Mechanism and Control of Interspecies Recombination in Escherichia coli. I. Mismatch Repair, Methylation, Recombination and Replication Functions

Corresponding author: Snjezana Stambuk, Laboratoire de Mutagénèse, Institut J. Monod, Université Paris 7, 2, place Jussieu, 75251, Paris, France., stambuk{at}ijm.jussieu.fr (E-mail).

Communicating editor: P. J. PUKKILA

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

A genetic analysis of interspecies recombination in Escherichia coli between the linear Hfr DNA from Salmonella typhimurium and the circular recipient chromosome reveals some fundamental aspects of recombination between related DNA sequences. The MutS and MutL mismatch binding proteins edit (prevent) homeologous recombination between these 16% diverged genomes by at least two distinct mechanisms. One is MutH independent and presumably acts by aborting the initiated recombination through the UvrD helicase activity. The RecBCD nuclease might contribute to this editing step, presumably by preventing reiterated initiations of recombination at a given locus. The other editing mechanism is MutH dependent, requires unmethylated GATC sequences, and probably corresponds to an incomplete long-patch mismatch repair process that does not depend on UvrD helicase activity. Insignificant effects of the Dam methylation of parental DNAs suggest that unmethylated GATC sequences involved in the MutH-dependent editing are newly synthesized in the course of recombination. This hypothetical, recombination-associated DNA synthesis involves PriA and RecF functions, which, therefore, determine the extent of MutH effect on interspecies recombination. Sequence divergence of recombining DNAs appears to limit the frequency, length, and stability of early heteroduplex intermediates, which can be stabilized, and the recombinants mature via the initiation of DNA replication.

HOMOLOGOUS genetic recombination is required for DNA repair and for meiotic crossovers involved in chromosome disjunction (KUCHERLAPATI and SMITH 1988 ). However, crossovers between interspersed repeated sequences cause deleterious chromosomal rearrangements (for review see RADMAN 1991 ). A balance between the positive and negative effects of homologous recombination is kept by cellular mechanisms that control its frequency and fidelity. The key role of homologous recombination in the repair of DNA probably sets the limits to the extent of negative control of recombination. For example, a cellular Rec^- phenotype could be favored to avoid chromosomal rearrangements, but it would lead to a great disadvantage because of the deficiency in DNA repair. The solution to this particular problem appears to be provided by the high fidelity of homologous recombination (i.e., its strict requirement for sequence identity): DNA repair can proceed by unrestricted precise recombination between the identical sister chromatids, whereas the nonidentity of repeated sequences prevents their recombination (RAYSSIGUIER et al. 1989 ; PETIT et al. 1991 ; RADMAN 1991 ; ABDULKARIM and HUGHES 1996 ).

Thus, homologous genetic recombination is largely controlled at the DNA substrate level by the degree and length of sequence identity shared by the two recombining DNAs (for review see RADMAN 1991 ). The decrease in DNA homology through sequence divergence is much more efficient in preventing recombination than is the decrease in the length of homology (SHEN and HUANG 1986 ). Even a low divergence, e.g., 1% or less, can severely inhibit homologous recombination in bacteria, yeast, and mammalian cells (DE WIND et al. 1995 ; DATTA et al. 1996 , DATTA et al. 1997 ; VULIC et al. 1997 ; ZAHRT and MALOY 1997 ). This high fidelity of genetic recombination is caused not only by the intrinsic properties of recombination enzymes, but also by the editing of recombination by the mismatch repair system, the same system that controls also the fidelity of DNA replication through the correction of base pair mismatches caused by replication errors (for review see RADMAN and WAGNER 1993A ; MODRICH and LAHUE 1996 ).

Deficiency in any of the four mismatch repair proteins, MutS, MutL, MutH, and UvrD (MutU), has equal effect on the correction of DNA replication errors, reflecting the requirement of all four proteins in the successful error correction process (for review see RADMAN and WAGNER 1986 ). This is generally not the case for the editing of homologous DNA recombination, where the effect of specific mut gene mutations depends on the nature of recombination substrates and events (PETIT et al. 1991 ; ABDULKARIM and HUGHES 1996 ). Interspecies recombination between Salmonella and Escherichia in conjugational and transductional crosses is increased ~1000-fold by mutS and/or mutL mutations, ~20-fold by mutH, and only about fivefold by a mutU mutation (RAYSSIGUIER et al. 1989 ; see Table 2).

We have been studying gene exchange between two related species, Escherichia coli and Salmonella typhimurium, mediated by homologous recombination enzymes as a model system for the definition of the genetic barriers at the molecular level (RAYSSIGUIER et al. 1989 ; MATIC et al. 1995 ; VULIC et al. 1997 ). The principal genetic barrier among enterobacteria is the recombinational barrier, whose structural element is the genomic sequence divergence and whose enzymatic element is the mismatch repair system. So far, the basic molecular rules for this interspecies recombination appear to apply also to other homologous recombination systems involving similar sequences (homeologous recombination) in bacteria, yeast, and mammals (RADMAN and WAGNER 1993A ; DE WIND et al. 1995 ; DATTA et al. 1996 , DATTA et al. 1997 ; VULIC et al. 1997 ). Although the editing of homeologous strand exchange by the MutS and MutL proteins has been reproduced in vitro (WORTH et al. 1994 ), the detailed mechanism of the editing of recombination by the mismatch repair system remains obscure.

