Transient and Heritable Mutators in Adaptive Evolution in the Lab and in Nature

Corresponding author: Susan M. Rosenberg, Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, smr{at}bcm.tmc.edu (E-mail).

	ABSTRACT

TOP ABSTRACT Stationary-Phase Mutation... Stationary-Phase Mutation in... Transient and Heritable Mutators... LITERATURE CITED

Major advances in understanding the molecular mechanism of recombination-dependent stationary-phase mutation in Escherichia coli occurred this past year. These advances are reviewed here, and we also present new evidence that the mutagenic state responsible is transient. We find that most stationary-phase mutants do not possess a heritable stationary-phase mutator phenotype, although a small proportion of heritable mutators was found previously. We outline similarities between this well-studied system and several recent examples of adaptive evolution associated with heritable mutator phenotype in a similarly small proportion of survivors of selection in nature and in the lab. We suggest the following: (1) Transient mutator states may also be a predominant source of adaptive mutations in these latter systems, the heritable mutators being a minority (ROSENBERG 1997 ); (2) heritable mutators may sometimes be a product of, rather than the cause of, hypermutation that gives rise to adaptive mutations.

ADAPTIVE mutations are those that allow organisms to succeed in the face of natural or artificial selections. These mutations can arise by any of several routes. "Adaptive mutation" has also been used to denote a particular set of mutational routes, described in bacteria and yeast, that differ from canonical growth-dependent mutations (e.g., LURIA and DELBRUCK 1943 ; NEWCOMB 1949 ; LEDERBERG and LEDERBERG 1952 ) in that the mutations occur after cells are exposed to selection, in cells apparently not dividing, and only had been found in genes whose functions were selected (CAIRNS et al. 1988 ; reviewed by DRAKE 1991 ; FOSTER 1993 ; HALL 1993 ; ROSENBERG 1994 , ROSENBERG 1997 ). This past year, the best-studied assay system for such mutations also has been shown to produce unselected (nonadaptive) mutations (reviewed by ROSENBERG 1997 ). Therefore, this special mode of mutation is now called stationary-phase, or stressful l ifestyle-associated mutation (SLAM), of hypermutation. It may be a major route to formation of adaptive mutations as defined here.

	Stationary-Phase Mutation pre-1997

TOP ABSTRACT Stationary-Phase Mutation... Stationary-Phase Mutation in... Transient and Heritable Mutators... LITERATURE CITED

The mere existence of a mutagenic mode that appeared to produce adaptive mutations preferentially was controversial because of the possibility that mutations might be directed, in a Lamarckian manner, specifically to the selected gene (CAIRNS et al. 1988 ). In one experimental system, the stationary-phase mutation was demonstrated to exist as a process distinct from normal growth-dependent mutation by virtue of using a different molecular mechanism of mutation. [ SEE MAENHAUT-MICHEL and SHAPIRO 1994 and MAENHAUT-MICHEL et al. 1997 for evidence that stationary-phase transposon-mediated deletion mutations are also mechanistically different from those occurring during growth.]

Recombination:
Stationary-phase reversion of a lac + 1 frameshift mutation carried on an F' sex plasmid in Escherichia coli (CAIRNS and FOSTER 1991 ) requires functional proteins of the RecBCD double-strand break-repair recombination system, whereas growth-dependent reversion of the same allele does not (HARRIS et al. 1994 , HARRIS et al. 1996 ; FOSTER et al. 1996 ). This implies that recombination is part of a unique mechanism by which the stationary-phase mutations form. Because RecBCD enzyme loads onto DNA only at double-strand ends, this also implicates DNA double-strand breaks (DSBs) as a molecular intermediate in the stationary-phase mutation mechanism (HARRIS et al. 1994 ). [ SEE TADDEI et al. 1995 , SEE TADDEI et al. 1997A for another stationary-phase mutation assay system with RecA- and RecB-dependence that may operate via a similar mechanism to that of lac system.]

One way that recombination might promote mutation is illustrated in Figure 1. A recombinational strand-exchange intermediate could prime DNA synthesis. Polymerase errors made during this synthesis could become mutations. That recombinational strand-exchange intermediates are also intermediates in recombination-dependent mutation is implied by the requirement in mutation for proteins that process such intermediates (FOSTER et al. 1996 ; HARRIS et al. 1996 ) and by evidence that a transient accumulation of these intermediates is mutagenic (HARRIS et al. 1996 ).

View larger version (6K): In this window In a new window Download PPT slide   Figure 1. —Hypothesis: DNA synthesis is primed by a recombinational strand-exchange intermediate (HARRIS et al. 1994 , HARRIS et al. 1996 ; FOSTER et al. 1996 ). This idea incorporates three known features of recombination-dependent mutation: the requirement for homologous genetic recombination, the involvement of DNA double-strand breaks (DSBs), and the evidence that the mutations result from DNA polymerase errors. Each line represents a single DNA strand, close parallel lines represent duplex DNA, arrowed ends represent 3' DNA ends, dashed line represents newly synthesized DNA, and "m" indicates a DNA polymerase mistake that can become a mutation when not corrected by postsynthesis mismatch repair. As an example of what DNAs could recombine, the homologous DNAs recombining are illustrated as sister molecules derived from replication which, although drawn as uncompleted replication products, need not be. Aspects of this hypothesis that have not been tested are discussed in the text. Figure reprinted with permission from ROSENBERG 1997 .

