow if crossovers occur in the following pairs: 1 and 2, 1 and 3, and 2 and 4. The genotypes of those strains will be, respectively: metB+ priA+ btuB3191::Tn10, metB+ priA2::kan btuB3191::Tn10, and metB+ priA+ btuB+. Double pairs of crossovers are also possible but would be expected to occur at a much lower frequency.

View larger version (31K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Agarose gel electrophoresis of PCR-generated DNA fragments with different DNA templates (i.e., different strains of bacteria) and PCR primers prSJS361 and prSJS362. (A) Lane 1: molecular weight markers (1-kb ladder purchased from Gibco). Lane 2: DM4000 (priA+). Lane 3: JC18983 (priA2::kan). (B) Lane 1: molecular weight markers (1-kb ladder purchased from Gibco). Lanes 2–17: independent transductants from the cross of JC19008 into SS182.

Theoretically, the cotransduction frequency between two markers can be predicted using the formula F = (1 - )³; where F is the cotransduction frequency, d is the distance between the markers in minutes, and L is the length of the bacterial DNA packaged in the transducing particle (which is ~2.1 min for P1; WU 1966 ; MASTERS 1996 ). By this formula, the predicted cotransduction frequency between metB (88.9 min) and priA (88.8 min; BERLYN 1998 ) is 86%. To determine experimentally the frequencies of cotransduction in a wild-type strain, a cross of JC19008 (priA2::kan metB⁺ btuB⁺) into JC19250 (btuB3191::Tn10 metB1) was performed. Table 2 shows that the measured cotransduction frequency between metB⁺ and priA23::kan was 44/56 or 78%. Thus the predicted and the observed results are in reasonable agreement.

To test if the combination of rep::kan and priA2::kan is synthetically lethal, JC19008 (priA2::kan metB⁺ btuB⁺) was transduced into SS182 (rep::kan priA⁺ metB1 btuB3191::Tn10). If rep::kan and priA2::kan are compatible mutations, one would expect to see a cotransduction frequency between metB⁺ and priA2::kan of ~80%. If the combination of rep::kan and priA2::kan was lethal, then the cotransduction frequency would be zero (0). Table 2 shows that of 32 Met⁺ transductants, none showed priA2::kan (the PCR analysis from 16 of the 32 transductants is shown in Fig 2B). Since the cotransduction frequency between metB and priA is not 100%, it is possible that the result attained could have occurred by random chance. To estimate this possibility, a chi-square distribution test was performed. It revealed that the chance of this result (no priA2::kan transductants) occurring by random chance is 3.4 x 10^-7. Therefore, it is likely that mutations in priA and rep are synthetically lethal.

It is also possible however that the reason why no priA2::kan transductants might have been seen in the above cross is that the rep::kan strain has an altered cotransduction frequency of nonselective markers. To test for this possibility, the allele at the btuB locus was tested for its cotransduction frequency with metB⁺. Using the same cross as above (Fig 1), but focusing on the btuB locus, one sees that the recipient is btuB3191::Tn10 and the donor is btuB⁺. Conversion of the btuB locus would produce a Tet^S phenotype. Table 2 shows that cotransduction frequencies between metB⁺ and btuB⁺ using JC19250 (rep⁺) and SS182 (rep::kan) as recipients are 55% and 53%, respectively. Hence, the rep::kan mutant inherits alleles by cotransduction at a frequency similar to wild type. These results support the conclusions drawn by SEIGNEUR et al. 1998 that the priA-rep double mutant is not viable.

dnaC809 does not rescue the synthetic lethality between rep::kan and priA2::kan:
dnaC809 is an extragenic suppressor of priA2::kan. It has been shown to suppress the high basal levels of SOS expression and the UV-sensitive, Rec^-, and poor viability phenotypes associated with priA2::kan mutants (SANDLER et al. 1996 ). To see if dnaC809 would suppress the synthetic lethality of the rep::kan and priA2::kan double mutant, transductions were done as described above except that the recipient strain (SS183) was additionally dnaC809. Table 2 shows that of 32 Met⁺ transductants, none cotransduced priA2::kan. Again the control experiments show that the transductional cross was technically successful and that btuB⁺ was cotransduced 75% of the time. Thus it is concluded that dnaC809 will not suppress the synthetic lethality produced by priA2::kan and rep::kan. To explain these results, it is hypothesized that rep function may be necessary for dnaC809 suppression of priA2::kan mutant phenotypes.