This is a study of the roles of DNA methylation and MutS/L vs. MutH functions in recombinational editing. By studying the fidelity of interspecies recombination, we found that (i) this recombination often involves DNA synthesis initiated by a pairing between 16% (on average) diverged parental sequences and that requires the PriA primosome function and some RecF functions and (ii) editing of recombination by mismatch repair proteins occurs by two mechanisms: one is MutH independent and the other is MutH dependent.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

Bacterial strains and plasmids:
The bacterial strains used are listed in Table 1. Figure 1 shows the position and direction of the transfer of the Hfr strains used in this study. SA965, SA977, SA534, and SU573 are S. typhimurium Hfr, while PK3 is E. coli Hfr. Plasmid pDDM6, which carries the gene coding for the Dam methylase enzyme (GUHA and GUSCHLBAUER 1992 ) under the control of the tac promoter, was propagated in mut⁺ and mutH (AB1157) strains in the presence of 3 mM IPTG to induce Dam methylase.

View larger version (15K): [in this window] [in a new window] [Download PPT slide]   Figure 1. The schematic representation of the position and direction of transfer of Hfr donors used in this study (see MATERIALS AND METHODS). E. coli Hfr strain is PK3, while S. typhimurium Hfr strains are SA977, SA965, SA534, and SU573.

Conjugational crosses:
Donor and recipient strains were grown in rich medium or minimal medium where indicated. M63 minimal medium contained glucose (0.4%), histidine, leucine, proline, arginine, threonine (100 µg/ml), and thiamine (30 µg/ml). Log-phase bacteria were grown to 2–4 x 10⁸ cells/ml, mixed in a 1:1 Hfr:F^- ratio, and filtered through a 0.45-µm pore size filter (Schleicher & Schuell, Keene, NH) that was incubated on prewarmed, rich medium agar for 40 or 60 min. Filters were resuspended in 10^-2 M MgSO₄, and bacteria were separated by vortexing. Bacteria were plated on M63 minimal medium supplemented with histidine, leucine, proline, and threonine (100 µg/ml), thiamine (30 µg/ml), glucose (0.4%), and nalidixic acid (40 µg/ml) to counterselect donor cells, but without arginine to select for transconjugants in crosses involving SA965 and SA977 donors. Resuspension of filters on which crosses with SU573 and SA534 donors were plated on similar plates, but to select for leucine prototrophy. In interspecies crosses where the recipients used were S. typhimurium and donor E. coli PK3, selection was on M63 minimal medium containing glucose (0.4%), thiamine (30 µg/ml), isoleucine, and valine (100 µg/ml), and Streptomycin (100 µg/ml) to counterselect donor cells. Colonies were counted after 4 days or more if scoring for the frequency of recombinants of priA cells was performed. In an intraspecies cross where the donor used was PK3, selection was performed for arginine marker.

To minimize the appearance of suppressor mutations because of the mutator effect of mutS and mutH recipients used, the priA and recF strains were grown in M63 minimal medium containing glucose (0.4%, MCPHEE 1993 ), histidine, leucine, proline, arginine, threonine (100 µg/ml), and thiamine (30 µg/ml) until their midexponential phase of regrowth, and were then agitated for 3 hr in rich medium before conjugation. Recombination frequency is expressed as the number of recombinants per Hfr donor. Values represent the means and standard deviations of the recombination frequencies obtained in independent crosses.

	RESULTS AND DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

MutS vs. MutH effects on interspecies recombination:
Presumably because of their mismatch recognition and binding activities (for review see MODRICH and LAHUE 1996 ), the MutS and MutL functions appear indispensable for all modes of the editing of DNA recombination and replication (this paper and unpublished results from our laboratory). The roles of MutH and UvrD in the editing of recombination are poorly understood. The key role of MutH in the mismatch repair process is to produce a single-strand scission (nick) 5' to an unmethylated GATC sequence. This MutH activity is somehow stimulated by the binding of MutS protein to a nearby mismatch (mispaired or unpaired bases in duplex DNA) followed by binding of MutL. The uvrD gene encodes DNA helicase II, whose role in mismatch repair is supposedly to peel off the nicked strand in the direction of the mismatch and expose the free single strand for destruction by single-strand-specific nucleases (for review see MODRICH and LAHUE 1996 ).

To determine the epistatic relationships of such mutations, we have studied the effect of a series of single and double mutations that affect mismatch repair and/or recombination machinery. Such analysis is useful for pathway assignments of genes that control the fidelity of genetic recombination. A large number of mutations studied makes the epistasis analysis notoriously difficult. Therefore, we chose to express the effects of mutS and mutH mutations in different mutational backgrounds (Table 2). Because the MutS function is involved in all whereas MutH is involved in only some aspects of editing of recombination by mismatch repair, the ratio of MutS to MutH effects shows the extent of MutH-independent editing (Table 2).

