Gene conversion in the Escherichia coli RecF pathway: a successive half crossing-over model.

Summary Gene conversion - apparently non-reciprocal transfer of sequence information between homologous DNA sequences - has been reported in various organisms. Frequent association of gene conversion with reciprocal exchange (crossing-over) of the flanking sequences in meiosis has formed the basis of the current view that gene conversion reflects events at the site of interaction during homologous recombination. In order to analyze mechanisms of gene conversion and homologous recombination in an Escherichia coli strain with an active RecF pathway (recBC sbcBC), we first established in cells of this strain a plasmid carrying two mutant neo genes, each deleted for a different gene segment, in inverted orientation. We then selected kanamycin-resistant plasmids that had reconstituted an intact neo ⁺ gene by homologous recombination. We found that all the neo ⁺ plasmids from these clones belonged to the gene-conversion type in the sense that they carried one neo ⁺ gene and retained one of the mutant neo genes. This apparent gene conversion was, however, only very rarely accompanied by apparent crossing-over of the flanking sequences. This is in contrast to the case in a rec ⁺ strain. or in a strain with an active RecE pathway (recBC sbcA). Our further analyses, especially comparisons with apparent gene conversion in the rec ⁺ strain, led us to propose a mechanism for this biased gene conversion. This successive half crossing-over model proposes that the elementary recombinational process is half crossing;-over in the sense that it generates only one recombinant DNA duplex molecule, and leaves one or two free end(s), out of two parental DNA duplexes. The resulting free end is, the model assumes, recombinogenic and frequently engages in a second round of half crossing-over with the recombinant duplex. The products resulting from such interaction involving two molecules of the plasmid would be classified as belonging to the gene-conversion type without crossing-over. We constructed a dimeric molecule that mimics the intermediate form hypothesized in this model and introduced it into cells. Biased gene conversion products were obtained in this reconstruction experiment. The half crossing-over mechanism can also explain formation of huge linear multimers of bacterial plasmids, the nature of transcribable recombination products in bacterial conjugation, chromosomal gene conversion not accompanied by flanking exchange (like that in yeast mating-type switching), and antigenic variation in microorganisms.     

Repair of heteroduplex DNA molecules with multibase loops in Escherichia coli.

The fate of heteroduplex molecules containing 5-, 7-, 9-, 192-, 410-, and 514-base loops after transformation of wild-type and various mutant strains of Escherichia coli has been examined. No evidence for repair was obtained for the wild type or for strains with mutations in the following genes: mutS, recA, recBC sbcBC, recD, recF, recJ, recN, recO, recR, recBC sbcBC recF uvrA, recG ruvC, ruvB, lexA3, lexA51, uvrA, nfo xth nth, polA(Ts), or pcnB. These results rule out the involvement of the SOS system and most known recombination and repair pathways. Repair of heteroduplex molecules containing 410- and 514-base loops was observed when a 1-base deletion-insertion mismatch was present nearby. The repair of both the mismatch and the loops was directed by the state of dam methylation of the DNA chains and was dependent on the product of the mutS gene. A high efficiency of repair (95%) was found even when the mismatch and the loops were 1,448 nucleotides apart. We conclude that multibase loops in DNA can be removed only as a consequence of corepair by dam-directed mismatch repair.     

Gene conversion: point-mutation heterozygosities lower heteroduplex formation.

The effects of heterozygosity on meiotic gene conversion characteristics have been studied in the fungus Ascobolus immersus. The non-Mendelian segregation patterns of seven white ascospore mutants of the b2 gene were established in the presence or the absence of additional neighbouring allelic mutations. These correspond to nine different double mutants with wild-type or pseudo-wild-type phenotypes, constituted by two +1, -1 frameshift mutations of complementary phases. When heterozygous, these double point mutations decrease, by an average of one third, the gene conversion frequencies of the mutants located on their right, toward the low conversion end of the gene. The decrease corresponds either to a reduction in all classes of non-Mendelian segregation (6:2, 5:3 and aberrant 4:4 asci) or to a reduction restricted to the single class of aberrant 4:4 asci. These modifications are explained by changes in hybrid DNA parameter values: frequencies of formation and modalities of distribution (asymmetric versus symmetric ratio). Besides the nature of the non-homology, point mutation versus deletion, which leads to quantitative differential effects, the region where the non-homology is located within the gene also appears to play an important role.     