That DNA synthesis is part of recombination-dependent mutation was implied by the DNA sequences of the stationary-phase Lac⁺ reversions, which are nearly all -1 deletions in small mononucleotide repeats (FOSTER and TRIMARCHI 1994 ; ROSENBERG et al. 1994 ). These are characteristic of DNA polymerase errors and are different from growth-dependent Lac⁺ reversions, which are heterogeneous (FOSTER and TRIMARCHI 1994 ; ROSENBERG et al. 1994 ). The major replicative polymerase of E. coli, PolIII, has been implicated in the stationary-phase mutagenesis (FOSTER et al. 1995 ; HARRIS et al. 1997A ).

The idea in Figure 1 is similar to that hypothesized to explain origin-independent, inducible stable DNA replication in E. coli (iSDR; KOGOMA 1997 ), which has similar, but not identical, genetic requirements to recombination-dependent stationary-phase mutation (ASAI and KOGOMA 1994 ; FOSTER et al. 1996 ; HARRIS et al. 1996 ). For both, features that have not been tested experimentally include the following: (1) Whether there is a direct association of recombination with the DNA synthesis that leads to mutation. Alternatively, synthesis might occur in one genomic region with recombination required elsewhere, for example, for cell survival, or mutation fixation; (2) whether 3' end invasions initiate the DNA synthesis (leading strand) as drawn. Alternatively, 5' end invasions, evidence for which also exists (DUTREIX et al. 1991 ; ROSENBERG and HASTINGS 1991 ; MIESEL and ROTH 1996 ; RAZAVY et al. 1996 ; ROCA and COX 1997 ), could allow assembly of whole replication forks; and (3) whether the homologous DNAs that recombine are sister molecules as drawn. TORKELSON et al. 1997 provide evidence that the homologies used are not regions of amplified DNA, and discuss other possible homology sources.

Are the mutations confined to sex plasmids?
Although a novel mutational mechanism was obvious in recombination-dependent stationary-phase mutation, doubt was cast on the generality of this mechanism because of the possibility that the mechanism might be specific to sex plasmids. First, F'-encoded transfer functions are required for efficient Lac⁺ reversion on the F' (FOSTER and TRIMARCHI 1995 ; GALITSKI and ROTH 1995 ). Second, one site on the E. coli chromosome (FOSTER and TRIMARCHI 1995 ; RADICELLA et al. 1995 ; R. S. HARRIS and S. M. ROSENBERG, unpublished results), and one on the Salmonella chromosome (GALITSKI and ROTH 1995 ), do not support recombination-dependent Lac⁺ reversion. Although conjugative transfer is not required for recombination-dependent mutation (FOSTER et al. 1995 ; ROSENBERG et al. 1995 ), the suspicion lingered that somehow bacterial sex must be responsible (e.g., GALITSKI and ROTH 1996 ; BENSON 1997 ).

An alternative possibility is that the transfer (Tra) proteins promote the DSBs required for recombination-dependent mutation. Tra proteins cause single-strand nicks at the F' origin of transfer, and these could become DSBs by any of several routes including endonuclease, or replication of the nicked template (KUZMINOV 1995 ; ROSENBERG et al. 1995 ; ROSENBERG et al. 1996 ). If DSBs were limiting in the chromosome, this could account for some sites being inactive and would not exclude the possibility that DSBs formed in other ways would allow other sites to be active for recombination-dependent mutation.

	Stationary-Phase Mutation in 1997

TOP ABSTRACT Stationary-Phase Mutation... Stationary-Phase Mutation in... Transient and Heritable Mutators... LITERATURE CITED

At least five important discoveries were made last year:

Not directed:
Two groups provided evidence that the recombination-dependent mutation mechanism is not directed in a Lamarckian manner specifically at genes whose functions are selected. FOSTER 1997 measured reversion of a frameshift mutation in a defective tetracycline-resistance gene in a transposon that was hopped onto the F' that carries lac. FOSTER found recombination-dependent accumulation of tetracycline-resistant (TetR) mutants over time after exposure to lactose as the sole carbon source, coincident with the accumulation of Lac⁺ mutants. Independently, TORKELSON et al. 1997 found high frequencies of unselected frameshift-reversion, loss-of-function, and temperature-sensitive mutations at sites on a cloning vector (pBR322)-style plasmid, in the bacterial chromosome, and on the F' during recombination-dependent stationary-phase Lac⁺ reversion. The frameshift-reversion targets used displayed a preference for -1 deletions in small mononucleotide repeats, the characteristic mutation spectrum observed previously in lac. These results demonstrate that the mutational process is not directed at the selected lac gene.

Not sex plasmid-specific:
Models in which recombination-dependent stationary phase mutation was held to be specific to bacterial sex plasmids predicted that the mutagenic process would not operate on other replicons (e.g., FOSTER and TRIMARCHI 1995 ; GALITSKI and ROTH 1995 ; RADICELLA et al. 1995 ; GALITSKI and ROTH 1996 ; BENSON 1997 ). This contrasts with the detection by TORKELSON et al. 1997 of high-level mutation in three replicons, including the bacterial chromosome. To compare the ability of chromosomal and F'-located sites to undergo mutation, TORKELSON et al. 1997 screened for mutations that confer resistance to the nucleotide analog, 5-fluorocytosine (FC). These can occur by loss of function of the F'-located codAB locus, or by loss of function of the chromosomal upp gene. The FC-resistant mutations fell about equally into these two classes, as demonstrated by mapping and by a separate phenotype specific to upp mutants. This argues for a mechanism of chromosomal mutation similar to that on the F'. However, the recombination-dependence of the chromosomal mutations still needs to be tested.