The ratio of the MutS to MutH effects is significantly decreased in uvrD, recD, recG, and recF mutants. LexA1 (a dominant mutation preventing the induction of SOS response), uvrD, recD, recG, and recF show a pronounced decrease in the MutS effect, whereas priA, recF, and lexA1 also show a decreased effect of MutH. The mutH effect on interspecies recombination is increased from ~20-fold in the wild type to ~120-fold in recD, 50-fold in uvrD, and 45-fold in recG (Table 2). Because the interspecies recombination frequencies are similar in the lexA1, recF, and priA mutants, it may be that the effects of RecF and PriA functions and/or of some components active in the same pathway are inducible because of the SOS response that is activated in E. coli during the course of conjugation with Salmonella (MATIC et al. 1995 ). The significance of these findings will be discussed below.

MutH-dependent editing of interspecies recombination requires unmethylated GATC sequences:
Methylation of the adenine moiety in the GATC sequence to 6-methyl-adenine by the Dam methylase prevents the MutH activity and, thus, prevents mismatch repair (for review see MODRICH and LAHUE 1996 ). However, internal free ends, such as nicks, can substitute for both the GATC sequence and MutH function (LAHUE et al. 1987 ; LANGLE-ROUAULT et al. 1987 ).

Table 3 shows that the overexpression of Dam methylase (in the F^- cells) during the course of S. typhimurium Hfr x E. coli F^- conjugational crosses increases interspecies recombination to the same extent as the mutH mutation. Furthermore, Dam overexpression has no effect on recombination in a mutH mutant (Table 3). This result suggests that the MutH-dependent editing of recombination requires unmethylated GATC sequences in one or both strands of the DNA heteroduplex, just as in the case of standard mismatch repair.

What is the origin of unmethylated GATC sequences in recombination: parental Hfr DNA, parental newly synthesized F^- DNA, or some particular DNA synthesis associated with the recombination process?

Methylation of parental DNAs does not determine the MutH effect on interspecies recombination:
To explore the effect of Dam methylation of parental DNAs on the MutH-dependent editing of recombination, the following interspecies crosses between E. coli Hfr and S. typhimurium F^- were carried out: Hfr (Dam⁺) x F^- (Dam⁺), Hfr (Dam⁺) x F^- (Dam^-), Hfr (Dam^-) x F^- (Dam⁺), and Hfr (Dam^-) x F^- (Dam^-). Note that the crosses in Table 4 are exceptions in this article in that E. coli is the Hfr. These crosses cannot be compared with those in the article by RAYSSIGUIER et al. 1989 , which were performed too long ago. The results in Table 4 show that the effect of MutH on interspecies recombination is not affected by the methylation status of parental DNAs in a way that could be expected from the known roles of the MutH protein and unmethylated GATC sequences. For example, in the interspecies cross between S. typhimurium F^- (Dam^-) and E. coli, Hfr (Dam⁺ or Dam^-), one would expect the highest effect of MutH-dependent rejection of the Hfr DNA, i.e., the highest ratio of recombination in mutH compared with mutH⁺ crosses. In the total absence of Dam methylation, even the inactivation of mismatched intermediates by the MutS, L, and H activities can be expected (DOUTRIAUX et al. 1986 ). However, this was not found: when both parents are Dam⁺, the MutH effect on recombination is 10-fold (line 1 vs. line 3), as compared to 13-fold when only the Hfr DNA is unmethylated (line 5 vs. line 7); when both DNAs are unmethylated, the MutH effect is also 10-fold (line 6 vs. line 8). Note that independently of the donor's DNA methylation, recombination frequency is about five times higher in dam recipients, presumably because of their constitutive expression of SOS response (PETERSON and MOUNT 1993 ).

PriA, recF, and lexA1 mutations affect MutH-dependent editing of interspecies recombination:
Because the MutH-dependent editing of recombination intermediates requires unmethylated GATC sequences (Table 3), which are probably not carried by the parental DNAs (Table 4), it may be that a DNA synthesis mechanistically associated with genetic recombination produces such sequences. The functions of two proteins are supposed to be involved in both DNA recombination and replication: the primosome assembly protein PriA, which is apparently required for replication dependent recombination (KOGOMA et al. 1996 ; SANDLER et al. 1996 ), and the RecF recombination protein, which was claimed to be required for resumption of replication forks disrupted by radiation damage to DNA (COURCELLE et al. 1997 ). PriA protein is hypothesized to recognize sites for loading of DnaB helicase onto the single-stranded DNA within a D loop structure produced by the invading DNA strand and, in that way, permit the primosome assembly system to initiate replication from the invading DNA strand (KOGOMA 1996 ; MCGLYNN et al. 1997 ).

If the MutH editing step of recombination requires de novo DNA synthesis associated with recombination, and if such DNA synthesis requires PriA and/or RecF functions, then the MutH effect on interspecies recombination should be diminished in priA and recF mutants. This was indeed observed (Table 2, lines 19–28): the mutH mutation increased interspecies recombination only twofold in a priA mutant (compared to the 11-fold increase in the corresponding wild-type background). Because the MutH effect is also similar in an intraspecies E. coli priA cross (twofold; data not shown), it appears that most recombination-associated DNA synthesis required for editing by MutH depends on PriA activity. The effect of RecF on editing by MutH protein is less pronounced (MutH effect is still ninefold). Perhaps by coincidence, it is quantitatively closer to the effects of lexA1 (11-fold).