dnaC809,820 suppresses the rep::kan and priA2::kan synthetic lethality:
As mentioned in the Introduction, dnaC809 only partially suppresses the phenotypes caused by the absence of priBC. Since dnaC809,820 could more fully suppress the absence of priBC than dnaC809, it was hypothesized that dnaC809 suppression (of priA2::kan) was partially priB and/or priC dependent (it was selected in a priA2::kan priB⁺ priC⁺ strain). Therefore, it is possible that rep might participate with priB and/or priC in the suppression of priA2::kan. If so, then dnaC809,820 might suppress the synthetic lethality caused by rep::kan and priA2::kan. Table 2 shows that when SS407 (rep::kan dnaC809,820) was used as a recipient in the cross with JC19008, priA::kan was cotransduced with metB⁺ in 14/16 transductants tested. Therefore dnaC809,820 suppressed the lethality caused by the absence of rep and priA. Furthermore, this supports that dnaC809 suppression of priA2::kan mutant phenotypes is rep dependent and that dnaC820 relieves this dependency.

priC, but not priB mutations, are synthetically lethal with priA2::kan:
The above experiment suggests that priB and/or priC may participate with rep in suppression of priA2::kan mutant phenotypes. If so, then priB and/or priC mutations may be lethal when placed in combination with a priA2::kan mutation. To test this first for priB, priA2::kan was transduced from JC19008 to a strain harboring del(priB)302 (JC19266). Selecting for kanamycin resistance directly, priA2::kan del(priB)302 double mutants were found at high frequency. To ascertain if dnaC809 suppression of priA2::kan mutant phenotypes was priB dependent, the priA2::kan del(priB)302 dnaC809 triple mutant was constructed. This triple mutant had UV resistance similar to that of a priA2::kan dnaC809 strain (data not shown). It is concluded that priA and priB null mutations are not synthetically lethal and that dnaC809 suppression of priA mutants is not priB dependent.

To test if priA2::kan was lethal when placed with priC303::kan, SS72 (btuB3191::Tn10 metB1 priC303::kan) was crossed with JC19008 (Fig 1). Met⁺ transductants were selected. Table 2 shows that like a rep::kan mutant, the priC303::kan mutant did not allow cotransduction of priA2::kan with metB⁺ (0/32 events tested). SS72, however, did allow cotransduction of btuB⁺ (Tet^S) with metB⁺ at high frequency. Therefore, priA::kan and priC303::kan form a lethal combination, like rep::kan and priA2::kan.

If dnaC809 suppression of priA2::kan mutant phenotypes is priC dependent (as hypothesized above), then dnaC809 should not be able to rescue the synthetic lethality of a priA2::kan priC303::kan double mutant (like rep::kan and priA2::kan). Similarly, if dnaC809,820 is priC independent, then it should rescue the priA2::kan and priC303::kan synthetic lethality. Table 2 shows that the data support both of these ideas. Using JC19008 as a donor, the cross into the dnaC809 priC303::kan strain (JC19244) shows no Met⁺ priA2::kan transductants while the cross into the dnaC809,820 priC303::kan strain (SS94) shows 5/8 Met⁺ transductants are also priA2::kan. It is concluded that dnaC809 is not able to rescue the priA2::kan and priC303::kan synthetic lethality. One interpretation of this result is that dnaC809 suppression of priA2::kan is priC dependent. Likewise, the doubly mutated dnaC809,820 gene can rescue the priA2::kan and priC303::kan synthetic lethality. This is consistent with the idea that the dnaC820 mutation makes dnaC809 suppression priC independent.

priC and rep double mutants are viable:
The above results are consistent with the model that PriC and Rep both participate in the pathway used for dnaC809 suppression of priA2::kan mutants. If this is true, then it should be possible to create the priC-rep double mutant. Since priC303::kan and rep::kan are both kanamycin resistant, a tetracycline resistance gene (zbb-3055::Tn10) was inserted near priC303::kan to facilitate construction of the double mutant. This strain (JC19281) could then be used as a donor in a cross with the SS403 (rep::kan mutant). Tetracycline resistance was selected and the cotransduction of the nonselected priC303::kan allele was detected by PCR. A total of 2/16 Tet^R transductants were also priC303::kan (data not shown). It is concluded that priC and rep are not synthetically lethal. This result is consistent with, but does not prove, that priC and rep are in the same pathway.

priC recBC^ts double mutants are viable at 42°:
As stated above, it has been shown that rep function is required for viability in a recBC^ts strain at 42°. If rep and priC are in the same pathway for cell viability and if these genes have only one role in the cell, then one would simply predict that priC function would also be required at 42° in a recBC^ts mutant. This was tested and the priC303::kan recBC^ts mutant (SS423) has a plating efficiency of 1.0 (cfu at 42°/cfu at 30°) whereas the rep::kan recBC^ts (SS422) shows a plating efficiency of <10^-4. It is concluded that unlike rep, priC is not required for viability in a recBC^ts mutant at 42°. This unexpected result is discussed in detail below.