Protein Z is not an endonuclease of the recF recombination pathway in Escherichia coli K12

A 74 kD protein was extracted from Escherichia coli cells and purified under the physiological conditions. The protein is able to catalyze the reactions of endonucleolytic degradation of plasmid DNA. The genetic determinant coding for the 74 KD protein synthesis has been localized between 17 and 27 min on Escherichia coli chromosomal map. The endonuclease previously described as a recF gene dependent "protein Z" (Krivonogov S. V., Novitskaja V. A. Mol. Gen. Genet., 1982, v, 187, p. 302) is shown to be independent of the integrity of Escherichia coli recF gene.     

Recombination following transformation of Escherichia coli by heteroduplex plasmid DNA molecules

Circular heteroduplex DNA molecules introduced into Escherichia coli-competent cells are converted to new recombinant plasmids as a result of enzymatic actions in vivo. A pair of plasmids with partial sequence homology were each linearized at a different position with restriction enzymes, and the termini were made flush with the single-strand-specific S1 nuclease. Duplex molecules were then formed by melting and annealing these plasmid DNAs together. In contrast to linear homoduplex molecules, heteroduplexes circularize and therefore transform E. coli efficiently. Unique DNA sequences on each of the parental strands in the transforming heteroduplexes can be selectively incorporated or deleted as a result of in vivo enzymatic activities in transformed cells. This method permits the generation of new recombinant sequences in vivo without relying solely on the presence of convenient restriction sites for manipulation of DNA fragments in vitro.     

RecOR suppression of recF mutant phenotypes in Escherichia coli K-12.

The recF, recO, and recR genes form the recFOR epistasis group for DNA repair. recF mutants are sensitive to UV irradiation and fail to properly induce the SOS response. Using plasmid derivatives that overexpress combinations of the recO+ and recR+ genes, we tested the hypothesis that high-level expression of recO+ and recR+ (recOR) in vivo will indirectly suppress the recF mutant phenotypes mentioned above. We found that overexpression of just recR+ from the plasmid will partially suppress both phenotypes. Expression of the chromosomal recO+ gene is essential for the recR+ suppression. Hence we call this RecOR suppression of recF mutant phenotypes. RecOR suppression of SOS induction is more efficient with recO+ expression from a plasmid than with recO+ expression from the chromosome. This is not true for RecOR suppression of UV sensitivity (the two are equal). Comparison of RecOR suppression with the suppression caused by recA801 and recA803 shows that RecOR suppression of UV sensitivity is more effective than recA803 suppression and that RecOR suppression of UV sensitivity, like recA801 suppression, requires recJ+. We present a model that explains the data and proposes a function for the recFOR epistasis group in the induction of the SOS response and recombinational DNA repair.     

recB recC-dependent processing of heteroduplex DNA stimulates recombination of an adjacent gene in Escherichia coli.

The effect of DNA mismatched repair on the genetic recombination of a gene adjacent to the mismatch site (MS) was tested by using four mismatch configurations. An MS was constructed in a well-characterized plasmid recombination substrate, and recombination with a resident compatible plasmid was measured after transformation of the mismatched plasmid into Escherichia coli. The mismatched plasmids were constructed such that one of the DNA strands was methylated by the DNA adenine methylase (Dam), while the other strand was unmethylated. The processing of a hemimethylated single-base-pair mismatch had no effect on the recombination of the adjacent gene, suggesting that the most efficient (Dam-instructed) mismatch repair process does not secondarily promote genetic recombination. However, mismatches that could form an ordered secondary structure resembling a cruciform increased the recombination of this adjacent gene at least 20-fold. An identical mismatch that could not form an ordered secondary structure had no effect in this system. The increased frequency of recombination observed was found to require the recB or recC gene product or both. Furthermore, the recombination appeared unidirectional, in that the cruciform-containing plasmid did not produce stable transformants. Our results support a model in which the cruciform-containing plasmid can participate in recombination with the resident plasmid but is unable to produce stable transformant progeny. A proposed role for the RecBCD enzyme (ExoV) in this process is discussed.     