Table 2 shows that a priA null mutation and a recF mutation inhibit interspecies recombination by ~15-fold in mutH background, while having a twofold and negligible effect, respectively, on intraspecies recombination (data not shown). Thus, PriA and RecF functions are more limiting for interspecies than intraspecies recombination in an otherwise wild-type genetic background. This effect is more pronounced in MutH-deficient bacteria, presumably because of the efficient MutH-dependent prevention of interspecies recombination involving DNA synthesis. The significance of this finding and the observation that the lexA1 and ruvA mutations also decrease interspecies more than intraspecies recombination (Table 2; MATIC et al. 1995 ; data not shown) are discussed below.

The difference between these results and those of a previous study on the implication of the recF gene product in interspecies recombination (MATIC et al. 1995 ) derives from the difference in leakiness of the two recF alleles used. It has been reported that the recF143 substitution mutation used in the previous study, among its other leaky phenotypes tested, exerts an effect in conjugational recombination in recBCsbcBC recipients 10 times weaker than that of the recF349 deletion allele used in this study (SANDLER and CLARK 1993 ).

A model for the editing of interspecies recombination:
These results can be organized within the frame of a model involving two mechanisms for editing of interspecies recombination in E. coli: one is MutH independent and the other is MutH dependent (Figure 2). These two mechanisms could act in parallel as two alternative mechanisms: one is DNA end dependent but MutH independent, and the other is dependent on MutH, unmethylated GATC sequences, and de novo DNA synthesis. However, because of the requirement of DNA strand transfer for de novo DNA synthesis, we hypothesize that the MutH-independent editing is an early-stage editing, whereas the MutH-dependent process is a late-stage editing. The key features of the model are as follows:

1. The recombination between diverged DNAs is limited by the number of minimum efficient processing segments (MEPS), i.e., minimal blocks of sequence identity that are sufficient for the efficient RecA-catalyzed initiation of the DNA strand transfer or exchange process (SHEN and HUANG 1986 ; VULIC et al. 1997 ). The outcome of the initiated strand invasion process (including its reversibility, RAYSSIGUIER et al. 1989 ) will depend on several recombination and repair proteins acting on intermediates of genetic recombination.

   
   

   View larger version (16K): [in this window] [in a new window] [Download PPT slide]   Figure 2. (a–g) A model for two mechanisms of editing of homologous recombination by the mismatch repair proteins. The MutH-independent (early-stage) editing of homologous recombination during the course of bacterial conjugation is directly related to the degree and pattern of sequence divergence between the two recombining DNAs. This editing (c–b) occurs before extensive branch migration and/or recombination-initiated DNA replication. The MutH-dependent (late-stage) editing (e to g) occurs after the initiation of DNA synthesis. Both mechanisms of editing by the mismatch repair components require MutS and MutL proteins: the early-stage editing requires specifically UvrD helicase and RecBCD nuclease, and the late-stage editing requires specifically the MutH protein (see text for details). Newly synthesized DNA strands are presented by thick lines, and mismatches are represented by triangular bumps in the duplex DNA. REC+ and REC- refer to high and low recombination frequencies, respectively. (+++) Overexpression of MutS, MutL, and Dam proteins.

2. Because the sequence divergence between the recombining DNAs limits the length and/or stability of the initial heteroduplex intermediate, the interspecies recombination becomes more dependent on recombination-associated DNA replication than the isogenic intraspecies recombination. The PriA- and RecF-dependent de novo initiation of DNA synthesis upon the early heteroduplex intermediates appears to prevent the dissociation of such unstable recombination initiation events, particularly when MutH is not active (Table 2; Figure 2E). Recombination-associated DNA synthesis has been postulated for some systems of homeologous recombination in yeast (PORTER et al. 1996 ; TRAN et al. 1997 ).

3. Direct maturation of the recombination process can be provided by branch migration of the Holliday junction away from the invading end into the domain of diverged parental sequences (Figure 2D). About a 10-fold decrease of interspecies recombination by the ruvA mutation (Table 2, lines 13–15) suggests that RuvAB provides the branch migration function (MATIC et al. 1995 ). This activity may be SOS inducible (SHURVINTON and LLOYD 1982 ; see the effect of the lexA1 mutation in Table 2, lines 16–18) because interspecies conjugation induces SOS response (MATIC et al. 1995 ).