The absence of PriA and Rep 3'-5' helicase activities is not synthetically lethal:
Both PriA and Rep have 3'-5' helicase activities (ZAVITZ and MARIANS 1993 ; LOHMAN and BJORNSON 1996 ). It is possible that the 3'-5' helicase activity is the redundant activity that, when missing, is responsible for the synthetic lethality. To test this idea, priA300 was placed on the chromosome (see MATERIALS AND METHODS) and combined with rep::kan. priA300 was created by Zavitz and Marians (originally called K230R) and changes a lysine to an arginine at codon 230 in the highly conserved P-loop motif (ZAVITZ and MARIANS 1992 ). PriA300 has no detectable ATPase and helicase activities, but is competent for the in vitro assembly of the X174-type primosome (ZAVITZ and MARIANS 1992 ) and in vivo recombination (KOGOMA et al. 1996 ; SANDLER et al. 1996 ).

JJC213 containing rep::kan was used as a donor in a cross with the recipient SS97 (priA300). Kanamycin-resistant transductants were selected and were recovered at high frequency (data not shown). Thus the priA300 rep::kan double mutant is viable. It was noted however, that the priA300 rep::kan double mutants make smaller colonies relative to either the donor or recipient strains (Fig 3). Thus while the priA300 rep::kan double mutant is viable, the absence of the two helicase activities does have some effects on the growth characteristics of the strain. A fuller characterization of the priA300 strain and the priA300 rep::kan double mutant will be done elsewhere.

View larger version (106K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Comparison of colony size in strains containing priA300 and rep::kan. These strains were grown for 30 hr at 37° on minimal media. The strains are DM4000 (priA+ rep+), SS97 (priA300 rep+), SS403 (priA+ rep::kan), and SS405 (priA300 rep::kan).

It is concluded that although the 3'-5' helicase activity of either PriA or Rep is important for optimal growth of the strain, the absence of this activity is not the only determining factor for the synthetic lethality seen between priA2::kan and rep::kan.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

It is shown that mutations in priC and rep are synthetically lethal with mutations in priA and that some types of priA suppressors (dnaC809,820) are able to suppress both types of lethality and some (dnaC809) cannot. These results are summarized in Table 3. One way to interpret the data suggests that priA is in one pathway needed for viability and priC and rep are used in another. Since priA is thought to be responsible for the initial step in a pathway that loads the replisome at a recombinational intermediate, it is therefore possible that rep and priC are needed in another, priA-independent pathway, for restart of replication forks. These data also support the hypothesis that restarting DNA replication is essential for growth. It is noteworthy that the PriA-dependent and the PriA-independent pathways are not equal to one another as priA2::kan mutants have many deficient phenotypes and priC303::kan and rep::kan single mutants have few (see Introduction).

Another way to interpret the data would be with a "stress" model. This is where the synthetic lethality between the two genes is due not to the elimination of two distinct pathways for the same essential function, but rather to the absence of two distinct functions or roles in the cell, which causes a stress or imbalance leading to cell death. This model proposes that the absence of rep or priC puts a stress on the cell such that it needs to rely more heavily on the priA function and that the absence of both (rep-priA and priC-priA) leads to lethality. A more specific model has been proposed for the rep-priA double mutant lethality. This is where the absence of Rep leads to more frequent replication fork arrest and thus a greater need for PriA to restart these arrested forks (B. MICHEL, personal communication; SEIGNEUR et al. 1998 ). Both models have their strong and weak points and will be compared here. Other models are also possible.

Testing whether different mutant alleles of dnaC could suppress the synthetic lethality involving priA2::kan turned out to be unexpectedly illuminating for discerning differences between two dnaC mutants. While both dnaC809 and dnaC809,820 suppress the phenotypes caused by the absence of priA (SANDLER et al. 1996 , SANDLER et al. 1999 ), only dnaC809,820 could suppress the lethality caused by the simultaneous absence of priA-rep and priA-priC genes. This suggests that dnaC809 suppression of priA2::kan phenotypes is priC and rep dependent. Formally using the "pathways" model (Fig 4 and below), dnaC809 suppression of priA2::kan phenotypes could occur by elevating utilization and/or modification of the PriA-independent PriC-Rep pathway. Whether any other proteins are needed to facilitate the loading of DnaB in a priA2::kan dnaC809,820 strain and whether they function in the PriA-independent pathways remain to be determined. Alternately, the stress model could suggest that the specific activity, stability, and/or specificity of DnaC809 is sufficient to suppress the absence of priA in a rep⁺ priC⁺ strain. However, if either priC or rep is also missing, then the addition of the dnaC820 mutation is necessary to change the character of DnaC809 to compensate for this additional strain.