Hypermutation in a subpopulation of stressed cells:
The way that TORKELSON et al. 1997 performed the screen for unselected mutations allowed us to observe that genome-wide hypermutation underlies stationary-phase Lac⁺ reversion and occurs in a subpopulation of the cells exposed to selection on lactose medium. They screened for the unselected mutations by replica plating. They compared the frequency of unselected mutations among recombination-dependent Lac⁺ mutant colonies with two important control populations: colonies from lac^- cells that were never exposed to selection (Lac^- unstressed cells) and colonies derived from Lac^- cells exposed to prolonged lactose selection, rescued from the same plates that generated Lac⁺ revertants, but which did not become Lac⁺ (Lac^- stressed cells). The Lac⁺ revertants displayed frequencies of unselected mutations 10–100 or more times greater than in either control population, indicating that they experienced hypermutation. Moreover, starvation on lactose is not sufficient to cause hypermutation, because only some of the starved cells (the Lac⁺ mutants) were hypermutated (Lac^- stressed cells were not).

Hypermutation in the subpopulation is transient:
Two possibilities for the basis of subpopulation hypermutability exist: The cells in the subpopulation could be either (1) transiently or (2) heritably hypermutable. Previously, LONGERICH et al. 1995 , and more recently TORKELSON et al. 1997 , found that most recombination-dependent Lac⁺ revertants do not possess heritable mutator phenotypes [although a small fraction of those examined by TORKELSON et al. 1997 did, as discussed below]. However, the assays they used measured growth-dependent mutation. At least two mutator mutations that are specific to recombination-dependent mutation have been described previously: recD (HARRIS et al. 1994 ) and recG (FOSTER et al. 1996 ; HARRIS et al. 1996 ) null alleles. Such mutators could have been missed in tests for growth-dependent mutators.

To ask whether Lac⁺ revertants are heritable stationary-phase mutators, we "recycled" 12 independent recombination-dependent stationary-phase Lac⁺ revertants to Lac^- and tested them in another round of stationary-phase mutation. This was done by replacing the Lac⁺-conferring sequences with the original lac^- frameshift mutation and then by measuring rates of stationary-phase Lac⁺ reversion in these strains (Table 1). None of the 12 recycled stationary-phase mutants possesses a mutator phenotype for stationary-phase mutation (Table 1). This indicates that most cells that have undergone recombination-dependent stationary-phase Lac⁺ reversion are descended from a transiently hypermutable subpopulation of all of the stressed cells. Two of the 12 are deficient in stationary-phase mutation (antimutators) (Table 1). Although the origin of the two antimutator mutations is not known, we found that the antimutator phenotype is not linked with the lac locus: 20 out of 20 transductants derived from each antimutator did not cotransduce the antimutator phenotype with a selectable transposon marker linked to lac (marker described in Table 1). It is possible that the antimutator phenotype results from unselected mutation(s) that occurred during the first round of stationary-phase mutation.

The hypermutable subpopulation:
The finding of TORKELSON et al. 1997 that unselected mutations were elevated only among Lac⁺ revertants does not conflict with FOSTER's (1997) finding of unselected mutants that were Lac^-, because FOSTER's unselected mutants occurred orders of magnitude less frequently than could have been detected in the replica plating screen of TORKELSON et al. 1997 . However, FOSTER's data do bear on the nature of the hypermutable subpopulation discovered by TORKELSON et al. 1997 . HALL 1990 suggested that stressed cells might enter a hypermutable state that they could leave only by generating the selected (adaptive) mutation, or by death if they failed to do so. This would produce survivors that all carry an adaptive mutation, in contrast with FOSTER's (1997) data. Thus, current data suggest that the hypermutable state is not lethal, or at least not always lethal, in the absence of an adaptive mutation (ROSENBERG 1997 ; TORKELSON et al. 1997 ).

Hot and cold sites:
The multiple chromosomal mutation targets assayed by TORKELSON et al. 1997 varied dramatically in their mutability, showing no relationship between target size and frequency of mutation. Two fermentation regulons displayed fewer loss-of-function mutations than one single gene target. This implies the existence of hot and cold regions for the recombination-dependent stationary-phase mutation mechanism. This finding could explain the previous failure to detect recombination-dependent mutation to Lac⁺ at the lac chromosomal locus (FOSTER and TRIMARCHI 1995 ; RADICELLA et al. 1995 ; R. S. HARRIS and S. M. ROSENBERG, unpublished results) and at one locus on the Salmonella chromosome (GALITSKI and ROTH 1995 ), and the previous lack of accumulation of a specific substitution mutation coincident with recombination-dependent Lac⁺ reversion (FOSTER 1994 ). Also, stationary-phase mutation systems that operate independently of recombination proteins (e.g., BRIDGES 1993 ; HALL 1995 ; GALITSKI and ROTH 1996 ) may show recombination-independence because their mutation targets reside in cold regions. [It is also possible that those with no known genetic requirements may simply assay growth-dependent mutations under circumstances in which growth is difficult to measure. However, the RecA-independent mutations of MCKENZIE et al. (1998) and GALITSKI and ROTH 1996 do not occur during normal growth or reflect normal growth-dependent mutation rates, respectively.] Because recombination-dependent mutation requires DSBs, region-to-region variability in DSB-formation might be responsible for mutational hot and cold sites (ROSENBERG et al. 1995 ; ROSENBERG 1997 ). DSBs can be caused by arrest of replication forks (MICHEL et al. 1997 ), by oxidative damage (BRIDGES 1997 ), and by other enzymatic and chemical means (ROSENBERG et al. 1996 ), any of which might occur heterogeneously in the bacterial chromosome. Understanding the cause of hot and cold regions will provide important insights into bacterial hypermutation.