Suppression of Escherichia coli recF mutations by recA-linked srfA mutations.

Suppressors of recF (srfA) were found by selection for resistance to mitomycin C and UV irradiation in a recB21 recC22 sbcB15 recF143 strain. srfA mutations map in recA and are dominant to srfA+. They suppress both the DNA repair and the recombination deficiencies due to recF mutations. Therefore, RecA protein which is altered by the srfA mutation can allow genetic recombination to proceed in the absence of recB, recC, and recF functions. recF is also required for induction of the SOS response after UV damage. We propose that recF+ normally functions to allow the expression of two recA activities, one that is required for the RecF pathway of recombination and another that is required for SOS induction. The two RecA activities are different and are separable by mutation since srfA mutations permit recombination to proceed but have not caused a dramatic increase in SOS induction in recF mutants. According to this hypothesis, one role for recF in DNA repair and recombination is to modulate RecA activities to allow RecA to participate in these recF-dependent processes.     

The Escherichia coli MutL protein stimulates binding of Vsr and MutS to heteroduplex DNA.

Vsr DNA mismatch endonuclease is the key enzyme of very short patch (VSP) DNA mismatch repair and nicks the T-containing strand at the site of a T-G mismatch in a sequence-dependent manner. MutS is part of the mutHLS repair system and binds to diverse mismatches in DNA. The function of the mutL gene product is currently unclear but mutations in the gene abolish mutHLS -dependent repair. The absence of MutL severely reduces VSP repair but does not abolish it. Purified MutL appears to act catalytically to bind Vsr to its substrate; one-hundredth of an equivalent of MutL is sufficient to bring about a significant effect. MutL enhances binding of MutS to its substrate 6-fold but does so in a stoichiometric manner. Mutational studies indicate that the MutL interaction region lies within the N-terminal 330 amino acids and that the MutL multimerization region is at the C-terminal end. MutL mutant monomeric forms can stimulate MutS binding.     

Genetic analysis of the recF pathway to genetic recombination in Escherichia coli k12: Isolation and characterization of mutants

A recombination proficient strain ofEscherichia coli which is recB^? recC^? sbcB^? has been subjected to mutagenesis by nitrosoguanidine. Among the recombination deficient mutants isolated one was sbcB⁺, three were recA^～ and 11 were mutants in at least four newrec genes: recF, recJ, recK and recL. recF143 and recL152 are cotransducible with ilv but they lie on opposite sides of the ilv operons as determined by F$\text{\$}\text{?}$studies. recF, recL and recK are not involved in the RecBC pathway of recombination since a recB⁺recC⁺sbcB^? strain carrying a mutation in one of these genes is recombination proficient. Hence the hypothesis that a RecF pathway of recombination can operate as a partially independent substitute for the RecBC pathway of recombination is supported. recF^?recB⁺ and recF⁺recB^? single mutants are sensitive to u.v. irradiation while the recF^?recB^? double mutant is more sensitive than either single mutant. The sensitivity of the recB^?recC^?sbcB^?recF^? strain approaches the sensitivity of a recA^? single mutant. This is interpreted to mean that there are partially independent RecF and RecBC pathways for the repair of u.v. damage. recJ and mutations were not mapped precisely; hence the mutant properties they confer can not be stated conclusively.     

Formation of heteroduplex DNA during mammalian intrachromosomal gene conversion.

We have studied intrachromosomal gene conversion in mouse Ltk- cells with a substrate designed to provide genetic evidence for heteroduplex DNA. Our recombination substrate consists of two defective chicken thymidine kinase genes arranged so as to favor the selection of gene conversion products. The gene intended to serve as the recipient in gene conversion differs from the donor sequence by virtue of a palindromic insertion that creates silent restriction site polymorphisms between the two genes. While selection for gene conversion at a XhoI linker insertion within the recipient gene results in coconversion of the nearby palindromic site in more than half of the convertants, 4% of convertant colonies show both parental and nonparental genotypes at the polymorphic site. We consider these mixed colonies to be the result of genotypic sectoring and interpret this sectoring to be a consequence of unrepaired heteroduplex DNA at the polymorphic palindromic site. DNA replication through the heteroduplex recombination intermediate generates genetically distinct daughter cells that comprise a single colony. We believe that the data provide the first compelling genetic evidence for the presence of heteroduplex DNA during chromosomal gene conversion in mammalian cells.     