4. MutH-independent or early-stage editing of interspecies recombination corresponds to the dissociation of the earliest RecA-catalyzed heteroduplex intermediate (Figure 2C) because of mismatch formation. This editing requires the key mismatch recognition and repair functions, MutS and MutL, as well as the UvrD (helicase II) function. Helicase II can become dispensable at very high concentrations of the MutS and MutL proteins (S. TAMBUK and M. RADMAN, unpublished results). This mechanism was probably reproduced in vitro in the study by WORTH et al. 1994 . The effect of the recG mutation on interspecies recombination in the mutS mutant (Table 2, lines 10–12) suggests that the RecG protein may be recombinogenic only in the mutS mutant (compare lines 3 and 12 in Table 2) or that the MutS-mediated, MutH-independent editing is ~10 times less efficient in the recG mutant (Table 2, line 12). Curiously, a functional interaction was postulated between a RecG homolog of S. pneumoniae and some undefined mismatch repair function (MARTIN et al. 1996 ). This (early) stage of editing presumably occurs before the initiation of DNA synthesis (and/or RuvAB-catalyzed branch migration) and does not involve the MutH function (Table 2; Figure 2; WORTH et al. 1994 ). The RecA protein could use the aborted single-stranded end to reinitiate the strand invasion process that leads eventually to the repeated (and therefore more successful) attempts to initiate recombination in the rare regions of sufficient sequence identity (MEPS, see SHEN and HUANG 1986 ). Inspired by the results of ZAHRT and MALOY 1997 , F. W. Stahl suggested to us that the RecBCD nuclease activity renders this early editing event at a given site irreversible by the breakdown of the dissociated single-stranded end (see below). (Note that unlike events in isogenic recombination, the destruction of a donor MEPS sequence condemns the RecABCD system to search for the next MEPS within the double-stranded Hfr DNA, which, at 16% divergence from the recipient chromosome, may be many kilobases away from the Hfr DNA end). Consequently, in a recD null mutant, the repeated strand invasions would decrease the global effect of MutS on recombination as observed (Table 2, lines 1–3 vs. 7–9). In this scheme, the RecBCD nuclease is required to improve the efficiency of the early-stage editing. Therefore, the frequency and stability of the intermediate (c) should increase in the recD mutant, and the impact of the late-stage editing (the MutH effect) is expected to increase, which was found to be the case (see the next paragraph). The role of RecBCD nuclease in early-stage editing is also reflected in the observation that the contribution of this mechanism is not significant in the RecF and RecE recombination pathways (S. TAMBUK and M. RADMAN, unpublished results).

5. Late-stage editing is defined operationally as the MutH-dependent editing that appears to act on the DNA strand that contains a mismatch in the heteroduplex region derived from the parental sequences and an unmethylated GATC sequence in the newly synthetized extension of the same strand (Figure 2E). The results in Table 2 Table 3 Table 4 support this conclusion. Curiously, UvrD (helicase II) function is not required for this editing step. In fact, the increase of interspecies recombination caused by the mutH mutation is even higher in the uvrD mutant (52-fold) than in the wild-type strain (22-fold; Table 2, lines 4–6). This difference may result from the higher incidence of the late-stage intermediates caused by the inefficient early-stage editing in the uvrD mutant and, hence, more impact of the MutH-dependent editing. [The same holds for the recD and recG mutants (Table 2, lines 7–12).] Alternatively, another helicase may replace the missing helicase II and be even more efficient in this late-stage editing step. This editing step resembles an incomplete mismatch repair process that leads to the separation of the two parental molecules (Figure 2G) involved in the attempted recombination event that has already initiated DNA synthesis. We postulate that the unwinding/excision tract extends from the unmethylated GATC to the Holliday junction, thus interrupting the recombination process by separating the two parental DNA molecules (Figure 2, e–g).

6. Coincident branch migration and DNA replication (Figure 2F) should lead to the productive one-end recombination event associated with an unscheduled replication fork. Equal and nonadditive effects of the mutH mutation and of the Dam methylase overexpression (Table 3) are well accounted for by this model (Figure 2H).

The lack of a coherent effect of Dam methylation of parental DNAs on MutH- (and unmethylated GATC-) dependent editing of recombination (Table 4) leads to the prediction that the functional interaction between the mismatch recognition by MutS/L and the GATC recognition by MutH cannot cross the Holliday junction (see Figure 2 and imagine a GATC sequence to the left of the Holliday junction). If it could, then the MutH effect should be at least as efficient in the case of unmethylated parental GATCs as in the case of newly synthesized GATCs because the mismatch repair process should necessarily destroy the three-stranded junction in the former case. This implies that for efficient MutS, L, and H activity, the mismatch and the GATC sequence must be strictly "in cis" on the same strand of a heteroduplex DNA. Experiments with structural barriers between the mismatch and the GATC sequence should test this prediction.

Comparison with homeologous recombination in other systems:
The conspicuous resemblance in the genetic requirements between the interspecies recombination and the adaptive frameshift mutagenesis in E. coli (FOSTER 1998 ; ROSENBERG et al. 1998 ) may be more than accidental. The key intermediates in the adaptive mutagenesis proposed by ROSENBERG and colleagues (HARRIS et al. 1996 ) are identical to the intermediates in Figure 2. Could it be that the privileged sites for adaptive mutations are created by recombination-associated DNA synthesis? Such a process seems to occur in yeast and involves the REV3 DNA polymerase (STRATHERN et al. 1995 ; HOLBECK and STRATHERN 1997 ).

In yeast, meiotic interspecies (S. cerevisiae x S. paradoxus) recombination (HUNTER et al. 1996 ) and several mitotic recombination systems showed effects of null mutations in gene homologs of bacterial mutS and mutL genes (ALANI et al. 1994 ; SELVA et al. 1995 ; DATTA et al. 1996 , DATTA et al. 1997 ) similar to those in bacterial crosses. However, in some yeast homeologous recombination systems associated with transformation, the effects of mismatch repair mutations appear to range from weak to nonexistent (MEZARD et al. 1992 ; PRIEBE et al. 1994 ; PORTER et al. 1996 ).

Gene replacement by homologous recombination between a linear donor DNA fragment and the chromosome in the mouse embryonic stem cells is highly sensitive to the natural sequence divergence (0.6%), and this effect can be totally accounted for by the activity involving a mutS homolog gene (MSH2, DE WIND et al. 1995 ). This result suggests that the practice of gene therapy by gene replacement will have to take into account the effects of DNA sequence divergence (polymorphism) that differentiate donor and recipient DNAs and those of mismatch repair systems. Last but not least, facilitated recombination between genes and/or entire genomes of related species may provide a tool for fast experimental evolution of new genes/proteins and new metabolic pathways (for review see RADMAN and WAGNER 1993B ).