View larger version (20K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Genetic pathways for assembly of a replication fork at a recombinational intermediate. The starting point of the pathways is a set of demised replication forks that have been fixed by recombination or other processes yet to be described. The proteins listed operate on these substrates to reload the replisome. The product of the pathway is then a reloaded replisome that is able to initiate DNA replication. The replisome is defined as all proteins needed for a functioning replication fork with both leading and lagging strand synthesis. The proteins shown here may be only a subset of the total proteins needed. The thickness of lines indicates the robustness of the pathway. Vertical lines intersecting with horizontal lines indicate branch points in the pathways. Dotted lines indicate pathways available in the presence of the suppressor mutation indicated. The reader should refer to the text for explanations of which suppressors function in the absence of certain gene product(s) and the other proteins required for suppression. Four different entry points into the pathways are shown: two are PriA dependent (PriA-B and PriA-C) and two are PriA independent (PriC/Rep and DnaC809,820). For clarity, many possible paths have not been shown (i.e., PriA-B DnaT DnaC809). Potential general functions or roles for the groups of proteins in replication restart are listed above those proteins. Note that the presence of DnaT in the PriA-dependent pathway is preliminary and will be published elsewhere.

Pathways model for restart of replication forks:
Fig 4 shows a pathways model for the relationship between the various genes thought to be involved in reloading the replisome at a recombinational intermediate and how they might be formed into biochemical pathways. The order of the proteins in these pathways is based on the X174 model system (KORNBERG and BAKER 1992 ; MARIANS 1992 ; LIU et al. 1996 ; NG and MARIANS 1996A , NG and MARIANS 1996B ). In a broad sense, Fig 4 suggests that there are multiple pathways or entry points that funnel DNA-protein substrates into a central pathway to load DnaB. Once DnaB is loaded, it can serve as a focal point for the loading of the replisome. The reason for multiple pathways or entry points is that it may be necessary to recognize different DNA structures/proteins at repaired forks for restart. Alternately, redundancy of essential processes may be evolutionarily advantageous.

In Fig 4, the pathways model has broken down the restart process into three stages and possible functions associated with gene products have been suggested:

1. Recognition and modulation: These proteins are responsible for recognizing the correct DNA-protein substrate. This must occur only at the correct DNA structures and in the right orientation. Since this DNA structure is thought to resemble a recombinational intermediate, it may be necessary to remove recombination and/or other proteins (i.e., RecA, SSB) associated with this structure or unwind some duplex DNA to load DnaB.

2. Loading DnaB: This is accomplished by DnaC or one type of mutant dnaC protein shown.

3. Loading the replisome: DnaB's known interactions with two other important parts of the replisome [Pol III holoenzyme (KIM et al. 1996 ) and DnaG (primase) (TOUGU and MARIANS 1996 )] are hypothesized to serve as a focal point for helping to load or form these protein complexes into a replication fork.

Fig 4 also broadly divides the pathways into PriA-dependent and PriA-independent pathways. Suppressor pathways are included in the PriA-independent pathways since both dnaC809 and dnaC809,820 can fully suppress the absence of priA. Although not shown, both dnaC mutant proteins can function in the PriA-dependent pathway (no phenotypes have yet been discovered for strains that contain only dnaC809 or dnaC809,820). DnaT is needed biochemically in the formation of the X174-type primosome (NG and MARIANS 1996A ). Preliminary evidence indicates that a newly constructed dnaT null mutant is phenotypically similar to a priA null mutant. This suggests that dnaT is only necessary for the PriA-dependent pathway.

If one assumes that restarting replication forks at recombinational intermediates is essential for cell growth, then this diagram explains many observations. It explains why priA mutants are viable. In priA mutant cells, there is an alternate pathway (PriC-Rep) available to fulfill the same function. This pathway, however, is less proficient and hence the priA2::kan mutants display many deficient phenotypes (see above). If the priC or rep gene is additionally removed in a priA2::kan strain, then the alternate pathway is eliminated and the cell dies. The inviability of priBC double mutants is readily explained by elimination of the PriA-PriB, PriA-PriC, and PriC-Rep pathways. This diagram also suggests that priA-priB and priC-rep double mutant strains should be viable (and they are—see above). The model claims that those cells use the PriC-Rep and PriA-PriB pathways, respectively.

The suppression pathways are also incorporated into this diagram. dnaC809 suppression of priA2::kan mutations pathway requires PriC, Rep, and DnaC809. Thus if priC or rep are mutant, the DnaC809 pathway cannot function and cannot rescue the inviability of priA-priC and priA-rep double mutants. The DnaC809,820 pathway does not require Rep and/or PriC and can suppress the inviability of priA-rep and priA-priC strains.