Purification and preliminary characterization of the Escherichia coli K-12 recF protein.

The recF gene of Escherichia coli is known to encode an Mr-40,000 protein that is involved in DNA recombinationa nd postreplication DNA repair. To characterize the role of the recF gene product in these processes, the recF gene was cloned downstream of a tac promoter to facilitate overproduction of the recF gene product. The RecF protein was overproduced and purified to apparent homogeneity. N-terminal protein sequence analysis demonstrated that the purified protein had the sequence that was predicted from the DNA sequence of the recF gene, except that the predicted N-terminal Met was not present. The RecF protein bound to single-stranded oligonucleotides in filter binding and gel filtration assays. Maximal binding required 2 to 3 min of incubation at 37 degrees C; the binding reaction had a pH optimum of 7.0, did not require divalent cations, and was inhibited by NaCl concentrations of greater than 250 mM. The Kd of RecF protein binding to a 59-base single-stranded oligonucleotide was on the order of 1.3 X 10(-7) M, and the reaction did not show cooperativity. Experiments measuring the binding to various DNA substrates and competition binding experiments with different DNA molecules demonstrated that RecF protein binds preferentially to single-stranded, linear DNA molecules.     

Gene conversion in Escherichia coli. Resolution of heteroallelic mismatched nucleotides by co-repair

We have constructed heteroduplex plasmid DNA that is similar in structure to the heteroduplex DNA expected to be produced during genetic recombination of plasmids, and studied its repair after transformation into different Escherichia coli strains. The heteroduplex DNA was constructed using two different parental plasmids, each of which contained a different ten-nucleotide insertion mutation. The effect of different defined states of dam-methylation on repair was also examined. We found that heteroduplex DNA repair occurred prior to the replication of the substrate DNA 60 to 80% of the time, regardless of the state of DNA methylation. Most excision/synthesis tracts covered two markers separated by 1243 base-pairs, and this process has been termed co-repair. The most efficient co-repair pathway was the Dam-instructed repair pathway that required the mutH, mutL, mutS and uvrD gene products and preferentially used the methylated strand as the template for DNA synthesis. If there was no methylation asymmetry, mismatch nucleotide repair occurred with a similar frequency; however, no strand bias was observed. Co-repair of symmetrically methylated heteroduplex DNA required the mutS and uvrD gene products, while repair of unmethylated heteroduplex DNA also required the mutL and mutH gene products.     

Plasmidic recombination in Escherichia coli K-12: the role of recF gene function

Interplasmidic and intraplasmidic recombination proficiencies were determined in E. coli bacterial strains carrying rec mutations. Our results defined the role of recF gene function, recB, recC, and sbcB gene products (exonuclease V and exonuclease I) in plasmidic recombination in wild-type E. coli cells and in cells in which the recE recombination pathway is activated. RecF gene function is required for interplasmidic recombination regardless of the recB recC genotype. Intraplasmidic recombination is recF dependent in cells having a functional exonuclease V, but not in recB recC mutants. Exonuclease V activity inhibits both interplasmidic and intraplasmidic recombination via the recE pathway.     

A minor pathway of postreplication repair in Escherichia coli is independent of the recB, recC and recF genes