	FOOTNOTES

M. RADMAN DEDICATES THIS PAPER TO JAN DRAKE, WHOSE EDITORIAL POLICY AND SHARP EDITORIAL PENCIL REJUVENATED Genetics.

	ACKNOWLEDGMENTS

We thank our colleagues Ivan Matic, Marin Vulic, and François Taddei for helpful discussions and for sharing strains. We also thank Susan M. Rosenberg for her continuing interest in this work. S.S. was supported by a contract with European Communities and by Centre National de la Recherche Scientifique (CNRS). This research was sponsored by CNRS (ACC-SV3 program) and by grants from Association pour la Recherche sur le Cancer and Ligue National Française contre le Cancer.

Manuscript received March 16, 1998; Accepted for publication July 6, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS AND DISCUSSION LITERATURE CITED

ABDULKARIM, F. and D. HUGHES, 1996  Homologous recombination between the tuf genes of Salmonella typhimurium. J. Mol. Biol. 260:506-522[Medline].

ALANI, E., R. REENAN, and R. KOLODNER, 1994  Interaction between mismatch repair and genetic recombination in Saccharomyces cerevisiae. Genetics 137:19-39[Abstract].

COURCELLE, J., C. CARSWELL-CRUMPTON, and P. HANAWALT, 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[Abstract/Free Full Text].

DATTA, A., A. ADJIRI, L. NEW, G. F. CROUSE, and S. JINKS-ROBERTSON, 1996  Mitotic crossovers between diverged sequences are regulated by mismatch repair proteins in Saccharomyces cerevisiae. Mol. Cell. Biol. 16:1085-1093[Abstract].

DATTA, A., M. HENDRIX, M. LIPSITCH, and S. JINKS-ROBERTSON, 1997  Dual roles for DNA sequence identity and the mismatch repair system in the regulation of mitotic crossing-over in yeast. Proc. Natl. Acad. Sci. USA 94:9757-9762[Abstract/Free Full Text].

DE WIND, N., M. DEKKER, A. BERNS, M. RADMAN, and H. TE RIELE, 1995  Inactivation of the mouse Msh2 gene results in mismatch repair deficiency, methylation tolerance, hyperrecombination, and predisposition to cancer. Cell 82:321-330[Medline].

DOUTRIAUX, M.-P., R. WAGNER, and M. RADMAN, 1986  Mismatch-stimulated killing. Proc. Natl. Acad. Sci. USA 83:2576-2578[Abstract/Free Full Text].

FOSTER, P. L., 1998  Adaptive mutation: Has the unicorn landed? Genetics 148:1453-1459[Abstract/Free Full Text].

GUHA, S. and W. GUSCHLBAUER, 1992  Expression of Escherichia coli dam gene in Bacillus subtilis provokes DNA damage response: N6-methyladenine is removed by two repair pathways. Biochim. Biophys. Acta 1132:309-310[Medline].

HARRIS, R. S., K. J. ROSS, and S. M. ROSENBERG, 1996  Opposing roles of the Holliday junctions processing systems of Escherichia coli in recombination-dependent adaptive mutation. Genetics 142:681-691[Abstract].

HOLBECK, S. L. and J. N. STRATHERN, 1997  A role for REV3 in mutagenesis during double-strand break repair in Saccharomyces cerevisiae. Genetics 147:1017-1024[Abstract].

HUNTER, N., S. R. CHAMBERS, E. J. LOUIS, and R. H. BORTS, 1996  The mismatch-repair system contributes to meiotic sterility in an interspecific yeast hybrid. EMBO J. 15:1726-1733[Medline].

KAHN, P. L., 1968  Isolation of high frequency recombining strains from Escherichia coli containing the V colicinogenic factor. J. Bacteriol. 96:205-214[Abstract/Free Full Text].

KOGOMA, T., 1996  Recombination by replication. Cell 85:625-627[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].

KUCHERLAPATI, R., and G. R. SMITH, 1988 Genetic Recombination. American Society for Microbiology, Washington, DC.

LAHUE, R. S., S.-S. SU, and P. MODRICH, 1987  Requirement for d(GATC) sequences in Escherichia coli mutHLS mismatch correction. Proc. Natl. Acad. Sci. USA 84:1482-1486[Abstract/Free Full Text].

LÄNGLE-ROUAULT, F., G. MAENHAUT-MICHEL, and M. RADMAN, 1987  GATC sequences, DNA nicks and the MutH function in Escherichia coli mismatch repair. EMBO J. 6:1121-1127[Medline].

MARINUS, M. G., M. CARRAWAY, A. Z. FREY, L. BROWN, and J. A. ARRAJ, 1983  Insertion mutations in the dam gene of Escherichia coli K-12. Mol. Gen. Genet. 192:288-289[Medline].

MARTIN, B., G. J. SHARPLES, O. HUMBERT, R. G. LLOYD, and J. P. CLAVERYS, 1996  The mmsA locus of Streptococcus pneumoniae encodes a RecG-like protein involved in DNA repair and in three-strand recombination. Mol. Microbiol. 19:1035-1045[Medline].