This pathways model reflects a situation that is strikingly similar to that seen for the relationship between the RecBCD and RecF pathways of recombination in E. coli (CLARK and SANDLER 1994 ; KOWALCZYKOWSKI et al. 1994 ). While both pathways exist simultaneously in the cell, they prefer to operate on different types of substrates. In the absence of the one pathway (RecBCD), suppressor mutations (sbcA and sbcB; reviewed in KOWALCZYKOWSKI et al. 1994 ) can allow the other pathway to better utilize a broader set of DNA substrates. The PriC-Rep pathway may have a different substrate specificity from that of the PriA pathway and may be used instead of PriA to restart stalled, arrested, or collapsed replication forks. In the absence of PriA, the PriC-Rep pathway can function, albeit poorly, on the normal PriA pathway substrates. In strains carrying dnaC809 and dnaC809,820, the efficiency of this pathway on normal PriA pathway substrates increases.

The stress model:
As mentioned above, the stress model can also explain the viability of priA2::kan single mutants and the priA-priC and priA-rep lethality and that the differences in the stability, activity, or specificity of the dnaC mutant proteins may be responsible for the presence or absence of priA2::kan suppression. Other variations on this theme are also possible. For instance, DnaC809 may aid in creating the lethality in a stressed priA-rep double mutant. Thus dnaC820 may rescue dnaC809 by removing a lethal function without impairing the mutant DnaC's ability to load DnaB. One disadvantage of the stress model (vs. the pathways model) is that it does not make clear predictions of whether or not priC-rep and priA-priB double mutants should be viable.

Additional pathways, other stresses?
One complicating observation that has been incorporated into the pathways diagram is the functional redundancy between priB and priC. Mutants in these genes (individually) have essentially wild-type phenotypes. Presumably this allows restart at a greater variety of substrates. This redundancy is diagramed as a fork in the PriA-dependent pathway such that PriA can interact with either PriB or PriC. Currently there is only biochemical data to support the interaction between PriA and PriB (LIU et al. 1996 ; NG and MARIANS 1996A , NG and MARIANS 1996B ). It is proposed that each of these complexes then interacts with DnaT and DnaCB, etc. to restart the replication fork.

Another issue is that one must propose a second DnaC809-dependent suppressor pathway in addition to the one shown in Fig 4. Evidence of its existence is heralded by two observations. The first is that dnaC809 increases the viability and genetic stability of priBC mutants (SANDLER et al. 1999 ). The second is that priABC dnaC809 strains are viable (SANDLER et al. 1999 ). In these two strains, the PriC-Rep-dependent DnaC809 pathway cannot function, yet dnaC809 can offer viability to the strain. This second DnaC809-dependent pathway (not shown in Fig 4) appears to function only in the absence of both PriB and PriC. It is conceivable that DnaC809 could load DnaB onto the proper substrate in the absence of PriB and PriC. Some support for this reaction has been attained from in vitro reactions (LIU et al. 1999 ).

PriA's helicase activity:
The observation that the priA300-rep double mutant is viable suggests that the 3'-5' helicase function is not solely the redundant function missing in the rep-priA double mutant that causes the synthetic lethality. This conclusion assumes that the lack of measurable helicase activity in vitro (ZAVITZ and MARIANS 1992 ) is reflective of the absence of helicase activity in vivo. If so, then what function could be responsible for the lethality? A likely candidate is the ability to load the replisome at a recombinational intermediate. While primosome assembly activity has been readily demonstrated for PriA⁺ and PriA300 (ZAVITZ and MARIANS 1992 ), it has not yet been demonstrated for Rep. It is conceivable that Rep alone may not have this activity and may require PriC and/or other proteins.

Except for a recent article (JONES and NAKAI 1999 ; see below), there had been no previous demonstrated role for the PriA 3'-5' helicase activity. This was due to the fact that priA300 on a multicopy plasmid could completely suppress the absence of priA (ZAVITZ and MARIANS 1992 ; KOGOMA et al. 1996 ; SANDLER et al. 1996 ). It is plausible that this was seen because the strains used were rep⁺. The fact that the priA300-rep double mutant has smaller colony size than either of the single mutants suggests that the priA and rep 3'-5' helicase activities completely substitute for one another (in an otherwise wild-type strain) and that their presence is required for optimal growth. At least one condition where the rep helicase may not completely substitute for the PriA helicase is in the growth of Mu phage. Here, Jones and Nakai show that Mu replication is attenuated in a strain expressing priA300 from a plasmid (JONES and NAKAI 1999 ).

What might be the in vivo role of this 3'-5' helicase activity? One idea is that it modulates (Fig 4) the DNA structure at a recombinationally repaired replication fork to facilitate replication fork restart. The helicase activity could unwind duplex DNA adjacent to a gap or it could change the position of Holliday structures by branch migration to make ssDNA more accessible for DnaC to load DnaB.

Why is rep, but not priC, required for viability at 42° in the recBC^ts strain?
As discussed in the Introduction, Rep was thought to only function in DNA replication before replication fork collapse. This idea was originally based on studies by Denhardt and co-workers (LANE and DENHARDT 1974 , LANE and DENHARDT 1975 ; COLASANTI and DENHARDT 1987 ) and has been expanded by MICHEL et al. 1997 to be important in preventing frequent arrest of replication forks. The study here shows that priC and rep mutations cause similar patterns of phenotypes when placed in combination with priA and different dnaC mutants. Yet a different pattern is seen when in combination with recBC^ts at 42° (Table 3).

A strict interpretation of the pathways model states that if Rep and PriC participate together in a single function for all their roles in the cell, then PriC should also be lethal when in combination with recBC^ts at 42°. This is not seen. One way the pathways model could accommodate this data is by proposing that rep has two roles in the cell. Perhaps one function could be in replication fork stability (needed in the absence of recBC and priC independent) and the other after replication fork collapse (needed in the absence of priA and likely to be priC dependent). This proposal is not in conflict with any known data. However, there is no biochemical evidence to support this assumption.

The stress model suggests that the absence of rep, but not priC, exerts a pressure on the cell that cannot be compensated for in the absence of recBC. This suggests that rep and priC are needed for different roles in the cell. While this is a much simpler explanation than the pathways model, the stress model does not suggest a reason why rep and priC have the same pattern of phenotypes with respect to priA and alleles of dnaC (Table 3). Thus neither model is fully up to the challenge of explaining all the observations concerning priC. Both models can be rescued in this one regard, however, by the proposal that there may be a redundant function for priC that is used in priC's absence for viability in a recBC^ts at 42°. A possibility for this redundant function may be encoded by priB (SANDLER et al. 1999 ).

Does RecF function with Rep and PriC in the absence of PriA?
Last, it has been shown that the absence of recF function in the cell decreases the viability of priA2::kan mutants. The reason for this synthetic lethality may be different than that seen with rep or priC because dnaC809 rescues the recF-priA inviability completely (SANDLER 1996 ).

	ACKNOWLEDGMENTS

I thank Ken Marians and Benedicte Michel for sending strains, many useful discussions, and critically reading this manuscript; Jaime Coutu for technical assistance; and Abby Drake for assistance on the statistical analysis. S.J.S. was supported by Start-up funds from the University of Massachusetts and grant RPG-99-194-01-GMC from the American Cancer Society.

Manuscript received September 22, 1999; Accepted for publication February 10, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALLEN, G. C. and A. KORNBERG, 1993  Assembly of the primosome of DNA replication in Escherichia coli.. J. Biol. Chem. 268:19204-19209[Abstract/Free Full Text].

BERLYN, M. K. B., 1998  Linkage map of Escherichia coli K-12, edition 10: the traditional map. Microbiol. Mol. Biol. Rev. 62:814-984[Abstract/Free Full Text].

CLARK, A. J. and S. J. SANDLER, 1994  Homologous genetic recombination. Crit. Rev. Microbiol. 20:125-142[Medline].

COLASANTI, J. and D. T. DENHARDT, 1987  The Escherichia coli rep mutation. X. Consequences of increased and decreased Rep protein levels. Mol. Gen. Genet. 209:382-390[Medline].

COX, M. M., M. F. GOODMAN, K. N. KREUZER, D. J. SHERRATT, and S. J. SANDLER et al., 2000  The importance of repairing stalled replication forks. Nature 404:37-41[Medline].

JONES, J. M. and H. NAKAI, 1997  The X174-type primosome promotes replisome assembly at the site of recombination in bacteriophage Mu transposition. EMBO J. 16:6886-6895[Medline].

JONES, J. M. and H. NAKAI, 1999  Duplex opening by primosome protein PriA for replisome assembly on a recombination intermediate. J. Mol. Biol. 289:503-516[Medline].

KIM, S., G. DALLMANN, C. C. MCHENRY, and K. J. MARIANS, 1996  Coupling of a replicative polymerase and helicase: a t-DnaB interaction mediates rapid replication fork movement. Cell 84:643-650[Medline].

KOGOMA, T., G. W. CADWELL, K. G. BARNARD, and T. ASAI, 1996  The DNA replication priming protein, PriA, is required for homologous recombination and double-strand break repair. J. Bacteriol. 178:1258-1264[Abstract/Free Full Text].

KORNBERG, A., and T. BAKER, 1992 DNA Replication. W. H. Freeman, San Francisco.

KOWALCZYKOWSKI, S. C., D. A. DIXON, A. K. EGGLESTON, S. D. LAUDER, and W. M. REHRAUER, 1994  Biochemistry of homologous recombination in Escherichia coli.. Microbiol. Rev. 58:401-465[Abstract/Free Full Text].

KUZMINOV, A., 1999  Recombinational repair of DNA damage in Escherichia coli and bacteriophage . Microbiol. Mol. Biol. Rev. 63:751-813[Abstract/Free Full Text].

LANE, H. E. and D. T. DENHARDT, 1974  The rep mutation. III. Altered structure of the replicating Escherichia coli chromosome. J. Bacteriol. 120:805-814[Abstract/Free Full Text].

LANE, H. E. and D. T. DENHARDT, 1975  The rep mutation. IV. Slower movement of replication forks in Escherichia coli rep strains. J. Mol. Biol. 97:99-112[Medline].

LEE, E. H. and A. KORNBERG, 1991  Replication deficiencies in priA mutants of Escherichia coli lacking the primosomal replication n' protein. Proc. Natl. Acad. Sci. USA 88:3029-3032[Abstract/Free Full Text].

LEE, E. H., H. MASAI, G. C. ALLEN, and A. KORNBERG, 1990  The priA gene encoding the primosomal replicative n' protein of Escherchia coli.. Proc. Natl. Acad. Sci. USA 87:4620-4624[Abstract/Free Full Text].

LINK, A. J., D. PHILLIPS, and G. M. CHURCH, 1997  Methods for generating precise deletions and insertions in the genome of wild-type Escherichia coli: application to open reading frame characterization. J. Bacteriol. 179:6228-6237[Abstract/Free Full Text].

LIU, J. and K. J. MARIANS, 1999  PriA-directed assembly of a primosome on D loop DNA. J. Biol. Chem. 274:25033-25041[Abstract/Free Full Text].

LIU, J., P. NURSE, and K. J. MARIANS, 1996  The ordered assembly of the X174-type primosome. III. PriB facilitates complex formation between PriA and DnaT. J. Biol. Chem. 271:15656-15661[Abstract/Free Full Text].

LIU, J., L. XU, S. J. SANDLER, and K. J. MARIANS, 1999  Replication fork assembly at recombination intermediates is required for bacterial growth. Proc. Natl. Acad. Sci. USA 96:3552-3555[Abstract/Free Full Text].

LOHMAN, T. M. and K. P. BJORNSON, 1996  Mechanisms of helicase-catalyzed DNA unwinding. Annu. Rev. Biochem. 65:169-214[Medline].

MARIANS, K. J., 1992  Prokaryotic DNA replication. Annu. Rev. Biochem. 61:673-719[Medline].

MASAI, H., T. ASAI, Y. KUBUTA, K.-I. ARAI, and T. KOGOMA, 1994  Escherichia coli PriA protein is essential for inducible and constitutive stable DNA replication. EMBO J. 13:5338-5345[Medline].

MASTERS, M., 1996 Generalized transduction, pp. 2421–2441 in Escherichia coli and Salmonella Cellular and Molecular Biology, edited by F. C. NEIDHARDT. American Society of Microbiology, Washington, DC.

MCGLYNN, P., A. AL-DEIB, J. LIU, K. MARIANS, and R. LLOYD, 1997  The DNA replication protein PriA and the recombination protein RecG bind D-loops. J. Mol. Biol. 270:212-221[Medline].

MICHEL, B., S. D. EHRLICH, and M. UZEST, 1997  DNA double-strand breaks caused by replication arrest. EMBO J. 16:430-438[Medline].

NG, J. Y. and K. J. MARIANS, 1996a  The ordered assembly of the X174-type primosome. I. Isolation and identification of intermediate protein-DNA complexes. J. Biol. Chem. 271:15642-15648[Abstract/Free Full Text].

NG, J. Y. and K. J. MARIANS, 1996b  The ordered assembly of the X174-type primosome. II. Preservation of primosome composition from assembly through replication. J. Biol. Chem. 271:15649-15655[Abstract/Free Full Text].

NURSE, P., R. J. DIGATE, K. H. ZAVITZ, and K. J. MARIANS, 1990  Molecular cloning and DNA sequence analysis of Escherichia coli priA, the gene encoding the primosomal protein replication factor Y. Proc. Natl. Acad. Sci. USA 87:4615-4619[Abstract/Free Full Text].

NURSE, P., K. H. ZAVITZ, and K. J. MARIANS, 1991  Inactivation of the Escherichia coli PriA DNA replication protein induces the SOS response. J. Bacteriol. 173:6686-6693[Abstract/Free Full Text].

NURSE, P., J. LIU, and K. J. MARIANS, 1999  Two modes of PriA binding to DNA. J. Biol. Chem. 274:25026-25032[Abstract/Free Full Text].