After ultraviolet (UV) irradiation, an Escherichia coli K12 uvrB5 recB21 recF143 strain (SR1203) was able to perform a limited amount of postreplication repair when incubated in minimal growth medium (MM), but not if incubated in a rich growth medium. Similarly, this strain showed a higher survival after UV irradiation if plated on MM versus rich growth medium (i.e., it showed minimal medium recovery (MMR]. In fact, its survival after UV irradiation on rich growth medium was similar to that of a uvrB5 recA56 strain, which does not show MMR or postreplication repair. The results obtained with a uvrB5 recF332::Tn3 delta recBC strain and a uvrB5 recF332::Tn3 recB21 recC22 strain were similar to those obtained for strain SR1203, suggesting that the recB21 and recF143 alleles are not leaky in strain SR1203. The treatment of UV-irradiated uvrB5 recB21 recF143 and uvrB5 recF332::Tn3 delta recBC cells with rifampicin for 2 h had no effect on survival or the repair of DNA daughter-strand gaps. Therefore, a pathway of postreplication repair has been demonstrated that is constitutive in nature, is inhibited by postirradiation incubation in rich growth medium, and does not require the recB, recC and recF gene products, which control the major pathways of postreplication repair.     

Gene replacement with linear DNA in electroporated wild-type Escherichia coli.

Gene replacement using linear double-stranded DNA fragments in wild-type Escherichia coli transformation is generally inefficient due to exonucleolytic degradation of incoming DNA. Recombination-proficient strains, in which the exonucleolytic activity of RecBCD is inactivated, have been used as transformation recipients to overcome this difficulty. Here we report that gene replacements using linear double-stranded donor DNA can be achieved in wild-type E.coli if electrocompetent cells are used. Using a plasmid target, we obtained 10(2)-10(3) gene replacement events/microgram linear DNA. Using an independent chromosomal target, approximately 60 gene replacement events/microgram linear DNA were obtained. The presence of Chi sites on the linear DNA, which are known to block DNA degradation and stimulate recombination in E.coli, had no effect on gene replacement efficiency in either case. RecBCD-mediated exonucleolytic activity was found to be diminished in electroporated cells. Electrotransformation thus provides a simple way to perform gene replacements in many E.coli strains.     

Genetic and physical mapping of recF in Escherichia coli K-12

Summary Two factor transductional crosses place recF at approximately 82 min on the E. coli chromosome; recF is highly cotransducible with dnaA and gyrB (cou). Transductional analysis with a series of tna specialized transducing phages carrying chromosomal DNA from the tnaA region place recF between dnaA and gyrB. This analysis also indicates that a gene lying in the same region and producing an easily detectable protein (estimated MW of 45 kD) is dnaN and not recF.     

Altered induction of the adaptive response to alkylation damage in Escherichia coli recF mutants.

Escherichia coli recF mutants are hypermutable when treated with methyl methanesulfonate (G. C. Walker, Mol. Gen. Genet. 152:93-103, 1977). In this study, methylation hypermutability of recF mutant strains was examined, and it was found that recF+ is required for normal induction of the adaptive response to alkylation damage. Although this regulatory effect of recF mutations results in reduced levels of enzymes that specifically repair methyl lesions in DNA, it only partially explains the hypermutability. Further examination showed that methylation hypermutability of recF mutant strains required a functional umuDC operon, a component of the SOS response. These results lead to the hypothesis that methylation hypermutability results from the effects of recF mutations on the induction of both the SOS response and the adaptive response.     

Stationary phase induction of dnaN and recF, two genes of Escherichia coli involved in DNA replication and repair.

The beta subunit of DNA polymerase III holoenzyme, the Escherichia coli chromosomal replicase, is a sliding DNA clamp responsible for tethering the polymerase to DNA and endowing it with high processivity. The gene encoding beta, dnaN, maps between dnaA and recF, which are involved in initiation of DNA replication at oriC and resumption of DNA replication at disrupted replication forks, respectively. In exponentially growing cells, dnaN and recF are expressed predominantly from the dnaA promoters. However, we have found that stationary phase induction of the dnaN promoters drastically changes the expression pattern of the dnaA operon genes. As a striking consequence, synthesis of the beta subunit and RecF protein increases when cell metabolism is slowing down. Such an induction is dependent on the stationary phase sigma factor, RpoS, although the accumulation of this factor alone is not sufficient to activate the dnaN promoters. These promoters are located in DNA regions without static bending, and the -35 hexamer element is essential for their RpoS-dependent induction. Our results suggest that stationary phase-dependent mechanisms have evolved in order to coordinate expression of dnaN and recF independently of the dnaA regulatory region. These mechanisms might be part of a developmental programme aimed at maintaining DNA integrity under stress conditions.