MASAI, H., T. ASAI, Y. KUBOTA, K. ARAI, and T. KOGOMA, 1994  Escherichia coli PriA protein is essential for inducible and constitutive stable DNA replication. EMBO J. 15:5338-5345.

MATIC, I., C. RAYSSIGUIER, and M. RADMAN, 1995  Interspecies gene exchange in bacteria: the role of SOS and mismatch repair systems in evolution of species. Cell 80:507-515[Medline].

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

MCPHEE, D. G., 1993  Directed mutation: paradigm postponed. Mutat. Res. 285:109-116[Medline].

MEZARD, C., D. POMPON, and A. NICOLAS, 1992  Recombination between similar but not identical DNA sequences during yeast transformation occurs within short stretches of identity. Cell 70:659-670[Medline].

MODRICH, P. and R. LAHUE, 1996  Mismatch repair in replication fidelity, genetic recombination, and cancer biology. Annu. Rev. Biochem. 65:101-133[Medline].

PANG, P. P., A. S. LUNDBERG, and G. C. WALKER, 1985  Identification and characterization of the mutL and mutS gene products of Salmonella typhimurium LT2. J. Bacteriol. 163:1007-1015[Abstract/Free Full Text].

PETERSON, K. R. and D. W. MOUNT, 1993  Analysis of the genetic requirements for viability of Escherichia coli K-12 DNA adenine methylase (dam) mutants. J. Bacteriol. 175:7505-7508[Abstract/Free Full Text].

PETIT, M. A., J. DIMPFL, M. RADMAN, and H. ECHOLS, 1991  Control of chromosomal rearrangements in E. coli by the mismatch repair system. Genetics 129:327-332[Abstract].

PORTER, G., J. WESTMORELAND, S. PRIEBE, and M. RESNICK, 1996  Homologous and homeologous intermolecular gene conversion are not differentially affected by mutations in the DNA damage or the mismatch repair genes RAD1, RAD50, RAD51, RAD52, RAD54, PMS1 and MSH2. Genetics 143:755-767[Abstract].

PRIEBE, S., J. WESTMORELAND, T. NILSSON-TILLGREN, and M. RESNICK, 1994  Induction of recombination between homologous and diverged DNAs by double-strand gaps and breaks and role of mismatch repair. Mol. Cell. Biol. 14:4802-4814[Abstract/Free Full Text].

RADMAN, M., 1991  Avoidance of inter-repeat recombination by sequence divergence and a mechanism of neutral evolution. Biochimie 73:357-361[Medline].

RADMAN, M. and R. WAGNER, 1986  Mismatch repair in Escherichia coli. Annu. Rev. Genet. 20:523-538[Medline].

RADMAN, M. and R. WAGNER, 1993a  Mismatch recognition in chromosomal interactions and speciation. Chromosoma 102:369-373[Medline].

RADMAN, M. and R. WAGNER, 1993b  DNA mismatch repair systems: mechanisms and applications in biotechnology. Biotechnol. Genet. Eng. Rev. 11:357-366[Medline].

RAYSSIGUIER, C., D. S. THALER, and M. RADMAN, 1989  The barrier to recombination between Escherichia coli and Salmonella typhimurium is disrupted in mismatch-repair mutants. Nature 342:396-401[Medline].

ROSENBERG, S. M., C. THULIN, and R. S. HARRIS, 1998  Transient and heritable mutators in adaptive evolution in the lab and in nature. Genetics 148:1559-1566[Abstract/Free Full Text].

SANDLER, S. J. and A. J. CLARK, 1993  Use of high and low level overexpression plasmids to test mutant alleles of the recF gene of Escherichia coli K-12 for partial activity. Genetics 135:643-654[Abstract].

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].

SELVA, E. M., L. NEW, G. F. CROUSE, and R. S. LAHUE, 1995  Mismatch correction acts as a barrier to homeologous recombination in Saccharomyces cerevisiae. Genetics 139:1175-1188[Abstract].

SHEN, P. and H. V. HUANG, 1986  Homologous recombination in Escherichia coli: dependence on substrate length and homology. Genetics 112:441-457[Abstract/Free Full Text].

SHURVINTON, C. E. and R. G. LLOYD, 1982  Damage to DNA induces expression of the ruv gene of Escherichia coli. Mol. Gen. Genet. 185:352-355[Medline].

STRATHERN, J. N., B. K. SHAFER, and C. B. MCGILL, 1995  DNA synthesis errors associated with double-strand-break repair. Genetics 140:965-972[Abstract].

TRAN, H., N. DEGTYAREVA, D. GORDENIN, and M. RESNICK, 1997  Altered replication and inverted repeats induce mismatch repair-independent recombination between highly diverged DNAs in yeast. Mol. Cell. Biol. 17:1027-1036[Abstract].

VACCARO, K. K. and E. C. SIEGEL, 1977  The frameshift mutability of polA1 and recA1 derivatives of mutator strains of Escherichia coli. Mutat. Res. 42:443-446[Medline].

VULIC, M., F. DIONISIO, F. TADDEI, and M. RADMAN, 1997  Molecular keys to speciation: DNA polymorphism and the control of genetic exchange in Enterobacteria. Proc. Natl. Acad. Sci. USA 94:9763-9767[Abstract/Free Full Text].