Mismatch repair modulation:
The post-replicative mismatch repair (MMR) system of E. coli (reviewed by MODRICH and LAHUE 1996 ) is the single largest contributor to mutation avoidance and influences multiple aspects of genetic stability, including genome rearrangement (PETIT et al. 1991 ), other forms of DNA repair (e.g., LIEB and SHEHNAZ 1995 ; MELLON and CHAMPE 1996 ), and interspecies gene transfer (RAYSSIGUIER et al. 1989 ; MATIC et al. 1995 , MATIC et al. 1996 ; ZAHRT and MALOY 1997 ). Homologs of the E. coli MMR proteins appear to play similar roles in other bacteria and in simple and complex eukaryotes (e.g., MODRICH 1995 ; RADMAN et al. 1995 ; KOLODNER 1996 ). A 1997 study of recombination-dependent stationary-phase mutation in E. coli has produced the first evidence from any organism that MMR can be modulated and is not constitutively active (HARRIS et al. 1997B ).

Transient inhibition of MMR during stationary-phase mutation was suggested by the sequences of the stationary-phase Lac⁺ reversions. These are nearly all -1 deletions in small mononucleotide repeats, unlike growth-dependent Lac reversions (FOSTER and TRIMARCHI 1994 ; ROSENBERG et al. 1994 ), but identical to growth-dependent reversions in MMR-defective cells (LONGERICH et al. 1995 ). Because most stationary-phase revertants are not heritably MMR-defective (LONGERICH et al. 1995 ; TORKELSON et al. 1997 , and Table 1 above), these results implied a transient insufficiency of MMR activity during stationary-phase mutation. Cellular levels of MutS and MutH MMR proteins were shown to decrease during the stationary phase (FENG et al. 1996 ), but did these proteins become limiting for MMR, or did they simply fall to levels appropriate for the decreased replication in the stationary phase?

HARRIS et al. 1997B overproduced MMR proteins during recombination-dependent stationary-phase mutation, reasoning that if one were limiting, its production would restore MMR function and decrease stationary-phase mutation. MutS gave no such effect, suggesting that its stationary-phase decline is to a level appropriate for the decreased replication. MutL overproduction decreased mutation (HARRIS et al. 1997B ), but the levels of measurable MutL protein remained constant during the stationary phase (FENG et al. 1996 ; HARRIS et al. 1997B ). We suggest several mechanisms by which MutL might be visible on Western blots but not active for mismatch repair during stationary-phase mutation (HARRIS et al. 1997B ). Whatever mechanism ultimately proves to be responsible, these results demonstrate a transient limitation of mismatch repair capacity, caused at the level of MutL protein, that is specific to stationary-phase mutation. (We show that MutL is not limiting during growth.) It remains to be tested whether the whole population or the subpopulation experiences down-modulated MMR. We have suggested previously that the subpopulation could be differentiated, not by MMR capacity, but rather by experiencing DSBs, as these are the sole loading sites for RecBCD and so represent a key control point for the ability to perform RecBCD-mediated recombination (HARRIS et al. 1994 ; ROSENBERG et al. 1995 ; ROSENBERG et al. 1996 ; ROSENBERG 1997 ; TORKELSON et al. 1997 ).

Stationary phase in bacteria is both a response to environmental conditions and a differentiated state (SEIGELE and KOLTER 1992 ). After regulation of MMR in response to environmental conditions or cell differentiation would have profound consequences for evolution, development, and cancer formation. See RICHARD et al. (1997) for a similar modulation in stationary-phase human cells.

A model:
A model for the mechanism of recombination-dependent stationary-phase mutation is shown in Figure 2. Some of the features of this model that have not been tested are discussed above.

View larger version (12K): In this window In a new window Download PPT slide   Figure 2. —A model for the mechanism of recombination-dependent stationary-phase mutation (ROSENBERG 1997 ). Untested aspects of this model are described in the text.

Further unknowns include how a cell enters the hypermutable subpopulation. DSBs may be part of the differentiation (ROSENBERG et al. 1994 , ROSENBERG et al. 1996 ) but may not account for all of it, and moreover, how DSBs form is unknown. The DSBs may trigger an SOS response, which can be induced by starvation (TADDEI et al. 1995 ) and which is required for efficient recombination-dependent hypermutation (HARRIS 1997 ), and so may be part of a signal transduction process leading to a hypermutable state. [The first links to SOS were provided by CAIRNS and FOSTER 1991 . Some of these data were retracted by FOSTER et al. 1996 . Data providing evidence for an SOS involvement are shown by HARRIS 1997 .] SOS may not be the whole signal transduction pathway because there are aspects of recombinational hypermutation (e.g., diminished mismatch repair) not seen in a normal SOS response. Note that standard SOS mutagenesis caused by UmuDC is not part of recombinational hypermutation (CAIRNS and FOSTER 1991 ). How hypermutation is regulated will have profound implications in biology and is a subject whose surface has barely been scratched.

	Transient and Heritable Mutators in Adaptive Evolution

TOP ABSTRACT Stationary-Phase Mutation... Stationary-Phase Mutation in... Transient and Heritable Mutators... LITERATURE CITED

Successful pathogenic (LECLERC et al. 1996 ) and commensal (MATIC et al. 1997 ) enteric bacteria in nature, survivors of long-term laboratory culture (SNIEGOWSKI et al. 1997 ) and of selection in laboratory conditions (MAO et al. 1997 ), as well as the winners in computer-simulated selection wars (TADDEI et al. 1997B ), carry a few to several percent heritable mutator mutants in the surviving populations. Those examined so far are defective in MMR (LECLERC et al. 1996 ; MATIC et al. 1997 ; SNIEGOWSKI et al. 1997 ). The heritable mutators are a minority. In these systems the mechanism(s) of accumulation of the adaptive mutations in the majority of winners has not yet been examined.

These systems seem similar to recombination-dependent hypermutation in their small percentage of heritable mutators. TORKELSON et al. 1997 found that 6 out of 55 Lac⁺ stationary-phase mutants with one or more associated unselected mutations were heritable mutator mutants. At least five of these have lost the function of an essential component of the MMR system. Four carry mutations in mutL, and one carries a mutation in mutS (TORKELSON 1997 ). We present evidence demonstrating that the mutability experienced by most cells undergoing recombination-dependent stationary-phase mutation is transient (Table 1), and that occurs an entire novel mechanism for transient mutability in this system (reviewed above).

Several points follow from comparison of recombinational hypermutation with these other natural and laboratory systems in which mutators have been found:

1. Because of the high frequency of unselected mutations among Lac⁺ stationary-phase mutants, it is possible that the few MMR-defective mutations observed are also unselected (neutral), rather than selected (beneficial) mutations. In this context it is noteworthy that antimutators are also relatively frequent (2 out of 12) among the recycled adaptive mutants in Table 1, although the frequency of antimutators among Lac^- stressed and unstressed (i.e., nonhypermutating populations) cells was not measured for comparison. In the other systems showing a similarly small proportion of heritable mutators, it is also not known whether these were the cause or consequence of mutability that generated adaptive mutations.

2. In all of these systems, the ability to alter genomes radically under stressful conditions via transient, increased mutability mechanisms may be the predominant source of adaptive mutations, heritable mutators being the minority (ROSENBERG 1997 ). That regulation, rather than heritable loss, of MMR would be more subtle and less risky has been noted (LECLERC et al. 1996 ). [See also NINIO 1991 for a specific model predicting the fraction of spontaneous mutations that should arise from transient and heritable mutators.]

3. With such normal circumstances as a commensal lifestyle being associated with apparent increased mutability, one wonders whether the laboratory may be the only environment in which selection for mutability is normally avoided.

4. Regarding the possible generality of strategies like recombination-dependent hypermutation in E. coli, events have been described in multicellular eukaryotes (e.g., MCCLINTOCK 1978 ) that looked similar to the radical genomic upheavals implied by the DSBs, recombination, and loss of mismatch repair. Also, association of mutations with recombination, sex, or both has been reported in bacteria (DEMEREC 1962 , DEMEREC 1963 ), in yeast (MAGNI and VON BORSTEL 1962 ; ESPOSITO and BRUSCHI 1993 ; STRATHERN et al. 1995 ), and in filamentous fungi (PASZEWSKI and SURZYCKI 1964 ). Finally, stationary-phase adaptive mutability has been reported in other bacteria (KASAK et al. 1997 ) and in yeast (HALL 1992 ; STEELE and JINKS-ROBERTSON 1992 ; BARANOWSKA et al. 1995 ; HEIDENREICH and WINTERSBERGER 1997 ), although whether these share mechanistic similarities with the system reviewed here remains to be seen.

Conclusions:
Recombination-dependent stationary-phase mutation was shown this past year to be neither directed in a Lamarckian way to selected genes, nor specific to sex plasmids. Transient, genome-wide hypermutation occurs in only a subpopulation of the stressed cells. Hot and cold sites may explain previous failures to find an unselected mutation and to find adaptive mutations in some chromosomal regions. The MMR system's activity is limited during recombination-dependent stationary-phase mutation by a lack of functional MutL, providing the first evidence that this important system for maintenance of genetic stability is not constitutive. A few heritable mutator mutants are found among stationary-phase mutants, and it is not clear whether these are selected or unselected products of the hypermutation that gives rise to the adaptive mutants. This small minority of heritable mutator mutants resembles that seen in several recent examples of adaptive evolution. This suggests that in these systems, too, most adaptive evolution may result from transient hypermutation that may occur using mechanisms similar to or overlapping with those being revealed in recombination-dependent stationary-phase mutation, a system in which the transient mutator state can be studied experimentally.

	FOOTNOTES

Although it is obvious that this issue of GENETICS is dedicated to JAN DRAKE, the authors wish to dedicate this article to JAN and PAM DRAKE with heartfelt appreciation for their scrupulous and exhaustive service to the genetics community, and with particular appreciation for Jan's support and encouragement of our efforts in entering the field of mutation.

	ACKNOWLEDGMENTS

Thanks to T. PAULSON for enthusing about doing the experiments reported sooner rather than later, to G. MCKENZIE for isolating the Lac⁺ revertants that were recycled to Lac^- here, to H. J. BULL and R. H. HAYNES for comments on the manuscript, and to P. J. HASTINGS and M.-J. LOMBARDO for discussions and comments on the manuscript. Supported by a grant from the National Cancer Institute of Canada funded by the Canadian Cancer Society, and by Public Health Service (USA) grants ROI GM53158. S.M.R. was supported in part as a Medical Research Council of Canada Scientist and an Alberta Heritage Foundation for Medical Research Senior Scholar.

	LITERATURE CITED

TOP ABSTRACT Stationary-Phase Mutation... Stationary-Phase Mutation in... Transient and Heritable Mutators... LITERATURE CITED

ASAI, T. and T. KOGOMA, 1994  Roles of ruvA, ruvC and recG gene functions in normal and DNA damage inducible replication of the Escherichia coli chromosome. Genetics 137:895-902[Abstract].

BARANOWSKA, H., Z. POLICINSKA, and W. J. JACHYMCZYK, 1995  Effects of the CDC2 gene on adaptive mutation in the yeast Saccharomyces cerevisiae. Curr. Genet. 28:521-525[Medline].

BENSON, S., 1997  Adaptive mutation: a general phenomenon or a special case? BioEssays 19:9-11[Medline].

BRIDGES, B. A., 1993  Spontaneous mutation in stationary-phase Escherichia coli WP2 carrying various DNA repair alleles. Mutat. Res. 302:173-176[Medline].

BRIDGES, B. A., 1997  Hypermutation under stress. Nature 387:557-558[Medline].

CAIRNS, J. and P. L. FOSTER, 1991  Adaptive reversion of a frameshift mutation in Escherichia coli. Genetics 128:695-701[Abstract].

CAIRNS, J., J. OVERBAUGH, and S. MILLER, 1988  The origin of mutants. Nature 335:142-145[Medline].

DEMEREC, M., 1962  "Selfers" attributed to unequal crossovers in Salmonella. Proc. Natl. Acad. Sci. USA 48:1695-1704.

DEMEREC, M., 1963  Selfer mutants of Salmonella typhimurium. Genetics 48:1519-1531[Free Full Text].

DRAKE, J. W., 1991  Spontaneous mutation. Annu. Rev. Genet. 25:125-146[Medline].

DUTREIX, M., B. J. RAO, and C. M. RADDING, 1991  The effects on strand exchange of 5' versus 3' ends of single-stranded DNA in RecA nucleoprotein filaments. J. Mol. Biol. 219:645-654[Medline].

ESPOSITO, M. S. and C. V. BRUSCHI, 1993  Diploid yeast cells yield homozygous spontaneous mutations. Curr. Genet. 23:430-434[Medline].

FENG, G., H.-C. T. TSUI, and M. E. WINKLER, 1996  Depletion of the cellular amounts of the MutS and MutH methyl-directed mismatch repair proteins in stationary-phase Escherichia coli K-12 cells. J. Bacteriol. 178:2388-2396[Abstract/Free Full Text].

FOSTER, P. L., 1993  Adaptive mutation: the uses of adveristy. Annu. Rev. Microbiol. 47:467-504[Medline].

FOSTER, P. L., 1994  Population dynamics of a Lac^- strain of Escherichia coli during selection for lactose utilization. Genetics 138:253-261[Abstract].

FOSTER, P. L., 1997  Nonadaptive mutations occur in the F' episome during adaptive mutation conditions in Escherichia coli. J. Bacteriol. 179:1550-1554[Abstract/Free Full Text].

FOSTER, P. L. and J. M. TRIMARCHI, 1994  Adaptive reversion of a frameshift mutation in Escherichia coli by simple base deletions in homopolymeric runs. Science 265:407-409[Abstract/Free Full Text].

FOSTER, P. L. and J. M. TRIMARCHI, 1995  Adaptive reversion of an episomal frameshift mutation in Escherichia coli requires conjugal functions but not actual conjugation. Proc. Natl. Acad. Sci. USA 92:5487-5490[Abstract/Free Full Text].

FOSTER, P. L., G. GUDMUNDSSON, J. M. TRIMARCHI, H. CAI, and M. F. GOODMAN, 1995  Proofreading-defective DNA polymerase II increases adaptive mutation in Escherichia coli. Proc. Natl. Acad. Sci. USA 92:7951-7955[Abstract/Free Full Text].

FOSTER, P. L., J. M. TRIMARCHI, and R. A. MAURER, 1996  Two enzymes, both of which process recombination intermediates, have opposite effects on adaptive mutation in Escherichia coli. Genetics 142:25-37[Abstract].

GALITSKI, T. and J. R. ROTH, 1995  Evidence that F' transfer replication underlies apparent adaptive mutation. Science 268:421-423[Abstract/Free Full Text].

GALITSKI, T. and J. R. ROTH, 1996  A search for a general phenomenon of adaptive mutability. Genetics 143:645-659[Abstract].

HALL, B. G., 1990  Spontaneous point mutations that occur more often when advantageous than when neutral. Genetics 126:5-16[Abstract].

HALL, B. G., 1992  Selection-induced mutations occur in yeast. Proc. Natl. Acad. Sci. USA 89:4300-4303[Abstract/Free Full Text].

HALL, B. G., 1993  Selection-induced mutations. Curr. Opin. Genet. and Dev. 2:943-946.

HALL, B. G., 1995  Genetics of selection-induced mutations: I. uvrA, uvrB, uvrC, and uvrD are selection-induced specific mutator loci. J. Mol. Evol. 40:86-93[Medline].

HARRIS, R. S., 1997 On a molecular mechanism for adaptive mutation in Escherichia coli. Ph.D. Thesis, Department of Genetics, University of Alberta, Edmonton.

HARRIS, R. S., S. LONGERICH, and S. M. ROSENBERG, 1994  Recombination in adaptive mutation. Science 264:258-260[Abstract/Free Full Text].

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

HARRIS, R. S., H. J. BULL, and S. M. ROSENBERG, 1997a  A direct role for DNA polymerase III in adaptive reversion of a frameshift mutation in Escherichia coli. Mutat. Res. 375:19-24[Medline].

HARRIS, R. S., G. FENG, C. THULIN, K. J. ROSS, and R. SIDHU et al., 1997b  Mismatch repair protein MutL becomes limiting during stationary-phase mutation. Genes Dev. 11:2427-2437.

HEIDENREICH, E. and U. WINTERSBERGER, 1997  Starvation for a specific amino acid induces high frequencies of rho^- mutants in Saccharomyces cerevisiae. Curr. Genet. 31:408-413[Medline].

KASAK, L., R. HORAK, and M. KIVISAAR, 1997  Promoter-creating mutations in Pseudomonas putida: a model system for the study of mutation in starving bacteria. Proc. Natl. Acad. Sci. USA 94:3134-3139[Abstract/Free Full Text].

KOGOMA, T., 1997  Stable DNA replication: interplay between DNA replication, homologous recombination, and transcription. Microbiol. Mol. Biol. Rev. 61:212-238[Abstract].

KOLODNER, R., 1996  Biochemistry and genetics of eukaryotic mismatch repair. Genes Dev. 10:1433-1442[Free Full Text].

KUZMINOV, A., 1995  Collapse and repair of replication forks in Escherichia coli. Mol. Microbiol. 16:373-384[Medline].

LECLERC, J. E., B. LI, W. L. PAYNE, and T. A. CEBULA, 1996  High mutation frequencies among Escherichia coli and Salmonella pathogens. Science 274:1208-1211[Abstract/Free Full Text].

LEDERBERG, J. and E. M. LEDERBERG, 1952  Replica plating and indirect selection of bacterial mutants. J. Bacteriol. 63:399-406[Free Full Text].

LIEB, M. and R. SHEHNAZ, 1995  Very short patch repair of T:G mismatches in vivo: importance of context and accessory proteins. J. Bacteriol. 177:660-666[Abstract/Free Full Text].

LONGERICH, S., A. M. GALLOWAY, R. S. HARRIS, C. WONG, and S. M. ROSENBERG, 1995  Adaptive mutation sequences reproduced by mismatch repair deficiency. Proc. Natl. Acad. Sci. USA 92:12017-12020[Abstract/Free Full Text].

LURIA, S. E. and M. DELBRÜCK, 1943  Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491-511[Free Full Text].

MAENHAUT-MICHEL, G. and J. A. SHAPIRO, 1994  The roles of starvation and selective substrates in the emergence of araB-lacZ fusion clones. EMBO J. 13:5229-5239[Medline].

MAENHAUT-MICHEL, G., C. E. BLAKE, D. R. LEACH, and J. A. SHAPIRO, 1997  Different structures of selected and unselected araB-lacZ fusions. Mol. Microbiol. 23:133-145.

MAGNI, G. E. and R. C. VON BORSTEL, 1962  Different rates of spontaneous mutation during mitosis and meiosis in yeast. Genetics 47:1097-1108[Free Full Text].

MAO, E. F., L. LANE, J. LEE, and J. H. MILLER, 1997  Proliferation of mutators in a cell population. J. Bacteriol. 179:417-422[Abstract/Free Full Text].

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].

MATIC, I., F. TADDEI, and M. RADMAN, 1996  Genetic barriers among bacteria. Trends Microbiol. 4:69-73[Medline].

MATIC, I., M. RADMAN, F. TADDEI, B. PICARD, and C. DOIT et al., 1997  High polymorphism of mutation rates in commensal and pathogenic Escherichia coli natural isolates. Science 277:1833-1834[Free Full Text].

MCCLINTOCK, B., 1978  Mechanisms that rapidly reorganize the genome. Stadler Symp. 10:25-47.

MELLON, I. and G. N. CHAMPE, 1996  Products of DNA mismatch repair genes mutS and mutL are required for transcription-coupled nucleotide-excision repair of the lactose operon in Escherichia coli. Proc. Natl. Acad. Sci. USA 93:1292-1297[Abstract/Free Full Text].

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

MIESEL, L. and J. R. ROTH, 1996  Evidence that functions of the "RecF pathway" contribute to RecBCD-dependent transductional recombination. J. Bacteriol. 178:3146-3155[Abstract/Free Full Text].

MODRICH, P., 1995  Mismatch repair, genetic stability and tumour avoidance. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 347:89-95[Medline].

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

NEWCOMB, H. B., 1949  Origin of bacterial variants. Nature 164:150[Medline].

NINIO, J., 1991  Transient mutators: a semiquantitative analysis of the influence of translation and transcription errors on mutation rates. Genetics 129:957-962[Abstract].

PASZEWSKI, A. and S. SURZYCKI, 1964  "Selfers" and high mutation rate during meiosis in Ascobolus immersus. Nature 204:809[Medline].

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

RADICELLA, J. P., P. U. PARK, and M. S. FOX, 1995  Adaptive mutation in Escherichia coli: A role for conjugation. Science 268:418-420[Abstract/Free Full Text].

RADMAN, M., I. MATIC, J. A. HALLIDAY, and F. TADDEI, 1995  Editing DNA replication and recombination by mismatch repair: from bacterial genetics to mechanisms of predisposition to cancer in humans. Philos. Trans. R. Soc. Lond. Biol. Sci. 347:97-103[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].

RAZAVY, H., S. K. SZIGETY, and S. M. ROSENBERG, 1996  Evidence for both 3' and 5' single-strand DNA ends in intermediates in Chi stimulated recombination in vivo. Genetics 142:333-339[Abstract].

RICHARDS, B., H. ZHANG, G. PEAR, and M. MEUTH, 1997  Conditional mutator phenotypes in hMSH2-deficient tumor cell lines. Science 227:1523-1526.

ROCA, A. I. and M. M. COX, 1997  RecA protein: structure, function, and role in recombinational DNA repair. Prog. Nucl. Acids Res. Mol. Biol. 56:129-223[Medline].

ROSENBERG, S. M., 1994  In pursuit of a molecular mechanism for adaptive mutation. Genome 37:893-899[Medline].

ROSENBERG, S. M., 1997  Mutation for survival. Curr. Opinion Genet. Devel. 7:829-834[Medline].

ROSENBERG, S. M. and P. J. HASTINGS, 1991  The split-end model for homologous recombination at double-strand breaks and at Chi. Biochimie 73:385-397[Medline].

ROSENBERG, S. M., S. LONGERICH, P. GEE, and R. S. HARRIS, 1994  Adaptive mutation by deletions in small mononucleotide repeats. Science 265:405-407[Abstract/Free Full Text].

ROSENBERG, S. M., R. S. HARRIS, and J. TORKELSON, 1995  Molecular handles on adaptive mutation. Mol. Microbiol. 18:185-189[Medline].

ROSENBERG, S. M., R. S. HARRIS, S. LONGERICH, and A. M. GALLOWAY, 1996  Recombination-dependent mutation in non-dividing cells. Mutat. Res. 350:69-76[Medline].

SEIGELE, D. A. and R. KOLTER, 1992  Life after log. J. Bacteriol. 174:345-348[Free Full Text].

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].

SNIEGOWSKI, P. D., P. J. GERRISH, and R. E. LENSKI, 1997  Evolution of high mutation rates in experimental populations of E. coli. Nature 387:703-705[Medline].

STEELE, D. F. and S. JINKS-ROBERTSON, 1992  An examination of adaptive reversion in Saccharomyces cerevisiae. Genetics 132:9-21[Abstract].

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].

TADDEI, F., I. MATIC, and M. RADMAN, 1995  AMP-dependent SOS induction and mutagenesis in resting bacterial populations. Proc. Natl. Acad. Sci. USA 92:11736-11740[Abstract/Free Full Text].

TADDEI, F., J. A. HALLIDAY, I. MATIC, and M. RADMAN, 1997a  Genetic analysis of mutagenesis in aging Escherichia coli colonies. Mol. Gen. Genet. 256:277-281[Medline].

TADDEI, F., M. RADMAN, J. MAYNARD-SMITH, B. TOUPANCE, and P. H. GOUYON et al., 1997b  Role of mutator alleles in adaptive evolution. Nature 387:700-702[Medline].

TORKELSON, J., 1997 Genome-wide hypermutation underlies adaptive mutation. M. Sc. Thesis, Department of Biological Sciences, University of Alberta, Edmonton, Canada.

TORKELSON, J., R. S. HARRIS, M.-J. LOMBARDO, J. NAGENDRAN, and C. THULIN et al., 1997  Genome-wide hypermutation in a subpopulation of stationary-phase cells underlies recombination-dependent adaptive mutation. EMBO J. 16:3303-3311[Medline].

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:

				L. E. Vidales, L. C. Cardenas, E. Robleto, R. E. Yasbin, and M. Pedraza-Reyes Defects in the Error Prevention Oxidized Guanine System Potentiate Stationary-Phase Mutagenesis in Bacillus subtilis J. Bacteriol., January 15, 2009; 191(2): 506 - 513. [Abstract] [Full Text] [PDF]

				E. Loire, F. Praz, D. Higuet, P. Netter, and G. Achaz Hypermutability of Genes in Homo sapiens Due to the Hosting of Long Mono-SSR Mol. Biol. Evol., January 1, 2009; 26(1): 111 - 121. [Abstract] [Full Text] [PDF]

				P. Vineis and M. Berwick The population dynamics of cancer: a Darwinian perspective Int. J. Epidemiol., October 1, 2006; 35(5): 1151 - 1159. [Abstract] [Full Text] [PDF]

				L. M. C. Hall and S. K. Henderson-Begg Hypermutable bacteria isolated from humans - a critical analysis. Microbiology, September 1, 2006; 152(Pt 9): 2505 - 2514. [Abstract] [Full Text] [PDF]

				M. Pedraza-Reyes and R. E. Yasbin Contribution of the Mismatch DNA Repair System to the Generation of Stationary-Phase-Induced Mutants of Bacillus subtilis J. Bacteriol., October 1, 2004; 186(19): 6485 - 6491. [Abstract] [Full Text] [PDF]