SANDLER, S. J., 1996  Overlapping functions for recF and priA in cell viability and UV-inducible SOS expression are distinguished by dnaC809 in E. coli K-12. Mol. Microbiol. 19:871-880[Medline].

SANDLER, S. J. and K. J. MARIANS, 2000  Role of PriA replication fork reactivation in Escherichia coli.. J. Bacteriol. 182:9-13[Free Full Text].

SANDLER, S. J., H. S. SAMRA, and A. J. CLARK, 1996  Differential suppression of priA2::kan phenotypes in Escherichia coli K-12 by mutations in priA, lexA, and dnaC.. Genetics 143:5-13[Abstract].

SANDLER, S. J., K. J. MARIANS, K. H. ZAVITZ, J. COUTU, and M. A. PARENT et al., 1999  DnaC mutations suppress defects in DNA replication and recombination associated functions in priB and priC double mutants in E. coli K-12. Mol. Microbiol. 34:91-101[Medline].

SEIGNEUR, M., V. BIDNENKO, S. D. EHRLICH, and B. MICHEL, 1998  RuvAB acts at arrested replication forks. Cell 95:419-430[Medline].

SINGER, M., T. A. BAKER, G. SCHNITZLER, S. M. DEISCHEL, and M. GOEL et al., 1989  A collection of strains containing genetically linked alternating antibiotic resistance elements for genetic mapping of Escherichia coli.. Microbiol. Rev. 53:1-24[Abstract/Free Full Text].

STAHL, F. W., 1994  The Holliday junction on its thirtieth anniversary. Genetics 138:241-246[Medline].

TOUGU, K. and K. MARIANS, 1996  The extreme C terminus of primase is required for interaction with DnaB at the replication fork. J. Biol. Chem. 271:21391-21397[Abstract/Free Full Text].

UZEST, M., S. D. EHRLICH, and B. MICHEL, 1995  Lethality of rep recB and rep recC double mutants of Escherichia coli.. Mol. Microbiol. 17:1177-1188[Medline].

WILLETTS, N. S., A. J. CLARK, and B. LOW, 1969  Genetic location of certain mutations conferring recombination deficiency in Escherichia coli. J. Bacteriol. 97:244-249[Abstract/Free Full Text].

WU, T. T., 1966  A model for three point analysis of random general transduction. Genetics 54:405-410[Free Full Text].

ZAVITZ, K. H. and K. J. MARIANS, 1992  ATPase-deficient mutants of the Escherichia coli DNA replication protein PriA are capable of catalyzing the assembly of active primosomes. J. Biol. Chem. 267:6933-6940[Abstract/Free Full Text].

ZAVITZ, K. H. and K. J. MARIANS, 1993  Helicase-deficient cysteine to glycine substitution mutants of Escherichia coli replication protein PriA retain single-stranded DNA-dependent ATPase activity. J. Biol. Chem. 268:4337-4346[Abstract/Free Full Text].

ZAVITZ, K. H., R. J. DIGATE, and K. J. MARIANS, 1991  The PriB and PriC replication proteins of Escherichia coli.. J. Biol. Chem. 266:13988-13995[Abstract/Free Full Text].

This article has been cited by other articles:

				B. Slominski, J. Calkiewicz, P. Golec, G. Wegrzyn, and B. Wrobel Plasmids derived from Gifsy-1/Gifsy-2, lambdoid prophages contributing to the virulence of Salmonella enterica serovar Typhimurium: implications for the evolution of replication initiation proteins of lambdoid phages and enterobacteria Microbiology, June 1, 2007; 153(6): 1884 - 1896. [Abstract] [Full Text] [PDF]

				E. V. Mirkin and S. M. Mirkin Replication Fork Stalling at Natural Impediments Microbiol. Mol. Biol. Rev., March 1, 2007; 71(1): 13 - 35. [Abstract] [Full Text] [PDF]

				I. Ivancic-Bace, I. Vlasic, G. Cogelja-Cajo, K. Brcic-Kostic, and E. Salaj-Smic Roles of PriA Protein and Double-Strand DNA Break Repair Functions in UV-Induced Restriction Alleviation in Escherichia coli Genetics, December 1, 2006; 174(4): 2137 - 2149. [Abstract] [Full Text] [PDF]

				E. A. Stohl and H. S. Seifert Neisseria gonorrhoeae DNA Recombination and Repair Enzymes Protect against Oxidative Damage Caused by Hydrogen Peroxide J. Bacteriol., November 1, 2006; 188(21): 7645 - 7651. [Abstract] [Full Text] [PDF]

C.-Y. Huang, C.-H. Hsu, Y.-J. Sun, H.-N. Wu, and C.-D. Hsiao Complexed crystal structure of replication restart primosome protein PriB reveals a novel single-stranded DNA-binding mode Nucleic Acids Res.,