WORTH, L., JR., S. CLARK, M. RADMAN, and P. MODRICH, 1994  Mismatch repair proteins MutS and MutL inhibit RecA-catalyzed strand transfer between diverged DNAs. Proc. Natl. Acad. Sci. USA 91:3238-3241[Abstract/Free Full Text].

ZAHRT, T. C. and S. MALOY, 1997  Barriers to recombination between closely related bacteria: MutS and RecBCD inhibit recombination between Salmonella typhimurium and Salmonella typhi. Proc. Natl. Acad. Sci. USA 94:9786-9791[Abstract/Free Full Text].

This article has been cited by other articles:

				J. E. Stone, R. G. Ozbirn, T. D. Petes, and S. Jinks-Robertson Role of Proliferating Cell Nuclear Antigen Interactions in the Mismatch Repair-Dependent Processing of Mitotic and Meiotic Recombination Intermediates in Yeast Genetics, March 1, 2008; 178(3): 1221 - 1236. [Abstract] [Full Text] [PDF]

				A. Alonso, M. G. Pucciarelli, N. Figueroa-Bossi, and F. Garcia-del Portillo Increased Excision of the Salmonella Prophage ST64B Caused by a Deficiency in Dam Methylase J. Bacteriol., December 1, 2005; 187(23): 7901 - 7911. [Abstract] [Full Text] [PDF]

				J. Rakonjac, P. W. O'Toole, and M. Lubbers Isolation of Lactococcal Prolate Phage-Phage Recombinants by an Enrichment Strategy Reveals Two Novel Host Range Determinants J. Bacteriol., May 1, 2005; 187(9): 3110 - 3121. [Abstract] [Full Text] [PDF]

				T. Goldfarb and E. Alani Distinct Roles for the Saccharomyces cerevisiae Mismatch Repair Proteins in Heteroduplex Rejection, Mismatch Repair and Nonhomologous Tail Removal Genetics, February 1, 2005; 169(2): 563 - 574. [Abstract] [Full Text] [PDF]

				P. Zaleski and A. Piekarowicz Characterization of a dam mutant of Haemophilus influenzae Rd Microbiology, November 1, 2004; 150(11): 3773 - 3781. [Abstract] [Full Text] [PDF]

				N. Costantino and D. L. Court Enhanced levels of {lambda} Red-mediated recombinants in mismatch repair mutants PNAS, December 23, 2003; 100(26): 15748 - 15753. [Abstract] [Full Text] [PDF]

				P. Salvatore, C. Bucci, C. Pagliarulo, M. Tredici, R. Colicchio, G. Cantalupo, M. Bardaro, L. Del Giudice, D. R. Massardo, A. Lavitola, et al. Phenotypes of a Naturally Defective recB Allele in Neisseria meningitidis Clinical Isolates Infect. Immun., August 1, 2002; 70(8): 4185 - 4195. [Abstract] [Full Text] [PDF]

				S. Chen, S. H. Bigner, and P. Modrich High rate of CAD gene amplification in human cells deficient in MLH1 or MSH6 PNAS, November 20, 2001; 98(24): 13802 - 13807. [Abstract] [Full Text] [PDF]

				P. Funchain, A. Yeung, J. Stewart, W. M. Clendenin, and J. H. Miller Amplification of Mutator Cells in a Population as a Result of Horizontal Transfer J. Bacteriol., June 15, 2001; 183(12): 3737 - 3741. [Abstract] [Full Text]

				A. Bucka and A. Stasiak RecA-mediated strand exchange traverses substitutional heterologies more easily than deletions or insertions Nucleic Acids Res., June 15, 2001; 29(12): 2464 - 2470. [Abstract] [Full Text] [PDF]

				E. Evans and E. Alani Roles for Mismatch Repair Factors in Regulating Genetic Recombination Mol. Cell. Biol., November 1, 2000; 20(21): 7839 - 7844. [Full Text]

				I. Matic, F. Taddei, and M. Radman No Genetic Barriers between Salmonella enterica Serovar Typhimurium and Escherichia coli in SOS-Induced Mismatch Repair-Deficient Cells J. Bacteriol., October 15, 2000; 182(20): 5922 - 5924. [Abstract] [Full Text]

				K. Yesland and W. A. Fonzi Allele-specific gene targeting in Candida albicans results from heterology between alleles Microbiology, September 1, 2000; 146(9): 2097 - 2104. [Abstract] [Full Text] [PDF]

				J. Majewski, P. Zawadzki, P. Pickerill, F. M. Cohan, and C. G. Dowson Barriers to Genetic Exchange between Bacterial Species: Streptococcus pneumoniae Transformation J. Bacteriol., February 15, 2000; 182(4): 1016 - 1023. [Abstract] [Full Text]

				T. Lukacsovich and A. S. Waldman Suppression of Intrachromosomal Gene Conversion in Mammalian Cells by Small Degrees of Sequence Divergence Genetics, April 1, 1999; 151(4): 1559 - 1568. [Abstract] [Full Text]

				A. Fabisiewicz and L. Worth Jr. Escherichia coli MutS,L Modulate RuvAB-dependent Branch Migration between Diverged DNA J. Biol. Chem., March 16, 2001; 276(12): 9413 - 9420. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal